diff --git "a/models/checkpoint-10000/trainer_state.json" "b/models/checkpoint-10000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/models/checkpoint-10000/trainer_state.json"
@@ -0,0 +1,70113 @@
+{
+  "best_metric": 2.5555005073547363,
+  "best_model_checkpoint": "./out/checkpoint-10000",
+  "epoch": 0.9052231375033946,
+  "eval_steps": 1000,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 9.052231375033946e-05,
+      "grad_norm": 6.121739864349365,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 8.7363,
+      "step": 1
+    },
+    {
+      "epoch": 0.0001810446275006789,
+      "grad_norm": 6.012884140014648,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 8.7408,
+      "step": 2
+    },
+    {
+      "epoch": 0.00027156694125101837,
+      "grad_norm": 5.92803955078125,
+      "learning_rate": 6e-06,
+      "loss": 8.7133,
+      "step": 3
+    },
+    {
+      "epoch": 0.0003620892550013578,
+      "grad_norm": 5.670593738555908,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 8.6591,
+      "step": 4
+    },
+    {
+      "epoch": 0.0004526115687516973,
+      "grad_norm": 4.359646320343018,
+      "learning_rate": 1e-05,
+      "loss": 8.6018,
+      "step": 5
+    },
+    {
+      "epoch": 0.0005431338825020367,
+      "grad_norm": 3.6603660583496094,
+      "learning_rate": 1.2e-05,
+      "loss": 8.5488,
+      "step": 6
+    },
+    {
+      "epoch": 0.0006336561962523762,
+      "grad_norm": 3.5484490394592285,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 8.4825,
+      "step": 7
+    },
+    {
+      "epoch": 0.0007241785100027156,
+      "grad_norm": 2.9020848274230957,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 8.4122,
+      "step": 8
+    },
+    {
+      "epoch": 0.0008147008237530552,
+      "grad_norm": 2.7795889377593994,
+      "learning_rate": 1.8e-05,
+      "loss": 8.3877,
+      "step": 9
+    },
+    {
+      "epoch": 0.0009052231375033946,
+      "grad_norm": 2.3020100593566895,
+      "learning_rate": 2e-05,
+      "loss": 8.3544,
+      "step": 10
+    },
+    {
+      "epoch": 0.000995745451253734,
+      "grad_norm": 2.105527400970459,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 8.3211,
+      "step": 11
+    },
+    {
+      "epoch": 0.0010862677650040735,
+      "grad_norm": 2.0495691299438477,
+      "learning_rate": 2.4e-05,
+      "loss": 8.3053,
+      "step": 12
+    },
+    {
+      "epoch": 0.001176790078754413,
+      "grad_norm": 1.9920520782470703,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 8.2805,
+      "step": 13
+    },
+    {
+      "epoch": 0.0012673123925047525,
+      "grad_norm": 1.8438405990600586,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 8.2162,
+      "step": 14
+    },
+    {
+      "epoch": 0.0013578347062550918,
+      "grad_norm": 1.7026909589767456,
+      "learning_rate": 3e-05,
+      "loss": 8.2363,
+      "step": 15
+    },
+    {
+      "epoch": 0.0014483570200054313,
+      "grad_norm": 1.6654070615768433,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 8.1967,
+      "step": 16
+    },
+    {
+      "epoch": 0.0015388793337557708,
+      "grad_norm": 1.606818437576294,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 8.1724,
+      "step": 17
+    },
+    {
+      "epoch": 0.0016294016475061103,
+      "grad_norm": 1.5517821311950684,
+      "learning_rate": 3.6e-05,
+      "loss": 8.13,
+      "step": 18
+    },
+    {
+      "epoch": 0.0017199239612564498,
+      "grad_norm": 1.511502981185913,
+      "learning_rate": 3.8e-05,
+      "loss": 8.0888,
+      "step": 19
+    },
+    {
+      "epoch": 0.001810446275006789,
+      "grad_norm": 1.5508986711502075,
+      "learning_rate": 4e-05,
+      "loss": 8.0653,
+      "step": 20
+    },
+    {
+      "epoch": 0.0019009685887571286,
+      "grad_norm": 1.5409382581710815,
+      "learning_rate": 4.2e-05,
+      "loss": 8.0266,
+      "step": 21
+    },
+    {
+      "epoch": 0.001991490902507468,
+      "grad_norm": 1.6338551044464111,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 7.995,
+      "step": 22
+    },
+    {
+      "epoch": 0.0020820132162578074,
+      "grad_norm": 1.7471034526824951,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 7.9505,
+      "step": 23
+    },
+    {
+      "epoch": 0.002172535530008147,
+      "grad_norm": 1.6112545728683472,
+      "learning_rate": 4.8e-05,
+      "loss": 7.8981,
+      "step": 24
+    },
+    {
+      "epoch": 0.0022630578437584864,
+      "grad_norm": 1.7129489183425903,
+      "learning_rate": 5e-05,
+      "loss": 7.8488,
+      "step": 25
+    },
+    {
+      "epoch": 0.002353580157508826,
+      "grad_norm": 1.6461535692214966,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 7.8213,
+      "step": 26
+    },
+    {
+      "epoch": 0.0024441024712591655,
+      "grad_norm": 1.9870665073394775,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 7.7754,
+      "step": 27
+    },
+    {
+      "epoch": 0.002534624785009505,
+      "grad_norm": 1.6220543384552002,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 7.6909,
+      "step": 28
+    },
+    {
+      "epoch": 0.0026251470987598445,
+      "grad_norm": 4.334130764007568,
+      "learning_rate": 5.8e-05,
+      "loss": 7.6899,
+      "step": 29
+    },
+    {
+      "epoch": 0.0027156694125101835,
+      "grad_norm": 3.4958887100219727,
+      "learning_rate": 6e-05,
+      "loss": 7.5933,
+      "step": 30
+    },
+    {
+      "epoch": 0.002806191726260523,
+      "grad_norm": 2.5281577110290527,
+      "learning_rate": 6.2e-05,
+      "loss": 7.546,
+      "step": 31
+    },
+    {
+      "epoch": 0.0028967140400108626,
+      "grad_norm": 4.500778675079346,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 7.5493,
+      "step": 32
+    },
+    {
+      "epoch": 0.002987236353761202,
+      "grad_norm": 2.2724337577819824,
+      "learning_rate": 6.6e-05,
+      "loss": 7.4798,
+      "step": 33
+    },
+    {
+      "epoch": 0.0030777586675115416,
+      "grad_norm": 4.736296653747559,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 7.4648,
+      "step": 34
+    },
+    {
+      "epoch": 0.003168280981261881,
+      "grad_norm": 2.454152822494507,
+      "learning_rate": 7e-05,
+      "loss": 7.3821,
+      "step": 35
+    },
+    {
+      "epoch": 0.0032588032950122206,
+      "grad_norm": 16.982067108154297,
+      "learning_rate": 7.2e-05,
+      "loss": 7.4612,
+      "step": 36
+    },
+    {
+      "epoch": 0.00334932560876256,
+      "grad_norm": 4.89740514755249,
+      "learning_rate": 7.4e-05,
+      "loss": 7.3308,
+      "step": 37
+    },
+    {
+      "epoch": 0.0034398479225128996,
+      "grad_norm": 3.0148491859436035,
+      "learning_rate": 7.6e-05,
+      "loss": 7.2499,
+      "step": 38
+    },
+    {
+      "epoch": 0.0035303702362632387,
+      "grad_norm": 3.8699262142181396,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 7.2579,
+      "step": 39
+    },
+    {
+      "epoch": 0.003620892550013578,
+      "grad_norm": 2.45534086227417,
+      "learning_rate": 8e-05,
+      "loss": 7.1558,
+      "step": 40
+    },
+    {
+      "epoch": 0.0037114148637639177,
+      "grad_norm": 5.5353827476501465,
+      "learning_rate": 8.2e-05,
+      "loss": 7.1483,
+      "step": 41
+    },
+    {
+      "epoch": 0.0038019371775142572,
+      "grad_norm": 2.3498878479003906,
+      "learning_rate": 8.4e-05,
+      "loss": 7.0903,
+      "step": 42
+    },
+    {
+      "epoch": 0.0038924594912645967,
+      "grad_norm": 3.148735284805298,
+      "learning_rate": 8.6e-05,
+      "loss": 7.0197,
+      "step": 43
+    },
+    {
+      "epoch": 0.003982981805014936,
+      "grad_norm": 3.3806612491607666,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 6.9801,
+      "step": 44
+    },
+    {
+      "epoch": 0.004073504118765276,
+      "grad_norm": 4.305522441864014,
+      "learning_rate": 9e-05,
+      "loss": 6.9286,
+      "step": 45
+    },
+    {
+      "epoch": 0.004164026432515615,
+      "grad_norm": 3.0494394302368164,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 6.9058,
+      "step": 46
+    },
+    {
+      "epoch": 0.004254548746265955,
+      "grad_norm": 1.9974802732467651,
+      "learning_rate": 9.4e-05,
+      "loss": 6.8774,
+      "step": 47
+    },
+    {
+      "epoch": 0.004345071060016294,
+      "grad_norm": 7.055619239807129,
+      "learning_rate": 9.6e-05,
+      "loss": 6.8402,
+      "step": 48
+    },
+    {
+      "epoch": 0.004435593373766634,
+      "grad_norm": 2.6249217987060547,
+      "learning_rate": 9.8e-05,
+      "loss": 6.7974,
+      "step": 49
+    },
+    {
+      "epoch": 0.004526115687516973,
+      "grad_norm": 2.5299758911132812,
+      "learning_rate": 0.0001,
+      "loss": 6.7687,
+      "step": 50
+    },
+    {
+      "epoch": 0.004616638001267313,
+      "grad_norm": 2.7093379497528076,
+      "learning_rate": 0.00010200000000000001,
+      "loss": 6.705,
+      "step": 51
+    },
+    {
+      "epoch": 0.004707160315017652,
+      "grad_norm": 2.7266974449157715,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 6.5869,
+      "step": 52
+    },
+    {
+      "epoch": 0.004797682628767991,
+      "grad_norm": 2.071244239807129,
+      "learning_rate": 0.00010600000000000002,
+      "loss": 6.5817,
+      "step": 53
+    },
+    {
+      "epoch": 0.004888204942518331,
+      "grad_norm": 2.33720064163208,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 6.582,
+      "step": 54
+    },
+    {
+      "epoch": 0.00497872725626867,
+      "grad_norm": 2.151409387588501,
+      "learning_rate": 0.00011000000000000002,
+      "loss": 6.5736,
+      "step": 55
+    },
+    {
+      "epoch": 0.00506924957001901,
+      "grad_norm": 3.529722213745117,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 6.4519,
+      "step": 56
+    },
+    {
+      "epoch": 0.005159771883769349,
+      "grad_norm": 3.7185935974121094,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 6.3824,
+      "step": 57
+    },
+    {
+      "epoch": 0.005250294197519689,
+      "grad_norm": 3.911526918411255,
+      "learning_rate": 0.000116,
+      "loss": 6.3547,
+      "step": 58
+    },
+    {
+      "epoch": 0.005340816511270028,
+      "grad_norm": 3.760375499725342,
+      "learning_rate": 0.000118,
+      "loss": 6.3178,
+      "step": 59
+    },
+    {
+      "epoch": 0.005431338825020367,
+      "grad_norm": 3.210420608520508,
+      "learning_rate": 0.00012,
+      "loss": 6.3338,
+      "step": 60
+    },
+    {
+      "epoch": 0.005521861138770707,
+      "grad_norm": 2.1828818321228027,
+      "learning_rate": 0.000122,
+      "loss": 6.2572,
+      "step": 61
+    },
+    {
+      "epoch": 0.005612383452521046,
+      "grad_norm": 2.459007740020752,
+      "learning_rate": 0.000124,
+      "loss": 6.2562,
+      "step": 62
+    },
+    {
+      "epoch": 0.005702905766271386,
+      "grad_norm": 2.0643117427825928,
+      "learning_rate": 0.000126,
+      "loss": 6.126,
+      "step": 63
+    },
+    {
+      "epoch": 0.005793428080021725,
+      "grad_norm": 3.102482557296753,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 6.1061,
+      "step": 64
+    },
+    {
+      "epoch": 0.005883950393772065,
+      "grad_norm": 2.2519826889038086,
+      "learning_rate": 0.00013000000000000002,
+      "loss": 6.1004,
+      "step": 65
+    },
+    {
+      "epoch": 0.005974472707522404,
+      "grad_norm": 1.8878542184829712,
+      "learning_rate": 0.000132,
+      "loss": 6.0569,
+      "step": 66
+    },
+    {
+      "epoch": 0.006064995021272744,
+      "grad_norm": 1.7067312002182007,
+      "learning_rate": 0.000134,
+      "loss": 6.0232,
+      "step": 67
+    },
+    {
+      "epoch": 0.006155517335023083,
+      "grad_norm": 2.0737111568450928,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 5.9652,
+      "step": 68
+    },
+    {
+      "epoch": 0.006246039648773422,
+      "grad_norm": 2.1083405017852783,
+      "learning_rate": 0.000138,
+      "loss": 5.9704,
+      "step": 69
+    },
+    {
+      "epoch": 0.006336561962523762,
+      "grad_norm": 1.5221211910247803,
+      "learning_rate": 0.00014,
+      "loss": 5.8621,
+      "step": 70
+    },
+    {
+      "epoch": 0.006427084276274101,
+      "grad_norm": 2.04142427444458,
+      "learning_rate": 0.000142,
+      "loss": 5.8764,
+      "step": 71
+    },
+    {
+      "epoch": 0.006517606590024441,
+      "grad_norm": 2.7947962284088135,
+      "learning_rate": 0.000144,
+      "loss": 5.8276,
+      "step": 72
+    },
+    {
+      "epoch": 0.00660812890377478,
+      "grad_norm": 2.515359878540039,
+      "learning_rate": 0.000146,
+      "loss": 5.7222,
+      "step": 73
+    },
+    {
+      "epoch": 0.00669865121752512,
+      "grad_norm": 1.4865795373916626,
+      "learning_rate": 0.000148,
+      "loss": 5.7296,
+      "step": 74
+    },
+    {
+      "epoch": 0.006789173531275459,
+      "grad_norm": 1.521973729133606,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 5.6688,
+      "step": 75
+    },
+    {
+      "epoch": 0.006879695845025799,
+      "grad_norm": 1.8078877925872803,
+      "learning_rate": 0.000152,
+      "loss": 5.7169,
+      "step": 76
+    },
+    {
+      "epoch": 0.006970218158776138,
+      "grad_norm": 1.7064664363861084,
+      "learning_rate": 0.000154,
+      "loss": 5.6,
+      "step": 77
+    },
+    {
+      "epoch": 0.007060740472526477,
+      "grad_norm": 1.5587918758392334,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 5.6158,
+      "step": 78
+    },
+    {
+      "epoch": 0.007151262786276817,
+      "grad_norm": 1.5954521894454956,
+      "learning_rate": 0.00015800000000000002,
+      "loss": 5.5937,
+      "step": 79
+    },
+    {
+      "epoch": 0.007241785100027156,
+      "grad_norm": 1.7314285039901733,
+      "learning_rate": 0.00016,
+      "loss": 5.6023,
+      "step": 80
+    },
+    {
+      "epoch": 0.007332307413777496,
+      "grad_norm": 1.9381115436553955,
+      "learning_rate": 0.000162,
+      "loss": 5.4748,
+      "step": 81
+    },
+    {
+      "epoch": 0.007422829727527835,
+      "grad_norm": 1.2395066022872925,
+      "learning_rate": 0.000164,
+      "loss": 5.4938,
+      "step": 82
+    },
+    {
+      "epoch": 0.007513352041278175,
+      "grad_norm": 1.3130717277526855,
+      "learning_rate": 0.000166,
+      "loss": 5.5345,
+      "step": 83
+    },
+    {
+      "epoch": 0.0076038743550285144,
+      "grad_norm": 1.5142767429351807,
+      "learning_rate": 0.000168,
+      "loss": 5.4502,
+      "step": 84
+    },
+    {
+      "epoch": 0.007694396668778854,
+      "grad_norm": 1.2253237962722778,
+      "learning_rate": 0.00017,
+      "loss": 5.3557,
+      "step": 85
+    },
+    {
+      "epoch": 0.0077849189825291935,
+      "grad_norm": 1.3859381675720215,
+      "learning_rate": 0.000172,
+      "loss": 5.3861,
+      "step": 86
+    },
+    {
+      "epoch": 0.007875441296279533,
+      "grad_norm": 1.8688181638717651,
+      "learning_rate": 0.000174,
+      "loss": 5.4363,
+      "step": 87
+    },
+    {
+      "epoch": 0.007965963610029872,
+      "grad_norm": 1.328377604484558,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 5.3266,
+      "step": 88
+    },
+    {
+      "epoch": 0.008056485923780212,
+      "grad_norm": 1.5786259174346924,
+      "learning_rate": 0.00017800000000000002,
+      "loss": 5.3736,
+      "step": 89
+    },
+    {
+      "epoch": 0.008147008237530552,
+      "grad_norm": 1.2112221717834473,
+      "learning_rate": 0.00018,
+      "loss": 5.2945,
+      "step": 90
+    },
+    {
+      "epoch": 0.00823753055128089,
+      "grad_norm": 1.210732340812683,
+      "learning_rate": 0.000182,
+      "loss": 5.2812,
+      "step": 91
+    },
+    {
+      "epoch": 0.00832805286503123,
+      "grad_norm": 1.3312040567398071,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 5.2365,
+      "step": 92
+    },
+    {
+      "epoch": 0.00841857517878157,
+      "grad_norm": 1.0963982343673706,
+      "learning_rate": 0.00018600000000000002,
+      "loss": 5.243,
+      "step": 93
+    },
+    {
+      "epoch": 0.00850909749253191,
+      "grad_norm": 1.202478051185608,
+      "learning_rate": 0.000188,
+      "loss": 5.2121,
+      "step": 94
+    },
+    {
+      "epoch": 0.008599619806282249,
+      "grad_norm": 1.3619744777679443,
+      "learning_rate": 0.00019,
+      "loss": 5.186,
+      "step": 95
+    },
+    {
+      "epoch": 0.008690142120032588,
+      "grad_norm": 1.3738235235214233,
+      "learning_rate": 0.000192,
+      "loss": 5.2303,
+      "step": 96
+    },
+    {
+      "epoch": 0.008780664433782927,
+      "grad_norm": 1.2414801120758057,
+      "learning_rate": 0.000194,
+      "loss": 5.222,
+      "step": 97
+    },
+    {
+      "epoch": 0.008871186747533268,
+      "grad_norm": 1.2269741296768188,
+      "learning_rate": 0.000196,
+      "loss": 5.1837,
+      "step": 98
+    },
+    {
+      "epoch": 0.008961709061283607,
+      "grad_norm": 1.3814342021942139,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 5.1261,
+      "step": 99
+    },
+    {
+      "epoch": 0.009052231375033946,
+      "grad_norm": 1.3333874940872192,
+      "learning_rate": 0.0002,
+      "loss": 5.135,
+      "step": 100
+    },
+    {
+      "epoch": 0.009142753688784285,
+      "grad_norm": 0.9502299427986145,
+      "learning_rate": 0.00019999999496500136,
+      "loss": 5.167,
+      "step": 101
+    },
+    {
+      "epoch": 0.009233276002534626,
+      "grad_norm": 1.2078323364257812,
+      "learning_rate": 0.00019999997986000598,
+      "loss": 5.1009,
+      "step": 102
+    },
+    {
+      "epoch": 0.009323798316284965,
+      "grad_norm": 1.2079190015792847,
+      "learning_rate": 0.00019999995468501536,
+      "loss": 5.1678,
+      "step": 103
+    },
+    {
+      "epoch": 0.009414320630035304,
+      "grad_norm": 1.362955927848816,
+      "learning_rate": 0.00019999991944003202,
+      "loss": 5.0459,
+      "step": 104
+    },
+    {
+      "epoch": 0.009504842943785643,
+      "grad_norm": 1.3274837732315063,
+      "learning_rate": 0.00019999987412505955,
+      "loss": 5.1548,
+      "step": 105
+    },
+    {
+      "epoch": 0.009595365257535982,
+      "grad_norm": 1.1168432235717773,
+      "learning_rate": 0.00019999981874010248,
+      "loss": 5.0671,
+      "step": 106
+    },
+    {
+      "epoch": 0.009685887571286323,
+      "grad_norm": 1.6353976726531982,
+      "learning_rate": 0.00019999975328516642,
+      "loss": 5.0539,
+      "step": 107
+    },
+    {
+      "epoch": 0.009776409885036662,
+      "grad_norm": 1.2709882259368896,
+      "learning_rate": 0.00019999967776025793,
+      "loss": 5.0878,
+      "step": 108
+    },
+    {
+      "epoch": 0.009866932198787001,
+      "grad_norm": 1.3899245262145996,
+      "learning_rate": 0.00019999959216538462,
+      "loss": 4.9687,
+      "step": 109
+    },
+    {
+      "epoch": 0.00995745451253734,
+      "grad_norm": 1.2088505029678345,
+      "learning_rate": 0.00019999949650055513,
+      "loss": 5.0465,
+      "step": 110
+    },
+    {
+      "epoch": 0.01004797682628768,
+      "grad_norm": 0.9921075105667114,
+      "learning_rate": 0.00019999939076577905,
+      "loss": 5.1357,
+      "step": 111
+    },
+    {
+      "epoch": 0.01013849914003802,
+      "grad_norm": 1.2015072107315063,
+      "learning_rate": 0.00019999927496106707,
+      "loss": 5.0771,
+      "step": 112
+    },
+    {
+      "epoch": 0.010229021453788359,
+      "grad_norm": 1.0013004541397095,
+      "learning_rate": 0.00019999914908643086,
+      "loss": 5.0229,
+      "step": 113
+    },
+    {
+      "epoch": 0.010319543767538698,
+      "grad_norm": 1.2613677978515625,
+      "learning_rate": 0.00019999901314188303,
+      "loss": 4.9829,
+      "step": 114
+    },
+    {
+      "epoch": 0.010410066081289037,
+      "grad_norm": 1.47295081615448,
+      "learning_rate": 0.00019999886712743732,
+      "loss": 4.9887,
+      "step": 115
+    },
+    {
+      "epoch": 0.010500588395039378,
+      "grad_norm": 1.5374642610549927,
+      "learning_rate": 0.00019999871104310846,
+      "loss": 5.0556,
+      "step": 116
+    },
+    {
+      "epoch": 0.010591110708789717,
+      "grad_norm": 1.3752161264419556,
+      "learning_rate": 0.00019999854488891213,
+      "loss": 5.0298,
+      "step": 117
+    },
+    {
+      "epoch": 0.010681633022540056,
+      "grad_norm": 1.0788917541503906,
+      "learning_rate": 0.00019999836866486503,
+      "loss": 4.9835,
+      "step": 118
+    },
+    {
+      "epoch": 0.010772155336290395,
+      "grad_norm": 1.0625488758087158,
+      "learning_rate": 0.00019999818237098496,
+      "loss": 4.9575,
+      "step": 119
+    },
+    {
+      "epoch": 0.010862677650040734,
+      "grad_norm": 1.0351204872131348,
+      "learning_rate": 0.00019999798600729064,
+      "loss": 4.9703,
+      "step": 120
+    },
+    {
+      "epoch": 0.010953199963791075,
+      "grad_norm": 0.9874302744865417,
+      "learning_rate": 0.0001999977795738019,
+      "loss": 5.0153,
+      "step": 121
+    },
+    {
+      "epoch": 0.011043722277541414,
+      "grad_norm": 1.0721620321273804,
+      "learning_rate": 0.00019999756307053948,
+      "loss": 4.9537,
+      "step": 122
+    },
+    {
+      "epoch": 0.011134244591291753,
+      "grad_norm": 1.1902644634246826,
+      "learning_rate": 0.0001999973364975252,
+      "loss": 5.0332,
+      "step": 123
+    },
+    {
+      "epoch": 0.011224766905042092,
+      "grad_norm": 0.9976598620414734,
+      "learning_rate": 0.00019999709985478188,
+      "loss": 4.8953,
+      "step": 124
+    },
+    {
+      "epoch": 0.011315289218792433,
+      "grad_norm": 1.1079658269882202,
+      "learning_rate": 0.0001999968531423333,
+      "loss": 4.9752,
+      "step": 125
+    },
+    {
+      "epoch": 0.011405811532542772,
+      "grad_norm": 1.037431240081787,
+      "learning_rate": 0.0001999965963602044,
+      "loss": 4.9653,
+      "step": 126
+    },
+    {
+      "epoch": 0.011496333846293111,
+      "grad_norm": 1.1416285037994385,
+      "learning_rate": 0.00019999632950842093,
+      "loss": 4.926,
+      "step": 127
+    },
+    {
+      "epoch": 0.01158685616004345,
+      "grad_norm": 1.3808517456054688,
+      "learning_rate": 0.00019999605258700983,
+      "loss": 4.9866,
+      "step": 128
+    },
+    {
+      "epoch": 0.01167737847379379,
+      "grad_norm": 1.0173901319503784,
+      "learning_rate": 0.00019999576559599902,
+      "loss": 5.002,
+      "step": 129
+    },
+    {
+      "epoch": 0.01176790078754413,
+      "grad_norm": 1.1562535762786865,
+      "learning_rate": 0.0001999954685354173,
+      "loss": 4.9624,
+      "step": 130
+    },
+    {
+      "epoch": 0.01185842310129447,
+      "grad_norm": 1.1456613540649414,
+      "learning_rate": 0.00019999516140529463,
+      "loss": 4.9042,
+      "step": 131
+    },
+    {
+      "epoch": 0.011948945415044808,
+      "grad_norm": 1.322360634803772,
+      "learning_rate": 0.00019999484420566197,
+      "loss": 4.804,
+      "step": 132
+    },
+    {
+      "epoch": 0.012039467728795147,
+      "grad_norm": 1.1253564357757568,
+      "learning_rate": 0.00019999451693655123,
+      "loss": 4.8905,
+      "step": 133
+    },
+    {
+      "epoch": 0.012129990042545488,
+      "grad_norm": 1.1110708713531494,
+      "learning_rate": 0.00019999417959799535,
+      "loss": 4.9735,
+      "step": 134
+    },
+    {
+      "epoch": 0.012220512356295827,
+      "grad_norm": 1.0920100212097168,
+      "learning_rate": 0.00019999383219002835,
+      "loss": 4.9384,
+      "step": 135
+    },
+    {
+      "epoch": 0.012311034670046166,
+      "grad_norm": 0.9289610981941223,
+      "learning_rate": 0.00019999347471268516,
+      "loss": 4.9417,
+      "step": 136
+    },
+    {
+      "epoch": 0.012401556983796505,
+      "grad_norm": 1.009425163269043,
+      "learning_rate": 0.0001999931071660018,
+      "loss": 4.9405,
+      "step": 137
+    },
+    {
+      "epoch": 0.012492079297546844,
+      "grad_norm": 1.14154851436615,
+      "learning_rate": 0.00019999272955001528,
+      "loss": 4.9205,
+      "step": 138
+    },
+    {
+      "epoch": 0.012582601611297185,
+      "grad_norm": 0.9523165822029114,
+      "learning_rate": 0.00019999234186476365,
+      "loss": 4.8967,
+      "step": 139
+    },
+    {
+      "epoch": 0.012673123925047524,
+      "grad_norm": 1.2908835411071777,
+      "learning_rate": 0.00019999194411028594,
+      "loss": 4.9088,
+      "step": 140
+    },
+    {
+      "epoch": 0.012763646238797863,
+      "grad_norm": 0.9540936946868896,
+      "learning_rate": 0.00019999153628662218,
+      "loss": 4.9449,
+      "step": 141
+    },
+    {
+      "epoch": 0.012854168552548203,
+      "grad_norm": 0.9404425621032715,
+      "learning_rate": 0.00019999111839381345,
+      "loss": 4.9067,
+      "step": 142
+    },
+    {
+      "epoch": 0.012944690866298543,
+      "grad_norm": 0.8981451988220215,
+      "learning_rate": 0.00019999069043190185,
+      "loss": 4.8862,
+      "step": 143
+    },
+    {
+      "epoch": 0.013035213180048882,
+      "grad_norm": 0.8868217468261719,
+      "learning_rate": 0.00019999025240093044,
+      "loss": 4.8235,
+      "step": 144
+    },
+    {
+      "epoch": 0.013125735493799221,
+      "grad_norm": 0.8762542605400085,
+      "learning_rate": 0.00019998980430094334,
+      "loss": 4.8599,
+      "step": 145
+    },
+    {
+      "epoch": 0.01321625780754956,
+      "grad_norm": 1.0478554964065552,
+      "learning_rate": 0.0001999893461319857,
+      "loss": 4.8078,
+      "step": 146
+    },
+    {
+      "epoch": 0.0133067801212999,
+      "grad_norm": 1.216299295425415,
+      "learning_rate": 0.00019998887789410364,
+      "loss": 4.8147,
+      "step": 147
+    },
+    {
+      "epoch": 0.01339730243505024,
+      "grad_norm": 0.9078794717788696,
+      "learning_rate": 0.0001999883995873443,
+      "loss": 4.8386,
+      "step": 148
+    },
+    {
+      "epoch": 0.01348782474880058,
+      "grad_norm": 0.9007813930511475,
+      "learning_rate": 0.0001999879112117559,
+      "loss": 4.8378,
+      "step": 149
+    },
+    {
+      "epoch": 0.013578347062550919,
+      "grad_norm": 0.8915342092514038,
+      "learning_rate": 0.00019998741276738754,
+      "loss": 4.8782,
+      "step": 150
+    },
+    {
+      "epoch": 0.013668869376301258,
+      "grad_norm": 0.9013955593109131,
+      "learning_rate": 0.00019998690425428943,
+      "loss": 4.8299,
+      "step": 151
+    },
+    {
+      "epoch": 0.013759391690051598,
+      "grad_norm": 0.9909464120864868,
+      "learning_rate": 0.0001999863856725128,
+      "loss": 4.7892,
+      "step": 152
+    },
+    {
+      "epoch": 0.013849914003801938,
+      "grad_norm": 1.0322654247283936,
+      "learning_rate": 0.0001999858570221099,
+      "loss": 4.7991,
+      "step": 153
+    },
+    {
+      "epoch": 0.013940436317552277,
+      "grad_norm": 0.9434685111045837,
+      "learning_rate": 0.00019998531830313395,
+      "loss": 4.8662,
+      "step": 154
+    },
+    {
+      "epoch": 0.014030958631302616,
+      "grad_norm": 0.8822855353355408,
+      "learning_rate": 0.00019998476951563915,
+      "loss": 4.808,
+      "step": 155
+    },
+    {
+      "epoch": 0.014121480945052955,
+      "grad_norm": 1.0918142795562744,
+      "learning_rate": 0.00019998421065968078,
+      "loss": 4.9409,
+      "step": 156
+    },
+    {
+      "epoch": 0.014212003258803296,
+      "grad_norm": 1.1329283714294434,
+      "learning_rate": 0.00019998364173531513,
+      "loss": 4.8035,
+      "step": 157
+    },
+    {
+      "epoch": 0.014302525572553635,
+      "grad_norm": 1.6153861284255981,
+      "learning_rate": 0.00019998306274259954,
+      "loss": 4.8325,
+      "step": 158
+    },
+    {
+      "epoch": 0.014393047886303974,
+      "grad_norm": 0.9851101636886597,
+      "learning_rate": 0.00019998247368159224,
+      "loss": 4.7919,
+      "step": 159
+    },
+    {
+      "epoch": 0.014483570200054313,
+      "grad_norm": 0.9160235524177551,
+      "learning_rate": 0.0001999818745523526,
+      "loss": 4.8134,
+      "step": 160
+    },
+    {
+      "epoch": 0.014574092513804654,
+      "grad_norm": 1.10393488407135,
+      "learning_rate": 0.00019998126535494088,
+      "loss": 4.8546,
+      "step": 161
+    },
+    {
+      "epoch": 0.014664614827554993,
+      "grad_norm": 1.088984489440918,
+      "learning_rate": 0.0001999806460894185,
+      "loss": 4.7318,
+      "step": 162
+    },
+    {
+      "epoch": 0.014755137141305332,
+      "grad_norm": 1.0326741933822632,
+      "learning_rate": 0.00019998001675584781,
+      "loss": 4.8165,
+      "step": 163
+    },
+    {
+      "epoch": 0.01484565945505567,
+      "grad_norm": 0.8954424262046814,
+      "learning_rate": 0.00019997937735429219,
+      "loss": 4.8473,
+      "step": 164
+    },
+    {
+      "epoch": 0.01493618176880601,
+      "grad_norm": 0.9650543332099915,
+      "learning_rate": 0.00019997872788481594,
+      "loss": 4.8017,
+      "step": 165
+    },
+    {
+      "epoch": 0.01502670408255635,
+      "grad_norm": 1.042865514755249,
+      "learning_rate": 0.00019997806834748456,
+      "loss": 4.8427,
+      "step": 166
+    },
+    {
+      "epoch": 0.01511722639630669,
+      "grad_norm": 0.8964381814002991,
+      "learning_rate": 0.00019997739874236443,
+      "loss": 4.7021,
+      "step": 167
+    },
+    {
+      "epoch": 0.015207748710057029,
+      "grad_norm": 1.068221092224121,
+      "learning_rate": 0.00019997671906952298,
+      "loss": 4.8155,
+      "step": 168
+    },
+    {
+      "epoch": 0.015298271023807368,
+      "grad_norm": 1.0170791149139404,
+      "learning_rate": 0.00019997602932902866,
+      "loss": 4.8067,
+      "step": 169
+    },
+    {
+      "epoch": 0.015388793337557709,
+      "grad_norm": 1.1236660480499268,
+      "learning_rate": 0.00019997532952095094,
+      "loss": 4.8021,
+      "step": 170
+    },
+    {
+      "epoch": 0.015479315651308048,
+      "grad_norm": 0.9010448455810547,
+      "learning_rate": 0.00019997461964536023,
+      "loss": 4.8208,
+      "step": 171
+    },
+    {
+      "epoch": 0.015569837965058387,
+      "grad_norm": 1.1579195261001587,
+      "learning_rate": 0.0001999738997023281,
+      "loss": 4.8075,
+      "step": 172
+    },
+    {
+      "epoch": 0.015660360278808728,
+      "grad_norm": 0.8007659912109375,
+      "learning_rate": 0.00019997316969192698,
+      "loss": 4.791,
+      "step": 173
+    },
+    {
+      "epoch": 0.015750882592559065,
+      "grad_norm": 0.9970986247062683,
+      "learning_rate": 0.00019997242961423042,
+      "loss": 4.711,
+      "step": 174
+    },
+    {
+      "epoch": 0.015841404906309406,
+      "grad_norm": 0.7657274603843689,
+      "learning_rate": 0.0001999716794693129,
+      "loss": 4.7906,
+      "step": 175
+    },
+    {
+      "epoch": 0.015931927220059743,
+      "grad_norm": 0.899833083152771,
+      "learning_rate": 0.00019997091925725004,
+      "loss": 4.8504,
+      "step": 176
+    },
+    {
+      "epoch": 0.016022449533810084,
+      "grad_norm": 1.044193983078003,
+      "learning_rate": 0.00019997014897811833,
+      "loss": 4.8014,
+      "step": 177
+    },
+    {
+      "epoch": 0.016112971847560425,
+      "grad_norm": 1.321266531944275,
+      "learning_rate": 0.00019996936863199535,
+      "loss": 4.807,
+      "step": 178
+    },
+    {
+      "epoch": 0.016203494161310762,
+      "grad_norm": 0.8726624250411987,
+      "learning_rate": 0.00019996857821895966,
+      "loss": 4.7457,
+      "step": 179
+    },
+    {
+      "epoch": 0.016294016475061103,
+      "grad_norm": 0.9373568892478943,
+      "learning_rate": 0.00019996777773909093,
+      "loss": 4.7909,
+      "step": 180
+    },
+    {
+      "epoch": 0.01638453878881144,
+      "grad_norm": 0.933677613735199,
+      "learning_rate": 0.00019996696719246969,
+      "loss": 4.7476,
+      "step": 181
+    },
+    {
+      "epoch": 0.01647506110256178,
+      "grad_norm": 0.8615386486053467,
+      "learning_rate": 0.00019996614657917758,
+      "loss": 4.757,
+      "step": 182
+    },
+    {
+      "epoch": 0.016565583416312122,
+      "grad_norm": 0.9267805218696594,
+      "learning_rate": 0.00019996531589929725,
+      "loss": 4.8215,
+      "step": 183
+    },
+    {
+      "epoch": 0.01665610573006246,
+      "grad_norm": 0.9681649804115295,
+      "learning_rate": 0.00019996447515291233,
+      "loss": 4.7463,
+      "step": 184
+    },
+    {
+      "epoch": 0.0167466280438128,
+      "grad_norm": 0.9593618512153625,
+      "learning_rate": 0.00019996362434010752,
+      "loss": 4.7061,
+      "step": 185
+    },
+    {
+      "epoch": 0.01683715035756314,
+      "grad_norm": 1.017175555229187,
+      "learning_rate": 0.00019996276346096847,
+      "loss": 4.7243,
+      "step": 186
+    },
+    {
+      "epoch": 0.01692767267131348,
+      "grad_norm": 1.036476731300354,
+      "learning_rate": 0.00019996189251558189,
+      "loss": 4.7767,
+      "step": 187
+    },
+    {
+      "epoch": 0.01701819498506382,
+      "grad_norm": 0.892395555973053,
+      "learning_rate": 0.00019996101150403543,
+      "loss": 4.745,
+      "step": 188
+    },
+    {
+      "epoch": 0.017108717298814156,
+      "grad_norm": 1.1248540878295898,
+      "learning_rate": 0.00019996012042641786,
+      "loss": 4.7538,
+      "step": 189
+    },
+    {
+      "epoch": 0.017199239612564497,
+      "grad_norm": 0.9392964243888855,
+      "learning_rate": 0.00019995921928281894,
+      "loss": 4.7655,
+      "step": 190
+    },
+    {
+      "epoch": 0.017289761926314838,
+      "grad_norm": 1.0880126953125,
+      "learning_rate": 0.00019995830807332932,
+      "loss": 4.7164,
+      "step": 191
+    },
+    {
+      "epoch": 0.017380284240065175,
+      "grad_norm": 0.9286646246910095,
+      "learning_rate": 0.00019995738679804085,
+      "loss": 4.7107,
+      "step": 192
+    },
+    {
+      "epoch": 0.017470806553815516,
+      "grad_norm": 1.0103776454925537,
+      "learning_rate": 0.00019995645545704623,
+      "loss": 4.6853,
+      "step": 193
+    },
+    {
+      "epoch": 0.017561328867565854,
+      "grad_norm": 1.0643168687820435,
+      "learning_rate": 0.00019995551405043932,
+      "loss": 4.7065,
+      "step": 194
+    },
+    {
+      "epoch": 0.017651851181316194,
+      "grad_norm": 0.9227851629257202,
+      "learning_rate": 0.00019995456257831484,
+      "loss": 4.7221,
+      "step": 195
+    },
+    {
+      "epoch": 0.017742373495066535,
+      "grad_norm": 0.8938180208206177,
+      "learning_rate": 0.00019995360104076867,
+      "loss": 4.7293,
+      "step": 196
+    },
+    {
+      "epoch": 0.017832895808816873,
+      "grad_norm": 1.2211713790893555,
+      "learning_rate": 0.0001999526294378976,
+      "loss": 4.707,
+      "step": 197
+    },
+    {
+      "epoch": 0.017923418122567213,
+      "grad_norm": 0.8459523916244507,
+      "learning_rate": 0.0001999516477697995,
+      "loss": 4.6857,
+      "step": 198
+    },
+    {
+      "epoch": 0.01801394043631755,
+      "grad_norm": 0.9301251769065857,
+      "learning_rate": 0.00019995065603657316,
+      "loss": 4.7267,
+      "step": 199
+    },
+    {
+      "epoch": 0.01810446275006789,
+      "grad_norm": 0.8725582957267761,
+      "learning_rate": 0.00019994965423831854,
+      "loss": 4.6945,
+      "step": 200
+    },
+    {
+      "epoch": 0.018194985063818232,
+      "grad_norm": 1.0342596769332886,
+      "learning_rate": 0.00019994864237513643,
+      "loss": 4.707,
+      "step": 201
+    },
+    {
+      "epoch": 0.01828550737756857,
+      "grad_norm": 0.8990219235420227,
+      "learning_rate": 0.0001999476204471288,
+      "loss": 4.7324,
+      "step": 202
+    },
+    {
+      "epoch": 0.01837602969131891,
+      "grad_norm": 0.9822607040405273,
+      "learning_rate": 0.00019994658845439852,
+      "loss": 4.7214,
+      "step": 203
+    },
+    {
+      "epoch": 0.01846655200506925,
+      "grad_norm": 1.0090265274047852,
+      "learning_rate": 0.0001999455463970495,
+      "loss": 4.6733,
+      "step": 204
+    },
+    {
+      "epoch": 0.01855707431881959,
+      "grad_norm": 0.8931706547737122,
+      "learning_rate": 0.0001999444942751867,
+      "loss": 4.673,
+      "step": 205
+    },
+    {
+      "epoch": 0.01864759663256993,
+      "grad_norm": 0.975825309753418,
+      "learning_rate": 0.00019994343208891606,
+      "loss": 4.6323,
+      "step": 206
+    },
+    {
+      "epoch": 0.018738118946320267,
+      "grad_norm": 0.967452347278595,
+      "learning_rate": 0.00019994235983834455,
+      "loss": 4.6975,
+      "step": 207
+    },
+    {
+      "epoch": 0.018828641260070608,
+      "grad_norm": 0.8899520039558411,
+      "learning_rate": 0.00019994127752358013,
+      "loss": 4.6753,
+      "step": 208
+    },
+    {
+      "epoch": 0.01891916357382095,
+      "grad_norm": 1.09125816822052,
+      "learning_rate": 0.00019994018514473183,
+      "loss": 4.644,
+      "step": 209
+    },
+    {
+      "epoch": 0.019009685887571286,
+      "grad_norm": 0.9174689054489136,
+      "learning_rate": 0.0001999390827019096,
+      "loss": 4.6589,
+      "step": 210
+    },
+    {
+      "epoch": 0.019100208201321626,
+      "grad_norm": 0.8886850476264954,
+      "learning_rate": 0.00019993797019522447,
+      "loss": 4.6273,
+      "step": 211
+    },
+    {
+      "epoch": 0.019190730515071964,
+      "grad_norm": 0.8056244850158691,
+      "learning_rate": 0.00019993684762478845,
+      "loss": 4.6332,
+      "step": 212
+    },
+    {
+      "epoch": 0.019281252828822305,
+      "grad_norm": 0.7783074975013733,
+      "learning_rate": 0.00019993571499071465,
+      "loss": 4.6484,
+      "step": 213
+    },
+    {
+      "epoch": 0.019371775142572645,
+      "grad_norm": 0.8422152400016785,
+      "learning_rate": 0.00019993457229311708,
+      "loss": 4.6529,
+      "step": 214
+    },
+    {
+      "epoch": 0.019462297456322983,
+      "grad_norm": 0.8490102887153625,
+      "learning_rate": 0.0001999334195321108,
+      "loss": 4.6784,
+      "step": 215
+    },
+    {
+      "epoch": 0.019552819770073324,
+      "grad_norm": 0.931522786617279,
+      "learning_rate": 0.0001999322567078119,
+      "loss": 4.6701,
+      "step": 216
+    },
+    {
+      "epoch": 0.01964334208382366,
+      "grad_norm": 0.825369119644165,
+      "learning_rate": 0.0001999310838203375,
+      "loss": 4.64,
+      "step": 217
+    },
+    {
+      "epoch": 0.019733864397574002,
+      "grad_norm": 1.099587082862854,
+      "learning_rate": 0.0001999299008698057,
+      "loss": 4.6228,
+      "step": 218
+    },
+    {
+      "epoch": 0.019824386711324343,
+      "grad_norm": 0.8843775391578674,
+      "learning_rate": 0.00019992870785633563,
+      "loss": 4.6485,
+      "step": 219
+    },
+    {
+      "epoch": 0.01991490902507468,
+      "grad_norm": 0.8883227705955505,
+      "learning_rate": 0.00019992750478004738,
+      "loss": 4.6917,
+      "step": 220
+    },
+    {
+      "epoch": 0.02000543133882502,
+      "grad_norm": 1.078590989112854,
+      "learning_rate": 0.0001999262916410621,
+      "loss": 4.6544,
+      "step": 221
+    },
+    {
+      "epoch": 0.02009595365257536,
+      "grad_norm": 1.051282286643982,
+      "learning_rate": 0.00019992506843950208,
+      "loss": 4.7503,
+      "step": 222
+    },
+    {
+      "epoch": 0.0201864759663257,
+      "grad_norm": 0.8993271589279175,
+      "learning_rate": 0.0001999238351754903,
+      "loss": 4.6718,
+      "step": 223
+    },
+    {
+      "epoch": 0.02027699828007604,
+      "grad_norm": 0.9291291832923889,
+      "learning_rate": 0.00019992259184915115,
+      "loss": 4.6306,
+      "step": 224
+    },
+    {
+      "epoch": 0.020367520593826377,
+      "grad_norm": 0.9018997550010681,
+      "learning_rate": 0.00019992133846060968,
+      "loss": 4.6095,
+      "step": 225
+    },
+    {
+      "epoch": 0.020458042907576718,
+      "grad_norm": 0.9196507930755615,
+      "learning_rate": 0.00019992007500999214,
+      "loss": 4.6376,
+      "step": 226
+    },
+    {
+      "epoch": 0.02054856522132706,
+      "grad_norm": 1.0362221002578735,
+      "learning_rate": 0.00019991880149742582,
+      "loss": 4.6297,
+      "step": 227
+    },
+    {
+      "epoch": 0.020639087535077396,
+      "grad_norm": 0.9928797483444214,
+      "learning_rate": 0.00019991751792303893,
+      "loss": 4.6679,
+      "step": 228
+    },
+    {
+      "epoch": 0.020729609848827737,
+      "grad_norm": 1.0466713905334473,
+      "learning_rate": 0.00019991622428696068,
+      "loss": 4.6161,
+      "step": 229
+    },
+    {
+      "epoch": 0.020820132162578074,
+      "grad_norm": 1.0954371690750122,
+      "learning_rate": 0.00019991492058932142,
+      "loss": 4.6043,
+      "step": 230
+    },
+    {
+      "epoch": 0.020910654476328415,
+      "grad_norm": 0.9638803601264954,
+      "learning_rate": 0.00019991360683025238,
+      "loss": 4.6289,
+      "step": 231
+    },
+    {
+      "epoch": 0.021001176790078756,
+      "grad_norm": 0.8331062197685242,
+      "learning_rate": 0.00019991228300988585,
+      "loss": 4.6163,
+      "step": 232
+    },
+    {
+      "epoch": 0.021091699103829093,
+      "grad_norm": 0.9988797307014465,
+      "learning_rate": 0.00019991094912835515,
+      "loss": 4.6584,
+      "step": 233
+    },
+    {
+      "epoch": 0.021182221417579434,
+      "grad_norm": 0.9852668046951294,
+      "learning_rate": 0.0001999096051857946,
+      "loss": 4.6209,
+      "step": 234
+    },
+    {
+      "epoch": 0.02127274373132977,
+      "grad_norm": 1.0406630039215088,
+      "learning_rate": 0.00019990825118233957,
+      "loss": 4.578,
+      "step": 235
+    },
+    {
+      "epoch": 0.021363266045080112,
+      "grad_norm": 0.97223299741745,
+      "learning_rate": 0.0001999068871181264,
+      "loss": 4.7119,
+      "step": 236
+    },
+    {
+      "epoch": 0.021453788358830453,
+      "grad_norm": 1.0562740564346313,
+      "learning_rate": 0.00019990551299329238,
+      "loss": 4.6474,
+      "step": 237
+    },
+    {
+      "epoch": 0.02154431067258079,
+      "grad_norm": 0.8278136849403381,
+      "learning_rate": 0.00019990412880797597,
+      "loss": 4.6211,
+      "step": 238
+    },
+    {
+      "epoch": 0.02163483298633113,
+      "grad_norm": 0.9848172664642334,
+      "learning_rate": 0.0001999027345623165,
+      "loss": 4.6811,
+      "step": 239
+    },
+    {
+      "epoch": 0.02172535530008147,
+      "grad_norm": 0.9529289603233337,
+      "learning_rate": 0.0001999013302564544,
+      "loss": 4.6545,
+      "step": 240
+    },
+    {
+      "epoch": 0.02181587761383181,
+      "grad_norm": 0.993384063243866,
+      "learning_rate": 0.00019989991589053108,
+      "loss": 4.6565,
+      "step": 241
+    },
+    {
+      "epoch": 0.02190639992758215,
+      "grad_norm": 0.9394677877426147,
+      "learning_rate": 0.00019989849146468896,
+      "loss": 4.5875,
+      "step": 242
+    },
+    {
+      "epoch": 0.021996922241332487,
+      "grad_norm": 0.9237437844276428,
+      "learning_rate": 0.00019989705697907149,
+      "loss": 4.5709,
+      "step": 243
+    },
+    {
+      "epoch": 0.022087444555082828,
+      "grad_norm": 0.8978863954544067,
+      "learning_rate": 0.00019989561243382312,
+      "loss": 4.6205,
+      "step": 244
+    },
+    {
+      "epoch": 0.02217796686883317,
+      "grad_norm": 0.8756672143936157,
+      "learning_rate": 0.00019989415782908928,
+      "loss": 4.6204,
+      "step": 245
+    },
+    {
+      "epoch": 0.022268489182583506,
+      "grad_norm": 0.9668368101119995,
+      "learning_rate": 0.0001998926931650165,
+      "loss": 4.5963,
+      "step": 246
+    },
+    {
+      "epoch": 0.022359011496333847,
+      "grad_norm": 0.9980696439743042,
+      "learning_rate": 0.00019989121844175224,
+      "loss": 4.5992,
+      "step": 247
+    },
+    {
+      "epoch": 0.022449533810084184,
+      "grad_norm": 1.0099190473556519,
+      "learning_rate": 0.00019988973365944507,
+      "loss": 4.5303,
+      "step": 248
+    },
+    {
+      "epoch": 0.022540056123834525,
+      "grad_norm": 0.9659672975540161,
+      "learning_rate": 0.00019988823881824438,
+      "loss": 4.5601,
+      "step": 249
+    },
+    {
+      "epoch": 0.022630578437584866,
+      "grad_norm": 1.0651359558105469,
+      "learning_rate": 0.0001998867339183008,
+      "loss": 4.609,
+      "step": 250
+    },
+    {
+      "epoch": 0.022721100751335203,
+      "grad_norm": 0.9525238275527954,
+      "learning_rate": 0.00019988521895976586,
+      "loss": 4.6038,
+      "step": 251
+    },
+    {
+      "epoch": 0.022811623065085544,
+      "grad_norm": 0.859809398651123,
+      "learning_rate": 0.0001998836939427921,
+      "loss": 4.6071,
+      "step": 252
+    },
+    {
+      "epoch": 0.02290214537883588,
+      "grad_norm": 1.0387052297592163,
+      "learning_rate": 0.00019988215886753308,
+      "loss": 4.6451,
+      "step": 253
+    },
+    {
+      "epoch": 0.022992667692586222,
+      "grad_norm": 0.9915367960929871,
+      "learning_rate": 0.0001998806137341434,
+      "loss": 4.589,
+      "step": 254
+    },
+    {
+      "epoch": 0.023083190006336563,
+      "grad_norm": 0.8341922163963318,
+      "learning_rate": 0.00019987905854277867,
+      "loss": 4.5568,
+      "step": 255
+    },
+    {
+      "epoch": 0.0231737123200869,
+      "grad_norm": 0.8853479027748108,
+      "learning_rate": 0.00019987749329359548,
+      "loss": 4.5839,
+      "step": 256
+    },
+    {
+      "epoch": 0.02326423463383724,
+      "grad_norm": 0.9336071610450745,
+      "learning_rate": 0.0001998759179867514,
+      "loss": 4.6152,
+      "step": 257
+    },
+    {
+      "epoch": 0.02335475694758758,
+      "grad_norm": 1.0642181634902954,
+      "learning_rate": 0.00019987433262240516,
+      "loss": 4.5539,
+      "step": 258
+    },
+    {
+      "epoch": 0.02344527926133792,
+      "grad_norm": 1.0101536512374878,
+      "learning_rate": 0.00019987273720071632,
+      "loss": 4.6234,
+      "step": 259
+    },
+    {
+      "epoch": 0.02353580157508826,
+      "grad_norm": 0.9692165851593018,
+      "learning_rate": 0.00019987113172184563,
+      "loss": 4.5625,
+      "step": 260
+    },
+    {
+      "epoch": 0.023626323888838598,
+      "grad_norm": 1.0215392112731934,
+      "learning_rate": 0.00019986951618595464,
+      "loss": 4.6085,
+      "step": 261
+    },
+    {
+      "epoch": 0.02371684620258894,
+      "grad_norm": 1.2795698642730713,
+      "learning_rate": 0.00019986789059320615,
+      "loss": 4.5254,
+      "step": 262
+    },
+    {
+      "epoch": 0.02380736851633928,
+      "grad_norm": 0.8494666814804077,
+      "learning_rate": 0.0001998662549437638,
+      "loss": 4.5586,
+      "step": 263
+    },
+    {
+      "epoch": 0.023897890830089617,
+      "grad_norm": 1.0478988885879517,
+      "learning_rate": 0.0001998646092377923,
+      "loss": 4.5894,
+      "step": 264
+    },
+    {
+      "epoch": 0.023988413143839957,
+      "grad_norm": 0.8754821419715881,
+      "learning_rate": 0.0001998629534754574,
+      "loss": 4.5618,
+      "step": 265
+    },
+    {
+      "epoch": 0.024078935457590295,
+      "grad_norm": 1.0455201864242554,
+      "learning_rate": 0.0001998612876569258,
+      "loss": 4.5031,
+      "step": 266
+    },
+    {
+      "epoch": 0.024169457771340636,
+      "grad_norm": 0.9330147504806519,
+      "learning_rate": 0.00019985961178236529,
+      "loss": 4.5686,
+      "step": 267
+    },
+    {
+      "epoch": 0.024259980085090976,
+      "grad_norm": 0.933779239654541,
+      "learning_rate": 0.00019985792585194457,
+      "loss": 4.5083,
+      "step": 268
+    },
+    {
+      "epoch": 0.024350502398841314,
+      "grad_norm": 0.9001979231834412,
+      "learning_rate": 0.00019985622986583347,
+      "loss": 4.4943,
+      "step": 269
+    },
+    {
+      "epoch": 0.024441024712591655,
+      "grad_norm": 1.0643647909164429,
+      "learning_rate": 0.00019985452382420275,
+      "loss": 4.5722,
+      "step": 270
+    },
+    {
+      "epoch": 0.024531547026341992,
+      "grad_norm": 0.876860499382019,
+      "learning_rate": 0.00019985280772722423,
+      "loss": 4.5179,
+      "step": 271
+    },
+    {
+      "epoch": 0.024622069340092333,
+      "grad_norm": 1.0777976512908936,
+      "learning_rate": 0.00019985108157507067,
+      "loss": 4.5369,
+      "step": 272
+    },
+    {
+      "epoch": 0.024712591653842673,
+      "grad_norm": 0.962979257106781,
+      "learning_rate": 0.00019984934536791595,
+      "loss": 4.6008,
+      "step": 273
+    },
+    {
+      "epoch": 0.02480311396759301,
+      "grad_norm": 0.8431717157363892,
+      "learning_rate": 0.00019984759910593488,
+      "loss": 4.4974,
+      "step": 274
+    },
+    {
+      "epoch": 0.02489363628134335,
+      "grad_norm": 0.915391206741333,
+      "learning_rate": 0.0001998458427893033,
+      "loss": 4.4514,
+      "step": 275
+    },
+    {
+      "epoch": 0.02498415859509369,
+      "grad_norm": 0.8736391067504883,
+      "learning_rate": 0.00019984407641819812,
+      "loss": 4.5691,
+      "step": 276
+    },
+    {
+      "epoch": 0.02507468090884403,
+      "grad_norm": 0.9402247667312622,
+      "learning_rate": 0.00019984229999279714,
+      "loss": 4.406,
+      "step": 277
+    },
+    {
+      "epoch": 0.02516520322259437,
+      "grad_norm": 0.9595934152603149,
+      "learning_rate": 0.0001998405135132793,
+      "loss": 4.5847,
+      "step": 278
+    },
+    {
+      "epoch": 0.025255725536344708,
+      "grad_norm": 0.8316798210144043,
+      "learning_rate": 0.00019983871697982445,
+      "loss": 4.4519,
+      "step": 279
+    },
+    {
+      "epoch": 0.02534624785009505,
+      "grad_norm": 0.8516654968261719,
+      "learning_rate": 0.00019983691039261357,
+      "loss": 4.5498,
+      "step": 280
+    },
+    {
+      "epoch": 0.02543677016384539,
+      "grad_norm": 0.915367603302002,
+      "learning_rate": 0.00019983509375182855,
+      "loss": 4.5475,
+      "step": 281
+    },
+    {
+      "epoch": 0.025527292477595727,
+      "grad_norm": 0.9939991235733032,
+      "learning_rate": 0.0001998332670576523,
+      "loss": 4.5124,
+      "step": 282
+    },
+    {
+      "epoch": 0.025617814791346068,
+      "grad_norm": 1.2104617357254028,
+      "learning_rate": 0.00019983143031026879,
+      "loss": 4.4694,
+      "step": 283
+    },
+    {
+      "epoch": 0.025708337105096405,
+      "grad_norm": 1.1700996160507202,
+      "learning_rate": 0.00019982958350986296,
+      "loss": 4.5415,
+      "step": 284
+    },
+    {
+      "epoch": 0.025798859418846746,
+      "grad_norm": 1.249987006187439,
+      "learning_rate": 0.00019982772665662083,
+      "loss": 4.5698,
+      "step": 285
+    },
+    {
+      "epoch": 0.025889381732597087,
+      "grad_norm": 0.9590091109275818,
+      "learning_rate": 0.00019982585975072937,
+      "loss": 4.567,
+      "step": 286
+    },
+    {
+      "epoch": 0.025979904046347424,
+      "grad_norm": 1.1153589487075806,
+      "learning_rate": 0.00019982398279237655,
+      "loss": 4.4925,
+      "step": 287
+    },
+    {
+      "epoch": 0.026070426360097765,
+      "grad_norm": 0.9221207499504089,
+      "learning_rate": 0.00019982209578175137,
+      "loss": 4.5124,
+      "step": 288
+    },
+    {
+      "epoch": 0.026160948673848102,
+      "grad_norm": 0.9073323607444763,
+      "learning_rate": 0.00019982019871904393,
+      "loss": 4.5402,
+      "step": 289
+    },
+    {
+      "epoch": 0.026251470987598443,
+      "grad_norm": 0.9508566856384277,
+      "learning_rate": 0.00019981829160444514,
+      "loss": 4.5094,
+      "step": 290
+    },
+    {
+      "epoch": 0.026341993301348784,
+      "grad_norm": 0.963015615940094,
+      "learning_rate": 0.00019981637443814717,
+      "loss": 4.5076,
+      "step": 291
+    },
+    {
+      "epoch": 0.02643251561509912,
+      "grad_norm": 1.0233488082885742,
+      "learning_rate": 0.000199814447220343,
+      "loss": 4.4913,
+      "step": 292
+    },
+    {
+      "epoch": 0.026523037928849462,
+      "grad_norm": 0.9522859454154968,
+      "learning_rate": 0.00019981250995122673,
+      "loss": 4.4579,
+      "step": 293
+    },
+    {
+      "epoch": 0.0266135602425998,
+      "grad_norm": 0.8754635453224182,
+      "learning_rate": 0.00019981056263099344,
+      "loss": 4.4501,
+      "step": 294
+    },
+    {
+      "epoch": 0.02670408255635014,
+      "grad_norm": 0.9538990259170532,
+      "learning_rate": 0.00019980860525983923,
+      "loss": 4.47,
+      "step": 295
+    },
+    {
+      "epoch": 0.02679460487010048,
+      "grad_norm": 0.9469854235649109,
+      "learning_rate": 0.00019980663783796118,
+      "loss": 4.4804,
+      "step": 296
+    },
+    {
+      "epoch": 0.026885127183850818,
+      "grad_norm": 0.8881922364234924,
+      "learning_rate": 0.00019980466036555744,
+      "loss": 4.437,
+      "step": 297
+    },
+    {
+      "epoch": 0.02697564949760116,
+      "grad_norm": 0.9970735311508179,
+      "learning_rate": 0.00019980267284282717,
+      "loss": 4.5339,
+      "step": 298
+    },
+    {
+      "epoch": 0.0270661718113515,
+      "grad_norm": 1.0509942770004272,
+      "learning_rate": 0.00019980067526997045,
+      "loss": 4.5365,
+      "step": 299
+    },
+    {
+      "epoch": 0.027156694125101837,
+      "grad_norm": 0.9530550241470337,
+      "learning_rate": 0.00019979866764718843,
+      "loss": 4.4565,
+      "step": 300
+    },
+    {
+      "epoch": 0.027247216438852178,
+      "grad_norm": 1.178865909576416,
+      "learning_rate": 0.00019979664997468337,
+      "loss": 4.5071,
+      "step": 301
+    },
+    {
+      "epoch": 0.027337738752602515,
+      "grad_norm": 0.932168185710907,
+      "learning_rate": 0.00019979462225265833,
+      "loss": 4.4988,
+      "step": 302
+    },
+    {
+      "epoch": 0.027428261066352856,
+      "grad_norm": 0.9907435774803162,
+      "learning_rate": 0.00019979258448131758,
+      "loss": 4.4837,
+      "step": 303
+    },
+    {
+      "epoch": 0.027518783380103197,
+      "grad_norm": 0.9297950267791748,
+      "learning_rate": 0.00019979053666086634,
+      "loss": 4.4823,
+      "step": 304
+    },
+    {
+      "epoch": 0.027609305693853534,
+      "grad_norm": 0.9497891664505005,
+      "learning_rate": 0.00019978847879151076,
+      "loss": 4.4843,
+      "step": 305
+    },
+    {
+      "epoch": 0.027699828007603875,
+      "grad_norm": 0.9658074975013733,
+      "learning_rate": 0.00019978641087345806,
+      "loss": 4.3628,
+      "step": 306
+    },
+    {
+      "epoch": 0.027790350321354212,
+      "grad_norm": 0.9158378839492798,
+      "learning_rate": 0.00019978433290691655,
+      "loss": 4.4123,
+      "step": 307
+    },
+    {
+      "epoch": 0.027880872635104553,
+      "grad_norm": 1.1265052556991577,
+      "learning_rate": 0.00019978224489209545,
+      "loss": 4.471,
+      "step": 308
+    },
+    {
+      "epoch": 0.027971394948854894,
+      "grad_norm": 0.8981154561042786,
+      "learning_rate": 0.000199780146829205,
+      "loss": 4.4544,
+      "step": 309
+    },
+    {
+      "epoch": 0.02806191726260523,
+      "grad_norm": 0.9838968515396118,
+      "learning_rate": 0.0001997780387184565,
+      "loss": 4.5014,
+      "step": 310
+    },
+    {
+      "epoch": 0.028152439576355572,
+      "grad_norm": 0.991176962852478,
+      "learning_rate": 0.00019977592056006228,
+      "loss": 4.3644,
+      "step": 311
+    },
+    {
+      "epoch": 0.02824296189010591,
+      "grad_norm": 0.986817479133606,
+      "learning_rate": 0.00019977379235423551,
+      "loss": 4.4363,
+      "step": 312
+    },
+    {
+      "epoch": 0.02833348420385625,
+      "grad_norm": 1.0609725713729858,
+      "learning_rate": 0.00019977165410119062,
+      "loss": 4.4627,
+      "step": 313
+    },
+    {
+      "epoch": 0.02842400651760659,
+      "grad_norm": 1.1476916074752808,
+      "learning_rate": 0.0001997695058011429,
+      "loss": 4.4855,
+      "step": 314
+    },
+    {
+      "epoch": 0.02851452883135693,
+      "grad_norm": 1.1370539665222168,
+      "learning_rate": 0.00019976734745430868,
+      "loss": 4.4234,
+      "step": 315
+    },
+    {
+      "epoch": 0.02860505114510727,
+      "grad_norm": 1.0598125457763672,
+      "learning_rate": 0.00019976517906090529,
+      "loss": 4.4257,
+      "step": 316
+    },
+    {
+      "epoch": 0.028695573458857607,
+      "grad_norm": 1.014866590499878,
+      "learning_rate": 0.0001997630006211511,
+      "loss": 4.3799,
+      "step": 317
+    },
+    {
+      "epoch": 0.028786095772607947,
+      "grad_norm": 0.9023067951202393,
+      "learning_rate": 0.00019976081213526545,
+      "loss": 4.4157,
+      "step": 318
+    },
+    {
+      "epoch": 0.02887661808635829,
+      "grad_norm": 1.0376299619674683,
+      "learning_rate": 0.00019975861360346876,
+      "loss": 4.403,
+      "step": 319
+    },
+    {
+      "epoch": 0.028967140400108626,
+      "grad_norm": 1.077791452407837,
+      "learning_rate": 0.00019975640502598244,
+      "loss": 4.3543,
+      "step": 320
+    },
+    {
+      "epoch": 0.029057662713858966,
+      "grad_norm": 0.9885241389274597,
+      "learning_rate": 0.00019975418640302882,
+      "loss": 4.3259,
+      "step": 321
+    },
+    {
+      "epoch": 0.029148185027609307,
+      "grad_norm": 1.0007898807525635,
+      "learning_rate": 0.0001997519577348314,
+      "loss": 4.4491,
+      "step": 322
+    },
+    {
+      "epoch": 0.029238707341359645,
+      "grad_norm": 0.9583142995834351,
+      "learning_rate": 0.00019974971902161453,
+      "loss": 4.4285,
+      "step": 323
+    },
+    {
+      "epoch": 0.029329229655109985,
+      "grad_norm": 1.1063108444213867,
+      "learning_rate": 0.00019974747026360373,
+      "loss": 4.3511,
+      "step": 324
+    },
+    {
+      "epoch": 0.029419751968860323,
+      "grad_norm": 1.0260143280029297,
+      "learning_rate": 0.00019974521146102537,
+      "loss": 4.3798,
+      "step": 325
+    },
+    {
+      "epoch": 0.029510274282610664,
+      "grad_norm": 0.9445636868476868,
+      "learning_rate": 0.00019974294261410695,
+      "loss": 4.4577,
+      "step": 326
+    },
+    {
+      "epoch": 0.029600796596361004,
+      "grad_norm": 0.9906353950500488,
+      "learning_rate": 0.00019974066372307694,
+      "loss": 4.427,
+      "step": 327
+    },
+    {
+      "epoch": 0.02969131891011134,
+      "grad_norm": 1.0205063819885254,
+      "learning_rate": 0.0001997383747881648,
+      "loss": 4.4002,
+      "step": 328
+    },
+    {
+      "epoch": 0.029781841223861683,
+      "grad_norm": 1.0043867826461792,
+      "learning_rate": 0.00019973607580960108,
+      "loss": 4.3951,
+      "step": 329
+    },
+    {
+      "epoch": 0.02987236353761202,
+      "grad_norm": 0.9098749160766602,
+      "learning_rate": 0.00019973376678761724,
+      "loss": 4.3715,
+      "step": 330
+    },
+    {
+      "epoch": 0.02996288585136236,
+      "grad_norm": 1.0196176767349243,
+      "learning_rate": 0.00019973144772244582,
+      "loss": 4.5046,
+      "step": 331
+    },
+    {
+      "epoch": 0.0300534081651127,
+      "grad_norm": 0.970935583114624,
+      "learning_rate": 0.00019972911861432032,
+      "loss": 4.4171,
+      "step": 332
+    },
+    {
+      "epoch": 0.03014393047886304,
+      "grad_norm": 1.073278546333313,
+      "learning_rate": 0.00019972677946347535,
+      "loss": 4.4271,
+      "step": 333
+    },
+    {
+      "epoch": 0.03023445279261338,
+      "grad_norm": 0.9405482411384583,
+      "learning_rate": 0.0001997244302701464,
+      "loss": 4.3853,
+      "step": 334
+    },
+    {
+      "epoch": 0.030324975106363717,
+      "grad_norm": 1.1856549978256226,
+      "learning_rate": 0.00019972207103457001,
+      "loss": 4.4458,
+      "step": 335
+    },
+    {
+      "epoch": 0.030415497420114058,
+      "grad_norm": 1.0638178586959839,
+      "learning_rate": 0.00019971970175698385,
+      "loss": 4.4305,
+      "step": 336
+    },
+    {
+      "epoch": 0.0305060197338644,
+      "grad_norm": 0.9577354788780212,
+      "learning_rate": 0.00019971732243762643,
+      "loss": 4.3614,
+      "step": 337
+    },
+    {
+      "epoch": 0.030596542047614736,
+      "grad_norm": 0.867456316947937,
+      "learning_rate": 0.00019971493307673735,
+      "loss": 4.4327,
+      "step": 338
+    },
+    {
+      "epoch": 0.030687064361365077,
+      "grad_norm": 0.9249234199523926,
+      "learning_rate": 0.00019971253367455727,
+      "loss": 4.3964,
+      "step": 339
+    },
+    {
+      "epoch": 0.030777586675115418,
+      "grad_norm": 0.9764379262924194,
+      "learning_rate": 0.00019971012423132775,
+      "loss": 4.3969,
+      "step": 340
+    },
+    {
+      "epoch": 0.030868108988865755,
+      "grad_norm": 0.9793526530265808,
+      "learning_rate": 0.00019970770474729145,
+      "loss": 4.3077,
+      "step": 341
+    },
+    {
+      "epoch": 0.030958631302616096,
+      "grad_norm": 0.996056318283081,
+      "learning_rate": 0.00019970527522269205,
+      "loss": 4.3773,
+      "step": 342
+    },
+    {
+      "epoch": 0.031049153616366433,
+      "grad_norm": 0.9756794571876526,
+      "learning_rate": 0.0001997028356577741,
+      "loss": 4.3676,
+      "step": 343
+    },
+    {
+      "epoch": 0.031139675930116774,
+      "grad_norm": 0.9667348265647888,
+      "learning_rate": 0.00019970038605278338,
+      "loss": 4.3651,
+      "step": 344
+    },
+    {
+      "epoch": 0.031230198243867115,
+      "grad_norm": 1.0161426067352295,
+      "learning_rate": 0.0001996979264079665,
+      "loss": 4.4087,
+      "step": 345
+    },
+    {
+      "epoch": 0.031320720557617455,
+      "grad_norm": 0.9502101540565491,
+      "learning_rate": 0.00019969545672357116,
+      "loss": 4.3836,
+      "step": 346
+    },
+    {
+      "epoch": 0.03141124287136779,
+      "grad_norm": 1.0001097917556763,
+      "learning_rate": 0.00019969297699984603,
+      "loss": 4.3798,
+      "step": 347
+    },
+    {
+      "epoch": 0.03150176518511813,
+      "grad_norm": 0.9770258069038391,
+      "learning_rate": 0.00019969048723704087,
+      "loss": 4.4348,
+      "step": 348
+    },
+    {
+      "epoch": 0.03159228749886847,
+      "grad_norm": 0.9103361964225769,
+      "learning_rate": 0.00019968798743540635,
+      "loss": 4.4196,
+      "step": 349
+    },
+    {
+      "epoch": 0.03168280981261881,
+      "grad_norm": 1.0442837476730347,
+      "learning_rate": 0.00019968547759519425,
+      "loss": 4.3155,
+      "step": 350
+    },
+    {
+      "epoch": 0.03177333212636915,
+      "grad_norm": 0.9110909700393677,
+      "learning_rate": 0.00019968295771665727,
+      "loss": 4.2968,
+      "step": 351
+    },
+    {
+      "epoch": 0.031863854440119486,
+      "grad_norm": 1.1041204929351807,
+      "learning_rate": 0.00019968042780004917,
+      "loss": 4.3275,
+      "step": 352
+    },
+    {
+      "epoch": 0.03195437675386983,
+      "grad_norm": 1.1094104051589966,
+      "learning_rate": 0.00019967788784562473,
+      "loss": 4.3965,
+      "step": 353
+    },
+    {
+      "epoch": 0.03204489906762017,
+      "grad_norm": 1.0326623916625977,
+      "learning_rate": 0.0001996753378536397,
+      "loss": 4.3922,
+      "step": 354
+    },
+    {
+      "epoch": 0.03213542138137051,
+      "grad_norm": 1.1457520723342896,
+      "learning_rate": 0.00019967277782435088,
+      "loss": 4.3337,
+      "step": 355
+    },
+    {
+      "epoch": 0.03222594369512085,
+      "grad_norm": 1.0908020734786987,
+      "learning_rate": 0.00019967020775801605,
+      "loss": 4.3584,
+      "step": 356
+    },
+    {
+      "epoch": 0.032316466008871184,
+      "grad_norm": 1.1035650968551636,
+      "learning_rate": 0.00019966762765489407,
+      "loss": 4.3424,
+      "step": 357
+    },
+    {
+      "epoch": 0.032406988322621524,
+      "grad_norm": 0.9557312726974487,
+      "learning_rate": 0.00019966503751524465,
+      "loss": 4.3228,
+      "step": 358
+    },
+    {
+      "epoch": 0.032497510636371865,
+      "grad_norm": 1.026123046875,
+      "learning_rate": 0.00019966243733932873,
+      "loss": 4.3184,
+      "step": 359
+    },
+    {
+      "epoch": 0.032588032950122206,
+      "grad_norm": 0.9995486736297607,
+      "learning_rate": 0.00019965982712740808,
+      "loss": 4.3935,
+      "step": 360
+    },
+    {
+      "epoch": 0.03267855526387255,
+      "grad_norm": 1.0720714330673218,
+      "learning_rate": 0.00019965720687974554,
+      "loss": 4.3373,
+      "step": 361
+    },
+    {
+      "epoch": 0.03276907757762288,
+      "grad_norm": 0.9495123028755188,
+      "learning_rate": 0.00019965457659660503,
+      "loss": 4.3709,
+      "step": 362
+    },
+    {
+      "epoch": 0.03285959989137322,
+      "grad_norm": 0.9384881854057312,
+      "learning_rate": 0.00019965193627825137,
+      "loss": 4.3316,
+      "step": 363
+    },
+    {
+      "epoch": 0.03295012220512356,
+      "grad_norm": 0.9817916750907898,
+      "learning_rate": 0.00019964928592495045,
+      "loss": 4.3621,
+      "step": 364
+    },
+    {
+      "epoch": 0.0330406445188739,
+      "grad_norm": 0.9114745855331421,
+      "learning_rate": 0.00019964662553696917,
+      "loss": 4.3202,
+      "step": 365
+    },
+    {
+      "epoch": 0.033131166832624244,
+      "grad_norm": 0.9784656763076782,
+      "learning_rate": 0.0001996439551145754,
+      "loss": 4.3443,
+      "step": 366
+    },
+    {
+      "epoch": 0.033221689146374585,
+      "grad_norm": 1.0070286989212036,
+      "learning_rate": 0.00019964127465803812,
+      "loss": 4.3719,
+      "step": 367
+    },
+    {
+      "epoch": 0.03331221146012492,
+      "grad_norm": 0.9459866285324097,
+      "learning_rate": 0.00019963858416762717,
+      "loss": 4.3708,
+      "step": 368
+    },
+    {
+      "epoch": 0.03340273377387526,
+      "grad_norm": 0.937214732170105,
+      "learning_rate": 0.00019963588364361353,
+      "loss": 4.353,
+      "step": 369
+    },
+    {
+      "epoch": 0.0334932560876256,
+      "grad_norm": 0.8671116828918457,
+      "learning_rate": 0.00019963317308626914,
+      "loss": 4.3555,
+      "step": 370
+    },
+    {
+      "epoch": 0.03358377840137594,
+      "grad_norm": 0.9910116791725159,
+      "learning_rate": 0.00019963045249586693,
+      "loss": 4.375,
+      "step": 371
+    },
+    {
+      "epoch": 0.03367430071512628,
+      "grad_norm": 0.9444873332977295,
+      "learning_rate": 0.00019962772187268093,
+      "loss": 4.3282,
+      "step": 372
+    },
+    {
+      "epoch": 0.033764823028876616,
+      "grad_norm": 0.8867988586425781,
+      "learning_rate": 0.000199624981216986,
+      "loss": 4.3491,
+      "step": 373
+    },
+    {
+      "epoch": 0.03385534534262696,
+      "grad_norm": 1.096407413482666,
+      "learning_rate": 0.0001996222305290582,
+      "loss": 4.3438,
+      "step": 374
+    },
+    {
+      "epoch": 0.0339458676563773,
+      "grad_norm": 0.9576050043106079,
+      "learning_rate": 0.00019961946980917456,
+      "loss": 4.3186,
+      "step": 375
+    },
+    {
+      "epoch": 0.03403638997012764,
+      "grad_norm": 1.0348565578460693,
+      "learning_rate": 0.00019961669905761302,
+      "loss": 4.2565,
+      "step": 376
+    },
+    {
+      "epoch": 0.03412691228387798,
+      "grad_norm": 0.8637972474098206,
+      "learning_rate": 0.0001996139182746526,
+      "loss": 4.4194,
+      "step": 377
+    },
+    {
+      "epoch": 0.03421743459762831,
+      "grad_norm": 0.9136695265769958,
+      "learning_rate": 0.0001996111274605733,
+      "loss": 4.3234,
+      "step": 378
+    },
+    {
+      "epoch": 0.034307956911378654,
+      "grad_norm": 1.0527623891830444,
+      "learning_rate": 0.00019960832661565622,
+      "loss": 4.3159,
+      "step": 379
+    },
+    {
+      "epoch": 0.034398479225128994,
+      "grad_norm": 0.9086984992027283,
+      "learning_rate": 0.0001996055157401834,
+      "loss": 4.3534,
+      "step": 380
+    },
+    {
+      "epoch": 0.034489001538879335,
+      "grad_norm": 0.9098812341690063,
+      "learning_rate": 0.00019960269483443782,
+      "loss": 4.3152,
+      "step": 381
+    },
+    {
+      "epoch": 0.034579523852629676,
+      "grad_norm": 0.9199385046958923,
+      "learning_rate": 0.00019959986389870364,
+      "loss": 4.313,
+      "step": 382
+    },
+    {
+      "epoch": 0.03467004616638001,
+      "grad_norm": 1.0625910758972168,
+      "learning_rate": 0.00019959702293326586,
+      "loss": 4.305,
+      "step": 383
+    },
+    {
+      "epoch": 0.03476056848013035,
+      "grad_norm": 1.0181784629821777,
+      "learning_rate": 0.00019959417193841063,
+      "loss": 4.3,
+      "step": 384
+    },
+    {
+      "epoch": 0.03485109079388069,
+      "grad_norm": 0.9671763181686401,
+      "learning_rate": 0.00019959131091442495,
+      "loss": 4.3007,
+      "step": 385
+    },
+    {
+      "epoch": 0.03494161310763103,
+      "grad_norm": 0.9672999978065491,
+      "learning_rate": 0.00019958843986159704,
+      "loss": 4.3513,
+      "step": 386
+    },
+    {
+      "epoch": 0.03503213542138137,
+      "grad_norm": 0.9663201570510864,
+      "learning_rate": 0.00019958555878021595,
+      "loss": 4.3138,
+      "step": 387
+    },
+    {
+      "epoch": 0.03512265773513171,
+      "grad_norm": 1.073101282119751,
+      "learning_rate": 0.0001995826676705718,
+      "loss": 4.3399,
+      "step": 388
+    },
+    {
+      "epoch": 0.03521318004888205,
+      "grad_norm": 0.956693708896637,
+      "learning_rate": 0.00019957976653295576,
+      "loss": 4.3208,
+      "step": 389
+    },
+    {
+      "epoch": 0.03530370236263239,
+      "grad_norm": 0.952547550201416,
+      "learning_rate": 0.00019957685536765995,
+      "loss": 4.3082,
+      "step": 390
+    },
+    {
+      "epoch": 0.03539422467638273,
+      "grad_norm": 0.9408294558525085,
+      "learning_rate": 0.0001995739341749775,
+      "loss": 4.3202,
+      "step": 391
+    },
+    {
+      "epoch": 0.03548474699013307,
+      "grad_norm": 1.056509017944336,
+      "learning_rate": 0.00019957100295520266,
+      "loss": 4.258,
+      "step": 392
+    },
+    {
+      "epoch": 0.035575269303883404,
+      "grad_norm": 1.1849064826965332,
+      "learning_rate": 0.00019956806170863049,
+      "loss": 4.2442,
+      "step": 393
+    },
+    {
+      "epoch": 0.035665791617633745,
+      "grad_norm": 0.8983389139175415,
+      "learning_rate": 0.00019956511043555728,
+      "loss": 4.2728,
+      "step": 394
+    },
+    {
+      "epoch": 0.035756313931384086,
+      "grad_norm": 1.0108262300491333,
+      "learning_rate": 0.00019956214913628015,
+      "loss": 4.249,
+      "step": 395
+    },
+    {
+      "epoch": 0.03584683624513443,
+      "grad_norm": 0.9574976563453674,
+      "learning_rate": 0.0001995591778110973,
+      "loss": 4.2356,
+      "step": 396
+    },
+    {
+      "epoch": 0.03593735855888477,
+      "grad_norm": 1.0257916450500488,
+      "learning_rate": 0.00019955619646030802,
+      "loss": 4.2957,
+      "step": 397
+    },
+    {
+      "epoch": 0.0360278808726351,
+      "grad_norm": 1.0345457792282104,
+      "learning_rate": 0.00019955320508421243,
+      "loss": 4.2462,
+      "step": 398
+    },
+    {
+      "epoch": 0.03611840318638544,
+      "grad_norm": 0.9709088802337646,
+      "learning_rate": 0.00019955020368311183,
+      "loss": 4.2942,
+      "step": 399
+    },
+    {
+      "epoch": 0.03620892550013578,
+      "grad_norm": 0.9114799499511719,
+      "learning_rate": 0.00019954719225730847,
+      "loss": 4.3045,
+      "step": 400
+    },
+    {
+      "epoch": 0.036299447813886124,
+      "grad_norm": 1.0316181182861328,
+      "learning_rate": 0.00019954417080710556,
+      "loss": 4.3309,
+      "step": 401
+    },
+    {
+      "epoch": 0.036389970127636465,
+      "grad_norm": 0.9053332209587097,
+      "learning_rate": 0.00019954113933280736,
+      "loss": 4.2553,
+      "step": 402
+    },
+    {
+      "epoch": 0.0364804924413868,
+      "grad_norm": 0.9040462970733643,
+      "learning_rate": 0.00019953809783471917,
+      "loss": 4.2645,
+      "step": 403
+    },
+    {
+      "epoch": 0.03657101475513714,
+      "grad_norm": 0.8616097569465637,
+      "learning_rate": 0.0001995350463131472,
+      "loss": 4.2518,
+      "step": 404
+    },
+    {
+      "epoch": 0.03666153706888748,
+      "grad_norm": 0.8178855180740356,
+      "learning_rate": 0.00019953198476839888,
+      "loss": 4.2194,
+      "step": 405
+    },
+    {
+      "epoch": 0.03675205938263782,
+      "grad_norm": 0.8645557761192322,
+      "learning_rate": 0.00019952891320078236,
+      "loss": 4.1956,
+      "step": 406
+    },
+    {
+      "epoch": 0.03684258169638816,
+      "grad_norm": 0.9388260841369629,
+      "learning_rate": 0.000199525831610607,
+      "loss": 4.2618,
+      "step": 407
+    },
+    {
+      "epoch": 0.0369331040101385,
+      "grad_norm": 0.992255687713623,
+      "learning_rate": 0.0001995227399981831,
+      "loss": 4.2326,
+      "step": 408
+    },
+    {
+      "epoch": 0.037023626323888836,
+      "grad_norm": 0.964103639125824,
+      "learning_rate": 0.00019951963836382206,
+      "loss": 4.3077,
+      "step": 409
+    },
+    {
+      "epoch": 0.03711414863763918,
+      "grad_norm": 0.9777807593345642,
+      "learning_rate": 0.00019951652670783615,
+      "loss": 4.2155,
+      "step": 410
+    },
+    {
+      "epoch": 0.03720467095138952,
+      "grad_norm": 0.9376230835914612,
+      "learning_rate": 0.00019951340503053872,
+      "loss": 4.2368,
+      "step": 411
+    },
+    {
+      "epoch": 0.03729519326513986,
+      "grad_norm": 1.0110722780227661,
+      "learning_rate": 0.0001995102733322441,
+      "loss": 4.2638,
+      "step": 412
+    },
+    {
+      "epoch": 0.0373857155788902,
+      "grad_norm": 1.151977300643921,
+      "learning_rate": 0.00019950713161326767,
+      "loss": 4.2924,
+      "step": 413
+    },
+    {
+      "epoch": 0.03747623789264053,
+      "grad_norm": 0.9603147506713867,
+      "learning_rate": 0.00019950397987392586,
+      "loss": 4.2671,
+      "step": 414
+    },
+    {
+      "epoch": 0.037566760206390874,
+      "grad_norm": 1.1345927715301514,
+      "learning_rate": 0.00019950081811453597,
+      "loss": 4.2101,
+      "step": 415
+    },
+    {
+      "epoch": 0.037657282520141215,
+      "grad_norm": 1.1134638786315918,
+      "learning_rate": 0.00019949764633541643,
+      "loss": 4.282,
+      "step": 416
+    },
+    {
+      "epoch": 0.037747804833891556,
+      "grad_norm": 1.0932708978652954,
+      "learning_rate": 0.00019949446453688662,
+      "loss": 4.3003,
+      "step": 417
+    },
+    {
+      "epoch": 0.0378383271476419,
+      "grad_norm": 0.9736213088035583,
+      "learning_rate": 0.00019949127271926695,
+      "loss": 4.328,
+      "step": 418
+    },
+    {
+      "epoch": 0.03792884946139223,
+      "grad_norm": 1.0969462394714355,
+      "learning_rate": 0.00019948807088287883,
+      "loss": 4.2475,
+      "step": 419
+    },
+    {
+      "epoch": 0.03801937177514257,
+      "grad_norm": 1.046892523765564,
+      "learning_rate": 0.0001994848590280447,
+      "loss": 4.1855,
+      "step": 420
+    },
+    {
+      "epoch": 0.03810989408889291,
+      "grad_norm": 1.1222968101501465,
+      "learning_rate": 0.00019948163715508802,
+      "loss": 4.2414,
+      "step": 421
+    },
+    {
+      "epoch": 0.03820041640264325,
+      "grad_norm": 1.0552254915237427,
+      "learning_rate": 0.00019947840526433314,
+      "loss": 4.2269,
+      "step": 422
+    },
+    {
+      "epoch": 0.038290938716393594,
+      "grad_norm": 1.12909996509552,
+      "learning_rate": 0.00019947516335610562,
+      "loss": 4.2008,
+      "step": 423
+    },
+    {
+      "epoch": 0.03838146103014393,
+      "grad_norm": 1.0824707746505737,
+      "learning_rate": 0.00019947191143073186,
+      "loss": 4.2587,
+      "step": 424
+    },
+    {
+      "epoch": 0.03847198334389427,
+      "grad_norm": 1.0686942338943481,
+      "learning_rate": 0.00019946864948853935,
+      "loss": 4.2781,
+      "step": 425
+    },
+    {
+      "epoch": 0.03856250565764461,
+      "grad_norm": 0.9873172640800476,
+      "learning_rate": 0.0001994653775298565,
+      "loss": 4.2174,
+      "step": 426
+    },
+    {
+      "epoch": 0.03865302797139495,
+      "grad_norm": 0.9392846822738647,
+      "learning_rate": 0.00019946209555501293,
+      "loss": 4.2374,
+      "step": 427
+    },
+    {
+      "epoch": 0.03874355028514529,
+      "grad_norm": 1.0531303882598877,
+      "learning_rate": 0.00019945880356433903,
+      "loss": 4.2903,
+      "step": 428
+    },
+    {
+      "epoch": 0.038834072598895625,
+      "grad_norm": 0.963618814945221,
+      "learning_rate": 0.00019945550155816633,
+      "loss": 4.2779,
+      "step": 429
+    },
+    {
+      "epoch": 0.038924594912645966,
+      "grad_norm": 0.9659363031387329,
+      "learning_rate": 0.00019945218953682734,
+      "loss": 4.1951,
+      "step": 430
+    },
+    {
+      "epoch": 0.039015117226396306,
+      "grad_norm": 0.9671744704246521,
+      "learning_rate": 0.00019944886750065558,
+      "loss": 4.2058,
+      "step": 431
+    },
+    {
+      "epoch": 0.03910563954014665,
+      "grad_norm": 1.0616800785064697,
+      "learning_rate": 0.00019944553544998562,
+      "loss": 4.1887,
+      "step": 432
+    },
+    {
+      "epoch": 0.03919616185389699,
+      "grad_norm": 0.9300423264503479,
+      "learning_rate": 0.00019944219338515292,
+      "loss": 4.1821,
+      "step": 433
+    },
+    {
+      "epoch": 0.03928668416764732,
+      "grad_norm": 1.0894814729690552,
+      "learning_rate": 0.00019943884130649409,
+      "loss": 4.2175,
+      "step": 434
+    },
+    {
+      "epoch": 0.03937720648139766,
+      "grad_norm": 0.9679089188575745,
+      "learning_rate": 0.00019943547921434665,
+      "loss": 4.2315,
+      "step": 435
+    },
+    {
+      "epoch": 0.039467728795148004,
+      "grad_norm": 1.0963354110717773,
+      "learning_rate": 0.00019943210710904918,
+      "loss": 4.2163,
+      "step": 436
+    },
+    {
+      "epoch": 0.039558251108898344,
+      "grad_norm": 1.0342007875442505,
+      "learning_rate": 0.00019942872499094123,
+      "loss": 4.1901,
+      "step": 437
+    },
+    {
+      "epoch": 0.039648773422648685,
+      "grad_norm": 0.9861835837364197,
+      "learning_rate": 0.0001994253328603634,
+      "loss": 4.183,
+      "step": 438
+    },
+    {
+      "epoch": 0.03973929573639902,
+      "grad_norm": 1.081486701965332,
+      "learning_rate": 0.0001994219307176573,
+      "loss": 4.1647,
+      "step": 439
+    },
+    {
+      "epoch": 0.03982981805014936,
+      "grad_norm": 1.042413592338562,
+      "learning_rate": 0.00019941851856316548,
+      "loss": 4.1848,
+      "step": 440
+    },
+    {
+      "epoch": 0.0399203403638997,
+      "grad_norm": 0.9466920495033264,
+      "learning_rate": 0.00019941509639723155,
+      "loss": 4.1853,
+      "step": 441
+    },
+    {
+      "epoch": 0.04001086267765004,
+      "grad_norm": 1.0339183807373047,
+      "learning_rate": 0.00019941166422020014,
+      "loss": 4.271,
+      "step": 442
+    },
+    {
+      "epoch": 0.04010138499140038,
+      "grad_norm": 1.083583116531372,
+      "learning_rate": 0.00019940822203241685,
+      "loss": 4.2329,
+      "step": 443
+    },
+    {
+      "epoch": 0.04019190730515072,
+      "grad_norm": 0.9255849123001099,
+      "learning_rate": 0.00019940476983422832,
+      "loss": 4.1842,
+      "step": 444
+    },
+    {
+      "epoch": 0.04028242961890106,
+      "grad_norm": 1.0079110860824585,
+      "learning_rate": 0.00019940130762598223,
+      "loss": 4.2266,
+      "step": 445
+    },
+    {
+      "epoch": 0.0403729519326514,
+      "grad_norm": 1.0702378749847412,
+      "learning_rate": 0.00019939783540802715,
+      "loss": 4.2801,
+      "step": 446
+    },
+    {
+      "epoch": 0.04046347424640174,
+      "grad_norm": 1.0222079753875732,
+      "learning_rate": 0.00019939435318071277,
+      "loss": 4.1854,
+      "step": 447
+    },
+    {
+      "epoch": 0.04055399656015208,
+      "grad_norm": 1.028741478919983,
+      "learning_rate": 0.00019939086094438975,
+      "loss": 4.2361,
+      "step": 448
+    },
+    {
+      "epoch": 0.04064451887390242,
+      "grad_norm": 1.023998498916626,
+      "learning_rate": 0.00019938735869940973,
+      "loss": 4.2604,
+      "step": 449
+    },
+    {
+      "epoch": 0.040735041187652754,
+      "grad_norm": 0.9994665384292603,
+      "learning_rate": 0.00019938384644612543,
+      "loss": 4.2244,
+      "step": 450
+    },
+    {
+      "epoch": 0.040825563501403095,
+      "grad_norm": 0.9782562851905823,
+      "learning_rate": 0.0001993803241848905,
+      "loss": 4.1538,
+      "step": 451
+    },
+    {
+      "epoch": 0.040916085815153436,
+      "grad_norm": 0.9541926383972168,
+      "learning_rate": 0.00019937679191605963,
+      "loss": 4.2414,
+      "step": 452
+    },
+    {
+      "epoch": 0.041006608128903776,
+      "grad_norm": 0.9733009934425354,
+      "learning_rate": 0.0001993732496399886,
+      "loss": 4.1626,
+      "step": 453
+    },
+    {
+      "epoch": 0.04109713044265412,
+      "grad_norm": 1.0683079957962036,
+      "learning_rate": 0.00019936969735703396,
+      "loss": 4.1847,
+      "step": 454
+    },
+    {
+      "epoch": 0.04118765275640445,
+      "grad_norm": 0.8915784955024719,
+      "learning_rate": 0.00019936613506755358,
+      "loss": 4.2098,
+      "step": 455
+    },
+    {
+      "epoch": 0.04127817507015479,
+      "grad_norm": 0.8858823776245117,
+      "learning_rate": 0.00019936256277190608,
+      "loss": 4.1784,
+      "step": 456
+    },
+    {
+      "epoch": 0.04136869738390513,
+      "grad_norm": 0.9432275891304016,
+      "learning_rate": 0.00019935898047045126,
+      "loss": 4.2451,
+      "step": 457
+    },
+    {
+      "epoch": 0.041459219697655474,
+      "grad_norm": 0.9765615463256836,
+      "learning_rate": 0.0001993553881635498,
+      "loss": 4.1778,
+      "step": 458
+    },
+    {
+      "epoch": 0.041549742011405814,
+      "grad_norm": 0.993130624294281,
+      "learning_rate": 0.00019935178585156347,
+      "loss": 4.1452,
+      "step": 459
+    },
+    {
+      "epoch": 0.04164026432515615,
+      "grad_norm": 0.9554489850997925,
+      "learning_rate": 0.00019934817353485501,
+      "loss": 4.2302,
+      "step": 460
+    },
+    {
+      "epoch": 0.04173078663890649,
+      "grad_norm": 0.972508430480957,
+      "learning_rate": 0.00019934455121378824,
+      "loss": 4.211,
+      "step": 461
+    },
+    {
+      "epoch": 0.04182130895265683,
+      "grad_norm": 0.9439429640769958,
+      "learning_rate": 0.00019934091888872786,
+      "loss": 4.1795,
+      "step": 462
+    },
+    {
+      "epoch": 0.04191183126640717,
+      "grad_norm": 1.0175976753234863,
+      "learning_rate": 0.00019933727656003963,
+      "loss": 4.1867,
+      "step": 463
+    },
+    {
+      "epoch": 0.04200235358015751,
+      "grad_norm": 0.9793443083763123,
+      "learning_rate": 0.0001993336242280904,
+      "loss": 4.1704,
+      "step": 464
+    },
+    {
+      "epoch": 0.042092875893907845,
+      "grad_norm": 1.0228900909423828,
+      "learning_rate": 0.00019932996189324796,
+      "loss": 4.1865,
+      "step": 465
+    },
+    {
+      "epoch": 0.042183398207658186,
+      "grad_norm": 0.9703587889671326,
+      "learning_rate": 0.00019932628955588103,
+      "loss": 4.1667,
+      "step": 466
+    },
+    {
+      "epoch": 0.04227392052140853,
+      "grad_norm": 1.0352518558502197,
+      "learning_rate": 0.00019932260721635946,
+      "loss": 4.1628,
+      "step": 467
+    },
+    {
+      "epoch": 0.04236444283515887,
+      "grad_norm": 0.9684962034225464,
+      "learning_rate": 0.0001993189148750541,
+      "loss": 4.1845,
+      "step": 468
+    },
+    {
+      "epoch": 0.04245496514890921,
+      "grad_norm": 1.0994378328323364,
+      "learning_rate": 0.00019931521253233669,
+      "loss": 4.2044,
+      "step": 469
+    },
+    {
+      "epoch": 0.04254548746265954,
+      "grad_norm": 0.972074568271637,
+      "learning_rate": 0.00019931150018858012,
+      "loss": 4.1366,
+      "step": 470
+    },
+    {
+      "epoch": 0.04263600977640988,
+      "grad_norm": 1.0492547750473022,
+      "learning_rate": 0.0001993077778441582,
+      "loss": 4.1926,
+      "step": 471
+    },
+    {
+      "epoch": 0.042726532090160224,
+      "grad_norm": 0.9738761186599731,
+      "learning_rate": 0.00019930404549944574,
+      "loss": 4.1006,
+      "step": 472
+    },
+    {
+      "epoch": 0.042817054403910565,
+      "grad_norm": 1.095558524131775,
+      "learning_rate": 0.00019930030315481862,
+      "loss": 4.1846,
+      "step": 473
+    },
+    {
+      "epoch": 0.042907576717660906,
+      "grad_norm": 1.0250678062438965,
+      "learning_rate": 0.0001992965508106537,
+      "loss": 4.1583,
+      "step": 474
+    },
+    {
+      "epoch": 0.04299809903141124,
+      "grad_norm": 1.0323116779327393,
+      "learning_rate": 0.00019929278846732884,
+      "loss": 4.1364,
+      "step": 475
+    },
+    {
+      "epoch": 0.04308862134516158,
+      "grad_norm": 0.9422488808631897,
+      "learning_rate": 0.00019928901612522288,
+      "loss": 4.1598,
+      "step": 476
+    },
+    {
+      "epoch": 0.04317914365891192,
+      "grad_norm": 1.027329683303833,
+      "learning_rate": 0.00019928523378471573,
+      "loss": 4.1284,
+      "step": 477
+    },
+    {
+      "epoch": 0.04326966597266226,
+      "grad_norm": 1.0527349710464478,
+      "learning_rate": 0.00019928144144618824,
+      "loss": 4.1378,
+      "step": 478
+    },
+    {
+      "epoch": 0.0433601882864126,
+      "grad_norm": 1.0269211530685425,
+      "learning_rate": 0.00019927763911002232,
+      "loss": 4.1888,
+      "step": 479
+    },
+    {
+      "epoch": 0.04345071060016294,
+      "grad_norm": 1.0055499076843262,
+      "learning_rate": 0.00019927382677660088,
+      "loss": 4.1982,
+      "step": 480
+    },
+    {
+      "epoch": 0.04354123291391328,
+      "grad_norm": 1.147194504737854,
+      "learning_rate": 0.00019927000444630775,
+      "loss": 4.1557,
+      "step": 481
+    },
+    {
+      "epoch": 0.04363175522766362,
+      "grad_norm": 1.041730284690857,
+      "learning_rate": 0.0001992661721195279,
+      "loss": 4.1445,
+      "step": 482
+    },
+    {
+      "epoch": 0.04372227754141396,
+      "grad_norm": 1.1168657541275024,
+      "learning_rate": 0.00019926232979664728,
+      "loss": 4.1335,
+      "step": 483
+    },
+    {
+      "epoch": 0.0438127998551643,
+      "grad_norm": 0.983162522315979,
+      "learning_rate": 0.00019925847747805274,
+      "loss": 4.2185,
+      "step": 484
+    },
+    {
+      "epoch": 0.04390332216891464,
+      "grad_norm": 0.9528363347053528,
+      "learning_rate": 0.00019925461516413223,
+      "loss": 4.1842,
+      "step": 485
+    },
+    {
+      "epoch": 0.043993844482664975,
+      "grad_norm": 1.0722746849060059,
+      "learning_rate": 0.00019925074285527468,
+      "loss": 4.1321,
+      "step": 486
+    },
+    {
+      "epoch": 0.044084366796415315,
+      "grad_norm": 1.1250028610229492,
+      "learning_rate": 0.00019924686055187003,
+      "loss": 4.075,
+      "step": 487
+    },
+    {
+      "epoch": 0.044174889110165656,
+      "grad_norm": 1.0873173475265503,
+      "learning_rate": 0.00019924296825430925,
+      "loss": 4.1771,
+      "step": 488
+    },
+    {
+      "epoch": 0.044265411423916,
+      "grad_norm": 1.05666983127594,
+      "learning_rate": 0.0001992390659629843,
+      "loss": 4.1521,
+      "step": 489
+    },
+    {
+      "epoch": 0.04435593373766634,
+      "grad_norm": 0.9849108457565308,
+      "learning_rate": 0.0001992351536782881,
+      "loss": 4.1375,
+      "step": 490
+    },
+    {
+      "epoch": 0.04444645605141667,
+      "grad_norm": 1.069494605064392,
+      "learning_rate": 0.00019923123140061467,
+      "loss": 4.1763,
+      "step": 491
+    },
+    {
+      "epoch": 0.04453697836516701,
+      "grad_norm": 0.9202231168746948,
+      "learning_rate": 0.00019922729913035895,
+      "loss": 4.1439,
+      "step": 492
+    },
+    {
+      "epoch": 0.04462750067891735,
+      "grad_norm": 0.9625434279441833,
+      "learning_rate": 0.0001992233568679169,
+      "loss": 4.1377,
+      "step": 493
+    },
+    {
+      "epoch": 0.044718022992667694,
+      "grad_norm": 0.9912998676300049,
+      "learning_rate": 0.0001992194046136855,
+      "loss": 4.1928,
+      "step": 494
+    },
+    {
+      "epoch": 0.044808545306418035,
+      "grad_norm": 0.9711489081382751,
+      "learning_rate": 0.00019921544236806282,
+      "loss": 4.1072,
+      "step": 495
+    },
+    {
+      "epoch": 0.04489906762016837,
+      "grad_norm": 0.9541346430778503,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 4.1439,
+      "step": 496
+    },
+    {
+      "epoch": 0.04498958993391871,
+      "grad_norm": 0.9553717374801636,
+      "learning_rate": 0.00019920748790424044,
+      "loss": 4.1336,
+      "step": 497
+    },
+    {
+      "epoch": 0.04508011224766905,
+      "grad_norm": 0.9162443280220032,
+      "learning_rate": 0.0001992034956868418,
+      "loss": 4.1435,
+      "step": 498
+    },
+    {
+      "epoch": 0.04517063456141939,
+      "grad_norm": 1.0411958694458008,
+      "learning_rate": 0.0001991994934796538,
+      "loss": 4.1484,
+      "step": 499
+    },
+    {
+      "epoch": 0.04526115687516973,
+      "grad_norm": 0.9627156257629395,
+      "learning_rate": 0.00019919548128307954,
+      "loss": 4.1246,
+      "step": 500
+    },
+    {
+      "epoch": 0.045351679188920066,
+      "grad_norm": 0.952808678150177,
+      "learning_rate": 0.00019919145909752303,
+      "loss": 4.0516,
+      "step": 501
+    },
+    {
+      "epoch": 0.04544220150267041,
+      "grad_norm": 0.9976837635040283,
+      "learning_rate": 0.00019918742692338933,
+      "loss": 4.093,
+      "step": 502
+    },
+    {
+      "epoch": 0.04553272381642075,
+      "grad_norm": 1.005296230316162,
+      "learning_rate": 0.0001991833847610844,
+      "loss": 4.1368,
+      "step": 503
+    },
+    {
+      "epoch": 0.04562324613017109,
+      "grad_norm": 0.9490670561790466,
+      "learning_rate": 0.0001991793326110154,
+      "loss": 4.1738,
+      "step": 504
+    },
+    {
+      "epoch": 0.04571376844392143,
+      "grad_norm": 1.0245599746704102,
+      "learning_rate": 0.00019917527047359028,
+      "loss": 4.1499,
+      "step": 505
+    },
+    {
+      "epoch": 0.04580429075767176,
+      "grad_norm": 1.0409306287765503,
+      "learning_rate": 0.00019917119834921815,
+      "loss": 4.1139,
+      "step": 506
+    },
+    {
+      "epoch": 0.045894813071422104,
+      "grad_norm": 1.1501810550689697,
+      "learning_rate": 0.00019916711623830903,
+      "loss": 4.1677,
+      "step": 507
+    },
+    {
+      "epoch": 0.045985335385172445,
+      "grad_norm": 0.9567729234695435,
+      "learning_rate": 0.00019916302414127408,
+      "loss": 4.1223,
+      "step": 508
+    },
+    {
+      "epoch": 0.046075857698922786,
+      "grad_norm": 0.9898899793624878,
+      "learning_rate": 0.00019915892205852527,
+      "loss": 4.085,
+      "step": 509
+    },
+    {
+      "epoch": 0.046166380012673126,
+      "grad_norm": 0.962177038192749,
+      "learning_rate": 0.00019915480999047573,
+      "loss": 4.0921,
+      "step": 510
+    },
+    {
+      "epoch": 0.04625690232642346,
+      "grad_norm": 1.188584566116333,
+      "learning_rate": 0.00019915068793753952,
+      "loss": 4.0904,
+      "step": 511
+    },
+    {
+      "epoch": 0.0463474246401738,
+      "grad_norm": 1.13594651222229,
+      "learning_rate": 0.00019914655590013176,
+      "loss": 4.1522,
+      "step": 512
+    },
+    {
+      "epoch": 0.04643794695392414,
+      "grad_norm": 1.0383785963058472,
+      "learning_rate": 0.00019914241387866856,
+      "loss": 4.1267,
+      "step": 513
+    },
+    {
+      "epoch": 0.04652846926767448,
+      "grad_norm": 1.0647755861282349,
+      "learning_rate": 0.00019913826187356696,
+      "loss": 4.1404,
+      "step": 514
+    },
+    {
+      "epoch": 0.04661899158142482,
+      "grad_norm": 0.9425531029701233,
+      "learning_rate": 0.00019913409988524513,
+      "loss": 4.1329,
+      "step": 515
+    },
+    {
+      "epoch": 0.04670951389517516,
+      "grad_norm": 1.001288890838623,
+      "learning_rate": 0.00019912992791412212,
+      "loss": 4.116,
+      "step": 516
+    },
+    {
+      "epoch": 0.0468000362089255,
+      "grad_norm": 1.1774042844772339,
+      "learning_rate": 0.0001991257459606181,
+      "loss": 4.1825,
+      "step": 517
+    },
+    {
+      "epoch": 0.04689055852267584,
+      "grad_norm": 1.0480451583862305,
+      "learning_rate": 0.00019912155402515417,
+      "loss": 4.0484,
+      "step": 518
+    },
+    {
+      "epoch": 0.04698108083642618,
+      "grad_norm": 1.034772515296936,
+      "learning_rate": 0.0001991173521081525,
+      "loss": 4.0266,
+      "step": 519
+    },
+    {
+      "epoch": 0.04707160315017652,
+      "grad_norm": 1.3288160562515259,
+      "learning_rate": 0.00019911314021003613,
+      "loss": 4.0988,
+      "step": 520
+    },
+    {
+      "epoch": 0.04716212546392686,
+      "grad_norm": 1.2328402996063232,
+      "learning_rate": 0.00019910891833122926,
+      "loss": 4.1125,
+      "step": 521
+    },
+    {
+      "epoch": 0.047252647777677195,
+      "grad_norm": 1.0993627309799194,
+      "learning_rate": 0.00019910468647215706,
+      "loss": 4.1599,
+      "step": 522
+    },
+    {
+      "epoch": 0.047343170091427536,
+      "grad_norm": 1.1606861352920532,
+      "learning_rate": 0.00019910044463324563,
+      "loss": 4.1601,
+      "step": 523
+    },
+    {
+      "epoch": 0.04743369240517788,
+      "grad_norm": 0.9817080497741699,
+      "learning_rate": 0.0001990961928149221,
+      "loss": 4.0921,
+      "step": 524
+    },
+    {
+      "epoch": 0.04752421471892822,
+      "grad_norm": 1.0352342128753662,
+      "learning_rate": 0.0001990919310176147,
+      "loss": 4.0273,
+      "step": 525
+    },
+    {
+      "epoch": 0.04761473703267856,
+      "grad_norm": 1.0226386785507202,
+      "learning_rate": 0.00019908765924175258,
+      "loss": 4.1009,
+      "step": 526
+    },
+    {
+      "epoch": 0.04770525934642889,
+      "grad_norm": 0.9978080987930298,
+      "learning_rate": 0.00019908337748776584,
+      "loss": 4.1242,
+      "step": 527
+    },
+    {
+      "epoch": 0.04779578166017923,
+      "grad_norm": 1.0166219472885132,
+      "learning_rate": 0.00019907908575608573,
+      "loss": 4.085,
+      "step": 528
+    },
+    {
+      "epoch": 0.047886303973929574,
+      "grad_norm": 1.0522500276565552,
+      "learning_rate": 0.00019907478404714436,
+      "loss": 4.1089,
+      "step": 529
+    },
+    {
+      "epoch": 0.047976826287679915,
+      "grad_norm": 0.9937741756439209,
+      "learning_rate": 0.00019907047236137498,
+      "loss": 4.1281,
+      "step": 530
+    },
+    {
+      "epoch": 0.048067348601430256,
+      "grad_norm": 0.9463094472885132,
+      "learning_rate": 0.00019906615069921173,
+      "loss": 4.0034,
+      "step": 531
+    },
+    {
+      "epoch": 0.04815787091518059,
+      "grad_norm": 1.009852647781372,
+      "learning_rate": 0.00019906181906108984,
+      "loss": 4.1578,
+      "step": 532
+    },
+    {
+      "epoch": 0.04824839322893093,
+      "grad_norm": 1.0535956621170044,
+      "learning_rate": 0.00019905747744744545,
+      "loss": 4.1425,
+      "step": 533
+    },
+    {
+      "epoch": 0.04833891554268127,
+      "grad_norm": 1.0467532873153687,
+      "learning_rate": 0.0001990531258587158,
+      "loss": 4.1166,
+      "step": 534
+    },
+    {
+      "epoch": 0.04842943785643161,
+      "grad_norm": 0.9849286079406738,
+      "learning_rate": 0.0001990487642953391,
+      "loss": 4.0509,
+      "step": 535
+    },
+    {
+      "epoch": 0.04851996017018195,
+      "grad_norm": 1.15632963180542,
+      "learning_rate": 0.00019904439275775452,
+      "loss": 4.1081,
+      "step": 536
+    },
+    {
+      "epoch": 0.04861048248393229,
+      "grad_norm": 1.0920475721359253,
+      "learning_rate": 0.0001990400112464023,
+      "loss": 4.0632,
+      "step": 537
+    },
+    {
+      "epoch": 0.04870100479768263,
+      "grad_norm": 1.0463149547576904,
+      "learning_rate": 0.00019903561976172368,
+      "loss": 4.0839,
+      "step": 538
+    },
+    {
+      "epoch": 0.04879152711143297,
+      "grad_norm": 1.1025164127349854,
+      "learning_rate": 0.00019903121830416084,
+      "loss": 4.0504,
+      "step": 539
+    },
+    {
+      "epoch": 0.04888204942518331,
+      "grad_norm": 1.0772355794906616,
+      "learning_rate": 0.00019902680687415705,
+      "loss": 4.125,
+      "step": 540
+    },
+    {
+      "epoch": 0.04897257173893365,
+      "grad_norm": 1.1654671430587769,
+      "learning_rate": 0.0001990223854721565,
+      "loss": 4.0562,
+      "step": 541
+    },
+    {
+      "epoch": 0.049063094052683984,
+      "grad_norm": 1.0062533617019653,
+      "learning_rate": 0.00019901795409860444,
+      "loss": 3.9895,
+      "step": 542
+    },
+    {
+      "epoch": 0.049153616366434325,
+      "grad_norm": 1.0412124395370483,
+      "learning_rate": 0.0001990135127539471,
+      "loss": 4.0952,
+      "step": 543
+    },
+    {
+      "epoch": 0.049244138680184665,
+      "grad_norm": 0.9877052903175354,
+      "learning_rate": 0.0001990090614386318,
+      "loss": 4.0693,
+      "step": 544
+    },
+    {
+      "epoch": 0.049334660993935006,
+      "grad_norm": 0.978701114654541,
+      "learning_rate": 0.00019900460015310665,
+      "loss": 3.9831,
+      "step": 545
+    },
+    {
+      "epoch": 0.04942518330768535,
+      "grad_norm": 1.1422040462493896,
+      "learning_rate": 0.00019900012889782098,
+      "loss": 4.1191,
+      "step": 546
+    },
+    {
+      "epoch": 0.04951570562143568,
+      "grad_norm": 1.0867984294891357,
+      "learning_rate": 0.00019899564767322507,
+      "loss": 4.1175,
+      "step": 547
+    },
+    {
+      "epoch": 0.04960622793518602,
+      "grad_norm": 1.0576975345611572,
+      "learning_rate": 0.00019899115647977015,
+      "loss": 4.0439,
+      "step": 548
+    },
+    {
+      "epoch": 0.04969675024893636,
+      "grad_norm": 1.073725700378418,
+      "learning_rate": 0.00019898665531790845,
+      "loss": 4.1013,
+      "step": 549
+    },
+    {
+      "epoch": 0.0497872725626867,
+      "grad_norm": 0.9969006180763245,
+      "learning_rate": 0.0001989821441880933,
+      "loss": 4.0446,
+      "step": 550
+    },
+    {
+      "epoch": 0.049877794876437044,
+      "grad_norm": 1.0461211204528809,
+      "learning_rate": 0.0001989776230907789,
+      "loss": 4.0519,
+      "step": 551
+    },
+    {
+      "epoch": 0.04996831719018738,
+      "grad_norm": 0.9178494215011597,
+      "learning_rate": 0.0001989730920264206,
+      "loss": 3.9729,
+      "step": 552
+    },
+    {
+      "epoch": 0.05005883950393772,
+      "grad_norm": 1.0089956521987915,
+      "learning_rate": 0.0001989685509954746,
+      "loss": 4.0438,
+      "step": 553
+    },
+    {
+      "epoch": 0.05014936181768806,
+      "grad_norm": 0.9153216481208801,
+      "learning_rate": 0.00019896399999839827,
+      "loss": 4.0706,
+      "step": 554
+    },
+    {
+      "epoch": 0.0502398841314384,
+      "grad_norm": 1.0593777894973755,
+      "learning_rate": 0.0001989594390356498,
+      "loss": 4.0807,
+      "step": 555
+    },
+    {
+      "epoch": 0.05033040644518874,
+      "grad_norm": 1.0582592487335205,
+      "learning_rate": 0.00019895486810768856,
+      "loss": 3.9998,
+      "step": 556
+    },
+    {
+      "epoch": 0.050420928758939075,
+      "grad_norm": 0.9722828269004822,
+      "learning_rate": 0.00019895028721497482,
+      "loss": 4.0521,
+      "step": 557
+    },
+    {
+      "epoch": 0.050511451072689416,
+      "grad_norm": 1.1836525201797485,
+      "learning_rate": 0.00019894569635796984,
+      "loss": 4.0514,
+      "step": 558
+    },
+    {
+      "epoch": 0.05060197338643976,
+      "grad_norm": 1.1598135232925415,
+      "learning_rate": 0.00019894109553713596,
+      "loss": 4.0248,
+      "step": 559
+    },
+    {
+      "epoch": 0.0506924957001901,
+      "grad_norm": 0.9481819868087769,
+      "learning_rate": 0.00019893648475293648,
+      "loss": 4.0499,
+      "step": 560
+    },
+    {
+      "epoch": 0.05078301801394044,
+      "grad_norm": 1.014451503753662,
+      "learning_rate": 0.00019893186400583566,
+      "loss": 4.0659,
+      "step": 561
+    },
+    {
+      "epoch": 0.05087354032769078,
+      "grad_norm": 0.9883650541305542,
+      "learning_rate": 0.00019892723329629887,
+      "loss": 4.0131,
+      "step": 562
+    },
+    {
+      "epoch": 0.05096406264144111,
+      "grad_norm": 0.9952863454818726,
+      "learning_rate": 0.00019892259262479237,
+      "loss": 4.0082,
+      "step": 563
+    },
+    {
+      "epoch": 0.051054584955191454,
+      "grad_norm": 1.0756782293319702,
+      "learning_rate": 0.0001989179419917835,
+      "loss": 4.0799,
+      "step": 564
+    },
+    {
+      "epoch": 0.051145107268941795,
+      "grad_norm": 1.066792368888855,
+      "learning_rate": 0.00019891328139774056,
+      "loss": 4.0465,
+      "step": 565
+    },
+    {
+      "epoch": 0.051235629582692135,
+      "grad_norm": 1.0339174270629883,
+      "learning_rate": 0.00019890861084313294,
+      "loss": 4.0085,
+      "step": 566
+    },
+    {
+      "epoch": 0.051326151896442476,
+      "grad_norm": 0.9658034443855286,
+      "learning_rate": 0.00019890393032843086,
+      "loss": 4.025,
+      "step": 567
+    },
+    {
+      "epoch": 0.05141667421019281,
+      "grad_norm": 1.0180926322937012,
+      "learning_rate": 0.00019889923985410576,
+      "loss": 4.0619,
+      "step": 568
+    },
+    {
+      "epoch": 0.05150719652394315,
+      "grad_norm": 0.9875718951225281,
+      "learning_rate": 0.00019889453942062985,
+      "loss": 4.0292,
+      "step": 569
+    },
+    {
+      "epoch": 0.05159771883769349,
+      "grad_norm": 0.9706630706787109,
+      "learning_rate": 0.00019888982902847656,
+      "loss": 4.0822,
+      "step": 570
+    },
+    {
+      "epoch": 0.05168824115144383,
+      "grad_norm": 0.981421172618866,
+      "learning_rate": 0.0001988851086781202,
+      "loss": 4.0864,
+      "step": 571
+    },
+    {
+      "epoch": 0.05177876346519417,
+      "grad_norm": 1.0634344816207886,
+      "learning_rate": 0.00019888037837003612,
+      "loss": 4.0156,
+      "step": 572
+    },
+    {
+      "epoch": 0.05186928577894451,
+      "grad_norm": 1.0097788572311401,
+      "learning_rate": 0.0001988756381047006,
+      "loss": 4.0723,
+      "step": 573
+    },
+    {
+      "epoch": 0.05195980809269485,
+      "grad_norm": 0.9124920964241028,
+      "learning_rate": 0.00019887088788259102,
+      "loss": 3.9745,
+      "step": 574
+    },
+    {
+      "epoch": 0.05205033040644519,
+      "grad_norm": 1.0202012062072754,
+      "learning_rate": 0.00019886612770418578,
+      "loss": 4.0043,
+      "step": 575
+    },
+    {
+      "epoch": 0.05214085272019553,
+      "grad_norm": 1.0590636730194092,
+      "learning_rate": 0.0001988613575699642,
+      "loss": 3.9986,
+      "step": 576
+    },
+    {
+      "epoch": 0.05223137503394587,
+      "grad_norm": 0.9344479441642761,
+      "learning_rate": 0.00019885657748040653,
+      "loss": 4.0404,
+      "step": 577
+    },
+    {
+      "epoch": 0.052321897347696204,
+      "grad_norm": 0.9335837364196777,
+      "learning_rate": 0.00019885178743599428,
+      "loss": 4.0202,
+      "step": 578
+    },
+    {
+      "epoch": 0.052412419661446545,
+      "grad_norm": 0.8614705204963684,
+      "learning_rate": 0.00019884698743720974,
+      "loss": 4.0153,
+      "step": 579
+    },
+    {
+      "epoch": 0.052502941975196886,
+      "grad_norm": 0.9567529559135437,
+      "learning_rate": 0.00019884217748453623,
+      "loss": 4.0822,
+      "step": 580
+    },
+    {
+      "epoch": 0.05259346428894723,
+      "grad_norm": 0.9799798130989075,
+      "learning_rate": 0.0001988373575784582,
+      "loss": 4.0124,
+      "step": 581
+    },
+    {
+      "epoch": 0.05268398660269757,
+      "grad_norm": 1.0360461473464966,
+      "learning_rate": 0.00019883252771946095,
+      "loss": 4.0128,
+      "step": 582
+    },
+    {
+      "epoch": 0.0527745089164479,
+      "grad_norm": 1.0541273355484009,
+      "learning_rate": 0.00019882768790803086,
+      "loss": 3.9391,
+      "step": 583
+    },
+    {
+      "epoch": 0.05286503123019824,
+      "grad_norm": 1.0942133665084839,
+      "learning_rate": 0.0001988228381446553,
+      "loss": 3.9941,
+      "step": 584
+    },
+    {
+      "epoch": 0.05295555354394858,
+      "grad_norm": 0.9243170619010925,
+      "learning_rate": 0.00019881797842982263,
+      "loss": 3.9886,
+      "step": 585
+    },
+    {
+      "epoch": 0.053046075857698924,
+      "grad_norm": 0.9738331437110901,
+      "learning_rate": 0.00019881310876402223,
+      "loss": 3.9795,
+      "step": 586
+    },
+    {
+      "epoch": 0.053136598171449265,
+      "grad_norm": 0.932464599609375,
+      "learning_rate": 0.00019880822914774453,
+      "loss": 3.9952,
+      "step": 587
+    },
+    {
+      "epoch": 0.0532271204851996,
+      "grad_norm": 1.0746235847473145,
+      "learning_rate": 0.00019880333958148083,
+      "loss": 3.9263,
+      "step": 588
+    },
+    {
+      "epoch": 0.05331764279894994,
+      "grad_norm": 0.9642815589904785,
+      "learning_rate": 0.00019879844006572353,
+      "loss": 4.0313,
+      "step": 589
+    },
+    {
+      "epoch": 0.05340816511270028,
+      "grad_norm": 1.041780710220337,
+      "learning_rate": 0.00019879353060096603,
+      "loss": 3.9453,
+      "step": 590
+    },
+    {
+      "epoch": 0.05349868742645062,
+      "grad_norm": 0.9562025666236877,
+      "learning_rate": 0.0001987886111877027,
+      "loss": 3.9492,
+      "step": 591
+    },
+    {
+      "epoch": 0.05358920974020096,
+      "grad_norm": 0.9852890372276306,
+      "learning_rate": 0.0001987836818264289,
+      "loss": 3.9696,
+      "step": 592
+    },
+    {
+      "epoch": 0.053679732053951296,
+      "grad_norm": 1.0013340711593628,
+      "learning_rate": 0.00019877874251764105,
+      "loss": 3.9991,
+      "step": 593
+    },
+    {
+      "epoch": 0.053770254367701636,
+      "grad_norm": 1.1419621706008911,
+      "learning_rate": 0.00019877379326183655,
+      "loss": 4.0529,
+      "step": 594
+    },
+    {
+      "epoch": 0.05386077668145198,
+      "grad_norm": 1.0771965980529785,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 4.094,
+      "step": 595
+    },
+    {
+      "epoch": 0.05395129899520232,
+      "grad_norm": 1.0541069507598877,
+      "learning_rate": 0.0001987638649111721,
+      "loss": 3.9807,
+      "step": 596
+    },
+    {
+      "epoch": 0.05404182130895266,
+      "grad_norm": 0.9966290593147278,
+      "learning_rate": 0.00019875888581731194,
+      "loss": 3.9641,
+      "step": 597
+    },
+    {
+      "epoch": 0.054132343622703,
+      "grad_norm": 1.0405066013336182,
+      "learning_rate": 0.0001987538967784347,
+      "loss": 4.0511,
+      "step": 598
+    },
+    {
+      "epoch": 0.054222865936453334,
+      "grad_norm": 0.9865603446960449,
+      "learning_rate": 0.00019874889779504274,
+      "loss": 3.9883,
+      "step": 599
+    },
+    {
+      "epoch": 0.054313388250203674,
+      "grad_norm": 1.097095251083374,
+      "learning_rate": 0.00019874388886763944,
+      "loss": 3.9818,
+      "step": 600
+    },
+    {
+      "epoch": 0.054403910563954015,
+      "grad_norm": 0.9284229278564453,
+      "learning_rate": 0.00019873886999672926,
+      "loss": 4.0203,
+      "step": 601
+    },
+    {
+      "epoch": 0.054494432877704356,
+      "grad_norm": 1.0191222429275513,
+      "learning_rate": 0.00019873384118281756,
+      "loss": 3.9996,
+      "step": 602
+    },
+    {
+      "epoch": 0.0545849551914547,
+      "grad_norm": 1.0141578912734985,
+      "learning_rate": 0.00019872880242641078,
+      "loss": 3.996,
+      "step": 603
+    },
+    {
+      "epoch": 0.05467547750520503,
+      "grad_norm": 1.0445934534072876,
+      "learning_rate": 0.0001987237537280163,
+      "loss": 3.9988,
+      "step": 604
+    },
+    {
+      "epoch": 0.05476599981895537,
+      "grad_norm": 1.0781155824661255,
+      "learning_rate": 0.0001987186950881425,
+      "loss": 4.0494,
+      "step": 605
+    },
+    {
+      "epoch": 0.05485652213270571,
+      "grad_norm": 0.9800688028335571,
+      "learning_rate": 0.0001987136265072988,
+      "loss": 4.0039,
+      "step": 606
+    },
+    {
+      "epoch": 0.05494704444645605,
+      "grad_norm": 0.9078426361083984,
+      "learning_rate": 0.00019870854798599563,
+      "loss": 3.954,
+      "step": 607
+    },
+    {
+      "epoch": 0.055037566760206394,
+      "grad_norm": 0.9343809485435486,
+      "learning_rate": 0.00019870345952474437,
+      "loss": 3.947,
+      "step": 608
+    },
+    {
+      "epoch": 0.05512808907395673,
+      "grad_norm": 1.0472208261489868,
+      "learning_rate": 0.00019869836112405742,
+      "loss": 4.0662,
+      "step": 609
+    },
+    {
+      "epoch": 0.05521861138770707,
+      "grad_norm": 0.9884335994720459,
+      "learning_rate": 0.00019869325278444824,
+      "loss": 3.9495,
+      "step": 610
+    },
+    {
+      "epoch": 0.05530913370145741,
+      "grad_norm": 0.9580947160720825,
+      "learning_rate": 0.00019868813450643117,
+      "loss": 4.0143,
+      "step": 611
+    },
+    {
+      "epoch": 0.05539965601520775,
+      "grad_norm": 1.0103482007980347,
+      "learning_rate": 0.0001986830062905217,
+      "loss": 3.9516,
+      "step": 612
+    },
+    {
+      "epoch": 0.05549017832895809,
+      "grad_norm": 0.9422419667243958,
+      "learning_rate": 0.00019867786813723614,
+      "loss": 4.0485,
+      "step": 613
+    },
+    {
+      "epoch": 0.055580700642708425,
+      "grad_norm": 1.042916178703308,
+      "learning_rate": 0.00019867272004709202,
+      "loss": 3.9367,
+      "step": 614
+    },
+    {
+      "epoch": 0.055671222956458766,
+      "grad_norm": 1.0430114269256592,
+      "learning_rate": 0.00019866756202060762,
+      "loss": 3.9835,
+      "step": 615
+    },
+    {
+      "epoch": 0.05576174527020911,
+      "grad_norm": 0.9182488322257996,
+      "learning_rate": 0.00019866239405830248,
+      "loss": 3.9871,
+      "step": 616
+    },
+    {
+      "epoch": 0.05585226758395945,
+      "grad_norm": 1.0735499858856201,
+      "learning_rate": 0.00019865721616069696,
+      "loss": 4.0137,
+      "step": 617
+    },
+    {
+      "epoch": 0.05594278989770979,
+      "grad_norm": 0.9811666011810303,
+      "learning_rate": 0.00019865202832831244,
+      "loss": 4.0045,
+      "step": 618
+    },
+    {
+      "epoch": 0.05603331221146012,
+      "grad_norm": 1.1695924997329712,
+      "learning_rate": 0.00019864683056167138,
+      "loss": 4.0147,
+      "step": 619
+    },
+    {
+      "epoch": 0.05612383452521046,
+      "grad_norm": 1.052254557609558,
+      "learning_rate": 0.0001986416228612972,
+      "loss": 4.0263,
+      "step": 620
+    },
+    {
+      "epoch": 0.056214356838960804,
+      "grad_norm": 1.2805979251861572,
+      "learning_rate": 0.00019863640522771429,
+      "loss": 3.9631,
+      "step": 621
+    },
+    {
+      "epoch": 0.056304879152711144,
+      "grad_norm": 1.036574363708496,
+      "learning_rate": 0.00019863117766144806,
+      "loss": 4.0074,
+      "step": 622
+    },
+    {
+      "epoch": 0.056395401466461485,
+      "grad_norm": 1.0256383419036865,
+      "learning_rate": 0.00019862594016302492,
+      "loss": 4.0937,
+      "step": 623
+    },
+    {
+      "epoch": 0.05648592378021182,
+      "grad_norm": 0.9368764758110046,
+      "learning_rate": 0.00019862069273297232,
+      "loss": 3.9692,
+      "step": 624
+    },
+    {
+      "epoch": 0.05657644609396216,
+      "grad_norm": 1.1771879196166992,
+      "learning_rate": 0.00019861543537181867,
+      "loss": 4.0685,
+      "step": 625
+    },
+    {
+      "epoch": 0.0566669684077125,
+      "grad_norm": 1.032165288925171,
+      "learning_rate": 0.00019861016808009335,
+      "loss": 3.8764,
+      "step": 626
+    },
+    {
+      "epoch": 0.05675749072146284,
+      "grad_norm": 0.970867395401001,
+      "learning_rate": 0.00019860489085832684,
+      "loss": 3.9917,
+      "step": 627
+    },
+    {
+      "epoch": 0.05684801303521318,
+      "grad_norm": 1.1051703691482544,
+      "learning_rate": 0.0001985996037070505,
+      "loss": 3.9425,
+      "step": 628
+    },
+    {
+      "epoch": 0.056938535348963516,
+      "grad_norm": 0.953762948513031,
+      "learning_rate": 0.00019859430662679678,
+      "loss": 3.912,
+      "step": 629
+    },
+    {
+      "epoch": 0.05702905766271386,
+      "grad_norm": 1.0868688821792603,
+      "learning_rate": 0.00019858899961809905,
+      "loss": 3.9425,
+      "step": 630
+    },
+    {
+      "epoch": 0.0571195799764642,
+      "grad_norm": 1.01759672164917,
+      "learning_rate": 0.0001985836826814918,
+      "loss": 4.0169,
+      "step": 631
+    },
+    {
+      "epoch": 0.05721010229021454,
+      "grad_norm": 0.9279404878616333,
+      "learning_rate": 0.00019857835581751037,
+      "loss": 3.9645,
+      "step": 632
+    },
+    {
+      "epoch": 0.05730062460396488,
+      "grad_norm": 0.9583117961883545,
+      "learning_rate": 0.00019857301902669118,
+      "loss": 4.0205,
+      "step": 633
+    },
+    {
+      "epoch": 0.05739114691771521,
+      "grad_norm": 1.0274845361709595,
+      "learning_rate": 0.00019856767230957173,
+      "loss": 4.0191,
+      "step": 634
+    },
+    {
+      "epoch": 0.057481669231465554,
+      "grad_norm": 0.970229983329773,
+      "learning_rate": 0.00019856231566669035,
+      "loss": 4.0247,
+      "step": 635
+    },
+    {
+      "epoch": 0.057572191545215895,
+      "grad_norm": 0.9355469346046448,
+      "learning_rate": 0.00019855694909858644,
+      "loss": 3.937,
+      "step": 636
+    },
+    {
+      "epoch": 0.057662713858966236,
+      "grad_norm": 1.0804182291030884,
+      "learning_rate": 0.0001985515726058005,
+      "loss": 3.9142,
+      "step": 637
+    },
+    {
+      "epoch": 0.05775323617271658,
+      "grad_norm": 0.9826698303222656,
+      "learning_rate": 0.0001985461861888739,
+      "loss": 3.9708,
+      "step": 638
+    },
+    {
+      "epoch": 0.05784375848646692,
+      "grad_norm": 1.0979808568954468,
+      "learning_rate": 0.00019854078984834903,
+      "loss": 3.9984,
+      "step": 639
+    },
+    {
+      "epoch": 0.05793428080021725,
+      "grad_norm": 1.0105253458023071,
+      "learning_rate": 0.00019853538358476932,
+      "loss": 3.9522,
+      "step": 640
+    },
+    {
+      "epoch": 0.05802480311396759,
+      "grad_norm": 0.9418004155158997,
+      "learning_rate": 0.00019852996739867917,
+      "loss": 3.9531,
+      "step": 641
+    },
+    {
+      "epoch": 0.05811532542771793,
+      "grad_norm": 0.9501124024391174,
+      "learning_rate": 0.000198524541290624,
+      "loss": 4.0334,
+      "step": 642
+    },
+    {
+      "epoch": 0.058205847741468274,
+      "grad_norm": 0.9833979606628418,
+      "learning_rate": 0.00019851910526115022,
+      "loss": 3.9814,
+      "step": 643
+    },
+    {
+      "epoch": 0.058296370055218615,
+      "grad_norm": 0.9670637249946594,
+      "learning_rate": 0.00019851365931080527,
+      "loss": 3.9314,
+      "step": 644
+    },
+    {
+      "epoch": 0.05838689236896895,
+      "grad_norm": 0.9459611773490906,
+      "learning_rate": 0.00019850820344013748,
+      "loss": 3.901,
+      "step": 645
+    },
+    {
+      "epoch": 0.05847741468271929,
+      "grad_norm": 1.0001896619796753,
+      "learning_rate": 0.00019850273764969632,
+      "loss": 3.9213,
+      "step": 646
+    },
+    {
+      "epoch": 0.05856793699646963,
+      "grad_norm": 0.947214663028717,
+      "learning_rate": 0.00019849726194003214,
+      "loss": 3.9845,
+      "step": 647
+    },
+    {
+      "epoch": 0.05865845931021997,
+      "grad_norm": 0.9455295205116272,
+      "learning_rate": 0.00019849177631169643,
+      "loss": 3.9425,
+      "step": 648
+    },
+    {
+      "epoch": 0.05874898162397031,
+      "grad_norm": 0.9829345345497131,
+      "learning_rate": 0.00019848628076524148,
+      "loss": 3.847,
+      "step": 649
+    },
+    {
+      "epoch": 0.058839503937720646,
+      "grad_norm": 0.9669824838638306,
+      "learning_rate": 0.00019848077530122083,
+      "loss": 3.9601,
+      "step": 650
+    },
+    {
+      "epoch": 0.058930026251470986,
+      "grad_norm": 0.9459212422370911,
+      "learning_rate": 0.00019847525992018877,
+      "loss": 3.9733,
+      "step": 651
+    },
+    {
+      "epoch": 0.05902054856522133,
+      "grad_norm": 1.0247117280960083,
+      "learning_rate": 0.0001984697346227007,
+      "loss": 3.9652,
+      "step": 652
+    },
+    {
+      "epoch": 0.05911107087897167,
+      "grad_norm": 0.9490055441856384,
+      "learning_rate": 0.00019846419940931304,
+      "loss": 4.0009,
+      "step": 653
+    },
+    {
+      "epoch": 0.05920159319272201,
+      "grad_norm": 1.0625797510147095,
+      "learning_rate": 0.00019845865428058326,
+      "loss": 3.9344,
+      "step": 654
+    },
+    {
+      "epoch": 0.05929211550647234,
+      "grad_norm": 0.9928398132324219,
+      "learning_rate": 0.00019845309923706962,
+      "loss": 3.9198,
+      "step": 655
+    },
+    {
+      "epoch": 0.05938263782022268,
+      "grad_norm": 0.9566262364387512,
+      "learning_rate": 0.00019844753427933164,
+      "loss": 3.9347,
+      "step": 656
+    },
+    {
+      "epoch": 0.059473160133973024,
+      "grad_norm": 0.9817463755607605,
+      "learning_rate": 0.0001984419594079296,
+      "loss": 3.9768,
+      "step": 657
+    },
+    {
+      "epoch": 0.059563682447723365,
+      "grad_norm": 0.9232942461967468,
+      "learning_rate": 0.00019843637462342497,
+      "loss": 3.9482,
+      "step": 658
+    },
+    {
+      "epoch": 0.059654204761473706,
+      "grad_norm": 1.021917462348938,
+      "learning_rate": 0.00019843077992638008,
+      "loss": 4.0121,
+      "step": 659
+    },
+    {
+      "epoch": 0.05974472707522404,
+      "grad_norm": 0.8719213604927063,
+      "learning_rate": 0.00019842517531735838,
+      "loss": 3.932,
+      "step": 660
+    },
+    {
+      "epoch": 0.05983524938897438,
+      "grad_norm": 0.9565935134887695,
+      "learning_rate": 0.0001984195607969242,
+      "loss": 3.8937,
+      "step": 661
+    },
+    {
+      "epoch": 0.05992577170272472,
+      "grad_norm": 0.9811224341392517,
+      "learning_rate": 0.00019841393636564294,
+      "loss": 3.9254,
+      "step": 662
+    },
+    {
+      "epoch": 0.06001629401647506,
+      "grad_norm": 0.9084984064102173,
+      "learning_rate": 0.000198408302024081,
+      "loss": 3.9867,
+      "step": 663
+    },
+    {
+      "epoch": 0.0601068163302254,
+      "grad_norm": 1.0378906726837158,
+      "learning_rate": 0.0001984026577728057,
+      "loss": 3.8996,
+      "step": 664
+    },
+    {
+      "epoch": 0.06019733864397574,
+      "grad_norm": 1.0800020694732666,
+      "learning_rate": 0.0001983970036123855,
+      "loss": 3.9582,
+      "step": 665
+    },
+    {
+      "epoch": 0.06028786095772608,
+      "grad_norm": 1.086745023727417,
+      "learning_rate": 0.0001983913395433897,
+      "loss": 3.8442,
+      "step": 666
+    },
+    {
+      "epoch": 0.06037838327147642,
+      "grad_norm": 1.1331897974014282,
+      "learning_rate": 0.0001983856655663887,
+      "loss": 3.9351,
+      "step": 667
+    },
+    {
+      "epoch": 0.06046890558522676,
+      "grad_norm": 1.074208378791809,
+      "learning_rate": 0.0001983799816819539,
+      "loss": 3.9473,
+      "step": 668
+    },
+    {
+      "epoch": 0.0605594278989771,
+      "grad_norm": 0.9405530691146851,
+      "learning_rate": 0.0001983742878906576,
+      "loss": 3.9648,
+      "step": 669
+    },
+    {
+      "epoch": 0.060649950212727434,
+      "grad_norm": 0.9946943521499634,
+      "learning_rate": 0.00019836858419307324,
+      "loss": 3.9814,
+      "step": 670
+    },
+    {
+      "epoch": 0.060740472526477775,
+      "grad_norm": 0.952600359916687,
+      "learning_rate": 0.0001983628705897751,
+      "loss": 3.986,
+      "step": 671
+    },
+    {
+      "epoch": 0.060830994840228116,
+      "grad_norm": 0.9927718639373779,
+      "learning_rate": 0.00019835714708133862,
+      "loss": 3.9474,
+      "step": 672
+    },
+    {
+      "epoch": 0.060921517153978456,
+      "grad_norm": 0.9685267806053162,
+      "learning_rate": 0.00019835141366834007,
+      "loss": 3.8756,
+      "step": 673
+    },
+    {
+      "epoch": 0.0610120394677288,
+      "grad_norm": 1.0644733905792236,
+      "learning_rate": 0.0001983456703513569,
+      "loss": 4.029,
+      "step": 674
+    },
+    {
+      "epoch": 0.06110256178147914,
+      "grad_norm": 0.8898302316665649,
+      "learning_rate": 0.0001983399171309674,
+      "loss": 3.9162,
+      "step": 675
+    },
+    {
+      "epoch": 0.06119308409522947,
+      "grad_norm": 1.0014737844467163,
+      "learning_rate": 0.00019833415400775093,
+      "loss": 3.8812,
+      "step": 676
+    },
+    {
+      "epoch": 0.06128360640897981,
+      "grad_norm": 0.9743993878364563,
+      "learning_rate": 0.00019832838098228785,
+      "loss": 3.9309,
+      "step": 677
+    },
+    {
+      "epoch": 0.061374128722730154,
+      "grad_norm": 0.9850934743881226,
+      "learning_rate": 0.0001983225980551595,
+      "loss": 4.0093,
+      "step": 678
+    },
+    {
+      "epoch": 0.061464651036480494,
+      "grad_norm": 1.0402182340621948,
+      "learning_rate": 0.00019831680522694822,
+      "loss": 4.009,
+      "step": 679
+    },
+    {
+      "epoch": 0.061555173350230835,
+      "grad_norm": 1.0211234092712402,
+      "learning_rate": 0.00019831100249823733,
+      "loss": 4.0221,
+      "step": 680
+    },
+    {
+      "epoch": 0.06164569566398117,
+      "grad_norm": 1.0483235120773315,
+      "learning_rate": 0.00019830518986961118,
+      "loss": 3.9018,
+      "step": 681
+    },
+    {
+      "epoch": 0.06173621797773151,
+      "grad_norm": 0.9993516802787781,
+      "learning_rate": 0.0001982993673416551,
+      "loss": 3.9225,
+      "step": 682
+    },
+    {
+      "epoch": 0.06182674029148185,
+      "grad_norm": 0.9728137254714966,
+      "learning_rate": 0.00019829353491495545,
+      "loss": 3.9108,
+      "step": 683
+    },
+    {
+      "epoch": 0.06191726260523219,
+      "grad_norm": 1.0375216007232666,
+      "learning_rate": 0.00019828769259009948,
+      "loss": 3.9321,
+      "step": 684
+    },
+    {
+      "epoch": 0.06200778491898253,
+      "grad_norm": 1.0920323133468628,
+      "learning_rate": 0.00019828184036767556,
+      "loss": 3.9325,
+      "step": 685
+    },
+    {
+      "epoch": 0.062098307232732866,
+      "grad_norm": 0.940028965473175,
+      "learning_rate": 0.00019827597824827303,
+      "loss": 3.8847,
+      "step": 686
+    },
+    {
+      "epoch": 0.06218882954648321,
+      "grad_norm": 1.0867862701416016,
+      "learning_rate": 0.00019827010623248216,
+      "loss": 3.8561,
+      "step": 687
+    },
+    {
+      "epoch": 0.06227935186023355,
+      "grad_norm": 1.0442867279052734,
+      "learning_rate": 0.0001982642243208943,
+      "loss": 3.9314,
+      "step": 688
+    },
+    {
+      "epoch": 0.06236987417398389,
+      "grad_norm": 0.9575579762458801,
+      "learning_rate": 0.00019825833251410174,
+      "loss": 3.9118,
+      "step": 689
+    },
+    {
+      "epoch": 0.06246039648773423,
+      "grad_norm": 0.9611737132072449,
+      "learning_rate": 0.00019825243081269774,
+      "loss": 3.9184,
+      "step": 690
+    },
+    {
+      "epoch": 0.06255091880148457,
+      "grad_norm": 1.049651026725769,
+      "learning_rate": 0.0001982465192172767,
+      "loss": 3.949,
+      "step": 691
+    },
+    {
+      "epoch": 0.06264144111523491,
+      "grad_norm": 1.005272388458252,
+      "learning_rate": 0.0001982405977284338,
+      "loss": 3.8511,
+      "step": 692
+    },
+    {
+      "epoch": 0.06273196342898525,
+      "grad_norm": 1.052721381187439,
+      "learning_rate": 0.00019823466634676543,
+      "loss": 3.8426,
+      "step": 693
+    },
+    {
+      "epoch": 0.06282248574273558,
+      "grad_norm": 0.9251368641853333,
+      "learning_rate": 0.0001982287250728689,
+      "loss": 3.9242,
+      "step": 694
+    },
+    {
+      "epoch": 0.06291300805648592,
+      "grad_norm": 1.048956274986267,
+      "learning_rate": 0.00019822277390734239,
+      "loss": 3.8853,
+      "step": 695
+    },
+    {
+      "epoch": 0.06300353037023626,
+      "grad_norm": 1.1769357919692993,
+      "learning_rate": 0.00019821681285078522,
+      "loss": 3.9212,
+      "step": 696
+    },
+    {
+      "epoch": 0.0630940526839866,
+      "grad_norm": 0.9820570349693298,
+      "learning_rate": 0.0001982108419037977,
+      "loss": 3.8646,
+      "step": 697
+    },
+    {
+      "epoch": 0.06318457499773694,
+      "grad_norm": 1.104885220527649,
+      "learning_rate": 0.00019820486106698112,
+      "loss": 3.9699,
+      "step": 698
+    },
+    {
+      "epoch": 0.06327509731148728,
+      "grad_norm": 0.9958544969558716,
+      "learning_rate": 0.00019819887034093768,
+      "loss": 3.8866,
+      "step": 699
+    },
+    {
+      "epoch": 0.06336561962523762,
+      "grad_norm": 1.0173392295837402,
+      "learning_rate": 0.00019819286972627066,
+      "loss": 3.9072,
+      "step": 700
+    },
+    {
+      "epoch": 0.06345614193898796,
+      "grad_norm": 0.9573622345924377,
+      "learning_rate": 0.0001981868592235844,
+      "loss": 3.932,
+      "step": 701
+    },
+    {
+      "epoch": 0.0635466642527383,
+      "grad_norm": 0.9838647246360779,
+      "learning_rate": 0.0001981808388334841,
+      "loss": 3.851,
+      "step": 702
+    },
+    {
+      "epoch": 0.06363718656648865,
+      "grad_norm": 0.9603915810585022,
+      "learning_rate": 0.00019817480855657599,
+      "loss": 3.9591,
+      "step": 703
+    },
+    {
+      "epoch": 0.06372770888023897,
+      "grad_norm": 1.0218600034713745,
+      "learning_rate": 0.00019816876839346735,
+      "loss": 3.8491,
+      "step": 704
+    },
+    {
+      "epoch": 0.06381823119398931,
+      "grad_norm": 1.0183830261230469,
+      "learning_rate": 0.00019816271834476642,
+      "loss": 3.8674,
+      "step": 705
+    },
+    {
+      "epoch": 0.06390875350773965,
+      "grad_norm": 1.0954585075378418,
+      "learning_rate": 0.00019815665841108241,
+      "loss": 3.9141,
+      "step": 706
+    },
+    {
+      "epoch": 0.06399927582149,
+      "grad_norm": 1.0194724798202515,
+      "learning_rate": 0.00019815058859302563,
+      "loss": 3.9022,
+      "step": 707
+    },
+    {
+      "epoch": 0.06408979813524034,
+      "grad_norm": 0.9629772901535034,
+      "learning_rate": 0.00019814450889120723,
+      "loss": 3.8725,
+      "step": 708
+    },
+    {
+      "epoch": 0.06418032044899068,
+      "grad_norm": 0.9155080914497375,
+      "learning_rate": 0.0001981384193062395,
+      "loss": 3.9373,
+      "step": 709
+    },
+    {
+      "epoch": 0.06427084276274102,
+      "grad_norm": 0.9478293657302856,
+      "learning_rate": 0.0001981323198387356,
+      "loss": 3.9601,
+      "step": 710
+    },
+    {
+      "epoch": 0.06436136507649136,
+      "grad_norm": 0.9674055576324463,
+      "learning_rate": 0.0001981262104893098,
+      "loss": 3.8731,
+      "step": 711
+    },
+    {
+      "epoch": 0.0644518873902417,
+      "grad_norm": 0.898500382900238,
+      "learning_rate": 0.00019812009125857728,
+      "loss": 3.8238,
+      "step": 712
+    },
+    {
+      "epoch": 0.06454240970399204,
+      "grad_norm": 0.8977572321891785,
+      "learning_rate": 0.00019811396214715427,
+      "loss": 3.8915,
+      "step": 713
+    },
+    {
+      "epoch": 0.06463293201774237,
+      "grad_norm": 0.8710629940032959,
+      "learning_rate": 0.00019810782315565794,
+      "loss": 3.8397,
+      "step": 714
+    },
+    {
+      "epoch": 0.06472345433149271,
+      "grad_norm": 0.9772562980651855,
+      "learning_rate": 0.00019810167428470653,
+      "loss": 3.8439,
+      "step": 715
+    },
+    {
+      "epoch": 0.06481397664524305,
+      "grad_norm": 0.8889133930206299,
+      "learning_rate": 0.00019809551553491916,
+      "loss": 3.8663,
+      "step": 716
+    },
+    {
+      "epoch": 0.06490449895899339,
+      "grad_norm": 0.9468355774879456,
+      "learning_rate": 0.0001980893469069161,
+      "loss": 3.8705,
+      "step": 717
+    },
+    {
+      "epoch": 0.06499502127274373,
+      "grad_norm": 0.9171603322029114,
+      "learning_rate": 0.00019808316840131846,
+      "loss": 3.8796,
+      "step": 718
+    },
+    {
+      "epoch": 0.06508554358649407,
+      "grad_norm": 0.943007230758667,
+      "learning_rate": 0.00019807698001874846,
+      "loss": 3.8835,
+      "step": 719
+    },
+    {
+      "epoch": 0.06517606590024441,
+      "grad_norm": 0.8981984257698059,
+      "learning_rate": 0.00019807078175982924,
+      "loss": 3.9003,
+      "step": 720
+    },
+    {
+      "epoch": 0.06526658821399475,
+      "grad_norm": 0.9815030694007874,
+      "learning_rate": 0.000198064573625185,
+      "loss": 3.8903,
+      "step": 721
+    },
+    {
+      "epoch": 0.0653571105277451,
+      "grad_norm": 0.9398093819618225,
+      "learning_rate": 0.00019805835561544086,
+      "loss": 3.8726,
+      "step": 722
+    },
+    {
+      "epoch": 0.06544763284149543,
+      "grad_norm": 0.9546122550964355,
+      "learning_rate": 0.00019805212773122303,
+      "loss": 3.8834,
+      "step": 723
+    },
+    {
+      "epoch": 0.06553815515524576,
+      "grad_norm": 0.9206355810165405,
+      "learning_rate": 0.00019804588997315858,
+      "loss": 3.8714,
+      "step": 724
+    },
+    {
+      "epoch": 0.0656286774689961,
+      "grad_norm": 1.0589860677719116,
+      "learning_rate": 0.0001980396423418757,
+      "loss": 3.8971,
+      "step": 725
+    },
+    {
+      "epoch": 0.06571919978274644,
+      "grad_norm": 0.9480148553848267,
+      "learning_rate": 0.00019803338483800353,
+      "loss": 3.8883,
+      "step": 726
+    },
+    {
+      "epoch": 0.06580972209649678,
+      "grad_norm": 1.0164626836776733,
+      "learning_rate": 0.00019802711746217218,
+      "loss": 3.8837,
+      "step": 727
+    },
+    {
+      "epoch": 0.06590024441024712,
+      "grad_norm": 0.9303282499313354,
+      "learning_rate": 0.0001980208402150128,
+      "loss": 3.8288,
+      "step": 728
+    },
+    {
+      "epoch": 0.06599076672399747,
+      "grad_norm": 0.9358525276184082,
+      "learning_rate": 0.00019801455309715748,
+      "loss": 3.8526,
+      "step": 729
+    },
+    {
+      "epoch": 0.0660812890377478,
+      "grad_norm": 0.9523688554763794,
+      "learning_rate": 0.00019800825610923934,
+      "loss": 3.8298,
+      "step": 730
+    },
+    {
+      "epoch": 0.06617181135149815,
+      "grad_norm": 0.9664517045021057,
+      "learning_rate": 0.0001980019492518925,
+      "loss": 3.8322,
+      "step": 731
+    },
+    {
+      "epoch": 0.06626233366524849,
+      "grad_norm": 0.9282121062278748,
+      "learning_rate": 0.00019799563252575206,
+      "loss": 3.8764,
+      "step": 732
+    },
+    {
+      "epoch": 0.06635285597899883,
+      "grad_norm": 0.9822096824645996,
+      "learning_rate": 0.0001979893059314541,
+      "loss": 3.8923,
+      "step": 733
+    },
+    {
+      "epoch": 0.06644337829274917,
+      "grad_norm": 1.003895878791809,
+      "learning_rate": 0.0001979829694696357,
+      "loss": 3.8908,
+      "step": 734
+    },
+    {
+      "epoch": 0.0665339006064995,
+      "grad_norm": 0.9121044874191284,
+      "learning_rate": 0.00019797662314093497,
+      "loss": 3.8674,
+      "step": 735
+    },
+    {
+      "epoch": 0.06662442292024984,
+      "grad_norm": 0.898923933506012,
+      "learning_rate": 0.00019797026694599098,
+      "loss": 3.8893,
+      "step": 736
+    },
+    {
+      "epoch": 0.06671494523400018,
+      "grad_norm": 0.9216980338096619,
+      "learning_rate": 0.00019796390088544377,
+      "loss": 3.7977,
+      "step": 737
+    },
+    {
+      "epoch": 0.06680546754775052,
+      "grad_norm": 0.9738581776618958,
+      "learning_rate": 0.0001979575249599344,
+      "loss": 3.8256,
+      "step": 738
+    },
+    {
+      "epoch": 0.06689598986150086,
+      "grad_norm": 0.8819999098777771,
+      "learning_rate": 0.000197951139170105,
+      "loss": 3.8756,
+      "step": 739
+    },
+    {
+      "epoch": 0.0669865121752512,
+      "grad_norm": 0.925317108631134,
+      "learning_rate": 0.00019794474351659852,
+      "loss": 3.8616,
+      "step": 740
+    },
+    {
+      "epoch": 0.06707703448900154,
+      "grad_norm": 0.9750755429267883,
+      "learning_rate": 0.00019793833800005908,
+      "loss": 3.8383,
+      "step": 741
+    },
+    {
+      "epoch": 0.06716755680275188,
+      "grad_norm": 0.9731130599975586,
+      "learning_rate": 0.00019793192262113166,
+      "loss": 3.8245,
+      "step": 742
+    },
+    {
+      "epoch": 0.06725807911650222,
+      "grad_norm": 0.9129505753517151,
+      "learning_rate": 0.00019792549738046234,
+      "loss": 3.8026,
+      "step": 743
+    },
+    {
+      "epoch": 0.06734860143025256,
+      "grad_norm": 0.971178412437439,
+      "learning_rate": 0.00019791906227869808,
+      "loss": 3.8382,
+      "step": 744
+    },
+    {
+      "epoch": 0.06743912374400289,
+      "grad_norm": 1.011927843093872,
+      "learning_rate": 0.00019791261731648695,
+      "loss": 3.8173,
+      "step": 745
+    },
+    {
+      "epoch": 0.06752964605775323,
+      "grad_norm": 0.9691061973571777,
+      "learning_rate": 0.00019790616249447794,
+      "loss": 3.8659,
+      "step": 746
+    },
+    {
+      "epoch": 0.06762016837150357,
+      "grad_norm": 0.9729804992675781,
+      "learning_rate": 0.000197899697813321,
+      "loss": 3.8626,
+      "step": 747
+    },
+    {
+      "epoch": 0.06771069068525391,
+      "grad_norm": 0.9716980457305908,
+      "learning_rate": 0.00019789322327366723,
+      "loss": 3.9086,
+      "step": 748
+    },
+    {
+      "epoch": 0.06780121299900425,
+      "grad_norm": 1.1275818347930908,
+      "learning_rate": 0.0001978867388761685,
+      "loss": 3.8068,
+      "step": 749
+    },
+    {
+      "epoch": 0.0678917353127546,
+      "grad_norm": 0.9165233969688416,
+      "learning_rate": 0.00019788024462147788,
+      "loss": 3.9002,
+      "step": 750
+    },
+    {
+      "epoch": 0.06798225762650494,
+      "grad_norm": 0.9921157956123352,
+      "learning_rate": 0.0001978737405102493,
+      "loss": 3.833,
+      "step": 751
+    },
+    {
+      "epoch": 0.06807277994025528,
+      "grad_norm": 0.9720535278320312,
+      "learning_rate": 0.00019786722654313772,
+      "loss": 3.8524,
+      "step": 752
+    },
+    {
+      "epoch": 0.06816330225400562,
+      "grad_norm": 0.9478955864906311,
+      "learning_rate": 0.00019786070272079912,
+      "loss": 3.8483,
+      "step": 753
+    },
+    {
+      "epoch": 0.06825382456775596,
+      "grad_norm": 1.005020260810852,
+      "learning_rate": 0.00019785416904389042,
+      "loss": 3.9172,
+      "step": 754
+    },
+    {
+      "epoch": 0.06834434688150628,
+      "grad_norm": 0.9765123128890991,
+      "learning_rate": 0.00019784762551306955,
+      "loss": 3.8409,
+      "step": 755
+    },
+    {
+      "epoch": 0.06843486919525663,
+      "grad_norm": 1.0998058319091797,
+      "learning_rate": 0.00019784107212899552,
+      "loss": 3.8233,
+      "step": 756
+    },
+    {
+      "epoch": 0.06852539150900697,
+      "grad_norm": 0.9845994710922241,
+      "learning_rate": 0.00019783450889232818,
+      "loss": 3.9168,
+      "step": 757
+    },
+    {
+      "epoch": 0.06861591382275731,
+      "grad_norm": 1.118572473526001,
+      "learning_rate": 0.00019782793580372848,
+      "loss": 3.9189,
+      "step": 758
+    },
+    {
+      "epoch": 0.06870643613650765,
+      "grad_norm": 1.011348009109497,
+      "learning_rate": 0.0001978213528638583,
+      "loss": 3.8837,
+      "step": 759
+    },
+    {
+      "epoch": 0.06879695845025799,
+      "grad_norm": 1.0644543170928955,
+      "learning_rate": 0.00019781476007338058,
+      "loss": 3.9323,
+      "step": 760
+    },
+    {
+      "epoch": 0.06888748076400833,
+      "grad_norm": 0.9996188282966614,
+      "learning_rate": 0.0001978081574329592,
+      "loss": 3.8424,
+      "step": 761
+    },
+    {
+      "epoch": 0.06897800307775867,
+      "grad_norm": 1.0625057220458984,
+      "learning_rate": 0.00019780154494325903,
+      "loss": 3.8176,
+      "step": 762
+    },
+    {
+      "epoch": 0.06906852539150901,
+      "grad_norm": 1.0283875465393066,
+      "learning_rate": 0.00019779492260494594,
+      "loss": 3.8401,
+      "step": 763
+    },
+    {
+      "epoch": 0.06915904770525935,
+      "grad_norm": 1.0000511407852173,
+      "learning_rate": 0.00019778829041868689,
+      "loss": 3.8379,
+      "step": 764
+    },
+    {
+      "epoch": 0.06924957001900968,
+      "grad_norm": 1.0903455018997192,
+      "learning_rate": 0.00019778164838514962,
+      "loss": 3.8621,
+      "step": 765
+    },
+    {
+      "epoch": 0.06934009233276002,
+      "grad_norm": 1.013375163078308,
+      "learning_rate": 0.000197774996505003,
+      "loss": 3.797,
+      "step": 766
+    },
+    {
+      "epoch": 0.06943061464651036,
+      "grad_norm": 1.1123466491699219,
+      "learning_rate": 0.00019776833477891696,
+      "loss": 3.8484,
+      "step": 767
+    },
+    {
+      "epoch": 0.0695211369602607,
+      "grad_norm": 0.9966886043548584,
+      "learning_rate": 0.00019776166320756227,
+      "loss": 3.8394,
+      "step": 768
+    },
+    {
+      "epoch": 0.06961165927401104,
+      "grad_norm": 1.0038275718688965,
+      "learning_rate": 0.00019775498179161076,
+      "loss": 3.8286,
+      "step": 769
+    },
+    {
+      "epoch": 0.06970218158776138,
+      "grad_norm": 1.0068305730819702,
+      "learning_rate": 0.00019774829053173526,
+      "loss": 3.918,
+      "step": 770
+    },
+    {
+      "epoch": 0.06979270390151172,
+      "grad_norm": 1.0238758325576782,
+      "learning_rate": 0.0001977415894286096,
+      "loss": 3.801,
+      "step": 771
+    },
+    {
+      "epoch": 0.06988322621526206,
+      "grad_norm": 1.0598841905593872,
+      "learning_rate": 0.00019773487848290854,
+      "loss": 3.8412,
+      "step": 772
+    },
+    {
+      "epoch": 0.0699737485290124,
+      "grad_norm": 1.0061222314834595,
+      "learning_rate": 0.0001977281576953079,
+      "loss": 3.8693,
+      "step": 773
+    },
+    {
+      "epoch": 0.07006427084276275,
+      "grad_norm": 0.9197167754173279,
+      "learning_rate": 0.00019772142706648443,
+      "loss": 3.8136,
+      "step": 774
+    },
+    {
+      "epoch": 0.07015479315651309,
+      "grad_norm": 0.9932407140731812,
+      "learning_rate": 0.00019771468659711595,
+      "loss": 3.8804,
+      "step": 775
+    },
+    {
+      "epoch": 0.07024531547026341,
+      "grad_norm": 0.91663658618927,
+      "learning_rate": 0.00019770793628788122,
+      "loss": 3.9397,
+      "step": 776
+    },
+    {
+      "epoch": 0.07033583778401375,
+      "grad_norm": 0.9589143991470337,
+      "learning_rate": 0.00019770117613945995,
+      "loss": 3.8571,
+      "step": 777
+    },
+    {
+      "epoch": 0.0704263600977641,
+      "grad_norm": 0.9023710489273071,
+      "learning_rate": 0.00019769440615253293,
+      "loss": 3.861,
+      "step": 778
+    },
+    {
+      "epoch": 0.07051688241151444,
+      "grad_norm": 0.9534058570861816,
+      "learning_rate": 0.00019768762632778187,
+      "loss": 3.875,
+      "step": 779
+    },
+    {
+      "epoch": 0.07060740472526478,
+      "grad_norm": 0.9815419316291809,
+      "learning_rate": 0.00019768083666588953,
+      "loss": 3.7904,
+      "step": 780
+    },
+    {
+      "epoch": 0.07069792703901512,
+      "grad_norm": 1.0916591882705688,
+      "learning_rate": 0.00019767403716753959,
+      "loss": 3.8464,
+      "step": 781
+    },
+    {
+      "epoch": 0.07078844935276546,
+      "grad_norm": 0.9650158882141113,
+      "learning_rate": 0.0001976672278334168,
+      "loss": 3.8239,
+      "step": 782
+    },
+    {
+      "epoch": 0.0708789716665158,
+      "grad_norm": 0.9654297232627869,
+      "learning_rate": 0.00019766040866420683,
+      "loss": 3.9072,
+      "step": 783
+    },
+    {
+      "epoch": 0.07096949398026614,
+      "grad_norm": 0.9741345643997192,
+      "learning_rate": 0.00019765357966059638,
+      "loss": 3.8536,
+      "step": 784
+    },
+    {
+      "epoch": 0.07106001629401648,
+      "grad_norm": 0.9063356518745422,
+      "learning_rate": 0.0001976467408232731,
+      "loss": 3.8186,
+      "step": 785
+    },
+    {
+      "epoch": 0.07115053860776681,
+      "grad_norm": 0.9848807454109192,
+      "learning_rate": 0.00019763989215292575,
+      "loss": 3.866,
+      "step": 786
+    },
+    {
+      "epoch": 0.07124106092151715,
+      "grad_norm": 1.0605283975601196,
+      "learning_rate": 0.0001976330336502439,
+      "loss": 3.9509,
+      "step": 787
+    },
+    {
+      "epoch": 0.07133158323526749,
+      "grad_norm": 0.9561129808425903,
+      "learning_rate": 0.00019762616531591824,
+      "loss": 3.8792,
+      "step": 788
+    },
+    {
+      "epoch": 0.07142210554901783,
+      "grad_norm": 0.9112474322319031,
+      "learning_rate": 0.0001976192871506404,
+      "loss": 3.8499,
+      "step": 789
+    },
+    {
+      "epoch": 0.07151262786276817,
+      "grad_norm": 1.0188720226287842,
+      "learning_rate": 0.00019761239915510302,
+      "loss": 3.7913,
+      "step": 790
+    },
+    {
+      "epoch": 0.07160315017651851,
+      "grad_norm": 0.914110541343689,
+      "learning_rate": 0.0001976055013299997,
+      "loss": 3.8441,
+      "step": 791
+    },
+    {
+      "epoch": 0.07169367249026885,
+      "grad_norm": 1.0345866680145264,
+      "learning_rate": 0.0001975985936760251,
+      "loss": 3.7915,
+      "step": 792
+    },
+    {
+      "epoch": 0.0717841948040192,
+      "grad_norm": 0.9036014676094055,
+      "learning_rate": 0.00019759167619387476,
+      "loss": 3.8819,
+      "step": 793
+    },
+    {
+      "epoch": 0.07187471711776953,
+      "grad_norm": 0.8982127904891968,
+      "learning_rate": 0.0001975847488842453,
+      "loss": 3.7775,
+      "step": 794
+    },
+    {
+      "epoch": 0.07196523943151988,
+      "grad_norm": 0.8711851835250854,
+      "learning_rate": 0.00019757781174783428,
+      "loss": 3.8353,
+      "step": 795
+    },
+    {
+      "epoch": 0.0720557617452702,
+      "grad_norm": 0.9486883878707886,
+      "learning_rate": 0.0001975708647853403,
+      "loss": 3.8049,
+      "step": 796
+    },
+    {
+      "epoch": 0.07214628405902054,
+      "grad_norm": 0.9494708776473999,
+      "learning_rate": 0.00019756390799746294,
+      "loss": 3.8221,
+      "step": 797
+    },
+    {
+      "epoch": 0.07223680637277088,
+      "grad_norm": 1.008435845375061,
+      "learning_rate": 0.0001975569413849027,
+      "loss": 3.8172,
+      "step": 798
+    },
+    {
+      "epoch": 0.07232732868652123,
+      "grad_norm": 0.9401419758796692,
+      "learning_rate": 0.0001975499649483611,
+      "loss": 3.8145,
+      "step": 799
+    },
+    {
+      "epoch": 0.07241785100027157,
+      "grad_norm": 0.8914586305618286,
+      "learning_rate": 0.00019754297868854073,
+      "loss": 3.7788,
+      "step": 800
+    },
+    {
+      "epoch": 0.0725083733140219,
+      "grad_norm": 0.9457998275756836,
+      "learning_rate": 0.00019753598260614506,
+      "loss": 3.8508,
+      "step": 801
+    },
+    {
+      "epoch": 0.07259889562777225,
+      "grad_norm": 0.9615713953971863,
+      "learning_rate": 0.0001975289767018786,
+      "loss": 3.8725,
+      "step": 802
+    },
+    {
+      "epoch": 0.07268941794152259,
+      "grad_norm": 0.942954421043396,
+      "learning_rate": 0.00019752196097644686,
+      "loss": 3.8043,
+      "step": 803
+    },
+    {
+      "epoch": 0.07277994025527293,
+      "grad_norm": 0.9813474416732788,
+      "learning_rate": 0.00019751493543055632,
+      "loss": 3.8567,
+      "step": 804
+    },
+    {
+      "epoch": 0.07287046256902327,
+      "grad_norm": 1.0663015842437744,
+      "learning_rate": 0.00019750790006491448,
+      "loss": 3.8372,
+      "step": 805
+    },
+    {
+      "epoch": 0.0729609848827736,
+      "grad_norm": 0.9190868139266968,
+      "learning_rate": 0.00019750085488022973,
+      "loss": 3.8005,
+      "step": 806
+    },
+    {
+      "epoch": 0.07305150719652394,
+      "grad_norm": 0.9721860289573669,
+      "learning_rate": 0.00019749379987721159,
+      "loss": 3.7956,
+      "step": 807
+    },
+    {
+      "epoch": 0.07314202951027428,
+      "grad_norm": 1.0076109170913696,
+      "learning_rate": 0.00019748673505657046,
+      "loss": 3.8849,
+      "step": 808
+    },
+    {
+      "epoch": 0.07323255182402462,
+      "grad_norm": 1.0171688795089722,
+      "learning_rate": 0.00019747966041901776,
+      "loss": 3.8125,
+      "step": 809
+    },
+    {
+      "epoch": 0.07332307413777496,
+      "grad_norm": 0.9697157740592957,
+      "learning_rate": 0.00019747257596526593,
+      "loss": 3.7726,
+      "step": 810
+    },
+    {
+      "epoch": 0.0734135964515253,
+      "grad_norm": 1.0516706705093384,
+      "learning_rate": 0.0001974654816960284,
+      "loss": 3.7536,
+      "step": 811
+    },
+    {
+      "epoch": 0.07350411876527564,
+      "grad_norm": 0.9496332406997681,
+      "learning_rate": 0.00019745837761201947,
+      "loss": 3.8467,
+      "step": 812
+    },
+    {
+      "epoch": 0.07359464107902598,
+      "grad_norm": 1.0452812910079956,
+      "learning_rate": 0.00019745126371395464,
+      "loss": 3.7487,
+      "step": 813
+    },
+    {
+      "epoch": 0.07368516339277632,
+      "grad_norm": 1.0293980836868286,
+      "learning_rate": 0.0001974441400025502,
+      "loss": 3.8284,
+      "step": 814
+    },
+    {
+      "epoch": 0.07377568570652666,
+      "grad_norm": 0.9970689415931702,
+      "learning_rate": 0.00019743700647852354,
+      "loss": 3.8618,
+      "step": 815
+    },
+    {
+      "epoch": 0.073866208020277,
+      "grad_norm": 0.9802175760269165,
+      "learning_rate": 0.00019742986314259299,
+      "loss": 3.8989,
+      "step": 816
+    },
+    {
+      "epoch": 0.07395673033402733,
+      "grad_norm": 0.945270299911499,
+      "learning_rate": 0.00019742270999547788,
+      "loss": 3.8061,
+      "step": 817
+    },
+    {
+      "epoch": 0.07404725264777767,
+      "grad_norm": 0.9834704399108887,
+      "learning_rate": 0.00019741554703789853,
+      "loss": 3.8592,
+      "step": 818
+    },
+    {
+      "epoch": 0.07413777496152801,
+      "grad_norm": 0.889618992805481,
+      "learning_rate": 0.00019740837427057625,
+      "loss": 3.7997,
+      "step": 819
+    },
+    {
+      "epoch": 0.07422829727527835,
+      "grad_norm": 0.9406912326812744,
+      "learning_rate": 0.00019740119169423337,
+      "loss": 3.827,
+      "step": 820
+    },
+    {
+      "epoch": 0.0743188195890287,
+      "grad_norm": 0.8791980147361755,
+      "learning_rate": 0.00019739399930959314,
+      "loss": 3.7941,
+      "step": 821
+    },
+    {
+      "epoch": 0.07440934190277904,
+      "grad_norm": 0.89346843957901,
+      "learning_rate": 0.00019738679711737983,
+      "loss": 3.7829,
+      "step": 822
+    },
+    {
+      "epoch": 0.07449986421652938,
+      "grad_norm": 0.9298461675643921,
+      "learning_rate": 0.00019737958511831875,
+      "loss": 3.8339,
+      "step": 823
+    },
+    {
+      "epoch": 0.07459038653027972,
+      "grad_norm": 0.9051544666290283,
+      "learning_rate": 0.00019737236331313608,
+      "loss": 3.8614,
+      "step": 824
+    },
+    {
+      "epoch": 0.07468090884403006,
+      "grad_norm": 1.0275598764419556,
+      "learning_rate": 0.00019736513170255911,
+      "loss": 3.8188,
+      "step": 825
+    },
+    {
+      "epoch": 0.0747714311577804,
+      "grad_norm": 1.0261467695236206,
+      "learning_rate": 0.00019735789028731604,
+      "loss": 3.7921,
+      "step": 826
+    },
+    {
+      "epoch": 0.07486195347153073,
+      "grad_norm": 0.916692852973938,
+      "learning_rate": 0.00019735063906813606,
+      "loss": 3.7529,
+      "step": 827
+    },
+    {
+      "epoch": 0.07495247578528107,
+      "grad_norm": 1.012143850326538,
+      "learning_rate": 0.00019734337804574943,
+      "loss": 3.8495,
+      "step": 828
+    },
+    {
+      "epoch": 0.07504299809903141,
+      "grad_norm": 0.9389269351959229,
+      "learning_rate": 0.00019733610722088725,
+      "loss": 3.8664,
+      "step": 829
+    },
+    {
+      "epoch": 0.07513352041278175,
+      "grad_norm": 0.9384870529174805,
+      "learning_rate": 0.00019732882659428177,
+      "loss": 3.7599,
+      "step": 830
+    },
+    {
+      "epoch": 0.07522404272653209,
+      "grad_norm": 0.9308289885520935,
+      "learning_rate": 0.00019732153616666608,
+      "loss": 3.877,
+      "step": 831
+    },
+    {
+      "epoch": 0.07531456504028243,
+      "grad_norm": 0.9849106669425964,
+      "learning_rate": 0.0001973142359387744,
+      "loss": 3.7972,
+      "step": 832
+    },
+    {
+      "epoch": 0.07540508735403277,
+      "grad_norm": 0.9762073159217834,
+      "learning_rate": 0.0001973069259113418,
+      "loss": 3.8222,
+      "step": 833
+    },
+    {
+      "epoch": 0.07549560966778311,
+      "grad_norm": 0.946532130241394,
+      "learning_rate": 0.00019729960608510445,
+      "loss": 3.8143,
+      "step": 834
+    },
+    {
+      "epoch": 0.07558613198153345,
+      "grad_norm": 0.9300776124000549,
+      "learning_rate": 0.00019729227646079938,
+      "loss": 3.7953,
+      "step": 835
+    },
+    {
+      "epoch": 0.0756766542952838,
+      "grad_norm": 0.9871339797973633,
+      "learning_rate": 0.00019728493703916478,
+      "loss": 3.7921,
+      "step": 836
+    },
+    {
+      "epoch": 0.07576717660903412,
+      "grad_norm": 0.932281494140625,
+      "learning_rate": 0.00019727758782093967,
+      "loss": 3.8262,
+      "step": 837
+    },
+    {
+      "epoch": 0.07585769892278446,
+      "grad_norm": 0.9992408752441406,
+      "learning_rate": 0.00019727022880686412,
+      "loss": 3.8204,
+      "step": 838
+    },
+    {
+      "epoch": 0.0759482212365348,
+      "grad_norm": 0.9957219958305359,
+      "learning_rate": 0.00019726285999767919,
+      "loss": 3.8209,
+      "step": 839
+    },
+    {
+      "epoch": 0.07603874355028514,
+      "grad_norm": 0.9381322264671326,
+      "learning_rate": 0.00019725548139412692,
+      "loss": 3.7818,
+      "step": 840
+    },
+    {
+      "epoch": 0.07612926586403548,
+      "grad_norm": 0.8691635727882385,
+      "learning_rate": 0.00019724809299695033,
+      "loss": 3.7704,
+      "step": 841
+    },
+    {
+      "epoch": 0.07621978817778582,
+      "grad_norm": 0.9367995858192444,
+      "learning_rate": 0.00019724069480689345,
+      "loss": 3.768,
+      "step": 842
+    },
+    {
+      "epoch": 0.07631031049153617,
+      "grad_norm": 0.9435257911682129,
+      "learning_rate": 0.00019723328682470125,
+      "loss": 3.7615,
+      "step": 843
+    },
+    {
+      "epoch": 0.0764008328052865,
+      "grad_norm": 0.9820596575737,
+      "learning_rate": 0.00019722586905111976,
+      "loss": 3.823,
+      "step": 844
+    },
+    {
+      "epoch": 0.07649135511903685,
+      "grad_norm": 1.0243360996246338,
+      "learning_rate": 0.00019721844148689588,
+      "loss": 3.8064,
+      "step": 845
+    },
+    {
+      "epoch": 0.07658187743278719,
+      "grad_norm": 1.0168771743774414,
+      "learning_rate": 0.00019721100413277762,
+      "loss": 3.856,
+      "step": 846
+    },
+    {
+      "epoch": 0.07667239974653753,
+      "grad_norm": 0.9951126575469971,
+      "learning_rate": 0.00019720355698951388,
+      "loss": 3.8397,
+      "step": 847
+    },
+    {
+      "epoch": 0.07676292206028786,
+      "grad_norm": 0.9725316166877747,
+      "learning_rate": 0.00019719610005785465,
+      "loss": 3.8152,
+      "step": 848
+    },
+    {
+      "epoch": 0.0768534443740382,
+      "grad_norm": 0.9000321626663208,
+      "learning_rate": 0.0001971886333385508,
+      "loss": 3.8176,
+      "step": 849
+    },
+    {
+      "epoch": 0.07694396668778854,
+      "grad_norm": 0.9348044991493225,
+      "learning_rate": 0.00019718115683235417,
+      "loss": 3.7787,
+      "step": 850
+    },
+    {
+      "epoch": 0.07703448900153888,
+      "grad_norm": 0.9273238778114319,
+      "learning_rate": 0.00019717367054001775,
+      "loss": 3.8014,
+      "step": 851
+    },
+    {
+      "epoch": 0.07712501131528922,
+      "grad_norm": 0.9387201070785522,
+      "learning_rate": 0.00019716617446229536,
+      "loss": 3.7383,
+      "step": 852
+    },
+    {
+      "epoch": 0.07721553362903956,
+      "grad_norm": 0.9373703002929688,
+      "learning_rate": 0.0001971586685999419,
+      "loss": 3.8054,
+      "step": 853
+    },
+    {
+      "epoch": 0.0773060559427899,
+      "grad_norm": 0.9412992596626282,
+      "learning_rate": 0.00019715115295371313,
+      "loss": 3.7198,
+      "step": 854
+    },
+    {
+      "epoch": 0.07739657825654024,
+      "grad_norm": 1.0220283269882202,
+      "learning_rate": 0.0001971436275243659,
+      "loss": 3.7441,
+      "step": 855
+    },
+    {
+      "epoch": 0.07748710057029058,
+      "grad_norm": 0.9344440698623657,
+      "learning_rate": 0.00019713609231265805,
+      "loss": 3.8045,
+      "step": 856
+    },
+    {
+      "epoch": 0.07757762288404092,
+      "grad_norm": 0.9369484186172485,
+      "learning_rate": 0.00019712854731934837,
+      "loss": 3.765,
+      "step": 857
+    },
+    {
+      "epoch": 0.07766814519779125,
+      "grad_norm": 0.8779152631759644,
+      "learning_rate": 0.00019712099254519665,
+      "loss": 3.7957,
+      "step": 858
+    },
+    {
+      "epoch": 0.07775866751154159,
+      "grad_norm": 0.9569523930549622,
+      "learning_rate": 0.00019711342799096361,
+      "loss": 3.7851,
+      "step": 859
+    },
+    {
+      "epoch": 0.07784918982529193,
+      "grad_norm": 0.930668294429779,
+      "learning_rate": 0.00019710585365741103,
+      "loss": 3.8031,
+      "step": 860
+    },
+    {
+      "epoch": 0.07793971213904227,
+      "grad_norm": 0.9617711305618286,
+      "learning_rate": 0.00019709826954530168,
+      "loss": 3.8352,
+      "step": 861
+    },
+    {
+      "epoch": 0.07803023445279261,
+      "grad_norm": 0.9281615018844604,
+      "learning_rate": 0.0001970906756553992,
+      "loss": 3.8318,
+      "step": 862
+    },
+    {
+      "epoch": 0.07812075676654295,
+      "grad_norm": 0.9399791359901428,
+      "learning_rate": 0.00019708307198846835,
+      "loss": 3.7991,
+      "step": 863
+    },
+    {
+      "epoch": 0.0782112790802933,
+      "grad_norm": 0.965398371219635,
+      "learning_rate": 0.0001970754585452748,
+      "loss": 3.7721,
+      "step": 864
+    },
+    {
+      "epoch": 0.07830180139404364,
+      "grad_norm": 0.9370290040969849,
+      "learning_rate": 0.00019706783532658526,
+      "loss": 3.8555,
+      "step": 865
+    },
+    {
+      "epoch": 0.07839232370779398,
+      "grad_norm": 0.9139227867126465,
+      "learning_rate": 0.00019706020233316735,
+      "loss": 3.7587,
+      "step": 866
+    },
+    {
+      "epoch": 0.07848284602154432,
+      "grad_norm": 0.9634401202201843,
+      "learning_rate": 0.00019705255956578972,
+      "loss": 3.7673,
+      "step": 867
+    },
+    {
+      "epoch": 0.07857336833529464,
+      "grad_norm": 1.0112007856369019,
+      "learning_rate": 0.00019704490702522197,
+      "loss": 3.7562,
+      "step": 868
+    },
+    {
+      "epoch": 0.07866389064904498,
+      "grad_norm": 0.9345536231994629,
+      "learning_rate": 0.00019703724471223475,
+      "loss": 3.8501,
+      "step": 869
+    },
+    {
+      "epoch": 0.07875441296279533,
+      "grad_norm": 0.9797864556312561,
+      "learning_rate": 0.00019702957262759965,
+      "loss": 3.7076,
+      "step": 870
+    },
+    {
+      "epoch": 0.07884493527654567,
+      "grad_norm": 0.9007889032363892,
+      "learning_rate": 0.00019702189077208922,
+      "loss": 3.7771,
+      "step": 871
+    },
+    {
+      "epoch": 0.07893545759029601,
+      "grad_norm": 0.9339497089385986,
+      "learning_rate": 0.0001970141991464771,
+      "loss": 3.8068,
+      "step": 872
+    },
+    {
+      "epoch": 0.07902597990404635,
+      "grad_norm": 0.9866631627082825,
+      "learning_rate": 0.00019700649775153774,
+      "loss": 3.733,
+      "step": 873
+    },
+    {
+      "epoch": 0.07911650221779669,
+      "grad_norm": 0.936284065246582,
+      "learning_rate": 0.00019699878658804672,
+      "loss": 3.7971,
+      "step": 874
+    },
+    {
+      "epoch": 0.07920702453154703,
+      "grad_norm": 1.0367450714111328,
+      "learning_rate": 0.0001969910656567805,
+      "loss": 3.8082,
+      "step": 875
+    },
+    {
+      "epoch": 0.07929754684529737,
+      "grad_norm": 0.9137488603591919,
+      "learning_rate": 0.0001969833349585167,
+      "loss": 3.779,
+      "step": 876
+    },
+    {
+      "epoch": 0.07938806915904771,
+      "grad_norm": 1.0584319829940796,
+      "learning_rate": 0.00019697559449403368,
+      "loss": 3.756,
+      "step": 877
+    },
+    {
+      "epoch": 0.07947859147279804,
+      "grad_norm": 0.9855899810791016,
+      "learning_rate": 0.00019696784426411097,
+      "loss": 3.774,
+      "step": 878
+    },
+    {
+      "epoch": 0.07956911378654838,
+      "grad_norm": 0.9397339820861816,
+      "learning_rate": 0.00019696008426952897,
+      "loss": 3.7419,
+      "step": 879
+    },
+    {
+      "epoch": 0.07965963610029872,
+      "grad_norm": 1.0694619417190552,
+      "learning_rate": 0.00019695231451106912,
+      "loss": 3.839,
+      "step": 880
+    },
+    {
+      "epoch": 0.07975015841404906,
+      "grad_norm": 0.9818793535232544,
+      "learning_rate": 0.0001969445349895139,
+      "loss": 3.8308,
+      "step": 881
+    },
+    {
+      "epoch": 0.0798406807277994,
+      "grad_norm": 1.0380256175994873,
+      "learning_rate": 0.00019693674570564663,
+      "loss": 3.7835,
+      "step": 882
+    },
+    {
+      "epoch": 0.07993120304154974,
+      "grad_norm": 0.9677836894989014,
+      "learning_rate": 0.00019692894666025176,
+      "loss": 3.7801,
+      "step": 883
+    },
+    {
+      "epoch": 0.08002172535530008,
+      "grad_norm": 0.9173345565795898,
+      "learning_rate": 0.00019692113785411456,
+      "loss": 3.6993,
+      "step": 884
+    },
+    {
+      "epoch": 0.08011224766905042,
+      "grad_norm": 0.9473771452903748,
+      "learning_rate": 0.00019691331928802143,
+      "loss": 3.7979,
+      "step": 885
+    },
+    {
+      "epoch": 0.08020276998280076,
+      "grad_norm": 1.0071412324905396,
+      "learning_rate": 0.00019690549096275972,
+      "loss": 3.8488,
+      "step": 886
+    },
+    {
+      "epoch": 0.0802932922965511,
+      "grad_norm": 0.9691148400306702,
+      "learning_rate": 0.00019689765287911773,
+      "loss": 3.7602,
+      "step": 887
+    },
+    {
+      "epoch": 0.08038381461030145,
+      "grad_norm": 0.9631767868995667,
+      "learning_rate": 0.00019688980503788475,
+      "loss": 3.764,
+      "step": 888
+    },
+    {
+      "epoch": 0.08047433692405177,
+      "grad_norm": 1.0345791578292847,
+      "learning_rate": 0.00019688194743985103,
+      "loss": 3.7938,
+      "step": 889
+    },
+    {
+      "epoch": 0.08056485923780211,
+      "grad_norm": 0.9610744714736938,
+      "learning_rate": 0.00019687408008580784,
+      "loss": 3.7792,
+      "step": 890
+    },
+    {
+      "epoch": 0.08065538155155245,
+      "grad_norm": 1.063826560974121,
+      "learning_rate": 0.00019686620297654748,
+      "loss": 3.6855,
+      "step": 891
+    },
+    {
+      "epoch": 0.0807459038653028,
+      "grad_norm": 0.9013906121253967,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 3.7464,
+      "step": 892
+    },
+    {
+      "epoch": 0.08083642617905314,
+      "grad_norm": 0.9041231274604797,
+      "learning_rate": 0.00019685041949554896,
+      "loss": 3.7574,
+      "step": 893
+    },
+    {
+      "epoch": 0.08092694849280348,
+      "grad_norm": 0.9630007743835449,
+      "learning_rate": 0.00019684251312540024,
+      "loss": 3.7775,
+      "step": 894
+    },
+    {
+      "epoch": 0.08101747080655382,
+      "grad_norm": 0.9465706944465637,
+      "learning_rate": 0.00019683459700321304,
+      "loss": 3.7591,
+      "step": 895
+    },
+    {
+      "epoch": 0.08110799312030416,
+      "grad_norm": 0.8408172726631165,
+      "learning_rate": 0.00019682667112978463,
+      "loss": 3.7055,
+      "step": 896
+    },
+    {
+      "epoch": 0.0811985154340545,
+      "grad_norm": 0.9609735608100891,
+      "learning_rate": 0.00019681873550591306,
+      "loss": 3.6991,
+      "step": 897
+    },
+    {
+      "epoch": 0.08128903774780484,
+      "grad_norm": 0.9495016932487488,
+      "learning_rate": 0.00019681079013239748,
+      "loss": 3.7452,
+      "step": 898
+    },
+    {
+      "epoch": 0.08137956006155517,
+      "grad_norm": 1.0434441566467285,
+      "learning_rate": 0.00019680283501003797,
+      "loss": 3.7501,
+      "step": 899
+    },
+    {
+      "epoch": 0.08147008237530551,
+      "grad_norm": 0.9953208565711975,
+      "learning_rate": 0.00019679487013963564,
+      "loss": 3.8621,
+      "step": 900
+    },
+    {
+      "epoch": 0.08156060468905585,
+      "grad_norm": 0.9249243140220642,
+      "learning_rate": 0.0001967868955219925,
+      "loss": 3.6988,
+      "step": 901
+    },
+    {
+      "epoch": 0.08165112700280619,
+      "grad_norm": 1.0066379308700562,
+      "learning_rate": 0.0001967789111579117,
+      "loss": 3.7775,
+      "step": 902
+    },
+    {
+      "epoch": 0.08174164931655653,
+      "grad_norm": 1.0791577100753784,
+      "learning_rate": 0.00019677091704819715,
+      "loss": 3.7848,
+      "step": 903
+    },
+    {
+      "epoch": 0.08183217163030687,
+      "grad_norm": 1.0767844915390015,
+      "learning_rate": 0.00019676291319365387,
+      "loss": 3.7873,
+      "step": 904
+    },
+    {
+      "epoch": 0.08192269394405721,
+      "grad_norm": 0.9516428112983704,
+      "learning_rate": 0.00019675489959508792,
+      "loss": 3.8179,
+      "step": 905
+    },
+    {
+      "epoch": 0.08201321625780755,
+      "grad_norm": 1.0235987901687622,
+      "learning_rate": 0.00019674687625330625,
+      "loss": 3.7205,
+      "step": 906
+    },
+    {
+      "epoch": 0.0821037385715579,
+      "grad_norm": 0.9475753307342529,
+      "learning_rate": 0.00019673884316911673,
+      "loss": 3.7991,
+      "step": 907
+    },
+    {
+      "epoch": 0.08219426088530823,
+      "grad_norm": 1.0284664630889893,
+      "learning_rate": 0.0001967308003433284,
+      "loss": 3.7888,
+      "step": 908
+    },
+    {
+      "epoch": 0.08228478319905856,
+      "grad_norm": 0.9725825190544128,
+      "learning_rate": 0.0001967227477767511,
+      "loss": 3.7334,
+      "step": 909
+    },
+    {
+      "epoch": 0.0823753055128089,
+      "grad_norm": 1.0133692026138306,
+      "learning_rate": 0.00019671468547019573,
+      "loss": 3.6991,
+      "step": 910
+    },
+    {
+      "epoch": 0.08246582782655924,
+      "grad_norm": 0.9631583094596863,
+      "learning_rate": 0.0001967066134244742,
+      "loss": 3.7733,
+      "step": 911
+    },
+    {
+      "epoch": 0.08255635014030958,
+      "grad_norm": 0.8878318667411804,
+      "learning_rate": 0.00019669853164039933,
+      "loss": 3.7456,
+      "step": 912
+    },
+    {
+      "epoch": 0.08264687245405992,
+      "grad_norm": 0.9535197615623474,
+      "learning_rate": 0.00019669044011878497,
+      "loss": 3.7512,
+      "step": 913
+    },
+    {
+      "epoch": 0.08273739476781027,
+      "grad_norm": 0.9916483759880066,
+      "learning_rate": 0.00019668233886044597,
+      "loss": 3.7768,
+      "step": 914
+    },
+    {
+      "epoch": 0.0828279170815606,
+      "grad_norm": 0.8816608786582947,
+      "learning_rate": 0.00019667422786619806,
+      "loss": 3.7579,
+      "step": 915
+    },
+    {
+      "epoch": 0.08291843939531095,
+      "grad_norm": 0.9805126786231995,
+      "learning_rate": 0.00019666610713685802,
+      "loss": 3.6392,
+      "step": 916
+    },
+    {
+      "epoch": 0.08300896170906129,
+      "grad_norm": 1.065384864807129,
+      "learning_rate": 0.0001966579766732437,
+      "loss": 3.8084,
+      "step": 917
+    },
+    {
+      "epoch": 0.08309948402281163,
+      "grad_norm": 0.9200580716133118,
+      "learning_rate": 0.00019664983647617375,
+      "loss": 3.6934,
+      "step": 918
+    },
+    {
+      "epoch": 0.08319000633656196,
+      "grad_norm": 0.911810040473938,
+      "learning_rate": 0.00019664168654646787,
+      "loss": 3.7117,
+      "step": 919
+    },
+    {
+      "epoch": 0.0832805286503123,
+      "grad_norm": 0.9342407584190369,
+      "learning_rate": 0.00019663352688494684,
+      "loss": 3.8123,
+      "step": 920
+    },
+    {
+      "epoch": 0.08337105096406264,
+      "grad_norm": 0.9220612645149231,
+      "learning_rate": 0.00019662535749243233,
+      "loss": 3.7095,
+      "step": 921
+    },
+    {
+      "epoch": 0.08346157327781298,
+      "grad_norm": 0.9277602434158325,
+      "learning_rate": 0.0001966171783697469,
+      "loss": 3.7442,
+      "step": 922
+    },
+    {
+      "epoch": 0.08355209559156332,
+      "grad_norm": 0.9696083068847656,
+      "learning_rate": 0.0001966089895177143,
+      "loss": 3.6917,
+      "step": 923
+    },
+    {
+      "epoch": 0.08364261790531366,
+      "grad_norm": 0.9408513307571411,
+      "learning_rate": 0.00019660079093715906,
+      "loss": 3.7603,
+      "step": 924
+    },
+    {
+      "epoch": 0.083733140219064,
+      "grad_norm": 0.9134141802787781,
+      "learning_rate": 0.00019659258262890683,
+      "loss": 3.7554,
+      "step": 925
+    },
+    {
+      "epoch": 0.08382366253281434,
+      "grad_norm": 1.1565461158752441,
+      "learning_rate": 0.0001965843645937842,
+      "loss": 3.7358,
+      "step": 926
+    },
+    {
+      "epoch": 0.08391418484656468,
+      "grad_norm": 0.9286847114562988,
+      "learning_rate": 0.00019657613683261865,
+      "loss": 3.7097,
+      "step": 927
+    },
+    {
+      "epoch": 0.08400470716031502,
+      "grad_norm": 0.9701700806617737,
+      "learning_rate": 0.00019656789934623881,
+      "loss": 3.6998,
+      "step": 928
+    },
+    {
+      "epoch": 0.08409522947406536,
+      "grad_norm": 0.9360522627830505,
+      "learning_rate": 0.0001965596521354741,
+      "loss": 3.8173,
+      "step": 929
+    },
+    {
+      "epoch": 0.08418575178781569,
+      "grad_norm": 0.9045924544334412,
+      "learning_rate": 0.0001965513952011551,
+      "loss": 3.7291,
+      "step": 930
+    },
+    {
+      "epoch": 0.08427627410156603,
+      "grad_norm": 1.012802243232727,
+      "learning_rate": 0.00019654312854411324,
+      "loss": 3.767,
+      "step": 931
+    },
+    {
+      "epoch": 0.08436679641531637,
+      "grad_norm": 1.0733472108840942,
+      "learning_rate": 0.00019653485216518095,
+      "loss": 3.727,
+      "step": 932
+    },
+    {
+      "epoch": 0.08445731872906671,
+      "grad_norm": 0.977625846862793,
+      "learning_rate": 0.00019652656606519172,
+      "loss": 3.7632,
+      "step": 933
+    },
+    {
+      "epoch": 0.08454784104281705,
+      "grad_norm": 0.9703370928764343,
+      "learning_rate": 0.0001965182702449799,
+      "loss": 3.7985,
+      "step": 934
+    },
+    {
+      "epoch": 0.0846383633565674,
+      "grad_norm": 1.000142216682434,
+      "learning_rate": 0.0001965099647053809,
+      "loss": 3.737,
+      "step": 935
+    },
+    {
+      "epoch": 0.08472888567031774,
+      "grad_norm": 0.9989868402481079,
+      "learning_rate": 0.00019650164944723115,
+      "loss": 3.7328,
+      "step": 936
+    },
+    {
+      "epoch": 0.08481940798406808,
+      "grad_norm": 1.0792580842971802,
+      "learning_rate": 0.0001964933244713679,
+      "loss": 3.7435,
+      "step": 937
+    },
+    {
+      "epoch": 0.08490993029781842,
+      "grad_norm": 0.9847734570503235,
+      "learning_rate": 0.00019648498977862952,
+      "loss": 3.7951,
+      "step": 938
+    },
+    {
+      "epoch": 0.08500045261156876,
+      "grad_norm": 1.0123600959777832,
+      "learning_rate": 0.00019647664536985536,
+      "loss": 3.7984,
+      "step": 939
+    },
+    {
+      "epoch": 0.08509097492531909,
+      "grad_norm": 0.9911985993385315,
+      "learning_rate": 0.0001964682912458856,
+      "loss": 3.7909,
+      "step": 940
+    },
+    {
+      "epoch": 0.08518149723906943,
+      "grad_norm": 0.9254298210144043,
+      "learning_rate": 0.00019645992740756154,
+      "loss": 3.8021,
+      "step": 941
+    },
+    {
+      "epoch": 0.08527201955281977,
+      "grad_norm": 0.912542462348938,
+      "learning_rate": 0.00019645155385572544,
+      "loss": 3.7413,
+      "step": 942
+    },
+    {
+      "epoch": 0.08536254186657011,
+      "grad_norm": 0.9094882607460022,
+      "learning_rate": 0.00019644317059122051,
+      "loss": 3.7391,
+      "step": 943
+    },
+    {
+      "epoch": 0.08545306418032045,
+      "grad_norm": 0.9295530319213867,
+      "learning_rate": 0.00019643477761489096,
+      "loss": 3.7333,
+      "step": 944
+    },
+    {
+      "epoch": 0.08554358649407079,
+      "grad_norm": 0.9964175224304199,
+      "learning_rate": 0.0001964263749275819,
+      "loss": 3.7959,
+      "step": 945
+    },
+    {
+      "epoch": 0.08563410880782113,
+      "grad_norm": 0.9350283145904541,
+      "learning_rate": 0.00019641796253013958,
+      "loss": 3.807,
+      "step": 946
+    },
+    {
+      "epoch": 0.08572463112157147,
+      "grad_norm": 0.9795020222663879,
+      "learning_rate": 0.00019640954042341103,
+      "loss": 3.6858,
+      "step": 947
+    },
+    {
+      "epoch": 0.08581515343532181,
+      "grad_norm": 0.9330901503562927,
+      "learning_rate": 0.0001964011086082444,
+      "loss": 3.8028,
+      "step": 948
+    },
+    {
+      "epoch": 0.08590567574907215,
+      "grad_norm": 0.9229695200920105,
+      "learning_rate": 0.00019639266708548878,
+      "loss": 3.6575,
+      "step": 949
+    },
+    {
+      "epoch": 0.08599619806282248,
+      "grad_norm": 0.99271160364151,
+      "learning_rate": 0.00019638421585599423,
+      "loss": 3.7792,
+      "step": 950
+    },
+    {
+      "epoch": 0.08608672037657282,
+      "grad_norm": 0.9463202357292175,
+      "learning_rate": 0.00019637575492061174,
+      "loss": 3.8157,
+      "step": 951
+    },
+    {
+      "epoch": 0.08617724269032316,
+      "grad_norm": 1.0383622646331787,
+      "learning_rate": 0.0001963672842801934,
+      "loss": 3.6997,
+      "step": 952
+    },
+    {
+      "epoch": 0.0862677650040735,
+      "grad_norm": 1.0432873964309692,
+      "learning_rate": 0.0001963588039355922,
+      "loss": 3.7557,
+      "step": 953
+    },
+    {
+      "epoch": 0.08635828731782384,
+      "grad_norm": 0.914127767086029,
+      "learning_rate": 0.00019635031388766204,
+      "loss": 3.7724,
+      "step": 954
+    },
+    {
+      "epoch": 0.08644880963157418,
+      "grad_norm": 0.897434651851654,
+      "learning_rate": 0.0001963418141372579,
+      "loss": 3.7032,
+      "step": 955
+    },
+    {
+      "epoch": 0.08653933194532452,
+      "grad_norm": 1.049588918685913,
+      "learning_rate": 0.00019633330468523573,
+      "loss": 3.7061,
+      "step": 956
+    },
+    {
+      "epoch": 0.08662985425907486,
+      "grad_norm": 0.9720751047134399,
+      "learning_rate": 0.00019632478553245243,
+      "loss": 3.7529,
+      "step": 957
+    },
+    {
+      "epoch": 0.0867203765728252,
+      "grad_norm": 0.8728967905044556,
+      "learning_rate": 0.00019631625667976583,
+      "loss": 3.6846,
+      "step": 958
+    },
+    {
+      "epoch": 0.08681089888657555,
+      "grad_norm": 0.9369397759437561,
+      "learning_rate": 0.00019630771812803482,
+      "loss": 3.7338,
+      "step": 959
+    },
+    {
+      "epoch": 0.08690142120032587,
+      "grad_norm": 0.9343727231025696,
+      "learning_rate": 0.00019629916987811926,
+      "loss": 3.7064,
+      "step": 960
+    },
+    {
+      "epoch": 0.08699194351407621,
+      "grad_norm": 0.8479828238487244,
+      "learning_rate": 0.0001962906119308799,
+      "loss": 3.6929,
+      "step": 961
+    },
+    {
+      "epoch": 0.08708246582782656,
+      "grad_norm": 0.8664740920066833,
+      "learning_rate": 0.00019628204428717857,
+      "loss": 3.7699,
+      "step": 962
+    },
+    {
+      "epoch": 0.0871729881415769,
+      "grad_norm": 0.9612951278686523,
+      "learning_rate": 0.00019627346694787797,
+      "loss": 3.7328,
+      "step": 963
+    },
+    {
+      "epoch": 0.08726351045532724,
+      "grad_norm": 0.9078434109687805,
+      "learning_rate": 0.00019626487991384196,
+      "loss": 3.7624,
+      "step": 964
+    },
+    {
+      "epoch": 0.08735403276907758,
+      "grad_norm": 1.0075331926345825,
+      "learning_rate": 0.00019625628318593514,
+      "loss": 3.7806,
+      "step": 965
+    },
+    {
+      "epoch": 0.08744455508282792,
+      "grad_norm": 0.8739338517189026,
+      "learning_rate": 0.0001962476767650232,
+      "loss": 3.7445,
+      "step": 966
+    },
+    {
+      "epoch": 0.08753507739657826,
+      "grad_norm": 0.9157695174217224,
+      "learning_rate": 0.00019623906065197288,
+      "loss": 3.7081,
+      "step": 967
+    },
+    {
+      "epoch": 0.0876255997103286,
+      "grad_norm": 0.9016348719596863,
+      "learning_rate": 0.0001962304348476518,
+      "loss": 3.7984,
+      "step": 968
+    },
+    {
+      "epoch": 0.08771612202407894,
+      "grad_norm": 0.9907736778259277,
+      "learning_rate": 0.00019622179935292855,
+      "loss": 3.7199,
+      "step": 969
+    },
+    {
+      "epoch": 0.08780664433782928,
+      "grad_norm": 0.8860751390457153,
+      "learning_rate": 0.00019621315416867274,
+      "loss": 3.6944,
+      "step": 970
+    },
+    {
+      "epoch": 0.08789716665157961,
+      "grad_norm": 0.8661206960678101,
+      "learning_rate": 0.00019620449929575492,
+      "loss": 3.6799,
+      "step": 971
+    },
+    {
+      "epoch": 0.08798768896532995,
+      "grad_norm": 0.8809290528297424,
+      "learning_rate": 0.00019619583473504668,
+      "loss": 3.7001,
+      "step": 972
+    },
+    {
+      "epoch": 0.08807821127908029,
+      "grad_norm": 0.8480473160743713,
+      "learning_rate": 0.00019618716048742048,
+      "loss": 3.7303,
+      "step": 973
+    },
+    {
+      "epoch": 0.08816873359283063,
+      "grad_norm": 0.9732685685157776,
+      "learning_rate": 0.0001961784765537499,
+      "loss": 3.7068,
+      "step": 974
+    },
+    {
+      "epoch": 0.08825925590658097,
+      "grad_norm": 0.9032775163650513,
+      "learning_rate": 0.0001961697829349093,
+      "loss": 3.6811,
+      "step": 975
+    },
+    {
+      "epoch": 0.08834977822033131,
+      "grad_norm": 0.9440922141075134,
+      "learning_rate": 0.00019616107963177425,
+      "loss": 3.7028,
+      "step": 976
+    },
+    {
+      "epoch": 0.08844030053408165,
+      "grad_norm": 0.967886745929718,
+      "learning_rate": 0.00019615236664522105,
+      "loss": 3.6888,
+      "step": 977
+    },
+    {
+      "epoch": 0.088530822847832,
+      "grad_norm": 0.9971091151237488,
+      "learning_rate": 0.00019614364397612722,
+      "loss": 3.7256,
+      "step": 978
+    },
+    {
+      "epoch": 0.08862134516158233,
+      "grad_norm": 0.9110148549079895,
+      "learning_rate": 0.00019613491162537105,
+      "loss": 3.7195,
+      "step": 979
+    },
+    {
+      "epoch": 0.08871186747533268,
+      "grad_norm": 0.9255290627479553,
+      "learning_rate": 0.0001961261695938319,
+      "loss": 3.7545,
+      "step": 980
+    },
+    {
+      "epoch": 0.088802389789083,
+      "grad_norm": 0.9252705574035645,
+      "learning_rate": 0.00019611741788239012,
+      "loss": 3.7555,
+      "step": 981
+    },
+    {
+      "epoch": 0.08889291210283334,
+      "grad_norm": 0.9513722062110901,
+      "learning_rate": 0.00019610865649192697,
+      "loss": 3.7356,
+      "step": 982
+    },
+    {
+      "epoch": 0.08898343441658368,
+      "grad_norm": 0.9082736372947693,
+      "learning_rate": 0.00019609988542332473,
+      "loss": 3.7165,
+      "step": 983
+    },
+    {
+      "epoch": 0.08907395673033403,
+      "grad_norm": 0.9556236267089844,
+      "learning_rate": 0.00019609110467746666,
+      "loss": 3.6943,
+      "step": 984
+    },
+    {
+      "epoch": 0.08916447904408437,
+      "grad_norm": 0.9560606479644775,
+      "learning_rate": 0.00019608231425523702,
+      "loss": 3.7364,
+      "step": 985
+    },
+    {
+      "epoch": 0.0892550013578347,
+      "grad_norm": 0.9406256675720215,
+      "learning_rate": 0.00019607351415752095,
+      "loss": 3.6523,
+      "step": 986
+    },
+    {
+      "epoch": 0.08934552367158505,
+      "grad_norm": 0.8736352324485779,
+      "learning_rate": 0.00019606470438520462,
+      "loss": 3.6718,
+      "step": 987
+    },
+    {
+      "epoch": 0.08943604598533539,
+      "grad_norm": 0.9579623341560364,
+      "learning_rate": 0.00019605588493917518,
+      "loss": 3.6735,
+      "step": 988
+    },
+    {
+      "epoch": 0.08952656829908573,
+      "grad_norm": 1.038737177848816,
+      "learning_rate": 0.00019604705582032078,
+      "loss": 3.7648,
+      "step": 989
+    },
+    {
+      "epoch": 0.08961709061283607,
+      "grad_norm": 0.8953343033790588,
+      "learning_rate": 0.00019603821702953046,
+      "loss": 3.7876,
+      "step": 990
+    },
+    {
+      "epoch": 0.0897076129265864,
+      "grad_norm": 0.9059159159660339,
+      "learning_rate": 0.0001960293685676943,
+      "loss": 3.7621,
+      "step": 991
+    },
+    {
+      "epoch": 0.08979813524033674,
+      "grad_norm": 0.9297139644622803,
+      "learning_rate": 0.0001960205104357034,
+      "loss": 3.7517,
+      "step": 992
+    },
+    {
+      "epoch": 0.08988865755408708,
+      "grad_norm": 1.069353461265564,
+      "learning_rate": 0.0001960116426344497,
+      "loss": 3.6651,
+      "step": 993
+    },
+    {
+      "epoch": 0.08997917986783742,
+      "grad_norm": 0.9858116507530212,
+      "learning_rate": 0.00019600276516482622,
+      "loss": 3.7341,
+      "step": 994
+    },
+    {
+      "epoch": 0.09006970218158776,
+      "grad_norm": 0.93048495054245,
+      "learning_rate": 0.0001959938780277269,
+      "loss": 3.7052,
+      "step": 995
+    },
+    {
+      "epoch": 0.0901602244953381,
+      "grad_norm": 0.8646350502967834,
+      "learning_rate": 0.00019598498122404672,
+      "loss": 3.6858,
+      "step": 996
+    },
+    {
+      "epoch": 0.09025074680908844,
+      "grad_norm": 0.9945192337036133,
+      "learning_rate": 0.00019597607475468153,
+      "loss": 3.6827,
+      "step": 997
+    },
+    {
+      "epoch": 0.09034126912283878,
+      "grad_norm": 0.9069668054580688,
+      "learning_rate": 0.00019596715862052824,
+      "loss": 3.7084,
+      "step": 998
+    },
+    {
+      "epoch": 0.09043179143658912,
+      "grad_norm": 0.8991177082061768,
+      "learning_rate": 0.00019595823282248472,
+      "loss": 3.6903,
+      "step": 999
+    },
+    {
+      "epoch": 0.09052231375033946,
+      "grad_norm": 0.8807759284973145,
+      "learning_rate": 0.00019594929736144976,
+      "loss": 3.6824,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09052231375033946,
+      "eval_loss": 3.6549718379974365,
+      "eval_runtime": 71.8769,
+      "eval_samples_per_second": 37.606,
+      "eval_steps_per_second": 3.144,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0906128360640898,
+      "grad_norm": 0.952815055847168,
+      "learning_rate": 0.0001959403522383232,
+      "loss": 3.6511,
+      "step": 1001
+    },
+    {
+      "epoch": 0.09070335837784013,
+      "grad_norm": 0.9049227237701416,
+      "learning_rate": 0.00019593139745400576,
+      "loss": 3.6608,
+      "step": 1002
+    },
+    {
+      "epoch": 0.09079388069159047,
+      "grad_norm": 0.9202782511711121,
+      "learning_rate": 0.00019592243300939926,
+      "loss": 3.7317,
+      "step": 1003
+    },
+    {
+      "epoch": 0.09088440300534081,
+      "grad_norm": 0.9427291750907898,
+      "learning_rate": 0.00019591345890540636,
+      "loss": 3.7335,
+      "step": 1004
+    },
+    {
+      "epoch": 0.09097492531909115,
+      "grad_norm": 0.8800294399261475,
+      "learning_rate": 0.00019590447514293078,
+      "loss": 3.6614,
+      "step": 1005
+    },
+    {
+      "epoch": 0.0910654476328415,
+      "grad_norm": 0.9174665212631226,
+      "learning_rate": 0.00019589548172287719,
+      "loss": 3.6391,
+      "step": 1006
+    },
+    {
+      "epoch": 0.09115596994659184,
+      "grad_norm": 0.8996450901031494,
+      "learning_rate": 0.00019588647864615118,
+      "loss": 3.7422,
+      "step": 1007
+    },
+    {
+      "epoch": 0.09124649226034218,
+      "grad_norm": 0.9029880166053772,
+      "learning_rate": 0.00019587746591365941,
+      "loss": 3.6324,
+      "step": 1008
+    },
+    {
+      "epoch": 0.09133701457409252,
+      "grad_norm": 0.9739851355552673,
+      "learning_rate": 0.00019586844352630943,
+      "loss": 3.7409,
+      "step": 1009
+    },
+    {
+      "epoch": 0.09142753688784286,
+      "grad_norm": 0.9311057925224304,
+      "learning_rate": 0.00019585941148500985,
+      "loss": 3.7517,
+      "step": 1010
+    },
+    {
+      "epoch": 0.0915180592015932,
+      "grad_norm": 0.9630916714668274,
+      "learning_rate": 0.00019585036979067013,
+      "loss": 3.6949,
+      "step": 1011
+    },
+    {
+      "epoch": 0.09160858151534353,
+      "grad_norm": 1.0906858444213867,
+      "learning_rate": 0.00019584131844420082,
+      "loss": 3.644,
+      "step": 1012
+    },
+    {
+      "epoch": 0.09169910382909387,
+      "grad_norm": 1.0149617195129395,
+      "learning_rate": 0.00019583225744651333,
+      "loss": 3.6535,
+      "step": 1013
+    },
+    {
+      "epoch": 0.09178962614284421,
+      "grad_norm": 1.0162272453308105,
+      "learning_rate": 0.00019582318679852017,
+      "loss": 3.7233,
+      "step": 1014
+    },
+    {
+      "epoch": 0.09188014845659455,
+      "grad_norm": 0.9386601448059082,
+      "learning_rate": 0.0001958141065011347,
+      "loss": 3.7117,
+      "step": 1015
+    },
+    {
+      "epoch": 0.09197067077034489,
+      "grad_norm": 1.0775753259658813,
+      "learning_rate": 0.00019580501655527133,
+      "loss": 3.7124,
+      "step": 1016
+    },
+    {
+      "epoch": 0.09206119308409523,
+      "grad_norm": 1.0624293088912964,
+      "learning_rate": 0.0001957959169618454,
+      "loss": 3.7581,
+      "step": 1017
+    },
+    {
+      "epoch": 0.09215171539784557,
+      "grad_norm": 0.9643511176109314,
+      "learning_rate": 0.00019578680772177327,
+      "loss": 3.6738,
+      "step": 1018
+    },
+    {
+      "epoch": 0.09224223771159591,
+      "grad_norm": 0.9284974336624146,
+      "learning_rate": 0.00019577768883597224,
+      "loss": 3.7531,
+      "step": 1019
+    },
+    {
+      "epoch": 0.09233276002534625,
+      "grad_norm": 0.9831129908561707,
+      "learning_rate": 0.00019576856030536054,
+      "loss": 3.738,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0924232823390966,
+      "grad_norm": 1.0086435079574585,
+      "learning_rate": 0.00019575942213085744,
+      "loss": 3.6976,
+      "step": 1021
+    },
+    {
+      "epoch": 0.09251380465284692,
+      "grad_norm": 1.039759874343872,
+      "learning_rate": 0.00019575027431338314,
+      "loss": 3.7301,
+      "step": 1022
+    },
+    {
+      "epoch": 0.09260432696659726,
+      "grad_norm": 1.0100873708724976,
+      "learning_rate": 0.00019574111685385887,
+      "loss": 3.6884,
+      "step": 1023
+    },
+    {
+      "epoch": 0.0926948492803476,
+      "grad_norm": 1.0146422386169434,
+      "learning_rate": 0.00019573194975320673,
+      "loss": 3.6353,
+      "step": 1024
+    },
+    {
+      "epoch": 0.09278537159409794,
+      "grad_norm": 1.0001540184020996,
+      "learning_rate": 0.00019572277301234986,
+      "loss": 3.7716,
+      "step": 1025
+    },
+    {
+      "epoch": 0.09287589390784828,
+      "grad_norm": 0.8770668506622314,
+      "learning_rate": 0.0001957135866322124,
+      "loss": 3.7044,
+      "step": 1026
+    },
+    {
+      "epoch": 0.09296641622159862,
+      "grad_norm": 0.9398157000541687,
+      "learning_rate": 0.00019570439061371937,
+      "loss": 3.7006,
+      "step": 1027
+    },
+    {
+      "epoch": 0.09305693853534897,
+      "grad_norm": 0.9291334748268127,
+      "learning_rate": 0.00019569518495779682,
+      "loss": 3.6211,
+      "step": 1028
+    },
+    {
+      "epoch": 0.0931474608490993,
+      "grad_norm": 0.9487553238868713,
+      "learning_rate": 0.00019568596966537175,
+      "loss": 3.7016,
+      "step": 1029
+    },
+    {
+      "epoch": 0.09323798316284965,
+      "grad_norm": 0.9963234066963196,
+      "learning_rate": 0.00019567674473737218,
+      "loss": 3.6843,
+      "step": 1030
+    },
+    {
+      "epoch": 0.09332850547659999,
+      "grad_norm": 0.9393078088760376,
+      "learning_rate": 0.00019566751017472702,
+      "loss": 3.7114,
+      "step": 1031
+    },
+    {
+      "epoch": 0.09341902779035031,
+      "grad_norm": 0.978520393371582,
+      "learning_rate": 0.0001956582659783662,
+      "loss": 3.6871,
+      "step": 1032
+    },
+    {
+      "epoch": 0.09350955010410066,
+      "grad_norm": 1.07383131980896,
+      "learning_rate": 0.00019564901214922062,
+      "loss": 3.7456,
+      "step": 1033
+    },
+    {
+      "epoch": 0.093600072417851,
+      "grad_norm": 0.89312744140625,
+      "learning_rate": 0.00019563974868822212,
+      "loss": 3.7178,
+      "step": 1034
+    },
+    {
+      "epoch": 0.09369059473160134,
+      "grad_norm": 0.9418454766273499,
+      "learning_rate": 0.00019563047559630357,
+      "loss": 3.6676,
+      "step": 1035
+    },
+    {
+      "epoch": 0.09378111704535168,
+      "grad_norm": 0.9237220287322998,
+      "learning_rate": 0.00019562119287439873,
+      "loss": 3.6844,
+      "step": 1036
+    },
+    {
+      "epoch": 0.09387163935910202,
+      "grad_norm": 0.8427319526672363,
+      "learning_rate": 0.00019561190052344238,
+      "loss": 3.6416,
+      "step": 1037
+    },
+    {
+      "epoch": 0.09396216167285236,
+      "grad_norm": 0.9645209312438965,
+      "learning_rate": 0.00019560259854437025,
+      "loss": 3.6922,
+      "step": 1038
+    },
+    {
+      "epoch": 0.0940526839866027,
+      "grad_norm": 0.9387515783309937,
+      "learning_rate": 0.00019559328693811908,
+      "loss": 3.6764,
+      "step": 1039
+    },
+    {
+      "epoch": 0.09414320630035304,
+      "grad_norm": 1.045303225517273,
+      "learning_rate": 0.0001955839657056265,
+      "loss": 3.6943,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09423372861410338,
+      "grad_norm": 0.9981943964958191,
+      "learning_rate": 0.00019557463484783124,
+      "loss": 3.6776,
+      "step": 1041
+    },
+    {
+      "epoch": 0.09432425092785372,
+      "grad_norm": 0.9997328519821167,
+      "learning_rate": 0.00019556529436567287,
+      "loss": 3.6776,
+      "step": 1042
+    },
+    {
+      "epoch": 0.09441477324160405,
+      "grad_norm": 0.9236479997634888,
+      "learning_rate": 0.00019555594426009192,
+      "loss": 3.7411,
+      "step": 1043
+    },
+    {
+      "epoch": 0.09450529555535439,
+      "grad_norm": 0.9315744638442993,
+      "learning_rate": 0.00019554658453203004,
+      "loss": 3.6653,
+      "step": 1044
+    },
+    {
+      "epoch": 0.09459581786910473,
+      "grad_norm": 0.9356442093849182,
+      "learning_rate": 0.00019553721518242968,
+      "loss": 3.736,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09468634018285507,
+      "grad_norm": 0.9334856271743774,
+      "learning_rate": 0.00019552783621223436,
+      "loss": 3.7031,
+      "step": 1046
+    },
+    {
+      "epoch": 0.09477686249660541,
+      "grad_norm": 1.04819655418396,
+      "learning_rate": 0.00019551844762238858,
+      "loss": 3.6478,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09486738481035575,
+      "grad_norm": 1.0170056819915771,
+      "learning_rate": 0.00019550904941383773,
+      "loss": 3.6761,
+      "step": 1048
+    },
+    {
+      "epoch": 0.0949579071241061,
+      "grad_norm": 0.9866304397583008,
+      "learning_rate": 0.00019549964158752822,
+      "loss": 3.6145,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09504842943785644,
+      "grad_norm": 0.9720956087112427,
+      "learning_rate": 0.0001954902241444074,
+      "loss": 3.6766,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09513895175160678,
+      "grad_norm": 0.9859123229980469,
+      "learning_rate": 0.00019548079708542364,
+      "loss": 3.7507,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09522947406535712,
+      "grad_norm": 1.0696841478347778,
+      "learning_rate": 0.0001954713604115262,
+      "loss": 3.7064,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09531999637910744,
+      "grad_norm": 1.0893000364303589,
+      "learning_rate": 0.00019546191412366543,
+      "loss": 3.7356,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09541051869285778,
+      "grad_norm": 0.9062834978103638,
+      "learning_rate": 0.00019545245822279243,
+      "loss": 3.7005,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09550104100660813,
+      "grad_norm": 1.0514686107635498,
+      "learning_rate": 0.00019544299270985956,
+      "loss": 3.7077,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09559156332035847,
+      "grad_norm": 0.9726833701133728,
+      "learning_rate": 0.00019543351758581994,
+      "loss": 3.6598,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09568208563410881,
+      "grad_norm": 0.9796419143676758,
+      "learning_rate": 0.0001954240328516277,
+      "loss": 3.5648,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09577260794785915,
+      "grad_norm": 0.9635905623435974,
+      "learning_rate": 0.00019541453850823793,
+      "loss": 3.669,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09586313026160949,
+      "grad_norm": 0.8129740953445435,
+      "learning_rate": 0.0001954050345566068,
+      "loss": 3.6587,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09595365257535983,
+      "grad_norm": 0.9021315574645996,
+      "learning_rate": 0.00019539552099769126,
+      "loss": 3.6621,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09604417488911017,
+      "grad_norm": 0.948771595954895,
+      "learning_rate": 0.00019538599783244939,
+      "loss": 3.6992,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09613469720286051,
+      "grad_norm": 0.9002098441123962,
+      "learning_rate": 0.00019537646506184015,
+      "loss": 3.6613,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09622521951661084,
+      "grad_norm": 0.9096206426620483,
+      "learning_rate": 0.00019536692268682347,
+      "loss": 3.7313,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09631574183036118,
+      "grad_norm": 0.9365320205688477,
+      "learning_rate": 0.00019535737070836028,
+      "loss": 3.6228,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09640626414411152,
+      "grad_norm": 0.9058887362480164,
+      "learning_rate": 0.0001953478091274125,
+      "loss": 3.703,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09649678645786186,
+      "grad_norm": 0.8754643797874451,
+      "learning_rate": 0.00019533823794494292,
+      "loss": 3.643,
+      "step": 1066
+    },
+    {
+      "epoch": 0.0965873087716122,
+      "grad_norm": 0.9260498881340027,
+      "learning_rate": 0.0001953286571619154,
+      "loss": 3.6299,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09667783108536254,
+      "grad_norm": 0.9123169779777527,
+      "learning_rate": 0.0001953190667792947,
+      "loss": 3.7424,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09676835339911288,
+      "grad_norm": 0.9547523856163025,
+      "learning_rate": 0.0001953094667980466,
+      "loss": 3.6272,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09685887571286322,
+      "grad_norm": 0.8996102809906006,
+      "learning_rate": 0.00019529985721913778,
+      "loss": 3.6572,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09694939802661356,
+      "grad_norm": 0.9247338771820068,
+      "learning_rate": 0.00019529023804353596,
+      "loss": 3.7018,
+      "step": 1071
+    },
+    {
+      "epoch": 0.0970399203403639,
+      "grad_norm": 0.9439358711242676,
+      "learning_rate": 0.0001952806092722098,
+      "loss": 3.641,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09713044265411423,
+      "grad_norm": 0.9879926443099976,
+      "learning_rate": 0.00019527097090612887,
+      "loss": 3.7011,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09722096496786457,
+      "grad_norm": 0.8817611932754517,
+      "learning_rate": 0.0001952613229462638,
+      "loss": 3.6628,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09731148728161491,
+      "grad_norm": 0.9118003845214844,
+      "learning_rate": 0.00019525166539358606,
+      "loss": 3.6269,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09740200959536525,
+      "grad_norm": 0.9752653241157532,
+      "learning_rate": 0.00019524199824906826,
+      "loss": 3.6148,
+      "step": 1076
+    },
+    {
+      "epoch": 0.0974925319091156,
+      "grad_norm": 0.8902285099029541,
+      "learning_rate": 0.00019523232151368383,
+      "loss": 3.6231,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09758305422286594,
+      "grad_norm": 1.024625301361084,
+      "learning_rate": 0.0001952226351884072,
+      "loss": 3.7061,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09767357653661628,
+      "grad_norm": 0.9171874523162842,
+      "learning_rate": 0.00019521293927421388,
+      "loss": 3.7018,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09776409885036662,
+      "grad_norm": 0.9678434133529663,
+      "learning_rate": 0.00019520323377208017,
+      "loss": 3.6325,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09785462116411696,
+      "grad_norm": 0.923358142375946,
+      "learning_rate": 0.00019519351868298336,
+      "loss": 3.5903,
+      "step": 1081
+    },
+    {
+      "epoch": 0.0979451434778673,
+      "grad_norm": 0.8904416561126709,
+      "learning_rate": 0.00019518379400790189,
+      "loss": 3.6347,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09803566579161764,
+      "grad_norm": 0.9834840893745422,
+      "learning_rate": 0.00019517405974781493,
+      "loss": 3.7031,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09812618810536797,
+      "grad_norm": 0.9426645040512085,
+      "learning_rate": 0.00019516431590370278,
+      "loss": 3.675,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09821671041911831,
+      "grad_norm": 0.8758911490440369,
+      "learning_rate": 0.0001951545624765466,
+      "loss": 3.7114,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09830723273286865,
+      "grad_norm": 0.9593923687934875,
+      "learning_rate": 0.0001951447994673286,
+      "loss": 3.6828,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09839775504661899,
+      "grad_norm": 0.8801664113998413,
+      "learning_rate": 0.00019513502687703189,
+      "loss": 3.6278,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09848827736036933,
+      "grad_norm": 0.9123901128768921,
+      "learning_rate": 0.00019512524470664057,
+      "loss": 3.5551,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09857879967411967,
+      "grad_norm": 0.8368750810623169,
+      "learning_rate": 0.00019511545295713972,
+      "loss": 3.6159,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09866932198787001,
+      "grad_norm": 0.8892086148262024,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 3.669,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09875984430162035,
+      "grad_norm": 0.8537901639938354,
+      "learning_rate": 0.0001950958407247545,
+      "loss": 3.5952,
+      "step": 1091
+    },
+    {
+      "epoch": 0.0988503666153707,
+      "grad_norm": 0.8904717564582825,
+      "learning_rate": 0.00019508602024384506,
+      "loss": 3.7375,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09894088892912103,
+      "grad_norm": 0.8963021636009216,
+      "learning_rate": 0.00019507619018777596,
+      "loss": 3.6784,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09903141124287136,
+      "grad_norm": 0.9099161028862,
+      "learning_rate": 0.00019506635055753716,
+      "loss": 3.6208,
+      "step": 1094
+    },
+    {
+      "epoch": 0.0991219335566217,
+      "grad_norm": 0.9092404842376709,
+      "learning_rate": 0.00019505650135411942,
+      "loss": 3.6476,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09921245587037204,
+      "grad_norm": 0.9447798728942871,
+      "learning_rate": 0.0001950466425785146,
+      "loss": 3.636,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09930297818412238,
+      "grad_norm": 0.8991760611534119,
+      "learning_rate": 0.0001950367742317155,
+      "loss": 3.6676,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09939350049787272,
+      "grad_norm": 0.8840291500091553,
+      "learning_rate": 0.00019502689631471583,
+      "loss": 3.635,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09948402281162307,
+      "grad_norm": 0.9281356930732727,
+      "learning_rate": 0.0001950170088285103,
+      "loss": 3.612,
+      "step": 1099
+    },
+    {
+      "epoch": 0.0995745451253734,
+      "grad_norm": 0.8732534646987915,
+      "learning_rate": 0.00019500711177409454,
+      "loss": 3.6033,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09966506743912375,
+      "grad_norm": 0.9408723711967468,
+      "learning_rate": 0.00019499720515246525,
+      "loss": 3.599,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09975558975287409,
+      "grad_norm": 0.9561828970909119,
+      "learning_rate": 0.00019498728896462,
+      "loss": 3.6431,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09984611206662443,
+      "grad_norm": 0.8556196093559265,
+      "learning_rate": 0.00019497736321155736,
+      "loss": 3.6944,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09993663438037476,
+      "grad_norm": 0.895474910736084,
+      "learning_rate": 0.00019496742789427683,
+      "loss": 3.6039,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1000271566941251,
+      "grad_norm": 0.908662736415863,
+      "learning_rate": 0.00019495748301377895,
+      "loss": 3.5707,
+      "step": 1105
+    },
+    {
+      "epoch": 0.10011767900787544,
+      "grad_norm": 0.9285642504692078,
+      "learning_rate": 0.0001949475285710651,
+      "loss": 3.6242,
+      "step": 1106
+    },
+    {
+      "epoch": 0.10020820132162578,
+      "grad_norm": 0.8637263178825378,
+      "learning_rate": 0.0001949375645671377,
+      "loss": 3.7278,
+      "step": 1107
+    },
+    {
+      "epoch": 0.10029872363537612,
+      "grad_norm": 0.9085043668746948,
+      "learning_rate": 0.00019492759100300019,
+      "loss": 3.6507,
+      "step": 1108
+    },
+    {
+      "epoch": 0.10038924594912646,
+      "grad_norm": 0.8560227155685425,
+      "learning_rate": 0.0001949176078796568,
+      "loss": 3.5991,
+      "step": 1109
+    },
+    {
+      "epoch": 0.1004797682628768,
+      "grad_norm": 0.845382034778595,
+      "learning_rate": 0.00019490761519811293,
+      "loss": 3.6392,
+      "step": 1110
+    },
+    {
+      "epoch": 0.10057029057662714,
+      "grad_norm": 0.9317412376403809,
+      "learning_rate": 0.00019489761295937483,
+      "loss": 3.687,
+      "step": 1111
+    },
+    {
+      "epoch": 0.10066081289037748,
+      "grad_norm": 0.8906388878822327,
+      "learning_rate": 0.00019488760116444966,
+      "loss": 3.6313,
+      "step": 1112
+    },
+    {
+      "epoch": 0.10075133520412782,
+      "grad_norm": 0.8965174555778503,
+      "learning_rate": 0.00019487757981434566,
+      "loss": 3.6316,
+      "step": 1113
+    },
+    {
+      "epoch": 0.10084185751787815,
+      "grad_norm": 0.8226318359375,
+      "learning_rate": 0.00019486754891007198,
+      "loss": 3.6078,
+      "step": 1114
+    },
+    {
+      "epoch": 0.10093237983162849,
+      "grad_norm": 0.8582698702812195,
+      "learning_rate": 0.00019485750845263874,
+      "loss": 3.6597,
+      "step": 1115
+    },
+    {
+      "epoch": 0.10102290214537883,
+      "grad_norm": 0.8765236139297485,
+      "learning_rate": 0.00019484745844305695,
+      "loss": 3.5721,
+      "step": 1116
+    },
+    {
+      "epoch": 0.10111342445912917,
+      "grad_norm": 0.9732540249824524,
+      "learning_rate": 0.00019483739888233873,
+      "loss": 3.6898,
+      "step": 1117
+    },
+    {
+      "epoch": 0.10120394677287951,
+      "grad_norm": 0.8816714286804199,
+      "learning_rate": 0.000194827329771497,
+      "loss": 3.6785,
+      "step": 1118
+    },
+    {
+      "epoch": 0.10129446908662985,
+      "grad_norm": 1.0106911659240723,
+      "learning_rate": 0.00019481725111154577,
+      "loss": 3.6438,
+      "step": 1119
+    },
+    {
+      "epoch": 0.1013849914003802,
+      "grad_norm": 0.943163275718689,
+      "learning_rate": 0.00019480716290349995,
+      "loss": 3.6332,
+      "step": 1120
+    },
+    {
+      "epoch": 0.10147551371413054,
+      "grad_norm": 0.9958202242851257,
+      "learning_rate": 0.00019479706514837544,
+      "loss": 3.7277,
+      "step": 1121
+    },
+    {
+      "epoch": 0.10156603602788088,
+      "grad_norm": 0.985831618309021,
+      "learning_rate": 0.00019478695784718903,
+      "loss": 3.6826,
+      "step": 1122
+    },
+    {
+      "epoch": 0.10165655834163122,
+      "grad_norm": 0.9224241971969604,
+      "learning_rate": 0.0001947768410009586,
+      "loss": 3.6623,
+      "step": 1123
+    },
+    {
+      "epoch": 0.10174708065538156,
+      "grad_norm": 1.0890220403671265,
+      "learning_rate": 0.00019476671461070283,
+      "loss": 3.6679,
+      "step": 1124
+    },
+    {
+      "epoch": 0.10183760296913189,
+      "grad_norm": 0.901580274105072,
+      "learning_rate": 0.0001947565786774415,
+      "loss": 3.6733,
+      "step": 1125
+    },
+    {
+      "epoch": 0.10192812528288223,
+      "grad_norm": 1.0238951444625854,
+      "learning_rate": 0.00019474643320219532,
+      "loss": 3.705,
+      "step": 1126
+    },
+    {
+      "epoch": 0.10201864759663257,
+      "grad_norm": 0.9749259948730469,
+      "learning_rate": 0.00019473627818598587,
+      "loss": 3.6251,
+      "step": 1127
+    },
+    {
+      "epoch": 0.10210916991038291,
+      "grad_norm": 1.0571074485778809,
+      "learning_rate": 0.0001947261136298358,
+      "loss": 3.5952,
+      "step": 1128
+    },
+    {
+      "epoch": 0.10219969222413325,
+      "grad_norm": 1.082695722579956,
+      "learning_rate": 0.00019471593953476873,
+      "loss": 3.6309,
+      "step": 1129
+    },
+    {
+      "epoch": 0.10229021453788359,
+      "grad_norm": 1.031955361366272,
+      "learning_rate": 0.0001947057559018091,
+      "loss": 3.6201,
+      "step": 1130
+    },
+    {
+      "epoch": 0.10238073685163393,
+      "grad_norm": 0.928398072719574,
+      "learning_rate": 0.00019469556273198245,
+      "loss": 3.6031,
+      "step": 1131
+    },
+    {
+      "epoch": 0.10247125916538427,
+      "grad_norm": 0.9015419483184814,
+      "learning_rate": 0.0001946853600263152,
+      "loss": 3.6516,
+      "step": 1132
+    },
+    {
+      "epoch": 0.10256178147913461,
+      "grad_norm": 0.9696022272109985,
+      "learning_rate": 0.00019467514778583484,
+      "loss": 3.6614,
+      "step": 1133
+    },
+    {
+      "epoch": 0.10265230379288495,
+      "grad_norm": 0.9159278273582458,
+      "learning_rate": 0.00019466492601156966,
+      "loss": 3.6942,
+      "step": 1134
+    },
+    {
+      "epoch": 0.10274282610663528,
+      "grad_norm": 0.9715399742126465,
+      "learning_rate": 0.000194654694704549,
+      "loss": 3.6107,
+      "step": 1135
+    },
+    {
+      "epoch": 0.10283334842038562,
+      "grad_norm": 0.9022889137268066,
+      "learning_rate": 0.0001946444538658032,
+      "loss": 3.6543,
+      "step": 1136
+    },
+    {
+      "epoch": 0.10292387073413596,
+      "grad_norm": 0.9374133944511414,
+      "learning_rate": 0.00019463420349636348,
+      "loss": 3.6396,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1030143930478863,
+      "grad_norm": 0.9193755984306335,
+      "learning_rate": 0.00019462394359726206,
+      "loss": 3.6156,
+      "step": 1138
+    },
+    {
+      "epoch": 0.10310491536163664,
+      "grad_norm": 0.9619964361190796,
+      "learning_rate": 0.00019461367416953208,
+      "loss": 3.7139,
+      "step": 1139
+    },
+    {
+      "epoch": 0.10319543767538698,
+      "grad_norm": 0.9839112162590027,
+      "learning_rate": 0.00019460339521420772,
+      "loss": 3.6013,
+      "step": 1140
+    },
+    {
+      "epoch": 0.10328595998913732,
+      "grad_norm": 0.9973909258842468,
+      "learning_rate": 0.00019459310673232408,
+      "loss": 3.6059,
+      "step": 1141
+    },
+    {
+      "epoch": 0.10337648230288767,
+      "grad_norm": 0.8987120985984802,
+      "learning_rate": 0.00019458280872491714,
+      "loss": 3.6464,
+      "step": 1142
+    },
+    {
+      "epoch": 0.103467004616638,
+      "grad_norm": 1.0310770273208618,
+      "learning_rate": 0.00019457250119302394,
+      "loss": 3.6645,
+      "step": 1143
+    },
+    {
+      "epoch": 0.10355752693038835,
+      "grad_norm": 0.913856565952301,
+      "learning_rate": 0.0001945621841376825,
+      "loss": 3.6143,
+      "step": 1144
+    },
+    {
+      "epoch": 0.10364804924413867,
+      "grad_norm": 1.0820924043655396,
+      "learning_rate": 0.0001945518575599317,
+      "loss": 3.6924,
+      "step": 1145
+    },
+    {
+      "epoch": 0.10373857155788901,
+      "grad_norm": 0.9650067687034607,
+      "learning_rate": 0.00019454152146081141,
+      "loss": 3.6596,
+      "step": 1146
+    },
+    {
+      "epoch": 0.10382909387163936,
+      "grad_norm": 0.8672732710838318,
+      "learning_rate": 0.0001945311758413625,
+      "loss": 3.6083,
+      "step": 1147
+    },
+    {
+      "epoch": 0.1039196161853897,
+      "grad_norm": 1.0733089447021484,
+      "learning_rate": 0.0001945208207026268,
+      "loss": 3.7095,
+      "step": 1148
+    },
+    {
+      "epoch": 0.10401013849914004,
+      "grad_norm": 0.9187482595443726,
+      "learning_rate": 0.000194510456045647,
+      "loss": 3.6309,
+      "step": 1149
+    },
+    {
+      "epoch": 0.10410066081289038,
+      "grad_norm": 1.012186050415039,
+      "learning_rate": 0.00019450008187146684,
+      "loss": 3.6064,
+      "step": 1150
+    },
+    {
+      "epoch": 0.10419118312664072,
+      "grad_norm": 0.9701893329620361,
+      "learning_rate": 0.00019448969818113107,
+      "loss": 3.6653,
+      "step": 1151
+    },
+    {
+      "epoch": 0.10428170544039106,
+      "grad_norm": 0.9445899128913879,
+      "learning_rate": 0.00019447930497568528,
+      "loss": 3.6369,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1043722277541414,
+      "grad_norm": 0.9546499848365784,
+      "learning_rate": 0.00019446890225617604,
+      "loss": 3.6069,
+      "step": 1153
+    },
+    {
+      "epoch": 0.10446275006789174,
+      "grad_norm": 0.9554566144943237,
+      "learning_rate": 0.00019445849002365094,
+      "loss": 3.6144,
+      "step": 1154
+    },
+    {
+      "epoch": 0.10455327238164208,
+      "grad_norm": 0.8900935649871826,
+      "learning_rate": 0.00019444806827915846,
+      "loss": 3.6298,
+      "step": 1155
+    },
+    {
+      "epoch": 0.10464379469539241,
+      "grad_norm": 1.0609067678451538,
+      "learning_rate": 0.00019443763702374812,
+      "loss": 3.6561,
+      "step": 1156
+    },
+    {
+      "epoch": 0.10473431700914275,
+      "grad_norm": 0.8893911838531494,
+      "learning_rate": 0.0001944271962584703,
+      "loss": 3.56,
+      "step": 1157
+    },
+    {
+      "epoch": 0.10482483932289309,
+      "grad_norm": 1.1273530721664429,
+      "learning_rate": 0.0001944167459843764,
+      "loss": 3.6363,
+      "step": 1158
+    },
+    {
+      "epoch": 0.10491536163664343,
+      "grad_norm": 0.8835474848747253,
+      "learning_rate": 0.00019440628620251874,
+      "loss": 3.5974,
+      "step": 1159
+    },
+    {
+      "epoch": 0.10500588395039377,
+      "grad_norm": 0.9661706686019897,
+      "learning_rate": 0.00019439581691395067,
+      "loss": 3.685,
+      "step": 1160
+    },
+    {
+      "epoch": 0.10509640626414411,
+      "grad_norm": 0.9183404445648193,
+      "learning_rate": 0.00019438533811972643,
+      "loss": 3.6391,
+      "step": 1161
+    },
+    {
+      "epoch": 0.10518692857789445,
+      "grad_norm": 0.9571143984794617,
+      "learning_rate": 0.0001943748498209012,
+      "loss": 3.5659,
+      "step": 1162
+    },
+    {
+      "epoch": 0.1052774508916448,
+      "grad_norm": 0.9471140503883362,
+      "learning_rate": 0.0001943643520185312,
+      "loss": 3.6097,
+      "step": 1163
+    },
+    {
+      "epoch": 0.10536797320539514,
+      "grad_norm": 1.0031734704971313,
+      "learning_rate": 0.0001943538447136735,
+      "loss": 3.5766,
+      "step": 1164
+    },
+    {
+      "epoch": 0.10545849551914548,
+      "grad_norm": 0.9944052696228027,
+      "learning_rate": 0.00019434332790738625,
+      "loss": 3.6271,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1055490178328958,
+      "grad_norm": 0.9067218899726868,
+      "learning_rate": 0.00019433280160072847,
+      "loss": 3.6163,
+      "step": 1166
+    },
+    {
+      "epoch": 0.10563954014664614,
+      "grad_norm": 0.9322523474693298,
+      "learning_rate": 0.0001943222657947601,
+      "loss": 3.6253,
+      "step": 1167
+    },
+    {
+      "epoch": 0.10573006246039648,
+      "grad_norm": 1.0816961526870728,
+      "learning_rate": 0.0001943117204905422,
+      "loss": 3.6209,
+      "step": 1168
+    },
+    {
+      "epoch": 0.10582058477414683,
+      "grad_norm": 0.9161190986633301,
+      "learning_rate": 0.00019430116568913656,
+      "loss": 3.6201,
+      "step": 1169
+    },
+    {
+      "epoch": 0.10591110708789717,
+      "grad_norm": 1.024245023727417,
+      "learning_rate": 0.00019429060139160618,
+      "loss": 3.5892,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1060016294016475,
+      "grad_norm": 0.9940409660339355,
+      "learning_rate": 0.00019428002759901478,
+      "loss": 3.5953,
+      "step": 1171
+    },
+    {
+      "epoch": 0.10609215171539785,
+      "grad_norm": 1.0768675804138184,
+      "learning_rate": 0.0001942694443124272,
+      "loss": 3.6088,
+      "step": 1172
+    },
+    {
+      "epoch": 0.10618267402914819,
+      "grad_norm": 0.9526625275611877,
+      "learning_rate": 0.00019425885153290918,
+      "loss": 3.6164,
+      "step": 1173
+    },
+    {
+      "epoch": 0.10627319634289853,
+      "grad_norm": 1.0202966928482056,
+      "learning_rate": 0.00019424824926152735,
+      "loss": 3.6156,
+      "step": 1174
+    },
+    {
+      "epoch": 0.10636371865664887,
+      "grad_norm": 1.0075091123580933,
+      "learning_rate": 0.0001942376374993494,
+      "loss": 3.6492,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1064542409703992,
+      "grad_norm": 1.0465952157974243,
+      "learning_rate": 0.00019422701624744395,
+      "loss": 3.6663,
+      "step": 1176
+    },
+    {
+      "epoch": 0.10654476328414954,
+      "grad_norm": 0.9357006549835205,
+      "learning_rate": 0.00019421638550688055,
+      "loss": 3.5978,
+      "step": 1177
+    },
+    {
+      "epoch": 0.10663528559789988,
+      "grad_norm": 0.9215185642242432,
+      "learning_rate": 0.00019420574527872968,
+      "loss": 3.5934,
+      "step": 1178
+    },
+    {
+      "epoch": 0.10672580791165022,
+      "grad_norm": 1.01131272315979,
+      "learning_rate": 0.00019419509556406285,
+      "loss": 3.6031,
+      "step": 1179
+    },
+    {
+      "epoch": 0.10681633022540056,
+      "grad_norm": 0.9412872195243835,
+      "learning_rate": 0.00019418443636395248,
+      "loss": 3.6491,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1069068525391509,
+      "grad_norm": 0.9983108639717102,
+      "learning_rate": 0.00019417376767947192,
+      "loss": 3.7015,
+      "step": 1181
+    },
+    {
+      "epoch": 0.10699737485290124,
+      "grad_norm": 0.9529833793640137,
+      "learning_rate": 0.00019416308951169557,
+      "loss": 3.6213,
+      "step": 1182
+    },
+    {
+      "epoch": 0.10708789716665158,
+      "grad_norm": 1.0248433351516724,
+      "learning_rate": 0.00019415240186169866,
+      "loss": 3.6436,
+      "step": 1183
+    },
+    {
+      "epoch": 0.10717841948040192,
+      "grad_norm": 0.9561635255813599,
+      "learning_rate": 0.00019414170473055746,
+      "loss": 3.6078,
+      "step": 1184
+    },
+    {
+      "epoch": 0.10726894179415226,
+      "grad_norm": 0.8160126209259033,
+      "learning_rate": 0.00019413099811934917,
+      "loss": 3.6372,
+      "step": 1185
+    },
+    {
+      "epoch": 0.10735946410790259,
+      "grad_norm": 0.9989318251609802,
+      "learning_rate": 0.00019412028202915198,
+      "loss": 3.6229,
+      "step": 1186
+    },
+    {
+      "epoch": 0.10744998642165293,
+      "grad_norm": 0.8841098546981812,
+      "learning_rate": 0.00019410955646104492,
+      "loss": 3.5736,
+      "step": 1187
+    },
+    {
+      "epoch": 0.10754050873540327,
+      "grad_norm": 0.9628979563713074,
+      "learning_rate": 0.00019409882141610812,
+      "loss": 3.587,
+      "step": 1188
+    },
+    {
+      "epoch": 0.10763103104915361,
+      "grad_norm": 0.873678982257843,
+      "learning_rate": 0.00019408807689542257,
+      "loss": 3.6213,
+      "step": 1189
+    },
+    {
+      "epoch": 0.10772155336290395,
+      "grad_norm": 0.8769716024398804,
+      "learning_rate": 0.00019407732290007023,
+      "loss": 3.6219,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1078120756766543,
+      "grad_norm": 0.9059527516365051,
+      "learning_rate": 0.00019406655943113408,
+      "loss": 3.6575,
+      "step": 1191
+    },
+    {
+      "epoch": 0.10790259799040464,
+      "grad_norm": 0.878751277923584,
+      "learning_rate": 0.00019405578648969796,
+      "loss": 3.6348,
+      "step": 1192
+    },
+    {
+      "epoch": 0.10799312030415498,
+      "grad_norm": 1.0593430995941162,
+      "learning_rate": 0.00019404500407684672,
+      "loss": 3.5751,
+      "step": 1193
+    },
+    {
+      "epoch": 0.10808364261790532,
+      "grad_norm": 0.9103216528892517,
+      "learning_rate": 0.00019403421219366611,
+      "loss": 3.577,
+      "step": 1194
+    },
+    {
+      "epoch": 0.10817416493165566,
+      "grad_norm": 1.0272597074508667,
+      "learning_rate": 0.00019402341084124295,
+      "loss": 3.6078,
+      "step": 1195
+    },
+    {
+      "epoch": 0.108264687245406,
+      "grad_norm": 0.8711608648300171,
+      "learning_rate": 0.00019401260002066488,
+      "loss": 3.6503,
+      "step": 1196
+    },
+    {
+      "epoch": 0.10835520955915633,
+      "grad_norm": 1.018958330154419,
+      "learning_rate": 0.00019400177973302054,
+      "loss": 3.6311,
+      "step": 1197
+    },
+    {
+      "epoch": 0.10844573187290667,
+      "grad_norm": 0.8852933049201965,
+      "learning_rate": 0.00019399094997939957,
+      "loss": 3.67,
+      "step": 1198
+    },
+    {
+      "epoch": 0.10853625418665701,
+      "grad_norm": 0.939679741859436,
+      "learning_rate": 0.00019398011076089252,
+      "loss": 3.5849,
+      "step": 1199
+    },
+    {
+      "epoch": 0.10862677650040735,
+      "grad_norm": 0.8842020630836487,
+      "learning_rate": 0.00019396926207859084,
+      "loss": 3.6535,
+      "step": 1200
+    },
+    {
+      "epoch": 0.10871729881415769,
+      "grad_norm": 0.9720060229301453,
+      "learning_rate": 0.00019395840393358706,
+      "loss": 3.636,
+      "step": 1201
+    },
+    {
+      "epoch": 0.10880782112790803,
+      "grad_norm": 0.930729329586029,
+      "learning_rate": 0.00019394753632697462,
+      "loss": 3.5714,
+      "step": 1202
+    },
+    {
+      "epoch": 0.10889834344165837,
+      "grad_norm": 0.9698287844657898,
+      "learning_rate": 0.0001939366592598478,
+      "loss": 3.5511,
+      "step": 1203
+    },
+    {
+      "epoch": 0.10898886575540871,
+      "grad_norm": 0.9683051109313965,
+      "learning_rate": 0.00019392577273330197,
+      "loss": 3.6155,
+      "step": 1204
+    },
+    {
+      "epoch": 0.10907938806915905,
+      "grad_norm": 0.9001908302307129,
+      "learning_rate": 0.00019391487674843338,
+      "loss": 3.5123,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1091699103829094,
+      "grad_norm": 0.9388008713722229,
+      "learning_rate": 0.0001939039713063393,
+      "loss": 3.56,
+      "step": 1206
+    },
+    {
+      "epoch": 0.10926043269665972,
+      "grad_norm": 1.036388635635376,
+      "learning_rate": 0.0001938930564081179,
+      "loss": 3.5918,
+      "step": 1207
+    },
+    {
+      "epoch": 0.10935095501041006,
+      "grad_norm": 1.0771197080612183,
+      "learning_rate": 0.00019388213205486822,
+      "loss": 3.6138,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1094414773241604,
+      "grad_norm": 0.922707736492157,
+      "learning_rate": 0.00019387119824769045,
+      "loss": 3.5791,
+      "step": 1209
+    },
+    {
+      "epoch": 0.10953199963791074,
+      "grad_norm": 1.0649242401123047,
+      "learning_rate": 0.00019386025498768558,
+      "loss": 3.5486,
+      "step": 1210
+    },
+    {
+      "epoch": 0.10962252195166108,
+      "grad_norm": 0.9694133400917053,
+      "learning_rate": 0.0001938493022759556,
+      "loss": 3.5445,
+      "step": 1211
+    },
+    {
+      "epoch": 0.10971304426541142,
+      "grad_norm": 0.9994919896125793,
+      "learning_rate": 0.00019383834011360346,
+      "loss": 3.577,
+      "step": 1212
+    },
+    {
+      "epoch": 0.10980356657916177,
+      "grad_norm": 0.8897203803062439,
+      "learning_rate": 0.00019382736850173302,
+      "loss": 3.5406,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1098940888929121,
+      "grad_norm": 0.9498850107192993,
+      "learning_rate": 0.00019381638744144915,
+      "loss": 3.5708,
+      "step": 1214
+    },
+    {
+      "epoch": 0.10998461120666245,
+      "grad_norm": 0.9750722050666809,
+      "learning_rate": 0.00019380539693385763,
+      "loss": 3.6341,
+      "step": 1215
+    },
+    {
+      "epoch": 0.11007513352041279,
+      "grad_norm": 0.979321300983429,
+      "learning_rate": 0.0001937943969800652,
+      "loss": 3.6495,
+      "step": 1216
+    },
+    {
+      "epoch": 0.11016565583416311,
+      "grad_norm": 0.9199374914169312,
+      "learning_rate": 0.00019378338758117958,
+      "loss": 3.5429,
+      "step": 1217
+    },
+    {
+      "epoch": 0.11025617814791346,
+      "grad_norm": 0.9436404705047607,
+      "learning_rate": 0.00019377236873830938,
+      "loss": 3.5996,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1103467004616638,
+      "grad_norm": 0.9604935646057129,
+      "learning_rate": 0.00019376134045256423,
+      "loss": 3.6583,
+      "step": 1219
+    },
+    {
+      "epoch": 0.11043722277541414,
+      "grad_norm": 0.9841674566268921,
+      "learning_rate": 0.00019375030272505463,
+      "loss": 3.6077,
+      "step": 1220
+    },
+    {
+      "epoch": 0.11052774508916448,
+      "grad_norm": 0.919344425201416,
+      "learning_rate": 0.00019373925555689212,
+      "loss": 3.5181,
+      "step": 1221
+    },
+    {
+      "epoch": 0.11061826740291482,
+      "grad_norm": 0.9568752646446228,
+      "learning_rate": 0.00019372819894918915,
+      "loss": 3.5464,
+      "step": 1222
+    },
+    {
+      "epoch": 0.11070878971666516,
+      "grad_norm": 0.9117941856384277,
+      "learning_rate": 0.0001937171329030591,
+      "loss": 3.6367,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1107993120304155,
+      "grad_norm": 0.9444876909255981,
+      "learning_rate": 0.00019370605741961635,
+      "loss": 3.5982,
+      "step": 1224
+    },
+    {
+      "epoch": 0.11088983434416584,
+      "grad_norm": 1.0101433992385864,
+      "learning_rate": 0.0001936949724999762,
+      "loss": 3.5492,
+      "step": 1225
+    },
+    {
+      "epoch": 0.11098035665791618,
+      "grad_norm": 0.9878042340278625,
+      "learning_rate": 0.00019368387814525483,
+      "loss": 3.6045,
+      "step": 1226
+    },
+    {
+      "epoch": 0.11107087897166651,
+      "grad_norm": 0.9536324143409729,
+      "learning_rate": 0.00019367277435656952,
+      "loss": 3.5784,
+      "step": 1227
+    },
+    {
+      "epoch": 0.11116140128541685,
+      "grad_norm": 1.0027134418487549,
+      "learning_rate": 0.0001936616611350384,
+      "loss": 3.5433,
+      "step": 1228
+    },
+    {
+      "epoch": 0.11125192359916719,
+      "grad_norm": 0.9173308610916138,
+      "learning_rate": 0.00019365053848178056,
+      "loss": 3.6148,
+      "step": 1229
+    },
+    {
+      "epoch": 0.11134244591291753,
+      "grad_norm": 1.0498443841934204,
+      "learning_rate": 0.00019363940639791606,
+      "loss": 3.686,
+      "step": 1230
+    },
+    {
+      "epoch": 0.11143296822666787,
+      "grad_norm": 1.124347448348999,
+      "learning_rate": 0.00019362826488456588,
+      "loss": 3.613,
+      "step": 1231
+    },
+    {
+      "epoch": 0.11152349054041821,
+      "grad_norm": 0.9275368452072144,
+      "learning_rate": 0.00019361711394285202,
+      "loss": 3.6151,
+      "step": 1232
+    },
+    {
+      "epoch": 0.11161401285416855,
+      "grad_norm": 1.0720618963241577,
+      "learning_rate": 0.00019360595357389735,
+      "loss": 3.5815,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1117045351679189,
+      "grad_norm": 1.006804347038269,
+      "learning_rate": 0.00019359478377882567,
+      "loss": 3.6022,
+      "step": 1234
+    },
+    {
+      "epoch": 0.11179505748166924,
+      "grad_norm": 0.9970232248306274,
+      "learning_rate": 0.00019358360455876188,
+      "loss": 3.5799,
+      "step": 1235
+    },
+    {
+      "epoch": 0.11188557979541958,
+      "grad_norm": 0.9625207185745239,
+      "learning_rate": 0.00019357241591483164,
+      "loss": 3.5385,
+      "step": 1236
+    },
+    {
+      "epoch": 0.11197610210916992,
+      "grad_norm": 1.0641515254974365,
+      "learning_rate": 0.00019356121784816166,
+      "loss": 3.6034,
+      "step": 1237
+    },
+    {
+      "epoch": 0.11206662442292024,
+      "grad_norm": 0.9369845390319824,
+      "learning_rate": 0.00019355001035987966,
+      "loss": 3.5279,
+      "step": 1238
+    },
+    {
+      "epoch": 0.11215714673667058,
+      "grad_norm": 1.0742141008377075,
+      "learning_rate": 0.00019353879345111413,
+      "loss": 3.6396,
+      "step": 1239
+    },
+    {
+      "epoch": 0.11224766905042093,
+      "grad_norm": 1.0029163360595703,
+      "learning_rate": 0.00019352756712299468,
+      "loss": 3.5808,
+      "step": 1240
+    },
+    {
+      "epoch": 0.11233819136417127,
+      "grad_norm": 0.9248932003974915,
+      "learning_rate": 0.00019351633137665175,
+      "loss": 3.6924,
+      "step": 1241
+    },
+    {
+      "epoch": 0.11242871367792161,
+      "grad_norm": 0.9621601700782776,
+      "learning_rate": 0.00019350508621321683,
+      "loss": 3.6178,
+      "step": 1242
+    },
+    {
+      "epoch": 0.11251923599167195,
+      "grad_norm": 0.9053172469139099,
+      "learning_rate": 0.00019349383163382228,
+      "loss": 3.5506,
+      "step": 1243
+    },
+    {
+      "epoch": 0.11260975830542229,
+      "grad_norm": 0.9937805533409119,
+      "learning_rate": 0.00019348256763960145,
+      "loss": 3.6734,
+      "step": 1244
+    },
+    {
+      "epoch": 0.11270028061917263,
+      "grad_norm": 0.9606733918190002,
+      "learning_rate": 0.00019347129423168862,
+      "loss": 3.5821,
+      "step": 1245
+    },
+    {
+      "epoch": 0.11279080293292297,
+      "grad_norm": 0.8962657451629639,
+      "learning_rate": 0.000193460011411219,
+      "loss": 3.5363,
+      "step": 1246
+    },
+    {
+      "epoch": 0.11288132524667331,
+      "grad_norm": 0.9315593838691711,
+      "learning_rate": 0.00019344871917932884,
+      "loss": 3.5507,
+      "step": 1247
+    },
+    {
+      "epoch": 0.11297184756042364,
+      "grad_norm": 0.899619996547699,
+      "learning_rate": 0.00019343741753715516,
+      "loss": 3.4878,
+      "step": 1248
+    },
+    {
+      "epoch": 0.11306236987417398,
+      "grad_norm": 0.9632608294487,
+      "learning_rate": 0.00019342610648583611,
+      "loss": 3.593,
+      "step": 1249
+    },
+    {
+      "epoch": 0.11315289218792432,
+      "grad_norm": 0.879357635974884,
+      "learning_rate": 0.00019341478602651069,
+      "loss": 3.5508,
+      "step": 1250
+    },
+    {
+      "epoch": 0.11324341450167466,
+      "grad_norm": 0.9848529696464539,
+      "learning_rate": 0.00019340345616031886,
+      "loss": 3.5655,
+      "step": 1251
+    },
+    {
+      "epoch": 0.113333936815425,
+      "grad_norm": 0.8929630517959595,
+      "learning_rate": 0.00019339211688840157,
+      "loss": 3.5551,
+      "step": 1252
+    },
+    {
+      "epoch": 0.11342445912917534,
+      "grad_norm": 0.863125741481781,
+      "learning_rate": 0.00019338076821190066,
+      "loss": 3.6034,
+      "step": 1253
+    },
+    {
+      "epoch": 0.11351498144292568,
+      "grad_norm": 0.8554606437683105,
+      "learning_rate": 0.00019336941013195892,
+      "loss": 3.5429,
+      "step": 1254
+    },
+    {
+      "epoch": 0.11360550375667602,
+      "grad_norm": 1.0379880666732788,
+      "learning_rate": 0.00019335804264972018,
+      "loss": 3.6066,
+      "step": 1255
+    },
+    {
+      "epoch": 0.11369602607042636,
+      "grad_norm": 0.9469103217124939,
+      "learning_rate": 0.00019334666576632906,
+      "loss": 3.6111,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1137865483841767,
+      "grad_norm": 0.9332228899002075,
+      "learning_rate": 0.00019333527948293128,
+      "loss": 3.6062,
+      "step": 1257
+    },
+    {
+      "epoch": 0.11387707069792703,
+      "grad_norm": 0.8677222728729248,
+      "learning_rate": 0.0001933238838006734,
+      "loss": 3.579,
+      "step": 1258
+    },
+    {
+      "epoch": 0.11396759301167737,
+      "grad_norm": 1.0136536359786987,
+      "learning_rate": 0.000193312478720703,
+      "loss": 3.511,
+      "step": 1259
+    },
+    {
+      "epoch": 0.11405811532542771,
+      "grad_norm": 0.87325119972229,
+      "learning_rate": 0.00019330106424416852,
+      "loss": 3.5735,
+      "step": 1260
+    },
+    {
+      "epoch": 0.11414863763917805,
+      "grad_norm": 0.9630944132804871,
+      "learning_rate": 0.00019328964037221942,
+      "loss": 3.554,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1142391599529284,
+      "grad_norm": 0.946215808391571,
+      "learning_rate": 0.0001932782071060061,
+      "loss": 3.5589,
+      "step": 1262
+    },
+    {
+      "epoch": 0.11432968226667874,
+      "grad_norm": 0.9987102150917053,
+      "learning_rate": 0.00019326676444667986,
+      "loss": 3.5423,
+      "step": 1263
+    },
+    {
+      "epoch": 0.11442020458042908,
+      "grad_norm": 1.0245448350906372,
+      "learning_rate": 0.000193255312395393,
+      "loss": 3.5212,
+      "step": 1264
+    },
+    {
+      "epoch": 0.11451072689417942,
+      "grad_norm": 1.0864923000335693,
+      "learning_rate": 0.00019324385095329874,
+      "loss": 3.5383,
+      "step": 1265
+    },
+    {
+      "epoch": 0.11460124920792976,
+      "grad_norm": 0.9032666683197021,
+      "learning_rate": 0.00019323238012155123,
+      "loss": 3.5436,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1146917715216801,
+      "grad_norm": 0.9453108906745911,
+      "learning_rate": 0.0001932208999013056,
+      "loss": 3.5882,
+      "step": 1267
+    },
+    {
+      "epoch": 0.11478229383543043,
+      "grad_norm": 1.0362881422042847,
+      "learning_rate": 0.0001932094102937179,
+      "loss": 3.5696,
+      "step": 1268
+    },
+    {
+      "epoch": 0.11487281614918077,
+      "grad_norm": 1.04246187210083,
+      "learning_rate": 0.00019319791129994513,
+      "loss": 3.6289,
+      "step": 1269
+    },
+    {
+      "epoch": 0.11496333846293111,
+      "grad_norm": 0.9332128167152405,
+      "learning_rate": 0.00019318640292114524,
+      "loss": 3.5249,
+      "step": 1270
+    },
+    {
+      "epoch": 0.11505386077668145,
+      "grad_norm": 0.9713038802146912,
+      "learning_rate": 0.00019317488515847715,
+      "loss": 3.5822,
+      "step": 1271
+    },
+    {
+      "epoch": 0.11514438309043179,
+      "grad_norm": 0.96926349401474,
+      "learning_rate": 0.00019316335801310063,
+      "loss": 3.545,
+      "step": 1272
+    },
+    {
+      "epoch": 0.11523490540418213,
+      "grad_norm": 1.0021302700042725,
+      "learning_rate": 0.00019315182148617655,
+      "loss": 3.6455,
+      "step": 1273
+    },
+    {
+      "epoch": 0.11532542771793247,
+      "grad_norm": 0.9882180690765381,
+      "learning_rate": 0.00019314027557886657,
+      "loss": 3.4918,
+      "step": 1274
+    },
+    {
+      "epoch": 0.11541595003168281,
+      "grad_norm": 0.9814496040344238,
+      "learning_rate": 0.00019312872029233339,
+      "loss": 3.5967,
+      "step": 1275
+    },
+    {
+      "epoch": 0.11550647234543315,
+      "grad_norm": 0.9681494235992432,
+      "learning_rate": 0.00019311715562774062,
+      "loss": 3.5891,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1155969946591835,
+      "grad_norm": 1.08894681930542,
+      "learning_rate": 0.00019310558158625285,
+      "loss": 3.5718,
+      "step": 1277
+    },
+    {
+      "epoch": 0.11568751697293383,
+      "grad_norm": 0.9990349411964417,
+      "learning_rate": 0.00019309399816903552,
+      "loss": 3.5909,
+      "step": 1278
+    },
+    {
+      "epoch": 0.11577803928668416,
+      "grad_norm": 0.9717885851860046,
+      "learning_rate": 0.00019308240537725517,
+      "loss": 3.4866,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1158685616004345,
+      "grad_norm": 0.9531372785568237,
+      "learning_rate": 0.00019307080321207912,
+      "loss": 3.6393,
+      "step": 1280
+    },
+    {
+      "epoch": 0.11595908391418484,
+      "grad_norm": 1.0083506107330322,
+      "learning_rate": 0.00019305919167467573,
+      "loss": 3.5573,
+      "step": 1281
+    },
+    {
+      "epoch": 0.11604960622793518,
+      "grad_norm": 0.9667612910270691,
+      "learning_rate": 0.0001930475707662143,
+      "loss": 3.5548,
+      "step": 1282
+    },
+    {
+      "epoch": 0.11614012854168553,
+      "grad_norm": 0.9590164422988892,
+      "learning_rate": 0.00019303594048786503,
+      "loss": 3.5589,
+      "step": 1283
+    },
+    {
+      "epoch": 0.11623065085543587,
+      "grad_norm": 0.9013192057609558,
+      "learning_rate": 0.0001930243008407991,
+      "loss": 3.5273,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1163211731691862,
+      "grad_norm": 1.077648639678955,
+      "learning_rate": 0.00019301265182618862,
+      "loss": 3.6351,
+      "step": 1285
+    },
+    {
+      "epoch": 0.11641169548293655,
+      "grad_norm": 0.9391737580299377,
+      "learning_rate": 0.00019300099344520666,
+      "loss": 3.6289,
+      "step": 1286
+    },
+    {
+      "epoch": 0.11650221779668689,
+      "grad_norm": 0.9416689276695251,
+      "learning_rate": 0.00019298932569902717,
+      "loss": 3.5554,
+      "step": 1287
+    },
+    {
+      "epoch": 0.11659274011043723,
+      "grad_norm": 0.9104056358337402,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 3.477,
+      "step": 1288
+    },
+    {
+      "epoch": 0.11668326242418756,
+      "grad_norm": 0.9194529056549072,
+      "learning_rate": 0.00019296596211577645,
+      "loss": 3.4682,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1167737847379379,
+      "grad_norm": 0.9856456518173218,
+      "learning_rate": 0.00019295426628105792,
+      "loss": 3.4858,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11686430705168824,
+      "grad_norm": 0.9236043691635132,
+      "learning_rate": 0.0001929425610858473,
+      "loss": 3.6041,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11695482936543858,
+      "grad_norm": 1.005852460861206,
+      "learning_rate": 0.00019293084653132335,
+      "loss": 3.5722,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11704535167918892,
+      "grad_norm": 0.9348894357681274,
+      "learning_rate": 0.00019291912261866568,
+      "loss": 3.6171,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11713587399293926,
+      "grad_norm": 1.0413389205932617,
+      "learning_rate": 0.00019290738934905492,
+      "loss": 3.5894,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1172263963066896,
+      "grad_norm": 1.0090665817260742,
+      "learning_rate": 0.00019289564672367258,
+      "loss": 3.5964,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11731691862043994,
+      "grad_norm": 1.0007750988006592,
+      "learning_rate": 0.00019288389474370117,
+      "loss": 3.5692,
+      "step": 1296
+    },
+    {
+      "epoch": 0.11740744093419028,
+      "grad_norm": 0.9936386942863464,
+      "learning_rate": 0.0001928721334103241,
+      "loss": 3.5052,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11749796324794062,
+      "grad_norm": 0.9362757802009583,
+      "learning_rate": 0.0001928603627247257,
+      "loss": 3.497,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11758848556169095,
+      "grad_norm": 1.0437519550323486,
+      "learning_rate": 0.00019284858268809137,
+      "loss": 3.5627,
+      "step": 1299
+    },
+    {
+      "epoch": 0.11767900787544129,
+      "grad_norm": 1.045013666152954,
+      "learning_rate": 0.00019283679330160726,
+      "loss": 3.5357,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11776953018919163,
+      "grad_norm": 1.0569078922271729,
+      "learning_rate": 0.00019282499456646065,
+      "loss": 3.5811,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11786005250294197,
+      "grad_norm": 0.9556296467781067,
+      "learning_rate": 0.00019281318648383956,
+      "loss": 3.5919,
+      "step": 1302
+    },
+    {
+      "epoch": 0.11795057481669231,
+      "grad_norm": 0.9206773042678833,
+      "learning_rate": 0.0001928013690549332,
+      "loss": 3.5818,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11804109713044265,
+      "grad_norm": 1.0458245277404785,
+      "learning_rate": 0.00019278954228093146,
+      "loss": 3.5062,
+      "step": 1304
+    },
+    {
+      "epoch": 0.118131619444193,
+      "grad_norm": 0.9117056727409363,
+      "learning_rate": 0.0001927777061630254,
+      "loss": 3.4908,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11822214175794334,
+      "grad_norm": 0.9894657135009766,
+      "learning_rate": 0.00019276586070240682,
+      "loss": 3.5476,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11831266407169368,
+      "grad_norm": 1.0137649774551392,
+      "learning_rate": 0.00019275400590026864,
+      "loss": 3.5573,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11840318638544402,
+      "grad_norm": 0.9838030934333801,
+      "learning_rate": 0.00019274214175780458,
+      "loss": 3.5925,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11849370869919436,
+      "grad_norm": 0.970636785030365,
+      "learning_rate": 0.00019273026827620941,
+      "loss": 3.5876,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11858423101294469,
+      "grad_norm": 1.0284696817398071,
+      "learning_rate": 0.00019271838545667876,
+      "loss": 3.508,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11867475332669503,
+      "grad_norm": 0.9239933490753174,
+      "learning_rate": 0.0001927064933004092,
+      "loss": 3.5244,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11876527564044537,
+      "grad_norm": 0.9314912557601929,
+      "learning_rate": 0.0001926945918085983,
+      "loss": 3.5688,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11885579795419571,
+      "grad_norm": 0.9962323307991028,
+      "learning_rate": 0.0001926826809824446,
+      "loss": 3.5586,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11894632026794605,
+      "grad_norm": 0.8930354714393616,
+      "learning_rate": 0.0001926707608231474,
+      "loss": 3.5189,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11903684258169639,
+      "grad_norm": 1.0008631944656372,
+      "learning_rate": 0.00019265883133190713,
+      "loss": 3.5565,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11912736489544673,
+      "grad_norm": 0.9283084869384766,
+      "learning_rate": 0.0001926468925099251,
+      "loss": 3.51,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11921788720919707,
+      "grad_norm": 0.9283885955810547,
+      "learning_rate": 0.00019263494435840355,
+      "loss": 3.5087,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11930840952294741,
+      "grad_norm": 0.9244400858879089,
+      "learning_rate": 0.0001926229868785456,
+      "loss": 3.4833,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11939893183669775,
+      "grad_norm": 0.9127993583679199,
+      "learning_rate": 0.0001926110200715554,
+      "loss": 3.5401,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11948945415044808,
+      "grad_norm": 0.8905998468399048,
+      "learning_rate": 0.00019259904393863802,
+      "loss": 3.5208,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11957997646419842,
+      "grad_norm": 0.915275514125824,
+      "learning_rate": 0.0001925870584809995,
+      "loss": 3.5384,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11967049877794876,
+      "grad_norm": 0.9308586716651917,
+      "learning_rate": 0.00019257506369984667,
+      "loss": 3.522,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1197610210916991,
+      "grad_norm": 0.981001079082489,
+      "learning_rate": 0.00019256305959638748,
+      "loss": 3.559,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11985154340544944,
+      "grad_norm": 0.9434746503829956,
+      "learning_rate": 0.0001925510461718307,
+      "loss": 3.5701,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11994206571919978,
+      "grad_norm": 0.897885799407959,
+      "learning_rate": 0.0001925390234273861,
+      "loss": 3.5868,
+      "step": 1325
+    },
+    {
+      "epoch": 0.12003258803295012,
+      "grad_norm": 0.9865405559539795,
+      "learning_rate": 0.0001925269913642644,
+      "loss": 3.5309,
+      "step": 1326
+    },
+    {
+      "epoch": 0.12012311034670047,
+      "grad_norm": 0.9099435806274414,
+      "learning_rate": 0.0001925149499836772,
+      "loss": 3.5296,
+      "step": 1327
+    },
+    {
+      "epoch": 0.1202136326604508,
+      "grad_norm": 1.001774787902832,
+      "learning_rate": 0.00019250289928683705,
+      "loss": 3.4915,
+      "step": 1328
+    },
+    {
+      "epoch": 0.12030415497420115,
+      "grad_norm": 0.9014139175415039,
+      "learning_rate": 0.00019249083927495747,
+      "loss": 3.555,
+      "step": 1329
+    },
+    {
+      "epoch": 0.12039467728795147,
+      "grad_norm": 1.0934901237487793,
+      "learning_rate": 0.00019247876994925292,
+      "loss": 3.5932,
+      "step": 1330
+    },
+    {
+      "epoch": 0.12048519960170181,
+      "grad_norm": 0.9719903469085693,
+      "learning_rate": 0.00019246669131093875,
+      "loss": 3.5252,
+      "step": 1331
+    },
+    {
+      "epoch": 0.12057572191545216,
+      "grad_norm": 1.0894511938095093,
+      "learning_rate": 0.00019245460336123134,
+      "loss": 3.5852,
+      "step": 1332
+    },
+    {
+      "epoch": 0.1206662442292025,
+      "grad_norm": 1.0272449254989624,
+      "learning_rate": 0.00019244250610134787,
+      "loss": 3.5173,
+      "step": 1333
+    },
+    {
+      "epoch": 0.12075676654295284,
+      "grad_norm": 1.0155720710754395,
+      "learning_rate": 0.00019243039953250655,
+      "loss": 3.6185,
+      "step": 1334
+    },
+    {
+      "epoch": 0.12084728885670318,
+      "grad_norm": 1.3178904056549072,
+      "learning_rate": 0.00019241828365592655,
+      "loss": 3.5569,
+      "step": 1335
+    },
+    {
+      "epoch": 0.12093781117045352,
+      "grad_norm": 0.923801600933075,
+      "learning_rate": 0.00019240615847282788,
+      "loss": 3.5147,
+      "step": 1336
+    },
+    {
+      "epoch": 0.12102833348420386,
+      "grad_norm": 0.9957558512687683,
+      "learning_rate": 0.0001923940239844316,
+      "loss": 3.6609,
+      "step": 1337
+    },
+    {
+      "epoch": 0.1211188557979542,
+      "grad_norm": 1.0449210405349731,
+      "learning_rate": 0.00019238188019195964,
+      "loss": 3.4863,
+      "step": 1338
+    },
+    {
+      "epoch": 0.12120937811170454,
+      "grad_norm": 0.9299371242523193,
+      "learning_rate": 0.00019236972709663487,
+      "loss": 3.5952,
+      "step": 1339
+    },
+    {
+      "epoch": 0.12129990042545487,
+      "grad_norm": 1.1433054208755493,
+      "learning_rate": 0.0001923575646996811,
+      "loss": 3.5433,
+      "step": 1340
+    },
+    {
+      "epoch": 0.12139042273920521,
+      "grad_norm": 1.0358242988586426,
+      "learning_rate": 0.00019234539300232314,
+      "loss": 3.5435,
+      "step": 1341
+    },
+    {
+      "epoch": 0.12148094505295555,
+      "grad_norm": 0.9494446516036987,
+      "learning_rate": 0.0001923332120057866,
+      "loss": 3.5099,
+      "step": 1342
+    },
+    {
+      "epoch": 0.12157146736670589,
+      "grad_norm": 1.068947196006775,
+      "learning_rate": 0.00019232102171129811,
+      "loss": 3.5744,
+      "step": 1343
+    },
+    {
+      "epoch": 0.12166198968045623,
+      "grad_norm": 0.9469134211540222,
+      "learning_rate": 0.00019230882212008528,
+      "loss": 3.562,
+      "step": 1344
+    },
+    {
+      "epoch": 0.12175251199420657,
+      "grad_norm": 0.9132084846496582,
+      "learning_rate": 0.00019229661323337657,
+      "loss": 3.5147,
+      "step": 1345
+    },
+    {
+      "epoch": 0.12184303430795691,
+      "grad_norm": 0.9203004837036133,
+      "learning_rate": 0.00019228439505240147,
+      "loss": 3.5774,
+      "step": 1346
+    },
+    {
+      "epoch": 0.12193355662170725,
+      "grad_norm": 0.9643454551696777,
+      "learning_rate": 0.00019227216757839027,
+      "loss": 3.5466,
+      "step": 1347
+    },
+    {
+      "epoch": 0.1220240789354576,
+      "grad_norm": 0.9245633482933044,
+      "learning_rate": 0.00019225993081257436,
+      "loss": 3.5785,
+      "step": 1348
+    },
+    {
+      "epoch": 0.12211460124920794,
+      "grad_norm": 0.9338047504425049,
+      "learning_rate": 0.0001922476847561859,
+      "loss": 3.4997,
+      "step": 1349
+    },
+    {
+      "epoch": 0.12220512356295828,
+      "grad_norm": 0.948529839515686,
+      "learning_rate": 0.00019223542941045817,
+      "loss": 3.4982,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1222956458767086,
+      "grad_norm": 0.8959832787513733,
+      "learning_rate": 0.00019222316477662517,
+      "loss": 3.5094,
+      "step": 1351
+    },
+    {
+      "epoch": 0.12238616819045894,
+      "grad_norm": 0.8953176736831665,
+      "learning_rate": 0.00019221089085592202,
+      "loss": 3.5108,
+      "step": 1352
+    },
+    {
+      "epoch": 0.12247669050420928,
+      "grad_norm": 0.9416437745094299,
+      "learning_rate": 0.00019219860764958465,
+      "loss": 3.4638,
+      "step": 1353
+    },
+    {
+      "epoch": 0.12256721281795963,
+      "grad_norm": 0.9912828803062439,
+      "learning_rate": 0.00019218631515885006,
+      "loss": 3.5506,
+      "step": 1354
+    },
+    {
+      "epoch": 0.12265773513170997,
+      "grad_norm": 0.9638131856918335,
+      "learning_rate": 0.00019217401338495605,
+      "loss": 3.5068,
+      "step": 1355
+    },
+    {
+      "epoch": 0.12274825744546031,
+      "grad_norm": 0.9816718697547913,
+      "learning_rate": 0.0001921617023291414,
+      "loss": 3.4654,
+      "step": 1356
+    },
+    {
+      "epoch": 0.12283877975921065,
+      "grad_norm": 0.9565276503562927,
+      "learning_rate": 0.00019214938199264581,
+      "loss": 3.495,
+      "step": 1357
+    },
+    {
+      "epoch": 0.12292930207296099,
+      "grad_norm": 0.9229384660720825,
+      "learning_rate": 0.00019213705237671006,
+      "loss": 3.5001,
+      "step": 1358
+    },
+    {
+      "epoch": 0.12301982438671133,
+      "grad_norm": 0.9728844165802002,
+      "learning_rate": 0.00019212471348257562,
+      "loss": 3.4909,
+      "step": 1359
+    },
+    {
+      "epoch": 0.12311034670046167,
+      "grad_norm": 0.9265088438987732,
+      "learning_rate": 0.000192112365311485,
+      "loss": 3.4382,
+      "step": 1360
+    },
+    {
+      "epoch": 0.123200869014212,
+      "grad_norm": 1.0222772359848022,
+      "learning_rate": 0.00019210000786468179,
+      "loss": 3.5261,
+      "step": 1361
+    },
+    {
+      "epoch": 0.12329139132796234,
+      "grad_norm": 0.9751477837562561,
+      "learning_rate": 0.00019208764114341027,
+      "loss": 3.5201,
+      "step": 1362
+    },
+    {
+      "epoch": 0.12338191364171268,
+      "grad_norm": 0.9564617872238159,
+      "learning_rate": 0.0001920752651489158,
+      "loss": 3.4582,
+      "step": 1363
+    },
+    {
+      "epoch": 0.12347243595546302,
+      "grad_norm": 0.923428475856781,
+      "learning_rate": 0.00019206287988244466,
+      "loss": 3.4767,
+      "step": 1364
+    },
+    {
+      "epoch": 0.12356295826921336,
+      "grad_norm": 0.9109767079353333,
+      "learning_rate": 0.00019205048534524406,
+      "loss": 3.5975,
+      "step": 1365
+    },
+    {
+      "epoch": 0.1236534805829637,
+      "grad_norm": 0.8991703987121582,
+      "learning_rate": 0.00019203808153856207,
+      "loss": 3.5281,
+      "step": 1366
+    },
+    {
+      "epoch": 0.12374400289671404,
+      "grad_norm": 0.9747706651687622,
+      "learning_rate": 0.0001920256684636478,
+      "loss": 3.6229,
+      "step": 1367
+    },
+    {
+      "epoch": 0.12383452521046438,
+      "grad_norm": 0.9236463308334351,
+      "learning_rate": 0.00019201324612175123,
+      "loss": 3.6039,
+      "step": 1368
+    },
+    {
+      "epoch": 0.12392504752421472,
+      "grad_norm": 0.970248281955719,
+      "learning_rate": 0.00019200081451412328,
+      "loss": 3.4691,
+      "step": 1369
+    },
+    {
+      "epoch": 0.12401556983796506,
+      "grad_norm": 0.8713831305503845,
+      "learning_rate": 0.00019198837364201585,
+      "loss": 3.4869,
+      "step": 1370
+    },
+    {
+      "epoch": 0.12410609215171539,
+      "grad_norm": 0.9409332871437073,
+      "learning_rate": 0.0001919759235066817,
+      "loss": 3.4984,
+      "step": 1371
+    },
+    {
+      "epoch": 0.12419661446546573,
+      "grad_norm": 0.9115921854972839,
+      "learning_rate": 0.00019196346410937457,
+      "loss": 3.5122,
+      "step": 1372
+    },
+    {
+      "epoch": 0.12428713677921607,
+      "grad_norm": 0.9855613112449646,
+      "learning_rate": 0.00019195099545134912,
+      "loss": 3.5239,
+      "step": 1373
+    },
+    {
+      "epoch": 0.12437765909296641,
+      "grad_norm": 0.8684623837471008,
+      "learning_rate": 0.00019193851753386093,
+      "loss": 3.5421,
+      "step": 1374
+    },
+    {
+      "epoch": 0.12446818140671675,
+      "grad_norm": 0.9192137122154236,
+      "learning_rate": 0.00019192603035816656,
+      "loss": 3.5047,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1245587037204671,
+      "grad_norm": 0.8745288848876953,
+      "learning_rate": 0.00019191353392552344,
+      "loss": 3.5405,
+      "step": 1376
+    },
+    {
+      "epoch": 0.12464922603421744,
+      "grad_norm": 0.8868715763092041,
+      "learning_rate": 0.00019190102823718998,
+      "loss": 3.4262,
+      "step": 1377
+    },
+    {
+      "epoch": 0.12473974834796778,
+      "grad_norm": 0.8993757963180542,
+      "learning_rate": 0.00019188851329442547,
+      "loss": 3.5437,
+      "step": 1378
+    },
+    {
+      "epoch": 0.12483027066171812,
+      "grad_norm": 0.8323911428451538,
+      "learning_rate": 0.0001918759890984902,
+      "loss": 3.5279,
+      "step": 1379
+    },
+    {
+      "epoch": 0.12492079297546846,
+      "grad_norm": 0.9292891025543213,
+      "learning_rate": 0.00019186345565064535,
+      "loss": 3.5397,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1250113152892188,
+      "grad_norm": 0.9611436724662781,
+      "learning_rate": 0.00019185091295215303,
+      "loss": 3.5072,
+      "step": 1381
+    },
+    {
+      "epoch": 0.12510183760296914,
+      "grad_norm": 0.918283998966217,
+      "learning_rate": 0.00019183836100427628,
+      "loss": 3.4667,
+      "step": 1382
+    },
+    {
+      "epoch": 0.12519235991671948,
+      "grad_norm": 1.1492328643798828,
+      "learning_rate": 0.00019182579980827908,
+      "loss": 3.5358,
+      "step": 1383
+    },
+    {
+      "epoch": 0.12528288223046982,
+      "grad_norm": 0.8876953125,
+      "learning_rate": 0.00019181322936542635,
+      "loss": 3.5318,
+      "step": 1384
+    },
+    {
+      "epoch": 0.12537340454422016,
+      "grad_norm": 0.9587846398353577,
+      "learning_rate": 0.00019180064967698398,
+      "loss": 3.5055,
+      "step": 1385
+    },
+    {
+      "epoch": 0.1254639268579705,
+      "grad_norm": 0.9714934229850769,
+      "learning_rate": 0.00019178806074421866,
+      "loss": 3.536,
+      "step": 1386
+    },
+    {
+      "epoch": 0.12555444917172082,
+      "grad_norm": 0.9751466512680054,
+      "learning_rate": 0.00019177546256839812,
+      "loss": 3.4873,
+      "step": 1387
+    },
+    {
+      "epoch": 0.12564497148547116,
+      "grad_norm": 1.019853115081787,
+      "learning_rate": 0.00019176285515079102,
+      "loss": 3.5236,
+      "step": 1388
+    },
+    {
+      "epoch": 0.1257354937992215,
+      "grad_norm": 0.9810512661933899,
+      "learning_rate": 0.00019175023849266697,
+      "loss": 3.5788,
+      "step": 1389
+    },
+    {
+      "epoch": 0.12582601611297184,
+      "grad_norm": 1.1488577127456665,
+      "learning_rate": 0.00019173761259529633,
+      "loss": 3.5419,
+      "step": 1390
+    },
+    {
+      "epoch": 0.12591653842672218,
+      "grad_norm": 1.1745120286941528,
+      "learning_rate": 0.00019172497745995065,
+      "loss": 3.5204,
+      "step": 1391
+    },
+    {
+      "epoch": 0.12600706074047252,
+      "grad_norm": 1.1298781633377075,
+      "learning_rate": 0.00019171233308790225,
+      "loss": 3.5831,
+      "step": 1392
+    },
+    {
+      "epoch": 0.12609758305422286,
+      "grad_norm": 1.0139622688293457,
+      "learning_rate": 0.00019169967948042444,
+      "loss": 3.5082,
+      "step": 1393
+    },
+    {
+      "epoch": 0.1261881053679732,
+      "grad_norm": 0.9887393116950989,
+      "learning_rate": 0.0001916870166387914,
+      "loss": 3.4952,
+      "step": 1394
+    },
+    {
+      "epoch": 0.12627862768172354,
+      "grad_norm": 1.0016969442367554,
+      "learning_rate": 0.00019167434456427828,
+      "loss": 3.444,
+      "step": 1395
+    },
+    {
+      "epoch": 0.12636914999547388,
+      "grad_norm": 0.8871667385101318,
+      "learning_rate": 0.00019166166325816118,
+      "loss": 3.5377,
+      "step": 1396
+    },
+    {
+      "epoch": 0.12645967230922422,
+      "grad_norm": 0.9886758923530579,
+      "learning_rate": 0.0001916489727217171,
+      "loss": 3.5954,
+      "step": 1397
+    },
+    {
+      "epoch": 0.12655019462297457,
+      "grad_norm": 1.1002118587493896,
+      "learning_rate": 0.00019163627295622397,
+      "loss": 3.6343,
+      "step": 1398
+    },
+    {
+      "epoch": 0.1266407169367249,
+      "grad_norm": 1.1045169830322266,
+      "learning_rate": 0.00019162356396296067,
+      "loss": 3.51,
+      "step": 1399
+    },
+    {
+      "epoch": 0.12673123925047525,
+      "grad_norm": 1.0856785774230957,
+      "learning_rate": 0.00019161084574320696,
+      "loss": 3.4861,
+      "step": 1400
+    },
+    {
+      "epoch": 0.1268217615642256,
+      "grad_norm": 0.9421377182006836,
+      "learning_rate": 0.00019159811829824364,
+      "loss": 3.4557,
+      "step": 1401
+    },
+    {
+      "epoch": 0.12691228387797593,
+      "grad_norm": 1.1086941957473755,
+      "learning_rate": 0.00019158538162935225,
+      "loss": 3.5516,
+      "step": 1402
+    },
+    {
+      "epoch": 0.12700280619172627,
+      "grad_norm": 0.9888288378715515,
+      "learning_rate": 0.0001915726357378155,
+      "loss": 3.5028,
+      "step": 1403
+    },
+    {
+      "epoch": 0.1270933285054766,
+      "grad_norm": 1.0222643613815308,
+      "learning_rate": 0.0001915598806249168,
+      "loss": 3.5185,
+      "step": 1404
+    },
+    {
+      "epoch": 0.12718385081922695,
+      "grad_norm": 1.073885202407837,
+      "learning_rate": 0.00019154711629194062,
+      "loss": 3.4822,
+      "step": 1405
+    },
+    {
+      "epoch": 0.1272743731329773,
+      "grad_norm": 1.0235117673873901,
+      "learning_rate": 0.00019153434274017236,
+      "loss": 3.4721,
+      "step": 1406
+    },
+    {
+      "epoch": 0.1273648954467276,
+      "grad_norm": 0.8984697461128235,
+      "learning_rate": 0.00019152155997089823,
+      "loss": 3.4577,
+      "step": 1407
+    },
+    {
+      "epoch": 0.12745541776047795,
+      "grad_norm": 1.0705209970474243,
+      "learning_rate": 0.0001915087679854056,
+      "loss": 3.6095,
+      "step": 1408
+    },
+    {
+      "epoch": 0.1275459400742283,
+      "grad_norm": 0.9822291135787964,
+      "learning_rate": 0.0001914959667849825,
+      "loss": 3.5591,
+      "step": 1409
+    },
+    {
+      "epoch": 0.12763646238797863,
+      "grad_norm": 0.9398566484451294,
+      "learning_rate": 0.00019148315637091803,
+      "loss": 3.53,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12772698470172897,
+      "grad_norm": 0.9197871088981628,
+      "learning_rate": 0.00019147033674450225,
+      "loss": 3.4776,
+      "step": 1411
+    },
+    {
+      "epoch": 0.1278175070154793,
+      "grad_norm": 0.966709315776825,
+      "learning_rate": 0.00019145750790702603,
+      "loss": 3.4558,
+      "step": 1412
+    },
+    {
+      "epoch": 0.12790802932922965,
+      "grad_norm": 0.8962054252624512,
+      "learning_rate": 0.00019144466985978128,
+      "loss": 3.4839,
+      "step": 1413
+    },
+    {
+      "epoch": 0.12799855164298,
+      "grad_norm": 0.9782243371009827,
+      "learning_rate": 0.0001914318226040608,
+      "loss": 3.5515,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12808907395673033,
+      "grad_norm": 0.97163987159729,
+      "learning_rate": 0.00019141896614115824,
+      "loss": 3.491,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12817959627048067,
+      "grad_norm": 0.8906465768814087,
+      "learning_rate": 0.00019140610047236833,
+      "loss": 3.4513,
+      "step": 1416
+    },
+    {
+      "epoch": 0.128270118584231,
+      "grad_norm": 0.9411565661430359,
+      "learning_rate": 0.00019139322559898658,
+      "loss": 3.5149,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12836064089798135,
+      "grad_norm": 1.0057343244552612,
+      "learning_rate": 0.00019138034152230954,
+      "loss": 3.5091,
+      "step": 1418
+    },
+    {
+      "epoch": 0.1284511632117317,
+      "grad_norm": 0.9373046159744263,
+      "learning_rate": 0.0001913674482436346,
+      "loss": 3.5215,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12854168552548204,
+      "grad_norm": 0.9086987972259521,
+      "learning_rate": 0.0001913545457642601,
+      "loss": 3.5094,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12863220783923238,
+      "grad_norm": 1.0151550769805908,
+      "learning_rate": 0.00019134163408548535,
+      "loss": 3.5051,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12872273015298272,
+      "grad_norm": 0.9737538695335388,
+      "learning_rate": 0.00019132871320861057,
+      "loss": 3.5201,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12881325246673306,
+      "grad_norm": 0.8971673250198364,
+      "learning_rate": 0.00019131578313493684,
+      "loss": 3.5268,
+      "step": 1423
+    },
+    {
+      "epoch": 0.1289037747804834,
+      "grad_norm": 0.9950623512268066,
+      "learning_rate": 0.00019130284386576624,
+      "loss": 3.5047,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12899429709423374,
+      "grad_norm": 0.9624702334403992,
+      "learning_rate": 0.00019128989540240178,
+      "loss": 3.4666,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12908481940798408,
+      "grad_norm": 0.9711193442344666,
+      "learning_rate": 0.00019127693774614738,
+      "loss": 3.5022,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12917534172173442,
+      "grad_norm": 1.091729998588562,
+      "learning_rate": 0.0001912639708983078,
+      "loss": 3.4588,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12926586403548473,
+      "grad_norm": 1.0445271730422974,
+      "learning_rate": 0.00019125099486018887,
+      "loss": 3.4819,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12935638634923508,
+      "grad_norm": 0.9670456051826477,
+      "learning_rate": 0.00019123800963309723,
+      "loss": 3.5112,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12944690866298542,
+      "grad_norm": 1.0354905128479004,
+      "learning_rate": 0.00019122501521834053,
+      "loss": 3.5502,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12953743097673576,
+      "grad_norm": 0.9035726189613342,
+      "learning_rate": 0.0001912120116172273,
+      "loss": 3.5438,
+      "step": 1431
+    },
+    {
+      "epoch": 0.1296279532904861,
+      "grad_norm": 1.0219416618347168,
+      "learning_rate": 0.000191198998831067,
+      "loss": 3.4761,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12971847560423644,
+      "grad_norm": 0.9801256060600281,
+      "learning_rate": 0.00019118597686117002,
+      "loss": 3.4326,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12980899791798678,
+      "grad_norm": 0.9733026027679443,
+      "learning_rate": 0.00019117294570884764,
+      "loss": 3.5113,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12989952023173712,
+      "grad_norm": 0.966882586479187,
+      "learning_rate": 0.00019115990537541216,
+      "loss": 3.5104,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12999004254548746,
+      "grad_norm": 0.934974730014801,
+      "learning_rate": 0.00019114685586217665,
+      "loss": 3.5122,
+      "step": 1436
+    },
+    {
+      "epoch": 0.1300805648592378,
+      "grad_norm": 0.9903215765953064,
+      "learning_rate": 0.00019113379717045529,
+      "loss": 3.471,
+      "step": 1437
+    },
+    {
+      "epoch": 0.13017108717298814,
+      "grad_norm": 1.0421541929244995,
+      "learning_rate": 0.00019112072930156302,
+      "loss": 3.5033,
+      "step": 1438
+    },
+    {
+      "epoch": 0.13026160948673848,
+      "grad_norm": 1.0076266527175903,
+      "learning_rate": 0.00019110765225681582,
+      "loss": 3.4898,
+      "step": 1439
+    },
+    {
+      "epoch": 0.13035213180048882,
+      "grad_norm": 0.9797717928886414,
+      "learning_rate": 0.0001910945660375305,
+      "loss": 3.4908,
+      "step": 1440
+    },
+    {
+      "epoch": 0.13044265411423916,
+      "grad_norm": 1.0033999681472778,
+      "learning_rate": 0.00019108147064502494,
+      "loss": 3.4859,
+      "step": 1441
+    },
+    {
+      "epoch": 0.1305331764279895,
+      "grad_norm": 0.9333057403564453,
+      "learning_rate": 0.00019106836608061772,
+      "loss": 3.4984,
+      "step": 1442
+    },
+    {
+      "epoch": 0.13062369874173985,
+      "grad_norm": 1.0323694944381714,
+      "learning_rate": 0.00019105525234562856,
+      "loss": 3.4998,
+      "step": 1443
+    },
+    {
+      "epoch": 0.1307142210554902,
+      "grad_norm": 0.9501261115074158,
+      "learning_rate": 0.00019104212944137796,
+      "loss": 3.4632,
+      "step": 1444
+    },
+    {
+      "epoch": 0.13080474336924053,
+      "grad_norm": 0.9542381763458252,
+      "learning_rate": 0.0001910289973691874,
+      "loss": 3.4215,
+      "step": 1445
+    },
+    {
+      "epoch": 0.13089526568299087,
+      "grad_norm": 1.1503580808639526,
+      "learning_rate": 0.00019101585613037934,
+      "loss": 3.53,
+      "step": 1446
+    },
+    {
+      "epoch": 0.1309857879967412,
+      "grad_norm": 0.8851205706596375,
+      "learning_rate": 0.00019100270572627704,
+      "loss": 3.5032,
+      "step": 1447
+    },
+    {
+      "epoch": 0.13107631031049152,
+      "grad_norm": 1.1074649095535278,
+      "learning_rate": 0.00019098954615820476,
+      "loss": 3.5025,
+      "step": 1448
+    },
+    {
+      "epoch": 0.13116683262424186,
+      "grad_norm": 1.0411019325256348,
+      "learning_rate": 0.00019097637742748766,
+      "loss": 3.5763,
+      "step": 1449
+    },
+    {
+      "epoch": 0.1312573549379922,
+      "grad_norm": 0.9861066937446594,
+      "learning_rate": 0.00019096319953545185,
+      "loss": 3.5056,
+      "step": 1450
+    },
+    {
+      "epoch": 0.13134787725174255,
+      "grad_norm": 1.028899073600769,
+      "learning_rate": 0.00019095001248342435,
+      "loss": 3.5157,
+      "step": 1451
+    },
+    {
+      "epoch": 0.13143839956549289,
+      "grad_norm": 0.9953543543815613,
+      "learning_rate": 0.00019093681627273306,
+      "loss": 3.5027,
+      "step": 1452
+    },
+    {
+      "epoch": 0.13152892187924323,
+      "grad_norm": 0.969194233417511,
+      "learning_rate": 0.00019092361090470688,
+      "loss": 3.4318,
+      "step": 1453
+    },
+    {
+      "epoch": 0.13161944419299357,
+      "grad_norm": 0.9358369708061218,
+      "learning_rate": 0.00019091039638067554,
+      "loss": 3.488,
+      "step": 1454
+    },
+    {
+      "epoch": 0.1317099665067439,
+      "grad_norm": 1.0051780939102173,
+      "learning_rate": 0.00019089717270196982,
+      "loss": 3.4563,
+      "step": 1455
+    },
+    {
+      "epoch": 0.13180048882049425,
+      "grad_norm": 0.9302476048469543,
+      "learning_rate": 0.00019088393986992124,
+      "loss": 3.4359,
+      "step": 1456
+    },
+    {
+      "epoch": 0.1318910111342446,
+      "grad_norm": 1.03059720993042,
+      "learning_rate": 0.00019087069788586243,
+      "loss": 3.4482,
+      "step": 1457
+    },
+    {
+      "epoch": 0.13198153344799493,
+      "grad_norm": 0.8484632968902588,
+      "learning_rate": 0.00019085744675112682,
+      "loss": 3.5185,
+      "step": 1458
+    },
+    {
+      "epoch": 0.13207205576174527,
+      "grad_norm": 0.9625447988510132,
+      "learning_rate": 0.00019084418646704882,
+      "loss": 3.4287,
+      "step": 1459
+    },
+    {
+      "epoch": 0.1321625780754956,
+      "grad_norm": 0.9219020009040833,
+      "learning_rate": 0.0001908309170349637,
+      "loss": 3.4854,
+      "step": 1460
+    },
+    {
+      "epoch": 0.13225310038924595,
+      "grad_norm": 0.8788402676582336,
+      "learning_rate": 0.00019081763845620777,
+      "loss": 3.4869,
+      "step": 1461
+    },
+    {
+      "epoch": 0.1323436227029963,
+      "grad_norm": 0.9385944604873657,
+      "learning_rate": 0.0001908043507321181,
+      "loss": 3.4948,
+      "step": 1462
+    },
+    {
+      "epoch": 0.13243414501674663,
+      "grad_norm": 0.9940325021743774,
+      "learning_rate": 0.00019079105386403283,
+      "loss": 3.4377,
+      "step": 1463
+    },
+    {
+      "epoch": 0.13252466733049698,
+      "grad_norm": 0.9103172421455383,
+      "learning_rate": 0.00019077774785329087,
+      "loss": 3.4933,
+      "step": 1464
+    },
+    {
+      "epoch": 0.13261518964424732,
+      "grad_norm": 0.9533010721206665,
+      "learning_rate": 0.00019076443270123222,
+      "loss": 3.4554,
+      "step": 1465
+    },
+    {
+      "epoch": 0.13270571195799766,
+      "grad_norm": 0.9723507761955261,
+      "learning_rate": 0.00019075110840919765,
+      "loss": 3.5272,
+      "step": 1466
+    },
+    {
+      "epoch": 0.132796234271748,
+      "grad_norm": 0.9802050590515137,
+      "learning_rate": 0.00019073777497852898,
+      "loss": 3.5226,
+      "step": 1467
+    },
+    {
+      "epoch": 0.13288675658549834,
+      "grad_norm": 1.023480772972107,
+      "learning_rate": 0.00019072443241056883,
+      "loss": 3.4561,
+      "step": 1468
+    },
+    {
+      "epoch": 0.13297727889924865,
+      "grad_norm": 1.0592081546783447,
+      "learning_rate": 0.00019071108070666082,
+      "loss": 3.5974,
+      "step": 1469
+    },
+    {
+      "epoch": 0.133067801212999,
+      "grad_norm": 1.0560920238494873,
+      "learning_rate": 0.00019069771986814947,
+      "loss": 3.4346,
+      "step": 1470
+    },
+    {
+      "epoch": 0.13315832352674933,
+      "grad_norm": 0.8262521624565125,
+      "learning_rate": 0.00019068434989638023,
+      "loss": 3.3677,
+      "step": 1471
+    },
+    {
+      "epoch": 0.13324884584049967,
+      "grad_norm": 1.0982695817947388,
+      "learning_rate": 0.00019067097079269942,
+      "loss": 3.4826,
+      "step": 1472
+    },
+    {
+      "epoch": 0.13333936815425002,
+      "grad_norm": 0.9667245149612427,
+      "learning_rate": 0.00019065758255845433,
+      "loss": 3.358,
+      "step": 1473
+    },
+    {
+      "epoch": 0.13342989046800036,
+      "grad_norm": 0.8388450741767883,
+      "learning_rate": 0.00019064418519499317,
+      "loss": 3.4301,
+      "step": 1474
+    },
+    {
+      "epoch": 0.1335204127817507,
+      "grad_norm": 0.9396435618400574,
+      "learning_rate": 0.000190630778703665,
+      "loss": 3.4448,
+      "step": 1475
+    },
+    {
+      "epoch": 0.13361093509550104,
+      "grad_norm": 0.8738861083984375,
+      "learning_rate": 0.00019061736308581995,
+      "loss": 3.544,
+      "step": 1476
+    },
+    {
+      "epoch": 0.13370145740925138,
+      "grad_norm": 0.9134858250617981,
+      "learning_rate": 0.00019060393834280888,
+      "loss": 3.4979,
+      "step": 1477
+    },
+    {
+      "epoch": 0.13379197972300172,
+      "grad_norm": 1.056847095489502,
+      "learning_rate": 0.0001905905044759837,
+      "loss": 3.4604,
+      "step": 1478
+    },
+    {
+      "epoch": 0.13388250203675206,
+      "grad_norm": 0.9403647184371948,
+      "learning_rate": 0.0001905770614866972,
+      "loss": 3.5034,
+      "step": 1479
+    },
+    {
+      "epoch": 0.1339730243505024,
+      "grad_norm": 1.0279815196990967,
+      "learning_rate": 0.0001905636093763031,
+      "loss": 3.4622,
+      "step": 1480
+    },
+    {
+      "epoch": 0.13406354666425274,
+      "grad_norm": 0.8829526305198669,
+      "learning_rate": 0.00019055014814615598,
+      "loss": 3.4812,
+      "step": 1481
+    },
+    {
+      "epoch": 0.13415406897800308,
+      "grad_norm": 0.9637207388877869,
+      "learning_rate": 0.00019053667779761146,
+      "loss": 3.5299,
+      "step": 1482
+    },
+    {
+      "epoch": 0.13424459129175342,
+      "grad_norm": 1.0193819999694824,
+      "learning_rate": 0.00019052319833202597,
+      "loss": 3.4526,
+      "step": 1483
+    },
+    {
+      "epoch": 0.13433511360550376,
+      "grad_norm": 0.8556128144264221,
+      "learning_rate": 0.00019050970975075684,
+      "loss": 3.4781,
+      "step": 1484
+    },
+    {
+      "epoch": 0.1344256359192541,
+      "grad_norm": 0.9707963466644287,
+      "learning_rate": 0.00019049621205516242,
+      "loss": 3.503,
+      "step": 1485
+    },
+    {
+      "epoch": 0.13451615823300445,
+      "grad_norm": 0.9474601745605469,
+      "learning_rate": 0.00019048270524660196,
+      "loss": 3.4621,
+      "step": 1486
+    },
+    {
+      "epoch": 0.1346066805467548,
+      "grad_norm": 0.9763953685760498,
+      "learning_rate": 0.00019046918932643555,
+      "loss": 3.4991,
+      "step": 1487
+    },
+    {
+      "epoch": 0.13469720286050513,
+      "grad_norm": 0.8932214975357056,
+      "learning_rate": 0.00019045566429602424,
+      "loss": 3.4698,
+      "step": 1488
+    },
+    {
+      "epoch": 0.13478772517425544,
+      "grad_norm": 0.9706962704658508,
+      "learning_rate": 0.00019044213015673,
+      "loss": 3.3941,
+      "step": 1489
+    },
+    {
+      "epoch": 0.13487824748800578,
+      "grad_norm": 0.9907398223876953,
+      "learning_rate": 0.00019042858690991574,
+      "loss": 3.4483,
+      "step": 1490
+    },
+    {
+      "epoch": 0.13496876980175612,
+      "grad_norm": 1.020414113998413,
+      "learning_rate": 0.00019041503455694524,
+      "loss": 3.4209,
+      "step": 1491
+    },
+    {
+      "epoch": 0.13505929211550646,
+      "grad_norm": 0.9534351229667664,
+      "learning_rate": 0.00019040147309918326,
+      "loss": 3.442,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1351498144292568,
+      "grad_norm": 0.9675535559654236,
+      "learning_rate": 0.00019038790253799535,
+      "loss": 3.568,
+      "step": 1493
+    },
+    {
+      "epoch": 0.13524033674300714,
+      "grad_norm": 0.9924291372299194,
+      "learning_rate": 0.0001903743228747482,
+      "loss": 3.4718,
+      "step": 1494
+    },
+    {
+      "epoch": 0.13533085905675749,
+      "grad_norm": 0.9964470863342285,
+      "learning_rate": 0.00019036073411080916,
+      "loss": 3.5284,
+      "step": 1495
+    },
+    {
+      "epoch": 0.13542138137050783,
+      "grad_norm": 1.0707465410232544,
+      "learning_rate": 0.00019034713624754672,
+      "loss": 3.5124,
+      "step": 1496
+    },
+    {
+      "epoch": 0.13551190368425817,
+      "grad_norm": 1.0072213411331177,
+      "learning_rate": 0.0001903335292863301,
+      "loss": 3.5582,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1356024259980085,
+      "grad_norm": 1.0074089765548706,
+      "learning_rate": 0.00019031991322852955,
+      "loss": 3.5065,
+      "step": 1498
+    },
+    {
+      "epoch": 0.13569294831175885,
+      "grad_norm": 1.06528902053833,
+      "learning_rate": 0.0001903062880755162,
+      "loss": 3.5091,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1357834706255092,
+      "grad_norm": 1.0103434324264526,
+      "learning_rate": 0.00019029265382866214,
+      "loss": 3.4023,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13587399293925953,
+      "grad_norm": 1.0442228317260742,
+      "learning_rate": 0.00019027901048934028,
+      "loss": 3.3906,
+      "step": 1501
+    },
+    {
+      "epoch": 0.13596451525300987,
+      "grad_norm": 0.9370065927505493,
+      "learning_rate": 0.00019026535805892456,
+      "loss": 3.5163,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1360550375667602,
+      "grad_norm": 0.9656304717063904,
+      "learning_rate": 0.00019025169653878973,
+      "loss": 3.5767,
+      "step": 1503
+    },
+    {
+      "epoch": 0.13614555988051055,
+      "grad_norm": 1.059404730796814,
+      "learning_rate": 0.00019023802593031154,
+      "loss": 3.3782,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1362360821942609,
+      "grad_norm": 0.9082018136978149,
+      "learning_rate": 0.0001902243462348666,
+      "loss": 3.4719,
+      "step": 1505
+    },
+    {
+      "epoch": 0.13632660450801123,
+      "grad_norm": 0.9529256820678711,
+      "learning_rate": 0.0001902106574538325,
+      "loss": 3.4799,
+      "step": 1506
+    },
+    {
+      "epoch": 0.13641712682176158,
+      "grad_norm": 1.0089010000228882,
+      "learning_rate": 0.0001901969595885876,
+      "loss": 3.4298,
+      "step": 1507
+    },
+    {
+      "epoch": 0.13650764913551192,
+      "grad_norm": 0.9658178687095642,
+      "learning_rate": 0.0001901832526405114,
+      "loss": 3.4837,
+      "step": 1508
+    },
+    {
+      "epoch": 0.13659817144926226,
+      "grad_norm": 0.8518688678741455,
+      "learning_rate": 0.0001901695366109841,
+      "loss": 3.4015,
+      "step": 1509
+    },
+    {
+      "epoch": 0.13668869376301257,
+      "grad_norm": 1.0004554986953735,
+      "learning_rate": 0.00019015581150138693,
+      "loss": 3.4847,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1367792160767629,
+      "grad_norm": 1.018133521080017,
+      "learning_rate": 0.00019014207731310202,
+      "loss": 3.4175,
+      "step": 1511
+    },
+    {
+      "epoch": 0.13686973839051325,
+      "grad_norm": 0.9288983941078186,
+      "learning_rate": 0.00019012833404751235,
+      "loss": 3.4276,
+      "step": 1512
+    },
+    {
+      "epoch": 0.1369602607042636,
+      "grad_norm": 1.074337124824524,
+      "learning_rate": 0.00019011458170600193,
+      "loss": 3.4433,
+      "step": 1513
+    },
+    {
+      "epoch": 0.13705078301801393,
+      "grad_norm": 0.8855864405632019,
+      "learning_rate": 0.0001901008202899556,
+      "loss": 3.4224,
+      "step": 1514
+    },
+    {
+      "epoch": 0.13714130533176427,
+      "grad_norm": 1.0131200551986694,
+      "learning_rate": 0.00019008704980075916,
+      "loss": 3.4568,
+      "step": 1515
+    },
+    {
+      "epoch": 0.13723182764551461,
+      "grad_norm": 0.971107006072998,
+      "learning_rate": 0.00019007327023979923,
+      "loss": 3.4019,
+      "step": 1516
+    },
+    {
+      "epoch": 0.13732234995926496,
+      "grad_norm": 0.9484931230545044,
+      "learning_rate": 0.00019005948160846344,
+      "loss": 3.4329,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1374128722730153,
+      "grad_norm": 1.075060486793518,
+      "learning_rate": 0.00019004568390814037,
+      "loss": 3.4524,
+      "step": 1518
+    },
+    {
+      "epoch": 0.13750339458676564,
+      "grad_norm": 1.008601427078247,
+      "learning_rate": 0.00019003187714021938,
+      "loss": 3.5014,
+      "step": 1519
+    },
+    {
+      "epoch": 0.13759391690051598,
+      "grad_norm": 0.9233218431472778,
+      "learning_rate": 0.0001900180613060908,
+      "loss": 3.4856,
+      "step": 1520
+    },
+    {
+      "epoch": 0.13768443921426632,
+      "grad_norm": 1.0939658880233765,
+      "learning_rate": 0.00019000423640714594,
+      "loss": 3.4835,
+      "step": 1521
+    },
+    {
+      "epoch": 0.13777496152801666,
+      "grad_norm": 0.95887291431427,
+      "learning_rate": 0.0001899904024447769,
+      "loss": 3.4504,
+      "step": 1522
+    },
+    {
+      "epoch": 0.137865483841767,
+      "grad_norm": 1.1146246194839478,
+      "learning_rate": 0.0001899765594203768,
+      "loss": 3.5101,
+      "step": 1523
+    },
+    {
+      "epoch": 0.13795600615551734,
+      "grad_norm": 0.9778677225112915,
+      "learning_rate": 0.00018996270733533964,
+      "loss": 3.4009,
+      "step": 1524
+    },
+    {
+      "epoch": 0.13804652846926768,
+      "grad_norm": 0.945473849773407,
+      "learning_rate": 0.00018994884619106031,
+      "loss": 3.4917,
+      "step": 1525
+    },
+    {
+      "epoch": 0.13813705078301802,
+      "grad_norm": 1.044701099395752,
+      "learning_rate": 0.00018993497598893466,
+      "loss": 3.3803,
+      "step": 1526
+    },
+    {
+      "epoch": 0.13822757309676836,
+      "grad_norm": 1.0342203378677368,
+      "learning_rate": 0.00018992109673035936,
+      "loss": 3.435,
+      "step": 1527
+    },
+    {
+      "epoch": 0.1383180954105187,
+      "grad_norm": 1.0020346641540527,
+      "learning_rate": 0.00018990720841673207,
+      "loss": 3.4686,
+      "step": 1528
+    },
+    {
+      "epoch": 0.13840861772426905,
+      "grad_norm": 0.9971665740013123,
+      "learning_rate": 0.00018989331104945137,
+      "loss": 3.4112,
+      "step": 1529
+    },
+    {
+      "epoch": 0.13849914003801936,
+      "grad_norm": 0.9247247576713562,
+      "learning_rate": 0.0001898794046299167,
+      "loss": 3.4628,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1385896623517697,
+      "grad_norm": 0.9933388233184814,
+      "learning_rate": 0.00018986548915952845,
+      "loss": 3.4662,
+      "step": 1531
+    },
+    {
+      "epoch": 0.13868018466552004,
+      "grad_norm": 0.9920099377632141,
+      "learning_rate": 0.0001898515646396879,
+      "loss": 3.4703,
+      "step": 1532
+    },
+    {
+      "epoch": 0.13877070697927038,
+      "grad_norm": 0.9866044521331787,
+      "learning_rate": 0.00018983763107179727,
+      "loss": 3.4612,
+      "step": 1533
+    },
+    {
+      "epoch": 0.13886122929302072,
+      "grad_norm": 0.9920278191566467,
+      "learning_rate": 0.0001898236884572596,
+      "loss": 3.4442,
+      "step": 1534
+    },
+    {
+      "epoch": 0.13895175160677106,
+      "grad_norm": 0.9837521910667419,
+      "learning_rate": 0.00018980973679747897,
+      "loss": 3.4224,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1390422739205214,
+      "grad_norm": 0.9028905034065247,
+      "learning_rate": 0.00018979577609386033,
+      "loss": 3.3831,
+      "step": 1536
+    },
+    {
+      "epoch": 0.13913279623427174,
+      "grad_norm": 1.0407545566558838,
+      "learning_rate": 0.0001897818063478095,
+      "loss": 3.4652,
+      "step": 1537
+    },
+    {
+      "epoch": 0.13922331854802208,
+      "grad_norm": 1.0021555423736572,
+      "learning_rate": 0.0001897678275607332,
+      "loss": 3.5014,
+      "step": 1538
+    },
+    {
+      "epoch": 0.13931384086177243,
+      "grad_norm": 1.2132344245910645,
+      "learning_rate": 0.00018975383973403914,
+      "loss": 3.4108,
+      "step": 1539
+    },
+    {
+      "epoch": 0.13940436317552277,
+      "grad_norm": 1.033569574356079,
+      "learning_rate": 0.00018973984286913584,
+      "loss": 3.4222,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1394948854892731,
+      "grad_norm": 0.9325762391090393,
+      "learning_rate": 0.00018972583696743285,
+      "loss": 3.4622,
+      "step": 1541
+    },
+    {
+      "epoch": 0.13958540780302345,
+      "grad_norm": 0.9818862080574036,
+      "learning_rate": 0.0001897118220303405,
+      "loss": 3.4873,
+      "step": 1542
+    },
+    {
+      "epoch": 0.1396759301167738,
+      "grad_norm": 1.0688445568084717,
+      "learning_rate": 0.00018969779805927018,
+      "loss": 3.459,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13976645243052413,
+      "grad_norm": 0.9991565346717834,
+      "learning_rate": 0.000189683765055634,
+      "loss": 3.4183,
+      "step": 1544
+    },
+    {
+      "epoch": 0.13985697474427447,
+      "grad_norm": 1.1223208904266357,
+      "learning_rate": 0.00018966972302084516,
+      "loss": 3.4349,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1399474970580248,
+      "grad_norm": 1.0523415803909302,
+      "learning_rate": 0.00018965567195631765,
+      "loss": 3.4106,
+      "step": 1546
+    },
+    {
+      "epoch": 0.14003801937177515,
+      "grad_norm": 1.0131930112838745,
+      "learning_rate": 0.00018964161186346646,
+      "loss": 3.3923,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1401285416855255,
+      "grad_norm": 0.9429630637168884,
+      "learning_rate": 0.0001896275427437074,
+      "loss": 3.3949,
+      "step": 1548
+    },
+    {
+      "epoch": 0.14021906399927583,
+      "grad_norm": 1.0264955759048462,
+      "learning_rate": 0.00018961346459845722,
+      "loss": 3.4737,
+      "step": 1549
+    },
+    {
+      "epoch": 0.14030958631302617,
+      "grad_norm": 1.078149437904358,
+      "learning_rate": 0.00018959937742913359,
+      "loss": 3.4399,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1404001086267765,
+      "grad_norm": 0.9134674072265625,
+      "learning_rate": 0.00018958528123715513,
+      "loss": 3.4693,
+      "step": 1551
+    },
+    {
+      "epoch": 0.14049063094052683,
+      "grad_norm": 1.0741221904754639,
+      "learning_rate": 0.0001895711760239413,
+      "loss": 3.415,
+      "step": 1552
+    },
+    {
+      "epoch": 0.14058115325427717,
+      "grad_norm": 0.9441749453544617,
+      "learning_rate": 0.00018955706179091248,
+      "loss": 3.4556,
+      "step": 1553
+    },
+    {
+      "epoch": 0.1406716755680275,
+      "grad_norm": 1.1377722024917603,
+      "learning_rate": 0.00018954293853949,
+      "loss": 3.4593,
+      "step": 1554
+    },
+    {
+      "epoch": 0.14076219788177785,
+      "grad_norm": 1.0758979320526123,
+      "learning_rate": 0.00018952880627109606,
+      "loss": 3.3585,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1408527201955282,
+      "grad_norm": 0.9907482266426086,
+      "learning_rate": 0.00018951466498715378,
+      "loss": 3.4805,
+      "step": 1556
+    },
+    {
+      "epoch": 0.14094324250927853,
+      "grad_norm": 0.977868378162384,
+      "learning_rate": 0.0001895005146890872,
+      "loss": 3.4679,
+      "step": 1557
+    },
+    {
+      "epoch": 0.14103376482302887,
+      "grad_norm": 0.994621753692627,
+      "learning_rate": 0.0001894863553783212,
+      "loss": 3.4203,
+      "step": 1558
+    },
+    {
+      "epoch": 0.14112428713677921,
+      "grad_norm": 0.9056159853935242,
+      "learning_rate": 0.00018947218705628167,
+      "loss": 3.3717,
+      "step": 1559
+    },
+    {
+      "epoch": 0.14121480945052955,
+      "grad_norm": 0.9723846316337585,
+      "learning_rate": 0.00018945800972439538,
+      "loss": 3.4537,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1413053317642799,
+      "grad_norm": 1.0195595026016235,
+      "learning_rate": 0.00018944382338408994,
+      "loss": 3.4347,
+      "step": 1561
+    },
+    {
+      "epoch": 0.14139585407803024,
+      "grad_norm": 0.9769614934921265,
+      "learning_rate": 0.00018942962803679395,
+      "loss": 3.4331,
+      "step": 1562
+    },
+    {
+      "epoch": 0.14148637639178058,
+      "grad_norm": 0.9903321862220764,
+      "learning_rate": 0.0001894154236839368,
+      "loss": 3.3952,
+      "step": 1563
+    },
+    {
+      "epoch": 0.14157689870553092,
+      "grad_norm": 1.066959261894226,
+      "learning_rate": 0.00018940121032694898,
+      "loss": 3.4105,
+      "step": 1564
+    },
+    {
+      "epoch": 0.14166742101928126,
+      "grad_norm": 0.9177101850509644,
+      "learning_rate": 0.00018938698796726176,
+      "loss": 3.4389,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1417579433330316,
+      "grad_norm": 0.9230473041534424,
+      "learning_rate": 0.00018937275660630727,
+      "loss": 3.4248,
+      "step": 1566
+    },
+    {
+      "epoch": 0.14184846564678194,
+      "grad_norm": 1.0110989809036255,
+      "learning_rate": 0.00018935851624551862,
+      "loss": 3.3903,
+      "step": 1567
+    },
+    {
+      "epoch": 0.14193898796053228,
+      "grad_norm": 0.9214289784431458,
+      "learning_rate": 0.00018934426688632986,
+      "loss": 3.4403,
+      "step": 1568
+    },
+    {
+      "epoch": 0.14202951027428262,
+      "grad_norm": 0.9641263484954834,
+      "learning_rate": 0.0001893300085301758,
+      "loss": 3.4886,
+      "step": 1569
+    },
+    {
+      "epoch": 0.14212003258803296,
+      "grad_norm": 1.065324306488037,
+      "learning_rate": 0.0001893157411784924,
+      "loss": 3.4602,
+      "step": 1570
+    },
+    {
+      "epoch": 0.14221055490178328,
+      "grad_norm": 1.0880188941955566,
+      "learning_rate": 0.00018930146483271627,
+      "loss": 3.4612,
+      "step": 1571
+    },
+    {
+      "epoch": 0.14230107721553362,
+      "grad_norm": 1.1627023220062256,
+      "learning_rate": 0.00018928717949428508,
+      "loss": 3.46,
+      "step": 1572
+    },
+    {
+      "epoch": 0.14239159952928396,
+      "grad_norm": 0.9831439256668091,
+      "learning_rate": 0.00018927288516463737,
+      "loss": 3.5054,
+      "step": 1573
+    },
+    {
+      "epoch": 0.1424821218430343,
+      "grad_norm": 1.106811761856079,
+      "learning_rate": 0.00018925858184521256,
+      "loss": 3.3911,
+      "step": 1574
+    },
+    {
+      "epoch": 0.14257264415678464,
+      "grad_norm": 0.993094265460968,
+      "learning_rate": 0.000189244269537451,
+      "loss": 3.4724,
+      "step": 1575
+    },
+    {
+      "epoch": 0.14266316647053498,
+      "grad_norm": 1.1177440881729126,
+      "learning_rate": 0.00018922994824279395,
+      "loss": 3.4473,
+      "step": 1576
+    },
+    {
+      "epoch": 0.14275368878428532,
+      "grad_norm": 0.9967995882034302,
+      "learning_rate": 0.00018921561796268353,
+      "loss": 3.4442,
+      "step": 1577
+    },
+    {
+      "epoch": 0.14284421109803566,
+      "grad_norm": 0.9261888265609741,
+      "learning_rate": 0.00018920127869856282,
+      "loss": 3.488,
+      "step": 1578
+    },
+    {
+      "epoch": 0.142934733411786,
+      "grad_norm": 0.9250870943069458,
+      "learning_rate": 0.0001891869304518758,
+      "loss": 3.4093,
+      "step": 1579
+    },
+    {
+      "epoch": 0.14302525572553634,
+      "grad_norm": 0.9998716115951538,
+      "learning_rate": 0.00018917257322406734,
+      "loss": 3.4525,
+      "step": 1580
+    },
+    {
+      "epoch": 0.14311577803928668,
+      "grad_norm": 0.977238118648529,
+      "learning_rate": 0.0001891582070165832,
+      "loss": 3.3836,
+      "step": 1581
+    },
+    {
+      "epoch": 0.14320630035303702,
+      "grad_norm": 0.9520090222358704,
+      "learning_rate": 0.00018914383183087002,
+      "loss": 3.4572,
+      "step": 1582
+    },
+    {
+      "epoch": 0.14329682266678737,
+      "grad_norm": 1.0415148735046387,
+      "learning_rate": 0.00018912944766837542,
+      "loss": 3.4392,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1433873449805377,
+      "grad_norm": 0.9595130681991577,
+      "learning_rate": 0.00018911505453054786,
+      "loss": 3.4015,
+      "step": 1584
+    },
+    {
+      "epoch": 0.14347786729428805,
+      "grad_norm": 0.9924153685569763,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 3.3964,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1435683896080384,
+      "grad_norm": 0.8610692024230957,
+      "learning_rate": 0.00018908624133469245,
+      "loss": 3.3727,
+      "step": 1586
+    },
+    {
+      "epoch": 0.14365891192178873,
+      "grad_norm": 1.0391387939453125,
+      "learning_rate": 0.00018907182127956605,
+      "loss": 3.5463,
+      "step": 1587
+    },
+    {
+      "epoch": 0.14374943423553907,
+      "grad_norm": 0.894811749458313,
+      "learning_rate": 0.00018905739225490967,
+      "loss": 3.465,
+      "step": 1588
+    },
+    {
+      "epoch": 0.1438399565492894,
+      "grad_norm": 0.9358307719230652,
+      "learning_rate": 0.00018904295426217633,
+      "loss": 3.4852,
+      "step": 1589
+    },
+    {
+      "epoch": 0.14393047886303975,
+      "grad_norm": 0.880241334438324,
+      "learning_rate": 0.00018902850730281992,
+      "loss": 3.3983,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1440210011767901,
+      "grad_norm": 0.9443881511688232,
+      "learning_rate": 0.00018901405137829527,
+      "loss": 3.4337,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1441115234905404,
+      "grad_norm": 0.9192736744880676,
+      "learning_rate": 0.0001889995864900581,
+      "loss": 3.4605,
+      "step": 1592
+    },
+    {
+      "epoch": 0.14420204580429075,
+      "grad_norm": 1.04524827003479,
+      "learning_rate": 0.000188985112639565,
+      "loss": 3.425,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1442925681180411,
+      "grad_norm": 0.9558252096176147,
+      "learning_rate": 0.00018897062982827344,
+      "loss": 3.4396,
+      "step": 1594
+    },
+    {
+      "epoch": 0.14438309043179143,
+      "grad_norm": 0.8985421657562256,
+      "learning_rate": 0.00018895613805764195,
+      "loss": 3.399,
+      "step": 1595
+    },
+    {
+      "epoch": 0.14447361274554177,
+      "grad_norm": 0.9235852956771851,
+      "learning_rate": 0.00018894163732912977,
+      "loss": 3.4708,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1445641350592921,
+      "grad_norm": 1.1525856256484985,
+      "learning_rate": 0.00018892712764419716,
+      "loss": 3.4202,
+      "step": 1597
+    },
+    {
+      "epoch": 0.14465465737304245,
+      "grad_norm": 0.9813726544380188,
+      "learning_rate": 0.0001889126090043052,
+      "loss": 3.4341,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1447451796867928,
+      "grad_norm": 0.9528858065605164,
+      "learning_rate": 0.00018889808141091597,
+      "loss": 3.4332,
+      "step": 1599
+    },
+    {
+      "epoch": 0.14483570200054313,
+      "grad_norm": 0.964176595211029,
+      "learning_rate": 0.00018888354486549237,
+      "loss": 3.427,
+      "step": 1600
+    },
+    {
+      "epoch": 0.14492622431429347,
+      "grad_norm": 0.9885776042938232,
+      "learning_rate": 0.00018886899936949823,
+      "loss": 3.4011,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1450167466280438,
+      "grad_norm": 0.9918528199195862,
+      "learning_rate": 0.00018885444492439829,
+      "loss": 3.4817,
+      "step": 1602
+    },
+    {
+      "epoch": 0.14510726894179415,
+      "grad_norm": 0.9491215944290161,
+      "learning_rate": 0.00018883988153165813,
+      "loss": 3.3567,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1451977912555445,
+      "grad_norm": 0.9561818242073059,
+      "learning_rate": 0.00018882530919274443,
+      "loss": 3.3902,
+      "step": 1604
+    },
+    {
+      "epoch": 0.14528831356929484,
+      "grad_norm": 0.9810082316398621,
+      "learning_rate": 0.00018881072790912445,
+      "loss": 3.3979,
+      "step": 1605
+    },
+    {
+      "epoch": 0.14537883588304518,
+      "grad_norm": 0.985683023929596,
+      "learning_rate": 0.0001887961376822666,
+      "loss": 3.3853,
+      "step": 1606
+    },
+    {
+      "epoch": 0.14546935819679552,
+      "grad_norm": 0.9815475940704346,
+      "learning_rate": 0.00018878153851364013,
+      "loss": 3.4275,
+      "step": 1607
+    },
+    {
+      "epoch": 0.14555988051054586,
+      "grad_norm": 1.012288212776184,
+      "learning_rate": 0.00018876693040471517,
+      "loss": 3.367,
+      "step": 1608
+    },
+    {
+      "epoch": 0.1456504028242962,
+      "grad_norm": 0.8871361613273621,
+      "learning_rate": 0.00018875231335696276,
+      "loss": 3.4282,
+      "step": 1609
+    },
+    {
+      "epoch": 0.14574092513804654,
+      "grad_norm": 0.9624701738357544,
+      "learning_rate": 0.0001887376873718548,
+      "loss": 3.4779,
+      "step": 1610
+    },
+    {
+      "epoch": 0.14583144745179688,
+      "grad_norm": 1.0587818622589111,
+      "learning_rate": 0.00018872305245086414,
+      "loss": 3.3712,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1459219697655472,
+      "grad_norm": 0.9896037578582764,
+      "learning_rate": 0.00018870840859546456,
+      "loss": 3.4185,
+      "step": 1612
+    },
+    {
+      "epoch": 0.14601249207929753,
+      "grad_norm": 0.9666143655776978,
+      "learning_rate": 0.00018869375580713062,
+      "loss": 3.4198,
+      "step": 1613
+    },
+    {
+      "epoch": 0.14610301439304788,
+      "grad_norm": 0.9257593154907227,
+      "learning_rate": 0.00018867909408733794,
+      "loss": 3.3719,
+      "step": 1614
+    },
+    {
+      "epoch": 0.14619353670679822,
+      "grad_norm": 0.8875561356544495,
+      "learning_rate": 0.00018866442343756288,
+      "loss": 3.4509,
+      "step": 1615
+    },
+    {
+      "epoch": 0.14628405902054856,
+      "grad_norm": 0.9463050365447998,
+      "learning_rate": 0.00018864974385928283,
+      "loss": 3.3477,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1463745813342989,
+      "grad_norm": 0.9890984296798706,
+      "learning_rate": 0.00018863505535397597,
+      "loss": 3.4441,
+      "step": 1617
+    },
+    {
+      "epoch": 0.14646510364804924,
+      "grad_norm": 1.125440001487732,
+      "learning_rate": 0.00018862035792312147,
+      "loss": 3.3712,
+      "step": 1618
+    },
+    {
+      "epoch": 0.14655562596179958,
+      "grad_norm": 1.0327622890472412,
+      "learning_rate": 0.00018860565156819935,
+      "loss": 3.4126,
+      "step": 1619
+    },
+    {
+      "epoch": 0.14664614827554992,
+      "grad_norm": 0.9664136171340942,
+      "learning_rate": 0.00018859093629069058,
+      "loss": 3.3958,
+      "step": 1620
+    },
+    {
+      "epoch": 0.14673667058930026,
+      "grad_norm": 1.0481597185134888,
+      "learning_rate": 0.00018857621209207692,
+      "loss": 3.4617,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1468271929030506,
+      "grad_norm": 1.0612270832061768,
+      "learning_rate": 0.0001885614789738411,
+      "loss": 3.429,
+      "step": 1622
+    },
+    {
+      "epoch": 0.14691771521680094,
+      "grad_norm": 1.0658665895462036,
+      "learning_rate": 0.0001885467369374668,
+      "loss": 3.4412,
+      "step": 1623
+    },
+    {
+      "epoch": 0.14700823753055128,
+      "grad_norm": 1.0005717277526855,
+      "learning_rate": 0.00018853198598443852,
+      "loss": 3.4411,
+      "step": 1624
+    },
+    {
+      "epoch": 0.14709875984430162,
+      "grad_norm": 0.9699280858039856,
+      "learning_rate": 0.00018851722611624164,
+      "loss": 3.3876,
+      "step": 1625
+    },
+    {
+      "epoch": 0.14718928215805197,
+      "grad_norm": 1.086168646812439,
+      "learning_rate": 0.00018850245733436255,
+      "loss": 3.3776,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1472798044718023,
+      "grad_norm": 1.1531916856765747,
+      "learning_rate": 0.00018848767964028845,
+      "loss": 3.4189,
+      "step": 1627
+    },
+    {
+      "epoch": 0.14737032678555265,
+      "grad_norm": 1.0791008472442627,
+      "learning_rate": 0.00018847289303550738,
+      "loss": 3.4613,
+      "step": 1628
+    },
+    {
+      "epoch": 0.147460849099303,
+      "grad_norm": 1.0842652320861816,
+      "learning_rate": 0.0001884580975215084,
+      "loss": 3.424,
+      "step": 1629
+    },
+    {
+      "epoch": 0.14755137141305333,
+      "grad_norm": 1.0409084558486938,
+      "learning_rate": 0.00018844329309978145,
+      "loss": 3.4281,
+      "step": 1630
+    },
+    {
+      "epoch": 0.14764189372680367,
+      "grad_norm": 1.0114672183990479,
+      "learning_rate": 0.0001884284797718173,
+      "loss": 3.4209,
+      "step": 1631
+    },
+    {
+      "epoch": 0.147732416040554,
+      "grad_norm": 0.9554311037063599,
+      "learning_rate": 0.00018841365753910765,
+      "loss": 3.4625,
+      "step": 1632
+    },
+    {
+      "epoch": 0.14782293835430432,
+      "grad_norm": 1.119260311126709,
+      "learning_rate": 0.0001883988264031451,
+      "loss": 3.4175,
+      "step": 1633
+    },
+    {
+      "epoch": 0.14791346066805466,
+      "grad_norm": 0.9265503883361816,
+      "learning_rate": 0.00018838398636542312,
+      "loss": 3.4327,
+      "step": 1634
+    },
+    {
+      "epoch": 0.148003982981805,
+      "grad_norm": 0.9685906767845154,
+      "learning_rate": 0.00018836913742743618,
+      "loss": 3.3286,
+      "step": 1635
+    },
+    {
+      "epoch": 0.14809450529555535,
+      "grad_norm": 0.9059945940971375,
+      "learning_rate": 0.0001883542795906795,
+      "loss": 3.3457,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1481850276093057,
+      "grad_norm": 1.1037979125976562,
+      "learning_rate": 0.0001883394128566493,
+      "loss": 3.3818,
+      "step": 1637
+    },
+    {
+      "epoch": 0.14827554992305603,
+      "grad_norm": 0.9228719472885132,
+      "learning_rate": 0.00018832453722684262,
+      "loss": 3.3509,
+      "step": 1638
+    },
+    {
+      "epoch": 0.14836607223680637,
+      "grad_norm": 1.0875881910324097,
+      "learning_rate": 0.00018830965270275746,
+      "loss": 3.4402,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1484565945505567,
+      "grad_norm": 1.1249608993530273,
+      "learning_rate": 0.00018829475928589271,
+      "loss": 3.4133,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14854711686430705,
+      "grad_norm": 1.0222041606903076,
+      "learning_rate": 0.00018827985697774811,
+      "loss": 3.3476,
+      "step": 1641
+    },
+    {
+      "epoch": 0.1486376391780574,
+      "grad_norm": 0.9464477896690369,
+      "learning_rate": 0.00018826494577982433,
+      "loss": 3.3605,
+      "step": 1642
+    },
+    {
+      "epoch": 0.14872816149180773,
+      "grad_norm": 1.1207705736160278,
+      "learning_rate": 0.0001882500256936229,
+      "loss": 3.3664,
+      "step": 1643
+    },
+    {
+      "epoch": 0.14881868380555807,
+      "grad_norm": 0.8890235424041748,
+      "learning_rate": 0.0001882350967206463,
+      "loss": 3.4928,
+      "step": 1644
+    },
+    {
+      "epoch": 0.1489092061193084,
+      "grad_norm": 1.437223196029663,
+      "learning_rate": 0.00018822015886239788,
+      "loss": 3.4072,
+      "step": 1645
+    },
+    {
+      "epoch": 0.14899972843305875,
+      "grad_norm": 1.0502667427062988,
+      "learning_rate": 0.00018820521212038188,
+      "loss": 3.3892,
+      "step": 1646
+    },
+    {
+      "epoch": 0.1490902507468091,
+      "grad_norm": 1.1289931535720825,
+      "learning_rate": 0.00018819025649610344,
+      "loss": 3.4068,
+      "step": 1647
+    },
+    {
+      "epoch": 0.14918077306055944,
+      "grad_norm": 0.9635894298553467,
+      "learning_rate": 0.0001881752919910686,
+      "loss": 3.4564,
+      "step": 1648
+    },
+    {
+      "epoch": 0.14927129537430978,
+      "grad_norm": 1.047222375869751,
+      "learning_rate": 0.00018816031860678423,
+      "loss": 3.3389,
+      "step": 1649
+    },
+    {
+      "epoch": 0.14936181768806012,
+      "grad_norm": 1.1027660369873047,
+      "learning_rate": 0.00018814533634475822,
+      "loss": 3.392,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14945234000181046,
+      "grad_norm": 0.9586750268936157,
+      "learning_rate": 0.0001881303452064992,
+      "loss": 3.4108,
+      "step": 1651
+    },
+    {
+      "epoch": 0.1495428623155608,
+      "grad_norm": 1.0597426891326904,
+      "learning_rate": 0.0001881153451935169,
+      "loss": 3.3901,
+      "step": 1652
+    },
+    {
+      "epoch": 0.1496333846293111,
+      "grad_norm": 1.084460973739624,
+      "learning_rate": 0.00018810033630732172,
+      "loss": 3.3989,
+      "step": 1653
+    },
+    {
+      "epoch": 0.14972390694306145,
+      "grad_norm": 1.0946100950241089,
+      "learning_rate": 0.0001880853185494251,
+      "loss": 3.4507,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1498144292568118,
+      "grad_norm": 0.8299400806427002,
+      "learning_rate": 0.00018807029192133927,
+      "loss": 3.3557,
+      "step": 1655
+    },
+    {
+      "epoch": 0.14990495157056213,
+      "grad_norm": 1.1521278619766235,
+      "learning_rate": 0.0001880552564245775,
+      "loss": 3.4572,
+      "step": 1656
+    },
+    {
+      "epoch": 0.14999547388431247,
+      "grad_norm": 0.9992600679397583,
+      "learning_rate": 0.00018804021206065378,
+      "loss": 3.3996,
+      "step": 1657
+    },
+    {
+      "epoch": 0.15008599619806282,
+      "grad_norm": 1.0150352716445923,
+      "learning_rate": 0.00018802515883108314,
+      "loss": 3.3961,
+      "step": 1658
+    },
+    {
+      "epoch": 0.15017651851181316,
+      "grad_norm": 0.9606907963752747,
+      "learning_rate": 0.00018801009673738138,
+      "loss": 3.4012,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1502670408255635,
+      "grad_norm": 0.9377285242080688,
+      "learning_rate": 0.00018799502578106534,
+      "loss": 3.3552,
+      "step": 1660
+    },
+    {
+      "epoch": 0.15035756313931384,
+      "grad_norm": 0.909257709980011,
+      "learning_rate": 0.00018797994596365255,
+      "loss": 3.3984,
+      "step": 1661
+    },
+    {
+      "epoch": 0.15044808545306418,
+      "grad_norm": 1.0251672267913818,
+      "learning_rate": 0.00018796485728666165,
+      "loss": 3.4319,
+      "step": 1662
+    },
+    {
+      "epoch": 0.15053860776681452,
+      "grad_norm": 0.857364296913147,
+      "learning_rate": 0.00018794975975161205,
+      "loss": 3.2863,
+      "step": 1663
+    },
+    {
+      "epoch": 0.15062913008056486,
+      "grad_norm": 1.043169379234314,
+      "learning_rate": 0.000187934653360024,
+      "loss": 3.4156,
+      "step": 1664
+    },
+    {
+      "epoch": 0.1507196523943152,
+      "grad_norm": 0.9212275743484497,
+      "learning_rate": 0.00018791953811341877,
+      "loss": 3.4092,
+      "step": 1665
+    },
+    {
+      "epoch": 0.15081017470806554,
+      "grad_norm": 1.0773905515670776,
+      "learning_rate": 0.00018790441401331847,
+      "loss": 3.4174,
+      "step": 1666
+    },
+    {
+      "epoch": 0.15090069702181588,
+      "grad_norm": 0.9565873146057129,
+      "learning_rate": 0.00018788928106124608,
+      "loss": 3.3829,
+      "step": 1667
+    },
+    {
+      "epoch": 0.15099121933556622,
+      "grad_norm": 1.0591626167297363,
+      "learning_rate": 0.0001878741392587255,
+      "loss": 3.4282,
+      "step": 1668
+    },
+    {
+      "epoch": 0.15108174164931656,
+      "grad_norm": 0.9780889749526978,
+      "learning_rate": 0.00018785898860728152,
+      "loss": 3.4514,
+      "step": 1669
+    },
+    {
+      "epoch": 0.1511722639630669,
+      "grad_norm": 0.9237357974052429,
+      "learning_rate": 0.00018784382910843976,
+      "loss": 3.3146,
+      "step": 1670
+    },
+    {
+      "epoch": 0.15126278627681725,
+      "grad_norm": 0.9899992942810059,
+      "learning_rate": 0.00018782866076372682,
+      "loss": 3.3556,
+      "step": 1671
+    },
+    {
+      "epoch": 0.1513533085905676,
+      "grad_norm": 1.0460927486419678,
+      "learning_rate": 0.00018781348357467013,
+      "loss": 3.4256,
+      "step": 1672
+    },
+    {
+      "epoch": 0.15144383090431793,
+      "grad_norm": 0.9416075944900513,
+      "learning_rate": 0.00018779829754279805,
+      "loss": 3.3589,
+      "step": 1673
+    },
+    {
+      "epoch": 0.15153435321806824,
+      "grad_norm": 1.208129644393921,
+      "learning_rate": 0.00018778310266963987,
+      "loss": 3.3374,
+      "step": 1674
+    },
+    {
+      "epoch": 0.15162487553181858,
+      "grad_norm": 0.9150928258895874,
+      "learning_rate": 0.00018776789895672558,
+      "loss": 3.3576,
+      "step": 1675
+    },
+    {
+      "epoch": 0.15171539784556892,
+      "grad_norm": 1.0763602256774902,
+      "learning_rate": 0.0001877526864055863,
+      "loss": 3.3803,
+      "step": 1676
+    },
+    {
+      "epoch": 0.15180592015931926,
+      "grad_norm": 0.9067040085792542,
+      "learning_rate": 0.00018773746501775387,
+      "loss": 3.404,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1518964424730696,
+      "grad_norm": 0.9757846593856812,
+      "learning_rate": 0.00018772223479476114,
+      "loss": 3.4245,
+      "step": 1678
+    },
+    {
+      "epoch": 0.15198696478681994,
+      "grad_norm": 0.9587690234184265,
+      "learning_rate": 0.00018770699573814176,
+      "loss": 3.3856,
+      "step": 1679
+    },
+    {
+      "epoch": 0.15207748710057029,
+      "grad_norm": 1.2289484739303589,
+      "learning_rate": 0.0001876917478494303,
+      "loss": 3.3734,
+      "step": 1680
+    },
+    {
+      "epoch": 0.15216800941432063,
+      "grad_norm": 0.899712860584259,
+      "learning_rate": 0.00018767649113016224,
+      "loss": 3.4229,
+      "step": 1681
+    },
+    {
+      "epoch": 0.15225853172807097,
+      "grad_norm": 1.2586640119552612,
+      "learning_rate": 0.00018766122558187395,
+      "loss": 3.3438,
+      "step": 1682
+    },
+    {
+      "epoch": 0.1523490540418213,
+      "grad_norm": 0.8806917667388916,
+      "learning_rate": 0.00018764595120610258,
+      "loss": 3.3843,
+      "step": 1683
+    },
+    {
+      "epoch": 0.15243957635557165,
+      "grad_norm": 1.154797077178955,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 3.4531,
+      "step": 1684
+    },
+    {
+      "epoch": 0.152530098669322,
+      "grad_norm": 0.9956049919128418,
+      "learning_rate": 0.00018761537597826425,
+      "loss": 3.4306,
+      "step": 1685
+    },
+    {
+      "epoch": 0.15262062098307233,
+      "grad_norm": 1.1457597017288208,
+      "learning_rate": 0.00018760007512927623,
+      "loss": 3.4051,
+      "step": 1686
+    },
+    {
+      "epoch": 0.15271114329682267,
+      "grad_norm": 0.9226441383361816,
+      "learning_rate": 0.000187584765458963,
+      "loss": 3.4094,
+      "step": 1687
+    },
+    {
+      "epoch": 0.152801665610573,
+      "grad_norm": 1.1360063552856445,
+      "learning_rate": 0.0001875694469688663,
+      "loss": 3.3587,
+      "step": 1688
+    },
+    {
+      "epoch": 0.15289218792432335,
+      "grad_norm": 0.8890431523323059,
+      "learning_rate": 0.00018755411966052865,
+      "loss": 3.3562,
+      "step": 1689
+    },
+    {
+      "epoch": 0.1529827102380737,
+      "grad_norm": 1.0502040386199951,
+      "learning_rate": 0.00018753878353549357,
+      "loss": 3.4775,
+      "step": 1690
+    },
+    {
+      "epoch": 0.15307323255182403,
+      "grad_norm": 0.9252305030822754,
+      "learning_rate": 0.00018752343859530538,
+      "loss": 3.3852,
+      "step": 1691
+    },
+    {
+      "epoch": 0.15316375486557438,
+      "grad_norm": 1.039414405822754,
+      "learning_rate": 0.00018750808484150935,
+      "loss": 3.3441,
+      "step": 1692
+    },
+    {
+      "epoch": 0.15325427717932472,
+      "grad_norm": 1.0017343759536743,
+      "learning_rate": 0.00018749272227565152,
+      "loss": 3.4088,
+      "step": 1693
+    },
+    {
+      "epoch": 0.15334479949307506,
+      "grad_norm": 0.9906306266784668,
+      "learning_rate": 0.00018747735089927897,
+      "loss": 3.3515,
+      "step": 1694
+    },
+    {
+      "epoch": 0.15343532180682537,
+      "grad_norm": 1.0180130004882812,
+      "learning_rate": 0.00018746197071393958,
+      "loss": 3.4191,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1535258441205757,
+      "grad_norm": 0.9270363450050354,
+      "learning_rate": 0.00018744658172118215,
+      "loss": 3.3679,
+      "step": 1696
+    },
+    {
+      "epoch": 0.15361636643432605,
+      "grad_norm": 1.0553703308105469,
+      "learning_rate": 0.0001874311839225563,
+      "loss": 3.4235,
+      "step": 1697
+    },
+    {
+      "epoch": 0.1537068887480764,
+      "grad_norm": 0.9562894105911255,
+      "learning_rate": 0.00018741577731961267,
+      "loss": 3.382,
+      "step": 1698
+    },
+    {
+      "epoch": 0.15379741106182673,
+      "grad_norm": 1.0290777683258057,
+      "learning_rate": 0.0001874003619139026,
+      "loss": 3.2955,
+      "step": 1699
+    },
+    {
+      "epoch": 0.15388793337557707,
+      "grad_norm": 1.0647125244140625,
+      "learning_rate": 0.00018738493770697852,
+      "loss": 3.4676,
+      "step": 1700
+    },
+    {
+      "epoch": 0.15397845568932741,
+      "grad_norm": 0.975814700126648,
+      "learning_rate": 0.00018736950470039357,
+      "loss": 3.3495,
+      "step": 1701
+    },
+    {
+      "epoch": 0.15406897800307776,
+      "grad_norm": 1.3109288215637207,
+      "learning_rate": 0.00018735406289570192,
+      "loss": 3.3745,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1541595003168281,
+      "grad_norm": 0.8713765740394592,
+      "learning_rate": 0.0001873386122944585,
+      "loss": 3.3465,
+      "step": 1703
+    },
+    {
+      "epoch": 0.15425002263057844,
+      "grad_norm": 0.9419022798538208,
+      "learning_rate": 0.00018732315289821921,
+      "loss": 3.4305,
+      "step": 1704
+    },
+    {
+      "epoch": 0.15434054494432878,
+      "grad_norm": 1.0088133811950684,
+      "learning_rate": 0.00018730768470854084,
+      "loss": 3.3117,
+      "step": 1705
+    },
+    {
+      "epoch": 0.15443106725807912,
+      "grad_norm": 0.9172151684761047,
+      "learning_rate": 0.00018729220772698097,
+      "loss": 3.4693,
+      "step": 1706
+    },
+    {
+      "epoch": 0.15452158957182946,
+      "grad_norm": 1.0178710222244263,
+      "learning_rate": 0.0001872767219550982,
+      "loss": 3.3724,
+      "step": 1707
+    },
+    {
+      "epoch": 0.1546121118855798,
+      "grad_norm": 0.9283190965652466,
+      "learning_rate": 0.0001872612273944519,
+      "loss": 3.3333,
+      "step": 1708
+    },
+    {
+      "epoch": 0.15470263419933014,
+      "grad_norm": 0.9490437507629395,
+      "learning_rate": 0.0001872457240466024,
+      "loss": 3.3637,
+      "step": 1709
+    },
+    {
+      "epoch": 0.15479315651308048,
+      "grad_norm": 0.9215477705001831,
+      "learning_rate": 0.0001872302119131109,
+      "loss": 3.3062,
+      "step": 1710
+    },
+    {
+      "epoch": 0.15488367882683082,
+      "grad_norm": 0.8653420805931091,
+      "learning_rate": 0.0001872146909955394,
+      "loss": 3.3447,
+      "step": 1711
+    },
+    {
+      "epoch": 0.15497420114058116,
+      "grad_norm": 0.9787905216217041,
+      "learning_rate": 0.00018719916129545093,
+      "loss": 3.4323,
+      "step": 1712
+    },
+    {
+      "epoch": 0.1550647234543315,
+      "grad_norm": 0.8503413200378418,
+      "learning_rate": 0.00018718362281440929,
+      "loss": 3.3701,
+      "step": 1713
+    },
+    {
+      "epoch": 0.15515524576808185,
+      "grad_norm": 1.0395936965942383,
+      "learning_rate": 0.0001871680755539792,
+      "loss": 3.3697,
+      "step": 1714
+    },
+    {
+      "epoch": 0.15524576808183216,
+      "grad_norm": 0.9945054650306702,
+      "learning_rate": 0.00018715251951572634,
+      "loss": 3.4051,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1553362903955825,
+      "grad_norm": 0.9538306593894958,
+      "learning_rate": 0.00018713695470121714,
+      "loss": 3.3483,
+      "step": 1716
+    },
+    {
+      "epoch": 0.15542681270933284,
+      "grad_norm": 0.9405638575553894,
+      "learning_rate": 0.00018712138111201895,
+      "loss": 3.3741,
+      "step": 1717
+    },
+    {
+      "epoch": 0.15551733502308318,
+      "grad_norm": 1.037540316581726,
+      "learning_rate": 0.00018710579874970008,
+      "loss": 3.2944,
+      "step": 1718
+    },
+    {
+      "epoch": 0.15560785733683352,
+      "grad_norm": 0.9655166268348694,
+      "learning_rate": 0.00018709020761582967,
+      "loss": 3.4324,
+      "step": 1719
+    },
+    {
+      "epoch": 0.15569837965058386,
+      "grad_norm": 0.9335793256759644,
+      "learning_rate": 0.00018707460771197774,
+      "loss": 3.3129,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1557889019643342,
+      "grad_norm": 0.9587303996086121,
+      "learning_rate": 0.00018705899903971517,
+      "loss": 3.3023,
+      "step": 1721
+    },
+    {
+      "epoch": 0.15587942427808454,
+      "grad_norm": 1.0624523162841797,
+      "learning_rate": 0.0001870433816006138,
+      "loss": 3.3677,
+      "step": 1722
+    },
+    {
+      "epoch": 0.15596994659183488,
+      "grad_norm": 0.9460078477859497,
+      "learning_rate": 0.0001870277553962463,
+      "loss": 3.4654,
+      "step": 1723
+    },
+    {
+      "epoch": 0.15606046890558523,
+      "grad_norm": 0.9337052702903748,
+      "learning_rate": 0.00018701212042818616,
+      "loss": 3.3522,
+      "step": 1724
+    },
+    {
+      "epoch": 0.15615099121933557,
+      "grad_norm": 1.0163861513137817,
+      "learning_rate": 0.0001869964766980079,
+      "loss": 3.3387,
+      "step": 1725
+    },
+    {
+      "epoch": 0.1562415135330859,
+      "grad_norm": 0.8806196451187134,
+      "learning_rate": 0.00018698082420728684,
+      "loss": 3.3855,
+      "step": 1726
+    },
+    {
+      "epoch": 0.15633203584683625,
+      "grad_norm": 0.9415055513381958,
+      "learning_rate": 0.00018696516295759914,
+      "loss": 3.3961,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1564225581605866,
+      "grad_norm": 1.0619957447052002,
+      "learning_rate": 0.0001869494929505219,
+      "loss": 3.4081,
+      "step": 1728
+    },
+    {
+      "epoch": 0.15651308047433693,
+      "grad_norm": 1.1068556308746338,
+      "learning_rate": 0.0001869338141876331,
+      "loss": 3.3572,
+      "step": 1729
+    },
+    {
+      "epoch": 0.15660360278808727,
+      "grad_norm": 0.9381948709487915,
+      "learning_rate": 0.00018691812667051162,
+      "loss": 3.3883,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1566941251018376,
+      "grad_norm": 1.0782184600830078,
+      "learning_rate": 0.00018690243040073713,
+      "loss": 3.3126,
+      "step": 1731
+    },
+    {
+      "epoch": 0.15678464741558795,
+      "grad_norm": 0.9588199853897095,
+      "learning_rate": 0.0001868867253798903,
+      "loss": 3.3538,
+      "step": 1732
+    },
+    {
+      "epoch": 0.1568751697293383,
+      "grad_norm": 1.112703561782837,
+      "learning_rate": 0.00018687101160955261,
+      "loss": 3.4172,
+      "step": 1733
+    },
+    {
+      "epoch": 0.15696569204308863,
+      "grad_norm": 1.008054256439209,
+      "learning_rate": 0.0001868552890913064,
+      "loss": 3.3506,
+      "step": 1734
+    },
+    {
+      "epoch": 0.15705621435683897,
+      "grad_norm": 1.138614535331726,
+      "learning_rate": 0.00018683955782673498,
+      "loss": 3.3616,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1571467366705893,
+      "grad_norm": 0.9115349054336548,
+      "learning_rate": 0.00018682381781742245,
+      "loss": 3.3093,
+      "step": 1736
+    },
+    {
+      "epoch": 0.15723725898433963,
+      "grad_norm": 1.0118136405944824,
+      "learning_rate": 0.00018680806906495382,
+      "loss": 3.3777,
+      "step": 1737
+    },
+    {
+      "epoch": 0.15732778129808997,
+      "grad_norm": 0.9229224920272827,
+      "learning_rate": 0.00018679231157091506,
+      "loss": 3.4082,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1574183036118403,
+      "grad_norm": 0.9689410924911499,
+      "learning_rate": 0.00018677654533689287,
+      "loss": 3.2666,
+      "step": 1739
+    },
+    {
+      "epoch": 0.15750882592559065,
+      "grad_norm": 1.0013811588287354,
+      "learning_rate": 0.00018676077036447494,
+      "loss": 3.4077,
+      "step": 1740
+    },
+    {
+      "epoch": 0.157599348239341,
+      "grad_norm": 0.9101596474647522,
+      "learning_rate": 0.0001867449866552498,
+      "loss": 3.3659,
+      "step": 1741
+    },
+    {
+      "epoch": 0.15768987055309133,
+      "grad_norm": 0.9114277958869934,
+      "learning_rate": 0.00018672919421080687,
+      "loss": 3.3438,
+      "step": 1742
+    },
+    {
+      "epoch": 0.15778039286684167,
+      "grad_norm": 1.0535948276519775,
+      "learning_rate": 0.00018671339303273648,
+      "loss": 3.4323,
+      "step": 1743
+    },
+    {
+      "epoch": 0.15787091518059201,
+      "grad_norm": 0.9524382948875427,
+      "learning_rate": 0.00018669758312262976,
+      "loss": 3.4158,
+      "step": 1744
+    },
+    {
+      "epoch": 0.15796143749434235,
+      "grad_norm": 1.006758213043213,
+      "learning_rate": 0.0001866817644820788,
+      "loss": 3.3891,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1580519598080927,
+      "grad_norm": 0.8595851063728333,
+      "learning_rate": 0.00018666593711267658,
+      "loss": 3.2898,
+      "step": 1746
+    },
+    {
+      "epoch": 0.15814248212184304,
+      "grad_norm": 0.932120680809021,
+      "learning_rate": 0.0001866501010160168,
+      "loss": 3.2822,
+      "step": 1747
+    },
+    {
+      "epoch": 0.15823300443559338,
+      "grad_norm": 0.9294149279594421,
+      "learning_rate": 0.00018663425619369425,
+      "loss": 3.4403,
+      "step": 1748
+    },
+    {
+      "epoch": 0.15832352674934372,
+      "grad_norm": 1.0117833614349365,
+      "learning_rate": 0.00018661840264730444,
+      "loss": 3.3145,
+      "step": 1749
+    },
+    {
+      "epoch": 0.15841404906309406,
+      "grad_norm": 0.9287446141242981,
+      "learning_rate": 0.00018660254037844388,
+      "loss": 3.3256,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1585045713768444,
+      "grad_norm": 1.0927577018737793,
+      "learning_rate": 0.00018658666938870986,
+      "loss": 3.3843,
+      "step": 1751
+    },
+    {
+      "epoch": 0.15859509369059474,
+      "grad_norm": 1.0403542518615723,
+      "learning_rate": 0.00018657078967970062,
+      "loss": 3.3488,
+      "step": 1752
+    },
+    {
+      "epoch": 0.15868561600434508,
+      "grad_norm": 0.9642882347106934,
+      "learning_rate": 0.00018655490125301521,
+      "loss": 3.4212,
+      "step": 1753
+    },
+    {
+      "epoch": 0.15877613831809542,
+      "grad_norm": 0.9047889709472656,
+      "learning_rate": 0.0001865390041102536,
+      "loss": 3.3679,
+      "step": 1754
+    },
+    {
+      "epoch": 0.15886666063184576,
+      "grad_norm": 0.9774285554885864,
+      "learning_rate": 0.0001865230982530167,
+      "loss": 3.2796,
+      "step": 1755
+    },
+    {
+      "epoch": 0.15895718294559608,
+      "grad_norm": 0.944681704044342,
+      "learning_rate": 0.0001865071836829061,
+      "loss": 3.3673,
+      "step": 1756
+    },
+    {
+      "epoch": 0.15904770525934642,
+      "grad_norm": 0.9720564484596252,
+      "learning_rate": 0.00018649126040152454,
+      "loss": 3.3176,
+      "step": 1757
+    },
+    {
+      "epoch": 0.15913822757309676,
+      "grad_norm": 1.0847887992858887,
+      "learning_rate": 0.00018647532841047537,
+      "loss": 3.3372,
+      "step": 1758
+    },
+    {
+      "epoch": 0.1592287498868471,
+      "grad_norm": 1.0074677467346191,
+      "learning_rate": 0.00018645938771136303,
+      "loss": 3.4203,
+      "step": 1759
+    },
+    {
+      "epoch": 0.15931927220059744,
+      "grad_norm": 1.074242115020752,
+      "learning_rate": 0.0001864434383057927,
+      "loss": 3.3453,
+      "step": 1760
+    },
+    {
+      "epoch": 0.15940979451434778,
+      "grad_norm": 0.9605053067207336,
+      "learning_rate": 0.0001864274801953705,
+      "loss": 3.3927,
+      "step": 1761
+    },
+    {
+      "epoch": 0.15950031682809812,
+      "grad_norm": 1.0063918828964233,
+      "learning_rate": 0.0001864115133817034,
+      "loss": 3.392,
+      "step": 1762
+    },
+    {
+      "epoch": 0.15959083914184846,
+      "grad_norm": 1.0003013610839844,
+      "learning_rate": 0.00018639553786639924,
+      "loss": 3.3505,
+      "step": 1763
+    },
+    {
+      "epoch": 0.1596813614555988,
+      "grad_norm": 1.046702265739441,
+      "learning_rate": 0.00018637955365106683,
+      "loss": 3.3625,
+      "step": 1764
+    },
+    {
+      "epoch": 0.15977188376934914,
+      "grad_norm": 0.8922204375267029,
+      "learning_rate": 0.0001863635607373157,
+      "loss": 3.3603,
+      "step": 1765
+    },
+    {
+      "epoch": 0.15986240608309948,
+      "grad_norm": 0.9413361549377441,
+      "learning_rate": 0.00018634755912675638,
+      "loss": 3.3674,
+      "step": 1766
+    },
+    {
+      "epoch": 0.15995292839684982,
+      "grad_norm": 0.9285749793052673,
+      "learning_rate": 0.00018633154882100023,
+      "loss": 3.4742,
+      "step": 1767
+    },
+    {
+      "epoch": 0.16004345071060017,
+      "grad_norm": 0.9485217332839966,
+      "learning_rate": 0.00018631552982165944,
+      "loss": 3.2931,
+      "step": 1768
+    },
+    {
+      "epoch": 0.1601339730243505,
+      "grad_norm": 0.9633521437644958,
+      "learning_rate": 0.00018629950213034722,
+      "loss": 3.4031,
+      "step": 1769
+    },
+    {
+      "epoch": 0.16022449533810085,
+      "grad_norm": 0.984691321849823,
+      "learning_rate": 0.00018628346574867745,
+      "loss": 3.3443,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1603150176518512,
+      "grad_norm": 1.0759167671203613,
+      "learning_rate": 0.00018626742067826506,
+      "loss": 3.407,
+      "step": 1771
+    },
+    {
+      "epoch": 0.16040553996560153,
+      "grad_norm": 1.0845353603363037,
+      "learning_rate": 0.00018625136692072575,
+      "loss": 3.4139,
+      "step": 1772
+    },
+    {
+      "epoch": 0.16049606227935187,
+      "grad_norm": 1.1085989475250244,
+      "learning_rate": 0.00018623530447767615,
+      "loss": 3.4532,
+      "step": 1773
+    },
+    {
+      "epoch": 0.1605865845931022,
+      "grad_norm": 0.9502083659172058,
+      "learning_rate": 0.00018621923335073376,
+      "loss": 3.3118,
+      "step": 1774
+    },
+    {
+      "epoch": 0.16067710690685255,
+      "grad_norm": 1.1858634948730469,
+      "learning_rate": 0.00018620315354151695,
+      "loss": 3.3605,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1607676292206029,
+      "grad_norm": 1.0164425373077393,
+      "learning_rate": 0.0001861870650516449,
+      "loss": 3.3414,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1608581515343532,
+      "grad_norm": 1.6314200162887573,
+      "learning_rate": 0.00018617096788273778,
+      "loss": 3.3694,
+      "step": 1777
+    },
+    {
+      "epoch": 0.16094867384810355,
+      "grad_norm": 1.19080650806427,
+      "learning_rate": 0.00018615486203641654,
+      "loss": 3.3305,
+      "step": 1778
+    },
+    {
+      "epoch": 0.1610391961618539,
+      "grad_norm": 1.3236929178237915,
+      "learning_rate": 0.00018613874751430306,
+      "loss": 3.4197,
+      "step": 1779
+    },
+    {
+      "epoch": 0.16112971847560423,
+      "grad_norm": 1.0939617156982422,
+      "learning_rate": 0.00018612262431802007,
+      "loss": 3.3496,
+      "step": 1780
+    },
+    {
+      "epoch": 0.16122024078935457,
+      "grad_norm": 0.9928971529006958,
+      "learning_rate": 0.00018610649244919114,
+      "loss": 3.3846,
+      "step": 1781
+    },
+    {
+      "epoch": 0.1613107631031049,
+      "grad_norm": 1.2084952592849731,
+      "learning_rate": 0.0001860903519094408,
+      "loss": 3.3893,
+      "step": 1782
+    },
+    {
+      "epoch": 0.16140128541685525,
+      "grad_norm": 1.0836037397384644,
+      "learning_rate": 0.0001860742027003944,
+      "loss": 3.3599,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1614918077306056,
+      "grad_norm": 1.016389012336731,
+      "learning_rate": 0.00018605804482367807,
+      "loss": 3.4159,
+      "step": 1784
+    },
+    {
+      "epoch": 0.16158233004435593,
+      "grad_norm": 0.9981551766395569,
+      "learning_rate": 0.00018604187828091906,
+      "loss": 3.391,
+      "step": 1785
+    },
+    {
+      "epoch": 0.16167285235810627,
+      "grad_norm": 0.9092329740524292,
+      "learning_rate": 0.0001860257030737452,
+      "loss": 3.3708,
+      "step": 1786
+    },
+    {
+      "epoch": 0.1617633746718566,
+      "grad_norm": 0.9498738646507263,
+      "learning_rate": 0.00018600951920378543,
+      "loss": 3.325,
+      "step": 1787
+    },
+    {
+      "epoch": 0.16185389698560695,
+      "grad_norm": 1.0748759508132935,
+      "learning_rate": 0.00018599332667266943,
+      "loss": 3.3166,
+      "step": 1788
+    },
+    {
+      "epoch": 0.1619444192993573,
+      "grad_norm": 1.01380455493927,
+      "learning_rate": 0.00018597712548202778,
+      "loss": 3.4486,
+      "step": 1789
+    },
+    {
+      "epoch": 0.16203494161310764,
+      "grad_norm": 1.2337299585342407,
+      "learning_rate": 0.00018596091563349192,
+      "loss": 3.3372,
+      "step": 1790
+    },
+    {
+      "epoch": 0.16212546392685798,
+      "grad_norm": 0.9750309586524963,
+      "learning_rate": 0.00018594469712869424,
+      "loss": 3.4084,
+      "step": 1791
+    },
+    {
+      "epoch": 0.16221598624060832,
+      "grad_norm": 1.2485698461532593,
+      "learning_rate": 0.00018592846996926793,
+      "loss": 3.3383,
+      "step": 1792
+    },
+    {
+      "epoch": 0.16230650855435866,
+      "grad_norm": 1.1208914518356323,
+      "learning_rate": 0.000185912234156847,
+      "loss": 3.388,
+      "step": 1793
+    },
+    {
+      "epoch": 0.162397030868109,
+      "grad_norm": 0.8985668420791626,
+      "learning_rate": 0.00018589598969306645,
+      "loss": 3.3437,
+      "step": 1794
+    },
+    {
+      "epoch": 0.16248755318185934,
+      "grad_norm": 1.1364415884017944,
+      "learning_rate": 0.00018587973657956211,
+      "loss": 3.3177,
+      "step": 1795
+    },
+    {
+      "epoch": 0.16257807549560968,
+      "grad_norm": 1.0473604202270508,
+      "learning_rate": 0.00018586347481797062,
+      "loss": 3.4173,
+      "step": 1796
+    },
+    {
+      "epoch": 0.16266859780936,
+      "grad_norm": 0.9839944839477539,
+      "learning_rate": 0.00018584720440992957,
+      "loss": 3.3088,
+      "step": 1797
+    },
+    {
+      "epoch": 0.16275912012311033,
+      "grad_norm": 1.1791011095046997,
+      "learning_rate": 0.0001858309253570774,
+      "loss": 3.3331,
+      "step": 1798
+    },
+    {
+      "epoch": 0.16284964243686068,
+      "grad_norm": 1.0361721515655518,
+      "learning_rate": 0.0001858146376610534,
+      "loss": 3.3211,
+      "step": 1799
+    },
+    {
+      "epoch": 0.16294016475061102,
+      "grad_norm": 0.9734877943992615,
+      "learning_rate": 0.00018579834132349772,
+      "loss": 3.3124,
+      "step": 1800
+    },
+    {
+      "epoch": 0.16303068706436136,
+      "grad_norm": 1.0755244493484497,
+      "learning_rate": 0.0001857820363460514,
+      "loss": 3.439,
+      "step": 1801
+    },
+    {
+      "epoch": 0.1631212093781117,
+      "grad_norm": 0.9513584971427917,
+      "learning_rate": 0.0001857657227303564,
+      "loss": 3.3119,
+      "step": 1802
+    },
+    {
+      "epoch": 0.16321173169186204,
+      "grad_norm": 0.9032912850379944,
+      "learning_rate": 0.00018574940047805548,
+      "loss": 3.3228,
+      "step": 1803
+    },
+    {
+      "epoch": 0.16330225400561238,
+      "grad_norm": 1.1157050132751465,
+      "learning_rate": 0.0001857330695907922,
+      "loss": 3.2961,
+      "step": 1804
+    },
+    {
+      "epoch": 0.16339277631936272,
+      "grad_norm": 1.0433858633041382,
+      "learning_rate": 0.00018571673007021123,
+      "loss": 3.3587,
+      "step": 1805
+    },
+    {
+      "epoch": 0.16348329863311306,
+      "grad_norm": 0.9127818942070007,
+      "learning_rate": 0.00018570038191795786,
+      "loss": 3.425,
+      "step": 1806
+    },
+    {
+      "epoch": 0.1635738209468634,
+      "grad_norm": 0.9649263024330139,
+      "learning_rate": 0.0001856840251356784,
+      "loss": 3.3319,
+      "step": 1807
+    },
+    {
+      "epoch": 0.16366434326061374,
+      "grad_norm": 0.983040988445282,
+      "learning_rate": 0.00018566765972501993,
+      "loss": 3.3198,
+      "step": 1808
+    },
+    {
+      "epoch": 0.16375486557436408,
+      "grad_norm": 0.9087714552879333,
+      "learning_rate": 0.00018565128568763048,
+      "loss": 3.2742,
+      "step": 1809
+    },
+    {
+      "epoch": 0.16384538788811442,
+      "grad_norm": 0.9332515597343445,
+      "learning_rate": 0.0001856349030251589,
+      "loss": 3.3727,
+      "step": 1810
+    },
+    {
+      "epoch": 0.16393591020186477,
+      "grad_norm": 0.8810479640960693,
+      "learning_rate": 0.00018561851173925495,
+      "loss": 3.2996,
+      "step": 1811
+    },
+    {
+      "epoch": 0.1640264325156151,
+      "grad_norm": 0.9994482398033142,
+      "learning_rate": 0.00018560211183156917,
+      "loss": 3.325,
+      "step": 1812
+    },
+    {
+      "epoch": 0.16411695482936545,
+      "grad_norm": 0.9770562052726746,
+      "learning_rate": 0.0001855857033037531,
+      "loss": 3.335,
+      "step": 1813
+    },
+    {
+      "epoch": 0.1642074771431158,
+      "grad_norm": 1.2734119892120361,
+      "learning_rate": 0.00018556928615745904,
+      "loss": 3.3162,
+      "step": 1814
+    },
+    {
+      "epoch": 0.16429799945686613,
+      "grad_norm": 0.9887378215789795,
+      "learning_rate": 0.0001855528603943402,
+      "loss": 3.2748,
+      "step": 1815
+    },
+    {
+      "epoch": 0.16438852177061647,
+      "grad_norm": 1.0900452136993408,
+      "learning_rate": 0.00018553642601605068,
+      "loss": 3.32,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1644790440843668,
+      "grad_norm": 0.8632405996322632,
+      "learning_rate": 0.00018551998302424538,
+      "loss": 3.3532,
+      "step": 1817
+    },
+    {
+      "epoch": 0.16456956639811712,
+      "grad_norm": 0.945196807384491,
+      "learning_rate": 0.00018550353142058015,
+      "loss": 3.3202,
+      "step": 1818
+    },
+    {
+      "epoch": 0.16466008871186746,
+      "grad_norm": 0.9298883080482483,
+      "learning_rate": 0.0001854870712067116,
+      "loss": 3.3807,
+      "step": 1819
+    },
+    {
+      "epoch": 0.1647506110256178,
+      "grad_norm": 0.828575849533081,
+      "learning_rate": 0.00018547060238429736,
+      "loss": 3.2874,
+      "step": 1820
+    },
+    {
+      "epoch": 0.16484113333936815,
+      "grad_norm": 1.0545551776885986,
+      "learning_rate": 0.00018545412495499578,
+      "loss": 3.3441,
+      "step": 1821
+    },
+    {
+      "epoch": 0.1649316556531185,
+      "grad_norm": 0.9633010625839233,
+      "learning_rate": 0.00018543763892046617,
+      "loss": 3.3648,
+      "step": 1822
+    },
+    {
+      "epoch": 0.16502217796686883,
+      "grad_norm": 0.9645661115646362,
+      "learning_rate": 0.00018542114428236864,
+      "loss": 3.3397,
+      "step": 1823
+    },
+    {
+      "epoch": 0.16511270028061917,
+      "grad_norm": 0.9543652534484863,
+      "learning_rate": 0.00018540464104236425,
+      "loss": 3.289,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1652032225943695,
+      "grad_norm": 0.8952108025550842,
+      "learning_rate": 0.0001853881292021148,
+      "loss": 3.3568,
+      "step": 1825
+    },
+    {
+      "epoch": 0.16529374490811985,
+      "grad_norm": 0.911792516708374,
+      "learning_rate": 0.00018537160876328312,
+      "loss": 3.3482,
+      "step": 1826
+    },
+    {
+      "epoch": 0.1653842672218702,
+      "grad_norm": 0.9409090876579285,
+      "learning_rate": 0.00018535507972753274,
+      "loss": 3.3525,
+      "step": 1827
+    },
+    {
+      "epoch": 0.16547478953562053,
+      "grad_norm": 0.9712650179862976,
+      "learning_rate": 0.00018533854209652818,
+      "loss": 3.3596,
+      "step": 1828
+    },
+    {
+      "epoch": 0.16556531184937087,
+      "grad_norm": 0.9322774410247803,
+      "learning_rate": 0.00018532199587193479,
+      "loss": 3.358,
+      "step": 1829
+    },
+    {
+      "epoch": 0.1656558341631212,
+      "grad_norm": 1.0519828796386719,
+      "learning_rate": 0.00018530544105541872,
+      "loss": 3.2452,
+      "step": 1830
+    },
+    {
+      "epoch": 0.16574635647687155,
+      "grad_norm": 0.9365170001983643,
+      "learning_rate": 0.0001852888776486471,
+      "loss": 3.3644,
+      "step": 1831
+    },
+    {
+      "epoch": 0.1658368787906219,
+      "grad_norm": 1.0104777812957764,
+      "learning_rate": 0.00018527230565328778,
+      "loss": 3.2377,
+      "step": 1832
+    },
+    {
+      "epoch": 0.16592740110437224,
+      "grad_norm": 0.9820732474327087,
+      "learning_rate": 0.00018525572507100964,
+      "loss": 3.3633,
+      "step": 1833
+    },
+    {
+      "epoch": 0.16601792341812258,
+      "grad_norm": 1.0812543630599976,
+      "learning_rate": 0.0001852391359034823,
+      "loss": 3.3757,
+      "step": 1834
+    },
+    {
+      "epoch": 0.16610844573187292,
+      "grad_norm": 0.9869741201400757,
+      "learning_rate": 0.00018522253815237636,
+      "loss": 3.2189,
+      "step": 1835
+    },
+    {
+      "epoch": 0.16619896804562326,
+      "grad_norm": 1.1047897338867188,
+      "learning_rate": 0.00018520593181936312,
+      "loss": 3.4102,
+      "step": 1836
+    },
+    {
+      "epoch": 0.1662894903593736,
+      "grad_norm": 1.1179860830307007,
+      "learning_rate": 0.0001851893169061149,
+      "loss": 3.2613,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1663800126731239,
+      "grad_norm": 0.9361429214477539,
+      "learning_rate": 0.00018517269341430476,
+      "loss": 3.3489,
+      "step": 1838
+    },
+    {
+      "epoch": 0.16647053498687425,
+      "grad_norm": 1.0511845350265503,
+      "learning_rate": 0.00018515606134560675,
+      "loss": 3.294,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1665610573006246,
+      "grad_norm": 0.9614317417144775,
+      "learning_rate": 0.0001851394207016957,
+      "loss": 3.3574,
+      "step": 1840
+    },
+    {
+      "epoch": 0.16665157961437493,
+      "grad_norm": 0.8911113142967224,
+      "learning_rate": 0.0001851227714842473,
+      "loss": 3.3029,
+      "step": 1841
+    },
+    {
+      "epoch": 0.16674210192812527,
+      "grad_norm": 1.0619572401046753,
+      "learning_rate": 0.00018510611369493816,
+      "loss": 3.3396,
+      "step": 1842
+    },
+    {
+      "epoch": 0.16683262424187562,
+      "grad_norm": 1.2229690551757812,
+      "learning_rate": 0.0001850894473354457,
+      "loss": 3.381,
+      "step": 1843
+    },
+    {
+      "epoch": 0.16692314655562596,
+      "grad_norm": 1.0011261701583862,
+      "learning_rate": 0.0001850727724074482,
+      "loss": 3.3347,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1670136688693763,
+      "grad_norm": 1.1166538000106812,
+      "learning_rate": 0.00018505608891262486,
+      "loss": 3.3346,
+      "step": 1845
+    },
+    {
+      "epoch": 0.16710419118312664,
+      "grad_norm": 0.881597638130188,
+      "learning_rate": 0.00018503939685265568,
+      "loss": 3.2962,
+      "step": 1846
+    },
+    {
+      "epoch": 0.16719471349687698,
+      "grad_norm": 0.9117144346237183,
+      "learning_rate": 0.00018502269622922154,
+      "loss": 3.3041,
+      "step": 1847
+    },
+    {
+      "epoch": 0.16728523581062732,
+      "grad_norm": 0.9797277450561523,
+      "learning_rate": 0.00018500598704400428,
+      "loss": 3.3918,
+      "step": 1848
+    },
+    {
+      "epoch": 0.16737575812437766,
+      "grad_norm": 0.9083805680274963,
+      "learning_rate": 0.00018498926929868642,
+      "loss": 3.3226,
+      "step": 1849
+    },
+    {
+      "epoch": 0.167466280438128,
+      "grad_norm": 1.0031794309616089,
+      "learning_rate": 0.00018497254299495146,
+      "loss": 3.3247,
+      "step": 1850
+    },
+    {
+      "epoch": 0.16755680275187834,
+      "grad_norm": 0.9302388429641724,
+      "learning_rate": 0.00018495580813448372,
+      "loss": 3.3599,
+      "step": 1851
+    },
+    {
+      "epoch": 0.16764732506562868,
+      "grad_norm": 0.9099136590957642,
+      "learning_rate": 0.00018493906471896848,
+      "loss": 3.3358,
+      "step": 1852
+    },
+    {
+      "epoch": 0.16773784737937902,
+      "grad_norm": 0.8923931121826172,
+      "learning_rate": 0.00018492231275009174,
+      "loss": 3.329,
+      "step": 1853
+    },
+    {
+      "epoch": 0.16782836969312936,
+      "grad_norm": 1.0236018896102905,
+      "learning_rate": 0.00018490555222954038,
+      "loss": 3.3552,
+      "step": 1854
+    },
+    {
+      "epoch": 0.1679188920068797,
+      "grad_norm": 0.9347558617591858,
+      "learning_rate": 0.00018488878315900227,
+      "loss": 3.3169,
+      "step": 1855
+    },
+    {
+      "epoch": 0.16800941432063005,
+      "grad_norm": 1.0411633253097534,
+      "learning_rate": 0.00018487200554016602,
+      "loss": 3.331,
+      "step": 1856
+    },
+    {
+      "epoch": 0.1680999366343804,
+      "grad_norm": 0.9088875651359558,
+      "learning_rate": 0.00018485521937472113,
+      "loss": 3.3023,
+      "step": 1857
+    },
+    {
+      "epoch": 0.16819045894813073,
+      "grad_norm": 0.9790009260177612,
+      "learning_rate": 0.00018483842466435798,
+      "loss": 3.3666,
+      "step": 1858
+    },
+    {
+      "epoch": 0.16828098126188104,
+      "grad_norm": 1.0297186374664307,
+      "learning_rate": 0.00018482162141076778,
+      "loss": 3.3311,
+      "step": 1859
+    },
+    {
+      "epoch": 0.16837150357563138,
+      "grad_norm": 0.9576818346977234,
+      "learning_rate": 0.0001848048096156426,
+      "loss": 3.2539,
+      "step": 1860
+    },
+    {
+      "epoch": 0.16846202588938172,
+      "grad_norm": 0.9573496580123901,
+      "learning_rate": 0.00018478798928067542,
+      "loss": 3.3519,
+      "step": 1861
+    },
+    {
+      "epoch": 0.16855254820313206,
+      "grad_norm": 0.907659649848938,
+      "learning_rate": 0.00018477116040756006,
+      "loss": 3.3051,
+      "step": 1862
+    },
+    {
+      "epoch": 0.1686430705168824,
+      "grad_norm": 0.9441897869110107,
+      "learning_rate": 0.00018475432299799117,
+      "loss": 3.2564,
+      "step": 1863
+    },
+    {
+      "epoch": 0.16873359283063274,
+      "grad_norm": 1.1152738332748413,
+      "learning_rate": 0.00018473747705366426,
+      "loss": 3.2523,
+      "step": 1864
+    },
+    {
+      "epoch": 0.16882411514438309,
+      "grad_norm": 1.0354543924331665,
+      "learning_rate": 0.00018472062257627575,
+      "loss": 3.3501,
+      "step": 1865
+    },
+    {
+      "epoch": 0.16891463745813343,
+      "grad_norm": 0.9793465733528137,
+      "learning_rate": 0.00018470375956752285,
+      "loss": 3.2576,
+      "step": 1866
+    },
+    {
+      "epoch": 0.16900515977188377,
+      "grad_norm": 0.9896705746650696,
+      "learning_rate": 0.0001846868880291037,
+      "loss": 3.3021,
+      "step": 1867
+    },
+    {
+      "epoch": 0.1690956820856341,
+      "grad_norm": 0.9611893892288208,
+      "learning_rate": 0.0001846700079627172,
+      "loss": 3.2927,
+      "step": 1868
+    },
+    {
+      "epoch": 0.16918620439938445,
+      "grad_norm": 0.9868037104606628,
+      "learning_rate": 0.00018465311937006321,
+      "loss": 3.2828,
+      "step": 1869
+    },
+    {
+      "epoch": 0.1692767267131348,
+      "grad_norm": 0.9744787812232971,
+      "learning_rate": 0.00018463622225284242,
+      "loss": 3.3267,
+      "step": 1870
+    },
+    {
+      "epoch": 0.16936724902688513,
+      "grad_norm": 1.002983570098877,
+      "learning_rate": 0.00018461931661275643,
+      "loss": 3.2383,
+      "step": 1871
+    },
+    {
+      "epoch": 0.16945777134063547,
+      "grad_norm": 0.876952052116394,
+      "learning_rate": 0.0001846024024515075,
+      "loss": 3.2725,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1695482936543858,
+      "grad_norm": 1.1321675777435303,
+      "learning_rate": 0.000184585479770799,
+      "loss": 3.3552,
+      "step": 1873
+    },
+    {
+      "epoch": 0.16963881596813615,
+      "grad_norm": 0.8953356742858887,
+      "learning_rate": 0.00018456854857233496,
+      "loss": 3.3455,
+      "step": 1874
+    },
+    {
+      "epoch": 0.1697293382818865,
+      "grad_norm": 1.0338746309280396,
+      "learning_rate": 0.00018455160885782045,
+      "loss": 3.3038,
+      "step": 1875
+    },
+    {
+      "epoch": 0.16981986059563683,
+      "grad_norm": 0.9554142951965332,
+      "learning_rate": 0.0001845346606289612,
+      "loss": 3.3413,
+      "step": 1876
+    },
+    {
+      "epoch": 0.16991038290938718,
+      "grad_norm": 0.9484089016914368,
+      "learning_rate": 0.00018451770388746398,
+      "loss": 3.3453,
+      "step": 1877
+    },
+    {
+      "epoch": 0.17000090522313752,
+      "grad_norm": 0.9068911075592041,
+      "learning_rate": 0.00018450073863503623,
+      "loss": 3.329,
+      "step": 1878
+    },
+    {
+      "epoch": 0.17009142753688783,
+      "grad_norm": 0.9792476296424866,
+      "learning_rate": 0.00018448376487338646,
+      "loss": 3.3212,
+      "step": 1879
+    },
+    {
+      "epoch": 0.17018194985063817,
+      "grad_norm": 0.9379374384880066,
+      "learning_rate": 0.00018446678260422385,
+      "loss": 3.3014,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1702724721643885,
+      "grad_norm": 0.9665486216545105,
+      "learning_rate": 0.00018444979182925854,
+      "loss": 3.337,
+      "step": 1881
+    },
+    {
+      "epoch": 0.17036299447813885,
+      "grad_norm": 1.1156326532363892,
+      "learning_rate": 0.00018443279255020152,
+      "loss": 3.3711,
+      "step": 1882
+    },
+    {
+      "epoch": 0.1704535167918892,
+      "grad_norm": 1.0622475147247314,
+      "learning_rate": 0.0001844157847687646,
+      "loss": 3.3345,
+      "step": 1883
+    },
+    {
+      "epoch": 0.17054403910563953,
+      "grad_norm": 1.058219313621521,
+      "learning_rate": 0.00018439876848666046,
+      "loss": 3.3102,
+      "step": 1884
+    },
+    {
+      "epoch": 0.17063456141938987,
+      "grad_norm": 1.2012033462524414,
+      "learning_rate": 0.00018438174370560264,
+      "loss": 3.3606,
+      "step": 1885
+    },
+    {
+      "epoch": 0.17072508373314021,
+      "grad_norm": 1.0893925428390503,
+      "learning_rate": 0.00018436471042730554,
+      "loss": 3.3498,
+      "step": 1886
+    },
+    {
+      "epoch": 0.17081560604689056,
+      "grad_norm": 1.0066337585449219,
+      "learning_rate": 0.0001843476686534844,
+      "loss": 3.3289,
+      "step": 1887
+    },
+    {
+      "epoch": 0.1709061283606409,
+      "grad_norm": 1.0413285493850708,
+      "learning_rate": 0.00018433061838585534,
+      "loss": 3.3485,
+      "step": 1888
+    },
+    {
+      "epoch": 0.17099665067439124,
+      "grad_norm": 1.0871161222457886,
+      "learning_rate": 0.00018431355962613536,
+      "loss": 3.3714,
+      "step": 1889
+    },
+    {
+      "epoch": 0.17108717298814158,
+      "grad_norm": 0.9766511917114258,
+      "learning_rate": 0.00018429649237604217,
+      "loss": 3.309,
+      "step": 1890
+    },
+    {
+      "epoch": 0.17117769530189192,
+      "grad_norm": 1.0211293697357178,
+      "learning_rate": 0.00018427941663729455,
+      "loss": 3.2944,
+      "step": 1891
+    },
+    {
+      "epoch": 0.17126821761564226,
+      "grad_norm": 1.0084211826324463,
+      "learning_rate": 0.00018426233241161191,
+      "loss": 3.3315,
+      "step": 1892
+    },
+    {
+      "epoch": 0.1713587399293926,
+      "grad_norm": 0.87370765209198,
+      "learning_rate": 0.00018424523970071477,
+      "loss": 3.3272,
+      "step": 1893
+    },
+    {
+      "epoch": 0.17144926224314294,
+      "grad_norm": 0.9169411063194275,
+      "learning_rate": 0.0001842281385063243,
+      "loss": 3.295,
+      "step": 1894
+    },
+    {
+      "epoch": 0.17153978455689328,
+      "grad_norm": 1.0767097473144531,
+      "learning_rate": 0.00018421102883016253,
+      "loss": 3.3453,
+      "step": 1895
+    },
+    {
+      "epoch": 0.17163030687064362,
+      "grad_norm": 1.1827503442764282,
+      "learning_rate": 0.00018419391067395248,
+      "loss": 3.2742,
+      "step": 1896
+    },
+    {
+      "epoch": 0.17172082918439396,
+      "grad_norm": 0.9097943305969238,
+      "learning_rate": 0.00018417678403941795,
+      "loss": 3.2716,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1718113514981443,
+      "grad_norm": 0.9012870788574219,
+      "learning_rate": 0.00018415964892828357,
+      "loss": 3.3063,
+      "step": 1898
+    },
+    {
+      "epoch": 0.17190187381189465,
+      "grad_norm": 1.1969982385635376,
+      "learning_rate": 0.00018414250534227485,
+      "loss": 3.2844,
+      "step": 1899
+    },
+    {
+      "epoch": 0.17199239612564496,
+      "grad_norm": 0.9854839444160461,
+      "learning_rate": 0.00018412535328311814,
+      "loss": 3.3285,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1720829184393953,
+      "grad_norm": 1.0452015399932861,
+      "learning_rate": 0.00018410819275254062,
+      "loss": 3.2647,
+      "step": 1901
+    },
+    {
+      "epoch": 0.17217344075314564,
+      "grad_norm": 0.9425656199455261,
+      "learning_rate": 0.00018409102375227044,
+      "loss": 3.358,
+      "step": 1902
+    },
+    {
+      "epoch": 0.17226396306689598,
+      "grad_norm": 0.9737415313720703,
+      "learning_rate": 0.00018407384628403643,
+      "loss": 3.3233,
+      "step": 1903
+    },
+    {
+      "epoch": 0.17235448538064632,
+      "grad_norm": 1.0962470769882202,
+      "learning_rate": 0.00018405666034956844,
+      "loss": 3.3832,
+      "step": 1904
+    },
+    {
+      "epoch": 0.17244500769439666,
+      "grad_norm": 1.1687324047088623,
+      "learning_rate": 0.00018403946595059703,
+      "loss": 3.3107,
+      "step": 1905
+    },
+    {
+      "epoch": 0.172535530008147,
+      "grad_norm": 0.9966987371444702,
+      "learning_rate": 0.00018402226308885368,
+      "loss": 3.3198,
+      "step": 1906
+    },
+    {
+      "epoch": 0.17262605232189734,
+      "grad_norm": 0.9648373126983643,
+      "learning_rate": 0.00018400505176607078,
+      "loss": 3.2521,
+      "step": 1907
+    },
+    {
+      "epoch": 0.17271657463564768,
+      "grad_norm": 1.113847255706787,
+      "learning_rate": 0.00018398783198398145,
+      "loss": 3.363,
+      "step": 1908
+    },
+    {
+      "epoch": 0.17280709694939803,
+      "grad_norm": 1.0072240829467773,
+      "learning_rate": 0.00018397060374431972,
+      "loss": 3.3407,
+      "step": 1909
+    },
+    {
+      "epoch": 0.17289761926314837,
+      "grad_norm": 0.9911056160926819,
+      "learning_rate": 0.0001839533670488205,
+      "loss": 3.2513,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1729881415768987,
+      "grad_norm": 1.049584984779358,
+      "learning_rate": 0.00018393612189921953,
+      "loss": 3.3049,
+      "step": 1911
+    },
+    {
+      "epoch": 0.17307866389064905,
+      "grad_norm": 0.9908673763275146,
+      "learning_rate": 0.00018391886829725334,
+      "loss": 3.3312,
+      "step": 1912
+    },
+    {
+      "epoch": 0.1731691862043994,
+      "grad_norm": 1.1561472415924072,
+      "learning_rate": 0.00018390160624465944,
+      "loss": 3.3125,
+      "step": 1913
+    },
+    {
+      "epoch": 0.17325970851814973,
+      "grad_norm": 1.0068906545639038,
+      "learning_rate": 0.0001838843357431761,
+      "loss": 3.3257,
+      "step": 1914
+    },
+    {
+      "epoch": 0.17335023083190007,
+      "grad_norm": 0.9671674370765686,
+      "learning_rate": 0.00018386705679454242,
+      "loss": 3.3392,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1734407531456504,
+      "grad_norm": 0.8658766150474548,
+      "learning_rate": 0.0001838497694004984,
+      "loss": 3.2903,
+      "step": 1916
+    },
+    {
+      "epoch": 0.17353127545940075,
+      "grad_norm": 1.09023118019104,
+      "learning_rate": 0.00018383247356278494,
+      "loss": 3.3672,
+      "step": 1917
+    },
+    {
+      "epoch": 0.1736217977731511,
+      "grad_norm": 0.9814264178276062,
+      "learning_rate": 0.00018381516928314367,
+      "loss": 3.3352,
+      "step": 1918
+    },
+    {
+      "epoch": 0.17371232008690143,
+      "grad_norm": 1.116402268409729,
+      "learning_rate": 0.00018379785656331713,
+      "loss": 3.3452,
+      "step": 1919
+    },
+    {
+      "epoch": 0.17380284240065175,
+      "grad_norm": 1.029109001159668,
+      "learning_rate": 0.00018378053540504873,
+      "loss": 3.3507,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1738933647144021,
+      "grad_norm": 0.9994332194328308,
+      "learning_rate": 0.00018376320581008272,
+      "loss": 3.3412,
+      "step": 1921
+    },
+    {
+      "epoch": 0.17398388702815243,
+      "grad_norm": 0.9789600372314453,
+      "learning_rate": 0.00018374586778016418,
+      "loss": 3.2994,
+      "step": 1922
+    },
+    {
+      "epoch": 0.17407440934190277,
+      "grad_norm": 1.317229151725769,
+      "learning_rate": 0.00018372852131703904,
+      "loss": 3.3474,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1741649316556531,
+      "grad_norm": 1.0556979179382324,
+      "learning_rate": 0.00018371116642245408,
+      "loss": 3.3084,
+      "step": 1924
+    },
+    {
+      "epoch": 0.17425545396940345,
+      "grad_norm": 1.17436683177948,
+      "learning_rate": 0.00018369380309815698,
+      "loss": 3.2949,
+      "step": 1925
+    },
+    {
+      "epoch": 0.1743459762831538,
+      "grad_norm": 1.2278040647506714,
+      "learning_rate": 0.00018367643134589617,
+      "loss": 3.2857,
+      "step": 1926
+    },
+    {
+      "epoch": 0.17443649859690413,
+      "grad_norm": 1.1209522485733032,
+      "learning_rate": 0.00018365905116742105,
+      "loss": 3.2659,
+      "step": 1927
+    },
+    {
+      "epoch": 0.17452702091065447,
+      "grad_norm": 1.10820472240448,
+      "learning_rate": 0.00018364166256448173,
+      "loss": 3.2915,
+      "step": 1928
+    },
+    {
+      "epoch": 0.17461754322440481,
+      "grad_norm": 0.9124554395675659,
+      "learning_rate": 0.0001836242655388293,
+      "loss": 3.2319,
+      "step": 1929
+    },
+    {
+      "epoch": 0.17470806553815516,
+      "grad_norm": 1.0809296369552612,
+      "learning_rate": 0.0001836068600922156,
+      "loss": 3.3081,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1747985878519055,
+      "grad_norm": 0.9386884570121765,
+      "learning_rate": 0.00018358944622639338,
+      "loss": 3.3148,
+      "step": 1931
+    },
+    {
+      "epoch": 0.17488911016565584,
+      "grad_norm": 0.9475247263908386,
+      "learning_rate": 0.00018357202394311624,
+      "loss": 3.2621,
+      "step": 1932
+    },
+    {
+      "epoch": 0.17497963247940618,
+      "grad_norm": 1.2415645122528076,
+      "learning_rate": 0.00018355459324413852,
+      "loss": 3.3089,
+      "step": 1933
+    },
+    {
+      "epoch": 0.17507015479315652,
+      "grad_norm": 0.897971510887146,
+      "learning_rate": 0.0001835371541312156,
+      "loss": 3.2613,
+      "step": 1934
+    },
+    {
+      "epoch": 0.17516067710690686,
+      "grad_norm": 0.8742809891700745,
+      "learning_rate": 0.0001835197066061035,
+      "loss": 3.2324,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1752511994206572,
+      "grad_norm": 0.9875279068946838,
+      "learning_rate": 0.00018350225067055925,
+      "loss": 3.2903,
+      "step": 1936
+    },
+    {
+      "epoch": 0.17534172173440754,
+      "grad_norm": 0.8936052918434143,
+      "learning_rate": 0.00018348478632634066,
+      "loss": 3.2961,
+      "step": 1937
+    },
+    {
+      "epoch": 0.17543224404815788,
+      "grad_norm": 0.9748427271842957,
+      "learning_rate": 0.00018346731357520637,
+      "loss": 3.3293,
+      "step": 1938
+    },
+    {
+      "epoch": 0.17552276636190822,
+      "grad_norm": 1.0512956380844116,
+      "learning_rate": 0.00018344983241891586,
+      "loss": 3.2362,
+      "step": 1939
+    },
+    {
+      "epoch": 0.17561328867565856,
+      "grad_norm": 0.950809121131897,
+      "learning_rate": 0.00018343234285922953,
+      "loss": 3.27,
+      "step": 1940
+    },
+    {
+      "epoch": 0.17570381098940888,
+      "grad_norm": 1.1645820140838623,
+      "learning_rate": 0.00018341484489790854,
+      "loss": 3.3205,
+      "step": 1941
+    },
+    {
+      "epoch": 0.17579433330315922,
+      "grad_norm": 1.1008210182189941,
+      "learning_rate": 0.00018339733853671496,
+      "loss": 3.2902,
+      "step": 1942
+    },
+    {
+      "epoch": 0.17588485561690956,
+      "grad_norm": 1.1445972919464111,
+      "learning_rate": 0.00018337982377741166,
+      "loss": 3.319,
+      "step": 1943
+    },
+    {
+      "epoch": 0.1759753779306599,
+      "grad_norm": 1.0789272785186768,
+      "learning_rate": 0.00018336230062176244,
+      "loss": 3.2692,
+      "step": 1944
+    },
+    {
+      "epoch": 0.17606590024441024,
+      "grad_norm": 1.190246820449829,
+      "learning_rate": 0.00018334476907153177,
+      "loss": 3.337,
+      "step": 1945
+    },
+    {
+      "epoch": 0.17615642255816058,
+      "grad_norm": 1.0693939924240112,
+      "learning_rate": 0.00018332722912848515,
+      "loss": 3.1858,
+      "step": 1946
+    },
+    {
+      "epoch": 0.17624694487191092,
+      "grad_norm": 1.0451300144195557,
+      "learning_rate": 0.00018330968079438887,
+      "loss": 3.2647,
+      "step": 1947
+    },
+    {
+      "epoch": 0.17633746718566126,
+      "grad_norm": 0.9115802645683289,
+      "learning_rate": 0.00018329212407100994,
+      "loss": 3.2782,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1764279894994116,
+      "grad_norm": 1.0351324081420898,
+      "learning_rate": 0.00018327455896011645,
+      "loss": 3.3408,
+      "step": 1949
+    },
+    {
+      "epoch": 0.17651851181316194,
+      "grad_norm": 0.8210245370864868,
+      "learning_rate": 0.00018325698546347715,
+      "loss": 3.3034,
+      "step": 1950
+    },
+    {
+      "epoch": 0.17660903412691228,
+      "grad_norm": 0.9735020399093628,
+      "learning_rate": 0.00018323940358286168,
+      "loss": 3.3017,
+      "step": 1951
+    },
+    {
+      "epoch": 0.17669955644066263,
+      "grad_norm": 0.914596676826477,
+      "learning_rate": 0.00018322181332004056,
+      "loss": 3.2619,
+      "step": 1952
+    },
+    {
+      "epoch": 0.17679007875441297,
+      "grad_norm": 0.8887321949005127,
+      "learning_rate": 0.00018320421467678507,
+      "loss": 3.2938,
+      "step": 1953
+    },
+    {
+      "epoch": 0.1768806010681633,
+      "grad_norm": 0.851830005645752,
+      "learning_rate": 0.00018318660765486748,
+      "loss": 3.2433,
+      "step": 1954
+    },
+    {
+      "epoch": 0.17697112338191365,
+      "grad_norm": 0.9544591903686523,
+      "learning_rate": 0.00018316899225606078,
+      "loss": 3.2891,
+      "step": 1955
+    },
+    {
+      "epoch": 0.177061645695664,
+      "grad_norm": 0.9891889691352844,
+      "learning_rate": 0.00018315136848213884,
+      "loss": 3.2969,
+      "step": 1956
+    },
+    {
+      "epoch": 0.17715216800941433,
+      "grad_norm": 0.9165775179862976,
+      "learning_rate": 0.00018313373633487633,
+      "loss": 3.185,
+      "step": 1957
+    },
+    {
+      "epoch": 0.17724269032316467,
+      "grad_norm": 0.8629693388938904,
+      "learning_rate": 0.00018311609581604887,
+      "loss": 3.3367,
+      "step": 1958
+    },
+    {
+      "epoch": 0.177333212636915,
+      "grad_norm": 0.9266963601112366,
+      "learning_rate": 0.00018309844692743283,
+      "loss": 3.3069,
+      "step": 1959
+    },
+    {
+      "epoch": 0.17742373495066535,
+      "grad_norm": 0.865626335144043,
+      "learning_rate": 0.00018308078967080546,
+      "loss": 3.2557,
+      "step": 1960
+    },
+    {
+      "epoch": 0.17751425726441566,
+      "grad_norm": 0.9198414087295532,
+      "learning_rate": 0.00018306312404794487,
+      "loss": 3.2345,
+      "step": 1961
+    },
+    {
+      "epoch": 0.177604779578166,
+      "grad_norm": 1.0007754564285278,
+      "learning_rate": 0.0001830454500606299,
+      "loss": 3.3282,
+      "step": 1962
+    },
+    {
+      "epoch": 0.17769530189191635,
+      "grad_norm": 0.9740175604820251,
+      "learning_rate": 0.00018302776771064044,
+      "loss": 3.3237,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1777858242056667,
+      "grad_norm": 0.9359791278839111,
+      "learning_rate": 0.00018301007699975702,
+      "loss": 3.281,
+      "step": 1964
+    },
+    {
+      "epoch": 0.17787634651941703,
+      "grad_norm": 1.0587137937545776,
+      "learning_rate": 0.00018299237792976112,
+      "loss": 3.3509,
+      "step": 1965
+    },
+    {
+      "epoch": 0.17796686883316737,
+      "grad_norm": 0.9631637334823608,
+      "learning_rate": 0.00018297467050243501,
+      "loss": 3.3646,
+      "step": 1966
+    },
+    {
+      "epoch": 0.1780573911469177,
+      "grad_norm": 0.9102620482444763,
+      "learning_rate": 0.00018295695471956189,
+      "loss": 3.3086,
+      "step": 1967
+    },
+    {
+      "epoch": 0.17814791346066805,
+      "grad_norm": 0.9048988819122314,
+      "learning_rate": 0.0001829392305829257,
+      "loss": 3.2488,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1782384357744184,
+      "grad_norm": 0.9322419166564941,
+      "learning_rate": 0.00018292149809431122,
+      "loss": 3.242,
+      "step": 1969
+    },
+    {
+      "epoch": 0.17832895808816873,
+      "grad_norm": 1.042884349822998,
+      "learning_rate": 0.00018290375725550417,
+      "loss": 3.2302,
+      "step": 1970
+    },
+    {
+      "epoch": 0.17841948040191907,
+      "grad_norm": 0.9759350419044495,
+      "learning_rate": 0.00018288600806829104,
+      "loss": 3.2822,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1785100027156694,
+      "grad_norm": 0.9325250387191772,
+      "learning_rate": 0.00018286825053445918,
+      "loss": 3.2232,
+      "step": 1972
+    },
+    {
+      "epoch": 0.17860052502941975,
+      "grad_norm": 0.9702731370925903,
+      "learning_rate": 0.00018285048465579672,
+      "loss": 3.2642,
+      "step": 1973
+    },
+    {
+      "epoch": 0.1786910473431701,
+      "grad_norm": 0.9540538787841797,
+      "learning_rate": 0.00018283271043409273,
+      "loss": 3.2451,
+      "step": 1974
+    },
+    {
+      "epoch": 0.17878156965692044,
+      "grad_norm": 0.9841457605361938,
+      "learning_rate": 0.00018281492787113708,
+      "loss": 3.3384,
+      "step": 1975
+    },
+    {
+      "epoch": 0.17887209197067078,
+      "grad_norm": 0.9560434222221375,
+      "learning_rate": 0.00018279713696872047,
+      "loss": 3.372,
+      "step": 1976
+    },
+    {
+      "epoch": 0.17896261428442112,
+      "grad_norm": 1.0200730562210083,
+      "learning_rate": 0.0001827793377286344,
+      "loss": 3.3252,
+      "step": 1977
+    },
+    {
+      "epoch": 0.17905313659817146,
+      "grad_norm": 0.9988753795623779,
+      "learning_rate": 0.00018276153015267134,
+      "loss": 3.2427,
+      "step": 1978
+    },
+    {
+      "epoch": 0.1791436589119218,
+      "grad_norm": 0.9856618642807007,
+      "learning_rate": 0.0001827437142426244,
+      "loss": 3.3344,
+      "step": 1979
+    },
+    {
+      "epoch": 0.17923418122567214,
+      "grad_norm": 0.9482274055480957,
+      "learning_rate": 0.00018272589000028772,
+      "loss": 3.2712,
+      "step": 1980
+    },
+    {
+      "epoch": 0.17932470353942248,
+      "grad_norm": 1.0072088241577148,
+      "learning_rate": 0.00018270805742745617,
+      "loss": 3.2766,
+      "step": 1981
+    },
+    {
+      "epoch": 0.1794152258531728,
+      "grad_norm": 1.0293302536010742,
+      "learning_rate": 0.00018269021652592553,
+      "loss": 3.3456,
+      "step": 1982
+    },
+    {
+      "epoch": 0.17950574816692313,
+      "grad_norm": 1.1364372968673706,
+      "learning_rate": 0.00018267236729749232,
+      "loss": 3.3901,
+      "step": 1983
+    },
+    {
+      "epoch": 0.17959627048067348,
+      "grad_norm": 1.0293986797332764,
+      "learning_rate": 0.00018265450974395403,
+      "loss": 3.2848,
+      "step": 1984
+    },
+    {
+      "epoch": 0.17968679279442382,
+      "grad_norm": 0.8650600910186768,
+      "learning_rate": 0.00018263664386710882,
+      "loss": 3.2819,
+      "step": 1985
+    },
+    {
+      "epoch": 0.17977731510817416,
+      "grad_norm": 0.9513899087905884,
+      "learning_rate": 0.00018261876966875584,
+      "loss": 3.3258,
+      "step": 1986
+    },
+    {
+      "epoch": 0.1798678374219245,
+      "grad_norm": 0.9329274892807007,
+      "learning_rate": 0.00018260088715069505,
+      "loss": 3.2609,
+      "step": 1987
+    },
+    {
+      "epoch": 0.17995835973567484,
+      "grad_norm": 1.0387474298477173,
+      "learning_rate": 0.00018258299631472715,
+      "loss": 3.2725,
+      "step": 1988
+    },
+    {
+      "epoch": 0.18004888204942518,
+      "grad_norm": 0.8893908858299255,
+      "learning_rate": 0.0001825650971626538,
+      "loss": 3.2557,
+      "step": 1989
+    },
+    {
+      "epoch": 0.18013940436317552,
+      "grad_norm": 1.0605710744857788,
+      "learning_rate": 0.0001825471896962774,
+      "loss": 3.1661,
+      "step": 1990
+    },
+    {
+      "epoch": 0.18022992667692586,
+      "grad_norm": 1.0651538372039795,
+      "learning_rate": 0.00018252927391740128,
+      "loss": 3.305,
+      "step": 1991
+    },
+    {
+      "epoch": 0.1803204489906762,
+      "grad_norm": 1.0231541395187378,
+      "learning_rate": 0.00018251134982782952,
+      "loss": 3.3347,
+      "step": 1992
+    },
+    {
+      "epoch": 0.18041097130442654,
+      "grad_norm": 1.2418352365493774,
+      "learning_rate": 0.0001824934174293671,
+      "loss": 3.284,
+      "step": 1993
+    },
+    {
+      "epoch": 0.18050149361817688,
+      "grad_norm": 1.0233439207077026,
+      "learning_rate": 0.0001824754767238198,
+      "loss": 3.3628,
+      "step": 1994
+    },
+    {
+      "epoch": 0.18059201593192722,
+      "grad_norm": 1.1309494972229004,
+      "learning_rate": 0.00018245752771299425,
+      "loss": 3.2799,
+      "step": 1995
+    },
+    {
+      "epoch": 0.18068253824567757,
+      "grad_norm": 1.152944564819336,
+      "learning_rate": 0.0001824395703986979,
+      "loss": 3.271,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1807730605594279,
+      "grad_norm": 0.96519535779953,
+      "learning_rate": 0.00018242160478273907,
+      "loss": 3.2316,
+      "step": 1997
+    },
+    {
+      "epoch": 0.18086358287317825,
+      "grad_norm": 0.9933448433876038,
+      "learning_rate": 0.00018240363086692693,
+      "loss": 3.3146,
+      "step": 1998
+    },
+    {
+      "epoch": 0.1809541051869286,
+      "grad_norm": 0.9097755551338196,
+      "learning_rate": 0.00018238564865307138,
+      "loss": 3.3065,
+      "step": 1999
+    },
+    {
+      "epoch": 0.18104462750067893,
+      "grad_norm": 1.1705602407455444,
+      "learning_rate": 0.0001823676581429833,
+      "loss": 3.2106,
+      "step": 2000
+    },
+    {
+      "epoch": 0.18104462750067893,
+      "eval_loss": 3.209501028060913,
+      "eval_runtime": 71.6327,
+      "eval_samples_per_second": 37.734,
+      "eval_steps_per_second": 3.155,
+      "step": 2000
+    },
+    {
+      "epoch": 0.18113514981442927,
+      "grad_norm": 1.0002061128616333,
+      "learning_rate": 0.00018234965933847428,
+      "loss": 3.2278,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1812256721281796,
+      "grad_norm": 1.1755338907241821,
+      "learning_rate": 0.00018233165224135678,
+      "loss": 3.2653,
+      "step": 2002
+    },
+    {
+      "epoch": 0.18131619444192992,
+      "grad_norm": 0.9064604043960571,
+      "learning_rate": 0.0001823136368534442,
+      "loss": 3.2967,
+      "step": 2003
+    },
+    {
+      "epoch": 0.18140671675568026,
+      "grad_norm": 0.9214131236076355,
+      "learning_rate": 0.00018229561317655062,
+      "loss": 3.3147,
+      "step": 2004
+    },
+    {
+      "epoch": 0.1814972390694306,
+      "grad_norm": 0.8722012042999268,
+      "learning_rate": 0.00018227758121249106,
+      "loss": 3.2075,
+      "step": 2005
+    },
+    {
+      "epoch": 0.18158776138318095,
+      "grad_norm": 0.9610263705253601,
+      "learning_rate": 0.00018225954096308131,
+      "loss": 3.2718,
+      "step": 2006
+    },
+    {
+      "epoch": 0.1816782836969313,
+      "grad_norm": 0.9243883490562439,
+      "learning_rate": 0.00018224149243013804,
+      "loss": 3.2456,
+      "step": 2007
+    },
+    {
+      "epoch": 0.18176880601068163,
+      "grad_norm": 0.9212303757667542,
+      "learning_rate": 0.00018222343561547874,
+      "loss": 3.2107,
+      "step": 2008
+    },
+    {
+      "epoch": 0.18185932832443197,
+      "grad_norm": 0.8368549942970276,
+      "learning_rate": 0.0001822053705209217,
+      "loss": 3.2671,
+      "step": 2009
+    },
+    {
+      "epoch": 0.1819498506381823,
+      "grad_norm": 0.9258269667625427,
+      "learning_rate": 0.00018218729714828612,
+      "loss": 3.2758,
+      "step": 2010
+    },
+    {
+      "epoch": 0.18204037295193265,
+      "grad_norm": 0.9486406445503235,
+      "learning_rate": 0.00018216921549939197,
+      "loss": 3.2163,
+      "step": 2011
+    },
+    {
+      "epoch": 0.182130895265683,
+      "grad_norm": 0.9499784111976624,
+      "learning_rate": 0.00018215112557606005,
+      "loss": 3.2586,
+      "step": 2012
+    },
+    {
+      "epoch": 0.18222141757943333,
+      "grad_norm": 0.9287562370300293,
+      "learning_rate": 0.00018213302738011203,
+      "loss": 3.3807,
+      "step": 2013
+    },
+    {
+      "epoch": 0.18231193989318367,
+      "grad_norm": 1.0922387838363647,
+      "learning_rate": 0.00018211492091337042,
+      "loss": 3.2507,
+      "step": 2014
+    },
+    {
+      "epoch": 0.182402462206934,
+      "grad_norm": 0.9361944198608398,
+      "learning_rate": 0.00018209680617765848,
+      "loss": 3.3209,
+      "step": 2015
+    },
+    {
+      "epoch": 0.18249298452068435,
+      "grad_norm": 0.9226075410842896,
+      "learning_rate": 0.00018207868317480046,
+      "loss": 3.2744,
+      "step": 2016
+    },
+    {
+      "epoch": 0.1825835068344347,
+      "grad_norm": 0.9788509607315063,
+      "learning_rate": 0.00018206055190662126,
+      "loss": 3.2184,
+      "step": 2017
+    },
+    {
+      "epoch": 0.18267402914818504,
+      "grad_norm": 1.0086486339569092,
+      "learning_rate": 0.00018204241237494673,
+      "loss": 3.3117,
+      "step": 2018
+    },
+    {
+      "epoch": 0.18276455146193538,
+      "grad_norm": 1.0230251550674438,
+      "learning_rate": 0.00018202426458160354,
+      "loss": 3.3178,
+      "step": 2019
+    },
+    {
+      "epoch": 0.18285507377568572,
+      "grad_norm": 0.9822840690612793,
+      "learning_rate": 0.00018200610852841913,
+      "loss": 3.309,
+      "step": 2020
+    },
+    {
+      "epoch": 0.18294559608943606,
+      "grad_norm": 1.0047194957733154,
+      "learning_rate": 0.00018198794421722182,
+      "loss": 3.222,
+      "step": 2021
+    },
+    {
+      "epoch": 0.1830361184031864,
+      "grad_norm": 0.923944354057312,
+      "learning_rate": 0.0001819697716498408,
+      "loss": 3.2258,
+      "step": 2022
+    },
+    {
+      "epoch": 0.1831266407169367,
+      "grad_norm": 0.9885908961296082,
+      "learning_rate": 0.000181951590828106,
+      "loss": 3.2468,
+      "step": 2023
+    },
+    {
+      "epoch": 0.18321716303068705,
+      "grad_norm": 0.9848500490188599,
+      "learning_rate": 0.00018193340175384824,
+      "loss": 3.2564,
+      "step": 2024
+    },
+    {
+      "epoch": 0.1833076853444374,
+      "grad_norm": 0.9301328659057617,
+      "learning_rate": 0.0001819152044288992,
+      "loss": 3.2512,
+      "step": 2025
+    },
+    {
+      "epoch": 0.18339820765818773,
+      "grad_norm": 0.9676833152770996,
+      "learning_rate": 0.00018189699885509127,
+      "loss": 3.3275,
+      "step": 2026
+    },
+    {
+      "epoch": 0.18348872997193807,
+      "grad_norm": 1.0562924146652222,
+      "learning_rate": 0.00018187878503425782,
+      "loss": 3.2531,
+      "step": 2027
+    },
+    {
+      "epoch": 0.18357925228568842,
+      "grad_norm": 0.8997370600700378,
+      "learning_rate": 0.00018186056296823298,
+      "loss": 3.2492,
+      "step": 2028
+    },
+    {
+      "epoch": 0.18366977459943876,
+      "grad_norm": 1.132249355316162,
+      "learning_rate": 0.00018184233265885165,
+      "loss": 3.263,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1837602969131891,
+      "grad_norm": 0.9985945224761963,
+      "learning_rate": 0.00018182409410794968,
+      "loss": 3.2512,
+      "step": 2030
+    },
+    {
+      "epoch": 0.18385081922693944,
+      "grad_norm": 0.9604045748710632,
+      "learning_rate": 0.00018180584731736365,
+      "loss": 3.3035,
+      "step": 2031
+    },
+    {
+      "epoch": 0.18394134154068978,
+      "grad_norm": 0.9413614273071289,
+      "learning_rate": 0.00018178759228893108,
+      "loss": 3.2909,
+      "step": 2032
+    },
+    {
+      "epoch": 0.18403186385444012,
+      "grad_norm": 1.2012258768081665,
+      "learning_rate": 0.00018176932902449016,
+      "loss": 3.2334,
+      "step": 2033
+    },
+    {
+      "epoch": 0.18412238616819046,
+      "grad_norm": 0.8994251489639282,
+      "learning_rate": 0.00018175105752588007,
+      "loss": 3.2546,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1842129084819408,
+      "grad_norm": 0.9620054960250854,
+      "learning_rate": 0.0001817327777949407,
+      "loss": 3.3143,
+      "step": 2035
+    },
+    {
+      "epoch": 0.18430343079569114,
+      "grad_norm": 0.9758648872375488,
+      "learning_rate": 0.00018171448983351284,
+      "loss": 3.2648,
+      "step": 2036
+    },
+    {
+      "epoch": 0.18439395310944148,
+      "grad_norm": 0.9170195460319519,
+      "learning_rate": 0.00018169619364343813,
+      "loss": 3.2282,
+      "step": 2037
+    },
+    {
+      "epoch": 0.18448447542319182,
+      "grad_norm": 0.9589044451713562,
+      "learning_rate": 0.00018167788922655894,
+      "loss": 3.2508,
+      "step": 2038
+    },
+    {
+      "epoch": 0.18457499773694216,
+      "grad_norm": 0.948966383934021,
+      "learning_rate": 0.00018165957658471853,
+      "loss": 3.2873,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1846655200506925,
+      "grad_norm": 1.168324589729309,
+      "learning_rate": 0.00018164125571976098,
+      "loss": 3.3223,
+      "step": 2040
+    },
+    {
+      "epoch": 0.18475604236444285,
+      "grad_norm": 0.9419066905975342,
+      "learning_rate": 0.00018162292663353124,
+      "loss": 3.2272,
+      "step": 2041
+    },
+    {
+      "epoch": 0.1848465646781932,
+      "grad_norm": 0.9817954301834106,
+      "learning_rate": 0.000181604589327875,
+      "loss": 3.1878,
+      "step": 2042
+    },
+    {
+      "epoch": 0.18493708699194353,
+      "grad_norm": 0.8973686099052429,
+      "learning_rate": 0.00018158624380463888,
+      "loss": 3.2144,
+      "step": 2043
+    },
+    {
+      "epoch": 0.18502760930569384,
+      "grad_norm": 1.1221078634262085,
+      "learning_rate": 0.0001815678900656702,
+      "loss": 3.2616,
+      "step": 2044
+    },
+    {
+      "epoch": 0.18511813161944418,
+      "grad_norm": 0.8626366853713989,
+      "learning_rate": 0.00018154952811281724,
+      "loss": 3.2119,
+      "step": 2045
+    },
+    {
+      "epoch": 0.18520865393319452,
+      "grad_norm": 0.9677149653434753,
+      "learning_rate": 0.00018153115794792904,
+      "loss": 3.2749,
+      "step": 2046
+    },
+    {
+      "epoch": 0.18529917624694486,
+      "grad_norm": 0.9512866139411926,
+      "learning_rate": 0.00018151277957285543,
+      "loss": 3.2477,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1853896985606952,
+      "grad_norm": 1.0640791654586792,
+      "learning_rate": 0.00018149439298944717,
+      "loss": 3.198,
+      "step": 2048
+    },
+    {
+      "epoch": 0.18548022087444554,
+      "grad_norm": 1.0518122911453247,
+      "learning_rate": 0.0001814759981995558,
+      "loss": 3.3341,
+      "step": 2049
+    },
+    {
+      "epoch": 0.18557074318819589,
+      "grad_norm": 0.9965335130691528,
+      "learning_rate": 0.00018145759520503358,
+      "loss": 3.2755,
+      "step": 2050
+    },
+    {
+      "epoch": 0.18566126550194623,
+      "grad_norm": 0.8979808688163757,
+      "learning_rate": 0.00018143918400773376,
+      "loss": 3.2551,
+      "step": 2051
+    },
+    {
+      "epoch": 0.18575178781569657,
+      "grad_norm": 0.93203204870224,
+      "learning_rate": 0.0001814207646095104,
+      "loss": 3.2214,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1858423101294469,
+      "grad_norm": 0.9714310169219971,
+      "learning_rate": 0.0001814023370122182,
+      "loss": 3.3333,
+      "step": 2053
+    },
+    {
+      "epoch": 0.18593283244319725,
+      "grad_norm": 0.8723704218864441,
+      "learning_rate": 0.00018138390121771295,
+      "loss": 3.2398,
+      "step": 2054
+    },
+    {
+      "epoch": 0.1860233547569476,
+      "grad_norm": 0.9395724534988403,
+      "learning_rate": 0.00018136545722785102,
+      "loss": 3.3092,
+      "step": 2055
+    },
+    {
+      "epoch": 0.18611387707069793,
+      "grad_norm": 0.9963365793228149,
+      "learning_rate": 0.0001813470050444898,
+      "loss": 3.2367,
+      "step": 2056
+    },
+    {
+      "epoch": 0.18620439938444827,
+      "grad_norm": 0.9642789363861084,
+      "learning_rate": 0.0001813285446694874,
+      "loss": 3.2365,
+      "step": 2057
+    },
+    {
+      "epoch": 0.1862949216981986,
+      "grad_norm": 1.004480242729187,
+      "learning_rate": 0.00018131007610470276,
+      "loss": 3.2573,
+      "step": 2058
+    },
+    {
+      "epoch": 0.18638544401194895,
+      "grad_norm": 0.8809323310852051,
+      "learning_rate": 0.00018129159935199572,
+      "loss": 3.1922,
+      "step": 2059
+    },
+    {
+      "epoch": 0.1864759663256993,
+      "grad_norm": 1.2615585327148438,
+      "learning_rate": 0.0001812731144132268,
+      "loss": 3.2372,
+      "step": 2060
+    },
+    {
+      "epoch": 0.18656648863944963,
+      "grad_norm": 0.8767595887184143,
+      "learning_rate": 0.00018125462129025752,
+      "loss": 3.1661,
+      "step": 2061
+    },
+    {
+      "epoch": 0.18665701095319998,
+      "grad_norm": 1.0361719131469727,
+      "learning_rate": 0.00018123611998495007,
+      "loss": 3.2573,
+      "step": 2062
+    },
+    {
+      "epoch": 0.18674753326695032,
+      "grad_norm": 0.9326725006103516,
+      "learning_rate": 0.0001812176104991676,
+      "loss": 3.1669,
+      "step": 2063
+    },
+    {
+      "epoch": 0.18683805558070063,
+      "grad_norm": 1.1481561660766602,
+      "learning_rate": 0.00018119909283477394,
+      "loss": 3.2874,
+      "step": 2064
+    },
+    {
+      "epoch": 0.18692857789445097,
+      "grad_norm": 0.9084659218788147,
+      "learning_rate": 0.00018118056699363387,
+      "loss": 3.3139,
+      "step": 2065
+    },
+    {
+      "epoch": 0.1870191002082013,
+      "grad_norm": 0.9566446542739868,
+      "learning_rate": 0.00018116203297761292,
+      "loss": 3.318,
+      "step": 2066
+    },
+    {
+      "epoch": 0.18710962252195165,
+      "grad_norm": 0.8745667338371277,
+      "learning_rate": 0.00018114349078857746,
+      "loss": 3.1519,
+      "step": 2067
+    },
+    {
+      "epoch": 0.187200144835702,
+      "grad_norm": 0.954801082611084,
+      "learning_rate": 0.0001811249404283947,
+      "loss": 3.221,
+      "step": 2068
+    },
+    {
+      "epoch": 0.18729066714945233,
+      "grad_norm": 0.8902572989463806,
+      "learning_rate": 0.00018110638189893267,
+      "loss": 3.2083,
+      "step": 2069
+    },
+    {
+      "epoch": 0.18738118946320267,
+      "grad_norm": 0.880317211151123,
+      "learning_rate": 0.0001810878152020602,
+      "loss": 3.1926,
+      "step": 2070
+    },
+    {
+      "epoch": 0.18747171177695302,
+      "grad_norm": 1.0060982704162598,
+      "learning_rate": 0.00018106924033964696,
+      "loss": 3.2319,
+      "step": 2071
+    },
+    {
+      "epoch": 0.18756223409070336,
+      "grad_norm": 0.9899522662162781,
+      "learning_rate": 0.00018105065731356343,
+      "loss": 3.1814,
+      "step": 2072
+    },
+    {
+      "epoch": 0.1876527564044537,
+      "grad_norm": 0.8588231205940247,
+      "learning_rate": 0.00018103206612568094,
+      "loss": 3.21,
+      "step": 2073
+    },
+    {
+      "epoch": 0.18774327871820404,
+      "grad_norm": 0.9145846962928772,
+      "learning_rate": 0.00018101346677787156,
+      "loss": 3.2141,
+      "step": 2074
+    },
+    {
+      "epoch": 0.18783380103195438,
+      "grad_norm": 0.9708083271980286,
+      "learning_rate": 0.00018099485927200836,
+      "loss": 3.2744,
+      "step": 2075
+    },
+    {
+      "epoch": 0.18792432334570472,
+      "grad_norm": 0.9268299341201782,
+      "learning_rate": 0.000180976243609965,
+      "loss": 3.1685,
+      "step": 2076
+    },
+    {
+      "epoch": 0.18801484565945506,
+      "grad_norm": 0.9239932894706726,
+      "learning_rate": 0.00018095761979361615,
+      "loss": 3.2955,
+      "step": 2077
+    },
+    {
+      "epoch": 0.1881053679732054,
+      "grad_norm": 0.9097129702568054,
+      "learning_rate": 0.00018093898782483718,
+      "loss": 3.3611,
+      "step": 2078
+    },
+    {
+      "epoch": 0.18819589028695574,
+      "grad_norm": 1.1154252290725708,
+      "learning_rate": 0.00018092034770550436,
+      "loss": 3.2398,
+      "step": 2079
+    },
+    {
+      "epoch": 0.18828641260070608,
+      "grad_norm": 0.9370711445808411,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 3.2075,
+      "step": 2080
+    },
+    {
+      "epoch": 0.18837693491445642,
+      "grad_norm": 0.9421517252922058,
+      "learning_rate": 0.00018088304302268623,
+      "loss": 3.1681,
+      "step": 2081
+    },
+    {
+      "epoch": 0.18846745722820676,
+      "grad_norm": 0.8839324712753296,
+      "learning_rate": 0.00018086437846295746,
+      "loss": 3.2357,
+      "step": 2082
+    },
+    {
+      "epoch": 0.1885579795419571,
+      "grad_norm": 1.0185707807540894,
+      "learning_rate": 0.000180845705760188,
+      "loss": 3.2913,
+      "step": 2083
+    },
+    {
+      "epoch": 0.18864850185570745,
+      "grad_norm": 0.9895525574684143,
+      "learning_rate": 0.0001808270249162582,
+      "loss": 3.3165,
+      "step": 2084
+    },
+    {
+      "epoch": 0.18873902416945776,
+      "grad_norm": 0.9668818712234497,
+      "learning_rate": 0.00018080833593304917,
+      "loss": 3.2791,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1888295464832081,
+      "grad_norm": 1.1404991149902344,
+      "learning_rate": 0.00018078963881244296,
+      "loss": 3.3115,
+      "step": 2086
+    },
+    {
+      "epoch": 0.18892006879695844,
+      "grad_norm": 0.9569199085235596,
+      "learning_rate": 0.0001807709335563223,
+      "loss": 3.265,
+      "step": 2087
+    },
+    {
+      "epoch": 0.18901059111070878,
+      "grad_norm": 0.8879489302635193,
+      "learning_rate": 0.00018075222016657088,
+      "loss": 3.3034,
+      "step": 2088
+    },
+    {
+      "epoch": 0.18910111342445912,
+      "grad_norm": 1.0276618003845215,
+      "learning_rate": 0.00018073349864507305,
+      "loss": 3.2423,
+      "step": 2089
+    },
+    {
+      "epoch": 0.18919163573820946,
+      "grad_norm": 1.0222197771072388,
+      "learning_rate": 0.00018071476899371414,
+      "loss": 3.2373,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1892821580519598,
+      "grad_norm": 0.9289442896842957,
+      "learning_rate": 0.00018069603121438022,
+      "loss": 3.173,
+      "step": 2091
+    },
+    {
+      "epoch": 0.18937268036571014,
+      "grad_norm": 0.9327288866043091,
+      "learning_rate": 0.00018067728530895815,
+      "loss": 3.2297,
+      "step": 2092
+    },
+    {
+      "epoch": 0.18946320267946049,
+      "grad_norm": 0.9465458393096924,
+      "learning_rate": 0.00018065853127933562,
+      "loss": 3.225,
+      "step": 2093
+    },
+    {
+      "epoch": 0.18955372499321083,
+      "grad_norm": 0.9022896885871887,
+      "learning_rate": 0.00018063976912740124,
+      "loss": 3.2729,
+      "step": 2094
+    },
+    {
+      "epoch": 0.18964424730696117,
+      "grad_norm": 0.9528231024742126,
+      "learning_rate": 0.0001806209988550443,
+      "loss": 3.268,
+      "step": 2095
+    },
+    {
+      "epoch": 0.1897347696207115,
+      "grad_norm": 0.9072818756103516,
+      "learning_rate": 0.000180602220464155,
+      "loss": 3.2786,
+      "step": 2096
+    },
+    {
+      "epoch": 0.18982529193446185,
+      "grad_norm": 0.8678036332130432,
+      "learning_rate": 0.00018058343395662427,
+      "loss": 3.2,
+      "step": 2097
+    },
+    {
+      "epoch": 0.1899158142482122,
+      "grad_norm": 0.9147240519523621,
+      "learning_rate": 0.00018056463933434398,
+      "loss": 3.2329,
+      "step": 2098
+    },
+    {
+      "epoch": 0.19000633656196253,
+      "grad_norm": 0.9179824590682983,
+      "learning_rate": 0.00018054583659920669,
+      "loss": 3.2009,
+      "step": 2099
+    },
+    {
+      "epoch": 0.19009685887571287,
+      "grad_norm": 0.9133867621421814,
+      "learning_rate": 0.00018052702575310588,
+      "loss": 3.1839,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1901873811894632,
+      "grad_norm": 0.9742360711097717,
+      "learning_rate": 0.00018050820679793575,
+      "loss": 3.1951,
+      "step": 2101
+    },
+    {
+      "epoch": 0.19027790350321355,
+      "grad_norm": 0.9586870670318604,
+      "learning_rate": 0.0001804893797355914,
+      "loss": 3.276,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1903684258169639,
+      "grad_norm": 1.0903187990188599,
+      "learning_rate": 0.00018047054456796877,
+      "loss": 3.1827,
+      "step": 2103
+    },
+    {
+      "epoch": 0.19045894813071423,
+      "grad_norm": 1.017938494682312,
+      "learning_rate": 0.0001804517012969644,
+      "loss": 3.2418,
+      "step": 2104
+    },
+    {
+      "epoch": 0.19054947044446455,
+      "grad_norm": 0.997882604598999,
+      "learning_rate": 0.00018043284992447602,
+      "loss": 3.2032,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1906399927582149,
+      "grad_norm": 1.032028079032898,
+      "learning_rate": 0.00018041399045240183,
+      "loss": 3.2286,
+      "step": 2106
+    },
+    {
+      "epoch": 0.19073051507196523,
+      "grad_norm": 1.2459874153137207,
+      "learning_rate": 0.00018039512288264096,
+      "loss": 3.2662,
+      "step": 2107
+    },
+    {
+      "epoch": 0.19082103738571557,
+      "grad_norm": 1.0763884782791138,
+      "learning_rate": 0.00018037624721709342,
+      "loss": 3.2846,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1909115596994659,
+      "grad_norm": 1.4135953187942505,
+      "learning_rate": 0.00018035736345765999,
+      "loss": 3.1769,
+      "step": 2109
+    },
+    {
+      "epoch": 0.19100208201321625,
+      "grad_norm": 1.0524734258651733,
+      "learning_rate": 0.00018033847160624225,
+      "loss": 3.2379,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1910926043269666,
+      "grad_norm": 1.238844633102417,
+      "learning_rate": 0.00018031957166474264,
+      "loss": 3.2223,
+      "step": 2111
+    },
+    {
+      "epoch": 0.19118312664071693,
+      "grad_norm": 1.1036057472229004,
+      "learning_rate": 0.00018030066363506437,
+      "loss": 3.3338,
+      "step": 2112
+    },
+    {
+      "epoch": 0.19127364895446727,
+      "grad_norm": 1.2267645597457886,
+      "learning_rate": 0.00018028174751911146,
+      "loss": 3.2718,
+      "step": 2113
+    },
+    {
+      "epoch": 0.19136417126821761,
+      "grad_norm": 1.2532908916473389,
+      "learning_rate": 0.00018026282331878876,
+      "loss": 3.2186,
+      "step": 2114
+    },
+    {
+      "epoch": 0.19145469358196796,
+      "grad_norm": 0.913084089756012,
+      "learning_rate": 0.00018024389103600197,
+      "loss": 3.2445,
+      "step": 2115
+    },
+    {
+      "epoch": 0.1915452158957183,
+      "grad_norm": 1.1660557985305786,
+      "learning_rate": 0.00018022495067265753,
+      "loss": 3.1916,
+      "step": 2116
+    },
+    {
+      "epoch": 0.19163573820946864,
+      "grad_norm": 0.9609684348106384,
+      "learning_rate": 0.00018020600223066276,
+      "loss": 3.2422,
+      "step": 2117
+    },
+    {
+      "epoch": 0.19172626052321898,
+      "grad_norm": 0.9922468066215515,
+      "learning_rate": 0.00018018704571192573,
+      "loss": 3.2267,
+      "step": 2118
+    },
+    {
+      "epoch": 0.19181678283696932,
+      "grad_norm": 1.00340735912323,
+      "learning_rate": 0.00018016808111835544,
+      "loss": 3.2951,
+      "step": 2119
+    },
+    {
+      "epoch": 0.19190730515071966,
+      "grad_norm": 0.9352942109107971,
+      "learning_rate": 0.00018014910845186153,
+      "loss": 3.2118,
+      "step": 2120
+    },
+    {
+      "epoch": 0.19199782746447,
+      "grad_norm": 0.9789060950279236,
+      "learning_rate": 0.0001801301277143546,
+      "loss": 3.3368,
+      "step": 2121
+    },
+    {
+      "epoch": 0.19208834977822034,
+      "grad_norm": 0.9243131279945374,
+      "learning_rate": 0.00018011113890774603,
+      "loss": 3.2403,
+      "step": 2122
+    },
+    {
+      "epoch": 0.19217887209197068,
+      "grad_norm": 0.943941593170166,
+      "learning_rate": 0.00018009214203394794,
+      "loss": 3.2127,
+      "step": 2123
+    },
+    {
+      "epoch": 0.19226939440572102,
+      "grad_norm": 0.9229180216789246,
+      "learning_rate": 0.00018007313709487334,
+      "loss": 3.2846,
+      "step": 2124
+    },
+    {
+      "epoch": 0.19235991671947136,
+      "grad_norm": 1.020640254020691,
+      "learning_rate": 0.00018005412409243606,
+      "loss": 3.2412,
+      "step": 2125
+    },
+    {
+      "epoch": 0.19245043903322168,
+      "grad_norm": 0.8972303867340088,
+      "learning_rate": 0.0001800351030285506,
+      "loss": 3.2709,
+      "step": 2126
+    },
+    {
+      "epoch": 0.19254096134697202,
+      "grad_norm": 0.9157093167304993,
+      "learning_rate": 0.00018001607390513255,
+      "loss": 3.2852,
+      "step": 2127
+    },
+    {
+      "epoch": 0.19263148366072236,
+      "grad_norm": 0.9684386849403381,
+      "learning_rate": 0.000179997036724098,
+      "loss": 3.2637,
+      "step": 2128
+    },
+    {
+      "epoch": 0.1927220059744727,
+      "grad_norm": 1.0669679641723633,
+      "learning_rate": 0.00017997799148736403,
+      "loss": 3.1715,
+      "step": 2129
+    },
+    {
+      "epoch": 0.19281252828822304,
+      "grad_norm": 0.9132657051086426,
+      "learning_rate": 0.0001799589381968485,
+      "loss": 3.2373,
+      "step": 2130
+    },
+    {
+      "epoch": 0.19290305060197338,
+      "grad_norm": 0.9766040444374084,
+      "learning_rate": 0.00017993987685447008,
+      "loss": 3.2287,
+      "step": 2131
+    },
+    {
+      "epoch": 0.19299357291572372,
+      "grad_norm": 0.9845170974731445,
+      "learning_rate": 0.0001799208074621483,
+      "loss": 3.2386,
+      "step": 2132
+    },
+    {
+      "epoch": 0.19308409522947406,
+      "grad_norm": 0.9120230674743652,
+      "learning_rate": 0.00017990173002180334,
+      "loss": 3.2479,
+      "step": 2133
+    },
+    {
+      "epoch": 0.1931746175432244,
+      "grad_norm": 0.8494027256965637,
+      "learning_rate": 0.0001798826445353564,
+      "loss": 3.1777,
+      "step": 2134
+    },
+    {
+      "epoch": 0.19326513985697474,
+      "grad_norm": 0.945387065410614,
+      "learning_rate": 0.00017986355100472928,
+      "loss": 3.2252,
+      "step": 2135
+    },
+    {
+      "epoch": 0.19335566217072508,
+      "grad_norm": 1.0038387775421143,
+      "learning_rate": 0.0001798444494318448,
+      "loss": 3.1945,
+      "step": 2136
+    },
+    {
+      "epoch": 0.19344618448447543,
+      "grad_norm": 0.9529719948768616,
+      "learning_rate": 0.00017982533981862644,
+      "loss": 3.1815,
+      "step": 2137
+    },
+    {
+      "epoch": 0.19353670679822577,
+      "grad_norm": 0.9867908954620361,
+      "learning_rate": 0.00017980622216699855,
+      "loss": 3.2314,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1936272291119761,
+      "grad_norm": 0.9234095215797424,
+      "learning_rate": 0.0001797870964788863,
+      "loss": 3.1586,
+      "step": 2139
+    },
+    {
+      "epoch": 0.19371775142572645,
+      "grad_norm": 0.9278466105461121,
+      "learning_rate": 0.00017976796275621555,
+      "loss": 3.2559,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1938082737394768,
+      "grad_norm": 0.9761223196983337,
+      "learning_rate": 0.00017974882100091317,
+      "loss": 3.2429,
+      "step": 2141
+    },
+    {
+      "epoch": 0.19389879605322713,
+      "grad_norm": 0.9403523206710815,
+      "learning_rate": 0.00017972967121490668,
+      "loss": 3.2106,
+      "step": 2142
+    },
+    {
+      "epoch": 0.19398931836697747,
+      "grad_norm": 0.9349404573440552,
+      "learning_rate": 0.0001797105134001245,
+      "loss": 3.2218,
+      "step": 2143
+    },
+    {
+      "epoch": 0.1940798406807278,
+      "grad_norm": 0.8340486884117126,
+      "learning_rate": 0.0001796913475584958,
+      "loss": 3.1981,
+      "step": 2144
+    },
+    {
+      "epoch": 0.19417036299447815,
+      "grad_norm": 0.920889139175415,
+      "learning_rate": 0.00017967217369195058,
+      "loss": 3.1651,
+      "step": 2145
+    },
+    {
+      "epoch": 0.19426088530822846,
+      "grad_norm": 0.8778406381607056,
+      "learning_rate": 0.00017965299180241963,
+      "loss": 3.2923,
+      "step": 2146
+    },
+    {
+      "epoch": 0.1943514076219788,
+      "grad_norm": 0.9380999207496643,
+      "learning_rate": 0.0001796338018918346,
+      "loss": 3.2636,
+      "step": 2147
+    },
+    {
+      "epoch": 0.19444192993572915,
+      "grad_norm": 0.9442136883735657,
+      "learning_rate": 0.0001796146039621279,
+      "loss": 3.2431,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1945324522494795,
+      "grad_norm": 0.8796325922012329,
+      "learning_rate": 0.00017959539801523278,
+      "loss": 3.1596,
+      "step": 2149
+    },
+    {
+      "epoch": 0.19462297456322983,
+      "grad_norm": 1.0169458389282227,
+      "learning_rate": 0.00017957618405308324,
+      "loss": 3.146,
+      "step": 2150
+    },
+    {
+      "epoch": 0.19471349687698017,
+      "grad_norm": 0.987337589263916,
+      "learning_rate": 0.00017955696207761412,
+      "loss": 3.1754,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1948040191907305,
+      "grad_norm": 1.0014207363128662,
+      "learning_rate": 0.0001795377320907611,
+      "loss": 3.2415,
+      "step": 2152
+    },
+    {
+      "epoch": 0.19489454150448085,
+      "grad_norm": 0.9423050880432129,
+      "learning_rate": 0.00017951849409446067,
+      "loss": 3.1492,
+      "step": 2153
+    },
+    {
+      "epoch": 0.1949850638182312,
+      "grad_norm": 1.1418346166610718,
+      "learning_rate": 0.00017949924809065003,
+      "loss": 3.2466,
+      "step": 2154
+    },
+    {
+      "epoch": 0.19507558613198153,
+      "grad_norm": 0.8354297876358032,
+      "learning_rate": 0.0001794799940812673,
+      "loss": 3.227,
+      "step": 2155
+    },
+    {
+      "epoch": 0.19516610844573187,
+      "grad_norm": 1.1176152229309082,
+      "learning_rate": 0.00017946073206825134,
+      "loss": 3.2342,
+      "step": 2156
+    },
+    {
+      "epoch": 0.1952566307594822,
+      "grad_norm": 0.9996242523193359,
+      "learning_rate": 0.00017944146205354182,
+      "loss": 3.3404,
+      "step": 2157
+    },
+    {
+      "epoch": 0.19534715307323255,
+      "grad_norm": 1.025205373764038,
+      "learning_rate": 0.00017942218403907924,
+      "loss": 3.1818,
+      "step": 2158
+    },
+    {
+      "epoch": 0.1954376753869829,
+      "grad_norm": 0.8781141638755798,
+      "learning_rate": 0.0001794028980268049,
+      "loss": 3.1567,
+      "step": 2159
+    },
+    {
+      "epoch": 0.19552819770073324,
+      "grad_norm": 0.9099708199501038,
+      "learning_rate": 0.00017938360401866093,
+      "loss": 3.2146,
+      "step": 2160
+    },
+    {
+      "epoch": 0.19561872001448358,
+      "grad_norm": 0.9119500517845154,
+      "learning_rate": 0.0001793643020165902,
+      "loss": 3.2271,
+      "step": 2161
+    },
+    {
+      "epoch": 0.19570924232823392,
+      "grad_norm": 0.8782615065574646,
+      "learning_rate": 0.00017934499202253642,
+      "loss": 3.1913,
+      "step": 2162
+    },
+    {
+      "epoch": 0.19579976464198426,
+      "grad_norm": 0.9160125255584717,
+      "learning_rate": 0.0001793256740384441,
+      "loss": 3.2224,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1958902869557346,
+      "grad_norm": 0.8718566298484802,
+      "learning_rate": 0.0001793063480662586,
+      "loss": 3.1911,
+      "step": 2164
+    },
+    {
+      "epoch": 0.19598080926948494,
+      "grad_norm": 0.9492814540863037,
+      "learning_rate": 0.00017928701410792602,
+      "loss": 3.1833,
+      "step": 2165
+    },
+    {
+      "epoch": 0.19607133158323528,
+      "grad_norm": 0.8263920545578003,
+      "learning_rate": 0.0001792676721653933,
+      "loss": 3.3056,
+      "step": 2166
+    },
+    {
+      "epoch": 0.1961618538969856,
+      "grad_norm": 0.9434691667556763,
+      "learning_rate": 0.0001792483222406081,
+      "loss": 3.2539,
+      "step": 2167
+    },
+    {
+      "epoch": 0.19625237621073593,
+      "grad_norm": 0.8706825375556946,
+      "learning_rate": 0.00017922896433551907,
+      "loss": 3.1788,
+      "step": 2168
+    },
+    {
+      "epoch": 0.19634289852448628,
+      "grad_norm": 0.9813132882118225,
+      "learning_rate": 0.0001792095984520755,
+      "loss": 3.2355,
+      "step": 2169
+    },
+    {
+      "epoch": 0.19643342083823662,
+      "grad_norm": 0.9587005972862244,
+      "learning_rate": 0.00017919022459222752,
+      "loss": 3.2232,
+      "step": 2170
+    },
+    {
+      "epoch": 0.19652394315198696,
+      "grad_norm": 0.9128755927085876,
+      "learning_rate": 0.0001791708427579261,
+      "loss": 3.1844,
+      "step": 2171
+    },
+    {
+      "epoch": 0.1966144654657373,
+      "grad_norm": 0.89957195520401,
+      "learning_rate": 0.00017915145295112296,
+      "loss": 3.1824,
+      "step": 2172
+    },
+    {
+      "epoch": 0.19670498777948764,
+      "grad_norm": 0.8944728374481201,
+      "learning_rate": 0.0001791320551737707,
+      "loss": 3.2404,
+      "step": 2173
+    },
+    {
+      "epoch": 0.19679551009323798,
+      "grad_norm": 1.0140613317489624,
+      "learning_rate": 0.00017911264942782262,
+      "loss": 3.2424,
+      "step": 2174
+    },
+    {
+      "epoch": 0.19688603240698832,
+      "grad_norm": 0.9342913627624512,
+      "learning_rate": 0.00017909323571523294,
+      "loss": 3.2279,
+      "step": 2175
+    },
+    {
+      "epoch": 0.19697655472073866,
+      "grad_norm": 0.9111455678939819,
+      "learning_rate": 0.00017907381403795656,
+      "loss": 3.2082,
+      "step": 2176
+    },
+    {
+      "epoch": 0.197067077034489,
+      "grad_norm": 0.9179486036300659,
+      "learning_rate": 0.00017905438439794929,
+      "loss": 3.1214,
+      "step": 2177
+    },
+    {
+      "epoch": 0.19715759934823934,
+      "grad_norm": 0.9826908111572266,
+      "learning_rate": 0.00017903494679716767,
+      "loss": 3.228,
+      "step": 2178
+    },
+    {
+      "epoch": 0.19724812166198968,
+      "grad_norm": 0.9230121374130249,
+      "learning_rate": 0.00017901550123756906,
+      "loss": 3.1306,
+      "step": 2179
+    },
+    {
+      "epoch": 0.19733864397574002,
+      "grad_norm": 0.9603817462921143,
+      "learning_rate": 0.00017899604772111163,
+      "loss": 3.2442,
+      "step": 2180
+    },
+    {
+      "epoch": 0.19742916628949037,
+      "grad_norm": 0.9203359484672546,
+      "learning_rate": 0.00017897658624975438,
+      "loss": 3.1501,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1975196886032407,
+      "grad_norm": 0.9163203835487366,
+      "learning_rate": 0.00017895711682545704,
+      "loss": 3.2631,
+      "step": 2182
+    },
+    {
+      "epoch": 0.19761021091699105,
+      "grad_norm": 0.8634765148162842,
+      "learning_rate": 0.0001789376394501802,
+      "loss": 3.2008,
+      "step": 2183
+    },
+    {
+      "epoch": 0.1977007332307414,
+      "grad_norm": 0.9373166561126709,
+      "learning_rate": 0.00017891815412588523,
+      "loss": 3.1603,
+      "step": 2184
+    },
+    {
+      "epoch": 0.19779125554449173,
+      "grad_norm": 1.0072497129440308,
+      "learning_rate": 0.00017889866085453427,
+      "loss": 3.2015,
+      "step": 2185
+    },
+    {
+      "epoch": 0.19788177785824207,
+      "grad_norm": 0.8856438398361206,
+      "learning_rate": 0.00017887915963809038,
+      "loss": 3.2296,
+      "step": 2186
+    },
+    {
+      "epoch": 0.19797230017199238,
+      "grad_norm": 0.9211776852607727,
+      "learning_rate": 0.0001788596504785172,
+      "loss": 3.2378,
+      "step": 2187
+    },
+    {
+      "epoch": 0.19806282248574272,
+      "grad_norm": 0.9734625220298767,
+      "learning_rate": 0.00017884013337777943,
+      "loss": 3.1718,
+      "step": 2188
+    },
+    {
+      "epoch": 0.19815334479949306,
+      "grad_norm": 0.9697534441947937,
+      "learning_rate": 0.00017882060833784237,
+      "loss": 3.2205,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1982438671132434,
+      "grad_norm": 0.9940907955169678,
+      "learning_rate": 0.00017880107536067218,
+      "loss": 3.2506,
+      "step": 2190
+    },
+    {
+      "epoch": 0.19833438942699375,
+      "grad_norm": 0.9850450158119202,
+      "learning_rate": 0.00017878153444823592,
+      "loss": 3.2165,
+      "step": 2191
+    },
+    {
+      "epoch": 0.1984249117407441,
+      "grad_norm": 0.9790923595428467,
+      "learning_rate": 0.00017876198560250127,
+      "loss": 3.1925,
+      "step": 2192
+    },
+    {
+      "epoch": 0.19851543405449443,
+      "grad_norm": 0.9181186556816101,
+      "learning_rate": 0.0001787424288254368,
+      "loss": 3.196,
+      "step": 2193
+    },
+    {
+      "epoch": 0.19860595636824477,
+      "grad_norm": 0.924216628074646,
+      "learning_rate": 0.00017872286411901191,
+      "loss": 3.1773,
+      "step": 2194
+    },
+    {
+      "epoch": 0.1986964786819951,
+      "grad_norm": 0.9113030433654785,
+      "learning_rate": 0.00017870329148519676,
+      "loss": 3.1877,
+      "step": 2195
+    },
+    {
+      "epoch": 0.19878700099574545,
+      "grad_norm": 1.0087077617645264,
+      "learning_rate": 0.00017868371092596232,
+      "loss": 3.3046,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1988775233094958,
+      "grad_norm": 0.9603467583656311,
+      "learning_rate": 0.00017866412244328035,
+      "loss": 3.2462,
+      "step": 2197
+    },
+    {
+      "epoch": 0.19896804562324613,
+      "grad_norm": 1.048497200012207,
+      "learning_rate": 0.00017864452603912339,
+      "loss": 3.1842,
+      "step": 2198
+    },
+    {
+      "epoch": 0.19905856793699647,
+      "grad_norm": 0.9084252715110779,
+      "learning_rate": 0.00017862492171546478,
+      "loss": 3.1238,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1991490902507468,
+      "grad_norm": 1.100671410560608,
+      "learning_rate": 0.00017860530947427875,
+      "loss": 3.2618,
+      "step": 2200
+    },
+    {
+      "epoch": 0.19923961256449715,
+      "grad_norm": 1.1458168029785156,
+      "learning_rate": 0.0001785856893175402,
+      "loss": 3.185,
+      "step": 2201
+    },
+    {
+      "epoch": 0.1993301348782475,
+      "grad_norm": 1.094252109527588,
+      "learning_rate": 0.0001785660612472249,
+      "loss": 3.2558,
+      "step": 2202
+    },
+    {
+      "epoch": 0.19942065719199784,
+      "grad_norm": 0.8893122673034668,
+      "learning_rate": 0.00017854642526530934,
+      "loss": 3.1477,
+      "step": 2203
+    },
+    {
+      "epoch": 0.19951117950574818,
+      "grad_norm": 1.0326915979385376,
+      "learning_rate": 0.00017852678137377092,
+      "loss": 3.2328,
+      "step": 2204
+    },
+    {
+      "epoch": 0.19960170181949852,
+      "grad_norm": 1.028706431388855,
+      "learning_rate": 0.00017850712957458774,
+      "loss": 3.1817,
+      "step": 2205
+    },
+    {
+      "epoch": 0.19969222413324886,
+      "grad_norm": 0.9901148080825806,
+      "learning_rate": 0.00017848746986973883,
+      "loss": 3.2454,
+      "step": 2206
+    },
+    {
+      "epoch": 0.1997827464469992,
+      "grad_norm": 0.8898138403892517,
+      "learning_rate": 0.0001784678022612038,
+      "loss": 3.1329,
+      "step": 2207
+    },
+    {
+      "epoch": 0.1998732687607495,
+      "grad_norm": 0.9994850158691406,
+      "learning_rate": 0.00017844812675096324,
+      "loss": 3.2728,
+      "step": 2208
+    },
+    {
+      "epoch": 0.19996379107449985,
+      "grad_norm": 0.9238694906234741,
+      "learning_rate": 0.00017842844334099847,
+      "loss": 3.2762,
+      "step": 2209
+    },
+    {
+      "epoch": 0.2000543133882502,
+      "grad_norm": 1.0947091579437256,
+      "learning_rate": 0.0001784087520332916,
+      "loss": 3.254,
+      "step": 2210
+    },
+    {
+      "epoch": 0.20014483570200053,
+      "grad_norm": 0.8965323567390442,
+      "learning_rate": 0.00017838905282982552,
+      "loss": 3.2118,
+      "step": 2211
+    },
+    {
+      "epoch": 0.20023535801575088,
+      "grad_norm": 0.9534854292869568,
+      "learning_rate": 0.000178369345732584,
+      "loss": 3.2671,
+      "step": 2212
+    },
+    {
+      "epoch": 0.20032588032950122,
+      "grad_norm": 0.9523543119430542,
+      "learning_rate": 0.0001783496307435515,
+      "loss": 3.1687,
+      "step": 2213
+    },
+    {
+      "epoch": 0.20041640264325156,
+      "grad_norm": 0.9066437482833862,
+      "learning_rate": 0.0001783299078647133,
+      "loss": 3.2303,
+      "step": 2214
+    },
+    {
+      "epoch": 0.2005069249570019,
+      "grad_norm": 1.025066614151001,
+      "learning_rate": 0.00017831017709805556,
+      "loss": 3.2281,
+      "step": 2215
+    },
+    {
+      "epoch": 0.20059744727075224,
+      "grad_norm": 0.8147643208503723,
+      "learning_rate": 0.0001782904384455651,
+      "loss": 3.2593,
+      "step": 2216
+    },
+    {
+      "epoch": 0.20068796958450258,
+      "grad_norm": 1.0472511053085327,
+      "learning_rate": 0.00017827069190922966,
+      "loss": 3.1703,
+      "step": 2217
+    },
+    {
+      "epoch": 0.20077849189825292,
+      "grad_norm": 0.9242010712623596,
+      "learning_rate": 0.00017825093749103765,
+      "loss": 3.1703,
+      "step": 2218
+    },
+    {
+      "epoch": 0.20086901421200326,
+      "grad_norm": 0.8766939640045166,
+      "learning_rate": 0.0001782311751929784,
+      "loss": 3.1491,
+      "step": 2219
+    },
+    {
+      "epoch": 0.2009595365257536,
+      "grad_norm": 0.9397275447845459,
+      "learning_rate": 0.00017821140501704194,
+      "loss": 3.2021,
+      "step": 2220
+    },
+    {
+      "epoch": 0.20105005883950394,
+      "grad_norm": 0.8902556300163269,
+      "learning_rate": 0.00017819162696521915,
+      "loss": 3.2027,
+      "step": 2221
+    },
+    {
+      "epoch": 0.20114058115325428,
+      "grad_norm": 0.9259093999862671,
+      "learning_rate": 0.00017817184103950163,
+      "loss": 3.2105,
+      "step": 2222
+    },
+    {
+      "epoch": 0.20123110346700462,
+      "grad_norm": 1.0207278728485107,
+      "learning_rate": 0.00017815204724188187,
+      "loss": 3.2432,
+      "step": 2223
+    },
+    {
+      "epoch": 0.20132162578075496,
+      "grad_norm": 0.9795804619789124,
+      "learning_rate": 0.00017813224557435312,
+      "loss": 3.2357,
+      "step": 2224
+    },
+    {
+      "epoch": 0.2014121480945053,
+      "grad_norm": 0.9147193431854248,
+      "learning_rate": 0.00017811243603890934,
+      "loss": 3.2376,
+      "step": 2225
+    },
+    {
+      "epoch": 0.20150267040825565,
+      "grad_norm": 0.8962690234184265,
+      "learning_rate": 0.00017809261863754544,
+      "loss": 3.1848,
+      "step": 2226
+    },
+    {
+      "epoch": 0.201593192722006,
+      "grad_norm": 0.907362163066864,
+      "learning_rate": 0.00017807279337225693,
+      "loss": 3.1929,
+      "step": 2227
+    },
+    {
+      "epoch": 0.2016837150357563,
+      "grad_norm": 0.8899882435798645,
+      "learning_rate": 0.00017805296024504024,
+      "loss": 3.317,
+      "step": 2228
+    },
+    {
+      "epoch": 0.20177423734950664,
+      "grad_norm": 0.934199869632721,
+      "learning_rate": 0.00017803311925789265,
+      "loss": 3.2336,
+      "step": 2229
+    },
+    {
+      "epoch": 0.20186475966325698,
+      "grad_norm": 0.8858028650283813,
+      "learning_rate": 0.00017801327041281207,
+      "loss": 3.1753,
+      "step": 2230
+    },
+    {
+      "epoch": 0.20195528197700732,
+      "grad_norm": 0.9815089702606201,
+      "learning_rate": 0.00017799341371179728,
+      "loss": 3.1928,
+      "step": 2231
+    },
+    {
+      "epoch": 0.20204580429075766,
+      "grad_norm": 1.0124167203903198,
+      "learning_rate": 0.0001779735491568479,
+      "loss": 3.2052,
+      "step": 2232
+    },
+    {
+      "epoch": 0.202136326604508,
+      "grad_norm": 0.9589294791221619,
+      "learning_rate": 0.0001779536767499642,
+      "loss": 3.2347,
+      "step": 2233
+    },
+    {
+      "epoch": 0.20222684891825835,
+      "grad_norm": 1.084632158279419,
+      "learning_rate": 0.00017793379649314744,
+      "loss": 3.269,
+      "step": 2234
+    },
+    {
+      "epoch": 0.20231737123200869,
+      "grad_norm": 1.1935856342315674,
+      "learning_rate": 0.00017791390838839944,
+      "loss": 3.1992,
+      "step": 2235
+    },
+    {
+      "epoch": 0.20240789354575903,
+      "grad_norm": 1.022628664970398,
+      "learning_rate": 0.00017789401243772305,
+      "loss": 3.2459,
+      "step": 2236
+    },
+    {
+      "epoch": 0.20249841585950937,
+      "grad_norm": 1.017520785331726,
+      "learning_rate": 0.00017787410864312172,
+      "loss": 3.2321,
+      "step": 2237
+    },
+    {
+      "epoch": 0.2025889381732597,
+      "grad_norm": 0.9736983776092529,
+      "learning_rate": 0.0001778541970065998,
+      "loss": 3.2685,
+      "step": 2238
+    },
+    {
+      "epoch": 0.20267946048701005,
+      "grad_norm": 0.9741368293762207,
+      "learning_rate": 0.00017783427753016232,
+      "loss": 3.2179,
+      "step": 2239
+    },
+    {
+      "epoch": 0.2027699828007604,
+      "grad_norm": 1.198488712310791,
+      "learning_rate": 0.00017781435021581527,
+      "loss": 3.2421,
+      "step": 2240
+    },
+    {
+      "epoch": 0.20286050511451073,
+      "grad_norm": 0.9209917783737183,
+      "learning_rate": 0.00017779441506556528,
+      "loss": 3.1841,
+      "step": 2241
+    },
+    {
+      "epoch": 0.20295102742826107,
+      "grad_norm": 0.995261549949646,
+      "learning_rate": 0.0001777744720814198,
+      "loss": 3.2257,
+      "step": 2242
+    },
+    {
+      "epoch": 0.2030415497420114,
+      "grad_norm": 1.0581942796707153,
+      "learning_rate": 0.0001777545212653871,
+      "loss": 3.131,
+      "step": 2243
+    },
+    {
+      "epoch": 0.20313207205576175,
+      "grad_norm": 0.9594766497612,
+      "learning_rate": 0.00017773456261947627,
+      "loss": 3.2499,
+      "step": 2244
+    },
+    {
+      "epoch": 0.2032225943695121,
+      "grad_norm": 1.0768234729766846,
+      "learning_rate": 0.0001777145961456971,
+      "loss": 3.2079,
+      "step": 2245
+    },
+    {
+      "epoch": 0.20331311668326243,
+      "grad_norm": 0.952381432056427,
+      "learning_rate": 0.00017769462184606024,
+      "loss": 3.1935,
+      "step": 2246
+    },
+    {
+      "epoch": 0.20340363899701278,
+      "grad_norm": 0.895455002784729,
+      "learning_rate": 0.00017767463972257706,
+      "loss": 3.2608,
+      "step": 2247
+    },
+    {
+      "epoch": 0.20349416131076312,
+      "grad_norm": 0.9744084477424622,
+      "learning_rate": 0.00017765464977725977,
+      "loss": 3.1806,
+      "step": 2248
+    },
+    {
+      "epoch": 0.20358468362451343,
+      "grad_norm": 0.9829551577568054,
+      "learning_rate": 0.0001776346520121214,
+      "loss": 3.2598,
+      "step": 2249
+    },
+    {
+      "epoch": 0.20367520593826377,
+      "grad_norm": 1.0411384105682373,
+      "learning_rate": 0.0001776146464291757,
+      "loss": 3.1889,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2037657282520141,
+      "grad_norm": 1.1049972772598267,
+      "learning_rate": 0.00017759463303043723,
+      "loss": 3.1938,
+      "step": 2251
+    },
+    {
+      "epoch": 0.20385625056576445,
+      "grad_norm": 0.9213403463363647,
+      "learning_rate": 0.0001775746118179213,
+      "loss": 3.1873,
+      "step": 2252
+    },
+    {
+      "epoch": 0.2039467728795148,
+      "grad_norm": 0.97609943151474,
+      "learning_rate": 0.00017755458279364411,
+      "loss": 3.1772,
+      "step": 2253
+    },
+    {
+      "epoch": 0.20403729519326513,
+      "grad_norm": 1.166957139968872,
+      "learning_rate": 0.00017753454595962256,
+      "loss": 3.2645,
+      "step": 2254
+    },
+    {
+      "epoch": 0.20412781750701547,
+      "grad_norm": 0.9787424802780151,
+      "learning_rate": 0.00017751450131787435,
+      "loss": 3.1883,
+      "step": 2255
+    },
+    {
+      "epoch": 0.20421833982076582,
+      "grad_norm": 0.8625527620315552,
+      "learning_rate": 0.00017749444887041799,
+      "loss": 3.1512,
+      "step": 2256
+    },
+    {
+      "epoch": 0.20430886213451616,
+      "grad_norm": 0.8910434246063232,
+      "learning_rate": 0.00017747438861927272,
+      "loss": 3.135,
+      "step": 2257
+    },
+    {
+      "epoch": 0.2043993844482665,
+      "grad_norm": 0.8338545560836792,
+      "learning_rate": 0.00017745432056645864,
+      "loss": 3.2219,
+      "step": 2258
+    },
+    {
+      "epoch": 0.20448990676201684,
+      "grad_norm": 1.1126137971878052,
+      "learning_rate": 0.00017743424471399662,
+      "loss": 3.2227,
+      "step": 2259
+    },
+    {
+      "epoch": 0.20458042907576718,
+      "grad_norm": 1.1436206102371216,
+      "learning_rate": 0.00017741416106390826,
+      "loss": 3.2326,
+      "step": 2260
+    },
+    {
+      "epoch": 0.20467095138951752,
+      "grad_norm": 1.0938067436218262,
+      "learning_rate": 0.00017739406961821602,
+      "loss": 3.1486,
+      "step": 2261
+    },
+    {
+      "epoch": 0.20476147370326786,
+      "grad_norm": 1.6595628261566162,
+      "learning_rate": 0.00017737397037894304,
+      "loss": 3.2298,
+      "step": 2262
+    },
+    {
+      "epoch": 0.2048519960170182,
+      "grad_norm": 0.9511428475379944,
+      "learning_rate": 0.00017735386334811343,
+      "loss": 3.2617,
+      "step": 2263
+    },
+    {
+      "epoch": 0.20494251833076854,
+      "grad_norm": 1.0140020847320557,
+      "learning_rate": 0.00017733374852775183,
+      "loss": 3.2335,
+      "step": 2264
+    },
+    {
+      "epoch": 0.20503304064451888,
+      "grad_norm": 0.9424805045127869,
+      "learning_rate": 0.0001773136259198839,
+      "loss": 3.1668,
+      "step": 2265
+    },
+    {
+      "epoch": 0.20512356295826922,
+      "grad_norm": 1.0711252689361572,
+      "learning_rate": 0.00017729349552653595,
+      "loss": 3.1674,
+      "step": 2266
+    },
+    {
+      "epoch": 0.20521408527201956,
+      "grad_norm": 0.9161908626556396,
+      "learning_rate": 0.00017727335734973512,
+      "loss": 3.1823,
+      "step": 2267
+    },
+    {
+      "epoch": 0.2053046075857699,
+      "grad_norm": 0.8851282000541687,
+      "learning_rate": 0.0001772532113915093,
+      "loss": 3.1713,
+      "step": 2268
+    },
+    {
+      "epoch": 0.20539512989952022,
+      "grad_norm": 0.9802316427230835,
+      "learning_rate": 0.00017723305765388725,
+      "loss": 3.1771,
+      "step": 2269
+    },
+    {
+      "epoch": 0.20548565221327056,
+      "grad_norm": 0.8938327431678772,
+      "learning_rate": 0.00017721289613889835,
+      "loss": 3.1648,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2055761745270209,
+      "grad_norm": 1.0969438552856445,
+      "learning_rate": 0.00017719272684857295,
+      "loss": 3.2681,
+      "step": 2271
+    },
+    {
+      "epoch": 0.20566669684077124,
+      "grad_norm": 0.9800640940666199,
+      "learning_rate": 0.0001771725497849421,
+      "loss": 3.2137,
+      "step": 2272
+    },
+    {
+      "epoch": 0.20575721915452158,
+      "grad_norm": 0.9212649464607239,
+      "learning_rate": 0.00017715236495003756,
+      "loss": 3.1973,
+      "step": 2273
+    },
+    {
+      "epoch": 0.20584774146827192,
+      "grad_norm": 0.9442010521888733,
+      "learning_rate": 0.000177132172345892,
+      "loss": 3.2673,
+      "step": 2274
+    },
+    {
+      "epoch": 0.20593826378202226,
+      "grad_norm": 0.9160739779472351,
+      "learning_rate": 0.00017711197197453878,
+      "loss": 3.2638,
+      "step": 2275
+    },
+    {
+      "epoch": 0.2060287860957726,
+      "grad_norm": 0.9333342909812927,
+      "learning_rate": 0.0001770917638380121,
+      "loss": 3.2041,
+      "step": 2276
+    },
+    {
+      "epoch": 0.20611930840952294,
+      "grad_norm": 0.9469937086105347,
+      "learning_rate": 0.0001770715479383469,
+      "loss": 3.1967,
+      "step": 2277
+    },
+    {
+      "epoch": 0.20620983072327329,
+      "grad_norm": 1.072282075881958,
+      "learning_rate": 0.00017705132427757895,
+      "loss": 3.2015,
+      "step": 2278
+    },
+    {
+      "epoch": 0.20630035303702363,
+      "grad_norm": 1.0128763914108276,
+      "learning_rate": 0.00017703109285774473,
+      "loss": 3.1652,
+      "step": 2279
+    },
+    {
+      "epoch": 0.20639087535077397,
+      "grad_norm": 0.9716585278511047,
+      "learning_rate": 0.00017701085368088156,
+      "loss": 3.2273,
+      "step": 2280
+    },
+    {
+      "epoch": 0.2064813976645243,
+      "grad_norm": 0.9625109434127808,
+      "learning_rate": 0.00017699060674902756,
+      "loss": 3.241,
+      "step": 2281
+    },
+    {
+      "epoch": 0.20657191997827465,
+      "grad_norm": 0.9801328778266907,
+      "learning_rate": 0.00017697035206422156,
+      "loss": 3.1929,
+      "step": 2282
+    },
+    {
+      "epoch": 0.206662442292025,
+      "grad_norm": 1.0571147203445435,
+      "learning_rate": 0.0001769500896285032,
+      "loss": 3.2106,
+      "step": 2283
+    },
+    {
+      "epoch": 0.20675296460577533,
+      "grad_norm": 0.8661196231842041,
+      "learning_rate": 0.00017692981944391294,
+      "loss": 3.1964,
+      "step": 2284
+    },
+    {
+      "epoch": 0.20684348691952567,
+      "grad_norm": 1.0298678874969482,
+      "learning_rate": 0.00017690954151249196,
+      "loss": 3.1557,
+      "step": 2285
+    },
+    {
+      "epoch": 0.206934009233276,
+      "grad_norm": 0.9180775880813599,
+      "learning_rate": 0.00017688925583628223,
+      "loss": 3.1943,
+      "step": 2286
+    },
+    {
+      "epoch": 0.20702453154702635,
+      "grad_norm": 1.0359828472137451,
+      "learning_rate": 0.00017686896241732658,
+      "loss": 3.1889,
+      "step": 2287
+    },
+    {
+      "epoch": 0.2071150538607767,
+      "grad_norm": 1.0745766162872314,
+      "learning_rate": 0.00017684866125766853,
+      "loss": 3.2023,
+      "step": 2288
+    },
+    {
+      "epoch": 0.20720557617452703,
+      "grad_norm": 1.0356098413467407,
+      "learning_rate": 0.00017682835235935236,
+      "loss": 3.1698,
+      "step": 2289
+    },
+    {
+      "epoch": 0.20729609848827735,
+      "grad_norm": 0.9535381197929382,
+      "learning_rate": 0.00017680803572442318,
+      "loss": 3.1851,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2073866208020277,
+      "grad_norm": 0.9482605457305908,
+      "learning_rate": 0.00017678771135492695,
+      "loss": 3.2179,
+      "step": 2291
+    },
+    {
+      "epoch": 0.20747714311577803,
+      "grad_norm": 1.3931949138641357,
+      "learning_rate": 0.00017676737925291028,
+      "loss": 3.2253,
+      "step": 2292
+    },
+    {
+      "epoch": 0.20756766542952837,
+      "grad_norm": 0.9733486175537109,
+      "learning_rate": 0.00017674703942042062,
+      "loss": 3.1814,
+      "step": 2293
+    },
+    {
+      "epoch": 0.2076581877432787,
+      "grad_norm": 1.0330384969711304,
+      "learning_rate": 0.0001767266918595062,
+      "loss": 3.2122,
+      "step": 2294
+    },
+    {
+      "epoch": 0.20774871005702905,
+      "grad_norm": 1.0049928426742554,
+      "learning_rate": 0.000176706336572216,
+      "loss": 3.1779,
+      "step": 2295
+    },
+    {
+      "epoch": 0.2078392323707794,
+      "grad_norm": 0.9967108368873596,
+      "learning_rate": 0.00017668597356059978,
+      "loss": 3.1691,
+      "step": 2296
+    },
+    {
+      "epoch": 0.20792975468452973,
+      "grad_norm": 1.0939202308654785,
+      "learning_rate": 0.00017666560282670815,
+      "loss": 3.1512,
+      "step": 2297
+    },
+    {
+      "epoch": 0.20802027699828007,
+      "grad_norm": 0.9251821637153625,
+      "learning_rate": 0.0001766452243725924,
+      "loss": 3.2136,
+      "step": 2298
+    },
+    {
+      "epoch": 0.20811079931203041,
+      "grad_norm": 1.0332478284835815,
+      "learning_rate": 0.00017662483820030466,
+      "loss": 3.185,
+      "step": 2299
+    },
+    {
+      "epoch": 0.20820132162578076,
+      "grad_norm": 0.8579204082489014,
+      "learning_rate": 0.0001766044443118978,
+      "loss": 3.1962,
+      "step": 2300
+    },
+    {
+      "epoch": 0.2082918439395311,
+      "grad_norm": 0.9324372410774231,
+      "learning_rate": 0.00017658404270942552,
+      "loss": 3.1525,
+      "step": 2301
+    },
+    {
+      "epoch": 0.20838236625328144,
+      "grad_norm": 0.8616467118263245,
+      "learning_rate": 0.0001765636333949422,
+      "loss": 3.1899,
+      "step": 2302
+    },
+    {
+      "epoch": 0.20847288856703178,
+      "grad_norm": 0.8347539901733398,
+      "learning_rate": 0.00017654321637050312,
+      "loss": 3.2244,
+      "step": 2303
+    },
+    {
+      "epoch": 0.20856341088078212,
+      "grad_norm": 0.8825692534446716,
+      "learning_rate": 0.00017652279163816423,
+      "loss": 3.1876,
+      "step": 2304
+    },
+    {
+      "epoch": 0.20865393319453246,
+      "grad_norm": 0.9334546327590942,
+      "learning_rate": 0.00017650235919998232,
+      "loss": 3.1664,
+      "step": 2305
+    },
+    {
+      "epoch": 0.2087444555082828,
+      "grad_norm": 0.8983275890350342,
+      "learning_rate": 0.00017648191905801492,
+      "loss": 3.1204,
+      "step": 2306
+    },
+    {
+      "epoch": 0.20883497782203314,
+      "grad_norm": 0.9988447427749634,
+      "learning_rate": 0.0001764614712143204,
+      "loss": 3.2325,
+      "step": 2307
+    },
+    {
+      "epoch": 0.20892550013578348,
+      "grad_norm": 0.883084774017334,
+      "learning_rate": 0.00017644101567095778,
+      "loss": 3.1446,
+      "step": 2308
+    },
+    {
+      "epoch": 0.20901602244953382,
+      "grad_norm": 1.0438385009765625,
+      "learning_rate": 0.000176420552429987,
+      "loss": 3.1989,
+      "step": 2309
+    },
+    {
+      "epoch": 0.20910654476328416,
+      "grad_norm": 0.9093191623687744,
+      "learning_rate": 0.00017640008149346866,
+      "loss": 3.1986,
+      "step": 2310
+    },
+    {
+      "epoch": 0.20919706707703448,
+      "grad_norm": 0.9882534146308899,
+      "learning_rate": 0.00017637960286346425,
+      "loss": 3.1743,
+      "step": 2311
+    },
+    {
+      "epoch": 0.20928758939078482,
+      "grad_norm": 1.0348436832427979,
+      "learning_rate": 0.00017635911654203588,
+      "loss": 3.195,
+      "step": 2312
+    },
+    {
+      "epoch": 0.20937811170453516,
+      "grad_norm": 0.8533861637115479,
+      "learning_rate": 0.00017633862253124655,
+      "loss": 3.1846,
+      "step": 2313
+    },
+    {
+      "epoch": 0.2094686340182855,
+      "grad_norm": 0.973780632019043,
+      "learning_rate": 0.00017631812083316003,
+      "loss": 3.2228,
+      "step": 2314
+    },
+    {
+      "epoch": 0.20955915633203584,
+      "grad_norm": 0.9459713697433472,
+      "learning_rate": 0.00017629761144984088,
+      "loss": 3.1582,
+      "step": 2315
+    },
+    {
+      "epoch": 0.20964967864578618,
+      "grad_norm": 0.9357503652572632,
+      "learning_rate": 0.00017627709438335426,
+      "loss": 3.1782,
+      "step": 2316
+    },
+    {
+      "epoch": 0.20974020095953652,
+      "grad_norm": 0.9655158519744873,
+      "learning_rate": 0.0001762565696357664,
+      "loss": 3.1712,
+      "step": 2317
+    },
+    {
+      "epoch": 0.20983072327328686,
+      "grad_norm": 0.8718851208686829,
+      "learning_rate": 0.00017623603720914402,
+      "loss": 3.1737,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2099212455870372,
+      "grad_norm": 0.9923802614212036,
+      "learning_rate": 0.0001762154971055548,
+      "loss": 3.1181,
+      "step": 2319
+    },
+    {
+      "epoch": 0.21001176790078754,
+      "grad_norm": 0.8967884182929993,
+      "learning_rate": 0.0001761949493270671,
+      "loss": 3.2443,
+      "step": 2320
+    },
+    {
+      "epoch": 0.21010229021453788,
+      "grad_norm": 0.9012006521224976,
+      "learning_rate": 0.00017617439387575007,
+      "loss": 3.201,
+      "step": 2321
+    },
+    {
+      "epoch": 0.21019281252828823,
+      "grad_norm": 0.8997471332550049,
+      "learning_rate": 0.0001761538307536737,
+      "loss": 3.1972,
+      "step": 2322
+    },
+    {
+      "epoch": 0.21028333484203857,
+      "grad_norm": 0.9365257024765015,
+      "learning_rate": 0.00017613325996290862,
+      "loss": 3.265,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2103738571557889,
+      "grad_norm": 1.175702691078186,
+      "learning_rate": 0.00017611268150552635,
+      "loss": 3.2106,
+      "step": 2324
+    },
+    {
+      "epoch": 0.21046437946953925,
+      "grad_norm": 0.9600235819816589,
+      "learning_rate": 0.00017609209538359917,
+      "loss": 3.1665,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2105549017832896,
+      "grad_norm": 1.3065840005874634,
+      "learning_rate": 0.0001760715015992,
+      "loss": 3.1699,
+      "step": 2326
+    },
+    {
+      "epoch": 0.21064542409703993,
+      "grad_norm": 0.8768181204795837,
+      "learning_rate": 0.00017605090015440275,
+      "loss": 3.2304,
+      "step": 2327
+    },
+    {
+      "epoch": 0.21073594641079027,
+      "grad_norm": 1.052873134613037,
+      "learning_rate": 0.00017603029105128193,
+      "loss": 3.1579,
+      "step": 2328
+    },
+    {
+      "epoch": 0.2108264687245406,
+      "grad_norm": 1.34178626537323,
+      "learning_rate": 0.0001760096742919129,
+      "loss": 3.23,
+      "step": 2329
+    },
+    {
+      "epoch": 0.21091699103829095,
+      "grad_norm": 1.0743076801300049,
+      "learning_rate": 0.0001759890498783717,
+      "loss": 3.2031,
+      "step": 2330
+    },
+    {
+      "epoch": 0.21100751335204126,
+      "grad_norm": 1.1340997219085693,
+      "learning_rate": 0.0001759684178127353,
+      "loss": 3.2033,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2110980356657916,
+      "grad_norm": 0.9930472373962402,
+      "learning_rate": 0.00017594777809708126,
+      "loss": 3.0984,
+      "step": 2332
+    },
+    {
+      "epoch": 0.21118855797954195,
+      "grad_norm": 1.214159607887268,
+      "learning_rate": 0.00017592713073348807,
+      "loss": 3.2372,
+      "step": 2333
+    },
+    {
+      "epoch": 0.2112790802932923,
+      "grad_norm": 0.992418646812439,
+      "learning_rate": 0.0001759064757240349,
+      "loss": 3.2189,
+      "step": 2334
+    },
+    {
+      "epoch": 0.21136960260704263,
+      "grad_norm": 1.2041786909103394,
+      "learning_rate": 0.00017588581307080169,
+      "loss": 3.1819,
+      "step": 2335
+    },
+    {
+      "epoch": 0.21146012492079297,
+      "grad_norm": 1.1528639793395996,
+      "learning_rate": 0.0001758651427758692,
+      "loss": 3.2333,
+      "step": 2336
+    },
+    {
+      "epoch": 0.2115506472345433,
+      "grad_norm": 1.0668143033981323,
+      "learning_rate": 0.0001758444648413189,
+      "loss": 3.1998,
+      "step": 2337
+    },
+    {
+      "epoch": 0.21164116954829365,
+      "grad_norm": 1.164767861366272,
+      "learning_rate": 0.00017582377926923305,
+      "loss": 3.1914,
+      "step": 2338
+    },
+    {
+      "epoch": 0.211731691862044,
+      "grad_norm": 1.0460022687911987,
+      "learning_rate": 0.00017580308606169473,
+      "loss": 3.1727,
+      "step": 2339
+    },
+    {
+      "epoch": 0.21182221417579433,
+      "grad_norm": 1.0179146528244019,
+      "learning_rate": 0.0001757823852207877,
+      "loss": 3.1831,
+      "step": 2340
+    },
+    {
+      "epoch": 0.21191273648954467,
+      "grad_norm": 1.1723949909210205,
+      "learning_rate": 0.00017576167674859656,
+      "loss": 3.1325,
+      "step": 2341
+    },
+    {
+      "epoch": 0.212003258803295,
+      "grad_norm": 1.135374903678894,
+      "learning_rate": 0.00017574096064720663,
+      "loss": 3.1349,
+      "step": 2342
+    },
+    {
+      "epoch": 0.21209378111704535,
+      "grad_norm": 0.9510069489479065,
+      "learning_rate": 0.00017572023691870403,
+      "loss": 3.093,
+      "step": 2343
+    },
+    {
+      "epoch": 0.2121843034307957,
+      "grad_norm": 1.171178936958313,
+      "learning_rate": 0.00017569950556517566,
+      "loss": 3.1266,
+      "step": 2344
+    },
+    {
+      "epoch": 0.21227482574454604,
+      "grad_norm": 0.9744029641151428,
+      "learning_rate": 0.00017567876658870914,
+      "loss": 3.169,
+      "step": 2345
+    },
+    {
+      "epoch": 0.21236534805829638,
+      "grad_norm": 0.8958362936973572,
+      "learning_rate": 0.0001756580199913929,
+      "loss": 3.1228,
+      "step": 2346
+    },
+    {
+      "epoch": 0.21245587037204672,
+      "grad_norm": 0.9392367601394653,
+      "learning_rate": 0.00017563726577531612,
+      "loss": 3.126,
+      "step": 2347
+    },
+    {
+      "epoch": 0.21254639268579706,
+      "grad_norm": 0.9655486941337585,
+      "learning_rate": 0.00017561650394256874,
+      "loss": 3.1772,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2126369149995474,
+      "grad_norm": 0.8699467778205872,
+      "learning_rate": 0.0001755957344952415,
+      "loss": 3.1088,
+      "step": 2349
+    },
+    {
+      "epoch": 0.21272743731329774,
+      "grad_norm": 0.9144679307937622,
+      "learning_rate": 0.00017557495743542585,
+      "loss": 3.1993,
+      "step": 2350
+    },
+    {
+      "epoch": 0.21281795962704808,
+      "grad_norm": 0.919262707233429,
+      "learning_rate": 0.00017555417276521405,
+      "loss": 3.1578,
+      "step": 2351
+    },
+    {
+      "epoch": 0.2129084819407984,
+      "grad_norm": 1.0309118032455444,
+      "learning_rate": 0.00017553338048669913,
+      "loss": 3.167,
+      "step": 2352
+    },
+    {
+      "epoch": 0.21299900425454873,
+      "grad_norm": 0.9247647523880005,
+      "learning_rate": 0.00017551258060197485,
+      "loss": 3.1737,
+      "step": 2353
+    },
+    {
+      "epoch": 0.21308952656829908,
+      "grad_norm": 0.9028778076171875,
+      "learning_rate": 0.00017549177311313577,
+      "loss": 3.1385,
+      "step": 2354
+    },
+    {
+      "epoch": 0.21318004888204942,
+      "grad_norm": 0.9501022100448608,
+      "learning_rate": 0.00017547095802227723,
+      "loss": 3.1866,
+      "step": 2355
+    },
+    {
+      "epoch": 0.21327057119579976,
+      "grad_norm": 0.8799710273742676,
+      "learning_rate": 0.00017545013533149523,
+      "loss": 3.1837,
+      "step": 2356
+    },
+    {
+      "epoch": 0.2133610935095501,
+      "grad_norm": 0.9339806437492371,
+      "learning_rate": 0.0001754293050428867,
+      "loss": 3.2179,
+      "step": 2357
+    },
+    {
+      "epoch": 0.21345161582330044,
+      "grad_norm": 0.891981840133667,
+      "learning_rate": 0.00017540846715854923,
+      "loss": 3.2324,
+      "step": 2358
+    },
+    {
+      "epoch": 0.21354213813705078,
+      "grad_norm": 0.8680137395858765,
+      "learning_rate": 0.00017538762168058117,
+      "loss": 3.1399,
+      "step": 2359
+    },
+    {
+      "epoch": 0.21363266045080112,
+      "grad_norm": 0.9303855299949646,
+      "learning_rate": 0.00017536676861108164,
+      "loss": 3.1699,
+      "step": 2360
+    },
+    {
+      "epoch": 0.21372318276455146,
+      "grad_norm": 0.9293773770332336,
+      "learning_rate": 0.0001753459079521506,
+      "loss": 3.203,
+      "step": 2361
+    },
+    {
+      "epoch": 0.2138137050783018,
+      "grad_norm": 1.0480334758758545,
+      "learning_rate": 0.0001753250397058887,
+      "loss": 3.2587,
+      "step": 2362
+    },
+    {
+      "epoch": 0.21390422739205214,
+      "grad_norm": 0.8495811223983765,
+      "learning_rate": 0.00017530416387439733,
+      "loss": 3.1403,
+      "step": 2363
+    },
+    {
+      "epoch": 0.21399474970580248,
+      "grad_norm": 1.0046278238296509,
+      "learning_rate": 0.00017528328045977876,
+      "loss": 3.1598,
+      "step": 2364
+    },
+    {
+      "epoch": 0.21408527201955282,
+      "grad_norm": 0.9166176915168762,
+      "learning_rate": 0.00017526238946413589,
+      "loss": 3.1074,
+      "step": 2365
+    },
+    {
+      "epoch": 0.21417579433330317,
+      "grad_norm": 0.816504716873169,
+      "learning_rate": 0.00017524149088957245,
+      "loss": 3.1762,
+      "step": 2366
+    },
+    {
+      "epoch": 0.2142663166470535,
+      "grad_norm": 0.9491996765136719,
+      "learning_rate": 0.00017522058473819294,
+      "loss": 3.199,
+      "step": 2367
+    },
+    {
+      "epoch": 0.21435683896080385,
+      "grad_norm": 0.9550940990447998,
+      "learning_rate": 0.0001751996710121026,
+      "loss": 3.2461,
+      "step": 2368
+    },
+    {
+      "epoch": 0.2144473612745542,
+      "grad_norm": 0.9655570983886719,
+      "learning_rate": 0.00017517874971340747,
+      "loss": 3.1665,
+      "step": 2369
+    },
+    {
+      "epoch": 0.21453788358830453,
+      "grad_norm": 0.899738073348999,
+      "learning_rate": 0.00017515782084421427,
+      "loss": 3.1383,
+      "step": 2370
+    },
+    {
+      "epoch": 0.21462840590205487,
+      "grad_norm": 0.9755668044090271,
+      "learning_rate": 0.0001751368844066306,
+      "loss": 3.0888,
+      "step": 2371
+    },
+    {
+      "epoch": 0.21471892821580518,
+      "grad_norm": 0.9048589468002319,
+      "learning_rate": 0.00017511594040276469,
+      "loss": 3.1741,
+      "step": 2372
+    },
+    {
+      "epoch": 0.21480945052955552,
+      "grad_norm": 0.9476847648620605,
+      "learning_rate": 0.00017509498883472565,
+      "loss": 3.2199,
+      "step": 2373
+    },
+    {
+      "epoch": 0.21489997284330586,
+      "grad_norm": 0.8622844219207764,
+      "learning_rate": 0.0001750740297046233,
+      "loss": 3.1287,
+      "step": 2374
+    },
+    {
+      "epoch": 0.2149904951570562,
+      "grad_norm": 0.9325116276741028,
+      "learning_rate": 0.00017505306301456822,
+      "loss": 3.1792,
+      "step": 2375
+    },
+    {
+      "epoch": 0.21508101747080655,
+      "grad_norm": 0.9339794516563416,
+      "learning_rate": 0.00017503208876667173,
+      "loss": 3.1535,
+      "step": 2376
+    },
+    {
+      "epoch": 0.2151715397845569,
+      "grad_norm": 0.8598697185516357,
+      "learning_rate": 0.00017501110696304596,
+      "loss": 3.1139,
+      "step": 2377
+    },
+    {
+      "epoch": 0.21526206209830723,
+      "grad_norm": 0.8950153589248657,
+      "learning_rate": 0.00017499011760580376,
+      "loss": 3.1469,
+      "step": 2378
+    },
+    {
+      "epoch": 0.21535258441205757,
+      "grad_norm": 0.9081557989120483,
+      "learning_rate": 0.0001749691206970588,
+      "loss": 3.1602,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2154431067258079,
+      "grad_norm": 0.8921812176704407,
+      "learning_rate": 0.0001749481162389254,
+      "loss": 3.187,
+      "step": 2380
+    },
+    {
+      "epoch": 0.21553362903955825,
+      "grad_norm": 0.8708139061927795,
+      "learning_rate": 0.00017492710423351878,
+      "loss": 3.1596,
+      "step": 2381
+    },
+    {
+      "epoch": 0.2156241513533086,
+      "grad_norm": 0.8599114418029785,
+      "learning_rate": 0.0001749060846829548,
+      "loss": 3.1284,
+      "step": 2382
+    },
+    {
+      "epoch": 0.21571467366705893,
+      "grad_norm": 0.9189428687095642,
+      "learning_rate": 0.00017488505758935015,
+      "loss": 3.0794,
+      "step": 2383
+    },
+    {
+      "epoch": 0.21580519598080927,
+      "grad_norm": 0.8451887965202332,
+      "learning_rate": 0.00017486402295482223,
+      "loss": 3.1743,
+      "step": 2384
+    },
+    {
+      "epoch": 0.2158957182945596,
+      "grad_norm": 0.8397995829582214,
+      "learning_rate": 0.00017484298078148926,
+      "loss": 3.1973,
+      "step": 2385
+    },
+    {
+      "epoch": 0.21598624060830995,
+      "grad_norm": 0.889971911907196,
+      "learning_rate": 0.00017482193107147014,
+      "loss": 3.1543,
+      "step": 2386
+    },
+    {
+      "epoch": 0.2160767629220603,
+      "grad_norm": 0.8662976622581482,
+      "learning_rate": 0.00017480087382688465,
+      "loss": 3.1575,
+      "step": 2387
+    },
+    {
+      "epoch": 0.21616728523581064,
+      "grad_norm": 0.8261364698410034,
+      "learning_rate": 0.0001747798090498532,
+      "loss": 3.1776,
+      "step": 2388
+    },
+    {
+      "epoch": 0.21625780754956098,
+      "grad_norm": 0.8544386029243469,
+      "learning_rate": 0.00017475873674249704,
+      "loss": 3.1272,
+      "step": 2389
+    },
+    {
+      "epoch": 0.21634832986331132,
+      "grad_norm": 0.8943440914154053,
+      "learning_rate": 0.0001747376569069381,
+      "loss": 3.1176,
+      "step": 2390
+    },
+    {
+      "epoch": 0.21643885217706166,
+      "grad_norm": 0.8617627620697021,
+      "learning_rate": 0.0001747165695452992,
+      "loss": 3.0768,
+      "step": 2391
+    },
+    {
+      "epoch": 0.216529374490812,
+      "grad_norm": 1.0215213298797607,
+      "learning_rate": 0.00017469547465970373,
+      "loss": 3.1451,
+      "step": 2392
+    },
+    {
+      "epoch": 0.2166198968045623,
+      "grad_norm": 0.8882916569709778,
+      "learning_rate": 0.00017467437225227606,
+      "loss": 3.1673,
+      "step": 2393
+    },
+    {
+      "epoch": 0.21671041911831265,
+      "grad_norm": 0.8575624227523804,
+      "learning_rate": 0.0001746532623251411,
+      "loss": 3.1365,
+      "step": 2394
+    },
+    {
+      "epoch": 0.216800941432063,
+      "grad_norm": 0.9205721616744995,
+      "learning_rate": 0.00017463214488042472,
+      "loss": 3.1569,
+      "step": 2395
+    },
+    {
+      "epoch": 0.21689146374581333,
+      "grad_norm": 0.9491685628890991,
+      "learning_rate": 0.00017461101992025335,
+      "loss": 3.1019,
+      "step": 2396
+    },
+    {
+      "epoch": 0.21698198605956368,
+      "grad_norm": 0.8524163365364075,
+      "learning_rate": 0.00017458988744675433,
+      "loss": 3.1516,
+      "step": 2397
+    },
+    {
+      "epoch": 0.21707250837331402,
+      "grad_norm": 0.9187124967575073,
+      "learning_rate": 0.00017456874746205568,
+      "loss": 3.186,
+      "step": 2398
+    },
+    {
+      "epoch": 0.21716303068706436,
+      "grad_norm": 0.9027589559555054,
+      "learning_rate": 0.00017454759996828623,
+      "loss": 3.1449,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2172535530008147,
+      "grad_norm": 0.9175606966018677,
+      "learning_rate": 0.0001745264449675755,
+      "loss": 3.1573,
+      "step": 2400
+    },
+    {
+      "epoch": 0.21734407531456504,
+      "grad_norm": 0.9693629741668701,
+      "learning_rate": 0.00017450528246205376,
+      "loss": 3.0802,
+      "step": 2401
+    },
+    {
+      "epoch": 0.21743459762831538,
+      "grad_norm": 0.867482602596283,
+      "learning_rate": 0.00017448411245385214,
+      "loss": 3.176,
+      "step": 2402
+    },
+    {
+      "epoch": 0.21752511994206572,
+      "grad_norm": 1.0069620609283447,
+      "learning_rate": 0.00017446293494510244,
+      "loss": 3.1445,
+      "step": 2403
+    },
+    {
+      "epoch": 0.21761564225581606,
+      "grad_norm": 0.9470597505569458,
+      "learning_rate": 0.0001744417499379372,
+      "loss": 3.1548,
+      "step": 2404
+    },
+    {
+      "epoch": 0.2177061645695664,
+      "grad_norm": 0.9411137700080872,
+      "learning_rate": 0.0001744205574344898,
+      "loss": 3.1723,
+      "step": 2405
+    },
+    {
+      "epoch": 0.21779668688331674,
+      "grad_norm": 0.9758121967315674,
+      "learning_rate": 0.00017439935743689431,
+      "loss": 3.2164,
+      "step": 2406
+    },
+    {
+      "epoch": 0.21788720919706708,
+      "grad_norm": 0.9710676670074463,
+      "learning_rate": 0.00017437814994728553,
+      "loss": 3.1502,
+      "step": 2407
+    },
+    {
+      "epoch": 0.21797773151081742,
+      "grad_norm": 0.9743636250495911,
+      "learning_rate": 0.00017435693496779911,
+      "loss": 3.1949,
+      "step": 2408
+    },
+    {
+      "epoch": 0.21806825382456776,
+      "grad_norm": 0.9133598208427429,
+      "learning_rate": 0.0001743357125005714,
+      "loss": 3.1189,
+      "step": 2409
+    },
+    {
+      "epoch": 0.2181587761383181,
+      "grad_norm": 0.8851455450057983,
+      "learning_rate": 0.00017431448254773944,
+      "loss": 3.1698,
+      "step": 2410
+    },
+    {
+      "epoch": 0.21824929845206845,
+      "grad_norm": 0.873092770576477,
+      "learning_rate": 0.0001742932451114411,
+      "loss": 3.156,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2183398207658188,
+      "grad_norm": 0.956868052482605,
+      "learning_rate": 0.00017427200019381504,
+      "loss": 3.202,
+      "step": 2412
+    },
+    {
+      "epoch": 0.2184303430795691,
+      "grad_norm": 0.9415931701660156,
+      "learning_rate": 0.00017425074779700063,
+      "loss": 3.1744,
+      "step": 2413
+    },
+    {
+      "epoch": 0.21852086539331944,
+      "grad_norm": 0.9146011471748352,
+      "learning_rate": 0.00017422948792313788,
+      "loss": 3.1592,
+      "step": 2414
+    },
+    {
+      "epoch": 0.21861138770706978,
+      "grad_norm": 1.0343977212905884,
+      "learning_rate": 0.00017420822057436777,
+      "loss": 3.1817,
+      "step": 2415
+    },
+    {
+      "epoch": 0.21870191002082012,
+      "grad_norm": 1.1891487836837769,
+      "learning_rate": 0.00017418694575283185,
+      "loss": 3.2632,
+      "step": 2416
+    },
+    {
+      "epoch": 0.21879243233457046,
+      "grad_norm": 0.9265338182449341,
+      "learning_rate": 0.00017416566346067255,
+      "loss": 3.1894,
+      "step": 2417
+    },
+    {
+      "epoch": 0.2188829546483208,
+      "grad_norm": 0.8732950687408447,
+      "learning_rate": 0.00017414437370003293,
+      "loss": 3.1959,
+      "step": 2418
+    },
+    {
+      "epoch": 0.21897347696207115,
+      "grad_norm": 0.9065073132514954,
+      "learning_rate": 0.00017412307647305696,
+      "loss": 3.1425,
+      "step": 2419
+    },
+    {
+      "epoch": 0.21906399927582149,
+      "grad_norm": 0.9114652872085571,
+      "learning_rate": 0.00017410177178188918,
+      "loss": 3.2211,
+      "step": 2420
+    },
+    {
+      "epoch": 0.21915452158957183,
+      "grad_norm": 1.0082124471664429,
+      "learning_rate": 0.000174080459628675,
+      "loss": 3.2108,
+      "step": 2421
+    },
+    {
+      "epoch": 0.21924504390332217,
+      "grad_norm": 1.0025688409805298,
+      "learning_rate": 0.0001740591400155606,
+      "loss": 3.144,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2193355662170725,
+      "grad_norm": 0.8575987815856934,
+      "learning_rate": 0.0001740378129446928,
+      "loss": 3.1099,
+      "step": 2423
+    },
+    {
+      "epoch": 0.21942608853082285,
+      "grad_norm": 0.9605771899223328,
+      "learning_rate": 0.00017401647841821928,
+      "loss": 3.0751,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2195166108445732,
+      "grad_norm": 1.021604061126709,
+      "learning_rate": 0.0001739951364382884,
+      "loss": 3.1323,
+      "step": 2425
+    },
+    {
+      "epoch": 0.21960713315832353,
+      "grad_norm": 0.9601428508758545,
+      "learning_rate": 0.0001739737870070493,
+      "loss": 3.3042,
+      "step": 2426
+    },
+    {
+      "epoch": 0.21969765547207387,
+      "grad_norm": 1.1104557514190674,
+      "learning_rate": 0.00017395243012665188,
+      "loss": 3.2057,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2197881777858242,
+      "grad_norm": 0.9566591382026672,
+      "learning_rate": 0.00017393106579924677,
+      "loss": 3.1467,
+      "step": 2428
+    },
+    {
+      "epoch": 0.21987870009957455,
+      "grad_norm": 0.9779771566390991,
+      "learning_rate": 0.00017390969402698535,
+      "loss": 3.1151,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2199692224133249,
+      "grad_norm": 0.9599485993385315,
+      "learning_rate": 0.00017388831481201977,
+      "loss": 3.1216,
+      "step": 2430
+    },
+    {
+      "epoch": 0.22005974472707523,
+      "grad_norm": 1.0270799398422241,
+      "learning_rate": 0.0001738669281565029,
+      "loss": 3.258,
+      "step": 2431
+    },
+    {
+      "epoch": 0.22015026704082558,
+      "grad_norm": 0.8845586180686951,
+      "learning_rate": 0.00017384553406258842,
+      "loss": 3.0924,
+      "step": 2432
+    },
+    {
+      "epoch": 0.22024078935457592,
+      "grad_norm": 0.826064944267273,
+      "learning_rate": 0.0001738241325324306,
+      "loss": 3.1836,
+      "step": 2433
+    },
+    {
+      "epoch": 0.22033131166832623,
+      "grad_norm": 0.9534816145896912,
+      "learning_rate": 0.00017380272356818473,
+      "loss": 3.2086,
+      "step": 2434
+    },
+    {
+      "epoch": 0.22042183398207657,
+      "grad_norm": 0.9598084688186646,
+      "learning_rate": 0.0001737813071720066,
+      "loss": 3.1166,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2205123562958269,
+      "grad_norm": 1.0776300430297852,
+      "learning_rate": 0.00017375988334605285,
+      "loss": 3.176,
+      "step": 2436
+    },
+    {
+      "epoch": 0.22060287860957725,
+      "grad_norm": 0.8963593244552612,
+      "learning_rate": 0.00017373845209248088,
+      "loss": 3.1777,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2206934009233276,
+      "grad_norm": 0.9389859437942505,
+      "learning_rate": 0.00017371701341344878,
+      "loss": 3.1094,
+      "step": 2438
+    },
+    {
+      "epoch": 0.22078392323707793,
+      "grad_norm": 0.9606500267982483,
+      "learning_rate": 0.00017369556731111545,
+      "loss": 3.131,
+      "step": 2439
+    },
+    {
+      "epoch": 0.22087444555082827,
+      "grad_norm": 0.8964757323265076,
+      "learning_rate": 0.0001736741137876405,
+      "loss": 3.1318,
+      "step": 2440
+    },
+    {
+      "epoch": 0.22096496786457862,
+      "grad_norm": 0.9549975395202637,
+      "learning_rate": 0.00017365265284518432,
+      "loss": 3.1401,
+      "step": 2441
+    },
+    {
+      "epoch": 0.22105549017832896,
+      "grad_norm": 0.9945970773696899,
+      "learning_rate": 0.00017363118448590801,
+      "loss": 3.1488,
+      "step": 2442
+    },
+    {
+      "epoch": 0.2211460124920793,
+      "grad_norm": 0.9644862413406372,
+      "learning_rate": 0.00017360970871197346,
+      "loss": 3.1731,
+      "step": 2443
+    },
+    {
+      "epoch": 0.22123653480582964,
+      "grad_norm": 0.8978458046913147,
+      "learning_rate": 0.00017358822552554322,
+      "loss": 3.2147,
+      "step": 2444
+    },
+    {
+      "epoch": 0.22132705711957998,
+      "grad_norm": 0.97486811876297,
+      "learning_rate": 0.0001735667349287807,
+      "loss": 3.1623,
+      "step": 2445
+    },
+    {
+      "epoch": 0.22141757943333032,
+      "grad_norm": 0.9704816937446594,
+      "learning_rate": 0.00017354523692385,
+      "loss": 3.1984,
+      "step": 2446
+    },
+    {
+      "epoch": 0.22150810174708066,
+      "grad_norm": 1.077000379562378,
+      "learning_rate": 0.00017352373151291591,
+      "loss": 3.1177,
+      "step": 2447
+    },
+    {
+      "epoch": 0.221598624060831,
+      "grad_norm": 0.8644207715988159,
+      "learning_rate": 0.00017350221869814411,
+      "loss": 3.1528,
+      "step": 2448
+    },
+    {
+      "epoch": 0.22168914637458134,
+      "grad_norm": 0.8725171089172363,
+      "learning_rate": 0.00017348069848170086,
+      "loss": 3.159,
+      "step": 2449
+    },
+    {
+      "epoch": 0.22177966868833168,
+      "grad_norm": 1.0187084674835205,
+      "learning_rate": 0.00017345917086575332,
+      "loss": 3.1971,
+      "step": 2450
+    },
+    {
+      "epoch": 0.22187019100208202,
+      "grad_norm": 0.8878188133239746,
+      "learning_rate": 0.00017343763585246928,
+      "loss": 3.1588,
+      "step": 2451
+    },
+    {
+      "epoch": 0.22196071331583236,
+      "grad_norm": 0.9226207137107849,
+      "learning_rate": 0.0001734160934440173,
+      "loss": 3.1223,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2220512356295827,
+      "grad_norm": 0.9456731677055359,
+      "learning_rate": 0.00017339454364256676,
+      "loss": 3.223,
+      "step": 2453
+    },
+    {
+      "epoch": 0.22214175794333302,
+      "grad_norm": 0.9416071176528931,
+      "learning_rate": 0.00017337298645028764,
+      "loss": 3.2436,
+      "step": 2454
+    },
+    {
+      "epoch": 0.22223228025708336,
+      "grad_norm": 0.8809584379196167,
+      "learning_rate": 0.00017335142186935081,
+      "loss": 3.1287,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2223228025708337,
+      "grad_norm": 0.905727744102478,
+      "learning_rate": 0.00017332984990192779,
+      "loss": 3.2025,
+      "step": 2456
+    },
+    {
+      "epoch": 0.22241332488458404,
+      "grad_norm": 1.0547999143600464,
+      "learning_rate": 0.00017330827055019092,
+      "loss": 3.1311,
+      "step": 2457
+    },
+    {
+      "epoch": 0.22250384719833438,
+      "grad_norm": 0.8692690134048462,
+      "learning_rate": 0.00017328668381631318,
+      "loss": 3.0641,
+      "step": 2458
+    },
+    {
+      "epoch": 0.22259436951208472,
+      "grad_norm": 1.1819257736206055,
+      "learning_rate": 0.0001732650897024684,
+      "loss": 3.1454,
+      "step": 2459
+    },
+    {
+      "epoch": 0.22268489182583506,
+      "grad_norm": 1.016374111175537,
+      "learning_rate": 0.0001732434882108311,
+      "loss": 3.194,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2227754141395854,
+      "grad_norm": 1.5239908695220947,
+      "learning_rate": 0.0001732218793435765,
+      "loss": 3.1774,
+      "step": 2461
+    },
+    {
+      "epoch": 0.22286593645333574,
+      "grad_norm": 0.9903870820999146,
+      "learning_rate": 0.00017320026310288072,
+      "loss": 3.2263,
+      "step": 2462
+    },
+    {
+      "epoch": 0.22295645876708609,
+      "grad_norm": 1.1110303401947021,
+      "learning_rate": 0.0001731786394909204,
+      "loss": 3.1876,
+      "step": 2463
+    },
+    {
+      "epoch": 0.22304698108083643,
+      "grad_norm": 1.109635829925537,
+      "learning_rate": 0.0001731570085098731,
+      "loss": 3.1677,
+      "step": 2464
+    },
+    {
+      "epoch": 0.22313750339458677,
+      "grad_norm": 0.9838115572929382,
+      "learning_rate": 0.00017313537016191706,
+      "loss": 3.127,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2232280257083371,
+      "grad_norm": 1.3351515531539917,
+      "learning_rate": 0.00017311372444923123,
+      "loss": 3.1621,
+      "step": 2466
+    },
+    {
+      "epoch": 0.22331854802208745,
+      "grad_norm": 0.900549590587616,
+      "learning_rate": 0.00017309207137399535,
+      "loss": 3.0983,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2234090703358378,
+      "grad_norm": 1.293603777885437,
+      "learning_rate": 0.00017307041093838986,
+      "loss": 3.1548,
+      "step": 2468
+    },
+    {
+      "epoch": 0.22349959264958813,
+      "grad_norm": 0.9529392123222351,
+      "learning_rate": 0.00017304874314459602,
+      "loss": 3.1299,
+      "step": 2469
+    },
+    {
+      "epoch": 0.22359011496333847,
+      "grad_norm": 1.1548309326171875,
+      "learning_rate": 0.00017302706799479574,
+      "loss": 3.1252,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2236806372770888,
+      "grad_norm": 1.0050303936004639,
+      "learning_rate": 0.0001730053854911717,
+      "loss": 3.1417,
+      "step": 2471
+    },
+    {
+      "epoch": 0.22377115959083915,
+      "grad_norm": 1.0455986261367798,
+      "learning_rate": 0.00017298369563590734,
+      "loss": 3.1961,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2238616819045895,
+      "grad_norm": 0.999272346496582,
+      "learning_rate": 0.00017296199843118683,
+      "loss": 3.1329,
+      "step": 2473
+    },
+    {
+      "epoch": 0.22395220421833983,
+      "grad_norm": 0.886278510093689,
+      "learning_rate": 0.00017294029387919508,
+      "loss": 3.1698,
+      "step": 2474
+    },
+    {
+      "epoch": 0.22404272653209015,
+      "grad_norm": 1.007181167602539,
+      "learning_rate": 0.00017291858198211773,
+      "loss": 3.1676,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2241332488458405,
+      "grad_norm": 0.8566113114356995,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 3.1895,
+      "step": 2476
+    },
+    {
+      "epoch": 0.22422377115959083,
+      "grad_norm": 0.9515550136566162,
+      "learning_rate": 0.0001728751361614525,
+      "loss": 3.1546,
+      "step": 2477
+    },
+    {
+      "epoch": 0.22431429347334117,
+      "grad_norm": 0.8973231911659241,
+      "learning_rate": 0.00017285340224223965,
+      "loss": 3.1885,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2244048157870915,
+      "grad_norm": 0.9097982048988342,
+      "learning_rate": 0.00017283166098669116,
+      "loss": 3.1111,
+      "step": 2479
+    },
+    {
+      "epoch": 0.22449533810084185,
+      "grad_norm": 0.9288604855537415,
+      "learning_rate": 0.00017280991239699642,
+      "loss": 3.1735,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2245858604145922,
+      "grad_norm": 0.8434594869613647,
+      "learning_rate": 0.00017278815647534548,
+      "loss": 3.0955,
+      "step": 2481
+    },
+    {
+      "epoch": 0.22467638272834253,
+      "grad_norm": 0.8894537687301636,
+      "learning_rate": 0.00017276639322392917,
+      "loss": 3.0957,
+      "step": 2482
+    },
+    {
+      "epoch": 0.22476690504209287,
+      "grad_norm": 0.9819609522819519,
+      "learning_rate": 0.0001727446226449391,
+      "loss": 3.1211,
+      "step": 2483
+    },
+    {
+      "epoch": 0.22485742735584321,
+      "grad_norm": 0.8595069646835327,
+      "learning_rate": 0.00017272284474056744,
+      "loss": 3.1588,
+      "step": 2484
+    },
+    {
+      "epoch": 0.22494794966959356,
+      "grad_norm": 0.9514359831809998,
+      "learning_rate": 0.00017270105951300738,
+      "loss": 3.1498,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2250384719833439,
+      "grad_norm": 0.878564178943634,
+      "learning_rate": 0.00017267926696445257,
+      "loss": 3.1062,
+      "step": 2486
+    },
+    {
+      "epoch": 0.22512899429709424,
+      "grad_norm": 0.9188651442527771,
+      "learning_rate": 0.0001726574670970976,
+      "loss": 3.1394,
+      "step": 2487
+    },
+    {
+      "epoch": 0.22521951661084458,
+      "grad_norm": 0.9147782325744629,
+      "learning_rate": 0.00017263565991313765,
+      "loss": 3.1907,
+      "step": 2488
+    },
+    {
+      "epoch": 0.22531003892459492,
+      "grad_norm": 0.8864535093307495,
+      "learning_rate": 0.00017261384541476877,
+      "loss": 3.195,
+      "step": 2489
+    },
+    {
+      "epoch": 0.22540056123834526,
+      "grad_norm": 0.8890230655670166,
+      "learning_rate": 0.00017259202360418762,
+      "loss": 3.1051,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2254910835520956,
+      "grad_norm": 0.9365044832229614,
+      "learning_rate": 0.00017257019448359168,
+      "loss": 3.1838,
+      "step": 2491
+    },
+    {
+      "epoch": 0.22558160586584594,
+      "grad_norm": 1.0051701068878174,
+      "learning_rate": 0.00017254835805517914,
+      "loss": 3.1432,
+      "step": 2492
+    },
+    {
+      "epoch": 0.22567212817959628,
+      "grad_norm": 0.9916554689407349,
+      "learning_rate": 0.00017252651432114892,
+      "loss": 3.1628,
+      "step": 2493
+    },
+    {
+      "epoch": 0.22576265049334662,
+      "grad_norm": 0.9559274911880493,
+      "learning_rate": 0.0001725046632837007,
+      "loss": 3.1104,
+      "step": 2494
+    },
+    {
+      "epoch": 0.22585317280709694,
+      "grad_norm": 1.0775694847106934,
+      "learning_rate": 0.00017248280494503487,
+      "loss": 3.2151,
+      "step": 2495
+    },
+    {
+      "epoch": 0.22594369512084728,
+      "grad_norm": 1.055009365081787,
+      "learning_rate": 0.0001724609393073526,
+      "loss": 3.1756,
+      "step": 2496
+    },
+    {
+      "epoch": 0.22603421743459762,
+      "grad_norm": 1.210542917251587,
+      "learning_rate": 0.00017243906637285568,
+      "loss": 3.1387,
+      "step": 2497
+    },
+    {
+      "epoch": 0.22612473974834796,
+      "grad_norm": 1.0770033597946167,
+      "learning_rate": 0.00017241718614374678,
+      "loss": 3.2614,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2262152620620983,
+      "grad_norm": 1.004407525062561,
+      "learning_rate": 0.00017239529862222918,
+      "loss": 3.1391,
+      "step": 2499
+    },
+    {
+      "epoch": 0.22630578437584864,
+      "grad_norm": 1.0340214967727661,
+      "learning_rate": 0.00017237340381050703,
+      "loss": 3.2116,
+      "step": 2500
+    },
+    {
+      "epoch": 0.22639630668959898,
+      "grad_norm": 1.049316167831421,
+      "learning_rate": 0.00017235150171078506,
+      "loss": 3.1887,
+      "step": 2501
+    },
+    {
+      "epoch": 0.22648682900334932,
+      "grad_norm": 0.9701083898544312,
+      "learning_rate": 0.0001723295923252689,
+      "loss": 3.1859,
+      "step": 2502
+    },
+    {
+      "epoch": 0.22657735131709966,
+      "grad_norm": 1.031803011894226,
+      "learning_rate": 0.00017230767565616474,
+      "loss": 3.2023,
+      "step": 2503
+    },
+    {
+      "epoch": 0.22666787363085,
+      "grad_norm": 0.998855710029602,
+      "learning_rate": 0.00017228575170567963,
+      "loss": 3.1229,
+      "step": 2504
+    },
+    {
+      "epoch": 0.22675839594460034,
+      "grad_norm": 0.975975513458252,
+      "learning_rate": 0.00017226382047602127,
+      "loss": 3.1604,
+      "step": 2505
+    },
+    {
+      "epoch": 0.22684891825835068,
+      "grad_norm": 0.9017311334609985,
+      "learning_rate": 0.00017224188196939818,
+      "loss": 3.1677,
+      "step": 2506
+    },
+    {
+      "epoch": 0.22693944057210103,
+      "grad_norm": 0.9536845088005066,
+      "learning_rate": 0.0001722199361880196,
+      "loss": 3.0564,
+      "step": 2507
+    },
+    {
+      "epoch": 0.22702996288585137,
+      "grad_norm": 0.8992807865142822,
+      "learning_rate": 0.00017219798313409537,
+      "loss": 3.0938,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2271204851996017,
+      "grad_norm": 0.9468969702720642,
+      "learning_rate": 0.00017217602280983623,
+      "loss": 3.2082,
+      "step": 2509
+    },
+    {
+      "epoch": 0.22721100751335205,
+      "grad_norm": 0.9722668528556824,
+      "learning_rate": 0.00017215405521745357,
+      "loss": 3.2445,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2273015298271024,
+      "grad_norm": 0.8341018557548523,
+      "learning_rate": 0.0001721320803591595,
+      "loss": 3.099,
+      "step": 2511
+    },
+    {
+      "epoch": 0.22739205214085273,
+      "grad_norm": 1.0530873537063599,
+      "learning_rate": 0.00017211009823716694,
+      "loss": 3.181,
+      "step": 2512
+    },
+    {
+      "epoch": 0.22748257445460307,
+      "grad_norm": 0.890476405620575,
+      "learning_rate": 0.00017208810885368947,
+      "loss": 3.1118,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2275730967683534,
+      "grad_norm": 1.012795329093933,
+      "learning_rate": 0.0001720661122109414,
+      "loss": 3.1775,
+      "step": 2514
+    },
+    {
+      "epoch": 0.22766361908210375,
+      "grad_norm": 0.9841723442077637,
+      "learning_rate": 0.00017204410831113778,
+      "loss": 3.1364,
+      "step": 2515
+    },
+    {
+      "epoch": 0.22775414139585407,
+      "grad_norm": 0.9760757684707642,
+      "learning_rate": 0.00017202209715649444,
+      "loss": 3.15,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2278446637096044,
+      "grad_norm": 1.123109221458435,
+      "learning_rate": 0.00017200007874922788,
+      "loss": 3.1347,
+      "step": 2517
+    },
+    {
+      "epoch": 0.22793518602335475,
+      "grad_norm": 0.9273945093154907,
+      "learning_rate": 0.00017197805309155536,
+      "loss": 3.1127,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2280257083371051,
+      "grad_norm": 1.2223721742630005,
+      "learning_rate": 0.00017195602018569488,
+      "loss": 3.1295,
+      "step": 2519
+    },
+    {
+      "epoch": 0.22811623065085543,
+      "grad_norm": 0.9610428214073181,
+      "learning_rate": 0.0001719339800338651,
+      "loss": 3.153,
+      "step": 2520
+    },
+    {
+      "epoch": 0.22820675296460577,
+      "grad_norm": 1.200606107711792,
+      "learning_rate": 0.00017191193263828554,
+      "loss": 3.1055,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2282972752783561,
+      "grad_norm": 1.0138325691223145,
+      "learning_rate": 0.00017188987800117634,
+      "loss": 3.1634,
+      "step": 2522
+    },
+    {
+      "epoch": 0.22838779759210645,
+      "grad_norm": 0.8626307845115662,
+      "learning_rate": 0.00017186781612475835,
+      "loss": 3.0981,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2284783199058568,
+      "grad_norm": 1.167111873626709,
+      "learning_rate": 0.00017184574701125326,
+      "loss": 3.1991,
+      "step": 2524
+    },
+    {
+      "epoch": 0.22856884221960713,
+      "grad_norm": 0.8812128901481628,
+      "learning_rate": 0.00017182367066288342,
+      "loss": 3.1453,
+      "step": 2525
+    },
+    {
+      "epoch": 0.22865936453335747,
+      "grad_norm": 0.9116116166114807,
+      "learning_rate": 0.00017180158708187188,
+      "loss": 3.1386,
+      "step": 2526
+    },
+    {
+      "epoch": 0.22874988684710781,
+      "grad_norm": 1.0221889019012451,
+      "learning_rate": 0.00017177949627044252,
+      "loss": 3.2163,
+      "step": 2527
+    },
+    {
+      "epoch": 0.22884040916085815,
+      "grad_norm": 0.9080731272697449,
+      "learning_rate": 0.00017175739823081985,
+      "loss": 3.0455,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2289309314746085,
+      "grad_norm": 0.9786624908447266,
+      "learning_rate": 0.00017173529296522912,
+      "loss": 3.1194,
+      "step": 2529
+    },
+    {
+      "epoch": 0.22902145378835884,
+      "grad_norm": 0.8339390158653259,
+      "learning_rate": 0.00017171318047589637,
+      "loss": 3.1286,
+      "step": 2530
+    },
+    {
+      "epoch": 0.22911197610210918,
+      "grad_norm": 1.001883625984192,
+      "learning_rate": 0.0001716910607650483,
+      "loss": 3.1536,
+      "step": 2531
+    },
+    {
+      "epoch": 0.22920249841585952,
+      "grad_norm": 0.9073264002799988,
+      "learning_rate": 0.00017166893383491235,
+      "loss": 3.1435,
+      "step": 2532
+    },
+    {
+      "epoch": 0.22929302072960986,
+      "grad_norm": 0.913053035736084,
+      "learning_rate": 0.00017164679968771675,
+      "loss": 3.1389,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2293835430433602,
+      "grad_norm": 0.9608319401741028,
+      "learning_rate": 0.00017162465832569037,
+      "loss": 3.1763,
+      "step": 2534
+    },
+    {
+      "epoch": 0.22947406535711054,
+      "grad_norm": 0.8648624420166016,
+      "learning_rate": 0.00017160250975106287,
+      "loss": 3.1384,
+      "step": 2535
+    },
+    {
+      "epoch": 0.22956458767086085,
+      "grad_norm": 1.0078084468841553,
+      "learning_rate": 0.00017158035396606458,
+      "loss": 3.1459,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2296551099846112,
+      "grad_norm": 0.9229320883750916,
+      "learning_rate": 0.0001715581909729266,
+      "loss": 3.1257,
+      "step": 2537
+    },
+    {
+      "epoch": 0.22974563229836154,
+      "grad_norm": 0.9432445168495178,
+      "learning_rate": 0.0001715360207738808,
+      "loss": 3.0445,
+      "step": 2538
+    },
+    {
+      "epoch": 0.22983615461211188,
+      "grad_norm": 0.9014883041381836,
+      "learning_rate": 0.00017151384337115962,
+      "loss": 3.1565,
+      "step": 2539
+    },
+    {
+      "epoch": 0.22992667692586222,
+      "grad_norm": 1.0331895351409912,
+      "learning_rate": 0.00017149165876699635,
+      "loss": 3.128,
+      "step": 2540
+    },
+    {
+      "epoch": 0.23001719923961256,
+      "grad_norm": 0.9233134388923645,
+      "learning_rate": 0.00017146946696362502,
+      "loss": 3.1212,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2301077215533629,
+      "grad_norm": 0.9826732277870178,
+      "learning_rate": 0.00017144726796328034,
+      "loss": 3.1716,
+      "step": 2542
+    },
+    {
+      "epoch": 0.23019824386711324,
+      "grad_norm": 0.9416763782501221,
+      "learning_rate": 0.0001714250617681977,
+      "loss": 3.1447,
+      "step": 2543
+    },
+    {
+      "epoch": 0.23028876618086358,
+      "grad_norm": 0.9064895510673523,
+      "learning_rate": 0.00017140284838061334,
+      "loss": 3.0929,
+      "step": 2544
+    },
+    {
+      "epoch": 0.23037928849461392,
+      "grad_norm": 0.9274477362632751,
+      "learning_rate": 0.00017138062780276404,
+      "loss": 3.1231,
+      "step": 2545
+    },
+    {
+      "epoch": 0.23046981080836426,
+      "grad_norm": 1.0109010934829712,
+      "learning_rate": 0.00017135840003688752,
+      "loss": 3.0921,
+      "step": 2546
+    },
+    {
+      "epoch": 0.2305603331221146,
+      "grad_norm": 0.9265955090522766,
+      "learning_rate": 0.00017133616508522206,
+      "loss": 3.0887,
+      "step": 2547
+    },
+    {
+      "epoch": 0.23065085543586494,
+      "grad_norm": 0.9467498064041138,
+      "learning_rate": 0.00017131392295000674,
+      "loss": 3.1189,
+      "step": 2548
+    },
+    {
+      "epoch": 0.23074137774961528,
+      "grad_norm": 0.9402139782905579,
+      "learning_rate": 0.0001712916736334813,
+      "loss": 3.1551,
+      "step": 2549
+    },
+    {
+      "epoch": 0.23083190006336562,
+      "grad_norm": 0.9011145830154419,
+      "learning_rate": 0.00017126941713788632,
+      "loss": 3.108,
+      "step": 2550
+    },
+    {
+      "epoch": 0.23092242237711597,
+      "grad_norm": 1.010998249053955,
+      "learning_rate": 0.00017124715346546293,
+      "loss": 3.1518,
+      "step": 2551
+    },
+    {
+      "epoch": 0.2310129446908663,
+      "grad_norm": 0.8391900062561035,
+      "learning_rate": 0.00017122488261845316,
+      "loss": 3.0762,
+      "step": 2552
+    },
+    {
+      "epoch": 0.23110346700461665,
+      "grad_norm": 0.9358366131782532,
+      "learning_rate": 0.00017120260459909967,
+      "loss": 3.12,
+      "step": 2553
+    },
+    {
+      "epoch": 0.231193989318367,
+      "grad_norm": 0.9052395820617676,
+      "learning_rate": 0.00017118031940964584,
+      "loss": 3.0629,
+      "step": 2554
+    },
+    {
+      "epoch": 0.23128451163211733,
+      "grad_norm": 0.8716477751731873,
+      "learning_rate": 0.00017115802705233577,
+      "loss": 3.1671,
+      "step": 2555
+    },
+    {
+      "epoch": 0.23137503394586767,
+      "grad_norm": 0.8970102667808533,
+      "learning_rate": 0.00017113572752941434,
+      "loss": 3.1503,
+      "step": 2556
+    },
+    {
+      "epoch": 0.23146555625961798,
+      "grad_norm": 0.9600470066070557,
+      "learning_rate": 0.00017111342084312708,
+      "loss": 3.1936,
+      "step": 2557
+    },
+    {
+      "epoch": 0.23155607857336832,
+      "grad_norm": 0.9360811114311218,
+      "learning_rate": 0.0001710911069957203,
+      "loss": 3.1651,
+      "step": 2558
+    },
+    {
+      "epoch": 0.23164660088711866,
+      "grad_norm": 0.9788040518760681,
+      "learning_rate": 0.000171068785989441,
+      "loss": 3.1569,
+      "step": 2559
+    },
+    {
+      "epoch": 0.231737123200869,
+      "grad_norm": 0.9954851269721985,
+      "learning_rate": 0.0001710464578265369,
+      "loss": 3.117,
+      "step": 2560
+    },
+    {
+      "epoch": 0.23182764551461935,
+      "grad_norm": 0.9106504321098328,
+      "learning_rate": 0.0001710241225092564,
+      "loss": 3.0862,
+      "step": 2561
+    },
+    {
+      "epoch": 0.2319181678283697,
+      "grad_norm": 0.9232210516929626,
+      "learning_rate": 0.00017100178003984874,
+      "loss": 3.1645,
+      "step": 2562
+    },
+    {
+      "epoch": 0.23200869014212003,
+      "grad_norm": 0.880970299243927,
+      "learning_rate": 0.00017097943042056378,
+      "loss": 3.1417,
+      "step": 2563
+    },
+    {
+      "epoch": 0.23209921245587037,
+      "grad_norm": 0.9681224226951599,
+      "learning_rate": 0.0001709570736536521,
+      "loss": 3.1376,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2321897347696207,
+      "grad_norm": 0.9339620471000671,
+      "learning_rate": 0.00017093470974136504,
+      "loss": 3.1156,
+      "step": 2565
+    },
+    {
+      "epoch": 0.23228025708337105,
+      "grad_norm": 0.9068838357925415,
+      "learning_rate": 0.00017091233868595467,
+      "loss": 3.1433,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2323707793971214,
+      "grad_norm": 0.8875805139541626,
+      "learning_rate": 0.00017088996048967374,
+      "loss": 3.1451,
+      "step": 2567
+    },
+    {
+      "epoch": 0.23246130171087173,
+      "grad_norm": 0.8893206119537354,
+      "learning_rate": 0.0001708675751547757,
+      "loss": 3.0724,
+      "step": 2568
+    },
+    {
+      "epoch": 0.23255182402462207,
+      "grad_norm": 0.8865343928337097,
+      "learning_rate": 0.00017084518268351478,
+      "loss": 3.0983,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2326423463383724,
+      "grad_norm": 0.8660908937454224,
+      "learning_rate": 0.00017082278307814592,
+      "loss": 3.0848,
+      "step": 2570
+    },
+    {
+      "epoch": 0.23273286865212275,
+      "grad_norm": 0.8525618314743042,
+      "learning_rate": 0.00017080037634092476,
+      "loss": 3.0969,
+      "step": 2571
+    },
+    {
+      "epoch": 0.2328233909658731,
+      "grad_norm": 0.9192307591438293,
+      "learning_rate": 0.0001707779624741076,
+      "loss": 3.223,
+      "step": 2572
+    },
+    {
+      "epoch": 0.23291391327962344,
+      "grad_norm": 0.9319369792938232,
+      "learning_rate": 0.0001707555414799516,
+      "loss": 3.075,
+      "step": 2573
+    },
+    {
+      "epoch": 0.23300443559337378,
+      "grad_norm": 0.9295352697372437,
+      "learning_rate": 0.00017073311336071446,
+      "loss": 3.1054,
+      "step": 2574
+    },
+    {
+      "epoch": 0.23309495790712412,
+      "grad_norm": 1.1705827713012695,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 3.1209,
+      "step": 2575
+    },
+    {
+      "epoch": 0.23318548022087446,
+      "grad_norm": 0.8612735271453857,
+      "learning_rate": 0.00017068823575603172,
+      "loss": 3.0827,
+      "step": 2576
+    },
+    {
+      "epoch": 0.23327600253462477,
+      "grad_norm": 0.9167778491973877,
+      "learning_rate": 0.00017066578627510527,
+      "loss": 3.0788,
+      "step": 2577
+    },
+    {
+      "epoch": 0.2333665248483751,
+      "grad_norm": 1.0142779350280762,
+      "learning_rate": 0.00017064332967813605,
+      "loss": 3.1004,
+      "step": 2578
+    },
+    {
+      "epoch": 0.23345704716212545,
+      "grad_norm": 0.8765705823898315,
+      "learning_rate": 0.0001706208659673855,
+      "loss": 3.0819,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2335475694758758,
+      "grad_norm": 0.9593830704689026,
+      "learning_rate": 0.00017059839514511565,
+      "loss": 3.1851,
+      "step": 2580
+    },
+    {
+      "epoch": 0.23363809178962613,
+      "grad_norm": 1.0137742757797241,
+      "learning_rate": 0.00017057591721358935,
+      "loss": 3.1235,
+      "step": 2581
+    },
+    {
+      "epoch": 0.23372861410337648,
+      "grad_norm": 0.873552680015564,
+      "learning_rate": 0.00017055343217507014,
+      "loss": 3.1555,
+      "step": 2582
+    },
+    {
+      "epoch": 0.23381913641712682,
+      "grad_norm": 0.8927232623100281,
+      "learning_rate": 0.0001705309400318222,
+      "loss": 3.1304,
+      "step": 2583
+    },
+    {
+      "epoch": 0.23390965873087716,
+      "grad_norm": 0.9595664739608765,
+      "learning_rate": 0.00017050844078611056,
+      "loss": 3.0944,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2340001810446275,
+      "grad_norm": 0.9072505831718445,
+      "learning_rate": 0.00017048593444020084,
+      "loss": 3.2111,
+      "step": 2585
+    },
+    {
+      "epoch": 0.23409070335837784,
+      "grad_norm": 0.8766895532608032,
+      "learning_rate": 0.00017046342099635948,
+      "loss": 3.1421,
+      "step": 2586
+    },
+    {
+      "epoch": 0.23418122567212818,
+      "grad_norm": 0.8676466941833496,
+      "learning_rate": 0.00017044090045685353,
+      "loss": 3.14,
+      "step": 2587
+    },
+    {
+      "epoch": 0.23427174798587852,
+      "grad_norm": 0.9251097440719604,
+      "learning_rate": 0.00017041837282395084,
+      "loss": 3.0815,
+      "step": 2588
+    },
+    {
+      "epoch": 0.23436227029962886,
+      "grad_norm": 0.8791188597679138,
+      "learning_rate": 0.00017039583809991992,
+      "loss": 3.1301,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2344527926133792,
+      "grad_norm": 0.9609639048576355,
+      "learning_rate": 0.00017037329628703004,
+      "loss": 3.1222,
+      "step": 2590
+    },
+    {
+      "epoch": 0.23454331492712954,
+      "grad_norm": 0.937321662902832,
+      "learning_rate": 0.00017035074738755115,
+      "loss": 3.0724,
+      "step": 2591
+    },
+    {
+      "epoch": 0.23463383724087988,
+      "grad_norm": 0.9102678894996643,
+      "learning_rate": 0.00017032819140375393,
+      "loss": 3.1259,
+      "step": 2592
+    },
+    {
+      "epoch": 0.23472435955463022,
+      "grad_norm": 0.9216590523719788,
+      "learning_rate": 0.00017030562833790977,
+      "loss": 3.1505,
+      "step": 2593
+    },
+    {
+      "epoch": 0.23481488186838056,
+      "grad_norm": 1.007556676864624,
+      "learning_rate": 0.00017028305819229074,
+      "loss": 3.1326,
+      "step": 2594
+    },
+    {
+      "epoch": 0.2349054041821309,
+      "grad_norm": 0.9005347490310669,
+      "learning_rate": 0.0001702604809691697,
+      "loss": 3.1449,
+      "step": 2595
+    },
+    {
+      "epoch": 0.23499592649588125,
+      "grad_norm": 1.139877438545227,
+      "learning_rate": 0.00017023789667082012,
+      "loss": 3.0806,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2350864488096316,
+      "grad_norm": 0.910658597946167,
+      "learning_rate": 0.00017021530529951625,
+      "loss": 3.1299,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2351769711233819,
+      "grad_norm": 0.9529627561569214,
+      "learning_rate": 0.0001701927068575331,
+      "loss": 3.1053,
+      "step": 2598
+    },
+    {
+      "epoch": 0.23526749343713224,
+      "grad_norm": 0.892433226108551,
+      "learning_rate": 0.00017017010134714626,
+      "loss": 3.0509,
+      "step": 2599
+    },
+    {
+      "epoch": 0.23535801575088258,
+      "grad_norm": 0.9055562019348145,
+      "learning_rate": 0.00017014748877063214,
+      "loss": 3.169,
+      "step": 2600
+    },
+    {
+      "epoch": 0.23544853806463292,
+      "grad_norm": 1.1402655839920044,
+      "learning_rate": 0.00017012486913026782,
+      "loss": 3.1448,
+      "step": 2601
+    },
+    {
+      "epoch": 0.23553906037838326,
+      "grad_norm": 0.8739327192306519,
+      "learning_rate": 0.0001701022424283311,
+      "loss": 3.1233,
+      "step": 2602
+    },
+    {
+      "epoch": 0.2356295826921336,
+      "grad_norm": 0.8963400721549988,
+      "learning_rate": 0.00017007960866710048,
+      "loss": 3.1718,
+      "step": 2603
+    },
+    {
+      "epoch": 0.23572010500588395,
+      "grad_norm": 1.0158209800720215,
+      "learning_rate": 0.00017005696784885518,
+      "loss": 3.1502,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2358106273196343,
+      "grad_norm": 0.9023421406745911,
+      "learning_rate": 0.00017003431997587515,
+      "loss": 3.1557,
+      "step": 2605
+    },
+    {
+      "epoch": 0.23590114963338463,
+      "grad_norm": 0.8342739343643188,
+      "learning_rate": 0.00017001166505044104,
+      "loss": 3.1859,
+      "step": 2606
+    },
+    {
+      "epoch": 0.23599167194713497,
+      "grad_norm": 0.8725590705871582,
+      "learning_rate": 0.00016998900307483412,
+      "loss": 3.0903,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2360821942608853,
+      "grad_norm": 0.9000881314277649,
+      "learning_rate": 0.00016996633405133655,
+      "loss": 3.1529,
+      "step": 2608
+    },
+    {
+      "epoch": 0.23617271657463565,
+      "grad_norm": 0.8588462471961975,
+      "learning_rate": 0.00016994365798223102,
+      "loss": 3.11,
+      "step": 2609
+    },
+    {
+      "epoch": 0.236263238888386,
+      "grad_norm": 0.8472119569778442,
+      "learning_rate": 0.00016992097486980107,
+      "loss": 3.093,
+      "step": 2610
+    },
+    {
+      "epoch": 0.23635376120213633,
+      "grad_norm": 0.881607711315155,
+      "learning_rate": 0.00016989828471633084,
+      "loss": 3.0996,
+      "step": 2611
+    },
+    {
+      "epoch": 0.23644428351588667,
+      "grad_norm": 0.9433134198188782,
+      "learning_rate": 0.0001698755875241053,
+      "loss": 3.1756,
+      "step": 2612
+    },
+    {
+      "epoch": 0.236534805829637,
+      "grad_norm": 0.8859644532203674,
+      "learning_rate": 0.00016985288329540997,
+      "loss": 3.0946,
+      "step": 2613
+    },
+    {
+      "epoch": 0.23662532814338735,
+      "grad_norm": 0.845602810382843,
+      "learning_rate": 0.00016983017203253122,
+      "loss": 3.1056,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2367158504571377,
+      "grad_norm": 0.9098728895187378,
+      "learning_rate": 0.00016980745373775604,
+      "loss": 3.1132,
+      "step": 2615
+    },
+    {
+      "epoch": 0.23680637277088804,
+      "grad_norm": 1.0199395418167114,
+      "learning_rate": 0.00016978472841337218,
+      "loss": 3.1743,
+      "step": 2616
+    },
+    {
+      "epoch": 0.23689689508463838,
+      "grad_norm": 1.2085365056991577,
+      "learning_rate": 0.00016976199606166809,
+      "loss": 3.0876,
+      "step": 2617
+    },
+    {
+      "epoch": 0.23698741739838872,
+      "grad_norm": 0.9905280470848083,
+      "learning_rate": 0.0001697392566849329,
+      "loss": 3.1126,
+      "step": 2618
+    },
+    {
+      "epoch": 0.23707793971213903,
+      "grad_norm": 1.0612152814865112,
+      "learning_rate": 0.00016971651028545648,
+      "loss": 3.1126,
+      "step": 2619
+    },
+    {
+      "epoch": 0.23716846202588937,
+      "grad_norm": 0.8954588770866394,
+      "learning_rate": 0.00016969375686552937,
+      "loss": 3.1352,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2372589843396397,
+      "grad_norm": 1.0112584829330444,
+      "learning_rate": 0.00016967099642744283,
+      "loss": 3.1017,
+      "step": 2621
+    },
+    {
+      "epoch": 0.23734950665339005,
+      "grad_norm": 0.9944405555725098,
+      "learning_rate": 0.0001696482289734889,
+      "loss": 3.1267,
+      "step": 2622
+    },
+    {
+      "epoch": 0.2374400289671404,
+      "grad_norm": 0.9250493049621582,
+      "learning_rate": 0.00016962545450596017,
+      "loss": 3.1542,
+      "step": 2623
+    },
+    {
+      "epoch": 0.23753055128089073,
+      "grad_norm": 1.0268713235855103,
+      "learning_rate": 0.0001696026730271501,
+      "loss": 3.1766,
+      "step": 2624
+    },
+    {
+      "epoch": 0.23762107359464107,
+      "grad_norm": 0.9544373750686646,
+      "learning_rate": 0.00016957988453935276,
+      "loss": 3.0965,
+      "step": 2625
+    },
+    {
+      "epoch": 0.23771159590839142,
+      "grad_norm": 0.8758040070533752,
+      "learning_rate": 0.00016955708904486296,
+      "loss": 3.1607,
+      "step": 2626
+    },
+    {
+      "epoch": 0.23780211822214176,
+      "grad_norm": 0.9426255822181702,
+      "learning_rate": 0.00016953428654597618,
+      "loss": 3.1236,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2378926405358921,
+      "grad_norm": 0.8743308782577515,
+      "learning_rate": 0.0001695114770449887,
+      "loss": 3.0598,
+      "step": 2628
+    },
+    {
+      "epoch": 0.23798316284964244,
+      "grad_norm": 0.9559704065322876,
+      "learning_rate": 0.00016948866054419733,
+      "loss": 3.0916,
+      "step": 2629
+    },
+    {
+      "epoch": 0.23807368516339278,
+      "grad_norm": 0.899918258190155,
+      "learning_rate": 0.00016946583704589973,
+      "loss": 3.153,
+      "step": 2630
+    },
+    {
+      "epoch": 0.23816420747714312,
+      "grad_norm": 1.040543556213379,
+      "learning_rate": 0.00016944300655239429,
+      "loss": 3.193,
+      "step": 2631
+    },
+    {
+      "epoch": 0.23825472979089346,
+      "grad_norm": 0.8428203463554382,
+      "learning_rate": 0.00016942016906597995,
+      "loss": 3.0544,
+      "step": 2632
+    },
+    {
+      "epoch": 0.2383452521046438,
+      "grad_norm": 0.9760965704917908,
+      "learning_rate": 0.0001693973245889565,
+      "loss": 3.0425,
+      "step": 2633
+    },
+    {
+      "epoch": 0.23843577441839414,
+      "grad_norm": 0.881698727607727,
+      "learning_rate": 0.00016937447312362437,
+      "loss": 3.0975,
+      "step": 2634
+    },
+    {
+      "epoch": 0.23852629673214448,
+      "grad_norm": 0.934918224811554,
+      "learning_rate": 0.00016935161467228465,
+      "loss": 3.1498,
+      "step": 2635
+    },
+    {
+      "epoch": 0.23861681904589482,
+      "grad_norm": 0.9894874691963196,
+      "learning_rate": 0.00016932874923723928,
+      "loss": 3.171,
+      "step": 2636
+    },
+    {
+      "epoch": 0.23870734135964516,
+      "grad_norm": 0.9783918857574463,
+      "learning_rate": 0.00016930587682079077,
+      "loss": 3.1563,
+      "step": 2637
+    },
+    {
+      "epoch": 0.2387978636733955,
+      "grad_norm": 0.9620823860168457,
+      "learning_rate": 0.00016928299742524234,
+      "loss": 3.1405,
+      "step": 2638
+    },
+    {
+      "epoch": 0.23888838598714582,
+      "grad_norm": 0.8668125867843628,
+      "learning_rate": 0.00016926011105289796,
+      "loss": 3.1277,
+      "step": 2639
+    },
+    {
+      "epoch": 0.23897890830089616,
+      "grad_norm": 1.424202799797058,
+      "learning_rate": 0.00016923721770606228,
+      "loss": 3.1993,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2390694306146465,
+      "grad_norm": 0.874039351940155,
+      "learning_rate": 0.0001692143173870407,
+      "loss": 3.1476,
+      "step": 2641
+    },
+    {
+      "epoch": 0.23915995292839684,
+      "grad_norm": 0.8891192674636841,
+      "learning_rate": 0.00016919141009813924,
+      "loss": 3.1399,
+      "step": 2642
+    },
+    {
+      "epoch": 0.23925047524214718,
+      "grad_norm": 0.954897940158844,
+      "learning_rate": 0.00016916849584166466,
+      "loss": 3.1055,
+      "step": 2643
+    },
+    {
+      "epoch": 0.23934099755589752,
+      "grad_norm": 0.9279910326004028,
+      "learning_rate": 0.00016914557461992447,
+      "loss": 3.1847,
+      "step": 2644
+    },
+    {
+      "epoch": 0.23943151986964786,
+      "grad_norm": 0.9511407017707825,
+      "learning_rate": 0.00016912264643522678,
+      "loss": 3.1576,
+      "step": 2645
+    },
+    {
+      "epoch": 0.2395220421833982,
+      "grad_norm": 0.9269322752952576,
+      "learning_rate": 0.00016909971128988049,
+      "loss": 3.1004,
+      "step": 2646
+    },
+    {
+      "epoch": 0.23961256449714854,
+      "grad_norm": 0.9328409433364868,
+      "learning_rate": 0.0001690767691861952,
+      "loss": 3.1329,
+      "step": 2647
+    },
+    {
+      "epoch": 0.23970308681089889,
+      "grad_norm": 0.8925078511238098,
+      "learning_rate": 0.00016905382012648109,
+      "loss": 3.1537,
+      "step": 2648
+    },
+    {
+      "epoch": 0.23979360912464923,
+      "grad_norm": 0.9077745079994202,
+      "learning_rate": 0.00016903086411304923,
+      "loss": 3.168,
+      "step": 2649
+    },
+    {
+      "epoch": 0.23988413143839957,
+      "grad_norm": 0.9630613327026367,
+      "learning_rate": 0.00016900790114821122,
+      "loss": 3.1688,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2399746537521499,
+      "grad_norm": 0.8002099990844727,
+      "learning_rate": 0.00016898493123427945,
+      "loss": 3.07,
+      "step": 2651
+    },
+    {
+      "epoch": 0.24006517606590025,
+      "grad_norm": 0.8331252336502075,
+      "learning_rate": 0.000168961954373567,
+      "loss": 3.0506,
+      "step": 2652
+    },
+    {
+      "epoch": 0.2401556983796506,
+      "grad_norm": 0.8587708473205566,
+      "learning_rate": 0.0001689389705683876,
+      "loss": 3.1304,
+      "step": 2653
+    },
+    {
+      "epoch": 0.24024622069340093,
+      "grad_norm": 0.9123610854148865,
+      "learning_rate": 0.0001689159798210558,
+      "loss": 3.1164,
+      "step": 2654
+    },
+    {
+      "epoch": 0.24033674300715127,
+      "grad_norm": 1.0066545009613037,
+      "learning_rate": 0.0001688929821338867,
+      "loss": 3.1512,
+      "step": 2655
+    },
+    {
+      "epoch": 0.2404272653209016,
+      "grad_norm": 0.8197920918464661,
+      "learning_rate": 0.00016886997750919619,
+      "loss": 3.1139,
+      "step": 2656
+    },
+    {
+      "epoch": 0.24051778763465195,
+      "grad_norm": 1.0735701322555542,
+      "learning_rate": 0.00016884696594930077,
+      "loss": 3.0909,
+      "step": 2657
+    },
+    {
+      "epoch": 0.2406083099484023,
+      "grad_norm": 1.0315901041030884,
+      "learning_rate": 0.00016882394745651783,
+      "loss": 3.0775,
+      "step": 2658
+    },
+    {
+      "epoch": 0.24069883226215263,
+      "grad_norm": 0.9101348519325256,
+      "learning_rate": 0.00016880092203316527,
+      "loss": 3.067,
+      "step": 2659
+    },
+    {
+      "epoch": 0.24078935457590295,
+      "grad_norm": 0.9687630534172058,
+      "learning_rate": 0.0001687778896815617,
+      "loss": 3.1138,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2408798768896533,
+      "grad_norm": 0.8488638997077942,
+      "learning_rate": 0.00016875485040402654,
+      "loss": 3.0299,
+      "step": 2661
+    },
+    {
+      "epoch": 0.24097039920340363,
+      "grad_norm": 0.9105443358421326,
+      "learning_rate": 0.0001687318042028798,
+      "loss": 3.0773,
+      "step": 2662
+    },
+    {
+      "epoch": 0.24106092151715397,
+      "grad_norm": 1.0244542360305786,
+      "learning_rate": 0.0001687087510804423,
+      "loss": 3.1329,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2411514438309043,
+      "grad_norm": 0.9051873683929443,
+      "learning_rate": 0.00016868569103903544,
+      "loss": 3.1798,
+      "step": 2664
+    },
+    {
+      "epoch": 0.24124196614465465,
+      "grad_norm": 0.916400134563446,
+      "learning_rate": 0.00016866262408098134,
+      "loss": 3.11,
+      "step": 2665
+    },
+    {
+      "epoch": 0.241332488458405,
+      "grad_norm": 1.0660948753356934,
+      "learning_rate": 0.0001686395502086029,
+      "loss": 3.1874,
+      "step": 2666
+    },
+    {
+      "epoch": 0.24142301077215533,
+      "grad_norm": 1.2193211317062378,
+      "learning_rate": 0.00016861646942422365,
+      "loss": 3.1298,
+      "step": 2667
+    },
+    {
+      "epoch": 0.24151353308590567,
+      "grad_norm": 0.9132906198501587,
+      "learning_rate": 0.0001685933817301678,
+      "loss": 3.1635,
+      "step": 2668
+    },
+    {
+      "epoch": 0.24160405539965601,
+      "grad_norm": 1.070044994354248,
+      "learning_rate": 0.00016857028712876024,
+      "loss": 3.1155,
+      "step": 2669
+    },
+    {
+      "epoch": 0.24169457771340636,
+      "grad_norm": 0.8624181747436523,
+      "learning_rate": 0.00016854718562232668,
+      "loss": 3.1299,
+      "step": 2670
+    },
+    {
+      "epoch": 0.2417851000271567,
+      "grad_norm": 0.9248359203338623,
+      "learning_rate": 0.0001685240772131934,
+      "loss": 3.086,
+      "step": 2671
+    },
+    {
+      "epoch": 0.24187562234090704,
+      "grad_norm": 0.9360045194625854,
+      "learning_rate": 0.0001685009619036874,
+      "loss": 3.0809,
+      "step": 2672
+    },
+    {
+      "epoch": 0.24196614465465738,
+      "grad_norm": 0.9229385852813721,
+      "learning_rate": 0.00016847783969613643,
+      "loss": 3.0644,
+      "step": 2673
+    },
+    {
+      "epoch": 0.24205666696840772,
+      "grad_norm": 1.0891914367675781,
+      "learning_rate": 0.00016845471059286887,
+      "loss": 3.1425,
+      "step": 2674
+    },
+    {
+      "epoch": 0.24214718928215806,
+      "grad_norm": 0.9290924668312073,
+      "learning_rate": 0.00016843157459621384,
+      "loss": 3.1359,
+      "step": 2675
+    },
+    {
+      "epoch": 0.2422377115959084,
+      "grad_norm": 1.0456758737564087,
+      "learning_rate": 0.0001684084317085011,
+      "loss": 3.1505,
+      "step": 2676
+    },
+    {
+      "epoch": 0.24232823390965874,
+      "grad_norm": 0.8679757714271545,
+      "learning_rate": 0.00016838528193206117,
+      "loss": 3.082,
+      "step": 2677
+    },
+    {
+      "epoch": 0.24241875622340908,
+      "grad_norm": 1.1940603256225586,
+      "learning_rate": 0.00016836212526922522,
+      "loss": 3.1754,
+      "step": 2678
+    },
+    {
+      "epoch": 0.24250927853715942,
+      "grad_norm": 0.8902338743209839,
+      "learning_rate": 0.00016833896172232512,
+      "loss": 3.0364,
+      "step": 2679
+    },
+    {
+      "epoch": 0.24259980085090974,
+      "grad_norm": 1.1721466779708862,
+      "learning_rate": 0.00016831579129369346,
+      "loss": 3.0953,
+      "step": 2680
+    },
+    {
+      "epoch": 0.24269032316466008,
+      "grad_norm": 0.9444583058357239,
+      "learning_rate": 0.00016829261398566348,
+      "loss": 3.1059,
+      "step": 2681
+    },
+    {
+      "epoch": 0.24278084547841042,
+      "grad_norm": 0.9138522744178772,
+      "learning_rate": 0.00016826942980056913,
+      "loss": 3.1372,
+      "step": 2682
+    },
+    {
+      "epoch": 0.24287136779216076,
+      "grad_norm": 0.9610081911087036,
+      "learning_rate": 0.00016824623874074507,
+      "loss": 3.0693,
+      "step": 2683
+    },
+    {
+      "epoch": 0.2429618901059111,
+      "grad_norm": 0.8005548715591431,
+      "learning_rate": 0.00016822304080852662,
+      "loss": 3.1198,
+      "step": 2684
+    },
+    {
+      "epoch": 0.24305241241966144,
+      "grad_norm": 1.113952875137329,
+      "learning_rate": 0.00016819983600624986,
+      "loss": 3.1332,
+      "step": 2685
+    },
+    {
+      "epoch": 0.24314293473341178,
+      "grad_norm": 0.8378565311431885,
+      "learning_rate": 0.00016817662433625148,
+      "loss": 3.0195,
+      "step": 2686
+    },
+    {
+      "epoch": 0.24323345704716212,
+      "grad_norm": 0.9602024555206299,
+      "learning_rate": 0.00016815340580086886,
+      "loss": 3.066,
+      "step": 2687
+    },
+    {
+      "epoch": 0.24332397936091246,
+      "grad_norm": 0.8759976029396057,
+      "learning_rate": 0.00016813018040244017,
+      "loss": 3.0734,
+      "step": 2688
+    },
+    {
+      "epoch": 0.2434145016746628,
+      "grad_norm": 0.8833613991737366,
+      "learning_rate": 0.00016810694814330416,
+      "loss": 3.1273,
+      "step": 2689
+    },
+    {
+      "epoch": 0.24350502398841314,
+      "grad_norm": 0.8197110295295715,
+      "learning_rate": 0.00016808370902580036,
+      "loss": 3.0442,
+      "step": 2690
+    },
+    {
+      "epoch": 0.24359554630216348,
+      "grad_norm": 0.891551673412323,
+      "learning_rate": 0.0001680604630522689,
+      "loss": 3.0359,
+      "step": 2691
+    },
+    {
+      "epoch": 0.24368606861591383,
+      "grad_norm": 0.8699698448181152,
+      "learning_rate": 0.00016803721022505067,
+      "loss": 3.1161,
+      "step": 2692
+    },
+    {
+      "epoch": 0.24377659092966417,
+      "grad_norm": 0.8947954177856445,
+      "learning_rate": 0.00016801395054648723,
+      "loss": 3.0821,
+      "step": 2693
+    },
+    {
+      "epoch": 0.2438671132434145,
+      "grad_norm": 0.9444195032119751,
+      "learning_rate": 0.00016799068401892087,
+      "loss": 3.0483,
+      "step": 2694
+    },
+    {
+      "epoch": 0.24395763555716485,
+      "grad_norm": 0.8962039351463318,
+      "learning_rate": 0.00016796741064469445,
+      "loss": 3.1249,
+      "step": 2695
+    },
+    {
+      "epoch": 0.2440481578709152,
+      "grad_norm": 0.8785931468009949,
+      "learning_rate": 0.00016794413042615168,
+      "loss": 3.141,
+      "step": 2696
+    },
+    {
+      "epoch": 0.24413868018466553,
+      "grad_norm": 0.9811086654663086,
+      "learning_rate": 0.00016792084336563677,
+      "loss": 3.0502,
+      "step": 2697
+    },
+    {
+      "epoch": 0.24422920249841587,
+      "grad_norm": 0.8483330607414246,
+      "learning_rate": 0.00016789754946549485,
+      "loss": 3.0703,
+      "step": 2698
+    },
+    {
+      "epoch": 0.2443197248121662,
+      "grad_norm": 1.045301079750061,
+      "learning_rate": 0.00016787424872807152,
+      "loss": 3.0932,
+      "step": 2699
+    },
+    {
+      "epoch": 0.24441024712591655,
+      "grad_norm": 0.9416995048522949,
+      "learning_rate": 0.00016785094115571322,
+      "loss": 3.1171,
+      "step": 2700
+    },
+    {
+      "epoch": 0.24450076943966687,
+      "grad_norm": 0.8743078112602234,
+      "learning_rate": 0.00016782762675076699,
+      "loss": 3.1714,
+      "step": 2701
+    },
+    {
+      "epoch": 0.2445912917534172,
+      "grad_norm": 0.9239590764045715,
+      "learning_rate": 0.00016780430551558063,
+      "loss": 3.064,
+      "step": 2702
+    },
+    {
+      "epoch": 0.24468181406716755,
+      "grad_norm": 0.8741731643676758,
+      "learning_rate": 0.00016778097745250253,
+      "loss": 3.1444,
+      "step": 2703
+    },
+    {
+      "epoch": 0.2447723363809179,
+      "grad_norm": 0.9072793126106262,
+      "learning_rate": 0.00016775764256388186,
+      "loss": 3.0598,
+      "step": 2704
+    },
+    {
+      "epoch": 0.24486285869466823,
+      "grad_norm": 0.8954857587814331,
+      "learning_rate": 0.00016773430085206847,
+      "loss": 3.0361,
+      "step": 2705
+    },
+    {
+      "epoch": 0.24495338100841857,
+      "grad_norm": 0.8716298341751099,
+      "learning_rate": 0.0001677109523194128,
+      "loss": 3.0546,
+      "step": 2706
+    },
+    {
+      "epoch": 0.2450439033221689,
+      "grad_norm": 1.0035932064056396,
+      "learning_rate": 0.00016768759696826608,
+      "loss": 3.1424,
+      "step": 2707
+    },
+    {
+      "epoch": 0.24513442563591925,
+      "grad_norm": 0.8431956768035889,
+      "learning_rate": 0.0001676642348009802,
+      "loss": 3.0365,
+      "step": 2708
+    },
+    {
+      "epoch": 0.2452249479496696,
+      "grad_norm": 0.8482758402824402,
+      "learning_rate": 0.00016764086581990775,
+      "loss": 3.109,
+      "step": 2709
+    },
+    {
+      "epoch": 0.24531547026341993,
+      "grad_norm": 0.9053434729576111,
+      "learning_rate": 0.00016761749002740193,
+      "loss": 3.056,
+      "step": 2710
+    },
+    {
+      "epoch": 0.24540599257717027,
+      "grad_norm": 0.9073513746261597,
+      "learning_rate": 0.00016759410742581677,
+      "loss": 3.2316,
+      "step": 2711
+    },
+    {
+      "epoch": 0.24549651489092061,
+      "grad_norm": 0.9571587443351746,
+      "learning_rate": 0.00016757071801750678,
+      "loss": 3.0679,
+      "step": 2712
+    },
+    {
+      "epoch": 0.24558703720467095,
+      "grad_norm": 0.9908713698387146,
+      "learning_rate": 0.0001675473218048274,
+      "loss": 3.0663,
+      "step": 2713
+    },
+    {
+      "epoch": 0.2456775595184213,
+      "grad_norm": 0.8795186877250671,
+      "learning_rate": 0.0001675239187901345,
+      "loss": 2.9681,
+      "step": 2714
+    },
+    {
+      "epoch": 0.24576808183217164,
+      "grad_norm": 0.8643364906311035,
+      "learning_rate": 0.00016750050897578484,
+      "loss": 3.1057,
+      "step": 2715
+    },
+    {
+      "epoch": 0.24585860414592198,
+      "grad_norm": 0.9430257678031921,
+      "learning_rate": 0.0001674770923641358,
+      "loss": 3.1365,
+      "step": 2716
+    },
+    {
+      "epoch": 0.24594912645967232,
+      "grad_norm": 0.9766716957092285,
+      "learning_rate": 0.0001674536689575454,
+      "loss": 3.0803,
+      "step": 2717
+    },
+    {
+      "epoch": 0.24603964877342266,
+      "grad_norm": 0.9355599284172058,
+      "learning_rate": 0.00016743023875837233,
+      "loss": 3.1598,
+      "step": 2718
+    },
+    {
+      "epoch": 0.246130171087173,
+      "grad_norm": 0.8610727190971375,
+      "learning_rate": 0.00016740680176897615,
+      "loss": 3.0801,
+      "step": 2719
+    },
+    {
+      "epoch": 0.24622069340092334,
+      "grad_norm": 1.0121809244155884,
+      "learning_rate": 0.00016738335799171682,
+      "loss": 3.0124,
+      "step": 2720
+    },
+    {
+      "epoch": 0.24631121571467365,
+      "grad_norm": 0.8912957310676575,
+      "learning_rate": 0.00016735990742895517,
+      "loss": 3.0346,
+      "step": 2721
+    },
+    {
+      "epoch": 0.246401738028424,
+      "grad_norm": 0.9681625962257385,
+      "learning_rate": 0.00016733645008305272,
+      "loss": 3.0641,
+      "step": 2722
+    },
+    {
+      "epoch": 0.24649226034217434,
+      "grad_norm": 0.9645764827728271,
+      "learning_rate": 0.00016731298595637156,
+      "loss": 3.1159,
+      "step": 2723
+    },
+    {
+      "epoch": 0.24658278265592468,
+      "grad_norm": 0.8538256883621216,
+      "learning_rate": 0.00016728951505127457,
+      "loss": 3.0658,
+      "step": 2724
+    },
+    {
+      "epoch": 0.24667330496967502,
+      "grad_norm": 0.8959582448005676,
+      "learning_rate": 0.00016726603737012529,
+      "loss": 3.0969,
+      "step": 2725
+    },
+    {
+      "epoch": 0.24676382728342536,
+      "grad_norm": 0.9084468483924866,
+      "learning_rate": 0.0001672425529152878,
+      "loss": 3.0568,
+      "step": 2726
+    },
+    {
+      "epoch": 0.2468543495971757,
+      "grad_norm": 0.8262922763824463,
+      "learning_rate": 0.00016721906168912717,
+      "loss": 3.1008,
+      "step": 2727
+    },
+    {
+      "epoch": 0.24694487191092604,
+      "grad_norm": 1.1768760681152344,
+      "learning_rate": 0.0001671955636940088,
+      "loss": 3.0462,
+      "step": 2728
+    },
+    {
+      "epoch": 0.24703539422467638,
+      "grad_norm": 0.8710028529167175,
+      "learning_rate": 0.00016717205893229903,
+      "loss": 3.0722,
+      "step": 2729
+    },
+    {
+      "epoch": 0.24712591653842672,
+      "grad_norm": 1.082719326019287,
+      "learning_rate": 0.00016714854740636478,
+      "loss": 3.0794,
+      "step": 2730
+    },
+    {
+      "epoch": 0.24721643885217706,
+      "grad_norm": 0.9447751641273499,
+      "learning_rate": 0.0001671250291185736,
+      "loss": 3.0471,
+      "step": 2731
+    },
+    {
+      "epoch": 0.2473069611659274,
+      "grad_norm": 1.1959116458892822,
+      "learning_rate": 0.00016710150407129386,
+      "loss": 3.2106,
+      "step": 2732
+    },
+    {
+      "epoch": 0.24739748347967774,
+      "grad_norm": 0.9239301085472107,
+      "learning_rate": 0.00016707797226689448,
+      "loss": 3.0975,
+      "step": 2733
+    },
+    {
+      "epoch": 0.24748800579342808,
+      "grad_norm": 0.8619546890258789,
+      "learning_rate": 0.00016705443370774515,
+      "loss": 3.2133,
+      "step": 2734
+    },
+    {
+      "epoch": 0.24757852810717842,
+      "grad_norm": 1.0291749238967896,
+      "learning_rate": 0.00016703088839621613,
+      "loss": 3.0778,
+      "step": 2735
+    },
+    {
+      "epoch": 0.24766905042092877,
+      "grad_norm": 0.9417911767959595,
+      "learning_rate": 0.00016700733633467853,
+      "loss": 3.0914,
+      "step": 2736
+    },
+    {
+      "epoch": 0.2477595727346791,
+      "grad_norm": 1.1757731437683105,
+      "learning_rate": 0.000166983777525504,
+      "loss": 3.0828,
+      "step": 2737
+    },
+    {
+      "epoch": 0.24785009504842945,
+      "grad_norm": 0.9673495888710022,
+      "learning_rate": 0.00016696021197106487,
+      "loss": 3.0929,
+      "step": 2738
+    },
+    {
+      "epoch": 0.2479406173621798,
+      "grad_norm": 0.9365146160125732,
+      "learning_rate": 0.00016693663967373422,
+      "loss": 3.0604,
+      "step": 2739
+    },
+    {
+      "epoch": 0.24803113967593013,
+      "grad_norm": 0.9483555555343628,
+      "learning_rate": 0.00016691306063588583,
+      "loss": 3.0947,
+      "step": 2740
+    },
+    {
+      "epoch": 0.24812166198968047,
+      "grad_norm": 0.8077372312545776,
+      "learning_rate": 0.00016688947485989406,
+      "loss": 3.0894,
+      "step": 2741
+    },
+    {
+      "epoch": 0.24821218430343078,
+      "grad_norm": 0.9072548747062683,
+      "learning_rate": 0.000166865882348134,
+      "loss": 3.0464,
+      "step": 2742
+    },
+    {
+      "epoch": 0.24830270661718112,
+      "grad_norm": 0.8407996892929077,
+      "learning_rate": 0.00016684228310298138,
+      "loss": 3.0608,
+      "step": 2743
+    },
+    {
+      "epoch": 0.24839322893093146,
+      "grad_norm": 0.9338646531105042,
+      "learning_rate": 0.0001668186771268127,
+      "loss": 3.1346,
+      "step": 2744
+    },
+    {
+      "epoch": 0.2484837512446818,
+      "grad_norm": 0.9149847626686096,
+      "learning_rate": 0.00016679506442200505,
+      "loss": 3.0997,
+      "step": 2745
+    },
+    {
+      "epoch": 0.24857427355843215,
+      "grad_norm": 0.8203171491622925,
+      "learning_rate": 0.00016677144499093626,
+      "loss": 3.0163,
+      "step": 2746
+    },
+    {
+      "epoch": 0.2486647958721825,
+      "grad_norm": 0.8672574758529663,
+      "learning_rate": 0.00016674781883598477,
+      "loss": 3.0743,
+      "step": 2747
+    },
+    {
+      "epoch": 0.24875531818593283,
+      "grad_norm": 0.9464472532272339,
+      "learning_rate": 0.0001667241859595298,
+      "loss": 3.1096,
+      "step": 2748
+    },
+    {
+      "epoch": 0.24884584049968317,
+      "grad_norm": 0.8735107779502869,
+      "learning_rate": 0.00016670054636395108,
+      "loss": 3.1216,
+      "step": 2749
+    },
+    {
+      "epoch": 0.2489363628134335,
+      "grad_norm": 1.1609880924224854,
+      "learning_rate": 0.00016667690005162916,
+      "loss": 3.0835,
+      "step": 2750
+    },
+    {
+      "epoch": 0.24902688512718385,
+      "grad_norm": 0.8689505457878113,
+      "learning_rate": 0.00016665324702494524,
+      "loss": 3.1091,
+      "step": 2751
+    },
+    {
+      "epoch": 0.2491174074409342,
+      "grad_norm": 0.8796523809432983,
+      "learning_rate": 0.0001666295872862812,
+      "loss": 3.0843,
+      "step": 2752
+    },
+    {
+      "epoch": 0.24920792975468453,
+      "grad_norm": 0.9133955836296082,
+      "learning_rate": 0.0001666059208380195,
+      "loss": 3.1447,
+      "step": 2753
+    },
+    {
+      "epoch": 0.24929845206843487,
+      "grad_norm": 0.8879693150520325,
+      "learning_rate": 0.00016658224768254342,
+      "loss": 3.047,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2493889743821852,
+      "grad_norm": 0.9240202307701111,
+      "learning_rate": 0.00016655856782223682,
+      "loss": 3.0566,
+      "step": 2755
+    },
+    {
+      "epoch": 0.24947949669593555,
+      "grad_norm": 0.9515302777290344,
+      "learning_rate": 0.00016653488125948425,
+      "loss": 3.1346,
+      "step": 2756
+    },
+    {
+      "epoch": 0.2495700190096859,
+      "grad_norm": 0.8590241074562073,
+      "learning_rate": 0.00016651118799667096,
+      "loss": 3.0785,
+      "step": 2757
+    },
+    {
+      "epoch": 0.24966054132343624,
+      "grad_norm": 0.9613081216812134,
+      "learning_rate": 0.00016648748803618286,
+      "loss": 3.1146,
+      "step": 2758
+    },
+    {
+      "epoch": 0.24975106363718658,
+      "grad_norm": 0.8211583495140076,
+      "learning_rate": 0.00016646378138040655,
+      "loss": 3.0034,
+      "step": 2759
+    },
+    {
+      "epoch": 0.24984158595093692,
+      "grad_norm": 0.9760122299194336,
+      "learning_rate": 0.00016644006803172924,
+      "loss": 3.0653,
+      "step": 2760
+    },
+    {
+      "epoch": 0.24993210826468726,
+      "grad_norm": 0.8821545243263245,
+      "learning_rate": 0.0001664163479925389,
+      "loss": 3.1247,
+      "step": 2761
+    },
+    {
+      "epoch": 0.2500226305784376,
+      "grad_norm": 1.0329774618148804,
+      "learning_rate": 0.00016639262126522418,
+      "loss": 3.0699,
+      "step": 2762
+    },
+    {
+      "epoch": 0.25011315289218794,
+      "grad_norm": 0.9172666668891907,
+      "learning_rate": 0.00016636888785217426,
+      "loss": 3.1346,
+      "step": 2763
+    },
+    {
+      "epoch": 0.2502036752059383,
+      "grad_norm": 0.9607560634613037,
+      "learning_rate": 0.0001663451477557792,
+      "loss": 3.1161,
+      "step": 2764
+    },
+    {
+      "epoch": 0.2502941975196886,
+      "grad_norm": 0.8451930284500122,
+      "learning_rate": 0.00016632140097842953,
+      "loss": 3.0687,
+      "step": 2765
+    },
+    {
+      "epoch": 0.25038471983343896,
+      "grad_norm": 1.0101604461669922,
+      "learning_rate": 0.00016629764752251665,
+      "loss": 3.1201,
+      "step": 2766
+    },
+    {
+      "epoch": 0.2504752421471893,
+      "grad_norm": 0.8747375011444092,
+      "learning_rate": 0.00016627388739043242,
+      "loss": 3.1586,
+      "step": 2767
+    },
+    {
+      "epoch": 0.25056576446093964,
+      "grad_norm": 0.87981778383255,
+      "learning_rate": 0.00016625012058456957,
+      "loss": 3.0545,
+      "step": 2768
+    },
+    {
+      "epoch": 0.25065628677469,
+      "grad_norm": 1.0310460329055786,
+      "learning_rate": 0.00016622634710732138,
+      "loss": 3.0795,
+      "step": 2769
+    },
+    {
+      "epoch": 0.2507468090884403,
+      "grad_norm": 0.8559186458587646,
+      "learning_rate": 0.00016620256696108188,
+      "loss": 3.0805,
+      "step": 2770
+    },
+    {
+      "epoch": 0.25083733140219067,
+      "grad_norm": 0.988170325756073,
+      "learning_rate": 0.00016617878014824563,
+      "loss": 3.1125,
+      "step": 2771
+    },
+    {
+      "epoch": 0.250927853715941,
+      "grad_norm": 0.917720377445221,
+      "learning_rate": 0.0001661549866712081,
+      "loss": 3.0562,
+      "step": 2772
+    },
+    {
+      "epoch": 0.2510183760296913,
+      "grad_norm": 0.9410594701766968,
+      "learning_rate": 0.00016613118653236518,
+      "loss": 3.1102,
+      "step": 2773
+    },
+    {
+      "epoch": 0.25110889834344163,
+      "grad_norm": 0.9203336834907532,
+      "learning_rate": 0.00016610737973411362,
+      "loss": 3.1359,
+      "step": 2774
+    },
+    {
+      "epoch": 0.251199420657192,
+      "grad_norm": 0.9632644057273865,
+      "learning_rate": 0.00016608356627885072,
+      "loss": 3.0328,
+      "step": 2775
+    },
+    {
+      "epoch": 0.2512899429709423,
+      "grad_norm": 0.8281729817390442,
+      "learning_rate": 0.00016605974616897449,
+      "loss": 3.0756,
+      "step": 2776
+    },
+    {
+      "epoch": 0.25138046528469266,
+      "grad_norm": 0.9495511651039124,
+      "learning_rate": 0.00016603591940688364,
+      "loss": 3.0499,
+      "step": 2777
+    },
+    {
+      "epoch": 0.251470987598443,
+      "grad_norm": 0.8314706683158875,
+      "learning_rate": 0.00016601208599497752,
+      "loss": 3.036,
+      "step": 2778
+    },
+    {
+      "epoch": 0.25156150991219334,
+      "grad_norm": 1.015807032585144,
+      "learning_rate": 0.0001659882459356561,
+      "loss": 3.0966,
+      "step": 2779
+    },
+    {
+      "epoch": 0.2516520322259437,
+      "grad_norm": 0.9160824418067932,
+      "learning_rate": 0.00016596439923132017,
+      "loss": 3.0643,
+      "step": 2780
+    },
+    {
+      "epoch": 0.251742554539694,
+      "grad_norm": 1.1869463920593262,
+      "learning_rate": 0.00016594054588437102,
+      "loss": 3.0994,
+      "step": 2781
+    },
+    {
+      "epoch": 0.25183307685344436,
+      "grad_norm": 1.000400185585022,
+      "learning_rate": 0.0001659166858972107,
+      "loss": 3.0737,
+      "step": 2782
+    },
+    {
+      "epoch": 0.2519235991671947,
+      "grad_norm": 1.0648994445800781,
+      "learning_rate": 0.00016589281927224193,
+      "loss": 3.0411,
+      "step": 2783
+    },
+    {
+      "epoch": 0.25201412148094504,
+      "grad_norm": 0.989122748374939,
+      "learning_rate": 0.00016586894601186805,
+      "loss": 3.089,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2521046437946954,
+      "grad_norm": 0.8601987361907959,
+      "learning_rate": 0.0001658450661184931,
+      "loss": 3.0815,
+      "step": 2785
+    },
+    {
+      "epoch": 0.2521951661084457,
+      "grad_norm": 1.0197594165802002,
+      "learning_rate": 0.00016582117959452184,
+      "loss": 3.129,
+      "step": 2786
+    },
+    {
+      "epoch": 0.25228568842219606,
+      "grad_norm": 0.9159774780273438,
+      "learning_rate": 0.00016579728644235955,
+      "loss": 3.1056,
+      "step": 2787
+    },
+    {
+      "epoch": 0.2523762107359464,
+      "grad_norm": 0.8355146050453186,
+      "learning_rate": 0.00016577338666441232,
+      "loss": 3.1091,
+      "step": 2788
+    },
+    {
+      "epoch": 0.25246673304969675,
+      "grad_norm": 0.92014080286026,
+      "learning_rate": 0.00016574948026308687,
+      "loss": 3.1115,
+      "step": 2789
+    },
+    {
+      "epoch": 0.2525572553634471,
+      "grad_norm": 0.9154512882232666,
+      "learning_rate": 0.00016572556724079056,
+      "loss": 3.1172,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2526477776771974,
+      "grad_norm": 0.859870970249176,
+      "learning_rate": 0.00016570164759993142,
+      "loss": 3.0903,
+      "step": 2791
+    },
+    {
+      "epoch": 0.25273829999094777,
+      "grad_norm": 0.946029782295227,
+      "learning_rate": 0.00016567772134291816,
+      "loss": 3.0855,
+      "step": 2792
+    },
+    {
+      "epoch": 0.2528288223046981,
+      "grad_norm": 0.9203336834907532,
+      "learning_rate": 0.00016565378847216016,
+      "loss": 3.0622,
+      "step": 2793
+    },
+    {
+      "epoch": 0.25291934461844845,
+      "grad_norm": 0.9952797293663025,
+      "learning_rate": 0.00016562984899006744,
+      "loss": 3.0854,
+      "step": 2794
+    },
+    {
+      "epoch": 0.2530098669321988,
+      "grad_norm": 0.849673330783844,
+      "learning_rate": 0.00016560590289905073,
+      "loss": 3.0319,
+      "step": 2795
+    },
+    {
+      "epoch": 0.25310038924594913,
+      "grad_norm": 0.857610821723938,
+      "learning_rate": 0.0001655819502015214,
+      "loss": 3.0494,
+      "step": 2796
+    },
+    {
+      "epoch": 0.25319091155969947,
+      "grad_norm": 1.0628398656845093,
+      "learning_rate": 0.0001655579908998915,
+      "loss": 3.1293,
+      "step": 2797
+    },
+    {
+      "epoch": 0.2532814338734498,
+      "grad_norm": 0.8674313426017761,
+      "learning_rate": 0.0001655340249965737,
+      "loss": 3.0552,
+      "step": 2798
+    },
+    {
+      "epoch": 0.25337195618720015,
+      "grad_norm": 0.9114541411399841,
+      "learning_rate": 0.00016551005249398134,
+      "loss": 3.0008,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2534624785009505,
+      "grad_norm": 0.9368698000907898,
+      "learning_rate": 0.00016548607339452853,
+      "loss": 3.0419,
+      "step": 2800
+    },
+    {
+      "epoch": 0.25355300081470084,
+      "grad_norm": 0.9874829649925232,
+      "learning_rate": 0.0001654620877006299,
+      "loss": 3.0858,
+      "step": 2801
+    },
+    {
+      "epoch": 0.2536435231284512,
+      "grad_norm": 0.9258869290351868,
+      "learning_rate": 0.00016543809541470084,
+      "loss": 3.0617,
+      "step": 2802
+    },
+    {
+      "epoch": 0.2537340454422015,
+      "grad_norm": 0.924634575843811,
+      "learning_rate": 0.00016541409653915735,
+      "loss": 3.0395,
+      "step": 2803
+    },
+    {
+      "epoch": 0.25382456775595186,
+      "grad_norm": 0.8764649033546448,
+      "learning_rate": 0.00016539009107641612,
+      "loss": 3.0184,
+      "step": 2804
+    },
+    {
+      "epoch": 0.2539150900697022,
+      "grad_norm": 0.8432656526565552,
+      "learning_rate": 0.00016536607902889453,
+      "loss": 3.0731,
+      "step": 2805
+    },
+    {
+      "epoch": 0.25400561238345254,
+      "grad_norm": 0.8797190189361572,
+      "learning_rate": 0.00016534206039901057,
+      "loss": 3.1247,
+      "step": 2806
+    },
+    {
+      "epoch": 0.2540961346972029,
+      "grad_norm": 0.9437577128410339,
+      "learning_rate": 0.00016531803518918289,
+      "loss": 3.0625,
+      "step": 2807
+    },
+    {
+      "epoch": 0.2541866570109532,
+      "grad_norm": 0.8880354166030884,
+      "learning_rate": 0.00016529400340183087,
+      "loss": 3.0751,
+      "step": 2808
+    },
+    {
+      "epoch": 0.25427717932470356,
+      "grad_norm": 0.8828173875808716,
+      "learning_rate": 0.00016526996503937448,
+      "loss": 3.0696,
+      "step": 2809
+    },
+    {
+      "epoch": 0.2543677016384539,
+      "grad_norm": 0.9315621852874756,
+      "learning_rate": 0.00016524592010423443,
+      "loss": 3.0851,
+      "step": 2810
+    },
+    {
+      "epoch": 0.25445822395220424,
+      "grad_norm": 0.8461422324180603,
+      "learning_rate": 0.000165221868598832,
+      "loss": 3.0212,
+      "step": 2811
+    },
+    {
+      "epoch": 0.2545487462659546,
+      "grad_norm": 0.902062714099884,
+      "learning_rate": 0.00016519781052558917,
+      "loss": 3.0147,
+      "step": 2812
+    },
+    {
+      "epoch": 0.2546392685797049,
+      "grad_norm": 0.9265572428703308,
+      "learning_rate": 0.0001651737458869286,
+      "loss": 3.0852,
+      "step": 2813
+    },
+    {
+      "epoch": 0.2547297908934552,
+      "grad_norm": 1.1529960632324219,
+      "learning_rate": 0.00016514967468527365,
+      "loss": 3.057,
+      "step": 2814
+    },
+    {
+      "epoch": 0.25482031320720555,
+      "grad_norm": 0.894597589969635,
+      "learning_rate": 0.0001651255969230482,
+      "loss": 2.9886,
+      "step": 2815
+    },
+    {
+      "epoch": 0.2549108355209559,
+      "grad_norm": 0.9490203261375427,
+      "learning_rate": 0.00016510151260267693,
+      "loss": 3.0935,
+      "step": 2816
+    },
+    {
+      "epoch": 0.25500135783470623,
+      "grad_norm": 0.8473650813102722,
+      "learning_rate": 0.0001650774217265851,
+      "loss": 3.1135,
+      "step": 2817
+    },
+    {
+      "epoch": 0.2550918801484566,
+      "grad_norm": 0.913254976272583,
+      "learning_rate": 0.0001650533242971987,
+      "loss": 3.1171,
+      "step": 2818
+    },
+    {
+      "epoch": 0.2551824024622069,
+      "grad_norm": 0.9239524006843567,
+      "learning_rate": 0.00016502922031694434,
+      "loss": 3.0479,
+      "step": 2819
+    },
+    {
+      "epoch": 0.25527292477595726,
+      "grad_norm": 0.9490577578544617,
+      "learning_rate": 0.00016500510978824926,
+      "loss": 3.0795,
+      "step": 2820
+    },
+    {
+      "epoch": 0.2553634470897076,
+      "grad_norm": 1.005797028541565,
+      "learning_rate": 0.0001649809927135414,
+      "loss": 3.0178,
+      "step": 2821
+    },
+    {
+      "epoch": 0.25545396940345794,
+      "grad_norm": 0.8729204535484314,
+      "learning_rate": 0.00016495686909524934,
+      "loss": 3.113,
+      "step": 2822
+    },
+    {
+      "epoch": 0.2555444917172083,
+      "grad_norm": 0.9689465165138245,
+      "learning_rate": 0.00016493273893580237,
+      "loss": 3.1293,
+      "step": 2823
+    },
+    {
+      "epoch": 0.2556350140309586,
+      "grad_norm": 0.9394911527633667,
+      "learning_rate": 0.00016490860223763036,
+      "loss": 2.996,
+      "step": 2824
+    },
+    {
+      "epoch": 0.25572553634470896,
+      "grad_norm": 1.1007540225982666,
+      "learning_rate": 0.00016488445900316386,
+      "loss": 3.0747,
+      "step": 2825
+    },
+    {
+      "epoch": 0.2558160586584593,
+      "grad_norm": 0.8632012009620667,
+      "learning_rate": 0.00016486030923483413,
+      "loss": 3.0662,
+      "step": 2826
+    },
+    {
+      "epoch": 0.25590658097220964,
+      "grad_norm": 0.9733211994171143,
+      "learning_rate": 0.00016483615293507304,
+      "loss": 3.0925,
+      "step": 2827
+    },
+    {
+      "epoch": 0.25599710328596,
+      "grad_norm": 0.9308933019638062,
+      "learning_rate": 0.0001648119901063131,
+      "loss": 3.0849,
+      "step": 2828
+    },
+    {
+      "epoch": 0.2560876255997103,
+      "grad_norm": 1.0638294219970703,
+      "learning_rate": 0.00016478782075098755,
+      "loss": 3.0717,
+      "step": 2829
+    },
+    {
+      "epoch": 0.25617814791346066,
+      "grad_norm": 0.9067211151123047,
+      "learning_rate": 0.00016476364487153023,
+      "loss": 3.0821,
+      "step": 2830
+    },
+    {
+      "epoch": 0.256268670227211,
+      "grad_norm": 0.9581461548805237,
+      "learning_rate": 0.00016473946247037562,
+      "loss": 3.0636,
+      "step": 2831
+    },
+    {
+      "epoch": 0.25635919254096134,
+      "grad_norm": 0.9932011365890503,
+      "learning_rate": 0.0001647152735499589,
+      "loss": 3.0906,
+      "step": 2832
+    },
+    {
+      "epoch": 0.2564497148547117,
+      "grad_norm": 0.9417665004730225,
+      "learning_rate": 0.00016469107811271593,
+      "loss": 3.0517,
+      "step": 2833
+    },
+    {
+      "epoch": 0.256540237168462,
+      "grad_norm": 1.0941578149795532,
+      "learning_rate": 0.00016466687616108316,
+      "loss": 3.0603,
+      "step": 2834
+    },
+    {
+      "epoch": 0.25663075948221237,
+      "grad_norm": 0.9311975836753845,
+      "learning_rate": 0.00016464266769749774,
+      "loss": 3.0018,
+      "step": 2835
+    },
+    {
+      "epoch": 0.2567212817959627,
+      "grad_norm": 1.0369527339935303,
+      "learning_rate": 0.00016461845272439741,
+      "loss": 3.0921,
+      "step": 2836
+    },
+    {
+      "epoch": 0.25681180410971305,
+      "grad_norm": 0.8587122559547424,
+      "learning_rate": 0.00016459423124422072,
+      "loss": 3.0498,
+      "step": 2837
+    },
+    {
+      "epoch": 0.2569023264234634,
+      "grad_norm": 0.8985700011253357,
+      "learning_rate": 0.00016457000325940667,
+      "loss": 3.088,
+      "step": 2838
+    },
+    {
+      "epoch": 0.25699284873721373,
+      "grad_norm": 0.8981794118881226,
+      "learning_rate": 0.00016454576877239507,
+      "loss": 3.0898,
+      "step": 2839
+    },
+    {
+      "epoch": 0.25708337105096407,
+      "grad_norm": 0.9229661226272583,
+      "learning_rate": 0.0001645215277856263,
+      "loss": 3.0688,
+      "step": 2840
+    },
+    {
+      "epoch": 0.2571738933647144,
+      "grad_norm": 0.9163200259208679,
+      "learning_rate": 0.00016449728030154147,
+      "loss": 3.0581,
+      "step": 2841
+    },
+    {
+      "epoch": 0.25726441567846475,
+      "grad_norm": 0.8850761651992798,
+      "learning_rate": 0.0001644730263225823,
+      "loss": 3.002,
+      "step": 2842
+    },
+    {
+      "epoch": 0.2573549379922151,
+      "grad_norm": 0.9454545378684998,
+      "learning_rate": 0.00016444876585119112,
+      "loss": 3.056,
+      "step": 2843
+    },
+    {
+      "epoch": 0.25744546030596543,
+      "grad_norm": 0.9575306177139282,
+      "learning_rate": 0.00016442449888981098,
+      "loss": 3.0659,
+      "step": 2844
+    },
+    {
+      "epoch": 0.2575359826197158,
+      "grad_norm": 0.8770989775657654,
+      "learning_rate": 0.00016440022544088553,
+      "loss": 3.0354,
+      "step": 2845
+    },
+    {
+      "epoch": 0.2576265049334661,
+      "grad_norm": 0.899128258228302,
+      "learning_rate": 0.0001643759455068592,
+      "loss": 3.1051,
+      "step": 2846
+    },
+    {
+      "epoch": 0.25771702724721646,
+      "grad_norm": 0.9273167252540588,
+      "learning_rate": 0.00016435165909017688,
+      "loss": 3.062,
+      "step": 2847
+    },
+    {
+      "epoch": 0.2578075495609668,
+      "grad_norm": 0.8725229501724243,
+      "learning_rate": 0.00016432736619328425,
+      "loss": 3.0726,
+      "step": 2848
+    },
+    {
+      "epoch": 0.25789807187471714,
+      "grad_norm": 0.8743504285812378,
+      "learning_rate": 0.00016430306681862765,
+      "loss": 3.0016,
+      "step": 2849
+    },
+    {
+      "epoch": 0.2579885941884675,
+      "grad_norm": 0.894152820110321,
+      "learning_rate": 0.00016427876096865394,
+      "loss": 3.1186,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2580791165022178,
+      "grad_norm": 0.8980875015258789,
+      "learning_rate": 0.00016425444864581076,
+      "loss": 3.1181,
+      "step": 2851
+    },
+    {
+      "epoch": 0.25816963881596816,
+      "grad_norm": 1.2219386100769043,
+      "learning_rate": 0.00016423012985254638,
+      "loss": 3.1049,
+      "step": 2852
+    },
+    {
+      "epoch": 0.2582601611297185,
+      "grad_norm": 0.9017558693885803,
+      "learning_rate": 0.00016420580459130965,
+      "loss": 3.055,
+      "step": 2853
+    },
+    {
+      "epoch": 0.25835068344346884,
+      "grad_norm": 0.9594643115997314,
+      "learning_rate": 0.0001641814728645502,
+      "loss": 3.0854,
+      "step": 2854
+    },
+    {
+      "epoch": 0.25844120575721913,
+      "grad_norm": 0.8878849148750305,
+      "learning_rate": 0.00016415713467471816,
+      "loss": 3.0405,
+      "step": 2855
+    },
+    {
+      "epoch": 0.25853172807096947,
+      "grad_norm": 0.8501585721969604,
+      "learning_rate": 0.0001641327900242644,
+      "loss": 3.0943,
+      "step": 2856
+    },
+    {
+      "epoch": 0.2586222503847198,
+      "grad_norm": 0.8996968865394592,
+      "learning_rate": 0.00016410843891564048,
+      "loss": 3.0927,
+      "step": 2857
+    },
+    {
+      "epoch": 0.25871277269847015,
+      "grad_norm": 0.9761808514595032,
+      "learning_rate": 0.0001640840813512985,
+      "loss": 3.1048,
+      "step": 2858
+    },
+    {
+      "epoch": 0.2588032950122205,
+      "grad_norm": 0.8404366970062256,
+      "learning_rate": 0.00016405971733369126,
+      "loss": 3.0006,
+      "step": 2859
+    },
+    {
+      "epoch": 0.25889381732597083,
+      "grad_norm": 0.8861578702926636,
+      "learning_rate": 0.00016403534686527225,
+      "loss": 3.0375,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2589843396397212,
+      "grad_norm": 0.8492283225059509,
+      "learning_rate": 0.00016401096994849557,
+      "loss": 3.0891,
+      "step": 2861
+    },
+    {
+      "epoch": 0.2590748619534715,
+      "grad_norm": 0.9136780500411987,
+      "learning_rate": 0.00016398658658581594,
+      "loss": 3.0869,
+      "step": 2862
+    },
+    {
+      "epoch": 0.25916538426722185,
+      "grad_norm": 0.903567910194397,
+      "learning_rate": 0.00016396219677968882,
+      "loss": 3.0415,
+      "step": 2863
+    },
+    {
+      "epoch": 0.2592559065809722,
+      "grad_norm": 0.8297886848449707,
+      "learning_rate": 0.0001639378005325702,
+      "loss": 3.0239,
+      "step": 2864
+    },
+    {
+      "epoch": 0.25934642889472254,
+      "grad_norm": 0.8206044435501099,
+      "learning_rate": 0.00016391339784691686,
+      "loss": 3.0838,
+      "step": 2865
+    },
+    {
+      "epoch": 0.2594369512084729,
+      "grad_norm": 0.8900150656700134,
+      "learning_rate": 0.0001638889887251861,
+      "loss": 3.0946,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2595274735222232,
+      "grad_norm": 0.8471210598945618,
+      "learning_rate": 0.0001638645731698359,
+      "loss": 2.9901,
+      "step": 2867
+    },
+    {
+      "epoch": 0.25961799583597356,
+      "grad_norm": 0.8738295435905457,
+      "learning_rate": 0.00016384015118332495,
+      "loss": 3.0481,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2597085181497239,
+      "grad_norm": 0.805209219455719,
+      "learning_rate": 0.00016381572276811252,
+      "loss": 3.0732,
+      "step": 2869
+    },
+    {
+      "epoch": 0.25979904046347424,
+      "grad_norm": 0.8056511282920837,
+      "learning_rate": 0.00016379128792665855,
+      "loss": 3.0464,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2598895627772246,
+      "grad_norm": 0.8855249881744385,
+      "learning_rate": 0.00016376684666142362,
+      "loss": 3.0513,
+      "step": 2871
+    },
+    {
+      "epoch": 0.2599800850909749,
+      "grad_norm": 0.8531777858734131,
+      "learning_rate": 0.000163742398974869,
+      "loss": 3.0807,
+      "step": 2872
+    },
+    {
+      "epoch": 0.26007060740472526,
+      "grad_norm": 0.8292708396911621,
+      "learning_rate": 0.0001637179448694565,
+      "loss": 3.1046,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2601611297184756,
+      "grad_norm": 0.8251423239707947,
+      "learning_rate": 0.00016369348434764876,
+      "loss": 3.0695,
+      "step": 2874
+    },
+    {
+      "epoch": 0.26025165203222594,
+      "grad_norm": 0.938856303691864,
+      "learning_rate": 0.00016366901741190882,
+      "loss": 3.0725,
+      "step": 2875
+    },
+    {
+      "epoch": 0.2603421743459763,
+      "grad_norm": 0.9786975979804993,
+      "learning_rate": 0.00016364454406470063,
+      "loss": 3.0873,
+      "step": 2876
+    },
+    {
+      "epoch": 0.2604326966597266,
+      "grad_norm": 0.8554605841636658,
+      "learning_rate": 0.00016362006430848856,
+      "loss": 3.1021,
+      "step": 2877
+    },
+    {
+      "epoch": 0.26052321897347697,
+      "grad_norm": 1.1426806449890137,
+      "learning_rate": 0.00016359557814573777,
+      "loss": 3.0669,
+      "step": 2878
+    },
+    {
+      "epoch": 0.2606137412872273,
+      "grad_norm": 0.8070796728134155,
+      "learning_rate": 0.00016357108557891401,
+      "loss": 3.1003,
+      "step": 2879
+    },
+    {
+      "epoch": 0.26070426360097765,
+      "grad_norm": 1.416195034980774,
+      "learning_rate": 0.00016354658661048364,
+      "loss": 3.0877,
+      "step": 2880
+    },
+    {
+      "epoch": 0.260794785914728,
+      "grad_norm": 0.86333167552948,
+      "learning_rate": 0.00016352208124291375,
+      "loss": 3.0707,
+      "step": 2881
+    },
+    {
+      "epoch": 0.26088530822847833,
+      "grad_norm": 0.8207772374153137,
+      "learning_rate": 0.000163497569478672,
+      "loss": 3.09,
+      "step": 2882
+    },
+    {
+      "epoch": 0.26097583054222867,
+      "grad_norm": 0.9560832381248474,
+      "learning_rate": 0.00016347305132022677,
+      "loss": 3.0405,
+      "step": 2883
+    },
+    {
+      "epoch": 0.261066352855979,
+      "grad_norm": 0.9158850312232971,
+      "learning_rate": 0.000163448526770047,
+      "loss": 3.1062,
+      "step": 2884
+    },
+    {
+      "epoch": 0.26115687516972935,
+      "grad_norm": 0.9456655383110046,
+      "learning_rate": 0.00016342399583060233,
+      "loss": 3.0498,
+      "step": 2885
+    },
+    {
+      "epoch": 0.2612473974834797,
+      "grad_norm": 0.9813029766082764,
+      "learning_rate": 0.000163399458504363,
+      "loss": 3.1253,
+      "step": 2886
+    },
+    {
+      "epoch": 0.26133791979723003,
+      "grad_norm": 0.9460070133209229,
+      "learning_rate": 0.00016337491479379994,
+      "loss": 3.1175,
+      "step": 2887
+    },
+    {
+      "epoch": 0.2614284421109804,
+      "grad_norm": 0.9098150730133057,
+      "learning_rate": 0.0001633503647013847,
+      "loss": 3.0655,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2615189644247307,
+      "grad_norm": 0.9258887767791748,
+      "learning_rate": 0.00016332580822958946,
+      "loss": 3.0744,
+      "step": 2889
+    },
+    {
+      "epoch": 0.26160948673848106,
+      "grad_norm": 0.8709380626678467,
+      "learning_rate": 0.00016330124538088705,
+      "loss": 3.0591,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2617000090522314,
+      "grad_norm": 0.8492916226387024,
+      "learning_rate": 0.00016327667615775098,
+      "loss": 3.0236,
+      "step": 2891
+    },
+    {
+      "epoch": 0.26179053136598174,
+      "grad_norm": 0.9202527403831482,
+      "learning_rate": 0.00016325210056265536,
+      "loss": 2.9596,
+      "step": 2892
+    },
+    {
+      "epoch": 0.2618810536797321,
+      "grad_norm": 0.8801531195640564,
+      "learning_rate": 0.0001632275185980749,
+      "loss": 3.0367,
+      "step": 2893
+    },
+    {
+      "epoch": 0.2619715759934824,
+      "grad_norm": 0.9154375195503235,
+      "learning_rate": 0.0001632029302664851,
+      "loss": 3.0255,
+      "step": 2894
+    },
+    {
+      "epoch": 0.26206209830723276,
+      "grad_norm": 0.9271922707557678,
+      "learning_rate": 0.00016317833557036193,
+      "loss": 3.033,
+      "step": 2895
+    },
+    {
+      "epoch": 0.26215262062098305,
+      "grad_norm": 1.0529191493988037,
+      "learning_rate": 0.0001631537345121821,
+      "loss": 3.0615,
+      "step": 2896
+    },
+    {
+      "epoch": 0.2622431429347334,
+      "grad_norm": 1.1552090644836426,
+      "learning_rate": 0.00016312912709442293,
+      "loss": 3.0434,
+      "step": 2897
+    },
+    {
+      "epoch": 0.2623336652484837,
+      "grad_norm": 1.091846227645874,
+      "learning_rate": 0.00016310451331956238,
+      "loss": 2.9952,
+      "step": 2898
+    },
+    {
+      "epoch": 0.26242418756223407,
+      "grad_norm": 1.0549595355987549,
+      "learning_rate": 0.00016307989319007908,
+      "loss": 3.0755,
+      "step": 2899
+    },
+    {
+      "epoch": 0.2625147098759844,
+      "grad_norm": 1.1350163221359253,
+      "learning_rate": 0.00016305526670845226,
+      "loss": 3.071,
+      "step": 2900
+    },
+    {
+      "epoch": 0.26260523218973475,
+      "grad_norm": 0.9817744493484497,
+      "learning_rate": 0.0001630306338771618,
+      "loss": 3.0365,
+      "step": 2901
+    },
+    {
+      "epoch": 0.2626957545034851,
+      "grad_norm": 1.173923134803772,
+      "learning_rate": 0.00016300599469868825,
+      "loss": 3.0985,
+      "step": 2902
+    },
+    {
+      "epoch": 0.26278627681723543,
+      "grad_norm": 0.8843302130699158,
+      "learning_rate": 0.00016298134917551273,
+      "loss": 2.9988,
+      "step": 2903
+    },
+    {
+      "epoch": 0.26287679913098577,
+      "grad_norm": 0.9259403347969055,
+      "learning_rate": 0.00016295669731011708,
+      "loss": 3.0112,
+      "step": 2904
+    },
+    {
+      "epoch": 0.2629673214447361,
+      "grad_norm": 0.9056614637374878,
+      "learning_rate": 0.00016293203910498376,
+      "loss": 3.0118,
+      "step": 2905
+    },
+    {
+      "epoch": 0.26305784375848645,
+      "grad_norm": 0.919861912727356,
+      "learning_rate": 0.0001629073745625958,
+      "loss": 3.0115,
+      "step": 2906
+    },
+    {
+      "epoch": 0.2631483660722368,
+      "grad_norm": 0.8886492848396301,
+      "learning_rate": 0.00016288270368543698,
+      "loss": 3.075,
+      "step": 2907
+    },
+    {
+      "epoch": 0.26323888838598714,
+      "grad_norm": 0.842409610748291,
+      "learning_rate": 0.00016285802647599156,
+      "loss": 3.08,
+      "step": 2908
+    },
+    {
+      "epoch": 0.2633294106997375,
+      "grad_norm": 0.838261067867279,
+      "learning_rate": 0.00016283334293674463,
+      "loss": 3.0069,
+      "step": 2909
+    },
+    {
+      "epoch": 0.2634199330134878,
+      "grad_norm": 0.8960436582565308,
+      "learning_rate": 0.00016280865307018177,
+      "loss": 3.0596,
+      "step": 2910
+    },
+    {
+      "epoch": 0.26351045532723816,
+      "grad_norm": 0.9022462964057922,
+      "learning_rate": 0.00016278395687878926,
+      "loss": 3.0169,
+      "step": 2911
+    },
+    {
+      "epoch": 0.2636009776409885,
+      "grad_norm": 1.000325322151184,
+      "learning_rate": 0.000162759254365054,
+      "loss": 3.051,
+      "step": 2912
+    },
+    {
+      "epoch": 0.26369149995473884,
+      "grad_norm": 0.8697109222412109,
+      "learning_rate": 0.00016273454553146358,
+      "loss": 3.046,
+      "step": 2913
+    },
+    {
+      "epoch": 0.2637820222684892,
+      "grad_norm": 0.9372190833091736,
+      "learning_rate": 0.00016270983038050614,
+      "loss": 2.9794,
+      "step": 2914
+    },
+    {
+      "epoch": 0.2638725445822395,
+      "grad_norm": 0.8693639039993286,
+      "learning_rate": 0.00016268510891467045,
+      "loss": 3.0849,
+      "step": 2915
+    },
+    {
+      "epoch": 0.26396306689598986,
+      "grad_norm": 0.8379819393157959,
+      "learning_rate": 0.00016266038113644607,
+      "loss": 3.0528,
+      "step": 2916
+    },
+    {
+      "epoch": 0.2640535892097402,
+      "grad_norm": 0.9772222638130188,
+      "learning_rate": 0.00016263564704832297,
+      "loss": 3.0461,
+      "step": 2917
+    },
+    {
+      "epoch": 0.26414411152349054,
+      "grad_norm": 0.941190242767334,
+      "learning_rate": 0.00016261090665279198,
+      "loss": 3.0726,
+      "step": 2918
+    },
+    {
+      "epoch": 0.2642346338372409,
+      "grad_norm": 0.8939282894134521,
+      "learning_rate": 0.00016258615995234433,
+      "loss": 3.0199,
+      "step": 2919
+    },
+    {
+      "epoch": 0.2643251561509912,
+      "grad_norm": 0.8429410457611084,
+      "learning_rate": 0.00016256140694947217,
+      "loss": 3.0208,
+      "step": 2920
+    },
+    {
+      "epoch": 0.26441567846474157,
+      "grad_norm": 0.8292998671531677,
+      "learning_rate": 0.00016253664764666797,
+      "loss": 3.0057,
+      "step": 2921
+    },
+    {
+      "epoch": 0.2645062007784919,
+      "grad_norm": 0.8581786751747131,
+      "learning_rate": 0.0001625118820464251,
+      "loss": 3.11,
+      "step": 2922
+    },
+    {
+      "epoch": 0.26459672309224225,
+      "grad_norm": 0.8744654059410095,
+      "learning_rate": 0.00016248711015123742,
+      "loss": 3.0704,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2646872454059926,
+      "grad_norm": 0.9087191820144653,
+      "learning_rate": 0.00016246233196359942,
+      "loss": 3.0857,
+      "step": 2924
+    },
+    {
+      "epoch": 0.26477776771974293,
+      "grad_norm": 0.8340759873390198,
+      "learning_rate": 0.00016243754748600635,
+      "loss": 3.0451,
+      "step": 2925
+    },
+    {
+      "epoch": 0.26486829003349327,
+      "grad_norm": 0.8676270246505737,
+      "learning_rate": 0.00016241275672095395,
+      "loss": 2.9594,
+      "step": 2926
+    },
+    {
+      "epoch": 0.2649588123472436,
+      "grad_norm": 0.9136918187141418,
+      "learning_rate": 0.00016238795967093864,
+      "loss": 3.0149,
+      "step": 2927
+    },
+    {
+      "epoch": 0.26504933466099395,
+      "grad_norm": 0.9369771480560303,
+      "learning_rate": 0.00016236315633845748,
+      "loss": 3.131,
+      "step": 2928
+    },
+    {
+      "epoch": 0.2651398569747443,
+      "grad_norm": 0.8263919353485107,
+      "learning_rate": 0.00016233834672600822,
+      "loss": 2.9978,
+      "step": 2929
+    },
+    {
+      "epoch": 0.26523037928849463,
+      "grad_norm": 0.8877750635147095,
+      "learning_rate": 0.00016231353083608912,
+      "loss": 3.0539,
+      "step": 2930
+    },
+    {
+      "epoch": 0.265320901602245,
+      "grad_norm": 0.860744059085846,
+      "learning_rate": 0.00016228870867119919,
+      "loss": 3.0041,
+      "step": 2931
+    },
+    {
+      "epoch": 0.2654114239159953,
+      "grad_norm": 0.9063569903373718,
+      "learning_rate": 0.000162263880233838,
+      "loss": 3.0642,
+      "step": 2932
+    },
+    {
+      "epoch": 0.26550194622974566,
+      "grad_norm": 0.8420007228851318,
+      "learning_rate": 0.00016223904552650575,
+      "loss": 3.0962,
+      "step": 2933
+    },
+    {
+      "epoch": 0.265592468543496,
+      "grad_norm": 0.939404308795929,
+      "learning_rate": 0.00016221420455170336,
+      "loss": 3.0918,
+      "step": 2934
+    },
+    {
+      "epoch": 0.26568299085724634,
+      "grad_norm": 0.9313684105873108,
+      "learning_rate": 0.00016218935731193224,
+      "loss": 3.1075,
+      "step": 2935
+    },
+    {
+      "epoch": 0.2657735131709967,
+      "grad_norm": 0.8911711573600769,
+      "learning_rate": 0.00016216450380969455,
+      "loss": 3.0071,
+      "step": 2936
+    },
+    {
+      "epoch": 0.26586403548474696,
+      "grad_norm": 0.8708754181861877,
+      "learning_rate": 0.00016213964404749302,
+      "loss": 3.0248,
+      "step": 2937
+    },
+    {
+      "epoch": 0.2659545577984973,
+      "grad_norm": 0.8700071573257446,
+      "learning_rate": 0.00016211477802783103,
+      "loss": 3.0178,
+      "step": 2938
+    },
+    {
+      "epoch": 0.26604508011224764,
+      "grad_norm": 0.8768590688705444,
+      "learning_rate": 0.00016208990575321264,
+      "loss": 3.0197,
+      "step": 2939
+    },
+    {
+      "epoch": 0.266135602425998,
+      "grad_norm": 0.9197354912757874,
+      "learning_rate": 0.00016206502722614238,
+      "loss": 2.9952,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2662261247397483,
+      "grad_norm": 0.8466536998748779,
+      "learning_rate": 0.00016204014244912562,
+      "loss": 2.9759,
+      "step": 2941
+    },
+    {
+      "epoch": 0.26631664705349867,
+      "grad_norm": 0.9619039297103882,
+      "learning_rate": 0.00016201525142466817,
+      "loss": 3.0296,
+      "step": 2942
+    },
+    {
+      "epoch": 0.266407169367249,
+      "grad_norm": 0.9434160590171814,
+      "learning_rate": 0.00016199035415527662,
+      "loss": 3.0937,
+      "step": 2943
+    },
+    {
+      "epoch": 0.26649769168099935,
+      "grad_norm": 0.861297070980072,
+      "learning_rate": 0.00016196545064345812,
+      "loss": 3.0342,
+      "step": 2944
+    },
+    {
+      "epoch": 0.2665882139947497,
+      "grad_norm": 1.038414478302002,
+      "learning_rate": 0.00016194054089172042,
+      "loss": 3.0606,
+      "step": 2945
+    },
+    {
+      "epoch": 0.26667873630850003,
+      "grad_norm": 0.9113991856575012,
+      "learning_rate": 0.00016191562490257196,
+      "loss": 3.1049,
+      "step": 2946
+    },
+    {
+      "epoch": 0.26676925862225037,
+      "grad_norm": 0.9192466735839844,
+      "learning_rate": 0.00016189070267852174,
+      "loss": 3.0546,
+      "step": 2947
+    },
+    {
+      "epoch": 0.2668597809360007,
+      "grad_norm": 1.0234183073043823,
+      "learning_rate": 0.00016186577422207946,
+      "loss": 3.0497,
+      "step": 2948
+    },
+    {
+      "epoch": 0.26695030324975105,
+      "grad_norm": 0.896352231502533,
+      "learning_rate": 0.0001618408395357554,
+      "loss": 3.0133,
+      "step": 2949
+    },
+    {
+      "epoch": 0.2670408255635014,
+      "grad_norm": 1.0323007106781006,
+      "learning_rate": 0.00016181589862206052,
+      "loss": 3.1426,
+      "step": 2950
+    },
+    {
+      "epoch": 0.26713134787725173,
+      "grad_norm": 0.8639165163040161,
+      "learning_rate": 0.00016179095148350633,
+      "loss": 3.0012,
+      "step": 2951
+    },
+    {
+      "epoch": 0.2672218701910021,
+      "grad_norm": 0.9051254987716675,
+      "learning_rate": 0.00016176599812260501,
+      "loss": 3.0118,
+      "step": 2952
+    },
+    {
+      "epoch": 0.2673123925047524,
+      "grad_norm": 0.844688892364502,
+      "learning_rate": 0.0001617410385418694,
+      "loss": 2.9979,
+      "step": 2953
+    },
+    {
+      "epoch": 0.26740291481850276,
+      "grad_norm": 0.9484454989433289,
+      "learning_rate": 0.00016171607274381285,
+      "loss": 3.0809,
+      "step": 2954
+    },
+    {
+      "epoch": 0.2674934371322531,
+      "grad_norm": 0.8286004662513733,
+      "learning_rate": 0.00016169110073094948,
+      "loss": 3.0456,
+      "step": 2955
+    },
+    {
+      "epoch": 0.26758395944600344,
+      "grad_norm": 0.8905054926872253,
+      "learning_rate": 0.00016166612250579395,
+      "loss": 3.1231,
+      "step": 2956
+    },
+    {
+      "epoch": 0.2676744817597538,
+      "grad_norm": 0.9593135118484497,
+      "learning_rate": 0.0001616411380708616,
+      "loss": 3.0714,
+      "step": 2957
+    },
+    {
+      "epoch": 0.2677650040735041,
+      "grad_norm": 0.9497634172439575,
+      "learning_rate": 0.00016161614742866832,
+      "loss": 3.0306,
+      "step": 2958
+    },
+    {
+      "epoch": 0.26785552638725446,
+      "grad_norm": 1.0392967462539673,
+      "learning_rate": 0.00016159115058173064,
+      "loss": 2.9926,
+      "step": 2959
+    },
+    {
+      "epoch": 0.2679460487010048,
+      "grad_norm": 0.874346137046814,
+      "learning_rate": 0.0001615661475325658,
+      "loss": 3.0302,
+      "step": 2960
+    },
+    {
+      "epoch": 0.26803657101475514,
+      "grad_norm": 1.0402164459228516,
+      "learning_rate": 0.00016154113828369164,
+      "loss": 3.0318,
+      "step": 2961
+    },
+    {
+      "epoch": 0.2681270933285055,
+      "grad_norm": 0.9201018810272217,
+      "learning_rate": 0.00016151612283762652,
+      "loss": 3.1107,
+      "step": 2962
+    },
+    {
+      "epoch": 0.2682176156422558,
+      "grad_norm": 0.9608359932899475,
+      "learning_rate": 0.00016149110119688948,
+      "loss": 3.0355,
+      "step": 2963
+    },
+    {
+      "epoch": 0.26830813795600617,
+      "grad_norm": 0.9042795896530151,
+      "learning_rate": 0.00016146607336400023,
+      "loss": 3.0887,
+      "step": 2964
+    },
+    {
+      "epoch": 0.2683986602697565,
+      "grad_norm": 0.9694705009460449,
+      "learning_rate": 0.0001614410393414791,
+      "loss": 2.9801,
+      "step": 2965
+    },
+    {
+      "epoch": 0.26848918258350685,
+      "grad_norm": 0.831889271736145,
+      "learning_rate": 0.00016141599913184699,
+      "loss": 3.0268,
+      "step": 2966
+    },
+    {
+      "epoch": 0.2685797048972572,
+      "grad_norm": 1.1728935241699219,
+      "learning_rate": 0.00016139095273762544,
+      "loss": 3.1047,
+      "step": 2967
+    },
+    {
+      "epoch": 0.26867022721100753,
+      "grad_norm": 0.8189383745193481,
+      "learning_rate": 0.00016136590016133662,
+      "loss": 3.0057,
+      "step": 2968
+    },
+    {
+      "epoch": 0.26876074952475787,
+      "grad_norm": 1.0662305355072021,
+      "learning_rate": 0.00016134084140550335,
+      "loss": 3.0184,
+      "step": 2969
+    },
+    {
+      "epoch": 0.2688512718385082,
+      "grad_norm": 0.9276952743530273,
+      "learning_rate": 0.00016131577647264902,
+      "loss": 3.0401,
+      "step": 2970
+    },
+    {
+      "epoch": 0.26894179415225855,
+      "grad_norm": 0.9908543825149536,
+      "learning_rate": 0.00016129070536529766,
+      "loss": 3.0723,
+      "step": 2971
+    },
+    {
+      "epoch": 0.2690323164660089,
+      "grad_norm": 1.025470495223999,
+      "learning_rate": 0.00016126562808597397,
+      "loss": 3.0269,
+      "step": 2972
+    },
+    {
+      "epoch": 0.26912283877975923,
+      "grad_norm": 0.9223266243934631,
+      "learning_rate": 0.0001612405446372032,
+      "loss": 3.1187,
+      "step": 2973
+    },
+    {
+      "epoch": 0.2692133610935096,
+      "grad_norm": 0.924764096736908,
+      "learning_rate": 0.00016121545502151125,
+      "loss": 3.0623,
+      "step": 2974
+    },
+    {
+      "epoch": 0.2693038834072599,
+      "grad_norm": 0.8277813792228699,
+      "learning_rate": 0.00016119035924142466,
+      "loss": 3.027,
+      "step": 2975
+    },
+    {
+      "epoch": 0.26939440572101025,
+      "grad_norm": 0.8711262345314026,
+      "learning_rate": 0.00016116525729947058,
+      "loss": 3.039,
+      "step": 2976
+    },
+    {
+      "epoch": 0.2694849280347606,
+      "grad_norm": 0.9595316052436829,
+      "learning_rate": 0.00016114014919817678,
+      "loss": 3.0453,
+      "step": 2977
+    },
+    {
+      "epoch": 0.2695754503485109,
+      "grad_norm": 0.9323618412017822,
+      "learning_rate": 0.0001611150349400716,
+      "loss": 3.0467,
+      "step": 2978
+    },
+    {
+      "epoch": 0.2696659726622612,
+      "grad_norm": 0.9700012803077698,
+      "learning_rate": 0.00016108991452768408,
+      "loss": 3.027,
+      "step": 2979
+    },
+    {
+      "epoch": 0.26975649497601156,
+      "grad_norm": 1.0660173892974854,
+      "learning_rate": 0.00016106478796354382,
+      "loss": 3.1133,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2698470172897619,
+      "grad_norm": 0.8464480638504028,
+      "learning_rate": 0.00016103965525018113,
+      "loss": 3.0249,
+      "step": 2981
+    },
+    {
+      "epoch": 0.26993753960351224,
+      "grad_norm": 1.045445203781128,
+      "learning_rate": 0.0001610145163901268,
+      "loss": 3.0867,
+      "step": 2982
+    },
+    {
+      "epoch": 0.2700280619172626,
+      "grad_norm": 0.9046902656555176,
+      "learning_rate": 0.0001609893713859123,
+      "loss": 3.0284,
+      "step": 2983
+    },
+    {
+      "epoch": 0.2701185842310129,
+      "grad_norm": 0.8621628880500793,
+      "learning_rate": 0.00016096422024006982,
+      "loss": 3.0301,
+      "step": 2984
+    },
+    {
+      "epoch": 0.27020910654476327,
+      "grad_norm": 1.079614520072937,
+      "learning_rate": 0.00016093906295513202,
+      "loss": 3.0707,
+      "step": 2985
+    },
+    {
+      "epoch": 0.2702996288585136,
+      "grad_norm": 0.805829644203186,
+      "learning_rate": 0.00016091389953363226,
+      "loss": 3.0931,
+      "step": 2986
+    },
+    {
+      "epoch": 0.27039015117226395,
+      "grad_norm": 0.9469945430755615,
+      "learning_rate": 0.00016088872997810445,
+      "loss": 3.121,
+      "step": 2987
+    },
+    {
+      "epoch": 0.2704806734860143,
+      "grad_norm": 0.8542191386222839,
+      "learning_rate": 0.0001608635542910832,
+      "loss": 3.0177,
+      "step": 2988
+    },
+    {
+      "epoch": 0.27057119579976463,
+      "grad_norm": 0.8838629126548767,
+      "learning_rate": 0.00016083837247510375,
+      "loss": 3.0072,
+      "step": 2989
+    },
+    {
+      "epoch": 0.27066171811351497,
+      "grad_norm": 1.1052781343460083,
+      "learning_rate": 0.0001608131845327018,
+      "loss": 3.0693,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2707522404272653,
+      "grad_norm": 0.8787950277328491,
+      "learning_rate": 0.00016078799046641385,
+      "loss": 3.0323,
+      "step": 2991
+    },
+    {
+      "epoch": 0.27084276274101565,
+      "grad_norm": 0.9619665741920471,
+      "learning_rate": 0.0001607627902787769,
+      "loss": 3.019,
+      "step": 2992
+    },
+    {
+      "epoch": 0.270933285054766,
+      "grad_norm": 0.943647027015686,
+      "learning_rate": 0.00016073758397232868,
+      "loss": 3.0532,
+      "step": 2993
+    },
+    {
+      "epoch": 0.27102380736851633,
+      "grad_norm": 0.9765943884849548,
+      "learning_rate": 0.00016071237154960738,
+      "loss": 3.0638,
+      "step": 2994
+    },
+    {
+      "epoch": 0.2711143296822667,
+      "grad_norm": 0.8682699799537659,
+      "learning_rate": 0.00016068715301315197,
+      "loss": 3.0934,
+      "step": 2995
+    },
+    {
+      "epoch": 0.271204851996017,
+      "grad_norm": 0.9122071862220764,
+      "learning_rate": 0.00016066192836550184,
+      "loss": 3.0495,
+      "step": 2996
+    },
+    {
+      "epoch": 0.27129537430976736,
+      "grad_norm": 0.935530960559845,
+      "learning_rate": 0.00016063669760919727,
+      "loss": 3.0794,
+      "step": 2997
+    },
+    {
+      "epoch": 0.2713858966235177,
+      "grad_norm": 0.9900686740875244,
+      "learning_rate": 0.00016061146074677885,
+      "loss": 3.0998,
+      "step": 2998
+    },
+    {
+      "epoch": 0.27147641893726804,
+      "grad_norm": 0.9189563393592834,
+      "learning_rate": 0.000160586217780788,
+      "loss": 3.0788,
+      "step": 2999
+    },
+    {
+      "epoch": 0.2715669412510184,
+      "grad_norm": 0.8959607481956482,
+      "learning_rate": 0.00016056096871376667,
+      "loss": 2.97,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2715669412510184,
+      "eval_loss": 2.978346109390259,
+      "eval_runtime": 71.5356,
+      "eval_samples_per_second": 37.785,
+      "eval_steps_per_second": 3.159,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2716574635647687,
+      "grad_norm": 0.8900837302207947,
+      "learning_rate": 0.00016053571354825746,
+      "loss": 2.992,
+      "step": 3001
+    },
+    {
+      "epoch": 0.27174798587851906,
+      "grad_norm": 0.8718365430831909,
+      "learning_rate": 0.00016051045228680354,
+      "loss": 3.0593,
+      "step": 3002
+    },
+    {
+      "epoch": 0.2718385081922694,
+      "grad_norm": 0.9835093021392822,
+      "learning_rate": 0.00016048518493194878,
+      "loss": 3.0406,
+      "step": 3003
+    },
+    {
+      "epoch": 0.27192903050601974,
+      "grad_norm": 0.8958598971366882,
+      "learning_rate": 0.0001604599114862375,
+      "loss": 3.0482,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2720195528197701,
+      "grad_norm": 0.9504435658454895,
+      "learning_rate": 0.0001604346319522148,
+      "loss": 3.0606,
+      "step": 3005
+    },
+    {
+      "epoch": 0.2721100751335204,
+      "grad_norm": 0.911404013633728,
+      "learning_rate": 0.00016040934633242634,
+      "loss": 3.1059,
+      "step": 3006
+    },
+    {
+      "epoch": 0.27220059744727076,
+      "grad_norm": 0.8670641183853149,
+      "learning_rate": 0.00016038405462941833,
+      "loss": 3.0696,
+      "step": 3007
+    },
+    {
+      "epoch": 0.2722911197610211,
+      "grad_norm": 0.8918884992599487,
+      "learning_rate": 0.00016035875684573765,
+      "loss": 3.0457,
+      "step": 3008
+    },
+    {
+      "epoch": 0.27238164207477145,
+      "grad_norm": 1.0598082542419434,
+      "learning_rate": 0.00016033345298393186,
+      "loss": 3.03,
+      "step": 3009
+    },
+    {
+      "epoch": 0.2724721643885218,
+      "grad_norm": 0.8834877014160156,
+      "learning_rate": 0.00016030814304654895,
+      "loss": 2.9908,
+      "step": 3010
+    },
+    {
+      "epoch": 0.27256268670227213,
+      "grad_norm": 0.9708331823348999,
+      "learning_rate": 0.00016028282703613772,
+      "loss": 3.0751,
+      "step": 3011
+    },
+    {
+      "epoch": 0.27265320901602247,
+      "grad_norm": 0.9399593472480774,
+      "learning_rate": 0.00016025750495524745,
+      "loss": 3.0301,
+      "step": 3012
+    },
+    {
+      "epoch": 0.2727437313297728,
+      "grad_norm": 0.8679122924804688,
+      "learning_rate": 0.0001602321768064281,
+      "loss": 3.0653,
+      "step": 3013
+    },
+    {
+      "epoch": 0.27283425364352315,
+      "grad_norm": 0.9808764457702637,
+      "learning_rate": 0.00016020684259223015,
+      "loss": 3.0514,
+      "step": 3014
+    },
+    {
+      "epoch": 0.2729247759572735,
+      "grad_norm": 0.8745496273040771,
+      "learning_rate": 0.00016018150231520486,
+      "loss": 3.0555,
+      "step": 3015
+    },
+    {
+      "epoch": 0.27301529827102383,
+      "grad_norm": 1.0140714645385742,
+      "learning_rate": 0.00016015615597790388,
+      "loss": 3.0285,
+      "step": 3016
+    },
+    {
+      "epoch": 0.2731058205847742,
+      "grad_norm": 0.9872329831123352,
+      "learning_rate": 0.00016013080358287963,
+      "loss": 3.0468,
+      "step": 3017
+    },
+    {
+      "epoch": 0.2731963428985245,
+      "grad_norm": 0.8958715200424194,
+      "learning_rate": 0.00016010544513268515,
+      "loss": 3.0561,
+      "step": 3018
+    },
+    {
+      "epoch": 0.2732868652122748,
+      "grad_norm": 0.9180740118026733,
+      "learning_rate": 0.00016008008062987398,
+      "loss": 3.0743,
+      "step": 3019
+    },
+    {
+      "epoch": 0.27337738752602514,
+      "grad_norm": 0.8599140048027039,
+      "learning_rate": 0.00016005471007700031,
+      "loss": 3.0579,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2734679098397755,
+      "grad_norm": 0.9585762023925781,
+      "learning_rate": 0.00016002933347661902,
+      "loss": 3.0443,
+      "step": 3021
+    },
+    {
+      "epoch": 0.2735584321535258,
+      "grad_norm": 0.8619949221611023,
+      "learning_rate": 0.00016000395083128543,
+      "loss": 3.0021,
+      "step": 3022
+    },
+    {
+      "epoch": 0.27364895446727616,
+      "grad_norm": 0.9174975156784058,
+      "learning_rate": 0.00015997856214355565,
+      "loss": 2.9539,
+      "step": 3023
+    },
+    {
+      "epoch": 0.2737394767810265,
+      "grad_norm": 1.008433222770691,
+      "learning_rate": 0.00015995316741598633,
+      "loss": 3.0251,
+      "step": 3024
+    },
+    {
+      "epoch": 0.27382999909477684,
+      "grad_norm": 0.8638203740119934,
+      "learning_rate": 0.0001599277666511347,
+      "loss": 3.0041,
+      "step": 3025
+    },
+    {
+      "epoch": 0.2739205214085272,
+      "grad_norm": 0.9356272220611572,
+      "learning_rate": 0.0001599023598515586,
+      "loss": 3.0161,
+      "step": 3026
+    },
+    {
+      "epoch": 0.2740110437222775,
+      "grad_norm": 0.875857412815094,
+      "learning_rate": 0.00015987694701981647,
+      "loss": 3.049,
+      "step": 3027
+    },
+    {
+      "epoch": 0.27410156603602787,
+      "grad_norm": 0.9533547759056091,
+      "learning_rate": 0.00015985152815846745,
+      "loss": 3.0531,
+      "step": 3028
+    },
+    {
+      "epoch": 0.2741920883497782,
+      "grad_norm": 0.9074561595916748,
+      "learning_rate": 0.00015982610327007116,
+      "loss": 2.9959,
+      "step": 3029
+    },
+    {
+      "epoch": 0.27428261066352855,
+      "grad_norm": 0.8548593521118164,
+      "learning_rate": 0.00015980067235718792,
+      "loss": 3.0434,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2743731329772789,
+      "grad_norm": 0.8892874121665955,
+      "learning_rate": 0.00015977523542237863,
+      "loss": 2.9971,
+      "step": 3031
+    },
+    {
+      "epoch": 0.27446365529102923,
+      "grad_norm": 0.9240937829017639,
+      "learning_rate": 0.0001597497924682047,
+      "loss": 3.1156,
+      "step": 3032
+    },
+    {
+      "epoch": 0.27455417760477957,
+      "grad_norm": 1.205907940864563,
+      "learning_rate": 0.0001597243434972284,
+      "loss": 2.9952,
+      "step": 3033
+    },
+    {
+      "epoch": 0.2746446999185299,
+      "grad_norm": 0.9608234763145447,
+      "learning_rate": 0.00015969888851201226,
+      "loss": 3.0614,
+      "step": 3034
+    },
+    {
+      "epoch": 0.27473522223228025,
+      "grad_norm": 0.9994444251060486,
+      "learning_rate": 0.0001596734275151197,
+      "loss": 3.0317,
+      "step": 3035
+    },
+    {
+      "epoch": 0.2748257445460306,
+      "grad_norm": 1.1104655265808105,
+      "learning_rate": 0.00015964796050911467,
+      "loss": 3.0698,
+      "step": 3036
+    },
+    {
+      "epoch": 0.27491626685978093,
+      "grad_norm": 1.007348656654358,
+      "learning_rate": 0.0001596224874965616,
+      "loss": 2.9821,
+      "step": 3037
+    },
+    {
+      "epoch": 0.2750067891735313,
+      "grad_norm": 1.0268911123275757,
+      "learning_rate": 0.00015959700848002567,
+      "loss": 3.0238,
+      "step": 3038
+    },
+    {
+      "epoch": 0.2750973114872816,
+      "grad_norm": 1.0308053493499756,
+      "learning_rate": 0.00015957152346207263,
+      "loss": 3.0165,
+      "step": 3039
+    },
+    {
+      "epoch": 0.27518783380103196,
+      "grad_norm": 0.9716582894325256,
+      "learning_rate": 0.0001595460324452688,
+      "loss": 3.0735,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2752783561147823,
+      "grad_norm": 1.2407675981521606,
+      "learning_rate": 0.00015952053543218114,
+      "loss": 3.0039,
+      "step": 3041
+    },
+    {
+      "epoch": 0.27536887842853264,
+      "grad_norm": 0.8391728401184082,
+      "learning_rate": 0.00015949503242537717,
+      "loss": 3.0505,
+      "step": 3042
+    },
+    {
+      "epoch": 0.275459400742283,
+      "grad_norm": 1.2026934623718262,
+      "learning_rate": 0.0001594695234274251,
+      "loss": 3.0462,
+      "step": 3043
+    },
+    {
+      "epoch": 0.2755499230560333,
+      "grad_norm": 1.088293433189392,
+      "learning_rate": 0.00015944400844089364,
+      "loss": 3.0565,
+      "step": 3044
+    },
+    {
+      "epoch": 0.27564044536978366,
+      "grad_norm": 0.9235910177230835,
+      "learning_rate": 0.00015941848746835215,
+      "loss": 3.0061,
+      "step": 3045
+    },
+    {
+      "epoch": 0.275730967683534,
+      "grad_norm": 0.9839001893997192,
+      "learning_rate": 0.0001593929605123706,
+      "loss": 3.0349,
+      "step": 3046
+    },
+    {
+      "epoch": 0.27582148999728434,
+      "grad_norm": 0.8767859935760498,
+      "learning_rate": 0.00015936742757551955,
+      "loss": 3.0455,
+      "step": 3047
+    },
+    {
+      "epoch": 0.2759120123110347,
+      "grad_norm": 0.973345160484314,
+      "learning_rate": 0.00015934188866037016,
+      "loss": 3.065,
+      "step": 3048
+    },
+    {
+      "epoch": 0.276002534624785,
+      "grad_norm": 0.9066287875175476,
+      "learning_rate": 0.0001593163437694942,
+      "loss": 3.0872,
+      "step": 3049
+    },
+    {
+      "epoch": 0.27609305693853536,
+      "grad_norm": 0.9278755784034729,
+      "learning_rate": 0.00015929079290546408,
+      "loss": 2.9691,
+      "step": 3050
+    },
+    {
+      "epoch": 0.2761835792522857,
+      "grad_norm": 0.88511061668396,
+      "learning_rate": 0.00015926523607085269,
+      "loss": 3.0507,
+      "step": 3051
+    },
+    {
+      "epoch": 0.27627410156603605,
+      "grad_norm": 0.9436579346656799,
+      "learning_rate": 0.00015923967326823368,
+      "loss": 2.9823,
+      "step": 3052
+    },
+    {
+      "epoch": 0.2763646238797864,
+      "grad_norm": 0.9194360971450806,
+      "learning_rate": 0.00015921410450018118,
+      "loss": 2.9886,
+      "step": 3053
+    },
+    {
+      "epoch": 0.2764551461935367,
+      "grad_norm": 0.8538716435432434,
+      "learning_rate": 0.00015918852976926998,
+      "loss": 2.9554,
+      "step": 3054
+    },
+    {
+      "epoch": 0.27654566850728707,
+      "grad_norm": 0.9728320837020874,
+      "learning_rate": 0.00015916294907807545,
+      "loss": 3.02,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2766361908210374,
+      "grad_norm": 0.8909928202629089,
+      "learning_rate": 0.00015913736242917356,
+      "loss": 3.0283,
+      "step": 3056
+    },
+    {
+      "epoch": 0.27672671313478775,
+      "grad_norm": 0.8910848498344421,
+      "learning_rate": 0.0001591117698251409,
+      "loss": 3.0063,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2768172354485381,
+      "grad_norm": 0.8911659717559814,
+      "learning_rate": 0.00015908617126855466,
+      "loss": 2.9937,
+      "step": 3058
+    },
+    {
+      "epoch": 0.27690775776228843,
+      "grad_norm": 0.8785828351974487,
+      "learning_rate": 0.00015906056676199255,
+      "loss": 3.0193,
+      "step": 3059
+    },
+    {
+      "epoch": 0.2769982800760387,
+      "grad_norm": 0.9020123481750488,
+      "learning_rate": 0.000159034956308033,
+      "loss": 2.9334,
+      "step": 3060
+    },
+    {
+      "epoch": 0.27708880238978906,
+      "grad_norm": 0.8372553586959839,
+      "learning_rate": 0.00015900933990925498,
+      "loss": 3.0273,
+      "step": 3061
+    },
+    {
+      "epoch": 0.2771793247035394,
+      "grad_norm": 0.9668348431587219,
+      "learning_rate": 0.000158983717568238,
+      "loss": 3.0453,
+      "step": 3062
+    },
+    {
+      "epoch": 0.27726984701728974,
+      "grad_norm": 0.8780279755592346,
+      "learning_rate": 0.0001589580892875623,
+      "loss": 2.9765,
+      "step": 3063
+    },
+    {
+      "epoch": 0.2773603693310401,
+      "grad_norm": 0.8655260801315308,
+      "learning_rate": 0.00015893245506980866,
+      "loss": 3.0367,
+      "step": 3064
+    },
+    {
+      "epoch": 0.2774508916447904,
+      "grad_norm": 0.9587757587432861,
+      "learning_rate": 0.00015890681491755837,
+      "loss": 3.0454,
+      "step": 3065
+    },
+    {
+      "epoch": 0.27754141395854076,
+      "grad_norm": 0.8964729905128479,
+      "learning_rate": 0.0001588811688333934,
+      "loss": 3.0603,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2776319362722911,
+      "grad_norm": 0.8597284555435181,
+      "learning_rate": 0.0001588555168198964,
+      "loss": 3.0304,
+      "step": 3067
+    },
+    {
+      "epoch": 0.27772245858604144,
+      "grad_norm": 0.9043926000595093,
+      "learning_rate": 0.00015882985887965043,
+      "loss": 3.0292,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2778129808997918,
+      "grad_norm": 0.9580448865890503,
+      "learning_rate": 0.00015880419501523927,
+      "loss": 3.1074,
+      "step": 3069
+    },
+    {
+      "epoch": 0.2779035032135421,
+      "grad_norm": 0.9260460138320923,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 3.0452,
+      "step": 3070
+    },
+    {
+      "epoch": 0.27799402552729247,
+      "grad_norm": 1.195544958114624,
+      "learning_rate": 0.0001587528495242595,
+      "loss": 3.0491,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2780845478410428,
+      "grad_norm": 1.0550506114959717,
+      "learning_rate": 0.0001587271679028613,
+      "loss": 2.9788,
+      "step": 3072
+    },
+    {
+      "epoch": 0.27817507015479315,
+      "grad_norm": 1.0724434852600098,
+      "learning_rate": 0.0001587014803676389,
+      "loss": 2.9796,
+      "step": 3073
+    },
+    {
+      "epoch": 0.2782655924685435,
+      "grad_norm": 0.8458510637283325,
+      "learning_rate": 0.000158675786921179,
+      "loss": 3.053,
+      "step": 3074
+    },
+    {
+      "epoch": 0.27835611478229383,
+      "grad_norm": 1.1424552202224731,
+      "learning_rate": 0.00015865008756606904,
+      "loss": 3.0551,
+      "step": 3075
+    },
+    {
+      "epoch": 0.27844663709604417,
+      "grad_norm": 0.9233999252319336,
+      "learning_rate": 0.0001586243823048968,
+      "loss": 2.9895,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2785371594097945,
+      "grad_norm": 0.8394502401351929,
+      "learning_rate": 0.0001585986711402509,
+      "loss": 3.0804,
+      "step": 3077
+    },
+    {
+      "epoch": 0.27862768172354485,
+      "grad_norm": 0.9870647192001343,
+      "learning_rate": 0.00015857295407472046,
+      "loss": 3.0542,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2787182040372952,
+      "grad_norm": 0.8402102589607239,
+      "learning_rate": 0.00015854723111089508,
+      "loss": 3.0197,
+      "step": 3079
+    },
+    {
+      "epoch": 0.27880872635104553,
+      "grad_norm": 0.8542654514312744,
+      "learning_rate": 0.00015852150225136518,
+      "loss": 3.0423,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2788992486647959,
+      "grad_norm": 0.9618365168571472,
+      "learning_rate": 0.00015849576749872157,
+      "loss": 3.0611,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2789897709785462,
+      "grad_norm": 0.8424764275550842,
+      "learning_rate": 0.00015847002685555578,
+      "loss": 3.0326,
+      "step": 3082
+    },
+    {
+      "epoch": 0.27908029329229656,
+      "grad_norm": 0.8958133459091187,
+      "learning_rate": 0.00015844428032445987,
+      "loss": 3.0293,
+      "step": 3083
+    },
+    {
+      "epoch": 0.2791708156060469,
+      "grad_norm": 0.9073579907417297,
+      "learning_rate": 0.00015841852790802652,
+      "loss": 2.996,
+      "step": 3084
+    },
+    {
+      "epoch": 0.27926133791979724,
+      "grad_norm": 0.9596017599105835,
+      "learning_rate": 0.00015839276960884905,
+      "loss": 2.999,
+      "step": 3085
+    },
+    {
+      "epoch": 0.2793518602335476,
+      "grad_norm": 0.985481858253479,
+      "learning_rate": 0.00015836700542952128,
+      "loss": 3.0469,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2794423825472979,
+      "grad_norm": 0.9854786992073059,
+      "learning_rate": 0.00015834123537263763,
+      "loss": 3.0807,
+      "step": 3087
+    },
+    {
+      "epoch": 0.27953290486104826,
+      "grad_norm": 0.9631576538085938,
+      "learning_rate": 0.0001583154594407932,
+      "loss": 2.9977,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2796234271747986,
+      "grad_norm": 1.4733103513717651,
+      "learning_rate": 0.00015828967763658358,
+      "loss": 3.0604,
+      "step": 3089
+    },
+    {
+      "epoch": 0.27971394948854894,
+      "grad_norm": 0.9754955768585205,
+      "learning_rate": 0.00015826388996260503,
+      "loss": 3.0514,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2798044718022993,
+      "grad_norm": 0.918458104133606,
+      "learning_rate": 0.00015823809642145434,
+      "loss": 2.9837,
+      "step": 3091
+    },
+    {
+      "epoch": 0.2798949941160496,
+      "grad_norm": 0.9138178825378418,
+      "learning_rate": 0.00015821229701572896,
+      "loss": 3.0117,
+      "step": 3092
+    },
+    {
+      "epoch": 0.27998551642979996,
+      "grad_norm": 0.9586968421936035,
+      "learning_rate": 0.00015818649174802684,
+      "loss": 3.0233,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2800760387435503,
+      "grad_norm": 0.9427858591079712,
+      "learning_rate": 0.0001581606806209466,
+      "loss": 3.0529,
+      "step": 3094
+    },
+    {
+      "epoch": 0.28016656105730064,
+      "grad_norm": 0.891553521156311,
+      "learning_rate": 0.0001581348636370874,
+      "loss": 3.0154,
+      "step": 3095
+    },
+    {
+      "epoch": 0.280257083371051,
+      "grad_norm": 0.8784413933753967,
+      "learning_rate": 0.00015810904079904904,
+      "loss": 3.0261,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2803476056848013,
+      "grad_norm": 0.9564874172210693,
+      "learning_rate": 0.00015808321210943188,
+      "loss": 3.0236,
+      "step": 3097
+    },
+    {
+      "epoch": 0.28043812799855167,
+      "grad_norm": 0.9087296724319458,
+      "learning_rate": 0.00015805737757083681,
+      "loss": 3.094,
+      "step": 3098
+    },
+    {
+      "epoch": 0.280528650312302,
+      "grad_norm": 0.9095321893692017,
+      "learning_rate": 0.00015803153718586544,
+      "loss": 2.9883,
+      "step": 3099
+    },
+    {
+      "epoch": 0.28061917262605235,
+      "grad_norm": 0.9614866971969604,
+      "learning_rate": 0.00015800569095711982,
+      "loss": 3.003,
+      "step": 3100
+    },
+    {
+      "epoch": 0.28070969493980263,
+      "grad_norm": 0.8452001810073853,
+      "learning_rate": 0.00015797983888720274,
+      "loss": 2.9554,
+      "step": 3101
+    },
+    {
+      "epoch": 0.280800217253553,
+      "grad_norm": 0.8980777263641357,
+      "learning_rate": 0.00015795398097871748,
+      "loss": 3.0503,
+      "step": 3102
+    },
+    {
+      "epoch": 0.2808907395673033,
+      "grad_norm": 0.8974297046661377,
+      "learning_rate": 0.0001579281172342679,
+      "loss": 2.9996,
+      "step": 3103
+    },
+    {
+      "epoch": 0.28098126188105366,
+      "grad_norm": 1.022741675376892,
+      "learning_rate": 0.00015790224765645849,
+      "loss": 2.9941,
+      "step": 3104
+    },
+    {
+      "epoch": 0.281071784194804,
+      "grad_norm": 0.9084989428520203,
+      "learning_rate": 0.00015787637224789434,
+      "loss": 3.0378,
+      "step": 3105
+    },
+    {
+      "epoch": 0.28116230650855434,
+      "grad_norm": 1.0861493349075317,
+      "learning_rate": 0.0001578504910111811,
+      "loss": 3.0639,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2812528288223047,
+      "grad_norm": 0.9200314283370972,
+      "learning_rate": 0.00015782460394892498,
+      "loss": 2.9762,
+      "step": 3107
+    },
+    {
+      "epoch": 0.281343351136055,
+      "grad_norm": 1.0727646350860596,
+      "learning_rate": 0.00015779871106373283,
+      "loss": 3.0473,
+      "step": 3108
+    },
+    {
+      "epoch": 0.28143387344980536,
+      "grad_norm": 0.9366647601127625,
+      "learning_rate": 0.00015777281235821206,
+      "loss": 3.0291,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2815243957635557,
+      "grad_norm": 1.0757999420166016,
+      "learning_rate": 0.00015774690783497067,
+      "loss": 3.0211,
+      "step": 3110
+    },
+    {
+      "epoch": 0.28161491807730604,
+      "grad_norm": 0.8640356659889221,
+      "learning_rate": 0.00015772099749661723,
+      "loss": 3.0259,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2817054403910564,
+      "grad_norm": 0.8835778832435608,
+      "learning_rate": 0.00015769508134576095,
+      "loss": 2.9957,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2817959627048067,
+      "grad_norm": 0.9119789600372314,
+      "learning_rate": 0.00015766915938501152,
+      "loss": 3.0188,
+      "step": 3113
+    },
+    {
+      "epoch": 0.28188648501855706,
+      "grad_norm": 0.8268768787384033,
+      "learning_rate": 0.00015764323161697935,
+      "loss": 2.9249,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2819770073323074,
+      "grad_norm": 0.8696132302284241,
+      "learning_rate": 0.0001576172980442753,
+      "loss": 2.9569,
+      "step": 3115
+    },
+    {
+      "epoch": 0.28206752964605775,
+      "grad_norm": 0.8999311923980713,
+      "learning_rate": 0.00015759135866951092,
+      "loss": 3.024,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2821580519598081,
+      "grad_norm": 0.8366394639015198,
+      "learning_rate": 0.00015756541349529833,
+      "loss": 2.9927,
+      "step": 3117
+    },
+    {
+      "epoch": 0.28224857427355843,
+      "grad_norm": 0.9423478841781616,
+      "learning_rate": 0.00015753946252425013,
+      "loss": 3.0575,
+      "step": 3118
+    },
+    {
+      "epoch": 0.28233909658730877,
+      "grad_norm": 0.8939400911331177,
+      "learning_rate": 0.00015751350575897964,
+      "loss": 3.0443,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2824296189010591,
+      "grad_norm": 0.8804583549499512,
+      "learning_rate": 0.00015748754320210072,
+      "loss": 3.0482,
+      "step": 3120
+    },
+    {
+      "epoch": 0.28252014121480945,
+      "grad_norm": 0.9358223676681519,
+      "learning_rate": 0.00015746157485622777,
+      "loss": 3.0573,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2826106635285598,
+      "grad_norm": 0.8463422060012817,
+      "learning_rate": 0.00015743560072397578,
+      "loss": 3.0755,
+      "step": 3122
+    },
+    {
+      "epoch": 0.28270118584231013,
+      "grad_norm": 0.9778035879135132,
+      "learning_rate": 0.0001574096208079604,
+      "loss": 3.059,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2827917081560605,
+      "grad_norm": 0.8894894123077393,
+      "learning_rate": 0.00015738363511079776,
+      "loss": 3.0353,
+      "step": 3124
+    },
+    {
+      "epoch": 0.2828822304698108,
+      "grad_norm": 0.9795463681221008,
+      "learning_rate": 0.0001573576436351046,
+      "loss": 3.0643,
+      "step": 3125
+    },
+    {
+      "epoch": 0.28297275278356115,
+      "grad_norm": 0.9977430105209351,
+      "learning_rate": 0.00015733164638349835,
+      "loss": 3.0345,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2830632750973115,
+      "grad_norm": 1.0255120992660522,
+      "learning_rate": 0.00015730564335859684,
+      "loss": 3.0646,
+      "step": 3127
+    },
+    {
+      "epoch": 0.28315379741106184,
+      "grad_norm": 0.920333981513977,
+      "learning_rate": 0.0001572796345630186,
+      "loss": 3.0291,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2832443197248122,
+      "grad_norm": 0.8356728553771973,
+      "learning_rate": 0.00015725361999938277,
+      "loss": 3.0165,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2833348420385625,
+      "grad_norm": 0.9003295302391052,
+      "learning_rate": 0.00015722759967030898,
+      "loss": 2.9575,
+      "step": 3130
+    },
+    {
+      "epoch": 0.28342536435231286,
+      "grad_norm": 0.9911743998527527,
+      "learning_rate": 0.00015720157357841744,
+      "loss": 3.0059,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2835158866660632,
+      "grad_norm": 0.8708944320678711,
+      "learning_rate": 0.00015717554172632904,
+      "loss": 2.9358,
+      "step": 3132
+    },
+    {
+      "epoch": 0.28360640897981354,
+      "grad_norm": 0.9072350859642029,
+      "learning_rate": 0.0001571495041166651,
+      "loss": 3.0155,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2836969312935639,
+      "grad_norm": 0.9872345924377441,
+      "learning_rate": 0.00015712346075204766,
+      "loss": 3.0427,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2837874536073142,
+      "grad_norm": 1.0371378660202026,
+      "learning_rate": 0.00015709741163509933,
+      "loss": 3.0016,
+      "step": 3135
+    },
+    {
+      "epoch": 0.28387797592106456,
+      "grad_norm": 1.0638636350631714,
+      "learning_rate": 0.0001570713567684432,
+      "loss": 3.0165,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2839684982348149,
+      "grad_norm": 0.879398763179779,
+      "learning_rate": 0.00015704529615470295,
+      "loss": 3.0285,
+      "step": 3137
+    },
+    {
+      "epoch": 0.28405902054856524,
+      "grad_norm": 1.0027191638946533,
+      "learning_rate": 0.000157019229796503,
+      "loss": 3.019,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2841495428623156,
+      "grad_norm": 0.9104477763175964,
+      "learning_rate": 0.00015699315769646815,
+      "loss": 3.0091,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2842400651760659,
+      "grad_norm": 0.9437276721000671,
+      "learning_rate": 0.0001569670798572239,
+      "loss": 3.0236,
+      "step": 3140
+    },
+    {
+      "epoch": 0.28433058748981627,
+      "grad_norm": 1.2169289588928223,
+      "learning_rate": 0.00015694099628139622,
+      "loss": 3.0437,
+      "step": 3141
+    },
+    {
+      "epoch": 0.28442110980356655,
+      "grad_norm": 0.9517204761505127,
+      "learning_rate": 0.00015691490697161182,
+      "loss": 3.016,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2845116321173169,
+      "grad_norm": 1.0102424621582031,
+      "learning_rate": 0.00015688881193049782,
+      "loss": 2.9728,
+      "step": 3143
+    },
+    {
+      "epoch": 0.28460215443106723,
+      "grad_norm": 0.8501645922660828,
+      "learning_rate": 0.000156862711160682,
+      "loss": 2.9421,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2846926767448176,
+      "grad_norm": 0.870749831199646,
+      "learning_rate": 0.00015683660466479275,
+      "loss": 2.9931,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2847831990585679,
+      "grad_norm": 0.9530490636825562,
+      "learning_rate": 0.000156810492445459,
+      "loss": 3.0254,
+      "step": 3146
+    },
+    {
+      "epoch": 0.28487372137231826,
+      "grad_norm": 0.902142345905304,
+      "learning_rate": 0.00015678437450531013,
+      "loss": 3.0355,
+      "step": 3147
+    },
+    {
+      "epoch": 0.2849642436860686,
+      "grad_norm": 0.9047490954399109,
+      "learning_rate": 0.00015675825084697636,
+      "loss": 3.0706,
+      "step": 3148
+    },
+    {
+      "epoch": 0.28505476599981894,
+      "grad_norm": 0.8961873650550842,
+      "learning_rate": 0.00015673212147308826,
+      "loss": 3.0495,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2851452883135693,
+      "grad_norm": 0.8525515198707581,
+      "learning_rate": 0.00015670598638627706,
+      "loss": 3.0943,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2852358106273196,
+      "grad_norm": 0.9663558602333069,
+      "learning_rate": 0.00015667984558917464,
+      "loss": 3.0162,
+      "step": 3151
+    },
+    {
+      "epoch": 0.28532633294106996,
+      "grad_norm": 0.9535746574401855,
+      "learning_rate": 0.00015665369908441328,
+      "loss": 3.0412,
+      "step": 3152
+    },
+    {
+      "epoch": 0.2854168552548203,
+      "grad_norm": 0.8646260499954224,
+      "learning_rate": 0.00015662754687462596,
+      "loss": 2.9971,
+      "step": 3153
+    },
+    {
+      "epoch": 0.28550737756857064,
+      "grad_norm": 0.9250417351722717,
+      "learning_rate": 0.00015660138896244624,
+      "loss": 2.9493,
+      "step": 3154
+    },
+    {
+      "epoch": 0.285597899882321,
+      "grad_norm": 0.9591944217681885,
+      "learning_rate": 0.00015657522535050822,
+      "loss": 3.0685,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2856884221960713,
+      "grad_norm": 0.9319788217544556,
+      "learning_rate": 0.00015654905604144657,
+      "loss": 3.0092,
+      "step": 3156
+    },
+    {
+      "epoch": 0.28577894450982166,
+      "grad_norm": 0.8835229873657227,
+      "learning_rate": 0.00015652288103789648,
+      "loss": 2.9717,
+      "step": 3157
+    },
+    {
+      "epoch": 0.285869466823572,
+      "grad_norm": 0.9241860508918762,
+      "learning_rate": 0.0001564967003424938,
+      "loss": 2.9682,
+      "step": 3158
+    },
+    {
+      "epoch": 0.28595998913732235,
+      "grad_norm": 0.8293352127075195,
+      "learning_rate": 0.00015647051395787498,
+      "loss": 2.9572,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2860505114510727,
+      "grad_norm": 0.8276857137680054,
+      "learning_rate": 0.00015644432188667695,
+      "loss": 3.0033,
+      "step": 3160
+    },
+    {
+      "epoch": 0.286141033764823,
+      "grad_norm": 0.891465425491333,
+      "learning_rate": 0.00015641812413153724,
+      "loss": 3.0832,
+      "step": 3161
+    },
+    {
+      "epoch": 0.28623155607857337,
+      "grad_norm": 0.823231041431427,
+      "learning_rate": 0.00015639192069509395,
+      "loss": 3.0404,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2863220783923237,
+      "grad_norm": 0.8683503270149231,
+      "learning_rate": 0.0001563657115799858,
+      "loss": 2.9718,
+      "step": 3163
+    },
+    {
+      "epoch": 0.28641260070607405,
+      "grad_norm": 0.8113049864768982,
+      "learning_rate": 0.00015633949678885208,
+      "loss": 3.0384,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2865031230198244,
+      "grad_norm": 0.8468487858772278,
+      "learning_rate": 0.0001563132763243325,
+      "loss": 3.0317,
+      "step": 3165
+    },
+    {
+      "epoch": 0.28659364533357473,
+      "grad_norm": 0.8657675385475159,
+      "learning_rate": 0.0001562870501890676,
+      "loss": 3.048,
+      "step": 3166
+    },
+    {
+      "epoch": 0.28668416764732507,
+      "grad_norm": 0.8631026148796082,
+      "learning_rate": 0.00015626081838569826,
+      "loss": 3.0143,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2867746899610754,
+      "grad_norm": 0.8840932846069336,
+      "learning_rate": 0.00015623458091686604,
+      "loss": 3.0532,
+      "step": 3168
+    },
+    {
+      "epoch": 0.28686521227482575,
+      "grad_norm": 0.8777768611907959,
+      "learning_rate": 0.00015620833778521307,
+      "loss": 3.0652,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2869557345885761,
+      "grad_norm": 1.0639394521713257,
+      "learning_rate": 0.00015618208899338202,
+      "loss": 3.0654,
+      "step": 3170
+    },
+    {
+      "epoch": 0.28704625690232644,
+      "grad_norm": 0.9420127868652344,
+      "learning_rate": 0.00015615583454401617,
+      "loss": 2.967,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2871367792160768,
+      "grad_norm": 0.9539990425109863,
+      "learning_rate": 0.0001561295744397593,
+      "loss": 3.0091,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2872273015298271,
+      "grad_norm": 0.9690334796905518,
+      "learning_rate": 0.0001561033086832558,
+      "loss": 3.0722,
+      "step": 3173
+    },
+    {
+      "epoch": 0.28731782384357746,
+      "grad_norm": 0.9461049437522888,
+      "learning_rate": 0.00015607703727715065,
+      "loss": 3.0282,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2874083461573278,
+      "grad_norm": 1.0628281831741333,
+      "learning_rate": 0.0001560507602240894,
+      "loss": 3.0082,
+      "step": 3175
+    },
+    {
+      "epoch": 0.28749886847107814,
+      "grad_norm": 0.8840575814247131,
+      "learning_rate": 0.00015602447752671814,
+      "loss": 3.0105,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2875893907848285,
+      "grad_norm": 0.9166979789733887,
+      "learning_rate": 0.00015599818918768352,
+      "loss": 2.9886,
+      "step": 3177
+    },
+    {
+      "epoch": 0.2876799130985788,
+      "grad_norm": 0.9768393039703369,
+      "learning_rate": 0.00015597189520963277,
+      "loss": 2.9479,
+      "step": 3178
+    },
+    {
+      "epoch": 0.28777043541232916,
+      "grad_norm": 0.9496955871582031,
+      "learning_rate": 0.0001559455955952137,
+      "loss": 3.0099,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2878609577260795,
+      "grad_norm": 0.886669933795929,
+      "learning_rate": 0.0001559192903470747,
+      "loss": 2.9841,
+      "step": 3180
+    },
+    {
+      "epoch": 0.28795148003982984,
+      "grad_norm": 0.8499637842178345,
+      "learning_rate": 0.00015589297946786467,
+      "loss": 2.9788,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2880420023535802,
+      "grad_norm": 0.917888879776001,
+      "learning_rate": 0.00015586666296023315,
+      "loss": 3.0087,
+      "step": 3182
+    },
+    {
+      "epoch": 0.28813252466733047,
+      "grad_norm": 0.8253499269485474,
+      "learning_rate": 0.00015584034082683022,
+      "loss": 3.009,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2882230469810808,
+      "grad_norm": 0.7821881771087646,
+      "learning_rate": 0.00015581401307030647,
+      "loss": 3.0094,
+      "step": 3184
+    },
+    {
+      "epoch": 0.28831356929483115,
+      "grad_norm": 0.9027956128120422,
+      "learning_rate": 0.00015578767969331314,
+      "loss": 3.0757,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2884040916085815,
+      "grad_norm": 0.8098450303077698,
+      "learning_rate": 0.000155761340698502,
+      "loss": 2.9402,
+      "step": 3186
+    },
+    {
+      "epoch": 0.28849461392233183,
+      "grad_norm": 0.8672653436660767,
+      "learning_rate": 0.00015573499608852537,
+      "loss": 3.0249,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2885851362360822,
+      "grad_norm": 0.9086789488792419,
+      "learning_rate": 0.00015570864586603612,
+      "loss": 3.0921,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2886756585498325,
+      "grad_norm": 0.8213782906532288,
+      "learning_rate": 0.0001556822900336878,
+      "loss": 2.9316,
+      "step": 3189
+    },
+    {
+      "epoch": 0.28876618086358286,
+      "grad_norm": 0.8321839570999146,
+      "learning_rate": 0.0001556559285941344,
+      "loss": 3.0399,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2888567031773332,
+      "grad_norm": 0.8627842664718628,
+      "learning_rate": 0.0001556295615500305,
+      "loss": 2.9849,
+      "step": 3191
+    },
+    {
+      "epoch": 0.28894722549108354,
+      "grad_norm": 0.8597263693809509,
+      "learning_rate": 0.00015560318890403126,
+      "loss": 2.9951,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2890377478048339,
+      "grad_norm": 0.8256116509437561,
+      "learning_rate": 0.00015557681065879244,
+      "loss": 2.9524,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2891282701185842,
+      "grad_norm": 0.9126301407814026,
+      "learning_rate": 0.0001555504268169703,
+      "loss": 2.949,
+      "step": 3194
+    },
+    {
+      "epoch": 0.28921879243233456,
+      "grad_norm": 0.8662191033363342,
+      "learning_rate": 0.0001555240373812217,
+      "loss": 3.0159,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2893093147460849,
+      "grad_norm": 0.852206289768219,
+      "learning_rate": 0.00015549764235420405,
+      "loss": 3.0154,
+      "step": 3196
+    },
+    {
+      "epoch": 0.28939983705983524,
+      "grad_norm": 0.8456076383590698,
+      "learning_rate": 0.00015547124173857535,
+      "loss": 3.0452,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2894903593735856,
+      "grad_norm": 0.8650831580162048,
+      "learning_rate": 0.00015544483553699408,
+      "loss": 3.0385,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2895808816873359,
+      "grad_norm": 0.8533116579055786,
+      "learning_rate": 0.00015541842375211946,
+      "loss": 3.0309,
+      "step": 3199
+    },
+    {
+      "epoch": 0.28967140400108626,
+      "grad_norm": 0.8483850955963135,
+      "learning_rate": 0.00015539200638661104,
+      "loss": 3.029,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2897619263148366,
+      "grad_norm": 0.7920590043067932,
+      "learning_rate": 0.00015536558344312907,
+      "loss": 2.9799,
+      "step": 3201
+    },
+    {
+      "epoch": 0.28985244862858695,
+      "grad_norm": 0.8692976832389832,
+      "learning_rate": 0.00015533915492433443,
+      "loss": 3.0048,
+      "step": 3202
+    },
+    {
+      "epoch": 0.2899429709423373,
+      "grad_norm": 0.8859113454818726,
+      "learning_rate": 0.00015531272083288834,
+      "loss": 3.0178,
+      "step": 3203
+    },
+    {
+      "epoch": 0.2900334932560876,
+      "grad_norm": 0.8331505060195923,
+      "learning_rate": 0.0001552862811714528,
+      "loss": 3.0049,
+      "step": 3204
+    },
+    {
+      "epoch": 0.29012401556983797,
+      "grad_norm": 0.8357052206993103,
+      "learning_rate": 0.00015525983594269027,
+      "loss": 2.9647,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2902145378835883,
+      "grad_norm": 0.9645798802375793,
+      "learning_rate": 0.00015523338514926376,
+      "loss": 2.9726,
+      "step": 3206
+    },
+    {
+      "epoch": 0.29030506019733865,
+      "grad_norm": 0.8409217000007629,
+      "learning_rate": 0.00015520692879383686,
+      "loss": 2.9647,
+      "step": 3207
+    },
+    {
+      "epoch": 0.290395582511089,
+      "grad_norm": 1.0742449760437012,
+      "learning_rate": 0.00015518046687907377,
+      "loss": 3.0654,
+      "step": 3208
+    },
+    {
+      "epoch": 0.29048610482483933,
+      "grad_norm": 0.9031012654304504,
+      "learning_rate": 0.00015515399940763916,
+      "loss": 2.9865,
+      "step": 3209
+    },
+    {
+      "epoch": 0.29057662713858967,
+      "grad_norm": 0.9780822396278381,
+      "learning_rate": 0.00015512752638219835,
+      "loss": 3.0079,
+      "step": 3210
+    },
+    {
+      "epoch": 0.29066714945234,
+      "grad_norm": 1.170419692993164,
+      "learning_rate": 0.0001551010478054171,
+      "loss": 3.0254,
+      "step": 3211
+    },
+    {
+      "epoch": 0.29075767176609035,
+      "grad_norm": 0.9002463817596436,
+      "learning_rate": 0.00015507456367996184,
+      "loss": 3.0676,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2908481940798407,
+      "grad_norm": 1.1873091459274292,
+      "learning_rate": 0.00015504807400849958,
+      "loss": 2.9774,
+      "step": 3213
+    },
+    {
+      "epoch": 0.29093871639359103,
+      "grad_norm": 0.9433950185775757,
+      "learning_rate": 0.0001550215787936977,
+      "loss": 2.9299,
+      "step": 3214
+    },
+    {
+      "epoch": 0.2910292387073414,
+      "grad_norm": 1.0682939291000366,
+      "learning_rate": 0.0001549950780382244,
+      "loss": 3.013,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2911197610210917,
+      "grad_norm": 0.9073317646980286,
+      "learning_rate": 0.00015496857174474824,
+      "loss": 3.016,
+      "step": 3216
+    },
+    {
+      "epoch": 0.29121028333484206,
+      "grad_norm": 0.9233403205871582,
+      "learning_rate": 0.0001549420599159384,
+      "loss": 3.0042,
+      "step": 3217
+    },
+    {
+      "epoch": 0.2913008056485924,
+      "grad_norm": 0.9346694350242615,
+      "learning_rate": 0.00015491554255446462,
+      "loss": 3.0543,
+      "step": 3218
+    },
+    {
+      "epoch": 0.29139132796234274,
+      "grad_norm": 0.8480331897735596,
+      "learning_rate": 0.0001548890196629972,
+      "loss": 3.0001,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2914818502760931,
+      "grad_norm": 0.9132717251777649,
+      "learning_rate": 0.000154862491244207,
+      "loss": 2.9961,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2915723725898434,
+      "grad_norm": 0.9685938358306885,
+      "learning_rate": 0.00015483595730076546,
+      "loss": 2.9909,
+      "step": 3221
+    },
+    {
+      "epoch": 0.29166289490359376,
+      "grad_norm": 0.8925626277923584,
+      "learning_rate": 0.00015480941783534448,
+      "loss": 3.0059,
+      "step": 3222
+    },
+    {
+      "epoch": 0.2917534172173441,
+      "grad_norm": 0.8972873687744141,
+      "learning_rate": 0.00015478287285061666,
+      "loss": 2.99,
+      "step": 3223
+    },
+    {
+      "epoch": 0.2918439395310944,
+      "grad_norm": 0.8935641646385193,
+      "learning_rate": 0.00015475632234925504,
+      "loss": 2.9713,
+      "step": 3224
+    },
+    {
+      "epoch": 0.29193446184484473,
+      "grad_norm": 0.8276620507240295,
+      "learning_rate": 0.00015472976633393326,
+      "loss": 2.9518,
+      "step": 3225
+    },
+    {
+      "epoch": 0.29202498415859507,
+      "grad_norm": 0.9621675610542297,
+      "learning_rate": 0.0001547032048073255,
+      "loss": 3.0256,
+      "step": 3226
+    },
+    {
+      "epoch": 0.2921155064723454,
+      "grad_norm": 0.996496856212616,
+      "learning_rate": 0.0001546766377721065,
+      "loss": 2.9528,
+      "step": 3227
+    },
+    {
+      "epoch": 0.29220602878609575,
+      "grad_norm": 0.8234094381332397,
+      "learning_rate": 0.00015465006523095157,
+      "loss": 3.024,
+      "step": 3228
+    },
+    {
+      "epoch": 0.2922965510998461,
+      "grad_norm": 0.9537985920906067,
+      "learning_rate": 0.00015462348718653658,
+      "loss": 3.0105,
+      "step": 3229
+    },
+    {
+      "epoch": 0.29238707341359643,
+      "grad_norm": 1.0219292640686035,
+      "learning_rate": 0.0001545969036415379,
+      "loss": 3.0254,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2924775957273468,
+      "grad_norm": 0.8733369708061218,
+      "learning_rate": 0.00015457031459863258,
+      "loss": 3.0066,
+      "step": 3231
+    },
+    {
+      "epoch": 0.2925681180410971,
+      "grad_norm": 0.8705121874809265,
+      "learning_rate": 0.00015454372006049803,
+      "loss": 3.016,
+      "step": 3232
+    },
+    {
+      "epoch": 0.29265864035484745,
+      "grad_norm": 0.9468245506286621,
+      "learning_rate": 0.00015451712002981235,
+      "loss": 2.9881,
+      "step": 3233
+    },
+    {
+      "epoch": 0.2927491626685978,
+      "grad_norm": 0.8147585391998291,
+      "learning_rate": 0.0001544905145092542,
+      "loss": 3.0279,
+      "step": 3234
+    },
+    {
+      "epoch": 0.29283968498234814,
+      "grad_norm": 0.8886904716491699,
+      "learning_rate": 0.00015446390350150273,
+      "loss": 2.9563,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2929302072960985,
+      "grad_norm": 0.8884991407394409,
+      "learning_rate": 0.00015443728700923766,
+      "loss": 3.0518,
+      "step": 3236
+    },
+    {
+      "epoch": 0.2930207296098488,
+      "grad_norm": 0.8598198294639587,
+      "learning_rate": 0.00015441066503513927,
+      "loss": 3.0198,
+      "step": 3237
+    },
+    {
+      "epoch": 0.29311125192359916,
+      "grad_norm": 0.8423158526420593,
+      "learning_rate": 0.0001543840375818884,
+      "loss": 3.0396,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2932017742373495,
+      "grad_norm": 0.9256920218467712,
+      "learning_rate": 0.00015435740465216644,
+      "loss": 3.0247,
+      "step": 3239
+    },
+    {
+      "epoch": 0.29329229655109984,
+      "grad_norm": 0.862433910369873,
+      "learning_rate": 0.00015433076624865531,
+      "loss": 3.0077,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2933828188648502,
+      "grad_norm": 0.930212140083313,
+      "learning_rate": 0.00015430412237403752,
+      "loss": 3.0163,
+      "step": 3241
+    },
+    {
+      "epoch": 0.2934733411786005,
+      "grad_norm": 0.8511956334114075,
+      "learning_rate": 0.0001542774730309961,
+      "loss": 2.8893,
+      "step": 3242
+    },
+    {
+      "epoch": 0.29356386349235086,
+      "grad_norm": 0.8834841847419739,
+      "learning_rate": 0.0001542508182222146,
+      "loss": 2.9634,
+      "step": 3243
+    },
+    {
+      "epoch": 0.2936543858061012,
+      "grad_norm": 0.9784421920776367,
+      "learning_rate": 0.0001542241579503772,
+      "loss": 2.9846,
+      "step": 3244
+    },
+    {
+      "epoch": 0.29374490811985154,
+      "grad_norm": 0.8660554885864258,
+      "learning_rate": 0.00015419749221816858,
+      "loss": 2.9127,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2938354304336019,
+      "grad_norm": 0.9871695637702942,
+      "learning_rate": 0.000154170821028274,
+      "loss": 2.9124,
+      "step": 3246
+    },
+    {
+      "epoch": 0.2939259527473522,
+      "grad_norm": 0.8965557217597961,
+      "learning_rate": 0.0001541441443833792,
+      "loss": 2.9299,
+      "step": 3247
+    },
+    {
+      "epoch": 0.29401647506110257,
+      "grad_norm": 0.8073522448539734,
+      "learning_rate": 0.00015411746228617055,
+      "loss": 2.9716,
+      "step": 3248
+    },
+    {
+      "epoch": 0.2941069973748529,
+      "grad_norm": 0.9049869775772095,
+      "learning_rate": 0.00015409077473933492,
+      "loss": 3.0341,
+      "step": 3249
+    },
+    {
+      "epoch": 0.29419751968860325,
+      "grad_norm": 0.9170584082603455,
+      "learning_rate": 0.00015406408174555976,
+      "loss": 3.0301,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2942880420023536,
+      "grad_norm": 0.9323590397834778,
+      "learning_rate": 0.00015403738330753306,
+      "loss": 3.0336,
+      "step": 3251
+    },
+    {
+      "epoch": 0.29437856431610393,
+      "grad_norm": 0.9274752736091614,
+      "learning_rate": 0.00015401067942794332,
+      "loss": 3.0528,
+      "step": 3252
+    },
+    {
+      "epoch": 0.29446908662985427,
+      "grad_norm": 0.8868407011032104,
+      "learning_rate": 0.00015398397010947965,
+      "loss": 2.9815,
+      "step": 3253
+    },
+    {
+      "epoch": 0.2945596089436046,
+      "grad_norm": 0.9790063500404358,
+      "learning_rate": 0.00015395725535483168,
+      "loss": 3.0448,
+      "step": 3254
+    },
+    {
+      "epoch": 0.29465013125735495,
+      "grad_norm": 0.8878175020217896,
+      "learning_rate": 0.00015393053516668954,
+      "loss": 2.9678,
+      "step": 3255
+    },
+    {
+      "epoch": 0.2947406535711053,
+      "grad_norm": 0.897851824760437,
+      "learning_rate": 0.000153903809547744,
+      "loss": 3.0092,
+      "step": 3256
+    },
+    {
+      "epoch": 0.29483117588485563,
+      "grad_norm": 0.9687259793281555,
+      "learning_rate": 0.0001538770785006863,
+      "loss": 2.9965,
+      "step": 3257
+    },
+    {
+      "epoch": 0.294921698198606,
+      "grad_norm": 0.9195958971977234,
+      "learning_rate": 0.0001538503420282083,
+      "loss": 3.0088,
+      "step": 3258
+    },
+    {
+      "epoch": 0.2950122205123563,
+      "grad_norm": 0.9068818092346191,
+      "learning_rate": 0.00015382360013300233,
+      "loss": 2.9513,
+      "step": 3259
+    },
+    {
+      "epoch": 0.29510274282610666,
+      "grad_norm": 0.8815194368362427,
+      "learning_rate": 0.00015379685281776125,
+      "loss": 2.9852,
+      "step": 3260
+    },
+    {
+      "epoch": 0.295193265139857,
+      "grad_norm": 0.8915935754776001,
+      "learning_rate": 0.0001537701000851786,
+      "loss": 2.9612,
+      "step": 3261
+    },
+    {
+      "epoch": 0.29528378745360734,
+      "grad_norm": 0.9282931685447693,
+      "learning_rate": 0.00015374334193794838,
+      "loss": 3.0762,
+      "step": 3262
+    },
+    {
+      "epoch": 0.2953743097673577,
+      "grad_norm": 0.9892238974571228,
+      "learning_rate": 0.00015371657837876502,
+      "loss": 3.0449,
+      "step": 3263
+    },
+    {
+      "epoch": 0.295464832081108,
+      "grad_norm": 0.9499996304512024,
+      "learning_rate": 0.00015368980941032372,
+      "loss": 2.9896,
+      "step": 3264
+    },
+    {
+      "epoch": 0.2955553543948583,
+      "grad_norm": 0.8943758606910706,
+      "learning_rate": 0.00015366303503532009,
+      "loss": 3.0555,
+      "step": 3265
+    },
+    {
+      "epoch": 0.29564587670860865,
+      "grad_norm": 0.926461935043335,
+      "learning_rate": 0.0001536362552564503,
+      "loss": 2.9596,
+      "step": 3266
+    },
+    {
+      "epoch": 0.295736399022359,
+      "grad_norm": 0.8656926155090332,
+      "learning_rate": 0.00015360947007641106,
+      "loss": 2.9488,
+      "step": 3267
+    },
+    {
+      "epoch": 0.2958269213361093,
+      "grad_norm": 0.9003334641456604,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 3.0083,
+      "step": 3268
+    },
+    {
+      "epoch": 0.29591744364985967,
+      "grad_norm": 0.9856387376785278,
+      "learning_rate": 0.0001535558835236139,
+      "loss": 3.0598,
+      "step": 3269
+    },
+    {
+      "epoch": 0.29600796596361,
+      "grad_norm": 0.8337308764457703,
+      "learning_rate": 0.00015352908215625214,
+      "loss": 2.9763,
+      "step": 3270
+    },
+    {
+      "epoch": 0.29609848827736035,
+      "grad_norm": 0.8456469774246216,
+      "learning_rate": 0.00015350227539851326,
+      "loss": 2.9611,
+      "step": 3271
+    },
+    {
+      "epoch": 0.2961890105911107,
+      "grad_norm": 0.9669084548950195,
+      "learning_rate": 0.00015347546325309672,
+      "loss": 2.9944,
+      "step": 3272
+    },
+    {
+      "epoch": 0.29627953290486103,
+      "grad_norm": 0.9577633738517761,
+      "learning_rate": 0.0001534486457227025,
+      "loss": 3.0196,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2963700552186114,
+      "grad_norm": 0.8991074562072754,
+      "learning_rate": 0.00015342182281003112,
+      "loss": 2.9924,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2964605775323617,
+      "grad_norm": 1.0249013900756836,
+      "learning_rate": 0.00015339499451778364,
+      "loss": 2.9924,
+      "step": 3275
+    },
+    {
+      "epoch": 0.29655109984611205,
+      "grad_norm": 1.0314909219741821,
+      "learning_rate": 0.00015336816084866165,
+      "loss": 3.0581,
+      "step": 3276
+    },
+    {
+      "epoch": 0.2966416221598624,
+      "grad_norm": 0.785307765007019,
+      "learning_rate": 0.00015334132180536733,
+      "loss": 2.9663,
+      "step": 3277
+    },
+    {
+      "epoch": 0.29673214447361274,
+      "grad_norm": 0.9974069595336914,
+      "learning_rate": 0.00015331447739060338,
+      "loss": 2.9589,
+      "step": 3278
+    },
+    {
+      "epoch": 0.2968226667873631,
+      "grad_norm": 0.9813283681869507,
+      "learning_rate": 0.000153287627607073,
+      "loss": 3.0866,
+      "step": 3279
+    },
+    {
+      "epoch": 0.2969131891011134,
+      "grad_norm": 0.8992034792900085,
+      "learning_rate": 0.00015326077245747999,
+      "loss": 3.0273,
+      "step": 3280
+    },
+    {
+      "epoch": 0.29700371141486376,
+      "grad_norm": 1.013922095298767,
+      "learning_rate": 0.00015323391194452864,
+      "loss": 3.0097,
+      "step": 3281
+    },
+    {
+      "epoch": 0.2970942337286141,
+      "grad_norm": 0.950112521648407,
+      "learning_rate": 0.00015320704607092384,
+      "loss": 2.9898,
+      "step": 3282
+    },
+    {
+      "epoch": 0.29718475604236444,
+      "grad_norm": 0.9233756065368652,
+      "learning_rate": 0.00015318017483937092,
+      "loss": 2.9998,
+      "step": 3283
+    },
+    {
+      "epoch": 0.2972752783561148,
+      "grad_norm": 0.9420918226242065,
+      "learning_rate": 0.0001531532982525759,
+      "loss": 2.9409,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2973658006698651,
+      "grad_norm": 0.8628808856010437,
+      "learning_rate": 0.00015312641631324515,
+      "loss": 3.0076,
+      "step": 3285
+    },
+    {
+      "epoch": 0.29745632298361546,
+      "grad_norm": 0.869685709476471,
+      "learning_rate": 0.00015309952902408576,
+      "loss": 3.0232,
+      "step": 3286
+    },
+    {
+      "epoch": 0.2975468452973658,
+      "grad_norm": 0.8513570427894592,
+      "learning_rate": 0.00015307263638780522,
+      "loss": 2.9768,
+      "step": 3287
+    },
+    {
+      "epoch": 0.29763736761111614,
+      "grad_norm": 0.9360085129737854,
+      "learning_rate": 0.00015304573840711167,
+      "loss": 3.0254,
+      "step": 3288
+    },
+    {
+      "epoch": 0.2977278899248665,
+      "grad_norm": 1.0327681303024292,
+      "learning_rate": 0.00015301883508471372,
+      "loss": 3.078,
+      "step": 3289
+    },
+    {
+      "epoch": 0.2978184122386168,
+      "grad_norm": 0.8258141279220581,
+      "learning_rate": 0.0001529919264233205,
+      "loss": 2.9808,
+      "step": 3290
+    },
+    {
+      "epoch": 0.29790893455236717,
+      "grad_norm": 0.9610850811004639,
+      "learning_rate": 0.00015296501242564173,
+      "loss": 3.0336,
+      "step": 3291
+    },
+    {
+      "epoch": 0.2979994568661175,
+      "grad_norm": 0.9272431135177612,
+      "learning_rate": 0.00015293809309438773,
+      "loss": 3.072,
+      "step": 3292
+    },
+    {
+      "epoch": 0.29808997917986785,
+      "grad_norm": 0.9149249792098999,
+      "learning_rate": 0.00015291116843226915,
+      "loss": 2.9522,
+      "step": 3293
+    },
+    {
+      "epoch": 0.2981805014936182,
+      "grad_norm": 0.9025709629058838,
+      "learning_rate": 0.00015288423844199734,
+      "loss": 2.9639,
+      "step": 3294
+    },
+    {
+      "epoch": 0.29827102380736853,
+      "grad_norm": 0.8880616426467896,
+      "learning_rate": 0.0001528573031262842,
+      "loss": 3.0719,
+      "step": 3295
+    },
+    {
+      "epoch": 0.29836154612111887,
+      "grad_norm": 0.8158171772956848,
+      "learning_rate": 0.00015283036248784208,
+      "loss": 2.9746,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2984520684348692,
+      "grad_norm": 0.8735789060592651,
+      "learning_rate": 0.0001528034165293839,
+      "loss": 3.0311,
+      "step": 3297
+    },
+    {
+      "epoch": 0.29854259074861955,
+      "grad_norm": 0.8365547060966492,
+      "learning_rate": 0.0001527764652536231,
+      "loss": 2.9659,
+      "step": 3298
+    },
+    {
+      "epoch": 0.2986331130623699,
+      "grad_norm": 0.8127195239067078,
+      "learning_rate": 0.0001527495086632737,
+      "loss": 3.0082,
+      "step": 3299
+    },
+    {
+      "epoch": 0.29872363537612023,
+      "grad_norm": 1.0759278535842896,
+      "learning_rate": 0.00015272254676105025,
+      "loss": 3.0071,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2988141576898706,
+      "grad_norm": 0.8265242576599121,
+      "learning_rate": 0.00015269557954966778,
+      "loss": 2.9862,
+      "step": 3301
+    },
+    {
+      "epoch": 0.2989046800036209,
+      "grad_norm": 0.8545429110527039,
+      "learning_rate": 0.00015266860703184188,
+      "loss": 2.998,
+      "step": 3302
+    },
+    {
+      "epoch": 0.29899520231737126,
+      "grad_norm": 0.9787036776542664,
+      "learning_rate": 0.00015264162921028865,
+      "loss": 2.9452,
+      "step": 3303
+    },
+    {
+      "epoch": 0.2990857246311216,
+      "grad_norm": 0.8430793881416321,
+      "learning_rate": 0.00015261464608772488,
+      "loss": 2.9786,
+      "step": 3304
+    },
+    {
+      "epoch": 0.29917624694487194,
+      "grad_norm": 0.9696670770645142,
+      "learning_rate": 0.00015258765766686761,
+      "loss": 3.0475,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2992667692586222,
+      "grad_norm": 0.8284124135971069,
+      "learning_rate": 0.00015256066395043471,
+      "loss": 2.9923,
+      "step": 3306
+    },
+    {
+      "epoch": 0.29935729157237256,
+      "grad_norm": 0.8875390291213989,
+      "learning_rate": 0.00015253366494114435,
+      "loss": 2.9976,
+      "step": 3307
+    },
+    {
+      "epoch": 0.2994478138861229,
+      "grad_norm": 0.9056249856948853,
+      "learning_rate": 0.00015250666064171538,
+      "loss": 2.9238,
+      "step": 3308
+    },
+    {
+      "epoch": 0.29953833619987325,
+      "grad_norm": 0.885613739490509,
+      "learning_rate": 0.00015247965105486712,
+      "loss": 2.9875,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2996288585136236,
+      "grad_norm": 0.9627782106399536,
+      "learning_rate": 0.00015245263618331945,
+      "loss": 3.0419,
+      "step": 3310
+    },
+    {
+      "epoch": 0.2997193808273739,
+      "grad_norm": 0.9762522578239441,
+      "learning_rate": 0.00015242561602979275,
+      "loss": 2.9858,
+      "step": 3311
+    },
+    {
+      "epoch": 0.29980990314112427,
+      "grad_norm": 0.9357553720474243,
+      "learning_rate": 0.00015239859059700794,
+      "loss": 2.9092,
+      "step": 3312
+    },
+    {
+      "epoch": 0.2999004254548746,
+      "grad_norm": 1.0730036497116089,
+      "learning_rate": 0.00015237155988768647,
+      "loss": 2.9937,
+      "step": 3313
+    },
+    {
+      "epoch": 0.29999094776862495,
+      "grad_norm": 0.917066752910614,
+      "learning_rate": 0.00015234452390455036,
+      "loss": 3.0989,
+      "step": 3314
+    },
+    {
+      "epoch": 0.3000814700823753,
+      "grad_norm": 1.0417499542236328,
+      "learning_rate": 0.00015231748265032215,
+      "loss": 2.9708,
+      "step": 3315
+    },
+    {
+      "epoch": 0.30017199239612563,
+      "grad_norm": 0.9708789587020874,
+      "learning_rate": 0.00015229043612772486,
+      "loss": 2.9939,
+      "step": 3316
+    },
+    {
+      "epoch": 0.30026251470987597,
+      "grad_norm": 0.9104913473129272,
+      "learning_rate": 0.0001522633843394821,
+      "loss": 2.9814,
+      "step": 3317
+    },
+    {
+      "epoch": 0.3003530370236263,
+      "grad_norm": 0.9186818599700928,
+      "learning_rate": 0.0001522363272883179,
+      "loss": 2.9769,
+      "step": 3318
+    },
+    {
+      "epoch": 0.30044355933737665,
+      "grad_norm": 0.9663811922073364,
+      "learning_rate": 0.00015220926497695704,
+      "loss": 3.058,
+      "step": 3319
+    },
+    {
+      "epoch": 0.300534081651127,
+      "grad_norm": 0.9680636525154114,
+      "learning_rate": 0.0001521821974081246,
+      "loss": 2.9354,
+      "step": 3320
+    },
+    {
+      "epoch": 0.30062460396487733,
+      "grad_norm": 0.9134761095046997,
+      "learning_rate": 0.00015215512458454633,
+      "loss": 2.9999,
+      "step": 3321
+    },
+    {
+      "epoch": 0.3007151262786277,
+      "grad_norm": 0.9421324729919434,
+      "learning_rate": 0.0001521280465089484,
+      "loss": 3.0146,
+      "step": 3322
+    },
+    {
+      "epoch": 0.300805648592378,
+      "grad_norm": 0.9263170957565308,
+      "learning_rate": 0.00015210096318405767,
+      "loss": 2.972,
+      "step": 3323
+    },
+    {
+      "epoch": 0.30089617090612836,
+      "grad_norm": 0.9598850011825562,
+      "learning_rate": 0.00015207387461260133,
+      "loss": 3.07,
+      "step": 3324
+    },
+    {
+      "epoch": 0.3009866932198787,
+      "grad_norm": 0.8545998334884644,
+      "learning_rate": 0.00015204678079730724,
+      "loss": 3.0162,
+      "step": 3325
+    },
+    {
+      "epoch": 0.30107721553362904,
+      "grad_norm": 0.9190890192985535,
+      "learning_rate": 0.00015201968174090373,
+      "loss": 2.9696,
+      "step": 3326
+    },
+    {
+      "epoch": 0.3011677378473794,
+      "grad_norm": 0.942925751209259,
+      "learning_rate": 0.00015199257744611977,
+      "loss": 3.0451,
+      "step": 3327
+    },
+    {
+      "epoch": 0.3012582601611297,
+      "grad_norm": 0.8846598267555237,
+      "learning_rate": 0.0001519654679156846,
+      "loss": 2.996,
+      "step": 3328
+    },
+    {
+      "epoch": 0.30134878247488006,
+      "grad_norm": 0.8187833428382874,
+      "learning_rate": 0.00015193835315232827,
+      "loss": 2.9353,
+      "step": 3329
+    },
+    {
+      "epoch": 0.3014393047886304,
+      "grad_norm": 0.8995598554611206,
+      "learning_rate": 0.00015191123315878123,
+      "loss": 2.992,
+      "step": 3330
+    },
+    {
+      "epoch": 0.30152982710238074,
+      "grad_norm": 0.856484055519104,
+      "learning_rate": 0.00015188410793777436,
+      "loss": 2.963,
+      "step": 3331
+    },
+    {
+      "epoch": 0.3016203494161311,
+      "grad_norm": 0.9061743021011353,
+      "learning_rate": 0.00015185697749203926,
+      "loss": 2.9959,
+      "step": 3332
+    },
+    {
+      "epoch": 0.3017108717298814,
+      "grad_norm": 0.9021839499473572,
+      "learning_rate": 0.000151829841824308,
+      "loss": 2.9489,
+      "step": 3333
+    },
+    {
+      "epoch": 0.30180139404363177,
+      "grad_norm": 0.9534750580787659,
+      "learning_rate": 0.00015180270093731303,
+      "loss": 3.0352,
+      "step": 3334
+    },
+    {
+      "epoch": 0.3018919163573821,
+      "grad_norm": 0.8187622427940369,
+      "learning_rate": 0.00015177555483378752,
+      "loss": 3.0018,
+      "step": 3335
+    },
+    {
+      "epoch": 0.30198243867113245,
+      "grad_norm": 0.8136058449745178,
+      "learning_rate": 0.00015174840351646503,
+      "loss": 2.9523,
+      "step": 3336
+    },
+    {
+      "epoch": 0.3020729609848828,
+      "grad_norm": 0.8742151260375977,
+      "learning_rate": 0.00015172124698807973,
+      "loss": 2.9309,
+      "step": 3337
+    },
+    {
+      "epoch": 0.30216348329863313,
+      "grad_norm": 0.9198973774909973,
+      "learning_rate": 0.0001516940852513663,
+      "loss": 2.9981,
+      "step": 3338
+    },
+    {
+      "epoch": 0.30225400561238347,
+      "grad_norm": 0.8850584030151367,
+      "learning_rate": 0.00015166691830905988,
+      "loss": 3.0065,
+      "step": 3339
+    },
+    {
+      "epoch": 0.3023445279261338,
+      "grad_norm": 0.9615797400474548,
+      "learning_rate": 0.0001516397461638962,
+      "loss": 3.0168,
+      "step": 3340
+    },
+    {
+      "epoch": 0.30243505023988415,
+      "grad_norm": 0.8681991696357727,
+      "learning_rate": 0.0001516125688186115,
+      "loss": 3.05,
+      "step": 3341
+    },
+    {
+      "epoch": 0.3025255725536345,
+      "grad_norm": 0.8902552723884583,
+      "learning_rate": 0.00015158538627594253,
+      "loss": 3.0239,
+      "step": 3342
+    },
+    {
+      "epoch": 0.30261609486738483,
+      "grad_norm": 0.9143476486206055,
+      "learning_rate": 0.00015155819853862655,
+      "loss": 2.9265,
+      "step": 3343
+    },
+    {
+      "epoch": 0.3027066171811352,
+      "grad_norm": 0.8717308044433594,
+      "learning_rate": 0.00015153100560940142,
+      "loss": 2.9903,
+      "step": 3344
+    },
+    {
+      "epoch": 0.3027971394948855,
+      "grad_norm": 0.8563142418861389,
+      "learning_rate": 0.00015150380749100545,
+      "loss": 3.0039,
+      "step": 3345
+    },
+    {
+      "epoch": 0.30288766180863586,
+      "grad_norm": 0.9677398800849915,
+      "learning_rate": 0.00015147660418617743,
+      "loss": 2.9603,
+      "step": 3346
+    },
+    {
+      "epoch": 0.30297818412238614,
+      "grad_norm": 0.8721147775650024,
+      "learning_rate": 0.0001514493956976568,
+      "loss": 2.9895,
+      "step": 3347
+    },
+    {
+      "epoch": 0.3030687064361365,
+      "grad_norm": 0.9216533899307251,
+      "learning_rate": 0.00015142218202818342,
+      "loss": 3.0475,
+      "step": 3348
+    },
+    {
+      "epoch": 0.3031592287498868,
+      "grad_norm": 0.9659701585769653,
+      "learning_rate": 0.00015139496318049774,
+      "loss": 2.9786,
+      "step": 3349
+    },
+    {
+      "epoch": 0.30324975106363716,
+      "grad_norm": 0.8679906129837036,
+      "learning_rate": 0.00015136773915734066,
+      "loss": 3.0225,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3033402733773875,
+      "grad_norm": 0.9853136539459229,
+      "learning_rate": 0.00015134050996145362,
+      "loss": 2.9994,
+      "step": 3351
+    },
+    {
+      "epoch": 0.30343079569113784,
+      "grad_norm": 1.0005536079406738,
+      "learning_rate": 0.00015131327559557867,
+      "loss": 3.0669,
+      "step": 3352
+    },
+    {
+      "epoch": 0.3035213180048882,
+      "grad_norm": 0.9564321041107178,
+      "learning_rate": 0.0001512860360624583,
+      "loss": 2.9842,
+      "step": 3353
+    },
+    {
+      "epoch": 0.3036118403186385,
+      "grad_norm": 0.9644863605499268,
+      "learning_rate": 0.00015125879136483543,
+      "loss": 2.9576,
+      "step": 3354
+    },
+    {
+      "epoch": 0.30370236263238887,
+      "grad_norm": 1.0262359380722046,
+      "learning_rate": 0.00015123154150545372,
+      "loss": 3.0161,
+      "step": 3355
+    },
+    {
+      "epoch": 0.3037928849461392,
+      "grad_norm": 0.8304927349090576,
+      "learning_rate": 0.00015120428648705717,
+      "loss": 2.9595,
+      "step": 3356
+    },
+    {
+      "epoch": 0.30388340725988955,
+      "grad_norm": 0.9257547855377197,
+      "learning_rate": 0.00015117702631239037,
+      "loss": 2.97,
+      "step": 3357
+    },
+    {
+      "epoch": 0.3039739295736399,
+      "grad_norm": 0.8537124395370483,
+      "learning_rate": 0.00015114976098419842,
+      "loss": 2.9598,
+      "step": 3358
+    },
+    {
+      "epoch": 0.30406445188739023,
+      "grad_norm": 0.8638425469398499,
+      "learning_rate": 0.00015112249050522694,
+      "loss": 2.9879,
+      "step": 3359
+    },
+    {
+      "epoch": 0.30415497420114057,
+      "grad_norm": 0.8485719561576843,
+      "learning_rate": 0.00015109521487822206,
+      "loss": 2.9358,
+      "step": 3360
+    },
+    {
+      "epoch": 0.3042454965148909,
+      "grad_norm": 0.8621719479560852,
+      "learning_rate": 0.00015106793410593045,
+      "loss": 3.004,
+      "step": 3361
+    },
+    {
+      "epoch": 0.30433601882864125,
+      "grad_norm": 0.8881914019584656,
+      "learning_rate": 0.00015104064819109926,
+      "loss": 3.0266,
+      "step": 3362
+    },
+    {
+      "epoch": 0.3044265411423916,
+      "grad_norm": 0.9412158727645874,
+      "learning_rate": 0.0001510133571364762,
+      "loss": 2.9925,
+      "step": 3363
+    },
+    {
+      "epoch": 0.30451706345614193,
+      "grad_norm": 0.8518620729446411,
+      "learning_rate": 0.00015098606094480948,
+      "loss": 2.9884,
+      "step": 3364
+    },
+    {
+      "epoch": 0.3046075857698923,
+      "grad_norm": 0.8724182844161987,
+      "learning_rate": 0.0001509587596188478,
+      "loss": 2.9559,
+      "step": 3365
+    },
+    {
+      "epoch": 0.3046981080836426,
+      "grad_norm": 0.8881733417510986,
+      "learning_rate": 0.00015093145316134044,
+      "loss": 2.9055,
+      "step": 3366
+    },
+    {
+      "epoch": 0.30478863039739296,
+      "grad_norm": 0.8453825116157532,
+      "learning_rate": 0.00015090414157503714,
+      "loss": 2.9716,
+      "step": 3367
+    },
+    {
+      "epoch": 0.3048791527111433,
+      "grad_norm": 1.0025969743728638,
+      "learning_rate": 0.00015087682486268817,
+      "loss": 2.9664,
+      "step": 3368
+    },
+    {
+      "epoch": 0.30496967502489364,
+      "grad_norm": 1.0106847286224365,
+      "learning_rate": 0.00015084950302704433,
+      "loss": 2.9872,
+      "step": 3369
+    },
+    {
+      "epoch": 0.305060197338644,
+      "grad_norm": 0.9144858717918396,
+      "learning_rate": 0.00015082217607085692,
+      "loss": 2.9294,
+      "step": 3370
+    },
+    {
+      "epoch": 0.3051507196523943,
+      "grad_norm": 0.943499743938446,
+      "learning_rate": 0.00015079484399687778,
+      "loss": 2.9154,
+      "step": 3371
+    },
+    {
+      "epoch": 0.30524124196614466,
+      "grad_norm": 0.9145709276199341,
+      "learning_rate": 0.00015076750680785922,
+      "loss": 2.9809,
+      "step": 3372
+    },
+    {
+      "epoch": 0.305331764279895,
+      "grad_norm": 1.0750885009765625,
+      "learning_rate": 0.00015074016450655411,
+      "loss": 2.9281,
+      "step": 3373
+    },
+    {
+      "epoch": 0.30542228659364534,
+      "grad_norm": 0.8634659051895142,
+      "learning_rate": 0.00015071281709571585,
+      "loss": 2.9821,
+      "step": 3374
+    },
+    {
+      "epoch": 0.3055128089073957,
+      "grad_norm": 0.9092392325401306,
+      "learning_rate": 0.0001506854645780983,
+      "loss": 2.9494,
+      "step": 3375
+    },
+    {
+      "epoch": 0.305603331221146,
+      "grad_norm": 0.9126806259155273,
+      "learning_rate": 0.00015065810695645584,
+      "loss": 2.964,
+      "step": 3376
+    },
+    {
+      "epoch": 0.30569385353489636,
+      "grad_norm": 0.9306450486183167,
+      "learning_rate": 0.00015063074423354338,
+      "loss": 2.9637,
+      "step": 3377
+    },
+    {
+      "epoch": 0.3057843758486467,
+      "grad_norm": 1.1827573776245117,
+      "learning_rate": 0.00015060337641211637,
+      "loss": 3.0492,
+      "step": 3378
+    },
+    {
+      "epoch": 0.30587489816239705,
+      "grad_norm": 0.8481791615486145,
+      "learning_rate": 0.00015057600349493077,
+      "loss": 2.9656,
+      "step": 3379
+    },
+    {
+      "epoch": 0.3059654204761474,
+      "grad_norm": 0.951055109500885,
+      "learning_rate": 0.000150548625484743,
+      "loss": 3.0363,
+      "step": 3380
+    },
+    {
+      "epoch": 0.30605594278989773,
+      "grad_norm": 0.8536085486412048,
+      "learning_rate": 0.00015052124238430995,
+      "loss": 2.9393,
+      "step": 3381
+    },
+    {
+      "epoch": 0.30614646510364807,
+      "grad_norm": 0.8343144655227661,
+      "learning_rate": 0.00015049385419638926,
+      "loss": 3.0001,
+      "step": 3382
+    },
+    {
+      "epoch": 0.3062369874173984,
+      "grad_norm": 0.9116124510765076,
+      "learning_rate": 0.00015046646092373883,
+      "loss": 3.0328,
+      "step": 3383
+    },
+    {
+      "epoch": 0.30632750973114875,
+      "grad_norm": 0.8201098442077637,
+      "learning_rate": 0.0001504390625691171,
+      "loss": 3.0223,
+      "step": 3384
+    },
+    {
+      "epoch": 0.3064180320448991,
+      "grad_norm": 0.8883117437362671,
+      "learning_rate": 0.0001504116591352832,
+      "loss": 3.0032,
+      "step": 3385
+    },
+    {
+      "epoch": 0.30650855435864943,
+      "grad_norm": 0.883526086807251,
+      "learning_rate": 0.0001503842506249966,
+      "loss": 3.0075,
+      "step": 3386
+    },
+    {
+      "epoch": 0.3065990766723998,
+      "grad_norm": 0.8869072198867798,
+      "learning_rate": 0.00015035683704101732,
+      "loss": 2.8875,
+      "step": 3387
+    },
+    {
+      "epoch": 0.3066895989861501,
+      "grad_norm": 0.8745526075363159,
+      "learning_rate": 0.00015032941838610597,
+      "loss": 3.0136,
+      "step": 3388
+    },
+    {
+      "epoch": 0.3067801212999004,
+      "grad_norm": 0.8355185389518738,
+      "learning_rate": 0.00015030199466302353,
+      "loss": 2.9815,
+      "step": 3389
+    },
+    {
+      "epoch": 0.30687064361365074,
+      "grad_norm": 0.8745567202568054,
+      "learning_rate": 0.0001502745658745316,
+      "loss": 2.9707,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3069611659274011,
+      "grad_norm": 0.813612699508667,
+      "learning_rate": 0.0001502471320233923,
+      "loss": 2.9419,
+      "step": 3391
+    },
+    {
+      "epoch": 0.3070516882411514,
+      "grad_norm": 0.9350422024726868,
+      "learning_rate": 0.00015021969311236814,
+      "loss": 2.9618,
+      "step": 3392
+    },
+    {
+      "epoch": 0.30714221055490176,
+      "grad_norm": 0.888590931892395,
+      "learning_rate": 0.00015019224914422226,
+      "loss": 2.976,
+      "step": 3393
+    },
+    {
+      "epoch": 0.3072327328686521,
+      "grad_norm": 0.88670414686203,
+      "learning_rate": 0.00015016480012171828,
+      "loss": 3.0039,
+      "step": 3394
+    },
+    {
+      "epoch": 0.30732325518240244,
+      "grad_norm": 0.8592669367790222,
+      "learning_rate": 0.0001501373460476203,
+      "loss": 2.9485,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3074137774961528,
+      "grad_norm": 0.9061694741249084,
+      "learning_rate": 0.00015010988692469295,
+      "loss": 2.9608,
+      "step": 3396
+    },
+    {
+      "epoch": 0.3075042998099031,
+      "grad_norm": 0.8457714319229126,
+      "learning_rate": 0.00015008242275570133,
+      "loss": 2.9619,
+      "step": 3397
+    },
+    {
+      "epoch": 0.30759482212365347,
+      "grad_norm": 0.8951268196105957,
+      "learning_rate": 0.00015005495354341114,
+      "loss": 2.9305,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3076853444374038,
+      "grad_norm": 0.8539029359817505,
+      "learning_rate": 0.00015002747929058848,
+      "loss": 3.001,
+      "step": 3399
+    },
+    {
+      "epoch": 0.30777586675115415,
+      "grad_norm": 0.8860955238342285,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 2.9903,
+      "step": 3400
+    },
+    {
+      "epoch": 0.3078663890649045,
+      "grad_norm": 0.8125213980674744,
+      "learning_rate": 0.00014997251567441293,
+      "loss": 2.9353,
+      "step": 3401
+    },
+    {
+      "epoch": 0.30795691137865483,
+      "grad_norm": 0.8126906156539917,
+      "learning_rate": 0.0001499450263165949,
+      "loss": 2.9851,
+      "step": 3402
+    },
+    {
+      "epoch": 0.30804743369240517,
+      "grad_norm": 0.8433899283409119,
+      "learning_rate": 0.00014991753192931405,
+      "loss": 2.9667,
+      "step": 3403
+    },
+    {
+      "epoch": 0.3081379560061555,
+      "grad_norm": 0.8762634992599487,
+      "learning_rate": 0.00014989003251533912,
+      "loss": 2.9581,
+      "step": 3404
+    },
+    {
+      "epoch": 0.30822847831990585,
+      "grad_norm": 0.891649067401886,
+      "learning_rate": 0.0001498625280774393,
+      "loss": 2.911,
+      "step": 3405
+    },
+    {
+      "epoch": 0.3083190006336562,
+      "grad_norm": 0.9227769374847412,
+      "learning_rate": 0.00014983501861838425,
+      "loss": 2.9672,
+      "step": 3406
+    },
+    {
+      "epoch": 0.30840952294740653,
+      "grad_norm": 0.9137094020843506,
+      "learning_rate": 0.00014980750414094417,
+      "loss": 2.9415,
+      "step": 3407
+    },
+    {
+      "epoch": 0.3085000452611569,
+      "grad_norm": 0.878205418586731,
+      "learning_rate": 0.00014977998464788984,
+      "loss": 2.9455,
+      "step": 3408
+    },
+    {
+      "epoch": 0.3085905675749072,
+      "grad_norm": 0.8257315754890442,
+      "learning_rate": 0.00014975246014199237,
+      "loss": 2.9471,
+      "step": 3409
+    },
+    {
+      "epoch": 0.30868108988865756,
+      "grad_norm": 0.8512585163116455,
+      "learning_rate": 0.00014972493062602354,
+      "loss": 3.0183,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3087716122024079,
+      "grad_norm": 0.9373469352722168,
+      "learning_rate": 0.00014969739610275556,
+      "loss": 2.9832,
+      "step": 3411
+    },
+    {
+      "epoch": 0.30886213451615824,
+      "grad_norm": 0.8257470726966858,
+      "learning_rate": 0.00014966985657496114,
+      "loss": 2.9597,
+      "step": 3412
+    },
+    {
+      "epoch": 0.3089526568299086,
+      "grad_norm": 0.8489750027656555,
+      "learning_rate": 0.00014964231204541353,
+      "loss": 3.0061,
+      "step": 3413
+    },
+    {
+      "epoch": 0.3090431791436589,
+      "grad_norm": 0.8522926568984985,
+      "learning_rate": 0.00014961476251688645,
+      "loss": 2.9881,
+      "step": 3414
+    },
+    {
+      "epoch": 0.30913370145740926,
+      "grad_norm": 0.901701807975769,
+      "learning_rate": 0.00014958720799215414,
+      "loss": 3.0353,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3092242237711596,
+      "grad_norm": 0.8479433655738831,
+      "learning_rate": 0.00014955964847399137,
+      "loss": 2.9602,
+      "step": 3416
+    },
+    {
+      "epoch": 0.30931474608490994,
+      "grad_norm": 0.9179433584213257,
+      "learning_rate": 0.00014953208396517332,
+      "loss": 3.0362,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3094052683986603,
+      "grad_norm": 0.8743851780891418,
+      "learning_rate": 0.00014950451446847578,
+      "loss": 2.9541,
+      "step": 3418
+    },
+    {
+      "epoch": 0.3094957907124106,
+      "grad_norm": 0.8852507472038269,
+      "learning_rate": 0.00014947693998667498,
+      "loss": 3.0251,
+      "step": 3419
+    },
+    {
+      "epoch": 0.30958631302616096,
+      "grad_norm": 1.0363515615463257,
+      "learning_rate": 0.0001494493605225477,
+      "loss": 2.9889,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3096768353399113,
+      "grad_norm": 0.7981639504432678,
+      "learning_rate": 0.00014942177607887115,
+      "loss": 2.9148,
+      "step": 3421
+    },
+    {
+      "epoch": 0.30976735765366165,
+      "grad_norm": 0.962907075881958,
+      "learning_rate": 0.0001493941866584231,
+      "loss": 2.9303,
+      "step": 3422
+    },
+    {
+      "epoch": 0.309857879967412,
+      "grad_norm": 0.8996175527572632,
+      "learning_rate": 0.00014936659226398181,
+      "loss": 2.994,
+      "step": 3423
+    },
+    {
+      "epoch": 0.3099484022811623,
+      "grad_norm": 0.9445270299911499,
+      "learning_rate": 0.00014933899289832603,
+      "loss": 2.9394,
+      "step": 3424
+    },
+    {
+      "epoch": 0.31003892459491267,
+      "grad_norm": 0.9439915418624878,
+      "learning_rate": 0.00014931138856423502,
+      "loss": 2.9735,
+      "step": 3425
+    },
+    {
+      "epoch": 0.310129446908663,
+      "grad_norm": 0.813269853591919,
+      "learning_rate": 0.00014928377926448855,
+      "loss": 2.9808,
+      "step": 3426
+    },
+    {
+      "epoch": 0.31021996922241335,
+      "grad_norm": 0.861720621585846,
+      "learning_rate": 0.0001492561650018668,
+      "loss": 3.0155,
+      "step": 3427
+    },
+    {
+      "epoch": 0.3103104915361637,
+      "grad_norm": 1.0294265747070312,
+      "learning_rate": 0.00014922854577915064,
+      "loss": 2.9706,
+      "step": 3428
+    },
+    {
+      "epoch": 0.31040101384991403,
+      "grad_norm": 0.9049494862556458,
+      "learning_rate": 0.00014920092159912126,
+      "loss": 2.96,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3104915361636643,
+      "grad_norm": 1.1088683605194092,
+      "learning_rate": 0.0001491732924645604,
+      "loss": 3.0453,
+      "step": 3430
+    },
+    {
+      "epoch": 0.31058205847741466,
+      "grad_norm": 0.9710911512374878,
+      "learning_rate": 0.00014914565837825032,
+      "loss": 2.9452,
+      "step": 3431
+    },
+    {
+      "epoch": 0.310672580791165,
+      "grad_norm": 0.9654075503349304,
+      "learning_rate": 0.00014911801934297381,
+      "loss": 2.9662,
+      "step": 3432
+    },
+    {
+      "epoch": 0.31076310310491534,
+      "grad_norm": 0.8625348806381226,
+      "learning_rate": 0.00014909037536151409,
+      "loss": 2.9671,
+      "step": 3433
+    },
+    {
+      "epoch": 0.3108536254186657,
+      "grad_norm": 0.9648244976997375,
+      "learning_rate": 0.0001490627264366549,
+      "loss": 2.9945,
+      "step": 3434
+    },
+    {
+      "epoch": 0.310944147732416,
+      "grad_norm": 0.937522828578949,
+      "learning_rate": 0.00014903507257118054,
+      "loss": 2.919,
+      "step": 3435
+    },
+    {
+      "epoch": 0.31103467004616636,
+      "grad_norm": 0.8191124200820923,
+      "learning_rate": 0.0001490074137678757,
+      "loss": 2.8935,
+      "step": 3436
+    },
+    {
+      "epoch": 0.3111251923599167,
+      "grad_norm": 0.9017074704170227,
+      "learning_rate": 0.0001489797500295256,
+      "loss": 3.0027,
+      "step": 3437
+    },
+    {
+      "epoch": 0.31121571467366704,
+      "grad_norm": 0.9712589383125305,
+      "learning_rate": 0.00014895208135891604,
+      "loss": 2.9805,
+      "step": 3438
+    },
+    {
+      "epoch": 0.3113062369874174,
+      "grad_norm": 0.9253582954406738,
+      "learning_rate": 0.00014892440775883322,
+      "loss": 3.0006,
+      "step": 3439
+    },
+    {
+      "epoch": 0.3113967593011677,
+      "grad_norm": 0.9078521132469177,
+      "learning_rate": 0.0001488967292320639,
+      "loss": 2.9813,
+      "step": 3440
+    },
+    {
+      "epoch": 0.31148728161491807,
+      "grad_norm": 0.8458201289176941,
+      "learning_rate": 0.00014886904578139526,
+      "loss": 3.0009,
+      "step": 3441
+    },
+    {
+      "epoch": 0.3115778039286684,
+      "grad_norm": 0.9842881560325623,
+      "learning_rate": 0.00014884135740961504,
+      "loss": 3.0177,
+      "step": 3442
+    },
+    {
+      "epoch": 0.31166832624241875,
+      "grad_norm": 0.8262262940406799,
+      "learning_rate": 0.00014881366411951147,
+      "loss": 2.9914,
+      "step": 3443
+    },
+    {
+      "epoch": 0.3117588485561691,
+      "grad_norm": 0.8502604961395264,
+      "learning_rate": 0.0001487859659138733,
+      "loss": 2.9886,
+      "step": 3444
+    },
+    {
+      "epoch": 0.31184937086991943,
+      "grad_norm": 1.1262938976287842,
+      "learning_rate": 0.00014875826279548963,
+      "loss": 3.0346,
+      "step": 3445
+    },
+    {
+      "epoch": 0.31193989318366977,
+      "grad_norm": 1.0363510847091675,
+      "learning_rate": 0.0001487305547671503,
+      "loss": 2.9927,
+      "step": 3446
+    },
+    {
+      "epoch": 0.3120304154974201,
+      "grad_norm": 1.2626910209655762,
+      "learning_rate": 0.00014870284183164537,
+      "loss": 2.9711,
+      "step": 3447
+    },
+    {
+      "epoch": 0.31212093781117045,
+      "grad_norm": 1.3493363857269287,
+      "learning_rate": 0.00014867512399176563,
+      "loss": 3.0042,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3122114601249208,
+      "grad_norm": 0.9481455087661743,
+      "learning_rate": 0.00014864740125030226,
+      "loss": 3.0049,
+      "step": 3449
+    },
+    {
+      "epoch": 0.31230198243867113,
+      "grad_norm": 1.1836961507797241,
+      "learning_rate": 0.00014861967361004687,
+      "loss": 2.9519,
+      "step": 3450
+    },
+    {
+      "epoch": 0.3123925047524215,
+      "grad_norm": 0.9749188423156738,
+      "learning_rate": 0.0001485919410737917,
+      "loss": 2.9565,
+      "step": 3451
+    },
+    {
+      "epoch": 0.3124830270661718,
+      "grad_norm": 0.8708198070526123,
+      "learning_rate": 0.0001485642036443294,
+      "loss": 3.0315,
+      "step": 3452
+    },
+    {
+      "epoch": 0.31257354937992216,
+      "grad_norm": 1.0292015075683594,
+      "learning_rate": 0.00014853646132445311,
+      "loss": 2.9577,
+      "step": 3453
+    },
+    {
+      "epoch": 0.3126640716936725,
+      "grad_norm": 0.9962379336357117,
+      "learning_rate": 0.0001485087141169565,
+      "loss": 2.9129,
+      "step": 3454
+    },
+    {
+      "epoch": 0.31275459400742284,
+      "grad_norm": 0.8681389093399048,
+      "learning_rate": 0.00014848096202463372,
+      "loss": 2.9596,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3128451163211732,
+      "grad_norm": 0.9183157682418823,
+      "learning_rate": 0.00014845320505027938,
+      "loss": 2.9673,
+      "step": 3456
+    },
+    {
+      "epoch": 0.3129356386349235,
+      "grad_norm": 0.9543888568878174,
+      "learning_rate": 0.0001484254431966886,
+      "loss": 3.0068,
+      "step": 3457
+    },
+    {
+      "epoch": 0.31302616094867386,
+      "grad_norm": 0.8184690475463867,
+      "learning_rate": 0.000148397676466657,
+      "loss": 2.9318,
+      "step": 3458
+    },
+    {
+      "epoch": 0.3131166832624242,
+      "grad_norm": 0.9538193941116333,
+      "learning_rate": 0.00014836990486298077,
+      "loss": 3.0353,
+      "step": 3459
+    },
+    {
+      "epoch": 0.31320720557617454,
+      "grad_norm": 0.9387879371643066,
+      "learning_rate": 0.00014834212838845637,
+      "loss": 3.0407,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3132977278899249,
+      "grad_norm": 0.9548669457435608,
+      "learning_rate": 0.00014831434704588102,
+      "loss": 2.9379,
+      "step": 3461
+    },
+    {
+      "epoch": 0.3133882502036752,
+      "grad_norm": 0.9051340222358704,
+      "learning_rate": 0.0001482865608380522,
+      "loss": 2.9099,
+      "step": 3462
+    },
+    {
+      "epoch": 0.31347877251742556,
+      "grad_norm": 0.8997891545295715,
+      "learning_rate": 0.00014825876976776807,
+      "loss": 2.979,
+      "step": 3463
+    },
+    {
+      "epoch": 0.3135692948311759,
+      "grad_norm": 0.9646556377410889,
+      "learning_rate": 0.00014823097383782715,
+      "loss": 3.0219,
+      "step": 3464
+    },
+    {
+      "epoch": 0.31365981714492625,
+      "grad_norm": 0.8742055296897888,
+      "learning_rate": 0.0001482031730510284,
+      "loss": 2.9746,
+      "step": 3465
+    },
+    {
+      "epoch": 0.3137503394586766,
+      "grad_norm": 0.8171219229698181,
+      "learning_rate": 0.00014817536741017152,
+      "loss": 2.9259,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3138408617724269,
+      "grad_norm": 0.9034594893455505,
+      "learning_rate": 0.00014814755691805646,
+      "loss": 2.9577,
+      "step": 3467
+    },
+    {
+      "epoch": 0.31393138408617727,
+      "grad_norm": 0.8481171727180481,
+      "learning_rate": 0.00014811974157748372,
+      "loss": 2.9841,
+      "step": 3468
+    },
+    {
+      "epoch": 0.3140219063999276,
+      "grad_norm": 1.0891640186309814,
+      "learning_rate": 0.0001480919213912543,
+      "loss": 2.9582,
+      "step": 3469
+    },
+    {
+      "epoch": 0.31411242871367795,
+      "grad_norm": 0.9174221754074097,
+      "learning_rate": 0.00014806409636216973,
+      "loss": 2.9967,
+      "step": 3470
+    },
+    {
+      "epoch": 0.31420295102742823,
+      "grad_norm": 0.9914358258247375,
+      "learning_rate": 0.00014803626649303198,
+      "loss": 3.0575,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3142934733411786,
+      "grad_norm": 0.8691591620445251,
+      "learning_rate": 0.0001480084317866435,
+      "loss": 2.9155,
+      "step": 3472
+    },
+    {
+      "epoch": 0.3143839956549289,
+      "grad_norm": 1.0562714338302612,
+      "learning_rate": 0.00014798059224580725,
+      "loss": 2.954,
+      "step": 3473
+    },
+    {
+      "epoch": 0.31447451796867926,
+      "grad_norm": 0.9047598242759705,
+      "learning_rate": 0.00014795274787332666,
+      "loss": 3.0319,
+      "step": 3474
+    },
+    {
+      "epoch": 0.3145650402824296,
+      "grad_norm": 0.9135189652442932,
+      "learning_rate": 0.0001479248986720057,
+      "loss": 2.9636,
+      "step": 3475
+    },
+    {
+      "epoch": 0.31465556259617994,
+      "grad_norm": 0.8667458891868591,
+      "learning_rate": 0.00014789704464464873,
+      "loss": 2.9684,
+      "step": 3476
+    },
+    {
+      "epoch": 0.3147460849099303,
+      "grad_norm": 0.9025455713272095,
+      "learning_rate": 0.0001478691857940607,
+      "loss": 2.9914,
+      "step": 3477
+    },
+    {
+      "epoch": 0.3148366072236806,
+      "grad_norm": 0.8761173486709595,
+      "learning_rate": 0.00014784132212304694,
+      "loss": 2.908,
+      "step": 3478
+    },
+    {
+      "epoch": 0.31492712953743096,
+      "grad_norm": 0.946856677532196,
+      "learning_rate": 0.0001478134536344134,
+      "loss": 2.9857,
+      "step": 3479
+    },
+    {
+      "epoch": 0.3150176518511813,
+      "grad_norm": 0.9196282029151917,
+      "learning_rate": 0.00014778558033096633,
+      "loss": 2.9855,
+      "step": 3480
+    },
+    {
+      "epoch": 0.31510817416493164,
+      "grad_norm": 0.9204413890838623,
+      "learning_rate": 0.00014775770221551264,
+      "loss": 3.0459,
+      "step": 3481
+    },
+    {
+      "epoch": 0.315198696478682,
+      "grad_norm": 0.971332848072052,
+      "learning_rate": 0.00014772981929085964,
+      "loss": 3.0045,
+      "step": 3482
+    },
+    {
+      "epoch": 0.3152892187924323,
+      "grad_norm": 0.8783162236213684,
+      "learning_rate": 0.0001477019315598152,
+      "loss": 2.9549,
+      "step": 3483
+    },
+    {
+      "epoch": 0.31537974110618266,
+      "grad_norm": 0.9069380164146423,
+      "learning_rate": 0.0001476740390251875,
+      "loss": 2.9589,
+      "step": 3484
+    },
+    {
+      "epoch": 0.315470263419933,
+      "grad_norm": 1.194395899772644,
+      "learning_rate": 0.0001476461416897854,
+      "loss": 3.0065,
+      "step": 3485
+    },
+    {
+      "epoch": 0.31556078573368335,
+      "grad_norm": 0.8641033172607422,
+      "learning_rate": 0.00014761823955641813,
+      "loss": 3.0169,
+      "step": 3486
+    },
+    {
+      "epoch": 0.3156513080474337,
+      "grad_norm": 1.0002996921539307,
+      "learning_rate": 0.00014759033262789543,
+      "loss": 2.9833,
+      "step": 3487
+    },
+    {
+      "epoch": 0.31574183036118403,
+      "grad_norm": 0.9187266230583191,
+      "learning_rate": 0.00014756242090702756,
+      "loss": 2.956,
+      "step": 3488
+    },
+    {
+      "epoch": 0.31583235267493437,
+      "grad_norm": 0.9133673906326294,
+      "learning_rate": 0.00014753450439662516,
+      "loss": 3.0216,
+      "step": 3489
+    },
+    {
+      "epoch": 0.3159228749886847,
+      "grad_norm": 0.9382879734039307,
+      "learning_rate": 0.0001475065830994995,
+      "loss": 2.9997,
+      "step": 3490
+    },
+    {
+      "epoch": 0.31601339730243505,
+      "grad_norm": 1.1736315488815308,
+      "learning_rate": 0.00014747865701846222,
+      "loss": 3.2377,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3161039196161854,
+      "grad_norm": 0.9480122327804565,
+      "learning_rate": 0.00014745072615632548,
+      "loss": 2.9493,
+      "step": 3492
+    },
+    {
+      "epoch": 0.31619444192993573,
+      "grad_norm": 0.9033775925636292,
+      "learning_rate": 0.0001474227905159019,
+      "loss": 2.9793,
+      "step": 3493
+    },
+    {
+      "epoch": 0.3162849642436861,
+      "grad_norm": 0.9714937210083008,
+      "learning_rate": 0.00014739485010000465,
+      "loss": 2.9726,
+      "step": 3494
+    },
+    {
+      "epoch": 0.3163754865574364,
+      "grad_norm": 1.0207247734069824,
+      "learning_rate": 0.00014736690491144725,
+      "loss": 2.986,
+      "step": 3495
+    },
+    {
+      "epoch": 0.31646600887118675,
+      "grad_norm": 0.9080100655555725,
+      "learning_rate": 0.0001473389549530438,
+      "loss": 2.9687,
+      "step": 3496
+    },
+    {
+      "epoch": 0.3165565311849371,
+      "grad_norm": 0.9480652809143066,
+      "learning_rate": 0.00014731100022760893,
+      "loss": 2.9902,
+      "step": 3497
+    },
+    {
+      "epoch": 0.31664705349868744,
+      "grad_norm": 0.9133635759353638,
+      "learning_rate": 0.00014728304073795764,
+      "loss": 2.9709,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3167375758124378,
+      "grad_norm": 0.9235483407974243,
+      "learning_rate": 0.00014725507648690543,
+      "loss": 3.0183,
+      "step": 3499
+    },
+    {
+      "epoch": 0.3168280981261881,
+      "grad_norm": 0.9264606833457947,
+      "learning_rate": 0.0001472271074772683,
+      "loss": 2.9804,
+      "step": 3500
+    },
+    {
+      "epoch": 0.31691862043993846,
+      "grad_norm": 0.8027808666229248,
+      "learning_rate": 0.00014719913371186275,
+      "loss": 2.9111,
+      "step": 3501
+    },
+    {
+      "epoch": 0.3170091427536888,
+      "grad_norm": 0.8905692100524902,
+      "learning_rate": 0.00014717115519350567,
+      "loss": 2.9191,
+      "step": 3502
+    },
+    {
+      "epoch": 0.31709966506743914,
+      "grad_norm": 1.0128405094146729,
+      "learning_rate": 0.00014714317192501461,
+      "loss": 2.9664,
+      "step": 3503
+    },
+    {
+      "epoch": 0.3171901873811895,
+      "grad_norm": 0.8433849811553955,
+      "learning_rate": 0.00014711518390920743,
+      "loss": 2.9312,
+      "step": 3504
+    },
+    {
+      "epoch": 0.3172807096949398,
+      "grad_norm": 0.9650546908378601,
+      "learning_rate": 0.00014708719114890252,
+      "loss": 2.9772,
+      "step": 3505
+    },
+    {
+      "epoch": 0.31737123200869016,
+      "grad_norm": 0.8981730341911316,
+      "learning_rate": 0.0001470591936469187,
+      "loss": 2.986,
+      "step": 3506
+    },
+    {
+      "epoch": 0.3174617543224405,
+      "grad_norm": 0.8478082418441772,
+      "learning_rate": 0.00014703119140607542,
+      "loss": 3.0154,
+      "step": 3507
+    },
+    {
+      "epoch": 0.31755227663619084,
+      "grad_norm": 0.9572350382804871,
+      "learning_rate": 0.00014700318442919242,
+      "loss": 3.0304,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3176427989499412,
+      "grad_norm": 0.8729496598243713,
+      "learning_rate": 0.00014697517271909007,
+      "loss": 2.9723,
+      "step": 3509
+    },
+    {
+      "epoch": 0.3177333212636915,
+      "grad_norm": 0.9278503060340881,
+      "learning_rate": 0.00014694715627858908,
+      "loss": 2.9353,
+      "step": 3510
+    },
+    {
+      "epoch": 0.31782384357744187,
+      "grad_norm": 0.9096142649650574,
+      "learning_rate": 0.00014691913511051077,
+      "loss": 2.9696,
+      "step": 3511
+    },
+    {
+      "epoch": 0.31791436589119215,
+      "grad_norm": 1.0005139112472534,
+      "learning_rate": 0.0001468911092176768,
+      "loss": 2.9708,
+      "step": 3512
+    },
+    {
+      "epoch": 0.3180048882049425,
+      "grad_norm": 0.8990221619606018,
+      "learning_rate": 0.00014686307860290945,
+      "loss": 2.9606,
+      "step": 3513
+    },
+    {
+      "epoch": 0.31809541051869283,
+      "grad_norm": 0.8603647351264954,
+      "learning_rate": 0.00014683504326903134,
+      "loss": 3.0335,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3181859328324432,
+      "grad_norm": 0.9636383652687073,
+      "learning_rate": 0.00014680700321886568,
+      "loss": 2.9187,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3182764551461935,
+      "grad_norm": 0.9207581877708435,
+      "learning_rate": 0.00014677895845523607,
+      "loss": 2.9667,
+      "step": 3516
+    },
+    {
+      "epoch": 0.31836697745994386,
+      "grad_norm": 0.8631317019462585,
+      "learning_rate": 0.0001467509089809666,
+      "loss": 2.9814,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3184574997736942,
+      "grad_norm": 0.836303174495697,
+      "learning_rate": 0.0001467228547988819,
+      "loss": 2.9347,
+      "step": 3518
+    },
+    {
+      "epoch": 0.31854802208744454,
+      "grad_norm": 0.8611830472946167,
+      "learning_rate": 0.00014669479591180695,
+      "loss": 2.9619,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3186385444011949,
+      "grad_norm": 0.8925729990005493,
+      "learning_rate": 0.00014666673232256738,
+      "loss": 2.9209,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3187290667149452,
+      "grad_norm": 0.8913902640342712,
+      "learning_rate": 0.00014663866403398913,
+      "loss": 2.9936,
+      "step": 3521
+    },
+    {
+      "epoch": 0.31881958902869556,
+      "grad_norm": 0.8865054249763489,
+      "learning_rate": 0.00014661059104889866,
+      "loss": 3.0011,
+      "step": 3522
+    },
+    {
+      "epoch": 0.3189101113424459,
+      "grad_norm": 0.8239423632621765,
+      "learning_rate": 0.000146582513370123,
+      "loss": 2.9246,
+      "step": 3523
+    },
+    {
+      "epoch": 0.31900063365619624,
+      "grad_norm": 0.9283320307731628,
+      "learning_rate": 0.00014655443100048952,
+      "loss": 2.9693,
+      "step": 3524
+    },
+    {
+      "epoch": 0.3190911559699466,
+      "grad_norm": 0.88014155626297,
+      "learning_rate": 0.00014652634394282608,
+      "loss": 2.8746,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3191816782836969,
+      "grad_norm": 1.0238566398620605,
+      "learning_rate": 0.00014649825219996106,
+      "loss": 2.9485,
+      "step": 3526
+    },
+    {
+      "epoch": 0.31927220059744726,
+      "grad_norm": 0.8470163941383362,
+      "learning_rate": 0.00014647015577472333,
+      "loss": 2.9672,
+      "step": 3527
+    },
+    {
+      "epoch": 0.3193627229111976,
+      "grad_norm": 0.8591601252555847,
+      "learning_rate": 0.0001464420546699422,
+      "loss": 2.8937,
+      "step": 3528
+    },
+    {
+      "epoch": 0.31945324522494795,
+      "grad_norm": 0.9024593830108643,
+      "learning_rate": 0.00014641394888844744,
+      "loss": 2.9542,
+      "step": 3529
+    },
+    {
+      "epoch": 0.3195437675386983,
+      "grad_norm": 0.901153564453125,
+      "learning_rate": 0.00014638583843306927,
+      "loss": 3.0101,
+      "step": 3530
+    },
+    {
+      "epoch": 0.3196342898524486,
+      "grad_norm": 0.8467450141906738,
+      "learning_rate": 0.0001463577233066385,
+      "loss": 2.9043,
+      "step": 3531
+    },
+    {
+      "epoch": 0.31972481216619897,
+      "grad_norm": 1.0799063444137573,
+      "learning_rate": 0.00014632960351198618,
+      "loss": 2.9837,
+      "step": 3532
+    },
+    {
+      "epoch": 0.3198153344799493,
+      "grad_norm": 0.9393935799598694,
+      "learning_rate": 0.0001463014790519441,
+      "loss": 2.9321,
+      "step": 3533
+    },
+    {
+      "epoch": 0.31990585679369965,
+      "grad_norm": 0.9979540109634399,
+      "learning_rate": 0.0001462733499293443,
+      "loss": 3.016,
+      "step": 3534
+    },
+    {
+      "epoch": 0.31999637910745,
+      "grad_norm": 1.284339189529419,
+      "learning_rate": 0.0001462452161470195,
+      "loss": 2.9151,
+      "step": 3535
+    },
+    {
+      "epoch": 0.32008690142120033,
+      "grad_norm": 0.9282693862915039,
+      "learning_rate": 0.00014621707770780268,
+      "loss": 2.9544,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3201774237349507,
+      "grad_norm": 0.9534181356430054,
+      "learning_rate": 0.00014618893461452735,
+      "loss": 2.9185,
+      "step": 3537
+    },
+    {
+      "epoch": 0.320267946048701,
+      "grad_norm": 0.9282907247543335,
+      "learning_rate": 0.0001461607868700276,
+      "loss": 2.8928,
+      "step": 3538
+    },
+    {
+      "epoch": 0.32035846836245135,
+      "grad_norm": 0.8647146224975586,
+      "learning_rate": 0.0001461326344771379,
+      "loss": 2.9864,
+      "step": 3539
+    },
+    {
+      "epoch": 0.3204489906762017,
+      "grad_norm": 0.9271777868270874,
+      "learning_rate": 0.00014610447743869314,
+      "loss": 2.9487,
+      "step": 3540
+    },
+    {
+      "epoch": 0.32053951298995204,
+      "grad_norm": 0.9051327109336853,
+      "learning_rate": 0.00014607631575752877,
+      "loss": 2.8988,
+      "step": 3541
+    },
+    {
+      "epoch": 0.3206300353037024,
+      "grad_norm": 0.9088444113731384,
+      "learning_rate": 0.00014604814943648065,
+      "loss": 2.9941,
+      "step": 3542
+    },
+    {
+      "epoch": 0.3207205576174527,
+      "grad_norm": 0.8992854952812195,
+      "learning_rate": 0.00014601997847838518,
+      "loss": 2.9576,
+      "step": 3543
+    },
+    {
+      "epoch": 0.32081107993120306,
+      "grad_norm": 0.9406589865684509,
+      "learning_rate": 0.0001459918028860791,
+      "loss": 2.9269,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3209016022449534,
+      "grad_norm": 0.821314811706543,
+      "learning_rate": 0.00014596362266239973,
+      "loss": 2.9641,
+      "step": 3545
+    },
+    {
+      "epoch": 0.32099212455870374,
+      "grad_norm": 0.873015284538269,
+      "learning_rate": 0.00014593543781018484,
+      "loss": 2.9098,
+      "step": 3546
+    },
+    {
+      "epoch": 0.3210826468724541,
+      "grad_norm": 0.8742039203643799,
+      "learning_rate": 0.0001459072483322726,
+      "loss": 3.0273,
+      "step": 3547
+    },
+    {
+      "epoch": 0.3211731691862044,
+      "grad_norm": 0.861594021320343,
+      "learning_rate": 0.0001458790542315017,
+      "loss": 2.9947,
+      "step": 3548
+    },
+    {
+      "epoch": 0.32126369149995476,
+      "grad_norm": 0.8861513733863831,
+      "learning_rate": 0.00014585085551071134,
+      "loss": 2.967,
+      "step": 3549
+    },
+    {
+      "epoch": 0.3213542138137051,
+      "grad_norm": 0.8987154960632324,
+      "learning_rate": 0.00014582265217274104,
+      "loss": 2.9136,
+      "step": 3550
+    },
+    {
+      "epoch": 0.32144473612745544,
+      "grad_norm": 0.8634107708930969,
+      "learning_rate": 0.00014579444422043093,
+      "loss": 2.9147,
+      "step": 3551
+    },
+    {
+      "epoch": 0.3215352584412058,
+      "grad_norm": 0.9557790160179138,
+      "learning_rate": 0.00014576623165662157,
+      "loss": 3.0121,
+      "step": 3552
+    },
+    {
+      "epoch": 0.32162578075495607,
+      "grad_norm": 0.8912983536720276,
+      "learning_rate": 0.00014573801448415389,
+      "loss": 2.9428,
+      "step": 3553
+    },
+    {
+      "epoch": 0.3217163030687064,
+      "grad_norm": 0.8403054475784302,
+      "learning_rate": 0.00014570979270586945,
+      "loss": 2.917,
+      "step": 3554
+    },
+    {
+      "epoch": 0.32180682538245675,
+      "grad_norm": 0.8585752844810486,
+      "learning_rate": 0.00014568156632461008,
+      "loss": 2.9762,
+      "step": 3555
+    },
+    {
+      "epoch": 0.3218973476962071,
+      "grad_norm": 0.9195684790611267,
+      "learning_rate": 0.00014565333534321826,
+      "loss": 2.8837,
+      "step": 3556
+    },
+    {
+      "epoch": 0.32198787000995743,
+      "grad_norm": 0.8550272583961487,
+      "learning_rate": 0.00014562509976453683,
+      "loss": 2.9835,
+      "step": 3557
+    },
+    {
+      "epoch": 0.3220783923237078,
+      "grad_norm": 0.8897809982299805,
+      "learning_rate": 0.00014559685959140907,
+      "loss": 2.8815,
+      "step": 3558
+    },
+    {
+      "epoch": 0.3221689146374581,
+      "grad_norm": 0.8674122095108032,
+      "learning_rate": 0.00014556861482667883,
+      "loss": 2.9212,
+      "step": 3559
+    },
+    {
+      "epoch": 0.32225943695120846,
+      "grad_norm": 0.8612362146377563,
+      "learning_rate": 0.00014554036547319033,
+      "loss": 2.9564,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3223499592649588,
+      "grad_norm": 0.8690678477287292,
+      "learning_rate": 0.00014551211153378824,
+      "loss": 3.0016,
+      "step": 3561
+    },
+    {
+      "epoch": 0.32244048157870914,
+      "grad_norm": 0.9328231811523438,
+      "learning_rate": 0.0001454838530113178,
+      "loss": 3.0187,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3225310038924595,
+      "grad_norm": 0.8388855457305908,
+      "learning_rate": 0.00014545558990862456,
+      "loss": 2.9387,
+      "step": 3563
+    },
+    {
+      "epoch": 0.3226215262062098,
+      "grad_norm": 0.806565523147583,
+      "learning_rate": 0.00014542732222855469,
+      "loss": 2.9359,
+      "step": 3564
+    },
+    {
+      "epoch": 0.32271204851996016,
+      "grad_norm": 0.892017126083374,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 2.8981,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3228025708337105,
+      "grad_norm": 0.837681770324707,
+      "learning_rate": 0.00014537077314767162,
+      "loss": 2.9683,
+      "step": 3566
+    },
+    {
+      "epoch": 0.32289309314746084,
+      "grad_norm": 0.8649284839630127,
+      "learning_rate": 0.00014534249175255295,
+      "loss": 2.9524,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3229836154612112,
+      "grad_norm": 0.842377245426178,
+      "learning_rate": 0.00014531420579144656,
+      "loss": 2.9824,
+      "step": 3568
+    },
+    {
+      "epoch": 0.3230741377749615,
+      "grad_norm": 0.8362056016921997,
+      "learning_rate": 0.0001452859152672009,
+      "loss": 2.9602,
+      "step": 3569
+    },
+    {
+      "epoch": 0.32316466008871186,
+      "grad_norm": 0.8706377148628235,
+      "learning_rate": 0.00014525762018266483,
+      "loss": 3.0274,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3232551824024622,
+      "grad_norm": 0.8478174805641174,
+      "learning_rate": 0.00014522932054068766,
+      "loss": 2.9577,
+      "step": 3571
+    },
+    {
+      "epoch": 0.32334570471621255,
+      "grad_norm": 0.8531762957572937,
+      "learning_rate": 0.0001452010163441191,
+      "loss": 2.9567,
+      "step": 3572
+    },
+    {
+      "epoch": 0.3234362270299629,
+      "grad_norm": 0.9084886908531189,
+      "learning_rate": 0.00014517270759580947,
+      "loss": 2.9484,
+      "step": 3573
+    },
+    {
+      "epoch": 0.3235267493437132,
+      "grad_norm": 0.803013026714325,
+      "learning_rate": 0.00014514439429860943,
+      "loss": 2.9553,
+      "step": 3574
+    },
+    {
+      "epoch": 0.32361727165746357,
+      "grad_norm": 0.8305002450942993,
+      "learning_rate": 0.0001451160764553701,
+      "loss": 2.9813,
+      "step": 3575
+    },
+    {
+      "epoch": 0.3237077939712139,
+      "grad_norm": 0.8824440240859985,
+      "learning_rate": 0.00014508775406894307,
+      "loss": 3.0241,
+      "step": 3576
+    },
+    {
+      "epoch": 0.32379831628496425,
+      "grad_norm": 0.8341369032859802,
+      "learning_rate": 0.0001450594271421805,
+      "loss": 2.9415,
+      "step": 3577
+    },
+    {
+      "epoch": 0.3238888385987146,
+      "grad_norm": 0.8482344150543213,
+      "learning_rate": 0.00014503109567793481,
+      "loss": 2.9265,
+      "step": 3578
+    },
+    {
+      "epoch": 0.32397936091246493,
+      "grad_norm": 0.8310097455978394,
+      "learning_rate": 0.00014500275967905903,
+      "loss": 2.9018,
+      "step": 3579
+    },
+    {
+      "epoch": 0.32406988322621527,
+      "grad_norm": 0.8658933639526367,
+      "learning_rate": 0.0001449744191484066,
+      "loss": 3.0183,
+      "step": 3580
+    },
+    {
+      "epoch": 0.3241604055399656,
+      "grad_norm": 0.8318408131599426,
+      "learning_rate": 0.00014494607408883136,
+      "loss": 2.9002,
+      "step": 3581
+    },
+    {
+      "epoch": 0.32425092785371595,
+      "grad_norm": 0.9183574914932251,
+      "learning_rate": 0.00014491772450318772,
+      "loss": 2.9481,
+      "step": 3582
+    },
+    {
+      "epoch": 0.3243414501674663,
+      "grad_norm": 0.9491264820098877,
+      "learning_rate": 0.0001448893703943304,
+      "loss": 2.9774,
+      "step": 3583
+    },
+    {
+      "epoch": 0.32443197248121664,
+      "grad_norm": 0.9588972330093384,
+      "learning_rate": 0.00014486101176511474,
+      "loss": 2.9391,
+      "step": 3584
+    },
+    {
+      "epoch": 0.324522494794967,
+      "grad_norm": 0.8931008577346802,
+      "learning_rate": 0.00014483264861839644,
+      "loss": 2.9383,
+      "step": 3585
+    },
+    {
+      "epoch": 0.3246130171087173,
+      "grad_norm": 1.0485310554504395,
+      "learning_rate": 0.00014480428095703165,
+      "loss": 2.9579,
+      "step": 3586
+    },
+    {
+      "epoch": 0.32470353942246766,
+      "grad_norm": 0.9095520973205566,
+      "learning_rate": 0.00014477590878387696,
+      "loss": 2.9845,
+      "step": 3587
+    },
+    {
+      "epoch": 0.324794061736218,
+      "grad_norm": 0.9263955950737,
+      "learning_rate": 0.00014474753210178953,
+      "loss": 2.9714,
+      "step": 3588
+    },
+    {
+      "epoch": 0.32488458404996834,
+      "grad_norm": 0.8608377575874329,
+      "learning_rate": 0.00014471915091362683,
+      "loss": 2.9234,
+      "step": 3589
+    },
+    {
+      "epoch": 0.3249751063637187,
+      "grad_norm": 0.8173804879188538,
+      "learning_rate": 0.0001446907652222468,
+      "loss": 2.9445,
+      "step": 3590
+    },
+    {
+      "epoch": 0.325065628677469,
+      "grad_norm": 0.9364265203475952,
+      "learning_rate": 0.000144662375030508,
+      "loss": 2.9693,
+      "step": 3591
+    },
+    {
+      "epoch": 0.32515615099121936,
+      "grad_norm": 0.8866042494773865,
+      "learning_rate": 0.0001446339803412692,
+      "loss": 2.9631,
+      "step": 3592
+    },
+    {
+      "epoch": 0.3252466733049697,
+      "grad_norm": 0.9168328642845154,
+      "learning_rate": 0.00014460558115738986,
+      "loss": 2.9383,
+      "step": 3593
+    },
+    {
+      "epoch": 0.32533719561872,
+      "grad_norm": 1.059483528137207,
+      "learning_rate": 0.0001445771774817297,
+      "loss": 2.9121,
+      "step": 3594
+    },
+    {
+      "epoch": 0.32542771793247033,
+      "grad_norm": 0.8932848572731018,
+      "learning_rate": 0.00014454876931714895,
+      "loss": 2.8651,
+      "step": 3595
+    },
+    {
+      "epoch": 0.32551824024622067,
+      "grad_norm": 1.0029690265655518,
+      "learning_rate": 0.0001445203566665084,
+      "loss": 2.9751,
+      "step": 3596
+    },
+    {
+      "epoch": 0.325608762559971,
+      "grad_norm": 0.8788452744483948,
+      "learning_rate": 0.00014449193953266912,
+      "loss": 2.9459,
+      "step": 3597
+    },
+    {
+      "epoch": 0.32569928487372135,
+      "grad_norm": 1.2407371997833252,
+      "learning_rate": 0.00014446351791849276,
+      "loss": 3.045,
+      "step": 3598
+    },
+    {
+      "epoch": 0.3257898071874717,
+      "grad_norm": 0.8738275766372681,
+      "learning_rate": 0.00014443509182684135,
+      "loss": 2.9144,
+      "step": 3599
+    },
+    {
+      "epoch": 0.32588032950122203,
+      "grad_norm": 1.4616142511367798,
+      "learning_rate": 0.00014440666126057744,
+      "loss": 2.9866,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3259708518149724,
+      "grad_norm": 0.9846639633178711,
+      "learning_rate": 0.00014437822622256392,
+      "loss": 2.8977,
+      "step": 3601
+    },
+    {
+      "epoch": 0.3260613741287227,
+      "grad_norm": 0.8873237371444702,
+      "learning_rate": 0.00014434978671566424,
+      "loss": 2.944,
+      "step": 3602
+    },
+    {
+      "epoch": 0.32615189644247305,
+      "grad_norm": 1.063632607460022,
+      "learning_rate": 0.00014432134274274224,
+      "loss": 2.9393,
+      "step": 3603
+    },
+    {
+      "epoch": 0.3262424187562234,
+      "grad_norm": 0.8502988815307617,
+      "learning_rate": 0.00014429289430666227,
+      "loss": 2.9085,
+      "step": 3604
+    },
+    {
+      "epoch": 0.32633294106997374,
+      "grad_norm": 0.8328250646591187,
+      "learning_rate": 0.00014426444141028906,
+      "loss": 2.892,
+      "step": 3605
+    },
+    {
+      "epoch": 0.3264234633837241,
+      "grad_norm": 0.9032594561576843,
+      "learning_rate": 0.00014423598405648776,
+      "loss": 2.9523,
+      "step": 3606
+    },
+    {
+      "epoch": 0.3265139856974744,
+      "grad_norm": 0.8867715001106262,
+      "learning_rate": 0.0001442075222481241,
+      "loss": 2.913,
+      "step": 3607
+    },
+    {
+      "epoch": 0.32660450801122476,
+      "grad_norm": 0.9769021272659302,
+      "learning_rate": 0.0001441790559880642,
+      "loss": 3.0404,
+      "step": 3608
+    },
+    {
+      "epoch": 0.3266950303249751,
+      "grad_norm": 0.9831045866012573,
+      "learning_rate": 0.00014415058527917452,
+      "loss": 2.949,
+      "step": 3609
+    },
+    {
+      "epoch": 0.32678555263872544,
+      "grad_norm": 0.8027966618537903,
+      "learning_rate": 0.00014412211012432212,
+      "loss": 2.9648,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3268760749524758,
+      "grad_norm": 0.9449635744094849,
+      "learning_rate": 0.00014409363052637447,
+      "loss": 2.9716,
+      "step": 3611
+    },
+    {
+      "epoch": 0.3269665972662261,
+      "grad_norm": 0.9377633929252625,
+      "learning_rate": 0.00014406514648819942,
+      "loss": 2.9289,
+      "step": 3612
+    },
+    {
+      "epoch": 0.32705711957997646,
+      "grad_norm": 0.971062183380127,
+      "learning_rate": 0.0001440366580126653,
+      "loss": 3.0385,
+      "step": 3613
+    },
+    {
+      "epoch": 0.3271476418937268,
+      "grad_norm": 0.9378071427345276,
+      "learning_rate": 0.00014400816510264092,
+      "loss": 2.9413,
+      "step": 3614
+    },
+    {
+      "epoch": 0.32723816420747714,
+      "grad_norm": 0.8906871676445007,
+      "learning_rate": 0.00014397966776099556,
+      "loss": 2.9619,
+      "step": 3615
+    },
+    {
+      "epoch": 0.3273286865212275,
+      "grad_norm": 0.8794304728507996,
+      "learning_rate": 0.0001439511659905988,
+      "loss": 2.9532,
+      "step": 3616
+    },
+    {
+      "epoch": 0.3274192088349778,
+      "grad_norm": 0.9370628595352173,
+      "learning_rate": 0.0001439226597943209,
+      "loss": 2.978,
+      "step": 3617
+    },
+    {
+      "epoch": 0.32750973114872817,
+      "grad_norm": 0.9121912717819214,
+      "learning_rate": 0.0001438941491750323,
+      "loss": 2.938,
+      "step": 3618
+    },
+    {
+      "epoch": 0.3276002534624785,
+      "grad_norm": 0.8118641376495361,
+      "learning_rate": 0.00014386563413560413,
+      "loss": 2.986,
+      "step": 3619
+    },
+    {
+      "epoch": 0.32769077577622885,
+      "grad_norm": 0.9029416441917419,
+      "learning_rate": 0.00014383711467890774,
+      "loss": 2.9635,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3277812980899792,
+      "grad_norm": 0.8763449192047119,
+      "learning_rate": 0.00014380859080781515,
+      "loss": 2.9133,
+      "step": 3621
+    },
+    {
+      "epoch": 0.32787182040372953,
+      "grad_norm": 0.881456732749939,
+      "learning_rate": 0.00014378006252519865,
+      "loss": 2.9245,
+      "step": 3622
+    },
+    {
+      "epoch": 0.32796234271747987,
+      "grad_norm": 0.8821005821228027,
+      "learning_rate": 0.00014375152983393106,
+      "loss": 2.9826,
+      "step": 3623
+    },
+    {
+      "epoch": 0.3280528650312302,
+      "grad_norm": 0.9180141091346741,
+      "learning_rate": 0.00014372299273688558,
+      "loss": 2.9819,
+      "step": 3624
+    },
+    {
+      "epoch": 0.32814338734498055,
+      "grad_norm": 0.9080764651298523,
+      "learning_rate": 0.00014369445123693596,
+      "loss": 3.0095,
+      "step": 3625
+    },
+    {
+      "epoch": 0.3282339096587309,
+      "grad_norm": 0.903130054473877,
+      "learning_rate": 0.0001436659053369563,
+      "loss": 2.902,
+      "step": 3626
+    },
+    {
+      "epoch": 0.32832443197248123,
+      "grad_norm": 0.8723325729370117,
+      "learning_rate": 0.00014363735503982114,
+      "loss": 2.9421,
+      "step": 3627
+    },
+    {
+      "epoch": 0.3284149542862316,
+      "grad_norm": 0.9515482187271118,
+      "learning_rate": 0.00014360880034840554,
+      "loss": 2.9796,
+      "step": 3628
+    },
+    {
+      "epoch": 0.3285054765999819,
+      "grad_norm": 0.8937515616416931,
+      "learning_rate": 0.00014358024126558493,
+      "loss": 2.9437,
+      "step": 3629
+    },
+    {
+      "epoch": 0.32859599891373226,
+      "grad_norm": 0.8733628392219543,
+      "learning_rate": 0.00014355167779423524,
+      "loss": 2.9753,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3286865212274826,
+      "grad_norm": 0.8773425221443176,
+      "learning_rate": 0.00014352310993723277,
+      "loss": 2.9253,
+      "step": 3631
+    },
+    {
+      "epoch": 0.32877704354123294,
+      "grad_norm": 0.8728827238082886,
+      "learning_rate": 0.0001434945376974543,
+      "loss": 2.9234,
+      "step": 3632
+    },
+    {
+      "epoch": 0.3288675658549833,
+      "grad_norm": 0.8740129470825195,
+      "learning_rate": 0.0001434659610777771,
+      "loss": 2.98,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3289580881687336,
+      "grad_norm": 0.831460177898407,
+      "learning_rate": 0.0001434373800810788,
+      "loss": 2.9551,
+      "step": 3634
+    },
+    {
+      "epoch": 0.3290486104824839,
+      "grad_norm": 0.8963636755943298,
+      "learning_rate": 0.0001434087947102375,
+      "loss": 2.9985,
+      "step": 3635
+    },
+    {
+      "epoch": 0.32913913279623425,
+      "grad_norm": 0.9177209138870239,
+      "learning_rate": 0.0001433802049681318,
+      "loss": 2.9544,
+      "step": 3636
+    },
+    {
+      "epoch": 0.3292296551099846,
+      "grad_norm": 0.8798718452453613,
+      "learning_rate": 0.00014335161085764062,
+      "loss": 2.9751,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3293201774237349,
+      "grad_norm": 0.8863023519515991,
+      "learning_rate": 0.00014332301238164342,
+      "loss": 2.8911,
+      "step": 3638
+    },
+    {
+      "epoch": 0.32941069973748527,
+      "grad_norm": 0.8351691961288452,
+      "learning_rate": 0.00014329440954302005,
+      "loss": 2.9239,
+      "step": 3639
+    },
+    {
+      "epoch": 0.3295012220512356,
+      "grad_norm": 0.8428190350532532,
+      "learning_rate": 0.00014326580234465085,
+      "loss": 2.9803,
+      "step": 3640
+    },
+    {
+      "epoch": 0.32959174436498595,
+      "grad_norm": 0.8503945469856262,
+      "learning_rate": 0.0001432371907894165,
+      "loss": 2.9194,
+      "step": 3641
+    },
+    {
+      "epoch": 0.3296822666787363,
+      "grad_norm": 0.9099516868591309,
+      "learning_rate": 0.00014320857488019824,
+      "loss": 2.9307,
+      "step": 3642
+    },
+    {
+      "epoch": 0.32977278899248663,
+      "grad_norm": 0.9085094928741455,
+      "learning_rate": 0.00014317995461987767,
+      "loss": 2.9746,
+      "step": 3643
+    },
+    {
+      "epoch": 0.329863311306237,
+      "grad_norm": 0.9462329149246216,
+      "learning_rate": 0.00014315133001133686,
+      "loss": 2.9018,
+      "step": 3644
+    },
+    {
+      "epoch": 0.3299538336199873,
+      "grad_norm": 0.9889357686042786,
+      "learning_rate": 0.0001431227010574583,
+      "loss": 3.0249,
+      "step": 3645
+    },
+    {
+      "epoch": 0.33004435593373765,
+      "grad_norm": 0.8533375859260559,
+      "learning_rate": 0.0001430940677611249,
+      "loss": 2.8802,
+      "step": 3646
+    },
+    {
+      "epoch": 0.330134878247488,
+      "grad_norm": 0.9745096564292908,
+      "learning_rate": 0.00014306543012522005,
+      "loss": 2.9591,
+      "step": 3647
+    },
+    {
+      "epoch": 0.33022540056123834,
+      "grad_norm": 0.843706488609314,
+      "learning_rate": 0.0001430367881526276,
+      "loss": 3.0582,
+      "step": 3648
+    },
+    {
+      "epoch": 0.3303159228749887,
+      "grad_norm": 1.0291938781738281,
+      "learning_rate": 0.00014300814184623172,
+      "loss": 2.9939,
+      "step": 3649
+    },
+    {
+      "epoch": 0.330406445188739,
+      "grad_norm": 1.0679194927215576,
+      "learning_rate": 0.00014297949120891718,
+      "loss": 2.9759,
+      "step": 3650
+    },
+    {
+      "epoch": 0.33049696750248936,
+      "grad_norm": 0.8248839974403381,
+      "learning_rate": 0.00014295083624356903,
+      "loss": 2.9712,
+      "step": 3651
+    },
+    {
+      "epoch": 0.3305874898162397,
+      "grad_norm": 1.0173978805541992,
+      "learning_rate": 0.00014292217695307285,
+      "loss": 2.9606,
+      "step": 3652
+    },
+    {
+      "epoch": 0.33067801212999004,
+      "grad_norm": 0.9261618852615356,
+      "learning_rate": 0.0001428935133403146,
+      "loss": 2.9399,
+      "step": 3653
+    },
+    {
+      "epoch": 0.3307685344437404,
+      "grad_norm": 1.108538031578064,
+      "learning_rate": 0.00014286484540818078,
+      "loss": 3.0039,
+      "step": 3654
+    },
+    {
+      "epoch": 0.3308590567574907,
+      "grad_norm": 0.9498209953308105,
+      "learning_rate": 0.00014283617315955814,
+      "loss": 2.9224,
+      "step": 3655
+    },
+    {
+      "epoch": 0.33094957907124106,
+      "grad_norm": 0.8508312106132507,
+      "learning_rate": 0.00014280749659733406,
+      "loss": 2.8872,
+      "step": 3656
+    },
+    {
+      "epoch": 0.3310401013849914,
+      "grad_norm": 0.9958304762840271,
+      "learning_rate": 0.00014277881572439627,
+      "loss": 2.9212,
+      "step": 3657
+    },
+    {
+      "epoch": 0.33113062369874174,
+      "grad_norm": 0.8798844814300537,
+      "learning_rate": 0.00014275013054363287,
+      "loss": 2.9737,
+      "step": 3658
+    },
+    {
+      "epoch": 0.3312211460124921,
+      "grad_norm": 0.913406252861023,
+      "learning_rate": 0.0001427214410579325,
+      "loss": 2.8793,
+      "step": 3659
+    },
+    {
+      "epoch": 0.3313116683262424,
+      "grad_norm": 0.8229892253875732,
+      "learning_rate": 0.0001426927472701842,
+      "loss": 3.0067,
+      "step": 3660
+    },
+    {
+      "epoch": 0.33140219063999277,
+      "grad_norm": 0.9121760725975037,
+      "learning_rate": 0.00014266404918327743,
+      "loss": 2.9986,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3314927129537431,
+      "grad_norm": 0.875041127204895,
+      "learning_rate": 0.00014263534680010204,
+      "loss": 2.934,
+      "step": 3662
+    },
+    {
+      "epoch": 0.33158323526749345,
+      "grad_norm": 0.8997043967247009,
+      "learning_rate": 0.00014260664012354842,
+      "loss": 2.9247,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3316737575812438,
+      "grad_norm": 0.8942994475364685,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 2.9651,
+      "step": 3664
+    },
+    {
+      "epoch": 0.33176427989499413,
+      "grad_norm": 0.8819488286972046,
+      "learning_rate": 0.00014254921390186986,
+      "loss": 2.9339,
+      "step": 3665
+    },
+    {
+      "epoch": 0.33185480220874447,
+      "grad_norm": 0.8902394771575928,
+      "learning_rate": 0.00014252049436252776,
+      "loss": 2.8851,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3319453245224948,
+      "grad_norm": 0.876203179359436,
+      "learning_rate": 0.00014249177054137305,
+      "loss": 2.9535,
+      "step": 3667
+    },
+    {
+      "epoch": 0.33203584683624515,
+      "grad_norm": 0.8925275802612305,
+      "learning_rate": 0.00014246304244129818,
+      "loss": 2.9907,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3321263691499955,
+      "grad_norm": 0.8686583042144775,
+      "learning_rate": 0.00014243431006519613,
+      "loss": 2.9287,
+      "step": 3669
+    },
+    {
+      "epoch": 0.33221689146374583,
+      "grad_norm": 0.939652681350708,
+      "learning_rate": 0.00014240557341596018,
+      "loss": 2.9866,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3323074137774962,
+      "grad_norm": 1.0006009340286255,
+      "learning_rate": 0.00014237683249648418,
+      "loss": 3.0263,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3323979360912465,
+      "grad_norm": 0.8614761233329773,
+      "learning_rate": 0.0001423480873096623,
+      "loss": 2.9568,
+      "step": 3672
+    },
+    {
+      "epoch": 0.33248845840499686,
+      "grad_norm": 0.9231300950050354,
+      "learning_rate": 0.00014231933785838918,
+      "loss": 2.9747,
+      "step": 3673
+    },
+    {
+      "epoch": 0.3325789807187472,
+      "grad_norm": 0.9635558128356934,
+      "learning_rate": 0.0001422905841455599,
+      "loss": 2.927,
+      "step": 3674
+    },
+    {
+      "epoch": 0.33266950303249754,
+      "grad_norm": 0.8838689923286438,
+      "learning_rate": 0.00014226182617406996,
+      "loss": 2.9381,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3327600253462478,
+      "grad_norm": 0.9083494544029236,
+      "learning_rate": 0.00014223306394681528,
+      "loss": 2.9369,
+      "step": 3676
+    },
+    {
+      "epoch": 0.33285054765999816,
+      "grad_norm": 0.9969063997268677,
+      "learning_rate": 0.00014220429746669222,
+      "loss": 2.9549,
+      "step": 3677
+    },
+    {
+      "epoch": 0.3329410699737485,
+      "grad_norm": 0.8796403408050537,
+      "learning_rate": 0.00014217552673659754,
+      "loss": 2.9092,
+      "step": 3678
+    },
+    {
+      "epoch": 0.33303159228749885,
+      "grad_norm": 0.8336253762245178,
+      "learning_rate": 0.00014214675175942847,
+      "loss": 2.9268,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3331221146012492,
+      "grad_norm": 0.9689157009124756,
+      "learning_rate": 0.00014211797253808268,
+      "loss": 2.971,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3332126369149995,
+      "grad_norm": 0.8374845385551453,
+      "learning_rate": 0.00014208918907545817,
+      "loss": 2.9561,
+      "step": 3681
+    },
+    {
+      "epoch": 0.33330315922874987,
+      "grad_norm": 0.9589225053787231,
+      "learning_rate": 0.00014206040137445348,
+      "loss": 2.9519,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3333936815425002,
+      "grad_norm": 0.977763295173645,
+      "learning_rate": 0.00014203160943796755,
+      "loss": 2.9803,
+      "step": 3683
+    },
+    {
+      "epoch": 0.33348420385625055,
+      "grad_norm": 0.8360752463340759,
+      "learning_rate": 0.0001420028132688997,
+      "loss": 2.9539,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3335747261700009,
+      "grad_norm": 0.9216536283493042,
+      "learning_rate": 0.0001419740128701497,
+      "loss": 2.8518,
+      "step": 3685
+    },
+    {
+      "epoch": 0.33366524848375123,
+      "grad_norm": 0.9173371195793152,
+      "learning_rate": 0.00014194520824461771,
+      "loss": 3.0058,
+      "step": 3686
+    },
+    {
+      "epoch": 0.33375577079750157,
+      "grad_norm": 0.841460645198822,
+      "learning_rate": 0.00014191639939520442,
+      "loss": 2.9655,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3338462931112519,
+      "grad_norm": 1.026754379272461,
+      "learning_rate": 0.0001418875863248109,
+      "loss": 2.9974,
+      "step": 3688
+    },
+    {
+      "epoch": 0.33393681542500225,
+      "grad_norm": 0.8212361335754395,
+      "learning_rate": 0.0001418587690363385,
+      "loss": 2.8847,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3340273377387526,
+      "grad_norm": 0.8820496201515198,
+      "learning_rate": 0.00014182994753268927,
+      "loss": 2.938,
+      "step": 3690
+    },
+    {
+      "epoch": 0.33411786005250294,
+      "grad_norm": 1.0030783414840698,
+      "learning_rate": 0.0001418011218167655,
+      "loss": 2.9654,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3342083823662533,
+      "grad_norm": 0.9057952165603638,
+      "learning_rate": 0.00014177229189146984,
+      "loss": 2.8773,
+      "step": 3692
+    },
+    {
+      "epoch": 0.3342989046800036,
+      "grad_norm": 0.8676903247833252,
+      "learning_rate": 0.00014174345775970552,
+      "loss": 2.9207,
+      "step": 3693
+    },
+    {
+      "epoch": 0.33438942699375396,
+      "grad_norm": 0.9278954863548279,
+      "learning_rate": 0.0001417146194243762,
+      "loss": 2.9738,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3344799493075043,
+      "grad_norm": 0.8062622547149658,
+      "learning_rate": 0.0001416857768883858,
+      "loss": 2.8546,
+      "step": 3695
+    },
+    {
+      "epoch": 0.33457047162125464,
+      "grad_norm": 0.8296931385993958,
+      "learning_rate": 0.00014165693015463885,
+      "loss": 2.9236,
+      "step": 3696
+    },
+    {
+      "epoch": 0.334660993935005,
+      "grad_norm": 0.8529160022735596,
+      "learning_rate": 0.00014162807922604012,
+      "loss": 2.9099,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3347515162487553,
+      "grad_norm": 0.876424252986908,
+      "learning_rate": 0.00014159922410549497,
+      "loss": 2.958,
+      "step": 3698
+    },
+    {
+      "epoch": 0.33484203856250566,
+      "grad_norm": 0.920254647731781,
+      "learning_rate": 0.00014157036479590913,
+      "loss": 2.9123,
+      "step": 3699
+    },
+    {
+      "epoch": 0.334932560876256,
+      "grad_norm": 0.9651416540145874,
+      "learning_rate": 0.00014154150130018866,
+      "loss": 2.9118,
+      "step": 3700
+    },
+    {
+      "epoch": 0.33502308319000634,
+      "grad_norm": 0.8106426000595093,
+      "learning_rate": 0.00014151263362124013,
+      "loss": 2.9436,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3351136055037567,
+      "grad_norm": 0.9680935144424438,
+      "learning_rate": 0.00014148376176197056,
+      "loss": 3.0007,
+      "step": 3702
+    },
+    {
+      "epoch": 0.335204127817507,
+      "grad_norm": 0.831698477268219,
+      "learning_rate": 0.0001414548857252873,
+      "loss": 2.8877,
+      "step": 3703
+    },
+    {
+      "epoch": 0.33529465013125737,
+      "grad_norm": 0.8796342015266418,
+      "learning_rate": 0.00014142600551409818,
+      "loss": 2.9198,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3353851724450077,
+      "grad_norm": 0.9696304798126221,
+      "learning_rate": 0.00014139712113131144,
+      "loss": 2.9221,
+      "step": 3705
+    },
+    {
+      "epoch": 0.33547569475875805,
+      "grad_norm": 0.8577775955200195,
+      "learning_rate": 0.00014136823257983577,
+      "loss": 2.9389,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3355662170725084,
+      "grad_norm": 0.9382533431053162,
+      "learning_rate": 0.00014133933986258017,
+      "loss": 2.8726,
+      "step": 3707
+    },
+    {
+      "epoch": 0.33565673938625873,
+      "grad_norm": 1.0489027500152588,
+      "learning_rate": 0.0001413104429824542,
+      "loss": 2.9656,
+      "step": 3708
+    },
+    {
+      "epoch": 0.33574726170000907,
+      "grad_norm": 0.8827459216117859,
+      "learning_rate": 0.0001412815419423677,
+      "loss": 2.9129,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3358377840137594,
+      "grad_norm": 1.0175361633300781,
+      "learning_rate": 0.00014125263674523114,
+      "loss": 2.9383,
+      "step": 3710
+    },
+    {
+      "epoch": 0.33592830632750975,
+      "grad_norm": 1.0674699544906616,
+      "learning_rate": 0.00014122372739395513,
+      "loss": 2.9685,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3360188286412601,
+      "grad_norm": 0.9076451659202576,
+      "learning_rate": 0.0001411948138914509,
+      "loss": 2.942,
+      "step": 3712
+    },
+    {
+      "epoch": 0.33610935095501043,
+      "grad_norm": 0.9945937395095825,
+      "learning_rate": 0.00014116589624063006,
+      "loss": 2.9611,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3361998732687608,
+      "grad_norm": 1.0060627460479736,
+      "learning_rate": 0.0001411369744444046,
+      "loss": 2.9929,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3362903955825111,
+      "grad_norm": 0.8711893558502197,
+      "learning_rate": 0.0001411080485056869,
+      "loss": 2.9307,
+      "step": 3715
+    },
+    {
+      "epoch": 0.33638091789626146,
+      "grad_norm": 0.9552099704742432,
+      "learning_rate": 0.00014107911842738987,
+      "loss": 2.961,
+      "step": 3716
+    },
+    {
+      "epoch": 0.33647144021001174,
+      "grad_norm": 1.0096842050552368,
+      "learning_rate": 0.00014105018421242674,
+      "loss": 2.8867,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3365619625237621,
+      "grad_norm": 0.9284431338310242,
+      "learning_rate": 0.0001410212458637112,
+      "loss": 2.9696,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3366524848375124,
+      "grad_norm": 0.9499095678329468,
+      "learning_rate": 0.00014099230338415728,
+      "loss": 2.9788,
+      "step": 3719
+    },
+    {
+      "epoch": 0.33674300715126276,
+      "grad_norm": 0.8471880555152893,
+      "learning_rate": 0.00014096335677667954,
+      "loss": 2.9692,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3368335294650131,
+      "grad_norm": 0.8633290529251099,
+      "learning_rate": 0.00014093440604419293,
+      "loss": 2.9557,
+      "step": 3721
+    },
+    {
+      "epoch": 0.33692405177876344,
+      "grad_norm": 0.855199933052063,
+      "learning_rate": 0.00014090545118961272,
+      "loss": 2.8918,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3370145740925138,
+      "grad_norm": 0.9451314806938171,
+      "learning_rate": 0.0001408764922158547,
+      "loss": 2.9487,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3371050964062641,
+      "grad_norm": 0.8324365615844727,
+      "learning_rate": 0.00014084752912583504,
+      "loss": 2.894,
+      "step": 3724
+    },
+    {
+      "epoch": 0.33719561872001447,
+      "grad_norm": 0.9008909463882446,
+      "learning_rate": 0.00014081856192247033,
+      "loss": 2.9572,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3372861410337648,
+      "grad_norm": 0.8862847089767456,
+      "learning_rate": 0.0001407895906086775,
+      "loss": 2.9045,
+      "step": 3726
+    },
+    {
+      "epoch": 0.33737666334751515,
+      "grad_norm": 0.8587580919265747,
+      "learning_rate": 0.00014076061518737405,
+      "loss": 2.9436,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3374671856612655,
+      "grad_norm": 0.8281906843185425,
+      "learning_rate": 0.00014073163566147775,
+      "loss": 2.9422,
+      "step": 3728
+    },
+    {
+      "epoch": 0.33755770797501583,
+      "grad_norm": 0.9020146727561951,
+      "learning_rate": 0.00014070265203390686,
+      "loss": 2.9748,
+      "step": 3729
+    },
+    {
+      "epoch": 0.33764823028876617,
+      "grad_norm": 0.9331706166267395,
+      "learning_rate": 0.00014067366430758004,
+      "loss": 2.9166,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3377387526025165,
+      "grad_norm": 0.8349210023880005,
+      "learning_rate": 0.00014064467248541633,
+      "loss": 2.8796,
+      "step": 3731
+    },
+    {
+      "epoch": 0.33782927491626685,
+      "grad_norm": 0.9263269305229187,
+      "learning_rate": 0.00014061567657033518,
+      "loss": 2.9703,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3379197972300172,
+      "grad_norm": 0.9281939268112183,
+      "learning_rate": 0.00014058667656525654,
+      "loss": 2.8827,
+      "step": 3733
+    },
+    {
+      "epoch": 0.33801031954376753,
+      "grad_norm": 0.9000682234764099,
+      "learning_rate": 0.0001405576724731007,
+      "loss": 2.915,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3381008418575179,
+      "grad_norm": 0.8951552510261536,
+      "learning_rate": 0.00014052866429678832,
+      "loss": 2.9386,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3381913641712682,
+      "grad_norm": 0.8973798155784607,
+      "learning_rate": 0.00014049965203924054,
+      "loss": 2.9231,
+      "step": 3736
+    },
+    {
+      "epoch": 0.33828188648501856,
+      "grad_norm": 0.9233877658843994,
+      "learning_rate": 0.00014047063570337895,
+      "loss": 2.9131,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3383724087987689,
+      "grad_norm": 0.7996513843536377,
+      "learning_rate": 0.00014044161529212543,
+      "loss": 2.9451,
+      "step": 3738
+    },
+    {
+      "epoch": 0.33846293111251924,
+      "grad_norm": 0.9365185499191284,
+      "learning_rate": 0.00014041259080840236,
+      "loss": 2.9335,
+      "step": 3739
+    },
+    {
+      "epoch": 0.3385534534262696,
+      "grad_norm": 0.8419017791748047,
+      "learning_rate": 0.00014038356225513248,
+      "loss": 2.9522,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3386439757400199,
+      "grad_norm": 0.8480751514434814,
+      "learning_rate": 0.00014035452963523902,
+      "loss": 2.9226,
+      "step": 3741
+    },
+    {
+      "epoch": 0.33873449805377026,
+      "grad_norm": 0.8409445881843567,
+      "learning_rate": 0.00014032549295164552,
+      "loss": 2.9347,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3388250203675206,
+      "grad_norm": 0.8422466516494751,
+      "learning_rate": 0.00014029645220727595,
+      "loss": 2.8973,
+      "step": 3743
+    },
+    {
+      "epoch": 0.33891554268127094,
+      "grad_norm": 0.7969680428504944,
+      "learning_rate": 0.0001402674074050548,
+      "loss": 2.8855,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3390060649950213,
+      "grad_norm": 0.8092168569564819,
+      "learning_rate": 0.0001402383585479068,
+      "loss": 2.94,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3390965873087716,
+      "grad_norm": 0.8328232765197754,
+      "learning_rate": 0.0001402093056387572,
+      "loss": 2.8787,
+      "step": 3746
+    },
+    {
+      "epoch": 0.33918710962252197,
+      "grad_norm": 0.8564672470092773,
+      "learning_rate": 0.0001401802486805316,
+      "loss": 2.9456,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3392776319362723,
+      "grad_norm": 0.8365373015403748,
+      "learning_rate": 0.00014015118767615606,
+      "loss": 2.933,
+      "step": 3748
+    },
+    {
+      "epoch": 0.33936815425002265,
+      "grad_norm": 0.8401426076889038,
+      "learning_rate": 0.00014012212262855706,
+      "loss": 2.8474,
+      "step": 3749
+    },
+    {
+      "epoch": 0.339458676563773,
+      "grad_norm": 0.8071068525314331,
+      "learning_rate": 0.00014009305354066137,
+      "loss": 2.9493,
+      "step": 3750
+    },
+    {
+      "epoch": 0.33954919887752333,
+      "grad_norm": 0.8695021867752075,
+      "learning_rate": 0.0001400639804153963,
+      "loss": 2.9483,
+      "step": 3751
+    },
+    {
+      "epoch": 0.33963972119127367,
+      "grad_norm": 0.81166672706604,
+      "learning_rate": 0.00014003490325568954,
+      "loss": 2.9202,
+      "step": 3752
+    },
+    {
+      "epoch": 0.339730243505024,
+      "grad_norm": 0.8853387832641602,
+      "learning_rate": 0.00014000582206446905,
+      "loss": 2.976,
+      "step": 3753
+    },
+    {
+      "epoch": 0.33982076581877435,
+      "grad_norm": 0.8278989791870117,
+      "learning_rate": 0.0001399767368446634,
+      "loss": 2.9156,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3399112881325247,
+      "grad_norm": 0.8867419958114624,
+      "learning_rate": 0.00013994764759920142,
+      "loss": 2.9845,
+      "step": 3755
+    },
+    {
+      "epoch": 0.34000181044627503,
+      "grad_norm": 0.8510578870773315,
+      "learning_rate": 0.00013991855433101246,
+      "loss": 2.95,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3400923327600254,
+      "grad_norm": 0.8585345149040222,
+      "learning_rate": 0.00013988945704302615,
+      "loss": 2.8936,
+      "step": 3757
+    },
+    {
+      "epoch": 0.34018285507377566,
+      "grad_norm": 0.854981005191803,
+      "learning_rate": 0.0001398603557381726,
+      "loss": 2.9409,
+      "step": 3758
+    },
+    {
+      "epoch": 0.340273377387526,
+      "grad_norm": 0.8987625241279602,
+      "learning_rate": 0.00013983125041938232,
+      "loss": 2.9962,
+      "step": 3759
+    },
+    {
+      "epoch": 0.34036389970127634,
+      "grad_norm": 0.9021677374839783,
+      "learning_rate": 0.00013980214108958624,
+      "loss": 2.8906,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3404544220150267,
+      "grad_norm": 0.9355793595314026,
+      "learning_rate": 0.0001397730277517156,
+      "loss": 2.9823,
+      "step": 3761
+    },
+    {
+      "epoch": 0.340544944328777,
+      "grad_norm": 0.8212774395942688,
+      "learning_rate": 0.00013974391040870218,
+      "loss": 2.927,
+      "step": 3762
+    },
+    {
+      "epoch": 0.34063546664252736,
+      "grad_norm": 0.9005094170570374,
+      "learning_rate": 0.00013971478906347806,
+      "loss": 2.9531,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3407259889562777,
+      "grad_norm": 1.0443588495254517,
+      "learning_rate": 0.0001396856637189758,
+      "loss": 2.9933,
+      "step": 3764
+    },
+    {
+      "epoch": 0.34081651127002804,
+      "grad_norm": 0.8536207675933838,
+      "learning_rate": 0.00013965653437812825,
+      "loss": 2.9121,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3409070335837784,
+      "grad_norm": 0.8683392405509949,
+      "learning_rate": 0.00013962740104386876,
+      "loss": 2.9358,
+      "step": 3766
+    },
+    {
+      "epoch": 0.3409975558975287,
+      "grad_norm": 0.8488772511482239,
+      "learning_rate": 0.00013959826371913111,
+      "loss": 2.9209,
+      "step": 3767
+    },
+    {
+      "epoch": 0.34108807821127907,
+      "grad_norm": 0.9025128483772278,
+      "learning_rate": 0.0001395691224068494,
+      "loss": 2.9691,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3411786005250294,
+      "grad_norm": 0.878285825252533,
+      "learning_rate": 0.0001395399771099581,
+      "loss": 2.9542,
+      "step": 3769
+    },
+    {
+      "epoch": 0.34126912283877975,
+      "grad_norm": 0.8517298698425293,
+      "learning_rate": 0.0001395108278313922,
+      "loss": 2.9952,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3413596451525301,
+      "grad_norm": 0.919712483882904,
+      "learning_rate": 0.00013948167457408702,
+      "loss": 3.0014,
+      "step": 3771
+    },
+    {
+      "epoch": 0.34145016746628043,
+      "grad_norm": 0.8718852400779724,
+      "learning_rate": 0.00013945251734097828,
+      "loss": 2.9527,
+      "step": 3772
+    },
+    {
+      "epoch": 0.34154068978003077,
+      "grad_norm": 0.8577459454536438,
+      "learning_rate": 0.00013942335613500214,
+      "loss": 2.918,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3416312120937811,
+      "grad_norm": 0.8327313661575317,
+      "learning_rate": 0.00013939419095909512,
+      "loss": 2.9399,
+      "step": 3774
+    },
+    {
+      "epoch": 0.34172173440753145,
+      "grad_norm": 0.9073922038078308,
+      "learning_rate": 0.00013936502181619416,
+      "loss": 2.9598,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3418122567212818,
+      "grad_norm": 0.8576645851135254,
+      "learning_rate": 0.00013933584870923655,
+      "loss": 2.9237,
+      "step": 3776
+    },
+    {
+      "epoch": 0.34190277903503213,
+      "grad_norm": 0.8886789679527283,
+      "learning_rate": 0.00013930667164116004,
+      "loss": 2.958,
+      "step": 3777
+    },
+    {
+      "epoch": 0.3419933013487825,
+      "grad_norm": 0.8276694416999817,
+      "learning_rate": 0.0001392774906149028,
+      "loss": 2.9362,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3420838236625328,
+      "grad_norm": 0.8300824761390686,
+      "learning_rate": 0.00013924830563340334,
+      "loss": 2.9551,
+      "step": 3779
+    },
+    {
+      "epoch": 0.34217434597628316,
+      "grad_norm": 0.8122133612632751,
+      "learning_rate": 0.00013921911669960055,
+      "loss": 2.9439,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3422648682900335,
+      "grad_norm": 0.8210098743438721,
+      "learning_rate": 0.0001391899238164338,
+      "loss": 2.9274,
+      "step": 3781
+    },
+    {
+      "epoch": 0.34235539060378384,
+      "grad_norm": 0.8149298429489136,
+      "learning_rate": 0.00013916072698684278,
+      "loss": 2.9039,
+      "step": 3782
+    },
+    {
+      "epoch": 0.3424459129175342,
+      "grad_norm": 0.8439550399780273,
+      "learning_rate": 0.00013913152621376765,
+      "loss": 2.9896,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3425364352312845,
+      "grad_norm": 0.8266587257385254,
+      "learning_rate": 0.00013910232150014885,
+      "loss": 2.9002,
+      "step": 3784
+    },
+    {
+      "epoch": 0.34262695754503486,
+      "grad_norm": 0.8575969934463501,
+      "learning_rate": 0.00013907311284892736,
+      "loss": 2.9798,
+      "step": 3785
+    },
+    {
+      "epoch": 0.3427174798587852,
+      "grad_norm": 0.8559625148773193,
+      "learning_rate": 0.00013904390026304452,
+      "loss": 2.9167,
+      "step": 3786
+    },
+    {
+      "epoch": 0.34280800217253554,
+      "grad_norm": 0.8668503761291504,
+      "learning_rate": 0.00013901468374544196,
+      "loss": 2.8825,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3428985244862859,
+      "grad_norm": 0.8820399045944214,
+      "learning_rate": 0.00013898546329906182,
+      "loss": 2.9226,
+      "step": 3788
+    },
+    {
+      "epoch": 0.3429890468000362,
+      "grad_norm": 0.9047626256942749,
+      "learning_rate": 0.00013895623892684657,
+      "loss": 2.9645,
+      "step": 3789
+    },
+    {
+      "epoch": 0.34307956911378656,
+      "grad_norm": 0.9389455318450928,
+      "learning_rate": 0.00013892701063173918,
+      "loss": 2.8754,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3431700914275369,
+      "grad_norm": 0.8222225904464722,
+      "learning_rate": 0.00013889777841668283,
+      "loss": 2.8993,
+      "step": 3791
+    },
+    {
+      "epoch": 0.34326061374128725,
+      "grad_norm": 0.8395286798477173,
+      "learning_rate": 0.00013886854228462126,
+      "loss": 2.9675,
+      "step": 3792
+    },
+    {
+      "epoch": 0.3433511360550376,
+      "grad_norm": 0.8453222513198853,
+      "learning_rate": 0.00013883930223849858,
+      "loss": 2.8964,
+      "step": 3793
+    },
+    {
+      "epoch": 0.3434416583687879,
+      "grad_norm": 0.8446892499923706,
+      "learning_rate": 0.00013881005828125922,
+      "loss": 2.8557,
+      "step": 3794
+    },
+    {
+      "epoch": 0.34353218068253827,
+      "grad_norm": 0.8886775970458984,
+      "learning_rate": 0.00013878081041584803,
+      "loss": 2.9072,
+      "step": 3795
+    },
+    {
+      "epoch": 0.3436227029962886,
+      "grad_norm": 0.8726206421852112,
+      "learning_rate": 0.0001387515586452103,
+      "loss": 2.9519,
+      "step": 3796
+    },
+    {
+      "epoch": 0.34371322531003895,
+      "grad_norm": 0.8880523443222046,
+      "learning_rate": 0.00013872230297229168,
+      "loss": 2.88,
+      "step": 3797
+    },
+    {
+      "epoch": 0.3438037476237893,
+      "grad_norm": 0.8805081248283386,
+      "learning_rate": 0.0001386930434000382,
+      "loss": 2.8936,
+      "step": 3798
+    },
+    {
+      "epoch": 0.3438942699375396,
+      "grad_norm": 0.8521276712417603,
+      "learning_rate": 0.00013866377993139635,
+      "loss": 2.9465,
+      "step": 3799
+    },
+    {
+      "epoch": 0.3439847922512899,
+      "grad_norm": 0.9059352874755859,
+      "learning_rate": 0.00013863451256931287,
+      "loss": 2.89,
+      "step": 3800
+    },
+    {
+      "epoch": 0.34407531456504026,
+      "grad_norm": 0.8595585823059082,
+      "learning_rate": 0.00013860524131673505,
+      "loss": 2.9224,
+      "step": 3801
+    },
+    {
+      "epoch": 0.3441658368787906,
+      "grad_norm": 0.8836684823036194,
+      "learning_rate": 0.00013857596617661047,
+      "loss": 2.9381,
+      "step": 3802
+    },
+    {
+      "epoch": 0.34425635919254094,
+      "grad_norm": 0.8365030884742737,
+      "learning_rate": 0.0001385466871518872,
+      "loss": 2.9124,
+      "step": 3803
+    },
+    {
+      "epoch": 0.3443468815062913,
+      "grad_norm": 0.8696910738945007,
+      "learning_rate": 0.0001385174042455135,
+      "loss": 2.8779,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3444374038200416,
+      "grad_norm": 0.8183446526527405,
+      "learning_rate": 0.00013848811746043835,
+      "loss": 2.8944,
+      "step": 3805
+    },
+    {
+      "epoch": 0.34452792613379196,
+      "grad_norm": 0.8655415177345276,
+      "learning_rate": 0.00013845882679961076,
+      "loss": 2.9028,
+      "step": 3806
+    },
+    {
+      "epoch": 0.3446184484475423,
+      "grad_norm": 0.9962408542633057,
+      "learning_rate": 0.00013842953226598037,
+      "loss": 2.9101,
+      "step": 3807
+    },
+    {
+      "epoch": 0.34470897076129264,
+      "grad_norm": 0.9712062478065491,
+      "learning_rate": 0.00013840023386249713,
+      "loss": 2.9491,
+      "step": 3808
+    },
+    {
+      "epoch": 0.344799493075043,
+      "grad_norm": 0.9880498051643372,
+      "learning_rate": 0.00013837093159211144,
+      "loss": 2.9419,
+      "step": 3809
+    },
+    {
+      "epoch": 0.3448900153887933,
+      "grad_norm": 0.8662852048873901,
+      "learning_rate": 0.00013834162545777395,
+      "loss": 2.8891,
+      "step": 3810
+    },
+    {
+      "epoch": 0.34498053770254367,
+      "grad_norm": 0.9303768277168274,
+      "learning_rate": 0.00013831231546243582,
+      "loss": 2.9551,
+      "step": 3811
+    },
+    {
+      "epoch": 0.345071060016294,
+      "grad_norm": 0.8257996439933777,
+      "learning_rate": 0.00013828300160904856,
+      "loss": 2.9602,
+      "step": 3812
+    },
+    {
+      "epoch": 0.34516158233004435,
+      "grad_norm": 0.9260730147361755,
+      "learning_rate": 0.00013825368390056416,
+      "loss": 2.9734,
+      "step": 3813
+    },
+    {
+      "epoch": 0.3452521046437947,
+      "grad_norm": 0.8937355875968933,
+      "learning_rate": 0.00013822436233993475,
+      "loss": 2.9492,
+      "step": 3814
+    },
+    {
+      "epoch": 0.34534262695754503,
+      "grad_norm": 0.8791353106498718,
+      "learning_rate": 0.00013819503693011313,
+      "loss": 2.863,
+      "step": 3815
+    },
+    {
+      "epoch": 0.34543314927129537,
+      "grad_norm": 0.8687493205070496,
+      "learning_rate": 0.00013816570767405232,
+      "loss": 2.8727,
+      "step": 3816
+    },
+    {
+      "epoch": 0.3455236715850457,
+      "grad_norm": 0.9130697250366211,
+      "learning_rate": 0.00013813637457470583,
+      "loss": 2.9826,
+      "step": 3817
+    },
+    {
+      "epoch": 0.34561419389879605,
+      "grad_norm": 0.9329674243927002,
+      "learning_rate": 0.00013810703763502744,
+      "loss": 2.9252,
+      "step": 3818
+    },
+    {
+      "epoch": 0.3457047162125464,
+      "grad_norm": 0.9105332493782043,
+      "learning_rate": 0.00013807769685797138,
+      "loss": 2.897,
+      "step": 3819
+    },
+    {
+      "epoch": 0.34579523852629673,
+      "grad_norm": 0.9339108467102051,
+      "learning_rate": 0.0001380483522464923,
+      "loss": 2.8901,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3458857608400471,
+      "grad_norm": 0.9092152118682861,
+      "learning_rate": 0.00013801900380354526,
+      "loss": 2.9041,
+      "step": 3821
+    },
+    {
+      "epoch": 0.3459762831537974,
+      "grad_norm": 1.0351412296295166,
+      "learning_rate": 0.0001379896515320855,
+      "loss": 2.9424,
+      "step": 3822
+    },
+    {
+      "epoch": 0.34606680546754776,
+      "grad_norm": 0.8615590333938599,
+      "learning_rate": 0.00013796029543506887,
+      "loss": 2.8604,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3461573277812981,
+      "grad_norm": 0.9683560132980347,
+      "learning_rate": 0.00013793093551545154,
+      "loss": 2.8971,
+      "step": 3824
+    },
+    {
+      "epoch": 0.34624785009504844,
+      "grad_norm": 0.9004538655281067,
+      "learning_rate": 0.00013790157177619004,
+      "loss": 2.9622,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3463383724087988,
+      "grad_norm": 0.9129707217216492,
+      "learning_rate": 0.00013787220422024134,
+      "loss": 2.9685,
+      "step": 3826
+    },
+    {
+      "epoch": 0.3464288947225491,
+      "grad_norm": 0.9671385288238525,
+      "learning_rate": 0.00013784283285056268,
+      "loss": 2.9015,
+      "step": 3827
+    },
+    {
+      "epoch": 0.34651941703629946,
+      "grad_norm": 0.9019637107849121,
+      "learning_rate": 0.00013781345767011177,
+      "loss": 2.9677,
+      "step": 3828
+    },
+    {
+      "epoch": 0.3466099393500498,
+      "grad_norm": 0.8544565439224243,
+      "learning_rate": 0.00013778407868184672,
+      "loss": 2.966,
+      "step": 3829
+    },
+    {
+      "epoch": 0.34670046166380014,
+      "grad_norm": 0.9372018575668335,
+      "learning_rate": 0.000137754695888726,
+      "loss": 2.9562,
+      "step": 3830
+    },
+    {
+      "epoch": 0.3467909839775505,
+      "grad_norm": 1.1464779376983643,
+      "learning_rate": 0.0001377253092937084,
+      "loss": 2.9491,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3468815062913008,
+      "grad_norm": 1.0488883256912231,
+      "learning_rate": 0.0001376959188997532,
+      "loss": 2.8893,
+      "step": 3832
+    },
+    {
+      "epoch": 0.34697202860505116,
+      "grad_norm": 0.9740673899650574,
+      "learning_rate": 0.00013766652470982002,
+      "loss": 2.9507,
+      "step": 3833
+    },
+    {
+      "epoch": 0.3470625509188015,
+      "grad_norm": 0.8552302718162537,
+      "learning_rate": 0.00013763712672686885,
+      "loss": 2.9664,
+      "step": 3834
+    },
+    {
+      "epoch": 0.34715307323255185,
+      "grad_norm": 1.0639338493347168,
+      "learning_rate": 0.00013760772495385998,
+      "loss": 3.0053,
+      "step": 3835
+    },
+    {
+      "epoch": 0.3472435955463022,
+      "grad_norm": 1.0619330406188965,
+      "learning_rate": 0.0001375783193937543,
+      "loss": 2.9125,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3473341178600525,
+      "grad_norm": 1.0379544496536255,
+      "learning_rate": 0.00013754891004951286,
+      "loss": 2.8785,
+      "step": 3837
+    },
+    {
+      "epoch": 0.34742464017380287,
+      "grad_norm": 0.863757848739624,
+      "learning_rate": 0.00013751949692409718,
+      "loss": 2.8689,
+      "step": 3838
+    },
+    {
+      "epoch": 0.3475151624875532,
+      "grad_norm": 0.9642022252082825,
+      "learning_rate": 0.00013749008002046921,
+      "loss": 2.9416,
+      "step": 3839
+    },
+    {
+      "epoch": 0.3476056848013035,
+      "grad_norm": 1.0625447034835815,
+      "learning_rate": 0.00013746065934159123,
+      "loss": 2.927,
+      "step": 3840
+    },
+    {
+      "epoch": 0.34769620711505383,
+      "grad_norm": 0.9343215227127075,
+      "learning_rate": 0.00013743123489042586,
+      "loss": 2.865,
+      "step": 3841
+    },
+    {
+      "epoch": 0.3477867294288042,
+      "grad_norm": 1.0552362203598022,
+      "learning_rate": 0.00013740180666993614,
+      "loss": 2.9259,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3478772517425545,
+      "grad_norm": 0.9143028855323792,
+      "learning_rate": 0.00013737237468308553,
+      "loss": 2.9009,
+      "step": 3843
+    },
+    {
+      "epoch": 0.34796777405630486,
+      "grad_norm": 0.9504222273826599,
+      "learning_rate": 0.00013734293893283783,
+      "loss": 2.9042,
+      "step": 3844
+    },
+    {
+      "epoch": 0.3480582963700552,
+      "grad_norm": 0.8866637349128723,
+      "learning_rate": 0.00013731349942215716,
+      "loss": 2.9043,
+      "step": 3845
+    },
+    {
+      "epoch": 0.34814881868380554,
+      "grad_norm": 0.9600756764411926,
+      "learning_rate": 0.00013728405615400814,
+      "loss": 2.9725,
+      "step": 3846
+    },
+    {
+      "epoch": 0.3482393409975559,
+      "grad_norm": 1.0086690187454224,
+      "learning_rate": 0.00013725460913135572,
+      "loss": 2.9477,
+      "step": 3847
+    },
+    {
+      "epoch": 0.3483298633113062,
+      "grad_norm": 0.8588940501213074,
+      "learning_rate": 0.00013722515835716515,
+      "loss": 2.9101,
+      "step": 3848
+    },
+    {
+      "epoch": 0.34842038562505656,
+      "grad_norm": 0.8638690710067749,
+      "learning_rate": 0.00013719570383440211,
+      "loss": 2.9316,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3485109079388069,
+      "grad_norm": 0.9232494235038757,
+      "learning_rate": 0.00013716624556603274,
+      "loss": 2.9143,
+      "step": 3850
+    },
+    {
+      "epoch": 0.34860143025255724,
+      "grad_norm": 0.9864001274108887,
+      "learning_rate": 0.00013713678355502351,
+      "loss": 2.9226,
+      "step": 3851
+    },
+    {
+      "epoch": 0.3486919525663076,
+      "grad_norm": 0.898056149482727,
+      "learning_rate": 0.00013710731780434114,
+      "loss": 2.8942,
+      "step": 3852
+    },
+    {
+      "epoch": 0.3487824748800579,
+      "grad_norm": 0.8752807378768921,
+      "learning_rate": 0.0001370778483169529,
+      "loss": 2.8418,
+      "step": 3853
+    },
+    {
+      "epoch": 0.34887299719380827,
+      "grad_norm": 0.9617156982421875,
+      "learning_rate": 0.00013704837509582633,
+      "loss": 2.9151,
+      "step": 3854
+    },
+    {
+      "epoch": 0.3489635195075586,
+      "grad_norm": 0.8508132100105286,
+      "learning_rate": 0.00013701889814392944,
+      "loss": 2.9283,
+      "step": 3855
+    },
+    {
+      "epoch": 0.34905404182130895,
+      "grad_norm": 0.8842729330062866,
+      "learning_rate": 0.00013698941746423046,
+      "loss": 2.9183,
+      "step": 3856
+    },
+    {
+      "epoch": 0.3491445641350593,
+      "grad_norm": 0.8900924324989319,
+      "learning_rate": 0.0001369599330596982,
+      "loss": 2.8878,
+      "step": 3857
+    },
+    {
+      "epoch": 0.34923508644880963,
+      "grad_norm": 0.8917263150215149,
+      "learning_rate": 0.00013693044493330166,
+      "loss": 2.8979,
+      "step": 3858
+    },
+    {
+      "epoch": 0.34932560876255997,
+      "grad_norm": 1.1581900119781494,
+      "learning_rate": 0.00013690095308801036,
+      "loss": 2.9183,
+      "step": 3859
+    },
+    {
+      "epoch": 0.3494161310763103,
+      "grad_norm": 0.900983452796936,
+      "learning_rate": 0.0001368714575267941,
+      "loss": 2.8954,
+      "step": 3860
+    },
+    {
+      "epoch": 0.34950665339006065,
+      "grad_norm": 0.8715013265609741,
+      "learning_rate": 0.00013684195825262302,
+      "loss": 2.8939,
+      "step": 3861
+    },
+    {
+      "epoch": 0.349597175703811,
+      "grad_norm": 1.0625563859939575,
+      "learning_rate": 0.00013681245526846783,
+      "loss": 2.9139,
+      "step": 3862
+    },
+    {
+      "epoch": 0.34968769801756133,
+      "grad_norm": 0.8163450956344604,
+      "learning_rate": 0.00013678294857729934,
+      "loss": 2.8957,
+      "step": 3863
+    },
+    {
+      "epoch": 0.3497782203313117,
+      "grad_norm": 1.280600905418396,
+      "learning_rate": 0.00013675343818208896,
+      "loss": 2.9895,
+      "step": 3864
+    },
+    {
+      "epoch": 0.349868742645062,
+      "grad_norm": 0.9801856875419617,
+      "learning_rate": 0.00013672392408580833,
+      "loss": 2.9835,
+      "step": 3865
+    },
+    {
+      "epoch": 0.34995926495881235,
+      "grad_norm": 0.8751368522644043,
+      "learning_rate": 0.0001366944062914296,
+      "loss": 2.9536,
+      "step": 3866
+    },
+    {
+      "epoch": 0.3500497872725627,
+      "grad_norm": 0.8888764977455139,
+      "learning_rate": 0.00013666488480192513,
+      "loss": 2.8834,
+      "step": 3867
+    },
+    {
+      "epoch": 0.35014030958631304,
+      "grad_norm": 0.8316412568092346,
+      "learning_rate": 0.00013663535962026778,
+      "loss": 2.8587,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3502308319000634,
+      "grad_norm": 0.8390792608261108,
+      "learning_rate": 0.00013660583074943068,
+      "loss": 2.9432,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3503213542138137,
+      "grad_norm": 0.8946274518966675,
+      "learning_rate": 0.00013657629819238746,
+      "loss": 2.9448,
+      "step": 3870
+    },
+    {
+      "epoch": 0.35041187652756406,
+      "grad_norm": 0.8634069561958313,
+      "learning_rate": 0.000136546761952112,
+      "loss": 2.9023,
+      "step": 3871
+    },
+    {
+      "epoch": 0.3505023988413144,
+      "grad_norm": 0.932811975479126,
+      "learning_rate": 0.00013651722203157862,
+      "loss": 2.9542,
+      "step": 3872
+    },
+    {
+      "epoch": 0.35059292115506474,
+      "grad_norm": 0.9042550921440125,
+      "learning_rate": 0.00013648767843376196,
+      "loss": 2.978,
+      "step": 3873
+    },
+    {
+      "epoch": 0.3506834434688151,
+      "grad_norm": 0.8619733452796936,
+      "learning_rate": 0.00013645813116163713,
+      "loss": 2.9299,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3507739657825654,
+      "grad_norm": 0.9075789451599121,
+      "learning_rate": 0.00013642858021817943,
+      "loss": 2.8829,
+      "step": 3875
+    },
+    {
+      "epoch": 0.35086448809631576,
+      "grad_norm": 0.8590002655982971,
+      "learning_rate": 0.00013639902560636473,
+      "loss": 2.8802,
+      "step": 3876
+    },
+    {
+      "epoch": 0.3509550104100661,
+      "grad_norm": 0.9196003079414368,
+      "learning_rate": 0.00013636946732916916,
+      "loss": 2.9321,
+      "step": 3877
+    },
+    {
+      "epoch": 0.35104553272381644,
+      "grad_norm": 0.9296711087226868,
+      "learning_rate": 0.0001363399053895692,
+      "loss": 2.9494,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3511360550375668,
+      "grad_norm": 1.036049723625183,
+      "learning_rate": 0.00013631033979054176,
+      "loss": 2.9524,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3512265773513171,
+      "grad_norm": 0.806705892086029,
+      "learning_rate": 0.0001362807705350641,
+      "loss": 2.9233,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3513170996650674,
+      "grad_norm": 0.9473732709884644,
+      "learning_rate": 0.00013625119762611387,
+      "loss": 2.9387,
+      "step": 3881
+    },
+    {
+      "epoch": 0.35140762197881775,
+      "grad_norm": 0.8431051969528198,
+      "learning_rate": 0.000136221621066669,
+      "loss": 2.9026,
+      "step": 3882
+    },
+    {
+      "epoch": 0.3514981442925681,
+      "grad_norm": 0.784507155418396,
+      "learning_rate": 0.00013619204085970788,
+      "loss": 2.8671,
+      "step": 3883
+    },
+    {
+      "epoch": 0.35158866660631843,
+      "grad_norm": 0.8995726704597473,
+      "learning_rate": 0.00013616245700820922,
+      "loss": 2.877,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3516791889200688,
+      "grad_norm": 0.8044691681861877,
+      "learning_rate": 0.00013613286951515217,
+      "loss": 2.926,
+      "step": 3885
+    },
+    {
+      "epoch": 0.3517697112338191,
+      "grad_norm": 0.886650025844574,
+      "learning_rate": 0.00013610327838351613,
+      "loss": 2.8995,
+      "step": 3886
+    },
+    {
+      "epoch": 0.35186023354756946,
+      "grad_norm": 0.8361202478408813,
+      "learning_rate": 0.00013607368361628094,
+      "loss": 2.8801,
+      "step": 3887
+    },
+    {
+      "epoch": 0.3519507558613198,
+      "grad_norm": 0.8228411078453064,
+      "learning_rate": 0.0001360440852164268,
+      "loss": 2.8873,
+      "step": 3888
+    },
+    {
+      "epoch": 0.35204127817507014,
+      "grad_norm": 0.8641404509544373,
+      "learning_rate": 0.0001360144831869343,
+      "loss": 2.9192,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3521318004888205,
+      "grad_norm": 0.862514078617096,
+      "learning_rate": 0.00013598487753078425,
+      "loss": 3.0059,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3522223228025708,
+      "grad_norm": 0.8352610468864441,
+      "learning_rate": 0.0001359552682509581,
+      "loss": 2.9455,
+      "step": 3891
+    },
+    {
+      "epoch": 0.35231284511632116,
+      "grad_norm": 0.8321881890296936,
+      "learning_rate": 0.00013592565535043737,
+      "loss": 2.874,
+      "step": 3892
+    },
+    {
+      "epoch": 0.3524033674300715,
+      "grad_norm": 0.9299854040145874,
+      "learning_rate": 0.00013589603883220415,
+      "loss": 2.9172,
+      "step": 3893
+    },
+    {
+      "epoch": 0.35249388974382184,
+      "grad_norm": 0.8699597716331482,
+      "learning_rate": 0.00013586641869924078,
+      "loss": 2.8922,
+      "step": 3894
+    },
+    {
+      "epoch": 0.3525844120575722,
+      "grad_norm": 0.9263263940811157,
+      "learning_rate": 0.00013583679495453,
+      "loss": 2.9072,
+      "step": 3895
+    },
+    {
+      "epoch": 0.3526749343713225,
+      "grad_norm": 0.9022011160850525,
+      "learning_rate": 0.00013580716760105502,
+      "loss": 2.9152,
+      "step": 3896
+    },
+    {
+      "epoch": 0.35276545668507286,
+      "grad_norm": 0.8681530356407166,
+      "learning_rate": 0.0001357775366417992,
+      "loss": 2.9911,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3528559789988232,
+      "grad_norm": 0.9377198815345764,
+      "learning_rate": 0.00013574790207974646,
+      "loss": 2.8987,
+      "step": 3898
+    },
+    {
+      "epoch": 0.35294650131257355,
+      "grad_norm": 0.9530036449432373,
+      "learning_rate": 0.00013571826391788093,
+      "loss": 2.9039,
+      "step": 3899
+    },
+    {
+      "epoch": 0.3530370236263239,
+      "grad_norm": 1.0229028463363647,
+      "learning_rate": 0.00013568862215918717,
+      "loss": 2.9019,
+      "step": 3900
+    },
+    {
+      "epoch": 0.35312754594007423,
+      "grad_norm": 1.012978434562683,
+      "learning_rate": 0.0001356589768066502,
+      "loss": 2.8668,
+      "step": 3901
+    },
+    {
+      "epoch": 0.35321806825382457,
+      "grad_norm": 1.0030837059020996,
+      "learning_rate": 0.00013562932786325522,
+      "loss": 2.9775,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3533085905675749,
+      "grad_norm": 1.00313401222229,
+      "learning_rate": 0.00013559967533198787,
+      "loss": 2.909,
+      "step": 3903
+    },
+    {
+      "epoch": 0.35339911288132525,
+      "grad_norm": 1.044822335243225,
+      "learning_rate": 0.0001355700192158342,
+      "loss": 2.938,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3534896351950756,
+      "grad_norm": 1.1762657165527344,
+      "learning_rate": 0.00013554035951778058,
+      "loss": 2.9599,
+      "step": 3905
+    },
+    {
+      "epoch": 0.35358015750882593,
+      "grad_norm": 0.9761765599250793,
+      "learning_rate": 0.0001355106962408137,
+      "loss": 2.9664,
+      "step": 3906
+    },
+    {
+      "epoch": 0.3536706798225763,
+      "grad_norm": 0.9626485109329224,
+      "learning_rate": 0.00013548102938792074,
+      "loss": 2.9368,
+      "step": 3907
+    },
+    {
+      "epoch": 0.3537612021363266,
+      "grad_norm": 1.0166648626327515,
+      "learning_rate": 0.00013545135896208906,
+      "loss": 2.9386,
+      "step": 3908
+    },
+    {
+      "epoch": 0.35385172445007695,
+      "grad_norm": 0.9419061541557312,
+      "learning_rate": 0.00013542168496630648,
+      "loss": 2.9433,
+      "step": 3909
+    },
+    {
+      "epoch": 0.3539422467638273,
+      "grad_norm": 0.870576024055481,
+      "learning_rate": 0.00013539200740356118,
+      "loss": 2.9445,
+      "step": 3910
+    },
+    {
+      "epoch": 0.35403276907757764,
+      "grad_norm": 0.9303431510925293,
+      "learning_rate": 0.00013536232627684176,
+      "loss": 2.8962,
+      "step": 3911
+    },
+    {
+      "epoch": 0.354123291391328,
+      "grad_norm": 1.0295835733413696,
+      "learning_rate": 0.00013533264158913704,
+      "loss": 2.9648,
+      "step": 3912
+    },
+    {
+      "epoch": 0.3542138137050783,
+      "grad_norm": 0.9527624249458313,
+      "learning_rate": 0.00013530295334343624,
+      "loss": 2.9421,
+      "step": 3913
+    },
+    {
+      "epoch": 0.35430433601882866,
+      "grad_norm": 1.1770457029342651,
+      "learning_rate": 0.000135273261542729,
+      "loss": 2.9074,
+      "step": 3914
+    },
+    {
+      "epoch": 0.354394858332579,
+      "grad_norm": 0.8796162605285645,
+      "learning_rate": 0.00013524356619000534,
+      "loss": 2.956,
+      "step": 3915
+    },
+    {
+      "epoch": 0.35448538064632934,
+      "grad_norm": 0.9670255780220032,
+      "learning_rate": 0.0001352138672882555,
+      "loss": 2.8812,
+      "step": 3916
+    },
+    {
+      "epoch": 0.3545759029600797,
+      "grad_norm": 0.9039412140846252,
+      "learning_rate": 0.00013518416484047018,
+      "loss": 2.9025,
+      "step": 3917
+    },
+    {
+      "epoch": 0.35466642527383,
+      "grad_norm": 0.8382848501205444,
+      "learning_rate": 0.00013515445884964045,
+      "loss": 2.914,
+      "step": 3918
+    },
+    {
+      "epoch": 0.35475694758758036,
+      "grad_norm": 0.9838911890983582,
+      "learning_rate": 0.00013512474931875765,
+      "loss": 2.9671,
+      "step": 3919
+    },
+    {
+      "epoch": 0.3548474699013307,
+      "grad_norm": 0.9354199171066284,
+      "learning_rate": 0.00013509503625081358,
+      "loss": 2.9004,
+      "step": 3920
+    },
+    {
+      "epoch": 0.35493799221508104,
+      "grad_norm": 0.8873734474182129,
+      "learning_rate": 0.0001350653196488003,
+      "loss": 2.879,
+      "step": 3921
+    },
+    {
+      "epoch": 0.35502851452883133,
+      "grad_norm": 1.0136916637420654,
+      "learning_rate": 0.0001350355995157103,
+      "loss": 2.914,
+      "step": 3922
+    },
+    {
+      "epoch": 0.35511903684258167,
+      "grad_norm": 0.8554311394691467,
+      "learning_rate": 0.00013500587585453638,
+      "loss": 2.9463,
+      "step": 3923
+    },
+    {
+      "epoch": 0.355209559156332,
+      "grad_norm": 0.8193634748458862,
+      "learning_rate": 0.00013497614866827173,
+      "loss": 2.9058,
+      "step": 3924
+    },
+    {
+      "epoch": 0.35530008147008235,
+      "grad_norm": 0.9600344300270081,
+      "learning_rate": 0.00013494641795990986,
+      "loss": 2.873,
+      "step": 3925
+    },
+    {
+      "epoch": 0.3553906037838327,
+      "grad_norm": 0.928473949432373,
+      "learning_rate": 0.00013491668373244464,
+      "loss": 2.8798,
+      "step": 3926
+    },
+    {
+      "epoch": 0.35548112609758303,
+      "grad_norm": 0.8732498288154602,
+      "learning_rate": 0.00013488694598887035,
+      "loss": 2.861,
+      "step": 3927
+    },
+    {
+      "epoch": 0.3555716484113334,
+      "grad_norm": 0.9893165230751038,
+      "learning_rate": 0.00013485720473218154,
+      "loss": 2.8755,
+      "step": 3928
+    },
+    {
+      "epoch": 0.3556621707250837,
+      "grad_norm": 0.9542164206504822,
+      "learning_rate": 0.00013482745996537315,
+      "loss": 2.9289,
+      "step": 3929
+    },
+    {
+      "epoch": 0.35575269303883406,
+      "grad_norm": 0.8913797736167908,
+      "learning_rate": 0.0001347977116914405,
+      "loss": 2.8931,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3558432153525844,
+      "grad_norm": 1.016385555267334,
+      "learning_rate": 0.0001347679599133793,
+      "loss": 2.8436,
+      "step": 3931
+    },
+    {
+      "epoch": 0.35593373766633474,
+      "grad_norm": 1.0167441368103027,
+      "learning_rate": 0.0001347382046341854,
+      "loss": 2.9751,
+      "step": 3932
+    },
+    {
+      "epoch": 0.3560242599800851,
+      "grad_norm": 0.8540645241737366,
+      "learning_rate": 0.0001347084458568553,
+      "loss": 2.95,
+      "step": 3933
+    },
+    {
+      "epoch": 0.3561147822938354,
+      "grad_norm": 0.943660318851471,
+      "learning_rate": 0.00013467868358438563,
+      "loss": 2.8982,
+      "step": 3934
+    },
+    {
+      "epoch": 0.35620530460758576,
+      "grad_norm": 0.9221616983413696,
+      "learning_rate": 0.00013464891781977349,
+      "loss": 2.9524,
+      "step": 3935
+    },
+    {
+      "epoch": 0.3562958269213361,
+      "grad_norm": 0.9135646224021912,
+      "learning_rate": 0.00013461914856601625,
+      "loss": 2.8953,
+      "step": 3936
+    },
+    {
+      "epoch": 0.35638634923508644,
+      "grad_norm": 1.021881103515625,
+      "learning_rate": 0.0001345893758261117,
+      "loss": 2.994,
+      "step": 3937
+    },
+    {
+      "epoch": 0.3564768715488368,
+      "grad_norm": 0.8870673775672913,
+      "learning_rate": 0.00013455959960305798,
+      "loss": 2.8441,
+      "step": 3938
+    },
+    {
+      "epoch": 0.3565673938625871,
+      "grad_norm": 0.9238022565841675,
+      "learning_rate": 0.00013452981989985348,
+      "loss": 2.8884,
+      "step": 3939
+    },
+    {
+      "epoch": 0.35665791617633746,
+      "grad_norm": 1.0125572681427002,
+      "learning_rate": 0.00013450003671949706,
+      "loss": 2.9548,
+      "step": 3940
+    },
+    {
+      "epoch": 0.3567484384900878,
+      "grad_norm": 0.8192005157470703,
+      "learning_rate": 0.00013447025006498793,
+      "loss": 2.9126,
+      "step": 3941
+    },
+    {
+      "epoch": 0.35683896080383815,
+      "grad_norm": 1.4117205142974854,
+      "learning_rate": 0.00013444045993932555,
+      "loss": 2.9368,
+      "step": 3942
+    },
+    {
+      "epoch": 0.3569294831175885,
+      "grad_norm": 0.9625622630119324,
+      "learning_rate": 0.00013441066634550978,
+      "loss": 2.8696,
+      "step": 3943
+    },
+    {
+      "epoch": 0.3570200054313388,
+      "grad_norm": 0.7912752032279968,
+      "learning_rate": 0.00013438086928654086,
+      "loss": 2.9093,
+      "step": 3944
+    },
+    {
+      "epoch": 0.35711052774508917,
+      "grad_norm": 1.269979476928711,
+      "learning_rate": 0.00013435106876541933,
+      "loss": 2.9174,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3572010500588395,
+      "grad_norm": 1.0200188159942627,
+      "learning_rate": 0.00013432126478514614,
+      "loss": 2.8732,
+      "step": 3946
+    },
+    {
+      "epoch": 0.35729157237258985,
+      "grad_norm": 0.8661880493164062,
+      "learning_rate": 0.0001342914573487225,
+      "loss": 2.8182,
+      "step": 3947
+    },
+    {
+      "epoch": 0.3573820946863402,
+      "grad_norm": 1.0338268280029297,
+      "learning_rate": 0.00013426164645915004,
+      "loss": 2.9328,
+      "step": 3948
+    },
+    {
+      "epoch": 0.35747261700009053,
+      "grad_norm": 1.0458835363388062,
+      "learning_rate": 0.00013423183211943074,
+      "loss": 2.8545,
+      "step": 3949
+    },
+    {
+      "epoch": 0.35756313931384087,
+      "grad_norm": 0.9032859206199646,
+      "learning_rate": 0.00013420201433256689,
+      "loss": 2.8888,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3576536616275912,
+      "grad_norm": 0.9008709788322449,
+      "learning_rate": 0.0001341721931015611,
+      "loss": 2.8591,
+      "step": 3951
+    },
+    {
+      "epoch": 0.35774418394134155,
+      "grad_norm": 0.9279823899269104,
+      "learning_rate": 0.00013414236842941644,
+      "loss": 2.8552,
+      "step": 3952
+    },
+    {
+      "epoch": 0.3578347062550919,
+      "grad_norm": 0.9996717572212219,
+      "learning_rate": 0.00013411254031913615,
+      "loss": 2.9944,
+      "step": 3953
+    },
+    {
+      "epoch": 0.35792522856884224,
+      "grad_norm": 0.8940361738204956,
+      "learning_rate": 0.00013408270877372406,
+      "loss": 2.8701,
+      "step": 3954
+    },
+    {
+      "epoch": 0.3580157508825926,
+      "grad_norm": 0.96931391954422,
+      "learning_rate": 0.0001340528737961841,
+      "loss": 2.9365,
+      "step": 3955
+    },
+    {
+      "epoch": 0.3581062731963429,
+      "grad_norm": 0.9654762148857117,
+      "learning_rate": 0.00013402303538952065,
+      "loss": 2.854,
+      "step": 3956
+    },
+    {
+      "epoch": 0.35819679551009326,
+      "grad_norm": 0.8370428681373596,
+      "learning_rate": 0.0001339931935567385,
+      "loss": 2.9161,
+      "step": 3957
+    },
+    {
+      "epoch": 0.3582873178238436,
+      "grad_norm": 0.8343283534049988,
+      "learning_rate": 0.0001339633483008427,
+      "loss": 2.8579,
+      "step": 3958
+    },
+    {
+      "epoch": 0.35837784013759394,
+      "grad_norm": 0.9394232630729675,
+      "learning_rate": 0.0001339334996248386,
+      "loss": 2.9178,
+      "step": 3959
+    },
+    {
+      "epoch": 0.3584683624513443,
+      "grad_norm": 0.946890652179718,
+      "learning_rate": 0.00013390364753173206,
+      "loss": 2.9488,
+      "step": 3960
+    },
+    {
+      "epoch": 0.3585588847650946,
+      "grad_norm": 0.8705549836158752,
+      "learning_rate": 0.00013387379202452917,
+      "loss": 2.8761,
+      "step": 3961
+    },
+    {
+      "epoch": 0.35864940707884496,
+      "grad_norm": 0.836525559425354,
+      "learning_rate": 0.0001338439331062363,
+      "loss": 2.8907,
+      "step": 3962
+    },
+    {
+      "epoch": 0.35873992939259525,
+      "grad_norm": 0.9729019403457642,
+      "learning_rate": 0.0001338140707798603,
+      "loss": 2.9337,
+      "step": 3963
+    },
+    {
+      "epoch": 0.3588304517063456,
+      "grad_norm": 0.8781609535217285,
+      "learning_rate": 0.00013378420504840828,
+      "loss": 3.0256,
+      "step": 3964
+    },
+    {
+      "epoch": 0.35892097402009593,
+      "grad_norm": 0.8746907114982605,
+      "learning_rate": 0.0001337543359148878,
+      "loss": 2.9021,
+      "step": 3965
+    },
+    {
+      "epoch": 0.35901149633384627,
+      "grad_norm": 0.8925855755805969,
+      "learning_rate": 0.00013372446338230656,
+      "loss": 2.9353,
+      "step": 3966
+    },
+    {
+      "epoch": 0.3591020186475966,
+      "grad_norm": 0.8417982459068298,
+      "learning_rate": 0.0001336945874536728,
+      "loss": 2.8828,
+      "step": 3967
+    },
+    {
+      "epoch": 0.35919254096134695,
+      "grad_norm": 0.8790850043296814,
+      "learning_rate": 0.000133664708131995,
+      "loss": 2.8381,
+      "step": 3968
+    },
+    {
+      "epoch": 0.3592830632750973,
+      "grad_norm": 0.9135075807571411,
+      "learning_rate": 0.000133634825420282,
+      "loss": 2.9152,
+      "step": 3969
+    },
+    {
+      "epoch": 0.35937358558884763,
+      "grad_norm": 0.939116358757019,
+      "learning_rate": 0.00013360493932154302,
+      "loss": 2.9089,
+      "step": 3970
+    },
+    {
+      "epoch": 0.359464107902598,
+      "grad_norm": 0.8897899985313416,
+      "learning_rate": 0.00013357504983878754,
+      "loss": 2.902,
+      "step": 3971
+    },
+    {
+      "epoch": 0.3595546302163483,
+      "grad_norm": 0.9135044813156128,
+      "learning_rate": 0.00013354515697502553,
+      "loss": 2.8635,
+      "step": 3972
+    },
+    {
+      "epoch": 0.35964515253009866,
+      "grad_norm": 0.8595693111419678,
+      "learning_rate": 0.00013351526073326707,
+      "loss": 2.9217,
+      "step": 3973
+    },
+    {
+      "epoch": 0.359735674843849,
+      "grad_norm": 0.9174153804779053,
+      "learning_rate": 0.00013348536111652278,
+      "loss": 2.8597,
+      "step": 3974
+    },
+    {
+      "epoch": 0.35982619715759934,
+      "grad_norm": 0.9224886894226074,
+      "learning_rate": 0.00013345545812780353,
+      "loss": 2.9499,
+      "step": 3975
+    },
+    {
+      "epoch": 0.3599167194713497,
+      "grad_norm": 0.8565012216567993,
+      "learning_rate": 0.0001334255517701206,
+      "loss": 2.9246,
+      "step": 3976
+    },
+    {
+      "epoch": 0.3600072417851,
+      "grad_norm": 0.8404529690742493,
+      "learning_rate": 0.0001333956420464855,
+      "loss": 2.8516,
+      "step": 3977
+    },
+    {
+      "epoch": 0.36009776409885036,
+      "grad_norm": 0.8397718071937561,
+      "learning_rate": 0.00013336572895991016,
+      "loss": 2.9095,
+      "step": 3978
+    },
+    {
+      "epoch": 0.3601882864126007,
+      "grad_norm": 0.8200170993804932,
+      "learning_rate": 0.0001333358125134068,
+      "loss": 2.9085,
+      "step": 3979
+    },
+    {
+      "epoch": 0.36027880872635104,
+      "grad_norm": 0.8960346579551697,
+      "learning_rate": 0.00013330589270998808,
+      "loss": 2.9098,
+      "step": 3980
+    },
+    {
+      "epoch": 0.3603693310401014,
+      "grad_norm": 0.8814414143562317,
+      "learning_rate": 0.00013327596955266686,
+      "loss": 2.8328,
+      "step": 3981
+    },
+    {
+      "epoch": 0.3604598533538517,
+      "grad_norm": 0.9030823707580566,
+      "learning_rate": 0.0001332460430444564,
+      "loss": 2.9183,
+      "step": 3982
+    },
+    {
+      "epoch": 0.36055037566760206,
+      "grad_norm": 0.9058015942573547,
+      "learning_rate": 0.00013321611318837032,
+      "loss": 2.9164,
+      "step": 3983
+    },
+    {
+      "epoch": 0.3606408979813524,
+      "grad_norm": 1.0026699304580688,
+      "learning_rate": 0.0001331861799874226,
+      "loss": 2.8641,
+      "step": 3984
+    },
+    {
+      "epoch": 0.36073142029510274,
+      "grad_norm": 0.8512104749679565,
+      "learning_rate": 0.0001331562434446274,
+      "loss": 2.8295,
+      "step": 3985
+    },
+    {
+      "epoch": 0.3608219426088531,
+      "grad_norm": 0.7946496605873108,
+      "learning_rate": 0.00013312630356299943,
+      "loss": 2.8758,
+      "step": 3986
+    },
+    {
+      "epoch": 0.3609124649226034,
+      "grad_norm": 0.9297618865966797,
+      "learning_rate": 0.00013309636034555355,
+      "loss": 2.8679,
+      "step": 3987
+    },
+    {
+      "epoch": 0.36100298723635377,
+      "grad_norm": 0.9128022789955139,
+      "learning_rate": 0.00013306641379530514,
+      "loss": 2.8875,
+      "step": 3988
+    },
+    {
+      "epoch": 0.3610935095501041,
+      "grad_norm": 0.9625574946403503,
+      "learning_rate": 0.00013303646391526973,
+      "loss": 2.9534,
+      "step": 3989
+    },
+    {
+      "epoch": 0.36118403186385445,
+      "grad_norm": 0.8782124519348145,
+      "learning_rate": 0.00013300651070846333,
+      "loss": 2.9566,
+      "step": 3990
+    },
+    {
+      "epoch": 0.3612745541776048,
+      "grad_norm": 0.8106698989868164,
+      "learning_rate": 0.0001329765541779022,
+      "loss": 2.8592,
+      "step": 3991
+    },
+    {
+      "epoch": 0.36136507649135513,
+      "grad_norm": 0.9018152356147766,
+      "learning_rate": 0.00013294659432660296,
+      "loss": 2.9032,
+      "step": 3992
+    },
+    {
+      "epoch": 0.36145559880510547,
+      "grad_norm": 0.8355795741081238,
+      "learning_rate": 0.0001329166311575826,
+      "loss": 2.9498,
+      "step": 3993
+    },
+    {
+      "epoch": 0.3615461211188558,
+      "grad_norm": 0.862778902053833,
+      "learning_rate": 0.00013288666467385833,
+      "loss": 2.8544,
+      "step": 3994
+    },
+    {
+      "epoch": 0.36163664343260615,
+      "grad_norm": 0.8779953718185425,
+      "learning_rate": 0.00013285669487844785,
+      "loss": 2.9377,
+      "step": 3995
+    },
+    {
+      "epoch": 0.3617271657463565,
+      "grad_norm": 0.9221465587615967,
+      "learning_rate": 0.00013282672177436912,
+      "loss": 2.8477,
+      "step": 3996
+    },
+    {
+      "epoch": 0.36181768806010683,
+      "grad_norm": 0.908944308757782,
+      "learning_rate": 0.00013279674536464036,
+      "loss": 2.8751,
+      "step": 3997
+    },
+    {
+      "epoch": 0.3619082103738572,
+      "grad_norm": 0.9321480989456177,
+      "learning_rate": 0.00013276676565228027,
+      "loss": 2.9282,
+      "step": 3998
+    },
+    {
+      "epoch": 0.3619987326876075,
+      "grad_norm": 0.9838185906410217,
+      "learning_rate": 0.00013273678264030778,
+      "loss": 2.905,
+      "step": 3999
+    },
+    {
+      "epoch": 0.36208925500135786,
+      "grad_norm": 0.9170008301734924,
+      "learning_rate": 0.00013270679633174218,
+      "loss": 2.9405,
+      "step": 4000
+    },
+    {
+      "epoch": 0.36208925500135786,
+      "eval_loss": 2.8359169960021973,
+      "eval_runtime": 71.472,
+      "eval_samples_per_second": 37.819,
+      "eval_steps_per_second": 3.162,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3621797773151082,
+      "grad_norm": 0.8533981442451477,
+      "learning_rate": 0.00013267680672960307,
+      "loss": 2.9021,
+      "step": 4001
+    },
+    {
+      "epoch": 0.36227029962885854,
+      "grad_norm": 0.9296990633010864,
+      "learning_rate": 0.0001326468138369104,
+      "loss": 2.9398,
+      "step": 4002
+    },
+    {
+      "epoch": 0.3623608219426089,
+      "grad_norm": 0.8591863512992859,
+      "learning_rate": 0.00013261681765668453,
+      "loss": 2.8718,
+      "step": 4003
+    },
+    {
+      "epoch": 0.3624513442563592,
+      "grad_norm": 0.9215024709701538,
+      "learning_rate": 0.00013258681819194594,
+      "loss": 2.9158,
+      "step": 4004
+    },
+    {
+      "epoch": 0.3625418665701095,
+      "grad_norm": 0.9291088581085205,
+      "learning_rate": 0.00013255681544571568,
+      "loss": 2.8731,
+      "step": 4005
+    },
+    {
+      "epoch": 0.36263238888385985,
+      "grad_norm": 0.9838840365409851,
+      "learning_rate": 0.000132526809421015,
+      "loss": 2.8812,
+      "step": 4006
+    },
+    {
+      "epoch": 0.3627229111976102,
+      "grad_norm": 0.8778908848762512,
+      "learning_rate": 0.0001324968001208655,
+      "loss": 2.8826,
+      "step": 4007
+    },
+    {
+      "epoch": 0.36281343351136053,
+      "grad_norm": 0.9058883190155029,
+      "learning_rate": 0.00013246678754828912,
+      "loss": 2.9113,
+      "step": 4008
+    },
+    {
+      "epoch": 0.36290395582511087,
+      "grad_norm": 0.9569624662399292,
+      "learning_rate": 0.0001324367717063081,
+      "loss": 2.9263,
+      "step": 4009
+    },
+    {
+      "epoch": 0.3629944781388612,
+      "grad_norm": 0.839088499546051,
+      "learning_rate": 0.00013240675259794507,
+      "loss": 2.8705,
+      "step": 4010
+    },
+    {
+      "epoch": 0.36308500045261155,
+      "grad_norm": 0.906187891960144,
+      "learning_rate": 0.00013237673022622296,
+      "loss": 2.842,
+      "step": 4011
+    },
+    {
+      "epoch": 0.3631755227663619,
+      "grad_norm": 0.8538724780082703,
+      "learning_rate": 0.00013234670459416498,
+      "loss": 2.8928,
+      "step": 4012
+    },
+    {
+      "epoch": 0.36326604508011223,
+      "grad_norm": 0.933245062828064,
+      "learning_rate": 0.00013231667570479474,
+      "loss": 2.9416,
+      "step": 4013
+    },
+    {
+      "epoch": 0.3633565673938626,
+      "grad_norm": 0.82557213306427,
+      "learning_rate": 0.00013228664356113612,
+      "loss": 2.8834,
+      "step": 4014
+    },
+    {
+      "epoch": 0.3634470897076129,
+      "grad_norm": 0.8684479594230652,
+      "learning_rate": 0.0001322566081662134,
+      "loss": 2.9387,
+      "step": 4015
+    },
+    {
+      "epoch": 0.36353761202136325,
+      "grad_norm": 0.8461321592330933,
+      "learning_rate": 0.00013222656952305113,
+      "loss": 2.9157,
+      "step": 4016
+    },
+    {
+      "epoch": 0.3636281343351136,
+      "grad_norm": 0.815130352973938,
+      "learning_rate": 0.00013219652763467416,
+      "loss": 2.831,
+      "step": 4017
+    },
+    {
+      "epoch": 0.36371865664886394,
+      "grad_norm": 0.835340142250061,
+      "learning_rate": 0.00013216648250410776,
+      "loss": 2.8644,
+      "step": 4018
+    },
+    {
+      "epoch": 0.3638091789626143,
+      "grad_norm": 0.8132588267326355,
+      "learning_rate": 0.00013213643413437746,
+      "loss": 2.9148,
+      "step": 4019
+    },
+    {
+      "epoch": 0.3638997012763646,
+      "grad_norm": 0.8714550733566284,
+      "learning_rate": 0.00013210638252850908,
+      "loss": 2.8805,
+      "step": 4020
+    },
+    {
+      "epoch": 0.36399022359011496,
+      "grad_norm": 0.8507872819900513,
+      "learning_rate": 0.0001320763276895289,
+      "loss": 2.901,
+      "step": 4021
+    },
+    {
+      "epoch": 0.3640807459038653,
+      "grad_norm": 0.8836880922317505,
+      "learning_rate": 0.0001320462696204634,
+      "loss": 2.9043,
+      "step": 4022
+    },
+    {
+      "epoch": 0.36417126821761564,
+      "grad_norm": 0.9315512776374817,
+      "learning_rate": 0.0001320162083243394,
+      "loss": 2.8655,
+      "step": 4023
+    },
+    {
+      "epoch": 0.364261790531366,
+      "grad_norm": 0.828289806842804,
+      "learning_rate": 0.00013198614380418412,
+      "loss": 2.8794,
+      "step": 4024
+    },
+    {
+      "epoch": 0.3643523128451163,
+      "grad_norm": 0.8746485114097595,
+      "learning_rate": 0.00013195607606302501,
+      "loss": 2.9045,
+      "step": 4025
+    },
+    {
+      "epoch": 0.36444283515886666,
+      "grad_norm": 0.8498311638832092,
+      "learning_rate": 0.00013192600510388992,
+      "loss": 2.8249,
+      "step": 4026
+    },
+    {
+      "epoch": 0.364533357472617,
+      "grad_norm": 0.8305221796035767,
+      "learning_rate": 0.00013189593092980702,
+      "loss": 2.8825,
+      "step": 4027
+    },
+    {
+      "epoch": 0.36462387978636734,
+      "grad_norm": 0.7950791716575623,
+      "learning_rate": 0.0001318658535438047,
+      "loss": 2.9084,
+      "step": 4028
+    },
+    {
+      "epoch": 0.3647144021001177,
+      "grad_norm": 0.888029158115387,
+      "learning_rate": 0.00013183577294891184,
+      "loss": 2.8292,
+      "step": 4029
+    },
+    {
+      "epoch": 0.364804924413868,
+      "grad_norm": 0.8210163712501526,
+      "learning_rate": 0.00013180568914815752,
+      "loss": 2.888,
+      "step": 4030
+    },
+    {
+      "epoch": 0.36489544672761837,
+      "grad_norm": 0.8236469030380249,
+      "learning_rate": 0.00013177560214457115,
+      "loss": 2.8584,
+      "step": 4031
+    },
+    {
+      "epoch": 0.3649859690413687,
+      "grad_norm": 0.8504825234413147,
+      "learning_rate": 0.0001317455119411825,
+      "loss": 2.8539,
+      "step": 4032
+    },
+    {
+      "epoch": 0.36507649135511905,
+      "grad_norm": 0.825000524520874,
+      "learning_rate": 0.0001317154185410217,
+      "loss": 2.8158,
+      "step": 4033
+    },
+    {
+      "epoch": 0.3651670136688694,
+      "grad_norm": 0.8093627095222473,
+      "learning_rate": 0.0001316853219471191,
+      "loss": 2.8809,
+      "step": 4034
+    },
+    {
+      "epoch": 0.36525753598261973,
+      "grad_norm": 0.9785566926002502,
+      "learning_rate": 0.00013165522216250543,
+      "loss": 2.9101,
+      "step": 4035
+    },
+    {
+      "epoch": 0.36534805829637007,
+      "grad_norm": 0.9431535005569458,
+      "learning_rate": 0.00013162511919021178,
+      "loss": 2.8807,
+      "step": 4036
+    },
+    {
+      "epoch": 0.3654385806101204,
+      "grad_norm": 0.8769155144691467,
+      "learning_rate": 0.0001315950130332695,
+      "loss": 2.9477,
+      "step": 4037
+    },
+    {
+      "epoch": 0.36552910292387075,
+      "grad_norm": 0.9615750908851624,
+      "learning_rate": 0.00013156490369471027,
+      "loss": 2.8857,
+      "step": 4038
+    },
+    {
+      "epoch": 0.3656196252376211,
+      "grad_norm": 0.8426896333694458,
+      "learning_rate": 0.00013153479117756608,
+      "loss": 2.8606,
+      "step": 4039
+    },
+    {
+      "epoch": 0.36571014755137143,
+      "grad_norm": 1.0978776216506958,
+      "learning_rate": 0.0001315046754848693,
+      "loss": 2.8393,
+      "step": 4040
+    },
+    {
+      "epoch": 0.3658006698651218,
+      "grad_norm": 0.8376330137252808,
+      "learning_rate": 0.00013147455661965255,
+      "loss": 2.9575,
+      "step": 4041
+    },
+    {
+      "epoch": 0.3658911921788721,
+      "grad_norm": 0.8953753709793091,
+      "learning_rate": 0.00013144443458494882,
+      "loss": 2.8782,
+      "step": 4042
+    },
+    {
+      "epoch": 0.36598171449262246,
+      "grad_norm": 0.9642663598060608,
+      "learning_rate": 0.00013141430938379134,
+      "loss": 2.8789,
+      "step": 4043
+    },
+    {
+      "epoch": 0.3660722368063728,
+      "grad_norm": 1.0093460083007812,
+      "learning_rate": 0.00013138418101921385,
+      "loss": 2.9142,
+      "step": 4044
+    },
+    {
+      "epoch": 0.36616275912012314,
+      "grad_norm": 0.83012455701828,
+      "learning_rate": 0.00013135404949425015,
+      "loss": 2.8987,
+      "step": 4045
+    },
+    {
+      "epoch": 0.3662532814338734,
+      "grad_norm": 1.1322907209396362,
+      "learning_rate": 0.00013132391481193451,
+      "loss": 2.8769,
+      "step": 4046
+    },
+    {
+      "epoch": 0.36634380374762376,
+      "grad_norm": 1.0075737237930298,
+      "learning_rate": 0.0001312937769753015,
+      "loss": 2.8329,
+      "step": 4047
+    },
+    {
+      "epoch": 0.3664343260613741,
+      "grad_norm": 1.0125104188919067,
+      "learning_rate": 0.00013126363598738603,
+      "loss": 2.8341,
+      "step": 4048
+    },
+    {
+      "epoch": 0.36652484837512445,
+      "grad_norm": 0.9636220932006836,
+      "learning_rate": 0.00013123349185122327,
+      "loss": 2.8841,
+      "step": 4049
+    },
+    {
+      "epoch": 0.3666153706888748,
+      "grad_norm": 0.9219857454299927,
+      "learning_rate": 0.0001312033445698487,
+      "loss": 2.9358,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3667058930026251,
+      "grad_norm": 0.8947569727897644,
+      "learning_rate": 0.00013117319414629824,
+      "loss": 2.9103,
+      "step": 4051
+    },
+    {
+      "epoch": 0.36679641531637547,
+      "grad_norm": 0.9168105125427246,
+      "learning_rate": 0.00013114304058360797,
+      "loss": 2.8693,
+      "step": 4052
+    },
+    {
+      "epoch": 0.3668869376301258,
+      "grad_norm": 0.8657991886138916,
+      "learning_rate": 0.00013111288388481434,
+      "loss": 2.9258,
+      "step": 4053
+    },
+    {
+      "epoch": 0.36697745994387615,
+      "grad_norm": 0.8736631274223328,
+      "learning_rate": 0.00013108272405295415,
+      "loss": 2.8122,
+      "step": 4054
+    },
+    {
+      "epoch": 0.3670679822576265,
+      "grad_norm": 0.8283593654632568,
+      "learning_rate": 0.0001310525610910645,
+      "loss": 2.9052,
+      "step": 4055
+    },
+    {
+      "epoch": 0.36715850457137683,
+      "grad_norm": 0.8502643704414368,
+      "learning_rate": 0.00013102239500218282,
+      "loss": 2.892,
+      "step": 4056
+    },
+    {
+      "epoch": 0.36724902688512717,
+      "grad_norm": 0.976543128490448,
+      "learning_rate": 0.0001309922257893468,
+      "loss": 2.9142,
+      "step": 4057
+    },
+    {
+      "epoch": 0.3673395491988775,
+      "grad_norm": 0.8439890742301941,
+      "learning_rate": 0.00013096205345559448,
+      "loss": 2.9241,
+      "step": 4058
+    },
+    {
+      "epoch": 0.36743007151262785,
+      "grad_norm": 0.8744709491729736,
+      "learning_rate": 0.00013093187800396425,
+      "loss": 2.8804,
+      "step": 4059
+    },
+    {
+      "epoch": 0.3675205938263782,
+      "grad_norm": 0.8625809550285339,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 2.8871,
+      "step": 4060
+    },
+    {
+      "epoch": 0.36761111614012854,
+      "grad_norm": 0.9283736348152161,
+      "learning_rate": 0.00013087151775922494,
+      "loss": 2.9417,
+      "step": 4061
+    },
+    {
+      "epoch": 0.3677016384538789,
+      "grad_norm": 0.8103830814361572,
+      "learning_rate": 0.00013084133297219417,
+      "loss": 2.7827,
+      "step": 4062
+    },
+    {
+      "epoch": 0.3677921607676292,
+      "grad_norm": 0.8377107381820679,
+      "learning_rate": 0.00013081114507944197,
+      "loss": 2.8471,
+      "step": 4063
+    },
+    {
+      "epoch": 0.36788268308137956,
+      "grad_norm": 0.8858683705329895,
+      "learning_rate": 0.00013078095408400835,
+      "loss": 2.8529,
+      "step": 4064
+    },
+    {
+      "epoch": 0.3679732053951299,
+      "grad_norm": 0.8910707235336304,
+      "learning_rate": 0.00013075075998893345,
+      "loss": 2.9123,
+      "step": 4065
+    },
+    {
+      "epoch": 0.36806372770888024,
+      "grad_norm": 0.8349747657775879,
+      "learning_rate": 0.00013072056279725788,
+      "loss": 2.837,
+      "step": 4066
+    },
+    {
+      "epoch": 0.3681542500226306,
+      "grad_norm": 1.0403574705123901,
+      "learning_rate": 0.0001306903625120225,
+      "loss": 2.898,
+      "step": 4067
+    },
+    {
+      "epoch": 0.3682447723363809,
+      "grad_norm": 0.9942634105682373,
+      "learning_rate": 0.00013066015913626843,
+      "loss": 2.8599,
+      "step": 4068
+    },
+    {
+      "epoch": 0.36833529465013126,
+      "grad_norm": 0.812485933303833,
+      "learning_rate": 0.0001306299526730372,
+      "loss": 2.8962,
+      "step": 4069
+    },
+    {
+      "epoch": 0.3684258169638816,
+      "grad_norm": 1.0433251857757568,
+      "learning_rate": 0.00013059974312537053,
+      "loss": 2.918,
+      "step": 4070
+    },
+    {
+      "epoch": 0.36851633927763194,
+      "grad_norm": 0.9210599064826965,
+      "learning_rate": 0.00013056953049631057,
+      "loss": 2.8613,
+      "step": 4071
+    },
+    {
+      "epoch": 0.3686068615913823,
+      "grad_norm": 0.8464783430099487,
+      "learning_rate": 0.00013053931478889975,
+      "loss": 2.8794,
+      "step": 4072
+    },
+    {
+      "epoch": 0.3686973839051326,
+      "grad_norm": 1.1169415712356567,
+      "learning_rate": 0.00013050909600618077,
+      "loss": 2.9114,
+      "step": 4073
+    },
+    {
+      "epoch": 0.36878790621888297,
+      "grad_norm": 0.8666006922721863,
+      "learning_rate": 0.00013047887415119664,
+      "loss": 2.8093,
+      "step": 4074
+    },
+    {
+      "epoch": 0.3688784285326333,
+      "grad_norm": 0.8614941835403442,
+      "learning_rate": 0.0001304486492269907,
+      "loss": 2.9002,
+      "step": 4075
+    },
+    {
+      "epoch": 0.36896895084638365,
+      "grad_norm": 0.951631486415863,
+      "learning_rate": 0.00013041842123660667,
+      "loss": 2.8685,
+      "step": 4076
+    },
+    {
+      "epoch": 0.369059473160134,
+      "grad_norm": 0.920071542263031,
+      "learning_rate": 0.0001303881901830884,
+      "loss": 2.8821,
+      "step": 4077
+    },
+    {
+      "epoch": 0.36914999547388433,
+      "grad_norm": 0.8674535751342773,
+      "learning_rate": 0.00013035795606948023,
+      "loss": 2.9443,
+      "step": 4078
+    },
+    {
+      "epoch": 0.36924051778763467,
+      "grad_norm": 1.0542681217193604,
+      "learning_rate": 0.00013032771889882674,
+      "loss": 2.9529,
+      "step": 4079
+    },
+    {
+      "epoch": 0.369331040101385,
+      "grad_norm": 0.8281850218772888,
+      "learning_rate": 0.00013029747867417276,
+      "loss": 2.8641,
+      "step": 4080
+    },
+    {
+      "epoch": 0.36942156241513535,
+      "grad_norm": 0.850652277469635,
+      "learning_rate": 0.00013026723539856348,
+      "loss": 2.8656,
+      "step": 4081
+    },
+    {
+      "epoch": 0.3695120847288857,
+      "grad_norm": 0.8725227117538452,
+      "learning_rate": 0.00013023698907504446,
+      "loss": 2.9565,
+      "step": 4082
+    },
+    {
+      "epoch": 0.36960260704263603,
+      "grad_norm": 0.832707405090332,
+      "learning_rate": 0.0001302067397066615,
+      "loss": 2.8305,
+      "step": 4083
+    },
+    {
+      "epoch": 0.3696931293563864,
+      "grad_norm": 0.8224778771400452,
+      "learning_rate": 0.0001301764872964606,
+      "loss": 2.9179,
+      "step": 4084
+    },
+    {
+      "epoch": 0.3697836516701367,
+      "grad_norm": 0.8339827060699463,
+      "learning_rate": 0.0001301462318474883,
+      "loss": 2.8407,
+      "step": 4085
+    },
+    {
+      "epoch": 0.36987417398388706,
+      "grad_norm": 0.8362826108932495,
+      "learning_rate": 0.00013011597336279126,
+      "loss": 2.7989,
+      "step": 4086
+    },
+    {
+      "epoch": 0.36996469629763734,
+      "grad_norm": 0.8587050437927246,
+      "learning_rate": 0.00013008571184541657,
+      "loss": 2.8354,
+      "step": 4087
+    },
+    {
+      "epoch": 0.3700552186113877,
+      "grad_norm": 0.885046660900116,
+      "learning_rate": 0.00013005544729841146,
+      "loss": 2.8989,
+      "step": 4088
+    },
+    {
+      "epoch": 0.370145740925138,
+      "grad_norm": 0.9273785948753357,
+      "learning_rate": 0.00013002517972482368,
+      "loss": 2.8913,
+      "step": 4089
+    },
+    {
+      "epoch": 0.37023626323888836,
+      "grad_norm": 0.8214293122291565,
+      "learning_rate": 0.00012999490912770107,
+      "loss": 2.8843,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3703267855526387,
+      "grad_norm": 0.9375723004341125,
+      "learning_rate": 0.00012996463551009197,
+      "loss": 2.8452,
+      "step": 4091
+    },
+    {
+      "epoch": 0.37041730786638905,
+      "grad_norm": 0.8647713661193848,
+      "learning_rate": 0.00012993435887504487,
+      "loss": 2.9241,
+      "step": 4092
+    },
+    {
+      "epoch": 0.3705078301801394,
+      "grad_norm": 0.8443884253501892,
+      "learning_rate": 0.00012990407922560868,
+      "loss": 2.8765,
+      "step": 4093
+    },
+    {
+      "epoch": 0.3705983524938897,
+      "grad_norm": 0.9264103770256042,
+      "learning_rate": 0.0001298737965648325,
+      "loss": 2.8917,
+      "step": 4094
+    },
+    {
+      "epoch": 0.37068887480764007,
+      "grad_norm": 0.7921240925788879,
+      "learning_rate": 0.00012984351089576586,
+      "loss": 2.7979,
+      "step": 4095
+    },
+    {
+      "epoch": 0.3707793971213904,
+      "grad_norm": 0.9128064513206482,
+      "learning_rate": 0.00012981322222145846,
+      "loss": 2.9056,
+      "step": 4096
+    },
+    {
+      "epoch": 0.37086991943514075,
+      "grad_norm": 1.012534499168396,
+      "learning_rate": 0.0001297829305449604,
+      "loss": 3.0145,
+      "step": 4097
+    },
+    {
+      "epoch": 0.3709604417488911,
+      "grad_norm": 0.9095678925514221,
+      "learning_rate": 0.00012975263586932208,
+      "loss": 2.8566,
+      "step": 4098
+    },
+    {
+      "epoch": 0.37105096406264143,
+      "grad_norm": 0.8481354713439941,
+      "learning_rate": 0.0001297223381975941,
+      "loss": 2.8792,
+      "step": 4099
+    },
+    {
+      "epoch": 0.37114148637639177,
+      "grad_norm": 0.9588367938995361,
+      "learning_rate": 0.0001296920375328275,
+      "loss": 2.8702,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3712320086901421,
+      "grad_norm": 0.9041178822517395,
+      "learning_rate": 0.0001296617338780735,
+      "loss": 2.8923,
+      "step": 4101
+    },
+    {
+      "epoch": 0.37132253100389245,
+      "grad_norm": 0.9123795628547668,
+      "learning_rate": 0.00012963142723638378,
+      "loss": 2.9847,
+      "step": 4102
+    },
+    {
+      "epoch": 0.3714130533176428,
+      "grad_norm": 0.8965758085250854,
+      "learning_rate": 0.00012960111761081008,
+      "loss": 2.8985,
+      "step": 4103
+    },
+    {
+      "epoch": 0.37150357563139313,
+      "grad_norm": 0.8802927732467651,
+      "learning_rate": 0.00012957080500440468,
+      "loss": 2.8288,
+      "step": 4104
+    },
+    {
+      "epoch": 0.3715940979451435,
+      "grad_norm": 0.8850862383842468,
+      "learning_rate": 0.00012954048942022002,
+      "loss": 2.878,
+      "step": 4105
+    },
+    {
+      "epoch": 0.3716846202588938,
+      "grad_norm": 0.8778537511825562,
+      "learning_rate": 0.0001295101708613089,
+      "loss": 2.8459,
+      "step": 4106
+    },
+    {
+      "epoch": 0.37177514257264416,
+      "grad_norm": 0.8377190232276917,
+      "learning_rate": 0.00012947984933072435,
+      "loss": 2.9462,
+      "step": 4107
+    },
+    {
+      "epoch": 0.3718656648863945,
+      "grad_norm": 1.0077134370803833,
+      "learning_rate": 0.00012944952483151978,
+      "loss": 2.8496,
+      "step": 4108
+    },
+    {
+      "epoch": 0.37195618720014484,
+      "grad_norm": 0.8665797114372253,
+      "learning_rate": 0.00012941919736674887,
+      "loss": 2.8963,
+      "step": 4109
+    },
+    {
+      "epoch": 0.3720467095138952,
+      "grad_norm": 0.8284353017807007,
+      "learning_rate": 0.0001293888669394656,
+      "loss": 2.9595,
+      "step": 4110
+    },
+    {
+      "epoch": 0.3721372318276455,
+      "grad_norm": 0.8614863753318787,
+      "learning_rate": 0.00012935853355272424,
+      "loss": 2.878,
+      "step": 4111
+    },
+    {
+      "epoch": 0.37222775414139586,
+      "grad_norm": 0.8768258690834045,
+      "learning_rate": 0.00012932819720957935,
+      "loss": 2.8506,
+      "step": 4112
+    },
+    {
+      "epoch": 0.3723182764551462,
+      "grad_norm": 0.8659483194351196,
+      "learning_rate": 0.0001292978579130858,
+      "loss": 2.883,
+      "step": 4113
+    },
+    {
+      "epoch": 0.37240879876889654,
+      "grad_norm": 0.8106630444526672,
+      "learning_rate": 0.00012926751566629875,
+      "loss": 2.8823,
+      "step": 4114
+    },
+    {
+      "epoch": 0.3724993210826469,
+      "grad_norm": 0.8505522012710571,
+      "learning_rate": 0.00012923717047227368,
+      "loss": 2.878,
+      "step": 4115
+    },
+    {
+      "epoch": 0.3725898433963972,
+      "grad_norm": 0.8515750765800476,
+      "learning_rate": 0.00012920682233406633,
+      "loss": 2.8721,
+      "step": 4116
+    },
+    {
+      "epoch": 0.37268036571014757,
+      "grad_norm": 0.893912672996521,
+      "learning_rate": 0.0001291764712547328,
+      "loss": 2.9565,
+      "step": 4117
+    },
+    {
+      "epoch": 0.3727708880238979,
+      "grad_norm": 0.937900185585022,
+      "learning_rate": 0.00012914611723732942,
+      "loss": 2.8862,
+      "step": 4118
+    },
+    {
+      "epoch": 0.37286141033764825,
+      "grad_norm": 0.921802818775177,
+      "learning_rate": 0.00012911576028491278,
+      "loss": 2.9014,
+      "step": 4119
+    },
+    {
+      "epoch": 0.3729519326513986,
+      "grad_norm": 0.8836219310760498,
+      "learning_rate": 0.0001290854004005399,
+      "loss": 2.9419,
+      "step": 4120
+    },
+    {
+      "epoch": 0.37304245496514893,
+      "grad_norm": 0.8305163383483887,
+      "learning_rate": 0.00012905503758726803,
+      "loss": 2.8919,
+      "step": 4121
+    },
+    {
+      "epoch": 0.37313297727889927,
+      "grad_norm": 0.992817759513855,
+      "learning_rate": 0.00012902467184815463,
+      "loss": 2.8757,
+      "step": 4122
+    },
+    {
+      "epoch": 0.3732234995926496,
+      "grad_norm": 0.8104541897773743,
+      "learning_rate": 0.00012899430318625757,
+      "loss": 2.8196,
+      "step": 4123
+    },
+    {
+      "epoch": 0.37331402190639995,
+      "grad_norm": 0.8181300163269043,
+      "learning_rate": 0.000128963931604635,
+      "loss": 2.8211,
+      "step": 4124
+    },
+    {
+      "epoch": 0.3734045442201503,
+      "grad_norm": 0.8287342190742493,
+      "learning_rate": 0.0001289335571063453,
+      "loss": 2.8883,
+      "step": 4125
+    },
+    {
+      "epoch": 0.37349506653390063,
+      "grad_norm": 0.8907531499862671,
+      "learning_rate": 0.00012890317969444716,
+      "loss": 2.926,
+      "step": 4126
+    },
+    {
+      "epoch": 0.373585588847651,
+      "grad_norm": 0.8196613192558289,
+      "learning_rate": 0.00012887279937199962,
+      "loss": 2.8536,
+      "step": 4127
+    },
+    {
+      "epoch": 0.37367611116140126,
+      "grad_norm": 0.84673011302948,
+      "learning_rate": 0.00012884241614206202,
+      "loss": 2.8441,
+      "step": 4128
+    },
+    {
+      "epoch": 0.3737666334751516,
+      "grad_norm": 0.9478344321250916,
+      "learning_rate": 0.00012881203000769388,
+      "loss": 2.8454,
+      "step": 4129
+    },
+    {
+      "epoch": 0.37385715578890194,
+      "grad_norm": 0.8046672344207764,
+      "learning_rate": 0.0001287816409719551,
+      "loss": 2.8328,
+      "step": 4130
+    },
+    {
+      "epoch": 0.3739476781026523,
+      "grad_norm": 0.924982488155365,
+      "learning_rate": 0.00012875124903790586,
+      "loss": 2.9235,
+      "step": 4131
+    },
+    {
+      "epoch": 0.3740382004164026,
+      "grad_norm": 0.8310782313346863,
+      "learning_rate": 0.00012872085420860665,
+      "loss": 2.9017,
+      "step": 4132
+    },
+    {
+      "epoch": 0.37412872273015296,
+      "grad_norm": 0.8438023328781128,
+      "learning_rate": 0.00012869045648711817,
+      "loss": 2.931,
+      "step": 4133
+    },
+    {
+      "epoch": 0.3742192450439033,
+      "grad_norm": 0.8599631786346436,
+      "learning_rate": 0.0001286600558765015,
+      "loss": 2.8513,
+      "step": 4134
+    },
+    {
+      "epoch": 0.37430976735765364,
+      "grad_norm": 0.9266738891601562,
+      "learning_rate": 0.00012862965237981803,
+      "loss": 2.9311,
+      "step": 4135
+    },
+    {
+      "epoch": 0.374400289671404,
+      "grad_norm": 0.918306827545166,
+      "learning_rate": 0.00012859924600012936,
+      "loss": 2.8969,
+      "step": 4136
+    },
+    {
+      "epoch": 0.3744908119851543,
+      "grad_norm": 0.8407638072967529,
+      "learning_rate": 0.00012856883674049736,
+      "loss": 2.8794,
+      "step": 4137
+    },
+    {
+      "epoch": 0.37458133429890467,
+      "grad_norm": 0.8753742575645447,
+      "learning_rate": 0.00012853842460398428,
+      "loss": 2.8764,
+      "step": 4138
+    },
+    {
+      "epoch": 0.374671856612655,
+      "grad_norm": 0.9728417992591858,
+      "learning_rate": 0.00012850800959365266,
+      "loss": 2.9482,
+      "step": 4139
+    },
+    {
+      "epoch": 0.37476237892640535,
+      "grad_norm": 0.832207977771759,
+      "learning_rate": 0.00012847759171256523,
+      "loss": 2.9202,
+      "step": 4140
+    },
+    {
+      "epoch": 0.3748529012401557,
+      "grad_norm": 0.8750010132789612,
+      "learning_rate": 0.00012844717096378508,
+      "loss": 2.9017,
+      "step": 4141
+    },
+    {
+      "epoch": 0.37494342355390603,
+      "grad_norm": 0.9617355465888977,
+      "learning_rate": 0.0001284167473503756,
+      "loss": 2.8529,
+      "step": 4142
+    },
+    {
+      "epoch": 0.37503394586765637,
+      "grad_norm": 0.8699533939361572,
+      "learning_rate": 0.00012838632087540047,
+      "loss": 2.8743,
+      "step": 4143
+    },
+    {
+      "epoch": 0.3751244681814067,
+      "grad_norm": 0.9251249432563782,
+      "learning_rate": 0.00012835589154192357,
+      "loss": 2.819,
+      "step": 4144
+    },
+    {
+      "epoch": 0.37521499049515705,
+      "grad_norm": 0.9185225963592529,
+      "learning_rate": 0.0001283254593530092,
+      "loss": 2.8883,
+      "step": 4145
+    },
+    {
+      "epoch": 0.3753055128089074,
+      "grad_norm": 0.92988520860672,
+      "learning_rate": 0.0001282950243117218,
+      "loss": 2.9043,
+      "step": 4146
+    },
+    {
+      "epoch": 0.37539603512265773,
+      "grad_norm": 0.8868668675422668,
+      "learning_rate": 0.0001282645864211263,
+      "loss": 2.9247,
+      "step": 4147
+    },
+    {
+      "epoch": 0.3754865574364081,
+      "grad_norm": 0.9574105739593506,
+      "learning_rate": 0.00012823414568428768,
+      "loss": 2.8973,
+      "step": 4148
+    },
+    {
+      "epoch": 0.3755770797501584,
+      "grad_norm": 0.936896026134491,
+      "learning_rate": 0.00012820370210427133,
+      "loss": 2.9021,
+      "step": 4149
+    },
+    {
+      "epoch": 0.37566760206390876,
+      "grad_norm": 0.8531725406646729,
+      "learning_rate": 0.00012817325568414297,
+      "loss": 2.7698,
+      "step": 4150
+    },
+    {
+      "epoch": 0.3757581243776591,
+      "grad_norm": 0.9539023041725159,
+      "learning_rate": 0.00012814280642696856,
+      "loss": 2.9243,
+      "step": 4151
+    },
+    {
+      "epoch": 0.37584864669140944,
+      "grad_norm": 0.8933930397033691,
+      "learning_rate": 0.00012811235433581427,
+      "loss": 2.8733,
+      "step": 4152
+    },
+    {
+      "epoch": 0.3759391690051598,
+      "grad_norm": 0.9392674565315247,
+      "learning_rate": 0.00012808189941374668,
+      "loss": 2.8682,
+      "step": 4153
+    },
+    {
+      "epoch": 0.3760296913189101,
+      "grad_norm": 0.8223789930343628,
+      "learning_rate": 0.00012805144166383263,
+      "loss": 2.861,
+      "step": 4154
+    },
+    {
+      "epoch": 0.37612021363266046,
+      "grad_norm": 0.9169121980667114,
+      "learning_rate": 0.00012802098108913914,
+      "loss": 2.8804,
+      "step": 4155
+    },
+    {
+      "epoch": 0.3762107359464108,
+      "grad_norm": 0.9235851168632507,
+      "learning_rate": 0.00012799051769273362,
+      "loss": 2.8841,
+      "step": 4156
+    },
+    {
+      "epoch": 0.37630125826016114,
+      "grad_norm": 0.8986383080482483,
+      "learning_rate": 0.00012796005147768376,
+      "loss": 2.8593,
+      "step": 4157
+    },
+    {
+      "epoch": 0.3763917805739115,
+      "grad_norm": 0.9925276637077332,
+      "learning_rate": 0.00012792958244705745,
+      "loss": 2.9139,
+      "step": 4158
+    },
+    {
+      "epoch": 0.3764823028876618,
+      "grad_norm": 0.8893771171569824,
+      "learning_rate": 0.00012789911060392294,
+      "loss": 2.7976,
+      "step": 4159
+    },
+    {
+      "epoch": 0.37657282520141216,
+      "grad_norm": 0.8771906495094299,
+      "learning_rate": 0.0001278686359513488,
+      "loss": 2.8583,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3766633475151625,
+      "grad_norm": 0.9729670882225037,
+      "learning_rate": 0.00012783815849240374,
+      "loss": 2.8408,
+      "step": 4161
+    },
+    {
+      "epoch": 0.37675386982891285,
+      "grad_norm": 0.9033991098403931,
+      "learning_rate": 0.0001278076782301569,
+      "loss": 2.8587,
+      "step": 4162
+    },
+    {
+      "epoch": 0.3768443921426632,
+      "grad_norm": 0.9147695302963257,
+      "learning_rate": 0.0001277771951676776,
+      "loss": 2.8858,
+      "step": 4163
+    },
+    {
+      "epoch": 0.37693491445641353,
+      "grad_norm": 0.849720299243927,
+      "learning_rate": 0.00012774670930803554,
+      "loss": 2.8747,
+      "step": 4164
+    },
+    {
+      "epoch": 0.37702543677016387,
+      "grad_norm": 0.8087278604507446,
+      "learning_rate": 0.0001277162206543006,
+      "loss": 2.8786,
+      "step": 4165
+    },
+    {
+      "epoch": 0.3771159590839142,
+      "grad_norm": 0.8310896158218384,
+      "learning_rate": 0.00012768572920954298,
+      "loss": 2.9031,
+      "step": 4166
+    },
+    {
+      "epoch": 0.37720648139766455,
+      "grad_norm": 0.8340770602226257,
+      "learning_rate": 0.00012765523497683322,
+      "loss": 2.8748,
+      "step": 4167
+    },
+    {
+      "epoch": 0.3772970037114149,
+      "grad_norm": 0.8991544842720032,
+      "learning_rate": 0.00012762473795924204,
+      "loss": 2.9399,
+      "step": 4168
+    },
+    {
+      "epoch": 0.3773875260251652,
+      "grad_norm": 0.9005395174026489,
+      "learning_rate": 0.0001275942381598405,
+      "loss": 2.9311,
+      "step": 4169
+    },
+    {
+      "epoch": 0.3774780483389155,
+      "grad_norm": 0.9556332230567932,
+      "learning_rate": 0.0001275637355816999,
+      "loss": 2.8791,
+      "step": 4170
+    },
+    {
+      "epoch": 0.37756857065266586,
+      "grad_norm": 0.8282430768013,
+      "learning_rate": 0.00012753323022789194,
+      "loss": 2.9248,
+      "step": 4171
+    },
+    {
+      "epoch": 0.3776590929664162,
+      "grad_norm": 0.8877475261688232,
+      "learning_rate": 0.00012750272210148843,
+      "loss": 2.8638,
+      "step": 4172
+    },
+    {
+      "epoch": 0.37774961528016654,
+      "grad_norm": 0.8167067170143127,
+      "learning_rate": 0.00012747221120556153,
+      "loss": 2.9017,
+      "step": 4173
+    },
+    {
+      "epoch": 0.3778401375939169,
+      "grad_norm": 0.8058724403381348,
+      "learning_rate": 0.00012744169754318375,
+      "loss": 2.8457,
+      "step": 4174
+    },
+    {
+      "epoch": 0.3779306599076672,
+      "grad_norm": 0.812492847442627,
+      "learning_rate": 0.00012741118111742777,
+      "loss": 2.8472,
+      "step": 4175
+    },
+    {
+      "epoch": 0.37802118222141756,
+      "grad_norm": 0.9392409920692444,
+      "learning_rate": 0.0001273806619313666,
+      "loss": 2.851,
+      "step": 4176
+    },
+    {
+      "epoch": 0.3781117045351679,
+      "grad_norm": 0.877270519733429,
+      "learning_rate": 0.0001273501399880735,
+      "loss": 2.8653,
+      "step": 4177
+    },
+    {
+      "epoch": 0.37820222684891824,
+      "grad_norm": 0.9208534359931946,
+      "learning_rate": 0.00012731961529062211,
+      "loss": 2.8362,
+      "step": 4178
+    },
+    {
+      "epoch": 0.3782927491626686,
+      "grad_norm": 0.9426068663597107,
+      "learning_rate": 0.00012728908784208616,
+      "loss": 2.8833,
+      "step": 4179
+    },
+    {
+      "epoch": 0.3783832714764189,
+      "grad_norm": 0.9093818068504333,
+      "learning_rate": 0.0001272585576455398,
+      "loss": 2.9297,
+      "step": 4180
+    },
+    {
+      "epoch": 0.37847379379016927,
+      "grad_norm": 0.953253984451294,
+      "learning_rate": 0.00012722802470405744,
+      "loss": 2.837,
+      "step": 4181
+    },
+    {
+      "epoch": 0.3785643161039196,
+      "grad_norm": 0.9242717623710632,
+      "learning_rate": 0.00012719748902071374,
+      "loss": 2.8074,
+      "step": 4182
+    },
+    {
+      "epoch": 0.37865483841766995,
+      "grad_norm": 0.8428967595100403,
+      "learning_rate": 0.00012716695059858366,
+      "loss": 2.8661,
+      "step": 4183
+    },
+    {
+      "epoch": 0.3787453607314203,
+      "grad_norm": 1.0141724348068237,
+      "learning_rate": 0.00012713640944074238,
+      "loss": 2.8884,
+      "step": 4184
+    },
+    {
+      "epoch": 0.37883588304517063,
+      "grad_norm": 0.8837010860443115,
+      "learning_rate": 0.0001271058655502654,
+      "loss": 2.903,
+      "step": 4185
+    },
+    {
+      "epoch": 0.37892640535892097,
+      "grad_norm": 0.8637431859970093,
+      "learning_rate": 0.00012707531893022854,
+      "loss": 2.8396,
+      "step": 4186
+    },
+    {
+      "epoch": 0.3790169276726713,
+      "grad_norm": 0.9034766554832458,
+      "learning_rate": 0.00012704476958370778,
+      "loss": 2.8339,
+      "step": 4187
+    },
+    {
+      "epoch": 0.37910744998642165,
+      "grad_norm": 0.8125141859054565,
+      "learning_rate": 0.00012701421751377945,
+      "loss": 2.8675,
+      "step": 4188
+    },
+    {
+      "epoch": 0.379197972300172,
+      "grad_norm": 0.9025378227233887,
+      "learning_rate": 0.00012698366272352019,
+      "loss": 2.8319,
+      "step": 4189
+    },
+    {
+      "epoch": 0.37928849461392233,
+      "grad_norm": 0.9068765044212341,
+      "learning_rate": 0.0001269531052160068,
+      "loss": 2.8712,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3793790169276727,
+      "grad_norm": 0.9077470302581787,
+      "learning_rate": 0.00012692254499431646,
+      "loss": 2.8898,
+      "step": 4191
+    },
+    {
+      "epoch": 0.379469539241423,
+      "grad_norm": 0.8503580093383789,
+      "learning_rate": 0.00012689198206152657,
+      "loss": 2.8703,
+      "step": 4192
+    },
+    {
+      "epoch": 0.37956006155517336,
+      "grad_norm": 0.8890337347984314,
+      "learning_rate": 0.00012686141642071486,
+      "loss": 2.8733,
+      "step": 4193
+    },
+    {
+      "epoch": 0.3796505838689237,
+      "grad_norm": 0.7975422739982605,
+      "learning_rate": 0.0001268308480749592,
+      "loss": 2.8605,
+      "step": 4194
+    },
+    {
+      "epoch": 0.37974110618267404,
+      "grad_norm": 0.8244346380233765,
+      "learning_rate": 0.0001268002770273379,
+      "loss": 2.8681,
+      "step": 4195
+    },
+    {
+      "epoch": 0.3798316284964244,
+      "grad_norm": 0.8913033604621887,
+      "learning_rate": 0.00012676970328092944,
+      "loss": 2.9361,
+      "step": 4196
+    },
+    {
+      "epoch": 0.3799221508101747,
+      "grad_norm": 0.8553683161735535,
+      "learning_rate": 0.0001267391268388126,
+      "loss": 2.8772,
+      "step": 4197
+    },
+    {
+      "epoch": 0.38001267312392506,
+      "grad_norm": 0.8335919976234436,
+      "learning_rate": 0.0001267085477040664,
+      "loss": 2.8831,
+      "step": 4198
+    },
+    {
+      "epoch": 0.3801031954376754,
+      "grad_norm": 0.8713265061378479,
+      "learning_rate": 0.00012667796587977016,
+      "loss": 2.9399,
+      "step": 4199
+    },
+    {
+      "epoch": 0.38019371775142574,
+      "grad_norm": 0.94398033618927,
+      "learning_rate": 0.00012664738136900348,
+      "loss": 2.8853,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3802842400651761,
+      "grad_norm": 0.7876652479171753,
+      "learning_rate": 0.0001266167941748463,
+      "loss": 2.8678,
+      "step": 4201
+    },
+    {
+      "epoch": 0.3803747623789264,
+      "grad_norm": 0.8114630579948425,
+      "learning_rate": 0.00012658620430037863,
+      "loss": 2.8559,
+      "step": 4202
+    },
+    {
+      "epoch": 0.38046528469267676,
+      "grad_norm": 0.9380217790603638,
+      "learning_rate": 0.00012655561174868088,
+      "loss": 2.8233,
+      "step": 4203
+    },
+    {
+      "epoch": 0.3805558070064271,
+      "grad_norm": 0.8938210606575012,
+      "learning_rate": 0.00012652501652283377,
+      "loss": 2.8636,
+      "step": 4204
+    },
+    {
+      "epoch": 0.38064632932017745,
+      "grad_norm": 0.9160013794898987,
+      "learning_rate": 0.00012649441862591827,
+      "loss": 2.8823,
+      "step": 4205
+    },
+    {
+      "epoch": 0.3807368516339278,
+      "grad_norm": 1.035874605178833,
+      "learning_rate": 0.0001264638180610155,
+      "loss": 2.9009,
+      "step": 4206
+    },
+    {
+      "epoch": 0.3808273739476781,
+      "grad_norm": 0.8780242204666138,
+      "learning_rate": 0.00012643321483120698,
+      "loss": 2.8823,
+      "step": 4207
+    },
+    {
+      "epoch": 0.38091789626142847,
+      "grad_norm": 0.9666605591773987,
+      "learning_rate": 0.00012640260893957445,
+      "loss": 2.8469,
+      "step": 4208
+    },
+    {
+      "epoch": 0.3810084185751788,
+      "grad_norm": 0.8629251718521118,
+      "learning_rate": 0.0001263720003891999,
+      "loss": 2.8137,
+      "step": 4209
+    },
+    {
+      "epoch": 0.3810989408889291,
+      "grad_norm": 0.9518691301345825,
+      "learning_rate": 0.00012634138918316568,
+      "loss": 2.9227,
+      "step": 4210
+    },
+    {
+      "epoch": 0.38118946320267943,
+      "grad_norm": 0.8704745769500732,
+      "learning_rate": 0.00012631077532455426,
+      "loss": 2.8615,
+      "step": 4211
+    },
+    {
+      "epoch": 0.3812799855164298,
+      "grad_norm": 0.9694190621376038,
+      "learning_rate": 0.0001262801588164485,
+      "loss": 2.9558,
+      "step": 4212
+    },
+    {
+      "epoch": 0.3813705078301801,
+      "grad_norm": 1.1124874353408813,
+      "learning_rate": 0.00012624953966193145,
+      "loss": 2.9274,
+      "step": 4213
+    },
+    {
+      "epoch": 0.38146103014393046,
+      "grad_norm": 0.9309592247009277,
+      "learning_rate": 0.00012621891786408648,
+      "loss": 2.8492,
+      "step": 4214
+    },
+    {
+      "epoch": 0.3815515524576808,
+      "grad_norm": 0.910110592842102,
+      "learning_rate": 0.00012618829342599718,
+      "loss": 2.8612,
+      "step": 4215
+    },
+    {
+      "epoch": 0.38164207477143114,
+      "grad_norm": 0.9568766355514526,
+      "learning_rate": 0.0001261576663507475,
+      "loss": 2.8755,
+      "step": 4216
+    },
+    {
+      "epoch": 0.3817325970851815,
+      "grad_norm": 0.9147237539291382,
+      "learning_rate": 0.0001261270366414215,
+      "loss": 2.8692,
+      "step": 4217
+    },
+    {
+      "epoch": 0.3818231193989318,
+      "grad_norm": 0.9557003378868103,
+      "learning_rate": 0.0001260964043011036,
+      "loss": 2.9074,
+      "step": 4218
+    },
+    {
+      "epoch": 0.38191364171268216,
+      "grad_norm": 0.9502111077308655,
+      "learning_rate": 0.00012606576933287854,
+      "loss": 2.9124,
+      "step": 4219
+    },
+    {
+      "epoch": 0.3820041640264325,
+      "grad_norm": 0.9677018523216248,
+      "learning_rate": 0.0001260351317398312,
+      "loss": 2.8416,
+      "step": 4220
+    },
+    {
+      "epoch": 0.38209468634018284,
+      "grad_norm": 0.9683780074119568,
+      "learning_rate": 0.00012600449152504683,
+      "loss": 2.9383,
+      "step": 4221
+    },
+    {
+      "epoch": 0.3821852086539332,
+      "grad_norm": 0.8879417777061462,
+      "learning_rate": 0.00012597384869161084,
+      "loss": 2.9172,
+      "step": 4222
+    },
+    {
+      "epoch": 0.3822757309676835,
+      "grad_norm": 0.9438839554786682,
+      "learning_rate": 0.00012594320324260905,
+      "loss": 2.8511,
+      "step": 4223
+    },
+    {
+      "epoch": 0.38236625328143387,
+      "grad_norm": 0.9300707578659058,
+      "learning_rate": 0.00012591255518112736,
+      "loss": 2.8671,
+      "step": 4224
+    },
+    {
+      "epoch": 0.3824567755951842,
+      "grad_norm": 0.9284319877624512,
+      "learning_rate": 0.00012588190451025207,
+      "loss": 2.9136,
+      "step": 4225
+    },
+    {
+      "epoch": 0.38254729790893455,
+      "grad_norm": 0.8524740934371948,
+      "learning_rate": 0.0001258512512330697,
+      "loss": 2.8853,
+      "step": 4226
+    },
+    {
+      "epoch": 0.3826378202226849,
+      "grad_norm": 0.935153603553772,
+      "learning_rate": 0.00012582059535266708,
+      "loss": 2.9154,
+      "step": 4227
+    },
+    {
+      "epoch": 0.38272834253643523,
+      "grad_norm": 0.9207292199134827,
+      "learning_rate": 0.00012578993687213118,
+      "loss": 2.8153,
+      "step": 4228
+    },
+    {
+      "epoch": 0.38281886485018557,
+      "grad_norm": 0.8508001565933228,
+      "learning_rate": 0.00012575927579454936,
+      "loss": 2.8845,
+      "step": 4229
+    },
+    {
+      "epoch": 0.3829093871639359,
+      "grad_norm": 0.9135047793388367,
+      "learning_rate": 0.00012572861212300918,
+      "loss": 2.9142,
+      "step": 4230
+    },
+    {
+      "epoch": 0.38299990947768625,
+      "grad_norm": 0.8793505430221558,
+      "learning_rate": 0.00012569794586059845,
+      "loss": 2.9614,
+      "step": 4231
+    },
+    {
+      "epoch": 0.3830904317914366,
+      "grad_norm": 0.8126338124275208,
+      "learning_rate": 0.00012566727701040527,
+      "loss": 2.859,
+      "step": 4232
+    },
+    {
+      "epoch": 0.38318095410518693,
+      "grad_norm": 0.8644139170646667,
+      "learning_rate": 0.000125636605575518,
+      "loss": 2.875,
+      "step": 4233
+    },
+    {
+      "epoch": 0.3832714764189373,
+      "grad_norm": 0.9225925207138062,
+      "learning_rate": 0.00012560593155902522,
+      "loss": 2.9098,
+      "step": 4234
+    },
+    {
+      "epoch": 0.3833619987326876,
+      "grad_norm": 0.8944104909896851,
+      "learning_rate": 0.00012557525496401589,
+      "loss": 2.8909,
+      "step": 4235
+    },
+    {
+      "epoch": 0.38345252104643796,
+      "grad_norm": 0.8773167729377747,
+      "learning_rate": 0.00012554457579357905,
+      "loss": 2.8494,
+      "step": 4236
+    },
+    {
+      "epoch": 0.3835430433601883,
+      "grad_norm": 0.9156031012535095,
+      "learning_rate": 0.00012551389405080415,
+      "loss": 2.8644,
+      "step": 4237
+    },
+    {
+      "epoch": 0.38363356567393864,
+      "grad_norm": 0.9683679342269897,
+      "learning_rate": 0.0001254832097387808,
+      "loss": 2.8617,
+      "step": 4238
+    },
+    {
+      "epoch": 0.383724087987689,
+      "grad_norm": 0.8678101301193237,
+      "learning_rate": 0.0001254525228605989,
+      "loss": 2.8842,
+      "step": 4239
+    },
+    {
+      "epoch": 0.3838146103014393,
+      "grad_norm": 0.8760818243026733,
+      "learning_rate": 0.00012542183341934872,
+      "loss": 2.8221,
+      "step": 4240
+    },
+    {
+      "epoch": 0.38390513261518966,
+      "grad_norm": 0.957080066204071,
+      "learning_rate": 0.00012539114141812057,
+      "loss": 2.8962,
+      "step": 4241
+    },
+    {
+      "epoch": 0.38399565492894,
+      "grad_norm": 0.8777123093605042,
+      "learning_rate": 0.00012536044686000514,
+      "loss": 2.8909,
+      "step": 4242
+    },
+    {
+      "epoch": 0.38408617724269034,
+      "grad_norm": 0.8769468665122986,
+      "learning_rate": 0.00012532974974809345,
+      "loss": 2.8519,
+      "step": 4243
+    },
+    {
+      "epoch": 0.3841766995564407,
+      "grad_norm": 0.9063210487365723,
+      "learning_rate": 0.0001252990500854766,
+      "loss": 2.8714,
+      "step": 4244
+    },
+    {
+      "epoch": 0.384267221870191,
+      "grad_norm": 0.9434245228767395,
+      "learning_rate": 0.00012526834787524615,
+      "loss": 2.8069,
+      "step": 4245
+    },
+    {
+      "epoch": 0.38435774418394136,
+      "grad_norm": 1.06622314453125,
+      "learning_rate": 0.00012523764312049376,
+      "loss": 2.8203,
+      "step": 4246
+    },
+    {
+      "epoch": 0.3844482664976917,
+      "grad_norm": 0.9174593687057495,
+      "learning_rate": 0.0001252069358243114,
+      "loss": 2.879,
+      "step": 4247
+    },
+    {
+      "epoch": 0.38453878881144204,
+      "grad_norm": 1.1018455028533936,
+      "learning_rate": 0.00012517622598979126,
+      "loss": 2.8449,
+      "step": 4248
+    },
+    {
+      "epoch": 0.3846293111251924,
+      "grad_norm": 0.8385401368141174,
+      "learning_rate": 0.00012514551362002586,
+      "loss": 2.7949,
+      "step": 4249
+    },
+    {
+      "epoch": 0.3847198334389427,
+      "grad_norm": 0.9275259971618652,
+      "learning_rate": 0.0001251147987181079,
+      "loss": 2.8635,
+      "step": 4250
+    },
+    {
+      "epoch": 0.384810355752693,
+      "grad_norm": 0.9716927409172058,
+      "learning_rate": 0.00012508408128713048,
+      "loss": 2.9095,
+      "step": 4251
+    },
+    {
+      "epoch": 0.38490087806644335,
+      "grad_norm": 0.8956915140151978,
+      "learning_rate": 0.00012505336133018672,
+      "loss": 2.8683,
+      "step": 4252
+    },
+    {
+      "epoch": 0.3849914003801937,
+      "grad_norm": 0.9020968675613403,
+      "learning_rate": 0.00012502263885037014,
+      "loss": 2.8835,
+      "step": 4253
+    },
+    {
+      "epoch": 0.38508192269394403,
+      "grad_norm": 1.0122654438018799,
+      "learning_rate": 0.0001249919138507745,
+      "loss": 2.9904,
+      "step": 4254
+    },
+    {
+      "epoch": 0.3851724450076944,
+      "grad_norm": 0.9834968447685242,
+      "learning_rate": 0.00012496118633449385,
+      "loss": 2.8318,
+      "step": 4255
+    },
+    {
+      "epoch": 0.3852629673214447,
+      "grad_norm": 0.9706962704658508,
+      "learning_rate": 0.0001249304563046224,
+      "loss": 2.9138,
+      "step": 4256
+    },
+    {
+      "epoch": 0.38535348963519506,
+      "grad_norm": 1.0231425762176514,
+      "learning_rate": 0.0001248997237642547,
+      "loss": 2.9679,
+      "step": 4257
+    },
+    {
+      "epoch": 0.3854440119489454,
+      "grad_norm": 0.9518309831619263,
+      "learning_rate": 0.0001248689887164855,
+      "loss": 2.891,
+      "step": 4258
+    },
+    {
+      "epoch": 0.38553453426269574,
+      "grad_norm": 0.9990224242210388,
+      "learning_rate": 0.00012483825116440983,
+      "loss": 2.8797,
+      "step": 4259
+    },
+    {
+      "epoch": 0.3856250565764461,
+      "grad_norm": 0.8865232467651367,
+      "learning_rate": 0.0001248075111111229,
+      "loss": 2.8957,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3857155788901964,
+      "grad_norm": 0.9017773270606995,
+      "learning_rate": 0.0001247767685597203,
+      "loss": 2.9159,
+      "step": 4261
+    },
+    {
+      "epoch": 0.38580610120394676,
+      "grad_norm": 0.8915837407112122,
+      "learning_rate": 0.0001247460235132978,
+      "loss": 2.9103,
+      "step": 4262
+    },
+    {
+      "epoch": 0.3858966235176971,
+      "grad_norm": 0.9616152048110962,
+      "learning_rate": 0.00012471527597495138,
+      "loss": 2.8551,
+      "step": 4263
+    },
+    {
+      "epoch": 0.38598714583144744,
+      "grad_norm": 0.9576265811920166,
+      "learning_rate": 0.00012468452594777737,
+      "loss": 2.9263,
+      "step": 4264
+    },
+    {
+      "epoch": 0.3860776681451978,
+      "grad_norm": 0.9476226568222046,
+      "learning_rate": 0.00012465377343487226,
+      "loss": 2.957,
+      "step": 4265
+    },
+    {
+      "epoch": 0.3861681904589481,
+      "grad_norm": 0.8279786705970764,
+      "learning_rate": 0.00012462301843933285,
+      "loss": 2.8588,
+      "step": 4266
+    },
+    {
+      "epoch": 0.38625871277269846,
+      "grad_norm": 0.8010810017585754,
+      "learning_rate": 0.00012459226096425615,
+      "loss": 2.8308,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3863492350864488,
+      "grad_norm": 0.7839717268943787,
+      "learning_rate": 0.00012456150101273944,
+      "loss": 2.8188,
+      "step": 4268
+    },
+    {
+      "epoch": 0.38643975740019915,
+      "grad_norm": 0.8168207406997681,
+      "learning_rate": 0.00012453073858788026,
+      "loss": 2.8888,
+      "step": 4269
+    },
+    {
+      "epoch": 0.3865302797139495,
+      "grad_norm": 0.8193057179450989,
+      "learning_rate": 0.0001244999736927764,
+      "loss": 2.8527,
+      "step": 4270
+    },
+    {
+      "epoch": 0.38662080202769983,
+      "grad_norm": 0.8369550704956055,
+      "learning_rate": 0.00012446920633052582,
+      "loss": 2.8992,
+      "step": 4271
+    },
+    {
+      "epoch": 0.38671132434145017,
+      "grad_norm": 0.8519963026046753,
+      "learning_rate": 0.00012443843650422685,
+      "loss": 2.8817,
+      "step": 4272
+    },
+    {
+      "epoch": 0.3868018466552005,
+      "grad_norm": 0.8299965262413025,
+      "learning_rate": 0.00012440766421697799,
+      "loss": 2.9291,
+      "step": 4273
+    },
+    {
+      "epoch": 0.38689236896895085,
+      "grad_norm": 0.826339066028595,
+      "learning_rate": 0.000124376889471878,
+      "loss": 2.8891,
+      "step": 4274
+    },
+    {
+      "epoch": 0.3869828912827012,
+      "grad_norm": 0.9041489958763123,
+      "learning_rate": 0.0001243461122720259,
+      "loss": 2.9446,
+      "step": 4275
+    },
+    {
+      "epoch": 0.38707341359645153,
+      "grad_norm": 0.8152336478233337,
+      "learning_rate": 0.00012431533262052098,
+      "loss": 2.9069,
+      "step": 4276
+    },
+    {
+      "epoch": 0.3871639359102019,
+      "grad_norm": 1.0291637182235718,
+      "learning_rate": 0.00012428455052046272,
+      "loss": 2.8704,
+      "step": 4277
+    },
+    {
+      "epoch": 0.3872544582239522,
+      "grad_norm": 0.8456107378005981,
+      "learning_rate": 0.0001242537659749509,
+      "loss": 2.9149,
+      "step": 4278
+    },
+    {
+      "epoch": 0.38734498053770255,
+      "grad_norm": 0.8081295490264893,
+      "learning_rate": 0.0001242229789870855,
+      "loss": 2.8514,
+      "step": 4279
+    },
+    {
+      "epoch": 0.3874355028514529,
+      "grad_norm": 0.9080233573913574,
+      "learning_rate": 0.00012419218955996676,
+      "loss": 2.8538,
+      "step": 4280
+    },
+    {
+      "epoch": 0.38752602516520324,
+      "grad_norm": 0.862369954586029,
+      "learning_rate": 0.00012416139769669523,
+      "loss": 2.8268,
+      "step": 4281
+    },
+    {
+      "epoch": 0.3876165474789536,
+      "grad_norm": 0.8651420474052429,
+      "learning_rate": 0.00012413060340037163,
+      "loss": 2.8342,
+      "step": 4282
+    },
+    {
+      "epoch": 0.3877070697927039,
+      "grad_norm": 0.8810906410217285,
+      "learning_rate": 0.00012409980667409687,
+      "loss": 2.8656,
+      "step": 4283
+    },
+    {
+      "epoch": 0.38779759210645426,
+      "grad_norm": 0.8607074022293091,
+      "learning_rate": 0.00012406900752097227,
+      "loss": 2.8539,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3878881144202046,
+      "grad_norm": 0.9719976186752319,
+      "learning_rate": 0.00012403820594409924,
+      "loss": 2.9161,
+      "step": 4285
+    },
+    {
+      "epoch": 0.38797863673395494,
+      "grad_norm": 0.8851466178894043,
+      "learning_rate": 0.00012400740194657957,
+      "loss": 2.8657,
+      "step": 4286
+    },
+    {
+      "epoch": 0.3880691590477053,
+      "grad_norm": 0.853900671005249,
+      "learning_rate": 0.00012397659553151514,
+      "loss": 2.8479,
+      "step": 4287
+    },
+    {
+      "epoch": 0.3881596813614556,
+      "grad_norm": 0.9867987036705017,
+      "learning_rate": 0.00012394578670200826,
+      "loss": 2.8907,
+      "step": 4288
+    },
+    {
+      "epoch": 0.38825020367520596,
+      "grad_norm": 0.9488958716392517,
+      "learning_rate": 0.00012391497546116126,
+      "loss": 2.8455,
+      "step": 4289
+    },
+    {
+      "epoch": 0.3883407259889563,
+      "grad_norm": 0.9296489953994751,
+      "learning_rate": 0.0001238841618120769,
+      "loss": 2.8706,
+      "step": 4290
+    },
+    {
+      "epoch": 0.38843124830270664,
+      "grad_norm": 0.8813881874084473,
+      "learning_rate": 0.0001238533457578581,
+      "loss": 2.9245,
+      "step": 4291
+    },
+    {
+      "epoch": 0.38852177061645693,
+      "grad_norm": 0.8357471823692322,
+      "learning_rate": 0.00012382252730160803,
+      "loss": 2.8462,
+      "step": 4292
+    },
+    {
+      "epoch": 0.38861229293020727,
+      "grad_norm": 0.8444204330444336,
+      "learning_rate": 0.00012379170644643013,
+      "loss": 2.8553,
+      "step": 4293
+    },
+    {
+      "epoch": 0.3887028152439576,
+      "grad_norm": 0.8784762024879456,
+      "learning_rate": 0.000123760883195428,
+      "loss": 2.8585,
+      "step": 4294
+    },
+    {
+      "epoch": 0.38879333755770795,
+      "grad_norm": 0.8281036019325256,
+      "learning_rate": 0.00012373005755170563,
+      "loss": 2.8572,
+      "step": 4295
+    },
+    {
+      "epoch": 0.3888838598714583,
+      "grad_norm": 0.8708269596099854,
+      "learning_rate": 0.0001236992295183671,
+      "loss": 2.8737,
+      "step": 4296
+    },
+    {
+      "epoch": 0.38897438218520863,
+      "grad_norm": 0.9682866930961609,
+      "learning_rate": 0.0001236683990985168,
+      "loss": 2.8974,
+      "step": 4297
+    },
+    {
+      "epoch": 0.389064904498959,
+      "grad_norm": 0.8646484017372131,
+      "learning_rate": 0.00012363756629525937,
+      "loss": 2.8173,
+      "step": 4298
+    },
+    {
+      "epoch": 0.3891554268127093,
+      "grad_norm": 0.8460427522659302,
+      "learning_rate": 0.00012360673111169965,
+      "loss": 2.8546,
+      "step": 4299
+    },
+    {
+      "epoch": 0.38924594912645966,
+      "grad_norm": 0.8345630168914795,
+      "learning_rate": 0.00012357589355094275,
+      "loss": 2.8514,
+      "step": 4300
+    },
+    {
+      "epoch": 0.38933647144021,
+      "grad_norm": 0.8849084377288818,
+      "learning_rate": 0.00012354505361609398,
+      "loss": 2.8739,
+      "step": 4301
+    },
+    {
+      "epoch": 0.38942699375396034,
+      "grad_norm": 0.825874388217926,
+      "learning_rate": 0.000123514211310259,
+      "loss": 2.8724,
+      "step": 4302
+    },
+    {
+      "epoch": 0.3895175160677107,
+      "grad_norm": 0.8098484873771667,
+      "learning_rate": 0.00012348336663654358,
+      "loss": 2.8817,
+      "step": 4303
+    },
+    {
+      "epoch": 0.389608038381461,
+      "grad_norm": 0.8967568874359131,
+      "learning_rate": 0.0001234525195980538,
+      "loss": 2.9141,
+      "step": 4304
+    },
+    {
+      "epoch": 0.38969856069521136,
+      "grad_norm": 0.8785738348960876,
+      "learning_rate": 0.0001234216701978959,
+      "loss": 2.8451,
+      "step": 4305
+    },
+    {
+      "epoch": 0.3897890830089617,
+      "grad_norm": 0.9527652263641357,
+      "learning_rate": 0.00012339081843917645,
+      "loss": 2.796,
+      "step": 4306
+    },
+    {
+      "epoch": 0.38987960532271204,
+      "grad_norm": 0.8243290781974792,
+      "learning_rate": 0.00012335996432500228,
+      "loss": 2.7992,
+      "step": 4307
+    },
+    {
+      "epoch": 0.3899701276364624,
+      "grad_norm": 0.8251200318336487,
+      "learning_rate": 0.0001233291078584803,
+      "loss": 2.839,
+      "step": 4308
+    },
+    {
+      "epoch": 0.3900606499502127,
+      "grad_norm": 0.8345006108283997,
+      "learning_rate": 0.00012329824904271777,
+      "loss": 2.854,
+      "step": 4309
+    },
+    {
+      "epoch": 0.39015117226396306,
+      "grad_norm": 0.8241994976997375,
+      "learning_rate": 0.00012326738788082223,
+      "loss": 2.8613,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3902416945777134,
+      "grad_norm": 0.7824759483337402,
+      "learning_rate": 0.00012323652437590138,
+      "loss": 2.8107,
+      "step": 4311
+    },
+    {
+      "epoch": 0.39033221689146375,
+      "grad_norm": 0.8252853751182556,
+      "learning_rate": 0.00012320565853106316,
+      "loss": 2.8466,
+      "step": 4312
+    },
+    {
+      "epoch": 0.3904227392052141,
+      "grad_norm": 0.8508025407791138,
+      "learning_rate": 0.00012317479034941573,
+      "loss": 2.8644,
+      "step": 4313
+    },
+    {
+      "epoch": 0.3905132615189644,
+      "grad_norm": 0.8455524444580078,
+      "learning_rate": 0.00012314391983406757,
+      "loss": 2.8706,
+      "step": 4314
+    },
+    {
+      "epoch": 0.39060378383271477,
+      "grad_norm": 0.8711053729057312,
+      "learning_rate": 0.0001231130469881273,
+      "loss": 2.8698,
+      "step": 4315
+    },
+    {
+      "epoch": 0.3906943061464651,
+      "grad_norm": 0.9837364554405212,
+      "learning_rate": 0.00012308217181470385,
+      "loss": 2.9033,
+      "step": 4316
+    },
+    {
+      "epoch": 0.39078482846021545,
+      "grad_norm": 0.8791324496269226,
+      "learning_rate": 0.00012305129431690632,
+      "loss": 2.8562,
+      "step": 4317
+    },
+    {
+      "epoch": 0.3908753507739658,
+      "grad_norm": 0.8958924412727356,
+      "learning_rate": 0.00012302041449784409,
+      "loss": 2.8738,
+      "step": 4318
+    },
+    {
+      "epoch": 0.39096587308771613,
+      "grad_norm": 0.8958149552345276,
+      "learning_rate": 0.00012298953236062677,
+      "loss": 2.9199,
+      "step": 4319
+    },
+    {
+      "epoch": 0.3910563954014665,
+      "grad_norm": 0.7985171675682068,
+      "learning_rate": 0.0001229586479083641,
+      "loss": 2.8096,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3911469177152168,
+      "grad_norm": 0.8733648061752319,
+      "learning_rate": 0.00012292776114416627,
+      "loss": 2.8864,
+      "step": 4321
+    },
+    {
+      "epoch": 0.39123744002896715,
+      "grad_norm": 0.8304138779640198,
+      "learning_rate": 0.0001228968720711435,
+      "loss": 2.8483,
+      "step": 4322
+    },
+    {
+      "epoch": 0.3913279623427175,
+      "grad_norm": 0.9201784729957581,
+      "learning_rate": 0.00012286598069240635,
+      "loss": 2.9073,
+      "step": 4323
+    },
+    {
+      "epoch": 0.39141848465646784,
+      "grad_norm": 0.9191954135894775,
+      "learning_rate": 0.00012283508701106557,
+      "loss": 2.8897,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3915090069702182,
+      "grad_norm": 0.8933663368225098,
+      "learning_rate": 0.00012280419103023217,
+      "loss": 2.8218,
+      "step": 4325
+    },
+    {
+      "epoch": 0.3915995292839685,
+      "grad_norm": 0.8663362860679626,
+      "learning_rate": 0.00012277329275301732,
+      "loss": 2.8768,
+      "step": 4326
+    },
+    {
+      "epoch": 0.39169005159771886,
+      "grad_norm": 0.8382437229156494,
+      "learning_rate": 0.00012274239218253255,
+      "loss": 2.8738,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3917805739114692,
+      "grad_norm": 0.8278548121452332,
+      "learning_rate": 0.0001227114893218895,
+      "loss": 2.8751,
+      "step": 4328
+    },
+    {
+      "epoch": 0.39187109622521954,
+      "grad_norm": 0.8287359476089478,
+      "learning_rate": 0.00012268058417420007,
+      "loss": 2.8383,
+      "step": 4329
+    },
+    {
+      "epoch": 0.3919616185389699,
+      "grad_norm": 0.8964946866035461,
+      "learning_rate": 0.00012264967674257646,
+      "loss": 2.8526,
+      "step": 4330
+    },
+    {
+      "epoch": 0.3920521408527202,
+      "grad_norm": 0.829495370388031,
+      "learning_rate": 0.00012261876703013102,
+      "loss": 2.8759,
+      "step": 4331
+    },
+    {
+      "epoch": 0.39214266316647056,
+      "grad_norm": 0.8812410235404968,
+      "learning_rate": 0.00012258785503997633,
+      "loss": 2.8668,
+      "step": 4332
+    },
+    {
+      "epoch": 0.39223318548022085,
+      "grad_norm": 0.8432778120040894,
+      "learning_rate": 0.00012255694077522528,
+      "loss": 2.857,
+      "step": 4333
+    },
+    {
+      "epoch": 0.3923237077939712,
+      "grad_norm": 0.8417193293571472,
+      "learning_rate": 0.0001225260242389909,
+      "loss": 2.8395,
+      "step": 4334
+    },
+    {
+      "epoch": 0.39241423010772153,
+      "grad_norm": 0.8582254648208618,
+      "learning_rate": 0.0001224951054343865,
+      "loss": 2.9089,
+      "step": 4335
+    },
+    {
+      "epoch": 0.39250475242147187,
+      "grad_norm": 0.8694718480110168,
+      "learning_rate": 0.00012246418436452562,
+      "loss": 2.8468,
+      "step": 4336
+    },
+    {
+      "epoch": 0.3925952747352222,
+      "grad_norm": 0.8391682505607605,
+      "learning_rate": 0.00012243326103252198,
+      "loss": 2.9174,
+      "step": 4337
+    },
+    {
+      "epoch": 0.39268579704897255,
+      "grad_norm": 0.8766013979911804,
+      "learning_rate": 0.00012240233544148955,
+      "loss": 2.7995,
+      "step": 4338
+    },
+    {
+      "epoch": 0.3927763193627229,
+      "grad_norm": 0.8532915711402893,
+      "learning_rate": 0.00012237140759454256,
+      "loss": 2.8328,
+      "step": 4339
+    },
+    {
+      "epoch": 0.39286684167647323,
+      "grad_norm": 0.8749273419380188,
+      "learning_rate": 0.00012234047749479544,
+      "loss": 2.8936,
+      "step": 4340
+    },
+    {
+      "epoch": 0.3929573639902236,
+      "grad_norm": 0.8936702013015747,
+      "learning_rate": 0.00012230954514536285,
+      "loss": 2.7884,
+      "step": 4341
+    },
+    {
+      "epoch": 0.3930478863039739,
+      "grad_norm": 0.7897341847419739,
+      "learning_rate": 0.00012227861054935968,
+      "loss": 2.8163,
+      "step": 4342
+    },
+    {
+      "epoch": 0.39313840861772426,
+      "grad_norm": 0.8324590921401978,
+      "learning_rate": 0.00012224767370990101,
+      "loss": 2.9101,
+      "step": 4343
+    },
+    {
+      "epoch": 0.3932289309314746,
+      "grad_norm": 0.787476122379303,
+      "learning_rate": 0.00012221673463010224,
+      "loss": 2.8887,
+      "step": 4344
+    },
+    {
+      "epoch": 0.39331945324522494,
+      "grad_norm": 0.8571562767028809,
+      "learning_rate": 0.00012218579331307888,
+      "loss": 2.8645,
+      "step": 4345
+    },
+    {
+      "epoch": 0.3934099755589753,
+      "grad_norm": 0.87290358543396,
+      "learning_rate": 0.00012215484976194676,
+      "loss": 2.8824,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3935004978727256,
+      "grad_norm": 0.8542999029159546,
+      "learning_rate": 0.00012212390397982185,
+      "loss": 2.8801,
+      "step": 4347
+    },
+    {
+      "epoch": 0.39359102018647596,
+      "grad_norm": 0.9487765431404114,
+      "learning_rate": 0.00012209295596982042,
+      "loss": 2.8601,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3936815425002263,
+      "grad_norm": 0.85524982213974,
+      "learning_rate": 0.00012206200573505895,
+      "loss": 2.8454,
+      "step": 4349
+    },
+    {
+      "epoch": 0.39377206481397664,
+      "grad_norm": 0.8825935125350952,
+      "learning_rate": 0.00012203105327865407,
+      "loss": 2.9284,
+      "step": 4350
+    },
+    {
+      "epoch": 0.393862587127727,
+      "grad_norm": 0.8699039816856384,
+      "learning_rate": 0.00012200009860372275,
+      "loss": 2.8432,
+      "step": 4351
+    },
+    {
+      "epoch": 0.3939531094414773,
+      "grad_norm": 0.8540199398994446,
+      "learning_rate": 0.0001219691417133821,
+      "loss": 2.9114,
+      "step": 4352
+    },
+    {
+      "epoch": 0.39404363175522766,
+      "grad_norm": 0.9271214604377747,
+      "learning_rate": 0.00012193818261074949,
+      "loss": 2.8886,
+      "step": 4353
+    },
+    {
+      "epoch": 0.394134154068978,
+      "grad_norm": 0.8362212777137756,
+      "learning_rate": 0.00012190722129894248,
+      "loss": 2.8184,
+      "step": 4354
+    },
+    {
+      "epoch": 0.39422467638272835,
+      "grad_norm": 0.8742707371711731,
+      "learning_rate": 0.0001218762577810789,
+      "loss": 2.9139,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3943151986964787,
+      "grad_norm": 0.8451969027519226,
+      "learning_rate": 0.00012184529206027673,
+      "loss": 2.8399,
+      "step": 4356
+    },
+    {
+      "epoch": 0.394405721010229,
+      "grad_norm": 0.8662227988243103,
+      "learning_rate": 0.00012181432413965428,
+      "loss": 2.8809,
+      "step": 4357
+    },
+    {
+      "epoch": 0.39449624332397937,
+      "grad_norm": 0.8754814267158508,
+      "learning_rate": 0.00012178335402232996,
+      "loss": 2.8007,
+      "step": 4358
+    },
+    {
+      "epoch": 0.3945867656377297,
+      "grad_norm": 0.7949616312980652,
+      "learning_rate": 0.00012175238171142249,
+      "loss": 2.8371,
+      "step": 4359
+    },
+    {
+      "epoch": 0.39467728795148005,
+      "grad_norm": 0.8271093368530273,
+      "learning_rate": 0.00012172140721005079,
+      "loss": 2.7895,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3947678102652304,
+      "grad_norm": 0.8187628388404846,
+      "learning_rate": 0.00012169043052133398,
+      "loss": 2.7996,
+      "step": 4361
+    },
+    {
+      "epoch": 0.39485833257898073,
+      "grad_norm": 0.9346163868904114,
+      "learning_rate": 0.00012165945164839138,
+      "loss": 2.8735,
+      "step": 4362
+    },
+    {
+      "epoch": 0.39494885489273107,
+      "grad_norm": 0.9251669645309448,
+      "learning_rate": 0.00012162847059434261,
+      "loss": 2.8397,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3950393772064814,
+      "grad_norm": 0.818914532661438,
+      "learning_rate": 0.00012159748736230747,
+      "loss": 2.8434,
+      "step": 4364
+    },
+    {
+      "epoch": 0.39512989952023175,
+      "grad_norm": 0.8839664459228516,
+      "learning_rate": 0.0001215665019554059,
+      "loss": 2.8257,
+      "step": 4365
+    },
+    {
+      "epoch": 0.3952204218339821,
+      "grad_norm": 0.8695526719093323,
+      "learning_rate": 0.00012153551437675821,
+      "loss": 2.8094,
+      "step": 4366
+    },
+    {
+      "epoch": 0.39531094414773243,
+      "grad_norm": 0.8801332712173462,
+      "learning_rate": 0.00012150452462948479,
+      "loss": 2.832,
+      "step": 4367
+    },
+    {
+      "epoch": 0.3954014664614828,
+      "grad_norm": 0.8525098562240601,
+      "learning_rate": 0.00012147353271670634,
+      "loss": 2.8797,
+      "step": 4368
+    },
+    {
+      "epoch": 0.3954919887752331,
+      "grad_norm": 1.0056629180908203,
+      "learning_rate": 0.00012144253864154372,
+      "loss": 2.9493,
+      "step": 4369
+    },
+    {
+      "epoch": 0.39558251108898346,
+      "grad_norm": 0.8409388065338135,
+      "learning_rate": 0.00012141154240711805,
+      "loss": 2.8124,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3956730334027338,
+      "grad_norm": 0.8684418201446533,
+      "learning_rate": 0.00012138054401655063,
+      "loss": 2.7768,
+      "step": 4371
+    },
+    {
+      "epoch": 0.39576355571648414,
+      "grad_norm": 0.8791725039482117,
+      "learning_rate": 0.00012134954347296305,
+      "loss": 2.8593,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3958540780302345,
+      "grad_norm": 0.9349090456962585,
+      "learning_rate": 0.00012131854077947699,
+      "loss": 2.8747,
+      "step": 4373
+    },
+    {
+      "epoch": 0.39594460034398477,
+      "grad_norm": 0.8536260724067688,
+      "learning_rate": 0.00012128753593921447,
+      "loss": 2.9138,
+      "step": 4374
+    },
+    {
+      "epoch": 0.3960351226577351,
+      "grad_norm": 0.8024949431419373,
+      "learning_rate": 0.00012125652895529766,
+      "loss": 2.8145,
+      "step": 4375
+    },
+    {
+      "epoch": 0.39612564497148545,
+      "grad_norm": 0.876518726348877,
+      "learning_rate": 0.000121225519830849,
+      "loss": 2.8944,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3962161672852358,
+      "grad_norm": 0.9107688069343567,
+      "learning_rate": 0.00012119450856899104,
+      "loss": 2.8592,
+      "step": 4377
+    },
+    {
+      "epoch": 0.39630668959898613,
+      "grad_norm": 0.8297701478004456,
+      "learning_rate": 0.00012116349517284665,
+      "loss": 2.8506,
+      "step": 4378
+    },
+    {
+      "epoch": 0.39639721191273647,
+      "grad_norm": 0.9151343703269958,
+      "learning_rate": 0.00012113247964553888,
+      "loss": 2.8746,
+      "step": 4379
+    },
+    {
+      "epoch": 0.3964877342264868,
+      "grad_norm": 0.9112572073936462,
+      "learning_rate": 0.000121101461990191,
+      "loss": 2.8462,
+      "step": 4380
+    },
+    {
+      "epoch": 0.39657825654023715,
+      "grad_norm": 0.879729151725769,
+      "learning_rate": 0.00012107044220992642,
+      "loss": 2.9074,
+      "step": 4381
+    },
+    {
+      "epoch": 0.3966687788539875,
+      "grad_norm": 0.9959706664085388,
+      "learning_rate": 0.00012103942030786894,
+      "loss": 2.8299,
+      "step": 4382
+    },
+    {
+      "epoch": 0.39675930116773783,
+      "grad_norm": 0.869488000869751,
+      "learning_rate": 0.0001210083962871424,
+      "loss": 2.8547,
+      "step": 4383
+    },
+    {
+      "epoch": 0.3968498234814882,
+      "grad_norm": 0.826572597026825,
+      "learning_rate": 0.00012097737015087094,
+      "loss": 2.8806,
+      "step": 4384
+    },
+    {
+      "epoch": 0.3969403457952385,
+      "grad_norm": 0.8823446035385132,
+      "learning_rate": 0.00012094634190217887,
+      "loss": 2.8588,
+      "step": 4385
+    },
+    {
+      "epoch": 0.39703086810898885,
+      "grad_norm": 0.9061761498451233,
+      "learning_rate": 0.00012091531154419072,
+      "loss": 2.8667,
+      "step": 4386
+    },
+    {
+      "epoch": 0.3971213904227392,
+      "grad_norm": 0.8355569243431091,
+      "learning_rate": 0.0001208842790800313,
+      "loss": 2.8116,
+      "step": 4387
+    },
+    {
+      "epoch": 0.39721191273648954,
+      "grad_norm": 0.8613497018814087,
+      "learning_rate": 0.00012085324451282553,
+      "loss": 2.8725,
+      "step": 4388
+    },
+    {
+      "epoch": 0.3973024350502399,
+      "grad_norm": 0.9011744856834412,
+      "learning_rate": 0.00012082220784569862,
+      "loss": 2.9125,
+      "step": 4389
+    },
+    {
+      "epoch": 0.3973929573639902,
+      "grad_norm": 0.8703922629356384,
+      "learning_rate": 0.00012079116908177593,
+      "loss": 2.845,
+      "step": 4390
+    },
+    {
+      "epoch": 0.39748347967774056,
+      "grad_norm": 0.8401715159416199,
+      "learning_rate": 0.00012076012822418312,
+      "loss": 2.8521,
+      "step": 4391
+    },
+    {
+      "epoch": 0.3975740019914909,
+      "grad_norm": 0.8488914370536804,
+      "learning_rate": 0.00012072908527604594,
+      "loss": 2.8931,
+      "step": 4392
+    },
+    {
+      "epoch": 0.39766452430524124,
+      "grad_norm": 0.9117351770401001,
+      "learning_rate": 0.00012069804024049043,
+      "loss": 2.7976,
+      "step": 4393
+    },
+    {
+      "epoch": 0.3977550466189916,
+      "grad_norm": 0.94156813621521,
+      "learning_rate": 0.00012066699312064286,
+      "loss": 2.9198,
+      "step": 4394
+    },
+    {
+      "epoch": 0.3978455689327419,
+      "grad_norm": 0.8953226804733276,
+      "learning_rate": 0.00012063594391962962,
+      "loss": 2.8474,
+      "step": 4395
+    },
+    {
+      "epoch": 0.39793609124649226,
+      "grad_norm": 0.8929262161254883,
+      "learning_rate": 0.00012060489264057742,
+      "loss": 2.8395,
+      "step": 4396
+    },
+    {
+      "epoch": 0.3980266135602426,
+      "grad_norm": 0.8781452178955078,
+      "learning_rate": 0.00012057383928661308,
+      "loss": 2.853,
+      "step": 4397
+    },
+    {
+      "epoch": 0.39811713587399294,
+      "grad_norm": 0.9336954951286316,
+      "learning_rate": 0.00012054278386086368,
+      "loss": 2.8811,
+      "step": 4398
+    },
+    {
+      "epoch": 0.3982076581877433,
+      "grad_norm": 0.8699139952659607,
+      "learning_rate": 0.00012051172636645654,
+      "loss": 2.9717,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3982981805014936,
+      "grad_norm": 0.8598851561546326,
+      "learning_rate": 0.00012048066680651908,
+      "loss": 2.8222,
+      "step": 4400
+    },
+    {
+      "epoch": 0.39838870281524397,
+      "grad_norm": 0.8180475831031799,
+      "learning_rate": 0.00012044960518417903,
+      "loss": 2.8072,
+      "step": 4401
+    },
+    {
+      "epoch": 0.3984792251289943,
+      "grad_norm": 0.9748538732528687,
+      "learning_rate": 0.00012041854150256433,
+      "loss": 2.8898,
+      "step": 4402
+    },
+    {
+      "epoch": 0.39856974744274465,
+      "grad_norm": 0.8849080204963684,
+      "learning_rate": 0.00012038747576480303,
+      "loss": 2.8782,
+      "step": 4403
+    },
+    {
+      "epoch": 0.398660269756495,
+      "grad_norm": 1.0537378787994385,
+      "learning_rate": 0.0001203564079740235,
+      "loss": 2.8096,
+      "step": 4404
+    },
+    {
+      "epoch": 0.39875079207024533,
+      "grad_norm": 0.8559945225715637,
+      "learning_rate": 0.00012032533813335423,
+      "loss": 2.8918,
+      "step": 4405
+    },
+    {
+      "epoch": 0.39884131438399567,
+      "grad_norm": 0.9511855244636536,
+      "learning_rate": 0.00012029426624592397,
+      "loss": 2.9183,
+      "step": 4406
+    },
+    {
+      "epoch": 0.398931836697746,
+      "grad_norm": 0.9862476587295532,
+      "learning_rate": 0.00012026319231486165,
+      "loss": 2.8062,
+      "step": 4407
+    },
+    {
+      "epoch": 0.39902235901149635,
+      "grad_norm": 0.9083656668663025,
+      "learning_rate": 0.00012023211634329643,
+      "loss": 2.8313,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3991128813252467,
+      "grad_norm": 0.9601117372512817,
+      "learning_rate": 0.00012020103833435766,
+      "loss": 2.8391,
+      "step": 4409
+    },
+    {
+      "epoch": 0.39920340363899703,
+      "grad_norm": 0.9977763295173645,
+      "learning_rate": 0.00012016995829117488,
+      "loss": 2.8204,
+      "step": 4410
+    },
+    {
+      "epoch": 0.3992939259527474,
+      "grad_norm": 0.8471393585205078,
+      "learning_rate": 0.00012013887621687783,
+      "loss": 2.905,
+      "step": 4411
+    },
+    {
+      "epoch": 0.3993844482664977,
+      "grad_norm": 0.9744412302970886,
+      "learning_rate": 0.00012010779211459648,
+      "loss": 2.8808,
+      "step": 4412
+    },
+    {
+      "epoch": 0.39947497058024806,
+      "grad_norm": 1.0381581783294678,
+      "learning_rate": 0.00012007670598746103,
+      "loss": 2.8895,
+      "step": 4413
+    },
+    {
+      "epoch": 0.3995654928939984,
+      "grad_norm": 0.8495534658432007,
+      "learning_rate": 0.00012004561783860186,
+      "loss": 2.8401,
+      "step": 4414
+    },
+    {
+      "epoch": 0.3996560152077487,
+      "grad_norm": 1.0044671297073364,
+      "learning_rate": 0.00012001452767114951,
+      "loss": 2.918,
+      "step": 4415
+    },
+    {
+      "epoch": 0.399746537521499,
+      "grad_norm": 0.9183343648910522,
+      "learning_rate": 0.00011998343548823474,
+      "loss": 2.8259,
+      "step": 4416
+    },
+    {
+      "epoch": 0.39983705983524936,
+      "grad_norm": 0.8820471167564392,
+      "learning_rate": 0.0001199523412929886,
+      "loss": 2.8592,
+      "step": 4417
+    },
+    {
+      "epoch": 0.3999275821489997,
+      "grad_norm": 0.9422193169593811,
+      "learning_rate": 0.00011992124508854223,
+      "loss": 2.8703,
+      "step": 4418
+    },
+    {
+      "epoch": 0.40001810446275005,
+      "grad_norm": 0.8565012216567993,
+      "learning_rate": 0.00011989014687802703,
+      "loss": 2.8063,
+      "step": 4419
+    },
+    {
+      "epoch": 0.4001086267765004,
+      "grad_norm": 0.8324585556983948,
+      "learning_rate": 0.00011985904666457455,
+      "loss": 2.8611,
+      "step": 4420
+    },
+    {
+      "epoch": 0.4001991490902507,
+      "grad_norm": 0.8732492923736572,
+      "learning_rate": 0.00011982794445131665,
+      "loss": 2.8615,
+      "step": 4421
+    },
+    {
+      "epoch": 0.40028967140400107,
+      "grad_norm": 0.7857765555381775,
+      "learning_rate": 0.00011979684024138529,
+      "loss": 2.8034,
+      "step": 4422
+    },
+    {
+      "epoch": 0.4003801937177514,
+      "grad_norm": 0.8531179428100586,
+      "learning_rate": 0.00011976573403791262,
+      "loss": 2.9058,
+      "step": 4423
+    },
+    {
+      "epoch": 0.40047071603150175,
+      "grad_norm": 0.9699780941009521,
+      "learning_rate": 0.0001197346258440311,
+      "loss": 2.8425,
+      "step": 4424
+    },
+    {
+      "epoch": 0.4005612383452521,
+      "grad_norm": 0.8353390097618103,
+      "learning_rate": 0.00011970351566287333,
+      "loss": 2.8082,
+      "step": 4425
+    },
+    {
+      "epoch": 0.40065176065900243,
+      "grad_norm": 0.9401875138282776,
+      "learning_rate": 0.00011967240349757203,
+      "loss": 2.9017,
+      "step": 4426
+    },
+    {
+      "epoch": 0.4007422829727528,
+      "grad_norm": 0.9973408579826355,
+      "learning_rate": 0.00011964128935126023,
+      "loss": 2.8623,
+      "step": 4427
+    },
+    {
+      "epoch": 0.4008328052865031,
+      "grad_norm": 0.8002395629882812,
+      "learning_rate": 0.00011961017322707113,
+      "loss": 2.8808,
+      "step": 4428
+    },
+    {
+      "epoch": 0.40092332760025345,
+      "grad_norm": 0.9130750298500061,
+      "learning_rate": 0.00011957905512813817,
+      "loss": 2.8638,
+      "step": 4429
+    },
+    {
+      "epoch": 0.4010138499140038,
+      "grad_norm": 0.9972096085548401,
+      "learning_rate": 0.00011954793505759483,
+      "loss": 2.9469,
+      "step": 4430
+    },
+    {
+      "epoch": 0.40110437222775414,
+      "grad_norm": 0.8289200663566589,
+      "learning_rate": 0.000119516813018575,
+      "loss": 2.7791,
+      "step": 4431
+    },
+    {
+      "epoch": 0.4011948945415045,
+      "grad_norm": 1.0694494247436523,
+      "learning_rate": 0.00011948568901421263,
+      "loss": 2.8447,
+      "step": 4432
+    },
+    {
+      "epoch": 0.4012854168552548,
+      "grad_norm": 1.0092542171478271,
+      "learning_rate": 0.00011945456304764192,
+      "loss": 2.8501,
+      "step": 4433
+    },
+    {
+      "epoch": 0.40137593916900516,
+      "grad_norm": 0.955417275428772,
+      "learning_rate": 0.0001194234351219972,
+      "loss": 2.8781,
+      "step": 4434
+    },
+    {
+      "epoch": 0.4014664614827555,
+      "grad_norm": 1.1501630544662476,
+      "learning_rate": 0.00011939230524041314,
+      "loss": 2.841,
+      "step": 4435
+    },
+    {
+      "epoch": 0.40155698379650584,
+      "grad_norm": 0.8584458827972412,
+      "learning_rate": 0.00011936117340602444,
+      "loss": 2.8188,
+      "step": 4436
+    },
+    {
+      "epoch": 0.4016475061102562,
+      "grad_norm": 0.9481405019760132,
+      "learning_rate": 0.00011933003962196613,
+      "loss": 2.8674,
+      "step": 4437
+    },
+    {
+      "epoch": 0.4017380284240065,
+      "grad_norm": 0.9717084169387817,
+      "learning_rate": 0.00011929890389137337,
+      "loss": 2.857,
+      "step": 4438
+    },
+    {
+      "epoch": 0.40182855073775686,
+      "grad_norm": 0.8797189593315125,
+      "learning_rate": 0.00011926776621738149,
+      "loss": 2.7974,
+      "step": 4439
+    },
+    {
+      "epoch": 0.4019190730515072,
+      "grad_norm": 0.9222303628921509,
+      "learning_rate": 0.00011923662660312611,
+      "loss": 2.8468,
+      "step": 4440
+    },
+    {
+      "epoch": 0.40200959536525754,
+      "grad_norm": 0.8810079097747803,
+      "learning_rate": 0.00011920548505174293,
+      "loss": 2.8549,
+      "step": 4441
+    },
+    {
+      "epoch": 0.4021001176790079,
+      "grad_norm": 0.9079646468162537,
+      "learning_rate": 0.00011917434156636793,
+      "loss": 2.7933,
+      "step": 4442
+    },
+    {
+      "epoch": 0.4021906399927582,
+      "grad_norm": 0.8858142495155334,
+      "learning_rate": 0.00011914319615013726,
+      "loss": 2.7967,
+      "step": 4443
+    },
+    {
+      "epoch": 0.40228116230650857,
+      "grad_norm": 0.8796921372413635,
+      "learning_rate": 0.00011911204880618729,
+      "loss": 2.8341,
+      "step": 4444
+    },
+    {
+      "epoch": 0.4023716846202589,
+      "grad_norm": 0.9303306937217712,
+      "learning_rate": 0.00011908089953765449,
+      "loss": 2.8224,
+      "step": 4445
+    },
+    {
+      "epoch": 0.40246220693400925,
+      "grad_norm": 0.9221446514129639,
+      "learning_rate": 0.00011904974834767563,
+      "loss": 2.8236,
+      "step": 4446
+    },
+    {
+      "epoch": 0.4025527292477596,
+      "grad_norm": 0.9513319730758667,
+      "learning_rate": 0.00011901859523938765,
+      "loss": 2.8596,
+      "step": 4447
+    },
+    {
+      "epoch": 0.40264325156150993,
+      "grad_norm": 0.7914127707481384,
+      "learning_rate": 0.00011898744021592767,
+      "loss": 2.8026,
+      "step": 4448
+    },
+    {
+      "epoch": 0.40273377387526027,
+      "grad_norm": 0.8371176719665527,
+      "learning_rate": 0.00011895628328043296,
+      "loss": 2.8296,
+      "step": 4449
+    },
+    {
+      "epoch": 0.4028242961890106,
+      "grad_norm": 0.9562647938728333,
+      "learning_rate": 0.00011892512443604102,
+      "loss": 2.8599,
+      "step": 4450
+    },
+    {
+      "epoch": 0.40291481850276095,
+      "grad_norm": 0.9588335156440735,
+      "learning_rate": 0.00011889396368588958,
+      "loss": 2.8484,
+      "step": 4451
+    },
+    {
+      "epoch": 0.4030053408165113,
+      "grad_norm": 0.9410818219184875,
+      "learning_rate": 0.00011886280103311656,
+      "loss": 2.8455,
+      "step": 4452
+    },
+    {
+      "epoch": 0.40309586313026163,
+      "grad_norm": 1.0550609827041626,
+      "learning_rate": 0.00011883163648085993,
+      "loss": 2.8471,
+      "step": 4453
+    },
+    {
+      "epoch": 0.403186385444012,
+      "grad_norm": 0.8768680095672607,
+      "learning_rate": 0.00011880047003225804,
+      "loss": 2.8065,
+      "step": 4454
+    },
+    {
+      "epoch": 0.4032769077577623,
+      "grad_norm": 0.815466582775116,
+      "learning_rate": 0.00011876930169044936,
+      "loss": 2.8442,
+      "step": 4455
+    },
+    {
+      "epoch": 0.4033674300715126,
+      "grad_norm": 0.9968241453170776,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 2.9243,
+      "step": 4456
+    },
+    {
+      "epoch": 0.40345795238526294,
+      "grad_norm": 0.8512164354324341,
+      "learning_rate": 0.00011870695933976628,
+      "loss": 2.8586,
+      "step": 4457
+    },
+    {
+      "epoch": 0.4035484746990133,
+      "grad_norm": 0.846886396408081,
+      "learning_rate": 0.00011867578533716979,
+      "loss": 2.8505,
+      "step": 4458
+    },
+    {
+      "epoch": 0.4036389970127636,
+      "grad_norm": 0.8526726365089417,
+      "learning_rate": 0.00011864460945392224,
+      "loss": 2.8297,
+      "step": 4459
+    },
+    {
+      "epoch": 0.40372951932651396,
+      "grad_norm": 0.8981117606163025,
+      "learning_rate": 0.00011861343169316301,
+      "loss": 2.8628,
+      "step": 4460
+    },
+    {
+      "epoch": 0.4038200416402643,
+      "grad_norm": 0.8755601644515991,
+      "learning_rate": 0.00011858225205803171,
+      "loss": 2.8063,
+      "step": 4461
+    },
+    {
+      "epoch": 0.40391056395401465,
+      "grad_norm": 0.9109624624252319,
+      "learning_rate": 0.00011855107055166814,
+      "loss": 2.8223,
+      "step": 4462
+    },
+    {
+      "epoch": 0.404001086267765,
+      "grad_norm": 0.8895991444587708,
+      "learning_rate": 0.00011851988717721232,
+      "loss": 2.8655,
+      "step": 4463
+    },
+    {
+      "epoch": 0.4040916085815153,
+      "grad_norm": 0.8829487562179565,
+      "learning_rate": 0.00011848870193780433,
+      "loss": 2.8666,
+      "step": 4464
+    },
+    {
+      "epoch": 0.40418213089526567,
+      "grad_norm": 0.9633542895317078,
+      "learning_rate": 0.00011845751483658453,
+      "loss": 2.8261,
+      "step": 4465
+    },
+    {
+      "epoch": 0.404272653209016,
+      "grad_norm": 0.8148611187934875,
+      "learning_rate": 0.00011842632587669353,
+      "loss": 2.8781,
+      "step": 4466
+    },
+    {
+      "epoch": 0.40436317552276635,
+      "grad_norm": 0.8730365633964539,
+      "learning_rate": 0.00011839513506127203,
+      "loss": 2.8309,
+      "step": 4467
+    },
+    {
+      "epoch": 0.4044536978365167,
+      "grad_norm": 0.90838623046875,
+      "learning_rate": 0.00011836394239346091,
+      "loss": 2.8738,
+      "step": 4468
+    },
+    {
+      "epoch": 0.40454422015026703,
+      "grad_norm": 0.8492640256881714,
+      "learning_rate": 0.00011833274787640129,
+      "loss": 2.8809,
+      "step": 4469
+    },
+    {
+      "epoch": 0.40463474246401737,
+      "grad_norm": 0.8511492013931274,
+      "learning_rate": 0.00011830155151323446,
+      "loss": 2.8329,
+      "step": 4470
+    },
+    {
+      "epoch": 0.4047252647777677,
+      "grad_norm": 0.8511828780174255,
+      "learning_rate": 0.00011827035330710192,
+      "loss": 2.9125,
+      "step": 4471
+    },
+    {
+      "epoch": 0.40481578709151805,
+      "grad_norm": 0.9062504172325134,
+      "learning_rate": 0.00011823915326114527,
+      "loss": 2.8957,
+      "step": 4472
+    },
+    {
+      "epoch": 0.4049063094052684,
+      "grad_norm": 0.927947461605072,
+      "learning_rate": 0.0001182079513785064,
+      "loss": 2.8272,
+      "step": 4473
+    },
+    {
+      "epoch": 0.40499683171901874,
+      "grad_norm": 0.8176203966140747,
+      "learning_rate": 0.00011817674766232734,
+      "loss": 2.8658,
+      "step": 4474
+    },
+    {
+      "epoch": 0.4050873540327691,
+      "grad_norm": 0.86082524061203,
+      "learning_rate": 0.00011814554211575027,
+      "loss": 2.8829,
+      "step": 4475
+    },
+    {
+      "epoch": 0.4051778763465194,
+      "grad_norm": 0.8323209881782532,
+      "learning_rate": 0.0001181143347419176,
+      "loss": 2.834,
+      "step": 4476
+    },
+    {
+      "epoch": 0.40526839866026976,
+      "grad_norm": 0.8598990440368652,
+      "learning_rate": 0.00011808312554397192,
+      "loss": 2.8747,
+      "step": 4477
+    },
+    {
+      "epoch": 0.4053589209740201,
+      "grad_norm": 0.887123167514801,
+      "learning_rate": 0.00011805191452505602,
+      "loss": 2.9156,
+      "step": 4478
+    },
+    {
+      "epoch": 0.40544944328777044,
+      "grad_norm": 0.9374668598175049,
+      "learning_rate": 0.00011802070168831279,
+      "loss": 2.9036,
+      "step": 4479
+    },
+    {
+      "epoch": 0.4055399656015208,
+      "grad_norm": 0.9349672198295593,
+      "learning_rate": 0.00011798948703688539,
+      "loss": 2.8519,
+      "step": 4480
+    },
+    {
+      "epoch": 0.4056304879152711,
+      "grad_norm": 0.8681858777999878,
+      "learning_rate": 0.00011795827057391715,
+      "loss": 2.8797,
+      "step": 4481
+    },
+    {
+      "epoch": 0.40572101022902146,
+      "grad_norm": 0.8855987191200256,
+      "learning_rate": 0.00011792705230255157,
+      "loss": 2.8729,
+      "step": 4482
+    },
+    {
+      "epoch": 0.4058115325427718,
+      "grad_norm": 0.8382616639137268,
+      "learning_rate": 0.00011789583222593227,
+      "loss": 2.7793,
+      "step": 4483
+    },
+    {
+      "epoch": 0.40590205485652214,
+      "grad_norm": 0.8915740251541138,
+      "learning_rate": 0.00011786461034720319,
+      "loss": 2.8429,
+      "step": 4484
+    },
+    {
+      "epoch": 0.4059925771702725,
+      "grad_norm": 0.9631752967834473,
+      "learning_rate": 0.00011783338666950831,
+      "loss": 2.899,
+      "step": 4485
+    },
+    {
+      "epoch": 0.4060830994840228,
+      "grad_norm": 0.8699368238449097,
+      "learning_rate": 0.00011780216119599192,
+      "loss": 2.8685,
+      "step": 4486
+    },
+    {
+      "epoch": 0.40617362179777317,
+      "grad_norm": 0.9649322032928467,
+      "learning_rate": 0.00011777093392979837,
+      "loss": 2.822,
+      "step": 4487
+    },
+    {
+      "epoch": 0.4062641441115235,
+      "grad_norm": 0.9097427725791931,
+      "learning_rate": 0.00011773970487407224,
+      "loss": 2.8564,
+      "step": 4488
+    },
+    {
+      "epoch": 0.40635466642527385,
+      "grad_norm": 0.8963183164596558,
+      "learning_rate": 0.00011770847403195834,
+      "loss": 2.8812,
+      "step": 4489
+    },
+    {
+      "epoch": 0.4064451887390242,
+      "grad_norm": 0.9345114827156067,
+      "learning_rate": 0.00011767724140660157,
+      "loss": 2.8541,
+      "step": 4490
+    },
+    {
+      "epoch": 0.40653571105277453,
+      "grad_norm": 0.9190236926078796,
+      "learning_rate": 0.00011764600700114708,
+      "loss": 2.9107,
+      "step": 4491
+    },
+    {
+      "epoch": 0.40662623336652487,
+      "grad_norm": 0.8682000041007996,
+      "learning_rate": 0.00011761477081874015,
+      "loss": 2.9114,
+      "step": 4492
+    },
+    {
+      "epoch": 0.4067167556802752,
+      "grad_norm": 0.9088363647460938,
+      "learning_rate": 0.00011758353286252632,
+      "loss": 2.8547,
+      "step": 4493
+    },
+    {
+      "epoch": 0.40680727799402555,
+      "grad_norm": 0.8862866759300232,
+      "learning_rate": 0.0001175522931356512,
+      "loss": 2.8347,
+      "step": 4494
+    },
+    {
+      "epoch": 0.4068978003077759,
+      "grad_norm": 0.8324484825134277,
+      "learning_rate": 0.00011752105164126062,
+      "loss": 2.8227,
+      "step": 4495
+    },
+    {
+      "epoch": 0.40698832262152623,
+      "grad_norm": 0.9196053743362427,
+      "learning_rate": 0.00011748980838250065,
+      "loss": 2.793,
+      "step": 4496
+    },
+    {
+      "epoch": 0.4070788449352765,
+      "grad_norm": 0.8978742361068726,
+      "learning_rate": 0.00011745856336251742,
+      "loss": 2.8878,
+      "step": 4497
+    },
+    {
+      "epoch": 0.40716936724902686,
+      "grad_norm": 0.8683122992515564,
+      "learning_rate": 0.00011742731658445738,
+      "loss": 2.855,
+      "step": 4498
+    },
+    {
+      "epoch": 0.4072598895627772,
+      "grad_norm": 0.8661417365074158,
+      "learning_rate": 0.00011739606805146701,
+      "loss": 2.8421,
+      "step": 4499
+    },
+    {
+      "epoch": 0.40735041187652754,
+      "grad_norm": 0.9605647921562195,
+      "learning_rate": 0.00011736481776669306,
+      "loss": 2.803,
+      "step": 4500
+    },
+    {
+      "epoch": 0.4074409341902779,
+      "grad_norm": 0.8836671710014343,
+      "learning_rate": 0.00011733356573328244,
+      "loss": 2.8395,
+      "step": 4501
+    },
+    {
+      "epoch": 0.4075314565040282,
+      "grad_norm": 0.9318768382072449,
+      "learning_rate": 0.00011730231195438222,
+      "loss": 2.9127,
+      "step": 4502
+    },
+    {
+      "epoch": 0.40762197881777856,
+      "grad_norm": 0.8878833651542664,
+      "learning_rate": 0.00011727105643313966,
+      "loss": 2.8762,
+      "step": 4503
+    },
+    {
+      "epoch": 0.4077125011315289,
+      "grad_norm": 0.8291527032852173,
+      "learning_rate": 0.00011723979917270218,
+      "loss": 2.8568,
+      "step": 4504
+    },
+    {
+      "epoch": 0.40780302344527924,
+      "grad_norm": 0.9366280436515808,
+      "learning_rate": 0.00011720854017621743,
+      "loss": 2.8106,
+      "step": 4505
+    },
+    {
+      "epoch": 0.4078935457590296,
+      "grad_norm": 0.8883543610572815,
+      "learning_rate": 0.00011717727944683313,
+      "loss": 2.8194,
+      "step": 4506
+    },
+    {
+      "epoch": 0.4079840680727799,
+      "grad_norm": 0.9248679280281067,
+      "learning_rate": 0.00011714601698769725,
+      "loss": 2.8272,
+      "step": 4507
+    },
+    {
+      "epoch": 0.40807459038653027,
+      "grad_norm": 0.9774189591407776,
+      "learning_rate": 0.00011711475280195794,
+      "loss": 2.8291,
+      "step": 4508
+    },
+    {
+      "epoch": 0.4081651127002806,
+      "grad_norm": 0.9434395432472229,
+      "learning_rate": 0.00011708348689276351,
+      "loss": 2.8495,
+      "step": 4509
+    },
+    {
+      "epoch": 0.40825563501403095,
+      "grad_norm": 0.8033177256584167,
+      "learning_rate": 0.0001170522192632624,
+      "loss": 2.8239,
+      "step": 4510
+    },
+    {
+      "epoch": 0.4083461573277813,
+      "grad_norm": 0.8647253513336182,
+      "learning_rate": 0.00011702094991660326,
+      "loss": 2.8936,
+      "step": 4511
+    },
+    {
+      "epoch": 0.40843667964153163,
+      "grad_norm": 0.8673940896987915,
+      "learning_rate": 0.00011698967885593495,
+      "loss": 2.7856,
+      "step": 4512
+    },
+    {
+      "epoch": 0.40852720195528197,
+      "grad_norm": 0.8677223920822144,
+      "learning_rate": 0.00011695840608440644,
+      "loss": 2.8896,
+      "step": 4513
+    },
+    {
+      "epoch": 0.4086177242690323,
+      "grad_norm": 0.996747612953186,
+      "learning_rate": 0.00011692713160516691,
+      "loss": 2.88,
+      "step": 4514
+    },
+    {
+      "epoch": 0.40870824658278265,
+      "grad_norm": 0.9442813992500305,
+      "learning_rate": 0.00011689585542136567,
+      "loss": 2.9284,
+      "step": 4515
+    },
+    {
+      "epoch": 0.408798768896533,
+      "grad_norm": 0.9169313311576843,
+      "learning_rate": 0.00011686457753615228,
+      "loss": 2.8345,
+      "step": 4516
+    },
+    {
+      "epoch": 0.40888929121028333,
+      "grad_norm": 0.847822904586792,
+      "learning_rate": 0.00011683329795267636,
+      "loss": 2.8906,
+      "step": 4517
+    },
+    {
+      "epoch": 0.4089798135240337,
+      "grad_norm": 0.9691492915153503,
+      "learning_rate": 0.00011680201667408783,
+      "loss": 2.8347,
+      "step": 4518
+    },
+    {
+      "epoch": 0.409070335837784,
+      "grad_norm": 0.8392035961151123,
+      "learning_rate": 0.00011677073370353666,
+      "loss": 2.8626,
+      "step": 4519
+    },
+    {
+      "epoch": 0.40916085815153436,
+      "grad_norm": 0.8107256293296814,
+      "learning_rate": 0.00011673944904417308,
+      "loss": 2.8532,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4092513804652847,
+      "grad_norm": 0.9076669812202454,
+      "learning_rate": 0.00011670816269914742,
+      "loss": 2.8716,
+      "step": 4521
+    },
+    {
+      "epoch": 0.40934190277903504,
+      "grad_norm": 0.8376779556274414,
+      "learning_rate": 0.00011667687467161024,
+      "loss": 2.8887,
+      "step": 4522
+    },
+    {
+      "epoch": 0.4094324250927854,
+      "grad_norm": 0.759242594242096,
+      "learning_rate": 0.00011664558496471223,
+      "loss": 2.84,
+      "step": 4523
+    },
+    {
+      "epoch": 0.4095229474065357,
+      "grad_norm": 0.815267026424408,
+      "learning_rate": 0.00011661429358160429,
+      "loss": 2.8227,
+      "step": 4524
+    },
+    {
+      "epoch": 0.40961346972028606,
+      "grad_norm": 0.854684591293335,
+      "learning_rate": 0.00011658300052543742,
+      "loss": 2.8768,
+      "step": 4525
+    },
+    {
+      "epoch": 0.4097039920340364,
+      "grad_norm": 0.8627126216888428,
+      "learning_rate": 0.00011655170579936286,
+      "loss": 2.8762,
+      "step": 4526
+    },
+    {
+      "epoch": 0.40979451434778674,
+      "grad_norm": 0.8167665004730225,
+      "learning_rate": 0.00011652040940653197,
+      "loss": 2.8455,
+      "step": 4527
+    },
+    {
+      "epoch": 0.4098850366615371,
+      "grad_norm": 0.9664933681488037,
+      "learning_rate": 0.00011648911135009634,
+      "loss": 2.8866,
+      "step": 4528
+    },
+    {
+      "epoch": 0.4099755589752874,
+      "grad_norm": 0.8510946035385132,
+      "learning_rate": 0.00011645781163320761,
+      "loss": 2.8484,
+      "step": 4529
+    },
+    {
+      "epoch": 0.41006608128903776,
+      "grad_norm": 0.9158300757408142,
+      "learning_rate": 0.00011642651025901772,
+      "loss": 2.9048,
+      "step": 4530
+    },
+    {
+      "epoch": 0.4101566036027881,
+      "grad_norm": 0.8029270172119141,
+      "learning_rate": 0.00011639520723067868,
+      "loss": 2.8317,
+      "step": 4531
+    },
+    {
+      "epoch": 0.41024712591653845,
+      "grad_norm": 0.962932288646698,
+      "learning_rate": 0.00011636390255134273,
+      "loss": 2.8754,
+      "step": 4532
+    },
+    {
+      "epoch": 0.4103376482302888,
+      "grad_norm": 0.9180269241333008,
+      "learning_rate": 0.00011633259622416224,
+      "loss": 2.869,
+      "step": 4533
+    },
+    {
+      "epoch": 0.41042817054403913,
+      "grad_norm": 0.9610112905502319,
+      "learning_rate": 0.00011630128825228974,
+      "loss": 2.878,
+      "step": 4534
+    },
+    {
+      "epoch": 0.41051869285778947,
+      "grad_norm": 0.8697507381439209,
+      "learning_rate": 0.000116269978638878,
+      "loss": 2.8602,
+      "step": 4535
+    },
+    {
+      "epoch": 0.4106092151715398,
+      "grad_norm": 0.9531928300857544,
+      "learning_rate": 0.00011623866738707983,
+      "loss": 2.8178,
+      "step": 4536
+    },
+    {
+      "epoch": 0.41069973748529015,
+      "grad_norm": 0.870388388633728,
+      "learning_rate": 0.00011620735450004829,
+      "loss": 2.8233,
+      "step": 4537
+    },
+    {
+      "epoch": 0.41079025979904044,
+      "grad_norm": 0.8597344160079956,
+      "learning_rate": 0.0001161760399809366,
+      "loss": 2.8641,
+      "step": 4538
+    },
+    {
+      "epoch": 0.4108807821127908,
+      "grad_norm": 0.8699856400489807,
+      "learning_rate": 0.00011614472383289814,
+      "loss": 2.8854,
+      "step": 4539
+    },
+    {
+      "epoch": 0.4109713044265411,
+      "grad_norm": 0.8118836283683777,
+      "learning_rate": 0.00011611340605908642,
+      "loss": 2.818,
+      "step": 4540
+    },
+    {
+      "epoch": 0.41106182674029146,
+      "grad_norm": 0.8394683003425598,
+      "learning_rate": 0.00011608208666265514,
+      "loss": 2.8377,
+      "step": 4541
+    },
+    {
+      "epoch": 0.4111523490540418,
+      "grad_norm": 0.9108496904373169,
+      "learning_rate": 0.00011605076564675818,
+      "loss": 2.8582,
+      "step": 4542
+    },
+    {
+      "epoch": 0.41124287136779214,
+      "grad_norm": 0.9241228699684143,
+      "learning_rate": 0.0001160194430145496,
+      "loss": 2.8557,
+      "step": 4543
+    },
+    {
+      "epoch": 0.4113333936815425,
+      "grad_norm": 0.8194548487663269,
+      "learning_rate": 0.0001159881187691835,
+      "loss": 2.8321,
+      "step": 4544
+    },
+    {
+      "epoch": 0.4114239159952928,
+      "grad_norm": 0.8807789087295532,
+      "learning_rate": 0.00011595679291381429,
+      "loss": 2.7702,
+      "step": 4545
+    },
+    {
+      "epoch": 0.41151443830904316,
+      "grad_norm": 0.9182827472686768,
+      "learning_rate": 0.00011592546545159645,
+      "loss": 2.8463,
+      "step": 4546
+    },
+    {
+      "epoch": 0.4116049606227935,
+      "grad_norm": 0.8201344013214111,
+      "learning_rate": 0.00011589413638568471,
+      "loss": 2.8441,
+      "step": 4547
+    },
+    {
+      "epoch": 0.41169548293654384,
+      "grad_norm": 0.9747979640960693,
+      "learning_rate": 0.00011586280571923384,
+      "loss": 2.8377,
+      "step": 4548
+    },
+    {
+      "epoch": 0.4117860052502942,
+      "grad_norm": 0.9699158072471619,
+      "learning_rate": 0.00011583147345539889,
+      "loss": 2.8487,
+      "step": 4549
+    },
+    {
+      "epoch": 0.4118765275640445,
+      "grad_norm": 0.8967488408088684,
+      "learning_rate": 0.000115800139597335,
+      "loss": 2.8408,
+      "step": 4550
+    },
+    {
+      "epoch": 0.41196704987779487,
+      "grad_norm": 0.9237414598464966,
+      "learning_rate": 0.00011576880414819749,
+      "loss": 2.8475,
+      "step": 4551
+    },
+    {
+      "epoch": 0.4120575721915452,
+      "grad_norm": 0.8308104276657104,
+      "learning_rate": 0.00011573746711114179,
+      "loss": 2.869,
+      "step": 4552
+    },
+    {
+      "epoch": 0.41214809450529555,
+      "grad_norm": 0.9393368363380432,
+      "learning_rate": 0.00011570612848932361,
+      "loss": 2.8297,
+      "step": 4553
+    },
+    {
+      "epoch": 0.4122386168190459,
+      "grad_norm": 0.8871431946754456,
+      "learning_rate": 0.00011567478828589875,
+      "loss": 2.7945,
+      "step": 4554
+    },
+    {
+      "epoch": 0.41232913913279623,
+      "grad_norm": 0.9611302018165588,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 2.8664,
+      "step": 4555
+    },
+    {
+      "epoch": 0.41241966144654657,
+      "grad_norm": 0.9113096594810486,
+      "learning_rate": 0.00011561210314685282,
+      "loss": 2.8138,
+      "step": 4556
+    },
+    {
+      "epoch": 0.4125101837602969,
+      "grad_norm": 0.8560320138931274,
+      "learning_rate": 0.00011558075821754417,
+      "loss": 2.8107,
+      "step": 4557
+    },
+    {
+      "epoch": 0.41260070607404725,
+      "grad_norm": 0.8099222183227539,
+      "learning_rate": 0.00011554941171925365,
+      "loss": 2.8509,
+      "step": 4558
+    },
+    {
+      "epoch": 0.4126912283877976,
+      "grad_norm": 0.9044903516769409,
+      "learning_rate": 0.00011551806365513773,
+      "loss": 2.8588,
+      "step": 4559
+    },
+    {
+      "epoch": 0.41278175070154793,
+      "grad_norm": 0.9626808762550354,
+      "learning_rate": 0.00011548671402835325,
+      "loss": 2.87,
+      "step": 4560
+    },
+    {
+      "epoch": 0.4128722730152983,
+      "grad_norm": 0.8645593523979187,
+      "learning_rate": 0.00011545536284205707,
+      "loss": 2.8122,
+      "step": 4561
+    },
+    {
+      "epoch": 0.4129627953290486,
+      "grad_norm": 0.92280113697052,
+      "learning_rate": 0.00011542401009940631,
+      "loss": 2.8386,
+      "step": 4562
+    },
+    {
+      "epoch": 0.41305331764279896,
+      "grad_norm": 0.8630074858665466,
+      "learning_rate": 0.00011539265580355814,
+      "loss": 2.7993,
+      "step": 4563
+    },
+    {
+      "epoch": 0.4131438399565493,
+      "grad_norm": 0.9118307828903198,
+      "learning_rate": 0.00011536129995766996,
+      "loss": 2.8264,
+      "step": 4564
+    },
+    {
+      "epoch": 0.41323436227029964,
+      "grad_norm": 0.9462095499038696,
+      "learning_rate": 0.00011532994256489927,
+      "loss": 2.8627,
+      "step": 4565
+    },
+    {
+      "epoch": 0.41332488458405,
+      "grad_norm": 0.9295179843902588,
+      "learning_rate": 0.00011529858362840382,
+      "loss": 2.8083,
+      "step": 4566
+    },
+    {
+      "epoch": 0.4134154068978003,
+      "grad_norm": 0.8588746190071106,
+      "learning_rate": 0.00011526722315134139,
+      "loss": 2.8716,
+      "step": 4567
+    },
+    {
+      "epoch": 0.41350592921155066,
+      "grad_norm": 0.9106554388999939,
+      "learning_rate": 0.00011523586113687001,
+      "loss": 2.9148,
+      "step": 4568
+    },
+    {
+      "epoch": 0.413596451525301,
+      "grad_norm": 0.8623448610305786,
+      "learning_rate": 0.00011520449758814784,
+      "loss": 2.8142,
+      "step": 4569
+    },
+    {
+      "epoch": 0.41368697383905134,
+      "grad_norm": 0.898600161075592,
+      "learning_rate": 0.00011517313250833317,
+      "loss": 2.887,
+      "step": 4570
+    },
+    {
+      "epoch": 0.4137774961528017,
+      "grad_norm": 0.8837737441062927,
+      "learning_rate": 0.00011514176590058448,
+      "loss": 2.7951,
+      "step": 4571
+    },
+    {
+      "epoch": 0.413868018466552,
+      "grad_norm": 0.8727512955665588,
+      "learning_rate": 0.00011511039776806037,
+      "loss": 2.868,
+      "step": 4572
+    },
+    {
+      "epoch": 0.41395854078030236,
+      "grad_norm": 0.9692105650901794,
+      "learning_rate": 0.00011507902811391961,
+      "loss": 2.8471,
+      "step": 4573
+    },
+    {
+      "epoch": 0.4140490630940527,
+      "grad_norm": 0.8482305407524109,
+      "learning_rate": 0.00011504765694132116,
+      "loss": 2.8946,
+      "step": 4574
+    },
+    {
+      "epoch": 0.41413958540780305,
+      "grad_norm": 0.7920590043067932,
+      "learning_rate": 0.00011501628425342403,
+      "loss": 2.8343,
+      "step": 4575
+    },
+    {
+      "epoch": 0.4142301077215534,
+      "grad_norm": 0.8491135835647583,
+      "learning_rate": 0.0001149849100533875,
+      "loss": 2.8269,
+      "step": 4576
+    },
+    {
+      "epoch": 0.4143206300353037,
+      "grad_norm": 0.8065090775489807,
+      "learning_rate": 0.00011495353434437098,
+      "loss": 2.8108,
+      "step": 4577
+    },
+    {
+      "epoch": 0.41441115234905407,
+      "grad_norm": 0.8313660025596619,
+      "learning_rate": 0.0001149221571295339,
+      "loss": 2.8532,
+      "step": 4578
+    },
+    {
+      "epoch": 0.41450167466280435,
+      "grad_norm": 0.8295189142227173,
+      "learning_rate": 0.00011489077841203602,
+      "loss": 2.8218,
+      "step": 4579
+    },
+    {
+      "epoch": 0.4145921969765547,
+      "grad_norm": 0.8401261568069458,
+      "learning_rate": 0.00011485939819503717,
+      "loss": 2.848,
+      "step": 4580
+    },
+    {
+      "epoch": 0.41468271929030504,
+      "grad_norm": 0.9519902467727661,
+      "learning_rate": 0.00011482801648169736,
+      "loss": 2.8726,
+      "step": 4581
+    },
+    {
+      "epoch": 0.4147732416040554,
+      "grad_norm": 1.0146515369415283,
+      "learning_rate": 0.00011479663327517667,
+      "loss": 2.8943,
+      "step": 4582
+    },
+    {
+      "epoch": 0.4148637639178057,
+      "grad_norm": 0.8038744926452637,
+      "learning_rate": 0.0001147652485786354,
+      "loss": 2.8749,
+      "step": 4583
+    },
+    {
+      "epoch": 0.41495428623155606,
+      "grad_norm": 0.9129198789596558,
+      "learning_rate": 0.000114733862395234,
+      "loss": 2.8427,
+      "step": 4584
+    },
+    {
+      "epoch": 0.4150448085453064,
+      "grad_norm": 0.8790321350097656,
+      "learning_rate": 0.00011470247472813309,
+      "loss": 2.8366,
+      "step": 4585
+    },
+    {
+      "epoch": 0.41513533085905674,
+      "grad_norm": 0.8764024972915649,
+      "learning_rate": 0.00011467108558049335,
+      "loss": 2.8418,
+      "step": 4586
+    },
+    {
+      "epoch": 0.4152258531728071,
+      "grad_norm": 0.8494692444801331,
+      "learning_rate": 0.0001146396949554757,
+      "loss": 2.8391,
+      "step": 4587
+    },
+    {
+      "epoch": 0.4153163754865574,
+      "grad_norm": 0.874064028263092,
+      "learning_rate": 0.00011460830285624118,
+      "loss": 2.7919,
+      "step": 4588
+    },
+    {
+      "epoch": 0.41540689780030776,
+      "grad_norm": 0.8704847693443298,
+      "learning_rate": 0.00011457690928595095,
+      "loss": 2.8265,
+      "step": 4589
+    },
+    {
+      "epoch": 0.4154974201140581,
+      "grad_norm": 0.8909250497817993,
+      "learning_rate": 0.00011454551424776637,
+      "loss": 2.7961,
+      "step": 4590
+    },
+    {
+      "epoch": 0.41558794242780844,
+      "grad_norm": 0.8759780526161194,
+      "learning_rate": 0.00011451411774484889,
+      "loss": 2.804,
+      "step": 4591
+    },
+    {
+      "epoch": 0.4156784647415588,
+      "grad_norm": 0.8733645081520081,
+      "learning_rate": 0.00011448271978036013,
+      "loss": 2.8749,
+      "step": 4592
+    },
+    {
+      "epoch": 0.4157689870553091,
+      "grad_norm": 0.9054057002067566,
+      "learning_rate": 0.0001144513203574619,
+      "loss": 2.8899,
+      "step": 4593
+    },
+    {
+      "epoch": 0.41585950936905947,
+      "grad_norm": 0.8773120641708374,
+      "learning_rate": 0.00011441991947931612,
+      "loss": 2.8509,
+      "step": 4594
+    },
+    {
+      "epoch": 0.4159500316828098,
+      "grad_norm": 0.9054686427116394,
+      "learning_rate": 0.00011438851714908483,
+      "loss": 2.8504,
+      "step": 4595
+    },
+    {
+      "epoch": 0.41604055399656015,
+      "grad_norm": 0.9160197377204895,
+      "learning_rate": 0.00011435711336993027,
+      "loss": 2.872,
+      "step": 4596
+    },
+    {
+      "epoch": 0.4161310763103105,
+      "grad_norm": 0.8991233706474304,
+      "learning_rate": 0.00011432570814501478,
+      "loss": 2.8094,
+      "step": 4597
+    },
+    {
+      "epoch": 0.41622159862406083,
+      "grad_norm": 0.9077282547950745,
+      "learning_rate": 0.00011429430147750087,
+      "loss": 2.9541,
+      "step": 4598
+    },
+    {
+      "epoch": 0.41631212093781117,
+      "grad_norm": 1.050952434539795,
+      "learning_rate": 0.00011426289337055119,
+      "loss": 2.9243,
+      "step": 4599
+    },
+    {
+      "epoch": 0.4164026432515615,
+      "grad_norm": 1.1110472679138184,
+      "learning_rate": 0.00011423148382732853,
+      "loss": 2.876,
+      "step": 4600
+    },
+    {
+      "epoch": 0.41649316556531185,
+      "grad_norm": 0.9317214488983154,
+      "learning_rate": 0.00011420007285099585,
+      "loss": 2.8084,
+      "step": 4601
+    },
+    {
+      "epoch": 0.4165836878790622,
+      "grad_norm": 0.9028974771499634,
+      "learning_rate": 0.00011416866044471622,
+      "loss": 2.859,
+      "step": 4602
+    },
+    {
+      "epoch": 0.41667421019281253,
+      "grad_norm": 0.9547843337059021,
+      "learning_rate": 0.00011413724661165286,
+      "loss": 2.8426,
+      "step": 4603
+    },
+    {
+      "epoch": 0.4167647325065629,
+      "grad_norm": 0.8694342970848083,
+      "learning_rate": 0.00011410583135496917,
+      "loss": 2.8205,
+      "step": 4604
+    },
+    {
+      "epoch": 0.4168552548203132,
+      "grad_norm": 0.872846245765686,
+      "learning_rate": 0.00011407441467782865,
+      "loss": 2.8493,
+      "step": 4605
+    },
+    {
+      "epoch": 0.41694577713406356,
+      "grad_norm": 1.0523523092269897,
+      "learning_rate": 0.00011404299658339493,
+      "loss": 2.8408,
+      "step": 4606
+    },
+    {
+      "epoch": 0.4170362994478139,
+      "grad_norm": 0.9270108342170715,
+      "learning_rate": 0.00011401157707483185,
+      "loss": 2.8591,
+      "step": 4607
+    },
+    {
+      "epoch": 0.41712682176156424,
+      "grad_norm": 0.9646519422531128,
+      "learning_rate": 0.00011398015615530337,
+      "loss": 2.8092,
+      "step": 4608
+    },
+    {
+      "epoch": 0.4172173440753146,
+      "grad_norm": 0.8431654572486877,
+      "learning_rate": 0.00011394873382797351,
+      "loss": 2.8707,
+      "step": 4609
+    },
+    {
+      "epoch": 0.4173078663890649,
+      "grad_norm": 1.0408530235290527,
+      "learning_rate": 0.00011391731009600654,
+      "loss": 2.7969,
+      "step": 4610
+    },
+    {
+      "epoch": 0.41739838870281526,
+      "grad_norm": 0.906665563583374,
+      "learning_rate": 0.00011388588496256685,
+      "loss": 2.861,
+      "step": 4611
+    },
+    {
+      "epoch": 0.4174889110165656,
+      "grad_norm": 0.9940811395645142,
+      "learning_rate": 0.00011385445843081892,
+      "loss": 2.8514,
+      "step": 4612
+    },
+    {
+      "epoch": 0.41757943333031594,
+      "grad_norm": 0.8722427487373352,
+      "learning_rate": 0.0001138230305039274,
+      "loss": 2.7887,
+      "step": 4613
+    },
+    {
+      "epoch": 0.4176699556440663,
+      "grad_norm": 0.9708684682846069,
+      "learning_rate": 0.00011379160118505705,
+      "loss": 2.8326,
+      "step": 4614
+    },
+    {
+      "epoch": 0.4177604779578166,
+      "grad_norm": 1.0897921323776245,
+      "learning_rate": 0.00011376017047737291,
+      "loss": 2.8131,
+      "step": 4615
+    },
+    {
+      "epoch": 0.41785100027156696,
+      "grad_norm": 0.9895443320274353,
+      "learning_rate": 0.00011372873838403993,
+      "loss": 2.8639,
+      "step": 4616
+    },
+    {
+      "epoch": 0.4179415225853173,
+      "grad_norm": 1.0044602155685425,
+      "learning_rate": 0.00011369730490822336,
+      "loss": 2.8073,
+      "step": 4617
+    },
+    {
+      "epoch": 0.41803204489906765,
+      "grad_norm": 1.0245896577835083,
+      "learning_rate": 0.00011366587005308858,
+      "loss": 2.8415,
+      "step": 4618
+    },
+    {
+      "epoch": 0.418122567212818,
+      "grad_norm": 0.8707718849182129,
+      "learning_rate": 0.00011363443382180108,
+      "loss": 2.8101,
+      "step": 4619
+    },
+    {
+      "epoch": 0.4182130895265683,
+      "grad_norm": 0.8844798803329468,
+      "learning_rate": 0.00011360299621752644,
+      "loss": 2.8005,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4183036118403186,
+      "grad_norm": 0.9705588817596436,
+      "learning_rate": 0.00011357155724343045,
+      "loss": 2.8566,
+      "step": 4621
+    },
+    {
+      "epoch": 0.41839413415406895,
+      "grad_norm": 0.8912902474403381,
+      "learning_rate": 0.00011354011690267903,
+      "loss": 2.8557,
+      "step": 4622
+    },
+    {
+      "epoch": 0.4184846564678193,
+      "grad_norm": 0.8974612951278687,
+      "learning_rate": 0.00011350867519843821,
+      "loss": 2.9134,
+      "step": 4623
+    },
+    {
+      "epoch": 0.41857517878156963,
+      "grad_norm": 0.8711275458335876,
+      "learning_rate": 0.00011347723213387416,
+      "loss": 2.7833,
+      "step": 4624
+    },
+    {
+      "epoch": 0.41866570109532,
+      "grad_norm": 0.876578152179718,
+      "learning_rate": 0.00011344578771215319,
+      "loss": 2.889,
+      "step": 4625
+    },
+    {
+      "epoch": 0.4187562234090703,
+      "grad_norm": 0.8725136518478394,
+      "learning_rate": 0.00011341434193644178,
+      "loss": 2.8994,
+      "step": 4626
+    },
+    {
+      "epoch": 0.41884674572282066,
+      "grad_norm": 0.9713298082351685,
+      "learning_rate": 0.00011338289480990652,
+      "loss": 2.8326,
+      "step": 4627
+    },
+    {
+      "epoch": 0.418937268036571,
+      "grad_norm": 0.9306853413581848,
+      "learning_rate": 0.00011335144633571411,
+      "loss": 2.8872,
+      "step": 4628
+    },
+    {
+      "epoch": 0.41902779035032134,
+      "grad_norm": 0.8581621646881104,
+      "learning_rate": 0.00011331999651703139,
+      "loss": 2.8281,
+      "step": 4629
+    },
+    {
+      "epoch": 0.4191183126640717,
+      "grad_norm": 0.8599483966827393,
+      "learning_rate": 0.00011328854535702543,
+      "loss": 2.8116,
+      "step": 4630
+    },
+    {
+      "epoch": 0.419208834977822,
+      "grad_norm": 0.8819389939308167,
+      "learning_rate": 0.00011325709285886328,
+      "loss": 2.824,
+      "step": 4631
+    },
+    {
+      "epoch": 0.41929935729157236,
+      "grad_norm": 0.8218708038330078,
+      "learning_rate": 0.00011322563902571226,
+      "loss": 2.8347,
+      "step": 4632
+    },
+    {
+      "epoch": 0.4193898796053227,
+      "grad_norm": 0.768627405166626,
+      "learning_rate": 0.00011319418386073972,
+      "loss": 2.783,
+      "step": 4633
+    },
+    {
+      "epoch": 0.41948040191907304,
+      "grad_norm": 0.8319827914237976,
+      "learning_rate": 0.00011316272736711329,
+      "loss": 2.8694,
+      "step": 4634
+    },
+    {
+      "epoch": 0.4195709242328234,
+      "grad_norm": 0.8202218413352966,
+      "learning_rate": 0.00011313126954800052,
+      "loss": 2.8062,
+      "step": 4635
+    },
+    {
+      "epoch": 0.4196614465465737,
+      "grad_norm": 0.8292239904403687,
+      "learning_rate": 0.0001130998104065693,
+      "loss": 2.9159,
+      "step": 4636
+    },
+    {
+      "epoch": 0.41975196886032407,
+      "grad_norm": 0.8331717848777771,
+      "learning_rate": 0.0001130683499459875,
+      "loss": 2.731,
+      "step": 4637
+    },
+    {
+      "epoch": 0.4198424911740744,
+      "grad_norm": 0.7963569164276123,
+      "learning_rate": 0.00011303688816942324,
+      "loss": 2.8067,
+      "step": 4638
+    },
+    {
+      "epoch": 0.41993301348782475,
+      "grad_norm": 0.8360899090766907,
+      "learning_rate": 0.00011300542508004468,
+      "loss": 2.872,
+      "step": 4639
+    },
+    {
+      "epoch": 0.4200235358015751,
+      "grad_norm": 0.8733784556388855,
+      "learning_rate": 0.00011297396068102017,
+      "loss": 2.8585,
+      "step": 4640
+    },
+    {
+      "epoch": 0.42011405811532543,
+      "grad_norm": 0.8134546875953674,
+      "learning_rate": 0.00011294249497551819,
+      "loss": 2.8574,
+      "step": 4641
+    },
+    {
+      "epoch": 0.42020458042907577,
+      "grad_norm": 0.7982373833656311,
+      "learning_rate": 0.00011291102796670734,
+      "loss": 2.8462,
+      "step": 4642
+    },
+    {
+      "epoch": 0.4202951027428261,
+      "grad_norm": 0.8143455982208252,
+      "learning_rate": 0.0001128795596577563,
+      "loss": 2.815,
+      "step": 4643
+    },
+    {
+      "epoch": 0.42038562505657645,
+      "grad_norm": 0.8223409056663513,
+      "learning_rate": 0.00011284809005183394,
+      "loss": 2.8082,
+      "step": 4644
+    },
+    {
+      "epoch": 0.4204761473703268,
+      "grad_norm": 0.9110050201416016,
+      "learning_rate": 0.0001128166191521093,
+      "loss": 2.8725,
+      "step": 4645
+    },
+    {
+      "epoch": 0.42056666968407713,
+      "grad_norm": 0.8185941576957703,
+      "learning_rate": 0.00011278514696175144,
+      "loss": 2.8505,
+      "step": 4646
+    },
+    {
+      "epoch": 0.4206571919978275,
+      "grad_norm": 0.9211212396621704,
+      "learning_rate": 0.00011275367348392962,
+      "loss": 2.8217,
+      "step": 4647
+    },
+    {
+      "epoch": 0.4207477143115778,
+      "grad_norm": 0.8780548572540283,
+      "learning_rate": 0.00011272219872181322,
+      "loss": 2.8891,
+      "step": 4648
+    },
+    {
+      "epoch": 0.42083823662532815,
+      "grad_norm": 0.9037941098213196,
+      "learning_rate": 0.00011269072267857182,
+      "loss": 2.8489,
+      "step": 4649
+    },
+    {
+      "epoch": 0.4209287589390785,
+      "grad_norm": 0.8477228283882141,
+      "learning_rate": 0.00011265924535737493,
+      "loss": 2.8404,
+      "step": 4650
+    },
+    {
+      "epoch": 0.42101928125282884,
+      "grad_norm": 0.8942874073982239,
+      "learning_rate": 0.0001126277667613924,
+      "loss": 2.7807,
+      "step": 4651
+    },
+    {
+      "epoch": 0.4211098035665792,
+      "grad_norm": 0.8547981381416321,
+      "learning_rate": 0.00011259628689379411,
+      "loss": 2.8997,
+      "step": 4652
+    },
+    {
+      "epoch": 0.4212003258803295,
+      "grad_norm": 0.8350889682769775,
+      "learning_rate": 0.00011256480575775007,
+      "loss": 2.8157,
+      "step": 4653
+    },
+    {
+      "epoch": 0.42129084819407986,
+      "grad_norm": 1.0040391683578491,
+      "learning_rate": 0.00011253332335643043,
+      "loss": 2.8049,
+      "step": 4654
+    },
+    {
+      "epoch": 0.4213813705078302,
+      "grad_norm": 0.8473368287086487,
+      "learning_rate": 0.00011250183969300548,
+      "loss": 2.8186,
+      "step": 4655
+    },
+    {
+      "epoch": 0.42147189282158054,
+      "grad_norm": 0.8104838728904724,
+      "learning_rate": 0.0001124703547706456,
+      "loss": 2.7801,
+      "step": 4656
+    },
+    {
+      "epoch": 0.4215624151353309,
+      "grad_norm": 0.8853965401649475,
+      "learning_rate": 0.00011243886859252135,
+      "loss": 2.7825,
+      "step": 4657
+    },
+    {
+      "epoch": 0.4216529374490812,
+      "grad_norm": 0.9112235307693481,
+      "learning_rate": 0.00011240738116180336,
+      "loss": 2.8799,
+      "step": 4658
+    },
+    {
+      "epoch": 0.42174345976283156,
+      "grad_norm": 0.8312155604362488,
+      "learning_rate": 0.00011237589248166242,
+      "loss": 2.8449,
+      "step": 4659
+    },
+    {
+      "epoch": 0.4218339820765819,
+      "grad_norm": 0.8818202018737793,
+      "learning_rate": 0.00011234440255526948,
+      "loss": 2.8088,
+      "step": 4660
+    },
+    {
+      "epoch": 0.42192450439033224,
+      "grad_norm": 0.9247568249702454,
+      "learning_rate": 0.00011231291138579552,
+      "loss": 2.7613,
+      "step": 4661
+    },
+    {
+      "epoch": 0.42201502670408253,
+      "grad_norm": 0.8678056597709656,
+      "learning_rate": 0.00011228141897641172,
+      "loss": 2.8426,
+      "step": 4662
+    },
+    {
+      "epoch": 0.42210554901783287,
+      "grad_norm": 0.8642814755439758,
+      "learning_rate": 0.00011224992533028935,
+      "loss": 2.7944,
+      "step": 4663
+    },
+    {
+      "epoch": 0.4221960713315832,
+      "grad_norm": 0.9176373481750488,
+      "learning_rate": 0.00011221843045059988,
+      "loss": 2.7694,
+      "step": 4664
+    },
+    {
+      "epoch": 0.42228659364533355,
+      "grad_norm": 0.8463912010192871,
+      "learning_rate": 0.00011218693434051475,
+      "loss": 2.8044,
+      "step": 4665
+    },
+    {
+      "epoch": 0.4223771159590839,
+      "grad_norm": 0.9299211502075195,
+      "learning_rate": 0.0001121554370032057,
+      "loss": 2.8042,
+      "step": 4666
+    },
+    {
+      "epoch": 0.42246763827283423,
+      "grad_norm": 0.9610387682914734,
+      "learning_rate": 0.00011212393844184445,
+      "loss": 2.8076,
+      "step": 4667
+    },
+    {
+      "epoch": 0.4225581605865846,
+      "grad_norm": 0.8744280338287354,
+      "learning_rate": 0.00011209243865960296,
+      "loss": 2.807,
+      "step": 4668
+    },
+    {
+      "epoch": 0.4226486829003349,
+      "grad_norm": 0.8455687761306763,
+      "learning_rate": 0.0001120609376596532,
+      "loss": 2.7714,
+      "step": 4669
+    },
+    {
+      "epoch": 0.42273920521408526,
+      "grad_norm": 0.8594902753829956,
+      "learning_rate": 0.00011202943544516736,
+      "loss": 2.7968,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4228297275278356,
+      "grad_norm": 0.8992725610733032,
+      "learning_rate": 0.00011199793201931767,
+      "loss": 2.8628,
+      "step": 4671
+    },
+    {
+      "epoch": 0.42292024984158594,
+      "grad_norm": 0.9300136566162109,
+      "learning_rate": 0.00011196642738527659,
+      "loss": 2.8668,
+      "step": 4672
+    },
+    {
+      "epoch": 0.4230107721553363,
+      "grad_norm": 0.8369343876838684,
+      "learning_rate": 0.0001119349215462166,
+      "loss": 2.8436,
+      "step": 4673
+    },
+    {
+      "epoch": 0.4231012944690866,
+      "grad_norm": 0.8958376049995422,
+      "learning_rate": 0.00011190341450531033,
+      "loss": 2.8824,
+      "step": 4674
+    },
+    {
+      "epoch": 0.42319181678283696,
+      "grad_norm": 0.8437206149101257,
+      "learning_rate": 0.00011187190626573052,
+      "loss": 2.7803,
+      "step": 4675
+    },
+    {
+      "epoch": 0.4232823390965873,
+      "grad_norm": 0.8179407119750977,
+      "learning_rate": 0.00011184039683065013,
+      "loss": 2.8214,
+      "step": 4676
+    },
+    {
+      "epoch": 0.42337286141033764,
+      "grad_norm": 0.8458150625228882,
+      "learning_rate": 0.00011180888620324205,
+      "loss": 2.8447,
+      "step": 4677
+    },
+    {
+      "epoch": 0.423463383724088,
+      "grad_norm": 0.9066711664199829,
+      "learning_rate": 0.00011177737438667948,
+      "loss": 2.7854,
+      "step": 4678
+    },
+    {
+      "epoch": 0.4235539060378383,
+      "grad_norm": 0.8020719289779663,
+      "learning_rate": 0.0001117458613841356,
+      "loss": 2.7944,
+      "step": 4679
+    },
+    {
+      "epoch": 0.42364442835158866,
+      "grad_norm": 0.8214885592460632,
+      "learning_rate": 0.00011171434719878384,
+      "loss": 2.8372,
+      "step": 4680
+    },
+    {
+      "epoch": 0.423734950665339,
+      "grad_norm": 0.8697049021720886,
+      "learning_rate": 0.00011168283183379762,
+      "loss": 2.8341,
+      "step": 4681
+    },
+    {
+      "epoch": 0.42382547297908935,
+      "grad_norm": 0.85956871509552,
+      "learning_rate": 0.00011165131529235056,
+      "loss": 2.7766,
+      "step": 4682
+    },
+    {
+      "epoch": 0.4239159952928397,
+      "grad_norm": 0.8642212748527527,
+      "learning_rate": 0.00011161979757761634,
+      "loss": 2.8029,
+      "step": 4683
+    },
+    {
+      "epoch": 0.42400651760659,
+      "grad_norm": 0.864029586315155,
+      "learning_rate": 0.00011158827869276887,
+      "loss": 2.8399,
+      "step": 4684
+    },
+    {
+      "epoch": 0.42409703992034037,
+      "grad_norm": 0.9287242293357849,
+      "learning_rate": 0.00011155675864098201,
+      "loss": 2.854,
+      "step": 4685
+    },
+    {
+      "epoch": 0.4241875622340907,
+      "grad_norm": 0.9051283001899719,
+      "learning_rate": 0.00011152523742542989,
+      "loss": 2.8412,
+      "step": 4686
+    },
+    {
+      "epoch": 0.42427808454784105,
+      "grad_norm": 0.9301141500473022,
+      "learning_rate": 0.00011149371504928668,
+      "loss": 2.8457,
+      "step": 4687
+    },
+    {
+      "epoch": 0.4243686068615914,
+      "grad_norm": 0.8932237029075623,
+      "learning_rate": 0.00011146219151572668,
+      "loss": 2.8585,
+      "step": 4688
+    },
+    {
+      "epoch": 0.42445912917534173,
+      "grad_norm": 0.844142496585846,
+      "learning_rate": 0.0001114306668279243,
+      "loss": 2.8338,
+      "step": 4689
+    },
+    {
+      "epoch": 0.4245496514890921,
+      "grad_norm": 0.8861607909202576,
+      "learning_rate": 0.00011139914098905406,
+      "loss": 2.7525,
+      "step": 4690
+    },
+    {
+      "epoch": 0.4246401738028424,
+      "grad_norm": 0.8443228602409363,
+      "learning_rate": 0.0001113676140022907,
+      "loss": 2.825,
+      "step": 4691
+    },
+    {
+      "epoch": 0.42473069611659275,
+      "grad_norm": 0.804202675819397,
+      "learning_rate": 0.00011133608587080886,
+      "loss": 2.8037,
+      "step": 4692
+    },
+    {
+      "epoch": 0.4248212184303431,
+      "grad_norm": 0.8334986567497253,
+      "learning_rate": 0.00011130455659778349,
+      "loss": 2.8234,
+      "step": 4693
+    },
+    {
+      "epoch": 0.42491174074409344,
+      "grad_norm": 0.8185308575630188,
+      "learning_rate": 0.0001112730261863896,
+      "loss": 2.8714,
+      "step": 4694
+    },
+    {
+      "epoch": 0.4250022630578438,
+      "grad_norm": 0.863042950630188,
+      "learning_rate": 0.00011124149463980229,
+      "loss": 2.8481,
+      "step": 4695
+    },
+    {
+      "epoch": 0.4250927853715941,
+      "grad_norm": 0.8510686755180359,
+      "learning_rate": 0.00011120996196119675,
+      "loss": 2.8536,
+      "step": 4696
+    },
+    {
+      "epoch": 0.42518330768534446,
+      "grad_norm": 0.9181560277938843,
+      "learning_rate": 0.00011117842815374835,
+      "loss": 2.8783,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4252738299990948,
+      "grad_norm": 0.8466564416885376,
+      "learning_rate": 0.00011114689322063255,
+      "loss": 2.8045,
+      "step": 4698
+    },
+    {
+      "epoch": 0.42536435231284514,
+      "grad_norm": 0.9046276211738586,
+      "learning_rate": 0.00011111535716502493,
+      "loss": 2.8196,
+      "step": 4699
+    },
+    {
+      "epoch": 0.4254548746265955,
+      "grad_norm": 1.0147709846496582,
+      "learning_rate": 0.00011108381999010111,
+      "loss": 2.8999,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4255453969403458,
+      "grad_norm": 0.8731804490089417,
+      "learning_rate": 0.00011105228169903695,
+      "loss": 2.7696,
+      "step": 4701
+    },
+    {
+      "epoch": 0.42563591925409616,
+      "grad_norm": 0.8598802089691162,
+      "learning_rate": 0.0001110207422950083,
+      "loss": 2.8292,
+      "step": 4702
+    },
+    {
+      "epoch": 0.42572644156784645,
+      "grad_norm": 0.8652425408363342,
+      "learning_rate": 0.0001109892017811912,
+      "loss": 2.7647,
+      "step": 4703
+    },
+    {
+      "epoch": 0.4258169638815968,
+      "grad_norm": 0.8402916193008423,
+      "learning_rate": 0.0001109576601607618,
+      "loss": 2.8358,
+      "step": 4704
+    },
+    {
+      "epoch": 0.42590748619534713,
+      "grad_norm": 0.843143105506897,
+      "learning_rate": 0.00011092611743689631,
+      "loss": 2.7721,
+      "step": 4705
+    },
+    {
+      "epoch": 0.42599800850909747,
+      "grad_norm": 0.8784768581390381,
+      "learning_rate": 0.00011089457361277113,
+      "loss": 2.8367,
+      "step": 4706
+    },
+    {
+      "epoch": 0.4260885308228478,
+      "grad_norm": 0.9330942034721375,
+      "learning_rate": 0.00011086302869156265,
+      "loss": 2.8551,
+      "step": 4707
+    },
+    {
+      "epoch": 0.42617905313659815,
+      "grad_norm": 0.8752138614654541,
+      "learning_rate": 0.00011083148267644747,
+      "loss": 2.876,
+      "step": 4708
+    },
+    {
+      "epoch": 0.4262695754503485,
+      "grad_norm": 0.8569071292877197,
+      "learning_rate": 0.0001107999355706023,
+      "loss": 2.7731,
+      "step": 4709
+    },
+    {
+      "epoch": 0.42636009776409883,
+      "grad_norm": 0.9110112190246582,
+      "learning_rate": 0.00011076838737720392,
+      "loss": 2.8479,
+      "step": 4710
+    },
+    {
+      "epoch": 0.4264506200778492,
+      "grad_norm": 0.8108072280883789,
+      "learning_rate": 0.0001107368380994292,
+      "loss": 2.8059,
+      "step": 4711
+    },
+    {
+      "epoch": 0.4265411423915995,
+      "grad_norm": 0.8465254902839661,
+      "learning_rate": 0.00011070528774045519,
+      "loss": 2.86,
+      "step": 4712
+    },
+    {
+      "epoch": 0.42663166470534986,
+      "grad_norm": 0.8941924571990967,
+      "learning_rate": 0.00011067373630345899,
+      "loss": 2.8472,
+      "step": 4713
+    },
+    {
+      "epoch": 0.4267221870191002,
+      "grad_norm": 0.8601610064506531,
+      "learning_rate": 0.00011064218379161786,
+      "loss": 2.8781,
+      "step": 4714
+    },
+    {
+      "epoch": 0.42681270933285054,
+      "grad_norm": 0.9912343621253967,
+      "learning_rate": 0.00011061063020810909,
+      "loss": 2.8762,
+      "step": 4715
+    },
+    {
+      "epoch": 0.4269032316466009,
+      "grad_norm": 0.81369948387146,
+      "learning_rate": 0.00011057907555611016,
+      "loss": 2.8324,
+      "step": 4716
+    },
+    {
+      "epoch": 0.4269937539603512,
+      "grad_norm": 0.8867669701576233,
+      "learning_rate": 0.00011054751983879859,
+      "loss": 2.8066,
+      "step": 4717
+    },
+    {
+      "epoch": 0.42708427627410156,
+      "grad_norm": 0.8661606907844543,
+      "learning_rate": 0.0001105159630593521,
+      "loss": 2.8492,
+      "step": 4718
+    },
+    {
+      "epoch": 0.4271747985878519,
+      "grad_norm": 0.9250156283378601,
+      "learning_rate": 0.00011048440522094838,
+      "loss": 2.8745,
+      "step": 4719
+    },
+    {
+      "epoch": 0.42726532090160224,
+      "grad_norm": 0.9255236983299255,
+      "learning_rate": 0.00011045284632676536,
+      "loss": 2.855,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4273558432153526,
+      "grad_norm": 0.9331503510475159,
+      "learning_rate": 0.00011042128637998099,
+      "loss": 2.8058,
+      "step": 4721
+    },
+    {
+      "epoch": 0.4274463655291029,
+      "grad_norm": 0.8795275092124939,
+      "learning_rate": 0.0001103897253837734,
+      "loss": 2.814,
+      "step": 4722
+    },
+    {
+      "epoch": 0.42753688784285326,
+      "grad_norm": 0.9849618077278137,
+      "learning_rate": 0.0001103581633413207,
+      "loss": 2.8216,
+      "step": 4723
+    },
+    {
+      "epoch": 0.4276274101566036,
+      "grad_norm": 0.8891952037811279,
+      "learning_rate": 0.00011032660025580125,
+      "loss": 2.8081,
+      "step": 4724
+    },
+    {
+      "epoch": 0.42771793247035395,
+      "grad_norm": 0.9136685729026794,
+      "learning_rate": 0.00011029503613039346,
+      "loss": 2.9267,
+      "step": 4725
+    },
+    {
+      "epoch": 0.4278084547841043,
+      "grad_norm": 0.8363056778907776,
+      "learning_rate": 0.00011026347096827578,
+      "loss": 2.8177,
+      "step": 4726
+    },
+    {
+      "epoch": 0.4278989770978546,
+      "grad_norm": 0.9110442399978638,
+      "learning_rate": 0.00011023190477262688,
+      "loss": 2.7913,
+      "step": 4727
+    },
+    {
+      "epoch": 0.42798949941160497,
+      "grad_norm": 0.7873501777648926,
+      "learning_rate": 0.00011020033754662543,
+      "loss": 2.7982,
+      "step": 4728
+    },
+    {
+      "epoch": 0.4280800217253553,
+      "grad_norm": 0.8280324339866638,
+      "learning_rate": 0.0001101687692934503,
+      "loss": 2.8528,
+      "step": 4729
+    },
+    {
+      "epoch": 0.42817054403910565,
+      "grad_norm": 0.9280153512954712,
+      "learning_rate": 0.00011013720001628035,
+      "loss": 2.881,
+      "step": 4730
+    },
+    {
+      "epoch": 0.428261066352856,
+      "grad_norm": 0.8496368527412415,
+      "learning_rate": 0.00011010562971829463,
+      "loss": 2.8228,
+      "step": 4731
+    },
+    {
+      "epoch": 0.42835158866660633,
+      "grad_norm": 0.8465971350669861,
+      "learning_rate": 0.00011007405840267228,
+      "loss": 2.8523,
+      "step": 4732
+    },
+    {
+      "epoch": 0.42844211098035667,
+      "grad_norm": 1.1676387786865234,
+      "learning_rate": 0.00011004248607259255,
+      "loss": 2.8543,
+      "step": 4733
+    },
+    {
+      "epoch": 0.428532633294107,
+      "grad_norm": 0.855336606502533,
+      "learning_rate": 0.00011001091273123473,
+      "loss": 2.8102,
+      "step": 4734
+    },
+    {
+      "epoch": 0.42862315560785735,
+      "grad_norm": 0.8129889965057373,
+      "learning_rate": 0.00010997933838177827,
+      "loss": 2.8427,
+      "step": 4735
+    },
+    {
+      "epoch": 0.4287136779216077,
+      "grad_norm": 0.8779901266098022,
+      "learning_rate": 0.0001099477630274027,
+      "loss": 2.8498,
+      "step": 4736
+    },
+    {
+      "epoch": 0.42880420023535804,
+      "grad_norm": 0.8442177176475525,
+      "learning_rate": 0.00010991618667128769,
+      "loss": 2.8006,
+      "step": 4737
+    },
+    {
+      "epoch": 0.4288947225491084,
+      "grad_norm": 0.8123629093170166,
+      "learning_rate": 0.00010988460931661295,
+      "loss": 2.7864,
+      "step": 4738
+    },
+    {
+      "epoch": 0.4289852448628587,
+      "grad_norm": 0.8855149149894714,
+      "learning_rate": 0.0001098530309665583,
+      "loss": 2.8527,
+      "step": 4739
+    },
+    {
+      "epoch": 0.42907576717660906,
+      "grad_norm": 0.9084986448287964,
+      "learning_rate": 0.00010982145162430373,
+      "loss": 2.8764,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4291662894903594,
+      "grad_norm": 0.8852027654647827,
+      "learning_rate": 0.00010978987129302922,
+      "loss": 2.8407,
+      "step": 4741
+    },
+    {
+      "epoch": 0.42925681180410974,
+      "grad_norm": 0.8793681263923645,
+      "learning_rate": 0.00010975828997591495,
+      "loss": 2.8483,
+      "step": 4742
+    },
+    {
+      "epoch": 0.4293473341178601,
+      "grad_norm": 0.9163318276405334,
+      "learning_rate": 0.00010972670767614114,
+      "loss": 2.8332,
+      "step": 4743
+    },
+    {
+      "epoch": 0.42943785643161037,
+      "grad_norm": 0.8697690963745117,
+      "learning_rate": 0.00010969512439688816,
+      "loss": 2.8413,
+      "step": 4744
+    },
+    {
+      "epoch": 0.4295283787453607,
+      "grad_norm": 0.8311997056007385,
+      "learning_rate": 0.0001096635401413364,
+      "loss": 2.7779,
+      "step": 4745
+    },
+    {
+      "epoch": 0.42961890105911105,
+      "grad_norm": 0.891202449798584,
+      "learning_rate": 0.0001096319549126664,
+      "loss": 2.8442,
+      "step": 4746
+    },
+    {
+      "epoch": 0.4297094233728614,
+      "grad_norm": 0.9648249745368958,
+      "learning_rate": 0.0001096003687140588,
+      "loss": 2.8438,
+      "step": 4747
+    },
+    {
+      "epoch": 0.42979994568661173,
+      "grad_norm": 0.8829810619354248,
+      "learning_rate": 0.00010956878154869437,
+      "loss": 2.8257,
+      "step": 4748
+    },
+    {
+      "epoch": 0.42989046800036207,
+      "grad_norm": 0.9004414677619934,
+      "learning_rate": 0.00010953719341975388,
+      "loss": 2.7777,
+      "step": 4749
+    },
+    {
+      "epoch": 0.4299809903141124,
+      "grad_norm": 0.892963171005249,
+      "learning_rate": 0.00010950560433041826,
+      "loss": 2.8466,
+      "step": 4750
+    },
+    {
+      "epoch": 0.43007151262786275,
+      "grad_norm": 1.0104292631149292,
+      "learning_rate": 0.00010947401428386856,
+      "loss": 2.8534,
+      "step": 4751
+    },
+    {
+      "epoch": 0.4301620349416131,
+      "grad_norm": 0.90214604139328,
+      "learning_rate": 0.0001094424232832859,
+      "loss": 2.8257,
+      "step": 4752
+    },
+    {
+      "epoch": 0.43025255725536343,
+      "grad_norm": 0.9649103283882141,
+      "learning_rate": 0.00010941083133185146,
+      "loss": 2.867,
+      "step": 4753
+    },
+    {
+      "epoch": 0.4303430795691138,
+      "grad_norm": 0.9712350964546204,
+      "learning_rate": 0.00010937923843274654,
+      "loss": 2.8453,
+      "step": 4754
+    },
+    {
+      "epoch": 0.4304336018828641,
+      "grad_norm": 1.0164681673049927,
+      "learning_rate": 0.00010934764458915258,
+      "loss": 2.8726,
+      "step": 4755
+    },
+    {
+      "epoch": 0.43052412419661445,
+      "grad_norm": 0.9008257389068604,
+      "learning_rate": 0.00010931604980425108,
+      "loss": 2.7884,
+      "step": 4756
+    },
+    {
+      "epoch": 0.4306146465103648,
+      "grad_norm": 0.9697628021240234,
+      "learning_rate": 0.00010928445408122361,
+      "loss": 2.7969,
+      "step": 4757
+    },
+    {
+      "epoch": 0.43070516882411514,
+      "grad_norm": 1.0028095245361328,
+      "learning_rate": 0.00010925285742325184,
+      "loss": 2.8293,
+      "step": 4758
+    },
+    {
+      "epoch": 0.4307956911378655,
+      "grad_norm": 0.8799826502799988,
+      "learning_rate": 0.0001092212598335176,
+      "loss": 2.8643,
+      "step": 4759
+    },
+    {
+      "epoch": 0.4308862134516158,
+      "grad_norm": 0.8830612897872925,
+      "learning_rate": 0.00010918966131520277,
+      "loss": 2.7758,
+      "step": 4760
+    },
+    {
+      "epoch": 0.43097673576536616,
+      "grad_norm": 0.9633815288543701,
+      "learning_rate": 0.00010915806187148927,
+      "loss": 2.8098,
+      "step": 4761
+    },
+    {
+      "epoch": 0.4310672580791165,
+      "grad_norm": 0.9045613408088684,
+      "learning_rate": 0.00010912646150555919,
+      "loss": 2.8306,
+      "step": 4762
+    },
+    {
+      "epoch": 0.43115778039286684,
+      "grad_norm": 0.7916074395179749,
+      "learning_rate": 0.0001090948602205947,
+      "loss": 2.8245,
+      "step": 4763
+    },
+    {
+      "epoch": 0.4312483027066172,
+      "grad_norm": 0.8476104736328125,
+      "learning_rate": 0.00010906325801977804,
+      "loss": 2.7464,
+      "step": 4764
+    },
+    {
+      "epoch": 0.4313388250203675,
+      "grad_norm": 0.9538005590438843,
+      "learning_rate": 0.00010903165490629153,
+      "loss": 2.8615,
+      "step": 4765
+    },
+    {
+      "epoch": 0.43142934733411786,
+      "grad_norm": 0.8170968294143677,
+      "learning_rate": 0.00010900005088331761,
+      "loss": 2.7865,
+      "step": 4766
+    },
+    {
+      "epoch": 0.4315198696478682,
+      "grad_norm": 0.8634709715843201,
+      "learning_rate": 0.00010896844595403881,
+      "loss": 2.8317,
+      "step": 4767
+    },
+    {
+      "epoch": 0.43161039196161854,
+      "grad_norm": 0.9003289341926575,
+      "learning_rate": 0.00010893684012163779,
+      "loss": 2.7993,
+      "step": 4768
+    },
+    {
+      "epoch": 0.4317009142753689,
+      "grad_norm": 0.8575468063354492,
+      "learning_rate": 0.00010890523338929719,
+      "loss": 2.7796,
+      "step": 4769
+    },
+    {
+      "epoch": 0.4317914365891192,
+      "grad_norm": 0.7894264459609985,
+      "learning_rate": 0.00010887362576019981,
+      "loss": 2.7963,
+      "step": 4770
+    },
+    {
+      "epoch": 0.43188195890286957,
+      "grad_norm": 0.9664896130561829,
+      "learning_rate": 0.00010884201723752861,
+      "loss": 2.7932,
+      "step": 4771
+    },
+    {
+      "epoch": 0.4319724812166199,
+      "grad_norm": 0.9773446917533875,
+      "learning_rate": 0.00010881040782446645,
+      "loss": 2.8373,
+      "step": 4772
+    },
+    {
+      "epoch": 0.43206300353037025,
+      "grad_norm": 1.0005125999450684,
+      "learning_rate": 0.00010877879752419651,
+      "loss": 2.8478,
+      "step": 4773
+    },
+    {
+      "epoch": 0.4321535258441206,
+      "grad_norm": 0.886049747467041,
+      "learning_rate": 0.00010874718633990189,
+      "loss": 2.7862,
+      "step": 4774
+    },
+    {
+      "epoch": 0.43224404815787093,
+      "grad_norm": 0.9218837022781372,
+      "learning_rate": 0.00010871557427476583,
+      "loss": 2.8233,
+      "step": 4775
+    },
+    {
+      "epoch": 0.43233457047162127,
+      "grad_norm": 0.8934755921363831,
+      "learning_rate": 0.0001086839613319717,
+      "loss": 2.8197,
+      "step": 4776
+    },
+    {
+      "epoch": 0.4324250927853716,
+      "grad_norm": 0.9310414791107178,
+      "learning_rate": 0.00010865234751470288,
+      "loss": 2.8478,
+      "step": 4777
+    },
+    {
+      "epoch": 0.43251561509912195,
+      "grad_norm": 0.8048056960105896,
+      "learning_rate": 0.00010862073282614291,
+      "loss": 2.8203,
+      "step": 4778
+    },
+    {
+      "epoch": 0.4326061374128723,
+      "grad_norm": 1.0201817750930786,
+      "learning_rate": 0.00010858911726947541,
+      "loss": 2.852,
+      "step": 4779
+    },
+    {
+      "epoch": 0.43269665972662263,
+      "grad_norm": 0.8759087920188904,
+      "learning_rate": 0.00010855750084788398,
+      "loss": 2.9003,
+      "step": 4780
+    },
+    {
+      "epoch": 0.432787182040373,
+      "grad_norm": 0.83902508020401,
+      "learning_rate": 0.00010852588356455249,
+      "loss": 2.7608,
+      "step": 4781
+    },
+    {
+      "epoch": 0.4328777043541233,
+      "grad_norm": 1.0640679597854614,
+      "learning_rate": 0.00010849426542266474,
+      "loss": 2.8459,
+      "step": 4782
+    },
+    {
+      "epoch": 0.43296822666787366,
+      "grad_norm": 0.893929660320282,
+      "learning_rate": 0.0001084626464254047,
+      "loss": 2.8119,
+      "step": 4783
+    },
+    {
+      "epoch": 0.433058748981624,
+      "grad_norm": 0.8554607629776001,
+      "learning_rate": 0.00010843102657595641,
+      "loss": 2.8471,
+      "step": 4784
+    },
+    {
+      "epoch": 0.4331492712953743,
+      "grad_norm": 0.989777684211731,
+      "learning_rate": 0.00010839940587750395,
+      "loss": 2.8205,
+      "step": 4785
+    },
+    {
+      "epoch": 0.4332397936091246,
+      "grad_norm": 0.926990807056427,
+      "learning_rate": 0.00010836778433323158,
+      "loss": 2.8905,
+      "step": 4786
+    },
+    {
+      "epoch": 0.43333031592287496,
+      "grad_norm": 0.9434003829956055,
+      "learning_rate": 0.00010833616194632353,
+      "loss": 2.8383,
+      "step": 4787
+    },
+    {
+      "epoch": 0.4334208382366253,
+      "grad_norm": 0.9875321388244629,
+      "learning_rate": 0.0001083045387199642,
+      "loss": 2.7727,
+      "step": 4788
+    },
+    {
+      "epoch": 0.43351136055037565,
+      "grad_norm": 0.8352621793746948,
+      "learning_rate": 0.00010827291465733804,
+      "loss": 2.812,
+      "step": 4789
+    },
+    {
+      "epoch": 0.433601882864126,
+      "grad_norm": 0.959829568862915,
+      "learning_rate": 0.00010824128976162964,
+      "loss": 2.899,
+      "step": 4790
+    },
+    {
+      "epoch": 0.43369240517787633,
+      "grad_norm": 0.8951857686042786,
+      "learning_rate": 0.00010820966403602355,
+      "loss": 2.817,
+      "step": 4791
+    },
+    {
+      "epoch": 0.43378292749162667,
+      "grad_norm": 0.8995950222015381,
+      "learning_rate": 0.00010817803748370452,
+      "loss": 2.8097,
+      "step": 4792
+    },
+    {
+      "epoch": 0.433873449805377,
+      "grad_norm": 0.9037355780601501,
+      "learning_rate": 0.00010814641010785733,
+      "loss": 2.8683,
+      "step": 4793
+    },
+    {
+      "epoch": 0.43396397211912735,
+      "grad_norm": 0.9580223560333252,
+      "learning_rate": 0.00010811478191166691,
+      "loss": 2.7849,
+      "step": 4794
+    },
+    {
+      "epoch": 0.4340544944328777,
+      "grad_norm": 0.8089432716369629,
+      "learning_rate": 0.00010808315289831814,
+      "loss": 2.8267,
+      "step": 4795
+    },
+    {
+      "epoch": 0.43414501674662803,
+      "grad_norm": 0.8201754689216614,
+      "learning_rate": 0.00010805152307099609,
+      "loss": 2.8147,
+      "step": 4796
+    },
+    {
+      "epoch": 0.4342355390603784,
+      "grad_norm": 0.9184460043907166,
+      "learning_rate": 0.00010801989243288589,
+      "loss": 2.7641,
+      "step": 4797
+    },
+    {
+      "epoch": 0.4343260613741287,
+      "grad_norm": 0.9155360460281372,
+      "learning_rate": 0.00010798826098717276,
+      "loss": 2.8113,
+      "step": 4798
+    },
+    {
+      "epoch": 0.43441658368787905,
+      "grad_norm": 0.834479808807373,
+      "learning_rate": 0.00010795662873704196,
+      "loss": 2.8262,
+      "step": 4799
+    },
+    {
+      "epoch": 0.4345071060016294,
+      "grad_norm": 0.8895245790481567,
+      "learning_rate": 0.00010792499568567884,
+      "loss": 2.8121,
+      "step": 4800
+    },
+    {
+      "epoch": 0.43459762831537974,
+      "grad_norm": 0.9145119190216064,
+      "learning_rate": 0.00010789336183626892,
+      "loss": 2.8212,
+      "step": 4801
+    },
+    {
+      "epoch": 0.4346881506291301,
+      "grad_norm": 0.9751086235046387,
+      "learning_rate": 0.00010786172719199763,
+      "loss": 2.767,
+      "step": 4802
+    },
+    {
+      "epoch": 0.4347786729428804,
+      "grad_norm": 0.8262462615966797,
+      "learning_rate": 0.00010783009175605065,
+      "loss": 2.7572,
+      "step": 4803
+    },
+    {
+      "epoch": 0.43486919525663076,
+      "grad_norm": 1.0077941417694092,
+      "learning_rate": 0.00010779845553161362,
+      "loss": 2.8179,
+      "step": 4804
+    },
+    {
+      "epoch": 0.4349597175703811,
+      "grad_norm": 0.9216823577880859,
+      "learning_rate": 0.00010776681852187238,
+      "loss": 2.7986,
+      "step": 4805
+    },
+    {
+      "epoch": 0.43505023988413144,
+      "grad_norm": 0.8575187921524048,
+      "learning_rate": 0.0001077351807300127,
+      "loss": 2.8272,
+      "step": 4806
+    },
+    {
+      "epoch": 0.4351407621978818,
+      "grad_norm": 1.0988225936889648,
+      "learning_rate": 0.00010770354215922052,
+      "loss": 2.9174,
+      "step": 4807
+    },
+    {
+      "epoch": 0.4352312845116321,
+      "grad_norm": 0.8845909833908081,
+      "learning_rate": 0.00010767190281268187,
+      "loss": 2.8137,
+      "step": 4808
+    },
+    {
+      "epoch": 0.43532180682538246,
+      "grad_norm": 0.9869620203971863,
+      "learning_rate": 0.00010764026269358282,
+      "loss": 2.8202,
+      "step": 4809
+    },
+    {
+      "epoch": 0.4354123291391328,
+      "grad_norm": 0.9068284034729004,
+      "learning_rate": 0.00010760862180510951,
+      "loss": 2.8021,
+      "step": 4810
+    },
+    {
+      "epoch": 0.43550285145288314,
+      "grad_norm": 1.0142253637313843,
+      "learning_rate": 0.00010757698015044822,
+      "loss": 2.9227,
+      "step": 4811
+    },
+    {
+      "epoch": 0.4355933737666335,
+      "grad_norm": 0.8755457997322083,
+      "learning_rate": 0.0001075453377327852,
+      "loss": 2.8365,
+      "step": 4812
+    },
+    {
+      "epoch": 0.4356838960803838,
+      "grad_norm": 1.2936592102050781,
+      "learning_rate": 0.0001075136945553069,
+      "loss": 2.9145,
+      "step": 4813
+    },
+    {
+      "epoch": 0.43577441839413417,
+      "grad_norm": 0.9290542006492615,
+      "learning_rate": 0.00010748205062119976,
+      "loss": 2.8088,
+      "step": 4814
+    },
+    {
+      "epoch": 0.4358649407078845,
+      "grad_norm": 0.8115638494491577,
+      "learning_rate": 0.00010745040593365032,
+      "loss": 2.7689,
+      "step": 4815
+    },
+    {
+      "epoch": 0.43595546302163485,
+      "grad_norm": 0.93105548620224,
+      "learning_rate": 0.00010741876049584523,
+      "loss": 2.8144,
+      "step": 4816
+    },
+    {
+      "epoch": 0.4360459853353852,
+      "grad_norm": 0.9035531282424927,
+      "learning_rate": 0.00010738711431097112,
+      "loss": 2.825,
+      "step": 4817
+    },
+    {
+      "epoch": 0.43613650764913553,
+      "grad_norm": 0.8488040566444397,
+      "learning_rate": 0.00010735546738221483,
+      "loss": 2.8077,
+      "step": 4818
+    },
+    {
+      "epoch": 0.43622702996288587,
+      "grad_norm": 0.9691907167434692,
+      "learning_rate": 0.00010732381971276318,
+      "loss": 2.8493,
+      "step": 4819
+    },
+    {
+      "epoch": 0.4363175522766362,
+      "grad_norm": 0.7879188060760498,
+      "learning_rate": 0.0001072921713058031,
+      "loss": 2.8162,
+      "step": 4820
+    },
+    {
+      "epoch": 0.43640807459038655,
+      "grad_norm": 1.0249758958816528,
+      "learning_rate": 0.00010726052216452151,
+      "loss": 2.8505,
+      "step": 4821
+    },
+    {
+      "epoch": 0.4364985969041369,
+      "grad_norm": 0.9572840332984924,
+      "learning_rate": 0.00010722887229210557,
+      "loss": 2.7912,
+      "step": 4822
+    },
+    {
+      "epoch": 0.43658911921788723,
+      "grad_norm": 0.8638061285018921,
+      "learning_rate": 0.00010719722169174238,
+      "loss": 2.7941,
+      "step": 4823
+    },
+    {
+      "epoch": 0.4366796415316376,
+      "grad_norm": 0.8792514801025391,
+      "learning_rate": 0.00010716557036661918,
+      "loss": 2.79,
+      "step": 4824
+    },
+    {
+      "epoch": 0.4367701638453879,
+      "grad_norm": 0.8902309536933899,
+      "learning_rate": 0.00010713391831992323,
+      "loss": 2.8306,
+      "step": 4825
+    },
+    {
+      "epoch": 0.4368606861591382,
+      "grad_norm": 1.0273114442825317,
+      "learning_rate": 0.0001071022655548419,
+      "loss": 2.8273,
+      "step": 4826
+    },
+    {
+      "epoch": 0.43695120847288854,
+      "grad_norm": 0.8674315214157104,
+      "learning_rate": 0.00010707061207456264,
+      "loss": 2.8312,
+      "step": 4827
+    },
+    {
+      "epoch": 0.4370417307866389,
+      "grad_norm": 0.846809446811676,
+      "learning_rate": 0.00010703895788227292,
+      "loss": 2.829,
+      "step": 4828
+    },
+    {
+      "epoch": 0.4371322531003892,
+      "grad_norm": 0.8628326654434204,
+      "learning_rate": 0.00010700730298116033,
+      "loss": 2.7867,
+      "step": 4829
+    },
+    {
+      "epoch": 0.43722277541413956,
+      "grad_norm": 0.8731932640075684,
+      "learning_rate": 0.00010697564737441252,
+      "loss": 2.7932,
+      "step": 4830
+    },
+    {
+      "epoch": 0.4373132977278899,
+      "grad_norm": 0.8147916197776794,
+      "learning_rate": 0.00010694399106521722,
+      "loss": 2.8206,
+      "step": 4831
+    },
+    {
+      "epoch": 0.43740382004164025,
+      "grad_norm": 0.861187219619751,
+      "learning_rate": 0.00010691233405676221,
+      "loss": 2.8254,
+      "step": 4832
+    },
+    {
+      "epoch": 0.4374943423553906,
+      "grad_norm": 0.8892345428466797,
+      "learning_rate": 0.00010688067635223536,
+      "loss": 2.819,
+      "step": 4833
+    },
+    {
+      "epoch": 0.4375848646691409,
+      "grad_norm": 0.8232401609420776,
+      "learning_rate": 0.00010684901795482456,
+      "loss": 2.8206,
+      "step": 4834
+    },
+    {
+      "epoch": 0.43767538698289127,
+      "grad_norm": 0.899243950843811,
+      "learning_rate": 0.00010681735886771788,
+      "loss": 2.8571,
+      "step": 4835
+    },
+    {
+      "epoch": 0.4377659092966416,
+      "grad_norm": 0.8655483722686768,
+      "learning_rate": 0.00010678569909410331,
+      "loss": 2.779,
+      "step": 4836
+    },
+    {
+      "epoch": 0.43785643161039195,
+      "grad_norm": 0.8032087683677673,
+      "learning_rate": 0.00010675403863716907,
+      "loss": 2.7326,
+      "step": 4837
+    },
+    {
+      "epoch": 0.4379469539241423,
+      "grad_norm": 0.8202515840530396,
+      "learning_rate": 0.00010672237750010329,
+      "loss": 2.7479,
+      "step": 4838
+    },
+    {
+      "epoch": 0.43803747623789263,
+      "grad_norm": 0.8518462181091309,
+      "learning_rate": 0.00010669071568609427,
+      "loss": 2.8267,
+      "step": 4839
+    },
+    {
+      "epoch": 0.43812799855164297,
+      "grad_norm": 0.8580194711685181,
+      "learning_rate": 0.00010665905319833041,
+      "loss": 2.7765,
+      "step": 4840
+    },
+    {
+      "epoch": 0.4382185208653933,
+      "grad_norm": 0.8346232771873474,
+      "learning_rate": 0.00010662739004000005,
+      "loss": 2.8643,
+      "step": 4841
+    },
+    {
+      "epoch": 0.43830904317914365,
+      "grad_norm": 0.8472459316253662,
+      "learning_rate": 0.00010659572621429166,
+      "loss": 2.7977,
+      "step": 4842
+    },
+    {
+      "epoch": 0.438399565492894,
+      "grad_norm": 0.8689875602722168,
+      "learning_rate": 0.00010656406172439386,
+      "loss": 2.7833,
+      "step": 4843
+    },
+    {
+      "epoch": 0.43849008780664434,
+      "grad_norm": 0.8687725067138672,
+      "learning_rate": 0.00010653239657349524,
+      "loss": 2.8504,
+      "step": 4844
+    },
+    {
+      "epoch": 0.4385806101203947,
+      "grad_norm": 0.8304334282875061,
+      "learning_rate": 0.00010650073076478442,
+      "loss": 2.7971,
+      "step": 4845
+    },
+    {
+      "epoch": 0.438671132434145,
+      "grad_norm": 0.8737621307373047,
+      "learning_rate": 0.00010646906430145018,
+      "loss": 2.8455,
+      "step": 4846
+    },
+    {
+      "epoch": 0.43876165474789536,
+      "grad_norm": 0.7990383505821228,
+      "learning_rate": 0.0001064373971866814,
+      "loss": 2.7958,
+      "step": 4847
+    },
+    {
+      "epoch": 0.4388521770616457,
+      "grad_norm": 0.8296318650245667,
+      "learning_rate": 0.00010640572942366686,
+      "loss": 2.8039,
+      "step": 4848
+    },
+    {
+      "epoch": 0.43894269937539604,
+      "grad_norm": 0.8771649599075317,
+      "learning_rate": 0.00010637406101559555,
+      "loss": 2.8536,
+      "step": 4849
+    },
+    {
+      "epoch": 0.4390332216891464,
+      "grad_norm": 0.885762095451355,
+      "learning_rate": 0.00010634239196565646,
+      "loss": 2.823,
+      "step": 4850
+    },
+    {
+      "epoch": 0.4391237440028967,
+      "grad_norm": 0.7665767073631287,
+      "learning_rate": 0.0001063107222770387,
+      "loss": 2.7319,
+      "step": 4851
+    },
+    {
+      "epoch": 0.43921426631664706,
+      "grad_norm": 0.8268359303474426,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 2.8446,
+      "step": 4852
+    },
+    {
+      "epoch": 0.4393047886303974,
+      "grad_norm": 0.9912901520729065,
+      "learning_rate": 0.00010624738099652365,
+      "loss": 2.857,
+      "step": 4853
+    },
+    {
+      "epoch": 0.43939531094414774,
+      "grad_norm": 0.8394913673400879,
+      "learning_rate": 0.00010621570941100485,
+      "loss": 2.7842,
+      "step": 4854
+    },
+    {
+      "epoch": 0.4394858332578981,
+      "grad_norm": 0.9693658351898193,
+      "learning_rate": 0.00010618403719956431,
+      "loss": 2.8659,
+      "step": 4855
+    },
+    {
+      "epoch": 0.4395763555716484,
+      "grad_norm": 0.8589224815368652,
+      "learning_rate": 0.00010615236436539137,
+      "loss": 2.8498,
+      "step": 4856
+    },
+    {
+      "epoch": 0.43966687788539877,
+      "grad_norm": 0.8704410195350647,
+      "learning_rate": 0.00010612069091167551,
+      "loss": 2.8244,
+      "step": 4857
+    },
+    {
+      "epoch": 0.4397574001991491,
+      "grad_norm": 0.9131534099578857,
+      "learning_rate": 0.00010608901684160624,
+      "loss": 2.7766,
+      "step": 4858
+    },
+    {
+      "epoch": 0.43984792251289945,
+      "grad_norm": 0.9123629927635193,
+      "learning_rate": 0.00010605734215837316,
+      "loss": 2.756,
+      "step": 4859
+    },
+    {
+      "epoch": 0.4399384448266498,
+      "grad_norm": 0.8991694450378418,
+      "learning_rate": 0.00010602566686516586,
+      "loss": 2.8209,
+      "step": 4860
+    },
+    {
+      "epoch": 0.44002896714040013,
+      "grad_norm": 1.0187561511993408,
+      "learning_rate": 0.00010599399096517408,
+      "loss": 2.8028,
+      "step": 4861
+    },
+    {
+      "epoch": 0.44011948945415047,
+      "grad_norm": 0.87259441614151,
+      "learning_rate": 0.0001059623144615876,
+      "loss": 2.9245,
+      "step": 4862
+    },
+    {
+      "epoch": 0.4402100117679008,
+      "grad_norm": 0.8925728797912598,
+      "learning_rate": 0.00010593063735759618,
+      "loss": 2.7712,
+      "step": 4863
+    },
+    {
+      "epoch": 0.44030053408165115,
+      "grad_norm": 1.0046817064285278,
+      "learning_rate": 0.00010589895965638976,
+      "loss": 2.807,
+      "step": 4864
+    },
+    {
+      "epoch": 0.4403910563954015,
+      "grad_norm": 0.8938895463943481,
+      "learning_rate": 0.00010586728136115824,
+      "loss": 2.7988,
+      "step": 4865
+    },
+    {
+      "epoch": 0.44048157870915183,
+      "grad_norm": 0.9129030108451843,
+      "learning_rate": 0.0001058356024750917,
+      "loss": 2.8089,
+      "step": 4866
+    },
+    {
+      "epoch": 0.4405721010229021,
+      "grad_norm": 1.0655113458633423,
+      "learning_rate": 0.00010580392300138011,
+      "loss": 2.8057,
+      "step": 4867
+    },
+    {
+      "epoch": 0.44066262333665246,
+      "grad_norm": 0.9552054405212402,
+      "learning_rate": 0.00010577224294321362,
+      "loss": 2.7412,
+      "step": 4868
+    },
+    {
+      "epoch": 0.4407531456504028,
+      "grad_norm": 0.8570512533187866,
+      "learning_rate": 0.00010574056230378245,
+      "loss": 2.7769,
+      "step": 4869
+    },
+    {
+      "epoch": 0.44084366796415314,
+      "grad_norm": 1.06633722782135,
+      "learning_rate": 0.00010570888108627681,
+      "loss": 2.8406,
+      "step": 4870
+    },
+    {
+      "epoch": 0.4409341902779035,
+      "grad_norm": 0.9155492782592773,
+      "learning_rate": 0.000105677199293887,
+      "loss": 2.8326,
+      "step": 4871
+    },
+    {
+      "epoch": 0.4410247125916538,
+      "grad_norm": 0.8963524699211121,
+      "learning_rate": 0.00010564551692980337,
+      "loss": 2.8356,
+      "step": 4872
+    },
+    {
+      "epoch": 0.44111523490540416,
+      "grad_norm": 0.9572543501853943,
+      "learning_rate": 0.00010561383399721633,
+      "loss": 2.787,
+      "step": 4873
+    },
+    {
+      "epoch": 0.4412057572191545,
+      "grad_norm": 0.8590235114097595,
+      "learning_rate": 0.00010558215049931638,
+      "loss": 2.7589,
+      "step": 4874
+    },
+    {
+      "epoch": 0.44129627953290484,
+      "grad_norm": 0.8373232483863831,
+      "learning_rate": 0.00010555046643929403,
+      "loss": 2.766,
+      "step": 4875
+    },
+    {
+      "epoch": 0.4413868018466552,
+      "grad_norm": 0.82489013671875,
+      "learning_rate": 0.00010551878182033985,
+      "loss": 2.7575,
+      "step": 4876
+    },
+    {
+      "epoch": 0.4414773241604055,
+      "grad_norm": 0.9751967787742615,
+      "learning_rate": 0.00010548709664564449,
+      "loss": 2.8225,
+      "step": 4877
+    },
+    {
+      "epoch": 0.44156784647415587,
+      "grad_norm": 0.8528650999069214,
+      "learning_rate": 0.00010545541091839867,
+      "loss": 2.7976,
+      "step": 4878
+    },
+    {
+      "epoch": 0.4416583687879062,
+      "grad_norm": 0.9051427841186523,
+      "learning_rate": 0.00010542372464179309,
+      "loss": 2.8601,
+      "step": 4879
+    },
+    {
+      "epoch": 0.44174889110165655,
+      "grad_norm": 0.8459887504577637,
+      "learning_rate": 0.00010539203781901861,
+      "loss": 2.8074,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4418394134154069,
+      "grad_norm": 0.7828327417373657,
+      "learning_rate": 0.00010536035045326606,
+      "loss": 2.8364,
+      "step": 4881
+    },
+    {
+      "epoch": 0.44192993572915723,
+      "grad_norm": 0.881540060043335,
+      "learning_rate": 0.00010532866254772638,
+      "loss": 2.8323,
+      "step": 4882
+    },
+    {
+      "epoch": 0.44202045804290757,
+      "grad_norm": 0.8367167711257935,
+      "learning_rate": 0.00010529697410559049,
+      "loss": 2.7518,
+      "step": 4883
+    },
+    {
+      "epoch": 0.4421109803566579,
+      "grad_norm": 0.7621573805809021,
+      "learning_rate": 0.00010526528513004947,
+      "loss": 2.795,
+      "step": 4884
+    },
+    {
+      "epoch": 0.44220150267040825,
+      "grad_norm": 0.8117220997810364,
+      "learning_rate": 0.0001052335956242944,
+      "loss": 2.8407,
+      "step": 4885
+    },
+    {
+      "epoch": 0.4422920249841586,
+      "grad_norm": 0.7706013321876526,
+      "learning_rate": 0.00010520190559151637,
+      "loss": 2.8473,
+      "step": 4886
+    },
+    {
+      "epoch": 0.44238254729790893,
+      "grad_norm": 0.8366645574569702,
+      "learning_rate": 0.0001051702150349066,
+      "loss": 2.822,
+      "step": 4887
+    },
+    {
+      "epoch": 0.4424730696116593,
+      "grad_norm": 0.8487375378608704,
+      "learning_rate": 0.00010513852395765631,
+      "loss": 2.8013,
+      "step": 4888
+    },
+    {
+      "epoch": 0.4425635919254096,
+      "grad_norm": 0.7707353234291077,
+      "learning_rate": 0.00010510683236295682,
+      "loss": 2.8227,
+      "step": 4889
+    },
+    {
+      "epoch": 0.44265411423915996,
+      "grad_norm": 0.850206196308136,
+      "learning_rate": 0.00010507514025399943,
+      "loss": 2.7895,
+      "step": 4890
+    },
+    {
+      "epoch": 0.4427446365529103,
+      "grad_norm": 0.9814115166664124,
+      "learning_rate": 0.00010504344763397556,
+      "loss": 2.8261,
+      "step": 4891
+    },
+    {
+      "epoch": 0.44283515886666064,
+      "grad_norm": 0.9729773998260498,
+      "learning_rate": 0.00010501175450607664,
+      "loss": 2.8185,
+      "step": 4892
+    },
+    {
+      "epoch": 0.442925681180411,
+      "grad_norm": 1.0523028373718262,
+      "learning_rate": 0.00010498006087349421,
+      "loss": 2.8141,
+      "step": 4893
+    },
+    {
+      "epoch": 0.4430162034941613,
+      "grad_norm": 0.9811918139457703,
+      "learning_rate": 0.00010494836673941977,
+      "loss": 2.8219,
+      "step": 4894
+    },
+    {
+      "epoch": 0.44310672580791166,
+      "grad_norm": 0.8823708891868591,
+      "learning_rate": 0.00010491667210704493,
+      "loss": 2.7874,
+      "step": 4895
+    },
+    {
+      "epoch": 0.443197248121662,
+      "grad_norm": 0.9686410427093506,
+      "learning_rate": 0.00010488497697956135,
+      "loss": 2.8533,
+      "step": 4896
+    },
+    {
+      "epoch": 0.44328777043541234,
+      "grad_norm": 0.9977758526802063,
+      "learning_rate": 0.00010485328136016071,
+      "loss": 2.8345,
+      "step": 4897
+    },
+    {
+      "epoch": 0.4433782927491627,
+      "grad_norm": 0.8284408450126648,
+      "learning_rate": 0.00010482158525203476,
+      "loss": 2.7972,
+      "step": 4898
+    },
+    {
+      "epoch": 0.443468815062913,
+      "grad_norm": 0.9240431785583496,
+      "learning_rate": 0.00010478988865837531,
+      "loss": 2.7735,
+      "step": 4899
+    },
+    {
+      "epoch": 0.44355933737666337,
+      "grad_norm": 0.9728517532348633,
+      "learning_rate": 0.00010475819158237425,
+      "loss": 2.8262,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4436498596904137,
+      "grad_norm": 0.9056679010391235,
+      "learning_rate": 0.00010472649402722339,
+      "loss": 2.8424,
+      "step": 4901
+    },
+    {
+      "epoch": 0.44374038200416405,
+      "grad_norm": 0.9230318665504456,
+      "learning_rate": 0.0001046947959961147,
+      "loss": 2.8233,
+      "step": 4902
+    },
+    {
+      "epoch": 0.4438309043179144,
+      "grad_norm": 0.8907014727592468,
+      "learning_rate": 0.00010466309749224019,
+      "loss": 2.7827,
+      "step": 4903
+    },
+    {
+      "epoch": 0.44392142663166473,
+      "grad_norm": 0.8172028660774231,
+      "learning_rate": 0.00010463139851879191,
+      "loss": 2.8291,
+      "step": 4904
+    },
+    {
+      "epoch": 0.44401194894541507,
+      "grad_norm": 0.9084460139274597,
+      "learning_rate": 0.00010459969907896192,
+      "loss": 2.7821,
+      "step": 4905
+    },
+    {
+      "epoch": 0.4441024712591654,
+      "grad_norm": 0.8679286241531372,
+      "learning_rate": 0.00010456799917594233,
+      "loss": 2.8449,
+      "step": 4906
+    },
+    {
+      "epoch": 0.44419299357291575,
+      "grad_norm": 0.9083068370819092,
+      "learning_rate": 0.00010453629881292538,
+      "loss": 2.7829,
+      "step": 4907
+    },
+    {
+      "epoch": 0.44428351588666604,
+      "grad_norm": 0.877331018447876,
+      "learning_rate": 0.00010450459799310326,
+      "loss": 2.7828,
+      "step": 4908
+    },
+    {
+      "epoch": 0.4443740382004164,
+      "grad_norm": 0.9195173978805542,
+      "learning_rate": 0.00010447289671966824,
+      "loss": 2.7903,
+      "step": 4909
+    },
+    {
+      "epoch": 0.4444645605141667,
+      "grad_norm": 0.8775318264961243,
+      "learning_rate": 0.00010444119499581261,
+      "loss": 2.8409,
+      "step": 4910
+    },
+    {
+      "epoch": 0.44455508282791706,
+      "grad_norm": 0.9003087282180786,
+      "learning_rate": 0.00010440949282472883,
+      "loss": 2.8174,
+      "step": 4911
+    },
+    {
+      "epoch": 0.4446456051416674,
+      "grad_norm": 0.8533322215080261,
+      "learning_rate": 0.00010437779020960921,
+      "loss": 2.7998,
+      "step": 4912
+    },
+    {
+      "epoch": 0.44473612745541774,
+      "grad_norm": 0.8779058456420898,
+      "learning_rate": 0.00010434608715364622,
+      "loss": 2.856,
+      "step": 4913
+    },
+    {
+      "epoch": 0.4448266497691681,
+      "grad_norm": 0.8800071477890015,
+      "learning_rate": 0.00010431438366003239,
+      "loss": 2.8337,
+      "step": 4914
+    },
+    {
+      "epoch": 0.4449171720829184,
+      "grad_norm": 0.875268816947937,
+      "learning_rate": 0.00010428267973196027,
+      "loss": 2.8456,
+      "step": 4915
+    },
+    {
+      "epoch": 0.44500769439666876,
+      "grad_norm": 0.9144801497459412,
+      "learning_rate": 0.00010425097537262236,
+      "loss": 2.8311,
+      "step": 4916
+    },
+    {
+      "epoch": 0.4450982167104191,
+      "grad_norm": 0.9461082220077515,
+      "learning_rate": 0.00010421927058521137,
+      "loss": 2.8462,
+      "step": 4917
+    },
+    {
+      "epoch": 0.44518873902416944,
+      "grad_norm": 0.9008672833442688,
+      "learning_rate": 0.00010418756537291996,
+      "loss": 2.7521,
+      "step": 4918
+    },
+    {
+      "epoch": 0.4452792613379198,
+      "grad_norm": 0.8437387943267822,
+      "learning_rate": 0.00010415585973894084,
+      "loss": 2.7871,
+      "step": 4919
+    },
+    {
+      "epoch": 0.4453697836516701,
+      "grad_norm": 0.8589797616004944,
+      "learning_rate": 0.00010412415368646673,
+      "loss": 2.7782,
+      "step": 4920
+    },
+    {
+      "epoch": 0.44546030596542047,
+      "grad_norm": 0.7951464653015137,
+      "learning_rate": 0.00010409244721869047,
+      "loss": 2.7879,
+      "step": 4921
+    },
+    {
+      "epoch": 0.4455508282791708,
+      "grad_norm": 0.8469287753105164,
+      "learning_rate": 0.00010406074033880491,
+      "loss": 2.7961,
+      "step": 4922
+    },
+    {
+      "epoch": 0.44564135059292115,
+      "grad_norm": 0.83460533618927,
+      "learning_rate": 0.00010402903305000287,
+      "loss": 2.8391,
+      "step": 4923
+    },
+    {
+      "epoch": 0.4457318729066715,
+      "grad_norm": 0.9511914253234863,
+      "learning_rate": 0.00010399732535547734,
+      "loss": 2.8332,
+      "step": 4924
+    },
+    {
+      "epoch": 0.44582239522042183,
+      "grad_norm": 0.8237845301628113,
+      "learning_rate": 0.00010396561725842124,
+      "loss": 2.7943,
+      "step": 4925
+    },
+    {
+      "epoch": 0.44591291753417217,
+      "grad_norm": 0.850771963596344,
+      "learning_rate": 0.0001039339087620276,
+      "loss": 2.8397,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4460034398479225,
+      "grad_norm": 0.7968292236328125,
+      "learning_rate": 0.00010390219986948946,
+      "loss": 2.8291,
+      "step": 4927
+    },
+    {
+      "epoch": 0.44609396216167285,
+      "grad_norm": 0.9098146557807922,
+      "learning_rate": 0.00010387049058399989,
+      "loss": 2.7927,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4461844844754232,
+      "grad_norm": 0.878772497177124,
+      "learning_rate": 0.00010383878090875201,
+      "loss": 2.8495,
+      "step": 4929
+    },
+    {
+      "epoch": 0.44627500678917353,
+      "grad_norm": 0.8625677227973938,
+      "learning_rate": 0.00010380707084693901,
+      "loss": 2.8378,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4463655291029239,
+      "grad_norm": 0.8826537132263184,
+      "learning_rate": 0.00010377536040175408,
+      "loss": 2.8177,
+      "step": 4931
+    },
+    {
+      "epoch": 0.4464560514166742,
+      "grad_norm": 0.865592360496521,
+      "learning_rate": 0.00010374364957639047,
+      "loss": 2.8599,
+      "step": 4932
+    },
+    {
+      "epoch": 0.44654657373042456,
+      "grad_norm": 0.8049623966217041,
+      "learning_rate": 0.00010371193837404143,
+      "loss": 2.7947,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4466370960441749,
+      "grad_norm": 0.868293821811676,
+      "learning_rate": 0.00010368022679790031,
+      "loss": 2.7748,
+      "step": 4934
+    },
+    {
+      "epoch": 0.44672761835792524,
+      "grad_norm": 0.9875622987747192,
+      "learning_rate": 0.00010364851485116048,
+      "loss": 2.8251,
+      "step": 4935
+    },
+    {
+      "epoch": 0.4468181406716756,
+      "grad_norm": 0.8716400265693665,
+      "learning_rate": 0.00010361680253701527,
+      "loss": 2.7795,
+      "step": 4936
+    },
+    {
+      "epoch": 0.4469086629854259,
+      "grad_norm": 0.808600902557373,
+      "learning_rate": 0.00010358508985865813,
+      "loss": 2.8153,
+      "step": 4937
+    },
+    {
+      "epoch": 0.44699918529917626,
+      "grad_norm": 0.8896029591560364,
+      "learning_rate": 0.00010355337681928254,
+      "loss": 2.7848,
+      "step": 4938
+    },
+    {
+      "epoch": 0.4470897076129266,
+      "grad_norm": 0.8998689651489258,
+      "learning_rate": 0.00010352166342208205,
+      "loss": 2.856,
+      "step": 4939
+    },
+    {
+      "epoch": 0.44718022992667694,
+      "grad_norm": 0.8900119066238403,
+      "learning_rate": 0.00010348994967025012,
+      "loss": 2.7832,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4472707522404273,
+      "grad_norm": 0.921204686164856,
+      "learning_rate": 0.00010345823556698034,
+      "loss": 2.8575,
+      "step": 4941
+    },
+    {
+      "epoch": 0.4473612745541776,
+      "grad_norm": 0.8895449042320251,
+      "learning_rate": 0.00010342652111546635,
+      "loss": 2.7814,
+      "step": 4942
+    },
+    {
+      "epoch": 0.44745179686792796,
+      "grad_norm": 0.8308514952659607,
+      "learning_rate": 0.00010339480631890178,
+      "loss": 2.8368,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4475423191816783,
+      "grad_norm": 0.8498168587684631,
+      "learning_rate": 0.00010336309118048029,
+      "loss": 2.8192,
+      "step": 4944
+    },
+    {
+      "epoch": 0.44763284149542865,
+      "grad_norm": 0.8537622690200806,
+      "learning_rate": 0.00010333137570339562,
+      "loss": 2.8891,
+      "step": 4945
+    },
+    {
+      "epoch": 0.447723363809179,
+      "grad_norm": 0.8644798994064331,
+      "learning_rate": 0.00010329965989084152,
+      "loss": 2.8316,
+      "step": 4946
+    },
+    {
+      "epoch": 0.4478138861229293,
+      "grad_norm": 0.8171406388282776,
+      "learning_rate": 0.00010326794374601174,
+      "loss": 2.7997,
+      "step": 4947
+    },
+    {
+      "epoch": 0.44790440843667967,
+      "grad_norm": 0.8642208576202393,
+      "learning_rate": 0.00010323622727210012,
+      "loss": 2.7823,
+      "step": 4948
+    },
+    {
+      "epoch": 0.44799493075042995,
+      "grad_norm": 0.8446072340011597,
+      "learning_rate": 0.00010320451047230051,
+      "loss": 2.8498,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4480854530641803,
+      "grad_norm": 0.8191630840301514,
+      "learning_rate": 0.00010317279334980678,
+      "loss": 2.8523,
+      "step": 4950
+    },
+    {
+      "epoch": 0.44817597537793064,
+      "grad_norm": 0.8264406323432922,
+      "learning_rate": 0.00010314107590781284,
+      "loss": 2.7282,
+      "step": 4951
+    },
+    {
+      "epoch": 0.448266497691681,
+      "grad_norm": 0.7761468887329102,
+      "learning_rate": 0.00010310935814951264,
+      "loss": 2.7908,
+      "step": 4952
+    },
+    {
+      "epoch": 0.4483570200054313,
+      "grad_norm": 0.8397097587585449,
+      "learning_rate": 0.00010307764007810017,
+      "loss": 2.7509,
+      "step": 4953
+    },
+    {
+      "epoch": 0.44844754231918166,
+      "grad_norm": 0.8186659812927246,
+      "learning_rate": 0.00010304592169676943,
+      "loss": 2.7241,
+      "step": 4954
+    },
+    {
+      "epoch": 0.448538064632932,
+      "grad_norm": 0.8386641144752502,
+      "learning_rate": 0.00010301420300871445,
+      "loss": 2.7638,
+      "step": 4955
+    },
+    {
+      "epoch": 0.44862858694668234,
+      "grad_norm": 0.8329310417175293,
+      "learning_rate": 0.00010298248401712931,
+      "loss": 2.8275,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4487191092604327,
+      "grad_norm": 0.8720096349716187,
+      "learning_rate": 0.00010295076472520812,
+      "loss": 2.753,
+      "step": 4957
+    },
+    {
+      "epoch": 0.448809631574183,
+      "grad_norm": 0.8473954200744629,
+      "learning_rate": 0.00010291904513614499,
+      "loss": 2.843,
+      "step": 4958
+    },
+    {
+      "epoch": 0.44890015388793336,
+      "grad_norm": 0.8658212423324585,
+      "learning_rate": 0.0001028873252531341,
+      "loss": 2.7285,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4489906762016837,
+      "grad_norm": 0.8777493238449097,
+      "learning_rate": 0.00010285560507936961,
+      "loss": 2.7865,
+      "step": 4960
+    },
+    {
+      "epoch": 0.44908119851543404,
+      "grad_norm": 0.9246544241905212,
+      "learning_rate": 0.00010282388461804584,
+      "loss": 2.7525,
+      "step": 4961
+    },
+    {
+      "epoch": 0.4491717208291844,
+      "grad_norm": 0.8622031211853027,
+      "learning_rate": 0.0001027921638723569,
+      "loss": 2.9143,
+      "step": 4962
+    },
+    {
+      "epoch": 0.4492622431429347,
+      "grad_norm": 0.9040536880493164,
+      "learning_rate": 0.00010276044284549716,
+      "loss": 2.8462,
+      "step": 4963
+    },
+    {
+      "epoch": 0.44935276545668507,
+      "grad_norm": 0.8143455386161804,
+      "learning_rate": 0.00010272872154066089,
+      "loss": 2.8701,
+      "step": 4964
+    },
+    {
+      "epoch": 0.4494432877704354,
+      "grad_norm": 0.8502079248428345,
+      "learning_rate": 0.00010269699996104246,
+      "loss": 2.8012,
+      "step": 4965
+    },
+    {
+      "epoch": 0.44953381008418575,
+      "grad_norm": 0.9193055629730225,
+      "learning_rate": 0.00010266527810983617,
+      "loss": 2.8005,
+      "step": 4966
+    },
+    {
+      "epoch": 0.4496243323979361,
+      "grad_norm": 0.8001332879066467,
+      "learning_rate": 0.00010263355599023645,
+      "loss": 2.7171,
+      "step": 4967
+    },
+    {
+      "epoch": 0.44971485471168643,
+      "grad_norm": 0.830764889717102,
+      "learning_rate": 0.00010260183360543773,
+      "loss": 2.8556,
+      "step": 4968
+    },
+    {
+      "epoch": 0.44980537702543677,
+      "grad_norm": 0.8240771889686584,
+      "learning_rate": 0.00010257011095863444,
+      "loss": 2.805,
+      "step": 4969
+    },
+    {
+      "epoch": 0.4498958993391871,
+      "grad_norm": 0.8989068269729614,
+      "learning_rate": 0.00010253838805302104,
+      "loss": 2.8138,
+      "step": 4970
+    },
+    {
+      "epoch": 0.44998642165293745,
+      "grad_norm": 0.7964334487915039,
+      "learning_rate": 0.00010250666489179204,
+      "loss": 2.7718,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4500769439666878,
+      "grad_norm": 0.8724195957183838,
+      "learning_rate": 0.00010247494147814196,
+      "loss": 2.809,
+      "step": 4972
+    },
+    {
+      "epoch": 0.45016746628043813,
+      "grad_norm": 0.9154969453811646,
+      "learning_rate": 0.00010244321781526533,
+      "loss": 2.8357,
+      "step": 4973
+    },
+    {
+      "epoch": 0.4502579885941885,
+      "grad_norm": 0.898012101650238,
+      "learning_rate": 0.00010241149390635674,
+      "loss": 2.744,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4503485109079388,
+      "grad_norm": 1.0105397701263428,
+      "learning_rate": 0.00010237976975461075,
+      "loss": 2.8061,
+      "step": 4975
+    },
+    {
+      "epoch": 0.45043903322168916,
+      "grad_norm": 0.8304058313369751,
+      "learning_rate": 0.00010234804536322204,
+      "loss": 2.816,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4505295555354395,
+      "grad_norm": 0.9987416863441467,
+      "learning_rate": 0.00010231632073538522,
+      "loss": 2.8618,
+      "step": 4977
+    },
+    {
+      "epoch": 0.45062007784918984,
+      "grad_norm": 0.8492810726165771,
+      "learning_rate": 0.00010228459587429497,
+      "loss": 2.7893,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4507106001629402,
+      "grad_norm": 0.8335104584693909,
+      "learning_rate": 0.00010225287078314596,
+      "loss": 2.8049,
+      "step": 4979
+    },
+    {
+      "epoch": 0.4508011224766905,
+      "grad_norm": 0.881281852722168,
+      "learning_rate": 0.00010222114546513295,
+      "loss": 2.8618,
+      "step": 4980
+    },
+    {
+      "epoch": 0.45089164479044086,
+      "grad_norm": 0.9159513115882874,
+      "learning_rate": 0.00010218941992345063,
+      "loss": 2.8276,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4509821671041912,
+      "grad_norm": 0.846339762210846,
+      "learning_rate": 0.0001021576941612938,
+      "loss": 2.7684,
+      "step": 4982
+    },
+    {
+      "epoch": 0.45107268941794154,
+      "grad_norm": 0.8389536142349243,
+      "learning_rate": 0.0001021259681818572,
+      "loss": 2.8375,
+      "step": 4983
+    },
+    {
+      "epoch": 0.4511632117316919,
+      "grad_norm": 0.8507980108261108,
+      "learning_rate": 0.0001020942419883357,
+      "loss": 2.8065,
+      "step": 4984
+    },
+    {
+      "epoch": 0.4512537340454422,
+      "grad_norm": 0.8403035998344421,
+      "learning_rate": 0.00010206251558392408,
+      "loss": 2.8151,
+      "step": 4985
+    },
+    {
+      "epoch": 0.45134425635919256,
+      "grad_norm": 0.8249539732933044,
+      "learning_rate": 0.00010203078897181717,
+      "loss": 2.7943,
+      "step": 4986
+    },
+    {
+      "epoch": 0.4514347786729429,
+      "grad_norm": 0.8872669339179993,
+      "learning_rate": 0.00010199906215520989,
+      "loss": 2.8219,
+      "step": 4987
+    },
+    {
+      "epoch": 0.45152530098669325,
+      "grad_norm": 0.9296691417694092,
+      "learning_rate": 0.0001019673351372971,
+      "loss": 2.8051,
+      "step": 4988
+    },
+    {
+      "epoch": 0.4516158233004436,
+      "grad_norm": 0.9113702774047852,
+      "learning_rate": 0.00010193560792127372,
+      "loss": 2.8503,
+      "step": 4989
+    },
+    {
+      "epoch": 0.45170634561419387,
+      "grad_norm": 0.9052483439445496,
+      "learning_rate": 0.00010190388051033466,
+      "loss": 2.8591,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4517968679279442,
+      "grad_norm": 0.844837486743927,
+      "learning_rate": 0.00010187215290767491,
+      "loss": 2.8988,
+      "step": 4991
+    },
+    {
+      "epoch": 0.45188739024169455,
+      "grad_norm": 0.8985702395439148,
+      "learning_rate": 0.00010184042511648942,
+      "loss": 2.8286,
+      "step": 4992
+    },
+    {
+      "epoch": 0.4519779125554449,
+      "grad_norm": 0.9050378203392029,
+      "learning_rate": 0.00010180869713997312,
+      "loss": 2.7684,
+      "step": 4993
+    },
+    {
+      "epoch": 0.45206843486919523,
+      "grad_norm": 0.8572180867195129,
+      "learning_rate": 0.0001017769689813211,
+      "loss": 2.7869,
+      "step": 4994
+    },
+    {
+      "epoch": 0.4521589571829456,
+      "grad_norm": 0.8583084940910339,
+      "learning_rate": 0.00010174524064372837,
+      "loss": 2.839,
+      "step": 4995
+    },
+    {
+      "epoch": 0.4522494794966959,
+      "grad_norm": 0.7894691228866577,
+      "learning_rate": 0.00010171351213038993,
+      "loss": 2.7837,
+      "step": 4996
+    },
+    {
+      "epoch": 0.45234000181044626,
+      "grad_norm": 0.843013346195221,
+      "learning_rate": 0.00010168178344450086,
+      "loss": 2.7827,
+      "step": 4997
+    },
+    {
+      "epoch": 0.4524305241241966,
+      "grad_norm": 0.9390647411346436,
+      "learning_rate": 0.00010165005458925626,
+      "loss": 2.7951,
+      "step": 4998
+    },
+    {
+      "epoch": 0.45252104643794694,
+      "grad_norm": 0.8602046966552734,
+      "learning_rate": 0.0001016183255678512,
+      "loss": 2.77,
+      "step": 4999
+    },
+    {
+      "epoch": 0.4526115687516973,
+      "grad_norm": 0.9842708110809326,
+      "learning_rate": 0.00010158659638348081,
+      "loss": 2.8039,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4526115687516973,
+      "eval_loss": 2.7390329837799072,
+      "eval_runtime": 71.5613,
+      "eval_samples_per_second": 37.772,
+      "eval_steps_per_second": 3.158,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4527020910654476,
+      "grad_norm": 0.8009204864501953,
+      "learning_rate": 0.00010155486703934018,
+      "loss": 2.8257,
+      "step": 5001
+    },
+    {
+      "epoch": 0.45279261337919796,
+      "grad_norm": 0.8216099143028259,
+      "learning_rate": 0.0001015231375386245,
+      "loss": 2.7155,
+      "step": 5002
+    },
+    {
+      "epoch": 0.4528831356929483,
+      "grad_norm": 0.879997193813324,
+      "learning_rate": 0.00010149140788452894,
+      "loss": 2.8294,
+      "step": 5003
+    },
+    {
+      "epoch": 0.45297365800669864,
+      "grad_norm": 0.7909499406814575,
+      "learning_rate": 0.00010145967808024859,
+      "loss": 2.7564,
+      "step": 5004
+    },
+    {
+      "epoch": 0.453064180320449,
+      "grad_norm": 0.8433207869529724,
+      "learning_rate": 0.00010142794812897873,
+      "loss": 2.8164,
+      "step": 5005
+    },
+    {
+      "epoch": 0.4531547026341993,
+      "grad_norm": 0.8329892754554749,
+      "learning_rate": 0.00010139621803391455,
+      "loss": 2.8001,
+      "step": 5006
+    },
+    {
+      "epoch": 0.45324522494794967,
+      "grad_norm": 0.9415286779403687,
+      "learning_rate": 0.0001013644877982512,
+      "loss": 2.8395,
+      "step": 5007
+    },
+    {
+      "epoch": 0.4533357472617,
+      "grad_norm": 0.8327991366386414,
+      "learning_rate": 0.00010133275742518403,
+      "loss": 2.8404,
+      "step": 5008
+    },
+    {
+      "epoch": 0.45342626957545035,
+      "grad_norm": 0.8747456073760986,
+      "learning_rate": 0.0001013010269179082,
+      "loss": 2.8421,
+      "step": 5009
+    },
+    {
+      "epoch": 0.4535167918892007,
+      "grad_norm": 0.8911159038543701,
+      "learning_rate": 0.00010126929627961896,
+      "loss": 2.7713,
+      "step": 5010
+    },
+    {
+      "epoch": 0.45360731420295103,
+      "grad_norm": 0.7843373417854309,
+      "learning_rate": 0.00010123756551351166,
+      "loss": 2.806,
+      "step": 5011
+    },
+    {
+      "epoch": 0.45369783651670137,
+      "grad_norm": 0.8249711990356445,
+      "learning_rate": 0.00010120583462278152,
+      "loss": 2.7439,
+      "step": 5012
+    },
+    {
+      "epoch": 0.4537883588304517,
+      "grad_norm": 0.8340107202529907,
+      "learning_rate": 0.00010117410361062387,
+      "loss": 2.7893,
+      "step": 5013
+    },
+    {
+      "epoch": 0.45387888114420205,
+      "grad_norm": 0.8179864287376404,
+      "learning_rate": 0.00010114237248023404,
+      "loss": 2.7871,
+      "step": 5014
+    },
+    {
+      "epoch": 0.4539694034579524,
+      "grad_norm": 0.8432429432868958,
+      "learning_rate": 0.00010111064123480733,
+      "loss": 2.8294,
+      "step": 5015
+    },
+    {
+      "epoch": 0.45405992577170273,
+      "grad_norm": 0.8422669768333435,
+      "learning_rate": 0.00010107890987753905,
+      "loss": 2.8206,
+      "step": 5016
+    },
+    {
+      "epoch": 0.4541504480854531,
+      "grad_norm": 0.8166409134864807,
+      "learning_rate": 0.00010104717841162458,
+      "loss": 2.81,
+      "step": 5017
+    },
+    {
+      "epoch": 0.4542409703992034,
+      "grad_norm": 0.8468937277793884,
+      "learning_rate": 0.00010101544684025931,
+      "loss": 2.7166,
+      "step": 5018
+    },
+    {
+      "epoch": 0.45433149271295376,
+      "grad_norm": 0.8151309490203857,
+      "learning_rate": 0.00010098371516663853,
+      "loss": 2.8096,
+      "step": 5019
+    },
+    {
+      "epoch": 0.4544220150267041,
+      "grad_norm": 0.8079573512077332,
+      "learning_rate": 0.00010095198339395769,
+      "loss": 2.7564,
+      "step": 5020
+    },
+    {
+      "epoch": 0.45451253734045444,
+      "grad_norm": 0.8220385313034058,
+      "learning_rate": 0.00010092025152541211,
+      "loss": 2.7707,
+      "step": 5021
+    },
+    {
+      "epoch": 0.4546030596542048,
+      "grad_norm": 0.7748722434043884,
+      "learning_rate": 0.00010088851956419728,
+      "loss": 2.7977,
+      "step": 5022
+    },
+    {
+      "epoch": 0.4546935819679551,
+      "grad_norm": 0.8827757835388184,
+      "learning_rate": 0.0001008567875135085,
+      "loss": 2.8557,
+      "step": 5023
+    },
+    {
+      "epoch": 0.45478410428170546,
+      "grad_norm": 0.852816104888916,
+      "learning_rate": 0.00010082505537654128,
+      "loss": 2.8645,
+      "step": 5024
+    },
+    {
+      "epoch": 0.4548746265954558,
+      "grad_norm": 0.845237672328949,
+      "learning_rate": 0.00010079332315649097,
+      "loss": 2.805,
+      "step": 5025
+    },
+    {
+      "epoch": 0.45496514890920614,
+      "grad_norm": 0.7926275134086609,
+      "learning_rate": 0.00010076159085655308,
+      "loss": 2.7563,
+      "step": 5026
+    },
+    {
+      "epoch": 0.4550556712229565,
+      "grad_norm": 0.8070298433303833,
+      "learning_rate": 0.000100729858479923,
+      "loss": 2.8283,
+      "step": 5027
+    },
+    {
+      "epoch": 0.4551461935367068,
+      "grad_norm": 0.8690354824066162,
+      "learning_rate": 0.00010069812602979615,
+      "loss": 2.8512,
+      "step": 5028
+    },
+    {
+      "epoch": 0.45523671585045716,
+      "grad_norm": 0.8458876013755798,
+      "learning_rate": 0.00010066639350936806,
+      "loss": 2.8317,
+      "step": 5029
+    },
+    {
+      "epoch": 0.4553272381642075,
+      "grad_norm": 0.8781901001930237,
+      "learning_rate": 0.0001006346609218342,
+      "loss": 2.7928,
+      "step": 5030
+    },
+    {
+      "epoch": 0.4554177604779578,
+      "grad_norm": 0.8437893986701965,
+      "learning_rate": 0.00010060292827038995,
+      "loss": 2.7485,
+      "step": 5031
+    },
+    {
+      "epoch": 0.45550828279170813,
+      "grad_norm": 0.8205193281173706,
+      "learning_rate": 0.00010057119555823085,
+      "loss": 2.8012,
+      "step": 5032
+    },
+    {
+      "epoch": 0.45559880510545847,
+      "grad_norm": 0.8345540165901184,
+      "learning_rate": 0.00010053946278855237,
+      "loss": 2.7236,
+      "step": 5033
+    },
+    {
+      "epoch": 0.4556893274192088,
+      "grad_norm": 0.8251396417617798,
+      "learning_rate": 0.00010050772996455002,
+      "loss": 2.8524,
+      "step": 5034
+    },
+    {
+      "epoch": 0.45577984973295915,
+      "grad_norm": 0.9152898192405701,
+      "learning_rate": 0.00010047599708941927,
+      "loss": 2.7933,
+      "step": 5035
+    },
+    {
+      "epoch": 0.4558703720467095,
+      "grad_norm": 0.8356764316558838,
+      "learning_rate": 0.00010044426416635562,
+      "loss": 2.7767,
+      "step": 5036
+    },
+    {
+      "epoch": 0.45596089436045983,
+      "grad_norm": 0.8549394607543945,
+      "learning_rate": 0.0001004125311985546,
+      "loss": 2.7789,
+      "step": 5037
+    },
+    {
+      "epoch": 0.4560514166742102,
+      "grad_norm": 0.8408095240592957,
+      "learning_rate": 0.00010038079818921166,
+      "loss": 2.7851,
+      "step": 5038
+    },
+    {
+      "epoch": 0.4561419389879605,
+      "grad_norm": 0.8904051184654236,
+      "learning_rate": 0.00010034906514152238,
+      "loss": 2.7786,
+      "step": 5039
+    },
+    {
+      "epoch": 0.45623246130171086,
+      "grad_norm": 0.8291855454444885,
+      "learning_rate": 0.00010031733205868224,
+      "loss": 2.7598,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4563229836154612,
+      "grad_norm": 0.8298647403717041,
+      "learning_rate": 0.0001002855989438868,
+      "loss": 2.7359,
+      "step": 5041
+    },
+    {
+      "epoch": 0.45641350592921154,
+      "grad_norm": 0.8610008358955383,
+      "learning_rate": 0.00010025386580033151,
+      "loss": 2.7943,
+      "step": 5042
+    },
+    {
+      "epoch": 0.4565040282429619,
+      "grad_norm": 0.8803578615188599,
+      "learning_rate": 0.00010022213263121197,
+      "loss": 2.7389,
+      "step": 5043
+    },
+    {
+      "epoch": 0.4565945505567122,
+      "grad_norm": 0.8574979901313782,
+      "learning_rate": 0.00010019039943972366,
+      "loss": 2.8107,
+      "step": 5044
+    },
+    {
+      "epoch": 0.45668507287046256,
+      "grad_norm": 0.8162574172019958,
+      "learning_rate": 0.00010015866622906217,
+      "loss": 2.843,
+      "step": 5045
+    },
+    {
+      "epoch": 0.4567755951842129,
+      "grad_norm": 0.9479469060897827,
+      "learning_rate": 0.00010012693300242296,
+      "loss": 2.7705,
+      "step": 5046
+    },
+    {
+      "epoch": 0.45686611749796324,
+      "grad_norm": 0.8440346121788025,
+      "learning_rate": 0.00010009519976300159,
+      "loss": 2.8239,
+      "step": 5047
+    },
+    {
+      "epoch": 0.4569566398117136,
+      "grad_norm": 0.8232409358024597,
+      "learning_rate": 0.00010006346651399363,
+      "loss": 2.7954,
+      "step": 5048
+    },
+    {
+      "epoch": 0.4570471621254639,
+      "grad_norm": 0.8686602115631104,
+      "learning_rate": 0.00010003173325859461,
+      "loss": 2.8405,
+      "step": 5049
+    },
+    {
+      "epoch": 0.45713768443921426,
+      "grad_norm": 0.8840144872665405,
+      "learning_rate": 0.0001,
+      "loss": 2.8045,
+      "step": 5050
+    },
+    {
+      "epoch": 0.4572282067529646,
+      "grad_norm": 0.8932952880859375,
+      "learning_rate": 9.996826674140544e-05,
+      "loss": 2.778,
+      "step": 5051
+    },
+    {
+      "epoch": 0.45731872906671495,
+      "grad_norm": 0.874193012714386,
+      "learning_rate": 9.99365334860064e-05,
+      "loss": 2.7877,
+      "step": 5052
+    },
+    {
+      "epoch": 0.4574092513804653,
+      "grad_norm": 0.7909197807312012,
+      "learning_rate": 9.990480023699845e-05,
+      "loss": 2.7235,
+      "step": 5053
+    },
+    {
+      "epoch": 0.45749977369421563,
+      "grad_norm": 0.8901601433753967,
+      "learning_rate": 9.987306699757707e-05,
+      "loss": 2.7525,
+      "step": 5054
+    },
+    {
+      "epoch": 0.45759029600796597,
+      "grad_norm": 0.8555707931518555,
+      "learning_rate": 9.984133377093789e-05,
+      "loss": 2.7974,
+      "step": 5055
+    },
+    {
+      "epoch": 0.4576808183217163,
+      "grad_norm": 0.8725164532661438,
+      "learning_rate": 9.980960056027636e-05,
+      "loss": 2.7639,
+      "step": 5056
+    },
+    {
+      "epoch": 0.45777134063546665,
+      "grad_norm": 0.8749620318412781,
+      "learning_rate": 9.977786736878808e-05,
+      "loss": 2.766,
+      "step": 5057
+    },
+    {
+      "epoch": 0.457861862949217,
+      "grad_norm": 0.8529465794563293,
+      "learning_rate": 9.974613419966851e-05,
+      "loss": 2.7849,
+      "step": 5058
+    },
+    {
+      "epoch": 0.45795238526296733,
+      "grad_norm": 0.8112756609916687,
+      "learning_rate": 9.971440105611325e-05,
+      "loss": 2.8221,
+      "step": 5059
+    },
+    {
+      "epoch": 0.4580429075767177,
+      "grad_norm": 0.8644946217536926,
+      "learning_rate": 9.968266794131777e-05,
+      "loss": 2.8586,
+      "step": 5060
+    },
+    {
+      "epoch": 0.458133429890468,
+      "grad_norm": 0.819271981716156,
+      "learning_rate": 9.965093485847767e-05,
+      "loss": 2.7961,
+      "step": 5061
+    },
+    {
+      "epoch": 0.45822395220421835,
+      "grad_norm": 0.8047432899475098,
+      "learning_rate": 9.961920181078835e-05,
+      "loss": 2.824,
+      "step": 5062
+    },
+    {
+      "epoch": 0.4583144745179687,
+      "grad_norm": 0.9896413087844849,
+      "learning_rate": 9.958746880144545e-05,
+      "loss": 2.887,
+      "step": 5063
+    },
+    {
+      "epoch": 0.45840499683171904,
+      "grad_norm": 0.8702191114425659,
+      "learning_rate": 9.95557358336444e-05,
+      "loss": 2.7741,
+      "step": 5064
+    },
+    {
+      "epoch": 0.4584955191454694,
+      "grad_norm": 0.8362398147583008,
+      "learning_rate": 9.952400291058077e-05,
+      "loss": 2.7465,
+      "step": 5065
+    },
+    {
+      "epoch": 0.4585860414592197,
+      "grad_norm": 0.8557057976722717,
+      "learning_rate": 9.949227003545e-05,
+      "loss": 2.7277,
+      "step": 5066
+    },
+    {
+      "epoch": 0.45867656377297006,
+      "grad_norm": 0.9606072902679443,
+      "learning_rate": 9.946053721144765e-05,
+      "loss": 2.7846,
+      "step": 5067
+    },
+    {
+      "epoch": 0.4587670860867204,
+      "grad_norm": 0.8010448217391968,
+      "learning_rate": 9.942880444176918e-05,
+      "loss": 2.7601,
+      "step": 5068
+    },
+    {
+      "epoch": 0.45885760840047074,
+      "grad_norm": 0.833601176738739,
+      "learning_rate": 9.93970717296101e-05,
+      "loss": 2.8154,
+      "step": 5069
+    },
+    {
+      "epoch": 0.4589481307142211,
+      "grad_norm": 0.8944669961929321,
+      "learning_rate": 9.936533907816584e-05,
+      "loss": 2.7353,
+      "step": 5070
+    },
+    {
+      "epoch": 0.4590386530279714,
+      "grad_norm": 0.9217827916145325,
+      "learning_rate": 9.933360649063195e-05,
+      "loss": 2.8062,
+      "step": 5071
+    },
+    {
+      "epoch": 0.4591291753417217,
+      "grad_norm": 0.7689042091369629,
+      "learning_rate": 9.930187397020386e-05,
+      "loss": 2.6995,
+      "step": 5072
+    },
+    {
+      "epoch": 0.45921969765547205,
+      "grad_norm": 0.8688762187957764,
+      "learning_rate": 9.927014152007706e-05,
+      "loss": 2.7928,
+      "step": 5073
+    },
+    {
+      "epoch": 0.4593102199692224,
+      "grad_norm": 0.8136226534843445,
+      "learning_rate": 9.923840914344695e-05,
+      "loss": 2.786,
+      "step": 5074
+    },
+    {
+      "epoch": 0.45940074228297273,
+      "grad_norm": 0.8458419442176819,
+      "learning_rate": 9.920667684350905e-05,
+      "loss": 2.7491,
+      "step": 5075
+    },
+    {
+      "epoch": 0.45949126459672307,
+      "grad_norm": 0.8307839035987854,
+      "learning_rate": 9.917494462345876e-05,
+      "loss": 2.7433,
+      "step": 5076
+    },
+    {
+      "epoch": 0.4595817869104734,
+      "grad_norm": 0.8397489786148071,
+      "learning_rate": 9.914321248649153e-05,
+      "loss": 2.8396,
+      "step": 5077
+    },
+    {
+      "epoch": 0.45967230922422375,
+      "grad_norm": 0.8725177645683289,
+      "learning_rate": 9.911148043580275e-05,
+      "loss": 2.802,
+      "step": 5078
+    },
+    {
+      "epoch": 0.4597628315379741,
+      "grad_norm": 0.9256505370140076,
+      "learning_rate": 9.907974847458791e-05,
+      "loss": 2.8105,
+      "step": 5079
+    },
+    {
+      "epoch": 0.45985335385172443,
+      "grad_norm": 0.8496103286743164,
+      "learning_rate": 9.904801660604234e-05,
+      "loss": 2.8376,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4599438761654748,
+      "grad_norm": 0.8865105509757996,
+      "learning_rate": 9.90162848333615e-05,
+      "loss": 2.842,
+      "step": 5081
+    },
+    {
+      "epoch": 0.4600343984792251,
+      "grad_norm": 0.862686038017273,
+      "learning_rate": 9.898455315974071e-05,
+      "loss": 2.7353,
+      "step": 5082
+    },
+    {
+      "epoch": 0.46012492079297546,
+      "grad_norm": 0.8391594290733337,
+      "learning_rate": 9.895282158837545e-05,
+      "loss": 2.8188,
+      "step": 5083
+    },
+    {
+      "epoch": 0.4602154431067258,
+      "grad_norm": 0.8564516305923462,
+      "learning_rate": 9.892109012246096e-05,
+      "loss": 2.8319,
+      "step": 5084
+    },
+    {
+      "epoch": 0.46030596542047614,
+      "grad_norm": 0.8370792269706726,
+      "learning_rate": 9.888935876519272e-05,
+      "loss": 2.8322,
+      "step": 5085
+    },
+    {
+      "epoch": 0.4603964877342265,
+      "grad_norm": 0.8662410974502563,
+      "learning_rate": 9.8857627519766e-05,
+      "loss": 2.7647,
+      "step": 5086
+    },
+    {
+      "epoch": 0.4604870100479768,
+      "grad_norm": 0.8782781362533569,
+      "learning_rate": 9.882589638937615e-05,
+      "loss": 2.8176,
+      "step": 5087
+    },
+    {
+      "epoch": 0.46057753236172716,
+      "grad_norm": 0.878341794013977,
+      "learning_rate": 9.87941653772185e-05,
+      "loss": 2.8341,
+      "step": 5088
+    },
+    {
+      "epoch": 0.4606680546754775,
+      "grad_norm": 0.8614301085472107,
+      "learning_rate": 9.876243448648839e-05,
+      "loss": 2.8229,
+      "step": 5089
+    },
+    {
+      "epoch": 0.46075857698922784,
+      "grad_norm": 0.9689937233924866,
+      "learning_rate": 9.873070372038105e-05,
+      "loss": 2.8517,
+      "step": 5090
+    },
+    {
+      "epoch": 0.4608490993029782,
+      "grad_norm": 0.8726419806480408,
+      "learning_rate": 9.869897308209186e-05,
+      "loss": 2.8126,
+      "step": 5091
+    },
+    {
+      "epoch": 0.4609396216167285,
+      "grad_norm": 0.8292773365974426,
+      "learning_rate": 9.8667242574816e-05,
+      "loss": 2.7537,
+      "step": 5092
+    },
+    {
+      "epoch": 0.46103014393047886,
+      "grad_norm": 0.8113095760345459,
+      "learning_rate": 9.863551220174881e-05,
+      "loss": 2.7456,
+      "step": 5093
+    },
+    {
+      "epoch": 0.4611206662442292,
+      "grad_norm": 0.8891957402229309,
+      "learning_rate": 9.860378196608549e-05,
+      "loss": 2.8136,
+      "step": 5094
+    },
+    {
+      "epoch": 0.46121118855797955,
+      "grad_norm": 0.8564567565917969,
+      "learning_rate": 9.85720518710213e-05,
+      "loss": 2.8245,
+      "step": 5095
+    },
+    {
+      "epoch": 0.4613017108717299,
+      "grad_norm": 0.8329167366027832,
+      "learning_rate": 9.85403219197514e-05,
+      "loss": 2.766,
+      "step": 5096
+    },
+    {
+      "epoch": 0.4613922331854802,
+      "grad_norm": 0.824367105960846,
+      "learning_rate": 9.85085921154711e-05,
+      "loss": 2.7998,
+      "step": 5097
+    },
+    {
+      "epoch": 0.46148275549923057,
+      "grad_norm": 0.8718504309654236,
+      "learning_rate": 9.847686246137551e-05,
+      "loss": 2.7728,
+      "step": 5098
+    },
+    {
+      "epoch": 0.4615732778129809,
+      "grad_norm": 0.8713845610618591,
+      "learning_rate": 9.844513296065984e-05,
+      "loss": 2.8486,
+      "step": 5099
+    },
+    {
+      "epoch": 0.46166380012673125,
+      "grad_norm": 0.7742902040481567,
+      "learning_rate": 9.84134036165192e-05,
+      "loss": 2.8051,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4617543224404816,
+      "grad_norm": 0.8396895527839661,
+      "learning_rate": 9.838167443214882e-05,
+      "loss": 2.803,
+      "step": 5101
+    },
+    {
+      "epoch": 0.46184484475423193,
+      "grad_norm": 0.8684070110321045,
+      "learning_rate": 9.834994541074375e-05,
+      "loss": 2.8388,
+      "step": 5102
+    },
+    {
+      "epoch": 0.46193536706798227,
+      "grad_norm": 0.8808414936065674,
+      "learning_rate": 9.831821655549916e-05,
+      "loss": 2.792,
+      "step": 5103
+    },
+    {
+      "epoch": 0.4620258893817326,
+      "grad_norm": 0.8871559500694275,
+      "learning_rate": 9.828648786961008e-05,
+      "loss": 2.8197,
+      "step": 5104
+    },
+    {
+      "epoch": 0.46211641169548295,
+      "grad_norm": 0.8918201923370361,
+      "learning_rate": 9.825475935627165e-05,
+      "loss": 2.8327,
+      "step": 5105
+    },
+    {
+      "epoch": 0.4622069340092333,
+      "grad_norm": 0.9378277659416199,
+      "learning_rate": 9.82230310186789e-05,
+      "loss": 2.7894,
+      "step": 5106
+    },
+    {
+      "epoch": 0.46229745632298364,
+      "grad_norm": 0.8530616760253906,
+      "learning_rate": 9.819130286002689e-05,
+      "loss": 2.8486,
+      "step": 5107
+    },
+    {
+      "epoch": 0.462387978636734,
+      "grad_norm": 0.8263341784477234,
+      "learning_rate": 9.81595748835106e-05,
+      "loss": 2.7534,
+      "step": 5108
+    },
+    {
+      "epoch": 0.4624785009504843,
+      "grad_norm": 0.9433355927467346,
+      "learning_rate": 9.81278470923251e-05,
+      "loss": 2.833,
+      "step": 5109
+    },
+    {
+      "epoch": 0.46256902326423466,
+      "grad_norm": 0.8154783844947815,
+      "learning_rate": 9.809611948966533e-05,
+      "loss": 2.8349,
+      "step": 5110
+    },
+    {
+      "epoch": 0.462659545577985,
+      "grad_norm": 0.8896995782852173,
+      "learning_rate": 9.80643920787263e-05,
+      "loss": 2.7907,
+      "step": 5111
+    },
+    {
+      "epoch": 0.46275006789173534,
+      "grad_norm": 0.8480448126792908,
+      "learning_rate": 9.803266486270288e-05,
+      "loss": 2.7842,
+      "step": 5112
+    },
+    {
+      "epoch": 0.4628405902054856,
+      "grad_norm": 0.844628095626831,
+      "learning_rate": 9.800093784479013e-05,
+      "loss": 2.8267,
+      "step": 5113
+    },
+    {
+      "epoch": 0.46293111251923597,
+      "grad_norm": 0.9172288775444031,
+      "learning_rate": 9.796921102818281e-05,
+      "loss": 2.8163,
+      "step": 5114
+    },
+    {
+      "epoch": 0.4630216348329863,
+      "grad_norm": 0.8828997015953064,
+      "learning_rate": 9.793748441607594e-05,
+      "loss": 2.7847,
+      "step": 5115
+    },
+    {
+      "epoch": 0.46311215714673665,
+      "grad_norm": 0.9007895588874817,
+      "learning_rate": 9.790575801166432e-05,
+      "loss": 2.8427,
+      "step": 5116
+    },
+    {
+      "epoch": 0.463202679460487,
+      "grad_norm": 0.8572633862495422,
+      "learning_rate": 9.787403181814281e-05,
+      "loss": 2.7562,
+      "step": 5117
+    },
+    {
+      "epoch": 0.46329320177423733,
+      "grad_norm": 0.8919517993927002,
+      "learning_rate": 9.784230583870621e-05,
+      "loss": 2.8008,
+      "step": 5118
+    },
+    {
+      "epoch": 0.46338372408798767,
+      "grad_norm": 0.8707318305969238,
+      "learning_rate": 9.781058007654939e-05,
+      "loss": 2.8122,
+      "step": 5119
+    },
+    {
+      "epoch": 0.463474246401738,
+      "grad_norm": 0.8310428857803345,
+      "learning_rate": 9.777885453486706e-05,
+      "loss": 2.807,
+      "step": 5120
+    },
+    {
+      "epoch": 0.46356476871548835,
+      "grad_norm": 0.8464776277542114,
+      "learning_rate": 9.774712921685407e-05,
+      "loss": 2.751,
+      "step": 5121
+    },
+    {
+      "epoch": 0.4636552910292387,
+      "grad_norm": 0.9811182618141174,
+      "learning_rate": 9.771540412570504e-05,
+      "loss": 2.8484,
+      "step": 5122
+    },
+    {
+      "epoch": 0.46374581334298903,
+      "grad_norm": 0.9233271479606628,
+      "learning_rate": 9.768367926461479e-05,
+      "loss": 2.7754,
+      "step": 5123
+    },
+    {
+      "epoch": 0.4638363356567394,
+      "grad_norm": 0.9555431604385376,
+      "learning_rate": 9.765195463677797e-05,
+      "loss": 2.8653,
+      "step": 5124
+    },
+    {
+      "epoch": 0.4639268579704897,
+      "grad_norm": 1.0181915760040283,
+      "learning_rate": 9.762023024538926e-05,
+      "loss": 2.8013,
+      "step": 5125
+    },
+    {
+      "epoch": 0.46401738028424006,
+      "grad_norm": 0.8504894971847534,
+      "learning_rate": 9.758850609364327e-05,
+      "loss": 2.7465,
+      "step": 5126
+    },
+    {
+      "epoch": 0.4641079025979904,
+      "grad_norm": 0.9759464263916016,
+      "learning_rate": 9.755678218473469e-05,
+      "loss": 2.8684,
+      "step": 5127
+    },
+    {
+      "epoch": 0.46419842491174074,
+      "grad_norm": 0.870422899723053,
+      "learning_rate": 9.752505852185805e-05,
+      "loss": 2.7721,
+      "step": 5128
+    },
+    {
+      "epoch": 0.4642889472254911,
+      "grad_norm": 0.8897386789321899,
+      "learning_rate": 9.749333510820798e-05,
+      "loss": 2.7581,
+      "step": 5129
+    },
+    {
+      "epoch": 0.4643794695392414,
+      "grad_norm": 0.8926811218261719,
+      "learning_rate": 9.746161194697895e-05,
+      "loss": 2.7956,
+      "step": 5130
+    },
+    {
+      "epoch": 0.46446999185299176,
+      "grad_norm": 0.797377347946167,
+      "learning_rate": 9.742988904136557e-05,
+      "loss": 2.8038,
+      "step": 5131
+    },
+    {
+      "epoch": 0.4645605141667421,
+      "grad_norm": 0.8459729552268982,
+      "learning_rate": 9.739816639456228e-05,
+      "loss": 2.7561,
+      "step": 5132
+    },
+    {
+      "epoch": 0.46465103648049244,
+      "grad_norm": 0.8637222051620483,
+      "learning_rate": 9.736644400976357e-05,
+      "loss": 2.772,
+      "step": 5133
+    },
+    {
+      "epoch": 0.4647415587942428,
+      "grad_norm": 0.7574160695075989,
+      "learning_rate": 9.733472189016383e-05,
+      "loss": 2.7673,
+      "step": 5134
+    },
+    {
+      "epoch": 0.4648320811079931,
+      "grad_norm": 0.8637992739677429,
+      "learning_rate": 9.730300003895758e-05,
+      "loss": 2.7775,
+      "step": 5135
+    },
+    {
+      "epoch": 0.46492260342174346,
+      "grad_norm": 0.8622542023658752,
+      "learning_rate": 9.727127845933915e-05,
+      "loss": 2.7724,
+      "step": 5136
+    },
+    {
+      "epoch": 0.4650131257354938,
+      "grad_norm": 0.9396885633468628,
+      "learning_rate": 9.723955715450287e-05,
+      "loss": 2.7755,
+      "step": 5137
+    },
+    {
+      "epoch": 0.46510364804924414,
+      "grad_norm": 0.8237531781196594,
+      "learning_rate": 9.720783612764314e-05,
+      "loss": 2.7245,
+      "step": 5138
+    },
+    {
+      "epoch": 0.4651941703629945,
+      "grad_norm": 1.1125627756118774,
+      "learning_rate": 9.717611538195419e-05,
+      "loss": 2.8233,
+      "step": 5139
+    },
+    {
+      "epoch": 0.4652846926767448,
+      "grad_norm": 0.836974024772644,
+      "learning_rate": 9.71443949206304e-05,
+      "loss": 2.7418,
+      "step": 5140
+    },
+    {
+      "epoch": 0.46537521499049517,
+      "grad_norm": 0.9694908857345581,
+      "learning_rate": 9.711267474686591e-05,
+      "loss": 2.8141,
+      "step": 5141
+    },
+    {
+      "epoch": 0.4654657373042455,
+      "grad_norm": 0.894636332988739,
+      "learning_rate": 9.708095486385504e-05,
+      "loss": 2.7566,
+      "step": 5142
+    },
+    {
+      "epoch": 0.46555625961799585,
+      "grad_norm": 0.8978986144065857,
+      "learning_rate": 9.704923527479189e-05,
+      "loss": 2.8253,
+      "step": 5143
+    },
+    {
+      "epoch": 0.4656467819317462,
+      "grad_norm": 0.8937349915504456,
+      "learning_rate": 9.701751598287074e-05,
+      "loss": 2.7733,
+      "step": 5144
+    },
+    {
+      "epoch": 0.46573730424549653,
+      "grad_norm": 0.8157208561897278,
+      "learning_rate": 9.698579699128557e-05,
+      "loss": 2.7197,
+      "step": 5145
+    },
+    {
+      "epoch": 0.46582782655924687,
+      "grad_norm": 0.8707994818687439,
+      "learning_rate": 9.69540783032306e-05,
+      "loss": 2.7368,
+      "step": 5146
+    },
+    {
+      "epoch": 0.4659183488729972,
+      "grad_norm": 0.8988774418830872,
+      "learning_rate": 9.692235992189985e-05,
+      "loss": 2.783,
+      "step": 5147
+    },
+    {
+      "epoch": 0.46600887118674755,
+      "grad_norm": 0.9029574990272522,
+      "learning_rate": 9.68906418504874e-05,
+      "loss": 2.8309,
+      "step": 5148
+    },
+    {
+      "epoch": 0.4660993935004979,
+      "grad_norm": 0.846811056137085,
+      "learning_rate": 9.685892409218717e-05,
+      "loss": 2.8838,
+      "step": 5149
+    },
+    {
+      "epoch": 0.46618991581424823,
+      "grad_norm": 0.8255429863929749,
+      "learning_rate": 9.682720665019325e-05,
+      "loss": 2.7979,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4662804381279986,
+      "grad_norm": 0.8471197485923767,
+      "learning_rate": 9.679548952769953e-05,
+      "loss": 2.7293,
+      "step": 5151
+    },
+    {
+      "epoch": 0.4663709604417489,
+      "grad_norm": 0.7674946188926697,
+      "learning_rate": 9.676377272789992e-05,
+      "loss": 2.7723,
+      "step": 5152
+    },
+    {
+      "epoch": 0.46646148275549926,
+      "grad_norm": 0.809223473072052,
+      "learning_rate": 9.673205625398827e-05,
+      "loss": 2.8182,
+      "step": 5153
+    },
+    {
+      "epoch": 0.46655200506924954,
+      "grad_norm": 0.9060366153717041,
+      "learning_rate": 9.670034010915852e-05,
+      "loss": 2.7695,
+      "step": 5154
+    },
+    {
+      "epoch": 0.4666425273829999,
+      "grad_norm": 0.8269774317741394,
+      "learning_rate": 9.66686242966044e-05,
+      "loss": 2.7963,
+      "step": 5155
+    },
+    {
+      "epoch": 0.4667330496967502,
+      "grad_norm": 0.8438358902931213,
+      "learning_rate": 9.663690881951975e-05,
+      "loss": 2.7485,
+      "step": 5156
+    },
+    {
+      "epoch": 0.46682357201050056,
+      "grad_norm": 0.9002498984336853,
+      "learning_rate": 9.660519368109823e-05,
+      "loss": 2.7802,
+      "step": 5157
+    },
+    {
+      "epoch": 0.4669140943242509,
+      "grad_norm": 0.8270025849342346,
+      "learning_rate": 9.657347888453367e-05,
+      "loss": 2.8007,
+      "step": 5158
+    },
+    {
+      "epoch": 0.46700461663800125,
+      "grad_norm": 0.892013669013977,
+      "learning_rate": 9.654176443301967e-05,
+      "loss": 2.8027,
+      "step": 5159
+    },
+    {
+      "epoch": 0.4670951389517516,
+      "grad_norm": 0.9057855606079102,
+      "learning_rate": 9.651005032974994e-05,
+      "loss": 2.7702,
+      "step": 5160
+    },
+    {
+      "epoch": 0.46718566126550193,
+      "grad_norm": 0.786109209060669,
+      "learning_rate": 9.647833657791797e-05,
+      "loss": 2.846,
+      "step": 5161
+    },
+    {
+      "epoch": 0.46727618357925227,
+      "grad_norm": 0.9512817859649658,
+      "learning_rate": 9.644662318071747e-05,
+      "loss": 2.8294,
+      "step": 5162
+    },
+    {
+      "epoch": 0.4673667058930026,
+      "grad_norm": 0.967814028263092,
+      "learning_rate": 9.641491014134189e-05,
+      "loss": 2.7562,
+      "step": 5163
+    },
+    {
+      "epoch": 0.46745722820675295,
+      "grad_norm": 0.8379111289978027,
+      "learning_rate": 9.638319746298478e-05,
+      "loss": 2.7692,
+      "step": 5164
+    },
+    {
+      "epoch": 0.4675477505205033,
+      "grad_norm": 0.8625251054763794,
+      "learning_rate": 9.635148514883956e-05,
+      "loss": 2.7676,
+      "step": 5165
+    },
+    {
+      "epoch": 0.46763827283425363,
+      "grad_norm": 0.921722412109375,
+      "learning_rate": 9.631977320209972e-05,
+      "loss": 2.8362,
+      "step": 5166
+    },
+    {
+      "epoch": 0.467728795148004,
+      "grad_norm": 0.9456198215484619,
+      "learning_rate": 9.628806162595858e-05,
+      "loss": 2.8482,
+      "step": 5167
+    },
+    {
+      "epoch": 0.4678193174617543,
+      "grad_norm": 0.9429085850715637,
+      "learning_rate": 9.625635042360957e-05,
+      "loss": 2.7669,
+      "step": 5168
+    },
+    {
+      "epoch": 0.46790983977550465,
+      "grad_norm": 1.4454035758972168,
+      "learning_rate": 9.622463959824593e-05,
+      "loss": 2.8269,
+      "step": 5169
+    },
+    {
+      "epoch": 0.468000362089255,
+      "grad_norm": 0.8234190940856934,
+      "learning_rate": 9.619292915306101e-05,
+      "loss": 2.8078,
+      "step": 5170
+    },
+    {
+      "epoch": 0.46809088440300534,
+      "grad_norm": 0.9734144806861877,
+      "learning_rate": 9.616121909124801e-05,
+      "loss": 2.7734,
+      "step": 5171
+    },
+    {
+      "epoch": 0.4681814067167557,
+      "grad_norm": 0.934136688709259,
+      "learning_rate": 9.612950941600016e-05,
+      "loss": 2.8085,
+      "step": 5172
+    },
+    {
+      "epoch": 0.468271929030506,
+      "grad_norm": 0.9146413207054138,
+      "learning_rate": 9.609780013051057e-05,
+      "loss": 2.7461,
+      "step": 5173
+    },
+    {
+      "epoch": 0.46836245134425636,
+      "grad_norm": 0.9391007423400879,
+      "learning_rate": 9.606609123797243e-05,
+      "loss": 2.8535,
+      "step": 5174
+    },
+    {
+      "epoch": 0.4684529736580067,
+      "grad_norm": 0.9421570301055908,
+      "learning_rate": 9.603438274157878e-05,
+      "loss": 2.8389,
+      "step": 5175
+    },
+    {
+      "epoch": 0.46854349597175704,
+      "grad_norm": 0.7810022830963135,
+      "learning_rate": 9.60026746445227e-05,
+      "loss": 2.7927,
+      "step": 5176
+    },
+    {
+      "epoch": 0.4686340182855074,
+      "grad_norm": 0.8332025408744812,
+      "learning_rate": 9.597096694999715e-05,
+      "loss": 2.769,
+      "step": 5177
+    },
+    {
+      "epoch": 0.4687245405992577,
+      "grad_norm": 0.841300368309021,
+      "learning_rate": 9.593925966119515e-05,
+      "loss": 2.8227,
+      "step": 5178
+    },
+    {
+      "epoch": 0.46881506291300806,
+      "grad_norm": 0.9895218014717102,
+      "learning_rate": 9.590755278130952e-05,
+      "loss": 2.8348,
+      "step": 5179
+    },
+    {
+      "epoch": 0.4689055852267584,
+      "grad_norm": 0.9044789671897888,
+      "learning_rate": 9.587584631353329e-05,
+      "loss": 2.8038,
+      "step": 5180
+    },
+    {
+      "epoch": 0.46899610754050874,
+      "grad_norm": 0.8770217299461365,
+      "learning_rate": 9.584414026105918e-05,
+      "loss": 2.8169,
+      "step": 5181
+    },
+    {
+      "epoch": 0.4690866298542591,
+      "grad_norm": 0.9499712586402893,
+      "learning_rate": 9.581243462708006e-05,
+      "loss": 2.8173,
+      "step": 5182
+    },
+    {
+      "epoch": 0.4691771521680094,
+      "grad_norm": 0.8335294127464294,
+      "learning_rate": 9.578072941478861e-05,
+      "loss": 2.7457,
+      "step": 5183
+    },
+    {
+      "epoch": 0.46926767448175977,
+      "grad_norm": 0.9775024056434631,
+      "learning_rate": 9.574902462737765e-05,
+      "loss": 2.8145,
+      "step": 5184
+    },
+    {
+      "epoch": 0.4693581967955101,
+      "grad_norm": 0.9454213380813599,
+      "learning_rate": 9.571732026803977e-05,
+      "loss": 2.766,
+      "step": 5185
+    },
+    {
+      "epoch": 0.46944871910926045,
+      "grad_norm": 0.9457453489303589,
+      "learning_rate": 9.568561633996763e-05,
+      "loss": 2.8457,
+      "step": 5186
+    },
+    {
+      "epoch": 0.4695392414230108,
+      "grad_norm": 0.8910146355628967,
+      "learning_rate": 9.565391284635378e-05,
+      "loss": 2.8096,
+      "step": 5187
+    },
+    {
+      "epoch": 0.46962976373676113,
+      "grad_norm": 0.8416481614112854,
+      "learning_rate": 9.562220979039082e-05,
+      "loss": 2.8409,
+      "step": 5188
+    },
+    {
+      "epoch": 0.46972028605051147,
+      "grad_norm": 0.9077867865562439,
+      "learning_rate": 9.559050717527119e-05,
+      "loss": 2.7308,
+      "step": 5189
+    },
+    {
+      "epoch": 0.4698108083642618,
+      "grad_norm": 0.8774727582931519,
+      "learning_rate": 9.55588050041874e-05,
+      "loss": 2.7031,
+      "step": 5190
+    },
+    {
+      "epoch": 0.46990133067801215,
+      "grad_norm": 0.9110966324806213,
+      "learning_rate": 9.552710328033176e-05,
+      "loss": 2.7335,
+      "step": 5191
+    },
+    {
+      "epoch": 0.4699918529917625,
+      "grad_norm": 0.9437413811683655,
+      "learning_rate": 9.549540200689676e-05,
+      "loss": 2.8008,
+      "step": 5192
+    },
+    {
+      "epoch": 0.47008237530551283,
+      "grad_norm": 0.9005468487739563,
+      "learning_rate": 9.546370118707463e-05,
+      "loss": 2.8006,
+      "step": 5193
+    },
+    {
+      "epoch": 0.4701728976192632,
+      "grad_norm": 0.9098567962646484,
+      "learning_rate": 9.543200082405768e-05,
+      "loss": 2.8465,
+      "step": 5194
+    },
+    {
+      "epoch": 0.47026341993301346,
+      "grad_norm": 1.0034912824630737,
+      "learning_rate": 9.540030092103809e-05,
+      "loss": 2.8063,
+      "step": 5195
+    },
+    {
+      "epoch": 0.4703539422467638,
+      "grad_norm": 0.873590350151062,
+      "learning_rate": 9.536860148120811e-05,
+      "loss": 2.7531,
+      "step": 5196
+    },
+    {
+      "epoch": 0.47044446456051414,
+      "grad_norm": 0.8259015679359436,
+      "learning_rate": 9.53369025077598e-05,
+      "loss": 2.8364,
+      "step": 5197
+    },
+    {
+      "epoch": 0.4705349868742645,
+      "grad_norm": 0.9385213851928711,
+      "learning_rate": 9.530520400388531e-05,
+      "loss": 2.7638,
+      "step": 5198
+    },
+    {
+      "epoch": 0.4706255091880148,
+      "grad_norm": 0.8609074354171753,
+      "learning_rate": 9.527350597277661e-05,
+      "loss": 2.7279,
+      "step": 5199
+    },
+    {
+      "epoch": 0.47071603150176516,
+      "grad_norm": 0.7989048957824707,
+      "learning_rate": 9.524180841762577e-05,
+      "loss": 2.7767,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4708065538155155,
+      "grad_norm": 0.8445473313331604,
+      "learning_rate": 9.521011134162467e-05,
+      "loss": 2.7703,
+      "step": 5201
+    },
+    {
+      "epoch": 0.47089707612926585,
+      "grad_norm": 0.9142276048660278,
+      "learning_rate": 9.517841474796526e-05,
+      "loss": 2.7968,
+      "step": 5202
+    },
+    {
+      "epoch": 0.4709875984430162,
+      "grad_norm": 0.8804418444633484,
+      "learning_rate": 9.51467186398393e-05,
+      "loss": 2.7839,
+      "step": 5203
+    },
+    {
+      "epoch": 0.4710781207567665,
+      "grad_norm": 0.9042705297470093,
+      "learning_rate": 9.511502302043868e-05,
+      "loss": 2.7856,
+      "step": 5204
+    },
+    {
+      "epoch": 0.47116864307051687,
+      "grad_norm": 0.8807315826416016,
+      "learning_rate": 9.508332789295508e-05,
+      "loss": 2.8576,
+      "step": 5205
+    },
+    {
+      "epoch": 0.4712591653842672,
+      "grad_norm": 0.9314857721328735,
+      "learning_rate": 9.505163326058027e-05,
+      "loss": 2.7471,
+      "step": 5206
+    },
+    {
+      "epoch": 0.47134968769801755,
+      "grad_norm": 0.8590176105499268,
+      "learning_rate": 9.50199391265058e-05,
+      "loss": 2.8074,
+      "step": 5207
+    },
+    {
+      "epoch": 0.4714402100117679,
+      "grad_norm": 0.8539712429046631,
+      "learning_rate": 9.498824549392337e-05,
+      "loss": 2.7981,
+      "step": 5208
+    },
+    {
+      "epoch": 0.47153073232551823,
+      "grad_norm": 0.8064674139022827,
+      "learning_rate": 9.495655236602444e-05,
+      "loss": 2.7815,
+      "step": 5209
+    },
+    {
+      "epoch": 0.4716212546392686,
+      "grad_norm": 0.8869786262512207,
+      "learning_rate": 9.492485974600059e-05,
+      "loss": 2.8347,
+      "step": 5210
+    },
+    {
+      "epoch": 0.4717117769530189,
+      "grad_norm": 0.8497887253761292,
+      "learning_rate": 9.48931676370432e-05,
+      "loss": 2.7493,
+      "step": 5211
+    },
+    {
+      "epoch": 0.47180229926676925,
+      "grad_norm": 0.8447049856185913,
+      "learning_rate": 9.486147604234371e-05,
+      "loss": 2.7629,
+      "step": 5212
+    },
+    {
+      "epoch": 0.4718928215805196,
+      "grad_norm": 0.8657582402229309,
+      "learning_rate": 9.482978496509341e-05,
+      "loss": 2.7943,
+      "step": 5213
+    },
+    {
+      "epoch": 0.47198334389426994,
+      "grad_norm": 0.8378434777259827,
+      "learning_rate": 9.479809440848364e-05,
+      "loss": 2.7565,
+      "step": 5214
+    },
+    {
+      "epoch": 0.4720738662080203,
+      "grad_norm": 0.8710640668869019,
+      "learning_rate": 9.476640437570562e-05,
+      "loss": 2.8118,
+      "step": 5215
+    },
+    {
+      "epoch": 0.4721643885217706,
+      "grad_norm": 0.866980254650116,
+      "learning_rate": 9.473471486995055e-05,
+      "loss": 2.794,
+      "step": 5216
+    },
+    {
+      "epoch": 0.47225491083552096,
+      "grad_norm": 0.9196212887763977,
+      "learning_rate": 9.470302589440952e-05,
+      "loss": 2.8236,
+      "step": 5217
+    },
+    {
+      "epoch": 0.4723454331492713,
+      "grad_norm": 0.9319663047790527,
+      "learning_rate": 9.467133745227366e-05,
+      "loss": 2.7992,
+      "step": 5218
+    },
+    {
+      "epoch": 0.47243595546302164,
+      "grad_norm": 0.8404010534286499,
+      "learning_rate": 9.463964954673396e-05,
+      "loss": 2.8101,
+      "step": 5219
+    },
+    {
+      "epoch": 0.472526477776772,
+      "grad_norm": 0.9603475332260132,
+      "learning_rate": 9.460796218098143e-05,
+      "loss": 2.7364,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4726170000905223,
+      "grad_norm": 0.8324125409126282,
+      "learning_rate": 9.457627535820695e-05,
+      "loss": 2.7999,
+      "step": 5221
+    },
+    {
+      "epoch": 0.47270752240427266,
+      "grad_norm": 0.9238003492355347,
+      "learning_rate": 9.454458908160135e-05,
+      "loss": 2.7564,
+      "step": 5222
+    },
+    {
+      "epoch": 0.472798044718023,
+      "grad_norm": 0.866795003414154,
+      "learning_rate": 9.451290335435553e-05,
+      "loss": 2.7623,
+      "step": 5223
+    },
+    {
+      "epoch": 0.47288856703177334,
+      "grad_norm": 0.8267090320587158,
+      "learning_rate": 9.448121817966018e-05,
+      "loss": 2.7862,
+      "step": 5224
+    },
+    {
+      "epoch": 0.4729790893455237,
+      "grad_norm": 0.8153204917907715,
+      "learning_rate": 9.444953356070601e-05,
+      "loss": 2.7832,
+      "step": 5225
+    },
+    {
+      "epoch": 0.473069611659274,
+      "grad_norm": 0.8457561731338501,
+      "learning_rate": 9.441784950068362e-05,
+      "loss": 2.8002,
+      "step": 5226
+    },
+    {
+      "epoch": 0.47316013397302437,
+      "grad_norm": 0.8408893346786499,
+      "learning_rate": 9.438616600278368e-05,
+      "loss": 2.7474,
+      "step": 5227
+    },
+    {
+      "epoch": 0.4732506562867747,
+      "grad_norm": 0.9174901843070984,
+      "learning_rate": 9.435448307019665e-05,
+      "loss": 2.828,
+      "step": 5228
+    },
+    {
+      "epoch": 0.47334117860052505,
+      "grad_norm": 0.8532944917678833,
+      "learning_rate": 9.432280070611305e-05,
+      "loss": 2.7933,
+      "step": 5229
+    },
+    {
+      "epoch": 0.4734317009142754,
+      "grad_norm": 0.8048722743988037,
+      "learning_rate": 9.42911189137232e-05,
+      "loss": 2.7738,
+      "step": 5230
+    },
+    {
+      "epoch": 0.47352222322802573,
+      "grad_norm": 0.8400250673294067,
+      "learning_rate": 9.425943769621757e-05,
+      "loss": 2.784,
+      "step": 5231
+    },
+    {
+      "epoch": 0.47361274554177607,
+      "grad_norm": 0.8464124798774719,
+      "learning_rate": 9.422775705678639e-05,
+      "loss": 2.8168,
+      "step": 5232
+    },
+    {
+      "epoch": 0.4737032678555264,
+      "grad_norm": 0.8276224136352539,
+      "learning_rate": 9.419607699861994e-05,
+      "loss": 2.8354,
+      "step": 5233
+    },
+    {
+      "epoch": 0.47379379016927675,
+      "grad_norm": 0.8101887702941895,
+      "learning_rate": 9.416439752490834e-05,
+      "loss": 2.7846,
+      "step": 5234
+    },
+    {
+      "epoch": 0.4738843124830271,
+      "grad_norm": 0.863121747970581,
+      "learning_rate": 9.413271863884178e-05,
+      "loss": 2.8169,
+      "step": 5235
+    },
+    {
+      "epoch": 0.47397483479677743,
+      "grad_norm": 0.89903724193573,
+      "learning_rate": 9.410104034361027e-05,
+      "loss": 2.7933,
+      "step": 5236
+    },
+    {
+      "epoch": 0.4740653571105277,
+      "grad_norm": 0.8720853924751282,
+      "learning_rate": 9.406936264240386e-05,
+      "loss": 2.8064,
+      "step": 5237
+    },
+    {
+      "epoch": 0.47415587942427806,
+      "grad_norm": 0.9143301844596863,
+      "learning_rate": 9.403768553841244e-05,
+      "loss": 2.8193,
+      "step": 5238
+    },
+    {
+      "epoch": 0.4742464017380284,
+      "grad_norm": 0.8418611288070679,
+      "learning_rate": 9.400600903482597e-05,
+      "loss": 2.7703,
+      "step": 5239
+    },
+    {
+      "epoch": 0.47433692405177874,
+      "grad_norm": 0.8874722719192505,
+      "learning_rate": 9.397433313483416e-05,
+      "loss": 2.8001,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4744274463655291,
+      "grad_norm": 0.9456201195716858,
+      "learning_rate": 9.394265784162689e-05,
+      "loss": 2.7166,
+      "step": 5241
+    },
+    {
+      "epoch": 0.4745179686792794,
+      "grad_norm": 0.8349551558494568,
+      "learning_rate": 9.39109831583938e-05,
+      "loss": 2.8146,
+      "step": 5242
+    },
+    {
+      "epoch": 0.47460849099302976,
+      "grad_norm": 0.8971189856529236,
+      "learning_rate": 9.387930908832454e-05,
+      "loss": 2.7563,
+      "step": 5243
+    },
+    {
+      "epoch": 0.4746990133067801,
+      "grad_norm": 0.8368891477584839,
+      "learning_rate": 9.384763563460865e-05,
+      "loss": 2.8169,
+      "step": 5244
+    },
+    {
+      "epoch": 0.47478953562053045,
+      "grad_norm": 0.7798503637313843,
+      "learning_rate": 9.381596280043573e-05,
+      "loss": 2.7759,
+      "step": 5245
+    },
+    {
+      "epoch": 0.4748800579342808,
+      "grad_norm": 0.8371593356132507,
+      "learning_rate": 9.378429058899516e-05,
+      "loss": 2.7602,
+      "step": 5246
+    },
+    {
+      "epoch": 0.4749705802480311,
+      "grad_norm": 0.9390305280685425,
+      "learning_rate": 9.37526190034764e-05,
+      "loss": 2.7563,
+      "step": 5247
+    },
+    {
+      "epoch": 0.47506110256178147,
+      "grad_norm": 0.8491203784942627,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 2.8111,
+      "step": 5248
+    },
+    {
+      "epoch": 0.4751516248755318,
+      "grad_norm": 0.9118536114692688,
+      "learning_rate": 9.368927772296134e-05,
+      "loss": 2.764,
+      "step": 5249
+    },
+    {
+      "epoch": 0.47524214718928215,
+      "grad_norm": 0.9842036962509155,
+      "learning_rate": 9.365760803434355e-05,
+      "loss": 2.8288,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4753326695030325,
+      "grad_norm": 1.0022794008255005,
+      "learning_rate": 9.36259389844045e-05,
+      "loss": 2.8153,
+      "step": 5251
+    },
+    {
+      "epoch": 0.47542319181678283,
+      "grad_norm": 0.8196097016334534,
+      "learning_rate": 9.359427057633316e-05,
+      "loss": 2.8021,
+      "step": 5252
+    },
+    {
+      "epoch": 0.47551371413053317,
+      "grad_norm": 0.839787483215332,
+      "learning_rate": 9.356260281331864e-05,
+      "loss": 2.7805,
+      "step": 5253
+    },
+    {
+      "epoch": 0.4756042364442835,
+      "grad_norm": 0.8965675234794617,
+      "learning_rate": 9.353093569854983e-05,
+      "loss": 2.7889,
+      "step": 5254
+    },
+    {
+      "epoch": 0.47569475875803385,
+      "grad_norm": 0.8917959928512573,
+      "learning_rate": 9.349926923521563e-05,
+      "loss": 2.8078,
+      "step": 5255
+    },
+    {
+      "epoch": 0.4757852810717842,
+      "grad_norm": 0.8598231077194214,
+      "learning_rate": 9.34676034265048e-05,
+      "loss": 2.7312,
+      "step": 5256
+    },
+    {
+      "epoch": 0.47587580338553453,
+      "grad_norm": 0.8513514995574951,
+      "learning_rate": 9.343593827560617e-05,
+      "loss": 2.787,
+      "step": 5257
+    },
+    {
+      "epoch": 0.4759663256992849,
+      "grad_norm": 0.81255704164505,
+      "learning_rate": 9.340427378570835e-05,
+      "loss": 2.7823,
+      "step": 5258
+    },
+    {
+      "epoch": 0.4760568480130352,
+      "grad_norm": 0.8444796800613403,
+      "learning_rate": 9.337260996000002e-05,
+      "loss": 2.7665,
+      "step": 5259
+    },
+    {
+      "epoch": 0.47614737032678556,
+      "grad_norm": 0.9492307901382446,
+      "learning_rate": 9.334094680166962e-05,
+      "loss": 2.7264,
+      "step": 5260
+    },
+    {
+      "epoch": 0.4762378926405359,
+      "grad_norm": 0.809389591217041,
+      "learning_rate": 9.330928431390575e-05,
+      "loss": 2.7932,
+      "step": 5261
+    },
+    {
+      "epoch": 0.47632841495428624,
+      "grad_norm": 0.8459622859954834,
+      "learning_rate": 9.327762249989673e-05,
+      "loss": 2.765,
+      "step": 5262
+    },
+    {
+      "epoch": 0.4764189372680366,
+      "grad_norm": 0.8855199217796326,
+      "learning_rate": 9.324596136283097e-05,
+      "loss": 2.7769,
+      "step": 5263
+    },
+    {
+      "epoch": 0.4765094595817869,
+      "grad_norm": 0.8350811004638672,
+      "learning_rate": 9.321430090589668e-05,
+      "loss": 2.736,
+      "step": 5264
+    },
+    {
+      "epoch": 0.47659998189553726,
+      "grad_norm": 0.8843063116073608,
+      "learning_rate": 9.318264113228214e-05,
+      "loss": 2.7474,
+      "step": 5265
+    },
+    {
+      "epoch": 0.4766905042092876,
+      "grad_norm": 0.8612480759620667,
+      "learning_rate": 9.315098204517543e-05,
+      "loss": 2.8543,
+      "step": 5266
+    },
+    {
+      "epoch": 0.47678102652303794,
+      "grad_norm": 0.8599689602851868,
+      "learning_rate": 9.311932364776466e-05,
+      "loss": 2.8309,
+      "step": 5267
+    },
+    {
+      "epoch": 0.4768715488367883,
+      "grad_norm": 0.7801754474639893,
+      "learning_rate": 9.308766594323778e-05,
+      "loss": 2.7796,
+      "step": 5268
+    },
+    {
+      "epoch": 0.4769620711505386,
+      "grad_norm": 0.853583037853241,
+      "learning_rate": 9.30560089347828e-05,
+      "loss": 2.7741,
+      "step": 5269
+    },
+    {
+      "epoch": 0.47705259346428897,
+      "grad_norm": 0.9441990852355957,
+      "learning_rate": 9.302435262558747e-05,
+      "loss": 2.7768,
+      "step": 5270
+    },
+    {
+      "epoch": 0.4771431157780393,
+      "grad_norm": 0.9938909411430359,
+      "learning_rate": 9.299269701883968e-05,
+      "loss": 2.8211,
+      "step": 5271
+    },
+    {
+      "epoch": 0.47723363809178965,
+      "grad_norm": 0.8555877804756165,
+      "learning_rate": 9.296104211772709e-05,
+      "loss": 2.7621,
+      "step": 5272
+    },
+    {
+      "epoch": 0.47732416040554,
+      "grad_norm": 0.938060998916626,
+      "learning_rate": 9.29293879254374e-05,
+      "loss": 2.7601,
+      "step": 5273
+    },
+    {
+      "epoch": 0.47741468271929033,
+      "grad_norm": 0.8729612827301025,
+      "learning_rate": 9.28977344451581e-05,
+      "loss": 2.7921,
+      "step": 5274
+    },
+    {
+      "epoch": 0.47750520503304067,
+      "grad_norm": 0.87042635679245,
+      "learning_rate": 9.286608168007678e-05,
+      "loss": 2.8061,
+      "step": 5275
+    },
+    {
+      "epoch": 0.477595727346791,
+      "grad_norm": 0.8272691965103149,
+      "learning_rate": 9.283442963338083e-05,
+      "loss": 2.7974,
+      "step": 5276
+    },
+    {
+      "epoch": 0.47768624966054135,
+      "grad_norm": 0.826310932636261,
+      "learning_rate": 9.280277830825763e-05,
+      "loss": 2.7249,
+      "step": 5277
+    },
+    {
+      "epoch": 0.47777677197429164,
+      "grad_norm": 0.8631619215011597,
+      "learning_rate": 9.277112770789443e-05,
+      "loss": 2.7685,
+      "step": 5278
+    },
+    {
+      "epoch": 0.477867294288042,
+      "grad_norm": 0.8807775974273682,
+      "learning_rate": 9.273947783547851e-05,
+      "loss": 2.8475,
+      "step": 5279
+    },
+    {
+      "epoch": 0.4779578166017923,
+      "grad_norm": 0.8230133056640625,
+      "learning_rate": 9.270782869419694e-05,
+      "loss": 2.7196,
+      "step": 5280
+    },
+    {
+      "epoch": 0.47804833891554266,
+      "grad_norm": 0.9338566064834595,
+      "learning_rate": 9.267618028723686e-05,
+      "loss": 2.8041,
+      "step": 5281
+    },
+    {
+      "epoch": 0.478138861229293,
+      "grad_norm": 0.8943211436271667,
+      "learning_rate": 9.264453261778517e-05,
+      "loss": 2.7414,
+      "step": 5282
+    },
+    {
+      "epoch": 0.47822938354304334,
+      "grad_norm": 0.8780472278594971,
+      "learning_rate": 9.261288568902889e-05,
+      "loss": 2.829,
+      "step": 5283
+    },
+    {
+      "epoch": 0.4783199058567937,
+      "grad_norm": 0.8478779792785645,
+      "learning_rate": 9.258123950415479e-05,
+      "loss": 2.8234,
+      "step": 5284
+    },
+    {
+      "epoch": 0.478410428170544,
+      "grad_norm": 0.841843843460083,
+      "learning_rate": 9.25495940663497e-05,
+      "loss": 2.7386,
+      "step": 5285
+    },
+    {
+      "epoch": 0.47850095048429436,
+      "grad_norm": 1.0312539339065552,
+      "learning_rate": 9.251794937880023e-05,
+      "loss": 2.8354,
+      "step": 5286
+    },
+    {
+      "epoch": 0.4785914727980447,
+      "grad_norm": 0.8610813617706299,
+      "learning_rate": 9.24863054446931e-05,
+      "loss": 2.769,
+      "step": 5287
+    },
+    {
+      "epoch": 0.47868199511179504,
+      "grad_norm": 0.8196991086006165,
+      "learning_rate": 9.24546622672148e-05,
+      "loss": 2.7291,
+      "step": 5288
+    },
+    {
+      "epoch": 0.4787725174255454,
+      "grad_norm": 0.8547167181968689,
+      "learning_rate": 9.242301984955182e-05,
+      "loss": 2.7811,
+      "step": 5289
+    },
+    {
+      "epoch": 0.4788630397392957,
+      "grad_norm": 0.8360316753387451,
+      "learning_rate": 9.239137819489047e-05,
+      "loss": 2.795,
+      "step": 5290
+    },
+    {
+      "epoch": 0.47895356205304607,
+      "grad_norm": 0.831993579864502,
+      "learning_rate": 9.235973730641719e-05,
+      "loss": 2.7939,
+      "step": 5291
+    },
+    {
+      "epoch": 0.4790440843667964,
+      "grad_norm": 0.8065147995948792,
+      "learning_rate": 9.232809718731814e-05,
+      "loss": 2.8107,
+      "step": 5292
+    },
+    {
+      "epoch": 0.47913460668054675,
+      "grad_norm": 0.8437579274177551,
+      "learning_rate": 9.229645784077949e-05,
+      "loss": 2.8158,
+      "step": 5293
+    },
+    {
+      "epoch": 0.4792251289942971,
+      "grad_norm": 0.876053512096405,
+      "learning_rate": 9.22648192699873e-05,
+      "loss": 2.8033,
+      "step": 5294
+    },
+    {
+      "epoch": 0.47931565130804743,
+      "grad_norm": 0.8478907346725464,
+      "learning_rate": 9.223318147812763e-05,
+      "loss": 2.8328,
+      "step": 5295
+    },
+    {
+      "epoch": 0.47940617362179777,
+      "grad_norm": 0.818863570690155,
+      "learning_rate": 9.220154446838637e-05,
+      "loss": 2.7338,
+      "step": 5296
+    },
+    {
+      "epoch": 0.4794966959355481,
+      "grad_norm": 0.8235446810722351,
+      "learning_rate": 9.216990824394937e-05,
+      "loss": 2.8377,
+      "step": 5297
+    },
+    {
+      "epoch": 0.47958721824929845,
+      "grad_norm": 0.9009406566619873,
+      "learning_rate": 9.213827280800236e-05,
+      "loss": 2.7701,
+      "step": 5298
+    },
+    {
+      "epoch": 0.4796777405630488,
+      "grad_norm": 0.8477178812026978,
+      "learning_rate": 9.21066381637311e-05,
+      "loss": 2.773,
+      "step": 5299
+    },
+    {
+      "epoch": 0.47976826287679913,
+      "grad_norm": 0.8886308073997498,
+      "learning_rate": 9.207500431432115e-05,
+      "loss": 2.7909,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4798587851905495,
+      "grad_norm": 0.8153219223022461,
+      "learning_rate": 9.204337126295807e-05,
+      "loss": 2.7347,
+      "step": 5301
+    },
+    {
+      "epoch": 0.4799493075042998,
+      "grad_norm": 0.8049747347831726,
+      "learning_rate": 9.201173901282724e-05,
+      "loss": 2.7898,
+      "step": 5302
+    },
+    {
+      "epoch": 0.48003982981805016,
+      "grad_norm": 0.8145269155502319,
+      "learning_rate": 9.198010756711412e-05,
+      "loss": 2.7629,
+      "step": 5303
+    },
+    {
+      "epoch": 0.4801303521318005,
+      "grad_norm": 0.909950315952301,
+      "learning_rate": 9.194847692900391e-05,
+      "loss": 2.8089,
+      "step": 5304
+    },
+    {
+      "epoch": 0.48022087444555084,
+      "grad_norm": 0.8205299377441406,
+      "learning_rate": 9.191684710168188e-05,
+      "loss": 2.7675,
+      "step": 5305
+    },
+    {
+      "epoch": 0.4803113967593012,
+      "grad_norm": 0.9104087352752686,
+      "learning_rate": 9.188521808833311e-05,
+      "loss": 2.824,
+      "step": 5306
+    },
+    {
+      "epoch": 0.4804019190730515,
+      "grad_norm": 0.956626296043396,
+      "learning_rate": 9.185358989214268e-05,
+      "loss": 2.7594,
+      "step": 5307
+    },
+    {
+      "epoch": 0.48049244138680186,
+      "grad_norm": 0.8074916005134583,
+      "learning_rate": 9.182196251629552e-05,
+      "loss": 2.8271,
+      "step": 5308
+    },
+    {
+      "epoch": 0.4805829637005522,
+      "grad_norm": 0.8834424018859863,
+      "learning_rate": 9.179033596397647e-05,
+      "loss": 2.8061,
+      "step": 5309
+    },
+    {
+      "epoch": 0.48067348601430254,
+      "grad_norm": 0.8030386567115784,
+      "learning_rate": 9.175871023837042e-05,
+      "loss": 2.7642,
+      "step": 5310
+    },
+    {
+      "epoch": 0.4807640083280529,
+      "grad_norm": 0.8542213439941406,
+      "learning_rate": 9.172708534266197e-05,
+      "loss": 2.8134,
+      "step": 5311
+    },
+    {
+      "epoch": 0.4808545306418032,
+      "grad_norm": 0.8158377408981323,
+      "learning_rate": 9.169546128003585e-05,
+      "loss": 2.7797,
+      "step": 5312
+    },
+    {
+      "epoch": 0.48094505295555356,
+      "grad_norm": 0.7623111605644226,
+      "learning_rate": 9.16638380536765e-05,
+      "loss": 2.7451,
+      "step": 5313
+    },
+    {
+      "epoch": 0.4810355752693039,
+      "grad_norm": 0.8422583937644958,
+      "learning_rate": 9.163221566676847e-05,
+      "loss": 2.8154,
+      "step": 5314
+    },
+    {
+      "epoch": 0.48112609758305425,
+      "grad_norm": 0.9185038805007935,
+      "learning_rate": 9.160059412249608e-05,
+      "loss": 2.7935,
+      "step": 5315
+    },
+    {
+      "epoch": 0.4812166198968046,
+      "grad_norm": 0.8019948601722717,
+      "learning_rate": 9.156897342404364e-05,
+      "loss": 2.7404,
+      "step": 5316
+    },
+    {
+      "epoch": 0.48130714221055493,
+      "grad_norm": 0.8030464053153992,
+      "learning_rate": 9.15373535745953e-05,
+      "loss": 2.7628,
+      "step": 5317
+    },
+    {
+      "epoch": 0.48139766452430527,
+      "grad_norm": 0.8811296820640564,
+      "learning_rate": 9.150573457733529e-05,
+      "loss": 2.8164,
+      "step": 5318
+    },
+    {
+      "epoch": 0.48148818683805555,
+      "grad_norm": 0.8141608834266663,
+      "learning_rate": 9.147411643544755e-05,
+      "loss": 2.826,
+      "step": 5319
+    },
+    {
+      "epoch": 0.4815787091518059,
+      "grad_norm": 0.8477288484573364,
+      "learning_rate": 9.144249915211605e-05,
+      "loss": 2.7852,
+      "step": 5320
+    },
+    {
+      "epoch": 0.48166923146555624,
+      "grad_norm": 0.811499297618866,
+      "learning_rate": 9.141088273052463e-05,
+      "loss": 2.7776,
+      "step": 5321
+    },
+    {
+      "epoch": 0.4817597537793066,
+      "grad_norm": 0.8425334095954895,
+      "learning_rate": 9.137926717385711e-05,
+      "loss": 2.797,
+      "step": 5322
+    },
+    {
+      "epoch": 0.4818502760930569,
+      "grad_norm": 0.9187545776367188,
+      "learning_rate": 9.134765248529715e-05,
+      "loss": 2.7419,
+      "step": 5323
+    },
+    {
+      "epoch": 0.48194079840680726,
+      "grad_norm": 0.8210468888282776,
+      "learning_rate": 9.131603866802835e-05,
+      "loss": 2.7954,
+      "step": 5324
+    },
+    {
+      "epoch": 0.4820313207205576,
+      "grad_norm": 0.844838559627533,
+      "learning_rate": 9.128442572523417e-05,
+      "loss": 2.7602,
+      "step": 5325
+    },
+    {
+      "epoch": 0.48212184303430794,
+      "grad_norm": 0.8497430086135864,
+      "learning_rate": 9.125281366009815e-05,
+      "loss": 2.7525,
+      "step": 5326
+    },
+    {
+      "epoch": 0.4822123653480583,
+      "grad_norm": 0.8585840463638306,
+      "learning_rate": 9.122120247580351e-05,
+      "loss": 2.803,
+      "step": 5327
+    },
+    {
+      "epoch": 0.4823028876618086,
+      "grad_norm": 0.9494702219963074,
+      "learning_rate": 9.118959217553358e-05,
+      "loss": 2.7816,
+      "step": 5328
+    },
+    {
+      "epoch": 0.48239340997555896,
+      "grad_norm": 0.9262629151344299,
+      "learning_rate": 9.115798276247143e-05,
+      "loss": 2.7696,
+      "step": 5329
+    },
+    {
+      "epoch": 0.4824839322893093,
+      "grad_norm": 0.8653173446655273,
+      "learning_rate": 9.112637423980021e-05,
+      "loss": 2.8015,
+      "step": 5330
+    },
+    {
+      "epoch": 0.48257445460305964,
+      "grad_norm": 0.841497540473938,
+      "learning_rate": 9.109476661070285e-05,
+      "loss": 2.7352,
+      "step": 5331
+    },
+    {
+      "epoch": 0.48266497691681,
+      "grad_norm": 0.8958382606506348,
+      "learning_rate": 9.106315987836225e-05,
+      "loss": 2.7476,
+      "step": 5332
+    },
+    {
+      "epoch": 0.4827554992305603,
+      "grad_norm": 0.8004329204559326,
+      "learning_rate": 9.10315540459612e-05,
+      "loss": 2.7025,
+      "step": 5333
+    },
+    {
+      "epoch": 0.48284602154431067,
+      "grad_norm": 0.8242617845535278,
+      "learning_rate": 9.099994911668242e-05,
+      "loss": 2.7859,
+      "step": 5334
+    },
+    {
+      "epoch": 0.482936543858061,
+      "grad_norm": 0.8296079635620117,
+      "learning_rate": 9.096834509370849e-05,
+      "loss": 2.7148,
+      "step": 5335
+    },
+    {
+      "epoch": 0.48302706617181135,
+      "grad_norm": 0.8059534430503845,
+      "learning_rate": 9.093674198022201e-05,
+      "loss": 2.7773,
+      "step": 5336
+    },
+    {
+      "epoch": 0.4831175884855617,
+      "grad_norm": 0.8140705227851868,
+      "learning_rate": 9.090513977940532e-05,
+      "loss": 2.7571,
+      "step": 5337
+    },
+    {
+      "epoch": 0.48320811079931203,
+      "grad_norm": 0.9303644299507141,
+      "learning_rate": 9.087353849444085e-05,
+      "loss": 2.839,
+      "step": 5338
+    },
+    {
+      "epoch": 0.48329863311306237,
+      "grad_norm": 0.8803795576095581,
+      "learning_rate": 9.084193812851075e-05,
+      "loss": 2.8175,
+      "step": 5339
+    },
+    {
+      "epoch": 0.4833891554268127,
+      "grad_norm": 0.8591445088386536,
+      "learning_rate": 9.081033868479727e-05,
+      "loss": 2.7367,
+      "step": 5340
+    },
+    {
+      "epoch": 0.48347967774056305,
+      "grad_norm": 0.819056510925293,
+      "learning_rate": 9.077874016648243e-05,
+      "loss": 2.784,
+      "step": 5341
+    },
+    {
+      "epoch": 0.4835702000543134,
+      "grad_norm": 0.8365514874458313,
+      "learning_rate": 9.07471425767482e-05,
+      "loss": 2.7453,
+      "step": 5342
+    },
+    {
+      "epoch": 0.48366072236806373,
+      "grad_norm": 0.8705887794494629,
+      "learning_rate": 9.071554591877643e-05,
+      "loss": 2.836,
+      "step": 5343
+    },
+    {
+      "epoch": 0.4837512446818141,
+      "grad_norm": 0.8518171310424805,
+      "learning_rate": 9.068395019574897e-05,
+      "loss": 2.827,
+      "step": 5344
+    },
+    {
+      "epoch": 0.4838417669955644,
+      "grad_norm": 0.8364830613136292,
+      "learning_rate": 9.065235541084744e-05,
+      "loss": 2.7149,
+      "step": 5345
+    },
+    {
+      "epoch": 0.48393228930931476,
+      "grad_norm": 0.8858799338340759,
+      "learning_rate": 9.06207615672535e-05,
+      "loss": 2.7244,
+      "step": 5346
+    },
+    {
+      "epoch": 0.4840228116230651,
+      "grad_norm": 0.8261039853096008,
+      "learning_rate": 9.058916866814858e-05,
+      "loss": 2.7948,
+      "step": 5347
+    },
+    {
+      "epoch": 0.48411333393681544,
+      "grad_norm": 0.8314043283462524,
+      "learning_rate": 9.055757671671413e-05,
+      "loss": 2.729,
+      "step": 5348
+    },
+    {
+      "epoch": 0.4842038562505658,
+      "grad_norm": 0.8006643652915955,
+      "learning_rate": 9.052598571613144e-05,
+      "loss": 2.7492,
+      "step": 5349
+    },
+    {
+      "epoch": 0.4842943785643161,
+      "grad_norm": 0.8151100277900696,
+      "learning_rate": 9.049439566958175e-05,
+      "loss": 2.7673,
+      "step": 5350
+    },
+    {
+      "epoch": 0.48438490087806646,
+      "grad_norm": 0.9020787477493286,
+      "learning_rate": 9.046280658024612e-05,
+      "loss": 2.7503,
+      "step": 5351
+    },
+    {
+      "epoch": 0.4844754231918168,
+      "grad_norm": 0.8793193101882935,
+      "learning_rate": 9.043121845130564e-05,
+      "loss": 2.7668,
+      "step": 5352
+    },
+    {
+      "epoch": 0.48456594550556714,
+      "grad_norm": 0.7916908860206604,
+      "learning_rate": 9.039963128594119e-05,
+      "loss": 2.7605,
+      "step": 5353
+    },
+    {
+      "epoch": 0.4846564678193175,
+      "grad_norm": 0.8754737377166748,
+      "learning_rate": 9.036804508733362e-05,
+      "loss": 2.7812,
+      "step": 5354
+    },
+    {
+      "epoch": 0.4847469901330678,
+      "grad_norm": 0.9203606843948364,
+      "learning_rate": 9.03364598586636e-05,
+      "loss": 2.7841,
+      "step": 5355
+    },
+    {
+      "epoch": 0.48483751244681816,
+      "grad_norm": 0.8462159633636475,
+      "learning_rate": 9.030487560311186e-05,
+      "loss": 2.784,
+      "step": 5356
+    },
+    {
+      "epoch": 0.4849280347605685,
+      "grad_norm": 0.8361531496047974,
+      "learning_rate": 9.027329232385887e-05,
+      "loss": 2.7941,
+      "step": 5357
+    },
+    {
+      "epoch": 0.48501855707431885,
+      "grad_norm": 0.8178485035896301,
+      "learning_rate": 9.024171002408506e-05,
+      "loss": 2.733,
+      "step": 5358
+    },
+    {
+      "epoch": 0.4851090793880692,
+      "grad_norm": 0.9467344284057617,
+      "learning_rate": 9.021012870697077e-05,
+      "loss": 2.8265,
+      "step": 5359
+    },
+    {
+      "epoch": 0.48519960170181947,
+      "grad_norm": 0.8394115567207336,
+      "learning_rate": 9.01785483756963e-05,
+      "loss": 2.8171,
+      "step": 5360
+    },
+    {
+      "epoch": 0.4852901240155698,
+      "grad_norm": 0.869340717792511,
+      "learning_rate": 9.014696903344169e-05,
+      "loss": 2.7675,
+      "step": 5361
+    },
+    {
+      "epoch": 0.48538064632932015,
+      "grad_norm": 0.8268486261367798,
+      "learning_rate": 9.011539068338708e-05,
+      "loss": 2.8036,
+      "step": 5362
+    },
+    {
+      "epoch": 0.4854711686430705,
+      "grad_norm": 0.7936989665031433,
+      "learning_rate": 9.008381332871229e-05,
+      "loss": 2.7107,
+      "step": 5363
+    },
+    {
+      "epoch": 0.48556169095682084,
+      "grad_norm": 0.849716067314148,
+      "learning_rate": 9.005223697259732e-05,
+      "loss": 2.7853,
+      "step": 5364
+    },
+    {
+      "epoch": 0.4856522132705712,
+      "grad_norm": 0.8900785446166992,
+      "learning_rate": 9.002066161822172e-05,
+      "loss": 2.7911,
+      "step": 5365
+    },
+    {
+      "epoch": 0.4857427355843215,
+      "grad_norm": 0.7993195652961731,
+      "learning_rate": 8.998908726876529e-05,
+      "loss": 2.7532,
+      "step": 5366
+    },
+    {
+      "epoch": 0.48583325789807186,
+      "grad_norm": 0.9737759828567505,
+      "learning_rate": 8.995751392740746e-05,
+      "loss": 2.7393,
+      "step": 5367
+    },
+    {
+      "epoch": 0.4859237802118222,
+      "grad_norm": 0.8641330599784851,
+      "learning_rate": 8.992594159732774e-05,
+      "loss": 2.7719,
+      "step": 5368
+    },
+    {
+      "epoch": 0.48601430252557254,
+      "grad_norm": 0.8608652353286743,
+      "learning_rate": 8.989437028170537e-05,
+      "loss": 2.7101,
+      "step": 5369
+    },
+    {
+      "epoch": 0.4861048248393229,
+      "grad_norm": 0.9764891862869263,
+      "learning_rate": 8.986279998371966e-05,
+      "loss": 2.7968,
+      "step": 5370
+    },
+    {
+      "epoch": 0.4861953471530732,
+      "grad_norm": 0.9630980491638184,
+      "learning_rate": 8.983123070654973e-05,
+      "loss": 2.7807,
+      "step": 5371
+    },
+    {
+      "epoch": 0.48628586946682356,
+      "grad_norm": 0.8285449147224426,
+      "learning_rate": 8.979966245337458e-05,
+      "loss": 2.7832,
+      "step": 5372
+    },
+    {
+      "epoch": 0.4863763917805739,
+      "grad_norm": 0.8406493067741394,
+      "learning_rate": 8.976809522737313e-05,
+      "loss": 2.8093,
+      "step": 5373
+    },
+    {
+      "epoch": 0.48646691409432424,
+      "grad_norm": 0.9877937436103821,
+      "learning_rate": 8.973652903172423e-05,
+      "loss": 2.7104,
+      "step": 5374
+    },
+    {
+      "epoch": 0.4865574364080746,
+      "grad_norm": 0.8676995038986206,
+      "learning_rate": 8.970496386960656e-05,
+      "loss": 2.8185,
+      "step": 5375
+    },
+    {
+      "epoch": 0.4866479587218249,
+      "grad_norm": 0.8179798126220703,
+      "learning_rate": 8.967339974419877e-05,
+      "loss": 2.812,
+      "step": 5376
+    },
+    {
+      "epoch": 0.48673848103557527,
+      "grad_norm": 0.8636631965637207,
+      "learning_rate": 8.96418366586793e-05,
+      "loss": 2.7825,
+      "step": 5377
+    },
+    {
+      "epoch": 0.4868290033493256,
+      "grad_norm": 0.9385567307472229,
+      "learning_rate": 8.961027461622664e-05,
+      "loss": 2.7539,
+      "step": 5378
+    },
+    {
+      "epoch": 0.48691952566307595,
+      "grad_norm": 0.8954267501831055,
+      "learning_rate": 8.957871362001902e-05,
+      "loss": 2.7582,
+      "step": 5379
+    },
+    {
+      "epoch": 0.4870100479768263,
+      "grad_norm": 0.869686484336853,
+      "learning_rate": 8.954715367323468e-05,
+      "loss": 2.8182,
+      "step": 5380
+    },
+    {
+      "epoch": 0.48710057029057663,
+      "grad_norm": 0.8564013838768005,
+      "learning_rate": 8.951559477905161e-05,
+      "loss": 2.7847,
+      "step": 5381
+    },
+    {
+      "epoch": 0.48719109260432697,
+      "grad_norm": 0.8536160588264465,
+      "learning_rate": 8.948403694064793e-05,
+      "loss": 2.7588,
+      "step": 5382
+    },
+    {
+      "epoch": 0.4872816149180773,
+      "grad_norm": 0.8505567312240601,
+      "learning_rate": 8.945248016120142e-05,
+      "loss": 2.7828,
+      "step": 5383
+    },
+    {
+      "epoch": 0.48737213723182765,
+      "grad_norm": 0.9134613275527954,
+      "learning_rate": 8.942092444388988e-05,
+      "loss": 2.7442,
+      "step": 5384
+    },
+    {
+      "epoch": 0.487462659545578,
+      "grad_norm": 0.9947112798690796,
+      "learning_rate": 8.938936979189091e-05,
+      "loss": 2.8192,
+      "step": 5385
+    },
+    {
+      "epoch": 0.48755318185932833,
+      "grad_norm": 0.7951664328575134,
+      "learning_rate": 8.935781620838216e-05,
+      "loss": 2.7398,
+      "step": 5386
+    },
+    {
+      "epoch": 0.4876437041730787,
+      "grad_norm": 0.8587552905082703,
+      "learning_rate": 8.932626369654101e-05,
+      "loss": 2.7777,
+      "step": 5387
+    },
+    {
+      "epoch": 0.487734226486829,
+      "grad_norm": 0.9508317708969116,
+      "learning_rate": 8.929471225954484e-05,
+      "loss": 2.7739,
+      "step": 5388
+    },
+    {
+      "epoch": 0.48782474880057936,
+      "grad_norm": 0.8943755030632019,
+      "learning_rate": 8.92631619005708e-05,
+      "loss": 2.8702,
+      "step": 5389
+    },
+    {
+      "epoch": 0.4879152711143297,
+      "grad_norm": 0.8466609716415405,
+      "learning_rate": 8.92316126227961e-05,
+      "loss": 2.7681,
+      "step": 5390
+    },
+    {
+      "epoch": 0.48800579342808004,
+      "grad_norm": 0.9310498833656311,
+      "learning_rate": 8.920006442939772e-05,
+      "loss": 2.7856,
+      "step": 5391
+    },
+    {
+      "epoch": 0.4880963157418304,
+      "grad_norm": 1.0089055299758911,
+      "learning_rate": 8.916851732355255e-05,
+      "loss": 2.8008,
+      "step": 5392
+    },
+    {
+      "epoch": 0.4881868380555807,
+      "grad_norm": 0.9262751936912537,
+      "learning_rate": 8.91369713084374e-05,
+      "loss": 2.7587,
+      "step": 5393
+    },
+    {
+      "epoch": 0.48827736036933106,
+      "grad_norm": 0.9704766869544983,
+      "learning_rate": 8.910542638722889e-05,
+      "loss": 2.7551,
+      "step": 5394
+    },
+    {
+      "epoch": 0.4883678826830814,
+      "grad_norm": 0.8155854344367981,
+      "learning_rate": 8.907388256310373e-05,
+      "loss": 2.7697,
+      "step": 5395
+    },
+    {
+      "epoch": 0.48845840499683174,
+      "grad_norm": 0.9788278341293335,
+      "learning_rate": 8.90423398392382e-05,
+      "loss": 2.8081,
+      "step": 5396
+    },
+    {
+      "epoch": 0.4885489273105821,
+      "grad_norm": 0.9229268431663513,
+      "learning_rate": 8.901079821880882e-05,
+      "loss": 2.7874,
+      "step": 5397
+    },
+    {
+      "epoch": 0.4886394496243324,
+      "grad_norm": 0.8876495361328125,
+      "learning_rate": 8.897925770499174e-05,
+      "loss": 2.7961,
+      "step": 5398
+    },
+    {
+      "epoch": 0.48872997193808276,
+      "grad_norm": 0.9457216262817383,
+      "learning_rate": 8.894771830096312e-05,
+      "loss": 2.7652,
+      "step": 5399
+    },
+    {
+      "epoch": 0.4888204942518331,
+      "grad_norm": 0.8485191464424133,
+      "learning_rate": 8.891618000989891e-05,
+      "loss": 2.7648,
+      "step": 5400
+    },
+    {
+      "epoch": 0.4889110165655834,
+      "grad_norm": 0.8234026432037354,
+      "learning_rate": 8.888464283497512e-05,
+      "loss": 2.731,
+      "step": 5401
+    },
+    {
+      "epoch": 0.48900153887933373,
+      "grad_norm": 0.9092859625816345,
+      "learning_rate": 8.885310677936746e-05,
+      "loss": 2.7226,
+      "step": 5402
+    },
+    {
+      "epoch": 0.48909206119308407,
+      "grad_norm": 0.8186323642730713,
+      "learning_rate": 8.882157184625169e-05,
+      "loss": 2.8257,
+      "step": 5403
+    },
+    {
+      "epoch": 0.4891825835068344,
+      "grad_norm": 0.7837226986885071,
+      "learning_rate": 8.879003803880326e-05,
+      "loss": 2.7632,
+      "step": 5404
+    },
+    {
+      "epoch": 0.48927310582058475,
+      "grad_norm": 0.953040599822998,
+      "learning_rate": 8.875850536019775e-05,
+      "loss": 2.804,
+      "step": 5405
+    },
+    {
+      "epoch": 0.4893636281343351,
+      "grad_norm": 0.921772301197052,
+      "learning_rate": 8.872697381361043e-05,
+      "loss": 2.8101,
+      "step": 5406
+    },
+    {
+      "epoch": 0.48945415044808543,
+      "grad_norm": 0.8325876593589783,
+      "learning_rate": 8.869544340221654e-05,
+      "loss": 2.7615,
+      "step": 5407
+    },
+    {
+      "epoch": 0.4895446727618358,
+      "grad_norm": 0.8860905170440674,
+      "learning_rate": 8.866391412919117e-05,
+      "loss": 2.7417,
+      "step": 5408
+    },
+    {
+      "epoch": 0.4896351950755861,
+      "grad_norm": 0.8813610672950745,
+      "learning_rate": 8.863238599770936e-05,
+      "loss": 2.7837,
+      "step": 5409
+    },
+    {
+      "epoch": 0.48972571738933646,
+      "grad_norm": 0.878919243812561,
+      "learning_rate": 8.860085901094595e-05,
+      "loss": 2.7313,
+      "step": 5410
+    },
+    {
+      "epoch": 0.4898162397030868,
+      "grad_norm": 0.884091854095459,
+      "learning_rate": 8.856933317207576e-05,
+      "loss": 2.7618,
+      "step": 5411
+    },
+    {
+      "epoch": 0.48990676201683714,
+      "grad_norm": 0.848335862159729,
+      "learning_rate": 8.853780848427334e-05,
+      "loss": 2.7053,
+      "step": 5412
+    },
+    {
+      "epoch": 0.4899972843305875,
+      "grad_norm": 0.8469763994216919,
+      "learning_rate": 8.850628495071336e-05,
+      "loss": 2.7734,
+      "step": 5413
+    },
+    {
+      "epoch": 0.4900878066443378,
+      "grad_norm": 0.8686950206756592,
+      "learning_rate": 8.847476257457013e-05,
+      "loss": 2.8046,
+      "step": 5414
+    },
+    {
+      "epoch": 0.49017832895808816,
+      "grad_norm": 1.0004031658172607,
+      "learning_rate": 8.844324135901802e-05,
+      "loss": 2.7847,
+      "step": 5415
+    },
+    {
+      "epoch": 0.4902688512718385,
+      "grad_norm": 0.8817363381385803,
+      "learning_rate": 8.841172130723115e-05,
+      "loss": 2.7698,
+      "step": 5416
+    },
+    {
+      "epoch": 0.49035937358558884,
+      "grad_norm": 0.8560882806777954,
+      "learning_rate": 8.838020242238367e-05,
+      "loss": 2.7825,
+      "step": 5417
+    },
+    {
+      "epoch": 0.4904498958993392,
+      "grad_norm": 0.8987289071083069,
+      "learning_rate": 8.834868470764948e-05,
+      "loss": 2.8044,
+      "step": 5418
+    },
+    {
+      "epoch": 0.4905404182130895,
+      "grad_norm": 0.8843113780021667,
+      "learning_rate": 8.831716816620243e-05,
+      "loss": 2.7462,
+      "step": 5419
+    },
+    {
+      "epoch": 0.49063094052683986,
+      "grad_norm": 0.9686959385871887,
+      "learning_rate": 8.828565280121617e-05,
+      "loss": 2.7268,
+      "step": 5420
+    },
+    {
+      "epoch": 0.4907214628405902,
+      "grad_norm": 0.8693787455558777,
+      "learning_rate": 8.825413861586442e-05,
+      "loss": 2.7575,
+      "step": 5421
+    },
+    {
+      "epoch": 0.49081198515434055,
+      "grad_norm": 0.8458731770515442,
+      "learning_rate": 8.822262561332056e-05,
+      "loss": 2.8095,
+      "step": 5422
+    },
+    {
+      "epoch": 0.4909025074680909,
+      "grad_norm": 0.8771583437919617,
+      "learning_rate": 8.819111379675798e-05,
+      "loss": 2.7571,
+      "step": 5423
+    },
+    {
+      "epoch": 0.49099302978184123,
+      "grad_norm": 0.8951244950294495,
+      "learning_rate": 8.81596031693499e-05,
+      "loss": 2.8171,
+      "step": 5424
+    },
+    {
+      "epoch": 0.49108355209559157,
+      "grad_norm": 0.9194262623786926,
+      "learning_rate": 8.812809373426951e-05,
+      "loss": 2.7763,
+      "step": 5425
+    },
+    {
+      "epoch": 0.4911740744093419,
+      "grad_norm": 1.0411819219589233,
+      "learning_rate": 8.80965854946897e-05,
+      "loss": 2.8314,
+      "step": 5426
+    },
+    {
+      "epoch": 0.49126459672309225,
+      "grad_norm": 0.8811046481132507,
+      "learning_rate": 8.806507845378343e-05,
+      "loss": 2.7847,
+      "step": 5427
+    },
+    {
+      "epoch": 0.4913551190368426,
+      "grad_norm": 0.9744120836257935,
+      "learning_rate": 8.803357261472343e-05,
+      "loss": 2.7625,
+      "step": 5428
+    },
+    {
+      "epoch": 0.49144564135059293,
+      "grad_norm": 0.8727533221244812,
+      "learning_rate": 8.800206798068235e-05,
+      "loss": 2.823,
+      "step": 5429
+    },
+    {
+      "epoch": 0.4915361636643433,
+      "grad_norm": 0.8588023781776428,
+      "learning_rate": 8.797056455483266e-05,
+      "loss": 2.7709,
+      "step": 5430
+    },
+    {
+      "epoch": 0.4916266859780936,
+      "grad_norm": 0.9703391790390015,
+      "learning_rate": 8.793906234034683e-05,
+      "loss": 2.8147,
+      "step": 5431
+    },
+    {
+      "epoch": 0.49171720829184395,
+      "grad_norm": 0.8462010622024536,
+      "learning_rate": 8.790756134039708e-05,
+      "loss": 2.7482,
+      "step": 5432
+    },
+    {
+      "epoch": 0.4918077306055943,
+      "grad_norm": 0.8772428631782532,
+      "learning_rate": 8.787606155815558e-05,
+      "loss": 2.7578,
+      "step": 5433
+    },
+    {
+      "epoch": 0.49189825291934464,
+      "grad_norm": 0.8778265118598938,
+      "learning_rate": 8.784456299679432e-05,
+      "loss": 2.7482,
+      "step": 5434
+    },
+    {
+      "epoch": 0.491988775233095,
+      "grad_norm": 0.8943318724632263,
+      "learning_rate": 8.781306565948528e-05,
+      "loss": 2.7438,
+      "step": 5435
+    },
+    {
+      "epoch": 0.4920792975468453,
+      "grad_norm": 0.8805944323539734,
+      "learning_rate": 8.778156954940016e-05,
+      "loss": 2.7882,
+      "step": 5436
+    },
+    {
+      "epoch": 0.49216981986059566,
+      "grad_norm": 1.0474109649658203,
+      "learning_rate": 8.775007466971067e-05,
+      "loss": 2.7527,
+      "step": 5437
+    },
+    {
+      "epoch": 0.492260342174346,
+      "grad_norm": 0.8754709362983704,
+      "learning_rate": 8.771858102358828e-05,
+      "loss": 2.7693,
+      "step": 5438
+    },
+    {
+      "epoch": 0.49235086448809634,
+      "grad_norm": 0.8586017489433289,
+      "learning_rate": 8.76870886142045e-05,
+      "loss": 2.7932,
+      "step": 5439
+    },
+    {
+      "epoch": 0.4924413868018467,
+      "grad_norm": 0.9212132692337036,
+      "learning_rate": 8.765559744473053e-05,
+      "loss": 2.7714,
+      "step": 5440
+    },
+    {
+      "epoch": 0.492531909115597,
+      "grad_norm": 0.9377496838569641,
+      "learning_rate": 8.762410751833759e-05,
+      "loss": 2.8411,
+      "step": 5441
+    },
+    {
+      "epoch": 0.4926224314293473,
+      "grad_norm": 0.8290513157844543,
+      "learning_rate": 8.759261883819664e-05,
+      "loss": 2.7639,
+      "step": 5442
+    },
+    {
+      "epoch": 0.49271295374309765,
+      "grad_norm": 0.8210857510566711,
+      "learning_rate": 8.756113140747867e-05,
+      "loss": 2.7487,
+      "step": 5443
+    },
+    {
+      "epoch": 0.492803476056848,
+      "grad_norm": 0.9162151217460632,
+      "learning_rate": 8.752964522935441e-05,
+      "loss": 2.8344,
+      "step": 5444
+    },
+    {
+      "epoch": 0.49289399837059833,
+      "grad_norm": 0.8273245096206665,
+      "learning_rate": 8.749816030699456e-05,
+      "loss": 2.7405,
+      "step": 5445
+    },
+    {
+      "epoch": 0.49298452068434867,
+      "grad_norm": 0.8682637214660645,
+      "learning_rate": 8.746667664356956e-05,
+      "loss": 2.7461,
+      "step": 5446
+    },
+    {
+      "epoch": 0.493075042998099,
+      "grad_norm": 0.8884209990501404,
+      "learning_rate": 8.743519424224995e-05,
+      "loss": 2.8145,
+      "step": 5447
+    },
+    {
+      "epoch": 0.49316556531184935,
+      "grad_norm": 0.9339919090270996,
+      "learning_rate": 8.74037131062059e-05,
+      "loss": 2.7779,
+      "step": 5448
+    },
+    {
+      "epoch": 0.4932560876255997,
+      "grad_norm": 0.9694758653640747,
+      "learning_rate": 8.737223323860762e-05,
+      "loss": 2.7847,
+      "step": 5449
+    },
+    {
+      "epoch": 0.49334660993935003,
+      "grad_norm": 0.7979462146759033,
+      "learning_rate": 8.734075464262507e-05,
+      "loss": 2.7319,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4934371322531004,
+      "grad_norm": 0.8110317587852478,
+      "learning_rate": 8.730927732142822e-05,
+      "loss": 2.6897,
+      "step": 5451
+    },
+    {
+      "epoch": 0.4935276545668507,
+      "grad_norm": 0.8965287208557129,
+      "learning_rate": 8.727780127818677e-05,
+      "loss": 2.6903,
+      "step": 5452
+    },
+    {
+      "epoch": 0.49361817688060106,
+      "grad_norm": 0.9517734050750732,
+      "learning_rate": 8.72463265160704e-05,
+      "loss": 2.8256,
+      "step": 5453
+    },
+    {
+      "epoch": 0.4937086991943514,
+      "grad_norm": 0.892761766910553,
+      "learning_rate": 8.721485303824857e-05,
+      "loss": 2.6702,
+      "step": 5454
+    },
+    {
+      "epoch": 0.49379922150810174,
+      "grad_norm": 0.9610890746116638,
+      "learning_rate": 8.718338084789072e-05,
+      "loss": 2.7717,
+      "step": 5455
+    },
+    {
+      "epoch": 0.4938897438218521,
+      "grad_norm": 0.9610509872436523,
+      "learning_rate": 8.715190994816607e-05,
+      "loss": 2.7269,
+      "step": 5456
+    },
+    {
+      "epoch": 0.4939802661356024,
+      "grad_norm": 0.9475899338722229,
+      "learning_rate": 8.712044034224374e-05,
+      "loss": 2.8009,
+      "step": 5457
+    },
+    {
+      "epoch": 0.49407078844935276,
+      "grad_norm": 0.93819659948349,
+      "learning_rate": 8.708897203329267e-05,
+      "loss": 2.785,
+      "step": 5458
+    },
+    {
+      "epoch": 0.4941613107631031,
+      "grad_norm": 0.957547128200531,
+      "learning_rate": 8.705750502448183e-05,
+      "loss": 2.7372,
+      "step": 5459
+    },
+    {
+      "epoch": 0.49425183307685344,
+      "grad_norm": 1.0170553922653198,
+      "learning_rate": 8.702603931897982e-05,
+      "loss": 2.7857,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4943423553906038,
+      "grad_norm": 0.9468273520469666,
+      "learning_rate": 8.699457491995534e-05,
+      "loss": 2.8072,
+      "step": 5461
+    },
+    {
+      "epoch": 0.4944328777043541,
+      "grad_norm": 0.9994382858276367,
+      "learning_rate": 8.696311183057678e-05,
+      "loss": 2.7599,
+      "step": 5462
+    },
+    {
+      "epoch": 0.49452340001810446,
+      "grad_norm": 0.8759012818336487,
+      "learning_rate": 8.693165005401254e-05,
+      "loss": 2.7841,
+      "step": 5463
+    },
+    {
+      "epoch": 0.4946139223318548,
+      "grad_norm": 0.894015908241272,
+      "learning_rate": 8.690018959343072e-05,
+      "loss": 2.7601,
+      "step": 5464
+    },
+    {
+      "epoch": 0.49470444464560515,
+      "grad_norm": 0.8741127848625183,
+      "learning_rate": 8.68687304519995e-05,
+      "loss": 2.7864,
+      "step": 5465
+    },
+    {
+      "epoch": 0.4947949669593555,
+      "grad_norm": 0.8510367274284363,
+      "learning_rate": 8.683727263288673e-05,
+      "loss": 2.737,
+      "step": 5466
+    },
+    {
+      "epoch": 0.4948854892731058,
+      "grad_norm": 0.8362459540367126,
+      "learning_rate": 8.680581613926029e-05,
+      "loss": 2.8298,
+      "step": 5467
+    },
+    {
+      "epoch": 0.49497601158685617,
+      "grad_norm": 0.9130518436431885,
+      "learning_rate": 8.677436097428775e-05,
+      "loss": 2.7198,
+      "step": 5468
+    },
+    {
+      "epoch": 0.4950665339006065,
+      "grad_norm": 0.9414628148078918,
+      "learning_rate": 8.674290714113675e-05,
+      "loss": 2.8174,
+      "step": 5469
+    },
+    {
+      "epoch": 0.49515705621435685,
+      "grad_norm": 0.8726707696914673,
+      "learning_rate": 8.67114546429746e-05,
+      "loss": 2.6835,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4952475785281072,
+      "grad_norm": 0.8614273071289062,
+      "learning_rate": 8.668000348296862e-05,
+      "loss": 2.7743,
+      "step": 5471
+    },
+    {
+      "epoch": 0.49533810084185753,
+      "grad_norm": 0.8497290015220642,
+      "learning_rate": 8.66485536642859e-05,
+      "loss": 2.7286,
+      "step": 5472
+    },
+    {
+      "epoch": 0.4954286231556079,
+      "grad_norm": 0.8182339072227478,
+      "learning_rate": 8.661710519009351e-05,
+      "loss": 2.7595,
+      "step": 5473
+    },
+    {
+      "epoch": 0.4955191454693582,
+      "grad_norm": 0.8943324089050293,
+      "learning_rate": 8.658565806355823e-05,
+      "loss": 2.7802,
+      "step": 5474
+    },
+    {
+      "epoch": 0.49560966778310855,
+      "grad_norm": 0.8168202042579651,
+      "learning_rate": 8.655421228784683e-05,
+      "loss": 2.7348,
+      "step": 5475
+    },
+    {
+      "epoch": 0.4957001900968589,
+      "grad_norm": 0.9240425825119019,
+      "learning_rate": 8.652276786612584e-05,
+      "loss": 2.7805,
+      "step": 5476
+    },
+    {
+      "epoch": 0.49579071241060924,
+      "grad_norm": 0.9528402090072632,
+      "learning_rate": 8.649132480156181e-05,
+      "loss": 2.7187,
+      "step": 5477
+    },
+    {
+      "epoch": 0.4958812347243596,
+      "grad_norm": 0.835676908493042,
+      "learning_rate": 8.6459883097321e-05,
+      "loss": 2.7786,
+      "step": 5478
+    },
+    {
+      "epoch": 0.4959717570381099,
+      "grad_norm": 0.9240402579307556,
+      "learning_rate": 8.642844275656957e-05,
+      "loss": 2.8179,
+      "step": 5479
+    },
+    {
+      "epoch": 0.49606227935186026,
+      "grad_norm": 1.0338637828826904,
+      "learning_rate": 8.639700378247361e-05,
+      "loss": 2.7696,
+      "step": 5480
+    },
+    {
+      "epoch": 0.4961528016656106,
+      "grad_norm": 0.9106727242469788,
+      "learning_rate": 8.636556617819896e-05,
+      "loss": 2.7819,
+      "step": 5481
+    },
+    {
+      "epoch": 0.49624332397936094,
+      "grad_norm": 0.9276056289672852,
+      "learning_rate": 8.633412994691144e-05,
+      "loss": 2.7896,
+      "step": 5482
+    },
+    {
+      "epoch": 0.4963338462931112,
+      "grad_norm": 0.92255699634552,
+      "learning_rate": 8.630269509177666e-05,
+      "loss": 2.7433,
+      "step": 5483
+    },
+    {
+      "epoch": 0.49642436860686157,
+      "grad_norm": 0.8409753441810608,
+      "learning_rate": 8.627126161596012e-05,
+      "loss": 2.7577,
+      "step": 5484
+    },
+    {
+      "epoch": 0.4965148909206119,
+      "grad_norm": 0.9331403374671936,
+      "learning_rate": 8.623982952262713e-05,
+      "loss": 2.7238,
+      "step": 5485
+    },
+    {
+      "epoch": 0.49660541323436225,
+      "grad_norm": 0.8792486190795898,
+      "learning_rate": 8.620839881494296e-05,
+      "loss": 2.7781,
+      "step": 5486
+    },
+    {
+      "epoch": 0.4966959355481126,
+      "grad_norm": 0.9217527508735657,
+      "learning_rate": 8.617696949607264e-05,
+      "loss": 2.7072,
+      "step": 5487
+    },
+    {
+      "epoch": 0.49678645786186293,
+      "grad_norm": 0.8589668273925781,
+      "learning_rate": 8.614554156918113e-05,
+      "loss": 2.6873,
+      "step": 5488
+    },
+    {
+      "epoch": 0.49687698017561327,
+      "grad_norm": 0.879298210144043,
+      "learning_rate": 8.611411503743319e-05,
+      "loss": 2.7492,
+      "step": 5489
+    },
+    {
+      "epoch": 0.4969675024893636,
+      "grad_norm": 0.8220292329788208,
+      "learning_rate": 8.608268990399349e-05,
+      "loss": 2.7402,
+      "step": 5490
+    },
+    {
+      "epoch": 0.49705802480311395,
+      "grad_norm": 0.840372622013092,
+      "learning_rate": 8.60512661720265e-05,
+      "loss": 2.7803,
+      "step": 5491
+    },
+    {
+      "epoch": 0.4971485471168643,
+      "grad_norm": 0.9626206755638123,
+      "learning_rate": 8.601984384469668e-05,
+      "loss": 2.6962,
+      "step": 5492
+    },
+    {
+      "epoch": 0.49723906943061463,
+      "grad_norm": 0.8860694169998169,
+      "learning_rate": 8.598842292516817e-05,
+      "loss": 2.7389,
+      "step": 5493
+    },
+    {
+      "epoch": 0.497329591744365,
+      "grad_norm": 0.9730173945426941,
+      "learning_rate": 8.595700341660511e-05,
+      "loss": 2.8225,
+      "step": 5494
+    },
+    {
+      "epoch": 0.4974201140581153,
+      "grad_norm": 0.8305181264877319,
+      "learning_rate": 8.592558532217139e-05,
+      "loss": 2.8373,
+      "step": 5495
+    },
+    {
+      "epoch": 0.49751063637186566,
+      "grad_norm": 0.8341726660728455,
+      "learning_rate": 8.589416864503087e-05,
+      "loss": 2.7348,
+      "step": 5496
+    },
+    {
+      "epoch": 0.497601158685616,
+      "grad_norm": 0.8894588351249695,
+      "learning_rate": 8.586275338834718e-05,
+      "loss": 2.7598,
+      "step": 5497
+    },
+    {
+      "epoch": 0.49769168099936634,
+      "grad_norm": 0.857745885848999,
+      "learning_rate": 8.583133955528383e-05,
+      "loss": 2.7502,
+      "step": 5498
+    },
+    {
+      "epoch": 0.4977822033131167,
+      "grad_norm": 0.8845468759536743,
+      "learning_rate": 8.579992714900417e-05,
+      "loss": 2.7651,
+      "step": 5499
+    },
+    {
+      "epoch": 0.497872725626867,
+      "grad_norm": 0.8758352398872375,
+      "learning_rate": 8.57685161726715e-05,
+      "loss": 2.7849,
+      "step": 5500
+    },
+    {
+      "epoch": 0.49796324794061736,
+      "grad_norm": 0.9248115420341492,
+      "learning_rate": 8.573710662944885e-05,
+      "loss": 2.7983,
+      "step": 5501
+    },
+    {
+      "epoch": 0.4980537702543677,
+      "grad_norm": 0.873688817024231,
+      "learning_rate": 8.57056985224992e-05,
+      "loss": 2.7343,
+      "step": 5502
+    },
+    {
+      "epoch": 0.49814429256811804,
+      "grad_norm": 1.0141209363937378,
+      "learning_rate": 8.567429185498524e-05,
+      "loss": 2.7702,
+      "step": 5503
+    },
+    {
+      "epoch": 0.4982348148818684,
+      "grad_norm": 0.8634735941886902,
+      "learning_rate": 8.564288663006977e-05,
+      "loss": 2.7663,
+      "step": 5504
+    },
+    {
+      "epoch": 0.4983253371956187,
+      "grad_norm": 1.0201530456542969,
+      "learning_rate": 8.56114828509152e-05,
+      "loss": 2.732,
+      "step": 5505
+    },
+    {
+      "epoch": 0.49841585950936906,
+      "grad_norm": 1.0170044898986816,
+      "learning_rate": 8.558008052068392e-05,
+      "loss": 2.7937,
+      "step": 5506
+    },
+    {
+      "epoch": 0.4985063818231194,
+      "grad_norm": 0.8283863663673401,
+      "learning_rate": 8.55486796425381e-05,
+      "loss": 2.7991,
+      "step": 5507
+    },
+    {
+      "epoch": 0.49859690413686975,
+      "grad_norm": 0.8377839922904968,
+      "learning_rate": 8.551728021963989e-05,
+      "loss": 2.741,
+      "step": 5508
+    },
+    {
+      "epoch": 0.4986874264506201,
+      "grad_norm": 0.9424071311950684,
+      "learning_rate": 8.548588225515115e-05,
+      "loss": 2.6981,
+      "step": 5509
+    },
+    {
+      "epoch": 0.4987779487643704,
+      "grad_norm": 1.026139259338379,
+      "learning_rate": 8.545448575223368e-05,
+      "loss": 2.7906,
+      "step": 5510
+    },
+    {
+      "epoch": 0.49886847107812077,
+      "grad_norm": 0.9049829840660095,
+      "learning_rate": 8.542309071404906e-05,
+      "loss": 2.7607,
+      "step": 5511
+    },
+    {
+      "epoch": 0.4989589933918711,
+      "grad_norm": 0.8872949481010437,
+      "learning_rate": 8.539169714375885e-05,
+      "loss": 2.7427,
+      "step": 5512
+    },
+    {
+      "epoch": 0.49904951570562145,
+      "grad_norm": 0.9735966324806213,
+      "learning_rate": 8.536030504452432e-05,
+      "loss": 2.7462,
+      "step": 5513
+    },
+    {
+      "epoch": 0.4991400380193718,
+      "grad_norm": 0.8961998820304871,
+      "learning_rate": 8.532891441950667e-05,
+      "loss": 2.7231,
+      "step": 5514
+    },
+    {
+      "epoch": 0.49923056033312213,
+      "grad_norm": 0.8910137414932251,
+      "learning_rate": 8.529752527186693e-05,
+      "loss": 2.7965,
+      "step": 5515
+    },
+    {
+      "epoch": 0.49932108264687247,
+      "grad_norm": 0.8779480457305908,
+      "learning_rate": 8.526613760476602e-05,
+      "loss": 2.7419,
+      "step": 5516
+    },
+    {
+      "epoch": 0.4994116049606228,
+      "grad_norm": 0.9542471766471863,
+      "learning_rate": 8.523475142136463e-05,
+      "loss": 2.7206,
+      "step": 5517
+    },
+    {
+      "epoch": 0.49950212727437315,
+      "grad_norm": 0.8974640369415283,
+      "learning_rate": 8.520336672482338e-05,
+      "loss": 2.7817,
+      "step": 5518
+    },
+    {
+      "epoch": 0.4995926495881235,
+      "grad_norm": 0.9307314157485962,
+      "learning_rate": 8.517198351830265e-05,
+      "loss": 2.7546,
+      "step": 5519
+    },
+    {
+      "epoch": 0.49968317190187383,
+      "grad_norm": 0.8345010876655579,
+      "learning_rate": 8.514060180496285e-05,
+      "loss": 2.7418,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4997736942156242,
+      "grad_norm": 0.8099145293235779,
+      "learning_rate": 8.510922158796398e-05,
+      "loss": 2.6789,
+      "step": 5521
+    },
+    {
+      "epoch": 0.4998642165293745,
+      "grad_norm": 0.9911144375801086,
+      "learning_rate": 8.507784287046612e-05,
+      "loss": 2.7712,
+      "step": 5522
+    },
+    {
+      "epoch": 0.49995473884312486,
+      "grad_norm": 0.8634917140007019,
+      "learning_rate": 8.504646565562906e-05,
+      "loss": 2.762,
+      "step": 5523
+    },
+    {
+      "epoch": 0.5000452611568752,
+      "grad_norm": 0.8461886644363403,
+      "learning_rate": 8.501508994661251e-05,
+      "loss": 2.7835,
+      "step": 5524
+    },
+    {
+      "epoch": 0.5001357834706255,
+      "grad_norm": 0.9123673439025879,
+      "learning_rate": 8.498371574657596e-05,
+      "loss": 2.7487,
+      "step": 5525
+    },
+    {
+      "epoch": 0.5002263057843759,
+      "grad_norm": 0.9670429825782776,
+      "learning_rate": 8.495234305867887e-05,
+      "loss": 2.712,
+      "step": 5526
+    },
+    {
+      "epoch": 0.5003168280981262,
+      "grad_norm": 0.7825371026992798,
+      "learning_rate": 8.49209718860804e-05,
+      "loss": 2.7777,
+      "step": 5527
+    },
+    {
+      "epoch": 0.5004073504118766,
+      "grad_norm": 0.8448870778083801,
+      "learning_rate": 8.488960223193967e-05,
+      "loss": 2.7192,
+      "step": 5528
+    },
+    {
+      "epoch": 0.5004978727256268,
+      "grad_norm": 0.9148772358894348,
+      "learning_rate": 8.485823409941553e-05,
+      "loss": 2.7715,
+      "step": 5529
+    },
+    {
+      "epoch": 0.5005883950393772,
+      "grad_norm": 0.8517109751701355,
+      "learning_rate": 8.482686749166686e-05,
+      "loss": 2.7457,
+      "step": 5530
+    },
+    {
+      "epoch": 0.5006789173531275,
+      "grad_norm": 0.9205785393714905,
+      "learning_rate": 8.479550241185218e-05,
+      "loss": 2.7768,
+      "step": 5531
+    },
+    {
+      "epoch": 0.5007694396668779,
+      "grad_norm": 0.8705542087554932,
+      "learning_rate": 8.476413886313001e-05,
+      "loss": 2.8299,
+      "step": 5532
+    },
+    {
+      "epoch": 0.5008599619806282,
+      "grad_norm": 0.8568559885025024,
+      "learning_rate": 8.473277684865862e-05,
+      "loss": 2.7363,
+      "step": 5533
+    },
+    {
+      "epoch": 0.5009504842943786,
+      "grad_norm": 0.8830341696739197,
+      "learning_rate": 8.47014163715962e-05,
+      "loss": 2.8005,
+      "step": 5534
+    },
+    {
+      "epoch": 0.5010410066081289,
+      "grad_norm": 0.8620105981826782,
+      "learning_rate": 8.467005743510073e-05,
+      "loss": 2.7425,
+      "step": 5535
+    },
+    {
+      "epoch": 0.5011315289218793,
+      "grad_norm": 0.8455480933189392,
+      "learning_rate": 8.463870004233008e-05,
+      "loss": 2.7436,
+      "step": 5536
+    },
+    {
+      "epoch": 0.5012220512356296,
+      "grad_norm": 0.8549380898475647,
+      "learning_rate": 8.460734419644185e-05,
+      "loss": 2.7569,
+      "step": 5537
+    },
+    {
+      "epoch": 0.50131257354938,
+      "grad_norm": 0.8382590413093567,
+      "learning_rate": 8.45759899005937e-05,
+      "loss": 2.7851,
+      "step": 5538
+    },
+    {
+      "epoch": 0.5014030958631303,
+      "grad_norm": 0.9025892615318298,
+      "learning_rate": 8.454463715794292e-05,
+      "loss": 2.7553,
+      "step": 5539
+    },
+    {
+      "epoch": 0.5014936181768807,
+      "grad_norm": 0.9311736226081848,
+      "learning_rate": 8.451328597164679e-05,
+      "loss": 2.8024,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5015841404906309,
+      "grad_norm": 0.8533101677894592,
+      "learning_rate": 8.448193634486226e-05,
+      "loss": 2.716,
+      "step": 5541
+    },
+    {
+      "epoch": 0.5016746628043813,
+      "grad_norm": 0.8735119104385376,
+      "learning_rate": 8.445058828074639e-05,
+      "loss": 2.7105,
+      "step": 5542
+    },
+    {
+      "epoch": 0.5017651851181316,
+      "grad_norm": 0.9033412337303162,
+      "learning_rate": 8.441924178245581e-05,
+      "loss": 2.7299,
+      "step": 5543
+    },
+    {
+      "epoch": 0.501855707431882,
+      "grad_norm": 0.8257886171340942,
+      "learning_rate": 8.438789685314721e-05,
+      "loss": 2.7265,
+      "step": 5544
+    },
+    {
+      "epoch": 0.5019462297456323,
+      "grad_norm": 0.9109305739402771,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 2.7337,
+      "step": 5545
+    },
+    {
+      "epoch": 0.5020367520593826,
+      "grad_norm": 0.8924078345298767,
+      "learning_rate": 8.432521171410128e-05,
+      "loss": 2.7947,
+      "step": 5546
+    },
+    {
+      "epoch": 0.502127274373133,
+      "grad_norm": 0.9318864345550537,
+      "learning_rate": 8.429387151067638e-05,
+      "loss": 2.7178,
+      "step": 5547
+    },
+    {
+      "epoch": 0.5022177966868833,
+      "grad_norm": 0.8161680102348328,
+      "learning_rate": 8.426253288885822e-05,
+      "loss": 2.7857,
+      "step": 5548
+    },
+    {
+      "epoch": 0.5023083190006337,
+      "grad_norm": 0.7949011921882629,
+      "learning_rate": 8.423119585180252e-05,
+      "loss": 2.7455,
+      "step": 5549
+    },
+    {
+      "epoch": 0.502398841314384,
+      "grad_norm": 0.8968280553817749,
+      "learning_rate": 8.4199860402665e-05,
+      "loss": 2.7814,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5024893636281343,
+      "grad_norm": 0.8965038061141968,
+      "learning_rate": 8.41685265446011e-05,
+      "loss": 2.7532,
+      "step": 5551
+    },
+    {
+      "epoch": 0.5025798859418846,
+      "grad_norm": 0.8199223875999451,
+      "learning_rate": 8.413719428076617e-05,
+      "loss": 2.7675,
+      "step": 5552
+    },
+    {
+      "epoch": 0.502670408255635,
+      "grad_norm": 0.8846849203109741,
+      "learning_rate": 8.410586361431528e-05,
+      "loss": 2.7393,
+      "step": 5553
+    },
+    {
+      "epoch": 0.5027609305693853,
+      "grad_norm": 0.895946741104126,
+      "learning_rate": 8.407453454840357e-05,
+      "loss": 2.8227,
+      "step": 5554
+    },
+    {
+      "epoch": 0.5028514528831357,
+      "grad_norm": 0.8147122859954834,
+      "learning_rate": 8.404320708618572e-05,
+      "loss": 2.7743,
+      "step": 5555
+    },
+    {
+      "epoch": 0.502941975196886,
+      "grad_norm": 0.7956191897392273,
+      "learning_rate": 8.401188123081653e-05,
+      "loss": 2.7715,
+      "step": 5556
+    },
+    {
+      "epoch": 0.5030324975106364,
+      "grad_norm": 0.9651154279708862,
+      "learning_rate": 8.398055698545043e-05,
+      "loss": 2.7298,
+      "step": 5557
+    },
+    {
+      "epoch": 0.5031230198243867,
+      "grad_norm": 0.9466992020606995,
+      "learning_rate": 8.394923435324183e-05,
+      "loss": 2.7663,
+      "step": 5558
+    },
+    {
+      "epoch": 0.5032135421381371,
+      "grad_norm": 0.9539802074432373,
+      "learning_rate": 8.391791333734484e-05,
+      "loss": 2.7655,
+      "step": 5559
+    },
+    {
+      "epoch": 0.5033040644518874,
+      "grad_norm": 0.9526113271713257,
+      "learning_rate": 8.38865939409136e-05,
+      "loss": 2.6986,
+      "step": 5560
+    },
+    {
+      "epoch": 0.5033945867656378,
+      "grad_norm": 0.8260806798934937,
+      "learning_rate": 8.385527616710188e-05,
+      "loss": 2.7494,
+      "step": 5561
+    },
+    {
+      "epoch": 0.503485109079388,
+      "grad_norm": 0.8653164505958557,
+      "learning_rate": 8.382396001906343e-05,
+      "loss": 2.7554,
+      "step": 5562
+    },
+    {
+      "epoch": 0.5035756313931384,
+      "grad_norm": 0.9280028939247131,
+      "learning_rate": 8.379264549995176e-05,
+      "loss": 2.776,
+      "step": 5563
+    },
+    {
+      "epoch": 0.5036661537068887,
+      "grad_norm": 0.8204590678215027,
+      "learning_rate": 8.37613326129202e-05,
+      "loss": 2.7261,
+      "step": 5564
+    },
+    {
+      "epoch": 0.5037566760206391,
+      "grad_norm": 0.8759326934814453,
+      "learning_rate": 8.373002136112205e-05,
+      "loss": 2.7027,
+      "step": 5565
+    },
+    {
+      "epoch": 0.5038471983343894,
+      "grad_norm": 0.8595099449157715,
+      "learning_rate": 8.369871174771027e-05,
+      "loss": 2.7887,
+      "step": 5566
+    },
+    {
+      "epoch": 0.5039377206481398,
+      "grad_norm": 0.8577147722244263,
+      "learning_rate": 8.366740377583781e-05,
+      "loss": 2.7618,
+      "step": 5567
+    },
+    {
+      "epoch": 0.5040282429618901,
+      "grad_norm": 0.8850293159484863,
+      "learning_rate": 8.363609744865728e-05,
+      "loss": 2.7329,
+      "step": 5568
+    },
+    {
+      "epoch": 0.5041187652756405,
+      "grad_norm": 0.8491239547729492,
+      "learning_rate": 8.360479276932135e-05,
+      "loss": 2.6908,
+      "step": 5569
+    },
+    {
+      "epoch": 0.5042092875893908,
+      "grad_norm": 0.9252224564552307,
+      "learning_rate": 8.357348974098231e-05,
+      "loss": 2.7958,
+      "step": 5570
+    },
+    {
+      "epoch": 0.5042998099031412,
+      "grad_norm": 0.8544613122940063,
+      "learning_rate": 8.354218836679243e-05,
+      "loss": 2.7999,
+      "step": 5571
+    },
+    {
+      "epoch": 0.5043903322168914,
+      "grad_norm": 0.9366018176078796,
+      "learning_rate": 8.351088864990368e-05,
+      "loss": 2.8178,
+      "step": 5572
+    },
+    {
+      "epoch": 0.5044808545306418,
+      "grad_norm": 0.8889404535293579,
+      "learning_rate": 8.347959059346804e-05,
+      "loss": 2.7446,
+      "step": 5573
+    },
+    {
+      "epoch": 0.5045713768443921,
+      "grad_norm": 0.8375549912452698,
+      "learning_rate": 8.344829420063715e-05,
+      "loss": 2.7979,
+      "step": 5574
+    },
+    {
+      "epoch": 0.5046618991581425,
+      "grad_norm": 0.8302408456802368,
+      "learning_rate": 8.34169994745626e-05,
+      "loss": 2.7855,
+      "step": 5575
+    },
+    {
+      "epoch": 0.5047524214718928,
+      "grad_norm": 0.8453435301780701,
+      "learning_rate": 8.338570641839573e-05,
+      "loss": 2.775,
+      "step": 5576
+    },
+    {
+      "epoch": 0.5048429437856432,
+      "grad_norm": 0.8431276082992554,
+      "learning_rate": 8.33544150352878e-05,
+      "loss": 2.7928,
+      "step": 5577
+    },
+    {
+      "epoch": 0.5049334660993935,
+      "grad_norm": 0.7774648666381836,
+      "learning_rate": 8.332312532838978e-05,
+      "loss": 2.7188,
+      "step": 5578
+    },
+    {
+      "epoch": 0.5050239884131439,
+      "grad_norm": 1.0299315452575684,
+      "learning_rate": 8.329183730085262e-05,
+      "loss": 2.6806,
+      "step": 5579
+    },
+    {
+      "epoch": 0.5051145107268942,
+      "grad_norm": 0.9037138819694519,
+      "learning_rate": 8.326055095582694e-05,
+      "loss": 2.7802,
+      "step": 5580
+    },
+    {
+      "epoch": 0.5052050330406446,
+      "grad_norm": 0.9699219465255737,
+      "learning_rate": 8.32292662964634e-05,
+      "loss": 2.7911,
+      "step": 5581
+    },
+    {
+      "epoch": 0.5052955553543949,
+      "grad_norm": 0.9145121574401855,
+      "learning_rate": 8.31979833259122e-05,
+      "loss": 2.7354,
+      "step": 5582
+    },
+    {
+      "epoch": 0.5053860776681453,
+      "grad_norm": 0.854351282119751,
+      "learning_rate": 8.316670204732366e-05,
+      "loss": 2.7455,
+      "step": 5583
+    },
+    {
+      "epoch": 0.5054765999818955,
+      "grad_norm": 0.9479194283485413,
+      "learning_rate": 8.313542246384775e-05,
+      "loss": 2.8164,
+      "step": 5584
+    },
+    {
+      "epoch": 0.5055671222956459,
+      "grad_norm": 0.8744751811027527,
+      "learning_rate": 8.310414457863437e-05,
+      "loss": 2.7129,
+      "step": 5585
+    },
+    {
+      "epoch": 0.5056576446093962,
+      "grad_norm": 0.8344112634658813,
+      "learning_rate": 8.307286839483311e-05,
+      "loss": 2.6956,
+      "step": 5586
+    },
+    {
+      "epoch": 0.5057481669231465,
+      "grad_norm": 0.8224936127662659,
+      "learning_rate": 8.304159391559359e-05,
+      "loss": 2.7589,
+      "step": 5587
+    },
+    {
+      "epoch": 0.5058386892368969,
+      "grad_norm": 0.8344510793685913,
+      "learning_rate": 8.301032114406507e-05,
+      "loss": 2.7799,
+      "step": 5588
+    },
+    {
+      "epoch": 0.5059292115506472,
+      "grad_norm": 0.8643276691436768,
+      "learning_rate": 8.297905008339677e-05,
+      "loss": 2.7249,
+      "step": 5589
+    },
+    {
+      "epoch": 0.5060197338643976,
+      "grad_norm": 0.8124856352806091,
+      "learning_rate": 8.294778073673762e-05,
+      "loss": 2.8106,
+      "step": 5590
+    },
+    {
+      "epoch": 0.5061102561781479,
+      "grad_norm": 0.869785487651825,
+      "learning_rate": 8.291651310723654e-05,
+      "loss": 2.7463,
+      "step": 5591
+    },
+    {
+      "epoch": 0.5062007784918983,
+      "grad_norm": 0.9260883927345276,
+      "learning_rate": 8.288524719804208e-05,
+      "loss": 2.7447,
+      "step": 5592
+    },
+    {
+      "epoch": 0.5062913008056485,
+      "grad_norm": 0.9370576739311218,
+      "learning_rate": 8.285398301230278e-05,
+      "loss": 2.6753,
+      "step": 5593
+    },
+    {
+      "epoch": 0.5063818231193989,
+      "grad_norm": 0.7781187891960144,
+      "learning_rate": 8.282272055316688e-05,
+      "loss": 2.7359,
+      "step": 5594
+    },
+    {
+      "epoch": 0.5064723454331492,
+      "grad_norm": 0.8413085341453552,
+      "learning_rate": 8.279145982378261e-05,
+      "loss": 2.7524,
+      "step": 5595
+    },
+    {
+      "epoch": 0.5065628677468996,
+      "grad_norm": 0.852820634841919,
+      "learning_rate": 8.276020082729783e-05,
+      "loss": 2.734,
+      "step": 5596
+    },
+    {
+      "epoch": 0.5066533900606499,
+      "grad_norm": 0.8400428891181946,
+      "learning_rate": 8.272894356686039e-05,
+      "loss": 2.7109,
+      "step": 5597
+    },
+    {
+      "epoch": 0.5067439123744003,
+      "grad_norm": 0.9511899352073669,
+      "learning_rate": 8.269768804561779e-05,
+      "loss": 2.7996,
+      "step": 5598
+    },
+    {
+      "epoch": 0.5068344346881506,
+      "grad_norm": 0.8015035390853882,
+      "learning_rate": 8.26664342667176e-05,
+      "loss": 2.7499,
+      "step": 5599
+    },
+    {
+      "epoch": 0.506924957001901,
+      "grad_norm": 0.83297199010849,
+      "learning_rate": 8.263518223330697e-05,
+      "loss": 2.7384,
+      "step": 5600
+    },
+    {
+      "epoch": 0.5070154793156513,
+      "grad_norm": 0.9492606520652771,
+      "learning_rate": 8.260393194853304e-05,
+      "loss": 2.7537,
+      "step": 5601
+    },
+    {
+      "epoch": 0.5071060016294017,
+      "grad_norm": 0.8790477514266968,
+      "learning_rate": 8.257268341554264e-05,
+      "loss": 2.7606,
+      "step": 5602
+    },
+    {
+      "epoch": 0.507196523943152,
+      "grad_norm": 0.8750600814819336,
+      "learning_rate": 8.25414366374826e-05,
+      "loss": 2.7937,
+      "step": 5603
+    },
+    {
+      "epoch": 0.5072870462569024,
+      "grad_norm": 0.8970506191253662,
+      "learning_rate": 8.251019161749938e-05,
+      "loss": 2.82,
+      "step": 5604
+    },
+    {
+      "epoch": 0.5073775685706526,
+      "grad_norm": 0.9380742311477661,
+      "learning_rate": 8.24789483587394e-05,
+      "loss": 2.7346,
+      "step": 5605
+    },
+    {
+      "epoch": 0.507468090884403,
+      "grad_norm": 0.9393888711929321,
+      "learning_rate": 8.244770686434881e-05,
+      "loss": 2.7351,
+      "step": 5606
+    },
+    {
+      "epoch": 0.5075586131981533,
+      "grad_norm": 0.7890758514404297,
+      "learning_rate": 8.24164671374737e-05,
+      "loss": 2.7012,
+      "step": 5607
+    },
+    {
+      "epoch": 0.5076491355119037,
+      "grad_norm": 0.881951630115509,
+      "learning_rate": 8.238522918125983e-05,
+      "loss": 2.8086,
+      "step": 5608
+    },
+    {
+      "epoch": 0.507739657825654,
+      "grad_norm": 1.0359519720077515,
+      "learning_rate": 8.235399299885293e-05,
+      "loss": 2.7548,
+      "step": 5609
+    },
+    {
+      "epoch": 0.5078301801394044,
+      "grad_norm": 0.8619797825813293,
+      "learning_rate": 8.232275859339841e-05,
+      "loss": 2.7076,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5079207024531547,
+      "grad_norm": 0.887519359588623,
+      "learning_rate": 8.229152596804168e-05,
+      "loss": 2.7639,
+      "step": 5611
+    },
+    {
+      "epoch": 0.5080112247669051,
+      "grad_norm": 0.9172592759132385,
+      "learning_rate": 8.226029512592777e-05,
+      "loss": 2.7452,
+      "step": 5612
+    },
+    {
+      "epoch": 0.5081017470806554,
+      "grad_norm": 0.8905028104782104,
+      "learning_rate": 8.222906607020165e-05,
+      "loss": 2.806,
+      "step": 5613
+    },
+    {
+      "epoch": 0.5081922693944058,
+      "grad_norm": 0.8358117341995239,
+      "learning_rate": 8.219783880400808e-05,
+      "loss": 2.7684,
+      "step": 5614
+    },
+    {
+      "epoch": 0.508282791708156,
+      "grad_norm": 0.9052249789237976,
+      "learning_rate": 8.21666133304917e-05,
+      "loss": 2.8095,
+      "step": 5615
+    },
+    {
+      "epoch": 0.5083733140219064,
+      "grad_norm": 0.8885976672172546,
+      "learning_rate": 8.213538965279682e-05,
+      "loss": 2.7564,
+      "step": 5616
+    },
+    {
+      "epoch": 0.5084638363356567,
+      "grad_norm": 0.9161496162414551,
+      "learning_rate": 8.210416777406774e-05,
+      "loss": 2.7672,
+      "step": 5617
+    },
+    {
+      "epoch": 0.5085543586494071,
+      "grad_norm": 0.8568997979164124,
+      "learning_rate": 8.207294769744847e-05,
+      "loss": 2.748,
+      "step": 5618
+    },
+    {
+      "epoch": 0.5086448809631574,
+      "grad_norm": 0.8771458864212036,
+      "learning_rate": 8.204172942608289e-05,
+      "loss": 2.7699,
+      "step": 5619
+    },
+    {
+      "epoch": 0.5087354032769078,
+      "grad_norm": 0.9072162508964539,
+      "learning_rate": 8.201051296311462e-05,
+      "loss": 2.7902,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5088259255906581,
+      "grad_norm": 0.8902173638343811,
+      "learning_rate": 8.197929831168724e-05,
+      "loss": 2.7359,
+      "step": 5621
+    },
+    {
+      "epoch": 0.5089164479044085,
+      "grad_norm": 0.8940975666046143,
+      "learning_rate": 8.194808547494401e-05,
+      "loss": 2.8079,
+      "step": 5622
+    },
+    {
+      "epoch": 0.5090069702181588,
+      "grad_norm": 0.8261607885360718,
+      "learning_rate": 8.19168744560281e-05,
+      "loss": 2.7931,
+      "step": 5623
+    },
+    {
+      "epoch": 0.5090974925319092,
+      "grad_norm": 0.8398805260658264,
+      "learning_rate": 8.18856652580824e-05,
+      "loss": 2.7469,
+      "step": 5624
+    },
+    {
+      "epoch": 0.5091880148456595,
+      "grad_norm": 0.8998643755912781,
+      "learning_rate": 8.185445788424974e-05,
+      "loss": 2.723,
+      "step": 5625
+    },
+    {
+      "epoch": 0.5092785371594098,
+      "grad_norm": 0.9138926863670349,
+      "learning_rate": 8.182325233767267e-05,
+      "loss": 2.7699,
+      "step": 5626
+    },
+    {
+      "epoch": 0.5093690594731601,
+      "grad_norm": 0.8297197818756104,
+      "learning_rate": 8.179204862149361e-05,
+      "loss": 2.7137,
+      "step": 5627
+    },
+    {
+      "epoch": 0.5094595817869104,
+      "grad_norm": 0.8896348476409912,
+      "learning_rate": 8.176084673885471e-05,
+      "loss": 2.7489,
+      "step": 5628
+    },
+    {
+      "epoch": 0.5095501041006608,
+      "grad_norm": 0.9025987386703491,
+      "learning_rate": 8.17296466928981e-05,
+      "loss": 2.7342,
+      "step": 5629
+    },
+    {
+      "epoch": 0.5096406264144111,
+      "grad_norm": 0.890731692314148,
+      "learning_rate": 8.169844848676554e-05,
+      "loss": 2.7439,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5097311487281615,
+      "grad_norm": 0.8023970127105713,
+      "learning_rate": 8.166725212359873e-05,
+      "loss": 2.794,
+      "step": 5631
+    },
+    {
+      "epoch": 0.5098216710419118,
+      "grad_norm": 0.8444201946258545,
+      "learning_rate": 8.16360576065391e-05,
+      "loss": 2.7278,
+      "step": 5632
+    },
+    {
+      "epoch": 0.5099121933556622,
+      "grad_norm": 0.8420068621635437,
+      "learning_rate": 8.160486493872798e-05,
+      "loss": 2.7529,
+      "step": 5633
+    },
+    {
+      "epoch": 0.5100027156694125,
+      "grad_norm": 0.8325995206832886,
+      "learning_rate": 8.157367412330647e-05,
+      "loss": 2.7908,
+      "step": 5634
+    },
+    {
+      "epoch": 0.5100932379831629,
+      "grad_norm": 0.9310572743415833,
+      "learning_rate": 8.154248516341548e-05,
+      "loss": 2.7331,
+      "step": 5635
+    },
+    {
+      "epoch": 0.5101837602969131,
+      "grad_norm": 0.8454694747924805,
+      "learning_rate": 8.151129806219568e-05,
+      "loss": 2.7153,
+      "step": 5636
+    },
+    {
+      "epoch": 0.5102742826106635,
+      "grad_norm": 0.842123806476593,
+      "learning_rate": 8.148011282278772e-05,
+      "loss": 2.79,
+      "step": 5637
+    },
+    {
+      "epoch": 0.5103648049244138,
+      "grad_norm": 0.9366497993469238,
+      "learning_rate": 8.144892944833184e-05,
+      "loss": 2.7858,
+      "step": 5638
+    },
+    {
+      "epoch": 0.5104553272381642,
+      "grad_norm": 0.8815160393714905,
+      "learning_rate": 8.14177479419683e-05,
+      "loss": 2.7287,
+      "step": 5639
+    },
+    {
+      "epoch": 0.5105458495519145,
+      "grad_norm": 0.8443790078163147,
+      "learning_rate": 8.1386568306837e-05,
+      "loss": 2.7358,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5106363718656649,
+      "grad_norm": 0.8315573930740356,
+      "learning_rate": 8.135539054607779e-05,
+      "loss": 2.7245,
+      "step": 5641
+    },
+    {
+      "epoch": 0.5107268941794152,
+      "grad_norm": 0.9345365762710571,
+      "learning_rate": 8.132421466283022e-05,
+      "loss": 2.7289,
+      "step": 5642
+    },
+    {
+      "epoch": 0.5108174164931656,
+      "grad_norm": 0.9540194869041443,
+      "learning_rate": 8.129304066023375e-05,
+      "loss": 2.7142,
+      "step": 5643
+    },
+    {
+      "epoch": 0.5109079388069159,
+      "grad_norm": 0.9122259020805359,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 2.8238,
+      "step": 5644
+    },
+    {
+      "epoch": 0.5109984611206663,
+      "grad_norm": 0.8844560980796814,
+      "learning_rate": 8.123069830955067e-05,
+      "loss": 2.7545,
+      "step": 5645
+    },
+    {
+      "epoch": 0.5110889834344166,
+      "grad_norm": 0.9123935699462891,
+      "learning_rate": 8.119952996774197e-05,
+      "loss": 2.74,
+      "step": 5646
+    },
+    {
+      "epoch": 0.511179505748167,
+      "grad_norm": 0.9042823314666748,
+      "learning_rate": 8.116836351914009e-05,
+      "loss": 2.7527,
+      "step": 5647
+    },
+    {
+      "epoch": 0.5112700280619172,
+      "grad_norm": 0.9140527248382568,
+      "learning_rate": 8.11371989668835e-05,
+      "loss": 2.7742,
+      "step": 5648
+    },
+    {
+      "epoch": 0.5113605503756676,
+      "grad_norm": 0.8494327664375305,
+      "learning_rate": 8.110603631411043e-05,
+      "loss": 2.7112,
+      "step": 5649
+    },
+    {
+      "epoch": 0.5114510726894179,
+      "grad_norm": 0.8976174592971802,
+      "learning_rate": 8.107487556395901e-05,
+      "loss": 2.7035,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5115415950031683,
+      "grad_norm": 0.9411106109619141,
+      "learning_rate": 8.104371671956707e-05,
+      "loss": 2.7734,
+      "step": 5651
+    },
+    {
+      "epoch": 0.5116321173169186,
+      "grad_norm": 0.8943339586257935,
+      "learning_rate": 8.101255978407237e-05,
+      "loss": 2.759,
+      "step": 5652
+    },
+    {
+      "epoch": 0.511722639630669,
+      "grad_norm": 0.8546929955482483,
+      "learning_rate": 8.098140476061236e-05,
+      "loss": 2.7437,
+      "step": 5653
+    },
+    {
+      "epoch": 0.5118131619444193,
+      "grad_norm": 0.8101754784584045,
+      "learning_rate": 8.095025165232439e-05,
+      "loss": 2.7083,
+      "step": 5654
+    },
+    {
+      "epoch": 0.5119036842581697,
+      "grad_norm": 0.8320404291152954,
+      "learning_rate": 8.091910046234552e-05,
+      "loss": 2.7549,
+      "step": 5655
+    },
+    {
+      "epoch": 0.51199420657192,
+      "grad_norm": 0.8851386308670044,
+      "learning_rate": 8.088795119381276e-05,
+      "loss": 2.755,
+      "step": 5656
+    },
+    {
+      "epoch": 0.5120847288856704,
+      "grad_norm": 0.8043539524078369,
+      "learning_rate": 8.085680384986276e-05,
+      "loss": 2.6904,
+      "step": 5657
+    },
+    {
+      "epoch": 0.5121752511994206,
+      "grad_norm": 0.8275747895240784,
+      "learning_rate": 8.08256584336321e-05,
+      "loss": 2.7432,
+      "step": 5658
+    },
+    {
+      "epoch": 0.512265773513171,
+      "grad_norm": 0.8125655055046082,
+      "learning_rate": 8.079451494825709e-05,
+      "loss": 2.7733,
+      "step": 5659
+    },
+    {
+      "epoch": 0.5123562958269213,
+      "grad_norm": 0.8181120157241821,
+      "learning_rate": 8.076337339687394e-05,
+      "loss": 2.7406,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5124468181406717,
+      "grad_norm": 0.8543720841407776,
+      "learning_rate": 8.073223378261852e-05,
+      "loss": 2.7117,
+      "step": 5661
+    },
+    {
+      "epoch": 0.512537340454422,
+      "grad_norm": 0.8811604976654053,
+      "learning_rate": 8.070109610862668e-05,
+      "loss": 2.8013,
+      "step": 5662
+    },
+    {
+      "epoch": 0.5126278627681724,
+      "grad_norm": 0.8130089640617371,
+      "learning_rate": 8.066996037803388e-05,
+      "loss": 2.7559,
+      "step": 5663
+    },
+    {
+      "epoch": 0.5127183850819227,
+      "grad_norm": 0.779761016368866,
+      "learning_rate": 8.063882659397557e-05,
+      "loss": 2.7497,
+      "step": 5664
+    },
+    {
+      "epoch": 0.5128089073956731,
+      "grad_norm": 0.8280618786811829,
+      "learning_rate": 8.060769475958689e-05,
+      "loss": 2.7415,
+      "step": 5665
+    },
+    {
+      "epoch": 0.5128994297094234,
+      "grad_norm": 0.8324514031410217,
+      "learning_rate": 8.057656487800282e-05,
+      "loss": 2.7464,
+      "step": 5666
+    },
+    {
+      "epoch": 0.5129899520231738,
+      "grad_norm": 0.794349193572998,
+      "learning_rate": 8.054543695235811e-05,
+      "loss": 2.7231,
+      "step": 5667
+    },
+    {
+      "epoch": 0.513080474336924,
+      "grad_norm": 0.8479178547859192,
+      "learning_rate": 8.051431098578741e-05,
+      "loss": 2.7078,
+      "step": 5668
+    },
+    {
+      "epoch": 0.5131709966506743,
+      "grad_norm": 0.864479660987854,
+      "learning_rate": 8.048318698142503e-05,
+      "loss": 2.743,
+      "step": 5669
+    },
+    {
+      "epoch": 0.5132615189644247,
+      "grad_norm": 0.8360306024551392,
+      "learning_rate": 8.045206494240521e-05,
+      "loss": 2.7561,
+      "step": 5670
+    },
+    {
+      "epoch": 0.513352041278175,
+      "grad_norm": 0.8737164735794067,
+      "learning_rate": 8.042094487186187e-05,
+      "loss": 2.7586,
+      "step": 5671
+    },
+    {
+      "epoch": 0.5134425635919254,
+      "grad_norm": 0.8805273175239563,
+      "learning_rate": 8.038982677292889e-05,
+      "loss": 2.7258,
+      "step": 5672
+    },
+    {
+      "epoch": 0.5135330859056757,
+      "grad_norm": 0.8032810688018799,
+      "learning_rate": 8.03587106487398e-05,
+      "loss": 2.6993,
+      "step": 5673
+    },
+    {
+      "epoch": 0.5136236082194261,
+      "grad_norm": 0.813574492931366,
+      "learning_rate": 8.032759650242802e-05,
+      "loss": 2.7844,
+      "step": 5674
+    },
+    {
+      "epoch": 0.5137141305331764,
+      "grad_norm": 0.8926960825920105,
+      "learning_rate": 8.029648433712671e-05,
+      "loss": 2.7381,
+      "step": 5675
+    },
+    {
+      "epoch": 0.5138046528469268,
+      "grad_norm": 0.8661999702453613,
+      "learning_rate": 8.026537415596895e-05,
+      "loss": 2.732,
+      "step": 5676
+    },
+    {
+      "epoch": 0.5138951751606771,
+      "grad_norm": 0.7916553616523743,
+      "learning_rate": 8.023426596208739e-05,
+      "loss": 2.7387,
+      "step": 5677
+    },
+    {
+      "epoch": 0.5139856974744275,
+      "grad_norm": 0.8442259430885315,
+      "learning_rate": 8.020315975861476e-05,
+      "loss": 2.7264,
+      "step": 5678
+    },
+    {
+      "epoch": 0.5140762197881777,
+      "grad_norm": 0.8810858130455017,
+      "learning_rate": 8.017205554868337e-05,
+      "loss": 2.7055,
+      "step": 5679
+    },
+    {
+      "epoch": 0.5141667421019281,
+      "grad_norm": 0.8070834279060364,
+      "learning_rate": 8.014095333542548e-05,
+      "loss": 2.7906,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5142572644156784,
+      "grad_norm": 0.8607498407363892,
+      "learning_rate": 8.0109853121973e-05,
+      "loss": 2.7554,
+      "step": 5681
+    },
+    {
+      "epoch": 0.5143477867294288,
+      "grad_norm": 0.8190198540687561,
+      "learning_rate": 8.007875491145781e-05,
+      "loss": 2.7636,
+      "step": 5682
+    },
+    {
+      "epoch": 0.5144383090431791,
+      "grad_norm": 0.832055389881134,
+      "learning_rate": 8.004765870701144e-05,
+      "loss": 2.7015,
+      "step": 5683
+    },
+    {
+      "epoch": 0.5145288313569295,
+      "grad_norm": 0.8482311964035034,
+      "learning_rate": 8.00165645117653e-05,
+      "loss": 2.762,
+      "step": 5684
+    },
+    {
+      "epoch": 0.5146193536706798,
+      "grad_norm": 0.8550591468811035,
+      "learning_rate": 7.998547232885052e-05,
+      "loss": 2.7264,
+      "step": 5685
+    },
+    {
+      "epoch": 0.5147098759844302,
+      "grad_norm": 0.8633624315261841,
+      "learning_rate": 7.995438216139818e-05,
+      "loss": 2.7324,
+      "step": 5686
+    },
+    {
+      "epoch": 0.5148003982981805,
+      "grad_norm": 0.8387537598609924,
+      "learning_rate": 7.992329401253898e-05,
+      "loss": 2.7605,
+      "step": 5687
+    },
+    {
+      "epoch": 0.5148909206119309,
+      "grad_norm": 0.7846046686172485,
+      "learning_rate": 7.989220788540355e-05,
+      "loss": 2.7169,
+      "step": 5688
+    },
+    {
+      "epoch": 0.5149814429256812,
+      "grad_norm": 0.8761247396469116,
+      "learning_rate": 7.986112378312221e-05,
+      "loss": 2.7233,
+      "step": 5689
+    },
+    {
+      "epoch": 0.5150719652394316,
+      "grad_norm": 0.888114333152771,
+      "learning_rate": 7.983004170882518e-05,
+      "loss": 2.7698,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5151624875531818,
+      "grad_norm": 0.8965131044387817,
+      "learning_rate": 7.979896166564237e-05,
+      "loss": 2.8113,
+      "step": 5691
+    },
+    {
+      "epoch": 0.5152530098669322,
+      "grad_norm": 0.8232806921005249,
+      "learning_rate": 7.97678836567036e-05,
+      "loss": 2.7355,
+      "step": 5692
+    },
+    {
+      "epoch": 0.5153435321806825,
+      "grad_norm": 0.8814472556114197,
+      "learning_rate": 7.973680768513835e-05,
+      "loss": 2.7442,
+      "step": 5693
+    },
+    {
+      "epoch": 0.5154340544944329,
+      "grad_norm": 0.8496522307395935,
+      "learning_rate": 7.970573375407604e-05,
+      "loss": 2.703,
+      "step": 5694
+    },
+    {
+      "epoch": 0.5155245768081832,
+      "grad_norm": 0.9013231992721558,
+      "learning_rate": 7.967466186664578e-05,
+      "loss": 2.7507,
+      "step": 5695
+    },
+    {
+      "epoch": 0.5156150991219336,
+      "grad_norm": 0.9093582034111023,
+      "learning_rate": 7.964359202597654e-05,
+      "loss": 2.753,
+      "step": 5696
+    },
+    {
+      "epoch": 0.5157056214356839,
+      "grad_norm": 0.8807817697525024,
+      "learning_rate": 7.961252423519696e-05,
+      "loss": 2.7535,
+      "step": 5697
+    },
+    {
+      "epoch": 0.5157961437494343,
+      "grad_norm": 0.8731459379196167,
+      "learning_rate": 7.958145849743569e-05,
+      "loss": 2.79,
+      "step": 5698
+    },
+    {
+      "epoch": 0.5158866660631846,
+      "grad_norm": 0.8574411273002625,
+      "learning_rate": 7.955039481582097e-05,
+      "loss": 2.6913,
+      "step": 5699
+    },
+    {
+      "epoch": 0.515977188376935,
+      "grad_norm": 0.8288585543632507,
+      "learning_rate": 7.951933319348095e-05,
+      "loss": 2.7044,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5160677106906852,
+      "grad_norm": 0.8653036952018738,
+      "learning_rate": 7.948827363354347e-05,
+      "loss": 2.6882,
+      "step": 5701
+    },
+    {
+      "epoch": 0.5161582330044356,
+      "grad_norm": 0.818445086479187,
+      "learning_rate": 7.945721613913633e-05,
+      "loss": 2.7755,
+      "step": 5702
+    },
+    {
+      "epoch": 0.5162487553181859,
+      "grad_norm": 0.7889580726623535,
+      "learning_rate": 7.942616071338693e-05,
+      "loss": 2.786,
+      "step": 5703
+    },
+    {
+      "epoch": 0.5163392776319363,
+      "grad_norm": 0.875316858291626,
+      "learning_rate": 7.939510735942262e-05,
+      "loss": 2.7799,
+      "step": 5704
+    },
+    {
+      "epoch": 0.5164297999456866,
+      "grad_norm": 0.8405930995941162,
+      "learning_rate": 7.936405608037036e-05,
+      "loss": 2.7441,
+      "step": 5705
+    },
+    {
+      "epoch": 0.516520322259437,
+      "grad_norm": 0.9210299253463745,
+      "learning_rate": 7.933300687935716e-05,
+      "loss": 2.7433,
+      "step": 5706
+    },
+    {
+      "epoch": 0.5166108445731873,
+      "grad_norm": 0.8364388346672058,
+      "learning_rate": 7.930195975950957e-05,
+      "loss": 2.7355,
+      "step": 5707
+    },
+    {
+      "epoch": 0.5167013668869377,
+      "grad_norm": 0.8821870684623718,
+      "learning_rate": 7.92709147239541e-05,
+      "loss": 2.7608,
+      "step": 5708
+    },
+    {
+      "epoch": 0.516791889200688,
+      "grad_norm": 0.9231151342391968,
+      "learning_rate": 7.923987177581689e-05,
+      "loss": 2.7756,
+      "step": 5709
+    },
+    {
+      "epoch": 0.5168824115144383,
+      "grad_norm": 0.8480895161628723,
+      "learning_rate": 7.920883091822408e-05,
+      "loss": 2.7919,
+      "step": 5710
+    },
+    {
+      "epoch": 0.5169729338281887,
+      "grad_norm": 0.8738788962364197,
+      "learning_rate": 7.917779215430139e-05,
+      "loss": 2.7691,
+      "step": 5711
+    },
+    {
+      "epoch": 0.5170634561419389,
+      "grad_norm": 0.8850705027580261,
+      "learning_rate": 7.914675548717449e-05,
+      "loss": 2.7242,
+      "step": 5712
+    },
+    {
+      "epoch": 0.5171539784556893,
+      "grad_norm": 0.9246595501899719,
+      "learning_rate": 7.911572091996873e-05,
+      "loss": 2.7948,
+      "step": 5713
+    },
+    {
+      "epoch": 0.5172445007694396,
+      "grad_norm": 0.8189140558242798,
+      "learning_rate": 7.90846884558093e-05,
+      "loss": 2.717,
+      "step": 5714
+    },
+    {
+      "epoch": 0.51733502308319,
+      "grad_norm": 0.9362362027168274,
+      "learning_rate": 7.905365809782114e-05,
+      "loss": 2.7988,
+      "step": 5715
+    },
+    {
+      "epoch": 0.5174255453969403,
+      "grad_norm": 0.9175571799278259,
+      "learning_rate": 7.902262984912909e-05,
+      "loss": 2.7469,
+      "step": 5716
+    },
+    {
+      "epoch": 0.5175160677106907,
+      "grad_norm": 0.8594390749931335,
+      "learning_rate": 7.899160371285761e-05,
+      "loss": 2.7248,
+      "step": 5717
+    },
+    {
+      "epoch": 0.517606590024441,
+      "grad_norm": 0.8442773818969727,
+      "learning_rate": 7.896057969213109e-05,
+      "loss": 2.7009,
+      "step": 5718
+    },
+    {
+      "epoch": 0.5176971123381914,
+      "grad_norm": 0.8920677900314331,
+      "learning_rate": 7.892955779007356e-05,
+      "loss": 2.7349,
+      "step": 5719
+    },
+    {
+      "epoch": 0.5177876346519417,
+      "grad_norm": 0.86868816614151,
+      "learning_rate": 7.889853800980904e-05,
+      "loss": 2.7457,
+      "step": 5720
+    },
+    {
+      "epoch": 0.5178781569656921,
+      "grad_norm": 0.9317598938941956,
+      "learning_rate": 7.886752035446114e-05,
+      "loss": 2.811,
+      "step": 5721
+    },
+    {
+      "epoch": 0.5179686792794423,
+      "grad_norm": 1.1110806465148926,
+      "learning_rate": 7.883650482715338e-05,
+      "loss": 2.8067,
+      "step": 5722
+    },
+    {
+      "epoch": 0.5180592015931927,
+      "grad_norm": 0.8412470817565918,
+      "learning_rate": 7.880549143100897e-05,
+      "loss": 2.7069,
+      "step": 5723
+    },
+    {
+      "epoch": 0.518149723906943,
+      "grad_norm": 0.9267826676368713,
+      "learning_rate": 7.877448016915104e-05,
+      "loss": 2.7886,
+      "step": 5724
+    },
+    {
+      "epoch": 0.5182402462206934,
+      "grad_norm": 1.0773874521255493,
+      "learning_rate": 7.874347104470234e-05,
+      "loss": 2.7465,
+      "step": 5725
+    },
+    {
+      "epoch": 0.5183307685344437,
+      "grad_norm": 1.025495171546936,
+      "learning_rate": 7.871246406078554e-05,
+      "loss": 2.7204,
+      "step": 5726
+    },
+    {
+      "epoch": 0.5184212908481941,
+      "grad_norm": 0.8501183390617371,
+      "learning_rate": 7.8681459220523e-05,
+      "loss": 2.7032,
+      "step": 5727
+    },
+    {
+      "epoch": 0.5185118131619444,
+      "grad_norm": 1.1093471050262451,
+      "learning_rate": 7.865045652703697e-05,
+      "loss": 2.7576,
+      "step": 5728
+    },
+    {
+      "epoch": 0.5186023354756948,
+      "grad_norm": 0.8723694086074829,
+      "learning_rate": 7.861945598344936e-05,
+      "loss": 2.7593,
+      "step": 5729
+    },
+    {
+      "epoch": 0.5186928577894451,
+      "grad_norm": 1.0160908699035645,
+      "learning_rate": 7.858845759288198e-05,
+      "loss": 2.7928,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5187833801031955,
+      "grad_norm": 0.7631496787071228,
+      "learning_rate": 7.855746135845628e-05,
+      "loss": 2.7705,
+      "step": 5731
+    },
+    {
+      "epoch": 0.5188739024169458,
+      "grad_norm": 0.882847785949707,
+      "learning_rate": 7.852646728329368e-05,
+      "loss": 2.7306,
+      "step": 5732
+    },
+    {
+      "epoch": 0.5189644247306961,
+      "grad_norm": 0.9211302995681763,
+      "learning_rate": 7.849547537051525e-05,
+      "loss": 2.7593,
+      "step": 5733
+    },
+    {
+      "epoch": 0.5190549470444464,
+      "grad_norm": 0.9857428073883057,
+      "learning_rate": 7.846448562324183e-05,
+      "loss": 2.7279,
+      "step": 5734
+    },
+    {
+      "epoch": 0.5191454693581968,
+      "grad_norm": 0.8712624311447144,
+      "learning_rate": 7.843349804459412e-05,
+      "loss": 2.7363,
+      "step": 5735
+    },
+    {
+      "epoch": 0.5192359916719471,
+      "grad_norm": 1.004382848739624,
+      "learning_rate": 7.840251263769255e-05,
+      "loss": 2.7636,
+      "step": 5736
+    },
+    {
+      "epoch": 0.5193265139856975,
+      "grad_norm": 0.9310490489006042,
+      "learning_rate": 7.837152940565741e-05,
+      "loss": 2.779,
+      "step": 5737
+    },
+    {
+      "epoch": 0.5194170362994478,
+      "grad_norm": 0.8835846781730652,
+      "learning_rate": 7.834054835160864e-05,
+      "loss": 2.7261,
+      "step": 5738
+    },
+    {
+      "epoch": 0.5195075586131982,
+      "grad_norm": 1.0388543605804443,
+      "learning_rate": 7.830956947866607e-05,
+      "loss": 2.7124,
+      "step": 5739
+    },
+    {
+      "epoch": 0.5195980809269485,
+      "grad_norm": 0.8679820895195007,
+      "learning_rate": 7.827859278994925e-05,
+      "loss": 2.7933,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5196886032406989,
+      "grad_norm": 1.0058894157409668,
+      "learning_rate": 7.824761828857755e-05,
+      "loss": 2.8184,
+      "step": 5741
+    },
+    {
+      "epoch": 0.5197791255544492,
+      "grad_norm": 0.9288240671157837,
+      "learning_rate": 7.821664597767006e-05,
+      "loss": 2.7192,
+      "step": 5742
+    },
+    {
+      "epoch": 0.5198696478681996,
+      "grad_norm": 0.8325408697128296,
+      "learning_rate": 7.818567586034577e-05,
+      "loss": 2.6912,
+      "step": 5743
+    },
+    {
+      "epoch": 0.5199601701819498,
+      "grad_norm": 0.9072100520133972,
+      "learning_rate": 7.815470793972331e-05,
+      "loss": 2.7642,
+      "step": 5744
+    },
+    {
+      "epoch": 0.5200506924957002,
+      "grad_norm": 0.8883663415908813,
+      "learning_rate": 7.812374221892115e-05,
+      "loss": 2.7666,
+      "step": 5745
+    },
+    {
+      "epoch": 0.5201412148094505,
+      "grad_norm": 0.8535317778587341,
+      "learning_rate": 7.809277870105753e-05,
+      "loss": 2.7037,
+      "step": 5746
+    },
+    {
+      "epoch": 0.5202317371232009,
+      "grad_norm": 0.8594459295272827,
+      "learning_rate": 7.806181738925055e-05,
+      "loss": 2.7063,
+      "step": 5747
+    },
+    {
+      "epoch": 0.5203222594369512,
+      "grad_norm": 0.9233312606811523,
+      "learning_rate": 7.803085828661792e-05,
+      "loss": 2.7687,
+      "step": 5748
+    },
+    {
+      "epoch": 0.5204127817507016,
+      "grad_norm": 0.8871405124664307,
+      "learning_rate": 7.799990139627729e-05,
+      "loss": 2.7712,
+      "step": 5749
+    },
+    {
+      "epoch": 0.5205033040644519,
+      "grad_norm": 0.8980066776275635,
+      "learning_rate": 7.796894672134594e-05,
+      "loss": 2.7603,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5205938263782022,
+      "grad_norm": 0.8637656569480896,
+      "learning_rate": 7.79379942649411e-05,
+      "loss": 2.7187,
+      "step": 5751
+    },
+    {
+      "epoch": 0.5206843486919526,
+      "grad_norm": 0.8965839147567749,
+      "learning_rate": 7.79070440301796e-05,
+      "loss": 2.7725,
+      "step": 5752
+    },
+    {
+      "epoch": 0.5207748710057029,
+      "grad_norm": 0.7956546545028687,
+      "learning_rate": 7.78760960201782e-05,
+      "loss": 2.6925,
+      "step": 5753
+    },
+    {
+      "epoch": 0.5208653933194533,
+      "grad_norm": 0.9585323333740234,
+      "learning_rate": 7.784515023805328e-05,
+      "loss": 2.8031,
+      "step": 5754
+    },
+    {
+      "epoch": 0.5209559156332035,
+      "grad_norm": 0.935881495475769,
+      "learning_rate": 7.781420668692116e-05,
+      "loss": 2.7467,
+      "step": 5755
+    },
+    {
+      "epoch": 0.5210464379469539,
+      "grad_norm": 0.8071610927581787,
+      "learning_rate": 7.778326536989779e-05,
+      "loss": 2.7042,
+      "step": 5756
+    },
+    {
+      "epoch": 0.5211369602607042,
+      "grad_norm": 0.9618384838104248,
+      "learning_rate": 7.775232629009904e-05,
+      "loss": 2.7429,
+      "step": 5757
+    },
+    {
+      "epoch": 0.5212274825744546,
+      "grad_norm": 0.9782554507255554,
+      "learning_rate": 7.772138945064035e-05,
+      "loss": 2.7997,
+      "step": 5758
+    },
+    {
+      "epoch": 0.5213180048882049,
+      "grad_norm": 0.9291982054710388,
+      "learning_rate": 7.769045485463719e-05,
+      "loss": 2.7496,
+      "step": 5759
+    },
+    {
+      "epoch": 0.5214085272019553,
+      "grad_norm": 0.8440256118774414,
+      "learning_rate": 7.765952250520459e-05,
+      "loss": 2.7811,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5214990495157056,
+      "grad_norm": 0.8678810596466064,
+      "learning_rate": 7.762859240545748e-05,
+      "loss": 2.7768,
+      "step": 5761
+    },
+    {
+      "epoch": 0.521589571829456,
+      "grad_norm": 0.8405066132545471,
+      "learning_rate": 7.759766455851046e-05,
+      "loss": 2.7836,
+      "step": 5762
+    },
+    {
+      "epoch": 0.5216800941432063,
+      "grad_norm": 0.8501881957054138,
+      "learning_rate": 7.756673896747805e-05,
+      "loss": 2.7211,
+      "step": 5763
+    },
+    {
+      "epoch": 0.5217706164569567,
+      "grad_norm": 0.9259394407272339,
+      "learning_rate": 7.753581563547441e-05,
+      "loss": 2.7372,
+      "step": 5764
+    },
+    {
+      "epoch": 0.521861138770707,
+      "grad_norm": 0.8363810181617737,
+      "learning_rate": 7.750489456561352e-05,
+      "loss": 2.7835,
+      "step": 5765
+    },
+    {
+      "epoch": 0.5219516610844573,
+      "grad_norm": 0.8172920942306519,
+      "learning_rate": 7.74739757610091e-05,
+      "loss": 2.8003,
+      "step": 5766
+    },
+    {
+      "epoch": 0.5220421833982076,
+      "grad_norm": 0.8599033355712891,
+      "learning_rate": 7.744305922477474e-05,
+      "loss": 2.7674,
+      "step": 5767
+    },
+    {
+      "epoch": 0.522132705711958,
+      "grad_norm": 0.853209376335144,
+      "learning_rate": 7.741214496002368e-05,
+      "loss": 2.7579,
+      "step": 5768
+    },
+    {
+      "epoch": 0.5222232280257083,
+      "grad_norm": 0.8500555753707886,
+      "learning_rate": 7.738123296986903e-05,
+      "loss": 2.7062,
+      "step": 5769
+    },
+    {
+      "epoch": 0.5223137503394587,
+      "grad_norm": 0.8476354479789734,
+      "learning_rate": 7.735032325742355e-05,
+      "loss": 2.818,
+      "step": 5770
+    },
+    {
+      "epoch": 0.522404272653209,
+      "grad_norm": 0.8359866738319397,
+      "learning_rate": 7.731941582579998e-05,
+      "loss": 2.7028,
+      "step": 5771
+    },
+    {
+      "epoch": 0.5224947949669594,
+      "grad_norm": 0.8658446073532104,
+      "learning_rate": 7.728851067811054e-05,
+      "loss": 2.7431,
+      "step": 5772
+    },
+    {
+      "epoch": 0.5225853172807097,
+      "grad_norm": 0.9646530151367188,
+      "learning_rate": 7.72576078174675e-05,
+      "loss": 2.705,
+      "step": 5773
+    },
+    {
+      "epoch": 0.5226758395944601,
+      "grad_norm": 0.8622293472290039,
+      "learning_rate": 7.72267072469827e-05,
+      "loss": 2.7561,
+      "step": 5774
+    },
+    {
+      "epoch": 0.5227663619082104,
+      "grad_norm": 0.7764016389846802,
+      "learning_rate": 7.719580896976788e-05,
+      "loss": 2.75,
+      "step": 5775
+    },
+    {
+      "epoch": 0.5228568842219607,
+      "grad_norm": 0.8442264199256897,
+      "learning_rate": 7.716491298893442e-05,
+      "loss": 2.6812,
+      "step": 5776
+    },
+    {
+      "epoch": 0.522947406535711,
+      "grad_norm": 0.9320610165596008,
+      "learning_rate": 7.713401930759365e-05,
+      "loss": 2.6775,
+      "step": 5777
+    },
+    {
+      "epoch": 0.5230379288494614,
+      "grad_norm": 0.8457524180412292,
+      "learning_rate": 7.71031279288565e-05,
+      "loss": 2.7454,
+      "step": 5778
+    },
+    {
+      "epoch": 0.5231284511632117,
+      "grad_norm": 0.8799865245819092,
+      "learning_rate": 7.707223885583375e-05,
+      "loss": 2.7764,
+      "step": 5779
+    },
+    {
+      "epoch": 0.5232189734769621,
+      "grad_norm": 0.8141556978225708,
+      "learning_rate": 7.704135209163589e-05,
+      "loss": 2.7949,
+      "step": 5780
+    },
+    {
+      "epoch": 0.5233094957907124,
+      "grad_norm": 0.8758169412612915,
+      "learning_rate": 7.701046763937328e-05,
+      "loss": 2.7202,
+      "step": 5781
+    },
+    {
+      "epoch": 0.5234000181044628,
+      "grad_norm": 0.8458023071289062,
+      "learning_rate": 7.697958550215592e-05,
+      "loss": 2.756,
+      "step": 5782
+    },
+    {
+      "epoch": 0.5234905404182131,
+      "grad_norm": 0.8262035846710205,
+      "learning_rate": 7.69487056830937e-05,
+      "loss": 2.7523,
+      "step": 5783
+    },
+    {
+      "epoch": 0.5235810627319635,
+      "grad_norm": 0.7745006680488586,
+      "learning_rate": 7.691782818529615e-05,
+      "loss": 2.7067,
+      "step": 5784
+    },
+    {
+      "epoch": 0.5236715850457138,
+      "grad_norm": 0.825136661529541,
+      "learning_rate": 7.68869530118727e-05,
+      "loss": 2.7196,
+      "step": 5785
+    },
+    {
+      "epoch": 0.5237621073594642,
+      "grad_norm": 0.8253481984138489,
+      "learning_rate": 7.685608016593244e-05,
+      "loss": 2.7117,
+      "step": 5786
+    },
+    {
+      "epoch": 0.5238526296732144,
+      "grad_norm": 0.8600527048110962,
+      "learning_rate": 7.682520965058428e-05,
+      "loss": 2.7612,
+      "step": 5787
+    },
+    {
+      "epoch": 0.5239431519869648,
+      "grad_norm": 0.7747024297714233,
+      "learning_rate": 7.679434146893685e-05,
+      "loss": 2.6913,
+      "step": 5788
+    },
+    {
+      "epoch": 0.5240336743007151,
+      "grad_norm": 0.8048658967018127,
+      "learning_rate": 7.676347562409863e-05,
+      "loss": 2.8,
+      "step": 5789
+    },
+    {
+      "epoch": 0.5241241966144655,
+      "grad_norm": 0.8808863162994385,
+      "learning_rate": 7.673261211917776e-05,
+      "loss": 2.7518,
+      "step": 5790
+    },
+    {
+      "epoch": 0.5242147189282158,
+      "grad_norm": 0.8000421524047852,
+      "learning_rate": 7.670175095728224e-05,
+      "loss": 2.7402,
+      "step": 5791
+    },
+    {
+      "epoch": 0.5243052412419661,
+      "grad_norm": 0.8699004054069519,
+      "learning_rate": 7.667089214151971e-05,
+      "loss": 2.8246,
+      "step": 5792
+    },
+    {
+      "epoch": 0.5243957635557165,
+      "grad_norm": 0.8132240176200867,
+      "learning_rate": 7.664003567499776e-05,
+      "loss": 2.7355,
+      "step": 5793
+    },
+    {
+      "epoch": 0.5244862858694668,
+      "grad_norm": 0.8190984725952148,
+      "learning_rate": 7.660918156082354e-05,
+      "loss": 2.6704,
+      "step": 5794
+    },
+    {
+      "epoch": 0.5245768081832172,
+      "grad_norm": 0.9430919885635376,
+      "learning_rate": 7.657832980210412e-05,
+      "loss": 2.7914,
+      "step": 5795
+    },
+    {
+      "epoch": 0.5246673304969675,
+      "grad_norm": 0.8266918659210205,
+      "learning_rate": 7.654748040194622e-05,
+      "loss": 2.7768,
+      "step": 5796
+    },
+    {
+      "epoch": 0.5247578528107179,
+      "grad_norm": 0.9308954477310181,
+      "learning_rate": 7.651663336345642e-05,
+      "loss": 2.7125,
+      "step": 5797
+    },
+    {
+      "epoch": 0.5248483751244681,
+      "grad_norm": 0.9079810380935669,
+      "learning_rate": 7.6485788689741e-05,
+      "loss": 2.761,
+      "step": 5798
+    },
+    {
+      "epoch": 0.5249388974382185,
+      "grad_norm": 0.9260097742080688,
+      "learning_rate": 7.645494638390603e-05,
+      "loss": 2.7153,
+      "step": 5799
+    },
+    {
+      "epoch": 0.5250294197519688,
+      "grad_norm": 0.8457943797111511,
+      "learning_rate": 7.642410644905726e-05,
+      "loss": 2.7854,
+      "step": 5800
+    },
+    {
+      "epoch": 0.5251199420657192,
+      "grad_norm": 0.8552992343902588,
+      "learning_rate": 7.639326888830039e-05,
+      "loss": 2.7727,
+      "step": 5801
+    },
+    {
+      "epoch": 0.5252104643794695,
+      "grad_norm": 0.9081202745437622,
+      "learning_rate": 7.636243370474066e-05,
+      "loss": 2.7699,
+      "step": 5802
+    },
+    {
+      "epoch": 0.5253009866932199,
+      "grad_norm": 0.9340787529945374,
+      "learning_rate": 7.633160090148323e-05,
+      "loss": 2.7689,
+      "step": 5803
+    },
+    {
+      "epoch": 0.5253915090069702,
+      "grad_norm": 0.8104767799377441,
+      "learning_rate": 7.630077048163292e-05,
+      "loss": 2.7085,
+      "step": 5804
+    },
+    {
+      "epoch": 0.5254820313207206,
+      "grad_norm": 0.907471776008606,
+      "learning_rate": 7.626994244829441e-05,
+      "loss": 2.7188,
+      "step": 5805
+    },
+    {
+      "epoch": 0.5255725536344709,
+      "grad_norm": 0.8070109486579895,
+      "learning_rate": 7.623911680457198e-05,
+      "loss": 2.6972,
+      "step": 5806
+    },
+    {
+      "epoch": 0.5256630759482213,
+      "grad_norm": 0.7986403107643127,
+      "learning_rate": 7.620829355356989e-05,
+      "loss": 2.7255,
+      "step": 5807
+    },
+    {
+      "epoch": 0.5257535982619715,
+      "grad_norm": 0.893733561038971,
+      "learning_rate": 7.617747269839198e-05,
+      "loss": 2.7295,
+      "step": 5808
+    },
+    {
+      "epoch": 0.5258441205757219,
+      "grad_norm": 1.015689730644226,
+      "learning_rate": 7.614665424214193e-05,
+      "loss": 2.7273,
+      "step": 5809
+    },
+    {
+      "epoch": 0.5259346428894722,
+      "grad_norm": 0.8221648931503296,
+      "learning_rate": 7.611583818792311e-05,
+      "loss": 2.7565,
+      "step": 5810
+    },
+    {
+      "epoch": 0.5260251652032226,
+      "grad_norm": 0.8100350499153137,
+      "learning_rate": 7.608502453883876e-05,
+      "loss": 2.7337,
+      "step": 5811
+    },
+    {
+      "epoch": 0.5261156875169729,
+      "grad_norm": 0.9273995161056519,
+      "learning_rate": 7.605421329799176e-05,
+      "loss": 2.7955,
+      "step": 5812
+    },
+    {
+      "epoch": 0.5262062098307233,
+      "grad_norm": 0.8550669550895691,
+      "learning_rate": 7.602340446848486e-05,
+      "loss": 2.7722,
+      "step": 5813
+    },
+    {
+      "epoch": 0.5262967321444736,
+      "grad_norm": 0.9372648596763611,
+      "learning_rate": 7.599259805342043e-05,
+      "loss": 2.7734,
+      "step": 5814
+    },
+    {
+      "epoch": 0.526387254458224,
+      "grad_norm": 0.8486389517784119,
+      "learning_rate": 7.596179405590076e-05,
+      "loss": 2.7522,
+      "step": 5815
+    },
+    {
+      "epoch": 0.5264777767719743,
+      "grad_norm": 0.9426712393760681,
+      "learning_rate": 7.593099247902775e-05,
+      "loss": 2.7529,
+      "step": 5816
+    },
+    {
+      "epoch": 0.5265682990857247,
+      "grad_norm": 0.8096632361412048,
+      "learning_rate": 7.590019332590315e-05,
+      "loss": 2.7477,
+      "step": 5817
+    },
+    {
+      "epoch": 0.526658821399475,
+      "grad_norm": 0.89849454164505,
+      "learning_rate": 7.58693965996284e-05,
+      "loss": 2.7255,
+      "step": 5818
+    },
+    {
+      "epoch": 0.5267493437132253,
+      "grad_norm": 0.8777953386306763,
+      "learning_rate": 7.583860230330478e-05,
+      "loss": 2.7483,
+      "step": 5819
+    },
+    {
+      "epoch": 0.5268398660269756,
+      "grad_norm": 0.8368728160858154,
+      "learning_rate": 7.580781044003324e-05,
+      "loss": 2.7536,
+      "step": 5820
+    },
+    {
+      "epoch": 0.526930388340726,
+      "grad_norm": 0.9228582978248596,
+      "learning_rate": 7.577702101291453e-05,
+      "loss": 2.7164,
+      "step": 5821
+    },
+    {
+      "epoch": 0.5270209106544763,
+      "grad_norm": 0.8236722350120544,
+      "learning_rate": 7.574623402504914e-05,
+      "loss": 2.7076,
+      "step": 5822
+    },
+    {
+      "epoch": 0.5271114329682267,
+      "grad_norm": 0.8082314133644104,
+      "learning_rate": 7.571544947953728e-05,
+      "loss": 2.7056,
+      "step": 5823
+    },
+    {
+      "epoch": 0.527201955281977,
+      "grad_norm": 0.9416561722755432,
+      "learning_rate": 7.568466737947905e-05,
+      "loss": 2.785,
+      "step": 5824
+    },
+    {
+      "epoch": 0.5272924775957274,
+      "grad_norm": 0.8790504336357117,
+      "learning_rate": 7.565388772797411e-05,
+      "loss": 2.767,
+      "step": 5825
+    },
+    {
+      "epoch": 0.5273829999094777,
+      "grad_norm": 0.875694215297699,
+      "learning_rate": 7.562311052812206e-05,
+      "loss": 2.7041,
+      "step": 5826
+    },
+    {
+      "epoch": 0.5274735222232281,
+      "grad_norm": 0.8280487656593323,
+      "learning_rate": 7.559233578302204e-05,
+      "loss": 2.6991,
+      "step": 5827
+    },
+    {
+      "epoch": 0.5275640445369784,
+      "grad_norm": 0.9365737438201904,
+      "learning_rate": 7.556156349577319e-05,
+      "loss": 2.8038,
+      "step": 5828
+    },
+    {
+      "epoch": 0.5276545668507288,
+      "grad_norm": 0.911470890045166,
+      "learning_rate": 7.553079366947421e-05,
+      "loss": 2.7747,
+      "step": 5829
+    },
+    {
+      "epoch": 0.527745089164479,
+      "grad_norm": 0.802963137626648,
+      "learning_rate": 7.550002630722366e-05,
+      "loss": 2.6946,
+      "step": 5830
+    },
+    {
+      "epoch": 0.5278356114782294,
+      "grad_norm": 0.9103678464889526,
+      "learning_rate": 7.546926141211974e-05,
+      "loss": 2.7042,
+      "step": 5831
+    },
+    {
+      "epoch": 0.5279261337919797,
+      "grad_norm": 0.9274582862854004,
+      "learning_rate": 7.54384989872606e-05,
+      "loss": 2.7517,
+      "step": 5832
+    },
+    {
+      "epoch": 0.52801665610573,
+      "grad_norm": 0.8614877462387085,
+      "learning_rate": 7.540773903574386e-05,
+      "loss": 2.7453,
+      "step": 5833
+    },
+    {
+      "epoch": 0.5281071784194804,
+      "grad_norm": 0.888545036315918,
+      "learning_rate": 7.537698156066718e-05,
+      "loss": 2.6807,
+      "step": 5834
+    },
+    {
+      "epoch": 0.5281977007332307,
+      "grad_norm": 0.8485361337661743,
+      "learning_rate": 7.534622656512776e-05,
+      "loss": 2.7216,
+      "step": 5835
+    },
+    {
+      "epoch": 0.5282882230469811,
+      "grad_norm": 0.947659432888031,
+      "learning_rate": 7.531547405222268e-05,
+      "loss": 2.794,
+      "step": 5836
+    },
+    {
+      "epoch": 0.5283787453607314,
+      "grad_norm": 0.8418532609939575,
+      "learning_rate": 7.528472402504862e-05,
+      "loss": 2.7598,
+      "step": 5837
+    },
+    {
+      "epoch": 0.5284692676744818,
+      "grad_norm": 0.8268518447875977,
+      "learning_rate": 7.525397648670225e-05,
+      "loss": 2.7805,
+      "step": 5838
+    },
+    {
+      "epoch": 0.528559789988232,
+      "grad_norm": 0.8547778129577637,
+      "learning_rate": 7.522323144027972e-05,
+      "loss": 2.7813,
+      "step": 5839
+    },
+    {
+      "epoch": 0.5286503123019825,
+      "grad_norm": 0.8886696696281433,
+      "learning_rate": 7.519248888887716e-05,
+      "loss": 2.736,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5287408346157327,
+      "grad_norm": 0.8827240467071533,
+      "learning_rate": 7.516174883559022e-05,
+      "loss": 2.6714,
+      "step": 5841
+    },
+    {
+      "epoch": 0.5288313569294831,
+      "grad_norm": 0.8961451649665833,
+      "learning_rate": 7.513101128351454e-05,
+      "loss": 2.7882,
+      "step": 5842
+    },
+    {
+      "epoch": 0.5289218792432334,
+      "grad_norm": 0.8667140007019043,
+      "learning_rate": 7.510027623574532e-05,
+      "loss": 2.7755,
+      "step": 5843
+    },
+    {
+      "epoch": 0.5290124015569838,
+      "grad_norm": 0.810696005821228,
+      "learning_rate": 7.506954369537763e-05,
+      "loss": 2.6993,
+      "step": 5844
+    },
+    {
+      "epoch": 0.5291029238707341,
+      "grad_norm": 0.8313196301460266,
+      "learning_rate": 7.503881366550617e-05,
+      "loss": 2.6605,
+      "step": 5845
+    },
+    {
+      "epoch": 0.5291934461844845,
+      "grad_norm": 0.8745372295379639,
+      "learning_rate": 7.500808614922551e-05,
+      "loss": 2.7422,
+      "step": 5846
+    },
+    {
+      "epoch": 0.5292839684982348,
+      "grad_norm": 0.8211199641227722,
+      "learning_rate": 7.497736114962989e-05,
+      "loss": 2.7424,
+      "step": 5847
+    },
+    {
+      "epoch": 0.5293744908119852,
+      "grad_norm": 0.8483652472496033,
+      "learning_rate": 7.494663866981335e-05,
+      "loss": 2.7724,
+      "step": 5848
+    },
+    {
+      "epoch": 0.5294650131257355,
+      "grad_norm": 0.8949604630470276,
+      "learning_rate": 7.491591871286954e-05,
+      "loss": 2.7255,
+      "step": 5849
+    },
+    {
+      "epoch": 0.5295555354394859,
+      "grad_norm": 0.8604280948638916,
+      "learning_rate": 7.488520128189209e-05,
+      "loss": 2.7332,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5296460577532361,
+      "grad_norm": 0.8617332577705383,
+      "learning_rate": 7.485448637997416e-05,
+      "loss": 2.7666,
+      "step": 5851
+    },
+    {
+      "epoch": 0.5297365800669865,
+      "grad_norm": 0.8203973770141602,
+      "learning_rate": 7.482377401020879e-05,
+      "loss": 2.7467,
+      "step": 5852
+    },
+    {
+      "epoch": 0.5298271023807368,
+      "grad_norm": 0.8294515013694763,
+      "learning_rate": 7.479306417568864e-05,
+      "loss": 2.7485,
+      "step": 5853
+    },
+    {
+      "epoch": 0.5299176246944872,
+      "grad_norm": 0.8514267802238464,
+      "learning_rate": 7.476235687950628e-05,
+      "loss": 2.7631,
+      "step": 5854
+    },
+    {
+      "epoch": 0.5300081470082375,
+      "grad_norm": 0.909512460231781,
+      "learning_rate": 7.473165212475386e-05,
+      "loss": 2.7867,
+      "step": 5855
+    },
+    {
+      "epoch": 0.5300986693219879,
+      "grad_norm": 0.8771156072616577,
+      "learning_rate": 7.470094991452341e-05,
+      "loss": 2.7723,
+      "step": 5856
+    },
+    {
+      "epoch": 0.5301891916357382,
+      "grad_norm": 0.9345921277999878,
+      "learning_rate": 7.467025025190657e-05,
+      "loss": 2.7522,
+      "step": 5857
+    },
+    {
+      "epoch": 0.5302797139494886,
+      "grad_norm": 0.858391523361206,
+      "learning_rate": 7.463955313999488e-05,
+      "loss": 2.7519,
+      "step": 5858
+    },
+    {
+      "epoch": 0.5303702362632389,
+      "grad_norm": 0.8921976685523987,
+      "learning_rate": 7.460885858187948e-05,
+      "loss": 2.7519,
+      "step": 5859
+    },
+    {
+      "epoch": 0.5304607585769893,
+      "grad_norm": 0.8551515340805054,
+      "learning_rate": 7.457816658065134e-05,
+      "loss": 2.6766,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5305512808907396,
+      "grad_norm": 0.8405719995498657,
+      "learning_rate": 7.454747713940107e-05,
+      "loss": 2.6896,
+      "step": 5861
+    },
+    {
+      "epoch": 0.53064180320449,
+      "grad_norm": 0.8666146397590637,
+      "learning_rate": 7.451679026121922e-05,
+      "loss": 2.7501,
+      "step": 5862
+    },
+    {
+      "epoch": 0.5307323255182402,
+      "grad_norm": 0.8197934031486511,
+      "learning_rate": 7.448610594919586e-05,
+      "loss": 2.7125,
+      "step": 5863
+    },
+    {
+      "epoch": 0.5308228478319906,
+      "grad_norm": 0.8025767803192139,
+      "learning_rate": 7.445542420642097e-05,
+      "loss": 2.772,
+      "step": 5864
+    },
+    {
+      "epoch": 0.5309133701457409,
+      "grad_norm": 0.8161811232566833,
+      "learning_rate": 7.442474503598411e-05,
+      "loss": 2.7286,
+      "step": 5865
+    },
+    {
+      "epoch": 0.5310038924594913,
+      "grad_norm": 0.8732130527496338,
+      "learning_rate": 7.439406844097479e-05,
+      "loss": 2.6955,
+      "step": 5866
+    },
+    {
+      "epoch": 0.5310944147732416,
+      "grad_norm": 0.8568199872970581,
+      "learning_rate": 7.436339442448201e-05,
+      "loss": 2.7652,
+      "step": 5867
+    },
+    {
+      "epoch": 0.531184937086992,
+      "grad_norm": 0.8628947138786316,
+      "learning_rate": 7.433272298959475e-05,
+      "loss": 2.7113,
+      "step": 5868
+    },
+    {
+      "epoch": 0.5312754594007423,
+      "grad_norm": 0.8824062347412109,
+      "learning_rate": 7.430205413940156e-05,
+      "loss": 2.726,
+      "step": 5869
+    },
+    {
+      "epoch": 0.5313659817144927,
+      "grad_norm": 0.8852974772453308,
+      "learning_rate": 7.427138787699086e-05,
+      "loss": 2.7462,
+      "step": 5870
+    },
+    {
+      "epoch": 0.531456504028243,
+      "grad_norm": 0.8664782047271729,
+      "learning_rate": 7.424072420545063e-05,
+      "loss": 2.7654,
+      "step": 5871
+    },
+    {
+      "epoch": 0.5315470263419934,
+      "grad_norm": 0.8074523210525513,
+      "learning_rate": 7.421006312786883e-05,
+      "loss": 2.6719,
+      "step": 5872
+    },
+    {
+      "epoch": 0.5316375486557436,
+      "grad_norm": 0.8614087104797363,
+      "learning_rate": 7.417940464733293e-05,
+      "loss": 2.7299,
+      "step": 5873
+    },
+    {
+      "epoch": 0.5317280709694939,
+      "grad_norm": 0.8256407380104065,
+      "learning_rate": 7.41487487669303e-05,
+      "loss": 2.7517,
+      "step": 5874
+    },
+    {
+      "epoch": 0.5318185932832443,
+      "grad_norm": 0.8972731232643127,
+      "learning_rate": 7.411809548974792e-05,
+      "loss": 2.7187,
+      "step": 5875
+    },
+    {
+      "epoch": 0.5319091155969946,
+      "grad_norm": 0.875149130821228,
+      "learning_rate": 7.408744481887266e-05,
+      "loss": 2.7454,
+      "step": 5876
+    },
+    {
+      "epoch": 0.531999637910745,
+      "grad_norm": 0.8577212691307068,
+      "learning_rate": 7.405679675739096e-05,
+      "loss": 2.7422,
+      "step": 5877
+    },
+    {
+      "epoch": 0.5320901602244953,
+      "grad_norm": 0.8802106380462646,
+      "learning_rate": 7.402615130838917e-05,
+      "loss": 2.7643,
+      "step": 5878
+    },
+    {
+      "epoch": 0.5321806825382457,
+      "grad_norm": 0.9183444976806641,
+      "learning_rate": 7.399550847495318e-05,
+      "loss": 2.7689,
+      "step": 5879
+    },
+    {
+      "epoch": 0.532271204851996,
+      "grad_norm": 0.7971477508544922,
+      "learning_rate": 7.39648682601688e-05,
+      "loss": 2.7532,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5323617271657464,
+      "grad_norm": 0.9035381078720093,
+      "learning_rate": 7.393423066712146e-05,
+      "loss": 2.7816,
+      "step": 5881
+    },
+    {
+      "epoch": 0.5324522494794967,
+      "grad_norm": 0.9476202130317688,
+      "learning_rate": 7.39035956988964e-05,
+      "loss": 2.7672,
+      "step": 5882
+    },
+    {
+      "epoch": 0.532542771793247,
+      "grad_norm": 0.8284057378768921,
+      "learning_rate": 7.387296335857851e-05,
+      "loss": 2.6749,
+      "step": 5883
+    },
+    {
+      "epoch": 0.5326332941069973,
+      "grad_norm": 0.8201116919517517,
+      "learning_rate": 7.384233364925253e-05,
+      "loss": 2.695,
+      "step": 5884
+    },
+    {
+      "epoch": 0.5327238164207477,
+      "grad_norm": 0.8144286274909973,
+      "learning_rate": 7.38117065740028e-05,
+      "loss": 2.7469,
+      "step": 5885
+    },
+    {
+      "epoch": 0.532814338734498,
+      "grad_norm": 0.8197963237762451,
+      "learning_rate": 7.378108213591355e-05,
+      "loss": 2.7029,
+      "step": 5886
+    },
+    {
+      "epoch": 0.5329048610482484,
+      "grad_norm": 0.8558396100997925,
+      "learning_rate": 7.375046033806854e-05,
+      "loss": 2.6771,
+      "step": 5887
+    },
+    {
+      "epoch": 0.5329953833619987,
+      "grad_norm": 0.8030669093132019,
+      "learning_rate": 7.371984118355152e-05,
+      "loss": 2.7112,
+      "step": 5888
+    },
+    {
+      "epoch": 0.5330859056757491,
+      "grad_norm": 0.8504447937011719,
+      "learning_rate": 7.368922467544575e-05,
+      "loss": 2.7953,
+      "step": 5889
+    },
+    {
+      "epoch": 0.5331764279894994,
+      "grad_norm": 0.8708440065383911,
+      "learning_rate": 7.365861081683433e-05,
+      "loss": 2.774,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5332669503032498,
+      "grad_norm": 0.9638001322746277,
+      "learning_rate": 7.362799961080006e-05,
+      "loss": 2.8029,
+      "step": 5891
+    },
+    {
+      "epoch": 0.5333574726170001,
+      "grad_norm": 0.8533604145050049,
+      "learning_rate": 7.359739106042556e-05,
+      "loss": 2.7299,
+      "step": 5892
+    },
+    {
+      "epoch": 0.5334479949307505,
+      "grad_norm": 0.8586344718933105,
+      "learning_rate": 7.356678516879303e-05,
+      "loss": 2.769,
+      "step": 5893
+    },
+    {
+      "epoch": 0.5335385172445007,
+      "grad_norm": 0.8530272841453552,
+      "learning_rate": 7.353618193898453e-05,
+      "loss": 2.7365,
+      "step": 5894
+    },
+    {
+      "epoch": 0.5336290395582511,
+      "grad_norm": 0.8625940680503845,
+      "learning_rate": 7.350558137408174e-05,
+      "loss": 2.7369,
+      "step": 5895
+    },
+    {
+      "epoch": 0.5337195618720014,
+      "grad_norm": 0.8640052676200867,
+      "learning_rate": 7.347498347716624e-05,
+      "loss": 2.7248,
+      "step": 5896
+    },
+    {
+      "epoch": 0.5338100841857518,
+      "grad_norm": 0.8243191242218018,
+      "learning_rate": 7.344438825131911e-05,
+      "loss": 2.7137,
+      "step": 5897
+    },
+    {
+      "epoch": 0.5339006064995021,
+      "grad_norm": 0.8464127779006958,
+      "learning_rate": 7.341379569962141e-05,
+      "loss": 2.7535,
+      "step": 5898
+    },
+    {
+      "epoch": 0.5339911288132525,
+      "grad_norm": 0.884215235710144,
+      "learning_rate": 7.338320582515374e-05,
+      "loss": 2.7813,
+      "step": 5899
+    },
+    {
+      "epoch": 0.5340816511270028,
+      "grad_norm": 0.8700332641601562,
+      "learning_rate": 7.335261863099651e-05,
+      "loss": 2.7107,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5341721734407532,
+      "grad_norm": 0.8220349550247192,
+      "learning_rate": 7.332203412022985e-05,
+      "loss": 2.6694,
+      "step": 5901
+    },
+    {
+      "epoch": 0.5342626957545035,
+      "grad_norm": 0.806977391242981,
+      "learning_rate": 7.329145229593364e-05,
+      "loss": 2.7473,
+      "step": 5902
+    },
+    {
+      "epoch": 0.5343532180682539,
+      "grad_norm": 0.8836818337440491,
+      "learning_rate": 7.326087316118744e-05,
+      "loss": 2.7675,
+      "step": 5903
+    },
+    {
+      "epoch": 0.5344437403820042,
+      "grad_norm": 0.8241246342658997,
+      "learning_rate": 7.323029671907061e-05,
+      "loss": 2.7076,
+      "step": 5904
+    },
+    {
+      "epoch": 0.5345342626957545,
+      "grad_norm": 0.876545250415802,
+      "learning_rate": 7.319972297266214e-05,
+      "loss": 2.7178,
+      "step": 5905
+    },
+    {
+      "epoch": 0.5346247850095048,
+      "grad_norm": 0.8125167489051819,
+      "learning_rate": 7.316915192504082e-05,
+      "loss": 2.7232,
+      "step": 5906
+    },
+    {
+      "epoch": 0.5347153073232552,
+      "grad_norm": 0.8703387379646301,
+      "learning_rate": 7.313858357928519e-05,
+      "loss": 2.7011,
+      "step": 5907
+    },
+    {
+      "epoch": 0.5348058296370055,
+      "grad_norm": 0.8583438396453857,
+      "learning_rate": 7.310801793847344e-05,
+      "loss": 2.7476,
+      "step": 5908
+    },
+    {
+      "epoch": 0.5348963519507559,
+      "grad_norm": 0.8597928285598755,
+      "learning_rate": 7.307745500568358e-05,
+      "loss": 2.751,
+      "step": 5909
+    },
+    {
+      "epoch": 0.5349868742645062,
+      "grad_norm": 0.8204509615898132,
+      "learning_rate": 7.304689478399323e-05,
+      "loss": 2.7064,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5350773965782566,
+      "grad_norm": 0.8251175880432129,
+      "learning_rate": 7.301633727647985e-05,
+      "loss": 2.8144,
+      "step": 5911
+    },
+    {
+      "epoch": 0.5351679188920069,
+      "grad_norm": 0.7730387449264526,
+      "learning_rate": 7.298578248622056e-05,
+      "loss": 2.7158,
+      "step": 5912
+    },
+    {
+      "epoch": 0.5352584412057573,
+      "grad_norm": 0.918010413646698,
+      "learning_rate": 7.295523041629227e-05,
+      "loss": 2.7091,
+      "step": 5913
+    },
+    {
+      "epoch": 0.5353489635195076,
+      "grad_norm": 0.8758893013000488,
+      "learning_rate": 7.292468106977148e-05,
+      "loss": 2.6953,
+      "step": 5914
+    },
+    {
+      "epoch": 0.5354394858332578,
+      "grad_norm": 0.8721521496772766,
+      "learning_rate": 7.289413444973461e-05,
+      "loss": 2.7949,
+      "step": 5915
+    },
+    {
+      "epoch": 0.5355300081470082,
+      "grad_norm": 0.9021723866462708,
+      "learning_rate": 7.286359055925765e-05,
+      "loss": 2.783,
+      "step": 5916
+    },
+    {
+      "epoch": 0.5356205304607585,
+      "grad_norm": 0.8621231913566589,
+      "learning_rate": 7.283304940141637e-05,
+      "loss": 2.793,
+      "step": 5917
+    },
+    {
+      "epoch": 0.5357110527745089,
+      "grad_norm": 0.7841774225234985,
+      "learning_rate": 7.280251097928625e-05,
+      "loss": 2.7259,
+      "step": 5918
+    },
+    {
+      "epoch": 0.5358015750882592,
+      "grad_norm": 0.8686859607696533,
+      "learning_rate": 7.277197529594257e-05,
+      "loss": 2.7544,
+      "step": 5919
+    },
+    {
+      "epoch": 0.5358920974020096,
+      "grad_norm": 0.8585182428359985,
+      "learning_rate": 7.274144235446023e-05,
+      "loss": 2.7618,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5359826197157599,
+      "grad_norm": 0.907319962978363,
+      "learning_rate": 7.27109121579139e-05,
+      "loss": 2.7342,
+      "step": 5921
+    },
+    {
+      "epoch": 0.5360731420295103,
+      "grad_norm": 0.81803959608078,
+      "learning_rate": 7.268038470937792e-05,
+      "loss": 2.6816,
+      "step": 5922
+    },
+    {
+      "epoch": 0.5361636643432606,
+      "grad_norm": 0.861193835735321,
+      "learning_rate": 7.26498600119265e-05,
+      "loss": 2.672,
+      "step": 5923
+    },
+    {
+      "epoch": 0.536254186657011,
+      "grad_norm": 0.8426436185836792,
+      "learning_rate": 7.261933806863342e-05,
+      "loss": 2.7048,
+      "step": 5924
+    },
+    {
+      "epoch": 0.5363447089707613,
+      "grad_norm": 0.8128862977027893,
+      "learning_rate": 7.258881888257227e-05,
+      "loss": 2.6564,
+      "step": 5925
+    },
+    {
+      "epoch": 0.5364352312845116,
+      "grad_norm": 0.8285736441612244,
+      "learning_rate": 7.255830245681626e-05,
+      "loss": 2.6867,
+      "step": 5926
+    },
+    {
+      "epoch": 0.5365257535982619,
+      "grad_norm": 0.8216138482093811,
+      "learning_rate": 7.252778879443849e-05,
+      "loss": 2.7369,
+      "step": 5927
+    },
+    {
+      "epoch": 0.5366162759120123,
+      "grad_norm": 0.7723904848098755,
+      "learning_rate": 7.249727789851158e-05,
+      "loss": 2.7313,
+      "step": 5928
+    },
+    {
+      "epoch": 0.5367067982257626,
+      "grad_norm": 0.8136193752288818,
+      "learning_rate": 7.24667697721081e-05,
+      "loss": 2.7245,
+      "step": 5929
+    },
+    {
+      "epoch": 0.536797320539513,
+      "grad_norm": 0.8116896748542786,
+      "learning_rate": 7.243626441830009e-05,
+      "loss": 2.699,
+      "step": 5930
+    },
+    {
+      "epoch": 0.5368878428532633,
+      "grad_norm": 0.7944446206092834,
+      "learning_rate": 7.240576184015954e-05,
+      "loss": 2.7374,
+      "step": 5931
+    },
+    {
+      "epoch": 0.5369783651670137,
+      "grad_norm": 0.8052989840507507,
+      "learning_rate": 7.237526204075797e-05,
+      "loss": 2.6733,
+      "step": 5932
+    },
+    {
+      "epoch": 0.537068887480764,
+      "grad_norm": 0.92603600025177,
+      "learning_rate": 7.234476502316682e-05,
+      "loss": 2.7779,
+      "step": 5933
+    },
+    {
+      "epoch": 0.5371594097945144,
+      "grad_norm": 0.9119099974632263,
+      "learning_rate": 7.231427079045703e-05,
+      "loss": 2.7242,
+      "step": 5934
+    },
+    {
+      "epoch": 0.5372499321082647,
+      "grad_norm": 0.8775193095207214,
+      "learning_rate": 7.228377934569945e-05,
+      "loss": 2.7134,
+      "step": 5935
+    },
+    {
+      "epoch": 0.5373404544220151,
+      "grad_norm": 0.8331400156021118,
+      "learning_rate": 7.225329069196448e-05,
+      "loss": 2.7262,
+      "step": 5936
+    },
+    {
+      "epoch": 0.5374309767357653,
+      "grad_norm": 0.9593919515609741,
+      "learning_rate": 7.222280483232242e-05,
+      "loss": 2.7241,
+      "step": 5937
+    },
+    {
+      "epoch": 0.5375214990495157,
+      "grad_norm": 0.8055837750434875,
+      "learning_rate": 7.219232176984314e-05,
+      "loss": 2.7373,
+      "step": 5938
+    },
+    {
+      "epoch": 0.537612021363266,
+      "grad_norm": 0.8983293175697327,
+      "learning_rate": 7.21618415075963e-05,
+      "loss": 2.7519,
+      "step": 5939
+    },
+    {
+      "epoch": 0.5377025436770164,
+      "grad_norm": 0.8614658117294312,
+      "learning_rate": 7.213136404865124e-05,
+      "loss": 2.7133,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5377930659907667,
+      "grad_norm": 0.8453959822654724,
+      "learning_rate": 7.210088939607708e-05,
+      "loss": 2.7399,
+      "step": 5941
+    },
+    {
+      "epoch": 0.5378835883045171,
+      "grad_norm": 0.8339153528213501,
+      "learning_rate": 7.207041755294259e-05,
+      "loss": 2.7262,
+      "step": 5942
+    },
+    {
+      "epoch": 0.5379741106182674,
+      "grad_norm": 0.8798535466194153,
+      "learning_rate": 7.20399485223163e-05,
+      "loss": 2.6945,
+      "step": 5943
+    },
+    {
+      "epoch": 0.5380646329320178,
+      "grad_norm": 0.8549126982688904,
+      "learning_rate": 7.200948230726639e-05,
+      "loss": 2.7325,
+      "step": 5944
+    },
+    {
+      "epoch": 0.5381551552457681,
+      "grad_norm": 0.9284637570381165,
+      "learning_rate": 7.19790189108609e-05,
+      "loss": 2.7726,
+      "step": 5945
+    },
+    {
+      "epoch": 0.5382456775595185,
+      "grad_norm": 0.8560376167297363,
+      "learning_rate": 7.194855833616739e-05,
+      "loss": 2.7303,
+      "step": 5946
+    },
+    {
+      "epoch": 0.5383361998732688,
+      "grad_norm": 0.8566827178001404,
+      "learning_rate": 7.191810058625333e-05,
+      "loss": 2.7196,
+      "step": 5947
+    },
+    {
+      "epoch": 0.5384267221870191,
+      "grad_norm": 0.8862281441688538,
+      "learning_rate": 7.188764566418572e-05,
+      "loss": 2.7211,
+      "step": 5948
+    },
+    {
+      "epoch": 0.5385172445007694,
+      "grad_norm": 0.8478960990905762,
+      "learning_rate": 7.185719357303146e-05,
+      "loss": 2.7088,
+      "step": 5949
+    },
+    {
+      "epoch": 0.5386077668145198,
+      "grad_norm": 0.8626987934112549,
+      "learning_rate": 7.182674431585704e-05,
+      "loss": 2.7561,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5386982891282701,
+      "grad_norm": 0.9624312520027161,
+      "learning_rate": 7.17962978957287e-05,
+      "loss": 2.7622,
+      "step": 5951
+    },
+    {
+      "epoch": 0.5387888114420205,
+      "grad_norm": 0.8094488382339478,
+      "learning_rate": 7.176585431571235e-05,
+      "loss": 2.6762,
+      "step": 5952
+    },
+    {
+      "epoch": 0.5388793337557708,
+      "grad_norm": 0.8915351033210754,
+      "learning_rate": 7.173541357887375e-05,
+      "loss": 2.7593,
+      "step": 5953
+    },
+    {
+      "epoch": 0.5389698560695212,
+      "grad_norm": 0.9364251494407654,
+      "learning_rate": 7.170497568827818e-05,
+      "loss": 2.7553,
+      "step": 5954
+    },
+    {
+      "epoch": 0.5390603783832715,
+      "grad_norm": 0.850823163986206,
+      "learning_rate": 7.167454064699083e-05,
+      "loss": 2.7318,
+      "step": 5955
+    },
+    {
+      "epoch": 0.5391509006970218,
+      "grad_norm": 0.8671566843986511,
+      "learning_rate": 7.16441084580764e-05,
+      "loss": 2.7674,
+      "step": 5956
+    },
+    {
+      "epoch": 0.5392414230107722,
+      "grad_norm": 0.8717043995857239,
+      "learning_rate": 7.161367912459954e-05,
+      "loss": 2.7091,
+      "step": 5957
+    },
+    {
+      "epoch": 0.5393319453245224,
+      "grad_norm": 0.8433679938316345,
+      "learning_rate": 7.158325264962438e-05,
+      "loss": 2.7505,
+      "step": 5958
+    },
+    {
+      "epoch": 0.5394224676382728,
+      "grad_norm": 0.8027213215827942,
+      "learning_rate": 7.155282903621492e-05,
+      "loss": 2.7469,
+      "step": 5959
+    },
+    {
+      "epoch": 0.5395129899520231,
+      "grad_norm": 0.823722779750824,
+      "learning_rate": 7.152240828743477e-05,
+      "loss": 2.7318,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5396035122657735,
+      "grad_norm": 0.8178278803825378,
+      "learning_rate": 7.149199040634737e-05,
+      "loss": 2.7116,
+      "step": 5961
+    },
+    {
+      "epoch": 0.5396940345795238,
+      "grad_norm": 0.8535356521606445,
+      "learning_rate": 7.14615753960157e-05,
+      "loss": 2.7306,
+      "step": 5962
+    },
+    {
+      "epoch": 0.5397845568932742,
+      "grad_norm": 0.8493401408195496,
+      "learning_rate": 7.143116325950265e-05,
+      "loss": 2.7386,
+      "step": 5963
+    },
+    {
+      "epoch": 0.5398750792070245,
+      "grad_norm": 0.7935865521430969,
+      "learning_rate": 7.140075399987068e-05,
+      "loss": 2.7214,
+      "step": 5964
+    },
+    {
+      "epoch": 0.5399656015207749,
+      "grad_norm": 0.9051488041877747,
+      "learning_rate": 7.137034762018198e-05,
+      "loss": 2.7434,
+      "step": 5965
+    },
+    {
+      "epoch": 0.5400561238345252,
+      "grad_norm": 0.8879326581954956,
+      "learning_rate": 7.133994412349847e-05,
+      "loss": 2.7408,
+      "step": 5966
+    },
+    {
+      "epoch": 0.5401466461482756,
+      "grad_norm": 0.9167154431343079,
+      "learning_rate": 7.130954351288184e-05,
+      "loss": 2.7936,
+      "step": 5967
+    },
+    {
+      "epoch": 0.5402371684620259,
+      "grad_norm": 0.9298393726348877,
+      "learning_rate": 7.127914579139338e-05,
+      "loss": 2.7735,
+      "step": 5968
+    },
+    {
+      "epoch": 0.5403276907757762,
+      "grad_norm": 0.845268964767456,
+      "learning_rate": 7.124875096209416e-05,
+      "loss": 2.7469,
+      "step": 5969
+    },
+    {
+      "epoch": 0.5404182130895265,
+      "grad_norm": 0.9079629182815552,
+      "learning_rate": 7.12183590280449e-05,
+      "loss": 2.7276,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5405087354032769,
+      "grad_norm": 0.8613841533660889,
+      "learning_rate": 7.118796999230614e-05,
+      "loss": 2.7374,
+      "step": 5971
+    },
+    {
+      "epoch": 0.5405992577170272,
+      "grad_norm": 0.8585695028305054,
+      "learning_rate": 7.115758385793799e-05,
+      "loss": 2.7661,
+      "step": 5972
+    },
+    {
+      "epoch": 0.5406897800307776,
+      "grad_norm": 0.8777322173118591,
+      "learning_rate": 7.112720062800038e-05,
+      "loss": 2.7605,
+      "step": 5973
+    },
+    {
+      "epoch": 0.5407803023445279,
+      "grad_norm": 0.8847030997276306,
+      "learning_rate": 7.109682030555283e-05,
+      "loss": 2.714,
+      "step": 5974
+    },
+    {
+      "epoch": 0.5408708246582783,
+      "grad_norm": 0.8644956350326538,
+      "learning_rate": 7.106644289365473e-05,
+      "loss": 2.7303,
+      "step": 5975
+    },
+    {
+      "epoch": 0.5409613469720286,
+      "grad_norm": 0.8250646591186523,
+      "learning_rate": 7.103606839536502e-05,
+      "loss": 2.7247,
+      "step": 5976
+    },
+    {
+      "epoch": 0.541051869285779,
+      "grad_norm": 0.8640917539596558,
+      "learning_rate": 7.100569681374245e-05,
+      "loss": 2.7778,
+      "step": 5977
+    },
+    {
+      "epoch": 0.5411423915995293,
+      "grad_norm": 0.864972710609436,
+      "learning_rate": 7.097532815184538e-05,
+      "loss": 2.6894,
+      "step": 5978
+    },
+    {
+      "epoch": 0.5412329139132797,
+      "grad_norm": 0.8219044804573059,
+      "learning_rate": 7.094496241273199e-05,
+      "loss": 2.7022,
+      "step": 5979
+    },
+    {
+      "epoch": 0.5413234362270299,
+      "grad_norm": 0.8397749066352844,
+      "learning_rate": 7.09145995994601e-05,
+      "loss": 2.7996,
+      "step": 5980
+    },
+    {
+      "epoch": 0.5414139585407803,
+      "grad_norm": 0.8366591930389404,
+      "learning_rate": 7.088423971508724e-05,
+      "loss": 2.7352,
+      "step": 5981
+    },
+    {
+      "epoch": 0.5415044808545306,
+      "grad_norm": 0.8768446445465088,
+      "learning_rate": 7.08538827626706e-05,
+      "loss": 2.7229,
+      "step": 5982
+    },
+    {
+      "epoch": 0.541595003168281,
+      "grad_norm": 0.8540074825286865,
+      "learning_rate": 7.082352874526721e-05,
+      "loss": 2.7058,
+      "step": 5983
+    },
+    {
+      "epoch": 0.5416855254820313,
+      "grad_norm": 0.8582448363304138,
+      "learning_rate": 7.079317766593365e-05,
+      "loss": 2.7219,
+      "step": 5984
+    },
+    {
+      "epoch": 0.5417760477957817,
+      "grad_norm": 0.9049850106239319,
+      "learning_rate": 7.076282952772633e-05,
+      "loss": 2.7804,
+      "step": 5985
+    },
+    {
+      "epoch": 0.541866570109532,
+      "grad_norm": 0.8414204716682434,
+      "learning_rate": 7.073248433370124e-05,
+      "loss": 2.7502,
+      "step": 5986
+    },
+    {
+      "epoch": 0.5419570924232824,
+      "grad_norm": 0.9655411839485168,
+      "learning_rate": 7.070214208691422e-05,
+      "loss": 2.7796,
+      "step": 5987
+    },
+    {
+      "epoch": 0.5420476147370327,
+      "grad_norm": 0.9256438612937927,
+      "learning_rate": 7.067180279042066e-05,
+      "loss": 2.7149,
+      "step": 5988
+    },
+    {
+      "epoch": 0.5421381370507831,
+      "grad_norm": 0.8467999696731567,
+      "learning_rate": 7.064146644727579e-05,
+      "loss": 2.6974,
+      "step": 5989
+    },
+    {
+      "epoch": 0.5422286593645333,
+      "grad_norm": 0.883070170879364,
+      "learning_rate": 7.061113306053443e-05,
+      "loss": 2.745,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5423191816782837,
+      "grad_norm": 0.9323094487190247,
+      "learning_rate": 7.058080263325115e-05,
+      "loss": 2.8148,
+      "step": 5991
+    },
+    {
+      "epoch": 0.542409703992034,
+      "grad_norm": 0.8571617007255554,
+      "learning_rate": 7.055047516848025e-05,
+      "loss": 2.7616,
+      "step": 5992
+    },
+    {
+      "epoch": 0.5425002263057844,
+      "grad_norm": 0.911531925201416,
+      "learning_rate": 7.052015066927568e-05,
+      "loss": 2.7568,
+      "step": 5993
+    },
+    {
+      "epoch": 0.5425907486195347,
+      "grad_norm": 0.8240039944648743,
+      "learning_rate": 7.048982913869116e-05,
+      "loss": 2.7021,
+      "step": 5994
+    },
+    {
+      "epoch": 0.5426812709332851,
+      "grad_norm": 0.8227641582489014,
+      "learning_rate": 7.045951057978e-05,
+      "loss": 2.7268,
+      "step": 5995
+    },
+    {
+      "epoch": 0.5427717932470354,
+      "grad_norm": 0.8382290601730347,
+      "learning_rate": 7.042919499559537e-05,
+      "loss": 2.7026,
+      "step": 5996
+    },
+    {
+      "epoch": 0.5428623155607857,
+      "grad_norm": 0.9181439876556396,
+      "learning_rate": 7.039888238918993e-05,
+      "loss": 2.7422,
+      "step": 5997
+    },
+    {
+      "epoch": 0.5429528378745361,
+      "grad_norm": 0.8383064866065979,
+      "learning_rate": 7.036857276361627e-05,
+      "loss": 2.7658,
+      "step": 5998
+    },
+    {
+      "epoch": 0.5430433601882864,
+      "grad_norm": 1.049307942390442,
+      "learning_rate": 7.03382661219265e-05,
+      "loss": 2.7369,
+      "step": 5999
+    },
+    {
+      "epoch": 0.5431338825020368,
+      "grad_norm": 0.8651420474052429,
+      "learning_rate": 7.030796246717255e-05,
+      "loss": 2.6735,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5431338825020368,
+      "eval_loss": 2.6711881160736084,
+      "eval_runtime": 71.67,
+      "eval_samples_per_second": 37.715,
+      "eval_steps_per_second": 3.153,
+      "step": 6000
+    },
+    {
+      "epoch": 0.543224404815787,
+      "grad_norm": 0.8109533786773682,
+      "learning_rate": 7.027766180240592e-05,
+      "loss": 2.7178,
+      "step": 6001
+    },
+    {
+      "epoch": 0.5433149271295374,
+      "grad_norm": 0.845350980758667,
+      "learning_rate": 7.024736413067796e-05,
+      "loss": 2.7571,
+      "step": 6002
+    },
+    {
+      "epoch": 0.5434054494432877,
+      "grad_norm": 0.82073575258255,
+      "learning_rate": 7.02170694550396e-05,
+      "loss": 2.7131,
+      "step": 6003
+    },
+    {
+      "epoch": 0.5434959717570381,
+      "grad_norm": 0.8369652032852173,
+      "learning_rate": 7.018677777854157e-05,
+      "loss": 2.7398,
+      "step": 6004
+    },
+    {
+      "epoch": 0.5435864940707884,
+      "grad_norm": 0.8532238602638245,
+      "learning_rate": 7.015648910423416e-05,
+      "loss": 2.7892,
+      "step": 6005
+    },
+    {
+      "epoch": 0.5436770163845388,
+      "grad_norm": 0.7869599461555481,
+      "learning_rate": 7.012620343516751e-05,
+      "loss": 2.6773,
+      "step": 6006
+    },
+    {
+      "epoch": 0.5437675386982891,
+      "grad_norm": 0.820217490196228,
+      "learning_rate": 7.009592077439134e-05,
+      "loss": 2.7094,
+      "step": 6007
+    },
+    {
+      "epoch": 0.5438580610120395,
+      "grad_norm": 0.8736022710800171,
+      "learning_rate": 7.006564112495516e-05,
+      "loss": 2.7607,
+      "step": 6008
+    },
+    {
+      "epoch": 0.5439485833257898,
+      "grad_norm": 0.8576371669769287,
+      "learning_rate": 7.003536448990804e-05,
+      "loss": 2.7191,
+      "step": 6009
+    },
+    {
+      "epoch": 0.5440391056395402,
+      "grad_norm": 0.8501553535461426,
+      "learning_rate": 7.000509087229895e-05,
+      "loss": 2.7255,
+      "step": 6010
+    },
+    {
+      "epoch": 0.5441296279532905,
+      "grad_norm": 0.8708006143569946,
+      "learning_rate": 6.997482027517637e-05,
+      "loss": 2.7492,
+      "step": 6011
+    },
+    {
+      "epoch": 0.5442201502670408,
+      "grad_norm": 0.9946036338806152,
+      "learning_rate": 6.994455270158858e-05,
+      "loss": 2.7547,
+      "step": 6012
+    },
+    {
+      "epoch": 0.5443106725807911,
+      "grad_norm": 0.8790095448493958,
+      "learning_rate": 6.991428815458347e-05,
+      "loss": 2.8004,
+      "step": 6013
+    },
+    {
+      "epoch": 0.5444011948945415,
+      "grad_norm": 0.8968866467475891,
+      "learning_rate": 6.988402663720877e-05,
+      "loss": 2.7249,
+      "step": 6014
+    },
+    {
+      "epoch": 0.5444917172082918,
+      "grad_norm": 0.847731351852417,
+      "learning_rate": 6.985376815251173e-05,
+      "loss": 2.6742,
+      "step": 6015
+    },
+    {
+      "epoch": 0.5445822395220422,
+      "grad_norm": 0.9313018918037415,
+      "learning_rate": 6.982351270353944e-05,
+      "loss": 2.6773,
+      "step": 6016
+    },
+    {
+      "epoch": 0.5446727618357925,
+      "grad_norm": 0.8626921772956848,
+      "learning_rate": 6.979326029333855e-05,
+      "loss": 2.6568,
+      "step": 6017
+    },
+    {
+      "epoch": 0.5447632841495429,
+      "grad_norm": 0.873760461807251,
+      "learning_rate": 6.976301092495556e-05,
+      "loss": 2.7645,
+      "step": 6018
+    },
+    {
+      "epoch": 0.5448538064632932,
+      "grad_norm": 0.8684285283088684,
+      "learning_rate": 6.973276460143653e-05,
+      "loss": 2.7056,
+      "step": 6019
+    },
+    {
+      "epoch": 0.5449443287770436,
+      "grad_norm": 0.8305872082710266,
+      "learning_rate": 6.970252132582728e-05,
+      "loss": 2.7385,
+      "step": 6020
+    },
+    {
+      "epoch": 0.5450348510907939,
+      "grad_norm": 0.8480159640312195,
+      "learning_rate": 6.967228110117328e-05,
+      "loss": 2.6805,
+      "step": 6021
+    },
+    {
+      "epoch": 0.5451253734045443,
+      "grad_norm": 0.8201212286949158,
+      "learning_rate": 6.964204393051981e-05,
+      "loss": 2.7454,
+      "step": 6022
+    },
+    {
+      "epoch": 0.5452158957182945,
+      "grad_norm": 0.8432485461235046,
+      "learning_rate": 6.961180981691162e-05,
+      "loss": 2.7223,
+      "step": 6023
+    },
+    {
+      "epoch": 0.5453064180320449,
+      "grad_norm": 0.9415342211723328,
+      "learning_rate": 6.958157876339337e-05,
+      "loss": 2.7548,
+      "step": 6024
+    },
+    {
+      "epoch": 0.5453969403457952,
+      "grad_norm": 0.8638390302658081,
+      "learning_rate": 6.955135077300931e-05,
+      "loss": 2.755,
+      "step": 6025
+    },
+    {
+      "epoch": 0.5454874626595456,
+      "grad_norm": 0.9359387159347534,
+      "learning_rate": 6.952112584880341e-05,
+      "loss": 2.7935,
+      "step": 6026
+    },
+    {
+      "epoch": 0.5455779849732959,
+      "grad_norm": 0.8699626326560974,
+      "learning_rate": 6.949090399381925e-05,
+      "loss": 2.7137,
+      "step": 6027
+    },
+    {
+      "epoch": 0.5456685072870463,
+      "grad_norm": 0.8389036655426025,
+      "learning_rate": 6.946068521110028e-05,
+      "loss": 2.7469,
+      "step": 6028
+    },
+    {
+      "epoch": 0.5457590296007966,
+      "grad_norm": 0.8703330159187317,
+      "learning_rate": 6.943046950368944e-05,
+      "loss": 2.7258,
+      "step": 6029
+    },
+    {
+      "epoch": 0.545849551914547,
+      "grad_norm": 0.8876437544822693,
+      "learning_rate": 6.940025687462952e-05,
+      "loss": 2.7863,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5459400742282973,
+      "grad_norm": 0.8767717480659485,
+      "learning_rate": 6.937004732696285e-05,
+      "loss": 2.7403,
+      "step": 6031
+    },
+    {
+      "epoch": 0.5460305965420477,
+      "grad_norm": 0.8453347086906433,
+      "learning_rate": 6.93398408637316e-05,
+      "loss": 2.7127,
+      "step": 6032
+    },
+    {
+      "epoch": 0.546121118855798,
+      "grad_norm": 0.8377524018287659,
+      "learning_rate": 6.930963748797754e-05,
+      "loss": 2.7375,
+      "step": 6033
+    },
+    {
+      "epoch": 0.5462116411695483,
+      "grad_norm": 0.8833092451095581,
+      "learning_rate": 6.927943720274215e-05,
+      "loss": 2.7563,
+      "step": 6034
+    },
+    {
+      "epoch": 0.5463021634832986,
+      "grad_norm": 0.8684247732162476,
+      "learning_rate": 6.924924001106654e-05,
+      "loss": 2.7916,
+      "step": 6035
+    },
+    {
+      "epoch": 0.546392685797049,
+      "grad_norm": 0.8185412287712097,
+      "learning_rate": 6.921904591599169e-05,
+      "loss": 2.6799,
+      "step": 6036
+    },
+    {
+      "epoch": 0.5464832081107993,
+      "grad_norm": 0.8553128838539124,
+      "learning_rate": 6.918885492055803e-05,
+      "loss": 2.6648,
+      "step": 6037
+    },
+    {
+      "epoch": 0.5465737304245496,
+      "grad_norm": 0.8351742625236511,
+      "learning_rate": 6.915866702780586e-05,
+      "loss": 2.7457,
+      "step": 6038
+    },
+    {
+      "epoch": 0.5466642527383,
+      "grad_norm": 0.8689993619918823,
+      "learning_rate": 6.912848224077506e-05,
+      "loss": 2.691,
+      "step": 6039
+    },
+    {
+      "epoch": 0.5467547750520503,
+      "grad_norm": 0.788202166557312,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 2.6985,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5468452973658007,
+      "grad_norm": 0.8690296411514282,
+      "learning_rate": 6.906812199603575e-05,
+      "loss": 2.7368,
+      "step": 6041
+    },
+    {
+      "epoch": 0.546935819679551,
+      "grad_norm": 0.8578805327415466,
+      "learning_rate": 6.903794654440551e-05,
+      "loss": 2.7074,
+      "step": 6042
+    },
+    {
+      "epoch": 0.5470263419933014,
+      "grad_norm": 0.8935035467147827,
+      "learning_rate": 6.900777421065319e-05,
+      "loss": 2.7643,
+      "step": 6043
+    },
+    {
+      "epoch": 0.5471168643070516,
+      "grad_norm": 0.7811352610588074,
+      "learning_rate": 6.897760499781718e-05,
+      "loss": 2.6852,
+      "step": 6044
+    },
+    {
+      "epoch": 0.547207386620802,
+      "grad_norm": 0.7939018607139587,
+      "learning_rate": 6.894743890893548e-05,
+      "loss": 2.7098,
+      "step": 6045
+    },
+    {
+      "epoch": 0.5472979089345523,
+      "grad_norm": 0.8887366056442261,
+      "learning_rate": 6.891727594704587e-05,
+      "loss": 2.7744,
+      "step": 6046
+    },
+    {
+      "epoch": 0.5473884312483027,
+      "grad_norm": 0.8419579267501831,
+      "learning_rate": 6.888711611518568e-05,
+      "loss": 2.7,
+      "step": 6047
+    },
+    {
+      "epoch": 0.547478953562053,
+      "grad_norm": 0.771739661693573,
+      "learning_rate": 6.885695941639207e-05,
+      "loss": 2.6992,
+      "step": 6048
+    },
+    {
+      "epoch": 0.5475694758758034,
+      "grad_norm": 0.9229654669761658,
+      "learning_rate": 6.882680585370178e-05,
+      "loss": 2.7491,
+      "step": 6049
+    },
+    {
+      "epoch": 0.5476599981895537,
+      "grad_norm": 0.8693364858627319,
+      "learning_rate": 6.87966554301513e-05,
+      "loss": 2.7429,
+      "step": 6050
+    },
+    {
+      "epoch": 0.5477505205033041,
+      "grad_norm": 0.8253883719444275,
+      "learning_rate": 6.876650814877674e-05,
+      "loss": 2.6733,
+      "step": 6051
+    },
+    {
+      "epoch": 0.5478410428170544,
+      "grad_norm": 0.8329223394393921,
+      "learning_rate": 6.873636401261401e-05,
+      "loss": 2.7187,
+      "step": 6052
+    },
+    {
+      "epoch": 0.5479315651308048,
+      "grad_norm": 0.8911111354827881,
+      "learning_rate": 6.87062230246985e-05,
+      "loss": 2.7515,
+      "step": 6053
+    },
+    {
+      "epoch": 0.548022087444555,
+      "grad_norm": 0.8329285979270935,
+      "learning_rate": 6.867608518806552e-05,
+      "loss": 2.7343,
+      "step": 6054
+    },
+    {
+      "epoch": 0.5481126097583054,
+      "grad_norm": 0.8729497194290161,
+      "learning_rate": 6.864595050574989e-05,
+      "loss": 2.7164,
+      "step": 6055
+    },
+    {
+      "epoch": 0.5482031320720557,
+      "grad_norm": 0.9158942699432373,
+      "learning_rate": 6.861581898078619e-05,
+      "loss": 2.7189,
+      "step": 6056
+    },
+    {
+      "epoch": 0.5482936543858061,
+      "grad_norm": 0.8531440496444702,
+      "learning_rate": 6.858569061620862e-05,
+      "loss": 2.6761,
+      "step": 6057
+    },
+    {
+      "epoch": 0.5483841766995564,
+      "grad_norm": 0.8701982498168945,
+      "learning_rate": 6.855556541505122e-05,
+      "loss": 2.7406,
+      "step": 6058
+    },
+    {
+      "epoch": 0.5484746990133068,
+      "grad_norm": 0.8322367668151855,
+      "learning_rate": 6.852544338034748e-05,
+      "loss": 2.7331,
+      "step": 6059
+    },
+    {
+      "epoch": 0.5485652213270571,
+      "grad_norm": 0.8104520440101624,
+      "learning_rate": 6.849532451513074e-05,
+      "loss": 2.6839,
+      "step": 6060
+    },
+    {
+      "epoch": 0.5486557436408075,
+      "grad_norm": 0.8525737524032593,
+      "learning_rate": 6.846520882243393e-05,
+      "loss": 2.6964,
+      "step": 6061
+    },
+    {
+      "epoch": 0.5487462659545578,
+      "grad_norm": 0.8493409156799316,
+      "learning_rate": 6.843509630528977e-05,
+      "loss": 2.756,
+      "step": 6062
+    },
+    {
+      "epoch": 0.5488367882683082,
+      "grad_norm": 0.829964816570282,
+      "learning_rate": 6.840498696673053e-05,
+      "loss": 2.7348,
+      "step": 6063
+    },
+    {
+      "epoch": 0.5489273105820585,
+      "grad_norm": 0.8463034629821777,
+      "learning_rate": 6.837488080978824e-05,
+      "loss": 2.6979,
+      "step": 6064
+    },
+    {
+      "epoch": 0.5490178328958089,
+      "grad_norm": 0.8787871599197388,
+      "learning_rate": 6.834477783749455e-05,
+      "loss": 2.7413,
+      "step": 6065
+    },
+    {
+      "epoch": 0.5491083552095591,
+      "grad_norm": 0.8388311862945557,
+      "learning_rate": 6.831467805288092e-05,
+      "loss": 2.6772,
+      "step": 6066
+    },
+    {
+      "epoch": 0.5491988775233095,
+      "grad_norm": 0.8327917456626892,
+      "learning_rate": 6.828458145897832e-05,
+      "loss": 2.6996,
+      "step": 6067
+    },
+    {
+      "epoch": 0.5492893998370598,
+      "grad_norm": 0.9191029071807861,
+      "learning_rate": 6.825448805881752e-05,
+      "loss": 2.7829,
+      "step": 6068
+    },
+    {
+      "epoch": 0.5493799221508102,
+      "grad_norm": 0.7828680872917175,
+      "learning_rate": 6.822439785542886e-05,
+      "loss": 2.7264,
+      "step": 6069
+    },
+    {
+      "epoch": 0.5494704444645605,
+      "grad_norm": 0.8679816722869873,
+      "learning_rate": 6.819431085184251e-05,
+      "loss": 2.689,
+      "step": 6070
+    },
+    {
+      "epoch": 0.5495609667783109,
+      "grad_norm": 0.913976788520813,
+      "learning_rate": 6.816422705108817e-05,
+      "loss": 2.7366,
+      "step": 6071
+    },
+    {
+      "epoch": 0.5496514890920612,
+      "grad_norm": 0.7898628115653992,
+      "learning_rate": 6.813414645619531e-05,
+      "loss": 2.6953,
+      "step": 6072
+    },
+    {
+      "epoch": 0.5497420114058116,
+      "grad_norm": 0.8191077709197998,
+      "learning_rate": 6.8104069070193e-05,
+      "loss": 2.6626,
+      "step": 6073
+    },
+    {
+      "epoch": 0.5498325337195619,
+      "grad_norm": 0.8287845849990845,
+      "learning_rate": 6.807399489611009e-05,
+      "loss": 2.6873,
+      "step": 6074
+    },
+    {
+      "epoch": 0.5499230560333123,
+      "grad_norm": 0.8519953489303589,
+      "learning_rate": 6.804392393697502e-05,
+      "loss": 2.7353,
+      "step": 6075
+    },
+    {
+      "epoch": 0.5500135783470625,
+      "grad_norm": 0.9047031402587891,
+      "learning_rate": 6.801385619581592e-05,
+      "loss": 2.6962,
+      "step": 6076
+    },
+    {
+      "epoch": 0.550104100660813,
+      "grad_norm": 0.8621162176132202,
+      "learning_rate": 6.798379167566064e-05,
+      "loss": 2.6707,
+      "step": 6077
+    },
+    {
+      "epoch": 0.5501946229745632,
+      "grad_norm": 0.8716885447502136,
+      "learning_rate": 6.795373037953663e-05,
+      "loss": 2.7579,
+      "step": 6078
+    },
+    {
+      "epoch": 0.5502851452883135,
+      "grad_norm": 0.777885377407074,
+      "learning_rate": 6.792367231047112e-05,
+      "loss": 2.6918,
+      "step": 6079
+    },
+    {
+      "epoch": 0.5503756676020639,
+      "grad_norm": 0.9372280836105347,
+      "learning_rate": 6.789361747149093e-05,
+      "loss": 2.7738,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5504661899158142,
+      "grad_norm": 0.8034177422523499,
+      "learning_rate": 6.786356586562259e-05,
+      "loss": 2.6903,
+      "step": 6081
+    },
+    {
+      "epoch": 0.5505567122295646,
+      "grad_norm": 0.7964674234390259,
+      "learning_rate": 6.783351749589225e-05,
+      "loss": 2.6633,
+      "step": 6082
+    },
+    {
+      "epoch": 0.5506472345433149,
+      "grad_norm": 0.8511160612106323,
+      "learning_rate": 6.780347236532588e-05,
+      "loss": 2.7443,
+      "step": 6083
+    },
+    {
+      "epoch": 0.5507377568570653,
+      "grad_norm": 0.8763010501861572,
+      "learning_rate": 6.77734304769489e-05,
+      "loss": 2.7221,
+      "step": 6084
+    },
+    {
+      "epoch": 0.5508282791708156,
+      "grad_norm": 0.8387385606765747,
+      "learning_rate": 6.774339183378663e-05,
+      "loss": 2.7448,
+      "step": 6085
+    },
+    {
+      "epoch": 0.550918801484566,
+      "grad_norm": 0.8287498950958252,
+      "learning_rate": 6.771335643886389e-05,
+      "loss": 2.6757,
+      "step": 6086
+    },
+    {
+      "epoch": 0.5510093237983162,
+      "grad_norm": 0.8377882242202759,
+      "learning_rate": 6.768332429520532e-05,
+      "loss": 2.6849,
+      "step": 6087
+    },
+    {
+      "epoch": 0.5510998461120666,
+      "grad_norm": 0.8893755078315735,
+      "learning_rate": 6.765329540583504e-05,
+      "loss": 2.7471,
+      "step": 6088
+    },
+    {
+      "epoch": 0.5511903684258169,
+      "grad_norm": 1.009269118309021,
+      "learning_rate": 6.762326977377709e-05,
+      "loss": 2.6596,
+      "step": 6089
+    },
+    {
+      "epoch": 0.5512808907395673,
+      "grad_norm": 0.9541810154914856,
+      "learning_rate": 6.759324740205495e-05,
+      "loss": 2.7035,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5513714130533176,
+      "grad_norm": 0.808933436870575,
+      "learning_rate": 6.756322829369195e-05,
+      "loss": 2.7264,
+      "step": 6091
+    },
+    {
+      "epoch": 0.551461935367068,
+      "grad_norm": 0.9128525257110596,
+      "learning_rate": 6.753321245171092e-05,
+      "loss": 2.7286,
+      "step": 6092
+    },
+    {
+      "epoch": 0.5515524576808183,
+      "grad_norm": 0.9354084134101868,
+      "learning_rate": 6.750319987913453e-05,
+      "loss": 2.7369,
+      "step": 6093
+    },
+    {
+      "epoch": 0.5516429799945687,
+      "grad_norm": 0.8727840781211853,
+      "learning_rate": 6.747319057898503e-05,
+      "loss": 2.7014,
+      "step": 6094
+    },
+    {
+      "epoch": 0.551733502308319,
+      "grad_norm": 0.8667216897010803,
+      "learning_rate": 6.744318455428436e-05,
+      "loss": 2.7354,
+      "step": 6095
+    },
+    {
+      "epoch": 0.5518240246220694,
+      "grad_norm": 1.0342938899993896,
+      "learning_rate": 6.741318180805407e-05,
+      "loss": 2.7443,
+      "step": 6096
+    },
+    {
+      "epoch": 0.5519145469358196,
+      "grad_norm": 0.7890323400497437,
+      "learning_rate": 6.738318234331554e-05,
+      "loss": 2.6711,
+      "step": 6097
+    },
+    {
+      "epoch": 0.55200506924957,
+      "grad_norm": 0.8307895064353943,
+      "learning_rate": 6.735318616308961e-05,
+      "loss": 2.6776,
+      "step": 6098
+    },
+    {
+      "epoch": 0.5520955915633203,
+      "grad_norm": 0.8395517468452454,
+      "learning_rate": 6.732319327039698e-05,
+      "loss": 2.6839,
+      "step": 6099
+    },
+    {
+      "epoch": 0.5521861138770707,
+      "grad_norm": 0.855904221534729,
+      "learning_rate": 6.729320366825784e-05,
+      "loss": 2.7223,
+      "step": 6100
+    },
+    {
+      "epoch": 0.552276636190821,
+      "grad_norm": 0.8447253108024597,
+      "learning_rate": 6.726321735969223e-05,
+      "loss": 2.7251,
+      "step": 6101
+    },
+    {
+      "epoch": 0.5523671585045714,
+      "grad_norm": 0.8225886225700378,
+      "learning_rate": 6.723323434771974e-05,
+      "loss": 2.7384,
+      "step": 6102
+    },
+    {
+      "epoch": 0.5524576808183217,
+      "grad_norm": 0.869172990322113,
+      "learning_rate": 6.720325463535966e-05,
+      "loss": 2.7413,
+      "step": 6103
+    },
+    {
+      "epoch": 0.5525482031320721,
+      "grad_norm": 0.8755947947502136,
+      "learning_rate": 6.717327822563089e-05,
+      "loss": 2.7926,
+      "step": 6104
+    },
+    {
+      "epoch": 0.5526387254458224,
+      "grad_norm": 0.8883076310157776,
+      "learning_rate": 6.714330512155216e-05,
+      "loss": 2.7569,
+      "step": 6105
+    },
+    {
+      "epoch": 0.5527292477595728,
+      "grad_norm": 0.8901398181915283,
+      "learning_rate": 6.711333532614168e-05,
+      "loss": 2.7178,
+      "step": 6106
+    },
+    {
+      "epoch": 0.5528197700733231,
+      "grad_norm": 0.856147050857544,
+      "learning_rate": 6.708336884241746e-05,
+      "loss": 2.6511,
+      "step": 6107
+    },
+    {
+      "epoch": 0.5529102923870735,
+      "grad_norm": 0.834158718585968,
+      "learning_rate": 6.705340567339704e-05,
+      "loss": 2.6735,
+      "step": 6108
+    },
+    {
+      "epoch": 0.5530008147008237,
+      "grad_norm": 0.8236615657806396,
+      "learning_rate": 6.702344582209782e-05,
+      "loss": 2.7029,
+      "step": 6109
+    },
+    {
+      "epoch": 0.5530913370145741,
+      "grad_norm": 0.8336913585662842,
+      "learning_rate": 6.699348929153668e-05,
+      "loss": 2.6636,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5531818593283244,
+      "grad_norm": 0.9094150066375732,
+      "learning_rate": 6.69635360847303e-05,
+      "loss": 2.7634,
+      "step": 6111
+    },
+    {
+      "epoch": 0.5532723816420748,
+      "grad_norm": 0.9049180150032043,
+      "learning_rate": 6.693358620469487e-05,
+      "loss": 2.7459,
+      "step": 6112
+    },
+    {
+      "epoch": 0.5533629039558251,
+      "grad_norm": 0.8091766834259033,
+      "learning_rate": 6.690363965444646e-05,
+      "loss": 2.7064,
+      "step": 6113
+    },
+    {
+      "epoch": 0.5534534262695755,
+      "grad_norm": 0.8895435929298401,
+      "learning_rate": 6.68736964370006e-05,
+      "loss": 2.7459,
+      "step": 6114
+    },
+    {
+      "epoch": 0.5535439485833258,
+      "grad_norm": 0.8452686071395874,
+      "learning_rate": 6.684375655537263e-05,
+      "loss": 2.7731,
+      "step": 6115
+    },
+    {
+      "epoch": 0.5536344708970762,
+      "grad_norm": 0.8960869312286377,
+      "learning_rate": 6.681382001257744e-05,
+      "loss": 2.6751,
+      "step": 6116
+    },
+    {
+      "epoch": 0.5537249932108265,
+      "grad_norm": 0.8500857949256897,
+      "learning_rate": 6.67838868116297e-05,
+      "loss": 2.7669,
+      "step": 6117
+    },
+    {
+      "epoch": 0.5538155155245769,
+      "grad_norm": 0.9451707005500793,
+      "learning_rate": 6.675395695554359e-05,
+      "loss": 2.6578,
+      "step": 6118
+    },
+    {
+      "epoch": 0.5539060378383271,
+      "grad_norm": 0.8644303679466248,
+      "learning_rate": 6.672403044733317e-05,
+      "loss": 2.7596,
+      "step": 6119
+    },
+    {
+      "epoch": 0.5539965601520774,
+      "grad_norm": 0.7932984232902527,
+      "learning_rate": 6.669410729001193e-05,
+      "loss": 2.7395,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5540870824658278,
+      "grad_norm": 0.9047853946685791,
+      "learning_rate": 6.666418748659321e-05,
+      "loss": 2.777,
+      "step": 6121
+    },
+    {
+      "epoch": 0.5541776047795781,
+      "grad_norm": 0.8457450866699219,
+      "learning_rate": 6.663427104008984e-05,
+      "loss": 2.7113,
+      "step": 6122
+    },
+    {
+      "epoch": 0.5542681270933285,
+      "grad_norm": 0.9217113256454468,
+      "learning_rate": 6.660435795351451e-05,
+      "loss": 2.7712,
+      "step": 6123
+    },
+    {
+      "epoch": 0.5543586494070788,
+      "grad_norm": 1.0403680801391602,
+      "learning_rate": 6.657444822987942e-05,
+      "loss": 2.7665,
+      "step": 6124
+    },
+    {
+      "epoch": 0.5544491717208292,
+      "grad_norm": 0.7791826128959656,
+      "learning_rate": 6.654454187219649e-05,
+      "loss": 2.6847,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5545396940345795,
+      "grad_norm": 0.8778238296508789,
+      "learning_rate": 6.651463888347722e-05,
+      "loss": 2.7482,
+      "step": 6126
+    },
+    {
+      "epoch": 0.5546302163483299,
+      "grad_norm": 0.7843742370605469,
+      "learning_rate": 6.648473926673295e-05,
+      "loss": 2.744,
+      "step": 6127
+    },
+    {
+      "epoch": 0.5547207386620802,
+      "grad_norm": 0.959306538105011,
+      "learning_rate": 6.64548430249745e-05,
+      "loss": 2.7685,
+      "step": 6128
+    },
+    {
+      "epoch": 0.5548112609758306,
+      "grad_norm": 0.8497962355613708,
+      "learning_rate": 6.642495016121246e-05,
+      "loss": 2.731,
+      "step": 6129
+    },
+    {
+      "epoch": 0.5549017832895808,
+      "grad_norm": 0.7902641892433167,
+      "learning_rate": 6.639506067845697e-05,
+      "loss": 2.7404,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5549923056033312,
+      "grad_norm": 0.910959780216217,
+      "learning_rate": 6.636517457971801e-05,
+      "loss": 2.7076,
+      "step": 6131
+    },
+    {
+      "epoch": 0.5550828279170815,
+      "grad_norm": 0.9271765351295471,
+      "learning_rate": 6.633529186800502e-05,
+      "loss": 2.7268,
+      "step": 6132
+    },
+    {
+      "epoch": 0.5551733502308319,
+      "grad_norm": 0.8239401578903198,
+      "learning_rate": 6.630541254632724e-05,
+      "loss": 2.7182,
+      "step": 6133
+    },
+    {
+      "epoch": 0.5552638725445822,
+      "grad_norm": 0.852472186088562,
+      "learning_rate": 6.627553661769345e-05,
+      "loss": 2.6915,
+      "step": 6134
+    },
+    {
+      "epoch": 0.5553543948583326,
+      "grad_norm": 0.8380687832832336,
+      "learning_rate": 6.624566408511223e-05,
+      "loss": 2.6643,
+      "step": 6135
+    },
+    {
+      "epoch": 0.5554449171720829,
+      "grad_norm": 0.7780315279960632,
+      "learning_rate": 6.621579495159171e-05,
+      "loss": 2.6831,
+      "step": 6136
+    },
+    {
+      "epoch": 0.5555354394858333,
+      "grad_norm": 0.9226695895195007,
+      "learning_rate": 6.618592922013973e-05,
+      "loss": 2.6992,
+      "step": 6137
+    },
+    {
+      "epoch": 0.5556259617995836,
+      "grad_norm": 0.894949734210968,
+      "learning_rate": 6.61560668937637e-05,
+      "loss": 2.7078,
+      "step": 6138
+    },
+    {
+      "epoch": 0.555716484113334,
+      "grad_norm": 0.911466658115387,
+      "learning_rate": 6.612620797547087e-05,
+      "loss": 2.7357,
+      "step": 6139
+    },
+    {
+      "epoch": 0.5558070064270842,
+      "grad_norm": 0.9253066778182983,
+      "learning_rate": 6.609635246826794e-05,
+      "loss": 2.7053,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5558975287408346,
+      "grad_norm": 0.8521486520767212,
+      "learning_rate": 6.60665003751614e-05,
+      "loss": 2.76,
+      "step": 6141
+    },
+    {
+      "epoch": 0.5559880510545849,
+      "grad_norm": 0.7954798340797424,
+      "learning_rate": 6.603665169915732e-05,
+      "loss": 2.6759,
+      "step": 6142
+    },
+    {
+      "epoch": 0.5560785733683353,
+      "grad_norm": 0.9189724326133728,
+      "learning_rate": 6.600680644326151e-05,
+      "loss": 2.7218,
+      "step": 6143
+    },
+    {
+      "epoch": 0.5561690956820856,
+      "grad_norm": 0.8937711119651794,
+      "learning_rate": 6.597696461047934e-05,
+      "loss": 2.7618,
+      "step": 6144
+    },
+    {
+      "epoch": 0.556259617995836,
+      "grad_norm": 0.8732416033744812,
+      "learning_rate": 6.594712620381594e-05,
+      "loss": 2.7768,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5563501403095863,
+      "grad_norm": 0.8561906218528748,
+      "learning_rate": 6.591729122627595e-05,
+      "loss": 2.7299,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5564406626233367,
+      "grad_norm": 0.959690272808075,
+      "learning_rate": 6.588745968086385e-05,
+      "loss": 2.7445,
+      "step": 6147
+    },
+    {
+      "epoch": 0.556531184937087,
+      "grad_norm": 0.8490232229232788,
+      "learning_rate": 6.585763157058358e-05,
+      "loss": 2.7148,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5566217072508374,
+      "grad_norm": 0.9139682650566101,
+      "learning_rate": 6.58278068984389e-05,
+      "loss": 2.676,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5567122295645877,
+      "grad_norm": 0.8509264588356018,
+      "learning_rate": 6.579798566743314e-05,
+      "loss": 2.7449,
+      "step": 6150
+    },
+    {
+      "epoch": 0.556802751878338,
+      "grad_norm": 0.8683857917785645,
+      "learning_rate": 6.576816788056928e-05,
+      "loss": 2.7018,
+      "step": 6151
+    },
+    {
+      "epoch": 0.5568932741920883,
+      "grad_norm": 0.9115766882896423,
+      "learning_rate": 6.573835354084996e-05,
+      "loss": 2.7474,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5569837965058387,
+      "grad_norm": 0.9348102807998657,
+      "learning_rate": 6.570854265127753e-05,
+      "loss": 2.7245,
+      "step": 6153
+    },
+    {
+      "epoch": 0.557074318819589,
+      "grad_norm": 0.8374444842338562,
+      "learning_rate": 6.567873521485389e-05,
+      "loss": 2.6585,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5571648411333394,
+      "grad_norm": 0.9902597665786743,
+      "learning_rate": 6.564893123458069e-05,
+      "loss": 2.7539,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5572553634470897,
+      "grad_norm": 0.9274411201477051,
+      "learning_rate": 6.561913071345915e-05,
+      "loss": 2.6765,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5573458857608401,
+      "grad_norm": 0.8324682712554932,
+      "learning_rate": 6.558933365449025e-05,
+      "loss": 2.711,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5574364080745904,
+      "grad_norm": 0.8855867385864258,
+      "learning_rate": 6.555954006067447e-05,
+      "loss": 2.6812,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5575269303883408,
+      "grad_norm": 0.8500581979751587,
+      "learning_rate": 6.552974993501211e-05,
+      "loss": 2.7387,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5576174527020911,
+      "grad_norm": 0.8339815735816956,
+      "learning_rate": 6.549996328050296e-05,
+      "loss": 2.7213,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5577079750158414,
+      "grad_norm": 0.9083566069602966,
+      "learning_rate": 6.547018010014654e-05,
+      "loss": 2.7218,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5577984973295917,
+      "grad_norm": 0.8667166829109192,
+      "learning_rate": 6.544040039694208e-05,
+      "loss": 2.7232,
+      "step": 6162
+    },
+    {
+      "epoch": 0.557889019643342,
+      "grad_norm": 0.8151534795761108,
+      "learning_rate": 6.541062417388833e-05,
+      "loss": 2.6772,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5579795419570924,
+      "grad_norm": 0.8102704286575317,
+      "learning_rate": 6.53808514339838e-05,
+      "loss": 2.6715,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5580700642708427,
+      "grad_norm": 0.9075278043746948,
+      "learning_rate": 6.535108218022654e-05,
+      "loss": 2.7122,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5581605865845931,
+      "grad_norm": 0.8931668996810913,
+      "learning_rate": 6.53213164156144e-05,
+      "loss": 2.7641,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5582511088983434,
+      "grad_norm": 0.8440109491348267,
+      "learning_rate": 6.529155414314472e-05,
+      "loss": 2.7849,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5583416312120938,
+      "grad_norm": 0.8298215270042419,
+      "learning_rate": 6.526179536581463e-05,
+      "loss": 2.7624,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5584321535258441,
+      "grad_norm": 0.9021025896072388,
+      "learning_rate": 6.523204008662074e-05,
+      "loss": 2.7398,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5585226758395945,
+      "grad_norm": 0.8510945439338684,
+      "learning_rate": 6.52022883085595e-05,
+      "loss": 2.7218,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5586131981533448,
+      "grad_norm": 0.8765257596969604,
+      "learning_rate": 6.517254003462686e-05,
+      "loss": 2.7187,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5587037204670952,
+      "grad_norm": 0.8824365735054016,
+      "learning_rate": 6.51427952678185e-05,
+      "loss": 2.7178,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5587942427808454,
+      "grad_norm": 0.8940009474754333,
+      "learning_rate": 6.511305401112967e-05,
+      "loss": 2.7676,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5588847650945958,
+      "grad_norm": 0.8891744613647461,
+      "learning_rate": 6.508331626755538e-05,
+      "loss": 2.6589,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5589752874083461,
+      "grad_norm": 0.8407256603240967,
+      "learning_rate": 6.505358204009017e-05,
+      "loss": 2.7429,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5590658097220965,
+      "grad_norm": 0.8492648005485535,
+      "learning_rate": 6.502385133172833e-05,
+      "loss": 2.7259,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5591563320358468,
+      "grad_norm": 1.0108580589294434,
+      "learning_rate": 6.499412414546362e-05,
+      "loss": 2.7615,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5592468543495972,
+      "grad_norm": 0.8850505948066711,
+      "learning_rate": 6.496440048428976e-05,
+      "loss": 2.723,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5593373766633475,
+      "grad_norm": 0.8815732598304749,
+      "learning_rate": 6.493468035119972e-05,
+      "loss": 2.6943,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5594278989770979,
+      "grad_norm": 0.8840451240539551,
+      "learning_rate": 6.490496374918647e-05,
+      "loss": 2.714,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5595184212908482,
+      "grad_norm": 0.8313421010971069,
+      "learning_rate": 6.487525068124237e-05,
+      "loss": 2.7207,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5596089436045986,
+      "grad_norm": 0.8717589378356934,
+      "learning_rate": 6.48455411503596e-05,
+      "loss": 2.7393,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5596994659183488,
+      "grad_norm": 0.9375311732292175,
+      "learning_rate": 6.481583515952983e-05,
+      "loss": 2.7661,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5597899882320992,
+      "grad_norm": 0.9299476146697998,
+      "learning_rate": 6.478613271174453e-05,
+      "loss": 2.7136,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5598805105458495,
+      "grad_norm": 0.9196488857269287,
+      "learning_rate": 6.475643380999468e-05,
+      "loss": 2.7096,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5599710328595999,
+      "grad_norm": 0.852522075176239,
+      "learning_rate": 6.472673845727102e-05,
+      "loss": 2.6871,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5600615551733502,
+      "grad_norm": 0.9577264785766602,
+      "learning_rate": 6.469704665656378e-05,
+      "loss": 2.8258,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5601520774871006,
+      "grad_norm": 0.9313653111457825,
+      "learning_rate": 6.466735841086302e-05,
+      "loss": 2.69,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5602425998008509,
+      "grad_norm": 0.8006998300552368,
+      "learning_rate": 6.463767372315827e-05,
+      "loss": 2.6518,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5603331221146013,
+      "grad_norm": 0.8366522789001465,
+      "learning_rate": 6.460799259643884e-05,
+      "loss": 2.7591,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5604236444283516,
+      "grad_norm": 0.8412538170814514,
+      "learning_rate": 6.457831503369354e-05,
+      "loss": 2.6933,
+      "step": 6191
+    },
+    {
+      "epoch": 0.560514166742102,
+      "grad_norm": 0.8227540254592896,
+      "learning_rate": 6.454864103791099e-05,
+      "loss": 2.7545,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5606046890558523,
+      "grad_norm": 0.8566422462463379,
+      "learning_rate": 6.45189706120793e-05,
+      "loss": 2.7272,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5606952113696027,
+      "grad_norm": 0.8530976176261902,
+      "learning_rate": 6.448930375918631e-05,
+      "loss": 2.6771,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5607857336833529,
+      "grad_norm": 0.8847705125808716,
+      "learning_rate": 6.445964048221944e-05,
+      "loss": 2.7028,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5608762559971033,
+      "grad_norm": 0.9251405000686646,
+      "learning_rate": 6.442998078416583e-05,
+      "loss": 2.7475,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5609667783108536,
+      "grad_norm": 0.8360767364501953,
+      "learning_rate": 6.440032466801215e-05,
+      "loss": 2.7187,
+      "step": 6197
+    },
+    {
+      "epoch": 0.561057300624604,
+      "grad_norm": 0.901405930519104,
+      "learning_rate": 6.437067213674483e-05,
+      "loss": 2.6693,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5611478229383543,
+      "grad_norm": 0.9669973850250244,
+      "learning_rate": 6.434102319334981e-05,
+      "loss": 2.7424,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5612383452521047,
+      "grad_norm": 0.8233455419540405,
+      "learning_rate": 6.431137784081282e-05,
+      "loss": 2.6649,
+      "step": 6200
+    },
+    {
+      "epoch": 0.561328867565855,
+      "grad_norm": 0.9279999732971191,
+      "learning_rate": 6.428173608211909e-05,
+      "loss": 2.7304,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5614193898796053,
+      "grad_norm": 0.88908451795578,
+      "learning_rate": 6.425209792025358e-05,
+      "loss": 2.7849,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5615099121933557,
+      "grad_norm": 0.9120383858680725,
+      "learning_rate": 6.422246335820079e-05,
+      "loss": 2.7655,
+      "step": 6203
+    },
+    {
+      "epoch": 0.561600434507106,
+      "grad_norm": 0.8763618469238281,
+      "learning_rate": 6.419283239894499e-05,
+      "loss": 2.7167,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5616909568208563,
+      "grad_norm": 0.8769980669021606,
+      "learning_rate": 6.416320504546997e-05,
+      "loss": 2.7123,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5617814791346066,
+      "grad_norm": 0.8798347115516663,
+      "learning_rate": 6.413358130075925e-05,
+      "loss": 2.7843,
+      "step": 6206
+    },
+    {
+      "epoch": 0.561872001448357,
+      "grad_norm": 0.8606103658676147,
+      "learning_rate": 6.410396116779587e-05,
+      "loss": 2.6994,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5619625237621073,
+      "grad_norm": 0.8350090980529785,
+      "learning_rate": 6.407434464956266e-05,
+      "loss": 2.7007,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5620530460758577,
+      "grad_norm": 0.841839075088501,
+      "learning_rate": 6.404473174904193e-05,
+      "loss": 2.6919,
+      "step": 6209
+    },
+    {
+      "epoch": 0.562143568389608,
+      "grad_norm": 0.9548243880271912,
+      "learning_rate": 6.401512246921576e-05,
+      "loss": 2.7083,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5622340907033584,
+      "grad_norm": 0.8236305117607117,
+      "learning_rate": 6.398551681306575e-05,
+      "loss": 2.7162,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5623246130171087,
+      "grad_norm": 0.8622593879699707,
+      "learning_rate": 6.395591478357324e-05,
+      "loss": 2.7296,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5624151353308591,
+      "grad_norm": 0.8808903694152832,
+      "learning_rate": 6.392631638371907e-05,
+      "loss": 2.7442,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5625056576446094,
+      "grad_norm": 0.84144526720047,
+      "learning_rate": 6.389672161648389e-05,
+      "loss": 2.7346,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5625961799583598,
+      "grad_norm": 0.8263358473777771,
+      "learning_rate": 6.386713048484785e-05,
+      "loss": 2.7278,
+      "step": 6215
+    },
+    {
+      "epoch": 0.56268670227211,
+      "grad_norm": 0.8334097266197205,
+      "learning_rate": 6.383754299179079e-05,
+      "loss": 2.7624,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5627772245858604,
+      "grad_norm": 0.8723674416542053,
+      "learning_rate": 6.380795914029213e-05,
+      "loss": 2.6862,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5628677468996107,
+      "grad_norm": 0.8597888946533203,
+      "learning_rate": 6.377837893333103e-05,
+      "loss": 2.7343,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5629582692133611,
+      "grad_norm": 0.819015383720398,
+      "learning_rate": 6.374880237388616e-05,
+      "loss": 2.6803,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5630487915271114,
+      "grad_norm": 0.9522851705551147,
+      "learning_rate": 6.371922946493591e-05,
+      "loss": 2.693,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5631393138408618,
+      "grad_norm": 0.8460712432861328,
+      "learning_rate": 6.368966020945824e-05,
+      "loss": 2.8005,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5632298361546121,
+      "grad_norm": 0.8678927421569824,
+      "learning_rate": 6.366009461043083e-05,
+      "loss": 2.7346,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5633203584683625,
+      "grad_norm": 0.8084684014320374,
+      "learning_rate": 6.363053267083086e-05,
+      "loss": 2.6989,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5634108807821128,
+      "grad_norm": 0.8358949422836304,
+      "learning_rate": 6.360097439363529e-05,
+      "loss": 2.7444,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5635014030958632,
+      "grad_norm": 0.8385283350944519,
+      "learning_rate": 6.357141978182056e-05,
+      "loss": 2.7455,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5635919254096134,
+      "grad_norm": 0.8100994229316711,
+      "learning_rate": 6.35418688383629e-05,
+      "loss": 2.6919,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5636824477233638,
+      "grad_norm": 0.9156104326248169,
+      "learning_rate": 6.351232156623803e-05,
+      "loss": 2.7142,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5637729700371141,
+      "grad_norm": 0.8520402312278748,
+      "learning_rate": 6.348277796842141e-05,
+      "loss": 2.7497,
+      "step": 6228
+    },
+    {
+      "epoch": 0.5638634923508645,
+      "grad_norm": 0.8622889518737793,
+      "learning_rate": 6.345323804788799e-05,
+      "loss": 2.6753,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5639540146646148,
+      "grad_norm": 0.8826088309288025,
+      "learning_rate": 6.342370180761256e-05,
+      "loss": 2.7352,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5640445369783652,
+      "grad_norm": 0.8279819488525391,
+      "learning_rate": 6.339416925056933e-05,
+      "loss": 2.6723,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5641350592921155,
+      "grad_norm": 0.8228965401649475,
+      "learning_rate": 6.336464037973226e-05,
+      "loss": 2.7167,
+      "step": 6232
+    },
+    {
+      "epoch": 0.5642255816058659,
+      "grad_norm": 0.921067476272583,
+      "learning_rate": 6.333511519807488e-05,
+      "loss": 2.7505,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5643161039196162,
+      "grad_norm": 0.8844332695007324,
+      "learning_rate": 6.330559370857043e-05,
+      "loss": 2.6383,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5644066262333666,
+      "grad_norm": 0.9329710602760315,
+      "learning_rate": 6.327607591419166e-05,
+      "loss": 2.7016,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5644971485471169,
+      "grad_norm": 0.8243615627288818,
+      "learning_rate": 6.324656181791108e-05,
+      "loss": 2.7539,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5645876708608673,
+      "grad_norm": 0.8301962018013,
+      "learning_rate": 6.321705142270067e-05,
+      "loss": 2.6993,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5646781931746175,
+      "grad_norm": 0.8087690472602844,
+      "learning_rate": 6.318754473153221e-05,
+      "loss": 2.7268,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5647687154883679,
+      "grad_norm": 0.8300209641456604,
+      "learning_rate": 6.315804174737697e-05,
+      "loss": 2.688,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5648592378021182,
+      "grad_norm": 0.8376254439353943,
+      "learning_rate": 6.312854247320595e-05,
+      "loss": 2.7136,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5649497601158686,
+      "grad_norm": 0.8355643153190613,
+      "learning_rate": 6.309904691198962e-05,
+      "loss": 2.6754,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5650402824296189,
+      "grad_norm": 0.8193901181221008,
+      "learning_rate": 6.306955506669836e-05,
+      "loss": 2.7298,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5651308047433692,
+      "grad_norm": 0.8663204908370972,
+      "learning_rate": 6.30400669403018e-05,
+      "loss": 2.7622,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5652213270571196,
+      "grad_norm": 0.8417595028877258,
+      "learning_rate": 6.301058253576955e-05,
+      "loss": 2.7069,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5653118493708699,
+      "grad_norm": 0.8490182757377625,
+      "learning_rate": 6.298110185607063e-05,
+      "loss": 2.7275,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5654023716846203,
+      "grad_norm": 0.8474552035331726,
+      "learning_rate": 6.29516249041737e-05,
+      "loss": 2.6955,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5654928939983705,
+      "grad_norm": 0.7780411839485168,
+      "learning_rate": 6.292215168304716e-05,
+      "loss": 2.7219,
+      "step": 6247
+    },
+    {
+      "epoch": 0.565583416312121,
+      "grad_norm": 0.8314462900161743,
+      "learning_rate": 6.28926821956589e-05,
+      "loss": 2.7067,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5656739386258712,
+      "grad_norm": 0.851689875125885,
+      "learning_rate": 6.286321644497655e-05,
+      "loss": 2.7218,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5657644609396216,
+      "grad_norm": 0.841330349445343,
+      "learning_rate": 6.283375443396726e-05,
+      "loss": 2.7342,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5658549832533719,
+      "grad_norm": 0.8683754801750183,
+      "learning_rate": 6.280429616559792e-05,
+      "loss": 2.7366,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5659455055671223,
+      "grad_norm": 0.7926754355430603,
+      "learning_rate": 6.277484164283489e-05,
+      "loss": 2.7295,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5660360278808726,
+      "grad_norm": 0.9038828611373901,
+      "learning_rate": 6.274539086864433e-05,
+      "loss": 2.7524,
+      "step": 6253
+    },
+    {
+      "epoch": 0.566126550194623,
+      "grad_norm": 0.8178255558013916,
+      "learning_rate": 6.271594384599188e-05,
+      "loss": 2.6828,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5662170725083733,
+      "grad_norm": 0.7750669121742249,
+      "learning_rate": 6.268650057784288e-05,
+      "loss": 2.6739,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5663075948221237,
+      "grad_norm": 0.8675525188446045,
+      "learning_rate": 6.26570610671622e-05,
+      "loss": 2.7148,
+      "step": 6256
+    },
+    {
+      "epoch": 0.566398117135874,
+      "grad_norm": 0.928869366645813,
+      "learning_rate": 6.262762531691451e-05,
+      "loss": 2.7006,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5664886394496244,
+      "grad_norm": 0.8617798686027527,
+      "learning_rate": 6.259819333006388e-05,
+      "loss": 2.745,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5665791617633746,
+      "grad_norm": 0.8284051418304443,
+      "learning_rate": 6.25687651095742e-05,
+      "loss": 2.6976,
+      "step": 6259
+    },
+    {
+      "epoch": 0.566669684077125,
+      "grad_norm": 0.9877228736877441,
+      "learning_rate": 6.25393406584088e-05,
+      "loss": 2.7422,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5667602063908753,
+      "grad_norm": 0.889123260974884,
+      "learning_rate": 6.250991997953082e-05,
+      "loss": 2.7538,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5668507287046257,
+      "grad_norm": 0.8846951723098755,
+      "learning_rate": 6.248050307590283e-05,
+      "loss": 2.7475,
+      "step": 6262
+    },
+    {
+      "epoch": 0.566941251018376,
+      "grad_norm": 0.8666096925735474,
+      "learning_rate": 6.245108995048719e-05,
+      "loss": 2.7457,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5670317733321264,
+      "grad_norm": 0.8434032201766968,
+      "learning_rate": 6.242168060624572e-05,
+      "loss": 2.7108,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5671222956458767,
+      "grad_norm": 0.8461436629295349,
+      "learning_rate": 6.239227504614003e-05,
+      "loss": 2.6978,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5672128179596271,
+      "grad_norm": 0.8737865686416626,
+      "learning_rate": 6.23628732731312e-05,
+      "loss": 2.6715,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5673033402733774,
+      "grad_norm": 0.9114745855331421,
+      "learning_rate": 6.233347529018e-05,
+      "loss": 2.6919,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5673938625871278,
+      "grad_norm": 0.8154740929603577,
+      "learning_rate": 6.230408110024679e-05,
+      "loss": 2.6827,
+      "step": 6268
+    },
+    {
+      "epoch": 0.567484384900878,
+      "grad_norm": 0.8688656091690063,
+      "learning_rate": 6.227469070629162e-05,
+      "loss": 2.6951,
+      "step": 6269
+    },
+    {
+      "epoch": 0.5675749072146284,
+      "grad_norm": 0.8917716145515442,
+      "learning_rate": 6.224530411127403e-05,
+      "loss": 2.749,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5676654295283787,
+      "grad_norm": 0.8169562816619873,
+      "learning_rate": 6.22159213181533e-05,
+      "loss": 2.7254,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5677559518421291,
+      "grad_norm": 0.8753547072410583,
+      "learning_rate": 6.218654232988824e-05,
+      "loss": 2.7687,
+      "step": 6272
+    },
+    {
+      "epoch": 0.5678464741558794,
+      "grad_norm": 0.8617856502532959,
+      "learning_rate": 6.215716714943738e-05,
+      "loss": 2.7017,
+      "step": 6273
+    },
+    {
+      "epoch": 0.5679369964696298,
+      "grad_norm": 0.8434569239616394,
+      "learning_rate": 6.21277957797587e-05,
+      "loss": 2.674,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5680275187833801,
+      "grad_norm": 0.8649165034294128,
+      "learning_rate": 6.209842822380998e-05,
+      "loss": 2.7317,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5681180410971305,
+      "grad_norm": 0.8554753661155701,
+      "learning_rate": 6.206906448454848e-05,
+      "loss": 2.7445,
+      "step": 6276
+    },
+    {
+      "epoch": 0.5682085634108808,
+      "grad_norm": 0.9278080463409424,
+      "learning_rate": 6.203970456493118e-05,
+      "loss": 2.7859,
+      "step": 6277
+    },
+    {
+      "epoch": 0.5682990857246312,
+      "grad_norm": 0.8859151005744934,
+      "learning_rate": 6.201034846791454e-05,
+      "loss": 2.6952,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5683896080383815,
+      "grad_norm": 0.9116597771644592,
+      "learning_rate": 6.198099619645481e-05,
+      "loss": 2.7023,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5684801303521319,
+      "grad_norm": 0.850883424282074,
+      "learning_rate": 6.19516477535077e-05,
+      "loss": 2.6934,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5685706526658821,
+      "grad_norm": 0.9140669703483582,
+      "learning_rate": 6.192230314202864e-05,
+      "loss": 2.6676,
+      "step": 6281
+    },
+    {
+      "epoch": 0.5686611749796325,
+      "grad_norm": 0.8315000534057617,
+      "learning_rate": 6.18929623649726e-05,
+      "loss": 2.6754,
+      "step": 6282
+    },
+    {
+      "epoch": 0.5687516972933828,
+      "grad_norm": 0.8378522396087646,
+      "learning_rate": 6.186362542529421e-05,
+      "loss": 2.7073,
+      "step": 6283
+    },
+    {
+      "epoch": 0.5688422196071331,
+      "grad_norm": 0.8570544123649597,
+      "learning_rate": 6.18342923259477e-05,
+      "loss": 2.6984,
+      "step": 6284
+    },
+    {
+      "epoch": 0.5689327419208835,
+      "grad_norm": 0.9210760593414307,
+      "learning_rate": 6.180496306988692e-05,
+      "loss": 2.7113,
+      "step": 6285
+    },
+    {
+      "epoch": 0.5690232642346338,
+      "grad_norm": 0.8755303025245667,
+      "learning_rate": 6.177563766006526e-05,
+      "loss": 2.6953,
+      "step": 6286
+    },
+    {
+      "epoch": 0.5691137865483842,
+      "grad_norm": 0.9134300351142883,
+      "learning_rate": 6.174631609943591e-05,
+      "loss": 2.7662,
+      "step": 6287
+    },
+    {
+      "epoch": 0.5692043088621345,
+      "grad_norm": 0.8612111210823059,
+      "learning_rate": 6.171699839095144e-05,
+      "loss": 2.7175,
+      "step": 6288
+    },
+    {
+      "epoch": 0.5692948311758849,
+      "grad_norm": 0.8268765211105347,
+      "learning_rate": 6.168768453756421e-05,
+      "loss": 2.685,
+      "step": 6289
+    },
+    {
+      "epoch": 0.5693853534896351,
+      "grad_norm": 0.9528706669807434,
+      "learning_rate": 6.165837454222608e-05,
+      "loss": 2.7083,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5694758758033855,
+      "grad_norm": 0.9075449705123901,
+      "learning_rate": 6.16290684078886e-05,
+      "loss": 2.6614,
+      "step": 6291
+    },
+    {
+      "epoch": 0.5695663981171358,
+      "grad_norm": 0.8905490636825562,
+      "learning_rate": 6.159976613750286e-05,
+      "loss": 2.7241,
+      "step": 6292
+    },
+    {
+      "epoch": 0.5696569204308862,
+      "grad_norm": 0.8218395113945007,
+      "learning_rate": 6.157046773401964e-05,
+      "loss": 2.6705,
+      "step": 6293
+    },
+    {
+      "epoch": 0.5697474427446365,
+      "grad_norm": 0.9417014122009277,
+      "learning_rate": 6.154117320038924e-05,
+      "loss": 2.7265,
+      "step": 6294
+    },
+    {
+      "epoch": 0.5698379650583869,
+      "grad_norm": 0.8464024066925049,
+      "learning_rate": 6.151188253956168e-05,
+      "loss": 2.7003,
+      "step": 6295
+    },
+    {
+      "epoch": 0.5699284873721372,
+      "grad_norm": 0.9145004153251648,
+      "learning_rate": 6.148259575448647e-05,
+      "loss": 2.7112,
+      "step": 6296
+    },
+    {
+      "epoch": 0.5700190096858876,
+      "grad_norm": 0.8569707870483398,
+      "learning_rate": 6.145331284811285e-05,
+      "loss": 2.7289,
+      "step": 6297
+    },
+    {
+      "epoch": 0.5701095319996379,
+      "grad_norm": 0.9021281003952026,
+      "learning_rate": 6.142403382338951e-05,
+      "loss": 2.6766,
+      "step": 6298
+    },
+    {
+      "epoch": 0.5702000543133883,
+      "grad_norm": 0.8854073286056519,
+      "learning_rate": 6.139475868326496e-05,
+      "loss": 2.7415,
+      "step": 6299
+    },
+    {
+      "epoch": 0.5702905766271386,
+      "grad_norm": 0.816065788269043,
+      "learning_rate": 6.136548743068713e-05,
+      "loss": 2.6619,
+      "step": 6300
+    },
+    {
+      "epoch": 0.570381098940889,
+      "grad_norm": 0.8643266558647156,
+      "learning_rate": 6.133622006860369e-05,
+      "loss": 2.7427,
+      "step": 6301
+    },
+    {
+      "epoch": 0.5704716212546392,
+      "grad_norm": 0.8317883014678955,
+      "learning_rate": 6.130695659996179e-05,
+      "loss": 2.651,
+      "step": 6302
+    },
+    {
+      "epoch": 0.5705621435683896,
+      "grad_norm": 0.8347657322883606,
+      "learning_rate": 6.127769702770834e-05,
+      "loss": 2.6842,
+      "step": 6303
+    },
+    {
+      "epoch": 0.5706526658821399,
+      "grad_norm": 0.8077671527862549,
+      "learning_rate": 6.12484413547897e-05,
+      "loss": 2.6599,
+      "step": 6304
+    },
+    {
+      "epoch": 0.5707431881958903,
+      "grad_norm": 0.8410141468048096,
+      "learning_rate": 6.121918958415199e-05,
+      "loss": 2.7244,
+      "step": 6305
+    },
+    {
+      "epoch": 0.5708337105096406,
+      "grad_norm": 0.8797140121459961,
+      "learning_rate": 6.118994171874082e-05,
+      "loss": 2.7024,
+      "step": 6306
+    },
+    {
+      "epoch": 0.570924232823391,
+      "grad_norm": 0.8294859528541565,
+      "learning_rate": 6.116069776150147e-05,
+      "loss": 2.7675,
+      "step": 6307
+    },
+    {
+      "epoch": 0.5710147551371413,
+      "grad_norm": 0.8064338564872742,
+      "learning_rate": 6.113145771537874e-05,
+      "loss": 2.7055,
+      "step": 6308
+    },
+    {
+      "epoch": 0.5711052774508917,
+      "grad_norm": 0.7874312996864319,
+      "learning_rate": 6.11022215833172e-05,
+      "loss": 2.6885,
+      "step": 6309
+    },
+    {
+      "epoch": 0.571195799764642,
+      "grad_norm": 0.8688129186630249,
+      "learning_rate": 6.107298936826086e-05,
+      "loss": 2.6763,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5712863220783924,
+      "grad_norm": 0.9120535850524902,
+      "learning_rate": 6.104376107315345e-05,
+      "loss": 2.6554,
+      "step": 6311
+    },
+    {
+      "epoch": 0.5713768443921426,
+      "grad_norm": 0.869500994682312,
+      "learning_rate": 6.10145367009382e-05,
+      "loss": 2.7581,
+      "step": 6312
+    },
+    {
+      "epoch": 0.571467366705893,
+      "grad_norm": 0.8980798721313477,
+      "learning_rate": 6.098531625455806e-05,
+      "loss": 2.663,
+      "step": 6313
+    },
+    {
+      "epoch": 0.5715578890196433,
+      "grad_norm": 0.8309452533721924,
+      "learning_rate": 6.0956099736955494e-05,
+      "loss": 2.7424,
+      "step": 6314
+    },
+    {
+      "epoch": 0.5716484113333937,
+      "grad_norm": 0.8424989581108093,
+      "learning_rate": 6.092688715107264e-05,
+      "loss": 2.7321,
+      "step": 6315
+    },
+    {
+      "epoch": 0.571738933647144,
+      "grad_norm": 0.8888978362083435,
+      "learning_rate": 6.089767849985114e-05,
+      "loss": 2.7492,
+      "step": 6316
+    },
+    {
+      "epoch": 0.5718294559608944,
+      "grad_norm": 0.8150400519371033,
+      "learning_rate": 6.0868473786232395e-05,
+      "loss": 2.7341,
+      "step": 6317
+    },
+    {
+      "epoch": 0.5719199782746447,
+      "grad_norm": 0.8388065695762634,
+      "learning_rate": 6.083927301315724e-05,
+      "loss": 2.7131,
+      "step": 6318
+    },
+    {
+      "epoch": 0.5720105005883951,
+      "grad_norm": 0.9214149117469788,
+      "learning_rate": 6.081007618356623e-05,
+      "loss": 2.7402,
+      "step": 6319
+    },
+    {
+      "epoch": 0.5721010229021454,
+      "grad_norm": 0.8912999629974365,
+      "learning_rate": 6.078088330039945e-05,
+      "loss": 2.7239,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5721915452158958,
+      "grad_norm": 0.7984560132026672,
+      "learning_rate": 6.075169436659669e-05,
+      "loss": 2.6878,
+      "step": 6321
+    },
+    {
+      "epoch": 0.572282067529646,
+      "grad_norm": 0.8435370326042175,
+      "learning_rate": 6.0722509385097205e-05,
+      "loss": 2.7153,
+      "step": 6322
+    },
+    {
+      "epoch": 0.5723725898433965,
+      "grad_norm": 0.8220704793930054,
+      "learning_rate": 6.0693328358839966e-05,
+      "loss": 2.658,
+      "step": 6323
+    },
+    {
+      "epoch": 0.5724631121571467,
+      "grad_norm": 0.8204720616340637,
+      "learning_rate": 6.066415129076346e-05,
+      "loss": 2.6718,
+      "step": 6324
+    },
+    {
+      "epoch": 0.572553634470897,
+      "grad_norm": 0.8146944046020508,
+      "learning_rate": 6.063497818380587e-05,
+      "loss": 2.6801,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5726441567846474,
+      "grad_norm": 0.7801823019981384,
+      "learning_rate": 6.0605809040904894e-05,
+      "loss": 2.6909,
+      "step": 6326
+    },
+    {
+      "epoch": 0.5727346790983977,
+      "grad_norm": 0.8066073656082153,
+      "learning_rate": 6.0576643864997875e-05,
+      "loss": 2.7067,
+      "step": 6327
+    },
+    {
+      "epoch": 0.5728252014121481,
+      "grad_norm": 0.8099483847618103,
+      "learning_rate": 6.0547482659021706e-05,
+      "loss": 2.6446,
+      "step": 6328
+    },
+    {
+      "epoch": 0.5729157237258984,
+      "grad_norm": 0.841600775718689,
+      "learning_rate": 6.051832542591299e-05,
+      "loss": 2.7368,
+      "step": 6329
+    },
+    {
+      "epoch": 0.5730062460396488,
+      "grad_norm": 0.8247181177139282,
+      "learning_rate": 6.048917216860781e-05,
+      "loss": 2.6526,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5730967683533991,
+      "grad_norm": 0.8284602165222168,
+      "learning_rate": 6.0460022890041934e-05,
+      "loss": 2.6957,
+      "step": 6331
+    },
+    {
+      "epoch": 0.5731872906671495,
+      "grad_norm": 0.8628219366073608,
+      "learning_rate": 6.043087759315066e-05,
+      "loss": 2.7116,
+      "step": 6332
+    },
+    {
+      "epoch": 0.5732778129808997,
+      "grad_norm": 0.754366934299469,
+      "learning_rate": 6.0401736280868895e-05,
+      "loss": 2.6362,
+      "step": 6333
+    },
+    {
+      "epoch": 0.5733683352946501,
+      "grad_norm": 0.8926199674606323,
+      "learning_rate": 6.0372598956131265e-05,
+      "loss": 2.7344,
+      "step": 6334
+    },
+    {
+      "epoch": 0.5734588576084004,
+      "grad_norm": 0.8918818235397339,
+      "learning_rate": 6.0343465621871776e-05,
+      "loss": 2.7467,
+      "step": 6335
+    },
+    {
+      "epoch": 0.5735493799221508,
+      "grad_norm": 0.8268258571624756,
+      "learning_rate": 6.031433628102425e-05,
+      "loss": 2.6649,
+      "step": 6336
+    },
+    {
+      "epoch": 0.5736399022359011,
+      "grad_norm": 0.8441937565803528,
+      "learning_rate": 6.0285210936521955e-05,
+      "loss": 2.7086,
+      "step": 6337
+    },
+    {
+      "epoch": 0.5737304245496515,
+      "grad_norm": 0.873117208480835,
+      "learning_rate": 6.025608959129785e-05,
+      "loss": 2.6915,
+      "step": 6338
+    },
+    {
+      "epoch": 0.5738209468634018,
+      "grad_norm": 0.8519162535667419,
+      "learning_rate": 6.02269722482844e-05,
+      "loss": 2.7296,
+      "step": 6339
+    },
+    {
+      "epoch": 0.5739114691771522,
+      "grad_norm": 0.9300065040588379,
+      "learning_rate": 6.019785891041381e-05,
+      "loss": 2.7292,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5740019914909025,
+      "grad_norm": 0.8156430125236511,
+      "learning_rate": 6.0168749580617686e-05,
+      "loss": 2.7259,
+      "step": 6341
+    },
+    {
+      "epoch": 0.5740925138046529,
+      "grad_norm": 0.9231059551239014,
+      "learning_rate": 6.0139644261827435e-05,
+      "loss": 2.7137,
+      "step": 6342
+    },
+    {
+      "epoch": 0.5741830361184032,
+      "grad_norm": 0.8606845140457153,
+      "learning_rate": 6.011054295697387e-05,
+      "loss": 2.6815,
+      "step": 6343
+    },
+    {
+      "epoch": 0.5742735584321536,
+      "grad_norm": 0.8757749795913696,
+      "learning_rate": 6.008144566898758e-05,
+      "loss": 2.6852,
+      "step": 6344
+    },
+    {
+      "epoch": 0.5743640807459038,
+      "grad_norm": 0.8971512317657471,
+      "learning_rate": 6.005235240079859e-05,
+      "loss": 2.6847,
+      "step": 6345
+    },
+    {
+      "epoch": 0.5744546030596542,
+      "grad_norm": 0.8060917854309082,
+      "learning_rate": 6.002326315533665e-05,
+      "loss": 2.6772,
+      "step": 6346
+    },
+    {
+      "epoch": 0.5745451253734045,
+      "grad_norm": 0.9402631521224976,
+      "learning_rate": 5.999417793553097e-05,
+      "loss": 2.745,
+      "step": 6347
+    },
+    {
+      "epoch": 0.5746356476871549,
+      "grad_norm": 0.8727978467941284,
+      "learning_rate": 5.9965096744310526e-05,
+      "loss": 2.6546,
+      "step": 6348
+    },
+    {
+      "epoch": 0.5747261700009052,
+      "grad_norm": 0.9258029460906982,
+      "learning_rate": 5.993601958460371e-05,
+      "loss": 2.7069,
+      "step": 6349
+    },
+    {
+      "epoch": 0.5748166923146556,
+      "grad_norm": 0.8712645769119263,
+      "learning_rate": 5.9906946459338656e-05,
+      "loss": 2.7036,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5749072146284059,
+      "grad_norm": 1.0271838903427124,
+      "learning_rate": 5.987787737144296e-05,
+      "loss": 2.717,
+      "step": 6351
+    },
+    {
+      "epoch": 0.5749977369421563,
+      "grad_norm": 0.8358034491539001,
+      "learning_rate": 5.984881232384394e-05,
+      "loss": 2.665,
+      "step": 6352
+    },
+    {
+      "epoch": 0.5750882592559066,
+      "grad_norm": 0.9175279140472412,
+      "learning_rate": 5.981975131946842e-05,
+      "loss": 2.6991,
+      "step": 6353
+    },
+    {
+      "epoch": 0.575178781569657,
+      "grad_norm": 0.9423436522483826,
+      "learning_rate": 5.979069436124285e-05,
+      "loss": 2.698,
+      "step": 6354
+    },
+    {
+      "epoch": 0.5752693038834072,
+      "grad_norm": 0.9727945923805237,
+      "learning_rate": 5.976164145209322e-05,
+      "loss": 2.7172,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5753598261971576,
+      "grad_norm": 0.8513864874839783,
+      "learning_rate": 5.973259259494524e-05,
+      "loss": 2.6729,
+      "step": 6356
+    },
+    {
+      "epoch": 0.5754503485109079,
+      "grad_norm": 1.0204179286956787,
+      "learning_rate": 5.9703547792724045e-05,
+      "loss": 2.7242,
+      "step": 6357
+    },
+    {
+      "epoch": 0.5755408708246583,
+      "grad_norm": 0.9762052893638611,
+      "learning_rate": 5.967450704835452e-05,
+      "loss": 2.7628,
+      "step": 6358
+    },
+    {
+      "epoch": 0.5756313931384086,
+      "grad_norm": 0.8357113003730774,
+      "learning_rate": 5.964547036476099e-05,
+      "loss": 2.7648,
+      "step": 6359
+    },
+    {
+      "epoch": 0.575721915452159,
+      "grad_norm": 0.8610087633132935,
+      "learning_rate": 5.9616437744867535e-05,
+      "loss": 2.6764,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5758124377659093,
+      "grad_norm": 0.916714608669281,
+      "learning_rate": 5.958740919159766e-05,
+      "loss": 2.6536,
+      "step": 6361
+    },
+    {
+      "epoch": 0.5759029600796597,
+      "grad_norm": 0.7818869352340698,
+      "learning_rate": 5.95583847078746e-05,
+      "loss": 2.7121,
+      "step": 6362
+    },
+    {
+      "epoch": 0.57599348239341,
+      "grad_norm": 0.9010102152824402,
+      "learning_rate": 5.952936429662106e-05,
+      "loss": 2.7471,
+      "step": 6363
+    },
+    {
+      "epoch": 0.5760840047071604,
+      "grad_norm": 0.8215236663818359,
+      "learning_rate": 5.950034796075947e-05,
+      "loss": 2.6007,
+      "step": 6364
+    },
+    {
+      "epoch": 0.5761745270209107,
+      "grad_norm": 0.8552883267402649,
+      "learning_rate": 5.94713357032117e-05,
+      "loss": 2.7062,
+      "step": 6365
+    },
+    {
+      "epoch": 0.5762650493346609,
+      "grad_norm": 0.7977161407470703,
+      "learning_rate": 5.944232752689936e-05,
+      "loss": 2.7083,
+      "step": 6366
+    },
+    {
+      "epoch": 0.5763555716484113,
+      "grad_norm": 0.8528048396110535,
+      "learning_rate": 5.941332343474349e-05,
+      "loss": 2.715,
+      "step": 6367
+    },
+    {
+      "epoch": 0.5764460939621616,
+      "grad_norm": 0.9308705925941467,
+      "learning_rate": 5.9384323429664845e-05,
+      "loss": 2.6994,
+      "step": 6368
+    },
+    {
+      "epoch": 0.576536616275912,
+      "grad_norm": 0.9029669761657715,
+      "learning_rate": 5.93553275145837e-05,
+      "loss": 2.7165,
+      "step": 6369
+    },
+    {
+      "epoch": 0.5766271385896623,
+      "grad_norm": 0.791476309299469,
+      "learning_rate": 5.9326335692419995e-05,
+      "loss": 2.7014,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5767176609034127,
+      "grad_norm": 0.8183833956718445,
+      "learning_rate": 5.929734796609315e-05,
+      "loss": 2.742,
+      "step": 6371
+    },
+    {
+      "epoch": 0.576808183217163,
+      "grad_norm": 0.9724903702735901,
+      "learning_rate": 5.926836433852227e-05,
+      "loss": 2.6981,
+      "step": 6372
+    },
+    {
+      "epoch": 0.5768987055309134,
+      "grad_norm": 0.8855651021003723,
+      "learning_rate": 5.9239384812625964e-05,
+      "loss": 2.6728,
+      "step": 6373
+    },
+    {
+      "epoch": 0.5769892278446637,
+      "grad_norm": 0.7812449336051941,
+      "learning_rate": 5.9210409391322516e-05,
+      "loss": 2.6848,
+      "step": 6374
+    },
+    {
+      "epoch": 0.5770797501584141,
+      "grad_norm": 0.9259555339813232,
+      "learning_rate": 5.918143807752972e-05,
+      "loss": 2.7506,
+      "step": 6375
+    },
+    {
+      "epoch": 0.5771702724721643,
+      "grad_norm": 0.9594706296920776,
+      "learning_rate": 5.9152470874165e-05,
+      "loss": 2.6974,
+      "step": 6376
+    },
+    {
+      "epoch": 0.5772607947859147,
+      "grad_norm": 0.9384106397628784,
+      "learning_rate": 5.912350778414531e-05,
+      "loss": 2.6748,
+      "step": 6377
+    },
+    {
+      "epoch": 0.577351317099665,
+      "grad_norm": 0.9170209765434265,
+      "learning_rate": 5.909454881038729e-05,
+      "loss": 2.664,
+      "step": 6378
+    },
+    {
+      "epoch": 0.5774418394134154,
+      "grad_norm": 0.915307343006134,
+      "learning_rate": 5.906559395580709e-05,
+      "loss": 2.7388,
+      "step": 6379
+    },
+    {
+      "epoch": 0.5775323617271657,
+      "grad_norm": 0.8271397352218628,
+      "learning_rate": 5.9036643223320475e-05,
+      "loss": 2.6895,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5776228840409161,
+      "grad_norm": 0.7941808700561523,
+      "learning_rate": 5.900769661584272e-05,
+      "loss": 2.6932,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5777134063546664,
+      "grad_norm": 0.9167299270629883,
+      "learning_rate": 5.897875413628884e-05,
+      "loss": 2.7084,
+      "step": 6382
+    },
+    {
+      "epoch": 0.5778039286684168,
+      "grad_norm": 0.8044254183769226,
+      "learning_rate": 5.894981578757327e-05,
+      "loss": 2.7183,
+      "step": 6383
+    },
+    {
+      "epoch": 0.5778944509821671,
+      "grad_norm": 0.9017865061759949,
+      "learning_rate": 5.892088157261015e-05,
+      "loss": 2.7275,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5779849732959175,
+      "grad_norm": 0.8614157438278198,
+      "learning_rate": 5.889195149431309e-05,
+      "loss": 2.7007,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5780754956096678,
+      "grad_norm": 0.9102115631103516,
+      "learning_rate": 5.886302555559543e-05,
+      "loss": 2.7378,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5781660179234182,
+      "grad_norm": 0.8719278573989868,
+      "learning_rate": 5.883410375936995e-05,
+      "loss": 2.7231,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5782565402371684,
+      "grad_norm": 0.9065318703651428,
+      "learning_rate": 5.8805186108549114e-05,
+      "loss": 2.707,
+      "step": 6388
+    },
+    {
+      "epoch": 0.5783470625509188,
+      "grad_norm": 0.8618296980857849,
+      "learning_rate": 5.877627260604488e-05,
+      "loss": 2.7146,
+      "step": 6389
+    },
+    {
+      "epoch": 0.5784375848646691,
+      "grad_norm": 0.9378622770309448,
+      "learning_rate": 5.8747363254768894e-05,
+      "loss": 2.7278,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5785281071784195,
+      "grad_norm": 0.9272425770759583,
+      "learning_rate": 5.8718458057632286e-05,
+      "loss": 2.6517,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5786186294921698,
+      "grad_norm": 0.8620819449424744,
+      "learning_rate": 5.868955701754584e-05,
+      "loss": 2.7031,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5787091518059202,
+      "grad_norm": 0.8210526704788208,
+      "learning_rate": 5.866066013741983e-05,
+      "loss": 2.6852,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5787996741196705,
+      "grad_norm": 0.8317428827285767,
+      "learning_rate": 5.863176742016425e-05,
+      "loss": 2.7107,
+      "step": 6394
+    },
+    {
+      "epoch": 0.5788901964334209,
+      "grad_norm": 0.854906439781189,
+      "learning_rate": 5.860287886868855e-05,
+      "loss": 2.7202,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5789807187471712,
+      "grad_norm": 0.8338930606842041,
+      "learning_rate": 5.857399448590183e-05,
+      "loss": 2.6836,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5790712410609216,
+      "grad_norm": 0.8774193525314331,
+      "learning_rate": 5.8545114274712695e-05,
+      "loss": 2.7342,
+      "step": 6397
+    },
+    {
+      "epoch": 0.5791617633746718,
+      "grad_norm": 0.912401020526886,
+      "learning_rate": 5.8516238238029476e-05,
+      "loss": 2.683,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5792522856884222,
+      "grad_norm": 0.835411548614502,
+      "learning_rate": 5.848736637875987e-05,
+      "loss": 2.7199,
+      "step": 6399
+    },
+    {
+      "epoch": 0.5793428080021725,
+      "grad_norm": 0.9073437452316284,
+      "learning_rate": 5.845849869981137e-05,
+      "loss": 2.734,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5794333303159229,
+      "grad_norm": 0.8086467385292053,
+      "learning_rate": 5.8429635204090904e-05,
+      "loss": 2.7081,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5795238526296732,
+      "grad_norm": 0.8387646675109863,
+      "learning_rate": 5.840077589450505e-05,
+      "loss": 2.7258,
+      "step": 6402
+    },
+    {
+      "epoch": 0.5796143749434236,
+      "grad_norm": 0.870087742805481,
+      "learning_rate": 5.83719207739599e-05,
+      "loss": 2.7826,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5797048972571739,
+      "grad_norm": 0.8552343249320984,
+      "learning_rate": 5.83430698453612e-05,
+      "loss": 2.7115,
+      "step": 6404
+    },
+    {
+      "epoch": 0.5797954195709243,
+      "grad_norm": 0.8185401558876038,
+      "learning_rate": 5.831422311161421e-05,
+      "loss": 2.6749,
+      "step": 6405
+    },
+    {
+      "epoch": 0.5798859418846746,
+      "grad_norm": 0.8774237036705017,
+      "learning_rate": 5.8285380575623826e-05,
+      "loss": 2.6766,
+      "step": 6406
+    },
+    {
+      "epoch": 0.5799764641984249,
+      "grad_norm": 0.9418449401855469,
+      "learning_rate": 5.825654224029447e-05,
+      "loss": 2.6728,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5800669865121753,
+      "grad_norm": 0.8073645830154419,
+      "learning_rate": 5.822770810853022e-05,
+      "loss": 2.6885,
+      "step": 6408
+    },
+    {
+      "epoch": 0.5801575088259255,
+      "grad_norm": 0.8460543751716614,
+      "learning_rate": 5.819887818323453e-05,
+      "loss": 2.7333,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5802480311396759,
+      "grad_norm": 0.9544316530227661,
+      "learning_rate": 5.817005246731073e-05,
+      "loss": 2.689,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5803385534534262,
+      "grad_norm": 0.8486777544021606,
+      "learning_rate": 5.814123096366148e-05,
+      "loss": 2.7374,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5804290757671766,
+      "grad_norm": 0.8409209251403809,
+      "learning_rate": 5.811241367518914e-05,
+      "loss": 2.7153,
+      "step": 6412
+    },
+    {
+      "epoch": 0.5805195980809269,
+      "grad_norm": 0.8512220978736877,
+      "learning_rate": 5.808360060479557e-05,
+      "loss": 2.708,
+      "step": 6413
+    },
+    {
+      "epoch": 0.5806101203946773,
+      "grad_norm": 0.9078315496444702,
+      "learning_rate": 5.805479175538229e-05,
+      "loss": 2.7111,
+      "step": 6414
+    },
+    {
+      "epoch": 0.5807006427084276,
+      "grad_norm": 0.8165913224220276,
+      "learning_rate": 5.802598712985032e-05,
+      "loss": 2.6777,
+      "step": 6415
+    },
+    {
+      "epoch": 0.580791165022178,
+      "grad_norm": 0.890654444694519,
+      "learning_rate": 5.799718673110035e-05,
+      "loss": 2.7412,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5808816873359283,
+      "grad_norm": 0.8421646356582642,
+      "learning_rate": 5.796839056203247e-05,
+      "loss": 2.7217,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5809722096496787,
+      "grad_norm": 0.8950380086898804,
+      "learning_rate": 5.793959862554652e-05,
+      "loss": 2.6949,
+      "step": 6418
+    },
+    {
+      "epoch": 0.581062731963429,
+      "grad_norm": 0.8083726167678833,
+      "learning_rate": 5.7910810924541844e-05,
+      "loss": 2.6173,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5811532542771793,
+      "grad_norm": 0.9236308336257935,
+      "learning_rate": 5.788202746191734e-05,
+      "loss": 2.7524,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5812437765909296,
+      "grad_norm": 0.8910950422286987,
+      "learning_rate": 5.785324824057157e-05,
+      "loss": 2.6792,
+      "step": 6421
+    },
+    {
+      "epoch": 0.58133429890468,
+      "grad_norm": 0.798999547958374,
+      "learning_rate": 5.782447326340247e-05,
+      "loss": 2.6614,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5814248212184303,
+      "grad_norm": 0.8792107701301575,
+      "learning_rate": 5.779570253330784e-05,
+      "loss": 2.6563,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5815153435321807,
+      "grad_norm": 0.9009785652160645,
+      "learning_rate": 5.776693605318476e-05,
+      "loss": 2.7166,
+      "step": 6424
+    },
+    {
+      "epoch": 0.581605865845931,
+      "grad_norm": 0.8150326013565063,
+      "learning_rate": 5.773817382593008e-05,
+      "loss": 2.738,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5816963881596814,
+      "grad_norm": 0.8171740770339966,
+      "learning_rate": 5.7709415854440116e-05,
+      "loss": 2.6949,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5817869104734317,
+      "grad_norm": 0.9186320900917053,
+      "learning_rate": 5.768066214161087e-05,
+      "loss": 2.686,
+      "step": 6427
+    },
+    {
+      "epoch": 0.5818774327871821,
+      "grad_norm": 0.8507105708122253,
+      "learning_rate": 5.7651912690337716e-05,
+      "loss": 2.7283,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5819679551009324,
+      "grad_norm": 0.8721489906311035,
+      "learning_rate": 5.762316750351586e-05,
+      "loss": 2.6998,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5820584774146828,
+      "grad_norm": 0.9127447009086609,
+      "learning_rate": 5.759442658403985e-05,
+      "loss": 2.696,
+      "step": 6430
+    },
+    {
+      "epoch": 0.582148999728433,
+      "grad_norm": 0.8901234865188599,
+      "learning_rate": 5.756568993480391e-05,
+      "loss": 2.6944,
+      "step": 6431
+    },
+    {
+      "epoch": 0.5822395220421834,
+      "grad_norm": 0.8674662113189697,
+      "learning_rate": 5.753695755870185e-05,
+      "loss": 2.6569,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5823300443559337,
+      "grad_norm": 0.8261517882347107,
+      "learning_rate": 5.7508229458627e-05,
+      "loss": 2.6845,
+      "step": 6433
+    },
+    {
+      "epoch": 0.5824205666696841,
+      "grad_norm": 0.8022430539131165,
+      "learning_rate": 5.7479505637472264e-05,
+      "loss": 2.6961,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5825110889834344,
+      "grad_norm": 0.9176723957061768,
+      "learning_rate": 5.7450786098130194e-05,
+      "loss": 2.6622,
+      "step": 6435
+    },
+    {
+      "epoch": 0.5826016112971848,
+      "grad_norm": 0.9195401072502136,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 2.7736,
+      "step": 6436
+    },
+    {
+      "epoch": 0.5826921336109351,
+      "grad_norm": 0.9636011719703674,
+      "learning_rate": 5.7393359876451634e-05,
+      "loss": 2.7399,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5827826559246855,
+      "grad_norm": 0.870815098285675,
+      "learning_rate": 5.7364653199898e-05,
+      "loss": 2.7254,
+      "step": 6438
+    },
+    {
+      "epoch": 0.5828731782384358,
+      "grad_norm": 0.881637692451477,
+      "learning_rate": 5.733595081672263e-05,
+      "loss": 2.6513,
+      "step": 6439
+    },
+    {
+      "epoch": 0.5829637005521862,
+      "grad_norm": 0.9924260973930359,
+      "learning_rate": 5.7307252729815833e-05,
+      "loss": 2.7274,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5830542228659364,
+      "grad_norm": 0.853918731212616,
+      "learning_rate": 5.727855894206752e-05,
+      "loss": 2.7194,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5831447451796868,
+      "grad_norm": 0.907598078250885,
+      "learning_rate": 5.7249869456367146e-05,
+      "loss": 2.7265,
+      "step": 6442
+    },
+    {
+      "epoch": 0.5832352674934371,
+      "grad_norm": 0.8251694440841675,
+      "learning_rate": 5.722118427560379e-05,
+      "loss": 2.7336,
+      "step": 6443
+    },
+    {
+      "epoch": 0.5833257898071875,
+      "grad_norm": 0.8961870670318604,
+      "learning_rate": 5.7192503402665944e-05,
+      "loss": 2.7209,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5834163121209378,
+      "grad_norm": 0.8750089406967163,
+      "learning_rate": 5.71638268404419e-05,
+      "loss": 2.7938,
+      "step": 6445
+    },
+    {
+      "epoch": 0.5835068344346882,
+      "grad_norm": 0.8658246397972107,
+      "learning_rate": 5.7135154591819287e-05,
+      "loss": 2.6955,
+      "step": 6446
+    },
+    {
+      "epoch": 0.5835973567484385,
+      "grad_norm": 0.8144567608833313,
+      "learning_rate": 5.710648665968543e-05,
+      "loss": 2.6923,
+      "step": 6447
+    },
+    {
+      "epoch": 0.5836878790621888,
+      "grad_norm": 0.8991824388504028,
+      "learning_rate": 5.707782304692719e-05,
+      "loss": 2.694,
+      "step": 6448
+    },
+    {
+      "epoch": 0.5837784013759392,
+      "grad_norm": 1.0232534408569336,
+      "learning_rate": 5.704916375643099e-05,
+      "loss": 2.7462,
+      "step": 6449
+    },
+    {
+      "epoch": 0.5838689236896895,
+      "grad_norm": 0.8977211117744446,
+      "learning_rate": 5.702050879108284e-05,
+      "loss": 2.74,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5839594460034399,
+      "grad_norm": 0.8499026894569397,
+      "learning_rate": 5.6991858153768305e-05,
+      "loss": 2.697,
+      "step": 6451
+    },
+    {
+      "epoch": 0.5840499683171901,
+      "grad_norm": 0.8971026539802551,
+      "learning_rate": 5.696321184737241e-05,
+      "loss": 2.6834,
+      "step": 6452
+    },
+    {
+      "epoch": 0.5841404906309405,
+      "grad_norm": 0.9016530513763428,
+      "learning_rate": 5.693456987477996e-05,
+      "loss": 2.7218,
+      "step": 6453
+    },
+    {
+      "epoch": 0.5842310129446908,
+      "grad_norm": 0.8605112433433533,
+      "learning_rate": 5.6905932238875123e-05,
+      "loss": 2.7338,
+      "step": 6454
+    },
+    {
+      "epoch": 0.5843215352584412,
+      "grad_norm": 0.8572389483451843,
+      "learning_rate": 5.6877298942541746e-05,
+      "loss": 2.7093,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5844120575721915,
+      "grad_norm": 0.8466581702232361,
+      "learning_rate": 5.684866998866316e-05,
+      "loss": 2.6993,
+      "step": 6456
+    },
+    {
+      "epoch": 0.5845025798859419,
+      "grad_norm": 0.8294211030006409,
+      "learning_rate": 5.682004538012233e-05,
+      "loss": 2.6486,
+      "step": 6457
+    },
+    {
+      "epoch": 0.5845931021996922,
+      "grad_norm": 0.7687167525291443,
+      "learning_rate": 5.679142511980175e-05,
+      "loss": 2.6596,
+      "step": 6458
+    },
+    {
+      "epoch": 0.5846836245134426,
+      "grad_norm": 0.7967802882194519,
+      "learning_rate": 5.6762809210583534e-05,
+      "loss": 2.7049,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5847741468271929,
+      "grad_norm": 0.905829906463623,
+      "learning_rate": 5.6734197655349156e-05,
+      "loss": 2.6424,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5848646691409433,
+      "grad_norm": 0.7958777546882629,
+      "learning_rate": 5.670559045697996e-05,
+      "loss": 2.664,
+      "step": 6461
+    },
+    {
+      "epoch": 0.5849551914546935,
+      "grad_norm": 0.8610337376594543,
+      "learning_rate": 5.66769876183566e-05,
+      "loss": 2.7594,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5850457137684439,
+      "grad_norm": 0.9278115034103394,
+      "learning_rate": 5.664838914235939e-05,
+      "loss": 2.7304,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5851362360821942,
+      "grad_norm": 0.9422511458396912,
+      "learning_rate": 5.661979503186821e-05,
+      "loss": 2.7364,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5852267583959446,
+      "grad_norm": 0.8577379584312439,
+      "learning_rate": 5.659120528976252e-05,
+      "loss": 2.8038,
+      "step": 6465
+    },
+    {
+      "epoch": 0.5853172807096949,
+      "grad_norm": 0.8961200714111328,
+      "learning_rate": 5.65626199189212e-05,
+      "loss": 2.6653,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5854078030234453,
+      "grad_norm": 0.8905138373374939,
+      "learning_rate": 5.653403892222293e-05,
+      "loss": 2.7148,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5854983253371956,
+      "grad_norm": 0.9227715134620667,
+      "learning_rate": 5.650546230254572e-05,
+      "loss": 2.73,
+      "step": 6468
+    },
+    {
+      "epoch": 0.585588847650946,
+      "grad_norm": 0.8112932443618774,
+      "learning_rate": 5.647689006276726e-05,
+      "loss": 2.6512,
+      "step": 6469
+    },
+    {
+      "epoch": 0.5856793699646963,
+      "grad_norm": 0.8007726073265076,
+      "learning_rate": 5.6448322205764794e-05,
+      "loss": 2.6787,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5857698922784467,
+      "grad_norm": 0.9057304859161377,
+      "learning_rate": 5.6419758734415075e-05,
+      "loss": 2.721,
+      "step": 6471
+    },
+    {
+      "epoch": 0.585860414592197,
+      "grad_norm": 0.8815562725067139,
+      "learning_rate": 5.639119965159446e-05,
+      "loss": 2.6815,
+      "step": 6472
+    },
+    {
+      "epoch": 0.5859509369059474,
+      "grad_norm": 0.8501884341239929,
+      "learning_rate": 5.636264496017889e-05,
+      "loss": 2.6578,
+      "step": 6473
+    },
+    {
+      "epoch": 0.5860414592196976,
+      "grad_norm": 0.9638636708259583,
+      "learning_rate": 5.63340946630437e-05,
+      "loss": 2.7108,
+      "step": 6474
+    },
+    {
+      "epoch": 0.586131981533448,
+      "grad_norm": 0.8438670635223389,
+      "learning_rate": 5.630554876306407e-05,
+      "loss": 2.6919,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5862225038471983,
+      "grad_norm": 0.8943795561790466,
+      "learning_rate": 5.6277007263114437e-05,
+      "loss": 2.6687,
+      "step": 6476
+    },
+    {
+      "epoch": 0.5863130261609487,
+      "grad_norm": 0.8650156259536743,
+      "learning_rate": 5.624847016606898e-05,
+      "loss": 2.7242,
+      "step": 6477
+    },
+    {
+      "epoch": 0.586403548474699,
+      "grad_norm": 0.949901819229126,
+      "learning_rate": 5.6219937474801366e-05,
+      "loss": 2.7432,
+      "step": 6478
+    },
+    {
+      "epoch": 0.5864940707884494,
+      "grad_norm": 0.8957086205482483,
+      "learning_rate": 5.619140919218486e-05,
+      "loss": 2.7215,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5865845931021997,
+      "grad_norm": 0.8338333964347839,
+      "learning_rate": 5.616288532109225e-05,
+      "loss": 2.7067,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5866751154159501,
+      "grad_norm": 0.7819648385047913,
+      "learning_rate": 5.613436586439593e-05,
+      "loss": 2.6851,
+      "step": 6481
+    },
+    {
+      "epoch": 0.5867656377297004,
+      "grad_norm": 0.9758778214454651,
+      "learning_rate": 5.610585082496769e-05,
+      "loss": 2.7379,
+      "step": 6482
+    },
+    {
+      "epoch": 0.5868561600434508,
+      "grad_norm": 0.9074897766113281,
+      "learning_rate": 5.607734020567914e-05,
+      "loss": 2.7016,
+      "step": 6483
+    },
+    {
+      "epoch": 0.586946682357201,
+      "grad_norm": 0.8512108325958252,
+      "learning_rate": 5.6048834009401196e-05,
+      "loss": 2.7295,
+      "step": 6484
+    },
+    {
+      "epoch": 0.5870372046709514,
+      "grad_norm": 0.8494038581848145,
+      "learning_rate": 5.602033223900447e-05,
+      "loss": 2.694,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5871277269847017,
+      "grad_norm": 0.8115109205245972,
+      "learning_rate": 5.599183489735907e-05,
+      "loss": 2.6858,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5872182492984521,
+      "grad_norm": 0.8463485836982727,
+      "learning_rate": 5.5963341987334704e-05,
+      "loss": 2.6405,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5873087716122024,
+      "grad_norm": 0.7917320728302002,
+      "learning_rate": 5.593485351180059e-05,
+      "loss": 2.674,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5873992939259527,
+      "grad_norm": 0.819711446762085,
+      "learning_rate": 5.590636947362557e-05,
+      "loss": 2.811,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5874898162397031,
+      "grad_norm": 0.92168128490448,
+      "learning_rate": 5.5877889875677845e-05,
+      "loss": 2.6875,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5875803385534534,
+      "grad_norm": 0.864594042301178,
+      "learning_rate": 5.584941472082549e-05,
+      "loss": 2.6684,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5876708608672038,
+      "grad_norm": 0.8601884245872498,
+      "learning_rate": 5.5820944011935826e-05,
+      "loss": 2.713,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5877613831809541,
+      "grad_norm": 0.8170925378799438,
+      "learning_rate": 5.5792477751875886e-05,
+      "loss": 2.7317,
+      "step": 6493
+    },
+    {
+      "epoch": 0.5878519054947045,
+      "grad_norm": 0.9108325839042664,
+      "learning_rate": 5.576401594351223e-05,
+      "loss": 2.765,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5879424278084547,
+      "grad_norm": 0.8704808354377747,
+      "learning_rate": 5.5735558589711e-05,
+      "loss": 2.7455,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5880329501222051,
+      "grad_norm": 0.868759036064148,
+      "learning_rate": 5.570710569333772e-05,
+      "loss": 2.7448,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5881234724359554,
+      "grad_norm": 0.884606122970581,
+      "learning_rate": 5.567865725725777e-05,
+      "loss": 2.6955,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5882139947497058,
+      "grad_norm": 0.891055703163147,
+      "learning_rate": 5.565021328433578e-05,
+      "loss": 2.6881,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5883045170634561,
+      "grad_norm": 0.8647257089614868,
+      "learning_rate": 5.56217737774361e-05,
+      "loss": 2.745,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5883950393772065,
+      "grad_norm": 0.8522151112556458,
+      "learning_rate": 5.559333873942259e-05,
+      "loss": 2.758,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5884855616909568,
+      "grad_norm": 0.8541638851165771,
+      "learning_rate": 5.556490817315865e-05,
+      "loss": 2.6753,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5885760840047072,
+      "grad_norm": 0.8651673197746277,
+      "learning_rate": 5.553648208150728e-05,
+      "loss": 2.6963,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5886666063184575,
+      "grad_norm": 0.8872950077056885,
+      "learning_rate": 5.5508060467330915e-05,
+      "loss": 2.7789,
+      "step": 6503
+    },
+    {
+      "epoch": 0.5887571286322079,
+      "grad_norm": 0.8444436192512512,
+      "learning_rate": 5.547964333349164e-05,
+      "loss": 2.7032,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5888476509459581,
+      "grad_norm": 0.835922122001648,
+      "learning_rate": 5.545123068285105e-05,
+      "loss": 2.7042,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5889381732597085,
+      "grad_norm": 0.8999233245849609,
+      "learning_rate": 5.542282251827034e-05,
+      "loss": 2.7133,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5890286955734588,
+      "grad_norm": 0.8679043650627136,
+      "learning_rate": 5.539441884261016e-05,
+      "loss": 2.6582,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5891192178872092,
+      "grad_norm": 0.8555617928504944,
+      "learning_rate": 5.5366019658730825e-05,
+      "loss": 2.6486,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5892097402009595,
+      "grad_norm": 0.8898964524269104,
+      "learning_rate": 5.5337624969492006e-05,
+      "loss": 2.7339,
+      "step": 6509
+    },
+    {
+      "epoch": 0.5893002625147099,
+      "grad_norm": 0.890902042388916,
+      "learning_rate": 5.530923477775323e-05,
+      "loss": 2.6978,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5893907848284602,
+      "grad_norm": 0.8486526012420654,
+      "learning_rate": 5.528084908637323e-05,
+      "loss": 2.6369,
+      "step": 6511
+    },
+    {
+      "epoch": 0.5894813071422106,
+      "grad_norm": 0.8361667990684509,
+      "learning_rate": 5.5252467898210505e-05,
+      "loss": 2.692,
+      "step": 6512
+    },
+    {
+      "epoch": 0.5895718294559609,
+      "grad_norm": 0.8661167025566101,
+      "learning_rate": 5.522409121612304e-05,
+      "loss": 2.7287,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5896623517697113,
+      "grad_norm": 0.9223380088806152,
+      "learning_rate": 5.5195719042968365e-05,
+      "loss": 2.7243,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5897528740834616,
+      "grad_norm": 0.8186379075050354,
+      "learning_rate": 5.516735138160356e-05,
+      "loss": 2.6714,
+      "step": 6515
+    },
+    {
+      "epoch": 0.589843396397212,
+      "grad_norm": 1.0045249462127686,
+      "learning_rate": 5.513898823488528e-05,
+      "loss": 2.7635,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5899339187109622,
+      "grad_norm": 0.8674926161766052,
+      "learning_rate": 5.5110629605669574e-05,
+      "loss": 2.7173,
+      "step": 6517
+    },
+    {
+      "epoch": 0.5900244410247126,
+      "grad_norm": 0.8774318695068359,
+      "learning_rate": 5.508227549681233e-05,
+      "loss": 2.7164,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5901149633384629,
+      "grad_norm": 0.972149133682251,
+      "learning_rate": 5.505392591116867e-05,
+      "loss": 2.7337,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5902054856522133,
+      "grad_norm": 0.847661018371582,
+      "learning_rate": 5.5025580851593436e-05,
+      "loss": 2.7372,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5902960079659636,
+      "grad_norm": 0.9635895490646362,
+      "learning_rate": 5.499724032094098e-05,
+      "loss": 2.7392,
+      "step": 6521
+    },
+    {
+      "epoch": 0.590386530279714,
+      "grad_norm": 0.8678410649299622,
+      "learning_rate": 5.496890432206524e-05,
+      "loss": 2.6731,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5904770525934643,
+      "grad_norm": 0.9082297682762146,
+      "learning_rate": 5.4940572857819516e-05,
+      "loss": 2.7296,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5905675749072147,
+      "grad_norm": 0.9244065880775452,
+      "learning_rate": 5.491224593105695e-05,
+      "loss": 2.7483,
+      "step": 6524
+    },
+    {
+      "epoch": 0.590658097220965,
+      "grad_norm": 0.8604749441146851,
+      "learning_rate": 5.4883923544629955e-05,
+      "loss": 2.7175,
+      "step": 6525
+    },
+    {
+      "epoch": 0.5907486195347154,
+      "grad_norm": 0.9151113033294678,
+      "learning_rate": 5.485560570139061e-05,
+      "loss": 2.6912,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5908391418484656,
+      "grad_norm": 0.8470368981361389,
+      "learning_rate": 5.482729240419056e-05,
+      "loss": 2.7286,
+      "step": 6527
+    },
+    {
+      "epoch": 0.590929664162216,
+      "grad_norm": 0.7878344058990479,
+      "learning_rate": 5.4798983655880896e-05,
+      "loss": 2.6645,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5910201864759663,
+      "grad_norm": 0.8447884917259216,
+      "learning_rate": 5.477067945931237e-05,
+      "loss": 2.6488,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5911107087897166,
+      "grad_norm": 0.8558040857315063,
+      "learning_rate": 5.474237981733521e-05,
+      "loss": 2.6722,
+      "step": 6530
+    },
+    {
+      "epoch": 0.591201231103467,
+      "grad_norm": 0.9111878871917725,
+      "learning_rate": 5.4714084732799084e-05,
+      "loss": 2.7111,
+      "step": 6531
+    },
+    {
+      "epoch": 0.5912917534172173,
+      "grad_norm": 0.8919588923454285,
+      "learning_rate": 5.4685794208553465e-05,
+      "loss": 2.7136,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5913822757309677,
+      "grad_norm": 0.8634358048439026,
+      "learning_rate": 5.465750824744711e-05,
+      "loss": 2.6038,
+      "step": 6533
+    },
+    {
+      "epoch": 0.591472798044718,
+      "grad_norm": 0.8753179311752319,
+      "learning_rate": 5.462922685232841e-05,
+      "loss": 2.6767,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5915633203584684,
+      "grad_norm": 0.8791088461875916,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 2.7181,
+      "step": 6535
+    },
+    {
+      "epoch": 0.5916538426722187,
+      "grad_norm": 0.8607524633407593,
+      "learning_rate": 5.4572677771445344e-05,
+      "loss": 2.6828,
+      "step": 6536
+    },
+    {
+      "epoch": 0.591744364985969,
+      "grad_norm": 0.9010187983512878,
+      "learning_rate": 5.4544410091375454e-05,
+      "loss": 2.6991,
+      "step": 6537
+    },
+    {
+      "epoch": 0.5918348872997193,
+      "grad_norm": 0.8918096423149109,
+      "learning_rate": 5.4516146988682285e-05,
+      "loss": 2.7083,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5919254096134697,
+      "grad_norm": 0.8350089192390442,
+      "learning_rate": 5.448788846621178e-05,
+      "loss": 2.7102,
+      "step": 6539
+    },
+    {
+      "epoch": 0.59201593192722,
+      "grad_norm": 0.8323737382888794,
+      "learning_rate": 5.445963452680973e-05,
+      "loss": 2.7275,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5921064542409704,
+      "grad_norm": 0.8640821576118469,
+      "learning_rate": 5.4431385173321204e-05,
+      "loss": 2.7141,
+      "step": 6541
+    },
+    {
+      "epoch": 0.5921969765547207,
+      "grad_norm": 0.8797996044158936,
+      "learning_rate": 5.440314040859094e-05,
+      "loss": 2.7086,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5922874988684711,
+      "grad_norm": 0.809884250164032,
+      "learning_rate": 5.43749002354632e-05,
+      "loss": 2.5982,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5923780211822214,
+      "grad_norm": 0.9129722714424133,
+      "learning_rate": 5.434666465678175e-05,
+      "loss": 2.8024,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5924685434959718,
+      "grad_norm": 0.9209685921669006,
+      "learning_rate": 5.431843367538992e-05,
+      "loss": 2.7221,
+      "step": 6545
+    },
+    {
+      "epoch": 0.5925590658097221,
+      "grad_norm": 0.8171425461769104,
+      "learning_rate": 5.4290207294130615e-05,
+      "loss": 2.7129,
+      "step": 6546
+    },
+    {
+      "epoch": 0.5926495881234725,
+      "grad_norm": 0.8546719551086426,
+      "learning_rate": 5.426198551584609e-05,
+      "loss": 2.6868,
+      "step": 6547
+    },
+    {
+      "epoch": 0.5927401104372227,
+      "grad_norm": 0.8493045568466187,
+      "learning_rate": 5.423376834337848e-05,
+      "loss": 2.6751,
+      "step": 6548
+    },
+    {
+      "epoch": 0.5928306327509731,
+      "grad_norm": 0.8081749677658081,
+      "learning_rate": 5.4205555779569075e-05,
+      "loss": 2.6307,
+      "step": 6549
+    },
+    {
+      "epoch": 0.5929211550647234,
+      "grad_norm": 0.9289596676826477,
+      "learning_rate": 5.417734782725896e-05,
+      "loss": 2.6709,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5930116773784738,
+      "grad_norm": 0.844437837600708,
+      "learning_rate": 5.414914448928867e-05,
+      "loss": 2.7338,
+      "step": 6551
+    },
+    {
+      "epoch": 0.5931021996922241,
+      "grad_norm": 0.905487060546875,
+      "learning_rate": 5.412094576849829e-05,
+      "loss": 2.6682,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5931927220059745,
+      "grad_norm": 1.018079400062561,
+      "learning_rate": 5.409275166772738e-05,
+      "loss": 2.7137,
+      "step": 6553
+    },
+    {
+      "epoch": 0.5932832443197248,
+      "grad_norm": 0.9324988126754761,
+      "learning_rate": 5.406456218981519e-05,
+      "loss": 2.7833,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5933737666334752,
+      "grad_norm": 0.8467705249786377,
+      "learning_rate": 5.403637733760024e-05,
+      "loss": 2.6998,
+      "step": 6555
+    },
+    {
+      "epoch": 0.5934642889472255,
+      "grad_norm": 0.878297746181488,
+      "learning_rate": 5.400819711392091e-05,
+      "loss": 2.6542,
+      "step": 6556
+    },
+    {
+      "epoch": 0.5935548112609759,
+      "grad_norm": 1.018547773361206,
+      "learning_rate": 5.398002152161484e-05,
+      "loss": 2.753,
+      "step": 6557
+    },
+    {
+      "epoch": 0.5936453335747262,
+      "grad_norm": 0.8300045132637024,
+      "learning_rate": 5.395185056351935e-05,
+      "loss": 2.6834,
+      "step": 6558
+    },
+    {
+      "epoch": 0.5937358558884765,
+      "grad_norm": 0.8211262822151184,
+      "learning_rate": 5.392368424247123e-05,
+      "loss": 2.6606,
+      "step": 6559
+    },
+    {
+      "epoch": 0.5938263782022268,
+      "grad_norm": 0.936378538608551,
+      "learning_rate": 5.38955225613069e-05,
+      "loss": 2.6774,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5939169005159772,
+      "grad_norm": 0.8802645206451416,
+      "learning_rate": 5.38673655228621e-05,
+      "loss": 2.6995,
+      "step": 6561
+    },
+    {
+      "epoch": 0.5940074228297275,
+      "grad_norm": 0.8793949484825134,
+      "learning_rate": 5.383921312997242e-05,
+      "loss": 2.7237,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5940979451434779,
+      "grad_norm": 0.8559528589248657,
+      "learning_rate": 5.381106538547266e-05,
+      "loss": 2.6689,
+      "step": 6563
+    },
+    {
+      "epoch": 0.5941884674572282,
+      "grad_norm": 0.9179282188415527,
+      "learning_rate": 5.3782922292197354e-05,
+      "loss": 2.6752,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5942789897709786,
+      "grad_norm": 0.9603455662727356,
+      "learning_rate": 5.375478385298052e-05,
+      "loss": 2.7392,
+      "step": 6565
+    },
+    {
+      "epoch": 0.5943695120847289,
+      "grad_norm": 0.8355276584625244,
+      "learning_rate": 5.372665007065567e-05,
+      "loss": 2.6832,
+      "step": 6566
+    },
+    {
+      "epoch": 0.5944600343984793,
+      "grad_norm": 0.8501160740852356,
+      "learning_rate": 5.3698520948055906e-05,
+      "loss": 2.6979,
+      "step": 6567
+    },
+    {
+      "epoch": 0.5945505567122296,
+      "grad_norm": 0.8548684120178223,
+      "learning_rate": 5.3670396488013854e-05,
+      "loss": 2.67,
+      "step": 6568
+    },
+    {
+      "epoch": 0.59464107902598,
+      "grad_norm": 0.8284228444099426,
+      "learning_rate": 5.364227669336153e-05,
+      "loss": 2.6429,
+      "step": 6569
+    },
+    {
+      "epoch": 0.5947316013397302,
+      "grad_norm": 0.8854268193244934,
+      "learning_rate": 5.361416156693075e-05,
+      "loss": 2.6237,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5948221236534805,
+      "grad_norm": 0.9293057322502136,
+      "learning_rate": 5.358605111155259e-05,
+      "loss": 2.704,
+      "step": 6571
+    },
+    {
+      "epoch": 0.5949126459672309,
+      "grad_norm": 0.917051374912262,
+      "learning_rate": 5.3557945330057813e-05,
+      "loss": 2.7112,
+      "step": 6572
+    },
+    {
+      "epoch": 0.5950031682809812,
+      "grad_norm": 0.8866673111915588,
+      "learning_rate": 5.352984422527667e-05,
+      "loss": 2.6689,
+      "step": 6573
+    },
+    {
+      "epoch": 0.5950936905947316,
+      "grad_norm": 0.8097630143165588,
+      "learning_rate": 5.3501747800038934e-05,
+      "loss": 2.6382,
+      "step": 6574
+    },
+    {
+      "epoch": 0.5951842129084819,
+      "grad_norm": 0.8683230876922607,
+      "learning_rate": 5.347365605717394e-05,
+      "loss": 2.7184,
+      "step": 6575
+    },
+    {
+      "epoch": 0.5952747352222323,
+      "grad_norm": 0.9342178106307983,
+      "learning_rate": 5.344556899951054e-05,
+      "loss": 2.7366,
+      "step": 6576
+    },
+    {
+      "epoch": 0.5953652575359826,
+      "grad_norm": 0.838149905204773,
+      "learning_rate": 5.3417486629876986e-05,
+      "loss": 2.737,
+      "step": 6577
+    },
+    {
+      "epoch": 0.595455779849733,
+      "grad_norm": 0.8259788751602173,
+      "learning_rate": 5.3389408951101336e-05,
+      "loss": 2.7081,
+      "step": 6578
+    },
+    {
+      "epoch": 0.5955463021634833,
+      "grad_norm": 0.8724661469459534,
+      "learning_rate": 5.33613359660109e-05,
+      "loss": 2.6943,
+      "step": 6579
+    },
+    {
+      "epoch": 0.5956368244772337,
+      "grad_norm": 0.8654960989952087,
+      "learning_rate": 5.333326767743263e-05,
+      "loss": 2.7003,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5957273467909839,
+      "grad_norm": 0.9523176550865173,
+      "learning_rate": 5.330520408819304e-05,
+      "loss": 2.6237,
+      "step": 6581
+    },
+    {
+      "epoch": 0.5958178691047343,
+      "grad_norm": 0.9099210500717163,
+      "learning_rate": 5.327714520111813e-05,
+      "loss": 2.6642,
+      "step": 6582
+    },
+    {
+      "epoch": 0.5959083914184846,
+      "grad_norm": 0.8548341393470764,
+      "learning_rate": 5.324909101903341e-05,
+      "loss": 2.6689,
+      "step": 6583
+    },
+    {
+      "epoch": 0.595998913732235,
+      "grad_norm": 0.9076147675514221,
+      "learning_rate": 5.322104154476399e-05,
+      "loss": 2.687,
+      "step": 6584
+    },
+    {
+      "epoch": 0.5960894360459853,
+      "grad_norm": 0.8491636514663696,
+      "learning_rate": 5.319299678113432e-05,
+      "loss": 2.7021,
+      "step": 6585
+    },
+    {
+      "epoch": 0.5961799583597357,
+      "grad_norm": 0.8565483689308167,
+      "learning_rate": 5.316495673096869e-05,
+      "loss": 2.7048,
+      "step": 6586
+    },
+    {
+      "epoch": 0.596270480673486,
+      "grad_norm": 0.8949451446533203,
+      "learning_rate": 5.313692139709061e-05,
+      "loss": 2.7275,
+      "step": 6587
+    },
+    {
+      "epoch": 0.5963610029872364,
+      "grad_norm": 0.8468038439750671,
+      "learning_rate": 5.31088907823232e-05,
+      "loss": 2.6339,
+      "step": 6588
+    },
+    {
+      "epoch": 0.5964515253009867,
+      "grad_norm": 0.8995200991630554,
+      "learning_rate": 5.308086488948929e-05,
+      "loss": 2.6561,
+      "step": 6589
+    },
+    {
+      "epoch": 0.5965420476147371,
+      "grad_norm": 0.8418900370597839,
+      "learning_rate": 5.305284372141095e-05,
+      "loss": 2.7061,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5966325699284873,
+      "grad_norm": 0.8721002340316772,
+      "learning_rate": 5.302482728090996e-05,
+      "loss": 2.727,
+      "step": 6591
+    },
+    {
+      "epoch": 0.5967230922422377,
+      "grad_norm": 0.9773480296134949,
+      "learning_rate": 5.299681557080759e-05,
+      "loss": 2.7723,
+      "step": 6592
+    },
+    {
+      "epoch": 0.596813614555988,
+      "grad_norm": 0.8725921511650085,
+      "learning_rate": 5.2968808593924604e-05,
+      "loss": 2.6763,
+      "step": 6593
+    },
+    {
+      "epoch": 0.5969041368697384,
+      "grad_norm": 0.8329956531524658,
+      "learning_rate": 5.2940806353081296e-05,
+      "loss": 2.6554,
+      "step": 6594
+    },
+    {
+      "epoch": 0.5969946591834887,
+      "grad_norm": 0.7967681288719177,
+      "learning_rate": 5.291280885109755e-05,
+      "loss": 2.7187,
+      "step": 6595
+    },
+    {
+      "epoch": 0.5970851814972391,
+      "grad_norm": 0.861714780330658,
+      "learning_rate": 5.288481609079259e-05,
+      "loss": 2.6465,
+      "step": 6596
+    },
+    {
+      "epoch": 0.5971757038109894,
+      "grad_norm": 0.9225656390190125,
+      "learning_rate": 5.285682807498542e-05,
+      "loss": 2.6922,
+      "step": 6597
+    },
+    {
+      "epoch": 0.5972662261247398,
+      "grad_norm": 0.8872184753417969,
+      "learning_rate": 5.282884480649435e-05,
+      "loss": 2.7618,
+      "step": 6598
+    },
+    {
+      "epoch": 0.5973567484384901,
+      "grad_norm": 0.8705012798309326,
+      "learning_rate": 5.280086628813731e-05,
+      "loss": 2.7644,
+      "step": 6599
+    },
+    {
+      "epoch": 0.5974472707522405,
+      "grad_norm": 0.8825952410697937,
+      "learning_rate": 5.277289252273174e-05,
+      "loss": 2.7019,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5975377930659908,
+      "grad_norm": 0.8400018811225891,
+      "learning_rate": 5.274492351309461e-05,
+      "loss": 2.654,
+      "step": 6601
+    },
+    {
+      "epoch": 0.5976283153797411,
+      "grad_norm": 0.8624444007873535,
+      "learning_rate": 5.2716959262042386e-05,
+      "loss": 2.6893,
+      "step": 6602
+    },
+    {
+      "epoch": 0.5977188376934914,
+      "grad_norm": 0.84315425157547,
+      "learning_rate": 5.26889997723911e-05,
+      "loss": 2.6576,
+      "step": 6603
+    },
+    {
+      "epoch": 0.5978093600072418,
+      "grad_norm": 0.960757315158844,
+      "learning_rate": 5.266104504695617e-05,
+      "loss": 2.66,
+      "step": 6604
+    },
+    {
+      "epoch": 0.5978998823209921,
+      "grad_norm": 0.8555940985679626,
+      "learning_rate": 5.26330950885528e-05,
+      "loss": 2.6179,
+      "step": 6605
+    },
+    {
+      "epoch": 0.5979904046347425,
+      "grad_norm": 0.9326249361038208,
+      "learning_rate": 5.2605149899995406e-05,
+      "loss": 2.6908,
+      "step": 6606
+    },
+    {
+      "epoch": 0.5980809269484928,
+      "grad_norm": 0.9245107769966125,
+      "learning_rate": 5.257720948409812e-05,
+      "loss": 2.7202,
+      "step": 6607
+    },
+    {
+      "epoch": 0.5981714492622432,
+      "grad_norm": 0.9167315363883972,
+      "learning_rate": 5.254927384367454e-05,
+      "loss": 2.6924,
+      "step": 6608
+    },
+    {
+      "epoch": 0.5982619715759935,
+      "grad_norm": 0.8900579214096069,
+      "learning_rate": 5.252134298153779e-05,
+      "loss": 2.6718,
+      "step": 6609
+    },
+    {
+      "epoch": 0.5983524938897439,
+      "grad_norm": 0.8357805609703064,
+      "learning_rate": 5.249341690050051e-05,
+      "loss": 2.6777,
+      "step": 6610
+    },
+    {
+      "epoch": 0.5984430162034942,
+      "grad_norm": 0.8500257134437561,
+      "learning_rate": 5.246549560337488e-05,
+      "loss": 2.669,
+      "step": 6611
+    },
+    {
+      "epoch": 0.5985335385172444,
+      "grad_norm": 0.9223446846008301,
+      "learning_rate": 5.243757909297247e-05,
+      "loss": 2.7136,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5986240608309948,
+      "grad_norm": 0.9422861337661743,
+      "learning_rate": 5.2409667372104607e-05,
+      "loss": 2.7072,
+      "step": 6613
+    },
+    {
+      "epoch": 0.5987145831447451,
+      "grad_norm": 0.841312050819397,
+      "learning_rate": 5.2381760443581916e-05,
+      "loss": 2.6323,
+      "step": 6614
+    },
+    {
+      "epoch": 0.5988051054584955,
+      "grad_norm": 0.9254672527313232,
+      "learning_rate": 5.2353858310214635e-05,
+      "loss": 2.7042,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5988956277722458,
+      "grad_norm": 0.8144631385803223,
+      "learning_rate": 5.232596097481251e-05,
+      "loss": 2.6905,
+      "step": 6616
+    },
+    {
+      "epoch": 0.5989861500859962,
+      "grad_norm": 0.8707183003425598,
+      "learning_rate": 5.229806844018487e-05,
+      "loss": 2.7542,
+      "step": 6617
+    },
+    {
+      "epoch": 0.5990766723997465,
+      "grad_norm": 0.9004245400428772,
+      "learning_rate": 5.227018070914034e-05,
+      "loss": 2.7192,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5991671947134969,
+      "grad_norm": 0.8687030673027039,
+      "learning_rate": 5.2242297784487395e-05,
+      "loss": 2.6497,
+      "step": 6619
+    },
+    {
+      "epoch": 0.5992577170272472,
+      "grad_norm": 0.8232517242431641,
+      "learning_rate": 5.221441966903371e-05,
+      "loss": 2.7078,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5993482393409976,
+      "grad_norm": 0.7970709204673767,
+      "learning_rate": 5.218654636558666e-05,
+      "loss": 2.7088,
+      "step": 6621
+    },
+    {
+      "epoch": 0.5994387616547479,
+      "grad_norm": 0.8401699662208557,
+      "learning_rate": 5.2158677876953075e-05,
+      "loss": 2.6806,
+      "step": 6622
+    },
+    {
+      "epoch": 0.5995292839684982,
+      "grad_norm": 0.8412536978721619,
+      "learning_rate": 5.213081420593933e-05,
+      "loss": 2.7362,
+      "step": 6623
+    },
+    {
+      "epoch": 0.5996198062822485,
+      "grad_norm": 0.835806131362915,
+      "learning_rate": 5.2102955355351276e-05,
+      "loss": 2.7277,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5997103285959989,
+      "grad_norm": 0.7974480986595154,
+      "learning_rate": 5.207510132799436e-05,
+      "loss": 2.6583,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5998008509097492,
+      "grad_norm": 0.8804001212120056,
+      "learning_rate": 5.204725212667334e-05,
+      "loss": 2.7269,
+      "step": 6626
+    },
+    {
+      "epoch": 0.5998913732234996,
+      "grad_norm": 0.8101835250854492,
+      "learning_rate": 5.201940775419278e-05,
+      "loss": 2.6732,
+      "step": 6627
+    },
+    {
+      "epoch": 0.5999818955372499,
+      "grad_norm": 0.815082311630249,
+      "learning_rate": 5.199156821335653e-05,
+      "loss": 2.6888,
+      "step": 6628
+    },
+    {
+      "epoch": 0.6000724178510003,
+      "grad_norm": 0.8119527101516724,
+      "learning_rate": 5.1963733506968046e-05,
+      "loss": 2.5854,
+      "step": 6629
+    },
+    {
+      "epoch": 0.6001629401647506,
+      "grad_norm": 0.841860830783844,
+      "learning_rate": 5.193590363783028e-05,
+      "loss": 2.6658,
+      "step": 6630
+    },
+    {
+      "epoch": 0.600253462478501,
+      "grad_norm": 0.8963244557380676,
+      "learning_rate": 5.1908078608745705e-05,
+      "loss": 2.7245,
+      "step": 6631
+    },
+    {
+      "epoch": 0.6003439847922513,
+      "grad_norm": 0.8063508868217468,
+      "learning_rate": 5.1880258422516294e-05,
+      "loss": 2.6822,
+      "step": 6632
+    },
+    {
+      "epoch": 0.6004345071060017,
+      "grad_norm": 0.830461859703064,
+      "learning_rate": 5.185244308194358e-05,
+      "loss": 2.6834,
+      "step": 6633
+    },
+    {
+      "epoch": 0.6005250294197519,
+      "grad_norm": 0.8261478543281555,
+      "learning_rate": 5.182463258982846e-05,
+      "loss": 2.6629,
+      "step": 6634
+    },
+    {
+      "epoch": 0.6006155517335023,
+      "grad_norm": 0.8018027544021606,
+      "learning_rate": 5.179682694897159e-05,
+      "loss": 2.6857,
+      "step": 6635
+    },
+    {
+      "epoch": 0.6007060740472526,
+      "grad_norm": 0.8765165209770203,
+      "learning_rate": 5.17690261621729e-05,
+      "loss": 2.6925,
+      "step": 6636
+    },
+    {
+      "epoch": 0.600796596361003,
+      "grad_norm": 0.8018073439598083,
+      "learning_rate": 5.174123023223194e-05,
+      "loss": 2.7014,
+      "step": 6637
+    },
+    {
+      "epoch": 0.6008871186747533,
+      "grad_norm": 0.8557896018028259,
+      "learning_rate": 5.1713439161947785e-05,
+      "loss": 2.6468,
+      "step": 6638
+    },
+    {
+      "epoch": 0.6009776409885037,
+      "grad_norm": 0.8501036167144775,
+      "learning_rate": 5.168565295411898e-05,
+      "loss": 2.6903,
+      "step": 6639
+    },
+    {
+      "epoch": 0.601068163302254,
+      "grad_norm": 0.8637775182723999,
+      "learning_rate": 5.1657871611543605e-05,
+      "loss": 2.6627,
+      "step": 6640
+    },
+    {
+      "epoch": 0.6011586856160044,
+      "grad_norm": 0.8541369438171387,
+      "learning_rate": 5.163009513701929e-05,
+      "loss": 2.7018,
+      "step": 6641
+    },
+    {
+      "epoch": 0.6012492079297547,
+      "grad_norm": 0.8165152668952942,
+      "learning_rate": 5.1602323533342964e-05,
+      "loss": 2.6724,
+      "step": 6642
+    },
+    {
+      "epoch": 0.6013397302435051,
+      "grad_norm": 0.965965747833252,
+      "learning_rate": 5.157455680331142e-05,
+      "loss": 2.6428,
+      "step": 6643
+    },
+    {
+      "epoch": 0.6014302525572554,
+      "grad_norm": 0.8963990211486816,
+      "learning_rate": 5.154679494972066e-05,
+      "loss": 2.68,
+      "step": 6644
+    },
+    {
+      "epoch": 0.6015207748710057,
+      "grad_norm": 0.8216471672058105,
+      "learning_rate": 5.15190379753663e-05,
+      "loss": 2.6938,
+      "step": 6645
+    },
+    {
+      "epoch": 0.601611297184756,
+      "grad_norm": 0.8308603167533875,
+      "learning_rate": 5.149128588304351e-05,
+      "loss": 2.6742,
+      "step": 6646
+    },
+    {
+      "epoch": 0.6017018194985064,
+      "grad_norm": 0.9278542995452881,
+      "learning_rate": 5.146353867554688e-05,
+      "loss": 2.7318,
+      "step": 6647
+    },
+    {
+      "epoch": 0.6017923418122567,
+      "grad_norm": 0.9083507061004639,
+      "learning_rate": 5.1435796355670594e-05,
+      "loss": 2.7105,
+      "step": 6648
+    },
+    {
+      "epoch": 0.6018828641260071,
+      "grad_norm": 0.865098237991333,
+      "learning_rate": 5.140805892620833e-05,
+      "loss": 2.6751,
+      "step": 6649
+    },
+    {
+      "epoch": 0.6019733864397574,
+      "grad_norm": 0.813866913318634,
+      "learning_rate": 5.138032638995315e-05,
+      "loss": 2.6353,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6020639087535078,
+      "grad_norm": 0.970386803150177,
+      "learning_rate": 5.135259874969778e-05,
+      "loss": 2.6791,
+      "step": 6651
+    },
+    {
+      "epoch": 0.6021544310672581,
+      "grad_norm": 0.9106497764587402,
+      "learning_rate": 5.132487600823438e-05,
+      "loss": 2.7343,
+      "step": 6652
+    },
+    {
+      "epoch": 0.6022449533810084,
+      "grad_norm": 0.8499709367752075,
+      "learning_rate": 5.129715816835463e-05,
+      "loss": 2.733,
+      "step": 6653
+    },
+    {
+      "epoch": 0.6023354756947588,
+      "grad_norm": 0.9132952094078064,
+      "learning_rate": 5.1269445232849734e-05,
+      "loss": 2.707,
+      "step": 6654
+    },
+    {
+      "epoch": 0.602425998008509,
+      "grad_norm": 0.9442603588104248,
+      "learning_rate": 5.1241737204510395e-05,
+      "loss": 2.7212,
+      "step": 6655
+    },
+    {
+      "epoch": 0.6025165203222594,
+      "grad_norm": 1.0592482089996338,
+      "learning_rate": 5.121403408612672e-05,
+      "loss": 2.7101,
+      "step": 6656
+    },
+    {
+      "epoch": 0.6026070426360097,
+      "grad_norm": 0.831420361995697,
+      "learning_rate": 5.1186335880488544e-05,
+      "loss": 2.6751,
+      "step": 6657
+    },
+    {
+      "epoch": 0.6026975649497601,
+      "grad_norm": 0.8643193244934082,
+      "learning_rate": 5.115864259038498e-05,
+      "loss": 2.71,
+      "step": 6658
+    },
+    {
+      "epoch": 0.6027880872635104,
+      "grad_norm": 0.8968473076820374,
+      "learning_rate": 5.1130954218604767e-05,
+      "loss": 2.6981,
+      "step": 6659
+    },
+    {
+      "epoch": 0.6028786095772608,
+      "grad_norm": 0.8939312696456909,
+      "learning_rate": 5.110327076793613e-05,
+      "loss": 2.7011,
+      "step": 6660
+    },
+    {
+      "epoch": 0.6029691318910111,
+      "grad_norm": 0.8195379376411438,
+      "learning_rate": 5.107559224116678e-05,
+      "loss": 2.6823,
+      "step": 6661
+    },
+    {
+      "epoch": 0.6030596542047615,
+      "grad_norm": 0.8442776799201965,
+      "learning_rate": 5.104791864108396e-05,
+      "loss": 2.696,
+      "step": 6662
+    },
+    {
+      "epoch": 0.6031501765185118,
+      "grad_norm": 0.9072765707969666,
+      "learning_rate": 5.102024997047443e-05,
+      "loss": 2.7176,
+      "step": 6663
+    },
+    {
+      "epoch": 0.6032406988322622,
+      "grad_norm": 0.8840109705924988,
+      "learning_rate": 5.099258623212431e-05,
+      "loss": 2.6998,
+      "step": 6664
+    },
+    {
+      "epoch": 0.6033312211460125,
+      "grad_norm": 0.8407348990440369,
+      "learning_rate": 5.096492742881949e-05,
+      "loss": 2.6591,
+      "step": 6665
+    },
+    {
+      "epoch": 0.6034217434597628,
+      "grad_norm": 0.8747027516365051,
+      "learning_rate": 5.09372735633451e-05,
+      "loss": 2.6933,
+      "step": 6666
+    },
+    {
+      "epoch": 0.6035122657735131,
+      "grad_norm": 0.8869914412498474,
+      "learning_rate": 5.090962463848592e-05,
+      "loss": 2.7229,
+      "step": 6667
+    },
+    {
+      "epoch": 0.6036027880872635,
+      "grad_norm": 0.8956210613250732,
+      "learning_rate": 5.0881980657026195e-05,
+      "loss": 2.6556,
+      "step": 6668
+    },
+    {
+      "epoch": 0.6036933104010138,
+      "grad_norm": 0.8411816358566284,
+      "learning_rate": 5.085434162174968e-05,
+      "loss": 2.707,
+      "step": 6669
+    },
+    {
+      "epoch": 0.6037838327147642,
+      "grad_norm": 0.8487591743469238,
+      "learning_rate": 5.082670753543961e-05,
+      "loss": 2.754,
+      "step": 6670
+    },
+    {
+      "epoch": 0.6038743550285145,
+      "grad_norm": 0.8434279561042786,
+      "learning_rate": 5.0799078400878786e-05,
+      "loss": 2.666,
+      "step": 6671
+    },
+    {
+      "epoch": 0.6039648773422649,
+      "grad_norm": 0.8222647309303284,
+      "learning_rate": 5.077145422084939e-05,
+      "loss": 2.7008,
+      "step": 6672
+    },
+    {
+      "epoch": 0.6040553996560152,
+      "grad_norm": 0.9104819893836975,
+      "learning_rate": 5.0743834998133196e-05,
+      "loss": 2.6976,
+      "step": 6673
+    },
+    {
+      "epoch": 0.6041459219697656,
+      "grad_norm": 0.829352617263794,
+      "learning_rate": 5.071622073551148e-05,
+      "loss": 2.6813,
+      "step": 6674
+    },
+    {
+      "epoch": 0.6042364442835159,
+      "grad_norm": 0.8321340680122375,
+      "learning_rate": 5.0688611435764975e-05,
+      "loss": 2.6406,
+      "step": 6675
+    },
+    {
+      "epoch": 0.6043269665972663,
+      "grad_norm": 0.8086342215538025,
+      "learning_rate": 5.066100710167401e-05,
+      "loss": 2.6696,
+      "step": 6676
+    },
+    {
+      "epoch": 0.6044174889110165,
+      "grad_norm": 0.8482884764671326,
+      "learning_rate": 5.063340773601818e-05,
+      "loss": 2.7286,
+      "step": 6677
+    },
+    {
+      "epoch": 0.6045080112247669,
+      "grad_norm": 0.857356607913971,
+      "learning_rate": 5.0605813341576924e-05,
+      "loss": 2.6962,
+      "step": 6678
+    },
+    {
+      "epoch": 0.6045985335385172,
+      "grad_norm": 0.9065026640892029,
+      "learning_rate": 5.0578223921128875e-05,
+      "loss": 2.7226,
+      "step": 6679
+    },
+    {
+      "epoch": 0.6046890558522676,
+      "grad_norm": 0.8281168937683105,
+      "learning_rate": 5.055063947745233e-05,
+      "loss": 2.6442,
+      "step": 6680
+    },
+    {
+      "epoch": 0.6047795781660179,
+      "grad_norm": 0.8826650977134705,
+      "learning_rate": 5.052306001332503e-05,
+      "loss": 2.7138,
+      "step": 6681
+    },
+    {
+      "epoch": 0.6048701004797683,
+      "grad_norm": 0.8792107105255127,
+      "learning_rate": 5.049548553152428e-05,
+      "loss": 2.7541,
+      "step": 6682
+    },
+    {
+      "epoch": 0.6049606227935186,
+      "grad_norm": 0.7904659509658813,
+      "learning_rate": 5.0467916034826704e-05,
+      "loss": 2.7115,
+      "step": 6683
+    },
+    {
+      "epoch": 0.605051145107269,
+      "grad_norm": 0.891960084438324,
+      "learning_rate": 5.044035152600869e-05,
+      "loss": 2.6897,
+      "step": 6684
+    },
+    {
+      "epoch": 0.6051416674210193,
+      "grad_norm": 0.9381482005119324,
+      "learning_rate": 5.04127920078459e-05,
+      "loss": 2.7353,
+      "step": 6685
+    },
+    {
+      "epoch": 0.6052321897347697,
+      "grad_norm": 0.8143098950386047,
+      "learning_rate": 5.038523748311359e-05,
+      "loss": 2.6349,
+      "step": 6686
+    },
+    {
+      "epoch": 0.60532271204852,
+      "grad_norm": 0.837303102016449,
+      "learning_rate": 5.0357687954586506e-05,
+      "loss": 2.6685,
+      "step": 6687
+    },
+    {
+      "epoch": 0.6054132343622703,
+      "grad_norm": 0.8285949230194092,
+      "learning_rate": 5.033014342503889e-05,
+      "loss": 2.6775,
+      "step": 6688
+    },
+    {
+      "epoch": 0.6055037566760206,
+      "grad_norm": 0.8302776217460632,
+      "learning_rate": 5.0302603897244474e-05,
+      "loss": 2.7217,
+      "step": 6689
+    },
+    {
+      "epoch": 0.605594278989771,
+      "grad_norm": 0.7810454368591309,
+      "learning_rate": 5.027506937397652e-05,
+      "loss": 2.6596,
+      "step": 6690
+    },
+    {
+      "epoch": 0.6056848013035213,
+      "grad_norm": 0.8588391542434692,
+      "learning_rate": 5.0247539858007644e-05,
+      "loss": 2.6877,
+      "step": 6691
+    },
+    {
+      "epoch": 0.6057753236172717,
+      "grad_norm": 0.9216744899749756,
+      "learning_rate": 5.022001535211022e-05,
+      "loss": 2.7234,
+      "step": 6692
+    },
+    {
+      "epoch": 0.605865845931022,
+      "grad_norm": 0.8668200373649597,
+      "learning_rate": 5.019249585905585e-05,
+      "loss": 2.6811,
+      "step": 6693
+    },
+    {
+      "epoch": 0.6059563682447723,
+      "grad_norm": 0.7832801938056946,
+      "learning_rate": 5.0164981381615786e-05,
+      "loss": 2.654,
+      "step": 6694
+    },
+    {
+      "epoch": 0.6060468905585227,
+      "grad_norm": 0.8727904558181763,
+      "learning_rate": 5.0137471922560734e-05,
+      "loss": 2.6633,
+      "step": 6695
+    },
+    {
+      "epoch": 0.606137412872273,
+      "grad_norm": 0.8319174647331238,
+      "learning_rate": 5.010996748466088e-05,
+      "loss": 2.733,
+      "step": 6696
+    },
+    {
+      "epoch": 0.6062279351860234,
+      "grad_norm": 0.8686563372612,
+      "learning_rate": 5.008246807068595e-05,
+      "loss": 2.7043,
+      "step": 6697
+    },
+    {
+      "epoch": 0.6063184574997736,
+      "grad_norm": 0.8815221786499023,
+      "learning_rate": 5.005497368340516e-05,
+      "loss": 2.7486,
+      "step": 6698
+    },
+    {
+      "epoch": 0.606408979813524,
+      "grad_norm": 0.8359525799751282,
+      "learning_rate": 5.002748432558707e-05,
+      "loss": 2.6408,
+      "step": 6699
+    },
+    {
+      "epoch": 0.6064995021272743,
+      "grad_norm": 0.8514461517333984,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 2.6483,
+      "step": 6700
+    },
+    {
+      "epoch": 0.6065900244410247,
+      "grad_norm": 0.890376091003418,
+      "learning_rate": 4.9972520709411565e-05,
+      "loss": 2.6174,
+      "step": 6701
+    },
+    {
+      "epoch": 0.606680546754775,
+      "grad_norm": 0.8972718715667725,
+      "learning_rate": 4.9945046456588905e-05,
+      "loss": 2.7029,
+      "step": 6702
+    },
+    {
+      "epoch": 0.6067710690685254,
+      "grad_norm": 0.8866444230079651,
+      "learning_rate": 4.9917577244298696e-05,
+      "loss": 2.718,
+      "step": 6703
+    },
+    {
+      "epoch": 0.6068615913822757,
+      "grad_norm": 0.8488792777061462,
+      "learning_rate": 4.9890113075307085e-05,
+      "loss": 2.7348,
+      "step": 6704
+    },
+    {
+      "epoch": 0.6069521136960261,
+      "grad_norm": 0.8102442622184753,
+      "learning_rate": 4.986265395237972e-05,
+      "loss": 2.6668,
+      "step": 6705
+    },
+    {
+      "epoch": 0.6070426360097764,
+      "grad_norm": 0.8155598640441895,
+      "learning_rate": 4.9835199878281765e-05,
+      "loss": 2.6679,
+      "step": 6706
+    },
+    {
+      "epoch": 0.6071331583235268,
+      "grad_norm": 0.8324066400527954,
+      "learning_rate": 4.980775085577775e-05,
+      "loss": 2.7047,
+      "step": 6707
+    },
+    {
+      "epoch": 0.607223680637277,
+      "grad_norm": 0.8296021223068237,
+      "learning_rate": 4.978030688763191e-05,
+      "loss": 2.6714,
+      "step": 6708
+    },
+    {
+      "epoch": 0.6073142029510274,
+      "grad_norm": 0.8644400238990784,
+      "learning_rate": 4.975286797660775e-05,
+      "loss": 2.7341,
+      "step": 6709
+    },
+    {
+      "epoch": 0.6074047252647777,
+      "grad_norm": 0.8654029369354248,
+      "learning_rate": 4.972543412546842e-05,
+      "loss": 2.6972,
+      "step": 6710
+    },
+    {
+      "epoch": 0.6074952475785281,
+      "grad_norm": 0.848343551158905,
+      "learning_rate": 4.969800533697649e-05,
+      "loss": 2.6639,
+      "step": 6711
+    },
+    {
+      "epoch": 0.6075857698922784,
+      "grad_norm": 0.8281153440475464,
+      "learning_rate": 4.9670581613894094e-05,
+      "loss": 2.6206,
+      "step": 6712
+    },
+    {
+      "epoch": 0.6076762922060288,
+      "grad_norm": 0.8595371246337891,
+      "learning_rate": 4.964316295898268e-05,
+      "loss": 2.6815,
+      "step": 6713
+    },
+    {
+      "epoch": 0.6077668145197791,
+      "grad_norm": 0.8331053256988525,
+      "learning_rate": 4.961574937500344e-05,
+      "loss": 2.6783,
+      "step": 6714
+    },
+    {
+      "epoch": 0.6078573368335295,
+      "grad_norm": 0.8491067290306091,
+      "learning_rate": 4.958834086471683e-05,
+      "loss": 2.7056,
+      "step": 6715
+    },
+    {
+      "epoch": 0.6079478591472798,
+      "grad_norm": 0.8933209776878357,
+      "learning_rate": 4.956093743088291e-05,
+      "loss": 2.7263,
+      "step": 6716
+    },
+    {
+      "epoch": 0.6080383814610302,
+      "grad_norm": 0.883208155632019,
+      "learning_rate": 4.953353907626122e-05,
+      "loss": 2.6177,
+      "step": 6717
+    },
+    {
+      "epoch": 0.6081289037747805,
+      "grad_norm": 0.8465838432312012,
+      "learning_rate": 4.950614580361076e-05,
+      "loss": 2.6835,
+      "step": 6718
+    },
+    {
+      "epoch": 0.6082194260885309,
+      "grad_norm": 0.8147262334823608,
+      "learning_rate": 4.947875761569003e-05,
+      "loss": 2.6652,
+      "step": 6719
+    },
+    {
+      "epoch": 0.6083099484022811,
+      "grad_norm": 0.9500305652618408,
+      "learning_rate": 4.945137451525707e-05,
+      "loss": 2.6889,
+      "step": 6720
+    },
+    {
+      "epoch": 0.6084004707160315,
+      "grad_norm": 0.9218015074729919,
+      "learning_rate": 4.942399650506923e-05,
+      "loss": 2.6836,
+      "step": 6721
+    },
+    {
+      "epoch": 0.6084909930297818,
+      "grad_norm": 0.7956727743148804,
+      "learning_rate": 4.939662358788364e-05,
+      "loss": 2.6653,
+      "step": 6722
+    },
+    {
+      "epoch": 0.6085815153435322,
+      "grad_norm": 0.9122270345687866,
+      "learning_rate": 4.936925576645663e-05,
+      "loss": 2.7115,
+      "step": 6723
+    },
+    {
+      "epoch": 0.6086720376572825,
+      "grad_norm": 0.8649951219558716,
+      "learning_rate": 4.9341893043544185e-05,
+      "loss": 2.7205,
+      "step": 6724
+    },
+    {
+      "epoch": 0.6087625599710329,
+      "grad_norm": 0.9105194211006165,
+      "learning_rate": 4.9314535421901716e-05,
+      "loss": 2.7186,
+      "step": 6725
+    },
+    {
+      "epoch": 0.6088530822847832,
+      "grad_norm": 0.8381279706954956,
+      "learning_rate": 4.9287182904284143e-05,
+      "loss": 2.7746,
+      "step": 6726
+    },
+    {
+      "epoch": 0.6089436045985336,
+      "grad_norm": 0.8129174113273621,
+      "learning_rate": 4.925983549344587e-05,
+      "loss": 2.6485,
+      "step": 6727
+    },
+    {
+      "epoch": 0.6090341269122839,
+      "grad_norm": 0.8500168323516846,
+      "learning_rate": 4.9232493192140806e-05,
+      "loss": 2.6848,
+      "step": 6728
+    },
+    {
+      "epoch": 0.6091246492260343,
+      "grad_norm": 0.9311719536781311,
+      "learning_rate": 4.920515600312222e-05,
+      "loss": 2.7047,
+      "step": 6729
+    },
+    {
+      "epoch": 0.6092151715397846,
+      "grad_norm": 0.7978945970535278,
+      "learning_rate": 4.9177823929143106e-05,
+      "loss": 2.6726,
+      "step": 6730
+    },
+    {
+      "epoch": 0.609305693853535,
+      "grad_norm": 0.8584246635437012,
+      "learning_rate": 4.9150496972955696e-05,
+      "loss": 2.6859,
+      "step": 6731
+    },
+    {
+      "epoch": 0.6093962161672852,
+      "grad_norm": 0.8876283764839172,
+      "learning_rate": 4.912317513731185e-05,
+      "loss": 2.6616,
+      "step": 6732
+    },
+    {
+      "epoch": 0.6094867384810356,
+      "grad_norm": 0.8028571605682373,
+      "learning_rate": 4.909585842496287e-05,
+      "loss": 2.7145,
+      "step": 6733
+    },
+    {
+      "epoch": 0.6095772607947859,
+      "grad_norm": 0.87955641746521,
+      "learning_rate": 4.906854683865956e-05,
+      "loss": 2.7036,
+      "step": 6734
+    },
+    {
+      "epoch": 0.6096677831085362,
+      "grad_norm": 0.8874965906143188,
+      "learning_rate": 4.9041240381152185e-05,
+      "loss": 2.725,
+      "step": 6735
+    },
+    {
+      "epoch": 0.6097583054222866,
+      "grad_norm": 0.9333867430686951,
+      "learning_rate": 4.901393905519055e-05,
+      "loss": 2.7052,
+      "step": 6736
+    },
+    {
+      "epoch": 0.6098488277360369,
+      "grad_norm": 0.8766831755638123,
+      "learning_rate": 4.898664286352378e-05,
+      "loss": 2.7184,
+      "step": 6737
+    },
+    {
+      "epoch": 0.6099393500497873,
+      "grad_norm": 0.8419327139854431,
+      "learning_rate": 4.8959351808900754e-05,
+      "loss": 2.7346,
+      "step": 6738
+    },
+    {
+      "epoch": 0.6100298723635376,
+      "grad_norm": 0.9690167307853699,
+      "learning_rate": 4.893206589406957e-05,
+      "loss": 2.8024,
+      "step": 6739
+    },
+    {
+      "epoch": 0.610120394677288,
+      "grad_norm": 0.8833349347114563,
+      "learning_rate": 4.890478512177795e-05,
+      "loss": 2.7829,
+      "step": 6740
+    },
+    {
+      "epoch": 0.6102109169910382,
+      "grad_norm": 0.8960972428321838,
+      "learning_rate": 4.8877509494773065e-05,
+      "loss": 2.7061,
+      "step": 6741
+    },
+    {
+      "epoch": 0.6103014393047886,
+      "grad_norm": 0.9377439022064209,
+      "learning_rate": 4.8850239015801625e-05,
+      "loss": 2.6516,
+      "step": 6742
+    },
+    {
+      "epoch": 0.6103919616185389,
+      "grad_norm": 0.9206840395927429,
+      "learning_rate": 4.882297368760963e-05,
+      "loss": 2.6613,
+      "step": 6743
+    },
+    {
+      "epoch": 0.6104824839322893,
+      "grad_norm": 0.9181665778160095,
+      "learning_rate": 4.8795713512942865e-05,
+      "loss": 2.6805,
+      "step": 6744
+    },
+    {
+      "epoch": 0.6105730062460396,
+      "grad_norm": 0.9120600819587708,
+      "learning_rate": 4.876845849454631e-05,
+      "loss": 2.6996,
+      "step": 6745
+    },
+    {
+      "epoch": 0.61066352855979,
+      "grad_norm": 0.8618726134300232,
+      "learning_rate": 4.8741208635164585e-05,
+      "loss": 2.7314,
+      "step": 6746
+    },
+    {
+      "epoch": 0.6107540508735403,
+      "grad_norm": 0.8103480935096741,
+      "learning_rate": 4.871396393754174e-05,
+      "loss": 2.6864,
+      "step": 6747
+    },
+    {
+      "epoch": 0.6108445731872907,
+      "grad_norm": 0.8761424422264099,
+      "learning_rate": 4.868672440442134e-05,
+      "loss": 2.6805,
+      "step": 6748
+    },
+    {
+      "epoch": 0.610935095501041,
+      "grad_norm": 0.8975646495819092,
+      "learning_rate": 4.865949003854637e-05,
+      "loss": 2.7368,
+      "step": 6749
+    },
+    {
+      "epoch": 0.6110256178147914,
+      "grad_norm": 0.9692808985710144,
+      "learning_rate": 4.8632260842659393e-05,
+      "loss": 2.7351,
+      "step": 6750
+    },
+    {
+      "epoch": 0.6111161401285417,
+      "grad_norm": 0.8636605739593506,
+      "learning_rate": 4.860503681950227e-05,
+      "loss": 2.653,
+      "step": 6751
+    },
+    {
+      "epoch": 0.611206662442292,
+      "grad_norm": 0.8828364014625549,
+      "learning_rate": 4.857781797181661e-05,
+      "loss": 2.7143,
+      "step": 6752
+    },
+    {
+      "epoch": 0.6112971847560423,
+      "grad_norm": 0.9609152674674988,
+      "learning_rate": 4.8550604302343227e-05,
+      "loss": 2.6827,
+      "step": 6753
+    },
+    {
+      "epoch": 0.6113877070697927,
+      "grad_norm": 1.0201287269592285,
+      "learning_rate": 4.852339581382258e-05,
+      "loss": 2.7189,
+      "step": 6754
+    },
+    {
+      "epoch": 0.611478229383543,
+      "grad_norm": 0.9233447313308716,
+      "learning_rate": 4.8496192508994576e-05,
+      "loss": 2.7395,
+      "step": 6755
+    },
+    {
+      "epoch": 0.6115687516972934,
+      "grad_norm": 0.8440586924552917,
+      "learning_rate": 4.8468994390598574e-05,
+      "loss": 2.6691,
+      "step": 6756
+    },
+    {
+      "epoch": 0.6116592740110437,
+      "grad_norm": 0.9626230001449585,
+      "learning_rate": 4.844180146137347e-05,
+      "loss": 2.6837,
+      "step": 6757
+    },
+    {
+      "epoch": 0.6117497963247941,
+      "grad_norm": 0.9260591268539429,
+      "learning_rate": 4.841461372405751e-05,
+      "loss": 2.7157,
+      "step": 6758
+    },
+    {
+      "epoch": 0.6118403186385444,
+      "grad_norm": 0.8617759943008423,
+      "learning_rate": 4.838743118138853e-05,
+      "loss": 2.6938,
+      "step": 6759
+    },
+    {
+      "epoch": 0.6119308409522948,
+      "grad_norm": 0.8547298908233643,
+      "learning_rate": 4.836025383610382e-05,
+      "loss": 2.705,
+      "step": 6760
+    },
+    {
+      "epoch": 0.6120213632660451,
+      "grad_norm": 0.8187577724456787,
+      "learning_rate": 4.833308169094014e-05,
+      "loss": 2.7203,
+      "step": 6761
+    },
+    {
+      "epoch": 0.6121118855797955,
+      "grad_norm": 0.8286332488059998,
+      "learning_rate": 4.830591474863372e-05,
+      "loss": 2.6601,
+      "step": 6762
+    },
+    {
+      "epoch": 0.6122024078935457,
+      "grad_norm": 0.8664364218711853,
+      "learning_rate": 4.8278753011920295e-05,
+      "loss": 2.7084,
+      "step": 6763
+    },
+    {
+      "epoch": 0.6122929302072961,
+      "grad_norm": 0.8383306860923767,
+      "learning_rate": 4.8251596483534965e-05,
+      "loss": 2.6873,
+      "step": 6764
+    },
+    {
+      "epoch": 0.6123834525210464,
+      "grad_norm": 0.8713318109512329,
+      "learning_rate": 4.822444516621252e-05,
+      "loss": 2.6643,
+      "step": 6765
+    },
+    {
+      "epoch": 0.6124739748347968,
+      "grad_norm": 0.8678194284439087,
+      "learning_rate": 4.8197299062686995e-05,
+      "loss": 2.6886,
+      "step": 6766
+    },
+    {
+      "epoch": 0.6125644971485471,
+      "grad_norm": 0.7932306528091431,
+      "learning_rate": 4.8170158175692035e-05,
+      "loss": 2.6389,
+      "step": 6767
+    },
+    {
+      "epoch": 0.6126550194622975,
+      "grad_norm": 0.8153578042984009,
+      "learning_rate": 4.8143022507960735e-05,
+      "loss": 2.6928,
+      "step": 6768
+    },
+    {
+      "epoch": 0.6127455417760478,
+      "grad_norm": 0.8641914129257202,
+      "learning_rate": 4.811589206222565e-05,
+      "loss": 2.715,
+      "step": 6769
+    },
+    {
+      "epoch": 0.6128360640897982,
+      "grad_norm": 0.8290460109710693,
+      "learning_rate": 4.808876684121881e-05,
+      "loss": 2.7175,
+      "step": 6770
+    },
+    {
+      "epoch": 0.6129265864035485,
+      "grad_norm": 0.8863869309425354,
+      "learning_rate": 4.806164684767177e-05,
+      "loss": 2.6914,
+      "step": 6771
+    },
+    {
+      "epoch": 0.6130171087172989,
+      "grad_norm": 0.8532030582427979,
+      "learning_rate": 4.80345320843154e-05,
+      "loss": 2.7317,
+      "step": 6772
+    },
+    {
+      "epoch": 0.6131076310310491,
+      "grad_norm": 0.8034926652908325,
+      "learning_rate": 4.800742255388029e-05,
+      "loss": 2.7586,
+      "step": 6773
+    },
+    {
+      "epoch": 0.6131981533447995,
+      "grad_norm": 0.8824958801269531,
+      "learning_rate": 4.798031825909627e-05,
+      "loss": 2.7163,
+      "step": 6774
+    },
+    {
+      "epoch": 0.6132886756585498,
+      "grad_norm": 0.8216482400894165,
+      "learning_rate": 4.795321920269279e-05,
+      "loss": 2.7146,
+      "step": 6775
+    },
+    {
+      "epoch": 0.6133791979723002,
+      "grad_norm": 0.8607343435287476,
+      "learning_rate": 4.79261253873987e-05,
+      "loss": 2.6636,
+      "step": 6776
+    },
+    {
+      "epoch": 0.6134697202860505,
+      "grad_norm": 0.9348889589309692,
+      "learning_rate": 4.78990368159424e-05,
+      "loss": 2.7435,
+      "step": 6777
+    },
+    {
+      "epoch": 0.6135602425998008,
+      "grad_norm": 0.8563113212585449,
+      "learning_rate": 4.787195349105159e-05,
+      "loss": 2.6861,
+      "step": 6778
+    },
+    {
+      "epoch": 0.6136507649135512,
+      "grad_norm": 0.8089884519577026,
+      "learning_rate": 4.7844875415453714e-05,
+      "loss": 2.7004,
+      "step": 6779
+    },
+    {
+      "epoch": 0.6137412872273015,
+      "grad_norm": 0.8360724449157715,
+      "learning_rate": 4.7817802591875426e-05,
+      "loss": 2.6961,
+      "step": 6780
+    },
+    {
+      "epoch": 0.6138318095410519,
+      "grad_norm": 0.7992052435874939,
+      "learning_rate": 4.779073502304299e-05,
+      "loss": 2.6625,
+      "step": 6781
+    },
+    {
+      "epoch": 0.6139223318548022,
+      "grad_norm": 0.7994864583015442,
+      "learning_rate": 4.776367271168209e-05,
+      "loss": 2.6471,
+      "step": 6782
+    },
+    {
+      "epoch": 0.6140128541685526,
+      "grad_norm": 0.8618261814117432,
+      "learning_rate": 4.773661566051793e-05,
+      "loss": 2.707,
+      "step": 6783
+    },
+    {
+      "epoch": 0.6141033764823028,
+      "grad_norm": 0.8256996273994446,
+      "learning_rate": 4.770956387227515e-05,
+      "loss": 2.6674,
+      "step": 6784
+    },
+    {
+      "epoch": 0.6141938987960532,
+      "grad_norm": 0.7925830483436584,
+      "learning_rate": 4.768251734967789e-05,
+      "loss": 2.7334,
+      "step": 6785
+    },
+    {
+      "epoch": 0.6142844211098035,
+      "grad_norm": 0.9798775315284729,
+      "learning_rate": 4.765547609544963e-05,
+      "loss": 2.6846,
+      "step": 6786
+    },
+    {
+      "epoch": 0.6143749434235539,
+      "grad_norm": 0.8514744639396667,
+      "learning_rate": 4.7628440112313556e-05,
+      "loss": 2.6665,
+      "step": 6787
+    },
+    {
+      "epoch": 0.6144654657373042,
+      "grad_norm": 0.9284042716026306,
+      "learning_rate": 4.7601409402992106e-05,
+      "loss": 2.6741,
+      "step": 6788
+    },
+    {
+      "epoch": 0.6145559880510546,
+      "grad_norm": 0.8871949315071106,
+      "learning_rate": 4.757438397020729e-05,
+      "loss": 2.7161,
+      "step": 6789
+    },
+    {
+      "epoch": 0.6146465103648049,
+      "grad_norm": 0.8882738351821899,
+      "learning_rate": 4.754736381668057e-05,
+      "loss": 2.7017,
+      "step": 6790
+    },
+    {
+      "epoch": 0.6147370326785553,
+      "grad_norm": 0.849459707736969,
+      "learning_rate": 4.7520348945132886e-05,
+      "loss": 2.7023,
+      "step": 6791
+    },
+    {
+      "epoch": 0.6148275549923056,
+      "grad_norm": 0.9086245894432068,
+      "learning_rate": 4.7493339358284625e-05,
+      "loss": 2.7293,
+      "step": 6792
+    },
+    {
+      "epoch": 0.614918077306056,
+      "grad_norm": 0.8342201709747314,
+      "learning_rate": 4.746633505885569e-05,
+      "loss": 2.6367,
+      "step": 6793
+    },
+    {
+      "epoch": 0.6150085996198063,
+      "grad_norm": 0.8293510675430298,
+      "learning_rate": 4.7439336049565295e-05,
+      "loss": 2.6908,
+      "step": 6794
+    },
+    {
+      "epoch": 0.6150991219335566,
+      "grad_norm": 0.831994354724884,
+      "learning_rate": 4.741234233313241e-05,
+      "loss": 2.6787,
+      "step": 6795
+    },
+    {
+      "epoch": 0.6151896442473069,
+      "grad_norm": 0.9759984016418457,
+      "learning_rate": 4.7385353912275165e-05,
+      "loss": 2.7383,
+      "step": 6796
+    },
+    {
+      "epoch": 0.6152801665610573,
+      "grad_norm": 0.92172771692276,
+      "learning_rate": 4.735837078971135e-05,
+      "loss": 2.702,
+      "step": 6797
+    },
+    {
+      "epoch": 0.6153706888748076,
+      "grad_norm": 0.892575204372406,
+      "learning_rate": 4.733139296815815e-05,
+      "loss": 2.6274,
+      "step": 6798
+    },
+    {
+      "epoch": 0.615461211188558,
+      "grad_norm": 0.8499800562858582,
+      "learning_rate": 4.7304420450332244e-05,
+      "loss": 2.6803,
+      "step": 6799
+    },
+    {
+      "epoch": 0.6155517335023083,
+      "grad_norm": 0.8400810360908508,
+      "learning_rate": 4.727745323894976e-05,
+      "loss": 2.6503,
+      "step": 6800
+    },
+    {
+      "epoch": 0.6156422558160587,
+      "grad_norm": 0.8385573625564575,
+      "learning_rate": 4.725049133672631e-05,
+      "loss": 2.718,
+      "step": 6801
+    },
+    {
+      "epoch": 0.615732778129809,
+      "grad_norm": 0.9191269278526306,
+      "learning_rate": 4.7223534746376884e-05,
+      "loss": 2.6797,
+      "step": 6802
+    },
+    {
+      "epoch": 0.6158233004435594,
+      "grad_norm": 1.0587358474731445,
+      "learning_rate": 4.719658347061613e-05,
+      "loss": 2.7238,
+      "step": 6803
+    },
+    {
+      "epoch": 0.6159138227573097,
+      "grad_norm": 0.9422360062599182,
+      "learning_rate": 4.716963751215794e-05,
+      "loss": 2.6894,
+      "step": 6804
+    },
+    {
+      "epoch": 0.61600434507106,
+      "grad_norm": 0.9692339301109314,
+      "learning_rate": 4.714269687371581e-05,
+      "loss": 2.6718,
+      "step": 6805
+    },
+    {
+      "epoch": 0.6160948673848103,
+      "grad_norm": 0.8721240162849426,
+      "learning_rate": 4.7115761558002646e-05,
+      "loss": 2.714,
+      "step": 6806
+    },
+    {
+      "epoch": 0.6161853896985607,
+      "grad_norm": 0.9523304104804993,
+      "learning_rate": 4.7088831567730894e-05,
+      "loss": 2.6986,
+      "step": 6807
+    },
+    {
+      "epoch": 0.616275912012311,
+      "grad_norm": 0.8966914415359497,
+      "learning_rate": 4.706190690561228e-05,
+      "loss": 2.671,
+      "step": 6808
+    },
+    {
+      "epoch": 0.6163664343260614,
+      "grad_norm": 0.8884119391441345,
+      "learning_rate": 4.7034987574358266e-05,
+      "loss": 2.6862,
+      "step": 6809
+    },
+    {
+      "epoch": 0.6164569566398117,
+      "grad_norm": 0.9186428785324097,
+      "learning_rate": 4.700807357667952e-05,
+      "loss": 2.6717,
+      "step": 6810
+    },
+    {
+      "epoch": 0.6165474789535621,
+      "grad_norm": 0.8419829607009888,
+      "learning_rate": 4.698116491528631e-05,
+      "loss": 2.6886,
+      "step": 6811
+    },
+    {
+      "epoch": 0.6166380012673124,
+      "grad_norm": 0.9079512357711792,
+      "learning_rate": 4.695426159288835e-05,
+      "loss": 2.6787,
+      "step": 6812
+    },
+    {
+      "epoch": 0.6167285235810628,
+      "grad_norm": 0.8453792333602905,
+      "learning_rate": 4.692736361219478e-05,
+      "loss": 2.6798,
+      "step": 6813
+    },
+    {
+      "epoch": 0.6168190458948131,
+      "grad_norm": 0.9105737805366516,
+      "learning_rate": 4.690047097591427e-05,
+      "loss": 2.6833,
+      "step": 6814
+    },
+    {
+      "epoch": 0.6169095682085635,
+      "grad_norm": 0.9034150242805481,
+      "learning_rate": 4.68735836867549e-05,
+      "loss": 2.7098,
+      "step": 6815
+    },
+    {
+      "epoch": 0.6170000905223137,
+      "grad_norm": 0.8991881608963013,
+      "learning_rate": 4.684670174742412e-05,
+      "loss": 2.6973,
+      "step": 6816
+    },
+    {
+      "epoch": 0.6170906128360641,
+      "grad_norm": 0.9135326743125916,
+      "learning_rate": 4.6819825160629096e-05,
+      "loss": 2.7298,
+      "step": 6817
+    },
+    {
+      "epoch": 0.6171811351498144,
+      "grad_norm": 0.8568701148033142,
+      "learning_rate": 4.679295392907619e-05,
+      "loss": 2.6618,
+      "step": 6818
+    },
+    {
+      "epoch": 0.6172716574635647,
+      "grad_norm": 0.8539561629295349,
+      "learning_rate": 4.676608805547137e-05,
+      "loss": 2.7173,
+      "step": 6819
+    },
+    {
+      "epoch": 0.6173621797773151,
+      "grad_norm": 0.8731486797332764,
+      "learning_rate": 4.673922754252002e-05,
+      "loss": 2.6987,
+      "step": 6820
+    },
+    {
+      "epoch": 0.6174527020910654,
+      "grad_norm": 0.8331453800201416,
+      "learning_rate": 4.6712372392927e-05,
+      "loss": 2.672,
+      "step": 6821
+    },
+    {
+      "epoch": 0.6175432244048158,
+      "grad_norm": 0.8664054870605469,
+      "learning_rate": 4.6685522609396626e-05,
+      "loss": 2.681,
+      "step": 6822
+    },
+    {
+      "epoch": 0.6176337467185661,
+      "grad_norm": 0.7990508079528809,
+      "learning_rate": 4.66586781946327e-05,
+      "loss": 2.6322,
+      "step": 6823
+    },
+    {
+      "epoch": 0.6177242690323165,
+      "grad_norm": 0.8614292740821838,
+      "learning_rate": 4.663183915133834e-05,
+      "loss": 2.6739,
+      "step": 6824
+    },
+    {
+      "epoch": 0.6178147913460668,
+      "grad_norm": 0.8222824931144714,
+      "learning_rate": 4.6605005482216405e-05,
+      "loss": 2.7179,
+      "step": 6825
+    },
+    {
+      "epoch": 0.6179053136598172,
+      "grad_norm": 0.8004710078239441,
+      "learning_rate": 4.657817718996891e-05,
+      "loss": 2.665,
+      "step": 6826
+    },
+    {
+      "epoch": 0.6179958359735674,
+      "grad_norm": 0.9650024771690369,
+      "learning_rate": 4.655135427729752e-05,
+      "loss": 2.6428,
+      "step": 6827
+    },
+    {
+      "epoch": 0.6180863582873178,
+      "grad_norm": 0.8087639212608337,
+      "learning_rate": 4.652453674690328e-05,
+      "loss": 2.6029,
+      "step": 6828
+    },
+    {
+      "epoch": 0.6181768806010681,
+      "grad_norm": 0.8862298130989075,
+      "learning_rate": 4.6497724601486735e-05,
+      "loss": 2.6962,
+      "step": 6829
+    },
+    {
+      "epoch": 0.6182674029148185,
+      "grad_norm": 0.8443248271942139,
+      "learning_rate": 4.647091784374785e-05,
+      "loss": 2.6529,
+      "step": 6830
+    },
+    {
+      "epoch": 0.6183579252285688,
+      "grad_norm": 0.7951116561889648,
+      "learning_rate": 4.6444116476386124e-05,
+      "loss": 2.655,
+      "step": 6831
+    },
+    {
+      "epoch": 0.6184484475423192,
+      "grad_norm": 0.9196255207061768,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 2.6802,
+      "step": 6832
+    },
+    {
+      "epoch": 0.6185389698560695,
+      "grad_norm": 0.9127574563026428,
+      "learning_rate": 4.6390529923588946e-05,
+      "loss": 2.6822,
+      "step": 6833
+    },
+    {
+      "epoch": 0.6186294921698199,
+      "grad_norm": 0.8869885802268982,
+      "learning_rate": 4.636374474354971e-05,
+      "loss": 2.73,
+      "step": 6834
+    },
+    {
+      "epoch": 0.6187200144835702,
+      "grad_norm": 0.8392091989517212,
+      "learning_rate": 4.633696496467991e-05,
+      "loss": 2.6734,
+      "step": 6835
+    },
+    {
+      "epoch": 0.6188105367973206,
+      "grad_norm": 0.8784205317497253,
+      "learning_rate": 4.631019058967627e-05,
+      "loss": 2.6801,
+      "step": 6836
+    },
+    {
+      "epoch": 0.6189010591110709,
+      "grad_norm": 0.9287988543510437,
+      "learning_rate": 4.628342162123501e-05,
+      "loss": 2.6924,
+      "step": 6837
+    },
+    {
+      "epoch": 0.6189915814248212,
+      "grad_norm": 0.8669235110282898,
+      "learning_rate": 4.625665806205164e-05,
+      "loss": 2.7482,
+      "step": 6838
+    },
+    {
+      "epoch": 0.6190821037385715,
+      "grad_norm": 0.89461350440979,
+      "learning_rate": 4.6229899914821414e-05,
+      "loss": 2.7027,
+      "step": 6839
+    },
+    {
+      "epoch": 0.6191726260523219,
+      "grad_norm": 0.9218267798423767,
+      "learning_rate": 4.620314718223876e-05,
+      "loss": 2.6773,
+      "step": 6840
+    },
+    {
+      "epoch": 0.6192631483660722,
+      "grad_norm": 0.8262689113616943,
+      "learning_rate": 4.617639986699771e-05,
+      "loss": 2.7483,
+      "step": 6841
+    },
+    {
+      "epoch": 0.6193536706798226,
+      "grad_norm": 0.8751800060272217,
+      "learning_rate": 4.614965797179172e-05,
+      "loss": 2.6881,
+      "step": 6842
+    },
+    {
+      "epoch": 0.6194441929935729,
+      "grad_norm": 0.9504587650299072,
+      "learning_rate": 4.612292149931369e-05,
+      "loss": 2.6459,
+      "step": 6843
+    },
+    {
+      "epoch": 0.6195347153073233,
+      "grad_norm": 0.8248963952064514,
+      "learning_rate": 4.609619045225604e-05,
+      "loss": 2.6531,
+      "step": 6844
+    },
+    {
+      "epoch": 0.6196252376210736,
+      "grad_norm": 0.851895272731781,
+      "learning_rate": 4.60694648333105e-05,
+      "loss": 2.6741,
+      "step": 6845
+    },
+    {
+      "epoch": 0.619715759934824,
+      "grad_norm": 0.8876405358314514,
+      "learning_rate": 4.6042744645168366e-05,
+      "loss": 2.7042,
+      "step": 6846
+    },
+    {
+      "epoch": 0.6198062822485743,
+      "grad_norm": 0.8736615180969238,
+      "learning_rate": 4.601602989052037e-05,
+      "loss": 2.7016,
+      "step": 6847
+    },
+    {
+      "epoch": 0.6198968045623247,
+      "grad_norm": 0.8399230241775513,
+      "learning_rate": 4.5989320572056696e-05,
+      "loss": 2.6293,
+      "step": 6848
+    },
+    {
+      "epoch": 0.6199873268760749,
+      "grad_norm": 0.9112479090690613,
+      "learning_rate": 4.5962616692466954e-05,
+      "loss": 2.6717,
+      "step": 6849
+    },
+    {
+      "epoch": 0.6200778491898253,
+      "grad_norm": 0.8878874778747559,
+      "learning_rate": 4.593591825444028e-05,
+      "loss": 2.6787,
+      "step": 6850
+    },
+    {
+      "epoch": 0.6201683715035756,
+      "grad_norm": 0.9500158429145813,
+      "learning_rate": 4.590922526066508e-05,
+      "loss": 2.7017,
+      "step": 6851
+    },
+    {
+      "epoch": 0.620258893817326,
+      "grad_norm": 1.0439828634262085,
+      "learning_rate": 4.588253771382949e-05,
+      "loss": 2.6546,
+      "step": 6852
+    },
+    {
+      "epoch": 0.6203494161310763,
+      "grad_norm": 0.9312981367111206,
+      "learning_rate": 4.585585561662082e-05,
+      "loss": 2.6252,
+      "step": 6853
+    },
+    {
+      "epoch": 0.6204399384448267,
+      "grad_norm": 0.8428036570549011,
+      "learning_rate": 4.582917897172603e-05,
+      "loss": 2.6421,
+      "step": 6854
+    },
+    {
+      "epoch": 0.620530460758577,
+      "grad_norm": 0.9673554301261902,
+      "learning_rate": 4.580250778183143e-05,
+      "loss": 2.6584,
+      "step": 6855
+    },
+    {
+      "epoch": 0.6206209830723274,
+      "grad_norm": 0.9068723320960999,
+      "learning_rate": 4.5775842049622806e-05,
+      "loss": 2.7353,
+      "step": 6856
+    },
+    {
+      "epoch": 0.6207115053860777,
+      "grad_norm": 0.8398150205612183,
+      "learning_rate": 4.574918177778541e-05,
+      "loss": 2.7008,
+      "step": 6857
+    },
+    {
+      "epoch": 0.6208020276998281,
+      "grad_norm": 0.8431996703147888,
+      "learning_rate": 4.5722526969003956e-05,
+      "loss": 2.6987,
+      "step": 6858
+    },
+    {
+      "epoch": 0.6208925500135783,
+      "grad_norm": 0.9009360074996948,
+      "learning_rate": 4.569587762596248e-05,
+      "loss": 2.6668,
+      "step": 6859
+    },
+    {
+      "epoch": 0.6209830723273286,
+      "grad_norm": 0.857540488243103,
+      "learning_rate": 4.566923375134472e-05,
+      "loss": 2.6374,
+      "step": 6860
+    },
+    {
+      "epoch": 0.621073594641079,
+      "grad_norm": 0.83136385679245,
+      "learning_rate": 4.5642595347833594e-05,
+      "loss": 2.652,
+      "step": 6861
+    },
+    {
+      "epoch": 0.6211641169548293,
+      "grad_norm": 0.8541494011878967,
+      "learning_rate": 4.5615962418111625e-05,
+      "loss": 2.6612,
+      "step": 6862
+    },
+    {
+      "epoch": 0.6212546392685797,
+      "grad_norm": 0.8624999523162842,
+      "learning_rate": 4.558933496486075e-05,
+      "loss": 2.6716,
+      "step": 6863
+    },
+    {
+      "epoch": 0.62134516158233,
+      "grad_norm": 0.9279386401176453,
+      "learning_rate": 4.5562712990762366e-05,
+      "loss": 2.7097,
+      "step": 6864
+    },
+    {
+      "epoch": 0.6214356838960804,
+      "grad_norm": 0.8684672713279724,
+      "learning_rate": 4.5536096498497295e-05,
+      "loss": 2.6706,
+      "step": 6865
+    },
+    {
+      "epoch": 0.6215262062098307,
+      "grad_norm": 0.8466289639472961,
+      "learning_rate": 4.550948549074584e-05,
+      "loss": 2.6747,
+      "step": 6866
+    },
+    {
+      "epoch": 0.6216167285235811,
+      "grad_norm": 0.7903226613998413,
+      "learning_rate": 4.548287997018764e-05,
+      "loss": 2.7077,
+      "step": 6867
+    },
+    {
+      "epoch": 0.6217072508373314,
+      "grad_norm": 0.86106938123703,
+      "learning_rate": 4.545627993950201e-05,
+      "loss": 2.6402,
+      "step": 6868
+    },
+    {
+      "epoch": 0.6217977731510818,
+      "grad_norm": 0.9263157248497009,
+      "learning_rate": 4.542968540136746e-05,
+      "loss": 2.7607,
+      "step": 6869
+    },
+    {
+      "epoch": 0.621888295464832,
+      "grad_norm": 0.9493169784545898,
+      "learning_rate": 4.5403096358462095e-05,
+      "loss": 2.7068,
+      "step": 6870
+    },
+    {
+      "epoch": 0.6219788177785824,
+      "grad_norm": 0.8483140468597412,
+      "learning_rate": 4.5376512813463434e-05,
+      "loss": 2.7293,
+      "step": 6871
+    },
+    {
+      "epoch": 0.6220693400923327,
+      "grad_norm": 0.8495675325393677,
+      "learning_rate": 4.534993476904848e-05,
+      "loss": 2.6592,
+      "step": 6872
+    },
+    {
+      "epoch": 0.6221598624060831,
+      "grad_norm": 0.8462209105491638,
+      "learning_rate": 4.532336222789352e-05,
+      "loss": 2.7176,
+      "step": 6873
+    },
+    {
+      "epoch": 0.6222503847198334,
+      "grad_norm": 0.9456688165664673,
+      "learning_rate": 4.529679519267456e-05,
+      "loss": 2.6491,
+      "step": 6874
+    },
+    {
+      "epoch": 0.6223409070335838,
+      "grad_norm": 0.843177080154419,
+      "learning_rate": 4.527023366606679e-05,
+      "loss": 2.6578,
+      "step": 6875
+    },
+    {
+      "epoch": 0.6224314293473341,
+      "grad_norm": 0.9310163855552673,
+      "learning_rate": 4.524367765074499e-05,
+      "loss": 2.7034,
+      "step": 6876
+    },
+    {
+      "epoch": 0.6225219516610845,
+      "grad_norm": 0.8866987228393555,
+      "learning_rate": 4.521712714938335e-05,
+      "loss": 2.6532,
+      "step": 6877
+    },
+    {
+      "epoch": 0.6226124739748348,
+      "grad_norm": 0.9527557492256165,
+      "learning_rate": 4.5190582164655525e-05,
+      "loss": 2.6735,
+      "step": 6878
+    },
+    {
+      "epoch": 0.6227029962885852,
+      "grad_norm": 0.9127796292304993,
+      "learning_rate": 4.5164042699234566e-05,
+      "loss": 2.6247,
+      "step": 6879
+    },
+    {
+      "epoch": 0.6227935186023354,
+      "grad_norm": 0.8668599724769592,
+      "learning_rate": 4.513750875579303e-05,
+      "loss": 2.6347,
+      "step": 6880
+    },
+    {
+      "epoch": 0.6228840409160858,
+      "grad_norm": 0.7990663647651672,
+      "learning_rate": 4.5110980337002806e-05,
+      "loss": 2.687,
+      "step": 6881
+    },
+    {
+      "epoch": 0.6229745632298361,
+      "grad_norm": 0.8905903697013855,
+      "learning_rate": 4.5084457445535435e-05,
+      "loss": 2.6849,
+      "step": 6882
+    },
+    {
+      "epoch": 0.6230650855435865,
+      "grad_norm": 0.9093679189682007,
+      "learning_rate": 4.505794008406165e-05,
+      "loss": 2.6596,
+      "step": 6883
+    },
+    {
+      "epoch": 0.6231556078573368,
+      "grad_norm": 0.8659344911575317,
+      "learning_rate": 4.5031428255251793e-05,
+      "loss": 2.7382,
+      "step": 6884
+    },
+    {
+      "epoch": 0.6232461301710872,
+      "grad_norm": 0.7988776564598083,
+      "learning_rate": 4.500492196177561e-05,
+      "loss": 2.6526,
+      "step": 6885
+    },
+    {
+      "epoch": 0.6233366524848375,
+      "grad_norm": 0.8617873787879944,
+      "learning_rate": 4.497842120630229e-05,
+      "loss": 2.7033,
+      "step": 6886
+    },
+    {
+      "epoch": 0.6234271747985879,
+      "grad_norm": 0.8665861487388611,
+      "learning_rate": 4.495192599150044e-05,
+      "loss": 2.6881,
+      "step": 6887
+    },
+    {
+      "epoch": 0.6235176971123382,
+      "grad_norm": 0.8218960762023926,
+      "learning_rate": 4.492543632003817e-05,
+      "loss": 2.6161,
+      "step": 6888
+    },
+    {
+      "epoch": 0.6236082194260886,
+      "grad_norm": 0.8584741950035095,
+      "learning_rate": 4.489895219458289e-05,
+      "loss": 2.7239,
+      "step": 6889
+    },
+    {
+      "epoch": 0.6236987417398389,
+      "grad_norm": 0.8423812389373779,
+      "learning_rate": 4.487247361780169e-05,
+      "loss": 2.679,
+      "step": 6890
+    },
+    {
+      "epoch": 0.6237892640535893,
+      "grad_norm": 0.8747096061706543,
+      "learning_rate": 4.484600059236085e-05,
+      "loss": 2.6855,
+      "step": 6891
+    },
+    {
+      "epoch": 0.6238797863673395,
+      "grad_norm": 0.8690003156661987,
+      "learning_rate": 4.4819533120926236e-05,
+      "loss": 2.6958,
+      "step": 6892
+    },
+    {
+      "epoch": 0.6239703086810899,
+      "grad_norm": 0.802766740322113,
+      "learning_rate": 4.479307120616313e-05,
+      "loss": 2.6509,
+      "step": 6893
+    },
+    {
+      "epoch": 0.6240608309948402,
+      "grad_norm": 0.8117000460624695,
+      "learning_rate": 4.476661485073624e-05,
+      "loss": 2.7068,
+      "step": 6894
+    },
+    {
+      "epoch": 0.6241513533085906,
+      "grad_norm": 0.8532359600067139,
+      "learning_rate": 4.474016405730973e-05,
+      "loss": 2.68,
+      "step": 6895
+    },
+    {
+      "epoch": 0.6242418756223409,
+      "grad_norm": 0.8388345241546631,
+      "learning_rate": 4.471371882854723e-05,
+      "loss": 2.7115,
+      "step": 6896
+    },
+    {
+      "epoch": 0.6243323979360913,
+      "grad_norm": 0.9178318381309509,
+      "learning_rate": 4.468727916711164e-05,
+      "loss": 2.6154,
+      "step": 6897
+    },
+    {
+      "epoch": 0.6244229202498416,
+      "grad_norm": 0.8816856145858765,
+      "learning_rate": 4.46608450756656e-05,
+      "loss": 2.7028,
+      "step": 6898
+    },
+    {
+      "epoch": 0.624513442563592,
+      "grad_norm": 0.8467722535133362,
+      "learning_rate": 4.463441655687092e-05,
+      "loss": 2.6352,
+      "step": 6899
+    },
+    {
+      "epoch": 0.6246039648773423,
+      "grad_norm": 0.8618403077125549,
+      "learning_rate": 4.4607993613388976e-05,
+      "loss": 2.695,
+      "step": 6900
+    },
+    {
+      "epoch": 0.6246944871910926,
+      "grad_norm": 0.9370056986808777,
+      "learning_rate": 4.458157624788056e-05,
+      "loss": 2.6996,
+      "step": 6901
+    },
+    {
+      "epoch": 0.624785009504843,
+      "grad_norm": 0.9463796019554138,
+      "learning_rate": 4.455516446300593e-05,
+      "loss": 2.6751,
+      "step": 6902
+    },
+    {
+      "epoch": 0.6248755318185932,
+      "grad_norm": 0.9075630307197571,
+      "learning_rate": 4.4528758261424655e-05,
+      "loss": 2.6395,
+      "step": 6903
+    },
+    {
+      "epoch": 0.6249660541323436,
+      "grad_norm": 0.9862688183784485,
+      "learning_rate": 4.4502357645795976e-05,
+      "loss": 2.6984,
+      "step": 6904
+    },
+    {
+      "epoch": 0.6250565764460939,
+      "grad_norm": 0.8373948931694031,
+      "learning_rate": 4.447596261877832e-05,
+      "loss": 2.6353,
+      "step": 6905
+    },
+    {
+      "epoch": 0.6251470987598443,
+      "grad_norm": 0.86262047290802,
+      "learning_rate": 4.444957318302973e-05,
+      "loss": 2.6511,
+      "step": 6906
+    },
+    {
+      "epoch": 0.6252376210735946,
+      "grad_norm": 0.8128232359886169,
+      "learning_rate": 4.442318934120758e-05,
+      "loss": 2.6396,
+      "step": 6907
+    },
+    {
+      "epoch": 0.625328143387345,
+      "grad_norm": 0.9029171466827393,
+      "learning_rate": 4.439681109596875e-05,
+      "loss": 2.6639,
+      "step": 6908
+    },
+    {
+      "epoch": 0.6254186657010953,
+      "grad_norm": 0.877526581287384,
+      "learning_rate": 4.437043844996952e-05,
+      "loss": 2.6135,
+      "step": 6909
+    },
+    {
+      "epoch": 0.6255091880148457,
+      "grad_norm": 0.8649809956550598,
+      "learning_rate": 4.434407140586565e-05,
+      "loss": 2.7198,
+      "step": 6910
+    },
+    {
+      "epoch": 0.625599710328596,
+      "grad_norm": 0.8621827363967896,
+      "learning_rate": 4.431770996631219e-05,
+      "loss": 2.6379,
+      "step": 6911
+    },
+    {
+      "epoch": 0.6256902326423464,
+      "grad_norm": 0.8444892764091492,
+      "learning_rate": 4.429135413396389e-05,
+      "loss": 2.71,
+      "step": 6912
+    },
+    {
+      "epoch": 0.6257807549560966,
+      "grad_norm": 0.8484147191047668,
+      "learning_rate": 4.426500391147468e-05,
+      "loss": 2.7082,
+      "step": 6913
+    },
+    {
+      "epoch": 0.625871277269847,
+      "grad_norm": 0.8146012425422668,
+      "learning_rate": 4.4238659301498045e-05,
+      "loss": 2.7241,
+      "step": 6914
+    },
+    {
+      "epoch": 0.6259617995835973,
+      "grad_norm": 0.866885244846344,
+      "learning_rate": 4.4212320306686884e-05,
+      "loss": 2.6802,
+      "step": 6915
+    },
+    {
+      "epoch": 0.6260523218973477,
+      "grad_norm": 0.8107602000236511,
+      "learning_rate": 4.4185986929693546e-05,
+      "loss": 2.7391,
+      "step": 6916
+    },
+    {
+      "epoch": 0.626142844211098,
+      "grad_norm": 0.9229167103767395,
+      "learning_rate": 4.41596591731698e-05,
+      "loss": 2.6994,
+      "step": 6917
+    },
+    {
+      "epoch": 0.6262333665248484,
+      "grad_norm": 0.9321455359458923,
+      "learning_rate": 4.4133337039766877e-05,
+      "loss": 2.7259,
+      "step": 6918
+    },
+    {
+      "epoch": 0.6263238888385987,
+      "grad_norm": 0.9236816167831421,
+      "learning_rate": 4.4107020532135314e-05,
+      "loss": 2.6672,
+      "step": 6919
+    },
+    {
+      "epoch": 0.6264144111523491,
+      "grad_norm": 0.831727147102356,
+      "learning_rate": 4.4080709652925336e-05,
+      "loss": 2.6458,
+      "step": 6920
+    },
+    {
+      "epoch": 0.6265049334660994,
+      "grad_norm": 0.8836252093315125,
+      "learning_rate": 4.405440440478632e-05,
+      "loss": 2.6761,
+      "step": 6921
+    },
+    {
+      "epoch": 0.6265954557798498,
+      "grad_norm": 0.8219419121742249,
+      "learning_rate": 4.402810479036725e-05,
+      "loss": 2.6527,
+      "step": 6922
+    },
+    {
+      "epoch": 0.6266859780936,
+      "grad_norm": 0.8386253118515015,
+      "learning_rate": 4.40018108123165e-05,
+      "loss": 2.7052,
+      "step": 6923
+    },
+    {
+      "epoch": 0.6267765004073504,
+      "grad_norm": 0.8826115727424622,
+      "learning_rate": 4.3975522473281875e-05,
+      "loss": 2.6718,
+      "step": 6924
+    },
+    {
+      "epoch": 0.6268670227211007,
+      "grad_norm": 0.8514438271522522,
+      "learning_rate": 4.394923977591059e-05,
+      "loss": 2.6277,
+      "step": 6925
+    },
+    {
+      "epoch": 0.6269575450348511,
+      "grad_norm": 0.8475186824798584,
+      "learning_rate": 4.392296272284937e-05,
+      "loss": 2.692,
+      "step": 6926
+    },
+    {
+      "epoch": 0.6270480673486014,
+      "grad_norm": 0.891557514667511,
+      "learning_rate": 4.389669131674419e-05,
+      "loss": 2.7062,
+      "step": 6927
+    },
+    {
+      "epoch": 0.6271385896623518,
+      "grad_norm": 0.832285463809967,
+      "learning_rate": 4.387042556024074e-05,
+      "loss": 2.6905,
+      "step": 6928
+    },
+    {
+      "epoch": 0.6272291119761021,
+      "grad_norm": 0.9494576454162598,
+      "learning_rate": 4.38441654559839e-05,
+      "loss": 2.6781,
+      "step": 6929
+    },
+    {
+      "epoch": 0.6273196342898525,
+      "grad_norm": 0.8865106701850891,
+      "learning_rate": 4.3817911006617986e-05,
+      "loss": 2.7055,
+      "step": 6930
+    },
+    {
+      "epoch": 0.6274101566036028,
+      "grad_norm": 0.8264738321304321,
+      "learning_rate": 4.379166221478697e-05,
+      "loss": 2.6233,
+      "step": 6931
+    },
+    {
+      "epoch": 0.6275006789173532,
+      "grad_norm": 0.8542665243148804,
+      "learning_rate": 4.3765419083134e-05,
+      "loss": 2.6594,
+      "step": 6932
+    },
+    {
+      "epoch": 0.6275912012311035,
+      "grad_norm": 0.8898385763168335,
+      "learning_rate": 4.3739181614301786e-05,
+      "loss": 2.6302,
+      "step": 6933
+    },
+    {
+      "epoch": 0.6276817235448539,
+      "grad_norm": 0.7890973687171936,
+      "learning_rate": 4.371294981093244e-05,
+      "loss": 2.6175,
+      "step": 6934
+    },
+    {
+      "epoch": 0.6277722458586041,
+      "grad_norm": 0.8115043640136719,
+      "learning_rate": 4.368672367566751e-05,
+      "loss": 2.6913,
+      "step": 6935
+    },
+    {
+      "epoch": 0.6278627681723545,
+      "grad_norm": 0.8468533158302307,
+      "learning_rate": 4.366050321114796e-05,
+      "loss": 2.6607,
+      "step": 6936
+    },
+    {
+      "epoch": 0.6279532904861048,
+      "grad_norm": 0.8599989414215088,
+      "learning_rate": 4.363428842001425e-05,
+      "loss": 2.6724,
+      "step": 6937
+    },
+    {
+      "epoch": 0.6280438127998552,
+      "grad_norm": 0.8746944069862366,
+      "learning_rate": 4.360807930490606e-05,
+      "loss": 2.7038,
+      "step": 6938
+    },
+    {
+      "epoch": 0.6281343351136055,
+      "grad_norm": 0.8745467066764832,
+      "learning_rate": 4.3581875868462815e-05,
+      "loss": 2.7207,
+      "step": 6939
+    },
+    {
+      "epoch": 0.6282248574273559,
+      "grad_norm": 0.8731328845024109,
+      "learning_rate": 4.355567811332311e-05,
+      "loss": 2.6887,
+      "step": 6940
+    },
+    {
+      "epoch": 0.6283153797411062,
+      "grad_norm": 0.8534673452377319,
+      "learning_rate": 4.3529486042125055e-05,
+      "loss": 2.6319,
+      "step": 6941
+    },
+    {
+      "epoch": 0.6284059020548565,
+      "grad_norm": 0.8535872101783752,
+      "learning_rate": 4.350329965750621e-05,
+      "loss": 2.6605,
+      "step": 6942
+    },
+    {
+      "epoch": 0.6284964243686069,
+      "grad_norm": 0.8605890870094299,
+      "learning_rate": 4.347711896210356e-05,
+      "loss": 2.7461,
+      "step": 6943
+    },
+    {
+      "epoch": 0.6285869466823572,
+      "grad_norm": 0.7937997579574585,
+      "learning_rate": 4.3450943958553466e-05,
+      "loss": 2.6408,
+      "step": 6944
+    },
+    {
+      "epoch": 0.6286774689961075,
+      "grad_norm": 0.8514191508293152,
+      "learning_rate": 4.342477464949182e-05,
+      "loss": 2.7152,
+      "step": 6945
+    },
+    {
+      "epoch": 0.6287679913098578,
+      "grad_norm": 0.8694384694099426,
+      "learning_rate": 4.339861103755374e-05,
+      "loss": 2.6926,
+      "step": 6946
+    },
+    {
+      "epoch": 0.6288585136236082,
+      "grad_norm": 0.9281476140022278,
+      "learning_rate": 4.3372453125374046e-05,
+      "loss": 2.6812,
+      "step": 6947
+    },
+    {
+      "epoch": 0.6289490359373585,
+      "grad_norm": 0.8273919224739075,
+      "learning_rate": 4.334630091558675e-05,
+      "loss": 2.6989,
+      "step": 6948
+    },
+    {
+      "epoch": 0.6290395582511089,
+      "grad_norm": 0.9093640446662903,
+      "learning_rate": 4.3320154410825386e-05,
+      "loss": 2.707,
+      "step": 6949
+    },
+    {
+      "epoch": 0.6291300805648592,
+      "grad_norm": 0.8211041688919067,
+      "learning_rate": 4.329401361372294e-05,
+      "loss": 2.661,
+      "step": 6950
+    },
+    {
+      "epoch": 0.6292206028786096,
+      "grad_norm": 0.860684871673584,
+      "learning_rate": 4.326787852691175e-05,
+      "loss": 2.7215,
+      "step": 6951
+    },
+    {
+      "epoch": 0.6293111251923599,
+      "grad_norm": 0.9397785663604736,
+      "learning_rate": 4.324174915302366e-05,
+      "loss": 2.7168,
+      "step": 6952
+    },
+    {
+      "epoch": 0.6294016475061103,
+      "grad_norm": 0.8350105285644531,
+      "learning_rate": 4.32156254946899e-05,
+      "loss": 2.6404,
+      "step": 6953
+    },
+    {
+      "epoch": 0.6294921698198606,
+      "grad_norm": 0.8832926750183105,
+      "learning_rate": 4.318950755454103e-05,
+      "loss": 2.6015,
+      "step": 6954
+    },
+    {
+      "epoch": 0.629582692133611,
+      "grad_norm": 0.8618656992912292,
+      "learning_rate": 4.3163395335207265e-05,
+      "loss": 2.6702,
+      "step": 6955
+    },
+    {
+      "epoch": 0.6296732144473612,
+      "grad_norm": 0.8379136323928833,
+      "learning_rate": 4.3137288839318014e-05,
+      "loss": 2.6714,
+      "step": 6956
+    },
+    {
+      "epoch": 0.6297637367611116,
+      "grad_norm": 0.8382007479667664,
+      "learning_rate": 4.3111188069502206e-05,
+      "loss": 2.6437,
+      "step": 6957
+    },
+    {
+      "epoch": 0.6298542590748619,
+      "grad_norm": 0.8624709844589233,
+      "learning_rate": 4.30850930283882e-05,
+      "loss": 2.7487,
+      "step": 6958
+    },
+    {
+      "epoch": 0.6299447813886123,
+      "grad_norm": 0.8321948647499084,
+      "learning_rate": 4.3059003718603776e-05,
+      "loss": 2.6971,
+      "step": 6959
+    },
+    {
+      "epoch": 0.6300353037023626,
+      "grad_norm": 0.8173766732215881,
+      "learning_rate": 4.3032920142776125e-05,
+      "loss": 2.6913,
+      "step": 6960
+    },
+    {
+      "epoch": 0.630125826016113,
+      "grad_norm": 0.8754605650901794,
+      "learning_rate": 4.300684230353188e-05,
+      "loss": 2.6797,
+      "step": 6961
+    },
+    {
+      "epoch": 0.6302163483298633,
+      "grad_norm": 0.8446570634841919,
+      "learning_rate": 4.2980770203497035e-05,
+      "loss": 2.6248,
+      "step": 6962
+    },
+    {
+      "epoch": 0.6303068706436137,
+      "grad_norm": 0.8704685568809509,
+      "learning_rate": 4.295470384529705e-05,
+      "loss": 2.6948,
+      "step": 6963
+    },
+    {
+      "epoch": 0.630397392957364,
+      "grad_norm": 0.8113224506378174,
+      "learning_rate": 4.2928643231556844e-05,
+      "loss": 2.6748,
+      "step": 6964
+    },
+    {
+      "epoch": 0.6304879152711144,
+      "grad_norm": 0.8907050490379333,
+      "learning_rate": 4.2902588364900695e-05,
+      "loss": 2.6532,
+      "step": 6965
+    },
+    {
+      "epoch": 0.6305784375848646,
+      "grad_norm": 0.8262181878089905,
+      "learning_rate": 4.287653924795234e-05,
+      "loss": 2.658,
+      "step": 6966
+    },
+    {
+      "epoch": 0.630668959898615,
+      "grad_norm": 0.8925417065620422,
+      "learning_rate": 4.285049588333495e-05,
+      "loss": 2.7295,
+      "step": 6967
+    },
+    {
+      "epoch": 0.6307594822123653,
+      "grad_norm": 0.8073951005935669,
+      "learning_rate": 4.2824458273671e-05,
+      "loss": 2.6385,
+      "step": 6968
+    },
+    {
+      "epoch": 0.6308500045261157,
+      "grad_norm": 0.9392690658569336,
+      "learning_rate": 4.2798426421582594e-05,
+      "loss": 2.6744,
+      "step": 6969
+    },
+    {
+      "epoch": 0.630940526839866,
+      "grad_norm": 0.8945513963699341,
+      "learning_rate": 4.277240032969105e-05,
+      "loss": 2.6396,
+      "step": 6970
+    },
+    {
+      "epoch": 0.6310310491536164,
+      "grad_norm": 0.8523789644241333,
+      "learning_rate": 4.2746380000617236e-05,
+      "loss": 2.6597,
+      "step": 6971
+    },
+    {
+      "epoch": 0.6311215714673667,
+      "grad_norm": 0.8517530560493469,
+      "learning_rate": 4.272036543698138e-05,
+      "loss": 2.6104,
+      "step": 6972
+    },
+    {
+      "epoch": 0.6312120937811171,
+      "grad_norm": 0.8112363815307617,
+      "learning_rate": 4.2694356641403165e-05,
+      "loss": 2.6842,
+      "step": 6973
+    },
+    {
+      "epoch": 0.6313026160948674,
+      "grad_norm": 0.8543645739555359,
+      "learning_rate": 4.2668353616501666e-05,
+      "loss": 2.6565,
+      "step": 6974
+    },
+    {
+      "epoch": 0.6313931384086178,
+      "grad_norm": 0.8598533272743225,
+      "learning_rate": 4.264235636489542e-05,
+      "loss": 2.6996,
+      "step": 6975
+    },
+    {
+      "epoch": 0.6314836607223681,
+      "grad_norm": 0.8872503042221069,
+      "learning_rate": 4.2616364889202254e-05,
+      "loss": 2.6053,
+      "step": 6976
+    },
+    {
+      "epoch": 0.6315741830361185,
+      "grad_norm": 0.8489211797714233,
+      "learning_rate": 4.259037919203964e-05,
+      "loss": 2.6973,
+      "step": 6977
+    },
+    {
+      "epoch": 0.6316647053498687,
+      "grad_norm": 0.8446736335754395,
+      "learning_rate": 4.256439927602424e-05,
+      "loss": 2.699,
+      "step": 6978
+    },
+    {
+      "epoch": 0.6317552276636191,
+      "grad_norm": 0.9012982845306396,
+      "learning_rate": 4.2538425143772256e-05,
+      "loss": 2.6917,
+      "step": 6979
+    },
+    {
+      "epoch": 0.6318457499773694,
+      "grad_norm": 0.8830162882804871,
+      "learning_rate": 4.251245679789928e-05,
+      "loss": 2.681,
+      "step": 6980
+    },
+    {
+      "epoch": 0.6319362722911198,
+      "grad_norm": 0.8840652108192444,
+      "learning_rate": 4.248649424102035e-05,
+      "loss": 2.7276,
+      "step": 6981
+    },
+    {
+      "epoch": 0.6320267946048701,
+      "grad_norm": 0.9230279922485352,
+      "learning_rate": 4.2460537475749864e-05,
+      "loss": 2.6295,
+      "step": 6982
+    },
+    {
+      "epoch": 0.6321173169186204,
+      "grad_norm": 0.8687030673027039,
+      "learning_rate": 4.243458650470171e-05,
+      "loss": 2.6694,
+      "step": 6983
+    },
+    {
+      "epoch": 0.6322078392323708,
+      "grad_norm": 0.8421883583068848,
+      "learning_rate": 4.2408641330489066e-05,
+      "loss": 2.6714,
+      "step": 6984
+    },
+    {
+      "epoch": 0.6322983615461211,
+      "grad_norm": 0.8303890824317932,
+      "learning_rate": 4.238270195572472e-05,
+      "loss": 2.6576,
+      "step": 6985
+    },
+    {
+      "epoch": 0.6323888838598715,
+      "grad_norm": 0.8518853783607483,
+      "learning_rate": 4.235676838302068e-05,
+      "loss": 2.666,
+      "step": 6986
+    },
+    {
+      "epoch": 0.6324794061736217,
+      "grad_norm": 0.9766391515731812,
+      "learning_rate": 4.2330840614988496e-05,
+      "loss": 2.7746,
+      "step": 6987
+    },
+    {
+      "epoch": 0.6325699284873721,
+      "grad_norm": 0.8131774663925171,
+      "learning_rate": 4.230491865423908e-05,
+      "loss": 2.7201,
+      "step": 6988
+    },
+    {
+      "epoch": 0.6326604508011224,
+      "grad_norm": 0.8058497905731201,
+      "learning_rate": 4.227900250338277e-05,
+      "loss": 2.6695,
+      "step": 6989
+    },
+    {
+      "epoch": 0.6327509731148728,
+      "grad_norm": 0.8577495217323303,
+      "learning_rate": 4.225309216502933e-05,
+      "loss": 2.6565,
+      "step": 6990
+    },
+    {
+      "epoch": 0.6328414954286231,
+      "grad_norm": 0.8953158259391785,
+      "learning_rate": 4.2227187641787966e-05,
+      "loss": 2.6916,
+      "step": 6991
+    },
+    {
+      "epoch": 0.6329320177423735,
+      "grad_norm": 0.918773889541626,
+      "learning_rate": 4.220128893626715e-05,
+      "loss": 2.7069,
+      "step": 6992
+    },
+    {
+      "epoch": 0.6330225400561238,
+      "grad_norm": 0.8627886772155762,
+      "learning_rate": 4.2175396051075035e-05,
+      "loss": 2.6335,
+      "step": 6993
+    },
+    {
+      "epoch": 0.6331130623698742,
+      "grad_norm": 0.8618264198303223,
+      "learning_rate": 4.214950898881892e-05,
+      "loss": 2.6784,
+      "step": 6994
+    },
+    {
+      "epoch": 0.6332035846836245,
+      "grad_norm": 0.8480239510536194,
+      "learning_rate": 4.2123627752105665e-05,
+      "loss": 2.6699,
+      "step": 6995
+    },
+    {
+      "epoch": 0.6332941069973749,
+      "grad_norm": 0.9663839340209961,
+      "learning_rate": 4.209775234354151e-05,
+      "loss": 2.6057,
+      "step": 6996
+    },
+    {
+      "epoch": 0.6333846293111252,
+      "grad_norm": 0.8858630657196045,
+      "learning_rate": 4.207188276573214e-05,
+      "loss": 2.6634,
+      "step": 6997
+    },
+    {
+      "epoch": 0.6334751516248756,
+      "grad_norm": 0.9681085348129272,
+      "learning_rate": 4.204601902128253e-05,
+      "loss": 2.7401,
+      "step": 6998
+    },
+    {
+      "epoch": 0.6335656739386258,
+      "grad_norm": 0.853465735912323,
+      "learning_rate": 4.2020161112797285e-05,
+      "loss": 2.6457,
+      "step": 6999
+    },
+    {
+      "epoch": 0.6336561962523762,
+      "grad_norm": 0.9195594191551208,
+      "learning_rate": 4.19943090428802e-05,
+      "loss": 2.6566,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6336561962523762,
+      "eval_loss": 2.6202468872070312,
+      "eval_runtime": 71.5629,
+      "eval_samples_per_second": 37.771,
+      "eval_steps_per_second": 3.158,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6337467185661265,
+      "grad_norm": 0.8611754775047302,
+      "learning_rate": 4.19684628141346e-05,
+      "loss": 2.6688,
+      "step": 7001
+    },
+    {
+      "epoch": 0.6338372408798769,
+      "grad_norm": 0.8615222573280334,
+      "learning_rate": 4.19426224291632e-05,
+      "loss": 2.6729,
+      "step": 7002
+    },
+    {
+      "epoch": 0.6339277631936272,
+      "grad_norm": 0.8649547696113586,
+      "learning_rate": 4.1916787890568146e-05,
+      "loss": 2.6855,
+      "step": 7003
+    },
+    {
+      "epoch": 0.6340182855073776,
+      "grad_norm": 0.874506413936615,
+      "learning_rate": 4.189095920095096e-05,
+      "loss": 2.6496,
+      "step": 7004
+    },
+    {
+      "epoch": 0.6341088078211279,
+      "grad_norm": 0.811741054058075,
+      "learning_rate": 4.186513636291263e-05,
+      "loss": 2.6205,
+      "step": 7005
+    },
+    {
+      "epoch": 0.6341993301348783,
+      "grad_norm": 0.8108596801757812,
+      "learning_rate": 4.18393193790534e-05,
+      "loss": 2.6642,
+      "step": 7006
+    },
+    {
+      "epoch": 0.6342898524486286,
+      "grad_norm": 0.8476105332374573,
+      "learning_rate": 4.181350825197319e-05,
+      "loss": 2.6937,
+      "step": 7007
+    },
+    {
+      "epoch": 0.634380374762379,
+      "grad_norm": 0.9545331001281738,
+      "learning_rate": 4.1787702984271074e-05,
+      "loss": 2.6579,
+      "step": 7008
+    },
+    {
+      "epoch": 0.6344708970761292,
+      "grad_norm": 0.8565835356712341,
+      "learning_rate": 4.176190357854567e-05,
+      "loss": 2.6092,
+      "step": 7009
+    },
+    {
+      "epoch": 0.6345614193898796,
+      "grad_norm": 0.8987926840782166,
+      "learning_rate": 4.173611003739498e-05,
+      "loss": 2.7727,
+      "step": 7010
+    },
+    {
+      "epoch": 0.6346519417036299,
+      "grad_norm": 1.0171281099319458,
+      "learning_rate": 4.171032236341642e-05,
+      "loss": 2.6363,
+      "step": 7011
+    },
+    {
+      "epoch": 0.6347424640173803,
+      "grad_norm": 0.918834388256073,
+      "learning_rate": 4.168454055920681e-05,
+      "loss": 2.6419,
+      "step": 7012
+    },
+    {
+      "epoch": 0.6348329863311306,
+      "grad_norm": 0.8396235108375549,
+      "learning_rate": 4.165876462736239e-05,
+      "loss": 2.6763,
+      "step": 7013
+    },
+    {
+      "epoch": 0.634923508644881,
+      "grad_norm": 0.8260104656219482,
+      "learning_rate": 4.1632994570478746e-05,
+      "loss": 2.6594,
+      "step": 7014
+    },
+    {
+      "epoch": 0.6350140309586313,
+      "grad_norm": 0.9070295691490173,
+      "learning_rate": 4.160723039115095e-05,
+      "loss": 2.6776,
+      "step": 7015
+    },
+    {
+      "epoch": 0.6351045532723817,
+      "grad_norm": 0.8622607588768005,
+      "learning_rate": 4.158147209197347e-05,
+      "loss": 2.6578,
+      "step": 7016
+    },
+    {
+      "epoch": 0.635195075586132,
+      "grad_norm": 0.8886837959289551,
+      "learning_rate": 4.155571967554014e-05,
+      "loss": 2.6415,
+      "step": 7017
+    },
+    {
+      "epoch": 0.6352855978998824,
+      "grad_norm": 0.9302876591682434,
+      "learning_rate": 4.152997314444428e-05,
+      "loss": 2.7068,
+      "step": 7018
+    },
+    {
+      "epoch": 0.6353761202136327,
+      "grad_norm": 0.8932064175605774,
+      "learning_rate": 4.150423250127845e-05,
+      "loss": 2.6586,
+      "step": 7019
+    },
+    {
+      "epoch": 0.635466642527383,
+      "grad_norm": 0.842445433139801,
+      "learning_rate": 4.147849774863488e-05,
+      "loss": 2.6829,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6355571648411333,
+      "grad_norm": 0.8959222435951233,
+      "learning_rate": 4.145276888910494e-05,
+      "loss": 2.6554,
+      "step": 7021
+    },
+    {
+      "epoch": 0.6356476871548837,
+      "grad_norm": 0.8270236253738403,
+      "learning_rate": 4.1427045925279586e-05,
+      "loss": 2.6167,
+      "step": 7022
+    },
+    {
+      "epoch": 0.635738209468634,
+      "grad_norm": 0.8666673302650452,
+      "learning_rate": 4.14013288597491e-05,
+      "loss": 2.7031,
+      "step": 7023
+    },
+    {
+      "epoch": 0.6358287317823843,
+      "grad_norm": 0.8924165368080139,
+      "learning_rate": 4.137561769510322e-05,
+      "loss": 2.651,
+      "step": 7024
+    },
+    {
+      "epoch": 0.6359192540961347,
+      "grad_norm": 0.8522528409957886,
+      "learning_rate": 4.134991243393097e-05,
+      "loss": 2.686,
+      "step": 7025
+    },
+    {
+      "epoch": 0.636009776409885,
+      "grad_norm": 0.8556380271911621,
+      "learning_rate": 4.1324213078821005e-05,
+      "loss": 2.6777,
+      "step": 7026
+    },
+    {
+      "epoch": 0.6361002987236354,
+      "grad_norm": 0.8803753852844238,
+      "learning_rate": 4.1298519632361144e-05,
+      "loss": 2.6631,
+      "step": 7027
+    },
+    {
+      "epoch": 0.6361908210373857,
+      "grad_norm": 0.8770917654037476,
+      "learning_rate": 4.1272832097138744e-05,
+      "loss": 2.6452,
+      "step": 7028
+    },
+    {
+      "epoch": 0.6362813433511361,
+      "grad_norm": 0.8415400385856628,
+      "learning_rate": 4.124715047574055e-05,
+      "loss": 2.6616,
+      "step": 7029
+    },
+    {
+      "epoch": 0.6363718656648863,
+      "grad_norm": 0.8721527457237244,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 2.6991,
+      "step": 7030
+    },
+    {
+      "epoch": 0.6364623879786367,
+      "grad_norm": 0.9462018609046936,
+      "learning_rate": 4.119580498476072e-05,
+      "loss": 2.6648,
+      "step": 7031
+    },
+    {
+      "epoch": 0.636552910292387,
+      "grad_norm": 0.8758288621902466,
+      "learning_rate": 4.117014112034961e-05,
+      "loss": 2.6952,
+      "step": 7032
+    },
+    {
+      "epoch": 0.6366434326061374,
+      "grad_norm": 0.8675450086593628,
+      "learning_rate": 4.114448318010361e-05,
+      "loss": 2.6529,
+      "step": 7033
+    },
+    {
+      "epoch": 0.6367339549198877,
+      "grad_norm": 0.8605788946151733,
+      "learning_rate": 4.111883116660662e-05,
+      "loss": 2.6896,
+      "step": 7034
+    },
+    {
+      "epoch": 0.6368244772336381,
+      "grad_norm": 0.8481203317642212,
+      "learning_rate": 4.1093185082441677e-05,
+      "loss": 2.6641,
+      "step": 7035
+    },
+    {
+      "epoch": 0.6369149995473884,
+      "grad_norm": 0.8389714360237122,
+      "learning_rate": 4.106754493019138e-05,
+      "loss": 2.7555,
+      "step": 7036
+    },
+    {
+      "epoch": 0.6370055218611388,
+      "grad_norm": 0.857469916343689,
+      "learning_rate": 4.10419107124377e-05,
+      "loss": 2.6379,
+      "step": 7037
+    },
+    {
+      "epoch": 0.6370960441748891,
+      "grad_norm": 0.8377165794372559,
+      "learning_rate": 4.101628243176201e-05,
+      "loss": 2.656,
+      "step": 7038
+    },
+    {
+      "epoch": 0.6371865664886395,
+      "grad_norm": 0.8693411350250244,
+      "learning_rate": 4.099066009074505e-05,
+      "loss": 2.6445,
+      "step": 7039
+    },
+    {
+      "epoch": 0.6372770888023898,
+      "grad_norm": 0.9868283867835999,
+      "learning_rate": 4.096504369196704e-05,
+      "loss": 2.6139,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6373676111161402,
+      "grad_norm": 0.8638676404953003,
+      "learning_rate": 4.093943323800745e-05,
+      "loss": 2.6561,
+      "step": 7041
+    },
+    {
+      "epoch": 0.6374581334298904,
+      "grad_norm": 0.844467043876648,
+      "learning_rate": 4.091382873144539e-05,
+      "loss": 2.6476,
+      "step": 7042
+    },
+    {
+      "epoch": 0.6375486557436408,
+      "grad_norm": 0.8906952738761902,
+      "learning_rate": 4.088823017485912e-05,
+      "loss": 2.6715,
+      "step": 7043
+    },
+    {
+      "epoch": 0.6376391780573911,
+      "grad_norm": 0.9263741374015808,
+      "learning_rate": 4.086263757082646e-05,
+      "loss": 2.6712,
+      "step": 7044
+    },
+    {
+      "epoch": 0.6377297003711415,
+      "grad_norm": 0.9978516101837158,
+      "learning_rate": 4.083705092192457e-05,
+      "loss": 2.6592,
+      "step": 7045
+    },
+    {
+      "epoch": 0.6378202226848918,
+      "grad_norm": 0.8400484323501587,
+      "learning_rate": 4.0811470230730045e-05,
+      "loss": 2.6687,
+      "step": 7046
+    },
+    {
+      "epoch": 0.6379107449986422,
+      "grad_norm": 0.812984049320221,
+      "learning_rate": 4.078589549981884e-05,
+      "loss": 2.65,
+      "step": 7047
+    },
+    {
+      "epoch": 0.6380012673123925,
+      "grad_norm": 0.8772035241127014,
+      "learning_rate": 4.0760326731766374e-05,
+      "loss": 2.6881,
+      "step": 7048
+    },
+    {
+      "epoch": 0.6380917896261429,
+      "grad_norm": 0.8191560506820679,
+      "learning_rate": 4.073476392914731e-05,
+      "loss": 2.6028,
+      "step": 7049
+    },
+    {
+      "epoch": 0.6381823119398932,
+      "grad_norm": 0.8499199748039246,
+      "learning_rate": 4.070920709453597e-05,
+      "loss": 2.6397,
+      "step": 7050
+    },
+    {
+      "epoch": 0.6382728342536436,
+      "grad_norm": 0.896375834941864,
+      "learning_rate": 4.068365623050582e-05,
+      "loss": 2.6906,
+      "step": 7051
+    },
+    {
+      "epoch": 0.6383633565673938,
+      "grad_norm": 0.8490253686904907,
+      "learning_rate": 4.065811133962987e-05,
+      "loss": 2.6626,
+      "step": 7052
+    },
+    {
+      "epoch": 0.6384538788811442,
+      "grad_norm": 0.8479554057121277,
+      "learning_rate": 4.0632572424480476e-05,
+      "loss": 2.6799,
+      "step": 7053
+    },
+    {
+      "epoch": 0.6385444011948945,
+      "grad_norm": 0.8941323161125183,
+      "learning_rate": 4.060703948762945e-05,
+      "loss": 2.7183,
+      "step": 7054
+    },
+    {
+      "epoch": 0.6386349235086449,
+      "grad_norm": 0.8600932955741882,
+      "learning_rate": 4.0581512531647857e-05,
+      "loss": 2.6362,
+      "step": 7055
+    },
+    {
+      "epoch": 0.6387254458223952,
+      "grad_norm": 0.8809225559234619,
+      "learning_rate": 4.055599155910639e-05,
+      "loss": 2.7255,
+      "step": 7056
+    },
+    {
+      "epoch": 0.6388159681361456,
+      "grad_norm": 0.8225449919700623,
+      "learning_rate": 4.0530476572574915e-05,
+      "loss": 2.6802,
+      "step": 7057
+    },
+    {
+      "epoch": 0.6389064904498959,
+      "grad_norm": 0.8979951739311218,
+      "learning_rate": 4.0504967574622824e-05,
+      "loss": 2.7077,
+      "step": 7058
+    },
+    {
+      "epoch": 0.6389970127636463,
+      "grad_norm": 0.8487021923065186,
+      "learning_rate": 4.047946456781887e-05,
+      "loss": 2.6969,
+      "step": 7059
+    },
+    {
+      "epoch": 0.6390875350773966,
+      "grad_norm": 0.8609391450881958,
+      "learning_rate": 4.045396755473121e-05,
+      "loss": 2.6395,
+      "step": 7060
+    },
+    {
+      "epoch": 0.639178057391147,
+      "grad_norm": 0.8995999693870544,
+      "learning_rate": 4.042847653792737e-05,
+      "loss": 2.6665,
+      "step": 7061
+    },
+    {
+      "epoch": 0.6392685797048973,
+      "grad_norm": 0.94828200340271,
+      "learning_rate": 4.0402991519974364e-05,
+      "loss": 2.6555,
+      "step": 7062
+    },
+    {
+      "epoch": 0.6393591020186477,
+      "grad_norm": 0.8435460329055786,
+      "learning_rate": 4.037751250343841e-05,
+      "loss": 2.6601,
+      "step": 7063
+    },
+    {
+      "epoch": 0.6394496243323979,
+      "grad_norm": 0.8438577055931091,
+      "learning_rate": 4.0352039490885384e-05,
+      "loss": 2.6738,
+      "step": 7064
+    },
+    {
+      "epoch": 0.6395401466461482,
+      "grad_norm": 0.8616018295288086,
+      "learning_rate": 4.032657248488031e-05,
+      "loss": 2.6628,
+      "step": 7065
+    },
+    {
+      "epoch": 0.6396306689598986,
+      "grad_norm": 0.8819381594657898,
+      "learning_rate": 4.030111148798775e-05,
+      "loss": 2.704,
+      "step": 7066
+    },
+    {
+      "epoch": 0.6397211912736489,
+      "grad_norm": 0.8637552261352539,
+      "learning_rate": 4.027565650277164e-05,
+      "loss": 2.7014,
+      "step": 7067
+    },
+    {
+      "epoch": 0.6398117135873993,
+      "grad_norm": 0.8700119853019714,
+      "learning_rate": 4.025020753179528e-05,
+      "loss": 2.6311,
+      "step": 7068
+    },
+    {
+      "epoch": 0.6399022359011496,
+      "grad_norm": 0.8614180684089661,
+      "learning_rate": 4.0224764577621396e-05,
+      "loss": 2.6712,
+      "step": 7069
+    },
+    {
+      "epoch": 0.6399927582149,
+      "grad_norm": 0.8416438698768616,
+      "learning_rate": 4.019932764281211e-05,
+      "loss": 2.6397,
+      "step": 7070
+    },
+    {
+      "epoch": 0.6400832805286503,
+      "grad_norm": 0.849120557308197,
+      "learning_rate": 4.017389672992883e-05,
+      "loss": 2.6249,
+      "step": 7071
+    },
+    {
+      "epoch": 0.6401738028424007,
+      "grad_norm": 0.91354900598526,
+      "learning_rate": 4.014847184153258e-05,
+      "loss": 2.6493,
+      "step": 7072
+    },
+    {
+      "epoch": 0.640264325156151,
+      "grad_norm": 0.8874666690826416,
+      "learning_rate": 4.012305298018354e-05,
+      "loss": 2.6636,
+      "step": 7073
+    },
+    {
+      "epoch": 0.6403548474699013,
+      "grad_norm": 0.9202023148536682,
+      "learning_rate": 4.009764014844143e-05,
+      "loss": 2.6334,
+      "step": 7074
+    },
+    {
+      "epoch": 0.6404453697836516,
+      "grad_norm": 0.8148588538169861,
+      "learning_rate": 4.007223334886531e-05,
+      "loss": 2.6099,
+      "step": 7075
+    },
+    {
+      "epoch": 0.640535892097402,
+      "grad_norm": 0.8374625444412231,
+      "learning_rate": 4.004683258401366e-05,
+      "loss": 2.6621,
+      "step": 7076
+    },
+    {
+      "epoch": 0.6406264144111523,
+      "grad_norm": 0.8390886187553406,
+      "learning_rate": 4.002143785644432e-05,
+      "loss": 2.6802,
+      "step": 7077
+    },
+    {
+      "epoch": 0.6407169367249027,
+      "grad_norm": 0.908154308795929,
+      "learning_rate": 3.9996049168714586e-05,
+      "loss": 2.6487,
+      "step": 7078
+    },
+    {
+      "epoch": 0.640807459038653,
+      "grad_norm": 0.8716943860054016,
+      "learning_rate": 3.9970666523380994e-05,
+      "loss": 2.6942,
+      "step": 7079
+    },
+    {
+      "epoch": 0.6408979813524034,
+      "grad_norm": 0.8854706287384033,
+      "learning_rate": 3.994528992299971e-05,
+      "loss": 2.6846,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6409885036661537,
+      "grad_norm": 1.0154533386230469,
+      "learning_rate": 3.9919919370126045e-05,
+      "loss": 2.6979,
+      "step": 7081
+    },
+    {
+      "epoch": 0.6410790259799041,
+      "grad_norm": 0.9278495907783508,
+      "learning_rate": 3.9894554867314856e-05,
+      "loss": 2.7322,
+      "step": 7082
+    },
+    {
+      "epoch": 0.6411695482936544,
+      "grad_norm": 0.881550133228302,
+      "learning_rate": 3.9869196417120355e-05,
+      "loss": 2.6633,
+      "step": 7083
+    },
+    {
+      "epoch": 0.6412600706074048,
+      "grad_norm": 0.8752927184104919,
+      "learning_rate": 3.9843844022096135e-05,
+      "loss": 2.6351,
+      "step": 7084
+    },
+    {
+      "epoch": 0.641350592921155,
+      "grad_norm": 0.8580919504165649,
+      "learning_rate": 3.981849768479517e-05,
+      "loss": 2.6034,
+      "step": 7085
+    },
+    {
+      "epoch": 0.6414411152349054,
+      "grad_norm": 0.9341000318527222,
+      "learning_rate": 3.9793157407769866e-05,
+      "loss": 2.7307,
+      "step": 7086
+    },
+    {
+      "epoch": 0.6415316375486557,
+      "grad_norm": 0.892649233341217,
+      "learning_rate": 3.9767823193571896e-05,
+      "loss": 2.6646,
+      "step": 7087
+    },
+    {
+      "epoch": 0.6416221598624061,
+      "grad_norm": 0.8354327082633972,
+      "learning_rate": 3.974249504475256e-05,
+      "loss": 2.6757,
+      "step": 7088
+    },
+    {
+      "epoch": 0.6417126821761564,
+      "grad_norm": 0.90390545129776,
+      "learning_rate": 3.9717172963862294e-05,
+      "loss": 2.6837,
+      "step": 7089
+    },
+    {
+      "epoch": 0.6418032044899068,
+      "grad_norm": 0.8547748923301697,
+      "learning_rate": 3.969185695345105e-05,
+      "loss": 2.6614,
+      "step": 7090
+    },
+    {
+      "epoch": 0.6418937268036571,
+      "grad_norm": 0.9019080996513367,
+      "learning_rate": 3.9666547016068167e-05,
+      "loss": 2.6596,
+      "step": 7091
+    },
+    {
+      "epoch": 0.6419842491174075,
+      "grad_norm": 0.9135777950286865,
+      "learning_rate": 3.964124315426237e-05,
+      "loss": 2.6858,
+      "step": 7092
+    },
+    {
+      "epoch": 0.6420747714311578,
+      "grad_norm": 0.8872252106666565,
+      "learning_rate": 3.961594537058169e-05,
+      "loss": 2.6434,
+      "step": 7093
+    },
+    {
+      "epoch": 0.6421652937449082,
+      "grad_norm": 0.948832094669342,
+      "learning_rate": 3.959065366757371e-05,
+      "loss": 2.6778,
+      "step": 7094
+    },
+    {
+      "epoch": 0.6422558160586584,
+      "grad_norm": 0.9001719951629639,
+      "learning_rate": 3.956536804778522e-05,
+      "loss": 2.7093,
+      "step": 7095
+    },
+    {
+      "epoch": 0.6423463383724088,
+      "grad_norm": 0.7865238189697266,
+      "learning_rate": 3.954008851376252e-05,
+      "loss": 2.6818,
+      "step": 7096
+    },
+    {
+      "epoch": 0.6424368606861591,
+      "grad_norm": 0.8909227252006531,
+      "learning_rate": 3.951481506805125e-05,
+      "loss": 2.6662,
+      "step": 7097
+    },
+    {
+      "epoch": 0.6425273829999095,
+      "grad_norm": 0.8204286098480225,
+      "learning_rate": 3.948954771319645e-05,
+      "loss": 2.6512,
+      "step": 7098
+    },
+    {
+      "epoch": 0.6426179053136598,
+      "grad_norm": 0.8769195079803467,
+      "learning_rate": 3.946428645174257e-05,
+      "loss": 2.6931,
+      "step": 7099
+    },
+    {
+      "epoch": 0.6427084276274102,
+      "grad_norm": 0.814003050327301,
+      "learning_rate": 3.943903128623335e-05,
+      "loss": 2.6869,
+      "step": 7100
+    },
+    {
+      "epoch": 0.6427989499411605,
+      "grad_norm": 0.8625674247741699,
+      "learning_rate": 3.9413782219212035e-05,
+      "loss": 2.6778,
+      "step": 7101
+    },
+    {
+      "epoch": 0.6428894722549109,
+      "grad_norm": 0.9159715175628662,
+      "learning_rate": 3.938853925322118e-05,
+      "loss": 2.6754,
+      "step": 7102
+    },
+    {
+      "epoch": 0.6429799945686612,
+      "grad_norm": 0.8626770377159119,
+      "learning_rate": 3.936330239080277e-05,
+      "loss": 2.6311,
+      "step": 7103
+    },
+    {
+      "epoch": 0.6430705168824116,
+      "grad_norm": 0.833588182926178,
+      "learning_rate": 3.933807163449815e-05,
+      "loss": 2.6601,
+      "step": 7104
+    },
+    {
+      "epoch": 0.6431610391961619,
+      "grad_norm": 0.8459279537200928,
+      "learning_rate": 3.9312846986848095e-05,
+      "loss": 2.7138,
+      "step": 7105
+    },
+    {
+      "epoch": 0.6432515615099121,
+      "grad_norm": 0.837768018245697,
+      "learning_rate": 3.928762845039261e-05,
+      "loss": 2.6789,
+      "step": 7106
+    },
+    {
+      "epoch": 0.6433420838236625,
+      "grad_norm": 0.8656929135322571,
+      "learning_rate": 3.9262416027671356e-05,
+      "loss": 2.694,
+      "step": 7107
+    },
+    {
+      "epoch": 0.6434326061374128,
+      "grad_norm": 0.8893898129463196,
+      "learning_rate": 3.923720972122311e-05,
+      "loss": 2.6351,
+      "step": 7108
+    },
+    {
+      "epoch": 0.6435231284511632,
+      "grad_norm": 0.7833218574523926,
+      "learning_rate": 3.921200953358618e-05,
+      "loss": 2.6247,
+      "step": 7109
+    },
+    {
+      "epoch": 0.6436136507649135,
+      "grad_norm": 0.8954211473464966,
+      "learning_rate": 3.918681546729822e-05,
+      "loss": 2.6661,
+      "step": 7110
+    },
+    {
+      "epoch": 0.6437041730786639,
+      "grad_norm": 0.822925865650177,
+      "learning_rate": 3.91616275248963e-05,
+      "loss": 2.6223,
+      "step": 7111
+    },
+    {
+      "epoch": 0.6437946953924142,
+      "grad_norm": 0.9061444401741028,
+      "learning_rate": 3.91364457089168e-05,
+      "loss": 2.7069,
+      "step": 7112
+    },
+    {
+      "epoch": 0.6438852177061646,
+      "grad_norm": 0.8984582424163818,
+      "learning_rate": 3.9111270021895596e-05,
+      "loss": 2.6528,
+      "step": 7113
+    },
+    {
+      "epoch": 0.6439757400199149,
+      "grad_norm": 0.8554080128669739,
+      "learning_rate": 3.908610046636776e-05,
+      "loss": 2.6859,
+      "step": 7114
+    },
+    {
+      "epoch": 0.6440662623336653,
+      "grad_norm": 0.8965380191802979,
+      "learning_rate": 3.906093704486802e-05,
+      "loss": 2.7286,
+      "step": 7115
+    },
+    {
+      "epoch": 0.6441567846474155,
+      "grad_norm": 0.8609764575958252,
+      "learning_rate": 3.903577975993021e-05,
+      "loss": 2.6072,
+      "step": 7116
+    },
+    {
+      "epoch": 0.6442473069611659,
+      "grad_norm": 0.8662620782852173,
+      "learning_rate": 3.9010628614087705e-05,
+      "loss": 2.6947,
+      "step": 7117
+    },
+    {
+      "epoch": 0.6443378292749162,
+      "grad_norm": 0.880328893661499,
+      "learning_rate": 3.8985483609873244e-05,
+      "loss": 2.636,
+      "step": 7118
+    },
+    {
+      "epoch": 0.6444283515886666,
+      "grad_norm": 0.897362470626831,
+      "learning_rate": 3.896034474981894e-05,
+      "loss": 2.6112,
+      "step": 7119
+    },
+    {
+      "epoch": 0.6445188739024169,
+      "grad_norm": 0.9112858176231384,
+      "learning_rate": 3.893521203645618e-05,
+      "loss": 2.7018,
+      "step": 7120
+    },
+    {
+      "epoch": 0.6446093962161673,
+      "grad_norm": 0.8464901447296143,
+      "learning_rate": 3.891008547231597e-05,
+      "loss": 2.6105,
+      "step": 7121
+    },
+    {
+      "epoch": 0.6446999185299176,
+      "grad_norm": 0.888416051864624,
+      "learning_rate": 3.8884965059928444e-05,
+      "loss": 2.7111,
+      "step": 7122
+    },
+    {
+      "epoch": 0.644790440843668,
+      "grad_norm": 0.8662939667701721,
+      "learning_rate": 3.885985080182327e-05,
+      "loss": 2.6843,
+      "step": 7123
+    },
+    {
+      "epoch": 0.6448809631574183,
+      "grad_norm": 0.9020508527755737,
+      "learning_rate": 3.8834742700529435e-05,
+      "loss": 2.7249,
+      "step": 7124
+    },
+    {
+      "epoch": 0.6449714854711687,
+      "grad_norm": 0.8338090777397156,
+      "learning_rate": 3.880964075857535e-05,
+      "loss": 2.6608,
+      "step": 7125
+    },
+    {
+      "epoch": 0.645062007784919,
+      "grad_norm": 0.7852875590324402,
+      "learning_rate": 3.8784544978488756e-05,
+      "loss": 2.6479,
+      "step": 7126
+    },
+    {
+      "epoch": 0.6451525300986694,
+      "grad_norm": 0.7826331853866577,
+      "learning_rate": 3.8759455362796846e-05,
+      "loss": 2.6882,
+      "step": 7127
+    },
+    {
+      "epoch": 0.6452430524124196,
+      "grad_norm": 0.9392991065979004,
+      "learning_rate": 3.8734371914026044e-05,
+      "loss": 2.7536,
+      "step": 7128
+    },
+    {
+      "epoch": 0.64533357472617,
+      "grad_norm": 0.8982208371162415,
+      "learning_rate": 3.8709294634702376e-05,
+      "loss": 2.7042,
+      "step": 7129
+    },
+    {
+      "epoch": 0.6454240970399203,
+      "grad_norm": 0.9267832040786743,
+      "learning_rate": 3.8684223527351025e-05,
+      "loss": 2.6595,
+      "step": 7130
+    },
+    {
+      "epoch": 0.6455146193536707,
+      "grad_norm": 0.8747368454933167,
+      "learning_rate": 3.865915859449669e-05,
+      "loss": 2.6925,
+      "step": 7131
+    },
+    {
+      "epoch": 0.645605141667421,
+      "grad_norm": 0.8629845380783081,
+      "learning_rate": 3.863409983866341e-05,
+      "loss": 2.711,
+      "step": 7132
+    },
+    {
+      "epoch": 0.6456956639811714,
+      "grad_norm": 0.8262534141540527,
+      "learning_rate": 3.8609047262374586e-05,
+      "loss": 2.6492,
+      "step": 7133
+    },
+    {
+      "epoch": 0.6457861862949217,
+      "grad_norm": 0.8800877332687378,
+      "learning_rate": 3.8584000868153025e-05,
+      "loss": 2.7027,
+      "step": 7134
+    },
+    {
+      "epoch": 0.6458767086086721,
+      "grad_norm": 0.9538904428482056,
+      "learning_rate": 3.855896065852094e-05,
+      "loss": 2.6911,
+      "step": 7135
+    },
+    {
+      "epoch": 0.6459672309224224,
+      "grad_norm": 0.8761887550354004,
+      "learning_rate": 3.853392663599976e-05,
+      "loss": 2.6475,
+      "step": 7136
+    },
+    {
+      "epoch": 0.6460577532361728,
+      "grad_norm": 0.9193413257598877,
+      "learning_rate": 3.850889880311055e-05,
+      "loss": 2.7227,
+      "step": 7137
+    },
+    {
+      "epoch": 0.646148275549923,
+      "grad_norm": 0.8643597960472107,
+      "learning_rate": 3.848387716237353e-05,
+      "loss": 2.6666,
+      "step": 7138
+    },
+    {
+      "epoch": 0.6462387978636734,
+      "grad_norm": 0.8625378012657166,
+      "learning_rate": 3.845886171630838e-05,
+      "loss": 2.6604,
+      "step": 7139
+    },
+    {
+      "epoch": 0.6463293201774237,
+      "grad_norm": 0.8720336556434631,
+      "learning_rate": 3.843385246743417e-05,
+      "loss": 2.6764,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6464198424911741,
+      "grad_norm": 0.8405011296272278,
+      "learning_rate": 3.8408849418269346e-05,
+      "loss": 2.685,
+      "step": 7141
+    },
+    {
+      "epoch": 0.6465103648049244,
+      "grad_norm": 0.882945716381073,
+      "learning_rate": 3.838385257133169e-05,
+      "loss": 2.7478,
+      "step": 7142
+    },
+    {
+      "epoch": 0.6466008871186748,
+      "grad_norm": 0.9414501786231995,
+      "learning_rate": 3.8358861929138436e-05,
+      "loss": 2.6697,
+      "step": 7143
+    },
+    {
+      "epoch": 0.6466914094324251,
+      "grad_norm": 0.8411595821380615,
+      "learning_rate": 3.833387749420603e-05,
+      "loss": 2.6961,
+      "step": 7144
+    },
+    {
+      "epoch": 0.6467819317461755,
+      "grad_norm": 0.8170534372329712,
+      "learning_rate": 3.8308899269050535e-05,
+      "loss": 2.6046,
+      "step": 7145
+    },
+    {
+      "epoch": 0.6468724540599258,
+      "grad_norm": 0.8441464900970459,
+      "learning_rate": 3.828392725618717e-05,
+      "loss": 2.6861,
+      "step": 7146
+    },
+    {
+      "epoch": 0.6469629763736761,
+      "grad_norm": 0.8740844130516052,
+      "learning_rate": 3.8258961458130625e-05,
+      "loss": 2.6563,
+      "step": 7147
+    },
+    {
+      "epoch": 0.6470534986874265,
+      "grad_norm": 0.8880472183227539,
+      "learning_rate": 3.823400187739499e-05,
+      "loss": 2.6645,
+      "step": 7148
+    },
+    {
+      "epoch": 0.6471440210011767,
+      "grad_norm": 0.854893684387207,
+      "learning_rate": 3.8209048516493706e-05,
+      "loss": 2.7327,
+      "step": 7149
+    },
+    {
+      "epoch": 0.6472345433149271,
+      "grad_norm": 0.8700079321861267,
+      "learning_rate": 3.8184101377939476e-05,
+      "loss": 2.6862,
+      "step": 7150
+    },
+    {
+      "epoch": 0.6473250656286774,
+      "grad_norm": 0.934057354927063,
+      "learning_rate": 3.8159160464244606e-05,
+      "loss": 2.6973,
+      "step": 7151
+    },
+    {
+      "epoch": 0.6474155879424278,
+      "grad_norm": 0.8303758502006531,
+      "learning_rate": 3.813422577792056e-05,
+      "loss": 2.6229,
+      "step": 7152
+    },
+    {
+      "epoch": 0.6475061102561781,
+      "grad_norm": 0.9065086245536804,
+      "learning_rate": 3.810929732147829e-05,
+      "loss": 2.6548,
+      "step": 7153
+    },
+    {
+      "epoch": 0.6475966325699285,
+      "grad_norm": 0.9324973225593567,
+      "learning_rate": 3.808437509742807e-05,
+      "loss": 2.6761,
+      "step": 7154
+    },
+    {
+      "epoch": 0.6476871548836788,
+      "grad_norm": 0.8872169256210327,
+      "learning_rate": 3.8059459108279594e-05,
+      "loss": 2.681,
+      "step": 7155
+    },
+    {
+      "epoch": 0.6477776771974292,
+      "grad_norm": 0.815177321434021,
+      "learning_rate": 3.8034549356541894e-05,
+      "loss": 2.6889,
+      "step": 7156
+    },
+    {
+      "epoch": 0.6478681995111795,
+      "grad_norm": 0.8200878500938416,
+      "learning_rate": 3.8009645844723406e-05,
+      "loss": 2.5912,
+      "step": 7157
+    },
+    {
+      "epoch": 0.6479587218249299,
+      "grad_norm": 0.8385193943977356,
+      "learning_rate": 3.798474857533182e-05,
+      "loss": 2.6814,
+      "step": 7158
+    },
+    {
+      "epoch": 0.6480492441386801,
+      "grad_norm": 0.8454488515853882,
+      "learning_rate": 3.795985755087442e-05,
+      "loss": 2.6542,
+      "step": 7159
+    },
+    {
+      "epoch": 0.6481397664524305,
+      "grad_norm": 0.8543698787689209,
+      "learning_rate": 3.7934972773857634e-05,
+      "loss": 2.6238,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6482302887661808,
+      "grad_norm": 0.8137564659118652,
+      "learning_rate": 3.79100942467874e-05,
+      "loss": 2.6324,
+      "step": 7161
+    },
+    {
+      "epoch": 0.6483208110799312,
+      "grad_norm": 1.0308117866516113,
+      "learning_rate": 3.788522197216897e-05,
+      "loss": 2.646,
+      "step": 7162
+    },
+    {
+      "epoch": 0.6484113333936815,
+      "grad_norm": 0.9226447939872742,
+      "learning_rate": 3.7860355952506985e-05,
+      "loss": 2.6695,
+      "step": 7163
+    },
+    {
+      "epoch": 0.6485018557074319,
+      "grad_norm": 0.8795346021652222,
+      "learning_rate": 3.7835496190305463e-05,
+      "loss": 2.6895,
+      "step": 7164
+    },
+    {
+      "epoch": 0.6485923780211822,
+      "grad_norm": 0.8603231310844421,
+      "learning_rate": 3.7810642688067796e-05,
+      "loss": 2.5942,
+      "step": 7165
+    },
+    {
+      "epoch": 0.6486829003349326,
+      "grad_norm": 0.8747546672821045,
+      "learning_rate": 3.7785795448296646e-05,
+      "loss": 2.6818,
+      "step": 7166
+    },
+    {
+      "epoch": 0.6487734226486829,
+      "grad_norm": 0.9285270571708679,
+      "learning_rate": 3.776095447349427e-05,
+      "loss": 2.7052,
+      "step": 7167
+    },
+    {
+      "epoch": 0.6488639449624333,
+      "grad_norm": 0.8840459585189819,
+      "learning_rate": 3.773611976616203e-05,
+      "loss": 2.7358,
+      "step": 7168
+    },
+    {
+      "epoch": 0.6489544672761836,
+      "grad_norm": 0.7858835458755493,
+      "learning_rate": 3.7711291328800824e-05,
+      "loss": 2.6761,
+      "step": 7169
+    },
+    {
+      "epoch": 0.649044989589934,
+      "grad_norm": 0.8488116264343262,
+      "learning_rate": 3.7686469163910885e-05,
+      "loss": 2.6549,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6491355119036842,
+      "grad_norm": 0.8749338388442993,
+      "learning_rate": 3.766165327399179e-05,
+      "loss": 2.6001,
+      "step": 7171
+    },
+    {
+      "epoch": 0.6492260342174346,
+      "grad_norm": 0.9855179190635681,
+      "learning_rate": 3.7636843661542506e-05,
+      "loss": 2.731,
+      "step": 7172
+    },
+    {
+      "epoch": 0.6493165565311849,
+      "grad_norm": 0.8725718855857849,
+      "learning_rate": 3.7612040329061405e-05,
+      "loss": 2.593,
+      "step": 7173
+    },
+    {
+      "epoch": 0.6494070788449353,
+      "grad_norm": 0.9173427224159241,
+      "learning_rate": 3.7587243279046056e-05,
+      "loss": 2.6827,
+      "step": 7174
+    },
+    {
+      "epoch": 0.6494976011586856,
+      "grad_norm": 0.9711499810218811,
+      "learning_rate": 3.7562452513993674e-05,
+      "loss": 2.6605,
+      "step": 7175
+    },
+    {
+      "epoch": 0.649588123472436,
+      "grad_norm": 0.8643674254417419,
+      "learning_rate": 3.7537668036400574e-05,
+      "loss": 2.6727,
+      "step": 7176
+    },
+    {
+      "epoch": 0.6496786457861863,
+      "grad_norm": 0.9335436224937439,
+      "learning_rate": 3.75128898487626e-05,
+      "loss": 2.6548,
+      "step": 7177
+    },
+    {
+      "epoch": 0.6497691680999367,
+      "grad_norm": 0.8806255459785461,
+      "learning_rate": 3.748811795357491e-05,
+      "loss": 2.5545,
+      "step": 7178
+    },
+    {
+      "epoch": 0.649859690413687,
+      "grad_norm": 0.8108535408973694,
+      "learning_rate": 3.746335235333203e-05,
+      "loss": 2.6547,
+      "step": 7179
+    },
+    {
+      "epoch": 0.6499502127274374,
+      "grad_norm": 0.9339863657951355,
+      "learning_rate": 3.7438593050527845e-05,
+      "loss": 2.6247,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6500407350411876,
+      "grad_norm": 0.8718104958534241,
+      "learning_rate": 3.7413840047655666e-05,
+      "loss": 2.6577,
+      "step": 7181
+    },
+    {
+      "epoch": 0.650131257354938,
+      "grad_norm": 0.8415114879608154,
+      "learning_rate": 3.7389093347208026e-05,
+      "loss": 2.735,
+      "step": 7182
+    },
+    {
+      "epoch": 0.6502217796686883,
+      "grad_norm": 0.8771581649780273,
+      "learning_rate": 3.736435295167704e-05,
+      "loss": 2.7182,
+      "step": 7183
+    },
+    {
+      "epoch": 0.6503123019824387,
+      "grad_norm": 0.828473687171936,
+      "learning_rate": 3.733961886355398e-05,
+      "loss": 2.6797,
+      "step": 7184
+    },
+    {
+      "epoch": 0.650402824296189,
+      "grad_norm": 0.8189043402671814,
+      "learning_rate": 3.731489108532954e-05,
+      "loss": 2.6999,
+      "step": 7185
+    },
+    {
+      "epoch": 0.6504933466099394,
+      "grad_norm": 0.913808286190033,
+      "learning_rate": 3.729016961949391e-05,
+      "loss": 2.6483,
+      "step": 7186
+    },
+    {
+      "epoch": 0.6505838689236897,
+      "grad_norm": 0.8551123738288879,
+      "learning_rate": 3.7265454468536456e-05,
+      "loss": 2.6946,
+      "step": 7187
+    },
+    {
+      "epoch": 0.65067439123744,
+      "grad_norm": 0.8683720827102661,
+      "learning_rate": 3.724074563494601e-05,
+      "loss": 2.7036,
+      "step": 7188
+    },
+    {
+      "epoch": 0.6507649135511904,
+      "grad_norm": 0.8683645725250244,
+      "learning_rate": 3.721604312121076e-05,
+      "loss": 2.6839,
+      "step": 7189
+    },
+    {
+      "epoch": 0.6508554358649407,
+      "grad_norm": 0.8150107860565186,
+      "learning_rate": 3.719134692981826e-05,
+      "loss": 2.6815,
+      "step": 7190
+    },
+    {
+      "epoch": 0.650945958178691,
+      "grad_norm": 0.8053891062736511,
+      "learning_rate": 3.71666570632554e-05,
+      "loss": 2.5942,
+      "step": 7191
+    },
+    {
+      "epoch": 0.6510364804924413,
+      "grad_norm": 0.8532904982566833,
+      "learning_rate": 3.714197352400849e-05,
+      "loss": 2.6201,
+      "step": 7192
+    },
+    {
+      "epoch": 0.6511270028061917,
+      "grad_norm": 0.8892574310302734,
+      "learning_rate": 3.711729631456305e-05,
+      "loss": 2.7435,
+      "step": 7193
+    },
+    {
+      "epoch": 0.651217525119942,
+      "grad_norm": 0.8301617503166199,
+      "learning_rate": 3.7092625437404236e-05,
+      "loss": 2.6718,
+      "step": 7194
+    },
+    {
+      "epoch": 0.6513080474336924,
+      "grad_norm": 0.801541268825531,
+      "learning_rate": 3.7067960895016275e-05,
+      "loss": 2.6546,
+      "step": 7195
+    },
+    {
+      "epoch": 0.6513985697474427,
+      "grad_norm": 0.8979970216751099,
+      "learning_rate": 3.704330268988293e-05,
+      "loss": 2.6777,
+      "step": 7196
+    },
+    {
+      "epoch": 0.6514890920611931,
+      "grad_norm": 0.8236854672431946,
+      "learning_rate": 3.7018650824487277e-05,
+      "loss": 2.6671,
+      "step": 7197
+    },
+    {
+      "epoch": 0.6515796143749434,
+      "grad_norm": 0.8789283633232117,
+      "learning_rate": 3.6994005301311777e-05,
+      "loss": 2.6919,
+      "step": 7198
+    },
+    {
+      "epoch": 0.6516701366886938,
+      "grad_norm": 0.8241438865661621,
+      "learning_rate": 3.6969366122838215e-05,
+      "loss": 2.6186,
+      "step": 7199
+    },
+    {
+      "epoch": 0.6517606590024441,
+      "grad_norm": 0.8289199471473694,
+      "learning_rate": 3.694473329154778e-05,
+      "loss": 2.6487,
+      "step": 7200
+    },
+    {
+      "epoch": 0.6518511813161945,
+      "grad_norm": 0.8322336077690125,
+      "learning_rate": 3.692010680992092e-05,
+      "loss": 2.6941,
+      "step": 7201
+    },
+    {
+      "epoch": 0.6519417036299447,
+      "grad_norm": 0.8611689805984497,
+      "learning_rate": 3.689548668043764e-05,
+      "loss": 2.6761,
+      "step": 7202
+    },
+    {
+      "epoch": 0.6520322259436951,
+      "grad_norm": 0.8806263208389282,
+      "learning_rate": 3.68708729055771e-05,
+      "loss": 2.6751,
+      "step": 7203
+    },
+    {
+      "epoch": 0.6521227482574454,
+      "grad_norm": 0.836117148399353,
+      "learning_rate": 3.684626548781792e-05,
+      "loss": 2.6244,
+      "step": 7204
+    },
+    {
+      "epoch": 0.6522132705711958,
+      "grad_norm": 0.9649540185928345,
+      "learning_rate": 3.682166442963809e-05,
+      "loss": 2.7419,
+      "step": 7205
+    },
+    {
+      "epoch": 0.6523037928849461,
+      "grad_norm": 0.8877153396606445,
+      "learning_rate": 3.679706973351491e-05,
+      "loss": 2.6511,
+      "step": 7206
+    },
+    {
+      "epoch": 0.6523943151986965,
+      "grad_norm": 0.9087035655975342,
+      "learning_rate": 3.67724814019251e-05,
+      "loss": 2.6311,
+      "step": 7207
+    },
+    {
+      "epoch": 0.6524848375124468,
+      "grad_norm": 0.9013026356697083,
+      "learning_rate": 3.674789943734469e-05,
+      "loss": 2.6748,
+      "step": 7208
+    },
+    {
+      "epoch": 0.6525753598261972,
+      "grad_norm": 0.8409872651100159,
+      "learning_rate": 3.6723323842249025e-05,
+      "loss": 2.7069,
+      "step": 7209
+    },
+    {
+      "epoch": 0.6526658821399475,
+      "grad_norm": 0.8533176183700562,
+      "learning_rate": 3.669875461911297e-05,
+      "loss": 2.6412,
+      "step": 7210
+    },
+    {
+      "epoch": 0.6527564044536979,
+      "grad_norm": 0.8591048717498779,
+      "learning_rate": 3.667419177041058e-05,
+      "loss": 2.7495,
+      "step": 7211
+    },
+    {
+      "epoch": 0.6528469267674482,
+      "grad_norm": 0.8098570108413696,
+      "learning_rate": 3.664963529861534e-05,
+      "loss": 2.6755,
+      "step": 7212
+    },
+    {
+      "epoch": 0.6529374490811986,
+      "grad_norm": 0.871677815914154,
+      "learning_rate": 3.662508520620008e-05,
+      "loss": 2.6289,
+      "step": 7213
+    },
+    {
+      "epoch": 0.6530279713949488,
+      "grad_norm": 0.895744800567627,
+      "learning_rate": 3.6600541495637055e-05,
+      "loss": 2.6593,
+      "step": 7214
+    },
+    {
+      "epoch": 0.6531184937086992,
+      "grad_norm": 0.836062490940094,
+      "learning_rate": 3.657600416939768e-05,
+      "loss": 2.6574,
+      "step": 7215
+    },
+    {
+      "epoch": 0.6532090160224495,
+      "grad_norm": 0.8524973392486572,
+      "learning_rate": 3.6551473229953037e-05,
+      "loss": 2.6401,
+      "step": 7216
+    },
+    {
+      "epoch": 0.6532995383361999,
+      "grad_norm": 0.8783817291259766,
+      "learning_rate": 3.6526948679773257e-05,
+      "loss": 2.6905,
+      "step": 7217
+    },
+    {
+      "epoch": 0.6533900606499502,
+      "grad_norm": 1.0054057836532593,
+      "learning_rate": 3.650243052132801e-05,
+      "loss": 2.6963,
+      "step": 7218
+    },
+    {
+      "epoch": 0.6534805829637006,
+      "grad_norm": 0.8324257135391235,
+      "learning_rate": 3.647791875708628e-05,
+      "loss": 2.6178,
+      "step": 7219
+    },
+    {
+      "epoch": 0.6535711052774509,
+      "grad_norm": 0.8131170272827148,
+      "learning_rate": 3.645341338951639e-05,
+      "loss": 2.6597,
+      "step": 7220
+    },
+    {
+      "epoch": 0.6536616275912013,
+      "grad_norm": 0.879077136516571,
+      "learning_rate": 3.642891442108602e-05,
+      "loss": 2.6612,
+      "step": 7221
+    },
+    {
+      "epoch": 0.6537521499049516,
+      "grad_norm": 0.9463702440261841,
+      "learning_rate": 3.640442185426228e-05,
+      "loss": 2.7035,
+      "step": 7222
+    },
+    {
+      "epoch": 0.653842672218702,
+      "grad_norm": 0.8919106721878052,
+      "learning_rate": 3.6379935691511447e-05,
+      "loss": 2.6789,
+      "step": 7223
+    },
+    {
+      "epoch": 0.6539331945324522,
+      "grad_norm": 0.8371802568435669,
+      "learning_rate": 3.635545593529941e-05,
+      "loss": 2.6433,
+      "step": 7224
+    },
+    {
+      "epoch": 0.6540237168462026,
+      "grad_norm": 0.8284990191459656,
+      "learning_rate": 3.6330982588091186e-05,
+      "loss": 2.6303,
+      "step": 7225
+    },
+    {
+      "epoch": 0.6541142391599529,
+      "grad_norm": 0.9169434905052185,
+      "learning_rate": 3.630651565235128e-05,
+      "loss": 2.7317,
+      "step": 7226
+    },
+    {
+      "epoch": 0.6542047614737033,
+      "grad_norm": 0.8090481758117676,
+      "learning_rate": 3.6282055130543505e-05,
+      "loss": 2.6255,
+      "step": 7227
+    },
+    {
+      "epoch": 0.6542952837874536,
+      "grad_norm": 0.9667547941207886,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 2.6606,
+      "step": 7228
+    },
+    {
+      "epoch": 0.6543858061012039,
+      "grad_norm": 0.868475079536438,
+      "learning_rate": 3.6233153338576384e-05,
+      "loss": 2.6585,
+      "step": 7229
+    },
+    {
+      "epoch": 0.6544763284149543,
+      "grad_norm": 0.8772515058517456,
+      "learning_rate": 3.62087120733415e-05,
+      "loss": 2.6808,
+      "step": 7230
+    },
+    {
+      "epoch": 0.6545668507287046,
+      "grad_norm": 0.8411134481430054,
+      "learning_rate": 3.618427723188749e-05,
+      "loss": 2.6591,
+      "step": 7231
+    },
+    {
+      "epoch": 0.654657373042455,
+      "grad_norm": 0.8320329189300537,
+      "learning_rate": 3.615984881667508e-05,
+      "loss": 2.6187,
+      "step": 7232
+    },
+    {
+      "epoch": 0.6547478953562053,
+      "grad_norm": 0.8759403824806213,
+      "learning_rate": 3.613542683016411e-05,
+      "loss": 2.7017,
+      "step": 7233
+    },
+    {
+      "epoch": 0.6548384176699557,
+      "grad_norm": 0.9727622270584106,
+      "learning_rate": 3.611101127481392e-05,
+      "loss": 2.6601,
+      "step": 7234
+    },
+    {
+      "epoch": 0.6549289399837059,
+      "grad_norm": 0.8831741213798523,
+      "learning_rate": 3.6086602153083146e-05,
+      "loss": 2.7137,
+      "step": 7235
+    },
+    {
+      "epoch": 0.6550194622974563,
+      "grad_norm": 0.8495936393737793,
+      "learning_rate": 3.606219946742978e-05,
+      "loss": 2.6793,
+      "step": 7236
+    },
+    {
+      "epoch": 0.6551099846112066,
+      "grad_norm": 0.8561792373657227,
+      "learning_rate": 3.6037803220311187e-05,
+      "loss": 2.6837,
+      "step": 7237
+    },
+    {
+      "epoch": 0.655200506924957,
+      "grad_norm": 0.8705282211303711,
+      "learning_rate": 3.601341341418408e-05,
+      "loss": 2.6161,
+      "step": 7238
+    },
+    {
+      "epoch": 0.6552910292387073,
+      "grad_norm": 0.8599856495857239,
+      "learning_rate": 3.5989030051504434e-05,
+      "loss": 2.7134,
+      "step": 7239
+    },
+    {
+      "epoch": 0.6553815515524577,
+      "grad_norm": 0.8800930380821228,
+      "learning_rate": 3.5964653134727776e-05,
+      "loss": 2.6341,
+      "step": 7240
+    },
+    {
+      "epoch": 0.655472073866208,
+      "grad_norm": 0.8483147621154785,
+      "learning_rate": 3.594028266630877e-05,
+      "loss": 2.6875,
+      "step": 7241
+    },
+    {
+      "epoch": 0.6555625961799584,
+      "grad_norm": 0.952504575252533,
+      "learning_rate": 3.5915918648701526e-05,
+      "loss": 2.6613,
+      "step": 7242
+    },
+    {
+      "epoch": 0.6556531184937087,
+      "grad_norm": 0.8933284878730774,
+      "learning_rate": 3.5891561084359535e-05,
+      "loss": 2.6707,
+      "step": 7243
+    },
+    {
+      "epoch": 0.6557436408074591,
+      "grad_norm": 0.9218231439590454,
+      "learning_rate": 3.5867209975735625e-05,
+      "loss": 2.6409,
+      "step": 7244
+    },
+    {
+      "epoch": 0.6558341631212093,
+      "grad_norm": 0.9200209379196167,
+      "learning_rate": 3.584286532528184e-05,
+      "loss": 2.6412,
+      "step": 7245
+    },
+    {
+      "epoch": 0.6559246854349597,
+      "grad_norm": 0.8185088634490967,
+      "learning_rate": 3.581852713544983e-05,
+      "loss": 2.6865,
+      "step": 7246
+    },
+    {
+      "epoch": 0.65601520774871,
+      "grad_norm": 0.8688246011734009,
+      "learning_rate": 3.579419540869036e-05,
+      "loss": 2.7028,
+      "step": 7247
+    },
+    {
+      "epoch": 0.6561057300624604,
+      "grad_norm": 0.9082852602005005,
+      "learning_rate": 3.5769870147453646e-05,
+      "loss": 2.6682,
+      "step": 7248
+    },
+    {
+      "epoch": 0.6561962523762107,
+      "grad_norm": 0.9646705985069275,
+      "learning_rate": 3.5745551354189244e-05,
+      "loss": 2.7612,
+      "step": 7249
+    },
+    {
+      "epoch": 0.6562867746899611,
+      "grad_norm": 0.9219226241111755,
+      "learning_rate": 3.5721239031346066e-05,
+      "loss": 2.573,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6563772970037114,
+      "grad_norm": 0.8789383769035339,
+      "learning_rate": 3.5696933181372364e-05,
+      "loss": 2.7032,
+      "step": 7251
+    },
+    {
+      "epoch": 0.6564678193174618,
+      "grad_norm": 0.9892781376838684,
+      "learning_rate": 3.567263380671576e-05,
+      "loss": 2.6416,
+      "step": 7252
+    },
+    {
+      "epoch": 0.6565583416312121,
+      "grad_norm": 0.8261721730232239,
+      "learning_rate": 3.5648340909823105e-05,
+      "loss": 2.6554,
+      "step": 7253
+    },
+    {
+      "epoch": 0.6566488639449625,
+      "grad_norm": 0.8541987538337708,
+      "learning_rate": 3.562405449314084e-05,
+      "loss": 2.7041,
+      "step": 7254
+    },
+    {
+      "epoch": 0.6567393862587128,
+      "grad_norm": 0.886324942111969,
+      "learning_rate": 3.5599774559114475e-05,
+      "loss": 2.6688,
+      "step": 7255
+    },
+    {
+      "epoch": 0.6568299085724632,
+      "grad_norm": 0.891233503818512,
+      "learning_rate": 3.557550111018906e-05,
+      "loss": 2.703,
+      "step": 7256
+    },
+    {
+      "epoch": 0.6569204308862134,
+      "grad_norm": 0.8668513894081116,
+      "learning_rate": 3.555123414880891e-05,
+      "loss": 2.6485,
+      "step": 7257
+    },
+    {
+      "epoch": 0.6570109531999638,
+      "grad_norm": 0.8383379578590393,
+      "learning_rate": 3.552697367741772e-05,
+      "loss": 2.6134,
+      "step": 7258
+    },
+    {
+      "epoch": 0.6571014755137141,
+      "grad_norm": 0.8313611149787903,
+      "learning_rate": 3.5502719698458517e-05,
+      "loss": 2.6524,
+      "step": 7259
+    },
+    {
+      "epoch": 0.6571919978274645,
+      "grad_norm": 0.9203676581382751,
+      "learning_rate": 3.547847221437372e-05,
+      "loss": 2.7042,
+      "step": 7260
+    },
+    {
+      "epoch": 0.6572825201412148,
+      "grad_norm": 0.8398598432540894,
+      "learning_rate": 3.545423122760493e-05,
+      "loss": 2.6851,
+      "step": 7261
+    },
+    {
+      "epoch": 0.6573730424549652,
+      "grad_norm": 1.0203299522399902,
+      "learning_rate": 3.542999674059335e-05,
+      "loss": 2.6645,
+      "step": 7262
+    },
+    {
+      "epoch": 0.6574635647687155,
+      "grad_norm": 0.8423504829406738,
+      "learning_rate": 3.5405768755779315e-05,
+      "loss": 2.6734,
+      "step": 7263
+    },
+    {
+      "epoch": 0.6575540870824659,
+      "grad_norm": 0.8030686378479004,
+      "learning_rate": 3.538154727560259e-05,
+      "loss": 2.6363,
+      "step": 7264
+    },
+    {
+      "epoch": 0.6576446093962162,
+      "grad_norm": 0.8241353034973145,
+      "learning_rate": 3.5357332302502274e-05,
+      "loss": 2.6287,
+      "step": 7265
+    },
+    {
+      "epoch": 0.6577351317099666,
+      "grad_norm": 0.8681092262268066,
+      "learning_rate": 3.5333123838916845e-05,
+      "loss": 2.6093,
+      "step": 7266
+    },
+    {
+      "epoch": 0.6578256540237168,
+      "grad_norm": 0.8543620705604553,
+      "learning_rate": 3.530892188728406e-05,
+      "loss": 2.6196,
+      "step": 7267
+    },
+    {
+      "epoch": 0.6579161763374672,
+      "grad_norm": 0.8001599311828613,
+      "learning_rate": 3.5284726450041116e-05,
+      "loss": 2.6523,
+      "step": 7268
+    },
+    {
+      "epoch": 0.6580066986512175,
+      "grad_norm": 0.8121172785758972,
+      "learning_rate": 3.526053752962441e-05,
+      "loss": 2.6775,
+      "step": 7269
+    },
+    {
+      "epoch": 0.6580972209649678,
+      "grad_norm": 0.8860118389129639,
+      "learning_rate": 3.523635512846981e-05,
+      "loss": 2.6323,
+      "step": 7270
+    },
+    {
+      "epoch": 0.6581877432787182,
+      "grad_norm": 0.937731921672821,
+      "learning_rate": 3.521217924901248e-05,
+      "loss": 2.675,
+      "step": 7271
+    },
+    {
+      "epoch": 0.6582782655924685,
+      "grad_norm": 0.868818461894989,
+      "learning_rate": 3.518800989368691e-05,
+      "loss": 2.6533,
+      "step": 7272
+    },
+    {
+      "epoch": 0.6583687879062189,
+      "grad_norm": 0.8203585147857666,
+      "learning_rate": 3.516384706492701e-05,
+      "loss": 2.6489,
+      "step": 7273
+    },
+    {
+      "epoch": 0.6584593102199692,
+      "grad_norm": 0.9456077814102173,
+      "learning_rate": 3.5139690765165876e-05,
+      "loss": 2.7232,
+      "step": 7274
+    },
+    {
+      "epoch": 0.6585498325337196,
+      "grad_norm": 0.8109578490257263,
+      "learning_rate": 3.5115540996836174e-05,
+      "loss": 2.6335,
+      "step": 7275
+    },
+    {
+      "epoch": 0.6586403548474699,
+      "grad_norm": 0.8843202590942383,
+      "learning_rate": 3.509139776236967e-05,
+      "loss": 2.6194,
+      "step": 7276
+    },
+    {
+      "epoch": 0.6587308771612203,
+      "grad_norm": 0.9179278612136841,
+      "learning_rate": 3.506726106419765e-05,
+      "loss": 2.6024,
+      "step": 7277
+    },
+    {
+      "epoch": 0.6588213994749705,
+      "grad_norm": 0.9402636289596558,
+      "learning_rate": 3.5043130904750665e-05,
+      "loss": 2.6987,
+      "step": 7278
+    },
+    {
+      "epoch": 0.6589119217887209,
+      "grad_norm": 0.8278437852859497,
+      "learning_rate": 3.501900728645865e-05,
+      "loss": 2.624,
+      "step": 7279
+    },
+    {
+      "epoch": 0.6590024441024712,
+      "grad_norm": 0.8150365352630615,
+      "learning_rate": 3.4994890211750754e-05,
+      "loss": 2.6596,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6590929664162216,
+      "grad_norm": 0.9417485594749451,
+      "learning_rate": 3.49707796830557e-05,
+      "loss": 2.6667,
+      "step": 7281
+    },
+    {
+      "epoch": 0.6591834887299719,
+      "grad_norm": 0.8826394081115723,
+      "learning_rate": 3.494667570280132e-05,
+      "loss": 2.6328,
+      "step": 7282
+    },
+    {
+      "epoch": 0.6592740110437223,
+      "grad_norm": 0.8603037595748901,
+      "learning_rate": 3.492257827341492e-05,
+      "loss": 2.727,
+      "step": 7283
+    },
+    {
+      "epoch": 0.6593645333574726,
+      "grad_norm": 0.8361364603042603,
+      "learning_rate": 3.489848739732311e-05,
+      "loss": 2.6093,
+      "step": 7284
+    },
+    {
+      "epoch": 0.659455055671223,
+      "grad_norm": 0.8012417554855347,
+      "learning_rate": 3.487440307695183e-05,
+      "loss": 2.6427,
+      "step": 7285
+    },
+    {
+      "epoch": 0.6595455779849733,
+      "grad_norm": 0.8624984622001648,
+      "learning_rate": 3.485032531472638e-05,
+      "loss": 2.6505,
+      "step": 7286
+    },
+    {
+      "epoch": 0.6596361002987237,
+      "grad_norm": 0.9410037398338318,
+      "learning_rate": 3.4826254113071434e-05,
+      "loss": 2.637,
+      "step": 7287
+    },
+    {
+      "epoch": 0.659726622612474,
+      "grad_norm": 0.862177848815918,
+      "learning_rate": 3.480218947441083e-05,
+      "loss": 2.6587,
+      "step": 7288
+    },
+    {
+      "epoch": 0.6598171449262243,
+      "grad_norm": 0.82010418176651,
+      "learning_rate": 3.477813140116805e-05,
+      "loss": 2.6426,
+      "step": 7289
+    },
+    {
+      "epoch": 0.6599076672399746,
+      "grad_norm": 0.9039086699485779,
+      "learning_rate": 3.47540798957656e-05,
+      "loss": 2.6614,
+      "step": 7290
+    },
+    {
+      "epoch": 0.659998189553725,
+      "grad_norm": 0.8869426250457764,
+      "learning_rate": 3.473003496062552e-05,
+      "loss": 2.5854,
+      "step": 7291
+    },
+    {
+      "epoch": 0.6600887118674753,
+      "grad_norm": 0.8362395763397217,
+      "learning_rate": 3.470599659816914e-05,
+      "loss": 2.6311,
+      "step": 7292
+    },
+    {
+      "epoch": 0.6601792341812257,
+      "grad_norm": 0.8808051943778992,
+      "learning_rate": 3.4681964810817124e-05,
+      "loss": 2.6611,
+      "step": 7293
+    },
+    {
+      "epoch": 0.660269756494976,
+      "grad_norm": 0.8643760085105896,
+      "learning_rate": 3.465793960098945e-05,
+      "loss": 2.6705,
+      "step": 7294
+    },
+    {
+      "epoch": 0.6603602788087264,
+      "grad_norm": 0.8985621333122253,
+      "learning_rate": 3.4633920971105515e-05,
+      "loss": 2.6765,
+      "step": 7295
+    },
+    {
+      "epoch": 0.6604508011224767,
+      "grad_norm": 0.9751450419425964,
+      "learning_rate": 3.460990892358388e-05,
+      "loss": 2.6361,
+      "step": 7296
+    },
+    {
+      "epoch": 0.6605413234362271,
+      "grad_norm": 0.8282762169837952,
+      "learning_rate": 3.458590346084269e-05,
+      "loss": 2.6246,
+      "step": 7297
+    },
+    {
+      "epoch": 0.6606318457499774,
+      "grad_norm": 0.8377431631088257,
+      "learning_rate": 3.45619045852992e-05,
+      "loss": 2.6649,
+      "step": 7298
+    },
+    {
+      "epoch": 0.6607223680637277,
+      "grad_norm": 0.9260903596878052,
+      "learning_rate": 3.4537912299370134e-05,
+      "loss": 2.7063,
+      "step": 7299
+    },
+    {
+      "epoch": 0.660812890377478,
+      "grad_norm": 0.9289246797561646,
+      "learning_rate": 3.45139266054715e-05,
+      "loss": 2.7076,
+      "step": 7300
+    },
+    {
+      "epoch": 0.6609034126912284,
+      "grad_norm": 0.8479653596878052,
+      "learning_rate": 3.4489947506018674e-05,
+      "loss": 2.6206,
+      "step": 7301
+    },
+    {
+      "epoch": 0.6609939350049787,
+      "grad_norm": 0.8131391406059265,
+      "learning_rate": 3.446597500342633e-05,
+      "loss": 2.6383,
+      "step": 7302
+    },
+    {
+      "epoch": 0.6610844573187291,
+      "grad_norm": 0.8845452666282654,
+      "learning_rate": 3.444200910010854e-05,
+      "loss": 2.6501,
+      "step": 7303
+    },
+    {
+      "epoch": 0.6611749796324794,
+      "grad_norm": 0.8882754445075989,
+      "learning_rate": 3.441804979847859e-05,
+      "loss": 2.6433,
+      "step": 7304
+    },
+    {
+      "epoch": 0.6612655019462298,
+      "grad_norm": 0.8090992569923401,
+      "learning_rate": 3.439409710094929e-05,
+      "loss": 2.6514,
+      "step": 7305
+    },
+    {
+      "epoch": 0.6613560242599801,
+      "grad_norm": 0.9277315735816956,
+      "learning_rate": 3.4370151009932584e-05,
+      "loss": 2.6344,
+      "step": 7306
+    },
+    {
+      "epoch": 0.6614465465737305,
+      "grad_norm": 0.8132951259613037,
+      "learning_rate": 3.434621152783987e-05,
+      "loss": 2.6609,
+      "step": 7307
+    },
+    {
+      "epoch": 0.6615370688874808,
+      "grad_norm": 0.8669840097427368,
+      "learning_rate": 3.4322278657081867e-05,
+      "loss": 2.6665,
+      "step": 7308
+    },
+    {
+      "epoch": 0.6616275912012312,
+      "grad_norm": 0.8841760754585266,
+      "learning_rate": 3.4298352400068635e-05,
+      "loss": 2.6327,
+      "step": 7309
+    },
+    {
+      "epoch": 0.6617181135149814,
+      "grad_norm": 0.8753495812416077,
+      "learning_rate": 3.4274432759209453e-05,
+      "loss": 2.6814,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6618086358287317,
+      "grad_norm": 0.8575315475463867,
+      "learning_rate": 3.4250519736913154e-05,
+      "loss": 2.6455,
+      "step": 7311
+    },
+    {
+      "epoch": 0.6618991581424821,
+      "grad_norm": 0.874495804309845,
+      "learning_rate": 3.4226613335587695e-05,
+      "loss": 2.6424,
+      "step": 7312
+    },
+    {
+      "epoch": 0.6619896804562324,
+      "grad_norm": 0.8438037633895874,
+      "learning_rate": 3.420271355764047e-05,
+      "loss": 2.7271,
+      "step": 7313
+    },
+    {
+      "epoch": 0.6620802027699828,
+      "grad_norm": 0.8370341658592224,
+      "learning_rate": 3.417882040547818e-05,
+      "loss": 2.7052,
+      "step": 7314
+    },
+    {
+      "epoch": 0.6621707250837331,
+      "grad_norm": 0.8512766361236572,
+      "learning_rate": 3.415493388150689e-05,
+      "loss": 2.6608,
+      "step": 7315
+    },
+    {
+      "epoch": 0.6622612473974835,
+      "grad_norm": 0.804831326007843,
+      "learning_rate": 3.413105398813195e-05,
+      "loss": 2.7129,
+      "step": 7316
+    },
+    {
+      "epoch": 0.6623517697112338,
+      "grad_norm": 0.8848623633384705,
+      "learning_rate": 3.41071807277581e-05,
+      "loss": 2.6712,
+      "step": 7317
+    },
+    {
+      "epoch": 0.6624422920249842,
+      "grad_norm": 0.8708617687225342,
+      "learning_rate": 3.408331410278929e-05,
+      "loss": 2.6331,
+      "step": 7318
+    },
+    {
+      "epoch": 0.6625328143387345,
+      "grad_norm": 0.8301901817321777,
+      "learning_rate": 3.4059454115629005e-05,
+      "loss": 2.6331,
+      "step": 7319
+    },
+    {
+      "epoch": 0.6626233366524849,
+      "grad_norm": 0.8876309394836426,
+      "learning_rate": 3.4035600768679855e-05,
+      "loss": 2.7056,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6627138589662351,
+      "grad_norm": 0.8508949875831604,
+      "learning_rate": 3.40117540643439e-05,
+      "loss": 2.7127,
+      "step": 7321
+    },
+    {
+      "epoch": 0.6628043812799855,
+      "grad_norm": 0.8443604707717896,
+      "learning_rate": 3.398791400502251e-05,
+      "loss": 2.6863,
+      "step": 7322
+    },
+    {
+      "epoch": 0.6628949035937358,
+      "grad_norm": 0.8200173377990723,
+      "learning_rate": 3.396408059311638e-05,
+      "loss": 2.6272,
+      "step": 7323
+    },
+    {
+      "epoch": 0.6629854259074862,
+      "grad_norm": 0.8505333662033081,
+      "learning_rate": 3.394025383102552e-05,
+      "loss": 2.6465,
+      "step": 7324
+    },
+    {
+      "epoch": 0.6630759482212365,
+      "grad_norm": 0.8726922273635864,
+      "learning_rate": 3.391643372114932e-05,
+      "loss": 2.6408,
+      "step": 7325
+    },
+    {
+      "epoch": 0.6631664705349869,
+      "grad_norm": 0.9224769473075867,
+      "learning_rate": 3.3892620265886376e-05,
+      "loss": 2.6597,
+      "step": 7326
+    },
+    {
+      "epoch": 0.6632569928487372,
+      "grad_norm": 0.9343586564064026,
+      "learning_rate": 3.386881346763483e-05,
+      "loss": 2.6887,
+      "step": 7327
+    },
+    {
+      "epoch": 0.6633475151624876,
+      "grad_norm": 0.8156290650367737,
+      "learning_rate": 3.384501332879192e-05,
+      "loss": 2.608,
+      "step": 7328
+    },
+    {
+      "epoch": 0.6634380374762379,
+      "grad_norm": 0.8800485730171204,
+      "learning_rate": 3.382121985175436e-05,
+      "loss": 2.7136,
+      "step": 7329
+    },
+    {
+      "epoch": 0.6635285597899883,
+      "grad_norm": 0.8803322315216064,
+      "learning_rate": 3.379743303891815e-05,
+      "loss": 2.6477,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6636190821037385,
+      "grad_norm": 0.8415948152542114,
+      "learning_rate": 3.377365289267862e-05,
+      "loss": 2.6389,
+      "step": 7331
+    },
+    {
+      "epoch": 0.6637096044174889,
+      "grad_norm": 0.9095073938369751,
+      "learning_rate": 3.374987941543043e-05,
+      "loss": 2.6926,
+      "step": 7332
+    },
+    {
+      "epoch": 0.6638001267312392,
+      "grad_norm": 0.8726220726966858,
+      "learning_rate": 3.372611260956761e-05,
+      "loss": 2.6237,
+      "step": 7333
+    },
+    {
+      "epoch": 0.6638906490449896,
+      "grad_norm": 0.8335959315299988,
+      "learning_rate": 3.370235247748337e-05,
+      "loss": 2.6272,
+      "step": 7334
+    },
+    {
+      "epoch": 0.6639811713587399,
+      "grad_norm": 0.7942869663238525,
+      "learning_rate": 3.367859902157048e-05,
+      "loss": 2.5965,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6640716936724903,
+      "grad_norm": 0.9050523042678833,
+      "learning_rate": 3.3654852244220826e-05,
+      "loss": 2.6389,
+      "step": 7336
+    },
+    {
+      "epoch": 0.6641622159862406,
+      "grad_norm": 0.8640265464782715,
+      "learning_rate": 3.363111214782574e-05,
+      "loss": 2.701,
+      "step": 7337
+    },
+    {
+      "epoch": 0.664252738299991,
+      "grad_norm": 0.8449823260307312,
+      "learning_rate": 3.360737873477584e-05,
+      "loss": 2.6787,
+      "step": 7338
+    },
+    {
+      "epoch": 0.6643432606137413,
+      "grad_norm": 0.885446310043335,
+      "learning_rate": 3.358365200746112e-05,
+      "loss": 2.6858,
+      "step": 7339
+    },
+    {
+      "epoch": 0.6644337829274917,
+      "grad_norm": 0.9015811085700989,
+      "learning_rate": 3.3559931968270753e-05,
+      "loss": 2.6176,
+      "step": 7340
+    },
+    {
+      "epoch": 0.664524305241242,
+      "grad_norm": 0.8146497011184692,
+      "learning_rate": 3.3536218619593496e-05,
+      "loss": 2.6218,
+      "step": 7341
+    },
+    {
+      "epoch": 0.6646148275549923,
+      "grad_norm": 0.7753192186355591,
+      "learning_rate": 3.351251196381716e-05,
+      "loss": 2.6111,
+      "step": 7342
+    },
+    {
+      "epoch": 0.6647053498687426,
+      "grad_norm": 0.8494303822517395,
+      "learning_rate": 3.348881200332905e-05,
+      "loss": 2.7045,
+      "step": 7343
+    },
+    {
+      "epoch": 0.664795872182493,
+      "grad_norm": 0.8519464731216431,
+      "learning_rate": 3.3465118740515764e-05,
+      "loss": 2.6458,
+      "step": 7344
+    },
+    {
+      "epoch": 0.6648863944962433,
+      "grad_norm": 0.7856410145759583,
+      "learning_rate": 3.344143217776319e-05,
+      "loss": 2.6726,
+      "step": 7345
+    },
+    {
+      "epoch": 0.6649769168099937,
+      "grad_norm": 0.8536087870597839,
+      "learning_rate": 3.3417752317456584e-05,
+      "loss": 2.6002,
+      "step": 7346
+    },
+    {
+      "epoch": 0.665067439123744,
+      "grad_norm": 0.8146572709083557,
+      "learning_rate": 3.339407916198052e-05,
+      "loss": 2.6429,
+      "step": 7347
+    },
+    {
+      "epoch": 0.6651579614374944,
+      "grad_norm": 0.8779633045196533,
+      "learning_rate": 3.33704127137188e-05,
+      "loss": 2.6807,
+      "step": 7348
+    },
+    {
+      "epoch": 0.6652484837512447,
+      "grad_norm": 0.8361524343490601,
+      "learning_rate": 3.334675297505476e-05,
+      "loss": 2.5803,
+      "step": 7349
+    },
+    {
+      "epoch": 0.6653390060649951,
+      "grad_norm": 0.8956053256988525,
+      "learning_rate": 3.332309994837085e-05,
+      "loss": 2.6896,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6654295283787454,
+      "grad_norm": 0.7889111638069153,
+      "learning_rate": 3.329945363604894e-05,
+      "loss": 2.6373,
+      "step": 7351
+    },
+    {
+      "epoch": 0.6655200506924956,
+      "grad_norm": 0.8257021307945251,
+      "learning_rate": 3.3275814040470234e-05,
+      "loss": 2.5832,
+      "step": 7352
+    },
+    {
+      "epoch": 0.665610573006246,
+      "grad_norm": 0.8282142877578735,
+      "learning_rate": 3.325218116401522e-05,
+      "loss": 2.65,
+      "step": 7353
+    },
+    {
+      "epoch": 0.6657010953199963,
+      "grad_norm": 0.8529660105705261,
+      "learning_rate": 3.322855500906373e-05,
+      "loss": 2.6685,
+      "step": 7354
+    },
+    {
+      "epoch": 0.6657916176337467,
+      "grad_norm": 0.9197294116020203,
+      "learning_rate": 3.320493557799497e-05,
+      "loss": 2.6648,
+      "step": 7355
+    },
+    {
+      "epoch": 0.665882139947497,
+      "grad_norm": 0.8231920003890991,
+      "learning_rate": 3.3181322873187326e-05,
+      "loss": 2.6398,
+      "step": 7356
+    },
+    {
+      "epoch": 0.6659726622612474,
+      "grad_norm": 0.8430575728416443,
+      "learning_rate": 3.3157716897018645e-05,
+      "loss": 2.6812,
+      "step": 7357
+    },
+    {
+      "epoch": 0.6660631845749977,
+      "grad_norm": 0.8932907581329346,
+      "learning_rate": 3.313411765186605e-05,
+      "loss": 2.6575,
+      "step": 7358
+    },
+    {
+      "epoch": 0.6661537068887481,
+      "grad_norm": 0.823968768119812,
+      "learning_rate": 3.311052514010596e-05,
+      "loss": 2.6438,
+      "step": 7359
+    },
+    {
+      "epoch": 0.6662442292024984,
+      "grad_norm": 0.8624204397201538,
+      "learning_rate": 3.308693936411421e-05,
+      "loss": 2.6686,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6663347515162488,
+      "grad_norm": 0.7924514412879944,
+      "learning_rate": 3.306336032626576e-05,
+      "loss": 2.6259,
+      "step": 7361
+    },
+    {
+      "epoch": 0.666425273829999,
+      "grad_norm": 0.843051552772522,
+      "learning_rate": 3.3039788028935156e-05,
+      "loss": 2.6942,
+      "step": 7362
+    },
+    {
+      "epoch": 0.6665157961437495,
+      "grad_norm": 0.8649163842201233,
+      "learning_rate": 3.301622247449604e-05,
+      "loss": 2.6264,
+      "step": 7363
+    },
+    {
+      "epoch": 0.6666063184574997,
+      "grad_norm": 0.8684132695198059,
+      "learning_rate": 3.299266366532149e-05,
+      "loss": 2.66,
+      "step": 7364
+    },
+    {
+      "epoch": 0.6666968407712501,
+      "grad_norm": 0.8766152858734131,
+      "learning_rate": 3.2969111603783875e-05,
+      "loss": 2.6308,
+      "step": 7365
+    },
+    {
+      "epoch": 0.6667873630850004,
+      "grad_norm": 0.9793428778648376,
+      "learning_rate": 3.294556629225488e-05,
+      "loss": 2.726,
+      "step": 7366
+    },
+    {
+      "epoch": 0.6668778853987508,
+      "grad_norm": 0.8449831008911133,
+      "learning_rate": 3.292202773310553e-05,
+      "loss": 2.6433,
+      "step": 7367
+    },
+    {
+      "epoch": 0.6669684077125011,
+      "grad_norm": 0.7769423723220825,
+      "learning_rate": 3.2898495928706185e-05,
+      "loss": 2.6087,
+      "step": 7368
+    },
+    {
+      "epoch": 0.6670589300262515,
+      "grad_norm": 0.8507257699966431,
+      "learning_rate": 3.2874970881426426e-05,
+      "loss": 2.6934,
+      "step": 7369
+    },
+    {
+      "epoch": 0.6671494523400018,
+      "grad_norm": 0.8462991714477539,
+      "learning_rate": 3.2851452593635266e-05,
+      "loss": 2.7214,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6672399746537522,
+      "grad_norm": 0.8643592000007629,
+      "learning_rate": 3.2827941067700996e-05,
+      "loss": 2.7112,
+      "step": 7371
+    },
+    {
+      "epoch": 0.6673304969675025,
+      "grad_norm": 0.9072657823562622,
+      "learning_rate": 3.2804436305991214e-05,
+      "loss": 2.6819,
+      "step": 7372
+    },
+    {
+      "epoch": 0.6674210192812529,
+      "grad_norm": 0.8652128577232361,
+      "learning_rate": 3.278093831087287e-05,
+      "loss": 2.6554,
+      "step": 7373
+    },
+    {
+      "epoch": 0.6675115415950031,
+      "grad_norm": 0.8562641739845276,
+      "learning_rate": 3.275744708471222e-05,
+      "loss": 2.6745,
+      "step": 7374
+    },
+    {
+      "epoch": 0.6676020639087535,
+      "grad_norm": 0.8770109415054321,
+      "learning_rate": 3.273396262987475e-05,
+      "loss": 2.6677,
+      "step": 7375
+    },
+    {
+      "epoch": 0.6676925862225038,
+      "grad_norm": 0.8314611315727234,
+      "learning_rate": 3.271048494872546e-05,
+      "loss": 2.6635,
+      "step": 7376
+    },
+    {
+      "epoch": 0.6677831085362542,
+      "grad_norm": 0.9163419604301453,
+      "learning_rate": 3.268701404362847e-05,
+      "loss": 2.634,
+      "step": 7377
+    },
+    {
+      "epoch": 0.6678736308500045,
+      "grad_norm": 0.8832092881202698,
+      "learning_rate": 3.266354991694732e-05,
+      "loss": 2.6737,
+      "step": 7378
+    },
+    {
+      "epoch": 0.6679641531637549,
+      "grad_norm": 0.9241970777511597,
+      "learning_rate": 3.264009257104486e-05,
+      "loss": 2.6664,
+      "step": 7379
+    },
+    {
+      "epoch": 0.6680546754775052,
+      "grad_norm": 0.9200668334960938,
+      "learning_rate": 3.2616642008283213e-05,
+      "loss": 2.6741,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6681451977912556,
+      "grad_norm": 0.8832535743713379,
+      "learning_rate": 3.259319823102389e-05,
+      "loss": 2.6622,
+      "step": 7381
+    },
+    {
+      "epoch": 0.6682357201050059,
+      "grad_norm": 0.8708347678184509,
+      "learning_rate": 3.2569761241627696e-05,
+      "loss": 2.6535,
+      "step": 7382
+    },
+    {
+      "epoch": 0.6683262424187563,
+      "grad_norm": 0.8515925407409668,
+      "learning_rate": 3.254633104245463e-05,
+      "loss": 2.6247,
+      "step": 7383
+    },
+    {
+      "epoch": 0.6684167647325066,
+      "grad_norm": 0.9260343909263611,
+      "learning_rate": 3.2522907635864244e-05,
+      "loss": 2.6386,
+      "step": 7384
+    },
+    {
+      "epoch": 0.668507287046257,
+      "grad_norm": 0.8982704877853394,
+      "learning_rate": 3.249949102421518e-05,
+      "loss": 2.697,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6685978093600072,
+      "grad_norm": 0.8327007293701172,
+      "learning_rate": 3.247608120986552e-05,
+      "loss": 2.6658,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6686883316737576,
+      "grad_norm": 0.8208869695663452,
+      "learning_rate": 3.2452678195172635e-05,
+      "loss": 2.6131,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6687788539875079,
+      "grad_norm": 0.830223023891449,
+      "learning_rate": 3.242928198249322e-05,
+      "loss": 2.6519,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6688693763012583,
+      "grad_norm": 0.8725573420524597,
+      "learning_rate": 3.240589257418325e-05,
+      "loss": 2.6519,
+      "step": 7389
+    },
+    {
+      "epoch": 0.6689598986150086,
+      "grad_norm": 0.8340583443641663,
+      "learning_rate": 3.238250997259808e-05,
+      "loss": 2.7087,
+      "step": 7390
+    },
+    {
+      "epoch": 0.669050420928759,
+      "grad_norm": 0.9096596837043762,
+      "learning_rate": 3.235913418009224e-05,
+      "loss": 2.6359,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6691409432425093,
+      "grad_norm": 0.7934027314186096,
+      "learning_rate": 3.233576519901981e-05,
+      "loss": 2.6322,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6692314655562596,
+      "grad_norm": 0.8696547746658325,
+      "learning_rate": 3.231240303173394e-05,
+      "loss": 2.639,
+      "step": 7393
+    },
+    {
+      "epoch": 0.66932198787001,
+      "grad_norm": 0.8256217837333679,
+      "learning_rate": 3.228904768058722e-05,
+      "loss": 2.619,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6694125101837602,
+      "grad_norm": 0.8816400170326233,
+      "learning_rate": 3.226569914793156e-05,
+      "loss": 2.6671,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6695030324975106,
+      "grad_norm": 0.8656291365623474,
+      "learning_rate": 3.224235743611814e-05,
+      "loss": 2.6264,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6695935548112609,
+      "grad_norm": 0.8678957223892212,
+      "learning_rate": 3.221902254749747e-05,
+      "loss": 2.7066,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6696840771250113,
+      "grad_norm": 0.870672345161438,
+      "learning_rate": 3.21956944844194e-05,
+      "loss": 2.6839,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6697745994387616,
+      "grad_norm": 0.8486281633377075,
+      "learning_rate": 3.217237324923299e-05,
+      "loss": 2.7099,
+      "step": 7399
+    },
+    {
+      "epoch": 0.669865121752512,
+      "grad_norm": 0.8596231937408447,
+      "learning_rate": 3.21490588442868e-05,
+      "loss": 2.6821,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6699556440662623,
+      "grad_norm": 0.8599048852920532,
+      "learning_rate": 3.2125751271928485e-05,
+      "loss": 2.6727,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6700461663800127,
+      "grad_norm": 0.8863310217857361,
+      "learning_rate": 3.210245053450517e-05,
+      "loss": 2.6249,
+      "step": 7402
+    },
+    {
+      "epoch": 0.670136688693763,
+      "grad_norm": 0.8507751822471619,
+      "learning_rate": 3.207915663436322e-05,
+      "loss": 2.6963,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6702272110075134,
+      "grad_norm": 0.8587643504142761,
+      "learning_rate": 3.205586957384838e-05,
+      "loss": 2.6832,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6703177333212637,
+      "grad_norm": 0.8385748267173767,
+      "learning_rate": 3.203258935530554e-05,
+      "loss": 2.6764,
+      "step": 7405
+    },
+    {
+      "epoch": 0.670408255635014,
+      "grad_norm": 0.8625391721725464,
+      "learning_rate": 3.200931598107916e-05,
+      "loss": 2.6947,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6704987779487643,
+      "grad_norm": 0.9186429977416992,
+      "learning_rate": 3.1986049453512766e-05,
+      "loss": 2.6733,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6705893002625147,
+      "grad_norm": 0.8060020804405212,
+      "learning_rate": 3.196278977494934e-05,
+      "loss": 2.6097,
+      "step": 7408
+    },
+    {
+      "epoch": 0.670679822576265,
+      "grad_norm": 0.8401286005973816,
+      "learning_rate": 3.193953694773112e-05,
+      "loss": 2.6956,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6707703448900154,
+      "grad_norm": 0.8545077443122864,
+      "learning_rate": 3.191629097419966e-05,
+      "loss": 2.6702,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6708608672037657,
+      "grad_norm": 0.9147730469703674,
+      "learning_rate": 3.1893051856695846e-05,
+      "loss": 2.6265,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6709513895175161,
+      "grad_norm": 0.8632221221923828,
+      "learning_rate": 3.186981959755987e-05,
+      "loss": 2.6475,
+      "step": 7412
+    },
+    {
+      "epoch": 0.6710419118312664,
+      "grad_norm": 0.846584141254425,
+      "learning_rate": 3.184659419913113e-05,
+      "loss": 2.6488,
+      "step": 7413
+    },
+    {
+      "epoch": 0.6711324341450168,
+      "grad_norm": 0.8977507948875427,
+      "learning_rate": 3.182337566374856e-05,
+      "loss": 2.5776,
+      "step": 7414
+    },
+    {
+      "epoch": 0.6712229564587671,
+      "grad_norm": 0.8261081576347351,
+      "learning_rate": 3.1800163993750166e-05,
+      "loss": 2.6674,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6713134787725175,
+      "grad_norm": 0.8828268647193909,
+      "learning_rate": 3.177695919147339e-05,
+      "loss": 2.6563,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6714040010862677,
+      "grad_norm": 0.8706898093223572,
+      "learning_rate": 3.1753761259254946e-05,
+      "loss": 2.651,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6714945234000181,
+      "grad_norm": 0.8818956613540649,
+      "learning_rate": 3.173057019943089e-05,
+      "loss": 2.6796,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6715850457137684,
+      "grad_norm": 0.8012150526046753,
+      "learning_rate": 3.1707386014336537e-05,
+      "loss": 2.6349,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6716755680275188,
+      "grad_norm": 0.8736044764518738,
+      "learning_rate": 3.1684208706306574e-05,
+      "loss": 2.6447,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6717660903412691,
+      "grad_norm": 0.8982002139091492,
+      "learning_rate": 3.166103827767487e-05,
+      "loss": 2.7035,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6718566126550195,
+      "grad_norm": 0.816577136516571,
+      "learning_rate": 3.1637874730774795e-05,
+      "loss": 2.6144,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6719471349687698,
+      "grad_norm": 0.8612717986106873,
+      "learning_rate": 3.161471806793884e-05,
+      "loss": 2.6686,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6720376572825202,
+      "grad_norm": 0.8321888446807861,
+      "learning_rate": 3.1591568291498904e-05,
+      "loss": 2.5793,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6721281795962705,
+      "grad_norm": 0.8054267168045044,
+      "learning_rate": 3.156842540378617e-05,
+      "loss": 2.6141,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6722187019100209,
+      "grad_norm": 0.8201277852058411,
+      "learning_rate": 3.154528940713113e-05,
+      "loss": 2.6211,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6723092242237712,
+      "grad_norm": 0.8517640233039856,
+      "learning_rate": 3.152216030386357e-05,
+      "loss": 2.6474,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6723997465375215,
+      "grad_norm": 0.9056238532066345,
+      "learning_rate": 3.1499038096312615e-05,
+      "loss": 2.6526,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6724902688512718,
+      "grad_norm": 0.9206411838531494,
+      "learning_rate": 3.14759227868066e-05,
+      "loss": 2.6644,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6725807911650222,
+      "grad_norm": 0.8964048027992249,
+      "learning_rate": 3.1452814377673346e-05,
+      "loss": 2.6055,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6726713134787725,
+      "grad_norm": 0.8579918146133423,
+      "learning_rate": 3.142971287123977e-05,
+      "loss": 2.6294,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6727618357925229,
+      "grad_norm": 0.8341345191001892,
+      "learning_rate": 3.140661826983223e-05,
+      "loss": 2.6064,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6728523581062732,
+      "grad_norm": 0.8820474147796631,
+      "learning_rate": 3.138353057577636e-05,
+      "loss": 2.6981,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6729428804200235,
+      "grad_norm": 0.8326157331466675,
+      "learning_rate": 3.136044979139712e-05,
+      "loss": 2.6532,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6730334027337739,
+      "grad_norm": 0.8076948523521423,
+      "learning_rate": 3.133737591901864e-05,
+      "loss": 2.6392,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6731239250475242,
+      "grad_norm": 0.809867799282074,
+      "learning_rate": 3.131430896096459e-05,
+      "loss": 2.6554,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6732144473612746,
+      "grad_norm": 0.8507243990898132,
+      "learning_rate": 3.129124891955771e-05,
+      "loss": 2.6328,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6733049696750248,
+      "grad_norm": 0.8950903415679932,
+      "learning_rate": 3.1268195797120195e-05,
+      "loss": 2.7115,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6733954919887752,
+      "grad_norm": 0.9369960427284241,
+      "learning_rate": 3.124514959597346e-05,
+      "loss": 2.6328,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6734860143025255,
+      "grad_norm": 0.8768920302391052,
+      "learning_rate": 3.1222110318438304e-05,
+      "loss": 2.6721,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6735765366162759,
+      "grad_norm": 0.8693579435348511,
+      "learning_rate": 3.119907796683479e-05,
+      "loss": 2.6859,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6736670589300262,
+      "grad_norm": 0.847797155380249,
+      "learning_rate": 3.1176052543482194e-05,
+      "loss": 2.6447,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6737575812437766,
+      "grad_norm": 0.819843053817749,
+      "learning_rate": 3.115303405069922e-05,
+      "loss": 2.6963,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6738481035575269,
+      "grad_norm": 0.8983978033065796,
+      "learning_rate": 3.113002249080386e-05,
+      "loss": 2.6088,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6739386258712773,
+      "grad_norm": 0.8833874464035034,
+      "learning_rate": 3.110701786611333e-05,
+      "loss": 2.6925,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6740291481850276,
+      "grad_norm": 0.7997896075248718,
+      "learning_rate": 3.108402017894422e-05,
+      "loss": 2.5894,
+      "step": 7446
+    },
+    {
+      "epoch": 0.674119670498778,
+      "grad_norm": 0.867850661277771,
+      "learning_rate": 3.106102943161242e-05,
+      "loss": 2.5973,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6742101928125283,
+      "grad_norm": 0.7908462882041931,
+      "learning_rate": 3.103804562643302e-05,
+      "loss": 2.6414,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6743007151262786,
+      "grad_norm": 0.8645251393318176,
+      "learning_rate": 3.101506876572059e-05,
+      "loss": 2.6771,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6743912374400289,
+      "grad_norm": 0.8117453455924988,
+      "learning_rate": 3.099209885178882e-05,
+      "loss": 2.6452,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6744817597537793,
+      "grad_norm": 0.8113788962364197,
+      "learning_rate": 3.096913588695081e-05,
+      "loss": 2.6444,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6745722820675296,
+      "grad_norm": 0.871590256690979,
+      "learning_rate": 3.094617987351892e-05,
+      "loss": 2.6551,
+      "step": 7452
+    },
+    {
+      "epoch": 0.67466280438128,
+      "grad_norm": 0.8343399167060852,
+      "learning_rate": 3.0923230813804825e-05,
+      "loss": 2.6521,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6747533266950303,
+      "grad_norm": 0.9126983880996704,
+      "learning_rate": 3.0900288710119504e-05,
+      "loss": 2.6849,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6748438490087807,
+      "grad_norm": 0.8040212392807007,
+      "learning_rate": 3.087735356477326e-05,
+      "loss": 2.6112,
+      "step": 7455
+    },
+    {
+      "epoch": 0.674934371322531,
+      "grad_norm": 0.8733232021331787,
+      "learning_rate": 3.0854425380075544e-05,
+      "loss": 2.6742,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6750248936362814,
+      "grad_norm": 0.8318789601325989,
+      "learning_rate": 3.083150415833537e-05,
+      "loss": 2.6729,
+      "step": 7457
+    },
+    {
+      "epoch": 0.6751154159500317,
+      "grad_norm": 0.884968638420105,
+      "learning_rate": 3.08085899018608e-05,
+      "loss": 2.6462,
+      "step": 7458
+    },
+    {
+      "epoch": 0.6752059382637821,
+      "grad_norm": 0.7779831886291504,
+      "learning_rate": 3.078568261295933e-05,
+      "loss": 2.6402,
+      "step": 7459
+    },
+    {
+      "epoch": 0.6752964605775323,
+      "grad_norm": 0.8294646143913269,
+      "learning_rate": 3.076278229393773e-05,
+      "loss": 2.6428,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6753869828912827,
+      "grad_norm": 0.8570337295532227,
+      "learning_rate": 3.073988894710209e-05,
+      "loss": 2.6938,
+      "step": 7461
+    },
+    {
+      "epoch": 0.675477505205033,
+      "grad_norm": 0.87303227186203,
+      "learning_rate": 3.071700257475768e-05,
+      "loss": 2.7059,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6755680275187834,
+      "grad_norm": 0.808883011341095,
+      "learning_rate": 3.0694123179209266e-05,
+      "loss": 2.6511,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6756585498325337,
+      "grad_norm": 0.9272127151489258,
+      "learning_rate": 3.067125076276073e-05,
+      "loss": 2.6945,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6757490721462841,
+      "grad_norm": 0.974225640296936,
+      "learning_rate": 3.064838532771535e-05,
+      "loss": 2.6187,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6758395944600344,
+      "grad_norm": 0.8074182868003845,
+      "learning_rate": 3.062552687637566e-05,
+      "loss": 2.6598,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6759301167737848,
+      "grad_norm": 0.8622375130653381,
+      "learning_rate": 3.0602675411043516e-05,
+      "loss": 2.6804,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6760206390875351,
+      "grad_norm": 0.8563747406005859,
+      "learning_rate": 3.0579830934020057e-05,
+      "loss": 2.6641,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6761111614012855,
+      "grad_norm": 0.8611782193183899,
+      "learning_rate": 3.0556993447605765e-05,
+      "loss": 2.653,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6762016837150358,
+      "grad_norm": 0.8958120942115784,
+      "learning_rate": 3.053416295410026e-05,
+      "loss": 2.6745,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6762922060287861,
+      "grad_norm": 0.883176326751709,
+      "learning_rate": 3.0511339455802724e-05,
+      "loss": 2.6106,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6763827283425364,
+      "grad_norm": 0.8690735101699829,
+      "learning_rate": 3.0488522955011357e-05,
+      "loss": 2.6656,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6764732506562868,
+      "grad_norm": 0.8291492462158203,
+      "learning_rate": 3.0465713454023826e-05,
+      "loss": 2.6225,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6765637729700371,
+      "grad_norm": 0.7962046265602112,
+      "learning_rate": 3.044291095513705e-05,
+      "loss": 2.6375,
+      "step": 7474
+    },
+    {
+      "epoch": 0.6766542952837874,
+      "grad_norm": 0.9379552602767944,
+      "learning_rate": 3.0420115460647235e-05,
+      "loss": 2.7519,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6767448175975378,
+      "grad_norm": 0.8300173878669739,
+      "learning_rate": 3.0397326972849892e-05,
+      "loss": 2.6709,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6768353399112881,
+      "grad_norm": 0.9035888314247131,
+      "learning_rate": 3.0374545494039852e-05,
+      "loss": 2.6305,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6769258622250385,
+      "grad_norm": 0.8968276381492615,
+      "learning_rate": 3.0351771026511122e-05,
+      "loss": 2.6216,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6770163845387888,
+      "grad_norm": 0.847926139831543,
+      "learning_rate": 3.032900357255719e-05,
+      "loss": 2.6709,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6771069068525392,
+      "grad_norm": 1.1153925657272339,
+      "learning_rate": 3.030624313447067e-05,
+      "loss": 2.6113,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6771974291662894,
+      "grad_norm": 0.9465881586074829,
+      "learning_rate": 3.0283489714543556e-05,
+      "loss": 2.6132,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6772879514800398,
+      "grad_norm": 0.8538044691085815,
+      "learning_rate": 3.026074331506712e-05,
+      "loss": 2.6816,
+      "step": 7482
+    },
+    {
+      "epoch": 0.6773784737937901,
+      "grad_norm": 0.8334993720054626,
+      "learning_rate": 3.0238003938331927e-05,
+      "loss": 2.655,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6774689961075405,
+      "grad_norm": 1.0370806455612183,
+      "learning_rate": 3.0215271586627825e-05,
+      "loss": 2.6343,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6775595184212908,
+      "grad_norm": 0.8620277047157288,
+      "learning_rate": 3.019254626224399e-05,
+      "loss": 2.6227,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6776500407350412,
+      "grad_norm": 0.8354658484458923,
+      "learning_rate": 3.016982796746879e-05,
+      "loss": 2.6554,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6777405630487915,
+      "grad_norm": 0.8558822274208069,
+      "learning_rate": 3.014711670459005e-05,
+      "loss": 2.643,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6778310853625419,
+      "grad_norm": 0.8511578440666199,
+      "learning_rate": 3.0124412475894726e-05,
+      "loss": 2.6077,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6779216076762922,
+      "grad_norm": 0.9501864910125732,
+      "learning_rate": 3.0101715283669153e-05,
+      "loss": 2.6687,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6780121299900426,
+      "grad_norm": 0.838162362575531,
+      "learning_rate": 3.0079025130198935e-05,
+      "loss": 2.6471,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6781026523037929,
+      "grad_norm": 0.837769091129303,
+      "learning_rate": 3.005634201776897e-05,
+      "loss": 2.6728,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6781931746175432,
+      "grad_norm": 0.8764926195144653,
+      "learning_rate": 3.0033665948663448e-05,
+      "loss": 2.67,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6782836969312935,
+      "grad_norm": 0.9114204049110413,
+      "learning_rate": 3.0010996925165892e-05,
+      "loss": 2.6779,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6783742192450439,
+      "grad_norm": 0.8063619136810303,
+      "learning_rate": 2.998833494955896e-05,
+      "loss": 2.5825,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6784647415587942,
+      "grad_norm": 0.8676156401634216,
+      "learning_rate": 2.9965680024124852e-05,
+      "loss": 2.6501,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6785552638725446,
+      "grad_norm": 0.8409841060638428,
+      "learning_rate": 2.9943032151144812e-05,
+      "loss": 2.6461,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6786457861862949,
+      "grad_norm": 0.8916147947311401,
+      "learning_rate": 2.992039133289952e-05,
+      "loss": 2.6078,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6787363085000453,
+      "grad_norm": 0.8280127644538879,
+      "learning_rate": 2.9897757571668905e-05,
+      "loss": 2.6796,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6788268308137956,
+      "grad_norm": 0.881831705570221,
+      "learning_rate": 2.9875130869732214e-05,
+      "loss": 2.6024,
+      "step": 7499
+    },
+    {
+      "epoch": 0.678917353127546,
+      "grad_norm": 0.8717087507247925,
+      "learning_rate": 2.9852511229367865e-05,
+      "loss": 2.6138,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6790078754412963,
+      "grad_norm": 0.8997431397438049,
+      "learning_rate": 2.9829898652853773e-05,
+      "loss": 2.6541,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6790983977550467,
+      "grad_norm": 0.9234827756881714,
+      "learning_rate": 2.980729314246694e-05,
+      "loss": 2.672,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6791889200687969,
+      "grad_norm": 0.8099880814552307,
+      "learning_rate": 2.9784694700483762e-05,
+      "loss": 2.6478,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6792794423825473,
+      "grad_norm": 0.8408997058868408,
+      "learning_rate": 2.9762103329179913e-05,
+      "loss": 2.6768,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6793699646962976,
+      "grad_norm": 0.8334234952926636,
+      "learning_rate": 2.9739519030830333e-05,
+      "loss": 2.6251,
+      "step": 7505
+    },
+    {
+      "epoch": 0.679460487010048,
+      "grad_norm": 0.8718681335449219,
+      "learning_rate": 2.9716941807709265e-05,
+      "loss": 2.6532,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6795510093237983,
+      "grad_norm": 0.9016917943954468,
+      "learning_rate": 2.969437166209027e-05,
+      "loss": 2.6318,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6796415316375487,
+      "grad_norm": 0.9001185297966003,
+      "learning_rate": 2.9671808596246053e-05,
+      "loss": 2.6418,
+      "step": 7508
+    },
+    {
+      "epoch": 0.679732053951299,
+      "grad_norm": 0.8468513488769531,
+      "learning_rate": 2.9649252612448863e-05,
+      "loss": 2.6544,
+      "step": 7509
+    },
+    {
+      "epoch": 0.6798225762650494,
+      "grad_norm": 0.8203318119049072,
+      "learning_rate": 2.962670371296996e-05,
+      "loss": 2.644,
+      "step": 7510
+    },
+    {
+      "epoch": 0.6799130985787997,
+      "grad_norm": 0.8951826691627502,
+      "learning_rate": 2.9604161900080087e-05,
+      "loss": 2.6286,
+      "step": 7511
+    },
+    {
+      "epoch": 0.6800036208925501,
+      "grad_norm": 0.8559239506721497,
+      "learning_rate": 2.9581627176049166e-05,
+      "loss": 2.6228,
+      "step": 7512
+    },
+    {
+      "epoch": 0.6800941432063004,
+      "grad_norm": 0.880177915096283,
+      "learning_rate": 2.9559099543146474e-05,
+      "loss": 2.6389,
+      "step": 7513
+    },
+    {
+      "epoch": 0.6801846655200507,
+      "grad_norm": 0.9628912210464478,
+      "learning_rate": 2.953657900364053e-05,
+      "loss": 2.6487,
+      "step": 7514
+    },
+    {
+      "epoch": 0.680275187833801,
+      "grad_norm": 0.7840589284896851,
+      "learning_rate": 2.9514065559799176e-05,
+      "loss": 2.642,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6803657101475513,
+      "grad_norm": 0.9465866088867188,
+      "learning_rate": 2.949155921388943e-05,
+      "loss": 2.6898,
+      "step": 7516
+    },
+    {
+      "epoch": 0.6804562324613017,
+      "grad_norm": 0.8815156817436218,
+      "learning_rate": 2.9469059968177815e-05,
+      "loss": 2.699,
+      "step": 7517
+    },
+    {
+      "epoch": 0.680546754775052,
+      "grad_norm": 0.9435743093490601,
+      "learning_rate": 2.9446567824929883e-05,
+      "loss": 2.6925,
+      "step": 7518
+    },
+    {
+      "epoch": 0.6806372770888024,
+      "grad_norm": 0.8814501166343689,
+      "learning_rate": 2.9424082786410656e-05,
+      "loss": 2.7095,
+      "step": 7519
+    },
+    {
+      "epoch": 0.6807277994025527,
+      "grad_norm": 0.7933807969093323,
+      "learning_rate": 2.9401604854884357e-05,
+      "loss": 2.6263,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6808183217163031,
+      "grad_norm": 0.8809409141540527,
+      "learning_rate": 2.9379134032614507e-05,
+      "loss": 2.6298,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6809088440300534,
+      "grad_norm": 0.8568894267082214,
+      "learning_rate": 2.9356670321863942e-05,
+      "loss": 2.6771,
+      "step": 7522
+    },
+    {
+      "epoch": 0.6809993663438038,
+      "grad_norm": 0.8187510371208191,
+      "learning_rate": 2.9334213724894767e-05,
+      "loss": 2.6327,
+      "step": 7523
+    },
+    {
+      "epoch": 0.681089888657554,
+      "grad_norm": 0.8523328900337219,
+      "learning_rate": 2.9311764243968275e-05,
+      "loss": 2.6121,
+      "step": 7524
+    },
+    {
+      "epoch": 0.6811804109713044,
+      "grad_norm": 0.9264262914657593,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 2.7426,
+      "step": 7525
+    },
+    {
+      "epoch": 0.6812709332850547,
+      "grad_norm": 0.9365399479866028,
+      "learning_rate": 2.9266886639285585e-05,
+      "loss": 2.6904,
+      "step": 7526
+    },
+    {
+      "epoch": 0.6813614555988051,
+      "grad_norm": 0.8703701496124268,
+      "learning_rate": 2.9244458520048434e-05,
+      "loss": 2.6645,
+      "step": 7527
+    },
+    {
+      "epoch": 0.6814519779125554,
+      "grad_norm": 0.8455237150192261,
+      "learning_rate": 2.922203752589243e-05,
+      "loss": 2.602,
+      "step": 7528
+    },
+    {
+      "epoch": 0.6815425002263058,
+      "grad_norm": 0.8652321696281433,
+      "learning_rate": 2.9199623659075283e-05,
+      "loss": 2.6605,
+      "step": 7529
+    },
+    {
+      "epoch": 0.6816330225400561,
+      "grad_norm": 0.8773972988128662,
+      "learning_rate": 2.91772169218541e-05,
+      "loss": 2.6237,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6817235448538065,
+      "grad_norm": 0.8873440623283386,
+      "learning_rate": 2.915481731648523e-05,
+      "loss": 2.6625,
+      "step": 7531
+    },
+    {
+      "epoch": 0.6818140671675568,
+      "grad_norm": 0.7923961877822876,
+      "learning_rate": 2.9132424845224317e-05,
+      "loss": 2.681,
+      "step": 7532
+    },
+    {
+      "epoch": 0.6819045894813072,
+      "grad_norm": 0.8112161755561829,
+      "learning_rate": 2.9110039510326293e-05,
+      "loss": 2.5817,
+      "step": 7533
+    },
+    {
+      "epoch": 0.6819951117950575,
+      "grad_norm": 0.874810516834259,
+      "learning_rate": 2.9087661314045366e-05,
+      "loss": 2.6135,
+      "step": 7534
+    },
+    {
+      "epoch": 0.6820856341088078,
+      "grad_norm": 0.7897319793701172,
+      "learning_rate": 2.9065290258634958e-05,
+      "loss": 2.6112,
+      "step": 7535
+    },
+    {
+      "epoch": 0.6821761564225581,
+      "grad_norm": 0.8886232972145081,
+      "learning_rate": 2.904292634634793e-05,
+      "loss": 2.6189,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6822666787363085,
+      "grad_norm": 0.8394643068313599,
+      "learning_rate": 2.9020569579436252e-05,
+      "loss": 2.6385,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6823572010500588,
+      "grad_norm": 0.8300806879997253,
+      "learning_rate": 2.8998219960151275e-05,
+      "loss": 2.6521,
+      "step": 7538
+    },
+    {
+      "epoch": 0.6824477233638092,
+      "grad_norm": 0.8635507822036743,
+      "learning_rate": 2.8975877490743607e-05,
+      "loss": 2.6268,
+      "step": 7539
+    },
+    {
+      "epoch": 0.6825382456775595,
+      "grad_norm": 0.7847825884819031,
+      "learning_rate": 2.8953542173463133e-05,
+      "loss": 2.6358,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6826287679913099,
+      "grad_norm": 0.9233072400093079,
+      "learning_rate": 2.8931214010559004e-05,
+      "loss": 2.6771,
+      "step": 7541
+    },
+    {
+      "epoch": 0.6827192903050602,
+      "grad_norm": 0.8705090880393982,
+      "learning_rate": 2.8908893004279723e-05,
+      "loss": 2.6798,
+      "step": 7542
+    },
+    {
+      "epoch": 0.6828098126188106,
+      "grad_norm": 0.8056427836418152,
+      "learning_rate": 2.8886579156872916e-05,
+      "loss": 2.6228,
+      "step": 7543
+    },
+    {
+      "epoch": 0.6829003349325609,
+      "grad_norm": 0.8417764902114868,
+      "learning_rate": 2.8864272470585696e-05,
+      "loss": 2.6361,
+      "step": 7544
+    },
+    {
+      "epoch": 0.6829908572463113,
+      "grad_norm": 0.9244621992111206,
+      "learning_rate": 2.8841972947664252e-05,
+      "loss": 2.6495,
+      "step": 7545
+    },
+    {
+      "epoch": 0.6830813795600615,
+      "grad_norm": 0.8148440718650818,
+      "learning_rate": 2.8819680590354202e-05,
+      "loss": 2.6521,
+      "step": 7546
+    },
+    {
+      "epoch": 0.6831719018738119,
+      "grad_norm": 0.8853485584259033,
+      "learning_rate": 2.879739540090036e-05,
+      "loss": 2.6394,
+      "step": 7547
+    },
+    {
+      "epoch": 0.6832624241875622,
+      "grad_norm": 0.8543866276741028,
+      "learning_rate": 2.8775117381546857e-05,
+      "loss": 2.6207,
+      "step": 7548
+    },
+    {
+      "epoch": 0.6833529465013126,
+      "grad_norm": 0.8939554691314697,
+      "learning_rate": 2.875284653453708e-05,
+      "loss": 2.653,
+      "step": 7549
+    },
+    {
+      "epoch": 0.6834434688150629,
+      "grad_norm": 0.8086675405502319,
+      "learning_rate": 2.8730582862113742e-05,
+      "loss": 2.6412,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6835339911288133,
+      "grad_norm": 0.8328844308853149,
+      "learning_rate": 2.870832636651869e-05,
+      "loss": 2.6295,
+      "step": 7551
+    },
+    {
+      "epoch": 0.6836245134425636,
+      "grad_norm": 0.8223507404327393,
+      "learning_rate": 2.8686077049993287e-05,
+      "loss": 2.6705,
+      "step": 7552
+    },
+    {
+      "epoch": 0.683715035756314,
+      "grad_norm": 0.8664432764053345,
+      "learning_rate": 2.8663834914777952e-05,
+      "loss": 2.6218,
+      "step": 7553
+    },
+    {
+      "epoch": 0.6838055580700643,
+      "grad_norm": 0.8756895661354065,
+      "learning_rate": 2.8641599963112488e-05,
+      "loss": 2.6909,
+      "step": 7554
+    },
+    {
+      "epoch": 0.6838960803838147,
+      "grad_norm": 0.8878546357154846,
+      "learning_rate": 2.861937219723595e-05,
+      "loss": 2.6313,
+      "step": 7555
+    },
+    {
+      "epoch": 0.683986602697565,
+      "grad_norm": 0.8262296319007874,
+      "learning_rate": 2.8597151619386707e-05,
+      "loss": 2.6287,
+      "step": 7556
+    },
+    {
+      "epoch": 0.6840771250113152,
+      "grad_norm": 0.8517236709594727,
+      "learning_rate": 2.8574938231802283e-05,
+      "loss": 2.6161,
+      "step": 7557
+    },
+    {
+      "epoch": 0.6841676473250656,
+      "grad_norm": 0.8476995229721069,
+      "learning_rate": 2.8552732036719687e-05,
+      "loss": 2.6321,
+      "step": 7558
+    },
+    {
+      "epoch": 0.6842581696388159,
+      "grad_norm": 0.8412792086601257,
+      "learning_rate": 2.8530533036374984e-05,
+      "loss": 2.6756,
+      "step": 7559
+    },
+    {
+      "epoch": 0.6843486919525663,
+      "grad_norm": 0.9044947624206543,
+      "learning_rate": 2.8508341233003654e-05,
+      "loss": 2.6733,
+      "step": 7560
+    },
+    {
+      "epoch": 0.6844392142663166,
+      "grad_norm": 0.8056501150131226,
+      "learning_rate": 2.848615662884041e-05,
+      "loss": 2.6212,
+      "step": 7561
+    },
+    {
+      "epoch": 0.684529736580067,
+      "grad_norm": 0.8448612689971924,
+      "learning_rate": 2.846397922611923e-05,
+      "loss": 2.6485,
+      "step": 7562
+    },
+    {
+      "epoch": 0.6846202588938173,
+      "grad_norm": 0.8401530385017395,
+      "learning_rate": 2.8441809027073386e-05,
+      "loss": 2.6439,
+      "step": 7563
+    },
+    {
+      "epoch": 0.6847107812075677,
+      "grad_norm": 0.846659243106842,
+      "learning_rate": 2.8419646033935444e-05,
+      "loss": 2.5858,
+      "step": 7564
+    },
+    {
+      "epoch": 0.684801303521318,
+      "grad_norm": 0.8078282475471497,
+      "learning_rate": 2.839749024893713e-05,
+      "loss": 2.6221,
+      "step": 7565
+    },
+    {
+      "epoch": 0.6848918258350684,
+      "grad_norm": 0.833645761013031,
+      "learning_rate": 2.8375341674309653e-05,
+      "loss": 2.6456,
+      "step": 7566
+    },
+    {
+      "epoch": 0.6849823481488186,
+      "grad_norm": 0.7956631779670715,
+      "learning_rate": 2.835320031228328e-05,
+      "loss": 2.6156,
+      "step": 7567
+    },
+    {
+      "epoch": 0.685072870462569,
+      "grad_norm": 0.8895859718322754,
+      "learning_rate": 2.833106616508766e-05,
+      "loss": 2.7153,
+      "step": 7568
+    },
+    {
+      "epoch": 0.6851633927763193,
+      "grad_norm": 0.8952557444572449,
+      "learning_rate": 2.8308939234951726e-05,
+      "loss": 2.6639,
+      "step": 7569
+    },
+    {
+      "epoch": 0.6852539150900697,
+      "grad_norm": 0.8234478235244751,
+      "learning_rate": 2.828681952410366e-05,
+      "loss": 2.6604,
+      "step": 7570
+    },
+    {
+      "epoch": 0.68534443740382,
+      "grad_norm": 0.7875887751579285,
+      "learning_rate": 2.826470703477089e-05,
+      "loss": 2.5946,
+      "step": 7571
+    },
+    {
+      "epoch": 0.6854349597175704,
+      "grad_norm": 0.7900510430335999,
+      "learning_rate": 2.82426017691802e-05,
+      "loss": 2.6325,
+      "step": 7572
+    },
+    {
+      "epoch": 0.6855254820313207,
+      "grad_norm": 0.8922678828239441,
+      "learning_rate": 2.822050372955748e-05,
+      "loss": 2.5736,
+      "step": 7573
+    },
+    {
+      "epoch": 0.6856160043450711,
+      "grad_norm": 0.8456141948699951,
+      "learning_rate": 2.8198412918128126e-05,
+      "loss": 2.7264,
+      "step": 7574
+    },
+    {
+      "epoch": 0.6857065266588214,
+      "grad_norm": 0.8549056053161621,
+      "learning_rate": 2.81763293371166e-05,
+      "loss": 2.6174,
+      "step": 7575
+    },
+    {
+      "epoch": 0.6857970489725718,
+      "grad_norm": 0.8470386266708374,
+      "learning_rate": 2.8154252988746755e-05,
+      "loss": 2.6436,
+      "step": 7576
+    },
+    {
+      "epoch": 0.685887571286322,
+      "grad_norm": 0.8629388809204102,
+      "learning_rate": 2.8132183875241657e-05,
+      "loss": 2.6735,
+      "step": 7577
+    },
+    {
+      "epoch": 0.6859780936000724,
+      "grad_norm": 0.9061052203178406,
+      "learning_rate": 2.811012199882368e-05,
+      "loss": 2.6579,
+      "step": 7578
+    },
+    {
+      "epoch": 0.6860686159138227,
+      "grad_norm": 0.8674964308738708,
+      "learning_rate": 2.8088067361714455e-05,
+      "loss": 2.6735,
+      "step": 7579
+    },
+    {
+      "epoch": 0.6861591382275731,
+      "grad_norm": 0.8112437725067139,
+      "learning_rate": 2.8066019966134904e-05,
+      "loss": 2.6401,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6862496605413234,
+      "grad_norm": 0.8892813324928284,
+      "learning_rate": 2.8043979814305123e-05,
+      "loss": 2.6504,
+      "step": 7581
+    },
+    {
+      "epoch": 0.6863401828550738,
+      "grad_norm": 0.8235838413238525,
+      "learning_rate": 2.802194690844465e-05,
+      "loss": 2.6437,
+      "step": 7582
+    },
+    {
+      "epoch": 0.6864307051688241,
+      "grad_norm": 0.8243260383605957,
+      "learning_rate": 2.799992125077213e-05,
+      "loss": 2.6016,
+      "step": 7583
+    },
+    {
+      "epoch": 0.6865212274825745,
+      "grad_norm": 0.8342082500457764,
+      "learning_rate": 2.7977902843505578e-05,
+      "loss": 2.6386,
+      "step": 7584
+    },
+    {
+      "epoch": 0.6866117497963248,
+      "grad_norm": 0.8364400267601013,
+      "learning_rate": 2.7955891688862236e-05,
+      "loss": 2.6256,
+      "step": 7585
+    },
+    {
+      "epoch": 0.6867022721100752,
+      "grad_norm": 0.9203447103500366,
+      "learning_rate": 2.7933887789058622e-05,
+      "loss": 2.6949,
+      "step": 7586
+    },
+    {
+      "epoch": 0.6867927944238255,
+      "grad_norm": 0.9918256998062134,
+      "learning_rate": 2.7911891146310544e-05,
+      "loss": 2.6685,
+      "step": 7587
+    },
+    {
+      "epoch": 0.6868833167375759,
+      "grad_norm": 0.8644828796386719,
+      "learning_rate": 2.7889901762833083e-05,
+      "loss": 2.5957,
+      "step": 7588
+    },
+    {
+      "epoch": 0.6869738390513261,
+      "grad_norm": 0.8854060173034668,
+      "learning_rate": 2.786791964084048e-05,
+      "loss": 2.604,
+      "step": 7589
+    },
+    {
+      "epoch": 0.6870643613650765,
+      "grad_norm": 0.8530220985412598,
+      "learning_rate": 2.7845944782546453e-05,
+      "loss": 2.678,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6871548836788268,
+      "grad_norm": 0.8707152009010315,
+      "learning_rate": 2.7823977190163786e-05,
+      "loss": 2.6487,
+      "step": 7591
+    },
+    {
+      "epoch": 0.6872454059925772,
+      "grad_norm": 0.8770884275436401,
+      "learning_rate": 2.7802016865904644e-05,
+      "loss": 2.6149,
+      "step": 7592
+    },
+    {
+      "epoch": 0.6873359283063275,
+      "grad_norm": 0.9286378026008606,
+      "learning_rate": 2.7780063811980417e-05,
+      "loss": 2.6784,
+      "step": 7593
+    },
+    {
+      "epoch": 0.6874264506200779,
+      "grad_norm": 0.8385622501373291,
+      "learning_rate": 2.775811803060183e-05,
+      "loss": 2.642,
+      "step": 7594
+    },
+    {
+      "epoch": 0.6875169729338282,
+      "grad_norm": 0.9218869805335999,
+      "learning_rate": 2.773617952397871e-05,
+      "loss": 2.6173,
+      "step": 7595
+    },
+    {
+      "epoch": 0.6876074952475786,
+      "grad_norm": 0.8965048789978027,
+      "learning_rate": 2.771424829432041e-05,
+      "loss": 2.6545,
+      "step": 7596
+    },
+    {
+      "epoch": 0.6876980175613289,
+      "grad_norm": 0.9724282622337341,
+      "learning_rate": 2.7692324343835286e-05,
+      "loss": 2.7086,
+      "step": 7597
+    },
+    {
+      "epoch": 0.6877885398750792,
+      "grad_norm": 0.9105857014656067,
+      "learning_rate": 2.7670407674731126e-05,
+      "loss": 2.6421,
+      "step": 7598
+    },
+    {
+      "epoch": 0.6878790621888295,
+      "grad_norm": 0.8620057106018066,
+      "learning_rate": 2.764849828921493e-05,
+      "loss": 2.6495,
+      "step": 7599
+    },
+    {
+      "epoch": 0.6879695845025798,
+      "grad_norm": 0.8677911162376404,
+      "learning_rate": 2.7626596189492983e-05,
+      "loss": 2.6579,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6880601068163302,
+      "grad_norm": 0.8879269361495972,
+      "learning_rate": 2.7604701377770816e-05,
+      "loss": 2.6584,
+      "step": 7601
+    },
+    {
+      "epoch": 0.6881506291300805,
+      "grad_norm": 0.8561745882034302,
+      "learning_rate": 2.7582813856253275e-05,
+      "loss": 2.6416,
+      "step": 7602
+    },
+    {
+      "epoch": 0.6882411514438309,
+      "grad_norm": 0.8197829723358154,
+      "learning_rate": 2.7560933627144325e-05,
+      "loss": 2.6112,
+      "step": 7603
+    },
+    {
+      "epoch": 0.6883316737575812,
+      "grad_norm": 0.8608514666557312,
+      "learning_rate": 2.7539060692647446e-05,
+      "loss": 2.6194,
+      "step": 7604
+    },
+    {
+      "epoch": 0.6884221960713316,
+      "grad_norm": 0.8288963437080383,
+      "learning_rate": 2.7517195054965138e-05,
+      "loss": 2.5699,
+      "step": 7605
+    },
+    {
+      "epoch": 0.6885127183850819,
+      "grad_norm": 0.8938080072402954,
+      "learning_rate": 2.7495336716299313e-05,
+      "loss": 2.7042,
+      "step": 7606
+    },
+    {
+      "epoch": 0.6886032406988323,
+      "grad_norm": 0.869457483291626,
+      "learning_rate": 2.7473485678851086e-05,
+      "loss": 2.6793,
+      "step": 7607
+    },
+    {
+      "epoch": 0.6886937630125826,
+      "grad_norm": 0.8588059544563293,
+      "learning_rate": 2.745164194482087e-05,
+      "loss": 2.6752,
+      "step": 7608
+    },
+    {
+      "epoch": 0.688784285326333,
+      "grad_norm": 0.8931800127029419,
+      "learning_rate": 2.7429805516408337e-05,
+      "loss": 2.6063,
+      "step": 7609
+    },
+    {
+      "epoch": 0.6888748076400832,
+      "grad_norm": 0.9317516684532166,
+      "learning_rate": 2.7407976395812418e-05,
+      "loss": 2.6434,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6889653299538336,
+      "grad_norm": 0.9000864028930664,
+      "learning_rate": 2.738615458523126e-05,
+      "loss": 2.646,
+      "step": 7611
+    },
+    {
+      "epoch": 0.6890558522675839,
+      "grad_norm": 0.8780214190483093,
+      "learning_rate": 2.736434008686235e-05,
+      "loss": 2.5954,
+      "step": 7612
+    },
+    {
+      "epoch": 0.6891463745813343,
+      "grad_norm": 0.8571871519088745,
+      "learning_rate": 2.734253290290242e-05,
+      "loss": 2.6304,
+      "step": 7613
+    },
+    {
+      "epoch": 0.6892368968950846,
+      "grad_norm": 0.8072302937507629,
+      "learning_rate": 2.732073303554742e-05,
+      "loss": 2.6923,
+      "step": 7614
+    },
+    {
+      "epoch": 0.689327419208835,
+      "grad_norm": 0.8645291924476624,
+      "learning_rate": 2.729894048699265e-05,
+      "loss": 2.7113,
+      "step": 7615
+    },
+    {
+      "epoch": 0.6894179415225853,
+      "grad_norm": 0.8110138773918152,
+      "learning_rate": 2.727715525943253e-05,
+      "loss": 2.6095,
+      "step": 7616
+    },
+    {
+      "epoch": 0.6895084638363357,
+      "grad_norm": 0.8943793773651123,
+      "learning_rate": 2.725537735506094e-05,
+      "loss": 2.6108,
+      "step": 7617
+    },
+    {
+      "epoch": 0.689598986150086,
+      "grad_norm": 0.844049334526062,
+      "learning_rate": 2.723360677607083e-05,
+      "loss": 2.6082,
+      "step": 7618
+    },
+    {
+      "epoch": 0.6896895084638364,
+      "grad_norm": 0.808815062046051,
+      "learning_rate": 2.7211843524654535e-05,
+      "loss": 2.6504,
+      "step": 7619
+    },
+    {
+      "epoch": 0.6897800307775867,
+      "grad_norm": 0.8212969899177551,
+      "learning_rate": 2.719008760300359e-05,
+      "loss": 2.6516,
+      "step": 7620
+    },
+    {
+      "epoch": 0.689870553091337,
+      "grad_norm": 0.8447639346122742,
+      "learning_rate": 2.7168339013308875e-05,
+      "loss": 2.6996,
+      "step": 7621
+    },
+    {
+      "epoch": 0.6899610754050873,
+      "grad_norm": 0.9643140435218811,
+      "learning_rate": 2.714659775776036e-05,
+      "loss": 2.6779,
+      "step": 7622
+    },
+    {
+      "epoch": 0.6900515977188377,
+      "grad_norm": 0.8468511700630188,
+      "learning_rate": 2.712486383854752e-05,
+      "loss": 2.6389,
+      "step": 7623
+    },
+    {
+      "epoch": 0.690142120032588,
+      "grad_norm": 0.8627493977546692,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 2.6873,
+      "step": 7624
+    },
+    {
+      "epoch": 0.6902326423463384,
+      "grad_norm": 0.9205848574638367,
+      "learning_rate": 2.70814180178823e-05,
+      "loss": 2.6157,
+      "step": 7625
+    },
+    {
+      "epoch": 0.6903231646600887,
+      "grad_norm": 0.857008695602417,
+      "learning_rate": 2.705970612080494e-05,
+      "loss": 2.6498,
+      "step": 7626
+    },
+    {
+      "epoch": 0.6904136869738391,
+      "grad_norm": 0.8849743604660034,
+      "learning_rate": 2.7038001568813175e-05,
+      "loss": 2.6083,
+      "step": 7627
+    },
+    {
+      "epoch": 0.6905042092875894,
+      "grad_norm": 0.8480933904647827,
+      "learning_rate": 2.701630436409267e-05,
+      "loss": 2.6904,
+      "step": 7628
+    },
+    {
+      "epoch": 0.6905947316013398,
+      "grad_norm": 0.8617257475852966,
+      "learning_rate": 2.699461450882833e-05,
+      "loss": 2.6551,
+      "step": 7629
+    },
+    {
+      "epoch": 0.6906852539150901,
+      "grad_norm": 0.9328704476356506,
+      "learning_rate": 2.6972932005204267e-05,
+      "loss": 2.6988,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6907757762288405,
+      "grad_norm": 0.8302156329154968,
+      "learning_rate": 2.6951256855404005e-05,
+      "loss": 2.574,
+      "step": 7631
+    },
+    {
+      "epoch": 0.6908662985425907,
+      "grad_norm": 0.8471900820732117,
+      "learning_rate": 2.692958906161015e-05,
+      "loss": 2.662,
+      "step": 7632
+    },
+    {
+      "epoch": 0.6909568208563411,
+      "grad_norm": 0.8453875184059143,
+      "learning_rate": 2.6907928626004674e-05,
+      "loss": 2.6366,
+      "step": 7633
+    },
+    {
+      "epoch": 0.6910473431700914,
+      "grad_norm": 0.8545379638671875,
+      "learning_rate": 2.688627555076879e-05,
+      "loss": 2.5942,
+      "step": 7634
+    },
+    {
+      "epoch": 0.6911378654838418,
+      "grad_norm": 0.8129159808158875,
+      "learning_rate": 2.6864629838082956e-05,
+      "loss": 2.6947,
+      "step": 7635
+    },
+    {
+      "epoch": 0.6912283877975921,
+      "grad_norm": 0.8285556435585022,
+      "learning_rate": 2.68429914901269e-05,
+      "loss": 2.6357,
+      "step": 7636
+    },
+    {
+      "epoch": 0.6913189101113425,
+      "grad_norm": 0.8780727386474609,
+      "learning_rate": 2.682136050907962e-05,
+      "loss": 2.642,
+      "step": 7637
+    },
+    {
+      "epoch": 0.6914094324250928,
+      "grad_norm": 0.9794698357582092,
+      "learning_rate": 2.6799736897119278e-05,
+      "loss": 2.7138,
+      "step": 7638
+    },
+    {
+      "epoch": 0.6914999547388431,
+      "grad_norm": 0.815741240978241,
+      "learning_rate": 2.677812065642349e-05,
+      "loss": 2.6004,
+      "step": 7639
+    },
+    {
+      "epoch": 0.6915904770525935,
+      "grad_norm": 0.84134441614151,
+      "learning_rate": 2.6756511789168925e-05,
+      "loss": 2.6911,
+      "step": 7640
+    },
+    {
+      "epoch": 0.6916809993663438,
+      "grad_norm": 1.0214581489562988,
+      "learning_rate": 2.6734910297531613e-05,
+      "loss": 2.5947,
+      "step": 7641
+    },
+    {
+      "epoch": 0.6917715216800941,
+      "grad_norm": 0.9078264832496643,
+      "learning_rate": 2.671331618368682e-05,
+      "loss": 2.6658,
+      "step": 7642
+    },
+    {
+      "epoch": 0.6918620439938444,
+      "grad_norm": 0.8683854341506958,
+      "learning_rate": 2.6691729449809088e-05,
+      "loss": 2.6305,
+      "step": 7643
+    },
+    {
+      "epoch": 0.6919525663075948,
+      "grad_norm": 0.9419749975204468,
+      "learning_rate": 2.66701500980722e-05,
+      "loss": 2.6673,
+      "step": 7644
+    },
+    {
+      "epoch": 0.6920430886213451,
+      "grad_norm": 0.9228997230529785,
+      "learning_rate": 2.6648578130649215e-05,
+      "loss": 2.6446,
+      "step": 7645
+    },
+    {
+      "epoch": 0.6921336109350955,
+      "grad_norm": 0.839359700679779,
+      "learning_rate": 2.6627013549712355e-05,
+      "loss": 2.633,
+      "step": 7646
+    },
+    {
+      "epoch": 0.6922241332488458,
+      "grad_norm": 0.8543779850006104,
+      "learning_rate": 2.6605456357433268e-05,
+      "loss": 2.619,
+      "step": 7647
+    },
+    {
+      "epoch": 0.6923146555625962,
+      "grad_norm": 0.9176219701766968,
+      "learning_rate": 2.6583906555982697e-05,
+      "loss": 2.6784,
+      "step": 7648
+    },
+    {
+      "epoch": 0.6924051778763465,
+      "grad_norm": 0.8997385501861572,
+      "learning_rate": 2.6562364147530727e-05,
+      "loss": 2.7032,
+      "step": 7649
+    },
+    {
+      "epoch": 0.6924957001900969,
+      "grad_norm": 0.868236243724823,
+      "learning_rate": 2.654082913424668e-05,
+      "loss": 2.6488,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6925862225038472,
+      "grad_norm": 0.8935613036155701,
+      "learning_rate": 2.6519301518299157e-05,
+      "loss": 2.6575,
+      "step": 7651
+    },
+    {
+      "epoch": 0.6926767448175976,
+      "grad_norm": 0.895734965801239,
+      "learning_rate": 2.6497781301855895e-05,
+      "loss": 2.6495,
+      "step": 7652
+    },
+    {
+      "epoch": 0.6927672671313478,
+      "grad_norm": 0.9116073250770569,
+      "learning_rate": 2.647626848708411e-05,
+      "loss": 2.6608,
+      "step": 7653
+    },
+    {
+      "epoch": 0.6928577894450982,
+      "grad_norm": 0.9018348455429077,
+      "learning_rate": 2.6454763076150046e-05,
+      "loss": 2.7109,
+      "step": 7654
+    },
+    {
+      "epoch": 0.6929483117588485,
+      "grad_norm": 0.83049076795578,
+      "learning_rate": 2.6433265071219327e-05,
+      "loss": 2.6045,
+      "step": 7655
+    },
+    {
+      "epoch": 0.6930388340725989,
+      "grad_norm": 0.8695515394210815,
+      "learning_rate": 2.6411774474456797e-05,
+      "loss": 2.6627,
+      "step": 7656
+    },
+    {
+      "epoch": 0.6931293563863492,
+      "grad_norm": 0.8901898860931396,
+      "learning_rate": 2.639029128802657e-05,
+      "loss": 2.5905,
+      "step": 7657
+    },
+    {
+      "epoch": 0.6932198787000996,
+      "grad_norm": 0.8204820156097412,
+      "learning_rate": 2.6368815514091992e-05,
+      "loss": 2.6049,
+      "step": 7658
+    },
+    {
+      "epoch": 0.6933104010138499,
+      "grad_norm": 0.8440626263618469,
+      "learning_rate": 2.6347347154815715e-05,
+      "loss": 2.6364,
+      "step": 7659
+    },
+    {
+      "epoch": 0.6934009233276003,
+      "grad_norm": 0.8276053071022034,
+      "learning_rate": 2.6325886212359498e-05,
+      "loss": 2.61,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6934914456413506,
+      "grad_norm": 0.8775001764297485,
+      "learning_rate": 2.6304432688884584e-05,
+      "loss": 2.691,
+      "step": 7661
+    },
+    {
+      "epoch": 0.693581967955101,
+      "grad_norm": 0.7714978456497192,
+      "learning_rate": 2.6282986586551262e-05,
+      "loss": 2.6292,
+      "step": 7662
+    },
+    {
+      "epoch": 0.6936724902688512,
+      "grad_norm": 0.8272567987442017,
+      "learning_rate": 2.6261547907519158e-05,
+      "loss": 2.6549,
+      "step": 7663
+    },
+    {
+      "epoch": 0.6937630125826016,
+      "grad_norm": 0.8909174203872681,
+      "learning_rate": 2.6240116653947166e-05,
+      "loss": 2.5959,
+      "step": 7664
+    },
+    {
+      "epoch": 0.6938535348963519,
+      "grad_norm": 0.8319750428199768,
+      "learning_rate": 2.6218692827993417e-05,
+      "loss": 2.6633,
+      "step": 7665
+    },
+    {
+      "epoch": 0.6939440572101023,
+      "grad_norm": 0.836725652217865,
+      "learning_rate": 2.6197276431815277e-05,
+      "loss": 2.6459,
+      "step": 7666
+    },
+    {
+      "epoch": 0.6940345795238526,
+      "grad_norm": 0.8594474196434021,
+      "learning_rate": 2.6175867467569405e-05,
+      "loss": 2.6501,
+      "step": 7667
+    },
+    {
+      "epoch": 0.694125101837603,
+      "grad_norm": 0.8767529129981995,
+      "learning_rate": 2.615446593741161e-05,
+      "loss": 2.5987,
+      "step": 7668
+    },
+    {
+      "epoch": 0.6942156241513533,
+      "grad_norm": 0.8602820038795471,
+      "learning_rate": 2.6133071843497126e-05,
+      "loss": 2.6252,
+      "step": 7669
+    },
+    {
+      "epoch": 0.6943061464651037,
+      "grad_norm": 0.9225953221321106,
+      "learning_rate": 2.6111685187980262e-05,
+      "loss": 2.6669,
+      "step": 7670
+    },
+    {
+      "epoch": 0.694396668778854,
+      "grad_norm": 0.8112370371818542,
+      "learning_rate": 2.6090305973014672e-05,
+      "loss": 2.6329,
+      "step": 7671
+    },
+    {
+      "epoch": 0.6944871910926044,
+      "grad_norm": 0.8147202134132385,
+      "learning_rate": 2.606893420075325e-05,
+      "loss": 2.6265,
+      "step": 7672
+    },
+    {
+      "epoch": 0.6945777134063547,
+      "grad_norm": 0.8251135349273682,
+      "learning_rate": 2.6047569873348133e-05,
+      "loss": 2.6144,
+      "step": 7673
+    },
+    {
+      "epoch": 0.694668235720105,
+      "grad_norm": 0.850727915763855,
+      "learning_rate": 2.6026212992950704e-05,
+      "loss": 2.6623,
+      "step": 7674
+    },
+    {
+      "epoch": 0.6947587580338553,
+      "grad_norm": 0.9017343521118164,
+      "learning_rate": 2.6004863561711635e-05,
+      "loss": 2.6588,
+      "step": 7675
+    },
+    {
+      "epoch": 0.6948492803476057,
+      "grad_norm": 0.8735215067863464,
+      "learning_rate": 2.5983521581780724e-05,
+      "loss": 2.6361,
+      "step": 7676
+    },
+    {
+      "epoch": 0.694939802661356,
+      "grad_norm": 0.7911974787712097,
+      "learning_rate": 2.5962187055307218e-05,
+      "loss": 2.5933,
+      "step": 7677
+    },
+    {
+      "epoch": 0.6950303249751064,
+      "grad_norm": 0.9292235970497131,
+      "learning_rate": 2.5940859984439424e-05,
+      "loss": 2.6769,
+      "step": 7678
+    },
+    {
+      "epoch": 0.6951208472888567,
+      "grad_norm": 0.8577764630317688,
+      "learning_rate": 2.5919540371325e-05,
+      "loss": 2.6119,
+      "step": 7679
+    },
+    {
+      "epoch": 0.695211369602607,
+      "grad_norm": 0.8263047933578491,
+      "learning_rate": 2.589822821811083e-05,
+      "loss": 2.6231,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6953018919163574,
+      "grad_norm": 0.825162947177887,
+      "learning_rate": 2.5876923526943055e-05,
+      "loss": 2.6699,
+      "step": 7681
+    },
+    {
+      "epoch": 0.6953924142301077,
+      "grad_norm": 0.8402916193008423,
+      "learning_rate": 2.585562629996705e-05,
+      "loss": 2.6313,
+      "step": 7682
+    },
+    {
+      "epoch": 0.6954829365438581,
+      "grad_norm": 0.832701563835144,
+      "learning_rate": 2.5834336539327486e-05,
+      "loss": 2.6451,
+      "step": 7683
+    },
+    {
+      "epoch": 0.6955734588576084,
+      "grad_norm": 0.8593753576278687,
+      "learning_rate": 2.5813054247168167e-05,
+      "loss": 2.6064,
+      "step": 7684
+    },
+    {
+      "epoch": 0.6956639811713587,
+      "grad_norm": 0.8329028487205505,
+      "learning_rate": 2.5791779425632255e-05,
+      "loss": 2.6614,
+      "step": 7685
+    },
+    {
+      "epoch": 0.695754503485109,
+      "grad_norm": 0.8595302700996399,
+      "learning_rate": 2.577051207686213e-05,
+      "loss": 2.5816,
+      "step": 7686
+    },
+    {
+      "epoch": 0.6958450257988594,
+      "grad_norm": 0.8601256012916565,
+      "learning_rate": 2.5749252202999408e-05,
+      "loss": 2.6177,
+      "step": 7687
+    },
+    {
+      "epoch": 0.6959355481126097,
+      "grad_norm": 0.8946163654327393,
+      "learning_rate": 2.5727999806184945e-05,
+      "loss": 2.6212,
+      "step": 7688
+    },
+    {
+      "epoch": 0.6960260704263601,
+      "grad_norm": 0.8692400455474854,
+      "learning_rate": 2.5706754888558914e-05,
+      "loss": 2.612,
+      "step": 7689
+    },
+    {
+      "epoch": 0.6961165927401104,
+      "grad_norm": 0.8570036888122559,
+      "learning_rate": 2.5685517452260567e-05,
+      "loss": 2.6734,
+      "step": 7690
+    },
+    {
+      "epoch": 0.6962071150538608,
+      "grad_norm": 0.854727566242218,
+      "learning_rate": 2.5664287499428642e-05,
+      "loss": 2.6159,
+      "step": 7691
+    },
+    {
+      "epoch": 0.6962976373676111,
+      "grad_norm": 0.8037676215171814,
+      "learning_rate": 2.5643065032200895e-05,
+      "loss": 2.6414,
+      "step": 7692
+    },
+    {
+      "epoch": 0.6963881596813615,
+      "grad_norm": 0.8172332644462585,
+      "learning_rate": 2.5621850052714468e-05,
+      "loss": 2.6629,
+      "step": 7693
+    },
+    {
+      "epoch": 0.6964786819951118,
+      "grad_norm": 0.9130670428276062,
+      "learning_rate": 2.5600642563105703e-05,
+      "loss": 2.6594,
+      "step": 7694
+    },
+    {
+      "epoch": 0.6965692043088622,
+      "grad_norm": 0.8906755447387695,
+      "learning_rate": 2.5579442565510203e-05,
+      "loss": 2.6459,
+      "step": 7695
+    },
+    {
+      "epoch": 0.6966597266226124,
+      "grad_norm": 0.8938328623771667,
+      "learning_rate": 2.5558250062062828e-05,
+      "loss": 2.664,
+      "step": 7696
+    },
+    {
+      "epoch": 0.6967502489363628,
+      "grad_norm": 1.0191668272018433,
+      "learning_rate": 2.5537065054897602e-05,
+      "loss": 2.6972,
+      "step": 7697
+    },
+    {
+      "epoch": 0.6968407712501131,
+      "grad_norm": 0.921165943145752,
+      "learning_rate": 2.5515887546147886e-05,
+      "loss": 2.6141,
+      "step": 7698
+    },
+    {
+      "epoch": 0.6969312935638635,
+      "grad_norm": 0.8833820819854736,
+      "learning_rate": 2.5494717537946256e-05,
+      "loss": 2.6725,
+      "step": 7699
+    },
+    {
+      "epoch": 0.6970218158776138,
+      "grad_norm": 0.8970410227775574,
+      "learning_rate": 2.5473555032424533e-05,
+      "loss": 2.6716,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6971123381913642,
+      "grad_norm": 0.8503276705741882,
+      "learning_rate": 2.5452400031713785e-05,
+      "loss": 2.6465,
+      "step": 7701
+    },
+    {
+      "epoch": 0.6972028605051145,
+      "grad_norm": 0.9117843508720398,
+      "learning_rate": 2.543125253794434e-05,
+      "loss": 2.6583,
+      "step": 7702
+    },
+    {
+      "epoch": 0.6972933828188649,
+      "grad_norm": 0.8610855937004089,
+      "learning_rate": 2.5410112553245657e-05,
+      "loss": 2.5539,
+      "step": 7703
+    },
+    {
+      "epoch": 0.6973839051326152,
+      "grad_norm": 0.8850502371788025,
+      "learning_rate": 2.538898007974667e-05,
+      "loss": 2.6254,
+      "step": 7704
+    },
+    {
+      "epoch": 0.6974744274463656,
+      "grad_norm": 0.9232295751571655,
+      "learning_rate": 2.536785511957531e-05,
+      "loss": 2.6897,
+      "step": 7705
+    },
+    {
+      "epoch": 0.6975649497601158,
+      "grad_norm": 0.8219690918922424,
+      "learning_rate": 2.53467376748589e-05,
+      "loss": 2.6201,
+      "step": 7706
+    },
+    {
+      "epoch": 0.6976554720738662,
+      "grad_norm": 0.8805753588676453,
+      "learning_rate": 2.532562774772397e-05,
+      "loss": 2.5956,
+      "step": 7707
+    },
+    {
+      "epoch": 0.6977459943876165,
+      "grad_norm": 0.8470893502235413,
+      "learning_rate": 2.530452534029627e-05,
+      "loss": 2.6556,
+      "step": 7708
+    },
+    {
+      "epoch": 0.6978365167013669,
+      "grad_norm": 0.8256682753562927,
+      "learning_rate": 2.528343045470083e-05,
+      "loss": 2.6603,
+      "step": 7709
+    },
+    {
+      "epoch": 0.6979270390151172,
+      "grad_norm": 0.8292689323425293,
+      "learning_rate": 2.5262343093061936e-05,
+      "loss": 2.6241,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6980175613288676,
+      "grad_norm": 0.8731260299682617,
+      "learning_rate": 2.5241263257502977e-05,
+      "loss": 2.6633,
+      "step": 7711
+    },
+    {
+      "epoch": 0.6981080836426179,
+      "grad_norm": 0.9130814075469971,
+      "learning_rate": 2.5220190950146827e-05,
+      "loss": 2.6724,
+      "step": 7712
+    },
+    {
+      "epoch": 0.6981986059563683,
+      "grad_norm": 0.8607717752456665,
+      "learning_rate": 2.5199126173115372e-05,
+      "loss": 2.6518,
+      "step": 7713
+    },
+    {
+      "epoch": 0.6982891282701186,
+      "grad_norm": 0.8219232559204102,
+      "learning_rate": 2.5178068928529864e-05,
+      "loss": 2.647,
+      "step": 7714
+    },
+    {
+      "epoch": 0.698379650583869,
+      "grad_norm": 0.8676030039787292,
+      "learning_rate": 2.515701921851077e-05,
+      "loss": 2.6465,
+      "step": 7715
+    },
+    {
+      "epoch": 0.6984701728976193,
+      "grad_norm": 0.8603833317756653,
+      "learning_rate": 2.5135977045177815e-05,
+      "loss": 2.6908,
+      "step": 7716
+    },
+    {
+      "epoch": 0.6985606952113697,
+      "grad_norm": 0.8424780368804932,
+      "learning_rate": 2.5114942410649878e-05,
+      "loss": 2.6428,
+      "step": 7717
+    },
+    {
+      "epoch": 0.6986512175251199,
+      "grad_norm": 0.8137457370758057,
+      "learning_rate": 2.5093915317045237e-05,
+      "loss": 2.723,
+      "step": 7718
+    },
+    {
+      "epoch": 0.6987417398388703,
+      "grad_norm": 0.8049132227897644,
+      "learning_rate": 2.5072895766481254e-05,
+      "loss": 2.668,
+      "step": 7719
+    },
+    {
+      "epoch": 0.6988322621526206,
+      "grad_norm": 0.8493908047676086,
+      "learning_rate": 2.5051883761074614e-05,
+      "loss": 2.5841,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6989227844663709,
+      "grad_norm": 0.949332058429718,
+      "learning_rate": 2.5030879302941222e-05,
+      "loss": 2.6522,
+      "step": 7721
+    },
+    {
+      "epoch": 0.6990133067801213,
+      "grad_norm": 0.8372514247894287,
+      "learning_rate": 2.500988239419624e-05,
+      "loss": 2.569,
+      "step": 7722
+    },
+    {
+      "epoch": 0.6991038290938716,
+      "grad_norm": 0.9176880717277527,
+      "learning_rate": 2.4988893036954043e-05,
+      "loss": 2.6696,
+      "step": 7723
+    },
+    {
+      "epoch": 0.699194351407622,
+      "grad_norm": 0.7949821352958679,
+      "learning_rate": 2.4967911233328302e-05,
+      "loss": 2.6849,
+      "step": 7724
+    },
+    {
+      "epoch": 0.6992848737213723,
+      "grad_norm": 0.9147369861602783,
+      "learning_rate": 2.494693698543179e-05,
+      "loss": 2.6272,
+      "step": 7725
+    },
+    {
+      "epoch": 0.6993753960351227,
+      "grad_norm": 0.8446216583251953,
+      "learning_rate": 2.4925970295376722e-05,
+      "loss": 2.6242,
+      "step": 7726
+    },
+    {
+      "epoch": 0.699465918348873,
+      "grad_norm": 0.8607931137084961,
+      "learning_rate": 2.4905011165274363e-05,
+      "loss": 2.6809,
+      "step": 7727
+    },
+    {
+      "epoch": 0.6995564406626233,
+      "grad_norm": 0.9009124636650085,
+      "learning_rate": 2.488405959723532e-05,
+      "loss": 2.6458,
+      "step": 7728
+    },
+    {
+      "epoch": 0.6996469629763736,
+      "grad_norm": 0.8539814949035645,
+      "learning_rate": 2.4863115593369423e-05,
+      "loss": 2.6098,
+      "step": 7729
+    },
+    {
+      "epoch": 0.699737485290124,
+      "grad_norm": 0.9067366123199463,
+      "learning_rate": 2.4842179155785737e-05,
+      "loss": 2.5878,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6998280076038743,
+      "grad_norm": 0.8943397998809814,
+      "learning_rate": 2.4821250286592546e-05,
+      "loss": 2.6324,
+      "step": 7731
+    },
+    {
+      "epoch": 0.6999185299176247,
+      "grad_norm": 0.8534390926361084,
+      "learning_rate": 2.4800328987897427e-05,
+      "loss": 2.6712,
+      "step": 7732
+    },
+    {
+      "epoch": 0.700009052231375,
+      "grad_norm": 0.8815813064575195,
+      "learning_rate": 2.477941526180706e-05,
+      "loss": 2.732,
+      "step": 7733
+    },
+    {
+      "epoch": 0.7000995745451254,
+      "grad_norm": 0.8730425834655762,
+      "learning_rate": 2.4758509110427575e-05,
+      "loss": 2.6302,
+      "step": 7734
+    },
+    {
+      "epoch": 0.7001900968588757,
+      "grad_norm": 0.7939863204956055,
+      "learning_rate": 2.473761053586414e-05,
+      "loss": 2.578,
+      "step": 7735
+    },
+    {
+      "epoch": 0.7002806191726261,
+      "grad_norm": 0.8331583738327026,
+      "learning_rate": 2.4716719540221268e-05,
+      "loss": 2.663,
+      "step": 7736
+    },
+    {
+      "epoch": 0.7003711414863764,
+      "grad_norm": 0.8007463216781616,
+      "learning_rate": 2.4695836125602667e-05,
+      "loss": 2.5627,
+      "step": 7737
+    },
+    {
+      "epoch": 0.7004616638001268,
+      "grad_norm": 0.7924402952194214,
+      "learning_rate": 2.4674960294111314e-05,
+      "loss": 2.6149,
+      "step": 7738
+    },
+    {
+      "epoch": 0.700552186113877,
+      "grad_norm": 0.8849868178367615,
+      "learning_rate": 2.46540920478494e-05,
+      "loss": 2.6322,
+      "step": 7739
+    },
+    {
+      "epoch": 0.7006427084276274,
+      "grad_norm": 0.9710038900375366,
+      "learning_rate": 2.4633231388918378e-05,
+      "loss": 2.614,
+      "step": 7740
+    },
+    {
+      "epoch": 0.7007332307413777,
+      "grad_norm": 0.8998742699623108,
+      "learning_rate": 2.461237831941884e-05,
+      "loss": 2.6947,
+      "step": 7741
+    },
+    {
+      "epoch": 0.7008237530551281,
+      "grad_norm": 0.7999228239059448,
+      "learning_rate": 2.4591532841450804e-05,
+      "loss": 2.6198,
+      "step": 7742
+    },
+    {
+      "epoch": 0.7009142753688784,
+      "grad_norm": 0.9255106449127197,
+      "learning_rate": 2.4570694957113304e-05,
+      "loss": 2.6402,
+      "step": 7743
+    },
+    {
+      "epoch": 0.7010047976826288,
+      "grad_norm": 0.8379217982292175,
+      "learning_rate": 2.4549864668504774e-05,
+      "loss": 2.6122,
+      "step": 7744
+    },
+    {
+      "epoch": 0.7010953199963791,
+      "grad_norm": 0.8677035570144653,
+      "learning_rate": 2.45290419777228e-05,
+      "loss": 2.6151,
+      "step": 7745
+    },
+    {
+      "epoch": 0.7011858423101295,
+      "grad_norm": 0.888058602809906,
+      "learning_rate": 2.450822688686426e-05,
+      "loss": 2.6339,
+      "step": 7746
+    },
+    {
+      "epoch": 0.7012763646238798,
+      "grad_norm": 0.825615644454956,
+      "learning_rate": 2.4487419398025157e-05,
+      "loss": 2.6411,
+      "step": 7747
+    },
+    {
+      "epoch": 0.7013668869376302,
+      "grad_norm": 0.8349601626396179,
+      "learning_rate": 2.4466619513300915e-05,
+      "loss": 2.6204,
+      "step": 7748
+    },
+    {
+      "epoch": 0.7014574092513804,
+      "grad_norm": 0.8351013660430908,
+      "learning_rate": 2.444582723478598e-05,
+      "loss": 2.6293,
+      "step": 7749
+    },
+    {
+      "epoch": 0.7015479315651308,
+      "grad_norm": 0.8708988428115845,
+      "learning_rate": 2.4425042564574184e-05,
+      "loss": 2.6312,
+      "step": 7750
+    },
+    {
+      "epoch": 0.7016384538788811,
+      "grad_norm": 0.8427244424819946,
+      "learning_rate": 2.4404265504758528e-05,
+      "loss": 2.6313,
+      "step": 7751
+    },
+    {
+      "epoch": 0.7017289761926315,
+      "grad_norm": 0.821071445941925,
+      "learning_rate": 2.4383496057431267e-05,
+      "loss": 2.6344,
+      "step": 7752
+    },
+    {
+      "epoch": 0.7018194985063818,
+      "grad_norm": 0.8288323283195496,
+      "learning_rate": 2.436273422468389e-05,
+      "loss": 2.6508,
+      "step": 7753
+    },
+    {
+      "epoch": 0.7019100208201322,
+      "grad_norm": 0.8814200758934021,
+      "learning_rate": 2.4341980008607136e-05,
+      "loss": 2.635,
+      "step": 7754
+    },
+    {
+      "epoch": 0.7020005431338825,
+      "grad_norm": 0.9194662570953369,
+      "learning_rate": 2.4321233411290868e-05,
+      "loss": 2.5983,
+      "step": 7755
+    },
+    {
+      "epoch": 0.7020910654476329,
+      "grad_norm": 0.820624589920044,
+      "learning_rate": 2.4300494434824373e-05,
+      "loss": 2.6291,
+      "step": 7756
+    },
+    {
+      "epoch": 0.7021815877613832,
+      "grad_norm": 0.8468902707099915,
+      "learning_rate": 2.4279763081295993e-05,
+      "loss": 2.6783,
+      "step": 7757
+    },
+    {
+      "epoch": 0.7022721100751336,
+      "grad_norm": 0.8257344365119934,
+      "learning_rate": 2.42590393527934e-05,
+      "loss": 2.5901,
+      "step": 7758
+    },
+    {
+      "epoch": 0.7023626323888839,
+      "grad_norm": 0.8577067852020264,
+      "learning_rate": 2.4238323251403473e-05,
+      "loss": 2.6071,
+      "step": 7759
+    },
+    {
+      "epoch": 0.7024531547026343,
+      "grad_norm": 0.8569862246513367,
+      "learning_rate": 2.4217614779212315e-05,
+      "loss": 2.6367,
+      "step": 7760
+    },
+    {
+      "epoch": 0.7025436770163845,
+      "grad_norm": 0.9399639964103699,
+      "learning_rate": 2.419691393830529e-05,
+      "loss": 2.5941,
+      "step": 7761
+    },
+    {
+      "epoch": 0.7026341993301348,
+      "grad_norm": 0.8270482420921326,
+      "learning_rate": 2.4176220730766974e-05,
+      "loss": 2.6653,
+      "step": 7762
+    },
+    {
+      "epoch": 0.7027247216438852,
+      "grad_norm": 0.8636314868927002,
+      "learning_rate": 2.4155535158681098e-05,
+      "loss": 2.693,
+      "step": 7763
+    },
+    {
+      "epoch": 0.7028152439576355,
+      "grad_norm": 0.9459649920463562,
+      "learning_rate": 2.4134857224130814e-05,
+      "loss": 2.648,
+      "step": 7764
+    },
+    {
+      "epoch": 0.7029057662713859,
+      "grad_norm": 0.9734020829200745,
+      "learning_rate": 2.4114186929198313e-05,
+      "loss": 2.6417,
+      "step": 7765
+    },
+    {
+      "epoch": 0.7029962885851362,
+      "grad_norm": 0.8456811904907227,
+      "learning_rate": 2.4093524275965097e-05,
+      "loss": 2.6391,
+      "step": 7766
+    },
+    {
+      "epoch": 0.7030868108988866,
+      "grad_norm": 0.8550230860710144,
+      "learning_rate": 2.407286926651192e-05,
+      "loss": 2.6139,
+      "step": 7767
+    },
+    {
+      "epoch": 0.7031773332126369,
+      "grad_norm": 0.8369646668434143,
+      "learning_rate": 2.4052221902918725e-05,
+      "loss": 2.6357,
+      "step": 7768
+    },
+    {
+      "epoch": 0.7032678555263873,
+      "grad_norm": 0.8603420853614807,
+      "learning_rate": 2.4031582187264713e-05,
+      "loss": 2.6607,
+      "step": 7769
+    },
+    {
+      "epoch": 0.7033583778401375,
+      "grad_norm": 0.9332559108734131,
+      "learning_rate": 2.4010950121628318e-05,
+      "loss": 2.6348,
+      "step": 7770
+    },
+    {
+      "epoch": 0.703448900153888,
+      "grad_norm": 0.7975991368293762,
+      "learning_rate": 2.3990325708087114e-05,
+      "loss": 2.621,
+      "step": 7771
+    },
+    {
+      "epoch": 0.7035394224676382,
+      "grad_norm": 0.816978931427002,
+      "learning_rate": 2.3969708948718084e-05,
+      "loss": 2.5824,
+      "step": 7772
+    },
+    {
+      "epoch": 0.7036299447813886,
+      "grad_norm": 0.9127275347709656,
+      "learning_rate": 2.3949099845597255e-05,
+      "loss": 2.6732,
+      "step": 7773
+    },
+    {
+      "epoch": 0.7037204670951389,
+      "grad_norm": 0.8123887181282043,
+      "learning_rate": 2.3928498400799993e-05,
+      "loss": 2.6204,
+      "step": 7774
+    },
+    {
+      "epoch": 0.7038109894088893,
+      "grad_norm": 0.871410608291626,
+      "learning_rate": 2.3907904616400854e-05,
+      "loss": 2.6454,
+      "step": 7775
+    },
+    {
+      "epoch": 0.7039015117226396,
+      "grad_norm": 0.8844181895256042,
+      "learning_rate": 2.3887318494473677e-05,
+      "loss": 2.6197,
+      "step": 7776
+    },
+    {
+      "epoch": 0.70399203403639,
+      "grad_norm": 0.845853865146637,
+      "learning_rate": 2.3866740037091386e-05,
+      "loss": 2.6927,
+      "step": 7777
+    },
+    {
+      "epoch": 0.7040825563501403,
+      "grad_norm": 0.8261255621910095,
+      "learning_rate": 2.3846169246326343e-05,
+      "loss": 2.666,
+      "step": 7778
+    },
+    {
+      "epoch": 0.7041730786638907,
+      "grad_norm": 0.8859426379203796,
+      "learning_rate": 2.3825606124249942e-05,
+      "loss": 2.5974,
+      "step": 7779
+    },
+    {
+      "epoch": 0.704263600977641,
+      "grad_norm": 0.8192964792251587,
+      "learning_rate": 2.3805050672932928e-05,
+      "loss": 2.6081,
+      "step": 7780
+    },
+    {
+      "epoch": 0.7043541232913914,
+      "grad_norm": 0.7949223518371582,
+      "learning_rate": 2.3784502894445214e-05,
+      "loss": 2.6564,
+      "step": 7781
+    },
+    {
+      "epoch": 0.7044446456051416,
+      "grad_norm": 0.8724743723869324,
+      "learning_rate": 2.376396279085599e-05,
+      "loss": 2.5728,
+      "step": 7782
+    },
+    {
+      "epoch": 0.704535167918892,
+      "grad_norm": 0.881767749786377,
+      "learning_rate": 2.3743430364233643e-05,
+      "loss": 2.6532,
+      "step": 7783
+    },
+    {
+      "epoch": 0.7046256902326423,
+      "grad_norm": 0.8147710561752319,
+      "learning_rate": 2.3722905616645752e-05,
+      "loss": 2.601,
+      "step": 7784
+    },
+    {
+      "epoch": 0.7047162125463927,
+      "grad_norm": 0.9021417498588562,
+      "learning_rate": 2.370238855015917e-05,
+      "loss": 2.6329,
+      "step": 7785
+    },
+    {
+      "epoch": 0.704806734860143,
+      "grad_norm": 0.8347455263137817,
+      "learning_rate": 2.368187916683997e-05,
+      "loss": 2.5916,
+      "step": 7786
+    },
+    {
+      "epoch": 0.7048972571738934,
+      "grad_norm": 0.8550873398780823,
+      "learning_rate": 2.3661377468753465e-05,
+      "loss": 2.6595,
+      "step": 7787
+    },
+    {
+      "epoch": 0.7049877794876437,
+      "grad_norm": 0.8691414594650269,
+      "learning_rate": 2.3640883457964148e-05,
+      "loss": 2.6136,
+      "step": 7788
+    },
+    {
+      "epoch": 0.7050783018013941,
+      "grad_norm": 0.8594337105751038,
+      "learning_rate": 2.362039713653581e-05,
+      "loss": 2.6274,
+      "step": 7789
+    },
+    {
+      "epoch": 0.7051688241151444,
+      "grad_norm": 0.9055542349815369,
+      "learning_rate": 2.3599918506531337e-05,
+      "loss": 2.657,
+      "step": 7790
+    },
+    {
+      "epoch": 0.7052593464288948,
+      "grad_norm": 0.9172478914260864,
+      "learning_rate": 2.3579447570013026e-05,
+      "loss": 2.7091,
+      "step": 7791
+    },
+    {
+      "epoch": 0.705349868742645,
+      "grad_norm": 0.896195650100708,
+      "learning_rate": 2.3558984329042233e-05,
+      "loss": 2.6851,
+      "step": 7792
+    },
+    {
+      "epoch": 0.7054403910563954,
+      "grad_norm": 0.8319085836410522,
+      "learning_rate": 2.3538528785679625e-05,
+      "loss": 2.5979,
+      "step": 7793
+    },
+    {
+      "epoch": 0.7055309133701457,
+      "grad_norm": 0.8286898732185364,
+      "learning_rate": 2.3518080941985076e-05,
+      "loss": 2.6195,
+      "step": 7794
+    },
+    {
+      "epoch": 0.7056214356838961,
+      "grad_norm": 0.8543409705162048,
+      "learning_rate": 2.3497640800017685e-05,
+      "loss": 2.6256,
+      "step": 7795
+    },
+    {
+      "epoch": 0.7057119579976464,
+      "grad_norm": 0.8427583575248718,
+      "learning_rate": 2.347720836183578e-05,
+      "loss": 2.6666,
+      "step": 7796
+    },
+    {
+      "epoch": 0.7058024803113968,
+      "grad_norm": 0.8994350433349609,
+      "learning_rate": 2.3456783629496915e-05,
+      "loss": 2.641,
+      "step": 7797
+    },
+    {
+      "epoch": 0.7058930026251471,
+      "grad_norm": 0.862696647644043,
+      "learning_rate": 2.343636660505779e-05,
+      "loss": 2.6623,
+      "step": 7798
+    },
+    {
+      "epoch": 0.7059835249388975,
+      "grad_norm": 0.9399603605270386,
+      "learning_rate": 2.341595729057451e-05,
+      "loss": 2.6399,
+      "step": 7799
+    },
+    {
+      "epoch": 0.7060740472526478,
+      "grad_norm": 0.8926723599433899,
+      "learning_rate": 2.339555568810221e-05,
+      "loss": 2.5605,
+      "step": 7800
+    },
+    {
+      "epoch": 0.7061645695663982,
+      "grad_norm": 0.8574604392051697,
+      "learning_rate": 2.3375161799695355e-05,
+      "loss": 2.6727,
+      "step": 7801
+    },
+    {
+      "epoch": 0.7062550918801485,
+      "grad_norm": 0.82745361328125,
+      "learning_rate": 2.335477562740761e-05,
+      "loss": 2.5692,
+      "step": 7802
+    },
+    {
+      "epoch": 0.7063456141938987,
+      "grad_norm": 0.8518375754356384,
+      "learning_rate": 2.3334397173291855e-05,
+      "loss": 2.6753,
+      "step": 7803
+    },
+    {
+      "epoch": 0.7064361365076491,
+      "grad_norm": 0.8378230333328247,
+      "learning_rate": 2.3314026439400217e-05,
+      "loss": 2.6279,
+      "step": 7804
+    },
+    {
+      "epoch": 0.7065266588213994,
+      "grad_norm": 0.850548505783081,
+      "learning_rate": 2.329366342778404e-05,
+      "loss": 2.6488,
+      "step": 7805
+    },
+    {
+      "epoch": 0.7066171811351498,
+      "grad_norm": 0.8931477069854736,
+      "learning_rate": 2.327330814049381e-05,
+      "loss": 2.7017,
+      "step": 7806
+    },
+    {
+      "epoch": 0.7067077034489001,
+      "grad_norm": 0.8398114442825317,
+      "learning_rate": 2.3252960579579397e-05,
+      "loss": 2.6148,
+      "step": 7807
+    },
+    {
+      "epoch": 0.7067982257626505,
+      "grad_norm": 0.9468042254447937,
+      "learning_rate": 2.3232620747089727e-05,
+      "loss": 2.6562,
+      "step": 7808
+    },
+    {
+      "epoch": 0.7068887480764008,
+      "grad_norm": 0.8379910588264465,
+      "learning_rate": 2.3212288645073054e-05,
+      "loss": 2.675,
+      "step": 7809
+    },
+    {
+      "epoch": 0.7069792703901512,
+      "grad_norm": 0.9158812165260315,
+      "learning_rate": 2.3191964275576805e-05,
+      "loss": 2.6877,
+      "step": 7810
+    },
+    {
+      "epoch": 0.7070697927039015,
+      "grad_norm": 0.8671376705169678,
+      "learning_rate": 2.3171647640647687e-05,
+      "loss": 2.5982,
+      "step": 7811
+    },
+    {
+      "epoch": 0.7071603150176519,
+      "grad_norm": 0.8869242668151855,
+      "learning_rate": 2.3151338742331496e-05,
+      "loss": 2.5712,
+      "step": 7812
+    },
+    {
+      "epoch": 0.7072508373314021,
+      "grad_norm": 0.7991974353790283,
+      "learning_rate": 2.3131037582673443e-05,
+      "loss": 2.6408,
+      "step": 7813
+    },
+    {
+      "epoch": 0.7073413596451525,
+      "grad_norm": 0.853305995464325,
+      "learning_rate": 2.3110744163717766e-05,
+      "loss": 2.6105,
+      "step": 7814
+    },
+    {
+      "epoch": 0.7074318819589028,
+      "grad_norm": 0.8013371825218201,
+      "learning_rate": 2.309045848750806e-05,
+      "loss": 2.6922,
+      "step": 7815
+    },
+    {
+      "epoch": 0.7075224042726532,
+      "grad_norm": 0.9263652563095093,
+      "learning_rate": 2.3070180556087074e-05,
+      "loss": 2.65,
+      "step": 7816
+    },
+    {
+      "epoch": 0.7076129265864035,
+      "grad_norm": 0.8769465088844299,
+      "learning_rate": 2.3049910371496797e-05,
+      "loss": 2.5925,
+      "step": 7817
+    },
+    {
+      "epoch": 0.7077034489001539,
+      "grad_norm": 0.7965158820152283,
+      "learning_rate": 2.302964793577844e-05,
+      "loss": 2.5404,
+      "step": 7818
+    },
+    {
+      "epoch": 0.7077939712139042,
+      "grad_norm": 0.8006182909011841,
+      "learning_rate": 2.3009393250972465e-05,
+      "loss": 2.6565,
+      "step": 7819
+    },
+    {
+      "epoch": 0.7078844935276546,
+      "grad_norm": 0.8500511050224304,
+      "learning_rate": 2.2989146319118425e-05,
+      "loss": 2.6867,
+      "step": 7820
+    },
+    {
+      "epoch": 0.7079750158414049,
+      "grad_norm": 0.9117669463157654,
+      "learning_rate": 2.2968907142255303e-05,
+      "loss": 2.658,
+      "step": 7821
+    },
+    {
+      "epoch": 0.7080655381551553,
+      "grad_norm": 0.9068896174430847,
+      "learning_rate": 2.2948675722421086e-05,
+      "loss": 2.6569,
+      "step": 7822
+    },
+    {
+      "epoch": 0.7081560604689056,
+      "grad_norm": 0.8441785573959351,
+      "learning_rate": 2.2928452061653127e-05,
+      "loss": 2.6124,
+      "step": 7823
+    },
+    {
+      "epoch": 0.708246582782656,
+      "grad_norm": 0.8266540169715881,
+      "learning_rate": 2.290823616198793e-05,
+      "loss": 2.6104,
+      "step": 7824
+    },
+    {
+      "epoch": 0.7083371050964062,
+      "grad_norm": 0.8758473992347717,
+      "learning_rate": 2.288802802546124e-05,
+      "loss": 2.5885,
+      "step": 7825
+    },
+    {
+      "epoch": 0.7084276274101566,
+      "grad_norm": 0.8110272884368896,
+      "learning_rate": 2.286782765410802e-05,
+      "loss": 2.6885,
+      "step": 7826
+    },
+    {
+      "epoch": 0.7085181497239069,
+      "grad_norm": 0.860407829284668,
+      "learning_rate": 2.284763504996248e-05,
+      "loss": 2.7112,
+      "step": 7827
+    },
+    {
+      "epoch": 0.7086086720376573,
+      "grad_norm": 0.8285262584686279,
+      "learning_rate": 2.282745021505791e-05,
+      "loss": 2.6778,
+      "step": 7828
+    },
+    {
+      "epoch": 0.7086991943514076,
+      "grad_norm": 0.8299581408500671,
+      "learning_rate": 2.2807273151427055e-05,
+      "loss": 2.5845,
+      "step": 7829
+    },
+    {
+      "epoch": 0.708789716665158,
+      "grad_norm": 0.808203935623169,
+      "learning_rate": 2.2787103861101655e-05,
+      "loss": 2.6503,
+      "step": 7830
+    },
+    {
+      "epoch": 0.7088802389789083,
+      "grad_norm": 0.9187443256378174,
+      "learning_rate": 2.2766942346112786e-05,
+      "loss": 2.6364,
+      "step": 7831
+    },
+    {
+      "epoch": 0.7089707612926587,
+      "grad_norm": 0.8716665506362915,
+      "learning_rate": 2.27467886084907e-05,
+      "loss": 2.6456,
+      "step": 7832
+    },
+    {
+      "epoch": 0.709061283606409,
+      "grad_norm": 0.8400218486785889,
+      "learning_rate": 2.2726642650264895e-05,
+      "loss": 2.6181,
+      "step": 7833
+    },
+    {
+      "epoch": 0.7091518059201594,
+      "grad_norm": 0.9854234457015991,
+      "learning_rate": 2.2706504473464063e-05,
+      "loss": 2.6259,
+      "step": 7834
+    },
+    {
+      "epoch": 0.7092423282339096,
+      "grad_norm": 1.016037106513977,
+      "learning_rate": 2.2686374080116136e-05,
+      "loss": 2.6556,
+      "step": 7835
+    },
+    {
+      "epoch": 0.70933285054766,
+      "grad_norm": 0.8779817819595337,
+      "learning_rate": 2.266625147224817e-05,
+      "loss": 2.655,
+      "step": 7836
+    },
+    {
+      "epoch": 0.7094233728614103,
+      "grad_norm": 0.8986814022064209,
+      "learning_rate": 2.2646136651886617e-05,
+      "loss": 2.6691,
+      "step": 7837
+    },
+    {
+      "epoch": 0.7095138951751607,
+      "grad_norm": 0.9211919903755188,
+      "learning_rate": 2.2626029621056967e-05,
+      "loss": 2.6743,
+      "step": 7838
+    },
+    {
+      "epoch": 0.709604417488911,
+      "grad_norm": 0.9180659651756287,
+      "learning_rate": 2.2605930381784013e-05,
+      "loss": 2.6589,
+      "step": 7839
+    },
+    {
+      "epoch": 0.7096949398026614,
+      "grad_norm": 0.8097158670425415,
+      "learning_rate": 2.2585838936091754e-05,
+      "loss": 2.5478,
+      "step": 7840
+    },
+    {
+      "epoch": 0.7097854621164117,
+      "grad_norm": 0.8261932134628296,
+      "learning_rate": 2.256575528600342e-05,
+      "loss": 2.6354,
+      "step": 7841
+    },
+    {
+      "epoch": 0.7098759844301621,
+      "grad_norm": 0.903597354888916,
+      "learning_rate": 2.2545679433541355e-05,
+      "loss": 2.6254,
+      "step": 7842
+    },
+    {
+      "epoch": 0.7099665067439124,
+      "grad_norm": 0.9137170910835266,
+      "learning_rate": 2.2525611380727307e-05,
+      "loss": 2.6587,
+      "step": 7843
+    },
+    {
+      "epoch": 0.7100570290576627,
+      "grad_norm": 0.8651295304298401,
+      "learning_rate": 2.2505551129582047e-05,
+      "loss": 2.6701,
+      "step": 7844
+    },
+    {
+      "epoch": 0.710147551371413,
+      "grad_norm": 0.8841049075126648,
+      "learning_rate": 2.2485498682125673e-05,
+      "loss": 2.6473,
+      "step": 7845
+    },
+    {
+      "epoch": 0.7102380736851633,
+      "grad_norm": 0.8099082708358765,
+      "learning_rate": 2.2465454040377455e-05,
+      "loss": 2.6234,
+      "step": 7846
+    },
+    {
+      "epoch": 0.7103285959989137,
+      "grad_norm": 0.8708997368812561,
+      "learning_rate": 2.24454172063559e-05,
+      "loss": 2.5894,
+      "step": 7847
+    },
+    {
+      "epoch": 0.710419118312664,
+      "grad_norm": 0.8280975222587585,
+      "learning_rate": 2.2425388182078698e-05,
+      "loss": 2.5893,
+      "step": 7848
+    },
+    {
+      "epoch": 0.7105096406264144,
+      "grad_norm": 0.79764723777771,
+      "learning_rate": 2.2405366969562823e-05,
+      "loss": 2.6345,
+      "step": 7849
+    },
+    {
+      "epoch": 0.7106001629401647,
+      "grad_norm": 0.8236714005470276,
+      "learning_rate": 2.2385353570824308e-05,
+      "loss": 2.5931,
+      "step": 7850
+    },
+    {
+      "epoch": 0.7106906852539151,
+      "grad_norm": 0.8521032929420471,
+      "learning_rate": 2.2365347987878625e-05,
+      "loss": 2.6221,
+      "step": 7851
+    },
+    {
+      "epoch": 0.7107812075676654,
+      "grad_norm": 0.8951077461242676,
+      "learning_rate": 2.2345350222740247e-05,
+      "loss": 2.6962,
+      "step": 7852
+    },
+    {
+      "epoch": 0.7108717298814158,
+      "grad_norm": 0.8242161273956299,
+      "learning_rate": 2.2325360277422967e-05,
+      "loss": 2.5837,
+      "step": 7853
+    },
+    {
+      "epoch": 0.7109622521951661,
+      "grad_norm": 0.8123531341552734,
+      "learning_rate": 2.2305378153939793e-05,
+      "loss": 2.6381,
+      "step": 7854
+    },
+    {
+      "epoch": 0.7110527745089165,
+      "grad_norm": 0.8669412732124329,
+      "learning_rate": 2.2285403854302912e-05,
+      "loss": 2.6961,
+      "step": 7855
+    },
+    {
+      "epoch": 0.7111432968226667,
+      "grad_norm": 0.8449344635009766,
+      "learning_rate": 2.2265437380523734e-05,
+      "loss": 2.6054,
+      "step": 7856
+    },
+    {
+      "epoch": 0.7112338191364171,
+      "grad_norm": 0.8434821963310242,
+      "learning_rate": 2.2245478734612914e-05,
+      "loss": 2.6581,
+      "step": 7857
+    },
+    {
+      "epoch": 0.7113243414501674,
+      "grad_norm": 0.8311582207679749,
+      "learning_rate": 2.2225527918580204e-05,
+      "loss": 2.6345,
+      "step": 7858
+    },
+    {
+      "epoch": 0.7114148637639178,
+      "grad_norm": 0.8839373588562012,
+      "learning_rate": 2.2205584934434753e-05,
+      "loss": 2.63,
+      "step": 7859
+    },
+    {
+      "epoch": 0.7115053860776681,
+      "grad_norm": 0.8330554366111755,
+      "learning_rate": 2.2185649784184746e-05,
+      "loss": 2.6272,
+      "step": 7860
+    },
+    {
+      "epoch": 0.7115959083914185,
+      "grad_norm": 0.8182799220085144,
+      "learning_rate": 2.216572246983768e-05,
+      "loss": 2.655,
+      "step": 7861
+    },
+    {
+      "epoch": 0.7116864307051688,
+      "grad_norm": 0.9063050746917725,
+      "learning_rate": 2.2145802993400223e-05,
+      "loss": 2.6717,
+      "step": 7862
+    },
+    {
+      "epoch": 0.7117769530189192,
+      "grad_norm": 0.8399935960769653,
+      "learning_rate": 2.212589135687828e-05,
+      "loss": 2.677,
+      "step": 7863
+    },
+    {
+      "epoch": 0.7118674753326695,
+      "grad_norm": 0.9330717921257019,
+      "learning_rate": 2.2105987562276953e-05,
+      "loss": 2.7948,
+      "step": 7864
+    },
+    {
+      "epoch": 0.7119579976464199,
+      "grad_norm": 0.8137609362602234,
+      "learning_rate": 2.2086091611600567e-05,
+      "loss": 2.6701,
+      "step": 7865
+    },
+    {
+      "epoch": 0.7120485199601702,
+      "grad_norm": 0.7958082556724548,
+      "learning_rate": 2.2066203506852566e-05,
+      "loss": 2.572,
+      "step": 7866
+    },
+    {
+      "epoch": 0.7121390422739206,
+      "grad_norm": 0.850689172744751,
+      "learning_rate": 2.2046323250035804e-05,
+      "loss": 2.6466,
+      "step": 7867
+    },
+    {
+      "epoch": 0.7122295645876708,
+      "grad_norm": 0.8957266807556152,
+      "learning_rate": 2.2026450843152157e-05,
+      "loss": 2.656,
+      "step": 7868
+    },
+    {
+      "epoch": 0.7123200869014212,
+      "grad_norm": 0.867348849773407,
+      "learning_rate": 2.200658628820271e-05,
+      "loss": 2.6563,
+      "step": 7869
+    },
+    {
+      "epoch": 0.7124106092151715,
+      "grad_norm": 0.833674967288971,
+      "learning_rate": 2.198672958718796e-05,
+      "loss": 2.6242,
+      "step": 7870
+    },
+    {
+      "epoch": 0.7125011315289219,
+      "grad_norm": 0.8591665625572205,
+      "learning_rate": 2.1966880742107364e-05,
+      "loss": 2.6653,
+      "step": 7871
+    },
+    {
+      "epoch": 0.7125916538426722,
+      "grad_norm": 0.8057499527931213,
+      "learning_rate": 2.1947039754959753e-05,
+      "loss": 2.5693,
+      "step": 7872
+    },
+    {
+      "epoch": 0.7126821761564226,
+      "grad_norm": 0.9254628419876099,
+      "learning_rate": 2.1927206627743102e-05,
+      "loss": 2.702,
+      "step": 7873
+    },
+    {
+      "epoch": 0.7127726984701729,
+      "grad_norm": 0.8389781713485718,
+      "learning_rate": 2.1907381362454605e-05,
+      "loss": 2.644,
+      "step": 7874
+    },
+    {
+      "epoch": 0.7128632207839233,
+      "grad_norm": 0.792872428894043,
+      "learning_rate": 2.1887563961090663e-05,
+      "loss": 2.6184,
+      "step": 7875
+    },
+    {
+      "epoch": 0.7129537430976736,
+      "grad_norm": 0.8513414859771729,
+      "learning_rate": 2.1867754425646926e-05,
+      "loss": 2.6076,
+      "step": 7876
+    },
+    {
+      "epoch": 0.713044265411424,
+      "grad_norm": 0.9116446375846863,
+      "learning_rate": 2.1847952758118117e-05,
+      "loss": 2.7136,
+      "step": 7877
+    },
+    {
+      "epoch": 0.7131347877251742,
+      "grad_norm": 0.8806061744689941,
+      "learning_rate": 2.182815896049839e-05,
+      "loss": 2.6,
+      "step": 7878
+    },
+    {
+      "epoch": 0.7132253100389246,
+      "grad_norm": 0.8659895658493042,
+      "learning_rate": 2.1808373034780893e-05,
+      "loss": 2.693,
+      "step": 7879
+    },
+    {
+      "epoch": 0.7133158323526749,
+      "grad_norm": 0.7976323962211609,
+      "learning_rate": 2.178859498295809e-05,
+      "loss": 2.6586,
+      "step": 7880
+    },
+    {
+      "epoch": 0.7134063546664253,
+      "grad_norm": 0.8185864686965942,
+      "learning_rate": 2.1768824807021624e-05,
+      "loss": 2.6354,
+      "step": 7881
+    },
+    {
+      "epoch": 0.7134968769801756,
+      "grad_norm": 0.8875325918197632,
+      "learning_rate": 2.174906250896237e-05,
+      "loss": 2.7474,
+      "step": 7882
+    },
+    {
+      "epoch": 0.713587399293926,
+      "grad_norm": 1.010949730873108,
+      "learning_rate": 2.1729308090770372e-05,
+      "loss": 2.6001,
+      "step": 7883
+    },
+    {
+      "epoch": 0.7136779216076763,
+      "grad_norm": 0.8385566473007202,
+      "learning_rate": 2.1709561554434933e-05,
+      "loss": 2.676,
+      "step": 7884
+    },
+    {
+      "epoch": 0.7137684439214266,
+      "grad_norm": 0.8849919438362122,
+      "learning_rate": 2.1689822901944457e-05,
+      "loss": 2.6793,
+      "step": 7885
+    },
+    {
+      "epoch": 0.713858966235177,
+      "grad_norm": 0.8859854340553284,
+      "learning_rate": 2.1670092135286723e-05,
+      "loss": 2.6753,
+      "step": 7886
+    },
+    {
+      "epoch": 0.7139494885489273,
+      "grad_norm": 0.8255252242088318,
+      "learning_rate": 2.1650369256448543e-05,
+      "loss": 2.6519,
+      "step": 7887
+    },
+    {
+      "epoch": 0.7140400108626777,
+      "grad_norm": 0.8009936809539795,
+      "learning_rate": 2.163065426741603e-05,
+      "loss": 2.6148,
+      "step": 7888
+    },
+    {
+      "epoch": 0.7141305331764279,
+      "grad_norm": 0.85800701379776,
+      "learning_rate": 2.1610947170174488e-05,
+      "loss": 2.617,
+      "step": 7889
+    },
+    {
+      "epoch": 0.7142210554901783,
+      "grad_norm": 0.8138502836227417,
+      "learning_rate": 2.159124796670843e-05,
+      "loss": 2.6566,
+      "step": 7890
+    },
+    {
+      "epoch": 0.7143115778039286,
+      "grad_norm": 0.8225932121276855,
+      "learning_rate": 2.1571556659001545e-05,
+      "loss": 2.6542,
+      "step": 7891
+    },
+    {
+      "epoch": 0.714402100117679,
+      "grad_norm": 0.9008784890174866,
+      "learning_rate": 2.1551873249036793e-05,
+      "loss": 2.6692,
+      "step": 7892
+    },
+    {
+      "epoch": 0.7144926224314293,
+      "grad_norm": 0.8721801042556763,
+      "learning_rate": 2.15321977387962e-05,
+      "loss": 2.71,
+      "step": 7893
+    },
+    {
+      "epoch": 0.7145831447451797,
+      "grad_norm": 0.8188298940658569,
+      "learning_rate": 2.151253013026121e-05,
+      "loss": 2.6062,
+      "step": 7894
+    },
+    {
+      "epoch": 0.71467366705893,
+      "grad_norm": 0.8610367178916931,
+      "learning_rate": 2.149287042541225e-05,
+      "loss": 2.7082,
+      "step": 7895
+    },
+    {
+      "epoch": 0.7147641893726804,
+      "grad_norm": 0.8750115633010864,
+      "learning_rate": 2.1473218626229095e-05,
+      "loss": 2.7165,
+      "step": 7896
+    },
+    {
+      "epoch": 0.7148547116864307,
+      "grad_norm": 0.827516496181488,
+      "learning_rate": 2.1453574734690672e-05,
+      "loss": 2.5671,
+      "step": 7897
+    },
+    {
+      "epoch": 0.7149452340001811,
+      "grad_norm": 0.9428842067718506,
+      "learning_rate": 2.143393875277513e-05,
+      "loss": 2.6636,
+      "step": 7898
+    },
+    {
+      "epoch": 0.7150357563139313,
+      "grad_norm": 0.8270500898361206,
+      "learning_rate": 2.1414310682459802e-05,
+      "loss": 2.6905,
+      "step": 7899
+    },
+    {
+      "epoch": 0.7151262786276817,
+      "grad_norm": 0.7811481356620789,
+      "learning_rate": 2.139469052572127e-05,
+      "loss": 2.6164,
+      "step": 7900
+    },
+    {
+      "epoch": 0.715216800941432,
+      "grad_norm": 0.8671733140945435,
+      "learning_rate": 2.13750782845352e-05,
+      "loss": 2.671,
+      "step": 7901
+    },
+    {
+      "epoch": 0.7153073232551824,
+      "grad_norm": 0.8288303017616272,
+      "learning_rate": 2.1355473960876637e-05,
+      "loss": 2.6416,
+      "step": 7902
+    },
+    {
+      "epoch": 0.7153978455689327,
+      "grad_norm": 0.8316097259521484,
+      "learning_rate": 2.1335877556719676e-05,
+      "loss": 2.6305,
+      "step": 7903
+    },
+    {
+      "epoch": 0.7154883678826831,
+      "grad_norm": 0.9389979243278503,
+      "learning_rate": 2.1316289074037687e-05,
+      "loss": 2.6453,
+      "step": 7904
+    },
+    {
+      "epoch": 0.7155788901964334,
+      "grad_norm": 0.9156227707862854,
+      "learning_rate": 2.1296708514803245e-05,
+      "loss": 2.6353,
+      "step": 7905
+    },
+    {
+      "epoch": 0.7156694125101838,
+      "grad_norm": 0.8820056319236755,
+      "learning_rate": 2.127713588098812e-05,
+      "loss": 2.6453,
+      "step": 7906
+    },
+    {
+      "epoch": 0.7157599348239341,
+      "grad_norm": 0.9267116785049438,
+      "learning_rate": 2.1257571174563206e-05,
+      "loss": 2.6122,
+      "step": 7907
+    },
+    {
+      "epoch": 0.7158504571376845,
+      "grad_norm": 0.8101881742477417,
+      "learning_rate": 2.1238014397498773e-05,
+      "loss": 2.6201,
+      "step": 7908
+    },
+    {
+      "epoch": 0.7159409794514348,
+      "grad_norm": 0.8202838897705078,
+      "learning_rate": 2.121846555176411e-05,
+      "loss": 2.6305,
+      "step": 7909
+    },
+    {
+      "epoch": 0.7160315017651852,
+      "grad_norm": 0.8424105048179626,
+      "learning_rate": 2.119892463932781e-05,
+      "loss": 2.6367,
+      "step": 7910
+    },
+    {
+      "epoch": 0.7161220240789354,
+      "grad_norm": 0.8372321724891663,
+      "learning_rate": 2.1179391662157643e-05,
+      "loss": 2.6069,
+      "step": 7911
+    },
+    {
+      "epoch": 0.7162125463926858,
+      "grad_norm": 0.8347082138061523,
+      "learning_rate": 2.115986662222058e-05,
+      "loss": 2.6203,
+      "step": 7912
+    },
+    {
+      "epoch": 0.7163030687064361,
+      "grad_norm": 0.9294868111610413,
+      "learning_rate": 2.1140349521482795e-05,
+      "loss": 2.6633,
+      "step": 7913
+    },
+    {
+      "epoch": 0.7163935910201865,
+      "grad_norm": 0.8243364691734314,
+      "learning_rate": 2.1120840361909676e-05,
+      "loss": 2.6449,
+      "step": 7914
+    },
+    {
+      "epoch": 0.7164841133339368,
+      "grad_norm": 0.8228632211685181,
+      "learning_rate": 2.1101339145465726e-05,
+      "loss": 2.6706,
+      "step": 7915
+    },
+    {
+      "epoch": 0.7165746356476872,
+      "grad_norm": 0.8205071687698364,
+      "learning_rate": 2.1081845874114815e-05,
+      "loss": 2.6565,
+      "step": 7916
+    },
+    {
+      "epoch": 0.7166651579614375,
+      "grad_norm": 0.9841361045837402,
+      "learning_rate": 2.1062360549819837e-05,
+      "loss": 2.7053,
+      "step": 7917
+    },
+    {
+      "epoch": 0.7167556802751879,
+      "grad_norm": 0.8670927882194519,
+      "learning_rate": 2.1042883174542992e-05,
+      "loss": 2.6442,
+      "step": 7918
+    },
+    {
+      "epoch": 0.7168462025889382,
+      "grad_norm": 0.8424184918403625,
+      "learning_rate": 2.1023413750245645e-05,
+      "loss": 2.6044,
+      "step": 7919
+    },
+    {
+      "epoch": 0.7169367249026886,
+      "grad_norm": 0.8506214618682861,
+      "learning_rate": 2.1003952278888382e-05,
+      "loss": 2.6469,
+      "step": 7920
+    },
+    {
+      "epoch": 0.7170272472164388,
+      "grad_norm": 0.8821026682853699,
+      "learning_rate": 2.098449876243096e-05,
+      "loss": 2.55,
+      "step": 7921
+    },
+    {
+      "epoch": 0.7171177695301892,
+      "grad_norm": 0.7802221179008484,
+      "learning_rate": 2.0965053202832375e-05,
+      "loss": 2.607,
+      "step": 7922
+    },
+    {
+      "epoch": 0.7172082918439395,
+      "grad_norm": 0.8895224928855896,
+      "learning_rate": 2.0945615602050716e-05,
+      "loss": 2.5846,
+      "step": 7923
+    },
+    {
+      "epoch": 0.7172988141576899,
+      "grad_norm": 0.8102734684944153,
+      "learning_rate": 2.0926185962043466e-05,
+      "loss": 2.5796,
+      "step": 7924
+    },
+    {
+      "epoch": 0.7173893364714402,
+      "grad_norm": 0.8378505706787109,
+      "learning_rate": 2.090676428476709e-05,
+      "loss": 2.5913,
+      "step": 7925
+    },
+    {
+      "epoch": 0.7174798587851905,
+      "grad_norm": 0.8973132371902466,
+      "learning_rate": 2.088735057217739e-05,
+      "loss": 2.6591,
+      "step": 7926
+    },
+    {
+      "epoch": 0.7175703810989409,
+      "grad_norm": 0.8435284495353699,
+      "learning_rate": 2.0867944826229324e-05,
+      "loss": 2.5978,
+      "step": 7927
+    },
+    {
+      "epoch": 0.7176609034126912,
+      "grad_norm": 0.9157330989837646,
+      "learning_rate": 2.0848547048877044e-05,
+      "loss": 2.6221,
+      "step": 7928
+    },
+    {
+      "epoch": 0.7177514257264416,
+      "grad_norm": 0.9741874933242798,
+      "learning_rate": 2.0829157242073915e-05,
+      "loss": 2.6642,
+      "step": 7929
+    },
+    {
+      "epoch": 0.7178419480401919,
+      "grad_norm": 0.8608724474906921,
+      "learning_rate": 2.0809775407772503e-05,
+      "loss": 2.629,
+      "step": 7930
+    },
+    {
+      "epoch": 0.7179324703539423,
+      "grad_norm": 0.8521125316619873,
+      "learning_rate": 2.0790401547924498e-05,
+      "loss": 2.6475,
+      "step": 7931
+    },
+    {
+      "epoch": 0.7180229926676925,
+      "grad_norm": 0.8497169017791748,
+      "learning_rate": 2.0771035664480942e-05,
+      "loss": 2.6299,
+      "step": 7932
+    },
+    {
+      "epoch": 0.7181135149814429,
+      "grad_norm": 0.8564878106117249,
+      "learning_rate": 2.0751677759391895e-05,
+      "loss": 2.7099,
+      "step": 7933
+    },
+    {
+      "epoch": 0.7182040372951932,
+      "grad_norm": 0.8833389282226562,
+      "learning_rate": 2.0732327834606735e-05,
+      "loss": 2.6199,
+      "step": 7934
+    },
+    {
+      "epoch": 0.7182945596089436,
+      "grad_norm": 0.9143378138542175,
+      "learning_rate": 2.0712985892073987e-05,
+      "loss": 2.6486,
+      "step": 7935
+    },
+    {
+      "epoch": 0.7183850819226939,
+      "grad_norm": 0.8474334478378296,
+      "learning_rate": 2.069365193374142e-05,
+      "loss": 2.636,
+      "step": 7936
+    },
+    {
+      "epoch": 0.7184756042364443,
+      "grad_norm": 0.8107338547706604,
+      "learning_rate": 2.0674325961555872e-05,
+      "loss": 2.6333,
+      "step": 7937
+    },
+    {
+      "epoch": 0.7185661265501946,
+      "grad_norm": 0.8914929032325745,
+      "learning_rate": 2.065500797746359e-05,
+      "loss": 2.6174,
+      "step": 7938
+    },
+    {
+      "epoch": 0.718656648863945,
+      "grad_norm": 0.8551090359687805,
+      "learning_rate": 2.0635697983409817e-05,
+      "loss": 2.6287,
+      "step": 7939
+    },
+    {
+      "epoch": 0.7187471711776953,
+      "grad_norm": 0.8404701352119446,
+      "learning_rate": 2.0616395981339075e-05,
+      "loss": 2.5802,
+      "step": 7940
+    },
+    {
+      "epoch": 0.7188376934914457,
+      "grad_norm": 0.8328677415847778,
+      "learning_rate": 2.059710197319509e-05,
+      "loss": 2.6099,
+      "step": 7941
+    },
+    {
+      "epoch": 0.718928215805196,
+      "grad_norm": 0.8687942028045654,
+      "learning_rate": 2.0577815960920754e-05,
+      "loss": 2.6319,
+      "step": 7942
+    },
+    {
+      "epoch": 0.7190187381189463,
+      "grad_norm": 0.8320063948631287,
+      "learning_rate": 2.0558537946458177e-05,
+      "loss": 2.6597,
+      "step": 7943
+    },
+    {
+      "epoch": 0.7191092604326966,
+      "grad_norm": 0.835174560546875,
+      "learning_rate": 2.0539267931748695e-05,
+      "loss": 2.6207,
+      "step": 7944
+    },
+    {
+      "epoch": 0.719199782746447,
+      "grad_norm": 0.7959899306297302,
+      "learning_rate": 2.0520005918732698e-05,
+      "loss": 2.5987,
+      "step": 7945
+    },
+    {
+      "epoch": 0.7192903050601973,
+      "grad_norm": 0.8572633862495422,
+      "learning_rate": 2.050075190934998e-05,
+      "loss": 2.6462,
+      "step": 7946
+    },
+    {
+      "epoch": 0.7193808273739477,
+      "grad_norm": 0.8058256506919861,
+      "learning_rate": 2.0481505905539357e-05,
+      "loss": 2.6276,
+      "step": 7947
+    },
+    {
+      "epoch": 0.719471349687698,
+      "grad_norm": 0.8863130807876587,
+      "learning_rate": 2.0462267909238896e-05,
+      "loss": 2.6207,
+      "step": 7948
+    },
+    {
+      "epoch": 0.7195618720014484,
+      "grad_norm": 0.808664083480835,
+      "learning_rate": 2.0443037922385888e-05,
+      "loss": 2.6078,
+      "step": 7949
+    },
+    {
+      "epoch": 0.7196523943151987,
+      "grad_norm": 0.9259856343269348,
+      "learning_rate": 2.042381594691678e-05,
+      "loss": 2.672,
+      "step": 7950
+    },
+    {
+      "epoch": 0.7197429166289491,
+      "grad_norm": 0.8126853704452515,
+      "learning_rate": 2.0404601984767234e-05,
+      "loss": 2.6178,
+      "step": 7951
+    },
+    {
+      "epoch": 0.7198334389426994,
+      "grad_norm": 0.8391700387001038,
+      "learning_rate": 2.038539603787212e-05,
+      "loss": 2.608,
+      "step": 7952
+    },
+    {
+      "epoch": 0.7199239612564498,
+      "grad_norm": 0.8788531422615051,
+      "learning_rate": 2.0366198108165412e-05,
+      "loss": 2.6365,
+      "step": 7953
+    },
+    {
+      "epoch": 0.7200144835702,
+      "grad_norm": 0.8273787498474121,
+      "learning_rate": 2.0347008197580374e-05,
+      "loss": 2.6462,
+      "step": 7954
+    },
+    {
+      "epoch": 0.7201050058839504,
+      "grad_norm": 0.8949300646781921,
+      "learning_rate": 2.0327826308049447e-05,
+      "loss": 2.5884,
+      "step": 7955
+    },
+    {
+      "epoch": 0.7201955281977007,
+      "grad_norm": 0.8484796285629272,
+      "learning_rate": 2.0308652441504217e-05,
+      "loss": 2.6097,
+      "step": 7956
+    },
+    {
+      "epoch": 0.7202860505114511,
+      "grad_norm": 0.8365808129310608,
+      "learning_rate": 2.0289486599875528e-05,
+      "loss": 2.6805,
+      "step": 7957
+    },
+    {
+      "epoch": 0.7203765728252014,
+      "grad_norm": 0.8657910823822021,
+      "learning_rate": 2.027032878509332e-05,
+      "loss": 2.6297,
+      "step": 7958
+    },
+    {
+      "epoch": 0.7204670951389518,
+      "grad_norm": 0.8437427878379822,
+      "learning_rate": 2.025117899908686e-05,
+      "loss": 2.6447,
+      "step": 7959
+    },
+    {
+      "epoch": 0.7205576174527021,
+      "grad_norm": 0.8482447266578674,
+      "learning_rate": 2.0232037243784475e-05,
+      "loss": 2.6771,
+      "step": 7960
+    },
+    {
+      "epoch": 0.7206481397664525,
+      "grad_norm": 0.8186309933662415,
+      "learning_rate": 2.0212903521113757e-05,
+      "loss": 2.6535,
+      "step": 7961
+    },
+    {
+      "epoch": 0.7207386620802028,
+      "grad_norm": 0.8866466283798218,
+      "learning_rate": 2.019377783300147e-05,
+      "loss": 2.6354,
+      "step": 7962
+    },
+    {
+      "epoch": 0.7208291843939532,
+      "grad_norm": 0.8010830283164978,
+      "learning_rate": 2.0174660181373595e-05,
+      "loss": 2.6233,
+      "step": 7963
+    },
+    {
+      "epoch": 0.7209197067077034,
+      "grad_norm": 0.9101557731628418,
+      "learning_rate": 2.01555505681552e-05,
+      "loss": 2.7038,
+      "step": 7964
+    },
+    {
+      "epoch": 0.7210102290214538,
+      "grad_norm": 0.8546535968780518,
+      "learning_rate": 2.013644899527074e-05,
+      "loss": 2.6985,
+      "step": 7965
+    },
+    {
+      "epoch": 0.7211007513352041,
+      "grad_norm": 0.8595011234283447,
+      "learning_rate": 2.011735546464365e-05,
+      "loss": 2.6894,
+      "step": 7966
+    },
+    {
+      "epoch": 0.7211912736489544,
+      "grad_norm": 0.8850213289260864,
+      "learning_rate": 2.0098269978196672e-05,
+      "loss": 2.6243,
+      "step": 7967
+    },
+    {
+      "epoch": 0.7212817959627048,
+      "grad_norm": 0.8743082284927368,
+      "learning_rate": 2.0079192537851733e-05,
+      "loss": 2.6324,
+      "step": 7968
+    },
+    {
+      "epoch": 0.7213723182764551,
+      "grad_norm": 0.8616811633110046,
+      "learning_rate": 2.0060123145529907e-05,
+      "loss": 2.6398,
+      "step": 7969
+    },
+    {
+      "epoch": 0.7214628405902055,
+      "grad_norm": 0.901737630367279,
+      "learning_rate": 2.0041061803151508e-05,
+      "loss": 2.6913,
+      "step": 7970
+    },
+    {
+      "epoch": 0.7215533629039558,
+      "grad_norm": 0.9253664612770081,
+      "learning_rate": 2.0022008512636016e-05,
+      "loss": 2.6259,
+      "step": 7971
+    },
+    {
+      "epoch": 0.7216438852177062,
+      "grad_norm": 0.8385868072509766,
+      "learning_rate": 2.000296327590202e-05,
+      "loss": 2.636,
+      "step": 7972
+    },
+    {
+      "epoch": 0.7217344075314565,
+      "grad_norm": 0.8327115178108215,
+      "learning_rate": 1.9983926094867488e-05,
+      "loss": 2.5869,
+      "step": 7973
+    },
+    {
+      "epoch": 0.7218249298452069,
+      "grad_norm": 0.8412270545959473,
+      "learning_rate": 1.9964896971449387e-05,
+      "loss": 2.6467,
+      "step": 7974
+    },
+    {
+      "epoch": 0.7219154521589571,
+      "grad_norm": 0.8899946808815002,
+      "learning_rate": 1.994587590756397e-05,
+      "loss": 2.6581,
+      "step": 7975
+    },
+    {
+      "epoch": 0.7220059744727075,
+      "grad_norm": 0.8444313406944275,
+      "learning_rate": 1.9926862905126665e-05,
+      "loss": 2.5909,
+      "step": 7976
+    },
+    {
+      "epoch": 0.7220964967864578,
+      "grad_norm": 0.8593295812606812,
+      "learning_rate": 1.9907857966052056e-05,
+      "loss": 2.6534,
+      "step": 7977
+    },
+    {
+      "epoch": 0.7221870191002082,
+      "grad_norm": 0.811881422996521,
+      "learning_rate": 1.9888861092253975e-05,
+      "loss": 2.6254,
+      "step": 7978
+    },
+    {
+      "epoch": 0.7222775414139585,
+      "grad_norm": 0.8473266363143921,
+      "learning_rate": 1.9869872285645406e-05,
+      "loss": 2.6193,
+      "step": 7979
+    },
+    {
+      "epoch": 0.7223680637277089,
+      "grad_norm": 0.8891638517379761,
+      "learning_rate": 1.985089154813846e-05,
+      "loss": 2.5915,
+      "step": 7980
+    },
+    {
+      "epoch": 0.7224585860414592,
+      "grad_norm": 0.8436593413352966,
+      "learning_rate": 1.9831918881644595e-05,
+      "loss": 2.6799,
+      "step": 7981
+    },
+    {
+      "epoch": 0.7225491083552096,
+      "grad_norm": 0.9255457520484924,
+      "learning_rate": 1.981295428807428e-05,
+      "loss": 2.6251,
+      "step": 7982
+    },
+    {
+      "epoch": 0.7226396306689599,
+      "grad_norm": 0.8497655987739563,
+      "learning_rate": 1.979399776933727e-05,
+      "loss": 2.603,
+      "step": 7983
+    },
+    {
+      "epoch": 0.7227301529827103,
+      "grad_norm": 0.8996449708938599,
+      "learning_rate": 1.9775049327342486e-05,
+      "loss": 2.7178,
+      "step": 7984
+    },
+    {
+      "epoch": 0.7228206752964605,
+      "grad_norm": 0.8296543955802917,
+      "learning_rate": 1.9756108963998054e-05,
+      "loss": 2.6355,
+      "step": 7985
+    },
+    {
+      "epoch": 0.7229111976102109,
+      "grad_norm": 0.843795120716095,
+      "learning_rate": 1.9737176681211244e-05,
+      "loss": 2.6516,
+      "step": 7986
+    },
+    {
+      "epoch": 0.7230017199239612,
+      "grad_norm": 0.8656344413757324,
+      "learning_rate": 1.9718252480888566e-05,
+      "loss": 2.577,
+      "step": 7987
+    },
+    {
+      "epoch": 0.7230922422377116,
+      "grad_norm": 0.8967494964599609,
+      "learning_rate": 1.9699336364935615e-05,
+      "loss": 2.623,
+      "step": 7988
+    },
+    {
+      "epoch": 0.7231827645514619,
+      "grad_norm": 0.8973367810249329,
+      "learning_rate": 1.968042833525736e-05,
+      "loss": 2.6147,
+      "step": 7989
+    },
+    {
+      "epoch": 0.7232732868652123,
+      "grad_norm": 0.831631064414978,
+      "learning_rate": 1.9661528393757744e-05,
+      "loss": 2.6482,
+      "step": 7990
+    },
+    {
+      "epoch": 0.7233638091789626,
+      "grad_norm": 0.8482058048248291,
+      "learning_rate": 1.964263654234001e-05,
+      "loss": 2.5898,
+      "step": 7991
+    },
+    {
+      "epoch": 0.723454331492713,
+      "grad_norm": 0.816674530506134,
+      "learning_rate": 1.962375278290659e-05,
+      "loss": 2.6284,
+      "step": 7992
+    },
+    {
+      "epoch": 0.7235448538064633,
+      "grad_norm": 0.9323936104774475,
+      "learning_rate": 1.9604877117359054e-05,
+      "loss": 2.6328,
+      "step": 7993
+    },
+    {
+      "epoch": 0.7236353761202137,
+      "grad_norm": 0.8940310478210449,
+      "learning_rate": 1.9586009547598205e-05,
+      "loss": 2.6816,
+      "step": 7994
+    },
+    {
+      "epoch": 0.723725898433964,
+      "grad_norm": 0.965698778629303,
+      "learning_rate": 1.956715007552401e-05,
+      "loss": 2.5628,
+      "step": 7995
+    },
+    {
+      "epoch": 0.7238164207477144,
+      "grad_norm": 0.8234290480613708,
+      "learning_rate": 1.954829870303555e-05,
+      "loss": 2.6205,
+      "step": 7996
+    },
+    {
+      "epoch": 0.7239069430614646,
+      "grad_norm": 0.8232055902481079,
+      "learning_rate": 1.9529455432031273e-05,
+      "loss": 2.6669,
+      "step": 7997
+    },
+    {
+      "epoch": 0.723997465375215,
+      "grad_norm": 0.8574815392494202,
+      "learning_rate": 1.9510620264408596e-05,
+      "loss": 2.6661,
+      "step": 7998
+    },
+    {
+      "epoch": 0.7240879876889653,
+      "grad_norm": 0.880766749382019,
+      "learning_rate": 1.9491793202064256e-05,
+      "loss": 2.6692,
+      "step": 7999
+    },
+    {
+      "epoch": 0.7241785100027157,
+      "grad_norm": 0.8221156597137451,
+      "learning_rate": 1.947297424689414e-05,
+      "loss": 2.6569,
+      "step": 8000
+    },
+    {
+      "epoch": 0.7241785100027157,
+      "eval_loss": 2.581277847290039,
+      "eval_runtime": 71.4998,
+      "eval_samples_per_second": 37.804,
+      "eval_steps_per_second": 3.161,
+      "step": 8000
+    },
+    {
+      "epoch": 0.724269032316466,
+      "grad_norm": 0.9000765085220337,
+      "learning_rate": 1.9454163400793334e-05,
+      "loss": 2.5595,
+      "step": 8001
+    },
+    {
+      "epoch": 0.7243595546302164,
+      "grad_norm": 0.8212129473686218,
+      "learning_rate": 1.943536066565603e-05,
+      "loss": 2.604,
+      "step": 8002
+    },
+    {
+      "epoch": 0.7244500769439667,
+      "grad_norm": 0.8661413788795471,
+      "learning_rate": 1.941656604337575e-05,
+      "loss": 2.6305,
+      "step": 8003
+    },
+    {
+      "epoch": 0.7245405992577171,
+      "grad_norm": 0.860085666179657,
+      "learning_rate": 1.9397779535845028e-05,
+      "loss": 2.6943,
+      "step": 8004
+    },
+    {
+      "epoch": 0.7246311215714674,
+      "grad_norm": 0.8990339636802673,
+      "learning_rate": 1.937900114495571e-05,
+      "loss": 2.6932,
+      "step": 8005
+    },
+    {
+      "epoch": 0.7247216438852178,
+      "grad_norm": 0.8864555358886719,
+      "learning_rate": 1.9360230872598773e-05,
+      "loss": 2.5889,
+      "step": 8006
+    },
+    {
+      "epoch": 0.724812166198968,
+      "grad_norm": 0.8880107998847961,
+      "learning_rate": 1.934146872066438e-05,
+      "loss": 2.6804,
+      "step": 8007
+    },
+    {
+      "epoch": 0.7249026885127184,
+      "grad_norm": 0.8222184777259827,
+      "learning_rate": 1.9322714691041878e-05,
+      "loss": 2.625,
+      "step": 8008
+    },
+    {
+      "epoch": 0.7249932108264687,
+      "grad_norm": 0.8501119613647461,
+      "learning_rate": 1.930396878561983e-05,
+      "loss": 2.6189,
+      "step": 8009
+    },
+    {
+      "epoch": 0.725083733140219,
+      "grad_norm": 0.8509238958358765,
+      "learning_rate": 1.9285231006285853e-05,
+      "loss": 2.5565,
+      "step": 8010
+    },
+    {
+      "epoch": 0.7251742554539694,
+      "grad_norm": 0.8645397424697876,
+      "learning_rate": 1.926650135492697e-05,
+      "loss": 2.6501,
+      "step": 8011
+    },
+    {
+      "epoch": 0.7252647777677197,
+      "grad_norm": 0.8589481115341187,
+      "learning_rate": 1.924777983342916e-05,
+      "loss": 2.6458,
+      "step": 8012
+    },
+    {
+      "epoch": 0.7253553000814701,
+      "grad_norm": 0.8989269137382507,
+      "learning_rate": 1.9229066443677713e-05,
+      "loss": 2.6505,
+      "step": 8013
+    },
+    {
+      "epoch": 0.7254458223952204,
+      "grad_norm": 0.924936830997467,
+      "learning_rate": 1.9210361187557057e-05,
+      "loss": 2.6867,
+      "step": 8014
+    },
+    {
+      "epoch": 0.7255363447089708,
+      "grad_norm": 0.8466819524765015,
+      "learning_rate": 1.9191664066950832e-05,
+      "loss": 2.6915,
+      "step": 8015
+    },
+    {
+      "epoch": 0.7256268670227211,
+      "grad_norm": 0.8410524129867554,
+      "learning_rate": 1.9172975083741817e-05,
+      "loss": 2.5731,
+      "step": 8016
+    },
+    {
+      "epoch": 0.7257173893364715,
+      "grad_norm": 0.8567287921905518,
+      "learning_rate": 1.915429423981202e-05,
+      "loss": 2.6707,
+      "step": 8017
+    },
+    {
+      "epoch": 0.7258079116502217,
+      "grad_norm": 0.8804045915603638,
+      "learning_rate": 1.9135621537042535e-05,
+      "loss": 2.6704,
+      "step": 8018
+    },
+    {
+      "epoch": 0.7258984339639721,
+      "grad_norm": 0.8276041746139526,
+      "learning_rate": 1.9116956977313804e-05,
+      "loss": 2.6637,
+      "step": 8019
+    },
+    {
+      "epoch": 0.7259889562777224,
+      "grad_norm": 0.8281861543655396,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 2.5908,
+      "step": 8020
+    },
+    {
+      "epoch": 0.7260794785914728,
+      "grad_norm": 0.8218677639961243,
+      "learning_rate": 1.907965229449564e-05,
+      "loss": 2.5981,
+      "step": 8021
+    },
+    {
+      "epoch": 0.7261700009052231,
+      "grad_norm": 0.7891271710395813,
+      "learning_rate": 1.9061012175162817e-05,
+      "loss": 2.6288,
+      "step": 8022
+    },
+    {
+      "epoch": 0.7262605232189735,
+      "grad_norm": 0.8426888585090637,
+      "learning_rate": 1.9042380206383858e-05,
+      "loss": 2.6242,
+      "step": 8023
+    },
+    {
+      "epoch": 0.7263510455327238,
+      "grad_norm": 0.8271638751029968,
+      "learning_rate": 1.9023756390034998e-05,
+      "loss": 2.6407,
+      "step": 8024
+    },
+    {
+      "epoch": 0.7264415678464742,
+      "grad_norm": 0.8045911192893982,
+      "learning_rate": 1.9005140727991677e-05,
+      "loss": 2.6306,
+      "step": 8025
+    },
+    {
+      "epoch": 0.7265320901602245,
+      "grad_norm": 0.8195641040802002,
+      "learning_rate": 1.8986533222128413e-05,
+      "loss": 2.587,
+      "step": 8026
+    },
+    {
+      "epoch": 0.7266226124739749,
+      "grad_norm": 0.8664358258247375,
+      "learning_rate": 1.896793387431909e-05,
+      "loss": 2.6334,
+      "step": 8027
+    },
+    {
+      "epoch": 0.7267131347877251,
+      "grad_norm": 0.8703563809394836,
+      "learning_rate": 1.8949342686436588e-05,
+      "loss": 2.6234,
+      "step": 8028
+    },
+    {
+      "epoch": 0.7268036571014755,
+      "grad_norm": 0.8076162934303284,
+      "learning_rate": 1.8930759660353048e-05,
+      "loss": 2.6312,
+      "step": 8029
+    },
+    {
+      "epoch": 0.7268941794152258,
+      "grad_norm": 0.8560411930084229,
+      "learning_rate": 1.8912184797939803e-05,
+      "loss": 2.6338,
+      "step": 8030
+    },
+    {
+      "epoch": 0.7269847017289762,
+      "grad_norm": 0.8122009038925171,
+      "learning_rate": 1.8893618101067355e-05,
+      "loss": 2.6406,
+      "step": 8031
+    },
+    {
+      "epoch": 0.7270752240427265,
+      "grad_norm": 0.8404006361961365,
+      "learning_rate": 1.8875059571605293e-05,
+      "loss": 2.6284,
+      "step": 8032
+    },
+    {
+      "epoch": 0.7271657463564769,
+      "grad_norm": 0.8854060769081116,
+      "learning_rate": 1.8856509211422557e-05,
+      "loss": 2.5995,
+      "step": 8033
+    },
+    {
+      "epoch": 0.7272562686702272,
+      "grad_norm": 0.8898727297782898,
+      "learning_rate": 1.8837967022387102e-05,
+      "loss": 2.611,
+      "step": 8034
+    },
+    {
+      "epoch": 0.7273467909839776,
+      "grad_norm": 0.9046933054924011,
+      "learning_rate": 1.881943300636615e-05,
+      "loss": 2.6472,
+      "step": 8035
+    },
+    {
+      "epoch": 0.7274373132977279,
+      "grad_norm": 0.8621687889099121,
+      "learning_rate": 1.8800907165226066e-05,
+      "loss": 2.6788,
+      "step": 8036
+    },
+    {
+      "epoch": 0.7275278356114783,
+      "grad_norm": 0.8203818798065186,
+      "learning_rate": 1.878238950083242e-05,
+      "loss": 2.6216,
+      "step": 8037
+    },
+    {
+      "epoch": 0.7276183579252286,
+      "grad_norm": 0.8922222852706909,
+      "learning_rate": 1.876388001504995e-05,
+      "loss": 2.6475,
+      "step": 8038
+    },
+    {
+      "epoch": 0.727708880238979,
+      "grad_norm": 0.8815889358520508,
+      "learning_rate": 1.874537870974251e-05,
+      "loss": 2.7283,
+      "step": 8039
+    },
+    {
+      "epoch": 0.7277994025527292,
+      "grad_norm": 0.8744126558303833,
+      "learning_rate": 1.8726885586773212e-05,
+      "loss": 2.6958,
+      "step": 8040
+    },
+    {
+      "epoch": 0.7278899248664796,
+      "grad_norm": 0.8782918453216553,
+      "learning_rate": 1.870840064800431e-05,
+      "loss": 2.6935,
+      "step": 8041
+    },
+    {
+      "epoch": 0.7279804471802299,
+      "grad_norm": 0.859200119972229,
+      "learning_rate": 1.8689923895297245e-05,
+      "loss": 2.6295,
+      "step": 8042
+    },
+    {
+      "epoch": 0.7280709694939803,
+      "grad_norm": 0.8936563730239868,
+      "learning_rate": 1.867145533051261e-05,
+      "loss": 2.6756,
+      "step": 8043
+    },
+    {
+      "epoch": 0.7281614918077306,
+      "grad_norm": 0.8279479742050171,
+      "learning_rate": 1.8652994955510227e-05,
+      "loss": 2.5914,
+      "step": 8044
+    },
+    {
+      "epoch": 0.728252014121481,
+      "grad_norm": 0.8420711159706116,
+      "learning_rate": 1.8634542772148978e-05,
+      "loss": 2.6247,
+      "step": 8045
+    },
+    {
+      "epoch": 0.7283425364352313,
+      "grad_norm": 0.8849563598632812,
+      "learning_rate": 1.8616098782287095e-05,
+      "loss": 2.5771,
+      "step": 8046
+    },
+    {
+      "epoch": 0.7284330587489817,
+      "grad_norm": 0.8260254263877869,
+      "learning_rate": 1.8597662987781816e-05,
+      "loss": 2.65,
+      "step": 8047
+    },
+    {
+      "epoch": 0.728523581062732,
+      "grad_norm": 0.8107094764709473,
+      "learning_rate": 1.8579235390489635e-05,
+      "loss": 2.6522,
+      "step": 8048
+    },
+    {
+      "epoch": 0.7286141033764824,
+      "grad_norm": 0.9501773715019226,
+      "learning_rate": 1.8560815992266234e-05,
+      "loss": 2.6627,
+      "step": 8049
+    },
+    {
+      "epoch": 0.7287046256902326,
+      "grad_norm": 0.8340609073638916,
+      "learning_rate": 1.854240479496643e-05,
+      "loss": 2.593,
+      "step": 8050
+    },
+    {
+      "epoch": 0.7287951480039829,
+      "grad_norm": 0.8562996983528137,
+      "learning_rate": 1.8524001800444235e-05,
+      "loss": 2.6483,
+      "step": 8051
+    },
+    {
+      "epoch": 0.7288856703177333,
+      "grad_norm": 0.8556627035140991,
+      "learning_rate": 1.8505607010552862e-05,
+      "loss": 2.6394,
+      "step": 8052
+    },
+    {
+      "epoch": 0.7289761926314836,
+      "grad_norm": 0.8138837218284607,
+      "learning_rate": 1.848722042714457e-05,
+      "loss": 2.6216,
+      "step": 8053
+    },
+    {
+      "epoch": 0.729066714945234,
+      "grad_norm": 0.8794299960136414,
+      "learning_rate": 1.846884205207101e-05,
+      "loss": 2.6405,
+      "step": 8054
+    },
+    {
+      "epoch": 0.7291572372589843,
+      "grad_norm": 0.8482553958892822,
+      "learning_rate": 1.8450471887182796e-05,
+      "loss": 2.648,
+      "step": 8055
+    },
+    {
+      "epoch": 0.7292477595727347,
+      "grad_norm": 0.8459094166755676,
+      "learning_rate": 1.8432109934329834e-05,
+      "loss": 2.6691,
+      "step": 8056
+    },
+    {
+      "epoch": 0.729338281886485,
+      "grad_norm": 0.8724743723869324,
+      "learning_rate": 1.8413756195361166e-05,
+      "loss": 2.5916,
+      "step": 8057
+    },
+    {
+      "epoch": 0.7294288042002354,
+      "grad_norm": 0.8487122654914856,
+      "learning_rate": 1.8395410672125035e-05,
+      "loss": 2.6292,
+      "step": 8058
+    },
+    {
+      "epoch": 0.7295193265139857,
+      "grad_norm": 0.8307660818099976,
+      "learning_rate": 1.8377073366468778e-05,
+      "loss": 2.5968,
+      "step": 8059
+    },
+    {
+      "epoch": 0.729609848827736,
+      "grad_norm": 0.8485986590385437,
+      "learning_rate": 1.835874428023905e-05,
+      "loss": 2.5854,
+      "step": 8060
+    },
+    {
+      "epoch": 0.7297003711414863,
+      "grad_norm": 0.8574190139770508,
+      "learning_rate": 1.8340423415281505e-05,
+      "loss": 2.6206,
+      "step": 8061
+    },
+    {
+      "epoch": 0.7297908934552367,
+      "grad_norm": 0.8343666791915894,
+      "learning_rate": 1.832211077344109e-05,
+      "loss": 2.6166,
+      "step": 8062
+    },
+    {
+      "epoch": 0.729881415768987,
+      "grad_norm": 0.8841292262077332,
+      "learning_rate": 1.8303806356561893e-05,
+      "loss": 2.6292,
+      "step": 8063
+    },
+    {
+      "epoch": 0.7299719380827374,
+      "grad_norm": 0.8868502974510193,
+      "learning_rate": 1.8285510166487152e-05,
+      "loss": 2.6101,
+      "step": 8064
+    },
+    {
+      "epoch": 0.7300624603964877,
+      "grad_norm": 0.8825874328613281,
+      "learning_rate": 1.8267222205059308e-05,
+      "loss": 2.6845,
+      "step": 8065
+    },
+    {
+      "epoch": 0.7301529827102381,
+      "grad_norm": 0.9012864828109741,
+      "learning_rate": 1.824894247411998e-05,
+      "loss": 2.6119,
+      "step": 8066
+    },
+    {
+      "epoch": 0.7302435050239884,
+      "grad_norm": 0.8223727345466614,
+      "learning_rate": 1.8230670975509844e-05,
+      "loss": 2.6667,
+      "step": 8067
+    },
+    {
+      "epoch": 0.7303340273377388,
+      "grad_norm": 0.8057619333267212,
+      "learning_rate": 1.8212407711068958e-05,
+      "loss": 2.6175,
+      "step": 8068
+    },
+    {
+      "epoch": 0.7304245496514891,
+      "grad_norm": 0.7814914584159851,
+      "learning_rate": 1.8194152682636356e-05,
+      "loss": 2.6271,
+      "step": 8069
+    },
+    {
+      "epoch": 0.7305150719652395,
+      "grad_norm": 0.8347715735435486,
+      "learning_rate": 1.817590589205035e-05,
+      "loss": 2.6295,
+      "step": 8070
+    },
+    {
+      "epoch": 0.7306055942789897,
+      "grad_norm": 0.8338614106178284,
+      "learning_rate": 1.8157667341148364e-05,
+      "loss": 2.612,
+      "step": 8071
+    },
+    {
+      "epoch": 0.7306961165927401,
+      "grad_norm": 0.8309128880500793,
+      "learning_rate": 1.8139437031767048e-05,
+      "loss": 2.6255,
+      "step": 8072
+    },
+    {
+      "epoch": 0.7307866389064904,
+      "grad_norm": 0.8655341863632202,
+      "learning_rate": 1.8121214965742182e-05,
+      "loss": 2.6858,
+      "step": 8073
+    },
+    {
+      "epoch": 0.7308771612202408,
+      "grad_norm": 0.8365732431411743,
+      "learning_rate": 1.810300114490875e-05,
+      "loss": 2.6609,
+      "step": 8074
+    },
+    {
+      "epoch": 0.7309676835339911,
+      "grad_norm": 0.9555655717849731,
+      "learning_rate": 1.808479557110081e-05,
+      "loss": 2.6157,
+      "step": 8075
+    },
+    {
+      "epoch": 0.7310582058477415,
+      "grad_norm": 0.834373950958252,
+      "learning_rate": 1.8066598246151768e-05,
+      "loss": 2.6308,
+      "step": 8076
+    },
+    {
+      "epoch": 0.7311487281614918,
+      "grad_norm": 0.9408546686172485,
+      "learning_rate": 1.804840917189402e-05,
+      "loss": 2.612,
+      "step": 8077
+    },
+    {
+      "epoch": 0.7312392504752422,
+      "grad_norm": 0.8463700413703918,
+      "learning_rate": 1.8030228350159228e-05,
+      "loss": 2.6008,
+      "step": 8078
+    },
+    {
+      "epoch": 0.7313297727889925,
+      "grad_norm": 0.880274772644043,
+      "learning_rate": 1.8012055782778194e-05,
+      "loss": 2.614,
+      "step": 8079
+    },
+    {
+      "epoch": 0.7314202951027429,
+      "grad_norm": 0.8945399522781372,
+      "learning_rate": 1.7993891471580893e-05,
+      "loss": 2.6185,
+      "step": 8080
+    },
+    {
+      "epoch": 0.7315108174164932,
+      "grad_norm": 0.8566452860832214,
+      "learning_rate": 1.797573541839648e-05,
+      "loss": 2.6112,
+      "step": 8081
+    },
+    {
+      "epoch": 0.7316013397302435,
+      "grad_norm": 0.8027698397636414,
+      "learning_rate": 1.7957587625053296e-05,
+      "loss": 2.6642,
+      "step": 8082
+    },
+    {
+      "epoch": 0.7316918620439938,
+      "grad_norm": 0.7911441922187805,
+      "learning_rate": 1.7939448093378743e-05,
+      "loss": 2.5205,
+      "step": 8083
+    },
+    {
+      "epoch": 0.7317823843577442,
+      "grad_norm": 0.9017260670661926,
+      "learning_rate": 1.7921316825199563e-05,
+      "loss": 2.5994,
+      "step": 8084
+    },
+    {
+      "epoch": 0.7318729066714945,
+      "grad_norm": 0.8945018649101257,
+      "learning_rate": 1.7903193822341512e-05,
+      "loss": 2.5853,
+      "step": 8085
+    },
+    {
+      "epoch": 0.7319634289852449,
+      "grad_norm": 0.8400217890739441,
+      "learning_rate": 1.78850790866296e-05,
+      "loss": 2.6382,
+      "step": 8086
+    },
+    {
+      "epoch": 0.7320539512989952,
+      "grad_norm": 0.8900841474533081,
+      "learning_rate": 1.7866972619887978e-05,
+      "loss": 2.6085,
+      "step": 8087
+    },
+    {
+      "epoch": 0.7321444736127456,
+      "grad_norm": 0.8915272355079651,
+      "learning_rate": 1.7848874423939966e-05,
+      "loss": 2.7206,
+      "step": 8088
+    },
+    {
+      "epoch": 0.7322349959264959,
+      "grad_norm": 0.832134485244751,
+      "learning_rate": 1.7830784500608044e-05,
+      "loss": 2.6791,
+      "step": 8089
+    },
+    {
+      "epoch": 0.7323255182402463,
+      "grad_norm": 0.8645135164260864,
+      "learning_rate": 1.7812702851713904e-05,
+      "loss": 2.5735,
+      "step": 8090
+    },
+    {
+      "epoch": 0.7324160405539966,
+      "grad_norm": 0.7954042553901672,
+      "learning_rate": 1.7794629479078318e-05,
+      "loss": 2.6569,
+      "step": 8091
+    },
+    {
+      "epoch": 0.7325065628677468,
+      "grad_norm": 0.8583080172538757,
+      "learning_rate": 1.777656438452129e-05,
+      "loss": 2.6162,
+      "step": 8092
+    },
+    {
+      "epoch": 0.7325970851814972,
+      "grad_norm": 0.8749803304672241,
+      "learning_rate": 1.7758507569861983e-05,
+      "loss": 2.6316,
+      "step": 8093
+    },
+    {
+      "epoch": 0.7326876074952475,
+      "grad_norm": 0.8284852504730225,
+      "learning_rate": 1.77404590369187e-05,
+      "loss": 2.5988,
+      "step": 8094
+    },
+    {
+      "epoch": 0.7327781298089979,
+      "grad_norm": 0.8743882775306702,
+      "learning_rate": 1.7722418787508955e-05,
+      "loss": 2.6676,
+      "step": 8095
+    },
+    {
+      "epoch": 0.7328686521227482,
+      "grad_norm": 0.8623575568199158,
+      "learning_rate": 1.7704386823449403e-05,
+      "loss": 2.6689,
+      "step": 8096
+    },
+    {
+      "epoch": 0.7329591744364986,
+      "grad_norm": 0.8969399333000183,
+      "learning_rate": 1.7686363146555805e-05,
+      "loss": 2.6124,
+      "step": 8097
+    },
+    {
+      "epoch": 0.7330496967502489,
+      "grad_norm": 0.8351315259933472,
+      "learning_rate": 1.7668347758643233e-05,
+      "loss": 2.5901,
+      "step": 8098
+    },
+    {
+      "epoch": 0.7331402190639993,
+      "grad_norm": 0.811880886554718,
+      "learning_rate": 1.7650340661525754e-05,
+      "loss": 2.6038,
+      "step": 8099
+    },
+    {
+      "epoch": 0.7332307413777496,
+      "grad_norm": 0.810198187828064,
+      "learning_rate": 1.763234185701673e-05,
+      "loss": 2.6428,
+      "step": 8100
+    },
+    {
+      "epoch": 0.7333212636915,
+      "grad_norm": 0.8915790319442749,
+      "learning_rate": 1.7614351346928627e-05,
+      "loss": 2.6327,
+      "step": 8101
+    },
+    {
+      "epoch": 0.7334117860052503,
+      "grad_norm": 1.0429893732070923,
+      "learning_rate": 1.7596369133073076e-05,
+      "loss": 2.6487,
+      "step": 8102
+    },
+    {
+      "epoch": 0.7335023083190007,
+      "grad_norm": 0.8270307779312134,
+      "learning_rate": 1.7578395217260912e-05,
+      "loss": 2.6318,
+      "step": 8103
+    },
+    {
+      "epoch": 0.7335928306327509,
+      "grad_norm": 0.8245958089828491,
+      "learning_rate": 1.756042960130212e-05,
+      "loss": 2.6318,
+      "step": 8104
+    },
+    {
+      "epoch": 0.7336833529465013,
+      "grad_norm": 0.8989834785461426,
+      "learning_rate": 1.754247228700575e-05,
+      "loss": 2.6562,
+      "step": 8105
+    },
+    {
+      "epoch": 0.7337738752602516,
+      "grad_norm": 0.8361373543739319,
+      "learning_rate": 1.7524523276180226e-05,
+      "loss": 2.5975,
+      "step": 8106
+    },
+    {
+      "epoch": 0.733864397574002,
+      "grad_norm": 0.8043052554130554,
+      "learning_rate": 1.7506582570632914e-05,
+      "loss": 2.5758,
+      "step": 8107
+    },
+    {
+      "epoch": 0.7339549198877523,
+      "grad_norm": 0.8632999658584595,
+      "learning_rate": 1.7488650172170496e-05,
+      "loss": 2.6127,
+      "step": 8108
+    },
+    {
+      "epoch": 0.7340454422015027,
+      "grad_norm": 0.8198879361152649,
+      "learning_rate": 1.7470726082598733e-05,
+      "loss": 2.6815,
+      "step": 8109
+    },
+    {
+      "epoch": 0.734135964515253,
+      "grad_norm": 0.8820613026618958,
+      "learning_rate": 1.74528103037226e-05,
+      "loss": 2.6031,
+      "step": 8110
+    },
+    {
+      "epoch": 0.7342264868290034,
+      "grad_norm": 0.8991879224777222,
+      "learning_rate": 1.743490283734621e-05,
+      "loss": 2.5671,
+      "step": 8111
+    },
+    {
+      "epoch": 0.7343170091427537,
+      "grad_norm": 0.7861236929893494,
+      "learning_rate": 1.741700368527288e-05,
+      "loss": 2.6053,
+      "step": 8112
+    },
+    {
+      "epoch": 0.7344075314565041,
+      "grad_norm": 0.8448014259338379,
+      "learning_rate": 1.7399112849304954e-05,
+      "loss": 2.6049,
+      "step": 8113
+    },
+    {
+      "epoch": 0.7344980537702543,
+      "grad_norm": 0.8373123407363892,
+      "learning_rate": 1.7381230331244168e-05,
+      "loss": 2.6315,
+      "step": 8114
+    },
+    {
+      "epoch": 0.7345885760840047,
+      "grad_norm": 0.8829507827758789,
+      "learning_rate": 1.7363356132891196e-05,
+      "loss": 2.6234,
+      "step": 8115
+    },
+    {
+      "epoch": 0.734679098397755,
+      "grad_norm": 0.9009426236152649,
+      "learning_rate": 1.7345490256045993e-05,
+      "loss": 2.6165,
+      "step": 8116
+    },
+    {
+      "epoch": 0.7347696207115054,
+      "grad_norm": 0.8082904815673828,
+      "learning_rate": 1.7327632702507667e-05,
+      "loss": 2.6422,
+      "step": 8117
+    },
+    {
+      "epoch": 0.7348601430252557,
+      "grad_norm": 0.8886256217956543,
+      "learning_rate": 1.730978347407447e-05,
+      "loss": 2.6248,
+      "step": 8118
+    },
+    {
+      "epoch": 0.7349506653390061,
+      "grad_norm": 0.8580721616744995,
+      "learning_rate": 1.7291942572543807e-05,
+      "loss": 2.6861,
+      "step": 8119
+    },
+    {
+      "epoch": 0.7350411876527564,
+      "grad_norm": 0.911753237247467,
+      "learning_rate": 1.7274109999712295e-05,
+      "loss": 2.6194,
+      "step": 8120
+    },
+    {
+      "epoch": 0.7351317099665068,
+      "grad_norm": 0.9159942269325256,
+      "learning_rate": 1.725628575737559e-05,
+      "loss": 2.5499,
+      "step": 8121
+    },
+    {
+      "epoch": 0.7352222322802571,
+      "grad_norm": 0.8776478171348572,
+      "learning_rate": 1.7238469847328697e-05,
+      "loss": 2.659,
+      "step": 8122
+    },
+    {
+      "epoch": 0.7353127545940075,
+      "grad_norm": 0.8453285098075867,
+      "learning_rate": 1.7220662271365618e-05,
+      "loss": 2.6071,
+      "step": 8123
+    },
+    {
+      "epoch": 0.7354032769077578,
+      "grad_norm": 0.9019317030906677,
+      "learning_rate": 1.7202863031279537e-05,
+      "loss": 2.6502,
+      "step": 8124
+    },
+    {
+      "epoch": 0.7354937992215081,
+      "grad_norm": 0.8425202965736389,
+      "learning_rate": 1.7185072128862933e-05,
+      "loss": 2.5551,
+      "step": 8125
+    },
+    {
+      "epoch": 0.7355843215352584,
+      "grad_norm": 0.8753564357757568,
+      "learning_rate": 1.7167289565907287e-05,
+      "loss": 2.6372,
+      "step": 8126
+    },
+    {
+      "epoch": 0.7356748438490088,
+      "grad_norm": 0.847804069519043,
+      "learning_rate": 1.7149515344203304e-05,
+      "loss": 2.5992,
+      "step": 8127
+    },
+    {
+      "epoch": 0.7357653661627591,
+      "grad_norm": 0.8290091156959534,
+      "learning_rate": 1.713174946554086e-05,
+      "loss": 2.6184,
+      "step": 8128
+    },
+    {
+      "epoch": 0.7358558884765095,
+      "grad_norm": 0.8304346203804016,
+      "learning_rate": 1.7113991931708972e-05,
+      "loss": 2.6377,
+      "step": 8129
+    },
+    {
+      "epoch": 0.7359464107902598,
+      "grad_norm": 0.881280243396759,
+      "learning_rate": 1.7096242744495837e-05,
+      "loss": 2.6182,
+      "step": 8130
+    },
+    {
+      "epoch": 0.7360369331040102,
+      "grad_norm": 0.846674382686615,
+      "learning_rate": 1.7078501905688805e-05,
+      "loss": 2.6258,
+      "step": 8131
+    },
+    {
+      "epoch": 0.7361274554177605,
+      "grad_norm": 0.860022783279419,
+      "learning_rate": 1.7060769417074317e-05,
+      "loss": 2.6807,
+      "step": 8132
+    },
+    {
+      "epoch": 0.7362179777315108,
+      "grad_norm": 0.9033744931221008,
+      "learning_rate": 1.704304528043813e-05,
+      "loss": 2.7004,
+      "step": 8133
+    },
+    {
+      "epoch": 0.7363085000452612,
+      "grad_norm": 0.8882016539573669,
+      "learning_rate": 1.702532949756499e-05,
+      "loss": 2.6703,
+      "step": 8134
+    },
+    {
+      "epoch": 0.7363990223590114,
+      "grad_norm": 0.880939245223999,
+      "learning_rate": 1.7007622070238906e-05,
+      "loss": 2.6365,
+      "step": 8135
+    },
+    {
+      "epoch": 0.7364895446727618,
+      "grad_norm": 0.847804844379425,
+      "learning_rate": 1.6989923000243e-05,
+      "loss": 2.6706,
+      "step": 8136
+    },
+    {
+      "epoch": 0.7365800669865121,
+      "grad_norm": 0.8592808246612549,
+      "learning_rate": 1.697223228935958e-05,
+      "loss": 2.6047,
+      "step": 8137
+    },
+    {
+      "epoch": 0.7366705893002625,
+      "grad_norm": 0.8524011969566345,
+      "learning_rate": 1.695454993937009e-05,
+      "loss": 2.6969,
+      "step": 8138
+    },
+    {
+      "epoch": 0.7367611116140128,
+      "grad_norm": 0.8169214725494385,
+      "learning_rate": 1.6936875952055175e-05,
+      "loss": 2.607,
+      "step": 8139
+    },
+    {
+      "epoch": 0.7368516339277632,
+      "grad_norm": 0.8056317567825317,
+      "learning_rate": 1.6919210329194533e-05,
+      "loss": 2.6452,
+      "step": 8140
+    },
+    {
+      "epoch": 0.7369421562415135,
+      "grad_norm": 0.8107497096061707,
+      "learning_rate": 1.690155307256719e-05,
+      "loss": 2.5957,
+      "step": 8141
+    },
+    {
+      "epoch": 0.7370326785552639,
+      "grad_norm": 0.9261689186096191,
+      "learning_rate": 1.688390418395115e-05,
+      "loss": 2.6895,
+      "step": 8142
+    },
+    {
+      "epoch": 0.7371232008690142,
+      "grad_norm": 0.8380181789398193,
+      "learning_rate": 1.6866263665123684e-05,
+      "loss": 2.6552,
+      "step": 8143
+    },
+    {
+      "epoch": 0.7372137231827646,
+      "grad_norm": 0.8535093069076538,
+      "learning_rate": 1.684863151786119e-05,
+      "loss": 2.6757,
+      "step": 8144
+    },
+    {
+      "epoch": 0.7373042454965149,
+      "grad_norm": 0.8620217442512512,
+      "learning_rate": 1.683100774393923e-05,
+      "loss": 2.6247,
+      "step": 8145
+    },
+    {
+      "epoch": 0.7373947678102653,
+      "grad_norm": 0.9317750930786133,
+      "learning_rate": 1.6813392345132518e-05,
+      "loss": 2.5605,
+      "step": 8146
+    },
+    {
+      "epoch": 0.7374852901240155,
+      "grad_norm": 0.8641749024391174,
+      "learning_rate": 1.679578532321494e-05,
+      "loss": 2.5937,
+      "step": 8147
+    },
+    {
+      "epoch": 0.7375758124377659,
+      "grad_norm": 0.8631426692008972,
+      "learning_rate": 1.6778186679959452e-05,
+      "loss": 2.7071,
+      "step": 8148
+    },
+    {
+      "epoch": 0.7376663347515162,
+      "grad_norm": 0.816089391708374,
+      "learning_rate": 1.6760596417138342e-05,
+      "loss": 2.6578,
+      "step": 8149
+    },
+    {
+      "epoch": 0.7377568570652666,
+      "grad_norm": 0.8169116377830505,
+      "learning_rate": 1.6743014536522873e-05,
+      "loss": 2.5996,
+      "step": 8150
+    },
+    {
+      "epoch": 0.7378473793790169,
+      "grad_norm": 0.8737078905105591,
+      "learning_rate": 1.672544103988356e-05,
+      "loss": 2.6331,
+      "step": 8151
+    },
+    {
+      "epoch": 0.7379379016927673,
+      "grad_norm": 0.8869850039482117,
+      "learning_rate": 1.6707875928990058e-05,
+      "loss": 2.7173,
+      "step": 8152
+    },
+    {
+      "epoch": 0.7380284240065176,
+      "grad_norm": 0.8866559863090515,
+      "learning_rate": 1.6690319205611194e-05,
+      "loss": 2.5917,
+      "step": 8153
+    },
+    {
+      "epoch": 0.738118946320268,
+      "grad_norm": 0.8431488871574402,
+      "learning_rate": 1.6672770871514854e-05,
+      "loss": 2.5925,
+      "step": 8154
+    },
+    {
+      "epoch": 0.7382094686340183,
+      "grad_norm": 0.8199314475059509,
+      "learning_rate": 1.6655230928468258e-05,
+      "loss": 2.6326,
+      "step": 8155
+    },
+    {
+      "epoch": 0.7382999909477687,
+      "grad_norm": 0.8784447312355042,
+      "learning_rate": 1.6637699378237605e-05,
+      "loss": 2.6156,
+      "step": 8156
+    },
+    {
+      "epoch": 0.7383905132615189,
+      "grad_norm": 0.7935834527015686,
+      "learning_rate": 1.6620176222588334e-05,
+      "loss": 2.6223,
+      "step": 8157
+    },
+    {
+      "epoch": 0.7384810355752693,
+      "grad_norm": 0.8735072016716003,
+      "learning_rate": 1.660266146328504e-05,
+      "loss": 2.6237,
+      "step": 8158
+    },
+    {
+      "epoch": 0.7385715578890196,
+      "grad_norm": 0.8265995383262634,
+      "learning_rate": 1.6585155102091467e-05,
+      "loss": 2.5732,
+      "step": 8159
+    },
+    {
+      "epoch": 0.73866208020277,
+      "grad_norm": 0.8630474209785461,
+      "learning_rate": 1.6567657140770475e-05,
+      "loss": 2.5965,
+      "step": 8160
+    },
+    {
+      "epoch": 0.7387526025165203,
+      "grad_norm": 0.8460192680358887,
+      "learning_rate": 1.6550167581084164e-05,
+      "loss": 2.701,
+      "step": 8161
+    },
+    {
+      "epoch": 0.7388431248302707,
+      "grad_norm": 0.8023285865783691,
+      "learning_rate": 1.6532686424793644e-05,
+      "loss": 2.6347,
+      "step": 8162
+    },
+    {
+      "epoch": 0.738933647144021,
+      "grad_norm": 0.8689784407615662,
+      "learning_rate": 1.6515213673659357e-05,
+      "loss": 2.6865,
+      "step": 8163
+    },
+    {
+      "epoch": 0.7390241694577714,
+      "grad_norm": 0.8490846157073975,
+      "learning_rate": 1.649774932944075e-05,
+      "loss": 2.6641,
+      "step": 8164
+    },
+    {
+      "epoch": 0.7391146917715217,
+      "grad_norm": 0.8280508518218994,
+      "learning_rate": 1.6480293393896507e-05,
+      "loss": 2.5669,
+      "step": 8165
+    },
+    {
+      "epoch": 0.7392052140852721,
+      "grad_norm": 0.9088148474693298,
+      "learning_rate": 1.6462845868784425e-05,
+      "loss": 2.7244,
+      "step": 8166
+    },
+    {
+      "epoch": 0.7392957363990224,
+      "grad_norm": 0.7922699451446533,
+      "learning_rate": 1.6445406755861482e-05,
+      "loss": 2.607,
+      "step": 8167
+    },
+    {
+      "epoch": 0.7393862587127727,
+      "grad_norm": 0.8359072804450989,
+      "learning_rate": 1.642797605688379e-05,
+      "loss": 2.6218,
+      "step": 8168
+    },
+    {
+      "epoch": 0.739476781026523,
+      "grad_norm": 0.8865196704864502,
+      "learning_rate": 1.6410553773606653e-05,
+      "loss": 2.6453,
+      "step": 8169
+    },
+    {
+      "epoch": 0.7395673033402734,
+      "grad_norm": 0.8127596378326416,
+      "learning_rate": 1.6393139907784404e-05,
+      "loss": 2.6738,
+      "step": 8170
+    },
+    {
+      "epoch": 0.7396578256540237,
+      "grad_norm": 0.8192434906959534,
+      "learning_rate": 1.6375734461170733e-05,
+      "loss": 2.6035,
+      "step": 8171
+    },
+    {
+      "epoch": 0.7397483479677741,
+      "grad_norm": 0.8248426914215088,
+      "learning_rate": 1.635833743551829e-05,
+      "loss": 2.6281,
+      "step": 8172
+    },
+    {
+      "epoch": 0.7398388702815244,
+      "grad_norm": 0.8829620480537415,
+      "learning_rate": 1.6340948832578983e-05,
+      "loss": 2.6225,
+      "step": 8173
+    },
+    {
+      "epoch": 0.7399293925952747,
+      "grad_norm": 0.8189470767974854,
+      "learning_rate": 1.632356865410384e-05,
+      "loss": 2.669,
+      "step": 8174
+    },
+    {
+      "epoch": 0.7400199149090251,
+      "grad_norm": 0.872611403465271,
+      "learning_rate": 1.630619690184303e-05,
+      "loss": 2.6179,
+      "step": 8175
+    },
+    {
+      "epoch": 0.7401104372227754,
+      "grad_norm": 0.8715521097183228,
+      "learning_rate": 1.6288833577545914e-05,
+      "loss": 2.6397,
+      "step": 8176
+    },
+    {
+      "epoch": 0.7402009595365258,
+      "grad_norm": 0.9405277371406555,
+      "learning_rate": 1.6271478682960993e-05,
+      "loss": 2.6259,
+      "step": 8177
+    },
+    {
+      "epoch": 0.740291481850276,
+      "grad_norm": 0.8795056939125061,
+      "learning_rate": 1.6254132219835816e-05,
+      "loss": 2.6157,
+      "step": 8178
+    },
+    {
+      "epoch": 0.7403820041640264,
+      "grad_norm": 0.8261298537254333,
+      "learning_rate": 1.6236794189917293e-05,
+      "loss": 2.5853,
+      "step": 8179
+    },
+    {
+      "epoch": 0.7404725264777767,
+      "grad_norm": 0.9246150255203247,
+      "learning_rate": 1.621946459495127e-05,
+      "loss": 2.601,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7405630487915271,
+      "grad_norm": 0.803239107131958,
+      "learning_rate": 1.6202143436682882e-05,
+      "loss": 2.6329,
+      "step": 8181
+    },
+    {
+      "epoch": 0.7406535711052774,
+      "grad_norm": 0.8190500736236572,
+      "learning_rate": 1.6184830716856347e-05,
+      "loss": 2.6416,
+      "step": 8182
+    },
+    {
+      "epoch": 0.7407440934190278,
+      "grad_norm": 0.945702075958252,
+      "learning_rate": 1.6167526437215095e-05,
+      "loss": 2.6504,
+      "step": 8183
+    },
+    {
+      "epoch": 0.7408346157327781,
+      "grad_norm": 0.8268640041351318,
+      "learning_rate": 1.6150230599501594e-05,
+      "loss": 2.6327,
+      "step": 8184
+    },
+    {
+      "epoch": 0.7409251380465285,
+      "grad_norm": 0.8651500344276428,
+      "learning_rate": 1.6132943205457606e-05,
+      "loss": 2.5888,
+      "step": 8185
+    },
+    {
+      "epoch": 0.7410156603602788,
+      "grad_norm": 0.8950450420379639,
+      "learning_rate": 1.6115664256823937e-05,
+      "loss": 2.6752,
+      "step": 8186
+    },
+    {
+      "epoch": 0.7411061826740292,
+      "grad_norm": 0.8721691370010376,
+      "learning_rate": 1.609839375534058e-05,
+      "loss": 2.6091,
+      "step": 8187
+    },
+    {
+      "epoch": 0.7411967049877795,
+      "grad_norm": 0.8327284455299377,
+      "learning_rate": 1.6081131702746667e-05,
+      "loss": 2.6489,
+      "step": 8188
+    },
+    {
+      "epoch": 0.7412872273015298,
+      "grad_norm": 0.7996222376823425,
+      "learning_rate": 1.6063878100780506e-05,
+      "loss": 2.6007,
+      "step": 8189
+    },
+    {
+      "epoch": 0.7413777496152801,
+      "grad_norm": 0.8431447148323059,
+      "learning_rate": 1.6046632951179508e-05,
+      "loss": 2.6216,
+      "step": 8190
+    },
+    {
+      "epoch": 0.7414682719290305,
+      "grad_norm": 0.8708969950675964,
+      "learning_rate": 1.6029396255680305e-05,
+      "loss": 2.614,
+      "step": 8191
+    },
+    {
+      "epoch": 0.7415587942427808,
+      "grad_norm": 0.8860566020011902,
+      "learning_rate": 1.601216801601857e-05,
+      "loss": 2.6083,
+      "step": 8192
+    },
+    {
+      "epoch": 0.7416493165565312,
+      "grad_norm": 0.8327545523643494,
+      "learning_rate": 1.5994948233929253e-05,
+      "loss": 2.6623,
+      "step": 8193
+    },
+    {
+      "epoch": 0.7417398388702815,
+      "grad_norm": 0.8494716286659241,
+      "learning_rate": 1.5977736911146324e-05,
+      "loss": 2.6268,
+      "step": 8194
+    },
+    {
+      "epoch": 0.7418303611840319,
+      "grad_norm": 0.8706437945365906,
+      "learning_rate": 1.5960534049402987e-05,
+      "loss": 2.6315,
+      "step": 8195
+    },
+    {
+      "epoch": 0.7419208834977822,
+      "grad_norm": 0.7802254557609558,
+      "learning_rate": 1.5943339650431576e-05,
+      "loss": 2.6167,
+      "step": 8196
+    },
+    {
+      "epoch": 0.7420114058115326,
+      "grad_norm": 0.8174785375595093,
+      "learning_rate": 1.5926153715963565e-05,
+      "loss": 2.5987,
+      "step": 8197
+    },
+    {
+      "epoch": 0.7421019281252829,
+      "grad_norm": 0.8315815329551697,
+      "learning_rate": 1.590897624772957e-05,
+      "loss": 2.6454,
+      "step": 8198
+    },
+    {
+      "epoch": 0.7421924504390333,
+      "grad_norm": 0.8264350295066833,
+      "learning_rate": 1.589180724745939e-05,
+      "loss": 2.6312,
+      "step": 8199
+    },
+    {
+      "epoch": 0.7422829727527835,
+      "grad_norm": 0.7961425185203552,
+      "learning_rate": 1.587464671688187e-05,
+      "loss": 2.5814,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7423734950665339,
+      "grad_norm": 0.7933376431465149,
+      "learning_rate": 1.5857494657725182e-05,
+      "loss": 2.6176,
+      "step": 8201
+    },
+    {
+      "epoch": 0.7424640173802842,
+      "grad_norm": 0.8313885927200317,
+      "learning_rate": 1.584035107171644e-05,
+      "loss": 2.6577,
+      "step": 8202
+    },
+    {
+      "epoch": 0.7425545396940346,
+      "grad_norm": 0.8346498012542725,
+      "learning_rate": 1.582321596058205e-05,
+      "loss": 2.5666,
+      "step": 8203
+    },
+    {
+      "epoch": 0.7426450620077849,
+      "grad_norm": 0.8698704838752747,
+      "learning_rate": 1.5806089326047512e-05,
+      "loss": 2.5968,
+      "step": 8204
+    },
+    {
+      "epoch": 0.7427355843215353,
+      "grad_norm": 0.792100727558136,
+      "learning_rate": 1.5788971169837473e-05,
+      "loss": 2.6165,
+      "step": 8205
+    },
+    {
+      "epoch": 0.7428261066352856,
+      "grad_norm": 0.8512359857559204,
+      "learning_rate": 1.5771861493675732e-05,
+      "loss": 2.6735,
+      "step": 8206
+    },
+    {
+      "epoch": 0.742916628949036,
+      "grad_norm": 0.9061763286590576,
+      "learning_rate": 1.5754760299285252e-05,
+      "loss": 2.6143,
+      "step": 8207
+    },
+    {
+      "epoch": 0.7430071512627863,
+      "grad_norm": 0.8766939043998718,
+      "learning_rate": 1.5737667588388093e-05,
+      "loss": 2.6372,
+      "step": 8208
+    },
+    {
+      "epoch": 0.7430976735765367,
+      "grad_norm": 0.9132415056228638,
+      "learning_rate": 1.5720583362705498e-05,
+      "loss": 2.6132,
+      "step": 8209
+    },
+    {
+      "epoch": 0.743188195890287,
+      "grad_norm": 0.7986286282539368,
+      "learning_rate": 1.5703507623957848e-05,
+      "loss": 2.6779,
+      "step": 8210
+    },
+    {
+      "epoch": 0.7432787182040373,
+      "grad_norm": 0.8528510332107544,
+      "learning_rate": 1.5686440373864675e-05,
+      "loss": 2.6198,
+      "step": 8211
+    },
+    {
+      "epoch": 0.7433692405177876,
+      "grad_norm": 0.8899943828582764,
+      "learning_rate": 1.5669381614144685e-05,
+      "loss": 2.6477,
+      "step": 8212
+    },
+    {
+      "epoch": 0.743459762831538,
+      "grad_norm": 0.8294568657875061,
+      "learning_rate": 1.5652331346515592e-05,
+      "loss": 2.6188,
+      "step": 8213
+    },
+    {
+      "epoch": 0.7435502851452883,
+      "grad_norm": 0.7675567269325256,
+      "learning_rate": 1.563528957269449e-05,
+      "loss": 2.5777,
+      "step": 8214
+    },
+    {
+      "epoch": 0.7436408074590386,
+      "grad_norm": 0.8410840034484863,
+      "learning_rate": 1.5618256294397382e-05,
+      "loss": 2.6778,
+      "step": 8215
+    },
+    {
+      "epoch": 0.743731329772789,
+      "grad_norm": 0.8316123485565186,
+      "learning_rate": 1.5601231513339565e-05,
+      "loss": 2.6331,
+      "step": 8216
+    },
+    {
+      "epoch": 0.7438218520865393,
+      "grad_norm": 0.7986269593238831,
+      "learning_rate": 1.558421523123542e-05,
+      "loss": 2.642,
+      "step": 8217
+    },
+    {
+      "epoch": 0.7439123744002897,
+      "grad_norm": 0.862214207649231,
+      "learning_rate": 1.5567207449798515e-05,
+      "loss": 2.6645,
+      "step": 8218
+    },
+    {
+      "epoch": 0.74400289671404,
+      "grad_norm": 0.8295678496360779,
+      "learning_rate": 1.5550208170741463e-05,
+      "loss": 2.628,
+      "step": 8219
+    },
+    {
+      "epoch": 0.7440934190277904,
+      "grad_norm": 0.8058247566223145,
+      "learning_rate": 1.553321739577619e-05,
+      "loss": 2.6442,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7441839413415406,
+      "grad_norm": 0.8965185284614563,
+      "learning_rate": 1.551623512661359e-05,
+      "loss": 2.6656,
+      "step": 8221
+    },
+    {
+      "epoch": 0.744274463655291,
+      "grad_norm": 0.8854925036430359,
+      "learning_rate": 1.5499261364963792e-05,
+      "loss": 2.6364,
+      "step": 8222
+    },
+    {
+      "epoch": 0.7443649859690413,
+      "grad_norm": 0.8158112168312073,
+      "learning_rate": 1.5482296112536067e-05,
+      "loss": 2.5834,
+      "step": 8223
+    },
+    {
+      "epoch": 0.7444555082827917,
+      "grad_norm": 0.9737561345100403,
+      "learning_rate": 1.546533937103881e-05,
+      "loss": 2.6548,
+      "step": 8224
+    },
+    {
+      "epoch": 0.744546030596542,
+      "grad_norm": 0.8352574110031128,
+      "learning_rate": 1.5448391142179574e-05,
+      "loss": 2.6044,
+      "step": 8225
+    },
+    {
+      "epoch": 0.7446365529102924,
+      "grad_norm": 0.8467913269996643,
+      "learning_rate": 1.5431451427665057e-05,
+      "loss": 2.6274,
+      "step": 8226
+    },
+    {
+      "epoch": 0.7447270752240427,
+      "grad_norm": 0.867701530456543,
+      "learning_rate": 1.541452022920101e-05,
+      "loss": 2.6024,
+      "step": 8227
+    },
+    {
+      "epoch": 0.7448175975377931,
+      "grad_norm": 0.8936637043952942,
+      "learning_rate": 1.5397597548492514e-05,
+      "loss": 2.6748,
+      "step": 8228
+    },
+    {
+      "epoch": 0.7449081198515434,
+      "grad_norm": 0.8338824510574341,
+      "learning_rate": 1.538068338724361e-05,
+      "loss": 2.6225,
+      "step": 8229
+    },
+    {
+      "epoch": 0.7449986421652938,
+      "grad_norm": 0.8347210884094238,
+      "learning_rate": 1.5363777747157572e-05,
+      "loss": 2.6783,
+      "step": 8230
+    },
+    {
+      "epoch": 0.745089164479044,
+      "grad_norm": 0.8179384469985962,
+      "learning_rate": 1.53468806299368e-05,
+      "loss": 2.615,
+      "step": 8231
+    },
+    {
+      "epoch": 0.7451796867927944,
+      "grad_norm": 0.8636762499809265,
+      "learning_rate": 1.5329992037282825e-05,
+      "loss": 2.685,
+      "step": 8232
+    },
+    {
+      "epoch": 0.7452702091065447,
+      "grad_norm": 0.793683648109436,
+      "learning_rate": 1.5313111970896345e-05,
+      "loss": 2.607,
+      "step": 8233
+    },
+    {
+      "epoch": 0.7453607314202951,
+      "grad_norm": 0.8522347211837769,
+      "learning_rate": 1.5296240432477195e-05,
+      "loss": 2.6216,
+      "step": 8234
+    },
+    {
+      "epoch": 0.7454512537340454,
+      "grad_norm": 0.8533903360366821,
+      "learning_rate": 1.5279377423724263e-05,
+      "loss": 2.5714,
+      "step": 8235
+    },
+    {
+      "epoch": 0.7455417760477958,
+      "grad_norm": 0.8542166948318481,
+      "learning_rate": 1.5262522946335755e-05,
+      "loss": 2.6102,
+      "step": 8236
+    },
+    {
+      "epoch": 0.7456322983615461,
+      "grad_norm": 0.9041294455528259,
+      "learning_rate": 1.524567700200884e-05,
+      "loss": 2.5996,
+      "step": 8237
+    },
+    {
+      "epoch": 0.7457228206752965,
+      "grad_norm": 0.8735194206237793,
+      "learning_rate": 1.5228839592439947e-05,
+      "loss": 2.6817,
+      "step": 8238
+    },
+    {
+      "epoch": 0.7458133429890468,
+      "grad_norm": 0.8800003528594971,
+      "learning_rate": 1.5212010719324577e-05,
+      "loss": 2.6788,
+      "step": 8239
+    },
+    {
+      "epoch": 0.7459038653027972,
+      "grad_norm": 0.791961133480072,
+      "learning_rate": 1.5195190384357404e-05,
+      "loss": 2.6099,
+      "step": 8240
+    },
+    {
+      "epoch": 0.7459943876165475,
+      "grad_norm": 0.8906918168067932,
+      "learning_rate": 1.5178378589232246e-05,
+      "loss": 2.6168,
+      "step": 8241
+    },
+    {
+      "epoch": 0.7460849099302979,
+      "grad_norm": 0.8241778612136841,
+      "learning_rate": 1.5161575335642064e-05,
+      "loss": 2.6929,
+      "step": 8242
+    },
+    {
+      "epoch": 0.7461754322440481,
+      "grad_norm": 0.8503902554512024,
+      "learning_rate": 1.5144780625278876e-05,
+      "loss": 2.5645,
+      "step": 8243
+    },
+    {
+      "epoch": 0.7462659545577985,
+      "grad_norm": 0.8449828028678894,
+      "learning_rate": 1.5127994459834006e-05,
+      "loss": 2.6398,
+      "step": 8244
+    },
+    {
+      "epoch": 0.7463564768715488,
+      "grad_norm": 0.8013359904289246,
+      "learning_rate": 1.5111216840997743e-05,
+      "loss": 2.6211,
+      "step": 8245
+    },
+    {
+      "epoch": 0.7464469991852992,
+      "grad_norm": 0.85307776927948,
+      "learning_rate": 1.509444777045963e-05,
+      "loss": 2.6471,
+      "step": 8246
+    },
+    {
+      "epoch": 0.7465375214990495,
+      "grad_norm": 0.829259991645813,
+      "learning_rate": 1.5077687249908302e-05,
+      "loss": 2.6416,
+      "step": 8247
+    },
+    {
+      "epoch": 0.7466280438127999,
+      "grad_norm": 0.8358555436134338,
+      "learning_rate": 1.5060935281031563e-05,
+      "loss": 2.6283,
+      "step": 8248
+    },
+    {
+      "epoch": 0.7467185661265502,
+      "grad_norm": 0.8651643991470337,
+      "learning_rate": 1.5044191865516266e-05,
+      "loss": 2.5934,
+      "step": 8249
+    },
+    {
+      "epoch": 0.7468090884403006,
+      "grad_norm": 0.828653872013092,
+      "learning_rate": 1.5027457005048573e-05,
+      "loss": 2.6596,
+      "step": 8250
+    },
+    {
+      "epoch": 0.7468996107540509,
+      "grad_norm": 0.8189446926116943,
+      "learning_rate": 1.5010730701313625e-05,
+      "loss": 2.6267,
+      "step": 8251
+    },
+    {
+      "epoch": 0.7469901330678013,
+      "grad_norm": 0.8088429570198059,
+      "learning_rate": 1.499401295599575e-05,
+      "loss": 2.6345,
+      "step": 8252
+    },
+    {
+      "epoch": 0.7470806553815516,
+      "grad_norm": 0.8520417213439941,
+      "learning_rate": 1.4977303770778462e-05,
+      "loss": 2.627,
+      "step": 8253
+    },
+    {
+      "epoch": 0.747171177695302,
+      "grad_norm": 0.7819222211837769,
+      "learning_rate": 1.4960603147344343e-05,
+      "loss": 2.5788,
+      "step": 8254
+    },
+    {
+      "epoch": 0.7472617000090522,
+      "grad_norm": 0.8902117609977722,
+      "learning_rate": 1.4943911087375173e-05,
+      "loss": 2.605,
+      "step": 8255
+    },
+    {
+      "epoch": 0.7473522223228025,
+      "grad_norm": 0.814746081829071,
+      "learning_rate": 1.492722759255184e-05,
+      "loss": 2.6328,
+      "step": 8256
+    },
+    {
+      "epoch": 0.7474427446365529,
+      "grad_norm": 0.8802046775817871,
+      "learning_rate": 1.4910552664554323e-05,
+      "loss": 2.6181,
+      "step": 8257
+    },
+    {
+      "epoch": 0.7475332669503032,
+      "grad_norm": 0.8508935570716858,
+      "learning_rate": 1.4893886305061865e-05,
+      "loss": 2.6653,
+      "step": 8258
+    },
+    {
+      "epoch": 0.7476237892640536,
+      "grad_norm": 0.8923085927963257,
+      "learning_rate": 1.4877228515752717e-05,
+      "loss": 2.6272,
+      "step": 8259
+    },
+    {
+      "epoch": 0.7477143115778039,
+      "grad_norm": 0.897159218788147,
+      "learning_rate": 1.4860579298304312e-05,
+      "loss": 2.5451,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7478048338915543,
+      "grad_norm": 0.8598203659057617,
+      "learning_rate": 1.4843938654393253e-05,
+      "loss": 2.6218,
+      "step": 8261
+    },
+    {
+      "epoch": 0.7478953562053046,
+      "grad_norm": 0.8585354089736938,
+      "learning_rate": 1.4827306585695234e-05,
+      "loss": 2.6232,
+      "step": 8262
+    },
+    {
+      "epoch": 0.747985878519055,
+      "grad_norm": 0.8278530836105347,
+      "learning_rate": 1.4810683093885126e-05,
+      "loss": 2.6651,
+      "step": 8263
+    },
+    {
+      "epoch": 0.7480764008328052,
+      "grad_norm": 0.8540686964988708,
+      "learning_rate": 1.479406818063691e-05,
+      "loss": 2.6294,
+      "step": 8264
+    },
+    {
+      "epoch": 0.7481669231465556,
+      "grad_norm": 0.8882604837417603,
+      "learning_rate": 1.4777461847623653e-05,
+      "loss": 2.6104,
+      "step": 8265
+    },
+    {
+      "epoch": 0.7482574454603059,
+      "grad_norm": 1.0102626085281372,
+      "learning_rate": 1.4760864096517701e-05,
+      "loss": 2.6443,
+      "step": 8266
+    },
+    {
+      "epoch": 0.7483479677740563,
+      "grad_norm": 0.8543476462364197,
+      "learning_rate": 1.4744274928990376e-05,
+      "loss": 2.647,
+      "step": 8267
+    },
+    {
+      "epoch": 0.7484384900878066,
+      "grad_norm": 0.8158116340637207,
+      "learning_rate": 1.472769434671224e-05,
+      "loss": 2.6096,
+      "step": 8268
+    },
+    {
+      "epoch": 0.748529012401557,
+      "grad_norm": 0.8607696890830994,
+      "learning_rate": 1.4711122351352947e-05,
+      "loss": 2.6002,
+      "step": 8269
+    },
+    {
+      "epoch": 0.7486195347153073,
+      "grad_norm": 0.8293615579605103,
+      "learning_rate": 1.4694558944581293e-05,
+      "loss": 2.6072,
+      "step": 8270
+    },
+    {
+      "epoch": 0.7487100570290577,
+      "grad_norm": 0.8597177863121033,
+      "learning_rate": 1.4678004128065225e-05,
+      "loss": 2.6456,
+      "step": 8271
+    },
+    {
+      "epoch": 0.748800579342808,
+      "grad_norm": 0.7982829213142395,
+      "learning_rate": 1.466145790347183e-05,
+      "loss": 2.5905,
+      "step": 8272
+    },
+    {
+      "epoch": 0.7488911016565584,
+      "grad_norm": 0.8462602496147156,
+      "learning_rate": 1.4644920272467244e-05,
+      "loss": 2.619,
+      "step": 8273
+    },
+    {
+      "epoch": 0.7489816239703087,
+      "grad_norm": 0.8661282658576965,
+      "learning_rate": 1.4628391236716899e-05,
+      "loss": 2.6635,
+      "step": 8274
+    },
+    {
+      "epoch": 0.749072146284059,
+      "grad_norm": 0.838843822479248,
+      "learning_rate": 1.4611870797885197e-05,
+      "loss": 2.6534,
+      "step": 8275
+    },
+    {
+      "epoch": 0.7491626685978093,
+      "grad_norm": 0.984654426574707,
+      "learning_rate": 1.4595358957635763e-05,
+      "loss": 2.6257,
+      "step": 8276
+    },
+    {
+      "epoch": 0.7492531909115597,
+      "grad_norm": 0.8167694807052612,
+      "learning_rate": 1.4578855717631357e-05,
+      "loss": 2.5533,
+      "step": 8277
+    },
+    {
+      "epoch": 0.74934371322531,
+      "grad_norm": 0.8685299754142761,
+      "learning_rate": 1.4562361079533849e-05,
+      "loss": 2.556,
+      "step": 8278
+    },
+    {
+      "epoch": 0.7494342355390604,
+      "grad_norm": 0.8326447606086731,
+      "learning_rate": 1.454587504500421e-05,
+      "loss": 2.6453,
+      "step": 8279
+    },
+    {
+      "epoch": 0.7495247578528107,
+      "grad_norm": 0.8826172947883606,
+      "learning_rate": 1.4529397615702656e-05,
+      "loss": 2.6536,
+      "step": 8280
+    },
+    {
+      "epoch": 0.7496152801665611,
+      "grad_norm": 0.8852538466453552,
+      "learning_rate": 1.4512928793288405e-05,
+      "loss": 2.6287,
+      "step": 8281
+    },
+    {
+      "epoch": 0.7497058024803114,
+      "grad_norm": 0.8458629846572876,
+      "learning_rate": 1.4496468579419886e-05,
+      "loss": 2.5647,
+      "step": 8282
+    },
+    {
+      "epoch": 0.7497963247940618,
+      "grad_norm": 0.9038055539131165,
+      "learning_rate": 1.4480016975754629e-05,
+      "loss": 2.6955,
+      "step": 8283
+    },
+    {
+      "epoch": 0.7498868471078121,
+      "grad_norm": 0.823957085609436,
+      "learning_rate": 1.4463573983949341e-05,
+      "loss": 2.6589,
+      "step": 8284
+    },
+    {
+      "epoch": 0.7499773694215625,
+      "grad_norm": 0.8328465223312378,
+      "learning_rate": 1.4447139605659799e-05,
+      "loss": 2.6246,
+      "step": 8285
+    },
+    {
+      "epoch": 0.7500678917353127,
+      "grad_norm": 0.8769001364707947,
+      "learning_rate": 1.4430713842540989e-05,
+      "loss": 2.6395,
+      "step": 8286
+    },
+    {
+      "epoch": 0.7501584140490631,
+      "grad_norm": 0.8344104290008545,
+      "learning_rate": 1.44142966962469e-05,
+      "loss": 2.6723,
+      "step": 8287
+    },
+    {
+      "epoch": 0.7502489363628134,
+      "grad_norm": 0.8472191095352173,
+      "learning_rate": 1.439788816843084e-05,
+      "loss": 2.6124,
+      "step": 8288
+    },
+    {
+      "epoch": 0.7503394586765638,
+      "grad_norm": 0.8723897933959961,
+      "learning_rate": 1.438148826074508e-05,
+      "loss": 2.672,
+      "step": 8289
+    },
+    {
+      "epoch": 0.7504299809903141,
+      "grad_norm": 0.822022020816803,
+      "learning_rate": 1.4365096974841108e-05,
+      "loss": 2.6048,
+      "step": 8290
+    },
+    {
+      "epoch": 0.7505205033040645,
+      "grad_norm": 0.8745297193527222,
+      "learning_rate": 1.4348714312369527e-05,
+      "loss": 2.6106,
+      "step": 8291
+    },
+    {
+      "epoch": 0.7506110256178148,
+      "grad_norm": 0.8441516160964966,
+      "learning_rate": 1.4332340274980072e-05,
+      "loss": 2.5962,
+      "step": 8292
+    },
+    {
+      "epoch": 0.7507015479315652,
+      "grad_norm": 0.8232734799385071,
+      "learning_rate": 1.4315974864321602e-05,
+      "loss": 2.6471,
+      "step": 8293
+    },
+    {
+      "epoch": 0.7507920702453155,
+      "grad_norm": 0.9545968770980835,
+      "learning_rate": 1.4299618082042143e-05,
+      "loss": 2.6994,
+      "step": 8294
+    },
+    {
+      "epoch": 0.7508825925590659,
+      "grad_norm": 0.8935458064079285,
+      "learning_rate": 1.4283269929788779e-05,
+      "loss": 2.6443,
+      "step": 8295
+    },
+    {
+      "epoch": 0.7509731148728161,
+      "grad_norm": 0.8546053767204285,
+      "learning_rate": 1.4266930409207791e-05,
+      "loss": 2.615,
+      "step": 8296
+    },
+    {
+      "epoch": 0.7510636371865664,
+      "grad_norm": 0.8507965803146362,
+      "learning_rate": 1.4250599521944563e-05,
+      "loss": 2.6345,
+      "step": 8297
+    },
+    {
+      "epoch": 0.7511541595003168,
+      "grad_norm": 0.799853503704071,
+      "learning_rate": 1.4234277269643614e-05,
+      "loss": 2.6021,
+      "step": 8298
+    },
+    {
+      "epoch": 0.7512446818140671,
+      "grad_norm": 0.8891720771789551,
+      "learning_rate": 1.4217963653948618e-05,
+      "loss": 2.5909,
+      "step": 8299
+    },
+    {
+      "epoch": 0.7513352041278175,
+      "grad_norm": 0.8445886969566345,
+      "learning_rate": 1.4201658676502294e-05,
+      "loss": 2.6488,
+      "step": 8300
+    },
+    {
+      "epoch": 0.7514257264415678,
+      "grad_norm": 0.8394327759742737,
+      "learning_rate": 1.418536233894664e-05,
+      "loss": 2.6577,
+      "step": 8301
+    },
+    {
+      "epoch": 0.7515162487553182,
+      "grad_norm": 0.7951192259788513,
+      "learning_rate": 1.416907464292262e-05,
+      "loss": 2.5963,
+      "step": 8302
+    },
+    {
+      "epoch": 0.7516067710690685,
+      "grad_norm": 0.8702280521392822,
+      "learning_rate": 1.4152795590070445e-05,
+      "loss": 2.6286,
+      "step": 8303
+    },
+    {
+      "epoch": 0.7516972933828189,
+      "grad_norm": 0.8904082179069519,
+      "learning_rate": 1.4136525182029392e-05,
+      "loss": 2.6705,
+      "step": 8304
+    },
+    {
+      "epoch": 0.7517878156965692,
+      "grad_norm": 0.8519523739814758,
+      "learning_rate": 1.4120263420437919e-05,
+      "loss": 2.5251,
+      "step": 8305
+    },
+    {
+      "epoch": 0.7518783380103196,
+      "grad_norm": 0.8387498259544373,
+      "learning_rate": 1.4104010306933557e-05,
+      "loss": 2.6455,
+      "step": 8306
+    },
+    {
+      "epoch": 0.7519688603240698,
+      "grad_norm": 0.8888024091720581,
+      "learning_rate": 1.4087765843153033e-05,
+      "loss": 2.6558,
+      "step": 8307
+    },
+    {
+      "epoch": 0.7520593826378202,
+      "grad_norm": 0.8241297006607056,
+      "learning_rate": 1.4071530030732095e-05,
+      "loss": 2.62,
+      "step": 8308
+    },
+    {
+      "epoch": 0.7521499049515705,
+      "grad_norm": 0.8831976652145386,
+      "learning_rate": 1.4055302871305775e-05,
+      "loss": 2.6161,
+      "step": 8309
+    },
+    {
+      "epoch": 0.7522404272653209,
+      "grad_norm": 0.9573201537132263,
+      "learning_rate": 1.4039084366508092e-05,
+      "loss": 2.6222,
+      "step": 8310
+    },
+    {
+      "epoch": 0.7523309495790712,
+      "grad_norm": 0.8343157768249512,
+      "learning_rate": 1.4022874517972251e-05,
+      "loss": 2.6234,
+      "step": 8311
+    },
+    {
+      "epoch": 0.7524214718928216,
+      "grad_norm": 0.8954187631607056,
+      "learning_rate": 1.4006673327330589e-05,
+      "loss": 2.66,
+      "step": 8312
+    },
+    {
+      "epoch": 0.7525119942065719,
+      "grad_norm": 0.8994336128234863,
+      "learning_rate": 1.39904807962146e-05,
+      "loss": 2.6723,
+      "step": 8313
+    },
+    {
+      "epoch": 0.7526025165203223,
+      "grad_norm": 0.8243401050567627,
+      "learning_rate": 1.3974296926254794e-05,
+      "loss": 2.6384,
+      "step": 8314
+    },
+    {
+      "epoch": 0.7526930388340726,
+      "grad_norm": 0.8115041851997375,
+      "learning_rate": 1.3958121719080986e-05,
+      "loss": 2.6071,
+      "step": 8315
+    },
+    {
+      "epoch": 0.752783561147823,
+      "grad_norm": 0.8584572076797485,
+      "learning_rate": 1.394195517632193e-05,
+      "loss": 2.6322,
+      "step": 8316
+    },
+    {
+      "epoch": 0.7528740834615733,
+      "grad_norm": 0.8973844647407532,
+      "learning_rate": 1.3925797299605647e-05,
+      "loss": 2.622,
+      "step": 8317
+    },
+    {
+      "epoch": 0.7529646057753236,
+      "grad_norm": 0.8394388556480408,
+      "learning_rate": 1.3909648090559212e-05,
+      "loss": 2.5859,
+      "step": 8318
+    },
+    {
+      "epoch": 0.7530551280890739,
+      "grad_norm": 0.8554968237876892,
+      "learning_rate": 1.389350755080886e-05,
+      "loss": 2.6101,
+      "step": 8319
+    },
+    {
+      "epoch": 0.7531456504028243,
+      "grad_norm": 0.8254625797271729,
+      "learning_rate": 1.3877375681979943e-05,
+      "loss": 2.609,
+      "step": 8320
+    },
+    {
+      "epoch": 0.7532361727165746,
+      "grad_norm": 0.9236030578613281,
+      "learning_rate": 1.3861252485696963e-05,
+      "loss": 2.7006,
+      "step": 8321
+    },
+    {
+      "epoch": 0.753326695030325,
+      "grad_norm": 0.9001591801643372,
+      "learning_rate": 1.3845137963583454e-05,
+      "loss": 2.5643,
+      "step": 8322
+    },
+    {
+      "epoch": 0.7534172173440753,
+      "grad_norm": 0.844906210899353,
+      "learning_rate": 1.3829032117262242e-05,
+      "loss": 2.6529,
+      "step": 8323
+    },
+    {
+      "epoch": 0.7535077396578257,
+      "grad_norm": 0.87331622838974,
+      "learning_rate": 1.3812934948355116e-05,
+      "loss": 2.6026,
+      "step": 8324
+    },
+    {
+      "epoch": 0.753598261971576,
+      "grad_norm": 0.814003586769104,
+      "learning_rate": 1.379684645848307e-05,
+      "loss": 2.6305,
+      "step": 8325
+    },
+    {
+      "epoch": 0.7536887842853264,
+      "grad_norm": 0.8583101034164429,
+      "learning_rate": 1.3780766649266242e-05,
+      "loss": 2.5818,
+      "step": 8326
+    },
+    {
+      "epoch": 0.7537793065990767,
+      "grad_norm": 0.792962908744812,
+      "learning_rate": 1.3764695522323845e-05,
+      "loss": 2.6312,
+      "step": 8327
+    },
+    {
+      "epoch": 0.7538698289128271,
+      "grad_norm": 0.8592725396156311,
+      "learning_rate": 1.3748633079274253e-05,
+      "loss": 2.546,
+      "step": 8328
+    },
+    {
+      "epoch": 0.7539603512265773,
+      "grad_norm": 0.873782753944397,
+      "learning_rate": 1.3732579321734972e-05,
+      "loss": 2.6973,
+      "step": 8329
+    },
+    {
+      "epoch": 0.7540508735403277,
+      "grad_norm": 0.8384394645690918,
+      "learning_rate": 1.3716534251322544e-05,
+      "loss": 2.6215,
+      "step": 8330
+    },
+    {
+      "epoch": 0.754141395854078,
+      "grad_norm": 0.9007170796394348,
+      "learning_rate": 1.3700497869652807e-05,
+      "loss": 2.6467,
+      "step": 8331
+    },
+    {
+      "epoch": 0.7542319181678284,
+      "grad_norm": 0.7979040145874023,
+      "learning_rate": 1.3684470178340548e-05,
+      "loss": 2.6331,
+      "step": 8332
+    },
+    {
+      "epoch": 0.7543224404815787,
+      "grad_norm": 0.8360491991043091,
+      "learning_rate": 1.3668451178999775e-05,
+      "loss": 2.5763,
+      "step": 8333
+    },
+    {
+      "epoch": 0.7544129627953291,
+      "grad_norm": 0.8180165886878967,
+      "learning_rate": 1.3652440873243622e-05,
+      "loss": 2.5277,
+      "step": 8334
+    },
+    {
+      "epoch": 0.7545034851090794,
+      "grad_norm": 0.9115409255027771,
+      "learning_rate": 1.3636439262684298e-05,
+      "loss": 2.616,
+      "step": 8335
+    },
+    {
+      "epoch": 0.7545940074228298,
+      "grad_norm": 0.8859723806381226,
+      "learning_rate": 1.362044634893318e-05,
+      "loss": 2.6118,
+      "step": 8336
+    },
+    {
+      "epoch": 0.7546845297365801,
+      "grad_norm": 0.8769333958625793,
+      "learning_rate": 1.360446213360077e-05,
+      "loss": 2.6341,
+      "step": 8337
+    },
+    {
+      "epoch": 0.7547750520503304,
+      "grad_norm": 0.8497744202613831,
+      "learning_rate": 1.3588486618296615e-05,
+      "loss": 2.6316,
+      "step": 8338
+    },
+    {
+      "epoch": 0.7548655743640807,
+      "grad_norm": 0.8281365633010864,
+      "learning_rate": 1.3572519804629536e-05,
+      "loss": 2.6588,
+      "step": 8339
+    },
+    {
+      "epoch": 0.754956096677831,
+      "grad_norm": 0.888231635093689,
+      "learning_rate": 1.3556561694207338e-05,
+      "loss": 2.6359,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7550466189915814,
+      "grad_norm": 0.893904447555542,
+      "learning_rate": 1.3540612288637e-05,
+      "loss": 2.5951,
+      "step": 8341
+    },
+    {
+      "epoch": 0.7551371413053317,
+      "grad_norm": 0.8344972729682922,
+      "learning_rate": 1.352467158952464e-05,
+      "loss": 2.5994,
+      "step": 8342
+    },
+    {
+      "epoch": 0.7552276636190821,
+      "grad_norm": 0.8195839524269104,
+      "learning_rate": 1.3508739598475506e-05,
+      "loss": 2.6477,
+      "step": 8343
+    },
+    {
+      "epoch": 0.7553181859328324,
+      "grad_norm": 0.8574609160423279,
+      "learning_rate": 1.3492816317093893e-05,
+      "loss": 2.7042,
+      "step": 8344
+    },
+    {
+      "epoch": 0.7554087082465828,
+      "grad_norm": 0.846380352973938,
+      "learning_rate": 1.3476901746983351e-05,
+      "loss": 2.5925,
+      "step": 8345
+    },
+    {
+      "epoch": 0.7554992305603331,
+      "grad_norm": 0.8791022300720215,
+      "learning_rate": 1.3460995889746408e-05,
+      "loss": 2.6581,
+      "step": 8346
+    },
+    {
+      "epoch": 0.7555897528740835,
+      "grad_norm": 0.8232809901237488,
+      "learning_rate": 1.3445098746984808e-05,
+      "loss": 2.6401,
+      "step": 8347
+    },
+    {
+      "epoch": 0.7556802751878338,
+      "grad_norm": 0.814947247505188,
+      "learning_rate": 1.3429210320299401e-05,
+      "loss": 2.6589,
+      "step": 8348
+    },
+    {
+      "epoch": 0.7557707975015842,
+      "grad_norm": 0.8401846885681152,
+      "learning_rate": 1.3413330611290154e-05,
+      "loss": 2.6416,
+      "step": 8349
+    },
+    {
+      "epoch": 0.7558613198153344,
+      "grad_norm": 0.8400335907936096,
+      "learning_rate": 1.339745962155613e-05,
+      "loss": 2.6249,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7559518421290848,
+      "grad_norm": 0.8805189728736877,
+      "learning_rate": 1.3381597352695586e-05,
+      "loss": 2.635,
+      "step": 8351
+    },
+    {
+      "epoch": 0.7560423644428351,
+      "grad_norm": 0.8336710929870605,
+      "learning_rate": 1.3365743806305774e-05,
+      "loss": 2.6311,
+      "step": 8352
+    },
+    {
+      "epoch": 0.7561328867565855,
+      "grad_norm": 0.9214432835578918,
+      "learning_rate": 1.3349898983983233e-05,
+      "loss": 2.6312,
+      "step": 8353
+    },
+    {
+      "epoch": 0.7562234090703358,
+      "grad_norm": 0.8631122708320618,
+      "learning_rate": 1.3334062887323461e-05,
+      "loss": 2.628,
+      "step": 8354
+    },
+    {
+      "epoch": 0.7563139313840862,
+      "grad_norm": 0.8307778835296631,
+      "learning_rate": 1.3318235517921196e-05,
+      "loss": 2.6126,
+      "step": 8355
+    },
+    {
+      "epoch": 0.7564044536978365,
+      "grad_norm": 0.902026355266571,
+      "learning_rate": 1.3302416877370239e-05,
+      "loss": 2.7022,
+      "step": 8356
+    },
+    {
+      "epoch": 0.7564949760115869,
+      "grad_norm": 0.8334587812423706,
+      "learning_rate": 1.3286606967263537e-05,
+      "loss": 2.6638,
+      "step": 8357
+    },
+    {
+      "epoch": 0.7565854983253372,
+      "grad_norm": 0.8361995816230774,
+      "learning_rate": 1.327080578919313e-05,
+      "loss": 2.6521,
+      "step": 8358
+    },
+    {
+      "epoch": 0.7566760206390876,
+      "grad_norm": 0.8585364818572998,
+      "learning_rate": 1.3255013344750233e-05,
+      "loss": 2.6088,
+      "step": 8359
+    },
+    {
+      "epoch": 0.7567665429528379,
+      "grad_norm": 0.8744394779205322,
+      "learning_rate": 1.3239229635525074e-05,
+      "loss": 2.6148,
+      "step": 8360
+    },
+    {
+      "epoch": 0.7568570652665882,
+      "grad_norm": 0.8813425898551941,
+      "learning_rate": 1.3223454663107172e-05,
+      "loss": 2.6994,
+      "step": 8361
+    },
+    {
+      "epoch": 0.7569475875803385,
+      "grad_norm": 0.8385540843009949,
+      "learning_rate": 1.3207688429084974e-05,
+      "loss": 2.6528,
+      "step": 8362
+    },
+    {
+      "epoch": 0.7570381098940889,
+      "grad_norm": 0.852018415927887,
+      "learning_rate": 1.3191930935046181e-05,
+      "loss": 2.6398,
+      "step": 8363
+    },
+    {
+      "epoch": 0.7571286322078392,
+      "grad_norm": 0.8620157241821289,
+      "learning_rate": 1.3176182182577567e-05,
+      "loss": 2.624,
+      "step": 8364
+    },
+    {
+      "epoch": 0.7572191545215896,
+      "grad_norm": 0.8437519669532776,
+      "learning_rate": 1.3160442173265031e-05,
+      "loss": 2.6425,
+      "step": 8365
+    },
+    {
+      "epoch": 0.7573096768353399,
+      "grad_norm": 1.0246024131774902,
+      "learning_rate": 1.3144710908693602e-05,
+      "loss": 2.6969,
+      "step": 8366
+    },
+    {
+      "epoch": 0.7574001991490903,
+      "grad_norm": 0.8751848340034485,
+      "learning_rate": 1.3128988390447416e-05,
+      "loss": 2.6285,
+      "step": 8367
+    },
+    {
+      "epoch": 0.7574907214628406,
+      "grad_norm": 0.8701299428939819,
+      "learning_rate": 1.3113274620109684e-05,
+      "loss": 2.6615,
+      "step": 8368
+    },
+    {
+      "epoch": 0.757581243776591,
+      "grad_norm": 0.9776567816734314,
+      "learning_rate": 1.3097569599262871e-05,
+      "loss": 2.6625,
+      "step": 8369
+    },
+    {
+      "epoch": 0.7576717660903413,
+      "grad_norm": 0.9123702049255371,
+      "learning_rate": 1.3081873329488392e-05,
+      "loss": 2.6172,
+      "step": 8370
+    },
+    {
+      "epoch": 0.7577622884040917,
+      "grad_norm": 0.8895208835601807,
+      "learning_rate": 1.3066185812366883e-05,
+      "loss": 2.6311,
+      "step": 8371
+    },
+    {
+      "epoch": 0.7578528107178419,
+      "grad_norm": 0.841701328754425,
+      "learning_rate": 1.30505070494781e-05,
+      "loss": 2.6267,
+      "step": 8372
+    },
+    {
+      "epoch": 0.7579433330315923,
+      "grad_norm": 0.8388078808784485,
+      "learning_rate": 1.3034837042400894e-05,
+      "loss": 2.6393,
+      "step": 8373
+    },
+    {
+      "epoch": 0.7580338553453426,
+      "grad_norm": 0.9033432602882385,
+      "learning_rate": 1.3019175792713167e-05,
+      "loss": 2.6579,
+      "step": 8374
+    },
+    {
+      "epoch": 0.758124377659093,
+      "grad_norm": 0.8371830582618713,
+      "learning_rate": 1.3003523301992104e-05,
+      "loss": 2.6308,
+      "step": 8375
+    },
+    {
+      "epoch": 0.7582148999728433,
+      "grad_norm": 0.8309195041656494,
+      "learning_rate": 1.2987879571813854e-05,
+      "loss": 2.6421,
+      "step": 8376
+    },
+    {
+      "epoch": 0.7583054222865937,
+      "grad_norm": 0.8373768329620361,
+      "learning_rate": 1.2972244603753736e-05,
+      "loss": 2.6414,
+      "step": 8377
+    },
+    {
+      "epoch": 0.758395944600344,
+      "grad_norm": 0.8145757913589478,
+      "learning_rate": 1.295661839938621e-05,
+      "loss": 2.6132,
+      "step": 8378
+    },
+    {
+      "epoch": 0.7584864669140943,
+      "grad_norm": 0.8609750866889954,
+      "learning_rate": 1.2941000960284832e-05,
+      "loss": 2.6015,
+      "step": 8379
+    },
+    {
+      "epoch": 0.7585769892278447,
+      "grad_norm": 0.8729307055473328,
+      "learning_rate": 1.2925392288022298e-05,
+      "loss": 2.6121,
+      "step": 8380
+    },
+    {
+      "epoch": 0.758667511541595,
+      "grad_norm": 0.9552813768386841,
+      "learning_rate": 1.290979238417035e-05,
+      "loss": 2.6482,
+      "step": 8381
+    },
+    {
+      "epoch": 0.7587580338553453,
+      "grad_norm": 0.8189707398414612,
+      "learning_rate": 1.289420125029993e-05,
+      "loss": 2.6284,
+      "step": 8382
+    },
+    {
+      "epoch": 0.7588485561690956,
+      "grad_norm": 0.8567572832107544,
+      "learning_rate": 1.2878618887981064e-05,
+      "loss": 2.642,
+      "step": 8383
+    },
+    {
+      "epoch": 0.758939078482846,
+      "grad_norm": 0.8377892374992371,
+      "learning_rate": 1.2863045298782895e-05,
+      "loss": 2.6071,
+      "step": 8384
+    },
+    {
+      "epoch": 0.7590296007965963,
+      "grad_norm": 0.903947114944458,
+      "learning_rate": 1.2847480484273666e-05,
+      "loss": 2.6266,
+      "step": 8385
+    },
+    {
+      "epoch": 0.7591201231103467,
+      "grad_norm": 0.8137821555137634,
+      "learning_rate": 1.2831924446020805e-05,
+      "loss": 2.6096,
+      "step": 8386
+    },
+    {
+      "epoch": 0.759210645424097,
+      "grad_norm": 0.935057520866394,
+      "learning_rate": 1.2816377185590711e-05,
+      "loss": 2.5789,
+      "step": 8387
+    },
+    {
+      "epoch": 0.7593011677378474,
+      "grad_norm": 0.9254269003868103,
+      "learning_rate": 1.28008387045491e-05,
+      "loss": 2.7207,
+      "step": 8388
+    },
+    {
+      "epoch": 0.7593916900515977,
+      "grad_norm": 0.8800764679908752,
+      "learning_rate": 1.2785309004460622e-05,
+      "loss": 2.6285,
+      "step": 8389
+    },
+    {
+      "epoch": 0.7594822123653481,
+      "grad_norm": 0.8191325664520264,
+      "learning_rate": 1.2769788086889134e-05,
+      "loss": 2.6175,
+      "step": 8390
+    },
+    {
+      "epoch": 0.7595727346790984,
+      "grad_norm": 0.9105868935585022,
+      "learning_rate": 1.275427595339761e-05,
+      "loss": 2.7079,
+      "step": 8391
+    },
+    {
+      "epoch": 0.7596632569928488,
+      "grad_norm": 0.90646892786026,
+      "learning_rate": 1.27387726055481e-05,
+      "loss": 2.6461,
+      "step": 8392
+    },
+    {
+      "epoch": 0.759753779306599,
+      "grad_norm": 0.8635348677635193,
+      "learning_rate": 1.272327804490181e-05,
+      "loss": 2.6116,
+      "step": 8393
+    },
+    {
+      "epoch": 0.7598443016203494,
+      "grad_norm": 0.8347007036209106,
+      "learning_rate": 1.2707792273019048e-05,
+      "loss": 2.6086,
+      "step": 8394
+    },
+    {
+      "epoch": 0.7599348239340997,
+      "grad_norm": 0.8188871741294861,
+      "learning_rate": 1.2692315291459178e-05,
+      "loss": 2.6129,
+      "step": 8395
+    },
+    {
+      "epoch": 0.7600253462478501,
+      "grad_norm": 0.8775956034660339,
+      "learning_rate": 1.267684710178081e-05,
+      "loss": 2.6255,
+      "step": 8396
+    },
+    {
+      "epoch": 0.7601158685616004,
+      "grad_norm": 0.8294050097465515,
+      "learning_rate": 1.2661387705541528e-05,
+      "loss": 2.6402,
+      "step": 8397
+    },
+    {
+      "epoch": 0.7602063908753508,
+      "grad_norm": 0.8578444123268127,
+      "learning_rate": 1.2645937104298111e-05,
+      "loss": 2.6559,
+      "step": 8398
+    },
+    {
+      "epoch": 0.7602969131891011,
+      "grad_norm": 0.8342133164405823,
+      "learning_rate": 1.2630495299606448e-05,
+      "loss": 2.6107,
+      "step": 8399
+    },
+    {
+      "epoch": 0.7603874355028515,
+      "grad_norm": 0.8250716328620911,
+      "learning_rate": 1.2615062293021507e-05,
+      "loss": 2.6742,
+      "step": 8400
+    },
+    {
+      "epoch": 0.7604779578166018,
+      "grad_norm": 0.8013415932655334,
+      "learning_rate": 1.2599638086097399e-05,
+      "loss": 2.5804,
+      "step": 8401
+    },
+    {
+      "epoch": 0.7605684801303522,
+      "grad_norm": 0.9561328291893005,
+      "learning_rate": 1.2584222680387381e-05,
+      "loss": 2.5898,
+      "step": 8402
+    },
+    {
+      "epoch": 0.7606590024441025,
+      "grad_norm": 0.8616392016410828,
+      "learning_rate": 1.2568816077443712e-05,
+      "loss": 2.5695,
+      "step": 8403
+    },
+    {
+      "epoch": 0.7607495247578528,
+      "grad_norm": 0.8710083365440369,
+      "learning_rate": 1.2553418278817874e-05,
+      "loss": 2.6597,
+      "step": 8404
+    },
+    {
+      "epoch": 0.7608400470716031,
+      "grad_norm": 0.926058292388916,
+      "learning_rate": 1.2538029286060426e-05,
+      "loss": 2.6538,
+      "step": 8405
+    },
+    {
+      "epoch": 0.7609305693853535,
+      "grad_norm": 0.9008700251579285,
+      "learning_rate": 1.252264910072104e-05,
+      "loss": 2.6652,
+      "step": 8406
+    },
+    {
+      "epoch": 0.7610210916991038,
+      "grad_norm": 0.8355394601821899,
+      "learning_rate": 1.2507277724348487e-05,
+      "loss": 2.5824,
+      "step": 8407
+    },
+    {
+      "epoch": 0.7611116140128542,
+      "grad_norm": 0.9210795760154724,
+      "learning_rate": 1.2491915158490697e-05,
+      "loss": 2.5588,
+      "step": 8408
+    },
+    {
+      "epoch": 0.7612021363266045,
+      "grad_norm": 0.8426117897033691,
+      "learning_rate": 1.247656140469462e-05,
+      "loss": 2.6513,
+      "step": 8409
+    },
+    {
+      "epoch": 0.7612926586403549,
+      "grad_norm": 0.8535647988319397,
+      "learning_rate": 1.2461216464506454e-05,
+      "loss": 2.6681,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7613831809541052,
+      "grad_norm": 0.9873753786087036,
+      "learning_rate": 1.2445880339471372e-05,
+      "loss": 2.615,
+      "step": 8411
+    },
+    {
+      "epoch": 0.7614737032678556,
+      "grad_norm": 0.8945762515068054,
+      "learning_rate": 1.243055303113374e-05,
+      "loss": 2.6129,
+      "step": 8412
+    },
+    {
+      "epoch": 0.7615642255816059,
+      "grad_norm": 0.9072584509849548,
+      "learning_rate": 1.2415234541037024e-05,
+      "loss": 2.573,
+      "step": 8413
+    },
+    {
+      "epoch": 0.7616547478953563,
+      "grad_norm": 0.8211448192596436,
+      "learning_rate": 1.23999248707238e-05,
+      "loss": 2.5749,
+      "step": 8414
+    },
+    {
+      "epoch": 0.7617452702091065,
+      "grad_norm": 0.9077050685882568,
+      "learning_rate": 1.2384624021735735e-05,
+      "loss": 2.692,
+      "step": 8415
+    },
+    {
+      "epoch": 0.7618357925228569,
+      "grad_norm": 0.8252149224281311,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 2.5867,
+      "step": 8416
+    },
+    {
+      "epoch": 0.7619263148366072,
+      "grad_norm": 0.9161756634712219,
+      "learning_rate": 1.235404879389741e-05,
+      "loss": 2.6664,
+      "step": 8417
+    },
+    {
+      "epoch": 0.7620168371503576,
+      "grad_norm": 0.9396390318870544,
+      "learning_rate": 1.2338774418126098e-05,
+      "loss": 2.6342,
+      "step": 8418
+    },
+    {
+      "epoch": 0.7621073594641079,
+      "grad_norm": 0.8113436102867126,
+      "learning_rate": 1.2323508869837775e-05,
+      "loss": 2.6522,
+      "step": 8419
+    },
+    {
+      "epoch": 0.7621978817778582,
+      "grad_norm": 0.9393036961555481,
+      "learning_rate": 1.230825215056971e-05,
+      "loss": 2.6139,
+      "step": 8420
+    },
+    {
+      "epoch": 0.7622884040916086,
+      "grad_norm": 0.8151003122329712,
+      "learning_rate": 1.2293004261858254e-05,
+      "loss": 2.6219,
+      "step": 8421
+    },
+    {
+      "epoch": 0.7623789264053589,
+      "grad_norm": 0.865507960319519,
+      "learning_rate": 1.2277765205238879e-05,
+      "loss": 2.6003,
+      "step": 8422
+    },
+    {
+      "epoch": 0.7624694487191093,
+      "grad_norm": 0.8855223059654236,
+      "learning_rate": 1.2262534982246132e-05,
+      "loss": 2.6697,
+      "step": 8423
+    },
+    {
+      "epoch": 0.7625599710328596,
+      "grad_norm": 0.9732133150100708,
+      "learning_rate": 1.2247313594413746e-05,
+      "loss": 2.5585,
+      "step": 8424
+    },
+    {
+      "epoch": 0.76265049334661,
+      "grad_norm": 0.8758246898651123,
+      "learning_rate": 1.2232101043274436e-05,
+      "loss": 2.6185,
+      "step": 8425
+    },
+    {
+      "epoch": 0.7627410156603602,
+      "grad_norm": 0.8588250279426575,
+      "learning_rate": 1.2216897330360178e-05,
+      "loss": 2.6275,
+      "step": 8426
+    },
+    {
+      "epoch": 0.7628315379741106,
+      "grad_norm": 0.9134284853935242,
+      "learning_rate": 1.2201702457201947e-05,
+      "loss": 2.7113,
+      "step": 8427
+    },
+    {
+      "epoch": 0.7629220602878609,
+      "grad_norm": 0.8985430598258972,
+      "learning_rate": 1.2186516425329874e-05,
+      "loss": 2.6369,
+      "step": 8428
+    },
+    {
+      "epoch": 0.7630125826016113,
+      "grad_norm": 0.8405027389526367,
+      "learning_rate": 1.2171339236273205e-05,
+      "loss": 2.613,
+      "step": 8429
+    },
+    {
+      "epoch": 0.7631031049153616,
+      "grad_norm": 0.8441949486732483,
+      "learning_rate": 1.2156170891560258e-05,
+      "loss": 2.6233,
+      "step": 8430
+    },
+    {
+      "epoch": 0.763193627229112,
+      "grad_norm": 0.84303218126297,
+      "learning_rate": 1.2141011392718504e-05,
+      "loss": 2.5981,
+      "step": 8431
+    },
+    {
+      "epoch": 0.7632841495428623,
+      "grad_norm": 0.9115005731582642,
+      "learning_rate": 1.212586074127452e-05,
+      "loss": 2.6523,
+      "step": 8432
+    },
+    {
+      "epoch": 0.7633746718566127,
+      "grad_norm": 0.8431458473205566,
+      "learning_rate": 1.2110718938753907e-05,
+      "loss": 2.6375,
+      "step": 8433
+    },
+    {
+      "epoch": 0.763465194170363,
+      "grad_norm": 0.9723434448242188,
+      "learning_rate": 1.2095585986681535e-05,
+      "loss": 2.6754,
+      "step": 8434
+    },
+    {
+      "epoch": 0.7635557164841134,
+      "grad_norm": 0.8683786392211914,
+      "learning_rate": 1.208046188658124e-05,
+      "loss": 2.6303,
+      "step": 8435
+    },
+    {
+      "epoch": 0.7636462387978636,
+      "grad_norm": 0.8498546481132507,
+      "learning_rate": 1.2065346639976016e-05,
+      "loss": 2.5578,
+      "step": 8436
+    },
+    {
+      "epoch": 0.763736761111614,
+      "grad_norm": 0.9147982001304626,
+      "learning_rate": 1.2050240248387978e-05,
+      "loss": 2.6561,
+      "step": 8437
+    },
+    {
+      "epoch": 0.7638272834253643,
+      "grad_norm": 0.8597583174705505,
+      "learning_rate": 1.2035142713338366e-05,
+      "loss": 2.5931,
+      "step": 8438
+    },
+    {
+      "epoch": 0.7639178057391147,
+      "grad_norm": 0.9011256694793701,
+      "learning_rate": 1.2020054036347439e-05,
+      "loss": 2.5613,
+      "step": 8439
+    },
+    {
+      "epoch": 0.764008328052865,
+      "grad_norm": 0.8712127804756165,
+      "learning_rate": 1.2004974218934695e-05,
+      "loss": 2.6386,
+      "step": 8440
+    },
+    {
+      "epoch": 0.7640988503666154,
+      "grad_norm": 0.8568911552429199,
+      "learning_rate": 1.1989903262618628e-05,
+      "loss": 2.6678,
+      "step": 8441
+    },
+    {
+      "epoch": 0.7641893726803657,
+      "grad_norm": 0.8371545076370239,
+      "learning_rate": 1.1974841168916884e-05,
+      "loss": 2.6751,
+      "step": 8442
+    },
+    {
+      "epoch": 0.7642798949941161,
+      "grad_norm": 0.8238829374313354,
+      "learning_rate": 1.1959787939346235e-05,
+      "loss": 2.6016,
+      "step": 8443
+    },
+    {
+      "epoch": 0.7643704173078664,
+      "grad_norm": 0.868220329284668,
+      "learning_rate": 1.1944743575422524e-05,
+      "loss": 2.5867,
+      "step": 8444
+    },
+    {
+      "epoch": 0.7644609396216168,
+      "grad_norm": 0.8624018430709839,
+      "learning_rate": 1.1929708078660728e-05,
+      "loss": 2.6194,
+      "step": 8445
+    },
+    {
+      "epoch": 0.764551461935367,
+      "grad_norm": 0.8850862979888916,
+      "learning_rate": 1.1914681450574949e-05,
+      "loss": 2.619,
+      "step": 8446
+    },
+    {
+      "epoch": 0.7646419842491174,
+      "grad_norm": 0.8665874600410461,
+      "learning_rate": 1.1899663692678287e-05,
+      "loss": 2.5906,
+      "step": 8447
+    },
+    {
+      "epoch": 0.7647325065628677,
+      "grad_norm": 0.8597021698951721,
+      "learning_rate": 1.188465480648312e-05,
+      "loss": 2.6345,
+      "step": 8448
+    },
+    {
+      "epoch": 0.7648230288766181,
+      "grad_norm": 0.8657194972038269,
+      "learning_rate": 1.1869654793500784e-05,
+      "loss": 2.6515,
+      "step": 8449
+    },
+    {
+      "epoch": 0.7649135511903684,
+      "grad_norm": 0.8500573635101318,
+      "learning_rate": 1.1854663655241805e-05,
+      "loss": 2.7176,
+      "step": 8450
+    },
+    {
+      "epoch": 0.7650040735041188,
+      "grad_norm": 0.8615239262580872,
+      "learning_rate": 1.1839681393215773e-05,
+      "loss": 2.56,
+      "step": 8451
+    },
+    {
+      "epoch": 0.7650945958178691,
+      "grad_norm": 0.7915595769882202,
+      "learning_rate": 1.1824708008931418e-05,
+      "loss": 2.5786,
+      "step": 8452
+    },
+    {
+      "epoch": 0.7651851181316195,
+      "grad_norm": 0.8926154971122742,
+      "learning_rate": 1.1809743503896554e-05,
+      "loss": 2.5991,
+      "step": 8453
+    },
+    {
+      "epoch": 0.7652756404453698,
+      "grad_norm": 0.8740158677101135,
+      "learning_rate": 1.179478787961813e-05,
+      "loss": 2.624,
+      "step": 8454
+    },
+    {
+      "epoch": 0.7653661627591202,
+      "grad_norm": 0.8418128490447998,
+      "learning_rate": 1.177984113760211e-05,
+      "loss": 2.5846,
+      "step": 8455
+    },
+    {
+      "epoch": 0.7654566850728705,
+      "grad_norm": 0.8274219632148743,
+      "learning_rate": 1.1764903279353712e-05,
+      "loss": 2.5793,
+      "step": 8456
+    },
+    {
+      "epoch": 0.7655472073866209,
+      "grad_norm": 0.8460656404495239,
+      "learning_rate": 1.1749974306377121e-05,
+      "loss": 2.5991,
+      "step": 8457
+    },
+    {
+      "epoch": 0.7656377297003711,
+      "grad_norm": 0.8882750868797302,
+      "learning_rate": 1.1735054220175711e-05,
+      "loss": 2.6413,
+      "step": 8458
+    },
+    {
+      "epoch": 0.7657282520141215,
+      "grad_norm": 0.895829975605011,
+      "learning_rate": 1.1720143022251917e-05,
+      "loss": 2.6272,
+      "step": 8459
+    },
+    {
+      "epoch": 0.7658187743278718,
+      "grad_norm": 0.8516156673431396,
+      "learning_rate": 1.1705240714107302e-05,
+      "loss": 2.635,
+      "step": 8460
+    },
+    {
+      "epoch": 0.7659092966416221,
+      "grad_norm": 0.8761586546897888,
+      "learning_rate": 1.1690347297242532e-05,
+      "loss": 2.6001,
+      "step": 8461
+    },
+    {
+      "epoch": 0.7659998189553725,
+      "grad_norm": 0.8189490437507629,
+      "learning_rate": 1.1675462773157398e-05,
+      "loss": 2.6178,
+      "step": 8462
+    },
+    {
+      "epoch": 0.7660903412691228,
+      "grad_norm": 0.8238328695297241,
+      "learning_rate": 1.166058714335071e-05,
+      "loss": 2.6499,
+      "step": 8463
+    },
+    {
+      "epoch": 0.7661808635828732,
+      "grad_norm": 0.8749627470970154,
+      "learning_rate": 1.1645720409320504e-05,
+      "loss": 2.6602,
+      "step": 8464
+    },
+    {
+      "epoch": 0.7662713858966235,
+      "grad_norm": 0.8527346253395081,
+      "learning_rate": 1.163086257256385e-05,
+      "loss": 2.6257,
+      "step": 8465
+    },
+    {
+      "epoch": 0.7663619082103739,
+      "grad_norm": 0.7932024002075195,
+      "learning_rate": 1.1616013634576873e-05,
+      "loss": 2.5665,
+      "step": 8466
+    },
+    {
+      "epoch": 0.7664524305241242,
+      "grad_norm": 0.8431447148323059,
+      "learning_rate": 1.1601173596854942e-05,
+      "loss": 2.698,
+      "step": 8467
+    },
+    {
+      "epoch": 0.7665429528378745,
+      "grad_norm": 0.8905431628227234,
+      "learning_rate": 1.1586342460892396e-05,
+      "loss": 2.5765,
+      "step": 8468
+    },
+    {
+      "epoch": 0.7666334751516248,
+      "grad_norm": 0.8475296497344971,
+      "learning_rate": 1.1571520228182742e-05,
+      "loss": 2.622,
+      "step": 8469
+    },
+    {
+      "epoch": 0.7667239974653752,
+      "grad_norm": 0.8639915585517883,
+      "learning_rate": 1.1556706900218572e-05,
+      "loss": 2.6325,
+      "step": 8470
+    },
+    {
+      "epoch": 0.7668145197791255,
+      "grad_norm": 0.8727537989616394,
+      "learning_rate": 1.1541902478491606e-05,
+      "loss": 2.6259,
+      "step": 8471
+    },
+    {
+      "epoch": 0.7669050420928759,
+      "grad_norm": 0.8499577045440674,
+      "learning_rate": 1.152710696449264e-05,
+      "loss": 2.627,
+      "step": 8472
+    },
+    {
+      "epoch": 0.7669955644066262,
+      "grad_norm": 0.7941082715988159,
+      "learning_rate": 1.1512320359711604e-05,
+      "loss": 2.6087,
+      "step": 8473
+    },
+    {
+      "epoch": 0.7670860867203766,
+      "grad_norm": 0.873528242111206,
+      "learning_rate": 1.1497542665637451e-05,
+      "loss": 2.5909,
+      "step": 8474
+    },
+    {
+      "epoch": 0.7671766090341269,
+      "grad_norm": 0.820582926273346,
+      "learning_rate": 1.1482773883758357e-05,
+      "loss": 2.6254,
+      "step": 8475
+    },
+    {
+      "epoch": 0.7672671313478773,
+      "grad_norm": 0.9067497253417969,
+      "learning_rate": 1.14680140155615e-05,
+      "loss": 2.6378,
+      "step": 8476
+    },
+    {
+      "epoch": 0.7673576536616276,
+      "grad_norm": 0.804765522480011,
+      "learning_rate": 1.1453263062533203e-05,
+      "loss": 2.6179,
+      "step": 8477
+    },
+    {
+      "epoch": 0.767448175975378,
+      "grad_norm": 0.8286309838294983,
+      "learning_rate": 1.14385210261589e-05,
+      "loss": 2.6059,
+      "step": 8478
+    },
+    {
+      "epoch": 0.7675386982891282,
+      "grad_norm": 0.8492158055305481,
+      "learning_rate": 1.14237879079231e-05,
+      "loss": 2.6408,
+      "step": 8479
+    },
+    {
+      "epoch": 0.7676292206028786,
+      "grad_norm": 0.8288713097572327,
+      "learning_rate": 1.1409063709309442e-05,
+      "loss": 2.5898,
+      "step": 8480
+    },
+    {
+      "epoch": 0.7677197429166289,
+      "grad_norm": 0.842309832572937,
+      "learning_rate": 1.139434843180066e-05,
+      "loss": 2.6221,
+      "step": 8481
+    },
+    {
+      "epoch": 0.7678102652303793,
+      "grad_norm": 0.9057137370109558,
+      "learning_rate": 1.1379642076878527e-05,
+      "loss": 2.6263,
+      "step": 8482
+    },
+    {
+      "epoch": 0.7679007875441296,
+      "grad_norm": 0.8420442938804626,
+      "learning_rate": 1.1364944646024045e-05,
+      "loss": 2.6164,
+      "step": 8483
+    },
+    {
+      "epoch": 0.76799130985788,
+      "grad_norm": 0.8523637056350708,
+      "learning_rate": 1.1350256140717197e-05,
+      "loss": 2.5767,
+      "step": 8484
+    },
+    {
+      "epoch": 0.7680818321716303,
+      "grad_norm": 0.9101349115371704,
+      "learning_rate": 1.1335576562437134e-05,
+      "loss": 2.6628,
+      "step": 8485
+    },
+    {
+      "epoch": 0.7681723544853807,
+      "grad_norm": 0.8948091268539429,
+      "learning_rate": 1.1320905912662084e-05,
+      "loss": 2.6112,
+      "step": 8486
+    },
+    {
+      "epoch": 0.768262876799131,
+      "grad_norm": 0.8722454905509949,
+      "learning_rate": 1.1306244192869386e-05,
+      "loss": 2.6569,
+      "step": 8487
+    },
+    {
+      "epoch": 0.7683533991128814,
+      "grad_norm": 0.8330963253974915,
+      "learning_rate": 1.1291591404535462e-05,
+      "loss": 2.6174,
+      "step": 8488
+    },
+    {
+      "epoch": 0.7684439214266316,
+      "grad_norm": 0.8809168934822083,
+      "learning_rate": 1.1276947549135875e-05,
+      "loss": 2.6485,
+      "step": 8489
+    },
+    {
+      "epoch": 0.768534443740382,
+      "grad_norm": 0.8144750595092773,
+      "learning_rate": 1.126231262814521e-05,
+      "loss": 2.608,
+      "step": 8490
+    },
+    {
+      "epoch": 0.7686249660541323,
+      "grad_norm": 0.8324765563011169,
+      "learning_rate": 1.124768664303727e-05,
+      "loss": 2.6431,
+      "step": 8491
+    },
+    {
+      "epoch": 0.7687154883678827,
+      "grad_norm": 0.8306261301040649,
+      "learning_rate": 1.123306959528484e-05,
+      "loss": 2.6201,
+      "step": 8492
+    },
+    {
+      "epoch": 0.768806010681633,
+      "grad_norm": 0.8392279148101807,
+      "learning_rate": 1.1218461486359877e-05,
+      "loss": 2.6706,
+      "step": 8493
+    },
+    {
+      "epoch": 0.7688965329953834,
+      "grad_norm": 0.8646116256713867,
+      "learning_rate": 1.1203862317733404e-05,
+      "loss": 2.6032,
+      "step": 8494
+    },
+    {
+      "epoch": 0.7689870553091337,
+      "grad_norm": 0.8765877485275269,
+      "learning_rate": 1.1189272090875591e-05,
+      "loss": 2.6159,
+      "step": 8495
+    },
+    {
+      "epoch": 0.7690775776228841,
+      "grad_norm": 0.8648416996002197,
+      "learning_rate": 1.1174690807255605e-05,
+      "loss": 2.6345,
+      "step": 8496
+    },
+    {
+      "epoch": 0.7691680999366344,
+      "grad_norm": 0.8127090334892273,
+      "learning_rate": 1.1160118468341863e-05,
+      "loss": 2.6969,
+      "step": 8497
+    },
+    {
+      "epoch": 0.7692586222503848,
+      "grad_norm": 0.9276617169380188,
+      "learning_rate": 1.1145555075601743e-05,
+      "loss": 2.6525,
+      "step": 8498
+    },
+    {
+      "epoch": 0.7693491445641351,
+      "grad_norm": 0.8718293905258179,
+      "learning_rate": 1.1131000630501797e-05,
+      "loss": 2.6579,
+      "step": 8499
+    },
+    {
+      "epoch": 0.7694396668778855,
+      "grad_norm": 0.8407883048057556,
+      "learning_rate": 1.1116455134507664e-05,
+      "loss": 2.6185,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7695301891916357,
+      "grad_norm": 0.8548136353492737,
+      "learning_rate": 1.110191858908406e-05,
+      "loss": 2.6078,
+      "step": 8501
+    },
+    {
+      "epoch": 0.769620711505386,
+      "grad_norm": 0.8696662783622742,
+      "learning_rate": 1.1087390995694813e-05,
+      "loss": 2.6446,
+      "step": 8502
+    },
+    {
+      "epoch": 0.7697112338191364,
+      "grad_norm": 0.8202139139175415,
+      "learning_rate": 1.1072872355802888e-05,
+      "loss": 2.5884,
+      "step": 8503
+    },
+    {
+      "epoch": 0.7698017561328867,
+      "grad_norm": 0.7680681347846985,
+      "learning_rate": 1.1058362670870249e-05,
+      "loss": 2.5989,
+      "step": 8504
+    },
+    {
+      "epoch": 0.7698922784466371,
+      "grad_norm": 0.8208710551261902,
+      "learning_rate": 1.104386194235808e-05,
+      "loss": 2.613,
+      "step": 8505
+    },
+    {
+      "epoch": 0.7699828007603874,
+      "grad_norm": 0.9006401300430298,
+      "learning_rate": 1.1029370171726571e-05,
+      "loss": 2.6549,
+      "step": 8506
+    },
+    {
+      "epoch": 0.7700733230741378,
+      "grad_norm": 0.8165719509124756,
+      "learning_rate": 1.1014887360435045e-05,
+      "loss": 2.6377,
+      "step": 8507
+    },
+    {
+      "epoch": 0.7701638453878881,
+      "grad_norm": 0.8723152875900269,
+      "learning_rate": 1.1000413509941931e-05,
+      "loss": 2.6185,
+      "step": 8508
+    },
+    {
+      "epoch": 0.7702543677016385,
+      "grad_norm": 0.8101234436035156,
+      "learning_rate": 1.0985948621704745e-05,
+      "loss": 2.6072,
+      "step": 8509
+    },
+    {
+      "epoch": 0.7703448900153888,
+      "grad_norm": 0.8769212961196899,
+      "learning_rate": 1.0971492697180096e-05,
+      "loss": 2.6771,
+      "step": 8510
+    },
+    {
+      "epoch": 0.7704354123291391,
+      "grad_norm": 0.9129467606544495,
+      "learning_rate": 1.0957045737823712e-05,
+      "loss": 2.6916,
+      "step": 8511
+    },
+    {
+      "epoch": 0.7705259346428894,
+      "grad_norm": 0.828231155872345,
+      "learning_rate": 1.094260774509035e-05,
+      "loss": 2.5834,
+      "step": 8512
+    },
+    {
+      "epoch": 0.7706164569566398,
+      "grad_norm": 0.9102683663368225,
+      "learning_rate": 1.092817872043399e-05,
+      "loss": 2.6544,
+      "step": 8513
+    },
+    {
+      "epoch": 0.7707069792703901,
+      "grad_norm": 0.9233009219169617,
+      "learning_rate": 1.0913758665307572e-05,
+      "loss": 2.6324,
+      "step": 8514
+    },
+    {
+      "epoch": 0.7707975015841405,
+      "grad_norm": 0.8687834143638611,
+      "learning_rate": 1.0899347581163221e-05,
+      "loss": 2.6166,
+      "step": 8515
+    },
+    {
+      "epoch": 0.7708880238978908,
+      "grad_norm": 0.8975607752799988,
+      "learning_rate": 1.0884945469452124e-05,
+      "loss": 2.6041,
+      "step": 8516
+    },
+    {
+      "epoch": 0.7709785462116412,
+      "grad_norm": 0.9307120442390442,
+      "learning_rate": 1.0870552331624594e-05,
+      "loss": 2.5894,
+      "step": 8517
+    },
+    {
+      "epoch": 0.7710690685253915,
+      "grad_norm": 0.7821042537689209,
+      "learning_rate": 1.085616816913e-05,
+      "loss": 2.5775,
+      "step": 8518
+    },
+    {
+      "epoch": 0.7711595908391419,
+      "grad_norm": 0.8686316609382629,
+      "learning_rate": 1.0841792983416842e-05,
+      "loss": 2.6168,
+      "step": 8519
+    },
+    {
+      "epoch": 0.7712501131528922,
+      "grad_norm": 0.8629463315010071,
+      "learning_rate": 1.0827426775932658e-05,
+      "loss": 2.6129,
+      "step": 8520
+    },
+    {
+      "epoch": 0.7713406354666426,
+      "grad_norm": 0.8287545442581177,
+      "learning_rate": 1.081306954812421e-05,
+      "loss": 2.6252,
+      "step": 8521
+    },
+    {
+      "epoch": 0.7714311577803928,
+      "grad_norm": 0.8263203501701355,
+      "learning_rate": 1.0798721301437187e-05,
+      "loss": 2.6089,
+      "step": 8522
+    },
+    {
+      "epoch": 0.7715216800941432,
+      "grad_norm": 0.8079046607017517,
+      "learning_rate": 1.0784382037316487e-05,
+      "loss": 2.5681,
+      "step": 8523
+    },
+    {
+      "epoch": 0.7716122024078935,
+      "grad_norm": 0.8208576440811157,
+      "learning_rate": 1.0770051757206079e-05,
+      "loss": 2.5862,
+      "step": 8524
+    },
+    {
+      "epoch": 0.7717027247216439,
+      "grad_norm": 0.8792392611503601,
+      "learning_rate": 1.0755730462549007e-05,
+      "loss": 2.6873,
+      "step": 8525
+    },
+    {
+      "epoch": 0.7717932470353942,
+      "grad_norm": 0.8938965797424316,
+      "learning_rate": 1.0741418154787442e-05,
+      "loss": 2.5957,
+      "step": 8526
+    },
+    {
+      "epoch": 0.7718837693491446,
+      "grad_norm": 0.850585401058197,
+      "learning_rate": 1.072711483536265e-05,
+      "loss": 2.5936,
+      "step": 8527
+    },
+    {
+      "epoch": 0.7719742916628949,
+      "grad_norm": 0.8643617630004883,
+      "learning_rate": 1.0712820505714916e-05,
+      "loss": 2.6053,
+      "step": 8528
+    },
+    {
+      "epoch": 0.7720648139766453,
+      "grad_norm": 0.8208019733428955,
+      "learning_rate": 1.069853516728374e-05,
+      "loss": 2.621,
+      "step": 8529
+    },
+    {
+      "epoch": 0.7721553362903956,
+      "grad_norm": 0.820443868637085,
+      "learning_rate": 1.068425882150762e-05,
+      "loss": 2.6543,
+      "step": 8530
+    },
+    {
+      "epoch": 0.772245858604146,
+      "grad_norm": 0.810172438621521,
+      "learning_rate": 1.0669991469824192e-05,
+      "loss": 2.5902,
+      "step": 8531
+    },
+    {
+      "epoch": 0.7723363809178962,
+      "grad_norm": 0.8411741256713867,
+      "learning_rate": 1.0655733113670174e-05,
+      "loss": 2.5858,
+      "step": 8532
+    },
+    {
+      "epoch": 0.7724269032316466,
+      "grad_norm": 0.8169674873352051,
+      "learning_rate": 1.0641483754481408e-05,
+      "loss": 2.6641,
+      "step": 8533
+    },
+    {
+      "epoch": 0.7725174255453969,
+      "grad_norm": 0.7835273146629333,
+      "learning_rate": 1.0627243393692743e-05,
+      "loss": 2.5762,
+      "step": 8534
+    },
+    {
+      "epoch": 0.7726079478591473,
+      "grad_norm": 0.8190977573394775,
+      "learning_rate": 1.0613012032738268e-05,
+      "loss": 2.6175,
+      "step": 8535
+    },
+    {
+      "epoch": 0.7726984701728976,
+      "grad_norm": 0.8712911605834961,
+      "learning_rate": 1.0598789673051014e-05,
+      "loss": 2.6005,
+      "step": 8536
+    },
+    {
+      "epoch": 0.772788992486648,
+      "grad_norm": 0.8913556933403015,
+      "learning_rate": 1.0584576316063188e-05,
+      "loss": 2.6676,
+      "step": 8537
+    },
+    {
+      "epoch": 0.7728795148003983,
+      "grad_norm": 0.8455060720443726,
+      "learning_rate": 1.0570371963206083e-05,
+      "loss": 2.6582,
+      "step": 8538
+    },
+    {
+      "epoch": 0.7729700371141487,
+      "grad_norm": 0.801141619682312,
+      "learning_rate": 1.0556176615910074e-05,
+      "loss": 2.6232,
+      "step": 8539
+    },
+    {
+      "epoch": 0.773060559427899,
+      "grad_norm": 0.8068770170211792,
+      "learning_rate": 1.054199027560463e-05,
+      "loss": 2.5629,
+      "step": 8540
+    },
+    {
+      "epoch": 0.7731510817416494,
+      "grad_norm": 0.7819279432296753,
+      "learning_rate": 1.052781294371834e-05,
+      "loss": 2.632,
+      "step": 8541
+    },
+    {
+      "epoch": 0.7732416040553997,
+      "grad_norm": 0.863209068775177,
+      "learning_rate": 1.051364462167881e-05,
+      "loss": 2.5913,
+      "step": 8542
+    },
+    {
+      "epoch": 0.7733321263691499,
+      "grad_norm": 0.8289285898208618,
+      "learning_rate": 1.0499485310912838e-05,
+      "loss": 2.5594,
+      "step": 8543
+    },
+    {
+      "epoch": 0.7734226486829003,
+      "grad_norm": 0.8970001935958862,
+      "learning_rate": 1.0485335012846242e-05,
+      "loss": 2.6596,
+      "step": 8544
+    },
+    {
+      "epoch": 0.7735131709966506,
+      "grad_norm": 0.8289165496826172,
+      "learning_rate": 1.047119372890395e-05,
+      "loss": 2.5846,
+      "step": 8545
+    },
+    {
+      "epoch": 0.773603693310401,
+      "grad_norm": 0.877833366394043,
+      "learning_rate": 1.045706146051001e-05,
+      "loss": 2.6525,
+      "step": 8546
+    },
+    {
+      "epoch": 0.7736942156241513,
+      "grad_norm": 0.8716258406639099,
+      "learning_rate": 1.0442938209087517e-05,
+      "loss": 2.5987,
+      "step": 8547
+    },
+    {
+      "epoch": 0.7737847379379017,
+      "grad_norm": 0.8302717804908752,
+      "learning_rate": 1.042882397605871e-05,
+      "loss": 2.5771,
+      "step": 8548
+    },
+    {
+      "epoch": 0.773875260251652,
+      "grad_norm": 0.8333648443222046,
+      "learning_rate": 1.0414718762844889e-05,
+      "loss": 2.62,
+      "step": 8549
+    },
+    {
+      "epoch": 0.7739657825654024,
+      "grad_norm": 0.8761650919914246,
+      "learning_rate": 1.0400622570866425e-05,
+      "loss": 2.6661,
+      "step": 8550
+    },
+    {
+      "epoch": 0.7740563048791527,
+      "grad_norm": 0.8990418910980225,
+      "learning_rate": 1.0386535401542808e-05,
+      "loss": 2.6772,
+      "step": 8551
+    },
+    {
+      "epoch": 0.7741468271929031,
+      "grad_norm": 0.85703045129776,
+      "learning_rate": 1.0372457256292634e-05,
+      "loss": 2.6688,
+      "step": 8552
+    },
+    {
+      "epoch": 0.7742373495066533,
+      "grad_norm": 0.8957107067108154,
+      "learning_rate": 1.0358388136533549e-05,
+      "loss": 2.5837,
+      "step": 8553
+    },
+    {
+      "epoch": 0.7743278718204037,
+      "grad_norm": 0.840560793876648,
+      "learning_rate": 1.034432804368235e-05,
+      "loss": 2.5957,
+      "step": 8554
+    },
+    {
+      "epoch": 0.774418394134154,
+      "grad_norm": 0.8776541948318481,
+      "learning_rate": 1.033027697915483e-05,
+      "loss": 2.6229,
+      "step": 8555
+    },
+    {
+      "epoch": 0.7745089164479044,
+      "grad_norm": 0.8002888560295105,
+      "learning_rate": 1.0316234944365999e-05,
+      "loss": 2.6228,
+      "step": 8556
+    },
+    {
+      "epoch": 0.7745994387616547,
+      "grad_norm": 0.896776020526886,
+      "learning_rate": 1.0302201940729838e-05,
+      "loss": 2.6482,
+      "step": 8557
+    },
+    {
+      "epoch": 0.7746899610754051,
+      "grad_norm": 0.8228409290313721,
+      "learning_rate": 1.0288177969659485e-05,
+      "loss": 2.6539,
+      "step": 8558
+    },
+    {
+      "epoch": 0.7747804833891554,
+      "grad_norm": 0.8731130361557007,
+      "learning_rate": 1.0274163032567163e-05,
+      "loss": 2.5874,
+      "step": 8559
+    },
+    {
+      "epoch": 0.7748710057029058,
+      "grad_norm": 0.8286214470863342,
+      "learning_rate": 1.026015713086418e-05,
+      "loss": 2.647,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7749615280166561,
+      "grad_norm": 0.8495842814445496,
+      "learning_rate": 1.024616026596088e-05,
+      "loss": 2.624,
+      "step": 8561
+    },
+    {
+      "epoch": 0.7750520503304065,
+      "grad_norm": 0.9225957989692688,
+      "learning_rate": 1.023217243926683e-05,
+      "loss": 2.6714,
+      "step": 8562
+    },
+    {
+      "epoch": 0.7751425726441568,
+      "grad_norm": 0.807245135307312,
+      "learning_rate": 1.0218193652190533e-05,
+      "loss": 2.6377,
+      "step": 8563
+    },
+    {
+      "epoch": 0.7752330949579072,
+      "grad_norm": 0.8943997025489807,
+      "learning_rate": 1.0204223906139686e-05,
+      "loss": 2.5683,
+      "step": 8564
+    },
+    {
+      "epoch": 0.7753236172716574,
+      "grad_norm": 0.8158625960350037,
+      "learning_rate": 1.0190263202521033e-05,
+      "loss": 2.6212,
+      "step": 8565
+    },
+    {
+      "epoch": 0.7754141395854078,
+      "grad_norm": 0.8700400590896606,
+      "learning_rate": 1.0176311542740413e-05,
+      "loss": 2.6376,
+      "step": 8566
+    },
+    {
+      "epoch": 0.7755046618991581,
+      "grad_norm": 0.8914387226104736,
+      "learning_rate": 1.016236892820276e-05,
+      "loss": 2.6217,
+      "step": 8567
+    },
+    {
+      "epoch": 0.7755951842129085,
+      "grad_norm": 0.8653732538223267,
+      "learning_rate": 1.0148435360312125e-05,
+      "loss": 2.5857,
+      "step": 8568
+    },
+    {
+      "epoch": 0.7756857065266588,
+      "grad_norm": 0.8333502411842346,
+      "learning_rate": 1.0134510840471556e-05,
+      "loss": 2.5917,
+      "step": 8569
+    },
+    {
+      "epoch": 0.7757762288404092,
+      "grad_norm": 0.9215497970581055,
+      "learning_rate": 1.0120595370083318e-05,
+      "loss": 2.6399,
+      "step": 8570
+    },
+    {
+      "epoch": 0.7758667511541595,
+      "grad_norm": 0.8121035695075989,
+      "learning_rate": 1.0106688950548649e-05,
+      "loss": 2.5964,
+      "step": 8571
+    },
+    {
+      "epoch": 0.7759572734679099,
+      "grad_norm": 0.848892867565155,
+      "learning_rate": 1.0092791583267936e-05,
+      "loss": 2.6607,
+      "step": 8572
+    },
+    {
+      "epoch": 0.7760477957816602,
+      "grad_norm": 0.8766636848449707,
+      "learning_rate": 1.0078903269640661e-05,
+      "loss": 2.6587,
+      "step": 8573
+    },
+    {
+      "epoch": 0.7761383180954106,
+      "grad_norm": 0.8555773496627808,
+      "learning_rate": 1.006502401106536e-05,
+      "loss": 2.5938,
+      "step": 8574
+    },
+    {
+      "epoch": 0.7762288404091608,
+      "grad_norm": 0.819665253162384,
+      "learning_rate": 1.0051153808939685e-05,
+      "loss": 2.6605,
+      "step": 8575
+    },
+    {
+      "epoch": 0.7763193627229112,
+      "grad_norm": 0.797521710395813,
+      "learning_rate": 1.003729266466038e-05,
+      "loss": 2.5949,
+      "step": 8576
+    },
+    {
+      "epoch": 0.7764098850366615,
+      "grad_norm": 0.8572009801864624,
+      "learning_rate": 1.00234405796232e-05,
+      "loss": 2.6371,
+      "step": 8577
+    },
+    {
+      "epoch": 0.7765004073504119,
+      "grad_norm": 0.8093275427818298,
+      "learning_rate": 1.0009597555223128e-05,
+      "loss": 2.6094,
+      "step": 8578
+    },
+    {
+      "epoch": 0.7765909296641622,
+      "grad_norm": 0.9169244170188904,
+      "learning_rate": 9.995763592854102e-06,
+      "loss": 2.6016,
+      "step": 8579
+    },
+    {
+      "epoch": 0.7766814519779126,
+      "grad_norm": 0.9339399933815002,
+      "learning_rate": 9.98193869390922e-06,
+      "loss": 2.6263,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7767719742916629,
+      "grad_norm": 0.8368563055992126,
+      "learning_rate": 9.968122859780648e-06,
+      "loss": 2.6526,
+      "step": 8581
+    },
+    {
+      "epoch": 0.7768624966054133,
+      "grad_norm": 0.9302917122840881,
+      "learning_rate": 9.95431609185965e-06,
+      "loss": 2.6532,
+      "step": 8582
+    },
+    {
+      "epoch": 0.7769530189191636,
+      "grad_norm": 0.8470931649208069,
+      "learning_rate": 9.940518391536547e-06,
+      "loss": 2.705,
+      "step": 8583
+    },
+    {
+      "epoch": 0.7770435412329139,
+      "grad_norm": 0.9310446977615356,
+      "learning_rate": 9.926729760200803e-06,
+      "loss": 2.6525,
+      "step": 8584
+    },
+    {
+      "epoch": 0.7771340635466643,
+      "grad_norm": 0.8245245218276978,
+      "learning_rate": 9.912950199240867e-06,
+      "loss": 2.5873,
+      "step": 8585
+    },
+    {
+      "epoch": 0.7772245858604145,
+      "grad_norm": 0.8588077425956726,
+      "learning_rate": 9.899179710044415e-06,
+      "loss": 2.6191,
+      "step": 8586
+    },
+    {
+      "epoch": 0.7773151081741649,
+      "grad_norm": 0.8226948380470276,
+      "learning_rate": 9.885418293998083e-06,
+      "loss": 2.5726,
+      "step": 8587
+    },
+    {
+      "epoch": 0.7774056304879152,
+      "grad_norm": 0.8949288725852966,
+      "learning_rate": 9.871665952487663e-06,
+      "loss": 2.587,
+      "step": 8588
+    },
+    {
+      "epoch": 0.7774961528016656,
+      "grad_norm": 0.8486374616622925,
+      "learning_rate": 9.857922686898024e-06,
+      "loss": 2.5791,
+      "step": 8589
+    },
+    {
+      "epoch": 0.7775866751154159,
+      "grad_norm": 0.8344716429710388,
+      "learning_rate": 9.844188498613116e-06,
+      "loss": 2.6232,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7776771974291663,
+      "grad_norm": 0.8915365934371948,
+      "learning_rate": 9.830463389015921e-06,
+      "loss": 2.6402,
+      "step": 8591
+    },
+    {
+      "epoch": 0.7777677197429166,
+      "grad_norm": 0.863557755947113,
+      "learning_rate": 9.816747359488632e-06,
+      "loss": 2.6397,
+      "step": 8592
+    },
+    {
+      "epoch": 0.777858242056667,
+      "grad_norm": 0.8383275866508484,
+      "learning_rate": 9.8030404114124e-06,
+      "loss": 2.6956,
+      "step": 8593
+    },
+    {
+      "epoch": 0.7779487643704173,
+      "grad_norm": 0.8920031189918518,
+      "learning_rate": 9.789342546167535e-06,
+      "loss": 2.6436,
+      "step": 8594
+    },
+    {
+      "epoch": 0.7780392866841677,
+      "grad_norm": 0.8083052039146423,
+      "learning_rate": 9.775653765133396e-06,
+      "loss": 2.6028,
+      "step": 8595
+    },
+    {
+      "epoch": 0.778129808997918,
+      "grad_norm": 0.8552556037902832,
+      "learning_rate": 9.761974069688461e-06,
+      "loss": 2.6034,
+      "step": 8596
+    },
+    {
+      "epoch": 0.7782203313116683,
+      "grad_norm": 0.9225404262542725,
+      "learning_rate": 9.748303461210273e-06,
+      "loss": 2.5879,
+      "step": 8597
+    },
+    {
+      "epoch": 0.7783108536254186,
+      "grad_norm": 0.8443930149078369,
+      "learning_rate": 9.734641941075474e-06,
+      "loss": 2.613,
+      "step": 8598
+    },
+    {
+      "epoch": 0.778401375939169,
+      "grad_norm": 0.8860672116279602,
+      "learning_rate": 9.720989510659717e-06,
+      "loss": 2.6089,
+      "step": 8599
+    },
+    {
+      "epoch": 0.7784918982529193,
+      "grad_norm": 0.9268253445625305,
+      "learning_rate": 9.707346171337894e-06,
+      "loss": 2.6889,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7785824205666697,
+      "grad_norm": 0.8402400016784668,
+      "learning_rate": 9.693711924483817e-06,
+      "loss": 2.6401,
+      "step": 8601
+    },
+    {
+      "epoch": 0.77867294288042,
+      "grad_norm": 0.865181028842926,
+      "learning_rate": 9.680086771470475e-06,
+      "loss": 2.6647,
+      "step": 8602
+    },
+    {
+      "epoch": 0.7787634651941704,
+      "grad_norm": 0.893332839012146,
+      "learning_rate": 9.666470713669918e-06,
+      "loss": 2.6299,
+      "step": 8603
+    },
+    {
+      "epoch": 0.7788539875079207,
+      "grad_norm": 0.7875241637229919,
+      "learning_rate": 9.6528637524533e-06,
+      "loss": 2.57,
+      "step": 8604
+    },
+    {
+      "epoch": 0.7789445098216711,
+      "grad_norm": 0.7910863161087036,
+      "learning_rate": 9.63926588919083e-06,
+      "loss": 2.6203,
+      "step": 8605
+    },
+    {
+      "epoch": 0.7790350321354214,
+      "grad_norm": 0.8562816381454468,
+      "learning_rate": 9.625677125251831e-06,
+      "loss": 2.6552,
+      "step": 8606
+    },
+    {
+      "epoch": 0.7791255544491718,
+      "grad_norm": 0.8345043659210205,
+      "learning_rate": 9.612097462004622e-06,
+      "loss": 2.6441,
+      "step": 8607
+    },
+    {
+      "epoch": 0.779216076762922,
+      "grad_norm": 0.786302924156189,
+      "learning_rate": 9.598526900816774e-06,
+      "loss": 2.6094,
+      "step": 8608
+    },
+    {
+      "epoch": 0.7793065990766724,
+      "grad_norm": 0.8758160471916199,
+      "learning_rate": 9.584965443054772e-06,
+      "loss": 2.5929,
+      "step": 8609
+    },
+    {
+      "epoch": 0.7793971213904227,
+      "grad_norm": 0.8279112577438354,
+      "learning_rate": 9.57141309008428e-06,
+      "loss": 2.6303,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7794876437041731,
+      "grad_norm": 0.845248281955719,
+      "learning_rate": 9.557869843270018e-06,
+      "loss": 2.6165,
+      "step": 8611
+    },
+    {
+      "epoch": 0.7795781660179234,
+      "grad_norm": 0.8265160322189331,
+      "learning_rate": 9.544335703975783e-06,
+      "loss": 2.6468,
+      "step": 8612
+    },
+    {
+      "epoch": 0.7796686883316738,
+      "grad_norm": 0.8631579279899597,
+      "learning_rate": 9.530810673564461e-06,
+      "loss": 2.6689,
+      "step": 8613
+    },
+    {
+      "epoch": 0.7797592106454241,
+      "grad_norm": 0.7934325933456421,
+      "learning_rate": 9.517294753398064e-06,
+      "loss": 2.6113,
+      "step": 8614
+    },
+    {
+      "epoch": 0.7798497329591745,
+      "grad_norm": 0.8415437340736389,
+      "learning_rate": 9.503787944837561e-06,
+      "loss": 2.6356,
+      "step": 8615
+    },
+    {
+      "epoch": 0.7799402552729248,
+      "grad_norm": 0.861663281917572,
+      "learning_rate": 9.490290249243172e-06,
+      "loss": 2.6316,
+      "step": 8616
+    },
+    {
+      "epoch": 0.7800307775866752,
+      "grad_norm": 0.7900444865226746,
+      "learning_rate": 9.476801667974067e-06,
+      "loss": 2.5699,
+      "step": 8617
+    },
+    {
+      "epoch": 0.7801212999004254,
+      "grad_norm": 0.8477349281311035,
+      "learning_rate": 9.463322202388546e-06,
+      "loss": 2.6082,
+      "step": 8618
+    },
+    {
+      "epoch": 0.7802118222141758,
+      "grad_norm": 0.8869505524635315,
+      "learning_rate": 9.449851853844005e-06,
+      "loss": 2.6699,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7803023445279261,
+      "grad_norm": 0.8460425138473511,
+      "learning_rate": 9.436390623696911e-06,
+      "loss": 2.5868,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7803928668416765,
+      "grad_norm": 0.8156192302703857,
+      "learning_rate": 9.422938513302804e-06,
+      "loss": 2.5906,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7804833891554268,
+      "grad_norm": 0.9294385313987732,
+      "learning_rate": 9.409495524016321e-06,
+      "loss": 2.6262,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7805739114691772,
+      "grad_norm": 0.8695874810218811,
+      "learning_rate": 9.396061657191135e-06,
+      "loss": 2.6308,
+      "step": 8623
+    },
+    {
+      "epoch": 0.7806644337829275,
+      "grad_norm": 0.8546766042709351,
+      "learning_rate": 9.382636914180087e-06,
+      "loss": 2.6711,
+      "step": 8624
+    },
+    {
+      "epoch": 0.7807549560966778,
+      "grad_norm": 0.8264554738998413,
+      "learning_rate": 9.369221296335006e-06,
+      "loss": 2.5733,
+      "step": 8625
+    },
+    {
+      "epoch": 0.7808454784104282,
+      "grad_norm": 0.8439817428588867,
+      "learning_rate": 9.355814805006858e-06,
+      "loss": 2.5912,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7809360007241785,
+      "grad_norm": 0.8114175796508789,
+      "learning_rate": 9.342417441545692e-06,
+      "loss": 2.6588,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7810265230379289,
+      "grad_norm": 0.8043602705001831,
+      "learning_rate": 9.32902920730061e-06,
+      "loss": 2.617,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7811170453516791,
+      "grad_norm": 0.8502074480056763,
+      "learning_rate": 9.315650103619778e-06,
+      "loss": 2.646,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7812075676654295,
+      "grad_norm": 0.9189439415931702,
+      "learning_rate": 9.302280131850539e-06,
+      "loss": 2.5986,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7812980899791798,
+      "grad_norm": 0.9692003726959229,
+      "learning_rate": 9.288919293339182e-06,
+      "loss": 2.5921,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7813886122929302,
+      "grad_norm": 0.86298668384552,
+      "learning_rate": 9.275567589431178e-06,
+      "loss": 2.6563,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7814791346066805,
+      "grad_norm": 0.8018830418586731,
+      "learning_rate": 9.262225021471038e-06,
+      "loss": 2.6627,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7815696569204309,
+      "grad_norm": 0.8384698033332825,
+      "learning_rate": 9.248891590802344e-06,
+      "loss": 2.5806,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7816601792341812,
+      "grad_norm": 0.8785240650177002,
+      "learning_rate": 9.23556729876781e-06,
+      "loss": 2.7003,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7817507015479316,
+      "grad_norm": 0.8848211765289307,
+      "learning_rate": 9.222252146709142e-06,
+      "loss": 2.7151,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7818412238616819,
+      "grad_norm": 0.8771471381187439,
+      "learning_rate": 9.208946135967211e-06,
+      "loss": 2.6156,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7819317461754323,
+      "grad_norm": 0.8841015100479126,
+      "learning_rate": 9.195649267881911e-06,
+      "loss": 2.6746,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7820222684891825,
+      "grad_norm": 0.822411835193634,
+      "learning_rate": 9.182361543792239e-06,
+      "loss": 2.625,
+      "step": 8639
+    },
+    {
+      "epoch": 0.782112790802933,
+      "grad_norm": 0.8606818318367004,
+      "learning_rate": 9.16908296503628e-06,
+      "loss": 2.5483,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7822033131166832,
+      "grad_norm": 0.8249856233596802,
+      "learning_rate": 9.155813532951208e-06,
+      "loss": 2.6611,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7822938354304336,
+      "grad_norm": 0.8641714453697205,
+      "learning_rate": 9.142553248873175e-06,
+      "loss": 2.5893,
+      "step": 8642
+    },
+    {
+      "epoch": 0.7823843577441839,
+      "grad_norm": 0.8504626154899597,
+      "learning_rate": 9.129302114137594e-06,
+      "loss": 2.6912,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7824748800579343,
+      "grad_norm": 0.8407044410705566,
+      "learning_rate": 9.116060130078775e-06,
+      "loss": 2.6577,
+      "step": 8644
+    },
+    {
+      "epoch": 0.7825654023716846,
+      "grad_norm": 0.8601648211479187,
+      "learning_rate": 9.102827298030226e-06,
+      "loss": 2.6203,
+      "step": 8645
+    },
+    {
+      "epoch": 0.782655924685435,
+      "grad_norm": 0.8728191256523132,
+      "learning_rate": 9.089603619324472e-06,
+      "loss": 2.6375,
+      "step": 8646
+    },
+    {
+      "epoch": 0.7827464469991853,
+      "grad_norm": 0.8072683215141296,
+      "learning_rate": 9.076389095293148e-06,
+      "loss": 2.6768,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7828369693129357,
+      "grad_norm": 0.8349645137786865,
+      "learning_rate": 9.063183727266956e-06,
+      "loss": 2.6109,
+      "step": 8648
+    },
+    {
+      "epoch": 0.782927491626686,
+      "grad_norm": 0.903693675994873,
+      "learning_rate": 9.049987516575687e-06,
+      "loss": 2.6401,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7830180139404364,
+      "grad_norm": 0.9026902914047241,
+      "learning_rate": 9.036800464548157e-06,
+      "loss": 2.5926,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7831085362541866,
+      "grad_norm": 0.8069325685501099,
+      "learning_rate": 9.023622572512369e-06,
+      "loss": 2.617,
+      "step": 8651
+    },
+    {
+      "epoch": 0.783199058567937,
+      "grad_norm": 0.8999775648117065,
+      "learning_rate": 9.010453841795274e-06,
+      "loss": 2.5992,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7832895808816873,
+      "grad_norm": 0.8389754891395569,
+      "learning_rate": 8.99729427372299e-06,
+      "loss": 2.6275,
+      "step": 8653
+    },
+    {
+      "epoch": 0.7833801031954377,
+      "grad_norm": 0.8419077396392822,
+      "learning_rate": 8.984143869620687e-06,
+      "loss": 2.6742,
+      "step": 8654
+    },
+    {
+      "epoch": 0.783470625509188,
+      "grad_norm": 0.8308020830154419,
+      "learning_rate": 8.971002630812619e-06,
+      "loss": 2.5836,
+      "step": 8655
+    },
+    {
+      "epoch": 0.7835611478229384,
+      "grad_norm": 0.8611428737640381,
+      "learning_rate": 8.957870558622061e-06,
+      "loss": 2.6362,
+      "step": 8656
+    },
+    {
+      "epoch": 0.7836516701366887,
+      "grad_norm": 0.8711309432983398,
+      "learning_rate": 8.944747654371478e-06,
+      "loss": 2.6592,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7837421924504391,
+      "grad_norm": 0.8325426578521729,
+      "learning_rate": 8.931633919382298e-06,
+      "loss": 2.6889,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7838327147641894,
+      "grad_norm": 0.8667068481445312,
+      "learning_rate": 8.918529354975091e-06,
+      "loss": 2.5751,
+      "step": 8659
+    },
+    {
+      "epoch": 0.7839232370779398,
+      "grad_norm": 0.8186068534851074,
+      "learning_rate": 8.905433962469489e-06,
+      "loss": 2.6027,
+      "step": 8660
+    },
+    {
+      "epoch": 0.78401375939169,
+      "grad_norm": 0.8946523070335388,
+      "learning_rate": 8.892347743184192e-06,
+      "loss": 2.6405,
+      "step": 8661
+    },
+    {
+      "epoch": 0.7841042817054404,
+      "grad_norm": 0.7675448060035706,
+      "learning_rate": 8.879270698436993e-06,
+      "loss": 2.6046,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7841948040191907,
+      "grad_norm": 0.841367781162262,
+      "learning_rate": 8.866202829544745e-06,
+      "loss": 2.5573,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7842853263329411,
+      "grad_norm": 0.939943253993988,
+      "learning_rate": 8.853144137823355e-06,
+      "loss": 2.697,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7843758486466914,
+      "grad_norm": 0.8685693144798279,
+      "learning_rate": 8.840094624587891e-06,
+      "loss": 2.6366,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7844663709604417,
+      "grad_norm": 0.8320167064666748,
+      "learning_rate": 8.827054291152381e-06,
+      "loss": 2.5981,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7845568932741921,
+      "grad_norm": 0.8161466717720032,
+      "learning_rate": 8.814023138830008e-06,
+      "loss": 2.5679,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7846474155879424,
+      "grad_norm": 0.8341578841209412,
+      "learning_rate": 8.80100116893301e-06,
+      "loss": 2.5952,
+      "step": 8668
+    },
+    {
+      "epoch": 0.7847379379016928,
+      "grad_norm": 0.8965204358100891,
+      "learning_rate": 8.787988382772705e-06,
+      "loss": 2.5905,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7848284602154431,
+      "grad_norm": 0.8038343191146851,
+      "learning_rate": 8.774984781659467e-06,
+      "loss": 2.5923,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7849189825291935,
+      "grad_norm": 0.8061921000480652,
+      "learning_rate": 8.761990366902783e-06,
+      "loss": 2.6416,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7850095048429437,
+      "grad_norm": 0.9051434993743896,
+      "learning_rate": 8.749005139811139e-06,
+      "loss": 2.6485,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7851000271566941,
+      "grad_norm": 0.8930591344833374,
+      "learning_rate": 8.736029101692223e-06,
+      "loss": 2.612,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7851905494704444,
+      "grad_norm": 0.9820253849029541,
+      "learning_rate": 8.723062253852654e-06,
+      "loss": 2.6247,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7852810717841948,
+      "grad_norm": 0.8085618615150452,
+      "learning_rate": 8.710104597598223e-06,
+      "loss": 2.5768,
+      "step": 8675
+    },
+    {
+      "epoch": 0.7853715940979451,
+      "grad_norm": 0.8512108325958252,
+      "learning_rate": 8.697156134233763e-06,
+      "loss": 2.6114,
+      "step": 8676
+    },
+    {
+      "epoch": 0.7854621164116955,
+      "grad_norm": 0.9089962244033813,
+      "learning_rate": 8.684216865063178e-06,
+      "loss": 2.6616,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7855526387254458,
+      "grad_norm": 0.8274604082107544,
+      "learning_rate": 8.671286791389455e-06,
+      "loss": 2.6086,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7856431610391962,
+      "grad_norm": 0.872837483882904,
+      "learning_rate": 8.658365914514665e-06,
+      "loss": 2.5432,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7857336833529465,
+      "grad_norm": 0.7837662696838379,
+      "learning_rate": 8.645454235739903e-06,
+      "loss": 2.6133,
+      "step": 8680
+    },
+    {
+      "epoch": 0.7858242056666969,
+      "grad_norm": 0.8648548126220703,
+      "learning_rate": 8.632551756365436e-06,
+      "loss": 2.632,
+      "step": 8681
+    },
+    {
+      "epoch": 0.7859147279804471,
+      "grad_norm": 0.8463937640190125,
+      "learning_rate": 8.619658477690485e-06,
+      "loss": 2.6317,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7860052502941975,
+      "grad_norm": 0.8059602975845337,
+      "learning_rate": 8.606774401013418e-06,
+      "loss": 2.6053,
+      "step": 8683
+    },
+    {
+      "epoch": 0.7860957726079478,
+      "grad_norm": 0.8308817744255066,
+      "learning_rate": 8.593899527631676e-06,
+      "loss": 2.6569,
+      "step": 8684
+    },
+    {
+      "epoch": 0.7861862949216982,
+      "grad_norm": 0.7859178185462952,
+      "learning_rate": 8.581033858841769e-06,
+      "loss": 2.6233,
+      "step": 8685
+    },
+    {
+      "epoch": 0.7862768172354485,
+      "grad_norm": 0.7964602708816528,
+      "learning_rate": 8.568177395939215e-06,
+      "loss": 2.6029,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7863673395491989,
+      "grad_norm": 0.8171819448471069,
+      "learning_rate": 8.555330140218732e-06,
+      "loss": 2.6731,
+      "step": 8687
+    },
+    {
+      "epoch": 0.7864578618629492,
+      "grad_norm": 0.8279446959495544,
+      "learning_rate": 8.542492092973987e-06,
+      "loss": 2.5903,
+      "step": 8688
+    },
+    {
+      "epoch": 0.7865483841766996,
+      "grad_norm": 0.7885525822639465,
+      "learning_rate": 8.529663255497778e-06,
+      "loss": 2.6323,
+      "step": 8689
+    },
+    {
+      "epoch": 0.7866389064904499,
+      "grad_norm": 0.8916425704956055,
+      "learning_rate": 8.516843629081984e-06,
+      "loss": 2.586,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7867294288042003,
+      "grad_norm": 0.8177568316459656,
+      "learning_rate": 8.504033215017527e-06,
+      "loss": 2.6363,
+      "step": 8691
+    },
+    {
+      "epoch": 0.7868199511179506,
+      "grad_norm": 0.8604661226272583,
+      "learning_rate": 8.491232014594418e-06,
+      "loss": 2.6184,
+      "step": 8692
+    },
+    {
+      "epoch": 0.786910473431701,
+      "grad_norm": 0.784534215927124,
+      "learning_rate": 8.478440029101764e-06,
+      "loss": 2.567,
+      "step": 8693
+    },
+    {
+      "epoch": 0.7870009957454512,
+      "grad_norm": 0.8089743852615356,
+      "learning_rate": 8.465657259827663e-06,
+      "loss": 2.608,
+      "step": 8694
+    },
+    {
+      "epoch": 0.7870915180592016,
+      "grad_norm": 0.8101866245269775,
+      "learning_rate": 8.4528837080594e-06,
+      "loss": 2.6037,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7871820403729519,
+      "grad_norm": 0.8513261675834656,
+      "learning_rate": 8.440119375083232e-06,
+      "loss": 2.6164,
+      "step": 8696
+    },
+    {
+      "epoch": 0.7872725626867023,
+      "grad_norm": 0.8743961453437805,
+      "learning_rate": 8.427364262184534e-06,
+      "loss": 2.6142,
+      "step": 8697
+    },
+    {
+      "epoch": 0.7873630850004526,
+      "grad_norm": 0.8597943186759949,
+      "learning_rate": 8.414618370647753e-06,
+      "loss": 2.6462,
+      "step": 8698
+    },
+    {
+      "epoch": 0.787453607314203,
+      "grad_norm": 0.795477569103241,
+      "learning_rate": 8.401881701756386e-06,
+      "loss": 2.5958,
+      "step": 8699
+    },
+    {
+      "epoch": 0.7875441296279533,
+      "grad_norm": 0.8336818218231201,
+      "learning_rate": 8.38915425679304e-06,
+      "loss": 2.613,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7876346519417037,
+      "grad_norm": 0.8857749104499817,
+      "learning_rate": 8.376436037039358e-06,
+      "loss": 2.6521,
+      "step": 8701
+    },
+    {
+      "epoch": 0.787725174255454,
+      "grad_norm": 0.81220942735672,
+      "learning_rate": 8.363727043776038e-06,
+      "loss": 2.6256,
+      "step": 8702
+    },
+    {
+      "epoch": 0.7878156965692044,
+      "grad_norm": 0.8130623698234558,
+      "learning_rate": 8.35102727828292e-06,
+      "loss": 2.5793,
+      "step": 8703
+    },
+    {
+      "epoch": 0.7879062188829546,
+      "grad_norm": 0.8743788599967957,
+      "learning_rate": 8.338336741838838e-06,
+      "loss": 2.6113,
+      "step": 8704
+    },
+    {
+      "epoch": 0.787996741196705,
+      "grad_norm": 0.8230921626091003,
+      "learning_rate": 8.325655435721735e-06,
+      "loss": 2.6658,
+      "step": 8705
+    },
+    {
+      "epoch": 0.7880872635104553,
+      "grad_norm": 0.8547943830490112,
+      "learning_rate": 8.312983361208615e-06,
+      "loss": 2.6389,
+      "step": 8706
+    },
+    {
+      "epoch": 0.7881777858242056,
+      "grad_norm": 0.8661034107208252,
+      "learning_rate": 8.300320519575567e-06,
+      "loss": 2.6049,
+      "step": 8707
+    },
+    {
+      "epoch": 0.788268308137956,
+      "grad_norm": 0.8751810193061829,
+      "learning_rate": 8.287666912097736e-06,
+      "loss": 2.6397,
+      "step": 8708
+    },
+    {
+      "epoch": 0.7883588304517063,
+      "grad_norm": 0.8370258212089539,
+      "learning_rate": 8.27502254004936e-06,
+      "loss": 2.6364,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7884493527654567,
+      "grad_norm": 0.8904893398284912,
+      "learning_rate": 8.262387404703653e-06,
+      "loss": 2.634,
+      "step": 8710
+    },
+    {
+      "epoch": 0.788539875079207,
+      "grad_norm": 0.8433648943901062,
+      "learning_rate": 8.249761507333076e-06,
+      "loss": 2.6143,
+      "step": 8711
+    },
+    {
+      "epoch": 0.7886303973929574,
+      "grad_norm": 0.8583589792251587,
+      "learning_rate": 8.237144849208978e-06,
+      "loss": 2.5688,
+      "step": 8712
+    },
+    {
+      "epoch": 0.7887209197067077,
+      "grad_norm": 0.8374467492103577,
+      "learning_rate": 8.224537431601886e-06,
+      "loss": 2.6051,
+      "step": 8713
+    },
+    {
+      "epoch": 0.788811442020458,
+      "grad_norm": 0.9252846240997314,
+      "learning_rate": 8.211939255781365e-06,
+      "loss": 2.6603,
+      "step": 8714
+    },
+    {
+      "epoch": 0.7889019643342083,
+      "grad_norm": 0.8124175071716309,
+      "learning_rate": 8.199350323016041e-06,
+      "loss": 2.6242,
+      "step": 8715
+    },
+    {
+      "epoch": 0.7889924866479587,
+      "grad_norm": 0.8528426885604858,
+      "learning_rate": 8.186770634573637e-06,
+      "loss": 2.6286,
+      "step": 8716
+    },
+    {
+      "epoch": 0.789083008961709,
+      "grad_norm": 0.8937164545059204,
+      "learning_rate": 8.174200191720938e-06,
+      "loss": 2.5274,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7891735312754594,
+      "grad_norm": 0.8798384666442871,
+      "learning_rate": 8.161638995723742e-06,
+      "loss": 2.6026,
+      "step": 8718
+    },
+    {
+      "epoch": 0.7892640535892097,
+      "grad_norm": 0.9604594111442566,
+      "learning_rate": 8.149087047846981e-06,
+      "loss": 2.6316,
+      "step": 8719
+    },
+    {
+      "epoch": 0.7893545759029601,
+      "grad_norm": 0.8138752579689026,
+      "learning_rate": 8.13654434935467e-06,
+      "loss": 2.5715,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7894450982167104,
+      "grad_norm": 0.8585542440414429,
+      "learning_rate": 8.124010901509793e-06,
+      "loss": 2.6234,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7895356205304608,
+      "grad_norm": 0.8834009766578674,
+      "learning_rate": 8.111486705574534e-06,
+      "loss": 2.6046,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7896261428442111,
+      "grad_norm": 0.940254807472229,
+      "learning_rate": 8.098971762810038e-06,
+      "loss": 2.6024,
+      "step": 8723
+    },
+    {
+      "epoch": 0.7897166651579615,
+      "grad_norm": 0.8321104645729065,
+      "learning_rate": 8.086466074476563e-06,
+      "loss": 2.6294,
+      "step": 8724
+    },
+    {
+      "epoch": 0.7898071874717117,
+      "grad_norm": 0.82854163646698,
+      "learning_rate": 8.073969641833445e-06,
+      "loss": 2.6374,
+      "step": 8725
+    },
+    {
+      "epoch": 0.7898977097854621,
+      "grad_norm": 0.8564167618751526,
+      "learning_rate": 8.06148246613907e-06,
+      "loss": 2.5884,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7899882320992124,
+      "grad_norm": 0.7891023755073547,
+      "learning_rate": 8.049004548650895e-06,
+      "loss": 2.6236,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7900787544129628,
+      "grad_norm": 0.8628188967704773,
+      "learning_rate": 8.036535890625463e-06,
+      "loss": 2.5974,
+      "step": 8728
+    },
+    {
+      "epoch": 0.7901692767267131,
+      "grad_norm": 0.9000905156135559,
+      "learning_rate": 8.024076493318312e-06,
+      "loss": 2.6091,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7902597990404635,
+      "grad_norm": 0.8640795946121216,
+      "learning_rate": 8.011626357984181e-06,
+      "loss": 2.6086,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7903503213542138,
+      "grad_norm": 0.8639902472496033,
+      "learning_rate": 7.999185485876737e-06,
+      "loss": 2.6104,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7904408436679642,
+      "grad_norm": 0.8826909065246582,
+      "learning_rate": 7.986753878248799e-06,
+      "loss": 2.5812,
+      "step": 8732
+    },
+    {
+      "epoch": 0.7905313659817145,
+      "grad_norm": 0.8098333477973938,
+      "learning_rate": 7.97433153635222e-06,
+      "loss": 2.6244,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7906218882954649,
+      "grad_norm": 0.8974426984786987,
+      "learning_rate": 7.961918461437946e-06,
+      "loss": 2.6268,
+      "step": 8734
+    },
+    {
+      "epoch": 0.7907124106092152,
+      "grad_norm": 0.8217973709106445,
+      "learning_rate": 7.949514654755962e-06,
+      "loss": 2.6167,
+      "step": 8735
+    },
+    {
+      "epoch": 0.7908029329229656,
+      "grad_norm": 0.8261401057243347,
+      "learning_rate": 7.937120117555363e-06,
+      "loss": 2.6662,
+      "step": 8736
+    },
+    {
+      "epoch": 0.7908934552367158,
+      "grad_norm": 0.8707113265991211,
+      "learning_rate": 7.924734851084203e-06,
+      "loss": 2.5881,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7909839775504662,
+      "grad_norm": 0.836617648601532,
+      "learning_rate": 7.912358856589751e-06,
+      "loss": 2.6031,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7910744998642165,
+      "grad_norm": 0.8788150548934937,
+      "learning_rate": 7.899992135318235e-06,
+      "loss": 2.7125,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7911650221779669,
+      "grad_norm": 0.9116982817649841,
+      "learning_rate": 7.887634688515e-06,
+      "loss": 2.549,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7912555444917172,
+      "grad_norm": 0.9610945582389832,
+      "learning_rate": 7.875286517424418e-06,
+      "loss": 2.6044,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7913460668054676,
+      "grad_norm": 0.8556106686592102,
+      "learning_rate": 7.862947623289962e-06,
+      "loss": 2.6286,
+      "step": 8742
+    },
+    {
+      "epoch": 0.7914365891192179,
+      "grad_norm": 0.8233311772346497,
+      "learning_rate": 7.850618007354172e-06,
+      "loss": 2.622,
+      "step": 8743
+    },
+    {
+      "epoch": 0.7915271114329683,
+      "grad_norm": 0.8949629068374634,
+      "learning_rate": 7.83829767085863e-06,
+      "loss": 2.5996,
+      "step": 8744
+    },
+    {
+      "epoch": 0.7916176337467186,
+      "grad_norm": 0.8206738233566284,
+      "learning_rate": 7.825986615043968e-06,
+      "loss": 2.5791,
+      "step": 8745
+    },
+    {
+      "epoch": 0.791708156060469,
+      "grad_norm": 0.8221138119697571,
+      "learning_rate": 7.81368484114996e-06,
+      "loss": 2.6101,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7917986783742192,
+      "grad_norm": 0.838756799697876,
+      "learning_rate": 7.80139235041535e-06,
+      "loss": 2.6361,
+      "step": 8747
+    },
+    {
+      "epoch": 0.7918892006879695,
+      "grad_norm": 0.8324940800666809,
+      "learning_rate": 7.789109144078e-06,
+      "loss": 2.5672,
+      "step": 8748
+    },
+    {
+      "epoch": 0.7919797230017199,
+      "grad_norm": 0.8779529333114624,
+      "learning_rate": 7.776835223374845e-06,
+      "loss": 2.6992,
+      "step": 8749
+    },
+    {
+      "epoch": 0.7920702453154702,
+      "grad_norm": 0.8620442152023315,
+      "learning_rate": 7.764570589541875e-06,
+      "loss": 2.6591,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7921607676292206,
+      "grad_norm": 0.883399486541748,
+      "learning_rate": 7.75231524381409e-06,
+      "loss": 2.6385,
+      "step": 8751
+    },
+    {
+      "epoch": 0.7922512899429709,
+      "grad_norm": 0.874740719795227,
+      "learning_rate": 7.74006918742567e-06,
+      "loss": 2.6073,
+      "step": 8752
+    },
+    {
+      "epoch": 0.7923418122567213,
+      "grad_norm": 0.8114060759544373,
+      "learning_rate": 7.727832421609738e-06,
+      "loss": 2.5893,
+      "step": 8753
+    },
+    {
+      "epoch": 0.7924323345704716,
+      "grad_norm": 0.7919555306434631,
+      "learning_rate": 7.715604947598564e-06,
+      "loss": 2.5594,
+      "step": 8754
+    },
+    {
+      "epoch": 0.792522856884222,
+      "grad_norm": 0.831693708896637,
+      "learning_rate": 7.703386766623444e-06,
+      "loss": 2.6075,
+      "step": 8755
+    },
+    {
+      "epoch": 0.7926133791979723,
+      "grad_norm": 0.803679347038269,
+      "learning_rate": 7.691177879914747e-06,
+      "loss": 2.635,
+      "step": 8756
+    },
+    {
+      "epoch": 0.7927039015117227,
+      "grad_norm": 0.8092373609542847,
+      "learning_rate": 7.67897828870191e-06,
+      "loss": 2.565,
+      "step": 8757
+    },
+    {
+      "epoch": 0.7927944238254729,
+      "grad_norm": 0.8561666011810303,
+      "learning_rate": 7.666787994213453e-06,
+      "loss": 2.6256,
+      "step": 8758
+    },
+    {
+      "epoch": 0.7928849461392233,
+      "grad_norm": 0.8929632306098938,
+      "learning_rate": 7.654606997676884e-06,
+      "loss": 2.6277,
+      "step": 8759
+    },
+    {
+      "epoch": 0.7929754684529736,
+      "grad_norm": 0.8307953476905823,
+      "learning_rate": 7.642435300318907e-06,
+      "loss": 2.6541,
+      "step": 8760
+    },
+    {
+      "epoch": 0.793065990766724,
+      "grad_norm": 0.8383509516716003,
+      "learning_rate": 7.630272903365131e-06,
+      "loss": 2.5554,
+      "step": 8761
+    },
+    {
+      "epoch": 0.7931565130804743,
+      "grad_norm": 0.8645136952400208,
+      "learning_rate": 7.618119808040369e-06,
+      "loss": 2.6302,
+      "step": 8762
+    },
+    {
+      "epoch": 0.7932470353942247,
+      "grad_norm": 0.9244964122772217,
+      "learning_rate": 7.605976015568394e-06,
+      "loss": 2.6323,
+      "step": 8763
+    },
+    {
+      "epoch": 0.793337557707975,
+      "grad_norm": 0.8274905681610107,
+      "learning_rate": 7.593841527172118e-06,
+      "loss": 2.6271,
+      "step": 8764
+    },
+    {
+      "epoch": 0.7934280800217254,
+      "grad_norm": 0.8871122598648071,
+      "learning_rate": 7.581716344073475e-06,
+      "loss": 2.5937,
+      "step": 8765
+    },
+    {
+      "epoch": 0.7935186023354757,
+      "grad_norm": 0.8496264219284058,
+      "learning_rate": 7.569600467493476e-06,
+      "loss": 2.5983,
+      "step": 8766
+    },
+    {
+      "epoch": 0.7936091246492261,
+      "grad_norm": 0.8048467636108398,
+      "learning_rate": 7.557493898652157e-06,
+      "loss": 2.6305,
+      "step": 8767
+    },
+    {
+      "epoch": 0.7936996469629763,
+      "grad_norm": 0.8628296256065369,
+      "learning_rate": 7.545396638768698e-06,
+      "loss": 2.5695,
+      "step": 8768
+    },
+    {
+      "epoch": 0.7937901692767267,
+      "grad_norm": 0.9485136866569519,
+      "learning_rate": 7.533308689061247e-06,
+      "loss": 2.6363,
+      "step": 8769
+    },
+    {
+      "epoch": 0.793880691590477,
+      "grad_norm": 0.8617843985557556,
+      "learning_rate": 7.521230050747086e-06,
+      "loss": 2.6322,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7939712139042274,
+      "grad_norm": 0.8456613421440125,
+      "learning_rate": 7.50916072504253e-06,
+      "loss": 2.5722,
+      "step": 8771
+    },
+    {
+      "epoch": 0.7940617362179777,
+      "grad_norm": 0.8934623003005981,
+      "learning_rate": 7.497100713162963e-06,
+      "loss": 2.6712,
+      "step": 8772
+    },
+    {
+      "epoch": 0.7941522585317281,
+      "grad_norm": 0.8487347960472107,
+      "learning_rate": 7.485050016322814e-06,
+      "loss": 2.6064,
+      "step": 8773
+    },
+    {
+      "epoch": 0.7942427808454784,
+      "grad_norm": 0.8945321440696716,
+      "learning_rate": 7.473008635735623e-06,
+      "loss": 2.6911,
+      "step": 8774
+    },
+    {
+      "epoch": 0.7943333031592288,
+      "grad_norm": 0.9113092422485352,
+      "learning_rate": 7.460976572613887e-06,
+      "loss": 2.6164,
+      "step": 8775
+    },
+    {
+      "epoch": 0.7944238254729791,
+      "grad_norm": 0.840405285358429,
+      "learning_rate": 7.448953828169314e-06,
+      "loss": 2.6195,
+      "step": 8776
+    },
+    {
+      "epoch": 0.7945143477867295,
+      "grad_norm": 0.8640242218971252,
+      "learning_rate": 7.436940403612547e-06,
+      "loss": 2.5575,
+      "step": 8777
+    },
+    {
+      "epoch": 0.7946048701004798,
+      "grad_norm": 0.8765424489974976,
+      "learning_rate": 7.42493630015334e-06,
+      "loss": 2.6034,
+      "step": 8778
+    },
+    {
+      "epoch": 0.7946953924142302,
+      "grad_norm": 0.8792320489883423,
+      "learning_rate": 7.412941519000527e-06,
+      "loss": 2.6519,
+      "step": 8779
+    },
+    {
+      "epoch": 0.7947859147279804,
+      "grad_norm": 0.8563889265060425,
+      "learning_rate": 7.400956061361974e-06,
+      "loss": 2.6077,
+      "step": 8780
+    },
+    {
+      "epoch": 0.7948764370417308,
+      "grad_norm": 0.9053653478622437,
+      "learning_rate": 7.388979928444584e-06,
+      "loss": 2.6558,
+      "step": 8781
+    },
+    {
+      "epoch": 0.7949669593554811,
+      "grad_norm": 0.8070489764213562,
+      "learning_rate": 7.377013121454412e-06,
+      "loss": 2.6156,
+      "step": 8782
+    },
+    {
+      "epoch": 0.7950574816692315,
+      "grad_norm": 0.8767279386520386,
+      "learning_rate": 7.365055641596474e-06,
+      "loss": 2.6594,
+      "step": 8783
+    },
+    {
+      "epoch": 0.7951480039829818,
+      "grad_norm": 0.8439067602157593,
+      "learning_rate": 7.353107490074895e-06,
+      "loss": 2.7274,
+      "step": 8784
+    },
+    {
+      "epoch": 0.7952385262967322,
+      "grad_norm": 0.8305009007453918,
+      "learning_rate": 7.341168668092857e-06,
+      "loss": 2.6025,
+      "step": 8785
+    },
+    {
+      "epoch": 0.7953290486104825,
+      "grad_norm": 0.8626893162727356,
+      "learning_rate": 7.329239176852598e-06,
+      "loss": 2.64,
+      "step": 8786
+    },
+    {
+      "epoch": 0.7954195709242329,
+      "grad_norm": 0.8474772572517395,
+      "learning_rate": 7.317319017555413e-06,
+      "loss": 2.636,
+      "step": 8787
+    },
+    {
+      "epoch": 0.7955100932379832,
+      "grad_norm": 0.9051727652549744,
+      "learning_rate": 7.305408191401697e-06,
+      "loss": 2.6179,
+      "step": 8788
+    },
+    {
+      "epoch": 0.7956006155517334,
+      "grad_norm": 0.8000675439834595,
+      "learning_rate": 7.293506699590791e-06,
+      "loss": 2.6536,
+      "step": 8789
+    },
+    {
+      "epoch": 0.7956911378654838,
+      "grad_norm": 0.8184769749641418,
+      "learning_rate": 7.281614543321269e-06,
+      "loss": 2.6171,
+      "step": 8790
+    },
+    {
+      "epoch": 0.7957816601792341,
+      "grad_norm": 0.8048648834228516,
+      "learning_rate": 7.269731723790607e-06,
+      "loss": 2.646,
+      "step": 8791
+    },
+    {
+      "epoch": 0.7958721824929845,
+      "grad_norm": 0.850342333316803,
+      "learning_rate": 7.257858242195414e-06,
+      "loss": 2.5731,
+      "step": 8792
+    },
+    {
+      "epoch": 0.7959627048067348,
+      "grad_norm": 0.8059483170509338,
+      "learning_rate": 7.245994099731368e-06,
+      "loss": 2.631,
+      "step": 8793
+    },
+    {
+      "epoch": 0.7960532271204852,
+      "grad_norm": 0.9179108738899231,
+      "learning_rate": 7.2341392975931785e-06,
+      "loss": 2.6504,
+      "step": 8794
+    },
+    {
+      "epoch": 0.7961437494342355,
+      "grad_norm": 0.8052546381950378,
+      "learning_rate": 7.222293836974614e-06,
+      "loss": 2.6354,
+      "step": 8795
+    },
+    {
+      "epoch": 0.7962342717479859,
+      "grad_norm": 0.8943376541137695,
+      "learning_rate": 7.210457719068542e-06,
+      "loss": 2.6342,
+      "step": 8796
+    },
+    {
+      "epoch": 0.7963247940617362,
+      "grad_norm": 0.9096117615699768,
+      "learning_rate": 7.198630945066809e-06,
+      "loss": 2.6045,
+      "step": 8797
+    },
+    {
+      "epoch": 0.7964153163754866,
+      "grad_norm": 0.8474493026733398,
+      "learning_rate": 7.18681351616044e-06,
+      "loss": 2.5857,
+      "step": 8798
+    },
+    {
+      "epoch": 0.7965058386892369,
+      "grad_norm": 0.9806895852088928,
+      "learning_rate": 7.175005433539383e-06,
+      "loss": 2.6353,
+      "step": 8799
+    },
+    {
+      "epoch": 0.7965963610029873,
+      "grad_norm": 0.879974365234375,
+      "learning_rate": 7.163206698392744e-06,
+      "loss": 2.6562,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7966868833167375,
+      "grad_norm": 0.8434916734695435,
+      "learning_rate": 7.151417311908648e-06,
+      "loss": 2.6339,
+      "step": 8801
+    },
+    {
+      "epoch": 0.7967774056304879,
+      "grad_norm": 0.8419424295425415,
+      "learning_rate": 7.139637275274291e-06,
+      "loss": 2.6013,
+      "step": 8802
+    },
+    {
+      "epoch": 0.7968679279442382,
+      "grad_norm": 0.8428929448127747,
+      "learning_rate": 7.1278665896759135e-06,
+      "loss": 2.6455,
+      "step": 8803
+    },
+    {
+      "epoch": 0.7969584502579886,
+      "grad_norm": 0.8516901731491089,
+      "learning_rate": 7.116105256298844e-06,
+      "loss": 2.6677,
+      "step": 8804
+    },
+    {
+      "epoch": 0.7970489725717389,
+      "grad_norm": 0.8902682662010193,
+      "learning_rate": 7.104353276327413e-06,
+      "loss": 2.669,
+      "step": 8805
+    },
+    {
+      "epoch": 0.7971394948854893,
+      "grad_norm": 0.8300895094871521,
+      "learning_rate": 7.092610650945086e-06,
+      "loss": 2.6027,
+      "step": 8806
+    },
+    {
+      "epoch": 0.7972300171992396,
+      "grad_norm": 0.8032069206237793,
+      "learning_rate": 7.0808773813343165e-06,
+      "loss": 2.5721,
+      "step": 8807
+    },
+    {
+      "epoch": 0.79732053951299,
+      "grad_norm": 0.8614211082458496,
+      "learning_rate": 7.069153468676648e-06,
+      "loss": 2.5475,
+      "step": 8808
+    },
+    {
+      "epoch": 0.7974110618267403,
+      "grad_norm": 0.893402099609375,
+      "learning_rate": 7.057438914152703e-06,
+      "loss": 2.6342,
+      "step": 8809
+    },
+    {
+      "epoch": 0.7975015841404907,
+      "grad_norm": 0.8314564228057861,
+      "learning_rate": 7.045733718942094e-06,
+      "loss": 2.5609,
+      "step": 8810
+    },
+    {
+      "epoch": 0.797592106454241,
+      "grad_norm": 0.9000319242477417,
+      "learning_rate": 7.0340378842235656e-06,
+      "loss": 2.6909,
+      "step": 8811
+    },
+    {
+      "epoch": 0.7976826287679913,
+      "grad_norm": 0.8975379467010498,
+      "learning_rate": 7.022351411174866e-06,
+      "loss": 2.6104,
+      "step": 8812
+    },
+    {
+      "epoch": 0.7977731510817416,
+      "grad_norm": 0.8559728264808655,
+      "learning_rate": 7.010674300972842e-06,
+      "loss": 2.6397,
+      "step": 8813
+    },
+    {
+      "epoch": 0.797863673395492,
+      "grad_norm": 0.8419605493545532,
+      "learning_rate": 6.999006554793374e-06,
+      "loss": 2.589,
+      "step": 8814
+    },
+    {
+      "epoch": 0.7979541957092423,
+      "grad_norm": 0.8505333662033081,
+      "learning_rate": 6.9873481738114145e-06,
+      "loss": 2.5767,
+      "step": 8815
+    },
+    {
+      "epoch": 0.7980447180229927,
+      "grad_norm": 0.7990020513534546,
+      "learning_rate": 6.97569915920091e-06,
+      "loss": 2.6203,
+      "step": 8816
+    },
+    {
+      "epoch": 0.798135240336743,
+      "grad_norm": 0.8316020965576172,
+      "learning_rate": 6.964059512135001e-06,
+      "loss": 2.6493,
+      "step": 8817
+    },
+    {
+      "epoch": 0.7982257626504934,
+      "grad_norm": 0.9669241309165955,
+      "learning_rate": 6.952429233785729e-06,
+      "loss": 2.6498,
+      "step": 8818
+    },
+    {
+      "epoch": 0.7983162849642437,
+      "grad_norm": 0.8680517673492432,
+      "learning_rate": 6.940808325324288e-06,
+      "loss": 2.6225,
+      "step": 8819
+    },
+    {
+      "epoch": 0.7984068072779941,
+      "grad_norm": 0.8824862837791443,
+      "learning_rate": 6.929196787920899e-06,
+      "loss": 2.676,
+      "step": 8820
+    },
+    {
+      "epoch": 0.7984973295917444,
+      "grad_norm": 0.900398850440979,
+      "learning_rate": 6.917594622744849e-06,
+      "loss": 2.6275,
+      "step": 8821
+    },
+    {
+      "epoch": 0.7985878519054947,
+      "grad_norm": 0.8329018354415894,
+      "learning_rate": 6.906001830964481e-06,
+      "loss": 2.659,
+      "step": 8822
+    },
+    {
+      "epoch": 0.798678374219245,
+      "grad_norm": 0.8776055574417114,
+      "learning_rate": 6.894418413747183e-06,
+      "loss": 2.623,
+      "step": 8823
+    },
+    {
+      "epoch": 0.7987688965329954,
+      "grad_norm": 0.7747718095779419,
+      "learning_rate": 6.8828443722593784e-06,
+      "loss": 2.5839,
+      "step": 8824
+    },
+    {
+      "epoch": 0.7988594188467457,
+      "grad_norm": 0.8642785549163818,
+      "learning_rate": 6.8712797076666335e-06,
+      "loss": 2.6411,
+      "step": 8825
+    },
+    {
+      "epoch": 0.7989499411604961,
+      "grad_norm": 0.962983250617981,
+      "learning_rate": 6.859724421133462e-06,
+      "loss": 2.6914,
+      "step": 8826
+    },
+    {
+      "epoch": 0.7990404634742464,
+      "grad_norm": 0.831626832485199,
+      "learning_rate": 6.848178513823478e-06,
+      "loss": 2.5524,
+      "step": 8827
+    },
+    {
+      "epoch": 0.7991309857879968,
+      "grad_norm": 0.9346736073493958,
+      "learning_rate": 6.836641986899383e-06,
+      "loss": 2.5418,
+      "step": 8828
+    },
+    {
+      "epoch": 0.7992215081017471,
+      "grad_norm": 0.8697760701179504,
+      "learning_rate": 6.825114841522884e-06,
+      "loss": 2.663,
+      "step": 8829
+    },
+    {
+      "epoch": 0.7993120304154974,
+      "grad_norm": 0.9198794364929199,
+      "learning_rate": 6.813597078854772e-06,
+      "loss": 2.5732,
+      "step": 8830
+    },
+    {
+      "epoch": 0.7994025527292478,
+      "grad_norm": 0.7944701313972473,
+      "learning_rate": 6.8020887000548985e-06,
+      "loss": 2.5865,
+      "step": 8831
+    },
+    {
+      "epoch": 0.799493075042998,
+      "grad_norm": 0.7876954078674316,
+      "learning_rate": 6.790589706282113e-06,
+      "loss": 2.6271,
+      "step": 8832
+    },
+    {
+      "epoch": 0.7995835973567484,
+      "grad_norm": 0.8922056555747986,
+      "learning_rate": 6.779100098694424e-06,
+      "loss": 2.6208,
+      "step": 8833
+    },
+    {
+      "epoch": 0.7996741196704987,
+      "grad_norm": 0.8670765161514282,
+      "learning_rate": 6.767619878448783e-06,
+      "loss": 2.5881,
+      "step": 8834
+    },
+    {
+      "epoch": 0.7997646419842491,
+      "grad_norm": 0.8833421468734741,
+      "learning_rate": 6.756149046701277e-06,
+      "loss": 2.6605,
+      "step": 8835
+    },
+    {
+      "epoch": 0.7998551642979994,
+      "grad_norm": 0.8624097108840942,
+      "learning_rate": 6.744687604607014e-06,
+      "loss": 2.5801,
+      "step": 8836
+    },
+    {
+      "epoch": 0.7999456866117498,
+      "grad_norm": 0.8142250776290894,
+      "learning_rate": 6.7332355533201495e-06,
+      "loss": 2.6091,
+      "step": 8837
+    },
+    {
+      "epoch": 0.8000362089255001,
+      "grad_norm": 0.835433304309845,
+      "learning_rate": 6.721792893993917e-06,
+      "loss": 2.6392,
+      "step": 8838
+    },
+    {
+      "epoch": 0.8001267312392505,
+      "grad_norm": 0.8677368760108948,
+      "learning_rate": 6.710359627780605e-06,
+      "loss": 2.6255,
+      "step": 8839
+    },
+    {
+      "epoch": 0.8002172535530008,
+      "grad_norm": 1.044707179069519,
+      "learning_rate": 6.698935755831492e-06,
+      "loss": 2.621,
+      "step": 8840
+    },
+    {
+      "epoch": 0.8003077758667512,
+      "grad_norm": 0.8238329291343689,
+      "learning_rate": 6.687521279297026e-06,
+      "loss": 2.5222,
+      "step": 8841
+    },
+    {
+      "epoch": 0.8003982981805015,
+      "grad_norm": 0.7824071049690247,
+      "learning_rate": 6.6761161993265985e-06,
+      "loss": 2.6161,
+      "step": 8842
+    },
+    {
+      "epoch": 0.8004888204942519,
+      "grad_norm": 0.8748955726623535,
+      "learning_rate": 6.664720517068723e-06,
+      "loss": 2.5776,
+      "step": 8843
+    },
+    {
+      "epoch": 0.8005793428080021,
+      "grad_norm": 0.8659281730651855,
+      "learning_rate": 6.653334233670927e-06,
+      "loss": 2.5945,
+      "step": 8844
+    },
+    {
+      "epoch": 0.8006698651217525,
+      "grad_norm": 0.8498925566673279,
+      "learning_rate": 6.6419573502798374e-06,
+      "loss": 2.6343,
+      "step": 8845
+    },
+    {
+      "epoch": 0.8007603874355028,
+      "grad_norm": 0.809684157371521,
+      "learning_rate": 6.63058986804106e-06,
+      "loss": 2.5778,
+      "step": 8846
+    },
+    {
+      "epoch": 0.8008509097492532,
+      "grad_norm": 0.8012062907218933,
+      "learning_rate": 6.619231788099356e-06,
+      "loss": 2.6529,
+      "step": 8847
+    },
+    {
+      "epoch": 0.8009414320630035,
+      "grad_norm": 0.8674015998840332,
+      "learning_rate": 6.607883111598445e-06,
+      "loss": 2.594,
+      "step": 8848
+    },
+    {
+      "epoch": 0.8010319543767539,
+      "grad_norm": 0.8957664966583252,
+      "learning_rate": 6.596543839681146e-06,
+      "loss": 2.6222,
+      "step": 8849
+    },
+    {
+      "epoch": 0.8011224766905042,
+      "grad_norm": 0.8477498292922974,
+      "learning_rate": 6.585213973489335e-06,
+      "loss": 2.6276,
+      "step": 8850
+    },
+    {
+      "epoch": 0.8012129990042546,
+      "grad_norm": 0.8417879939079285,
+      "learning_rate": 6.573893514163909e-06,
+      "loss": 2.6097,
+      "step": 8851
+    },
+    {
+      "epoch": 0.8013035213180049,
+      "grad_norm": 0.8202842473983765,
+      "learning_rate": 6.5625824628448575e-06,
+      "loss": 2.6,
+      "step": 8852
+    },
+    {
+      "epoch": 0.8013940436317553,
+      "grad_norm": 0.8281559944152832,
+      "learning_rate": 6.551280820671201e-06,
+      "loss": 2.5725,
+      "step": 8853
+    },
+    {
+      "epoch": 0.8014845659455055,
+      "grad_norm": 1.8357462882995605,
+      "learning_rate": 6.539988588780988e-06,
+      "loss": 2.8031,
+      "step": 8854
+    },
+    {
+      "epoch": 0.8015750882592559,
+      "grad_norm": 0.9159720540046692,
+      "learning_rate": 6.528705768311394e-06,
+      "loss": 2.6284,
+      "step": 8855
+    },
+    {
+      "epoch": 0.8016656105730062,
+      "grad_norm": 0.8227183222770691,
+      "learning_rate": 6.517432360398556e-06,
+      "loss": 2.6235,
+      "step": 8856
+    },
+    {
+      "epoch": 0.8017561328867566,
+      "grad_norm": 0.828522801399231,
+      "learning_rate": 6.506168366177723e-06,
+      "loss": 2.5962,
+      "step": 8857
+    },
+    {
+      "epoch": 0.8018466552005069,
+      "grad_norm": 0.836940586566925,
+      "learning_rate": 6.494913786783185e-06,
+      "loss": 2.6592,
+      "step": 8858
+    },
+    {
+      "epoch": 0.8019371775142573,
+      "grad_norm": 0.9068986177444458,
+      "learning_rate": 6.48366862334826e-06,
+      "loss": 2.6665,
+      "step": 8859
+    },
+    {
+      "epoch": 0.8020276998280076,
+      "grad_norm": 0.8480276465415955,
+      "learning_rate": 6.472432877005341e-06,
+      "loss": 2.6148,
+      "step": 8860
+    },
+    {
+      "epoch": 0.802118222141758,
+      "grad_norm": 0.8470519185066223,
+      "learning_rate": 6.4612065488859004e-06,
+      "loss": 2.6465,
+      "step": 8861
+    },
+    {
+      "epoch": 0.8022087444555083,
+      "grad_norm": 0.853524923324585,
+      "learning_rate": 6.449989640120368e-06,
+      "loss": 2.6106,
+      "step": 8862
+    },
+    {
+      "epoch": 0.8022992667692587,
+      "grad_norm": 0.8823791146278381,
+      "learning_rate": 6.438782151838341e-06,
+      "loss": 2.5735,
+      "step": 8863
+    },
+    {
+      "epoch": 0.802389789083009,
+      "grad_norm": 0.8680602312088013,
+      "learning_rate": 6.427584085168381e-06,
+      "loss": 2.6031,
+      "step": 8864
+    },
+    {
+      "epoch": 0.8024803113967593,
+      "grad_norm": 0.820585310459137,
+      "learning_rate": 6.416395441238143e-06,
+      "loss": 2.6024,
+      "step": 8865
+    },
+    {
+      "epoch": 0.8025708337105096,
+      "grad_norm": 0.8056566715240479,
+      "learning_rate": 6.405216221174326e-06,
+      "loss": 2.613,
+      "step": 8866
+    },
+    {
+      "epoch": 0.80266135602426,
+      "grad_norm": 0.8219767808914185,
+      "learning_rate": 6.394046426102674e-06,
+      "loss": 2.6172,
+      "step": 8867
+    },
+    {
+      "epoch": 0.8027518783380103,
+      "grad_norm": 0.8536273241043091,
+      "learning_rate": 6.382886057147985e-06,
+      "loss": 2.6725,
+      "step": 8868
+    },
+    {
+      "epoch": 0.8028424006517607,
+      "grad_norm": 0.8024052381515503,
+      "learning_rate": 6.371735115434119e-06,
+      "loss": 2.6119,
+      "step": 8869
+    },
+    {
+      "epoch": 0.802932922965511,
+      "grad_norm": 0.8551531434059143,
+      "learning_rate": 6.360593602083942e-06,
+      "loss": 2.6366,
+      "step": 8870
+    },
+    {
+      "epoch": 0.8030234452792613,
+      "grad_norm": 0.8735834956169128,
+      "learning_rate": 6.349461518219446e-06,
+      "loss": 2.6373,
+      "step": 8871
+    },
+    {
+      "epoch": 0.8031139675930117,
+      "grad_norm": 0.8546817302703857,
+      "learning_rate": 6.338338864961612e-06,
+      "loss": 2.525,
+      "step": 8872
+    },
+    {
+      "epoch": 0.803204489906762,
+      "grad_norm": 0.839828372001648,
+      "learning_rate": 6.327225643430479e-06,
+      "loss": 2.572,
+      "step": 8873
+    },
+    {
+      "epoch": 0.8032950122205124,
+      "grad_norm": 0.7989261746406555,
+      "learning_rate": 6.3161218547451716e-06,
+      "loss": 2.574,
+      "step": 8874
+    },
+    {
+      "epoch": 0.8033855345342626,
+      "grad_norm": 0.7898634672164917,
+      "learning_rate": 6.3050275000238414e-06,
+      "loss": 2.5904,
+      "step": 8875
+    },
+    {
+      "epoch": 0.803476056848013,
+      "grad_norm": 0.87517911195755,
+      "learning_rate": 6.293942580383649e-06,
+      "loss": 2.6511,
+      "step": 8876
+    },
+    {
+      "epoch": 0.8035665791617633,
+      "grad_norm": 0.8695262670516968,
+      "learning_rate": 6.282867096940903e-06,
+      "loss": 2.5497,
+      "step": 8877
+    },
+    {
+      "epoch": 0.8036571014755137,
+      "grad_norm": 0.8134534358978271,
+      "learning_rate": 6.2718010508108545e-06,
+      "loss": 2.6483,
+      "step": 8878
+    },
+    {
+      "epoch": 0.803747623789264,
+      "grad_norm": 0.8465479016304016,
+      "learning_rate": 6.26074444310788e-06,
+      "loss": 2.6658,
+      "step": 8879
+    },
+    {
+      "epoch": 0.8038381461030144,
+      "grad_norm": 0.8220434188842773,
+      "learning_rate": 6.2496972749453766e-06,
+      "loss": 2.6187,
+      "step": 8880
+    },
+    {
+      "epoch": 0.8039286684167647,
+      "grad_norm": 0.8323163986206055,
+      "learning_rate": 6.238659547435799e-06,
+      "loss": 2.669,
+      "step": 8881
+    },
+    {
+      "epoch": 0.8040191907305151,
+      "grad_norm": 0.851899266242981,
+      "learning_rate": 6.227631261690636e-06,
+      "loss": 2.6631,
+      "step": 8882
+    },
+    {
+      "epoch": 0.8041097130442654,
+      "grad_norm": 0.8882586359977722,
+      "learning_rate": 6.2166124188204444e-06,
+      "loss": 2.6435,
+      "step": 8883
+    },
+    {
+      "epoch": 0.8042002353580158,
+      "grad_norm": 0.8469085693359375,
+      "learning_rate": 6.205603019934791e-06,
+      "loss": 2.5918,
+      "step": 8884
+    },
+    {
+      "epoch": 0.804290757671766,
+      "grad_norm": 0.7756091356277466,
+      "learning_rate": 6.194603066142379e-06,
+      "loss": 2.5813,
+      "step": 8885
+    },
+    {
+      "epoch": 0.8043812799855165,
+      "grad_norm": 0.8727147579193115,
+      "learning_rate": 6.183612558550867e-06,
+      "loss": 2.6397,
+      "step": 8886
+    },
+    {
+      "epoch": 0.8044718022992667,
+      "grad_norm": 0.810825526714325,
+      "learning_rate": 6.1726314982669896e-06,
+      "loss": 2.5986,
+      "step": 8887
+    },
+    {
+      "epoch": 0.8045623246130171,
+      "grad_norm": 0.8421832323074341,
+      "learning_rate": 6.161659886396554e-06,
+      "loss": 2.626,
+      "step": 8888
+    },
+    {
+      "epoch": 0.8046528469267674,
+      "grad_norm": 0.8276161551475525,
+      "learning_rate": 6.1506977240444074e-06,
+      "loss": 2.6342,
+      "step": 8889
+    },
+    {
+      "epoch": 0.8047433692405178,
+      "grad_norm": 0.8567266464233398,
+      "learning_rate": 6.139745012314424e-06,
+      "loss": 2.653,
+      "step": 8890
+    },
+    {
+      "epoch": 0.8048338915542681,
+      "grad_norm": 0.8193711638450623,
+      "learning_rate": 6.128801752309565e-06,
+      "loss": 2.5809,
+      "step": 8891
+    },
+    {
+      "epoch": 0.8049244138680185,
+      "grad_norm": 0.8752370476722717,
+      "learning_rate": 6.1178679451317945e-06,
+      "loss": 2.6181,
+      "step": 8892
+    },
+    {
+      "epoch": 0.8050149361817688,
+      "grad_norm": 0.9479153752326965,
+      "learning_rate": 6.106943591882153e-06,
+      "loss": 2.7413,
+      "step": 8893
+    },
+    {
+      "epoch": 0.8051054584955192,
+      "grad_norm": 0.8616109490394592,
+      "learning_rate": 6.096028693660716e-06,
+      "loss": 2.5879,
+      "step": 8894
+    },
+    {
+      "epoch": 0.8051959808092695,
+      "grad_norm": 0.8333169221878052,
+      "learning_rate": 6.085123251566616e-06,
+      "loss": 2.5674,
+      "step": 8895
+    },
+    {
+      "epoch": 0.8052865031230199,
+      "grad_norm": 0.8511015176773071,
+      "learning_rate": 6.0742272666980625e-06,
+      "loss": 2.6387,
+      "step": 8896
+    },
+    {
+      "epoch": 0.8053770254367701,
+      "grad_norm": 0.778052568435669,
+      "learning_rate": 6.063340740152213e-06,
+      "loss": 2.6394,
+      "step": 8897
+    },
+    {
+      "epoch": 0.8054675477505205,
+      "grad_norm": 0.8331275582313538,
+      "learning_rate": 6.05246367302541e-06,
+      "loss": 2.6225,
+      "step": 8898
+    },
+    {
+      "epoch": 0.8055580700642708,
+      "grad_norm": 0.8217616081237793,
+      "learning_rate": 6.041596066412936e-06,
+      "loss": 2.6483,
+      "step": 8899
+    },
+    {
+      "epoch": 0.8056485923780212,
+      "grad_norm": 0.8371033668518066,
+      "learning_rate": 6.030737921409169e-06,
+      "loss": 2.5995,
+      "step": 8900
+    },
+    {
+      "epoch": 0.8057391146917715,
+      "grad_norm": 0.770965039730072,
+      "learning_rate": 6.019889239107512e-06,
+      "loss": 2.638,
+      "step": 8901
+    },
+    {
+      "epoch": 0.8058296370055219,
+      "grad_norm": 0.8548250198364258,
+      "learning_rate": 6.009050020600459e-06,
+      "loss": 2.6621,
+      "step": 8902
+    },
+    {
+      "epoch": 0.8059201593192722,
+      "grad_norm": 0.8284469842910767,
+      "learning_rate": 5.998220266979471e-06,
+      "loss": 2.6562,
+      "step": 8903
+    },
+    {
+      "epoch": 0.8060106816330226,
+      "grad_norm": 0.8728049397468567,
+      "learning_rate": 5.987399979335151e-06,
+      "loss": 2.6278,
+      "step": 8904
+    },
+    {
+      "epoch": 0.8061012039467729,
+      "grad_norm": 0.8274415731430054,
+      "learning_rate": 5.976589158757073e-06,
+      "loss": 2.6045,
+      "step": 8905
+    },
+    {
+      "epoch": 0.8061917262605233,
+      "grad_norm": 0.9814711809158325,
+      "learning_rate": 5.965787806333889e-06,
+      "loss": 2.6218,
+      "step": 8906
+    },
+    {
+      "epoch": 0.8062822485742736,
+      "grad_norm": 0.882437527179718,
+      "learning_rate": 5.954995923153306e-06,
+      "loss": 2.6554,
+      "step": 8907
+    },
+    {
+      "epoch": 0.806372770888024,
+      "grad_norm": 0.9071118235588074,
+      "learning_rate": 5.944213510302054e-06,
+      "loss": 2.5828,
+      "step": 8908
+    },
+    {
+      "epoch": 0.8064632932017742,
+      "grad_norm": 0.8062431812286377,
+      "learning_rate": 5.933440568865933e-06,
+      "loss": 2.6099,
+      "step": 8909
+    },
+    {
+      "epoch": 0.8065538155155246,
+      "grad_norm": 0.8858148455619812,
+      "learning_rate": 5.922677099929786e-06,
+      "loss": 2.5753,
+      "step": 8910
+    },
+    {
+      "epoch": 0.8066443378292749,
+      "grad_norm": 0.8028451800346375,
+      "learning_rate": 5.911923104577455e-06,
+      "loss": 2.5819,
+      "step": 8911
+    },
+    {
+      "epoch": 0.8067348601430252,
+      "grad_norm": 0.843224048614502,
+      "learning_rate": 5.901178583891909e-06,
+      "loss": 2.6249,
+      "step": 8912
+    },
+    {
+      "epoch": 0.8068253824567756,
+      "grad_norm": 0.8451189994812012,
+      "learning_rate": 5.890443538955104e-06,
+      "loss": 2.6,
+      "step": 8913
+    },
+    {
+      "epoch": 0.8069159047705259,
+      "grad_norm": 0.8016216158866882,
+      "learning_rate": 5.879717970848053e-06,
+      "loss": 2.6323,
+      "step": 8914
+    },
+    {
+      "epoch": 0.8070064270842763,
+      "grad_norm": 0.8591460585594177,
+      "learning_rate": 5.869001880650826e-06,
+      "loss": 2.6226,
+      "step": 8915
+    },
+    {
+      "epoch": 0.8070969493980266,
+      "grad_norm": 0.8800707459449768,
+      "learning_rate": 5.8582952694425374e-06,
+      "loss": 2.6134,
+      "step": 8916
+    },
+    {
+      "epoch": 0.807187471711777,
+      "grad_norm": 0.8347895741462708,
+      "learning_rate": 5.847598138301347e-06,
+      "loss": 2.6334,
+      "step": 8917
+    },
+    {
+      "epoch": 0.8072779940255272,
+      "grad_norm": 0.845406711101532,
+      "learning_rate": 5.836910488304448e-06,
+      "loss": 2.6321,
+      "step": 8918
+    },
+    {
+      "epoch": 0.8073685163392776,
+      "grad_norm": 0.8826831579208374,
+      "learning_rate": 5.826232320528069e-06,
+      "loss": 2.6547,
+      "step": 8919
+    },
+    {
+      "epoch": 0.8074590386530279,
+      "grad_norm": 0.8191585540771484,
+      "learning_rate": 5.8155636360475385e-06,
+      "loss": 2.5802,
+      "step": 8920
+    },
+    {
+      "epoch": 0.8075495609667783,
+      "grad_norm": 0.8471720218658447,
+      "learning_rate": 5.804904435937164e-06,
+      "loss": 2.6293,
+      "step": 8921
+    },
+    {
+      "epoch": 0.8076400832805286,
+      "grad_norm": 0.84287029504776,
+      "learning_rate": 5.7942547212703315e-06,
+      "loss": 2.6357,
+      "step": 8922
+    },
+    {
+      "epoch": 0.807730605594279,
+      "grad_norm": 0.8456634283065796,
+      "learning_rate": 5.783614493119471e-06,
+      "loss": 2.5488,
+      "step": 8923
+    },
+    {
+      "epoch": 0.8078211279080293,
+      "grad_norm": 0.8603562116622925,
+      "learning_rate": 5.77298375255606e-06,
+      "loss": 2.6228,
+      "step": 8924
+    },
+    {
+      "epoch": 0.8079116502217797,
+      "grad_norm": 0.8196355104446411,
+      "learning_rate": 5.7623625006505975e-06,
+      "loss": 2.6893,
+      "step": 8925
+    },
+    {
+      "epoch": 0.80800217253553,
+      "grad_norm": 0.8516672253608704,
+      "learning_rate": 5.751750738472672e-06,
+      "loss": 2.6273,
+      "step": 8926
+    },
+    {
+      "epoch": 0.8080926948492804,
+      "grad_norm": 0.8056585788726807,
+      "learning_rate": 5.74114846709084e-06,
+      "loss": 2.6184,
+      "step": 8927
+    },
+    {
+      "epoch": 0.8081832171630307,
+      "grad_norm": 0.9525690078735352,
+      "learning_rate": 5.730555687572803e-06,
+      "loss": 2.5773,
+      "step": 8928
+    },
+    {
+      "epoch": 0.808273739476781,
+      "grad_norm": 0.8798109889030457,
+      "learning_rate": 5.719972400985219e-06,
+      "loss": 2.6305,
+      "step": 8929
+    },
+    {
+      "epoch": 0.8083642617905313,
+      "grad_norm": 0.9028758406639099,
+      "learning_rate": 5.709398608393835e-06,
+      "loss": 2.6212,
+      "step": 8930
+    },
+    {
+      "epoch": 0.8084547841042817,
+      "grad_norm": 0.8423622250556946,
+      "learning_rate": 5.698834310863432e-06,
+      "loss": 2.6162,
+      "step": 8931
+    },
+    {
+      "epoch": 0.808545306418032,
+      "grad_norm": 0.8932363390922546,
+      "learning_rate": 5.688279509457828e-06,
+      "loss": 2.6524,
+      "step": 8932
+    },
+    {
+      "epoch": 0.8086358287317824,
+      "grad_norm": 0.787690281867981,
+      "learning_rate": 5.6777342052399045e-06,
+      "loss": 2.5985,
+      "step": 8933
+    },
+    {
+      "epoch": 0.8087263510455327,
+      "grad_norm": 0.8639971613883972,
+      "learning_rate": 5.667198399271567e-06,
+      "loss": 2.6242,
+      "step": 8934
+    },
+    {
+      "epoch": 0.8088168733592831,
+      "grad_norm": 0.7899857759475708,
+      "learning_rate": 5.656672092613757e-06,
+      "loss": 2.6458,
+      "step": 8935
+    },
+    {
+      "epoch": 0.8089073956730334,
+      "grad_norm": 0.8420873880386353,
+      "learning_rate": 5.646155286326504e-06,
+      "loss": 2.624,
+      "step": 8936
+    },
+    {
+      "epoch": 0.8089979179867838,
+      "grad_norm": 0.8478801250457764,
+      "learning_rate": 5.635647981468817e-06,
+      "loss": 2.6155,
+      "step": 8937
+    },
+    {
+      "epoch": 0.8090884403005341,
+      "grad_norm": 0.845515787601471,
+      "learning_rate": 5.625150179098804e-06,
+      "loss": 2.649,
+      "step": 8938
+    },
+    {
+      "epoch": 0.8091789626142845,
+      "grad_norm": 0.8689040541648865,
+      "learning_rate": 5.614661880273575e-06,
+      "loss": 2.6855,
+      "step": 8939
+    },
+    {
+      "epoch": 0.8092694849280347,
+      "grad_norm": 0.857174277305603,
+      "learning_rate": 5.604183086049342e-06,
+      "loss": 2.5821,
+      "step": 8940
+    },
+    {
+      "epoch": 0.8093600072417851,
+      "grad_norm": 0.7921069264411926,
+      "learning_rate": 5.593713797481248e-06,
+      "loss": 2.5377,
+      "step": 8941
+    },
+    {
+      "epoch": 0.8094505295555354,
+      "grad_norm": 0.8913686275482178,
+      "learning_rate": 5.583254015623618e-06,
+      "loss": 2.6787,
+      "step": 8942
+    },
+    {
+      "epoch": 0.8095410518692858,
+      "grad_norm": 0.8169851303100586,
+      "learning_rate": 5.572803741529719e-06,
+      "loss": 2.5624,
+      "step": 8943
+    },
+    {
+      "epoch": 0.8096315741830361,
+      "grad_norm": 0.8717596530914307,
+      "learning_rate": 5.562362976251901e-06,
+      "loss": 2.5997,
+      "step": 8944
+    },
+    {
+      "epoch": 0.8097220964967865,
+      "grad_norm": 0.8305097222328186,
+      "learning_rate": 5.551931720841541e-06,
+      "loss": 2.6424,
+      "step": 8945
+    },
+    {
+      "epoch": 0.8098126188105368,
+      "grad_norm": 0.8397768139839172,
+      "learning_rate": 5.54150997634908e-06,
+      "loss": 2.6499,
+      "step": 8946
+    },
+    {
+      "epoch": 0.8099031411242872,
+      "grad_norm": 0.8684827089309692,
+      "learning_rate": 5.531097743823965e-06,
+      "loss": 2.5971,
+      "step": 8947
+    },
+    {
+      "epoch": 0.8099936634380375,
+      "grad_norm": 0.8356045484542847,
+      "learning_rate": 5.520695024314748e-06,
+      "loss": 2.6024,
+      "step": 8948
+    },
+    {
+      "epoch": 0.8100841857517879,
+      "grad_norm": 0.8520987629890442,
+      "learning_rate": 5.510301818868935e-06,
+      "loss": 2.6148,
+      "step": 8949
+    },
+    {
+      "epoch": 0.8101747080655382,
+      "grad_norm": 0.8510938882827759,
+      "learning_rate": 5.499918128533155e-06,
+      "loss": 2.654,
+      "step": 8950
+    },
+    {
+      "epoch": 0.8102652303792885,
+      "grad_norm": 0.8644112944602966,
+      "learning_rate": 5.489543954353027e-06,
+      "loss": 2.5872,
+      "step": 8951
+    },
+    {
+      "epoch": 0.8103557526930388,
+      "grad_norm": 0.7893914580345154,
+      "learning_rate": 5.479179297373238e-06,
+      "loss": 2.6502,
+      "step": 8952
+    },
+    {
+      "epoch": 0.8104462750067891,
+      "grad_norm": 0.8841590881347656,
+      "learning_rate": 5.468824158637509e-06,
+      "loss": 2.6004,
+      "step": 8953
+    },
+    {
+      "epoch": 0.8105367973205395,
+      "grad_norm": 0.8805847764015198,
+      "learning_rate": 5.458478539188605e-06,
+      "loss": 2.5836,
+      "step": 8954
+    },
+    {
+      "epoch": 0.8106273196342898,
+      "grad_norm": 0.8868323564529419,
+      "learning_rate": 5.448142440068316e-06,
+      "loss": 2.6562,
+      "step": 8955
+    },
+    {
+      "epoch": 0.8107178419480402,
+      "grad_norm": 0.8586719632148743,
+      "learning_rate": 5.437815862317519e-06,
+      "loss": 2.6409,
+      "step": 8956
+    },
+    {
+      "epoch": 0.8108083642617905,
+      "grad_norm": 0.8426293134689331,
+      "learning_rate": 5.427498806976039e-06,
+      "loss": 2.584,
+      "step": 8957
+    },
+    {
+      "epoch": 0.8108988865755409,
+      "grad_norm": 0.8411415219306946,
+      "learning_rate": 5.417191275082878e-06,
+      "loss": 2.6113,
+      "step": 8958
+    },
+    {
+      "epoch": 0.8109894088892912,
+      "grad_norm": 0.8480225801467896,
+      "learning_rate": 5.40689326767595e-06,
+      "loss": 2.5914,
+      "step": 8959
+    },
+    {
+      "epoch": 0.8110799312030416,
+      "grad_norm": 0.8543084263801575,
+      "learning_rate": 5.396604785792281e-06,
+      "loss": 2.6394,
+      "step": 8960
+    },
+    {
+      "epoch": 0.8111704535167918,
+      "grad_norm": 0.8379021883010864,
+      "learning_rate": 5.38632583046792e-06,
+      "loss": 2.6218,
+      "step": 8961
+    },
+    {
+      "epoch": 0.8112609758305422,
+      "grad_norm": 0.8481666445732117,
+      "learning_rate": 5.3760564027379615e-06,
+      "loss": 2.614,
+      "step": 8962
+    },
+    {
+      "epoch": 0.8113514981442925,
+      "grad_norm": 0.8670530319213867,
+      "learning_rate": 5.3657965036365335e-06,
+      "loss": 2.6602,
+      "step": 8963
+    },
+    {
+      "epoch": 0.8114420204580429,
+      "grad_norm": 0.8096442818641663,
+      "learning_rate": 5.3555461341968205e-06,
+      "loss": 2.6109,
+      "step": 8964
+    },
+    {
+      "epoch": 0.8115325427717932,
+      "grad_norm": 0.8703358173370361,
+      "learning_rate": 5.345305295450997e-06,
+      "loss": 2.6763,
+      "step": 8965
+    },
+    {
+      "epoch": 0.8116230650855436,
+      "grad_norm": 0.8103675842285156,
+      "learning_rate": 5.335073988430372e-06,
+      "loss": 2.5879,
+      "step": 8966
+    },
+    {
+      "epoch": 0.8117135873992939,
+      "grad_norm": 0.8090990781784058,
+      "learning_rate": 5.324852214165188e-06,
+      "loss": 2.5768,
+      "step": 8967
+    },
+    {
+      "epoch": 0.8118041097130443,
+      "grad_norm": 0.8572064638137817,
+      "learning_rate": 5.314639973684787e-06,
+      "loss": 2.5946,
+      "step": 8968
+    },
+    {
+      "epoch": 0.8118946320267946,
+      "grad_norm": 0.8513726592063904,
+      "learning_rate": 5.3044372680175595e-06,
+      "loss": 2.5761,
+      "step": 8969
+    },
+    {
+      "epoch": 0.811985154340545,
+      "grad_norm": 0.8478163480758667,
+      "learning_rate": 5.294244098190926e-06,
+      "loss": 2.6156,
+      "step": 8970
+    },
+    {
+      "epoch": 0.8120756766542953,
+      "grad_norm": 0.9168113470077515,
+      "learning_rate": 5.284060465231288e-06,
+      "loss": 2.6562,
+      "step": 8971
+    },
+    {
+      "epoch": 0.8121661989680456,
+      "grad_norm": 0.8734987378120422,
+      "learning_rate": 5.273886370164194e-06,
+      "loss": 2.6128,
+      "step": 8972
+    },
+    {
+      "epoch": 0.8122567212817959,
+      "grad_norm": 0.8242517709732056,
+      "learning_rate": 5.2637218140141444e-06,
+      "loss": 2.5921,
+      "step": 8973
+    },
+    {
+      "epoch": 0.8123472435955463,
+      "grad_norm": 0.8394028544425964,
+      "learning_rate": 5.25356679780471e-06,
+      "loss": 2.5752,
+      "step": 8974
+    },
+    {
+      "epoch": 0.8124377659092966,
+      "grad_norm": 0.8377310037612915,
+      "learning_rate": 5.243421322558506e-06,
+      "loss": 2.6753,
+      "step": 8975
+    },
+    {
+      "epoch": 0.812528288223047,
+      "grad_norm": 0.8019624352455139,
+      "learning_rate": 5.233285389297182e-06,
+      "loss": 2.5654,
+      "step": 8976
+    },
+    {
+      "epoch": 0.8126188105367973,
+      "grad_norm": 0.8036606311798096,
+      "learning_rate": 5.223158999041444e-06,
+      "loss": 2.5713,
+      "step": 8977
+    },
+    {
+      "epoch": 0.8127093328505477,
+      "grad_norm": 0.8323128819465637,
+      "learning_rate": 5.213042152810988e-06,
+      "loss": 2.6211,
+      "step": 8978
+    },
+    {
+      "epoch": 0.812799855164298,
+      "grad_norm": 0.8095520734786987,
+      "learning_rate": 5.202934851624586e-06,
+      "loss": 2.5408,
+      "step": 8979
+    },
+    {
+      "epoch": 0.8128903774780484,
+      "grad_norm": 0.8063791394233704,
+      "learning_rate": 5.192837096500058e-06,
+      "loss": 2.6181,
+      "step": 8980
+    },
+    {
+      "epoch": 0.8129808997917987,
+      "grad_norm": 0.8308850526809692,
+      "learning_rate": 5.1827488884542365e-06,
+      "loss": 2.5909,
+      "step": 8981
+    },
+    {
+      "epoch": 0.8130714221055491,
+      "grad_norm": 0.804786741733551,
+      "learning_rate": 5.172670228503007e-06,
+      "loss": 2.6033,
+      "step": 8982
+    },
+    {
+      "epoch": 0.8131619444192993,
+      "grad_norm": 0.8006044030189514,
+      "learning_rate": 5.162601117661314e-06,
+      "loss": 2.5865,
+      "step": 8983
+    },
+    {
+      "epoch": 0.8132524667330497,
+      "grad_norm": 0.8728293180465698,
+      "learning_rate": 5.152541556943058e-06,
+      "loss": 2.6456,
+      "step": 8984
+    },
+    {
+      "epoch": 0.8133429890468,
+      "grad_norm": 0.8906390070915222,
+      "learning_rate": 5.1424915473612945e-06,
+      "loss": 2.5856,
+      "step": 8985
+    },
+    {
+      "epoch": 0.8134335113605504,
+      "grad_norm": 0.8838881254196167,
+      "learning_rate": 5.132451089928026e-06,
+      "loss": 2.652,
+      "step": 8986
+    },
+    {
+      "epoch": 0.8135240336743007,
+      "grad_norm": 0.7686001062393188,
+      "learning_rate": 5.1224201856543416e-06,
+      "loss": 2.5519,
+      "step": 8987
+    },
+    {
+      "epoch": 0.8136145559880511,
+      "grad_norm": 0.8629806637763977,
+      "learning_rate": 5.1123988355503475e-06,
+      "loss": 2.6648,
+      "step": 8988
+    },
+    {
+      "epoch": 0.8137050783018014,
+      "grad_norm": 0.841788649559021,
+      "learning_rate": 5.1023870406252e-06,
+      "loss": 2.6663,
+      "step": 8989
+    },
+    {
+      "epoch": 0.8137956006155518,
+      "grad_norm": 0.8157702684402466,
+      "learning_rate": 5.092384801887074e-06,
+      "loss": 2.6437,
+      "step": 8990
+    },
+    {
+      "epoch": 0.8138861229293021,
+      "grad_norm": 0.8443762063980103,
+      "learning_rate": 5.082392120343215e-06,
+      "loss": 2.5729,
+      "step": 8991
+    },
+    {
+      "epoch": 0.8139766452430525,
+      "grad_norm": 0.8237950801849365,
+      "learning_rate": 5.072408996999844e-06,
+      "loss": 2.6317,
+      "step": 8992
+    },
+    {
+      "epoch": 0.8140671675568028,
+      "grad_norm": 0.8749138712882996,
+      "learning_rate": 5.0624354328623225e-06,
+      "loss": 2.6091,
+      "step": 8993
+    },
+    {
+      "epoch": 0.814157689870553,
+      "grad_norm": 0.8353228569030762,
+      "learning_rate": 5.052471428934935e-06,
+      "loss": 2.6076,
+      "step": 8994
+    },
+    {
+      "epoch": 0.8142482121843034,
+      "grad_norm": 0.8347374796867371,
+      "learning_rate": 5.04251698622108e-06,
+      "loss": 2.611,
+      "step": 8995
+    },
+    {
+      "epoch": 0.8143387344980537,
+      "grad_norm": 0.8166795969009399,
+      "learning_rate": 5.032572105723165e-06,
+      "loss": 2.6171,
+      "step": 8996
+    },
+    {
+      "epoch": 0.8144292568118041,
+      "grad_norm": 0.846067488193512,
+      "learning_rate": 5.022636788442659e-06,
+      "loss": 2.6073,
+      "step": 8997
+    },
+    {
+      "epoch": 0.8145197791255544,
+      "grad_norm": 0.8584829568862915,
+      "learning_rate": 5.0127110353799915e-06,
+      "loss": 2.6213,
+      "step": 8998
+    },
+    {
+      "epoch": 0.8146103014393048,
+      "grad_norm": 0.8557177186012268,
+      "learning_rate": 5.002794847534764e-06,
+      "loss": 2.6213,
+      "step": 8999
+    },
+    {
+      "epoch": 0.8147008237530551,
+      "grad_norm": 0.9503716230392456,
+      "learning_rate": 4.992888225905468e-06,
+      "loss": 2.6209,
+      "step": 9000
+    },
+    {
+      "epoch": 0.8147008237530551,
+      "eval_loss": 2.5586037635803223,
+      "eval_runtime": 71.4633,
+      "eval_samples_per_second": 37.824,
+      "eval_steps_per_second": 3.162,
+      "step": 9000
+    },
+    {
+      "epoch": 0.8147913460668055,
+      "grad_norm": 0.8258231282234192,
+      "learning_rate": 4.982991171489737e-06,
+      "loss": 2.638,
+      "step": 9001
+    },
+    {
+      "epoch": 0.8148818683805558,
+      "grad_norm": 0.9333593845367432,
+      "learning_rate": 4.973103685284197e-06,
+      "loss": 2.6165,
+      "step": 9002
+    },
+    {
+      "epoch": 0.8149723906943062,
+      "grad_norm": 0.8705248236656189,
+      "learning_rate": 4.963225768284507e-06,
+      "loss": 2.6147,
+      "step": 9003
+    },
+    {
+      "epoch": 0.8150629130080564,
+      "grad_norm": 0.8360617756843567,
+      "learning_rate": 4.953357421485394e-06,
+      "loss": 2.6,
+      "step": 9004
+    },
+    {
+      "epoch": 0.8151534353218068,
+      "grad_norm": 0.8686841726303101,
+      "learning_rate": 4.943498645880595e-06,
+      "loss": 2.5953,
+      "step": 9005
+    },
+    {
+      "epoch": 0.8152439576355571,
+      "grad_norm": 0.8963146805763245,
+      "learning_rate": 4.933649442462851e-06,
+      "loss": 2.5785,
+      "step": 9006
+    },
+    {
+      "epoch": 0.8153344799493075,
+      "grad_norm": 0.853178858757019,
+      "learning_rate": 4.923809812224045e-06,
+      "loss": 2.6126,
+      "step": 9007
+    },
+    {
+      "epoch": 0.8154250022630578,
+      "grad_norm": 0.8550891876220703,
+      "learning_rate": 4.913979756154963e-06,
+      "loss": 2.6948,
+      "step": 9008
+    },
+    {
+      "epoch": 0.8155155245768082,
+      "grad_norm": 0.815525233745575,
+      "learning_rate": 4.904159275245524e-06,
+      "loss": 2.6115,
+      "step": 9009
+    },
+    {
+      "epoch": 0.8156060468905585,
+      "grad_norm": 0.8543317914009094,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 2.6508,
+      "step": 9010
+    },
+    {
+      "epoch": 0.8156965692043089,
+      "grad_norm": 0.7768651843070984,
+      "learning_rate": 4.884547042860288e-06,
+      "loss": 2.6331,
+      "step": 9011
+    },
+    {
+      "epoch": 0.8157870915180592,
+      "grad_norm": 0.805082380771637,
+      "learning_rate": 4.874755293359434e-06,
+      "loss": 2.6249,
+      "step": 9012
+    },
+    {
+      "epoch": 0.8158776138318096,
+      "grad_norm": 0.8653188943862915,
+      "learning_rate": 4.86497312296813e-06,
+      "loss": 2.6314,
+      "step": 9013
+    },
+    {
+      "epoch": 0.8159681361455599,
+      "grad_norm": 0.8116848468780518,
+      "learning_rate": 4.8552005326714e-06,
+      "loss": 2.6135,
+      "step": 9014
+    },
+    {
+      "epoch": 0.8160586584593102,
+      "grad_norm": 0.8101407885551453,
+      "learning_rate": 4.845437523453411e-06,
+      "loss": 2.5825,
+      "step": 9015
+    },
+    {
+      "epoch": 0.8161491807730605,
+      "grad_norm": 0.8034658432006836,
+      "learning_rate": 4.835684096297244e-06,
+      "loss": 2.575,
+      "step": 9016
+    },
+    {
+      "epoch": 0.8162397030868109,
+      "grad_norm": 0.9002031087875366,
+      "learning_rate": 4.82594025218508e-06,
+      "loss": 2.6026,
+      "step": 9017
+    },
+    {
+      "epoch": 0.8163302254005612,
+      "grad_norm": 0.952223002910614,
+      "learning_rate": 4.8162059920981324e-06,
+      "loss": 2.6649,
+      "step": 9018
+    },
+    {
+      "epoch": 0.8164207477143116,
+      "grad_norm": 0.9982672333717346,
+      "learning_rate": 4.806481317016631e-06,
+      "loss": 2.6041,
+      "step": 9019
+    },
+    {
+      "epoch": 0.8165112700280619,
+      "grad_norm": 0.8889972567558289,
+      "learning_rate": 4.796766227919857e-06,
+      "loss": 2.5816,
+      "step": 9020
+    },
+    {
+      "epoch": 0.8166017923418123,
+      "grad_norm": 0.8761163353919983,
+      "learning_rate": 4.7870607257861415e-06,
+      "loss": 2.6037,
+      "step": 9021
+    },
+    {
+      "epoch": 0.8166923146555626,
+      "grad_norm": 0.8455398678779602,
+      "learning_rate": 4.777364811592766e-06,
+      "loss": 2.6092,
+      "step": 9022
+    },
+    {
+      "epoch": 0.816782836969313,
+      "grad_norm": 0.9293164610862732,
+      "learning_rate": 4.767678486316185e-06,
+      "loss": 2.6139,
+      "step": 9023
+    },
+    {
+      "epoch": 0.8168733592830633,
+      "grad_norm": 0.8474072217941284,
+      "learning_rate": 4.758001750931762e-06,
+      "loss": 2.591,
+      "step": 9024
+    },
+    {
+      "epoch": 0.8169638815968137,
+      "grad_norm": 0.8196378350257874,
+      "learning_rate": 4.748334606413951e-06,
+      "loss": 2.5984,
+      "step": 9025
+    },
+    {
+      "epoch": 0.8170544039105639,
+      "grad_norm": 0.8233461976051331,
+      "learning_rate": 4.73867705373624e-06,
+      "loss": 2.5775,
+      "step": 9026
+    },
+    {
+      "epoch": 0.8171449262243143,
+      "grad_norm": 0.8707271218299866,
+      "learning_rate": 4.729029093871151e-06,
+      "loss": 2.5932,
+      "step": 9027
+    },
+    {
+      "epoch": 0.8172354485380646,
+      "grad_norm": 0.865720272064209,
+      "learning_rate": 4.719390727790218e-06,
+      "loss": 2.6235,
+      "step": 9028
+    },
+    {
+      "epoch": 0.817325970851815,
+      "grad_norm": 0.8621955513954163,
+      "learning_rate": 4.709761956464043e-06,
+      "loss": 2.6102,
+      "step": 9029
+    },
+    {
+      "epoch": 0.8174164931655653,
+      "grad_norm": 0.8045375347137451,
+      "learning_rate": 4.700142780862205e-06,
+      "loss": 2.6161,
+      "step": 9030
+    },
+    {
+      "epoch": 0.8175070154793157,
+      "grad_norm": 0.8365339040756226,
+      "learning_rate": 4.690533201953407e-06,
+      "loss": 2.6275,
+      "step": 9031
+    },
+    {
+      "epoch": 0.817597537793066,
+      "grad_norm": 0.872138261795044,
+      "learning_rate": 4.680933220705308e-06,
+      "loss": 2.5851,
+      "step": 9032
+    },
+    {
+      "epoch": 0.8176880601068164,
+      "grad_norm": 0.7883930802345276,
+      "learning_rate": 4.671342838084614e-06,
+      "loss": 2.5809,
+      "step": 9033
+    },
+    {
+      "epoch": 0.8177785824205667,
+      "grad_norm": 0.8309215307235718,
+      "learning_rate": 4.661762055057084e-06,
+      "loss": 2.6601,
+      "step": 9034
+    },
+    {
+      "epoch": 0.817869104734317,
+      "grad_norm": 0.875511109828949,
+      "learning_rate": 4.652190872587525e-06,
+      "loss": 2.5819,
+      "step": 9035
+    },
+    {
+      "epoch": 0.8179596270480674,
+      "grad_norm": 0.9092219471931458,
+      "learning_rate": 4.6426292916397105e-06,
+      "loss": 2.6716,
+      "step": 9036
+    },
+    {
+      "epoch": 0.8180501493618176,
+      "grad_norm": 0.8753784894943237,
+      "learning_rate": 4.633077313176548e-06,
+      "loss": 2.5837,
+      "step": 9037
+    },
+    {
+      "epoch": 0.818140671675568,
+      "grad_norm": 0.8911800384521484,
+      "learning_rate": 4.623534938159868e-06,
+      "loss": 2.5886,
+      "step": 9038
+    },
+    {
+      "epoch": 0.8182311939893183,
+      "grad_norm": 0.8206343054771423,
+      "learning_rate": 4.614002167550613e-06,
+      "loss": 2.6194,
+      "step": 9039
+    },
+    {
+      "epoch": 0.8183217163030687,
+      "grad_norm": 0.9463024735450745,
+      "learning_rate": 4.604479002308737e-06,
+      "loss": 2.5975,
+      "step": 9040
+    },
+    {
+      "epoch": 0.818412238616819,
+      "grad_norm": 0.8719370365142822,
+      "learning_rate": 4.594965443393206e-06,
+      "loss": 2.6995,
+      "step": 9041
+    },
+    {
+      "epoch": 0.8185027609305694,
+      "grad_norm": 0.8398632407188416,
+      "learning_rate": 4.585461491762044e-06,
+      "loss": 2.5941,
+      "step": 9042
+    },
+    {
+      "epoch": 0.8185932832443197,
+      "grad_norm": 0.8209207653999329,
+      "learning_rate": 4.575967148372317e-06,
+      "loss": 2.6073,
+      "step": 9043
+    },
+    {
+      "epoch": 0.8186838055580701,
+      "grad_norm": 0.9303069710731506,
+      "learning_rate": 4.5664824141800624e-06,
+      "loss": 2.6029,
+      "step": 9044
+    },
+    {
+      "epoch": 0.8187743278718204,
+      "grad_norm": 0.8226931095123291,
+      "learning_rate": 4.557007290140447e-06,
+      "loss": 2.697,
+      "step": 9045
+    },
+    {
+      "epoch": 0.8188648501855708,
+      "grad_norm": 0.9183313250541687,
+      "learning_rate": 4.547541777207565e-06,
+      "loss": 2.6763,
+      "step": 9046
+    },
+    {
+      "epoch": 0.818955372499321,
+      "grad_norm": 0.8149845004081726,
+      "learning_rate": 4.538085876334619e-06,
+      "loss": 2.6456,
+      "step": 9047
+    },
+    {
+      "epoch": 0.8190458948130714,
+      "grad_norm": 0.8685399293899536,
+      "learning_rate": 4.528639588473804e-06,
+      "loss": 2.5925,
+      "step": 9048
+    },
+    {
+      "epoch": 0.8191364171268217,
+      "grad_norm": 0.830268144607544,
+      "learning_rate": 4.519202914576381e-06,
+      "loss": 2.5512,
+      "step": 9049
+    },
+    {
+      "epoch": 0.8192269394405721,
+      "grad_norm": 0.8395816683769226,
+      "learning_rate": 4.509775855592613e-06,
+      "loss": 2.6046,
+      "step": 9050
+    },
+    {
+      "epoch": 0.8193174617543224,
+      "grad_norm": 0.8282546401023865,
+      "learning_rate": 4.5003584124718055e-06,
+      "loss": 2.6571,
+      "step": 9051
+    },
+    {
+      "epoch": 0.8194079840680728,
+      "grad_norm": 0.8863653540611267,
+      "learning_rate": 4.490950586162279e-06,
+      "loss": 2.6142,
+      "step": 9052
+    },
+    {
+      "epoch": 0.8194985063818231,
+      "grad_norm": 0.8952196836471558,
+      "learning_rate": 4.481552377611431e-06,
+      "loss": 2.6454,
+      "step": 9053
+    },
+    {
+      "epoch": 0.8195890286955735,
+      "grad_norm": 0.845206081867218,
+      "learning_rate": 4.4721637877656375e-06,
+      "loss": 2.6916,
+      "step": 9054
+    },
+    {
+      "epoch": 0.8196795510093238,
+      "grad_norm": 0.8556833863258362,
+      "learning_rate": 4.462784817570331e-06,
+      "loss": 2.6503,
+      "step": 9055
+    },
+    {
+      "epoch": 0.8197700733230742,
+      "grad_norm": 0.7939481735229492,
+      "learning_rate": 4.45341546796999e-06,
+      "loss": 2.6606,
+      "step": 9056
+    },
+    {
+      "epoch": 0.8198605956368245,
+      "grad_norm": 0.8105207085609436,
+      "learning_rate": 4.444055739908082e-06,
+      "loss": 2.6122,
+      "step": 9057
+    },
+    {
+      "epoch": 0.8199511179505748,
+      "grad_norm": 0.8216555714607239,
+      "learning_rate": 4.434705634327163e-06,
+      "loss": 2.5746,
+      "step": 9058
+    },
+    {
+      "epoch": 0.8200416402643251,
+      "grad_norm": 0.9625558853149414,
+      "learning_rate": 4.42536515216877e-06,
+      "loss": 2.6698,
+      "step": 9059
+    },
+    {
+      "epoch": 0.8201321625780755,
+      "grad_norm": 0.9395487308502197,
+      "learning_rate": 4.416034294373472e-06,
+      "loss": 2.6948,
+      "step": 9060
+    },
+    {
+      "epoch": 0.8202226848918258,
+      "grad_norm": 0.8587431907653809,
+      "learning_rate": 4.406713061880941e-06,
+      "loss": 2.5942,
+      "step": 9061
+    },
+    {
+      "epoch": 0.8203132072055762,
+      "grad_norm": 0.9051746129989624,
+      "learning_rate": 4.3974014556297685e-06,
+      "loss": 2.6551,
+      "step": 9062
+    },
+    {
+      "epoch": 0.8204037295193265,
+      "grad_norm": 0.8491767644882202,
+      "learning_rate": 4.388099476557639e-06,
+      "loss": 2.5966,
+      "step": 9063
+    },
+    {
+      "epoch": 0.8204942518330769,
+      "grad_norm": 0.877068042755127,
+      "learning_rate": 4.378807125601303e-06,
+      "loss": 2.6041,
+      "step": 9064
+    },
+    {
+      "epoch": 0.8205847741468272,
+      "grad_norm": 0.8690078258514404,
+      "learning_rate": 4.369524403696457e-06,
+      "loss": 2.6046,
+      "step": 9065
+    },
+    {
+      "epoch": 0.8206752964605776,
+      "grad_norm": 0.8469387292861938,
+      "learning_rate": 4.360251311777896e-06,
+      "loss": 2.6946,
+      "step": 9066
+    },
+    {
+      "epoch": 0.8207658187743279,
+      "grad_norm": 0.8528283834457397,
+      "learning_rate": 4.3509878507793975e-06,
+      "loss": 2.5908,
+      "step": 9067
+    },
+    {
+      "epoch": 0.8208563410880783,
+      "grad_norm": 0.8180561065673828,
+      "learning_rate": 4.341734021633814e-06,
+      "loss": 2.6254,
+      "step": 9068
+    },
+    {
+      "epoch": 0.8209468634018285,
+      "grad_norm": 0.8620453476905823,
+      "learning_rate": 4.33248982527299e-06,
+      "loss": 2.6312,
+      "step": 9069
+    },
+    {
+      "epoch": 0.8210373857155789,
+      "grad_norm": 0.8564789295196533,
+      "learning_rate": 4.323255262627846e-06,
+      "loss": 2.5788,
+      "step": 9070
+    },
+    {
+      "epoch": 0.8211279080293292,
+      "grad_norm": 0.8219857215881348,
+      "learning_rate": 4.314030334628249e-06,
+      "loss": 2.6346,
+      "step": 9071
+    },
+    {
+      "epoch": 0.8212184303430796,
+      "grad_norm": 0.8396607637405396,
+      "learning_rate": 4.304815042203203e-06,
+      "loss": 2.6268,
+      "step": 9072
+    },
+    {
+      "epoch": 0.8213089526568299,
+      "grad_norm": 0.8367118239402771,
+      "learning_rate": 4.2956093862806525e-06,
+      "loss": 2.5875,
+      "step": 9073
+    },
+    {
+      "epoch": 0.8213994749705803,
+      "grad_norm": 0.8491257429122925,
+      "learning_rate": 4.286413367787612e-06,
+      "loss": 2.653,
+      "step": 9074
+    },
+    {
+      "epoch": 0.8214899972843306,
+      "grad_norm": 0.8212398886680603,
+      "learning_rate": 4.277226987650129e-06,
+      "loss": 2.597,
+      "step": 9075
+    },
+    {
+      "epoch": 0.8215805195980809,
+      "grad_norm": 0.857515811920166,
+      "learning_rate": 4.268050246793276e-06,
+      "loss": 2.6104,
+      "step": 9076
+    },
+    {
+      "epoch": 0.8216710419118313,
+      "grad_norm": 0.8828648328781128,
+      "learning_rate": 4.258883146141135e-06,
+      "loss": 2.6992,
+      "step": 9077
+    },
+    {
+      "epoch": 0.8217615642255816,
+      "grad_norm": 0.8497927188873291,
+      "learning_rate": 4.249725686616867e-06,
+      "loss": 2.6917,
+      "step": 9078
+    },
+    {
+      "epoch": 0.821852086539332,
+      "grad_norm": 0.921812117099762,
+      "learning_rate": 4.24057786914257e-06,
+      "loss": 2.6864,
+      "step": 9079
+    },
+    {
+      "epoch": 0.8219426088530822,
+      "grad_norm": 0.901273787021637,
+      "learning_rate": 4.231439694639483e-06,
+      "loss": 2.6079,
+      "step": 9080
+    },
+    {
+      "epoch": 0.8220331311668326,
+      "grad_norm": 0.863243043422699,
+      "learning_rate": 4.2223111640277815e-06,
+      "loss": 2.596,
+      "step": 9081
+    },
+    {
+      "epoch": 0.8221236534805829,
+      "grad_norm": 0.8987923860549927,
+      "learning_rate": 4.2131922782267405e-06,
+      "loss": 2.6031,
+      "step": 9082
+    },
+    {
+      "epoch": 0.8222141757943333,
+      "grad_norm": 0.8319920897483826,
+      "learning_rate": 4.2040830381546045e-06,
+      "loss": 2.6171,
+      "step": 9083
+    },
+    {
+      "epoch": 0.8223046981080836,
+      "grad_norm": 0.8061295747756958,
+      "learning_rate": 4.194983444728684e-06,
+      "loss": 2.5421,
+      "step": 9084
+    },
+    {
+      "epoch": 0.822395220421834,
+      "grad_norm": 0.8673340082168579,
+      "learning_rate": 4.185893498865323e-06,
+      "loss": 2.6519,
+      "step": 9085
+    },
+    {
+      "epoch": 0.8224857427355843,
+      "grad_norm": 0.8772735595703125,
+      "learning_rate": 4.176813201479867e-06,
+      "loss": 2.6119,
+      "step": 9086
+    },
+    {
+      "epoch": 0.8225762650493347,
+      "grad_norm": 0.9388923048973083,
+      "learning_rate": 4.167742553486675e-06,
+      "loss": 2.6075,
+      "step": 9087
+    },
+    {
+      "epoch": 0.822666787363085,
+      "grad_norm": 0.8103634715080261,
+      "learning_rate": 4.158681555799204e-06,
+      "loss": 2.5887,
+      "step": 9088
+    },
+    {
+      "epoch": 0.8227573096768354,
+      "grad_norm": 0.8761494159698486,
+      "learning_rate": 4.14963020932988e-06,
+      "loss": 2.6375,
+      "step": 9089
+    },
+    {
+      "epoch": 0.8228478319905856,
+      "grad_norm": 0.8322147130966187,
+      "learning_rate": 4.140588514990162e-06,
+      "loss": 2.6086,
+      "step": 9090
+    },
+    {
+      "epoch": 0.822938354304336,
+      "grad_norm": 0.823414146900177,
+      "learning_rate": 4.131556473690557e-06,
+      "loss": 2.6468,
+      "step": 9091
+    },
+    {
+      "epoch": 0.8230288766180863,
+      "grad_norm": 0.8451811075210571,
+      "learning_rate": 4.122534086340612e-06,
+      "loss": 2.6127,
+      "step": 9092
+    },
+    {
+      "epoch": 0.8231193989318367,
+      "grad_norm": 0.7874065637588501,
+      "learning_rate": 4.113521353848826e-06,
+      "loss": 2.6115,
+      "step": 9093
+    },
+    {
+      "epoch": 0.823209921245587,
+      "grad_norm": 0.9588502645492554,
+      "learning_rate": 4.104518277122848e-06,
+      "loss": 2.6193,
+      "step": 9094
+    },
+    {
+      "epoch": 0.8233004435593374,
+      "grad_norm": 0.8688252568244934,
+      "learning_rate": 4.095524857069244e-06,
+      "loss": 2.6243,
+      "step": 9095
+    },
+    {
+      "epoch": 0.8233909658730877,
+      "grad_norm": 0.8532291650772095,
+      "learning_rate": 4.086541094593666e-06,
+      "loss": 2.5685,
+      "step": 9096
+    },
+    {
+      "epoch": 0.8234814881868381,
+      "grad_norm": 0.866753876209259,
+      "learning_rate": 4.077566990600767e-06,
+      "loss": 2.5997,
+      "step": 9097
+    },
+    {
+      "epoch": 0.8235720105005884,
+      "grad_norm": 0.8232748508453369,
+      "learning_rate": 4.068602545994249e-06,
+      "loss": 2.623,
+      "step": 9098
+    },
+    {
+      "epoch": 0.8236625328143388,
+      "grad_norm": 0.8832443952560425,
+      "learning_rate": 4.059647761676833e-06,
+      "loss": 2.6458,
+      "step": 9099
+    },
+    {
+      "epoch": 0.823753055128089,
+      "grad_norm": 0.845489501953125,
+      "learning_rate": 4.050702638550275e-06,
+      "loss": 2.5738,
+      "step": 9100
+    },
+    {
+      "epoch": 0.8238435774418394,
+      "grad_norm": 0.8758887052536011,
+      "learning_rate": 4.04176717751531e-06,
+      "loss": 2.6469,
+      "step": 9101
+    },
+    {
+      "epoch": 0.8239340997555897,
+      "grad_norm": 0.8528258800506592,
+      "learning_rate": 4.0328413794717855e-06,
+      "loss": 2.6441,
+      "step": 9102
+    },
+    {
+      "epoch": 0.8240246220693401,
+      "grad_norm": 0.8793032765388489,
+      "learning_rate": 4.023925245318494e-06,
+      "loss": 2.64,
+      "step": 9103
+    },
+    {
+      "epoch": 0.8241151443830904,
+      "grad_norm": 0.8211391568183899,
+      "learning_rate": 4.015018775953317e-06,
+      "loss": 2.608,
+      "step": 9104
+    },
+    {
+      "epoch": 0.8242056666968408,
+      "grad_norm": 0.842514157295227,
+      "learning_rate": 4.006121972273113e-06,
+      "loss": 2.6211,
+      "step": 9105
+    },
+    {
+      "epoch": 0.8242961890105911,
+      "grad_norm": 0.7817758917808533,
+      "learning_rate": 3.997234835173802e-06,
+      "loss": 2.5894,
+      "step": 9106
+    },
+    {
+      "epoch": 0.8243867113243415,
+      "grad_norm": 0.8630059361457825,
+      "learning_rate": 3.9883573655503106e-06,
+      "loss": 2.609,
+      "step": 9107
+    },
+    {
+      "epoch": 0.8244772336380918,
+      "grad_norm": 0.871557354927063,
+      "learning_rate": 3.979489564296624e-06,
+      "loss": 2.6673,
+      "step": 9108
+    },
+    {
+      "epoch": 0.8245677559518422,
+      "grad_norm": 0.9041063189506531,
+      "learning_rate": 3.970631432305694e-06,
+      "loss": 2.627,
+      "step": 9109
+    },
+    {
+      "epoch": 0.8246582782655925,
+      "grad_norm": 0.9307136535644531,
+      "learning_rate": 3.961782970469563e-06,
+      "loss": 2.6341,
+      "step": 9110
+    },
+    {
+      "epoch": 0.8247488005793429,
+      "grad_norm": 0.8669264316558838,
+      "learning_rate": 3.952944179679252e-06,
+      "loss": 2.5751,
+      "step": 9111
+    },
+    {
+      "epoch": 0.8248393228930931,
+      "grad_norm": 0.8764011263847351,
+      "learning_rate": 3.944115060824826e-06,
+      "loss": 2.6495,
+      "step": 9112
+    },
+    {
+      "epoch": 0.8249298452068435,
+      "grad_norm": 0.8889691829681396,
+      "learning_rate": 3.935295614795398e-06,
+      "loss": 2.6362,
+      "step": 9113
+    },
+    {
+      "epoch": 0.8250203675205938,
+      "grad_norm": 0.8333461284637451,
+      "learning_rate": 3.926485842479066e-06,
+      "loss": 2.5991,
+      "step": 9114
+    },
+    {
+      "epoch": 0.8251108898343442,
+      "grad_norm": 0.796550989151001,
+      "learning_rate": 3.917685744762989e-06,
+      "loss": 2.6287,
+      "step": 9115
+    },
+    {
+      "epoch": 0.8252014121480945,
+      "grad_norm": 0.8666393756866455,
+      "learning_rate": 3.908895322533335e-06,
+      "loss": 2.6685,
+      "step": 9116
+    },
+    {
+      "epoch": 0.8252919344618448,
+      "grad_norm": 0.8646209836006165,
+      "learning_rate": 3.900114576675262e-06,
+      "loss": 2.6127,
+      "step": 9117
+    },
+    {
+      "epoch": 0.8253824567755952,
+      "grad_norm": 0.8975469470024109,
+      "learning_rate": 3.891343508073053e-06,
+      "loss": 2.5959,
+      "step": 9118
+    },
+    {
+      "epoch": 0.8254729790893455,
+      "grad_norm": 0.8263643980026245,
+      "learning_rate": 3.882582117609912e-06,
+      "loss": 2.6019,
+      "step": 9119
+    },
+    {
+      "epoch": 0.8255635014030959,
+      "grad_norm": 0.8185175061225891,
+      "learning_rate": 3.873830406168111e-06,
+      "loss": 2.6134,
+      "step": 9120
+    },
+    {
+      "epoch": 0.8256540237168462,
+      "grad_norm": 0.8060017824172974,
+      "learning_rate": 3.865088374628967e-06,
+      "loss": 2.5985,
+      "step": 9121
+    },
+    {
+      "epoch": 0.8257445460305965,
+      "grad_norm": 0.812060534954071,
+      "learning_rate": 3.856356023872798e-06,
+      "loss": 2.58,
+      "step": 9122
+    },
+    {
+      "epoch": 0.8258350683443468,
+      "grad_norm": 0.8128660917282104,
+      "learning_rate": 3.847633354778935e-06,
+      "loss": 2.612,
+      "step": 9123
+    },
+    {
+      "epoch": 0.8259255906580972,
+      "grad_norm": 0.8915673494338989,
+      "learning_rate": 3.838920368225784e-06,
+      "loss": 2.5464,
+      "step": 9124
+    },
+    {
+      "epoch": 0.8260161129718475,
+      "grad_norm": 0.802297830581665,
+      "learning_rate": 3.830217065090702e-06,
+      "loss": 2.5796,
+      "step": 9125
+    },
+    {
+      "epoch": 0.8261066352855979,
+      "grad_norm": 0.8497213125228882,
+      "learning_rate": 3.821523446250142e-06,
+      "loss": 2.7114,
+      "step": 9126
+    },
+    {
+      "epoch": 0.8261971575993482,
+      "grad_norm": 0.8551464676856995,
+      "learning_rate": 3.8128395125795246e-06,
+      "loss": 2.6046,
+      "step": 9127
+    },
+    {
+      "epoch": 0.8262876799130986,
+      "grad_norm": 0.8229303359985352,
+      "learning_rate": 3.8041652649533412e-06,
+      "loss": 2.6217,
+      "step": 9128
+    },
+    {
+      "epoch": 0.8263782022268489,
+      "grad_norm": 0.8738320469856262,
+      "learning_rate": 3.795500704245092e-06,
+      "loss": 2.6069,
+      "step": 9129
+    },
+    {
+      "epoch": 0.8264687245405993,
+      "grad_norm": 0.815537691116333,
+      "learning_rate": 3.7868458313272904e-06,
+      "loss": 2.6014,
+      "step": 9130
+    },
+    {
+      "epoch": 0.8265592468543496,
+      "grad_norm": 0.8061794638633728,
+      "learning_rate": 3.7782006470714616e-06,
+      "loss": 2.6608,
+      "step": 9131
+    },
+    {
+      "epoch": 0.8266497691681,
+      "grad_norm": 0.8272706866264343,
+      "learning_rate": 3.76956515234822e-06,
+      "loss": 2.65,
+      "step": 9132
+    },
+    {
+      "epoch": 0.8267402914818502,
+      "grad_norm": 0.8943510055541992,
+      "learning_rate": 3.760939348027115e-06,
+      "loss": 2.6195,
+      "step": 9133
+    },
+    {
+      "epoch": 0.8268308137956006,
+      "grad_norm": 0.8860402703285217,
+      "learning_rate": 3.7523232349767957e-06,
+      "loss": 2.5961,
+      "step": 9134
+    },
+    {
+      "epoch": 0.8269213361093509,
+      "grad_norm": 0.8261328935623169,
+      "learning_rate": 3.7437168140648904e-06,
+      "loss": 2.57,
+      "step": 9135
+    },
+    {
+      "epoch": 0.8270118584231013,
+      "grad_norm": 0.8775939345359802,
+      "learning_rate": 3.7351200861580617e-06,
+      "loss": 2.6482,
+      "step": 9136
+    },
+    {
+      "epoch": 0.8271023807368516,
+      "grad_norm": 0.8148878216743469,
+      "learning_rate": 3.7265330521220056e-06,
+      "loss": 2.6488,
+      "step": 9137
+    },
+    {
+      "epoch": 0.827192903050602,
+      "grad_norm": 0.791273295879364,
+      "learning_rate": 3.7179557128214526e-06,
+      "loss": 2.6304,
+      "step": 9138
+    },
+    {
+      "epoch": 0.8272834253643523,
+      "grad_norm": 0.8352766633033752,
+      "learning_rate": 3.709388069120101e-06,
+      "loss": 2.6171,
+      "step": 9139
+    },
+    {
+      "epoch": 0.8273739476781027,
+      "grad_norm": 0.8089897036552429,
+      "learning_rate": 3.7008301218807716e-06,
+      "loss": 2.5845,
+      "step": 9140
+    },
+    {
+      "epoch": 0.827464469991853,
+      "grad_norm": 0.8363292217254639,
+      "learning_rate": 3.692281871965186e-06,
+      "loss": 2.6664,
+      "step": 9141
+    },
+    {
+      "epoch": 0.8275549923056034,
+      "grad_norm": 0.8899134993553162,
+      "learning_rate": 3.68374332023419e-06,
+      "loss": 2.6082,
+      "step": 9142
+    },
+    {
+      "epoch": 0.8276455146193537,
+      "grad_norm": 0.8281762003898621,
+      "learning_rate": 3.675214467547594e-06,
+      "loss": 2.6403,
+      "step": 9143
+    },
+    {
+      "epoch": 0.827736036933104,
+      "grad_norm": 0.795846700668335,
+      "learning_rate": 3.6666953147642792e-06,
+      "loss": 2.6201,
+      "step": 9144
+    },
+    {
+      "epoch": 0.8278265592468543,
+      "grad_norm": 0.8525198698043823,
+      "learning_rate": 3.6581858627421027e-06,
+      "loss": 2.6095,
+      "step": 9145
+    },
+    {
+      "epoch": 0.8279170815606047,
+      "grad_norm": 0.8217182159423828,
+      "learning_rate": 3.649686112337991e-06,
+      "loss": 2.6264,
+      "step": 9146
+    },
+    {
+      "epoch": 0.828007603874355,
+      "grad_norm": 0.8377102613449097,
+      "learning_rate": 3.6411960644078367e-06,
+      "loss": 2.6227,
+      "step": 9147
+    },
+    {
+      "epoch": 0.8280981261881054,
+      "grad_norm": 0.8462782502174377,
+      "learning_rate": 3.632715719806601e-06,
+      "loss": 2.6041,
+      "step": 9148
+    },
+    {
+      "epoch": 0.8281886485018557,
+      "grad_norm": 0.8169267773628235,
+      "learning_rate": 3.6242450793882555e-06,
+      "loss": 2.6211,
+      "step": 9149
+    },
+    {
+      "epoch": 0.8282791708156061,
+      "grad_norm": 0.8321042060852051,
+      "learning_rate": 3.615784144005796e-06,
+      "loss": 2.6057,
+      "step": 9150
+    },
+    {
+      "epoch": 0.8283696931293564,
+      "grad_norm": 0.8683964014053345,
+      "learning_rate": 3.607332914511241e-06,
+      "loss": 2.5884,
+      "step": 9151
+    },
+    {
+      "epoch": 0.8284602154431068,
+      "grad_norm": 0.8136768937110901,
+      "learning_rate": 3.59889139175561e-06,
+      "loss": 2.5525,
+      "step": 9152
+    },
+    {
+      "epoch": 0.8285507377568571,
+      "grad_norm": 0.8591272234916687,
+      "learning_rate": 3.5904595765890005e-06,
+      "loss": 2.6401,
+      "step": 9153
+    },
+    {
+      "epoch": 0.8286412600706075,
+      "grad_norm": 0.7958996295928955,
+      "learning_rate": 3.5820374698604555e-06,
+      "loss": 2.6246,
+      "step": 9154
+    },
+    {
+      "epoch": 0.8287317823843577,
+      "grad_norm": 0.8030251860618591,
+      "learning_rate": 3.5736250724180966e-06,
+      "loss": 2.6206,
+      "step": 9155
+    },
+    {
+      "epoch": 0.8288223046981081,
+      "grad_norm": 0.904325544834137,
+      "learning_rate": 3.565222385109068e-06,
+      "loss": 2.5609,
+      "step": 9156
+    },
+    {
+      "epoch": 0.8289128270118584,
+      "grad_norm": 0.8910101056098938,
+      "learning_rate": 3.5568294087795053e-06,
+      "loss": 2.6318,
+      "step": 9157
+    },
+    {
+      "epoch": 0.8290033493256087,
+      "grad_norm": 0.8450800776481628,
+      "learning_rate": 3.5484461442745643e-06,
+      "loss": 2.621,
+      "step": 9158
+    },
+    {
+      "epoch": 0.8290938716393591,
+      "grad_norm": 0.8555982708930969,
+      "learning_rate": 3.5400725924384813e-06,
+      "loss": 2.6358,
+      "step": 9159
+    },
+    {
+      "epoch": 0.8291843939531094,
+      "grad_norm": 0.8615543842315674,
+      "learning_rate": 3.5317087541144377e-06,
+      "loss": 2.63,
+      "step": 9160
+    },
+    {
+      "epoch": 0.8292749162668598,
+      "grad_norm": 0.83702552318573,
+      "learning_rate": 3.5233546301446817e-06,
+      "loss": 2.5729,
+      "step": 9161
+    },
+    {
+      "epoch": 0.8293654385806101,
+      "grad_norm": 0.8201742768287659,
+      "learning_rate": 3.515010221370474e-06,
+      "loss": 2.5653,
+      "step": 9162
+    },
+    {
+      "epoch": 0.8294559608943605,
+      "grad_norm": 0.8219901323318481,
+      "learning_rate": 3.5066755286320975e-06,
+      "loss": 2.5807,
+      "step": 9163
+    },
+    {
+      "epoch": 0.8295464832081108,
+      "grad_norm": 0.8451256155967712,
+      "learning_rate": 3.4983505527688586e-06,
+      "loss": 2.6275,
+      "step": 9164
+    },
+    {
+      "epoch": 0.8296370055218611,
+      "grad_norm": 0.8181697130203247,
+      "learning_rate": 3.4900352946190874e-06,
+      "loss": 2.5696,
+      "step": 9165
+    },
+    {
+      "epoch": 0.8297275278356114,
+      "grad_norm": 0.8448930382728577,
+      "learning_rate": 3.4817297550200913e-06,
+      "loss": 2.6381,
+      "step": 9166
+    },
+    {
+      "epoch": 0.8298180501493618,
+      "grad_norm": 0.9296396374702454,
+      "learning_rate": 3.4734339348083012e-06,
+      "loss": 2.6578,
+      "step": 9167
+    },
+    {
+      "epoch": 0.8299085724631121,
+      "grad_norm": 0.8755658268928528,
+      "learning_rate": 3.4651478348190603e-06,
+      "loss": 2.6097,
+      "step": 9168
+    },
+    {
+      "epoch": 0.8299990947768625,
+      "grad_norm": 0.8138020038604736,
+      "learning_rate": 3.45687145588679e-06,
+      "loss": 2.6322,
+      "step": 9169
+    },
+    {
+      "epoch": 0.8300896170906128,
+      "grad_norm": 0.821542501449585,
+      "learning_rate": 3.448604798844912e-06,
+      "loss": 2.602,
+      "step": 9170
+    },
+    {
+      "epoch": 0.8301801394043632,
+      "grad_norm": 0.8439465165138245,
+      "learning_rate": 3.4403478645259057e-06,
+      "loss": 2.6411,
+      "step": 9171
+    },
+    {
+      "epoch": 0.8302706617181135,
+      "grad_norm": 0.8149715662002563,
+      "learning_rate": 3.4321006537612165e-06,
+      "loss": 2.5726,
+      "step": 9172
+    },
+    {
+      "epoch": 0.8303611840318639,
+      "grad_norm": 0.8409690856933594,
+      "learning_rate": 3.4238631673813582e-06,
+      "loss": 2.6176,
+      "step": 9173
+    },
+    {
+      "epoch": 0.8304517063456142,
+      "grad_norm": 0.8181410431861877,
+      "learning_rate": 3.4156354062158226e-06,
+      "loss": 2.595,
+      "step": 9174
+    },
+    {
+      "epoch": 0.8305422286593646,
+      "grad_norm": 0.8303077220916748,
+      "learning_rate": 3.40741737109318e-06,
+      "loss": 2.561,
+      "step": 9175
+    },
+    {
+      "epoch": 0.8306327509731148,
+      "grad_norm": 0.8174563050270081,
+      "learning_rate": 3.3992090628409466e-06,
+      "loss": 2.6311,
+      "step": 9176
+    },
+    {
+      "epoch": 0.8307232732868652,
+      "grad_norm": 0.8620786666870117,
+      "learning_rate": 3.391010482285728e-06,
+      "loss": 2.6014,
+      "step": 9177
+    },
+    {
+      "epoch": 0.8308137956006155,
+      "grad_norm": 0.9524500370025635,
+      "learning_rate": 3.3828216302531078e-06,
+      "loss": 2.6493,
+      "step": 9178
+    },
+    {
+      "epoch": 0.8309043179143659,
+      "grad_norm": 0.8139269948005676,
+      "learning_rate": 3.3746425075676933e-06,
+      "loss": 2.6455,
+      "step": 9179
+    },
+    {
+      "epoch": 0.8309948402281162,
+      "grad_norm": 0.8140713572502136,
+      "learning_rate": 3.3664731150531482e-06,
+      "loss": 2.5223,
+      "step": 9180
+    },
+    {
+      "epoch": 0.8310853625418666,
+      "grad_norm": 0.8157796859741211,
+      "learning_rate": 3.358313453532125e-06,
+      "loss": 2.6053,
+      "step": 9181
+    },
+    {
+      "epoch": 0.8311758848556169,
+      "grad_norm": 0.8004946112632751,
+      "learning_rate": 3.3501635238262663e-06,
+      "loss": 2.6019,
+      "step": 9182
+    },
+    {
+      "epoch": 0.8312664071693673,
+      "grad_norm": 0.8402578830718994,
+      "learning_rate": 3.3420233267563272e-06,
+      "loss": 2.5942,
+      "step": 9183
+    },
+    {
+      "epoch": 0.8313569294831176,
+      "grad_norm": 0.8298675417900085,
+      "learning_rate": 3.333892863141974e-06,
+      "loss": 2.6567,
+      "step": 9184
+    },
+    {
+      "epoch": 0.831447451796868,
+      "grad_norm": 0.8852568864822388,
+      "learning_rate": 3.325772133801963e-06,
+      "loss": 2.6207,
+      "step": 9185
+    },
+    {
+      "epoch": 0.8315379741106182,
+      "grad_norm": 0.8615192770957947,
+      "learning_rate": 3.3176611395540626e-06,
+      "loss": 2.5879,
+      "step": 9186
+    },
+    {
+      "epoch": 0.8316284964243686,
+      "grad_norm": 0.9066726565361023,
+      "learning_rate": 3.3095598812150406e-06,
+      "loss": 2.634,
+      "step": 9187
+    },
+    {
+      "epoch": 0.8317190187381189,
+      "grad_norm": 0.8623664975166321,
+      "learning_rate": 3.301468359600679e-06,
+      "loss": 2.6659,
+      "step": 9188
+    },
+    {
+      "epoch": 0.8318095410518693,
+      "grad_norm": 0.7866799831390381,
+      "learning_rate": 3.293386575525825e-06,
+      "loss": 2.6064,
+      "step": 9189
+    },
+    {
+      "epoch": 0.8319000633656196,
+      "grad_norm": 0.8399327993392944,
+      "learning_rate": 3.2853145298042953e-06,
+      "loss": 2.6452,
+      "step": 9190
+    },
+    {
+      "epoch": 0.83199058567937,
+      "grad_norm": 0.8688633441925049,
+      "learning_rate": 3.2772522232489387e-06,
+      "loss": 2.687,
+      "step": 9191
+    },
+    {
+      "epoch": 0.8320811079931203,
+      "grad_norm": 0.8381837606430054,
+      "learning_rate": 3.269199656671629e-06,
+      "loss": 2.6317,
+      "step": 9192
+    },
+    {
+      "epoch": 0.8321716303068707,
+      "grad_norm": 0.8940417170524597,
+      "learning_rate": 3.261156830883283e-06,
+      "loss": 2.6023,
+      "step": 9193
+    },
+    {
+      "epoch": 0.832262152620621,
+      "grad_norm": 0.8516998291015625,
+      "learning_rate": 3.253123746693787e-06,
+      "loss": 2.6418,
+      "step": 9194
+    },
+    {
+      "epoch": 0.8323526749343714,
+      "grad_norm": 0.8648186922073364,
+      "learning_rate": 3.245100404912094e-06,
+      "loss": 2.6774,
+      "step": 9195
+    },
+    {
+      "epoch": 0.8324431972481217,
+      "grad_norm": 0.9397299289703369,
+      "learning_rate": 3.2370868063461236e-06,
+      "loss": 2.6327,
+      "step": 9196
+    },
+    {
+      "epoch": 0.832533719561872,
+      "grad_norm": 0.8416707515716553,
+      "learning_rate": 3.2290829518028862e-06,
+      "loss": 2.6227,
+      "step": 9197
+    },
+    {
+      "epoch": 0.8326242418756223,
+      "grad_norm": 0.9227898716926575,
+      "learning_rate": 3.221088842088338e-06,
+      "loss": 2.6949,
+      "step": 9198
+    },
+    {
+      "epoch": 0.8327147641893726,
+      "grad_norm": 0.8900390863418579,
+      "learning_rate": 3.213104478007489e-06,
+      "loss": 2.5867,
+      "step": 9199
+    },
+    {
+      "epoch": 0.832805286503123,
+      "grad_norm": 0.8820727467536926,
+      "learning_rate": 3.2051298603643753e-06,
+      "loss": 2.5983,
+      "step": 9200
+    },
+    {
+      "epoch": 0.8328958088168733,
+      "grad_norm": 0.8546134829521179,
+      "learning_rate": 3.1971649899620317e-06,
+      "loss": 2.5818,
+      "step": 9201
+    },
+    {
+      "epoch": 0.8329863311306237,
+      "grad_norm": 0.8514634966850281,
+      "learning_rate": 3.1892098676025274e-06,
+      "loss": 2.6578,
+      "step": 9202
+    },
+    {
+      "epoch": 0.833076853444374,
+      "grad_norm": 0.9180089235305786,
+      "learning_rate": 3.1812644940869552e-06,
+      "loss": 2.5797,
+      "step": 9203
+    },
+    {
+      "epoch": 0.8331673757581244,
+      "grad_norm": 0.8001142144203186,
+      "learning_rate": 3.173328870215375e-06,
+      "loss": 2.574,
+      "step": 9204
+    },
+    {
+      "epoch": 0.8332578980718747,
+      "grad_norm": 0.9282077550888062,
+      "learning_rate": 3.165402996786948e-06,
+      "loss": 2.5877,
+      "step": 9205
+    },
+    {
+      "epoch": 0.8333484203856251,
+      "grad_norm": 0.8573855757713318,
+      "learning_rate": 3.157486874599791e-06,
+      "loss": 2.607,
+      "step": 9206
+    },
+    {
+      "epoch": 0.8334389426993754,
+      "grad_norm": 0.8349403142929077,
+      "learning_rate": 3.149580504451044e-06,
+      "loss": 2.6184,
+      "step": 9207
+    },
+    {
+      "epoch": 0.8335294650131257,
+      "grad_norm": 0.8084670305252075,
+      "learning_rate": 3.1416838871368924e-06,
+      "loss": 2.623,
+      "step": 9208
+    },
+    {
+      "epoch": 0.833619987326876,
+      "grad_norm": 0.8632099032402039,
+      "learning_rate": 3.1337970234525226e-06,
+      "loss": 2.6386,
+      "step": 9209
+    },
+    {
+      "epoch": 0.8337105096406264,
+      "grad_norm": 0.8786114454269409,
+      "learning_rate": 3.1259199141921435e-06,
+      "loss": 2.5755,
+      "step": 9210
+    },
+    {
+      "epoch": 0.8338010319543767,
+      "grad_norm": 0.8364348411560059,
+      "learning_rate": 3.1180525601489873e-06,
+      "loss": 2.6017,
+      "step": 9211
+    },
+    {
+      "epoch": 0.8338915542681271,
+      "grad_norm": 0.8469467163085938,
+      "learning_rate": 3.1101949621152647e-06,
+      "loss": 2.6735,
+      "step": 9212
+    },
+    {
+      "epoch": 0.8339820765818774,
+      "grad_norm": 0.8727800846099854,
+      "learning_rate": 3.1023471208822765e-06,
+      "loss": 2.5964,
+      "step": 9213
+    },
+    {
+      "epoch": 0.8340725988956278,
+      "grad_norm": 0.9073167443275452,
+      "learning_rate": 3.0945090372402785e-06,
+      "loss": 2.6447,
+      "step": 9214
+    },
+    {
+      "epoch": 0.8341631212093781,
+      "grad_norm": 0.8227028250694275,
+      "learning_rate": 3.0866807119785734e-06,
+      "loss": 2.6402,
+      "step": 9215
+    },
+    {
+      "epoch": 0.8342536435231285,
+      "grad_norm": 0.8529277443885803,
+      "learning_rate": 3.0788621458854527e-06,
+      "loss": 2.6093,
+      "step": 9216
+    },
+    {
+      "epoch": 0.8343441658368788,
+      "grad_norm": 0.8378677368164062,
+      "learning_rate": 3.0710533397482866e-06,
+      "loss": 2.654,
+      "step": 9217
+    },
+    {
+      "epoch": 0.8344346881506292,
+      "grad_norm": 0.8516073226928711,
+      "learning_rate": 3.063254294353368e-06,
+      "loss": 2.5382,
+      "step": 9218
+    },
+    {
+      "epoch": 0.8345252104643794,
+      "grad_norm": 0.831871509552002,
+      "learning_rate": 3.0554650104861136e-06,
+      "loss": 2.6137,
+      "step": 9219
+    },
+    {
+      "epoch": 0.8346157327781298,
+      "grad_norm": 0.8696051836013794,
+      "learning_rate": 3.047685488930874e-06,
+      "loss": 2.5878,
+      "step": 9220
+    },
+    {
+      "epoch": 0.8347062550918801,
+      "grad_norm": 0.8664183020591736,
+      "learning_rate": 3.0399157304710545e-06,
+      "loss": 2.6619,
+      "step": 9221
+    },
+    {
+      "epoch": 0.8347967774056305,
+      "grad_norm": 0.7911030650138855,
+      "learning_rate": 3.032155735889053e-06,
+      "loss": 2.5723,
+      "step": 9222
+    },
+    {
+      "epoch": 0.8348872997193808,
+      "grad_norm": 0.9079735279083252,
+      "learning_rate": 3.024405505966332e-06,
+      "loss": 2.7033,
+      "step": 9223
+    },
+    {
+      "epoch": 0.8349778220331312,
+      "grad_norm": 0.8091928958892822,
+      "learning_rate": 3.016665041483313e-06,
+      "loss": 2.5664,
+      "step": 9224
+    },
+    {
+      "epoch": 0.8350683443468815,
+      "grad_norm": 0.9017247557640076,
+      "learning_rate": 3.0089343432194826e-06,
+      "loss": 2.6433,
+      "step": 9225
+    },
+    {
+      "epoch": 0.8351588666606319,
+      "grad_norm": 0.8352522253990173,
+      "learning_rate": 3.0012134119532964e-06,
+      "loss": 2.6499,
+      "step": 9226
+    },
+    {
+      "epoch": 0.8352493889743822,
+      "grad_norm": 0.8612746596336365,
+      "learning_rate": 2.9935022484622766e-06,
+      "loss": 2.6483,
+      "step": 9227
+    },
+    {
+      "epoch": 0.8353399112881326,
+      "grad_norm": 0.9562451243400574,
+      "learning_rate": 2.985800853522924e-06,
+      "loss": 2.6012,
+      "step": 9228
+    },
+    {
+      "epoch": 0.8354304336018828,
+      "grad_norm": 0.8161827325820923,
+      "learning_rate": 2.9781092279107637e-06,
+      "loss": 2.5909,
+      "step": 9229
+    },
+    {
+      "epoch": 0.8355209559156332,
+      "grad_norm": 0.9282371401786804,
+      "learning_rate": 2.970427372400353e-06,
+      "loss": 2.5909,
+      "step": 9230
+    },
+    {
+      "epoch": 0.8356114782293835,
+      "grad_norm": 0.8111222386360168,
+      "learning_rate": 2.962755287765251e-06,
+      "loss": 2.6167,
+      "step": 9231
+    },
+    {
+      "epoch": 0.8357020005431339,
+      "grad_norm": 0.7850275635719299,
+      "learning_rate": 2.955092974778051e-06,
+      "loss": 2.5852,
+      "step": 9232
+    },
+    {
+      "epoch": 0.8357925228568842,
+      "grad_norm": 0.8214910626411438,
+      "learning_rate": 2.9474404342103135e-06,
+      "loss": 2.643,
+      "step": 9233
+    },
+    {
+      "epoch": 0.8358830451706346,
+      "grad_norm": 0.8709685802459717,
+      "learning_rate": 2.939797666832678e-06,
+      "loss": 2.5581,
+      "step": 9234
+    },
+    {
+      "epoch": 0.8359735674843849,
+      "grad_norm": 0.8599022030830383,
+      "learning_rate": 2.9321646734147502e-06,
+      "loss": 2.6064,
+      "step": 9235
+    },
+    {
+      "epoch": 0.8360640897981353,
+      "grad_norm": 0.9196804761886597,
+      "learning_rate": 2.9245414547251935e-06,
+      "loss": 2.7081,
+      "step": 9236
+    },
+    {
+      "epoch": 0.8361546121118856,
+      "grad_norm": 0.8670064210891724,
+      "learning_rate": 2.91692801153165e-06,
+      "loss": 2.5799,
+      "step": 9237
+    },
+    {
+      "epoch": 0.836245134425636,
+      "grad_norm": 0.8151206970214844,
+      "learning_rate": 2.9093243446008166e-06,
+      "loss": 2.6471,
+      "step": 9238
+    },
+    {
+      "epoch": 0.8363356567393863,
+      "grad_norm": 0.7814372181892395,
+      "learning_rate": 2.901730454698337e-06,
+      "loss": 2.6122,
+      "step": 9239
+    },
+    {
+      "epoch": 0.8364261790531367,
+      "grad_norm": 0.8052501678466797,
+      "learning_rate": 2.894146342588977e-06,
+      "loss": 2.5885,
+      "step": 9240
+    },
+    {
+      "epoch": 0.8365167013668869,
+      "grad_norm": 0.830434262752533,
+      "learning_rate": 2.8865720090364034e-06,
+      "loss": 2.657,
+      "step": 9241
+    },
+    {
+      "epoch": 0.8366072236806372,
+      "grad_norm": 0.855962336063385,
+      "learning_rate": 2.879007454803373e-06,
+      "loss": 2.61,
+      "step": 9242
+    },
+    {
+      "epoch": 0.8366977459943876,
+      "grad_norm": 0.8213224411010742,
+      "learning_rate": 2.8714526806516316e-06,
+      "loss": 2.6103,
+      "step": 9243
+    },
+    {
+      "epoch": 0.8367882683081379,
+      "grad_norm": 0.8698161840438843,
+      "learning_rate": 2.863907687341949e-06,
+      "loss": 2.5893,
+      "step": 9244
+    },
+    {
+      "epoch": 0.8368787906218883,
+      "grad_norm": 0.8918118476867676,
+      "learning_rate": 2.856372475634106e-06,
+      "loss": 2.595,
+      "step": 9245
+    },
+    {
+      "epoch": 0.8369693129356386,
+      "grad_norm": 0.8260166645050049,
+      "learning_rate": 2.8488470462869065e-06,
+      "loss": 2.5987,
+      "step": 9246
+    },
+    {
+      "epoch": 0.837059835249389,
+      "grad_norm": 0.8333005309104919,
+      "learning_rate": 2.8413314000581338e-06,
+      "loss": 2.5971,
+      "step": 9247
+    },
+    {
+      "epoch": 0.8371503575631393,
+      "grad_norm": 0.8795336484909058,
+      "learning_rate": 2.8338255377046375e-06,
+      "loss": 2.5984,
+      "step": 9248
+    },
+    {
+      "epoch": 0.8372408798768897,
+      "grad_norm": 0.7897392511367798,
+      "learning_rate": 2.8263294599822577e-06,
+      "loss": 2.6101,
+      "step": 9249
+    },
+    {
+      "epoch": 0.83733140219064,
+      "grad_norm": 0.7922962307929993,
+      "learning_rate": 2.818843167645835e-06,
+      "loss": 2.6432,
+      "step": 9250
+    },
+    {
+      "epoch": 0.8374219245043903,
+      "grad_norm": 0.8887566328048706,
+      "learning_rate": 2.811366661449244e-06,
+      "loss": 2.5893,
+      "step": 9251
+    },
+    {
+      "epoch": 0.8375124468181406,
+      "grad_norm": 0.8131504654884338,
+      "learning_rate": 2.8038999421453826e-06,
+      "loss": 2.6805,
+      "step": 9252
+    },
+    {
+      "epoch": 0.837602969131891,
+      "grad_norm": 0.8904218673706055,
+      "learning_rate": 2.7964430104861274e-06,
+      "loss": 2.6164,
+      "step": 9253
+    },
+    {
+      "epoch": 0.8376934914456413,
+      "grad_norm": 0.8426781892776489,
+      "learning_rate": 2.7889958672224104e-06,
+      "loss": 2.6167,
+      "step": 9254
+    },
+    {
+      "epoch": 0.8377840137593917,
+      "grad_norm": 0.7983620762825012,
+      "learning_rate": 2.7815585131041435e-06,
+      "loss": 2.5987,
+      "step": 9255
+    },
+    {
+      "epoch": 0.837874536073142,
+      "grad_norm": 0.8963645100593567,
+      "learning_rate": 2.7741309488802712e-06,
+      "loss": 2.7137,
+      "step": 9256
+    },
+    {
+      "epoch": 0.8379650583868924,
+      "grad_norm": 0.8122845888137817,
+      "learning_rate": 2.7667131752987517e-06,
+      "loss": 2.5951,
+      "step": 9257
+    },
+    {
+      "epoch": 0.8380555807006427,
+      "grad_norm": 0.83088618516922,
+      "learning_rate": 2.759305193106554e-06,
+      "loss": 2.663,
+      "step": 9258
+    },
+    {
+      "epoch": 0.8381461030143931,
+      "grad_norm": 0.8195231556892395,
+      "learning_rate": 2.7519070030496695e-06,
+      "loss": 2.59,
+      "step": 9259
+    },
+    {
+      "epoch": 0.8382366253281434,
+      "grad_norm": 0.8513022065162659,
+      "learning_rate": 2.744518605873092e-06,
+      "loss": 2.6079,
+      "step": 9260
+    },
+    {
+      "epoch": 0.8383271476418938,
+      "grad_norm": 0.830094039440155,
+      "learning_rate": 2.737140002320815e-06,
+      "loss": 2.668,
+      "step": 9261
+    },
+    {
+      "epoch": 0.838417669955644,
+      "grad_norm": 0.8253921270370483,
+      "learning_rate": 2.7297711931358993e-06,
+      "loss": 2.6399,
+      "step": 9262
+    },
+    {
+      "epoch": 0.8385081922693944,
+      "grad_norm": 0.9423378705978394,
+      "learning_rate": 2.7224121790603517e-06,
+      "loss": 2.6048,
+      "step": 9263
+    },
+    {
+      "epoch": 0.8385987145831447,
+      "grad_norm": 0.8792223930358887,
+      "learning_rate": 2.715062960835235e-06,
+      "loss": 2.626,
+      "step": 9264
+    },
+    {
+      "epoch": 0.8386892368968951,
+      "grad_norm": 0.8084904551506042,
+      "learning_rate": 2.707723539200613e-06,
+      "loss": 2.5398,
+      "step": 9265
+    },
+    {
+      "epoch": 0.8387797592106454,
+      "grad_norm": 0.8560845255851746,
+      "learning_rate": 2.7003939148955715e-06,
+      "loss": 2.6488,
+      "step": 9266
+    },
+    {
+      "epoch": 0.8388702815243958,
+      "grad_norm": 0.9024186730384827,
+      "learning_rate": 2.6930740886581985e-06,
+      "loss": 2.6038,
+      "step": 9267
+    },
+    {
+      "epoch": 0.8389608038381461,
+      "grad_norm": 0.8976023197174072,
+      "learning_rate": 2.685764061225615e-06,
+      "loss": 2.5994,
+      "step": 9268
+    },
+    {
+      "epoch": 0.8390513261518965,
+      "grad_norm": 0.8648386001586914,
+      "learning_rate": 2.678463833333911e-06,
+      "loss": 2.5852,
+      "step": 9269
+    },
+    {
+      "epoch": 0.8391418484656468,
+      "grad_norm": 0.8111081123352051,
+      "learning_rate": 2.6711734057182415e-06,
+      "loss": 2.6155,
+      "step": 9270
+    },
+    {
+      "epoch": 0.8392323707793972,
+      "grad_norm": 0.808134138584137,
+      "learning_rate": 2.6638927791127534e-06,
+      "loss": 2.5815,
+      "step": 9271
+    },
+    {
+      "epoch": 0.8393228930931474,
+      "grad_norm": 0.8514514565467834,
+      "learning_rate": 2.6566219542505933e-06,
+      "loss": 2.6199,
+      "step": 9272
+    },
+    {
+      "epoch": 0.8394134154068978,
+      "grad_norm": 0.8413440585136414,
+      "learning_rate": 2.649360931863942e-06,
+      "loss": 2.584,
+      "step": 9273
+    },
+    {
+      "epoch": 0.8395039377206481,
+      "grad_norm": 0.8079906105995178,
+      "learning_rate": 2.6421097126839712e-06,
+      "loss": 2.5947,
+      "step": 9274
+    },
+    {
+      "epoch": 0.8395944600343985,
+      "grad_norm": 0.8232433199882507,
+      "learning_rate": 2.6348682974408955e-06,
+      "loss": 2.6103,
+      "step": 9275
+    },
+    {
+      "epoch": 0.8396849823481488,
+      "grad_norm": 0.9182198643684387,
+      "learning_rate": 2.6276366868639325e-06,
+      "loss": 2.6223,
+      "step": 9276
+    },
+    {
+      "epoch": 0.8397755046618992,
+      "grad_norm": 0.8715531826019287,
+      "learning_rate": 2.6204148816812657e-06,
+      "loss": 2.671,
+      "step": 9277
+    },
+    {
+      "epoch": 0.8398660269756495,
+      "grad_norm": 0.8018813729286194,
+      "learning_rate": 2.6132028826201695e-06,
+      "loss": 2.6102,
+      "step": 9278
+    },
+    {
+      "epoch": 0.8399565492893999,
+      "grad_norm": 0.8041639924049377,
+      "learning_rate": 2.6060006904068844e-06,
+      "loss": 2.5943,
+      "step": 9279
+    },
+    {
+      "epoch": 0.8400470716031502,
+      "grad_norm": 0.9004766345024109,
+      "learning_rate": 2.5988083057666533e-06,
+      "loss": 2.5938,
+      "step": 9280
+    },
+    {
+      "epoch": 0.8401375939169006,
+      "grad_norm": 0.8807133436203003,
+      "learning_rate": 2.5916257294237633e-06,
+      "loss": 2.6039,
+      "step": 9281
+    },
+    {
+      "epoch": 0.8402281162306509,
+      "grad_norm": 0.8423158526420593,
+      "learning_rate": 2.5844529621015023e-06,
+      "loss": 2.5367,
+      "step": 9282
+    },
+    {
+      "epoch": 0.8403186385444011,
+      "grad_norm": 0.8630332946777344,
+      "learning_rate": 2.577290004522148e-06,
+      "loss": 2.5719,
+      "step": 9283
+    },
+    {
+      "epoch": 0.8404091608581515,
+      "grad_norm": 0.8227906823158264,
+      "learning_rate": 2.5701368574070463e-06,
+      "loss": 2.6222,
+      "step": 9284
+    },
+    {
+      "epoch": 0.8404996831719018,
+      "grad_norm": 0.8309725522994995,
+      "learning_rate": 2.5629935214764865e-06,
+      "loss": 2.6203,
+      "step": 9285
+    },
+    {
+      "epoch": 0.8405902054856522,
+      "grad_norm": 0.8114734888076782,
+      "learning_rate": 2.5558599974498054e-06,
+      "loss": 2.5701,
+      "step": 9286
+    },
+    {
+      "epoch": 0.8406807277994025,
+      "grad_norm": 0.9079375267028809,
+      "learning_rate": 2.548736286045361e-06,
+      "loss": 2.5715,
+      "step": 9287
+    },
+    {
+      "epoch": 0.8407712501131529,
+      "grad_norm": 0.9126922488212585,
+      "learning_rate": 2.5416223879805134e-06,
+      "loss": 2.6241,
+      "step": 9288
+    },
+    {
+      "epoch": 0.8408617724269032,
+      "grad_norm": 0.8237440586090088,
+      "learning_rate": 2.5345183039716112e-06,
+      "loss": 2.6061,
+      "step": 9289
+    },
+    {
+      "epoch": 0.8409522947406536,
+      "grad_norm": 0.8415271043777466,
+      "learning_rate": 2.5274240347340717e-06,
+      "loss": 2.5721,
+      "step": 9290
+    },
+    {
+      "epoch": 0.8410428170544039,
+      "grad_norm": 0.8436958193778992,
+      "learning_rate": 2.520339580982234e-06,
+      "loss": 2.6532,
+      "step": 9291
+    },
+    {
+      "epoch": 0.8411333393681543,
+      "grad_norm": 0.911311686038971,
+      "learning_rate": 2.5132649434295606e-06,
+      "loss": 2.6995,
+      "step": 9292
+    },
+    {
+      "epoch": 0.8412238616819046,
+      "grad_norm": 0.9336774349212646,
+      "learning_rate": 2.506200122788427e-06,
+      "loss": 2.6893,
+      "step": 9293
+    },
+    {
+      "epoch": 0.841314383995655,
+      "grad_norm": 0.8583858013153076,
+      "learning_rate": 2.499145119770274e-06,
+      "loss": 2.5763,
+      "step": 9294
+    },
+    {
+      "epoch": 0.8414049063094052,
+      "grad_norm": 0.8368796706199646,
+      "learning_rate": 2.4920999350855458e-06,
+      "loss": 2.5478,
+      "step": 9295
+    },
+    {
+      "epoch": 0.8414954286231556,
+      "grad_norm": 0.8354606628417969,
+      "learning_rate": 2.4850645694436736e-06,
+      "loss": 2.655,
+      "step": 9296
+    },
+    {
+      "epoch": 0.8415859509369059,
+      "grad_norm": 0.8495997786521912,
+      "learning_rate": 2.478039023553136e-06,
+      "loss": 2.6892,
+      "step": 9297
+    },
+    {
+      "epoch": 0.8416764732506563,
+      "grad_norm": 0.8512294292449951,
+      "learning_rate": 2.471023298121422e-06,
+      "loss": 2.6317,
+      "step": 9298
+    },
+    {
+      "epoch": 0.8417669955644066,
+      "grad_norm": 0.8798067569732666,
+      "learning_rate": 2.464017393854956e-06,
+      "loss": 2.6381,
+      "step": 9299
+    },
+    {
+      "epoch": 0.841857517878157,
+      "grad_norm": 0.8893240690231323,
+      "learning_rate": 2.4570213114592954e-06,
+      "loss": 2.625,
+      "step": 9300
+    },
+    {
+      "epoch": 0.8419480401919073,
+      "grad_norm": 0.7912008166313171,
+      "learning_rate": 2.45003505163891e-06,
+      "loss": 2.5689,
+      "step": 9301
+    },
+    {
+      "epoch": 0.8420385625056577,
+      "grad_norm": 0.850443959236145,
+      "learning_rate": 2.443058615097338e-06,
+      "loss": 2.6325,
+      "step": 9302
+    },
+    {
+      "epoch": 0.842129084819408,
+      "grad_norm": 0.857284665107727,
+      "learning_rate": 2.4360920025370827e-06,
+      "loss": 2.679,
+      "step": 9303
+    },
+    {
+      "epoch": 0.8422196071331584,
+      "grad_norm": 0.8806108832359314,
+      "learning_rate": 2.4291352146596945e-06,
+      "loss": 2.6189,
+      "step": 9304
+    },
+    {
+      "epoch": 0.8423101294469086,
+      "grad_norm": 0.8154246211051941,
+      "learning_rate": 2.4221882521657136e-06,
+      "loss": 2.5786,
+      "step": 9305
+    },
+    {
+      "epoch": 0.842400651760659,
+      "grad_norm": 0.8218053579330444,
+      "learning_rate": 2.4152511157547244e-06,
+      "loss": 2.6728,
+      "step": 9306
+    },
+    {
+      "epoch": 0.8424911740744093,
+      "grad_norm": 0.8055351376533508,
+      "learning_rate": 2.4083238061252567e-06,
+      "loss": 2.6068,
+      "step": 9307
+    },
+    {
+      "epoch": 0.8425816963881597,
+      "grad_norm": 0.9141035676002502,
+      "learning_rate": 2.401406323974931e-06,
+      "loss": 2.6306,
+      "step": 9308
+    },
+    {
+      "epoch": 0.84267221870191,
+      "grad_norm": 0.8625081777572632,
+      "learning_rate": 2.3944986700003004e-06,
+      "loss": 2.5936,
+      "step": 9309
+    },
+    {
+      "epoch": 0.8427627410156604,
+      "grad_norm": 0.8520241975784302,
+      "learning_rate": 2.3876008448969976e-06,
+      "loss": 2.5859,
+      "step": 9310
+    },
+    {
+      "epoch": 0.8428532633294107,
+      "grad_norm": 0.8802344799041748,
+      "learning_rate": 2.380712849359612e-06,
+      "loss": 2.5918,
+      "step": 9311
+    },
+    {
+      "epoch": 0.8429437856431611,
+      "grad_norm": 0.9308227300643921,
+      "learning_rate": 2.3738346840817773e-06,
+      "loss": 2.6696,
+      "step": 9312
+    },
+    {
+      "epoch": 0.8430343079569114,
+      "grad_norm": 0.782535195350647,
+      "learning_rate": 2.366966349756106e-06,
+      "loss": 2.6541,
+      "step": 9313
+    },
+    {
+      "epoch": 0.8431248302706618,
+      "grad_norm": 0.7898374795913696,
+      "learning_rate": 2.3601078470742664e-06,
+      "loss": 2.6019,
+      "step": 9314
+    },
+    {
+      "epoch": 0.843215352584412,
+      "grad_norm": 0.8801725506782532,
+      "learning_rate": 2.3532591767268853e-06,
+      "loss": 2.605,
+      "step": 9315
+    },
+    {
+      "epoch": 0.8433058748981624,
+      "grad_norm": 0.8103786706924438,
+      "learning_rate": 2.3464203394036322e-06,
+      "loss": 2.5527,
+      "step": 9316
+    },
+    {
+      "epoch": 0.8433963972119127,
+      "grad_norm": 0.8946208357810974,
+      "learning_rate": 2.3395913357931786e-06,
+      "loss": 2.5804,
+      "step": 9317
+    },
+    {
+      "epoch": 0.8434869195256631,
+      "grad_norm": 0.8092337846755981,
+      "learning_rate": 2.332772166583208e-06,
+      "loss": 2.6068,
+      "step": 9318
+    },
+    {
+      "epoch": 0.8435774418394134,
+      "grad_norm": 0.8034257292747498,
+      "learning_rate": 2.3259628324604155e-06,
+      "loss": 2.6107,
+      "step": 9319
+    },
+    {
+      "epoch": 0.8436679641531638,
+      "grad_norm": 0.9613232016563416,
+      "learning_rate": 2.3191633341104856e-06,
+      "loss": 2.6176,
+      "step": 9320
+    },
+    {
+      "epoch": 0.8437584864669141,
+      "grad_norm": 0.8717933297157288,
+      "learning_rate": 2.3123736722181375e-06,
+      "loss": 2.6105,
+      "step": 9321
+    },
+    {
+      "epoch": 0.8438490087806645,
+      "grad_norm": 0.8280078172683716,
+      "learning_rate": 2.3055938474670915e-06,
+      "loss": 2.5993,
+      "step": 9322
+    },
+    {
+      "epoch": 0.8439395310944148,
+      "grad_norm": 0.8411900401115417,
+      "learning_rate": 2.298823860540056e-06,
+      "loss": 2.6247,
+      "step": 9323
+    },
+    {
+      "epoch": 0.8440300534081651,
+      "grad_norm": 0.8409061431884766,
+      "learning_rate": 2.292063712118797e-06,
+      "loss": 2.5859,
+      "step": 9324
+    },
+    {
+      "epoch": 0.8441205757219155,
+      "grad_norm": 0.8324682116508484,
+      "learning_rate": 2.2853134028840594e-06,
+      "loss": 2.6155,
+      "step": 9325
+    },
+    {
+      "epoch": 0.8442110980356657,
+      "grad_norm": 0.8437857627868652,
+      "learning_rate": 2.2785729335155657e-06,
+      "loss": 2.6489,
+      "step": 9326
+    },
+    {
+      "epoch": 0.8443016203494161,
+      "grad_norm": 0.8383326530456543,
+      "learning_rate": 2.271842304692118e-06,
+      "loss": 2.6049,
+      "step": 9327
+    },
+    {
+      "epoch": 0.8443921426631664,
+      "grad_norm": 0.8164210915565491,
+      "learning_rate": 2.265121517091473e-06,
+      "loss": 2.6312,
+      "step": 9328
+    },
+    {
+      "epoch": 0.8444826649769168,
+      "grad_norm": 0.865288496017456,
+      "learning_rate": 2.2584105713904125e-06,
+      "loss": 2.6321,
+      "step": 9329
+    },
+    {
+      "epoch": 0.8445731872906671,
+      "grad_norm": 0.8106584548950195,
+      "learning_rate": 2.2517094682647397e-06,
+      "loss": 2.6074,
+      "step": 9330
+    },
+    {
+      "epoch": 0.8446637096044175,
+      "grad_norm": 0.8246323466300964,
+      "learning_rate": 2.245018208389249e-06,
+      "loss": 2.5897,
+      "step": 9331
+    },
+    {
+      "epoch": 0.8447542319181678,
+      "grad_norm": 0.8730551600456238,
+      "learning_rate": 2.2383367924377452e-06,
+      "loss": 2.6694,
+      "step": 9332
+    },
+    {
+      "epoch": 0.8448447542319182,
+      "grad_norm": 0.8684254288673401,
+      "learning_rate": 2.2316652210830568e-06,
+      "loss": 2.5865,
+      "step": 9333
+    },
+    {
+      "epoch": 0.8449352765456685,
+      "grad_norm": 0.860093891620636,
+      "learning_rate": 2.2250034949969913e-06,
+      "loss": 2.6367,
+      "step": 9334
+    },
+    {
+      "epoch": 0.8450257988594189,
+      "grad_norm": 0.878279447555542,
+      "learning_rate": 2.2183516148504226e-06,
+      "loss": 2.6596,
+      "step": 9335
+    },
+    {
+      "epoch": 0.8451163211731691,
+      "grad_norm": 0.864924967288971,
+      "learning_rate": 2.2117095813131484e-06,
+      "loss": 2.6327,
+      "step": 9336
+    },
+    {
+      "epoch": 0.8452068434869195,
+      "grad_norm": 0.8816613554954529,
+      "learning_rate": 2.205077395054056e-06,
+      "loss": 2.6484,
+      "step": 9337
+    },
+    {
+      "epoch": 0.8452973658006698,
+      "grad_norm": 0.8233687877655029,
+      "learning_rate": 2.198455056740989e-06,
+      "loss": 2.5487,
+      "step": 9338
+    },
+    {
+      "epoch": 0.8453878881144202,
+      "grad_norm": 0.9007396697998047,
+      "learning_rate": 2.1918425670408247e-06,
+      "loss": 2.6456,
+      "step": 9339
+    },
+    {
+      "epoch": 0.8454784104281705,
+      "grad_norm": 0.8072501420974731,
+      "learning_rate": 2.1852399266194314e-06,
+      "loss": 2.5781,
+      "step": 9340
+    },
+    {
+      "epoch": 0.8455689327419209,
+      "grad_norm": 0.879115104675293,
+      "learning_rate": 2.1786471361417206e-06,
+      "loss": 2.5904,
+      "step": 9341
+    },
+    {
+      "epoch": 0.8456594550556712,
+      "grad_norm": 0.9695716500282288,
+      "learning_rate": 2.1720641962715393e-06,
+      "loss": 2.6004,
+      "step": 9342
+    },
+    {
+      "epoch": 0.8457499773694216,
+      "grad_norm": 0.8915279507637024,
+      "learning_rate": 2.1654911076718355e-06,
+      "loss": 2.5751,
+      "step": 9343
+    },
+    {
+      "epoch": 0.8458404996831719,
+      "grad_norm": 0.8442407846450806,
+      "learning_rate": 2.1589278710045013e-06,
+      "loss": 2.6255,
+      "step": 9344
+    },
+    {
+      "epoch": 0.8459310219969223,
+      "grad_norm": 0.8827381134033203,
+      "learning_rate": 2.152374486930442e-06,
+      "loss": 2.6682,
+      "step": 9345
+    },
+    {
+      "epoch": 0.8460215443106726,
+      "grad_norm": 0.8156737089157104,
+      "learning_rate": 2.145830956109596e-06,
+      "loss": 2.6366,
+      "step": 9346
+    },
+    {
+      "epoch": 0.846112066624423,
+      "grad_norm": 0.7898461222648621,
+      "learning_rate": 2.1392972792009026e-06,
+      "loss": 2.5952,
+      "step": 9347
+    },
+    {
+      "epoch": 0.8462025889381732,
+      "grad_norm": 0.8535096645355225,
+      "learning_rate": 2.1327734568622914e-06,
+      "loss": 2.6272,
+      "step": 9348
+    },
+    {
+      "epoch": 0.8462931112519236,
+      "grad_norm": 0.8594303131103516,
+      "learning_rate": 2.126259489750715e-06,
+      "loss": 2.6145,
+      "step": 9349
+    },
+    {
+      "epoch": 0.8463836335656739,
+      "grad_norm": 0.8821359276771545,
+      "learning_rate": 2.119755378522137e-06,
+      "loss": 2.6136,
+      "step": 9350
+    },
+    {
+      "epoch": 0.8464741558794243,
+      "grad_norm": 0.817482590675354,
+      "learning_rate": 2.1132611238315003e-06,
+      "loss": 2.6069,
+      "step": 9351
+    },
+    {
+      "epoch": 0.8465646781931746,
+      "grad_norm": 0.8469284772872925,
+      "learning_rate": 2.1067767263327933e-06,
+      "loss": 2.6322,
+      "step": 9352
+    },
+    {
+      "epoch": 0.846655200506925,
+      "grad_norm": 0.7845486998558044,
+      "learning_rate": 2.100302186679004e-06,
+      "loss": 2.6096,
+      "step": 9353
+    },
+    {
+      "epoch": 0.8467457228206753,
+      "grad_norm": 0.7727720737457275,
+      "learning_rate": 2.0938375055220893e-06,
+      "loss": 2.6145,
+      "step": 9354
+    },
+    {
+      "epoch": 0.8468362451344257,
+      "grad_norm": 0.7955880165100098,
+      "learning_rate": 2.087382683513073e-06,
+      "loss": 2.6062,
+      "step": 9355
+    },
+    {
+      "epoch": 0.846926767448176,
+      "grad_norm": 0.9472482204437256,
+      "learning_rate": 2.0809377213019234e-06,
+      "loss": 2.6432,
+      "step": 9356
+    },
+    {
+      "epoch": 0.8470172897619264,
+      "grad_norm": 0.8384003639221191,
+      "learning_rate": 2.074502619537688e-06,
+      "loss": 2.6339,
+      "step": 9357
+    },
+    {
+      "epoch": 0.8471078120756766,
+      "grad_norm": 0.9159187078475952,
+      "learning_rate": 2.0680773788683494e-06,
+      "loss": 2.5945,
+      "step": 9358
+    },
+    {
+      "epoch": 0.847198334389427,
+      "grad_norm": 0.8477405309677124,
+      "learning_rate": 2.0616619999409337e-06,
+      "loss": 2.6826,
+      "step": 9359
+    },
+    {
+      "epoch": 0.8472888567031773,
+      "grad_norm": 0.8061385750770569,
+      "learning_rate": 2.05525648340148e-06,
+      "loss": 2.5887,
+      "step": 9360
+    },
+    {
+      "epoch": 0.8473793790169277,
+      "grad_norm": 0.7954801917076111,
+      "learning_rate": 2.048860829895016e-06,
+      "loss": 2.6384,
+      "step": 9361
+    },
+    {
+      "epoch": 0.847469901330678,
+      "grad_norm": 0.8151822090148926,
+      "learning_rate": 2.0424750400655947e-06,
+      "loss": 2.6229,
+      "step": 9362
+    },
+    {
+      "epoch": 0.8475604236444284,
+      "grad_norm": 0.8000811338424683,
+      "learning_rate": 2.0360991145562557e-06,
+      "loss": 2.5853,
+      "step": 9363
+    },
+    {
+      "epoch": 0.8476509459581787,
+      "grad_norm": 0.8454170823097229,
+      "learning_rate": 2.029733054009042e-06,
+      "loss": 2.6348,
+      "step": 9364
+    },
+    {
+      "epoch": 0.847741468271929,
+      "grad_norm": 0.8597866892814636,
+      "learning_rate": 2.0233768590650403e-06,
+      "loss": 2.5944,
+      "step": 9365
+    },
+    {
+      "epoch": 0.8478319905856794,
+      "grad_norm": 0.8464025259017944,
+      "learning_rate": 2.0170305303643054e-06,
+      "loss": 2.5851,
+      "step": 9366
+    },
+    {
+      "epoch": 0.8479225128994297,
+      "grad_norm": 0.8239722847938538,
+      "learning_rate": 2.0106940685459154e-06,
+      "loss": 2.6316,
+      "step": 9367
+    },
+    {
+      "epoch": 0.84801303521318,
+      "grad_norm": 0.8130668997764587,
+      "learning_rate": 2.004367474247948e-06,
+      "loss": 2.615,
+      "step": 9368
+    },
+    {
+      "epoch": 0.8481035575269303,
+      "grad_norm": 0.8077114820480347,
+      "learning_rate": 1.998050748107505e-06,
+      "loss": 2.6375,
+      "step": 9369
+    },
+    {
+      "epoch": 0.8481940798406807,
+      "grad_norm": 0.8757891058921814,
+      "learning_rate": 1.9917438907606556e-06,
+      "loss": 2.5772,
+      "step": 9370
+    },
+    {
+      "epoch": 0.848284602154431,
+      "grad_norm": 0.8093799948692322,
+      "learning_rate": 1.985446902842536e-06,
+      "loss": 2.6391,
+      "step": 9371
+    },
+    {
+      "epoch": 0.8483751244681814,
+      "grad_norm": 0.8270487785339355,
+      "learning_rate": 1.9791597849872057e-06,
+      "loss": 2.5764,
+      "step": 9372
+    },
+    {
+      "epoch": 0.8484656467819317,
+      "grad_norm": 0.9486851692199707,
+      "learning_rate": 1.9728825378278246e-06,
+      "loss": 2.6505,
+      "step": 9373
+    },
+    {
+      "epoch": 0.8485561690956821,
+      "grad_norm": 0.9480065107345581,
+      "learning_rate": 1.966615161996477e-06,
+      "loss": 2.5907,
+      "step": 9374
+    },
+    {
+      "epoch": 0.8486466914094324,
+      "grad_norm": 0.8524114489555359,
+      "learning_rate": 1.960357658124301e-06,
+      "loss": 2.6317,
+      "step": 9375
+    },
+    {
+      "epoch": 0.8487372137231828,
+      "grad_norm": 0.908987820148468,
+      "learning_rate": 1.954110026841427e-06,
+      "loss": 2.5713,
+      "step": 9376
+    },
+    {
+      "epoch": 0.8488277360369331,
+      "grad_norm": 0.8268985152244568,
+      "learning_rate": 1.947872268776996e-06,
+      "loss": 2.6577,
+      "step": 9377
+    },
+    {
+      "epoch": 0.8489182583506835,
+      "grad_norm": 0.7865447998046875,
+      "learning_rate": 1.941644384559138e-06,
+      "loss": 2.5872,
+      "step": 9378
+    },
+    {
+      "epoch": 0.8490087806644337,
+      "grad_norm": 0.8748701810836792,
+      "learning_rate": 1.935426374815008e-06,
+      "loss": 2.6353,
+      "step": 9379
+    },
+    {
+      "epoch": 0.8490993029781841,
+      "grad_norm": 0.8506055474281311,
+      "learning_rate": 1.9292182401707603e-06,
+      "loss": 2.6137,
+      "step": 9380
+    },
+    {
+      "epoch": 0.8491898252919344,
+      "grad_norm": 0.7830395698547363,
+      "learning_rate": 1.9230199812515616e-06,
+      "loss": 2.5969,
+      "step": 9381
+    },
+    {
+      "epoch": 0.8492803476056848,
+      "grad_norm": 0.7881369590759277,
+      "learning_rate": 1.9168315986815567e-06,
+      "loss": 2.6085,
+      "step": 9382
+    },
+    {
+      "epoch": 0.8493708699194351,
+      "grad_norm": 0.891779363155365,
+      "learning_rate": 1.910653093083925e-06,
+      "loss": 2.6918,
+      "step": 9383
+    },
+    {
+      "epoch": 0.8494613922331855,
+      "grad_norm": 0.819579541683197,
+      "learning_rate": 1.904484465080847e-06,
+      "loss": 2.5821,
+      "step": 9384
+    },
+    {
+      "epoch": 0.8495519145469358,
+      "grad_norm": 0.8453068137168884,
+      "learning_rate": 1.898325715293503e-06,
+      "loss": 2.5648,
+      "step": 9385
+    },
+    {
+      "epoch": 0.8496424368606862,
+      "grad_norm": 0.8719899654388428,
+      "learning_rate": 1.8921768443420528e-06,
+      "loss": 2.6316,
+      "step": 9386
+    },
+    {
+      "epoch": 0.8497329591744365,
+      "grad_norm": 0.9381957054138184,
+      "learning_rate": 1.8860378528457456e-06,
+      "loss": 2.6796,
+      "step": 9387
+    },
+    {
+      "epoch": 0.8498234814881869,
+      "grad_norm": 0.8671666979789734,
+      "learning_rate": 1.87990874142272e-06,
+      "loss": 2.6189,
+      "step": 9388
+    },
+    {
+      "epoch": 0.8499140038019372,
+      "grad_norm": 0.8676027655601501,
+      "learning_rate": 1.8737895106902048e-06,
+      "loss": 2.5648,
+      "step": 9389
+    },
+    {
+      "epoch": 0.8500045261156876,
+      "grad_norm": 0.8236902356147766,
+      "learning_rate": 1.8676801612643957e-06,
+      "loss": 2.5913,
+      "step": 9390
+    },
+    {
+      "epoch": 0.8500950484294378,
+      "grad_norm": 0.769631028175354,
+      "learning_rate": 1.8615806937605118e-06,
+      "loss": 2.6084,
+      "step": 9391
+    },
+    {
+      "epoch": 0.8501855707431882,
+      "grad_norm": 0.8675589561462402,
+      "learning_rate": 1.8554911087927618e-06,
+      "loss": 2.6249,
+      "step": 9392
+    },
+    {
+      "epoch": 0.8502760930569385,
+      "grad_norm": 0.8161866664886475,
+      "learning_rate": 1.8494114069743885e-06,
+      "loss": 2.6269,
+      "step": 9393
+    },
+    {
+      "epoch": 0.8503666153706889,
+      "grad_norm": 0.8564107418060303,
+      "learning_rate": 1.8433415889175799e-06,
+      "loss": 2.6545,
+      "step": 9394
+    },
+    {
+      "epoch": 0.8504571376844392,
+      "grad_norm": 0.8710647225379944,
+      "learning_rate": 1.8372816552336026e-06,
+      "loss": 2.6265,
+      "step": 9395
+    },
+    {
+      "epoch": 0.8505476599981896,
+      "grad_norm": 0.8734665513038635,
+      "learning_rate": 1.8312316065326796e-06,
+      "loss": 2.6363,
+      "step": 9396
+    },
+    {
+      "epoch": 0.8506381823119399,
+      "grad_norm": 0.8522061109542847,
+      "learning_rate": 1.8251914434240347e-06,
+      "loss": 2.5858,
+      "step": 9397
+    },
+    {
+      "epoch": 0.8507287046256903,
+      "grad_norm": 0.838045060634613,
+      "learning_rate": 1.819161166515937e-06,
+      "loss": 2.6064,
+      "step": 9398
+    },
+    {
+      "epoch": 0.8508192269394406,
+      "grad_norm": 0.8066017031669617,
+      "learning_rate": 1.8131407764156117e-06,
+      "loss": 2.6138,
+      "step": 9399
+    },
+    {
+      "epoch": 0.850909749253191,
+      "grad_norm": 0.8111808896064758,
+      "learning_rate": 1.8071302737293295e-06,
+      "loss": 2.5855,
+      "step": 9400
+    },
+    {
+      "epoch": 0.8510002715669412,
+      "grad_norm": 0.8490340709686279,
+      "learning_rate": 1.8011296590623506e-06,
+      "loss": 2.5868,
+      "step": 9401
+    },
+    {
+      "epoch": 0.8510907938806916,
+      "grad_norm": 0.8828288316726685,
+      "learning_rate": 1.7951389330189029e-06,
+      "loss": 2.6405,
+      "step": 9402
+    },
+    {
+      "epoch": 0.8511813161944419,
+      "grad_norm": 0.918430507183075,
+      "learning_rate": 1.7891580962023036e-06,
+      "loss": 2.6249,
+      "step": 9403
+    },
+    {
+      "epoch": 0.8512718385081923,
+      "grad_norm": 0.7756317257881165,
+      "learning_rate": 1.7831871492147934e-06,
+      "loss": 2.6389,
+      "step": 9404
+    },
+    {
+      "epoch": 0.8513623608219426,
+      "grad_norm": 0.9137489795684814,
+      "learning_rate": 1.7772260926576356e-06,
+      "loss": 2.6336,
+      "step": 9405
+    },
+    {
+      "epoch": 0.8514528831356929,
+      "grad_norm": 0.827333390712738,
+      "learning_rate": 1.771274927131139e-06,
+      "loss": 2.5928,
+      "step": 9406
+    },
+    {
+      "epoch": 0.8515434054494433,
+      "grad_norm": 0.8530473113059998,
+      "learning_rate": 1.7653336532345577e-06,
+      "loss": 2.6192,
+      "step": 9407
+    },
+    {
+      "epoch": 0.8516339277631936,
+      "grad_norm": 0.8667789101600647,
+      "learning_rate": 1.7594022715661906e-06,
+      "loss": 2.6038,
+      "step": 9408
+    },
+    {
+      "epoch": 0.851724450076944,
+      "grad_norm": 0.8074167370796204,
+      "learning_rate": 1.753480782723338e-06,
+      "loss": 2.6365,
+      "step": 9409
+    },
+    {
+      "epoch": 0.8518149723906943,
+      "grad_norm": 0.8859494924545288,
+      "learning_rate": 1.747569187302267e-06,
+      "loss": 2.6868,
+      "step": 9410
+    },
+    {
+      "epoch": 0.8519054947044447,
+      "grad_norm": 0.8398774862289429,
+      "learning_rate": 1.7416674858983018e-06,
+      "loss": 2.7059,
+      "step": 9411
+    },
+    {
+      "epoch": 0.8519960170181949,
+      "grad_norm": 0.9021377563476562,
+      "learning_rate": 1.7357756791057334e-06,
+      "loss": 2.6566,
+      "step": 9412
+    },
+    {
+      "epoch": 0.8520865393319453,
+      "grad_norm": 0.8221820592880249,
+      "learning_rate": 1.7298937675178428e-06,
+      "loss": 2.6093,
+      "step": 9413
+    },
+    {
+      "epoch": 0.8521770616456956,
+      "grad_norm": 0.9722798466682434,
+      "learning_rate": 1.7240217517269897e-06,
+      "loss": 2.6304,
+      "step": 9414
+    },
+    {
+      "epoch": 0.852267583959446,
+      "grad_norm": 0.820933997631073,
+      "learning_rate": 1.7181596323244454e-06,
+      "loss": 2.6182,
+      "step": 9415
+    },
+    {
+      "epoch": 0.8523581062731963,
+      "grad_norm": 0.914440393447876,
+      "learning_rate": 1.7123074099005377e-06,
+      "loss": 2.6131,
+      "step": 9416
+    },
+    {
+      "epoch": 0.8524486285869467,
+      "grad_norm": 0.9035472273826599,
+      "learning_rate": 1.706465085044584e-06,
+      "loss": 2.5809,
+      "step": 9417
+    },
+    {
+      "epoch": 0.852539150900697,
+      "grad_norm": 0.9354585409164429,
+      "learning_rate": 1.7006326583449029e-06,
+      "loss": 2.6365,
+      "step": 9418
+    },
+    {
+      "epoch": 0.8526296732144474,
+      "grad_norm": 0.8287830948829651,
+      "learning_rate": 1.6948101303888243e-06,
+      "loss": 2.6085,
+      "step": 9419
+    },
+    {
+      "epoch": 0.8527201955281977,
+      "grad_norm": 0.8384172320365906,
+      "learning_rate": 1.6889975017626903e-06,
+      "loss": 2.6083,
+      "step": 9420
+    },
+    {
+      "epoch": 0.8528107178419481,
+      "grad_norm": 0.817358136177063,
+      "learning_rate": 1.6831947730517883e-06,
+      "loss": 2.6818,
+      "step": 9421
+    },
+    {
+      "epoch": 0.8529012401556983,
+      "grad_norm": 0.8479032516479492,
+      "learning_rate": 1.6774019448405064e-06,
+      "loss": 2.5817,
+      "step": 9422
+    },
+    {
+      "epoch": 0.8529917624694487,
+      "grad_norm": 0.8063598275184631,
+      "learning_rate": 1.671619017712156e-06,
+      "loss": 2.5778,
+      "step": 9423
+    },
+    {
+      "epoch": 0.853082284783199,
+      "grad_norm": 0.840354323387146,
+      "learning_rate": 1.665845992249071e-06,
+      "loss": 2.5911,
+      "step": 9424
+    },
+    {
+      "epoch": 0.8531728070969494,
+      "grad_norm": 0.8489820957183838,
+      "learning_rate": 1.6600828690326087e-06,
+      "loss": 2.6492,
+      "step": 9425
+    },
+    {
+      "epoch": 0.8532633294106997,
+      "grad_norm": 0.8331730961799622,
+      "learning_rate": 1.6543296486431159e-06,
+      "loss": 2.6017,
+      "step": 9426
+    },
+    {
+      "epoch": 0.8533538517244501,
+      "grad_norm": 0.8832200169563293,
+      "learning_rate": 1.6485863316599294e-06,
+      "loss": 2.5786,
+      "step": 9427
+    },
+    {
+      "epoch": 0.8534443740382004,
+      "grad_norm": 0.8762169480323792,
+      "learning_rate": 1.6428529186614195e-06,
+      "loss": 2.6089,
+      "step": 9428
+    },
+    {
+      "epoch": 0.8535348963519508,
+      "grad_norm": 0.8387589454650879,
+      "learning_rate": 1.6371294102249134e-06,
+      "loss": 2.6026,
+      "step": 9429
+    },
+    {
+      "epoch": 0.8536254186657011,
+      "grad_norm": 0.8397504687309265,
+      "learning_rate": 1.6314158069267948e-06,
+      "loss": 2.5771,
+      "step": 9430
+    },
+    {
+      "epoch": 0.8537159409794515,
+      "grad_norm": 0.8257327675819397,
+      "learning_rate": 1.6257121093424033e-06,
+      "loss": 2.5911,
+      "step": 9431
+    },
+    {
+      "epoch": 0.8538064632932018,
+      "grad_norm": 0.836053729057312,
+      "learning_rate": 1.6200183180461236e-06,
+      "loss": 2.6332,
+      "step": 9432
+    },
+    {
+      "epoch": 0.8538969856069522,
+      "grad_norm": 0.8341732621192932,
+      "learning_rate": 1.6143344336112976e-06,
+      "loss": 2.5628,
+      "step": 9433
+    },
+    {
+      "epoch": 0.8539875079207024,
+      "grad_norm": 0.8637101054191589,
+      "learning_rate": 1.6086604566103002e-06,
+      "loss": 2.5738,
+      "step": 9434
+    },
+    {
+      "epoch": 0.8540780302344528,
+      "grad_norm": 0.9812245965003967,
+      "learning_rate": 1.6029963876145083e-06,
+      "loss": 2.6578,
+      "step": 9435
+    },
+    {
+      "epoch": 0.8541685525482031,
+      "grad_norm": 0.8124945163726807,
+      "learning_rate": 1.5973422271942985e-06,
+      "loss": 2.6427,
+      "step": 9436
+    },
+    {
+      "epoch": 0.8542590748619535,
+      "grad_norm": 0.8565968871116638,
+      "learning_rate": 1.5916979759190264e-06,
+      "loss": 2.6265,
+      "step": 9437
+    },
+    {
+      "epoch": 0.8543495971757038,
+      "grad_norm": 0.8052038550376892,
+      "learning_rate": 1.5860636343570711e-06,
+      "loss": 2.6221,
+      "step": 9438
+    },
+    {
+      "epoch": 0.8544401194894542,
+      "grad_norm": 0.9391239881515503,
+      "learning_rate": 1.580439203075812e-06,
+      "loss": 2.64,
+      "step": 9439
+    },
+    {
+      "epoch": 0.8545306418032045,
+      "grad_norm": 0.8188319206237793,
+      "learning_rate": 1.574824682641629e-06,
+      "loss": 2.5393,
+      "step": 9440
+    },
+    {
+      "epoch": 0.8546211641169549,
+      "grad_norm": 0.8291501402854919,
+      "learning_rate": 1.5692200736199148e-06,
+      "loss": 2.6161,
+      "step": 9441
+    },
+    {
+      "epoch": 0.8547116864307052,
+      "grad_norm": 0.9298816919326782,
+      "learning_rate": 1.5636253765750508e-06,
+      "loss": 2.6896,
+      "step": 9442
+    },
+    {
+      "epoch": 0.8548022087444556,
+      "grad_norm": 0.8298754692077637,
+      "learning_rate": 1.5580405920704088e-06,
+      "loss": 2.6438,
+      "step": 9443
+    },
+    {
+      "epoch": 0.8548927310582058,
+      "grad_norm": 0.9584684371948242,
+      "learning_rate": 1.5524657206683947e-06,
+      "loss": 2.6297,
+      "step": 9444
+    },
+    {
+      "epoch": 0.8549832533719562,
+      "grad_norm": 0.8696407675743103,
+      "learning_rate": 1.5469007629303812e-06,
+      "loss": 2.6289,
+      "step": 9445
+    },
+    {
+      "epoch": 0.8550737756857065,
+      "grad_norm": 0.8324605822563171,
+      "learning_rate": 1.5413457194167757e-06,
+      "loss": 2.5645,
+      "step": 9446
+    },
+    {
+      "epoch": 0.8551642979994568,
+      "grad_norm": 0.8564126491546631,
+      "learning_rate": 1.5358005906869532e-06,
+      "loss": 2.6157,
+      "step": 9447
+    },
+    {
+      "epoch": 0.8552548203132072,
+      "grad_norm": 0.8332357406616211,
+      "learning_rate": 1.5302653772993225e-06,
+      "loss": 2.5958,
+      "step": 9448
+    },
+    {
+      "epoch": 0.8553453426269575,
+      "grad_norm": 0.8862999677658081,
+      "learning_rate": 1.5247400798112598e-06,
+      "loss": 2.5999,
+      "step": 9449
+    },
+    {
+      "epoch": 0.8554358649407079,
+      "grad_norm": 0.8509512543678284,
+      "learning_rate": 1.5192246987791981e-06,
+      "loss": 2.6174,
+      "step": 9450
+    },
+    {
+      "epoch": 0.8555263872544582,
+      "grad_norm": 0.7944133877754211,
+      "learning_rate": 1.513719234758504e-06,
+      "loss": 2.564,
+      "step": 9451
+    },
+    {
+      "epoch": 0.8556169095682086,
+      "grad_norm": 0.8499475121498108,
+      "learning_rate": 1.5082236883035895e-06,
+      "loss": 2.6325,
+      "step": 9452
+    },
+    {
+      "epoch": 0.8557074318819589,
+      "grad_norm": 0.8515414595603943,
+      "learning_rate": 1.5027380599678564e-06,
+      "loss": 2.6154,
+      "step": 9453
+    },
+    {
+      "epoch": 0.8557979541957093,
+      "grad_norm": 0.9373891949653625,
+      "learning_rate": 1.4972623503036965e-06,
+      "loss": 2.6027,
+      "step": 9454
+    },
+    {
+      "epoch": 0.8558884765094595,
+      "grad_norm": 0.8476731777191162,
+      "learning_rate": 1.4917965598625349e-06,
+      "loss": 2.6462,
+      "step": 9455
+    },
+    {
+      "epoch": 0.8559789988232099,
+      "grad_norm": 0.9011625051498413,
+      "learning_rate": 1.486340689194754e-06,
+      "loss": 2.6312,
+      "step": 9456
+    },
+    {
+      "epoch": 0.8560695211369602,
+      "grad_norm": 0.8020340800285339,
+      "learning_rate": 1.4808947388497808e-06,
+      "loss": 2.5642,
+      "step": 9457
+    },
+    {
+      "epoch": 0.8561600434507106,
+      "grad_norm": 0.8751192092895508,
+      "learning_rate": 1.4754587093760097e-06,
+      "loss": 2.6344,
+      "step": 9458
+    },
+    {
+      "epoch": 0.8562505657644609,
+      "grad_norm": 0.8079261183738708,
+      "learning_rate": 1.4700326013208365e-06,
+      "loss": 2.6598,
+      "step": 9459
+    },
+    {
+      "epoch": 0.8563410880782113,
+      "grad_norm": 0.8567526340484619,
+      "learning_rate": 1.4646164152307018e-06,
+      "loss": 2.6714,
+      "step": 9460
+    },
+    {
+      "epoch": 0.8564316103919616,
+      "grad_norm": 0.8911241888999939,
+      "learning_rate": 1.4592101516509914e-06,
+      "loss": 2.6807,
+      "step": 9461
+    },
+    {
+      "epoch": 0.856522132705712,
+      "grad_norm": 0.8518723249435425,
+      "learning_rate": 1.4538138111261257e-06,
+      "loss": 2.6163,
+      "step": 9462
+    },
+    {
+      "epoch": 0.8566126550194623,
+      "grad_norm": 0.8393850922584534,
+      "learning_rate": 1.448427394199503e-06,
+      "loss": 2.6404,
+      "step": 9463
+    },
+    {
+      "epoch": 0.8567031773332127,
+      "grad_norm": 0.8219265341758728,
+      "learning_rate": 1.4430509014135563e-06,
+      "loss": 2.6566,
+      "step": 9464
+    },
+    {
+      "epoch": 0.856793699646963,
+      "grad_norm": 0.8629975914955139,
+      "learning_rate": 1.4376843333096747e-06,
+      "loss": 2.6258,
+      "step": 9465
+    },
+    {
+      "epoch": 0.8568842219607133,
+      "grad_norm": 0.8664360046386719,
+      "learning_rate": 1.4323276904283034e-06,
+      "loss": 2.5728,
+      "step": 9466
+    },
+    {
+      "epoch": 0.8569747442744636,
+      "grad_norm": 0.7899994254112244,
+      "learning_rate": 1.4269809733088112e-06,
+      "loss": 2.6007,
+      "step": 9467
+    },
+    {
+      "epoch": 0.857065266588214,
+      "grad_norm": 0.8480337858200073,
+      "learning_rate": 1.4216441824896565e-06,
+      "loss": 2.6015,
+      "step": 9468
+    },
+    {
+      "epoch": 0.8571557889019643,
+      "grad_norm": 0.8864121437072754,
+      "learning_rate": 1.4163173185082312e-06,
+      "loss": 2.5481,
+      "step": 9469
+    },
+    {
+      "epoch": 0.8572463112157147,
+      "grad_norm": 0.8953933119773865,
+      "learning_rate": 1.411000381900951e-06,
+      "loss": 2.6592,
+      "step": 9470
+    },
+    {
+      "epoch": 0.857336833529465,
+      "grad_norm": 0.8673043251037598,
+      "learning_rate": 1.405693373203243e-06,
+      "loss": 2.6351,
+      "step": 9471
+    },
+    {
+      "epoch": 0.8574273558432154,
+      "grad_norm": 0.8806537389755249,
+      "learning_rate": 1.400396292949513e-06,
+      "loss": 2.6091,
+      "step": 9472
+    },
+    {
+      "epoch": 0.8575178781569657,
+      "grad_norm": 0.8178626298904419,
+      "learning_rate": 1.3951091416731676e-06,
+      "loss": 2.6218,
+      "step": 9473
+    },
+    {
+      "epoch": 0.8576084004707161,
+      "grad_norm": 0.8824848532676697,
+      "learning_rate": 1.3898319199066478e-06,
+      "loss": 2.6469,
+      "step": 9474
+    },
+    {
+      "epoch": 0.8576989227844664,
+      "grad_norm": 0.8050047755241394,
+      "learning_rate": 1.3845646281813507e-06,
+      "loss": 2.636,
+      "step": 9475
+    },
+    {
+      "epoch": 0.8577894450982168,
+      "grad_norm": 0.936113178730011,
+      "learning_rate": 1.3793072670276851e-06,
+      "loss": 2.5698,
+      "step": 9476
+    },
+    {
+      "epoch": 0.857879967411967,
+      "grad_norm": 0.8579187393188477,
+      "learning_rate": 1.3740598369750945e-06,
+      "loss": 2.6198,
+      "step": 9477
+    },
+    {
+      "epoch": 0.8579704897257174,
+      "grad_norm": 0.8076322674751282,
+      "learning_rate": 1.3688223385519672e-06,
+      "loss": 2.6443,
+      "step": 9478
+    },
+    {
+      "epoch": 0.8580610120394677,
+      "grad_norm": 0.8612182140350342,
+      "learning_rate": 1.3635947722857367e-06,
+      "loss": 2.6371,
+      "step": 9479
+    },
+    {
+      "epoch": 0.8581515343532181,
+      "grad_norm": 0.8125132918357849,
+      "learning_rate": 1.3583771387028265e-06,
+      "loss": 2.5982,
+      "step": 9480
+    },
+    {
+      "epoch": 0.8582420566669684,
+      "grad_norm": 0.824117124080658,
+      "learning_rate": 1.3531694383286165e-06,
+      "loss": 2.572,
+      "step": 9481
+    },
+    {
+      "epoch": 0.8583325789807188,
+      "grad_norm": 0.8419950008392334,
+      "learning_rate": 1.3479716716875645e-06,
+      "loss": 2.5942,
+      "step": 9482
+    },
+    {
+      "epoch": 0.8584231012944691,
+      "grad_norm": 0.8302912712097168,
+      "learning_rate": 1.3427838393030633e-06,
+      "loss": 2.5714,
+      "step": 9483
+    },
+    {
+      "epoch": 0.8585136236082195,
+      "grad_norm": 0.8524641990661621,
+      "learning_rate": 1.3376059416975172e-06,
+      "loss": 2.6421,
+      "step": 9484
+    },
+    {
+      "epoch": 0.8586041459219698,
+      "grad_norm": 0.8621329069137573,
+      "learning_rate": 1.3324379793923647e-06,
+      "loss": 2.5743,
+      "step": 9485
+    },
+    {
+      "epoch": 0.8586946682357202,
+      "grad_norm": 0.9065933227539062,
+      "learning_rate": 1.3272799529080004e-06,
+      "loss": 2.6904,
+      "step": 9486
+    },
+    {
+      "epoch": 0.8587851905494704,
+      "grad_norm": 0.8736260533332825,
+      "learning_rate": 1.3221318627638534e-06,
+      "loss": 2.5828,
+      "step": 9487
+    },
+    {
+      "epoch": 0.8588757128632207,
+      "grad_norm": 0.8250309824943542,
+      "learning_rate": 1.3169937094783314e-06,
+      "loss": 2.5656,
+      "step": 9488
+    },
+    {
+      "epoch": 0.8589662351769711,
+      "grad_norm": 0.9094919562339783,
+      "learning_rate": 1.3118654935688313e-06,
+      "loss": 2.6435,
+      "step": 9489
+    },
+    {
+      "epoch": 0.8590567574907214,
+      "grad_norm": 0.857205331325531,
+      "learning_rate": 1.3067472155517735e-06,
+      "loss": 2.5497,
+      "step": 9490
+    },
+    {
+      "epoch": 0.8591472798044718,
+      "grad_norm": 0.844194769859314,
+      "learning_rate": 1.301638875942579e-06,
+      "loss": 2.6107,
+      "step": 9491
+    },
+    {
+      "epoch": 0.8592378021182221,
+      "grad_norm": 0.8576357960700989,
+      "learning_rate": 1.2965404752556476e-06,
+      "loss": 2.5993,
+      "step": 9492
+    },
+    {
+      "epoch": 0.8593283244319725,
+      "grad_norm": 0.9494067430496216,
+      "learning_rate": 1.2914520140043906e-06,
+      "loss": 2.6781,
+      "step": 9493
+    },
+    {
+      "epoch": 0.8594188467457228,
+      "grad_norm": 0.8041919469833374,
+      "learning_rate": 1.2863734927012095e-06,
+      "loss": 2.5825,
+      "step": 9494
+    },
+    {
+      "epoch": 0.8595093690594732,
+      "grad_norm": 0.8312264680862427,
+      "learning_rate": 1.2813049118575282e-06,
+      "loss": 2.5765,
+      "step": 9495
+    },
+    {
+      "epoch": 0.8595998913732235,
+      "grad_norm": 0.9145568013191223,
+      "learning_rate": 1.2762462719837275e-06,
+      "loss": 2.7119,
+      "step": 9496
+    },
+    {
+      "epoch": 0.8596904136869739,
+      "grad_norm": 0.8223174810409546,
+      "learning_rate": 1.2711975735892334e-06,
+      "loss": 2.5868,
+      "step": 9497
+    },
+    {
+      "epoch": 0.8597809360007241,
+      "grad_norm": 0.7928852438926697,
+      "learning_rate": 1.266158817182439e-06,
+      "loss": 2.5961,
+      "step": 9498
+    },
+    {
+      "epoch": 0.8598714583144745,
+      "grad_norm": 0.8287642598152161,
+      "learning_rate": 1.2611300032707496e-06,
+      "loss": 2.6338,
+      "step": 9499
+    },
+    {
+      "epoch": 0.8599619806282248,
+      "grad_norm": 0.8964071869850159,
+      "learning_rate": 1.2561111323605712e-06,
+      "loss": 2.5713,
+      "step": 9500
+    },
+    {
+      "epoch": 0.8600525029419752,
+      "grad_norm": 0.7847150564193726,
+      "learning_rate": 1.2511022049572996e-06,
+      "loss": 2.6324,
+      "step": 9501
+    },
+    {
+      "epoch": 0.8601430252557255,
+      "grad_norm": 0.8039739727973938,
+      "learning_rate": 1.2461032215653311e-06,
+      "loss": 2.5915,
+      "step": 9502
+    },
+    {
+      "epoch": 0.8602335475694759,
+      "grad_norm": 0.8426544666290283,
+      "learning_rate": 1.2411141826880636e-06,
+      "loss": 2.6034,
+      "step": 9503
+    },
+    {
+      "epoch": 0.8603240698832262,
+      "grad_norm": 0.8430858254432678,
+      "learning_rate": 1.2361350888279056e-06,
+      "loss": 2.543,
+      "step": 9504
+    },
+    {
+      "epoch": 0.8604145921969766,
+      "grad_norm": 0.8529922366142273,
+      "learning_rate": 1.231165940486234e-06,
+      "loss": 2.6163,
+      "step": 9505
+    },
+    {
+      "epoch": 0.8605051145107269,
+      "grad_norm": 0.8032079339027405,
+      "learning_rate": 1.226206738163449e-06,
+      "loss": 2.5339,
+      "step": 9506
+    },
+    {
+      "epoch": 0.8605956368244773,
+      "grad_norm": 0.9089851379394531,
+      "learning_rate": 1.2212574823589507e-06,
+      "loss": 2.6063,
+      "step": 9507
+    },
+    {
+      "epoch": 0.8606861591382275,
+      "grad_norm": 0.7995228171348572,
+      "learning_rate": 1.2163181735711072e-06,
+      "loss": 2.6312,
+      "step": 9508
+    },
+    {
+      "epoch": 0.8607766814519779,
+      "grad_norm": 0.8813963532447815,
+      "learning_rate": 1.2113888122973317e-06,
+      "loss": 2.6057,
+      "step": 9509
+    },
+    {
+      "epoch": 0.8608672037657282,
+      "grad_norm": 0.8219525814056396,
+      "learning_rate": 1.2064693990339936e-06,
+      "loss": 2.6038,
+      "step": 9510
+    },
+    {
+      "epoch": 0.8609577260794786,
+      "grad_norm": 0.8565325140953064,
+      "learning_rate": 1.2015599342764862e-06,
+      "loss": 2.623,
+      "step": 9511
+    },
+    {
+      "epoch": 0.8610482483932289,
+      "grad_norm": 0.8099278211593628,
+      "learning_rate": 1.196660418519191e-06,
+      "loss": 2.6101,
+      "step": 9512
+    },
+    {
+      "epoch": 0.8611387707069793,
+      "grad_norm": 0.9025592803955078,
+      "learning_rate": 1.1917708522554915e-06,
+      "loss": 2.6319,
+      "step": 9513
+    },
+    {
+      "epoch": 0.8612292930207296,
+      "grad_norm": 0.7932490706443787,
+      "learning_rate": 1.1868912359777607e-06,
+      "loss": 2.6129,
+      "step": 9514
+    },
+    {
+      "epoch": 0.86131981533448,
+      "grad_norm": 0.8897316455841064,
+      "learning_rate": 1.1820215701773829e-06,
+      "loss": 2.6638,
+      "step": 9515
+    },
+    {
+      "epoch": 0.8614103376482303,
+      "grad_norm": 0.8247643709182739,
+      "learning_rate": 1.1771618553447216e-06,
+      "loss": 2.6295,
+      "step": 9516
+    },
+    {
+      "epoch": 0.8615008599619807,
+      "grad_norm": 0.8464657068252563,
+      "learning_rate": 1.1723120919691634e-06,
+      "loss": 2.631,
+      "step": 9517
+    },
+    {
+      "epoch": 0.861591382275731,
+      "grad_norm": 0.8722254037857056,
+      "learning_rate": 1.1674722805390726e-06,
+      "loss": 2.6014,
+      "step": 9518
+    },
+    {
+      "epoch": 0.8616819045894814,
+      "grad_norm": 0.9596865177154541,
+      "learning_rate": 1.1626424215418151e-06,
+      "loss": 2.6208,
+      "step": 9519
+    },
+    {
+      "epoch": 0.8617724269032316,
+      "grad_norm": 0.816199779510498,
+      "learning_rate": 1.157822515463758e-06,
+      "loss": 2.5962,
+      "step": 9520
+    },
+    {
+      "epoch": 0.861862949216982,
+      "grad_norm": 0.8612105250358582,
+      "learning_rate": 1.1530125627902788e-06,
+      "loss": 2.6487,
+      "step": 9521
+    },
+    {
+      "epoch": 0.8619534715307323,
+      "grad_norm": 0.9207258820533752,
+      "learning_rate": 1.148212564005724e-06,
+      "loss": 2.6646,
+      "step": 9522
+    },
+    {
+      "epoch": 0.8620439938444827,
+      "grad_norm": 0.8424425721168518,
+      "learning_rate": 1.1434225195934622e-06,
+      "loss": 2.6143,
+      "step": 9523
+    },
+    {
+      "epoch": 0.862134516158233,
+      "grad_norm": 0.8309099078178406,
+      "learning_rate": 1.138642430035841e-06,
+      "loss": 2.6148,
+      "step": 9524
+    },
+    {
+      "epoch": 0.8622250384719834,
+      "grad_norm": 0.839776873588562,
+      "learning_rate": 1.1338722958142312e-06,
+      "loss": 2.6355,
+      "step": 9525
+    },
+    {
+      "epoch": 0.8623155607857337,
+      "grad_norm": 0.8244553208351135,
+      "learning_rate": 1.1291121174089703e-06,
+      "loss": 2.6388,
+      "step": 9526
+    },
+    {
+      "epoch": 0.8624060830994841,
+      "grad_norm": 0.8162111639976501,
+      "learning_rate": 1.1243618952994195e-06,
+      "loss": 2.5933,
+      "step": 9527
+    },
+    {
+      "epoch": 0.8624966054132344,
+      "grad_norm": 0.7930101156234741,
+      "learning_rate": 1.1196216299639072e-06,
+      "loss": 2.5523,
+      "step": 9528
+    },
+    {
+      "epoch": 0.8625871277269846,
+      "grad_norm": 0.8353174328804016,
+      "learning_rate": 1.1148913218798075e-06,
+      "loss": 2.6007,
+      "step": 9529
+    },
+    {
+      "epoch": 0.862677650040735,
+      "grad_norm": 0.8778246641159058,
+      "learning_rate": 1.1101709715234386e-06,
+      "loss": 2.602,
+      "step": 9530
+    },
+    {
+      "epoch": 0.8627681723544853,
+      "grad_norm": 0.8321190476417542,
+      "learning_rate": 1.1054605793701545e-06,
+      "loss": 2.6676,
+      "step": 9531
+    },
+    {
+      "epoch": 0.8628586946682357,
+      "grad_norm": 0.7573698163032532,
+      "learning_rate": 1.1007601458942752e-06,
+      "loss": 2.563,
+      "step": 9532
+    },
+    {
+      "epoch": 0.862949216981986,
+      "grad_norm": 0.8202477097511292,
+      "learning_rate": 1.0960696715691443e-06,
+      "loss": 2.5636,
+      "step": 9533
+    },
+    {
+      "epoch": 0.8630397392957364,
+      "grad_norm": 0.7851460576057434,
+      "learning_rate": 1.0913891568670842e-06,
+      "loss": 2.6107,
+      "step": 9534
+    },
+    {
+      "epoch": 0.8631302616094867,
+      "grad_norm": 0.8061966896057129,
+      "learning_rate": 1.08671860225944e-06,
+      "loss": 2.5635,
+      "step": 9535
+    },
+    {
+      "epoch": 0.8632207839232371,
+      "grad_norm": 0.8170107007026672,
+      "learning_rate": 1.0820580082165244e-06,
+      "loss": 2.577,
+      "step": 9536
+    },
+    {
+      "epoch": 0.8633113062369874,
+      "grad_norm": 0.8730764985084534,
+      "learning_rate": 1.0774073752076618e-06,
+      "loss": 2.6012,
+      "step": 9537
+    },
+    {
+      "epoch": 0.8634018285507378,
+      "grad_norm": 0.8004407286643982,
+      "learning_rate": 1.0727667037011668e-06,
+      "loss": 2.6438,
+      "step": 9538
+    },
+    {
+      "epoch": 0.8634923508644881,
+      "grad_norm": 0.8690145611763,
+      "learning_rate": 1.0681359941643654e-06,
+      "loss": 2.6304,
+      "step": 9539
+    },
+    {
+      "epoch": 0.8635828731782385,
+      "grad_norm": 0.8464412689208984,
+      "learning_rate": 1.0635152470635512e-06,
+      "loss": 2.6322,
+      "step": 9540
+    },
+    {
+      "epoch": 0.8636733954919887,
+      "grad_norm": 0.86949622631073,
+      "learning_rate": 1.0589044628640522e-06,
+      "loss": 2.6291,
+      "step": 9541
+    },
+    {
+      "epoch": 0.8637639178057391,
+      "grad_norm": 0.8569595217704773,
+      "learning_rate": 1.0543036420301634e-06,
+      "loss": 2.5688,
+      "step": 9542
+    },
+    {
+      "epoch": 0.8638544401194894,
+      "grad_norm": 0.817255437374115,
+      "learning_rate": 1.049712785025192e-06,
+      "loss": 2.6486,
+      "step": 9543
+    },
+    {
+      "epoch": 0.8639449624332398,
+      "grad_norm": 0.8158665299415588,
+      "learning_rate": 1.045131892311435e-06,
+      "loss": 2.579,
+      "step": 9544
+    },
+    {
+      "epoch": 0.8640354847469901,
+      "grad_norm": 0.783174455165863,
+      "learning_rate": 1.0405609643501902e-06,
+      "loss": 2.5935,
+      "step": 9545
+    },
+    {
+      "epoch": 0.8641260070607405,
+      "grad_norm": 0.7993530631065369,
+      "learning_rate": 1.0360000016017445e-06,
+      "loss": 2.5839,
+      "step": 9546
+    },
+    {
+      "epoch": 0.8642165293744908,
+      "grad_norm": 0.8038840889930725,
+      "learning_rate": 1.0314490045253977e-06,
+      "loss": 2.5646,
+      "step": 9547
+    },
+    {
+      "epoch": 0.8643070516882412,
+      "grad_norm": 0.7852585315704346,
+      "learning_rate": 1.0269079735794164e-06,
+      "loss": 2.5571,
+      "step": 9548
+    },
+    {
+      "epoch": 0.8643975740019915,
+      "grad_norm": 0.8642681837081909,
+      "learning_rate": 1.0223769092211012e-06,
+      "loss": 2.6293,
+      "step": 9549
+    },
+    {
+      "epoch": 0.8644880963157419,
+      "grad_norm": 0.820342481136322,
+      "learning_rate": 1.0178558119067315e-06,
+      "loss": 2.59,
+      "step": 9550
+    },
+    {
+      "epoch": 0.8645786186294921,
+      "grad_norm": 0.8889754414558411,
+      "learning_rate": 1.013344682091555e-06,
+      "loss": 2.7136,
+      "step": 9551
+    },
+    {
+      "epoch": 0.8646691409432425,
+      "grad_norm": 0.8811084032058716,
+      "learning_rate": 1.008843520229874e-06,
+      "loss": 2.659,
+      "step": 9552
+    },
+    {
+      "epoch": 0.8647596632569928,
+      "grad_norm": 0.7807613015174866,
+      "learning_rate": 1.0043523267749489e-06,
+      "loss": 2.5764,
+      "step": 9553
+    },
+    {
+      "epoch": 0.8648501855707432,
+      "grad_norm": 0.8308903574943542,
+      "learning_rate": 9.998711021790174e-07,
+      "loss": 2.6319,
+      "step": 9554
+    },
+    {
+      "epoch": 0.8649407078844935,
+      "grad_norm": 0.8310360312461853,
+      "learning_rate": 9.953998468933634e-07,
+      "loss": 2.6007,
+      "step": 9555
+    },
+    {
+      "epoch": 0.8650312301982439,
+      "grad_norm": 0.7842338681221008,
+      "learning_rate": 9.909385613682375e-07,
+      "loss": 2.6078,
+      "step": 9556
+    },
+    {
+      "epoch": 0.8651217525119942,
+      "grad_norm": 0.8118513226509094,
+      "learning_rate": 9.864872460528918e-07,
+      "loss": 2.5622,
+      "step": 9557
+    },
+    {
+      "epoch": 0.8652122748257446,
+      "grad_norm": 0.7811709642410278,
+      "learning_rate": 9.820459013955674e-07,
+      "loss": 2.5342,
+      "step": 9558
+    },
+    {
+      "epoch": 0.8653027971394949,
+      "grad_norm": 0.8299935460090637,
+      "learning_rate": 9.776145278435178e-07,
+      "loss": 2.5751,
+      "step": 9559
+    },
+    {
+      "epoch": 0.8653933194532453,
+      "grad_norm": 0.8128529191017151,
+      "learning_rate": 9.731931258429638e-07,
+      "loss": 2.6198,
+      "step": 9560
+    },
+    {
+      "epoch": 0.8654838417669956,
+      "grad_norm": 0.8885383009910583,
+      "learning_rate": 9.687816958391716e-07,
+      "loss": 2.5648,
+      "step": 9561
+    },
+    {
+      "epoch": 0.865574364080746,
+      "grad_norm": 0.8305163383483887,
+      "learning_rate": 9.6438023827633e-07,
+      "loss": 2.586,
+      "step": 9562
+    },
+    {
+      "epoch": 0.8656648863944962,
+      "grad_norm": 0.8077706098556519,
+      "learning_rate": 9.59988753597707e-07,
+      "loss": 2.5863,
+      "step": 9563
+    },
+    {
+      "epoch": 0.8657554087082466,
+      "grad_norm": 0.8040973544120789,
+      "learning_rate": 9.55607242245493e-07,
+      "loss": 2.6233,
+      "step": 9564
+    },
+    {
+      "epoch": 0.8658459310219969,
+      "grad_norm": 0.8089603781700134,
+      "learning_rate": 9.512357046609244e-07,
+      "loss": 2.6533,
+      "step": 9565
+    },
+    {
+      "epoch": 0.8659364533357473,
+      "grad_norm": 0.7722705006599426,
+      "learning_rate": 9.468741412842152e-07,
+      "loss": 2.5909,
+      "step": 9566
+    },
+    {
+      "epoch": 0.8660269756494976,
+      "grad_norm": 0.8577929735183716,
+      "learning_rate": 9.4252255255457e-07,
+      "loss": 2.5763,
+      "step": 9567
+    },
+    {
+      "epoch": 0.866117497963248,
+      "grad_norm": 0.848434567451477,
+      "learning_rate": 9.381809389101825e-07,
+      "loss": 2.6147,
+      "step": 9568
+    },
+    {
+      "epoch": 0.8662080202769983,
+      "grad_norm": 0.8390416502952576,
+      "learning_rate": 9.338493007882809e-07,
+      "loss": 2.6285,
+      "step": 9569
+    },
+    {
+      "epoch": 0.8662985425907486,
+      "grad_norm": 0.8751125931739807,
+      "learning_rate": 9.295276386250274e-07,
+      "loss": 2.6566,
+      "step": 9570
+    },
+    {
+      "epoch": 0.866389064904499,
+      "grad_norm": 0.8329317569732666,
+      "learning_rate": 9.252159528556403e-07,
+      "loss": 2.6553,
+      "step": 9571
+    },
+    {
+      "epoch": 0.8664795872182492,
+      "grad_norm": 0.9101200103759766,
+      "learning_rate": 9.209142439142948e-07,
+      "loss": 2.6609,
+      "step": 9572
+    },
+    {
+      "epoch": 0.8665701095319996,
+      "grad_norm": 0.8253322243690491,
+      "learning_rate": 9.166225122341665e-07,
+      "loss": 2.606,
+      "step": 9573
+    },
+    {
+      "epoch": 0.8666606318457499,
+      "grad_norm": 0.8649923205375671,
+      "learning_rate": 9.123407582474541e-07,
+      "loss": 2.6011,
+      "step": 9574
+    },
+    {
+      "epoch": 0.8667511541595003,
+      "grad_norm": 0.9202775955200195,
+      "learning_rate": 9.080689823853017e-07,
+      "loss": 2.6167,
+      "step": 9575
+    },
+    {
+      "epoch": 0.8668416764732506,
+      "grad_norm": 0.8605530261993408,
+      "learning_rate": 9.038071850778984e-07,
+      "loss": 2.6093,
+      "step": 9576
+    },
+    {
+      "epoch": 0.866932198787001,
+      "grad_norm": 0.8165022730827332,
+      "learning_rate": 8.99555366754401e-07,
+      "loss": 2.5998,
+      "step": 9577
+    },
+    {
+      "epoch": 0.8670227211007513,
+      "grad_norm": 0.832669198513031,
+      "learning_rate": 8.953135278429558e-07,
+      "loss": 2.6118,
+      "step": 9578
+    },
+    {
+      "epoch": 0.8671132434145017,
+      "grad_norm": 0.8316287398338318,
+      "learning_rate": 8.910816687707435e-07,
+      "loss": 2.6511,
+      "step": 9579
+    },
+    {
+      "epoch": 0.867203765728252,
+      "grad_norm": 0.8216667771339417,
+      "learning_rate": 8.868597899638898e-07,
+      "loss": 2.6116,
+      "step": 9580
+    },
+    {
+      "epoch": 0.8672942880420024,
+      "grad_norm": 0.8596590757369995,
+      "learning_rate": 8.826478918475323e-07,
+      "loss": 2.6065,
+      "step": 9581
+    },
+    {
+      "epoch": 0.8673848103557527,
+      "grad_norm": 0.8234314918518066,
+      "learning_rate": 8.784459748458318e-07,
+      "loss": 2.6449,
+      "step": 9582
+    },
+    {
+      "epoch": 0.867475332669503,
+      "grad_norm": 0.8860653042793274,
+      "learning_rate": 8.742540393819054e-07,
+      "loss": 2.6161,
+      "step": 9583
+    },
+    {
+      "epoch": 0.8675658549832533,
+      "grad_norm": 0.8453433513641357,
+      "learning_rate": 8.700720858778821e-07,
+      "loss": 2.633,
+      "step": 9584
+    },
+    {
+      "epoch": 0.8676563772970037,
+      "grad_norm": 0.8935437202453613,
+      "learning_rate": 8.659001147548918e-07,
+      "loss": 2.6248,
+      "step": 9585
+    },
+    {
+      "epoch": 0.867746899610754,
+      "grad_norm": 0.8853086233139038,
+      "learning_rate": 8.617381264330426e-07,
+      "loss": 2.7061,
+      "step": 9586
+    },
+    {
+      "epoch": 0.8678374219245044,
+      "grad_norm": 0.8435954451560974,
+      "learning_rate": 8.575861213314551e-07,
+      "loss": 2.6205,
+      "step": 9587
+    },
+    {
+      "epoch": 0.8679279442382547,
+      "grad_norm": 0.8238980770111084,
+      "learning_rate": 8.534440998682391e-07,
+      "loss": 2.5941,
+      "step": 9588
+    },
+    {
+      "epoch": 0.8680184665520051,
+      "grad_norm": 0.784957766532898,
+      "learning_rate": 8.493120624604833e-07,
+      "loss": 2.5959,
+      "step": 9589
+    },
+    {
+      "epoch": 0.8681089888657554,
+      "grad_norm": 1.014724612236023,
+      "learning_rate": 8.451900095242881e-07,
+      "loss": 2.6458,
+      "step": 9590
+    },
+    {
+      "epoch": 0.8681995111795058,
+      "grad_norm": 0.7878062725067139,
+      "learning_rate": 8.410779414747439e-07,
+      "loss": 2.6508,
+      "step": 9591
+    },
+    {
+      "epoch": 0.8682900334932561,
+      "grad_norm": 0.9261922836303711,
+      "learning_rate": 8.369758587259413e-07,
+      "loss": 2.6285,
+      "step": 9592
+    },
+    {
+      "epoch": 0.8683805558070065,
+      "grad_norm": 0.8362093567848206,
+      "learning_rate": 8.328837616909613e-07,
+      "loss": 2.5912,
+      "step": 9593
+    },
+    {
+      "epoch": 0.8684710781207567,
+      "grad_norm": 0.8119001984596252,
+      "learning_rate": 8.288016507818742e-07,
+      "loss": 2.6457,
+      "step": 9594
+    },
+    {
+      "epoch": 0.8685616004345071,
+      "grad_norm": 0.7739616632461548,
+      "learning_rate": 8.247295264097288e-07,
+      "loss": 2.6102,
+      "step": 9595
+    },
+    {
+      "epoch": 0.8686521227482574,
+      "grad_norm": 0.8285651206970215,
+      "learning_rate": 8.206673889846306e-07,
+      "loss": 2.6539,
+      "step": 9596
+    },
+    {
+      "epoch": 0.8687426450620078,
+      "grad_norm": 0.8595246076583862,
+      "learning_rate": 8.166152389155967e-07,
+      "loss": 2.5955,
+      "step": 9597
+    },
+    {
+      "epoch": 0.8688331673757581,
+      "grad_norm": 0.8347133994102478,
+      "learning_rate": 8.125730766107009e-07,
+      "loss": 2.645,
+      "step": 9598
+    },
+    {
+      "epoch": 0.8689236896895085,
+      "grad_norm": 0.8522047996520996,
+      "learning_rate": 8.08540902476973e-07,
+      "loss": 2.591,
+      "step": 9599
+    },
+    {
+      "epoch": 0.8690142120032588,
+      "grad_norm": 0.8081803917884827,
+      "learning_rate": 8.04518716920466e-07,
+      "loss": 2.5815,
+      "step": 9600
+    },
+    {
+      "epoch": 0.8691047343170092,
+      "grad_norm": 0.8424069881439209,
+      "learning_rate": 8.005065203462225e-07,
+      "loss": 2.6368,
+      "step": 9601
+    },
+    {
+      "epoch": 0.8691952566307595,
+      "grad_norm": 0.8203180432319641,
+      "learning_rate": 7.965043131582529e-07,
+      "loss": 2.5874,
+      "step": 9602
+    },
+    {
+      "epoch": 0.8692857789445099,
+      "grad_norm": 0.7955147624015808,
+      "learning_rate": 7.925120957595678e-07,
+      "loss": 2.6271,
+      "step": 9603
+    },
+    {
+      "epoch": 0.8693763012582602,
+      "grad_norm": 0.8208197355270386,
+      "learning_rate": 7.885298685522235e-07,
+      "loss": 2.5695,
+      "step": 9604
+    },
+    {
+      "epoch": 0.8694668235720105,
+      "grad_norm": 0.8572072982788086,
+      "learning_rate": 7.845576319371883e-07,
+      "loss": 2.5843,
+      "step": 9605
+    },
+    {
+      "epoch": 0.8695573458857608,
+      "grad_norm": 0.9009573459625244,
+      "learning_rate": 7.805953863144977e-07,
+      "loss": 2.5855,
+      "step": 9606
+    },
+    {
+      "epoch": 0.8696478681995112,
+      "grad_norm": 0.8166123032569885,
+      "learning_rate": 7.766431320831324e-07,
+      "loss": 2.6133,
+      "step": 9607
+    },
+    {
+      "epoch": 0.8697383905132615,
+      "grad_norm": 0.831098198890686,
+      "learning_rate": 7.727008696410854e-07,
+      "loss": 2.6089,
+      "step": 9608
+    },
+    {
+      "epoch": 0.8698289128270119,
+      "grad_norm": 0.8021360039710999,
+      "learning_rate": 7.6876859938535e-07,
+      "loss": 2.6223,
+      "step": 9609
+    },
+    {
+      "epoch": 0.8699194351407622,
+      "grad_norm": 0.8092429041862488,
+      "learning_rate": 7.648463217118984e-07,
+      "loss": 2.5619,
+      "step": 9610
+    },
+    {
+      "epoch": 0.8700099574545125,
+      "grad_norm": 0.8074584603309631,
+      "learning_rate": 7.609340370157037e-07,
+      "loss": 2.6306,
+      "step": 9611
+    },
+    {
+      "epoch": 0.8701004797682629,
+      "grad_norm": 0.898284912109375,
+      "learning_rate": 7.570317456907506e-07,
+      "loss": 2.6498,
+      "step": 9612
+    },
+    {
+      "epoch": 0.8701910020820132,
+      "grad_norm": 0.8530883193016052,
+      "learning_rate": 7.531394481299691e-07,
+      "loss": 2.6105,
+      "step": 9613
+    },
+    {
+      "epoch": 0.8702815243957636,
+      "grad_norm": 0.8278710842132568,
+      "learning_rate": 7.49257144725346e-07,
+      "loss": 2.6035,
+      "step": 9614
+    },
+    {
+      "epoch": 0.8703720467095138,
+      "grad_norm": 0.8579922914505005,
+      "learning_rate": 7.453848358678017e-07,
+      "loss": 2.6313,
+      "step": 9615
+    },
+    {
+      "epoch": 0.8704625690232642,
+      "grad_norm": 0.831552267074585,
+      "learning_rate": 7.415225219472799e-07,
+      "loss": 2.6281,
+      "step": 9616
+    },
+    {
+      "epoch": 0.8705530913370145,
+      "grad_norm": 0.8504630327224731,
+      "learning_rate": 7.37670203352736e-07,
+      "loss": 2.6271,
+      "step": 9617
+    },
+    {
+      "epoch": 0.8706436136507649,
+      "grad_norm": 0.8295220732688904,
+      "learning_rate": 7.338278804720933e-07,
+      "loss": 2.583,
+      "step": 9618
+    },
+    {
+      "epoch": 0.8707341359645152,
+      "grad_norm": 0.8532193303108215,
+      "learning_rate": 7.299955536922531e-07,
+      "loss": 2.6143,
+      "step": 9619
+    },
+    {
+      "epoch": 0.8708246582782656,
+      "grad_norm": 0.8124157786369324,
+      "learning_rate": 7.261732233991513e-07,
+      "loss": 2.6051,
+      "step": 9620
+    },
+    {
+      "epoch": 0.8709151805920159,
+      "grad_norm": 0.8071900010108948,
+      "learning_rate": 7.223608899776912e-07,
+      "loss": 2.5306,
+      "step": 9621
+    },
+    {
+      "epoch": 0.8710057029057663,
+      "grad_norm": 0.8964101672172546,
+      "learning_rate": 7.185585538117657e-07,
+      "loss": 2.5625,
+      "step": 9622
+    },
+    {
+      "epoch": 0.8710962252195166,
+      "grad_norm": 0.8295537829399109,
+      "learning_rate": 7.147662152842794e-07,
+      "loss": 2.6173,
+      "step": 9623
+    },
+    {
+      "epoch": 0.871186747533267,
+      "grad_norm": 0.835732638835907,
+      "learning_rate": 7.109838747771269e-07,
+      "loss": 2.6326,
+      "step": 9624
+    },
+    {
+      "epoch": 0.8712772698470173,
+      "grad_norm": 0.8687270879745483,
+      "learning_rate": 7.072115326711704e-07,
+      "loss": 2.5738,
+      "step": 9625
+    },
+    {
+      "epoch": 0.8713677921607677,
+      "grad_norm": 0.8733930587768555,
+      "learning_rate": 7.034491893463058e-07,
+      "loss": 2.6143,
+      "step": 9626
+    },
+    {
+      "epoch": 0.8714583144745179,
+      "grad_norm": 0.8681638240814209,
+      "learning_rate": 6.996968451813857e-07,
+      "loss": 2.6106,
+      "step": 9627
+    },
+    {
+      "epoch": 0.8715488367882683,
+      "grad_norm": 0.8479701280593872,
+      "learning_rate": 6.959545005542744e-07,
+      "loss": 2.6332,
+      "step": 9628
+    },
+    {
+      "epoch": 0.8716393591020186,
+      "grad_norm": 0.7846943140029907,
+      "learning_rate": 6.922221558418263e-07,
+      "loss": 2.6283,
+      "step": 9629
+    },
+    {
+      "epoch": 0.871729881415769,
+      "grad_norm": 0.8460354208946228,
+      "learning_rate": 6.884998114198959e-07,
+      "loss": 2.6348,
+      "step": 9630
+    },
+    {
+      "epoch": 0.8718204037295193,
+      "grad_norm": 0.9053749442100525,
+      "learning_rate": 6.847874676633171e-07,
+      "loss": 2.6853,
+      "step": 9631
+    },
+    {
+      "epoch": 0.8719109260432697,
+      "grad_norm": 0.8776060938835144,
+      "learning_rate": 6.810851249459238e-07,
+      "loss": 2.6069,
+      "step": 9632
+    },
+    {
+      "epoch": 0.87200144835702,
+      "grad_norm": 0.843860924243927,
+      "learning_rate": 6.773927836405291e-07,
+      "loss": 2.6399,
+      "step": 9633
+    },
+    {
+      "epoch": 0.8720919706707704,
+      "grad_norm": 0.8224048614501953,
+      "learning_rate": 6.737104441189801e-07,
+      "loss": 2.6078,
+      "step": 9634
+    },
+    {
+      "epoch": 0.8721824929845207,
+      "grad_norm": 0.8930675983428955,
+      "learning_rate": 6.700381067520578e-07,
+      "loss": 2.6118,
+      "step": 9635
+    },
+    {
+      "epoch": 0.8722730152982711,
+      "grad_norm": 0.8419080376625061,
+      "learning_rate": 6.663757719095886e-07,
+      "loss": 2.5938,
+      "step": 9636
+    },
+    {
+      "epoch": 0.8723635376120213,
+      "grad_norm": 0.8541986346244812,
+      "learning_rate": 6.627234399603555e-07,
+      "loss": 2.6475,
+      "step": 9637
+    },
+    {
+      "epoch": 0.8724540599257717,
+      "grad_norm": 0.8377965688705444,
+      "learning_rate": 6.59081111272164e-07,
+      "loss": 2.6409,
+      "step": 9638
+    },
+    {
+      "epoch": 0.872544582239522,
+      "grad_norm": 0.8927517533302307,
+      "learning_rate": 6.554487862117764e-07,
+      "loss": 2.6301,
+      "step": 9639
+    },
+    {
+      "epoch": 0.8726351045532724,
+      "grad_norm": 0.8720734119415283,
+      "learning_rate": 6.518264651449779e-07,
+      "loss": 2.6155,
+      "step": 9640
+    },
+    {
+      "epoch": 0.8727256268670227,
+      "grad_norm": 0.8335932493209839,
+      "learning_rate": 6.482141484365434e-07,
+      "loss": 2.6052,
+      "step": 9641
+    },
+    {
+      "epoch": 0.8728161491807731,
+      "grad_norm": 0.7936855554580688,
+      "learning_rate": 6.446118364502152e-07,
+      "loss": 2.6495,
+      "step": 9642
+    },
+    {
+      "epoch": 0.8729066714945234,
+      "grad_norm": 0.8104913234710693,
+      "learning_rate": 6.410195295487698e-07,
+      "loss": 2.571,
+      "step": 9643
+    },
+    {
+      "epoch": 0.8729971938082738,
+      "grad_norm": 0.8593425154685974,
+      "learning_rate": 6.374372280939289e-07,
+      "loss": 2.5517,
+      "step": 9644
+    },
+    {
+      "epoch": 0.8730877161220241,
+      "grad_norm": 0.8080833554267883,
+      "learning_rate": 6.338649324464374e-07,
+      "loss": 2.617,
+      "step": 9645
+    },
+    {
+      "epoch": 0.8731782384357745,
+      "grad_norm": 0.8711926341056824,
+      "learning_rate": 6.303026429660408e-07,
+      "loss": 2.627,
+      "step": 9646
+    },
+    {
+      "epoch": 0.8732687607495248,
+      "grad_norm": 0.8949486017227173,
+      "learning_rate": 6.267503600114411e-07,
+      "loss": 2.6759,
+      "step": 9647
+    },
+    {
+      "epoch": 0.8733592830632751,
+      "grad_norm": 0.8196110129356384,
+      "learning_rate": 6.232080839403631e-07,
+      "loss": 2.6175,
+      "step": 9648
+    },
+    {
+      "epoch": 0.8734498053770254,
+      "grad_norm": 0.8647054433822632,
+      "learning_rate": 6.196758151095106e-07,
+      "loss": 2.6176,
+      "step": 9649
+    },
+    {
+      "epoch": 0.8735403276907758,
+      "grad_norm": 0.7463350296020508,
+      "learning_rate": 6.161535538745878e-07,
+      "loss": 2.586,
+      "step": 9650
+    },
+    {
+      "epoch": 0.8736308500045261,
+      "grad_norm": 0.8691142797470093,
+      "learning_rate": 6.126413005902775e-07,
+      "loss": 2.6337,
+      "step": 9651
+    },
+    {
+      "epoch": 0.8737213723182764,
+      "grad_norm": 0.8492142558097839,
+      "learning_rate": 6.09139055610275e-07,
+      "loss": 2.5886,
+      "step": 9652
+    },
+    {
+      "epoch": 0.8738118946320268,
+      "grad_norm": 0.8391124606132507,
+      "learning_rate": 6.056468192872422e-07,
+      "loss": 2.6409,
+      "step": 9653
+    },
+    {
+      "epoch": 0.8739024169457771,
+      "grad_norm": 0.8021348714828491,
+      "learning_rate": 6.021645919728647e-07,
+      "loss": 2.5925,
+      "step": 9654
+    },
+    {
+      "epoch": 0.8739929392595275,
+      "grad_norm": 0.7921338081359863,
+      "learning_rate": 5.986923740177842e-07,
+      "loss": 2.609,
+      "step": 9655
+    },
+    {
+      "epoch": 0.8740834615732778,
+      "grad_norm": 0.8511329889297485,
+      "learning_rate": 5.952301657716652e-07,
+      "loss": 2.6105,
+      "step": 9656
+    },
+    {
+      "epoch": 0.8741739838870282,
+      "grad_norm": 0.8938495516777039,
+      "learning_rate": 5.917779675831403e-07,
+      "loss": 2.5969,
+      "step": 9657
+    },
+    {
+      "epoch": 0.8742645062007784,
+      "grad_norm": 0.9117132425308228,
+      "learning_rate": 5.883357797998757e-07,
+      "loss": 2.6351,
+      "step": 9658
+    },
+    {
+      "epoch": 0.8743550285145288,
+      "grad_norm": 0.8959320187568665,
+      "learning_rate": 5.849036027684606e-07,
+      "loss": 2.5781,
+      "step": 9659
+    },
+    {
+      "epoch": 0.8744455508282791,
+      "grad_norm": 0.8768128156661987,
+      "learning_rate": 5.814814368345412e-07,
+      "loss": 2.6333,
+      "step": 9660
+    },
+    {
+      "epoch": 0.8745360731420295,
+      "grad_norm": 0.809253454208374,
+      "learning_rate": 5.780692823427191e-07,
+      "loss": 2.5255,
+      "step": 9661
+    },
+    {
+      "epoch": 0.8746265954557798,
+      "grad_norm": 0.9544652104377747,
+      "learning_rate": 5.746671396365977e-07,
+      "loss": 2.6322,
+      "step": 9662
+    },
+    {
+      "epoch": 0.8747171177695302,
+      "grad_norm": 0.8581130504608154,
+      "learning_rate": 5.712750090587693e-07,
+      "loss": 2.6322,
+      "step": 9663
+    },
+    {
+      "epoch": 0.8748076400832805,
+      "grad_norm": 0.8829495906829834,
+      "learning_rate": 5.678928909508275e-07,
+      "loss": 2.586,
+      "step": 9664
+    },
+    {
+      "epoch": 0.8748981623970309,
+      "grad_norm": 0.7909978032112122,
+      "learning_rate": 5.645207856533552e-07,
+      "loss": 2.6083,
+      "step": 9665
+    },
+    {
+      "epoch": 0.8749886847107812,
+      "grad_norm": 0.861017644405365,
+      "learning_rate": 5.611586935059255e-07,
+      "loss": 2.6235,
+      "step": 9666
+    },
+    {
+      "epoch": 0.8750792070245316,
+      "grad_norm": 0.8640226125717163,
+      "learning_rate": 5.578066148470895e-07,
+      "loss": 2.5969,
+      "step": 9667
+    },
+    {
+      "epoch": 0.8751697293382819,
+      "grad_norm": 0.8930851221084595,
+      "learning_rate": 5.544645500143997e-07,
+      "loss": 2.5674,
+      "step": 9668
+    },
+    {
+      "epoch": 0.8752602516520323,
+      "grad_norm": 0.8401358127593994,
+      "learning_rate": 5.511324993444201e-07,
+      "loss": 2.6385,
+      "step": 9669
+    },
+    {
+      "epoch": 0.8753507739657825,
+      "grad_norm": 0.8609418869018555,
+      "learning_rate": 5.478104631726711e-07,
+      "loss": 2.6145,
+      "step": 9670
+    },
+    {
+      "epoch": 0.8754412962795329,
+      "grad_norm": 0.8679646253585815,
+      "learning_rate": 5.444984418336852e-07,
+      "loss": 2.6633,
+      "step": 9671
+    },
+    {
+      "epoch": 0.8755318185932832,
+      "grad_norm": 0.8041927218437195,
+      "learning_rate": 5.411964356609845e-07,
+      "loss": 2.6179,
+      "step": 9672
+    },
+    {
+      "epoch": 0.8756223409070336,
+      "grad_norm": 0.8592411279678345,
+      "learning_rate": 5.379044449870807e-07,
+      "loss": 2.598,
+      "step": 9673
+    },
+    {
+      "epoch": 0.8757128632207839,
+      "grad_norm": 0.8296803832054138,
+      "learning_rate": 5.346224701434866e-07,
+      "loss": 2.6365,
+      "step": 9674
+    },
+    {
+      "epoch": 0.8758033855345343,
+      "grad_norm": 0.8249205946922302,
+      "learning_rate": 5.31350511460682e-07,
+      "loss": 2.613,
+      "step": 9675
+    },
+    {
+      "epoch": 0.8758939078482846,
+      "grad_norm": 0.80314040184021,
+      "learning_rate": 5.280885692681592e-07,
+      "loss": 2.5824,
+      "step": 9676
+    },
+    {
+      "epoch": 0.875984430162035,
+      "grad_norm": 0.8617088794708252,
+      "learning_rate": 5.248366438943996e-07,
+      "loss": 2.5894,
+      "step": 9677
+    },
+    {
+      "epoch": 0.8760749524757853,
+      "grad_norm": 0.8568413257598877,
+      "learning_rate": 5.215947356668638e-07,
+      "loss": 2.7115,
+      "step": 9678
+    },
+    {
+      "epoch": 0.8761654747895357,
+      "grad_norm": 0.8277208209037781,
+      "learning_rate": 5.183628449120126e-07,
+      "loss": 2.6778,
+      "step": 9679
+    },
+    {
+      "epoch": 0.8762559971032859,
+      "grad_norm": 0.9010827541351318,
+      "learning_rate": 5.151409719553079e-07,
+      "loss": 2.6317,
+      "step": 9680
+    },
+    {
+      "epoch": 0.8763465194170363,
+      "grad_norm": 0.8497141599655151,
+      "learning_rate": 5.119291171211793e-07,
+      "loss": 2.6076,
+      "step": 9681
+    },
+    {
+      "epoch": 0.8764370417307866,
+      "grad_norm": 0.8685660362243652,
+      "learning_rate": 5.087272807330679e-07,
+      "loss": 2.6325,
+      "step": 9682
+    },
+    {
+      "epoch": 0.876527564044537,
+      "grad_norm": 0.804347813129425,
+      "learning_rate": 5.055354631134047e-07,
+      "loss": 2.6702,
+      "step": 9683
+    },
+    {
+      "epoch": 0.8766180863582873,
+      "grad_norm": 0.8076506853103638,
+      "learning_rate": 5.02353664583588e-07,
+      "loss": 2.6483,
+      "step": 9684
+    },
+    {
+      "epoch": 0.8767086086720377,
+      "grad_norm": 0.790969967842102,
+      "learning_rate": 4.991818854640395e-07,
+      "loss": 2.5682,
+      "step": 9685
+    },
+    {
+      "epoch": 0.876799130985788,
+      "grad_norm": 0.8944148421287537,
+      "learning_rate": 4.960201260741593e-07,
+      "loss": 2.605,
+      "step": 9686
+    },
+    {
+      "epoch": 0.8768896532995384,
+      "grad_norm": 0.8538900017738342,
+      "learning_rate": 4.928683867323147e-07,
+      "loss": 2.6175,
+      "step": 9687
+    },
+    {
+      "epoch": 0.8769801756132887,
+      "grad_norm": 0.8444143533706665,
+      "learning_rate": 4.897266677559187e-07,
+      "loss": 2.6092,
+      "step": 9688
+    },
+    {
+      "epoch": 0.8770706979270391,
+      "grad_norm": 0.8742196559906006,
+      "learning_rate": 4.865949694613181e-07,
+      "loss": 2.6217,
+      "step": 9689
+    },
+    {
+      "epoch": 0.8771612202407894,
+      "grad_norm": 0.8844677805900574,
+      "learning_rate": 4.834732921638719e-07,
+      "loss": 2.5721,
+      "step": 9690
+    },
+    {
+      "epoch": 0.8772517425545397,
+      "grad_norm": 0.8656468391418457,
+      "learning_rate": 4.803616361779506e-07,
+      "loss": 2.6361,
+      "step": 9691
+    },
+    {
+      "epoch": 0.87734226486829,
+      "grad_norm": 0.8689473271369934,
+      "learning_rate": 4.772600018168816e-07,
+      "loss": 2.6226,
+      "step": 9692
+    },
+    {
+      "epoch": 0.8774327871820403,
+      "grad_norm": 0.8834279179573059,
+      "learning_rate": 4.7416838939301487e-07,
+      "loss": 2.5908,
+      "step": 9693
+    },
+    {
+      "epoch": 0.8775233094957907,
+      "grad_norm": 0.8247548937797546,
+      "learning_rate": 4.710867992176682e-07,
+      "loss": 2.6217,
+      "step": 9694
+    },
+    {
+      "epoch": 0.877613831809541,
+      "grad_norm": 0.8003450036048889,
+      "learning_rate": 4.680152316011488e-07,
+      "loss": 2.5869,
+      "step": 9695
+    },
+    {
+      "epoch": 0.8777043541232914,
+      "grad_norm": 0.8029552698135376,
+      "learning_rate": 4.649536868527871e-07,
+      "loss": 2.6065,
+      "step": 9696
+    },
+    {
+      "epoch": 0.8777948764370417,
+      "grad_norm": 0.8596997857093811,
+      "learning_rate": 4.619021652808586e-07,
+      "loss": 2.7067,
+      "step": 9697
+    },
+    {
+      "epoch": 0.8778853987507921,
+      "grad_norm": 0.7986694574356079,
+      "learning_rate": 4.58860667192651e-07,
+      "loss": 2.5868,
+      "step": 9698
+    },
+    {
+      "epoch": 0.8779759210645424,
+      "grad_norm": 0.8311411142349243,
+      "learning_rate": 4.5582919289446357e-07,
+      "loss": 2.5162,
+      "step": 9699
+    },
+    {
+      "epoch": 0.8780664433782928,
+      "grad_norm": 0.7954468131065369,
+      "learning_rate": 4.5280774269154115e-07,
+      "loss": 2.5697,
+      "step": 9700
+    },
+    {
+      "epoch": 0.878156965692043,
+      "grad_norm": 0.9038304090499878,
+      "learning_rate": 4.497963168881625e-07,
+      "loss": 2.5705,
+      "step": 9701
+    },
+    {
+      "epoch": 0.8782474880057934,
+      "grad_norm": 0.856782078742981,
+      "learning_rate": 4.4679491578756285e-07,
+      "loss": 2.6178,
+      "step": 9702
+    },
+    {
+      "epoch": 0.8783380103195437,
+      "grad_norm": 0.8561943769454956,
+      "learning_rate": 4.438035396920004e-07,
+      "loss": 2.6096,
+      "step": 9703
+    },
+    {
+      "epoch": 0.8784285326332941,
+      "grad_norm": 0.8387544751167297,
+      "learning_rate": 4.408221889026898e-07,
+      "loss": 2.5912,
+      "step": 9704
+    },
+    {
+      "epoch": 0.8785190549470444,
+      "grad_norm": 0.8454294204711914,
+      "learning_rate": 4.378508637198686e-07,
+      "loss": 2.5926,
+      "step": 9705
+    },
+    {
+      "epoch": 0.8786095772607948,
+      "grad_norm": 0.9286566972732544,
+      "learning_rate": 4.348895644427309e-07,
+      "loss": 2.6156,
+      "step": 9706
+    },
+    {
+      "epoch": 0.8787000995745451,
+      "grad_norm": 0.8351259231567383,
+      "learning_rate": 4.319382913695047e-07,
+      "loss": 2.651,
+      "step": 9707
+    },
+    {
+      "epoch": 0.8787906218882955,
+      "grad_norm": 0.8249180912971497,
+      "learning_rate": 4.289970447973635e-07,
+      "loss": 2.5576,
+      "step": 9708
+    },
+    {
+      "epoch": 0.8788811442020458,
+      "grad_norm": 0.8286994695663452,
+      "learning_rate": 4.2606582502248137e-07,
+      "loss": 2.6025,
+      "step": 9709
+    },
+    {
+      "epoch": 0.8789716665157962,
+      "grad_norm": 0.8584800362586975,
+      "learning_rate": 4.2314463234005565e-07,
+      "loss": 2.5787,
+      "step": 9710
+    },
+    {
+      "epoch": 0.8790621888295465,
+      "grad_norm": 0.8871709108352661,
+      "learning_rate": 4.2023346704425094e-07,
+      "loss": 2.6003,
+      "step": 9711
+    },
+    {
+      "epoch": 0.8791527111432968,
+      "grad_norm": 0.8227541446685791,
+      "learning_rate": 4.173323294281994e-07,
+      "loss": 2.5901,
+      "step": 9712
+    },
+    {
+      "epoch": 0.8792432334570471,
+      "grad_norm": 0.8498722910881042,
+      "learning_rate": 4.144412197840564e-07,
+      "loss": 2.5865,
+      "step": 9713
+    },
+    {
+      "epoch": 0.8793337557707975,
+      "grad_norm": 0.8699567317962646,
+      "learning_rate": 4.115601384029666e-07,
+      "loss": 2.6844,
+      "step": 9714
+    },
+    {
+      "epoch": 0.8794242780845478,
+      "grad_norm": 0.9470655918121338,
+      "learning_rate": 4.0868908557504247e-07,
+      "loss": 2.643,
+      "step": 9715
+    },
+    {
+      "epoch": 0.8795148003982982,
+      "grad_norm": 0.8927464485168457,
+      "learning_rate": 4.058280615893972e-07,
+      "loss": 2.5728,
+      "step": 9716
+    },
+    {
+      "epoch": 0.8796053227120485,
+      "grad_norm": 0.8503594398498535,
+      "learning_rate": 4.029770667341448e-07,
+      "loss": 2.6019,
+      "step": 9717
+    },
+    {
+      "epoch": 0.8796958450257989,
+      "grad_norm": 0.885911226272583,
+      "learning_rate": 4.001361012963778e-07,
+      "loss": 2.6893,
+      "step": 9718
+    },
+    {
+      "epoch": 0.8797863673395492,
+      "grad_norm": 0.7960912585258484,
+      "learning_rate": 3.9730516556217845e-07,
+      "loss": 2.5933,
+      "step": 9719
+    },
+    {
+      "epoch": 0.8798768896532996,
+      "grad_norm": 0.8120215535163879,
+      "learning_rate": 3.9448425981661876e-07,
+      "loss": 2.6384,
+      "step": 9720
+    },
+    {
+      "epoch": 0.8799674119670499,
+      "grad_norm": 0.8336339592933655,
+      "learning_rate": 3.916733843437825e-07,
+      "loss": 2.6157,
+      "step": 9721
+    },
+    {
+      "epoch": 0.8800579342808003,
+      "grad_norm": 0.8533289432525635,
+      "learning_rate": 3.88872539426699e-07,
+      "loss": 2.5811,
+      "step": 9722
+    },
+    {
+      "epoch": 0.8801484565945505,
+      "grad_norm": 0.8294474482536316,
+      "learning_rate": 3.8608172534743135e-07,
+      "loss": 2.6509,
+      "step": 9723
+    },
+    {
+      "epoch": 0.8802389789083009,
+      "grad_norm": 0.9152242541313171,
+      "learning_rate": 3.833009423869993e-07,
+      "loss": 2.6402,
+      "step": 9724
+    },
+    {
+      "epoch": 0.8803295012220512,
+      "grad_norm": 1.01529061794281,
+      "learning_rate": 3.805301908254455e-07,
+      "loss": 2.6434,
+      "step": 9725
+    },
+    {
+      "epoch": 0.8804200235358016,
+      "grad_norm": 0.8746420741081238,
+      "learning_rate": 3.7776947094176893e-07,
+      "loss": 2.6068,
+      "step": 9726
+    },
+    {
+      "epoch": 0.8805105458495519,
+      "grad_norm": 0.875525712966919,
+      "learning_rate": 3.750187830139917e-07,
+      "loss": 2.5965,
+      "step": 9727
+    },
+    {
+      "epoch": 0.8806010681633023,
+      "grad_norm": 0.9152287244796753,
+      "learning_rate": 3.7227812731909227e-07,
+      "loss": 2.5523,
+      "step": 9728
+    },
+    {
+      "epoch": 0.8806915904770526,
+      "grad_norm": 0.7987388968467712,
+      "learning_rate": 3.6954750413306096e-07,
+      "loss": 2.6045,
+      "step": 9729
+    },
+    {
+      "epoch": 0.880782112790803,
+      "grad_norm": 0.8325386047363281,
+      "learning_rate": 3.6682691373086665e-07,
+      "loss": 2.6529,
+      "step": 9730
+    },
+    {
+      "epoch": 0.8808726351045533,
+      "grad_norm": 0.7871389985084534,
+      "learning_rate": 3.64116356386468e-07,
+      "loss": 2.5837,
+      "step": 9731
+    },
+    {
+      "epoch": 0.8809631574183037,
+      "grad_norm": 0.83063805103302,
+      "learning_rate": 3.614158323728356e-07,
+      "loss": 2.607,
+      "step": 9732
+    },
+    {
+      "epoch": 0.881053679732054,
+      "grad_norm": 0.859025239944458,
+      "learning_rate": 3.587253419619074e-07,
+      "loss": 2.6318,
+      "step": 9733
+    },
+    {
+      "epoch": 0.8811442020458042,
+      "grad_norm": 0.8289056420326233,
+      "learning_rate": 3.5604488542460014e-07,
+      "loss": 2.5765,
+      "step": 9734
+    },
+    {
+      "epoch": 0.8812347243595546,
+      "grad_norm": 0.826295793056488,
+      "learning_rate": 3.533744630308533e-07,
+      "loss": 2.6251,
+      "step": 9735
+    },
+    {
+      "epoch": 0.8813252466733049,
+      "grad_norm": 0.8544880151748657,
+      "learning_rate": 3.50714075049563e-07,
+      "loss": 2.5817,
+      "step": 9736
+    },
+    {
+      "epoch": 0.8814157689870553,
+      "grad_norm": 0.8992463946342468,
+      "learning_rate": 3.480637217486482e-07,
+      "loss": 2.7122,
+      "step": 9737
+    },
+    {
+      "epoch": 0.8815062913008056,
+      "grad_norm": 0.9246089458465576,
+      "learning_rate": 3.4542340339498437e-07,
+      "loss": 2.6458,
+      "step": 9738
+    },
+    {
+      "epoch": 0.881596813614556,
+      "grad_norm": 0.7926107048988342,
+      "learning_rate": 3.4279312025445876e-07,
+      "loss": 2.6255,
+      "step": 9739
+    },
+    {
+      "epoch": 0.8816873359283063,
+      "grad_norm": 0.8976580500602722,
+      "learning_rate": 3.401728725919373e-07,
+      "loss": 2.6193,
+      "step": 9740
+    },
+    {
+      "epoch": 0.8817778582420567,
+      "grad_norm": 0.8773987889289856,
+      "learning_rate": 3.375626606712867e-07,
+      "loss": 2.6014,
+      "step": 9741
+    },
+    {
+      "epoch": 0.881868380555807,
+      "grad_norm": 0.8479148149490356,
+      "learning_rate": 3.349624847553412e-07,
+      "loss": 2.6462,
+      "step": 9742
+    },
+    {
+      "epoch": 0.8819589028695574,
+      "grad_norm": 0.8767992854118347,
+      "learning_rate": 3.3237234510595793e-07,
+      "loss": 2.5869,
+      "step": 9743
+    },
+    {
+      "epoch": 0.8820494251833076,
+      "grad_norm": 0.8235431909561157,
+      "learning_rate": 3.2979224198395055e-07,
+      "loss": 2.6434,
+      "step": 9744
+    },
+    {
+      "epoch": 0.882139947497058,
+      "grad_norm": 0.8444194197654724,
+      "learning_rate": 3.2722217564912226e-07,
+      "loss": 2.6202,
+      "step": 9745
+    },
+    {
+      "epoch": 0.8822304698108083,
+      "grad_norm": 0.8630514144897461,
+      "learning_rate": 3.2466214636031056e-07,
+      "loss": 2.5635,
+      "step": 9746
+    },
+    {
+      "epoch": 0.8823209921245587,
+      "grad_norm": 0.8077291250228882,
+      "learning_rate": 3.2211215437528694e-07,
+      "loss": 2.6431,
+      "step": 9747
+    },
+    {
+      "epoch": 0.882411514438309,
+      "grad_norm": 0.9250985383987427,
+      "learning_rate": 3.195721999508461e-07,
+      "loss": 2.6194,
+      "step": 9748
+    },
+    {
+      "epoch": 0.8825020367520594,
+      "grad_norm": 0.8408756256103516,
+      "learning_rate": 3.1704228334273887e-07,
+      "loss": 2.6216,
+      "step": 9749
+    },
+    {
+      "epoch": 0.8825925590658097,
+      "grad_norm": 0.8261474967002869,
+      "learning_rate": 3.145224048057727e-07,
+      "loss": 2.6363,
+      "step": 9750
+    },
+    {
+      "epoch": 0.8826830813795601,
+      "grad_norm": 0.8455504775047302,
+      "learning_rate": 3.1201256459365557e-07,
+      "loss": 2.6032,
+      "step": 9751
+    },
+    {
+      "epoch": 0.8827736036933104,
+      "grad_norm": 0.7715721130371094,
+      "learning_rate": 3.095127629591521e-07,
+      "loss": 2.5847,
+      "step": 9752
+    },
+    {
+      "epoch": 0.8828641260070608,
+      "grad_norm": 0.8211907744407654,
+      "learning_rate": 3.070230001539831e-07,
+      "loss": 2.6066,
+      "step": 9753
+    },
+    {
+      "epoch": 0.882954648320811,
+      "grad_norm": 0.8626493215560913,
+      "learning_rate": 3.045432764288703e-07,
+      "loss": 2.6688,
+      "step": 9754
+    },
+    {
+      "epoch": 0.8830451706345614,
+      "grad_norm": 0.8915566205978394,
+      "learning_rate": 3.020735920335138e-07,
+      "loss": 2.663,
+      "step": 9755
+    },
+    {
+      "epoch": 0.8831356929483117,
+      "grad_norm": 0.8275890946388245,
+      "learning_rate": 2.9961394721663703e-07,
+      "loss": 2.6331,
+      "step": 9756
+    },
+    {
+      "epoch": 0.8832262152620621,
+      "grad_norm": 0.837154746055603,
+      "learning_rate": 2.971643422258974e-07,
+      "loss": 2.6798,
+      "step": 9757
+    },
+    {
+      "epoch": 0.8833167375758124,
+      "grad_norm": 0.9589747190475464,
+      "learning_rate": 2.947247773079753e-07,
+      "loss": 2.5884,
+      "step": 9758
+    },
+    {
+      "epoch": 0.8834072598895628,
+      "grad_norm": 0.8292232751846313,
+      "learning_rate": 2.922952527085521e-07,
+      "loss": 2.6457,
+      "step": 9759
+    },
+    {
+      "epoch": 0.8834977822033131,
+      "grad_norm": 0.8505760431289673,
+      "learning_rate": 2.898757686722542e-07,
+      "loss": 2.5983,
+      "step": 9760
+    },
+    {
+      "epoch": 0.8835883045170635,
+      "grad_norm": 0.7981782555580139,
+      "learning_rate": 2.874663254427534e-07,
+      "loss": 2.6458,
+      "step": 9761
+    },
+    {
+      "epoch": 0.8836788268308138,
+      "grad_norm": 0.85322105884552,
+      "learning_rate": 2.850669232626557e-07,
+      "loss": 2.6481,
+      "step": 9762
+    },
+    {
+      "epoch": 0.8837693491445642,
+      "grad_norm": 0.9375588297843933,
+      "learning_rate": 2.8267756237359e-07,
+      "loss": 2.642,
+      "step": 9763
+    },
+    {
+      "epoch": 0.8838598714583145,
+      "grad_norm": 0.8446762561798096,
+      "learning_rate": 2.802982430161749e-07,
+      "loss": 2.6055,
+      "step": 9764
+    },
+    {
+      "epoch": 0.8839503937720649,
+      "grad_norm": 0.8564532399177551,
+      "learning_rate": 2.779289654299855e-07,
+      "loss": 2.6305,
+      "step": 9765
+    },
+    {
+      "epoch": 0.8840409160858151,
+      "grad_norm": 0.8674356341362,
+      "learning_rate": 2.7556972985363085e-07,
+      "loss": 2.639,
+      "step": 9766
+    },
+    {
+      "epoch": 0.8841314383995655,
+      "grad_norm": 0.8495703935623169,
+      "learning_rate": 2.732205365246654e-07,
+      "loss": 2.6171,
+      "step": 9767
+    },
+    {
+      "epoch": 0.8842219607133158,
+      "grad_norm": 0.828662633895874,
+      "learning_rate": 2.708813856796666e-07,
+      "loss": 2.6002,
+      "step": 9768
+    },
+    {
+      "epoch": 0.8843124830270662,
+      "grad_norm": 0.8758851289749146,
+      "learning_rate": 2.685522775541904e-07,
+      "loss": 2.6089,
+      "step": 9769
+    },
+    {
+      "epoch": 0.8844030053408165,
+      "grad_norm": 0.8468634486198425,
+      "learning_rate": 2.6623321238277157e-07,
+      "loss": 2.6449,
+      "step": 9770
+    },
+    {
+      "epoch": 0.8844935276545669,
+      "grad_norm": 0.8085419535636902,
+      "learning_rate": 2.6392419039892314e-07,
+      "loss": 2.6476,
+      "step": 9771
+    },
+    {
+      "epoch": 0.8845840499683172,
+      "grad_norm": 0.9288700222969055,
+      "learning_rate": 2.616252118352036e-07,
+      "loss": 2.6527,
+      "step": 9772
+    },
+    {
+      "epoch": 0.8846745722820676,
+      "grad_norm": 0.796964704990387,
+      "learning_rate": 2.5933627692308336e-07,
+      "loss": 2.6187,
+      "step": 9773
+    },
+    {
+      "epoch": 0.8847650945958179,
+      "grad_norm": 0.8337318897247314,
+      "learning_rate": 2.5705738589306696e-07,
+      "loss": 2.5765,
+      "step": 9774
+    },
+    {
+      "epoch": 0.8848556169095682,
+      "grad_norm": 0.9105563163757324,
+      "learning_rate": 2.547885389746485e-07,
+      "loss": 2.5931,
+      "step": 9775
+    },
+    {
+      "epoch": 0.8849461392233186,
+      "grad_norm": 0.8183169364929199,
+      "learning_rate": 2.5252973639628976e-07,
+      "loss": 2.5773,
+      "step": 9776
+    },
+    {
+      "epoch": 0.8850366615370688,
+      "grad_norm": 0.8156883120536804,
+      "learning_rate": 2.5028097838546425e-07,
+      "loss": 2.6185,
+      "step": 9777
+    },
+    {
+      "epoch": 0.8851271838508192,
+      "grad_norm": 0.9505560398101807,
+      "learning_rate": 2.480422651686132e-07,
+      "loss": 2.5851,
+      "step": 9778
+    },
+    {
+      "epoch": 0.8852177061645695,
+      "grad_norm": 0.9275203943252563,
+      "learning_rate": 2.4581359697116724e-07,
+      "loss": 2.6243,
+      "step": 9779
+    },
+    {
+      "epoch": 0.8853082284783199,
+      "grad_norm": 0.9110391139984131,
+      "learning_rate": 2.4359497401758024e-07,
+      "loss": 2.6535,
+      "step": 9780
+    },
+    {
+      "epoch": 0.8853987507920702,
+      "grad_norm": 0.8156226873397827,
+      "learning_rate": 2.413863965312402e-07,
+      "loss": 2.5498,
+      "step": 9781
+    },
+    {
+      "epoch": 0.8854892731058206,
+      "grad_norm": 0.8637019991874695,
+      "learning_rate": 2.3918786473455803e-07,
+      "loss": 2.6342,
+      "step": 9782
+    },
+    {
+      "epoch": 0.8855797954195709,
+      "grad_norm": 0.7892611622810364,
+      "learning_rate": 2.3699937884892332e-07,
+      "loss": 2.6476,
+      "step": 9783
+    },
+    {
+      "epoch": 0.8856703177333213,
+      "grad_norm": 0.8169649243354797,
+      "learning_rate": 2.3482093909473756e-07,
+      "loss": 2.6645,
+      "step": 9784
+    },
+    {
+      "epoch": 0.8857608400470716,
+      "grad_norm": 0.8641905188560486,
+      "learning_rate": 2.3265254569133644e-07,
+      "loss": 2.65,
+      "step": 9785
+    },
+    {
+      "epoch": 0.885851362360822,
+      "grad_norm": 0.9224991798400879,
+      "learning_rate": 2.3049419885711187e-07,
+      "loss": 2.5947,
+      "step": 9786
+    },
+    {
+      "epoch": 0.8859418846745722,
+      "grad_norm": 0.8599008321762085,
+      "learning_rate": 2.2834589880937894e-07,
+      "loss": 2.6237,
+      "step": 9787
+    },
+    {
+      "epoch": 0.8860324069883226,
+      "grad_norm": 0.8519898653030396,
+      "learning_rate": 2.2620764576448683e-07,
+      "loss": 2.6718,
+      "step": 9788
+    },
+    {
+      "epoch": 0.8861229293020729,
+      "grad_norm": 0.8302574157714844,
+      "learning_rate": 2.2407943993775214e-07,
+      "loss": 2.5672,
+      "step": 9789
+    },
+    {
+      "epoch": 0.8862134516158233,
+      "grad_norm": 0.8853892087936401,
+      "learning_rate": 2.219612815434924e-07,
+      "loss": 2.6739,
+      "step": 9790
+    },
+    {
+      "epoch": 0.8863039739295736,
+      "grad_norm": 0.81326824426651,
+      "learning_rate": 2.1985317079500356e-07,
+      "loss": 2.6567,
+      "step": 9791
+    },
+    {
+      "epoch": 0.886394496243324,
+      "grad_norm": 0.8466541767120361,
+      "learning_rate": 2.1775510790456033e-07,
+      "loss": 2.6393,
+      "step": 9792
+    },
+    {
+      "epoch": 0.8864850185570743,
+      "grad_norm": 0.8405859470367432,
+      "learning_rate": 2.1566709308346034e-07,
+      "loss": 2.6045,
+      "step": 9793
+    },
+    {
+      "epoch": 0.8865755408708247,
+      "grad_norm": 0.867979109287262,
+      "learning_rate": 2.135891265419465e-07,
+      "loss": 2.6394,
+      "step": 9794
+    },
+    {
+      "epoch": 0.886666063184575,
+      "grad_norm": 0.8394297361373901,
+      "learning_rate": 2.115212084892737e-07,
+      "loss": 2.5794,
+      "step": 9795
+    },
+    {
+      "epoch": 0.8867565854983254,
+      "grad_norm": 0.9299885034561157,
+      "learning_rate": 2.0946333913368643e-07,
+      "loss": 2.6022,
+      "step": 9796
+    },
+    {
+      "epoch": 0.8868471078120757,
+      "grad_norm": 0.8341772556304932,
+      "learning_rate": 2.0741551868241893e-07,
+      "loss": 2.6508,
+      "step": 9797
+    },
+    {
+      "epoch": 0.886937630125826,
+      "grad_norm": 0.8111425042152405,
+      "learning_rate": 2.0537774734167293e-07,
+      "loss": 2.6191,
+      "step": 9798
+    },
+    {
+      "epoch": 0.8870281524395763,
+      "grad_norm": 0.9102740287780762,
+      "learning_rate": 2.0335002531665092e-07,
+      "loss": 2.5984,
+      "step": 9799
+    },
+    {
+      "epoch": 0.8871186747533267,
+      "grad_norm": 0.8335270881652832,
+      "learning_rate": 2.0133235281156736e-07,
+      "loss": 2.535,
+      "step": 9800
+    },
+    {
+      "epoch": 0.887209197067077,
+      "grad_norm": 0.8242570161819458,
+      "learning_rate": 1.9932473002957087e-07,
+      "loss": 2.6613,
+      "step": 9801
+    },
+    {
+      "epoch": 0.8872997193808274,
+      "grad_norm": 0.8383133411407471,
+      "learning_rate": 1.973271571728441e-07,
+      "loss": 2.6395,
+      "step": 9802
+    },
+    {
+      "epoch": 0.8873902416945777,
+      "grad_norm": 0.8448271751403809,
+      "learning_rate": 1.9533963444254843e-07,
+      "loss": 2.6522,
+      "step": 9803
+    },
+    {
+      "epoch": 0.8874807640083281,
+      "grad_norm": 0.84769207239151,
+      "learning_rate": 1.9336216203881262e-07,
+      "loss": 2.6113,
+      "step": 9804
+    },
+    {
+      "epoch": 0.8875712863220784,
+      "grad_norm": 0.8611244559288025,
+      "learning_rate": 1.913947401607774e-07,
+      "loss": 2.5958,
+      "step": 9805
+    },
+    {
+      "epoch": 0.8876618086358288,
+      "grad_norm": 0.8302865624427795,
+      "learning_rate": 1.8943736900656206e-07,
+      "loss": 2.6314,
+      "step": 9806
+    },
+    {
+      "epoch": 0.8877523309495791,
+      "grad_norm": 0.8019070029258728,
+      "learning_rate": 1.8749004877327558e-07,
+      "loss": 2.648,
+      "step": 9807
+    },
+    {
+      "epoch": 0.8878428532633295,
+      "grad_norm": 0.856889009475708,
+      "learning_rate": 1.8555277965701668e-07,
+      "loss": 2.6345,
+      "step": 9808
+    },
+    {
+      "epoch": 0.8879333755770797,
+      "grad_norm": 0.8450315594673157,
+      "learning_rate": 1.836255618528515e-07,
+      "loss": 2.6494,
+      "step": 9809
+    },
+    {
+      "epoch": 0.8880238978908301,
+      "grad_norm": 0.8036577701568604,
+      "learning_rate": 1.817083955548693e-07,
+      "loss": 2.6248,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8881144202045804,
+      "grad_norm": 0.8193379044532776,
+      "learning_rate": 1.798012809561045e-07,
+      "loss": 2.5855,
+      "step": 9811
+    },
+    {
+      "epoch": 0.8882049425183308,
+      "grad_norm": 0.7942414879798889,
+      "learning_rate": 1.7790421824863678e-07,
+      "loss": 2.591,
+      "step": 9812
+    },
+    {
+      "epoch": 0.8882954648320811,
+      "grad_norm": 0.9019108414649963,
+      "learning_rate": 1.7601720762346897e-07,
+      "loss": 2.6131,
+      "step": 9813
+    },
+    {
+      "epoch": 0.8883859871458315,
+      "grad_norm": 0.9382195472717285,
+      "learning_rate": 1.7414024927064897e-07,
+      "loss": 2.6733,
+      "step": 9814
+    },
+    {
+      "epoch": 0.8884765094595818,
+      "grad_norm": 0.8000298142433167,
+      "learning_rate": 1.7227334337917011e-07,
+      "loss": 2.5882,
+      "step": 9815
+    },
+    {
+      "epoch": 0.8885670317733321,
+      "grad_norm": 0.7768236994743347,
+      "learning_rate": 1.7041649013703753e-07,
+      "loss": 2.6274,
+      "step": 9816
+    },
+    {
+      "epoch": 0.8886575540870825,
+      "grad_norm": 0.9013504981994629,
+      "learning_rate": 1.685696897312239e-07,
+      "loss": 2.5695,
+      "step": 9817
+    },
+    {
+      "epoch": 0.8887480764008328,
+      "grad_norm": 0.8952003121376038,
+      "learning_rate": 1.6673294234771374e-07,
+      "loss": 2.5901,
+      "step": 9818
+    },
+    {
+      "epoch": 0.8888385987145832,
+      "grad_norm": 0.8011465072631836,
+      "learning_rate": 1.6490624817147026e-07,
+      "loss": 2.6135,
+      "step": 9819
+    },
+    {
+      "epoch": 0.8889291210283334,
+      "grad_norm": 0.7969688773155212,
+      "learning_rate": 1.630896073864352e-07,
+      "loss": 2.6103,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8890196433420838,
+      "grad_norm": 0.853882372379303,
+      "learning_rate": 1.6128302017553998e-07,
+      "loss": 2.6324,
+      "step": 9821
+    },
+    {
+      "epoch": 0.8891101656558341,
+      "grad_norm": 0.8941799402236938,
+      "learning_rate": 1.5948648672071687e-07,
+      "loss": 2.5876,
+      "step": 9822
+    },
+    {
+      "epoch": 0.8892006879695845,
+      "grad_norm": 0.8067328929901123,
+      "learning_rate": 1.577000072028656e-07,
+      "loss": 2.5852,
+      "step": 9823
+    },
+    {
+      "epoch": 0.8892912102833348,
+      "grad_norm": 0.83470618724823,
+      "learning_rate": 1.5592358180189782e-07,
+      "loss": 2.6015,
+      "step": 9824
+    },
+    {
+      "epoch": 0.8893817325970852,
+      "grad_norm": 0.8342416882514954,
+      "learning_rate": 1.5415721069669264e-07,
+      "loss": 2.5621,
+      "step": 9825
+    },
+    {
+      "epoch": 0.8894722549108355,
+      "grad_norm": 0.7842913866043091,
+      "learning_rate": 1.5240089406513003e-07,
+      "loss": 2.6539,
+      "step": 9826
+    },
+    {
+      "epoch": 0.8895627772245859,
+      "grad_norm": 0.8189299702644348,
+      "learning_rate": 1.506546320840574e-07,
+      "loss": 2.586,
+      "step": 9827
+    },
+    {
+      "epoch": 0.8896532995383362,
+      "grad_norm": 0.8747230768203735,
+      "learning_rate": 1.4891842492933404e-07,
+      "loss": 2.6314,
+      "step": 9828
+    },
+    {
+      "epoch": 0.8897438218520866,
+      "grad_norm": 0.8976212739944458,
+      "learning_rate": 1.4719227277578685e-07,
+      "loss": 2.5779,
+      "step": 9829
+    },
+    {
+      "epoch": 0.8898343441658368,
+      "grad_norm": 0.9021840691566467,
+      "learning_rate": 1.4547617579725449e-07,
+      "loss": 2.5662,
+      "step": 9830
+    },
+    {
+      "epoch": 0.8899248664795872,
+      "grad_norm": 0.8414490222930908,
+      "learning_rate": 1.4377013416654317e-07,
+      "loss": 2.5859,
+      "step": 9831
+    },
+    {
+      "epoch": 0.8900153887933375,
+      "grad_norm": 0.8306783437728882,
+      "learning_rate": 1.4207414805543774e-07,
+      "loss": 2.6371,
+      "step": 9832
+    },
+    {
+      "epoch": 0.8901059111070879,
+      "grad_norm": 0.8770350813865662,
+      "learning_rate": 1.4038821763473485e-07,
+      "loss": 2.6405,
+      "step": 9833
+    },
+    {
+      "epoch": 0.8901964334208382,
+      "grad_norm": 0.8089725971221924,
+      "learning_rate": 1.3871234307420989e-07,
+      "loss": 2.5898,
+      "step": 9834
+    },
+    {
+      "epoch": 0.8902869557345886,
+      "grad_norm": 0.8233959078788757,
+      "learning_rate": 1.3704652454261668e-07,
+      "loss": 2.62,
+      "step": 9835
+    },
+    {
+      "epoch": 0.8903774780483389,
+      "grad_norm": 0.828133761882782,
+      "learning_rate": 1.3539076220769887e-07,
+      "loss": 2.6081,
+      "step": 9836
+    },
+    {
+      "epoch": 0.8904680003620893,
+      "grad_norm": 0.8399420976638794,
+      "learning_rate": 1.3374505623621192e-07,
+      "loss": 2.5688,
+      "step": 9837
+    },
+    {
+      "epoch": 0.8905585226758396,
+      "grad_norm": 0.8425469398498535,
+      "learning_rate": 1.3210940679385664e-07,
+      "loss": 2.6351,
+      "step": 9838
+    },
+    {
+      "epoch": 0.89064904498959,
+      "grad_norm": 0.8105515241622925,
+      "learning_rate": 1.3048381404535682e-07,
+      "loss": 2.5983,
+      "step": 9839
+    },
+    {
+      "epoch": 0.8907395673033403,
+      "grad_norm": 0.8184119462966919,
+      "learning_rate": 1.2886827815440372e-07,
+      "loss": 2.6031,
+      "step": 9840
+    },
+    {
+      "epoch": 0.8908300896170906,
+      "grad_norm": 0.8745843768119812,
+      "learning_rate": 1.2726279928367835e-07,
+      "loss": 2.6651,
+      "step": 9841
+    },
+    {
+      "epoch": 0.8909206119308409,
+      "grad_norm": 0.8466340899467468,
+      "learning_rate": 1.2566737759485136e-07,
+      "loss": 2.6278,
+      "step": 9842
+    },
+    {
+      "epoch": 0.8910111342445913,
+      "grad_norm": 0.8394578695297241,
+      "learning_rate": 1.2408201324859425e-07,
+      "loss": 2.5863,
+      "step": 9843
+    },
+    {
+      "epoch": 0.8911016565583416,
+      "grad_norm": 0.8000689148902893,
+      "learning_rate": 1.2250670640454597e-07,
+      "loss": 2.6167,
+      "step": 9844
+    },
+    {
+      "epoch": 0.891192178872092,
+      "grad_norm": 0.784951388835907,
+      "learning_rate": 1.2094145722134632e-07,
+      "loss": 2.5898,
+      "step": 9845
+    },
+    {
+      "epoch": 0.8912827011858423,
+      "grad_norm": 0.8008831739425659,
+      "learning_rate": 1.193862658566025e-07,
+      "loss": 2.6132,
+      "step": 9846
+    },
+    {
+      "epoch": 0.8913732234995927,
+      "grad_norm": 0.8564389944076538,
+      "learning_rate": 1.1784113246692263e-07,
+      "loss": 2.6358,
+      "step": 9847
+    },
+    {
+      "epoch": 0.891463745813343,
+      "grad_norm": 0.8283603191375732,
+      "learning_rate": 1.1630605720791554e-07,
+      "loss": 2.59,
+      "step": 9848
+    },
+    {
+      "epoch": 0.8915542681270934,
+      "grad_norm": 0.796347439289093,
+      "learning_rate": 1.1478104023414648e-07,
+      "loss": 2.5929,
+      "step": 9849
+    },
+    {
+      "epoch": 0.8916447904408437,
+      "grad_norm": 0.7919137477874756,
+      "learning_rate": 1.1326608169920372e-07,
+      "loss": 2.5491,
+      "step": 9850
+    },
+    {
+      "epoch": 0.8917353127545941,
+      "grad_norm": 0.9107193946838379,
+      "learning_rate": 1.1176118175563188e-07,
+      "loss": 2.6295,
+      "step": 9851
+    },
+    {
+      "epoch": 0.8918258350683443,
+      "grad_norm": 0.7748007774353027,
+      "learning_rate": 1.1026634055497642e-07,
+      "loss": 2.572,
+      "step": 9852
+    },
+    {
+      "epoch": 0.8919163573820947,
+      "grad_norm": 0.852328896522522,
+      "learning_rate": 1.0878155824776137e-07,
+      "loss": 2.6392,
+      "step": 9853
+    },
+    {
+      "epoch": 0.892006879695845,
+      "grad_norm": 0.7925384044647217,
+      "learning_rate": 1.0730683498351157e-07,
+      "loss": 2.5562,
+      "step": 9854
+    },
+    {
+      "epoch": 0.8920974020095954,
+      "grad_norm": 0.8247081637382507,
+      "learning_rate": 1.0584217091073046e-07,
+      "loss": 2.6082,
+      "step": 9855
+    },
+    {
+      "epoch": 0.8921879243233457,
+      "grad_norm": 0.8525282740592957,
+      "learning_rate": 1.0438756617691115e-07,
+      "loss": 2.622,
+      "step": 9856
+    },
+    {
+      "epoch": 0.892278446637096,
+      "grad_norm": 0.8287916779518127,
+      "learning_rate": 1.0294302092853647e-07,
+      "loss": 2.6699,
+      "step": 9857
+    },
+    {
+      "epoch": 0.8923689689508464,
+      "grad_norm": 0.8627311587333679,
+      "learning_rate": 1.0150853531105675e-07,
+      "loss": 2.6384,
+      "step": 9858
+    },
+    {
+      "epoch": 0.8924594912645967,
+      "grad_norm": 0.8527313470840454,
+      "learning_rate": 1.000841094689342e-07,
+      "loss": 2.5526,
+      "step": 9859
+    },
+    {
+      "epoch": 0.8925500135783471,
+      "grad_norm": 0.9218466281890869,
+      "learning_rate": 9.866974354560965e-08,
+      "loss": 2.6543,
+      "step": 9860
+    },
+    {
+      "epoch": 0.8926405358920974,
+      "grad_norm": 0.8103916049003601,
+      "learning_rate": 9.72654376835136e-08,
+      "loss": 2.5987,
+      "step": 9861
+    },
+    {
+      "epoch": 0.8927310582058477,
+      "grad_norm": 0.7997682094573975,
+      "learning_rate": 9.587119202405515e-08,
+      "loss": 2.6328,
+      "step": 9862
+    },
+    {
+      "epoch": 0.892821580519598,
+      "grad_norm": 0.9294595122337341,
+      "learning_rate": 9.4487006707622e-08,
+      "loss": 2.6306,
+      "step": 9863
+    },
+    {
+      "epoch": 0.8929121028333484,
+      "grad_norm": 0.8661568760871887,
+      "learning_rate": 9.311288187362488e-08,
+      "loss": 2.6281,
+      "step": 9864
+    },
+    {
+      "epoch": 0.8930026251470987,
+      "grad_norm": 0.8452597856521606,
+      "learning_rate": 9.174881766043087e-08,
+      "loss": 2.6432,
+      "step": 9865
+    },
+    {
+      "epoch": 0.8930931474608491,
+      "grad_norm": 0.8144384622573853,
+      "learning_rate": 9.039481420538565e-08,
+      "loss": 2.6192,
+      "step": 9866
+    },
+    {
+      "epoch": 0.8931836697745994,
+      "grad_norm": 0.8570119738578796,
+      "learning_rate": 8.905087164485793e-08,
+      "loss": 2.6034,
+      "step": 9867
+    },
+    {
+      "epoch": 0.8932741920883498,
+      "grad_norm": 0.8554793000221252,
+      "learning_rate": 8.771699011416168e-08,
+      "loss": 2.6499,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8933647144021001,
+      "grad_norm": 0.8595912456512451,
+      "learning_rate": 8.6393169747645e-08,
+      "loss": 2.6522,
+      "step": 9869
+    },
+    {
+      "epoch": 0.8934552367158505,
+      "grad_norm": 0.7925248146057129,
+      "learning_rate": 8.507941067859016e-08,
+      "loss": 2.6335,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8935457590296008,
+      "grad_norm": 0.9258356690406799,
+      "learning_rate": 8.377571303931353e-08,
+      "loss": 2.6616,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8936362813433512,
+      "grad_norm": 0.8388485908508301,
+      "learning_rate": 8.24820769610879e-08,
+      "loss": 2.5649,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8937268036571014,
+      "grad_norm": 0.8156565427780151,
+      "learning_rate": 8.119850257417571e-08,
+      "loss": 2.6408,
+      "step": 9873
+    },
+    {
+      "epoch": 0.8938173259708518,
+      "grad_norm": 0.8444782495498657,
+      "learning_rate": 7.992499000785136e-08,
+      "loss": 2.6462,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8939078482846021,
+      "grad_norm": 0.7647959589958191,
+      "learning_rate": 7.866153939033449e-08,
+      "loss": 2.5655,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8939983705983525,
+      "grad_norm": 0.9548771381378174,
+      "learning_rate": 7.740815084887887e-08,
+      "loss": 2.6601,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8940888929121028,
+      "grad_norm": 0.8189795613288879,
+      "learning_rate": 7.616482450968354e-08,
+      "loss": 2.6318,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8941794152258532,
+      "grad_norm": 0.9351743459701538,
+      "learning_rate": 7.493156049794836e-08,
+      "loss": 2.5698,
+      "step": 9878
+    },
+    {
+      "epoch": 0.8942699375396035,
+      "grad_norm": 0.8982877135276794,
+      "learning_rate": 7.370835893788508e-08,
+      "loss": 2.6731,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8943604598533539,
+      "grad_norm": 0.9087199568748474,
+      "learning_rate": 7.249521995263964e-08,
+      "loss": 2.6435,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8944509821671042,
+      "grad_norm": 0.8815861940383911,
+      "learning_rate": 7.129214366440317e-08,
+      "loss": 2.6479,
+      "step": 9881
+    },
+    {
+      "epoch": 0.8945415044808546,
+      "grad_norm": 0.8453705906867981,
+      "learning_rate": 7.009913019431213e-08,
+      "loss": 2.6292,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8946320267946049,
+      "grad_norm": 0.8382077813148499,
+      "learning_rate": 6.891617966250374e-08,
+      "loss": 2.6051,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8947225491083552,
+      "grad_norm": 0.8215992450714111,
+      "learning_rate": 6.774329218810493e-08,
+      "loss": 2.6244,
+      "step": 9884
+    },
+    {
+      "epoch": 0.8948130714221055,
+      "grad_norm": 0.9238488078117371,
+      "learning_rate": 6.658046788921012e-08,
+      "loss": 2.6096,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8949035937358559,
+      "grad_norm": 0.8317505717277527,
+      "learning_rate": 6.542770688293676e-08,
+      "loss": 2.5954,
+      "step": 9886
+    },
+    {
+      "epoch": 0.8949941160496062,
+      "grad_norm": 0.8946189284324646,
+      "learning_rate": 6.428500928535864e-08,
+      "loss": 2.6525,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8950846383633566,
+      "grad_norm": 0.8472013473510742,
+      "learning_rate": 6.315237521155038e-08,
+      "loss": 2.5973,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8951751606771069,
+      "grad_norm": 0.8401451706886292,
+      "learning_rate": 6.202980477555408e-08,
+      "loss": 2.6031,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8952656829908573,
+      "grad_norm": 0.8763166666030884,
+      "learning_rate": 6.09172980904238e-08,
+      "loss": 2.6654,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8953562053046076,
+      "grad_norm": 0.8638229370117188,
+      "learning_rate": 5.981485526819208e-08,
+      "loss": 2.6553,
+      "step": 9891
+    },
+    {
+      "epoch": 0.895446727618358,
+      "grad_norm": 0.8521655797958374,
+      "learning_rate": 5.872247641987016e-08,
+      "loss": 2.579,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8955372499321083,
+      "grad_norm": 0.8707503080368042,
+      "learning_rate": 5.764016165545894e-08,
+      "loss": 2.592,
+      "step": 9893
+    },
+    {
+      "epoch": 0.8956277722458587,
+      "grad_norm": 0.7827308177947998,
+      "learning_rate": 5.6567911083937883e-08,
+      "loss": 2.5608,
+      "step": 9894
+    },
+    {
+      "epoch": 0.8957182945596089,
+      "grad_norm": 0.8894588351249695,
+      "learning_rate": 5.5505724813309514e-08,
+      "loss": 2.6634,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8958088168733593,
+      "grad_norm": 0.8418501019477844,
+      "learning_rate": 5.4453602950510494e-08,
+      "loss": 2.5637,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8958993391871096,
+      "grad_norm": 0.8499705791473389,
+      "learning_rate": 5.341154560150052e-08,
+      "loss": 2.6132,
+      "step": 9897
+    },
+    {
+      "epoch": 0.8959898615008599,
+      "grad_norm": 0.871951699256897,
+      "learning_rate": 5.2379552871217875e-08,
+      "loss": 2.5509,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8960803838146103,
+      "grad_norm": 0.8677495121955872,
+      "learning_rate": 5.1357624863579426e-08,
+      "loss": 2.6282,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8961709061283606,
+      "grad_norm": 0.8658422827720642,
+      "learning_rate": 5.0345761681491746e-08,
+      "loss": 2.6401,
+      "step": 9900
+    },
+    {
+      "epoch": 0.896261428442111,
+      "grad_norm": 0.8718020915985107,
+      "learning_rate": 4.934396342684e-08,
+      "loss": 2.5931,
+      "step": 9901
+    },
+    {
+      "epoch": 0.8963519507558613,
+      "grad_norm": 0.8454814553260803,
+      "learning_rate": 4.8352230200532366e-08,
+      "loss": 2.5815,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8964424730696117,
+      "grad_norm": 0.8844153881072998,
+      "learning_rate": 4.73705621024112e-08,
+      "loss": 2.6454,
+      "step": 9903
+    },
+    {
+      "epoch": 0.896532995383362,
+      "grad_norm": 0.7848690748214722,
+      "learning_rate": 4.639895923134185e-08,
+      "loss": 2.5818,
+      "step": 9904
+    },
+    {
+      "epoch": 0.8966235176971123,
+      "grad_norm": 0.8615793585777283,
+      "learning_rate": 4.543742168516829e-08,
+      "loss": 2.6557,
+      "step": 9905
+    },
+    {
+      "epoch": 0.8967140400108626,
+      "grad_norm": 0.924583375453949,
+      "learning_rate": 4.4485949560701955e-08,
+      "loss": 2.6373,
+      "step": 9906
+    },
+    {
+      "epoch": 0.896804562324613,
+      "grad_norm": 0.8576784133911133,
+      "learning_rate": 4.3544542953766196e-08,
+      "loss": 2.6288,
+      "step": 9907
+    },
+    {
+      "epoch": 0.8968950846383633,
+      "grad_norm": 0.9127756953239441,
+      "learning_rate": 4.2613201959162964e-08,
+      "loss": 2.6422,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8969856069521137,
+      "grad_norm": 0.8672379851341248,
+      "learning_rate": 4.169192667067279e-08,
+      "loss": 2.5805,
+      "step": 9909
+    },
+    {
+      "epoch": 0.897076129265864,
+      "grad_norm": 0.8124669194221497,
+      "learning_rate": 4.078071718107701e-08,
+      "loss": 2.5549,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8971666515796144,
+      "grad_norm": 0.9412040114402771,
+      "learning_rate": 3.987957358212446e-08,
+      "loss": 2.62,
+      "step": 9911
+    },
+    {
+      "epoch": 0.8972571738933647,
+      "grad_norm": 0.8409717679023743,
+      "learning_rate": 3.898849596456478e-08,
+      "loss": 2.6287,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8973476962071151,
+      "grad_norm": 0.8697944283485413,
+      "learning_rate": 3.810748441812617e-08,
+      "loss": 2.6004,
+      "step": 9913
+    },
+    {
+      "epoch": 0.8974382185208654,
+      "grad_norm": 0.8681595921516418,
+      "learning_rate": 3.723653903152657e-08,
+      "loss": 2.5771,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8975287408346158,
+      "grad_norm": 0.8452757000923157,
+      "learning_rate": 3.63756598924736e-08,
+      "loss": 2.6115,
+      "step": 9915
+    },
+    {
+      "epoch": 0.897619263148366,
+      "grad_norm": 0.8613131642341614,
+      "learning_rate": 3.552484708766457e-08,
+      "loss": 2.6312,
+      "step": 9916
+    },
+    {
+      "epoch": 0.8977097854621164,
+      "grad_norm": 0.8760083913803101,
+      "learning_rate": 3.468410070276429e-08,
+      "loss": 2.6234,
+      "step": 9917
+    },
+    {
+      "epoch": 0.8978003077758667,
+      "grad_norm": 0.8526800274848938,
+      "learning_rate": 3.3853420822438367e-08,
+      "loss": 2.5904,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8978908300896171,
+      "grad_norm": 0.7996740937232971,
+      "learning_rate": 3.3032807530331e-08,
+      "loss": 2.5986,
+      "step": 9919
+    },
+    {
+      "epoch": 0.8979813524033674,
+      "grad_norm": 0.865635335445404,
+      "learning_rate": 3.2222260909087196e-08,
+      "loss": 2.5428,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8980718747171178,
+      "grad_norm": 0.8549889922142029,
+      "learning_rate": 3.1421781040330555e-08,
+      "loss": 2.6553,
+      "step": 9921
+    },
+    {
+      "epoch": 0.8981623970308681,
+      "grad_norm": 0.8013193011283875,
+      "learning_rate": 3.0631368004663263e-08,
+      "loss": 2.5791,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8982529193446185,
+      "grad_norm": 0.8656326532363892,
+      "learning_rate": 2.985102188168831e-08,
+      "loss": 2.6103,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8983434416583688,
+      "grad_norm": 0.8901556730270386,
+      "learning_rate": 2.9080742749976188e-08,
+      "loss": 2.6084,
+      "step": 9924
+    },
+    {
+      "epoch": 0.8984339639721192,
+      "grad_norm": 0.7727611064910889,
+      "learning_rate": 2.8320530687098166e-08,
+      "loss": 2.5956,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8985244862858695,
+      "grad_norm": 0.9226417541503906,
+      "learning_rate": 2.7570385769604135e-08,
+      "loss": 2.6375,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8986150085996198,
+      "grad_norm": 0.8745355010032654,
+      "learning_rate": 2.683030807303366e-08,
+      "loss": 2.6177,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8987055309133701,
+      "grad_norm": 0.8516587615013123,
+      "learning_rate": 2.610029767191602e-08,
+      "loss": 2.6181,
+      "step": 9928
+    },
+    {
+      "epoch": 0.8987960532271205,
+      "grad_norm": 0.8466259241104126,
+      "learning_rate": 2.5380354639770176e-08,
+      "loss": 2.6352,
+      "step": 9929
+    },
+    {
+      "epoch": 0.8988865755408708,
+      "grad_norm": 0.8286177515983582,
+      "learning_rate": 2.4670479049082597e-08,
+      "loss": 2.5683,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8989770978546212,
+      "grad_norm": 0.9245983958244324,
+      "learning_rate": 2.397067097135164e-08,
+      "loss": 2.6485,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8990676201683715,
+      "grad_norm": 0.8311428427696228,
+      "learning_rate": 2.3280930477032058e-08,
+      "loss": 2.6075,
+      "step": 9932
+    },
+    {
+      "epoch": 0.8991581424821219,
+      "grad_norm": 0.8732925057411194,
+      "learning_rate": 2.2601257635579408e-08,
+      "loss": 2.6051,
+      "step": 9933
+    },
+    {
+      "epoch": 0.8992486647958722,
+      "grad_norm": 0.8483322858810425,
+      "learning_rate": 2.193165251545004e-08,
+      "loss": 2.5679,
+      "step": 9934
+    },
+    {
+      "epoch": 0.8993391871096226,
+      "grad_norm": 0.8530345559120178,
+      "learning_rate": 2.1272115184067797e-08,
+      "loss": 2.5918,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8994297094233729,
+      "grad_norm": 0.8855475783348083,
+      "learning_rate": 2.0622645707857326e-08,
+      "loss": 2.6783,
+      "step": 9936
+    },
+    {
+      "epoch": 0.8995202317371233,
+      "grad_norm": 0.9028958082199097,
+      "learning_rate": 1.9983244152199654e-08,
+      "loss": 2.6063,
+      "step": 9937
+    },
+    {
+      "epoch": 0.8996107540508735,
+      "grad_norm": 0.846435010433197,
+      "learning_rate": 1.9353910581498826e-08,
+      "loss": 2.5958,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8997012763646238,
+      "grad_norm": 0.9395351409912109,
+      "learning_rate": 1.8734645059115265e-08,
+      "loss": 2.6846,
+      "step": 9939
+    },
+    {
+      "epoch": 0.8997917986783742,
+      "grad_norm": 0.7862410545349121,
+      "learning_rate": 1.81254476474213e-08,
+      "loss": 2.5467,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8998823209921245,
+      "grad_norm": 0.8430321216583252,
+      "learning_rate": 1.752631840776786e-08,
+      "loss": 2.6722,
+      "step": 9941
+    },
+    {
+      "epoch": 0.8999728433058749,
+      "grad_norm": 0.8460858464241028,
+      "learning_rate": 1.693725740046226e-08,
+      "loss": 2.5722,
+      "step": 9942
+    },
+    {
+      "epoch": 0.9000633656196252,
+      "grad_norm": 0.7821448445320129,
+      "learning_rate": 1.6358264684857016e-08,
+      "loss": 2.5701,
+      "step": 9943
+    },
+    {
+      "epoch": 0.9001538879333756,
+      "grad_norm": 0.8697470426559448,
+      "learning_rate": 1.578934031922774e-08,
+      "loss": 2.6128,
+      "step": 9944
+    },
+    {
+      "epoch": 0.9002444102471259,
+      "grad_norm": 0.8176562190055847,
+      "learning_rate": 1.5230484360873044e-08,
+      "loss": 2.5756,
+      "step": 9945
+    },
+    {
+      "epoch": 0.9003349325608763,
+      "grad_norm": 0.8593572974205017,
+      "learning_rate": 1.4681696866081229e-08,
+      "loss": 2.6,
+      "step": 9946
+    },
+    {
+      "epoch": 0.9004254548746266,
+      "grad_norm": 0.7897228002548218,
+      "learning_rate": 1.41429778901081e-08,
+      "loss": 2.6036,
+      "step": 9947
+    },
+    {
+      "epoch": 0.900515977188377,
+      "grad_norm": 0.851230800151825,
+      "learning_rate": 1.361432748718805e-08,
+      "loss": 2.5607,
+      "step": 9948
+    },
+    {
+      "epoch": 0.9006064995021272,
+      "grad_norm": 0.855710506439209,
+      "learning_rate": 1.3095745710578477e-08,
+      "loss": 2.636,
+      "step": 9949
+    },
+    {
+      "epoch": 0.9006970218158776,
+      "grad_norm": 0.796705961227417,
+      "learning_rate": 1.2587232612493172e-08,
+      "loss": 2.5435,
+      "step": 9950
+    },
+    {
+      "epoch": 0.9007875441296279,
+      "grad_norm": 0.7984212040901184,
+      "learning_rate": 1.2088788244135618e-08,
+      "loss": 2.5907,
+      "step": 9951
+    },
+    {
+      "epoch": 0.9008780664433783,
+      "grad_norm": 0.8091149926185608,
+      "learning_rate": 1.1600412655710102e-08,
+      "loss": 2.6241,
+      "step": 9952
+    },
+    {
+      "epoch": 0.9009685887571286,
+      "grad_norm": 0.828718900680542,
+      "learning_rate": 1.1122105896377299e-08,
+      "loss": 2.6062,
+      "step": 9953
+    },
+    {
+      "epoch": 0.901059111070879,
+      "grad_norm": 0.8733603358268738,
+      "learning_rate": 1.0653868014309786e-08,
+      "loss": 2.5837,
+      "step": 9954
+    },
+    {
+      "epoch": 0.9011496333846293,
+      "grad_norm": 0.9175240993499756,
+      "learning_rate": 1.0195699056669838e-08,
+      "loss": 2.6383,
+      "step": 9955
+    },
+    {
+      "epoch": 0.9012401556983797,
+      "grad_norm": 0.8264015913009644,
+      "learning_rate": 9.747599069576119e-09,
+      "loss": 2.6036,
+      "step": 9956
+    },
+    {
+      "epoch": 0.90133067801213,
+      "grad_norm": 0.8627322912216187,
+      "learning_rate": 9.309568098170296e-09,
+      "loss": 2.5881,
+      "step": 9957
+    },
+    {
+      "epoch": 0.9014212003258804,
+      "grad_norm": 0.8578746914863586,
+      "learning_rate": 8.881606186561531e-09,
+      "loss": 2.6102,
+      "step": 9958
+    },
+    {
+      "epoch": 0.9015117226396306,
+      "grad_norm": 0.8336324095726013,
+      "learning_rate": 8.463713377826476e-09,
+      "loss": 2.6608,
+      "step": 9959
+    },
+    {
+      "epoch": 0.901602244953381,
+      "grad_norm": 0.8321331143379211,
+      "learning_rate": 8.055889714064791e-09,
+      "loss": 2.5883,
+      "step": 9960
+    },
+    {
+      "epoch": 0.9016927672671313,
+      "grad_norm": 0.822594165802002,
+      "learning_rate": 7.658135236343623e-09,
+      "loss": 2.6386,
+      "step": 9961
+    },
+    {
+      "epoch": 0.9017832895808817,
+      "grad_norm": 0.9133470058441162,
+      "learning_rate": 7.270449984708716e-09,
+      "loss": 2.6727,
+      "step": 9962
+    },
+    {
+      "epoch": 0.901873811894632,
+      "grad_norm": 0.8246998190879822,
+      "learning_rate": 6.892833998206616e-09,
+      "loss": 2.6284,
+      "step": 9963
+    },
+    {
+      "epoch": 0.9019643342083824,
+      "grad_norm": 0.9324564337730408,
+      "learning_rate": 6.525287314851358e-09,
+      "loss": 2.698,
+      "step": 9964
+    },
+    {
+      "epoch": 0.9020548565221327,
+      "grad_norm": 0.9247385859489441,
+      "learning_rate": 6.167809971668881e-09,
+      "loss": 2.6434,
+      "step": 9965
+    },
+    {
+      "epoch": 0.9021453788358831,
+      "grad_norm": 0.8682430982589722,
+      "learning_rate": 5.820402004652614e-09,
+      "loss": 2.6659,
+      "step": 9966
+    },
+    {
+      "epoch": 0.9022359011496334,
+      "grad_norm": 0.8353280425071716,
+      "learning_rate": 5.483063448785686e-09,
+      "loss": 2.6197,
+      "step": 9967
+    },
+    {
+      "epoch": 0.9023264234633838,
+      "grad_norm": 0.8058872222900391,
+      "learning_rate": 5.15579433804092e-09,
+      "loss": 2.5601,
+      "step": 9968
+    },
+    {
+      "epoch": 0.902416945777134,
+      "grad_norm": 0.9338617324829102,
+      "learning_rate": 4.838594705369737e-09,
+      "loss": 2.6476,
+      "step": 9969
+    },
+    {
+      "epoch": 0.9025074680908844,
+      "grad_norm": 0.8994864821434021,
+      "learning_rate": 4.531464582713252e-09,
+      "loss": 2.66,
+      "step": 9970
+    },
+    {
+      "epoch": 0.9025979904046347,
+      "grad_norm": 0.8840702772140503,
+      "learning_rate": 4.234404001002279e-09,
+      "loss": 2.578,
+      "step": 9971
+    },
+    {
+      "epoch": 0.9026885127183851,
+      "grad_norm": 0.9012619256973267,
+      "learning_rate": 3.947412990157329e-09,
+      "loss": 2.6063,
+      "step": 9972
+    },
+    {
+      "epoch": 0.9027790350321354,
+      "grad_norm": 0.8753988742828369,
+      "learning_rate": 3.670491579066404e-09,
+      "loss": 2.6057,
+      "step": 9973
+    },
+    {
+      "epoch": 0.9028695573458858,
+      "grad_norm": 0.8586761951446533,
+      "learning_rate": 3.4036397956183076e-09,
+      "loss": 2.6562,
+      "step": 9974
+    },
+    {
+      "epoch": 0.9029600796596361,
+      "grad_norm": 0.7856614589691162,
+      "learning_rate": 3.1468576666915383e-09,
+      "loss": 2.6084,
+      "step": 9975
+    },
+    {
+      "epoch": 0.9030506019733865,
+      "grad_norm": 0.837960958480835,
+      "learning_rate": 2.900145218143191e-09,
+      "loss": 2.648,
+      "step": 9976
+    },
+    {
+      "epoch": 0.9031411242871368,
+      "grad_norm": 0.9321572184562683,
+      "learning_rate": 2.6635024748089545e-09,
+      "loss": 2.6036,
+      "step": 9977
+    },
+    {
+      "epoch": 0.9032316466008872,
+      "grad_norm": 0.8802992701530457,
+      "learning_rate": 2.4369294605253166e-09,
+      "loss": 2.5803,
+      "step": 9978
+    },
+    {
+      "epoch": 0.9033221689146375,
+      "grad_norm": 0.8097015619277954,
+      "learning_rate": 2.2204261981073615e-09,
+      "loss": 2.6154,
+      "step": 9979
+    },
+    {
+      "epoch": 0.9034126912283877,
+      "grad_norm": 0.8596858978271484,
+      "learning_rate": 2.0139927093487664e-09,
+      "loss": 2.6293,
+      "step": 9980
+    },
+    {
+      "epoch": 0.9035032135421381,
+      "grad_norm": 0.8549774289131165,
+      "learning_rate": 1.8176290150551112e-09,
+      "loss": 2.6386,
+      "step": 9981
+    },
+    {
+      "epoch": 0.9035937358558884,
+      "grad_norm": 0.9319025278091431,
+      "learning_rate": 1.6313351349883655e-09,
+      "loss": 2.6025,
+      "step": 9982
+    },
+    {
+      "epoch": 0.9036842581696388,
+      "grad_norm": 0.8225411772727966,
+      "learning_rate": 1.455111087900196e-09,
+      "loss": 2.5595,
+      "step": 9983
+    },
+    {
+      "epoch": 0.9037747804833891,
+      "grad_norm": 0.8542320132255554,
+      "learning_rate": 1.2889568915541717e-09,
+      "loss": 2.6098,
+      "step": 9984
+    },
+    {
+      "epoch": 0.9038653027971395,
+      "grad_norm": 0.9284166693687439,
+      "learning_rate": 1.132872562681353e-09,
+      "loss": 2.6031,
+      "step": 9985
+    },
+    {
+      "epoch": 0.9039558251108898,
+      "grad_norm": 0.8331528902053833,
+      "learning_rate": 9.868581169802938e-10,
+      "loss": 2.6571,
+      "step": 9986
+    },
+    {
+      "epoch": 0.9040463474246402,
+      "grad_norm": 0.7966681122779846,
+      "learning_rate": 8.509135691725512e-10,
+      "loss": 2.6511,
+      "step": 9987
+    },
+    {
+      "epoch": 0.9041368697383905,
+      "grad_norm": 0.8282911777496338,
+      "learning_rate": 7.250389329471751e-10,
+      "loss": 2.5352,
+      "step": 9988
+    },
+    {
+      "epoch": 0.9042273920521409,
+      "grad_norm": 0.9159926176071167,
+      "learning_rate": 6.092342209607083e-10,
+      "loss": 2.5998,
+      "step": 9989
+    },
+    {
+      "epoch": 0.9043179143658912,
+      "grad_norm": 0.8755436539649963,
+      "learning_rate": 5.034994448926967e-10,
+      "loss": 2.6352,
+      "step": 9990
+    },
+    {
+      "epoch": 0.9044084366796415,
+      "grad_norm": 0.81312096118927,
+      "learning_rate": 4.078346153901791e-10,
+      "loss": 2.6202,
+      "step": 9991
+    },
+    {
+      "epoch": 0.9044989589933918,
+      "grad_norm": 0.7991079688072205,
+      "learning_rate": 3.2223974207878925e-10,
+      "loss": 2.6106,
+      "step": 9992
+    },
+    {
+      "epoch": 0.9045894813071422,
+      "grad_norm": 0.7968497276306152,
+      "learning_rate": 2.4671483358496007e-10,
+      "loss": 2.5344,
+      "step": 9993
+    },
+    {
+      "epoch": 0.9046800036208925,
+      "grad_norm": 0.7982305884361267,
+      "learning_rate": 1.812598975137192e-10,
+      "loss": 2.6351,
+      "step": 9994
+    },
+    {
+      "epoch": 0.9047705259346429,
+      "grad_norm": 0.8206318020820618,
+      "learning_rate": 1.258749404486892e-10,
+      "loss": 2.6411,
+      "step": 9995
+    },
+    {
+      "epoch": 0.9048610482483932,
+      "grad_norm": 0.7778028249740601,
+      "learning_rate": 8.055996797429188e-11,
+      "loss": 2.6404,
+      "step": 9996
+    },
+    {
+      "epoch": 0.9049515705621436,
+      "grad_norm": 0.8508470058441162,
+      "learning_rate": 4.531498464244166e-11,
+      "loss": 2.5897,
+      "step": 9997
+    },
+    {
+      "epoch": 0.9050420928758939,
+      "grad_norm": 0.8271351456642151,
+      "learning_rate": 2.0139994016954434e-11,
+      "loss": 2.5587,
+      "step": 9998
+    },
+    {
+      "epoch": 0.9051326151896443,
+      "grad_norm": 0.8943343758583069,
+      "learning_rate": 5.034998629138698e-12,
+      "loss": 2.5531,
+      "step": 9999
+    },
+    {
+      "epoch": 0.9052231375033946,
+      "grad_norm": 0.8616811633110046,
+      "learning_rate": 0.0,
+      "loss": 2.6833,
+      "step": 10000
+    },
+    {
+      "epoch": 0.9052231375033946,
+      "eval_loss": 2.5555005073547363,
+      "eval_runtime": 71.4851,
+      "eval_samples_per_second": 37.812,
+      "eval_steps_per_second": 3.161,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 10000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1494808064e+17,
+  "train_batch_size": 12,
+  "trial_name": null,
+  "trial_params": null
+}