diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,47477 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.6447368421052633,
+  "eval_steps": 30,
+  "global_step": 6500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00025303643724696357,
+      "grad_norm": 9.6875,
+      "learning_rate": 7.591093117408908e-09,
+      "loss": 3.9346,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005060728744939271,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.5182186234817815e-08,
+      "loss": 3.9477,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007591093117408907,
+      "grad_norm": 9.125,
+      "learning_rate": 2.277327935222672e-08,
+      "loss": 3.8734,
+      "step": 3
+    },
+    {
+      "epoch": 0.0010121457489878543,
+      "grad_norm": 9.5625,
+      "learning_rate": 3.036437246963563e-08,
+      "loss": 3.8829,
+      "step": 4
+    },
+    {
+      "epoch": 0.0012651821862348178,
+      "grad_norm": 9.5625,
+      "learning_rate": 3.795546558704453e-08,
+      "loss": 3.9816,
+      "step": 5
+    },
+    {
+      "epoch": 0.0015182186234817814,
+      "grad_norm": 9.625,
+      "learning_rate": 4.554655870445344e-08,
+      "loss": 3.8713,
+      "step": 6
+    },
+    {
+      "epoch": 0.001771255060728745,
+      "grad_norm": 9.9375,
+      "learning_rate": 5.313765182186235e-08,
+      "loss": 3.978,
+      "step": 7
+    },
+    {
+      "epoch": 0.0020242914979757085,
+      "grad_norm": 9.9375,
+      "learning_rate": 6.072874493927126e-08,
+      "loss": 4.0109,
+      "step": 8
+    },
+    {
+      "epoch": 0.002277327935222672,
+      "grad_norm": 9.75,
+      "learning_rate": 6.831983805668016e-08,
+      "loss": 3.9153,
+      "step": 9
+    },
+    {
+      "epoch": 0.0025303643724696357,
+      "grad_norm": 9.375,
+      "learning_rate": 7.591093117408907e-08,
+      "loss": 3.8863,
+      "step": 10
+    },
+    {
+      "epoch": 0.002783400809716599,
+      "grad_norm": 9.3125,
+      "learning_rate": 8.350202429149797e-08,
+      "loss": 3.8835,
+      "step": 11
+    },
+    {
+      "epoch": 0.003036437246963563,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.109311740890688e-08,
+      "loss": 3.931,
+      "step": 12
+    },
+    {
+      "epoch": 0.003289473684210526,
+      "grad_norm": 9.625,
+      "learning_rate": 9.868421052631579e-08,
+      "loss": 4.0463,
+      "step": 13
+    },
+    {
+      "epoch": 0.00354251012145749,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.062753036437247e-07,
+      "loss": 4.0129,
+      "step": 14
+    },
+    {
+      "epoch": 0.0037955465587044533,
+      "grad_norm": 10.0,
+      "learning_rate": 1.138663967611336e-07,
+      "loss": 3.9607,
+      "step": 15
+    },
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 9.625,
+      "learning_rate": 1.2145748987854252e-07,
+      "loss": 3.8953,
+      "step": 16
+    },
+    {
+      "epoch": 0.00430161943319838,
+      "grad_norm": 9.5,
+      "learning_rate": 1.2904858299595142e-07,
+      "loss": 3.9134,
+      "step": 17
+    },
+    {
+      "epoch": 0.004554655870445344,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.3663967611336033e-07,
+      "loss": 4.054,
+      "step": 18
+    },
+    {
+      "epoch": 0.004807692307692308,
+      "grad_norm": 10.0625,
+      "learning_rate": 1.4423076923076925e-07,
+      "loss": 4.0016,
+      "step": 19
+    },
+    {
+      "epoch": 0.005060728744939271,
+      "grad_norm": 9.875,
+      "learning_rate": 1.5182186234817813e-07,
+      "loss": 3.7957,
+      "step": 20
+    },
+    {
+      "epoch": 0.005313765182186235,
+      "grad_norm": 9.25,
+      "learning_rate": 1.5941295546558703e-07,
+      "loss": 3.8804,
+      "step": 21
+    },
+    {
+      "epoch": 0.005566801619433198,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.6700404858299594e-07,
+      "loss": 4.021,
+      "step": 22
+    },
+    {
+      "epoch": 0.005819838056680162,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.7459514170040487e-07,
+      "loss": 3.937,
+      "step": 23
+    },
+    {
+      "epoch": 0.006072874493927126,
+      "grad_norm": 9.75,
+      "learning_rate": 1.8218623481781377e-07,
+      "loss": 3.9381,
+      "step": 24
+    },
+    {
+      "epoch": 0.006325910931174089,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.8977732793522267e-07,
+      "loss": 3.7875,
+      "step": 25
+    },
+    {
+      "epoch": 0.006578947368421052,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.9736842105263157e-07,
+      "loss": 4.0147,
+      "step": 26
+    },
+    {
+      "epoch": 0.0068319838056680165,
+      "grad_norm": 10.0,
+      "learning_rate": 2.049595141700405e-07,
+      "loss": 3.9926,
+      "step": 27
+    },
+    {
+      "epoch": 0.00708502024291498,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.125506072874494e-07,
+      "loss": 3.7789,
+      "step": 28
+    },
+    {
+      "epoch": 0.007338056680161943,
+      "grad_norm": 9.0,
+      "learning_rate": 2.201417004048583e-07,
+      "loss": 3.6699,
+      "step": 29
+    },
+    {
+      "epoch": 0.0075910931174089065,
+      "grad_norm": 8.375,
+      "learning_rate": 2.277327935222672e-07,
+      "loss": 3.8175,
+      "step": 30
+    },
+    {
+      "epoch": 0.0075910931174089065,
+      "eval_loss": 4.026721000671387,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.8529,
+      "eval_samples_per_second": 9.707,
+      "eval_steps_per_second": 9.707,
+      "step": 30
+    },
+    {
+      "epoch": 0.00784412955465587,
+      "grad_norm": 9.3125,
+      "learning_rate": 2.353238866396761e-07,
+      "loss": 3.9732,
+      "step": 31
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 9.25,
+      "learning_rate": 2.4291497975708504e-07,
+      "loss": 3.9454,
+      "step": 32
+    },
+    {
+      "epoch": 0.008350202429149798,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.5050607287449394e-07,
+      "loss": 3.7822,
+      "step": 33
+    },
+    {
+      "epoch": 0.00860323886639676,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.5809716599190285e-07,
+      "loss": 3.9836,
+      "step": 34
+    },
+    {
+      "epoch": 0.008856275303643725,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.6568825910931175e-07,
+      "loss": 3.7134,
+      "step": 35
+    },
+    {
+      "epoch": 0.009109311740890687,
+      "grad_norm": 9.4375,
+      "learning_rate": 2.7327935222672065e-07,
+      "loss": 3.9914,
+      "step": 36
+    },
+    {
+      "epoch": 0.009362348178137652,
+      "grad_norm": 9.5,
+      "learning_rate": 2.8087044534412955e-07,
+      "loss": 3.8909,
+      "step": 37
+    },
+    {
+      "epoch": 0.009615384615384616,
+      "grad_norm": 8.5,
+      "learning_rate": 2.884615384615385e-07,
+      "loss": 3.7381,
+      "step": 38
+    },
+    {
+      "epoch": 0.009868421052631578,
+      "grad_norm": 9.25,
+      "learning_rate": 2.9605263157894736e-07,
+      "loss": 3.9282,
+      "step": 39
+    },
+    {
+      "epoch": 0.010121457489878543,
+      "grad_norm": 9.1875,
+      "learning_rate": 3.0364372469635626e-07,
+      "loss": 3.9731,
+      "step": 40
+    },
+    {
+      "epoch": 0.010374493927125507,
+      "grad_norm": 9.125,
+      "learning_rate": 3.112348178137652e-07,
+      "loss": 3.943,
+      "step": 41
+    },
+    {
+      "epoch": 0.01062753036437247,
+      "grad_norm": 8.9375,
+      "learning_rate": 3.1882591093117407e-07,
+      "loss": 3.7826,
+      "step": 42
+    },
+    {
+      "epoch": 0.010880566801619434,
+      "grad_norm": 9.3125,
+      "learning_rate": 3.26417004048583e-07,
+      "loss": 3.9675,
+      "step": 43
+    },
+    {
+      "epoch": 0.011133603238866396,
+      "grad_norm": 9.0625,
+      "learning_rate": 3.3400809716599187e-07,
+      "loss": 3.8407,
+      "step": 44
+    },
+    {
+      "epoch": 0.01138663967611336,
+      "grad_norm": 8.75,
+      "learning_rate": 3.4159919028340083e-07,
+      "loss": 3.8558,
+      "step": 45
+    },
+    {
+      "epoch": 0.011639676113360324,
+      "grad_norm": 8.625,
+      "learning_rate": 3.4919028340080973e-07,
+      "loss": 3.8936,
+      "step": 46
+    },
+    {
+      "epoch": 0.011892712550607287,
+      "grad_norm": 9.0625,
+      "learning_rate": 3.5678137651821863e-07,
+      "loss": 3.8767,
+      "step": 47
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 9.0,
+      "learning_rate": 3.6437246963562754e-07,
+      "loss": 3.9194,
+      "step": 48
+    },
+    {
+      "epoch": 0.012398785425101215,
+      "grad_norm": 8.75,
+      "learning_rate": 3.719635627530365e-07,
+      "loss": 3.8251,
+      "step": 49
+    },
+    {
+      "epoch": 0.012651821862348178,
+      "grad_norm": 8.625,
+      "learning_rate": 3.7955465587044534e-07,
+      "loss": 3.8566,
+      "step": 50
+    },
+    {
+      "epoch": 0.012904858299595142,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.871457489878543e-07,
+      "loss": 3.6928,
+      "step": 51
+    },
+    {
+      "epoch": 0.013157894736842105,
+      "grad_norm": 8.8125,
+      "learning_rate": 3.9473684210526315e-07,
+      "loss": 3.9343,
+      "step": 52
+    },
+    {
+      "epoch": 0.013410931174089069,
+      "grad_norm": 8.75,
+      "learning_rate": 4.0232793522267205e-07,
+      "loss": 3.8763,
+      "step": 53
+    },
+    {
+      "epoch": 0.013663967611336033,
+      "grad_norm": 8.125,
+      "learning_rate": 4.09919028340081e-07,
+      "loss": 3.7227,
+      "step": 54
+    },
+    {
+      "epoch": 0.013917004048582995,
+      "grad_norm": 8.25,
+      "learning_rate": 4.1751012145748985e-07,
+      "loss": 3.8447,
+      "step": 55
+    },
+    {
+      "epoch": 0.01417004048582996,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.251012145748988e-07,
+      "loss": 3.8632,
+      "step": 56
+    },
+    {
+      "epoch": 0.014423076923076924,
+      "grad_norm": 8.5,
+      "learning_rate": 4.326923076923077e-07,
+      "loss": 3.8124,
+      "step": 57
+    },
+    {
+      "epoch": 0.014676113360323886,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.402834008097166e-07,
+      "loss": 3.7154,
+      "step": 58
+    },
+    {
+      "epoch": 0.01492914979757085,
+      "grad_norm": 8.25,
+      "learning_rate": 4.478744939271255e-07,
+      "loss": 3.8865,
+      "step": 59
+    },
+    {
+      "epoch": 0.015182186234817813,
+      "grad_norm": 8.375,
+      "learning_rate": 4.554655870445344e-07,
+      "loss": 3.8622,
+      "step": 60
+    },
+    {
+      "epoch": 0.015182186234817813,
+      "eval_loss": 3.8989968299865723,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.7706,
+      "eval_samples_per_second": 9.746,
+      "eval_steps_per_second": 9.746,
+      "step": 60
+    },
+    {
+      "epoch": 0.015435222672064777,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.630566801619433e-07,
+      "loss": 3.5651,
+      "step": 61
+    },
+    {
+      "epoch": 0.01568825910931174,
+      "grad_norm": 8.0,
+      "learning_rate": 4.706477732793522e-07,
+      "loss": 3.7991,
+      "step": 62
+    },
+    {
+      "epoch": 0.015941295546558706,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.782388663967612e-07,
+      "loss": 3.8332,
+      "step": 63
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.858299595141701e-07,
+      "loss": 3.7036,
+      "step": 64
+    },
+    {
+      "epoch": 0.01644736842105263,
+      "grad_norm": 8.75,
+      "learning_rate": 4.934210526315789e-07,
+      "loss": 3.8723,
+      "step": 65
+    },
+    {
+      "epoch": 0.016700404858299597,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.010121457489879e-07,
+      "loss": 3.811,
+      "step": 66
+    },
+    {
+      "epoch": 0.01695344129554656,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.086032388663967e-07,
+      "loss": 3.806,
+      "step": 67
+    },
+    {
+      "epoch": 0.01720647773279352,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.161943319838057e-07,
+      "loss": 3.8396,
+      "step": 68
+    },
+    {
+      "epoch": 0.017459514170040488,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.237854251012146e-07,
+      "loss": 3.8137,
+      "step": 69
+    },
+    {
+      "epoch": 0.01771255060728745,
+      "grad_norm": 7.46875,
+      "learning_rate": 5.313765182186235e-07,
+      "loss": 3.7385,
+      "step": 70
+    },
+    {
+      "epoch": 0.017965587044534412,
+      "grad_norm": 7.78125,
+      "learning_rate": 5.389676113360323e-07,
+      "loss": 3.8198,
+      "step": 71
+    },
+    {
+      "epoch": 0.018218623481781375,
+      "grad_norm": 7.1875,
+      "learning_rate": 5.465587044534413e-07,
+      "loss": 3.5681,
+      "step": 72
+    },
+    {
+      "epoch": 0.01847165991902834,
+      "grad_norm": 7.28125,
+      "learning_rate": 5.541497975708503e-07,
+      "loss": 3.729,
+      "step": 73
+    },
+    {
+      "epoch": 0.018724696356275303,
+      "grad_norm": 6.75,
+      "learning_rate": 5.617408906882591e-07,
+      "loss": 3.6065,
+      "step": 74
+    },
+    {
+      "epoch": 0.018977732793522266,
+      "grad_norm": 7.4375,
+      "learning_rate": 5.69331983805668e-07,
+      "loss": 3.7124,
+      "step": 75
+    },
+    {
+      "epoch": 0.019230769230769232,
+      "grad_norm": 7.21875,
+      "learning_rate": 5.76923076923077e-07,
+      "loss": 3.7887,
+      "step": 76
+    },
+    {
+      "epoch": 0.019483805668016194,
+      "grad_norm": 7.46875,
+      "learning_rate": 5.845141700404859e-07,
+      "loss": 3.8091,
+      "step": 77
+    },
+    {
+      "epoch": 0.019736842105263157,
+      "grad_norm": 7.3125,
+      "learning_rate": 5.921052631578947e-07,
+      "loss": 3.7451,
+      "step": 78
+    },
+    {
+      "epoch": 0.019989878542510123,
+      "grad_norm": 6.5,
+      "learning_rate": 5.996963562753037e-07,
+      "loss": 3.4476,
+      "step": 79
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 6.5,
+      "learning_rate": 6.072874493927125e-07,
+      "loss": 3.4508,
+      "step": 80
+    },
+    {
+      "epoch": 0.020495951417004048,
+      "grad_norm": 6.875,
+      "learning_rate": 6.148785425101215e-07,
+      "loss": 3.7592,
+      "step": 81
+    },
+    {
+      "epoch": 0.020748987854251014,
+      "grad_norm": 6.4375,
+      "learning_rate": 6.224696356275304e-07,
+      "loss": 3.661,
+      "step": 82
+    },
+    {
+      "epoch": 0.021002024291497976,
+      "grad_norm": 7.15625,
+      "learning_rate": 6.300607287449393e-07,
+      "loss": 3.7193,
+      "step": 83
+    },
+    {
+      "epoch": 0.02125506072874494,
+      "grad_norm": 7.0,
+      "learning_rate": 6.376518218623481e-07,
+      "loss": 3.6798,
+      "step": 84
+    },
+    {
+      "epoch": 0.0215080971659919,
+      "grad_norm": 6.65625,
+      "learning_rate": 6.452429149797571e-07,
+      "loss": 3.5788,
+      "step": 85
+    },
+    {
+      "epoch": 0.021761133603238867,
+      "grad_norm": 6.96875,
+      "learning_rate": 6.52834008097166e-07,
+      "loss": 3.721,
+      "step": 86
+    },
+    {
+      "epoch": 0.02201417004048583,
+      "grad_norm": 6.25,
+      "learning_rate": 6.604251012145749e-07,
+      "loss": 3.2765,
+      "step": 87
+    },
+    {
+      "epoch": 0.022267206477732792,
+      "grad_norm": 7.15625,
+      "learning_rate": 6.680161943319837e-07,
+      "loss": 3.5858,
+      "step": 88
+    },
+    {
+      "epoch": 0.022520242914979758,
+      "grad_norm": 6.5,
+      "learning_rate": 6.756072874493928e-07,
+      "loss": 3.6337,
+      "step": 89
+    },
+    {
+      "epoch": 0.02277327935222672,
+      "grad_norm": 6.34375,
+      "learning_rate": 6.831983805668017e-07,
+      "loss": 3.6173,
+      "step": 90
+    },
+    {
+      "epoch": 0.02277327935222672,
+      "eval_loss": 3.722729444503784,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.189,
+      "eval_samples_per_second": 9.731,
+      "eval_steps_per_second": 9.731,
+      "step": 90
+    },
+    {
+      "epoch": 0.023026315789473683,
+      "grad_norm": 6.59375,
+      "learning_rate": 6.907894736842105e-07,
+      "loss": 3.674,
+      "step": 91
+    },
+    {
+      "epoch": 0.02327935222672065,
+      "grad_norm": 6.5,
+      "learning_rate": 6.983805668016195e-07,
+      "loss": 3.6546,
+      "step": 92
+    },
+    {
+      "epoch": 0.02353238866396761,
+      "grad_norm": 5.90625,
+      "learning_rate": 7.059716599190283e-07,
+      "loss": 3.6028,
+      "step": 93
+    },
+    {
+      "epoch": 0.023785425101214574,
+      "grad_norm": 6.25,
+      "learning_rate": 7.135627530364373e-07,
+      "loss": 3.5596,
+      "step": 94
+    },
+    {
+      "epoch": 0.02403846153846154,
+      "grad_norm": 6.25,
+      "learning_rate": 7.211538461538462e-07,
+      "loss": 3.6763,
+      "step": 95
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 5.65625,
+      "learning_rate": 7.287449392712551e-07,
+      "loss": 3.5643,
+      "step": 96
+    },
+    {
+      "epoch": 0.024544534412955465,
+      "grad_norm": 5.90625,
+      "learning_rate": 7.363360323886639e-07,
+      "loss": 3.6113,
+      "step": 97
+    },
+    {
+      "epoch": 0.02479757085020243,
+      "grad_norm": 6.25,
+      "learning_rate": 7.43927125506073e-07,
+      "loss": 3.6049,
+      "step": 98
+    },
+    {
+      "epoch": 0.025050607287449393,
+      "grad_norm": 5.90625,
+      "learning_rate": 7.515182186234818e-07,
+      "loss": 3.6362,
+      "step": 99
+    },
+    {
+      "epoch": 0.025303643724696356,
+      "grad_norm": 5.40625,
+      "learning_rate": 7.591093117408907e-07,
+      "loss": 3.5375,
+      "step": 100
+    },
+    {
+      "epoch": 0.025556680161943318,
+      "grad_norm": 6.1875,
+      "learning_rate": 7.667004048582995e-07,
+      "loss": 3.6903,
+      "step": 101
+    },
+    {
+      "epoch": 0.025809716599190284,
+      "grad_norm": 5.75,
+      "learning_rate": 7.742914979757086e-07,
+      "loss": 3.6619,
+      "step": 102
+    },
+    {
+      "epoch": 0.026062753036437247,
+      "grad_norm": 5.71875,
+      "learning_rate": 7.818825910931174e-07,
+      "loss": 3.718,
+      "step": 103
+    },
+    {
+      "epoch": 0.02631578947368421,
+      "grad_norm": 4.65625,
+      "learning_rate": 7.894736842105263e-07,
+      "loss": 3.3288,
+      "step": 104
+    },
+    {
+      "epoch": 0.026568825910931175,
+      "grad_norm": 5.84375,
+      "learning_rate": 7.970647773279352e-07,
+      "loss": 3.589,
+      "step": 105
+    },
+    {
+      "epoch": 0.026821862348178137,
+      "grad_norm": 5.125,
+      "learning_rate": 8.046558704453441e-07,
+      "loss": 3.4999,
+      "step": 106
+    },
+    {
+      "epoch": 0.0270748987854251,
+      "grad_norm": 5.5625,
+      "learning_rate": 8.12246963562753e-07,
+      "loss": 3.5566,
+      "step": 107
+    },
+    {
+      "epoch": 0.027327935222672066,
+      "grad_norm": 5.21875,
+      "learning_rate": 8.19838056680162e-07,
+      "loss": 3.6094,
+      "step": 108
+    },
+    {
+      "epoch": 0.02758097165991903,
+      "grad_norm": 5.53125,
+      "learning_rate": 8.274291497975709e-07,
+      "loss": 3.5878,
+      "step": 109
+    },
+    {
+      "epoch": 0.02783400809716599,
+      "grad_norm": 5.03125,
+      "learning_rate": 8.350202429149797e-07,
+      "loss": 3.4998,
+      "step": 110
+    },
+    {
+      "epoch": 0.028087044534412957,
+      "grad_norm": 4.9375,
+      "learning_rate": 8.426113360323888e-07,
+      "loss": 3.582,
+      "step": 111
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 5.1875,
+      "learning_rate": 8.502024291497976e-07,
+      "loss": 3.6525,
+      "step": 112
+    },
+    {
+      "epoch": 0.028593117408906882,
+      "grad_norm": 5.25,
+      "learning_rate": 8.577935222672065e-07,
+      "loss": 3.5774,
+      "step": 113
+    },
+    {
+      "epoch": 0.028846153846153848,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.653846153846154e-07,
+      "loss": 3.5114,
+      "step": 114
+    },
+    {
+      "epoch": 0.02909919028340081,
+      "grad_norm": 4.75,
+      "learning_rate": 8.729757085020244e-07,
+      "loss": 3.5236,
+      "step": 115
+    },
+    {
+      "epoch": 0.029352226720647773,
+      "grad_norm": 4.59375,
+      "learning_rate": 8.805668016194332e-07,
+      "loss": 3.5327,
+      "step": 116
+    },
+    {
+      "epoch": 0.029605263157894735,
+      "grad_norm": 4.90625,
+      "learning_rate": 8.881578947368421e-07,
+      "loss": 3.5597,
+      "step": 117
+    },
+    {
+      "epoch": 0.0298582995951417,
+      "grad_norm": 4.71875,
+      "learning_rate": 8.95748987854251e-07,
+      "loss": 3.6605,
+      "step": 118
+    },
+    {
+      "epoch": 0.030111336032388664,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.033400809716599e-07,
+      "loss": 3.5254,
+      "step": 119
+    },
+    {
+      "epoch": 0.030364372469635626,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.109311740890688e-07,
+      "loss": 3.5048,
+      "step": 120
+    },
+    {
+      "epoch": 0.030364372469635626,
+      "eval_loss": 3.594075918197632,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.9103,
+      "eval_samples_per_second": 9.741,
+      "eval_steps_per_second": 9.741,
+      "step": 120
+    },
+    {
+      "epoch": 0.030617408906882592,
+      "grad_norm": 4.5,
+      "learning_rate": 9.185222672064778e-07,
+      "loss": 3.5779,
+      "step": 121
+    },
+    {
+      "epoch": 0.030870445344129555,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.261133603238866e-07,
+      "loss": 3.5045,
+      "step": 122
+    },
+    {
+      "epoch": 0.031123481781376517,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.337044534412955e-07,
+      "loss": 3.483,
+      "step": 123
+    },
+    {
+      "epoch": 0.03137651821862348,
+      "grad_norm": 4.75,
+      "learning_rate": 9.412955465587044e-07,
+      "loss": 3.6197,
+      "step": 124
+    },
+    {
+      "epoch": 0.031629554655870445,
+      "grad_norm": 4.375,
+      "learning_rate": 9.488866396761134e-07,
+      "loss": 3.5454,
+      "step": 125
+    },
+    {
+      "epoch": 0.03188259109311741,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.564777327935225e-07,
+      "loss": 3.448,
+      "step": 126
+    },
+    {
+      "epoch": 0.03213562753036437,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.64068825910931e-07,
+      "loss": 3.3637,
+      "step": 127
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 4.5,
+      "learning_rate": 9.716599190283402e-07,
+      "loss": 3.4938,
+      "step": 128
+    },
+    {
+      "epoch": 0.0326417004048583,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.79251012145749e-07,
+      "loss": 3.548,
+      "step": 129
+    },
+    {
+      "epoch": 0.03289473684210526,
+      "grad_norm": 4.125,
+      "learning_rate": 9.868421052631579e-07,
+      "loss": 3.418,
+      "step": 130
+    },
+    {
+      "epoch": 0.03314777327935223,
+      "grad_norm": 4.25,
+      "learning_rate": 9.94433198380567e-07,
+      "loss": 3.4737,
+      "step": 131
+    },
+    {
+      "epoch": 0.03340080971659919,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.0020242914979758e-06,
+      "loss": 3.5622,
+      "step": 132
+    },
+    {
+      "epoch": 0.03365384615384615,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.0096153846153846e-06,
+      "loss": 3.5176,
+      "step": 133
+    },
+    {
+      "epoch": 0.03390688259109312,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.0172064777327935e-06,
+      "loss": 3.5446,
+      "step": 134
+    },
+    {
+      "epoch": 0.034159919028340084,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.0247975708502025e-06,
+      "loss": 3.5028,
+      "step": 135
+    },
+    {
+      "epoch": 0.03441295546558704,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.0323886639676114e-06,
+      "loss": 3.5673,
+      "step": 136
+    },
+    {
+      "epoch": 0.03466599190283401,
+      "grad_norm": 4.65625,
+      "learning_rate": 1.0399797570850202e-06,
+      "loss": 3.5705,
+      "step": 137
+    },
+    {
+      "epoch": 0.034919028340080975,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.0475708502024293e-06,
+      "loss": 3.3797,
+      "step": 138
+    },
+    {
+      "epoch": 0.035172064777327934,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.0551619433198381e-06,
+      "loss": 3.3866,
+      "step": 139
+    },
+    {
+      "epoch": 0.0354251012145749,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.062753036437247e-06,
+      "loss": 3.5144,
+      "step": 140
+    },
+    {
+      "epoch": 0.03567813765182186,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.0703441295546558e-06,
+      "loss": 3.5378,
+      "step": 141
+    },
+    {
+      "epoch": 0.035931174089068825,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.0779352226720647e-06,
+      "loss": 3.4701,
+      "step": 142
+    },
+    {
+      "epoch": 0.03618421052631579,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.0855263157894738e-06,
+      "loss": 3.3188,
+      "step": 143
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 4.3125,
+      "learning_rate": 1.0931174089068826e-06,
+      "loss": 3.5538,
+      "step": 144
+    },
+    {
+      "epoch": 0.036690283400809716,
+      "grad_norm": 3.75,
+      "learning_rate": 1.1007085020242915e-06,
+      "loss": 3.3608,
+      "step": 145
+    },
+    {
+      "epoch": 0.03694331983805668,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.1082995951417005e-06,
+      "loss": 3.5807,
+      "step": 146
+    },
+    {
+      "epoch": 0.03719635627530364,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.1158906882591092e-06,
+      "loss": 3.2498,
+      "step": 147
+    },
+    {
+      "epoch": 0.03744939271255061,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.1234817813765182e-06,
+      "loss": 3.5459,
+      "step": 148
+    },
+    {
+      "epoch": 0.03770242914979757,
+      "grad_norm": 4.25,
+      "learning_rate": 1.1310728744939273e-06,
+      "loss": 3.4524,
+      "step": 149
+    },
+    {
+      "epoch": 0.03795546558704453,
+      "grad_norm": 4.25,
+      "learning_rate": 1.138663967611336e-06,
+      "loss": 3.4575,
+      "step": 150
+    },
+    {
+      "epoch": 0.03795546558704453,
+      "eval_loss": 3.521191358566284,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.6529,
+      "eval_samples_per_second": 9.751,
+      "eval_steps_per_second": 9.751,
+      "step": 150
+    },
+    {
+      "epoch": 0.0382085020242915,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.146255060728745e-06,
+      "loss": 3.4378,
+      "step": 151
+    },
+    {
+      "epoch": 0.038461538461538464,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 3.4296,
+      "step": 152
+    },
+    {
+      "epoch": 0.03871457489878542,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.1614372469635627e-06,
+      "loss": 3.4832,
+      "step": 153
+    },
+    {
+      "epoch": 0.03896761133603239,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.1690283400809717e-06,
+      "loss": 3.4565,
+      "step": 154
+    },
+    {
+      "epoch": 0.039220647773279355,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.1766194331983806e-06,
+      "loss": 3.386,
+      "step": 155
+    },
+    {
+      "epoch": 0.039473684210526314,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.1842105263157894e-06,
+      "loss": 3.2917,
+      "step": 156
+    },
+    {
+      "epoch": 0.03972672064777328,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.1918016194331985e-06,
+      "loss": 3.4754,
+      "step": 157
+    },
+    {
+      "epoch": 0.039979757085020245,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.1993927125506073e-06,
+      "loss": 3.536,
+      "step": 158
+    },
+    {
+      "epoch": 0.040232793522267205,
+      "grad_norm": 4.0,
+      "learning_rate": 1.2069838056680162e-06,
+      "loss": 3.3942,
+      "step": 159
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 3.875,
+      "learning_rate": 1.214574898785425e-06,
+      "loss": 3.396,
+      "step": 160
+    },
+    {
+      "epoch": 0.040738866396761136,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.2221659919028341e-06,
+      "loss": 3.4066,
+      "step": 161
+    },
+    {
+      "epoch": 0.040991902834008095,
+      "grad_norm": 4.125,
+      "learning_rate": 1.229757085020243e-06,
+      "loss": 3.4321,
+      "step": 162
+    },
+    {
+      "epoch": 0.04124493927125506,
+      "grad_norm": 4.0,
+      "learning_rate": 1.2373481781376518e-06,
+      "loss": 3.4505,
+      "step": 163
+    },
+    {
+      "epoch": 0.04149797570850203,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.2449392712550609e-06,
+      "loss": 3.393,
+      "step": 164
+    },
+    {
+      "epoch": 0.041751012145748986,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.2525303643724697e-06,
+      "loss": 3.3221,
+      "step": 165
+    },
+    {
+      "epoch": 0.04200404858299595,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.2601214574898786e-06,
+      "loss": 3.3035,
+      "step": 166
+    },
+    {
+      "epoch": 0.04225708502024292,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.2677125506072876e-06,
+      "loss": 3.3424,
+      "step": 167
+    },
+    {
+      "epoch": 0.04251012145748988,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.2753036437246963e-06,
+      "loss": 3.3755,
+      "step": 168
+    },
+    {
+      "epoch": 0.04276315789473684,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.2828947368421053e-06,
+      "loss": 3.4524,
+      "step": 169
+    },
+    {
+      "epoch": 0.0430161943319838,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.2904858299595142e-06,
+      "loss": 3.5399,
+      "step": 170
+    },
+    {
+      "epoch": 0.04326923076923077,
+      "grad_norm": 4.125,
+      "learning_rate": 1.298076923076923e-06,
+      "loss": 3.551,
+      "step": 171
+    },
+    {
+      "epoch": 0.043522267206477734,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.305668016194332e-06,
+      "loss": 3.4368,
+      "step": 172
+    },
+    {
+      "epoch": 0.04377530364372469,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.3132591093117407e-06,
+      "loss": 3.264,
+      "step": 173
+    },
+    {
+      "epoch": 0.04402834008097166,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.3208502024291498e-06,
+      "loss": 3.4775,
+      "step": 174
+    },
+    {
+      "epoch": 0.044281376518218625,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.3284412955465589e-06,
+      "loss": 3.4838,
+      "step": 175
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.3360323886639675e-06,
+      "loss": 3.3844,
+      "step": 176
+    },
+    {
+      "epoch": 0.04478744939271255,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.3436234817813765e-06,
+      "loss": 3.4116,
+      "step": 177
+    },
+    {
+      "epoch": 0.045040485829959516,
+      "grad_norm": 3.75,
+      "learning_rate": 1.3512145748987856e-06,
+      "loss": 3.3178,
+      "step": 178
+    },
+    {
+      "epoch": 0.045293522267206475,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.3588056680161942e-06,
+      "loss": 3.4294,
+      "step": 179
+    },
+    {
+      "epoch": 0.04554655870445344,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.3663967611336033e-06,
+      "loss": 3.3183,
+      "step": 180
+    },
+    {
+      "epoch": 0.04554655870445344,
+      "eval_loss": 3.4689724445343018,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.2997,
+      "eval_samples_per_second": 9.763,
+      "eval_steps_per_second": 9.763,
+      "step": 180
+    },
+    {
+      "epoch": 0.04579959514170041,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.3739878542510122e-06,
+      "loss": 3.3741,
+      "step": 181
+    },
+    {
+      "epoch": 0.046052631578947366,
+      "grad_norm": 4.25,
+      "learning_rate": 1.381578947368421e-06,
+      "loss": 3.3538,
+      "step": 182
+    },
+    {
+      "epoch": 0.04630566801619433,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.38917004048583e-06,
+      "loss": 3.4457,
+      "step": 183
+    },
+    {
+      "epoch": 0.0465587044534413,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.396761133603239e-06,
+      "loss": 3.4034,
+      "step": 184
+    },
+    {
+      "epoch": 0.04681174089068826,
+      "grad_norm": 4.0,
+      "learning_rate": 1.4043522267206478e-06,
+      "loss": 3.4732,
+      "step": 185
+    },
+    {
+      "epoch": 0.04706477732793522,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.4119433198380566e-06,
+      "loss": 3.4386,
+      "step": 186
+    },
+    {
+      "epoch": 0.04731781376518219,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.4195344129554657e-06,
+      "loss": 3.4672,
+      "step": 187
+    },
+    {
+      "epoch": 0.04757085020242915,
+      "grad_norm": 3.625,
+      "learning_rate": 1.4271255060728745e-06,
+      "loss": 3.2735,
+      "step": 188
+    },
+    {
+      "epoch": 0.047823886639676114,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.4347165991902834e-06,
+      "loss": 3.3614,
+      "step": 189
+    },
+    {
+      "epoch": 0.04807692307692308,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.4423076923076924e-06,
+      "loss": 3.4163,
+      "step": 190
+    },
+    {
+      "epoch": 0.04832995951417004,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.4498987854251013e-06,
+      "loss": 3.4352,
+      "step": 191
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.4574898785425101e-06,
+      "loss": 3.394,
+      "step": 192
+    },
+    {
+      "epoch": 0.04883603238866397,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.4650809716599192e-06,
+      "loss": 3.4913,
+      "step": 193
+    },
+    {
+      "epoch": 0.04908906882591093,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.4726720647773278e-06,
+      "loss": 3.397,
+      "step": 194
+    },
+    {
+      "epoch": 0.049342105263157895,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.480263157894737e-06,
+      "loss": 3.4548,
+      "step": 195
+    },
+    {
+      "epoch": 0.04959514170040486,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.487854251012146e-06,
+      "loss": 3.196,
+      "step": 196
+    },
+    {
+      "epoch": 0.04984817813765182,
+      "grad_norm": 4.0,
+      "learning_rate": 1.4954453441295546e-06,
+      "loss": 3.387,
+      "step": 197
+    },
+    {
+      "epoch": 0.050101214574898786,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.5030364372469637e-06,
+      "loss": 3.4757,
+      "step": 198
+    },
+    {
+      "epoch": 0.05035425101214575,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.5106275303643725e-06,
+      "loss": 3.3087,
+      "step": 199
+    },
+    {
+      "epoch": 0.05060728744939271,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.5182186234817814e-06,
+      "loss": 3.4974,
+      "step": 200
+    },
+    {
+      "epoch": 0.05086032388663968,
+      "grad_norm": 4.0,
+      "learning_rate": 1.5258097165991904e-06,
+      "loss": 3.3488,
+      "step": 201
+    },
+    {
+      "epoch": 0.051113360323886636,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.533400809716599e-06,
+      "loss": 3.1754,
+      "step": 202
+    },
+    {
+      "epoch": 0.0513663967611336,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.5409919028340081e-06,
+      "loss": 3.1885,
+      "step": 203
+    },
+    {
+      "epoch": 0.05161943319838057,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.5485829959514172e-06,
+      "loss": 3.3669,
+      "step": 204
+    },
+    {
+      "epoch": 0.05187246963562753,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.5561740890688258e-06,
+      "loss": 3.3768,
+      "step": 205
+    },
+    {
+      "epoch": 0.05212550607287449,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.5637651821862349e-06,
+      "loss": 3.3708,
+      "step": 206
+    },
+    {
+      "epoch": 0.05237854251012146,
+      "grad_norm": 80.5,
+      "learning_rate": 1.5713562753036437e-06,
+      "loss": 3.3298,
+      "step": 207
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.5789473684210526e-06,
+      "loss": 3.3535,
+      "step": 208
+    },
+    {
+      "epoch": 0.052884615384615384,
+      "grad_norm": 4.0,
+      "learning_rate": 1.5865384615384616e-06,
+      "loss": 3.3483,
+      "step": 209
+    },
+    {
+      "epoch": 0.05313765182186235,
+      "grad_norm": 4.0,
+      "learning_rate": 1.5941295546558705e-06,
+      "loss": 3.4232,
+      "step": 210
+    },
+    {
+      "epoch": 0.05313765182186235,
+      "eval_loss": 3.428436040878296,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.3637,
+      "eval_samples_per_second": 9.761,
+      "eval_steps_per_second": 9.761,
+      "step": 210
+    },
+    {
+      "epoch": 0.05339068825910931,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.6017206477732793e-06,
+      "loss": 3.226,
+      "step": 211
+    },
+    {
+      "epoch": 0.053643724696356275,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.6093117408906882e-06,
+      "loss": 3.3584,
+      "step": 212
+    },
+    {
+      "epoch": 0.05389676113360324,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.6169028340080973e-06,
+      "loss": 3.3947,
+      "step": 213
+    },
+    {
+      "epoch": 0.0541497975708502,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.624493927125506e-06,
+      "loss": 3.3236,
+      "step": 214
+    },
+    {
+      "epoch": 0.054402834008097166,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.632085020242915e-06,
+      "loss": 3.3499,
+      "step": 215
+    },
+    {
+      "epoch": 0.05465587044534413,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.639676113360324e-06,
+      "loss": 3.2439,
+      "step": 216
+    },
+    {
+      "epoch": 0.05490890688259109,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.6472672064777329e-06,
+      "loss": 3.1813,
+      "step": 217
+    },
+    {
+      "epoch": 0.05516194331983806,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.6548582995951417e-06,
+      "loss": 3.0514,
+      "step": 218
+    },
+    {
+      "epoch": 0.05541497975708502,
+      "grad_norm": 4.28125,
+      "learning_rate": 1.6624493927125508e-06,
+      "loss": 3.4213,
+      "step": 219
+    },
+    {
+      "epoch": 0.05566801619433198,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.6700404858299594e-06,
+      "loss": 3.3607,
+      "step": 220
+    },
+    {
+      "epoch": 0.05592105263157895,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.6776315789473685e-06,
+      "loss": 3.3745,
+      "step": 221
+    },
+    {
+      "epoch": 0.056174089068825914,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.6852226720647775e-06,
+      "loss": 3.4646,
+      "step": 222
+    },
+    {
+      "epoch": 0.05642712550607287,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.6928137651821862e-06,
+      "loss": 3.4045,
+      "step": 223
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.7004048582995952e-06,
+      "loss": 3.3526,
+      "step": 224
+    },
+    {
+      "epoch": 0.056933198380566805,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.707995951417004e-06,
+      "loss": 3.3529,
+      "step": 225
+    },
+    {
+      "epoch": 0.057186234817813764,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.715587044534413e-06,
+      "loss": 3.3143,
+      "step": 226
+    },
+    {
+      "epoch": 0.05743927125506073,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.723178137651822e-06,
+      "loss": 3.383,
+      "step": 227
+    },
+    {
+      "epoch": 0.057692307692307696,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.7307692307692308e-06,
+      "loss": 3.3154,
+      "step": 228
+    },
+    {
+      "epoch": 0.057945344129554655,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.7383603238866397e-06,
+      "loss": 3.2319,
+      "step": 229
+    },
+    {
+      "epoch": 0.05819838056680162,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.7459514170040488e-06,
+      "loss": 3.4175,
+      "step": 230
+    },
+    {
+      "epoch": 0.058451417004048586,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.7535425101214576e-06,
+      "loss": 3.3109,
+      "step": 231
+    },
+    {
+      "epoch": 0.058704453441295545,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.7611336032388665e-06,
+      "loss": 3.3509,
+      "step": 232
+    },
+    {
+      "epoch": 0.05895748987854251,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.7687246963562753e-06,
+      "loss": 3.343,
+      "step": 233
+    },
+    {
+      "epoch": 0.05921052631578947,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.7763157894736842e-06,
+      "loss": 3.3576,
+      "step": 234
+    },
+    {
+      "epoch": 0.059463562753036436,
+      "grad_norm": 3.75,
+      "learning_rate": 1.7839068825910932e-06,
+      "loss": 3.1804,
+      "step": 235
+    },
+    {
+      "epoch": 0.0597165991902834,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.791497975708502e-06,
+      "loss": 3.3756,
+      "step": 236
+    },
+    {
+      "epoch": 0.05996963562753036,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.799089068825911e-06,
+      "loss": 3.2225,
+      "step": 237
+    },
+    {
+      "epoch": 0.06022267206477733,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.8066801619433198e-06,
+      "loss": 3.3377,
+      "step": 238
+    },
+    {
+      "epoch": 0.06047570850202429,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.8142712550607288e-06,
+      "loss": 3.3492,
+      "step": 239
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.8218623481781377e-06,
+      "loss": 3.3506,
+      "step": 240
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "eval_loss": 3.394364833831787,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 265.6171,
+      "eval_samples_per_second": 9.789,
+      "eval_steps_per_second": 9.789,
+      "step": 240
+    },
+    {
+      "epoch": 0.06098178137651822,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.8294534412955465e-06,
+      "loss": 3.1323,
+      "step": 241
+    },
+    {
+      "epoch": 0.061234817813765184,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.8370445344129556e-06,
+      "loss": 3.2968,
+      "step": 242
+    },
+    {
+      "epoch": 0.06148785425101214,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.8446356275303644e-06,
+      "loss": 3.3939,
+      "step": 243
+    },
+    {
+      "epoch": 0.06174089068825911,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.8522267206477733e-06,
+      "loss": 3.2642,
+      "step": 244
+    },
+    {
+      "epoch": 0.061993927125506075,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.8598178137651823e-06,
+      "loss": 3.3424,
+      "step": 245
+    },
+    {
+      "epoch": 0.062246963562753034,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.867408906882591e-06,
+      "loss": 3.2232,
+      "step": 246
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.875e-06,
+      "loss": 3.2869,
+      "step": 247
+    },
+    {
+      "epoch": 0.06275303643724696,
+      "grad_norm": 4.125,
+      "learning_rate": 1.882591093117409e-06,
+      "loss": 3.2594,
+      "step": 248
+    },
+    {
+      "epoch": 0.06300607287449393,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.890182186234818e-06,
+      "loss": 3.3061,
+      "step": 249
+    },
+    {
+      "epoch": 0.06325910931174089,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.8977732793522268e-06,
+      "loss": 3.3135,
+      "step": 250
+    },
+    {
+      "epoch": 0.06351214574898785,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.9053643724696354e-06,
+      "loss": 3.364,
+      "step": 251
+    },
+    {
+      "epoch": 0.06376518218623482,
+      "grad_norm": 35.75,
+      "learning_rate": 1.912955465587045e-06,
+      "loss": 3.2256,
+      "step": 252
+    },
+    {
+      "epoch": 0.06401821862348178,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.9205465587044534e-06,
+      "loss": 3.3095,
+      "step": 253
+    },
+    {
+      "epoch": 0.06427125506072874,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.928137651821862e-06,
+      "loss": 3.2367,
+      "step": 254
+    },
+    {
+      "epoch": 0.06452429149797571,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.9357287449392715e-06,
+      "loss": 3.1417,
+      "step": 255
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.9433198380566803e-06,
+      "loss": 3.3548,
+      "step": 256
+    },
+    {
+      "epoch": 0.06503036437246963,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.950910931174089e-06,
+      "loss": 3.445,
+      "step": 257
+    },
+    {
+      "epoch": 0.0652834008097166,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.958502024291498e-06,
+      "loss": 3.232,
+      "step": 258
+    },
+    {
+      "epoch": 0.06553643724696356,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.966093117408907e-06,
+      "loss": 3.4024,
+      "step": 259
+    },
+    {
+      "epoch": 0.06578947368421052,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.9736842105263157e-06,
+      "loss": 3.2631,
+      "step": 260
+    },
+    {
+      "epoch": 0.0660425101214575,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.981275303643725e-06,
+      "loss": 3.355,
+      "step": 261
+    },
+    {
+      "epoch": 0.06629554655870445,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.988866396761134e-06,
+      "loss": 3.3463,
+      "step": 262
+    },
+    {
+      "epoch": 0.06654858299595141,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.9964574898785423e-06,
+      "loss": 3.2892,
+      "step": 263
+    },
+    {
+      "epoch": 0.06680161943319839,
+      "grad_norm": 4.40625,
+      "learning_rate": 2.0040485829959516e-06,
+      "loss": 3.3025,
+      "step": 264
+    },
+    {
+      "epoch": 0.06705465587044535,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.0116396761133604e-06,
+      "loss": 3.3443,
+      "step": 265
+    },
+    {
+      "epoch": 0.0673076923076923,
+      "grad_norm": 4.0,
+      "learning_rate": 2.0192307692307692e-06,
+      "loss": 3.3377,
+      "step": 266
+    },
+    {
+      "epoch": 0.06756072874493928,
+      "grad_norm": 4.125,
+      "learning_rate": 2.0268218623481785e-06,
+      "loss": 3.3298,
+      "step": 267
+    },
+    {
+      "epoch": 0.06781376518218624,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.034412955465587e-06,
+      "loss": 3.373,
+      "step": 268
+    },
+    {
+      "epoch": 0.0680668016194332,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.042004048582996e-06,
+      "loss": 3.2904,
+      "step": 269
+    },
+    {
+      "epoch": 0.06831983805668017,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.049595141700405e-06,
+      "loss": 3.3605,
+      "step": 270
+    },
+    {
+      "epoch": 0.06831983805668017,
+      "eval_loss": 3.36352801322937,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 265.8172,
+      "eval_samples_per_second": 9.781,
+      "eval_steps_per_second": 9.781,
+      "step": 270
+    },
+    {
+      "epoch": 0.06857287449392713,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.057186234817814e-06,
+      "loss": 3.3621,
+      "step": 271
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.0647773279352228e-06,
+      "loss": 3.3485,
+      "step": 272
+    },
+    {
+      "epoch": 0.06907894736842106,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.0723684210526316e-06,
+      "loss": 3.3451,
+      "step": 273
+    },
+    {
+      "epoch": 0.06933198380566802,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.0799595141700405e-06,
+      "loss": 3.2057,
+      "step": 274
+    },
+    {
+      "epoch": 0.06958502024291498,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.0875506072874493e-06,
+      "loss": 3.4109,
+      "step": 275
+    },
+    {
+      "epoch": 0.06983805668016195,
+      "grad_norm": 3.625,
+      "learning_rate": 2.0951417004048586e-06,
+      "loss": 3.1479,
+      "step": 276
+    },
+    {
+      "epoch": 0.07009109311740891,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.1027327935222674e-06,
+      "loss": 3.2703,
+      "step": 277
+    },
+    {
+      "epoch": 0.07034412955465587,
+      "grad_norm": 4.46875,
+      "learning_rate": 2.1103238866396763e-06,
+      "loss": 3.3497,
+      "step": 278
+    },
+    {
+      "epoch": 0.07059716599190283,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.1179149797570847e-06,
+      "loss": 3.3317,
+      "step": 279
+    },
+    {
+      "epoch": 0.0708502024291498,
+      "grad_norm": 3.875,
+      "learning_rate": 2.125506072874494e-06,
+      "loss": 3.3002,
+      "step": 280
+    },
+    {
+      "epoch": 0.07110323886639676,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.133097165991903e-06,
+      "loss": 3.2763,
+      "step": 281
+    },
+    {
+      "epoch": 0.07135627530364372,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.1406882591093117e-06,
+      "loss": 3.2701,
+      "step": 282
+    },
+    {
+      "epoch": 0.07160931174089069,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.148279352226721e-06,
+      "loss": 3.2716,
+      "step": 283
+    },
+    {
+      "epoch": 0.07186234817813765,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.1558704453441294e-06,
+      "loss": 3.2635,
+      "step": 284
+    },
+    {
+      "epoch": 0.07211538461538461,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.1634615384615382e-06,
+      "loss": 3.2941,
+      "step": 285
+    },
+    {
+      "epoch": 0.07236842105263158,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.1710526315789475e-06,
+      "loss": 3.3379,
+      "step": 286
+    },
+    {
+      "epoch": 0.07262145748987854,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.1786437246963564e-06,
+      "loss": 3.3231,
+      "step": 287
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.1862348178137652e-06,
+      "loss": 3.1944,
+      "step": 288
+    },
+    {
+      "epoch": 0.07312753036437247,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.193825910931174e-06,
+      "loss": 3.3514,
+      "step": 289
+    },
+    {
+      "epoch": 0.07338056680161943,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.201417004048583e-06,
+      "loss": 2.9866,
+      "step": 290
+    },
+    {
+      "epoch": 0.07363360323886639,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.2090080971659918e-06,
+      "loss": 3.3327,
+      "step": 291
+    },
+    {
+      "epoch": 0.07388663967611336,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.216599190283401e-06,
+      "loss": 3.2911,
+      "step": 292
+    },
+    {
+      "epoch": 0.07413967611336032,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.22419028340081e-06,
+      "loss": 3.3534,
+      "step": 293
+    },
+    {
+      "epoch": 0.07439271255060728,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.2317813765182183e-06,
+      "loss": 3.2998,
+      "step": 294
+    },
+    {
+      "epoch": 0.07464574898785425,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.2393724696356276e-06,
+      "loss": 3.3745,
+      "step": 295
+    },
+    {
+      "epoch": 0.07489878542510121,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.2469635627530364e-06,
+      "loss": 3.317,
+      "step": 296
+    },
+    {
+      "epoch": 0.07515182186234817,
+      "grad_norm": 4.34375,
+      "learning_rate": 2.2545546558704453e-06,
+      "loss": 3.3992,
+      "step": 297
+    },
+    {
+      "epoch": 0.07540485829959515,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.2621457489878546e-06,
+      "loss": 3.3001,
+      "step": 298
+    },
+    {
+      "epoch": 0.0756578947368421,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.2697368421052634e-06,
+      "loss": 3.2192,
+      "step": 299
+    },
+    {
+      "epoch": 0.07591093117408906,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.277327935222672e-06,
+      "loss": 3.3411,
+      "step": 300
+    },
+    {
+      "epoch": 0.07591093117408906,
+      "eval_loss": 3.335643768310547,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 265.6147,
+      "eval_samples_per_second": 9.789,
+      "eval_steps_per_second": 9.789,
+      "step": 300
+    },
+    {
+      "epoch": 0.07616396761133604,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.284919028340081e-06,
+      "loss": 3.248,
+      "step": 301
+    },
+    {
+      "epoch": 0.076417004048583,
+      "grad_norm": 3.5,
+      "learning_rate": 2.29251012145749e-06,
+      "loss": 3.1704,
+      "step": 302
+    },
+    {
+      "epoch": 0.07667004048582995,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.300101214574899e-06,
+      "loss": 3.2629,
+      "step": 303
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 3.3536,
+      "step": 304
+    },
+    {
+      "epoch": 0.07717611336032389,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.3152834008097165e-06,
+      "loss": 3.3183,
+      "step": 305
+    },
+    {
+      "epoch": 0.07742914979757085,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.3228744939271254e-06,
+      "loss": 3.1306,
+      "step": 306
+    },
+    {
+      "epoch": 0.07768218623481782,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.3304655870445346e-06,
+      "loss": 3.0119,
+      "step": 307
+    },
+    {
+      "epoch": 0.07793522267206478,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.3380566801619435e-06,
+      "loss": 3.1086,
+      "step": 308
+    },
+    {
+      "epoch": 0.07818825910931174,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.3456477732793523e-06,
+      "loss": 3.2373,
+      "step": 309
+    },
+    {
+      "epoch": 0.07844129554655871,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.353238866396761e-06,
+      "loss": 3.2396,
+      "step": 310
+    },
+    {
+      "epoch": 0.07869433198380567,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.36082995951417e-06,
+      "loss": 3.2353,
+      "step": 311
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.368421052631579e-06,
+      "loss": 3.2537,
+      "step": 312
+    },
+    {
+      "epoch": 0.0792004048582996,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.376012145748988e-06,
+      "loss": 3.2361,
+      "step": 313
+    },
+    {
+      "epoch": 0.07945344129554656,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.383603238866397e-06,
+      "loss": 3.3014,
+      "step": 314
+    },
+    {
+      "epoch": 0.07970647773279352,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.3911943319838054e-06,
+      "loss": 3.2054,
+      "step": 315
+    },
+    {
+      "epoch": 0.07995951417004049,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.3987854251012147e-06,
+      "loss": 3.3034,
+      "step": 316
+    },
+    {
+      "epoch": 0.08021255060728745,
+      "grad_norm": 3.875,
+      "learning_rate": 2.4063765182186235e-06,
+      "loss": 3.3018,
+      "step": 317
+    },
+    {
+      "epoch": 0.08046558704453441,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.4139676113360324e-06,
+      "loss": 2.9881,
+      "step": 318
+    },
+    {
+      "epoch": 0.08071862348178138,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.4215587044534417e-06,
+      "loss": 3.2021,
+      "step": 319
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.42914979757085e-06,
+      "loss": 3.2639,
+      "step": 320
+    },
+    {
+      "epoch": 0.0812246963562753,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.436740890688259e-06,
+      "loss": 3.2662,
+      "step": 321
+    },
+    {
+      "epoch": 0.08147773279352227,
+      "grad_norm": 3.625,
+      "learning_rate": 2.4443319838056682e-06,
+      "loss": 3.0858,
+      "step": 322
+    },
+    {
+      "epoch": 0.08173076923076923,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.451923076923077e-06,
+      "loss": 3.2958,
+      "step": 323
+    },
+    {
+      "epoch": 0.08198380566801619,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.459514170040486e-06,
+      "loss": 3.2128,
+      "step": 324
+    },
+    {
+      "epoch": 0.08223684210526316,
+      "grad_norm": 3.875,
+      "learning_rate": 2.4671052631578948e-06,
+      "loss": 3.2558,
+      "step": 325
+    },
+    {
+      "epoch": 0.08248987854251012,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.4746963562753036e-06,
+      "loss": 3.2995,
+      "step": 326
+    },
+    {
+      "epoch": 0.08274291497975708,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.4822874493927125e-06,
+      "loss": 3.2499,
+      "step": 327
+    },
+    {
+      "epoch": 0.08299595141700405,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.4898785425101217e-06,
+      "loss": 3.3091,
+      "step": 328
+    },
+    {
+      "epoch": 0.08324898785425101,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.4974696356275306e-06,
+      "loss": 3.3163,
+      "step": 329
+    },
+    {
+      "epoch": 0.08350202429149797,
+      "grad_norm": 4.125,
+      "learning_rate": 2.5050607287449394e-06,
+      "loss": 3.2892,
+      "step": 330
+    },
+    {
+      "epoch": 0.08350202429149797,
+      "eval_loss": 3.3100743293762207,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.0577,
+      "eval_samples_per_second": 9.772,
+      "eval_steps_per_second": 9.772,
+      "step": 330
+    },
+    {
+      "epoch": 0.08375506072874495,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.5126518218623483e-06,
+      "loss": 3.2173,
+      "step": 331
+    },
+    {
+      "epoch": 0.0840080971659919,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.520242914979757e-06,
+      "loss": 3.1206,
+      "step": 332
+    },
+    {
+      "epoch": 0.08426113360323886,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.527834008097166e-06,
+      "loss": 3.2805,
+      "step": 333
+    },
+    {
+      "epoch": 0.08451417004048584,
+      "grad_norm": 4.0,
+      "learning_rate": 2.5354251012145753e-06,
+      "loss": 3.2389,
+      "step": 334
+    },
+    {
+      "epoch": 0.0847672064777328,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.543016194331984e-06,
+      "loss": 3.251,
+      "step": 335
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.5506072874493925e-06,
+      "loss": 3.0921,
+      "step": 336
+    },
+    {
+      "epoch": 0.08527327935222673,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.558198380566802e-06,
+      "loss": 3.3231,
+      "step": 337
+    },
+    {
+      "epoch": 0.08552631578947369,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.5657894736842107e-06,
+      "loss": 3.1829,
+      "step": 338
+    },
+    {
+      "epoch": 0.08577935222672065,
+      "grad_norm": 4.375,
+      "learning_rate": 2.5733805668016195e-06,
+      "loss": 3.2374,
+      "step": 339
+    },
+    {
+      "epoch": 0.0860323886639676,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.5809716599190284e-06,
+      "loss": 3.1373,
+      "step": 340
+    },
+    {
+      "epoch": 0.08628542510121458,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.588562753036437e-06,
+      "loss": 3.2256,
+      "step": 341
+    },
+    {
+      "epoch": 0.08653846153846154,
+      "grad_norm": 4.0,
+      "learning_rate": 2.596153846153846e-06,
+      "loss": 3.2534,
+      "step": 342
+    },
+    {
+      "epoch": 0.0867914979757085,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.603744939271255e-06,
+      "loss": 3.2242,
+      "step": 343
+    },
+    {
+      "epoch": 0.08704453441295547,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.611336032388664e-06,
+      "loss": 3.1236,
+      "step": 344
+    },
+    {
+      "epoch": 0.08729757085020243,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.618927125506073e-06,
+      "loss": 3.2309,
+      "step": 345
+    },
+    {
+      "epoch": 0.08755060728744939,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.6265182186234815e-06,
+      "loss": 3.1701,
+      "step": 346
+    },
+    {
+      "epoch": 0.08780364372469636,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.6341093117408907e-06,
+      "loss": 3.195,
+      "step": 347
+    },
+    {
+      "epoch": 0.08805668016194332,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.6417004048582996e-06,
+      "loss": 3.2068,
+      "step": 348
+    },
+    {
+      "epoch": 0.08830971659919028,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.6492914979757084e-06,
+      "loss": 3.1687,
+      "step": 349
+    },
+    {
+      "epoch": 0.08856275303643725,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.6568825910931177e-06,
+      "loss": 3.2299,
+      "step": 350
+    },
+    {
+      "epoch": 0.08881578947368421,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.6644736842105266e-06,
+      "loss": 3.2512,
+      "step": 351
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.672064777327935e-06,
+      "loss": 3.1597,
+      "step": 352
+    },
+    {
+      "epoch": 0.08932186234817814,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.6796558704453442e-06,
+      "loss": 3.2883,
+      "step": 353
+    },
+    {
+      "epoch": 0.0895748987854251,
+      "grad_norm": 3.875,
+      "learning_rate": 2.687246963562753e-06,
+      "loss": 3.2693,
+      "step": 354
+    },
+    {
+      "epoch": 0.08982793522267206,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.694838056680162e-06,
+      "loss": 3.2533,
+      "step": 355
+    },
+    {
+      "epoch": 0.09008097165991903,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.7024291497975712e-06,
+      "loss": 3.1419,
+      "step": 356
+    },
+    {
+      "epoch": 0.09033400809716599,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.7100202429149796e-06,
+      "loss": 3.2672,
+      "step": 357
+    },
+    {
+      "epoch": 0.09058704453441295,
+      "grad_norm": 3.625,
+      "learning_rate": 2.7176113360323885e-06,
+      "loss": 3.2232,
+      "step": 358
+    },
+    {
+      "epoch": 0.09084008097165992,
+      "grad_norm": 4.125,
+      "learning_rate": 2.7252024291497978e-06,
+      "loss": 3.268,
+      "step": 359
+    },
+    {
+      "epoch": 0.09109311740890688,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.7327935222672066e-06,
+      "loss": 3.3207,
+      "step": 360
+    },
+    {
+      "epoch": 0.09109311740890688,
+      "eval_loss": 3.287496328353882,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.7793,
+      "eval_samples_per_second": 9.746,
+      "eval_steps_per_second": 9.746,
+      "step": 360
+    },
+    {
+      "epoch": 0.09134615384615384,
+      "grad_norm": 3.5,
+      "learning_rate": 2.7403846153846155e-06,
+      "loss": 3.2149,
+      "step": 361
+    },
+    {
+      "epoch": 0.09159919028340081,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.7479757085020243e-06,
+      "loss": 3.2364,
+      "step": 362
+    },
+    {
+      "epoch": 0.09185222672064777,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.755566801619433e-06,
+      "loss": 3.1369,
+      "step": 363
+    },
+    {
+      "epoch": 0.09210526315789473,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.763157894736842e-06,
+      "loss": 3.1973,
+      "step": 364
+    },
+    {
+      "epoch": 0.0923582995951417,
+      "grad_norm": 3.875,
+      "learning_rate": 2.7707489878542513e-06,
+      "loss": 3.3327,
+      "step": 365
+    },
+    {
+      "epoch": 0.09261133603238866,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.77834008097166e-06,
+      "loss": 3.2696,
+      "step": 366
+    },
+    {
+      "epoch": 0.09286437246963562,
+      "grad_norm": 3.96875,
+      "learning_rate": 2.7859311740890686e-06,
+      "loss": 3.2108,
+      "step": 367
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.793522267206478e-06,
+      "loss": 3.1963,
+      "step": 368
+    },
+    {
+      "epoch": 0.09337044534412955,
+      "grad_norm": 4.21875,
+      "learning_rate": 2.8011133603238867e-06,
+      "loss": 3.2062,
+      "step": 369
+    },
+    {
+      "epoch": 0.09362348178137651,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.8087044534412955e-06,
+      "loss": 3.1829,
+      "step": 370
+    },
+    {
+      "epoch": 0.09387651821862349,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.816295546558705e-06,
+      "loss": 3.1628,
+      "step": 371
+    },
+    {
+      "epoch": 0.09412955465587045,
+      "grad_norm": 3.75,
+      "learning_rate": 2.8238866396761132e-06,
+      "loss": 3.2366,
+      "step": 372
+    },
+    {
+      "epoch": 0.0943825910931174,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.831477732793522e-06,
+      "loss": 2.9478,
+      "step": 373
+    },
+    {
+      "epoch": 0.09463562753036438,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.8390688259109314e-06,
+      "loss": 3.2031,
+      "step": 374
+    },
+    {
+      "epoch": 0.09488866396761134,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.8466599190283402e-06,
+      "loss": 3.0563,
+      "step": 375
+    },
+    {
+      "epoch": 0.0951417004048583,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.854251012145749e-06,
+      "loss": 3.0132,
+      "step": 376
+    },
+    {
+      "epoch": 0.09539473684210527,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.861842105263158e-06,
+      "loss": 3.2841,
+      "step": 377
+    },
+    {
+      "epoch": 0.09564777327935223,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.8694331983805668e-06,
+      "loss": 3.0711,
+      "step": 378
+    },
+    {
+      "epoch": 0.09590080971659919,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.8770242914979756e-06,
+      "loss": 3.0532,
+      "step": 379
+    },
+    {
+      "epoch": 0.09615384615384616,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.884615384615385e-06,
+      "loss": 3.1008,
+      "step": 380
+    },
+    {
+      "epoch": 0.09640688259109312,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.8922064777327937e-06,
+      "loss": 2.8251,
+      "step": 381
+    },
+    {
+      "epoch": 0.09665991902834008,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.8997975708502026e-06,
+      "loss": 3.2038,
+      "step": 382
+    },
+    {
+      "epoch": 0.09691295546558705,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.9073886639676114e-06,
+      "loss": 3.0765,
+      "step": 383
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.9149797570850203e-06,
+      "loss": 3.2462,
+      "step": 384
+    },
+    {
+      "epoch": 0.09741902834008097,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.922570850202429e-06,
+      "loss": 3.0698,
+      "step": 385
+    },
+    {
+      "epoch": 0.09767206477732794,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.9301619433198384e-06,
+      "loss": 3.1953,
+      "step": 386
+    },
+    {
+      "epoch": 0.0979251012145749,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.9377530364372473e-06,
+      "loss": 3.0546,
+      "step": 387
+    },
+    {
+      "epoch": 0.09817813765182186,
+      "grad_norm": 3.875,
+      "learning_rate": 2.9453441295546557e-06,
+      "loss": 3.1914,
+      "step": 388
+    },
+    {
+      "epoch": 0.09843117408906883,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.952935222672065e-06,
+      "loss": 3.257,
+      "step": 389
+    },
+    {
+      "epoch": 0.09868421052631579,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.960526315789474e-06,
+      "loss": 3.1663,
+      "step": 390
+    },
+    {
+      "epoch": 0.09868421052631579,
+      "eval_loss": 3.2687041759490967,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.9717,
+      "eval_samples_per_second": 9.739,
+      "eval_steps_per_second": 9.739,
+      "step": 390
+    },
+    {
+      "epoch": 0.09893724696356275,
+      "grad_norm": 4.0,
+      "learning_rate": 2.9681174089068827e-06,
+      "loss": 3.2455,
+      "step": 391
+    },
+    {
+      "epoch": 0.09919028340080972,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.975708502024292e-06,
+      "loss": 3.2181,
+      "step": 392
+    },
+    {
+      "epoch": 0.09944331983805668,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.9832995951417004e-06,
+      "loss": 3.1625,
+      "step": 393
+    },
+    {
+      "epoch": 0.09969635627530364,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.990890688259109e-06,
+      "loss": 3.3208,
+      "step": 394
+    },
+    {
+      "epoch": 0.09994939271255061,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.9984817813765185e-06,
+      "loss": 3.1644,
+      "step": 395
+    },
+    {
+      "epoch": 0.10020242914979757,
+      "grad_norm": 3.984375,
+      "learning_rate": 3.0060728744939273e-06,
+      "loss": 3.1736,
+      "step": 396
+    },
+    {
+      "epoch": 0.10045546558704453,
+      "grad_norm": 3.953125,
+      "learning_rate": 3.013663967611336e-06,
+      "loss": 3.1653,
+      "step": 397
+    },
+    {
+      "epoch": 0.1007085020242915,
+      "grad_norm": 3.984375,
+      "learning_rate": 3.021255060728745e-06,
+      "loss": 3.2299,
+      "step": 398
+    },
+    {
+      "epoch": 0.10096153846153846,
+      "grad_norm": 4.125,
+      "learning_rate": 3.028846153846154e-06,
+      "loss": 3.2327,
+      "step": 399
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 4.125,
+      "learning_rate": 3.0364372469635627e-06,
+      "loss": 3.2087,
+      "step": 400
+    },
+    {
+      "epoch": 0.1014676113360324,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.044028340080972e-06,
+      "loss": 3.1914,
+      "step": 401
+    },
+    {
+      "epoch": 0.10172064777327935,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.051619433198381e-06,
+      "loss": 3.201,
+      "step": 402
+    },
+    {
+      "epoch": 0.10197368421052631,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.0592105263157897e-06,
+      "loss": 3.1036,
+      "step": 403
+    },
+    {
+      "epoch": 0.10222672064777327,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.066801619433198e-06,
+      "loss": 3.2281,
+      "step": 404
+    },
+    {
+      "epoch": 0.10247975708502025,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.0743927125506074e-06,
+      "loss": 3.1966,
+      "step": 405
+    },
+    {
+      "epoch": 0.1027327935222672,
+      "grad_norm": 3.890625,
+      "learning_rate": 3.0819838056680162e-06,
+      "loss": 3.2849,
+      "step": 406
+    },
+    {
+      "epoch": 0.10298582995951416,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.089574898785425e-06,
+      "loss": 3.2518,
+      "step": 407
+    },
+    {
+      "epoch": 0.10323886639676114,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.0971659919028344e-06,
+      "loss": 3.0795,
+      "step": 408
+    },
+    {
+      "epoch": 0.1034919028340081,
+      "grad_norm": 3.625,
+      "learning_rate": 3.104757085020243e-06,
+      "loss": 3.3405,
+      "step": 409
+    },
+    {
+      "epoch": 0.10374493927125505,
+      "grad_norm": 3.765625,
+      "learning_rate": 3.1123481781376516e-06,
+      "loss": 3.2025,
+      "step": 410
+    },
+    {
+      "epoch": 0.10399797570850203,
+      "grad_norm": 4.28125,
+      "learning_rate": 3.119939271255061e-06,
+      "loss": 3.3012,
+      "step": 411
+    },
+    {
+      "epoch": 0.10425101214574899,
+      "grad_norm": 4.25,
+      "learning_rate": 3.1275303643724698e-06,
+      "loss": 3.2545,
+      "step": 412
+    },
+    {
+      "epoch": 0.10450404858299595,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.1351214574898786e-06,
+      "loss": 3.1081,
+      "step": 413
+    },
+    {
+      "epoch": 0.10475708502024292,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.1427125506072875e-06,
+      "loss": 3.1993,
+      "step": 414
+    },
+    {
+      "epoch": 0.10501012145748988,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.1503036437246963e-06,
+      "loss": 3.1923,
+      "step": 415
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.157894736842105e-06,
+      "loss": 3.2214,
+      "step": 416
+    },
+    {
+      "epoch": 0.10551619433198381,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.1654858299595144e-06,
+      "loss": 3.2199,
+      "step": 417
+    },
+    {
+      "epoch": 0.10576923076923077,
+      "grad_norm": 3.75,
+      "learning_rate": 3.1730769230769233e-06,
+      "loss": 3.1278,
+      "step": 418
+    },
+    {
+      "epoch": 0.10602226720647773,
+      "grad_norm": 3.75,
+      "learning_rate": 3.1806680161943317e-06,
+      "loss": 3.1004,
+      "step": 419
+    },
+    {
+      "epoch": 0.1062753036437247,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.188259109311741e-06,
+      "loss": 3.1078,
+      "step": 420
+    },
+    {
+      "epoch": 0.1062753036437247,
+      "eval_loss": 3.249948263168335,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 265.8859,
+      "eval_samples_per_second": 9.779,
+      "eval_steps_per_second": 9.779,
+      "step": 420
+    },
+    {
+      "epoch": 0.10652834008097166,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.19585020242915e-06,
+      "loss": 3.1628,
+      "step": 421
+    },
+    {
+      "epoch": 0.10678137651821862,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.2034412955465587e-06,
+      "loss": 3.2632,
+      "step": 422
+    },
+    {
+      "epoch": 0.10703441295546559,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.211032388663968e-06,
+      "loss": 3.0979,
+      "step": 423
+    },
+    {
+      "epoch": 0.10728744939271255,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.2186234817813764e-06,
+      "loss": 2.99,
+      "step": 424
+    },
+    {
+      "epoch": 0.10754048582995951,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.2262145748987852e-06,
+      "loss": 3.2533,
+      "step": 425
+    },
+    {
+      "epoch": 0.10779352226720648,
+      "grad_norm": 3.953125,
+      "learning_rate": 3.2338056680161945e-06,
+      "loss": 3.184,
+      "step": 426
+    },
+    {
+      "epoch": 0.10804655870445344,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.2413967611336034e-06,
+      "loss": 3.1788,
+      "step": 427
+    },
+    {
+      "epoch": 0.1082995951417004,
+      "grad_norm": 3.75,
+      "learning_rate": 3.248987854251012e-06,
+      "loss": 3.1937,
+      "step": 428
+    },
+    {
+      "epoch": 0.10855263157894737,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.256578947368421e-06,
+      "loss": 3.2438,
+      "step": 429
+    },
+    {
+      "epoch": 0.10880566801619433,
+      "grad_norm": 4.0,
+      "learning_rate": 3.26417004048583e-06,
+      "loss": 3.2015,
+      "step": 430
+    },
+    {
+      "epoch": 0.10905870445344129,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.2717611336032388e-06,
+      "loss": 3.2239,
+      "step": 431
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 4.53125,
+      "learning_rate": 3.279352226720648e-06,
+      "loss": 3.1564,
+      "step": 432
+    },
+    {
+      "epoch": 0.10956477732793522,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.286943319838057e-06,
+      "loss": 3.2057,
+      "step": 433
+    },
+    {
+      "epoch": 0.10981781376518218,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.2945344129554657e-06,
+      "loss": 3.0739,
+      "step": 434
+    },
+    {
+      "epoch": 0.11007085020242915,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.3021255060728746e-06,
+      "loss": 3.2605,
+      "step": 435
+    },
+    {
+      "epoch": 0.11032388663967611,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.3097165991902834e-06,
+      "loss": 3.1526,
+      "step": 436
+    },
+    {
+      "epoch": 0.11057692307692307,
+      "grad_norm": 3.984375,
+      "learning_rate": 3.3173076923076923e-06,
+      "loss": 3.1645,
+      "step": 437
+    },
+    {
+      "epoch": 0.11082995951417005,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.3248987854251016e-06,
+      "loss": 3.0687,
+      "step": 438
+    },
+    {
+      "epoch": 0.111082995951417,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.3324898785425104e-06,
+      "loss": 3.144,
+      "step": 439
+    },
+    {
+      "epoch": 0.11133603238866396,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.340080971659919e-06,
+      "loss": 3.2517,
+      "step": 440
+    },
+    {
+      "epoch": 0.11158906882591094,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.347672064777328e-06,
+      "loss": 3.226,
+      "step": 441
+    },
+    {
+      "epoch": 0.1118421052631579,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.355263157894737e-06,
+      "loss": 3.1567,
+      "step": 442
+    },
+    {
+      "epoch": 0.11209514170040485,
+      "grad_norm": 3.875,
+      "learning_rate": 3.362854251012146e-06,
+      "loss": 3.1997,
+      "step": 443
+    },
+    {
+      "epoch": 0.11234817813765183,
+      "grad_norm": 3.625,
+      "learning_rate": 3.370445344129555e-06,
+      "loss": 3.0442,
+      "step": 444
+    },
+    {
+      "epoch": 0.11260121457489879,
+      "grad_norm": 3.75,
+      "learning_rate": 3.3780364372469635e-06,
+      "loss": 3.1687,
+      "step": 445
+    },
+    {
+      "epoch": 0.11285425101214575,
+      "grad_norm": 3.953125,
+      "learning_rate": 3.3856275303643723e-06,
+      "loss": 3.2081,
+      "step": 446
+    },
+    {
+      "epoch": 0.11310728744939272,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.3932186234817816e-06,
+      "loss": 3.176,
+      "step": 447
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 4.125,
+      "learning_rate": 3.4008097165991905e-06,
+      "loss": 3.2223,
+      "step": 448
+    },
+    {
+      "epoch": 0.11361336032388664,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.4084008097165993e-06,
+      "loss": 3.1781,
+      "step": 449
+    },
+    {
+      "epoch": 0.11386639676113361,
+      "grad_norm": 6.0,
+      "learning_rate": 3.415991902834008e-06,
+      "loss": 3.0695,
+      "step": 450
+    },
+    {
+      "epoch": 0.11386639676113361,
+      "eval_loss": 3.2298877239227295,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.0294,
+      "eval_samples_per_second": 9.737,
+      "eval_steps_per_second": 9.737,
+      "step": 450
+    },
+    {
+      "epoch": 0.11411943319838057,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.423582995951417e-06,
+      "loss": 3.2869,
+      "step": 451
+    },
+    {
+      "epoch": 0.11437246963562753,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.431174089068826e-06,
+      "loss": 3.1212,
+      "step": 452
+    },
+    {
+      "epoch": 0.1146255060728745,
+      "grad_norm": 3.625,
+      "learning_rate": 3.438765182186235e-06,
+      "loss": 3.1894,
+      "step": 453
+    },
+    {
+      "epoch": 0.11487854251012146,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.446356275303644e-06,
+      "loss": 3.1481,
+      "step": 454
+    },
+    {
+      "epoch": 0.11513157894736842,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.453947368421053e-06,
+      "loss": 3.183,
+      "step": 455
+    },
+    {
+      "epoch": 0.11538461538461539,
+      "grad_norm": 4.3125,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 3.2638,
+      "step": 456
+    },
+    {
+      "epoch": 0.11563765182186235,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.4691295546558705e-06,
+      "loss": 3.0402,
+      "step": 457
+    },
+    {
+      "epoch": 0.11589068825910931,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.4767206477732794e-06,
+      "loss": 3.2754,
+      "step": 458
+    },
+    {
+      "epoch": 0.11614372469635628,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.4843117408906887e-06,
+      "loss": 3.1621,
+      "step": 459
+    },
+    {
+      "epoch": 0.11639676113360324,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.4919028340080975e-06,
+      "loss": 3.1904,
+      "step": 460
+    },
+    {
+      "epoch": 0.1166497975708502,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.499493927125506e-06,
+      "loss": 3.1502,
+      "step": 461
+    },
+    {
+      "epoch": 0.11690283400809717,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.5070850202429152e-06,
+      "loss": 3.1225,
+      "step": 462
+    },
+    {
+      "epoch": 0.11715587044534413,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.514676113360324e-06,
+      "loss": 3.0442,
+      "step": 463
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 4.28125,
+      "learning_rate": 3.522267206477733e-06,
+      "loss": 3.2107,
+      "step": 464
+    },
+    {
+      "epoch": 0.11766194331983805,
+      "grad_norm": 4.125,
+      "learning_rate": 3.5298582995951418e-06,
+      "loss": 3.2197,
+      "step": 465
+    },
+    {
+      "epoch": 0.11791497975708502,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.5374493927125506e-06,
+      "loss": 3.0518,
+      "step": 466
+    },
+    {
+      "epoch": 0.11816801619433198,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.5450404858299595e-06,
+      "loss": 3.1683,
+      "step": 467
+    },
+    {
+      "epoch": 0.11842105263157894,
+      "grad_norm": 4.125,
+      "learning_rate": 3.5526315789473683e-06,
+      "loss": 3.2341,
+      "step": 468
+    },
+    {
+      "epoch": 0.11867408906882591,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.5602226720647776e-06,
+      "loss": 3.1617,
+      "step": 469
+    },
+    {
+      "epoch": 0.11892712550607287,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.5678137651821864e-06,
+      "loss": 3.0863,
+      "step": 470
+    },
+    {
+      "epoch": 0.11918016194331983,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.575404858299595e-06,
+      "loss": 3.1213,
+      "step": 471
+    },
+    {
+      "epoch": 0.1194331983805668,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.582995951417004e-06,
+      "loss": 3.1579,
+      "step": 472
+    },
+    {
+      "epoch": 0.11968623481781376,
+      "grad_norm": 3.59375,
+      "learning_rate": 3.590587044534413e-06,
+      "loss": 3.0113,
+      "step": 473
+    },
+    {
+      "epoch": 0.11993927125506072,
+      "grad_norm": 4.34375,
+      "learning_rate": 3.598178137651822e-06,
+      "loss": 3.1977,
+      "step": 474
+    },
+    {
+      "epoch": 0.1201923076923077,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.605769230769231e-06,
+      "loss": 3.1362,
+      "step": 475
+    },
+    {
+      "epoch": 0.12044534412955465,
+      "grad_norm": 4.125,
+      "learning_rate": 3.6133603238866395e-06,
+      "loss": 3.1401,
+      "step": 476
+    },
+    {
+      "epoch": 0.12069838056680161,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.6209514170040484e-06,
+      "loss": 3.1889,
+      "step": 477
+    },
+    {
+      "epoch": 0.12095141700404859,
+      "grad_norm": 3.796875,
+      "learning_rate": 3.6285425101214577e-06,
+      "loss": 3.1723,
+      "step": 478
+    },
+    {
+      "epoch": 0.12120445344129555,
+      "grad_norm": 4.28125,
+      "learning_rate": 3.6361336032388665e-06,
+      "loss": 3.143,
+      "step": 479
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.6437246963562754e-06,
+      "loss": 3.155,
+      "step": 480
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "eval_loss": 3.212629556655884,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.4211,
+      "eval_samples_per_second": 9.722,
+      "eval_steps_per_second": 9.722,
+      "step": 480
+    },
+    {
+      "epoch": 0.12171052631578948,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.651315789473684e-06,
+      "loss": 3.121,
+      "step": 481
+    },
+    {
+      "epoch": 0.12196356275303644,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.658906882591093e-06,
+      "loss": 3.1192,
+      "step": 482
+    },
+    {
+      "epoch": 0.1222165991902834,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.666497975708502e-06,
+      "loss": 3.1597,
+      "step": 483
+    },
+    {
+      "epoch": 0.12246963562753037,
+      "grad_norm": 3.984375,
+      "learning_rate": 3.674089068825911e-06,
+      "loss": 3.1496,
+      "step": 484
+    },
+    {
+      "epoch": 0.12272267206477733,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.68168016194332e-06,
+      "loss": 3.1908,
+      "step": 485
+    },
+    {
+      "epoch": 0.12297570850202429,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.689271255060729e-06,
+      "loss": 3.2245,
+      "step": 486
+    },
+    {
+      "epoch": 0.12322874493927126,
+      "grad_norm": 3.71875,
+      "learning_rate": 3.6968623481781377e-06,
+      "loss": 3.1459,
+      "step": 487
+    },
+    {
+      "epoch": 0.12348178137651822,
+      "grad_norm": 4.0,
+      "learning_rate": 3.7044534412955466e-06,
+      "loss": 3.0791,
+      "step": 488
+    },
+    {
+      "epoch": 0.12373481781376518,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.7120445344129554e-06,
+      "loss": 3.1752,
+      "step": 489
+    },
+    {
+      "epoch": 0.12398785425101215,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.7196356275303647e-06,
+      "loss": 3.1211,
+      "step": 490
+    },
+    {
+      "epoch": 0.12424089068825911,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.7272267206477735e-06,
+      "loss": 3.1053,
+      "step": 491
+    },
+    {
+      "epoch": 0.12449392712550607,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.734817813765182e-06,
+      "loss": 3.1269,
+      "step": 492
+    },
+    {
+      "epoch": 0.12474696356275304,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.7424089068825912e-06,
+      "loss": 3.1721,
+      "step": 493
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 4.0,
+      "learning_rate": 3.75e-06,
+      "loss": 3.1291,
+      "step": 494
+    },
+    {
+      "epoch": 0.12525303643724697,
+      "grad_norm": 3.953125,
+      "learning_rate": 3.7575910931174094e-06,
+      "loss": 3.1698,
+      "step": 495
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 4.15625,
+      "learning_rate": 3.765182186234818e-06,
+      "loss": 3.2239,
+      "step": 496
+    },
+    {
+      "epoch": 0.1257591093117409,
+      "grad_norm": 4.0,
+      "learning_rate": 3.7727732793522266e-06,
+      "loss": 3.0746,
+      "step": 497
+    },
+    {
+      "epoch": 0.12601214574898786,
+      "grad_norm": 3.96875,
+      "learning_rate": 3.780364372469636e-06,
+      "loss": 3.0945,
+      "step": 498
+    },
+    {
+      "epoch": 0.1262651821862348,
+      "grad_norm": 3.875,
+      "learning_rate": 3.7879554655870443e-06,
+      "loss": 3.084,
+      "step": 499
+    },
+    {
+      "epoch": 0.12651821862348178,
+      "grad_norm": 4.15625,
+      "learning_rate": 3.7955465587044536e-06,
+      "loss": 3.1654,
+      "step": 500
+    },
+    {
+      "epoch": 0.12677125506072875,
+      "grad_norm": 3.890625,
+      "learning_rate": 3.803137651821863e-06,
+      "loss": 3.1123,
+      "step": 501
+    },
+    {
+      "epoch": 0.1270242914979757,
+      "grad_norm": 4.3125,
+      "learning_rate": 3.810728744939271e-06,
+      "loss": 3.184,
+      "step": 502
+    },
+    {
+      "epoch": 0.12727732793522267,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.818319838056681e-06,
+      "loss": 3.159,
+      "step": 503
+    },
+    {
+      "epoch": 0.12753036437246965,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.82591093117409e-06,
+      "loss": 3.1984,
+      "step": 504
+    },
+    {
+      "epoch": 0.1277834008097166,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.8335020242914974e-06,
+      "loss": 3.0401,
+      "step": 505
+    },
+    {
+      "epoch": 0.12803643724696356,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.841093117408907e-06,
+      "loss": 3.1182,
+      "step": 506
+    },
+    {
+      "epoch": 0.12828947368421054,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.848684210526316e-06,
+      "loss": 3.0156,
+      "step": 507
+    },
+    {
+      "epoch": 0.12854251012145748,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.856275303643724e-06,
+      "loss": 3.0785,
+      "step": 508
+    },
+    {
+      "epoch": 0.12879554655870445,
+      "grad_norm": 4.125,
+      "learning_rate": 3.863866396761134e-06,
+      "loss": 3.1563,
+      "step": 509
+    },
+    {
+      "epoch": 0.12904858299595143,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.871457489878543e-06,
+      "loss": 3.1553,
+      "step": 510
+    },
+    {
+      "epoch": 0.12904858299595143,
+      "eval_loss": 3.1954586505889893,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.515,
+      "eval_samples_per_second": 9.506,
+      "eval_steps_per_second": 9.506,
+      "step": 510
+    },
+    {
+      "epoch": 0.12930161943319837,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.879048582995951e-06,
+      "loss": 2.9564,
+      "step": 511
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.886639676113361e-06,
+      "loss": 2.9836,
+      "step": 512
+    },
+    {
+      "epoch": 0.12980769230769232,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.89423076923077e-06,
+      "loss": 2.9772,
+      "step": 513
+    },
+    {
+      "epoch": 0.13006072874493926,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.901821862348178e-06,
+      "loss": 3.1529,
+      "step": 514
+    },
+    {
+      "epoch": 0.13031376518218624,
+      "grad_norm": 3.890625,
+      "learning_rate": 3.909412955465587e-06,
+      "loss": 3.1729,
+      "step": 515
+    },
+    {
+      "epoch": 0.1305668016194332,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.917004048582996e-06,
+      "loss": 3.0057,
+      "step": 516
+    },
+    {
+      "epoch": 0.13081983805668015,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.9245951417004045e-06,
+      "loss": 3.2076,
+      "step": 517
+    },
+    {
+      "epoch": 0.13107287449392713,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.932186234817814e-06,
+      "loss": 3.1654,
+      "step": 518
+    },
+    {
+      "epoch": 0.1313259109311741,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.939777327935223e-06,
+      "loss": 3.1709,
+      "step": 519
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 3.8125,
+      "learning_rate": 3.9473684210526315e-06,
+      "loss": 3.1499,
+      "step": 520
+    },
+    {
+      "epoch": 0.13183198380566802,
+      "grad_norm": 3.90625,
+      "learning_rate": 3.954959514170041e-06,
+      "loss": 3.1163,
+      "step": 521
+    },
+    {
+      "epoch": 0.132085020242915,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.96255060728745e-06,
+      "loss": 2.9688,
+      "step": 522
+    },
+    {
+      "epoch": 0.13233805668016194,
+      "grad_norm": 3.5,
+      "learning_rate": 3.9701417004048584e-06,
+      "loss": 3.091,
+      "step": 523
+    },
+    {
+      "epoch": 0.1325910931174089,
+      "grad_norm": 4.125,
+      "learning_rate": 3.977732793522268e-06,
+      "loss": 3.126,
+      "step": 524
+    },
+    {
+      "epoch": 0.13284412955465588,
+      "grad_norm": 3.828125,
+      "learning_rate": 3.985323886639677e-06,
+      "loss": 3.0701,
+      "step": 525
+    },
+    {
+      "epoch": 0.13309716599190283,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.9929149797570846e-06,
+      "loss": 3.1333,
+      "step": 526
+    },
+    {
+      "epoch": 0.1333502024291498,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.000506072874494e-06,
+      "loss": 3.1486,
+      "step": 527
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.008097165991903e-06,
+      "loss": 3.0486,
+      "step": 528
+    },
+    {
+      "epoch": 0.13385627530364372,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.0156882591093115e-06,
+      "loss": 3.0774,
+      "step": 529
+    },
+    {
+      "epoch": 0.1341093117408907,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.023279352226721e-06,
+      "loss": 2.9626,
+      "step": 530
+    },
+    {
+      "epoch": 0.13436234817813766,
+      "grad_norm": 3.625,
+      "learning_rate": 4.03087044534413e-06,
+      "loss": 3.0822,
+      "step": 531
+    },
+    {
+      "epoch": 0.1346153846153846,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.0384615384615385e-06,
+      "loss": 3.0098,
+      "step": 532
+    },
+    {
+      "epoch": 0.13486842105263158,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.046052631578948e-06,
+      "loss": 3.1393,
+      "step": 533
+    },
+    {
+      "epoch": 0.13512145748987855,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.053643724696357e-06,
+      "loss": 3.1223,
+      "step": 534
+    },
+    {
+      "epoch": 0.1353744939271255,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.0612348178137655e-06,
+      "loss": 2.9687,
+      "step": 535
+    },
+    {
+      "epoch": 0.13562753036437247,
+      "grad_norm": 73.0,
+      "learning_rate": 4.068825910931174e-06,
+      "loss": 3.1831,
+      "step": 536
+    },
+    {
+      "epoch": 0.13588056680161945,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.076417004048583e-06,
+      "loss": 3.1062,
+      "step": 537
+    },
+    {
+      "epoch": 0.1361336032388664,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.084008097165992e-06,
+      "loss": 3.1839,
+      "step": 538
+    },
+    {
+      "epoch": 0.13638663967611336,
+      "grad_norm": 3.875,
+      "learning_rate": 4.091599190283401e-06,
+      "loss": 3.2083,
+      "step": 539
+    },
+    {
+      "epoch": 0.13663967611336034,
+      "grad_norm": 3.75,
+      "learning_rate": 4.09919028340081e-06,
+      "loss": 3.1296,
+      "step": 540
+    },
+    {
+      "epoch": 0.13663967611336034,
+      "eval_loss": 3.1818010807037354,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.2506,
+      "eval_samples_per_second": 9.692,
+      "eval_steps_per_second": 9.692,
+      "step": 540
+    },
+    {
+      "epoch": 0.13689271255060728,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.1067813765182186e-06,
+      "loss": 3.1835,
+      "step": 541
+    },
+    {
+      "epoch": 0.13714574898785425,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.114372469635628e-06,
+      "loss": 2.9988,
+      "step": 542
+    },
+    {
+      "epoch": 0.13739878542510123,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.121963562753037e-06,
+      "loss": 3.1326,
+      "step": 543
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.1295546558704455e-06,
+      "loss": 3.17,
+      "step": 544
+    },
+    {
+      "epoch": 0.13790485829959515,
+      "grad_norm": 104.0,
+      "learning_rate": 4.137145748987855e-06,
+      "loss": 3.1791,
+      "step": 545
+    },
+    {
+      "epoch": 0.13815789473684212,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.144736842105263e-06,
+      "loss": 3.0552,
+      "step": 546
+    },
+    {
+      "epoch": 0.13841093117408906,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.152327935222672e-06,
+      "loss": 3.1413,
+      "step": 547
+    },
+    {
+      "epoch": 0.13866396761133604,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.159919028340081e-06,
+      "loss": 3.0183,
+      "step": 548
+    },
+    {
+      "epoch": 0.138917004048583,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.16751012145749e-06,
+      "loss": 3.1236,
+      "step": 549
+    },
+    {
+      "epoch": 0.13917004048582995,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.175101214574899e-06,
+      "loss": 3.147,
+      "step": 550
+    },
+    {
+      "epoch": 0.13942307692307693,
+      "grad_norm": 3.75,
+      "learning_rate": 4.182692307692308e-06,
+      "loss": 3.1591,
+      "step": 551
+    },
+    {
+      "epoch": 0.1396761133603239,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.190283400809717e-06,
+      "loss": 3.1326,
+      "step": 552
+    },
+    {
+      "epoch": 0.13992914979757085,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.197874493927126e-06,
+      "loss": 3.1756,
+      "step": 553
+    },
+    {
+      "epoch": 0.14018218623481782,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.205465587044535e-06,
+      "loss": 3.1109,
+      "step": 554
+    },
+    {
+      "epoch": 0.1404352226720648,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.213056680161944e-06,
+      "loss": 3.1741,
+      "step": 555
+    },
+    {
+      "epoch": 0.14068825910931174,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.220647773279353e-06,
+      "loss": 3.1985,
+      "step": 556
+    },
+    {
+      "epoch": 0.1409412955465587,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.228238866396761e-06,
+      "loss": 3.0959,
+      "step": 557
+    },
+    {
+      "epoch": 0.14119433198380565,
+      "grad_norm": 3.875,
+      "learning_rate": 4.2358299595141694e-06,
+      "loss": 3.0607,
+      "step": 558
+    },
+    {
+      "epoch": 0.14144736842105263,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.243421052631579e-06,
+      "loss": 3.031,
+      "step": 559
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.251012145748988e-06,
+      "loss": 3.1635,
+      "step": 560
+    },
+    {
+      "epoch": 0.14195344129554655,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.258603238866396e-06,
+      "loss": 3.0872,
+      "step": 561
+    },
+    {
+      "epoch": 0.14220647773279352,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.266194331983806e-06,
+      "loss": 3.1187,
+      "step": 562
+    },
+    {
+      "epoch": 0.1424595141700405,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.273785425101215e-06,
+      "loss": 2.9782,
+      "step": 563
+    },
+    {
+      "epoch": 0.14271255060728744,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.281376518218623e-06,
+      "loss": 3.058,
+      "step": 564
+    },
+    {
+      "epoch": 0.1429655870445344,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.288967611336033e-06,
+      "loss": 3.0383,
+      "step": 565
+    },
+    {
+      "epoch": 0.14321862348178138,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.296558704453442e-06,
+      "loss": 3.0373,
+      "step": 566
+    },
+    {
+      "epoch": 0.14347165991902833,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.3041497975708495e-06,
+      "loss": 3.1107,
+      "step": 567
+    },
+    {
+      "epoch": 0.1437246963562753,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.311740890688259e-06,
+      "loss": 3.0741,
+      "step": 568
+    },
+    {
+      "epoch": 0.14397773279352227,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.319331983805668e-06,
+      "loss": 3.1091,
+      "step": 569
+    },
+    {
+      "epoch": 0.14423076923076922,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.3269230769230765e-06,
+      "loss": 3.0664,
+      "step": 570
+    },
+    {
+      "epoch": 0.14423076923076922,
+      "eval_loss": 3.168433666229248,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.5097,
+      "eval_samples_per_second": 9.683,
+      "eval_steps_per_second": 9.683,
+      "step": 570
+    },
+    {
+      "epoch": 0.1444838056680162,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.334514170040486e-06,
+      "loss": 3.1092,
+      "step": 571
+    },
+    {
+      "epoch": 0.14473684210526316,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.342105263157895e-06,
+      "loss": 3.0604,
+      "step": 572
+    },
+    {
+      "epoch": 0.1449898785425101,
+      "grad_norm": 87.5,
+      "learning_rate": 4.3496963562753035e-06,
+      "loss": 3.2488,
+      "step": 573
+    },
+    {
+      "epoch": 0.14524291497975708,
+      "grad_norm": 4.25,
+      "learning_rate": 4.357287449392713e-06,
+      "loss": 3.1374,
+      "step": 574
+    },
+    {
+      "epoch": 0.14549595141700405,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.364878542510122e-06,
+      "loss": 3.0683,
+      "step": 575
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.3724696356275304e-06,
+      "loss": 3.1137,
+      "step": 576
+    },
+    {
+      "epoch": 0.14600202429149797,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.38006072874494e-06,
+      "loss": 3.1209,
+      "step": 577
+    },
+    {
+      "epoch": 0.14625506072874495,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.387651821862348e-06,
+      "loss": 3.1248,
+      "step": 578
+    },
+    {
+      "epoch": 0.1465080971659919,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.3952429149797565e-06,
+      "loss": 3.2215,
+      "step": 579
+    },
+    {
+      "epoch": 0.14676113360323886,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.402834008097166e-06,
+      "loss": 3.053,
+      "step": 580
+    },
+    {
+      "epoch": 0.14701417004048584,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.410425101214575e-06,
+      "loss": 3.1138,
+      "step": 581
+    },
+    {
+      "epoch": 0.14726720647773278,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.4180161943319835e-06,
+      "loss": 2.9672,
+      "step": 582
+    },
+    {
+      "epoch": 0.14752024291497975,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.425607287449393e-06,
+      "loss": 3.0974,
+      "step": 583
+    },
+    {
+      "epoch": 0.14777327935222673,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.433198380566802e-06,
+      "loss": 3.0713,
+      "step": 584
+    },
+    {
+      "epoch": 0.14802631578947367,
+      "grad_norm": 3.8125,
+      "learning_rate": 4.4407894736842105e-06,
+      "loss": 3.0797,
+      "step": 585
+    },
+    {
+      "epoch": 0.14827935222672065,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.44838056680162e-06,
+      "loss": 3.0963,
+      "step": 586
+    },
+    {
+      "epoch": 0.14853238866396762,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.455971659919029e-06,
+      "loss": 3.0901,
+      "step": 587
+    },
+    {
+      "epoch": 0.14878542510121456,
+      "grad_norm": 67.5,
+      "learning_rate": 4.463562753036437e-06,
+      "loss": 3.0422,
+      "step": 588
+    },
+    {
+      "epoch": 0.14903846153846154,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.471153846153846e-06,
+      "loss": 3.118,
+      "step": 589
+    },
+    {
+      "epoch": 0.1492914979757085,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.478744939271255e-06,
+      "loss": 3.1169,
+      "step": 590
+    },
+    {
+      "epoch": 0.14954453441295545,
+      "grad_norm": 3.75,
+      "learning_rate": 4.486336032388664e-06,
+      "loss": 3.157,
+      "step": 591
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.493927125506073e-06,
+      "loss": 3.1131,
+      "step": 592
+    },
+    {
+      "epoch": 0.1500506072874494,
+      "grad_norm": 4.125,
+      "learning_rate": 4.501518218623482e-06,
+      "loss": 3.0682,
+      "step": 593
+    },
+    {
+      "epoch": 0.15030364372469635,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.5091093117408906e-06,
+      "loss": 3.055,
+      "step": 594
+    },
+    {
+      "epoch": 0.15055668016194332,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.5167004048583e-06,
+      "loss": 3.1462,
+      "step": 595
+    },
+    {
+      "epoch": 0.1508097165991903,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.524291497975709e-06,
+      "loss": 3.0887,
+      "step": 596
+    },
+    {
+      "epoch": 0.15106275303643724,
+      "grad_norm": 4.0,
+      "learning_rate": 4.5318825910931175e-06,
+      "loss": 3.1579,
+      "step": 597
+    },
+    {
+      "epoch": 0.1513157894736842,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.539473684210527e-06,
+      "loss": 3.0464,
+      "step": 598
+    },
+    {
+      "epoch": 0.15156882591093118,
+      "grad_norm": 3.75,
+      "learning_rate": 4.547064777327935e-06,
+      "loss": 3.1028,
+      "step": 599
+    },
+    {
+      "epoch": 0.15182186234817813,
+      "grad_norm": 3.75,
+      "learning_rate": 4.554655870445344e-06,
+      "loss": 3.0915,
+      "step": 600
+    },
+    {
+      "epoch": 0.15182186234817813,
+      "eval_loss": 3.1530938148498535,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.2905,
+      "eval_samples_per_second": 9.727,
+      "eval_steps_per_second": 9.727,
+      "step": 600
+    },
+    {
+      "epoch": 0.1520748987854251,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.562246963562753e-06,
+      "loss": 2.9882,
+      "step": 601
+    },
+    {
+      "epoch": 0.15232793522267207,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.569838056680162e-06,
+      "loss": 3.0956,
+      "step": 602
+    },
+    {
+      "epoch": 0.15258097165991902,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.577429149797571e-06,
+      "loss": 3.0202,
+      "step": 603
+    },
+    {
+      "epoch": 0.152834008097166,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.58502024291498e-06,
+      "loss": 3.0887,
+      "step": 604
+    },
+    {
+      "epoch": 0.15308704453441296,
+      "grad_norm": 4.0625,
+      "learning_rate": 4.592611336032389e-06,
+      "loss": 3.1785,
+      "step": 605
+    },
+    {
+      "epoch": 0.1533400809716599,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.600202429149798e-06,
+      "loss": 3.1281,
+      "step": 606
+    },
+    {
+      "epoch": 0.15359311740890688,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.607793522267207e-06,
+      "loss": 3.0125,
+      "step": 607
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 3.087,
+      "step": 608
+    },
+    {
+      "epoch": 0.1540991902834008,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.622975708502024e-06,
+      "loss": 3.0733,
+      "step": 609
+    },
+    {
+      "epoch": 0.15435222672064777,
+      "grad_norm": 3.890625,
+      "learning_rate": 4.630566801619433e-06,
+      "loss": 3.0493,
+      "step": 610
+    },
+    {
+      "epoch": 0.15460526315789475,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.638157894736842e-06,
+      "loss": 3.1533,
+      "step": 611
+    },
+    {
+      "epoch": 0.1548582995951417,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.645748987854251e-06,
+      "loss": 3.0949,
+      "step": 612
+    },
+    {
+      "epoch": 0.15511133603238866,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.65334008097166e-06,
+      "loss": 3.1561,
+      "step": 613
+    },
+    {
+      "epoch": 0.15536437246963564,
+      "grad_norm": 3.9375,
+      "learning_rate": 4.660931174089069e-06,
+      "loss": 3.0958,
+      "step": 614
+    },
+    {
+      "epoch": 0.15561740890688258,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.668522267206478e-06,
+      "loss": 3.0825,
+      "step": 615
+    },
+    {
+      "epoch": 0.15587044534412955,
+      "grad_norm": 3.5,
+      "learning_rate": 4.676113360323887e-06,
+      "loss": 2.8651,
+      "step": 616
+    },
+    {
+      "epoch": 0.15612348178137653,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.683704453441296e-06,
+      "loss": 3.0734,
+      "step": 617
+    },
+    {
+      "epoch": 0.15637651821862347,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.691295546558705e-06,
+      "loss": 2.9816,
+      "step": 618
+    },
+    {
+      "epoch": 0.15662955465587045,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.698886639676113e-06,
+      "loss": 3.1125,
+      "step": 619
+    },
+    {
+      "epoch": 0.15688259109311742,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.706477732793522e-06,
+      "loss": 3.1129,
+      "step": 620
+    },
+    {
+      "epoch": 0.15713562753036436,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.714068825910931e-06,
+      "loss": 3.0456,
+      "step": 621
+    },
+    {
+      "epoch": 0.15738866396761134,
+      "grad_norm": 3.640625,
+      "learning_rate": 4.72165991902834e-06,
+      "loss": 3.1425,
+      "step": 622
+    },
+    {
+      "epoch": 0.1576417004048583,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.729251012145749e-06,
+      "loss": 3.0115,
+      "step": 623
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.736842105263158e-06,
+      "loss": 3.101,
+      "step": 624
+    },
+    {
+      "epoch": 0.15814777327935223,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.744433198380567e-06,
+      "loss": 3.0613,
+      "step": 625
+    },
+    {
+      "epoch": 0.1584008097165992,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.752024291497976e-06,
+      "loss": 3.1223,
+      "step": 626
+    },
+    {
+      "epoch": 0.15865384615384615,
+      "grad_norm": 3.75,
+      "learning_rate": 4.759615384615385e-06,
+      "loss": 3.0761,
+      "step": 627
+    },
+    {
+      "epoch": 0.15890688259109312,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.767206477732794e-06,
+      "loss": 3.1282,
+      "step": 628
+    },
+    {
+      "epoch": 0.1591599190283401,
+      "grad_norm": 3.875,
+      "learning_rate": 4.774797570850203e-06,
+      "loss": 3.1324,
+      "step": 629
+    },
+    {
+      "epoch": 0.15941295546558704,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.782388663967611e-06,
+      "loss": 3.061,
+      "step": 630
+    },
+    {
+      "epoch": 0.15941295546558704,
+      "eval_loss": 3.140883684158325,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.7712,
+      "eval_samples_per_second": 9.746,
+      "eval_steps_per_second": 9.746,
+      "step": 630
+    },
+    {
+      "epoch": 0.159665991902834,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.78997975708502e-06,
+      "loss": 3.0887,
+      "step": 631
+    },
+    {
+      "epoch": 0.15991902834008098,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.797570850202429e-06,
+      "loss": 3.0449,
+      "step": 632
+    },
+    {
+      "epoch": 0.16017206477732793,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.805161943319838e-06,
+      "loss": 2.9738,
+      "step": 633
+    },
+    {
+      "epoch": 0.1604251012145749,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.812753036437247e-06,
+      "loss": 2.9846,
+      "step": 634
+    },
+    {
+      "epoch": 0.16067813765182187,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.820344129554656e-06,
+      "loss": 3.1175,
+      "step": 635
+    },
+    {
+      "epoch": 0.16093117408906882,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.827935222672065e-06,
+      "loss": 3.1026,
+      "step": 636
+    },
+    {
+      "epoch": 0.1611842105263158,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.835526315789474e-06,
+      "loss": 3.0183,
+      "step": 637
+    },
+    {
+      "epoch": 0.16143724696356276,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.843117408906883e-06,
+      "loss": 3.0738,
+      "step": 638
+    },
+    {
+      "epoch": 0.1616902834008097,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.850708502024292e-06,
+      "loss": 3.0521,
+      "step": 639
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 3.625,
+      "learning_rate": 4.8582995951417e-06,
+      "loss": 2.8303,
+      "step": 640
+    },
+    {
+      "epoch": 0.16219635627530365,
+      "grad_norm": 3.734375,
+      "learning_rate": 4.8658906882591095e-06,
+      "loss": 3.026,
+      "step": 641
+    },
+    {
+      "epoch": 0.1624493927125506,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.873481781376518e-06,
+      "loss": 3.1411,
+      "step": 642
+    },
+    {
+      "epoch": 0.16270242914979757,
+      "grad_norm": 3.78125,
+      "learning_rate": 4.881072874493927e-06,
+      "loss": 3.0336,
+      "step": 643
+    },
+    {
+      "epoch": 0.16295546558704455,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.8886639676113364e-06,
+      "loss": 3.1253,
+      "step": 644
+    },
+    {
+      "epoch": 0.1632085020242915,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.896255060728745e-06,
+      "loss": 3.0606,
+      "step": 645
+    },
+    {
+      "epoch": 0.16346153846153846,
+      "grad_norm": 3.5,
+      "learning_rate": 4.903846153846154e-06,
+      "loss": 3.0763,
+      "step": 646
+    },
+    {
+      "epoch": 0.16371457489878544,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.911437246963563e-06,
+      "loss": 3.1025,
+      "step": 647
+    },
+    {
+      "epoch": 0.16396761133603238,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.919028340080972e-06,
+      "loss": 3.1177,
+      "step": 648
+    },
+    {
+      "epoch": 0.16422064777327935,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.926619433198381e-06,
+      "loss": 3.0712,
+      "step": 649
+    },
+    {
+      "epoch": 0.16447368421052633,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.9342105263157895e-06,
+      "loss": 3.1288,
+      "step": 650
+    },
+    {
+      "epoch": 0.16472672064777327,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.941801619433198e-06,
+      "loss": 3.136,
+      "step": 651
+    },
+    {
+      "epoch": 0.16497975708502025,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.949392712550607e-06,
+      "loss": 2.9896,
+      "step": 652
+    },
+    {
+      "epoch": 0.16523279352226722,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.9569838056680165e-06,
+      "loss": 3.0562,
+      "step": 653
+    },
+    {
+      "epoch": 0.16548582995951416,
+      "grad_norm": 3.796875,
+      "learning_rate": 4.964574898785425e-06,
+      "loss": 2.9791,
+      "step": 654
+    },
+    {
+      "epoch": 0.16573886639676114,
+      "grad_norm": 3.671875,
+      "learning_rate": 4.972165991902834e-06,
+      "loss": 3.0434,
+      "step": 655
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.9797570850202435e-06,
+      "loss": 3.0241,
+      "step": 656
+    },
+    {
+      "epoch": 0.16624493927125505,
+      "grad_norm": 3.5,
+      "learning_rate": 4.987348178137652e-06,
+      "loss": 3.0723,
+      "step": 657
+    },
+    {
+      "epoch": 0.16649797570850203,
+      "grad_norm": 3.5,
+      "learning_rate": 4.994939271255061e-06,
+      "loss": 2.8955,
+      "step": 658
+    },
+    {
+      "epoch": 0.166751012145749,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.0025303643724704e-06,
+      "loss": 3.0076,
+      "step": 659
+    },
+    {
+      "epoch": 0.16700404858299595,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.010121457489879e-06,
+      "loss": 3.1446,
+      "step": 660
+    },
+    {
+      "epoch": 0.16700404858299595,
+      "eval_loss": 3.130033493041992,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.0463,
+      "eval_samples_per_second": 9.773,
+      "eval_steps_per_second": 9.773,
+      "step": 660
+    },
+    {
+      "epoch": 0.16725708502024292,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.017712550607287e-06,
+      "loss": 3.0264,
+      "step": 661
+    },
+    {
+      "epoch": 0.1675101214574899,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.0253036437246966e-06,
+      "loss": 3.1447,
+      "step": 662
+    },
+    {
+      "epoch": 0.16776315789473684,
+      "grad_norm": 3.625,
+      "learning_rate": 5.032894736842105e-06,
+      "loss": 3.0166,
+      "step": 663
+    },
+    {
+      "epoch": 0.1680161943319838,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.040485829959514e-06,
+      "loss": 3.0525,
+      "step": 664
+    },
+    {
+      "epoch": 0.16826923076923078,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.0480769230769235e-06,
+      "loss": 3.0068,
+      "step": 665
+    },
+    {
+      "epoch": 0.16852226720647773,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.055668016194332e-06,
+      "loss": 3.1059,
+      "step": 666
+    },
+    {
+      "epoch": 0.1687753036437247,
+      "grad_norm": 4.15625,
+      "learning_rate": 5.063259109311741e-06,
+      "loss": 3.0786,
+      "step": 667
+    },
+    {
+      "epoch": 0.16902834008097167,
+      "grad_norm": 3.625,
+      "learning_rate": 5.0708502024291505e-06,
+      "loss": 3.0473,
+      "step": 668
+    },
+    {
+      "epoch": 0.16928137651821862,
+      "grad_norm": 3.828125,
+      "learning_rate": 5.078441295546559e-06,
+      "loss": 3.0347,
+      "step": 669
+    },
+    {
+      "epoch": 0.1695344129554656,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.086032388663968e-06,
+      "loss": 3.0839,
+      "step": 670
+    },
+    {
+      "epoch": 0.16978744939271256,
+      "grad_norm": 3.625,
+      "learning_rate": 5.093623481781377e-06,
+      "loss": 3.122,
+      "step": 671
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.101214574898785e-06,
+      "loss": 2.9409,
+      "step": 672
+    },
+    {
+      "epoch": 0.17029352226720648,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.108805668016194e-06,
+      "loss": 3.0706,
+      "step": 673
+    },
+    {
+      "epoch": 0.17054655870445345,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.116396761133604e-06,
+      "loss": 2.9949,
+      "step": 674
+    },
+    {
+      "epoch": 0.1707995951417004,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.123987854251012e-06,
+      "loss": 3.0063,
+      "step": 675
+    },
+    {
+      "epoch": 0.17105263157894737,
+      "grad_norm": 3.75,
+      "learning_rate": 5.131578947368421e-06,
+      "loss": 3.0772,
+      "step": 676
+    },
+    {
+      "epoch": 0.17130566801619435,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.139170040485831e-06,
+      "loss": 2.8769,
+      "step": 677
+    },
+    {
+      "epoch": 0.1715587044534413,
+      "grad_norm": 4.28125,
+      "learning_rate": 5.146761133603239e-06,
+      "loss": 3.0646,
+      "step": 678
+    },
+    {
+      "epoch": 0.17181174089068826,
+      "grad_norm": 3.75,
+      "learning_rate": 5.154352226720648e-06,
+      "loss": 3.0382,
+      "step": 679
+    },
+    {
+      "epoch": 0.1720647773279352,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.161943319838057e-06,
+      "loss": 3.0102,
+      "step": 680
+    },
+    {
+      "epoch": 0.17231781376518218,
+      "grad_norm": 3.5,
+      "learning_rate": 5.169534412955466e-06,
+      "loss": 2.9247,
+      "step": 681
+    },
+    {
+      "epoch": 0.17257085020242915,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.177125506072874e-06,
+      "loss": 3.0248,
+      "step": 682
+    },
+    {
+      "epoch": 0.1728238866396761,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.184716599190283e-06,
+      "loss": 3.0027,
+      "step": 683
+    },
+    {
+      "epoch": 0.17307692307692307,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.192307692307692e-06,
+      "loss": 3.0657,
+      "step": 684
+    },
+    {
+      "epoch": 0.17332995951417005,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.199898785425101e-06,
+      "loss": 3.0676,
+      "step": 685
+    },
+    {
+      "epoch": 0.173582995951417,
+      "grad_norm": 4.3125,
+      "learning_rate": 5.20748987854251e-06,
+      "loss": 3.0786,
+      "step": 686
+    },
+    {
+      "epoch": 0.17383603238866396,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.215080971659919e-06,
+      "loss": 3.0408,
+      "step": 687
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 3.984375,
+      "learning_rate": 5.222672064777328e-06,
+      "loss": 3.0954,
+      "step": 688
+    },
+    {
+      "epoch": 0.17434210526315788,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.230263157894737e-06,
+      "loss": 3.0865,
+      "step": 689
+    },
+    {
+      "epoch": 0.17459514170040485,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.237854251012146e-06,
+      "loss": 2.9847,
+      "step": 690
+    },
+    {
+      "epoch": 0.17459514170040485,
+      "eval_loss": 3.117781162261963,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.951,
+      "eval_samples_per_second": 9.74,
+      "eval_steps_per_second": 9.74,
+      "step": 690
+    },
+    {
+      "epoch": 0.17484817813765183,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.245445344129555e-06,
+      "loss": 3.0076,
+      "step": 691
+    },
+    {
+      "epoch": 0.17510121457489877,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.253036437246963e-06,
+      "loss": 3.0273,
+      "step": 692
+    },
+    {
+      "epoch": 0.17535425101214575,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.260627530364372e-06,
+      "loss": 2.9823,
+      "step": 693
+    },
+    {
+      "epoch": 0.17560728744939272,
+      "grad_norm": 3.875,
+      "learning_rate": 5.2682186234817815e-06,
+      "loss": 2.9925,
+      "step": 694
+    },
+    {
+      "epoch": 0.17586032388663966,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.27580971659919e-06,
+      "loss": 2.9655,
+      "step": 695
+    },
+    {
+      "epoch": 0.17611336032388664,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.283400809716599e-06,
+      "loss": 2.9784,
+      "step": 696
+    },
+    {
+      "epoch": 0.1763663967611336,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.2909919028340084e-06,
+      "loss": 3.0142,
+      "step": 697
+    },
+    {
+      "epoch": 0.17661943319838055,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.298582995951417e-06,
+      "loss": 3.1247,
+      "step": 698
+    },
+    {
+      "epoch": 0.17687246963562753,
+      "grad_norm": 3.75,
+      "learning_rate": 5.306174089068826e-06,
+      "loss": 2.9389,
+      "step": 699
+    },
+    {
+      "epoch": 0.1771255060728745,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.313765182186235e-06,
+      "loss": 2.9812,
+      "step": 700
+    },
+    {
+      "epoch": 0.17737854251012145,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.321356275303644e-06,
+      "loss": 3.0161,
+      "step": 701
+    },
+    {
+      "epoch": 0.17763157894736842,
+      "grad_norm": 3.921875,
+      "learning_rate": 5.328947368421053e-06,
+      "loss": 3.1051,
+      "step": 702
+    },
+    {
+      "epoch": 0.1778846153846154,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.3365384615384615e-06,
+      "loss": 3.0691,
+      "step": 703
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.34412955465587e-06,
+      "loss": 3.009,
+      "step": 704
+    },
+    {
+      "epoch": 0.1783906882591093,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.351720647773279e-06,
+      "loss": 2.9665,
+      "step": 705
+    },
+    {
+      "epoch": 0.17864372469635628,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.3593117408906885e-06,
+      "loss": 3.0794,
+      "step": 706
+    },
+    {
+      "epoch": 0.17889676113360323,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.366902834008097e-06,
+      "loss": 3.0233,
+      "step": 707
+    },
+    {
+      "epoch": 0.1791497975708502,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.374493927125506e-06,
+      "loss": 3.0635,
+      "step": 708
+    },
+    {
+      "epoch": 0.17940283400809717,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.3820850202429155e-06,
+      "loss": 2.9993,
+      "step": 709
+    },
+    {
+      "epoch": 0.17965587044534412,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.389676113360324e-06,
+      "loss": 3.0302,
+      "step": 710
+    },
+    {
+      "epoch": 0.1799089068825911,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.397267206477733e-06,
+      "loss": 3.1214,
+      "step": 711
+    },
+    {
+      "epoch": 0.18016194331983806,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.4048582995951424e-06,
+      "loss": 2.9722,
+      "step": 712
+    },
+    {
+      "epoch": 0.180414979757085,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.41244939271255e-06,
+      "loss": 3.0057,
+      "step": 713
+    },
+    {
+      "epoch": 0.18066801619433198,
+      "grad_norm": 3.75,
+      "learning_rate": 5.420040485829959e-06,
+      "loss": 3.0673,
+      "step": 714
+    },
+    {
+      "epoch": 0.18092105263157895,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.4276315789473686e-06,
+      "loss": 2.981,
+      "step": 715
+    },
+    {
+      "epoch": 0.1811740890688259,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.435222672064777e-06,
+      "loss": 3.105,
+      "step": 716
+    },
+    {
+      "epoch": 0.18142712550607287,
+      "grad_norm": 4.0,
+      "learning_rate": 5.442813765182186e-06,
+      "loss": 3.1006,
+      "step": 717
+    },
+    {
+      "epoch": 0.18168016194331985,
+      "grad_norm": 3.828125,
+      "learning_rate": 5.4504048582995955e-06,
+      "loss": 3.0558,
+      "step": 718
+    },
+    {
+      "epoch": 0.1819331983805668,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.457995951417004e-06,
+      "loss": 3.0101,
+      "step": 719
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.465587044534413e-06,
+      "loss": 3.0869,
+      "step": 720
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "eval_loss": 3.1021931171417236,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.8959,
+      "eval_samples_per_second": 9.705,
+      "eval_steps_per_second": 9.705,
+      "step": 720
+    },
+    {
+      "epoch": 0.18243927125506074,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.4731781376518225e-06,
+      "loss": 2.9934,
+      "step": 721
+    },
+    {
+      "epoch": 0.18269230769230768,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.480769230769231e-06,
+      "loss": 3.0707,
+      "step": 722
+    },
+    {
+      "epoch": 0.18294534412955465,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.488360323886639e-06,
+      "loss": 3.0158,
+      "step": 723
+    },
+    {
+      "epoch": 0.18319838056680163,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.495951417004049e-06,
+      "loss": 2.9708,
+      "step": 724
+    },
+    {
+      "epoch": 0.18345141700404857,
+      "grad_norm": 4.3125,
+      "learning_rate": 5.503542510121457e-06,
+      "loss": 3.0792,
+      "step": 725
+    },
+    {
+      "epoch": 0.18370445344129555,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.511133603238866e-06,
+      "loss": 3.0909,
+      "step": 726
+    },
+    {
+      "epoch": 0.18395748987854252,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.518724696356276e-06,
+      "loss": 2.9544,
+      "step": 727
+    },
+    {
+      "epoch": 0.18421052631578946,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.526315789473684e-06,
+      "loss": 3.0776,
+      "step": 728
+    },
+    {
+      "epoch": 0.18446356275303644,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.533906882591093e-06,
+      "loss": 3.0672,
+      "step": 729
+    },
+    {
+      "epoch": 0.1847165991902834,
+      "grad_norm": 3.75,
+      "learning_rate": 5.541497975708503e-06,
+      "loss": 3.0432,
+      "step": 730
+    },
+    {
+      "epoch": 0.18496963562753035,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.549089068825911e-06,
+      "loss": 3.082,
+      "step": 731
+    },
+    {
+      "epoch": 0.18522267206477733,
+      "grad_norm": 3.75,
+      "learning_rate": 5.55668016194332e-06,
+      "loss": 3.0966,
+      "step": 732
+    },
+    {
+      "epoch": 0.1854757085020243,
+      "grad_norm": 4.3125,
+      "learning_rate": 5.5642712550607296e-06,
+      "loss": 3.0711,
+      "step": 733
+    },
+    {
+      "epoch": 0.18572874493927125,
+      "grad_norm": 3.75,
+      "learning_rate": 5.571862348178137e-06,
+      "loss": 2.9743,
+      "step": 734
+    },
+    {
+      "epoch": 0.18598178137651822,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.579453441295546e-06,
+      "loss": 3.0936,
+      "step": 735
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.587044534412956e-06,
+      "loss": 3.1479,
+      "step": 736
+    },
+    {
+      "epoch": 0.18648785425101214,
+      "grad_norm": 3.875,
+      "learning_rate": 5.594635627530364e-06,
+      "loss": 3.1551,
+      "step": 737
+    },
+    {
+      "epoch": 0.1867408906882591,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.602226720647773e-06,
+      "loss": 3.0295,
+      "step": 738
+    },
+    {
+      "epoch": 0.18699392712550608,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.609817813765183e-06,
+      "loss": 3.0215,
+      "step": 739
+    },
+    {
+      "epoch": 0.18724696356275303,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.617408906882591e-06,
+      "loss": 3.0727,
+      "step": 740
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 3.96875,
+      "learning_rate": 5.625e-06,
+      "loss": 3.057,
+      "step": 741
+    },
+    {
+      "epoch": 0.18775303643724697,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.63259109311741e-06,
+      "loss": 3.0477,
+      "step": 742
+    },
+    {
+      "epoch": 0.18800607287449392,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.640182186234818e-06,
+      "loss": 2.9999,
+      "step": 743
+    },
+    {
+      "epoch": 0.1882591093117409,
+      "grad_norm": 3.125,
+      "learning_rate": 5.6477732793522265e-06,
+      "loss": 2.7204,
+      "step": 744
+    },
+    {
+      "epoch": 0.18851214574898786,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.655364372469636e-06,
+      "loss": 2.9973,
+      "step": 745
+    },
+    {
+      "epoch": 0.1887651821862348,
+      "grad_norm": 4.84375,
+      "learning_rate": 5.662955465587044e-06,
+      "loss": 3.1209,
+      "step": 746
+    },
+    {
+      "epoch": 0.18901821862348178,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.6705465587044535e-06,
+      "loss": 3.0985,
+      "step": 747
+    },
+    {
+      "epoch": 0.18927125506072875,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.678137651821863e-06,
+      "loss": 3.1077,
+      "step": 748
+    },
+    {
+      "epoch": 0.1895242914979757,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.685728744939271e-06,
+      "loss": 2.9847,
+      "step": 749
+    },
+    {
+      "epoch": 0.18977732793522267,
+      "grad_norm": 3.890625,
+      "learning_rate": 5.6933198380566804e-06,
+      "loss": 3.0552,
+      "step": 750
+    },
+    {
+      "epoch": 0.18977732793522267,
+      "eval_loss": 3.0940282344818115,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.1887,
+      "eval_samples_per_second": 9.731,
+      "eval_steps_per_second": 9.731,
+      "step": 750
+    },
+    {
+      "epoch": 0.19003036437246965,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.70091093117409e-06,
+      "loss": 2.9666,
+      "step": 751
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.708502024291498e-06,
+      "loss": 3.0029,
+      "step": 752
+    },
+    {
+      "epoch": 0.19053643724696356,
+      "grad_norm": 3.984375,
+      "learning_rate": 5.716093117408907e-06,
+      "loss": 3.0812,
+      "step": 753
+    },
+    {
+      "epoch": 0.19078947368421054,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.723684210526316e-06,
+      "loss": 3.0695,
+      "step": 754
+    },
+    {
+      "epoch": 0.19104251012145748,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.731275303643724e-06,
+      "loss": 2.9936,
+      "step": 755
+    },
+    {
+      "epoch": 0.19129554655870445,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.7388663967611335e-06,
+      "loss": 3.0603,
+      "step": 756
+    },
+    {
+      "epoch": 0.19154858299595143,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.746457489878543e-06,
+      "loss": 3.1014,
+      "step": 757
+    },
+    {
+      "epoch": 0.19180161943319837,
+      "grad_norm": 3.71875,
+      "learning_rate": 5.754048582995951e-06,
+      "loss": 2.9489,
+      "step": 758
+    },
+    {
+      "epoch": 0.19205465587044535,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.7616396761133605e-06,
+      "loss": 2.9537,
+      "step": 759
+    },
+    {
+      "epoch": 0.19230769230769232,
+      "grad_norm": 3.671875,
+      "learning_rate": 5.76923076923077e-06,
+      "loss": 3.1126,
+      "step": 760
+    },
+    {
+      "epoch": 0.19256072874493926,
+      "grad_norm": 3.53125,
+      "learning_rate": 5.776821862348178e-06,
+      "loss": 2.9752,
+      "step": 761
+    },
+    {
+      "epoch": 0.19281376518218624,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.7844129554655875e-06,
+      "loss": 3.0233,
+      "step": 762
+    },
+    {
+      "epoch": 0.1930668016194332,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.792004048582997e-06,
+      "loss": 2.9834,
+      "step": 763
+    },
+    {
+      "epoch": 0.19331983805668015,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.799595141700405e-06,
+      "loss": 2.9245,
+      "step": 764
+    },
+    {
+      "epoch": 0.19357287449392713,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.807186234817814e-06,
+      "loss": 2.9947,
+      "step": 765
+    },
+    {
+      "epoch": 0.1938259109311741,
+      "grad_norm": 3.890625,
+      "learning_rate": 5.814777327935223e-06,
+      "loss": 3.0783,
+      "step": 766
+    },
+    {
+      "epoch": 0.19407894736842105,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.822368421052631e-06,
+      "loss": 3.0248,
+      "step": 767
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 3.546875,
+      "learning_rate": 5.8299595141700406e-06,
+      "loss": 2.9846,
+      "step": 768
+    },
+    {
+      "epoch": 0.194585020242915,
+      "grad_norm": 3.859375,
+      "learning_rate": 5.83755060728745e-06,
+      "loss": 2.9627,
+      "step": 769
+    },
+    {
+      "epoch": 0.19483805668016194,
+      "grad_norm": 3.6875,
+      "learning_rate": 5.845141700404858e-06,
+      "loss": 2.8714,
+      "step": 770
+    },
+    {
+      "epoch": 0.1950910931174089,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.8527327935222675e-06,
+      "loss": 3.061,
+      "step": 771
+    },
+    {
+      "epoch": 0.19534412955465588,
+      "grad_norm": 4.125,
+      "learning_rate": 5.860323886639677e-06,
+      "loss": 3.08,
+      "step": 772
+    },
+    {
+      "epoch": 0.19559716599190283,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.867914979757085e-06,
+      "loss": 3.0533,
+      "step": 773
+    },
+    {
+      "epoch": 0.1958502024291498,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.8755060728744945e-06,
+      "loss": 3.0369,
+      "step": 774
+    },
+    {
+      "epoch": 0.19610323886639677,
+      "grad_norm": 3.703125,
+      "learning_rate": 5.883097165991903e-06,
+      "loss": 2.9837,
+      "step": 775
+    },
+    {
+      "epoch": 0.19635627530364372,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.890688259109311e-06,
+      "loss": 3.1276,
+      "step": 776
+    },
+    {
+      "epoch": 0.1966093117408907,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.898279352226721e-06,
+      "loss": 2.7699,
+      "step": 777
+    },
+    {
+      "epoch": 0.19686234817813766,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.90587044534413e-06,
+      "loss": 2.868,
+      "step": 778
+    },
+    {
+      "epoch": 0.1971153846153846,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.913461538461538e-06,
+      "loss": 3.0327,
+      "step": 779
+    },
+    {
+      "epoch": 0.19736842105263158,
+      "grad_norm": 3.625,
+      "learning_rate": 5.921052631578948e-06,
+      "loss": 2.9681,
+      "step": 780
+    },
+    {
+      "epoch": 0.19736842105263158,
+      "eval_loss": 3.0829808712005615,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.8615,
+      "eval_samples_per_second": 9.743,
+      "eval_steps_per_second": 9.743,
+      "step": 780
+    },
+    {
+      "epoch": 0.19762145748987855,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.928643724696357e-06,
+      "loss": 3.0904,
+      "step": 781
+    },
+    {
+      "epoch": 0.1978744939271255,
+      "grad_norm": 3.84375,
+      "learning_rate": 5.936234817813765e-06,
+      "loss": 2.8908,
+      "step": 782
+    },
+    {
+      "epoch": 0.19812753036437247,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.943825910931175e-06,
+      "loss": 2.9214,
+      "step": 783
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.951417004048584e-06,
+      "loss": 2.9019,
+      "step": 784
+    },
+    {
+      "epoch": 0.1986336032388664,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.959008097165992e-06,
+      "loss": 3.1392,
+      "step": 785
+    },
+    {
+      "epoch": 0.19888663967611336,
+      "grad_norm": 3.734375,
+      "learning_rate": 5.966599190283401e-06,
+      "loss": 3.0227,
+      "step": 786
+    },
+    {
+      "epoch": 0.19913967611336034,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.97419028340081e-06,
+      "loss": 3.0191,
+      "step": 787
+    },
+    {
+      "epoch": 0.19939271255060728,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.981781376518218e-06,
+      "loss": 2.9431,
+      "step": 788
+    },
+    {
+      "epoch": 0.19964574898785425,
+      "grad_norm": 3.4375,
+      "learning_rate": 5.989372469635628e-06,
+      "loss": 2.8405,
+      "step": 789
+    },
+    {
+      "epoch": 0.19989878542510123,
+      "grad_norm": 3.828125,
+      "learning_rate": 5.996963562753037e-06,
+      "loss": 3.0976,
+      "step": 790
+    },
+    {
+      "epoch": 0.20015182186234817,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.004554655870445e-06,
+      "loss": 2.9744,
+      "step": 791
+    },
+    {
+      "epoch": 0.20040485829959515,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.012145748987855e-06,
+      "loss": 2.8994,
+      "step": 792
+    },
+    {
+      "epoch": 0.20065789473684212,
+      "grad_norm": 3.875,
+      "learning_rate": 6.019736842105264e-06,
+      "loss": 3.0855,
+      "step": 793
+    },
+    {
+      "epoch": 0.20091093117408906,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.027327935222672e-06,
+      "loss": 3.0096,
+      "step": 794
+    },
+    {
+      "epoch": 0.20116396761133604,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.034919028340082e-06,
+      "loss": 2.8351,
+      "step": 795
+    },
+    {
+      "epoch": 0.201417004048583,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.04251012145749e-06,
+      "loss": 3.0369,
+      "step": 796
+    },
+    {
+      "epoch": 0.20167004048582995,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.0501012145748985e-06,
+      "loss": 2.818,
+      "step": 797
+    },
+    {
+      "epoch": 0.20192307692307693,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.057692307692308e-06,
+      "loss": 3.0294,
+      "step": 798
+    },
+    {
+      "epoch": 0.2021761133603239,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.065283400809717e-06,
+      "loss": 3.0175,
+      "step": 799
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 3.875,
+      "learning_rate": 6.0728744939271254e-06,
+      "loss": 3.011,
+      "step": 800
+    },
+    {
+      "epoch": 0.20268218623481782,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.080465587044535e-06,
+      "loss": 3.0326,
+      "step": 801
+    },
+    {
+      "epoch": 0.2029352226720648,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.088056680161944e-06,
+      "loss": 3.0943,
+      "step": 802
+    },
+    {
+      "epoch": 0.20318825910931174,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.095647773279352e-06,
+      "loss": 3.0385,
+      "step": 803
+    },
+    {
+      "epoch": 0.2034412955465587,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.103238866396762e-06,
+      "loss": 2.992,
+      "step": 804
+    },
+    {
+      "epoch": 0.20369433198380565,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.11082995951417e-06,
+      "loss": 2.9779,
+      "step": 805
+    },
+    {
+      "epoch": 0.20394736842105263,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.118421052631579e-06,
+      "loss": 3.0143,
+      "step": 806
+    },
+    {
+      "epoch": 0.2042004048582996,
+      "grad_norm": 3.75,
+      "learning_rate": 6.126012145748988e-06,
+      "loss": 2.9599,
+      "step": 807
+    },
+    {
+      "epoch": 0.20445344129554655,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.133603238866396e-06,
+      "loss": 3.0264,
+      "step": 808
+    },
+    {
+      "epoch": 0.20470647773279352,
+      "grad_norm": 3.796875,
+      "learning_rate": 6.1411943319838055e-06,
+      "loss": 3.0623,
+      "step": 809
+    },
+    {
+      "epoch": 0.2049595141700405,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.148785425101215e-06,
+      "loss": 3.0844,
+      "step": 810
+    },
+    {
+      "epoch": 0.2049595141700405,
+      "eval_loss": 3.072495222091675,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 266.4795,
+      "eval_samples_per_second": 9.757,
+      "eval_steps_per_second": 9.757,
+      "step": 810
+    },
+    {
+      "epoch": 0.20521255060728744,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.156376518218623e-06,
+      "loss": 3.0002,
+      "step": 811
+    },
+    {
+      "epoch": 0.2054655870445344,
+      "grad_norm": 4.0625,
+      "learning_rate": 6.1639676113360325e-06,
+      "loss": 3.0671,
+      "step": 812
+    },
+    {
+      "epoch": 0.20571862348178138,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.171558704453442e-06,
+      "loss": 3.0119,
+      "step": 813
+    },
+    {
+      "epoch": 0.20597165991902833,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.17914979757085e-06,
+      "loss": 2.8645,
+      "step": 814
+    },
+    {
+      "epoch": 0.2062246963562753,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.1867408906882595e-06,
+      "loss": 2.9894,
+      "step": 815
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 3.78125,
+      "learning_rate": 6.194331983805669e-06,
+      "loss": 3.0083,
+      "step": 816
+    },
+    {
+      "epoch": 0.20673076923076922,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.201923076923076e-06,
+      "loss": 3.0582,
+      "step": 817
+    },
+    {
+      "epoch": 0.2069838056680162,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.209514170040486e-06,
+      "loss": 3.0771,
+      "step": 818
+    },
+    {
+      "epoch": 0.20723684210526316,
+      "grad_norm": 3.9375,
+      "learning_rate": 6.217105263157895e-06,
+      "loss": 2.9437,
+      "step": 819
+    },
+    {
+      "epoch": 0.2074898785425101,
+      "grad_norm": 3.875,
+      "learning_rate": 6.224696356275303e-06,
+      "loss": 3.0453,
+      "step": 820
+    },
+    {
+      "epoch": 0.20774291497975708,
+      "grad_norm": 4.0,
+      "learning_rate": 6.2322874493927126e-06,
+      "loss": 3.0598,
+      "step": 821
+    },
+    {
+      "epoch": 0.20799595141700405,
+      "grad_norm": 3.75,
+      "learning_rate": 6.239878542510122e-06,
+      "loss": 2.9714,
+      "step": 822
+    },
+    {
+      "epoch": 0.208248987854251,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.24746963562753e-06,
+      "loss": 2.9837,
+      "step": 823
+    },
+    {
+      "epoch": 0.20850202429149797,
+      "grad_norm": 3.71875,
+      "learning_rate": 6.2550607287449395e-06,
+      "loss": 3.0414,
+      "step": 824
+    },
+    {
+      "epoch": 0.20875506072874495,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.262651821862349e-06,
+      "loss": 3.0496,
+      "step": 825
+    },
+    {
+      "epoch": 0.2090080971659919,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.270242914979757e-06,
+      "loss": 3.014,
+      "step": 826
+    },
+    {
+      "epoch": 0.20926113360323886,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.277834008097166e-06,
+      "loss": 2.9257,
+      "step": 827
+    },
+    {
+      "epoch": 0.20951417004048584,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.285425101214575e-06,
+      "loss": 2.9727,
+      "step": 828
+    },
+    {
+      "epoch": 0.20976720647773278,
+      "grad_norm": 4.09375,
+      "learning_rate": 6.293016194331983e-06,
+      "loss": 3.0289,
+      "step": 829
+    },
+    {
+      "epoch": 0.21002024291497975,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.300607287449393e-06,
+      "loss": 3.0804,
+      "step": 830
+    },
+    {
+      "epoch": 0.21027327935222673,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.308198380566802e-06,
+      "loss": 2.8335,
+      "step": 831
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.31578947368421e-06,
+      "loss": 3.0829,
+      "step": 832
+    },
+    {
+      "epoch": 0.21077935222672065,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.32338056680162e-06,
+      "loss": 3.109,
+      "step": 833
+    },
+    {
+      "epoch": 0.21103238866396762,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.330971659919029e-06,
+      "loss": 2.9035,
+      "step": 834
+    },
+    {
+      "epoch": 0.21128542510121456,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.338562753036437e-06,
+      "loss": 2.9701,
+      "step": 835
+    },
+    {
+      "epoch": 0.21153846153846154,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.3461538461538466e-06,
+      "loss": 2.9422,
+      "step": 836
+    },
+    {
+      "epoch": 0.2117914979757085,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.353744939271256e-06,
+      "loss": 2.9055,
+      "step": 837
+    },
+    {
+      "epoch": 0.21204453441295545,
+      "grad_norm": 3.921875,
+      "learning_rate": 6.3613360323886634e-06,
+      "loss": 3.0181,
+      "step": 838
+    },
+    {
+      "epoch": 0.21229757085020243,
+      "grad_norm": 3.46875,
+      "learning_rate": 6.368927125506073e-06,
+      "loss": 2.8627,
+      "step": 839
+    },
+    {
+      "epoch": 0.2125506072874494,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.376518218623482e-06,
+      "loss": 3.0431,
+      "step": 840
+    },
+    {
+      "epoch": 0.2125506072874494,
+      "eval_loss": 3.0625393390655518,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.1783,
+      "eval_samples_per_second": 9.731,
+      "eval_steps_per_second": 9.731,
+      "step": 840
+    },
+    {
+      "epoch": 0.21280364372469635,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.38410931174089e-06,
+      "loss": 2.981,
+      "step": 841
+    },
+    {
+      "epoch": 0.21305668016194332,
+      "grad_norm": 3.75,
+      "learning_rate": 6.3917004048583e-06,
+      "loss": 3.124,
+      "step": 842
+    },
+    {
+      "epoch": 0.2133097165991903,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.399291497975709e-06,
+      "loss": 2.7321,
+      "step": 843
+    },
+    {
+      "epoch": 0.21356275303643724,
+      "grad_norm": 3.75,
+      "learning_rate": 6.406882591093117e-06,
+      "loss": 2.8939,
+      "step": 844
+    },
+    {
+      "epoch": 0.2138157894736842,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.414473684210527e-06,
+      "loss": 2.9942,
+      "step": 845
+    },
+    {
+      "epoch": 0.21406882591093118,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.422064777327936e-06,
+      "loss": 3.0342,
+      "step": 846
+    },
+    {
+      "epoch": 0.21432186234817813,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.429655870445344e-06,
+      "loss": 3.0008,
+      "step": 847
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 3.71875,
+      "learning_rate": 6.437246963562753e-06,
+      "loss": 3.0006,
+      "step": 848
+    },
+    {
+      "epoch": 0.21482793522267207,
+      "grad_norm": 3.75,
+      "learning_rate": 6.444838056680162e-06,
+      "loss": 2.8308,
+      "step": 849
+    },
+    {
+      "epoch": 0.21508097165991902,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.4524291497975705e-06,
+      "loss": 2.9409,
+      "step": 850
+    },
+    {
+      "epoch": 0.215334008097166,
+      "grad_norm": 4.0,
+      "learning_rate": 6.46002024291498e-06,
+      "loss": 2.9297,
+      "step": 851
+    },
+    {
+      "epoch": 0.21558704453441296,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.467611336032389e-06,
+      "loss": 2.9578,
+      "step": 852
+    },
+    {
+      "epoch": 0.2158400809716599,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.4752024291497974e-06,
+      "loss": 3.0171,
+      "step": 853
+    },
+    {
+      "epoch": 0.21609311740890688,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.482793522267207e-06,
+      "loss": 2.9582,
+      "step": 854
+    },
+    {
+      "epoch": 0.21634615384615385,
+      "grad_norm": 3.375,
+      "learning_rate": 6.490384615384616e-06,
+      "loss": 2.9292,
+      "step": 855
+    },
+    {
+      "epoch": 0.2165991902834008,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.497975708502024e-06,
+      "loss": 2.834,
+      "step": 856
+    },
+    {
+      "epoch": 0.21685222672064777,
+      "grad_norm": 3.875,
+      "learning_rate": 6.505566801619434e-06,
+      "loss": 2.9733,
+      "step": 857
+    },
+    {
+      "epoch": 0.21710526315789475,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.513157894736842e-06,
+      "loss": 3.0931,
+      "step": 858
+    },
+    {
+      "epoch": 0.2173582995951417,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.5207489878542505e-06,
+      "loss": 2.9161,
+      "step": 859
+    },
+    {
+      "epoch": 0.21761133603238866,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.52834008097166e-06,
+      "loss": 3.0451,
+      "step": 860
+    },
+    {
+      "epoch": 0.21786437246963564,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.535931174089069e-06,
+      "loss": 2.9235,
+      "step": 861
+    },
+    {
+      "epoch": 0.21811740890688258,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.5435222672064775e-06,
+      "loss": 2.9254,
+      "step": 862
+    },
+    {
+      "epoch": 0.21837044534412955,
+      "grad_norm": 3.8125,
+      "learning_rate": 6.551113360323887e-06,
+      "loss": 2.9421,
+      "step": 863
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 3.875,
+      "learning_rate": 6.558704453441296e-06,
+      "loss": 3.0251,
+      "step": 864
+    },
+    {
+      "epoch": 0.21887651821862347,
+      "grad_norm": 3.59375,
+      "learning_rate": 6.5662955465587045e-06,
+      "loss": 3.0222,
+      "step": 865
+    },
+    {
+      "epoch": 0.21912955465587045,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.573886639676114e-06,
+      "loss": 3.005,
+      "step": 866
+    },
+    {
+      "epoch": 0.21938259109311742,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.581477732793523e-06,
+      "loss": 3.0496,
+      "step": 867
+    },
+    {
+      "epoch": 0.21963562753036436,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.5890688259109315e-06,
+      "loss": 3.0029,
+      "step": 868
+    },
+    {
+      "epoch": 0.21988866396761134,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.59665991902834e-06,
+      "loss": 2.8275,
+      "step": 869
+    },
+    {
+      "epoch": 0.2201417004048583,
+      "grad_norm": 3.78125,
+      "learning_rate": 6.604251012145749e-06,
+      "loss": 2.8357,
+      "step": 870
+    },
+    {
+      "epoch": 0.2201417004048583,
+      "eval_loss": 3.0529282093048096,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.0794,
+      "eval_samples_per_second": 9.735,
+      "eval_steps_per_second": 9.735,
+      "step": 870
+    },
+    {
+      "epoch": 0.22039473684210525,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.611842105263158e-06,
+      "loss": 3.0159,
+      "step": 871
+    },
+    {
+      "epoch": 0.22064777327935223,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.619433198380567e-06,
+      "loss": 2.884,
+      "step": 872
+    },
+    {
+      "epoch": 0.2209008097165992,
+      "grad_norm": 3.625,
+      "learning_rate": 6.627024291497976e-06,
+      "loss": 2.9771,
+      "step": 873
+    },
+    {
+      "epoch": 0.22115384615384615,
+      "grad_norm": 3.75,
+      "learning_rate": 6.6346153846153846e-06,
+      "loss": 2.9626,
+      "step": 874
+    },
+    {
+      "epoch": 0.22140688259109312,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.642206477732794e-06,
+      "loss": 3.0141,
+      "step": 875
+    },
+    {
+      "epoch": 0.2216599190283401,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.649797570850203e-06,
+      "loss": 2.9443,
+      "step": 876
+    },
+    {
+      "epoch": 0.22191295546558704,
+      "grad_norm": 3.90625,
+      "learning_rate": 6.6573886639676115e-06,
+      "loss": 2.9439,
+      "step": 877
+    },
+    {
+      "epoch": 0.222165991902834,
+      "grad_norm": 4.125,
+      "learning_rate": 6.664979757085021e-06,
+      "loss": 3.0542,
+      "step": 878
+    },
+    {
+      "epoch": 0.22241902834008098,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.672570850202429e-06,
+      "loss": 3.0561,
+      "step": 879
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.680161943319838e-06,
+      "loss": 2.8712,
+      "step": 880
+    },
+    {
+      "epoch": 0.2229251012145749,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.687753036437247e-06,
+      "loss": 3.0827,
+      "step": 881
+    },
+    {
+      "epoch": 0.22317813765182187,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.695344129554656e-06,
+      "loss": 2.9287,
+      "step": 882
+    },
+    {
+      "epoch": 0.22343117408906882,
+      "grad_norm": 3.625,
+      "learning_rate": 6.702935222672065e-06,
+      "loss": 2.947,
+      "step": 883
+    },
+    {
+      "epoch": 0.2236842105263158,
+      "grad_norm": 3.625,
+      "learning_rate": 6.710526315789474e-06,
+      "loss": 2.9827,
+      "step": 884
+    },
+    {
+      "epoch": 0.22393724696356276,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.718117408906883e-06,
+      "loss": 2.797,
+      "step": 885
+    },
+    {
+      "epoch": 0.2241902834008097,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.725708502024292e-06,
+      "loss": 2.9667,
+      "step": 886
+    },
+    {
+      "epoch": 0.22444331983805668,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.733299595141701e-06,
+      "loss": 2.9356,
+      "step": 887
+    },
+    {
+      "epoch": 0.22469635627530365,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.74089068825911e-06,
+      "loss": 3.0176,
+      "step": 888
+    },
+    {
+      "epoch": 0.2249493927125506,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.7484817813765186e-06,
+      "loss": 2.9644,
+      "step": 889
+    },
+    {
+      "epoch": 0.22520242914979757,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.756072874493927e-06,
+      "loss": 2.9709,
+      "step": 890
+    },
+    {
+      "epoch": 0.22545546558704455,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.763663967611336e-06,
+      "loss": 2.8666,
+      "step": 891
+    },
+    {
+      "epoch": 0.2257085020242915,
+      "grad_norm": 3.78125,
+      "learning_rate": 6.771255060728745e-06,
+      "loss": 2.9297,
+      "step": 892
+    },
+    {
+      "epoch": 0.22596153846153846,
+      "grad_norm": 3.828125,
+      "learning_rate": 6.778846153846154e-06,
+      "loss": 2.9836,
+      "step": 893
+    },
+    {
+      "epoch": 0.22621457489878544,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.786437246963563e-06,
+      "loss": 3.1099,
+      "step": 894
+    },
+    {
+      "epoch": 0.22646761133603238,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.794028340080972e-06,
+      "loss": 2.8909,
+      "step": 895
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 3.5,
+      "learning_rate": 6.801619433198381e-06,
+      "loss": 2.9913,
+      "step": 896
+    },
+    {
+      "epoch": 0.22697368421052633,
+      "grad_norm": 3.5,
+      "learning_rate": 6.80921052631579e-06,
+      "loss": 2.8606,
+      "step": 897
+    },
+    {
+      "epoch": 0.22722672064777327,
+      "grad_norm": 3.375,
+      "learning_rate": 6.816801619433199e-06,
+      "loss": 2.8436,
+      "step": 898
+    },
+    {
+      "epoch": 0.22747975708502025,
+      "grad_norm": 3.796875,
+      "learning_rate": 6.824392712550608e-06,
+      "loss": 3.0114,
+      "step": 899
+    },
+    {
+      "epoch": 0.22773279352226722,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.831983805668016e-06,
+      "loss": 2.8732,
+      "step": 900
+    },
+    {
+      "epoch": 0.22773279352226722,
+      "eval_loss": 3.0436346530914307,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.7748,
+      "eval_samples_per_second": 9.71,
+      "eval_steps_per_second": 9.71,
+      "step": 900
+    },
+    {
+      "epoch": 0.22798582995951416,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.839574898785425e-06,
+      "loss": 3.0461,
+      "step": 901
+    },
+    {
+      "epoch": 0.22823886639676114,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.847165991902834e-06,
+      "loss": 2.9108,
+      "step": 902
+    },
+    {
+      "epoch": 0.2284919028340081,
+      "grad_norm": 3.5625,
+      "learning_rate": 6.854757085020243e-06,
+      "loss": 2.8658,
+      "step": 903
+    },
+    {
+      "epoch": 0.22874493927125505,
+      "grad_norm": 3.9375,
+      "learning_rate": 6.862348178137652e-06,
+      "loss": 3.0313,
+      "step": 904
+    },
+    {
+      "epoch": 0.22899797570850203,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.869939271255061e-06,
+      "loss": 2.9759,
+      "step": 905
+    },
+    {
+      "epoch": 0.229251012145749,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.87753036437247e-06,
+      "loss": 3.0148,
+      "step": 906
+    },
+    {
+      "epoch": 0.22950404858299595,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.885121457489879e-06,
+      "loss": 3.0494,
+      "step": 907
+    },
+    {
+      "epoch": 0.22975708502024292,
+      "grad_norm": 3.625,
+      "learning_rate": 6.892712550607288e-06,
+      "loss": 2.9803,
+      "step": 908
+    },
+    {
+      "epoch": 0.2300101214574899,
+      "grad_norm": 3.84375,
+      "learning_rate": 6.900303643724697e-06,
+      "loss": 3.002,
+      "step": 909
+    },
+    {
+      "epoch": 0.23026315789473684,
+      "grad_norm": 3.625,
+      "learning_rate": 6.907894736842106e-06,
+      "loss": 3.0024,
+      "step": 910
+    },
+    {
+      "epoch": 0.2305161943319838,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.915485829959514e-06,
+      "loss": 3.0073,
+      "step": 911
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 2.9849,
+      "step": 912
+    },
+    {
+      "epoch": 0.23102226720647773,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.930668016194332e-06,
+      "loss": 3.0274,
+      "step": 913
+    },
+    {
+      "epoch": 0.2312753036437247,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.938259109311741e-06,
+      "loss": 2.9412,
+      "step": 914
+    },
+    {
+      "epoch": 0.23152834008097167,
+      "grad_norm": 4.1875,
+      "learning_rate": 6.94585020242915e-06,
+      "loss": 3.0406,
+      "step": 915
+    },
+    {
+      "epoch": 0.23178137651821862,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.953441295546559e-06,
+      "loss": 2.7815,
+      "step": 916
+    },
+    {
+      "epoch": 0.2320344129554656,
+      "grad_norm": 3.546875,
+      "learning_rate": 6.961032388663968e-06,
+      "loss": 2.8122,
+      "step": 917
+    },
+    {
+      "epoch": 0.23228744939271256,
+      "grad_norm": 4.4375,
+      "learning_rate": 6.968623481781377e-06,
+      "loss": 3.0035,
+      "step": 918
+    },
+    {
+      "epoch": 0.2325404858299595,
+      "grad_norm": 3.609375,
+      "learning_rate": 6.976214574898786e-06,
+      "loss": 3.0236,
+      "step": 919
+    },
+    {
+      "epoch": 0.23279352226720648,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.983805668016195e-06,
+      "loss": 2.8639,
+      "step": 920
+    },
+    {
+      "epoch": 0.23304655870445345,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.9913967611336035e-06,
+      "loss": 2.9381,
+      "step": 921
+    },
+    {
+      "epoch": 0.2332995951417004,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.998987854251012e-06,
+      "loss": 2.9312,
+      "step": 922
+    },
+    {
+      "epoch": 0.23355263157894737,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.006578947368421e-06,
+      "loss": 2.9508,
+      "step": 923
+    },
+    {
+      "epoch": 0.23380566801619435,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.0141700404858304e-06,
+      "loss": 2.975,
+      "step": 924
+    },
+    {
+      "epoch": 0.2340587044534413,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.021761133603239e-06,
+      "loss": 2.9075,
+      "step": 925
+    },
+    {
+      "epoch": 0.23431174089068826,
+      "grad_norm": 3.875,
+      "learning_rate": 7.029352226720648e-06,
+      "loss": 3.0003,
+      "step": 926
+    },
+    {
+      "epoch": 0.2345647773279352,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.0369433198380566e-06,
+      "loss": 2.9249,
+      "step": 927
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.044534412955466e-06,
+      "loss": 2.8985,
+      "step": 928
+    },
+    {
+      "epoch": 0.23507085020242915,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.052125506072875e-06,
+      "loss": 2.9831,
+      "step": 929
+    },
+    {
+      "epoch": 0.2353238866396761,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.0597165991902835e-06,
+      "loss": 2.8375,
+      "step": 930
+    },
+    {
+      "epoch": 0.2353238866396761,
+      "eval_loss": 3.03554368019104,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.187,
+      "eval_samples_per_second": 9.695,
+      "eval_steps_per_second": 9.695,
+      "step": 930
+    },
+    {
+      "epoch": 0.23557692307692307,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.067307692307692e-06,
+      "loss": 2.9941,
+      "step": 931
+    },
+    {
+      "epoch": 0.23582995951417005,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.074898785425101e-06,
+      "loss": 2.97,
+      "step": 932
+    },
+    {
+      "epoch": 0.236082995951417,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.08248987854251e-06,
+      "loss": 2.958,
+      "step": 933
+    },
+    {
+      "epoch": 0.23633603238866396,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.090080971659919e-06,
+      "loss": 2.9556,
+      "step": 934
+    },
+    {
+      "epoch": 0.23658906882591094,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.097672064777328e-06,
+      "loss": 2.8294,
+      "step": 935
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.105263157894737e-06,
+      "loss": 2.9415,
+      "step": 936
+    },
+    {
+      "epoch": 0.23709514170040485,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.112854251012146e-06,
+      "loss": 3.0067,
+      "step": 937
+    },
+    {
+      "epoch": 0.23734817813765183,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.120445344129555e-06,
+      "loss": 2.9962,
+      "step": 938
+    },
+    {
+      "epoch": 0.23760121457489877,
+      "grad_norm": 3.921875,
+      "learning_rate": 7.128036437246964e-06,
+      "loss": 2.9955,
+      "step": 939
+    },
+    {
+      "epoch": 0.23785425101214575,
+      "grad_norm": 3.75,
+      "learning_rate": 7.135627530364373e-06,
+      "loss": 2.983,
+      "step": 940
+    },
+    {
+      "epoch": 0.23810728744939272,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.143218623481782e-06,
+      "loss": 2.8788,
+      "step": 941
+    },
+    {
+      "epoch": 0.23836032388663966,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.15080971659919e-06,
+      "loss": 2.8914,
+      "step": 942
+    },
+    {
+      "epoch": 0.23861336032388664,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.158400809716599e-06,
+      "loss": 3.0111,
+      "step": 943
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 3.75,
+      "learning_rate": 7.165991902834008e-06,
+      "loss": 3.0225,
+      "step": 944
+    },
+    {
+      "epoch": 0.23911943319838055,
+      "grad_norm": 3.90625,
+      "learning_rate": 7.173582995951417e-06,
+      "loss": 2.9513,
+      "step": 945
+    },
+    {
+      "epoch": 0.23937246963562753,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.181174089068826e-06,
+      "loss": 3.0164,
+      "step": 946
+    },
+    {
+      "epoch": 0.2396255060728745,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.188765182186235e-06,
+      "loss": 3.0156,
+      "step": 947
+    },
+    {
+      "epoch": 0.23987854251012145,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.196356275303644e-06,
+      "loss": 2.887,
+      "step": 948
+    },
+    {
+      "epoch": 0.24013157894736842,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.203947368421053e-06,
+      "loss": 2.9307,
+      "step": 949
+    },
+    {
+      "epoch": 0.2403846153846154,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.211538461538462e-06,
+      "loss": 2.8112,
+      "step": 950
+    },
+    {
+      "epoch": 0.24063765182186234,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.219129554655871e-06,
+      "loss": 2.9613,
+      "step": 951
+    },
+    {
+      "epoch": 0.2408906882591093,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.226720647773279e-06,
+      "loss": 3.056,
+      "step": 952
+    },
+    {
+      "epoch": 0.24114372469635628,
+      "grad_norm": 3.921875,
+      "learning_rate": 7.234311740890688e-06,
+      "loss": 3.0328,
+      "step": 953
+    },
+    {
+      "epoch": 0.24139676113360323,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.241902834008097e-06,
+      "loss": 3.0353,
+      "step": 954
+    },
+    {
+      "epoch": 0.2416497975708502,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.249493927125506e-06,
+      "loss": 2.9346,
+      "step": 955
+    },
+    {
+      "epoch": 0.24190283400809717,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.257085020242915e-06,
+      "loss": 2.8569,
+      "step": 956
+    },
+    {
+      "epoch": 0.24215587044534412,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.264676113360324e-06,
+      "loss": 2.7848,
+      "step": 957
+    },
+    {
+      "epoch": 0.2424089068825911,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.272267206477733e-06,
+      "loss": 3.0763,
+      "step": 958
+    },
+    {
+      "epoch": 0.24266194331983806,
+      "grad_norm": 3.75,
+      "learning_rate": 7.279858299595142e-06,
+      "loss": 2.9264,
+      "step": 959
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.287449392712551e-06,
+      "loss": 2.9192,
+      "step": 960
+    },
+    {
+      "epoch": 0.242914979757085,
+      "eval_loss": 3.024563789367676,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.5823,
+      "eval_samples_per_second": 9.68,
+      "eval_steps_per_second": 9.68,
+      "step": 960
+    },
+    {
+      "epoch": 0.24316801619433198,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.29504048582996e-06,
+      "loss": 2.9254,
+      "step": 961
+    },
+    {
+      "epoch": 0.24342105263157895,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.302631578947368e-06,
+      "loss": 2.7953,
+      "step": 962
+    },
+    {
+      "epoch": 0.2436740890688259,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.310222672064777e-06,
+      "loss": 2.8842,
+      "step": 963
+    },
+    {
+      "epoch": 0.24392712550607287,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.317813765182186e-06,
+      "loss": 3.001,
+      "step": 964
+    },
+    {
+      "epoch": 0.24418016194331985,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.325404858299595e-06,
+      "loss": 2.9251,
+      "step": 965
+    },
+    {
+      "epoch": 0.2444331983805668,
+      "grad_norm": 3.875,
+      "learning_rate": 7.332995951417004e-06,
+      "loss": 3.0608,
+      "step": 966
+    },
+    {
+      "epoch": 0.24468623481781376,
+      "grad_norm": 4.34375,
+      "learning_rate": 7.340587044534413e-06,
+      "loss": 3.0556,
+      "step": 967
+    },
+    {
+      "epoch": 0.24493927125506074,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.348178137651822e-06,
+      "loss": 2.9872,
+      "step": 968
+    },
+    {
+      "epoch": 0.24519230769230768,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.355769230769231e-06,
+      "loss": 2.9104,
+      "step": 969
+    },
+    {
+      "epoch": 0.24544534412955465,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.36336032388664e-06,
+      "loss": 2.9506,
+      "step": 970
+    },
+    {
+      "epoch": 0.24569838056680163,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.370951417004049e-06,
+      "loss": 2.9584,
+      "step": 971
+    },
+    {
+      "epoch": 0.24595141700404857,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.378542510121458e-06,
+      "loss": 2.9835,
+      "step": 972
+    },
+    {
+      "epoch": 0.24620445344129555,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.386133603238866e-06,
+      "loss": 2.9173,
+      "step": 973
+    },
+    {
+      "epoch": 0.24645748987854252,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.3937246963562754e-06,
+      "loss": 2.9387,
+      "step": 974
+    },
+    {
+      "epoch": 0.24671052631578946,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.401315789473684e-06,
+      "loss": 2.7419,
+      "step": 975
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.408906882591093e-06,
+      "loss": 2.9607,
+      "step": 976
+    },
+    {
+      "epoch": 0.2472165991902834,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.416497975708502e-06,
+      "loss": 2.9138,
+      "step": 977
+    },
+    {
+      "epoch": 0.24746963562753035,
+      "grad_norm": 3.5,
+      "learning_rate": 7.424089068825911e-06,
+      "loss": 2.9303,
+      "step": 978
+    },
+    {
+      "epoch": 0.24772267206477733,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.43168016194332e-06,
+      "loss": 2.9095,
+      "step": 979
+    },
+    {
+      "epoch": 0.2479757085020243,
+      "grad_norm": 3.859375,
+      "learning_rate": 7.439271255060729e-06,
+      "loss": 2.9842,
+      "step": 980
+    },
+    {
+      "epoch": 0.24822874493927125,
+      "grad_norm": 78.0,
+      "learning_rate": 7.446862348178138e-06,
+      "loss": 3.0913,
+      "step": 981
+    },
+    {
+      "epoch": 0.24848178137651822,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.454453441295547e-06,
+      "loss": 2.8525,
+      "step": 982
+    },
+    {
+      "epoch": 0.2487348178137652,
+      "grad_norm": 4.25,
+      "learning_rate": 7.4620445344129555e-06,
+      "loss": 2.9152,
+      "step": 983
+    },
+    {
+      "epoch": 0.24898785425101214,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.469635627530364e-06,
+      "loss": 2.9036,
+      "step": 984
+    },
+    {
+      "epoch": 0.2492408906882591,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.477226720647773e-06,
+      "loss": 3.0915,
+      "step": 985
+    },
+    {
+      "epoch": 0.24949392712550608,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.4848178137651825e-06,
+      "loss": 2.7673,
+      "step": 986
+    },
+    {
+      "epoch": 0.24974696356275303,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.492408906882591e-06,
+      "loss": 2.8971,
+      "step": 987
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.5e-06,
+      "loss": 2.939,
+      "step": 988
+    },
+    {
+      "epoch": 0.25025303643724695,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.507591093117409e-06,
+      "loss": 2.8988,
+      "step": 989
+    },
+    {
+      "epoch": 0.25050607287449395,
+      "grad_norm": 3.625,
+      "learning_rate": 7.515182186234819e-06,
+      "loss": 2.9906,
+      "step": 990
+    },
+    {
+      "epoch": 0.25050607287449395,
+      "eval_loss": 3.020565986633301,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.5881,
+      "eval_samples_per_second": 9.716,
+      "eval_steps_per_second": 9.716,
+      "step": 990
+    },
+    {
+      "epoch": 0.2507591093117409,
+      "grad_norm": 69.5,
+      "learning_rate": 7.522773279352227e-06,
+      "loss": 3.0859,
+      "step": 991
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 3.875,
+      "learning_rate": 7.530364372469636e-06,
+      "loss": 2.9378,
+      "step": 992
+    },
+    {
+      "epoch": 0.25126518218623484,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.537955465587046e-06,
+      "loss": 3.0314,
+      "step": 993
+    },
+    {
+      "epoch": 0.2515182186234818,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.545546558704453e-06,
+      "loss": 2.773,
+      "step": 994
+    },
+    {
+      "epoch": 0.2517712550607287,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.553137651821862e-06,
+      "loss": 2.9076,
+      "step": 995
+    },
+    {
+      "epoch": 0.2520242914979757,
+      "grad_norm": 3.625,
+      "learning_rate": 7.560728744939272e-06,
+      "loss": 2.8687,
+      "step": 996
+    },
+    {
+      "epoch": 0.2522773279352227,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.56831983805668e-06,
+      "loss": 2.9919,
+      "step": 997
+    },
+    {
+      "epoch": 0.2525303643724696,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.575910931174089e-06,
+      "loss": 2.9074,
+      "step": 998
+    },
+    {
+      "epoch": 0.2527834008097166,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.583502024291499e-06,
+      "loss": 2.9948,
+      "step": 999
+    },
+    {
+      "epoch": 0.25303643724696356,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.591093117408907e-06,
+      "loss": 3.0021,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2532894736842105,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.598684210526316e-06,
+      "loss": 2.8317,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2535425101214575,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.606275303643726e-06,
+      "loss": 2.9785,
+      "step": 1002
+    },
+    {
+      "epoch": 0.25379554655870445,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.613866396761134e-06,
+      "loss": 2.6755,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2540485829959514,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.621457489878542e-06,
+      "loss": 2.7472,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2543016194331984,
+      "grad_norm": 3.75,
+      "learning_rate": 7.629048582995952e-06,
+      "loss": 3.0165,
+      "step": 1005
+    },
+    {
+      "epoch": 0.25455465587044535,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.636639676113361e-06,
+      "loss": 2.8595,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2548076923076923,
+      "grad_norm": 3.765625,
+      "learning_rate": 7.64423076923077e-06,
+      "loss": 3.013,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.65182186234818e-06,
+      "loss": 2.9154,
+      "step": 1008
+    },
+    {
+      "epoch": 0.25531376518218624,
+      "grad_norm": 3.75,
+      "learning_rate": 7.659412955465586e-06,
+      "loss": 2.9328,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2555668016194332,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.667004048582995e-06,
+      "loss": 2.9184,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2558198380566802,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.674595141700405e-06,
+      "loss": 2.9668,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2560728744939271,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.682186234817813e-06,
+      "loss": 2.9786,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2563259109311741,
+      "grad_norm": 3.921875,
+      "learning_rate": 7.689777327935222e-06,
+      "loss": 2.8943,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2565789473684211,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.697368421052632e-06,
+      "loss": 2.8885,
+      "step": 1014
+    },
+    {
+      "epoch": 0.256831983805668,
+      "grad_norm": 68.5,
+      "learning_rate": 7.70495951417004e-06,
+      "loss": 2.9029,
+      "step": 1015
+    },
+    {
+      "epoch": 0.25708502024291496,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.712550607287449e-06,
+      "loss": 2.8988,
+      "step": 1016
+    },
+    {
+      "epoch": 0.25733805668016196,
+      "grad_norm": 3.859375,
+      "learning_rate": 7.720141700404859e-06,
+      "loss": 3.0451,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2575910931174089,
+      "grad_norm": 1256.0,
+      "learning_rate": 7.727732793522267e-06,
+      "loss": 2.9005,
+      "step": 1018
+    },
+    {
+      "epoch": 0.25784412955465585,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.735323886639676e-06,
+      "loss": 2.9125,
+      "step": 1019
+    },
+    {
+      "epoch": 0.25809716599190285,
+      "grad_norm": 3.75,
+      "learning_rate": 7.742914979757086e-06,
+      "loss": 3.0063,
+      "step": 1020
+    },
+    {
+      "epoch": 0.25809716599190285,
+      "eval_loss": 3.0151379108428955,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.3946,
+      "eval_samples_per_second": 9.687,
+      "eval_steps_per_second": 9.687,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2583502024291498,
+      "grad_norm": 3.75,
+      "learning_rate": 7.750506072874494e-06,
+      "loss": 2.8461,
+      "step": 1021
+    },
+    {
+      "epoch": 0.25860323886639675,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.758097165991903e-06,
+      "loss": 2.9027,
+      "step": 1022
+    },
+    {
+      "epoch": 0.25885627530364375,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.765688259109313e-06,
+      "loss": 2.8541,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.773279352226721e-06,
+      "loss": 3.0479,
+      "step": 1024
+    },
+    {
+      "epoch": 0.25936234817813764,
+      "grad_norm": 3.609375,
+      "learning_rate": 7.78087044534413e-06,
+      "loss": 2.9251,
+      "step": 1025
+    },
+    {
+      "epoch": 0.25961538461538464,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.78846153846154e-06,
+      "loss": 2.7012,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2598684210526316,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.796052631578948e-06,
+      "loss": 3.0136,
+      "step": 1027
+    },
+    {
+      "epoch": 0.2601214574898785,
+      "grad_norm": 4.125,
+      "learning_rate": 7.803643724696357e-06,
+      "loss": 2.9506,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2603744939271255,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.811234817813767e-06,
+      "loss": 3.0497,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2606275303643725,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.818825910931174e-06,
+      "loss": 2.9997,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2608805668016194,
+      "grad_norm": 3.5,
+      "learning_rate": 7.826417004048582e-06,
+      "loss": 2.8981,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2611336032388664,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.834008097165992e-06,
+      "loss": 2.8598,
+      "step": 1032
+    },
+    {
+      "epoch": 0.26138663967611336,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.8415991902834e-06,
+      "loss": 2.919,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2616396761133603,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.849190283400809e-06,
+      "loss": 2.9649,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2618927125506073,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.856781376518219e-06,
+      "loss": 2.9317,
+      "step": 1035
+    },
+    {
+      "epoch": 0.26214574898785425,
+      "grad_norm": 3.75,
+      "learning_rate": 7.864372469635628e-06,
+      "loss": 2.9782,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2623987854251012,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.871963562753036e-06,
+      "loss": 2.8636,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2626518218623482,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.879554655870446e-06,
+      "loss": 3.0152,
+      "step": 1038
+    },
+    {
+      "epoch": 0.26290485829959515,
+      "grad_norm": 3.859375,
+      "learning_rate": 7.887145748987854e-06,
+      "loss": 3.0107,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.894736842105263e-06,
+      "loss": 2.858,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2634109311740891,
+      "grad_norm": 3.65625,
+      "learning_rate": 7.902327935222673e-06,
+      "loss": 2.968,
+      "step": 1041
+    },
+    {
+      "epoch": 0.26366396761133604,
+      "grad_norm": 3.671875,
+      "learning_rate": 7.909919028340081e-06,
+      "loss": 3.0074,
+      "step": 1042
+    },
+    {
+      "epoch": 0.263917004048583,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.91751012145749e-06,
+      "loss": 3.0013,
+      "step": 1043
+    },
+    {
+      "epoch": 0.26417004048583,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.9251012145749e-06,
+      "loss": 2.8247,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2644230769230769,
+      "grad_norm": 3.78125,
+      "learning_rate": 7.932692307692308e-06,
+      "loss": 2.9814,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2646761133603239,
+      "grad_norm": 3.625,
+      "learning_rate": 7.940283400809717e-06,
+      "loss": 2.9166,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2649291497975709,
+      "grad_norm": 3.921875,
+      "learning_rate": 7.947874493927127e-06,
+      "loss": 3.0169,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2651821862348178,
+      "grad_norm": 3.375,
+      "learning_rate": 7.955465587044535e-06,
+      "loss": 2.9492,
+      "step": 1048
+    },
+    {
+      "epoch": 0.26543522267206476,
+      "grad_norm": 3.703125,
+      "learning_rate": 7.963056680161944e-06,
+      "loss": 2.8242,
+      "step": 1049
+    },
+    {
+      "epoch": 0.26568825910931176,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.970647773279354e-06,
+      "loss": 2.9637,
+      "step": 1050
+    },
+    {
+      "epoch": 0.26568825910931176,
+      "eval_loss": 3.0039262771606445,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.7128,
+      "eval_samples_per_second": 9.676,
+      "eval_steps_per_second": 9.676,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2659412955465587,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.97823886639676e-06,
+      "loss": 2.9906,
+      "step": 1051
+    },
+    {
+      "epoch": 0.26619433198380565,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.985829959514169e-06,
+      "loss": 2.8283,
+      "step": 1052
+    },
+    {
+      "epoch": 0.26644736842105265,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.99342105263158e-06,
+      "loss": 2.9329,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2667004048582996,
+      "grad_norm": 4.03125,
+      "learning_rate": 8.001012145748988e-06,
+      "loss": 2.9089,
+      "step": 1054
+    },
+    {
+      "epoch": 0.26695344129554655,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.008603238866396e-06,
+      "loss": 2.8665,
+      "step": 1055
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.016194331983806e-06,
+      "loss": 3.0315,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2674595141700405,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.023785425101215e-06,
+      "loss": 2.9099,
+      "step": 1057
+    },
+    {
+      "epoch": 0.26771255060728744,
+      "grad_norm": 3.875,
+      "learning_rate": 8.031376518218623e-06,
+      "loss": 3.0005,
+      "step": 1058
+    },
+    {
+      "epoch": 0.26796558704453444,
+      "grad_norm": 3.828125,
+      "learning_rate": 8.038967611336033e-06,
+      "loss": 2.9868,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2682186234817814,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.046558704453442e-06,
+      "loss": 2.9623,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2684716599190283,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.05414979757085e-06,
+      "loss": 2.8315,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2687246963562753,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.06174089068826e-06,
+      "loss": 2.9681,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2689777327935223,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.069331983805669e-06,
+      "loss": 2.9399,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2692307692307692,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 2.9838,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2694838056680162,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.084514170040487e-06,
+      "loss": 2.9407,
+      "step": 1065
+    },
+    {
+      "epoch": 0.26973684210526316,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.092105263157896e-06,
+      "loss": 2.8966,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2699898785425101,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.099696356275304e-06,
+      "loss": 2.8865,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2702429149797571,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.107287449392714e-06,
+      "loss": 2.8476,
+      "step": 1068
+    },
+    {
+      "epoch": 0.27049595141700405,
+      "grad_norm": 4.0,
+      "learning_rate": 8.114878542510123e-06,
+      "loss": 2.9218,
+      "step": 1069
+    },
+    {
+      "epoch": 0.270748987854251,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.122469635627531e-06,
+      "loss": 2.9156,
+      "step": 1070
+    },
+    {
+      "epoch": 0.271002024291498,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.130060728744941e-06,
+      "loss": 2.9703,
+      "step": 1071
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.137651821862348e-06,
+      "loss": 2.865,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2715080971659919,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.145242914979756e-06,
+      "loss": 2.8317,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2717611336032389,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.152834008097166e-06,
+      "loss": 2.9029,
+      "step": 1074
+    },
+    {
+      "epoch": 0.27201417004048584,
+      "grad_norm": 3.71875,
+      "learning_rate": 8.160425101214575e-06,
+      "loss": 2.9405,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2722672064777328,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.168016194331983e-06,
+      "loss": 2.9246,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2725202429149798,
+      "grad_norm": 3.484375,
+      "learning_rate": 8.175607287449393e-06,
+      "loss": 2.968,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2727732793522267,
+      "grad_norm": 3.8125,
+      "learning_rate": 8.183198380566802e-06,
+      "loss": 2.8986,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2730263157894737,
+      "grad_norm": 3.8125,
+      "learning_rate": 8.19078947368421e-06,
+      "loss": 2.848,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2732793522267207,
+      "grad_norm": 3.875,
+      "learning_rate": 8.19838056680162e-06,
+      "loss": 2.97,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2732793522267207,
+      "eval_loss": 2.9967875480651855,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.7671,
+      "eval_samples_per_second": 9.638,
+      "eval_steps_per_second": 9.638,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2735323886639676,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.205971659919029e-06,
+      "loss": 2.9221,
+      "step": 1081
+    },
+    {
+      "epoch": 0.27378542510121456,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.213562753036437e-06,
+      "loss": 2.8981,
+      "step": 1082
+    },
+    {
+      "epoch": 0.27403846153846156,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.221153846153847e-06,
+      "loss": 2.8552,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2742914979757085,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.228744939271256e-06,
+      "loss": 3.0222,
+      "step": 1084
+    },
+    {
+      "epoch": 0.27454453441295545,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.236336032388664e-06,
+      "loss": 2.8938,
+      "step": 1085
+    },
+    {
+      "epoch": 0.27479757085020245,
+      "grad_norm": 3.59375,
+      "learning_rate": 8.243927125506074e-06,
+      "loss": 2.8943,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2750506072874494,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.251518218623483e-06,
+      "loss": 2.9536,
+      "step": 1087
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 3.859375,
+      "learning_rate": 8.259109311740891e-06,
+      "loss": 2.9555,
+      "step": 1088
+    },
+    {
+      "epoch": 0.27555668016194335,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.266700404858301e-06,
+      "loss": 3.0256,
+      "step": 1089
+    },
+    {
+      "epoch": 0.2758097165991903,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.27429149797571e-06,
+      "loss": 2.8147,
+      "step": 1090
+    },
+    {
+      "epoch": 0.27606275303643724,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.281882591093118e-06,
+      "loss": 2.9216,
+      "step": 1091
+    },
+    {
+      "epoch": 0.27631578947368424,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.289473684210526e-06,
+      "loss": 2.8393,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2765688259109312,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.297064777327935e-06,
+      "loss": 2.903,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2768218623481781,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.304655870445343e-06,
+      "loss": 2.9203,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2770748987854251,
+      "grad_norm": 3.75,
+      "learning_rate": 8.312246963562753e-06,
+      "loss": 2.9468,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2773279352226721,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.319838056680162e-06,
+      "loss": 2.8718,
+      "step": 1096
+    },
+    {
+      "epoch": 0.277580971659919,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.32742914979757e-06,
+      "loss": 2.8175,
+      "step": 1097
+    },
+    {
+      "epoch": 0.277834008097166,
+      "grad_norm": 3.71875,
+      "learning_rate": 8.33502024291498e-06,
+      "loss": 2.9469,
+      "step": 1098
+    },
+    {
+      "epoch": 0.27808704453441296,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.342611336032389e-06,
+      "loss": 2.9725,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2783400809716599,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.350202429149797e-06,
+      "loss": 2.8763,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2785931174089069,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.357793522267207e-06,
+      "loss": 2.9102,
+      "step": 1101
+    },
+    {
+      "epoch": 0.27884615384615385,
+      "grad_norm": 4.03125,
+      "learning_rate": 8.365384615384616e-06,
+      "loss": 2.9574,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2790991902834008,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.372975708502024e-06,
+      "loss": 2.9552,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 3.8125,
+      "learning_rate": 8.380566801619434e-06,
+      "loss": 2.9349,
+      "step": 1104
+    },
+    {
+      "epoch": 0.27960526315789475,
+      "grad_norm": 3.875,
+      "learning_rate": 8.388157894736843e-06,
+      "loss": 2.934,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2798582995951417,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.395748987854251e-06,
+      "loss": 2.9886,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2801113360323887,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.403340080971661e-06,
+      "loss": 2.89,
+      "step": 1107
+    },
+    {
+      "epoch": 0.28036437246963564,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.41093117408907e-06,
+      "loss": 2.8822,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2806174089068826,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.418522267206478e-06,
+      "loss": 2.9851,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2808704453441296,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.426113360323888e-06,
+      "loss": 2.9545,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2808704453441296,
+      "eval_loss": 2.9902408123016357,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.5078,
+      "eval_samples_per_second": 9.647,
+      "eval_steps_per_second": 9.647,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2811234817813765,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.433704453441297e-06,
+      "loss": 2.8796,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2813765182186235,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.441295546558705e-06,
+      "loss": 2.9779,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2816295546558704,
+      "grad_norm": 3.59375,
+      "learning_rate": 8.448886639676112e-06,
+      "loss": 2.893,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2818825910931174,
+      "grad_norm": 3.75,
+      "learning_rate": 8.456477732793522e-06,
+      "loss": 2.9463,
+      "step": 1114
+    },
+    {
+      "epoch": 0.28213562753036436,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.46406882591093e-06,
+      "loss": 2.9213,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2823886639676113,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.471659919028339e-06,
+      "loss": 2.9211,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2826417004048583,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.479251012145749e-06,
+      "loss": 2.9084,
+      "step": 1117
+    },
+    {
+      "epoch": 0.28289473684210525,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.486842105263157e-06,
+      "loss": 2.8402,
+      "step": 1118
+    },
+    {
+      "epoch": 0.2831477732793522,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.494433198380566e-06,
+      "loss": 2.837,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 3.8125,
+      "learning_rate": 8.502024291497976e-06,
+      "loss": 2.8996,
+      "step": 1120
+    },
+    {
+      "epoch": 0.28365384615384615,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.509615384615384e-06,
+      "loss": 2.8869,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2839068825910931,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.517206477732793e-06,
+      "loss": 2.7914,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2841599190283401,
+      "grad_norm": 3.75,
+      "learning_rate": 8.524797570850203e-06,
+      "loss": 2.8916,
+      "step": 1123
+    },
+    {
+      "epoch": 0.28441295546558704,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.532388663967611e-06,
+      "loss": 2.7972,
+      "step": 1124
+    },
+    {
+      "epoch": 0.284665991902834,
+      "grad_norm": 3.875,
+      "learning_rate": 8.53997975708502e-06,
+      "loss": 3.0252,
+      "step": 1125
+    },
+    {
+      "epoch": 0.284919028340081,
+      "grad_norm": 3.890625,
+      "learning_rate": 8.54757085020243e-06,
+      "loss": 3.022,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2851720647773279,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.555161943319838e-06,
+      "loss": 2.9377,
+      "step": 1127
+    },
+    {
+      "epoch": 0.2854251012145749,
+      "grad_norm": 3.828125,
+      "learning_rate": 8.562753036437247e-06,
+      "loss": 2.8383,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2856781376518219,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.570344129554657e-06,
+      "loss": 2.7892,
+      "step": 1129
+    },
+    {
+      "epoch": 0.2859311740890688,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.577935222672065e-06,
+      "loss": 2.8677,
+      "step": 1130
+    },
+    {
+      "epoch": 0.28618421052631576,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.585526315789474e-06,
+      "loss": 2.9374,
+      "step": 1131
+    },
+    {
+      "epoch": 0.28643724696356276,
+      "grad_norm": 3.953125,
+      "learning_rate": 8.593117408906884e-06,
+      "loss": 3.0515,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2866902834008097,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.600708502024292e-06,
+      "loss": 2.9482,
+      "step": 1133
+    },
+    {
+      "epoch": 0.28694331983805665,
+      "grad_norm": 3.75,
+      "learning_rate": 8.608299595141699e-06,
+      "loss": 2.8495,
+      "step": 1134
+    },
+    {
+      "epoch": 0.28719635627530365,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.615890688259109e-06,
+      "loss": 2.8341,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 3.71875,
+      "learning_rate": 8.623481781376518e-06,
+      "loss": 2.9296,
+      "step": 1136
+    },
+    {
+      "epoch": 0.28770242914979755,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.631072874493926e-06,
+      "loss": 2.8134,
+      "step": 1137
+    },
+    {
+      "epoch": 0.28795546558704455,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.638663967611336e-06,
+      "loss": 2.9784,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2882085020242915,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.646255060728745e-06,
+      "loss": 2.9854,
+      "step": 1139
+    },
+    {
+      "epoch": 0.28846153846153844,
+      "grad_norm": 3.796875,
+      "learning_rate": 8.653846153846153e-06,
+      "loss": 2.9908,
+      "step": 1140
+    },
+    {
+      "epoch": 0.28846153846153844,
+      "eval_loss": 2.9807090759277344,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.7982,
+      "eval_samples_per_second": 9.673,
+      "eval_steps_per_second": 9.673,
+      "step": 1140
+    },
+    {
+      "epoch": 0.28871457489878544,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.661437246963563e-06,
+      "loss": 2.8478,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2889676113360324,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.669028340080972e-06,
+      "loss": 2.9619,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2892206477732793,
+      "grad_norm": 3.953125,
+      "learning_rate": 8.67661943319838e-06,
+      "loss": 2.7608,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2894736842105263,
+      "grad_norm": 3.25,
+      "learning_rate": 8.68421052631579e-06,
+      "loss": 2.7415,
+      "step": 1144
+    },
+    {
+      "epoch": 0.2897267206477733,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.691801619433198e-06,
+      "loss": 2.9274,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2899797570850202,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.699392712550607e-06,
+      "loss": 2.86,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2902327935222672,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.706983805668017e-06,
+      "loss": 2.7974,
+      "step": 1147
+    },
+    {
+      "epoch": 0.29048582995951416,
+      "grad_norm": 3.9375,
+      "learning_rate": 8.714574898785425e-06,
+      "loss": 2.9064,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2907388663967611,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.722165991902834e-06,
+      "loss": 2.8908,
+      "step": 1149
+    },
+    {
+      "epoch": 0.2909919028340081,
+      "grad_norm": 3.625,
+      "learning_rate": 8.729757085020244e-06,
+      "loss": 2.8807,
+      "step": 1150
+    },
+    {
+      "epoch": 0.29124493927125505,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.737348178137652e-06,
+      "loss": 2.8538,
+      "step": 1151
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 3.59375,
+      "learning_rate": 8.744939271255061e-06,
+      "loss": 2.8773,
+      "step": 1152
+    },
+    {
+      "epoch": 0.291751012145749,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.752530364372471e-06,
+      "loss": 2.6735,
+      "step": 1153
+    },
+    {
+      "epoch": 0.29200404858299595,
+      "grad_norm": 3.796875,
+      "learning_rate": 8.76012145748988e-06,
+      "loss": 2.9757,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2922570850202429,
+      "grad_norm": 3.90625,
+      "learning_rate": 8.767712550607286e-06,
+      "loss": 2.984,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2925101214574899,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.775303643724696e-06,
+      "loss": 3.0195,
+      "step": 1156
+    },
+    {
+      "epoch": 0.29276315789473684,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.782894736842105e-06,
+      "loss": 2.5399,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2930161943319838,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.790485829959513e-06,
+      "loss": 2.9217,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2932692307692308,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.798076923076923e-06,
+      "loss": 2.9491,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2935222672064777,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.805668016194332e-06,
+      "loss": 2.889,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2937753036437247,
+      "grad_norm": 3.8125,
+      "learning_rate": 8.81325910931174e-06,
+      "loss": 2.9357,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2940283400809717,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.82085020242915e-06,
+      "loss": 2.9689,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2942813765182186,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.828441295546559e-06,
+      "loss": 2.8745,
+      "step": 1163
+    },
+    {
+      "epoch": 0.29453441295546556,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.836032388663967e-06,
+      "loss": 2.9195,
+      "step": 1164
+    },
+    {
+      "epoch": 0.29478744939271256,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.843623481781377e-06,
+      "loss": 2.8411,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2950404858299595,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.851214574898786e-06,
+      "loss": 2.9161,
+      "step": 1166
+    },
+    {
+      "epoch": 0.29529352226720645,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.858805668016194e-06,
+      "loss": 2.9135,
+      "step": 1167
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.866396761133604e-06,
+      "loss": 2.8757,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2957995951417004,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.873987854251013e-06,
+      "loss": 2.8632,
+      "step": 1169
+    },
+    {
+      "epoch": 0.29605263157894735,
+      "grad_norm": 3.484375,
+      "learning_rate": 8.881578947368421e-06,
+      "loss": 2.7023,
+      "step": 1170
+    },
+    {
+      "epoch": 0.29605263157894735,
+      "eval_loss": 2.975774049758911,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.8235,
+      "eval_samples_per_second": 9.708,
+      "eval_steps_per_second": 9.708,
+      "step": 1170
+    },
+    {
+      "epoch": 0.29630566801619435,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.889170040485831e-06,
+      "loss": 2.9302,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2965587044534413,
+      "grad_norm": 3.9375,
+      "learning_rate": 8.89676113360324e-06,
+      "loss": 2.9749,
+      "step": 1172
+    },
+    {
+      "epoch": 0.29681174089068824,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.904352226720648e-06,
+      "loss": 2.7189,
+      "step": 1173
+    },
+    {
+      "epoch": 0.29706477732793524,
+      "grad_norm": 3.515625,
+      "learning_rate": 8.911943319838058e-06,
+      "loss": 2.7457,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2973178137651822,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.919534412955467e-06,
+      "loss": 2.8857,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2975708502024291,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.927125506072873e-06,
+      "loss": 2.9132,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2978238866396761,
+      "grad_norm": 61.75,
+      "learning_rate": 8.934716599190283e-06,
+      "loss": 3.0332,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2980769230769231,
+      "grad_norm": 5.34375,
+      "learning_rate": 8.942307692307692e-06,
+      "loss": 2.8025,
+      "step": 1178
+    },
+    {
+      "epoch": 0.29832995951417,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.9498987854251e-06,
+      "loss": 2.9328,
+      "step": 1179
+    },
+    {
+      "epoch": 0.298582995951417,
+      "grad_norm": 3.71875,
+      "learning_rate": 8.95748987854251e-06,
+      "loss": 2.9371,
+      "step": 1180
+    },
+    {
+      "epoch": 0.29883603238866396,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.965080971659919e-06,
+      "loss": 2.8515,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2990890688259109,
+      "grad_norm": 3.734375,
+      "learning_rate": 8.972672064777327e-06,
+      "loss": 2.9722,
+      "step": 1182
+    },
+    {
+      "epoch": 0.2993421052631579,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.980263157894737e-06,
+      "loss": 2.9224,
+      "step": 1183
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.987854251012146e-06,
+      "loss": 2.8186,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2998481781376518,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.995445344129554e-06,
+      "loss": 2.8148,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3001012145748988,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.003036437246964e-06,
+      "loss": 2.9725,
+      "step": 1186
+    },
+    {
+      "epoch": 0.30035425101214575,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.010627530364373e-06,
+      "loss": 2.9366,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3006072874493927,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.018218623481781e-06,
+      "loss": 2.8987,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3008603238866397,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.025809716599191e-06,
+      "loss": 2.8621,
+      "step": 1189
+    },
+    {
+      "epoch": 0.30111336032388664,
+      "grad_norm": 3.625,
+      "learning_rate": 9.0334008097166e-06,
+      "loss": 2.8806,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3013663967611336,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.040991902834008e-06,
+      "loss": 2.8258,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3016194331983806,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.048582995951418e-06,
+      "loss": 2.9464,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3018724696356275,
+      "grad_norm": 3.75,
+      "learning_rate": 9.056174089068827e-06,
+      "loss": 2.8922,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3021255060728745,
+      "grad_norm": 3.5,
+      "learning_rate": 9.063765182186235e-06,
+      "loss": 2.8025,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3023785425101215,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.071356275303645e-06,
+      "loss": 2.8231,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3026315789473684,
+      "grad_norm": 3.625,
+      "learning_rate": 9.078947368421054e-06,
+      "loss": 2.8489,
+      "step": 1196
+    },
+    {
+      "epoch": 0.30288461538461536,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.08653846153846e-06,
+      "loss": 2.9189,
+      "step": 1197
+    },
+    {
+      "epoch": 0.30313765182186236,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.09412955465587e-06,
+      "loss": 2.9256,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3033906882591093,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.101720647773279e-06,
+      "loss": 2.9023,
+      "step": 1199
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.109311740890687e-06,
+      "loss": 2.9562,
+      "step": 1200
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "eval_loss": 2.9728212356567383,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.6452,
+      "eval_samples_per_second": 9.642,
+      "eval_steps_per_second": 9.642,
+      "step": 1200
+    },
+    {
+      "epoch": 0.30389676113360325,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.116902834008097e-06,
+      "loss": 2.9531,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3041497975708502,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.124493927125506e-06,
+      "loss": 2.8684,
+      "step": 1202
+    },
+    {
+      "epoch": 0.30440283400809715,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.132085020242914e-06,
+      "loss": 2.779,
+      "step": 1203
+    },
+    {
+      "epoch": 0.30465587044534415,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.139676113360324e-06,
+      "loss": 2.9962,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3049089068825911,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.147267206477733e-06,
+      "loss": 2.869,
+      "step": 1205
+    },
+    {
+      "epoch": 0.30516194331983804,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.154858299595141e-06,
+      "loss": 2.9377,
+      "step": 1206
+    },
+    {
+      "epoch": 0.30541497975708504,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.162449392712551e-06,
+      "loss": 2.947,
+      "step": 1207
+    },
+    {
+      "epoch": 0.305668016194332,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.17004048582996e-06,
+      "loss": 2.9212,
+      "step": 1208
+    },
+    {
+      "epoch": 0.3059210526315789,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.177631578947368e-06,
+      "loss": 2.962,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3061740890688259,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.185222672064778e-06,
+      "loss": 2.9391,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3064271255060729,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.192813765182187e-06,
+      "loss": 2.9115,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3066801619433198,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.200404858299595e-06,
+      "loss": 2.933,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3069331983805668,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.207995951417005e-06,
+      "loss": 2.8577,
+      "step": 1213
+    },
+    {
+      "epoch": 0.30718623481781376,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.215587044534414e-06,
+      "loss": 2.9098,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3074392712550607,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.223178137651822e-06,
+      "loss": 2.671,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 2.885,
+      "step": 1216
+    },
+    {
+      "epoch": 0.30794534412955465,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.238360323886639e-06,
+      "loss": 2.9135,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3081983805668016,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.245951417004047e-06,
+      "loss": 2.9043,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3084514170040486,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.253542510121458e-06,
+      "loss": 2.9284,
+      "step": 1219
+    },
+    {
+      "epoch": 0.30870445344129555,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.261133603238866e-06,
+      "loss": 2.8474,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3089574898785425,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.268724696356274e-06,
+      "loss": 3.0091,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3092105263157895,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.276315789473685e-06,
+      "loss": 2.9275,
+      "step": 1222
+    },
+    {
+      "epoch": 0.30946356275303644,
+      "grad_norm": 3.5,
+      "learning_rate": 9.283906882591093e-06,
+      "loss": 2.9643,
+      "step": 1223
+    },
+    {
+      "epoch": 0.3097165991902834,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.291497975708501e-06,
+      "loss": 2.92,
+      "step": 1224
+    },
+    {
+      "epoch": 0.3099696356275304,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.299089068825912e-06,
+      "loss": 2.8997,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3102226720647773,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.30668016194332e-06,
+      "loss": 2.8654,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3104757085020243,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.314271255060728e-06,
+      "loss": 2.9156,
+      "step": 1227
+    },
+    {
+      "epoch": 0.3107287449392713,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.321862348178139e-06,
+      "loss": 2.9395,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3109817813765182,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.329453441295547e-06,
+      "loss": 2.7711,
+      "step": 1229
+    },
+    {
+      "epoch": 0.31123481781376516,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.337044534412955e-06,
+      "loss": 2.7797,
+      "step": 1230
+    },
+    {
+      "epoch": 0.31123481781376516,
+      "eval_loss": 2.969989538192749,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.3684,
+      "eval_samples_per_second": 9.688,
+      "eval_steps_per_second": 9.688,
+      "step": 1230
+    },
+    {
+      "epoch": 0.31148785425101216,
+      "grad_norm": 3.5,
+      "learning_rate": 9.344635627530365e-06,
+      "loss": 2.8949,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.352226720647774e-06,
+      "loss": 2.9452,
+      "step": 1232
+    },
+    {
+      "epoch": 0.31199392712550605,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.359817813765182e-06,
+      "loss": 2.848,
+      "step": 1233
+    },
+    {
+      "epoch": 0.31224696356275305,
+      "grad_norm": 3.75,
+      "learning_rate": 9.367408906882592e-06,
+      "loss": 2.8329,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 2.8923,
+      "step": 1235
+    },
+    {
+      "epoch": 0.31275303643724695,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.38259109311741e-06,
+      "loss": 2.7194,
+      "step": 1236
+    },
+    {
+      "epoch": 0.31300607287449395,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.39018218623482e-06,
+      "loss": 2.7406,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3132591093117409,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.397773279352226e-06,
+      "loss": 2.8938,
+      "step": 1238
+    },
+    {
+      "epoch": 0.31351214574898784,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.405364372469635e-06,
+      "loss": 2.8918,
+      "step": 1239
+    },
+    {
+      "epoch": 0.31376518218623484,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.412955465587045e-06,
+      "loss": 2.8985,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3140182186234818,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.420546558704453e-06,
+      "loss": 2.868,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3142712550607287,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.428137651821862e-06,
+      "loss": 2.9559,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3145242914979757,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.435728744939272e-06,
+      "loss": 2.9137,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3147773279352227,
+      "grad_norm": 3.375,
+      "learning_rate": 9.44331983805668e-06,
+      "loss": 2.8524,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3150303643724696,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.450910931174089e-06,
+      "loss": 2.9542,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3152834008097166,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.458502024291499e-06,
+      "loss": 2.9046,
+      "step": 1246
+    },
+    {
+      "epoch": 0.31553643724696356,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.466093117408907e-06,
+      "loss": 2.706,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.473684210526315e-06,
+      "loss": 2.965,
+      "step": 1248
+    },
+    {
+      "epoch": 0.3160425101214575,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.481275303643726e-06,
+      "loss": 2.8791,
+      "step": 1249
+    },
+    {
+      "epoch": 0.31629554655870445,
+      "grad_norm": 3.625,
+      "learning_rate": 9.488866396761134e-06,
+      "loss": 2.9766,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3165485829959514,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.496457489878542e-06,
+      "loss": 2.9323,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3168016194331984,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.504048582995953e-06,
+      "loss": 2.8652,
+      "step": 1252
+    },
+    {
+      "epoch": 0.31705465587044535,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.511639676113361e-06,
+      "loss": 2.8157,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3173076923076923,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.51923076923077e-06,
+      "loss": 2.5954,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3175607287449393,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.52682186234818e-06,
+      "loss": 2.7252,
+      "step": 1255
+    },
+    {
+      "epoch": 0.31781376518218624,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.534412955465588e-06,
+      "loss": 2.73,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3180668016194332,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.542004048582996e-06,
+      "loss": 2.8047,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3183198380566802,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.549595141700407e-06,
+      "loss": 2.9599,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3185728744939271,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.557186234817813e-06,
+      "loss": 2.8608,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3188259109311741,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.564777327935222e-06,
+      "loss": 2.9451,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3188259109311741,
+      "eval_loss": 2.9592697620391846,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 267.9572,
+      "eval_samples_per_second": 9.703,
+      "eval_steps_per_second": 9.703,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3190789473684211,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.572368421052632e-06,
+      "loss": 2.9176,
+      "step": 1261
+    },
+    {
+      "epoch": 0.319331983805668,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.57995951417004e-06,
+      "loss": 2.9645,
+      "step": 1262
+    },
+    {
+      "epoch": 0.31958502024291496,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.587550607287449e-06,
+      "loss": 2.7708,
+      "step": 1263
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.595141700404859e-06,
+      "loss": 2.8215,
+      "step": 1264
+    },
+    {
+      "epoch": 0.3200910931174089,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.602732793522267e-06,
+      "loss": 2.8962,
+      "step": 1265
+    },
+    {
+      "epoch": 0.32034412955465585,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.610323886639676e-06,
+      "loss": 2.9603,
+      "step": 1266
+    },
+    {
+      "epoch": 0.32059716599190285,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.617914979757086e-06,
+      "loss": 2.8116,
+      "step": 1267
+    },
+    {
+      "epoch": 0.3208502024291498,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.625506072874494e-06,
+      "loss": 2.8412,
+      "step": 1268
+    },
+    {
+      "epoch": 0.32110323886639675,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.633097165991903e-06,
+      "loss": 2.8511,
+      "step": 1269
+    },
+    {
+      "epoch": 0.32135627530364375,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.640688259109313e-06,
+      "loss": 2.9943,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3216093117408907,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.648279352226721e-06,
+      "loss": 2.8589,
+      "step": 1271
+    },
+    {
+      "epoch": 0.32186234817813764,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.65587044534413e-06,
+      "loss": 2.8062,
+      "step": 1272
+    },
+    {
+      "epoch": 0.32211538461538464,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.66346153846154e-06,
+      "loss": 2.7886,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3223684210526316,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.671052631578948e-06,
+      "loss": 2.839,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3226214574898785,
+      "grad_norm": 3.5,
+      "learning_rate": 9.678643724696357e-06,
+      "loss": 2.7592,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3228744939271255,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.686234817813767e-06,
+      "loss": 2.9175,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3231275303643725,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.693825910931175e-06,
+      "loss": 2.9578,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3233805668016194,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.701417004048584e-06,
+      "loss": 2.7927,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3236336032388664,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.709008097165994e-06,
+      "loss": 2.9072,
+      "step": 1279
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.7165991902834e-06,
+      "loss": 2.8897,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3241396761133603,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.724190283400809e-06,
+      "loss": 2.8433,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3243927125506073,
+      "grad_norm": 3.5,
+      "learning_rate": 9.731781376518219e-06,
+      "loss": 2.9016,
+      "step": 1282
+    },
+    {
+      "epoch": 0.32464574898785425,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.739372469635627e-06,
+      "loss": 2.8512,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3248987854251012,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.746963562753036e-06,
+      "loss": 2.7053,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3251518218623482,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.754554655870446e-06,
+      "loss": 2.8891,
+      "step": 1285
+    },
+    {
+      "epoch": 0.32540485829959515,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.762145748987854e-06,
+      "loss": 2.8945,
+      "step": 1286
+    },
+    {
+      "epoch": 0.3256578947368421,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.769736842105263e-06,
+      "loss": 2.8821,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3259109311740891,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.777327935222673e-06,
+      "loss": 2.9117,
+      "step": 1288
+    },
+    {
+      "epoch": 0.32616396761133604,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.784919028340081e-06,
+      "loss": 2.8433,
+      "step": 1289
+    },
+    {
+      "epoch": 0.326417004048583,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.79251012145749e-06,
+      "loss": 2.832,
+      "step": 1290
+    },
+    {
+      "epoch": 0.326417004048583,
+      "eval_loss": 2.951796531677246,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.7158,
+      "eval_samples_per_second": 9.676,
+      "eval_steps_per_second": 9.676,
+      "step": 1290
+    },
+    {
+      "epoch": 0.32667004048583,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.8001012145749e-06,
+      "loss": 2.9285,
+      "step": 1291
+    },
+    {
+      "epoch": 0.3269230769230769,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.807692307692308e-06,
+      "loss": 2.8487,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3271761133603239,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.815283400809717e-06,
+      "loss": 2.8962,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3274291497975709,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.822874493927127e-06,
+      "loss": 2.7684,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3276821862348178,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.830465587044535e-06,
+      "loss": 2.7719,
+      "step": 1295
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.838056680161944e-06,
+      "loss": 2.9372,
+      "step": 1296
+    },
+    {
+      "epoch": 0.32818825910931176,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.845647773279354e-06,
+      "loss": 2.9095,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3284412955465587,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.853238866396762e-06,
+      "loss": 2.9565,
+      "step": 1298
+    },
+    {
+      "epoch": 0.32869433198380565,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.86082995951417e-06,
+      "loss": 2.9341,
+      "step": 1299
+    },
+    {
+      "epoch": 0.32894736842105265,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.868421052631579e-06,
+      "loss": 2.9021,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3292004048582996,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.876012145748987e-06,
+      "loss": 2.7993,
+      "step": 1301
+    },
+    {
+      "epoch": 0.32945344129554655,
+      "grad_norm": 3.125,
+      "learning_rate": 9.883603238866396e-06,
+      "loss": 2.6463,
+      "step": 1302
+    },
+    {
+      "epoch": 0.32970647773279355,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.891194331983806e-06,
+      "loss": 2.8827,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3299595141700405,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.898785425101214e-06,
+      "loss": 2.9348,
+      "step": 1304
+    },
+    {
+      "epoch": 0.33021255060728744,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.906376518218623e-06,
+      "loss": 2.8589,
+      "step": 1305
+    },
+    {
+      "epoch": 0.33046558704453444,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.913967611336033e-06,
+      "loss": 2.9056,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3307186234817814,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.921558704453441e-06,
+      "loss": 2.9085,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3309716599190283,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.92914979757085e-06,
+      "loss": 2.7755,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3312246963562753,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.93674089068826e-06,
+      "loss": 2.8342,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3314777327935223,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.944331983805668e-06,
+      "loss": 2.9059,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3317307692307692,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.951923076923077e-06,
+      "loss": 2.9841,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 3.75,
+      "learning_rate": 9.959514170040487e-06,
+      "loss": 2.8514,
+      "step": 1312
+    },
+    {
+      "epoch": 0.33223684210526316,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.967105263157895e-06,
+      "loss": 2.7822,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3324898785425101,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.974696356275304e-06,
+      "loss": 2.9084,
+      "step": 1314
+    },
+    {
+      "epoch": 0.3327429149797571,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.982287449392714e-06,
+      "loss": 2.7868,
+      "step": 1315
+    },
+    {
+      "epoch": 0.33299595141700405,
+      "grad_norm": 3.75,
+      "learning_rate": 9.989878542510122e-06,
+      "loss": 2.9473,
+      "step": 1316
+    },
+    {
+      "epoch": 0.333248987854251,
+      "grad_norm": 3.5,
+      "learning_rate": 9.99746963562753e-06,
+      "loss": 2.8763,
+      "step": 1317
+    },
+    {
+      "epoch": 0.333502024291498,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.0005060728744941e-05,
+      "loss": 2.9091,
+      "step": 1318
+    },
+    {
+      "epoch": 0.33375506072874495,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.001265182186235e-05,
+      "loss": 2.8586,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3340080971659919,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.0020242914979758e-05,
+      "loss": 2.7651,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3340080971659919,
+      "eval_loss": 2.946282148361206,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.0399,
+      "eval_samples_per_second": 9.664,
+      "eval_steps_per_second": 9.664,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3342611336032389,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0027834008097166e-05,
+      "loss": 2.8397,
+      "step": 1321
+    },
+    {
+      "epoch": 0.33451417004048584,
+      "grad_norm": 60.25,
+      "learning_rate": 1.0035425101214575e-05,
+      "loss": 3.1112,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3347672064777328,
+      "grad_norm": 3.625,
+      "learning_rate": 1.0043016194331983e-05,
+      "loss": 2.9344,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3350202429149798,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.0050607287449393e-05,
+      "loss": 2.9408,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3352732793522267,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.0058198380566802e-05,
+      "loss": 2.7242,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3355263157894737,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.006578947368421e-05,
+      "loss": 2.8357,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3357793522267207,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.007338056680162e-05,
+      "loss": 2.8453,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.0080971659919029e-05,
+      "loss": 2.8035,
+      "step": 1328
+    },
+    {
+      "epoch": 0.33628542510121456,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.0088562753036437e-05,
+      "loss": 2.8931,
+      "step": 1329
+    },
+    {
+      "epoch": 0.33653846153846156,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.0096153846153847e-05,
+      "loss": 2.6974,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3367914979757085,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.0103744939271256e-05,
+      "loss": 2.8439,
+      "step": 1331
+    },
+    {
+      "epoch": 0.33704453441295545,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.0111336032388664e-05,
+      "loss": 2.9268,
+      "step": 1332
+    },
+    {
+      "epoch": 0.33729757085020245,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0118927125506074e-05,
+      "loss": 2.834,
+      "step": 1333
+    },
+    {
+      "epoch": 0.3375506072874494,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.0126518218623482e-05,
+      "loss": 2.9913,
+      "step": 1334
+    },
+    {
+      "epoch": 0.33780364372469635,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.0134109311740891e-05,
+      "loss": 2.8761,
+      "step": 1335
+    },
+    {
+      "epoch": 0.33805668016194335,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0141700404858301e-05,
+      "loss": 2.8479,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3383097165991903,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.014929149797571e-05,
+      "loss": 2.841,
+      "step": 1337
+    },
+    {
+      "epoch": 0.33856275303643724,
+      "grad_norm": 3.25,
+      "learning_rate": 1.0156882591093118e-05,
+      "loss": 2.8142,
+      "step": 1338
+    },
+    {
+      "epoch": 0.33881578947368424,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.0164473684210528e-05,
+      "loss": 2.7111,
+      "step": 1339
+    },
+    {
+      "epoch": 0.3390688259109312,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.0172064777327936e-05,
+      "loss": 2.8632,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3393218623481781,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.0179655870445345e-05,
+      "loss": 2.8494,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3395748987854251,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.0187246963562753e-05,
+      "loss": 2.906,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3398279352226721,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0194838056680162e-05,
+      "loss": 2.7345,
+      "step": 1343
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.020242914979757e-05,
+      "loss": 2.8431,
+      "step": 1344
+    },
+    {
+      "epoch": 0.340334008097166,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.021002024291498e-05,
+      "loss": 2.971,
+      "step": 1345
+    },
+    {
+      "epoch": 0.34058704453441296,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.0217611336032389e-05,
+      "loss": 2.8655,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3408400809716599,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.0225202429149797e-05,
+      "loss": 2.7949,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3410931174089069,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0232793522267207e-05,
+      "loss": 2.8569,
+      "step": 1348
+    },
+    {
+      "epoch": 0.34134615384615385,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.0240384615384616e-05,
+      "loss": 2.7065,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3415991902834008,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.0247975708502024e-05,
+      "loss": 2.7166,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3415991902834008,
+      "eval_loss": 2.9426627159118652,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.652,
+      "eval_samples_per_second": 9.678,
+      "eval_steps_per_second": 9.678,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3418522267206478,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.0255566801619434e-05,
+      "loss": 2.8355,
+      "step": 1351
+    },
+    {
+      "epoch": 0.34210526315789475,
+      "grad_norm": 3.75,
+      "learning_rate": 1.0263157894736843e-05,
+      "loss": 2.9467,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3423582995951417,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.0270748987854251e-05,
+      "loss": 2.8529,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3426113360323887,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.0278340080971661e-05,
+      "loss": 2.7004,
+      "step": 1354
+    },
+    {
+      "epoch": 0.34286437246963564,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.028593117408907e-05,
+      "loss": 2.8509,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3431174089068826,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.0293522267206478e-05,
+      "loss": 2.8657,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3433704453441296,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0301113360323888e-05,
+      "loss": 2.7872,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3436234817813765,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0308704453441297e-05,
+      "loss": 2.8088,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3438765182186235,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.0316295546558705e-05,
+      "loss": 2.8266,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.0323886639676113e-05,
+      "loss": 2.8759,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3443825910931174,
+      "grad_norm": 3.875,
+      "learning_rate": 1.0331477732793524e-05,
+      "loss": 2.8379,
+      "step": 1361
+    },
+    {
+      "epoch": 0.34463562753036436,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0339068825910932e-05,
+      "loss": 2.7769,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3448886639676113,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.0346659919028339e-05,
+      "loss": 2.5801,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3451417004048583,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.0354251012145749e-05,
+      "loss": 2.8774,
+      "step": 1364
+    },
+    {
+      "epoch": 0.34539473684210525,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.0361842105263157e-05,
+      "loss": 2.9477,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3456477732793522,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.0369433198380566e-05,
+      "loss": 2.8543,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3459008097165992,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.0377024291497976e-05,
+      "loss": 2.8973,
+      "step": 1367
+    },
+    {
+      "epoch": 0.34615384615384615,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.0384615384615384e-05,
+      "loss": 2.8318,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3464068825910931,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.0392206477732793e-05,
+      "loss": 2.7504,
+      "step": 1369
+    },
+    {
+      "epoch": 0.3466599190283401,
+      "grad_norm": 3.5,
+      "learning_rate": 1.0399797570850203e-05,
+      "loss": 2.8182,
+      "step": 1370
+    },
+    {
+      "epoch": 0.34691295546558704,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.0407388663967611e-05,
+      "loss": 2.9677,
+      "step": 1371
+    },
+    {
+      "epoch": 0.347165991902834,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.041497975708502e-05,
+      "loss": 2.7656,
+      "step": 1372
+    },
+    {
+      "epoch": 0.347419028340081,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.042257085020243e-05,
+      "loss": 2.8334,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3476720647773279,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.0430161943319838e-05,
+      "loss": 3.0108,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3479251012145749,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.0437753036437247e-05,
+      "loss": 2.859,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.0445344129554657e-05,
+      "loss": 2.8655,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3484311740890688,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0452935222672065e-05,
+      "loss": 2.7174,
+      "step": 1377
+    },
+    {
+      "epoch": 0.34868421052631576,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.0460526315789474e-05,
+      "loss": 2.9229,
+      "step": 1378
+    },
+    {
+      "epoch": 0.34893724696356276,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.0468117408906884e-05,
+      "loss": 2.9334,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3491902834008097,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.0475708502024292e-05,
+      "loss": 2.8117,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3491902834008097,
+      "eval_loss": 2.9350547790527344,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.6023,
+      "eval_samples_per_second": 9.68,
+      "eval_steps_per_second": 9.68,
+      "step": 1380
+    },
+    {
+      "epoch": 0.34944331983805665,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.04832995951417e-05,
+      "loss": 2.8085,
+      "step": 1381
+    },
+    {
+      "epoch": 0.34969635627530365,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.049089068825911e-05,
+      "loss": 2.7902,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3499493927125506,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.0498481781376519e-05,
+      "loss": 2.8877,
+      "step": 1383
+    },
+    {
+      "epoch": 0.35020242914979755,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.0506072874493926e-05,
+      "loss": 2.8458,
+      "step": 1384
+    },
+    {
+      "epoch": 0.35045546558704455,
+      "grad_norm": 3.875,
+      "learning_rate": 1.0513663967611336e-05,
+      "loss": 2.8659,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3507085020242915,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0521255060728744e-05,
+      "loss": 2.7834,
+      "step": 1386
+    },
+    {
+      "epoch": 0.35096153846153844,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.0528846153846153e-05,
+      "loss": 2.8242,
+      "step": 1387
+    },
+    {
+      "epoch": 0.35121457489878544,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.0536437246963563e-05,
+      "loss": 2.8893,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3514676113360324,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.0544028340080971e-05,
+      "loss": 2.8367,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3517206477732793,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.055161943319838e-05,
+      "loss": 2.794,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3519736842105263,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.055921052631579e-05,
+      "loss": 2.8147,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.0566801619433198e-05,
+      "loss": 2.7026,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3524797570850202,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.0574392712550607e-05,
+      "loss": 2.9095,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3527327935222672,
+      "grad_norm": 3.625,
+      "learning_rate": 1.0581983805668017e-05,
+      "loss": 2.8642,
+      "step": 1394
+    },
+    {
+      "epoch": 0.35298582995951416,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.0589574898785425e-05,
+      "loss": 2.5124,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3532388663967611,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.0597165991902834e-05,
+      "loss": 2.8208,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3534919028340081,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0604757085020244e-05,
+      "loss": 2.8159,
+      "step": 1397
+    },
+    {
+      "epoch": 0.35374493927125505,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.0612348178137652e-05,
+      "loss": 2.809,
+      "step": 1398
+    },
+    {
+      "epoch": 0.353997975708502,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.061993927125506e-05,
+      "loss": 2.9135,
+      "step": 1399
+    },
+    {
+      "epoch": 0.354251012145749,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.062753036437247e-05,
+      "loss": 2.961,
+      "step": 1400
+    },
+    {
+      "epoch": 0.35450404858299595,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.063512145748988e-05,
+      "loss": 2.8732,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3547570850202429,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.0642712550607288e-05,
+      "loss": 2.7591,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3550101214574899,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.0650303643724698e-05,
+      "loss": 2.764,
+      "step": 1403
+    },
+    {
+      "epoch": 0.35526315789473684,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.0657894736842106e-05,
+      "loss": 2.8278,
+      "step": 1404
+    },
+    {
+      "epoch": 0.3555161943319838,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0665485829959513e-05,
+      "loss": 2.8579,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3557692307692308,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.0673076923076923e-05,
+      "loss": 2.8857,
+      "step": 1406
+    },
+    {
+      "epoch": 0.3560222672064777,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.0680668016194331e-05,
+      "loss": 2.8245,
+      "step": 1407
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.068825910931174e-05,
+      "loss": 2.8507,
+      "step": 1408
+    },
+    {
+      "epoch": 0.3565283400809717,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.069585020242915e-05,
+      "loss": 2.8143,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3567813765182186,
+      "grad_norm": 3.75,
+      "learning_rate": 1.0703441295546558e-05,
+      "loss": 2.9167,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3567813765182186,
+      "eval_loss": 2.9286928176879883,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.0041,
+      "eval_samples_per_second": 9.665,
+      "eval_steps_per_second": 9.665,
+      "step": 1410
+    },
+    {
+      "epoch": 0.35703441295546556,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.0711032388663967e-05,
+      "loss": 2.8636,
+      "step": 1411
+    },
+    {
+      "epoch": 0.35728744939271256,
+      "grad_norm": 3.5,
+      "learning_rate": 1.0718623481781377e-05,
+      "loss": 2.8143,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3575404858299595,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.0726214574898785e-05,
+      "loss": 2.7405,
+      "step": 1413
+    },
+    {
+      "epoch": 0.35779352226720645,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.0733805668016194e-05,
+      "loss": 2.8193,
+      "step": 1414
+    },
+    {
+      "epoch": 0.35804655870445345,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.0741396761133604e-05,
+      "loss": 2.8564,
+      "step": 1415
+    },
+    {
+      "epoch": 0.3582995951417004,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.0748987854251012e-05,
+      "loss": 2.8234,
+      "step": 1416
+    },
+    {
+      "epoch": 0.35855263157894735,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.075657894736842e-05,
+      "loss": 2.8796,
+      "step": 1417
+    },
+    {
+      "epoch": 0.35880566801619435,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0764170040485831e-05,
+      "loss": 2.7677,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3590587044534413,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.077176113360324e-05,
+      "loss": 2.8517,
+      "step": 1419
+    },
+    {
+      "epoch": 0.35931174089068824,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0779352226720648e-05,
+      "loss": 2.7473,
+      "step": 1420
+    },
+    {
+      "epoch": 0.35956477732793524,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.0786943319838058e-05,
+      "loss": 2.8827,
+      "step": 1421
+    },
+    {
+      "epoch": 0.3598178137651822,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.0794534412955466e-05,
+      "loss": 2.9077,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3600708502024291,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0802125506072875e-05,
+      "loss": 2.8406,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.0809716599190285e-05,
+      "loss": 2.8587,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3605769230769231,
+      "grad_norm": 3.5,
+      "learning_rate": 1.0817307692307692e-05,
+      "loss": 2.7601,
+      "step": 1425
+    },
+    {
+      "epoch": 0.36082995951417,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.08248987854251e-05,
+      "loss": 2.8667,
+      "step": 1426
+    },
+    {
+      "epoch": 0.361082995951417,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.083248987854251e-05,
+      "loss": 2.8781,
+      "step": 1427
+    },
+    {
+      "epoch": 0.36133603238866396,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.0840080971659919e-05,
+      "loss": 2.9437,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3615890688259109,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.0847672064777327e-05,
+      "loss": 2.8342,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3618421052631579,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.0855263157894737e-05,
+      "loss": 2.8404,
+      "step": 1430
+    },
+    {
+      "epoch": 0.36209514170040485,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.0862854251012146e-05,
+      "loss": 2.922,
+      "step": 1431
+    },
+    {
+      "epoch": 0.3623481781376518,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.0870445344129554e-05,
+      "loss": 2.8158,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3626012145748988,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.0878036437246964e-05,
+      "loss": 2.6841,
+      "step": 1433
+    },
+    {
+      "epoch": 0.36285425101214575,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.0885627530364373e-05,
+      "loss": 2.9586,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3631072874493927,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0893218623481781e-05,
+      "loss": 2.8144,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3633603238866397,
+      "grad_norm": 3.75,
+      "learning_rate": 1.0900809716599191e-05,
+      "loss": 2.7828,
+      "step": 1436
+    },
+    {
+      "epoch": 0.36361336032388664,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.09084008097166e-05,
+      "loss": 2.7052,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3638663967611336,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.0915991902834008e-05,
+      "loss": 2.6739,
+      "step": 1438
+    },
+    {
+      "epoch": 0.3641194331983806,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.0923582995951418e-05,
+      "loss": 2.9572,
+      "step": 1439
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.0931174089068826e-05,
+      "loss": 2.8852,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "eval_loss": 2.926607847213745,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 271.2634,
+      "eval_samples_per_second": 9.585,
+      "eval_steps_per_second": 9.585,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3646255060728745,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.0938765182186235e-05,
+      "loss": 2.8541,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3648785425101215,
+      "grad_norm": 3.5,
+      "learning_rate": 1.0946356275303645e-05,
+      "loss": 2.8645,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3651315789473684,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.0953947368421053e-05,
+      "loss": 2.6961,
+      "step": 1443
+    },
+    {
+      "epoch": 0.36538461538461536,
+      "grad_norm": 3.75,
+      "learning_rate": 1.0961538461538462e-05,
+      "loss": 2.8807,
+      "step": 1444
+    },
+    {
+      "epoch": 0.36563765182186236,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.0969129554655872e-05,
+      "loss": 2.8228,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3658906882591093,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.0976720647773279e-05,
+      "loss": 2.876,
+      "step": 1446
+    },
+    {
+      "epoch": 0.36614372469635625,
+      "grad_norm": 3.625,
+      "learning_rate": 1.0984311740890687e-05,
+      "loss": 2.9007,
+      "step": 1447
+    },
+    {
+      "epoch": 0.36639676113360325,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.0991902834008097e-05,
+      "loss": 2.8447,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3666497975708502,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.0999493927125506e-05,
+      "loss": 2.8818,
+      "step": 1449
+    },
+    {
+      "epoch": 0.36690283400809715,
+      "grad_norm": 3.875,
+      "learning_rate": 1.1007085020242914e-05,
+      "loss": 2.9338,
+      "step": 1450
+    },
+    {
+      "epoch": 0.36715587044534415,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.1014676113360324e-05,
+      "loss": 2.767,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3674089068825911,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.1022267206477733e-05,
+      "loss": 2.5305,
+      "step": 1452
+    },
+    {
+      "epoch": 0.36766194331983804,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.1029858299595141e-05,
+      "loss": 2.8967,
+      "step": 1453
+    },
+    {
+      "epoch": 0.36791497975708504,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.1037449392712551e-05,
+      "loss": 2.671,
+      "step": 1454
+    },
+    {
+      "epoch": 0.368168016194332,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.104504048582996e-05,
+      "loss": 2.8511,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.1052631578947368e-05,
+      "loss": 2.9191,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3686740890688259,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.1060222672064778e-05,
+      "loss": 2.8002,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3689271255060729,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.1067813765182187e-05,
+      "loss": 2.764,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3691801619433198,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.1075404858299595e-05,
+      "loss": 2.8574,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3694331983805668,
+      "grad_norm": 3.625,
+      "learning_rate": 1.1082995951417005e-05,
+      "loss": 2.8377,
+      "step": 1460
+    },
+    {
+      "epoch": 0.36968623481781376,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.1090587044534414e-05,
+      "loss": 2.8743,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3699392712550607,
+      "grad_norm": 3.75,
+      "learning_rate": 1.1098178137651822e-05,
+      "loss": 2.8348,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3701923076923077,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.1105769230769232e-05,
+      "loss": 2.9052,
+      "step": 1463
+    },
+    {
+      "epoch": 0.37044534412955465,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.111336032388664e-05,
+      "loss": 2.8297,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3706983805668016,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.1120951417004049e-05,
+      "loss": 2.7709,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3709514170040486,
+      "grad_norm": 4.0,
+      "learning_rate": 1.1128542510121459e-05,
+      "loss": 2.8865,
+      "step": 1466
+    },
+    {
+      "epoch": 0.37120445344129555,
+      "grad_norm": 3.875,
+      "learning_rate": 1.1136133603238866e-05,
+      "loss": 2.8291,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3714574898785425,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.1143724696356274e-05,
+      "loss": 2.7908,
+      "step": 1468
+    },
+    {
+      "epoch": 0.3717105263157895,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.1151315789473684e-05,
+      "loss": 2.8248,
+      "step": 1469
+    },
+    {
+      "epoch": 0.37196356275303644,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.1158906882591093e-05,
+      "loss": 2.804,
+      "step": 1470
+    },
+    {
+      "epoch": 0.37196356275303644,
+      "eval_loss": 2.92094349861145,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.2273,
+      "eval_samples_per_second": 9.622,
+      "eval_steps_per_second": 9.622,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3722165991902834,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.1166497975708501e-05,
+      "loss": 2.9034,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 3.5,
+      "learning_rate": 1.1174089068825911e-05,
+      "loss": 2.6808,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3727226720647773,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.118168016194332e-05,
+      "loss": 2.9686,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3729757085020243,
+      "grad_norm": 3.375,
+      "learning_rate": 1.1189271255060728e-05,
+      "loss": 2.7669,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3732287449392713,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.1196862348178138e-05,
+      "loss": 2.817,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3734817813765182,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.1204453441295547e-05,
+      "loss": 2.805,
+      "step": 1476
+    },
+    {
+      "epoch": 0.37373481781376516,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.1212044534412955e-05,
+      "loss": 2.7742,
+      "step": 1477
+    },
+    {
+      "epoch": 0.37398785425101216,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.1219635627530365e-05,
+      "loss": 2.8538,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3742408906882591,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.1227226720647774e-05,
+      "loss": 2.8451,
+      "step": 1479
+    },
+    {
+      "epoch": 0.37449392712550605,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.1234817813765182e-05,
+      "loss": 2.8786,
+      "step": 1480
+    },
+    {
+      "epoch": 0.37474696356275305,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.1242408906882592e-05,
+      "loss": 2.8712,
+      "step": 1481
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 3.375,
+      "learning_rate": 1.125e-05,
+      "loss": 2.6997,
+      "step": 1482
+    },
+    {
+      "epoch": 0.37525303643724695,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.1257591093117409e-05,
+      "loss": 2.9534,
+      "step": 1483
+    },
+    {
+      "epoch": 0.37550607287449395,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.126518218623482e-05,
+      "loss": 2.8909,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3757591093117409,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.1272773279352228e-05,
+      "loss": 2.9082,
+      "step": 1485
+    },
+    {
+      "epoch": 0.37601214574898784,
+      "grad_norm": 3.5,
+      "learning_rate": 1.1280364372469636e-05,
+      "loss": 2.7357,
+      "step": 1486
+    },
+    {
+      "epoch": 0.37626518218623484,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.1287955465587046e-05,
+      "loss": 2.8565,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.1295546558704453e-05,
+      "loss": 2.9368,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3767712550607287,
+      "grad_norm": 3.75,
+      "learning_rate": 1.1303137651821861e-05,
+      "loss": 2.9694,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3770242914979757,
+      "grad_norm": 63.75,
+      "learning_rate": 1.1310728744939272e-05,
+      "loss": 3.0654,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3772773279352227,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.131831983805668e-05,
+      "loss": 2.7856,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3775303643724696,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.1325910931174088e-05,
+      "loss": 2.8992,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3777834008097166,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.1333502024291498e-05,
+      "loss": 2.8053,
+      "step": 1493
+    },
+    {
+      "epoch": 0.37803643724696356,
+      "grad_norm": 3.375,
+      "learning_rate": 1.1341093117408907e-05,
+      "loss": 2.6646,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3782894736842105,
+      "grad_norm": 3.75,
+      "learning_rate": 1.1348684210526315e-05,
+      "loss": 2.9447,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3785425101214575,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.1356275303643725e-05,
+      "loss": 2.8585,
+      "step": 1496
+    },
+    {
+      "epoch": 0.37879554655870445,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.1363866396761134e-05,
+      "loss": 2.8932,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3790485829959514,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.1371457489878542e-05,
+      "loss": 2.8241,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3793016194331984,
+      "grad_norm": 3.625,
+      "learning_rate": 1.1379048582995952e-05,
+      "loss": 2.8251,
+      "step": 1499
+    },
+    {
+      "epoch": 0.37955465587044535,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.1386639676113361e-05,
+      "loss": 2.7365,
+      "step": 1500
+    },
+    {
+      "epoch": 0.37955465587044535,
+      "eval_loss": 2.9152729511260986,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.2125,
+      "eval_samples_per_second": 9.658,
+      "eval_steps_per_second": 9.658,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3798076923076923,
+      "grad_norm": 18.625,
+      "learning_rate": 1.139423076923077e-05,
+      "loss": 2.8237,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3800607287449393,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.140182186234818e-05,
+      "loss": 2.9073,
+      "step": 1502
+    },
+    {
+      "epoch": 0.38031376518218624,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.1409412955465588e-05,
+      "loss": 2.9235,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.1417004048582996e-05,
+      "loss": 2.6491,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3808198380566802,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.1424595141700406e-05,
+      "loss": 2.7553,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3810728744939271,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.1432186234817815e-05,
+      "loss": 2.792,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3813259109311741,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.1439777327935223e-05,
+      "loss": 2.9344,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3815789473684211,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.1447368421052632e-05,
+      "loss": 2.8372,
+      "step": 1508
+    },
+    {
+      "epoch": 0.381831983805668,
+      "grad_norm": 3.5,
+      "learning_rate": 1.145495951417004e-05,
+      "loss": 2.7377,
+      "step": 1509
+    },
+    {
+      "epoch": 0.38208502024291496,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.1462550607287448e-05,
+      "loss": 2.8707,
+      "step": 1510
+    },
+    {
+      "epoch": 0.38233805668016196,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.1470141700404859e-05,
+      "loss": 2.9014,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3825910931174089,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.1477732793522267e-05,
+      "loss": 2.8576,
+      "step": 1512
+    },
+    {
+      "epoch": 0.38284412955465585,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.1485323886639675e-05,
+      "loss": 2.8876,
+      "step": 1513
+    },
+    {
+      "epoch": 0.38309716599190285,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.1492914979757086e-05,
+      "loss": 2.9444,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3833502024291498,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.1500506072874494e-05,
+      "loss": 2.7722,
+      "step": 1515
+    },
+    {
+      "epoch": 0.38360323886639675,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.1508097165991902e-05,
+      "loss": 2.9122,
+      "step": 1516
+    },
+    {
+      "epoch": 0.38385627530364375,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.1515688259109313e-05,
+      "loss": 2.9499,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3841093117408907,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.1523279352226721e-05,
+      "loss": 2.9018,
+      "step": 1518
+    },
+    {
+      "epoch": 0.38436234817813764,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.153087044534413e-05,
+      "loss": 2.7236,
+      "step": 1519
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.153846153846154e-05,
+      "loss": 2.8688,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3848684210526316,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.1546052631578948e-05,
+      "loss": 2.8422,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3851214574898785,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.1553643724696356e-05,
+      "loss": 2.7436,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3853744939271255,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.1561234817813767e-05,
+      "loss": 2.8171,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3856275303643725,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.1568825910931175e-05,
+      "loss": 2.7756,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3858805668016194,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.1576417004048583e-05,
+      "loss": 2.789,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3861336032388664,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.1584008097165993e-05,
+      "loss": 2.8218,
+      "step": 1526
+    },
+    {
+      "epoch": 0.38638663967611336,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.1591599190283402e-05,
+      "loss": 2.7186,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3866396761133603,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.159919028340081e-05,
+      "loss": 2.8542,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3868927125506073,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.1606781376518219e-05,
+      "loss": 2.8415,
+      "step": 1529
+    },
+    {
+      "epoch": 0.38714574898785425,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.1614372469635627e-05,
+      "loss": 2.825,
+      "step": 1530
+    },
+    {
+      "epoch": 0.38714574898785425,
+      "eval_loss": 2.90696382522583,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 278.9887,
+      "eval_samples_per_second": 9.319,
+      "eval_steps_per_second": 9.319,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3873987854251012,
+      "grad_norm": 3.953125,
+      "learning_rate": 1.1621963562753036e-05,
+      "loss": 2.9127,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3876518218623482,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.1629554655870446e-05,
+      "loss": 2.8184,
+      "step": 1532
+    },
+    {
+      "epoch": 0.38790485829959515,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.1637145748987854e-05,
+      "loss": 2.8663,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3881578947368421,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.1644736842105263e-05,
+      "loss": 2.729,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3884109311740891,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.1652327935222673e-05,
+      "loss": 2.8114,
+      "step": 1535
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.1659919028340081e-05,
+      "loss": 2.8122,
+      "step": 1536
+    },
+    {
+      "epoch": 0.388917004048583,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.166751012145749e-05,
+      "loss": 2.7703,
+      "step": 1537
+    },
+    {
+      "epoch": 0.38917004048583,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.16751012145749e-05,
+      "loss": 2.7739,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3894230769230769,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.1682692307692308e-05,
+      "loss": 2.8136,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3896761133603239,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.1690283400809717e-05,
+      "loss": 2.8355,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3899291497975709,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.1697874493927127e-05,
+      "loss": 2.7877,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3901821862348178,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.1705465587044535e-05,
+      "loss": 2.8882,
+      "step": 1542
+    },
+    {
+      "epoch": 0.39043522267206476,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.1713056680161944e-05,
+      "loss": 2.7209,
+      "step": 1543
+    },
+    {
+      "epoch": 0.39068825910931176,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.1720647773279354e-05,
+      "loss": 2.8433,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3909412955465587,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.1728238866396762e-05,
+      "loss": 2.8866,
+      "step": 1545
+    },
+    {
+      "epoch": 0.39119433198380565,
+      "grad_norm": 3.625,
+      "learning_rate": 1.173582995951417e-05,
+      "loss": 2.8449,
+      "step": 1546
+    },
+    {
+      "epoch": 0.39144736842105265,
+      "grad_norm": 3.625,
+      "learning_rate": 1.174342105263158e-05,
+      "loss": 2.8344,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3917004048582996,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.1751012145748989e-05,
+      "loss": 2.813,
+      "step": 1548
+    },
+    {
+      "epoch": 0.39195344129554655,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.1758603238866397e-05,
+      "loss": 2.9144,
+      "step": 1549
+    },
+    {
+      "epoch": 0.39220647773279355,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.1766194331983806e-05,
+      "loss": 2.8645,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3924595141700405,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.1773785425101214e-05,
+      "loss": 2.8244,
+      "step": 1551
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.1781376518218623e-05,
+      "loss": 2.8698,
+      "step": 1552
+    },
+    {
+      "epoch": 0.39296558704453444,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.1788967611336033e-05,
+      "loss": 2.7727,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3932186234817814,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.1796558704453441e-05,
+      "loss": 2.7804,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3934716599190283,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.180414979757085e-05,
+      "loss": 2.7495,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3937246963562753,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.181174089068826e-05,
+      "loss": 2.8342,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3939777327935223,
+      "grad_norm": 3.75,
+      "learning_rate": 1.1819331983805668e-05,
+      "loss": 2.871,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3942307692307692,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.1826923076923077e-05,
+      "loss": 2.9023,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3944838056680162,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.1834514170040487e-05,
+      "loss": 2.7016,
+      "step": 1559
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.1842105263157895e-05,
+      "loss": 2.741,
+      "step": 1560
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "eval_loss": 2.9025158882141113,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.4919,
+      "eval_samples_per_second": 9.684,
+      "eval_steps_per_second": 9.684,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3949898785425101,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.1849696356275304e-05,
+      "loss": 2.6743,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3952429149797571,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.1857287449392714e-05,
+      "loss": 2.7851,
+      "step": 1562
+    },
+    {
+      "epoch": 0.39549595141700405,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.1864878542510122e-05,
+      "loss": 2.8429,
+      "step": 1563
+    },
+    {
+      "epoch": 0.395748987854251,
+      "grad_norm": 3.375,
+      "learning_rate": 1.187246963562753e-05,
+      "loss": 2.7432,
+      "step": 1564
+    },
+    {
+      "epoch": 0.396002024291498,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.188006072874494e-05,
+      "loss": 2.8312,
+      "step": 1565
+    },
+    {
+      "epoch": 0.39625506072874495,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.188765182186235e-05,
+      "loss": 2.7963,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3965080971659919,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.1895242914979758e-05,
+      "loss": 2.7492,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.1902834008097168e-05,
+      "loss": 2.8769,
+      "step": 1568
+    },
+    {
+      "epoch": 0.39701417004048584,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.1910425101214576e-05,
+      "loss": 2.7517,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3972672064777328,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.1918016194331985e-05,
+      "loss": 2.8682,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3975202429149798,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.1925607287449393e-05,
+      "loss": 2.7705,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3977732793522267,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.1933198380566801e-05,
+      "loss": 2.7716,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3980263157894737,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.194078947368421e-05,
+      "loss": 2.9163,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3982793522267207,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.194838056680162e-05,
+      "loss": 2.7644,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3985323886639676,
+      "grad_norm": 3.875,
+      "learning_rate": 1.1955971659919028e-05,
+      "loss": 2.8489,
+      "step": 1575
+    },
+    {
+      "epoch": 0.39878542510121456,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.1963562753036437e-05,
+      "loss": 2.8999,
+      "step": 1576
+    },
+    {
+      "epoch": 0.39903846153846156,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.1971153846153847e-05,
+      "loss": 2.735,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3992914979757085,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.1978744939271255e-05,
+      "loss": 2.7831,
+      "step": 1578
+    },
+    {
+      "epoch": 0.39954453441295545,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.1986336032388664e-05,
+      "loss": 2.8476,
+      "step": 1579
+    },
+    {
+      "epoch": 0.39979757085020245,
+      "grad_norm": 4.125,
+      "learning_rate": 1.1993927125506074e-05,
+      "loss": 2.8794,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4000506072874494,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.2001518218623482e-05,
+      "loss": 2.8348,
+      "step": 1581
+    },
+    {
+      "epoch": 0.40030364372469635,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.200910931174089e-05,
+      "loss": 2.7297,
+      "step": 1582
+    },
+    {
+      "epoch": 0.40055668016194335,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.2016700404858301e-05,
+      "loss": 2.9067,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.202429149797571e-05,
+      "loss": 2.9049,
+      "step": 1584
+    },
+    {
+      "epoch": 0.40106275303643724,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.2031882591093118e-05,
+      "loss": 2.8656,
+      "step": 1585
+    },
+    {
+      "epoch": 0.40131578947368424,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.2039473684210528e-05,
+      "loss": 2.7989,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4015688259109312,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.2047064777327936e-05,
+      "loss": 2.8132,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4018218623481781,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.2054655870445345e-05,
+      "loss": 2.7121,
+      "step": 1588
+    },
+    {
+      "epoch": 0.4020748987854251,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.2062246963562755e-05,
+      "loss": 2.6906,
+      "step": 1589
+    },
+    {
+      "epoch": 0.4023279352226721,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.2069838056680163e-05,
+      "loss": 2.5701,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4023279352226721,
+      "eval_loss": 2.900104522705078,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.6712,
+      "eval_samples_per_second": 9.641,
+      "eval_steps_per_second": 9.641,
+      "step": 1590
+    },
+    {
+      "epoch": 0.402580971659919,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.2077429149797572e-05,
+      "loss": 2.905,
+      "step": 1591
+    },
+    {
+      "epoch": 0.402834008097166,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.208502024291498e-05,
+      "loss": 2.7971,
+      "step": 1592
+    },
+    {
+      "epoch": 0.40308704453441296,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.2092611336032389e-05,
+      "loss": 2.8316,
+      "step": 1593
+    },
+    {
+      "epoch": 0.4033400809716599,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.2100202429149797e-05,
+      "loss": 2.8387,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4035931174089069,
+      "grad_norm": 3.375,
+      "learning_rate": 1.2107793522267207e-05,
+      "loss": 2.7883,
+      "step": 1595
+    },
+    {
+      "epoch": 0.40384615384615385,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.2115384615384615e-05,
+      "loss": 2.7888,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4040991902834008,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.2122975708502024e-05,
+      "loss": 2.8423,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4043522267206478,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.2130566801619434e-05,
+      "loss": 2.8819,
+      "step": 1598
+    },
+    {
+      "epoch": 0.40460526315789475,
+      "grad_norm": 3.625,
+      "learning_rate": 1.2138157894736842e-05,
+      "loss": 2.8597,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 3.5,
+      "learning_rate": 1.2145748987854251e-05,
+      "loss": 2.7793,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4051113360323887,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.2153340080971661e-05,
+      "loss": 2.7847,
+      "step": 1601
+    },
+    {
+      "epoch": 0.40536437246963564,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.216093117408907e-05,
+      "loss": 2.9012,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4056174089068826,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.2168522267206478e-05,
+      "loss": 2.7534,
+      "step": 1603
+    },
+    {
+      "epoch": 0.4058704453441296,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.2176113360323888e-05,
+      "loss": 2.7773,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4061234817813765,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.2183704453441296e-05,
+      "loss": 2.8639,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4063765182186235,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.2191295546558705e-05,
+      "loss": 2.8615,
+      "step": 1606
+    },
+    {
+      "epoch": 0.4066295546558704,
+      "grad_norm": 3.375,
+      "learning_rate": 1.2198886639676113e-05,
+      "loss": 2.8151,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4068825910931174,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.2206477732793523e-05,
+      "loss": 2.8742,
+      "step": 1608
+    },
+    {
+      "epoch": 0.40713562753036436,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.2214068825910932e-05,
+      "loss": 2.8922,
+      "step": 1609
+    },
+    {
+      "epoch": 0.4073886639676113,
+      "grad_norm": 3.125,
+      "learning_rate": 1.222165991902834e-05,
+      "loss": 2.638,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4076417004048583,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.222925101214575e-05,
+      "loss": 2.9211,
+      "step": 1611
+    },
+    {
+      "epoch": 0.40789473684210525,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.2236842105263159e-05,
+      "loss": 2.8352,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4081477732793522,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.2244433198380566e-05,
+      "loss": 2.8167,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4084008097165992,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.2252024291497976e-05,
+      "loss": 2.6673,
+      "step": 1614
+    },
+    {
+      "epoch": 0.40865384615384615,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.2259615384615384e-05,
+      "loss": 2.8453,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.2267206477732792e-05,
+      "loss": 2.8345,
+      "step": 1616
+    },
+    {
+      "epoch": 0.4091599190283401,
+      "grad_norm": 3.75,
+      "learning_rate": 1.2274797570850203e-05,
+      "loss": 2.8189,
+      "step": 1617
+    },
+    {
+      "epoch": 0.40941295546558704,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.2282388663967611e-05,
+      "loss": 2.8883,
+      "step": 1618
+    },
+    {
+      "epoch": 0.409665991902834,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.228997975708502e-05,
+      "loss": 2.7965,
+      "step": 1619
+    },
+    {
+      "epoch": 0.409919028340081,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.229757085020243e-05,
+      "loss": 2.8352,
+      "step": 1620
+    },
+    {
+      "epoch": 0.409919028340081,
+      "eval_loss": 2.8934783935546875,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.4479,
+      "eval_samples_per_second": 9.649,
+      "eval_steps_per_second": 9.649,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4101720647773279,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.2305161943319838e-05,
+      "loss": 2.677,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4104251012145749,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.2312753036437246e-05,
+      "loss": 2.8939,
+      "step": 1622
+    },
+    {
+      "epoch": 0.4106781376518219,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.2320344129554657e-05,
+      "loss": 2.7119,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4109311740890688,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.2327935222672065e-05,
+      "loss": 2.7852,
+      "step": 1624
+    },
+    {
+      "epoch": 0.41118421052631576,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.2335526315789473e-05,
+      "loss": 2.8578,
+      "step": 1625
+    },
+    {
+      "epoch": 0.41143724696356276,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.2343117408906884e-05,
+      "loss": 2.8729,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4116902834008097,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.2350708502024292e-05,
+      "loss": 2.835,
+      "step": 1627
+    },
+    {
+      "epoch": 0.41194331983805665,
+      "grad_norm": 3.5,
+      "learning_rate": 1.23582995951417e-05,
+      "loss": 2.8308,
+      "step": 1628
+    },
+    {
+      "epoch": 0.41219635627530365,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.236589068825911e-05,
+      "loss": 2.7723,
+      "step": 1629
+    },
+    {
+      "epoch": 0.4124493927125506,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.2373481781376519e-05,
+      "loss": 2.774,
+      "step": 1630
+    },
+    {
+      "epoch": 0.41270242914979755,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.2381072874493927e-05,
+      "loss": 2.7447,
+      "step": 1631
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.2388663967611337e-05,
+      "loss": 2.7746,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4132085020242915,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.2396255060728744e-05,
+      "loss": 2.8931,
+      "step": 1633
+    },
+    {
+      "epoch": 0.41346153846153844,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.2403846153846153e-05,
+      "loss": 2.6253,
+      "step": 1634
+    },
+    {
+      "epoch": 0.41371457489878544,
+      "grad_norm": 3.5,
+      "learning_rate": 1.2411437246963563e-05,
+      "loss": 2.8367,
+      "step": 1635
+    },
+    {
+      "epoch": 0.4139676113360324,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.2419028340080971e-05,
+      "loss": 2.7907,
+      "step": 1636
+    },
+    {
+      "epoch": 0.4142206477732793,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.242661943319838e-05,
+      "loss": 2.8687,
+      "step": 1637
+    },
+    {
+      "epoch": 0.4144736842105263,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.243421052631579e-05,
+      "loss": 2.8176,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4147267206477733,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.2441801619433198e-05,
+      "loss": 2.7583,
+      "step": 1639
+    },
+    {
+      "epoch": 0.4149797570850202,
+      "grad_norm": 33.5,
+      "learning_rate": 1.2449392712550607e-05,
+      "loss": 3.0522,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4152327935222672,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.2456983805668017e-05,
+      "loss": 2.9121,
+      "step": 1641
+    },
+    {
+      "epoch": 0.41548582995951416,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.2464574898785425e-05,
+      "loss": 2.8386,
+      "step": 1642
+    },
+    {
+      "epoch": 0.4157388663967611,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.2472165991902834e-05,
+      "loss": 2.871,
+      "step": 1643
+    },
+    {
+      "epoch": 0.4159919028340081,
+      "grad_norm": 3.5,
+      "learning_rate": 1.2479757085020244e-05,
+      "loss": 2.8777,
+      "step": 1644
+    },
+    {
+      "epoch": 0.41624493927125505,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.2487348178137652e-05,
+      "loss": 2.7821,
+      "step": 1645
+    },
+    {
+      "epoch": 0.416497975708502,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.249493927125506e-05,
+      "loss": 2.8033,
+      "step": 1646
+    },
+    {
+      "epoch": 0.416751012145749,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.250253036437247e-05,
+      "loss": 2.7877,
+      "step": 1647
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.2510121457489879e-05,
+      "loss": 2.9266,
+      "step": 1648
+    },
+    {
+      "epoch": 0.4172570850202429,
+      "grad_norm": 96.0,
+      "learning_rate": 1.2517712550607287e-05,
+      "loss": 2.965,
+      "step": 1649
+    },
+    {
+      "epoch": 0.4175101214574899,
+      "grad_norm": 3.625,
+      "learning_rate": 1.2525303643724698e-05,
+      "loss": 2.8189,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4175101214574899,
+      "eval_loss": 2.8919875621795654,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 268.8293,
+      "eval_samples_per_second": 9.672,
+      "eval_steps_per_second": 9.672,
+      "step": 1650
+    },
+    {
+      "epoch": 0.41776315789473684,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.2532894736842106e-05,
+      "loss": 2.8668,
+      "step": 1651
+    },
+    {
+      "epoch": 0.4180161943319838,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.2540485829959514e-05,
+      "loss": 2.6681,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4182692307692308,
+      "grad_norm": 4.0,
+      "learning_rate": 1.2548076923076925e-05,
+      "loss": 2.7907,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4185222672064777,
+      "grad_norm": 3.5,
+      "learning_rate": 1.2555668016194331e-05,
+      "loss": 2.848,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4187753036437247,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.256325910931174e-05,
+      "loss": 2.8012,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4190283400809717,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.257085020242915e-05,
+      "loss": 2.8126,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4192813765182186,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.2578441295546558e-05,
+      "loss": 2.8214,
+      "step": 1657
+    },
+    {
+      "epoch": 0.41953441295546556,
+      "grad_norm": 3.25,
+      "learning_rate": 1.2586032388663967e-05,
+      "loss": 2.7707,
+      "step": 1658
+    },
+    {
+      "epoch": 0.41978744939271256,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.2593623481781377e-05,
+      "loss": 2.7897,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4200404858299595,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.2601214574898785e-05,
+      "loss": 2.7209,
+      "step": 1660
+    },
+    {
+      "epoch": 0.42029352226720645,
+      "grad_norm": 3.625,
+      "learning_rate": 1.2608805668016194e-05,
+      "loss": 2.8552,
+      "step": 1661
+    },
+    {
+      "epoch": 0.42054655870445345,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.2616396761133604e-05,
+      "loss": 2.8378,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4207995951417004,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.2623987854251012e-05,
+      "loss": 2.9254,
+      "step": 1663
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.263157894736842e-05,
+      "loss": 2.8671,
+      "step": 1664
+    },
+    {
+      "epoch": 0.42130566801619435,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.263917004048583e-05,
+      "loss": 2.8284,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4215587044534413,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.264676113360324e-05,
+      "loss": 2.7532,
+      "step": 1666
+    },
+    {
+      "epoch": 0.42181174089068824,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.2654352226720648e-05,
+      "loss": 2.8398,
+      "step": 1667
+    },
+    {
+      "epoch": 0.42206477732793524,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.2661943319838058e-05,
+      "loss": 2.8384,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4223178137651822,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.2669534412955466e-05,
+      "loss": 2.7879,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4225708502024291,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.2677125506072875e-05,
+      "loss": 2.8829,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4228238866396761,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.2684716599190285e-05,
+      "loss": 2.7978,
+      "step": 1671
+    },
+    {
+      "epoch": 0.4230769230769231,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.2692307692307693e-05,
+      "loss": 2.6966,
+      "step": 1672
+    },
+    {
+      "epoch": 0.42332995951417,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.2699898785425102e-05,
+      "loss": 2.7141,
+      "step": 1673
+    },
+    {
+      "epoch": 0.423582995951417,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.2707489878542512e-05,
+      "loss": 2.8197,
+      "step": 1674
+    },
+    {
+      "epoch": 0.42383603238866396,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.2715080971659918e-05,
+      "loss": 2.8335,
+      "step": 1675
+    },
+    {
+      "epoch": 0.4240890688259109,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.2722672064777327e-05,
+      "loss": 2.7349,
+      "step": 1676
+    },
+    {
+      "epoch": 0.4243421052631579,
+      "grad_norm": 3.921875,
+      "learning_rate": 1.2730263157894737e-05,
+      "loss": 2.7991,
+      "step": 1677
+    },
+    {
+      "epoch": 0.42459514170040485,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.2737854251012145e-05,
+      "loss": 2.7348,
+      "step": 1678
+    },
+    {
+      "epoch": 0.4248481781376518,
+      "grad_norm": 3.5,
+      "learning_rate": 1.2745445344129554e-05,
+      "loss": 2.77,
+      "step": 1679
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.2753036437246964e-05,
+      "loss": 2.8516,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "eval_loss": 2.8866162300109863,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.8187,
+      "eval_samples_per_second": 9.636,
+      "eval_steps_per_second": 9.636,
+      "step": 1680
+    },
+    {
+      "epoch": 0.42535425101214575,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.2760627530364372e-05,
+      "loss": 2.862,
+      "step": 1681
+    },
+    {
+      "epoch": 0.4256072874493927,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.276821862348178e-05,
+      "loss": 2.9164,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4258603238866397,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.2775809716599191e-05,
+      "loss": 2.739,
+      "step": 1683
+    },
+    {
+      "epoch": 0.42611336032388664,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.27834008097166e-05,
+      "loss": 2.8434,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4263663967611336,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.2790991902834008e-05,
+      "loss": 2.7795,
+      "step": 1685
+    },
+    {
+      "epoch": 0.4266194331983806,
+      "grad_norm": 3.625,
+      "learning_rate": 1.2798582995951418e-05,
+      "loss": 2.6812,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4268724696356275,
+      "grad_norm": 3.625,
+      "learning_rate": 1.2806174089068826e-05,
+      "loss": 2.8003,
+      "step": 1687
+    },
+    {
+      "epoch": 0.4271255060728745,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.2813765182186235e-05,
+      "loss": 2.9023,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4273785425101215,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.2821356275303645e-05,
+      "loss": 2.8196,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4276315789473684,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.2828947368421053e-05,
+      "loss": 2.8494,
+      "step": 1690
+    },
+    {
+      "epoch": 0.42788461538461536,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.2836538461538462e-05,
+      "loss": 2.874,
+      "step": 1691
+    },
+    {
+      "epoch": 0.42813765182186236,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.2844129554655872e-05,
+      "loss": 2.7877,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4283906882591093,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.285172064777328e-05,
+      "loss": 2.8116,
+      "step": 1693
+    },
+    {
+      "epoch": 0.42864372469635625,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.2859311740890689e-05,
+      "loss": 2.7781,
+      "step": 1694
+    },
+    {
+      "epoch": 0.42889676113360325,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.2866902834008099e-05,
+      "loss": 2.7908,
+      "step": 1695
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.2874493927125506e-05,
+      "loss": 2.8076,
+      "step": 1696
+    },
+    {
+      "epoch": 0.42940283400809715,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.2882085020242914e-05,
+      "loss": 2.6652,
+      "step": 1697
+    },
+    {
+      "epoch": 0.42965587044534415,
+      "grad_norm": 3.375,
+      "learning_rate": 1.2889676113360324e-05,
+      "loss": 2.7936,
+      "step": 1698
+    },
+    {
+      "epoch": 0.4299089068825911,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.2897267206477733e-05,
+      "loss": 2.7058,
+      "step": 1699
+    },
+    {
+      "epoch": 0.43016194331983804,
+      "grad_norm": 3.125,
+      "learning_rate": 1.2904858299595141e-05,
+      "loss": 2.6534,
+      "step": 1700
+    },
+    {
+      "epoch": 0.43041497975708504,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.2912449392712551e-05,
+      "loss": 2.8685,
+      "step": 1701
+    },
+    {
+      "epoch": 0.430668016194332,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.292004048582996e-05,
+      "loss": 2.8472,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4309210526315789,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.2927631578947368e-05,
+      "loss": 2.781,
+      "step": 1703
+    },
+    {
+      "epoch": 0.4311740890688259,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.2935222672064778e-05,
+      "loss": 2.7427,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4314271255060729,
+      "grad_norm": 4.625,
+      "learning_rate": 1.2942813765182186e-05,
+      "loss": 2.7475,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4316801619433198,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.2950404858299595e-05,
+      "loss": 2.8945,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4319331983805668,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.2957995951417005e-05,
+      "loss": 2.8314,
+      "step": 1707
+    },
+    {
+      "epoch": 0.43218623481781376,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.2965587044534413e-05,
+      "loss": 2.7303,
+      "step": 1708
+    },
+    {
+      "epoch": 0.4324392712550607,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.2973178137651822e-05,
+      "loss": 2.8612,
+      "step": 1709
+    },
+    {
+      "epoch": 0.4326923076923077,
+      "grad_norm": 3.75,
+      "learning_rate": 1.2980769230769232e-05,
+      "loss": 2.7864,
+      "step": 1710
+    },
+    {
+      "epoch": 0.4326923076923077,
+      "eval_loss": 2.8780622482299805,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.637,
+      "eval_samples_per_second": 9.643,
+      "eval_steps_per_second": 9.643,
+      "step": 1710
+    },
+    {
+      "epoch": 0.43294534412955465,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.298836032388664e-05,
+      "loss": 2.8317,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.2995951417004049e-05,
+      "loss": 2.7129,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4334514170040486,
+      "grad_norm": 3.75,
+      "learning_rate": 1.3003542510121459e-05,
+      "loss": 2.7798,
+      "step": 1713
+    },
+    {
+      "epoch": 0.43370445344129555,
+      "grad_norm": 3.25,
+      "learning_rate": 1.3011133603238867e-05,
+      "loss": 2.6732,
+      "step": 1714
+    },
+    {
+      "epoch": 0.4339574898785425,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.3018724696356276e-05,
+      "loss": 2.8461,
+      "step": 1715
+    },
+    {
+      "epoch": 0.4342105263157895,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.3026315789473684e-05,
+      "loss": 2.8107,
+      "step": 1716
+    },
+    {
+      "epoch": 0.43446356275303644,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.3033906882591093e-05,
+      "loss": 2.7023,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4347165991902834,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.3041497975708501e-05,
+      "loss": 2.7048,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4349696356275304,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.3049089068825911e-05,
+      "loss": 2.6912,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4352226720647773,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.305668016194332e-05,
+      "loss": 2.7664,
+      "step": 1720
+    },
+    {
+      "epoch": 0.4354757085020243,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.3064271255060728e-05,
+      "loss": 2.8273,
+      "step": 1721
+    },
+    {
+      "epoch": 0.4357287449392713,
+      "grad_norm": 3.625,
+      "learning_rate": 1.3071862348178138e-05,
+      "loss": 2.7948,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4359817813765182,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.3079453441295547e-05,
+      "loss": 2.821,
+      "step": 1723
+    },
+    {
+      "epoch": 0.43623481781376516,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.3087044534412955e-05,
+      "loss": 2.7543,
+      "step": 1724
+    },
+    {
+      "epoch": 0.43648785425101216,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.3094635627530365e-05,
+      "loss": 2.725,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4367408906882591,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.3102226720647774e-05,
+      "loss": 2.7909,
+      "step": 1726
+    },
+    {
+      "epoch": 0.43699392712550605,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.3109817813765182e-05,
+      "loss": 2.7978,
+      "step": 1727
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.3117408906882592e-05,
+      "loss": 2.7499,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.3125e-05,
+      "loss": 2.813,
+      "step": 1729
+    },
+    {
+      "epoch": 0.43775303643724695,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.3132591093117409e-05,
+      "loss": 2.6622,
+      "step": 1730
+    },
+    {
+      "epoch": 0.43800607287449395,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.3140182186234819e-05,
+      "loss": 2.8186,
+      "step": 1731
+    },
+    {
+      "epoch": 0.4382591093117409,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.3147773279352228e-05,
+      "loss": 2.6178,
+      "step": 1732
+    },
+    {
+      "epoch": 0.43851214574898784,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.3155364372469636e-05,
+      "loss": 2.8087,
+      "step": 1733
+    },
+    {
+      "epoch": 0.43876518218623484,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.3162955465587046e-05,
+      "loss": 2.8008,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4390182186234818,
+      "grad_norm": 3.375,
+      "learning_rate": 1.3170546558704454e-05,
+      "loss": 2.6989,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4392712550607287,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.3178137651821863e-05,
+      "loss": 2.8431,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4395242914979757,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.3185728744939271e-05,
+      "loss": 2.8202,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4397773279352227,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.319331983805668e-05,
+      "loss": 2.9207,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4400303643724696,
+      "grad_norm": 3.75,
+      "learning_rate": 1.3200910931174088e-05,
+      "loss": 2.7985,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4402834008097166,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.3208502024291498e-05,
+      "loss": 2.8151,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4402834008097166,
+      "eval_loss": 2.8696091175079346,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.6065,
+      "eval_samples_per_second": 9.608,
+      "eval_steps_per_second": 9.608,
+      "step": 1740
+    },
+    {
+      "epoch": 0.44053643724696356,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.3216093117408907e-05,
+      "loss": 2.7438,
+      "step": 1741
+    },
+    {
+      "epoch": 0.4407894736842105,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.3223684210526315e-05,
+      "loss": 2.8484,
+      "step": 1742
+    },
+    {
+      "epoch": 0.4410425101214575,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.3231275303643725e-05,
+      "loss": 2.8843,
+      "step": 1743
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.3238866396761134e-05,
+      "loss": 2.8246,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4415485829959514,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.3246457489878542e-05,
+      "loss": 2.7782,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4418016194331984,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.3254048582995952e-05,
+      "loss": 2.7651,
+      "step": 1746
+    },
+    {
+      "epoch": 0.44205465587044535,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.326163967611336e-05,
+      "loss": 2.8362,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4423076923076923,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.3269230769230769e-05,
+      "loss": 2.7673,
+      "step": 1748
+    },
+    {
+      "epoch": 0.4425607287449393,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.327682186234818e-05,
+      "loss": 2.7203,
+      "step": 1749
+    },
+    {
+      "epoch": 0.44281376518218624,
+      "grad_norm": 3.625,
+      "learning_rate": 1.3284412955465588e-05,
+      "loss": 2.7911,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4430668016194332,
+      "grad_norm": 3.625,
+      "learning_rate": 1.3292004048582996e-05,
+      "loss": 2.7805,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4433198380566802,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.3299595141700406e-05,
+      "loss": 2.7872,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4435728744939271,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.3307186234817815e-05,
+      "loss": 2.8186,
+      "step": 1753
+    },
+    {
+      "epoch": 0.4438259109311741,
+      "grad_norm": 3.75,
+      "learning_rate": 1.3314777327935223e-05,
+      "loss": 2.6563,
+      "step": 1754
+    },
+    {
+      "epoch": 0.4440789473684211,
+      "grad_norm": 3.625,
+      "learning_rate": 1.3322368421052633e-05,
+      "loss": 2.8683,
+      "step": 1755
+    },
+    {
+      "epoch": 0.444331983805668,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.3329959514170042e-05,
+      "loss": 2.8161,
+      "step": 1756
+    },
+    {
+      "epoch": 0.44458502024291496,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.333755060728745e-05,
+      "loss": 2.8853,
+      "step": 1757
+    },
+    {
+      "epoch": 0.44483805668016196,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.3345141700404858e-05,
+      "loss": 2.8712,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4450910931174089,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.3352732793522267e-05,
+      "loss": 2.7442,
+      "step": 1759
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 3.625,
+      "learning_rate": 1.3360323886639675e-05,
+      "loss": 2.8117,
+      "step": 1760
+    },
+    {
+      "epoch": 0.44559716599190285,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.3367914979757085e-05,
+      "loss": 2.7859,
+      "step": 1761
+    },
+    {
+      "epoch": 0.4458502024291498,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.3375506072874494e-05,
+      "loss": 2.7918,
+      "step": 1762
+    },
+    {
+      "epoch": 0.44610323886639675,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.3383097165991902e-05,
+      "loss": 2.8249,
+      "step": 1763
+    },
+    {
+      "epoch": 0.44635627530364375,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.3390688259109312e-05,
+      "loss": 2.8135,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4466093117408907,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.339827935222672e-05,
+      "loss": 2.834,
+      "step": 1765
+    },
+    {
+      "epoch": 0.44686234817813764,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.340587044534413e-05,
+      "loss": 2.7493,
+      "step": 1766
+    },
+    {
+      "epoch": 0.44711538461538464,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.341346153846154e-05,
+      "loss": 2.7271,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4473684210526316,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.3421052631578948e-05,
+      "loss": 2.8074,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4476214574898785,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.3428643724696356e-05,
+      "loss": 2.7987,
+      "step": 1769
+    },
+    {
+      "epoch": 0.4478744939271255,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.3436234817813766e-05,
+      "loss": 2.842,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4478744939271255,
+      "eval_loss": 2.8738222122192383,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.047,
+      "eval_samples_per_second": 9.664,
+      "eval_steps_per_second": 9.664,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4481275303643725,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.3443825910931175e-05,
+      "loss": 2.9191,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4483805668016194,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.3451417004048583e-05,
+      "loss": 2.7682,
+      "step": 1772
+    },
+    {
+      "epoch": 0.4486336032388664,
+      "grad_norm": 3.25,
+      "learning_rate": 1.3459008097165993e-05,
+      "loss": 2.7883,
+      "step": 1773
+    },
+    {
+      "epoch": 0.44888663967611336,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.3466599190283402e-05,
+      "loss": 2.9011,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4491396761133603,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.347419028340081e-05,
+      "loss": 2.7988,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 4.09375,
+      "learning_rate": 1.348178137651822e-05,
+      "loss": 2.8708,
+      "step": 1776
+    },
+    {
+      "epoch": 0.44964574898785425,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.3489372469635629e-05,
+      "loss": 2.8539,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4498987854251012,
+      "grad_norm": 3.375,
+      "learning_rate": 1.3496963562753037e-05,
+      "loss": 2.7114,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4501518218623482,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.3504554655870446e-05,
+      "loss": 2.6399,
+      "step": 1779
+    },
+    {
+      "epoch": 0.45040485829959515,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.3512145748987854e-05,
+      "loss": 2.7249,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4506578947368421,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.3519736842105262e-05,
+      "loss": 2.8675,
+      "step": 1781
+    },
+    {
+      "epoch": 0.4509109311740891,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.3527327935222673e-05,
+      "loss": 2.8413,
+      "step": 1782
+    },
+    {
+      "epoch": 0.45116396761133604,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.3534919028340081e-05,
+      "loss": 2.7766,
+      "step": 1783
+    },
+    {
+      "epoch": 0.451417004048583,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.354251012145749e-05,
+      "loss": 2.7733,
+      "step": 1784
+    },
+    {
+      "epoch": 0.45167004048583,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.35501012145749e-05,
+      "loss": 2.8002,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4519230769230769,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.3557692307692308e-05,
+      "loss": 2.8285,
+      "step": 1786
+    },
+    {
+      "epoch": 0.4521761133603239,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.3565283400809716e-05,
+      "loss": 2.769,
+      "step": 1787
+    },
+    {
+      "epoch": 0.4524291497975709,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.3572874493927126e-05,
+      "loss": 2.8364,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4526821862348178,
+      "grad_norm": 3.375,
+      "learning_rate": 1.3580465587044535e-05,
+      "loss": 2.8214,
+      "step": 1789
+    },
+    {
+      "epoch": 0.45293522267206476,
+      "grad_norm": 44.25,
+      "learning_rate": 1.3588056680161943e-05,
+      "loss": 2.9123,
+      "step": 1790
+    },
+    {
+      "epoch": 0.45318825910931176,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.3595647773279353e-05,
+      "loss": 2.5778,
+      "step": 1791
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.3603238866396762e-05,
+      "loss": 2.838,
+      "step": 1792
+    },
+    {
+      "epoch": 0.45369433198380565,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.361082995951417e-05,
+      "loss": 2.7671,
+      "step": 1793
+    },
+    {
+      "epoch": 0.45394736842105265,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.361842105263158e-05,
+      "loss": 2.6776,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4542004048582996,
+      "grad_norm": 3.5,
+      "learning_rate": 1.3626012145748989e-05,
+      "loss": 2.8128,
+      "step": 1795
+    },
+    {
+      "epoch": 0.45445344129554655,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.3633603238866397e-05,
+      "loss": 2.8619,
+      "step": 1796
+    },
+    {
+      "epoch": 0.45470647773279355,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.3641194331983807e-05,
+      "loss": 2.7514,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4549595141700405,
+      "grad_norm": 3.25,
+      "learning_rate": 1.3648785425101216e-05,
+      "loss": 2.6166,
+      "step": 1798
+    },
+    {
+      "epoch": 0.45521255060728744,
+      "grad_norm": 3.625,
+      "learning_rate": 1.3656376518218624e-05,
+      "loss": 2.8515,
+      "step": 1799
+    },
+    {
+      "epoch": 0.45546558704453444,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.3663967611336033e-05,
+      "loss": 2.7927,
+      "step": 1800
+    },
+    {
+      "epoch": 0.45546558704453444,
+      "eval_loss": 2.86922025680542,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.8759,
+      "eval_samples_per_second": 9.634,
+      "eval_steps_per_second": 9.634,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4557186234817814,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.3671558704453441e-05,
+      "loss": 2.881,
+      "step": 1801
+    },
+    {
+      "epoch": 0.4559716599190283,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.367914979757085e-05,
+      "loss": 2.8805,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4562246963562753,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.368674089068826e-05,
+      "loss": 2.749,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4564777327935223,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.3694331983805668e-05,
+      "loss": 2.7047,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4567307692307692,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.3701923076923077e-05,
+      "loss": 2.8215,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4569838056680162,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.3709514170040487e-05,
+      "loss": 2.7696,
+      "step": 1806
+    },
+    {
+      "epoch": 0.45723684210526316,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.3717105263157895e-05,
+      "loss": 2.7519,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.3724696356275303e-05,
+      "loss": 2.819,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4577429149797571,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.3732287449392714e-05,
+      "loss": 2.7637,
+      "step": 1809
+    },
+    {
+      "epoch": 0.45799595141700405,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.3739878542510122e-05,
+      "loss": 2.8179,
+      "step": 1810
+    },
+    {
+      "epoch": 0.458248987854251,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.374746963562753e-05,
+      "loss": 2.8034,
+      "step": 1811
+    },
+    {
+      "epoch": 0.458502024291498,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.375506072874494e-05,
+      "loss": 2.7085,
+      "step": 1812
+    },
+    {
+      "epoch": 0.45875506072874495,
+      "grad_norm": 3.875,
+      "learning_rate": 1.3762651821862349e-05,
+      "loss": 2.8351,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4590080971659919,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.3770242914979757e-05,
+      "loss": 2.7643,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4592611336032389,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.3777834008097168e-05,
+      "loss": 2.8117,
+      "step": 1815
+    },
+    {
+      "epoch": 0.45951417004048584,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.3785425101214576e-05,
+      "loss": 2.8394,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4597672064777328,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.3793016194331984e-05,
+      "loss": 2.7334,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4600202429149798,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.3800607287449395e-05,
+      "loss": 2.7619,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4602732793522267,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.3808198380566803e-05,
+      "loss": 2.7128,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4605263157894737,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.3815789473684211e-05,
+      "loss": 2.7561,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4607793522267207,
+      "grad_norm": 3.625,
+      "learning_rate": 1.382338056680162e-05,
+      "loss": 2.7923,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4610323886639676,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.3830971659919028e-05,
+      "loss": 2.8677,
+      "step": 1822
+    },
+    {
+      "epoch": 0.46128542510121456,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.3838562753036437e-05,
+      "loss": 2.6625,
+      "step": 1823
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 2.829,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4617914979757085,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.3853744939271255e-05,
+      "loss": 2.783,
+      "step": 1825
+    },
+    {
+      "epoch": 0.46204453441295545,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.3861336032388664e-05,
+      "loss": 2.7215,
+      "step": 1826
+    },
+    {
+      "epoch": 0.46229757085020245,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.3868927125506074e-05,
+      "loss": 2.7401,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4625506072874494,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.3876518218623482e-05,
+      "loss": 2.787,
+      "step": 1828
+    },
+    {
+      "epoch": 0.46280364372469635,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.388410931174089e-05,
+      "loss": 2.6234,
+      "step": 1829
+    },
+    {
+      "epoch": 0.46305668016194335,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.38917004048583e-05,
+      "loss": 2.8896,
+      "step": 1830
+    },
+    {
+      "epoch": 0.46305668016194335,
+      "eval_loss": 2.8619534969329834,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.5339,
+      "eval_samples_per_second": 9.611,
+      "eval_steps_per_second": 9.611,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4633097165991903,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.3899291497975709e-05,
+      "loss": 2.7498,
+      "step": 1831
+    },
+    {
+      "epoch": 0.46356275303643724,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.3906882591093118e-05,
+      "loss": 2.7884,
+      "step": 1832
+    },
+    {
+      "epoch": 0.46381578947368424,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.3914473684210528e-05,
+      "loss": 2.7501,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4640688259109312,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.3922064777327936e-05,
+      "loss": 2.8667,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4643218623481781,
+      "grad_norm": 4.0625,
+      "learning_rate": 1.3929655870445345e-05,
+      "loss": 2.7711,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4645748987854251,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.3937246963562755e-05,
+      "loss": 2.791,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4648279352226721,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.3944838056680163e-05,
+      "loss": 2.7554,
+      "step": 1837
+    },
+    {
+      "epoch": 0.465080971659919,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.3952429149797572e-05,
+      "loss": 2.6637,
+      "step": 1838
+    },
+    {
+      "epoch": 0.465334008097166,
+      "grad_norm": 3.75,
+      "learning_rate": 1.3960020242914982e-05,
+      "loss": 2.85,
+      "step": 1839
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.396761133603239e-05,
+      "loss": 2.8089,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4658400809716599,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.3975202429149797e-05,
+      "loss": 2.7586,
+      "step": 1841
+    },
+    {
+      "epoch": 0.4660931174089069,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.3982793522267207e-05,
+      "loss": 2.7293,
+      "step": 1842
+    },
+    {
+      "epoch": 0.46634615384615385,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.3990384615384615e-05,
+      "loss": 2.7026,
+      "step": 1843
+    },
+    {
+      "epoch": 0.4665991902834008,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.3997975708502024e-05,
+      "loss": 2.7838,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4668522267206478,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4005566801619434e-05,
+      "loss": 2.8014,
+      "step": 1845
+    },
+    {
+      "epoch": 0.46710526315789475,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4013157894736842e-05,
+      "loss": 2.7835,
+      "step": 1846
+    },
+    {
+      "epoch": 0.4673582995951417,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.402074898785425e-05,
+      "loss": 2.8522,
+      "step": 1847
+    },
+    {
+      "epoch": 0.4676113360323887,
+      "grad_norm": 3.375,
+      "learning_rate": 1.4028340080971661e-05,
+      "loss": 2.741,
+      "step": 1848
+    },
+    {
+      "epoch": 0.46786437246963564,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.403593117408907e-05,
+      "loss": 2.7992,
+      "step": 1849
+    },
+    {
+      "epoch": 0.4681174089068826,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.4043522267206478e-05,
+      "loss": 2.7978,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4683704453441296,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4051113360323888e-05,
+      "loss": 2.8458,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4686234817813765,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.4058704453441296e-05,
+      "loss": 2.8924,
+      "step": 1852
+    },
+    {
+      "epoch": 0.4688765182186235,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.4066295546558705e-05,
+      "loss": 2.6945,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4691295546558704,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.4073886639676113e-05,
+      "loss": 2.7706,
+      "step": 1854
+    },
+    {
+      "epoch": 0.4693825910931174,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4081477732793523e-05,
+      "loss": 2.7956,
+      "step": 1855
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.4089068825910932e-05,
+      "loss": 2.7591,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4698886639676113,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.409665991902834e-05,
+      "loss": 2.6951,
+      "step": 1857
+    },
+    {
+      "epoch": 0.4701417004048583,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.410425101214575e-05,
+      "loss": 2.69,
+      "step": 1858
+    },
+    {
+      "epoch": 0.47039473684210525,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4111842105263159e-05,
+      "loss": 2.7752,
+      "step": 1859
+    },
+    {
+      "epoch": 0.4706477732793522,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.4119433198380567e-05,
+      "loss": 2.698,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4706477732793522,
+      "eval_loss": 2.855579137802124,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.7917,
+      "eval_samples_per_second": 9.637,
+      "eval_steps_per_second": 9.637,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4709008097165992,
+      "grad_norm": 11.5,
+      "learning_rate": 1.4127024291497977e-05,
+      "loss": 2.7808,
+      "step": 1861
+    },
+    {
+      "epoch": 0.47115384615384615,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4134615384615384e-05,
+      "loss": 2.7772,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4714068825910931,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.4142206477732792e-05,
+      "loss": 2.8215,
+      "step": 1863
+    },
+    {
+      "epoch": 0.4716599190283401,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4149797570850202e-05,
+      "loss": 2.8044,
+      "step": 1864
+    },
+    {
+      "epoch": 0.47191295546558704,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4157388663967611e-05,
+      "loss": 2.845,
+      "step": 1865
+    },
+    {
+      "epoch": 0.472165991902834,
+      "grad_norm": 3.375,
+      "learning_rate": 1.416497975708502e-05,
+      "loss": 2.5937,
+      "step": 1866
+    },
+    {
+      "epoch": 0.472419028340081,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.417257085020243e-05,
+      "loss": 2.804,
+      "step": 1867
+    },
+    {
+      "epoch": 0.4726720647773279,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.4180161943319838e-05,
+      "loss": 2.7933,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4729251012145749,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.4187753036437246e-05,
+      "loss": 2.7579,
+      "step": 1869
+    },
+    {
+      "epoch": 0.4731781376518219,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.4195344129554656e-05,
+      "loss": 2.8189,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4734311740890688,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4202935222672065e-05,
+      "loss": 2.7896,
+      "step": 1871
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.4210526315789473e-05,
+      "loss": 2.8078,
+      "step": 1872
+    },
+    {
+      "epoch": 0.47393724696356276,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.4218117408906883e-05,
+      "loss": 2.7065,
+      "step": 1873
+    },
+    {
+      "epoch": 0.4741902834008097,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.4225708502024292e-05,
+      "loss": 2.8103,
+      "step": 1874
+    },
+    {
+      "epoch": 0.47444331983805665,
+      "grad_norm": 3.625,
+      "learning_rate": 1.42332995951417e-05,
+      "loss": 2.7062,
+      "step": 1875
+    },
+    {
+      "epoch": 0.47469635627530365,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.424089068825911e-05,
+      "loss": 2.7585,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4749493927125506,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.4248481781376519e-05,
+      "loss": 2.7362,
+      "step": 1877
+    },
+    {
+      "epoch": 0.47520242914979755,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.4256072874493927e-05,
+      "loss": 2.7729,
+      "step": 1878
+    },
+    {
+      "epoch": 0.47545546558704455,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.4263663967611337e-05,
+      "loss": 2.7493,
+      "step": 1879
+    },
+    {
+      "epoch": 0.4757085020242915,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.4271255060728746e-05,
+      "loss": 2.7199,
+      "step": 1880
+    },
+    {
+      "epoch": 0.47596153846153844,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.4278846153846154e-05,
+      "loss": 2.8012,
+      "step": 1881
+    },
+    {
+      "epoch": 0.47621457489878544,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.4286437246963564e-05,
+      "loss": 2.8299,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4764676113360324,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.4294028340080971e-05,
+      "loss": 2.925,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4767206477732793,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.430161943319838e-05,
+      "loss": 2.6653,
+      "step": 1884
+    },
+    {
+      "epoch": 0.4769736842105263,
+      "grad_norm": 3.625,
+      "learning_rate": 1.430921052631579e-05,
+      "loss": 2.7298,
+      "step": 1885
+    },
+    {
+      "epoch": 0.4772267206477733,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.4316801619433198e-05,
+      "loss": 2.7855,
+      "step": 1886
+    },
+    {
+      "epoch": 0.4774797570850202,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.4324392712550606e-05,
+      "loss": 2.8968,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.4331983805668017e-05,
+      "loss": 2.7006,
+      "step": 1888
+    },
+    {
+      "epoch": 0.47798582995951416,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4339574898785425e-05,
+      "loss": 2.812,
+      "step": 1889
+    },
+    {
+      "epoch": 0.4782388663967611,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.4347165991902833e-05,
+      "loss": 2.8128,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4782388663967611,
+      "eval_loss": 2.852397918701172,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.9349,
+      "eval_samples_per_second": 9.632,
+      "eval_steps_per_second": 9.632,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4784919028340081,
+      "grad_norm": 3.5,
+      "learning_rate": 1.4354757085020244e-05,
+      "loss": 2.8404,
+      "step": 1891
+    },
+    {
+      "epoch": 0.47874493927125505,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.4362348178137652e-05,
+      "loss": 2.8492,
+      "step": 1892
+    },
+    {
+      "epoch": 0.478997975708502,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.436993927125506e-05,
+      "loss": 2.7302,
+      "step": 1893
+    },
+    {
+      "epoch": 0.479251012145749,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.437753036437247e-05,
+      "loss": 2.8129,
+      "step": 1894
+    },
+    {
+      "epoch": 0.47950404858299595,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.4385121457489879e-05,
+      "loss": 2.7041,
+      "step": 1895
+    },
+    {
+      "epoch": 0.4797570850202429,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.4392712550607287e-05,
+      "loss": 2.8511,
+      "step": 1896
+    },
+    {
+      "epoch": 0.4800101214574899,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.4400303643724697e-05,
+      "loss": 2.7468,
+      "step": 1897
+    },
+    {
+      "epoch": 0.48026315789473684,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.4407894736842106e-05,
+      "loss": 2.7369,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4805161943319838,
+      "grad_norm": 3.375,
+      "learning_rate": 1.4415485829959514e-05,
+      "loss": 2.7546,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4807692307692308,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.4423076923076924e-05,
+      "loss": 2.6223,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4810222672064777,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.4430668016194333e-05,
+      "loss": 2.6967,
+      "step": 1901
+    },
+    {
+      "epoch": 0.4812753036437247,
+      "grad_norm": 3.5,
+      "learning_rate": 1.4438259109311741e-05,
+      "loss": 2.8054,
+      "step": 1902
+    },
+    {
+      "epoch": 0.4815283400809717,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.4445850202429151e-05,
+      "loss": 2.5712,
+      "step": 1903
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.4453441295546558e-05,
+      "loss": 2.8019,
+      "step": 1904
+    },
+    {
+      "epoch": 0.48203441295546556,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.4461032388663967e-05,
+      "loss": 2.7461,
+      "step": 1905
+    },
+    {
+      "epoch": 0.48228744939271256,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.4468623481781377e-05,
+      "loss": 2.7389,
+      "step": 1906
+    },
+    {
+      "epoch": 0.4825404858299595,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.4476214574898785e-05,
+      "loss": 2.5593,
+      "step": 1907
+    },
+    {
+      "epoch": 0.48279352226720645,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.4483805668016194e-05,
+      "loss": 2.7621,
+      "step": 1908
+    },
+    {
+      "epoch": 0.48304655870445345,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.4491396761133604e-05,
+      "loss": 2.8061,
+      "step": 1909
+    },
+    {
+      "epoch": 0.4832995951417004,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.4498987854251012e-05,
+      "loss": 2.7372,
+      "step": 1910
+    },
+    {
+      "epoch": 0.48355263157894735,
+      "grad_norm": 3.625,
+      "learning_rate": 1.450657894736842e-05,
+      "loss": 2.8362,
+      "step": 1911
+    },
+    {
+      "epoch": 0.48380566801619435,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.451417004048583e-05,
+      "loss": 2.8772,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4840587044534413,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.4521761133603239e-05,
+      "loss": 2.8708,
+      "step": 1913
+    },
+    {
+      "epoch": 0.48431174089068824,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4529352226720647e-05,
+      "loss": 2.6938,
+      "step": 1914
+    },
+    {
+      "epoch": 0.48456477732793524,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.4536943319838058e-05,
+      "loss": 2.8284,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4848178137651822,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.4544534412955466e-05,
+      "loss": 2.8037,
+      "step": 1916
+    },
+    {
+      "epoch": 0.4850708502024291,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.4552125506072874e-05,
+      "loss": 2.8297,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4853238866396761,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.4559716599190285e-05,
+      "loss": 2.7058,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4855769230769231,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.4567307692307693e-05,
+      "loss": 2.7252,
+      "step": 1919
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4574898785425101e-05,
+      "loss": 2.8939,
+      "step": 1920
+    },
+    {
+      "epoch": 0.48582995951417,
+      "eval_loss": 2.8510003089904785,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.1988,
+      "eval_samples_per_second": 9.658,
+      "eval_steps_per_second": 9.658,
+      "step": 1920
+    },
+    {
+      "epoch": 0.486082995951417,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4582489878542512e-05,
+      "loss": 2.8459,
+      "step": 1921
+    },
+    {
+      "epoch": 0.48633603238866396,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.459008097165992e-05,
+      "loss": 2.6948,
+      "step": 1922
+    },
+    {
+      "epoch": 0.4865890688259109,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4597672064777328e-05,
+      "loss": 2.8693,
+      "step": 1923
+    },
+    {
+      "epoch": 0.4868421052631579,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.4605263157894737e-05,
+      "loss": 2.7819,
+      "step": 1924
+    },
+    {
+      "epoch": 0.48709514170040485,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.4612854251012145e-05,
+      "loss": 2.7865,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4873481781376518,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4620445344129554e-05,
+      "loss": 2.8231,
+      "step": 1926
+    },
+    {
+      "epoch": 0.4876012145748988,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4628036437246964e-05,
+      "loss": 2.7703,
+      "step": 1927
+    },
+    {
+      "epoch": 0.48785425101214575,
+      "grad_norm": 3.625,
+      "learning_rate": 1.4635627530364372e-05,
+      "loss": 2.8082,
+      "step": 1928
+    },
+    {
+      "epoch": 0.4881072874493927,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.464321862348178e-05,
+      "loss": 2.7333,
+      "step": 1929
+    },
+    {
+      "epoch": 0.4883603238866397,
+      "grad_norm": 3.5,
+      "learning_rate": 1.465080971659919e-05,
+      "loss": 2.8076,
+      "step": 1930
+    },
+    {
+      "epoch": 0.48861336032388664,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.46584008097166e-05,
+      "loss": 2.7742,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4888663967611336,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4665991902834008e-05,
+      "loss": 2.752,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4891194331983806,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4673582995951418e-05,
+      "loss": 2.8268,
+      "step": 1933
+    },
+    {
+      "epoch": 0.4893724696356275,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.4681174089068826e-05,
+      "loss": 2.798,
+      "step": 1934
+    },
+    {
+      "epoch": 0.4896255060728745,
+      "grad_norm": 3.5,
+      "learning_rate": 1.4688765182186235e-05,
+      "loss": 2.7706,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.4696356275303645e-05,
+      "loss": 2.6532,
+      "step": 1936
+    },
+    {
+      "epoch": 0.4901315789473684,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.4703947368421053e-05,
+      "loss": 2.7889,
+      "step": 1937
+    },
+    {
+      "epoch": 0.49038461538461536,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4711538461538462e-05,
+      "loss": 2.6895,
+      "step": 1938
+    },
+    {
+      "epoch": 0.49063765182186236,
+      "grad_norm": 41.5,
+      "learning_rate": 1.4719129554655872e-05,
+      "loss": 2.8725,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4908906882591093,
+      "grad_norm": 4.125,
+      "learning_rate": 1.472672064777328e-05,
+      "loss": 2.8383,
+      "step": 1940
+    },
+    {
+      "epoch": 0.49114372469635625,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.4734311740890689e-05,
+      "loss": 2.7652,
+      "step": 1941
+    },
+    {
+      "epoch": 0.49139676113360325,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4741902834008099e-05,
+      "loss": 2.7934,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4916497975708502,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.4749493927125507e-05,
+      "loss": 2.7261,
+      "step": 1943
+    },
+    {
+      "epoch": 0.49190283400809715,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4757085020242915e-05,
+      "loss": 2.8112,
+      "step": 1944
+    },
+    {
+      "epoch": 0.49215587044534415,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4764676113360324e-05,
+      "loss": 2.7245,
+      "step": 1945
+    },
+    {
+      "epoch": 0.4924089068825911,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4772267206477732e-05,
+      "loss": 2.8077,
+      "step": 1946
+    },
+    {
+      "epoch": 0.49266194331983804,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.477985829959514e-05,
+      "loss": 2.7702,
+      "step": 1947
+    },
+    {
+      "epoch": 0.49291497975708504,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.4787449392712551e-05,
+      "loss": 2.7904,
+      "step": 1948
+    },
+    {
+      "epoch": 0.493168016194332,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.479504048582996e-05,
+      "loss": 2.5896,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4934210526315789,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.4802631578947368e-05,
+      "loss": 2.8204,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4934210526315789,
+      "eval_loss": 2.8465347290039062,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.354,
+      "eval_samples_per_second": 9.617,
+      "eval_steps_per_second": 9.617,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4936740890688259,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.4810222672064778e-05,
+      "loss": 2.8293,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4817813765182186e-05,
+      "loss": 2.7906,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4941801619433198,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.4825404858299595e-05,
+      "loss": 2.7507,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4944331983805668,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.4832995951417005e-05,
+      "loss": 2.7623,
+      "step": 1954
+    },
+    {
+      "epoch": 0.49468623481781376,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.4840587044534413e-05,
+      "loss": 2.7314,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4949392712550607,
+      "grad_norm": 3.625,
+      "learning_rate": 1.4848178137651822e-05,
+      "loss": 2.824,
+      "step": 1956
+    },
+    {
+      "epoch": 0.4951923076923077,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.4855769230769232e-05,
+      "loss": 2.4762,
+      "step": 1957
+    },
+    {
+      "epoch": 0.49544534412955465,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.486336032388664e-05,
+      "loss": 2.6091,
+      "step": 1958
+    },
+    {
+      "epoch": 0.4956983805668016,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.4870951417004049e-05,
+      "loss": 2.8759,
+      "step": 1959
+    },
+    {
+      "epoch": 0.4959514170040486,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.4878542510121459e-05,
+      "loss": 2.7919,
+      "step": 1960
+    },
+    {
+      "epoch": 0.49620445344129555,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.4886133603238867e-05,
+      "loss": 2.9196,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4964574898785425,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.4893724696356276e-05,
+      "loss": 2.6763,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4967105263157895,
+      "grad_norm": 3.5,
+      "learning_rate": 1.4901315789473686e-05,
+      "loss": 2.7987,
+      "step": 1963
+    },
+    {
+      "epoch": 0.49696356275303644,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.4908906882591094e-05,
+      "loss": 2.8038,
+      "step": 1964
+    },
+    {
+      "epoch": 0.4972165991902834,
+      "grad_norm": 3.96875,
+      "learning_rate": 1.4916497975708503e-05,
+      "loss": 2.7977,
+      "step": 1965
+    },
+    {
+      "epoch": 0.4974696356275304,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.4924089068825911e-05,
+      "loss": 2.8488,
+      "step": 1966
+    },
+    {
+      "epoch": 0.4977226720647773,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.493168016194332e-05,
+      "loss": 2.8237,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4939271255060728e-05,
+      "loss": 2.7703,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4982287449392713,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.4946862348178138e-05,
+      "loss": 2.7558,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4984817813765182,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.4954453441295546e-05,
+      "loss": 2.7187,
+      "step": 1970
+    },
+    {
+      "epoch": 0.49873481781376516,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.4962044534412955e-05,
+      "loss": 2.8425,
+      "step": 1971
+    },
+    {
+      "epoch": 0.49898785425101216,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.4969635627530365e-05,
+      "loss": 2.795,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4992408906882591,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.4977226720647773e-05,
+      "loss": 2.7879,
+      "step": 1973
+    },
+    {
+      "epoch": 0.49949392712550605,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.4984817813765182e-05,
+      "loss": 2.7878,
+      "step": 1974
+    },
+    {
+      "epoch": 0.49974696356275305,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.4992408906882592e-05,
+      "loss": 2.8405,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.5e-05,
+      "loss": 2.7937,
+      "step": 1976
+    },
+    {
+      "epoch": 0.500253036437247,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.5007591093117409e-05,
+      "loss": 2.7787,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5005060728744939,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.5015182186234817e-05,
+      "loss": 2.7404,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5007591093117408,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.5022773279352226e-05,
+      "loss": 2.7649,
+      "step": 1979
+    },
+    {
+      "epoch": 0.5010121457489879,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.5030364372469637e-05,
+      "loss": 2.7641,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5010121457489879,
+      "eval_loss": 2.8414037227630615,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.0649,
+      "eval_samples_per_second": 9.627,
+      "eval_steps_per_second": 9.627,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5012651821862348,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.5037955465587046e-05,
+      "loss": 2.7896,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5015182186234818,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.5045546558704454e-05,
+      "loss": 2.7682,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5017712550607287,
+      "grad_norm": 3.5,
+      "learning_rate": 1.5053137651821863e-05,
+      "loss": 2.8287,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.5060728744939271e-05,
+      "loss": 2.7517,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5022773279352226,
+      "grad_norm": 3.75,
+      "learning_rate": 1.506831983805668e-05,
+      "loss": 2.7687,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5025303643724697,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.5075910931174091e-05,
+      "loss": 2.7143,
+      "step": 1986
+    },
+    {
+      "epoch": 0.5027834008097166,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.5083502024291498e-05,
+      "loss": 2.7488,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5030364372469636,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.5091093117408907e-05,
+      "loss": 2.7904,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5032894736842105,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.5098684210526315e-05,
+      "loss": 2.5496,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5035425101214575,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.5106275303643723e-05,
+      "loss": 2.7644,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5037955465587044,
+      "grad_norm": 3.375,
+      "learning_rate": 1.5113866396761132e-05,
+      "loss": 2.7222,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5040485829959515,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.5121457489878544e-05,
+      "loss": 2.6273,
+      "step": 1992
+    },
+    {
+      "epoch": 0.5043016194331984,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.5129048582995952e-05,
+      "loss": 2.7509,
+      "step": 1993
+    },
+    {
+      "epoch": 0.5045546558704453,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.513663967611336e-05,
+      "loss": 2.7443,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5048076923076923,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.5144230769230769e-05,
+      "loss": 2.7898,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5050607287449392,
+      "grad_norm": 46.5,
+      "learning_rate": 1.5151821862348177e-05,
+      "loss": 2.9922,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5053137651821862,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.5159412955465586e-05,
+      "loss": 2.7922,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5055668016194332,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.5167004048582998e-05,
+      "loss": 2.6955,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5058198380566802,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.5174595141700406e-05,
+      "loss": 2.7534,
+      "step": 1999
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.5182186234817814e-05,
+      "loss": 2.8545,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5063259109311741,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.5189777327935223e-05,
+      "loss": 2.6663,
+      "step": 2001
+    },
+    {
+      "epoch": 0.506578947368421,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.5197368421052631e-05,
+      "loss": 2.8825,
+      "step": 2002
+    },
+    {
+      "epoch": 0.506831983805668,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.520495951417004e-05,
+      "loss": 2.8568,
+      "step": 2003
+    },
+    {
+      "epoch": 0.507085020242915,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.5212550607287452e-05,
+      "loss": 2.6715,
+      "step": 2004
+    },
+    {
+      "epoch": 0.507338056680162,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.522014170040486e-05,
+      "loss": 2.7363,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5075910931174089,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.5227732793522268e-05,
+      "loss": 2.8163,
+      "step": 2006
+    },
+    {
+      "epoch": 0.5078441295546559,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.5235323886639677e-05,
+      "loss": 2.6603,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5080971659919028,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.5242914979757084e-05,
+      "loss": 2.8167,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5083502024291497,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.5250506072874492e-05,
+      "loss": 2.7045,
+      "step": 2009
+    },
+    {
+      "epoch": 0.5086032388663968,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.5258097165991904e-05,
+      "loss": 2.8118,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5086032388663968,
+      "eval_loss": 2.839552402496338,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.2163,
+      "eval_samples_per_second": 9.413,
+      "eval_steps_per_second": 9.413,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5088562753036437,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.5265688259109312e-05,
+      "loss": 2.8463,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5091093117408907,
+      "grad_norm": 3.375,
+      "learning_rate": 1.5273279352226722e-05,
+      "loss": 2.6929,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5093623481781376,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.528087044534413e-05,
+      "loss": 2.7964,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5096153846153846,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.528846153846154e-05,
+      "loss": 2.6365,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5098684210526315,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.5296052631578946e-05,
+      "loss": 2.8001,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.530364372469636e-05,
+      "loss": 2.8843,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5103744939271255,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.5311234817813766e-05,
+      "loss": 2.7549,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5106275303643725,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.5318825910931173e-05,
+      "loss": 2.7458,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5108805668016194,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.5326417004048583e-05,
+      "loss": 2.7859,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5111336032388664,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.533400809716599e-05,
+      "loss": 2.7912,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5113866396761133,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.53415991902834e-05,
+      "loss": 2.6596,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5116396761133604,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.534919028340081e-05,
+      "loss": 2.6199,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5118927125506073,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.535678137651822e-05,
+      "loss": 2.7808,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5121457489878543,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.5364372469635627e-05,
+      "loss": 2.6743,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5123987854251012,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.5371963562753037e-05,
+      "loss": 2.6813,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5126518218623481,
+      "grad_norm": 3.75,
+      "learning_rate": 1.5379554655870444e-05,
+      "loss": 2.8256,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5129048582995951,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.5387145748987854e-05,
+      "loss": 2.7008,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5131578947368421,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.5394736842105264e-05,
+      "loss": 2.7512,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5134109311740891,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.5402327935222674e-05,
+      "loss": 2.7232,
+      "step": 2029
+    },
+    {
+      "epoch": 0.513663967611336,
+      "grad_norm": 3.5,
+      "learning_rate": 1.540991902834008e-05,
+      "loss": 2.7432,
+      "step": 2030
+    },
+    {
+      "epoch": 0.513917004048583,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.541751012145749e-05,
+      "loss": 2.8801,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.5425101214574898e-05,
+      "loss": 2.7671,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5144230769230769,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.5432692307692308e-05,
+      "loss": 2.7173,
+      "step": 2033
+    },
+    {
+      "epoch": 0.5146761133603239,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.5440283400809718e-05,
+      "loss": 2.7215,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5149291497975709,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.5447874493927128e-05,
+      "loss": 2.8015,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5151821862348178,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.5455465587044535e-05,
+      "loss": 2.9221,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5154352226720648,
+      "grad_norm": 3.5,
+      "learning_rate": 1.5463056680161945e-05,
+      "loss": 2.7518,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5156882591093117,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.547064777327935e-05,
+      "loss": 2.6568,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5159412955465587,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.547823886639676e-05,
+      "loss": 2.7185,
+      "step": 2039
+    },
+    {
+      "epoch": 0.5161943319838057,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.5485829959514172e-05,
+      "loss": 2.6579,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5161943319838057,
+      "eval_loss": 2.8350841999053955,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 278.6379,
+      "eval_samples_per_second": 9.331,
+      "eval_steps_per_second": 9.331,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5164473684210527,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.549342105263158e-05,
+      "loss": 2.7292,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5167004048582996,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.550101214574899e-05,
+      "loss": 2.8785,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5169534412955465,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.5508603238866395e-05,
+      "loss": 2.8221,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5172064777327935,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.5516194331983806e-05,
+      "loss": 2.6208,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5174595141700404,
+      "grad_norm": 3.375,
+      "learning_rate": 1.5523785425101212e-05,
+      "loss": 2.7962,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5177125506072875,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.5531376518218626e-05,
+      "loss": 2.7519,
+      "step": 2046
+    },
+    {
+      "epoch": 0.5179655870445344,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.5538967611336033e-05,
+      "loss": 2.7252,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.5546558704453443e-05,
+      "loss": 2.5043,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5184716599190283,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.555414979757085e-05,
+      "loss": 2.7925,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5187246963562753,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.556174089068826e-05,
+      "loss": 2.7042,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5189777327935222,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.5569331983805666e-05,
+      "loss": 2.6854,
+      "step": 2051
+    },
+    {
+      "epoch": 0.5192307692307693,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.557692307692308e-05,
+      "loss": 2.823,
+      "step": 2052
+    },
+    {
+      "epoch": 0.5194838056680162,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.5584514170040486e-05,
+      "loss": 2.7536,
+      "step": 2053
+    },
+    {
+      "epoch": 0.5197368421052632,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.5592105263157897e-05,
+      "loss": 2.7724,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5199898785425101,
+      "grad_norm": 3.5,
+      "learning_rate": 1.5599696356275303e-05,
+      "loss": 2.7553,
+      "step": 2055
+    },
+    {
+      "epoch": 0.520242914979757,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.5607287449392713e-05,
+      "loss": 2.7167,
+      "step": 2056
+    },
+    {
+      "epoch": 0.520495951417004,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.561487854251012e-05,
+      "loss": 2.6856,
+      "step": 2057
+    },
+    {
+      "epoch": 0.520748987854251,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.5622469635627534e-05,
+      "loss": 2.7564,
+      "step": 2058
+    },
+    {
+      "epoch": 0.521002024291498,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.563006072874494e-05,
+      "loss": 2.729,
+      "step": 2059
+    },
+    {
+      "epoch": 0.521255060728745,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.5637651821862347e-05,
+      "loss": 2.7201,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5215080971659919,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.5645242914979757e-05,
+      "loss": 2.7485,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5217611336032388,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.5652834008097164e-05,
+      "loss": 2.8055,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5220141700404858,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.5660425101214574e-05,
+      "loss": 2.6645,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.5668016194331984e-05,
+      "loss": 2.7524,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5225202429149798,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.5675607287449394e-05,
+      "loss": 2.7824,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5227732793522267,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.56831983805668e-05,
+      "loss": 2.7572,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5230263157894737,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.569078947368421e-05,
+      "loss": 2.8254,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5232793522267206,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.5698380566801618e-05,
+      "loss": 2.8107,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5235323886639676,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.5705971659919028e-05,
+      "loss": 2.6537,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5237854251012146,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.5713562753036438e-05,
+      "loss": 2.785,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5237854251012146,
+      "eval_loss": 2.831545829772949,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.3779,
+      "eval_samples_per_second": 9.616,
+      "eval_steps_per_second": 9.616,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5240384615384616,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.5721153846153848e-05,
+      "loss": 2.854,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5242914979757085,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.5728744939271255e-05,
+      "loss": 2.745,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5245445344129555,
+      "grad_norm": 3.75,
+      "learning_rate": 1.5736336032388665e-05,
+      "loss": 2.7818,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5247975708502024,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.5743927125506072e-05,
+      "loss": 2.6562,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5250506072874493,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.5751518218623482e-05,
+      "loss": 2.7594,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5253036437246964,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.5759109311740892e-05,
+      "loss": 2.5598,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5255566801619433,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.5766700404858302e-05,
+      "loss": 2.7486,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5258097165991903,
+      "grad_norm": 3.625,
+      "learning_rate": 1.577429149797571e-05,
+      "loss": 2.7373,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5260627530364372,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.578188259109312e-05,
+      "loss": 2.844,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 3.375,
+      "learning_rate": 1.5789473684210526e-05,
+      "loss": 2.7126,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5265688259109311,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.5797064777327933e-05,
+      "loss": 2.5914,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5268218623481782,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.5804655870445346e-05,
+      "loss": 2.7841,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5270748987854251,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.5812246963562753e-05,
+      "loss": 2.8018,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5273279352226721,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.5819838056680163e-05,
+      "loss": 2.7334,
+      "step": 2084
+    },
+    {
+      "epoch": 0.527580971659919,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.582742914979757e-05,
+      "loss": 2.8023,
+      "step": 2085
+    },
+    {
+      "epoch": 0.527834008097166,
+      "grad_norm": 3.5,
+      "learning_rate": 1.583502024291498e-05,
+      "loss": 2.7143,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5280870445344129,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.5842611336032387e-05,
+      "loss": 2.7554,
+      "step": 2087
+    },
+    {
+      "epoch": 0.52834008097166,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.58502024291498e-05,
+      "loss": 2.9647,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5285931174089069,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.5857793522267207e-05,
+      "loss": 2.7786,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5288461538461539,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.5865384615384617e-05,
+      "loss": 2.786,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5290991902834008,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.5872975708502024e-05,
+      "loss": 2.7759,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5293522267206477,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.5880566801619434e-05,
+      "loss": 2.6992,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5296052631578947,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.588815789473684e-05,
+      "loss": 2.8013,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5298582995951417,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.5895748987854254e-05,
+      "loss": 2.7741,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5301113360323887,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.590334008097166e-05,
+      "loss": 2.786,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.591093117408907e-05,
+      "loss": 2.7299,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5306174089068826,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.5918522267206478e-05,
+      "loss": 2.8205,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5308704453441295,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.5926113360323888e-05,
+      "loss": 2.7798,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5311234817813765,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.5933704453441294e-05,
+      "loss": 2.8824,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5313765182186235,
+      "grad_norm": 4.0,
+      "learning_rate": 1.5941295546558708e-05,
+      "loss": 2.6203,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5313765182186235,
+      "eval_loss": 2.830051898956299,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.6831,
+      "eval_samples_per_second": 9.641,
+      "eval_steps_per_second": 9.641,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5316295546558705,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.5948886639676115e-05,
+      "loss": 2.6483,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5318825910931174,
+      "grad_norm": 3.5,
+      "learning_rate": 1.595647773279352e-05,
+      "loss": 2.7409,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5321356275303644,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.596406882591093e-05,
+      "loss": 2.8094,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5323886639676113,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.5971659919028338e-05,
+      "loss": 2.7684,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5326417004048583,
+      "grad_norm": 3.5,
+      "learning_rate": 1.597925101214575e-05,
+      "loss": 2.5918,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5328947368421053,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.598684210526316e-05,
+      "loss": 2.7339,
+      "step": 2106
+    },
+    {
+      "epoch": 0.5331477732793523,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.599443319838057e-05,
+      "loss": 2.7743,
+      "step": 2107
+    },
+    {
+      "epoch": 0.5334008097165992,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.6002024291497975e-05,
+      "loss": 2.8056,
+      "step": 2108
+    },
+    {
+      "epoch": 0.5336538461538461,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.6009615384615385e-05,
+      "loss": 2.7493,
+      "step": 2109
+    },
+    {
+      "epoch": 0.5339068825910931,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.6017206477732792e-05,
+      "loss": 2.6141,
+      "step": 2110
+    },
+    {
+      "epoch": 0.53415991902834,
+      "grad_norm": 3.5,
+      "learning_rate": 1.6024797570850202e-05,
+      "loss": 2.7438,
+      "step": 2111
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.6032388663967612e-05,
+      "loss": 2.687,
+      "step": 2112
+    },
+    {
+      "epoch": 0.534665991902834,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.6039979757085023e-05,
+      "loss": 2.8384,
+      "step": 2113
+    },
+    {
+      "epoch": 0.534919028340081,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.604757085020243e-05,
+      "loss": 2.5207,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5351720647773279,
+      "grad_norm": 3.75,
+      "learning_rate": 1.605516194331984e-05,
+      "loss": 2.8408,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5354251012145749,
+      "grad_norm": 3.5,
+      "learning_rate": 1.6062753036437246e-05,
+      "loss": 2.5123,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5356781376518218,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.6070344129554656e-05,
+      "loss": 2.6737,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5359311740890689,
+      "grad_norm": 3.5,
+      "learning_rate": 1.6077935222672066e-05,
+      "loss": 2.7567,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5361842105263158,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.6085526315789476e-05,
+      "loss": 2.7681,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5364372469635628,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.6093117408906883e-05,
+      "loss": 2.8274,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5366902834008097,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.6100708502024293e-05,
+      "loss": 2.7744,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5369433198380567,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.61082995951417e-05,
+      "loss": 2.7833,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5371963562753036,
+      "grad_norm": 3.75,
+      "learning_rate": 1.6115890688259107e-05,
+      "loss": 2.6587,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5374493927125507,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.612348178137652e-05,
+      "loss": 2.7009,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5377024291497976,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.6131072874493927e-05,
+      "loss": 2.7993,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5379554655870445,
+      "grad_norm": 3.625,
+      "learning_rate": 1.6138663967611337e-05,
+      "loss": 2.7901,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5382085020242915,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.6146255060728744e-05,
+      "loss": 2.4543,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 2.7119,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5387145748987854,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.616143724696356e-05,
+      "loss": 2.4396,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5389676113360324,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.6169028340080974e-05,
+      "loss": 2.665,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5389676113360324,
+      "eval_loss": 2.823349714279175,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.9312,
+      "eval_samples_per_second": 9.632,
+      "eval_steps_per_second": 9.632,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5392206477732794,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.617661943319838e-05,
+      "loss": 2.795,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5394736842105263,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.618421052631579e-05,
+      "loss": 2.8233,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5397267206477733,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.6191801619433198e-05,
+      "loss": 2.656,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5399797570850202,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.6199392712550608e-05,
+      "loss": 2.692,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5402327935222672,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.6206983805668015e-05,
+      "loss": 2.8802,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5404858299595142,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.6214574898785428e-05,
+      "loss": 2.7377,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5407388663967612,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.6222165991902835e-05,
+      "loss": 2.6904,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5409919028340081,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.6229757085020245e-05,
+      "loss": 2.7981,
+      "step": 2138
+    },
+    {
+      "epoch": 0.541244939271255,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.6237348178137652e-05,
+      "loss": 2.8857,
+      "step": 2139
+    },
+    {
+      "epoch": 0.541497975708502,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.6244939271255062e-05,
+      "loss": 2.8512,
+      "step": 2140
+    },
+    {
+      "epoch": 0.541751012145749,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.625253036437247e-05,
+      "loss": 2.7707,
+      "step": 2141
+    },
+    {
+      "epoch": 0.542004048582996,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.6260121457489882e-05,
+      "loss": 2.7688,
+      "step": 2142
+    },
+    {
+      "epoch": 0.542257085020243,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.626771255060729e-05,
+      "loss": 2.7687,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.6275303643724696e-05,
+      "loss": 2.7007,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5427631578947368,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.6282894736842106e-05,
+      "loss": 2.618,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5430161943319838,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.6290485829959512e-05,
+      "loss": 2.6143,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5432692307692307,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.6298076923076923e-05,
+      "loss": 2.7079,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5435222672064778,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.6305668016194333e-05,
+      "loss": 2.7924,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5437753036437247,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.6313259109311743e-05,
+      "loss": 2.6737,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5440283400809717,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.632085020242915e-05,
+      "loss": 2.7443,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5442813765182186,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.632844129554656e-05,
+      "loss": 2.7817,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5445344129554656,
+      "grad_norm": 3.625,
+      "learning_rate": 1.6336032388663966e-05,
+      "loss": 2.8466,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5447874493927125,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.6343623481781377e-05,
+      "loss": 2.7845,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5450404858299596,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.6351214574898787e-05,
+      "loss": 2.8555,
+      "step": 2154
+    },
+    {
+      "epoch": 0.5452935222672065,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.6358805668016197e-05,
+      "loss": 2.6963,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5455465587044535,
+      "grad_norm": 3.5,
+      "learning_rate": 1.6366396761133603e-05,
+      "loss": 2.6888,
+      "step": 2156
+    },
+    {
+      "epoch": 0.5457995951417004,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.6373987854251014e-05,
+      "loss": 2.7136,
+      "step": 2157
+    },
+    {
+      "epoch": 0.5460526315789473,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.638157894736842e-05,
+      "loss": 2.8026,
+      "step": 2158
+    },
+    {
+      "epoch": 0.5463056680161943,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.638917004048583e-05,
+      "loss": 2.7424,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.639676113360324e-05,
+      "loss": 2.7106,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "eval_loss": 2.818882942199707,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.2492,
+      "eval_samples_per_second": 9.621,
+      "eval_steps_per_second": 9.621,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5468117408906883,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.640435222672065e-05,
+      "loss": 2.6388,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5470647773279352,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.6411943319838057e-05,
+      "loss": 2.6687,
+      "step": 2162
+    },
+    {
+      "epoch": 0.5473178137651822,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.6419534412955468e-05,
+      "loss": 2.6712,
+      "step": 2163
+    },
+    {
+      "epoch": 0.5475708502024291,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.6427125506072874e-05,
+      "loss": 2.7431,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5478238866396761,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.643471659919028e-05,
+      "loss": 2.7706,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5480769230769231,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.6442307692307695e-05,
+      "loss": 2.7377,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5483299595141701,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.64498987854251e-05,
+      "loss": 2.5928,
+      "step": 2167
+    },
+    {
+      "epoch": 0.548582995951417,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.645748987854251e-05,
+      "loss": 2.8041,
+      "step": 2168
+    },
+    {
+      "epoch": 0.548836032388664,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.6465080971659918e-05,
+      "loss": 2.7308,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5490890688259109,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.6472672064777328e-05,
+      "loss": 2.8866,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5493421052631579,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.6480263157894735e-05,
+      "loss": 2.7983,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5495951417004049,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.648785425101215e-05,
+      "loss": 2.7152,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5498481781376519,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.6495445344129555e-05,
+      "loss": 2.6972,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5501012145748988,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.6503036437246965e-05,
+      "loss": 2.8287,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5503542510121457,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.6510627530364372e-05,
+      "loss": 2.7556,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.6518218623481782e-05,
+      "loss": 2.8178,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5508603238866396,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.652580971659919e-05,
+      "loss": 2.8134,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5511133603238867,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.6533400809716602e-05,
+      "loss": 2.7028,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5513663967611336,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.654099190283401e-05,
+      "loss": 2.7672,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5516194331983806,
+      "grad_norm": 3.375,
+      "learning_rate": 1.654858299595142e-05,
+      "loss": 2.6655,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5518724696356275,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.6556174089068826e-05,
+      "loss": 2.8061,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5521255060728745,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.6563765182186236e-05,
+      "loss": 2.7377,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5523785425101214,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.6571356275303643e-05,
+      "loss": 2.6905,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.6578947368421053e-05,
+      "loss": 2.7598,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5528846153846154,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.6586538461538463e-05,
+      "loss": 2.5939,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5531376518218624,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.659412955465587e-05,
+      "loss": 2.733,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5533906882591093,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.660172064777328e-05,
+      "loss": 2.8178,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5536437246963563,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.6609311740890687e-05,
+      "loss": 2.793,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5538967611336032,
+      "grad_norm": 3.25,
+      "learning_rate": 1.6616902834008097e-05,
+      "loss": 2.766,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5541497975708503,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.6624493927125507e-05,
+      "loss": 2.7422,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5541497975708503,
+      "eval_loss": 2.81355619430542,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.0231,
+      "eval_samples_per_second": 9.665,
+      "eval_steps_per_second": 9.665,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5544028340080972,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.6632085020242917e-05,
+      "loss": 2.8014,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.6639676113360324e-05,
+      "loss": 2.7896,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5549089068825911,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.6647267206477734e-05,
+      "loss": 2.7373,
+      "step": 2193
+    },
+    {
+      "epoch": 0.555161943319838,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.665485829959514e-05,
+      "loss": 2.7008,
+      "step": 2194
+    },
+    {
+      "epoch": 0.555414979757085,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.666244939271255e-05,
+      "loss": 2.8135,
+      "step": 2195
+    },
+    {
+      "epoch": 0.555668016194332,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.667004048582996e-05,
+      "loss": 2.7074,
+      "step": 2196
+    },
+    {
+      "epoch": 0.555921052631579,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.667763157894737e-05,
+      "loss": 2.6969,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5561740890688259,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.6685222672064778e-05,
+      "loss": 2.6619,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5564271255060729,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.6692813765182188e-05,
+      "loss": 2.6592,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5566801619433198,
+      "grad_norm": 3.375,
+      "learning_rate": 1.6700404858299595e-05,
+      "loss": 2.6958,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5569331983805668,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.6707995951417005e-05,
+      "loss": 2.7204,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5571862348178138,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.6715587044534415e-05,
+      "loss": 2.7316,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5574392712550608,
+      "grad_norm": 3.5,
+      "learning_rate": 1.6723178137651825e-05,
+      "loss": 2.7552,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5576923076923077,
+      "grad_norm": 3.5,
+      "learning_rate": 1.673076923076923e-05,
+      "loss": 2.7641,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5579453441295547,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.673836032388664e-05,
+      "loss": 2.7312,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5581983805668016,
+      "grad_norm": 3.625,
+      "learning_rate": 1.674595141700405e-05,
+      "loss": 2.7822,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5584514170040485,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.6753542510121455e-05,
+      "loss": 2.704,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.676113360323887e-05,
+      "loss": 2.7231,
+      "step": 2208
+    },
+    {
+      "epoch": 0.5589574898785425,
+      "grad_norm": 3.5,
+      "learning_rate": 1.6768724696356275e-05,
+      "loss": 2.7335,
+      "step": 2209
+    },
+    {
+      "epoch": 0.5592105263157895,
+      "grad_norm": 3.5,
+      "learning_rate": 1.6776315789473686e-05,
+      "loss": 2.8011,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5594635627530364,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.6783906882591092e-05,
+      "loss": 2.7771,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5597165991902834,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.6791497975708502e-05,
+      "loss": 2.7715,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5599696356275303,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.679908906882591e-05,
+      "loss": 2.6649,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5602226720647774,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.6806680161943323e-05,
+      "loss": 2.686,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5604757085020243,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.681427125506073e-05,
+      "loss": 2.6366,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5607287449392713,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.682186234817814e-05,
+      "loss": 2.6119,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5609817813765182,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.6829453441295546e-05,
+      "loss": 2.7101,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5612348178137652,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.6837044534412956e-05,
+      "loss": 2.6691,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5614878542510121,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.6844635627530363e-05,
+      "loss": 2.7281,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5617408906882592,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.6852226720647777e-05,
+      "loss": 2.8306,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5617408906882592,
+      "eval_loss": 2.813448905944824,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.1158,
+      "eval_samples_per_second": 9.626,
+      "eval_steps_per_second": 9.626,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5619939271255061,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.6859817813765183e-05,
+      "loss": 2.835,
+      "step": 2221
+    },
+    {
+      "epoch": 0.562246963562753,
+      "grad_norm": 3.75,
+      "learning_rate": 1.6867408906882593e-05,
+      "loss": 2.7916,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.6875e-05,
+      "loss": 2.7588,
+      "step": 2223
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.688259109311741e-05,
+      "loss": 2.785,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5630060728744939,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.6890182186234817e-05,
+      "loss": 2.8396,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5632591093117408,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.6897773279352224e-05,
+      "loss": 2.7375,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5635121457489879,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.6905364372469637e-05,
+      "loss": 2.8193,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5637651821862348,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.6912955465587044e-05,
+      "loss": 2.7063,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5640182186234818,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.6920546558704454e-05,
+      "loss": 2.7886,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5642712550607287,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.692813765182186e-05,
+      "loss": 2.7943,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5645242914979757,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.693572874493927e-05,
+      "loss": 2.7491,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5647773279352226,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.6943319838056678e-05,
+      "loss": 2.6827,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5650303643724697,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.695091093117409e-05,
+      "loss": 2.7792,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5652834008097166,
+      "grad_norm": 3.625,
+      "learning_rate": 1.6958502024291498e-05,
+      "loss": 2.8075,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5655364372469636,
+      "grad_norm": 3.375,
+      "learning_rate": 1.6966093117408908e-05,
+      "loss": 2.6515,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5657894736842105,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.6973684210526315e-05,
+      "loss": 2.7458,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5660425101214575,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.6981275303643725e-05,
+      "loss": 2.7717,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5662955465587044,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.698886639676113e-05,
+      "loss": 2.7329,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5665485829959515,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.6996457489878545e-05,
+      "loss": 2.7832,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.7004048582995952e-05,
+      "loss": 2.6994,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5670546558704453,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.7011639676113362e-05,
+      "loss": 2.7588,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5673076923076923,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.701923076923077e-05,
+      "loss": 2.7956,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5675607287449392,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.702682186234818e-05,
+      "loss": 2.6725,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5678137651821862,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.7034412955465586e-05,
+      "loss": 2.7974,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5680668016194332,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.7042004048583e-05,
+      "loss": 2.6609,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5683198380566802,
+      "grad_norm": 3.375,
+      "learning_rate": 1.7049595141700406e-05,
+      "loss": 2.5255,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5685728744939271,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.7057186234817813e-05,
+      "loss": 2.8332,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5688259109311741,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.7064777327935223e-05,
+      "loss": 2.6042,
+      "step": 2248
+    },
+    {
+      "epoch": 0.569078947368421,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.707236842105263e-05,
+      "loss": 2.4928,
+      "step": 2249
+    },
+    {
+      "epoch": 0.569331983805668,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.707995951417004e-05,
+      "loss": 2.845,
+      "step": 2250
+    },
+    {
+      "epoch": 0.569331983805668,
+      "eval_loss": 2.807511329650879,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 271.5018,
+      "eval_samples_per_second": 9.576,
+      "eval_steps_per_second": 9.576,
+      "step": 2250
+    },
+    {
+      "epoch": 0.569585020242915,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.708755060728745e-05,
+      "loss": 2.78,
+      "step": 2251
+    },
+    {
+      "epoch": 0.569838056680162,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.709514170040486e-05,
+      "loss": 2.778,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5700910931174089,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.7102732793522267e-05,
+      "loss": 2.8197,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5703441295546559,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.7110323886639677e-05,
+      "loss": 2.828,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5705971659919028,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.7117914979757083e-05,
+      "loss": 2.7828,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.7125506072874494e-05,
+      "loss": 2.8196,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5711032388663968,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.7133097165991904e-05,
+      "loss": 2.6943,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5713562753036437,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.7140688259109314e-05,
+      "loss": 2.7463,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5716093117408907,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.714827935222672e-05,
+      "loss": 2.6227,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5718623481781376,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.715587044534413e-05,
+      "loss": 2.8672,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5721153846153846,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.7163461538461537e-05,
+      "loss": 2.9132,
+      "step": 2261
+    },
+    {
+      "epoch": 0.5723684210526315,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.7171052631578947e-05,
+      "loss": 2.757,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5726214574898786,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.7178643724696358e-05,
+      "loss": 2.7222,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5728744939271255,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.7186234817813768e-05,
+      "loss": 2.6505,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5731275303643725,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.7193825910931174e-05,
+      "loss": 2.732,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5733805668016194,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.7201417004048585e-05,
+      "loss": 2.7684,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5736336032388664,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.720900809716599e-05,
+      "loss": 2.7585,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5738866396761133,
+      "grad_norm": 3.25,
+      "learning_rate": 1.7216599190283398e-05,
+      "loss": 2.6153,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5741396761133604,
+      "grad_norm": 3.625,
+      "learning_rate": 1.722419028340081e-05,
+      "loss": 2.6918,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5743927125506073,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.7231781376518218e-05,
+      "loss": 2.7548,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5746457489878543,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.723937246963563e-05,
+      "loss": 2.5726,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.7246963562753035e-05,
+      "loss": 2.6712,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5751518218623481,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.7254554655870445e-05,
+      "loss": 2.8225,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5754048582995951,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.7262145748987852e-05,
+      "loss": 2.7124,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5756578947368421,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.7269736842105265e-05,
+      "loss": 2.5403,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5759109311740891,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.7277327935222672e-05,
+      "loss": 2.7302,
+      "step": 2276
+    },
+    {
+      "epoch": 0.576163967611336,
+      "grad_norm": 3.5,
+      "learning_rate": 1.7284919028340082e-05,
+      "loss": 2.8183,
+      "step": 2277
+    },
+    {
+      "epoch": 0.576417004048583,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.729251012145749e-05,
+      "loss": 2.8727,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5766700404858299,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.73001012145749e-05,
+      "loss": 2.6771,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5769230769230769,
+      "grad_norm": 3.5,
+      "learning_rate": 1.7307692307692306e-05,
+      "loss": 2.7066,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5769230769230769,
+      "eval_loss": 2.8080427646636963,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.7523,
+      "eval_samples_per_second": 9.638,
+      "eval_steps_per_second": 9.638,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5771761133603239,
+      "grad_norm": 3.25,
+      "learning_rate": 1.731528340080972e-05,
+      "loss": 2.6825,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5774291497975709,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.7322874493927126e-05,
+      "loss": 2.6673,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5776821862348178,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.7330465587044536e-05,
+      "loss": 2.7148,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5779352226720648,
+      "grad_norm": 3.875,
+      "learning_rate": 1.7338056680161943e-05,
+      "loss": 2.8065,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5781882591093117,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.7345647773279353e-05,
+      "loss": 2.7531,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5784412955465587,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.735323886639676e-05,
+      "loss": 2.703,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5786943319838057,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.7360829959514173e-05,
+      "loss": 2.5873,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 3.75,
+      "learning_rate": 1.736842105263158e-05,
+      "loss": 2.777,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5792004048582996,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.7376012145748987e-05,
+      "loss": 2.7695,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5794534412955465,
+      "grad_norm": 3.75,
+      "learning_rate": 1.7383603238866397e-05,
+      "loss": 2.7932,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5797064777327935,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.7391194331983804e-05,
+      "loss": 2.7781,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5799595141700404,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.7398785425101214e-05,
+      "loss": 2.6408,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5802125506072875,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.7406376518218624e-05,
+      "loss": 2.8109,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5804655870445344,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.7413967611336034e-05,
+      "loss": 2.6685,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5807186234817814,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.742155870445344e-05,
+      "loss": 2.7033,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5809716599190283,
+      "grad_norm": 19.875,
+      "learning_rate": 1.742914979757085e-05,
+      "loss": 2.7972,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5812246963562753,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.7436740890688258e-05,
+      "loss": 2.6014,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5814777327935222,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.7444331983805668e-05,
+      "loss": 2.7568,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5817307692307693,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.7451923076923078e-05,
+      "loss": 2.7854,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5819838056680162,
+      "grad_norm": 3.75,
+      "learning_rate": 1.7459514170040488e-05,
+      "loss": 2.6236,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5822368421052632,
+      "grad_norm": 3.625,
+      "learning_rate": 1.7467105263157895e-05,
+      "loss": 2.802,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5824898785425101,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.7474696356275305e-05,
+      "loss": 2.5859,
+      "step": 2302
+    },
+    {
+      "epoch": 0.582742914979757,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.748228744939271e-05,
+      "loss": 2.7995,
+      "step": 2303
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.7489878542510122e-05,
+      "loss": 2.7496,
+      "step": 2304
+    },
+    {
+      "epoch": 0.583248987854251,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.7497469635627532e-05,
+      "loss": 2.7133,
+      "step": 2305
+    },
+    {
+      "epoch": 0.583502024291498,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.7505060728744942e-05,
+      "loss": 2.6982,
+      "step": 2306
+    },
+    {
+      "epoch": 0.583755060728745,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.751265182186235e-05,
+      "loss": 2.7203,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5840080971659919,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.752024291497976e-05,
+      "loss": 2.6868,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5842611336032388,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.7527834008097166e-05,
+      "loss": 2.787,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5845141700404858,
+      "grad_norm": 2.875,
+      "learning_rate": 1.7535425101214572e-05,
+      "loss": 2.4712,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5845141700404858,
+      "eval_loss": 2.803760528564453,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.3867,
+      "eval_samples_per_second": 9.616,
+      "eval_steps_per_second": 9.616,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5847672064777328,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.7543016194331986e-05,
+      "loss": 2.7468,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5850202429149798,
+      "grad_norm": 3.5,
+      "learning_rate": 1.7550607287449392e-05,
+      "loss": 2.6524,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5852732793522267,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.7558198380566803e-05,
+      "loss": 2.803,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5855263157894737,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.756578947368421e-05,
+      "loss": 2.6927,
+      "step": 2314
+    },
+    {
+      "epoch": 0.5857793522267206,
+      "grad_norm": 3.25,
+      "learning_rate": 1.757338056680162e-05,
+      "loss": 2.6246,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5860323886639676,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.7580971659919026e-05,
+      "loss": 2.7251,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5862854251012146,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.758856275303644e-05,
+      "loss": 2.6965,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5865384615384616,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.7596153846153846e-05,
+      "loss": 2.7042,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5867914979757085,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.7603744939271257e-05,
+      "loss": 2.7918,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.7611336032388663e-05,
+      "loss": 2.8028,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5872975708502024,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.7618927125506073e-05,
+      "loss": 2.7631,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5875506072874493,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.762651821862348e-05,
+      "loss": 2.7109,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5878036437246964,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.7634109311740894e-05,
+      "loss": 2.6967,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5880566801619433,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.76417004048583e-05,
+      "loss": 2.5758,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5883097165991903,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.764929149797571e-05,
+      "loss": 2.7038,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5885627530364372,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.7656882591093117e-05,
+      "loss": 2.7392,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5888157894736842,
+      "grad_norm": 3.5,
+      "learning_rate": 1.7664473684210527e-05,
+      "loss": 2.6737,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5890688259109311,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.7672064777327934e-05,
+      "loss": 2.6791,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5893218623481782,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.7679655870445348e-05,
+      "loss": 2.7886,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5895748987854251,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.7687246963562754e-05,
+      "loss": 2.7114,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5898279352226721,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.769483805668016e-05,
+      "loss": 2.775,
+      "step": 2331
+    },
+    {
+      "epoch": 0.590080971659919,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.770242914979757e-05,
+      "loss": 2.7289,
+      "step": 2332
+    },
+    {
+      "epoch": 0.590334008097166,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.7710020242914978e-05,
+      "loss": 2.6843,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5905870445344129,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.7717611336032388e-05,
+      "loss": 2.7771,
+      "step": 2334
+    },
+    {
+      "epoch": 0.59084008097166,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.7725202429149798e-05,
+      "loss": 2.6002,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.7732793522267208e-05,
+      "loss": 2.8286,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5913461538461539,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.7740384615384615e-05,
+      "loss": 2.8244,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5915991902834008,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.7747975708502025e-05,
+      "loss": 2.7432,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5918522267206477,
+      "grad_norm": 3.125,
+      "learning_rate": 1.7755566801619432e-05,
+      "loss": 2.7504,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5921052631578947,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.7763157894736842e-05,
+      "loss": 2.64,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5921052631578947,
+      "eval_loss": 2.8002872467041016,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.4995,
+      "eval_samples_per_second": 9.648,
+      "eval_steps_per_second": 9.648,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5923582995951417,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.7770748987854252e-05,
+      "loss": 2.631,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5926113360323887,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.7778340080971662e-05,
+      "loss": 2.6851,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5928643724696356,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.778593117408907e-05,
+      "loss": 2.7618,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5931174089068826,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.779352226720648e-05,
+      "loss": 2.8008,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5933704453441295,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.7801113360323886e-05,
+      "loss": 2.7961,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5936234817813765,
+      "grad_norm": 3.375,
+      "learning_rate": 1.7808704453441296e-05,
+      "loss": 2.7697,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5938765182186235,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.7816295546558706e-05,
+      "loss": 2.7272,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5941295546558705,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.7823886639676116e-05,
+      "loss": 2.6584,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5943825910931174,
+      "grad_norm": 3.25,
+      "learning_rate": 1.7831477732793523e-05,
+      "loss": 2.6121,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5946356275303644,
+      "grad_norm": 3.25,
+      "learning_rate": 1.7839068825910933e-05,
+      "loss": 2.6719,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5948886639676113,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.784665991902834e-05,
+      "loss": 2.7496,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.7854251012145746e-05,
+      "loss": 2.8646,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5953947368421053,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.786184210526316e-05,
+      "loss": 2.7988,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5956477732793523,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.7869433198380567e-05,
+      "loss": 2.7369,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5959008097165992,
+      "grad_norm": 3.375,
+      "learning_rate": 1.7877024291497977e-05,
+      "loss": 2.6908,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5961538461538461,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.7884615384615384e-05,
+      "loss": 2.6893,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5964068825910931,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.7892206477732794e-05,
+      "loss": 2.7141,
+      "step": 2357
+    },
+    {
+      "epoch": 0.59665991902834,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.78997975708502e-05,
+      "loss": 2.6322,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5969129554655871,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.7907388663967614e-05,
+      "loss": 2.7891,
+      "step": 2359
+    },
+    {
+      "epoch": 0.597165991902834,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.791497975708502e-05,
+      "loss": 2.7066,
+      "step": 2360
+    },
+    {
+      "epoch": 0.597419028340081,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.792257085020243e-05,
+      "loss": 2.7309,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5976720647773279,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.7930161943319838e-05,
+      "loss": 2.7833,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5979251012145749,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.7937753036437248e-05,
+      "loss": 2.8182,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5981781376518218,
+      "grad_norm": 3.25,
+      "learning_rate": 1.7945344129554654e-05,
+      "loss": 2.6843,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5984311740890689,
+      "grad_norm": 3.375,
+      "learning_rate": 1.7952935222672068e-05,
+      "loss": 2.647,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5986842105263158,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.7960526315789475e-05,
+      "loss": 2.8126,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5989372469635628,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.7968117408906885e-05,
+      "loss": 2.6686,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.797570850202429e-05,
+      "loss": 2.6542,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5994433198380567,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.79832995951417e-05,
+      "loss": 2.6317,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5996963562753036,
+      "grad_norm": 3.75,
+      "learning_rate": 1.799089068825911e-05,
+      "loss": 2.7823,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5996963562753036,
+      "eval_loss": 2.800524950027466,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.4687,
+      "eval_samples_per_second": 9.613,
+      "eval_steps_per_second": 9.613,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5999493927125507,
+      "grad_norm": 3.5,
+      "learning_rate": 1.799848178137652e-05,
+      "loss": 2.74,
+      "step": 2371
+    },
+    {
+      "epoch": 0.6002024291497976,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.800607287449393e-05,
+      "loss": 2.7188,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6004554655870445,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.8013663967611335e-05,
+      "loss": 2.693,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6007085020242915,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.8021255060728745e-05,
+      "loss": 2.7309,
+      "step": 2374
+    },
+    {
+      "epoch": 0.6009615384615384,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.8028846153846152e-05,
+      "loss": 2.7572,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6012145748987854,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.8036437246963562e-05,
+      "loss": 2.6699,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6014676113360324,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.8044028340080972e-05,
+      "loss": 2.698,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6017206477732794,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.8051619433198382e-05,
+      "loss": 2.5747,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6019736842105263,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.805921052631579e-05,
+      "loss": 2.7919,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6022267206477733,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.80668016194332e-05,
+      "loss": 2.7579,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6024797570850202,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.8074392712550606e-05,
+      "loss": 2.6619,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6027327935222672,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.8081983805668016e-05,
+      "loss": 2.7772,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6029858299595142,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.8089574898785426e-05,
+      "loss": 2.5128,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.8097165991902836e-05,
+      "loss": 2.7817,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6034919028340081,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.8104757085020243e-05,
+      "loss": 2.6912,
+      "step": 2385
+    },
+    {
+      "epoch": 0.603744939271255,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.8112348178137653e-05,
+      "loss": 2.7133,
+      "step": 2386
+    },
+    {
+      "epoch": 0.603997975708502,
+      "grad_norm": 3.5,
+      "learning_rate": 1.811993927125506e-05,
+      "loss": 2.7825,
+      "step": 2387
+    },
+    {
+      "epoch": 0.604251012145749,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.812753036437247e-05,
+      "loss": 2.587,
+      "step": 2388
+    },
+    {
+      "epoch": 0.604504048582996,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.813512145748988e-05,
+      "loss": 2.5965,
+      "step": 2389
+    },
+    {
+      "epoch": 0.604757085020243,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.814271255060729e-05,
+      "loss": 2.6714,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6050101214574899,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.8150303643724697e-05,
+      "loss": 2.7213,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6052631578947368,
+      "grad_norm": 3.125,
+      "learning_rate": 1.8157894736842107e-05,
+      "loss": 2.631,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6055161943319838,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.8165485829959514e-05,
+      "loss": 2.7723,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6057692307692307,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.817307692307692e-05,
+      "loss": 2.6218,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6060222672064778,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.8180668016194334e-05,
+      "loss": 2.7066,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6062753036437247,
+      "grad_norm": 3.625,
+      "learning_rate": 1.818825910931174e-05,
+      "loss": 2.7827,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6065283400809717,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.819585020242915e-05,
+      "loss": 2.7559,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6067813765182186,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.8203441295546558e-05,
+      "loss": 2.7221,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6070344129554656,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.8211032388663968e-05,
+      "loss": 2.7097,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.8218623481781375e-05,
+      "loss": 2.7583,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "eval_loss": 2.7977170944213867,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.818,
+      "eval_samples_per_second": 9.601,
+      "eval_steps_per_second": 9.601,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6075404858299596,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.8226214574898788e-05,
+      "loss": 2.7525,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6077935222672065,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.8233805668016195e-05,
+      "loss": 2.7143,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6080465587044535,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.8241396761133605e-05,
+      "loss": 2.6177,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6082995951417004,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.8248987854251012e-05,
+      "loss": 2.7347,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6085526315789473,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.8256578947368422e-05,
+      "loss": 2.6489,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6088056680161943,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.826417004048583e-05,
+      "loss": 2.6984,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6090587044534413,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.8271761133603242e-05,
+      "loss": 2.7782,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6093117408906883,
+      "grad_norm": 3.375,
+      "learning_rate": 1.827935222672065e-05,
+      "loss": 2.6904,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6095647773279352,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.828694331983806e-05,
+      "loss": 2.814,
+      "step": 2409
+    },
+    {
+      "epoch": 0.6098178137651822,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.8294534412955466e-05,
+      "loss": 2.5486,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6100708502024291,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.8302125506072876e-05,
+      "loss": 2.7052,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6103238866396761,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.8309716599190283e-05,
+      "loss": 2.646,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6105769230769231,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.8317307692307693e-05,
+      "loss": 2.6064,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6108299595141701,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.8324898785425103e-05,
+      "loss": 2.5634,
+      "step": 2414
+    },
+    {
+      "epoch": 0.611082995951417,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.833248987854251e-05,
+      "loss": 2.7027,
+      "step": 2415
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.834008097165992e-05,
+      "loss": 2.7305,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6115890688259109,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.8347672064777326e-05,
+      "loss": 2.6501,
+      "step": 2417
+    },
+    {
+      "epoch": 0.6118421052631579,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.8355263157894736e-05,
+      "loss": 2.7252,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6120951417004049,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.8362854251012147e-05,
+      "loss": 2.6262,
+      "step": 2419
+    },
+    {
+      "epoch": 0.6123481781376519,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.8370445344129557e-05,
+      "loss": 2.6392,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6126012145748988,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.8378036437246963e-05,
+      "loss": 2.6298,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6128542510121457,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.8385627530364374e-05,
+      "loss": 2.6744,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6131072874493927,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.839321862348178e-05,
+      "loss": 2.777,
+      "step": 2423
+    },
+    {
+      "epoch": 0.6133603238866396,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.840080971659919e-05,
+      "loss": 2.7262,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6136133603238867,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.84084008097166e-05,
+      "loss": 2.7211,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6138663967611336,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.841599190283401e-05,
+      "loss": 2.6977,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6141194331983806,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.8423582995951417e-05,
+      "loss": 2.7392,
+      "step": 2427
+    },
+    {
+      "epoch": 0.6143724696356275,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.8431174089068828e-05,
+      "loss": 2.8151,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6146255060728745,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.8438765182186234e-05,
+      "loss": 2.7987,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6148785425101214,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.8446356275303644e-05,
+      "loss": 2.6767,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6148785425101214,
+      "eval_loss": 2.791733741760254,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 271.0019,
+      "eval_samples_per_second": 9.594,
+      "eval_steps_per_second": 9.594,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6151315789473685,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.8453947368421054e-05,
+      "loss": 2.6578,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 2.8046,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6156376518218624,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.846912955465587e-05,
+      "loss": 2.7235,
+      "step": 2433
+    },
+    {
+      "epoch": 0.6158906882591093,
+      "grad_norm": 3.25,
+      "learning_rate": 1.8476720647773278e-05,
+      "loss": 2.7565,
+      "step": 2434
+    },
+    {
+      "epoch": 0.6161437246963563,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.8484311740890688e-05,
+      "loss": 2.7876,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6163967611336032,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.8491902834008095e-05,
+      "loss": 2.6641,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6166497975708503,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.849949392712551e-05,
+      "loss": 2.6689,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6169028340080972,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.8507085020242915e-05,
+      "loss": 2.7437,
+      "step": 2438
+    },
+    {
+      "epoch": 0.6171558704453441,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.8514676113360325e-05,
+      "loss": 2.6882,
+      "step": 2439
+    },
+    {
+      "epoch": 0.6174089068825911,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.8522267206477732e-05,
+      "loss": 2.7074,
+      "step": 2440
+    },
+    {
+      "epoch": 0.617661943319838,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.8529858299595142e-05,
+      "loss": 2.7448,
+      "step": 2441
+    },
+    {
+      "epoch": 0.617914979757085,
+      "grad_norm": 3.75,
+      "learning_rate": 1.853744939271255e-05,
+      "loss": 2.8363,
+      "step": 2442
+    },
+    {
+      "epoch": 0.618168016194332,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.8545040485829962e-05,
+      "loss": 2.7897,
+      "step": 2443
+    },
+    {
+      "epoch": 0.618421052631579,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.855263157894737e-05,
+      "loss": 2.7273,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6186740890688259,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.856022267206478e-05,
+      "loss": 2.7695,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6189271255060729,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.8567813765182186e-05,
+      "loss": 2.678,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6191801619433198,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.8575404858299596e-05,
+      "loss": 2.6386,
+      "step": 2447
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.8582995951417003e-05,
+      "loss": 2.7061,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6196862348178138,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.8590587044534416e-05,
+      "loss": 2.7125,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6199392712550608,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.8598178137651823e-05,
+      "loss": 2.5551,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6201923076923077,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.8605769230769233e-05,
+      "loss": 2.7078,
+      "step": 2451
+    },
+    {
+      "epoch": 0.6204453441295547,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.861336032388664e-05,
+      "loss": 2.7192,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6206983805668016,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.862095141700405e-05,
+      "loss": 2.7192,
+      "step": 2453
+    },
+    {
+      "epoch": 0.6209514170040485,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.8628542510121457e-05,
+      "loss": 2.643,
+      "step": 2454
+    },
+    {
+      "epoch": 0.6212044534412956,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.8636133603238867e-05,
+      "loss": 2.6915,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6214574898785425,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.8643724696356277e-05,
+      "loss": 2.6879,
+      "step": 2456
+    },
+    {
+      "epoch": 0.6217105263157895,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.8651315789473684e-05,
+      "loss": 2.6758,
+      "step": 2457
+    },
+    {
+      "epoch": 0.6219635627530364,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.8658906882591094e-05,
+      "loss": 2.6493,
+      "step": 2458
+    },
+    {
+      "epoch": 0.6222165991902834,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.86664979757085e-05,
+      "loss": 2.7547,
+      "step": 2459
+    },
+    {
+      "epoch": 0.6224696356275303,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.867408906882591e-05,
+      "loss": 2.8063,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6224696356275303,
+      "eval_loss": 2.7824583053588867,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.4006,
+      "eval_samples_per_second": 9.615,
+      "eval_steps_per_second": 9.615,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6227226720647774,
+      "grad_norm": 4.0,
+      "learning_rate": 1.868168016194332e-05,
+      "loss": 2.6841,
+      "step": 2461
+    },
+    {
+      "epoch": 0.6229757085020243,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.868927125506073e-05,
+      "loss": 2.6806,
+      "step": 2462
+    },
+    {
+      "epoch": 0.6232287449392713,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.8696862348178138e-05,
+      "loss": 2.5602,
+      "step": 2463
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 3.375,
+      "learning_rate": 1.8704453441295548e-05,
+      "loss": 2.7092,
+      "step": 2464
+    },
+    {
+      "epoch": 0.6237348178137652,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.8712044534412955e-05,
+      "loss": 2.7351,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6239878542510121,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.8719635627530365e-05,
+      "loss": 2.6935,
+      "step": 2466
+    },
+    {
+      "epoch": 0.6242408906882592,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.8727226720647775e-05,
+      "loss": 2.7218,
+      "step": 2467
+    },
+    {
+      "epoch": 0.6244939271255061,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.8734817813765185e-05,
+      "loss": 2.7164,
+      "step": 2468
+    },
+    {
+      "epoch": 0.624746963562753,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.874240890688259e-05,
+      "loss": 2.7049,
+      "step": 2469
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 2.6651,
+      "step": 2470
+    },
+    {
+      "epoch": 0.625253036437247,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.875759109311741e-05,
+      "loss": 2.7113,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6255060728744939,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.876518218623482e-05,
+      "loss": 2.6729,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6257591093117408,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.8772773279352225e-05,
+      "loss": 2.7537,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6260121457489879,
+      "grad_norm": 3.375,
+      "learning_rate": 1.878036437246964e-05,
+      "loss": 2.6865,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6262651821862348,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.8787955465587046e-05,
+      "loss": 2.719,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6265182186234818,
+      "grad_norm": 2.875,
+      "learning_rate": 1.8795546558704452e-05,
+      "loss": 2.6417,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6267712550607287,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.8803137651821862e-05,
+      "loss": 2.7007,
+      "step": 2477
+    },
+    {
+      "epoch": 0.6270242914979757,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.881072874493927e-05,
+      "loss": 2.6647,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6272773279352226,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.881831983805668e-05,
+      "loss": 2.7162,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.882591093117409e-05,
+      "loss": 2.693,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6277834008097166,
+      "grad_norm": 3.375,
+      "learning_rate": 1.88335020242915e-05,
+      "loss": 2.7786,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6280364372469636,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.8841093117408906e-05,
+      "loss": 2.6728,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6282894736842105,
+      "grad_norm": 3.5,
+      "learning_rate": 1.8848684210526316e-05,
+      "loss": 2.7737,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6285425101214575,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.8856275303643723e-05,
+      "loss": 2.7166,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6287955465587044,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.8863866396761133e-05,
+      "loss": 2.6586,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6290485829959515,
+      "grad_norm": 3.25,
+      "learning_rate": 1.8871457489878543e-05,
+      "loss": 2.6989,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6293016194331984,
+      "grad_norm": 3.5,
+      "learning_rate": 1.8879048582995953e-05,
+      "loss": 2.7697,
+      "step": 2487
+    },
+    {
+      "epoch": 0.6295546558704453,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.888663967611336e-05,
+      "loss": 2.5894,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6298076923076923,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.889423076923077e-05,
+      "loss": 2.6725,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6300607287449392,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.8901821862348177e-05,
+      "loss": 2.645,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6300607287449392,
+      "eval_loss": 2.7853190898895264,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 269.7373,
+      "eval_samples_per_second": 9.639,
+      "eval_steps_per_second": 9.639,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6303137651821862,
+      "grad_norm": 3.5,
+      "learning_rate": 1.8909412955465587e-05,
+      "loss": 2.6821,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6305668016194332,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.8917004048582997e-05,
+      "loss": 2.7791,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6308198380566802,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.8924595141700407e-05,
+      "loss": 2.6197,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6310728744939271,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.8932186234817814e-05,
+      "loss": 2.6851,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6313259109311741,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.8939777327935224e-05,
+      "loss": 2.4624,
+      "step": 2495
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.894736842105263e-05,
+      "loss": 2.7872,
+      "step": 2496
+    },
+    {
+      "epoch": 0.631831983805668,
+      "grad_norm": 3.375,
+      "learning_rate": 1.8954959514170038e-05,
+      "loss": 2.7524,
+      "step": 2497
+    },
+    {
+      "epoch": 0.632085020242915,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.896255060728745e-05,
+      "loss": 2.696,
+      "step": 2498
+    },
+    {
+      "epoch": 0.632338056680162,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.8970141700404858e-05,
+      "loss": 2.7251,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6325910931174089,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.8977732793522268e-05,
+      "loss": 2.6873,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6328441295546559,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.8985323886639675e-05,
+      "loss": 2.693,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6330971659919028,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.8992914979757085e-05,
+      "loss": 2.6485,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6333502024291497,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.900050607287449e-05,
+      "loss": 2.6003,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6336032388663968,
+      "grad_norm": 3.25,
+      "learning_rate": 1.9008097165991905e-05,
+      "loss": 2.6491,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6338562753036437,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.9015688259109312e-05,
+      "loss": 2.8167,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6341093117408907,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.9023279352226722e-05,
+      "loss": 2.7026,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6343623481781376,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.903087044534413e-05,
+      "loss": 2.5662,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6346153846153846,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.903846153846154e-05,
+      "loss": 2.7748,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6348684210526315,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.9046052631578946e-05,
+      "loss": 2.6329,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6351214574898786,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.905364372469636e-05,
+      "loss": 2.7613,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6353744939271255,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.9061234817813766e-05,
+      "loss": 2.6565,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.9068825910931176e-05,
+      "loss": 2.7783,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6358805668016194,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.9076417004048583e-05,
+      "loss": 2.6875,
+      "step": 2513
+    },
+    {
+      "epoch": 0.6361336032388664,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.9084008097165993e-05,
+      "loss": 2.5635,
+      "step": 2514
+    },
+    {
+      "epoch": 0.6363866396761133,
+      "grad_norm": 3.90625,
+      "learning_rate": 1.90915991902834e-05,
+      "loss": 2.7697,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6366396761133604,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.9099190283400813e-05,
+      "loss": 2.7845,
+      "step": 2516
+    },
+    {
+      "epoch": 0.6368927125506073,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.910678137651822e-05,
+      "loss": 2.6654,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6371457489878543,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.9114372469635627e-05,
+      "loss": 2.7386,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6373987854251012,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.9121963562753037e-05,
+      "loss": 2.7484,
+      "step": 2519
+    },
+    {
+      "epoch": 0.6376518218623481,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.9129554655870443e-05,
+      "loss": 2.6883,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6376518218623481,
+      "eval_loss": 2.7815966606140137,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.336,
+      "eval_samples_per_second": 9.409,
+      "eval_steps_per_second": 9.409,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6379048582995951,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.9137145748987854e-05,
+      "loss": 2.6878,
+      "step": 2521
+    },
+    {
+      "epoch": 0.6381578947368421,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.9144736842105264e-05,
+      "loss": 2.7267,
+      "step": 2522
+    },
+    {
+      "epoch": 0.6384109311740891,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.9152327935222674e-05,
+      "loss": 2.6871,
+      "step": 2523
+    },
+    {
+      "epoch": 0.638663967611336,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.915991902834008e-05,
+      "loss": 2.7285,
+      "step": 2524
+    },
+    {
+      "epoch": 0.638917004048583,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.916751012145749e-05,
+      "loss": 2.7587,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6391700404858299,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.9175101214574897e-05,
+      "loss": 2.6619,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6394230769230769,
+      "grad_norm": 3.5,
+      "learning_rate": 1.9182692307692307e-05,
+      "loss": 2.6968,
+      "step": 2527
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 3.75,
+      "learning_rate": 1.9190283400809718e-05,
+      "loss": 2.7882,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6399291497975709,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.9197874493927128e-05,
+      "loss": 2.7327,
+      "step": 2529
+    },
+    {
+      "epoch": 0.6401821862348178,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.9205465587044534e-05,
+      "loss": 2.7337,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6404352226720648,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.9213056680161945e-05,
+      "loss": 2.7251,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6406882591093117,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.922064777327935e-05,
+      "loss": 2.6844,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6409412955465587,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.922823886639676e-05,
+      "loss": 2.7212,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6411943319838057,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.923582995951417e-05,
+      "loss": 2.5194,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6414473684210527,
+      "grad_norm": 3.84375,
+      "learning_rate": 1.924342105263158e-05,
+      "loss": 2.6839,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6417004048582996,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.925101214574899e-05,
+      "loss": 2.7261,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6419534412955465,
+      "grad_norm": 3.75,
+      "learning_rate": 1.92586032388664e-05,
+      "loss": 2.8029,
+      "step": 2537
+    },
+    {
+      "epoch": 0.6422064777327935,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.9266194331983805e-05,
+      "loss": 2.7179,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6424595141700404,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.9273785425101212e-05,
+      "loss": 2.6661,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6427125506072875,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.9281376518218625e-05,
+      "loss": 2.8468,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6429655870445344,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.9288967611336032e-05,
+      "loss": 2.6942,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6432186234817814,
+      "grad_norm": 3.75,
+      "learning_rate": 1.9296558704453442e-05,
+      "loss": 2.7215,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6434716599190283,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.930414979757085e-05,
+      "loss": 2.7931,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.931174089068826e-05,
+      "loss": 2.6542,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6439777327935222,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.9319331983805666e-05,
+      "loss": 2.7556,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6442307692307693,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.932692307692308e-05,
+      "loss": 2.6541,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6444838056680162,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.9334514170040486e-05,
+      "loss": 2.687,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6447368421052632,
+      "grad_norm": 3.375,
+      "learning_rate": 1.9342105263157896e-05,
+      "loss": 2.6333,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6449898785425101,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.9349696356275303e-05,
+      "loss": 2.7354,
+      "step": 2549
+    },
+    {
+      "epoch": 0.645242914979757,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.9357287449392713e-05,
+      "loss": 2.7433,
+      "step": 2550
+    },
+    {
+      "epoch": 0.645242914979757,
+      "eval_loss": 2.7797868251800537,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.5801,
+      "eval_samples_per_second": 9.538,
+      "eval_steps_per_second": 9.538,
+      "step": 2550
+    },
+    {
+      "epoch": 0.645495951417004,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.936487854251012e-05,
+      "loss": 2.6686,
+      "step": 2551
+    },
+    {
+      "epoch": 0.645748987854251,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.9372469635627533e-05,
+      "loss": 2.6723,
+      "step": 2552
+    },
+    {
+      "epoch": 0.646002024291498,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.938006072874494e-05,
+      "loss": 2.7732,
+      "step": 2553
+    },
+    {
+      "epoch": 0.646255060728745,
+      "grad_norm": 3.75,
+      "learning_rate": 1.938765182186235e-05,
+      "loss": 2.7697,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6465080971659919,
+      "grad_norm": 3.71875,
+      "learning_rate": 1.9395242914979757e-05,
+      "loss": 2.7898,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6467611336032388,
+      "grad_norm": 3.5,
+      "learning_rate": 1.9402834008097167e-05,
+      "loss": 2.616,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6470141700404858,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.9410425101214574e-05,
+      "loss": 2.5594,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6472672064777328,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.9418016194331987e-05,
+      "loss": 2.645,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6475202429149798,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.9425607287449394e-05,
+      "loss": 2.7707,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.94331983805668e-05,
+      "loss": 2.665,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6480263157894737,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.944078947368421e-05,
+      "loss": 2.7754,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6482793522267206,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.9448380566801618e-05,
+      "loss": 2.5732,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6485323886639676,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.9455971659919028e-05,
+      "loss": 2.6646,
+      "step": 2563
+    },
+    {
+      "epoch": 0.6487854251012146,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.9463562753036438e-05,
+      "loss": 2.8008,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6490384615384616,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.9471153846153848e-05,
+      "loss": 2.6927,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6492914979757085,
+      "grad_norm": 3.5,
+      "learning_rate": 1.9478744939271255e-05,
+      "loss": 2.7421,
+      "step": 2566
+    },
+    {
+      "epoch": 0.6495445344129555,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.9486336032388665e-05,
+      "loss": 2.8153,
+      "step": 2567
+    },
+    {
+      "epoch": 0.6497975708502024,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.949392712550607e-05,
+      "loss": 2.6461,
+      "step": 2568
+    },
+    {
+      "epoch": 0.6500506072874493,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.950151821862348e-05,
+      "loss": 2.648,
+      "step": 2569
+    },
+    {
+      "epoch": 0.6503036437246964,
+      "grad_norm": 3.75,
+      "learning_rate": 1.9509109311740892e-05,
+      "loss": 2.7606,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6505566801619433,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.9516700404858302e-05,
+      "loss": 2.7942,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6508097165991903,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.952429149797571e-05,
+      "loss": 2.7476,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6510627530364372,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.953188259109312e-05,
+      "loss": 2.8211,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6513157894736842,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.9539473684210525e-05,
+      "loss": 2.6356,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6515688259109311,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.9547064777327936e-05,
+      "loss": 2.7167,
+      "step": 2575
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.9554655870445346e-05,
+      "loss": 2.609,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6520748987854251,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.9562246963562756e-05,
+      "loss": 2.6392,
+      "step": 2577
+    },
+    {
+      "epoch": 0.6523279352226721,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.9569838056680163e-05,
+      "loss": 2.6849,
+      "step": 2578
+    },
+    {
+      "epoch": 0.652580971659919,
+      "grad_norm": 3.875,
+      "learning_rate": 1.9577429149797573e-05,
+      "loss": 2.7943,
+      "step": 2579
+    },
+    {
+      "epoch": 0.652834008097166,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.958502024291498e-05,
+      "loss": 2.6856,
+      "step": 2580
+    },
+    {
+      "epoch": 0.652834008097166,
+      "eval_loss": 2.7746686935424805,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 270.5224,
+      "eval_samples_per_second": 9.611,
+      "eval_steps_per_second": 9.611,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6530870445344129,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.9592611336032386e-05,
+      "loss": 2.7232,
+      "step": 2581
+    },
+    {
+      "epoch": 0.65334008097166,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.96002024291498e-05,
+      "loss": 2.7368,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6535931174089069,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.9607793522267206e-05,
+      "loss": 2.7535,
+      "step": 2583
+    },
+    {
+      "epoch": 0.6538461538461539,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.9615384615384617e-05,
+      "loss": 2.5894,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6540991902834008,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.9622975708502023e-05,
+      "loss": 2.7252,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6543522267206477,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.9630566801619433e-05,
+      "loss": 2.6281,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6546052631578947,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.963815789473684e-05,
+      "loss": 2.6577,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6548582995951417,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.9645748987854254e-05,
+      "loss": 2.7819,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6551113360323887,
+      "grad_norm": 3.375,
+      "learning_rate": 1.965334008097166e-05,
+      "loss": 2.7533,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6553643724696356,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.966093117408907e-05,
+      "loss": 2.6417,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6556174089068826,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.9668522267206477e-05,
+      "loss": 2.7211,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.9676113360323887e-05,
+      "loss": 2.604,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6561234817813765,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.9683704453441294e-05,
+      "loss": 2.6935,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6563765182186235,
+      "grad_norm": 3.25,
+      "learning_rate": 1.9691295546558708e-05,
+      "loss": 2.7228,
+      "step": 2594
+    },
+    {
+      "epoch": 0.6566295546558705,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.9698886639676114e-05,
+      "loss": 2.7768,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6568825910931174,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.9706477732793524e-05,
+      "loss": 2.5887,
+      "step": 2596
+    },
+    {
+      "epoch": 0.6571356275303644,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.971406882591093e-05,
+      "loss": 2.6935,
+      "step": 2597
+    },
+    {
+      "epoch": 0.6573886639676113,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.972165991902834e-05,
+      "loss": 2.7656,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6576417004048583,
+      "grad_norm": 3.125,
+      "learning_rate": 1.9729251012145748e-05,
+      "loss": 2.4758,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.9736842105263158e-05,
+      "loss": 2.7172,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6581477732793523,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.9744433198380568e-05,
+      "loss": 2.7701,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6584008097165992,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.9752024291497975e-05,
+      "loss": 2.7345,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6586538461538461,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.9759615384615385e-05,
+      "loss": 2.6908,
+      "step": 2603
+    },
+    {
+      "epoch": 0.6589068825910931,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.9767206477732792e-05,
+      "loss": 2.672,
+      "step": 2604
+    },
+    {
+      "epoch": 0.65915991902834,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.9774797570850202e-05,
+      "loss": 2.8723,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6594129554655871,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.9782388663967612e-05,
+      "loss": 2.6126,
+      "step": 2606
+    },
+    {
+      "epoch": 0.659665991902834,
+      "grad_norm": 3.25,
+      "learning_rate": 1.9789979757085022e-05,
+      "loss": 2.6174,
+      "step": 2607
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.979757085020243e-05,
+      "loss": 2.6174,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6601720647773279,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.980516194331984e-05,
+      "loss": 2.7482,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6604251012145749,
+      "grad_norm": 3.25,
+      "learning_rate": 1.9812753036437246e-05,
+      "loss": 2.6905,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6604251012145749,
+      "eval_loss": 2.771817445755005,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 271.4144,
+      "eval_samples_per_second": 9.579,
+      "eval_steps_per_second": 9.579,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6606781376518218,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.9820344129554656e-05,
+      "loss": 2.6489,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6609311740890689,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.9827935222672066e-05,
+      "loss": 2.7442,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6611842105263158,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.9835526315789476e-05,
+      "loss": 2.668,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6614372469635628,
+      "grad_norm": 3.25,
+      "learning_rate": 1.9843117408906883e-05,
+      "loss": 2.6717,
+      "step": 2614
+    },
+    {
+      "epoch": 0.6616902834008097,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.9850708502024293e-05,
+      "loss": 2.7673,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6619433198380567,
+      "grad_norm": 3.25,
+      "learning_rate": 1.98582995951417e-05,
+      "loss": 2.618,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6621963562753036,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.986589068825911e-05,
+      "loss": 2.7673,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6624493927125507,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.987348178137652e-05,
+      "loss": 2.6533,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6627024291497976,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.988107287449393e-05,
+      "loss": 2.6555,
+      "step": 2619
+    },
+    {
+      "epoch": 0.6629554655870445,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.9888663967611337e-05,
+      "loss": 2.5897,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6632085020242915,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.9896255060728744e-05,
+      "loss": 2.8354,
+      "step": 2621
+    },
+    {
+      "epoch": 0.6634615384615384,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.9903846153846154e-05,
+      "loss": 2.6845,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6637145748987854,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.991143724696356e-05,
+      "loss": 2.7531,
+      "step": 2623
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.9919028340080974e-05,
+      "loss": 2.7067,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6642206477732794,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.992661943319838e-05,
+      "loss": 2.769,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6644736842105263,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.993421052631579e-05,
+      "loss": 2.6759,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6647267206477733,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.9941801619433197e-05,
+      "loss": 2.6675,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6649797570850202,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.9949392712550608e-05,
+      "loss": 2.6762,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6652327935222672,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.9956983805668014e-05,
+      "loss": 2.6862,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6654858299595142,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.9964574898785428e-05,
+      "loss": 2.812,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6657388663967612,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.9972165991902835e-05,
+      "loss": 2.6683,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6659919028340081,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.9979757085020245e-05,
+      "loss": 2.7017,
+      "step": 2632
+    },
+    {
+      "epoch": 0.666244939271255,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.998734817813765e-05,
+      "loss": 2.6968,
+      "step": 2633
+    },
+    {
+      "epoch": 0.666497975708502,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.999493927125506e-05,
+      "loss": 2.7024,
+      "step": 2634
+    },
+    {
+      "epoch": 0.666751012145749,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.0002530364372468e-05,
+      "loss": 2.7638,
+      "step": 2635
+    },
+    {
+      "epoch": 0.667004048582996,
+      "grad_norm": 3.375,
+      "learning_rate": 2.0010121457489882e-05,
+      "loss": 2.6031,
+      "step": 2636
+    },
+    {
+      "epoch": 0.667257085020243,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.001771255060729e-05,
+      "loss": 2.6902,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6675101214574899,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.00253036437247e-05,
+      "loss": 2.6246,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6677631578947368,
+      "grad_norm": 3.25,
+      "learning_rate": 2.0032894736842105e-05,
+      "loss": 2.707,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.0040485829959516e-05,
+      "loss": 2.8292,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "eval_loss": 2.7737364768981934,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 271.6894,
+      "eval_samples_per_second": 9.57,
+      "eval_steps_per_second": 9.57,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6682692307692307,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.0048076923076922e-05,
+      "loss": 2.6753,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6685222672064778,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.0055668016194332e-05,
+      "loss": 2.6742,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6687753036437247,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.0063259109311742e-05,
+      "loss": 2.5727,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6690283400809717,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.007085020242915e-05,
+      "loss": 2.4476,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6692813765182186,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.007844129554656e-05,
+      "loss": 2.7737,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6695344129554656,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.0086032388663966e-05,
+      "loss": 2.603,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6697874493927125,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.0093623481781376e-05,
+      "loss": 2.5815,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6700404858299596,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.0101214574898786e-05,
+      "loss": 2.7284,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6702935222672065,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.0108805668016196e-05,
+      "loss": 2.7486,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6705465587044535,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.0116396761133603e-05,
+      "loss": 2.7009,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6707995951417004,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.0123987854251013e-05,
+      "loss": 2.7148,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6710526315789473,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.013157894736842e-05,
+      "loss": 2.7262,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6713056680161943,
+      "grad_norm": 3.375,
+      "learning_rate": 2.013917004048583e-05,
+      "loss": 2.6047,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6715587044534413,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.014676113360324e-05,
+      "loss": 2.8203,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6718117408906883,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.015435222672065e-05,
+      "loss": 2.5905,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.0161943319838057e-05,
+      "loss": 2.5538,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6723178137651822,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.0169534412955467e-05,
+      "loss": 2.6577,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6725708502024291,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.0177125506072874e-05,
+      "loss": 2.7618,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6728238866396761,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.0184716599190284e-05,
+      "loss": 2.7374,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6730769230769231,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.0192307692307694e-05,
+      "loss": 2.5723,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6733299595141701,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.0199898785425104e-05,
+      "loss": 2.7015,
+      "step": 2661
+    },
+    {
+      "epoch": 0.673582995951417,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.020748987854251e-05,
+      "loss": 2.6137,
+      "step": 2662
+    },
+    {
+      "epoch": 0.673836032388664,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.0215080971659918e-05,
+      "loss": 2.7264,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6740890688259109,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.0222672064777328e-05,
+      "loss": 2.6576,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6743421052631579,
+      "grad_norm": 3.375,
+      "learning_rate": 2.0230263157894735e-05,
+      "loss": 2.7774,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6745951417004049,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.0237854251012148e-05,
+      "loss": 2.6908,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6748481781376519,
+      "grad_norm": 3.375,
+      "learning_rate": 2.0245445344129555e-05,
+      "loss": 2.6673,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6751012145748988,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.0253036437246965e-05,
+      "loss": 2.7336,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6753542510121457,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.0260627530364372e-05,
+      "loss": 2.7329,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6756072874493927,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.0268218623481782e-05,
+      "loss": 2.733,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6756072874493927,
+      "eval_loss": 2.76603102684021,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.1128,
+      "eval_samples_per_second": 9.555,
+      "eval_steps_per_second": 9.555,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6758603238866396,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.027580971659919e-05,
+      "loss": 2.6984,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.0283400809716602e-05,
+      "loss": 2.7178,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6763663967611336,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.029099190283401e-05,
+      "loss": 2.5938,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6766194331983806,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.029858299595142e-05,
+      "loss": 2.5594,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6768724696356275,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.0306174089068826e-05,
+      "loss": 2.6079,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6771255060728745,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.0313765182186236e-05,
+      "loss": 2.7015,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6773785425101214,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.0321356275303643e-05,
+      "loss": 2.68,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6776315789473685,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.0328947368421056e-05,
+      "loss": 2.6071,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6778846153846154,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.0336538461538463e-05,
+      "loss": 2.5496,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6781376518218624,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.0344129554655873e-05,
+      "loss": 2.6091,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6783906882591093,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.035172064777328e-05,
+      "loss": 2.6298,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6786437246963563,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.035931174089069e-05,
+      "loss": 2.6278,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6788967611336032,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.0366902834008096e-05,
+      "loss": 2.7161,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6791497975708503,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.0374493927125507e-05,
+      "loss": 2.55,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6794028340080972,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.0382085020242917e-05,
+      "loss": 2.7116,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6796558704453441,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.0389676113360323e-05,
+      "loss": 2.7439,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6799089068825911,
+      "grad_norm": 3.5,
+      "learning_rate": 2.0397267206477734e-05,
+      "loss": 2.7164,
+      "step": 2687
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.040485829959514e-05,
+      "loss": 2.5955,
+      "step": 2688
+    },
+    {
+      "epoch": 0.680414979757085,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.041244939271255e-05,
+      "loss": 2.6129,
+      "step": 2689
+    },
+    {
+      "epoch": 0.680668016194332,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.042004048582996e-05,
+      "loss": 2.5393,
+      "step": 2690
+    },
+    {
+      "epoch": 0.680921052631579,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.042763157894737e-05,
+      "loss": 2.7037,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6811740890688259,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.0435222672064777e-05,
+      "loss": 2.7189,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6814271255060729,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.0442813765182187e-05,
+      "loss": 2.7368,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6816801619433198,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.0450404858299594e-05,
+      "loss": 2.6509,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6819331983805668,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.0457995951417004e-05,
+      "loss": 2.6683,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6821862348178138,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.0465587044534414e-05,
+      "loss": 2.6913,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6824392712550608,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.0473178137651825e-05,
+      "loss": 2.7446,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6826923076923077,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.048076923076923e-05,
+      "loss": 2.7395,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6829453441295547,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.048836032388664e-05,
+      "loss": 2.6669,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6831983805668016,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.0495951417004048e-05,
+      "loss": 2.7617,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6831983805668016,
+      "eval_loss": 2.7627296447753906,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.9286,
+      "eval_samples_per_second": 9.526,
+      "eval_steps_per_second": 9.526,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6834514170040485,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.0503542510121458e-05,
+      "loss": 2.7221,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6837044534412956,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.051113360323887e-05,
+      "loss": 2.5682,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6839574898785425,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.051872469635628e-05,
+      "loss": 2.6126,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.0526315789473685e-05,
+      "loss": 2.7101,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6844635627530364,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.0533906882591092e-05,
+      "loss": 2.6584,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6847165991902834,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.0541497975708502e-05,
+      "loss": 2.5904,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6849696356275303,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.054908906882591e-05,
+      "loss": 2.6287,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6852226720647774,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.0556680161943322e-05,
+      "loss": 2.742,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6854757085020243,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.056427125506073e-05,
+      "loss": 2.6108,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6857287449392713,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.057186234817814e-05,
+      "loss": 2.7564,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6859817813765182,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.0579453441295546e-05,
+      "loss": 2.5756,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6862348178137652,
+      "grad_norm": 3.5,
+      "learning_rate": 2.0587044534412956e-05,
+      "loss": 2.7171,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6864878542510121,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.0594635627530363e-05,
+      "loss": 2.6815,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6867408906882592,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.0602226720647776e-05,
+      "loss": 2.7384,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6869939271255061,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.0609817813765183e-05,
+      "loss": 2.6803,
+      "step": 2715
+    },
+    {
+      "epoch": 0.687246963562753,
+      "grad_norm": 3.5,
+      "learning_rate": 2.0617408906882593e-05,
+      "loss": 2.6391,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.0625e-05,
+      "loss": 2.734,
+      "step": 2717
+    },
+    {
+      "epoch": 0.687753036437247,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.063259109311741e-05,
+      "loss": 2.7375,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6880060728744939,
+      "grad_norm": 3.375,
+      "learning_rate": 2.0640182186234817e-05,
+      "loss": 2.7372,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 3.25,
+      "learning_rate": 2.0647773279352227e-05,
+      "loss": 2.7131,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6885121457489879,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.0655364372469637e-05,
+      "loss": 2.7336,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6887651821862348,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.0662955465587047e-05,
+      "loss": 2.7719,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6890182186234818,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.0670546558704454e-05,
+      "loss": 2.7418,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6892712550607287,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.0678137651821864e-05,
+      "loss": 2.6381,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6895242914979757,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.068572874493927e-05,
+      "loss": 2.6754,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6897773279352226,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.0693319838056677e-05,
+      "loss": 2.6671,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6900303643724697,
+      "grad_norm": 3.25,
+      "learning_rate": 2.070091093117409e-05,
+      "loss": 2.6093,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6902834008097166,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.0708502024291498e-05,
+      "loss": 2.6927,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6905364372469636,
+      "grad_norm": 3.25,
+      "learning_rate": 2.0716093117408908e-05,
+      "loss": 2.6803,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6907894736842105,
+      "grad_norm": 3.25,
+      "learning_rate": 2.0723684210526315e-05,
+      "loss": 2.7233,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6907894736842105,
+      "eval_loss": 2.7633814811706543,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 271.6837,
+      "eval_samples_per_second": 9.57,
+      "eval_steps_per_second": 9.57,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6910425101214575,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.0731275303643725e-05,
+      "loss": 2.7106,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6912955465587044,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.073886639676113e-05,
+      "loss": 2.7097,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6915485829959515,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.0746457489878545e-05,
+      "loss": 2.713,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6918016194331984,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.075404858299595e-05,
+      "loss": 2.6902,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6920546558704453,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.0761639676113362e-05,
+      "loss": 2.7034,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.076923076923077e-05,
+      "loss": 2.6836,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6925607287449392,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.077682186234818e-05,
+      "loss": 2.7314,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6928137651821862,
+      "grad_norm": 3.5,
+      "learning_rate": 2.0784412955465585e-05,
+      "loss": 2.735,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6930668016194332,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.0792004048583e-05,
+      "loss": 2.5949,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6933198380566802,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.0799595141700406e-05,
+      "loss": 2.6709,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6935728744939271,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.0807186234817816e-05,
+      "loss": 2.6965,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6938259109311741,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.0814777327935222e-05,
+      "loss": 2.7467,
+      "step": 2742
+    },
+    {
+      "epoch": 0.694078947368421,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.0822368421052633e-05,
+      "loss": 2.6937,
+      "step": 2743
+    },
+    {
+      "epoch": 0.694331983805668,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.082995951417004e-05,
+      "loss": 2.7164,
+      "step": 2744
+    },
+    {
+      "epoch": 0.694585020242915,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.0837550607287453e-05,
+      "loss": 2.7152,
+      "step": 2745
+    },
+    {
+      "epoch": 0.694838056680162,
+      "grad_norm": 3.5,
+      "learning_rate": 2.084514170040486e-05,
+      "loss": 2.9045,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6950910931174089,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.0852732793522266e-05,
+      "loss": 2.7219,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6953441295546559,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.0860323886639676e-05,
+      "loss": 2.6749,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6955971659919028,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.0867914979757083e-05,
+      "loss": 2.5928,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6958502024291497,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.0875506072874493e-05,
+      "loss": 2.5565,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6961032388663968,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.0883097165991903e-05,
+      "loss": 2.7026,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.0890688259109313e-05,
+      "loss": 2.7672,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6966093117408907,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.089827935222672e-05,
+      "loss": 2.7344,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6968623481781376,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.090587044534413e-05,
+      "loss": 2.6969,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6971153846153846,
+      "grad_norm": 3.5,
+      "learning_rate": 2.0913461538461537e-05,
+      "loss": 2.7518,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6973684210526315,
+      "grad_norm": 3.75,
+      "learning_rate": 2.0921052631578947e-05,
+      "loss": 2.7275,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6976214574898786,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.0928643724696357e-05,
+      "loss": 2.7109,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6978744939271255,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.0936234817813767e-05,
+      "loss": 2.546,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6981275303643725,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.0943825910931174e-05,
+      "loss": 2.6653,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6983805668016194,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.0951417004048584e-05,
+      "loss": 2.5624,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6983805668016194,
+      "eval_loss": 2.7570416927337646,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.1047,
+      "eval_samples_per_second": 9.52,
+      "eval_steps_per_second": 9.52,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6986336032388664,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.095900809716599e-05,
+      "loss": 2.6843,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6988866396761133,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.09665991902834e-05,
+      "loss": 2.6163,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6991396761133604,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.097419028340081e-05,
+      "loss": 2.7182,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6993927125506073,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.098178137651822e-05,
+      "loss": 2.6765,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6996457489878543,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.0989372469635628e-05,
+      "loss": 2.5953,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6998987854251012,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.0996963562753038e-05,
+      "loss": 2.8072,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7001518218623481,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.1004554655870445e-05,
+      "loss": 2.6903,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.101214574898785e-05,
+      "loss": 2.6236,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7006578947368421,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.1019736842105265e-05,
+      "loss": 2.7476,
+      "step": 2769
+    },
+    {
+      "epoch": 0.7009109311740891,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.1027327935222672e-05,
+      "loss": 2.5206,
+      "step": 2770
+    },
+    {
+      "epoch": 0.701163967611336,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.1034919028340082e-05,
+      "loss": 2.7132,
+      "step": 2771
+    },
+    {
+      "epoch": 0.701417004048583,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.104251012145749e-05,
+      "loss": 2.7719,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7016700404858299,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.10501012145749e-05,
+      "loss": 2.5694,
+      "step": 2773
+    },
+    {
+      "epoch": 0.7019230769230769,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.1057692307692306e-05,
+      "loss": 2.6988,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7021761133603239,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.106528340080972e-05,
+      "loss": 2.763,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7024291497975709,
+      "grad_norm": 3.5,
+      "learning_rate": 2.1072874493927126e-05,
+      "loss": 2.7688,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7026821862348178,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.1080465587044536e-05,
+      "loss": 2.7775,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7029352226720648,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.1088056680161943e-05,
+      "loss": 2.7603,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7031882591093117,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.1095647773279353e-05,
+      "loss": 2.6624,
+      "step": 2779
+    },
+    {
+      "epoch": 0.7034412955465587,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.110323886639676e-05,
+      "loss": 2.6676,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7036943319838057,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.1110829959514173e-05,
+      "loss": 2.7025,
+      "step": 2781
+    },
+    {
+      "epoch": 0.7039473684210527,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.111842105263158e-05,
+      "loss": 2.7127,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7042004048582996,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.112601214574899e-05,
+      "loss": 2.7786,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.1133603238866397e-05,
+      "loss": 2.7241,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7047064777327935,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.1141194331983807e-05,
+      "loss": 2.5566,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7049595141700404,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.1148785425101213e-05,
+      "loss": 2.7343,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7052125506072875,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.1156376518218624e-05,
+      "loss": 2.7055,
+      "step": 2787
+    },
+    {
+      "epoch": 0.7054655870445344,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.1163967611336034e-05,
+      "loss": 2.7302,
+      "step": 2788
+    },
+    {
+      "epoch": 0.7057186234817814,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.117155870445344e-05,
+      "loss": 2.7486,
+      "step": 2789
+    },
+    {
+      "epoch": 0.7059716599190283,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.117914979757085e-05,
+      "loss": 2.5934,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7059716599190283,
+      "eval_loss": 2.7598583698272705,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.6078,
+      "eval_samples_per_second": 9.503,
+      "eval_steps_per_second": 9.503,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7062246963562753,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.1186740890688257e-05,
+      "loss": 2.6817,
+      "step": 2791
+    },
+    {
+      "epoch": 0.7064777327935222,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.1194331983805667e-05,
+      "loss": 2.6879,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7067307692307693,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.1201923076923078e-05,
+      "loss": 2.6132,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7069838056680162,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.1209514170040488e-05,
+      "loss": 2.6573,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7072368421052632,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.1217105263157894e-05,
+      "loss": 2.7212,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7074898785425101,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.1224696356275305e-05,
+      "loss": 2.7364,
+      "step": 2796
+    },
+    {
+      "epoch": 0.707742914979757,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.123228744939271e-05,
+      "loss": 2.6262,
+      "step": 2797
+    },
+    {
+      "epoch": 0.707995951417004,
+      "grad_norm": 3.375,
+      "learning_rate": 2.123987854251012e-05,
+      "loss": 2.691,
+      "step": 2798
+    },
+    {
+      "epoch": 0.708248987854251,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.124746963562753e-05,
+      "loss": 2.6557,
+      "step": 2799
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 3.375,
+      "learning_rate": 2.125506072874494e-05,
+      "loss": 2.7539,
+      "step": 2800
+    },
+    {
+      "epoch": 0.708755060728745,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.126265182186235e-05,
+      "loss": 2.7606,
+      "step": 2801
+    },
+    {
+      "epoch": 0.7090080971659919,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.127024291497976e-05,
+      "loss": 2.7671,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7092611336032388,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.1277834008097165e-05,
+      "loss": 2.6724,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7095141700404858,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.1285425101214575e-05,
+      "loss": 2.7362,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7097672064777328,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.1293016194331985e-05,
+      "loss": 2.6364,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7100202429149798,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.1300607287449396e-05,
+      "loss": 2.7417,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7102732793522267,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.1308198380566802e-05,
+      "loss": 2.6849,
+      "step": 2807
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.1315789473684212e-05,
+      "loss": 2.6543,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7107793522267206,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.132338056680162e-05,
+      "loss": 2.7323,
+      "step": 2809
+    },
+    {
+      "epoch": 0.7110323886639676,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.1330971659919026e-05,
+      "loss": 2.8224,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7112854251012146,
+      "grad_norm": 3.375,
+      "learning_rate": 2.133856275303644e-05,
+      "loss": 2.7535,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7115384615384616,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.1346153846153846e-05,
+      "loss": 2.5543,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7117914979757085,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.1353744939271256e-05,
+      "loss": 2.7055,
+      "step": 2813
+    },
+    {
+      "epoch": 0.7120445344129555,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.1361336032388663e-05,
+      "loss": 2.7704,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7122975708502024,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.1368927125506073e-05,
+      "loss": 2.6683,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.137651821862348e-05,
+      "loss": 2.691,
+      "step": 2816
+    },
+    {
+      "epoch": 0.7128036437246964,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.1384109311740893e-05,
+      "loss": 2.5891,
+      "step": 2817
+    },
+    {
+      "epoch": 0.7130566801619433,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.13917004048583e-05,
+      "loss": 2.7275,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7133097165991903,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.139929149797571e-05,
+      "loss": 2.7564,
+      "step": 2819
+    },
+    {
+      "epoch": 0.7135627530364372,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.1406882591093117e-05,
+      "loss": 2.6291,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7135627530364372,
+      "eval_loss": 2.75575852394104,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.1361,
+      "eval_samples_per_second": 9.484,
+      "eval_steps_per_second": 9.484,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7138157894736842,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.1414473684210527e-05,
+      "loss": 2.6431,
+      "step": 2821
+    },
+    {
+      "epoch": 0.7140688259109311,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.1422064777327934e-05,
+      "loss": 2.7668,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7143218623481782,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.1429655870445347e-05,
+      "loss": 2.6307,
+      "step": 2823
+    },
+    {
+      "epoch": 0.7145748987854251,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.1437246963562754e-05,
+      "loss": 2.6048,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7148279352226721,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.1444838056680164e-05,
+      "loss": 2.7548,
+      "step": 2825
+    },
+    {
+      "epoch": 0.715080971659919,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.145242914979757e-05,
+      "loss": 2.6745,
+      "step": 2826
+    },
+    {
+      "epoch": 0.715334008097166,
+      "grad_norm": 4.3125,
+      "learning_rate": 2.146002024291498e-05,
+      "loss": 2.8345,
+      "step": 2827
+    },
+    {
+      "epoch": 0.7155870445344129,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.1467611336032388e-05,
+      "loss": 2.6478,
+      "step": 2828
+    },
+    {
+      "epoch": 0.71584008097166,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.1475202429149798e-05,
+      "loss": 2.594,
+      "step": 2829
+    },
+    {
+      "epoch": 0.7160931174089069,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.1482793522267208e-05,
+      "loss": 2.758,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7163461538461539,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.1490384615384615e-05,
+      "loss": 2.764,
+      "step": 2831
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.1497975708502025e-05,
+      "loss": 2.4837,
+      "step": 2832
+    },
+    {
+      "epoch": 0.7168522267206477,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.150556680161943e-05,
+      "loss": 2.5991,
+      "step": 2833
+    },
+    {
+      "epoch": 0.7171052631578947,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.151315789473684e-05,
+      "loss": 2.5935,
+      "step": 2834
+    },
+    {
+      "epoch": 0.7173582995951417,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.1520748987854252e-05,
+      "loss": 2.7424,
+      "step": 2835
+    },
+    {
+      "epoch": 0.7176113360323887,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.1528340080971662e-05,
+      "loss": 2.7154,
+      "step": 2836
+    },
+    {
+      "epoch": 0.7178643724696356,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.153593117408907e-05,
+      "loss": 2.6973,
+      "step": 2837
+    },
+    {
+      "epoch": 0.7181174089068826,
+      "grad_norm": 3.25,
+      "learning_rate": 2.154352226720648e-05,
+      "loss": 2.5021,
+      "step": 2838
+    },
+    {
+      "epoch": 0.7183704453441295,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.1551113360323885e-05,
+      "loss": 2.6294,
+      "step": 2839
+    },
+    {
+      "epoch": 0.7186234817813765,
+      "grad_norm": 3.25,
+      "learning_rate": 2.1558704453441296e-05,
+      "loss": 2.61,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7188765182186235,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.1566295546558706e-05,
+      "loss": 2.656,
+      "step": 2841
+    },
+    {
+      "epoch": 0.7191295546558705,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.1573886639676116e-05,
+      "loss": 2.6275,
+      "step": 2842
+    },
+    {
+      "epoch": 0.7193825910931174,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.1581477732793523e-05,
+      "loss": 2.6283,
+      "step": 2843
+    },
+    {
+      "epoch": 0.7196356275303644,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.1589068825910933e-05,
+      "loss": 2.565,
+      "step": 2844
+    },
+    {
+      "epoch": 0.7198886639676113,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.159665991902834e-05,
+      "loss": 2.7348,
+      "step": 2845
+    },
+    {
+      "epoch": 0.7201417004048583,
+      "grad_norm": 3.5,
+      "learning_rate": 2.160425101214575e-05,
+      "loss": 2.723,
+      "step": 2846
+    },
+    {
+      "epoch": 0.7203947368421053,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.161184210526316e-05,
+      "loss": 2.7373,
+      "step": 2847
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.161943319838057e-05,
+      "loss": 2.6842,
+      "step": 2848
+    },
+    {
+      "epoch": 0.7209008097165992,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.1627024291497977e-05,
+      "loss": 2.7039,
+      "step": 2849
+    },
+    {
+      "epoch": 0.7211538461538461,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1634615384615383e-05,
+      "loss": 2.7244,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7211538461538461,
+      "eval_loss": 2.7532737255096436,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.2828,
+      "eval_samples_per_second": 9.549,
+      "eval_steps_per_second": 9.549,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7214068825910931,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.1642206477732793e-05,
+      "loss": 2.6509,
+      "step": 2851
+    },
+    {
+      "epoch": 0.72165991902834,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.16497975708502e-05,
+      "loss": 2.6735,
+      "step": 2852
+    },
+    {
+      "epoch": 0.7219129554655871,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.1657388663967614e-05,
+      "loss": 2.6871,
+      "step": 2853
+    },
+    {
+      "epoch": 0.722165991902834,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.166497975708502e-05,
+      "loss": 2.8069,
+      "step": 2854
+    },
+    {
+      "epoch": 0.722419028340081,
+      "grad_norm": 3.625,
+      "learning_rate": 2.167257085020243e-05,
+      "loss": 2.6154,
+      "step": 2855
+    },
+    {
+      "epoch": 0.7226720647773279,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.1680161943319837e-05,
+      "loss": 2.5459,
+      "step": 2856
+    },
+    {
+      "epoch": 0.7229251012145749,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.1687753036437247e-05,
+      "loss": 2.6766,
+      "step": 2857
+    },
+    {
+      "epoch": 0.7231781376518218,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.1695344129554654e-05,
+      "loss": 2.6905,
+      "step": 2858
+    },
+    {
+      "epoch": 0.7234311740890689,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.1702935222672068e-05,
+      "loss": 2.7054,
+      "step": 2859
+    },
+    {
+      "epoch": 0.7236842105263158,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.1710526315789474e-05,
+      "loss": 2.6871,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7239372469635628,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.1718117408906884e-05,
+      "loss": 2.762,
+      "step": 2861
+    },
+    {
+      "epoch": 0.7241902834008097,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.172570850202429e-05,
+      "loss": 2.7294,
+      "step": 2862
+    },
+    {
+      "epoch": 0.7244433198380567,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.17332995951417e-05,
+      "loss": 2.7832,
+      "step": 2863
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.1740890688259108e-05,
+      "loss": 2.657,
+      "step": 2864
+    },
+    {
+      "epoch": 0.7249493927125507,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.174848178137652e-05,
+      "loss": 2.7658,
+      "step": 2865
+    },
+    {
+      "epoch": 0.7252024291497976,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.1756072874493928e-05,
+      "loss": 2.5579,
+      "step": 2866
+    },
+    {
+      "epoch": 0.7254554655870445,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.176366396761134e-05,
+      "loss": 2.6157,
+      "step": 2867
+    },
+    {
+      "epoch": 0.7257085020242915,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.1771255060728745e-05,
+      "loss": 2.6364,
+      "step": 2868
+    },
+    {
+      "epoch": 0.7259615384615384,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.1778846153846155e-05,
+      "loss": 2.7367,
+      "step": 2869
+    },
+    {
+      "epoch": 0.7262145748987854,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.1786437246963562e-05,
+      "loss": 2.651,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7264676113360324,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.1794028340080972e-05,
+      "loss": 2.7496,
+      "step": 2871
+    },
+    {
+      "epoch": 0.7267206477732794,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.1801619433198382e-05,
+      "loss": 2.6799,
+      "step": 2872
+    },
+    {
+      "epoch": 0.7269736842105263,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.180921052631579e-05,
+      "loss": 2.767,
+      "step": 2873
+    },
+    {
+      "epoch": 0.7272267206477733,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.18168016194332e-05,
+      "loss": 2.6498,
+      "step": 2874
+    },
+    {
+      "epoch": 0.7274797570850202,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.1824392712550606e-05,
+      "loss": 2.6531,
+      "step": 2875
+    },
+    {
+      "epoch": 0.7277327935222672,
+      "grad_norm": 3.5,
+      "learning_rate": 2.1831983805668016e-05,
+      "loss": 2.6582,
+      "step": 2876
+    },
+    {
+      "epoch": 0.7279858299595142,
+      "grad_norm": 3.375,
+      "learning_rate": 2.1839574898785426e-05,
+      "loss": 2.7734,
+      "step": 2877
+    },
+    {
+      "epoch": 0.7282388663967612,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.1847165991902836e-05,
+      "loss": 2.6556,
+      "step": 2878
+    },
+    {
+      "epoch": 0.7284919028340081,
+      "grad_norm": 3.25,
+      "learning_rate": 2.1854757085020243e-05,
+      "loss": 2.542,
+      "step": 2879
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.1862348178137653e-05,
+      "loss": 2.6901,
+      "step": 2880
+    },
+    {
+      "epoch": 0.728744939271255,
+      "eval_loss": 2.750654697418213,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.2037,
+      "eval_samples_per_second": 9.552,
+      "eval_steps_per_second": 9.552,
+      "step": 2880
+    },
+    {
+      "epoch": 0.728997975708502,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.186993927125506e-05,
+      "loss": 2.7081,
+      "step": 2881
+    },
+    {
+      "epoch": 0.729251012145749,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.187753036437247e-05,
+      "loss": 2.643,
+      "step": 2882
+    },
+    {
+      "epoch": 0.729504048582996,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.188512145748988e-05,
+      "loss": 2.6852,
+      "step": 2883
+    },
+    {
+      "epoch": 0.729757085020243,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.189271255060729e-05,
+      "loss": 2.4994,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7300101214574899,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.1900303643724697e-05,
+      "loss": 2.6501,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7302631578947368,
+      "grad_norm": 3.25,
+      "learning_rate": 2.1907894736842107e-05,
+      "loss": 2.6897,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7305161943319838,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.1915485829959514e-05,
+      "loss": 2.6917,
+      "step": 2887
+    },
+    {
+      "epoch": 0.7307692307692307,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.1923076923076924e-05,
+      "loss": 2.7631,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7310222672064778,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.1930668016194334e-05,
+      "loss": 2.6645,
+      "step": 2889
+    },
+    {
+      "epoch": 0.7312753036437247,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.1938259109311744e-05,
+      "loss": 2.7775,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7315283400809717,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.194585020242915e-05,
+      "loss": 2.6326,
+      "step": 2891
+    },
+    {
+      "epoch": 0.7317813765182186,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.1953441295546557e-05,
+      "loss": 2.5335,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7320344129554656,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.1961032388663968e-05,
+      "loss": 2.5355,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7322874493927125,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.1968623481781374e-05,
+      "loss": 2.5983,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7325404858299596,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.1976214574898788e-05,
+      "loss": 2.6637,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 3.375,
+      "learning_rate": 2.1983805668016195e-05,
+      "loss": 2.8059,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7330465587044535,
+      "grad_norm": 4.625,
+      "learning_rate": 2.1991396761133605e-05,
+      "loss": 2.6365,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7332995951417004,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.199898785425101e-05,
+      "loss": 2.6638,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7335526315789473,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.200657894736842e-05,
+      "loss": 2.5447,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7338056680161943,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.2014170040485828e-05,
+      "loss": 2.7612,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7340587044534413,
+      "grad_norm": 3.125,
+      "learning_rate": 2.2021761133603242e-05,
+      "loss": 2.5848,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7343117408906883,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.202935222672065e-05,
+      "loss": 2.7104,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7345647773279352,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.203694331983806e-05,
+      "loss": 2.6832,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7348178137651822,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.2044534412955465e-05,
+      "loss": 2.6666,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7350708502024291,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.2052125506072875e-05,
+      "loss": 2.6999,
+      "step": 2905
+    },
+    {
+      "epoch": 0.7353238866396761,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.2059716599190282e-05,
+      "loss": 2.7413,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7355769230769231,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.2067307692307696e-05,
+      "loss": 2.6747,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7358299595141701,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.2074898785425102e-05,
+      "loss": 2.7822,
+      "step": 2908
+    },
+    {
+      "epoch": 0.736082995951417,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.2082489878542513e-05,
+      "loss": 2.5901,
+      "step": 2909
+    },
+    {
+      "epoch": 0.736336032388664,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.209008097165992e-05,
+      "loss": 2.4631,
+      "step": 2910
+    },
+    {
+      "epoch": 0.736336032388664,
+      "eval_loss": 2.744507312774658,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.9198,
+      "eval_samples_per_second": 9.527,
+      "eval_steps_per_second": 9.527,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7365890688259109,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.209767206477733e-05,
+      "loss": 2.7354,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.2105263157894736e-05,
+      "loss": 2.7899,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7370951417004049,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.2112854251012146e-05,
+      "loss": 2.6446,
+      "step": 2913
+    },
+    {
+      "epoch": 0.7373481781376519,
+      "grad_norm": 3.125,
+      "learning_rate": 2.2120445344129556e-05,
+      "loss": 2.6442,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7376012145748988,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.2128036437246963e-05,
+      "loss": 2.5936,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7378542510121457,
+      "grad_norm": 3.5,
+      "learning_rate": 2.2135627530364373e-05,
+      "loss": 2.5958,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7381072874493927,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.214321862348178e-05,
+      "loss": 2.6408,
+      "step": 2917
+    },
+    {
+      "epoch": 0.7383603238866396,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.215080971659919e-05,
+      "loss": 2.5802,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7386133603238867,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.21584008097166e-05,
+      "loss": 2.6899,
+      "step": 2919
+    },
+    {
+      "epoch": 0.7388663967611336,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.216599190283401e-05,
+      "loss": 2.7789,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7391194331983806,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.2173582995951417e-05,
+      "loss": 2.6594,
+      "step": 2921
+    },
+    {
+      "epoch": 0.7393724696356275,
+      "grad_norm": 3.5,
+      "learning_rate": 2.2181174089068827e-05,
+      "loss": 2.6978,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7396255060728745,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.2188765182186234e-05,
+      "loss": 2.6057,
+      "step": 2923
+    },
+    {
+      "epoch": 0.7398785425101214,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.2196356275303644e-05,
+      "loss": 2.6894,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7401315789473685,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.2203947368421054e-05,
+      "loss": 2.7073,
+      "step": 2925
+    },
+    {
+      "epoch": 0.7403846153846154,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.2211538461538464e-05,
+      "loss": 2.7854,
+      "step": 2926
+    },
+    {
+      "epoch": 0.7406376518218624,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.221912955465587e-05,
+      "loss": 2.7096,
+      "step": 2927
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.222672064777328e-05,
+      "loss": 2.7113,
+      "step": 2928
+    },
+    {
+      "epoch": 0.7411437246963563,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.2234311740890688e-05,
+      "loss": 2.68,
+      "step": 2929
+    },
+    {
+      "epoch": 0.7413967611336032,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.2241902834008098e-05,
+      "loss": 2.6537,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7416497975708503,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.2249493927125508e-05,
+      "loss": 2.6443,
+      "step": 2931
+    },
+    {
+      "epoch": 0.7419028340080972,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.2257085020242918e-05,
+      "loss": 2.7096,
+      "step": 2932
+    },
+    {
+      "epoch": 0.7421558704453441,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.2264676113360325e-05,
+      "loss": 2.7953,
+      "step": 2933
+    },
+    {
+      "epoch": 0.7424089068825911,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.227226720647773e-05,
+      "loss": 2.5465,
+      "step": 2934
+    },
+    {
+      "epoch": 0.742661943319838,
+      "grad_norm": 3.375,
+      "learning_rate": 2.2279858299595142e-05,
+      "loss": 2.7294,
+      "step": 2935
+    },
+    {
+      "epoch": 0.742914979757085,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.228744939271255e-05,
+      "loss": 2.7265,
+      "step": 2936
+    },
+    {
+      "epoch": 0.743168016194332,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.2295040485829962e-05,
+      "loss": 2.5341,
+      "step": 2937
+    },
+    {
+      "epoch": 0.743421052631579,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.230263157894737e-05,
+      "loss": 2.6954,
+      "step": 2938
+    },
+    {
+      "epoch": 0.7436740890688259,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.231022267206478e-05,
+      "loss": 2.7404,
+      "step": 2939
+    },
+    {
+      "epoch": 0.7439271255060729,
+      "grad_norm": 3.125,
+      "learning_rate": 2.2317813765182186e-05,
+      "loss": 2.5806,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7439271255060729,
+      "eval_loss": 2.7456486225128174,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.632,
+      "eval_samples_per_second": 9.502,
+      "eval_steps_per_second": 9.502,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7441801619433198,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.2325404858299596e-05,
+      "loss": 2.7099,
+      "step": 2941
+    },
+    {
+      "epoch": 0.7444331983805668,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.2332995951417002e-05,
+      "loss": 2.6002,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7446862348178138,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.2340587044534416e-05,
+      "loss": 2.7895,
+      "step": 2943
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 3.375,
+      "learning_rate": 2.2348178137651823e-05,
+      "loss": 2.6617,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7451923076923077,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.2355769230769233e-05,
+      "loss": 2.6433,
+      "step": 2945
+    },
+    {
+      "epoch": 0.7454453441295547,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.236336032388664e-05,
+      "loss": 2.6041,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7456983805668016,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.237095141700405e-05,
+      "loss": 2.7637,
+      "step": 2947
+    },
+    {
+      "epoch": 0.7459514170040485,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.2378542510121456e-05,
+      "loss": 2.7245,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7462044534412956,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.238613360323887e-05,
+      "loss": 2.7809,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7464574898785425,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.2393724696356277e-05,
+      "loss": 2.6254,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7467105263157895,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.2401315789473687e-05,
+      "loss": 2.6753,
+      "step": 2951
+    },
+    {
+      "epoch": 0.7469635627530364,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.2408906882591094e-05,
+      "loss": 2.7314,
+      "step": 2952
+    },
+    {
+      "epoch": 0.7472165991902834,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.2416497975708504e-05,
+      "loss": 2.6581,
+      "step": 2953
+    },
+    {
+      "epoch": 0.7474696356275303,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.242408906882591e-05,
+      "loss": 2.6775,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7477226720647774,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.243168016194332e-05,
+      "loss": 2.6376,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7479757085020243,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.243927125506073e-05,
+      "loss": 2.6928,
+      "step": 2956
+    },
+    {
+      "epoch": 0.7482287449392713,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.2446862348178137e-05,
+      "loss": 2.7039,
+      "step": 2957
+    },
+    {
+      "epoch": 0.7484817813765182,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.2454453441295547e-05,
+      "loss": 2.7081,
+      "step": 2958
+    },
+    {
+      "epoch": 0.7487348178137652,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.2462044534412954e-05,
+      "loss": 2.7644,
+      "step": 2959
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.2469635627530364e-05,
+      "loss": 2.7337,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7492408906882592,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.2477226720647774e-05,
+      "loss": 2.575,
+      "step": 2961
+    },
+    {
+      "epoch": 0.7494939271255061,
+      "grad_norm": 3.5,
+      "learning_rate": 2.2484817813765185e-05,
+      "loss": 2.6202,
+      "step": 2962
+    },
+    {
+      "epoch": 0.749746963562753,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.249240890688259e-05,
+      "loss": 2.7013,
+      "step": 2963
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.25,
+      "learning_rate": 2.25e-05,
+      "loss": 2.7428,
+      "step": 2964
+    },
+    {
+      "epoch": 0.750253036437247,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.2507591093117408e-05,
+      "loss": 2.7051,
+      "step": 2965
+    },
+    {
+      "epoch": 0.7505060728744939,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.2515182186234818e-05,
+      "loss": 2.6076,
+      "step": 2966
+    },
+    {
+      "epoch": 0.7507591093117408,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.2522773279352225e-05,
+      "loss": 2.7031,
+      "step": 2967
+    },
+    {
+      "epoch": 0.7510121457489879,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.253036437246964e-05,
+      "loss": 2.6307,
+      "step": 2968
+    },
+    {
+      "epoch": 0.7512651821862348,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.2537955465587045e-05,
+      "loss": 2.6485,
+      "step": 2969
+    },
+    {
+      "epoch": 0.7515182186234818,
+      "grad_norm": 3.375,
+      "learning_rate": 2.2545546558704455e-05,
+      "loss": 2.7141,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7515182186234818,
+      "eval_loss": 2.7401885986328125,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.9357,
+      "eval_samples_per_second": 9.526,
+      "eval_steps_per_second": 9.526,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7517712550607287,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.2553137651821862e-05,
+      "loss": 2.5962,
+      "step": 2971
+    },
+    {
+      "epoch": 0.7520242914979757,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.2560728744939272e-05,
+      "loss": 2.5962,
+      "step": 2972
+    },
+    {
+      "epoch": 0.7522773279352226,
+      "grad_norm": 26.625,
+      "learning_rate": 2.256831983805668e-05,
+      "loss": 2.6827,
+      "step": 2973
+    },
+    {
+      "epoch": 0.7525303643724697,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.2575910931174092e-05,
+      "loss": 2.51,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7527834008097166,
+      "grad_norm": 3.5,
+      "learning_rate": 2.25835020242915e-05,
+      "loss": 2.7446,
+      "step": 2975
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.2591093117408906e-05,
+      "loss": 2.5697,
+      "step": 2976
+    },
+    {
+      "epoch": 0.7532894736842105,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.2598684210526316e-05,
+      "loss": 2.7084,
+      "step": 2977
+    },
+    {
+      "epoch": 0.7535425101214575,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.2606275303643723e-05,
+      "loss": 2.7081,
+      "step": 2978
+    },
+    {
+      "epoch": 0.7537955465587044,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.2613866396761133e-05,
+      "loss": 2.5676,
+      "step": 2979
+    },
+    {
+      "epoch": 0.7540485829959515,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.2621457489878543e-05,
+      "loss": 2.6802,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7543016194331984,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.2629048582995953e-05,
+      "loss": 2.629,
+      "step": 2981
+    },
+    {
+      "epoch": 0.7545546558704453,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.263663967611336e-05,
+      "loss": 2.5816,
+      "step": 2982
+    },
+    {
+      "epoch": 0.7548076923076923,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.264423076923077e-05,
+      "loss": 2.6787,
+      "step": 2983
+    },
+    {
+      "epoch": 0.7550607287449392,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.2651821862348177e-05,
+      "loss": 2.6635,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7553137651821862,
+      "grad_norm": 3.25,
+      "learning_rate": 2.2659412955465587e-05,
+      "loss": 2.6774,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7555668016194332,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.2667004048582997e-05,
+      "loss": 2.6784,
+      "step": 2986
+    },
+    {
+      "epoch": 0.7558198380566802,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.2674595141700407e-05,
+      "loss": 2.6618,
+      "step": 2987
+    },
+    {
+      "epoch": 0.7560728744939271,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.2682186234817814e-05,
+      "loss": 2.6535,
+      "step": 2988
+    },
+    {
+      "epoch": 0.7563259109311741,
+      "grad_norm": 3.25,
+      "learning_rate": 2.2689777327935224e-05,
+      "loss": 2.6947,
+      "step": 2989
+    },
+    {
+      "epoch": 0.756578947368421,
+      "grad_norm": 3.125,
+      "learning_rate": 2.269736842105263e-05,
+      "loss": 2.6462,
+      "step": 2990
+    },
+    {
+      "epoch": 0.756831983805668,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.270495951417004e-05,
+      "loss": 2.544,
+      "step": 2991
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.271255060728745e-05,
+      "loss": 2.7613,
+      "step": 2992
+    },
+    {
+      "epoch": 0.757338056680162,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.272014170040486e-05,
+      "loss": 2.6655,
+      "step": 2993
+    },
+    {
+      "epoch": 0.7575910931174089,
+      "grad_norm": 3.25,
+      "learning_rate": 2.2727732793522268e-05,
+      "loss": 2.6883,
+      "step": 2994
+    },
+    {
+      "epoch": 0.7578441295546559,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.2735323886639678e-05,
+      "loss": 2.6445,
+      "step": 2995
+    },
+    {
+      "epoch": 0.7580971659919028,
+      "grad_norm": 3.25,
+      "learning_rate": 2.2742914979757085e-05,
+      "loss": 2.6854,
+      "step": 2996
+    },
+    {
+      "epoch": 0.7583502024291497,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.275050607287449e-05,
+      "loss": 2.5352,
+      "step": 2997
+    },
+    {
+      "epoch": 0.7586032388663968,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.2758097165991905e-05,
+      "loss": 2.7088,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7588562753036437,
+      "grad_norm": 3.125,
+      "learning_rate": 2.276568825910931e-05,
+      "loss": 2.572,
+      "step": 2999
+    },
+    {
+      "epoch": 0.7591093117408907,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.2773279352226722e-05,
+      "loss": 2.6304,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7591093117408907,
+      "eval_loss": 2.7373597621917725,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.5685,
+      "eval_samples_per_second": 9.504,
+      "eval_steps_per_second": 9.504,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7593623481781376,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.278087044534413e-05,
+      "loss": 2.6217,
+      "step": 3001
+    },
+    {
+      "epoch": 0.7596153846153846,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.278846153846154e-05,
+      "loss": 2.5675,
+      "step": 3002
+    },
+    {
+      "epoch": 0.7598684210526315,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.2796052631578945e-05,
+      "loss": 2.7347,
+      "step": 3003
+    },
+    {
+      "epoch": 0.7601214574898786,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.280364372469636e-05,
+      "loss": 2.5945,
+      "step": 3004
+    },
+    {
+      "epoch": 0.7603744939271255,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.2811234817813766e-05,
+      "loss": 2.5772,
+      "step": 3005
+    },
+    {
+      "epoch": 0.7606275303643725,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.2818825910931176e-05,
+      "loss": 2.7264,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7608805668016194,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.2826417004048582e-05,
+      "loss": 2.7172,
+      "step": 3007
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.2834008097165992e-05,
+      "loss": 2.6576,
+      "step": 3008
+    },
+    {
+      "epoch": 0.7613866396761133,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.28415991902834e-05,
+      "loss": 2.7249,
+      "step": 3009
+    },
+    {
+      "epoch": 0.7616396761133604,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.2849190283400813e-05,
+      "loss": 2.7307,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7618927125506073,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.285678137651822e-05,
+      "loss": 2.6843,
+      "step": 3011
+    },
+    {
+      "epoch": 0.7621457489878543,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.286437246963563e-05,
+      "loss": 2.7428,
+      "step": 3012
+    },
+    {
+      "epoch": 0.7623987854251012,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.2871963562753036e-05,
+      "loss": 2.6363,
+      "step": 3013
+    },
+    {
+      "epoch": 0.7626518218623481,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.2879554655870446e-05,
+      "loss": 2.7052,
+      "step": 3014
+    },
+    {
+      "epoch": 0.7629048582995951,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.2887145748987853e-05,
+      "loss": 2.6947,
+      "step": 3015
+    },
+    {
+      "epoch": 0.7631578947368421,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.2894736842105263e-05,
+      "loss": 2.7009,
+      "step": 3016
+    },
+    {
+      "epoch": 0.7634109311740891,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.2902327935222673e-05,
+      "loss": 2.6455,
+      "step": 3017
+    },
+    {
+      "epoch": 0.763663967611336,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.290991902834008e-05,
+      "loss": 2.6835,
+      "step": 3018
+    },
+    {
+      "epoch": 0.763917004048583,
+      "grad_norm": 3.25,
+      "learning_rate": 2.291751012145749e-05,
+      "loss": 2.6234,
+      "step": 3019
+    },
+    {
+      "epoch": 0.7641700404858299,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.2925101214574897e-05,
+      "loss": 2.7171,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7644230769230769,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.2932692307692307e-05,
+      "loss": 2.7018,
+      "step": 3021
+    },
+    {
+      "epoch": 0.7646761133603239,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.2940283400809717e-05,
+      "loss": 2.6499,
+      "step": 3022
+    },
+    {
+      "epoch": 0.7649291497975709,
+      "grad_norm": 3.375,
+      "learning_rate": 2.2947874493927127e-05,
+      "loss": 2.7174,
+      "step": 3023
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.2955465587044534e-05,
+      "loss": 2.7102,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7654352226720648,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.2963056680161944e-05,
+      "loss": 2.6628,
+      "step": 3025
+    },
+    {
+      "epoch": 0.7656882591093117,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.297064777327935e-05,
+      "loss": 2.7316,
+      "step": 3026
+    },
+    {
+      "epoch": 0.7659412955465587,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.297823886639676e-05,
+      "loss": 2.6976,
+      "step": 3027
+    },
+    {
+      "epoch": 0.7661943319838057,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.298582995951417e-05,
+      "loss": 2.7464,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7664473684210527,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.299342105263158e-05,
+      "loss": 2.6874,
+      "step": 3029
+    },
+    {
+      "epoch": 0.7667004048582996,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.3001012145748988e-05,
+      "loss": 2.6333,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7667004048582996,
+      "eval_loss": 2.7381997108459473,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 283.5684,
+      "eval_samples_per_second": 9.169,
+      "eval_steps_per_second": 9.169,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7669534412955465,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.3008603238866398e-05,
+      "loss": 2.7163,
+      "step": 3031
+    },
+    {
+      "epoch": 0.7672064777327935,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.3016194331983805e-05,
+      "loss": 2.6555,
+      "step": 3032
+    },
+    {
+      "epoch": 0.7674595141700404,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.3023785425101215e-05,
+      "loss": 2.7413,
+      "step": 3033
+    },
+    {
+      "epoch": 0.7677125506072875,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.3031376518218625e-05,
+      "loss": 2.6561,
+      "step": 3034
+    },
+    {
+      "epoch": 0.7679655870445344,
+      "grad_norm": 3.25,
+      "learning_rate": 2.3038967611336035e-05,
+      "loss": 2.7231,
+      "step": 3035
+    },
+    {
+      "epoch": 0.7682186234817814,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.3046558704453442e-05,
+      "loss": 2.7183,
+      "step": 3036
+    },
+    {
+      "epoch": 0.7684716599190283,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.305414979757085e-05,
+      "loss": 2.5411,
+      "step": 3037
+    },
+    {
+      "epoch": 0.7687246963562753,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.306174089068826e-05,
+      "loss": 2.7295,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7689777327935222,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.3069331983805666e-05,
+      "loss": 2.6876,
+      "step": 3039
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.307692307692308e-05,
+      "loss": 2.6906,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7694838056680162,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.3084514170040486e-05,
+      "loss": 2.656,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7697368421052632,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.3092105263157896e-05,
+      "loss": 2.6865,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7699898785425101,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.3099696356275303e-05,
+      "loss": 2.7094,
+      "step": 3043
+    },
+    {
+      "epoch": 0.770242914979757,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.3107287449392713e-05,
+      "loss": 2.6725,
+      "step": 3044
+    },
+    {
+      "epoch": 0.770495951417004,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.311487854251012e-05,
+      "loss": 2.6593,
+      "step": 3045
+    },
+    {
+      "epoch": 0.770748987854251,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.3122469635627533e-05,
+      "loss": 2.6555,
+      "step": 3046
+    },
+    {
+      "epoch": 0.771002024291498,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.313006072874494e-05,
+      "loss": 2.7423,
+      "step": 3047
+    },
+    {
+      "epoch": 0.771255060728745,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.313765182186235e-05,
+      "loss": 2.7735,
+      "step": 3048
+    },
+    {
+      "epoch": 0.7715080971659919,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.3145242914979757e-05,
+      "loss": 2.672,
+      "step": 3049
+    },
+    {
+      "epoch": 0.7717611336032388,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.3152834008097167e-05,
+      "loss": 2.7184,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7720141700404858,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.3160425101214573e-05,
+      "loss": 2.6368,
+      "step": 3051
+    },
+    {
+      "epoch": 0.7722672064777328,
+      "grad_norm": 4.125,
+      "learning_rate": 2.3168016194331987e-05,
+      "loss": 2.6559,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7725202429149798,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.3175607287449394e-05,
+      "loss": 2.5762,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7727732793522267,
+      "grad_norm": 3.375,
+      "learning_rate": 2.3183198380566804e-05,
+      "loss": 2.6946,
+      "step": 3054
+    },
+    {
+      "epoch": 0.7730263157894737,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.319078947368421e-05,
+      "loss": 2.6929,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.319838056680162e-05,
+      "loss": 2.6336,
+      "step": 3056
+    },
+    {
+      "epoch": 0.7735323886639676,
+      "grad_norm": 3.125,
+      "learning_rate": 2.3205971659919027e-05,
+      "loss": 2.5908,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7737854251012146,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.3213562753036438e-05,
+      "loss": 2.6303,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7740384615384616,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.3221153846153848e-05,
+      "loss": 2.6548,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7742914979757085,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.3228744939271254e-05,
+      "loss": 2.6603,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7742914979757085,
+      "eval_loss": 2.743887186050415,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.9482,
+      "eval_samples_per_second": 9.526,
+      "eval_steps_per_second": 9.526,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7745445344129555,
+      "grad_norm": 3.625,
+      "learning_rate": 2.3236336032388664e-05,
+      "loss": 2.5837,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7747975708502024,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.324392712550607e-05,
+      "loss": 2.6816,
+      "step": 3062
+    },
+    {
+      "epoch": 0.7750506072874493,
+      "grad_norm": 3.375,
+      "learning_rate": 2.325151821862348e-05,
+      "loss": 2.6991,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7753036437246964,
+      "grad_norm": 3.953125,
+      "learning_rate": 2.325910931174089e-05,
+      "loss": 2.6406,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7755566801619433,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.32667004048583e-05,
+      "loss": 2.6758,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7758097165991903,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.327429149797571e-05,
+      "loss": 2.7374,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7760627530364372,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.328188259109312e-05,
+      "loss": 2.6563,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7763157894736842,
+      "grad_norm": 3.125,
+      "learning_rate": 2.3289473684210525e-05,
+      "loss": 2.5318,
+      "step": 3068
+    },
+    {
+      "epoch": 0.7765688259109311,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.3297064777327935e-05,
+      "loss": 2.7115,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7768218623481782,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.3304655870445345e-05,
+      "loss": 2.6664,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7770748987854251,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.3312246963562756e-05,
+      "loss": 2.5547,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.3319838056680162e-05,
+      "loss": 2.715,
+      "step": 3072
+    },
+    {
+      "epoch": 0.777580971659919,
+      "grad_norm": 3.5,
+      "learning_rate": 2.3327429149797572e-05,
+      "loss": 2.6837,
+      "step": 3073
+    },
+    {
+      "epoch": 0.777834008097166,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.333502024291498e-05,
+      "loss": 2.5746,
+      "step": 3074
+    },
+    {
+      "epoch": 0.7780870445344129,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.334261133603239e-05,
+      "loss": 2.6328,
+      "step": 3075
+    },
+    {
+      "epoch": 0.77834008097166,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.33502024291498e-05,
+      "loss": 2.4921,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7785931174089069,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.335779352226721e-05,
+      "loss": 2.6029,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7788461538461539,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.3365384615384616e-05,
+      "loss": 2.7748,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7790991902834008,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.3372975708502023e-05,
+      "loss": 2.7683,
+      "step": 3079
+    },
+    {
+      "epoch": 0.7793522267206477,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.3380566801619433e-05,
+      "loss": 2.7002,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7796052631578947,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.338815789473684e-05,
+      "loss": 2.5665,
+      "step": 3081
+    },
+    {
+      "epoch": 0.7798582995951417,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.3395748987854253e-05,
+      "loss": 2.7156,
+      "step": 3082
+    },
+    {
+      "epoch": 0.7801113360323887,
+      "grad_norm": 3.5,
+      "learning_rate": 2.340334008097166e-05,
+      "loss": 2.7988,
+      "step": 3083
+    },
+    {
+      "epoch": 0.7803643724696356,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.341093117408907e-05,
+      "loss": 2.5862,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7806174089068826,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.3418522267206477e-05,
+      "loss": 2.4331,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7808704453441295,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.3426113360323887e-05,
+      "loss": 2.6431,
+      "step": 3086
+    },
+    {
+      "epoch": 0.7811234817813765,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.3433704453441294e-05,
+      "loss": 2.5984,
+      "step": 3087
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.3441295546558707e-05,
+      "loss": 2.5673,
+      "step": 3088
+    },
+    {
+      "epoch": 0.7816295546558705,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.3448886639676114e-05,
+      "loss": 2.5863,
+      "step": 3089
+    },
+    {
+      "epoch": 0.7818825910931174,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.3456477732793524e-05,
+      "loss": 2.6174,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7818825910931174,
+      "eval_loss": 2.7358498573303223,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.3037,
+      "eval_samples_per_second": 9.513,
+      "eval_steps_per_second": 9.513,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7821356275303644,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.346406882591093e-05,
+      "loss": 2.7769,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7823886639676113,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.347165991902834e-05,
+      "loss": 2.6834,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7826417004048583,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.3479251012145748e-05,
+      "loss": 2.5575,
+      "step": 3093
+    },
+    {
+      "epoch": 0.7828947368421053,
+      "grad_norm": 3.25,
+      "learning_rate": 2.348684210526316e-05,
+      "loss": 2.6571,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7831477732793523,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.3494433198380568e-05,
+      "loss": 2.7951,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7834008097165992,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.3502024291497978e-05,
+      "loss": 2.6466,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7836538461538461,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.3509615384615385e-05,
+      "loss": 2.6756,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7839068825910931,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.3517206477732795e-05,
+      "loss": 2.6699,
+      "step": 3098
+    },
+    {
+      "epoch": 0.78415991902834,
+      "grad_norm": 3.375,
+      "learning_rate": 2.35247975708502e-05,
+      "loss": 2.698,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7844129554655871,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.3532388663967612e-05,
+      "loss": 2.7044,
+      "step": 3100
+    },
+    {
+      "epoch": 0.784665991902834,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.3539979757085022e-05,
+      "loss": 2.6821,
+      "step": 3101
+    },
+    {
+      "epoch": 0.784919028340081,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.354757085020243e-05,
+      "loss": 2.6564,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7851720647773279,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.355516194331984e-05,
+      "loss": 2.5695,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.3562753036437245e-05,
+      "loss": 2.6984,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7856781376518218,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.3570344129554656e-05,
+      "loss": 2.6344,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7859311740890689,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.3577935222672066e-05,
+      "loss": 2.7127,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7861842105263158,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.3585526315789476e-05,
+      "loss": 2.6057,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7864372469635628,
+      "grad_norm": 3.125,
+      "learning_rate": 2.3593117408906883e-05,
+      "loss": 2.6294,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7866902834008097,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.3600708502024293e-05,
+      "loss": 2.613,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7869433198380567,
+      "grad_norm": 3.375,
+      "learning_rate": 2.36082995951417e-05,
+      "loss": 2.547,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7871963562753036,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.361589068825911e-05,
+      "loss": 2.72,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7874493927125507,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.362348178137652e-05,
+      "loss": 2.5475,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7877024291497976,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.363107287449393e-05,
+      "loss": 2.4908,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7879554655870445,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.3638663967611336e-05,
+      "loss": 2.6907,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7882085020242915,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.3646255060728747e-05,
+      "loss": 2.6731,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7884615384615384,
+      "grad_norm": 3.125,
+      "learning_rate": 2.3653846153846153e-05,
+      "loss": 2.6995,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7887145748987854,
+      "grad_norm": 3.375,
+      "learning_rate": 2.3661437246963563e-05,
+      "loss": 2.6606,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7889676113360324,
+      "grad_norm": 3.25,
+      "learning_rate": 2.3669028340080974e-05,
+      "loss": 2.6009,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7892206477732794,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.3676619433198384e-05,
+      "loss": 2.655,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.368421052631579e-05,
+      "loss": 2.6498,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "eval_loss": 2.7270102500915527,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.4137,
+      "eval_samples_per_second": 9.544,
+      "eval_steps_per_second": 9.544,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7897267206477733,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.3691801619433197e-05,
+      "loss": 2.4799,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7899797570850202,
+      "grad_norm": 3.375,
+      "learning_rate": 2.3699392712550607e-05,
+      "loss": 2.5748,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7902327935222672,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.3706983805668014e-05,
+      "loss": 2.6259,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7904858299595142,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.3714574898785428e-05,
+      "loss": 2.645,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7907388663967612,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.3722165991902834e-05,
+      "loss": 2.5014,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7909919028340081,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.3729757085020244e-05,
+      "loss": 2.7325,
+      "step": 3126
+    },
+    {
+      "epoch": 0.791244939271255,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.373734817813765e-05,
+      "loss": 2.6699,
+      "step": 3127
+    },
+    {
+      "epoch": 0.791497975708502,
+      "grad_norm": 3.625,
+      "learning_rate": 2.374493927125506e-05,
+      "loss": 2.6479,
+      "step": 3128
+    },
+    {
+      "epoch": 0.791751012145749,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.3752530364372468e-05,
+      "loss": 2.6588,
+      "step": 3129
+    },
+    {
+      "epoch": 0.792004048582996,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.376012145748988e-05,
+      "loss": 2.6149,
+      "step": 3130
+    },
+    {
+      "epoch": 0.792257085020243,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.3767712550607288e-05,
+      "loss": 2.611,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7925101214574899,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.37753036437247e-05,
+      "loss": 2.7205,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7927631578947368,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.3782894736842105e-05,
+      "loss": 2.7005,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7930161943319838,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.3790485829959515e-05,
+      "loss": 2.7091,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7932692307692307,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.3798076923076922e-05,
+      "loss": 2.6711,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 31.125,
+      "learning_rate": 2.3805668016194335e-05,
+      "loss": 2.8105,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7937753036437247,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.3813259109311742e-05,
+      "loss": 2.6487,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7940283400809717,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.3820850202429152e-05,
+      "loss": 2.6887,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7942813765182186,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.382844129554656e-05,
+      "loss": 2.7009,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7945344129554656,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.383603238866397e-05,
+      "loss": 2.6726,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7947874493927125,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.3843623481781376e-05,
+      "loss": 2.7101,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7950404858299596,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.3851214574898786e-05,
+      "loss": 2.5466,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7952935222672065,
+      "grad_norm": 3.125,
+      "learning_rate": 2.3858805668016196e-05,
+      "loss": 2.6769,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7955465587044535,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.3866396761133603e-05,
+      "loss": 2.5509,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7957995951417004,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.3873987854251013e-05,
+      "loss": 2.6377,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7960526315789473,
+      "grad_norm": 3.625,
+      "learning_rate": 2.388157894736842e-05,
+      "loss": 2.6793,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7963056680161943,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.388917004048583e-05,
+      "loss": 2.7519,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7965587044534413,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.389676113360324e-05,
+      "loss": 2.4965,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7968117408906883,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.390435222672065e-05,
+      "loss": 2.6932,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7970647773279352,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.3911943319838057e-05,
+      "loss": 2.6626,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7970647773279352,
+      "eval_loss": 2.72641921043396,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.7795,
+      "eval_samples_per_second": 9.532,
+      "eval_steps_per_second": 9.532,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7973178137651822,
+      "grad_norm": 3.5,
+      "learning_rate": 2.3919534412955467e-05,
+      "loss": 2.6788,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.3927125506072874e-05,
+      "loss": 2.712,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7978238866396761,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.3934716599190284e-05,
+      "loss": 2.6585,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7980769230769231,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.3942307692307694e-05,
+      "loss": 2.6479,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7983299595141701,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.3949898785425104e-05,
+      "loss": 2.593,
+      "step": 3155
+    },
+    {
+      "epoch": 0.798582995951417,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.395748987854251e-05,
+      "loss": 2.6663,
+      "step": 3156
+    },
+    {
+      "epoch": 0.798836032388664,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.396508097165992e-05,
+      "loss": 2.6726,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7990890688259109,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.3972672064777328e-05,
+      "loss": 2.6702,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7993421052631579,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.3980263157894738e-05,
+      "loss": 2.6952,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7995951417004049,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.3987854251012148e-05,
+      "loss": 2.7573,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7998481781376519,
+      "grad_norm": 3.375,
+      "learning_rate": 2.3995445344129558e-05,
+      "loss": 2.7681,
+      "step": 3161
+    },
+    {
+      "epoch": 0.8001012145748988,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.4003036437246965e-05,
+      "loss": 2.7472,
+      "step": 3162
+    },
+    {
+      "epoch": 0.8003542510121457,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.401062753036437e-05,
+      "loss": 2.6449,
+      "step": 3163
+    },
+    {
+      "epoch": 0.8006072874493927,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.401821862348178e-05,
+      "loss": 2.6413,
+      "step": 3164
+    },
+    {
+      "epoch": 0.8008603238866396,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.4025809716599188e-05,
+      "loss": 2.642,
+      "step": 3165
+    },
+    {
+      "epoch": 0.8011133603238867,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.4033400809716602e-05,
+      "loss": 2.668,
+      "step": 3166
+    },
+    {
+      "epoch": 0.8013663967611336,
+      "grad_norm": 3.375,
+      "learning_rate": 2.404099190283401e-05,
+      "loss": 2.7315,
+      "step": 3167
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.404858299595142e-05,
+      "loss": 2.5857,
+      "step": 3168
+    },
+    {
+      "epoch": 0.8018724696356275,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.4056174089068825e-05,
+      "loss": 2.6473,
+      "step": 3169
+    },
+    {
+      "epoch": 0.8021255060728745,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.4063765182186235e-05,
+      "loss": 2.5963,
+      "step": 3170
+    },
+    {
+      "epoch": 0.8023785425101214,
+      "grad_norm": 3.25,
+      "learning_rate": 2.4071356275303642e-05,
+      "loss": 2.6495,
+      "step": 3171
+    },
+    {
+      "epoch": 0.8026315789473685,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.4078947368421056e-05,
+      "loss": 2.6958,
+      "step": 3172
+    },
+    {
+      "epoch": 0.8028846153846154,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.4086538461538462e-05,
+      "loss": 2.5456,
+      "step": 3173
+    },
+    {
+      "epoch": 0.8031376518218624,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.4094129554655873e-05,
+      "loss": 2.7156,
+      "step": 3174
+    },
+    {
+      "epoch": 0.8033906882591093,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.410172064777328e-05,
+      "loss": 2.5895,
+      "step": 3175
+    },
+    {
+      "epoch": 0.8036437246963563,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.410931174089069e-05,
+      "loss": 2.5756,
+      "step": 3176
+    },
+    {
+      "epoch": 0.8038967611336032,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.4116902834008096e-05,
+      "loss": 2.6139,
+      "step": 3177
+    },
+    {
+      "epoch": 0.8041497975708503,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.412449392712551e-05,
+      "loss": 2.6657,
+      "step": 3178
+    },
+    {
+      "epoch": 0.8044028340080972,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.4132085020242916e-05,
+      "loss": 2.7314,
+      "step": 3179
+    },
+    {
+      "epoch": 0.8046558704453441,
+      "grad_norm": 2.875,
+      "learning_rate": 2.4139676113360326e-05,
+      "loss": 2.4163,
+      "step": 3180
+    },
+    {
+      "epoch": 0.8046558704453441,
+      "eval_loss": 2.727024793624878,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.8787,
+      "eval_samples_per_second": 9.528,
+      "eval_steps_per_second": 9.528,
+      "step": 3180
+    },
+    {
+      "epoch": 0.8049089068825911,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.4147267206477733e-05,
+      "loss": 2.6841,
+      "step": 3181
+    },
+    {
+      "epoch": 0.805161943319838,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.4154858299595143e-05,
+      "loss": 2.5821,
+      "step": 3182
+    },
+    {
+      "epoch": 0.805414979757085,
+      "grad_norm": 3.25,
+      "learning_rate": 2.416244939271255e-05,
+      "loss": 2.6068,
+      "step": 3183
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.417004048582996e-05,
+      "loss": 2.6752,
+      "step": 3184
+    },
+    {
+      "epoch": 0.805921052631579,
+      "grad_norm": 3.5,
+      "learning_rate": 2.417763157894737e-05,
+      "loss": 2.7015,
+      "step": 3185
+    },
+    {
+      "epoch": 0.8061740890688259,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.4185222672064777e-05,
+      "loss": 2.5892,
+      "step": 3186
+    },
+    {
+      "epoch": 0.8064271255060729,
+      "grad_norm": 3.25,
+      "learning_rate": 2.4192813765182187e-05,
+      "loss": 2.6669,
+      "step": 3187
+    },
+    {
+      "epoch": 0.8066801619433198,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.4200404858299594e-05,
+      "loss": 2.5491,
+      "step": 3188
+    },
+    {
+      "epoch": 0.8069331983805668,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.4207995951417004e-05,
+      "loss": 2.6799,
+      "step": 3189
+    },
+    {
+      "epoch": 0.8071862348178138,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.4215587044534414e-05,
+      "loss": 2.692,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8074392712550608,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.4223178137651824e-05,
+      "loss": 2.6633,
+      "step": 3191
+    },
+    {
+      "epoch": 0.8076923076923077,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.423076923076923e-05,
+      "loss": 2.5712,
+      "step": 3192
+    },
+    {
+      "epoch": 0.8079453441295547,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.423836032388664e-05,
+      "loss": 2.5755,
+      "step": 3193
+    },
+    {
+      "epoch": 0.8081983805668016,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.4245951417004048e-05,
+      "loss": 2.6761,
+      "step": 3194
+    },
+    {
+      "epoch": 0.8084514170040485,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.4253542510121458e-05,
+      "loss": 2.6519,
+      "step": 3195
+    },
+    {
+      "epoch": 0.8087044534412956,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.4261133603238868e-05,
+      "loss": 2.6596,
+      "step": 3196
+    },
+    {
+      "epoch": 0.8089574898785425,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.4268724696356278e-05,
+      "loss": 2.7481,
+      "step": 3197
+    },
+    {
+      "epoch": 0.8092105263157895,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.4276315789473685e-05,
+      "loss": 2.7195,
+      "step": 3198
+    },
+    {
+      "epoch": 0.8094635627530364,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.4283906882591095e-05,
+      "loss": 2.6269,
+      "step": 3199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 4.125,
+      "learning_rate": 2.4291497975708502e-05,
+      "loss": 2.6272,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8099696356275303,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.4299089068825912e-05,
+      "loss": 2.7026,
+      "step": 3201
+    },
+    {
+      "epoch": 0.8102226720647774,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.4306680161943322e-05,
+      "loss": 2.6861,
+      "step": 3202
+    },
+    {
+      "epoch": 0.8104757085020243,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.431427125506073e-05,
+      "loss": 2.691,
+      "step": 3203
+    },
+    {
+      "epoch": 0.8107287449392713,
+      "grad_norm": 4.15625,
+      "learning_rate": 2.432186234817814e-05,
+      "loss": 2.5759,
+      "step": 3204
+    },
+    {
+      "epoch": 0.8109817813765182,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.4329453441295546e-05,
+      "loss": 2.7333,
+      "step": 3205
+    },
+    {
+      "epoch": 0.8112348178137652,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.4337044534412956e-05,
+      "loss": 2.6477,
+      "step": 3206
+    },
+    {
+      "epoch": 0.8114878542510121,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.4344635627530362e-05,
+      "loss": 2.6616,
+      "step": 3207
+    },
+    {
+      "epoch": 0.8117408906882592,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.4352226720647776e-05,
+      "loss": 2.6712,
+      "step": 3208
+    },
+    {
+      "epoch": 0.8119939271255061,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.4359817813765183e-05,
+      "loss": 2.5797,
+      "step": 3209
+    },
+    {
+      "epoch": 0.812246963562753,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.4367408906882593e-05,
+      "loss": 2.7074,
+      "step": 3210
+    },
+    {
+      "epoch": 0.812246963562753,
+      "eval_loss": 2.7210934162139893,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.9003,
+      "eval_samples_per_second": 9.493,
+      "eval_steps_per_second": 9.493,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.4375e-05,
+      "loss": 2.6775,
+      "step": 3211
+    },
+    {
+      "epoch": 0.812753036437247,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.438259109311741e-05,
+      "loss": 2.6801,
+      "step": 3212
+    },
+    {
+      "epoch": 0.8130060728744939,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.4390182186234816e-05,
+      "loss": 2.6502,
+      "step": 3213
+    },
+    {
+      "epoch": 0.8132591093117408,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.4397773279352227e-05,
+      "loss": 2.6,
+      "step": 3214
+    },
+    {
+      "epoch": 0.8135121457489879,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.4405364372469637e-05,
+      "loss": 2.5501,
+      "step": 3215
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.4412955465587047e-05,
+      "loss": 2.6769,
+      "step": 3216
+    },
+    {
+      "epoch": 0.8140182186234818,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.4420546558704454e-05,
+      "loss": 2.6591,
+      "step": 3217
+    },
+    {
+      "epoch": 0.8142712550607287,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.4428137651821864e-05,
+      "loss": 2.6202,
+      "step": 3218
+    },
+    {
+      "epoch": 0.8145242914979757,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.443572874493927e-05,
+      "loss": 2.6495,
+      "step": 3219
+    },
+    {
+      "epoch": 0.8147773279352226,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.444331983805668e-05,
+      "loss": 2.6601,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8150303643724697,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.445091093117409e-05,
+      "loss": 2.7302,
+      "step": 3221
+    },
+    {
+      "epoch": 0.8152834008097166,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.44585020242915e-05,
+      "loss": 2.7584,
+      "step": 3222
+    },
+    {
+      "epoch": 0.8155364372469636,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.4466093117408907e-05,
+      "loss": 2.6438,
+      "step": 3223
+    },
+    {
+      "epoch": 0.8157894736842105,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.4473684210526318e-05,
+      "loss": 2.6312,
+      "step": 3224
+    },
+    {
+      "epoch": 0.8160425101214575,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.4481275303643724e-05,
+      "loss": 2.5302,
+      "step": 3225
+    },
+    {
+      "epoch": 0.8162955465587044,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.448886639676113e-05,
+      "loss": 2.6503,
+      "step": 3226
+    },
+    {
+      "epoch": 0.8165485829959515,
+      "grad_norm": 3.25,
+      "learning_rate": 2.4496457489878545e-05,
+      "loss": 2.6506,
+      "step": 3227
+    },
+    {
+      "epoch": 0.8168016194331984,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.450404858299595e-05,
+      "loss": 2.6895,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8170546558704453,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.451163967611336e-05,
+      "loss": 2.6265,
+      "step": 3229
+    },
+    {
+      "epoch": 0.8173076923076923,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.4519230769230768e-05,
+      "loss": 2.5588,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8175607287449392,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.4526821862348178e-05,
+      "loss": 2.6254,
+      "step": 3231
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 3.25,
+      "learning_rate": 2.4534412955465585e-05,
+      "loss": 2.6313,
+      "step": 3232
+    },
+    {
+      "epoch": 0.8180668016194332,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.4542004048583e-05,
+      "loss": 2.6883,
+      "step": 3233
+    },
+    {
+      "epoch": 0.8183198380566802,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.4549595141700405e-05,
+      "loss": 2.6758,
+      "step": 3234
+    },
+    {
+      "epoch": 0.8185728744939271,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.4557186234817815e-05,
+      "loss": 2.6907,
+      "step": 3235
+    },
+    {
+      "epoch": 0.8188259109311741,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.4564777327935222e-05,
+      "loss": 2.7503,
+      "step": 3236
+    },
+    {
+      "epoch": 0.819078947368421,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.4572368421052632e-05,
+      "loss": 2.6554,
+      "step": 3237
+    },
+    {
+      "epoch": 0.819331983805668,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.457995951417004e-05,
+      "loss": 2.6659,
+      "step": 3238
+    },
+    {
+      "epoch": 0.819585020242915,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.4587550607287452e-05,
+      "loss": 2.6404,
+      "step": 3239
+    },
+    {
+      "epoch": 0.819838056680162,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.459514170040486e-05,
+      "loss": 2.5942,
+      "step": 3240
+    },
+    {
+      "epoch": 0.819838056680162,
+      "eval_loss": 2.7273881435394287,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.9242,
+      "eval_samples_per_second": 9.526,
+      "eval_steps_per_second": 9.526,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8200910931174089,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.460273279352227e-05,
+      "loss": 2.6758,
+      "step": 3241
+    },
+    {
+      "epoch": 0.8203441295546559,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.4610323886639676e-05,
+      "loss": 2.6966,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8205971659919028,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.4617914979757086e-05,
+      "loss": 2.7694,
+      "step": 3243
+    },
+    {
+      "epoch": 0.8208502024291497,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.4625506072874493e-05,
+      "loss": 2.6806,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8211032388663968,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.4633097165991903e-05,
+      "loss": 2.6943,
+      "step": 3245
+    },
+    {
+      "epoch": 0.8213562753036437,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.4640688259109313e-05,
+      "loss": 2.6872,
+      "step": 3246
+    },
+    {
+      "epoch": 0.8216093117408907,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.464827935222672e-05,
+      "loss": 2.6417,
+      "step": 3247
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.465587044534413e-05,
+      "loss": 2.5997,
+      "step": 3248
+    },
+    {
+      "epoch": 0.8221153846153846,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.4663461538461537e-05,
+      "loss": 2.7127,
+      "step": 3249
+    },
+    {
+      "epoch": 0.8223684210526315,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.4671052631578947e-05,
+      "loss": 2.6515,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8226214574898786,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.4678643724696357e-05,
+      "loss": 2.6184,
+      "step": 3251
+    },
+    {
+      "epoch": 0.8228744939271255,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.4686234817813767e-05,
+      "loss": 2.6758,
+      "step": 3252
+    },
+    {
+      "epoch": 0.8231275303643725,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.4693825910931174e-05,
+      "loss": 2.7089,
+      "step": 3253
+    },
+    {
+      "epoch": 0.8233805668016194,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.4701417004048584e-05,
+      "loss": 2.5652,
+      "step": 3254
+    },
+    {
+      "epoch": 0.8236336032388664,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.470900809716599e-05,
+      "loss": 2.6858,
+      "step": 3255
+    },
+    {
+      "epoch": 0.8238866396761133,
+      "grad_norm": 3.25,
+      "learning_rate": 2.47165991902834e-05,
+      "loss": 2.6272,
+      "step": 3256
+    },
+    {
+      "epoch": 0.8241396761133604,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.472419028340081e-05,
+      "loss": 2.6607,
+      "step": 3257
+    },
+    {
+      "epoch": 0.8243927125506073,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.473178137651822e-05,
+      "loss": 2.5742,
+      "step": 3258
+    },
+    {
+      "epoch": 0.8246457489878543,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.4739372469635628e-05,
+      "loss": 2.6328,
+      "step": 3259
+    },
+    {
+      "epoch": 0.8248987854251012,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.4746963562753038e-05,
+      "loss": 2.6295,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8251518218623481,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.4754554655870445e-05,
+      "loss": 2.7427,
+      "step": 3261
+    },
+    {
+      "epoch": 0.8254048582995951,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.4762145748987855e-05,
+      "loss": 2.712,
+      "step": 3262
+    },
+    {
+      "epoch": 0.8256578947368421,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.4769736842105265e-05,
+      "loss": 2.7367,
+      "step": 3263
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.4777327935222675e-05,
+      "loss": 2.6535,
+      "step": 3264
+    },
+    {
+      "epoch": 0.826163967611336,
+      "grad_norm": 3.375,
+      "learning_rate": 2.478491902834008e-05,
+      "loss": 2.7074,
+      "step": 3265
+    },
+    {
+      "epoch": 0.826417004048583,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.479251012145749e-05,
+      "loss": 2.6661,
+      "step": 3266
+    },
+    {
+      "epoch": 0.8266700404858299,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.48001012145749e-05,
+      "loss": 2.639,
+      "step": 3267
+    },
+    {
+      "epoch": 0.8269230769230769,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.4807692307692305e-05,
+      "loss": 2.6649,
+      "step": 3268
+    },
+    {
+      "epoch": 0.8271761133603239,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.481528340080972e-05,
+      "loss": 2.4975,
+      "step": 3269
+    },
+    {
+      "epoch": 0.8274291497975709,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.4822874493927125e-05,
+      "loss": 2.7,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8274291497975709,
+      "eval_loss": 2.7184789180755615,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.8744,
+      "eval_samples_per_second": 9.528,
+      "eval_steps_per_second": 9.528,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8276821862348178,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.4830465587044536e-05,
+      "loss": 2.5769,
+      "step": 3271
+    },
+    {
+      "epoch": 0.8279352226720648,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.4838056680161942e-05,
+      "loss": 2.65,
+      "step": 3272
+    },
+    {
+      "epoch": 0.8281882591093117,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.4845647773279352e-05,
+      "loss": 2.6563,
+      "step": 3273
+    },
+    {
+      "epoch": 0.8284412955465587,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.485323886639676e-05,
+      "loss": 2.7384,
+      "step": 3274
+    },
+    {
+      "epoch": 0.8286943319838057,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.4860829959514173e-05,
+      "loss": 2.677,
+      "step": 3275
+    },
+    {
+      "epoch": 0.8289473684210527,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.486842105263158e-05,
+      "loss": 2.6122,
+      "step": 3276
+    },
+    {
+      "epoch": 0.8292004048582996,
+      "grad_norm": 3.25,
+      "learning_rate": 2.487601214574899e-05,
+      "loss": 2.6212,
+      "step": 3277
+    },
+    {
+      "epoch": 0.8294534412955465,
+      "grad_norm": 3.0,
+      "learning_rate": 2.4883603238866396e-05,
+      "loss": 2.6121,
+      "step": 3278
+    },
+    {
+      "epoch": 0.8297064777327935,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.4891194331983806e-05,
+      "loss": 2.6202,
+      "step": 3279
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.4898785425101213e-05,
+      "loss": 2.6299,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8302125506072875,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.4906376518218627e-05,
+      "loss": 2.6546,
+      "step": 3281
+    },
+    {
+      "epoch": 0.8304655870445344,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.4913967611336033e-05,
+      "loss": 2.6601,
+      "step": 3282
+    },
+    {
+      "epoch": 0.8307186234817814,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.4921558704453444e-05,
+      "loss": 2.6394,
+      "step": 3283
+    },
+    {
+      "epoch": 0.8309716599190283,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.492914979757085e-05,
+      "loss": 2.5496,
+      "step": 3284
+    },
+    {
+      "epoch": 0.8312246963562753,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.493674089068826e-05,
+      "loss": 2.7346,
+      "step": 3285
+    },
+    {
+      "epoch": 0.8314777327935222,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.4944331983805667e-05,
+      "loss": 2.6257,
+      "step": 3286
+    },
+    {
+      "epoch": 0.8317307692307693,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.4951923076923077e-05,
+      "loss": 2.632,
+      "step": 3287
+    },
+    {
+      "epoch": 0.8319838056680162,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.4959514170040487e-05,
+      "loss": 2.6403,
+      "step": 3288
+    },
+    {
+      "epoch": 0.8322368421052632,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.4967105263157894e-05,
+      "loss": 2.6339,
+      "step": 3289
+    },
+    {
+      "epoch": 0.8324898785425101,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.4974696356275304e-05,
+      "loss": 2.5957,
+      "step": 3290
+    },
+    {
+      "epoch": 0.832742914979757,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.498228744939271e-05,
+      "loss": 2.6917,
+      "step": 3291
+    },
+    {
+      "epoch": 0.832995951417004,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.498987854251012e-05,
+      "loss": 2.7348,
+      "step": 3292
+    },
+    {
+      "epoch": 0.833248987854251,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.499746963562753e-05,
+      "loss": 2.6493,
+      "step": 3293
+    },
+    {
+      "epoch": 0.833502024291498,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.500506072874494e-05,
+      "loss": 2.6108,
+      "step": 3294
+    },
+    {
+      "epoch": 0.833755060728745,
+      "grad_norm": 26.75,
+      "learning_rate": 2.5012651821862348e-05,
+      "loss": 2.7342,
+      "step": 3295
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.5020242914979758e-05,
+      "loss": 2.614,
+      "step": 3296
+    },
+    {
+      "epoch": 0.8342611336032388,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.5027834008097165e-05,
+      "loss": 2.66,
+      "step": 3297
+    },
+    {
+      "epoch": 0.8345141700404858,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.5035425101214575e-05,
+      "loss": 2.5925,
+      "step": 3298
+    },
+    {
+      "epoch": 0.8347672064777328,
+      "grad_norm": 3.25,
+      "learning_rate": 2.5043016194331985e-05,
+      "loss": 2.6336,
+      "step": 3299
+    },
+    {
+      "epoch": 0.8350202429149798,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.5050607287449395e-05,
+      "loss": 2.7202,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8350202429149798,
+      "eval_loss": 2.7177040576934814,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.7772,
+      "eval_samples_per_second": 9.497,
+      "eval_steps_per_second": 9.497,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8352732793522267,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.5058198380566802e-05,
+      "loss": 2.5604,
+      "step": 3301
+    },
+    {
+      "epoch": 0.8355263157894737,
+      "grad_norm": 11.5625,
+      "learning_rate": 2.5065789473684212e-05,
+      "loss": 2.4907,
+      "step": 3302
+    },
+    {
+      "epoch": 0.8357793522267206,
+      "grad_norm": 3.25,
+      "learning_rate": 2.507338056680162e-05,
+      "loss": 2.6611,
+      "step": 3303
+    },
+    {
+      "epoch": 0.8360323886639676,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.508097165991903e-05,
+      "loss": 2.736,
+      "step": 3304
+    },
+    {
+      "epoch": 0.8362854251012146,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.508856275303644e-05,
+      "loss": 2.6213,
+      "step": 3305
+    },
+    {
+      "epoch": 0.8365384615384616,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.509615384615385e-05,
+      "loss": 2.5736,
+      "step": 3306
+    },
+    {
+      "epoch": 0.8367914979757085,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.5103744939271256e-05,
+      "loss": 2.4639,
+      "step": 3307
+    },
+    {
+      "epoch": 0.8370445344129555,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.5111336032388663e-05,
+      "loss": 2.6258,
+      "step": 3308
+    },
+    {
+      "epoch": 0.8372975708502024,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.5118927125506073e-05,
+      "loss": 2.6493,
+      "step": 3309
+    },
+    {
+      "epoch": 0.8375506072874493,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.512651821862348e-05,
+      "loss": 2.6649,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8378036437246964,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.5134109311740893e-05,
+      "loss": 2.7054,
+      "step": 3311
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.51417004048583e-05,
+      "loss": 2.6246,
+      "step": 3312
+    },
+    {
+      "epoch": 0.8383097165991903,
+      "grad_norm": 2.875,
+      "learning_rate": 2.514929149797571e-05,
+      "loss": 2.5766,
+      "step": 3313
+    },
+    {
+      "epoch": 0.8385627530364372,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.5156882591093117e-05,
+      "loss": 2.6652,
+      "step": 3314
+    },
+    {
+      "epoch": 0.8388157894736842,
+      "grad_norm": 3.25,
+      "learning_rate": 2.5164473684210527e-05,
+      "loss": 2.661,
+      "step": 3315
+    },
+    {
+      "epoch": 0.8390688259109311,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.5172064777327933e-05,
+      "loss": 2.5871,
+      "step": 3316
+    },
+    {
+      "epoch": 0.8393218623481782,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.5179655870445347e-05,
+      "loss": 2.6887,
+      "step": 3317
+    },
+    {
+      "epoch": 0.8395748987854251,
+      "grad_norm": 3.125,
+      "learning_rate": 2.5187246963562754e-05,
+      "loss": 2.7186,
+      "step": 3318
+    },
+    {
+      "epoch": 0.8398279352226721,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.5194838056680164e-05,
+      "loss": 2.6459,
+      "step": 3319
+    },
+    {
+      "epoch": 0.840080971659919,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.520242914979757e-05,
+      "loss": 2.4653,
+      "step": 3320
+    },
+    {
+      "epoch": 0.840334008097166,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.521002024291498e-05,
+      "loss": 2.6475,
+      "step": 3321
+    },
+    {
+      "epoch": 0.8405870445344129,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.5217611336032387e-05,
+      "loss": 2.596,
+      "step": 3322
+    },
+    {
+      "epoch": 0.84084008097166,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.52252024291498e-05,
+      "loss": 2.5841,
+      "step": 3323
+    },
+    {
+      "epoch": 0.8410931174089069,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.5232793522267208e-05,
+      "loss": 2.7541,
+      "step": 3324
+    },
+    {
+      "epoch": 0.8413461538461539,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.5240384615384618e-05,
+      "loss": 2.5862,
+      "step": 3325
+    },
+    {
+      "epoch": 0.8415991902834008,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.5247975708502024e-05,
+      "loss": 2.5889,
+      "step": 3326
+    },
+    {
+      "epoch": 0.8418522267206477,
+      "grad_norm": 3.125,
+      "learning_rate": 2.5255566801619435e-05,
+      "loss": 2.5138,
+      "step": 3327
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.526315789473684e-05,
+      "loss": 2.5977,
+      "step": 3328
+    },
+    {
+      "epoch": 0.8423582995951417,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.527074898785425e-05,
+      "loss": 2.7739,
+      "step": 3329
+    },
+    {
+      "epoch": 0.8426113360323887,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.527834008097166e-05,
+      "loss": 2.5754,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8426113360323887,
+      "eval_loss": 2.7215864658355713,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.2327,
+      "eval_samples_per_second": 9.516,
+      "eval_steps_per_second": 9.516,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8428643724696356,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.5285931174089068e-05,
+      "loss": 2.6762,
+      "step": 3331
+    },
+    {
+      "epoch": 0.8431174089068826,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.529352226720648e-05,
+      "loss": 2.7099,
+      "step": 3332
+    },
+    {
+      "epoch": 0.8433704453441295,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.5301113360323885e-05,
+      "loss": 2.6255,
+      "step": 3333
+    },
+    {
+      "epoch": 0.8436234817813765,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.5308704453441295e-05,
+      "loss": 2.6949,
+      "step": 3334
+    },
+    {
+      "epoch": 0.8438765182186235,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.5316295546558705e-05,
+      "loss": 2.691,
+      "step": 3335
+    },
+    {
+      "epoch": 0.8441295546558705,
+      "grad_norm": 3.875,
+      "learning_rate": 2.5323886639676116e-05,
+      "loss": 2.654,
+      "step": 3336
+    },
+    {
+      "epoch": 0.8443825910931174,
+      "grad_norm": 3.25,
+      "learning_rate": 2.5331477732793522e-05,
+      "loss": 2.642,
+      "step": 3337
+    },
+    {
+      "epoch": 0.8446356275303644,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.5339068825910932e-05,
+      "loss": 2.692,
+      "step": 3338
+    },
+    {
+      "epoch": 0.8448886639676113,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.534665991902834e-05,
+      "loss": 2.7067,
+      "step": 3339
+    },
+    {
+      "epoch": 0.8451417004048583,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.535425101214575e-05,
+      "loss": 2.5796,
+      "step": 3340
+    },
+    {
+      "epoch": 0.8453947368421053,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.536184210526316e-05,
+      "loss": 2.7225,
+      "step": 3341
+    },
+    {
+      "epoch": 0.8456477732793523,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.536943319838057e-05,
+      "loss": 2.6224,
+      "step": 3342
+    },
+    {
+      "epoch": 0.8459008097165992,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.5377024291497976e-05,
+      "loss": 2.6139,
+      "step": 3343
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.5384615384615386e-05,
+      "loss": 2.6268,
+      "step": 3344
+    },
+    {
+      "epoch": 0.8464068825910931,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.5392206477732793e-05,
+      "loss": 2.6779,
+      "step": 3345
+    },
+    {
+      "epoch": 0.84665991902834,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.5399797570850203e-05,
+      "loss": 2.6981,
+      "step": 3346
+    },
+    {
+      "epoch": 0.8469129554655871,
+      "grad_norm": 3.0,
+      "learning_rate": 2.5407388663967613e-05,
+      "loss": 2.561,
+      "step": 3347
+    },
+    {
+      "epoch": 0.847165991902834,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.5414979757085023e-05,
+      "loss": 2.6115,
+      "step": 3348
+    },
+    {
+      "epoch": 0.847419028340081,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.542257085020243e-05,
+      "loss": 2.653,
+      "step": 3349
+    },
+    {
+      "epoch": 0.8476720647773279,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.5430161943319837e-05,
+      "loss": 2.6776,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8479251012145749,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.5437753036437247e-05,
+      "loss": 2.5684,
+      "step": 3351
+    },
+    {
+      "epoch": 0.8481781376518218,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.5445344129554654e-05,
+      "loss": 2.6272,
+      "step": 3352
+    },
+    {
+      "epoch": 0.8484311740890689,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.5452935222672067e-05,
+      "loss": 2.559,
+      "step": 3353
+    },
+    {
+      "epoch": 0.8486842105263158,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.5460526315789474e-05,
+      "loss": 2.66,
+      "step": 3354
+    },
+    {
+      "epoch": 0.8489372469635628,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.5468117408906884e-05,
+      "loss": 2.6439,
+      "step": 3355
+    },
+    {
+      "epoch": 0.8491902834008097,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.547570850202429e-05,
+      "loss": 2.5904,
+      "step": 3356
+    },
+    {
+      "epoch": 0.8494433198380567,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.54832995951417e-05,
+      "loss": 2.5135,
+      "step": 3357
+    },
+    {
+      "epoch": 0.8496963562753036,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.5490890688259108e-05,
+      "loss": 2.6052,
+      "step": 3358
+    },
+    {
+      "epoch": 0.8499493927125507,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.549848178137652e-05,
+      "loss": 2.7415,
+      "step": 3359
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.5506072874493928e-05,
+      "loss": 2.5845,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "eval_loss": 2.712719440460205,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.8913,
+      "eval_samples_per_second": 9.493,
+      "eval_steps_per_second": 9.493,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8504554655870445,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.5513663967611338e-05,
+      "loss": 2.5788,
+      "step": 3361
+    },
+    {
+      "epoch": 0.8507085020242915,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.5521255060728745e-05,
+      "loss": 2.6442,
+      "step": 3362
+    },
+    {
+      "epoch": 0.8509615384615384,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.5528846153846155e-05,
+      "loss": 2.5689,
+      "step": 3363
+    },
+    {
+      "epoch": 0.8512145748987854,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.553643724696356e-05,
+      "loss": 2.5644,
+      "step": 3364
+    },
+    {
+      "epoch": 0.8514676113360324,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.5544028340080975e-05,
+      "loss": 2.626,
+      "step": 3365
+    },
+    {
+      "epoch": 0.8517206477732794,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.5551619433198382e-05,
+      "loss": 2.6063,
+      "step": 3366
+    },
+    {
+      "epoch": 0.8519736842105263,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.5559210526315792e-05,
+      "loss": 2.4552,
+      "step": 3367
+    },
+    {
+      "epoch": 0.8522267206477733,
+      "grad_norm": 3.375,
+      "learning_rate": 2.55668016194332e-05,
+      "loss": 2.6015,
+      "step": 3368
+    },
+    {
+      "epoch": 0.8524797570850202,
+      "grad_norm": 3.125,
+      "learning_rate": 2.557439271255061e-05,
+      "loss": 2.5908,
+      "step": 3369
+    },
+    {
+      "epoch": 0.8527327935222672,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.5581983805668016e-05,
+      "loss": 2.4084,
+      "step": 3370
+    },
+    {
+      "epoch": 0.8529858299595142,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.5589574898785426e-05,
+      "loss": 2.6872,
+      "step": 3371
+    },
+    {
+      "epoch": 0.8532388663967612,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.5597165991902836e-05,
+      "loss": 2.6854,
+      "step": 3372
+    },
+    {
+      "epoch": 0.8534919028340081,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.5604757085020243e-05,
+      "loss": 2.7338,
+      "step": 3373
+    },
+    {
+      "epoch": 0.853744939271255,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.5612348178137653e-05,
+      "loss": 2.6674,
+      "step": 3374
+    },
+    {
+      "epoch": 0.853997975708502,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.561993927125506e-05,
+      "loss": 2.7458,
+      "step": 3375
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 3.5,
+      "learning_rate": 2.562753036437247e-05,
+      "loss": 2.6589,
+      "step": 3376
+    },
+    {
+      "epoch": 0.854504048582996,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.563512145748988e-05,
+      "loss": 2.5972,
+      "step": 3377
+    },
+    {
+      "epoch": 0.854757085020243,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.564271255060729e-05,
+      "loss": 2.6199,
+      "step": 3378
+    },
+    {
+      "epoch": 0.8550101214574899,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.5650303643724696e-05,
+      "loss": 2.6506,
+      "step": 3379
+    },
+    {
+      "epoch": 0.8552631578947368,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.5657894736842107e-05,
+      "loss": 2.7615,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8555161943319838,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.5665485829959513e-05,
+      "loss": 2.689,
+      "step": 3381
+    },
+    {
+      "epoch": 0.8557692307692307,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.5673076923076923e-05,
+      "loss": 2.633,
+      "step": 3382
+    },
+    {
+      "epoch": 0.8560222672064778,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.5680668016194334e-05,
+      "loss": 2.6855,
+      "step": 3383
+    },
+    {
+      "epoch": 0.8562753036437247,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.5688259109311744e-05,
+      "loss": 2.7124,
+      "step": 3384
+    },
+    {
+      "epoch": 0.8565283400809717,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.569585020242915e-05,
+      "loss": 2.6903,
+      "step": 3385
+    },
+    {
+      "epoch": 0.8567813765182186,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.570344129554656e-05,
+      "loss": 2.5831,
+      "step": 3386
+    },
+    {
+      "epoch": 0.8570344129554656,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.5711032388663967e-05,
+      "loss": 2.7171,
+      "step": 3387
+    },
+    {
+      "epoch": 0.8572874493927125,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.5718623481781377e-05,
+      "loss": 2.5662,
+      "step": 3388
+    },
+    {
+      "epoch": 0.8575404858299596,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.5726214574898787e-05,
+      "loss": 2.6321,
+      "step": 3389
+    },
+    {
+      "epoch": 0.8577935222672065,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.5733805668016198e-05,
+      "loss": 2.7918,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8577935222672065,
+      "eval_loss": 2.7117137908935547,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.8749,
+      "eval_samples_per_second": 9.528,
+      "eval_steps_per_second": 9.528,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8580465587044535,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.5741396761133604e-05,
+      "loss": 2.6046,
+      "step": 3391
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.574898785425101e-05,
+      "loss": 2.5441,
+      "step": 3392
+    },
+    {
+      "epoch": 0.8585526315789473,
+      "grad_norm": 3.125,
+      "learning_rate": 2.575657894736842e-05,
+      "loss": 2.6764,
+      "step": 3393
+    },
+    {
+      "epoch": 0.8588056680161943,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.5764170040485828e-05,
+      "loss": 2.7157,
+      "step": 3394
+    },
+    {
+      "epoch": 0.8590587044534413,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.577176113360324e-05,
+      "loss": 2.6619,
+      "step": 3395
+    },
+    {
+      "epoch": 0.8593117408906883,
+      "grad_norm": 24.5,
+      "learning_rate": 2.5779352226720648e-05,
+      "loss": 2.708,
+      "step": 3396
+    },
+    {
+      "epoch": 0.8595647773279352,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.5786943319838058e-05,
+      "loss": 2.6838,
+      "step": 3397
+    },
+    {
+      "epoch": 0.8598178137651822,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.5794534412955465e-05,
+      "loss": 2.7312,
+      "step": 3398
+    },
+    {
+      "epoch": 0.8600708502024291,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.5802125506072875e-05,
+      "loss": 2.6206,
+      "step": 3399
+    },
+    {
+      "epoch": 0.8603238866396761,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.5809716599190282e-05,
+      "loss": 2.5343,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8605769230769231,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.5817307692307695e-05,
+      "loss": 2.527,
+      "step": 3401
+    },
+    {
+      "epoch": 0.8608299595141701,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.5824898785425102e-05,
+      "loss": 2.6705,
+      "step": 3402
+    },
+    {
+      "epoch": 0.861082995951417,
+      "grad_norm": 3.375,
+      "learning_rate": 2.5832489878542512e-05,
+      "loss": 2.578,
+      "step": 3403
+    },
+    {
+      "epoch": 0.861336032388664,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.584008097165992e-05,
+      "loss": 2.5653,
+      "step": 3404
+    },
+    {
+      "epoch": 0.8615890688259109,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.584767206477733e-05,
+      "loss": 2.5684,
+      "step": 3405
+    },
+    {
+      "epoch": 0.8618421052631579,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.5855263157894736e-05,
+      "loss": 2.7302,
+      "step": 3406
+    },
+    {
+      "epoch": 0.8620951417004049,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.586285425101215e-05,
+      "loss": 2.66,
+      "step": 3407
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 3.375,
+      "learning_rate": 2.5870445344129556e-05,
+      "loss": 2.6089,
+      "step": 3408
+    },
+    {
+      "epoch": 0.8626012145748988,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.5878036437246966e-05,
+      "loss": 2.5895,
+      "step": 3409
+    },
+    {
+      "epoch": 0.8628542510121457,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.5885627530364373e-05,
+      "loss": 2.6473,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8631072874493927,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.5893218623481783e-05,
+      "loss": 2.6968,
+      "step": 3411
+    },
+    {
+      "epoch": 0.8633603238866396,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.590080971659919e-05,
+      "loss": 2.6519,
+      "step": 3412
+    },
+    {
+      "epoch": 0.8636133603238867,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.59084008097166e-05,
+      "loss": 2.5916,
+      "step": 3413
+    },
+    {
+      "epoch": 0.8638663967611336,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.591599190283401e-05,
+      "loss": 2.6935,
+      "step": 3414
+    },
+    {
+      "epoch": 0.8641194331983806,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.5923582995951417e-05,
+      "loss": 2.6232,
+      "step": 3415
+    },
+    {
+      "epoch": 0.8643724696356275,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.5931174089068827e-05,
+      "loss": 2.6706,
+      "step": 3416
+    },
+    {
+      "epoch": 0.8646255060728745,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.5938765182186234e-05,
+      "loss": 2.617,
+      "step": 3417
+    },
+    {
+      "epoch": 0.8648785425101214,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.5946356275303644e-05,
+      "loss": 2.594,
+      "step": 3418
+    },
+    {
+      "epoch": 0.8651315789473685,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.5953947368421054e-05,
+      "loss": 2.5196,
+      "step": 3419
+    },
+    {
+      "epoch": 0.8653846153846154,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.5961538461538464e-05,
+      "loss": 2.589,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8653846153846154,
+      "eval_loss": 2.7065744400024414,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.6521,
+      "eval_samples_per_second": 9.501,
+      "eval_steps_per_second": 9.501,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8656376518218624,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.596912955465587e-05,
+      "loss": 2.6553,
+      "step": 3421
+    },
+    {
+      "epoch": 0.8658906882591093,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.597672064777328e-05,
+      "loss": 2.4889,
+      "step": 3422
+    },
+    {
+      "epoch": 0.8661437246963563,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.5984311740890688e-05,
+      "loss": 2.6928,
+      "step": 3423
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.5991902834008098e-05,
+      "loss": 2.5934,
+      "step": 3424
+    },
+    {
+      "epoch": 0.8666497975708503,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.5999493927125508e-05,
+      "loss": 2.6516,
+      "step": 3425
+    },
+    {
+      "epoch": 0.8669028340080972,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.6007085020242918e-05,
+      "loss": 2.6627,
+      "step": 3426
+    },
+    {
+      "epoch": 0.8671558704453441,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.6014676113360325e-05,
+      "loss": 2.6447,
+      "step": 3427
+    },
+    {
+      "epoch": 0.8674089068825911,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.6022267206477735e-05,
+      "loss": 2.7226,
+      "step": 3428
+    },
+    {
+      "epoch": 0.867661943319838,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.602985829959514e-05,
+      "loss": 2.6891,
+      "step": 3429
+    },
+    {
+      "epoch": 0.867914979757085,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.603744939271255e-05,
+      "loss": 2.682,
+      "step": 3430
+    },
+    {
+      "epoch": 0.868168016194332,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.6045040485829962e-05,
+      "loss": 2.6236,
+      "step": 3431
+    },
+    {
+      "epoch": 0.868421052631579,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.605263157894737e-05,
+      "loss": 2.6599,
+      "step": 3432
+    },
+    {
+      "epoch": 0.8686740890688259,
+      "grad_norm": 3.625,
+      "learning_rate": 2.606022267206478e-05,
+      "loss": 2.6443,
+      "step": 3433
+    },
+    {
+      "epoch": 0.8689271255060729,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.6067813765182185e-05,
+      "loss": 2.639,
+      "step": 3434
+    },
+    {
+      "epoch": 0.8691801619433198,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.6075404858299595e-05,
+      "loss": 2.6636,
+      "step": 3435
+    },
+    {
+      "epoch": 0.8694331983805668,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.6082995951417002e-05,
+      "loss": 2.6951,
+      "step": 3436
+    },
+    {
+      "epoch": 0.8696862348178138,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.6090587044534416e-05,
+      "loss": 2.6553,
+      "step": 3437
+    },
+    {
+      "epoch": 0.8699392712550608,
+      "grad_norm": 3.125,
+      "learning_rate": 2.6098178137651822e-05,
+      "loss": 2.6431,
+      "step": 3438
+    },
+    {
+      "epoch": 0.8701923076923077,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.6105769230769233e-05,
+      "loss": 2.5646,
+      "step": 3439
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 3.375,
+      "learning_rate": 2.611336032388664e-05,
+      "loss": 2.6135,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8706983805668016,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.612095141700405e-05,
+      "loss": 2.6298,
+      "step": 3441
+    },
+    {
+      "epoch": 0.8709514170040485,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.6128542510121456e-05,
+      "loss": 2.7184,
+      "step": 3442
+    },
+    {
+      "epoch": 0.8712044534412956,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.613613360323887e-05,
+      "loss": 2.5679,
+      "step": 3443
+    },
+    {
+      "epoch": 0.8714574898785425,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.6143724696356276e-05,
+      "loss": 2.6353,
+      "step": 3444
+    },
+    {
+      "epoch": 0.8717105263157895,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.6151315789473686e-05,
+      "loss": 2.6035,
+      "step": 3445
+    },
+    {
+      "epoch": 0.8719635627530364,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.6158906882591093e-05,
+      "loss": 2.7376,
+      "step": 3446
+    },
+    {
+      "epoch": 0.8722165991902834,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.6166497975708503e-05,
+      "loss": 2.7411,
+      "step": 3447
+    },
+    {
+      "epoch": 0.8724696356275303,
+      "grad_norm": 3.375,
+      "learning_rate": 2.617408906882591e-05,
+      "loss": 2.5782,
+      "step": 3448
+    },
+    {
+      "epoch": 0.8727226720647774,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.6181680161943324e-05,
+      "loss": 2.6588,
+      "step": 3449
+    },
+    {
+      "epoch": 0.8729757085020243,
+      "grad_norm": 3.0,
+      "learning_rate": 2.618927125506073e-05,
+      "loss": 2.6353,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8729757085020243,
+      "eval_loss": 2.707172155380249,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.8766,
+      "eval_samples_per_second": 9.528,
+      "eval_steps_per_second": 9.528,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8732287449392713,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.619686234817814e-05,
+      "loss": 2.5364,
+      "step": 3451
+    },
+    {
+      "epoch": 0.8734817813765182,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.6204453441295547e-05,
+      "loss": 2.5979,
+      "step": 3452
+    },
+    {
+      "epoch": 0.8737348178137652,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.6212044534412954e-05,
+      "loss": 2.6045,
+      "step": 3453
+    },
+    {
+      "epoch": 0.8739878542510121,
+      "grad_norm": 3.125,
+      "learning_rate": 2.6219635627530364e-05,
+      "loss": 2.7273,
+      "step": 3454
+    },
+    {
+      "epoch": 0.8742408906882592,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.6227226720647774e-05,
+      "loss": 2.7638,
+      "step": 3455
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.6234817813765184e-05,
+      "loss": 2.5415,
+      "step": 3456
+    },
+    {
+      "epoch": 0.874746963562753,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.624240890688259e-05,
+      "loss": 2.5306,
+      "step": 3457
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.625e-05,
+      "loss": 2.5947,
+      "step": 3458
+    },
+    {
+      "epoch": 0.875253036437247,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.6257591093117408e-05,
+      "loss": 2.5988,
+      "step": 3459
+    },
+    {
+      "epoch": 0.8755060728744939,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.6265182186234818e-05,
+      "loss": 2.6205,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8757591093117408,
+      "grad_norm": 3.5,
+      "learning_rate": 2.6272773279352225e-05,
+      "loss": 2.6582,
+      "step": 3461
+    },
+    {
+      "epoch": 0.8760121457489879,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.6280364372469638e-05,
+      "loss": 2.5973,
+      "step": 3462
+    },
+    {
+      "epoch": 0.8762651821862348,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.6287955465587045e-05,
+      "loss": 2.6606,
+      "step": 3463
+    },
+    {
+      "epoch": 0.8765182186234818,
+      "grad_norm": 3.25,
+      "learning_rate": 2.6295546558704455e-05,
+      "loss": 2.6677,
+      "step": 3464
+    },
+    {
+      "epoch": 0.8767712550607287,
+      "grad_norm": 3.125,
+      "learning_rate": 2.6303137651821862e-05,
+      "loss": 2.6471,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8770242914979757,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.6310728744939272e-05,
+      "loss": 2.644,
+      "step": 3466
+    },
+    {
+      "epoch": 0.8772773279352226,
+      "grad_norm": 3.0,
+      "learning_rate": 2.631831983805668e-05,
+      "loss": 2.454,
+      "step": 3467
+    },
+    {
+      "epoch": 0.8775303643724697,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.6325910931174092e-05,
+      "loss": 2.6516,
+      "step": 3468
+    },
+    {
+      "epoch": 0.8777834008097166,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.63335020242915e-05,
+      "loss": 2.6352,
+      "step": 3469
+    },
+    {
+      "epoch": 0.8780364372469636,
+      "grad_norm": 3.375,
+      "learning_rate": 2.634109311740891e-05,
+      "loss": 2.5892,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8782894736842105,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.6348684210526316e-05,
+      "loss": 2.5626,
+      "step": 3471
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.6356275303643726e-05,
+      "loss": 2.5418,
+      "step": 3472
+    },
+    {
+      "epoch": 0.8787955465587044,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.6363866396761133e-05,
+      "loss": 2.7036,
+      "step": 3473
+    },
+    {
+      "epoch": 0.8790485829959515,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.6371457489878543e-05,
+      "loss": 2.6823,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8793016194331984,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.6379048582995953e-05,
+      "loss": 2.6029,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8795546558704453,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.638663967611336e-05,
+      "loss": 2.5555,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8798076923076923,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.639423076923077e-05,
+      "loss": 2.7367,
+      "step": 3477
+    },
+    {
+      "epoch": 0.8800607287449392,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.6401821862348176e-05,
+      "loss": 2.6418,
+      "step": 3478
+    },
+    {
+      "epoch": 0.8803137651821862,
+      "grad_norm": 2.875,
+      "learning_rate": 2.6409412955465587e-05,
+      "loss": 2.5136,
+      "step": 3479
+    },
+    {
+      "epoch": 0.8805668016194332,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.6417004048582997e-05,
+      "loss": 2.5513,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8805668016194332,
+      "eval_loss": 2.708397388458252,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.4766,
+      "eval_samples_per_second": 9.473,
+      "eval_steps_per_second": 9.473,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8808198380566802,
+      "grad_norm": 3.375,
+      "learning_rate": 2.6424595141700407e-05,
+      "loss": 2.6116,
+      "step": 3481
+    },
+    {
+      "epoch": 0.8810728744939271,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.6432186234817813e-05,
+      "loss": 2.6512,
+      "step": 3482
+    },
+    {
+      "epoch": 0.8813259109311741,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.6439777327935224e-05,
+      "loss": 2.6236,
+      "step": 3483
+    },
+    {
+      "epoch": 0.881578947368421,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.644736842105263e-05,
+      "loss": 2.6207,
+      "step": 3484
+    },
+    {
+      "epoch": 0.881831983805668,
+      "grad_norm": 3.625,
+      "learning_rate": 2.645495951417004e-05,
+      "loss": 2.6522,
+      "step": 3485
+    },
+    {
+      "epoch": 0.882085020242915,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.646255060728745e-05,
+      "loss": 2.4925,
+      "step": 3486
+    },
+    {
+      "epoch": 0.882338056680162,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.647014170040486e-05,
+      "loss": 2.5908,
+      "step": 3487
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.6477732793522267e-05,
+      "loss": 2.6808,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8828441295546559,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.6485323886639678e-05,
+      "loss": 2.5424,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8830971659919028,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.6492914979757084e-05,
+      "loss": 2.5802,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8833502024291497,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.6500506072874494e-05,
+      "loss": 2.7062,
+      "step": 3491
+    },
+    {
+      "epoch": 0.8836032388663968,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.6508097165991905e-05,
+      "loss": 2.5654,
+      "step": 3492
+    },
+    {
+      "epoch": 0.8838562753036437,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.6515688259109315e-05,
+      "loss": 2.6993,
+      "step": 3493
+    },
+    {
+      "epoch": 0.8841093117408907,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.652327935222672e-05,
+      "loss": 2.6376,
+      "step": 3494
+    },
+    {
+      "epoch": 0.8843623481781376,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.6530870445344128e-05,
+      "loss": 2.75,
+      "step": 3495
+    },
+    {
+      "epoch": 0.8846153846153846,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.6538461538461538e-05,
+      "loss": 2.6001,
+      "step": 3496
+    },
+    {
+      "epoch": 0.8848684210526315,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.6546052631578945e-05,
+      "loss": 2.675,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8851214574898786,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.655364372469636e-05,
+      "loss": 2.5406,
+      "step": 3498
+    },
+    {
+      "epoch": 0.8853744939271255,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.6561234817813765e-05,
+      "loss": 2.5378,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8856275303643725,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.6568825910931175e-05,
+      "loss": 2.6099,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8858805668016194,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.6576417004048582e-05,
+      "loss": 2.6668,
+      "step": 3501
+    },
+    {
+      "epoch": 0.8861336032388664,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.6584008097165992e-05,
+      "loss": 2.5763,
+      "step": 3502
+    },
+    {
+      "epoch": 0.8863866396761133,
+      "grad_norm": 19.25,
+      "learning_rate": 2.65915991902834e-05,
+      "loss": 2.8718,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.6599190283400812e-05,
+      "loss": 2.6512,
+      "step": 3504
+    },
+    {
+      "epoch": 0.8868927125506073,
+      "grad_norm": 3.5,
+      "learning_rate": 2.660678137651822e-05,
+      "loss": 2.7142,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8871457489878543,
+      "grad_norm": 3.75,
+      "learning_rate": 2.661437246963563e-05,
+      "loss": 2.6472,
+      "step": 3506
+    },
+    {
+      "epoch": 0.8873987854251012,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.6621963562753036e-05,
+      "loss": 2.63,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8876518218623481,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.6629554655870446e-05,
+      "loss": 2.5757,
+      "step": 3508
+    },
+    {
+      "epoch": 0.8879048582995951,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.6637145748987853e-05,
+      "loss": 2.6057,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8881578947368421,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.6644736842105266e-05,
+      "loss": 2.6503,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8881578947368421,
+      "eval_loss": 2.70592999458313,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 291.5969,
+      "eval_samples_per_second": 8.916,
+      "eval_steps_per_second": 8.916,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8884109311740891,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.6652327935222673e-05,
+      "loss": 2.6583,
+      "step": 3511
+    },
+    {
+      "epoch": 0.888663967611336,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.6659919028340083e-05,
+      "loss": 2.6652,
+      "step": 3512
+    },
+    {
+      "epoch": 0.888917004048583,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.666751012145749e-05,
+      "loss": 2.6631,
+      "step": 3513
+    },
+    {
+      "epoch": 0.8891700404858299,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.66751012145749e-05,
+      "loss": 2.6801,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8894230769230769,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.6682692307692307e-05,
+      "loss": 2.6361,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8896761133603239,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.6690283400809717e-05,
+      "loss": 2.6205,
+      "step": 3516
+    },
+    {
+      "epoch": 0.8899291497975709,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.6697874493927127e-05,
+      "loss": 2.5886,
+      "step": 3517
+    },
+    {
+      "epoch": 0.8901821862348178,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.6705465587044534e-05,
+      "loss": 2.6754,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8904352226720648,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.6713056680161944e-05,
+      "loss": 2.6026,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 3.125,
+      "learning_rate": 2.672064777327935e-05,
+      "loss": 2.651,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8909412955465587,
+      "grad_norm": 3.5,
+      "learning_rate": 2.672823886639676e-05,
+      "loss": 2.6491,
+      "step": 3521
+    },
+    {
+      "epoch": 0.8911943319838057,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.673582995951417e-05,
+      "loss": 2.4485,
+      "step": 3522
+    },
+    {
+      "epoch": 0.8914473684210527,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.674342105263158e-05,
+      "loss": 2.4341,
+      "step": 3523
+    },
+    {
+      "epoch": 0.8917004048582996,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.6751012145748988e-05,
+      "loss": 2.6622,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8919534412955465,
+      "grad_norm": 3.875,
+      "learning_rate": 2.6758603238866398e-05,
+      "loss": 2.5638,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8922064777327935,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.6766194331983805e-05,
+      "loss": 2.6337,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8924595141700404,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.6773785425101215e-05,
+      "loss": 2.5664,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8927125506072875,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.6781376518218625e-05,
+      "loss": 2.5342,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8929655870445344,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.6788967611336035e-05,
+      "loss": 2.6602,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8932186234817814,
+      "grad_norm": 2.875,
+      "learning_rate": 2.679655870445344e-05,
+      "loss": 2.5177,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8934716599190283,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.6804149797570852e-05,
+      "loss": 2.6129,
+      "step": 3531
+    },
+    {
+      "epoch": 0.8937246963562753,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.681174089068826e-05,
+      "loss": 2.674,
+      "step": 3532
+    },
+    {
+      "epoch": 0.8939777327935222,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.681933198380567e-05,
+      "loss": 2.6136,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8942307692307693,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.682692307692308e-05,
+      "loss": 2.6421,
+      "step": 3534
+    },
+    {
+      "epoch": 0.8944838056680162,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.683451417004049e-05,
+      "loss": 2.7034,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.6842105263157896e-05,
+      "loss": 2.6488,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8949898785425101,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.6849696356275302e-05,
+      "loss": 2.5549,
+      "step": 3537
+    },
+    {
+      "epoch": 0.895242914979757,
+      "grad_norm": 3.125,
+      "learning_rate": 2.6857287449392712e-05,
+      "loss": 2.5861,
+      "step": 3538
+    },
+    {
+      "epoch": 0.895495951417004,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.686487854251012e-05,
+      "loss": 2.5518,
+      "step": 3539
+    },
+    {
+      "epoch": 0.895748987854251,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.6872469635627533e-05,
+      "loss": 2.7215,
+      "step": 3540
+    },
+    {
+      "epoch": 0.895748987854251,
+      "eval_loss": 2.7047250270843506,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.8442,
+      "eval_samples_per_second": 9.494,
+      "eval_steps_per_second": 9.494,
+      "step": 3540
+    },
+    {
+      "epoch": 0.896002024291498,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.688006072874494e-05,
+      "loss": 2.5625,
+      "step": 3541
+    },
+    {
+      "epoch": 0.896255060728745,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.688765182186235e-05,
+      "loss": 2.6913,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8965080971659919,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.6895242914979756e-05,
+      "loss": 2.6555,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8967611336032388,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.6902834008097166e-05,
+      "loss": 2.402,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8970141700404858,
+      "grad_norm": 3.25,
+      "learning_rate": 2.6910425101214573e-05,
+      "loss": 2.6175,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8972672064777328,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.6918016194331987e-05,
+      "loss": 2.5939,
+      "step": 3546
+    },
+    {
+      "epoch": 0.8975202429149798,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.6925607287449393e-05,
+      "loss": 2.6434,
+      "step": 3547
+    },
+    {
+      "epoch": 0.8977732793522267,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.6933198380566803e-05,
+      "loss": 2.6643,
+      "step": 3548
+    },
+    {
+      "epoch": 0.8980263157894737,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.694078947368421e-05,
+      "loss": 2.6218,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8982793522267206,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.694838056680162e-05,
+      "loss": 2.6279,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8985323886639676,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.6955971659919027e-05,
+      "loss": 2.605,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.696356275303644e-05,
+      "loss": 2.636,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8990384615384616,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.6971153846153847e-05,
+      "loss": 2.6375,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8992914979757085,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.6978744939271257e-05,
+      "loss": 2.6116,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8995445344129555,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.6986336032388664e-05,
+      "loss": 2.6603,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8997975708502024,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.6993927125506074e-05,
+      "loss": 2.5667,
+      "step": 3556
+    },
+    {
+      "epoch": 0.9000506072874493,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.700151821862348e-05,
+      "loss": 2.5898,
+      "step": 3557
+    },
+    {
+      "epoch": 0.9003036437246964,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.700910931174089e-05,
+      "loss": 2.5338,
+      "step": 3558
+    },
+    {
+      "epoch": 0.9005566801619433,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.70167004048583e-05,
+      "loss": 2.5866,
+      "step": 3559
+    },
+    {
+      "epoch": 0.9008097165991903,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.7024291497975708e-05,
+      "loss": 2.5661,
+      "step": 3560
+    },
+    {
+      "epoch": 0.9010627530364372,
+      "grad_norm": 3.25,
+      "learning_rate": 2.7031882591093118e-05,
+      "loss": 2.7327,
+      "step": 3561
+    },
+    {
+      "epoch": 0.9013157894736842,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.7039473684210525e-05,
+      "loss": 2.6416,
+      "step": 3562
+    },
+    {
+      "epoch": 0.9015688259109311,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.7047064777327935e-05,
+      "loss": 2.6637,
+      "step": 3563
+    },
+    {
+      "epoch": 0.9018218623481782,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.7054655870445345e-05,
+      "loss": 2.4248,
+      "step": 3564
+    },
+    {
+      "epoch": 0.9020748987854251,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.7062246963562755e-05,
+      "loss": 2.6011,
+      "step": 3565
+    },
+    {
+      "epoch": 0.9023279352226721,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.7069838056680162e-05,
+      "loss": 2.5859,
+      "step": 3566
+    },
+    {
+      "epoch": 0.902580971659919,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.7077429149797572e-05,
+      "loss": 2.6601,
+      "step": 3567
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.708502024291498e-05,
+      "loss": 2.4959,
+      "step": 3568
+    },
+    {
+      "epoch": 0.9030870445344129,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.709261133603239e-05,
+      "loss": 2.6432,
+      "step": 3569
+    },
+    {
+      "epoch": 0.90334008097166,
+      "grad_norm": 3.0,
+      "learning_rate": 2.71002024291498e-05,
+      "loss": 2.6309,
+      "step": 3570
+    },
+    {
+      "epoch": 0.90334008097166,
+      "eval_loss": 2.6983773708343506,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.095,
+      "eval_samples_per_second": 9.52,
+      "eval_steps_per_second": 9.52,
+      "step": 3570
+    },
+    {
+      "epoch": 0.9035931174089069,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.710779352226721e-05,
+      "loss": 2.5034,
+      "step": 3571
+    },
+    {
+      "epoch": 0.9038461538461539,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.7115384615384616e-05,
+      "loss": 2.625,
+      "step": 3572
+    },
+    {
+      "epoch": 0.9040991902834008,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.7122975708502026e-05,
+      "loss": 2.7253,
+      "step": 3573
+    },
+    {
+      "epoch": 0.9043522267206477,
+      "grad_norm": 3.625,
+      "learning_rate": 2.7130566801619433e-05,
+      "loss": 2.641,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9046052631578947,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.7138157894736843e-05,
+      "loss": 2.6846,
+      "step": 3575
+    },
+    {
+      "epoch": 0.9048582995951417,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.7145748987854253e-05,
+      "loss": 2.5593,
+      "step": 3576
+    },
+    {
+      "epoch": 0.9051113360323887,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.7153340080971663e-05,
+      "loss": 2.6793,
+      "step": 3577
+    },
+    {
+      "epoch": 0.9053643724696356,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.716093117408907e-05,
+      "loss": 2.672,
+      "step": 3578
+    },
+    {
+      "epoch": 0.9056174089068826,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.7168522267206477e-05,
+      "loss": 2.6168,
+      "step": 3579
+    },
+    {
+      "epoch": 0.9058704453441295,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.7176113360323887e-05,
+      "loss": 2.6193,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9061234817813765,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.7183704453441293e-05,
+      "loss": 2.6023,
+      "step": 3581
+    },
+    {
+      "epoch": 0.9063765182186235,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.7191295546558707e-05,
+      "loss": 2.6793,
+      "step": 3582
+    },
+    {
+      "epoch": 0.9066295546558705,
+      "grad_norm": 3.375,
+      "learning_rate": 2.7198886639676114e-05,
+      "loss": 2.6854,
+      "step": 3583
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.7206477732793524e-05,
+      "loss": 2.6914,
+      "step": 3584
+    },
+    {
+      "epoch": 0.9071356275303644,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.721406882591093e-05,
+      "loss": 2.5109,
+      "step": 3585
+    },
+    {
+      "epoch": 0.9073886639676113,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.722165991902834e-05,
+      "loss": 2.6537,
+      "step": 3586
+    },
+    {
+      "epoch": 0.9076417004048583,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.7229251012145747e-05,
+      "loss": 2.7258,
+      "step": 3587
+    },
+    {
+      "epoch": 0.9078947368421053,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.723684210526316e-05,
+      "loss": 2.3961,
+      "step": 3588
+    },
+    {
+      "epoch": 0.9081477732793523,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.7244433198380568e-05,
+      "loss": 2.6372,
+      "step": 3589
+    },
+    {
+      "epoch": 0.9084008097165992,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7252024291497978e-05,
+      "loss": 2.4976,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9086538461538461,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.7259615384615384e-05,
+      "loss": 2.6253,
+      "step": 3591
+    },
+    {
+      "epoch": 0.9089068825910931,
+      "grad_norm": 4.03125,
+      "learning_rate": 2.7267206477732795e-05,
+      "loss": 2.7333,
+      "step": 3592
+    },
+    {
+      "epoch": 0.90915991902834,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.72747975708502e-05,
+      "loss": 2.6592,
+      "step": 3593
+    },
+    {
+      "epoch": 0.9094129554655871,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.7282388663967615e-05,
+      "loss": 2.3379,
+      "step": 3594
+    },
+    {
+      "epoch": 0.909665991902834,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.728997975708502e-05,
+      "loss": 2.6244,
+      "step": 3595
+    },
+    {
+      "epoch": 0.909919028340081,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.729757085020243e-05,
+      "loss": 2.676,
+      "step": 3596
+    },
+    {
+      "epoch": 0.9101720647773279,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.730516194331984e-05,
+      "loss": 2.6403,
+      "step": 3597
+    },
+    {
+      "epoch": 0.9104251012145749,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.731275303643725e-05,
+      "loss": 2.6563,
+      "step": 3598
+    },
+    {
+      "epoch": 0.9106781376518218,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.7320344129554655e-05,
+      "loss": 2.6485,
+      "step": 3599
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 3.625,
+      "learning_rate": 2.7327935222672065e-05,
+      "loss": 2.6559,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "eval_loss": 2.6967318058013916,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.2138,
+      "eval_samples_per_second": 9.516,
+      "eval_steps_per_second": 9.516,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9111842105263158,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.7335526315789475e-05,
+      "loss": 2.729,
+      "step": 3601
+    },
+    {
+      "epoch": 0.9114372469635628,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.7343117408906882e-05,
+      "loss": 2.6945,
+      "step": 3602
+    },
+    {
+      "epoch": 0.9116902834008097,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.7350708502024292e-05,
+      "loss": 2.6277,
+      "step": 3603
+    },
+    {
+      "epoch": 0.9119433198380567,
+      "grad_norm": 3.0,
+      "learning_rate": 2.73582995951417e-05,
+      "loss": 2.6749,
+      "step": 3604
+    },
+    {
+      "epoch": 0.9121963562753036,
+      "grad_norm": 3.375,
+      "learning_rate": 2.736589068825911e-05,
+      "loss": 2.7646,
+      "step": 3605
+    },
+    {
+      "epoch": 0.9124493927125507,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.737348178137652e-05,
+      "loss": 2.6335,
+      "step": 3606
+    },
+    {
+      "epoch": 0.9127024291497976,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.738107287449393e-05,
+      "loss": 2.6974,
+      "step": 3607
+    },
+    {
+      "epoch": 0.9129554655870445,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.7388663967611336e-05,
+      "loss": 2.5916,
+      "step": 3608
+    },
+    {
+      "epoch": 0.9132085020242915,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.7396255060728746e-05,
+      "loss": 2.5949,
+      "step": 3609
+    },
+    {
+      "epoch": 0.9134615384615384,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.7403846153846153e-05,
+      "loss": 2.6252,
+      "step": 3610
+    },
+    {
+      "epoch": 0.9137145748987854,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.7411437246963563e-05,
+      "loss": 2.5912,
+      "step": 3611
+    },
+    {
+      "epoch": 0.9139676113360324,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.7419028340080973e-05,
+      "loss": 2.7055,
+      "step": 3612
+    },
+    {
+      "epoch": 0.9142206477732794,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.7426619433198383e-05,
+      "loss": 2.5808,
+      "step": 3613
+    },
+    {
+      "epoch": 0.9144736842105263,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.743421052631579e-05,
+      "loss": 2.6213,
+      "step": 3614
+    },
+    {
+      "epoch": 0.9147267206477733,
+      "grad_norm": 3.25,
+      "learning_rate": 2.74418016194332e-05,
+      "loss": 2.6425,
+      "step": 3615
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.7449392712550607e-05,
+      "loss": 2.6411,
+      "step": 3616
+    },
+    {
+      "epoch": 0.9152327935222672,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.7456983805668017e-05,
+      "loss": 2.6744,
+      "step": 3617
+    },
+    {
+      "epoch": 0.9154858299595142,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.7464574898785427e-05,
+      "loss": 2.6407,
+      "step": 3618
+    },
+    {
+      "epoch": 0.9157388663967612,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.7472165991902837e-05,
+      "loss": 2.6177,
+      "step": 3619
+    },
+    {
+      "epoch": 0.9159919028340081,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.7479757085020244e-05,
+      "loss": 2.5688,
+      "step": 3620
+    },
+    {
+      "epoch": 0.916244939271255,
+      "grad_norm": 3.25,
+      "learning_rate": 2.748734817813765e-05,
+      "loss": 2.6526,
+      "step": 3621
+    },
+    {
+      "epoch": 0.916497975708502,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.749493927125506e-05,
+      "loss": 2.6114,
+      "step": 3622
+    },
+    {
+      "epoch": 0.916751012145749,
+      "grad_norm": 3.125,
+      "learning_rate": 2.7502530364372468e-05,
+      "loss": 2.6036,
+      "step": 3623
+    },
+    {
+      "epoch": 0.917004048582996,
+      "grad_norm": 3.5,
+      "learning_rate": 2.751012145748988e-05,
+      "loss": 2.651,
+      "step": 3624
+    },
+    {
+      "epoch": 0.917257085020243,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.7517712550607288e-05,
+      "loss": 2.5904,
+      "step": 3625
+    },
+    {
+      "epoch": 0.9175101214574899,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.7525303643724698e-05,
+      "loss": 2.6714,
+      "step": 3626
+    },
+    {
+      "epoch": 0.9177631578947368,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.7532894736842105e-05,
+      "loss": 2.6004,
+      "step": 3627
+    },
+    {
+      "epoch": 0.9180161943319838,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.7540485829959515e-05,
+      "loss": 2.6114,
+      "step": 3628
+    },
+    {
+      "epoch": 0.9182692307692307,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.754807692307692e-05,
+      "loss": 2.6986,
+      "step": 3629
+    },
+    {
+      "epoch": 0.9185222672064778,
+      "grad_norm": 22.375,
+      "learning_rate": 2.7555668016194335e-05,
+      "loss": 2.6762,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9185222672064778,
+      "eval_loss": 2.6968343257904053,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.4532,
+      "eval_samples_per_second": 9.543,
+      "eval_steps_per_second": 9.543,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9187753036437247,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.7563259109311742e-05,
+      "loss": 2.5874,
+      "step": 3631
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.7570850202429152e-05,
+      "loss": 2.511,
+      "step": 3632
+    },
+    {
+      "epoch": 0.9192813765182186,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.757844129554656e-05,
+      "loss": 2.6754,
+      "step": 3633
+    },
+    {
+      "epoch": 0.9195344129554656,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.758603238866397e-05,
+      "loss": 2.6607,
+      "step": 3634
+    },
+    {
+      "epoch": 0.9197874493927125,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.7593623481781376e-05,
+      "loss": 2.74,
+      "step": 3635
+    },
+    {
+      "epoch": 0.9200404858299596,
+      "grad_norm": 3.5,
+      "learning_rate": 2.760121457489879e-05,
+      "loss": 2.6624,
+      "step": 3636
+    },
+    {
+      "epoch": 0.9202935222672065,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.7608805668016196e-05,
+      "loss": 2.6059,
+      "step": 3637
+    },
+    {
+      "epoch": 0.9205465587044535,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.7616396761133606e-05,
+      "loss": 2.615,
+      "step": 3638
+    },
+    {
+      "epoch": 0.9207995951417004,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.7623987854251013e-05,
+      "loss": 2.6294,
+      "step": 3639
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.7631578947368423e-05,
+      "loss": 2.7273,
+      "step": 3640
+    },
+    {
+      "epoch": 0.9213056680161943,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.763917004048583e-05,
+      "loss": 2.7014,
+      "step": 3641
+    },
+    {
+      "epoch": 0.9215587044534413,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.764676113360324e-05,
+      "loss": 2.5411,
+      "step": 3642
+    },
+    {
+      "epoch": 0.9218117408906883,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.765435222672065e-05,
+      "loss": 2.613,
+      "step": 3643
+    },
+    {
+      "epoch": 0.9220647773279352,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.7661943319838056e-05,
+      "loss": 2.5512,
+      "step": 3644
+    },
+    {
+      "epoch": 0.9223178137651822,
+      "grad_norm": 3.625,
+      "learning_rate": 2.7669534412955467e-05,
+      "loss": 2.6929,
+      "step": 3645
+    },
+    {
+      "epoch": 0.9225708502024291,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.7677125506072873e-05,
+      "loss": 2.6038,
+      "step": 3646
+    },
+    {
+      "epoch": 0.9228238866396761,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.7684716599190283e-05,
+      "loss": 2.6024,
+      "step": 3647
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 2.688,
+      "step": 3648
+    },
+    {
+      "epoch": 0.9233299595141701,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.7699898785425104e-05,
+      "loss": 2.6426,
+      "step": 3649
+    },
+    {
+      "epoch": 0.923582995951417,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.770748987854251e-05,
+      "loss": 2.5892,
+      "step": 3650
+    },
+    {
+      "epoch": 0.923836032388664,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.771508097165992e-05,
+      "loss": 2.6937,
+      "step": 3651
+    },
+    {
+      "epoch": 0.9240890688259109,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.7722672064777327e-05,
+      "loss": 2.6098,
+      "step": 3652
+    },
+    {
+      "epoch": 0.9243421052631579,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.7730263157894737e-05,
+      "loss": 2.6013,
+      "step": 3653
+    },
+    {
+      "epoch": 0.9245951417004049,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.7737854251012147e-05,
+      "loss": 2.6161,
+      "step": 3654
+    },
+    {
+      "epoch": 0.9248481781376519,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.7745445344129558e-05,
+      "loss": 2.6235,
+      "step": 3655
+    },
+    {
+      "epoch": 0.9251012145748988,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.7753036437246964e-05,
+      "loss": 2.5404,
+      "step": 3656
+    },
+    {
+      "epoch": 0.9253542510121457,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.7760627530364374e-05,
+      "loss": 2.5621,
+      "step": 3657
+    },
+    {
+      "epoch": 0.9256072874493927,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.776821862348178e-05,
+      "loss": 2.6151,
+      "step": 3658
+    },
+    {
+      "epoch": 0.9258603238866396,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.777580971659919e-05,
+      "loss": 2.633,
+      "step": 3659
+    },
+    {
+      "epoch": 0.9261133603238867,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.77834008097166e-05,
+      "loss": 2.7158,
+      "step": 3660
+    },
+    {
+      "epoch": 0.9261133603238867,
+      "eval_loss": 2.6880276203155518,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.1139,
+      "eval_samples_per_second": 9.52,
+      "eval_steps_per_second": 9.52,
+      "step": 3660
+    },
+    {
+      "epoch": 0.9263663967611336,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.7790991902834008e-05,
+      "loss": 2.6158,
+      "step": 3661
+    },
+    {
+      "epoch": 0.9266194331983806,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.7798582995951418e-05,
+      "loss": 2.6992,
+      "step": 3662
+    },
+    {
+      "epoch": 0.9268724696356275,
+      "grad_norm": 3.375,
+      "learning_rate": 2.7806174089068825e-05,
+      "loss": 2.7274,
+      "step": 3663
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.7813765182186235e-05,
+      "loss": 2.5208,
+      "step": 3664
+    },
+    {
+      "epoch": 0.9273785425101214,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.7821356275303642e-05,
+      "loss": 2.5015,
+      "step": 3665
+    },
+    {
+      "epoch": 0.9276315789473685,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.7828947368421055e-05,
+      "loss": 2.5438,
+      "step": 3666
+    },
+    {
+      "epoch": 0.9278846153846154,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.7836538461538462e-05,
+      "loss": 2.65,
+      "step": 3667
+    },
+    {
+      "epoch": 0.9281376518218624,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.7844129554655872e-05,
+      "loss": 2.4493,
+      "step": 3668
+    },
+    {
+      "epoch": 0.9283906882591093,
+      "grad_norm": 3.0,
+      "learning_rate": 2.785172064777328e-05,
+      "loss": 2.4782,
+      "step": 3669
+    },
+    {
+      "epoch": 0.9286437246963563,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.785931174089069e-05,
+      "loss": 2.7033,
+      "step": 3670
+    },
+    {
+      "epoch": 0.9288967611336032,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.7866902834008096e-05,
+      "loss": 2.5873,
+      "step": 3671
+    },
+    {
+      "epoch": 0.9291497975708503,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.787449392712551e-05,
+      "loss": 2.54,
+      "step": 3672
+    },
+    {
+      "epoch": 0.9294028340080972,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.7882085020242916e-05,
+      "loss": 2.6447,
+      "step": 3673
+    },
+    {
+      "epoch": 0.9296558704453441,
+      "grad_norm": 14.375,
+      "learning_rate": 2.7889676113360326e-05,
+      "loss": 2.7413,
+      "step": 3674
+    },
+    {
+      "epoch": 0.9299089068825911,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.7897267206477733e-05,
+      "loss": 2.6237,
+      "step": 3675
+    },
+    {
+      "epoch": 0.930161943319838,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.7904858299595143e-05,
+      "loss": 2.705,
+      "step": 3676
+    },
+    {
+      "epoch": 0.930414979757085,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.791244939271255e-05,
+      "loss": 2.6641,
+      "step": 3677
+    },
+    {
+      "epoch": 0.930668016194332,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.7920040485829963e-05,
+      "loss": 2.6811,
+      "step": 3678
+    },
+    {
+      "epoch": 0.930921052631579,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.792763157894737e-05,
+      "loss": 2.6437,
+      "step": 3679
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.793522267206478e-05,
+      "loss": 2.6611,
+      "step": 3680
+    },
+    {
+      "epoch": 0.9314271255060729,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.7942813765182187e-05,
+      "loss": 2.5925,
+      "step": 3681
+    },
+    {
+      "epoch": 0.9316801619433198,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.7950404858299594e-05,
+      "loss": 2.5969,
+      "step": 3682
+    },
+    {
+      "epoch": 0.9319331983805668,
+      "grad_norm": 23.125,
+      "learning_rate": 2.7957995951417004e-05,
+      "loss": 2.8055,
+      "step": 3683
+    },
+    {
+      "epoch": 0.9321862348178138,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.7965587044534414e-05,
+      "loss": 2.5086,
+      "step": 3684
+    },
+    {
+      "epoch": 0.9324392712550608,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.7973178137651824e-05,
+      "loss": 2.6495,
+      "step": 3685
+    },
+    {
+      "epoch": 0.9326923076923077,
+      "grad_norm": 3.25,
+      "learning_rate": 2.798076923076923e-05,
+      "loss": 2.6193,
+      "step": 3686
+    },
+    {
+      "epoch": 0.9329453441295547,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.798836032388664e-05,
+      "loss": 2.659,
+      "step": 3687
+    },
+    {
+      "epoch": 0.9331983805668016,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.7995951417004048e-05,
+      "loss": 2.6561,
+      "step": 3688
+    },
+    {
+      "epoch": 0.9334514170040485,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8003542510121458e-05,
+      "loss": 2.6129,
+      "step": 3689
+    },
+    {
+      "epoch": 0.9337044534412956,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8011133603238868e-05,
+      "loss": 2.629,
+      "step": 3690
+    },
+    {
+      "epoch": 0.9337044534412956,
+      "eval_loss": 2.6927950382232666,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.5627,
+      "eval_samples_per_second": 9.504,
+      "eval_steps_per_second": 9.504,
+      "step": 3690
+    },
+    {
+      "epoch": 0.9339574898785425,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.8018724696356278e-05,
+      "loss": 2.6978,
+      "step": 3691
+    },
+    {
+      "epoch": 0.9342105263157895,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8026315789473685e-05,
+      "loss": 2.6998,
+      "step": 3692
+    },
+    {
+      "epoch": 0.9344635627530364,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8033906882591095e-05,
+      "loss": 2.5354,
+      "step": 3693
+    },
+    {
+      "epoch": 0.9347165991902834,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.80414979757085e-05,
+      "loss": 2.6627,
+      "step": 3694
+    },
+    {
+      "epoch": 0.9349696356275303,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.804908906882591e-05,
+      "loss": 2.5971,
+      "step": 3695
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8056680161943322e-05,
+      "loss": 2.5437,
+      "step": 3696
+    },
+    {
+      "epoch": 0.9354757085020243,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8064271255060732e-05,
+      "loss": 2.567,
+      "step": 3697
+    },
+    {
+      "epoch": 0.9357287449392713,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.807186234817814e-05,
+      "loss": 2.6067,
+      "step": 3698
+    },
+    {
+      "epoch": 0.9359817813765182,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.807945344129555e-05,
+      "loss": 2.5841,
+      "step": 3699
+    },
+    {
+      "epoch": 0.9362348178137652,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8087044534412955e-05,
+      "loss": 2.4943,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9364878542510121,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.8094635627530366e-05,
+      "loss": 2.7151,
+      "step": 3701
+    },
+    {
+      "epoch": 0.9367408906882592,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8102226720647776e-05,
+      "loss": 2.6614,
+      "step": 3702
+    },
+    {
+      "epoch": 0.9369939271255061,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8109817813765182e-05,
+      "loss": 2.509,
+      "step": 3703
+    },
+    {
+      "epoch": 0.937246963562753,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8117408906882592e-05,
+      "loss": 2.6805,
+      "step": 3704
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8125e-05,
+      "loss": 2.6677,
+      "step": 3705
+    },
+    {
+      "epoch": 0.937753036437247,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.813259109311741e-05,
+      "loss": 2.6368,
+      "step": 3706
+    },
+    {
+      "epoch": 0.9380060728744939,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8140182186234816e-05,
+      "loss": 2.629,
+      "step": 3707
+    },
+    {
+      "epoch": 0.9382591093117408,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8147773279352226e-05,
+      "loss": 2.7141,
+      "step": 3708
+    },
+    {
+      "epoch": 0.9385121457489879,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8155364372469636e-05,
+      "loss": 2.516,
+      "step": 3709
+    },
+    {
+      "epoch": 0.9387651821862348,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8162955465587046e-05,
+      "loss": 2.6467,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9390182186234818,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8170546558704453e-05,
+      "loss": 2.2925,
+      "step": 3711
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8178137651821863e-05,
+      "loss": 2.5632,
+      "step": 3712
+    },
+    {
+      "epoch": 0.9395242914979757,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.818572874493927e-05,
+      "loss": 2.6218,
+      "step": 3713
+    },
+    {
+      "epoch": 0.9397773279352226,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.819331983805668e-05,
+      "loss": 2.6011,
+      "step": 3714
+    },
+    {
+      "epoch": 0.9400303643724697,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.820091093117409e-05,
+      "loss": 2.5792,
+      "step": 3715
+    },
+    {
+      "epoch": 0.9402834008097166,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.82085020242915e-05,
+      "loss": 2.5417,
+      "step": 3716
+    },
+    {
+      "epoch": 0.9405364372469636,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8216093117408907e-05,
+      "loss": 2.6364,
+      "step": 3717
+    },
+    {
+      "epoch": 0.9407894736842105,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.8223684210526317e-05,
+      "loss": 2.7048,
+      "step": 3718
+    },
+    {
+      "epoch": 0.9410425101214575,
+      "grad_norm": 22.5,
+      "learning_rate": 2.8231275303643724e-05,
+      "loss": 2.6389,
+      "step": 3719
+    },
+    {
+      "epoch": 0.9412955465587044,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8238866396761134e-05,
+      "loss": 2.5868,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9412955465587044,
+      "eval_loss": 2.690774917602539,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.1925,
+      "eval_samples_per_second": 9.552,
+      "eval_steps_per_second": 9.552,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9415485829959515,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8246457489878544e-05,
+      "loss": 2.6074,
+      "step": 3721
+    },
+    {
+      "epoch": 0.9418016194331984,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8254048582995954e-05,
+      "loss": 2.6009,
+      "step": 3722
+    },
+    {
+      "epoch": 0.9420546558704453,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.826163967611336e-05,
+      "loss": 2.5117,
+      "step": 3723
+    },
+    {
+      "epoch": 0.9423076923076923,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8269230769230768e-05,
+      "loss": 2.6332,
+      "step": 3724
+    },
+    {
+      "epoch": 0.9425607287449392,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8276821862348178e-05,
+      "loss": 2.7116,
+      "step": 3725
+    },
+    {
+      "epoch": 0.9428137651821862,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8284412955465585e-05,
+      "loss": 2.4826,
+      "step": 3726
+    },
+    {
+      "epoch": 0.9430668016194332,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8292004048582998e-05,
+      "loss": 2.6342,
+      "step": 3727
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8299595141700405e-05,
+      "loss": 2.5791,
+      "step": 3728
+    },
+    {
+      "epoch": 0.9435728744939271,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.8307186234817815e-05,
+      "loss": 2.6174,
+      "step": 3729
+    },
+    {
+      "epoch": 0.9438259109311741,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.8314777327935222e-05,
+      "loss": 2.7486,
+      "step": 3730
+    },
+    {
+      "epoch": 0.944078947368421,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8322368421052632e-05,
+      "loss": 2.6102,
+      "step": 3731
+    },
+    {
+      "epoch": 0.944331983805668,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.832995951417004e-05,
+      "loss": 2.6684,
+      "step": 3732
+    },
+    {
+      "epoch": 0.944585020242915,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8337550607287452e-05,
+      "loss": 2.7531,
+      "step": 3733
+    },
+    {
+      "epoch": 0.944838056680162,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.834514170040486e-05,
+      "loss": 2.5705,
+      "step": 3734
+    },
+    {
+      "epoch": 0.9450910931174089,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.835273279352227e-05,
+      "loss": 2.6568,
+      "step": 3735
+    },
+    {
+      "epoch": 0.9453441295546559,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8360323886639676e-05,
+      "loss": 2.5193,
+      "step": 3736
+    },
+    {
+      "epoch": 0.9455971659919028,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.8367914979757086e-05,
+      "loss": 2.6882,
+      "step": 3737
+    },
+    {
+      "epoch": 0.9458502024291497,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8375506072874493e-05,
+      "loss": 2.6113,
+      "step": 3738
+    },
+    {
+      "epoch": 0.9461032388663968,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8383097165991906e-05,
+      "loss": 2.4923,
+      "step": 3739
+    },
+    {
+      "epoch": 0.9463562753036437,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8390688259109313e-05,
+      "loss": 2.6127,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9466093117408907,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8398279352226723e-05,
+      "loss": 2.5845,
+      "step": 3741
+    },
+    {
+      "epoch": 0.9468623481781376,
+      "grad_norm": 3.125,
+      "learning_rate": 2.840587044534413e-05,
+      "loss": 2.5675,
+      "step": 3742
+    },
+    {
+      "epoch": 0.9471153846153846,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.841346153846154e-05,
+      "loss": 2.6139,
+      "step": 3743
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8421052631578946e-05,
+      "loss": 2.4076,
+      "step": 3744
+    },
+    {
+      "epoch": 0.9476214574898786,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.8428643724696357e-05,
+      "loss": 2.7533,
+      "step": 3745
+    },
+    {
+      "epoch": 0.9478744939271255,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8436234817813767e-05,
+      "loss": 2.6934,
+      "step": 3746
+    },
+    {
+      "epoch": 0.9481275303643725,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8443825910931173e-05,
+      "loss": 2.4391,
+      "step": 3747
+    },
+    {
+      "epoch": 0.9483805668016194,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8451417004048584e-05,
+      "loss": 2.6784,
+      "step": 3748
+    },
+    {
+      "epoch": 0.9486336032388664,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.845900809716599e-05,
+      "loss": 2.6749,
+      "step": 3749
+    },
+    {
+      "epoch": 0.9488866396761133,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.84665991902834e-05,
+      "loss": 2.632,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9488866396761133,
+      "eval_loss": 2.68782377243042,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.1751,
+      "eval_samples_per_second": 9.518,
+      "eval_steps_per_second": 9.518,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9491396761133604,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.847419028340081e-05,
+      "loss": 2.5705,
+      "step": 3751
+    },
+    {
+      "epoch": 0.9493927125506073,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.848178137651822e-05,
+      "loss": 2.659,
+      "step": 3752
+    },
+    {
+      "epoch": 0.9496457489878543,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8489372469635627e-05,
+      "loss": 2.6153,
+      "step": 3753
+    },
+    {
+      "epoch": 0.9498987854251012,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.8496963562753038e-05,
+      "loss": 2.5507,
+      "step": 3754
+    },
+    {
+      "epoch": 0.9501518218623481,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8504554655870444e-05,
+      "loss": 2.5134,
+      "step": 3755
+    },
+    {
+      "epoch": 0.9504048582995951,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.8512145748987854e-05,
+      "loss": 2.5443,
+      "step": 3756
+    },
+    {
+      "epoch": 0.9506578947368421,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.8519736842105264e-05,
+      "loss": 2.7248,
+      "step": 3757
+    },
+    {
+      "epoch": 0.9509109311740891,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.8527327935222675e-05,
+      "loss": 2.6289,
+      "step": 3758
+    },
+    {
+      "epoch": 0.951163967611336,
+      "grad_norm": 3.5,
+      "learning_rate": 2.853491902834008e-05,
+      "loss": 2.5167,
+      "step": 3759
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.854251012145749e-05,
+      "loss": 2.5634,
+      "step": 3760
+    },
+    {
+      "epoch": 0.9516700404858299,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.8550101214574898e-05,
+      "loss": 2.6432,
+      "step": 3761
+    },
+    {
+      "epoch": 0.9519230769230769,
+      "grad_norm": 3.984375,
+      "learning_rate": 2.855769230769231e-05,
+      "loss": 2.6474,
+      "step": 3762
+    },
+    {
+      "epoch": 0.9521761133603239,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.856528340080972e-05,
+      "loss": 2.589,
+      "step": 3763
+    },
+    {
+      "epoch": 0.9524291497975709,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.857287449392713e-05,
+      "loss": 2.6213,
+      "step": 3764
+    },
+    {
+      "epoch": 0.9526821862348178,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8580465587044535e-05,
+      "loss": 2.5997,
+      "step": 3765
+    },
+    {
+      "epoch": 0.9529352226720648,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8588056680161942e-05,
+      "loss": 2.6315,
+      "step": 3766
+    },
+    {
+      "epoch": 0.9531882591093117,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8595647773279352e-05,
+      "loss": 2.572,
+      "step": 3767
+    },
+    {
+      "epoch": 0.9534412955465587,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.860323886639676e-05,
+      "loss": 2.5626,
+      "step": 3768
+    },
+    {
+      "epoch": 0.9536943319838057,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8610829959514172e-05,
+      "loss": 2.6059,
+      "step": 3769
+    },
+    {
+      "epoch": 0.9539473684210527,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.861842105263158e-05,
+      "loss": 2.6168,
+      "step": 3770
+    },
+    {
+      "epoch": 0.9542004048582996,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.862601214574899e-05,
+      "loss": 2.6471,
+      "step": 3771
+    },
+    {
+      "epoch": 0.9544534412955465,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8633603238866396e-05,
+      "loss": 2.6342,
+      "step": 3772
+    },
+    {
+      "epoch": 0.9547064777327935,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8641194331983806e-05,
+      "loss": 2.4512,
+      "step": 3773
+    },
+    {
+      "epoch": 0.9549595141700404,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8648785425101213e-05,
+      "loss": 2.5316,
+      "step": 3774
+    },
+    {
+      "epoch": 0.9552125506072875,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8656376518218626e-05,
+      "loss": 2.5996,
+      "step": 3775
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.8663967611336033e-05,
+      "loss": 2.5505,
+      "step": 3776
+    },
+    {
+      "epoch": 0.9557186234817814,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.8671558704453443e-05,
+      "loss": 2.6195,
+      "step": 3777
+    },
+    {
+      "epoch": 0.9559716599190283,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.867914979757085e-05,
+      "loss": 2.6958,
+      "step": 3778
+    },
+    {
+      "epoch": 0.9562246963562753,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.868674089068826e-05,
+      "loss": 2.6529,
+      "step": 3779
+    },
+    {
+      "epoch": 0.9564777327935222,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8694331983805667e-05,
+      "loss": 2.6694,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9564777327935222,
+      "eval_loss": 2.6841161251068115,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.8538,
+      "eval_samples_per_second": 9.529,
+      "eval_steps_per_second": 9.529,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9567307692307693,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.870192307692308e-05,
+      "loss": 2.6799,
+      "step": 3781
+    },
+    {
+      "epoch": 0.9569838056680162,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8709514170040487e-05,
+      "loss": 2.6476,
+      "step": 3782
+    },
+    {
+      "epoch": 0.9572368421052632,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8717105263157897e-05,
+      "loss": 2.6378,
+      "step": 3783
+    },
+    {
+      "epoch": 0.9574898785425101,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.8724696356275304e-05,
+      "loss": 2.7041,
+      "step": 3784
+    },
+    {
+      "epoch": 0.957742914979757,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8732287449392714e-05,
+      "loss": 2.5822,
+      "step": 3785
+    },
+    {
+      "epoch": 0.957995951417004,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.873987854251012e-05,
+      "loss": 2.7698,
+      "step": 3786
+    },
+    {
+      "epoch": 0.958248987854251,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.874746963562753e-05,
+      "loss": 2.5793,
+      "step": 3787
+    },
+    {
+      "epoch": 0.958502024291498,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.875506072874494e-05,
+      "loss": 2.7375,
+      "step": 3788
+    },
+    {
+      "epoch": 0.958755060728745,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.8762651821862348e-05,
+      "loss": 2.6112,
+      "step": 3789
+    },
+    {
+      "epoch": 0.9590080971659919,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8770242914979758e-05,
+      "loss": 2.6321,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9592611336032388,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.8777834008097165e-05,
+      "loss": 2.6719,
+      "step": 3791
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8785425101214575e-05,
+      "loss": 2.6606,
+      "step": 3792
+    },
+    {
+      "epoch": 0.9597672064777328,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8793016194331985e-05,
+      "loss": 2.5706,
+      "step": 3793
+    },
+    {
+      "epoch": 0.9600202429149798,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8800607287449395e-05,
+      "loss": 2.6936,
+      "step": 3794
+    },
+    {
+      "epoch": 0.9602732793522267,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.88081983805668e-05,
+      "loss": 2.5342,
+      "step": 3795
+    },
+    {
+      "epoch": 0.9605263157894737,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8815789473684212e-05,
+      "loss": 2.6042,
+      "step": 3796
+    },
+    {
+      "epoch": 0.9607793522267206,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.882338056680162e-05,
+      "loss": 2.6416,
+      "step": 3797
+    },
+    {
+      "epoch": 0.9610323886639676,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.883097165991903e-05,
+      "loss": 2.6256,
+      "step": 3798
+    },
+    {
+      "epoch": 0.9612854251012146,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.883856275303644e-05,
+      "loss": 2.5029,
+      "step": 3799
+    },
+    {
+      "epoch": 0.9615384615384616,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.884615384615385e-05,
+      "loss": 2.6764,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9617914979757085,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.8853744939271256e-05,
+      "loss": 2.662,
+      "step": 3801
+    },
+    {
+      "epoch": 0.9620445344129555,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8861336032388666e-05,
+      "loss": 2.6004,
+      "step": 3802
+    },
+    {
+      "epoch": 0.9622975708502024,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8868927125506072e-05,
+      "loss": 2.5372,
+      "step": 3803
+    },
+    {
+      "epoch": 0.9625506072874493,
+      "grad_norm": 2.75,
+      "learning_rate": 2.8876518218623483e-05,
+      "loss": 2.5495,
+      "step": 3804
+    },
+    {
+      "epoch": 0.9628036437246964,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8884109311740893e-05,
+      "loss": 2.5369,
+      "step": 3805
+    },
+    {
+      "epoch": 0.9630566801619433,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8891700404858303e-05,
+      "loss": 2.5674,
+      "step": 3806
+    },
+    {
+      "epoch": 0.9633097165991903,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.889929149797571e-05,
+      "loss": 2.5501,
+      "step": 3807
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8906882591093116e-05,
+      "loss": 2.554,
+      "step": 3808
+    },
+    {
+      "epoch": 0.9638157894736842,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.8914473684210526e-05,
+      "loss": 2.6543,
+      "step": 3809
+    },
+    {
+      "epoch": 0.9640688259109311,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8922064777327933e-05,
+      "loss": 2.6103,
+      "step": 3810
+    },
+    {
+      "epoch": 0.9640688259109311,
+      "eval_loss": 2.6853597164154053,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.1218,
+      "eval_samples_per_second": 9.52,
+      "eval_steps_per_second": 9.52,
+      "step": 3810
+    },
+    {
+      "epoch": 0.9643218623481782,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.8929655870445347e-05,
+      "loss": 2.6231,
+      "step": 3811
+    },
+    {
+      "epoch": 0.9645748987854251,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8937246963562753e-05,
+      "loss": 2.5393,
+      "step": 3812
+    },
+    {
+      "epoch": 0.9648279352226721,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8944838056680163e-05,
+      "loss": 2.6194,
+      "step": 3813
+    },
+    {
+      "epoch": 0.965080971659919,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.895242914979757e-05,
+      "loss": 2.5631,
+      "step": 3814
+    },
+    {
+      "epoch": 0.965334008097166,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.896002024291498e-05,
+      "loss": 2.6094,
+      "step": 3815
+    },
+    {
+      "epoch": 0.9655870445344129,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.8967611336032387e-05,
+      "loss": 2.5405,
+      "step": 3816
+    },
+    {
+      "epoch": 0.96584008097166,
+      "grad_norm": 3.375,
+      "learning_rate": 2.89752024291498e-05,
+      "loss": 2.6169,
+      "step": 3817
+    },
+    {
+      "epoch": 0.9660931174089069,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8982793522267207e-05,
+      "loss": 2.608,
+      "step": 3818
+    },
+    {
+      "epoch": 0.9663461538461539,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8990384615384617e-05,
+      "loss": 2.6371,
+      "step": 3819
+    },
+    {
+      "epoch": 0.9665991902834008,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8997975708502024e-05,
+      "loss": 2.4742,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9668522267206477,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.9005566801619434e-05,
+      "loss": 2.5516,
+      "step": 3821
+    },
+    {
+      "epoch": 0.9671052631578947,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.901315789473684e-05,
+      "loss": 2.6451,
+      "step": 3822
+    },
+    {
+      "epoch": 0.9673582995951417,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9020748987854254e-05,
+      "loss": 2.5688,
+      "step": 3823
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.902834008097166e-05,
+      "loss": 2.6036,
+      "step": 3824
+    },
+    {
+      "epoch": 0.9678643724696356,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.903593117408907e-05,
+      "loss": 2.7176,
+      "step": 3825
+    },
+    {
+      "epoch": 0.9681174089068826,
+      "grad_norm": 3.625,
+      "learning_rate": 2.9043522267206478e-05,
+      "loss": 2.6809,
+      "step": 3826
+    },
+    {
+      "epoch": 0.9683704453441295,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9051113360323888e-05,
+      "loss": 2.6278,
+      "step": 3827
+    },
+    {
+      "epoch": 0.9686234817813765,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9058704453441295e-05,
+      "loss": 2.3821,
+      "step": 3828
+    },
+    {
+      "epoch": 0.9688765182186235,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9066295546558705e-05,
+      "loss": 2.6495,
+      "step": 3829
+    },
+    {
+      "epoch": 0.9691295546558705,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9073886639676115e-05,
+      "loss": 2.6141,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9693825910931174,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.9081477732793522e-05,
+      "loss": 2.6451,
+      "step": 3831
+    },
+    {
+      "epoch": 0.9696356275303644,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9089068825910932e-05,
+      "loss": 2.7243,
+      "step": 3832
+    },
+    {
+      "epoch": 0.9698886639676113,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.909665991902834e-05,
+      "loss": 2.5632,
+      "step": 3833
+    },
+    {
+      "epoch": 0.9701417004048583,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.910425101214575e-05,
+      "loss": 2.5856,
+      "step": 3834
+    },
+    {
+      "epoch": 0.9703947368421053,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.911184210526316e-05,
+      "loss": 2.5207,
+      "step": 3835
+    },
+    {
+      "epoch": 0.9706477732793523,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.911943319838057e-05,
+      "loss": 2.6308,
+      "step": 3836
+    },
+    {
+      "epoch": 0.9709008097165992,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9127024291497976e-05,
+      "loss": 2.6501,
+      "step": 3837
+    },
+    {
+      "epoch": 0.9711538461538461,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9134615384615386e-05,
+      "loss": 2.679,
+      "step": 3838
+    },
+    {
+      "epoch": 0.9714068825910931,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.9142206477732793e-05,
+      "loss": 2.5966,
+      "step": 3839
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9149797570850203e-05,
+      "loss": 2.5398,
+      "step": 3840
+    },
+    {
+      "epoch": 0.97165991902834,
+      "eval_loss": 2.681217670440674,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.6643,
+      "eval_samples_per_second": 9.501,
+      "eval_steps_per_second": 9.501,
+      "step": 3840
+    },
+    {
+      "epoch": 0.9719129554655871,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9157388663967613e-05,
+      "loss": 2.4106,
+      "step": 3841
+    },
+    {
+      "epoch": 0.972165991902834,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9164979757085023e-05,
+      "loss": 2.5529,
+      "step": 3842
+    },
+    {
+      "epoch": 0.972419028340081,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.917257085020243e-05,
+      "loss": 2.614,
+      "step": 3843
+    },
+    {
+      "epoch": 0.9726720647773279,
+      "grad_norm": 3.0,
+      "learning_rate": 2.918016194331984e-05,
+      "loss": 2.602,
+      "step": 3844
+    },
+    {
+      "epoch": 0.9729251012145749,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9187753036437247e-05,
+      "loss": 2.5795,
+      "step": 3845
+    },
+    {
+      "epoch": 0.9731781376518218,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9195344129554657e-05,
+      "loss": 2.5884,
+      "step": 3846
+    },
+    {
+      "epoch": 0.9734311740890689,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9202935222672067e-05,
+      "loss": 2.4898,
+      "step": 3847
+    },
+    {
+      "epoch": 0.9736842105263158,
+      "grad_norm": 3.5,
+      "learning_rate": 2.9210526315789474e-05,
+      "loss": 2.681,
+      "step": 3848
+    },
+    {
+      "epoch": 0.9739372469635628,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9218117408906884e-05,
+      "loss": 2.5165,
+      "step": 3849
+    },
+    {
+      "epoch": 0.9741902834008097,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.922570850202429e-05,
+      "loss": 2.5839,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9744433198380567,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.92332995951417e-05,
+      "loss": 2.5451,
+      "step": 3851
+    },
+    {
+      "epoch": 0.9746963562753036,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9240890688259107e-05,
+      "loss": 2.5144,
+      "step": 3852
+    },
+    {
+      "epoch": 0.9749493927125507,
+      "grad_norm": 3.25,
+      "learning_rate": 2.924848178137652e-05,
+      "loss": 2.5683,
+      "step": 3853
+    },
+    {
+      "epoch": 0.9752024291497976,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9256072874493928e-05,
+      "loss": 2.5681,
+      "step": 3854
+    },
+    {
+      "epoch": 0.9754554655870445,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9263663967611338e-05,
+      "loss": 2.4646,
+      "step": 3855
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9271255060728744e-05,
+      "loss": 2.6929,
+      "step": 3856
+    },
+    {
+      "epoch": 0.9759615384615384,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9278846153846155e-05,
+      "loss": 2.6017,
+      "step": 3857
+    },
+    {
+      "epoch": 0.9762145748987854,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.928643724696356e-05,
+      "loss": 2.5725,
+      "step": 3858
+    },
+    {
+      "epoch": 0.9764676113360324,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9294028340080975e-05,
+      "loss": 2.6263,
+      "step": 3859
+    },
+    {
+      "epoch": 0.9767206477732794,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.930161943319838e-05,
+      "loss": 2.4921,
+      "step": 3860
+    },
+    {
+      "epoch": 0.9769736842105263,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.930921052631579e-05,
+      "loss": 2.6504,
+      "step": 3861
+    },
+    {
+      "epoch": 0.9772267206477733,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.93168016194332e-05,
+      "loss": 2.5732,
+      "step": 3862
+    },
+    {
+      "epoch": 0.9774797570850202,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.932439271255061e-05,
+      "loss": 2.6375,
+      "step": 3863
+    },
+    {
+      "epoch": 0.9777327935222672,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9331983805668015e-05,
+      "loss": 2.7028,
+      "step": 3864
+    },
+    {
+      "epoch": 0.9779858299595142,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.933957489878543e-05,
+      "loss": 2.583,
+      "step": 3865
+    },
+    {
+      "epoch": 0.9782388663967612,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9347165991902835e-05,
+      "loss": 2.6117,
+      "step": 3866
+    },
+    {
+      "epoch": 0.9784919028340081,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9354757085020246e-05,
+      "loss": 2.612,
+      "step": 3867
+    },
+    {
+      "epoch": 0.978744939271255,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9362348178137652e-05,
+      "loss": 2.6117,
+      "step": 3868
+    },
+    {
+      "epoch": 0.978997975708502,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.936993927125506e-05,
+      "loss": 2.4952,
+      "step": 3869
+    },
+    {
+      "epoch": 0.979251012145749,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.937753036437247e-05,
+      "loss": 2.6321,
+      "step": 3870
+    },
+    {
+      "epoch": 0.979251012145749,
+      "eval_loss": 2.678889513015747,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.6922,
+      "eval_samples_per_second": 9.5,
+      "eval_steps_per_second": 9.5,
+      "step": 3870
+    },
+    {
+      "epoch": 0.979504048582996,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.938512145748988e-05,
+      "loss": 2.6651,
+      "step": 3871
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.939271255060729e-05,
+      "loss": 2.5781,
+      "step": 3872
+    },
+    {
+      "epoch": 0.9800101214574899,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9400303643724696e-05,
+      "loss": 2.644,
+      "step": 3873
+    },
+    {
+      "epoch": 0.9802631578947368,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9407894736842106e-05,
+      "loss": 2.5312,
+      "step": 3874
+    },
+    {
+      "epoch": 0.9805161943319838,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9415485829959513e-05,
+      "loss": 2.6322,
+      "step": 3875
+    },
+    {
+      "epoch": 0.9807692307692307,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9423076923076923e-05,
+      "loss": 2.4694,
+      "step": 3876
+    },
+    {
+      "epoch": 0.9810222672064778,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9430668016194333e-05,
+      "loss": 2.6484,
+      "step": 3877
+    },
+    {
+      "epoch": 0.9812753036437247,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.9438259109311743e-05,
+      "loss": 2.6733,
+      "step": 3878
+    },
+    {
+      "epoch": 0.9815283400809717,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.944585020242915e-05,
+      "loss": 2.5951,
+      "step": 3879
+    },
+    {
+      "epoch": 0.9817813765182186,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.945344129554656e-05,
+      "loss": 2.5886,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9820344129554656,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9461032388663967e-05,
+      "loss": 2.3848,
+      "step": 3881
+    },
+    {
+      "epoch": 0.9822874493927125,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9468623481781377e-05,
+      "loss": 2.6208,
+      "step": 3882
+    },
+    {
+      "epoch": 0.9825404858299596,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9476214574898787e-05,
+      "loss": 2.5741,
+      "step": 3883
+    },
+    {
+      "epoch": 0.9827935222672065,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9483805668016197e-05,
+      "loss": 2.6442,
+      "step": 3884
+    },
+    {
+      "epoch": 0.9830465587044535,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9491396761133604e-05,
+      "loss": 2.66,
+      "step": 3885
+    },
+    {
+      "epoch": 0.9832995951417004,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9498987854251014e-05,
+      "loss": 2.6296,
+      "step": 3886
+    },
+    {
+      "epoch": 0.9835526315789473,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.950657894736842e-05,
+      "loss": 2.6554,
+      "step": 3887
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.951417004048583e-05,
+      "loss": 2.5902,
+      "step": 3888
+    },
+    {
+      "epoch": 0.9840587044534413,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.952176113360324e-05,
+      "loss": 2.5708,
+      "step": 3889
+    },
+    {
+      "epoch": 0.9843117408906883,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9529352226720648e-05,
+      "loss": 2.5416,
+      "step": 3890
+    },
+    {
+      "epoch": 0.9845647773279352,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9536943319838058e-05,
+      "loss": 2.7167,
+      "step": 3891
+    },
+    {
+      "epoch": 0.9848178137651822,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9544534412955465e-05,
+      "loss": 2.7135,
+      "step": 3892
+    },
+    {
+      "epoch": 0.9850708502024291,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9552125506072875e-05,
+      "loss": 2.4476,
+      "step": 3893
+    },
+    {
+      "epoch": 0.9853238866396761,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.955971659919028e-05,
+      "loss": 2.6768,
+      "step": 3894
+    },
+    {
+      "epoch": 0.9855769230769231,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9567307692307695e-05,
+      "loss": 2.6595,
+      "step": 3895
+    },
+    {
+      "epoch": 0.9858299595141701,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.9574898785425102e-05,
+      "loss": 2.6881,
+      "step": 3896
+    },
+    {
+      "epoch": 0.986082995951417,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9582489878542512e-05,
+      "loss": 2.7068,
+      "step": 3897
+    },
+    {
+      "epoch": 0.986336032388664,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.959008097165992e-05,
+      "loss": 2.6263,
+      "step": 3898
+    },
+    {
+      "epoch": 0.9865890688259109,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.959767206477733e-05,
+      "loss": 2.6024,
+      "step": 3899
+    },
+    {
+      "epoch": 0.9868421052631579,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9605263157894735e-05,
+      "loss": 2.3899,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9868421052631579,
+      "eval_loss": 2.675966501235962,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.4262,
+      "eval_samples_per_second": 9.474,
+      "eval_steps_per_second": 9.474,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9870951417004049,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.961285425101215e-05,
+      "loss": 2.5015,
+      "step": 3901
+    },
+    {
+      "epoch": 0.9873481781376519,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.9620445344129556e-05,
+      "loss": 2.5926,
+      "step": 3902
+    },
+    {
+      "epoch": 0.9876012145748988,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9628036437246966e-05,
+      "loss": 2.6507,
+      "step": 3903
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9635627530364373e-05,
+      "loss": 2.725,
+      "step": 3904
+    },
+    {
+      "epoch": 0.9881072874493927,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9643218623481783e-05,
+      "loss": 2.5366,
+      "step": 3905
+    },
+    {
+      "epoch": 0.9883603238866396,
+      "grad_norm": 3.25,
+      "learning_rate": 2.965080971659919e-05,
+      "loss": 2.5585,
+      "step": 3906
+    },
+    {
+      "epoch": 0.9886133603238867,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9658400809716603e-05,
+      "loss": 2.5786,
+      "step": 3907
+    },
+    {
+      "epoch": 0.9888663967611336,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.966599190283401e-05,
+      "loss": 2.6128,
+      "step": 3908
+    },
+    {
+      "epoch": 0.9891194331983806,
+      "grad_norm": 3.125,
+      "learning_rate": 2.967358299595142e-05,
+      "loss": 2.5969,
+      "step": 3909
+    },
+    {
+      "epoch": 0.9893724696356275,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9681174089068827e-05,
+      "loss": 2.5925,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9896255060728745,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9688765182186233e-05,
+      "loss": 2.4698,
+      "step": 3911
+    },
+    {
+      "epoch": 0.9898785425101214,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9696356275303643e-05,
+      "loss": 2.6467,
+      "step": 3912
+    },
+    {
+      "epoch": 0.9901315789473685,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.9703947368421054e-05,
+      "loss": 2.6444,
+      "step": 3913
+    },
+    {
+      "epoch": 0.9903846153846154,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9711538461538464e-05,
+      "loss": 2.683,
+      "step": 3914
+    },
+    {
+      "epoch": 0.9906376518218624,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.971912955465587e-05,
+      "loss": 2.626,
+      "step": 3915
+    },
+    {
+      "epoch": 0.9908906882591093,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.972672064777328e-05,
+      "loss": 2.7258,
+      "step": 3916
+    },
+    {
+      "epoch": 0.9911437246963563,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9734311740890687e-05,
+      "loss": 2.6507,
+      "step": 3917
+    },
+    {
+      "epoch": 0.9913967611336032,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9741902834008097e-05,
+      "loss": 2.6743,
+      "step": 3918
+    },
+    {
+      "epoch": 0.9916497975708503,
+      "grad_norm": 3.859375,
+      "learning_rate": 2.9749493927125507e-05,
+      "loss": 2.5844,
+      "step": 3919
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9757085020242918e-05,
+      "loss": 2.5976,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9921558704453441,
+      "grad_norm": 3.625,
+      "learning_rate": 2.9764676113360324e-05,
+      "loss": 2.6835,
+      "step": 3921
+    },
+    {
+      "epoch": 0.9924089068825911,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9772267206477734e-05,
+      "loss": 2.666,
+      "step": 3922
+    },
+    {
+      "epoch": 0.992661943319838,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.977985829959514e-05,
+      "loss": 2.5814,
+      "step": 3923
+    },
+    {
+      "epoch": 0.992914979757085,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.978744939271255e-05,
+      "loss": 2.6441,
+      "step": 3924
+    },
+    {
+      "epoch": 0.993168016194332,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.979504048582996e-05,
+      "loss": 2.6596,
+      "step": 3925
+    },
+    {
+      "epoch": 0.993421052631579,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.980263157894737e-05,
+      "loss": 2.6221,
+      "step": 3926
+    },
+    {
+      "epoch": 0.9936740890688259,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9810222672064778e-05,
+      "loss": 2.6913,
+      "step": 3927
+    },
+    {
+      "epoch": 0.9939271255060729,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.981781376518219e-05,
+      "loss": 2.6154,
+      "step": 3928
+    },
+    {
+      "epoch": 0.9941801619433198,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9825404858299595e-05,
+      "loss": 2.5227,
+      "step": 3929
+    },
+    {
+      "epoch": 0.9944331983805668,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9832995951417005e-05,
+      "loss": 2.6063,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9944331983805668,
+      "eval_loss": 2.6725821495056152,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.5374,
+      "eval_samples_per_second": 9.505,
+      "eval_steps_per_second": 9.505,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9946862348178138,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9840587044534415e-05,
+      "loss": 2.6587,
+      "step": 3931
+    },
+    {
+      "epoch": 0.9949392712550608,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9848178137651822e-05,
+      "loss": 2.6488,
+      "step": 3932
+    },
+    {
+      "epoch": 0.9951923076923077,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9855769230769232e-05,
+      "loss": 2.4669,
+      "step": 3933
+    },
+    {
+      "epoch": 0.9954453441295547,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.986336032388664e-05,
+      "loss": 2.5917,
+      "step": 3934
+    },
+    {
+      "epoch": 0.9956983805668016,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.987095141700405e-05,
+      "loss": 2.6169,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9878542510121456e-05,
+      "loss": 2.5354,
+      "step": 3936
+    },
+    {
+      "epoch": 0.9962044534412956,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.988613360323887e-05,
+      "loss": 2.5976,
+      "step": 3937
+    },
+    {
+      "epoch": 0.9964574898785425,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9893724696356276e-05,
+      "loss": 2.559,
+      "step": 3938
+    },
+    {
+      "epoch": 0.9967105263157895,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9901315789473686e-05,
+      "loss": 2.7599,
+      "step": 3939
+    },
+    {
+      "epoch": 0.9969635627530364,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9908906882591093e-05,
+      "loss": 2.6618,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9972165991902834,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.9916497975708503e-05,
+      "loss": 2.607,
+      "step": 3941
+    },
+    {
+      "epoch": 0.9974696356275303,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.992408906882591e-05,
+      "loss": 2.5762,
+      "step": 3942
+    },
+    {
+      "epoch": 0.9977226720647774,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9931680161943323e-05,
+      "loss": 2.6357,
+      "step": 3943
+    },
+    {
+      "epoch": 0.9979757085020243,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.993927125506073e-05,
+      "loss": 2.6336,
+      "step": 3944
+    },
+    {
+      "epoch": 0.9982287449392713,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.994686234817814e-05,
+      "loss": 2.6648,
+      "step": 3945
+    },
+    {
+      "epoch": 0.9984817813765182,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9954453441295547e-05,
+      "loss": 2.5701,
+      "step": 3946
+    },
+    {
+      "epoch": 0.9987348178137652,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.9962044534412957e-05,
+      "loss": 2.5429,
+      "step": 3947
+    },
+    {
+      "epoch": 0.9989878542510121,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9969635627530364e-05,
+      "loss": 2.6094,
+      "step": 3948
+    },
+    {
+      "epoch": 0.9992408906882592,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9977226720647777e-05,
+      "loss": 2.5291,
+      "step": 3949
+    },
+    {
+      "epoch": 0.9994939271255061,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9984817813765184e-05,
+      "loss": 2.6153,
+      "step": 3950
+    },
+    {
+      "epoch": 0.999746963562753,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9992408906882594e-05,
+      "loss": 2.6148,
+      "step": 3951
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.28125,
+      "learning_rate": 3e-05,
+      "loss": 2.6897,
+      "step": 3952
+    },
+    {
+      "epoch": 1.000253036437247,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9999156545209177e-05,
+      "loss": 2.658,
+      "step": 3953
+    },
+    {
+      "epoch": 1.000506072874494,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.9998313090418354e-05,
+      "loss": 2.5216,
+      "step": 3954
+    },
+    {
+      "epoch": 1.000759109311741,
+      "grad_norm": 3.0,
+      "learning_rate": 2.999746963562753e-05,
+      "loss": 2.5581,
+      "step": 3955
+    },
+    {
+      "epoch": 1.0010121457489878,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9996626180836707e-05,
+      "loss": 2.5578,
+      "step": 3956
+    },
+    {
+      "epoch": 1.0012651821862348,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9995782726045887e-05,
+      "loss": 2.5525,
+      "step": 3957
+    },
+    {
+      "epoch": 1.0015182186234817,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.999493927125506e-05,
+      "loss": 2.589,
+      "step": 3958
+    },
+    {
+      "epoch": 1.0017712550607287,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9994095816464237e-05,
+      "loss": 2.5104,
+      "step": 3959
+    },
+    {
+      "epoch": 1.0020242914979758,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9993252361673417e-05,
+      "loss": 2.5078,
+      "step": 3960
+    },
+    {
+      "epoch": 1.0020242914979758,
+      "eval_loss": 2.6707966327667236,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.1314,
+      "eval_samples_per_second": 9.485,
+      "eval_steps_per_second": 9.485,
+      "step": 3960
+    },
+    {
+      "epoch": 1.0022773279352226,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9992408906882594e-05,
+      "loss": 2.5578,
+      "step": 3961
+    },
+    {
+      "epoch": 1.0025303643724697,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9991565452091767e-05,
+      "loss": 2.551,
+      "step": 3962
+    },
+    {
+      "epoch": 1.0027834008097165,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9990721997300944e-05,
+      "loss": 2.5181,
+      "step": 3963
+    },
+    {
+      "epoch": 1.0030364372469636,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9989878542510124e-05,
+      "loss": 2.4117,
+      "step": 3964
+    },
+    {
+      "epoch": 1.0032894736842106,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9989035087719297e-05,
+      "loss": 2.5741,
+      "step": 3965
+    },
+    {
+      "epoch": 1.0035425101214575,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9988191632928474e-05,
+      "loss": 2.5281,
+      "step": 3966
+    },
+    {
+      "epoch": 1.0037955465587045,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9987348178137654e-05,
+      "loss": 2.5275,
+      "step": 3967
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.998650472334683e-05,
+      "loss": 2.5837,
+      "step": 3968
+    },
+    {
+      "epoch": 1.0043016194331984,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9985661268556004e-05,
+      "loss": 2.5747,
+      "step": 3969
+    },
+    {
+      "epoch": 1.0045546558704452,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9984817813765184e-05,
+      "loss": 2.5312,
+      "step": 3970
+    },
+    {
+      "epoch": 1.0048076923076923,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.998397435897436e-05,
+      "loss": 2.4804,
+      "step": 3971
+    },
+    {
+      "epoch": 1.0050607287449393,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9983130904183537e-05,
+      "loss": 2.5972,
+      "step": 3972
+    },
+    {
+      "epoch": 1.0053137651821862,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9982287449392714e-05,
+      "loss": 2.5121,
+      "step": 3973
+    },
+    {
+      "epoch": 1.0055668016194332,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.998144399460189e-05,
+      "loss": 2.5273,
+      "step": 3974
+    },
+    {
+      "epoch": 1.00581983805668,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9980600539811067e-05,
+      "loss": 2.5207,
+      "step": 3975
+    },
+    {
+      "epoch": 1.0060728744939271,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9979757085020244e-05,
+      "loss": 2.6085,
+      "step": 3976
+    },
+    {
+      "epoch": 1.0063259109311742,
+      "grad_norm": 3.0,
+      "learning_rate": 2.997891363022942e-05,
+      "loss": 2.3951,
+      "step": 3977
+    },
+    {
+      "epoch": 1.006578947368421,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9978070175438597e-05,
+      "loss": 2.441,
+      "step": 3978
+    },
+    {
+      "epoch": 1.006831983805668,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9977226720647777e-05,
+      "loss": 2.5027,
+      "step": 3979
+    },
+    {
+      "epoch": 1.007085020242915,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.997638326585695e-05,
+      "loss": 2.4787,
+      "step": 3980
+    },
+    {
+      "epoch": 1.007338056680162,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9975539811066127e-05,
+      "loss": 2.621,
+      "step": 3981
+    },
+    {
+      "epoch": 1.0075910931174088,
+      "grad_norm": 6.03125,
+      "learning_rate": 2.9974696356275304e-05,
+      "loss": 2.5949,
+      "step": 3982
+    },
+    {
+      "epoch": 1.0078441295546559,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.997385290148448e-05,
+      "loss": 2.5108,
+      "step": 3983
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9973009446693657e-05,
+      "loss": 2.5907,
+      "step": 3984
+    },
+    {
+      "epoch": 1.0083502024291497,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9972165991902834e-05,
+      "loss": 2.473,
+      "step": 3985
+    },
+    {
+      "epoch": 1.0086032388663968,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9971322537112014e-05,
+      "loss": 2.549,
+      "step": 3986
+    },
+    {
+      "epoch": 1.0088562753036436,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9970479082321187e-05,
+      "loss": 2.5758,
+      "step": 3987
+    },
+    {
+      "epoch": 1.0091093117408907,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.9969635627530364e-05,
+      "loss": 2.3256,
+      "step": 3988
+    },
+    {
+      "epoch": 1.0093623481781377,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9968792172739544e-05,
+      "loss": 2.5471,
+      "step": 3989
+    },
+    {
+      "epoch": 1.0096153846153846,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.996794871794872e-05,
+      "loss": 2.4892,
+      "step": 3990
+    },
+    {
+      "epoch": 1.0096153846153846,
+      "eval_loss": 2.672454833984375,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.6795,
+      "eval_samples_per_second": 9.466,
+      "eval_steps_per_second": 9.466,
+      "step": 3990
+    },
+    {
+      "epoch": 1.0098684210526316,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9967105263157894e-05,
+      "loss": 2.5713,
+      "step": 3991
+    },
+    {
+      "epoch": 1.0101214574898785,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9966261808367074e-05,
+      "loss": 2.6152,
+      "step": 3992
+    },
+    {
+      "epoch": 1.0103744939271255,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.996541835357625e-05,
+      "loss": 2.5533,
+      "step": 3993
+    },
+    {
+      "epoch": 1.0106275303643724,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9964574898785424e-05,
+      "loss": 2.5424,
+      "step": 3994
+    },
+    {
+      "epoch": 1.0108805668016194,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.9963731443994604e-05,
+      "loss": 2.6467,
+      "step": 3995
+    },
+    {
+      "epoch": 1.0111336032388665,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.996288798920378e-05,
+      "loss": 2.6653,
+      "step": 3996
+    },
+    {
+      "epoch": 1.0113866396761133,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.9962044534412957e-05,
+      "loss": 2.583,
+      "step": 3997
+    },
+    {
+      "epoch": 1.0116396761133604,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.996120107962213e-05,
+      "loss": 2.3707,
+      "step": 3998
+    },
+    {
+      "epoch": 1.0118927125506072,
+      "grad_norm": 3.25,
+      "learning_rate": 2.996035762483131e-05,
+      "loss": 2.5087,
+      "step": 3999
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9959514170040487e-05,
+      "loss": 2.5202,
+      "step": 4000
+    },
+    {
+      "epoch": 1.0123987854251013,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9958670715249664e-05,
+      "loss": 2.5383,
+      "step": 4001
+    },
+    {
+      "epoch": 1.0126518218623481,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.995782726045884e-05,
+      "loss": 2.6339,
+      "step": 4002
+    },
+    {
+      "epoch": 1.0129048582995952,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9956983805668017e-05,
+      "loss": 2.4961,
+      "step": 4003
+    },
+    {
+      "epoch": 1.013157894736842,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9956140350877194e-05,
+      "loss": 2.3227,
+      "step": 4004
+    },
+    {
+      "epoch": 1.013410931174089,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.995529689608637e-05,
+      "loss": 2.4314,
+      "step": 4005
+    },
+    {
+      "epoch": 1.013663967611336,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.9954453441295547e-05,
+      "loss": 2.5944,
+      "step": 4006
+    },
+    {
+      "epoch": 1.013917004048583,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9953609986504723e-05,
+      "loss": 2.5237,
+      "step": 4007
+    },
+    {
+      "epoch": 1.01417004048583,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9952766531713904e-05,
+      "loss": 2.6349,
+      "step": 4008
+    },
+    {
+      "epoch": 1.0144230769230769,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9951923076923077e-05,
+      "loss": 2.5043,
+      "step": 4009
+    },
+    {
+      "epoch": 1.014676113360324,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9951079622132253e-05,
+      "loss": 2.426,
+      "step": 4010
+    },
+    {
+      "epoch": 1.0149291497975708,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9950236167341433e-05,
+      "loss": 2.4049,
+      "step": 4011
+    },
+    {
+      "epoch": 1.0151821862348178,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.9949392712550607e-05,
+      "loss": 2.5848,
+      "step": 4012
+    },
+    {
+      "epoch": 1.0154352226720649,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9948549257759783e-05,
+      "loss": 2.5649,
+      "step": 4013
+    },
+    {
+      "epoch": 1.0156882591093117,
+      "grad_norm": 20.25,
+      "learning_rate": 2.9947705802968963e-05,
+      "loss": 2.7005,
+      "step": 4014
+    },
+    {
+      "epoch": 1.0159412955465588,
+      "grad_norm": 3.375,
+      "learning_rate": 2.994686234817814e-05,
+      "loss": 2.5405,
+      "step": 4015
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9946018893387313e-05,
+      "loss": 2.5481,
+      "step": 4016
+    },
+    {
+      "epoch": 1.0164473684210527,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.994517543859649e-05,
+      "loss": 2.5016,
+      "step": 4017
+    },
+    {
+      "epoch": 1.0167004048582995,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.994433198380567e-05,
+      "loss": 2.462,
+      "step": 4018
+    },
+    {
+      "epoch": 1.0169534412955465,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9943488529014847e-05,
+      "loss": 2.5484,
+      "step": 4019
+    },
+    {
+      "epoch": 1.0172064777327936,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.994264507422402e-05,
+      "loss": 2.5968,
+      "step": 4020
+    },
+    {
+      "epoch": 1.0172064777327936,
+      "eval_loss": 2.6727564334869385,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 280.1234,
+      "eval_samples_per_second": 9.282,
+      "eval_steps_per_second": 9.282,
+      "step": 4020
+    },
+    {
+      "epoch": 1.0174595141700404,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.99418016194332e-05,
+      "loss": 2.514,
+      "step": 4021
+    },
+    {
+      "epoch": 1.0177125506072875,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9940958164642377e-05,
+      "loss": 2.5574,
+      "step": 4022
+    },
+    {
+      "epoch": 1.0179655870445343,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.994011470985155e-05,
+      "loss": 2.5433,
+      "step": 4023
+    },
+    {
+      "epoch": 1.0182186234817814,
+      "grad_norm": 3.0,
+      "learning_rate": 2.993927125506073e-05,
+      "loss": 2.5484,
+      "step": 4024
+    },
+    {
+      "epoch": 1.0184716599190284,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9938427800269907e-05,
+      "loss": 2.6035,
+      "step": 4025
+    },
+    {
+      "epoch": 1.0187246963562753,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9937584345479083e-05,
+      "loss": 2.6331,
+      "step": 4026
+    },
+    {
+      "epoch": 1.0189777327935223,
+      "grad_norm": 22.125,
+      "learning_rate": 2.993674089068826e-05,
+      "loss": 2.703,
+      "step": 4027
+    },
+    {
+      "epoch": 1.0192307692307692,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9935897435897437e-05,
+      "loss": 2.5025,
+      "step": 4028
+    },
+    {
+      "epoch": 1.0194838056680162,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9935053981106613e-05,
+      "loss": 2.6822,
+      "step": 4029
+    },
+    {
+      "epoch": 1.019736842105263,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9934210526315793e-05,
+      "loss": 2.5201,
+      "step": 4030
+    },
+    {
+      "epoch": 1.01998987854251,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9933367071524967e-05,
+      "loss": 2.5519,
+      "step": 4031
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.9932523616734143e-05,
+      "loss": 2.4969,
+      "step": 4032
+    },
+    {
+      "epoch": 1.020495951417004,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9931680161943323e-05,
+      "loss": 2.5465,
+      "step": 4033
+    },
+    {
+      "epoch": 1.020748987854251,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9930836707152496e-05,
+      "loss": 2.38,
+      "step": 4034
+    },
+    {
+      "epoch": 1.021002024291498,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.9929993252361673e-05,
+      "loss": 2.462,
+      "step": 4035
+    },
+    {
+      "epoch": 1.021255060728745,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.992914979757085e-05,
+      "loss": 2.5928,
+      "step": 4036
+    },
+    {
+      "epoch": 1.021508097165992,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.992830634278003e-05,
+      "loss": 2.5063,
+      "step": 4037
+    },
+    {
+      "epoch": 1.0217611336032388,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.9927462887989203e-05,
+      "loss": 2.5178,
+      "step": 4038
+    },
+    {
+      "epoch": 1.022014170040486,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.992661943319838e-05,
+      "loss": 2.5162,
+      "step": 4039
+    },
+    {
+      "epoch": 1.0222672064777327,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.992577597840756e-05,
+      "loss": 2.5954,
+      "step": 4040
+    },
+    {
+      "epoch": 1.0225202429149798,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9924932523616736e-05,
+      "loss": 2.5449,
+      "step": 4041
+    },
+    {
+      "epoch": 1.0227732793522266,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.992408906882591e-05,
+      "loss": 2.6255,
+      "step": 4042
+    },
+    {
+      "epoch": 1.0230263157894737,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.992324561403509e-05,
+      "loss": 2.5431,
+      "step": 4043
+    },
+    {
+      "epoch": 1.0232793522267207,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9922402159244266e-05,
+      "loss": 2.6016,
+      "step": 4044
+    },
+    {
+      "epoch": 1.0235323886639676,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.992155870445344e-05,
+      "loss": 2.5555,
+      "step": 4045
+    },
+    {
+      "epoch": 1.0237854251012146,
+      "grad_norm": 3.0,
+      "learning_rate": 2.992071524966262e-05,
+      "loss": 2.5544,
+      "step": 4046
+    },
+    {
+      "epoch": 1.0240384615384615,
+      "grad_norm": 3.8125,
+      "learning_rate": 2.9919871794871796e-05,
+      "loss": 2.5919,
+      "step": 4047
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9919028340080973e-05,
+      "loss": 2.6899,
+      "step": 4048
+    },
+    {
+      "epoch": 1.0245445344129556,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.991818488529015e-05,
+      "loss": 2.5249,
+      "step": 4049
+    },
+    {
+      "epoch": 1.0247975708502024,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9917341430499326e-05,
+      "loss": 2.5063,
+      "step": 4050
+    },
+    {
+      "epoch": 1.0247975708502024,
+      "eval_loss": 2.6735618114471436,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.2061,
+      "eval_samples_per_second": 9.413,
+      "eval_steps_per_second": 9.413,
+      "step": 4050
+    },
+    {
+      "epoch": 1.0250506072874495,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9916497975708503e-05,
+      "loss": 2.5103,
+      "step": 4051
+    },
+    {
+      "epoch": 1.0253036437246963,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9915654520917676e-05,
+      "loss": 2.6078,
+      "step": 4052
+    },
+    {
+      "epoch": 1.0255566801619433,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.9914811066126856e-05,
+      "loss": 2.5093,
+      "step": 4053
+    },
+    {
+      "epoch": 1.0258097165991902,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9913967611336033e-05,
+      "loss": 2.6444,
+      "step": 4054
+    },
+    {
+      "epoch": 1.0260627530364372,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.991312415654521e-05,
+      "loss": 2.6411,
+      "step": 4055
+    },
+    {
+      "epoch": 1.0263157894736843,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.9912280701754386e-05,
+      "loss": 2.5303,
+      "step": 4056
+    },
+    {
+      "epoch": 1.0265688259109311,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9911437246963563e-05,
+      "loss": 2.4731,
+      "step": 4057
+    },
+    {
+      "epoch": 1.0268218623481782,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.991059379217274e-05,
+      "loss": 2.5036,
+      "step": 4058
+    },
+    {
+      "epoch": 1.027074898785425,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.990975033738192e-05,
+      "loss": 2.6393,
+      "step": 4059
+    },
+    {
+      "epoch": 1.027327935222672,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9908906882591093e-05,
+      "loss": 2.6332,
+      "step": 4060
+    },
+    {
+      "epoch": 1.0275809716599191,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.990806342780027e-05,
+      "loss": 2.5075,
+      "step": 4061
+    },
+    {
+      "epoch": 1.027834008097166,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.990721997300945e-05,
+      "loss": 2.6123,
+      "step": 4062
+    },
+    {
+      "epoch": 1.028087044534413,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9906376518218623e-05,
+      "loss": 2.572,
+      "step": 4063
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.99055330634278e-05,
+      "loss": 2.4799,
+      "step": 4064
+    },
+    {
+      "epoch": 1.028593117408907,
+      "grad_norm": 3.125,
+      "learning_rate": 2.990468960863698e-05,
+      "loss": 2.5313,
+      "step": 4065
+    },
+    {
+      "epoch": 1.0288461538461537,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9903846153846156e-05,
+      "loss": 2.37,
+      "step": 4066
+    },
+    {
+      "epoch": 1.0290991902834008,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.990300269905533e-05,
+      "loss": 2.4242,
+      "step": 4067
+    },
+    {
+      "epoch": 1.0293522267206479,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.990215924426451e-05,
+      "loss": 2.5186,
+      "step": 4068
+    },
+    {
+      "epoch": 1.0296052631578947,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9901315789473686e-05,
+      "loss": 2.5813,
+      "step": 4069
+    },
+    {
+      "epoch": 1.0298582995951417,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9900472334682863e-05,
+      "loss": 2.6251,
+      "step": 4070
+    },
+    {
+      "epoch": 1.0301113360323886,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.9899628879892036e-05,
+      "loss": 2.3771,
+      "step": 4071
+    },
+    {
+      "epoch": 1.0303643724696356,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9898785425101216e-05,
+      "loss": 2.3989,
+      "step": 4072
+    },
+    {
+      "epoch": 1.0306174089068827,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9897941970310393e-05,
+      "loss": 2.5867,
+      "step": 4073
+    },
+    {
+      "epoch": 1.0308704453441295,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9897098515519566e-05,
+      "loss": 2.6074,
+      "step": 4074
+    },
+    {
+      "epoch": 1.0311234817813766,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9896255060728746e-05,
+      "loss": 2.5821,
+      "step": 4075
+    },
+    {
+      "epoch": 1.0313765182186234,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9895411605937923e-05,
+      "loss": 2.4522,
+      "step": 4076
+    },
+    {
+      "epoch": 1.0316295546558705,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.98945681511471e-05,
+      "loss": 2.4618,
+      "step": 4077
+    },
+    {
+      "epoch": 1.0318825910931173,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9893724696356276e-05,
+      "loss": 2.5797,
+      "step": 4078
+    },
+    {
+      "epoch": 1.0321356275303644,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9892881241565453e-05,
+      "loss": 2.4199,
+      "step": 4079
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.989203778677463e-05,
+      "loss": 2.558,
+      "step": 4080
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "eval_loss": 2.663304090499878,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.2028,
+      "eval_samples_per_second": 9.448,
+      "eval_steps_per_second": 9.448,
+      "step": 4080
+    },
+    {
+      "epoch": 1.0326417004048583,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.989119433198381e-05,
+      "loss": 2.5467,
+      "step": 4081
+    },
+    {
+      "epoch": 1.0328947368421053,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9890350877192983e-05,
+      "loss": 2.4193,
+      "step": 4082
+    },
+    {
+      "epoch": 1.0331477732793521,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.988950742240216e-05,
+      "loss": 2.5574,
+      "step": 4083
+    },
+    {
+      "epoch": 1.0334008097165992,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.988866396761134e-05,
+      "loss": 2.5798,
+      "step": 4084
+    },
+    {
+      "epoch": 1.0336538461538463,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9887820512820513e-05,
+      "loss": 2.4787,
+      "step": 4085
+    },
+    {
+      "epoch": 1.033906882591093,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.988697705802969e-05,
+      "loss": 2.418,
+      "step": 4086
+    },
+    {
+      "epoch": 1.0341599190283401,
+      "grad_norm": 3.0,
+      "learning_rate": 2.988613360323887e-05,
+      "loss": 2.5114,
+      "step": 4087
+    },
+    {
+      "epoch": 1.034412955465587,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9885290148448046e-05,
+      "loss": 2.5831,
+      "step": 4088
+    },
+    {
+      "epoch": 1.034665991902834,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.988444669365722e-05,
+      "loss": 2.5291,
+      "step": 4089
+    },
+    {
+      "epoch": 1.0349190283400809,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9883603238866396e-05,
+      "loss": 2.4418,
+      "step": 4090
+    },
+    {
+      "epoch": 1.035172064777328,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9882759784075576e-05,
+      "loss": 2.6277,
+      "step": 4091
+    },
+    {
+      "epoch": 1.035425101214575,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.988191632928475e-05,
+      "loss": 2.6055,
+      "step": 4092
+    },
+    {
+      "epoch": 1.0356781376518218,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9881072874493926e-05,
+      "loss": 2.5809,
+      "step": 4093
+    },
+    {
+      "epoch": 1.0359311740890689,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9880229419703106e-05,
+      "loss": 2.3725,
+      "step": 4094
+    },
+    {
+      "epoch": 1.0361842105263157,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9879385964912283e-05,
+      "loss": 2.3405,
+      "step": 4095
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9878542510121456e-05,
+      "loss": 2.5359,
+      "step": 4096
+    },
+    {
+      "epoch": 1.0366902834008098,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9877699055330636e-05,
+      "loss": 2.5518,
+      "step": 4097
+    },
+    {
+      "epoch": 1.0369433198380567,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9876855600539812e-05,
+      "loss": 2.5519,
+      "step": 4098
+    },
+    {
+      "epoch": 1.0371963562753037,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.987601214574899e-05,
+      "loss": 2.5492,
+      "step": 4099
+    },
+    {
+      "epoch": 1.0374493927125505,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9875168690958166e-05,
+      "loss": 2.5663,
+      "step": 4100
+    },
+    {
+      "epoch": 1.0377024291497976,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9874325236167342e-05,
+      "loss": 2.5296,
+      "step": 4101
+    },
+    {
+      "epoch": 1.0379554655870444,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.987348178137652e-05,
+      "loss": 2.6133,
+      "step": 4102
+    },
+    {
+      "epoch": 1.0382085020242915,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.9872638326585696e-05,
+      "loss": 2.4861,
+      "step": 4103
+    },
+    {
+      "epoch": 1.0384615384615385,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9871794871794872e-05,
+      "loss": 2.653,
+      "step": 4104
+    },
+    {
+      "epoch": 1.0387145748987854,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.987095141700405e-05,
+      "loss": 2.5319,
+      "step": 4105
+    },
+    {
+      "epoch": 1.0389676113360324,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9870107962213226e-05,
+      "loss": 2.4756,
+      "step": 4106
+    },
+    {
+      "epoch": 1.0392206477732793,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9869264507422402e-05,
+      "loss": 2.5107,
+      "step": 4107
+    },
+    {
+      "epoch": 1.0394736842105263,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.986842105263158e-05,
+      "loss": 2.5537,
+      "step": 4108
+    },
+    {
+      "epoch": 1.0397267206477734,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9867577597840756e-05,
+      "loss": 2.6251,
+      "step": 4109
+    },
+    {
+      "epoch": 1.0399797570850202,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9866734143049936e-05,
+      "loss": 2.527,
+      "step": 4110
+    },
+    {
+      "epoch": 1.0399797570850202,
+      "eval_loss": 2.669276475906372,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.2116,
+      "eval_samples_per_second": 9.447,
+      "eval_steps_per_second": 9.447,
+      "step": 4110
+    },
+    {
+      "epoch": 1.0402327935222673,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.986589068825911e-05,
+      "loss": 2.6064,
+      "step": 4111
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.9865047233468286e-05,
+      "loss": 2.5796,
+      "step": 4112
+    },
+    {
+      "epoch": 1.0407388663967612,
+      "grad_norm": 3.625,
+      "learning_rate": 2.9864203778677466e-05,
+      "loss": 2.5859,
+      "step": 4113
+    },
+    {
+      "epoch": 1.040991902834008,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.986336032388664e-05,
+      "loss": 2.4497,
+      "step": 4114
+    },
+    {
+      "epoch": 1.041244939271255,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9862516869095816e-05,
+      "loss": 2.4769,
+      "step": 4115
+    },
+    {
+      "epoch": 1.041497975708502,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.9861673414304996e-05,
+      "loss": 2.6317,
+      "step": 4116
+    },
+    {
+      "epoch": 1.041751012145749,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9860829959514172e-05,
+      "loss": 2.542,
+      "step": 4117
+    },
+    {
+      "epoch": 1.042004048582996,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9859986504723346e-05,
+      "loss": 2.5564,
+      "step": 4118
+    },
+    {
+      "epoch": 1.0422570850202428,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9859143049932526e-05,
+      "loss": 2.5818,
+      "step": 4119
+    },
+    {
+      "epoch": 1.04251012145749,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.9858299595141702e-05,
+      "loss": 2.3117,
+      "step": 4120
+    },
+    {
+      "epoch": 1.042763157894737,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.985745614035088e-05,
+      "loss": 2.5312,
+      "step": 4121
+    },
+    {
+      "epoch": 1.0430161943319838,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9856612685560056e-05,
+      "loss": 2.4934,
+      "step": 4122
+    },
+    {
+      "epoch": 1.0432692307692308,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.9855769230769232e-05,
+      "loss": 2.6403,
+      "step": 4123
+    },
+    {
+      "epoch": 1.0435222672064777,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.985492577597841e-05,
+      "loss": 2.5089,
+      "step": 4124
+    },
+    {
+      "epoch": 1.0437753036437247,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9854082321187582e-05,
+      "loss": 2.5574,
+      "step": 4125
+    },
+    {
+      "epoch": 1.0440283400809716,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9853238866396762e-05,
+      "loss": 2.4938,
+      "step": 4126
+    },
+    {
+      "epoch": 1.0442813765182186,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.985239541160594e-05,
+      "loss": 2.5835,
+      "step": 4127
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9851551956815115e-05,
+      "loss": 2.5449,
+      "step": 4128
+    },
+    {
+      "epoch": 1.0447874493927125,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9850708502024292e-05,
+      "loss": 2.5627,
+      "step": 4129
+    },
+    {
+      "epoch": 1.0450404858299596,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.984986504723347e-05,
+      "loss": 2.4407,
+      "step": 4130
+    },
+    {
+      "epoch": 1.0452935222672064,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9849021592442645e-05,
+      "loss": 2.5484,
+      "step": 4131
+    },
+    {
+      "epoch": 1.0455465587044535,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9848178137651822e-05,
+      "loss": 2.5021,
+      "step": 4132
+    },
+    {
+      "epoch": 1.0457995951417005,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9847334682861e-05,
+      "loss": 2.4918,
+      "step": 4133
+    },
+    {
+      "epoch": 1.0460526315789473,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.9846491228070175e-05,
+      "loss": 2.5768,
+      "step": 4134
+    },
+    {
+      "epoch": 1.0463056680161944,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9845647773279355e-05,
+      "loss": 2.4994,
+      "step": 4135
+    },
+    {
+      "epoch": 1.0465587044534412,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.984480431848853e-05,
+      "loss": 2.5677,
+      "step": 4136
+    },
+    {
+      "epoch": 1.0468117408906883,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9843960863697705e-05,
+      "loss": 2.5444,
+      "step": 4137
+    },
+    {
+      "epoch": 1.0470647773279351,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9843117408906885e-05,
+      "loss": 2.5814,
+      "step": 4138
+    },
+    {
+      "epoch": 1.0473178137651822,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.9842273954116062e-05,
+      "loss": 2.6205,
+      "step": 4139
+    },
+    {
+      "epoch": 1.0475708502024292,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9841430499325235e-05,
+      "loss": 2.5229,
+      "step": 4140
+    },
+    {
+      "epoch": 1.0475708502024292,
+      "eval_loss": 2.661015510559082,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.6012,
+      "eval_samples_per_second": 9.468,
+      "eval_steps_per_second": 9.468,
+      "step": 4140
+    },
+    {
+      "epoch": 1.047823886639676,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9840587044534415e-05,
+      "loss": 2.6253,
+      "step": 4141
+    },
+    {
+      "epoch": 1.0480769230769231,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9839743589743592e-05,
+      "loss": 2.6939,
+      "step": 4142
+    },
+    {
+      "epoch": 1.04832995951417,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9838900134952765e-05,
+      "loss": 2.5211,
+      "step": 4143
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9838056680161942e-05,
+      "loss": 2.5509,
+      "step": 4144
+    },
+    {
+      "epoch": 1.048836032388664,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9837213225371122e-05,
+      "loss": 2.5622,
+      "step": 4145
+    },
+    {
+      "epoch": 1.049089068825911,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.98363697705803e-05,
+      "loss": 2.6078,
+      "step": 4146
+    },
+    {
+      "epoch": 1.049342105263158,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9835526315789472e-05,
+      "loss": 2.3304,
+      "step": 4147
+    },
+    {
+      "epoch": 1.0495951417004048,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9834682860998652e-05,
+      "loss": 2.5965,
+      "step": 4148
+    },
+    {
+      "epoch": 1.0498481781376519,
+      "grad_norm": 3.125,
+      "learning_rate": 2.983383940620783e-05,
+      "loss": 2.5784,
+      "step": 4149
+    },
+    {
+      "epoch": 1.0501012145748987,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.9832995951417005e-05,
+      "loss": 2.5687,
+      "step": 4150
+    },
+    {
+      "epoch": 1.0503542510121457,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9832152496626182e-05,
+      "loss": 2.4057,
+      "step": 4151
+    },
+    {
+      "epoch": 1.0506072874493928,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.983130904183536e-05,
+      "loss": 2.6667,
+      "step": 4152
+    },
+    {
+      "epoch": 1.0508603238866396,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.9830465587044535e-05,
+      "loss": 2.5518,
+      "step": 4153
+    },
+    {
+      "epoch": 1.0511133603238867,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9829622132253712e-05,
+      "loss": 2.6093,
+      "step": 4154
+    },
+    {
+      "epoch": 1.0513663967611335,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.982877867746289e-05,
+      "loss": 2.4444,
+      "step": 4155
+    },
+    {
+      "epoch": 1.0516194331983806,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9827935222672065e-05,
+      "loss": 2.6072,
+      "step": 4156
+    },
+    {
+      "epoch": 1.0518724696356276,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9827091767881245e-05,
+      "loss": 2.59,
+      "step": 4157
+    },
+    {
+      "epoch": 1.0521255060728745,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.982624831309042e-05,
+      "loss": 2.4709,
+      "step": 4158
+    },
+    {
+      "epoch": 1.0523785425101215,
+      "grad_norm": 3.75,
+      "learning_rate": 2.9825404858299595e-05,
+      "loss": 2.5442,
+      "step": 4159
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9824561403508772e-05,
+      "loss": 2.4481,
+      "step": 4160
+    },
+    {
+      "epoch": 1.0528846153846154,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.982371794871795e-05,
+      "loss": 2.5787,
+      "step": 4161
+    },
+    {
+      "epoch": 1.0531376518218623,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9822874493927125e-05,
+      "loss": 2.4864,
+      "step": 4162
+    },
+    {
+      "epoch": 1.0533906882591093,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9822031039136302e-05,
+      "loss": 2.4291,
+      "step": 4163
+    },
+    {
+      "epoch": 1.0536437246963564,
+      "grad_norm": 20.125,
+      "learning_rate": 2.9821187584345482e-05,
+      "loss": 2.6158,
+      "step": 4164
+    },
+    {
+      "epoch": 1.0538967611336032,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9820344129554655e-05,
+      "loss": 2.5794,
+      "step": 4165
+    },
+    {
+      "epoch": 1.0541497975708503,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.981950067476383e-05,
+      "loss": 2.6893,
+      "step": 4166
+    },
+    {
+      "epoch": 1.054402834008097,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9818657219973012e-05,
+      "loss": 2.5211,
+      "step": 4167
+    },
+    {
+      "epoch": 1.0546558704453441,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.981781376518219e-05,
+      "loss": 2.6379,
+      "step": 4168
+    },
+    {
+      "epoch": 1.054908906882591,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.981697031039136e-05,
+      "loss": 2.5262,
+      "step": 4169
+    },
+    {
+      "epoch": 1.055161943319838,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.981612685560054e-05,
+      "loss": 2.6122,
+      "step": 4170
+    },
+    {
+      "epoch": 1.055161943319838,
+      "eval_loss": 2.6581344604492188,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.3325,
+      "eval_samples_per_second": 9.478,
+      "eval_steps_per_second": 9.478,
+      "step": 4170
+    },
+    {
+      "epoch": 1.055414979757085,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.981528340080972e-05,
+      "loss": 2.4524,
+      "step": 4171
+    },
+    {
+      "epoch": 1.055668016194332,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.981443994601889e-05,
+      "loss": 2.5337,
+      "step": 4172
+    },
+    {
+      "epoch": 1.055921052631579,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.981359649122807e-05,
+      "loss": 2.5887,
+      "step": 4173
+    },
+    {
+      "epoch": 1.0561740890688258,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9812753036437248e-05,
+      "loss": 2.537,
+      "step": 4174
+    },
+    {
+      "epoch": 1.0564271255060729,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9811909581646425e-05,
+      "loss": 2.5488,
+      "step": 4175
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.98110661268556e-05,
+      "loss": 2.5756,
+      "step": 4176
+    },
+    {
+      "epoch": 1.0569331983805668,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9810222672064778e-05,
+      "loss": 2.4902,
+      "step": 4177
+    },
+    {
+      "epoch": 1.0571862348178138,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9809379217273955e-05,
+      "loss": 2.6018,
+      "step": 4178
+    },
+    {
+      "epoch": 1.0574392712550607,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.980853576248313e-05,
+      "loss": 2.54,
+      "step": 4179
+    },
+    {
+      "epoch": 1.0576923076923077,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9807692307692308e-05,
+      "loss": 2.6139,
+      "step": 4180
+    },
+    {
+      "epoch": 1.0579453441295548,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9806848852901485e-05,
+      "loss": 2.4853,
+      "step": 4181
+    },
+    {
+      "epoch": 1.0581983805668016,
+      "grad_norm": 3.125,
+      "learning_rate": 2.980600539811066e-05,
+      "loss": 2.5307,
+      "step": 4182
+    },
+    {
+      "epoch": 1.0584514170040487,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9805161943319838e-05,
+      "loss": 2.4705,
+      "step": 4183
+    },
+    {
+      "epoch": 1.0587044534412955,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9804318488529015e-05,
+      "loss": 2.5814,
+      "step": 4184
+    },
+    {
+      "epoch": 1.0589574898785425,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.980347503373819e-05,
+      "loss": 2.5951,
+      "step": 4185
+    },
+    {
+      "epoch": 1.0592105263157894,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.980263157894737e-05,
+      "loss": 2.6388,
+      "step": 4186
+    },
+    {
+      "epoch": 1.0594635627530364,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9801788124156545e-05,
+      "loss": 2.4726,
+      "step": 4187
+    },
+    {
+      "epoch": 1.0597165991902835,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.980094466936572e-05,
+      "loss": 2.6078,
+      "step": 4188
+    },
+    {
+      "epoch": 1.0599696356275303,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.98001012145749e-05,
+      "loss": 2.5275,
+      "step": 4189
+    },
+    {
+      "epoch": 1.0602226720647774,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9799257759784078e-05,
+      "loss": 2.4812,
+      "step": 4190
+    },
+    {
+      "epoch": 1.0604757085020242,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.979841430499325e-05,
+      "loss": 2.585,
+      "step": 4191
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.979757085020243e-05,
+      "loss": 2.4503,
+      "step": 4192
+    },
+    {
+      "epoch": 1.060981781376518,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.9796727395411608e-05,
+      "loss": 2.7194,
+      "step": 4193
+    },
+    {
+      "epoch": 1.0612348178137652,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.979588394062078e-05,
+      "loss": 2.5749,
+      "step": 4194
+    },
+    {
+      "epoch": 1.0614878542510122,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.979504048582996e-05,
+      "loss": 2.5007,
+      "step": 4195
+    },
+    {
+      "epoch": 1.061740890688259,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9794197031039138e-05,
+      "loss": 2.5397,
+      "step": 4196
+    },
+    {
+      "epoch": 1.061993927125506,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9793353576248315e-05,
+      "loss": 2.4877,
+      "step": 4197
+    },
+    {
+      "epoch": 1.062246963562753,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9792510121457488e-05,
+      "loss": 2.3997,
+      "step": 4198
+    },
+    {
+      "epoch": 1.0625,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9791666666666668e-05,
+      "loss": 2.6275,
+      "step": 4199
+    },
+    {
+      "epoch": 1.062753036437247,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9790823211875845e-05,
+      "loss": 2.5029,
+      "step": 4200
+    },
+    {
+      "epoch": 1.062753036437247,
+      "eval_loss": 2.6579222679138184,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.4153,
+      "eval_samples_per_second": 9.475,
+      "eval_steps_per_second": 9.475,
+      "step": 4200
+    },
+    {
+      "epoch": 1.063006072874494,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.978997975708502e-05,
+      "loss": 2.5612,
+      "step": 4201
+    },
+    {
+      "epoch": 1.063259109311741,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.9789136302294198e-05,
+      "loss": 2.3169,
+      "step": 4202
+    },
+    {
+      "epoch": 1.0635121457489878,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9788292847503375e-05,
+      "loss": 2.5658,
+      "step": 4203
+    },
+    {
+      "epoch": 1.0637651821862348,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.978744939271255e-05,
+      "loss": 2.3314,
+      "step": 4204
+    },
+    {
+      "epoch": 1.064018218623482,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.9786605937921728e-05,
+      "loss": 2.3437,
+      "step": 4205
+    },
+    {
+      "epoch": 1.0642712550607287,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9785762483130905e-05,
+      "loss": 2.4701,
+      "step": 4206
+    },
+    {
+      "epoch": 1.0645242914979758,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.978491902834008e-05,
+      "loss": 2.5309,
+      "step": 4207
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.978407557354926e-05,
+      "loss": 2.4808,
+      "step": 4208
+    },
+    {
+      "epoch": 1.0650303643724697,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9783232118758435e-05,
+      "loss": 2.6048,
+      "step": 4209
+    },
+    {
+      "epoch": 1.0652834008097165,
+      "grad_norm": 3.0,
+      "learning_rate": 2.978238866396761e-05,
+      "loss": 2.4563,
+      "step": 4210
+    },
+    {
+      "epoch": 1.0655364372469636,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.978154520917679e-05,
+      "loss": 2.5349,
+      "step": 4211
+    },
+    {
+      "epoch": 1.0657894736842106,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9780701754385965e-05,
+      "loss": 2.5737,
+      "step": 4212
+    },
+    {
+      "epoch": 1.0660425101214575,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.977985829959514e-05,
+      "loss": 2.5399,
+      "step": 4213
+    },
+    {
+      "epoch": 1.0662955465587045,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.977901484480432e-05,
+      "loss": 2.4766,
+      "step": 4214
+    },
+    {
+      "epoch": 1.0665485829959513,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9778171390013498e-05,
+      "loss": 2.5528,
+      "step": 4215
+    },
+    {
+      "epoch": 1.0668016194331984,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.977732793522267e-05,
+      "loss": 2.5901,
+      "step": 4216
+    },
+    {
+      "epoch": 1.0670546558704452,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9776484480431848e-05,
+      "loss": 2.5203,
+      "step": 4217
+    },
+    {
+      "epoch": 1.0673076923076923,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9775641025641028e-05,
+      "loss": 2.3745,
+      "step": 4218
+    },
+    {
+      "epoch": 1.0675607287449393,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9774797570850204e-05,
+      "loss": 2.5673,
+      "step": 4219
+    },
+    {
+      "epoch": 1.0678137651821862,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9773954116059378e-05,
+      "loss": 2.5812,
+      "step": 4220
+    },
+    {
+      "epoch": 1.0680668016194332,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9773110661268558e-05,
+      "loss": 2.5594,
+      "step": 4221
+    },
+    {
+      "epoch": 1.06831983805668,
+      "grad_norm": 3.5,
+      "learning_rate": 2.9772267206477734e-05,
+      "loss": 2.6681,
+      "step": 4222
+    },
+    {
+      "epoch": 1.0685728744939271,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9771423751686908e-05,
+      "loss": 2.5575,
+      "step": 4223
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9770580296896088e-05,
+      "loss": 2.5518,
+      "step": 4224
+    },
+    {
+      "epoch": 1.069078947368421,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9769736842105264e-05,
+      "loss": 2.5258,
+      "step": 4225
+    },
+    {
+      "epoch": 1.069331983805668,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.976889338731444e-05,
+      "loss": 2.4572,
+      "step": 4226
+    },
+    {
+      "epoch": 1.069585020242915,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9768049932523618e-05,
+      "loss": 2.529,
+      "step": 4227
+    },
+    {
+      "epoch": 1.069838056680162,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9767206477732794e-05,
+      "loss": 2.5127,
+      "step": 4228
+    },
+    {
+      "epoch": 1.070091093117409,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.976636302294197e-05,
+      "loss": 2.3597,
+      "step": 4229
+    },
+    {
+      "epoch": 1.0703441295546559,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.976551956815115e-05,
+      "loss": 2.6265,
+      "step": 4230
+    },
+    {
+      "epoch": 1.0703441295546559,
+      "eval_loss": 2.6599621772766113,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.5023,
+      "eval_samples_per_second": 9.472,
+      "eval_steps_per_second": 9.472,
+      "step": 4230
+    },
+    {
+      "epoch": 1.070597165991903,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9764676113360324e-05,
+      "loss": 2.5475,
+      "step": 4231
+    },
+    {
+      "epoch": 1.0708502024291497,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.97638326585695e-05,
+      "loss": 2.4883,
+      "step": 4232
+    },
+    {
+      "epoch": 1.0711032388663968,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9762989203778678e-05,
+      "loss": 2.4223,
+      "step": 4233
+    },
+    {
+      "epoch": 1.0713562753036436,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9762145748987854e-05,
+      "loss": 2.5975,
+      "step": 4234
+    },
+    {
+      "epoch": 1.0716093117408907,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.976130229419703e-05,
+      "loss": 2.4883,
+      "step": 4235
+    },
+    {
+      "epoch": 1.0718623481781377,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.9760458839406208e-05,
+      "loss": 2.5381,
+      "step": 4236
+    },
+    {
+      "epoch": 1.0721153846153846,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9759615384615388e-05,
+      "loss": 2.583,
+      "step": 4237
+    },
+    {
+      "epoch": 1.0723684210526316,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.975877192982456e-05,
+      "loss": 2.5227,
+      "step": 4238
+    },
+    {
+      "epoch": 1.0726214574898785,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9757928475033738e-05,
+      "loss": 2.4414,
+      "step": 4239
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.9757085020242918e-05,
+      "loss": 2.5998,
+      "step": 4240
+    },
+    {
+      "epoch": 1.0731275303643724,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.975624156545209e-05,
+      "loss": 2.5917,
+      "step": 4241
+    },
+    {
+      "epoch": 1.0733805668016194,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9755398110661267e-05,
+      "loss": 2.4811,
+      "step": 4242
+    },
+    {
+      "epoch": 1.0736336032388665,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.9754554655870448e-05,
+      "loss": 2.4447,
+      "step": 4243
+    },
+    {
+      "epoch": 1.0738866396761133,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9753711201079624e-05,
+      "loss": 2.5413,
+      "step": 4244
+    },
+    {
+      "epoch": 1.0741396761133604,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9752867746288797e-05,
+      "loss": 2.4458,
+      "step": 4245
+    },
+    {
+      "epoch": 1.0743927125506072,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9752024291497977e-05,
+      "loss": 2.5469,
+      "step": 4246
+    },
+    {
+      "epoch": 1.0746457489878543,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9751180836707154e-05,
+      "loss": 2.575,
+      "step": 4247
+    },
+    {
+      "epoch": 1.0748987854251013,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.975033738191633e-05,
+      "loss": 2.6264,
+      "step": 4248
+    },
+    {
+      "epoch": 1.0751518218623481,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9749493927125507e-05,
+      "loss": 2.5696,
+      "step": 4249
+    },
+    {
+      "epoch": 1.0754048582995952,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9748650472334684e-05,
+      "loss": 2.4756,
+      "step": 4250
+    },
+    {
+      "epoch": 1.075657894736842,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.974780701754386e-05,
+      "loss": 2.5714,
+      "step": 4251
+    },
+    {
+      "epoch": 1.075910931174089,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9746963562753034e-05,
+      "loss": 2.3994,
+      "step": 4252
+    },
+    {
+      "epoch": 1.0761639676113361,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9746120107962214e-05,
+      "loss": 2.4242,
+      "step": 4253
+    },
+    {
+      "epoch": 1.076417004048583,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.974527665317139e-05,
+      "loss": 2.5894,
+      "step": 4254
+    },
+    {
+      "epoch": 1.07667004048583,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9744433198380567e-05,
+      "loss": 2.6142,
+      "step": 4255
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9743589743589744e-05,
+      "loss": 2.4868,
+      "step": 4256
+    },
+    {
+      "epoch": 1.077176113360324,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.974274628879892e-05,
+      "loss": 2.5947,
+      "step": 4257
+    },
+    {
+      "epoch": 1.0774291497975708,
+      "grad_norm": 3.5,
+      "learning_rate": 2.9741902834008097e-05,
+      "loss": 2.5782,
+      "step": 4258
+    },
+    {
+      "epoch": 1.0776821862348178,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9741059379217277e-05,
+      "loss": 2.6418,
+      "step": 4259
+    },
+    {
+      "epoch": 1.0779352226720649,
+      "grad_norm": 3.125,
+      "learning_rate": 2.974021592442645e-05,
+      "loss": 2.5577,
+      "step": 4260
+    },
+    {
+      "epoch": 1.0779352226720649,
+      "eval_loss": 2.6558804512023926,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.2718,
+      "eval_samples_per_second": 9.514,
+      "eval_steps_per_second": 9.514,
+      "step": 4260
+    },
+    {
+      "epoch": 1.0781882591093117,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9739372469635627e-05,
+      "loss": 2.5099,
+      "step": 4261
+    },
+    {
+      "epoch": 1.0784412955465588,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9738529014844807e-05,
+      "loss": 2.467,
+      "step": 4262
+    },
+    {
+      "epoch": 1.0786943319838056,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.973768556005398e-05,
+      "loss": 2.3692,
+      "step": 4263
+    },
+    {
+      "epoch": 1.0789473684210527,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9736842105263157e-05,
+      "loss": 2.6205,
+      "step": 4264
+    },
+    {
+      "epoch": 1.0792004048582995,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9735998650472337e-05,
+      "loss": 2.5724,
+      "step": 4265
+    },
+    {
+      "epoch": 1.0794534412955465,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9735155195681514e-05,
+      "loss": 2.4945,
+      "step": 4266
+    },
+    {
+      "epoch": 1.0797064777327936,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9734311740890687e-05,
+      "loss": 2.5327,
+      "step": 4267
+    },
+    {
+      "epoch": 1.0799595141700404,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9733468286099867e-05,
+      "loss": 2.526,
+      "step": 4268
+    },
+    {
+      "epoch": 1.0802125506072875,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9732624831309044e-05,
+      "loss": 2.6114,
+      "step": 4269
+    },
+    {
+      "epoch": 1.0804655870445343,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.973178137651822e-05,
+      "loss": 2.374,
+      "step": 4270
+    },
+    {
+      "epoch": 1.0807186234817814,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9730937921727394e-05,
+      "loss": 2.4157,
+      "step": 4271
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9730094466936574e-05,
+      "loss": 2.4415,
+      "step": 4272
+    },
+    {
+      "epoch": 1.0812246963562753,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.972925101214575e-05,
+      "loss": 2.4375,
+      "step": 4273
+    },
+    {
+      "epoch": 1.0814777327935223,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9728407557354924e-05,
+      "loss": 2.5395,
+      "step": 4274
+    },
+    {
+      "epoch": 1.0817307692307692,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.9727564102564104e-05,
+      "loss": 2.5748,
+      "step": 4275
+    },
+    {
+      "epoch": 1.0819838056680162,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.972672064777328e-05,
+      "loss": 2.5341,
+      "step": 4276
+    },
+    {
+      "epoch": 1.0822368421052633,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9725877192982457e-05,
+      "loss": 2.1984,
+      "step": 4277
+    },
+    {
+      "epoch": 1.08248987854251,
+      "grad_norm": 3.5,
+      "learning_rate": 2.9725033738191634e-05,
+      "loss": 2.5309,
+      "step": 4278
+    },
+    {
+      "epoch": 1.0827429149797572,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.972419028340081e-05,
+      "loss": 2.6218,
+      "step": 4279
+    },
+    {
+      "epoch": 1.082995951417004,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9723346828609987e-05,
+      "loss": 2.415,
+      "step": 4280
+    },
+    {
+      "epoch": 1.083248987854251,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9722503373819164e-05,
+      "loss": 2.5485,
+      "step": 4281
+    },
+    {
+      "epoch": 1.083502024291498,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.972165991902834e-05,
+      "loss": 2.4199,
+      "step": 4282
+    },
+    {
+      "epoch": 1.083755060728745,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9720816464237517e-05,
+      "loss": 2.5741,
+      "step": 4283
+    },
+    {
+      "epoch": 1.084008097165992,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9719973009446697e-05,
+      "loss": 2.5544,
+      "step": 4284
+    },
+    {
+      "epoch": 1.0842611336032388,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.971912955465587e-05,
+      "loss": 2.6655,
+      "step": 4285
+    },
+    {
+      "epoch": 1.084514170040486,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9718286099865047e-05,
+      "loss": 2.4714,
+      "step": 4286
+    },
+    {
+      "epoch": 1.0847672064777327,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9717442645074224e-05,
+      "loss": 2.4504,
+      "step": 4287
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9716599190283404e-05,
+      "loss": 2.5013,
+      "step": 4288
+    },
+    {
+      "epoch": 1.0852732793522266,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9715755735492577e-05,
+      "loss": 2.4706,
+      "step": 4289
+    },
+    {
+      "epoch": 1.0855263157894737,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9714912280701754e-05,
+      "loss": 2.5107,
+      "step": 4290
+    },
+    {
+      "epoch": 1.0855263157894737,
+      "eval_loss": 2.6505627632141113,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.2272,
+      "eval_samples_per_second": 9.447,
+      "eval_steps_per_second": 9.447,
+      "step": 4290
+    },
+    {
+      "epoch": 1.0857793522267207,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9714068825910934e-05,
+      "loss": 2.5102,
+      "step": 4291
+    },
+    {
+      "epoch": 1.0860323886639676,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9713225371120107e-05,
+      "loss": 2.4444,
+      "step": 4292
+    },
+    {
+      "epoch": 1.0862854251012146,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.9712381916329284e-05,
+      "loss": 2.5574,
+      "step": 4293
+    },
+    {
+      "epoch": 1.0865384615384615,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9711538461538464e-05,
+      "loss": 2.6163,
+      "step": 4294
+    },
+    {
+      "epoch": 1.0867914979757085,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.971069500674764e-05,
+      "loss": 2.4876,
+      "step": 4295
+    },
+    {
+      "epoch": 1.0870445344129556,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.9709851551956814e-05,
+      "loss": 2.6661,
+      "step": 4296
+    },
+    {
+      "epoch": 1.0872975708502024,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9709008097165994e-05,
+      "loss": 2.5292,
+      "step": 4297
+    },
+    {
+      "epoch": 1.0875506072874495,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.970816464237517e-05,
+      "loss": 2.3948,
+      "step": 4298
+    },
+    {
+      "epoch": 1.0878036437246963,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9707321187584347e-05,
+      "loss": 2.5847,
+      "step": 4299
+    },
+    {
+      "epoch": 1.0880566801619433,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9706477732793524e-05,
+      "loss": 2.5329,
+      "step": 4300
+    },
+    {
+      "epoch": 1.0883097165991902,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.97056342780027e-05,
+      "loss": 2.5278,
+      "step": 4301
+    },
+    {
+      "epoch": 1.0885627530364372,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9704790823211877e-05,
+      "loss": 2.5007,
+      "step": 4302
+    },
+    {
+      "epoch": 1.0888157894736843,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9703947368421054e-05,
+      "loss": 2.4743,
+      "step": 4303
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.970310391363023e-05,
+      "loss": 2.5187,
+      "step": 4304
+    },
+    {
+      "epoch": 1.0893218623481782,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9702260458839407e-05,
+      "loss": 2.5166,
+      "step": 4305
+    },
+    {
+      "epoch": 1.089574898785425,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9701417004048583e-05,
+      "loss": 2.5593,
+      "step": 4306
+    },
+    {
+      "epoch": 1.089827935222672,
+      "grad_norm": 3.25,
+      "learning_rate": 2.970057354925776e-05,
+      "loss": 2.5706,
+      "step": 4307
+    },
+    {
+      "epoch": 1.0900809716599191,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9699730094466937e-05,
+      "loss": 2.5925,
+      "step": 4308
+    },
+    {
+      "epoch": 1.090334008097166,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9698886639676113e-05,
+      "loss": 2.6102,
+      "step": 4309
+    },
+    {
+      "epoch": 1.090587044534413,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9698043184885293e-05,
+      "loss": 2.5449,
+      "step": 4310
+    },
+    {
+      "epoch": 1.0908400809716599,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9697199730094467e-05,
+      "loss": 2.4308,
+      "step": 4311
+    },
+    {
+      "epoch": 1.091093117408907,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9696356275303643e-05,
+      "loss": 2.5077,
+      "step": 4312
+    },
+    {
+      "epoch": 1.0913461538461537,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.9695512820512823e-05,
+      "loss": 2.4353,
+      "step": 4313
+    },
+    {
+      "epoch": 1.0915991902834008,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9694669365721997e-05,
+      "loss": 2.5222,
+      "step": 4314
+    },
+    {
+      "epoch": 1.0918522267206479,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9693825910931173e-05,
+      "loss": 2.5165,
+      "step": 4315
+    },
+    {
+      "epoch": 1.0921052631578947,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9692982456140353e-05,
+      "loss": 2.5881,
+      "step": 4316
+    },
+    {
+      "epoch": 1.0923582995951417,
+      "grad_norm": 3.0,
+      "learning_rate": 2.969213900134953e-05,
+      "loss": 2.5776,
+      "step": 4317
+    },
+    {
+      "epoch": 1.0926113360323886,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9691295546558703e-05,
+      "loss": 2.6104,
+      "step": 4318
+    },
+    {
+      "epoch": 1.0928643724696356,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9690452091767883e-05,
+      "loss": 2.5876,
+      "step": 4319
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.968960863697706e-05,
+      "loss": 2.6755,
+      "step": 4320
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "eval_loss": 2.651526927947998,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.1047,
+      "eval_samples_per_second": 9.451,
+      "eval_steps_per_second": 9.451,
+      "step": 4320
+    },
+    {
+      "epoch": 1.0933704453441295,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9688765182186233e-05,
+      "loss": 2.5221,
+      "step": 4321
+    },
+    {
+      "epoch": 1.0936234817813766,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.9687921727395413e-05,
+      "loss": 2.6166,
+      "step": 4322
+    },
+    {
+      "epoch": 1.0938765182186234,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.968707827260459e-05,
+      "loss": 2.467,
+      "step": 4323
+    },
+    {
+      "epoch": 1.0941295546558705,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9686234817813767e-05,
+      "loss": 2.5383,
+      "step": 4324
+    },
+    {
+      "epoch": 1.0943825910931173,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.968539136302294e-05,
+      "loss": 2.5235,
+      "step": 4325
+    },
+    {
+      "epoch": 1.0946356275303644,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.968454790823212e-05,
+      "loss": 2.4918,
+      "step": 4326
+    },
+    {
+      "epoch": 1.0948886639676114,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9683704453441297e-05,
+      "loss": 2.5217,
+      "step": 4327
+    },
+    {
+      "epoch": 1.0951417004048583,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9682860998650473e-05,
+      "loss": 2.5678,
+      "step": 4328
+    },
+    {
+      "epoch": 1.0953947368421053,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.968201754385965e-05,
+      "loss": 2.617,
+      "step": 4329
+    },
+    {
+      "epoch": 1.0956477732793521,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9681174089068827e-05,
+      "loss": 2.5367,
+      "step": 4330
+    },
+    {
+      "epoch": 1.0959008097165992,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9680330634278003e-05,
+      "loss": 2.5098,
+      "step": 4331
+    },
+    {
+      "epoch": 1.0961538461538463,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.967948717948718e-05,
+      "loss": 2.6251,
+      "step": 4332
+    },
+    {
+      "epoch": 1.096406882591093,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9678643724696356e-05,
+      "loss": 2.4889,
+      "step": 4333
+    },
+    {
+      "epoch": 1.0966599190283401,
+      "grad_norm": 3.5,
+      "learning_rate": 2.9677800269905533e-05,
+      "loss": 2.5437,
+      "step": 4334
+    },
+    {
+      "epoch": 1.096912955465587,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9676956815114713e-05,
+      "loss": 2.6669,
+      "step": 4335
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9676113360323886e-05,
+      "loss": 2.5544,
+      "step": 4336
+    },
+    {
+      "epoch": 1.0974190283400809,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.9675269905533063e-05,
+      "loss": 2.5122,
+      "step": 4337
+    },
+    {
+      "epoch": 1.097672064777328,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9674426450742243e-05,
+      "loss": 2.4527,
+      "step": 4338
+    },
+    {
+      "epoch": 1.097925101214575,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.967358299595142e-05,
+      "loss": 2.5623,
+      "step": 4339
+    },
+    {
+      "epoch": 1.0981781376518218,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.9672739541160593e-05,
+      "loss": 2.3525,
+      "step": 4340
+    },
+    {
+      "epoch": 1.0984311740890689,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.967189608636977e-05,
+      "loss": 2.5951,
+      "step": 4341
+    },
+    {
+      "epoch": 1.0986842105263157,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.967105263157895e-05,
+      "loss": 2.6134,
+      "step": 4342
+    },
+    {
+      "epoch": 1.0989372469635628,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9670209176788123e-05,
+      "loss": 2.5487,
+      "step": 4343
+    },
+    {
+      "epoch": 1.0991902834008098,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.96693657219973e-05,
+      "loss": 2.5002,
+      "step": 4344
+    },
+    {
+      "epoch": 1.0994433198380567,
+      "grad_norm": 3.125,
+      "learning_rate": 2.966852226720648e-05,
+      "loss": 2.5354,
+      "step": 4345
+    },
+    {
+      "epoch": 1.0996963562753037,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.9667678812415656e-05,
+      "loss": 2.5398,
+      "step": 4346
+    },
+    {
+      "epoch": 1.0999493927125505,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.966683535762483e-05,
+      "loss": 2.5632,
+      "step": 4347
+    },
+    {
+      "epoch": 1.1002024291497976,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.966599190283401e-05,
+      "loss": 2.3093,
+      "step": 4348
+    },
+    {
+      "epoch": 1.1004554655870444,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9665148448043186e-05,
+      "loss": 2.5698,
+      "step": 4349
+    },
+    {
+      "epoch": 1.1007085020242915,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9664304993252363e-05,
+      "loss": 2.4973,
+      "step": 4350
+    },
+    {
+      "epoch": 1.1007085020242915,
+      "eval_loss": 2.6540446281433105,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.0564,
+      "eval_samples_per_second": 9.453,
+      "eval_steps_per_second": 9.453,
+      "step": 4350
+    },
+    {
+      "epoch": 1.1009615384615385,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.966346153846154e-05,
+      "loss": 2.5233,
+      "step": 4351
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9662618083670716e-05,
+      "loss": 2.5482,
+      "step": 4352
+    },
+    {
+      "epoch": 1.1014676113360324,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9661774628879893e-05,
+      "loss": 2.5486,
+      "step": 4353
+    },
+    {
+      "epoch": 1.1017206477732793,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.966093117408907e-05,
+      "loss": 2.4042,
+      "step": 4354
+    },
+    {
+      "epoch": 1.1019736842105263,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.9660087719298246e-05,
+      "loss": 2.5848,
+      "step": 4355
+    },
+    {
+      "epoch": 1.1022267206477734,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9659244264507423e-05,
+      "loss": 2.4852,
+      "step": 4356
+    },
+    {
+      "epoch": 1.1024797570850202,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9658400809716603e-05,
+      "loss": 2.5027,
+      "step": 4357
+    },
+    {
+      "epoch": 1.1027327935222673,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9657557354925776e-05,
+      "loss": 2.5419,
+      "step": 4358
+    },
+    {
+      "epoch": 1.102985829959514,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9656713900134953e-05,
+      "loss": 2.5937,
+      "step": 4359
+    },
+    {
+      "epoch": 1.1032388663967612,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.965587044534413e-05,
+      "loss": 2.6344,
+      "step": 4360
+    },
+    {
+      "epoch": 1.103491902834008,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.9655026990553306e-05,
+      "loss": 2.5204,
+      "step": 4361
+    },
+    {
+      "epoch": 1.103744939271255,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9654183535762483e-05,
+      "loss": 2.4961,
+      "step": 4362
+    },
+    {
+      "epoch": 1.103997975708502,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.965334008097166e-05,
+      "loss": 2.6255,
+      "step": 4363
+    },
+    {
+      "epoch": 1.104251012145749,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.965249662618084e-05,
+      "loss": 2.5388,
+      "step": 4364
+    },
+    {
+      "epoch": 1.104504048582996,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9651653171390013e-05,
+      "loss": 2.5436,
+      "step": 4365
+    },
+    {
+      "epoch": 1.1047570850202428,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.965080971659919e-05,
+      "loss": 2.5246,
+      "step": 4366
+    },
+    {
+      "epoch": 1.10501012145749,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.964996626180837e-05,
+      "loss": 2.5834,
+      "step": 4367
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9649122807017546e-05,
+      "loss": 2.5459,
+      "step": 4368
+    },
+    {
+      "epoch": 1.1055161943319838,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.964827935222672e-05,
+      "loss": 2.6023,
+      "step": 4369
+    },
+    {
+      "epoch": 1.1057692307692308,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.96474358974359e-05,
+      "loss": 2.5513,
+      "step": 4370
+    },
+    {
+      "epoch": 1.1060222672064777,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9646592442645076e-05,
+      "loss": 2.2214,
+      "step": 4371
+    },
+    {
+      "epoch": 1.1062753036437247,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.964574898785425e-05,
+      "loss": 2.5244,
+      "step": 4372
+    },
+    {
+      "epoch": 1.1065283400809716,
+      "grad_norm": 3.25,
+      "learning_rate": 2.964490553306343e-05,
+      "loss": 2.5858,
+      "step": 4373
+    },
+    {
+      "epoch": 1.1067813765182186,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9644062078272606e-05,
+      "loss": 2.5934,
+      "step": 4374
+    },
+    {
+      "epoch": 1.1070344129554657,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9643218623481783e-05,
+      "loss": 2.4975,
+      "step": 4375
+    },
+    {
+      "epoch": 1.1072874493927125,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.964237516869096e-05,
+      "loss": 2.4044,
+      "step": 4376
+    },
+    {
+      "epoch": 1.1075404858299596,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9641531713900136e-05,
+      "loss": 2.5952,
+      "step": 4377
+    },
+    {
+      "epoch": 1.1077935222672064,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9640688259109313e-05,
+      "loss": 2.6045,
+      "step": 4378
+    },
+    {
+      "epoch": 1.1080465587044535,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.963984480431849e-05,
+      "loss": 2.5658,
+      "step": 4379
+    },
+    {
+      "epoch": 1.1082995951417005,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.9639001349527666e-05,
+      "loss": 2.408,
+      "step": 4380
+    },
+    {
+      "epoch": 1.1082995951417005,
+      "eval_loss": 2.6503822803497314,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.9773,
+      "eval_samples_per_second": 9.49,
+      "eval_steps_per_second": 9.49,
+      "step": 4380
+    },
+    {
+      "epoch": 1.1085526315789473,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.9638157894736843e-05,
+      "loss": 2.4645,
+      "step": 4381
+    },
+    {
+      "epoch": 1.1088056680161944,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.963731443994602e-05,
+      "loss": 2.4695,
+      "step": 4382
+    },
+    {
+      "epoch": 1.1090587044534412,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9636470985155196e-05,
+      "loss": 2.5141,
+      "step": 4383
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.9635627530364373e-05,
+      "loss": 2.5019,
+      "step": 4384
+    },
+    {
+      "epoch": 1.1095647773279351,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.963478407557355e-05,
+      "loss": 2.6437,
+      "step": 4385
+    },
+    {
+      "epoch": 1.1098178137651822,
+      "grad_norm": 4.0,
+      "learning_rate": 2.963394062078273e-05,
+      "loss": 2.51,
+      "step": 4386
+    },
+    {
+      "epoch": 1.1100708502024292,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9633097165991903e-05,
+      "loss": 2.4374,
+      "step": 4387
+    },
+    {
+      "epoch": 1.110323886639676,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.963225371120108e-05,
+      "loss": 2.3046,
+      "step": 4388
+    },
+    {
+      "epoch": 1.1105769230769231,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.963141025641026e-05,
+      "loss": 2.5364,
+      "step": 4389
+    },
+    {
+      "epoch": 1.11082995951417,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.9630566801619436e-05,
+      "loss": 2.4972,
+      "step": 4390
+    },
+    {
+      "epoch": 1.111082995951417,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.962972334682861e-05,
+      "loss": 2.4303,
+      "step": 4391
+    },
+    {
+      "epoch": 1.1113360323886639,
+      "grad_norm": 3.125,
+      "learning_rate": 2.962887989203779e-05,
+      "loss": 2.5249,
+      "step": 4392
+    },
+    {
+      "epoch": 1.111589068825911,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9628036437246966e-05,
+      "loss": 2.5446,
+      "step": 4393
+    },
+    {
+      "epoch": 1.111842105263158,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.962719298245614e-05,
+      "loss": 2.5165,
+      "step": 4394
+    },
+    {
+      "epoch": 1.1120951417004048,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.9626349527665316e-05,
+      "loss": 2.5834,
+      "step": 4395
+    },
+    {
+      "epoch": 1.1123481781376519,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9625506072874496e-05,
+      "loss": 2.408,
+      "step": 4396
+    },
+    {
+      "epoch": 1.1126012145748987,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9624662618083672e-05,
+      "loss": 2.5624,
+      "step": 4397
+    },
+    {
+      "epoch": 1.1128542510121457,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9623819163292846e-05,
+      "loss": 2.4968,
+      "step": 4398
+    },
+    {
+      "epoch": 1.1131072874493928,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.9622975708502026e-05,
+      "loss": 2.498,
+      "step": 4399
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9622132253711202e-05,
+      "loss": 2.5622,
+      "step": 4400
+    },
+    {
+      "epoch": 1.1136133603238867,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9621288798920376e-05,
+      "loss": 2.5379,
+      "step": 4401
+    },
+    {
+      "epoch": 1.1138663967611335,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9620445344129556e-05,
+      "loss": 2.5008,
+      "step": 4402
+    },
+    {
+      "epoch": 1.1141194331983806,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9619601889338732e-05,
+      "loss": 2.478,
+      "step": 4403
+    },
+    {
+      "epoch": 1.1143724696356276,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.961875843454791e-05,
+      "loss": 2.3533,
+      "step": 4404
+    },
+    {
+      "epoch": 1.1146255060728745,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9617914979757086e-05,
+      "loss": 2.5956,
+      "step": 4405
+    },
+    {
+      "epoch": 1.1148785425101215,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9617071524966262e-05,
+      "loss": 2.4939,
+      "step": 4406
+    },
+    {
+      "epoch": 1.1151315789473684,
+      "grad_norm": 3.25,
+      "learning_rate": 2.961622807017544e-05,
+      "loss": 2.4995,
+      "step": 4407
+    },
+    {
+      "epoch": 1.1153846153846154,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.961538461538462e-05,
+      "loss": 2.5931,
+      "step": 4408
+    },
+    {
+      "epoch": 1.1156376518218623,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9614541160593792e-05,
+      "loss": 2.411,
+      "step": 4409
+    },
+    {
+      "epoch": 1.1158906882591093,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.961369770580297e-05,
+      "loss": 2.4764,
+      "step": 4410
+    },
+    {
+      "epoch": 1.1158906882591093,
+      "eval_loss": 2.6466033458709717,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.5147,
+      "eval_samples_per_second": 9.471,
+      "eval_steps_per_second": 9.471,
+      "step": 4410
+    },
+    {
+      "epoch": 1.1161437246963564,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.961285425101215e-05,
+      "loss": 2.5615,
+      "step": 4411
+    },
+    {
+      "epoch": 1.1163967611336032,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9612010796221322e-05,
+      "loss": 2.6456,
+      "step": 4412
+    },
+    {
+      "epoch": 1.1166497975708503,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.96111673414305e-05,
+      "loss": 2.5396,
+      "step": 4413
+    },
+    {
+      "epoch": 1.116902834008097,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9610323886639676e-05,
+      "loss": 2.541,
+      "step": 4414
+    },
+    {
+      "epoch": 1.1171558704453441,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9609480431848856e-05,
+      "loss": 2.5242,
+      "step": 4415
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.960863697705803e-05,
+      "loss": 2.5006,
+      "step": 4416
+    },
+    {
+      "epoch": 1.117661943319838,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9607793522267206e-05,
+      "loss": 2.4502,
+      "step": 4417
+    },
+    {
+      "epoch": 1.117914979757085,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9606950067476386e-05,
+      "loss": 2.571,
+      "step": 4418
+    },
+    {
+      "epoch": 1.118168016194332,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.9606106612685562e-05,
+      "loss": 2.6389,
+      "step": 4419
+    },
+    {
+      "epoch": 1.118421052631579,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9605263157894735e-05,
+      "loss": 2.5453,
+      "step": 4420
+    },
+    {
+      "epoch": 1.1186740890688258,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9604419703103916e-05,
+      "loss": 2.6179,
+      "step": 4421
+    },
+    {
+      "epoch": 1.1189271255060729,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9603576248313092e-05,
+      "loss": 2.5375,
+      "step": 4422
+    },
+    {
+      "epoch": 1.11918016194332,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9602732793522265e-05,
+      "loss": 2.4387,
+      "step": 4423
+    },
+    {
+      "epoch": 1.1194331983805668,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9601889338731445e-05,
+      "loss": 2.5432,
+      "step": 4424
+    },
+    {
+      "epoch": 1.1196862348178138,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9601045883940622e-05,
+      "loss": 2.539,
+      "step": 4425
+    },
+    {
+      "epoch": 1.1199392712550607,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.96002024291498e-05,
+      "loss": 2.6214,
+      "step": 4426
+    },
+    {
+      "epoch": 1.1201923076923077,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9599358974358975e-05,
+      "loss": 2.567,
+      "step": 4427
+    },
+    {
+      "epoch": 1.1204453441295548,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9598515519568152e-05,
+      "loss": 2.5005,
+      "step": 4428
+    },
+    {
+      "epoch": 1.1206983805668016,
+      "grad_norm": 2.875,
+      "learning_rate": 2.959767206477733e-05,
+      "loss": 2.5428,
+      "step": 4429
+    },
+    {
+      "epoch": 1.1209514170040487,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9596828609986505e-05,
+      "loss": 2.5278,
+      "step": 4430
+    },
+    {
+      "epoch": 1.1212044534412955,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9595985155195682e-05,
+      "loss": 2.6241,
+      "step": 4431
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.959514170040486e-05,
+      "loss": 2.4117,
+      "step": 4432
+    },
+    {
+      "epoch": 1.1217105263157894,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9594298245614035e-05,
+      "loss": 2.5231,
+      "step": 4433
+    },
+    {
+      "epoch": 1.1219635627530364,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9593454790823212e-05,
+      "loss": 2.5684,
+      "step": 4434
+    },
+    {
+      "epoch": 1.1222165991902835,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.959261133603239e-05,
+      "loss": 2.5561,
+      "step": 4435
+    },
+    {
+      "epoch": 1.1224696356275303,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9591767881241565e-05,
+      "loss": 2.5031,
+      "step": 4436
+    },
+    {
+      "epoch": 1.1227226720647774,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9590924426450745e-05,
+      "loss": 2.5152,
+      "step": 4437
+    },
+    {
+      "epoch": 1.1229757085020242,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.959008097165992e-05,
+      "loss": 2.5172,
+      "step": 4438
+    },
+    {
+      "epoch": 1.1232287449392713,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.9589237516869095e-05,
+      "loss": 2.5308,
+      "step": 4439
+    },
+    {
+      "epoch": 1.123481781376518,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9588394062078275e-05,
+      "loss": 2.5266,
+      "step": 4440
+    },
+    {
+      "epoch": 1.123481781376518,
+      "eval_loss": 2.6453845500946045,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.4065,
+      "eval_samples_per_second": 9.475,
+      "eval_steps_per_second": 9.475,
+      "step": 4440
+    },
+    {
+      "epoch": 1.1237348178137652,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.958755060728745e-05,
+      "loss": 2.5934,
+      "step": 4441
+    },
+    {
+      "epoch": 1.1239878542510122,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9586707152496625e-05,
+      "loss": 2.4858,
+      "step": 4442
+    },
+    {
+      "epoch": 1.124240890688259,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9585863697705805e-05,
+      "loss": 2.4222,
+      "step": 4443
+    },
+    {
+      "epoch": 1.124493927125506,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9585020242914982e-05,
+      "loss": 2.5265,
+      "step": 4444
+    },
+    {
+      "epoch": 1.124746963562753,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9584176788124155e-05,
+      "loss": 2.5291,
+      "step": 4445
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9583333333333335e-05,
+      "loss": 2.5261,
+      "step": 4446
+    },
+    {
+      "epoch": 1.125253036437247,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9582489878542512e-05,
+      "loss": 2.4551,
+      "step": 4447
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.958164642375169e-05,
+      "loss": 2.5719,
+      "step": 4448
+    },
+    {
+      "epoch": 1.125759109311741,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9580802968960865e-05,
+      "loss": 2.5648,
+      "step": 4449
+    },
+    {
+      "epoch": 1.1260121457489878,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9579959514170042e-05,
+      "loss": 2.5331,
+      "step": 4450
+    },
+    {
+      "epoch": 1.1262651821862348,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.957911605937922e-05,
+      "loss": 2.4182,
+      "step": 4451
+    },
+    {
+      "epoch": 1.126518218623482,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9578272604588392e-05,
+      "loss": 2.4915,
+      "step": 4452
+    },
+    {
+      "epoch": 1.1267712550607287,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9577429149797572e-05,
+      "loss": 2.6679,
+      "step": 4453
+    },
+    {
+      "epoch": 1.1270242914979758,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.957658569500675e-05,
+      "loss": 2.4469,
+      "step": 4454
+    },
+    {
+      "epoch": 1.1272773279352226,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.9575742240215925e-05,
+      "loss": 2.5748,
+      "step": 4455
+    },
+    {
+      "epoch": 1.1275303643724697,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9574898785425102e-05,
+      "loss": 2.5872,
+      "step": 4456
+    },
+    {
+      "epoch": 1.1277834008097165,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.957405533063428e-05,
+      "loss": 2.5514,
+      "step": 4457
+    },
+    {
+      "epoch": 1.1280364372469636,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9573211875843455e-05,
+      "loss": 2.6453,
+      "step": 4458
+    },
+    {
+      "epoch": 1.1282894736842106,
+      "grad_norm": 3.625,
+      "learning_rate": 2.9572368421052635e-05,
+      "loss": 2.4048,
+      "step": 4459
+    },
+    {
+      "epoch": 1.1285425101214575,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.957152496626181e-05,
+      "loss": 2.4182,
+      "step": 4460
+    },
+    {
+      "epoch": 1.1287955465587045,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9570681511470985e-05,
+      "loss": 2.5844,
+      "step": 4461
+    },
+    {
+      "epoch": 1.1290485829959513,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9569838056680165e-05,
+      "loss": 2.5115,
+      "step": 4462
+    },
+    {
+      "epoch": 1.1293016194331984,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.956899460188934e-05,
+      "loss": 2.564,
+      "step": 4463
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 3.890625,
+      "learning_rate": 2.9568151147098515e-05,
+      "loss": 2.5782,
+      "step": 4464
+    },
+    {
+      "epoch": 1.1298076923076923,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9567307692307695e-05,
+      "loss": 2.5399,
+      "step": 4465
+    },
+    {
+      "epoch": 1.1300607287449393,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9566464237516872e-05,
+      "loss": 2.4143,
+      "step": 4466
+    },
+    {
+      "epoch": 1.1303137651821862,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9565620782726045e-05,
+      "loss": 2.4198,
+      "step": 4467
+    },
+    {
+      "epoch": 1.1305668016194332,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.956477732793522e-05,
+      "loss": 2.4403,
+      "step": 4468
+    },
+    {
+      "epoch": 1.13081983805668,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.95639338731444e-05,
+      "loss": 2.5644,
+      "step": 4469
+    },
+    {
+      "epoch": 1.1310728744939271,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9563090418353575e-05,
+      "loss": 2.5166,
+      "step": 4470
+    },
+    {
+      "epoch": 1.1310728744939271,
+      "eval_loss": 2.6394100189208984,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.7614,
+      "eval_samples_per_second": 9.497,
+      "eval_steps_per_second": 9.497,
+      "step": 4470
+    },
+    {
+      "epoch": 1.1313259109311742,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.956224696356275e-05,
+      "loss": 2.5551,
+      "step": 4471
+    },
+    {
+      "epoch": 1.131578947368421,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.956140350877193e-05,
+      "loss": 2.5505,
+      "step": 4472
+    },
+    {
+      "epoch": 1.131831983805668,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.9560560053981108e-05,
+      "loss": 2.4277,
+      "step": 4473
+    },
+    {
+      "epoch": 1.132085020242915,
+      "grad_norm": 3.375,
+      "learning_rate": 2.955971659919028e-05,
+      "loss": 2.6127,
+      "step": 4474
+    },
+    {
+      "epoch": 1.132338056680162,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.955887314439946e-05,
+      "loss": 2.5104,
+      "step": 4475
+    },
+    {
+      "epoch": 1.132591093117409,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9558029689608638e-05,
+      "loss": 2.5188,
+      "step": 4476
+    },
+    {
+      "epoch": 1.1328441295546559,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9557186234817815e-05,
+      "loss": 2.6102,
+      "step": 4477
+    },
+    {
+      "epoch": 1.133097165991903,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.955634278002699e-05,
+      "loss": 2.4305,
+      "step": 4478
+    },
+    {
+      "epoch": 1.1333502024291497,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9555499325236168e-05,
+      "loss": 2.6092,
+      "step": 4479
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9554655870445345e-05,
+      "loss": 2.5019,
+      "step": 4480
+    },
+    {
+      "epoch": 1.1338562753036436,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.955381241565452e-05,
+      "loss": 2.4384,
+      "step": 4481
+    },
+    {
+      "epoch": 1.1341093117408907,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9552968960863698e-05,
+      "loss": 2.5528,
+      "step": 4482
+    },
+    {
+      "epoch": 1.1343623481781377,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9552125506072875e-05,
+      "loss": 2.473,
+      "step": 4483
+    },
+    {
+      "epoch": 1.1346153846153846,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9551282051282055e-05,
+      "loss": 2.5235,
+      "step": 4484
+    },
+    {
+      "epoch": 1.1348684210526316,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9550438596491228e-05,
+      "loss": 2.3471,
+      "step": 4485
+    },
+    {
+      "epoch": 1.1351214574898785,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9549595141700405e-05,
+      "loss": 2.5375,
+      "step": 4486
+    },
+    {
+      "epoch": 1.1353744939271255,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.954875168690958e-05,
+      "loss": 2.55,
+      "step": 4487
+    },
+    {
+      "epoch": 1.1356275303643724,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.954790823211876e-05,
+      "loss": 2.5335,
+      "step": 4488
+    },
+    {
+      "epoch": 1.1358805668016194,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9547064777327935e-05,
+      "loss": 2.4174,
+      "step": 4489
+    },
+    {
+      "epoch": 1.1361336032388665,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.954622132253711e-05,
+      "loss": 2.5029,
+      "step": 4490
+    },
+    {
+      "epoch": 1.1363866396761133,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.954537786774629e-05,
+      "loss": 2.5549,
+      "step": 4491
+    },
+    {
+      "epoch": 1.1366396761133604,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9544534412955465e-05,
+      "loss": 2.4796,
+      "step": 4492
+    },
+    {
+      "epoch": 1.1368927125506072,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.954369095816464e-05,
+      "loss": 2.4249,
+      "step": 4493
+    },
+    {
+      "epoch": 1.1371457489878543,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.954284750337382e-05,
+      "loss": 2.4663,
+      "step": 4494
+    },
+    {
+      "epoch": 1.1373987854251013,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9542004048582998e-05,
+      "loss": 2.564,
+      "step": 4495
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.954116059379217e-05,
+      "loss": 2.5826,
+      "step": 4496
+    },
+    {
+      "epoch": 1.1379048582995952,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.954031713900135e-05,
+      "loss": 2.5924,
+      "step": 4497
+    },
+    {
+      "epoch": 1.138157894736842,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9539473684210528e-05,
+      "loss": 2.5291,
+      "step": 4498
+    },
+    {
+      "epoch": 1.138410931174089,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9538630229419705e-05,
+      "loss": 2.6304,
+      "step": 4499
+    },
+    {
+      "epoch": 1.1386639676113361,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.953778677462888e-05,
+      "loss": 2.4944,
+      "step": 4500
+    },
+    {
+      "epoch": 1.1386639676113361,
+      "eval_loss": 2.6451447010040283,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.1128,
+      "eval_samples_per_second": 9.52,
+      "eval_steps_per_second": 9.52,
+      "step": 4500
+    },
+    {
+      "epoch": 1.138917004048583,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.9536943319838058e-05,
+      "loss": 2.5205,
+      "step": 4501
+    },
+    {
+      "epoch": 1.13917004048583,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9536099865047235e-05,
+      "loss": 2.5359,
+      "step": 4502
+    },
+    {
+      "epoch": 1.1394230769230769,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.953525641025641e-05,
+      "loss": 2.5735,
+      "step": 4503
+    },
+    {
+      "epoch": 1.139676113360324,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.9534412955465588e-05,
+      "loss": 2.676,
+      "step": 4504
+    },
+    {
+      "epoch": 1.1399291497975708,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9533569500674765e-05,
+      "loss": 2.5444,
+      "step": 4505
+    },
+    {
+      "epoch": 1.1401821862348178,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.953272604588394e-05,
+      "loss": 2.5236,
+      "step": 4506
+    },
+    {
+      "epoch": 1.1404352226720649,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9531882591093118e-05,
+      "loss": 2.5345,
+      "step": 4507
+    },
+    {
+      "epoch": 1.1406882591093117,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9531039136302295e-05,
+      "loss": 2.5178,
+      "step": 4508
+    },
+    {
+      "epoch": 1.1409412955465588,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.953019568151147e-05,
+      "loss": 2.4517,
+      "step": 4509
+    },
+    {
+      "epoch": 1.1411943319838056,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9529352226720648e-05,
+      "loss": 2.5218,
+      "step": 4510
+    },
+    {
+      "epoch": 1.1414473684210527,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9528508771929825e-05,
+      "loss": 2.5612,
+      "step": 4511
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9527665317139e-05,
+      "loss": 2.5906,
+      "step": 4512
+    },
+    {
+      "epoch": 1.1419534412955465,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.952682186234818e-05,
+      "loss": 2.3618,
+      "step": 4513
+    },
+    {
+      "epoch": 1.1422064777327936,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9525978407557354e-05,
+      "loss": 2.3431,
+      "step": 4514
+    },
+    {
+      "epoch": 1.1424595141700404,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.952513495276653e-05,
+      "loss": 2.5418,
+      "step": 4515
+    },
+    {
+      "epoch": 1.1427125506072875,
+      "grad_norm": 3.125,
+      "learning_rate": 2.952429149797571e-05,
+      "loss": 2.6255,
+      "step": 4516
+    },
+    {
+      "epoch": 1.1429655870445343,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9523448043184888e-05,
+      "loss": 2.4406,
+      "step": 4517
+    },
+    {
+      "epoch": 1.1432186234817814,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.952260458839406e-05,
+      "loss": 2.248,
+      "step": 4518
+    },
+    {
+      "epoch": 1.1434716599190282,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.952176113360324e-05,
+      "loss": 2.5025,
+      "step": 4519
+    },
+    {
+      "epoch": 1.1437246963562753,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9520917678812418e-05,
+      "loss": 2.5326,
+      "step": 4520
+    },
+    {
+      "epoch": 1.1439777327935223,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.952007422402159e-05,
+      "loss": 2.5095,
+      "step": 4521
+    },
+    {
+      "epoch": 1.1442307692307692,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9519230769230768e-05,
+      "loss": 2.483,
+      "step": 4522
+    },
+    {
+      "epoch": 1.1444838056680162,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9518387314439948e-05,
+      "loss": 2.5503,
+      "step": 4523
+    },
+    {
+      "epoch": 1.1447368421052633,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9517543859649124e-05,
+      "loss": 2.4551,
+      "step": 4524
+    },
+    {
+      "epoch": 1.14498987854251,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9516700404858298e-05,
+      "loss": 2.4839,
+      "step": 4525
+    },
+    {
+      "epoch": 1.1452429149797572,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9515856950067478e-05,
+      "loss": 2.534,
+      "step": 4526
+    },
+    {
+      "epoch": 1.145495951417004,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9515013495276654e-05,
+      "loss": 2.4912,
+      "step": 4527
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.951417004048583e-05,
+      "loss": 2.5049,
+      "step": 4528
+    },
+    {
+      "epoch": 1.146002024291498,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9513326585695008e-05,
+      "loss": 2.67,
+      "step": 4529
+    },
+    {
+      "epoch": 1.146255060728745,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9512483130904184e-05,
+      "loss": 2.3003,
+      "step": 4530
+    },
+    {
+      "epoch": 1.146255060728745,
+      "eval_loss": 2.640401840209961,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.8955,
+      "eval_samples_per_second": 9.493,
+      "eval_steps_per_second": 9.493,
+      "step": 4530
+    },
+    {
+      "epoch": 1.146508097165992,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.951163967611336e-05,
+      "loss": 2.5238,
+      "step": 4531
+    },
+    {
+      "epoch": 1.1467611336032388,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9510796221322538e-05,
+      "loss": 2.4143,
+      "step": 4532
+    },
+    {
+      "epoch": 1.147014170040486,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9509952766531714e-05,
+      "loss": 2.6004,
+      "step": 4533
+    },
+    {
+      "epoch": 1.1472672064777327,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.950910931174089e-05,
+      "loss": 2.5183,
+      "step": 4534
+    },
+    {
+      "epoch": 1.1475202429149798,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.950826585695007e-05,
+      "loss": 2.5341,
+      "step": 4535
+    },
+    {
+      "epoch": 1.1477732793522266,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9507422402159244e-05,
+      "loss": 2.5822,
+      "step": 4536
+    },
+    {
+      "epoch": 1.1480263157894737,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.950657894736842e-05,
+      "loss": 2.5546,
+      "step": 4537
+    },
+    {
+      "epoch": 1.1482793522267207,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.95057354925776e-05,
+      "loss": 2.6188,
+      "step": 4538
+    },
+    {
+      "epoch": 1.1485323886639676,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9504892037786778e-05,
+      "loss": 2.6145,
+      "step": 4539
+    },
+    {
+      "epoch": 1.1487854251012146,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.950404858299595e-05,
+      "loss": 2.4295,
+      "step": 4540
+    },
+    {
+      "epoch": 1.1490384615384615,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9503205128205127e-05,
+      "loss": 2.6332,
+      "step": 4541
+    },
+    {
+      "epoch": 1.1492914979757085,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.9502361673414308e-05,
+      "loss": 2.4533,
+      "step": 4542
+    },
+    {
+      "epoch": 1.1495445344129553,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.950151821862348e-05,
+      "loss": 2.5541,
+      "step": 4543
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9500674763832657e-05,
+      "loss": 2.4472,
+      "step": 4544
+    },
+    {
+      "epoch": 1.1500506072874495,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9499831309041837e-05,
+      "loss": 2.527,
+      "step": 4545
+    },
+    {
+      "epoch": 1.1503036437246963,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.9498987854251014e-05,
+      "loss": 2.4155,
+      "step": 4546
+    },
+    {
+      "epoch": 1.1505566801619433,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.9498144399460187e-05,
+      "loss": 2.4656,
+      "step": 4547
+    },
+    {
+      "epoch": 1.1508097165991904,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9497300944669367e-05,
+      "loss": 2.5103,
+      "step": 4548
+    },
+    {
+      "epoch": 1.1510627530364372,
+      "grad_norm": 2.75,
+      "learning_rate": 2.9496457489878544e-05,
+      "loss": 2.4416,
+      "step": 4549
+    },
+    {
+      "epoch": 1.1513157894736843,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9495614035087717e-05,
+      "loss": 2.5476,
+      "step": 4550
+    },
+    {
+      "epoch": 1.1515688259109311,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9494770580296897e-05,
+      "loss": 2.475,
+      "step": 4551
+    },
+    {
+      "epoch": 1.1518218623481782,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9493927125506074e-05,
+      "loss": 2.6356,
+      "step": 4552
+    },
+    {
+      "epoch": 1.152074898785425,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.949308367071525e-05,
+      "loss": 2.4209,
+      "step": 4553
+    },
+    {
+      "epoch": 1.152327935222672,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9492240215924427e-05,
+      "loss": 2.5741,
+      "step": 4554
+    },
+    {
+      "epoch": 1.1525809716599191,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9491396761133604e-05,
+      "loss": 2.4859,
+      "step": 4555
+    },
+    {
+      "epoch": 1.152834008097166,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.949055330634278e-05,
+      "loss": 2.6011,
+      "step": 4556
+    },
+    {
+      "epoch": 1.153087044534413,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.948970985155196e-05,
+      "loss": 2.546,
+      "step": 4557
+    },
+    {
+      "epoch": 1.1533400809716599,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9488866396761134e-05,
+      "loss": 2.578,
+      "step": 4558
+    },
+    {
+      "epoch": 1.153593117408907,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.948802294197031e-05,
+      "loss": 2.5861,
+      "step": 4559
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9487179487179487e-05,
+      "loss": 2.5395,
+      "step": 4560
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "eval_loss": 2.636711359024048,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.1875,
+      "eval_samples_per_second": 9.517,
+      "eval_steps_per_second": 9.517,
+      "step": 4560
+    },
+    {
+      "epoch": 1.1540991902834008,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9486336032388664e-05,
+      "loss": 2.5658,
+      "step": 4561
+    },
+    {
+      "epoch": 1.1543522267206479,
+      "grad_norm": 3.25,
+      "learning_rate": 2.948549257759784e-05,
+      "loss": 2.4944,
+      "step": 4562
+    },
+    {
+      "epoch": 1.1546052631578947,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9484649122807017e-05,
+      "loss": 2.505,
+      "step": 4563
+    },
+    {
+      "epoch": 1.1548582995951417,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9483805668016197e-05,
+      "loss": 2.5184,
+      "step": 4564
+    },
+    {
+      "epoch": 1.1551113360323886,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.948296221322537e-05,
+      "loss": 2.5363,
+      "step": 4565
+    },
+    {
+      "epoch": 1.1553643724696356,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9482118758434547e-05,
+      "loss": 2.552,
+      "step": 4566
+    },
+    {
+      "epoch": 1.1556174089068825,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9481275303643727e-05,
+      "loss": 2.6467,
+      "step": 4567
+    },
+    {
+      "epoch": 1.1558704453441295,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9480431848852904e-05,
+      "loss": 2.4808,
+      "step": 4568
+    },
+    {
+      "epoch": 1.1561234817813766,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.9479588394062077e-05,
+      "loss": 2.4397,
+      "step": 4569
+    },
+    {
+      "epoch": 1.1563765182186234,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9478744939271257e-05,
+      "loss": 2.5965,
+      "step": 4570
+    },
+    {
+      "epoch": 1.1566295546558705,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9477901484480434e-05,
+      "loss": 2.5636,
+      "step": 4571
+    },
+    {
+      "epoch": 1.1568825910931175,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9477058029689607e-05,
+      "loss": 2.4536,
+      "step": 4572
+    },
+    {
+      "epoch": 1.1571356275303644,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9476214574898787e-05,
+      "loss": 2.4692,
+      "step": 4573
+    },
+    {
+      "epoch": 1.1573886639676114,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9475371120107964e-05,
+      "loss": 2.5399,
+      "step": 4574
+    },
+    {
+      "epoch": 1.1576417004048583,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.947452766531714e-05,
+      "loss": 2.6286,
+      "step": 4575
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9473684210526314e-05,
+      "loss": 2.4543,
+      "step": 4576
+    },
+    {
+      "epoch": 1.1581477732793521,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9472840755735494e-05,
+      "loss": 2.5254,
+      "step": 4577
+    },
+    {
+      "epoch": 1.1584008097165992,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.947199730094467e-05,
+      "loss": 2.4881,
+      "step": 4578
+    },
+    {
+      "epoch": 1.1586538461538463,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9471153846153847e-05,
+      "loss": 2.4328,
+      "step": 4579
+    },
+    {
+      "epoch": 1.158906882591093,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9470310391363024e-05,
+      "loss": 2.5573,
+      "step": 4580
+    },
+    {
+      "epoch": 1.1591599190283401,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.94694669365722e-05,
+      "loss": 2.5725,
+      "step": 4581
+    },
+    {
+      "epoch": 1.159412955465587,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9468623481781377e-05,
+      "loss": 2.5146,
+      "step": 4582
+    },
+    {
+      "epoch": 1.159665991902834,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9467780026990554e-05,
+      "loss": 2.5771,
+      "step": 4583
+    },
+    {
+      "epoch": 1.1599190283400809,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.946693657219973e-05,
+      "loss": 2.5635,
+      "step": 4584
+    },
+    {
+      "epoch": 1.160172064777328,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9466093117408907e-05,
+      "loss": 2.5462,
+      "step": 4585
+    },
+    {
+      "epoch": 1.160425101214575,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9465249662618087e-05,
+      "loss": 2.5683,
+      "step": 4586
+    },
+    {
+      "epoch": 1.1606781376518218,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.946440620782726e-05,
+      "loss": 2.5799,
+      "step": 4587
+    },
+    {
+      "epoch": 1.1609311740890689,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9463562753036437e-05,
+      "loss": 2.5196,
+      "step": 4588
+    },
+    {
+      "epoch": 1.1611842105263157,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9462719298245617e-05,
+      "loss": 2.5205,
+      "step": 4589
+    },
+    {
+      "epoch": 1.1614372469635628,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.946187584345479e-05,
+      "loss": 2.4195,
+      "step": 4590
+    },
+    {
+      "epoch": 1.1614372469635628,
+      "eval_loss": 2.6394684314727783,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 272.843,
+      "eval_samples_per_second": 9.529,
+      "eval_steps_per_second": 9.529,
+      "step": 4590
+    },
+    {
+      "epoch": 1.1616902834008096,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9461032388663967e-05,
+      "loss": 2.3849,
+      "step": 4591
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9460188933873147e-05,
+      "loss": 2.5463,
+      "step": 4592
+    },
+    {
+      "epoch": 1.1621963562753037,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9459345479082324e-05,
+      "loss": 2.5343,
+      "step": 4593
+    },
+    {
+      "epoch": 1.1624493927125505,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9458502024291497e-05,
+      "loss": 2.5098,
+      "step": 4594
+    },
+    {
+      "epoch": 1.1627024291497976,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9457658569500674e-05,
+      "loss": 2.5395,
+      "step": 4595
+    },
+    {
+      "epoch": 1.1629554655870447,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9456815114709854e-05,
+      "loss": 2.6133,
+      "step": 4596
+    },
+    {
+      "epoch": 1.1632085020242915,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.945597165991903e-05,
+      "loss": 2.4549,
+      "step": 4597
+    },
+    {
+      "epoch": 1.1634615384615385,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9455128205128204e-05,
+      "loss": 2.4323,
+      "step": 4598
+    },
+    {
+      "epoch": 1.1637145748987854,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9454284750337384e-05,
+      "loss": 2.4223,
+      "step": 4599
+    },
+    {
+      "epoch": 1.1639676113360324,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.945344129554656e-05,
+      "loss": 2.5664,
+      "step": 4600
+    },
+    {
+      "epoch": 1.1642206477732793,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9452597840755733e-05,
+      "loss": 2.5332,
+      "step": 4601
+    },
+    {
+      "epoch": 1.1644736842105263,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9451754385964914e-05,
+      "loss": 2.4117,
+      "step": 4602
+    },
+    {
+      "epoch": 1.1647267206477734,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.945091093117409e-05,
+      "loss": 2.4223,
+      "step": 4603
+    },
+    {
+      "epoch": 1.1649797570850202,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.9450067476383267e-05,
+      "loss": 2.3429,
+      "step": 4604
+    },
+    {
+      "epoch": 1.1652327935222673,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9449224021592443e-05,
+      "loss": 2.6575,
+      "step": 4605
+    },
+    {
+      "epoch": 1.165485829959514,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.944838056680162e-05,
+      "loss": 2.4837,
+      "step": 4606
+    },
+    {
+      "epoch": 1.1657388663967612,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9447537112010797e-05,
+      "loss": 2.4611,
+      "step": 4607
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9446693657219977e-05,
+      "loss": 2.5308,
+      "step": 4608
+    },
+    {
+      "epoch": 1.166244939271255,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.944585020242915e-05,
+      "loss": 2.4899,
+      "step": 4609
+    },
+    {
+      "epoch": 1.166497975708502,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.9445006747638327e-05,
+      "loss": 2.5054,
+      "step": 4610
+    },
+    {
+      "epoch": 1.166751012145749,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9444163292847507e-05,
+      "loss": 2.504,
+      "step": 4611
+    },
+    {
+      "epoch": 1.167004048582996,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.944331983805668e-05,
+      "loss": 2.562,
+      "step": 4612
+    },
+    {
+      "epoch": 1.1672570850202428,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9442476383265857e-05,
+      "loss": 2.4252,
+      "step": 4613
+    },
+    {
+      "epoch": 1.16751012145749,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9441632928475033e-05,
+      "loss": 2.5482,
+      "step": 4614
+    },
+    {
+      "epoch": 1.1677631578947367,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9440789473684213e-05,
+      "loss": 2.4086,
+      "step": 4615
+    },
+    {
+      "epoch": 1.1680161943319838,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9439946018893387e-05,
+      "loss": 2.5251,
+      "step": 4616
+    },
+    {
+      "epoch": 1.1682692307692308,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9439102564102563e-05,
+      "loss": 2.5436,
+      "step": 4617
+    },
+    {
+      "epoch": 1.1685222672064777,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9438259109311743e-05,
+      "loss": 2.2951,
+      "step": 4618
+    },
+    {
+      "epoch": 1.1687753036437247,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.943741565452092e-05,
+      "loss": 2.537,
+      "step": 4619
+    },
+    {
+      "epoch": 1.1690283400809718,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9436572199730093e-05,
+      "loss": 2.6035,
+      "step": 4620
+    },
+    {
+      "epoch": 1.1690283400809718,
+      "eval_loss": 2.635045051574707,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 273.1221,
+      "eval_samples_per_second": 9.52,
+      "eval_steps_per_second": 9.52,
+      "step": 4620
+    },
+    {
+      "epoch": 1.1692813765182186,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9435728744939273e-05,
+      "loss": 2.48,
+      "step": 4621
+    },
+    {
+      "epoch": 1.1695344129554657,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.943488529014845e-05,
+      "loss": 2.4594,
+      "step": 4622
+    },
+    {
+      "epoch": 1.1697874493927125,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9434041835357623e-05,
+      "loss": 2.4761,
+      "step": 4623
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9433198380566803e-05,
+      "loss": 2.5857,
+      "step": 4624
+    },
+    {
+      "epoch": 1.1702935222672064,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.943235492577598e-05,
+      "loss": 2.4704,
+      "step": 4625
+    },
+    {
+      "epoch": 1.1705465587044535,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9431511470985157e-05,
+      "loss": 2.4594,
+      "step": 4626
+    },
+    {
+      "epoch": 1.1707995951417005,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9430668016194333e-05,
+      "loss": 2.5909,
+      "step": 4627
+    },
+    {
+      "epoch": 1.1710526315789473,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.942982456140351e-05,
+      "loss": 2.5964,
+      "step": 4628
+    },
+    {
+      "epoch": 1.1713056680161944,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9428981106612687e-05,
+      "loss": 2.5466,
+      "step": 4629
+    },
+    {
+      "epoch": 1.1715587044534412,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.942813765182186e-05,
+      "loss": 2.5965,
+      "step": 4630
+    },
+    {
+      "epoch": 1.1718117408906883,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.942729419703104e-05,
+      "loss": 2.5725,
+      "step": 4631
+    },
+    {
+      "epoch": 1.1720647773279351,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9426450742240216e-05,
+      "loss": 2.5311,
+      "step": 4632
+    },
+    {
+      "epoch": 1.1723178137651822,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9425607287449393e-05,
+      "loss": 2.5318,
+      "step": 4633
+    },
+    {
+      "epoch": 1.1725708502024292,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.942476383265857e-05,
+      "loss": 2.5385,
+      "step": 4634
+    },
+    {
+      "epoch": 1.172823886639676,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9423920377867746e-05,
+      "loss": 2.5784,
+      "step": 4635
+    },
+    {
+      "epoch": 1.1730769230769231,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.9423076923076923e-05,
+      "loss": 2.4903,
+      "step": 4636
+    },
+    {
+      "epoch": 1.17332995951417,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9422233468286103e-05,
+      "loss": 2.4588,
+      "step": 4637
+    },
+    {
+      "epoch": 1.173582995951417,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9421390013495276e-05,
+      "loss": 2.5875,
+      "step": 4638
+    },
+    {
+      "epoch": 1.1738360323886639,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9420546558704453e-05,
+      "loss": 2.542,
+      "step": 4639
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9419703103913633e-05,
+      "loss": 2.4703,
+      "step": 4640
+    },
+    {
+      "epoch": 1.174342105263158,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9418859649122806e-05,
+      "loss": 2.5105,
+      "step": 4641
+    },
+    {
+      "epoch": 1.1745951417004048,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9418016194331983e-05,
+      "loss": 2.5791,
+      "step": 4642
+    },
+    {
+      "epoch": 1.1748481781376519,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9417172739541163e-05,
+      "loss": 2.5078,
+      "step": 4643
+    },
+    {
+      "epoch": 1.1751012145748987,
+      "grad_norm": 3.25,
+      "learning_rate": 2.941632928475034e-05,
+      "loss": 2.5191,
+      "step": 4644
+    },
+    {
+      "epoch": 1.1753542510121457,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.9415485829959513e-05,
+      "loss": 2.4219,
+      "step": 4645
+    },
+    {
+      "epoch": 1.1756072874493928,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.9414642375168693e-05,
+      "loss": 2.5068,
+      "step": 4646
+    },
+    {
+      "epoch": 1.1758603238866396,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.941379892037787e-05,
+      "loss": 2.504,
+      "step": 4647
+    },
+    {
+      "epoch": 1.1761133603238867,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9412955465587046e-05,
+      "loss": 2.46,
+      "step": 4648
+    },
+    {
+      "epoch": 1.1763663967611335,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.941211201079622e-05,
+      "loss": 2.5361,
+      "step": 4649
+    },
+    {
+      "epoch": 1.1766194331983806,
+      "grad_norm": 3.0,
+      "learning_rate": 2.94112685560054e-05,
+      "loss": 2.5188,
+      "step": 4650
+    },
+    {
+      "epoch": 1.1766194331983806,
+      "eval_loss": 2.6337063312530518,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.8113,
+      "eval_samples_per_second": 9.461,
+      "eval_steps_per_second": 9.461,
+      "step": 4650
+    },
+    {
+      "epoch": 1.1768724696356276,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9410425101214576e-05,
+      "loss": 2.5212,
+      "step": 4651
+    },
+    {
+      "epoch": 1.1771255060728745,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.940958164642375e-05,
+      "loss": 2.4507,
+      "step": 4652
+    },
+    {
+      "epoch": 1.1773785425101215,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.940873819163293e-05,
+      "loss": 2.4589,
+      "step": 4653
+    },
+    {
+      "epoch": 1.1776315789473684,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.9407894736842106e-05,
+      "loss": 2.6023,
+      "step": 4654
+    },
+    {
+      "epoch": 1.1778846153846154,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9407051282051283e-05,
+      "loss": 2.5048,
+      "step": 4655
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.940620782726046e-05,
+      "loss": 2.5618,
+      "step": 4656
+    },
+    {
+      "epoch": 1.1783906882591093,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9405364372469636e-05,
+      "loss": 2.5021,
+      "step": 4657
+    },
+    {
+      "epoch": 1.1786437246963564,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9404520917678813e-05,
+      "loss": 2.2845,
+      "step": 4658
+    },
+    {
+      "epoch": 1.1788967611336032,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.940367746288799e-05,
+      "loss": 2.5132,
+      "step": 4659
+    },
+    {
+      "epoch": 1.1791497975708503,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9402834008097166e-05,
+      "loss": 2.4496,
+      "step": 4660
+    },
+    {
+      "epoch": 1.179402834008097,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.9401990553306343e-05,
+      "loss": 2.3573,
+      "step": 4661
+    },
+    {
+      "epoch": 1.1796558704453441,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9401147098515523e-05,
+      "loss": 2.604,
+      "step": 4662
+    },
+    {
+      "epoch": 1.179908906882591,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9400303643724696e-05,
+      "loss": 2.5534,
+      "step": 4663
+    },
+    {
+      "epoch": 1.180161943319838,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9399460188933873e-05,
+      "loss": 2.4836,
+      "step": 4664
+    },
+    {
+      "epoch": 1.180414979757085,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9398616734143053e-05,
+      "loss": 2.5233,
+      "step": 4665
+    },
+    {
+      "epoch": 1.180668016194332,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.939777327935223e-05,
+      "loss": 2.5989,
+      "step": 4666
+    },
+    {
+      "epoch": 1.180921052631579,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9396929824561403e-05,
+      "loss": 2.6706,
+      "step": 4667
+    },
+    {
+      "epoch": 1.1811740890688258,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.939608636977058e-05,
+      "loss": 2.4549,
+      "step": 4668
+    },
+    {
+      "epoch": 1.1814271255060729,
+      "grad_norm": 3.125,
+      "learning_rate": 2.939524291497976e-05,
+      "loss": 2.554,
+      "step": 4669
+    },
+    {
+      "epoch": 1.18168016194332,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9394399460188933e-05,
+      "loss": 2.5173,
+      "step": 4670
+    },
+    {
+      "epoch": 1.1819331983805668,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.939355600539811e-05,
+      "loss": 2.6119,
+      "step": 4671
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.939271255060729e-05,
+      "loss": 2.6175,
+      "step": 4672
+    },
+    {
+      "epoch": 1.1824392712550607,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9391869095816466e-05,
+      "loss": 2.7129,
+      "step": 4673
+    },
+    {
+      "epoch": 1.1826923076923077,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.939102564102564e-05,
+      "loss": 2.542,
+      "step": 4674
+    },
+    {
+      "epoch": 1.1829453441295548,
+      "grad_norm": 3.375,
+      "learning_rate": 2.939018218623482e-05,
+      "loss": 2.4945,
+      "step": 4675
+    },
+    {
+      "epoch": 1.1831983805668016,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9389338731443996e-05,
+      "loss": 2.5687,
+      "step": 4676
+    },
+    {
+      "epoch": 1.1834514170040487,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9388495276653173e-05,
+      "loss": 2.439,
+      "step": 4677
+    },
+    {
+      "epoch": 1.1837044534412955,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.938765182186235e-05,
+      "loss": 2.5121,
+      "step": 4678
+    },
+    {
+      "epoch": 1.1839574898785425,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.9386808367071526e-05,
+      "loss": 2.6099,
+      "step": 4679
+    },
+    {
+      "epoch": 1.1842105263157894,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9385964912280703e-05,
+      "loss": 2.4131,
+      "step": 4680
+    },
+    {
+      "epoch": 1.1842105263157894,
+      "eval_loss": 2.6312525272369385,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.5913,
+      "eval_samples_per_second": 9.434,
+      "eval_steps_per_second": 9.434,
+      "step": 4680
+    },
+    {
+      "epoch": 1.1844635627530364,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.938512145748988e-05,
+      "loss": 2.4543,
+      "step": 4681
+    },
+    {
+      "epoch": 1.1847165991902835,
+      "grad_norm": 2.625,
+      "learning_rate": 2.9384278002699056e-05,
+      "loss": 2.4684,
+      "step": 4682
+    },
+    {
+      "epoch": 1.1849696356275303,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9383434547908233e-05,
+      "loss": 2.4078,
+      "step": 4683
+    },
+    {
+      "epoch": 1.1852226720647774,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.938259109311741e-05,
+      "loss": 2.3823,
+      "step": 4684
+    },
+    {
+      "epoch": 1.1854757085020242,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9381747638326586e-05,
+      "loss": 2.5756,
+      "step": 4685
+    },
+    {
+      "epoch": 1.1857287449392713,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9380904183535763e-05,
+      "loss": 2.4975,
+      "step": 4686
+    },
+    {
+      "epoch": 1.185981781376518,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.938006072874494e-05,
+      "loss": 2.5194,
+      "step": 4687
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.937921727395412e-05,
+      "loss": 2.4922,
+      "step": 4688
+    },
+    {
+      "epoch": 1.1864878542510122,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9378373819163293e-05,
+      "loss": 2.435,
+      "step": 4689
+    },
+    {
+      "epoch": 1.186740890688259,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.937753036437247e-05,
+      "loss": 2.477,
+      "step": 4690
+    },
+    {
+      "epoch": 1.186993927125506,
+      "grad_norm": 3.125,
+      "learning_rate": 2.937668690958165e-05,
+      "loss": 2.4847,
+      "step": 4691
+    },
+    {
+      "epoch": 1.187246963562753,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9375843454790822e-05,
+      "loss": 2.5542,
+      "step": 4692
+    },
+    {
+      "epoch": 1.1875,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9375e-05,
+      "loss": 2.5112,
+      "step": 4693
+    },
+    {
+      "epoch": 1.187753036437247,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.937415654520918e-05,
+      "loss": 2.5594,
+      "step": 4694
+    },
+    {
+      "epoch": 1.188006072874494,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9373313090418356e-05,
+      "loss": 2.6269,
+      "step": 4695
+    },
+    {
+      "epoch": 1.188259109311741,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.937246963562753e-05,
+      "loss": 2.4627,
+      "step": 4696
+    },
+    {
+      "epoch": 1.1885121457489878,
+      "grad_norm": 3.125,
+      "learning_rate": 2.937162618083671e-05,
+      "loss": 2.5548,
+      "step": 4697
+    },
+    {
+      "epoch": 1.1887651821862348,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9370782726045886e-05,
+      "loss": 2.5505,
+      "step": 4698
+    },
+    {
+      "epoch": 1.189018218623482,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.936993927125506e-05,
+      "loss": 2.5508,
+      "step": 4699
+    },
+    {
+      "epoch": 1.1892712550607287,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.936909581646424e-05,
+      "loss": 2.5443,
+      "step": 4700
+    },
+    {
+      "epoch": 1.1895242914979758,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9368252361673416e-05,
+      "loss": 2.5023,
+      "step": 4701
+    },
+    {
+      "epoch": 1.1897773279352226,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9367408906882592e-05,
+      "loss": 2.4273,
+      "step": 4702
+    },
+    {
+      "epoch": 1.1900303643724697,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9366565452091766e-05,
+      "loss": 2.4398,
+      "step": 4703
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9365721997300946e-05,
+      "loss": 2.4159,
+      "step": 4704
+    },
+    {
+      "epoch": 1.1905364372469636,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9364878542510122e-05,
+      "loss": 2.6037,
+      "step": 4705
+    },
+    {
+      "epoch": 1.1907894736842106,
+      "grad_norm": 2.875,
+      "learning_rate": 2.93640350877193e-05,
+      "loss": 2.5363,
+      "step": 4706
+    },
+    {
+      "epoch": 1.1910425101214575,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9363191632928476e-05,
+      "loss": 2.4333,
+      "step": 4707
+    },
+    {
+      "epoch": 1.1912955465587045,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9362348178137652e-05,
+      "loss": 2.5114,
+      "step": 4708
+    },
+    {
+      "epoch": 1.1915485829959513,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.936150472334683e-05,
+      "loss": 2.3945,
+      "step": 4709
+    },
+    {
+      "epoch": 1.1918016194331984,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9360661268556006e-05,
+      "loss": 2.5151,
+      "step": 4710
+    },
+    {
+      "epoch": 1.1918016194331984,
+      "eval_loss": 2.6289331912994385,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.923,
+      "eval_samples_per_second": 9.457,
+      "eval_steps_per_second": 9.457,
+      "step": 4710
+    },
+    {
+      "epoch": 1.1920546558704452,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9359817813765182e-05,
+      "loss": 2.5644,
+      "step": 4711
+    },
+    {
+      "epoch": 1.1923076923076923,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.935897435897436e-05,
+      "loss": 2.5758,
+      "step": 4712
+    },
+    {
+      "epoch": 1.1925607287449393,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.935813090418354e-05,
+      "loss": 2.5207,
+      "step": 4713
+    },
+    {
+      "epoch": 1.1928137651821862,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9357287449392712e-05,
+      "loss": 2.3805,
+      "step": 4714
+    },
+    {
+      "epoch": 1.1930668016194332,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.935644399460189e-05,
+      "loss": 2.5483,
+      "step": 4715
+    },
+    {
+      "epoch": 1.19331983805668,
+      "grad_norm": 3.25,
+      "learning_rate": 2.935560053981107e-05,
+      "loss": 2.5577,
+      "step": 4716
+    },
+    {
+      "epoch": 1.1935728744939271,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9354757085020246e-05,
+      "loss": 2.5779,
+      "step": 4717
+    },
+    {
+      "epoch": 1.1938259109311742,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.935391363022942e-05,
+      "loss": 2.5034,
+      "step": 4718
+    },
+    {
+      "epoch": 1.194078947368421,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.93530701754386e-05,
+      "loss": 2.4866,
+      "step": 4719
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9352226720647776e-05,
+      "loss": 2.3953,
+      "step": 4720
+    },
+    {
+      "epoch": 1.194585020242915,
+      "grad_norm": 3.125,
+      "learning_rate": 2.935138326585695e-05,
+      "loss": 2.5263,
+      "step": 4721
+    },
+    {
+      "epoch": 1.194838056680162,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9350539811066125e-05,
+      "loss": 2.4901,
+      "step": 4722
+    },
+    {
+      "epoch": 1.195091093117409,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9349696356275305e-05,
+      "loss": 2.5012,
+      "step": 4723
+    },
+    {
+      "epoch": 1.1953441295546559,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.9348852901484482e-05,
+      "loss": 2.3702,
+      "step": 4724
+    },
+    {
+      "epoch": 1.195597165991903,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9348009446693655e-05,
+      "loss": 2.3787,
+      "step": 4725
+    },
+    {
+      "epoch": 1.1958502024291497,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.9347165991902835e-05,
+      "loss": 2.4036,
+      "step": 4726
+    },
+    {
+      "epoch": 1.1961032388663968,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9346322537112012e-05,
+      "loss": 2.5155,
+      "step": 4727
+    },
+    {
+      "epoch": 1.1963562753036436,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.934547908232119e-05,
+      "loss": 2.4754,
+      "step": 4728
+    },
+    {
+      "epoch": 1.1966093117408907,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9344635627530365e-05,
+      "loss": 2.4325,
+      "step": 4729
+    },
+    {
+      "epoch": 1.1968623481781377,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.9343792172739542e-05,
+      "loss": 2.5288,
+      "step": 4730
+    },
+    {
+      "epoch": 1.1971153846153846,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.934294871794872e-05,
+      "loss": 2.3732,
+      "step": 4731
+    },
+    {
+      "epoch": 1.1973684210526316,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9342105263157895e-05,
+      "loss": 2.432,
+      "step": 4732
+    },
+    {
+      "epoch": 1.1976214574898785,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.9341261808367072e-05,
+      "loss": 2.5505,
+      "step": 4733
+    },
+    {
+      "epoch": 1.1978744939271255,
+      "grad_norm": 3.125,
+      "learning_rate": 2.934041835357625e-05,
+      "loss": 2.4809,
+      "step": 4734
+    },
+    {
+      "epoch": 1.1981275303643724,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.933957489878543e-05,
+      "loss": 2.5434,
+      "step": 4735
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9338731443994602e-05,
+      "loss": 2.5342,
+      "step": 4736
+    },
+    {
+      "epoch": 1.1986336032388665,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.933788798920378e-05,
+      "loss": 2.5835,
+      "step": 4737
+    },
+    {
+      "epoch": 1.1988866396761133,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.933704453441296e-05,
+      "loss": 2.5954,
+      "step": 4738
+    },
+    {
+      "epoch": 1.1991396761133604,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9336201079622132e-05,
+      "loss": 2.4328,
+      "step": 4739
+    },
+    {
+      "epoch": 1.1993927125506072,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.933535762483131e-05,
+      "loss": 2.5172,
+      "step": 4740
+    },
+    {
+      "epoch": 1.1993927125506072,
+      "eval_loss": 2.6253373622894287,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.0263,
+      "eval_samples_per_second": 9.488,
+      "eval_steps_per_second": 9.488,
+      "step": 4740
+    },
+    {
+      "epoch": 1.1996457489878543,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9334514170040485e-05,
+      "loss": 2.6028,
+      "step": 4741
+    },
+    {
+      "epoch": 1.1998987854251013,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.9333670715249665e-05,
+      "loss": 2.3887,
+      "step": 4742
+    },
+    {
+      "epoch": 1.2001518218623481,
+      "grad_norm": 3.25,
+      "learning_rate": 2.933282726045884e-05,
+      "loss": 2.4286,
+      "step": 4743
+    },
+    {
+      "epoch": 1.2004048582995952,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9331983805668015e-05,
+      "loss": 2.6043,
+      "step": 4744
+    },
+    {
+      "epoch": 1.200657894736842,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9331140350877195e-05,
+      "loss": 2.5337,
+      "step": 4745
+    },
+    {
+      "epoch": 1.200910931174089,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9330296896086372e-05,
+      "loss": 2.3406,
+      "step": 4746
+    },
+    {
+      "epoch": 1.2011639676113361,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9329453441295545e-05,
+      "loss": 2.5251,
+      "step": 4747
+    },
+    {
+      "epoch": 1.201417004048583,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.9328609986504725e-05,
+      "loss": 2.5896,
+      "step": 4748
+    },
+    {
+      "epoch": 1.20167004048583,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.9327766531713902e-05,
+      "loss": 2.4282,
+      "step": 4749
+    },
+    {
+      "epoch": 1.2019230769230769,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9326923076923075e-05,
+      "loss": 2.4458,
+      "step": 4750
+    },
+    {
+      "epoch": 1.202176113360324,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9326079622132255e-05,
+      "loss": 2.5674,
+      "step": 4751
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9325236167341432e-05,
+      "loss": 2.5531,
+      "step": 4752
+    },
+    {
+      "epoch": 1.2026821862348178,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.932439271255061e-05,
+      "loss": 2.5186,
+      "step": 4753
+    },
+    {
+      "epoch": 1.2029352226720649,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9323549257759785e-05,
+      "loss": 2.4762,
+      "step": 4754
+    },
+    {
+      "epoch": 1.2031882591093117,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.9322705802968962e-05,
+      "loss": 2.4619,
+      "step": 4755
+    },
+    {
+      "epoch": 1.2034412955465588,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.932186234817814e-05,
+      "loss": 2.5175,
+      "step": 4756
+    },
+    {
+      "epoch": 1.2036943319838056,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9321018893387315e-05,
+      "loss": 2.4957,
+      "step": 4757
+    },
+    {
+      "epoch": 1.2039473684210527,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9320175438596492e-05,
+      "loss": 2.5917,
+      "step": 4758
+    },
+    {
+      "epoch": 1.2042004048582995,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.931933198380567e-05,
+      "loss": 2.5233,
+      "step": 4759
+    },
+    {
+      "epoch": 1.2044534412955465,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9318488529014845e-05,
+      "loss": 2.5211,
+      "step": 4760
+    },
+    {
+      "epoch": 1.2047064777327936,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9317645074224022e-05,
+      "loss": 2.6007,
+      "step": 4761
+    },
+    {
+      "epoch": 1.2049595141700404,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.93168016194332e-05,
+      "loss": 2.508,
+      "step": 4762
+    },
+    {
+      "epoch": 1.2052125506072875,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9315958164642375e-05,
+      "loss": 2.4842,
+      "step": 4763
+    },
+    {
+      "epoch": 1.2054655870445343,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9315114709851555e-05,
+      "loss": 2.4998,
+      "step": 4764
+    },
+    {
+      "epoch": 1.2057186234817814,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.931427125506073e-05,
+      "loss": 2.5823,
+      "step": 4765
+    },
+    {
+      "epoch": 1.2059716599190282,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9313427800269905e-05,
+      "loss": 2.5026,
+      "step": 4766
+    },
+    {
+      "epoch": 1.2062246963562753,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9312584345479085e-05,
+      "loss": 2.633,
+      "step": 4767
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.931174089068826e-05,
+      "loss": 2.6065,
+      "step": 4768
+    },
+    {
+      "epoch": 1.2067307692307692,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9310897435897435e-05,
+      "loss": 2.492,
+      "step": 4769
+    },
+    {
+      "epoch": 1.2069838056680162,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9310053981106615e-05,
+      "loss": 2.548,
+      "step": 4770
+    },
+    {
+      "epoch": 1.2069838056680162,
+      "eval_loss": 2.6264219284057617,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.5146,
+      "eval_samples_per_second": 9.437,
+      "eval_steps_per_second": 9.437,
+      "step": 4770
+    },
+    {
+      "epoch": 1.2072368421052633,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.930921052631579e-05,
+      "loss": 2.4923,
+      "step": 4771
+    },
+    {
+      "epoch": 1.20748987854251,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9308367071524965e-05,
+      "loss": 2.6134,
+      "step": 4772
+    },
+    {
+      "epoch": 1.2077429149797572,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9307523616734145e-05,
+      "loss": 2.4916,
+      "step": 4773
+    },
+    {
+      "epoch": 1.207995951417004,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.930668016194332e-05,
+      "loss": 2.4906,
+      "step": 4774
+    },
+    {
+      "epoch": 1.208248987854251,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.9305836707152498e-05,
+      "loss": 2.4725,
+      "step": 4775
+    },
+    {
+      "epoch": 1.208502024291498,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.930499325236167e-05,
+      "loss": 2.4687,
+      "step": 4776
+    },
+    {
+      "epoch": 1.208755060728745,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.930414979757085e-05,
+      "loss": 2.4417,
+      "step": 4777
+    },
+    {
+      "epoch": 1.209008097165992,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9303306342780028e-05,
+      "loss": 2.5353,
+      "step": 4778
+    },
+    {
+      "epoch": 1.2092611336032388,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.93024628879892e-05,
+      "loss": 2.4296,
+      "step": 4779
+    },
+    {
+      "epoch": 1.209514170040486,
+      "grad_norm": 3.625,
+      "learning_rate": 2.930161943319838e-05,
+      "loss": 2.5006,
+      "step": 4780
+    },
+    {
+      "epoch": 1.2097672064777327,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9300775978407558e-05,
+      "loss": 2.4784,
+      "step": 4781
+    },
+    {
+      "epoch": 1.2100202429149798,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9299932523616735e-05,
+      "loss": 2.5242,
+      "step": 4782
+    },
+    {
+      "epoch": 1.2102732793522266,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.929908906882591e-05,
+      "loss": 2.5191,
+      "step": 4783
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9298245614035088e-05,
+      "loss": 2.5031,
+      "step": 4784
+    },
+    {
+      "epoch": 1.2107793522267207,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9297402159244265e-05,
+      "loss": 2.5261,
+      "step": 4785
+    },
+    {
+      "epoch": 1.2110323886639676,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9296558704453445e-05,
+      "loss": 2.5285,
+      "step": 4786
+    },
+    {
+      "epoch": 1.2112854251012146,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9295715249662618e-05,
+      "loss": 2.508,
+      "step": 4787
+    },
+    {
+      "epoch": 1.2115384615384615,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9294871794871795e-05,
+      "loss": 2.4294,
+      "step": 4788
+    },
+    {
+      "epoch": 1.2117914979757085,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9294028340080975e-05,
+      "loss": 2.4544,
+      "step": 4789
+    },
+    {
+      "epoch": 1.2120445344129553,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9293184885290148e-05,
+      "loss": 2.6036,
+      "step": 4790
+    },
+    {
+      "epoch": 1.2122975708502024,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9292341430499325e-05,
+      "loss": 2.4634,
+      "step": 4791
+    },
+    {
+      "epoch": 1.2125506072874495,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9291497975708505e-05,
+      "loss": 2.4254,
+      "step": 4792
+    },
+    {
+      "epoch": 1.2128036437246963,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.929065452091768e-05,
+      "loss": 2.6041,
+      "step": 4793
+    },
+    {
+      "epoch": 1.2130566801619433,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9289811066126855e-05,
+      "loss": 2.5233,
+      "step": 4794
+    },
+    {
+      "epoch": 1.2133097165991904,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.928896761133603e-05,
+      "loss": 2.5501,
+      "step": 4795
+    },
+    {
+      "epoch": 1.2135627530364372,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.928812415654521e-05,
+      "loss": 2.4565,
+      "step": 4796
+    },
+    {
+      "epoch": 1.2138157894736843,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9287280701754388e-05,
+      "loss": 2.5408,
+      "step": 4797
+    },
+    {
+      "epoch": 1.2140688259109311,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.928643724696356e-05,
+      "loss": 2.3655,
+      "step": 4798
+    },
+    {
+      "epoch": 1.2143218623481782,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.928559379217274e-05,
+      "loss": 2.4616,
+      "step": 4799
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9284750337381918e-05,
+      "loss": 2.5094,
+      "step": 4800
+    },
+    {
+      "epoch": 1.214574898785425,
+      "eval_loss": 2.624061107635498,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.4069,
+      "eval_samples_per_second": 9.475,
+      "eval_steps_per_second": 9.475,
+      "step": 4800
+    },
+    {
+      "epoch": 1.214827935222672,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.928390688259109e-05,
+      "loss": 2.4628,
+      "step": 4801
+    },
+    {
+      "epoch": 1.2150809716599191,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.928306342780027e-05,
+      "loss": 2.472,
+      "step": 4802
+    },
+    {
+      "epoch": 1.215334008097166,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9282219973009448e-05,
+      "loss": 2.5937,
+      "step": 4803
+    },
+    {
+      "epoch": 1.215587044534413,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9281376518218625e-05,
+      "loss": 2.6353,
+      "step": 4804
+    },
+    {
+      "epoch": 1.2158400809716599,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.92805330634278e-05,
+      "loss": 2.5176,
+      "step": 4805
+    },
+    {
+      "epoch": 1.216093117408907,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9279689608636978e-05,
+      "loss": 2.4868,
+      "step": 4806
+    },
+    {
+      "epoch": 1.2163461538461537,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9278846153846155e-05,
+      "loss": 2.5965,
+      "step": 4807
+    },
+    {
+      "epoch": 1.2165991902834008,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.9278002699055335e-05,
+      "loss": 2.5728,
+      "step": 4808
+    },
+    {
+      "epoch": 1.2168522267206479,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9277159244264508e-05,
+      "loss": 2.5152,
+      "step": 4809
+    },
+    {
+      "epoch": 1.2171052631578947,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9276315789473684e-05,
+      "loss": 2.6274,
+      "step": 4810
+    },
+    {
+      "epoch": 1.2173582995951417,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.927547233468286e-05,
+      "loss": 2.5386,
+      "step": 4811
+    },
+    {
+      "epoch": 1.2176113360323886,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.9274628879892038e-05,
+      "loss": 2.2323,
+      "step": 4812
+    },
+    {
+      "epoch": 1.2178643724696356,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9273785425101214e-05,
+      "loss": 2.5679,
+      "step": 4813
+    },
+    {
+      "epoch": 1.2181174089068825,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.927294197031039e-05,
+      "loss": 2.2249,
+      "step": 4814
+    },
+    {
+      "epoch": 1.2183704453441295,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.927209851551957e-05,
+      "loss": 2.5246,
+      "step": 4815
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.9271255060728744e-05,
+      "loss": 2.6056,
+      "step": 4816
+    },
+    {
+      "epoch": 1.2188765182186234,
+      "grad_norm": 3.0,
+      "learning_rate": 2.927041160593792e-05,
+      "loss": 2.5809,
+      "step": 4817
+    },
+    {
+      "epoch": 1.2191295546558705,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.92695681511471e-05,
+      "loss": 2.5838,
+      "step": 4818
+    },
+    {
+      "epoch": 1.2193825910931175,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.9268724696356274e-05,
+      "loss": 2.5213,
+      "step": 4819
+    },
+    {
+      "epoch": 1.2196356275303644,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.926788124156545e-05,
+      "loss": 2.5089,
+      "step": 4820
+    },
+    {
+      "epoch": 1.2198886639676114,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.926703778677463e-05,
+      "loss": 2.5378,
+      "step": 4821
+    },
+    {
+      "epoch": 1.2201417004048583,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.9266194331983808e-05,
+      "loss": 2.4427,
+      "step": 4822
+    },
+    {
+      "epoch": 1.2203947368421053,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.926535087719298e-05,
+      "loss": 2.4293,
+      "step": 4823
+    },
+    {
+      "epoch": 1.2206477732793521,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.926450742240216e-05,
+      "loss": 2.3961,
+      "step": 4824
+    },
+    {
+      "epoch": 1.2209008097165992,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.9263663967611338e-05,
+      "loss": 2.527,
+      "step": 4825
+    },
+    {
+      "epoch": 1.2211538461538463,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9262820512820514e-05,
+      "loss": 2.5503,
+      "step": 4826
+    },
+    {
+      "epoch": 1.221406882591093,
+      "grad_norm": 2.875,
+      "learning_rate": 2.926197705802969e-05,
+      "loss": 2.4417,
+      "step": 4827
+    },
+    {
+      "epoch": 1.2216599190283401,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9261133603238868e-05,
+      "loss": 2.5912,
+      "step": 4828
+    },
+    {
+      "epoch": 1.221912955465587,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9260290148448044e-05,
+      "loss": 2.6031,
+      "step": 4829
+    },
+    {
+      "epoch": 1.222165991902834,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9259446693657218e-05,
+      "loss": 2.519,
+      "step": 4830
+    },
+    {
+      "epoch": 1.222165991902834,
+      "eval_loss": 2.6208016872406006,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.4335,
+      "eval_samples_per_second": 9.44,
+      "eval_steps_per_second": 9.44,
+      "step": 4830
+    },
+    {
+      "epoch": 1.2224190283400809,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9258603238866398e-05,
+      "loss": 2.44,
+      "step": 4831
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9257759784075574e-05,
+      "loss": 2.5485,
+      "step": 4832
+    },
+    {
+      "epoch": 1.222925101214575,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.925691632928475e-05,
+      "loss": 2.4329,
+      "step": 4833
+    },
+    {
+      "epoch": 1.2231781376518218,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9256072874493928e-05,
+      "loss": 2.4702,
+      "step": 4834
+    },
+    {
+      "epoch": 1.2234311740890689,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9255229419703104e-05,
+      "loss": 2.5533,
+      "step": 4835
+    },
+    {
+      "epoch": 1.2236842105263157,
+      "grad_norm": 3.125,
+      "learning_rate": 2.925438596491228e-05,
+      "loss": 2.4876,
+      "step": 4836
+    },
+    {
+      "epoch": 1.2239372469635628,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.925354251012146e-05,
+      "loss": 2.4812,
+      "step": 4837
+    },
+    {
+      "epoch": 1.2241902834008096,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9252699055330634e-05,
+      "loss": 2.4367,
+      "step": 4838
+    },
+    {
+      "epoch": 1.2244433198380567,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.925185560053981e-05,
+      "loss": 2.4898,
+      "step": 4839
+    },
+    {
+      "epoch": 1.2246963562753037,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.925101214574899e-05,
+      "loss": 2.5466,
+      "step": 4840
+    },
+    {
+      "epoch": 1.2249493927125505,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9250168690958164e-05,
+      "loss": 2.6015,
+      "step": 4841
+    },
+    {
+      "epoch": 1.2252024291497976,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.924932523616734e-05,
+      "loss": 2.4715,
+      "step": 4842
+    },
+    {
+      "epoch": 1.2254554655870447,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.924848178137652e-05,
+      "loss": 2.5601,
+      "step": 4843
+    },
+    {
+      "epoch": 1.2257085020242915,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9247638326585697e-05,
+      "loss": 2.4781,
+      "step": 4844
+    },
+    {
+      "epoch": 1.2259615384615385,
+      "grad_norm": 3.375,
+      "learning_rate": 2.924679487179487e-05,
+      "loss": 2.5608,
+      "step": 4845
+    },
+    {
+      "epoch": 1.2262145748987854,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.924595141700405e-05,
+      "loss": 2.3792,
+      "step": 4846
+    },
+    {
+      "epoch": 1.2264676113360324,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9245107962213227e-05,
+      "loss": 2.3864,
+      "step": 4847
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9244264507422404e-05,
+      "loss": 2.3797,
+      "step": 4848
+    },
+    {
+      "epoch": 1.2269736842105263,
+      "grad_norm": 18.125,
+      "learning_rate": 2.9243421052631577e-05,
+      "loss": 2.3448,
+      "step": 4849
+    },
+    {
+      "epoch": 1.2272267206477734,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.9242577597840757e-05,
+      "loss": 2.603,
+      "step": 4850
+    },
+    {
+      "epoch": 1.2274797570850202,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9241734143049934e-05,
+      "loss": 2.5406,
+      "step": 4851
+    },
+    {
+      "epoch": 1.2277327935222673,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.9240890688259107e-05,
+      "loss": 2.3968,
+      "step": 4852
+    },
+    {
+      "epoch": 1.227985829959514,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9240047233468287e-05,
+      "loss": 2.6096,
+      "step": 4853
+    },
+    {
+      "epoch": 1.2282388663967612,
+      "grad_norm": 3.765625,
+      "learning_rate": 2.9239203778677464e-05,
+      "loss": 2.5565,
+      "step": 4854
+    },
+    {
+      "epoch": 1.228491902834008,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.923836032388664e-05,
+      "loss": 2.4172,
+      "step": 4855
+    },
+    {
+      "epoch": 1.228744939271255,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9237516869095817e-05,
+      "loss": 2.5686,
+      "step": 4856
+    },
+    {
+      "epoch": 1.228997975708502,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9236673414304994e-05,
+      "loss": 2.5272,
+      "step": 4857
+    },
+    {
+      "epoch": 1.229251012145749,
+      "grad_norm": 2.875,
+      "learning_rate": 2.923582995951417e-05,
+      "loss": 2.4822,
+      "step": 4858
+    },
+    {
+      "epoch": 1.229504048582996,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9234986504723347e-05,
+      "loss": 2.3344,
+      "step": 4859
+    },
+    {
+      "epoch": 1.2297570850202428,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9234143049932524e-05,
+      "loss": 2.3893,
+      "step": 4860
+    },
+    {
+      "epoch": 1.2297570850202428,
+      "eval_loss": 2.6210267543792725,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.2042,
+      "eval_samples_per_second": 9.413,
+      "eval_steps_per_second": 9.413,
+      "step": 4860
+    },
+    {
+      "epoch": 1.23001012145749,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.92332995951417e-05,
+      "loss": 2.5404,
+      "step": 4861
+    },
+    {
+      "epoch": 1.2302631578947367,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.923245614035088e-05,
+      "loss": 2.4004,
+      "step": 4862
+    },
+    {
+      "epoch": 1.2305161943319838,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9231612685560054e-05,
+      "loss": 2.5744,
+      "step": 4863
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.923076923076923e-05,
+      "loss": 2.5018,
+      "step": 4864
+    },
+    {
+      "epoch": 1.2310222672064777,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9229925775978407e-05,
+      "loss": 2.5363,
+      "step": 4865
+    },
+    {
+      "epoch": 1.2312753036437247,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9229082321187587e-05,
+      "loss": 2.5358,
+      "step": 4866
+    },
+    {
+      "epoch": 1.2315283400809718,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.922823886639676e-05,
+      "loss": 2.3875,
+      "step": 4867
+    },
+    {
+      "epoch": 1.2317813765182186,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.9227395411605937e-05,
+      "loss": 2.5477,
+      "step": 4868
+    },
+    {
+      "epoch": 1.2320344129554657,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9226551956815117e-05,
+      "loss": 2.4382,
+      "step": 4869
+    },
+    {
+      "epoch": 1.2322874493927125,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.922570850202429e-05,
+      "loss": 2.5732,
+      "step": 4870
+    },
+    {
+      "epoch": 1.2325404858299596,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9224865047233467e-05,
+      "loss": 2.5862,
+      "step": 4871
+    },
+    {
+      "epoch": 1.2327935222672064,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9224021592442647e-05,
+      "loss": 2.4288,
+      "step": 4872
+    },
+    {
+      "epoch": 1.2330465587044535,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9223178137651824e-05,
+      "loss": 2.5342,
+      "step": 4873
+    },
+    {
+      "epoch": 1.2332995951417005,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9222334682860997e-05,
+      "loss": 2.4708,
+      "step": 4874
+    },
+    {
+      "epoch": 1.2335526315789473,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.9221491228070177e-05,
+      "loss": 2.4225,
+      "step": 4875
+    },
+    {
+      "epoch": 1.2338056680161944,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9220647773279354e-05,
+      "loss": 2.4088,
+      "step": 4876
+    },
+    {
+      "epoch": 1.2340587044534412,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.921980431848853e-05,
+      "loss": 2.5011,
+      "step": 4877
+    },
+    {
+      "epoch": 1.2343117408906883,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9218960863697707e-05,
+      "loss": 2.4753,
+      "step": 4878
+    },
+    {
+      "epoch": 1.2345647773279351,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9218117408906884e-05,
+      "loss": 2.534,
+      "step": 4879
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.921727395411606e-05,
+      "loss": 2.3167,
+      "step": 4880
+    },
+    {
+      "epoch": 1.2350708502024292,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9216430499325237e-05,
+      "loss": 2.5121,
+      "step": 4881
+    },
+    {
+      "epoch": 1.235323886639676,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9215587044534414e-05,
+      "loss": 2.3891,
+      "step": 4882
+    },
+    {
+      "epoch": 1.2355769230769231,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.921474358974359e-05,
+      "loss": 2.5465,
+      "step": 4883
+    },
+    {
+      "epoch": 1.23582995951417,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.9213900134952767e-05,
+      "loss": 2.5552,
+      "step": 4884
+    },
+    {
+      "epoch": 1.236082995951417,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9213056680161944e-05,
+      "loss": 2.5455,
+      "step": 4885
+    },
+    {
+      "epoch": 1.2363360323886639,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.921221322537112e-05,
+      "loss": 2.4993,
+      "step": 4886
+    },
+    {
+      "epoch": 1.236589068825911,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9211369770580297e-05,
+      "loss": 2.5463,
+      "step": 4887
+    },
+    {
+      "epoch": 1.236842105263158,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9210526315789474e-05,
+      "loss": 2.4387,
+      "step": 4888
+    },
+    {
+      "epoch": 1.2370951417004048,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.920968286099865e-05,
+      "loss": 2.4168,
+      "step": 4889
+    },
+    {
+      "epoch": 1.2373481781376519,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9208839406207827e-05,
+      "loss": 2.4858,
+      "step": 4890
+    },
+    {
+      "epoch": 1.2373481781376519,
+      "eval_loss": 2.6216318607330322,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.6793,
+      "eval_samples_per_second": 9.431,
+      "eval_steps_per_second": 9.431,
+      "step": 4890
+    },
+    {
+      "epoch": 1.2376012145748987,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9207995951417007e-05,
+      "loss": 2.4315,
+      "step": 4891
+    },
+    {
+      "epoch": 1.2378542510121457,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.920715249662618e-05,
+      "loss": 2.5463,
+      "step": 4892
+    },
+    {
+      "epoch": 1.2381072874493928,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9206309041835357e-05,
+      "loss": 2.5655,
+      "step": 4893
+    },
+    {
+      "epoch": 1.2383603238866396,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.9205465587044537e-05,
+      "loss": 2.4437,
+      "step": 4894
+    },
+    {
+      "epoch": 1.2386133603238867,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9204622132253714e-05,
+      "loss": 2.6049,
+      "step": 4895
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9203778677462887e-05,
+      "loss": 2.5757,
+      "step": 4896
+    },
+    {
+      "epoch": 1.2391194331983806,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9202935222672067e-05,
+      "loss": 2.5064,
+      "step": 4897
+    },
+    {
+      "epoch": 1.2393724696356276,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9202091767881244e-05,
+      "loss": 2.4922,
+      "step": 4898
+    },
+    {
+      "epoch": 1.2396255060728745,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9201248313090417e-05,
+      "loss": 2.3647,
+      "step": 4899
+    },
+    {
+      "epoch": 1.2398785425101215,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9200404858299597e-05,
+      "loss": 2.5583,
+      "step": 4900
+    },
+    {
+      "epoch": 1.2401315789473684,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9199561403508774e-05,
+      "loss": 2.5413,
+      "step": 4901
+    },
+    {
+      "epoch": 1.2403846153846154,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.919871794871795e-05,
+      "loss": 2.5112,
+      "step": 4902
+    },
+    {
+      "epoch": 1.2406376518218623,
+      "grad_norm": 2.625,
+      "learning_rate": 2.9197874493927123e-05,
+      "loss": 2.4141,
+      "step": 4903
+    },
+    {
+      "epoch": 1.2408906882591093,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9197031039136303e-05,
+      "loss": 2.4808,
+      "step": 4904
+    },
+    {
+      "epoch": 1.2411437246963564,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.919618758434548e-05,
+      "loss": 2.5092,
+      "step": 4905
+    },
+    {
+      "epoch": 1.2413967611336032,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9195344129554657e-05,
+      "loss": 2.4827,
+      "step": 4906
+    },
+    {
+      "epoch": 1.2416497975708503,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9194500674763833e-05,
+      "loss": 2.5604,
+      "step": 4907
+    },
+    {
+      "epoch": 1.241902834008097,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.919365721997301e-05,
+      "loss": 2.5438,
+      "step": 4908
+    },
+    {
+      "epoch": 1.2421558704453441,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9192813765182187e-05,
+      "loss": 2.4495,
+      "step": 4909
+    },
+    {
+      "epoch": 1.242408906882591,
+      "grad_norm": 2.75,
+      "learning_rate": 2.9191970310391363e-05,
+      "loss": 2.3564,
+      "step": 4910
+    },
+    {
+      "epoch": 1.242661943319838,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.919112685560054e-05,
+      "loss": 2.4295,
+      "step": 4911
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9190283400809717e-05,
+      "loss": 2.4607,
+      "step": 4912
+    },
+    {
+      "epoch": 1.243168016194332,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9189439946018897e-05,
+      "loss": 2.4642,
+      "step": 4913
+    },
+    {
+      "epoch": 1.243421052631579,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.918859649122807e-05,
+      "loss": 2.5681,
+      "step": 4914
+    },
+    {
+      "epoch": 1.2436740890688258,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9187753036437247e-05,
+      "loss": 2.5654,
+      "step": 4915
+    },
+    {
+      "epoch": 1.2439271255060729,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9186909581646427e-05,
+      "loss": 2.4058,
+      "step": 4916
+    },
+    {
+      "epoch": 1.24418016194332,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.9186066126855603e-05,
+      "loss": 2.4839,
+      "step": 4917
+    },
+    {
+      "epoch": 1.2444331983805668,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9185222672064777e-05,
+      "loss": 2.4778,
+      "step": 4918
+    },
+    {
+      "epoch": 1.2446862348178138,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9184379217273953e-05,
+      "loss": 2.4988,
+      "step": 4919
+    },
+    {
+      "epoch": 1.2449392712550607,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9183535762483133e-05,
+      "loss": 2.5613,
+      "step": 4920
+    },
+    {
+      "epoch": 1.2449392712550607,
+      "eval_loss": 2.6197807788848877,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.61,
+      "eval_samples_per_second": 9.434,
+      "eval_steps_per_second": 9.434,
+      "step": 4920
+    },
+    {
+      "epoch": 1.2451923076923077,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9182692307692307e-05,
+      "loss": 2.4286,
+      "step": 4921
+    },
+    {
+      "epoch": 1.2454453441295548,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9181848852901483e-05,
+      "loss": 2.4183,
+      "step": 4922
+    },
+    {
+      "epoch": 1.2456983805668016,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9181005398110663e-05,
+      "loss": 2.4899,
+      "step": 4923
+    },
+    {
+      "epoch": 1.2459514170040487,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.918016194331984e-05,
+      "loss": 2.5233,
+      "step": 4924
+    },
+    {
+      "epoch": 1.2462044534412955,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.9179318488529013e-05,
+      "loss": 2.5501,
+      "step": 4925
+    },
+    {
+      "epoch": 1.2464574898785425,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9178475033738193e-05,
+      "loss": 2.569,
+      "step": 4926
+    },
+    {
+      "epoch": 1.2467105263157894,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.917763157894737e-05,
+      "loss": 2.5336,
+      "step": 4927
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9176788124156547e-05,
+      "loss": 2.5982,
+      "step": 4928
+    },
+    {
+      "epoch": 1.2472165991902835,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9175944669365723e-05,
+      "loss": 2.5049,
+      "step": 4929
+    },
+    {
+      "epoch": 1.2474696356275303,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.91751012145749e-05,
+      "loss": 2.6052,
+      "step": 4930
+    },
+    {
+      "epoch": 1.2477226720647774,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9174257759784076e-05,
+      "loss": 2.6721,
+      "step": 4931
+    },
+    {
+      "epoch": 1.2479757085020242,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.9173414304993253e-05,
+      "loss": 2.4868,
+      "step": 4932
+    },
+    {
+      "epoch": 1.2482287449392713,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.917257085020243e-05,
+      "loss": 2.5401,
+      "step": 4933
+    },
+    {
+      "epoch": 1.248481781376518,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9171727395411606e-05,
+      "loss": 2.5003,
+      "step": 4934
+    },
+    {
+      "epoch": 1.2487348178137652,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9170883940620786e-05,
+      "loss": 2.3742,
+      "step": 4935
+    },
+    {
+      "epoch": 1.2489878542510122,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.917004048582996e-05,
+      "loss": 2.4795,
+      "step": 4936
+    },
+    {
+      "epoch": 1.249240890688259,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9169197031039136e-05,
+      "loss": 2.4812,
+      "step": 4937
+    },
+    {
+      "epoch": 1.249493927125506,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.9168353576248313e-05,
+      "loss": 2.4475,
+      "step": 4938
+    },
+    {
+      "epoch": 1.249746963562753,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.916751012145749e-05,
+      "loss": 2.459,
+      "step": 4939
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9166666666666666e-05,
+      "loss": 2.5519,
+      "step": 4940
+    },
+    {
+      "epoch": 1.2502530364372468,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9165823211875843e-05,
+      "loss": 2.52,
+      "step": 4941
+    },
+    {
+      "epoch": 1.250506072874494,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9164979757085023e-05,
+      "loss": 2.4046,
+      "step": 4942
+    },
+    {
+      "epoch": 1.250759109311741,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9164136302294196e-05,
+      "loss": 2.4615,
+      "step": 4943
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9163292847503373e-05,
+      "loss": 2.5401,
+      "step": 4944
+    },
+    {
+      "epoch": 1.2512651821862348,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9162449392712553e-05,
+      "loss": 2.5569,
+      "step": 4945
+    },
+    {
+      "epoch": 1.251518218623482,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.916160593792173e-05,
+      "loss": 2.654,
+      "step": 4946
+    },
+    {
+      "epoch": 1.2517712550607287,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9160762483130903e-05,
+      "loss": 2.5555,
+      "step": 4947
+    },
+    {
+      "epoch": 1.2520242914979758,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9159919028340083e-05,
+      "loss": 2.5042,
+      "step": 4948
+    },
+    {
+      "epoch": 1.2522773279352226,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.915907557354926e-05,
+      "loss": 2.461,
+      "step": 4949
+    },
+    {
+      "epoch": 1.2525303643724697,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.9158232118758433e-05,
+      "loss": 2.3116,
+      "step": 4950
+    },
+    {
+      "epoch": 1.2525303643724697,
+      "eval_loss": 2.6171419620513916,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.8495,
+      "eval_samples_per_second": 9.425,
+      "eval_steps_per_second": 9.425,
+      "step": 4950
+    },
+    {
+      "epoch": 1.2527834008097165,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9157388663967613e-05,
+      "loss": 2.5063,
+      "step": 4951
+    },
+    {
+      "epoch": 1.2530364372469636,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.915654520917679e-05,
+      "loss": 2.3825,
+      "step": 4952
+    },
+    {
+      "epoch": 1.2532894736842106,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9155701754385966e-05,
+      "loss": 2.3704,
+      "step": 4953
+    },
+    {
+      "epoch": 1.2535425101214575,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.9154858299595143e-05,
+      "loss": 2.5014,
+      "step": 4954
+    },
+    {
+      "epoch": 1.2537955465587045,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.915401484480432e-05,
+      "loss": 2.6898,
+      "step": 4955
+    },
+    {
+      "epoch": 1.2540485829959513,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9153171390013496e-05,
+      "loss": 2.5514,
+      "step": 4956
+    },
+    {
+      "epoch": 1.2543016194331984,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.9152327935222673e-05,
+      "loss": 2.2733,
+      "step": 4957
+    },
+    {
+      "epoch": 1.2545546558704452,
+      "grad_norm": 3.25,
+      "learning_rate": 2.915148448043185e-05,
+      "loss": 2.5272,
+      "step": 4958
+    },
+    {
+      "epoch": 1.2548076923076923,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9150641025641026e-05,
+      "loss": 2.6367,
+      "step": 4959
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9149797570850203e-05,
+      "loss": 2.4212,
+      "step": 4960
+    },
+    {
+      "epoch": 1.2553137651821862,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.914895411605938e-05,
+      "loss": 2.5188,
+      "step": 4961
+    },
+    {
+      "epoch": 1.2555668016194332,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9148110661268556e-05,
+      "loss": 2.5167,
+      "step": 4962
+    },
+    {
+      "epoch": 1.2558198380566803,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9147267206477733e-05,
+      "loss": 2.5982,
+      "step": 4963
+    },
+    {
+      "epoch": 1.2560728744939271,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9146423751686913e-05,
+      "loss": 2.5853,
+      "step": 4964
+    },
+    {
+      "epoch": 1.256325910931174,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9145580296896086e-05,
+      "loss": 2.5485,
+      "step": 4965
+    },
+    {
+      "epoch": 1.256578947368421,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9144736842105263e-05,
+      "loss": 2.3347,
+      "step": 4966
+    },
+    {
+      "epoch": 1.256831983805668,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9143893387314443e-05,
+      "loss": 2.3572,
+      "step": 4967
+    },
+    {
+      "epoch": 1.257085020242915,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9143049932523616e-05,
+      "loss": 2.4284,
+      "step": 4968
+    },
+    {
+      "epoch": 1.257338056680162,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9142206477732793e-05,
+      "loss": 2.5323,
+      "step": 4969
+    },
+    {
+      "epoch": 1.257591093117409,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9141363022941973e-05,
+      "loss": 2.4026,
+      "step": 4970
+    },
+    {
+      "epoch": 1.2578441295546559,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.914051956815115e-05,
+      "loss": 2.4919,
+      "step": 4971
+    },
+    {
+      "epoch": 1.258097165991903,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9139676113360323e-05,
+      "loss": 2.4947,
+      "step": 4972
+    },
+    {
+      "epoch": 1.2583502024291497,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9138832658569503e-05,
+      "loss": 2.4618,
+      "step": 4973
+    },
+    {
+      "epoch": 1.2586032388663968,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.913798920377868e-05,
+      "loss": 2.5379,
+      "step": 4974
+    },
+    {
+      "epoch": 1.2588562753036436,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.9137145748987856e-05,
+      "loss": 2.4762,
+      "step": 4975
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.913630229419703e-05,
+      "loss": 2.4616,
+      "step": 4976
+    },
+    {
+      "epoch": 1.2593623481781377,
+      "grad_norm": 18.375,
+      "learning_rate": 2.913545883940621e-05,
+      "loss": 2.534,
+      "step": 4977
+    },
+    {
+      "epoch": 1.2596153846153846,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9134615384615386e-05,
+      "loss": 2.5831,
+      "step": 4978
+    },
+    {
+      "epoch": 1.2598684210526316,
+      "grad_norm": 2.875,
+      "learning_rate": 2.913377192982456e-05,
+      "loss": 2.3882,
+      "step": 4979
+    },
+    {
+      "epoch": 1.2601214574898785,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.913292847503374e-05,
+      "loss": 2.5314,
+      "step": 4980
+    },
+    {
+      "epoch": 1.2601214574898785,
+      "eval_loss": 2.6181039810180664,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.0573,
+      "eval_samples_per_second": 9.453,
+      "eval_steps_per_second": 9.453,
+      "step": 4980
+    },
+    {
+      "epoch": 1.2603744939271255,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.9132085020242916e-05,
+      "loss": 2.6143,
+      "step": 4981
+    },
+    {
+      "epoch": 1.2606275303643724,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9131241565452093e-05,
+      "loss": 2.5937,
+      "step": 4982
+    },
+    {
+      "epoch": 1.2608805668016194,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.913039811066127e-05,
+      "loss": 2.5286,
+      "step": 4983
+    },
+    {
+      "epoch": 1.2611336032388665,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9129554655870446e-05,
+      "loss": 2.4558,
+      "step": 4984
+    },
+    {
+      "epoch": 1.2613866396761133,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9128711201079623e-05,
+      "loss": 2.4286,
+      "step": 4985
+    },
+    {
+      "epoch": 1.2616396761133604,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9127867746288803e-05,
+      "loss": 2.4181,
+      "step": 4986
+    },
+    {
+      "epoch": 1.2618927125506074,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9127024291497976e-05,
+      "loss": 2.6668,
+      "step": 4987
+    },
+    {
+      "epoch": 1.2621457489878543,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9126180836707153e-05,
+      "loss": 2.531,
+      "step": 4988
+    },
+    {
+      "epoch": 1.262398785425101,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.9125337381916333e-05,
+      "loss": 2.3565,
+      "step": 4989
+    },
+    {
+      "epoch": 1.2626518218623481,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9124493927125506e-05,
+      "loss": 2.5077,
+      "step": 4990
+    },
+    {
+      "epoch": 1.2629048582995952,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9123650472334682e-05,
+      "loss": 2.5384,
+      "step": 4991
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.912280701754386e-05,
+      "loss": 2.5578,
+      "step": 4992
+    },
+    {
+      "epoch": 1.263410931174089,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.912196356275304e-05,
+      "loss": 2.3894,
+      "step": 4993
+    },
+    {
+      "epoch": 1.2636639676113361,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9121120107962212e-05,
+      "loss": 2.4435,
+      "step": 4994
+    },
+    {
+      "epoch": 1.263917004048583,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.912027665317139e-05,
+      "loss": 2.5564,
+      "step": 4995
+    },
+    {
+      "epoch": 1.26417004048583,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.911943319838057e-05,
+      "loss": 2.4714,
+      "step": 4996
+    },
+    {
+      "epoch": 1.2644230769230769,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9118589743589746e-05,
+      "loss": 2.4784,
+      "step": 4997
+    },
+    {
+      "epoch": 1.264676113360324,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.911774628879892e-05,
+      "loss": 2.5149,
+      "step": 4998
+    },
+    {
+      "epoch": 1.2649291497975708,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.91169028340081e-05,
+      "loss": 2.4569,
+      "step": 4999
+    },
+    {
+      "epoch": 1.2651821862348178,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9116059379217276e-05,
+      "loss": 2.5072,
+      "step": 5000
+    },
+    {
+      "epoch": 1.2654352226720649,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.911521592442645e-05,
+      "loss": 2.3544,
+      "step": 5001
+    },
+    {
+      "epoch": 1.2656882591093117,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.911437246963563e-05,
+      "loss": 2.5336,
+      "step": 5002
+    },
+    {
+      "epoch": 1.2659412955465588,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9113529014844806e-05,
+      "loss": 2.4622,
+      "step": 5003
+    },
+    {
+      "epoch": 1.2661943319838056,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9112685560053982e-05,
+      "loss": 2.5079,
+      "step": 5004
+    },
+    {
+      "epoch": 1.2664473684210527,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.911184210526316e-05,
+      "loss": 2.3634,
+      "step": 5005
+    },
+    {
+      "epoch": 1.2667004048582995,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9110998650472336e-05,
+      "loss": 2.5807,
+      "step": 5006
+    },
+    {
+      "epoch": 1.2669534412955465,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9110155195681512e-05,
+      "loss": 2.5075,
+      "step": 5007
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.910931174089069e-05,
+      "loss": 2.5389,
+      "step": 5008
+    },
+    {
+      "epoch": 1.2674595141700404,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9108468286099866e-05,
+      "loss": 2.3963,
+      "step": 5009
+    },
+    {
+      "epoch": 1.2677125506072875,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9107624831309042e-05,
+      "loss": 2.5469,
+      "step": 5010
+    },
+    {
+      "epoch": 1.2677125506072875,
+      "eval_loss": 2.6138052940368652,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 288.5222,
+      "eval_samples_per_second": 9.011,
+      "eval_steps_per_second": 9.011,
+      "step": 5010
+    },
+    {
+      "epoch": 1.2679655870445345,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.910678137651822e-05,
+      "loss": 2.407,
+      "step": 5011
+    },
+    {
+      "epoch": 1.2682186234817814,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9105937921727396e-05,
+      "loss": 2.5038,
+      "step": 5012
+    },
+    {
+      "epoch": 1.2684716599190282,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9105094466936572e-05,
+      "loss": 2.493,
+      "step": 5013
+    },
+    {
+      "epoch": 1.2687246963562753,
+      "grad_norm": 2.875,
+      "learning_rate": 2.910425101214575e-05,
+      "loss": 2.5166,
+      "step": 5014
+    },
+    {
+      "epoch": 1.2689777327935223,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.910340755735493e-05,
+      "loss": 2.5263,
+      "step": 5015
+    },
+    {
+      "epoch": 1.2692307692307692,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9102564102564102e-05,
+      "loss": 2.6044,
+      "step": 5016
+    },
+    {
+      "epoch": 1.2694838056680162,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.910172064777328e-05,
+      "loss": 2.492,
+      "step": 5017
+    },
+    {
+      "epoch": 1.2697368421052633,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.910087719298246e-05,
+      "loss": 2.5883,
+      "step": 5018
+    },
+    {
+      "epoch": 1.26998987854251,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9100033738191632e-05,
+      "loss": 2.5413,
+      "step": 5019
+    },
+    {
+      "epoch": 1.2702429149797572,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.909919028340081e-05,
+      "loss": 2.4661,
+      "step": 5020
+    },
+    {
+      "epoch": 1.270495951417004,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.909834682860999e-05,
+      "loss": 2.4783,
+      "step": 5021
+    },
+    {
+      "epoch": 1.270748987854251,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9097503373819165e-05,
+      "loss": 2.5434,
+      "step": 5022
+    },
+    {
+      "epoch": 1.271002024291498,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.909665991902834e-05,
+      "loss": 2.6008,
+      "step": 5023
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.909581646423752e-05,
+      "loss": 2.4518,
+      "step": 5024
+    },
+    {
+      "epoch": 1.271508097165992,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9094973009446695e-05,
+      "loss": 2.5939,
+      "step": 5025
+    },
+    {
+      "epoch": 1.2717611336032388,
+      "grad_norm": 3.375,
+      "learning_rate": 2.9094129554655872e-05,
+      "loss": 2.5017,
+      "step": 5026
+    },
+    {
+      "epoch": 1.272014170040486,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.909328609986505e-05,
+      "loss": 2.3098,
+      "step": 5027
+    },
+    {
+      "epoch": 1.2722672064777327,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.9092442645074225e-05,
+      "loss": 2.3618,
+      "step": 5028
+    },
+    {
+      "epoch": 1.2725202429149798,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9091599190283402e-05,
+      "loss": 2.492,
+      "step": 5029
+    },
+    {
+      "epoch": 1.2727732793522266,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.9090755735492575e-05,
+      "loss": 2.3856,
+      "step": 5030
+    },
+    {
+      "epoch": 1.2730263157894737,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.9089912280701755e-05,
+      "loss": 2.4685,
+      "step": 5031
+    },
+    {
+      "epoch": 1.2732793522267207,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.9089068825910932e-05,
+      "loss": 2.5135,
+      "step": 5032
+    },
+    {
+      "epoch": 1.2735323886639676,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.908822537112011e-05,
+      "loss": 2.5513,
+      "step": 5033
+    },
+    {
+      "epoch": 1.2737854251012146,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9087381916329285e-05,
+      "loss": 2.497,
+      "step": 5034
+    },
+    {
+      "epoch": 1.2740384615384617,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.9086538461538462e-05,
+      "loss": 2.4853,
+      "step": 5035
+    },
+    {
+      "epoch": 1.2742914979757085,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.908569500674764e-05,
+      "loss": 2.5659,
+      "step": 5036
+    },
+    {
+      "epoch": 1.2745445344129553,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.908485155195682e-05,
+      "loss": 2.5594,
+      "step": 5037
+    },
+    {
+      "epoch": 1.2747975708502024,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.9084008097165992e-05,
+      "loss": 2.3321,
+      "step": 5038
+    },
+    {
+      "epoch": 1.2750506072874495,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.908316464237517e-05,
+      "loss": 2.442,
+      "step": 5039
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.908232118758435e-05,
+      "loss": 2.6083,
+      "step": 5040
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "eval_loss": 2.616400718688965,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.3705,
+      "eval_samples_per_second": 9.442,
+      "eval_steps_per_second": 9.442,
+      "step": 5040
+    },
+    {
+      "epoch": 1.2755566801619433,
+      "grad_norm": 2.75,
+      "learning_rate": 2.9081477732793522e-05,
+      "loss": 2.4976,
+      "step": 5041
+    },
+    {
+      "epoch": 1.2758097165991904,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.90806342780027e-05,
+      "loss": 2.554,
+      "step": 5042
+    },
+    {
+      "epoch": 1.2760627530364372,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.907979082321188e-05,
+      "loss": 2.5153,
+      "step": 5043
+    },
+    {
+      "epoch": 1.2763157894736843,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9078947368421055e-05,
+      "loss": 2.5358,
+      "step": 5044
+    },
+    {
+      "epoch": 1.2765688259109311,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.907810391363023e-05,
+      "loss": 2.4817,
+      "step": 5045
+    },
+    {
+      "epoch": 1.2768218623481782,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.9077260458839405e-05,
+      "loss": 2.5372,
+      "step": 5046
+    },
+    {
+      "epoch": 1.277074898785425,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9076417004048585e-05,
+      "loss": 2.4524,
+      "step": 5047
+    },
+    {
+      "epoch": 1.277327935222672,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.907557354925776e-05,
+      "loss": 2.4322,
+      "step": 5048
+    },
+    {
+      "epoch": 1.2775809716599191,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.9074730094466935e-05,
+      "loss": 2.4957,
+      "step": 5049
+    },
+    {
+      "epoch": 1.277834008097166,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.9073886639676115e-05,
+      "loss": 2.4335,
+      "step": 5050
+    },
+    {
+      "epoch": 1.278087044534413,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9073043184885292e-05,
+      "loss": 2.505,
+      "step": 5051
+    },
+    {
+      "epoch": 1.2783400809716599,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9072199730094465e-05,
+      "loss": 2.4473,
+      "step": 5052
+    },
+    {
+      "epoch": 1.278593117408907,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9071356275303645e-05,
+      "loss": 2.5496,
+      "step": 5053
+    },
+    {
+      "epoch": 1.2788461538461537,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.9070512820512822e-05,
+      "loss": 2.5192,
+      "step": 5054
+    },
+    {
+      "epoch": 1.2790991902834008,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9069669365722e-05,
+      "loss": 2.4995,
+      "step": 5055
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9068825910931175e-05,
+      "loss": 2.5877,
+      "step": 5056
+    },
+    {
+      "epoch": 1.2796052631578947,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.9067982456140352e-05,
+      "loss": 2.6183,
+      "step": 5057
+    },
+    {
+      "epoch": 1.2798582995951417,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.906713900134953e-05,
+      "loss": 2.5217,
+      "step": 5058
+    },
+    {
+      "epoch": 1.2801113360323888,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.9066295546558705e-05,
+      "loss": 2.5002,
+      "step": 5059
+    },
+    {
+      "epoch": 1.2803643724696356,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9065452091767882e-05,
+      "loss": 2.5123,
+      "step": 5060
+    },
+    {
+      "epoch": 1.2806174089068825,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.906460863697706e-05,
+      "loss": 2.5531,
+      "step": 5061
+    },
+    {
+      "epoch": 1.2808704453441295,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.906376518218624e-05,
+      "loss": 2.5125,
+      "step": 5062
+    },
+    {
+      "epoch": 1.2811234817813766,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.906292172739541e-05,
+      "loss": 2.471,
+      "step": 5063
+    },
+    {
+      "epoch": 1.2813765182186234,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.906207827260459e-05,
+      "loss": 2.5583,
+      "step": 5064
+    },
+    {
+      "epoch": 1.2816295546558705,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9061234817813765e-05,
+      "loss": 2.4873,
+      "step": 5065
+    },
+    {
+      "epoch": 1.2818825910931175,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9060391363022945e-05,
+      "loss": 2.3429,
+      "step": 5066
+    },
+    {
+      "epoch": 1.2821356275303644,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9059547908232118e-05,
+      "loss": 2.4037,
+      "step": 5067
+    },
+    {
+      "epoch": 1.2823886639676112,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9058704453441295e-05,
+      "loss": 2.3922,
+      "step": 5068
+    },
+    {
+      "epoch": 1.2826417004048583,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.9057860998650475e-05,
+      "loss": 2.4698,
+      "step": 5069
+    },
+    {
+      "epoch": 1.2828947368421053,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9057017543859648e-05,
+      "loss": 2.5981,
+      "step": 5070
+    },
+    {
+      "epoch": 1.2828947368421053,
+      "eval_loss": 2.6198151111602783,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.0079,
+      "eval_samples_per_second": 9.42,
+      "eval_steps_per_second": 9.42,
+      "step": 5070
+    },
+    {
+      "epoch": 1.2831477732793521,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.9056174089068825e-05,
+      "loss": 2.4058,
+      "step": 5071
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9055330634278005e-05,
+      "loss": 2.6139,
+      "step": 5072
+    },
+    {
+      "epoch": 1.2836538461538463,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.905448717948718e-05,
+      "loss": 2.3736,
+      "step": 5073
+    },
+    {
+      "epoch": 1.283906882591093,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9053643724696355e-05,
+      "loss": 2.4882,
+      "step": 5074
+    },
+    {
+      "epoch": 1.2841599190283401,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9052800269905535e-05,
+      "loss": 2.4906,
+      "step": 5075
+    },
+    {
+      "epoch": 1.284412955465587,
+      "grad_norm": 2.875,
+      "learning_rate": 2.905195681511471e-05,
+      "loss": 2.4465,
+      "step": 5076
+    },
+    {
+      "epoch": 1.284665991902834,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9051113360323888e-05,
+      "loss": 2.477,
+      "step": 5077
+    },
+    {
+      "epoch": 1.2849190283400809,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9050269905533065e-05,
+      "loss": 2.4182,
+      "step": 5078
+    },
+    {
+      "epoch": 1.285172064777328,
+      "grad_norm": 3.5,
+      "learning_rate": 2.904942645074224e-05,
+      "loss": 2.5358,
+      "step": 5079
+    },
+    {
+      "epoch": 1.285425101214575,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9048582995951418e-05,
+      "loss": 2.4893,
+      "step": 5080
+    },
+    {
+      "epoch": 1.2856781376518218,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9047739541160595e-05,
+      "loss": 2.5081,
+      "step": 5081
+    },
+    {
+      "epoch": 1.2859311740890689,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.904689608636977e-05,
+      "loss": 2.4435,
+      "step": 5082
+    },
+    {
+      "epoch": 1.2861842105263157,
+      "grad_norm": 4.125,
+      "learning_rate": 2.9046052631578948e-05,
+      "loss": 2.5354,
+      "step": 5083
+    },
+    {
+      "epoch": 1.2864372469635628,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.9045209176788125e-05,
+      "loss": 2.5224,
+      "step": 5084
+    },
+    {
+      "epoch": 1.2866902834008096,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.90443657219973e-05,
+      "loss": 2.5071,
+      "step": 5085
+    },
+    {
+      "epoch": 1.2869433198380567,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9043522267206478e-05,
+      "loss": 2.5637,
+      "step": 5086
+    },
+    {
+      "epoch": 1.2871963562753037,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.9042678812415655e-05,
+      "loss": 2.5551,
+      "step": 5087
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 3.375,
+      "learning_rate": 2.904183535762483e-05,
+      "loss": 2.5732,
+      "step": 5088
+    },
+    {
+      "epoch": 1.2877024291497976,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.9040991902834008e-05,
+      "loss": 2.5108,
+      "step": 5089
+    },
+    {
+      "epoch": 1.2879554655870447,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.9040148448043185e-05,
+      "loss": 2.4997,
+      "step": 5090
+    },
+    {
+      "epoch": 1.2882085020242915,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.9039304993252365e-05,
+      "loss": 2.459,
+      "step": 5091
+    },
+    {
+      "epoch": 1.2884615384615383,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9038461538461538e-05,
+      "loss": 2.5273,
+      "step": 5092
+    },
+    {
+      "epoch": 1.2887145748987854,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9037618083670715e-05,
+      "loss": 2.4629,
+      "step": 5093
+    },
+    {
+      "epoch": 1.2889676113360324,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9036774628879895e-05,
+      "loss": 2.5678,
+      "step": 5094
+    },
+    {
+      "epoch": 1.2892206477732793,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.903593117408907e-05,
+      "loss": 2.5412,
+      "step": 5095
+    },
+    {
+      "epoch": 1.2894736842105263,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9035087719298245e-05,
+      "loss": 2.5287,
+      "step": 5096
+    },
+    {
+      "epoch": 1.2897267206477734,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.9034244264507425e-05,
+      "loss": 2.3764,
+      "step": 5097
+    },
+    {
+      "epoch": 1.2899797570850202,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.90334008097166e-05,
+      "loss": 2.5581,
+      "step": 5098
+    },
+    {
+      "epoch": 1.2902327935222673,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.9032557354925775e-05,
+      "loss": 2.4284,
+      "step": 5099
+    },
+    {
+      "epoch": 1.290485829959514,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.903171390013495e-05,
+      "loss": 2.4671,
+      "step": 5100
+    },
+    {
+      "epoch": 1.290485829959514,
+      "eval_loss": 2.611436367034912,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 277.2996,
+      "eval_samples_per_second": 9.376,
+      "eval_steps_per_second": 9.376,
+      "step": 5100
+    },
+    {
+      "epoch": 1.2907388663967612,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.903087044534413e-05,
+      "loss": 2.5297,
+      "step": 5101
+    },
+    {
+      "epoch": 1.290991902834008,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9030026990553308e-05,
+      "loss": 2.4737,
+      "step": 5102
+    },
+    {
+      "epoch": 1.291244939271255,
+      "grad_norm": 3.0,
+      "learning_rate": 2.902918353576248e-05,
+      "loss": 2.4972,
+      "step": 5103
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.902834008097166e-05,
+      "loss": 2.3964,
+      "step": 5104
+    },
+    {
+      "epoch": 1.291751012145749,
+      "grad_norm": 3.0,
+      "learning_rate": 2.9027496626180838e-05,
+      "loss": 2.4075,
+      "step": 5105
+    },
+    {
+      "epoch": 1.292004048582996,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9026653171390015e-05,
+      "loss": 2.6131,
+      "step": 5106
+    },
+    {
+      "epoch": 1.2922570850202428,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.902580971659919e-05,
+      "loss": 2.4967,
+      "step": 5107
+    },
+    {
+      "epoch": 1.29251012145749,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9024966261808368e-05,
+      "loss": 2.4236,
+      "step": 5108
+    },
+    {
+      "epoch": 1.2927631578947367,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9024122807017544e-05,
+      "loss": 2.5149,
+      "step": 5109
+    },
+    {
+      "epoch": 1.2930161943319838,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.902327935222672e-05,
+      "loss": 2.5746,
+      "step": 5110
+    },
+    {
+      "epoch": 1.2932692307692308,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9022435897435898e-05,
+      "loss": 2.6163,
+      "step": 5111
+    },
+    {
+      "epoch": 1.2935222672064777,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9021592442645074e-05,
+      "loss": 2.3532,
+      "step": 5112
+    },
+    {
+      "epoch": 1.2937753036437247,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.9020748987854254e-05,
+      "loss": 2.4822,
+      "step": 5113
+    },
+    {
+      "epoch": 1.2940283400809718,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9019905533063428e-05,
+      "loss": 2.4796,
+      "step": 5114
+    },
+    {
+      "epoch": 1.2942813765182186,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9019062078272604e-05,
+      "loss": 2.5354,
+      "step": 5115
+    },
+    {
+      "epoch": 1.2945344129554655,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.9018218623481784e-05,
+      "loss": 2.4648,
+      "step": 5116
+    },
+    {
+      "epoch": 1.2947874493927125,
+      "grad_norm": 3.25,
+      "learning_rate": 2.9017375168690958e-05,
+      "loss": 2.5273,
+      "step": 5117
+    },
+    {
+      "epoch": 1.2950404858299596,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.9016531713900134e-05,
+      "loss": 2.3696,
+      "step": 5118
+    },
+    {
+      "epoch": 1.2952935222672064,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.901568825910931e-05,
+      "loss": 2.4062,
+      "step": 5119
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.901484480431849e-05,
+      "loss": 2.6017,
+      "step": 5120
+    },
+    {
+      "epoch": 1.2957995951417005,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9014001349527664e-05,
+      "loss": 2.479,
+      "step": 5121
+    },
+    {
+      "epoch": 1.2960526315789473,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.901315789473684e-05,
+      "loss": 2.6132,
+      "step": 5122
+    },
+    {
+      "epoch": 1.2963056680161944,
+      "grad_norm": 2.75,
+      "learning_rate": 2.901231443994602e-05,
+      "loss": 2.4306,
+      "step": 5123
+    },
+    {
+      "epoch": 1.2965587044534412,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.9011470985155198e-05,
+      "loss": 2.5228,
+      "step": 5124
+    },
+    {
+      "epoch": 1.2968117408906883,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.901062753036437e-05,
+      "loss": 2.4612,
+      "step": 5125
+    },
+    {
+      "epoch": 1.2970647773279351,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.900978407557355e-05,
+      "loss": 2.6297,
+      "step": 5126
+    },
+    {
+      "epoch": 1.2973178137651822,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9008940620782728e-05,
+      "loss": 2.4507,
+      "step": 5127
+    },
+    {
+      "epoch": 1.2975708502024292,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.90080971659919e-05,
+      "loss": 2.4633,
+      "step": 5128
+    },
+    {
+      "epoch": 1.297823886639676,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.900725371120108e-05,
+      "loss": 2.4752,
+      "step": 5129
+    },
+    {
+      "epoch": 1.2980769230769231,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9006410256410258e-05,
+      "loss": 2.519,
+      "step": 5130
+    },
+    {
+      "epoch": 1.2980769230769231,
+      "eval_loss": 2.6121344566345215,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.9425,
+      "eval_samples_per_second": 9.388,
+      "eval_steps_per_second": 9.388,
+      "step": 5130
+    },
+    {
+      "epoch": 1.29832995951417,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9005566801619434e-05,
+      "loss": 2.5506,
+      "step": 5131
+    },
+    {
+      "epoch": 1.298582995951417,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.900472334682861e-05,
+      "loss": 2.3371,
+      "step": 5132
+    },
+    {
+      "epoch": 1.2988360323886639,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.9003879892037788e-05,
+      "loss": 2.6133,
+      "step": 5133
+    },
+    {
+      "epoch": 1.299089068825911,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9003036437246964e-05,
+      "loss": 2.4029,
+      "step": 5134
+    },
+    {
+      "epoch": 1.299342105263158,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9002192982456144e-05,
+      "loss": 2.4493,
+      "step": 5135
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9001349527665318e-05,
+      "loss": 2.4517,
+      "step": 5136
+    },
+    {
+      "epoch": 1.2998481781376519,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.9000506072874494e-05,
+      "loss": 2.6227,
+      "step": 5137
+    },
+    {
+      "epoch": 1.300101214574899,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.899966261808367e-05,
+      "loss": 2.4387,
+      "step": 5138
+    },
+    {
+      "epoch": 1.3003542510121457,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8998819163292847e-05,
+      "loss": 2.5569,
+      "step": 5139
+    },
+    {
+      "epoch": 1.3006072874493926,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8997975708502024e-05,
+      "loss": 2.5553,
+      "step": 5140
+    },
+    {
+      "epoch": 1.3008603238866396,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.89971322537112e-05,
+      "loss": 2.4712,
+      "step": 5141
+    },
+    {
+      "epoch": 1.3011133603238867,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.899628879892038e-05,
+      "loss": 2.6475,
+      "step": 5142
+    },
+    {
+      "epoch": 1.3013663967611335,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8995445344129554e-05,
+      "loss": 2.476,
+      "step": 5143
+    },
+    {
+      "epoch": 1.3016194331983806,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.899460188933873e-05,
+      "loss": 2.5488,
+      "step": 5144
+    },
+    {
+      "epoch": 1.3018724696356276,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.899375843454791e-05,
+      "loss": 2.4586,
+      "step": 5145
+    },
+    {
+      "epoch": 1.3021255060728745,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8992914979757087e-05,
+      "loss": 2.5598,
+      "step": 5146
+    },
+    {
+      "epoch": 1.3023785425101215,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.899207152496626e-05,
+      "loss": 2.526,
+      "step": 5147
+    },
+    {
+      "epoch": 1.3026315789473684,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.899122807017544e-05,
+      "loss": 2.5099,
+      "step": 5148
+    },
+    {
+      "epoch": 1.3028846153846154,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8990384615384617e-05,
+      "loss": 2.476,
+      "step": 5149
+    },
+    {
+      "epoch": 1.3031376518218623,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.898954116059379e-05,
+      "loss": 2.5827,
+      "step": 5150
+    },
+    {
+      "epoch": 1.3033906882591093,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.898869770580297e-05,
+      "loss": 2.4778,
+      "step": 5151
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8987854251012147e-05,
+      "loss": 2.5187,
+      "step": 5152
+    },
+    {
+      "epoch": 1.3038967611336032,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8987010796221324e-05,
+      "loss": 2.4226,
+      "step": 5153
+    },
+    {
+      "epoch": 1.3041497975708503,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8986167341430497e-05,
+      "loss": 2.5199,
+      "step": 5154
+    },
+    {
+      "epoch": 1.304402834008097,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8985323886639677e-05,
+      "loss": 2.5066,
+      "step": 5155
+    },
+    {
+      "epoch": 1.3046558704453441,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8984480431848854e-05,
+      "loss": 2.4781,
+      "step": 5156
+    },
+    {
+      "epoch": 1.304908906882591,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.898363697705803e-05,
+      "loss": 2.5143,
+      "step": 5157
+    },
+    {
+      "epoch": 1.305161943319838,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8982793522267207e-05,
+      "loss": 2.5438,
+      "step": 5158
+    },
+    {
+      "epoch": 1.305414979757085,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8981950067476384e-05,
+      "loss": 2.4745,
+      "step": 5159
+    },
+    {
+      "epoch": 1.305668016194332,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.898110661268556e-05,
+      "loss": 2.5527,
+      "step": 5160
+    },
+    {
+      "epoch": 1.305668016194332,
+      "eval_loss": 2.6091344356536865,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 277.4697,
+      "eval_samples_per_second": 9.37,
+      "eval_steps_per_second": 9.37,
+      "step": 5160
+    },
+    {
+      "epoch": 1.305921052631579,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8980263157894737e-05,
+      "loss": 2.4781,
+      "step": 5161
+    },
+    {
+      "epoch": 1.306174089068826,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8979419703103914e-05,
+      "loss": 2.4476,
+      "step": 5162
+    },
+    {
+      "epoch": 1.3064271255060729,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.897857624831309e-05,
+      "loss": 2.4764,
+      "step": 5163
+    },
+    {
+      "epoch": 1.3066801619433197,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.897773279352227e-05,
+      "loss": 2.5581,
+      "step": 5164
+    },
+    {
+      "epoch": 1.3069331983805668,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8976889338731444e-05,
+      "loss": 2.2826,
+      "step": 5165
+    },
+    {
+      "epoch": 1.3071862348178138,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.897604588394062e-05,
+      "loss": 2.4566,
+      "step": 5166
+    },
+    {
+      "epoch": 1.3074392712550607,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.89752024291498e-05,
+      "loss": 2.4857,
+      "step": 5167
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8974358974358974e-05,
+      "loss": 2.4116,
+      "step": 5168
+    },
+    {
+      "epoch": 1.3079453441295548,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.897351551956815e-05,
+      "loss": 2.4775,
+      "step": 5169
+    },
+    {
+      "epoch": 1.3081983805668016,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.897267206477733e-05,
+      "loss": 2.4918,
+      "step": 5170
+    },
+    {
+      "epoch": 1.3084514170040487,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8971828609986507e-05,
+      "loss": 2.571,
+      "step": 5171
+    },
+    {
+      "epoch": 1.3087044534412955,
+      "grad_norm": 3.125,
+      "learning_rate": 2.897098515519568e-05,
+      "loss": 2.5362,
+      "step": 5172
+    },
+    {
+      "epoch": 1.3089574898785425,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.8970141700404857e-05,
+      "loss": 2.4339,
+      "step": 5173
+    },
+    {
+      "epoch": 1.3092105263157894,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8969298245614037e-05,
+      "loss": 2.5286,
+      "step": 5174
+    },
+    {
+      "epoch": 1.3094635627530364,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8968454790823214e-05,
+      "loss": 2.6043,
+      "step": 5175
+    },
+    {
+      "epoch": 1.3097165991902835,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8967611336032387e-05,
+      "loss": 2.544,
+      "step": 5176
+    },
+    {
+      "epoch": 1.3099696356275303,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8966767881241567e-05,
+      "loss": 2.5995,
+      "step": 5177
+    },
+    {
+      "epoch": 1.3102226720647774,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8965924426450744e-05,
+      "loss": 2.5037,
+      "step": 5178
+    },
+    {
+      "epoch": 1.3104757085020242,
+      "grad_norm": 2.75,
+      "learning_rate": 2.8965080971659917e-05,
+      "loss": 2.3728,
+      "step": 5179
+    },
+    {
+      "epoch": 1.3107287449392713,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8964237516869097e-05,
+      "loss": 2.4775,
+      "step": 5180
+    },
+    {
+      "epoch": 1.310981781376518,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8963394062078274e-05,
+      "loss": 2.4281,
+      "step": 5181
+    },
+    {
+      "epoch": 1.3112348178137652,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.896255060728745e-05,
+      "loss": 2.5036,
+      "step": 5182
+    },
+    {
+      "epoch": 1.3114878542510122,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.8961707152496627e-05,
+      "loss": 2.518,
+      "step": 5183
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.8960863697705804e-05,
+      "loss": 2.299,
+      "step": 5184
+    },
+    {
+      "epoch": 1.311993927125506,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.896002024291498e-05,
+      "loss": 2.4381,
+      "step": 5185
+    },
+    {
+      "epoch": 1.3122469635627532,
+      "grad_norm": 3.125,
+      "learning_rate": 2.895917678812416e-05,
+      "loss": 2.4535,
+      "step": 5186
+    },
+    {
+      "epoch": 1.3125,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8958333333333334e-05,
+      "loss": 2.6241,
+      "step": 5187
+    },
+    {
+      "epoch": 1.3127530364372468,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.895748987854251e-05,
+      "loss": 2.4958,
+      "step": 5188
+    },
+    {
+      "epoch": 1.313006072874494,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.895664642375169e-05,
+      "loss": 2.4129,
+      "step": 5189
+    },
+    {
+      "epoch": 1.313259109311741,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8955802968960864e-05,
+      "loss": 2.4483,
+      "step": 5190
+    },
+    {
+      "epoch": 1.313259109311741,
+      "eval_loss": 2.607276439666748,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.9134,
+      "eval_samples_per_second": 9.389,
+      "eval_steps_per_second": 9.389,
+      "step": 5190
+    },
+    {
+      "epoch": 1.3135121457489878,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.895495951417004e-05,
+      "loss": 2.331,
+      "step": 5191
+    },
+    {
+      "epoch": 1.3137651821862348,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8954116059379217e-05,
+      "loss": 2.4495,
+      "step": 5192
+    },
+    {
+      "epoch": 1.314018218623482,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8953272604588397e-05,
+      "loss": 2.5483,
+      "step": 5193
+    },
+    {
+      "epoch": 1.3142712550607287,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.895242914979757e-05,
+      "loss": 2.461,
+      "step": 5194
+    },
+    {
+      "epoch": 1.3145242914979758,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8951585695006747e-05,
+      "loss": 2.4765,
+      "step": 5195
+    },
+    {
+      "epoch": 1.3147773279352226,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8950742240215927e-05,
+      "loss": 2.5364,
+      "step": 5196
+    },
+    {
+      "epoch": 1.3150303643724697,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.89498987854251e-05,
+      "loss": 2.4148,
+      "step": 5197
+    },
+    {
+      "epoch": 1.3152834008097165,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8949055330634277e-05,
+      "loss": 2.5244,
+      "step": 5198
+    },
+    {
+      "epoch": 1.3155364372469636,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8948211875843457e-05,
+      "loss": 2.4139,
+      "step": 5199
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8947368421052634e-05,
+      "loss": 2.4937,
+      "step": 5200
+    },
+    {
+      "epoch": 1.3160425101214575,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8946524966261807e-05,
+      "loss": 2.4446,
+      "step": 5201
+    },
+    {
+      "epoch": 1.3162955465587045,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8945681511470987e-05,
+      "loss": 2.3923,
+      "step": 5202
+    },
+    {
+      "epoch": 1.3165485829959513,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8944838056680163e-05,
+      "loss": 2.5282,
+      "step": 5203
+    },
+    {
+      "epoch": 1.3168016194331984,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.894399460188934e-05,
+      "loss": 2.4256,
+      "step": 5204
+    },
+    {
+      "epoch": 1.3170546558704452,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8943151147098517e-05,
+      "loss": 2.4788,
+      "step": 5205
+    },
+    {
+      "epoch": 1.3173076923076923,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8942307692307693e-05,
+      "loss": 2.5216,
+      "step": 5206
+    },
+    {
+      "epoch": 1.3175607287449393,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.894146423751687e-05,
+      "loss": 2.4287,
+      "step": 5207
+    },
+    {
+      "epoch": 1.3178137651821862,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.8940620782726047e-05,
+      "loss": 2.4186,
+      "step": 5208
+    },
+    {
+      "epoch": 1.3180668016194332,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8939777327935223e-05,
+      "loss": 2.5179,
+      "step": 5209
+    },
+    {
+      "epoch": 1.3183198380566803,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.89389338731444e-05,
+      "loss": 2.5376,
+      "step": 5210
+    },
+    {
+      "epoch": 1.3185728744939271,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8938090418353577e-05,
+      "loss": 2.4771,
+      "step": 5211
+    },
+    {
+      "epoch": 1.318825910931174,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8937246963562753e-05,
+      "loss": 2.4929,
+      "step": 5212
+    },
+    {
+      "epoch": 1.319078947368421,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.893640350877193e-05,
+      "loss": 2.587,
+      "step": 5213
+    },
+    {
+      "epoch": 1.319331983805668,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8935560053981107e-05,
+      "loss": 2.4872,
+      "step": 5214
+    },
+    {
+      "epoch": 1.319585020242915,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8934716599190287e-05,
+      "loss": 2.4541,
+      "step": 5215
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.893387314439946e-05,
+      "loss": 2.5216,
+      "step": 5216
+    },
+    {
+      "epoch": 1.320091093117409,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8933029689608637e-05,
+      "loss": 2.5178,
+      "step": 5217
+    },
+    {
+      "epoch": 1.3203441295546559,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8932186234817817e-05,
+      "loss": 2.5427,
+      "step": 5218
+    },
+    {
+      "epoch": 1.320597165991903,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.893134278002699e-05,
+      "loss": 2.4543,
+      "step": 5219
+    },
+    {
+      "epoch": 1.3208502024291497,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8930499325236167e-05,
+      "loss": 2.5977,
+      "step": 5220
+    },
+    {
+      "epoch": 1.3208502024291497,
+      "eval_loss": 2.6040070056915283,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.0435,
+      "eval_samples_per_second": 9.419,
+      "eval_steps_per_second": 9.419,
+      "step": 5220
+    },
+    {
+      "epoch": 1.3211032388663968,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.8929655870445347e-05,
+      "loss": 2.4637,
+      "step": 5221
+    },
+    {
+      "epoch": 1.3213562753036436,
+      "grad_norm": 2.75,
+      "learning_rate": 2.8928812415654523e-05,
+      "loss": 2.3994,
+      "step": 5222
+    },
+    {
+      "epoch": 1.3216093117408907,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8927968960863697e-05,
+      "loss": 2.4785,
+      "step": 5223
+    },
+    {
+      "epoch": 1.3218623481781377,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8927125506072877e-05,
+      "loss": 2.5264,
+      "step": 5224
+    },
+    {
+      "epoch": 1.3221153846153846,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8926282051282053e-05,
+      "loss": 2.5146,
+      "step": 5225
+    },
+    {
+      "epoch": 1.3223684210526316,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.892543859649123e-05,
+      "loss": 2.4746,
+      "step": 5226
+    },
+    {
+      "epoch": 1.3226214574898785,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8924595141700403e-05,
+      "loss": 2.3457,
+      "step": 5227
+    },
+    {
+      "epoch": 1.3228744939271255,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8923751686909583e-05,
+      "loss": 2.5484,
+      "step": 5228
+    },
+    {
+      "epoch": 1.3231275303643724,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.892290823211876e-05,
+      "loss": 2.5869,
+      "step": 5229
+    },
+    {
+      "epoch": 1.3233805668016194,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8922064777327933e-05,
+      "loss": 2.4683,
+      "step": 5230
+    },
+    {
+      "epoch": 1.3236336032388665,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8921221322537113e-05,
+      "loss": 2.472,
+      "step": 5231
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.892037786774629e-05,
+      "loss": 2.5743,
+      "step": 5232
+    },
+    {
+      "epoch": 1.3241396761133604,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8919534412955466e-05,
+      "loss": 2.541,
+      "step": 5233
+    },
+    {
+      "epoch": 1.3243927125506074,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8918690958164643e-05,
+      "loss": 2.4916,
+      "step": 5234
+    },
+    {
+      "epoch": 1.3246457489878543,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.891784750337382e-05,
+      "loss": 2.4212,
+      "step": 5235
+    },
+    {
+      "epoch": 1.324898785425101,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8917004048582996e-05,
+      "loss": 2.5816,
+      "step": 5236
+    },
+    {
+      "epoch": 1.3251518218623481,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8916160593792173e-05,
+      "loss": 2.4253,
+      "step": 5237
+    },
+    {
+      "epoch": 1.3254048582995952,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.891531713900135e-05,
+      "loss": 2.4702,
+      "step": 5238
+    },
+    {
+      "epoch": 1.325657894736842,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.8914473684210526e-05,
+      "loss": 2.6239,
+      "step": 5239
+    },
+    {
+      "epoch": 1.325910931174089,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8913630229419706e-05,
+      "loss": 2.3834,
+      "step": 5240
+    },
+    {
+      "epoch": 1.3261639676113361,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.891278677462888e-05,
+      "loss": 2.4737,
+      "step": 5241
+    },
+    {
+      "epoch": 1.326417004048583,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8911943319838056e-05,
+      "loss": 2.5732,
+      "step": 5242
+    },
+    {
+      "epoch": 1.32667004048583,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8911099865047236e-05,
+      "loss": 2.4692,
+      "step": 5243
+    },
+    {
+      "epoch": 1.3269230769230769,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8910256410256413e-05,
+      "loss": 2.5309,
+      "step": 5244
+    },
+    {
+      "epoch": 1.327176113360324,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8909412955465586e-05,
+      "loss": 2.3571,
+      "step": 5245
+    },
+    {
+      "epoch": 1.3274291497975708,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8908569500674763e-05,
+      "loss": 2.2871,
+      "step": 5246
+    },
+    {
+      "epoch": 1.3276821862348178,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8907726045883943e-05,
+      "loss": 2.4866,
+      "step": 5247
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8906882591093116e-05,
+      "loss": 2.3756,
+      "step": 5248
+    },
+    {
+      "epoch": 1.3281882591093117,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8906039136302293e-05,
+      "loss": 2.5011,
+      "step": 5249
+    },
+    {
+      "epoch": 1.3284412955465588,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8905195681511473e-05,
+      "loss": 2.5467,
+      "step": 5250
+    },
+    {
+      "epoch": 1.3284412955465588,
+      "eval_loss": 2.6036531925201416,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.3837,
+      "eval_samples_per_second": 9.407,
+      "eval_steps_per_second": 9.407,
+      "step": 5250
+    },
+    {
+      "epoch": 1.3286943319838056,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.890435222672065e-05,
+      "loss": 2.5008,
+      "step": 5251
+    },
+    {
+      "epoch": 1.3289473684210527,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8903508771929823e-05,
+      "loss": 2.3344,
+      "step": 5252
+    },
+    {
+      "epoch": 1.3292004048582995,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8902665317139003e-05,
+      "loss": 2.3392,
+      "step": 5253
+    },
+    {
+      "epoch": 1.3294534412955465,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.890182186234818e-05,
+      "loss": 2.4704,
+      "step": 5254
+    },
+    {
+      "epoch": 1.3297064777327936,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8900978407557356e-05,
+      "loss": 2.5606,
+      "step": 5255
+    },
+    {
+      "epoch": 1.3299595141700404,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8900134952766533e-05,
+      "loss": 2.5752,
+      "step": 5256
+    },
+    {
+      "epoch": 1.3302125506072875,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.889929149797571e-05,
+      "loss": 2.4954,
+      "step": 5257
+    },
+    {
+      "epoch": 1.3304655870445345,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8898448043184886e-05,
+      "loss": 2.4443,
+      "step": 5258
+    },
+    {
+      "epoch": 1.3307186234817814,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8897604588394063e-05,
+      "loss": 2.5357,
+      "step": 5259
+    },
+    {
+      "epoch": 1.3309716599190282,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.889676113360324e-05,
+      "loss": 2.493,
+      "step": 5260
+    },
+    {
+      "epoch": 1.3312246963562753,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8895917678812416e-05,
+      "loss": 2.5251,
+      "step": 5261
+    },
+    {
+      "epoch": 1.3314777327935223,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8895074224021596e-05,
+      "loss": 2.4526,
+      "step": 5262
+    },
+    {
+      "epoch": 1.3317307692307692,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.889423076923077e-05,
+      "loss": 2.452,
+      "step": 5263
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8893387314439946e-05,
+      "loss": 2.5161,
+      "step": 5264
+    },
+    {
+      "epoch": 1.3322368421052633,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.8892543859649123e-05,
+      "loss": 2.2828,
+      "step": 5265
+    },
+    {
+      "epoch": 1.33248987854251,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8891700404858303e-05,
+      "loss": 2.3777,
+      "step": 5266
+    },
+    {
+      "epoch": 1.3327429149797572,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8890856950067476e-05,
+      "loss": 2.4705,
+      "step": 5267
+    },
+    {
+      "epoch": 1.332995951417004,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8890013495276653e-05,
+      "loss": 2.4815,
+      "step": 5268
+    },
+    {
+      "epoch": 1.333248987854251,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8889170040485833e-05,
+      "loss": 2.5513,
+      "step": 5269
+    },
+    {
+      "epoch": 1.333502024291498,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8888326585695006e-05,
+      "loss": 2.4578,
+      "step": 5270
+    },
+    {
+      "epoch": 1.333755060728745,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8887483130904183e-05,
+      "loss": 2.5667,
+      "step": 5271
+    },
+    {
+      "epoch": 1.334008097165992,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8886639676113363e-05,
+      "loss": 2.5211,
+      "step": 5272
+    },
+    {
+      "epoch": 1.3342611336032388,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.888579622132254e-05,
+      "loss": 2.4533,
+      "step": 5273
+    },
+    {
+      "epoch": 1.334514170040486,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8884952766531713e-05,
+      "loss": 2.4491,
+      "step": 5274
+    },
+    {
+      "epoch": 1.3347672064777327,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8884109311740893e-05,
+      "loss": 2.519,
+      "step": 5275
+    },
+    {
+      "epoch": 1.3350202429149798,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.888326585695007e-05,
+      "loss": 2.4107,
+      "step": 5276
+    },
+    {
+      "epoch": 1.3352732793522266,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8882422402159243e-05,
+      "loss": 2.3278,
+      "step": 5277
+    },
+    {
+      "epoch": 1.3355263157894737,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8881578947368423e-05,
+      "loss": 2.5678,
+      "step": 5278
+    },
+    {
+      "epoch": 1.3357793522267207,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.88807354925776e-05,
+      "loss": 2.3971,
+      "step": 5279
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8879892037786776e-05,
+      "loss": 2.4324,
+      "step": 5280
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "eval_loss": 2.6017208099365234,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.9209,
+      "eval_samples_per_second": 9.423,
+      "eval_steps_per_second": 9.423,
+      "step": 5280
+    },
+    {
+      "epoch": 1.3362854251012146,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.887904858299595e-05,
+      "loss": 2.5392,
+      "step": 5281
+    },
+    {
+      "epoch": 1.3365384615384617,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.887820512820513e-05,
+      "loss": 2.488,
+      "step": 5282
+    },
+    {
+      "epoch": 1.3367914979757085,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8877361673414306e-05,
+      "loss": 2.45,
+      "step": 5283
+    },
+    {
+      "epoch": 1.3370445344129553,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8876518218623483e-05,
+      "loss": 2.4638,
+      "step": 5284
+    },
+    {
+      "epoch": 1.3372975708502024,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.887567476383266e-05,
+      "loss": 2.4603,
+      "step": 5285
+    },
+    {
+      "epoch": 1.3375506072874495,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8874831309041836e-05,
+      "loss": 2.4521,
+      "step": 5286
+    },
+    {
+      "epoch": 1.3378036437246963,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8873987854251013e-05,
+      "loss": 2.5006,
+      "step": 5287
+    },
+    {
+      "epoch": 1.3380566801619433,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.887314439946019e-05,
+      "loss": 2.3839,
+      "step": 5288
+    },
+    {
+      "epoch": 1.3383097165991904,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8872300944669366e-05,
+      "loss": 2.5188,
+      "step": 5289
+    },
+    {
+      "epoch": 1.3385627530364372,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8871457489878542e-05,
+      "loss": 2.4151,
+      "step": 5290
+    },
+    {
+      "epoch": 1.3388157894736843,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8870614035087723e-05,
+      "loss": 2.4392,
+      "step": 5291
+    },
+    {
+      "epoch": 1.3390688259109311,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8869770580296896e-05,
+      "loss": 2.5545,
+      "step": 5292
+    },
+    {
+      "epoch": 1.3393218623481782,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8868927125506072e-05,
+      "loss": 2.4866,
+      "step": 5293
+    },
+    {
+      "epoch": 1.339574898785425,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8868083670715252e-05,
+      "loss": 2.4128,
+      "step": 5294
+    },
+    {
+      "epoch": 1.339827935222672,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.886724021592443e-05,
+      "loss": 2.5153,
+      "step": 5295
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8866396761133602e-05,
+      "loss": 2.4854,
+      "step": 5296
+    },
+    {
+      "epoch": 1.340334008097166,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8865553306342782e-05,
+      "loss": 2.5195,
+      "step": 5297
+    },
+    {
+      "epoch": 1.340587044534413,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.886470985155196e-05,
+      "loss": 2.5326,
+      "step": 5298
+    },
+    {
+      "epoch": 1.3408400809716599,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8863866396761132e-05,
+      "loss": 2.4863,
+      "step": 5299
+    },
+    {
+      "epoch": 1.341093117408907,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.886302294197031e-05,
+      "loss": 2.4308,
+      "step": 5300
+    },
+    {
+      "epoch": 1.3413461538461537,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.886217948717949e-05,
+      "loss": 2.5052,
+      "step": 5301
+    },
+    {
+      "epoch": 1.3415991902834008,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8861336032388666e-05,
+      "loss": 2.5191,
+      "step": 5302
+    },
+    {
+      "epoch": 1.3418522267206479,
+      "grad_norm": 3.625,
+      "learning_rate": 2.886049257759784e-05,
+      "loss": 2.5327,
+      "step": 5303
+    },
+    {
+      "epoch": 1.3421052631578947,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.885964912280702e-05,
+      "loss": 2.5112,
+      "step": 5304
+    },
+    {
+      "epoch": 1.3423582995951417,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8858805668016196e-05,
+      "loss": 2.4805,
+      "step": 5305
+    },
+    {
+      "epoch": 1.3426113360323888,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8857962213225372e-05,
+      "loss": 2.5899,
+      "step": 5306
+    },
+    {
+      "epoch": 1.3428643724696356,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.885711875843455e-05,
+      "loss": 2.5251,
+      "step": 5307
+    },
+    {
+      "epoch": 1.3431174089068825,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8856275303643726e-05,
+      "loss": 2.4962,
+      "step": 5308
+    },
+    {
+      "epoch": 1.3433704453441295,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8855431848852902e-05,
+      "loss": 2.47,
+      "step": 5309
+    },
+    {
+      "epoch": 1.3436234817813766,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.885458839406208e-05,
+      "loss": 2.5015,
+      "step": 5310
+    },
+    {
+      "epoch": 1.3436234817813766,
+      "eval_loss": 2.604715347290039,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.6453,
+      "eval_samples_per_second": 9.398,
+      "eval_steps_per_second": 9.398,
+      "step": 5310
+    },
+    {
+      "epoch": 1.3438765182186234,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8853744939271256e-05,
+      "loss": 2.3988,
+      "step": 5311
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8852901484480432e-05,
+      "loss": 2.5435,
+      "step": 5312
+    },
+    {
+      "epoch": 1.3443825910931175,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8852058029689612e-05,
+      "loss": 2.495,
+      "step": 5313
+    },
+    {
+      "epoch": 1.3446356275303644,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8851214574898786e-05,
+      "loss": 2.4148,
+      "step": 5314
+    },
+    {
+      "epoch": 1.3448886639676112,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8850371120107962e-05,
+      "loss": 2.4684,
+      "step": 5315
+    },
+    {
+      "epoch": 1.3451417004048583,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8849527665317142e-05,
+      "loss": 2.555,
+      "step": 5316
+    },
+    {
+      "epoch": 1.3453947368421053,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.8848684210526315e-05,
+      "loss": 2.474,
+      "step": 5317
+    },
+    {
+      "epoch": 1.3456477732793521,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8847840755735492e-05,
+      "loss": 2.5851,
+      "step": 5318
+    },
+    {
+      "epoch": 1.3459008097165992,
+      "grad_norm": 3.25,
+      "learning_rate": 2.884699730094467e-05,
+      "loss": 2.5288,
+      "step": 5319
+    },
+    {
+      "epoch": 1.3461538461538463,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.884615384615385e-05,
+      "loss": 2.4822,
+      "step": 5320
+    },
+    {
+      "epoch": 1.346406882591093,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8845310391363022e-05,
+      "loss": 2.4674,
+      "step": 5321
+    },
+    {
+      "epoch": 1.3466599190283401,
+      "grad_norm": 3.0,
+      "learning_rate": 2.88444669365722e-05,
+      "loss": 2.453,
+      "step": 5322
+    },
+    {
+      "epoch": 1.346912955465587,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.884362348178138e-05,
+      "loss": 2.3558,
+      "step": 5323
+    },
+    {
+      "epoch": 1.347165991902834,
+      "grad_norm": 2.625,
+      "learning_rate": 2.8842780026990555e-05,
+      "loss": 2.4019,
+      "step": 5324
+    },
+    {
+      "epoch": 1.3474190283400809,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.884193657219973e-05,
+      "loss": 2.5411,
+      "step": 5325
+    },
+    {
+      "epoch": 1.347672064777328,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.884109311740891e-05,
+      "loss": 2.2658,
+      "step": 5326
+    },
+    {
+      "epoch": 1.347925101214575,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8840249662618085e-05,
+      "loss": 2.3355,
+      "step": 5327
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.883940620782726e-05,
+      "loss": 2.373,
+      "step": 5328
+    },
+    {
+      "epoch": 1.3484311740890689,
+      "grad_norm": 3.0,
+      "learning_rate": 2.883856275303644e-05,
+      "loss": 2.4526,
+      "step": 5329
+    },
+    {
+      "epoch": 1.3486842105263157,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8837719298245615e-05,
+      "loss": 2.542,
+      "step": 5330
+    },
+    {
+      "epoch": 1.3489372469635628,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8836875843454792e-05,
+      "loss": 2.5486,
+      "step": 5331
+    },
+    {
+      "epoch": 1.3491902834008096,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.883603238866397e-05,
+      "loss": 2.5153,
+      "step": 5332
+    },
+    {
+      "epoch": 1.3494433198380567,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8835188933873145e-05,
+      "loss": 2.4512,
+      "step": 5333
+    },
+    {
+      "epoch": 1.3496963562753037,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8834345479082322e-05,
+      "loss": 2.5022,
+      "step": 5334
+    },
+    {
+      "epoch": 1.3499493927125505,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.88335020242915e-05,
+      "loss": 2.5651,
+      "step": 5335
+    },
+    {
+      "epoch": 1.3502024291497976,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8832658569500675e-05,
+      "loss": 2.5078,
+      "step": 5336
+    },
+    {
+      "epoch": 1.3504554655870447,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8831815114709852e-05,
+      "loss": 2.4202,
+      "step": 5337
+    },
+    {
+      "epoch": 1.3507085020242915,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.883097165991903e-05,
+      "loss": 2.418,
+      "step": 5338
+    },
+    {
+      "epoch": 1.3509615384615383,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8830128205128205e-05,
+      "loss": 2.5966,
+      "step": 5339
+    },
+    {
+      "epoch": 1.3512145748987854,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8829284750337382e-05,
+      "loss": 2.5753,
+      "step": 5340
+    },
+    {
+      "epoch": 1.3512145748987854,
+      "eval_loss": 2.598905563354492,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.0104,
+      "eval_samples_per_second": 9.42,
+      "eval_steps_per_second": 9.42,
+      "step": 5340
+    },
+    {
+      "epoch": 1.3514676113360324,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.882844129554656e-05,
+      "loss": 2.557,
+      "step": 5341
+    },
+    {
+      "epoch": 1.3517206477732793,
+      "grad_norm": 3.0,
+      "learning_rate": 2.882759784075574e-05,
+      "loss": 2.6223,
+      "step": 5342
+    },
+    {
+      "epoch": 1.3519736842105263,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8826754385964912e-05,
+      "loss": 2.5377,
+      "step": 5343
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 3.0,
+      "learning_rate": 2.882591093117409e-05,
+      "loss": 2.6035,
+      "step": 5344
+    },
+    {
+      "epoch": 1.3524797570850202,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.882506747638327e-05,
+      "loss": 2.442,
+      "step": 5345
+    },
+    {
+      "epoch": 1.3527327935222673,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8824224021592445e-05,
+      "loss": 2.5151,
+      "step": 5346
+    },
+    {
+      "epoch": 1.352985829959514,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.882338056680162e-05,
+      "loss": 2.4624,
+      "step": 5347
+    },
+    {
+      "epoch": 1.3532388663967612,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.88225371120108e-05,
+      "loss": 2.4557,
+      "step": 5348
+    },
+    {
+      "epoch": 1.353491902834008,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8821693657219975e-05,
+      "loss": 2.5884,
+      "step": 5349
+    },
+    {
+      "epoch": 1.353744939271255,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.882085020242915e-05,
+      "loss": 2.4372,
+      "step": 5350
+    },
+    {
+      "epoch": 1.353997975708502,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.882000674763833e-05,
+      "loss": 2.4146,
+      "step": 5351
+    },
+    {
+      "epoch": 1.354251012145749,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.8819163292847505e-05,
+      "loss": 2.4756,
+      "step": 5352
+    },
+    {
+      "epoch": 1.354504048582996,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8818319838056682e-05,
+      "loss": 2.4873,
+      "step": 5353
+    },
+    {
+      "epoch": 1.3547570850202428,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8817476383265855e-05,
+      "loss": 2.542,
+      "step": 5354
+    },
+    {
+      "epoch": 1.35501012145749,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.8816632928475035e-05,
+      "loss": 2.4992,
+      "step": 5355
+    },
+    {
+      "epoch": 1.3552631578947367,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8815789473684212e-05,
+      "loss": 2.5456,
+      "step": 5356
+    },
+    {
+      "epoch": 1.3555161943319838,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8814946018893385e-05,
+      "loss": 2.4234,
+      "step": 5357
+    },
+    {
+      "epoch": 1.3557692307692308,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8814102564102565e-05,
+      "loss": 2.6009,
+      "step": 5358
+    },
+    {
+      "epoch": 1.3560222672064777,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.8813259109311742e-05,
+      "loss": 2.441,
+      "step": 5359
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.881241565452092e-05,
+      "loss": 2.4527,
+      "step": 5360
+    },
+    {
+      "epoch": 1.3565283400809718,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8811572199730095e-05,
+      "loss": 2.5149,
+      "step": 5361
+    },
+    {
+      "epoch": 1.3567813765182186,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.881072874493927e-05,
+      "loss": 2.5415,
+      "step": 5362
+    },
+    {
+      "epoch": 1.3570344129554655,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.880988529014845e-05,
+      "loss": 2.5225,
+      "step": 5363
+    },
+    {
+      "epoch": 1.3572874493927125,
+      "grad_norm": 3.125,
+      "learning_rate": 2.880904183535763e-05,
+      "loss": 2.5529,
+      "step": 5364
+    },
+    {
+      "epoch": 1.3575404858299596,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.88081983805668e-05,
+      "loss": 2.5951,
+      "step": 5365
+    },
+    {
+      "epoch": 1.3577935222672064,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8807354925775978e-05,
+      "loss": 2.582,
+      "step": 5366
+    },
+    {
+      "epoch": 1.3580465587044535,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.880651147098516e-05,
+      "loss": 2.5186,
+      "step": 5367
+    },
+    {
+      "epoch": 1.3582995951417005,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.880566801619433e-05,
+      "loss": 2.4222,
+      "step": 5368
+    },
+    {
+      "epoch": 1.3585526315789473,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8804824561403508e-05,
+      "loss": 2.4781,
+      "step": 5369
+    },
+    {
+      "epoch": 1.3588056680161944,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8803981106612688e-05,
+      "loss": 2.5274,
+      "step": 5370
+    },
+    {
+      "epoch": 1.3588056680161944,
+      "eval_loss": 2.5972201824188232,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.922,
+      "eval_samples_per_second": 9.423,
+      "eval_steps_per_second": 9.423,
+      "step": 5370
+    },
+    {
+      "epoch": 1.3590587044534412,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8803137651821865e-05,
+      "loss": 2.438,
+      "step": 5371
+    },
+    {
+      "epoch": 1.3593117408906883,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8802294197031038e-05,
+      "loss": 2.5218,
+      "step": 5372
+    },
+    {
+      "epoch": 1.3595647773279351,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.8801450742240215e-05,
+      "loss": 2.3056,
+      "step": 5373
+    },
+    {
+      "epoch": 1.3598178137651822,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8800607287449395e-05,
+      "loss": 2.5347,
+      "step": 5374
+    },
+    {
+      "epoch": 1.3600708502024292,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.879976383265857e-05,
+      "loss": 2.4632,
+      "step": 5375
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8798920377867745e-05,
+      "loss": 2.5204,
+      "step": 5376
+    },
+    {
+      "epoch": 1.3605769230769231,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8798076923076925e-05,
+      "loss": 2.3834,
+      "step": 5377
+    },
+    {
+      "epoch": 1.36082995951417,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.87972334682861e-05,
+      "loss": 2.5949,
+      "step": 5378
+    },
+    {
+      "epoch": 1.361082995951417,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8796390013495275e-05,
+      "loss": 2.5902,
+      "step": 5379
+    },
+    {
+      "epoch": 1.3613360323886639,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8795546558704455e-05,
+      "loss": 2.4576,
+      "step": 5380
+    },
+    {
+      "epoch": 1.361589068825911,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.879470310391363e-05,
+      "loss": 2.5014,
+      "step": 5381
+    },
+    {
+      "epoch": 1.361842105263158,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8793859649122808e-05,
+      "loss": 2.4597,
+      "step": 5382
+    },
+    {
+      "epoch": 1.3620951417004048,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8793016194331985e-05,
+      "loss": 2.415,
+      "step": 5383
+    },
+    {
+      "epoch": 1.3623481781376519,
+      "grad_norm": 3.0,
+      "learning_rate": 2.879217273954116e-05,
+      "loss": 2.4942,
+      "step": 5384
+    },
+    {
+      "epoch": 1.362601214574899,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8791329284750338e-05,
+      "loss": 2.6306,
+      "step": 5385
+    },
+    {
+      "epoch": 1.3628542510121457,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8790485829959515e-05,
+      "loss": 2.4332,
+      "step": 5386
+    },
+    {
+      "epoch": 1.3631072874493926,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.878964237516869e-05,
+      "loss": 2.4397,
+      "step": 5387
+    },
+    {
+      "epoch": 1.3633603238866396,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8788798920377868e-05,
+      "loss": 2.5063,
+      "step": 5388
+    },
+    {
+      "epoch": 1.3636133603238867,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8787955465587045e-05,
+      "loss": 2.5059,
+      "step": 5389
+    },
+    {
+      "epoch": 1.3638663967611335,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.878711201079622e-05,
+      "loss": 2.6009,
+      "step": 5390
+    },
+    {
+      "epoch": 1.3641194331983806,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.8786268556005398e-05,
+      "loss": 2.4904,
+      "step": 5391
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8785425101214575e-05,
+      "loss": 2.5084,
+      "step": 5392
+    },
+    {
+      "epoch": 1.3646255060728745,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8784581646423755e-05,
+      "loss": 2.4875,
+      "step": 5393
+    },
+    {
+      "epoch": 1.3648785425101215,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8783738191632928e-05,
+      "loss": 2.5669,
+      "step": 5394
+    },
+    {
+      "epoch": 1.3651315789473684,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8782894736842105e-05,
+      "loss": 2.5612,
+      "step": 5395
+    },
+    {
+      "epoch": 1.3653846153846154,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8782051282051285e-05,
+      "loss": 2.5426,
+      "step": 5396
+    },
+    {
+      "epoch": 1.3656376518218623,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8781207827260458e-05,
+      "loss": 2.4916,
+      "step": 5397
+    },
+    {
+      "epoch": 1.3658906882591093,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8780364372469635e-05,
+      "loss": 2.4726,
+      "step": 5398
+    },
+    {
+      "epoch": 1.3661437246963564,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.8779520917678815e-05,
+      "loss": 2.5317,
+      "step": 5399
+    },
+    {
+      "epoch": 1.3663967611336032,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.877867746288799e-05,
+      "loss": 2.4096,
+      "step": 5400
+    },
+    {
+      "epoch": 1.3663967611336032,
+      "eval_loss": 2.6006250381469727,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.4153,
+      "eval_samples_per_second": 9.406,
+      "eval_steps_per_second": 9.406,
+      "step": 5400
+    },
+    {
+      "epoch": 1.3666497975708503,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8777834008097165e-05,
+      "loss": 2.48,
+      "step": 5401
+    },
+    {
+      "epoch": 1.366902834008097,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8776990553306345e-05,
+      "loss": 2.5046,
+      "step": 5402
+    },
+    {
+      "epoch": 1.3671558704453441,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.877614709851552e-05,
+      "loss": 2.4094,
+      "step": 5403
+    },
+    {
+      "epoch": 1.367408906882591,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8775303643724698e-05,
+      "loss": 2.3974,
+      "step": 5404
+    },
+    {
+      "epoch": 1.367661943319838,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8774460188933875e-05,
+      "loss": 2.3998,
+      "step": 5405
+    },
+    {
+      "epoch": 1.367914979757085,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.877361673414305e-05,
+      "loss": 2.4437,
+      "step": 5406
+    },
+    {
+      "epoch": 1.368168016194332,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8772773279352228e-05,
+      "loss": 2.4629,
+      "step": 5407
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 3.125,
+      "learning_rate": 2.87719298245614e-05,
+      "loss": 2.5127,
+      "step": 5408
+    },
+    {
+      "epoch": 1.368674089068826,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.877108636977058e-05,
+      "loss": 2.3635,
+      "step": 5409
+    },
+    {
+      "epoch": 1.3689271255060729,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8770242914979758e-05,
+      "loss": 2.5198,
+      "step": 5410
+    },
+    {
+      "epoch": 1.3691801619433197,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8769399460188934e-05,
+      "loss": 2.5164,
+      "step": 5411
+    },
+    {
+      "epoch": 1.3694331983805668,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.876855600539811e-05,
+      "loss": 2.5028,
+      "step": 5412
+    },
+    {
+      "epoch": 1.3696862348178138,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8767712550607288e-05,
+      "loss": 2.5436,
+      "step": 5413
+    },
+    {
+      "epoch": 1.3699392712550607,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8766869095816464e-05,
+      "loss": 2.52,
+      "step": 5414
+    },
+    {
+      "epoch": 1.3701923076923077,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.8766025641025644e-05,
+      "loss": 2.4749,
+      "step": 5415
+    },
+    {
+      "epoch": 1.3704453441295548,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8765182186234818e-05,
+      "loss": 2.4232,
+      "step": 5416
+    },
+    {
+      "epoch": 1.3706983805668016,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8764338731443994e-05,
+      "loss": 2.5144,
+      "step": 5417
+    },
+    {
+      "epoch": 1.3709514170040487,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8763495276653174e-05,
+      "loss": 2.5424,
+      "step": 5418
+    },
+    {
+      "epoch": 1.3712044534412955,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8762651821862348e-05,
+      "loss": 2.5134,
+      "step": 5419
+    },
+    {
+      "epoch": 1.3714574898785425,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8761808367071524e-05,
+      "loss": 2.4333,
+      "step": 5420
+    },
+    {
+      "epoch": 1.3717105263157894,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8760964912280704e-05,
+      "loss": 2.3949,
+      "step": 5421
+    },
+    {
+      "epoch": 1.3719635627530364,
+      "grad_norm": 35.0,
+      "learning_rate": 2.876012145748988e-05,
+      "loss": 2.6663,
+      "step": 5422
+    },
+    {
+      "epoch": 1.3722165991902835,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8759278002699054e-05,
+      "loss": 2.4854,
+      "step": 5423
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8758434547908234e-05,
+      "loss": 2.4491,
+      "step": 5424
+    },
+    {
+      "epoch": 1.3727226720647774,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.875759109311741e-05,
+      "loss": 2.5484,
+      "step": 5425
+    },
+    {
+      "epoch": 1.3729757085020242,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8756747638326584e-05,
+      "loss": 2.6163,
+      "step": 5426
+    },
+    {
+      "epoch": 1.3732287449392713,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.875590418353576e-05,
+      "loss": 2.5137,
+      "step": 5427
+    },
+    {
+      "epoch": 1.373481781376518,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.875506072874494e-05,
+      "loss": 2.4943,
+      "step": 5428
+    },
+    {
+      "epoch": 1.3737348178137652,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8754217273954118e-05,
+      "loss": 2.3767,
+      "step": 5429
+    },
+    {
+      "epoch": 1.3739878542510122,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.875337381916329e-05,
+      "loss": 2.4915,
+      "step": 5430
+    },
+    {
+      "epoch": 1.3739878542510122,
+      "eval_loss": 2.5967421531677246,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.0242,
+      "eval_samples_per_second": 9.419,
+      "eval_steps_per_second": 9.419,
+      "step": 5430
+    },
+    {
+      "epoch": 1.374240890688259,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.875253036437247e-05,
+      "loss": 2.4363,
+      "step": 5431
+    },
+    {
+      "epoch": 1.374493927125506,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8751686909581648e-05,
+      "loss": 2.4453,
+      "step": 5432
+    },
+    {
+      "epoch": 1.3747469635627532,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8750843454790824e-05,
+      "loss": 2.4872,
+      "step": 5433
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.875e-05,
+      "loss": 2.4415,
+      "step": 5434
+    },
+    {
+      "epoch": 1.3752530364372468,
+      "grad_norm": 19.25,
+      "learning_rate": 2.8749156545209178e-05,
+      "loss": 2.5562,
+      "step": 5435
+    },
+    {
+      "epoch": 1.375506072874494,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.8748313090418354e-05,
+      "loss": 2.5642,
+      "step": 5436
+    },
+    {
+      "epoch": 1.375759109311741,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.874746963562753e-05,
+      "loss": 2.5235,
+      "step": 5437
+    },
+    {
+      "epoch": 1.3760121457489878,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8746626180836707e-05,
+      "loss": 2.5052,
+      "step": 5438
+    },
+    {
+      "epoch": 1.3762651821862348,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.8745782726045884e-05,
+      "loss": 2.499,
+      "step": 5439
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8744939271255064e-05,
+      "loss": 2.45,
+      "step": 5440
+    },
+    {
+      "epoch": 1.3767712550607287,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8744095816464237e-05,
+      "loss": 2.4619,
+      "step": 5441
+    },
+    {
+      "epoch": 1.3770242914979758,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8743252361673414e-05,
+      "loss": 2.5002,
+      "step": 5442
+    },
+    {
+      "epoch": 1.3772773279352226,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8742408906882594e-05,
+      "loss": 2.5265,
+      "step": 5443
+    },
+    {
+      "epoch": 1.3775303643724697,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.874156545209177e-05,
+      "loss": 2.4995,
+      "step": 5444
+    },
+    {
+      "epoch": 1.3777834008097165,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8740721997300944e-05,
+      "loss": 2.4264,
+      "step": 5445
+    },
+    {
+      "epoch": 1.3780364372469636,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.873987854251012e-05,
+      "loss": 2.3486,
+      "step": 5446
+    },
+    {
+      "epoch": 1.3782894736842106,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.87390350877193e-05,
+      "loss": 2.4857,
+      "step": 5447
+    },
+    {
+      "epoch": 1.3785425101214575,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8738191632928474e-05,
+      "loss": 2.4648,
+      "step": 5448
+    },
+    {
+      "epoch": 1.3787955465587045,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.873734817813765e-05,
+      "loss": 2.5463,
+      "step": 5449
+    },
+    {
+      "epoch": 1.3790485829959513,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.873650472334683e-05,
+      "loss": 2.4588,
+      "step": 5450
+    },
+    {
+      "epoch": 1.3793016194331984,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8735661268556007e-05,
+      "loss": 2.4139,
+      "step": 5451
+    },
+    {
+      "epoch": 1.3795546558704452,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.873481781376518e-05,
+      "loss": 2.547,
+      "step": 5452
+    },
+    {
+      "epoch": 1.3798076923076923,
+      "grad_norm": 3.125,
+      "learning_rate": 2.873397435897436e-05,
+      "loss": 2.5642,
+      "step": 5453
+    },
+    {
+      "epoch": 1.3800607287449393,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8733130904183537e-05,
+      "loss": 2.3457,
+      "step": 5454
+    },
+    {
+      "epoch": 1.3803137651821862,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8732287449392714e-05,
+      "loss": 2.4699,
+      "step": 5455
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.873144399460189e-05,
+      "loss": 2.5593,
+      "step": 5456
+    },
+    {
+      "epoch": 1.3808198380566803,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.8730600539811067e-05,
+      "loss": 2.4777,
+      "step": 5457
+    },
+    {
+      "epoch": 1.3810728744939271,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8729757085020244e-05,
+      "loss": 2.4499,
+      "step": 5458
+    },
+    {
+      "epoch": 1.381325910931174,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.872891363022942e-05,
+      "loss": 2.4818,
+      "step": 5459
+    },
+    {
+      "epoch": 1.381578947368421,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.8728070175438597e-05,
+      "loss": 2.5454,
+      "step": 5460
+    },
+    {
+      "epoch": 1.381578947368421,
+      "eval_loss": 2.5943753719329834,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.3521,
+      "eval_samples_per_second": 9.442,
+      "eval_steps_per_second": 9.442,
+      "step": 5460
+    },
+    {
+      "epoch": 1.381831983805668,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8727226720647774e-05,
+      "loss": 2.4586,
+      "step": 5461
+    },
+    {
+      "epoch": 1.382085020242915,
+      "grad_norm": 2.875,
+      "learning_rate": 2.872638326585695e-05,
+      "loss": 2.5091,
+      "step": 5462
+    },
+    {
+      "epoch": 1.382338056680162,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8725539811066127e-05,
+      "loss": 2.4048,
+      "step": 5463
+    },
+    {
+      "epoch": 1.382591093117409,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8724696356275304e-05,
+      "loss": 2.4671,
+      "step": 5464
+    },
+    {
+      "epoch": 1.3828441295546559,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.872385290148448e-05,
+      "loss": 2.4563,
+      "step": 5465
+    },
+    {
+      "epoch": 1.383097165991903,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8723009446693657e-05,
+      "loss": 2.4722,
+      "step": 5466
+    },
+    {
+      "epoch": 1.3833502024291497,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8722165991902834e-05,
+      "loss": 2.434,
+      "step": 5467
+    },
+    {
+      "epoch": 1.3836032388663968,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.872132253711201e-05,
+      "loss": 2.464,
+      "step": 5468
+    },
+    {
+      "epoch": 1.3838562753036436,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.872047908232119e-05,
+      "loss": 2.5098,
+      "step": 5469
+    },
+    {
+      "epoch": 1.3841093117408907,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8719635627530364e-05,
+      "loss": 2.4234,
+      "step": 5470
+    },
+    {
+      "epoch": 1.3843623481781377,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.871879217273954e-05,
+      "loss": 2.456,
+      "step": 5471
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.871794871794872e-05,
+      "loss": 2.548,
+      "step": 5472
+    },
+    {
+      "epoch": 1.3848684210526316,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8717105263157897e-05,
+      "loss": 2.5367,
+      "step": 5473
+    },
+    {
+      "epoch": 1.3851214574898785,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.871626180836707e-05,
+      "loss": 2.3563,
+      "step": 5474
+    },
+    {
+      "epoch": 1.3853744939271255,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.871541835357625e-05,
+      "loss": 2.5081,
+      "step": 5475
+    },
+    {
+      "epoch": 1.3856275303643724,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8714574898785427e-05,
+      "loss": 2.5451,
+      "step": 5476
+    },
+    {
+      "epoch": 1.3858805668016194,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.87137314439946e-05,
+      "loss": 2.523,
+      "step": 5477
+    },
+    {
+      "epoch": 1.3861336032388665,
+      "grad_norm": 2.875,
+      "learning_rate": 2.871288798920378e-05,
+      "loss": 2.5588,
+      "step": 5478
+    },
+    {
+      "epoch": 1.3863866396761133,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8712044534412957e-05,
+      "loss": 2.5286,
+      "step": 5479
+    },
+    {
+      "epoch": 1.3866396761133604,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8711201079622134e-05,
+      "loss": 2.4222,
+      "step": 5480
+    },
+    {
+      "epoch": 1.3868927125506074,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8710357624831307e-05,
+      "loss": 2.5415,
+      "step": 5481
+    },
+    {
+      "epoch": 1.3871457489878543,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8709514170040487e-05,
+      "loss": 2.4949,
+      "step": 5482
+    },
+    {
+      "epoch": 1.387398785425101,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8708670715249664e-05,
+      "loss": 2.4765,
+      "step": 5483
+    },
+    {
+      "epoch": 1.3876518218623481,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.870782726045884e-05,
+      "loss": 2.467,
+      "step": 5484
+    },
+    {
+      "epoch": 1.3879048582995952,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8706983805668017e-05,
+      "loss": 2.425,
+      "step": 5485
+    },
+    {
+      "epoch": 1.388157894736842,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8706140350877194e-05,
+      "loss": 2.5104,
+      "step": 5486
+    },
+    {
+      "epoch": 1.388410931174089,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.870529689608637e-05,
+      "loss": 2.5492,
+      "step": 5487
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8704453441295547e-05,
+      "loss": 2.5458,
+      "step": 5488
+    },
+    {
+      "epoch": 1.388917004048583,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8703609986504724e-05,
+      "loss": 2.5379,
+      "step": 5489
+    },
+    {
+      "epoch": 1.38917004048583,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.87027665317139e-05,
+      "loss": 2.4935,
+      "step": 5490
+    },
+    {
+      "epoch": 1.38917004048583,
+      "eval_loss": 2.5939886569976807,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 274.9974,
+      "eval_samples_per_second": 9.455,
+      "eval_steps_per_second": 9.455,
+      "step": 5490
+    },
+    {
+      "epoch": 1.3894230769230769,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.870192307692308e-05,
+      "loss": 2.3619,
+      "step": 5491
+    },
+    {
+      "epoch": 1.389676113360324,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8701079622132254e-05,
+      "loss": 2.5322,
+      "step": 5492
+    },
+    {
+      "epoch": 1.3899291497975708,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.870023616734143e-05,
+      "loss": 2.5313,
+      "step": 5493
+    },
+    {
+      "epoch": 1.3901821862348178,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.869939271255061e-05,
+      "loss": 2.5427,
+      "step": 5494
+    },
+    {
+      "epoch": 1.3904352226720649,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8698549257759787e-05,
+      "loss": 2.4854,
+      "step": 5495
+    },
+    {
+      "epoch": 1.3906882591093117,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.869770580296896e-05,
+      "loss": 2.3899,
+      "step": 5496
+    },
+    {
+      "epoch": 1.3909412955465588,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.869686234817814e-05,
+      "loss": 2.5257,
+      "step": 5497
+    },
+    {
+      "epoch": 1.3911943319838056,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8696018893387317e-05,
+      "loss": 2.4374,
+      "step": 5498
+    },
+    {
+      "epoch": 1.3914473684210527,
+      "grad_norm": 2.875,
+      "learning_rate": 2.869517543859649e-05,
+      "loss": 2.4384,
+      "step": 5499
+    },
+    {
+      "epoch": 1.3917004048582995,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8694331983805667e-05,
+      "loss": 2.4153,
+      "step": 5500
+    },
+    {
+      "epoch": 1.3919534412955465,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8693488529014847e-05,
+      "loss": 2.5023,
+      "step": 5501
+    },
+    {
+      "epoch": 1.3922064777327936,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8692645074224023e-05,
+      "loss": 2.5475,
+      "step": 5502
+    },
+    {
+      "epoch": 1.3924595141700404,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8691801619433197e-05,
+      "loss": 2.3903,
+      "step": 5503
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8690958164642377e-05,
+      "loss": 2.3761,
+      "step": 5504
+    },
+    {
+      "epoch": 1.3929655870445345,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8690114709851553e-05,
+      "loss": 2.5176,
+      "step": 5505
+    },
+    {
+      "epoch": 1.3932186234817814,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8689271255060727e-05,
+      "loss": 2.4488,
+      "step": 5506
+    },
+    {
+      "epoch": 1.3934716599190282,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8688427800269907e-05,
+      "loss": 2.5454,
+      "step": 5507
+    },
+    {
+      "epoch": 1.3937246963562753,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8687584345479083e-05,
+      "loss": 2.3375,
+      "step": 5508
+    },
+    {
+      "epoch": 1.3939777327935223,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.868674089068826e-05,
+      "loss": 2.4985,
+      "step": 5509
+    },
+    {
+      "epoch": 1.3942307692307692,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8685897435897437e-05,
+      "loss": 2.4634,
+      "step": 5510
+    },
+    {
+      "epoch": 1.3944838056680162,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8685053981106613e-05,
+      "loss": 2.3379,
+      "step": 5511
+    },
+    {
+      "epoch": 1.3947368421052633,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.868421052631579e-05,
+      "loss": 2.4707,
+      "step": 5512
+    },
+    {
+      "epoch": 1.39498987854251,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.868336707152497e-05,
+      "loss": 2.4285,
+      "step": 5513
+    },
+    {
+      "epoch": 1.3952429149797572,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8682523616734143e-05,
+      "loss": 2.5313,
+      "step": 5514
+    },
+    {
+      "epoch": 1.395495951417004,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.868168016194332e-05,
+      "loss": 2.6056,
+      "step": 5515
+    },
+    {
+      "epoch": 1.395748987854251,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8680836707152497e-05,
+      "loss": 2.3872,
+      "step": 5516
+    },
+    {
+      "epoch": 1.396002024291498,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8679993252361673e-05,
+      "loss": 2.4876,
+      "step": 5517
+    },
+    {
+      "epoch": 1.396255060728745,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.867914979757085e-05,
+      "loss": 2.4895,
+      "step": 5518
+    },
+    {
+      "epoch": 1.396508097165992,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8678306342780027e-05,
+      "loss": 2.4456,
+      "step": 5519
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8677462887989207e-05,
+      "loss": 2.499,
+      "step": 5520
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "eval_loss": 2.5953054428100586,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 288.8827,
+      "eval_samples_per_second": 9.0,
+      "eval_steps_per_second": 9.0,
+      "step": 5520
+    },
+    {
+      "epoch": 1.397014170040486,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.867661943319838e-05,
+      "loss": 2.4475,
+      "step": 5521
+    },
+    {
+      "epoch": 1.3972672064777327,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.8675775978407557e-05,
+      "loss": 2.5202,
+      "step": 5522
+    },
+    {
+      "epoch": 1.3975202429149798,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8674932523616737e-05,
+      "loss": 2.5031,
+      "step": 5523
+    },
+    {
+      "epoch": 1.3977732793522266,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8674089068825913e-05,
+      "loss": 2.5652,
+      "step": 5524
+    },
+    {
+      "epoch": 1.3980263157894737,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8673245614035086e-05,
+      "loss": 2.4741,
+      "step": 5525
+    },
+    {
+      "epoch": 1.3982793522267207,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8672402159244267e-05,
+      "loss": 2.4385,
+      "step": 5526
+    },
+    {
+      "epoch": 1.3985323886639676,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8671558704453443e-05,
+      "loss": 2.4222,
+      "step": 5527
+    },
+    {
+      "epoch": 1.3987854251012146,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8670715249662616e-05,
+      "loss": 2.5274,
+      "step": 5528
+    },
+    {
+      "epoch": 1.3990384615384617,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8669871794871796e-05,
+      "loss": 2.496,
+      "step": 5529
+    },
+    {
+      "epoch": 1.3992914979757085,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8669028340080973e-05,
+      "loss": 2.3773,
+      "step": 5530
+    },
+    {
+      "epoch": 1.3995445344129553,
+      "grad_norm": 3.75,
+      "learning_rate": 2.866818488529015e-05,
+      "loss": 2.5241,
+      "step": 5531
+    },
+    {
+      "epoch": 1.3997975708502024,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8667341430499326e-05,
+      "loss": 2.4507,
+      "step": 5532
+    },
+    {
+      "epoch": 1.4000506072874495,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8666497975708503e-05,
+      "loss": 2.5109,
+      "step": 5533
+    },
+    {
+      "epoch": 1.4003036437246963,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.866565452091768e-05,
+      "loss": 2.484,
+      "step": 5534
+    },
+    {
+      "epoch": 1.4005566801619433,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8664811066126856e-05,
+      "loss": 2.4766,
+      "step": 5535
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8663967611336033e-05,
+      "loss": 2.488,
+      "step": 5536
+    },
+    {
+      "epoch": 1.4010627530364372,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.866312415654521e-05,
+      "loss": 2.5446,
+      "step": 5537
+    },
+    {
+      "epoch": 1.4013157894736843,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8662280701754386e-05,
+      "loss": 2.537,
+      "step": 5538
+    },
+    {
+      "epoch": 1.4015688259109311,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8661437246963563e-05,
+      "loss": 2.5425,
+      "step": 5539
+    },
+    {
+      "epoch": 1.4018218623481782,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.866059379217274e-05,
+      "loss": 2.5064,
+      "step": 5540
+    },
+    {
+      "epoch": 1.402074898785425,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.8659750337381916e-05,
+      "loss": 2.3599,
+      "step": 5541
+    },
+    {
+      "epoch": 1.402327935222672,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8658906882591096e-05,
+      "loss": 2.4421,
+      "step": 5542
+    },
+    {
+      "epoch": 1.4025809716599191,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.865806342780027e-05,
+      "loss": 2.4702,
+      "step": 5543
+    },
+    {
+      "epoch": 1.402834008097166,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8657219973009446e-05,
+      "loss": 2.4317,
+      "step": 5544
+    },
+    {
+      "epoch": 1.403087044534413,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8656376518218626e-05,
+      "loss": 2.4607,
+      "step": 5545
+    },
+    {
+      "epoch": 1.4033400809716599,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.86555330634278e-05,
+      "loss": 2.56,
+      "step": 5546
+    },
+    {
+      "epoch": 1.403593117408907,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8654689608636976e-05,
+      "loss": 2.4927,
+      "step": 5547
+    },
+    {
+      "epoch": 1.4038461538461537,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8653846153846156e-05,
+      "loss": 2.4775,
+      "step": 5548
+    },
+    {
+      "epoch": 1.4040991902834008,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8653002699055333e-05,
+      "loss": 2.4806,
+      "step": 5549
+    },
+    {
+      "epoch": 1.4043522267206479,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8652159244264506e-05,
+      "loss": 2.4612,
+      "step": 5550
+    },
+    {
+      "epoch": 1.4043522267206479,
+      "eval_loss": 2.5933802127838135,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.0123,
+      "eval_samples_per_second": 9.454,
+      "eval_steps_per_second": 9.454,
+      "step": 5550
+    },
+    {
+      "epoch": 1.4046052631578947,
+      "grad_norm": 3.5,
+      "learning_rate": 2.8651315789473686e-05,
+      "loss": 2.5444,
+      "step": 5551
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8650472334682863e-05,
+      "loss": 2.3354,
+      "step": 5552
+    },
+    {
+      "epoch": 1.4051113360323888,
+      "grad_norm": 3.0,
+      "learning_rate": 2.864962887989204e-05,
+      "loss": 2.4033,
+      "step": 5553
+    },
+    {
+      "epoch": 1.4053643724696356,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8648785425101213e-05,
+      "loss": 2.497,
+      "step": 5554
+    },
+    {
+      "epoch": 1.4056174089068825,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8647941970310393e-05,
+      "loss": 2.5036,
+      "step": 5555
+    },
+    {
+      "epoch": 1.4058704453441295,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.864709851551957e-05,
+      "loss": 2.4637,
+      "step": 5556
+    },
+    {
+      "epoch": 1.4061234817813766,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8646255060728743e-05,
+      "loss": 2.5654,
+      "step": 5557
+    },
+    {
+      "epoch": 1.4063765182186234,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8645411605937923e-05,
+      "loss": 2.4501,
+      "step": 5558
+    },
+    {
+      "epoch": 1.4066295546558705,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.86445681511471e-05,
+      "loss": 2.4979,
+      "step": 5559
+    },
+    {
+      "epoch": 1.4068825910931175,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8643724696356276e-05,
+      "loss": 2.5221,
+      "step": 5560
+    },
+    {
+      "epoch": 1.4071356275303644,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8642881241565453e-05,
+      "loss": 2.4981,
+      "step": 5561
+    },
+    {
+      "epoch": 1.4073886639676112,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.864203778677463e-05,
+      "loss": 2.5873,
+      "step": 5562
+    },
+    {
+      "epoch": 1.4076417004048583,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8641194331983806e-05,
+      "loss": 2.6279,
+      "step": 5563
+    },
+    {
+      "epoch": 1.4078947368421053,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8640350877192986e-05,
+      "loss": 2.4772,
+      "step": 5564
+    },
+    {
+      "epoch": 1.4081477732793521,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.863950742240216e-05,
+      "loss": 2.5333,
+      "step": 5565
+    },
+    {
+      "epoch": 1.4084008097165992,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8638663967611336e-05,
+      "loss": 2.5058,
+      "step": 5566
+    },
+    {
+      "epoch": 1.4086538461538463,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8637820512820516e-05,
+      "loss": 2.5961,
+      "step": 5567
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.863697705802969e-05,
+      "loss": 2.4536,
+      "step": 5568
+    },
+    {
+      "epoch": 1.4091599190283401,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8636133603238866e-05,
+      "loss": 2.5503,
+      "step": 5569
+    },
+    {
+      "epoch": 1.409412955465587,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8635290148448043e-05,
+      "loss": 2.3905,
+      "step": 5570
+    },
+    {
+      "epoch": 1.409665991902834,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8634446693657223e-05,
+      "loss": 2.5709,
+      "step": 5571
+    },
+    {
+      "epoch": 1.4099190283400809,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8633603238866396e-05,
+      "loss": 2.4467,
+      "step": 5572
+    },
+    {
+      "epoch": 1.410172064777328,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8632759784075573e-05,
+      "loss": 2.4864,
+      "step": 5573
+    },
+    {
+      "epoch": 1.410425101214575,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8631916329284753e-05,
+      "loss": 2.4284,
+      "step": 5574
+    },
+    {
+      "epoch": 1.4106781376518218,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.863107287449393e-05,
+      "loss": 2.4667,
+      "step": 5575
+    },
+    {
+      "epoch": 1.4109311740890689,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8630229419703103e-05,
+      "loss": 2.5179,
+      "step": 5576
+    },
+    {
+      "epoch": 1.4111842105263157,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8629385964912283e-05,
+      "loss": 2.5569,
+      "step": 5577
+    },
+    {
+      "epoch": 1.4114372469635628,
+      "grad_norm": 3.25,
+      "learning_rate": 2.862854251012146e-05,
+      "loss": 2.5403,
+      "step": 5578
+    },
+    {
+      "epoch": 1.4116902834008096,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8627699055330633e-05,
+      "loss": 2.3656,
+      "step": 5579
+    },
+    {
+      "epoch": 1.4119433198380567,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8626855600539813e-05,
+      "loss": 2.3478,
+      "step": 5580
+    },
+    {
+      "epoch": 1.4119433198380567,
+      "eval_loss": 2.5878636837005615,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.9806,
+      "eval_samples_per_second": 9.421,
+      "eval_steps_per_second": 9.421,
+      "step": 5580
+    },
+    {
+      "epoch": 1.4121963562753037,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.862601214574899e-05,
+      "loss": 2.4491,
+      "step": 5581
+    },
+    {
+      "epoch": 1.4124493927125505,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8625168690958166e-05,
+      "loss": 2.5043,
+      "step": 5582
+    },
+    {
+      "epoch": 1.4127024291497976,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8624325236167343e-05,
+      "loss": 2.5108,
+      "step": 5583
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.862348178137652e-05,
+      "loss": 2.5478,
+      "step": 5584
+    },
+    {
+      "epoch": 1.4132085020242915,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8622638326585696e-05,
+      "loss": 2.444,
+      "step": 5585
+    },
+    {
+      "epoch": 1.4134615384615383,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8621794871794873e-05,
+      "loss": 2.3735,
+      "step": 5586
+    },
+    {
+      "epoch": 1.4137145748987854,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.862095141700405e-05,
+      "loss": 2.4598,
+      "step": 5587
+    },
+    {
+      "epoch": 1.4139676113360324,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8620107962213226e-05,
+      "loss": 2.4264,
+      "step": 5588
+    },
+    {
+      "epoch": 1.4142206477732793,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8619264507422402e-05,
+      "loss": 2.4404,
+      "step": 5589
+    },
+    {
+      "epoch": 1.4144736842105263,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.861842105263158e-05,
+      "loss": 2.4722,
+      "step": 5590
+    },
+    {
+      "epoch": 1.4147267206477734,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8617577597840756e-05,
+      "loss": 2.4538,
+      "step": 5591
+    },
+    {
+      "epoch": 1.4149797570850202,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8616734143049932e-05,
+      "loss": 2.535,
+      "step": 5592
+    },
+    {
+      "epoch": 1.4152327935222673,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8615890688259112e-05,
+      "loss": 2.466,
+      "step": 5593
+    },
+    {
+      "epoch": 1.415485829959514,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.8615047233468286e-05,
+      "loss": 2.546,
+      "step": 5594
+    },
+    {
+      "epoch": 1.4157388663967612,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8614203778677462e-05,
+      "loss": 2.4784,
+      "step": 5595
+    },
+    {
+      "epoch": 1.415991902834008,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8613360323886642e-05,
+      "loss": 2.4197,
+      "step": 5596
+    },
+    {
+      "epoch": 1.416244939271255,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8612516869095816e-05,
+      "loss": 2.409,
+      "step": 5597
+    },
+    {
+      "epoch": 1.416497975708502,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8611673414304992e-05,
+      "loss": 2.5224,
+      "step": 5598
+    },
+    {
+      "epoch": 1.416751012145749,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8610829959514172e-05,
+      "loss": 2.4355,
+      "step": 5599
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.860998650472335e-05,
+      "loss": 2.5385,
+      "step": 5600
+    },
+    {
+      "epoch": 1.4172570850202428,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8609143049932522e-05,
+      "loss": 2.4609,
+      "step": 5601
+    },
+    {
+      "epoch": 1.41751012145749,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8608299595141702e-05,
+      "loss": 2.4303,
+      "step": 5602
+    },
+    {
+      "epoch": 1.4177631578947367,
+      "grad_norm": 3.125,
+      "learning_rate": 2.860745614035088e-05,
+      "loss": 2.5028,
+      "step": 5603
+    },
+    {
+      "epoch": 1.4180161943319838,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8606612685560056e-05,
+      "loss": 2.5431,
+      "step": 5604
+    },
+    {
+      "epoch": 1.4182692307692308,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.8605769230769232e-05,
+      "loss": 2.3013,
+      "step": 5605
+    },
+    {
+      "epoch": 1.4185222672064777,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.860492577597841e-05,
+      "loss": 2.451,
+      "step": 5606
+    },
+    {
+      "epoch": 1.4187753036437247,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8604082321187586e-05,
+      "loss": 2.5158,
+      "step": 5607
+    },
+    {
+      "epoch": 1.4190283400809718,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.860323886639676e-05,
+      "loss": 2.3718,
+      "step": 5608
+    },
+    {
+      "epoch": 1.4192813765182186,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.860239541160594e-05,
+      "loss": 2.4353,
+      "step": 5609
+    },
+    {
+      "epoch": 1.4195344129554655,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8601551956815116e-05,
+      "loss": 2.5263,
+      "step": 5610
+    },
+    {
+      "epoch": 1.4195344129554655,
+      "eval_loss": 2.5863492488861084,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.0252,
+      "eval_samples_per_second": 9.454,
+      "eval_steps_per_second": 9.454,
+      "step": 5610
+    },
+    {
+      "epoch": 1.4197874493927125,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8600708502024292e-05,
+      "loss": 2.4626,
+      "step": 5611
+    },
+    {
+      "epoch": 1.4200404858299596,
+      "grad_norm": 2.875,
+      "learning_rate": 2.859986504723347e-05,
+      "loss": 2.4039,
+      "step": 5612
+    },
+    {
+      "epoch": 1.4202935222672064,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8599021592442646e-05,
+      "loss": 2.347,
+      "step": 5613
+    },
+    {
+      "epoch": 1.4205465587044535,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8598178137651822e-05,
+      "loss": 2.4788,
+      "step": 5614
+    },
+    {
+      "epoch": 1.4207995951417005,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8597334682861e-05,
+      "loss": 2.509,
+      "step": 5615
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8596491228070175e-05,
+      "loss": 2.3106,
+      "step": 5616
+    },
+    {
+      "epoch": 1.4213056680161944,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8595647773279352e-05,
+      "loss": 2.4209,
+      "step": 5617
+    },
+    {
+      "epoch": 1.4215587044534412,
+      "grad_norm": 2.625,
+      "learning_rate": 2.8594804318488532e-05,
+      "loss": 2.3647,
+      "step": 5618
+    },
+    {
+      "epoch": 1.4218117408906883,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8593960863697705e-05,
+      "loss": 2.3807,
+      "step": 5619
+    },
+    {
+      "epoch": 1.4220647773279351,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8593117408906882e-05,
+      "loss": 2.3685,
+      "step": 5620
+    },
+    {
+      "epoch": 1.4223178137651822,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8592273954116062e-05,
+      "loss": 2.4592,
+      "step": 5621
+    },
+    {
+      "epoch": 1.4225708502024292,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.859143049932524e-05,
+      "loss": 2.514,
+      "step": 5622
+    },
+    {
+      "epoch": 1.422823886639676,
+      "grad_norm": 3.609375,
+      "learning_rate": 2.8590587044534412e-05,
+      "loss": 2.4804,
+      "step": 5623
+    },
+    {
+      "epoch": 1.4230769230769231,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.858974358974359e-05,
+      "loss": 2.5243,
+      "step": 5624
+    },
+    {
+      "epoch": 1.42332995951417,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.858890013495277e-05,
+      "loss": 2.5822,
+      "step": 5625
+    },
+    {
+      "epoch": 1.423582995951417,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8588056680161942e-05,
+      "loss": 2.4923,
+      "step": 5626
+    },
+    {
+      "epoch": 1.4238360323886639,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.858721322537112e-05,
+      "loss": 2.5205,
+      "step": 5627
+    },
+    {
+      "epoch": 1.424089068825911,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.85863697705803e-05,
+      "loss": 2.5443,
+      "step": 5628
+    },
+    {
+      "epoch": 1.424342105263158,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8585526315789475e-05,
+      "loss": 2.505,
+      "step": 5629
+    },
+    {
+      "epoch": 1.4245951417004048,
+      "grad_norm": 3.25,
+      "learning_rate": 2.858468286099865e-05,
+      "loss": 2.5259,
+      "step": 5630
+    },
+    {
+      "epoch": 1.4248481781376519,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.858383940620783e-05,
+      "loss": 2.4454,
+      "step": 5631
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8582995951417005e-05,
+      "loss": 2.5299,
+      "step": 5632
+    },
+    {
+      "epoch": 1.4253542510121457,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8582152496626182e-05,
+      "loss": 2.3898,
+      "step": 5633
+    },
+    {
+      "epoch": 1.4256072874493926,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.858130904183536e-05,
+      "loss": 2.5317,
+      "step": 5634
+    },
+    {
+      "epoch": 1.4258603238866396,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8580465587044535e-05,
+      "loss": 2.4938,
+      "step": 5635
+    },
+    {
+      "epoch": 1.4261133603238867,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8579622132253712e-05,
+      "loss": 2.481,
+      "step": 5636
+    },
+    {
+      "epoch": 1.4263663967611335,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.857877867746289e-05,
+      "loss": 2.4266,
+      "step": 5637
+    },
+    {
+      "epoch": 1.4266194331983806,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8577935222672065e-05,
+      "loss": 2.5727,
+      "step": 5638
+    },
+    {
+      "epoch": 1.4268724696356276,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8577091767881242e-05,
+      "loss": 2.6257,
+      "step": 5639
+    },
+    {
+      "epoch": 1.4271255060728745,
+      "grad_norm": 21.0,
+      "learning_rate": 2.8576248313090422e-05,
+      "loss": 2.6945,
+      "step": 5640
+    },
+    {
+      "epoch": 1.4271255060728745,
+      "eval_loss": 2.585735559463501,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.1737,
+      "eval_samples_per_second": 9.414,
+      "eval_steps_per_second": 9.414,
+      "step": 5640
+    },
+    {
+      "epoch": 1.4273785425101215,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8575404858299595e-05,
+      "loss": 2.5301,
+      "step": 5641
+    },
+    {
+      "epoch": 1.4276315789473684,
+      "grad_norm": 11.25,
+      "learning_rate": 2.8574561403508772e-05,
+      "loss": 2.5136,
+      "step": 5642
+    },
+    {
+      "epoch": 1.4278846153846154,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.857371794871795e-05,
+      "loss": 2.5125,
+      "step": 5643
+    },
+    {
+      "epoch": 1.4281376518218623,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.857287449392713e-05,
+      "loss": 2.4118,
+      "step": 5644
+    },
+    {
+      "epoch": 1.4283906882591093,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.8572031039136302e-05,
+      "loss": 2.5436,
+      "step": 5645
+    },
+    {
+      "epoch": 1.4286437246963564,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.857118758434548e-05,
+      "loss": 2.4664,
+      "step": 5646
+    },
+    {
+      "epoch": 1.4288967611336032,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.857034412955466e-05,
+      "loss": 2.4289,
+      "step": 5647
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8569500674763832e-05,
+      "loss": 2.5674,
+      "step": 5648
+    },
+    {
+      "epoch": 1.429402834008097,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.856865721997301e-05,
+      "loss": 2.4174,
+      "step": 5649
+    },
+    {
+      "epoch": 1.4296558704453441,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.856781376518219e-05,
+      "loss": 2.4061,
+      "step": 5650
+    },
+    {
+      "epoch": 1.429908906882591,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8566970310391365e-05,
+      "loss": 2.446,
+      "step": 5651
+    },
+    {
+      "epoch": 1.430161943319838,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.856612685560054e-05,
+      "loss": 2.551,
+      "step": 5652
+    },
+    {
+      "epoch": 1.430414979757085,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.856528340080972e-05,
+      "loss": 2.2122,
+      "step": 5653
+    },
+    {
+      "epoch": 1.430668016194332,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8564439946018895e-05,
+      "loss": 2.4601,
+      "step": 5654
+    },
+    {
+      "epoch": 1.430921052631579,
+      "grad_norm": 3.0,
+      "learning_rate": 2.856359649122807e-05,
+      "loss": 2.4794,
+      "step": 5655
+    },
+    {
+      "epoch": 1.431174089068826,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.856275303643725e-05,
+      "loss": 2.3805,
+      "step": 5656
+    },
+    {
+      "epoch": 1.4314271255060729,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8561909581646425e-05,
+      "loss": 2.3825,
+      "step": 5657
+    },
+    {
+      "epoch": 1.4316801619433197,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8561066126855602e-05,
+      "loss": 2.547,
+      "step": 5658
+    },
+    {
+      "epoch": 1.4319331983805668,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.856022267206478e-05,
+      "loss": 2.3666,
+      "step": 5659
+    },
+    {
+      "epoch": 1.4321862348178138,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8559379217273955e-05,
+      "loss": 2.4457,
+      "step": 5660
+    },
+    {
+      "epoch": 1.4324392712550607,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.855853576248313e-05,
+      "loss": 2.4778,
+      "step": 5661
+    },
+    {
+      "epoch": 1.4326923076923077,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.855769230769231e-05,
+      "loss": 2.4399,
+      "step": 5662
+    },
+    {
+      "epoch": 1.4329453441295548,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8556848852901485e-05,
+      "loss": 2.4875,
+      "step": 5663
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.855600539811066e-05,
+      "loss": 2.3818,
+      "step": 5664
+    },
+    {
+      "epoch": 1.4334514170040487,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8555161943319838e-05,
+      "loss": 2.4211,
+      "step": 5665
+    },
+    {
+      "epoch": 1.4337044534412955,
+      "grad_norm": 2.75,
+      "learning_rate": 2.8554318488529015e-05,
+      "loss": 2.4491,
+      "step": 5666
+    },
+    {
+      "epoch": 1.4339574898785425,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.855347503373819e-05,
+      "loss": 2.3582,
+      "step": 5667
+    },
+    {
+      "epoch": 1.4342105263157894,
+      "grad_norm": 3.6875,
+      "learning_rate": 2.8552631578947368e-05,
+      "loss": 2.2664,
+      "step": 5668
+    },
+    {
+      "epoch": 1.4344635627530364,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8551788124156548e-05,
+      "loss": 2.5613,
+      "step": 5669
+    },
+    {
+      "epoch": 1.4347165991902835,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.855094466936572e-05,
+      "loss": 2.4973,
+      "step": 5670
+    },
+    {
+      "epoch": 1.4347165991902835,
+      "eval_loss": 2.586474895477295,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.822,
+      "eval_samples_per_second": 9.392,
+      "eval_steps_per_second": 9.392,
+      "step": 5670
+    },
+    {
+      "epoch": 1.4349696356275303,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8550101214574898e-05,
+      "loss": 2.4886,
+      "step": 5671
+    },
+    {
+      "epoch": 1.4352226720647774,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8549257759784078e-05,
+      "loss": 2.4589,
+      "step": 5672
+    },
+    {
+      "epoch": 1.4354757085020242,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.8548414304993255e-05,
+      "loss": 2.3214,
+      "step": 5673
+    },
+    {
+      "epoch": 1.4357287449392713,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8547570850202428e-05,
+      "loss": 2.4501,
+      "step": 5674
+    },
+    {
+      "epoch": 1.435981781376518,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8546727395411608e-05,
+      "loss": 2.4421,
+      "step": 5675
+    },
+    {
+      "epoch": 1.4362348178137652,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8545883940620785e-05,
+      "loss": 2.5333,
+      "step": 5676
+    },
+    {
+      "epoch": 1.4364878542510122,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8545040485829958e-05,
+      "loss": 2.4638,
+      "step": 5677
+    },
+    {
+      "epoch": 1.436740890688259,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.8544197031039135e-05,
+      "loss": 2.5418,
+      "step": 5678
+    },
+    {
+      "epoch": 1.436993927125506,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8543353576248315e-05,
+      "loss": 2.4331,
+      "step": 5679
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.854251012145749e-05,
+      "loss": 2.419,
+      "step": 5680
+    },
+    {
+      "epoch": 1.4375,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.8541666666666665e-05,
+      "loss": 2.4689,
+      "step": 5681
+    },
+    {
+      "epoch": 1.4377530364372468,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8540823211875845e-05,
+      "loss": 2.4562,
+      "step": 5682
+    },
+    {
+      "epoch": 1.438006072874494,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.853997975708502e-05,
+      "loss": 2.4876,
+      "step": 5683
+    },
+    {
+      "epoch": 1.438259109311741,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8539136302294198e-05,
+      "loss": 2.4665,
+      "step": 5684
+    },
+    {
+      "epoch": 1.4385121457489878,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.8538292847503375e-05,
+      "loss": 2.4669,
+      "step": 5685
+    },
+    {
+      "epoch": 1.4387651821862348,
+      "grad_norm": 3.0,
+      "learning_rate": 2.853744939271255e-05,
+      "loss": 2.5609,
+      "step": 5686
+    },
+    {
+      "epoch": 1.439018218623482,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8536605937921728e-05,
+      "loss": 2.3835,
+      "step": 5687
+    },
+    {
+      "epoch": 1.4392712550607287,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8535762483130905e-05,
+      "loss": 2.4026,
+      "step": 5688
+    },
+    {
+      "epoch": 1.4395242914979758,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.853491902834008e-05,
+      "loss": 2.3028,
+      "step": 5689
+    },
+    {
+      "epoch": 1.4397773279352226,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8534075573549258e-05,
+      "loss": 2.4973,
+      "step": 5690
+    },
+    {
+      "epoch": 1.4400303643724697,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8533232118758438e-05,
+      "loss": 2.2723,
+      "step": 5691
+    },
+    {
+      "epoch": 1.4402834008097165,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.853238866396761e-05,
+      "loss": 2.4868,
+      "step": 5692
+    },
+    {
+      "epoch": 1.4405364372469636,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8531545209176788e-05,
+      "loss": 2.4983,
+      "step": 5693
+    },
+    {
+      "epoch": 1.4407894736842106,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8530701754385968e-05,
+      "loss": 2.6268,
+      "step": 5694
+    },
+    {
+      "epoch": 1.4410425101214575,
+      "grad_norm": 2.875,
+      "learning_rate": 2.852985829959514e-05,
+      "loss": 2.5127,
+      "step": 5695
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8529014844804318e-05,
+      "loss": 2.4506,
+      "step": 5696
+    },
+    {
+      "epoch": 1.4415485829959513,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.8528171390013495e-05,
+      "loss": 2.5002,
+      "step": 5697
+    },
+    {
+      "epoch": 1.4418016194331984,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8527327935222675e-05,
+      "loss": 2.6156,
+      "step": 5698
+    },
+    {
+      "epoch": 1.4420546558704452,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8526484480431848e-05,
+      "loss": 2.42,
+      "step": 5699
+    },
+    {
+      "epoch": 1.4423076923076923,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8525641025641025e-05,
+      "loss": 2.5253,
+      "step": 5700
+    },
+    {
+      "epoch": 1.4423076923076923,
+      "eval_loss": 2.5833022594451904,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.1781,
+      "eval_samples_per_second": 9.414,
+      "eval_steps_per_second": 9.414,
+      "step": 5700
+    },
+    {
+      "epoch": 1.4425607287449393,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8524797570850205e-05,
+      "loss": 2.4825,
+      "step": 5701
+    },
+    {
+      "epoch": 1.4428137651821862,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.852395411605938e-05,
+      "loss": 2.4182,
+      "step": 5702
+    },
+    {
+      "epoch": 1.4430668016194332,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8523110661268554e-05,
+      "loss": 2.4882,
+      "step": 5703
+    },
+    {
+      "epoch": 1.4433198380566803,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8522267206477735e-05,
+      "loss": 2.4018,
+      "step": 5704
+    },
+    {
+      "epoch": 1.4435728744939271,
+      "grad_norm": 2.75,
+      "learning_rate": 2.852142375168691e-05,
+      "loss": 2.4379,
+      "step": 5705
+    },
+    {
+      "epoch": 1.443825910931174,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8520580296896084e-05,
+      "loss": 2.4972,
+      "step": 5706
+    },
+    {
+      "epoch": 1.444078947368421,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8519736842105264e-05,
+      "loss": 2.4438,
+      "step": 5707
+    },
+    {
+      "epoch": 1.444331983805668,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.851889338731444e-05,
+      "loss": 2.4646,
+      "step": 5708
+    },
+    {
+      "epoch": 1.444585020242915,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8518049932523618e-05,
+      "loss": 2.5041,
+      "step": 5709
+    },
+    {
+      "epoch": 1.444838056680162,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8517206477732794e-05,
+      "loss": 2.2652,
+      "step": 5710
+    },
+    {
+      "epoch": 1.445091093117409,
+      "grad_norm": 2.75,
+      "learning_rate": 2.851636302294197e-05,
+      "loss": 2.2834,
+      "step": 5711
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8515519568151148e-05,
+      "loss": 2.4888,
+      "step": 5712
+    },
+    {
+      "epoch": 1.445597165991903,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8514676113360328e-05,
+      "loss": 2.4713,
+      "step": 5713
+    },
+    {
+      "epoch": 1.4458502024291497,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.85138326585695e-05,
+      "loss": 2.4481,
+      "step": 5714
+    },
+    {
+      "epoch": 1.4461032388663968,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8512989203778678e-05,
+      "loss": 2.2896,
+      "step": 5715
+    },
+    {
+      "epoch": 1.4463562753036436,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8512145748987854e-05,
+      "loss": 2.4637,
+      "step": 5716
+    },
+    {
+      "epoch": 1.4466093117408907,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.851130229419703e-05,
+      "loss": 2.4942,
+      "step": 5717
+    },
+    {
+      "epoch": 1.4468623481781377,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8510458839406208e-05,
+      "loss": 2.4844,
+      "step": 5718
+    },
+    {
+      "epoch": 1.4471153846153846,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8509615384615384e-05,
+      "loss": 2.5146,
+      "step": 5719
+    },
+    {
+      "epoch": 1.4473684210526316,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8508771929824564e-05,
+      "loss": 2.469,
+      "step": 5720
+    },
+    {
+      "epoch": 1.4476214574898785,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8507928475033738e-05,
+      "loss": 2.4561,
+      "step": 5721
+    },
+    {
+      "epoch": 1.4478744939271255,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.8507085020242914e-05,
+      "loss": 2.5052,
+      "step": 5722
+    },
+    {
+      "epoch": 1.4481275303643724,
+      "grad_norm": 16.25,
+      "learning_rate": 2.8506241565452094e-05,
+      "loss": 2.4677,
+      "step": 5723
+    },
+    {
+      "epoch": 1.4483805668016194,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.850539811066127e-05,
+      "loss": 2.4552,
+      "step": 5724
+    },
+    {
+      "epoch": 1.4486336032388665,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8504554655870444e-05,
+      "loss": 2.4588,
+      "step": 5725
+    },
+    {
+      "epoch": 1.4488866396761133,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8503711201079624e-05,
+      "loss": 2.4258,
+      "step": 5726
+    },
+    {
+      "epoch": 1.4491396761133604,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.85028677462888e-05,
+      "loss": 2.5852,
+      "step": 5727
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8502024291497974e-05,
+      "loss": 2.4945,
+      "step": 5728
+    },
+    {
+      "epoch": 1.4496457489878543,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8501180836707154e-05,
+      "loss": 2.527,
+      "step": 5729
+    },
+    {
+      "epoch": 1.449898785425101,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.850033738191633e-05,
+      "loss": 2.4947,
+      "step": 5730
+    },
+    {
+      "epoch": 1.449898785425101,
+      "eval_loss": 2.583387613296509,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 275.148,
+      "eval_samples_per_second": 9.449,
+      "eval_steps_per_second": 9.449,
+      "step": 5730
+    },
+    {
+      "epoch": 1.4501518218623481,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8499493927125508e-05,
+      "loss": 2.5343,
+      "step": 5731
+    },
+    {
+      "epoch": 1.4504048582995952,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8498650472334684e-05,
+      "loss": 2.4974,
+      "step": 5732
+    },
+    {
+      "epoch": 1.450657894736842,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.849780701754386e-05,
+      "loss": 2.5266,
+      "step": 5733
+    },
+    {
+      "epoch": 1.450910931174089,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8496963562753038e-05,
+      "loss": 2.355,
+      "step": 5734
+    },
+    {
+      "epoch": 1.4511639676113361,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.849612010796221e-05,
+      "loss": 2.5112,
+      "step": 5735
+    },
+    {
+      "epoch": 1.451417004048583,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.849527665317139e-05,
+      "loss": 2.4965,
+      "step": 5736
+    },
+    {
+      "epoch": 1.45167004048583,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8494433198380567e-05,
+      "loss": 2.5391,
+      "step": 5737
+    },
+    {
+      "epoch": 1.4519230769230769,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8493589743589744e-05,
+      "loss": 2.4644,
+      "step": 5738
+    },
+    {
+      "epoch": 1.452176113360324,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.849274628879892e-05,
+      "loss": 2.38,
+      "step": 5739
+    },
+    {
+      "epoch": 1.4524291497975708,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8491902834008097e-05,
+      "loss": 2.4881,
+      "step": 5740
+    },
+    {
+      "epoch": 1.4526821862348178,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.8491059379217274e-05,
+      "loss": 2.4356,
+      "step": 5741
+    },
+    {
+      "epoch": 1.4529352226720649,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8490215924426454e-05,
+      "loss": 2.5405,
+      "step": 5742
+    },
+    {
+      "epoch": 1.4531882591093117,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8489372469635627e-05,
+      "loss": 2.5106,
+      "step": 5743
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8488529014844804e-05,
+      "loss": 2.5405,
+      "step": 5744
+    },
+    {
+      "epoch": 1.4536943319838056,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8487685560053984e-05,
+      "loss": 2.4785,
+      "step": 5745
+    },
+    {
+      "epoch": 1.4539473684210527,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8486842105263157e-05,
+      "loss": 2.5057,
+      "step": 5746
+    },
+    {
+      "epoch": 1.4542004048582995,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8485998650472334e-05,
+      "loss": 2.4419,
+      "step": 5747
+    },
+    {
+      "epoch": 1.4544534412955465,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8485155195681514e-05,
+      "loss": 2.4677,
+      "step": 5748
+    },
+    {
+      "epoch": 1.4547064777327936,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.848431174089069e-05,
+      "loss": 2.5775,
+      "step": 5749
+    },
+    {
+      "epoch": 1.4549595141700404,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8483468286099864e-05,
+      "loss": 2.5784,
+      "step": 5750
+    },
+    {
+      "epoch": 1.4552125506072875,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.848262483130904e-05,
+      "loss": 2.3211,
+      "step": 5751
+    },
+    {
+      "epoch": 1.4554655870445345,
+      "grad_norm": 3.0,
+      "learning_rate": 2.848178137651822e-05,
+      "loss": 2.5749,
+      "step": 5752
+    },
+    {
+      "epoch": 1.4557186234817814,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.8480937921727397e-05,
+      "loss": 2.469,
+      "step": 5753
+    },
+    {
+      "epoch": 1.4559716599190282,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.848009446693657e-05,
+      "loss": 2.4641,
+      "step": 5754
+    },
+    {
+      "epoch": 1.4562246963562753,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.847925101214575e-05,
+      "loss": 2.5647,
+      "step": 5755
+    },
+    {
+      "epoch": 1.4564777327935223,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8478407557354927e-05,
+      "loss": 2.522,
+      "step": 5756
+    },
+    {
+      "epoch": 1.4567307692307692,
+      "grad_norm": 3.125,
+      "learning_rate": 2.84775641025641e-05,
+      "loss": 2.4612,
+      "step": 5757
+    },
+    {
+      "epoch": 1.4569838056680162,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.847672064777328e-05,
+      "loss": 2.478,
+      "step": 5758
+    },
+    {
+      "epoch": 1.4572368421052633,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8475877192982457e-05,
+      "loss": 2.4704,
+      "step": 5759
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8475033738191634e-05,
+      "loss": 2.52,
+      "step": 5760
+    },
+    {
+      "epoch": 1.45748987854251,
+      "eval_loss": 2.5835964679718018,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 277.2156,
+      "eval_samples_per_second": 9.379,
+      "eval_steps_per_second": 9.379,
+      "step": 5760
+    },
+    {
+      "epoch": 1.4577429149797572,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.847419028340081e-05,
+      "loss": 2.5172,
+      "step": 5761
+    },
+    {
+      "epoch": 1.457995951417004,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8473346828609987e-05,
+      "loss": 2.5195,
+      "step": 5762
+    },
+    {
+      "epoch": 1.458248987854251,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8472503373819164e-05,
+      "loss": 2.5159,
+      "step": 5763
+    },
+    {
+      "epoch": 1.458502024291498,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.847165991902834e-05,
+      "loss": 2.4314,
+      "step": 5764
+    },
+    {
+      "epoch": 1.458755060728745,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8470816464237517e-05,
+      "loss": 2.5082,
+      "step": 5765
+    },
+    {
+      "epoch": 1.459008097165992,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8469973009446694e-05,
+      "loss": 2.4664,
+      "step": 5766
+    },
+    {
+      "epoch": 1.4592611336032388,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8469129554655874e-05,
+      "loss": 2.5298,
+      "step": 5767
+    },
+    {
+      "epoch": 1.459514170040486,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8468286099865047e-05,
+      "loss": 2.3651,
+      "step": 5768
+    },
+    {
+      "epoch": 1.4597672064777327,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8467442645074224e-05,
+      "loss": 2.4123,
+      "step": 5769
+    },
+    {
+      "epoch": 1.4600202429149798,
+      "grad_norm": 3.0,
+      "learning_rate": 2.84665991902834e-05,
+      "loss": 2.511,
+      "step": 5770
+    },
+    {
+      "epoch": 1.4602732793522266,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.846575573549258e-05,
+      "loss": 2.5344,
+      "step": 5771
+    },
+    {
+      "epoch": 1.4605263157894737,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8464912280701754e-05,
+      "loss": 2.3816,
+      "step": 5772
+    },
+    {
+      "epoch": 1.4607793522267207,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.846406882591093e-05,
+      "loss": 2.4809,
+      "step": 5773
+    },
+    {
+      "epoch": 1.4610323886639676,
+      "grad_norm": 3.375,
+      "learning_rate": 2.846322537112011e-05,
+      "loss": 2.4871,
+      "step": 5774
+    },
+    {
+      "epoch": 1.4612854251012146,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8462381916329284e-05,
+      "loss": 2.4975,
+      "step": 5775
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.846153846153846e-05,
+      "loss": 2.4816,
+      "step": 5776
+    },
+    {
+      "epoch": 1.4617914979757085,
+      "grad_norm": 2.875,
+      "learning_rate": 2.846069500674764e-05,
+      "loss": 2.4285,
+      "step": 5777
+    },
+    {
+      "epoch": 1.4620445344129553,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8459851551956817e-05,
+      "loss": 2.4341,
+      "step": 5778
+    },
+    {
+      "epoch": 1.4622975708502024,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.845900809716599e-05,
+      "loss": 2.4656,
+      "step": 5779
+    },
+    {
+      "epoch": 1.4625506072874495,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.845816464237517e-05,
+      "loss": 2.2871,
+      "step": 5780
+    },
+    {
+      "epoch": 1.4628036437246963,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8457321187584347e-05,
+      "loss": 2.4774,
+      "step": 5781
+    },
+    {
+      "epoch": 1.4630566801619433,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8456477732793524e-05,
+      "loss": 2.4734,
+      "step": 5782
+    },
+    {
+      "epoch": 1.4633097165991904,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.84556342780027e-05,
+      "loss": 2.4654,
+      "step": 5783
+    },
+    {
+      "epoch": 1.4635627530364372,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8454790823211877e-05,
+      "loss": 2.4225,
+      "step": 5784
+    },
+    {
+      "epoch": 1.4638157894736843,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8453947368421054e-05,
+      "loss": 2.4346,
+      "step": 5785
+    },
+    {
+      "epoch": 1.4640688259109311,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.845310391363023e-05,
+      "loss": 2.4591,
+      "step": 5786
+    },
+    {
+      "epoch": 1.4643218623481782,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8452260458839407e-05,
+      "loss": 2.4488,
+      "step": 5787
+    },
+    {
+      "epoch": 1.464574898785425,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.8451417004048584e-05,
+      "loss": 2.4973,
+      "step": 5788
+    },
+    {
+      "epoch": 1.464827935222672,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.845057354925776e-05,
+      "loss": 2.2202,
+      "step": 5789
+    },
+    {
+      "epoch": 1.4650809716599191,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8449730094466937e-05,
+      "loss": 2.4674,
+      "step": 5790
+    },
+    {
+      "epoch": 1.4650809716599191,
+      "eval_loss": 2.5826256275177,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.821,
+      "eval_samples_per_second": 9.392,
+      "eval_steps_per_second": 9.392,
+      "step": 5790
+    },
+    {
+      "epoch": 1.465334008097166,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8448886639676114e-05,
+      "loss": 2.5362,
+      "step": 5791
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.844804318488529e-05,
+      "loss": 2.5361,
+      "step": 5792
+    },
+    {
+      "epoch": 1.4658400809716599,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.844719973009447e-05,
+      "loss": 2.473,
+      "step": 5793
+    },
+    {
+      "epoch": 1.466093117408907,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8446356275303643e-05,
+      "loss": 2.5435,
+      "step": 5794
+    },
+    {
+      "epoch": 1.4663461538461537,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.844551282051282e-05,
+      "loss": 2.4656,
+      "step": 5795
+    },
+    {
+      "epoch": 1.4665991902834008,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8444669365722e-05,
+      "loss": 2.5954,
+      "step": 5796
+    },
+    {
+      "epoch": 1.4668522267206479,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8443825910931173e-05,
+      "loss": 2.4172,
+      "step": 5797
+    },
+    {
+      "epoch": 1.4671052631578947,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.844298245614035e-05,
+      "loss": 2.3689,
+      "step": 5798
+    },
+    {
+      "epoch": 1.4673582995951417,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.844213900134953e-05,
+      "loss": 2.435,
+      "step": 5799
+    },
+    {
+      "epoch": 1.4676113360323888,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8441295546558707e-05,
+      "loss": 2.5473,
+      "step": 5800
+    },
+    {
+      "epoch": 1.4678643724696356,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.844045209176788e-05,
+      "loss": 2.485,
+      "step": 5801
+    },
+    {
+      "epoch": 1.4681174089068825,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.843960863697706e-05,
+      "loss": 2.5483,
+      "step": 5802
+    },
+    {
+      "epoch": 1.4683704453441295,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8438765182186237e-05,
+      "loss": 2.4968,
+      "step": 5803
+    },
+    {
+      "epoch": 1.4686234817813766,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.8437921727395413e-05,
+      "loss": 2.5325,
+      "step": 5804
+    },
+    {
+      "epoch": 1.4688765182186234,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8437078272604587e-05,
+      "loss": 2.5293,
+      "step": 5805
+    },
+    {
+      "epoch": 1.4691295546558705,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8436234817813767e-05,
+      "loss": 2.5165,
+      "step": 5806
+    },
+    {
+      "epoch": 1.4693825910931175,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8435391363022943e-05,
+      "loss": 2.5174,
+      "step": 5807
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8434547908232117e-05,
+      "loss": 2.518,
+      "step": 5808
+    },
+    {
+      "epoch": 1.4698886639676112,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8433704453441297e-05,
+      "loss": 2.412,
+      "step": 5809
+    },
+    {
+      "epoch": 1.4701417004048583,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8432860998650473e-05,
+      "loss": 2.4819,
+      "step": 5810
+    },
+    {
+      "epoch": 1.4703947368421053,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.843201754385965e-05,
+      "loss": 2.3748,
+      "step": 5811
+    },
+    {
+      "epoch": 1.4706477732793521,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8431174089068827e-05,
+      "loss": 2.2885,
+      "step": 5812
+    },
+    {
+      "epoch": 1.4709008097165992,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8430330634278003e-05,
+      "loss": 2.4327,
+      "step": 5813
+    },
+    {
+      "epoch": 1.4711538461538463,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.842948717948718e-05,
+      "loss": 2.4164,
+      "step": 5814
+    },
+    {
+      "epoch": 1.471406882591093,
+      "grad_norm": 18.25,
+      "learning_rate": 2.8428643724696357e-05,
+      "loss": 2.639,
+      "step": 5815
+    },
+    {
+      "epoch": 1.4716599190283401,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8427800269905533e-05,
+      "loss": 2.373,
+      "step": 5816
+    },
+    {
+      "epoch": 1.471912955465587,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.842695681511471e-05,
+      "loss": 2.5408,
+      "step": 5817
+    },
+    {
+      "epoch": 1.472165991902834,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.842611336032389e-05,
+      "loss": 2.4497,
+      "step": 5818
+    },
+    {
+      "epoch": 1.4724190283400809,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8425269905533063e-05,
+      "loss": 2.4578,
+      "step": 5819
+    },
+    {
+      "epoch": 1.472672064777328,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.842442645074224e-05,
+      "loss": 2.4464,
+      "step": 5820
+    },
+    {
+      "epoch": 1.472672064777328,
+      "eval_loss": 2.5831332206726074,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 278.6892,
+      "eval_samples_per_second": 9.329,
+      "eval_steps_per_second": 9.329,
+      "step": 5820
+    },
+    {
+      "epoch": 1.472925101214575,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.842358299595142e-05,
+      "loss": 2.4277,
+      "step": 5821
+    },
+    {
+      "epoch": 1.4731781376518218,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8422739541160597e-05,
+      "loss": 2.5062,
+      "step": 5822
+    },
+    {
+      "epoch": 1.4734311740890689,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.842189608636977e-05,
+      "loss": 2.3734,
+      "step": 5823
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8421052631578946e-05,
+      "loss": 2.5589,
+      "step": 5824
+    },
+    {
+      "epoch": 1.4739372469635628,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8420209176788127e-05,
+      "loss": 2.5866,
+      "step": 5825
+    },
+    {
+      "epoch": 1.4741902834008096,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.84193657219973e-05,
+      "loss": 2.4988,
+      "step": 5826
+    },
+    {
+      "epoch": 1.4744433198380567,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8418522267206476e-05,
+      "loss": 2.5619,
+      "step": 5827
+    },
+    {
+      "epoch": 1.4746963562753037,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8417678812415656e-05,
+      "loss": 2.5225,
+      "step": 5828
+    },
+    {
+      "epoch": 1.4749493927125505,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8416835357624833e-05,
+      "loss": 2.4181,
+      "step": 5829
+    },
+    {
+      "epoch": 1.4752024291497976,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.8415991902834006e-05,
+      "loss": 2.4243,
+      "step": 5830
+    },
+    {
+      "epoch": 1.4754554655870447,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8415148448043186e-05,
+      "loss": 2.3252,
+      "step": 5831
+    },
+    {
+      "epoch": 1.4757085020242915,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8414304993252363e-05,
+      "loss": 2.5289,
+      "step": 5832
+    },
+    {
+      "epoch": 1.4759615384615383,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.841346153846154e-05,
+      "loss": 2.4007,
+      "step": 5833
+    },
+    {
+      "epoch": 1.4762145748987854,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8412618083670716e-05,
+      "loss": 2.3971,
+      "step": 5834
+    },
+    {
+      "epoch": 1.4764676113360324,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8411774628879893e-05,
+      "loss": 2.4882,
+      "step": 5835
+    },
+    {
+      "epoch": 1.4767206477732793,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.841093117408907e-05,
+      "loss": 2.44,
+      "step": 5836
+    },
+    {
+      "epoch": 1.4769736842105263,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8410087719298246e-05,
+      "loss": 2.4491,
+      "step": 5837
+    },
+    {
+      "epoch": 1.4772267206477734,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8409244264507423e-05,
+      "loss": 2.4799,
+      "step": 5838
+    },
+    {
+      "epoch": 1.4774797570850202,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.84084008097166e-05,
+      "loss": 2.5037,
+      "step": 5839
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.840755735492578e-05,
+      "loss": 2.4251,
+      "step": 5840
+    },
+    {
+      "epoch": 1.477985829959514,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8406713900134953e-05,
+      "loss": 2.4876,
+      "step": 5841
+    },
+    {
+      "epoch": 1.4782388663967612,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.840587044534413e-05,
+      "loss": 2.3812,
+      "step": 5842
+    },
+    {
+      "epoch": 1.478491902834008,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8405026990553306e-05,
+      "loss": 2.4963,
+      "step": 5843
+    },
+    {
+      "epoch": 1.478744939271255,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8404183535762483e-05,
+      "loss": 2.4318,
+      "step": 5844
+    },
+    {
+      "epoch": 1.478997975708502,
+      "grad_norm": 2.75,
+      "learning_rate": 2.840334008097166e-05,
+      "loss": 2.4273,
+      "step": 5845
+    },
+    {
+      "epoch": 1.479251012145749,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8402496626180836e-05,
+      "loss": 2.4446,
+      "step": 5846
+    },
+    {
+      "epoch": 1.479504048582996,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8401653171390016e-05,
+      "loss": 2.4659,
+      "step": 5847
+    },
+    {
+      "epoch": 1.4797570850202428,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.840080971659919e-05,
+      "loss": 2.5185,
+      "step": 5848
+    },
+    {
+      "epoch": 1.48001012145749,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8399966261808366e-05,
+      "loss": 2.5604,
+      "step": 5849
+    },
+    {
+      "epoch": 1.4802631578947367,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8399122807017546e-05,
+      "loss": 2.4647,
+      "step": 5850
+    },
+    {
+      "epoch": 1.4802631578947367,
+      "eval_loss": 2.578089475631714,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 277.5287,
+      "eval_samples_per_second": 9.368,
+      "eval_steps_per_second": 9.368,
+      "step": 5850
+    },
+    {
+      "epoch": 1.4805161943319838,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8398279352226723e-05,
+      "loss": 2.4432,
+      "step": 5851
+    },
+    {
+      "epoch": 1.4807692307692308,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8397435897435896e-05,
+      "loss": 2.4459,
+      "step": 5852
+    },
+    {
+      "epoch": 1.4810222672064777,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8396592442645076e-05,
+      "loss": 2.4325,
+      "step": 5853
+    },
+    {
+      "epoch": 1.4812753036437247,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8395748987854253e-05,
+      "loss": 2.5175,
+      "step": 5854
+    },
+    {
+      "epoch": 1.4815283400809718,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8394905533063426e-05,
+      "loss": 2.4146,
+      "step": 5855
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8394062078272606e-05,
+      "loss": 2.516,
+      "step": 5856
+    },
+    {
+      "epoch": 1.4820344129554655,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8393218623481783e-05,
+      "loss": 2.3161,
+      "step": 5857
+    },
+    {
+      "epoch": 1.4822874493927125,
+      "grad_norm": 3.0,
+      "learning_rate": 2.839237516869096e-05,
+      "loss": 2.5488,
+      "step": 5858
+    },
+    {
+      "epoch": 1.4825404858299596,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8391531713900133e-05,
+      "loss": 2.3183,
+      "step": 5859
+    },
+    {
+      "epoch": 1.4827935222672064,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8390688259109313e-05,
+      "loss": 2.5639,
+      "step": 5860
+    },
+    {
+      "epoch": 1.4830465587044535,
+      "grad_norm": 2.875,
+      "learning_rate": 2.838984480431849e-05,
+      "loss": 2.456,
+      "step": 5861
+    },
+    {
+      "epoch": 1.4832995951417005,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8389001349527666e-05,
+      "loss": 2.4296,
+      "step": 5862
+    },
+    {
+      "epoch": 1.4835526315789473,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8388157894736843e-05,
+      "loss": 2.4452,
+      "step": 5863
+    },
+    {
+      "epoch": 1.4838056680161944,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.838731443994602e-05,
+      "loss": 2.4634,
+      "step": 5864
+    },
+    {
+      "epoch": 1.4840587044534412,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8386470985155196e-05,
+      "loss": 2.4012,
+      "step": 5865
+    },
+    {
+      "epoch": 1.4843117408906883,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8385627530364373e-05,
+      "loss": 2.4686,
+      "step": 5866
+    },
+    {
+      "epoch": 1.4845647773279351,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.838478407557355e-05,
+      "loss": 2.4873,
+      "step": 5867
+    },
+    {
+      "epoch": 1.4848178137651822,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8383940620782726e-05,
+      "loss": 2.4722,
+      "step": 5868
+    },
+    {
+      "epoch": 1.4850708502024292,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8383097165991906e-05,
+      "loss": 2.589,
+      "step": 5869
+    },
+    {
+      "epoch": 1.485323886639676,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.838225371120108e-05,
+      "loss": 2.41,
+      "step": 5870
+    },
+    {
+      "epoch": 1.4855769230769231,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.8381410256410256e-05,
+      "loss": 2.3705,
+      "step": 5871
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8380566801619436e-05,
+      "loss": 2.4133,
+      "step": 5872
+    },
+    {
+      "epoch": 1.486082995951417,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8379723346828613e-05,
+      "loss": 2.4582,
+      "step": 5873
+    },
+    {
+      "epoch": 1.4863360323886639,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8378879892037786e-05,
+      "loss": 2.4875,
+      "step": 5874
+    },
+    {
+      "epoch": 1.486589068825911,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8378036437246966e-05,
+      "loss": 2.4131,
+      "step": 5875
+    },
+    {
+      "epoch": 1.486842105263158,
+      "grad_norm": 3.375,
+      "learning_rate": 2.8377192982456143e-05,
+      "loss": 2.4697,
+      "step": 5876
+    },
+    {
+      "epoch": 1.4870951417004048,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8376349527665316e-05,
+      "loss": 2.5466,
+      "step": 5877
+    },
+    {
+      "epoch": 1.4873481781376519,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.8375506072874493e-05,
+      "loss": 2.594,
+      "step": 5878
+    },
+    {
+      "epoch": 1.487601214574899,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8374662618083673e-05,
+      "loss": 2.4481,
+      "step": 5879
+    },
+    {
+      "epoch": 1.4878542510121457,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.837381916329285e-05,
+      "loss": 2.3939,
+      "step": 5880
+    },
+    {
+      "epoch": 1.4878542510121457,
+      "eval_loss": 2.5790834426879883,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 277.6706,
+      "eval_samples_per_second": 9.364,
+      "eval_steps_per_second": 9.364,
+      "step": 5880
+    },
+    {
+      "epoch": 1.4881072874493926,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8372975708502022e-05,
+      "loss": 2.4838,
+      "step": 5881
+    },
+    {
+      "epoch": 1.4883603238866396,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8372132253711203e-05,
+      "loss": 2.5687,
+      "step": 5882
+    },
+    {
+      "epoch": 1.4886133603238867,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.837128879892038e-05,
+      "loss": 2.3626,
+      "step": 5883
+    },
+    {
+      "epoch": 1.4888663967611335,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8370445344129552e-05,
+      "loss": 2.4404,
+      "step": 5884
+    },
+    {
+      "epoch": 1.4891194331983806,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8369601889338733e-05,
+      "loss": 2.5216,
+      "step": 5885
+    },
+    {
+      "epoch": 1.4893724696356276,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.836875843454791e-05,
+      "loss": 2.4778,
+      "step": 5886
+    },
+    {
+      "epoch": 1.4896255060728745,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8367914979757086e-05,
+      "loss": 2.391,
+      "step": 5887
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8367071524966262e-05,
+      "loss": 2.4999,
+      "step": 5888
+    },
+    {
+      "epoch": 1.4901315789473684,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.836622807017544e-05,
+      "loss": 2.5312,
+      "step": 5889
+    },
+    {
+      "epoch": 1.4903846153846154,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8365384615384616e-05,
+      "loss": 2.4547,
+      "step": 5890
+    },
+    {
+      "epoch": 1.4906376518218623,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8364541160593796e-05,
+      "loss": 2.4611,
+      "step": 5891
+    },
+    {
+      "epoch": 1.4908906882591093,
+      "grad_norm": 2.625,
+      "learning_rate": 2.836369770580297e-05,
+      "loss": 2.3641,
+      "step": 5892
+    },
+    {
+      "epoch": 1.4911437246963564,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8362854251012146e-05,
+      "loss": 2.4311,
+      "step": 5893
+    },
+    {
+      "epoch": 1.4913967611336032,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8362010796221326e-05,
+      "loss": 2.3158,
+      "step": 5894
+    },
+    {
+      "epoch": 1.4916497975708503,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.83611673414305e-05,
+      "loss": 2.3361,
+      "step": 5895
+    },
+    {
+      "epoch": 1.491902834008097,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8360323886639676e-05,
+      "loss": 2.5542,
+      "step": 5896
+    },
+    {
+      "epoch": 1.4921558704453441,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8359480431848852e-05,
+      "loss": 2.4767,
+      "step": 5897
+    },
+    {
+      "epoch": 1.492408906882591,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8358636977058032e-05,
+      "loss": 2.4513,
+      "step": 5898
+    },
+    {
+      "epoch": 1.492661943319838,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8357793522267206e-05,
+      "loss": 2.5457,
+      "step": 5899
+    },
+    {
+      "epoch": 1.492914979757085,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8356950067476382e-05,
+      "loss": 2.5629,
+      "step": 5900
+    },
+    {
+      "epoch": 1.493168016194332,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.8356106612685562e-05,
+      "loss": 2.3922,
+      "step": 5901
+    },
+    {
+      "epoch": 1.493421052631579,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.835526315789474e-05,
+      "loss": 2.4935,
+      "step": 5902
+    },
+    {
+      "epoch": 1.493674089068826,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8354419703103912e-05,
+      "loss": 2.4602,
+      "step": 5903
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8353576248313092e-05,
+      "loss": 2.517,
+      "step": 5904
+    },
+    {
+      "epoch": 1.4941801619433197,
+      "grad_norm": 3.0,
+      "learning_rate": 2.835273279352227e-05,
+      "loss": 2.5292,
+      "step": 5905
+    },
+    {
+      "epoch": 1.4944331983805668,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.8351889338731442e-05,
+      "loss": 2.4935,
+      "step": 5906
+    },
+    {
+      "epoch": 1.4946862348178138,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8351045883940622e-05,
+      "loss": 2.3356,
+      "step": 5907
+    },
+    {
+      "epoch": 1.4949392712550607,
+      "grad_norm": 3.375,
+      "learning_rate": 2.83502024291498e-05,
+      "loss": 2.5854,
+      "step": 5908
+    },
+    {
+      "epoch": 1.4951923076923077,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.8349358974358976e-05,
+      "loss": 2.4202,
+      "step": 5909
+    },
+    {
+      "epoch": 1.4954453441295548,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8348515519568152e-05,
+      "loss": 2.3042,
+      "step": 5910
+    },
+    {
+      "epoch": 1.4954453441295548,
+      "eval_loss": 2.5754177570343018,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 277.241,
+      "eval_samples_per_second": 9.378,
+      "eval_steps_per_second": 9.378,
+      "step": 5910
+    },
+    {
+      "epoch": 1.4956983805668016,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.834767206477733e-05,
+      "loss": 2.3414,
+      "step": 5911
+    },
+    {
+      "epoch": 1.4959514170040487,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8346828609986506e-05,
+      "loss": 2.5469,
+      "step": 5912
+    },
+    {
+      "epoch": 1.4962044534412955,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8345985155195682e-05,
+      "loss": 2.3496,
+      "step": 5913
+    },
+    {
+      "epoch": 1.4964574898785425,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.834514170040486e-05,
+      "loss": 2.5248,
+      "step": 5914
+    },
+    {
+      "epoch": 1.4967105263157894,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.8344298245614035e-05,
+      "loss": 2.542,
+      "step": 5915
+    },
+    {
+      "epoch": 1.4969635627530364,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8343454790823212e-05,
+      "loss": 2.4789,
+      "step": 5916
+    },
+    {
+      "epoch": 1.4972165991902835,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.834261133603239e-05,
+      "loss": 2.5214,
+      "step": 5917
+    },
+    {
+      "epoch": 1.4974696356275303,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8341767881241565e-05,
+      "loss": 2.3998,
+      "step": 5918
+    },
+    {
+      "epoch": 1.4977226720647774,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8340924426450742e-05,
+      "loss": 2.5176,
+      "step": 5919
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8340080971659922e-05,
+      "loss": 2.4432,
+      "step": 5920
+    },
+    {
+      "epoch": 1.4982287449392713,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8339237516869095e-05,
+      "loss": 2.5635,
+      "step": 5921
+    },
+    {
+      "epoch": 1.498481781376518,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8338394062078272e-05,
+      "loss": 2.5199,
+      "step": 5922
+    },
+    {
+      "epoch": 1.4987348178137652,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.8337550607287452e-05,
+      "loss": 2.3607,
+      "step": 5923
+    },
+    {
+      "epoch": 1.4989878542510122,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8336707152496625e-05,
+      "loss": 2.5311,
+      "step": 5924
+    },
+    {
+      "epoch": 1.499240890688259,
+      "grad_norm": 2.75,
+      "learning_rate": 2.8335863697705802e-05,
+      "loss": 2.3421,
+      "step": 5925
+    },
+    {
+      "epoch": 1.499493927125506,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8335020242914982e-05,
+      "loss": 2.5123,
+      "step": 5926
+    },
+    {
+      "epoch": 1.4997469635627532,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.833417678812416e-05,
+      "loss": 2.4789,
+      "step": 5927
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8333333333333332e-05,
+      "loss": 2.4349,
+      "step": 5928
+    },
+    {
+      "epoch": 1.5002530364372468,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8332489878542512e-05,
+      "loss": 2.4378,
+      "step": 5929
+    },
+    {
+      "epoch": 1.500506072874494,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.833164642375169e-05,
+      "loss": 2.5601,
+      "step": 5930
+    },
+    {
+      "epoch": 1.500759109311741,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8330802968960865e-05,
+      "loss": 2.5161,
+      "step": 5931
+    },
+    {
+      "epoch": 1.5010121457489878,
+      "grad_norm": 2.875,
+      "learning_rate": 2.832995951417004e-05,
+      "loss": 2.3922,
+      "step": 5932
+    },
+    {
+      "epoch": 1.5012651821862348,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.832911605937922e-05,
+      "loss": 2.4332,
+      "step": 5933
+    },
+    {
+      "epoch": 1.501518218623482,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8328272604588395e-05,
+      "loss": 2.4829,
+      "step": 5934
+    },
+    {
+      "epoch": 1.5017712550607287,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.832742914979757e-05,
+      "loss": 2.4714,
+      "step": 5935
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.832658569500675e-05,
+      "loss": 2.5121,
+      "step": 5936
+    },
+    {
+      "epoch": 1.5022773279352226,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8325742240215925e-05,
+      "loss": 2.478,
+      "step": 5937
+    },
+    {
+      "epoch": 1.5025303643724697,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.8324898785425102e-05,
+      "loss": 2.4949,
+      "step": 5938
+    },
+    {
+      "epoch": 1.5027834008097165,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.832405533063428e-05,
+      "loss": 2.4714,
+      "step": 5939
+    },
+    {
+      "epoch": 1.5030364372469636,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8323211875843455e-05,
+      "loss": 2.5049,
+      "step": 5940
+    },
+    {
+      "epoch": 1.5030364372469636,
+      "eval_loss": 2.5780413150787354,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 277.9453,
+      "eval_samples_per_second": 9.354,
+      "eval_steps_per_second": 9.354,
+      "step": 5940
+    },
+    {
+      "epoch": 1.5032894736842106,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8322368421052632e-05,
+      "loss": 2.5064,
+      "step": 5941
+    },
+    {
+      "epoch": 1.5035425101214575,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8321524966261812e-05,
+      "loss": 2.5167,
+      "step": 5942
+    },
+    {
+      "epoch": 1.5037955465587043,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8320681511470985e-05,
+      "loss": 2.4803,
+      "step": 5943
+    },
+    {
+      "epoch": 1.5040485829959516,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8319838056680162e-05,
+      "loss": 2.3815,
+      "step": 5944
+    },
+    {
+      "epoch": 1.5043016194331984,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8318994601889342e-05,
+      "loss": 2.3612,
+      "step": 5945
+    },
+    {
+      "epoch": 1.5045546558704452,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8318151147098515e-05,
+      "loss": 2.4966,
+      "step": 5946
+    },
+    {
+      "epoch": 1.5048076923076923,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8317307692307692e-05,
+      "loss": 2.4321,
+      "step": 5947
+    },
+    {
+      "epoch": 1.5050607287449393,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8316464237516872e-05,
+      "loss": 2.4395,
+      "step": 5948
+    },
+    {
+      "epoch": 1.5053137651821862,
+      "grad_norm": 2.75,
+      "learning_rate": 2.831562078272605e-05,
+      "loss": 2.4577,
+      "step": 5949
+    },
+    {
+      "epoch": 1.5055668016194332,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8314777327935222e-05,
+      "loss": 2.4539,
+      "step": 5950
+    },
+    {
+      "epoch": 1.5058198380566803,
+      "grad_norm": 2.875,
+      "learning_rate": 2.83139338731444e-05,
+      "loss": 2.4937,
+      "step": 5951
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.831309041835358e-05,
+      "loss": 2.4766,
+      "step": 5952
+    },
+    {
+      "epoch": 1.506325910931174,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8312246963562755e-05,
+      "loss": 2.4643,
+      "step": 5953
+    },
+    {
+      "epoch": 1.506578947368421,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.831140350877193e-05,
+      "loss": 2.648,
+      "step": 5954
+    },
+    {
+      "epoch": 1.506831983805668,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.831056005398111e-05,
+      "loss": 2.3613,
+      "step": 5955
+    },
+    {
+      "epoch": 1.507085020242915,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8309716599190285e-05,
+      "loss": 2.4496,
+      "step": 5956
+    },
+    {
+      "epoch": 1.507338056680162,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.830887314439946e-05,
+      "loss": 2.5195,
+      "step": 5957
+    },
+    {
+      "epoch": 1.507591093117409,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.830802968960864e-05,
+      "loss": 2.5306,
+      "step": 5958
+    },
+    {
+      "epoch": 1.5078441295546559,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8307186234817815e-05,
+      "loss": 2.5358,
+      "step": 5959
+    },
+    {
+      "epoch": 1.5080971659919027,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.830634278002699e-05,
+      "loss": 2.3758,
+      "step": 5960
+    },
+    {
+      "epoch": 1.5083502024291497,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.830549932523617e-05,
+      "loss": 2.5307,
+      "step": 5961
+    },
+    {
+      "epoch": 1.5086032388663968,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8304655870445345e-05,
+      "loss": 2.443,
+      "step": 5962
+    },
+    {
+      "epoch": 1.5088562753036436,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.830381241565452e-05,
+      "loss": 2.4712,
+      "step": 5963
+    },
+    {
+      "epoch": 1.5091093117408907,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8302968960863698e-05,
+      "loss": 2.5525,
+      "step": 5964
+    },
+    {
+      "epoch": 1.5093623481781377,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8302125506072875e-05,
+      "loss": 2.3883,
+      "step": 5965
+    },
+    {
+      "epoch": 1.5096153846153846,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.830128205128205e-05,
+      "loss": 2.4648,
+      "step": 5966
+    },
+    {
+      "epoch": 1.5098684210526314,
+      "grad_norm": 3.125,
+      "learning_rate": 2.830043859649123e-05,
+      "loss": 2.4672,
+      "step": 5967
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8299595141700405e-05,
+      "loss": 2.5117,
+      "step": 5968
+    },
+    {
+      "epoch": 1.5103744939271255,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.829875168690958e-05,
+      "loss": 2.4136,
+      "step": 5969
+    },
+    {
+      "epoch": 1.5106275303643724,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8297908232118758e-05,
+      "loss": 2.4545,
+      "step": 5970
+    },
+    {
+      "epoch": 1.5106275303643724,
+      "eval_loss": 2.5759289264678955,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 277.3728,
+      "eval_samples_per_second": 9.374,
+      "eval_steps_per_second": 9.374,
+      "step": 5970
+    },
+    {
+      "epoch": 1.5108805668016194,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8297064777327938e-05,
+      "loss": 2.4136,
+      "step": 5971
+    },
+    {
+      "epoch": 1.5111336032388665,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.829622132253711e-05,
+      "loss": 2.4432,
+      "step": 5972
+    },
+    {
+      "epoch": 1.5113866396761133,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8295377867746288e-05,
+      "loss": 2.5604,
+      "step": 5973
+    },
+    {
+      "epoch": 1.5116396761133604,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8294534412955468e-05,
+      "loss": 2.5072,
+      "step": 5974
+    },
+    {
+      "epoch": 1.5118927125506074,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.829369095816464e-05,
+      "loss": 2.5286,
+      "step": 5975
+    },
+    {
+      "epoch": 1.5121457489878543,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8292847503373818e-05,
+      "loss": 2.5489,
+      "step": 5976
+    },
+    {
+      "epoch": 1.512398785425101,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8292004048582998e-05,
+      "loss": 2.5126,
+      "step": 5977
+    },
+    {
+      "epoch": 1.5126518218623481,
+      "grad_norm": 3.375,
+      "learning_rate": 2.8291160593792175e-05,
+      "loss": 2.5287,
+      "step": 5978
+    },
+    {
+      "epoch": 1.5129048582995952,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8290317139001348e-05,
+      "loss": 2.3929,
+      "step": 5979
+    },
+    {
+      "epoch": 1.513157894736842,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8289473684210528e-05,
+      "loss": 2.4795,
+      "step": 5980
+    },
+    {
+      "epoch": 1.513410931174089,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8288630229419705e-05,
+      "loss": 2.4651,
+      "step": 5981
+    },
+    {
+      "epoch": 1.5136639676113361,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.828778677462888e-05,
+      "loss": 2.4882,
+      "step": 5982
+    },
+    {
+      "epoch": 1.513917004048583,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8286943319838058e-05,
+      "loss": 2.4846,
+      "step": 5983
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8286099865047235e-05,
+      "loss": 2.4871,
+      "step": 5984
+    },
+    {
+      "epoch": 1.5144230769230769,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.828525641025641e-05,
+      "loss": 2.4279,
+      "step": 5985
+    },
+    {
+      "epoch": 1.514676113360324,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8284412955465585e-05,
+      "loss": 2.4845,
+      "step": 5986
+    },
+    {
+      "epoch": 1.5149291497975708,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8283569500674765e-05,
+      "loss": 2.4342,
+      "step": 5987
+    },
+    {
+      "epoch": 1.5151821862348178,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.828272604588394e-05,
+      "loss": 2.485,
+      "step": 5988
+    },
+    {
+      "epoch": 1.5154352226720649,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8281882591093118e-05,
+      "loss": 2.3489,
+      "step": 5989
+    },
+    {
+      "epoch": 1.5156882591093117,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8281039136302295e-05,
+      "loss": 2.4904,
+      "step": 5990
+    },
+    {
+      "epoch": 1.5159412955465585,
+      "grad_norm": 3.125,
+      "learning_rate": 2.828019568151147e-05,
+      "loss": 2.5336,
+      "step": 5991
+    },
+    {
+      "epoch": 1.5161943319838058,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.8279352226720648e-05,
+      "loss": 2.3326,
+      "step": 5992
+    },
+    {
+      "epoch": 1.5164473684210527,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8278508771929828e-05,
+      "loss": 2.5127,
+      "step": 5993
+    },
+    {
+      "epoch": 1.5167004048582995,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8277665317139e-05,
+      "loss": 2.4879,
+      "step": 5994
+    },
+    {
+      "epoch": 1.5169534412955465,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8276821862348178e-05,
+      "loss": 2.4477,
+      "step": 5995
+    },
+    {
+      "epoch": 1.5172064777327936,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8275978407557358e-05,
+      "loss": 2.4356,
+      "step": 5996
+    },
+    {
+      "epoch": 1.5174595141700404,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.827513495276653e-05,
+      "loss": 2.4603,
+      "step": 5997
+    },
+    {
+      "epoch": 1.5177125506072875,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8274291497975708e-05,
+      "loss": 2.4832,
+      "step": 5998
+    },
+    {
+      "epoch": 1.5179655870445345,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8273448043184888e-05,
+      "loss": 2.429,
+      "step": 5999
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8272604588394065e-05,
+      "loss": 2.5285,
+      "step": 6000
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "eval_loss": 2.573819160461426,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 276.5471,
+      "eval_samples_per_second": 9.402,
+      "eval_steps_per_second": 9.402,
+      "step": 6000
+    },
+    {
+      "epoch": 1.5184716599190282,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.8271761133603238e-05,
+      "loss": 2.4864,
+      "step": 6001
+    },
+    {
+      "epoch": 1.5187246963562753,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8270917678812418e-05,
+      "loss": 2.4149,
+      "step": 6002
+    },
+    {
+      "epoch": 1.5189777327935223,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8270074224021595e-05,
+      "loss": 2.4334,
+      "step": 6003
+    },
+    {
+      "epoch": 1.5192307692307692,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8269230769230768e-05,
+      "loss": 2.4008,
+      "step": 6004
+    },
+    {
+      "epoch": 1.5194838056680162,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8268387314439944e-05,
+      "loss": 2.5149,
+      "step": 6005
+    },
+    {
+      "epoch": 1.5197368421052633,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.8267543859649124e-05,
+      "loss": 2.3914,
+      "step": 6006
+    },
+    {
+      "epoch": 1.51998987854251,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.82667004048583e-05,
+      "loss": 2.4652,
+      "step": 6007
+    },
+    {
+      "epoch": 1.520242914979757,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8265856950067474e-05,
+      "loss": 2.4738,
+      "step": 6008
+    },
+    {
+      "epoch": 1.520495951417004,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8265013495276654e-05,
+      "loss": 2.3956,
+      "step": 6009
+    },
+    {
+      "epoch": 1.520748987854251,
+      "grad_norm": 2.75,
+      "learning_rate": 2.826417004048583e-05,
+      "loss": 2.4087,
+      "step": 6010
+    },
+    {
+      "epoch": 1.521002024291498,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8263326585695008e-05,
+      "loss": 2.391,
+      "step": 6011
+    },
+    {
+      "epoch": 1.521255060728745,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8262483130904184e-05,
+      "loss": 2.3583,
+      "step": 6012
+    },
+    {
+      "epoch": 1.521508097165992,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.826163967611336e-05,
+      "loss": 2.4818,
+      "step": 6013
+    },
+    {
+      "epoch": 1.5217611336032388,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8260796221322538e-05,
+      "loss": 2.5353,
+      "step": 6014
+    },
+    {
+      "epoch": 1.5220141700404857,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8259952766531714e-05,
+      "loss": 2.3712,
+      "step": 6015
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.825910931174089e-05,
+      "loss": 2.4982,
+      "step": 6016
+    },
+    {
+      "epoch": 1.5225202429149798,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8258265856950068e-05,
+      "loss": 2.5366,
+      "step": 6017
+    },
+    {
+      "epoch": 1.5227732793522266,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8257422402159248e-05,
+      "loss": 2.3968,
+      "step": 6018
+    },
+    {
+      "epoch": 1.5230263157894737,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.825657894736842e-05,
+      "loss": 2.4476,
+      "step": 6019
+    },
+    {
+      "epoch": 1.5232793522267207,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8255735492577598e-05,
+      "loss": 2.4746,
+      "step": 6020
+    },
+    {
+      "epoch": 1.5235323886639676,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8254892037786778e-05,
+      "loss": 2.5154,
+      "step": 6021
+    },
+    {
+      "epoch": 1.5237854251012146,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8254048582995954e-05,
+      "loss": 2.4363,
+      "step": 6022
+    },
+    {
+      "epoch": 1.5240384615384617,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.8253205128205128e-05,
+      "loss": 2.3932,
+      "step": 6023
+    },
+    {
+      "epoch": 1.5242914979757085,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8252361673414304e-05,
+      "loss": 2.5062,
+      "step": 6024
+    },
+    {
+      "epoch": 1.5245445344129553,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8251518218623484e-05,
+      "loss": 2.4795,
+      "step": 6025
+    },
+    {
+      "epoch": 1.5247975708502024,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8250674763832658e-05,
+      "loss": 2.5355,
+      "step": 6026
+    },
+    {
+      "epoch": 1.5250506072874495,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8249831309041834e-05,
+      "loss": 2.4698,
+      "step": 6027
+    },
+    {
+      "epoch": 1.5253036437246963,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8248987854251014e-05,
+      "loss": 2.4706,
+      "step": 6028
+    },
+    {
+      "epoch": 1.5255566801619433,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.824814439946019e-05,
+      "loss": 2.4836,
+      "step": 6029
+    },
+    {
+      "epoch": 1.5258097165991904,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8247300944669364e-05,
+      "loss": 2.5202,
+      "step": 6030
+    },
+    {
+      "epoch": 1.5258097165991904,
+      "eval_loss": 2.5752227306365967,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 287.5732,
+      "eval_samples_per_second": 9.041,
+      "eval_steps_per_second": 9.041,
+      "step": 6030
+    },
+    {
+      "epoch": 1.5260627530364372,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8246457489878544e-05,
+      "loss": 2.4723,
+      "step": 6031
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.824561403508772e-05,
+      "loss": 2.5167,
+      "step": 6032
+    },
+    {
+      "epoch": 1.5265688259109311,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8244770580296898e-05,
+      "loss": 2.4638,
+      "step": 6033
+    },
+    {
+      "epoch": 1.5268218623481782,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8243927125506074e-05,
+      "loss": 2.4073,
+      "step": 6034
+    },
+    {
+      "epoch": 1.527074898785425,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.824308367071525e-05,
+      "loss": 2.4162,
+      "step": 6035
+    },
+    {
+      "epoch": 1.527327935222672,
+      "grad_norm": 2.75,
+      "learning_rate": 2.8242240215924427e-05,
+      "loss": 2.4605,
+      "step": 6036
+    },
+    {
+      "epoch": 1.5275809716599191,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8241396761133604e-05,
+      "loss": 2.4884,
+      "step": 6037
+    },
+    {
+      "epoch": 1.527834008097166,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.824055330634278e-05,
+      "loss": 2.4699,
+      "step": 6038
+    },
+    {
+      "epoch": 1.5280870445344128,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8239709851551957e-05,
+      "loss": 2.5012,
+      "step": 6039
+    },
+    {
+      "epoch": 1.52834008097166,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8238866396761134e-05,
+      "loss": 2.4338,
+      "step": 6040
+    },
+    {
+      "epoch": 1.528593117408907,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.823802294197031e-05,
+      "loss": 2.4811,
+      "step": 6041
+    },
+    {
+      "epoch": 1.5288461538461537,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8237179487179487e-05,
+      "loss": 2.559,
+      "step": 6042
+    },
+    {
+      "epoch": 1.5290991902834008,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8236336032388664e-05,
+      "loss": 2.3524,
+      "step": 6043
+    },
+    {
+      "epoch": 1.5293522267206479,
+      "grad_norm": 16.375,
+      "learning_rate": 2.823549257759784e-05,
+      "loss": 2.5293,
+      "step": 6044
+    },
+    {
+      "epoch": 1.5296052631578947,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8234649122807017e-05,
+      "loss": 2.506,
+      "step": 6045
+    },
+    {
+      "epoch": 1.5298582995951417,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8233805668016194e-05,
+      "loss": 2.4643,
+      "step": 6046
+    },
+    {
+      "epoch": 1.5301113360323888,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8232962213225374e-05,
+      "loss": 2.4107,
+      "step": 6047
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8232118758434547e-05,
+      "loss": 2.574,
+      "step": 6048
+    },
+    {
+      "epoch": 1.5306174089068825,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8231275303643724e-05,
+      "loss": 2.366,
+      "step": 6049
+    },
+    {
+      "epoch": 1.5308704453441295,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8230431848852904e-05,
+      "loss": 2.5111,
+      "step": 6050
+    },
+    {
+      "epoch": 1.5311234817813766,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.822958839406208e-05,
+      "loss": 2.4732,
+      "step": 6051
+    },
+    {
+      "epoch": 1.5313765182186234,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8228744939271254e-05,
+      "loss": 2.3773,
+      "step": 6052
+    },
+    {
+      "epoch": 1.5316295546558705,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8227901484480434e-05,
+      "loss": 2.5688,
+      "step": 6053
+    },
+    {
+      "epoch": 1.5318825910931175,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.822705802968961e-05,
+      "loss": 2.409,
+      "step": 6054
+    },
+    {
+      "epoch": 1.5321356275303644,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8226214574898784e-05,
+      "loss": 2.5426,
+      "step": 6055
+    },
+    {
+      "epoch": 1.5323886639676112,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8225371120107964e-05,
+      "loss": 2.5024,
+      "step": 6056
+    },
+    {
+      "epoch": 1.5326417004048583,
+      "grad_norm": 2.875,
+      "learning_rate": 2.822452766531714e-05,
+      "loss": 2.414,
+      "step": 6057
+    },
+    {
+      "epoch": 1.5328947368421053,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8223684210526317e-05,
+      "loss": 2.4948,
+      "step": 6058
+    },
+    {
+      "epoch": 1.5331477732793521,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.822284075573549e-05,
+      "loss": 2.516,
+      "step": 6059
+    },
+    {
+      "epoch": 1.5334008097165992,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.822199730094467e-05,
+      "loss": 2.4791,
+      "step": 6060
+    },
+    {
+      "epoch": 1.5334008097165992,
+      "eval_loss": 2.573637008666992,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 278.555,
+      "eval_samples_per_second": 9.334,
+      "eval_steps_per_second": 9.334,
+      "step": 6060
+    },
+    {
+      "epoch": 1.5336538461538463,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8221153846153847e-05,
+      "loss": 2.4536,
+      "step": 6061
+    },
+    {
+      "epoch": 1.533906882591093,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.8220310391363024e-05,
+      "loss": 2.3245,
+      "step": 6062
+    },
+    {
+      "epoch": 1.53415991902834,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.82194669365722e-05,
+      "loss": 2.5445,
+      "step": 6063
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8218623481781377e-05,
+      "loss": 2.4269,
+      "step": 6064
+    },
+    {
+      "epoch": 1.534665991902834,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8217780026990554e-05,
+      "loss": 2.5577,
+      "step": 6065
+    },
+    {
+      "epoch": 1.5349190283400809,
+      "grad_norm": 3.0,
+      "learning_rate": 2.821693657219973e-05,
+      "loss": 2.4485,
+      "step": 6066
+    },
+    {
+      "epoch": 1.535172064777328,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8216093117408907e-05,
+      "loss": 2.4904,
+      "step": 6067
+    },
+    {
+      "epoch": 1.535425101214575,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8215249662618084e-05,
+      "loss": 2.6056,
+      "step": 6068
+    },
+    {
+      "epoch": 1.5356781376518218,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8214406207827264e-05,
+      "loss": 2.5475,
+      "step": 6069
+    },
+    {
+      "epoch": 1.5359311740890689,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8213562753036437e-05,
+      "loss": 2.4756,
+      "step": 6070
+    },
+    {
+      "epoch": 1.536184210526316,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8212719298245614e-05,
+      "loss": 2.4704,
+      "step": 6071
+    },
+    {
+      "epoch": 1.5364372469635628,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8211875843454794e-05,
+      "loss": 2.4826,
+      "step": 6072
+    },
+    {
+      "epoch": 1.5366902834008096,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8211032388663967e-05,
+      "loss": 2.5109,
+      "step": 6073
+    },
+    {
+      "epoch": 1.5369433198380567,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8210188933873144e-05,
+      "loss": 2.4768,
+      "step": 6074
+    },
+    {
+      "epoch": 1.5371963562753037,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8209345479082324e-05,
+      "loss": 2.5319,
+      "step": 6075
+    },
+    {
+      "epoch": 1.5374493927125505,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.82085020242915e-05,
+      "loss": 2.5059,
+      "step": 6076
+    },
+    {
+      "epoch": 1.5377024291497976,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8207658569500674e-05,
+      "loss": 2.405,
+      "step": 6077
+    },
+    {
+      "epoch": 1.5379554655870447,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.820681511470985e-05,
+      "loss": 2.4441,
+      "step": 6078
+    },
+    {
+      "epoch": 1.5382085020242915,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.820597165991903e-05,
+      "loss": 2.5174,
+      "step": 6079
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8205128205128207e-05,
+      "loss": 2.514,
+      "step": 6080
+    },
+    {
+      "epoch": 1.5387145748987854,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.820428475033738e-05,
+      "loss": 2.4765,
+      "step": 6081
+    },
+    {
+      "epoch": 1.5389676113360324,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.820344129554656e-05,
+      "loss": 2.4602,
+      "step": 6082
+    },
+    {
+      "epoch": 1.5392206477732793,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8202597840755737e-05,
+      "loss": 2.548,
+      "step": 6083
+    },
+    {
+      "epoch": 1.5394736842105263,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.820175438596491e-05,
+      "loss": 2.3457,
+      "step": 6084
+    },
+    {
+      "epoch": 1.5397267206477734,
+      "grad_norm": 2.875,
+      "learning_rate": 2.820091093117409e-05,
+      "loss": 2.373,
+      "step": 6085
+    },
+    {
+      "epoch": 1.5399797570850202,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8200067476383267e-05,
+      "loss": 2.4141,
+      "step": 6086
+    },
+    {
+      "epoch": 1.540232793522267,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8199224021592444e-05,
+      "loss": 2.4155,
+      "step": 6087
+    },
+    {
+      "epoch": 1.5404858299595143,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.819838056680162e-05,
+      "loss": 2.4884,
+      "step": 6088
+    },
+    {
+      "epoch": 1.5407388663967612,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8197537112010797e-05,
+      "loss": 2.4825,
+      "step": 6089
+    },
+    {
+      "epoch": 1.540991902834008,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8196693657219974e-05,
+      "loss": 2.4925,
+      "step": 6090
+    },
+    {
+      "epoch": 1.540991902834008,
+      "eval_loss": 2.5765902996063232,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 278.6845,
+      "eval_samples_per_second": 9.33,
+      "eval_steps_per_second": 9.33,
+      "step": 6090
+    },
+    {
+      "epoch": 1.541244939271255,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8195850202429154e-05,
+      "loss": 2.4291,
+      "step": 6091
+    },
+    {
+      "epoch": 1.541497975708502,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8195006747638327e-05,
+      "loss": 2.517,
+      "step": 6092
+    },
+    {
+      "epoch": 1.541751012145749,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8194163292847503e-05,
+      "loss": 2.4972,
+      "step": 6093
+    },
+    {
+      "epoch": 1.542004048582996,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.819331983805668e-05,
+      "loss": 2.5711,
+      "step": 6094
+    },
+    {
+      "epoch": 1.542257085020243,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8192476383265857e-05,
+      "loss": 2.4792,
+      "step": 6095
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8191632928475033e-05,
+      "loss": 2.4473,
+      "step": 6096
+    },
+    {
+      "epoch": 1.5427631578947367,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.819078947368421e-05,
+      "loss": 2.3884,
+      "step": 6097
+    },
+    {
+      "epoch": 1.5430161943319838,
+      "grad_norm": 2.875,
+      "learning_rate": 2.818994601889339e-05,
+      "loss": 2.4883,
+      "step": 6098
+    },
+    {
+      "epoch": 1.5432692307692308,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8189102564102563e-05,
+      "loss": 2.3866,
+      "step": 6099
+    },
+    {
+      "epoch": 1.5435222672064777,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.818825910931174e-05,
+      "loss": 2.3471,
+      "step": 6100
+    },
+    {
+      "epoch": 1.5437753036437247,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.818741565452092e-05,
+      "loss": 2.5045,
+      "step": 6101
+    },
+    {
+      "epoch": 1.5440283400809718,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8186572199730097e-05,
+      "loss": 2.4905,
+      "step": 6102
+    },
+    {
+      "epoch": 1.5442813765182186,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.818572874493927e-05,
+      "loss": 2.4475,
+      "step": 6103
+    },
+    {
+      "epoch": 1.5445344129554655,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.818488529014845e-05,
+      "loss": 2.4576,
+      "step": 6104
+    },
+    {
+      "epoch": 1.5447874493927125,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8184041835357627e-05,
+      "loss": 2.4169,
+      "step": 6105
+    },
+    {
+      "epoch": 1.5450404858299596,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.81831983805668e-05,
+      "loss": 2.4768,
+      "step": 6106
+    },
+    {
+      "epoch": 1.5452935222672064,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.818235492577598e-05,
+      "loss": 2.5543,
+      "step": 6107
+    },
+    {
+      "epoch": 1.5455465587044535,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8181511470985157e-05,
+      "loss": 2.5633,
+      "step": 6108
+    },
+    {
+      "epoch": 1.5457995951417005,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8180668016194333e-05,
+      "loss": 2.4435,
+      "step": 6109
+    },
+    {
+      "epoch": 1.5460526315789473,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.817982456140351e-05,
+      "loss": 2.4473,
+      "step": 6110
+    },
+    {
+      "epoch": 1.5463056680161942,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8178981106612687e-05,
+      "loss": 2.4152,
+      "step": 6111
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8178137651821863e-05,
+      "loss": 2.4676,
+      "step": 6112
+    },
+    {
+      "epoch": 1.5468117408906883,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.817729419703104e-05,
+      "loss": 2.4425,
+      "step": 6113
+    },
+    {
+      "epoch": 1.5470647773279351,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.8176450742240217e-05,
+      "loss": 2.5077,
+      "step": 6114
+    },
+    {
+      "epoch": 1.5473178137651822,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8175607287449393e-05,
+      "loss": 2.5132,
+      "step": 6115
+    },
+    {
+      "epoch": 1.5475708502024292,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.817476383265857e-05,
+      "loss": 2.5754,
+      "step": 6116
+    },
+    {
+      "epoch": 1.547823886639676,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.8173920377867747e-05,
+      "loss": 2.3869,
+      "step": 6117
+    },
+    {
+      "epoch": 1.5480769230769231,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8173076923076923e-05,
+      "loss": 2.211,
+      "step": 6118
+    },
+    {
+      "epoch": 1.5483299595141702,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.81722334682861e-05,
+      "loss": 2.4331,
+      "step": 6119
+    },
+    {
+      "epoch": 1.548582995951417,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.817139001349528e-05,
+      "loss": 2.4204,
+      "step": 6120
+    },
+    {
+      "epoch": 1.548582995951417,
+      "eval_loss": 2.5716726779937744,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 279.0058,
+      "eval_samples_per_second": 9.319,
+      "eval_steps_per_second": 9.319,
+      "step": 6120
+    },
+    {
+      "epoch": 1.5488360323886639,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8170546558704453e-05,
+      "loss": 2.4031,
+      "step": 6121
+    },
+    {
+      "epoch": 1.549089068825911,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.816970310391363e-05,
+      "loss": 2.4367,
+      "step": 6122
+    },
+    {
+      "epoch": 1.549342105263158,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.816885964912281e-05,
+      "loss": 2.4873,
+      "step": 6123
+    },
+    {
+      "epoch": 1.5495951417004048,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8168016194331983e-05,
+      "loss": 2.4683,
+      "step": 6124
+    },
+    {
+      "epoch": 1.5498481781376519,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.816717273954116e-05,
+      "loss": 2.4717,
+      "step": 6125
+    },
+    {
+      "epoch": 1.550101214574899,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.816632928475034e-05,
+      "loss": 2.4095,
+      "step": 6126
+    },
+    {
+      "epoch": 1.5503542510121457,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8165485829959516e-05,
+      "loss": 2.3384,
+      "step": 6127
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.816464237516869e-05,
+      "loss": 2.5154,
+      "step": 6128
+    },
+    {
+      "epoch": 1.5508603238866396,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.816379892037787e-05,
+      "loss": 2.4355,
+      "step": 6129
+    },
+    {
+      "epoch": 1.5511133603238867,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8162955465587046e-05,
+      "loss": 2.5339,
+      "step": 6130
+    },
+    {
+      "epoch": 1.5513663967611335,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8162112010796223e-05,
+      "loss": 2.4276,
+      "step": 6131
+    },
+    {
+      "epoch": 1.5516194331983806,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8161268556005396e-05,
+      "loss": 2.377,
+      "step": 6132
+    },
+    {
+      "epoch": 1.5518724696356276,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8160425101214576e-05,
+      "loss": 2.4343,
+      "step": 6133
+    },
+    {
+      "epoch": 1.5521255060728745,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8159581646423753e-05,
+      "loss": 2.4228,
+      "step": 6134
+    },
+    {
+      "epoch": 1.5523785425101213,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8158738191632926e-05,
+      "loss": 2.4952,
+      "step": 6135
+    },
+    {
+      "epoch": 1.5526315789473686,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.8157894736842106e-05,
+      "loss": 2.4038,
+      "step": 6136
+    },
+    {
+      "epoch": 1.5528846153846154,
+      "grad_norm": 3.71875,
+      "learning_rate": 2.8157051282051283e-05,
+      "loss": 2.5521,
+      "step": 6137
+    },
+    {
+      "epoch": 1.5531376518218623,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.815620782726046e-05,
+      "loss": 2.5388,
+      "step": 6138
+    },
+    {
+      "epoch": 1.5533906882591093,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.8155364372469636e-05,
+      "loss": 2.372,
+      "step": 6139
+    },
+    {
+      "epoch": 1.5536437246963564,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8154520917678813e-05,
+      "loss": 2.3926,
+      "step": 6140
+    },
+    {
+      "epoch": 1.5538967611336032,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.815367746288799e-05,
+      "loss": 2.3872,
+      "step": 6141
+    },
+    {
+      "epoch": 1.5541497975708503,
+      "grad_norm": 3.390625,
+      "learning_rate": 2.815283400809717e-05,
+      "loss": 2.4428,
+      "step": 6142
+    },
+    {
+      "epoch": 1.5544028340080973,
+      "grad_norm": 3.5,
+      "learning_rate": 2.8151990553306343e-05,
+      "loss": 2.438,
+      "step": 6143
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 3.25,
+      "learning_rate": 2.815114709851552e-05,
+      "loss": 2.5084,
+      "step": 6144
+    },
+    {
+      "epoch": 1.554908906882591,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.81503036437247e-05,
+      "loss": 2.5461,
+      "step": 6145
+    },
+    {
+      "epoch": 1.555161943319838,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8149460188933873e-05,
+      "loss": 2.3894,
+      "step": 6146
+    },
+    {
+      "epoch": 1.555414979757085,
+      "grad_norm": 3.125,
+      "learning_rate": 2.814861673414305e-05,
+      "loss": 2.5167,
+      "step": 6147
+    },
+    {
+      "epoch": 1.555668016194332,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8147773279352226e-05,
+      "loss": 2.4959,
+      "step": 6148
+    },
+    {
+      "epoch": 1.555921052631579,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8146929824561406e-05,
+      "loss": 2.5099,
+      "step": 6149
+    },
+    {
+      "epoch": 1.556174089068826,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.814608636977058e-05,
+      "loss": 2.4605,
+      "step": 6150
+    },
+    {
+      "epoch": 1.556174089068826,
+      "eval_loss": 2.571333646774292,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 278.8206,
+      "eval_samples_per_second": 9.325,
+      "eval_steps_per_second": 9.325,
+      "step": 6150
+    },
+    {
+      "epoch": 1.5564271255060729,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8145242914979756e-05,
+      "loss": 2.439,
+      "step": 6151
+    },
+    {
+      "epoch": 1.5566801619433197,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8144399460188936e-05,
+      "loss": 2.3731,
+      "step": 6152
+    },
+    {
+      "epoch": 1.5569331983805668,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.814355600539811e-05,
+      "loss": 2.5116,
+      "step": 6153
+    },
+    {
+      "epoch": 1.5571862348178138,
+      "grad_norm": 18.125,
+      "learning_rate": 2.8142712550607286e-05,
+      "loss": 2.4059,
+      "step": 6154
+    },
+    {
+      "epoch": 1.5574392712550607,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8141869095816466e-05,
+      "loss": 2.4037,
+      "step": 6155
+    },
+    {
+      "epoch": 1.5576923076923077,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8141025641025643e-05,
+      "loss": 2.5348,
+      "step": 6156
+    },
+    {
+      "epoch": 1.5579453441295548,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.8140182186234816e-05,
+      "loss": 2.4721,
+      "step": 6157
+    },
+    {
+      "epoch": 1.5581983805668016,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8139338731443996e-05,
+      "loss": 2.5604,
+      "step": 6158
+    },
+    {
+      "epoch": 1.5584514170040484,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.8138495276653173e-05,
+      "loss": 2.5931,
+      "step": 6159
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.813765182186235e-05,
+      "loss": 2.5373,
+      "step": 6160
+    },
+    {
+      "epoch": 1.5589574898785425,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8136808367071526e-05,
+      "loss": 2.4503,
+      "step": 6161
+    },
+    {
+      "epoch": 1.5592105263157894,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8135964912280703e-05,
+      "loss": 2.5011,
+      "step": 6162
+    },
+    {
+      "epoch": 1.5594635627530364,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.813512145748988e-05,
+      "loss": 2.4203,
+      "step": 6163
+    },
+    {
+      "epoch": 1.5597165991902835,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8134278002699056e-05,
+      "loss": 2.4299,
+      "step": 6164
+    },
+    {
+      "epoch": 1.5599696356275303,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8133434547908233e-05,
+      "loss": 2.3928,
+      "step": 6165
+    },
+    {
+      "epoch": 1.5602226720647774,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.813259109311741e-05,
+      "loss": 2.4795,
+      "step": 6166
+    },
+    {
+      "epoch": 1.5604757085020244,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8131747638326586e-05,
+      "loss": 2.5036,
+      "step": 6167
+    },
+    {
+      "epoch": 1.5607287449392713,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8130904183535763e-05,
+      "loss": 2.4468,
+      "step": 6168
+    },
+    {
+      "epoch": 1.560981781376518,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.813006072874494e-05,
+      "loss": 2.5998,
+      "step": 6169
+    },
+    {
+      "epoch": 1.5612348178137652,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8129217273954116e-05,
+      "loss": 2.4622,
+      "step": 6170
+    },
+    {
+      "epoch": 1.5614878542510122,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8128373819163296e-05,
+      "loss": 2.4419,
+      "step": 6171
+    },
+    {
+      "epoch": 1.561740890688259,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.812753036437247e-05,
+      "loss": 2.4662,
+      "step": 6172
+    },
+    {
+      "epoch": 1.561993927125506,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8126686909581646e-05,
+      "loss": 2.3547,
+      "step": 6173
+    },
+    {
+      "epoch": 1.5622469635627532,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8125843454790826e-05,
+      "loss": 2.3157,
+      "step": 6174
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8125e-05,
+      "loss": 2.4867,
+      "step": 6175
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8124156545209176e-05,
+      "loss": 2.4629,
+      "step": 6176
+    },
+    {
+      "epoch": 1.563006072874494,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8123313090418356e-05,
+      "loss": 2.4967,
+      "step": 6177
+    },
+    {
+      "epoch": 1.563259109311741,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8122469635627533e-05,
+      "loss": 2.5011,
+      "step": 6178
+    },
+    {
+      "epoch": 1.5635121457489878,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8121626180836706e-05,
+      "loss": 2.5055,
+      "step": 6179
+    },
+    {
+      "epoch": 1.5637651821862348,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.8120782726045886e-05,
+      "loss": 2.444,
+      "step": 6180
+    },
+    {
+      "epoch": 1.5637651821862348,
+      "eval_loss": 2.5714807510375977,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 280.8397,
+      "eval_samples_per_second": 9.258,
+      "eval_steps_per_second": 9.258,
+      "step": 6180
+    },
+    {
+      "epoch": 1.564018218623482,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.8119939271255063e-05,
+      "loss": 2.4688,
+      "step": 6181
+    },
+    {
+      "epoch": 1.5642712550607287,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.811909581646424e-05,
+      "loss": 2.4215,
+      "step": 6182
+    },
+    {
+      "epoch": 1.5645242914979756,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8118252361673416e-05,
+      "loss": 2.4446,
+      "step": 6183
+    },
+    {
+      "epoch": 1.5647773279352226,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8117408906882592e-05,
+      "loss": 2.4845,
+      "step": 6184
+    },
+    {
+      "epoch": 1.5650303643724697,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.811656545209177e-05,
+      "loss": 2.4886,
+      "step": 6185
+    },
+    {
+      "epoch": 1.5652834008097165,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8115721997300942e-05,
+      "loss": 2.4463,
+      "step": 6186
+    },
+    {
+      "epoch": 1.5655364372469636,
+      "grad_norm": 3.375,
+      "learning_rate": 2.8114878542510122e-05,
+      "loss": 2.5581,
+      "step": 6187
+    },
+    {
+      "epoch": 1.5657894736842106,
+      "grad_norm": 3.0,
+      "learning_rate": 2.81140350877193e-05,
+      "loss": 2.4786,
+      "step": 6188
+    },
+    {
+      "epoch": 1.5660425101214575,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8113191632928476e-05,
+      "loss": 2.5631,
+      "step": 6189
+    },
+    {
+      "epoch": 1.5662955465587043,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8112348178137652e-05,
+      "loss": 2.3914,
+      "step": 6190
+    },
+    {
+      "epoch": 1.5665485829959516,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.811150472334683e-05,
+      "loss": 2.4293,
+      "step": 6191
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.8110661268556006e-05,
+      "loss": 2.4577,
+      "step": 6192
+    },
+    {
+      "epoch": 1.5670546558704452,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8109817813765182e-05,
+      "loss": 2.5159,
+      "step": 6193
+    },
+    {
+      "epoch": 1.5673076923076923,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.810897435897436e-05,
+      "loss": 2.5436,
+      "step": 6194
+    },
+    {
+      "epoch": 1.5675607287449393,
+      "grad_norm": 6.375,
+      "learning_rate": 2.8108130904183536e-05,
+      "loss": 2.4041,
+      "step": 6195
+    },
+    {
+      "epoch": 1.5678137651821862,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8107287449392716e-05,
+      "loss": 2.4837,
+      "step": 6196
+    },
+    {
+      "epoch": 1.5680668016194332,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.810644399460189e-05,
+      "loss": 2.4149,
+      "step": 6197
+    },
+    {
+      "epoch": 1.5683198380566803,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8105600539811066e-05,
+      "loss": 2.4511,
+      "step": 6198
+    },
+    {
+      "epoch": 1.5685728744939271,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8104757085020246e-05,
+      "loss": 2.4105,
+      "step": 6199
+    },
+    {
+      "epoch": 1.568825910931174,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8103913630229422e-05,
+      "loss": 2.4625,
+      "step": 6200
+    },
+    {
+      "epoch": 1.569078947368421,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8103070175438596e-05,
+      "loss": 2.4365,
+      "step": 6201
+    },
+    {
+      "epoch": 1.569331983805668,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8102226720647776e-05,
+      "loss": 2.4345,
+      "step": 6202
+    },
+    {
+      "epoch": 1.569585020242915,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8101383265856952e-05,
+      "loss": 2.4257,
+      "step": 6203
+    },
+    {
+      "epoch": 1.569838056680162,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8100539811066126e-05,
+      "loss": 2.5511,
+      "step": 6204
+    },
+    {
+      "epoch": 1.570091093117409,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8099696356275302e-05,
+      "loss": 2.5249,
+      "step": 6205
+    },
+    {
+      "epoch": 1.5703441295546559,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8098852901484482e-05,
+      "loss": 2.4452,
+      "step": 6206
+    },
+    {
+      "epoch": 1.5705971659919027,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.809800944669366e-05,
+      "loss": 2.4855,
+      "step": 6207
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8097165991902832e-05,
+      "loss": 2.5473,
+      "step": 6208
+    },
+    {
+      "epoch": 1.5711032388663968,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8096322537112012e-05,
+      "loss": 2.4486,
+      "step": 6209
+    },
+    {
+      "epoch": 1.5713562753036436,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.809547908232119e-05,
+      "loss": 2.3649,
+      "step": 6210
+    },
+    {
+      "epoch": 1.5713562753036436,
+      "eval_loss": 2.5713274478912354,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 278.8038,
+      "eval_samples_per_second": 9.326,
+      "eval_steps_per_second": 9.326,
+      "step": 6210
+    },
+    {
+      "epoch": 1.5716093117408907,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8094635627530366e-05,
+      "loss": 2.463,
+      "step": 6211
+    },
+    {
+      "epoch": 1.5718623481781377,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8093792172739542e-05,
+      "loss": 2.2639,
+      "step": 6212
+    },
+    {
+      "epoch": 1.5721153846153846,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.809294871794872e-05,
+      "loss": 2.4422,
+      "step": 6213
+    },
+    {
+      "epoch": 1.5723684210526314,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.8092105263157895e-05,
+      "loss": 2.4813,
+      "step": 6214
+    },
+    {
+      "epoch": 1.5726214574898787,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8091261808367072e-05,
+      "loss": 2.4334,
+      "step": 6215
+    },
+    {
+      "epoch": 1.5728744939271255,
+      "grad_norm": 3.0,
+      "learning_rate": 2.809041835357625e-05,
+      "loss": 2.497,
+      "step": 6216
+    },
+    {
+      "epoch": 1.5731275303643724,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.8089574898785425e-05,
+      "loss": 2.508,
+      "step": 6217
+    },
+    {
+      "epoch": 1.5733805668016194,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8088731443994605e-05,
+      "loss": 2.4699,
+      "step": 6218
+    },
+    {
+      "epoch": 1.5736336032388665,
+      "grad_norm": 3.0,
+      "learning_rate": 2.808788798920378e-05,
+      "loss": 2.417,
+      "step": 6219
+    },
+    {
+      "epoch": 1.5738866396761133,
+      "grad_norm": 3.375,
+      "learning_rate": 2.8087044534412955e-05,
+      "loss": 2.5163,
+      "step": 6220
+    },
+    {
+      "epoch": 1.5741396761133604,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.8086201079622132e-05,
+      "loss": 2.4801,
+      "step": 6221
+    },
+    {
+      "epoch": 1.5743927125506074,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8085357624831312e-05,
+      "loss": 2.4928,
+      "step": 6222
+    },
+    {
+      "epoch": 1.5746457489878543,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.8084514170040485e-05,
+      "loss": 2.432,
+      "step": 6223
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.8083670715249662e-05,
+      "loss": 2.4199,
+      "step": 6224
+    },
+    {
+      "epoch": 1.5751518218623481,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8082827260458842e-05,
+      "loss": 2.4333,
+      "step": 6225
+    },
+    {
+      "epoch": 1.5754048582995952,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8081983805668015e-05,
+      "loss": 2.4026,
+      "step": 6226
+    },
+    {
+      "epoch": 1.575657894736842,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8081140350877192e-05,
+      "loss": 2.4125,
+      "step": 6227
+    },
+    {
+      "epoch": 1.575910931174089,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8080296896086372e-05,
+      "loss": 2.4525,
+      "step": 6228
+    },
+    {
+      "epoch": 1.5761639676113361,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.807945344129555e-05,
+      "loss": 2.4189,
+      "step": 6229
+    },
+    {
+      "epoch": 1.576417004048583,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8078609986504722e-05,
+      "loss": 2.4201,
+      "step": 6230
+    },
+    {
+      "epoch": 1.5766700404858298,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8077766531713902e-05,
+      "loss": 2.4248,
+      "step": 6231
+    },
+    {
+      "epoch": 1.5769230769230769,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.807692307692308e-05,
+      "loss": 2.54,
+      "step": 6232
+    },
+    {
+      "epoch": 1.577176113360324,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8076079622132252e-05,
+      "loss": 2.5096,
+      "step": 6233
+    },
+    {
+      "epoch": 1.5774291497975708,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8075236167341432e-05,
+      "loss": 2.3816,
+      "step": 6234
+    },
+    {
+      "epoch": 1.5776821862348178,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.807439271255061e-05,
+      "loss": 2.5086,
+      "step": 6235
+    },
+    {
+      "epoch": 1.5779352226720649,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8073549257759785e-05,
+      "loss": 2.4105,
+      "step": 6236
+    },
+    {
+      "epoch": 1.5781882591093117,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8072705802968962e-05,
+      "loss": 2.4394,
+      "step": 6237
+    },
+    {
+      "epoch": 1.5784412955465585,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.807186234817814e-05,
+      "loss": 2.4589,
+      "step": 6238
+    },
+    {
+      "epoch": 1.5786943319838058,
+      "grad_norm": 2.625,
+      "learning_rate": 2.8071018893387315e-05,
+      "loss": 2.4057,
+      "step": 6239
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8070175438596492e-05,
+      "loss": 2.466,
+      "step": 6240
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "eval_loss": 2.5700247287750244,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 279.049,
+      "eval_samples_per_second": 9.317,
+      "eval_steps_per_second": 9.317,
+      "step": 6240
+    },
+    {
+      "epoch": 1.5792004048582995,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.806933198380567e-05,
+      "loss": 2.4518,
+      "step": 6241
+    },
+    {
+      "epoch": 1.5794534412955465,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.8068488529014845e-05,
+      "loss": 2.4694,
+      "step": 6242
+    },
+    {
+      "epoch": 1.5797064777327936,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8067645074224022e-05,
+      "loss": 2.4641,
+      "step": 6243
+    },
+    {
+      "epoch": 1.5799595141700404,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.80668016194332e-05,
+      "loss": 2.4553,
+      "step": 6244
+    },
+    {
+      "epoch": 1.5802125506072875,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8065958164642375e-05,
+      "loss": 2.5999,
+      "step": 6245
+    },
+    {
+      "epoch": 1.5804655870445345,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8065114709851552e-05,
+      "loss": 2.4874,
+      "step": 6246
+    },
+    {
+      "epoch": 1.5807186234817814,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8064271255060732e-05,
+      "loss": 2.4377,
+      "step": 6247
+    },
+    {
+      "epoch": 1.5809716599190282,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8063427800269905e-05,
+      "loss": 2.4574,
+      "step": 6248
+    },
+    {
+      "epoch": 1.5812246963562753,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.8062584345479082e-05,
+      "loss": 2.4603,
+      "step": 6249
+    },
+    {
+      "epoch": 1.5814777327935223,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8061740890688262e-05,
+      "loss": 2.3561,
+      "step": 6250
+    },
+    {
+      "epoch": 1.5817307692307692,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.806089743589744e-05,
+      "loss": 2.3797,
+      "step": 6251
+    },
+    {
+      "epoch": 1.5819838056680162,
+      "grad_norm": 3.5,
+      "learning_rate": 2.8060053981106612e-05,
+      "loss": 2.5423,
+      "step": 6252
+    },
+    {
+      "epoch": 1.5822368421052633,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8059210526315792e-05,
+      "loss": 2.5127,
+      "step": 6253
+    },
+    {
+      "epoch": 1.58248987854251,
+      "grad_norm": 2.875,
+      "learning_rate": 2.805836707152497e-05,
+      "loss": 2.5447,
+      "step": 6254
+    },
+    {
+      "epoch": 1.582742914979757,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.805752361673414e-05,
+      "loss": 2.3854,
+      "step": 6255
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8056680161943322e-05,
+      "loss": 2.3783,
+      "step": 6256
+    },
+    {
+      "epoch": 1.583248987854251,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.80558367071525e-05,
+      "loss": 2.4295,
+      "step": 6257
+    },
+    {
+      "epoch": 1.583502024291498,
+      "grad_norm": 2.875,
+      "learning_rate": 2.8054993252361675e-05,
+      "loss": 2.4035,
+      "step": 6258
+    },
+    {
+      "epoch": 1.583755060728745,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.8054149797570848e-05,
+      "loss": 2.5616,
+      "step": 6259
+    },
+    {
+      "epoch": 1.584008097165992,
+      "grad_norm": 2.875,
+      "learning_rate": 2.805330634278003e-05,
+      "loss": 2.2532,
+      "step": 6260
+    },
+    {
+      "epoch": 1.5842611336032388,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8052462887989205e-05,
+      "loss": 2.3478,
+      "step": 6261
+    },
+    {
+      "epoch": 1.5845141700404857,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.805161943319838e-05,
+      "loss": 2.3673,
+      "step": 6262
+    },
+    {
+      "epoch": 1.584767206477733,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.8050775978407558e-05,
+      "loss": 2.5199,
+      "step": 6263
+    },
+    {
+      "epoch": 1.5850202429149798,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8049932523616735e-05,
+      "loss": 2.4436,
+      "step": 6264
+    },
+    {
+      "epoch": 1.5852732793522266,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.804908906882591e-05,
+      "loss": 2.4338,
+      "step": 6265
+    },
+    {
+      "epoch": 1.5855263157894737,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8048245614035088e-05,
+      "loss": 2.512,
+      "step": 6266
+    },
+    {
+      "epoch": 1.5857793522267207,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8047402159244265e-05,
+      "loss": 2.4056,
+      "step": 6267
+    },
+    {
+      "epoch": 1.5860323886639676,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.804655870445344e-05,
+      "loss": 2.3638,
+      "step": 6268
+    },
+    {
+      "epoch": 1.5862854251012146,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.804571524966262e-05,
+      "loss": 2.5872,
+      "step": 6269
+    },
+    {
+      "epoch": 1.5865384615384617,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8044871794871795e-05,
+      "loss": 2.4778,
+      "step": 6270
+    },
+    {
+      "epoch": 1.5865384615384617,
+      "eval_loss": 2.567826509475708,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 277.3503,
+      "eval_samples_per_second": 9.374,
+      "eval_steps_per_second": 9.374,
+      "step": 6270
+    },
+    {
+      "epoch": 1.5867914979757085,
+      "grad_norm": 3.25,
+      "learning_rate": 2.804402834008097e-05,
+      "loss": 2.5323,
+      "step": 6271
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.804318488529015e-05,
+      "loss": 2.4515,
+      "step": 6272
+    },
+    {
+      "epoch": 1.5872975708502024,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8042341430499325e-05,
+      "loss": 2.458,
+      "step": 6273
+    },
+    {
+      "epoch": 1.5875506072874495,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.80414979757085e-05,
+      "loss": 2.4838,
+      "step": 6274
+    },
+    {
+      "epoch": 1.5878036437246963,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8040654520917678e-05,
+      "loss": 2.3637,
+      "step": 6275
+    },
+    {
+      "epoch": 1.5880566801619433,
+      "grad_norm": 17.75,
+      "learning_rate": 2.8039811066126858e-05,
+      "loss": 2.5389,
+      "step": 6276
+    },
+    {
+      "epoch": 1.5883097165991904,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.803896761133603e-05,
+      "loss": 2.4317,
+      "step": 6277
+    },
+    {
+      "epoch": 1.5885627530364372,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8038124156545208e-05,
+      "loss": 2.4757,
+      "step": 6278
+    },
+    {
+      "epoch": 1.588815789473684,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8037280701754388e-05,
+      "loss": 2.3486,
+      "step": 6279
+    },
+    {
+      "epoch": 1.5890688259109311,
+      "grad_norm": 2.75,
+      "learning_rate": 2.8036437246963565e-05,
+      "loss": 2.3351,
+      "step": 6280
+    },
+    {
+      "epoch": 1.5893218623481782,
+      "grad_norm": 3.125,
+      "learning_rate": 2.8035593792172738e-05,
+      "loss": 2.4752,
+      "step": 6281
+    },
+    {
+      "epoch": 1.589574898785425,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8034750337381918e-05,
+      "loss": 2.3867,
+      "step": 6282
+    },
+    {
+      "epoch": 1.589827935222672,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8033906882591095e-05,
+      "loss": 2.2967,
+      "step": 6283
+    },
+    {
+      "epoch": 1.5900809716599191,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8033063427800268e-05,
+      "loss": 2.4212,
+      "step": 6284
+    },
+    {
+      "epoch": 1.590334008097166,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8032219973009448e-05,
+      "loss": 2.3679,
+      "step": 6285
+    },
+    {
+      "epoch": 1.5905870445344128,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.8031376518218625e-05,
+      "loss": 2.4585,
+      "step": 6286
+    },
+    {
+      "epoch": 1.59084008097166,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.80305330634278e-05,
+      "loss": 2.4583,
+      "step": 6287
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.8029689608636978e-05,
+      "loss": 2.5018,
+      "step": 6288
+    },
+    {
+      "epoch": 1.5913461538461537,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.8028846153846155e-05,
+      "loss": 2.4699,
+      "step": 6289
+    },
+    {
+      "epoch": 1.5915991902834008,
+      "grad_norm": 2.875,
+      "learning_rate": 2.802800269905533e-05,
+      "loss": 2.4547,
+      "step": 6290
+    },
+    {
+      "epoch": 1.5918522267206479,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.802715924426451e-05,
+      "loss": 2.49,
+      "step": 6291
+    },
+    {
+      "epoch": 1.5921052631578947,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.8026315789473685e-05,
+      "loss": 2.5144,
+      "step": 6292
+    },
+    {
+      "epoch": 1.5923582995951417,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.802547233468286e-05,
+      "loss": 2.5318,
+      "step": 6293
+    },
+    {
+      "epoch": 1.5926113360323888,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.8024628879892038e-05,
+      "loss": 2.3804,
+      "step": 6294
+    },
+    {
+      "epoch": 1.5928643724696356,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8023785425101215e-05,
+      "loss": 2.4682,
+      "step": 6295
+    },
+    {
+      "epoch": 1.5931174089068825,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.802294197031039e-05,
+      "loss": 2.4384,
+      "step": 6296
+    },
+    {
+      "epoch": 1.5933704453441295,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8022098515519568e-05,
+      "loss": 2.4024,
+      "step": 6297
+    },
+    {
+      "epoch": 1.5936234817813766,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8021255060728748e-05,
+      "loss": 2.4662,
+      "step": 6298
+    },
+    {
+      "epoch": 1.5938765182186234,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.802041160593792e-05,
+      "loss": 2.4317,
+      "step": 6299
+    },
+    {
+      "epoch": 1.5941295546558705,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8019568151147098e-05,
+      "loss": 2.4354,
+      "step": 6300
+    },
+    {
+      "epoch": 1.5941295546558705,
+      "eval_loss": 2.5685713291168213,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 278.8867,
+      "eval_samples_per_second": 9.323,
+      "eval_steps_per_second": 9.323,
+      "step": 6300
+    },
+    {
+      "epoch": 1.5943825910931175,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8018724696356278e-05,
+      "loss": 2.4527,
+      "step": 6301
+    },
+    {
+      "epoch": 1.5946356275303644,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.801788124156545e-05,
+      "loss": 2.4359,
+      "step": 6302
+    },
+    {
+      "epoch": 1.5948886639676112,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.8017037786774628e-05,
+      "loss": 2.4465,
+      "step": 6303
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.8016194331983808e-05,
+      "loss": 2.5358,
+      "step": 6304
+    },
+    {
+      "epoch": 1.5953947368421053,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.8015350877192984e-05,
+      "loss": 2.3751,
+      "step": 6305
+    },
+    {
+      "epoch": 1.5956477732793521,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.8014507422402158e-05,
+      "loss": 2.4536,
+      "step": 6306
+    },
+    {
+      "epoch": 1.5959008097165992,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.8013663967611338e-05,
+      "loss": 2.4306,
+      "step": 6307
+    },
+    {
+      "epoch": 1.5961538461538463,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8012820512820514e-05,
+      "loss": 2.4693,
+      "step": 6308
+    },
+    {
+      "epoch": 1.596406882591093,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.801197705802969e-05,
+      "loss": 2.3714,
+      "step": 6309
+    },
+    {
+      "epoch": 1.59665991902834,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.8011133603238868e-05,
+      "loss": 2.4221,
+      "step": 6310
+    },
+    {
+      "epoch": 1.5969129554655872,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.8010290148448044e-05,
+      "loss": 2.4656,
+      "step": 6311
+    },
+    {
+      "epoch": 1.597165991902834,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.800944669365722e-05,
+      "loss": 2.3662,
+      "step": 6312
+    },
+    {
+      "epoch": 1.5974190283400809,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8008603238866394e-05,
+      "loss": 2.4424,
+      "step": 6313
+    },
+    {
+      "epoch": 1.597672064777328,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.8007759784075574e-05,
+      "loss": 2.4208,
+      "step": 6314
+    },
+    {
+      "epoch": 1.597925101214575,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.800691632928475e-05,
+      "loss": 2.3127,
+      "step": 6315
+    },
+    {
+      "epoch": 1.5981781376518218,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.8006072874493928e-05,
+      "loss": 2.474,
+      "step": 6316
+    },
+    {
+      "epoch": 1.5984311740890689,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.8005229419703104e-05,
+      "loss": 2.5204,
+      "step": 6317
+    },
+    {
+      "epoch": 1.598684210526316,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.800438596491228e-05,
+      "loss": 2.4834,
+      "step": 6318
+    },
+    {
+      "epoch": 1.5989372469635628,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.8003542510121458e-05,
+      "loss": 2.4737,
+      "step": 6319
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8002699055330638e-05,
+      "loss": 2.4592,
+      "step": 6320
+    },
+    {
+      "epoch": 1.5994433198380567,
+      "grad_norm": 3.375,
+      "learning_rate": 2.800185560053981e-05,
+      "loss": 2.4838,
+      "step": 6321
+    },
+    {
+      "epoch": 1.5996963562753037,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8001012145748988e-05,
+      "loss": 2.5266,
+      "step": 6322
+    },
+    {
+      "epoch": 1.5999493927125505,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.8000168690958168e-05,
+      "loss": 2.3435,
+      "step": 6323
+    },
+    {
+      "epoch": 1.6002024291497976,
+      "grad_norm": 2.75,
+      "learning_rate": 2.799932523616734e-05,
+      "loss": 2.4836,
+      "step": 6324
+    },
+    {
+      "epoch": 1.6004554655870447,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.7998481781376518e-05,
+      "loss": 2.4888,
+      "step": 6325
+    },
+    {
+      "epoch": 1.6007085020242915,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7997638326585698e-05,
+      "loss": 2.3811,
+      "step": 6326
+    },
+    {
+      "epoch": 1.6009615384615383,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.7996794871794874e-05,
+      "loss": 2.5088,
+      "step": 6327
+    },
+    {
+      "epoch": 1.6012145748987854,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.7995951417004048e-05,
+      "loss": 2.4739,
+      "step": 6328
+    },
+    {
+      "epoch": 1.6014676113360324,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.7995107962213224e-05,
+      "loss": 2.4048,
+      "step": 6329
+    },
+    {
+      "epoch": 1.6017206477732793,
+      "grad_norm": 2.75,
+      "learning_rate": 2.7994264507422404e-05,
+      "loss": 2.4877,
+      "step": 6330
+    },
+    {
+      "epoch": 1.6017206477732793,
+      "eval_loss": 2.5640411376953125,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 277.9985,
+      "eval_samples_per_second": 9.353,
+      "eval_steps_per_second": 9.353,
+      "step": 6330
+    },
+    {
+      "epoch": 1.6019736842105263,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.799342105263158e-05,
+      "loss": 2.4432,
+      "step": 6331
+    },
+    {
+      "epoch": 1.6022267206477734,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7992577597840754e-05,
+      "loss": 2.4146,
+      "step": 6332
+    },
+    {
+      "epoch": 1.6024797570850202,
+      "grad_norm": 2.875,
+      "learning_rate": 2.7991734143049934e-05,
+      "loss": 2.4531,
+      "step": 6333
+    },
+    {
+      "epoch": 1.602732793522267,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.799089068825911e-05,
+      "loss": 2.3144,
+      "step": 6334
+    },
+    {
+      "epoch": 1.6029858299595143,
+      "grad_norm": 2.875,
+      "learning_rate": 2.7990047233468284e-05,
+      "loss": 2.416,
+      "step": 6335
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.7989203778677464e-05,
+      "loss": 2.4803,
+      "step": 6336
+    },
+    {
+      "epoch": 1.603491902834008,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.798836032388664e-05,
+      "loss": 2.4223,
+      "step": 6337
+    },
+    {
+      "epoch": 1.603744939271255,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.7987516869095817e-05,
+      "loss": 2.4605,
+      "step": 6338
+    },
+    {
+      "epoch": 1.603997975708502,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.7986673414304994e-05,
+      "loss": 2.4828,
+      "step": 6339
+    },
+    {
+      "epoch": 1.604251012145749,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.798582995951417e-05,
+      "loss": 2.2471,
+      "step": 6340
+    },
+    {
+      "epoch": 1.604504048582996,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.7984986504723347e-05,
+      "loss": 2.573,
+      "step": 6341
+    },
+    {
+      "epoch": 1.604757085020243,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.7984143049932524e-05,
+      "loss": 2.5136,
+      "step": 6342
+    },
+    {
+      "epoch": 1.60501012145749,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.79832995951417e-05,
+      "loss": 2.453,
+      "step": 6343
+    },
+    {
+      "epoch": 1.6052631578947367,
+      "grad_norm": 3.5,
+      "learning_rate": 2.7982456140350877e-05,
+      "loss": 2.4968,
+      "step": 6344
+    },
+    {
+      "epoch": 1.6055161943319838,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.7981612685560057e-05,
+      "loss": 2.53,
+      "step": 6345
+    },
+    {
+      "epoch": 1.6057692307692308,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.798076923076923e-05,
+      "loss": 2.4765,
+      "step": 6346
+    },
+    {
+      "epoch": 1.6060222672064777,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.7979925775978407e-05,
+      "loss": 2.3718,
+      "step": 6347
+    },
+    {
+      "epoch": 1.6062753036437247,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.7979082321187584e-05,
+      "loss": 2.287,
+      "step": 6348
+    },
+    {
+      "epoch": 1.6065283400809718,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.7978238866396764e-05,
+      "loss": 2.5132,
+      "step": 6349
+    },
+    {
+      "epoch": 1.6067813765182186,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.7977395411605937e-05,
+      "loss": 2.4844,
+      "step": 6350
+    },
+    {
+      "epoch": 1.6070344129554655,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.7976551956815114e-05,
+      "loss": 2.4029,
+      "step": 6351
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.7975708502024294e-05,
+      "loss": 2.2693,
+      "step": 6352
+    },
+    {
+      "epoch": 1.6075404858299596,
+      "grad_norm": 3.421875,
+      "learning_rate": 2.7974865047233467e-05,
+      "loss": 2.4981,
+      "step": 6353
+    },
+    {
+      "epoch": 1.6077935222672064,
+      "grad_norm": 3.375,
+      "learning_rate": 2.7974021592442644e-05,
+      "loss": 2.4857,
+      "step": 6354
+    },
+    {
+      "epoch": 1.6080465587044535,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.7973178137651824e-05,
+      "loss": 2.5015,
+      "step": 6355
+    },
+    {
+      "epoch": 1.6082995951417005,
+      "grad_norm": 2.875,
+      "learning_rate": 2.7972334682861e-05,
+      "loss": 2.2937,
+      "step": 6356
+    },
+    {
+      "epoch": 1.6085526315789473,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.7971491228070174e-05,
+      "loss": 2.5307,
+      "step": 6357
+    },
+    {
+      "epoch": 1.6088056680161942,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.7970647773279354e-05,
+      "loss": 2.3655,
+      "step": 6358
+    },
+    {
+      "epoch": 1.6090587044534415,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.796980431848853e-05,
+      "loss": 2.4223,
+      "step": 6359
+    },
+    {
+      "epoch": 1.6093117408906883,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.7968960863697707e-05,
+      "loss": 2.3032,
+      "step": 6360
+    },
+    {
+      "epoch": 1.6093117408906883,
+      "eval_loss": 2.567202568054199,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 279.1434,
+      "eval_samples_per_second": 9.314,
+      "eval_steps_per_second": 9.314,
+      "step": 6360
+    },
+    {
+      "epoch": 1.6095647773279351,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.7968117408906884e-05,
+      "loss": 2.3739,
+      "step": 6361
+    },
+    {
+      "epoch": 1.6098178137651822,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.796727395411606e-05,
+      "loss": 2.4624,
+      "step": 6362
+    },
+    {
+      "epoch": 1.6100708502024292,
+      "grad_norm": 3.0,
+      "learning_rate": 2.7966430499325237e-05,
+      "loss": 2.4969,
+      "step": 6363
+    },
+    {
+      "epoch": 1.610323886639676,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.7965587044534414e-05,
+      "loss": 2.5055,
+      "step": 6364
+    },
+    {
+      "epoch": 1.6105769230769231,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.796474358974359e-05,
+      "loss": 2.4991,
+      "step": 6365
+    },
+    {
+      "epoch": 1.6108299595141702,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.7963900134952767e-05,
+      "loss": 2.4742,
+      "step": 6366
+    },
+    {
+      "epoch": 1.611082995951417,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.7963056680161944e-05,
+      "loss": 2.4578,
+      "step": 6367
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.796221322537112e-05,
+      "loss": 2.4325,
+      "step": 6368
+    },
+    {
+      "epoch": 1.611589068825911,
+      "grad_norm": 2.875,
+      "learning_rate": 2.7961369770580297e-05,
+      "loss": 2.3972,
+      "step": 6369
+    },
+    {
+      "epoch": 1.611842105263158,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.7960526315789474e-05,
+      "loss": 2.4278,
+      "step": 6370
+    },
+    {
+      "epoch": 1.6120951417004048,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7959682860998654e-05,
+      "loss": 2.4628,
+      "step": 6371
+    },
+    {
+      "epoch": 1.6123481781376519,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.7958839406207827e-05,
+      "loss": 2.3827,
+      "step": 6372
+    },
+    {
+      "epoch": 1.612601214574899,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.7957995951417004e-05,
+      "loss": 2.5676,
+      "step": 6373
+    },
+    {
+      "epoch": 1.6128542510121457,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7957152496626184e-05,
+      "loss": 2.3664,
+      "step": 6374
+    },
+    {
+      "epoch": 1.6131072874493926,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.7956309041835357e-05,
+      "loss": 2.5169,
+      "step": 6375
+    },
+    {
+      "epoch": 1.6133603238866396,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.7955465587044534e-05,
+      "loss": 2.4675,
+      "step": 6376
+    },
+    {
+      "epoch": 1.6136133603238867,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.7954622132253714e-05,
+      "loss": 2.5055,
+      "step": 6377
+    },
+    {
+      "epoch": 1.6138663967611335,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.795377867746289e-05,
+      "loss": 2.4652,
+      "step": 6378
+    },
+    {
+      "epoch": 1.6141194331983806,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.7952935222672064e-05,
+      "loss": 2.3744,
+      "step": 6379
+    },
+    {
+      "epoch": 1.6143724696356276,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.7952091767881244e-05,
+      "loss": 2.4467,
+      "step": 6380
+    },
+    {
+      "epoch": 1.6146255060728745,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.795124831309042e-05,
+      "loss": 2.3649,
+      "step": 6381
+    },
+    {
+      "epoch": 1.6148785425101213,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.7950404858299594e-05,
+      "loss": 2.4855,
+      "step": 6382
+    },
+    {
+      "epoch": 1.6151315789473686,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.794956140350877e-05,
+      "loss": 2.2997,
+      "step": 6383
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.794871794871795e-05,
+      "loss": 2.3985,
+      "step": 6384
+    },
+    {
+      "epoch": 1.6156376518218623,
+      "grad_norm": 3.125,
+      "learning_rate": 2.7947874493927127e-05,
+      "loss": 2.3947,
+      "step": 6385
+    },
+    {
+      "epoch": 1.6158906882591093,
+      "grad_norm": 3.0,
+      "learning_rate": 2.79470310391363e-05,
+      "loss": 2.4584,
+      "step": 6386
+    },
+    {
+      "epoch": 1.6161437246963564,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.794618758434548e-05,
+      "loss": 2.3538,
+      "step": 6387
+    },
+    {
+      "epoch": 1.6163967611336032,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.7945344129554657e-05,
+      "loss": 2.5055,
+      "step": 6388
+    },
+    {
+      "epoch": 1.6166497975708503,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.7944500674763834e-05,
+      "loss": 2.4817,
+      "step": 6389
+    },
+    {
+      "epoch": 1.6169028340080973,
+      "grad_norm": 3.0,
+      "learning_rate": 2.794365721997301e-05,
+      "loss": 2.4882,
+      "step": 6390
+    },
+    {
+      "epoch": 1.6169028340080973,
+      "eval_loss": 2.5633435249328613,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 278.3132,
+      "eval_samples_per_second": 9.342,
+      "eval_steps_per_second": 9.342,
+      "step": 6390
+    },
+    {
+      "epoch": 1.6171558704453441,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.7942813765182187e-05,
+      "loss": 2.418,
+      "step": 6391
+    },
+    {
+      "epoch": 1.617408906882591,
+      "grad_norm": 3.0,
+      "learning_rate": 2.7941970310391363e-05,
+      "loss": 2.4857,
+      "step": 6392
+    },
+    {
+      "epoch": 1.617661943319838,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.794112685560054e-05,
+      "loss": 2.4495,
+      "step": 6393
+    },
+    {
+      "epoch": 1.617914979757085,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.7940283400809717e-05,
+      "loss": 2.5025,
+      "step": 6394
+    },
+    {
+      "epoch": 1.618168016194332,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.7939439946018893e-05,
+      "loss": 2.5278,
+      "step": 6395
+    },
+    {
+      "epoch": 1.618421052631579,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.7938596491228073e-05,
+      "loss": 2.536,
+      "step": 6396
+    },
+    {
+      "epoch": 1.618674089068826,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.7937753036437247e-05,
+      "loss": 2.4805,
+      "step": 6397
+    },
+    {
+      "epoch": 1.6189271255060729,
+      "grad_norm": 2.75,
+      "learning_rate": 2.7936909581646423e-05,
+      "loss": 2.416,
+      "step": 6398
+    },
+    {
+      "epoch": 1.6191801619433197,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.7936066126855603e-05,
+      "loss": 2.5395,
+      "step": 6399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.793522267206478e-05,
+      "loss": 2.5021,
+      "step": 6400
+    },
+    {
+      "epoch": 1.6196862348178138,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7934379217273953e-05,
+      "loss": 2.5286,
+      "step": 6401
+    },
+    {
+      "epoch": 1.6199392712550607,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.793353576248313e-05,
+      "loss": 2.5184,
+      "step": 6402
+    },
+    {
+      "epoch": 1.6201923076923077,
+      "grad_norm": 3.0,
+      "learning_rate": 2.793269230769231e-05,
+      "loss": 2.5265,
+      "step": 6403
+    },
+    {
+      "epoch": 1.6204453441295548,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.7931848852901483e-05,
+      "loss": 2.4257,
+      "step": 6404
+    },
+    {
+      "epoch": 1.6206983805668016,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.793100539811066e-05,
+      "loss": 2.4936,
+      "step": 6405
+    },
+    {
+      "epoch": 1.6209514170040484,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.793016194331984e-05,
+      "loss": 2.415,
+      "step": 6406
+    },
+    {
+      "epoch": 1.6212044534412957,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.7929318488529017e-05,
+      "loss": 2.4464,
+      "step": 6407
+    },
+    {
+      "epoch": 1.6214574898785425,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.792847503373819e-05,
+      "loss": 2.3375,
+      "step": 6408
+    },
+    {
+      "epoch": 1.6217105263157894,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.792763157894737e-05,
+      "loss": 2.4626,
+      "step": 6409
+    },
+    {
+      "epoch": 1.6219635627530364,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.7926788124156547e-05,
+      "loss": 2.412,
+      "step": 6410
+    },
+    {
+      "epoch": 1.6222165991902835,
+      "grad_norm": 3.0,
+      "learning_rate": 2.7925944669365723e-05,
+      "loss": 2.5045,
+      "step": 6411
+    },
+    {
+      "epoch": 1.6224696356275303,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.79251012145749e-05,
+      "loss": 2.3349,
+      "step": 6412
+    },
+    {
+      "epoch": 1.6227226720647774,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.7924257759784077e-05,
+      "loss": 2.5079,
+      "step": 6413
+    },
+    {
+      "epoch": 1.6229757085020244,
+      "grad_norm": 3.0,
+      "learning_rate": 2.7923414304993253e-05,
+      "loss": 2.5167,
+      "step": 6414
+    },
+    {
+      "epoch": 1.6232287449392713,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.792257085020243e-05,
+      "loss": 2.49,
+      "step": 6415
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.7921727395411607e-05,
+      "loss": 2.4006,
+      "step": 6416
+    },
+    {
+      "epoch": 1.6237348178137652,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.7920883940620783e-05,
+      "loss": 2.4257,
+      "step": 6417
+    },
+    {
+      "epoch": 1.6239878542510122,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7920040485829963e-05,
+      "loss": 2.3887,
+      "step": 6418
+    },
+    {
+      "epoch": 1.624240890688259,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.7919197031039137e-05,
+      "loss": 2.2757,
+      "step": 6419
+    },
+    {
+      "epoch": 1.624493927125506,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.7918353576248313e-05,
+      "loss": 2.5356,
+      "step": 6420
+    },
+    {
+      "epoch": 1.624493927125506,
+      "eval_loss": 2.563068389892578,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 291.2488,
+      "eval_samples_per_second": 8.927,
+      "eval_steps_per_second": 8.927,
+      "step": 6420
+    },
+    {
+      "epoch": 1.6247469635627532,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.791751012145749e-05,
+      "loss": 2.3253,
+      "step": 6421
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 3.0,
+      "learning_rate": 2.7916666666666666e-05,
+      "loss": 2.512,
+      "step": 6422
+    },
+    {
+      "epoch": 1.6252530364372468,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.7915823211875843e-05,
+      "loss": 2.4286,
+      "step": 6423
+    },
+    {
+      "epoch": 1.625506072874494,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.791497975708502e-05,
+      "loss": 2.3305,
+      "step": 6424
+    },
+    {
+      "epoch": 1.625759109311741,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.79141363022942e-05,
+      "loss": 2.2894,
+      "step": 6425
+    },
+    {
+      "epoch": 1.6260121457489878,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.7913292847503373e-05,
+      "loss": 2.3919,
+      "step": 6426
+    },
+    {
+      "epoch": 1.6262651821862348,
+      "grad_norm": 2.875,
+      "learning_rate": 2.791244939271255e-05,
+      "loss": 2.4501,
+      "step": 6427
+    },
+    {
+      "epoch": 1.626518218623482,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.791160593792173e-05,
+      "loss": 2.4409,
+      "step": 6428
+    },
+    {
+      "epoch": 1.6267712550607287,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.7910762483130906e-05,
+      "loss": 2.3609,
+      "step": 6429
+    },
+    {
+      "epoch": 1.6270242914979756,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.790991902834008e-05,
+      "loss": 2.3663,
+      "step": 6430
+    },
+    {
+      "epoch": 1.6272773279352226,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.790907557354926e-05,
+      "loss": 2.5333,
+      "step": 6431
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.7908232118758436e-05,
+      "loss": 2.3919,
+      "step": 6432
+    },
+    {
+      "epoch": 1.6277834008097165,
+      "grad_norm": 3.0,
+      "learning_rate": 2.790738866396761e-05,
+      "loss": 2.4811,
+      "step": 6433
+    },
+    {
+      "epoch": 1.6280364372469636,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.790654520917679e-05,
+      "loss": 2.3939,
+      "step": 6434
+    },
+    {
+      "epoch": 1.6282894736842106,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.7905701754385966e-05,
+      "loss": 2.3335,
+      "step": 6435
+    },
+    {
+      "epoch": 1.6285425101214575,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.7904858299595143e-05,
+      "loss": 2.4868,
+      "step": 6436
+    },
+    {
+      "epoch": 1.6287955465587043,
+      "grad_norm": 3.0,
+      "learning_rate": 2.790401484480432e-05,
+      "loss": 2.4315,
+      "step": 6437
+    },
+    {
+      "epoch": 1.6290485829959516,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.7903171390013496e-05,
+      "loss": 2.3852,
+      "step": 6438
+    },
+    {
+      "epoch": 1.6293016194331984,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.7902327935222673e-05,
+      "loss": 2.4902,
+      "step": 6439
+    },
+    {
+      "epoch": 1.6295546558704452,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.790148448043185e-05,
+      "loss": 2.4227,
+      "step": 6440
+    },
+    {
+      "epoch": 1.6298076923076923,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.7900641025641026e-05,
+      "loss": 2.4492,
+      "step": 6441
+    },
+    {
+      "epoch": 1.6300607287449393,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.7899797570850203e-05,
+      "loss": 2.3384,
+      "step": 6442
+    },
+    {
+      "epoch": 1.6303137651821862,
+      "grad_norm": 2.875,
+      "learning_rate": 2.789895411605938e-05,
+      "loss": 2.4395,
+      "step": 6443
+    },
+    {
+      "epoch": 1.6305668016194332,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7898110661268556e-05,
+      "loss": 2.44,
+      "step": 6444
+    },
+    {
+      "epoch": 1.6308198380566803,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.7897267206477733e-05,
+      "loss": 2.3677,
+      "step": 6445
+    },
+    {
+      "epoch": 1.6310728744939271,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.789642375168691e-05,
+      "loss": 2.5487,
+      "step": 6446
+    },
+    {
+      "epoch": 1.631325910931174,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.789558029689609e-05,
+      "loss": 2.543,
+      "step": 6447
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.7894736842105263e-05,
+      "loss": 2.4945,
+      "step": 6448
+    },
+    {
+      "epoch": 1.631831983805668,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.789389338731444e-05,
+      "loss": 2.5239,
+      "step": 6449
+    },
+    {
+      "epoch": 1.632085020242915,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.789304993252362e-05,
+      "loss": 2.3759,
+      "step": 6450
+    },
+    {
+      "epoch": 1.632085020242915,
+      "eval_loss": 2.563199043273926,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 295.3523,
+      "eval_samples_per_second": 8.803,
+      "eval_steps_per_second": 8.803,
+      "step": 6450
+    },
+    {
+      "epoch": 1.632338056680162,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.7892206477732796e-05,
+      "loss": 2.3976,
+      "step": 6451
+    },
+    {
+      "epoch": 1.632591093117409,
+      "grad_norm": 3.0,
+      "learning_rate": 2.789136302294197e-05,
+      "loss": 2.4333,
+      "step": 6452
+    },
+    {
+      "epoch": 1.6328441295546559,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.789051956815115e-05,
+      "loss": 2.4044,
+      "step": 6453
+    },
+    {
+      "epoch": 1.6330971659919027,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7889676113360326e-05,
+      "loss": 2.4263,
+      "step": 6454
+    },
+    {
+      "epoch": 1.6333502024291497,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.78888326585695e-05,
+      "loss": 2.419,
+      "step": 6455
+    },
+    {
+      "epoch": 1.6336032388663968,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7887989203778676e-05,
+      "loss": 2.4339,
+      "step": 6456
+    },
+    {
+      "epoch": 1.6338562753036436,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7887145748987856e-05,
+      "loss": 2.4614,
+      "step": 6457
+    },
+    {
+      "epoch": 1.6341093117408907,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.7886302294197033e-05,
+      "loss": 2.396,
+      "step": 6458
+    },
+    {
+      "epoch": 1.6343623481781377,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.7885458839406206e-05,
+      "loss": 2.5559,
+      "step": 6459
+    },
+    {
+      "epoch": 1.6346153846153846,
+      "grad_norm": 19.25,
+      "learning_rate": 2.7884615384615386e-05,
+      "loss": 2.6979,
+      "step": 6460
+    },
+    {
+      "epoch": 1.6348684210526314,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.7883771929824563e-05,
+      "loss": 2.4883,
+      "step": 6461
+    },
+    {
+      "epoch": 1.6351214574898787,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7882928475033736e-05,
+      "loss": 2.4103,
+      "step": 6462
+    },
+    {
+      "epoch": 1.6353744939271255,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.7882085020242916e-05,
+      "loss": 2.3775,
+      "step": 6463
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.7881241565452093e-05,
+      "loss": 2.4463,
+      "step": 6464
+    },
+    {
+      "epoch": 1.6358805668016194,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.788039811066127e-05,
+      "loss": 2.2132,
+      "step": 6465
+    },
+    {
+      "epoch": 1.6361336032388665,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.7879554655870446e-05,
+      "loss": 2.4432,
+      "step": 6466
+    },
+    {
+      "epoch": 1.6363866396761133,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.7878711201079623e-05,
+      "loss": 2.5506,
+      "step": 6467
+    },
+    {
+      "epoch": 1.6366396761133604,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.78778677462888e-05,
+      "loss": 2.4464,
+      "step": 6468
+    },
+    {
+      "epoch": 1.6368927125506074,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.787702429149798e-05,
+      "loss": 2.4078,
+      "step": 6469
+    },
+    {
+      "epoch": 1.6371457489878543,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.7876180836707153e-05,
+      "loss": 2.364,
+      "step": 6470
+    },
+    {
+      "epoch": 1.637398785425101,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.787533738191633e-05,
+      "loss": 2.5078,
+      "step": 6471
+    },
+    {
+      "epoch": 1.6376518218623481,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.787449392712551e-05,
+      "loss": 2.4798,
+      "step": 6472
+    },
+    {
+      "epoch": 1.6379048582995952,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.7873650472334683e-05,
+      "loss": 2.3765,
+      "step": 6473
+    },
+    {
+      "epoch": 1.638157894736842,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.787280701754386e-05,
+      "loss": 2.4889,
+      "step": 6474
+    },
+    {
+      "epoch": 1.638410931174089,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7871963562753036e-05,
+      "loss": 2.5256,
+      "step": 6475
+    },
+    {
+      "epoch": 1.6386639676113361,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.7871120107962216e-05,
+      "loss": 2.3179,
+      "step": 6476
+    },
+    {
+      "epoch": 1.638917004048583,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.787027665317139e-05,
+      "loss": 2.4592,
+      "step": 6477
+    },
+    {
+      "epoch": 1.6391700404858298,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.7869433198380566e-05,
+      "loss": 2.332,
+      "step": 6478
+    },
+    {
+      "epoch": 1.6394230769230769,
+      "grad_norm": 3.0,
+      "learning_rate": 2.7868589743589746e-05,
+      "loss": 2.5009,
+      "step": 6479
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7867746288798923e-05,
+      "loss": 2.4756,
+      "step": 6480
+    },
+    {
+      "epoch": 1.639676113360324,
+      "eval_loss": 2.558899164199829,
+      "eval_model_preparation_time": 0.3365,
+      "eval_runtime": 284.8114,
+      "eval_samples_per_second": 9.129,
+      "eval_steps_per_second": 9.129,
+      "step": 6480
+    },
+    {
+      "epoch": 1.6399291497975708,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.7866902834008096e-05,
+      "loss": 2.3809,
+      "step": 6481
+    },
+    {
+      "epoch": 1.6401821862348178,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7866059379217276e-05,
+      "loss": 2.4314,
+      "step": 6482
+    },
+    {
+      "epoch": 1.6404352226720649,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.7865215924426452e-05,
+      "loss": 2.3195,
+      "step": 6483
+    },
+    {
+      "epoch": 1.6406882591093117,
+      "grad_norm": 2.875,
+      "learning_rate": 2.7864372469635626e-05,
+      "loss": 2.4304,
+      "step": 6484
+    },
+    {
+      "epoch": 1.6409412955465585,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.7863529014844806e-05,
+      "loss": 2.4807,
+      "step": 6485
+    },
+    {
+      "epoch": 1.6411943319838058,
+      "grad_norm": 3.0,
+      "learning_rate": 2.7862685560053982e-05,
+      "loss": 2.4925,
+      "step": 6486
+    },
+    {
+      "epoch": 1.6414473684210527,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.786184210526316e-05,
+      "loss": 2.3494,
+      "step": 6487
+    },
+    {
+      "epoch": 1.6417004048582995,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.7860998650472336e-05,
+      "loss": 2.515,
+      "step": 6488
+    },
+    {
+      "epoch": 1.6419534412955465,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.7860155195681512e-05,
+      "loss": 2.4318,
+      "step": 6489
+    },
+    {
+      "epoch": 1.6422064777327936,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.785931174089069e-05,
+      "loss": 2.5142,
+      "step": 6490
+    },
+    {
+      "epoch": 1.6424595141700404,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.7858468286099866e-05,
+      "loss": 2.4678,
+      "step": 6491
+    },
+    {
+      "epoch": 1.6427125506072875,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.7857624831309042e-05,
+      "loss": 2.2306,
+      "step": 6492
+    },
+    {
+      "epoch": 1.6429655870445345,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.785678137651822e-05,
+      "loss": 2.4297,
+      "step": 6493
+    },
+    {
+      "epoch": 1.6432186234817814,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.7855937921727396e-05,
+      "loss": 2.5432,
+      "step": 6494
+    },
+    {
+      "epoch": 1.6434716599190282,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.7855094466936572e-05,
+      "loss": 2.3669,
+      "step": 6495
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.785425101214575e-05,
+      "loss": 2.3253,
+      "step": 6496
+    },
+    {
+      "epoch": 1.6439777327935223,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.7853407557354926e-05,
+      "loss": 2.4658,
+      "step": 6497
+    },
+    {
+      "epoch": 1.6442307692307692,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7852564102564106e-05,
+      "loss": 2.4611,
+      "step": 6498
+    },
+    {
+      "epoch": 1.6444838056680162,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.785172064777328e-05,
+      "loss": 2.42,
+      "step": 6499
+    },
+    {
+      "epoch": 1.6447368421052633,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7850877192982456e-05,
+      "loss": 2.4055,
+      "step": 6500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 39520,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.6186190334794547e+18,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}