diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10387 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 184,
+  "global_step": 1470,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0013605442176870747,
+      "grad_norm": 0.24701461672354935,
+      "learning_rate": 1.36986301369863e-07,
+      "loss": 1.6736,
+      "step": 1
+    },
+    {
+      "epoch": 0.0013605442176870747,
+      "eval_loss": 1.7904456853866577,
+      "eval_runtime": 75.582,
+      "eval_samples_per_second": 53.888,
+      "eval_steps_per_second": 6.748,
+      "step": 1
+    },
+    {
+      "epoch": 0.0027210884353741495,
+      "grad_norm": 0.21437113339785932,
+      "learning_rate": 2.73972602739726e-07,
+      "loss": 1.5884,
+      "step": 2
+    },
+    {
+      "epoch": 0.004081632653061225,
+      "grad_norm": 0.3228668200940542,
+      "learning_rate": 4.1095890410958903e-07,
+      "loss": 1.6821,
+      "step": 3
+    },
+    {
+      "epoch": 0.005442176870748299,
+      "grad_norm": 0.19408831616689562,
+      "learning_rate": 5.47945205479452e-07,
+      "loss": 1.8146,
+      "step": 4
+    },
+    {
+      "epoch": 0.006802721088435374,
+      "grad_norm": 0.18446566294319683,
+      "learning_rate": 6.849315068493151e-07,
+      "loss": 1.6316,
+      "step": 5
+    },
+    {
+      "epoch": 0.00816326530612245,
+      "grad_norm": 0.26237580245842185,
+      "learning_rate": 8.219178082191781e-07,
+      "loss": 1.7544,
+      "step": 6
+    },
+    {
+      "epoch": 0.009523809523809525,
+      "grad_norm": 0.1659195721310037,
+      "learning_rate": 9.589041095890411e-07,
+      "loss": 1.8325,
+      "step": 7
+    },
+    {
+      "epoch": 0.010884353741496598,
+      "grad_norm": 0.14112003912821341,
+      "learning_rate": 1.095890410958904e-06,
+      "loss": 1.8533,
+      "step": 8
+    },
+    {
+      "epoch": 0.012244897959183673,
+      "grad_norm": 0.22295406766041573,
+      "learning_rate": 1.2328767123287673e-06,
+      "loss": 1.7309,
+      "step": 9
+    },
+    {
+      "epoch": 0.013605442176870748,
+      "grad_norm": 0.20855919407710727,
+      "learning_rate": 1.3698630136986302e-06,
+      "loss": 1.4983,
+      "step": 10
+    },
+    {
+      "epoch": 0.014965986394557823,
+      "grad_norm": 0.39634451341504184,
+      "learning_rate": 1.5068493150684932e-06,
+      "loss": 1.71,
+      "step": 11
+    },
+    {
+      "epoch": 0.0163265306122449,
+      "grad_norm": 0.2918296142957545,
+      "learning_rate": 1.6438356164383561e-06,
+      "loss": 1.6983,
+      "step": 12
+    },
+    {
+      "epoch": 0.017687074829931974,
+      "grad_norm": 0.3333249210865954,
+      "learning_rate": 1.7808219178082193e-06,
+      "loss": 1.6435,
+      "step": 13
+    },
+    {
+      "epoch": 0.01904761904761905,
+      "grad_norm": 0.3288930419026758,
+      "learning_rate": 1.9178082191780823e-06,
+      "loss": 1.9445,
+      "step": 14
+    },
+    {
+      "epoch": 0.02040816326530612,
+      "grad_norm": 0.3311742875918285,
+      "learning_rate": 2.0547945205479454e-06,
+      "loss": 1.8007,
+      "step": 15
+    },
+    {
+      "epoch": 0.021768707482993196,
+      "grad_norm": 0.24222843258421317,
+      "learning_rate": 2.191780821917808e-06,
+      "loss": 1.8698,
+      "step": 16
+    },
+    {
+      "epoch": 0.02312925170068027,
+      "grad_norm": 0.2863215351075517,
+      "learning_rate": 2.3287671232876713e-06,
+      "loss": 1.8295,
+      "step": 17
+    },
+    {
+      "epoch": 0.024489795918367346,
+      "grad_norm": 0.37000991286313667,
+      "learning_rate": 2.4657534246575345e-06,
+      "loss": 1.7748,
+      "step": 18
+    },
+    {
+      "epoch": 0.02585034013605442,
+      "grad_norm": 0.305345665951125,
+      "learning_rate": 2.6027397260273973e-06,
+      "loss": 1.7799,
+      "step": 19
+    },
+    {
+      "epoch": 0.027210884353741496,
+      "grad_norm": 0.276577651886119,
+      "learning_rate": 2.7397260273972604e-06,
+      "loss": 1.5892,
+      "step": 20
+    },
+    {
+      "epoch": 0.02857142857142857,
+      "grad_norm": 0.40747672861545675,
+      "learning_rate": 2.876712328767123e-06,
+      "loss": 1.797,
+      "step": 21
+    },
+    {
+      "epoch": 0.029931972789115645,
+      "grad_norm": 0.1663214297242309,
+      "learning_rate": 3.0136986301369864e-06,
+      "loss": 1.8254,
+      "step": 22
+    },
+    {
+      "epoch": 0.031292517006802724,
+      "grad_norm": 0.34875514252556655,
+      "learning_rate": 3.1506849315068495e-06,
+      "loss": 1.5967,
+      "step": 23
+    },
+    {
+      "epoch": 0.0326530612244898,
+      "grad_norm": 0.31488445168418,
+      "learning_rate": 3.2876712328767123e-06,
+      "loss": 1.8033,
+      "step": 24
+    },
+    {
+      "epoch": 0.034013605442176874,
+      "grad_norm": 0.9585107293220959,
+      "learning_rate": 3.4246575342465754e-06,
+      "loss": 1.9985,
+      "step": 25
+    },
+    {
+      "epoch": 0.03537414965986395,
+      "grad_norm": 0.4719659909416967,
+      "learning_rate": 3.5616438356164386e-06,
+      "loss": 1.6673,
+      "step": 26
+    },
+    {
+      "epoch": 0.036734693877551024,
+      "grad_norm": 0.5206398105101208,
+      "learning_rate": 3.6986301369863014e-06,
+      "loss": 1.7832,
+      "step": 27
+    },
+    {
+      "epoch": 0.0380952380952381,
+      "grad_norm": 0.5525391513084628,
+      "learning_rate": 3.8356164383561645e-06,
+      "loss": 1.8033,
+      "step": 28
+    },
+    {
+      "epoch": 0.03945578231292517,
+      "grad_norm": 0.5864368554335787,
+      "learning_rate": 3.972602739726027e-06,
+      "loss": 1.637,
+      "step": 29
+    },
+    {
+      "epoch": 0.04081632653061224,
+      "grad_norm": 0.18211390682659326,
+      "learning_rate": 4.109589041095891e-06,
+      "loss": 1.6996,
+      "step": 30
+    },
+    {
+      "epoch": 0.04217687074829932,
+      "grad_norm": 0.26324481615027445,
+      "learning_rate": 4.246575342465754e-06,
+      "loss": 1.7077,
+      "step": 31
+    },
+    {
+      "epoch": 0.04353741496598639,
+      "grad_norm": 0.487665052197852,
+      "learning_rate": 4.383561643835616e-06,
+      "loss": 1.5757,
+      "step": 32
+    },
+    {
+      "epoch": 0.044897959183673466,
+      "grad_norm": 0.5110956602957011,
+      "learning_rate": 4.52054794520548e-06,
+      "loss": 1.6525,
+      "step": 33
+    },
+    {
+      "epoch": 0.04625850340136054,
+      "grad_norm": 0.41488349790070234,
+      "learning_rate": 4.657534246575343e-06,
+      "loss": 1.7469,
+      "step": 34
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 0.3205549447320179,
+      "learning_rate": 4.7945205479452054e-06,
+      "loss": 1.6621,
+      "step": 35
+    },
+    {
+      "epoch": 0.04897959183673469,
+      "grad_norm": 0.2759804237950767,
+      "learning_rate": 4.931506849315069e-06,
+      "loss": 1.8364,
+      "step": 36
+    },
+    {
+      "epoch": 0.050340136054421766,
+      "grad_norm": 0.4070079284746193,
+      "learning_rate": 5.068493150684932e-06,
+      "loss": 1.7928,
+      "step": 37
+    },
+    {
+      "epoch": 0.05170068027210884,
+      "grad_norm": 0.3162452736080499,
+      "learning_rate": 5.2054794520547945e-06,
+      "loss": 1.8174,
+      "step": 38
+    },
+    {
+      "epoch": 0.053061224489795916,
+      "grad_norm": 0.339190852848117,
+      "learning_rate": 5.342465753424658e-06,
+      "loss": 1.8372,
+      "step": 39
+    },
+    {
+      "epoch": 0.05442176870748299,
+      "grad_norm": 0.31599130496764827,
+      "learning_rate": 5.479452054794521e-06,
+      "loss": 1.7265,
+      "step": 40
+    },
+    {
+      "epoch": 0.055782312925170066,
+      "grad_norm": 0.18290357316608127,
+      "learning_rate": 5.6164383561643845e-06,
+      "loss": 1.7055,
+      "step": 41
+    },
+    {
+      "epoch": 0.05714285714285714,
+      "grad_norm": 0.26450493295787797,
+      "learning_rate": 5.753424657534246e-06,
+      "loss": 1.859,
+      "step": 42
+    },
+    {
+      "epoch": 0.058503401360544216,
+      "grad_norm": 0.2105468885683211,
+      "learning_rate": 5.89041095890411e-06,
+      "loss": 1.7903,
+      "step": 43
+    },
+    {
+      "epoch": 0.05986394557823129,
+      "grad_norm": 0.21904274744659627,
+      "learning_rate": 6.027397260273973e-06,
+      "loss": 1.7112,
+      "step": 44
+    },
+    {
+      "epoch": 0.061224489795918366,
+      "grad_norm": 0.2766631664495227,
+      "learning_rate": 6.164383561643836e-06,
+      "loss": 1.626,
+      "step": 45
+    },
+    {
+      "epoch": 0.06258503401360545,
+      "grad_norm": 0.27137304801321466,
+      "learning_rate": 6.301369863013699e-06,
+      "loss": 1.8546,
+      "step": 46
+    },
+    {
+      "epoch": 0.06394557823129252,
+      "grad_norm": 0.17562873404669305,
+      "learning_rate": 6.438356164383563e-06,
+      "loss": 1.8687,
+      "step": 47
+    },
+    {
+      "epoch": 0.0653061224489796,
+      "grad_norm": 0.23608638226381062,
+      "learning_rate": 6.5753424657534245e-06,
+      "loss": 1.5768,
+      "step": 48
+    },
+    {
+      "epoch": 0.06666666666666667,
+      "grad_norm": 0.12395160133391969,
+      "learning_rate": 6.712328767123288e-06,
+      "loss": 1.8217,
+      "step": 49
+    },
+    {
+      "epoch": 0.06802721088435375,
+      "grad_norm": 0.21069127406909471,
+      "learning_rate": 6.849315068493151e-06,
+      "loss": 1.7057,
+      "step": 50
+    },
+    {
+      "epoch": 0.06938775510204082,
+      "grad_norm": 0.17153884217244356,
+      "learning_rate": 6.9863013698630145e-06,
+      "loss": 1.9143,
+      "step": 51
+    },
+    {
+      "epoch": 0.0707482993197279,
+      "grad_norm": 0.3084343242877715,
+      "learning_rate": 7.123287671232877e-06,
+      "loss": 1.8398,
+      "step": 52
+    },
+    {
+      "epoch": 0.07210884353741497,
+      "grad_norm": 0.14644662918576262,
+      "learning_rate": 7.260273972602741e-06,
+      "loss": 1.6646,
+      "step": 53
+    },
+    {
+      "epoch": 0.07346938775510205,
+      "grad_norm": 0.3001793602079481,
+      "learning_rate": 7.397260273972603e-06,
+      "loss": 1.689,
+      "step": 54
+    },
+    {
+      "epoch": 0.07482993197278912,
+      "grad_norm": 0.301851334470962,
+      "learning_rate": 7.534246575342466e-06,
+      "loss": 1.5179,
+      "step": 55
+    },
+    {
+      "epoch": 0.0761904761904762,
+      "grad_norm": 0.33200247196496224,
+      "learning_rate": 7.671232876712329e-06,
+      "loss": 1.8986,
+      "step": 56
+    },
+    {
+      "epoch": 0.07755102040816327,
+      "grad_norm": 0.18181195505623798,
+      "learning_rate": 7.808219178082192e-06,
+      "loss": 1.6426,
+      "step": 57
+    },
+    {
+      "epoch": 0.07891156462585033,
+      "grad_norm": 0.12250708549849011,
+      "learning_rate": 7.945205479452055e-06,
+      "loss": 1.6214,
+      "step": 58
+    },
+    {
+      "epoch": 0.08027210884353742,
+      "grad_norm": 0.09796847494385076,
+      "learning_rate": 8.082191780821919e-06,
+      "loss": 1.6547,
+      "step": 59
+    },
+    {
+      "epoch": 0.08163265306122448,
+      "grad_norm": 0.12998919923759888,
+      "learning_rate": 8.219178082191782e-06,
+      "loss": 1.7818,
+      "step": 60
+    },
+    {
+      "epoch": 0.08299319727891157,
+      "grad_norm": 0.2260386111575877,
+      "learning_rate": 8.356164383561644e-06,
+      "loss": 1.7807,
+      "step": 61
+    },
+    {
+      "epoch": 0.08435374149659863,
+      "grad_norm": 0.33754760373428094,
+      "learning_rate": 8.493150684931507e-06,
+      "loss": 1.617,
+      "step": 62
+    },
+    {
+      "epoch": 0.08571428571428572,
+      "grad_norm": 0.35962963555168737,
+      "learning_rate": 8.63013698630137e-06,
+      "loss": 1.6799,
+      "step": 63
+    },
+    {
+      "epoch": 0.08707482993197278,
+      "grad_norm": 0.32506967541048193,
+      "learning_rate": 8.767123287671233e-06,
+      "loss": 1.6454,
+      "step": 64
+    },
+    {
+      "epoch": 0.08843537414965986,
+      "grad_norm": 0.21523079823600388,
+      "learning_rate": 8.904109589041097e-06,
+      "loss": 1.8856,
+      "step": 65
+    },
+    {
+      "epoch": 0.08979591836734693,
+      "grad_norm": 0.5363358811064897,
+      "learning_rate": 9.04109589041096e-06,
+      "loss": 1.6952,
+      "step": 66
+    },
+    {
+      "epoch": 0.09115646258503401,
+      "grad_norm": 0.14306066721600327,
+      "learning_rate": 9.178082191780823e-06,
+      "loss": 1.8208,
+      "step": 67
+    },
+    {
+      "epoch": 0.09251700680272108,
+      "grad_norm": 0.18646957264381078,
+      "learning_rate": 9.315068493150685e-06,
+      "loss": 1.7517,
+      "step": 68
+    },
+    {
+      "epoch": 0.09387755102040816,
+      "grad_norm": 0.19137982075531637,
+      "learning_rate": 9.452054794520548e-06,
+      "loss": 1.6456,
+      "step": 69
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 0.15987203027468555,
+      "learning_rate": 9.589041095890411e-06,
+      "loss": 1.7148,
+      "step": 70
+    },
+    {
+      "epoch": 0.09659863945578231,
+      "grad_norm": 0.16311504243422864,
+      "learning_rate": 9.726027397260275e-06,
+      "loss": 1.6627,
+      "step": 71
+    },
+    {
+      "epoch": 0.09795918367346938,
+      "grad_norm": 0.10186314299964105,
+      "learning_rate": 9.863013698630138e-06,
+      "loss": 1.5856,
+      "step": 72
+    },
+    {
+      "epoch": 0.09931972789115646,
+      "grad_norm": 0.13469761876363148,
+      "learning_rate": 1e-05,
+      "loss": 1.6557,
+      "step": 73
+    },
+    {
+      "epoch": 0.10068027210884353,
+      "grad_norm": 0.11568418682806415,
+      "learning_rate": 9.999987357098372e-06,
+      "loss": 1.7807,
+      "step": 74
+    },
+    {
+      "epoch": 0.10204081632653061,
+      "grad_norm": 0.11288388506482096,
+      "learning_rate": 9.999949428457423e-06,
+      "loss": 1.8232,
+      "step": 75
+    },
+    {
+      "epoch": 0.10340136054421768,
+      "grad_norm": 0.16329859637421754,
+      "learning_rate": 9.999886214268967e-06,
+      "loss": 1.7462,
+      "step": 76
+    },
+    {
+      "epoch": 0.10476190476190476,
+      "grad_norm": 0.20231664635671653,
+      "learning_rate": 9.999797714852686e-06,
+      "loss": 1.5938,
+      "step": 77
+    },
+    {
+      "epoch": 0.10612244897959183,
+      "grad_norm": 0.34538065180937266,
+      "learning_rate": 9.999683930656135e-06,
+      "loss": 1.8806,
+      "step": 78
+    },
+    {
+      "epoch": 0.10748299319727891,
+      "grad_norm": 0.13354157904043504,
+      "learning_rate": 9.999544862254743e-06,
+      "loss": 1.801,
+      "step": 79
+    },
+    {
+      "epoch": 0.10884353741496598,
+      "grad_norm": 0.13220305876865404,
+      "learning_rate": 9.999380510351796e-06,
+      "loss": 1.6805,
+      "step": 80
+    },
+    {
+      "epoch": 0.11020408163265306,
+      "grad_norm": 0.13768110879863274,
+      "learning_rate": 9.999190875778452e-06,
+      "loss": 1.7481,
+      "step": 81
+    },
+    {
+      "epoch": 0.11156462585034013,
+      "grad_norm": 0.11222690770456831,
+      "learning_rate": 9.998975959493722e-06,
+      "loss": 1.7894,
+      "step": 82
+    },
+    {
+      "epoch": 0.11292517006802721,
+      "grad_norm": 0.11775170157819592,
+      "learning_rate": 9.998735762584471e-06,
+      "loss": 1.8592,
+      "step": 83
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 0.20855277686570553,
+      "learning_rate": 9.998470286265415e-06,
+      "loss": 1.7145,
+      "step": 84
+    },
+    {
+      "epoch": 0.11564625850340136,
+      "grad_norm": 0.10682809945125131,
+      "learning_rate": 9.998179531879112e-06,
+      "loss": 1.7563,
+      "step": 85
+    },
+    {
+      "epoch": 0.11700680272108843,
+      "grad_norm": 0.1332681057101403,
+      "learning_rate": 9.99786350089595e-06,
+      "loss": 1.6698,
+      "step": 86
+    },
+    {
+      "epoch": 0.11836734693877551,
+      "grad_norm": 0.1442352006249483,
+      "learning_rate": 9.99752219491415e-06,
+      "loss": 1.542,
+      "step": 87
+    },
+    {
+      "epoch": 0.11972789115646258,
+      "grad_norm": 0.09723976872539679,
+      "learning_rate": 9.997155615659753e-06,
+      "loss": 1.5545,
+      "step": 88
+    },
+    {
+      "epoch": 0.12108843537414966,
+      "grad_norm": 0.15078850009122496,
+      "learning_rate": 9.996763764986606e-06,
+      "loss": 1.6872,
+      "step": 89
+    },
+    {
+      "epoch": 0.12244897959183673,
+      "grad_norm": 0.09880013032692718,
+      "learning_rate": 9.996346644876363e-06,
+      "loss": 1.5761,
+      "step": 90
+    },
+    {
+      "epoch": 0.12380952380952381,
+      "grad_norm": 0.1797981570168221,
+      "learning_rate": 9.995904257438467e-06,
+      "loss": 1.5885,
+      "step": 91
+    },
+    {
+      "epoch": 0.1251700680272109,
+      "grad_norm": 0.14066405347976094,
+      "learning_rate": 9.995436604910142e-06,
+      "loss": 1.7558,
+      "step": 92
+    },
+    {
+      "epoch": 0.12653061224489795,
+      "grad_norm": 0.2804984380485241,
+      "learning_rate": 9.994943689656381e-06,
+      "loss": 1.5653,
+      "step": 93
+    },
+    {
+      "epoch": 0.12789115646258503,
+      "grad_norm": 0.09802426112688165,
+      "learning_rate": 9.994425514169938e-06,
+      "loss": 1.8666,
+      "step": 94
+    },
+    {
+      "epoch": 0.1292517006802721,
+      "grad_norm": 0.2640163991220947,
+      "learning_rate": 9.993882081071307e-06,
+      "loss": 1.8331,
+      "step": 95
+    },
+    {
+      "epoch": 0.1306122448979592,
+      "grad_norm": 0.12584718580988416,
+      "learning_rate": 9.99331339310872e-06,
+      "loss": 1.7264,
+      "step": 96
+    },
+    {
+      "epoch": 0.13197278911564625,
+      "grad_norm": 0.11723300893007116,
+      "learning_rate": 9.99271945315812e-06,
+      "loss": 1.774,
+      "step": 97
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 0.11104245778454394,
+      "learning_rate": 9.992100264223156e-06,
+      "loss": 1.7154,
+      "step": 98
+    },
+    {
+      "epoch": 0.1346938775510204,
+      "grad_norm": 0.0915644970371204,
+      "learning_rate": 9.99145582943517e-06,
+      "loss": 1.6768,
+      "step": 99
+    },
+    {
+      "epoch": 0.1360544217687075,
+      "grad_norm": 0.11971918094721708,
+      "learning_rate": 9.990786152053169e-06,
+      "loss": 1.895,
+      "step": 100
+    },
+    {
+      "epoch": 0.13741496598639455,
+      "grad_norm": 0.13849974347702929,
+      "learning_rate": 9.99009123546382e-06,
+      "loss": 1.9232,
+      "step": 101
+    },
+    {
+      "epoch": 0.13877551020408163,
+      "grad_norm": 0.0832290902024341,
+      "learning_rate": 9.98937108318143e-06,
+      "loss": 1.419,
+      "step": 102
+    },
+    {
+      "epoch": 0.1401360544217687,
+      "grad_norm": 0.09490309244168035,
+      "learning_rate": 9.988625698847921e-06,
+      "loss": 1.6096,
+      "step": 103
+    },
+    {
+      "epoch": 0.1414965986394558,
+      "grad_norm": 0.08634281151584555,
+      "learning_rate": 9.987855086232824e-06,
+      "loss": 1.6766,
+      "step": 104
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.12657846070776754,
+      "learning_rate": 9.98705924923325e-06,
+      "loss": 1.7755,
+      "step": 105
+    },
+    {
+      "epoch": 0.14421768707482993,
+      "grad_norm": 0.1730231080244019,
+      "learning_rate": 9.986238191873874e-06,
+      "loss": 1.671,
+      "step": 106
+    },
+    {
+      "epoch": 0.145578231292517,
+      "grad_norm": 0.11653855558191023,
+      "learning_rate": 9.985391918306915e-06,
+      "loss": 1.6012,
+      "step": 107
+    },
+    {
+      "epoch": 0.1469387755102041,
+      "grad_norm": 0.09868922955378823,
+      "learning_rate": 9.984520432812117e-06,
+      "loss": 1.8218,
+      "step": 108
+    },
+    {
+      "epoch": 0.14829931972789115,
+      "grad_norm": 0.08718149041105193,
+      "learning_rate": 9.983623739796718e-06,
+      "loss": 1.6361,
+      "step": 109
+    },
+    {
+      "epoch": 0.14965986394557823,
+      "grad_norm": 0.08536190731319725,
+      "learning_rate": 9.982701843795441e-06,
+      "loss": 1.8356,
+      "step": 110
+    },
+    {
+      "epoch": 0.1510204081632653,
+      "grad_norm": 0.1778419657439268,
+      "learning_rate": 9.981754749470463e-06,
+      "loss": 1.6968,
+      "step": 111
+    },
+    {
+      "epoch": 0.1523809523809524,
+      "grad_norm": 0.12982223254146993,
+      "learning_rate": 9.980782461611391e-06,
+      "loss": 1.8005,
+      "step": 112
+    },
+    {
+      "epoch": 0.15374149659863945,
+      "grad_norm": 0.08982117932691205,
+      "learning_rate": 9.979784985135239e-06,
+      "loss": 1.7645,
+      "step": 113
+    },
+    {
+      "epoch": 0.15510204081632653,
+      "grad_norm": 0.12460716696891104,
+      "learning_rate": 9.978762325086408e-06,
+      "loss": 1.6455,
+      "step": 114
+    },
+    {
+      "epoch": 0.1564625850340136,
+      "grad_norm": 0.09362932823935477,
+      "learning_rate": 9.977714486636657e-06,
+      "loss": 1.8083,
+      "step": 115
+    },
+    {
+      "epoch": 0.15782312925170067,
+      "grad_norm": 0.09099536634076917,
+      "learning_rate": 9.976641475085067e-06,
+      "loss": 1.7776,
+      "step": 116
+    },
+    {
+      "epoch": 0.15918367346938775,
+      "grad_norm": 0.08568595730791906,
+      "learning_rate": 9.975543295858035e-06,
+      "loss": 1.8846,
+      "step": 117
+    },
+    {
+      "epoch": 0.16054421768707483,
+      "grad_norm": 0.1310404323604523,
+      "learning_rate": 9.974419954509225e-06,
+      "loss": 1.5725,
+      "step": 118
+    },
+    {
+      "epoch": 0.1619047619047619,
+      "grad_norm": 0.11863021260862251,
+      "learning_rate": 9.97327145671956e-06,
+      "loss": 1.6409,
+      "step": 119
+    },
+    {
+      "epoch": 0.16326530612244897,
+      "grad_norm": 0.11864941995819639,
+      "learning_rate": 9.972097808297174e-06,
+      "loss": 1.7081,
+      "step": 120
+    },
+    {
+      "epoch": 0.16462585034013605,
+      "grad_norm": 0.08013610894171046,
+      "learning_rate": 9.970899015177398e-06,
+      "loss": 1.7804,
+      "step": 121
+    },
+    {
+      "epoch": 0.16598639455782313,
+      "grad_norm": 0.12399055095582327,
+      "learning_rate": 9.969675083422719e-06,
+      "loss": 1.6848,
+      "step": 122
+    },
+    {
+      "epoch": 0.1673469387755102,
+      "grad_norm": 0.1433779964353759,
+      "learning_rate": 9.96842601922276e-06,
+      "loss": 1.6888,
+      "step": 123
+    },
+    {
+      "epoch": 0.16870748299319727,
+      "grad_norm": 0.09915990901687576,
+      "learning_rate": 9.967151828894234e-06,
+      "loss": 1.7802,
+      "step": 124
+    },
+    {
+      "epoch": 0.17006802721088435,
+      "grad_norm": 0.10206449162778881,
+      "learning_rate": 9.965852518880931e-06,
+      "loss": 1.806,
+      "step": 125
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 0.09282143748721522,
+      "learning_rate": 9.964528095753669e-06,
+      "loss": 1.5987,
+      "step": 126
+    },
+    {
+      "epoch": 0.1727891156462585,
+      "grad_norm": 0.16907020113729054,
+      "learning_rate": 9.963178566210268e-06,
+      "loss": 1.7569,
+      "step": 127
+    },
+    {
+      "epoch": 0.17414965986394557,
+      "grad_norm": 0.08207848199751772,
+      "learning_rate": 9.961803937075516e-06,
+      "loss": 1.6724,
+      "step": 128
+    },
+    {
+      "epoch": 0.17551020408163265,
+      "grad_norm": 0.07319670048822571,
+      "learning_rate": 9.960404215301133e-06,
+      "loss": 1.7498,
+      "step": 129
+    },
+    {
+      "epoch": 0.17687074829931973,
+      "grad_norm": 0.08159880339274488,
+      "learning_rate": 9.958979407965738e-06,
+      "loss": 1.65,
+      "step": 130
+    },
+    {
+      "epoch": 0.1782312925170068,
+      "grad_norm": 0.09730054828361595,
+      "learning_rate": 9.95752952227481e-06,
+      "loss": 1.7796,
+      "step": 131
+    },
+    {
+      "epoch": 0.17959183673469387,
+      "grad_norm": 0.2629732363287427,
+      "learning_rate": 9.956054565560653e-06,
+      "loss": 1.6904,
+      "step": 132
+    },
+    {
+      "epoch": 0.18095238095238095,
+      "grad_norm": 0.07651749890098045,
+      "learning_rate": 9.954554545282363e-06,
+      "loss": 1.7809,
+      "step": 133
+    },
+    {
+      "epoch": 0.18231292517006803,
+      "grad_norm": 0.09628395581138101,
+      "learning_rate": 9.953029469025777e-06,
+      "loss": 1.8135,
+      "step": 134
+    },
+    {
+      "epoch": 0.1836734693877551,
+      "grad_norm": 0.09612376832963275,
+      "learning_rate": 9.951479344503459e-06,
+      "loss": 1.6617,
+      "step": 135
+    },
+    {
+      "epoch": 0.18503401360544217,
+      "grad_norm": 0.08107993061371403,
+      "learning_rate": 9.949904179554632e-06,
+      "loss": 1.6634,
+      "step": 136
+    },
+    {
+      "epoch": 0.18639455782312925,
+      "grad_norm": 0.07754512965459885,
+      "learning_rate": 9.94830398214516e-06,
+      "loss": 1.7732,
+      "step": 137
+    },
+    {
+      "epoch": 0.18775510204081633,
+      "grad_norm": 0.07265030754659244,
+      "learning_rate": 9.946678760367498e-06,
+      "loss": 1.7905,
+      "step": 138
+    },
+    {
+      "epoch": 0.1891156462585034,
+      "grad_norm": 0.09088517967487394,
+      "learning_rate": 9.945028522440654e-06,
+      "loss": 1.49,
+      "step": 139
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 0.21999926224724559,
+      "learning_rate": 9.943353276710146e-06,
+      "loss": 2.0726,
+      "step": 140
+    },
+    {
+      "epoch": 0.19183673469387755,
+      "grad_norm": 0.07397509235485085,
+      "learning_rate": 9.941653031647963e-06,
+      "loss": 1.6069,
+      "step": 141
+    },
+    {
+      "epoch": 0.19319727891156463,
+      "grad_norm": 0.17678430730401373,
+      "learning_rate": 9.939927795852513e-06,
+      "loss": 1.8128,
+      "step": 142
+    },
+    {
+      "epoch": 0.1945578231292517,
+      "grad_norm": 0.09311447920236875,
+      "learning_rate": 9.938177578048593e-06,
+      "loss": 1.682,
+      "step": 143
+    },
+    {
+      "epoch": 0.19591836734693877,
+      "grad_norm": 0.08923483853542422,
+      "learning_rate": 9.936402387087339e-06,
+      "loss": 1.7808,
+      "step": 144
+    },
+    {
+      "epoch": 0.19727891156462585,
+      "grad_norm": 0.3457260004062318,
+      "learning_rate": 9.93460223194617e-06,
+      "loss": 1.921,
+      "step": 145
+    },
+    {
+      "epoch": 0.19863945578231293,
+      "grad_norm": 0.11272086420065035,
+      "learning_rate": 9.932777121728765e-06,
+      "loss": 1.627,
+      "step": 146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0828154138118513,
+      "learning_rate": 9.930927065664997e-06,
+      "loss": 1.85,
+      "step": 147
+    },
+    {
+      "epoch": 0.20136054421768707,
+      "grad_norm": 0.10871781486388528,
+      "learning_rate": 9.929052073110897e-06,
+      "loss": 1.8526,
+      "step": 148
+    },
+    {
+      "epoch": 0.20272108843537415,
+      "grad_norm": 0.08372164326475892,
+      "learning_rate": 9.927152153548605e-06,
+      "loss": 1.6184,
+      "step": 149
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 0.2301749352348319,
+      "learning_rate": 9.925227316586316e-06,
+      "loss": 1.6416,
+      "step": 150
+    },
+    {
+      "epoch": 0.2054421768707483,
+      "grad_norm": 0.11534866479323268,
+      "learning_rate": 9.923277571958245e-06,
+      "loss": 1.6587,
+      "step": 151
+    },
+    {
+      "epoch": 0.20680272108843537,
+      "grad_norm": 0.1411655046905855,
+      "learning_rate": 9.921302929524561e-06,
+      "loss": 1.671,
+      "step": 152
+    },
+    {
+      "epoch": 0.20816326530612245,
+      "grad_norm": 0.07211757999248616,
+      "learning_rate": 9.919303399271348e-06,
+      "loss": 1.7163,
+      "step": 153
+    },
+    {
+      "epoch": 0.20952380952380953,
+      "grad_norm": 0.0873746156242924,
+      "learning_rate": 9.917278991310553e-06,
+      "loss": 1.6367,
+      "step": 154
+    },
+    {
+      "epoch": 0.2108843537414966,
+      "grad_norm": 0.0819591281688772,
+      "learning_rate": 9.915229715879928e-06,
+      "loss": 1.6989,
+      "step": 155
+    },
+    {
+      "epoch": 0.21224489795918366,
+      "grad_norm": 0.08552981032847369,
+      "learning_rate": 9.913155583342994e-06,
+      "loss": 1.5244,
+      "step": 156
+    },
+    {
+      "epoch": 0.21360544217687075,
+      "grad_norm": 0.13550974122069206,
+      "learning_rate": 9.91105660418897e-06,
+      "loss": 1.7495,
+      "step": 157
+    },
+    {
+      "epoch": 0.21496598639455783,
+      "grad_norm": 0.07091163304804983,
+      "learning_rate": 9.908932789032729e-06,
+      "loss": 1.7387,
+      "step": 158
+    },
+    {
+      "epoch": 0.2163265306122449,
+      "grad_norm": 0.0838103140533003,
+      "learning_rate": 9.906784148614745e-06,
+      "loss": 1.7076,
+      "step": 159
+    },
+    {
+      "epoch": 0.21768707482993196,
+      "grad_norm": 0.11349611508198672,
+      "learning_rate": 9.904610693801042e-06,
+      "loss": 1.6596,
+      "step": 160
+    },
+    {
+      "epoch": 0.21904761904761905,
+      "grad_norm": 0.07733122749252737,
+      "learning_rate": 9.902412435583127e-06,
+      "loss": 1.6503,
+      "step": 161
+    },
+    {
+      "epoch": 0.22040816326530613,
+      "grad_norm": 0.14625572340923682,
+      "learning_rate": 9.900189385077948e-06,
+      "loss": 1.564,
+      "step": 162
+    },
+    {
+      "epoch": 0.2217687074829932,
+      "grad_norm": 0.09712144690644532,
+      "learning_rate": 9.897941553527823e-06,
+      "loss": 1.7217,
+      "step": 163
+    },
+    {
+      "epoch": 0.22312925170068026,
+      "grad_norm": 0.0712274015908157,
+      "learning_rate": 9.895668952300403e-06,
+      "loss": 1.6412,
+      "step": 164
+    },
+    {
+      "epoch": 0.22448979591836735,
+      "grad_norm": 0.08811945291100708,
+      "learning_rate": 9.893371592888594e-06,
+      "loss": 1.6192,
+      "step": 165
+    },
+    {
+      "epoch": 0.22585034013605443,
+      "grad_norm": 0.07563751954927482,
+      "learning_rate": 9.891049486910513e-06,
+      "loss": 1.6283,
+      "step": 166
+    },
+    {
+      "epoch": 0.2272108843537415,
+      "grad_norm": 0.07473029887668768,
+      "learning_rate": 9.888702646109423e-06,
+      "loss": 1.6979,
+      "step": 167
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 0.07966661835478112,
+      "learning_rate": 9.886331082353673e-06,
+      "loss": 1.6951,
+      "step": 168
+    },
+    {
+      "epoch": 0.22993197278911565,
+      "grad_norm": 0.08625904148958655,
+      "learning_rate": 9.883934807636645e-06,
+      "loss": 1.6239,
+      "step": 169
+    },
+    {
+      "epoch": 0.23129251700680273,
+      "grad_norm": 0.06615144618602906,
+      "learning_rate": 9.881513834076683e-06,
+      "loss": 1.7456,
+      "step": 170
+    },
+    {
+      "epoch": 0.23265306122448978,
+      "grad_norm": 0.14491038831608893,
+      "learning_rate": 9.87906817391704e-06,
+      "loss": 1.7035,
+      "step": 171
+    },
+    {
+      "epoch": 0.23401360544217686,
+      "grad_norm": 0.0832300629302243,
+      "learning_rate": 9.876597839525814e-06,
+      "loss": 1.6672,
+      "step": 172
+    },
+    {
+      "epoch": 0.23537414965986395,
+      "grad_norm": 0.0917489076009908,
+      "learning_rate": 9.87410284339588e-06,
+      "loss": 1.6075,
+      "step": 173
+    },
+    {
+      "epoch": 0.23673469387755103,
+      "grad_norm": 0.06866742872418008,
+      "learning_rate": 9.871583198144836e-06,
+      "loss": 1.7646,
+      "step": 174
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 0.08465371920438951,
+      "learning_rate": 9.869038916514932e-06,
+      "loss": 1.6692,
+      "step": 175
+    },
+    {
+      "epoch": 0.23945578231292516,
+      "grad_norm": 0.09375415555940526,
+      "learning_rate": 9.866470011373009e-06,
+      "loss": 1.778,
+      "step": 176
+    },
+    {
+      "epoch": 0.24081632653061225,
+      "grad_norm": 0.07220432655814331,
+      "learning_rate": 9.863876495710433e-06,
+      "loss": 1.6857,
+      "step": 177
+    },
+    {
+      "epoch": 0.24217687074829933,
+      "grad_norm": 0.0797192184377915,
+      "learning_rate": 9.86125838264303e-06,
+      "loss": 1.7893,
+      "step": 178
+    },
+    {
+      "epoch": 0.24353741496598638,
+      "grad_norm": 0.07597718520214916,
+      "learning_rate": 9.858615685411018e-06,
+      "loss": 1.8848,
+      "step": 179
+    },
+    {
+      "epoch": 0.24489795918367346,
+      "grad_norm": 0.08003681020814803,
+      "learning_rate": 9.85594841737894e-06,
+      "loss": 1.8,
+      "step": 180
+    },
+    {
+      "epoch": 0.24625850340136055,
+      "grad_norm": 0.09696384289585193,
+      "learning_rate": 9.853256592035602e-06,
+      "loss": 1.7965,
+      "step": 181
+    },
+    {
+      "epoch": 0.24761904761904763,
+      "grad_norm": 0.12333580747104468,
+      "learning_rate": 9.850540222993994e-06,
+      "loss": 1.6365,
+      "step": 182
+    },
+    {
+      "epoch": 0.24897959183673468,
+      "grad_norm": 0.07310272321033273,
+      "learning_rate": 9.847799323991234e-06,
+      "loss": 1.5765,
+      "step": 183
+    },
+    {
+      "epoch": 0.2503401360544218,
+      "grad_norm": 0.12923131777808997,
+      "learning_rate": 9.845033908888485e-06,
+      "loss": 1.8017,
+      "step": 184
+    },
+    {
+      "epoch": 0.2503401360544218,
+      "eval_loss": 1.7241544723510742,
+      "eval_runtime": 76.6185,
+      "eval_samples_per_second": 53.159,
+      "eval_steps_per_second": 6.656,
+      "step": 184
+    },
+    {
+      "epoch": 0.25170068027210885,
+      "grad_norm": 0.06809715986288661,
+      "learning_rate": 9.842243991670899e-06,
+      "loss": 1.79,
+      "step": 185
+    },
+    {
+      "epoch": 0.2530612244897959,
+      "grad_norm": 0.08842474286261379,
+      "learning_rate": 9.839429586447534e-06,
+      "loss": 1.6168,
+      "step": 186
+    },
+    {
+      "epoch": 0.254421768707483,
+      "grad_norm": 0.10935159810036835,
+      "learning_rate": 9.836590707451287e-06,
+      "loss": 1.8505,
+      "step": 187
+    },
+    {
+      "epoch": 0.25578231292517006,
+      "grad_norm": 0.12270237138661655,
+      "learning_rate": 9.833727369038827e-06,
+      "loss": 1.635,
+      "step": 188
+    },
+    {
+      "epoch": 0.2571428571428571,
+      "grad_norm": 0.10719376109260877,
+      "learning_rate": 9.830839585690519e-06,
+      "loss": 1.8374,
+      "step": 189
+    },
+    {
+      "epoch": 0.2585034013605442,
+      "grad_norm": 0.07990417660478581,
+      "learning_rate": 9.827927372010343e-06,
+      "loss": 1.5681,
+      "step": 190
+    },
+    {
+      "epoch": 0.2598639455782313,
+      "grad_norm": 0.09668140137221073,
+      "learning_rate": 9.824990742725835e-06,
+      "loss": 1.6568,
+      "step": 191
+    },
+    {
+      "epoch": 0.2612244897959184,
+      "grad_norm": 0.07968515377548961,
+      "learning_rate": 9.822029712687999e-06,
+      "loss": 1.6007,
+      "step": 192
+    },
+    {
+      "epoch": 0.26258503401360545,
+      "grad_norm": 0.08837508173810749,
+      "learning_rate": 9.81904429687124e-06,
+      "loss": 1.6621,
+      "step": 193
+    },
+    {
+      "epoch": 0.2639455782312925,
+      "grad_norm": 0.08728025707185973,
+      "learning_rate": 9.816034510373287e-06,
+      "loss": 1.8335,
+      "step": 194
+    },
+    {
+      "epoch": 0.2653061224489796,
+      "grad_norm": 0.07961309839255727,
+      "learning_rate": 9.81300036841511e-06,
+      "loss": 1.7037,
+      "step": 195
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.09980119114286523,
+      "learning_rate": 9.809941886340854e-06,
+      "loss": 1.5664,
+      "step": 196
+    },
+    {
+      "epoch": 0.2680272108843537,
+      "grad_norm": 0.07147953268098678,
+      "learning_rate": 9.806859079617757e-06,
+      "loss": 1.7601,
+      "step": 197
+    },
+    {
+      "epoch": 0.2693877551020408,
+      "grad_norm": 0.08653388534305975,
+      "learning_rate": 9.803751963836065e-06,
+      "loss": 1.54,
+      "step": 198
+    },
+    {
+      "epoch": 0.2707482993197279,
+      "grad_norm": 0.0776685121413518,
+      "learning_rate": 9.800620554708962e-06,
+      "loss": 1.5557,
+      "step": 199
+    },
+    {
+      "epoch": 0.272108843537415,
+      "grad_norm": 0.07711567735740478,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 1.7101,
+      "step": 200
+    },
+    {
+      "epoch": 0.27346938775510204,
+      "grad_norm": 0.09355853262847387,
+      "learning_rate": 9.794284919885456e-06,
+      "loss": 1.7454,
+      "step": 201
+    },
+    {
+      "epoch": 0.2748299319727891,
+      "grad_norm": 0.0975587232648776,
+      "learning_rate": 9.791080726229376e-06,
+      "loss": 1.7479,
+      "step": 202
+    },
+    {
+      "epoch": 0.2761904761904762,
+      "grad_norm": 0.07709180794261607,
+      "learning_rate": 9.78785230330837e-06,
+      "loss": 1.8086,
+      "step": 203
+    },
+    {
+      "epoch": 0.27755102040816326,
+      "grad_norm": 0.09748041740615765,
+      "learning_rate": 9.784599667449088e-06,
+      "loss": 1.683,
+      "step": 204
+    },
+    {
+      "epoch": 0.2789115646258503,
+      "grad_norm": 0.09608384188874226,
+      "learning_rate": 9.781322835100639e-06,
+      "loss": 1.7985,
+      "step": 205
+    },
+    {
+      "epoch": 0.2802721088435374,
+      "grad_norm": 0.24417626356607502,
+      "learning_rate": 9.778021822834484e-06,
+      "loss": 1.721,
+      "step": 206
+    },
+    {
+      "epoch": 0.2816326530612245,
+      "grad_norm": 0.08000490559142356,
+      "learning_rate": 9.774696647344376e-06,
+      "loss": 1.5646,
+      "step": 207
+    },
+    {
+      "epoch": 0.2829931972789116,
+      "grad_norm": 0.08448656773098678,
+      "learning_rate": 9.771347325446261e-06,
+      "loss": 1.7897,
+      "step": 208
+    },
+    {
+      "epoch": 0.28435374149659864,
+      "grad_norm": 0.09777185072102372,
+      "learning_rate": 9.767973874078196e-06,
+      "loss": 1.8829,
+      "step": 209
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.07490737419370372,
+      "learning_rate": 9.764576310300268e-06,
+      "loss": 1.8031,
+      "step": 210
+    },
+    {
+      "epoch": 0.2870748299319728,
+      "grad_norm": 0.07567900709987288,
+      "learning_rate": 9.761154651294505e-06,
+      "loss": 1.752,
+      "step": 211
+    },
+    {
+      "epoch": 0.28843537414965986,
+      "grad_norm": 0.08322325294858353,
+      "learning_rate": 9.757708914364784e-06,
+      "loss": 1.6328,
+      "step": 212
+    },
+    {
+      "epoch": 0.2897959183673469,
+      "grad_norm": 0.19949313188011264,
+      "learning_rate": 9.75423911693675e-06,
+      "loss": 1.8577,
+      "step": 213
+    },
+    {
+      "epoch": 0.291156462585034,
+      "grad_norm": 0.07497191387563905,
+      "learning_rate": 9.750745276557725e-06,
+      "loss": 1.4911,
+      "step": 214
+    },
+    {
+      "epoch": 0.2925170068027211,
+      "grad_norm": 0.09096003185892962,
+      "learning_rate": 9.747227410896624e-06,
+      "loss": 1.5857,
+      "step": 215
+    },
+    {
+      "epoch": 0.2938775510204082,
+      "grad_norm": 0.06790968657114778,
+      "learning_rate": 9.743685537743856e-06,
+      "loss": 1.6452,
+      "step": 216
+    },
+    {
+      "epoch": 0.29523809523809524,
+      "grad_norm": 0.08487437350333037,
+      "learning_rate": 9.740119675011246e-06,
+      "loss": 1.674,
+      "step": 217
+    },
+    {
+      "epoch": 0.2965986394557823,
+      "grad_norm": 0.07405858389177783,
+      "learning_rate": 9.73652984073193e-06,
+      "loss": 1.7461,
+      "step": 218
+    },
+    {
+      "epoch": 0.2979591836734694,
+      "grad_norm": 0.07067520018576251,
+      "learning_rate": 9.73291605306028e-06,
+      "loss": 1.7163,
+      "step": 219
+    },
+    {
+      "epoch": 0.29931972789115646,
+      "grad_norm": 0.0791420166635673,
+      "learning_rate": 9.7292783302718e-06,
+      "loss": 1.6668,
+      "step": 220
+    },
+    {
+      "epoch": 0.3006802721088435,
+      "grad_norm": 0.07323313348575836,
+      "learning_rate": 9.72561669076304e-06,
+      "loss": 1.6398,
+      "step": 221
+    },
+    {
+      "epoch": 0.3020408163265306,
+      "grad_norm": 0.07634813609350796,
+      "learning_rate": 9.721931153051497e-06,
+      "loss": 1.6447,
+      "step": 222
+    },
+    {
+      "epoch": 0.3034013605442177,
+      "grad_norm": 0.07950888254230533,
+      "learning_rate": 9.718221735775527e-06,
+      "loss": 1.7845,
+      "step": 223
+    },
+    {
+      "epoch": 0.3047619047619048,
+      "grad_norm": 0.1408011781580588,
+      "learning_rate": 9.714488457694252e-06,
+      "loss": 1.7427,
+      "step": 224
+    },
+    {
+      "epoch": 0.30612244897959184,
+      "grad_norm": 0.20352696620915875,
+      "learning_rate": 9.710731337687457e-06,
+      "loss": 1.7789,
+      "step": 225
+    },
+    {
+      "epoch": 0.3074829931972789,
+      "grad_norm": 0.3632349628769343,
+      "learning_rate": 9.7069503947555e-06,
+      "loss": 1.7108,
+      "step": 226
+    },
+    {
+      "epoch": 0.308843537414966,
+      "grad_norm": 0.08426345577870951,
+      "learning_rate": 9.70314564801922e-06,
+      "loss": 1.5991,
+      "step": 227
+    },
+    {
+      "epoch": 0.31020408163265306,
+      "grad_norm": 0.07487391878859809,
+      "learning_rate": 9.699317116719831e-06,
+      "loss": 1.6637,
+      "step": 228
+    },
+    {
+      "epoch": 0.3115646258503401,
+      "grad_norm": 0.11408922719163124,
+      "learning_rate": 9.695464820218829e-06,
+      "loss": 1.734,
+      "step": 229
+    },
+    {
+      "epoch": 0.3129251700680272,
+      "grad_norm": 0.08379918487601977,
+      "learning_rate": 9.6915887779979e-06,
+      "loss": 1.5813,
+      "step": 230
+    },
+    {
+      "epoch": 0.3142857142857143,
+      "grad_norm": 0.08052280305893883,
+      "learning_rate": 9.68768900965881e-06,
+      "loss": 1.7164,
+      "step": 231
+    },
+    {
+      "epoch": 0.31564625850340133,
+      "grad_norm": 0.09154693865406958,
+      "learning_rate": 9.683765534923315e-06,
+      "loss": 1.5906,
+      "step": 232
+    },
+    {
+      "epoch": 0.31700680272108844,
+      "grad_norm": 0.1156976576631727,
+      "learning_rate": 9.679818373633054e-06,
+      "loss": 1.5045,
+      "step": 233
+    },
+    {
+      "epoch": 0.3183673469387755,
+      "grad_norm": 0.08364003073959134,
+      "learning_rate": 9.67584754574946e-06,
+      "loss": 1.6152,
+      "step": 234
+    },
+    {
+      "epoch": 0.3197278911564626,
+      "grad_norm": 0.15089067639580192,
+      "learning_rate": 9.671853071353645e-06,
+      "loss": 1.7127,
+      "step": 235
+    },
+    {
+      "epoch": 0.32108843537414966,
+      "grad_norm": 0.08835383872875176,
+      "learning_rate": 9.667834970646309e-06,
+      "loss": 1.609,
+      "step": 236
+    },
+    {
+      "epoch": 0.3224489795918367,
+      "grad_norm": 0.08056944247122935,
+      "learning_rate": 9.663793263947631e-06,
+      "loss": 1.6126,
+      "step": 237
+    },
+    {
+      "epoch": 0.3238095238095238,
+      "grad_norm": 0.09989875040942321,
+      "learning_rate": 9.659727971697173e-06,
+      "loss": 1.8035,
+      "step": 238
+    },
+    {
+      "epoch": 0.3251700680272109,
+      "grad_norm": 0.07493814768096231,
+      "learning_rate": 9.655639114453771e-06,
+      "loss": 1.813,
+      "step": 239
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 0.0879835131172389,
+      "learning_rate": 9.651526712895431e-06,
+      "loss": 1.6926,
+      "step": 240
+    },
+    {
+      "epoch": 0.32789115646258504,
+      "grad_norm": 0.1315730713741491,
+      "learning_rate": 9.647390787819232e-06,
+      "loss": 1.6993,
+      "step": 241
+    },
+    {
+      "epoch": 0.3292517006802721,
+      "grad_norm": 0.08144482513159658,
+      "learning_rate": 9.643231360141205e-06,
+      "loss": 1.5821,
+      "step": 242
+    },
+    {
+      "epoch": 0.3306122448979592,
+      "grad_norm": 0.11213233561578728,
+      "learning_rate": 9.639048450896251e-06,
+      "loss": 1.6491,
+      "step": 243
+    },
+    {
+      "epoch": 0.33197278911564626,
+      "grad_norm": 0.10094097099195502,
+      "learning_rate": 9.63484208123801e-06,
+      "loss": 1.5492,
+      "step": 244
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.06549445299823263,
+      "learning_rate": 9.630612272438771e-06,
+      "loss": 1.6378,
+      "step": 245
+    },
+    {
+      "epoch": 0.3346938775510204,
+      "grad_norm": 0.09866738594204459,
+      "learning_rate": 9.626359045889356e-06,
+      "loss": 1.6712,
+      "step": 246
+    },
+    {
+      "epoch": 0.3360544217687075,
+      "grad_norm": 0.08348510716160104,
+      "learning_rate": 9.622082423099013e-06,
+      "loss": 1.6177,
+      "step": 247
+    },
+    {
+      "epoch": 0.33741496598639453,
+      "grad_norm": 0.0912979636888496,
+      "learning_rate": 9.617782425695314e-06,
+      "loss": 1.7233,
+      "step": 248
+    },
+    {
+      "epoch": 0.33877551020408164,
+      "grad_norm": 0.0730054447204009,
+      "learning_rate": 9.613459075424033e-06,
+      "loss": 1.8876,
+      "step": 249
+    },
+    {
+      "epoch": 0.3401360544217687,
+      "grad_norm": 0.07708564338244163,
+      "learning_rate": 9.609112394149052e-06,
+      "loss": 1.5562,
+      "step": 250
+    },
+    {
+      "epoch": 0.3414965986394558,
+      "grad_norm": 0.09288554194206548,
+      "learning_rate": 9.604742403852232e-06,
+      "loss": 1.7512,
+      "step": 251
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 0.09029592521774728,
+      "learning_rate": 9.600349126633317e-06,
+      "loss": 1.4964,
+      "step": 252
+    },
+    {
+      "epoch": 0.3442176870748299,
+      "grad_norm": 0.09268667220104135,
+      "learning_rate": 9.595932584709815e-06,
+      "loss": 1.5166,
+      "step": 253
+    },
+    {
+      "epoch": 0.345578231292517,
+      "grad_norm": 0.09895745636981777,
+      "learning_rate": 9.59149280041689e-06,
+      "loss": 1.5862,
+      "step": 254
+    },
+    {
+      "epoch": 0.3469387755102041,
+      "grad_norm": 0.0918028367563306,
+      "learning_rate": 9.587029796207246e-06,
+      "loss": 1.6704,
+      "step": 255
+    },
+    {
+      "epoch": 0.34829931972789113,
+      "grad_norm": 0.08184214411162406,
+      "learning_rate": 9.582543594651006e-06,
+      "loss": 1.5767,
+      "step": 256
+    },
+    {
+      "epoch": 0.34965986394557824,
+      "grad_norm": 0.10380891219881462,
+      "learning_rate": 9.578034218435618e-06,
+      "loss": 1.7974,
+      "step": 257
+    },
+    {
+      "epoch": 0.3510204081632653,
+      "grad_norm": 0.07763979596015386,
+      "learning_rate": 9.573501690365718e-06,
+      "loss": 1.6754,
+      "step": 258
+    },
+    {
+      "epoch": 0.3523809523809524,
+      "grad_norm": 0.0974676538104556,
+      "learning_rate": 9.568946033363032e-06,
+      "loss": 1.7312,
+      "step": 259
+    },
+    {
+      "epoch": 0.35374149659863946,
+      "grad_norm": 0.08571433404959614,
+      "learning_rate": 9.564367270466247e-06,
+      "loss": 1.5805,
+      "step": 260
+    },
+    {
+      "epoch": 0.3551020408163265,
+      "grad_norm": 0.0818876254257716,
+      "learning_rate": 9.559765424830903e-06,
+      "loss": 1.7883,
+      "step": 261
+    },
+    {
+      "epoch": 0.3564625850340136,
+      "grad_norm": 0.0803950448066755,
+      "learning_rate": 9.555140519729273e-06,
+      "loss": 1.7474,
+      "step": 262
+    },
+    {
+      "epoch": 0.3578231292517007,
+      "grad_norm": 0.0843722550488927,
+      "learning_rate": 9.550492578550246e-06,
+      "loss": 1.5564,
+      "step": 263
+    },
+    {
+      "epoch": 0.35918367346938773,
+      "grad_norm": 0.07803324012301414,
+      "learning_rate": 9.545821624799205e-06,
+      "loss": 1.4879,
+      "step": 264
+    },
+    {
+      "epoch": 0.36054421768707484,
+      "grad_norm": 0.08256837435665124,
+      "learning_rate": 9.541127682097916e-06,
+      "loss": 1.7395,
+      "step": 265
+    },
+    {
+      "epoch": 0.3619047619047619,
+      "grad_norm": 0.08789940175813588,
+      "learning_rate": 9.536410774184397e-06,
+      "loss": 1.6602,
+      "step": 266
+    },
+    {
+      "epoch": 0.363265306122449,
+      "grad_norm": 0.10680679462820718,
+      "learning_rate": 9.531670924912814e-06,
+      "loss": 1.4675,
+      "step": 267
+    },
+    {
+      "epoch": 0.36462585034013606,
+      "grad_norm": 0.07750163404143413,
+      "learning_rate": 9.526908158253345e-06,
+      "loss": 1.7119,
+      "step": 268
+    },
+    {
+      "epoch": 0.3659863945578231,
+      "grad_norm": 0.07081517755227287,
+      "learning_rate": 9.522122498292066e-06,
+      "loss": 1.6457,
+      "step": 269
+    },
+    {
+      "epoch": 0.3673469387755102,
+      "grad_norm": 0.10501830739522407,
+      "learning_rate": 9.517313969230826e-06,
+      "loss": 1.6398,
+      "step": 270
+    },
+    {
+      "epoch": 0.3687074829931973,
+      "grad_norm": 0.08143325631533527,
+      "learning_rate": 9.512482595387131e-06,
+      "loss": 1.6122,
+      "step": 271
+    },
+    {
+      "epoch": 0.37006802721088433,
+      "grad_norm": 0.08749817297460229,
+      "learning_rate": 9.507628401194015e-06,
+      "loss": 1.7328,
+      "step": 272
+    },
+    {
+      "epoch": 0.37142857142857144,
+      "grad_norm": 0.10516442293102286,
+      "learning_rate": 9.50275141119992e-06,
+      "loss": 1.5773,
+      "step": 273
+    },
+    {
+      "epoch": 0.3727891156462585,
+      "grad_norm": 0.07808789214492048,
+      "learning_rate": 9.497851650068561e-06,
+      "loss": 1.6635,
+      "step": 274
+    },
+    {
+      "epoch": 0.3741496598639456,
+      "grad_norm": 0.17789525917278,
+      "learning_rate": 9.492929142578823e-06,
+      "loss": 1.9121,
+      "step": 275
+    },
+    {
+      "epoch": 0.37551020408163266,
+      "grad_norm": 0.11269638220817121,
+      "learning_rate": 9.487983913624616e-06,
+      "loss": 1.7355,
+      "step": 276
+    },
+    {
+      "epoch": 0.3768707482993197,
+      "grad_norm": 0.08409991854502387,
+      "learning_rate": 9.483015988214757e-06,
+      "loss": 1.7628,
+      "step": 277
+    },
+    {
+      "epoch": 0.3782312925170068,
+      "grad_norm": 0.07365655913118094,
+      "learning_rate": 9.478025391472841e-06,
+      "loss": 1.8144,
+      "step": 278
+    },
+    {
+      "epoch": 0.3795918367346939,
+      "grad_norm": 0.12778690022382183,
+      "learning_rate": 9.473012148637121e-06,
+      "loss": 1.4851,
+      "step": 279
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 0.0809734685514322,
+      "learning_rate": 9.467976285060369e-06,
+      "loss": 1.7698,
+      "step": 280
+    },
+    {
+      "epoch": 0.38231292517006804,
+      "grad_norm": 0.08513218534341072,
+      "learning_rate": 9.462917826209757e-06,
+      "loss": 1.6411,
+      "step": 281
+    },
+    {
+      "epoch": 0.3836734693877551,
+      "grad_norm": 0.09008342709764051,
+      "learning_rate": 9.457836797666722e-06,
+      "loss": 1.694,
+      "step": 282
+    },
+    {
+      "epoch": 0.38503401360544215,
+      "grad_norm": 0.10502121760541401,
+      "learning_rate": 9.452733225126845e-06,
+      "loss": 1.6999,
+      "step": 283
+    },
+    {
+      "epoch": 0.38639455782312926,
+      "grad_norm": 0.07343666145039363,
+      "learning_rate": 9.44760713439971e-06,
+      "loss": 1.8164,
+      "step": 284
+    },
+    {
+      "epoch": 0.3877551020408163,
+      "grad_norm": 0.08974960620233877,
+      "learning_rate": 9.442458551408784e-06,
+      "loss": 1.8539,
+      "step": 285
+    },
+    {
+      "epoch": 0.3891156462585034,
+      "grad_norm": 0.07202421778040075,
+      "learning_rate": 9.437287502191275e-06,
+      "loss": 1.5453,
+      "step": 286
+    },
+    {
+      "epoch": 0.3904761904761905,
+      "grad_norm": 0.09076340883522513,
+      "learning_rate": 9.43209401289801e-06,
+      "loss": 1.7088,
+      "step": 287
+    },
+    {
+      "epoch": 0.39183673469387753,
+      "grad_norm": 0.08425026537963505,
+      "learning_rate": 9.426878109793301e-06,
+      "loss": 1.4451,
+      "step": 288
+    },
+    {
+      "epoch": 0.39319727891156464,
+      "grad_norm": 0.09510405982528822,
+      "learning_rate": 9.421639819254806e-06,
+      "loss": 1.7913,
+      "step": 289
+    },
+    {
+      "epoch": 0.3945578231292517,
+      "grad_norm": 0.09616833843974483,
+      "learning_rate": 9.416379167773403e-06,
+      "loss": 1.649,
+      "step": 290
+    },
+    {
+      "epoch": 0.39591836734693875,
+      "grad_norm": 0.08286660774561835,
+      "learning_rate": 9.41109618195305e-06,
+      "loss": 1.9277,
+      "step": 291
+    },
+    {
+      "epoch": 0.39727891156462586,
+      "grad_norm": 0.07769697723139755,
+      "learning_rate": 9.405790888510655e-06,
+      "loss": 1.7279,
+      "step": 292
+    },
+    {
+      "epoch": 0.3986394557823129,
+      "grad_norm": 0.09006259560100753,
+      "learning_rate": 9.400463314275942e-06,
+      "loss": 1.6039,
+      "step": 293
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.09012228611455256,
+      "learning_rate": 9.39511348619131e-06,
+      "loss": 1.7865,
+      "step": 294
+    },
+    {
+      "epoch": 0.4013605442176871,
+      "grad_norm": 0.09086893554478499,
+      "learning_rate": 9.389741431311694e-06,
+      "loss": 1.6225,
+      "step": 295
+    },
+    {
+      "epoch": 0.40272108843537413,
+      "grad_norm": 0.1067223732758171,
+      "learning_rate": 9.384347176804441e-06,
+      "loss": 1.8657,
+      "step": 296
+    },
+    {
+      "epoch": 0.40408163265306124,
+      "grad_norm": 0.09585514004003831,
+      "learning_rate": 9.378930749949166e-06,
+      "loss": 1.6826,
+      "step": 297
+    },
+    {
+      "epoch": 0.4054421768707483,
+      "grad_norm": 0.08230460323355868,
+      "learning_rate": 9.373492178137606e-06,
+      "loss": 1.8107,
+      "step": 298
+    },
+    {
+      "epoch": 0.40680272108843535,
+      "grad_norm": 0.08574613920900533,
+      "learning_rate": 9.368031488873492e-06,
+      "loss": 1.5687,
+      "step": 299
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 0.07911826770354155,
+      "learning_rate": 9.36254870977241e-06,
+      "loss": 1.8716,
+      "step": 300
+    },
+    {
+      "epoch": 0.4095238095238095,
+      "grad_norm": 0.08540782320726534,
+      "learning_rate": 9.357043868561653e-06,
+      "loss": 1.7997,
+      "step": 301
+    },
+    {
+      "epoch": 0.4108843537414966,
+      "grad_norm": 0.08850925770828139,
+      "learning_rate": 9.351516993080088e-06,
+      "loss": 1.6299,
+      "step": 302
+    },
+    {
+      "epoch": 0.4122448979591837,
+      "grad_norm": 0.10611011164886108,
+      "learning_rate": 9.34596811127801e-06,
+      "loss": 1.5621,
+      "step": 303
+    },
+    {
+      "epoch": 0.41360544217687073,
+      "grad_norm": 0.08382183266579103,
+      "learning_rate": 9.340397251217009e-06,
+      "loss": 1.4407,
+      "step": 304
+    },
+    {
+      "epoch": 0.41496598639455784,
+      "grad_norm": 0.09119095993947843,
+      "learning_rate": 9.334804441069819e-06,
+      "loss": 1.7161,
+      "step": 305
+    },
+    {
+      "epoch": 0.4163265306122449,
+      "grad_norm": 0.11113995697266435,
+      "learning_rate": 9.329189709120175e-06,
+      "loss": 1.4126,
+      "step": 306
+    },
+    {
+      "epoch": 0.41768707482993195,
+      "grad_norm": 0.08080147586730375,
+      "learning_rate": 9.323553083762681e-06,
+      "loss": 1.7303,
+      "step": 307
+    },
+    {
+      "epoch": 0.41904761904761906,
+      "grad_norm": 0.10411476203370171,
+      "learning_rate": 9.31789459350266e-06,
+      "loss": 1.6408,
+      "step": 308
+    },
+    {
+      "epoch": 0.4204081632653061,
+      "grad_norm": 0.0859822277343485,
+      "learning_rate": 9.312214266956003e-06,
+      "loss": 1.6534,
+      "step": 309
+    },
+    {
+      "epoch": 0.4217687074829932,
+      "grad_norm": 0.08604098070322257,
+      "learning_rate": 9.306512132849035e-06,
+      "loss": 1.6252,
+      "step": 310
+    },
+    {
+      "epoch": 0.4231292517006803,
+      "grad_norm": 0.0876425921649466,
+      "learning_rate": 9.300788220018363e-06,
+      "loss": 1.5096,
+      "step": 311
+    },
+    {
+      "epoch": 0.42448979591836733,
+      "grad_norm": 0.08371274563563173,
+      "learning_rate": 9.295042557410736e-06,
+      "loss": 1.7352,
+      "step": 312
+    },
+    {
+      "epoch": 0.42585034013605444,
+      "grad_norm": 0.13500827487489858,
+      "learning_rate": 9.28927517408289e-06,
+      "loss": 1.7902,
+      "step": 313
+    },
+    {
+      "epoch": 0.4272108843537415,
+      "grad_norm": 0.08754620765852711,
+      "learning_rate": 9.28348609920141e-06,
+      "loss": 1.6862,
+      "step": 314
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.1028946128162606,
+      "learning_rate": 9.27767536204258e-06,
+      "loss": 1.6907,
+      "step": 315
+    },
+    {
+      "epoch": 0.42993197278911566,
+      "grad_norm": 0.08692259223714764,
+      "learning_rate": 9.271842991992231e-06,
+      "loss": 1.638,
+      "step": 316
+    },
+    {
+      "epoch": 0.4312925170068027,
+      "grad_norm": 0.09542502295403,
+      "learning_rate": 9.26598901854559e-06,
+      "loss": 1.7157,
+      "step": 317
+    },
+    {
+      "epoch": 0.4326530612244898,
+      "grad_norm": 0.09080935991935338,
+      "learning_rate": 9.260113471307148e-06,
+      "loss": 1.5851,
+      "step": 318
+    },
+    {
+      "epoch": 0.4340136054421769,
+      "grad_norm": 0.09865728458566181,
+      "learning_rate": 9.254216379990487e-06,
+      "loss": 1.8897,
+      "step": 319
+    },
+    {
+      "epoch": 0.43537414965986393,
+      "grad_norm": 0.10060290482375228,
+      "learning_rate": 9.248297774418147e-06,
+      "loss": 1.5605,
+      "step": 320
+    },
+    {
+      "epoch": 0.43673469387755104,
+      "grad_norm": 0.09013488977054508,
+      "learning_rate": 9.242357684521467e-06,
+      "loss": 1.5582,
+      "step": 321
+    },
+    {
+      "epoch": 0.4380952380952381,
+      "grad_norm": 0.12978917401977683,
+      "learning_rate": 9.236396140340435e-06,
+      "loss": 1.5953,
+      "step": 322
+    },
+    {
+      "epoch": 0.43945578231292515,
+      "grad_norm": 0.09012186897343348,
+      "learning_rate": 9.230413172023538e-06,
+      "loss": 1.6678,
+      "step": 323
+    },
+    {
+      "epoch": 0.44081632653061226,
+      "grad_norm": 0.07610784588472097,
+      "learning_rate": 9.224408809827609e-06,
+      "loss": 1.6697,
+      "step": 324
+    },
+    {
+      "epoch": 0.4421768707482993,
+      "grad_norm": 0.08134414493064245,
+      "learning_rate": 9.218383084117671e-06,
+      "loss": 1.6543,
+      "step": 325
+    },
+    {
+      "epoch": 0.4435374149659864,
+      "grad_norm": 0.07969199857220131,
+      "learning_rate": 9.212336025366789e-06,
+      "loss": 1.7372,
+      "step": 326
+    },
+    {
+      "epoch": 0.4448979591836735,
+      "grad_norm": 0.08833397568245774,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 1.5033,
+      "step": 327
+    },
+    {
+      "epoch": 0.44625850340136053,
+      "grad_norm": 0.09013321784578471,
+      "learning_rate": 9.200178031173706e-06,
+      "loss": 1.7467,
+      "step": 328
+    },
+    {
+      "epoch": 0.44761904761904764,
+      "grad_norm": 0.08492129873211993,
+      "learning_rate": 9.194067157216436e-06,
+      "loss": 1.6346,
+      "step": 329
+    },
+    {
+      "epoch": 0.4489795918367347,
+      "grad_norm": 0.08924441822794496,
+      "learning_rate": 9.187935073187768e-06,
+      "loss": 1.5647,
+      "step": 330
+    },
+    {
+      "epoch": 0.45034013605442175,
+      "grad_norm": 0.2238269401441882,
+      "learning_rate": 9.181781810098638e-06,
+      "loss": 1.9641,
+      "step": 331
+    },
+    {
+      "epoch": 0.45170068027210886,
+      "grad_norm": 0.08505567098719835,
+      "learning_rate": 9.175607399067086e-06,
+      "loss": 1.723,
+      "step": 332
+    },
+    {
+      "epoch": 0.4530612244897959,
+      "grad_norm": 0.09479499636885363,
+      "learning_rate": 9.1694118713181e-06,
+      "loss": 1.7358,
+      "step": 333
+    },
+    {
+      "epoch": 0.454421768707483,
+      "grad_norm": 0.09532298638421347,
+      "learning_rate": 9.163195258183457e-06,
+      "loss": 1.652,
+      "step": 334
+    },
+    {
+      "epoch": 0.4557823129251701,
+      "grad_norm": 0.1435078453546247,
+      "learning_rate": 9.156957591101573e-06,
+      "loss": 1.8876,
+      "step": 335
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 0.08623460392050768,
+      "learning_rate": 9.150698901617326e-06,
+      "loss": 1.6408,
+      "step": 336
+    },
+    {
+      "epoch": 0.45850340136054424,
+      "grad_norm": 0.08103596406817255,
+      "learning_rate": 9.144419221381919e-06,
+      "loss": 1.582,
+      "step": 337
+    },
+    {
+      "epoch": 0.4598639455782313,
+      "grad_norm": 0.09624169989634061,
+      "learning_rate": 9.138118582152704e-06,
+      "loss": 1.7272,
+      "step": 338
+    },
+    {
+      "epoch": 0.46122448979591835,
+      "grad_norm": 0.06914577882609961,
+      "learning_rate": 9.131797015793026e-06,
+      "loss": 1.6864,
+      "step": 339
+    },
+    {
+      "epoch": 0.46258503401360546,
+      "grad_norm": 0.09682023357095138,
+      "learning_rate": 9.125454554272057e-06,
+      "loss": 1.5849,
+      "step": 340
+    },
+    {
+      "epoch": 0.4639455782312925,
+      "grad_norm": 0.08865054891775723,
+      "learning_rate": 9.119091229664648e-06,
+      "loss": 1.4716,
+      "step": 341
+    },
+    {
+      "epoch": 0.46530612244897956,
+      "grad_norm": 0.09112991799939003,
+      "learning_rate": 9.112707074151152e-06,
+      "loss": 1.6393,
+      "step": 342
+    },
+    {
+      "epoch": 0.4666666666666667,
+      "grad_norm": 0.09319883560181061,
+      "learning_rate": 9.106302120017272e-06,
+      "loss": 1.7619,
+      "step": 343
+    },
+    {
+      "epoch": 0.46802721088435373,
+      "grad_norm": 0.101061665585339,
+      "learning_rate": 9.099876399653885e-06,
+      "loss": 1.6286,
+      "step": 344
+    },
+    {
+      "epoch": 0.46938775510204084,
+      "grad_norm": 0.09755047037445551,
+      "learning_rate": 9.093429945556895e-06,
+      "loss": 1.6591,
+      "step": 345
+    },
+    {
+      "epoch": 0.4707482993197279,
+      "grad_norm": 0.0831755062746902,
+      "learning_rate": 9.086962790327057e-06,
+      "loss": 1.7728,
+      "step": 346
+    },
+    {
+      "epoch": 0.47210884353741495,
+      "grad_norm": 0.0981445280966388,
+      "learning_rate": 9.08047496666981e-06,
+      "loss": 1.6489,
+      "step": 347
+    },
+    {
+      "epoch": 0.47346938775510206,
+      "grad_norm": 0.0965447984049988,
+      "learning_rate": 9.073966507395123e-06,
+      "loss": 1.7807,
+      "step": 348
+    },
+    {
+      "epoch": 0.4748299319727891,
+      "grad_norm": 0.08805192634428297,
+      "learning_rate": 9.06743744541732e-06,
+      "loss": 1.6932,
+      "step": 349
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.07486129688956443,
+      "learning_rate": 9.060887813754914e-06,
+      "loss": 1.6828,
+      "step": 350
+    },
+    {
+      "epoch": 0.4775510204081633,
+      "grad_norm": 0.08844318223023526,
+      "learning_rate": 9.054317645530449e-06,
+      "loss": 1.5791,
+      "step": 351
+    },
+    {
+      "epoch": 0.47891156462585033,
+      "grad_norm": 0.08590901507611927,
+      "learning_rate": 9.047726973970317e-06,
+      "loss": 1.8916,
+      "step": 352
+    },
+    {
+      "epoch": 0.48027210884353744,
+      "grad_norm": 0.08275711528371654,
+      "learning_rate": 9.041115832404605e-06,
+      "loss": 1.6376,
+      "step": 353
+    },
+    {
+      "epoch": 0.4816326530612245,
+      "grad_norm": 0.13882142863507996,
+      "learning_rate": 9.03448425426692e-06,
+      "loss": 1.5496,
+      "step": 354
+    },
+    {
+      "epoch": 0.48299319727891155,
+      "grad_norm": 0.09033930791413211,
+      "learning_rate": 9.027832273094213e-06,
+      "loss": 1.8207,
+      "step": 355
+    },
+    {
+      "epoch": 0.48435374149659866,
+      "grad_norm": 0.08776380923196873,
+      "learning_rate": 9.021159922526623e-06,
+      "loss": 1.734,
+      "step": 356
+    },
+    {
+      "epoch": 0.4857142857142857,
+      "grad_norm": 0.0871797042611384,
+      "learning_rate": 9.014467236307303e-06,
+      "loss": 1.7255,
+      "step": 357
+    },
+    {
+      "epoch": 0.48707482993197276,
+      "grad_norm": 0.0832915611192237,
+      "learning_rate": 9.007754248282236e-06,
+      "loss": 1.6354,
+      "step": 358
+    },
+    {
+      "epoch": 0.4884353741496599,
+      "grad_norm": 0.11961492483949776,
+      "learning_rate": 9.001020992400086e-06,
+      "loss": 1.4193,
+      "step": 359
+    },
+    {
+      "epoch": 0.4897959183673469,
+      "grad_norm": 0.0794404046382169,
+      "learning_rate": 8.994267502712007e-06,
+      "loss": 1.728,
+      "step": 360
+    },
+    {
+      "epoch": 0.49115646258503404,
+      "grad_norm": 0.13335980644840648,
+      "learning_rate": 8.987493813371481e-06,
+      "loss": 1.6729,
+      "step": 361
+    },
+    {
+      "epoch": 0.4925170068027211,
+      "grad_norm": 0.08449457187355984,
+      "learning_rate": 8.980699958634147e-06,
+      "loss": 1.6142,
+      "step": 362
+    },
+    {
+      "epoch": 0.49387755102040815,
+      "grad_norm": 0.09270436024065673,
+      "learning_rate": 8.973885972857616e-06,
+      "loss": 1.6753,
+      "step": 363
+    },
+    {
+      "epoch": 0.49523809523809526,
+      "grad_norm": 0.08033630058062084,
+      "learning_rate": 8.96705189050131e-06,
+      "loss": 1.6269,
+      "step": 364
+    },
+    {
+      "epoch": 0.4965986394557823,
+      "grad_norm": 0.4381615422407444,
+      "learning_rate": 8.96019774612628e-06,
+      "loss": 1.6131,
+      "step": 365
+    },
+    {
+      "epoch": 0.49795918367346936,
+      "grad_norm": 0.09462169653835782,
+      "learning_rate": 8.953323574395038e-06,
+      "loss": 1.5629,
+      "step": 366
+    },
+    {
+      "epoch": 0.4993197278911565,
+      "grad_norm": 0.13595506352030012,
+      "learning_rate": 8.946429410071373e-06,
+      "loss": 1.5593,
+      "step": 367
+    },
+    {
+      "epoch": 0.5006802721088436,
+      "grad_norm": 0.10253342646979322,
+      "learning_rate": 8.939515288020182e-06,
+      "loss": 1.6281,
+      "step": 368
+    },
+    {
+      "epoch": 0.5006802721088436,
+      "eval_loss": 1.7046507596969604,
+      "eval_runtime": 76.5686,
+      "eval_samples_per_second": 53.194,
+      "eval_steps_per_second": 6.661,
+      "step": 368
+    },
+    {
+      "epoch": 0.5020408163265306,
+      "grad_norm": 0.08764026597495381,
+      "learning_rate": 8.932581243207289e-06,
+      "loss": 1.5909,
+      "step": 369
+    },
+    {
+      "epoch": 0.5034013605442177,
+      "grad_norm": 0.07898445718193145,
+      "learning_rate": 8.925627310699275e-06,
+      "loss": 1.761,
+      "step": 370
+    },
+    {
+      "epoch": 0.5047619047619047,
+      "grad_norm": 0.10577594540101636,
+      "learning_rate": 8.918653525663295e-06,
+      "loss": 1.695,
+      "step": 371
+    },
+    {
+      "epoch": 0.5061224489795918,
+      "grad_norm": 0.110620128076388,
+      "learning_rate": 8.911659923366897e-06,
+      "loss": 1.7043,
+      "step": 372
+    },
+    {
+      "epoch": 0.507482993197279,
+      "grad_norm": 0.08992699882848652,
+      "learning_rate": 8.904646539177852e-06,
+      "loss": 1.674,
+      "step": 373
+    },
+    {
+      "epoch": 0.508843537414966,
+      "grad_norm": 0.0971954506200858,
+      "learning_rate": 8.897613408563972e-06,
+      "loss": 1.6565,
+      "step": 374
+    },
+    {
+      "epoch": 0.5102040816326531,
+      "grad_norm": 0.14199688300843188,
+      "learning_rate": 8.89056056709293e-06,
+      "loss": 1.4402,
+      "step": 375
+    },
+    {
+      "epoch": 0.5115646258503401,
+      "grad_norm": 0.1034458287611268,
+      "learning_rate": 8.883488050432073e-06,
+      "loss": 1.6606,
+      "step": 376
+    },
+    {
+      "epoch": 0.5129251700680272,
+      "grad_norm": 0.09270052668783225,
+      "learning_rate": 8.87639589434826e-06,
+      "loss": 1.7067,
+      "step": 377
+    },
+    {
+      "epoch": 0.5142857142857142,
+      "grad_norm": 0.08063860406008341,
+      "learning_rate": 8.869284134707659e-06,
+      "loss": 1.7683,
+      "step": 378
+    },
+    {
+      "epoch": 0.5156462585034014,
+      "grad_norm": 0.08700140267251555,
+      "learning_rate": 8.862152807475584e-06,
+      "loss": 1.7135,
+      "step": 379
+    },
+    {
+      "epoch": 0.5170068027210885,
+      "grad_norm": 0.09374652236807894,
+      "learning_rate": 8.8550019487163e-06,
+      "loss": 1.5927,
+      "step": 380
+    },
+    {
+      "epoch": 0.5183673469387755,
+      "grad_norm": 0.09628602949564133,
+      "learning_rate": 8.847831594592851e-06,
+      "loss": 1.6169,
+      "step": 381
+    },
+    {
+      "epoch": 0.5197278911564626,
+      "grad_norm": 0.09133344115171436,
+      "learning_rate": 8.840641781366867e-06,
+      "loss": 1.6077,
+      "step": 382
+    },
+    {
+      "epoch": 0.5210884353741496,
+      "grad_norm": 0.0898264889709745,
+      "learning_rate": 8.83343254539839e-06,
+      "loss": 1.7476,
+      "step": 383
+    },
+    {
+      "epoch": 0.5224489795918368,
+      "grad_norm": 0.09463509154870942,
+      "learning_rate": 8.826203923145687e-06,
+      "loss": 1.6178,
+      "step": 384
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 0.086933970419064,
+      "learning_rate": 8.818955951165059e-06,
+      "loss": 1.6544,
+      "step": 385
+    },
+    {
+      "epoch": 0.5251700680272109,
+      "grad_norm": 0.08686966519278672,
+      "learning_rate": 8.811688666110663e-06,
+      "loss": 1.7268,
+      "step": 386
+    },
+    {
+      "epoch": 0.5265306122448979,
+      "grad_norm": 0.10219001251324068,
+      "learning_rate": 8.80440210473433e-06,
+      "loss": 1.6538,
+      "step": 387
+    },
+    {
+      "epoch": 0.527891156462585,
+      "grad_norm": 0.08199431219832622,
+      "learning_rate": 8.797096303885374e-06,
+      "loss": 1.6524,
+      "step": 388
+    },
+    {
+      "epoch": 0.5292517006802722,
+      "grad_norm": 0.09053864014656055,
+      "learning_rate": 8.789771300510397e-06,
+      "loss": 1.5971,
+      "step": 389
+    },
+    {
+      "epoch": 0.5306122448979592,
+      "grad_norm": 0.08221661934582344,
+      "learning_rate": 8.782427131653121e-06,
+      "loss": 1.6643,
+      "step": 390
+    },
+    {
+      "epoch": 0.5319727891156463,
+      "grad_norm": 0.08888879198967926,
+      "learning_rate": 8.77506383445419e-06,
+      "loss": 1.7855,
+      "step": 391
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 0.08638631079426494,
+      "learning_rate": 8.767681446150977e-06,
+      "loss": 1.8028,
+      "step": 392
+    },
+    {
+      "epoch": 0.5346938775510204,
+      "grad_norm": 0.07669664542559732,
+      "learning_rate": 8.76028000407741e-06,
+      "loss": 1.5725,
+      "step": 393
+    },
+    {
+      "epoch": 0.5360544217687074,
+      "grad_norm": 0.098633841165777,
+      "learning_rate": 8.752859545663766e-06,
+      "loss": 1.6692,
+      "step": 394
+    },
+    {
+      "epoch": 0.5374149659863946,
+      "grad_norm": 0.09002401371176637,
+      "learning_rate": 8.745420108436498e-06,
+      "loss": 1.7636,
+      "step": 395
+    },
+    {
+      "epoch": 0.5387755102040817,
+      "grad_norm": 0.0952379865665774,
+      "learning_rate": 8.737961730018034e-06,
+      "loss": 1.5664,
+      "step": 396
+    },
+    {
+      "epoch": 0.5401360544217687,
+      "grad_norm": 0.09623958141543322,
+      "learning_rate": 8.730484448126594e-06,
+      "loss": 1.5345,
+      "step": 397
+    },
+    {
+      "epoch": 0.5414965986394558,
+      "grad_norm": 0.1502927654981511,
+      "learning_rate": 8.722988300575992e-06,
+      "loss": 1.7841,
+      "step": 398
+    },
+    {
+      "epoch": 0.5428571428571428,
+      "grad_norm": 0.09222431588896472,
+      "learning_rate": 8.71547332527545e-06,
+      "loss": 1.9559,
+      "step": 399
+    },
+    {
+      "epoch": 0.54421768707483,
+      "grad_norm": 0.0937414921537921,
+      "learning_rate": 8.707939560229406e-06,
+      "loss": 1.7022,
+      "step": 400
+    },
+    {
+      "epoch": 0.545578231292517,
+      "grad_norm": 0.11025083403359445,
+      "learning_rate": 8.700387043537319e-06,
+      "loss": 1.4365,
+      "step": 401
+    },
+    {
+      "epoch": 0.5469387755102041,
+      "grad_norm": 0.08441282686563067,
+      "learning_rate": 8.692815813393483e-06,
+      "loss": 1.6488,
+      "step": 402
+    },
+    {
+      "epoch": 0.5482993197278911,
+      "grad_norm": 0.3399602286969607,
+      "learning_rate": 8.68522590808682e-06,
+      "loss": 1.6829,
+      "step": 403
+    },
+    {
+      "epoch": 0.5496598639455782,
+      "grad_norm": 0.0924431892667502,
+      "learning_rate": 8.677617366000705e-06,
+      "loss": 1.7404,
+      "step": 404
+    },
+    {
+      "epoch": 0.5510204081632653,
+      "grad_norm": 0.09778253267340173,
+      "learning_rate": 8.669990225612754e-06,
+      "loss": 1.7674,
+      "step": 405
+    },
+    {
+      "epoch": 0.5523809523809524,
+      "grad_norm": 0.09471837896307483,
+      "learning_rate": 8.662344525494643e-06,
+      "loss": 1.6406,
+      "step": 406
+    },
+    {
+      "epoch": 0.5537414965986395,
+      "grad_norm": 0.08997298756417596,
+      "learning_rate": 8.654680304311908e-06,
+      "loss": 1.7875,
+      "step": 407
+    },
+    {
+      "epoch": 0.5551020408163265,
+      "grad_norm": 0.08849454584462031,
+      "learning_rate": 8.646997600823743e-06,
+      "loss": 1.5942,
+      "step": 408
+    },
+    {
+      "epoch": 0.5564625850340136,
+      "grad_norm": 0.10974758415682516,
+      "learning_rate": 8.639296453882816e-06,
+      "loss": 1.5229,
+      "step": 409
+    },
+    {
+      "epoch": 0.5578231292517006,
+      "grad_norm": 0.09669599431293863,
+      "learning_rate": 8.631576902435063e-06,
+      "loss": 1.7031,
+      "step": 410
+    },
+    {
+      "epoch": 0.5591836734693878,
+      "grad_norm": 0.10337335565704198,
+      "learning_rate": 8.623838985519498e-06,
+      "loss": 1.6138,
+      "step": 411
+    },
+    {
+      "epoch": 0.5605442176870749,
+      "grad_norm": 0.08963980024729686,
+      "learning_rate": 8.616082742268005e-06,
+      "loss": 1.6527,
+      "step": 412
+    },
+    {
+      "epoch": 0.5619047619047619,
+      "grad_norm": 0.13730489095006465,
+      "learning_rate": 8.608308211905159e-06,
+      "loss": 1.5823,
+      "step": 413
+    },
+    {
+      "epoch": 0.563265306122449,
+      "grad_norm": 0.08745105418294691,
+      "learning_rate": 8.600515433748003e-06,
+      "loss": 1.6647,
+      "step": 414
+    },
+    {
+      "epoch": 0.564625850340136,
+      "grad_norm": 0.19321095274209943,
+      "learning_rate": 8.592704447205872e-06,
+      "loss": 1.5218,
+      "step": 415
+    },
+    {
+      "epoch": 0.5659863945578232,
+      "grad_norm": 0.10518406439497321,
+      "learning_rate": 8.584875291780178e-06,
+      "loss": 1.5199,
+      "step": 416
+    },
+    {
+      "epoch": 0.5673469387755102,
+      "grad_norm": 0.09495102640071197,
+      "learning_rate": 8.577028007064218e-06,
+      "loss": 1.6623,
+      "step": 417
+    },
+    {
+      "epoch": 0.5687074829931973,
+      "grad_norm": 0.13232264409241218,
+      "learning_rate": 8.569162632742973e-06,
+      "loss": 1.606,
+      "step": 418
+    },
+    {
+      "epoch": 0.5700680272108843,
+      "grad_norm": 0.09382708523657708,
+      "learning_rate": 8.561279208592902e-06,
+      "loss": 1.6563,
+      "step": 419
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.10154097299186025,
+      "learning_rate": 8.553377774481748e-06,
+      "loss": 1.5177,
+      "step": 420
+    },
+    {
+      "epoch": 0.5727891156462585,
+      "grad_norm": 0.09008372077918206,
+      "learning_rate": 8.545458370368336e-06,
+      "loss": 1.5358,
+      "step": 421
+    },
+    {
+      "epoch": 0.5741496598639456,
+      "grad_norm": 0.11141318964471479,
+      "learning_rate": 8.53752103630236e-06,
+      "loss": 1.5975,
+      "step": 422
+    },
+    {
+      "epoch": 0.5755102040816327,
+      "grad_norm": 0.11594947645844504,
+      "learning_rate": 8.529565812424195e-06,
+      "loss": 1.5417,
+      "step": 423
+    },
+    {
+      "epoch": 0.5768707482993197,
+      "grad_norm": 0.1115243557715104,
+      "learning_rate": 8.521592738964689e-06,
+      "loss": 1.6912,
+      "step": 424
+    },
+    {
+      "epoch": 0.5782312925170068,
+      "grad_norm": 0.10294673717555994,
+      "learning_rate": 8.513601856244951e-06,
+      "loss": 1.6883,
+      "step": 425
+    },
+    {
+      "epoch": 0.5795918367346938,
+      "grad_norm": 0.12234481799540493,
+      "learning_rate": 8.505593204676162e-06,
+      "loss": 1.6903,
+      "step": 426
+    },
+    {
+      "epoch": 0.580952380952381,
+      "grad_norm": 0.1164390253206002,
+      "learning_rate": 8.497566824759359e-06,
+      "loss": 1.6433,
+      "step": 427
+    },
+    {
+      "epoch": 0.582312925170068,
+      "grad_norm": 0.10041564805248225,
+      "learning_rate": 8.489522757085234e-06,
+      "loss": 1.5482,
+      "step": 428
+    },
+    {
+      "epoch": 0.5836734693877551,
+      "grad_norm": 0.10770970938141446,
+      "learning_rate": 8.481461042333929e-06,
+      "loss": 1.6092,
+      "step": 429
+    },
+    {
+      "epoch": 0.5850340136054422,
+      "grad_norm": 0.08821548122605594,
+      "learning_rate": 8.473381721274832e-06,
+      "loss": 1.5793,
+      "step": 430
+    },
+    {
+      "epoch": 0.5863945578231292,
+      "grad_norm": 0.089232378501521,
+      "learning_rate": 8.465284834766365e-06,
+      "loss": 1.6233,
+      "step": 431
+    },
+    {
+      "epoch": 0.5877551020408164,
+      "grad_norm": 0.10250742081774146,
+      "learning_rate": 8.457170423755786e-06,
+      "loss": 1.5625,
+      "step": 432
+    },
+    {
+      "epoch": 0.5891156462585034,
+      "grad_norm": 0.1202037568410257,
+      "learning_rate": 8.449038529278976e-06,
+      "loss": 1.5843,
+      "step": 433
+    },
+    {
+      "epoch": 0.5904761904761905,
+      "grad_norm": 0.1323622148062186,
+      "learning_rate": 8.440889192460232e-06,
+      "loss": 1.808,
+      "step": 434
+    },
+    {
+      "epoch": 0.5918367346938775,
+      "grad_norm": 0.12495144537173114,
+      "learning_rate": 8.432722454512057e-06,
+      "loss": 1.9389,
+      "step": 435
+    },
+    {
+      "epoch": 0.5931972789115646,
+      "grad_norm": 0.09699121967438322,
+      "learning_rate": 8.424538356734957e-06,
+      "loss": 1.7367,
+      "step": 436
+    },
+    {
+      "epoch": 0.5945578231292517,
+      "grad_norm": 0.09801443283628167,
+      "learning_rate": 8.416336940517229e-06,
+      "loss": 1.6276,
+      "step": 437
+    },
+    {
+      "epoch": 0.5959183673469388,
+      "grad_norm": 0.10487807929341668,
+      "learning_rate": 8.408118247334755e-06,
+      "loss": 1.5578,
+      "step": 438
+    },
+    {
+      "epoch": 0.5972789115646259,
+      "grad_norm": 0.09364043239294455,
+      "learning_rate": 8.399882318750785e-06,
+      "loss": 1.5889,
+      "step": 439
+    },
+    {
+      "epoch": 0.5986394557823129,
+      "grad_norm": 0.10020651855110954,
+      "learning_rate": 8.391629196415733e-06,
+      "loss": 1.607,
+      "step": 440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0899206896336364,
+      "learning_rate": 8.383358922066965e-06,
+      "loss": 1.5508,
+      "step": 441
+    },
+    {
+      "epoch": 0.601360544217687,
+      "grad_norm": 0.09009526459055692,
+      "learning_rate": 8.375071537528587e-06,
+      "loss": 1.6629,
+      "step": 442
+    },
+    {
+      "epoch": 0.6027210884353742,
+      "grad_norm": 0.09413295136140958,
+      "learning_rate": 8.366767084711232e-06,
+      "loss": 1.6568,
+      "step": 443
+    },
+    {
+      "epoch": 0.6040816326530613,
+      "grad_norm": 0.22548893033360673,
+      "learning_rate": 8.358445605611856e-06,
+      "loss": 1.7594,
+      "step": 444
+    },
+    {
+      "epoch": 0.6054421768707483,
+      "grad_norm": 0.12757276275079013,
+      "learning_rate": 8.350107142313513e-06,
+      "loss": 1.4311,
+      "step": 445
+    },
+    {
+      "epoch": 0.6068027210884354,
+      "grad_norm": 0.09965051028904835,
+      "learning_rate": 8.34175173698515e-06,
+      "loss": 1.6618,
+      "step": 446
+    },
+    {
+      "epoch": 0.6081632653061224,
+      "grad_norm": 0.09666753861412604,
+      "learning_rate": 8.333379431881398e-06,
+      "loss": 1.6729,
+      "step": 447
+    },
+    {
+      "epoch": 0.6095238095238096,
+      "grad_norm": 0.09235423529974252,
+      "learning_rate": 8.324990269342345e-06,
+      "loss": 1.7872,
+      "step": 448
+    },
+    {
+      "epoch": 0.6108843537414966,
+      "grad_norm": 0.0969813250335363,
+      "learning_rate": 8.316584291793337e-06,
+      "loss": 1.5299,
+      "step": 449
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 0.08468781546097646,
+      "learning_rate": 8.30816154174475e-06,
+      "loss": 1.7958,
+      "step": 450
+    },
+    {
+      "epoch": 0.6136054421768707,
+      "grad_norm": 0.08812390382328499,
+      "learning_rate": 8.299722061791788e-06,
+      "loss": 1.7292,
+      "step": 451
+    },
+    {
+      "epoch": 0.6149659863945578,
+      "grad_norm": 0.08809745700239427,
+      "learning_rate": 8.291265894614253e-06,
+      "loss": 1.758,
+      "step": 452
+    },
+    {
+      "epoch": 0.6163265306122448,
+      "grad_norm": 0.09490957390276747,
+      "learning_rate": 8.282793082976343e-06,
+      "loss": 1.5475,
+      "step": 453
+    },
+    {
+      "epoch": 0.617687074829932,
+      "grad_norm": 0.11060404799058053,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 1.6792,
+      "step": 454
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 0.09540512150005957,
+      "learning_rate": 8.265797697796831e-06,
+      "loss": 1.6685,
+      "step": 455
+    },
+    {
+      "epoch": 0.6204081632653061,
+      "grad_norm": 0.12873788369707956,
+      "learning_rate": 8.257275210203621e-06,
+      "loss": 1.5043,
+      "step": 456
+    },
+    {
+      "epoch": 0.6217687074829932,
+      "grad_norm": 0.6499840705042578,
+      "learning_rate": 8.248736250046389e-06,
+      "loss": 1.7548,
+      "step": 457
+    },
+    {
+      "epoch": 0.6231292517006802,
+      "grad_norm": 0.08484061508587025,
+      "learning_rate": 8.240180860508027e-06,
+      "loss": 1.8159,
+      "step": 458
+    },
+    {
+      "epoch": 0.6244897959183674,
+      "grad_norm": 0.09906794908825851,
+      "learning_rate": 8.231609084854513e-06,
+      "loss": 1.7116,
+      "step": 459
+    },
+    {
+      "epoch": 0.6258503401360545,
+      "grad_norm": 0.0939998149984381,
+      "learning_rate": 8.223020966434695e-06,
+      "loss": 1.7448,
+      "step": 460
+    },
+    {
+      "epoch": 0.6272108843537415,
+      "grad_norm": 0.09983102806108725,
+      "learning_rate": 8.214416548680065e-06,
+      "loss": 1.7284,
+      "step": 461
+    },
+    {
+      "epoch": 0.6285714285714286,
+      "grad_norm": 0.1276929064149538,
+      "learning_rate": 8.205795875104549e-06,
+      "loss": 1.5541,
+      "step": 462
+    },
+    {
+      "epoch": 0.6299319727891156,
+      "grad_norm": 0.0971007724405045,
+      "learning_rate": 8.197158989304277e-06,
+      "loss": 1.749,
+      "step": 463
+    },
+    {
+      "epoch": 0.6312925170068027,
+      "grad_norm": 0.09931797410533018,
+      "learning_rate": 8.188505934957368e-06,
+      "loss": 1.7908,
+      "step": 464
+    },
+    {
+      "epoch": 0.6326530612244898,
+      "grad_norm": 0.10420469743883767,
+      "learning_rate": 8.179836755823707e-06,
+      "loss": 1.7156,
+      "step": 465
+    },
+    {
+      "epoch": 0.6340136054421769,
+      "grad_norm": 0.09471973183090822,
+      "learning_rate": 8.171151495744726e-06,
+      "loss": 1.6598,
+      "step": 466
+    },
+    {
+      "epoch": 0.6353741496598639,
+      "grad_norm": 0.08008463828970452,
+      "learning_rate": 8.162450198643184e-06,
+      "loss": 1.8476,
+      "step": 467
+    },
+    {
+      "epoch": 0.636734693877551,
+      "grad_norm": 0.09336869405905367,
+      "learning_rate": 8.153732908522933e-06,
+      "loss": 1.677,
+      "step": 468
+    },
+    {
+      "epoch": 0.638095238095238,
+      "grad_norm": 0.10567938348565759,
+      "learning_rate": 8.144999669468714e-06,
+      "loss": 1.6987,
+      "step": 469
+    },
+    {
+      "epoch": 0.6394557823129252,
+      "grad_norm": 0.08460548198303047,
+      "learning_rate": 8.136250525645916e-06,
+      "loss": 1.8206,
+      "step": 470
+    },
+    {
+      "epoch": 0.6408163265306123,
+      "grad_norm": 0.0993816872028852,
+      "learning_rate": 8.127485521300366e-06,
+      "loss": 1.6618,
+      "step": 471
+    },
+    {
+      "epoch": 0.6421768707482993,
+      "grad_norm": 0.1043492135258874,
+      "learning_rate": 8.118704700758103e-06,
+      "loss": 1.641,
+      "step": 472
+    },
+    {
+      "epoch": 0.6435374149659864,
+      "grad_norm": 0.0880711093010387,
+      "learning_rate": 8.109908108425142e-06,
+      "loss": 1.8376,
+      "step": 473
+    },
+    {
+      "epoch": 0.6448979591836734,
+      "grad_norm": 0.09271039929379037,
+      "learning_rate": 8.101095788787266e-06,
+      "loss": 1.6914,
+      "step": 474
+    },
+    {
+      "epoch": 0.6462585034013606,
+      "grad_norm": 0.10199949996500506,
+      "learning_rate": 8.092267786409788e-06,
+      "loss": 1.6264,
+      "step": 475
+    },
+    {
+      "epoch": 0.6476190476190476,
+      "grad_norm": 0.09927764715965239,
+      "learning_rate": 8.08342414593734e-06,
+      "loss": 1.7495,
+      "step": 476
+    },
+    {
+      "epoch": 0.6489795918367347,
+      "grad_norm": 0.11011814904081185,
+      "learning_rate": 8.07456491209363e-06,
+      "loss": 1.8044,
+      "step": 477
+    },
+    {
+      "epoch": 0.6503401360544218,
+      "grad_norm": 0.10592895331787452,
+      "learning_rate": 8.065690129681224e-06,
+      "loss": 1.5279,
+      "step": 478
+    },
+    {
+      "epoch": 0.6517006802721088,
+      "grad_norm": 0.09546781848020944,
+      "learning_rate": 8.056799843581326e-06,
+      "loss": 1.5599,
+      "step": 479
+    },
+    {
+      "epoch": 0.6530612244897959,
+      "grad_norm": 0.09305244725756194,
+      "learning_rate": 8.04789409875354e-06,
+      "loss": 1.7328,
+      "step": 480
+    },
+    {
+      "epoch": 0.654421768707483,
+      "grad_norm": 0.10356673246511783,
+      "learning_rate": 8.038972940235647e-06,
+      "loss": 1.6317,
+      "step": 481
+    },
+    {
+      "epoch": 0.6557823129251701,
+      "grad_norm": 0.08355158958046802,
+      "learning_rate": 8.030036413143382e-06,
+      "loss": 1.823,
+      "step": 482
+    },
+    {
+      "epoch": 0.6571428571428571,
+      "grad_norm": 0.13429081910577273,
+      "learning_rate": 8.021084562670193e-06,
+      "loss": 1.765,
+      "step": 483
+    },
+    {
+      "epoch": 0.6585034013605442,
+      "grad_norm": 0.10977848927929611,
+      "learning_rate": 8.012117434087032e-06,
+      "loss": 1.7983,
+      "step": 484
+    },
+    {
+      "epoch": 0.6598639455782312,
+      "grad_norm": 0.10127247770991282,
+      "learning_rate": 8.003135072742106e-06,
+      "loss": 1.7146,
+      "step": 485
+    },
+    {
+      "epoch": 0.6612244897959184,
+      "grad_norm": 0.08420649737520919,
+      "learning_rate": 7.994137524060656e-06,
+      "loss": 1.7273,
+      "step": 486
+    },
+    {
+      "epoch": 0.6625850340136055,
+      "grad_norm": 0.12998243453204658,
+      "learning_rate": 7.985124833544737e-06,
+      "loss": 1.7116,
+      "step": 487
+    },
+    {
+      "epoch": 0.6639455782312925,
+      "grad_norm": 0.12572177154996375,
+      "learning_rate": 7.976097046772971e-06,
+      "loss": 1.5875,
+      "step": 488
+    },
+    {
+      "epoch": 0.6653061224489796,
+      "grad_norm": 0.09257333856558611,
+      "learning_rate": 7.967054209400325e-06,
+      "loss": 1.6259,
+      "step": 489
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.15599333098996487,
+      "learning_rate": 7.95799636715788e-06,
+      "loss": 1.6602,
+      "step": 490
+    },
+    {
+      "epoch": 0.6680272108843538,
+      "grad_norm": 0.09770606447991517,
+      "learning_rate": 7.948923565852597e-06,
+      "loss": 1.6867,
+      "step": 491
+    },
+    {
+      "epoch": 0.6693877551020408,
+      "grad_norm": 0.1303849167483497,
+      "learning_rate": 7.939835851367097e-06,
+      "loss": 1.6583,
+      "step": 492
+    },
+    {
+      "epoch": 0.6707482993197279,
+      "grad_norm": 0.10864319676746689,
+      "learning_rate": 7.930733269659405e-06,
+      "loss": 1.6832,
+      "step": 493
+    },
+    {
+      "epoch": 0.672108843537415,
+      "grad_norm": 0.11083699474291044,
+      "learning_rate": 7.921615866762743e-06,
+      "loss": 1.7117,
+      "step": 494
+    },
+    {
+      "epoch": 0.673469387755102,
+      "grad_norm": 0.13148294366762545,
+      "learning_rate": 7.912483688785281e-06,
+      "loss": 1.5234,
+      "step": 495
+    },
+    {
+      "epoch": 0.6748299319727891,
+      "grad_norm": 0.11365843655270903,
+      "learning_rate": 7.903336781909911e-06,
+      "loss": 1.7783,
+      "step": 496
+    },
+    {
+      "epoch": 0.6761904761904762,
+      "grad_norm": 0.12277794049039219,
+      "learning_rate": 7.89417519239401e-06,
+      "loss": 1.6908,
+      "step": 497
+    },
+    {
+      "epoch": 0.6775510204081633,
+      "grad_norm": 0.11413949336950495,
+      "learning_rate": 7.884998966569206e-06,
+      "loss": 1.5654,
+      "step": 498
+    },
+    {
+      "epoch": 0.6789115646258503,
+      "grad_norm": 0.10494845825770326,
+      "learning_rate": 7.87580815084115e-06,
+      "loss": 1.5383,
+      "step": 499
+    },
+    {
+      "epoch": 0.6802721088435374,
+      "grad_norm": 0.09020097927859841,
+      "learning_rate": 7.866602791689272e-06,
+      "loss": 1.596,
+      "step": 500
+    },
+    {
+      "epoch": 0.6816326530612244,
+      "grad_norm": 0.08585696598298209,
+      "learning_rate": 7.857382935666554e-06,
+      "loss": 1.7307,
+      "step": 501
+    },
+    {
+      "epoch": 0.6829931972789116,
+      "grad_norm": 0.1010505486947337,
+      "learning_rate": 7.848148629399287e-06,
+      "loss": 1.6699,
+      "step": 502
+    },
+    {
+      "epoch": 0.6843537414965987,
+      "grad_norm": 0.11201229130009283,
+      "learning_rate": 7.838899919586841e-06,
+      "loss": 1.6521,
+      "step": 503
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 0.10009251853964633,
+      "learning_rate": 7.82963685300143e-06,
+      "loss": 1.6464,
+      "step": 504
+    },
+    {
+      "epoch": 0.6870748299319728,
+      "grad_norm": 0.10676770701676057,
+      "learning_rate": 7.820359476487866e-06,
+      "loss": 1.4472,
+      "step": 505
+    },
+    {
+      "epoch": 0.6884353741496598,
+      "grad_norm": 0.10987902455707375,
+      "learning_rate": 7.811067836963337e-06,
+      "loss": 1.6637,
+      "step": 506
+    },
+    {
+      "epoch": 0.689795918367347,
+      "grad_norm": 0.10593286251613945,
+      "learning_rate": 7.801761981417152e-06,
+      "loss": 1.714,
+      "step": 507
+    },
+    {
+      "epoch": 0.691156462585034,
+      "grad_norm": 0.11171983777047709,
+      "learning_rate": 7.792441956910523e-06,
+      "loss": 1.5948,
+      "step": 508
+    },
+    {
+      "epoch": 0.6925170068027211,
+      "grad_norm": 0.10208550757531555,
+      "learning_rate": 7.783107810576306e-06,
+      "loss": 1.7267,
+      "step": 509
+    },
+    {
+      "epoch": 0.6938775510204082,
+      "grad_norm": 0.11315560578865806,
+      "learning_rate": 7.773759589618782e-06,
+      "loss": 1.5995,
+      "step": 510
+    },
+    {
+      "epoch": 0.6952380952380952,
+      "grad_norm": 0.10501036944532048,
+      "learning_rate": 7.764397341313403e-06,
+      "loss": 1.4624,
+      "step": 511
+    },
+    {
+      "epoch": 0.6965986394557823,
+      "grad_norm": 0.08619282794989483,
+      "learning_rate": 7.755021113006567e-06,
+      "loss": 1.7983,
+      "step": 512
+    },
+    {
+      "epoch": 0.6979591836734694,
+      "grad_norm": 0.09109224074659802,
+      "learning_rate": 7.745630952115365e-06,
+      "loss": 1.6753,
+      "step": 513
+    },
+    {
+      "epoch": 0.6993197278911565,
+      "grad_norm": 0.11332860880884088,
+      "learning_rate": 7.736226906127344e-06,
+      "loss": 1.7472,
+      "step": 514
+    },
+    {
+      "epoch": 0.7006802721088435,
+      "grad_norm": 0.11046061973302813,
+      "learning_rate": 7.726809022600284e-06,
+      "loss": 1.6219,
+      "step": 515
+    },
+    {
+      "epoch": 0.7020408163265306,
+      "grad_norm": 0.09720494400811149,
+      "learning_rate": 7.71737734916193e-06,
+      "loss": 1.7941,
+      "step": 516
+    },
+    {
+      "epoch": 0.7034013605442176,
+      "grad_norm": 0.13391377607878455,
+      "learning_rate": 7.70793193350977e-06,
+      "loss": 1.7904,
+      "step": 517
+    },
+    {
+      "epoch": 0.7047619047619048,
+      "grad_norm": 0.12625091217599146,
+      "learning_rate": 7.69847282341079e-06,
+      "loss": 1.663,
+      "step": 518
+    },
+    {
+      "epoch": 0.7061224489795919,
+      "grad_norm": 0.12406607142823292,
+      "learning_rate": 7.68900006670123e-06,
+      "loss": 1.6766,
+      "step": 519
+    },
+    {
+      "epoch": 0.7074829931972789,
+      "grad_norm": 0.09475581974666519,
+      "learning_rate": 7.679513711286338e-06,
+      "loss": 1.7449,
+      "step": 520
+    },
+    {
+      "epoch": 0.708843537414966,
+      "grad_norm": 0.12489019591501085,
+      "learning_rate": 7.670013805140143e-06,
+      "loss": 1.7526,
+      "step": 521
+    },
+    {
+      "epoch": 0.710204081632653,
+      "grad_norm": 0.09091443476585943,
+      "learning_rate": 7.660500396305194e-06,
+      "loss": 1.66,
+      "step": 522
+    },
+    {
+      "epoch": 0.7115646258503401,
+      "grad_norm": 0.0843896278465122,
+      "learning_rate": 7.650973532892325e-06,
+      "loss": 1.5741,
+      "step": 523
+    },
+    {
+      "epoch": 0.7129251700680272,
+      "grad_norm": 0.12842839378581072,
+      "learning_rate": 7.641433263080418e-06,
+      "loss": 1.5639,
+      "step": 524
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.12573336181400996,
+      "learning_rate": 7.631879635116152e-06,
+      "loss": 1.5261,
+      "step": 525
+    },
+    {
+      "epoch": 0.7156462585034014,
+      "grad_norm": 0.08357817960907306,
+      "learning_rate": 7.622312697313754e-06,
+      "loss": 1.624,
+      "step": 526
+    },
+    {
+      "epoch": 0.7170068027210884,
+      "grad_norm": 0.08826301097458839,
+      "learning_rate": 7.612732498054769e-06,
+      "loss": 1.7131,
+      "step": 527
+    },
+    {
+      "epoch": 0.7183673469387755,
+      "grad_norm": 0.1030553177752236,
+      "learning_rate": 7.603139085787801e-06,
+      "loss": 1.76,
+      "step": 528
+    },
+    {
+      "epoch": 0.7197278911564626,
+      "grad_norm": 0.10248942600565085,
+      "learning_rate": 7.5935325090282785e-06,
+      "loss": 1.6537,
+      "step": 529
+    },
+    {
+      "epoch": 0.7210884353741497,
+      "grad_norm": 0.10935378440777248,
+      "learning_rate": 7.583912816358203e-06,
+      "loss": 1.7441,
+      "step": 530
+    },
+    {
+      "epoch": 0.7224489795918367,
+      "grad_norm": 0.10474278332332895,
+      "learning_rate": 7.574280056425907e-06,
+      "loss": 1.5672,
+      "step": 531
+    },
+    {
+      "epoch": 0.7238095238095238,
+      "grad_norm": 0.10337823669221702,
+      "learning_rate": 7.564634277945803e-06,
+      "loss": 1.7301,
+      "step": 532
+    },
+    {
+      "epoch": 0.7251700680272108,
+      "grad_norm": 0.1021232974428954,
+      "learning_rate": 7.554975529698143e-06,
+      "loss": 1.8401,
+      "step": 533
+    },
+    {
+      "epoch": 0.726530612244898,
+      "grad_norm": 0.10873875060914909,
+      "learning_rate": 7.54530386052877e-06,
+      "loss": 1.6832,
+      "step": 534
+    },
+    {
+      "epoch": 0.7278911564625851,
+      "grad_norm": 0.13142427274605614,
+      "learning_rate": 7.5356193193488655e-06,
+      "loss": 1.6824,
+      "step": 535
+    },
+    {
+      "epoch": 0.7292517006802721,
+      "grad_norm": 0.10578967924155344,
+      "learning_rate": 7.525921955134714e-06,
+      "loss": 1.6128,
+      "step": 536
+    },
+    {
+      "epoch": 0.7306122448979592,
+      "grad_norm": 0.1428684712147148,
+      "learning_rate": 7.5162118169274424e-06,
+      "loss": 1.6909,
+      "step": 537
+    },
+    {
+      "epoch": 0.7319727891156462,
+      "grad_norm": 0.12364317671003774,
+      "learning_rate": 7.506488953832779e-06,
+      "loss": 1.5894,
+      "step": 538
+    },
+    {
+      "epoch": 0.7333333333333333,
+      "grad_norm": 0.12070521447402052,
+      "learning_rate": 7.4967534150208066e-06,
+      "loss": 1.6316,
+      "step": 539
+    },
+    {
+      "epoch": 0.7346938775510204,
+      "grad_norm": 0.11805699210709567,
+      "learning_rate": 7.487005249725705e-06,
+      "loss": 1.773,
+      "step": 540
+    },
+    {
+      "epoch": 0.7360544217687075,
+      "grad_norm": 0.10518788665559951,
+      "learning_rate": 7.477244507245517e-06,
+      "loss": 1.5496,
+      "step": 541
+    },
+    {
+      "epoch": 0.7374149659863946,
+      "grad_norm": 0.10115679657141723,
+      "learning_rate": 7.4674712369418815e-06,
+      "loss": 1.6332,
+      "step": 542
+    },
+    {
+      "epoch": 0.7387755102040816,
+      "grad_norm": 0.13249620123305414,
+      "learning_rate": 7.457685488239799e-06,
+      "loss": 1.5464,
+      "step": 543
+    },
+    {
+      "epoch": 0.7401360544217687,
+      "grad_norm": 0.10912963076663591,
+      "learning_rate": 7.44788731062737e-06,
+      "loss": 1.6537,
+      "step": 544
+    },
+    {
+      "epoch": 0.7414965986394558,
+      "grad_norm": 0.0903784233977396,
+      "learning_rate": 7.438076753655557e-06,
+      "loss": 1.7509,
+      "step": 545
+    },
+    {
+      "epoch": 0.7428571428571429,
+      "grad_norm": 0.11534137833637581,
+      "learning_rate": 7.4282538669379186e-06,
+      "loss": 1.8423,
+      "step": 546
+    },
+    {
+      "epoch": 0.7442176870748299,
+      "grad_norm": 0.10673090155703877,
+      "learning_rate": 7.418418700150373e-06,
+      "loss": 1.5147,
+      "step": 547
+    },
+    {
+      "epoch": 0.745578231292517,
+      "grad_norm": 0.11078677610408977,
+      "learning_rate": 7.408571303030939e-06,
+      "loss": 1.598,
+      "step": 548
+    },
+    {
+      "epoch": 0.746938775510204,
+      "grad_norm": 0.20089268434188828,
+      "learning_rate": 7.398711725379486e-06,
+      "loss": 1.6854,
+      "step": 549
+    },
+    {
+      "epoch": 0.7482993197278912,
+      "grad_norm": 0.11465528276050316,
+      "learning_rate": 7.388840017057479e-06,
+      "loss": 1.7166,
+      "step": 550
+    },
+    {
+      "epoch": 0.7496598639455783,
+      "grad_norm": 0.17554597081264026,
+      "learning_rate": 7.378956227987738e-06,
+      "loss": 1.7621,
+      "step": 551
+    },
+    {
+      "epoch": 0.7510204081632653,
+      "grad_norm": 0.12001150745168689,
+      "learning_rate": 7.369060408154166e-06,
+      "loss": 1.6292,
+      "step": 552
+    },
+    {
+      "epoch": 0.7510204081632653,
+      "eval_loss": 1.695604681968689,
+      "eval_runtime": 76.6065,
+      "eval_samples_per_second": 53.168,
+      "eval_steps_per_second": 6.657,
+      "step": 552
+    },
+    {
+      "epoch": 0.7523809523809524,
+      "grad_norm": 0.08810105735536967,
+      "learning_rate": 7.35915260760152e-06,
+      "loss": 1.7169,
+      "step": 553
+    },
+    {
+      "epoch": 0.7537414965986394,
+      "grad_norm": 0.1670593915051014,
+      "learning_rate": 7.349232876435135e-06,
+      "loss": 1.5579,
+      "step": 554
+    },
+    {
+      "epoch": 0.7551020408163265,
+      "grad_norm": 0.11454927724613997,
+      "learning_rate": 7.3393012648206865e-06,
+      "loss": 1.7283,
+      "step": 555
+    },
+    {
+      "epoch": 0.7564625850340136,
+      "grad_norm": 0.969049897753043,
+      "learning_rate": 7.329357822983929e-06,
+      "loss": 1.7205,
+      "step": 556
+    },
+    {
+      "epoch": 0.7578231292517007,
+      "grad_norm": 0.14156929561935405,
+      "learning_rate": 7.319402601210448e-06,
+      "loss": 1.6642,
+      "step": 557
+    },
+    {
+      "epoch": 0.7591836734693878,
+      "grad_norm": 0.0984076585901956,
+      "learning_rate": 7.3094356498453955e-06,
+      "loss": 1.5543,
+      "step": 558
+    },
+    {
+      "epoch": 0.7605442176870748,
+      "grad_norm": 0.1065633653807634,
+      "learning_rate": 7.299457019293248e-06,
+      "loss": 1.6024,
+      "step": 559
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.10385154170413329,
+      "learning_rate": 7.289466760017543e-06,
+      "loss": 1.6121,
+      "step": 560
+    },
+    {
+      "epoch": 0.763265306122449,
+      "grad_norm": 0.09706098665043214,
+      "learning_rate": 7.279464922540626e-06,
+      "loss": 1.6291,
+      "step": 561
+    },
+    {
+      "epoch": 0.7646258503401361,
+      "grad_norm": 0.1005953850437904,
+      "learning_rate": 7.269451557443396e-06,
+      "loss": 1.5871,
+      "step": 562
+    },
+    {
+      "epoch": 0.7659863945578231,
+      "grad_norm": 0.12409919864299088,
+      "learning_rate": 7.2594267153650525e-06,
+      "loss": 1.8507,
+      "step": 563
+    },
+    {
+      "epoch": 0.7673469387755102,
+      "grad_norm": 0.17577412545797552,
+      "learning_rate": 7.249390447002827e-06,
+      "loss": 1.6741,
+      "step": 564
+    },
+    {
+      "epoch": 0.7687074829931972,
+      "grad_norm": 0.10625351066460387,
+      "learning_rate": 7.239342803111744e-06,
+      "loss": 1.6995,
+      "step": 565
+    },
+    {
+      "epoch": 0.7700680272108843,
+      "grad_norm": 0.10141675080832888,
+      "learning_rate": 7.229283834504351e-06,
+      "loss": 1.7018,
+      "step": 566
+    },
+    {
+      "epoch": 0.7714285714285715,
+      "grad_norm": 0.10351854356175742,
+      "learning_rate": 7.21921359205047e-06,
+      "loss": 1.6347,
+      "step": 567
+    },
+    {
+      "epoch": 0.7727891156462585,
+      "grad_norm": 0.097783522808633,
+      "learning_rate": 7.209132126676934e-06,
+      "loss": 1.63,
+      "step": 568
+    },
+    {
+      "epoch": 0.7741496598639456,
+      "grad_norm": 0.1032937811881391,
+      "learning_rate": 7.199039489367334e-06,
+      "loss": 1.7088,
+      "step": 569
+    },
+    {
+      "epoch": 0.7755102040816326,
+      "grad_norm": 0.12703981258728564,
+      "learning_rate": 7.188935731161756e-06,
+      "loss": 1.5488,
+      "step": 570
+    },
+    {
+      "epoch": 0.7768707482993197,
+      "grad_norm": 0.09603678350259663,
+      "learning_rate": 7.178820903156532e-06,
+      "loss": 1.7006,
+      "step": 571
+    },
+    {
+      "epoch": 0.7782312925170068,
+      "grad_norm": 0.10137848978882107,
+      "learning_rate": 7.168695056503967e-06,
+      "loss": 1.5343,
+      "step": 572
+    },
+    {
+      "epoch": 0.7795918367346939,
+      "grad_norm": 0.09168277145084121,
+      "learning_rate": 7.1585582424121005e-06,
+      "loss": 1.7654,
+      "step": 573
+    },
+    {
+      "epoch": 0.780952380952381,
+      "grad_norm": 0.10323000544095898,
+      "learning_rate": 7.148410512144425e-06,
+      "loss": 1.6613,
+      "step": 574
+    },
+    {
+      "epoch": 0.782312925170068,
+      "grad_norm": 0.09038228211904249,
+      "learning_rate": 7.138251917019645e-06,
+      "loss": 1.7182,
+      "step": 575
+    },
+    {
+      "epoch": 0.7836734693877551,
+      "grad_norm": 0.12949706977298758,
+      "learning_rate": 7.1280825084114065e-06,
+      "loss": 1.5075,
+      "step": 576
+    },
+    {
+      "epoch": 0.7850340136054422,
+      "grad_norm": 0.10182930523451113,
+      "learning_rate": 7.117902337748045e-06,
+      "loss": 1.5249,
+      "step": 577
+    },
+    {
+      "epoch": 0.7863945578231293,
+      "grad_norm": 0.1014784602992118,
+      "learning_rate": 7.107711456512316e-06,
+      "loss": 1.5699,
+      "step": 578
+    },
+    {
+      "epoch": 0.7877551020408163,
+      "grad_norm": 0.08920528739055578,
+      "learning_rate": 7.097509916241145e-06,
+      "loss": 1.7604,
+      "step": 579
+    },
+    {
+      "epoch": 0.7891156462585034,
+      "grad_norm": 0.14771070265391437,
+      "learning_rate": 7.08729776852536e-06,
+      "loss": 1.8294,
+      "step": 580
+    },
+    {
+      "epoch": 0.7904761904761904,
+      "grad_norm": 0.10437668016493229,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 1.5263,
+      "step": 581
+    },
+    {
+      "epoch": 0.7918367346938775,
+      "grad_norm": 0.09453693869246346,
+      "learning_rate": 7.066841857391215e-06,
+      "loss": 1.7625,
+      "step": 582
+    },
+    {
+      "epoch": 0.7931972789115647,
+      "grad_norm": 0.1024749155019567,
+      "learning_rate": 7.056598197421686e-06,
+      "loss": 1.6953,
+      "step": 583
+    },
+    {
+      "epoch": 0.7945578231292517,
+      "grad_norm": 0.09967417247994723,
+      "learning_rate": 7.046344136904675e-06,
+      "loss": 1.5067,
+      "step": 584
+    },
+    {
+      "epoch": 0.7959183673469388,
+      "grad_norm": 0.10667510272444436,
+      "learning_rate": 7.036079727696618e-06,
+      "loss": 1.6966,
+      "step": 585
+    },
+    {
+      "epoch": 0.7972789115646258,
+      "grad_norm": 0.13727868268941767,
+      "learning_rate": 7.025805021706276e-06,
+      "loss": 1.6554,
+      "step": 586
+    },
+    {
+      "epoch": 0.7986394557823129,
+      "grad_norm": 0.09241703194497365,
+      "learning_rate": 7.0155200708944915e-06,
+      "loss": 1.7987,
+      "step": 587
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.11984006315877656,
+      "learning_rate": 7.005224927273913e-06,
+      "loss": 1.7059,
+      "step": 588
+    },
+    {
+      "epoch": 0.8013605442176871,
+      "grad_norm": 0.12763094516083248,
+      "learning_rate": 6.9949196429087355e-06,
+      "loss": 1.8147,
+      "step": 589
+    },
+    {
+      "epoch": 0.8027210884353742,
+      "grad_norm": 0.13522443685391508,
+      "learning_rate": 6.984604269914437e-06,
+      "loss": 1.63,
+      "step": 590
+    },
+    {
+      "epoch": 0.8040816326530612,
+      "grad_norm": 0.10116021604335325,
+      "learning_rate": 6.974278860457515e-06,
+      "loss": 1.5963,
+      "step": 591
+    },
+    {
+      "epoch": 0.8054421768707483,
+      "grad_norm": 0.10234671686214797,
+      "learning_rate": 6.963943466755225e-06,
+      "loss": 1.7491,
+      "step": 592
+    },
+    {
+      "epoch": 0.8068027210884354,
+      "grad_norm": 0.09993833464748478,
+      "learning_rate": 6.953598141075315e-06,
+      "loss": 1.8742,
+      "step": 593
+    },
+    {
+      "epoch": 0.8081632653061225,
+      "grad_norm": 0.10194114128873782,
+      "learning_rate": 6.943242935735757e-06,
+      "loss": 1.8295,
+      "step": 594
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 0.11091877191061177,
+      "learning_rate": 6.932877903104487e-06,
+      "loss": 1.7282,
+      "step": 595
+    },
+    {
+      "epoch": 0.8108843537414966,
+      "grad_norm": 0.11885618706334389,
+      "learning_rate": 6.922503095599142e-06,
+      "loss": 1.7013,
+      "step": 596
+    },
+    {
+      "epoch": 0.8122448979591836,
+      "grad_norm": 0.128832076499377,
+      "learning_rate": 6.912118565686789e-06,
+      "loss": 1.6604,
+      "step": 597
+    },
+    {
+      "epoch": 0.8136054421768707,
+      "grad_norm": 0.149361446054299,
+      "learning_rate": 6.901724365883665e-06,
+      "loss": 1.5922,
+      "step": 598
+    },
+    {
+      "epoch": 0.8149659863945579,
+      "grad_norm": 0.10349101338251292,
+      "learning_rate": 6.89132054875491e-06,
+      "loss": 1.7742,
+      "step": 599
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 0.10814162257394357,
+      "learning_rate": 6.8809071669142946e-06,
+      "loss": 1.6099,
+      "step": 600
+    },
+    {
+      "epoch": 0.817687074829932,
+      "grad_norm": 0.09463224320256956,
+      "learning_rate": 6.870484273023967e-06,
+      "loss": 1.5986,
+      "step": 601
+    },
+    {
+      "epoch": 0.819047619047619,
+      "grad_norm": 0.10081423766781461,
+      "learning_rate": 6.8600519197941725e-06,
+      "loss": 1.7488,
+      "step": 602
+    },
+    {
+      "epoch": 0.8204081632653061,
+      "grad_norm": 0.10845349706729722,
+      "learning_rate": 6.849610159983003e-06,
+      "loss": 1.6419,
+      "step": 603
+    },
+    {
+      "epoch": 0.8217687074829932,
+      "grad_norm": 0.1220236426902432,
+      "learning_rate": 6.839159046396109e-06,
+      "loss": 1.6193,
+      "step": 604
+    },
+    {
+      "epoch": 0.8231292517006803,
+      "grad_norm": 0.11987974076038253,
+      "learning_rate": 6.828698631886455e-06,
+      "loss": 1.6836,
+      "step": 605
+    },
+    {
+      "epoch": 0.8244897959183674,
+      "grad_norm": 0.12488732229202766,
+      "learning_rate": 6.8182289693540375e-06,
+      "loss": 1.6057,
+      "step": 606
+    },
+    {
+      "epoch": 0.8258503401360544,
+      "grad_norm": 0.10061110967809365,
+      "learning_rate": 6.807750111745619e-06,
+      "loss": 1.6481,
+      "step": 607
+    },
+    {
+      "epoch": 0.8272108843537415,
+      "grad_norm": 0.10585066880846151,
+      "learning_rate": 6.797262112054469e-06,
+      "loss": 1.6665,
+      "step": 608
+    },
+    {
+      "epoch": 0.8285714285714286,
+      "grad_norm": 0.10134141638102989,
+      "learning_rate": 6.786765023320085e-06,
+      "loss": 1.5092,
+      "step": 609
+    },
+    {
+      "epoch": 0.8299319727891157,
+      "grad_norm": 0.11227140982751639,
+      "learning_rate": 6.776258898627932e-06,
+      "loss": 1.7522,
+      "step": 610
+    },
+    {
+      "epoch": 0.8312925170068027,
+      "grad_norm": 0.08967360374579285,
+      "learning_rate": 6.765743791109172e-06,
+      "loss": 1.7738,
+      "step": 611
+    },
+    {
+      "epoch": 0.8326530612244898,
+      "grad_norm": 0.12577528209737293,
+      "learning_rate": 6.755219753940389e-06,
+      "loss": 1.6958,
+      "step": 612
+    },
+    {
+      "epoch": 0.8340136054421768,
+      "grad_norm": 0.1026220373525454,
+      "learning_rate": 6.744686840343333e-06,
+      "loss": 1.7081,
+      "step": 613
+    },
+    {
+      "epoch": 0.8353741496598639,
+      "grad_norm": 0.1616576311932069,
+      "learning_rate": 6.734145103584638e-06,
+      "loss": 1.7878,
+      "step": 614
+    },
+    {
+      "epoch": 0.8367346938775511,
+      "grad_norm": 0.11606031099035814,
+      "learning_rate": 6.72359459697556e-06,
+      "loss": 1.704,
+      "step": 615
+    },
+    {
+      "epoch": 0.8380952380952381,
+      "grad_norm": 0.11405134687997359,
+      "learning_rate": 6.713035373871711e-06,
+      "loss": 1.6157,
+      "step": 616
+    },
+    {
+      "epoch": 0.8394557823129252,
+      "grad_norm": 0.11855263544564604,
+      "learning_rate": 6.702467487672771e-06,
+      "loss": 1.7325,
+      "step": 617
+    },
+    {
+      "epoch": 0.8408163265306122,
+      "grad_norm": 0.10438275658978799,
+      "learning_rate": 6.691890991822243e-06,
+      "loss": 1.6522,
+      "step": 618
+    },
+    {
+      "epoch": 0.8421768707482993,
+      "grad_norm": 0.10374577213787556,
+      "learning_rate": 6.681305939807165e-06,
+      "loss": 1.6307,
+      "step": 619
+    },
+    {
+      "epoch": 0.8435374149659864,
+      "grad_norm": 0.09336351222681238,
+      "learning_rate": 6.670712385157846e-06,
+      "loss": 1.5821,
+      "step": 620
+    },
+    {
+      "epoch": 0.8448979591836735,
+      "grad_norm": 0.12997638800995778,
+      "learning_rate": 6.660110381447593e-06,
+      "loss": 1.672,
+      "step": 621
+    },
+    {
+      "epoch": 0.8462585034013606,
+      "grad_norm": 0.09931623670142671,
+      "learning_rate": 6.649499982292441e-06,
+      "loss": 1.6305,
+      "step": 622
+    },
+    {
+      "epoch": 0.8476190476190476,
+      "grad_norm": 0.1164641533977229,
+      "learning_rate": 6.638881241350884e-06,
+      "loss": 1.6891,
+      "step": 623
+    },
+    {
+      "epoch": 0.8489795918367347,
+      "grad_norm": 0.09080847058543648,
+      "learning_rate": 6.628254212323601e-06,
+      "loss": 1.5685,
+      "step": 624
+    },
+    {
+      "epoch": 0.8503401360544217,
+      "grad_norm": 0.09554345130938086,
+      "learning_rate": 6.617618948953186e-06,
+      "loss": 1.8238,
+      "step": 625
+    },
+    {
+      "epoch": 0.8517006802721089,
+      "grad_norm": 0.11850979230452785,
+      "learning_rate": 6.606975505023874e-06,
+      "loss": 1.5686,
+      "step": 626
+    },
+    {
+      "epoch": 0.8530612244897959,
+      "grad_norm": 0.11391698206139707,
+      "learning_rate": 6.596323934361268e-06,
+      "loss": 1.6122,
+      "step": 627
+    },
+    {
+      "epoch": 0.854421768707483,
+      "grad_norm": 0.1259510837772197,
+      "learning_rate": 6.5856642908320745e-06,
+      "loss": 1.5638,
+      "step": 628
+    },
+    {
+      "epoch": 0.85578231292517,
+      "grad_norm": 0.3467765021073813,
+      "learning_rate": 6.574996628343824e-06,
+      "loss": 1.7503,
+      "step": 629
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.11342921408966115,
+      "learning_rate": 6.564321000844598e-06,
+      "loss": 1.6653,
+      "step": 630
+    },
+    {
+      "epoch": 0.8585034013605443,
+      "grad_norm": 0.10000357784567925,
+      "learning_rate": 6.553637462322759e-06,
+      "loss": 1.5783,
+      "step": 631
+    },
+    {
+      "epoch": 0.8598639455782313,
+      "grad_norm": 0.1067459993698176,
+      "learning_rate": 6.5429460668066825e-06,
+      "loss": 1.7222,
+      "step": 632
+    },
+    {
+      "epoch": 0.8612244897959184,
+      "grad_norm": 0.1669410376966875,
+      "learning_rate": 6.5322468683644665e-06,
+      "loss": 1.6325,
+      "step": 633
+    },
+    {
+      "epoch": 0.8625850340136054,
+      "grad_norm": 0.107409115579869,
+      "learning_rate": 6.5215399211036815e-06,
+      "loss": 1.5369,
+      "step": 634
+    },
+    {
+      "epoch": 0.8639455782312925,
+      "grad_norm": 0.1076970404147157,
+      "learning_rate": 6.510825279171077e-06,
+      "loss": 1.7722,
+      "step": 635
+    },
+    {
+      "epoch": 0.8653061224489796,
+      "grad_norm": 0.15949100738404212,
+      "learning_rate": 6.5001029967523195e-06,
+      "loss": 1.5295,
+      "step": 636
+    },
+    {
+      "epoch": 0.8666666666666667,
+      "grad_norm": 0.13488187893377007,
+      "learning_rate": 6.489373128071714e-06,
+      "loss": 1.6053,
+      "step": 637
+    },
+    {
+      "epoch": 0.8680272108843538,
+      "grad_norm": 0.1117979139478637,
+      "learning_rate": 6.4786357273919296e-06,
+      "loss": 1.6219,
+      "step": 638
+    },
+    {
+      "epoch": 0.8693877551020408,
+      "grad_norm": 0.11161396243370707,
+      "learning_rate": 6.467890849013728e-06,
+      "loss": 1.6193,
+      "step": 639
+    },
+    {
+      "epoch": 0.8707482993197279,
+      "grad_norm": 0.10619453607937132,
+      "learning_rate": 6.4571385472756835e-06,
+      "loss": 1.6587,
+      "step": 640
+    },
+    {
+      "epoch": 0.8721088435374149,
+      "grad_norm": 0.1092897734952719,
+      "learning_rate": 6.446378876553914e-06,
+      "loss": 1.5463,
+      "step": 641
+    },
+    {
+      "epoch": 0.8734693877551021,
+      "grad_norm": 0.10030800864141241,
+      "learning_rate": 6.4356118912618025e-06,
+      "loss": 1.6678,
+      "step": 642
+    },
+    {
+      "epoch": 0.8748299319727891,
+      "grad_norm": 0.10968908293389731,
+      "learning_rate": 6.424837645849724e-06,
+      "loss": 1.6558,
+      "step": 643
+    },
+    {
+      "epoch": 0.8761904761904762,
+      "grad_norm": 0.10586492564834864,
+      "learning_rate": 6.41405619480477e-06,
+      "loss": 1.6116,
+      "step": 644
+    },
+    {
+      "epoch": 0.8775510204081632,
+      "grad_norm": 0.10414471406609932,
+      "learning_rate": 6.403267592650466e-06,
+      "loss": 1.5987,
+      "step": 645
+    },
+    {
+      "epoch": 0.8789115646258503,
+      "grad_norm": 0.11102655800428667,
+      "learning_rate": 6.39247189394651e-06,
+      "loss": 1.5676,
+      "step": 646
+    },
+    {
+      "epoch": 0.8802721088435375,
+      "grad_norm": 0.1414944802203078,
+      "learning_rate": 6.381669153288485e-06,
+      "loss": 1.5632,
+      "step": 647
+    },
+    {
+      "epoch": 0.8816326530612245,
+      "grad_norm": 0.11327036672893112,
+      "learning_rate": 6.370859425307583e-06,
+      "loss": 1.6175,
+      "step": 648
+    },
+    {
+      "epoch": 0.8829931972789116,
+      "grad_norm": 0.10646525745122841,
+      "learning_rate": 6.360042764670337e-06,
+      "loss": 1.6644,
+      "step": 649
+    },
+    {
+      "epoch": 0.8843537414965986,
+      "grad_norm": 0.13888676328597274,
+      "learning_rate": 6.349219226078338e-06,
+      "loss": 1.707,
+      "step": 650
+    },
+    {
+      "epoch": 0.8857142857142857,
+      "grad_norm": 0.12373888508470542,
+      "learning_rate": 6.3383888642679585e-06,
+      "loss": 1.562,
+      "step": 651
+    },
+    {
+      "epoch": 0.8870748299319728,
+      "grad_norm": 0.1150370310262787,
+      "learning_rate": 6.327551734010079e-06,
+      "loss": 1.5981,
+      "step": 652
+    },
+    {
+      "epoch": 0.8884353741496599,
+      "grad_norm": 0.1333929036077325,
+      "learning_rate": 6.3167078901098064e-06,
+      "loss": 1.6216,
+      "step": 653
+    },
+    {
+      "epoch": 0.889795918367347,
+      "grad_norm": 0.09937346566987916,
+      "learning_rate": 6.305857387406204e-06,
+      "loss": 1.7385,
+      "step": 654
+    },
+    {
+      "epoch": 0.891156462585034,
+      "grad_norm": 0.10542763073830541,
+      "learning_rate": 6.295000280772004e-06,
+      "loss": 1.5687,
+      "step": 655
+    },
+    {
+      "epoch": 0.8925170068027211,
+      "grad_norm": 0.11666549211549145,
+      "learning_rate": 6.2841366251133405e-06,
+      "loss": 1.674,
+      "step": 656
+    },
+    {
+      "epoch": 0.8938775510204081,
+      "grad_norm": 0.1010611811906902,
+      "learning_rate": 6.273266475369466e-06,
+      "loss": 1.8506,
+      "step": 657
+    },
+    {
+      "epoch": 0.8952380952380953,
+      "grad_norm": 0.12229341089698531,
+      "learning_rate": 6.262389886512475e-06,
+      "loss": 1.6744,
+      "step": 658
+    },
+    {
+      "epoch": 0.8965986394557823,
+      "grad_norm": 0.1127681300190243,
+      "learning_rate": 6.251506913547021e-06,
+      "loss": 1.5399,
+      "step": 659
+    },
+    {
+      "epoch": 0.8979591836734694,
+      "grad_norm": 0.12397997048973994,
+      "learning_rate": 6.240617611510049e-06,
+      "loss": 1.5651,
+      "step": 660
+    },
+    {
+      "epoch": 0.8993197278911564,
+      "grad_norm": 0.11295244013894742,
+      "learning_rate": 6.229722035470509e-06,
+      "loss": 1.6198,
+      "step": 661
+    },
+    {
+      "epoch": 0.9006802721088435,
+      "grad_norm": 0.1075660765121038,
+      "learning_rate": 6.21882024052908e-06,
+      "loss": 1.6066,
+      "step": 662
+    },
+    {
+      "epoch": 0.9020408163265307,
+      "grad_norm": 0.1202834632226026,
+      "learning_rate": 6.2079122818178885e-06,
+      "loss": 1.7857,
+      "step": 663
+    },
+    {
+      "epoch": 0.9034013605442177,
+      "grad_norm": 0.10694066024862514,
+      "learning_rate": 6.196998214500236e-06,
+      "loss": 1.7661,
+      "step": 664
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 0.10432718122763324,
+      "learning_rate": 6.186078093770312e-06,
+      "loss": 1.5971,
+      "step": 665
+    },
+    {
+      "epoch": 0.9061224489795918,
+      "grad_norm": 0.11861992306540457,
+      "learning_rate": 6.1751519748529235e-06,
+      "loss": 1.5868,
+      "step": 666
+    },
+    {
+      "epoch": 0.9074829931972789,
+      "grad_norm": 0.10675270335750815,
+      "learning_rate": 6.164219913003208e-06,
+      "loss": 1.7003,
+      "step": 667
+    },
+    {
+      "epoch": 0.908843537414966,
+      "grad_norm": 0.11325117184223481,
+      "learning_rate": 6.153281963506359e-06,
+      "loss": 1.5944,
+      "step": 668
+    },
+    {
+      "epoch": 0.9102040816326531,
+      "grad_norm": 0.10366546297175218,
+      "learning_rate": 6.142338181677344e-06,
+      "loss": 1.8128,
+      "step": 669
+    },
+    {
+      "epoch": 0.9115646258503401,
+      "grad_norm": 0.10444277193505327,
+      "learning_rate": 6.131388622860627e-06,
+      "loss": 1.8767,
+      "step": 670
+    },
+    {
+      "epoch": 0.9129251700680272,
+      "grad_norm": 0.10972076945530858,
+      "learning_rate": 6.1204333424298835e-06,
+      "loss": 1.7049,
+      "step": 671
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 0.10751143314986424,
+      "learning_rate": 6.10947239578773e-06,
+      "loss": 1.7047,
+      "step": 672
+    },
+    {
+      "epoch": 0.9156462585034013,
+      "grad_norm": 0.10664892548031508,
+      "learning_rate": 6.098505838365431e-06,
+      "loss": 1.7452,
+      "step": 673
+    },
+    {
+      "epoch": 0.9170068027210885,
+      "grad_norm": 0.10573801367962304,
+      "learning_rate": 6.087533725622631e-06,
+      "loss": 1.6404,
+      "step": 674
+    },
+    {
+      "epoch": 0.9183673469387755,
+      "grad_norm": 0.11119140472745474,
+      "learning_rate": 6.076556113047066e-06,
+      "loss": 1.7246,
+      "step": 675
+    },
+    {
+      "epoch": 0.9197278911564626,
+      "grad_norm": 0.09707326009557425,
+      "learning_rate": 6.065573056154289e-06,
+      "loss": 1.6736,
+      "step": 676
+    },
+    {
+      "epoch": 0.9210884353741496,
+      "grad_norm": 0.14286264459704873,
+      "learning_rate": 6.05458461048738e-06,
+      "loss": 1.7604,
+      "step": 677
+    },
+    {
+      "epoch": 0.9224489795918367,
+      "grad_norm": 0.10186789751886853,
+      "learning_rate": 6.043590831616677e-06,
+      "loss": 1.52,
+      "step": 678
+    },
+    {
+      "epoch": 0.9238095238095239,
+      "grad_norm": 0.08610978660077238,
+      "learning_rate": 6.032591775139483e-06,
+      "loss": 1.6948,
+      "step": 679
+    },
+    {
+      "epoch": 0.9251700680272109,
+      "grad_norm": 0.14039977263434622,
+      "learning_rate": 6.0215874966797935e-06,
+      "loss": 1.7652,
+      "step": 680
+    },
+    {
+      "epoch": 0.926530612244898,
+      "grad_norm": 0.09777556431038441,
+      "learning_rate": 6.0105780518880156e-06,
+      "loss": 1.5695,
+      "step": 681
+    },
+    {
+      "epoch": 0.927891156462585,
+      "grad_norm": 0.10076506193217513,
+      "learning_rate": 5.999563496440678e-06,
+      "loss": 1.5797,
+      "step": 682
+    },
+    {
+      "epoch": 0.9292517006802721,
+      "grad_norm": 0.13645736281296353,
+      "learning_rate": 5.988543886040157e-06,
+      "loss": 1.6124,
+      "step": 683
+    },
+    {
+      "epoch": 0.9306122448979591,
+      "grad_norm": 0.18938296093195647,
+      "learning_rate": 5.977519276414393e-06,
+      "loss": 1.7377,
+      "step": 684
+    },
+    {
+      "epoch": 0.9319727891156463,
+      "grad_norm": 0.11671931887431021,
+      "learning_rate": 5.966489723316609e-06,
+      "loss": 1.57,
+      "step": 685
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 0.13192750753599222,
+      "learning_rate": 5.955455282525027e-06,
+      "loss": 1.5089,
+      "step": 686
+    },
+    {
+      "epoch": 0.9346938775510204,
+      "grad_norm": 0.1279138939941876,
+      "learning_rate": 5.944416009842585e-06,
+      "loss": 1.4862,
+      "step": 687
+    },
+    {
+      "epoch": 0.9360544217687075,
+      "grad_norm": 0.1368328421716675,
+      "learning_rate": 5.933371961096661e-06,
+      "loss": 1.7591,
+      "step": 688
+    },
+    {
+      "epoch": 0.9374149659863945,
+      "grad_norm": 0.09459226556926907,
+      "learning_rate": 5.92232319213878e-06,
+      "loss": 1.7315,
+      "step": 689
+    },
+    {
+      "epoch": 0.9387755102040817,
+      "grad_norm": 0.0944762256396259,
+      "learning_rate": 5.9112697588443456e-06,
+      "loss": 1.6664,
+      "step": 690
+    },
+    {
+      "epoch": 0.9401360544217687,
+      "grad_norm": 0.11165381365887841,
+      "learning_rate": 5.900211717112343e-06,
+      "loss": 1.512,
+      "step": 691
+    },
+    {
+      "epoch": 0.9414965986394558,
+      "grad_norm": 0.10026925859750246,
+      "learning_rate": 5.889149122865067e-06,
+      "loss": 1.8164,
+      "step": 692
+    },
+    {
+      "epoch": 0.9428571428571428,
+      "grad_norm": 0.14168969817408006,
+      "learning_rate": 5.8780820320478325e-06,
+      "loss": 1.7176,
+      "step": 693
+    },
+    {
+      "epoch": 0.9442176870748299,
+      "grad_norm": 0.11535232171729691,
+      "learning_rate": 5.867010500628698e-06,
+      "loss": 1.6684,
+      "step": 694
+    },
+    {
+      "epoch": 0.9455782312925171,
+      "grad_norm": 0.10755021118232244,
+      "learning_rate": 5.855934584598175e-06,
+      "loss": 1.7584,
+      "step": 695
+    },
+    {
+      "epoch": 0.9469387755102041,
+      "grad_norm": 0.11779529743978591,
+      "learning_rate": 5.844854339968952e-06,
+      "loss": 1.6906,
+      "step": 696
+    },
+    {
+      "epoch": 0.9482993197278912,
+      "grad_norm": 0.11394418134579978,
+      "learning_rate": 5.8337698227756035e-06,
+      "loss": 1.6403,
+      "step": 697
+    },
+    {
+      "epoch": 0.9496598639455782,
+      "grad_norm": 0.1136124987259348,
+      "learning_rate": 5.822681089074315e-06,
+      "loss": 1.5563,
+      "step": 698
+    },
+    {
+      "epoch": 0.9510204081632653,
+      "grad_norm": 0.10296990279179531,
+      "learning_rate": 5.811588194942593e-06,
+      "loss": 1.7407,
+      "step": 699
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.11976200431163776,
+      "learning_rate": 5.800491196478989e-06,
+      "loss": 1.4828,
+      "step": 700
+    },
+    {
+      "epoch": 0.9537414965986395,
+      "grad_norm": 0.11272704179233839,
+      "learning_rate": 5.789390149802802e-06,
+      "loss": 1.602,
+      "step": 701
+    },
+    {
+      "epoch": 0.9551020408163265,
+      "grad_norm": 0.23740154863087562,
+      "learning_rate": 5.778285111053812e-06,
+      "loss": 1.6265,
+      "step": 702
+    },
+    {
+      "epoch": 0.9564625850340136,
+      "grad_norm": 0.11973657664057448,
+      "learning_rate": 5.767176136391982e-06,
+      "loss": 1.5886,
+      "step": 703
+    },
+    {
+      "epoch": 0.9578231292517007,
+      "grad_norm": 0.11604087204409168,
+      "learning_rate": 5.756063281997183e-06,
+      "loss": 1.6891,
+      "step": 704
+    },
+    {
+      "epoch": 0.9591836734693877,
+      "grad_norm": 0.10628214364332488,
+      "learning_rate": 5.744946604068904e-06,
+      "loss": 1.6309,
+      "step": 705
+    },
+    {
+      "epoch": 0.9605442176870749,
+      "grad_norm": 0.09908640691690514,
+      "learning_rate": 5.733826158825973e-06,
+      "loss": 1.6741,
+      "step": 706
+    },
+    {
+      "epoch": 0.9619047619047619,
+      "grad_norm": 0.10347727809388245,
+      "learning_rate": 5.722702002506264e-06,
+      "loss": 1.6104,
+      "step": 707
+    },
+    {
+      "epoch": 0.963265306122449,
+      "grad_norm": 0.10424856238627535,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 1.6592,
+      "step": 708
+    },
+    {
+      "epoch": 0.964625850340136,
+      "grad_norm": 0.09391458584535348,
+      "learning_rate": 5.700442781681588e-06,
+      "loss": 1.7451,
+      "step": 709
+    },
+    {
+      "epoch": 0.9659863945578231,
+      "grad_norm": 0.1114698322409562,
+      "learning_rate": 5.689307829745074e-06,
+      "loss": 1.5695,
+      "step": 710
+    },
+    {
+      "epoch": 0.9673469387755103,
+      "grad_norm": 0.1263756308873154,
+      "learning_rate": 5.678169391868128e-06,
+      "loss": 1.7918,
+      "step": 711
+    },
+    {
+      "epoch": 0.9687074829931973,
+      "grad_norm": 0.1068286657604504,
+      "learning_rate": 5.6670275243796194e-06,
+      "loss": 1.6695,
+      "step": 712
+    },
+    {
+      "epoch": 0.9700680272108844,
+      "grad_norm": 0.08766062542171628,
+      "learning_rate": 5.65588228362576e-06,
+      "loss": 1.8529,
+      "step": 713
+    },
+    {
+      "epoch": 0.9714285714285714,
+      "grad_norm": 0.10061266375776706,
+      "learning_rate": 5.6447337259698245e-06,
+      "loss": 1.8285,
+      "step": 714
+    },
+    {
+      "epoch": 0.9727891156462585,
+      "grad_norm": 0.12075281384282141,
+      "learning_rate": 5.633581907791858e-06,
+      "loss": 1.7784,
+      "step": 715
+    },
+    {
+      "epoch": 0.9741496598639455,
+      "grad_norm": 0.1332470465992645,
+      "learning_rate": 5.6224268854884e-06,
+      "loss": 1.675,
+      "step": 716
+    },
+    {
+      "epoch": 0.9755102040816327,
+      "grad_norm": 0.1283354676957953,
+      "learning_rate": 5.611268715472187e-06,
+      "loss": 1.4725,
+      "step": 717
+    },
+    {
+      "epoch": 0.9768707482993197,
+      "grad_norm": 0.16904870188773163,
+      "learning_rate": 5.600107454171879e-06,
+      "loss": 1.6237,
+      "step": 718
+    },
+    {
+      "epoch": 0.9782312925170068,
+      "grad_norm": 0.1259321663471572,
+      "learning_rate": 5.5889431580317655e-06,
+      "loss": 1.663,
+      "step": 719
+    },
+    {
+      "epoch": 0.9795918367346939,
+      "grad_norm": 0.10967294130093679,
+      "learning_rate": 5.577775883511489e-06,
+      "loss": 1.6294,
+      "step": 720
+    },
+    {
+      "epoch": 0.9809523809523809,
+      "grad_norm": 0.17546450235130967,
+      "learning_rate": 5.566605687085749e-06,
+      "loss": 1.5841,
+      "step": 721
+    },
+    {
+      "epoch": 0.9823129251700681,
+      "grad_norm": 0.12265102828756684,
+      "learning_rate": 5.555432625244024e-06,
+      "loss": 1.4919,
+      "step": 722
+    },
+    {
+      "epoch": 0.9836734693877551,
+      "grad_norm": 0.10886604673890071,
+      "learning_rate": 5.5442567544902805e-06,
+      "loss": 1.6385,
+      "step": 723
+    },
+    {
+      "epoch": 0.9850340136054422,
+      "grad_norm": 0.11421831431853076,
+      "learning_rate": 5.533078131342695e-06,
+      "loss": 1.6341,
+      "step": 724
+    },
+    {
+      "epoch": 0.9863945578231292,
+      "grad_norm": 0.09054819655341287,
+      "learning_rate": 5.5218968123333594e-06,
+      "loss": 1.624,
+      "step": 725
+    },
+    {
+      "epoch": 0.9877551020408163,
+      "grad_norm": 0.10971827184516575,
+      "learning_rate": 5.510712854008001e-06,
+      "loss": 1.5447,
+      "step": 726
+    },
+    {
+      "epoch": 0.9891156462585035,
+      "grad_norm": 0.12112012725777507,
+      "learning_rate": 5.499526312925693e-06,
+      "loss": 1.7353,
+      "step": 727
+    },
+    {
+      "epoch": 0.9904761904761905,
+      "grad_norm": 0.10541432689931088,
+      "learning_rate": 5.488337245658569e-06,
+      "loss": 1.6583,
+      "step": 728
+    },
+    {
+      "epoch": 0.9918367346938776,
+      "grad_norm": 0.14736400599380917,
+      "learning_rate": 5.477145708791543e-06,
+      "loss": 1.6641,
+      "step": 729
+    },
+    {
+      "epoch": 0.9931972789115646,
+      "grad_norm": 0.10683105881710352,
+      "learning_rate": 5.4659517589220135e-06,
+      "loss": 1.4082,
+      "step": 730
+    },
+    {
+      "epoch": 0.9945578231292517,
+      "grad_norm": 0.1212803888010618,
+      "learning_rate": 5.454755452659583e-06,
+      "loss": 1.7298,
+      "step": 731
+    },
+    {
+      "epoch": 0.9959183673469387,
+      "grad_norm": 0.114898199928806,
+      "learning_rate": 5.443556846625773e-06,
+      "loss": 1.6922,
+      "step": 732
+    },
+    {
+      "epoch": 0.9972789115646259,
+      "grad_norm": 0.17977169651896993,
+      "learning_rate": 5.432355997453729e-06,
+      "loss": 1.6933,
+      "step": 733
+    },
+    {
+      "epoch": 0.998639455782313,
+      "grad_norm": 0.13478171843239625,
+      "learning_rate": 5.42115296178795e-06,
+      "loss": 1.758,
+      "step": 734
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.10983192009638694,
+      "learning_rate": 5.409947796283982e-06,
+      "loss": 1.6745,
+      "step": 735
+    },
+    {
+      "epoch": 1.0013605442176872,
+      "grad_norm": 0.15728756550496747,
+      "learning_rate": 5.398740557608151e-06,
+      "loss": 1.5976,
+      "step": 736
+    },
+    {
+      "epoch": 1.0013605442176872,
+      "eval_loss": 1.6908553838729858,
+      "eval_runtime": 76.8223,
+      "eval_samples_per_second": 53.018,
+      "eval_steps_per_second": 6.639,
+      "step": 736
+    },
+    {
+      "epoch": 1.002721088435374,
+      "grad_norm": 0.10538368349304568,
+      "learning_rate": 5.38753130243726e-06,
+      "loss": 1.6615,
+      "step": 737
+    },
+    {
+      "epoch": 1.0040816326530613,
+      "grad_norm": 0.11186773529092461,
+      "learning_rate": 5.376320087458316e-06,
+      "loss": 1.686,
+      "step": 738
+    },
+    {
+      "epoch": 1.0054421768707482,
+      "grad_norm": 0.09819212487389627,
+      "learning_rate": 5.365106969368235e-06,
+      "loss": 1.6144,
+      "step": 739
+    },
+    {
+      "epoch": 1.0068027210884354,
+      "grad_norm": 0.10858676538086111,
+      "learning_rate": 5.353892004873554e-06,
+      "loss": 1.7423,
+      "step": 740
+    },
+    {
+      "epoch": 1.0081632653061225,
+      "grad_norm": 0.11968981856525805,
+      "learning_rate": 5.34267525069015e-06,
+      "loss": 1.6532,
+      "step": 741
+    },
+    {
+      "epoch": 1.0095238095238095,
+      "grad_norm": 0.09936611311145167,
+      "learning_rate": 5.331456763542954e-06,
+      "loss": 1.8078,
+      "step": 742
+    },
+    {
+      "epoch": 1.0108843537414967,
+      "grad_norm": 0.10022311903878614,
+      "learning_rate": 5.3202366001656535e-06,
+      "loss": 1.5739,
+      "step": 743
+    },
+    {
+      "epoch": 1.0122448979591836,
+      "grad_norm": 0.4103772585061572,
+      "learning_rate": 5.309014817300422e-06,
+      "loss": 1.6617,
+      "step": 744
+    },
+    {
+      "epoch": 1.0136054421768708,
+      "grad_norm": 0.10960071410597617,
+      "learning_rate": 5.297791471697614e-06,
+      "loss": 1.5742,
+      "step": 745
+    },
+    {
+      "epoch": 1.014965986394558,
+      "grad_norm": 0.1310811762519516,
+      "learning_rate": 5.286566620115493e-06,
+      "loss": 1.7022,
+      "step": 746
+    },
+    {
+      "epoch": 1.0163265306122449,
+      "grad_norm": 0.15568202046790616,
+      "learning_rate": 5.2753403193199374e-06,
+      "loss": 1.592,
+      "step": 747
+    },
+    {
+      "epoch": 1.017687074829932,
+      "grad_norm": 0.11383339823280172,
+      "learning_rate": 5.264112626084153e-06,
+      "loss": 1.6331,
+      "step": 748
+    },
+    {
+      "epoch": 1.019047619047619,
+      "grad_norm": 0.11927282820236593,
+      "learning_rate": 5.2528835971883876e-06,
+      "loss": 1.7091,
+      "step": 749
+    },
+    {
+      "epoch": 1.0204081632653061,
+      "grad_norm": 0.11606987091212463,
+      "learning_rate": 5.241653289419647e-06,
+      "loss": 1.8403,
+      "step": 750
+    },
+    {
+      "epoch": 1.021768707482993,
+      "grad_norm": 0.16512841788961088,
+      "learning_rate": 5.230421759571398e-06,
+      "loss": 1.785,
+      "step": 751
+    },
+    {
+      "epoch": 1.0231292517006803,
+      "grad_norm": 0.12135979152340173,
+      "learning_rate": 5.219189064443296e-06,
+      "loss": 1.5237,
+      "step": 752
+    },
+    {
+      "epoch": 1.0244897959183674,
+      "grad_norm": 0.12001388963231435,
+      "learning_rate": 5.207955260840879e-06,
+      "loss": 1.6265,
+      "step": 753
+    },
+    {
+      "epoch": 1.0258503401360544,
+      "grad_norm": 0.14241973320076035,
+      "learning_rate": 5.1967204055753e-06,
+      "loss": 1.6843,
+      "step": 754
+    },
+    {
+      "epoch": 1.0272108843537415,
+      "grad_norm": 0.3503178069734055,
+      "learning_rate": 5.185484555463026e-06,
+      "loss": 1.8022,
+      "step": 755
+    },
+    {
+      "epoch": 1.0285714285714285,
+      "grad_norm": 0.10584406401171531,
+      "learning_rate": 5.17424776732556e-06,
+      "loss": 1.6713,
+      "step": 756
+    },
+    {
+      "epoch": 1.0299319727891156,
+      "grad_norm": 0.11525477562466327,
+      "learning_rate": 5.163010097989138e-06,
+      "loss": 1.73,
+      "step": 757
+    },
+    {
+      "epoch": 1.0312925170068028,
+      "grad_norm": 0.13500765537475423,
+      "learning_rate": 5.151771604284465e-06,
+      "loss": 1.405,
+      "step": 758
+    },
+    {
+      "epoch": 1.0326530612244897,
+      "grad_norm": 0.18732245961092447,
+      "learning_rate": 5.140532343046406e-06,
+      "loss": 1.5587,
+      "step": 759
+    },
+    {
+      "epoch": 1.034013605442177,
+      "grad_norm": 0.10827851940146976,
+      "learning_rate": 5.129292371113712e-06,
+      "loss": 1.7328,
+      "step": 760
+    },
+    {
+      "epoch": 1.0353741496598639,
+      "grad_norm": 0.11027741963001209,
+      "learning_rate": 5.118051745328725e-06,
+      "loss": 1.6382,
+      "step": 761
+    },
+    {
+      "epoch": 1.036734693877551,
+      "grad_norm": 0.09919441836217119,
+      "learning_rate": 5.1068105225370975e-06,
+      "loss": 1.6855,
+      "step": 762
+    },
+    {
+      "epoch": 1.0380952380952382,
+      "grad_norm": 0.13830797432010328,
+      "learning_rate": 5.095568759587497e-06,
+      "loss": 1.7411,
+      "step": 763
+    },
+    {
+      "epoch": 1.0394557823129251,
+      "grad_norm": 0.0972282280148536,
+      "learning_rate": 5.084326513331328e-06,
+      "loss": 1.621,
+      "step": 764
+    },
+    {
+      "epoch": 1.0408163265306123,
+      "grad_norm": 0.10578341535327669,
+      "learning_rate": 5.0730838406224324e-06,
+      "loss": 1.6273,
+      "step": 765
+    },
+    {
+      "epoch": 1.0421768707482992,
+      "grad_norm": 0.10463652066026084,
+      "learning_rate": 5.061840798316815e-06,
+      "loss": 1.725,
+      "step": 766
+    },
+    {
+      "epoch": 1.0435374149659864,
+      "grad_norm": 0.11942338757679548,
+      "learning_rate": 5.0505974432723445e-06,
+      "loss": 1.5898,
+      "step": 767
+    },
+    {
+      "epoch": 1.0448979591836736,
+      "grad_norm": 0.13689962498964267,
+      "learning_rate": 5.039353832348477e-06,
+      "loss": 1.5068,
+      "step": 768
+    },
+    {
+      "epoch": 1.0462585034013605,
+      "grad_norm": 0.13678139029155267,
+      "learning_rate": 5.028110022405955e-06,
+      "loss": 1.7158,
+      "step": 769
+    },
+    {
+      "epoch": 1.0476190476190477,
+      "grad_norm": 0.1000363922457913,
+      "learning_rate": 5.0168660703065354e-06,
+      "loss": 1.741,
+      "step": 770
+    },
+    {
+      "epoch": 1.0489795918367346,
+      "grad_norm": 0.1188815368304704,
+      "learning_rate": 5.005622032912687e-06,
+      "loss": 1.6623,
+      "step": 771
+    },
+    {
+      "epoch": 1.0503401360544218,
+      "grad_norm": 0.12019243813821356,
+      "learning_rate": 4.994377967087316e-06,
+      "loss": 1.5774,
+      "step": 772
+    },
+    {
+      "epoch": 1.051700680272109,
+      "grad_norm": 0.12450888223025358,
+      "learning_rate": 4.983133929693467e-06,
+      "loss": 1.5663,
+      "step": 773
+    },
+    {
+      "epoch": 1.0530612244897959,
+      "grad_norm": 0.10816898093198718,
+      "learning_rate": 4.971889977594048e-06,
+      "loss": 1.6911,
+      "step": 774
+    },
+    {
+      "epoch": 1.054421768707483,
+      "grad_norm": 1.729544315460825,
+      "learning_rate": 4.960646167651524e-06,
+      "loss": 1.7524,
+      "step": 775
+    },
+    {
+      "epoch": 1.05578231292517,
+      "grad_norm": 0.1133964092151596,
+      "learning_rate": 4.949402556727655e-06,
+      "loss": 1.7612,
+      "step": 776
+    },
+    {
+      "epoch": 1.0571428571428572,
+      "grad_norm": 0.12041549603751739,
+      "learning_rate": 4.9381592016831856e-06,
+      "loss": 1.5116,
+      "step": 777
+    },
+    {
+      "epoch": 1.0585034013605443,
+      "grad_norm": 0.11576436512487191,
+      "learning_rate": 4.9269161593775675e-06,
+      "loss": 1.5329,
+      "step": 778
+    },
+    {
+      "epoch": 1.0598639455782313,
+      "grad_norm": 0.12679277289105279,
+      "learning_rate": 4.915673486668673e-06,
+      "loss": 1.5506,
+      "step": 779
+    },
+    {
+      "epoch": 1.0612244897959184,
+      "grad_norm": 0.11679477613827106,
+      "learning_rate": 4.904431240412503e-06,
+      "loss": 1.5008,
+      "step": 780
+    },
+    {
+      "epoch": 1.0625850340136054,
+      "grad_norm": 0.10065169200286987,
+      "learning_rate": 4.893189477462905e-06,
+      "loss": 1.7685,
+      "step": 781
+    },
+    {
+      "epoch": 1.0639455782312925,
+      "grad_norm": 0.10287181892423168,
+      "learning_rate": 4.881948254671277e-06,
+      "loss": 1.6379,
+      "step": 782
+    },
+    {
+      "epoch": 1.0653061224489795,
+      "grad_norm": 0.10896624508987876,
+      "learning_rate": 4.870707628886291e-06,
+      "loss": 1.5234,
+      "step": 783
+    },
+    {
+      "epoch": 1.0666666666666667,
+      "grad_norm": 0.10915491188092791,
+      "learning_rate": 4.859467656953596e-06,
+      "loss": 1.5865,
+      "step": 784
+    },
+    {
+      "epoch": 1.0680272108843538,
+      "grad_norm": 0.1275517896298314,
+      "learning_rate": 4.8482283957155355e-06,
+      "loss": 1.7069,
+      "step": 785
+    },
+    {
+      "epoch": 1.0693877551020408,
+      "grad_norm": 0.10862109015230759,
+      "learning_rate": 4.836989902010863e-06,
+      "loss": 1.7682,
+      "step": 786
+    },
+    {
+      "epoch": 1.070748299319728,
+      "grad_norm": 0.09805361474836685,
+      "learning_rate": 4.825752232674441e-06,
+      "loss": 1.6228,
+      "step": 787
+    },
+    {
+      "epoch": 1.0721088435374149,
+      "grad_norm": 0.12481176224206536,
+      "learning_rate": 4.814515444536975e-06,
+      "loss": 1.5027,
+      "step": 788
+    },
+    {
+      "epoch": 1.073469387755102,
+      "grad_norm": 0.11390040747547601,
+      "learning_rate": 4.8032795944247e-06,
+      "loss": 1.5168,
+      "step": 789
+    },
+    {
+      "epoch": 1.0748299319727892,
+      "grad_norm": 0.13458744184978433,
+      "learning_rate": 4.792044739159124e-06,
+      "loss": 1.5188,
+      "step": 790
+    },
+    {
+      "epoch": 1.0761904761904761,
+      "grad_norm": 0.12294480568667183,
+      "learning_rate": 4.780810935556707e-06,
+      "loss": 1.3946,
+      "step": 791
+    },
+    {
+      "epoch": 1.0775510204081633,
+      "grad_norm": 0.13732274998725952,
+      "learning_rate": 4.7695782404286045e-06,
+      "loss": 1.6201,
+      "step": 792
+    },
+    {
+      "epoch": 1.0789115646258503,
+      "grad_norm": 0.09919081899157949,
+      "learning_rate": 4.758346710580355e-06,
+      "loss": 1.7815,
+      "step": 793
+    },
+    {
+      "epoch": 1.0802721088435374,
+      "grad_norm": 0.10000541950613748,
+      "learning_rate": 4.747116402811612e-06,
+      "loss": 1.8098,
+      "step": 794
+    },
+    {
+      "epoch": 1.0816326530612246,
+      "grad_norm": 0.11420683474567415,
+      "learning_rate": 4.735887373915848e-06,
+      "loss": 1.4835,
+      "step": 795
+    },
+    {
+      "epoch": 1.0829931972789115,
+      "grad_norm": 0.15140465279516127,
+      "learning_rate": 4.724659680680063e-06,
+      "loss": 1.5029,
+      "step": 796
+    },
+    {
+      "epoch": 1.0843537414965987,
+      "grad_norm": 0.11208117898027617,
+      "learning_rate": 4.713433379884508e-06,
+      "loss": 1.7194,
+      "step": 797
+    },
+    {
+      "epoch": 1.0857142857142856,
+      "grad_norm": 0.16855436651508154,
+      "learning_rate": 4.7022085283023875e-06,
+      "loss": 1.7491,
+      "step": 798
+    },
+    {
+      "epoch": 1.0870748299319728,
+      "grad_norm": 0.11595366549169164,
+      "learning_rate": 4.690985182699581e-06,
+      "loss": 1.6313,
+      "step": 799
+    },
+    {
+      "epoch": 1.08843537414966,
+      "grad_norm": 0.09503431401550515,
+      "learning_rate": 4.679763399834347e-06,
+      "loss": 1.5984,
+      "step": 800
+    },
+    {
+      "epoch": 1.089795918367347,
+      "grad_norm": 0.10896290135710643,
+      "learning_rate": 4.668543236457049e-06,
+      "loss": 1.7379,
+      "step": 801
+    },
+    {
+      "epoch": 1.091156462585034,
+      "grad_norm": 0.11312340154871599,
+      "learning_rate": 4.657324749309851e-06,
+      "loss": 1.7817,
+      "step": 802
+    },
+    {
+      "epoch": 1.092517006802721,
+      "grad_norm": 0.11398862476288807,
+      "learning_rate": 4.646107995126447e-06,
+      "loss": 1.6113,
+      "step": 803
+    },
+    {
+      "epoch": 1.0938775510204082,
+      "grad_norm": 0.11528891846651022,
+      "learning_rate": 4.634893030631767e-06,
+      "loss": 1.6745,
+      "step": 804
+    },
+    {
+      "epoch": 1.0952380952380953,
+      "grad_norm": 0.11540329784321814,
+      "learning_rate": 4.623679912541683e-06,
+      "loss": 1.6443,
+      "step": 805
+    },
+    {
+      "epoch": 1.0965986394557823,
+      "grad_norm": 0.11457602895195093,
+      "learning_rate": 4.612468697562741e-06,
+      "loss": 1.5109,
+      "step": 806
+    },
+    {
+      "epoch": 1.0979591836734695,
+      "grad_norm": 0.108466968677861,
+      "learning_rate": 4.6012594423918505e-06,
+      "loss": 1.6285,
+      "step": 807
+    },
+    {
+      "epoch": 1.0993197278911564,
+      "grad_norm": 0.11392368141416775,
+      "learning_rate": 4.5900522037160205e-06,
+      "loss": 1.524,
+      "step": 808
+    },
+    {
+      "epoch": 1.1006802721088436,
+      "grad_norm": 0.10657614384924384,
+      "learning_rate": 4.578847038212052e-06,
+      "loss": 1.7741,
+      "step": 809
+    },
+    {
+      "epoch": 1.1020408163265305,
+      "grad_norm": 0.11158460621214396,
+      "learning_rate": 4.567644002546273e-06,
+      "loss": 1.6648,
+      "step": 810
+    },
+    {
+      "epoch": 1.1034013605442177,
+      "grad_norm": 0.11249423664710494,
+      "learning_rate": 4.556443153374229e-06,
+      "loss": 1.5484,
+      "step": 811
+    },
+    {
+      "epoch": 1.1047619047619048,
+      "grad_norm": 0.18597431248705018,
+      "learning_rate": 4.5452445473404175e-06,
+      "loss": 1.4591,
+      "step": 812
+    },
+    {
+      "epoch": 1.1061224489795918,
+      "grad_norm": 0.1146701110937394,
+      "learning_rate": 4.534048241077987e-06,
+      "loss": 1.7267,
+      "step": 813
+    },
+    {
+      "epoch": 1.107482993197279,
+      "grad_norm": 0.10820118015434177,
+      "learning_rate": 4.522854291208458e-06,
+      "loss": 1.7739,
+      "step": 814
+    },
+    {
+      "epoch": 1.1088435374149659,
+      "grad_norm": 0.09878497379061389,
+      "learning_rate": 4.511662754341433e-06,
+      "loss": 1.6488,
+      "step": 815
+    },
+    {
+      "epoch": 1.110204081632653,
+      "grad_norm": 0.12970442140488536,
+      "learning_rate": 4.50047368707431e-06,
+      "loss": 1.4903,
+      "step": 816
+    },
+    {
+      "epoch": 1.1115646258503402,
+      "grad_norm": 0.10382954289821261,
+      "learning_rate": 4.489287145992002e-06,
+      "loss": 1.8014,
+      "step": 817
+    },
+    {
+      "epoch": 1.1129251700680272,
+      "grad_norm": 0.12667728251125698,
+      "learning_rate": 4.478103187666642e-06,
+      "loss": 1.6478,
+      "step": 818
+    },
+    {
+      "epoch": 1.1142857142857143,
+      "grad_norm": 0.13500564587776376,
+      "learning_rate": 4.4669218686573065e-06,
+      "loss": 1.6239,
+      "step": 819
+    },
+    {
+      "epoch": 1.1156462585034013,
+      "grad_norm": 0.1244506846499162,
+      "learning_rate": 4.45574324550972e-06,
+      "loss": 1.718,
+      "step": 820
+    },
+    {
+      "epoch": 1.1170068027210884,
+      "grad_norm": 0.11873853458664813,
+      "learning_rate": 4.444567374755978e-06,
+      "loss": 1.7092,
+      "step": 821
+    },
+    {
+      "epoch": 1.1183673469387756,
+      "grad_norm": 0.12208418986043022,
+      "learning_rate": 4.433394312914253e-06,
+      "loss": 1.7201,
+      "step": 822
+    },
+    {
+      "epoch": 1.1197278911564625,
+      "grad_norm": 0.13523682008047516,
+      "learning_rate": 4.4222241164885114e-06,
+      "loss": 1.6248,
+      "step": 823
+    },
+    {
+      "epoch": 1.1210884353741497,
+      "grad_norm": 0.11246329856395026,
+      "learning_rate": 4.411056841968236e-06,
+      "loss": 1.5008,
+      "step": 824
+    },
+    {
+      "epoch": 1.1224489795918366,
+      "grad_norm": 0.12873507179822666,
+      "learning_rate": 4.3998925458281225e-06,
+      "loss": 1.6518,
+      "step": 825
+    },
+    {
+      "epoch": 1.1238095238095238,
+      "grad_norm": 0.13059910987895082,
+      "learning_rate": 4.388731284527816e-06,
+      "loss": 1.6547,
+      "step": 826
+    },
+    {
+      "epoch": 1.125170068027211,
+      "grad_norm": 0.13485735751834116,
+      "learning_rate": 4.377573114511602e-06,
+      "loss": 1.5989,
+      "step": 827
+    },
+    {
+      "epoch": 1.126530612244898,
+      "grad_norm": 0.12101663226880303,
+      "learning_rate": 4.366418092208144e-06,
+      "loss": 1.6142,
+      "step": 828
+    },
+    {
+      "epoch": 1.127891156462585,
+      "grad_norm": 0.11607507412451191,
+      "learning_rate": 4.355266274030177e-06,
+      "loss": 1.6316,
+      "step": 829
+    },
+    {
+      "epoch": 1.129251700680272,
+      "grad_norm": 0.10911266136424244,
+      "learning_rate": 4.344117716374241e-06,
+      "loss": 1.5342,
+      "step": 830
+    },
+    {
+      "epoch": 1.1306122448979592,
+      "grad_norm": 0.12731437170330562,
+      "learning_rate": 4.332972475620381e-06,
+      "loss": 1.6973,
+      "step": 831
+    },
+    {
+      "epoch": 1.1319727891156464,
+      "grad_norm": 0.09762084699859862,
+      "learning_rate": 4.321830608131872e-06,
+      "loss": 1.6633,
+      "step": 832
+    },
+    {
+      "epoch": 1.1333333333333333,
+      "grad_norm": 0.12978891884133562,
+      "learning_rate": 4.310692170254927e-06,
+      "loss": 1.6098,
+      "step": 833
+    },
+    {
+      "epoch": 1.1346938775510205,
+      "grad_norm": 0.12415908074295716,
+      "learning_rate": 4.299557218318413e-06,
+      "loss": 1.6307,
+      "step": 834
+    },
+    {
+      "epoch": 1.1360544217687074,
+      "grad_norm": 0.2173948925361782,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 1.5907,
+      "step": 835
+    },
+    {
+      "epoch": 1.1374149659863946,
+      "grad_norm": 0.11912082507711948,
+      "learning_rate": 4.277297997493737e-06,
+      "loss": 1.5837,
+      "step": 836
+    },
+    {
+      "epoch": 1.1387755102040815,
+      "grad_norm": 0.19450908692643373,
+      "learning_rate": 4.266173841174031e-06,
+      "loss": 1.7324,
+      "step": 837
+    },
+    {
+      "epoch": 1.1401360544217687,
+      "grad_norm": 0.11032655579441553,
+      "learning_rate": 4.255053395931097e-06,
+      "loss": 1.7134,
+      "step": 838
+    },
+    {
+      "epoch": 1.1414965986394559,
+      "grad_norm": 0.10178083734022528,
+      "learning_rate": 4.243936718002818e-06,
+      "loss": 1.6472,
+      "step": 839
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.10102775840788815,
+      "learning_rate": 4.23282386360802e-06,
+      "loss": 1.6893,
+      "step": 840
+    },
+    {
+      "epoch": 1.14421768707483,
+      "grad_norm": 0.13340734842779356,
+      "learning_rate": 4.22171488894619e-06,
+      "loss": 1.582,
+      "step": 841
+    },
+    {
+      "epoch": 1.1455782312925171,
+      "grad_norm": 0.39827971211004304,
+      "learning_rate": 4.2106098501972e-06,
+      "loss": 1.5918,
+      "step": 842
+    },
+    {
+      "epoch": 1.146938775510204,
+      "grad_norm": 0.12809647826224843,
+      "learning_rate": 4.1995088035210126e-06,
+      "loss": 1.6786,
+      "step": 843
+    },
+    {
+      "epoch": 1.1482993197278912,
+      "grad_norm": 0.11881853753159395,
+      "learning_rate": 4.1884118050574084e-06,
+      "loss": 1.6218,
+      "step": 844
+    },
+    {
+      "epoch": 1.1496598639455782,
+      "grad_norm": 0.13188862048286334,
+      "learning_rate": 4.177318910925686e-06,
+      "loss": 1.6943,
+      "step": 845
+    },
+    {
+      "epoch": 1.1510204081632653,
+      "grad_norm": 0.11973362482121355,
+      "learning_rate": 4.1662301772244e-06,
+      "loss": 1.3518,
+      "step": 846
+    },
+    {
+      "epoch": 1.1523809523809523,
+      "grad_norm": 0.1265246682564999,
+      "learning_rate": 4.15514566003105e-06,
+      "loss": 1.717,
+      "step": 847
+    },
+    {
+      "epoch": 1.1537414965986394,
+      "grad_norm": 0.12163131262108975,
+      "learning_rate": 4.144065415401825e-06,
+      "loss": 1.6591,
+      "step": 848
+    },
+    {
+      "epoch": 1.1551020408163266,
+      "grad_norm": 0.11281334037077397,
+      "learning_rate": 4.132989499371303e-06,
+      "loss": 1.7488,
+      "step": 849
+    },
+    {
+      "epoch": 1.1564625850340136,
+      "grad_norm": 0.116733746797477,
+      "learning_rate": 4.1219179679521675e-06,
+      "loss": 1.4784,
+      "step": 850
+    },
+    {
+      "epoch": 1.1578231292517007,
+      "grad_norm": 0.10824656908525278,
+      "learning_rate": 4.110850877134935e-06,
+      "loss": 1.6377,
+      "step": 851
+    },
+    {
+      "epoch": 1.1591836734693877,
+      "grad_norm": 0.13121005190396953,
+      "learning_rate": 4.099788282887658e-06,
+      "loss": 1.7325,
+      "step": 852
+    },
+    {
+      "epoch": 1.1605442176870748,
+      "grad_norm": 0.10756631934140032,
+      "learning_rate": 4.088730241155657e-06,
+      "loss": 1.7462,
+      "step": 853
+    },
+    {
+      "epoch": 1.161904761904762,
+      "grad_norm": 0.17347249647835658,
+      "learning_rate": 4.077676807861221e-06,
+      "loss": 1.4988,
+      "step": 854
+    },
+    {
+      "epoch": 1.163265306122449,
+      "grad_norm": 0.10923737210751697,
+      "learning_rate": 4.066628038903341e-06,
+      "loss": 1.753,
+      "step": 855
+    },
+    {
+      "epoch": 1.164625850340136,
+      "grad_norm": 0.12401952453107581,
+      "learning_rate": 4.055583990157416e-06,
+      "loss": 1.6061,
+      "step": 856
+    },
+    {
+      "epoch": 1.165986394557823,
+      "grad_norm": 0.16613834484616688,
+      "learning_rate": 4.044544717474974e-06,
+      "loss": 1.6004,
+      "step": 857
+    },
+    {
+      "epoch": 1.1673469387755102,
+      "grad_norm": 0.10830889852190581,
+      "learning_rate": 4.033510276683392e-06,
+      "loss": 1.8508,
+      "step": 858
+    },
+    {
+      "epoch": 1.1687074829931974,
+      "grad_norm": 0.11570096470891132,
+      "learning_rate": 4.022480723585608e-06,
+      "loss": 1.7405,
+      "step": 859
+    },
+    {
+      "epoch": 1.1700680272108843,
+      "grad_norm": 0.13085816560126146,
+      "learning_rate": 4.011456113959845e-06,
+      "loss": 1.6882,
+      "step": 860
+    },
+    {
+      "epoch": 1.1714285714285715,
+      "grad_norm": 0.09457110643687838,
+      "learning_rate": 4.000436503559324e-06,
+      "loss": 1.7382,
+      "step": 861
+    },
+    {
+      "epoch": 1.1727891156462584,
+      "grad_norm": 0.11447572869844899,
+      "learning_rate": 3.989421948111987e-06,
+      "loss": 1.7447,
+      "step": 862
+    },
+    {
+      "epoch": 1.1741496598639456,
+      "grad_norm": 0.10219706136235482,
+      "learning_rate": 3.978412503320207e-06,
+      "loss": 1.6079,
+      "step": 863
+    },
+    {
+      "epoch": 1.1755102040816325,
+      "grad_norm": 0.09697745892488446,
+      "learning_rate": 3.967408224860518e-06,
+      "loss": 1.7658,
+      "step": 864
+    },
+    {
+      "epoch": 1.1768707482993197,
+      "grad_norm": 0.11935090111442251,
+      "learning_rate": 3.956409168383325e-06,
+      "loss": 1.642,
+      "step": 865
+    },
+    {
+      "epoch": 1.1782312925170069,
+      "grad_norm": 0.09593894653017256,
+      "learning_rate": 3.94541538951262e-06,
+      "loss": 1.6837,
+      "step": 866
+    },
+    {
+      "epoch": 1.1795918367346938,
+      "grad_norm": 0.17924060912455603,
+      "learning_rate": 3.934426943845712e-06,
+      "loss": 1.7668,
+      "step": 867
+    },
+    {
+      "epoch": 1.180952380952381,
+      "grad_norm": 0.11737602024296209,
+      "learning_rate": 3.923443886952934e-06,
+      "loss": 1.5348,
+      "step": 868
+    },
+    {
+      "epoch": 1.1823129251700681,
+      "grad_norm": 0.09917281103783915,
+      "learning_rate": 3.912466274377371e-06,
+      "loss": 1.6659,
+      "step": 869
+    },
+    {
+      "epoch": 1.183673469387755,
+      "grad_norm": 0.10815570929741564,
+      "learning_rate": 3.901494161634571e-06,
+      "loss": 1.5029,
+      "step": 870
+    },
+    {
+      "epoch": 1.1850340136054422,
+      "grad_norm": 0.09777863761181164,
+      "learning_rate": 3.890527604212273e-06,
+      "loss": 1.7809,
+      "step": 871
+    },
+    {
+      "epoch": 1.1863945578231292,
+      "grad_norm": 0.12446664048031521,
+      "learning_rate": 3.879566657570118e-06,
+      "loss": 1.6918,
+      "step": 872
+    },
+    {
+      "epoch": 1.1877551020408164,
+      "grad_norm": 0.11809808630659835,
+      "learning_rate": 3.868611377139375e-06,
+      "loss": 1.5293,
+      "step": 873
+    },
+    {
+      "epoch": 1.1891156462585033,
+      "grad_norm": 0.11626193699249888,
+      "learning_rate": 3.857661818322657e-06,
+      "loss": 1.5724,
+      "step": 874
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 0.10675553850073687,
+      "learning_rate": 3.846718036493642e-06,
+      "loss": 1.7026,
+      "step": 875
+    },
+    {
+      "epoch": 1.1918367346938776,
+      "grad_norm": 0.25785039953724753,
+      "learning_rate": 3.835780086996794e-06,
+      "loss": 1.816,
+      "step": 876
+    },
+    {
+      "epoch": 1.1931972789115646,
+      "grad_norm": 0.09919492633891115,
+      "learning_rate": 3.824848025147078e-06,
+      "loss": 1.8633,
+      "step": 877
+    },
+    {
+      "epoch": 1.1945578231292517,
+      "grad_norm": 0.13021628412390385,
+      "learning_rate": 3.81392190622969e-06,
+      "loss": 1.6292,
+      "step": 878
+    },
+    {
+      "epoch": 1.1959183673469387,
+      "grad_norm": 0.1063338899612945,
+      "learning_rate": 3.8030017854997654e-06,
+      "loss": 1.7619,
+      "step": 879
+    },
+    {
+      "epoch": 1.1972789115646258,
+      "grad_norm": 0.1400157862890337,
+      "learning_rate": 3.7920877181821136e-06,
+      "loss": 1.5179,
+      "step": 880
+    },
+    {
+      "epoch": 1.198639455782313,
+      "grad_norm": 0.1386445404000312,
+      "learning_rate": 3.781179759470921e-06,
+      "loss": 1.6624,
+      "step": 881
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.1069092489728771,
+      "learning_rate": 3.7702779645294907e-06,
+      "loss": 1.8113,
+      "step": 882
+    },
+    {
+      "epoch": 1.2013605442176871,
+      "grad_norm": 0.1190549983688599,
+      "learning_rate": 3.759382388489952e-06,
+      "loss": 1.6425,
+      "step": 883
+    },
+    {
+      "epoch": 1.202721088435374,
+      "grad_norm": 0.11079642382956639,
+      "learning_rate": 3.74849308645298e-06,
+      "loss": 1.6609,
+      "step": 884
+    },
+    {
+      "epoch": 1.2040816326530612,
+      "grad_norm": 0.14105617751955826,
+      "learning_rate": 3.7376101134875278e-06,
+      "loss": 1.55,
+      "step": 885
+    },
+    {
+      "epoch": 1.2054421768707484,
+      "grad_norm": 0.14448379899642935,
+      "learning_rate": 3.7267335246305346e-06,
+      "loss": 1.778,
+      "step": 886
+    },
+    {
+      "epoch": 1.2068027210884353,
+      "grad_norm": 0.10874354603342407,
+      "learning_rate": 3.715863374886661e-06,
+      "loss": 1.6611,
+      "step": 887
+    },
+    {
+      "epoch": 1.2081632653061225,
+      "grad_norm": 0.10802721342705796,
+      "learning_rate": 3.7049997192279976e-06,
+      "loss": 1.5966,
+      "step": 888
+    },
+    {
+      "epoch": 1.2095238095238094,
+      "grad_norm": 0.12233010145027158,
+      "learning_rate": 3.6941426125937992e-06,
+      "loss": 1.5311,
+      "step": 889
+    },
+    {
+      "epoch": 1.2108843537414966,
+      "grad_norm": 0.10009553518935768,
+      "learning_rate": 3.6832921098901952e-06,
+      "loss": 1.5145,
+      "step": 890
+    },
+    {
+      "epoch": 1.2122448979591836,
+      "grad_norm": 0.11901267626986793,
+      "learning_rate": 3.6724482659899226e-06,
+      "loss": 1.7466,
+      "step": 891
+    },
+    {
+      "epoch": 1.2136054421768707,
+      "grad_norm": 0.12962797644138863,
+      "learning_rate": 3.661611135732043e-06,
+      "loss": 1.5964,
+      "step": 892
+    },
+    {
+      "epoch": 1.2149659863945579,
+      "grad_norm": 0.10733667423090783,
+      "learning_rate": 3.6507807739216628e-06,
+      "loss": 1.7763,
+      "step": 893
+    },
+    {
+      "epoch": 1.2163265306122448,
+      "grad_norm": 0.11144056418847463,
+      "learning_rate": 3.6399572353296642e-06,
+      "loss": 1.7047,
+      "step": 894
+    },
+    {
+      "epoch": 1.217687074829932,
+      "grad_norm": 0.08875006973132216,
+      "learning_rate": 3.6291405746924186e-06,
+      "loss": 1.7604,
+      "step": 895
+    },
+    {
+      "epoch": 1.2190476190476192,
+      "grad_norm": 0.13024552194401284,
+      "learning_rate": 3.6183308467115174e-06,
+      "loss": 1.6061,
+      "step": 896
+    },
+    {
+      "epoch": 1.220408163265306,
+      "grad_norm": 0.12857972357080566,
+      "learning_rate": 3.6075281060534917e-06,
+      "loss": 1.6149,
+      "step": 897
+    },
+    {
+      "epoch": 1.2217687074829933,
+      "grad_norm": 0.11603998479755155,
+      "learning_rate": 3.5967324073495363e-06,
+      "loss": 1.6111,
+      "step": 898
+    },
+    {
+      "epoch": 1.2231292517006802,
+      "grad_norm": 0.11178507529251955,
+      "learning_rate": 3.585943805195232e-06,
+      "loss": 1.6971,
+      "step": 899
+    },
+    {
+      "epoch": 1.2244897959183674,
+      "grad_norm": 0.14802404367937794,
+      "learning_rate": 3.575162354150276e-06,
+      "loss": 1.7452,
+      "step": 900
+    },
+    {
+      "epoch": 1.2258503401360543,
+      "grad_norm": 0.09565639349194041,
+      "learning_rate": 3.5643881087381983e-06,
+      "loss": 1.7094,
+      "step": 901
+    },
+    {
+      "epoch": 1.2272108843537415,
+      "grad_norm": 0.12259182420540694,
+      "learning_rate": 3.553621123446087e-06,
+      "loss": 1.4945,
+      "step": 902
+    },
+    {
+      "epoch": 1.2285714285714286,
+      "grad_norm": 0.09891155424272953,
+      "learning_rate": 3.542861452724318e-06,
+      "loss": 1.7481,
+      "step": 903
+    },
+    {
+      "epoch": 1.2299319727891156,
+      "grad_norm": 0.12577917367601982,
+      "learning_rate": 3.5321091509862733e-06,
+      "loss": 1.658,
+      "step": 904
+    },
+    {
+      "epoch": 1.2312925170068028,
+      "grad_norm": 0.11110224185718393,
+      "learning_rate": 3.521364272608071e-06,
+      "loss": 1.7805,
+      "step": 905
+    },
+    {
+      "epoch": 1.2326530612244897,
+      "grad_norm": 1.029133559209144,
+      "learning_rate": 3.5106268719282863e-06,
+      "loss": 1.6974,
+      "step": 906
+    },
+    {
+      "epoch": 1.2340136054421769,
+      "grad_norm": 0.11635973055276655,
+      "learning_rate": 3.499897003247682e-06,
+      "loss": 1.6067,
+      "step": 907
+    },
+    {
+      "epoch": 1.235374149659864,
+      "grad_norm": 0.1374434820581917,
+      "learning_rate": 3.489174720828924e-06,
+      "loss": 1.4329,
+      "step": 908
+    },
+    {
+      "epoch": 1.236734693877551,
+      "grad_norm": 0.12005946386037955,
+      "learning_rate": 3.4784600788963197e-06,
+      "loss": 1.6061,
+      "step": 909
+    },
+    {
+      "epoch": 1.2380952380952381,
+      "grad_norm": 0.2566914570489289,
+      "learning_rate": 3.4677531316355343e-06,
+      "loss": 1.6285,
+      "step": 910
+    },
+    {
+      "epoch": 1.239455782312925,
+      "grad_norm": 0.12805153195567712,
+      "learning_rate": 3.4570539331933196e-06,
+      "loss": 1.6518,
+      "step": 911
+    },
+    {
+      "epoch": 1.2408163265306122,
+      "grad_norm": 0.12193109943671782,
+      "learning_rate": 3.4463625376772415e-06,
+      "loss": 1.7769,
+      "step": 912
+    },
+    {
+      "epoch": 1.2421768707482994,
+      "grad_norm": 0.11785291933334519,
+      "learning_rate": 3.4356789991554036e-06,
+      "loss": 1.7037,
+      "step": 913
+    },
+    {
+      "epoch": 1.2435374149659864,
+      "grad_norm": 0.13098314516857928,
+      "learning_rate": 3.425003371656178e-06,
+      "loss": 1.6332,
+      "step": 914
+    },
+    {
+      "epoch": 1.2448979591836735,
+      "grad_norm": 0.11058534722726451,
+      "learning_rate": 3.4143357091679276e-06,
+      "loss": 1.8928,
+      "step": 915
+    },
+    {
+      "epoch": 1.2462585034013605,
+      "grad_norm": 0.1304628400810422,
+      "learning_rate": 3.403676065638735e-06,
+      "loss": 1.5842,
+      "step": 916
+    },
+    {
+      "epoch": 1.2476190476190476,
+      "grad_norm": 0.12141593155806699,
+      "learning_rate": 3.393024494976128e-06,
+      "loss": 1.6872,
+      "step": 917
+    },
+    {
+      "epoch": 1.2489795918367346,
+      "grad_norm": 0.1180901086883869,
+      "learning_rate": 3.3823810510468146e-06,
+      "loss": 1.4999,
+      "step": 918
+    },
+    {
+      "epoch": 1.2503401360544217,
+      "grad_norm": 0.1426231115449843,
+      "learning_rate": 3.3717457876763994e-06,
+      "loss": 1.7262,
+      "step": 919
+    },
+    {
+      "epoch": 1.251700680272109,
+      "grad_norm": 0.11420848866165159,
+      "learning_rate": 3.361118758649116e-06,
+      "loss": 1.617,
+      "step": 920
+    },
+    {
+      "epoch": 1.251700680272109,
+      "eval_loss": 1.6881320476531982,
+      "eval_runtime": 76.6095,
+      "eval_samples_per_second": 53.166,
+      "eval_steps_per_second": 6.657,
+      "step": 920
+    },
+    {
+      "epoch": 1.2530612244897958,
+      "grad_norm": 0.1528479382450589,
+      "learning_rate": 3.3505000177075604e-06,
+      "loss": 1.6073,
+      "step": 921
+    },
+    {
+      "epoch": 1.254421768707483,
+      "grad_norm": 0.13709231144088224,
+      "learning_rate": 3.3398896185524084e-06,
+      "loss": 1.6827,
+      "step": 922
+    },
+    {
+      "epoch": 1.2557823129251702,
+      "grad_norm": 0.11431374619526342,
+      "learning_rate": 3.329287614842155e-06,
+      "loss": 1.5719,
+      "step": 923
+    },
+    {
+      "epoch": 1.2571428571428571,
+      "grad_norm": 0.1326188319367622,
+      "learning_rate": 3.3186940601928352e-06,
+      "loss": 1.6046,
+      "step": 924
+    },
+    {
+      "epoch": 1.2585034013605443,
+      "grad_norm": 0.11858562580408218,
+      "learning_rate": 3.3081090081777577e-06,
+      "loss": 1.5719,
+      "step": 925
+    },
+    {
+      "epoch": 1.2598639455782312,
+      "grad_norm": 0.1133691965363108,
+      "learning_rate": 3.2975325123272304e-06,
+      "loss": 1.6144,
+      "step": 926
+    },
+    {
+      "epoch": 1.2612244897959184,
+      "grad_norm": 0.12690673847281403,
+      "learning_rate": 3.286964626128292e-06,
+      "loss": 1.6217,
+      "step": 927
+    },
+    {
+      "epoch": 1.2625850340136053,
+      "grad_norm": 0.1329474778691663,
+      "learning_rate": 3.2764054030244406e-06,
+      "loss": 1.7347,
+      "step": 928
+    },
+    {
+      "epoch": 1.2639455782312925,
+      "grad_norm": 0.12616283586928403,
+      "learning_rate": 3.2658548964153636e-06,
+      "loss": 1.7557,
+      "step": 929
+    },
+    {
+      "epoch": 1.2653061224489797,
+      "grad_norm": 0.1153724209206064,
+      "learning_rate": 3.25531315965667e-06,
+      "loss": 1.568,
+      "step": 930
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 0.19514499732232182,
+      "learning_rate": 3.2447802460596124e-06,
+      "loss": 1.6412,
+      "step": 931
+    },
+    {
+      "epoch": 1.2680272108843538,
+      "grad_norm": 0.1324775105635469,
+      "learning_rate": 3.234256208890831e-06,
+      "loss": 1.4851,
+      "step": 932
+    },
+    {
+      "epoch": 1.269387755102041,
+      "grad_norm": 0.16422315509218421,
+      "learning_rate": 3.2237411013720686e-06,
+      "loss": 1.614,
+      "step": 933
+    },
+    {
+      "epoch": 1.2707482993197279,
+      "grad_norm": 0.40727066672865614,
+      "learning_rate": 3.2132349766799166e-06,
+      "loss": 1.6335,
+      "step": 934
+    },
+    {
+      "epoch": 1.272108843537415,
+      "grad_norm": 0.11728481208845543,
+      "learning_rate": 3.202737887945533e-06,
+      "loss": 1.6401,
+      "step": 935
+    },
+    {
+      "epoch": 1.273469387755102,
+      "grad_norm": 0.11606746557775659,
+      "learning_rate": 3.1922498882543808e-06,
+      "loss": 1.5821,
+      "step": 936
+    },
+    {
+      "epoch": 1.2748299319727892,
+      "grad_norm": 0.13755764701050108,
+      "learning_rate": 3.1817710306459638e-06,
+      "loss": 1.5831,
+      "step": 937
+    },
+    {
+      "epoch": 1.276190476190476,
+      "grad_norm": 0.11936848733746346,
+      "learning_rate": 3.171301368113545e-06,
+      "loss": 1.5987,
+      "step": 938
+    },
+    {
+      "epoch": 1.2775510204081633,
+      "grad_norm": 0.09685227260059284,
+      "learning_rate": 3.160840953603892e-06,
+      "loss": 1.7107,
+      "step": 939
+    },
+    {
+      "epoch": 1.2789115646258504,
+      "grad_norm": 0.11226497972441736,
+      "learning_rate": 3.1503898400169986e-06,
+      "loss": 1.7636,
+      "step": 940
+    },
+    {
+      "epoch": 1.2802721088435374,
+      "grad_norm": 0.19072705706329435,
+      "learning_rate": 3.139948080205828e-06,
+      "loss": 1.5306,
+      "step": 941
+    },
+    {
+      "epoch": 1.2816326530612245,
+      "grad_norm": 0.11221444877520503,
+      "learning_rate": 3.1295157269760347e-06,
+      "loss": 1.5718,
+      "step": 942
+    },
+    {
+      "epoch": 1.2829931972789117,
+      "grad_norm": 0.25831949084527644,
+      "learning_rate": 3.1190928330857063e-06,
+      "loss": 1.8018,
+      "step": 943
+    },
+    {
+      "epoch": 1.2843537414965986,
+      "grad_norm": 0.12949049150025038,
+      "learning_rate": 3.108679451245091e-06,
+      "loss": 1.6306,
+      "step": 944
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.11975166110577949,
+      "learning_rate": 3.0982756341163344e-06,
+      "loss": 1.5288,
+      "step": 945
+    },
+    {
+      "epoch": 1.2870748299319728,
+      "grad_norm": 0.12063089261265142,
+      "learning_rate": 3.087881434313212e-06,
+      "loss": 1.8165,
+      "step": 946
+    },
+    {
+      "epoch": 1.28843537414966,
+      "grad_norm": 0.10837632174717707,
+      "learning_rate": 3.0774969044008596e-06,
+      "loss": 1.7264,
+      "step": 947
+    },
+    {
+      "epoch": 1.2897959183673469,
+      "grad_norm": 0.13219269814229098,
+      "learning_rate": 3.067122096895515e-06,
+      "loss": 1.6243,
+      "step": 948
+    },
+    {
+      "epoch": 1.291156462585034,
+      "grad_norm": 0.11897597832335097,
+      "learning_rate": 3.0567570642642453e-06,
+      "loss": 1.7471,
+      "step": 949
+    },
+    {
+      "epoch": 1.2925170068027212,
+      "grad_norm": 0.11968690206341515,
+      "learning_rate": 3.0464018589246867e-06,
+      "loss": 1.544,
+      "step": 950
+    },
+    {
+      "epoch": 1.2938775510204081,
+      "grad_norm": 0.12640952696385496,
+      "learning_rate": 3.0360565332447754e-06,
+      "loss": 1.6696,
+      "step": 951
+    },
+    {
+      "epoch": 1.2952380952380953,
+      "grad_norm": 0.1598172355650226,
+      "learning_rate": 3.0257211395424846e-06,
+      "loss": 1.7097,
+      "step": 952
+    },
+    {
+      "epoch": 1.2965986394557822,
+      "grad_norm": 0.13149924526184206,
+      "learning_rate": 3.015395730085565e-06,
+      "loss": 1.6884,
+      "step": 953
+    },
+    {
+      "epoch": 1.2979591836734694,
+      "grad_norm": 0.14095309321731594,
+      "learning_rate": 3.0050803570912644e-06,
+      "loss": 1.766,
+      "step": 954
+    },
+    {
+      "epoch": 1.2993197278911564,
+      "grad_norm": 0.11707205160488976,
+      "learning_rate": 2.994775072726087e-06,
+      "loss": 1.6714,
+      "step": 955
+    },
+    {
+      "epoch": 1.3006802721088435,
+      "grad_norm": 0.11113529704832883,
+      "learning_rate": 2.984479929105508e-06,
+      "loss": 1.6734,
+      "step": 956
+    },
+    {
+      "epoch": 1.3020408163265307,
+      "grad_norm": 0.12550353089395527,
+      "learning_rate": 2.9741949782937252e-06,
+      "loss": 1.5402,
+      "step": 957
+    },
+    {
+      "epoch": 1.3034013605442176,
+      "grad_norm": 0.15756100509185716,
+      "learning_rate": 2.9639202723033843e-06,
+      "loss": 1.7012,
+      "step": 958
+    },
+    {
+      "epoch": 1.3047619047619048,
+      "grad_norm": 0.145899594151416,
+      "learning_rate": 2.9536558630953262e-06,
+      "loss": 1.6383,
+      "step": 959
+    },
+    {
+      "epoch": 1.306122448979592,
+      "grad_norm": 0.12913992593059825,
+      "learning_rate": 2.943401802578315e-06,
+      "loss": 1.5585,
+      "step": 960
+    },
+    {
+      "epoch": 1.307482993197279,
+      "grad_norm": 0.1278808504306049,
+      "learning_rate": 2.9331581426087847e-06,
+      "loss": 1.6784,
+      "step": 961
+    },
+    {
+      "epoch": 1.308843537414966,
+      "grad_norm": 0.15484400934294676,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 1.5304,
+      "step": 962
+    },
+    {
+      "epoch": 1.310204081632653,
+      "grad_norm": 0.11410951165140533,
+      "learning_rate": 2.912702231474639e-06,
+      "loss": 1.6496,
+      "step": 963
+    },
+    {
+      "epoch": 1.3115646258503402,
+      "grad_norm": 0.11183117701615923,
+      "learning_rate": 2.902490083758856e-06,
+      "loss": 1.6686,
+      "step": 964
+    },
+    {
+      "epoch": 1.3129251700680271,
+      "grad_norm": 0.10575601089764713,
+      "learning_rate": 2.8922885434876846e-06,
+      "loss": 1.6357,
+      "step": 965
+    },
+    {
+      "epoch": 1.3142857142857143,
+      "grad_norm": 0.12071397882258213,
+      "learning_rate": 2.882097662251956e-06,
+      "loss": 1.7448,
+      "step": 966
+    },
+    {
+      "epoch": 1.3156462585034014,
+      "grad_norm": 0.12145862487643339,
+      "learning_rate": 2.8719174915885935e-06,
+      "loss": 1.5548,
+      "step": 967
+    },
+    {
+      "epoch": 1.3170068027210884,
+      "grad_norm": 0.11127135568946671,
+      "learning_rate": 2.8617480829803574e-06,
+      "loss": 1.6849,
+      "step": 968
+    },
+    {
+      "epoch": 1.3183673469387756,
+      "grad_norm": 0.11147070098606063,
+      "learning_rate": 2.8515894878555766e-06,
+      "loss": 1.518,
+      "step": 969
+    },
+    {
+      "epoch": 1.3197278911564627,
+      "grad_norm": 0.12955537422802776,
+      "learning_rate": 2.841441757587901e-06,
+      "loss": 1.6614,
+      "step": 970
+    },
+    {
+      "epoch": 1.3210884353741497,
+      "grad_norm": 0.10292257541425158,
+      "learning_rate": 2.831304943496033e-06,
+      "loss": 1.7421,
+      "step": 971
+    },
+    {
+      "epoch": 1.3224489795918366,
+      "grad_norm": 0.10623781986886265,
+      "learning_rate": 2.8211790968434692e-06,
+      "loss": 1.6891,
+      "step": 972
+    },
+    {
+      "epoch": 1.3238095238095238,
+      "grad_norm": 0.13299637838745879,
+      "learning_rate": 2.8110642688382455e-06,
+      "loss": 1.7151,
+      "step": 973
+    },
+    {
+      "epoch": 1.325170068027211,
+      "grad_norm": 0.12676062857059178,
+      "learning_rate": 2.800960510632668e-06,
+      "loss": 1.7884,
+      "step": 974
+    },
+    {
+      "epoch": 1.3265306122448979,
+      "grad_norm": 0.10370471610066626,
+      "learning_rate": 2.790867873323067e-06,
+      "loss": 1.7972,
+      "step": 975
+    },
+    {
+      "epoch": 1.327891156462585,
+      "grad_norm": 0.11469407246054221,
+      "learning_rate": 2.7807864079495308e-06,
+      "loss": 1.7044,
+      "step": 976
+    },
+    {
+      "epoch": 1.3292517006802722,
+      "grad_norm": 0.127975064009094,
+      "learning_rate": 2.7707161654956516e-06,
+      "loss": 1.588,
+      "step": 977
+    },
+    {
+      "epoch": 1.3306122448979592,
+      "grad_norm": 0.1431937066191807,
+      "learning_rate": 2.760657196888259e-06,
+      "loss": 1.6176,
+      "step": 978
+    },
+    {
+      "epoch": 1.3319727891156463,
+      "grad_norm": 0.12699844201617452,
+      "learning_rate": 2.750609552997173e-06,
+      "loss": 1.6426,
+      "step": 979
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.12207870723428496,
+      "learning_rate": 2.740573284634949e-06,
+      "loss": 1.6522,
+      "step": 980
+    },
+    {
+      "epoch": 1.3346938775510204,
+      "grad_norm": 0.12255489213356785,
+      "learning_rate": 2.7305484425566033e-06,
+      "loss": 1.6833,
+      "step": 981
+    },
+    {
+      "epoch": 1.3360544217687074,
+      "grad_norm": 0.10634070168511277,
+      "learning_rate": 2.7205350774593757e-06,
+      "loss": 1.6049,
+      "step": 982
+    },
+    {
+      "epoch": 1.3374149659863945,
+      "grad_norm": 0.13904029518050998,
+      "learning_rate": 2.7105332399824567e-06,
+      "loss": 1.5175,
+      "step": 983
+    },
+    {
+      "epoch": 1.3387755102040817,
+      "grad_norm": 0.3143888763552485,
+      "learning_rate": 2.700542980706754e-06,
+      "loss": 1.6381,
+      "step": 984
+    },
+    {
+      "epoch": 1.3401360544217686,
+      "grad_norm": 0.11248808835798864,
+      "learning_rate": 2.6905643501546066e-06,
+      "loss": 1.6597,
+      "step": 985
+    },
+    {
+      "epoch": 1.3414965986394558,
+      "grad_norm": 0.12165354663856978,
+      "learning_rate": 2.680597398789554e-06,
+      "loss": 1.6165,
+      "step": 986
+    },
+    {
+      "epoch": 1.342857142857143,
+      "grad_norm": 0.13469020590289893,
+      "learning_rate": 2.670642177016071e-06,
+      "loss": 1.6989,
+      "step": 987
+    },
+    {
+      "epoch": 1.34421768707483,
+      "grad_norm": 0.10403924107277954,
+      "learning_rate": 2.6606987351793134e-06,
+      "loss": 1.6892,
+      "step": 988
+    },
+    {
+      "epoch": 1.345578231292517,
+      "grad_norm": 0.12657716162900132,
+      "learning_rate": 2.6507671235648667e-06,
+      "loss": 1.8571,
+      "step": 989
+    },
+    {
+      "epoch": 1.346938775510204,
+      "grad_norm": 0.14141908912174195,
+      "learning_rate": 2.6408473923984813e-06,
+      "loss": 1.5353,
+      "step": 990
+    },
+    {
+      "epoch": 1.3482993197278912,
+      "grad_norm": 0.1474252826828531,
+      "learning_rate": 2.630939591845834e-06,
+      "loss": 1.5924,
+      "step": 991
+    },
+    {
+      "epoch": 1.3496598639455781,
+      "grad_norm": 0.12543913644142501,
+      "learning_rate": 2.6210437720122633e-06,
+      "loss": 1.6871,
+      "step": 992
+    },
+    {
+      "epoch": 1.3510204081632653,
+      "grad_norm": 0.13883630578353837,
+      "learning_rate": 2.6111599829425228e-06,
+      "loss": 1.6789,
+      "step": 993
+    },
+    {
+      "epoch": 1.3523809523809525,
+      "grad_norm": 0.12846573886694015,
+      "learning_rate": 2.6012882746205164e-06,
+      "loss": 1.6067,
+      "step": 994
+    },
+    {
+      "epoch": 1.3537414965986394,
+      "grad_norm": 0.12008961120991984,
+      "learning_rate": 2.591428696969062e-06,
+      "loss": 1.5903,
+      "step": 995
+    },
+    {
+      "epoch": 1.3551020408163266,
+      "grad_norm": 0.11903477733151935,
+      "learning_rate": 2.581581299849627e-06,
+      "loss": 1.6369,
+      "step": 996
+    },
+    {
+      "epoch": 1.3564625850340137,
+      "grad_norm": 0.13999706538475443,
+      "learning_rate": 2.571746133062082e-06,
+      "loss": 1.5367,
+      "step": 997
+    },
+    {
+      "epoch": 1.3578231292517007,
+      "grad_norm": 0.11963009361233483,
+      "learning_rate": 2.5619232463444453e-06,
+      "loss": 1.4038,
+      "step": 998
+    },
+    {
+      "epoch": 1.3591836734693876,
+      "grad_norm": 0.12364356934092328,
+      "learning_rate": 2.5521126893726287e-06,
+      "loss": 1.5599,
+      "step": 999
+    },
+    {
+      "epoch": 1.3605442176870748,
+      "grad_norm": 0.1339584043443562,
+      "learning_rate": 2.542314511760203e-06,
+      "loss": 1.5419,
+      "step": 1000
+    },
+    {
+      "epoch": 1.361904761904762,
+      "grad_norm": 0.11610116574614443,
+      "learning_rate": 2.5325287630581197e-06,
+      "loss": 1.7461,
+      "step": 1001
+    },
+    {
+      "epoch": 1.363265306122449,
+      "grad_norm": 0.11167585561833832,
+      "learning_rate": 2.5227554927544863e-06,
+      "loss": 1.7333,
+      "step": 1002
+    },
+    {
+      "epoch": 1.364625850340136,
+      "grad_norm": 0.1443619458869183,
+      "learning_rate": 2.512994750274295e-06,
+      "loss": 1.5944,
+      "step": 1003
+    },
+    {
+      "epoch": 1.3659863945578232,
+      "grad_norm": 0.248168814176006,
+      "learning_rate": 2.5032465849791964e-06,
+      "loss": 1.781,
+      "step": 1004
+    },
+    {
+      "epoch": 1.3673469387755102,
+      "grad_norm": 0.23421079605400402,
+      "learning_rate": 2.4935110461672225e-06,
+      "loss": 1.635,
+      "step": 1005
+    },
+    {
+      "epoch": 1.3687074829931973,
+      "grad_norm": 0.1446962355000272,
+      "learning_rate": 2.4837881830725584e-06,
+      "loss": 1.4852,
+      "step": 1006
+    },
+    {
+      "epoch": 1.3700680272108843,
+      "grad_norm": 0.1291998509141991,
+      "learning_rate": 2.4740780448652866e-06,
+      "loss": 1.6777,
+      "step": 1007
+    },
+    {
+      "epoch": 1.3714285714285714,
+      "grad_norm": 0.09864313043191297,
+      "learning_rate": 2.4643806806511344e-06,
+      "loss": 1.6307,
+      "step": 1008
+    },
+    {
+      "epoch": 1.3727891156462584,
+      "grad_norm": 0.11658909231376899,
+      "learning_rate": 2.4546961394712324e-06,
+      "loss": 1.6741,
+      "step": 1009
+    },
+    {
+      "epoch": 1.3741496598639455,
+      "grad_norm": 0.12095557653741797,
+      "learning_rate": 2.4450244703018587e-06,
+      "loss": 1.6758,
+      "step": 1010
+    },
+    {
+      "epoch": 1.3755102040816327,
+      "grad_norm": 0.13782074238898825,
+      "learning_rate": 2.4353657220541986e-06,
+      "loss": 1.4454,
+      "step": 1011
+    },
+    {
+      "epoch": 1.3768707482993197,
+      "grad_norm": 0.12061869165588096,
+      "learning_rate": 2.4257199435740947e-06,
+      "loss": 1.6681,
+      "step": 1012
+    },
+    {
+      "epoch": 1.3782312925170068,
+      "grad_norm": 0.10755743305983602,
+      "learning_rate": 2.4160871836417992e-06,
+      "loss": 1.7763,
+      "step": 1013
+    },
+    {
+      "epoch": 1.379591836734694,
+      "grad_norm": 0.13833342392035397,
+      "learning_rate": 2.406467490971724e-06,
+      "loss": 1.4646,
+      "step": 1014
+    },
+    {
+      "epoch": 1.380952380952381,
+      "grad_norm": 0.14780468168264022,
+      "learning_rate": 2.396860914212199e-06,
+      "loss": 1.5136,
+      "step": 1015
+    },
+    {
+      "epoch": 1.382312925170068,
+      "grad_norm": 0.11362232721903939,
+      "learning_rate": 2.387267501945233e-06,
+      "loss": 1.644,
+      "step": 1016
+    },
+    {
+      "epoch": 1.383673469387755,
+      "grad_norm": 0.227770381091171,
+      "learning_rate": 2.3776873026862466e-06,
+      "loss": 1.7492,
+      "step": 1017
+    },
+    {
+      "epoch": 1.3850340136054422,
+      "grad_norm": 0.11603935670163865,
+      "learning_rate": 2.368120364883851e-06,
+      "loss": 1.5114,
+      "step": 1018
+    },
+    {
+      "epoch": 1.3863945578231291,
+      "grad_norm": 0.1168850031428384,
+      "learning_rate": 2.3585667369195815e-06,
+      "loss": 1.623,
+      "step": 1019
+    },
+    {
+      "epoch": 1.3877551020408163,
+      "grad_norm": 0.11197564926295547,
+      "learning_rate": 2.349026467107677e-06,
+      "loss": 1.8705,
+      "step": 1020
+    },
+    {
+      "epoch": 1.3891156462585035,
+      "grad_norm": 0.144237418189964,
+      "learning_rate": 2.3394996036948094e-06,
+      "loss": 1.6448,
+      "step": 1021
+    },
+    {
+      "epoch": 1.3904761904761904,
+      "grad_norm": 0.11173691780000068,
+      "learning_rate": 2.3299861948598585e-06,
+      "loss": 1.6681,
+      "step": 1022
+    },
+    {
+      "epoch": 1.3918367346938776,
+      "grad_norm": 0.15293585820058914,
+      "learning_rate": 2.320486288713662e-06,
+      "loss": 1.6767,
+      "step": 1023
+    },
+    {
+      "epoch": 1.3931972789115648,
+      "grad_norm": 0.11604159237097413,
+      "learning_rate": 2.3109999332987714e-06,
+      "loss": 1.7475,
+      "step": 1024
+    },
+    {
+      "epoch": 1.3945578231292517,
+      "grad_norm": 0.11608326681410934,
+      "learning_rate": 2.301527176589211e-06,
+      "loss": 1.7182,
+      "step": 1025
+    },
+    {
+      "epoch": 1.3959183673469386,
+      "grad_norm": 0.11802001766377367,
+      "learning_rate": 2.2920680664902305e-06,
+      "loss": 1.6692,
+      "step": 1026
+    },
+    {
+      "epoch": 1.3972789115646258,
+      "grad_norm": 0.14326115262175532,
+      "learning_rate": 2.282622650838071e-06,
+      "loss": 1.5178,
+      "step": 1027
+    },
+    {
+      "epoch": 1.398639455782313,
+      "grad_norm": 0.12464124111634249,
+      "learning_rate": 2.273190977399717e-06,
+      "loss": 1.7335,
+      "step": 1028
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.12412136945656944,
+      "learning_rate": 2.2637730938726578e-06,
+      "loss": 1.6765,
+      "step": 1029
+    },
+    {
+      "epoch": 1.401360544217687,
+      "grad_norm": 0.11072133022020544,
+      "learning_rate": 2.254369047884639e-06,
+      "loss": 1.6853,
+      "step": 1030
+    },
+    {
+      "epoch": 1.4027210884353742,
+      "grad_norm": 0.11744625816603456,
+      "learning_rate": 2.2449788869934332e-06,
+      "loss": 1.7241,
+      "step": 1031
+    },
+    {
+      "epoch": 1.4040816326530612,
+      "grad_norm": 0.11305715625400635,
+      "learning_rate": 2.2356026586865976e-06,
+      "loss": 1.828,
+      "step": 1032
+    },
+    {
+      "epoch": 1.4054421768707483,
+      "grad_norm": 0.11411548926790505,
+      "learning_rate": 2.2262404103812197e-06,
+      "loss": 1.7667,
+      "step": 1033
+    },
+    {
+      "epoch": 1.4068027210884353,
+      "grad_norm": 0.13159909159310862,
+      "learning_rate": 2.2168921894236963e-06,
+      "loss": 1.772,
+      "step": 1034
+    },
+    {
+      "epoch": 1.4081632653061225,
+      "grad_norm": 0.10869740493369748,
+      "learning_rate": 2.207558043089478e-06,
+      "loss": 1.6783,
+      "step": 1035
+    },
+    {
+      "epoch": 1.4095238095238094,
+      "grad_norm": 0.18634482438552744,
+      "learning_rate": 2.1982380185828485e-06,
+      "loss": 1.6199,
+      "step": 1036
+    },
+    {
+      "epoch": 1.4108843537414966,
+      "grad_norm": 0.12490228313384627,
+      "learning_rate": 2.1889321630366645e-06,
+      "loss": 1.6002,
+      "step": 1037
+    },
+    {
+      "epoch": 1.4122448979591837,
+      "grad_norm": 0.13407480013102485,
+      "learning_rate": 2.179640523512136e-06,
+      "loss": 1.4965,
+      "step": 1038
+    },
+    {
+      "epoch": 1.4136054421768707,
+      "grad_norm": 0.12540761630148858,
+      "learning_rate": 2.1703631469985713e-06,
+      "loss": 1.5602,
+      "step": 1039
+    },
+    {
+      "epoch": 1.4149659863945578,
+      "grad_norm": 0.11826915965312917,
+      "learning_rate": 2.1611000804131588e-06,
+      "loss": 1.6845,
+      "step": 1040
+    },
+    {
+      "epoch": 1.416326530612245,
+      "grad_norm": 0.11638811328551839,
+      "learning_rate": 2.1518513706007154e-06,
+      "loss": 1.7135,
+      "step": 1041
+    },
+    {
+      "epoch": 1.417687074829932,
+      "grad_norm": 0.12121992820341544,
+      "learning_rate": 2.1426170643334477e-06,
+      "loss": 1.6918,
+      "step": 1042
+    },
+    {
+      "epoch": 1.4190476190476191,
+      "grad_norm": 0.127097445433669,
+      "learning_rate": 2.133397208310728e-06,
+      "loss": 1.5874,
+      "step": 1043
+    },
+    {
+      "epoch": 1.420408163265306,
+      "grad_norm": 0.1120249401560636,
+      "learning_rate": 2.1241918491588503e-06,
+      "loss": 1.6122,
+      "step": 1044
+    },
+    {
+      "epoch": 1.4217687074829932,
+      "grad_norm": 0.13980832815015817,
+      "learning_rate": 2.1150010334307954e-06,
+      "loss": 1.7686,
+      "step": 1045
+    },
+    {
+      "epoch": 1.4231292517006802,
+      "grad_norm": 0.16218374995611207,
+      "learning_rate": 2.1058248076059916e-06,
+      "loss": 1.7442,
+      "step": 1046
+    },
+    {
+      "epoch": 1.4244897959183673,
+      "grad_norm": 0.11333274239691936,
+      "learning_rate": 2.09666321809009e-06,
+      "loss": 1.7052,
+      "step": 1047
+    },
+    {
+      "epoch": 1.4258503401360545,
+      "grad_norm": 0.11296676049286146,
+      "learning_rate": 2.087516311214719e-06,
+      "loss": 1.6709,
+      "step": 1048
+    },
+    {
+      "epoch": 1.4272108843537414,
+      "grad_norm": 0.16489322883195276,
+      "learning_rate": 2.0783841332372567e-06,
+      "loss": 1.6074,
+      "step": 1049
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.138226915964094,
+      "learning_rate": 2.0692667303405962e-06,
+      "loss": 1.6082,
+      "step": 1050
+    },
+    {
+      "epoch": 1.4299319727891158,
+      "grad_norm": 0.15926811847519332,
+      "learning_rate": 2.060164148632903e-06,
+      "loss": 1.5277,
+      "step": 1051
+    },
+    {
+      "epoch": 1.4312925170068027,
+      "grad_norm": 0.16960130944371998,
+      "learning_rate": 2.0510764341474032e-06,
+      "loss": 1.6701,
+      "step": 1052
+    },
+    {
+      "epoch": 1.4326530612244899,
+      "grad_norm": 0.11553461200820175,
+      "learning_rate": 2.042003632842123e-06,
+      "loss": 1.604,
+      "step": 1053
+    },
+    {
+      "epoch": 1.4340136054421768,
+      "grad_norm": 0.12461036286243339,
+      "learning_rate": 2.0329457905996785e-06,
+      "loss": 1.7384,
+      "step": 1054
+    },
+    {
+      "epoch": 1.435374149659864,
+      "grad_norm": 0.12232104998061016,
+      "learning_rate": 2.023902953227029e-06,
+      "loss": 1.639,
+      "step": 1055
+    },
+    {
+      "epoch": 1.436734693877551,
+      "grad_norm": 0.11267495872797326,
+      "learning_rate": 2.014875166455264e-06,
+      "loss": 1.7093,
+      "step": 1056
+    },
+    {
+      "epoch": 1.438095238095238,
+      "grad_norm": 0.11804588705775097,
+      "learning_rate": 2.005862475939344e-06,
+      "loss": 1.7411,
+      "step": 1057
+    },
+    {
+      "epoch": 1.4394557823129253,
+      "grad_norm": 0.10870950926259154,
+      "learning_rate": 1.9968649272578955e-06,
+      "loss": 1.8004,
+      "step": 1058
+    },
+    {
+      "epoch": 1.4408163265306122,
+      "grad_norm": 0.15480060013281374,
+      "learning_rate": 1.987882565912968e-06,
+      "loss": 1.5773,
+      "step": 1059
+    },
+    {
+      "epoch": 1.4421768707482994,
+      "grad_norm": 0.13989692137884865,
+      "learning_rate": 1.978915437329806e-06,
+      "loss": 1.6361,
+      "step": 1060
+    },
+    {
+      "epoch": 1.4435374149659865,
+      "grad_norm": 0.10649783390764571,
+      "learning_rate": 1.96996358685662e-06,
+      "loss": 1.6262,
+      "step": 1061
+    },
+    {
+      "epoch": 1.4448979591836735,
+      "grad_norm": 0.14326122461794985,
+      "learning_rate": 1.961027059764354e-06,
+      "loss": 1.5732,
+      "step": 1062
+    },
+    {
+      "epoch": 1.4462585034013604,
+      "grad_norm": 0.12408176967614352,
+      "learning_rate": 1.952105901246461e-06,
+      "loss": 1.5585,
+      "step": 1063
+    },
+    {
+      "epoch": 1.4476190476190476,
+      "grad_norm": 0.16522994922872408,
+      "learning_rate": 1.943200156418674e-06,
+      "loss": 1.6795,
+      "step": 1064
+    },
+    {
+      "epoch": 1.4489795918367347,
+      "grad_norm": 0.3681795789769912,
+      "learning_rate": 1.9343098703187777e-06,
+      "loss": 1.5695,
+      "step": 1065
+    },
+    {
+      "epoch": 1.4503401360544217,
+      "grad_norm": 0.13093870638144472,
+      "learning_rate": 1.925435087906373e-06,
+      "loss": 1.4991,
+      "step": 1066
+    },
+    {
+      "epoch": 1.4517006802721089,
+      "grad_norm": 0.11132092495609136,
+      "learning_rate": 1.9165758540626595e-06,
+      "loss": 1.7495,
+      "step": 1067
+    },
+    {
+      "epoch": 1.453061224489796,
+      "grad_norm": 0.1320841952456979,
+      "learning_rate": 1.907732213590212e-06,
+      "loss": 1.6193,
+      "step": 1068
+    },
+    {
+      "epoch": 1.454421768707483,
+      "grad_norm": 0.13072820770934512,
+      "learning_rate": 1.898904211212736e-06,
+      "loss": 1.6072,
+      "step": 1069
+    },
+    {
+      "epoch": 1.4557823129251701,
+      "grad_norm": 0.1309250057364391,
+      "learning_rate": 1.8900918915748612e-06,
+      "loss": 1.6395,
+      "step": 1070
+    },
+    {
+      "epoch": 1.457142857142857,
+      "grad_norm": 0.12551858156974602,
+      "learning_rate": 1.881295299241898e-06,
+      "loss": 1.6186,
+      "step": 1071
+    },
+    {
+      "epoch": 1.4585034013605442,
+      "grad_norm": 0.13244662699660312,
+      "learning_rate": 1.8725144786996347e-06,
+      "loss": 1.7768,
+      "step": 1072
+    },
+    {
+      "epoch": 1.4598639455782312,
+      "grad_norm": 0.12869848294692346,
+      "learning_rate": 1.8637494743540851e-06,
+      "loss": 1.8306,
+      "step": 1073
+    },
+    {
+      "epoch": 1.4612244897959183,
+      "grad_norm": 0.10453495592437635,
+      "learning_rate": 1.855000330531289e-06,
+      "loss": 1.7562,
+      "step": 1074
+    },
+    {
+      "epoch": 1.4625850340136055,
+      "grad_norm": 0.11842884214674637,
+      "learning_rate": 1.846267091477067e-06,
+      "loss": 1.6221,
+      "step": 1075
+    },
+    {
+      "epoch": 1.4639455782312925,
+      "grad_norm": 0.1366451554577879,
+      "learning_rate": 1.8375498013568166e-06,
+      "loss": 1.4415,
+      "step": 1076
+    },
+    {
+      "epoch": 1.4653061224489796,
+      "grad_norm": 0.11721529234776305,
+      "learning_rate": 1.8288485042552739e-06,
+      "loss": 1.6414,
+      "step": 1077
+    },
+    {
+      "epoch": 1.4666666666666668,
+      "grad_norm": 0.1296793815948946,
+      "learning_rate": 1.820163244176294e-06,
+      "loss": 1.5815,
+      "step": 1078
+    },
+    {
+      "epoch": 1.4680272108843537,
+      "grad_norm": 0.10030127712618647,
+      "learning_rate": 1.8114940650426332e-06,
+      "loss": 1.7649,
+      "step": 1079
+    },
+    {
+      "epoch": 1.469387755102041,
+      "grad_norm": 0.14298941316959418,
+      "learning_rate": 1.8028410106957239e-06,
+      "loss": 1.6847,
+      "step": 1080
+    },
+    {
+      "epoch": 1.4707482993197278,
+      "grad_norm": 0.140985161297301,
+      "learning_rate": 1.7942041248954534e-06,
+      "loss": 1.6794,
+      "step": 1081
+    },
+    {
+      "epoch": 1.472108843537415,
+      "grad_norm": 0.11886070737447318,
+      "learning_rate": 1.7855834513199365e-06,
+      "loss": 1.6796,
+      "step": 1082
+    },
+    {
+      "epoch": 1.473469387755102,
+      "grad_norm": 0.12803178657461758,
+      "learning_rate": 1.7769790335653076e-06,
+      "loss": 1.543,
+      "step": 1083
+    },
+    {
+      "epoch": 1.474829931972789,
+      "grad_norm": 0.10256161260658421,
+      "learning_rate": 1.7683909151454888e-06,
+      "loss": 1.7032,
+      "step": 1084
+    },
+    {
+      "epoch": 1.4761904761904763,
+      "grad_norm": 0.21096092474063058,
+      "learning_rate": 1.7598191394919738e-06,
+      "loss": 1.5423,
+      "step": 1085
+    },
+    {
+      "epoch": 1.4775510204081632,
+      "grad_norm": 0.12083343658033356,
+      "learning_rate": 1.7512637499536123e-06,
+      "loss": 1.6169,
+      "step": 1086
+    },
+    {
+      "epoch": 1.4789115646258504,
+      "grad_norm": 0.11608713517334812,
+      "learning_rate": 1.7427247897963784e-06,
+      "loss": 1.7293,
+      "step": 1087
+    },
+    {
+      "epoch": 1.4802721088435375,
+      "grad_norm": 0.1319069173143083,
+      "learning_rate": 1.7342023022031712e-06,
+      "loss": 1.455,
+      "step": 1088
+    },
+    {
+      "epoch": 1.4816326530612245,
+      "grad_norm": 0.09994450469778726,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 1.8571,
+      "step": 1089
+    },
+    {
+      "epoch": 1.4829931972789114,
+      "grad_norm": 0.11652396846241518,
+      "learning_rate": 1.71720691702366e-06,
+      "loss": 1.7252,
+      "step": 1090
+    },
+    {
+      "epoch": 1.4843537414965986,
+      "grad_norm": 0.14171795087480568,
+      "learning_rate": 1.7087341053857475e-06,
+      "loss": 1.5228,
+      "step": 1091
+    },
+    {
+      "epoch": 1.4857142857142858,
+      "grad_norm": 0.11989790454950104,
+      "learning_rate": 1.700277938208214e-06,
+      "loss": 1.577,
+      "step": 1092
+    },
+    {
+      "epoch": 1.4870748299319727,
+      "grad_norm": 0.10858297624497283,
+      "learning_rate": 1.6918384582552504e-06,
+      "loss": 1.7052,
+      "step": 1093
+    },
+    {
+      "epoch": 1.4884353741496599,
+      "grad_norm": 0.10917059280407902,
+      "learning_rate": 1.6834157082066638e-06,
+      "loss": 1.7003,
+      "step": 1094
+    },
+    {
+      "epoch": 1.489795918367347,
+      "grad_norm": 0.0982808829346907,
+      "learning_rate": 1.6750097306576552e-06,
+      "loss": 1.6922,
+      "step": 1095
+    },
+    {
+      "epoch": 1.491156462585034,
+      "grad_norm": 0.13697494235902768,
+      "learning_rate": 1.6666205681186032e-06,
+      "loss": 1.5286,
+      "step": 1096
+    },
+    {
+      "epoch": 1.4925170068027211,
+      "grad_norm": 0.12713384229112287,
+      "learning_rate": 1.6582482630148516e-06,
+      "loss": 1.5449,
+      "step": 1097
+    },
+    {
+      "epoch": 1.493877551020408,
+      "grad_norm": 0.28082296494568143,
+      "learning_rate": 1.6498928576864892e-06,
+      "loss": 1.6008,
+      "step": 1098
+    },
+    {
+      "epoch": 1.4952380952380953,
+      "grad_norm": 0.12441418528009313,
+      "learning_rate": 1.6415543943881457e-06,
+      "loss": 1.6326,
+      "step": 1099
+    },
+    {
+      "epoch": 1.4965986394557822,
+      "grad_norm": 0.12076209117459938,
+      "learning_rate": 1.6332329152887683e-06,
+      "loss": 1.6048,
+      "step": 1100
+    },
+    {
+      "epoch": 1.4979591836734694,
+      "grad_norm": 0.10773083863289847,
+      "learning_rate": 1.624928462471414e-06,
+      "loss": 1.7003,
+      "step": 1101
+    },
+    {
+      "epoch": 1.4993197278911565,
+      "grad_norm": 0.12873545682401677,
+      "learning_rate": 1.6166410779330372e-06,
+      "loss": 1.7761,
+      "step": 1102
+    },
+    {
+      "epoch": 1.5006802721088435,
+      "grad_norm": 0.12486712488322993,
+      "learning_rate": 1.6083708035842683e-06,
+      "loss": 1.6096,
+      "step": 1103
+    },
+    {
+      "epoch": 1.5020408163265306,
+      "grad_norm": 0.11819276519267044,
+      "learning_rate": 1.6001176812492164e-06,
+      "loss": 1.644,
+      "step": 1104
+    },
+    {
+      "epoch": 1.5020408163265306,
+      "eval_loss": 1.686686635017395,
+      "eval_runtime": 76.5083,
+      "eval_samples_per_second": 53.236,
+      "eval_steps_per_second": 6.666,
+      "step": 1104
+    },
+    {
+      "epoch": 1.5034013605442178,
+      "grad_norm": 0.12565443858963696,
+      "learning_rate": 1.5918817526652457e-06,
+      "loss": 1.6855,
+      "step": 1105
+    },
+    {
+      "epoch": 1.5047619047619047,
+      "grad_norm": 0.1235458965699543,
+      "learning_rate": 1.583663059482773e-06,
+      "loss": 1.5379,
+      "step": 1106
+    },
+    {
+      "epoch": 1.5061224489795917,
+      "grad_norm": 0.11893034993481316,
+      "learning_rate": 1.5754616432650443e-06,
+      "loss": 1.6431,
+      "step": 1107
+    },
+    {
+      "epoch": 1.507482993197279,
+      "grad_norm": 0.11339427750946436,
+      "learning_rate": 1.5672775454879458e-06,
+      "loss": 1.7272,
+      "step": 1108
+    },
+    {
+      "epoch": 1.508843537414966,
+      "grad_norm": 0.1845008212172551,
+      "learning_rate": 1.5591108075397698e-06,
+      "loss": 1.6404,
+      "step": 1109
+    },
+    {
+      "epoch": 1.510204081632653,
+      "grad_norm": 0.12462997782891976,
+      "learning_rate": 1.5509614707210235e-06,
+      "loss": 1.7679,
+      "step": 1110
+    },
+    {
+      "epoch": 1.5115646258503401,
+      "grad_norm": 0.12287064399894775,
+      "learning_rate": 1.542829576244213e-06,
+      "loss": 1.7342,
+      "step": 1111
+    },
+    {
+      "epoch": 1.5129251700680273,
+      "grad_norm": 0.1480464966747393,
+      "learning_rate": 1.5347151652336351e-06,
+      "loss": 1.552,
+      "step": 1112
+    },
+    {
+      "epoch": 1.5142857142857142,
+      "grad_norm": 0.11081261407520034,
+      "learning_rate": 1.5266182787251705e-06,
+      "loss": 1.8438,
+      "step": 1113
+    },
+    {
+      "epoch": 1.5156462585034014,
+      "grad_norm": 0.1029266924311367,
+      "learning_rate": 1.5185389576660725e-06,
+      "loss": 1.7652,
+      "step": 1114
+    },
+    {
+      "epoch": 1.5170068027210886,
+      "grad_norm": 0.12456172646180011,
+      "learning_rate": 1.5104772429147674e-06,
+      "loss": 1.5471,
+      "step": 1115
+    },
+    {
+      "epoch": 1.5183673469387755,
+      "grad_norm": 0.1371815920370596,
+      "learning_rate": 1.5024331752406418e-06,
+      "loss": 1.7099,
+      "step": 1116
+    },
+    {
+      "epoch": 1.5197278911564625,
+      "grad_norm": 0.12573407133507336,
+      "learning_rate": 1.4944067953238396e-06,
+      "loss": 1.4374,
+      "step": 1117
+    },
+    {
+      "epoch": 1.5210884353741496,
+      "grad_norm": 0.13604275704368526,
+      "learning_rate": 1.48639814375505e-06,
+      "loss": 1.5731,
+      "step": 1118
+    },
+    {
+      "epoch": 1.5224489795918368,
+      "grad_norm": 0.25410293091386754,
+      "learning_rate": 1.4784072610353129e-06,
+      "loss": 1.7197,
+      "step": 1119
+    },
+    {
+      "epoch": 1.5238095238095237,
+      "grad_norm": 0.114015489104318,
+      "learning_rate": 1.4704341875758055e-06,
+      "loss": 1.766,
+      "step": 1120
+    },
+    {
+      "epoch": 1.525170068027211,
+      "grad_norm": 0.12807572724335445,
+      "learning_rate": 1.4624789636976411e-06,
+      "loss": 1.4984,
+      "step": 1121
+    },
+    {
+      "epoch": 1.526530612244898,
+      "grad_norm": 0.12697960510593037,
+      "learning_rate": 1.4545416296316667e-06,
+      "loss": 1.4318,
+      "step": 1122
+    },
+    {
+      "epoch": 1.527891156462585,
+      "grad_norm": 0.2166265032856606,
+      "learning_rate": 1.446622225518251e-06,
+      "loss": 1.5213,
+      "step": 1123
+    },
+    {
+      "epoch": 1.5292517006802722,
+      "grad_norm": 0.12831405795176154,
+      "learning_rate": 1.4387207914070995e-06,
+      "loss": 1.8113,
+      "step": 1124
+    },
+    {
+      "epoch": 1.5306122448979593,
+      "grad_norm": 0.11776616874418332,
+      "learning_rate": 1.4308373672570286e-06,
+      "loss": 1.6389,
+      "step": 1125
+    },
+    {
+      "epoch": 1.5319727891156463,
+      "grad_norm": 0.11476693865274581,
+      "learning_rate": 1.4229719929357838e-06,
+      "loss": 1.6624,
+      "step": 1126
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 0.11217140275465112,
+      "learning_rate": 1.4151247082198223e-06,
+      "loss": 1.4694,
+      "step": 1127
+    },
+    {
+      "epoch": 1.5346938775510204,
+      "grad_norm": 0.12808950619446724,
+      "learning_rate": 1.4072955527941283e-06,
+      "loss": 1.5511,
+      "step": 1128
+    },
+    {
+      "epoch": 1.5360544217687075,
+      "grad_norm": 0.10834567787069439,
+      "learning_rate": 1.3994845662519985e-06,
+      "loss": 1.7348,
+      "step": 1129
+    },
+    {
+      "epoch": 1.5374149659863945,
+      "grad_norm": 0.12572410397799513,
+      "learning_rate": 1.3916917880948434e-06,
+      "loss": 1.6245,
+      "step": 1130
+    },
+    {
+      "epoch": 1.5387755102040817,
+      "grad_norm": 0.13586524305503017,
+      "learning_rate": 1.3839172577319953e-06,
+      "loss": 1.6267,
+      "step": 1131
+    },
+    {
+      "epoch": 1.5401360544217688,
+      "grad_norm": 0.1286100284581808,
+      "learning_rate": 1.3761610144805037e-06,
+      "loss": 1.6301,
+      "step": 1132
+    },
+    {
+      "epoch": 1.5414965986394558,
+      "grad_norm": 0.1380668839742671,
+      "learning_rate": 1.3684230975649387e-06,
+      "loss": 1.5868,
+      "step": 1133
+    },
+    {
+      "epoch": 1.5428571428571427,
+      "grad_norm": 0.12199901535322918,
+      "learning_rate": 1.3607035461171858e-06,
+      "loss": 1.4813,
+      "step": 1134
+    },
+    {
+      "epoch": 1.54421768707483,
+      "grad_norm": 0.14412311583495385,
+      "learning_rate": 1.3530023991762582e-06,
+      "loss": 1.6484,
+      "step": 1135
+    },
+    {
+      "epoch": 1.545578231292517,
+      "grad_norm": 0.13408653590603778,
+      "learning_rate": 1.3453196956880932e-06,
+      "loss": 1.4982,
+      "step": 1136
+    },
+    {
+      "epoch": 1.546938775510204,
+      "grad_norm": 0.10781564217676724,
+      "learning_rate": 1.3376554745053566e-06,
+      "loss": 1.6386,
+      "step": 1137
+    },
+    {
+      "epoch": 1.5482993197278911,
+      "grad_norm": 0.12672977099050545,
+      "learning_rate": 1.3300097743872476e-06,
+      "loss": 1.6502,
+      "step": 1138
+    },
+    {
+      "epoch": 1.5496598639455783,
+      "grad_norm": 0.2591283699743902,
+      "learning_rate": 1.3223826339992973e-06,
+      "loss": 1.7084,
+      "step": 1139
+    },
+    {
+      "epoch": 1.5510204081632653,
+      "grad_norm": 0.11440960704306805,
+      "learning_rate": 1.3147740919131814e-06,
+      "loss": 1.7807,
+      "step": 1140
+    },
+    {
+      "epoch": 1.5523809523809524,
+      "grad_norm": 0.11225169501875551,
+      "learning_rate": 1.3071841866065194e-06,
+      "loss": 1.7457,
+      "step": 1141
+    },
+    {
+      "epoch": 1.5537414965986396,
+      "grad_norm": 0.11995542951043948,
+      "learning_rate": 1.299612956462683e-06,
+      "loss": 1.6081,
+      "step": 1142
+    },
+    {
+      "epoch": 1.5551020408163265,
+      "grad_norm": 0.12456556248904507,
+      "learning_rate": 1.2920604397705955e-06,
+      "loss": 1.5734,
+      "step": 1143
+    },
+    {
+      "epoch": 1.5564625850340135,
+      "grad_norm": 0.12126055030258995,
+      "learning_rate": 1.2845266747245528e-06,
+      "loss": 1.7169,
+      "step": 1144
+    },
+    {
+      "epoch": 1.5578231292517006,
+      "grad_norm": 0.11877363967721631,
+      "learning_rate": 1.27701169942401e-06,
+      "loss": 1.8646,
+      "step": 1145
+    },
+    {
+      "epoch": 1.5591836734693878,
+      "grad_norm": 0.1274047299203738,
+      "learning_rate": 1.269515551873407e-06,
+      "loss": 1.5867,
+      "step": 1146
+    },
+    {
+      "epoch": 1.5605442176870747,
+      "grad_norm": 0.24308037811468208,
+      "learning_rate": 1.262038269981966e-06,
+      "loss": 1.7842,
+      "step": 1147
+    },
+    {
+      "epoch": 1.561904761904762,
+      "grad_norm": 0.101172361144347,
+      "learning_rate": 1.254579891563502e-06,
+      "loss": 1.7421,
+      "step": 1148
+    },
+    {
+      "epoch": 1.563265306122449,
+      "grad_norm": 0.11594265987998417,
+      "learning_rate": 1.2471404543362353e-06,
+      "loss": 1.6775,
+      "step": 1149
+    },
+    {
+      "epoch": 1.564625850340136,
+      "grad_norm": 0.13220275668191248,
+      "learning_rate": 1.2397199959225914e-06,
+      "loss": 1.8161,
+      "step": 1150
+    },
+    {
+      "epoch": 1.5659863945578232,
+      "grad_norm": 0.18599776723683806,
+      "learning_rate": 1.232318553849023e-06,
+      "loss": 1.6873,
+      "step": 1151
+    },
+    {
+      "epoch": 1.5673469387755103,
+      "grad_norm": 0.12379825057112667,
+      "learning_rate": 1.2249361655458104e-06,
+      "loss": 1.5921,
+      "step": 1152
+    },
+    {
+      "epoch": 1.5687074829931973,
+      "grad_norm": 0.1388929946755345,
+      "learning_rate": 1.2175728683468802e-06,
+      "loss": 1.601,
+      "step": 1153
+    },
+    {
+      "epoch": 1.5700680272108842,
+      "grad_norm": 0.11853041576236652,
+      "learning_rate": 1.210228699489605e-06,
+      "loss": 1.6044,
+      "step": 1154
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.10770432816956119,
+      "learning_rate": 1.2029036961146284e-06,
+      "loss": 1.6532,
+      "step": 1155
+    },
+    {
+      "epoch": 1.5727891156462586,
+      "grad_norm": 0.11627754549370563,
+      "learning_rate": 1.19559789526567e-06,
+      "loss": 1.573,
+      "step": 1156
+    },
+    {
+      "epoch": 1.5741496598639455,
+      "grad_norm": 0.12390094032474464,
+      "learning_rate": 1.1883113338893376e-06,
+      "loss": 1.6402,
+      "step": 1157
+    },
+    {
+      "epoch": 1.5755102040816327,
+      "grad_norm": 0.13389994045491743,
+      "learning_rate": 1.181044048834944e-06,
+      "loss": 1.7001,
+      "step": 1158
+    },
+    {
+      "epoch": 1.5768707482993198,
+      "grad_norm": 0.13235829797643509,
+      "learning_rate": 1.173796076854315e-06,
+      "loss": 1.5359,
+      "step": 1159
+    },
+    {
+      "epoch": 1.5782312925170068,
+      "grad_norm": 0.14770782101504495,
+      "learning_rate": 1.1665674546016103e-06,
+      "loss": 1.618,
+      "step": 1160
+    },
+    {
+      "epoch": 1.5795918367346937,
+      "grad_norm": 0.12522035555164912,
+      "learning_rate": 1.1593582186331342e-06,
+      "loss": 1.7929,
+      "step": 1161
+    },
+    {
+      "epoch": 1.580952380952381,
+      "grad_norm": 0.13198188480502632,
+      "learning_rate": 1.1521684054071524e-06,
+      "loss": 1.6348,
+      "step": 1162
+    },
+    {
+      "epoch": 1.582312925170068,
+      "grad_norm": 0.12666436947708926,
+      "learning_rate": 1.144998051283701e-06,
+      "loss": 1.5635,
+      "step": 1163
+    },
+    {
+      "epoch": 1.583673469387755,
+      "grad_norm": 0.13525407680642348,
+      "learning_rate": 1.1378471925244172e-06,
+      "loss": 1.8656,
+      "step": 1164
+    },
+    {
+      "epoch": 1.5850340136054422,
+      "grad_norm": 0.15049864837576243,
+      "learning_rate": 1.1307158652923427e-06,
+      "loss": 1.4551,
+      "step": 1165
+    },
+    {
+      "epoch": 1.5863945578231293,
+      "grad_norm": 0.1353796405332583,
+      "learning_rate": 1.1236041056517416e-06,
+      "loss": 1.7247,
+      "step": 1166
+    },
+    {
+      "epoch": 1.5877551020408163,
+      "grad_norm": 0.13147241896533723,
+      "learning_rate": 1.1165119495679272e-06,
+      "loss": 1.5845,
+      "step": 1167
+    },
+    {
+      "epoch": 1.5891156462585034,
+      "grad_norm": 0.11125489339677437,
+      "learning_rate": 1.1094394329070713e-06,
+      "loss": 1.6689,
+      "step": 1168
+    },
+    {
+      "epoch": 1.5904761904761906,
+      "grad_norm": 0.14732043400502948,
+      "learning_rate": 1.1023865914360288e-06,
+      "loss": 1.4794,
+      "step": 1169
+    },
+    {
+      "epoch": 1.5918367346938775,
+      "grad_norm": 0.1455095242199474,
+      "learning_rate": 1.095353460822149e-06,
+      "loss": 1.6856,
+      "step": 1170
+    },
+    {
+      "epoch": 1.5931972789115645,
+      "grad_norm": 0.14264573128464317,
+      "learning_rate": 1.0883400766331047e-06,
+      "loss": 1.7051,
+      "step": 1171
+    },
+    {
+      "epoch": 1.5945578231292517,
+      "grad_norm": 0.2311325753008129,
+      "learning_rate": 1.081346474336707e-06,
+      "loss": 1.6173,
+      "step": 1172
+    },
+    {
+      "epoch": 1.5959183673469388,
+      "grad_norm": 0.14054024648190236,
+      "learning_rate": 1.0743726893007257e-06,
+      "loss": 1.7158,
+      "step": 1173
+    },
+    {
+      "epoch": 1.5972789115646258,
+      "grad_norm": 0.16087280596160825,
+      "learning_rate": 1.067418756792713e-06,
+      "loss": 1.5578,
+      "step": 1174
+    },
+    {
+      "epoch": 1.598639455782313,
+      "grad_norm": 0.1203122510282115,
+      "learning_rate": 1.0604847119798212e-06,
+      "loss": 1.668,
+      "step": 1175
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.11094545607591003,
+      "learning_rate": 1.0535705899286292e-06,
+      "loss": 1.6685,
+      "step": 1176
+    },
+    {
+      "epoch": 1.601360544217687,
+      "grad_norm": 0.14762381520202184,
+      "learning_rate": 1.0466764256049632e-06,
+      "loss": 1.536,
+      "step": 1177
+    },
+    {
+      "epoch": 1.6027210884353742,
+      "grad_norm": 0.11571415514742311,
+      "learning_rate": 1.039802253873722e-06,
+      "loss": 1.7084,
+      "step": 1178
+    },
+    {
+      "epoch": 1.6040816326530614,
+      "grad_norm": 0.13034797124348868,
+      "learning_rate": 1.0329481094986921e-06,
+      "loss": 1.7737,
+      "step": 1179
+    },
+    {
+      "epoch": 1.6054421768707483,
+      "grad_norm": 0.1202706060362133,
+      "learning_rate": 1.0261140271423858e-06,
+      "loss": 1.6395,
+      "step": 1180
+    },
+    {
+      "epoch": 1.6068027210884352,
+      "grad_norm": 0.36927305650043224,
+      "learning_rate": 1.0193000413658538e-06,
+      "loss": 1.5686,
+      "step": 1181
+    },
+    {
+      "epoch": 1.6081632653061224,
+      "grad_norm": 0.11834428054470161,
+      "learning_rate": 1.0125061866285186e-06,
+      "loss": 1.5722,
+      "step": 1182
+    },
+    {
+      "epoch": 1.6095238095238096,
+      "grad_norm": 0.11481074155938689,
+      "learning_rate": 1.0057324972879933e-06,
+      "loss": 1.6238,
+      "step": 1183
+    },
+    {
+      "epoch": 1.6108843537414965,
+      "grad_norm": 0.22756149512694648,
+      "learning_rate": 9.989790075999145e-07,
+      "loss": 1.5934,
+      "step": 1184
+    },
+    {
+      "epoch": 1.6122448979591837,
+      "grad_norm": 0.11551066672925502,
+      "learning_rate": 9.922457517177648e-07,
+      "loss": 1.6947,
+      "step": 1185
+    },
+    {
+      "epoch": 1.6136054421768709,
+      "grad_norm": 0.17228114247643034,
+      "learning_rate": 9.855327636926992e-07,
+      "loss": 1.6467,
+      "step": 1186
+    },
+    {
+      "epoch": 1.6149659863945578,
+      "grad_norm": 0.13261272278043656,
+      "learning_rate": 9.78840077473377e-07,
+      "loss": 1.6139,
+      "step": 1187
+    },
+    {
+      "epoch": 1.6163265306122447,
+      "grad_norm": 0.14256320918658308,
+      "learning_rate": 9.721677269057882e-07,
+      "loss": 1.8835,
+      "step": 1188
+    },
+    {
+      "epoch": 1.6176870748299321,
+      "grad_norm": 0.13515325397992245,
+      "learning_rate": 9.655157457330816e-07,
+      "loss": 1.6236,
+      "step": 1189
+    },
+    {
+      "epoch": 1.619047619047619,
+      "grad_norm": 0.1213632301061294,
+      "learning_rate": 9.588841675953952e-07,
+      "loss": 1.6718,
+      "step": 1190
+    },
+    {
+      "epoch": 1.620408163265306,
+      "grad_norm": 0.12023247373916009,
+      "learning_rate": 9.522730260296836e-07,
+      "loss": 1.7103,
+      "step": 1191
+    },
+    {
+      "epoch": 1.6217687074829932,
+      "grad_norm": 0.12033848582541032,
+      "learning_rate": 9.45682354469552e-07,
+      "loss": 1.6756,
+      "step": 1192
+    },
+    {
+      "epoch": 1.6231292517006803,
+      "grad_norm": 0.15874394674606557,
+      "learning_rate": 9.391121862450858e-07,
+      "loss": 1.5691,
+      "step": 1193
+    },
+    {
+      "epoch": 1.6244897959183673,
+      "grad_norm": 0.16013714918796576,
+      "learning_rate": 9.325625545826822e-07,
+      "loss": 1.8024,
+      "step": 1194
+    },
+    {
+      "epoch": 1.6258503401360545,
+      "grad_norm": 0.27054822930134187,
+      "learning_rate": 9.260334926048787e-07,
+      "loss": 1.7083,
+      "step": 1195
+    },
+    {
+      "epoch": 1.6272108843537416,
+      "grad_norm": 0.11952723384185746,
+      "learning_rate": 9.195250333301913e-07,
+      "loss": 1.6187,
+      "step": 1196
+    },
+    {
+      "epoch": 1.6285714285714286,
+      "grad_norm": 0.1313632997033654,
+      "learning_rate": 9.130372096729446e-07,
+      "loss": 1.4098,
+      "step": 1197
+    },
+    {
+      "epoch": 1.6299319727891155,
+      "grad_norm": 0.12834344479681326,
+      "learning_rate": 9.06570054443105e-07,
+      "loss": 1.661,
+      "step": 1198
+    },
+    {
+      "epoch": 1.6312925170068027,
+      "grad_norm": 0.1090830592683463,
+      "learning_rate": 9.00123600346115e-07,
+      "loss": 1.6595,
+      "step": 1199
+    },
+    {
+      "epoch": 1.6326530612244898,
+      "grad_norm": 0.16043162392680574,
+      "learning_rate": 8.936978799827295e-07,
+      "loss": 1.7272,
+      "step": 1200
+    },
+    {
+      "epoch": 1.6340136054421768,
+      "grad_norm": 0.11867938317150763,
+      "learning_rate": 8.872929258488489e-07,
+      "loss": 1.827,
+      "step": 1201
+    },
+    {
+      "epoch": 1.635374149659864,
+      "grad_norm": 0.12762812337248913,
+      "learning_rate": 8.809087703353536e-07,
+      "loss": 1.6494,
+      "step": 1202
+    },
+    {
+      "epoch": 1.636734693877551,
+      "grad_norm": 0.1292715431110236,
+      "learning_rate": 8.74545445727944e-07,
+      "loss": 1.7392,
+      "step": 1203
+    },
+    {
+      "epoch": 1.638095238095238,
+      "grad_norm": 0.14455856202047512,
+      "learning_rate": 8.682029842069761e-07,
+      "loss": 1.5346,
+      "step": 1204
+    },
+    {
+      "epoch": 1.6394557823129252,
+      "grad_norm": 0.1074436319272245,
+      "learning_rate": 8.618814178472973e-07,
+      "loss": 1.7282,
+      "step": 1205
+    },
+    {
+      "epoch": 1.6408163265306124,
+      "grad_norm": 0.12115387945531858,
+      "learning_rate": 8.555807786180814e-07,
+      "loss": 1.7237,
+      "step": 1206
+    },
+    {
+      "epoch": 1.6421768707482993,
+      "grad_norm": 0.12805781303993397,
+      "learning_rate": 8.493010983826749e-07,
+      "loss": 1.634,
+      "step": 1207
+    },
+    {
+      "epoch": 1.6435374149659863,
+      "grad_norm": 0.12535314476626302,
+      "learning_rate": 8.430424088984295e-07,
+      "loss": 1.5882,
+      "step": 1208
+    },
+    {
+      "epoch": 1.6448979591836734,
+      "grad_norm": 0.12698774386082093,
+      "learning_rate": 8.368047418165437e-07,
+      "loss": 1.6526,
+      "step": 1209
+    },
+    {
+      "epoch": 1.6462585034013606,
+      "grad_norm": 0.13808871614579815,
+      "learning_rate": 8.305881286819029e-07,
+      "loss": 1.5524,
+      "step": 1210
+    },
+    {
+      "epoch": 1.6476190476190475,
+      "grad_norm": 0.11270379177937306,
+      "learning_rate": 8.243926009329156e-07,
+      "loss": 1.6414,
+      "step": 1211
+    },
+    {
+      "epoch": 1.6489795918367347,
+      "grad_norm": 0.1108976444955171,
+      "learning_rate": 8.18218189901363e-07,
+      "loss": 1.6002,
+      "step": 1212
+    },
+    {
+      "epoch": 1.6503401360544219,
+      "grad_norm": 0.12054751150197075,
+      "learning_rate": 8.120649268122333e-07,
+      "loss": 1.5422,
+      "step": 1213
+    },
+    {
+      "epoch": 1.6517006802721088,
+      "grad_norm": 0.14215811110011028,
+      "learning_rate": 8.059328427835672e-07,
+      "loss": 1.6454,
+      "step": 1214
+    },
+    {
+      "epoch": 1.6530612244897958,
+      "grad_norm": 0.11698885610808762,
+      "learning_rate": 7.998219688262971e-07,
+      "loss": 1.5852,
+      "step": 1215
+    },
+    {
+      "epoch": 1.6544217687074831,
+      "grad_norm": 0.1344113229524319,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 1.7622,
+      "step": 1216
+    },
+    {
+      "epoch": 1.65578231292517,
+      "grad_norm": 0.13434164059919554,
+      "learning_rate": 7.876639746332132e-07,
+      "loss": 1.7963,
+      "step": 1217
+    },
+    {
+      "epoch": 1.657142857142857,
+      "grad_norm": 0.17809396042944695,
+      "learning_rate": 7.816169158823295e-07,
+      "loss": 1.6858,
+      "step": 1218
+    },
+    {
+      "epoch": 1.6585034013605442,
+      "grad_norm": 0.13306970694049092,
+      "learning_rate": 7.755911901723917e-07,
+      "loss": 1.7431,
+      "step": 1219
+    },
+    {
+      "epoch": 1.6598639455782314,
+      "grad_norm": 0.11805560858596509,
+      "learning_rate": 7.695868279764623e-07,
+      "loss": 1.7348,
+      "step": 1220
+    },
+    {
+      "epoch": 1.6612244897959183,
+      "grad_norm": 0.138038505695151,
+      "learning_rate": 7.636038596595669e-07,
+      "loss": 1.6465,
+      "step": 1221
+    },
+    {
+      "epoch": 1.6625850340136055,
+      "grad_norm": 0.10405731691151054,
+      "learning_rate": 7.576423154785345e-07,
+      "loss": 1.6682,
+      "step": 1222
+    },
+    {
+      "epoch": 1.6639455782312926,
+      "grad_norm": 0.11496805932975722,
+      "learning_rate": 7.51702225581854e-07,
+      "loss": 1.5804,
+      "step": 1223
+    },
+    {
+      "epoch": 1.6653061224489796,
+      "grad_norm": 0.12087556838995649,
+      "learning_rate": 7.457836200095137e-07,
+      "loss": 1.6466,
+      "step": 1224
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.14333497087068797,
+      "learning_rate": 7.398865286928531e-07,
+      "loss": 1.7441,
+      "step": 1225
+    },
+    {
+      "epoch": 1.668027210884354,
+      "grad_norm": 0.13232481621976905,
+      "learning_rate": 7.34010981454411e-07,
+      "loss": 1.65,
+      "step": 1226
+    },
+    {
+      "epoch": 1.6693877551020408,
+      "grad_norm": 0.24932955649343116,
+      "learning_rate": 7.281570080077716e-07,
+      "loss": 1.4175,
+      "step": 1227
+    },
+    {
+      "epoch": 1.6707482993197278,
+      "grad_norm": 0.11924561800840275,
+      "learning_rate": 7.223246379574206e-07,
+      "loss": 1.6962,
+      "step": 1228
+    },
+    {
+      "epoch": 1.672108843537415,
+      "grad_norm": 0.12513856122351408,
+      "learning_rate": 7.165139007985899e-07,
+      "loss": 1.656,
+      "step": 1229
+    },
+    {
+      "epoch": 1.6734693877551021,
+      "grad_norm": 0.10354825149011261,
+      "learning_rate": 7.107248259171118e-07,
+      "loss": 1.7198,
+      "step": 1230
+    },
+    {
+      "epoch": 1.674829931972789,
+      "grad_norm": 0.1473977816911567,
+      "learning_rate": 7.049574425892663e-07,
+      "loss": 1.5774,
+      "step": 1231
+    },
+    {
+      "epoch": 1.6761904761904762,
+      "grad_norm": 0.12018009205318546,
+      "learning_rate": 6.992117799816383e-07,
+      "loss": 1.6315,
+      "step": 1232
+    },
+    {
+      "epoch": 1.6775510204081634,
+      "grad_norm": 0.36801389059595907,
+      "learning_rate": 6.934878671509665e-07,
+      "loss": 1.5693,
+      "step": 1233
+    },
+    {
+      "epoch": 1.6789115646258503,
+      "grad_norm": 0.12229393665572565,
+      "learning_rate": 6.87785733043998e-07,
+      "loss": 1.6481,
+      "step": 1234
+    },
+    {
+      "epoch": 1.6802721088435373,
+      "grad_norm": 0.1056291236057399,
+      "learning_rate": 6.821054064973415e-07,
+      "loss": 1.7811,
+      "step": 1235
+    },
+    {
+      "epoch": 1.6816326530612244,
+      "grad_norm": 0.1290627059039082,
+      "learning_rate": 6.764469162373183e-07,
+      "loss": 1.5862,
+      "step": 1236
+    },
+    {
+      "epoch": 1.6829931972789116,
+      "grad_norm": 0.14140062572587017,
+      "learning_rate": 6.708102908798269e-07,
+      "loss": 1.5937,
+      "step": 1237
+    },
+    {
+      "epoch": 1.6843537414965986,
+      "grad_norm": 0.12372660811883815,
+      "learning_rate": 6.651955589301839e-07,
+      "loss": 1.5442,
+      "step": 1238
+    },
+    {
+      "epoch": 1.6857142857142857,
+      "grad_norm": 0.11398935678427591,
+      "learning_rate": 6.596027487829915e-07,
+      "loss": 1.6802,
+      "step": 1239
+    },
+    {
+      "epoch": 1.6870748299319729,
+      "grad_norm": 0.1125362910385827,
+      "learning_rate": 6.540318887219899e-07,
+      "loss": 1.5756,
+      "step": 1240
+    },
+    {
+      "epoch": 1.6884353741496598,
+      "grad_norm": 0.12182979669358265,
+      "learning_rate": 6.484830069199132e-07,
+      "loss": 1.7316,
+      "step": 1241
+    },
+    {
+      "epoch": 1.689795918367347,
+      "grad_norm": 0.1757220644211557,
+      "learning_rate": 6.429561314383475e-07,
+      "loss": 1.5394,
+      "step": 1242
+    },
+    {
+      "epoch": 1.6911564625850342,
+      "grad_norm": 0.7508195823704957,
+      "learning_rate": 6.374512902275903e-07,
+      "loss": 1.5372,
+      "step": 1243
+    },
+    {
+      "epoch": 1.692517006802721,
+      "grad_norm": 0.11573367958160206,
+      "learning_rate": 6.319685111265078e-07,
+      "loss": 1.6907,
+      "step": 1244
+    },
+    {
+      "epoch": 1.693877551020408,
+      "grad_norm": 0.12247233986880039,
+      "learning_rate": 6.265078218623949e-07,
+      "loss": 1.6788,
+      "step": 1245
+    },
+    {
+      "epoch": 1.6952380952380952,
+      "grad_norm": 0.1420406063065376,
+      "learning_rate": 6.210692500508359e-07,
+      "loss": 1.7363,
+      "step": 1246
+    },
+    {
+      "epoch": 1.6965986394557824,
+      "grad_norm": 0.11985428532614184,
+      "learning_rate": 6.1565282319556e-07,
+      "loss": 1.5928,
+      "step": 1247
+    },
+    {
+      "epoch": 1.6979591836734693,
+      "grad_norm": 0.10678231200017475,
+      "learning_rate": 6.102585686883078e-07,
+      "loss": 1.7134,
+      "step": 1248
+    },
+    {
+      "epoch": 1.6993197278911565,
+      "grad_norm": 0.11899866385551988,
+      "learning_rate": 6.048865138086929e-07,
+      "loss": 1.5433,
+      "step": 1249
+    },
+    {
+      "epoch": 1.7006802721088436,
+      "grad_norm": 0.15494511895804758,
+      "learning_rate": 5.995366857240592e-07,
+      "loss": 1.6062,
+      "step": 1250
+    },
+    {
+      "epoch": 1.7020408163265306,
+      "grad_norm": 0.11026358843797994,
+      "learning_rate": 5.942091114893455e-07,
+      "loss": 1.7031,
+      "step": 1251
+    },
+    {
+      "epoch": 1.7034013605442175,
+      "grad_norm": 0.22132054192752112,
+      "learning_rate": 5.889038180469509e-07,
+      "loss": 1.6915,
+      "step": 1252
+    },
+    {
+      "epoch": 1.704761904761905,
+      "grad_norm": 0.1342507596457615,
+      "learning_rate": 5.836208322265991e-07,
+      "loss": 1.7164,
+      "step": 1253
+    },
+    {
+      "epoch": 1.7061224489795919,
+      "grad_norm": 0.12737785530895251,
+      "learning_rate": 5.783601807451949e-07,
+      "loss": 1.5374,
+      "step": 1254
+    },
+    {
+      "epoch": 1.7074829931972788,
+      "grad_norm": 0.12036712264849207,
+      "learning_rate": 5.73121890206701e-07,
+      "loss": 1.5903,
+      "step": 1255
+    },
+    {
+      "epoch": 1.708843537414966,
+      "grad_norm": 0.12472273002304834,
+      "learning_rate": 5.679059871019904e-07,
+      "loss": 1.5398,
+      "step": 1256
+    },
+    {
+      "epoch": 1.7102040816326531,
+      "grad_norm": 0.19389386529197455,
+      "learning_rate": 5.627124978087267e-07,
+      "loss": 1.6439,
+      "step": 1257
+    },
+    {
+      "epoch": 1.71156462585034,
+      "grad_norm": 0.1428822953499787,
+      "learning_rate": 5.575414485912173e-07,
+      "loss": 1.6785,
+      "step": 1258
+    },
+    {
+      "epoch": 1.7129251700680272,
+      "grad_norm": 0.12043634562801463,
+      "learning_rate": 5.523928656002897e-07,
+      "loss": 1.6439,
+      "step": 1259
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.19566443134501496,
+      "learning_rate": 5.472667748731552e-07,
+      "loss": 1.6816,
+      "step": 1260
+    },
+    {
+      "epoch": 1.7156462585034014,
+      "grad_norm": 0.12298818504704644,
+      "learning_rate": 5.421632023332779e-07,
+      "loss": 1.8452,
+      "step": 1261
+    },
+    {
+      "epoch": 1.7170068027210883,
+      "grad_norm": 0.12457908114663241,
+      "learning_rate": 5.370821737902455e-07,
+      "loss": 1.648,
+      "step": 1262
+    },
+    {
+      "epoch": 1.7183673469387755,
+      "grad_norm": 0.12905422383417112,
+      "learning_rate": 5.320237149396329e-07,
+      "loss": 1.7109,
+      "step": 1263
+    },
+    {
+      "epoch": 1.7197278911564626,
+      "grad_norm": 0.14174623921254864,
+      "learning_rate": 5.269878513628806e-07,
+      "loss": 1.6398,
+      "step": 1264
+    },
+    {
+      "epoch": 1.7210884353741496,
+      "grad_norm": 0.11019848785148706,
+      "learning_rate": 5.219746085271599e-07,
+      "loss": 1.6362,
+      "step": 1265
+    },
+    {
+      "epoch": 1.7224489795918367,
+      "grad_norm": 0.15843308364174372,
+      "learning_rate": 5.169840117852454e-07,
+      "loss": 1.6709,
+      "step": 1266
+    },
+    {
+      "epoch": 1.723809523809524,
+      "grad_norm": 0.12183022697550175,
+      "learning_rate": 5.120160863753859e-07,
+      "loss": 1.722,
+      "step": 1267
+    },
+    {
+      "epoch": 1.7251700680272108,
+      "grad_norm": 0.11328748286322111,
+      "learning_rate": 5.070708574211769e-07,
+      "loss": 1.7427,
+      "step": 1268
+    },
+    {
+      "epoch": 1.726530612244898,
+      "grad_norm": 0.5467870469075437,
+      "learning_rate": 5.0214834993144e-07,
+      "loss": 1.7051,
+      "step": 1269
+    },
+    {
+      "epoch": 1.7278911564625852,
+      "grad_norm": 0.1142321717186957,
+      "learning_rate": 4.972485888000822e-07,
+      "loss": 1.663,
+      "step": 1270
+    },
+    {
+      "epoch": 1.7292517006802721,
+      "grad_norm": 0.15669214354154648,
+      "learning_rate": 4.923715988059851e-07,
+      "loss": 1.7283,
+      "step": 1271
+    },
+    {
+      "epoch": 1.730612244897959,
+      "grad_norm": 0.12321158474736231,
+      "learning_rate": 4.875174046128684e-07,
+      "loss": 1.7658,
+      "step": 1272
+    },
+    {
+      "epoch": 1.7319727891156462,
+      "grad_norm": 0.12592275959820878,
+      "learning_rate": 4.826860307691749e-07,
+      "loss": 1.7346,
+      "step": 1273
+    },
+    {
+      "epoch": 1.7333333333333334,
+      "grad_norm": 0.11852385189098655,
+      "learning_rate": 4.778775017079357e-07,
+      "loss": 1.7388,
+      "step": 1274
+    },
+    {
+      "epoch": 1.7346938775510203,
+      "grad_norm": 0.11468824991687426,
+      "learning_rate": 4.730918417466551e-07,
+      "loss": 1.7385,
+      "step": 1275
+    },
+    {
+      "epoch": 1.7360544217687075,
+      "grad_norm": 0.15371819150117932,
+      "learning_rate": 4.683290750871855e-07,
+      "loss": 1.589,
+      "step": 1276
+    },
+    {
+      "epoch": 1.7374149659863947,
+      "grad_norm": 0.1990093146846139,
+      "learning_rate": 4.635892258156027e-07,
+      "loss": 1.6007,
+      "step": 1277
+    },
+    {
+      "epoch": 1.7387755102040816,
+      "grad_norm": 0.1663021965076431,
+      "learning_rate": 4.588723179020865e-07,
+      "loss": 1.5926,
+      "step": 1278
+    },
+    {
+      "epoch": 1.7401360544217686,
+      "grad_norm": 0.21861874038762558,
+      "learning_rate": 4.5417837520079667e-07,
+      "loss": 1.7584,
+      "step": 1279
+    },
+    {
+      "epoch": 1.741496598639456,
+      "grad_norm": 0.11717800217118578,
+      "learning_rate": 4.4950742144975525e-07,
+      "loss": 1.5905,
+      "step": 1280
+    },
+    {
+      "epoch": 1.7428571428571429,
+      "grad_norm": 0.12603146125560444,
+      "learning_rate": 4.4485948027072734e-07,
+      "loss": 1.6514,
+      "step": 1281
+    },
+    {
+      "epoch": 1.7442176870748298,
+      "grad_norm": 0.12095214576683595,
+      "learning_rate": 4.402345751690984e-07,
+      "loss": 1.6237,
+      "step": 1282
+    },
+    {
+      "epoch": 1.745578231292517,
+      "grad_norm": 0.12347022358642107,
+      "learning_rate": 4.3563272953375426e-07,
+      "loss": 1.6195,
+      "step": 1283
+    },
+    {
+      "epoch": 1.7469387755102042,
+      "grad_norm": 0.10437898422428948,
+      "learning_rate": 4.3105396663696896e-07,
+      "loss": 1.6093,
+      "step": 1284
+    },
+    {
+      "epoch": 1.748299319727891,
+      "grad_norm": 0.11098541618904957,
+      "learning_rate": 4.2649830963428227e-07,
+      "loss": 1.6999,
+      "step": 1285
+    },
+    {
+      "epoch": 1.7496598639455783,
+      "grad_norm": 0.23522787615940605,
+      "learning_rate": 4.2196578156438293e-07,
+      "loss": 1.6786,
+      "step": 1286
+    },
+    {
+      "epoch": 1.7510204081632654,
+      "grad_norm": 0.11988459025584766,
+      "learning_rate": 4.174564053489949e-07,
+      "loss": 1.5861,
+      "step": 1287
+    },
+    {
+      "epoch": 1.7523809523809524,
+      "grad_norm": 0.12717182374610603,
+      "learning_rate": 4.129702037927558e-07,
+      "loss": 1.5982,
+      "step": 1288
+    },
+    {
+      "epoch": 1.7523809523809524,
+      "eval_loss": 1.686065673828125,
+      "eval_runtime": 76.6201,
+      "eval_samples_per_second": 53.158,
+      "eval_steps_per_second": 6.656,
+      "step": 1288
+    },
+    {
+      "epoch": 1.7537414965986393,
+      "grad_norm": 0.13809405448384865,
+      "learning_rate": 4.0850719958311024e-07,
+      "loss": 1.5886,
+      "step": 1289
+    },
+    {
+      "epoch": 1.7551020408163265,
+      "grad_norm": 0.11274789712003737,
+      "learning_rate": 4.040674152901852e-07,
+      "loss": 1.694,
+      "step": 1290
+    },
+    {
+      "epoch": 1.7564625850340136,
+      "grad_norm": 0.1882765689460891,
+      "learning_rate": 3.9965087336668485e-07,
+      "loss": 1.5427,
+      "step": 1291
+    },
+    {
+      "epoch": 1.7578231292517006,
+      "grad_norm": 0.14271738534965617,
+      "learning_rate": 3.9525759614776894e-07,
+      "loss": 1.5454,
+      "step": 1292
+    },
+    {
+      "epoch": 1.7591836734693878,
+      "grad_norm": 0.1405107274242078,
+      "learning_rate": 3.9088760585094976e-07,
+      "loss": 1.5758,
+      "step": 1293
+    },
+    {
+      "epoch": 1.760544217687075,
+      "grad_norm": 0.13748980285277887,
+      "learning_rate": 3.8654092457596714e-07,
+      "loss": 1.7284,
+      "step": 1294
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 0.13551778981512638,
+      "learning_rate": 3.822175743046874e-07,
+      "loss": 1.4861,
+      "step": 1295
+    },
+    {
+      "epoch": 1.763265306122449,
+      "grad_norm": 0.12395464181970267,
+      "learning_rate": 3.7791757690098797e-07,
+      "loss": 1.6397,
+      "step": 1296
+    },
+    {
+      "epoch": 1.7646258503401362,
+      "grad_norm": 0.1305188086923226,
+      "learning_rate": 3.7364095411064614e-07,
+      "loss": 1.6356,
+      "step": 1297
+    },
+    {
+      "epoch": 1.7659863945578231,
+      "grad_norm": 0.11815051876585439,
+      "learning_rate": 3.693877275612312e-07,
+      "loss": 1.6778,
+      "step": 1298
+    },
+    {
+      "epoch": 1.76734693877551,
+      "grad_norm": 0.13907578783569868,
+      "learning_rate": 3.6515791876199136e-07,
+      "loss": 1.6404,
+      "step": 1299
+    },
+    {
+      "epoch": 1.7687074829931972,
+      "grad_norm": 0.11634068765275582,
+      "learning_rate": 3.6095154910375076e-07,
+      "loss": 1.7575,
+      "step": 1300
+    },
+    {
+      "epoch": 1.7700680272108844,
+      "grad_norm": 0.12322320540073207,
+      "learning_rate": 3.567686398587955e-07,
+      "loss": 1.7204,
+      "step": 1301
+    },
+    {
+      "epoch": 1.7714285714285714,
+      "grad_norm": 0.2519201944711608,
+      "learning_rate": 3.5260921218077116e-07,
+      "loss": 1.6642,
+      "step": 1302
+    },
+    {
+      "epoch": 1.7727891156462585,
+      "grad_norm": 0.12072706338657095,
+      "learning_rate": 3.484732871045704e-07,
+      "loss": 1.6003,
+      "step": 1303
+    },
+    {
+      "epoch": 1.7741496598639457,
+      "grad_norm": 0.12368254206995616,
+      "learning_rate": 3.443608855462294e-07,
+      "loss": 1.4418,
+      "step": 1304
+    },
+    {
+      "epoch": 1.7755102040816326,
+      "grad_norm": 0.1349862699749685,
+      "learning_rate": 3.402720283028277e-07,
+      "loss": 1.4863,
+      "step": 1305
+    },
+    {
+      "epoch": 1.7768707482993196,
+      "grad_norm": 0.1586268302178325,
+      "learning_rate": 3.3620673605236907e-07,
+      "loss": 1.6977,
+      "step": 1306
+    },
+    {
+      "epoch": 1.778231292517007,
+      "grad_norm": 0.10301426967234653,
+      "learning_rate": 3.3216502935369265e-07,
+      "loss": 1.7133,
+      "step": 1307
+    },
+    {
+      "epoch": 1.779591836734694,
+      "grad_norm": 0.1074501899935655,
+      "learning_rate": 3.2814692864635513e-07,
+      "loss": 1.6611,
+      "step": 1308
+    },
+    {
+      "epoch": 1.7809523809523808,
+      "grad_norm": 0.14501762775546292,
+      "learning_rate": 3.2415245425054087e-07,
+      "loss": 1.5968,
+      "step": 1309
+    },
+    {
+      "epoch": 1.782312925170068,
+      "grad_norm": 0.11188619893577698,
+      "learning_rate": 3.201816263669461e-07,
+      "loss": 1.4696,
+      "step": 1310
+    },
+    {
+      "epoch": 1.7836734693877552,
+      "grad_norm": 0.1444196373172706,
+      "learning_rate": 3.162344650766874e-07,
+      "loss": 1.6787,
+      "step": 1311
+    },
+    {
+      "epoch": 1.7850340136054421,
+      "grad_norm": 0.11758282717179587,
+      "learning_rate": 3.123109903411903e-07,
+      "loss": 1.6935,
+      "step": 1312
+    },
+    {
+      "epoch": 1.7863945578231293,
+      "grad_norm": 0.11743931898274525,
+      "learning_rate": 3.0841122200210014e-07,
+      "loss": 1.6681,
+      "step": 1313
+    },
+    {
+      "epoch": 1.7877551020408164,
+      "grad_norm": 0.15092126786575402,
+      "learning_rate": 3.045351797811713e-07,
+      "loss": 1.7548,
+      "step": 1314
+    },
+    {
+      "epoch": 1.7891156462585034,
+      "grad_norm": 0.13374065025706797,
+      "learning_rate": 3.006828832801706e-07,
+      "loss": 1.6463,
+      "step": 1315
+    },
+    {
+      "epoch": 1.7904761904761903,
+      "grad_norm": 0.1070307983897091,
+      "learning_rate": 2.9685435198078095e-07,
+      "loss": 1.6918,
+      "step": 1316
+    },
+    {
+      "epoch": 1.7918367346938775,
+      "grad_norm": 0.1043359426709795,
+      "learning_rate": 2.9304960524450034e-07,
+      "loss": 1.8017,
+      "step": 1317
+    },
+    {
+      "epoch": 1.7931972789115647,
+      "grad_norm": 0.12298508287191963,
+      "learning_rate": 2.892686623125446e-07,
+      "loss": 1.3275,
+      "step": 1318
+    },
+    {
+      "epoch": 1.7945578231292516,
+      "grad_norm": 0.12081576860619693,
+      "learning_rate": 2.855115423057492e-07,
+      "loss": 1.7012,
+      "step": 1319
+    },
+    {
+      "epoch": 1.7959183673469388,
+      "grad_norm": 0.11298191984110575,
+      "learning_rate": 2.817782642244732e-07,
+      "loss": 1.7513,
+      "step": 1320
+    },
+    {
+      "epoch": 1.797278911564626,
+      "grad_norm": 0.10476335704776595,
+      "learning_rate": 2.780688469485038e-07,
+      "loss": 1.7937,
+      "step": 1321
+    },
+    {
+      "epoch": 1.7986394557823129,
+      "grad_norm": 0.13240769456180743,
+      "learning_rate": 2.743833092369608e-07,
+      "loss": 1.5026,
+      "step": 1322
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.1349026266688947,
+      "learning_rate": 2.7072166972820005e-07,
+      "loss": 1.478,
+      "step": 1323
+    },
+    {
+      "epoch": 1.8013605442176872,
+      "grad_norm": 0.1581587709409474,
+      "learning_rate": 2.6708394693971963e-07,
+      "loss": 1.6953,
+      "step": 1324
+    },
+    {
+      "epoch": 1.8027210884353742,
+      "grad_norm": 0.1413036863393829,
+      "learning_rate": 2.6347015926807053e-07,
+      "loss": 1.7564,
+      "step": 1325
+    },
+    {
+      "epoch": 1.804081632653061,
+      "grad_norm": 0.10947438586849194,
+      "learning_rate": 2.5988032498875604e-07,
+      "loss": 1.693,
+      "step": 1326
+    },
+    {
+      "epoch": 1.8054421768707483,
+      "grad_norm": 0.18719949874204633,
+      "learning_rate": 2.563144622561453e-07,
+      "loss": 1.5172,
+      "step": 1327
+    },
+    {
+      "epoch": 1.8068027210884354,
+      "grad_norm": 0.12603171462141458,
+      "learning_rate": 2.527725891033772e-07,
+      "loss": 1.5421,
+      "step": 1328
+    },
+    {
+      "epoch": 1.8081632653061224,
+      "grad_norm": 0.11373462842692589,
+      "learning_rate": 2.49254723442276e-07,
+      "loss": 1.6575,
+      "step": 1329
+    },
+    {
+      "epoch": 1.8095238095238095,
+      "grad_norm": 0.12320391589520092,
+      "learning_rate": 2.4576088306325184e-07,
+      "loss": 1.6266,
+      "step": 1330
+    },
+    {
+      "epoch": 1.8108843537414967,
+      "grad_norm": 0.1362145220231982,
+      "learning_rate": 2.422910856352173e-07,
+      "loss": 1.6477,
+      "step": 1331
+    },
+    {
+      "epoch": 1.8122448979591836,
+      "grad_norm": 0.12607909355245228,
+      "learning_rate": 2.3884534870549534e-07,
+      "loss": 1.6541,
+      "step": 1332
+    },
+    {
+      "epoch": 1.8136054421768706,
+      "grad_norm": 0.12201919923215693,
+      "learning_rate": 2.354236896997314e-07,
+      "loss": 1.6405,
+      "step": 1333
+    },
+    {
+      "epoch": 1.814965986394558,
+      "grad_norm": 0.1475512645133478,
+      "learning_rate": 2.3202612592180485e-07,
+      "loss": 1.5981,
+      "step": 1334
+    },
+    {
+      "epoch": 1.816326530612245,
+      "grad_norm": 0.1410369523336787,
+      "learning_rate": 2.2865267455374042e-07,
+      "loss": 1.6012,
+      "step": 1335
+    },
+    {
+      "epoch": 1.8176870748299319,
+      "grad_norm": 0.14726614058677065,
+      "learning_rate": 2.2530335265562475e-07,
+      "loss": 1.5929,
+      "step": 1336
+    },
+    {
+      "epoch": 1.819047619047619,
+      "grad_norm": 0.11101939816547764,
+      "learning_rate": 2.2197817716551663e-07,
+      "loss": 1.7944,
+      "step": 1337
+    },
+    {
+      "epoch": 1.8204081632653062,
+      "grad_norm": 0.31120839304837183,
+      "learning_rate": 2.1867716489936297e-07,
+      "loss": 1.7441,
+      "step": 1338
+    },
+    {
+      "epoch": 1.8217687074829931,
+      "grad_norm": 0.1087680907822782,
+      "learning_rate": 2.1540033255091187e-07,
+      "loss": 1.7397,
+      "step": 1339
+    },
+    {
+      "epoch": 1.8231292517006803,
+      "grad_norm": 0.15824282985610105,
+      "learning_rate": 2.1214769669163137e-07,
+      "loss": 1.5673,
+      "step": 1340
+    },
+    {
+      "epoch": 1.8244897959183675,
+      "grad_norm": 0.11523860122450502,
+      "learning_rate": 2.0891927377062526e-07,
+      "loss": 1.7194,
+      "step": 1341
+    },
+    {
+      "epoch": 1.8258503401360544,
+      "grad_norm": 0.16694763381862118,
+      "learning_rate": 2.0571508011454467e-07,
+      "loss": 1.7396,
+      "step": 1342
+    },
+    {
+      "epoch": 1.8272108843537413,
+      "grad_norm": 0.12107304038026233,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 1.5903,
+      "step": 1343
+    },
+    {
+      "epoch": 1.8285714285714287,
+      "grad_norm": 0.12306189922695104,
+      "learning_rate": 1.993794452910386e-07,
+      "loss": 1.6681,
+      "step": 1344
+    },
+    {
+      "epoch": 1.8299319727891157,
+      "grad_norm": 0.13264203303883543,
+      "learning_rate": 1.962480361639363e-07,
+      "loss": 1.5602,
+      "step": 1345
+    },
+    {
+      "epoch": 1.8312925170068026,
+      "grad_norm": 0.19210662239119836,
+      "learning_rate": 1.9314092038224376e-07,
+      "loss": 1.7383,
+      "step": 1346
+    },
+    {
+      "epoch": 1.8326530612244898,
+      "grad_norm": 0.11196653339216565,
+      "learning_rate": 1.9005811365914561e-07,
+      "loss": 1.6738,
+      "step": 1347
+    },
+    {
+      "epoch": 1.834013605442177,
+      "grad_norm": 0.13383855955903828,
+      "learning_rate": 1.8699963158489097e-07,
+      "loss": 1.6137,
+      "step": 1348
+    },
+    {
+      "epoch": 1.835374149659864,
+      "grad_norm": 0.13158068933328537,
+      "learning_rate": 1.8396548962671456e-07,
+      "loss": 1.7042,
+      "step": 1349
+    },
+    {
+      "epoch": 1.836734693877551,
+      "grad_norm": 0.11719394942534897,
+      "learning_rate": 1.8095570312876066e-07,
+      "loss": 1.5956,
+      "step": 1350
+    },
+    {
+      "epoch": 1.8380952380952382,
+      "grad_norm": 0.11938431133510057,
+      "learning_rate": 1.779702873120015e-07,
+      "loss": 1.665,
+      "step": 1351
+    },
+    {
+      "epoch": 1.8394557823129252,
+      "grad_norm": 0.12054936992306543,
+      "learning_rate": 1.7500925727416517e-07,
+      "loss": 1.7007,
+      "step": 1352
+    },
+    {
+      "epoch": 1.8408163265306121,
+      "grad_norm": 0.12302278736973048,
+      "learning_rate": 1.7207262798965617e-07,
+      "loss": 1.8256,
+      "step": 1353
+    },
+    {
+      "epoch": 1.8421768707482993,
+      "grad_norm": 0.20874731105186642,
+      "learning_rate": 1.691604143094816e-07,
+      "loss": 1.6506,
+      "step": 1354
+    },
+    {
+      "epoch": 1.8435374149659864,
+      "grad_norm": 0.11883930933955864,
+      "learning_rate": 1.6627263096117286e-07,
+      "loss": 1.4408,
+      "step": 1355
+    },
+    {
+      "epoch": 1.8448979591836734,
+      "grad_norm": 0.13925197256417077,
+      "learning_rate": 1.6340929254871406e-07,
+      "loss": 1.597,
+      "step": 1356
+    },
+    {
+      "epoch": 1.8462585034013606,
+      "grad_norm": 0.12520385420034488,
+      "learning_rate": 1.6057041355246827e-07,
+      "loss": 1.7058,
+      "step": 1357
+    },
+    {
+      "epoch": 1.8476190476190477,
+      "grad_norm": 0.14534052106943013,
+      "learning_rate": 1.5775600832910187e-07,
+      "loss": 1.7239,
+      "step": 1358
+    },
+    {
+      "epoch": 1.8489795918367347,
+      "grad_norm": 0.10822664134568488,
+      "learning_rate": 1.5496609111151583e-07,
+      "loss": 1.6079,
+      "step": 1359
+    },
+    {
+      "epoch": 1.8503401360544216,
+      "grad_norm": 0.12860377257334074,
+      "learning_rate": 1.5220067600876686e-07,
+      "loss": 1.605,
+      "step": 1360
+    },
+    {
+      "epoch": 1.851700680272109,
+      "grad_norm": 0.12116010186228711,
+      "learning_rate": 1.4945977700600634e-07,
+      "loss": 1.5666,
+      "step": 1361
+    },
+    {
+      "epoch": 1.853061224489796,
+      "grad_norm": 0.15034373846159615,
+      "learning_rate": 1.4674340796439924e-07,
+      "loss": 1.5519,
+      "step": 1362
+    },
+    {
+      "epoch": 1.8544217687074829,
+      "grad_norm": 0.1249592480463045,
+      "learning_rate": 1.4405158262106144e-07,
+      "loss": 1.6923,
+      "step": 1363
+    },
+    {
+      "epoch": 1.85578231292517,
+      "grad_norm": 0.12093711468318252,
+      "learning_rate": 1.4138431458898372e-07,
+      "loss": 1.5712,
+      "step": 1364
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.1535628658201845,
+      "learning_rate": 1.3874161735697166e-07,
+      "loss": 1.5705,
+      "step": 1365
+    },
+    {
+      "epoch": 1.8585034013605441,
+      "grad_norm": 0.11441660994747492,
+      "learning_rate": 1.3612350428956865e-07,
+      "loss": 1.7873,
+      "step": 1366
+    },
+    {
+      "epoch": 1.8598639455782313,
+      "grad_norm": 0.1334231424049127,
+      "learning_rate": 1.3352998862699252e-07,
+      "loss": 1.6774,
+      "step": 1367
+    },
+    {
+      "epoch": 1.8612244897959185,
+      "grad_norm": 0.12313352697990386,
+      "learning_rate": 1.3096108348506942e-07,
+      "loss": 1.6795,
+      "step": 1368
+    },
+    {
+      "epoch": 1.8625850340136054,
+      "grad_norm": 0.13161417747426524,
+      "learning_rate": 1.2841680185516515e-07,
+      "loss": 1.6117,
+      "step": 1369
+    },
+    {
+      "epoch": 1.8639455782312924,
+      "grad_norm": 0.11729319664592838,
+      "learning_rate": 1.2589715660412117e-07,
+      "loss": 1.7447,
+      "step": 1370
+    },
+    {
+      "epoch": 1.8653061224489798,
+      "grad_norm": 0.17489311720035736,
+      "learning_rate": 1.2340216047418697e-07,
+      "loss": 1.6488,
+      "step": 1371
+    },
+    {
+      "epoch": 1.8666666666666667,
+      "grad_norm": 0.1323254687842618,
+      "learning_rate": 1.2093182608296007e-07,
+      "loss": 1.5905,
+      "step": 1372
+    },
+    {
+      "epoch": 1.8680272108843536,
+      "grad_norm": 0.13009049948669843,
+      "learning_rate": 1.1848616592331718e-07,
+      "loss": 1.7288,
+      "step": 1373
+    },
+    {
+      "epoch": 1.8693877551020408,
+      "grad_norm": 0.13326010230560859,
+      "learning_rate": 1.1606519236335601e-07,
+      "loss": 1.4685,
+      "step": 1374
+    },
+    {
+      "epoch": 1.870748299319728,
+      "grad_norm": 0.13195963848274078,
+      "learning_rate": 1.1366891764632792e-07,
+      "loss": 1.6705,
+      "step": 1375
+    },
+    {
+      "epoch": 1.872108843537415,
+      "grad_norm": 0.20403157309863337,
+      "learning_rate": 1.1129735389057872e-07,
+      "loss": 1.5451,
+      "step": 1376
+    },
+    {
+      "epoch": 1.873469387755102,
+      "grad_norm": 0.13333819468646058,
+      "learning_rate": 1.0895051308948857e-07,
+      "loss": 1.6161,
+      "step": 1377
+    },
+    {
+      "epoch": 1.8748299319727892,
+      "grad_norm": 0.11437641155105588,
+      "learning_rate": 1.0662840711140765e-07,
+      "loss": 1.6297,
+      "step": 1378
+    },
+    {
+      "epoch": 1.8761904761904762,
+      "grad_norm": 0.14869492133046525,
+      "learning_rate": 1.0433104769959901e-07,
+      "loss": 1.6101,
+      "step": 1379
+    },
+    {
+      "epoch": 1.8775510204081631,
+      "grad_norm": 0.10691630289178386,
+      "learning_rate": 1.0205844647217745e-07,
+      "loss": 1.7722,
+      "step": 1380
+    },
+    {
+      "epoch": 1.8789115646258503,
+      "grad_norm": 0.12576056527826357,
+      "learning_rate": 9.981061492205457e-08,
+      "loss": 1.6726,
+      "step": 1381
+    },
+    {
+      "epoch": 1.8802721088435375,
+      "grad_norm": 0.14899620665692015,
+      "learning_rate": 9.758756441687333e-08,
+      "loss": 1.7,
+      "step": 1382
+    },
+    {
+      "epoch": 1.8816326530612244,
+      "grad_norm": 0.11208257992745001,
+      "learning_rate": 9.538930619895914e-08,
+      "loss": 1.7984,
+      "step": 1383
+    },
+    {
+      "epoch": 1.8829931972789116,
+      "grad_norm": 0.1224966922554696,
+      "learning_rate": 9.321585138525546e-08,
+      "loss": 1.6945,
+      "step": 1384
+    },
+    {
+      "epoch": 1.8843537414965987,
+      "grad_norm": 0.11370992913550665,
+      "learning_rate": 9.106721096727334e-08,
+      "loss": 1.6014,
+      "step": 1385
+    },
+    {
+      "epoch": 1.8857142857142857,
+      "grad_norm": 0.1374350456716246,
+      "learning_rate": 8.89433958110325e-08,
+      "loss": 1.5331,
+      "step": 1386
+    },
+    {
+      "epoch": 1.8870748299319728,
+      "grad_norm": 0.1295417008752131,
+      "learning_rate": 8.684441665700704e-08,
+      "loss": 1.6951,
+      "step": 1387
+    },
+    {
+      "epoch": 1.88843537414966,
+      "grad_norm": 0.14326716925245933,
+      "learning_rate": 8.477028412007204e-08,
+      "loss": 1.4568,
+      "step": 1388
+    },
+    {
+      "epoch": 1.889795918367347,
+      "grad_norm": 0.11941708812661872,
+      "learning_rate": 8.272100868944865e-08,
+      "loss": 1.6917,
+      "step": 1389
+    },
+    {
+      "epoch": 1.891156462585034,
+      "grad_norm": 0.11037035678246186,
+      "learning_rate": 8.069660072865304e-08,
+      "loss": 1.7508,
+      "step": 1390
+    },
+    {
+      "epoch": 1.892517006802721,
+      "grad_norm": 0.12587422498732503,
+      "learning_rate": 7.869707047543973e-08,
+      "loss": 1.7974,
+      "step": 1391
+    },
+    {
+      "epoch": 1.8938775510204082,
+      "grad_norm": 0.1266780218730049,
+      "learning_rate": 7.6722428041755e-08,
+      "loss": 1.953,
+      "step": 1392
+    },
+    {
+      "epoch": 1.8952380952380952,
+      "grad_norm": 0.12200593740003406,
+      "learning_rate": 7.47726834136836e-08,
+      "loss": 1.7449,
+      "step": 1393
+    },
+    {
+      "epoch": 1.8965986394557823,
+      "grad_norm": 0.15472206202543426,
+      "learning_rate": 7.284784645139653e-08,
+      "loss": 1.5177,
+      "step": 1394
+    },
+    {
+      "epoch": 1.8979591836734695,
+      "grad_norm": 0.1624808633869621,
+      "learning_rate": 7.094792688910446e-08,
+      "loss": 1.4361,
+      "step": 1395
+    },
+    {
+      "epoch": 1.8993197278911564,
+      "grad_norm": 0.11848244574224435,
+      "learning_rate": 6.907293433500494e-08,
+      "loss": 1.8216,
+      "step": 1396
+    },
+    {
+      "epoch": 1.9006802721088434,
+      "grad_norm": 0.16167022314859605,
+      "learning_rate": 6.722287827123697e-08,
+      "loss": 1.6308,
+      "step": 1397
+    },
+    {
+      "epoch": 1.9020408163265308,
+      "grad_norm": 0.1190014648907852,
+      "learning_rate": 6.539776805383146e-08,
+      "loss": 1.5514,
+      "step": 1398
+    },
+    {
+      "epoch": 1.9034013605442177,
+      "grad_norm": 0.14368372012309139,
+      "learning_rate": 6.359761291266365e-08,
+      "loss": 1.6309,
+      "step": 1399
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.11701809111487846,
+      "learning_rate": 6.182242195140686e-08,
+      "loss": 1.5709,
+      "step": 1400
+    },
+    {
+      "epoch": 1.9061224489795918,
+      "grad_norm": 0.10607705943146,
+      "learning_rate": 6.007220414748772e-08,
+      "loss": 1.6508,
+      "step": 1401
+    },
+    {
+      "epoch": 1.907482993197279,
+      "grad_norm": 0.12434317488105485,
+      "learning_rate": 5.8346968352038794e-08,
+      "loss": 1.6759,
+      "step": 1402
+    },
+    {
+      "epoch": 1.908843537414966,
+      "grad_norm": 0.12404880236130861,
+      "learning_rate": 5.6646723289853767e-08,
+      "loss": 1.5871,
+      "step": 1403
+    },
+    {
+      "epoch": 1.910204081632653,
+      "grad_norm": 0.1032281528541635,
+      "learning_rate": 5.4971477559346286e-08,
+      "loss": 1.8326,
+      "step": 1404
+    },
+    {
+      "epoch": 1.9115646258503403,
+      "grad_norm": 0.13405302158392976,
+      "learning_rate": 5.33212396325028e-08,
+      "loss": 1.6984,
+      "step": 1405
+    },
+    {
+      "epoch": 1.9129251700680272,
+      "grad_norm": 0.10242389969549208,
+      "learning_rate": 5.1696017854841463e-08,
+      "loss": 1.8544,
+      "step": 1406
+    },
+    {
+      "epoch": 1.9142857142857141,
+      "grad_norm": 0.1445888697947125,
+      "learning_rate": 5.0095820445369424e-08,
+      "loss": 1.5013,
+      "step": 1407
+    },
+    {
+      "epoch": 1.9156462585034013,
+      "grad_norm": 0.20630737260948012,
+      "learning_rate": 4.852065549654228e-08,
+      "loss": 1.4415,
+      "step": 1408
+    },
+    {
+      "epoch": 1.9170068027210885,
+      "grad_norm": 0.1271819648110208,
+      "learning_rate": 4.697053097422299e-08,
+      "loss": 1.712,
+      "step": 1409
+    },
+    {
+      "epoch": 1.9183673469387754,
+      "grad_norm": 0.12993476660803524,
+      "learning_rate": 4.544545471763917e-08,
+      "loss": 1.7681,
+      "step": 1410
+    },
+    {
+      "epoch": 1.9197278911564626,
+      "grad_norm": 0.12496814049626584,
+      "learning_rate": 4.394543443934751e-08,
+      "loss": 1.7471,
+      "step": 1411
+    },
+    {
+      "epoch": 1.9210884353741497,
+      "grad_norm": 0.13393563592633062,
+      "learning_rate": 4.247047772519108e-08,
+      "loss": 1.4003,
+      "step": 1412
+    },
+    {
+      "epoch": 1.9224489795918367,
+      "grad_norm": 0.11852283012443121,
+      "learning_rate": 4.102059203426267e-08,
+      "loss": 1.7373,
+      "step": 1413
+    },
+    {
+      "epoch": 1.9238095238095239,
+      "grad_norm": 0.15124695275803585,
+      "learning_rate": 3.95957846988676e-08,
+      "loss": 1.8101,
+      "step": 1414
+    },
+    {
+      "epoch": 1.925170068027211,
+      "grad_norm": 0.1329717102509104,
+      "learning_rate": 3.819606292448541e-08,
+      "loss": 1.6143,
+      "step": 1415
+    },
+    {
+      "epoch": 1.926530612244898,
+      "grad_norm": 0.1287044226542165,
+      "learning_rate": 3.682143378973269e-08,
+      "loss": 1.5269,
+      "step": 1416
+    },
+    {
+      "epoch": 1.927891156462585,
+      "grad_norm": 0.13046094115226015,
+      "learning_rate": 3.5471904246331955e-08,
+      "loss": 1.7016,
+      "step": 1417
+    },
+    {
+      "epoch": 1.929251700680272,
+      "grad_norm": 0.1464904620501286,
+      "learning_rate": 3.414748111907007e-08,
+      "loss": 1.582,
+      "step": 1418
+    },
+    {
+      "epoch": 1.9306122448979592,
+      "grad_norm": 0.11902086875565884,
+      "learning_rate": 3.2848171105766545e-08,
+      "loss": 1.6504,
+      "step": 1419
+    },
+    {
+      "epoch": 1.9319727891156462,
+      "grad_norm": 0.1112077111298243,
+      "learning_rate": 3.15739807772425e-08,
+      "loss": 1.7393,
+      "step": 1420
+    },
+    {
+      "epoch": 1.9333333333333333,
+      "grad_norm": 0.12536056522513223,
+      "learning_rate": 3.032491657728176e-08,
+      "loss": 1.6239,
+      "step": 1421
+    },
+    {
+      "epoch": 1.9346938775510205,
+      "grad_norm": 0.120512032686707,
+      "learning_rate": 2.9100984822603705e-08,
+      "loss": 1.6468,
+      "step": 1422
+    },
+    {
+      "epoch": 1.9360544217687075,
+      "grad_norm": 0.14047085918823868,
+      "learning_rate": 2.7902191702827152e-08,
+      "loss": 1.7043,
+      "step": 1423
+    },
+    {
+      "epoch": 1.9374149659863944,
+      "grad_norm": 0.18275770385842988,
+      "learning_rate": 2.67285432804415e-08,
+      "loss": 1.6855,
+      "step": 1424
+    },
+    {
+      "epoch": 1.9387755102040818,
+      "grad_norm": 0.15317594155825304,
+      "learning_rate": 2.558004549077564e-08,
+      "loss": 1.6121,
+      "step": 1425
+    },
+    {
+      "epoch": 1.9401360544217687,
+      "grad_norm": 0.15316596709144686,
+      "learning_rate": 2.4456704141967437e-08,
+      "loss": 1.3756,
+      "step": 1426
+    },
+    {
+      "epoch": 1.9414965986394557,
+      "grad_norm": 0.12215537441906983,
+      "learning_rate": 2.335852491493429e-08,
+      "loss": 1.5432,
+      "step": 1427
+    },
+    {
+      "epoch": 1.9428571428571428,
+      "grad_norm": 0.11247818731447141,
+      "learning_rate": 2.2285513363344833e-08,
+      "loss": 1.4894,
+      "step": 1428
+    },
+    {
+      "epoch": 1.94421768707483,
+      "grad_norm": 0.14019650679504936,
+      "learning_rate": 2.1237674913591744e-08,
+      "loss": 1.6705,
+      "step": 1429
+    },
+    {
+      "epoch": 1.945578231292517,
+      "grad_norm": 0.1282593607045711,
+      "learning_rate": 2.0215014864761185e-08,
+      "loss": 1.5348,
+      "step": 1430
+    },
+    {
+      "epoch": 1.9469387755102041,
+      "grad_norm": 0.12284414079551015,
+      "learning_rate": 1.9217538388610625e-08,
+      "loss": 1.6945,
+      "step": 1431
+    },
+    {
+      "epoch": 1.9482993197278913,
+      "grad_norm": 0.11971853998705126,
+      "learning_rate": 1.824525052953774e-08,
+      "loss": 1.6959,
+      "step": 1432
+    },
+    {
+      "epoch": 1.9496598639455782,
+      "grad_norm": 0.12822616099558357,
+      "learning_rate": 1.7298156204559324e-08,
+      "loss": 1.7549,
+      "step": 1433
+    },
+    {
+      "epoch": 1.9510204081632652,
+      "grad_norm": 0.12499088858434286,
+      "learning_rate": 1.6376260203282422e-08,
+      "loss": 1.5989,
+      "step": 1434
+    },
+    {
+      "epoch": 1.9523809523809523,
+      "grad_norm": 0.12657981100016752,
+      "learning_rate": 1.5479567187884346e-08,
+      "loss": 1.6822,
+      "step": 1435
+    },
+    {
+      "epoch": 1.9537414965986395,
+      "grad_norm": 0.10376271150021693,
+      "learning_rate": 1.460808169308492e-08,
+      "loss": 1.861,
+      "step": 1436
+    },
+    {
+      "epoch": 1.9551020408163264,
+      "grad_norm": 0.12572469181134285,
+      "learning_rate": 1.3761808126126486e-08,
+      "loss": 1.7238,
+      "step": 1437
+    },
+    {
+      "epoch": 1.9564625850340136,
+      "grad_norm": 0.19569694241617253,
+      "learning_rate": 1.2940750766751164e-08,
+      "loss": 1.6449,
+      "step": 1438
+    },
+    {
+      "epoch": 1.9578231292517008,
+      "grad_norm": 0.133934432892983,
+      "learning_rate": 1.2144913767176968e-08,
+      "loss": 1.6028,
+      "step": 1439
+    },
+    {
+      "epoch": 1.9591836734693877,
+      "grad_norm": 0.276184773751762,
+      "learning_rate": 1.1374301152079491e-08,
+      "loss": 1.7483,
+      "step": 1440
+    },
+    {
+      "epoch": 1.9605442176870749,
+      "grad_norm": 0.12157965543272396,
+      "learning_rate": 1.0628916818571366e-08,
+      "loss": 1.7897,
+      "step": 1441
+    },
+    {
+      "epoch": 1.961904761904762,
+      "grad_norm": 0.12393150387814673,
+      "learning_rate": 9.90876453618006e-09,
+      "loss": 1.6346,
+      "step": 1442
+    },
+    {
+      "epoch": 1.963265306122449,
+      "grad_norm": 0.16519695397062947,
+      "learning_rate": 9.213847946832333e-09,
+      "loss": 1.6296,
+      "step": 1443
+    },
+    {
+      "epoch": 1.964625850340136,
+      "grad_norm": 0.16082589408015513,
+      "learning_rate": 8.544170564831478e-09,
+      "loss": 1.5966,
+      "step": 1444
+    },
+    {
+      "epoch": 1.965986394557823,
+      "grad_norm": 0.12676850389062208,
+      "learning_rate": 7.899735776845108e-09,
+      "loss": 1.5393,
+      "step": 1445
+    },
+    {
+      "epoch": 1.9673469387755103,
+      "grad_norm": 0.11804845339511627,
+      "learning_rate": 7.280546841882396e-09,
+      "loss": 1.6414,
+      "step": 1446
+    },
+    {
+      "epoch": 1.9687074829931972,
+      "grad_norm": 0.10847051602573662,
+      "learning_rate": 6.6866068912824215e-09,
+      "loss": 1.6769,
+      "step": 1447
+    },
+    {
+      "epoch": 1.9700680272108844,
+      "grad_norm": 0.10927297829948371,
+      "learning_rate": 6.117918928693623e-09,
+      "loss": 1.6144,
+      "step": 1448
+    },
+    {
+      "epoch": 1.9714285714285715,
+      "grad_norm": 0.14038993677085151,
+      "learning_rate": 5.574485830063258e-09,
+      "loss": 1.5951,
+      "step": 1449
+    },
+    {
+      "epoch": 1.9727891156462585,
+      "grad_norm": 0.13001361414839283,
+      "learning_rate": 5.056310343619641e-09,
+      "loss": 1.5813,
+      "step": 1450
+    },
+    {
+      "epoch": 1.9741496598639454,
+      "grad_norm": 0.11240349755485105,
+      "learning_rate": 4.563395089859368e-09,
+      "loss": 1.7346,
+      "step": 1451
+    },
+    {
+      "epoch": 1.9755102040816328,
+      "grad_norm": 0.13111702124771224,
+      "learning_rate": 4.095742561533444e-09,
+      "loss": 1.531,
+      "step": 1452
+    },
+    {
+      "epoch": 1.9768707482993197,
+      "grad_norm": 0.15379785521980685,
+      "learning_rate": 3.6533551236372923e-09,
+      "loss": 1.6275,
+      "step": 1453
+    },
+    {
+      "epoch": 1.9782312925170067,
+      "grad_norm": 0.1134944447962432,
+      "learning_rate": 3.2362350133940956e-09,
+      "loss": 1.6481,
+      "step": 1454
+    },
+    {
+      "epoch": 1.9795918367346939,
+      "grad_norm": 0.13403963298652696,
+      "learning_rate": 2.844384340247586e-09,
+      "loss": 1.4958,
+      "step": 1455
+    },
+    {
+      "epoch": 1.980952380952381,
+      "grad_norm": 0.13134186683472326,
+      "learning_rate": 2.4778050858492718e-09,
+      "loss": 1.8136,
+      "step": 1456
+    },
+    {
+      "epoch": 1.982312925170068,
+      "grad_norm": 0.11530031332268623,
+      "learning_rate": 2.136499104050116e-09,
+      "loss": 1.672,
+      "step": 1457
+    },
+    {
+      "epoch": 1.9836734693877551,
+      "grad_norm": 0.14400234513580018,
+      "learning_rate": 1.8204681208883191e-09,
+      "loss": 1.5871,
+      "step": 1458
+    },
+    {
+      "epoch": 1.9850340136054423,
+      "grad_norm": 0.10792037572917504,
+      "learning_rate": 1.5297137345843261e-09,
+      "loss": 1.7173,
+      "step": 1459
+    },
+    {
+      "epoch": 1.9863945578231292,
+      "grad_norm": 0.12254785014104333,
+      "learning_rate": 1.264237415529168e-09,
+      "loss": 1.6579,
+      "step": 1460
+    },
+    {
+      "epoch": 1.9877551020408162,
+      "grad_norm": 0.13854826500139353,
+      "learning_rate": 1.0240405062794666e-09,
+      "loss": 1.5945,
+      "step": 1461
+    },
+    {
+      "epoch": 1.9891156462585036,
+      "grad_norm": 0.124820740824841,
+      "learning_rate": 8.091242215491068e-10,
+      "loss": 1.6419,
+      "step": 1462
+    },
+    {
+      "epoch": 1.9904761904761905,
+      "grad_norm": 0.1529299716450345,
+      "learning_rate": 6.194896482047963e-10,
+      "loss": 1.6038,
+      "step": 1463
+    },
+    {
+      "epoch": 1.9918367346938775,
+      "grad_norm": 0.11316485439035386,
+      "learning_rate": 4.5513774525884904e-10,
+      "loss": 1.7278,
+      "step": 1464
+    },
+    {
+      "epoch": 1.9931972789115646,
+      "grad_norm": 0.16737814919567837,
+      "learning_rate": 3.1606934386529954e-10,
+      "loss": 1.745,
+      "step": 1465
+    },
+    {
+      "epoch": 1.9945578231292518,
+      "grad_norm": 0.1260365200841888,
+      "learning_rate": 2.022851473154619e-10,
+      "loss": 1.5775,
+      "step": 1466
+    },
+    {
+      "epoch": 1.9959183673469387,
+      "grad_norm": 0.14412575459341273,
+      "learning_rate": 1.13785731034044e-10,
+      "loss": 1.5964,
+      "step": 1467
+    },
+    {
+      "epoch": 1.997278911564626,
+      "grad_norm": 0.10844606500028223,
+      "learning_rate": 5.057154257692709e-11,
+      "loss": 1.6392,
+      "step": 1468
+    },
+    {
+      "epoch": 1.998639455782313,
+      "grad_norm": 0.136002057114523,
+      "learning_rate": 1.2642901628390214e-11,
+      "loss": 1.6056,
+      "step": 1469
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.10516095768661124,
+      "learning_rate": 0.0,
+      "loss": 1.6715,
+      "step": 1470
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1470,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 184,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1646792128633242e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}