diff --git "a/checkpoint-1000/trainer_state.json" "b/checkpoint-1000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1000/trainer_state.json"
@@ -0,0 +1,7034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.663003663003663,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003663003663003663,
+      "grad_norm": 33.24192428588867,
+      "learning_rate": 0.0,
+      "loss": 0.9555,
+      "step": 1
+    },
+    {
+      "epoch": 0.007326007326007326,
+      "grad_norm": 23.005327224731445,
+      "learning_rate": 2.1978021978021978e-07,
+      "loss": 0.7557,
+      "step": 2
+    },
+    {
+      "epoch": 0.01098901098901099,
+      "grad_norm": 12.516372680664062,
+      "learning_rate": 4.3956043956043957e-07,
+      "loss": 0.2322,
+      "step": 3
+    },
+    {
+      "epoch": 0.014652014652014652,
+      "grad_norm": 22.350322723388672,
+      "learning_rate": 6.593406593406594e-07,
+      "loss": 0.5263,
+      "step": 4
+    },
+    {
+      "epoch": 0.018315018315018316,
+      "grad_norm": 37.14425277709961,
+      "learning_rate": 8.791208791208791e-07,
+      "loss": 0.547,
+      "step": 5
+    },
+    {
+      "epoch": 0.02197802197802198,
+      "grad_norm": 27.73367691040039,
+      "learning_rate": 1.098901098901099e-06,
+      "loss": 0.5922,
+      "step": 6
+    },
+    {
+      "epoch": 0.02564102564102564,
+      "grad_norm": 28.463964462280273,
+      "learning_rate": 1.3186813186813187e-06,
+      "loss": 1.0195,
+      "step": 7
+    },
+    {
+      "epoch": 0.029304029304029304,
+      "grad_norm": 12.688858032226562,
+      "learning_rate": 1.5384615384615385e-06,
+      "loss": 0.1519,
+      "step": 8
+    },
+    {
+      "epoch": 0.03296703296703297,
+      "grad_norm": 24.222930908203125,
+      "learning_rate": 1.7582417582417583e-06,
+      "loss": 0.8008,
+      "step": 9
+    },
+    {
+      "epoch": 0.03663003663003663,
+      "grad_norm": 22.45709800720215,
+      "learning_rate": 1.9780219780219782e-06,
+      "loss": 1.1024,
+      "step": 10
+    },
+    {
+      "epoch": 0.040293040293040296,
+      "grad_norm": 23.01483917236328,
+      "learning_rate": 2.197802197802198e-06,
+      "loss": 0.3072,
+      "step": 11
+    },
+    {
+      "epoch": 0.04395604395604396,
+      "grad_norm": 24.276216506958008,
+      "learning_rate": 2.4175824175824177e-06,
+      "loss": 0.8937,
+      "step": 12
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 24.501638412475586,
+      "learning_rate": 2.6373626373626375e-06,
+      "loss": 0.3748,
+      "step": 13
+    },
+    {
+      "epoch": 0.05128205128205128,
+      "grad_norm": 11.965837478637695,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 0.2221,
+      "step": 14
+    },
+    {
+      "epoch": 0.054945054945054944,
+      "grad_norm": 8.884313583374023,
+      "learning_rate": 3.076923076923077e-06,
+      "loss": 0.1682,
+      "step": 15
+    },
+    {
+      "epoch": 0.05860805860805861,
+      "grad_norm": 13.486218452453613,
+      "learning_rate": 3.2967032967032968e-06,
+      "loss": 0.3324,
+      "step": 16
+    },
+    {
+      "epoch": 0.06227106227106227,
+      "grad_norm": 29.47451400756836,
+      "learning_rate": 3.5164835164835165e-06,
+      "loss": 0.9247,
+      "step": 17
+    },
+    {
+      "epoch": 0.06593406593406594,
+      "grad_norm": 38.8739128112793,
+      "learning_rate": 3.7362637362637363e-06,
+      "loss": 1.3591,
+      "step": 18
+    },
+    {
+      "epoch": 0.0695970695970696,
+      "grad_norm": 24.181066513061523,
+      "learning_rate": 3.9560439560439565e-06,
+      "loss": 0.4257,
+      "step": 19
+    },
+    {
+      "epoch": 0.07326007326007326,
+      "grad_norm": 18.25806427001953,
+      "learning_rate": 4.175824175824176e-06,
+      "loss": 0.3534,
+      "step": 20
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 4.121458053588867,
+      "learning_rate": 4.395604395604396e-06,
+      "loss": 0.0459,
+      "step": 21
+    },
+    {
+      "epoch": 0.08058608058608059,
+      "grad_norm": 17.89643096923828,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.3707,
+      "step": 22
+    },
+    {
+      "epoch": 0.08424908424908426,
+      "grad_norm": 43.25539016723633,
+      "learning_rate": 4.8351648351648355e-06,
+      "loss": 1.139,
+      "step": 23
+    },
+    {
+      "epoch": 0.08791208791208792,
+      "grad_norm": 19.56612205505371,
+      "learning_rate": 5.054945054945056e-06,
+      "loss": 0.3819,
+      "step": 24
+    },
+    {
+      "epoch": 0.09157509157509157,
+      "grad_norm": 18.20578956604004,
+      "learning_rate": 5.274725274725275e-06,
+      "loss": 0.516,
+      "step": 25
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 23.16927146911621,
+      "learning_rate": 5.494505494505494e-06,
+      "loss": 0.7161,
+      "step": 26
+    },
+    {
+      "epoch": 0.0989010989010989,
+      "grad_norm": 10.449734687805176,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.3049,
+      "step": 27
+    },
+    {
+      "epoch": 0.10256410256410256,
+      "grad_norm": 33.13974380493164,
+      "learning_rate": 5.934065934065934e-06,
+      "loss": 1.0178,
+      "step": 28
+    },
+    {
+      "epoch": 0.10622710622710622,
+      "grad_norm": 34.373470306396484,
+      "learning_rate": 6.153846153846154e-06,
+      "loss": 1.0162,
+      "step": 29
+    },
+    {
+      "epoch": 0.10989010989010989,
+      "grad_norm": 22.710988998413086,
+      "learning_rate": 6.373626373626373e-06,
+      "loss": 0.5866,
+      "step": 30
+    },
+    {
+      "epoch": 0.11355311355311355,
+      "grad_norm": 23.314502716064453,
+      "learning_rate": 6.5934065934065935e-06,
+      "loss": 0.6159,
+      "step": 31
+    },
+    {
+      "epoch": 0.11721611721611722,
+      "grad_norm": 23.481319427490234,
+      "learning_rate": 6.813186813186814e-06,
+      "loss": 0.5441,
+      "step": 32
+    },
+    {
+      "epoch": 0.12087912087912088,
+      "grad_norm": 35.16271209716797,
+      "learning_rate": 7.032967032967033e-06,
+      "loss": 0.9091,
+      "step": 33
+    },
+    {
+      "epoch": 0.12454212454212454,
+      "grad_norm": 32.2298698425293,
+      "learning_rate": 7.252747252747253e-06,
+      "loss": 0.5156,
+      "step": 34
+    },
+    {
+      "epoch": 0.1282051282051282,
+      "grad_norm": 36.708953857421875,
+      "learning_rate": 7.4725274725274726e-06,
+      "loss": 1.5839,
+      "step": 35
+    },
+    {
+      "epoch": 0.13186813186813187,
+      "grad_norm": 34.64887619018555,
+      "learning_rate": 7.692307692307692e-06,
+      "loss": 1.2861,
+      "step": 36
+    },
+    {
+      "epoch": 0.13553113553113552,
+      "grad_norm": 20.94220733642578,
+      "learning_rate": 7.912087912087913e-06,
+      "loss": 0.5027,
+      "step": 37
+    },
+    {
+      "epoch": 0.1391941391941392,
+      "grad_norm": 30.93832015991211,
+      "learning_rate": 8.131868131868132e-06,
+      "loss": 0.3584,
+      "step": 38
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 19.195362091064453,
+      "learning_rate": 8.351648351648352e-06,
+      "loss": 0.6912,
+      "step": 39
+    },
+    {
+      "epoch": 0.14652014652014653,
+      "grad_norm": 21.054162979125977,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 0.8027,
+      "step": 40
+    },
+    {
+      "epoch": 0.15018315018315018,
+      "grad_norm": 16.64535903930664,
+      "learning_rate": 8.791208791208792e-06,
+      "loss": 0.3004,
+      "step": 41
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 12.1064453125,
+      "learning_rate": 9.010989010989011e-06,
+      "loss": 0.2158,
+      "step": 42
+    },
+    {
+      "epoch": 0.1575091575091575,
+      "grad_norm": 16.20220947265625,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.4137,
+      "step": 43
+    },
+    {
+      "epoch": 0.16117216117216118,
+      "grad_norm": 25.698654174804688,
+      "learning_rate": 9.45054945054945e-06,
+      "loss": 0.7716,
+      "step": 44
+    },
+    {
+      "epoch": 0.16483516483516483,
+      "grad_norm": 7.480422019958496,
+      "learning_rate": 9.670329670329671e-06,
+      "loss": 0.1046,
+      "step": 45
+    },
+    {
+      "epoch": 0.1684981684981685,
+      "grad_norm": 38.25539016723633,
+      "learning_rate": 9.89010989010989e-06,
+      "loss": 1.3913,
+      "step": 46
+    },
+    {
+      "epoch": 0.17216117216117216,
+      "grad_norm": 24.113954544067383,
+      "learning_rate": 1.0109890109890111e-05,
+      "loss": 0.4632,
+      "step": 47
+    },
+    {
+      "epoch": 0.17582417582417584,
+      "grad_norm": 22.136140823364258,
+      "learning_rate": 1.032967032967033e-05,
+      "loss": 0.6634,
+      "step": 48
+    },
+    {
+      "epoch": 0.1794871794871795,
+      "grad_norm": 19.417444229125977,
+      "learning_rate": 1.054945054945055e-05,
+      "loss": 0.3991,
+      "step": 49
+    },
+    {
+      "epoch": 0.18315018315018314,
+      "grad_norm": 13.265430450439453,
+      "learning_rate": 1.076923076923077e-05,
+      "loss": 0.2613,
+      "step": 50
+    },
+    {
+      "epoch": 0.18681318681318682,
+      "grad_norm": 25.118703842163086,
+      "learning_rate": 1.0989010989010989e-05,
+      "loss": 0.9231,
+      "step": 51
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 34.06997299194336,
+      "learning_rate": 1.120879120879121e-05,
+      "loss": 1.5809,
+      "step": 52
+    },
+    {
+      "epoch": 0.19413919413919414,
+      "grad_norm": 40.32486343383789,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.4601,
+      "step": 53
+    },
+    {
+      "epoch": 0.1978021978021978,
+      "grad_norm": 18.847017288208008,
+      "learning_rate": 1.1648351648351648e-05,
+      "loss": 0.2345,
+      "step": 54
+    },
+    {
+      "epoch": 0.20146520146520147,
+      "grad_norm": 37.98270034790039,
+      "learning_rate": 1.1868131868131868e-05,
+      "loss": 0.9792,
+      "step": 55
+    },
+    {
+      "epoch": 0.20512820512820512,
+      "grad_norm": 35.72782897949219,
+      "learning_rate": 1.2087912087912089e-05,
+      "loss": 1.1561,
+      "step": 56
+    },
+    {
+      "epoch": 0.2087912087912088,
+      "grad_norm": 18.577186584472656,
+      "learning_rate": 1.2307692307692308e-05,
+      "loss": 0.5577,
+      "step": 57
+    },
+    {
+      "epoch": 0.21245421245421245,
+      "grad_norm": 23.086456298828125,
+      "learning_rate": 1.2527472527472529e-05,
+      "loss": 0.5807,
+      "step": 58
+    },
+    {
+      "epoch": 0.21611721611721613,
+      "grad_norm": 20.053525924682617,
+      "learning_rate": 1.2747252747252747e-05,
+      "loss": 0.7024,
+      "step": 59
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 22.25934410095215,
+      "learning_rate": 1.2967032967032968e-05,
+      "loss": 1.1033,
+      "step": 60
+    },
+    {
+      "epoch": 0.22344322344322345,
+      "grad_norm": 17.981454849243164,
+      "learning_rate": 1.3186813186813187e-05,
+      "loss": 0.2774,
+      "step": 61
+    },
+    {
+      "epoch": 0.2271062271062271,
+      "grad_norm": 11.286524772644043,
+      "learning_rate": 1.3406593406593408e-05,
+      "loss": 0.1802,
+      "step": 62
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 25.822996139526367,
+      "learning_rate": 1.3626373626373627e-05,
+      "loss": 0.651,
+      "step": 63
+    },
+    {
+      "epoch": 0.23443223443223443,
+      "grad_norm": 16.457286834716797,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 0.2946,
+      "step": 64
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 26.712799072265625,
+      "learning_rate": 1.4065934065934066e-05,
+      "loss": 0.7763,
+      "step": 65
+    },
+    {
+      "epoch": 0.24175824175824176,
+      "grad_norm": 21.4671630859375,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.4132,
+      "step": 66
+    },
+    {
+      "epoch": 0.2454212454212454,
+      "grad_norm": 21.834922790527344,
+      "learning_rate": 1.4505494505494506e-05,
+      "loss": 0.6544,
+      "step": 67
+    },
+    {
+      "epoch": 0.2490842490842491,
+      "grad_norm": 15.396453857421875,
+      "learning_rate": 1.4725274725274726e-05,
+      "loss": 0.2426,
+      "step": 68
+    },
+    {
+      "epoch": 0.25274725274725274,
+      "grad_norm": 8.851480484008789,
+      "learning_rate": 1.4945054945054945e-05,
+      "loss": 0.125,
+      "step": 69
+    },
+    {
+      "epoch": 0.2564102564102564,
+      "grad_norm": 22.21581268310547,
+      "learning_rate": 1.5164835164835164e-05,
+      "loss": 0.2585,
+      "step": 70
+    },
+    {
+      "epoch": 0.2600732600732601,
+      "grad_norm": 23.589736938476562,
+      "learning_rate": 1.5384615384615384e-05,
+      "loss": 0.386,
+      "step": 71
+    },
+    {
+      "epoch": 0.26373626373626374,
+      "grad_norm": 51.82280731201172,
+      "learning_rate": 1.5604395604395605e-05,
+      "loss": 1.1802,
+      "step": 72
+    },
+    {
+      "epoch": 0.2673992673992674,
+      "grad_norm": 36.43033981323242,
+      "learning_rate": 1.5824175824175826e-05,
+      "loss": 0.5574,
+      "step": 73
+    },
+    {
+      "epoch": 0.27106227106227104,
+      "grad_norm": 46.151885986328125,
+      "learning_rate": 1.6043956043956043e-05,
+      "loss": 0.9113,
+      "step": 74
+    },
+    {
+      "epoch": 0.27472527472527475,
+      "grad_norm": 34.090213775634766,
+      "learning_rate": 1.6263736263736265e-05,
+      "loss": 1.2161,
+      "step": 75
+    },
+    {
+      "epoch": 0.2783882783882784,
+      "grad_norm": 15.469125747680664,
+      "learning_rate": 1.6483516483516486e-05,
+      "loss": 0.1833,
+      "step": 76
+    },
+    {
+      "epoch": 0.28205128205128205,
+      "grad_norm": 26.77261734008789,
+      "learning_rate": 1.6703296703296703e-05,
+      "loss": 0.4095,
+      "step": 77
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 8.46114444732666,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 0.0724,
+      "step": 78
+    },
+    {
+      "epoch": 0.2893772893772894,
+      "grad_norm": 7.954617500305176,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.057,
+      "step": 79
+    },
+    {
+      "epoch": 0.29304029304029305,
+      "grad_norm": 32.47618103027344,
+      "learning_rate": 1.7362637362637366e-05,
+      "loss": 0.8099,
+      "step": 80
+    },
+    {
+      "epoch": 0.2967032967032967,
+      "grad_norm": 34.506927490234375,
+      "learning_rate": 1.7582417582417584e-05,
+      "loss": 0.5867,
+      "step": 81
+    },
+    {
+      "epoch": 0.30036630036630035,
+      "grad_norm": 18.276355743408203,
+      "learning_rate": 1.78021978021978e-05,
+      "loss": 0.4387,
+      "step": 82
+    },
+    {
+      "epoch": 0.304029304029304,
+      "grad_norm": 35.61729431152344,
+      "learning_rate": 1.8021978021978023e-05,
+      "loss": 0.9711,
+      "step": 83
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 14.001388549804688,
+      "learning_rate": 1.824175824175824e-05,
+      "loss": 0.1431,
+      "step": 84
+    },
+    {
+      "epoch": 0.31135531135531136,
+      "grad_norm": 27.521188735961914,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.3686,
+      "step": 85
+    },
+    {
+      "epoch": 0.315018315018315,
+      "grad_norm": 38.0133171081543,
+      "learning_rate": 1.8681318681318682e-05,
+      "loss": 1.3866,
+      "step": 86
+    },
+    {
+      "epoch": 0.31868131868131866,
+      "grad_norm": 30.895553588867188,
+      "learning_rate": 1.89010989010989e-05,
+      "loss": 0.6676,
+      "step": 87
+    },
+    {
+      "epoch": 0.32234432234432236,
+      "grad_norm": 26.165082931518555,
+      "learning_rate": 1.912087912087912e-05,
+      "loss": 0.4763,
+      "step": 88
+    },
+    {
+      "epoch": 0.326007326007326,
+      "grad_norm": 25.6451473236084,
+      "learning_rate": 1.9340659340659342e-05,
+      "loss": 0.6921,
+      "step": 89
+    },
+    {
+      "epoch": 0.32967032967032966,
+      "grad_norm": 31.52683448791504,
+      "learning_rate": 1.9560439560439563e-05,
+      "loss": 0.8449,
+      "step": 90
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 27.559072494506836,
+      "learning_rate": 1.978021978021978e-05,
+      "loss": 0.9726,
+      "step": 91
+    },
+    {
+      "epoch": 0.336996336996337,
+      "grad_norm": 38.23103713989258,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 0.2568,
+      "step": 92
+    },
+    {
+      "epoch": 0.34065934065934067,
+      "grad_norm": 28.575313568115234,
+      "learning_rate": 2.0219780219780223e-05,
+      "loss": 0.7039,
+      "step": 93
+    },
+    {
+      "epoch": 0.3443223443223443,
+      "grad_norm": 31.54847526550293,
+      "learning_rate": 2.043956043956044e-05,
+      "loss": 0.835,
+      "step": 94
+    },
+    {
+      "epoch": 0.34798534798534797,
+      "grad_norm": 34.27505111694336,
+      "learning_rate": 2.065934065934066e-05,
+      "loss": 1.0304,
+      "step": 95
+    },
+    {
+      "epoch": 0.3516483516483517,
+      "grad_norm": 23.972553253173828,
+      "learning_rate": 2.087912087912088e-05,
+      "loss": 0.775,
+      "step": 96
+    },
+    {
+      "epoch": 0.3553113553113553,
+      "grad_norm": 18.46526527404785,
+      "learning_rate": 2.10989010989011e-05,
+      "loss": 0.2856,
+      "step": 97
+    },
+    {
+      "epoch": 0.358974358974359,
+      "grad_norm": 22.087251663208008,
+      "learning_rate": 2.131868131868132e-05,
+      "loss": 0.6849,
+      "step": 98
+    },
+    {
+      "epoch": 0.3626373626373626,
+      "grad_norm": 13.144533157348633,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.2766,
+      "step": 99
+    },
+    {
+      "epoch": 0.3663003663003663,
+      "grad_norm": 14.740280151367188,
+      "learning_rate": 2.175824175824176e-05,
+      "loss": 0.27,
+      "step": 100
+    },
+    {
+      "epoch": 0.36996336996337,
+      "grad_norm": 17.15272331237793,
+      "learning_rate": 2.1978021978021977e-05,
+      "loss": 0.446,
+      "step": 101
+    },
+    {
+      "epoch": 0.37362637362637363,
+      "grad_norm": 45.865509033203125,
+      "learning_rate": 2.21978021978022e-05,
+      "loss": 2.4265,
+      "step": 102
+    },
+    {
+      "epoch": 0.3772893772893773,
+      "grad_norm": 22.298274993896484,
+      "learning_rate": 2.241758241758242e-05,
+      "loss": 1.5021,
+      "step": 103
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 20.314172744750977,
+      "learning_rate": 2.2637362637362637e-05,
+      "loss": 0.508,
+      "step": 104
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 11.217910766601562,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.2282,
+      "step": 105
+    },
+    {
+      "epoch": 0.3882783882783883,
+      "grad_norm": 21.36184310913086,
+      "learning_rate": 2.307692307692308e-05,
+      "loss": 0.4684,
+      "step": 106
+    },
+    {
+      "epoch": 0.39194139194139194,
+      "grad_norm": 12.759861946105957,
+      "learning_rate": 2.3296703296703297e-05,
+      "loss": 0.3076,
+      "step": 107
+    },
+    {
+      "epoch": 0.3956043956043956,
+      "grad_norm": 24.42287254333496,
+      "learning_rate": 2.3516483516483518e-05,
+      "loss": 1.3607,
+      "step": 108
+    },
+    {
+      "epoch": 0.3992673992673993,
+      "grad_norm": 13.014902114868164,
+      "learning_rate": 2.3736263736263735e-05,
+      "loss": 0.4984,
+      "step": 109
+    },
+    {
+      "epoch": 0.40293040293040294,
+      "grad_norm": 12.8681640625,
+      "learning_rate": 2.395604395604396e-05,
+      "loss": 0.4529,
+      "step": 110
+    },
+    {
+      "epoch": 0.4065934065934066,
+      "grad_norm": 21.19939422607422,
+      "learning_rate": 2.4175824175824177e-05,
+      "loss": 1.0197,
+      "step": 111
+    },
+    {
+      "epoch": 0.41025641025641024,
+      "grad_norm": 20.60430145263672,
+      "learning_rate": 2.4395604395604395e-05,
+      "loss": 0.5367,
+      "step": 112
+    },
+    {
+      "epoch": 0.4139194139194139,
+      "grad_norm": 34.49782943725586,
+      "learning_rate": 2.4615384615384616e-05,
+      "loss": 1.9045,
+      "step": 113
+    },
+    {
+      "epoch": 0.4175824175824176,
+      "grad_norm": 28.380966186523438,
+      "learning_rate": 2.4835164835164834e-05,
+      "loss": 0.9019,
+      "step": 114
+    },
+    {
+      "epoch": 0.42124542124542125,
+      "grad_norm": 18.234045028686523,
+      "learning_rate": 2.5054945054945058e-05,
+      "loss": 0.5529,
+      "step": 115
+    },
+    {
+      "epoch": 0.4249084249084249,
+      "grad_norm": 18.759784698486328,
+      "learning_rate": 2.5274725274725276e-05,
+      "loss": 0.85,
+      "step": 116
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 15.784387588500977,
+      "learning_rate": 2.5494505494505493e-05,
+      "loss": 0.429,
+      "step": 117
+    },
+    {
+      "epoch": 0.43223443223443225,
+      "grad_norm": 23.149036407470703,
+      "learning_rate": 2.5714285714285714e-05,
+      "loss": 0.8784,
+      "step": 118
+    },
+    {
+      "epoch": 0.4358974358974359,
+      "grad_norm": 18.77080535888672,
+      "learning_rate": 2.5934065934065935e-05,
+      "loss": 0.537,
+      "step": 119
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 24.311708450317383,
+      "learning_rate": 2.6153846153846157e-05,
+      "loss": 0.74,
+      "step": 120
+    },
+    {
+      "epoch": 0.4432234432234432,
+      "grad_norm": 15.09874439239502,
+      "learning_rate": 2.6373626373626374e-05,
+      "loss": 0.2978,
+      "step": 121
+    },
+    {
+      "epoch": 0.4468864468864469,
+      "grad_norm": 19.65829086303711,
+      "learning_rate": 2.6593406593406592e-05,
+      "loss": 0.8287,
+      "step": 122
+    },
+    {
+      "epoch": 0.45054945054945056,
+      "grad_norm": 21.237165451049805,
+      "learning_rate": 2.6813186813186816e-05,
+      "loss": 1.1967,
+      "step": 123
+    },
+    {
+      "epoch": 0.4542124542124542,
+      "grad_norm": 25.737913131713867,
+      "learning_rate": 2.7032967032967034e-05,
+      "loss": 0.9414,
+      "step": 124
+    },
+    {
+      "epoch": 0.45787545787545786,
+      "grad_norm": 22.84954833984375,
+      "learning_rate": 2.7252747252747255e-05,
+      "loss": 0.398,
+      "step": 125
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 35.505027770996094,
+      "learning_rate": 2.7472527472527473e-05,
+      "loss": 1.0497,
+      "step": 126
+    },
+    {
+      "epoch": 0.4652014652014652,
+      "grad_norm": 6.610748291015625,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.0491,
+      "step": 127
+    },
+    {
+      "epoch": 0.46886446886446886,
+      "grad_norm": 33.34388732910156,
+      "learning_rate": 2.7912087912087915e-05,
+      "loss": 0.8991,
+      "step": 128
+    },
+    {
+      "epoch": 0.4725274725274725,
+      "grad_norm": 17.098581314086914,
+      "learning_rate": 2.8131868131868132e-05,
+      "loss": 0.3217,
+      "step": 129
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 11.438309669494629,
+      "learning_rate": 2.8351648351648353e-05,
+      "loss": 0.4301,
+      "step": 130
+    },
+    {
+      "epoch": 0.47985347985347987,
+      "grad_norm": 25.803213119506836,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 0.8937,
+      "step": 131
+    },
+    {
+      "epoch": 0.4835164835164835,
+      "grad_norm": 16.61037826538086,
+      "learning_rate": 2.8791208791208792e-05,
+      "loss": 0.3603,
+      "step": 132
+    },
+    {
+      "epoch": 0.48717948717948717,
+      "grad_norm": 21.329975128173828,
+      "learning_rate": 2.9010989010989013e-05,
+      "loss": 0.4332,
+      "step": 133
+    },
+    {
+      "epoch": 0.4908424908424908,
+      "grad_norm": 24.83706283569336,
+      "learning_rate": 2.923076923076923e-05,
+      "loss": 0.3967,
+      "step": 134
+    },
+    {
+      "epoch": 0.4945054945054945,
+      "grad_norm": 8.3758544921875,
+      "learning_rate": 2.945054945054945e-05,
+      "loss": 0.1197,
+      "step": 135
+    },
+    {
+      "epoch": 0.4981684981684982,
+      "grad_norm": 31.096702575683594,
+      "learning_rate": 2.9670329670329673e-05,
+      "loss": 2.2867,
+      "step": 136
+    },
+    {
+      "epoch": 0.5018315018315018,
+      "grad_norm": 17.094390869140625,
+      "learning_rate": 2.989010989010989e-05,
+      "loss": 0.3064,
+      "step": 137
+    },
+    {
+      "epoch": 0.5054945054945055,
+      "grad_norm": 23.401243209838867,
+      "learning_rate": 3.010989010989011e-05,
+      "loss": 0.9779,
+      "step": 138
+    },
+    {
+      "epoch": 0.5091575091575091,
+      "grad_norm": 19.55811309814453,
+      "learning_rate": 3.032967032967033e-05,
+      "loss": 0.5665,
+      "step": 139
+    },
+    {
+      "epoch": 0.5128205128205128,
+      "grad_norm": 18.668622970581055,
+      "learning_rate": 3.0549450549450547e-05,
+      "loss": 0.7068,
+      "step": 140
+    },
+    {
+      "epoch": 0.5164835164835165,
+      "grad_norm": 9.49342155456543,
+      "learning_rate": 3.076923076923077e-05,
+      "loss": 0.2228,
+      "step": 141
+    },
+    {
+      "epoch": 0.5201465201465202,
+      "grad_norm": 17.131006240844727,
+      "learning_rate": 3.0989010989010995e-05,
+      "loss": 0.8947,
+      "step": 142
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 14.087484359741211,
+      "learning_rate": 3.120879120879121e-05,
+      "loss": 0.4394,
+      "step": 143
+    },
+    {
+      "epoch": 0.5274725274725275,
+      "grad_norm": 14.246976852416992,
+      "learning_rate": 3.142857142857143e-05,
+      "loss": 0.7608,
+      "step": 144
+    },
+    {
+      "epoch": 0.5311355311355311,
+      "grad_norm": 27.454071044921875,
+      "learning_rate": 3.164835164835165e-05,
+      "loss": 1.8982,
+      "step": 145
+    },
+    {
+      "epoch": 0.5347985347985348,
+      "grad_norm": 8.580923080444336,
+      "learning_rate": 3.1868131868131866e-05,
+      "loss": 0.2199,
+      "step": 146
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 12.200552940368652,
+      "learning_rate": 3.208791208791209e-05,
+      "loss": 0.4007,
+      "step": 147
+    },
+    {
+      "epoch": 0.5421245421245421,
+      "grad_norm": 11.350752830505371,
+      "learning_rate": 3.230769230769231e-05,
+      "loss": 0.5359,
+      "step": 148
+    },
+    {
+      "epoch": 0.5457875457875457,
+      "grad_norm": 21.45020866394043,
+      "learning_rate": 3.252747252747253e-05,
+      "loss": 1.4639,
+      "step": 149
+    },
+    {
+      "epoch": 0.5494505494505495,
+      "grad_norm": 29.84933090209961,
+      "learning_rate": 3.274725274725274e-05,
+      "loss": 0.8764,
+      "step": 150
+    },
+    {
+      "epoch": 0.5531135531135531,
+      "grad_norm": 14.899048805236816,
+      "learning_rate": 3.296703296703297e-05,
+      "loss": 0.3817,
+      "step": 151
+    },
+    {
+      "epoch": 0.5567765567765568,
+      "grad_norm": 14.95295238494873,
+      "learning_rate": 3.318681318681319e-05,
+      "loss": 1.0153,
+      "step": 152
+    },
+    {
+      "epoch": 0.5604395604395604,
+      "grad_norm": 13.904314994812012,
+      "learning_rate": 3.3406593406593406e-05,
+      "loss": 0.9891,
+      "step": 153
+    },
+    {
+      "epoch": 0.5641025641025641,
+      "grad_norm": 14.465546607971191,
+      "learning_rate": 3.362637362637363e-05,
+      "loss": 0.4935,
+      "step": 154
+    },
+    {
+      "epoch": 0.5677655677655677,
+      "grad_norm": 15.22211742401123,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.4973,
+      "step": 155
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 19.977941513061523,
+      "learning_rate": 3.406593406593406e-05,
+      "loss": 0.5768,
+      "step": 156
+    },
+    {
+      "epoch": 0.575091575091575,
+      "grad_norm": 21.778785705566406,
+      "learning_rate": 3.4285714285714284e-05,
+      "loss": 0.541,
+      "step": 157
+    },
+    {
+      "epoch": 0.5787545787545788,
+      "grad_norm": 7.957052707672119,
+      "learning_rate": 3.4505494505494505e-05,
+      "loss": 0.1676,
+      "step": 158
+    },
+    {
+      "epoch": 0.5824175824175825,
+      "grad_norm": 10.105476379394531,
+      "learning_rate": 3.472527472527473e-05,
+      "loss": 0.14,
+      "step": 159
+    },
+    {
+      "epoch": 0.5860805860805861,
+      "grad_norm": 13.895249366760254,
+      "learning_rate": 3.494505494505495e-05,
+      "loss": 0.2135,
+      "step": 160
+    },
+    {
+      "epoch": 0.5897435897435898,
+      "grad_norm": 15.14104175567627,
+      "learning_rate": 3.516483516483517e-05,
+      "loss": 0.2299,
+      "step": 161
+    },
+    {
+      "epoch": 0.5934065934065934,
+      "grad_norm": 27.537504196166992,
+      "learning_rate": 3.538461538461539e-05,
+      "loss": 0.4517,
+      "step": 162
+    },
+    {
+      "epoch": 0.5970695970695971,
+      "grad_norm": 22.290597915649414,
+      "learning_rate": 3.56043956043956e-05,
+      "loss": 0.2144,
+      "step": 163
+    },
+    {
+      "epoch": 0.6007326007326007,
+      "grad_norm": 24.176603317260742,
+      "learning_rate": 3.5824175824175824e-05,
+      "loss": 0.4184,
+      "step": 164
+    },
+    {
+      "epoch": 0.6043956043956044,
+      "grad_norm": 43.716552734375,
+      "learning_rate": 3.6043956043956045e-05,
+      "loss": 0.7672,
+      "step": 165
+    },
+    {
+      "epoch": 0.608058608058608,
+      "grad_norm": 5.516793727874756,
+      "learning_rate": 3.626373626373626e-05,
+      "loss": 0.0332,
+      "step": 166
+    },
+    {
+      "epoch": 0.6117216117216118,
+      "grad_norm": 13.202600479125977,
+      "learning_rate": 3.648351648351648e-05,
+      "loss": 0.1388,
+      "step": 167
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 8.389626502990723,
+      "learning_rate": 3.670329670329671e-05,
+      "loss": 0.0284,
+      "step": 168
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 11.500190734863281,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.1778,
+      "step": 169
+    },
+    {
+      "epoch": 0.6227106227106227,
+      "grad_norm": 49.76407241821289,
+      "learning_rate": 3.7142857142857143e-05,
+      "loss": 0.8075,
+      "step": 170
+    },
+    {
+      "epoch": 0.6263736263736264,
+      "grad_norm": 49.758705139160156,
+      "learning_rate": 3.7362637362637365e-05,
+      "loss": 1.3106,
+      "step": 171
+    },
+    {
+      "epoch": 0.63003663003663,
+      "grad_norm": 7.655544281005859,
+      "learning_rate": 3.7582417582417586e-05,
+      "loss": 0.1362,
+      "step": 172
+    },
+    {
+      "epoch": 0.6336996336996337,
+      "grad_norm": 29.778133392333984,
+      "learning_rate": 3.78021978021978e-05,
+      "loss": 0.2411,
+      "step": 173
+    },
+    {
+      "epoch": 0.6373626373626373,
+      "grad_norm": 23.79543113708496,
+      "learning_rate": 3.802197802197802e-05,
+      "loss": 0.5665,
+      "step": 174
+    },
+    {
+      "epoch": 0.6410256410256411,
+      "grad_norm": 25.333166122436523,
+      "learning_rate": 3.824175824175824e-05,
+      "loss": 0.5821,
+      "step": 175
+    },
+    {
+      "epoch": 0.6446886446886447,
+      "grad_norm": 38.367759704589844,
+      "learning_rate": 3.846153846153846e-05,
+      "loss": 1.1098,
+      "step": 176
+    },
+    {
+      "epoch": 0.6483516483516484,
+      "grad_norm": 31.53361701965332,
+      "learning_rate": 3.8681318681318684e-05,
+      "loss": 1.5399,
+      "step": 177
+    },
+    {
+      "epoch": 0.652014652014652,
+      "grad_norm": 8.453901290893555,
+      "learning_rate": 3.8901098901098905e-05,
+      "loss": 0.1327,
+      "step": 178
+    },
+    {
+      "epoch": 0.6556776556776557,
+      "grad_norm": 32.465980529785156,
+      "learning_rate": 3.9120879120879126e-05,
+      "loss": 0.8133,
+      "step": 179
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 21.503114700317383,
+      "learning_rate": 3.934065934065934e-05,
+      "loss": 0.2472,
+      "step": 180
+    },
+    {
+      "epoch": 0.663003663003663,
+      "grad_norm": 28.240659713745117,
+      "learning_rate": 3.956043956043956e-05,
+      "loss": 0.4718,
+      "step": 181
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 6.919331073760986,
+      "learning_rate": 3.978021978021978e-05,
+      "loss": 0.0947,
+      "step": 182
+    },
+    {
+      "epoch": 0.6703296703296703,
+      "grad_norm": 20.96783447265625,
+      "learning_rate": 3.9999999999999996e-05,
+      "loss": 1.1602,
+      "step": 183
+    },
+    {
+      "epoch": 0.673992673992674,
+      "grad_norm": 17.967914581298828,
+      "learning_rate": 4.021978021978022e-05,
+      "loss": 0.3684,
+      "step": 184
+    },
+    {
+      "epoch": 0.6776556776556777,
+      "grad_norm": 29.837678909301758,
+      "learning_rate": 4.0439560439560445e-05,
+      "loss": 0.5452,
+      "step": 185
+    },
+    {
+      "epoch": 0.6813186813186813,
+      "grad_norm": 37.0803108215332,
+      "learning_rate": 4.065934065934066e-05,
+      "loss": 0.5983,
+      "step": 186
+    },
+    {
+      "epoch": 0.684981684981685,
+      "grad_norm": 23.339448928833008,
+      "learning_rate": 4.087912087912088e-05,
+      "loss": 0.6255,
+      "step": 187
+    },
+    {
+      "epoch": 0.6886446886446886,
+      "grad_norm": 13.779767036437988,
+      "learning_rate": 4.10989010989011e-05,
+      "loss": 0.3705,
+      "step": 188
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 15.792436599731445,
+      "learning_rate": 4.131868131868132e-05,
+      "loss": 0.4128,
+      "step": 189
+    },
+    {
+      "epoch": 0.6959706959706959,
+      "grad_norm": 14.106623649597168,
+      "learning_rate": 4.153846153846154e-05,
+      "loss": 0.2914,
+      "step": 190
+    },
+    {
+      "epoch": 0.6996336996336996,
+      "grad_norm": 34.428951263427734,
+      "learning_rate": 4.175824175824176e-05,
+      "loss": 1.2232,
+      "step": 191
+    },
+    {
+      "epoch": 0.7032967032967034,
+      "grad_norm": 15.847033500671387,
+      "learning_rate": 4.197802197802198e-05,
+      "loss": 0.4129,
+      "step": 192
+    },
+    {
+      "epoch": 0.706959706959707,
+      "grad_norm": 17.834794998168945,
+      "learning_rate": 4.21978021978022e-05,
+      "loss": 0.4158,
+      "step": 193
+    },
+    {
+      "epoch": 0.7106227106227107,
+      "grad_norm": 29.807823181152344,
+      "learning_rate": 4.241758241758242e-05,
+      "loss": 0.9741,
+      "step": 194
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 15.9482421875,
+      "learning_rate": 4.263736263736264e-05,
+      "loss": 0.1953,
+      "step": 195
+    },
+    {
+      "epoch": 0.717948717948718,
+      "grad_norm": 37.89487075805664,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 1.1018,
+      "step": 196
+    },
+    {
+      "epoch": 0.7216117216117216,
+      "grad_norm": 24.060779571533203,
+      "learning_rate": 4.307692307692308e-05,
+      "loss": 0.4774,
+      "step": 197
+    },
+    {
+      "epoch": 0.7252747252747253,
+      "grad_norm": 18.701725006103516,
+      "learning_rate": 4.32967032967033e-05,
+      "loss": 0.2641,
+      "step": 198
+    },
+    {
+      "epoch": 0.7289377289377289,
+      "grad_norm": 32.18348693847656,
+      "learning_rate": 4.351648351648352e-05,
+      "loss": 0.6958,
+      "step": 199
+    },
+    {
+      "epoch": 0.7326007326007326,
+      "grad_norm": 16.504337310791016,
+      "learning_rate": 4.3736263736263734e-05,
+      "loss": 0.1933,
+      "step": 200
+    },
+    {
+      "epoch": 0.7362637362637363,
+      "grad_norm": 34.5928840637207,
+      "learning_rate": 4.3956043956043955e-05,
+      "loss": 0.3712,
+      "step": 201
+    },
+    {
+      "epoch": 0.73992673992674,
+      "grad_norm": 47.998512268066406,
+      "learning_rate": 4.417582417582418e-05,
+      "loss": 1.4578,
+      "step": 202
+    },
+    {
+      "epoch": 0.7435897435897436,
+      "grad_norm": 29.871829986572266,
+      "learning_rate": 4.43956043956044e-05,
+      "loss": 0.7628,
+      "step": 203
+    },
+    {
+      "epoch": 0.7472527472527473,
+      "grad_norm": 53.70481491088867,
+      "learning_rate": 4.461538461538462e-05,
+      "loss": 1.4017,
+      "step": 204
+    },
+    {
+      "epoch": 0.7509157509157509,
+      "grad_norm": 58.087646484375,
+      "learning_rate": 4.483516483516484e-05,
+      "loss": 1.3168,
+      "step": 205
+    },
+    {
+      "epoch": 0.7545787545787546,
+      "grad_norm": 44.62531280517578,
+      "learning_rate": 4.505494505494505e-05,
+      "loss": 0.8959,
+      "step": 206
+    },
+    {
+      "epoch": 0.7582417582417582,
+      "grad_norm": 18.427953720092773,
+      "learning_rate": 4.5274725274725274e-05,
+      "loss": 0.4202,
+      "step": 207
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 32.799434661865234,
+      "learning_rate": 4.5494505494505495e-05,
+      "loss": 0.5432,
+      "step": 208
+    },
+    {
+      "epoch": 0.7655677655677655,
+      "grad_norm": 22.136354446411133,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 1.0474,
+      "step": 209
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 14.09807014465332,
+      "learning_rate": 4.593406593406593e-05,
+      "loss": 0.4048,
+      "step": 210
+    },
+    {
+      "epoch": 0.7728937728937729,
+      "grad_norm": 16.818132400512695,
+      "learning_rate": 4.615384615384616e-05,
+      "loss": 0.4772,
+      "step": 211
+    },
+    {
+      "epoch": 0.7765567765567766,
+      "grad_norm": 36.87644577026367,
+      "learning_rate": 4.637362637362638e-05,
+      "loss": 1.0203,
+      "step": 212
+    },
+    {
+      "epoch": 0.7802197802197802,
+      "grad_norm": 23.279033660888672,
+      "learning_rate": 4.6593406593406593e-05,
+      "loss": 0.8223,
+      "step": 213
+    },
+    {
+      "epoch": 0.7838827838827839,
+      "grad_norm": 21.23172378540039,
+      "learning_rate": 4.6813186813186814e-05,
+      "loss": 0.6838,
+      "step": 214
+    },
+    {
+      "epoch": 0.7875457875457875,
+      "grad_norm": 15.129582405090332,
+      "learning_rate": 4.7032967032967035e-05,
+      "loss": 0.3939,
+      "step": 215
+    },
+    {
+      "epoch": 0.7912087912087912,
+      "grad_norm": 38.20903778076172,
+      "learning_rate": 4.725274725274725e-05,
+      "loss": 0.4395,
+      "step": 216
+    },
+    {
+      "epoch": 0.7948717948717948,
+      "grad_norm": 23.428571701049805,
+      "learning_rate": 4.747252747252747e-05,
+      "loss": 0.6657,
+      "step": 217
+    },
+    {
+      "epoch": 0.7985347985347986,
+      "grad_norm": 15.892741203308105,
+      "learning_rate": 4.769230769230769e-05,
+      "loss": 0.3867,
+      "step": 218
+    },
+    {
+      "epoch": 0.8021978021978022,
+      "grad_norm": 44.7977180480957,
+      "learning_rate": 4.791208791208792e-05,
+      "loss": 1.4335,
+      "step": 219
+    },
+    {
+      "epoch": 0.8058608058608059,
+      "grad_norm": 18.13700294494629,
+      "learning_rate": 4.8131868131868134e-05,
+      "loss": 0.3965,
+      "step": 220
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 23.00497817993164,
+      "learning_rate": 4.8351648351648355e-05,
+      "loss": 1.1319,
+      "step": 221
+    },
+    {
+      "epoch": 0.8131868131868132,
+      "grad_norm": 27.63648796081543,
+      "learning_rate": 4.8571428571428576e-05,
+      "loss": 0.7782,
+      "step": 222
+    },
+    {
+      "epoch": 0.8168498168498168,
+      "grad_norm": 23.91630744934082,
+      "learning_rate": 4.879120879120879e-05,
+      "loss": 0.7277,
+      "step": 223
+    },
+    {
+      "epoch": 0.8205128205128205,
+      "grad_norm": 27.157682418823242,
+      "learning_rate": 4.901098901098901e-05,
+      "loss": 0.8309,
+      "step": 224
+    },
+    {
+      "epoch": 0.8241758241758241,
+      "grad_norm": 20.686105728149414,
+      "learning_rate": 4.923076923076923e-05,
+      "loss": 0.4645,
+      "step": 225
+    },
+    {
+      "epoch": 0.8278388278388278,
+      "grad_norm": 18.44706916809082,
+      "learning_rate": 4.9450549450549446e-05,
+      "loss": 0.6298,
+      "step": 226
+    },
+    {
+      "epoch": 0.8315018315018315,
+      "grad_norm": 34.66194152832031,
+      "learning_rate": 4.967032967032967e-05,
+      "loss": 1.3282,
+      "step": 227
+    },
+    {
+      "epoch": 0.8351648351648352,
+      "grad_norm": 26.68456268310547,
+      "learning_rate": 4.9890109890109895e-05,
+      "loss": 0.8652,
+      "step": 228
+    },
+    {
+      "epoch": 0.8388278388278388,
+      "grad_norm": 18.36819839477539,
+      "learning_rate": 5.0109890109890116e-05,
+      "loss": 0.425,
+      "step": 229
+    },
+    {
+      "epoch": 0.8424908424908425,
+      "grad_norm": 10.212838172912598,
+      "learning_rate": 5.032967032967033e-05,
+      "loss": 0.2183,
+      "step": 230
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 28.40265464782715,
+      "learning_rate": 5.054945054945055e-05,
+      "loss": 1.6894,
+      "step": 231
+    },
+    {
+      "epoch": 0.8498168498168498,
+      "grad_norm": 48.70882797241211,
+      "learning_rate": 5.076923076923077e-05,
+      "loss": 0.8564,
+      "step": 232
+    },
+    {
+      "epoch": 0.8534798534798534,
+      "grad_norm": 38.576541900634766,
+      "learning_rate": 5.098901098901099e-05,
+      "loss": 0.8013,
+      "step": 233
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 20.17264747619629,
+      "learning_rate": 5.120879120879121e-05,
+      "loss": 0.4553,
+      "step": 234
+    },
+    {
+      "epoch": 0.8608058608058609,
+      "grad_norm": 33.383182525634766,
+      "learning_rate": 5.142857142857143e-05,
+      "loss": 0.9591,
+      "step": 235
+    },
+    {
+      "epoch": 0.8644688644688645,
+      "grad_norm": 22.734106063842773,
+      "learning_rate": 5.164835164835165e-05,
+      "loss": 0.589,
+      "step": 236
+    },
+    {
+      "epoch": 0.8681318681318682,
+      "grad_norm": 19.77442741394043,
+      "learning_rate": 5.186813186813187e-05,
+      "loss": 0.7066,
+      "step": 237
+    },
+    {
+      "epoch": 0.8717948717948718,
+      "grad_norm": 32.36431884765625,
+      "learning_rate": 5.208791208791209e-05,
+      "loss": 0.8878,
+      "step": 238
+    },
+    {
+      "epoch": 0.8754578754578755,
+      "grad_norm": 37.60574722290039,
+      "learning_rate": 5.230769230769231e-05,
+      "loss": 1.0034,
+      "step": 239
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 28.051666259765625,
+      "learning_rate": 5.252747252747253e-05,
+      "loss": 0.9695,
+      "step": 240
+    },
+    {
+      "epoch": 0.8827838827838828,
+      "grad_norm": 31.55886459350586,
+      "learning_rate": 5.274725274725275e-05,
+      "loss": 0.5416,
+      "step": 241
+    },
+    {
+      "epoch": 0.8864468864468864,
+      "grad_norm": 17.856632232666016,
+      "learning_rate": 5.296703296703297e-05,
+      "loss": 0.3647,
+      "step": 242
+    },
+    {
+      "epoch": 0.8901098901098901,
+      "grad_norm": 42.52962112426758,
+      "learning_rate": 5.3186813186813184e-05,
+      "loss": 1.3661,
+      "step": 243
+    },
+    {
+      "epoch": 0.8937728937728938,
+      "grad_norm": 26.439769744873047,
+      "learning_rate": 5.3406593406593405e-05,
+      "loss": 0.6629,
+      "step": 244
+    },
+    {
+      "epoch": 0.8974358974358975,
+      "grad_norm": 37.46576690673828,
+      "learning_rate": 5.362637362637363e-05,
+      "loss": 0.9631,
+      "step": 245
+    },
+    {
+      "epoch": 0.9010989010989011,
+      "grad_norm": 29.706708908081055,
+      "learning_rate": 5.384615384615385e-05,
+      "loss": 1.0034,
+      "step": 246
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 33.62871551513672,
+      "learning_rate": 5.406593406593407e-05,
+      "loss": 0.8036,
+      "step": 247
+    },
+    {
+      "epoch": 0.9084249084249084,
+      "grad_norm": 41.97051239013672,
+      "learning_rate": 5.428571428571429e-05,
+      "loss": 1.309,
+      "step": 248
+    },
+    {
+      "epoch": 0.9120879120879121,
+      "grad_norm": 37.57841110229492,
+      "learning_rate": 5.450549450549451e-05,
+      "loss": 1.2444,
+      "step": 249
+    },
+    {
+      "epoch": 0.9157509157509157,
+      "grad_norm": 21.220727920532227,
+      "learning_rate": 5.4725274725274724e-05,
+      "loss": 0.6556,
+      "step": 250
+    },
+    {
+      "epoch": 0.9194139194139194,
+      "grad_norm": 19.963764190673828,
+      "learning_rate": 5.4945054945054945e-05,
+      "loss": 0.7328,
+      "step": 251
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 21.196062088012695,
+      "learning_rate": 5.5164835164835166e-05,
+      "loss": 0.5752,
+      "step": 252
+    },
+    {
+      "epoch": 0.9267399267399268,
+      "grad_norm": 23.587268829345703,
+      "learning_rate": 5.538461538461539e-05,
+      "loss": 0.4801,
+      "step": 253
+    },
+    {
+      "epoch": 0.9304029304029304,
+      "grad_norm": 16.09604263305664,
+      "learning_rate": 5.560439560439561e-05,
+      "loss": 0.4795,
+      "step": 254
+    },
+    {
+      "epoch": 0.9340659340659341,
+      "grad_norm": 22.61296272277832,
+      "learning_rate": 5.582417582417583e-05,
+      "loss": 0.5807,
+      "step": 255
+    },
+    {
+      "epoch": 0.9377289377289377,
+      "grad_norm": 28.715890884399414,
+      "learning_rate": 5.604395604395604e-05,
+      "loss": 1.3141,
+      "step": 256
+    },
+    {
+      "epoch": 0.9413919413919414,
+      "grad_norm": 37.11213684082031,
+      "learning_rate": 5.6263736263736264e-05,
+      "loss": 1.7168,
+      "step": 257
+    },
+    {
+      "epoch": 0.945054945054945,
+      "grad_norm": 13.693246841430664,
+      "learning_rate": 5.6483516483516485e-05,
+      "loss": 0.3207,
+      "step": 258
+    },
+    {
+      "epoch": 0.9487179487179487,
+      "grad_norm": 18.186216354370117,
+      "learning_rate": 5.6703296703296706e-05,
+      "loss": 0.6265,
+      "step": 259
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 23.68426513671875,
+      "learning_rate": 5.692307692307692e-05,
+      "loss": 0.5226,
+      "step": 260
+    },
+    {
+      "epoch": 0.9560439560439561,
+      "grad_norm": 19.154836654663086,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 1.0116,
+      "step": 261
+    },
+    {
+      "epoch": 0.9597069597069597,
+      "grad_norm": 17.64719009399414,
+      "learning_rate": 5.736263736263737e-05,
+      "loss": 0.5992,
+      "step": 262
+    },
+    {
+      "epoch": 0.9633699633699634,
+      "grad_norm": 25.542757034301758,
+      "learning_rate": 5.7582417582417584e-05,
+      "loss": 0.8129,
+      "step": 263
+    },
+    {
+      "epoch": 0.967032967032967,
+      "grad_norm": 25.94204330444336,
+      "learning_rate": 5.7802197802197805e-05,
+      "loss": 1.2194,
+      "step": 264
+    },
+    {
+      "epoch": 0.9706959706959707,
+      "grad_norm": 13.693342208862305,
+      "learning_rate": 5.8021978021978026e-05,
+      "loss": 0.2565,
+      "step": 265
+    },
+    {
+      "epoch": 0.9743589743589743,
+      "grad_norm": 20.760122299194336,
+      "learning_rate": 5.824175824175824e-05,
+      "loss": 0.4023,
+      "step": 266
+    },
+    {
+      "epoch": 0.978021978021978,
+      "grad_norm": 20.00895118713379,
+      "learning_rate": 5.846153846153846e-05,
+      "loss": 0.2468,
+      "step": 267
+    },
+    {
+      "epoch": 0.9816849816849816,
+      "grad_norm": 25.56069564819336,
+      "learning_rate": 5.868131868131868e-05,
+      "loss": 0.5648,
+      "step": 268
+    },
+    {
+      "epoch": 0.9853479853479854,
+      "grad_norm": 38.19970703125,
+      "learning_rate": 5.89010989010989e-05,
+      "loss": 0.544,
+      "step": 269
+    },
+    {
+      "epoch": 0.989010989010989,
+      "grad_norm": 37.63619613647461,
+      "learning_rate": 5.9120879120879124e-05,
+      "loss": 0.7556,
+      "step": 270
+    },
+    {
+      "epoch": 0.9926739926739927,
+      "grad_norm": 10.586868286132812,
+      "learning_rate": 5.9340659340659345e-05,
+      "loss": 0.1003,
+      "step": 271
+    },
+    {
+      "epoch": 0.9963369963369964,
+      "grad_norm": 17.579208374023438,
+      "learning_rate": 5.9560439560439566e-05,
+      "loss": 0.2931,
+      "step": 272
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 24.657121658325195,
+      "learning_rate": 5.978021978021978e-05,
+      "loss": 0.2372,
+      "step": 273
+    },
+    {
+      "epoch": 1.0036630036630036,
+      "grad_norm": 29.52134895324707,
+      "learning_rate": 6e-05,
+      "loss": 0.5077,
+      "step": 274
+    },
+    {
+      "epoch": 1.0073260073260073,
+      "grad_norm": 51.900062561035156,
+      "learning_rate": 5.997557997557998e-05,
+      "loss": 0.4404,
+      "step": 275
+    },
+    {
+      "epoch": 1.010989010989011,
+      "grad_norm": 18.682769775390625,
+      "learning_rate": 5.995115995115995e-05,
+      "loss": 0.2405,
+      "step": 276
+    },
+    {
+      "epoch": 1.0146520146520146,
+      "grad_norm": 87.95014953613281,
+      "learning_rate": 5.992673992673993e-05,
+      "loss": 2.8585,
+      "step": 277
+    },
+    {
+      "epoch": 1.0183150183150182,
+      "grad_norm": 67.03990936279297,
+      "learning_rate": 5.990231990231991e-05,
+      "loss": 0.9746,
+      "step": 278
+    },
+    {
+      "epoch": 1.021978021978022,
+      "grad_norm": 47.63545227050781,
+      "learning_rate": 5.987789987789988e-05,
+      "loss": 0.241,
+      "step": 279
+    },
+    {
+      "epoch": 1.0256410256410255,
+      "grad_norm": 33.62876892089844,
+      "learning_rate": 5.985347985347986e-05,
+      "loss": 1.0003,
+      "step": 280
+    },
+    {
+      "epoch": 1.0293040293040292,
+      "grad_norm": 30.26620864868164,
+      "learning_rate": 5.982905982905983e-05,
+      "loss": 0.7767,
+      "step": 281
+    },
+    {
+      "epoch": 1.032967032967033,
+      "grad_norm": 33.785770416259766,
+      "learning_rate": 5.98046398046398e-05,
+      "loss": 0.899,
+      "step": 282
+    },
+    {
+      "epoch": 1.0366300366300367,
+      "grad_norm": 33.753849029541016,
+      "learning_rate": 5.978021978021978e-05,
+      "loss": 1.8225,
+      "step": 283
+    },
+    {
+      "epoch": 1.0402930402930404,
+      "grad_norm": 16.58989143371582,
+      "learning_rate": 5.975579975579976e-05,
+      "loss": 0.6211,
+      "step": 284
+    },
+    {
+      "epoch": 1.043956043956044,
+      "grad_norm": 23.08768653869629,
+      "learning_rate": 5.973137973137973e-05,
+      "loss": 0.7541,
+      "step": 285
+    },
+    {
+      "epoch": 1.0476190476190477,
+      "grad_norm": 24.57805824279785,
+      "learning_rate": 5.970695970695971e-05,
+      "loss": 0.8278,
+      "step": 286
+    },
+    {
+      "epoch": 1.0512820512820513,
+      "grad_norm": 25.1593017578125,
+      "learning_rate": 5.968253968253968e-05,
+      "loss": 0.6932,
+      "step": 287
+    },
+    {
+      "epoch": 1.054945054945055,
+      "grad_norm": 29.984054565429688,
+      "learning_rate": 5.965811965811966e-05,
+      "loss": 0.6987,
+      "step": 288
+    },
+    {
+      "epoch": 1.0586080586080586,
+      "grad_norm": 28.183151245117188,
+      "learning_rate": 5.963369963369964e-05,
+      "loss": 0.8771,
+      "step": 289
+    },
+    {
+      "epoch": 1.0622710622710623,
+      "grad_norm": 15.349969863891602,
+      "learning_rate": 5.960927960927961e-05,
+      "loss": 0.2906,
+      "step": 290
+    },
+    {
+      "epoch": 1.065934065934066,
+      "grad_norm": 17.618196487426758,
+      "learning_rate": 5.958485958485959e-05,
+      "loss": 0.595,
+      "step": 291
+    },
+    {
+      "epoch": 1.0695970695970696,
+      "grad_norm": 40.537925720214844,
+      "learning_rate": 5.9560439560439566e-05,
+      "loss": 1.3881,
+      "step": 292
+    },
+    {
+      "epoch": 1.0732600732600732,
+      "grad_norm": 41.12261962890625,
+      "learning_rate": 5.953601953601954e-05,
+      "loss": 0.5402,
+      "step": 293
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 38.4654655456543,
+      "learning_rate": 5.951159951159951e-05,
+      "loss": 0.3097,
+      "step": 294
+    },
+    {
+      "epoch": 1.0805860805860805,
+      "grad_norm": 34.19886016845703,
+      "learning_rate": 5.948717948717949e-05,
+      "loss": 1.0228,
+      "step": 295
+    },
+    {
+      "epoch": 1.0842490842490842,
+      "grad_norm": 19.727413177490234,
+      "learning_rate": 5.946275946275946e-05,
+      "loss": 0.1755,
+      "step": 296
+    },
+    {
+      "epoch": 1.0879120879120878,
+      "grad_norm": 33.413352966308594,
+      "learning_rate": 5.943833943833944e-05,
+      "loss": 0.8087,
+      "step": 297
+    },
+    {
+      "epoch": 1.0915750915750915,
+      "grad_norm": 29.848875045776367,
+      "learning_rate": 5.941391941391942e-05,
+      "loss": 0.673,
+      "step": 298
+    },
+    {
+      "epoch": 1.0952380952380953,
+      "grad_norm": 18.643922805786133,
+      "learning_rate": 5.938949938949939e-05,
+      "loss": 0.4759,
+      "step": 299
+    },
+    {
+      "epoch": 1.098901098901099,
+      "grad_norm": 28.923099517822266,
+      "learning_rate": 5.936507936507937e-05,
+      "loss": 0.6555,
+      "step": 300
+    },
+    {
+      "epoch": 1.1025641025641026,
+      "grad_norm": 26.4990177154541,
+      "learning_rate": 5.9340659340659345e-05,
+      "loss": 0.4679,
+      "step": 301
+    },
+    {
+      "epoch": 1.1062271062271063,
+      "grad_norm": 43.54881286621094,
+      "learning_rate": 5.931623931623932e-05,
+      "loss": 1.0861,
+      "step": 302
+    },
+    {
+      "epoch": 1.10989010989011,
+      "grad_norm": 32.66098403930664,
+      "learning_rate": 5.9291819291819295e-05,
+      "loss": 0.677,
+      "step": 303
+    },
+    {
+      "epoch": 1.1135531135531136,
+      "grad_norm": 43.79314422607422,
+      "learning_rate": 5.9267399267399274e-05,
+      "loss": 0.8883,
+      "step": 304
+    },
+    {
+      "epoch": 1.1172161172161172,
+      "grad_norm": 44.49085235595703,
+      "learning_rate": 5.9242979242979245e-05,
+      "loss": 0.9553,
+      "step": 305
+    },
+    {
+      "epoch": 1.120879120879121,
+      "grad_norm": 31.713787078857422,
+      "learning_rate": 5.9218559218559224e-05,
+      "loss": 0.6352,
+      "step": 306
+    },
+    {
+      "epoch": 1.1245421245421245,
+      "grad_norm": 19.930402755737305,
+      "learning_rate": 5.9194139194139196e-05,
+      "loss": 0.7023,
+      "step": 307
+    },
+    {
+      "epoch": 1.1282051282051282,
+      "grad_norm": 20.157196044921875,
+      "learning_rate": 5.916971916971917e-05,
+      "loss": 0.6241,
+      "step": 308
+    },
+    {
+      "epoch": 1.1318681318681318,
+      "grad_norm": 26.819135665893555,
+      "learning_rate": 5.9145299145299146e-05,
+      "loss": 0.4788,
+      "step": 309
+    },
+    {
+      "epoch": 1.1355311355311355,
+      "grad_norm": 24.948625564575195,
+      "learning_rate": 5.9120879120879124e-05,
+      "loss": 0.698,
+      "step": 310
+    },
+    {
+      "epoch": 1.1391941391941391,
+      "grad_norm": 15.883389472961426,
+      "learning_rate": 5.9096459096459096e-05,
+      "loss": 0.3325,
+      "step": 311
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 25.214584350585938,
+      "learning_rate": 5.9072039072039074e-05,
+      "loss": 0.4776,
+      "step": 312
+    },
+    {
+      "epoch": 1.1465201465201464,
+      "grad_norm": 27.4523983001709,
+      "learning_rate": 5.9047619047619046e-05,
+      "loss": 0.6155,
+      "step": 313
+    },
+    {
+      "epoch": 1.15018315018315,
+      "grad_norm": 48.60593795776367,
+      "learning_rate": 5.9023199023199024e-05,
+      "loss": 1.7225,
+      "step": 314
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 27.19314193725586,
+      "learning_rate": 5.8998778998779e-05,
+      "loss": 0.6805,
+      "step": 315
+    },
+    {
+      "epoch": 1.1575091575091574,
+      "grad_norm": 44.678768157958984,
+      "learning_rate": 5.8974358974358975e-05,
+      "loss": 0.5721,
+      "step": 316
+    },
+    {
+      "epoch": 1.1611721611721613,
+      "grad_norm": 12.109644889831543,
+      "learning_rate": 5.894993894993895e-05,
+      "loss": 0.1079,
+      "step": 317
+    },
+    {
+      "epoch": 1.164835164835165,
+      "grad_norm": 45.254730224609375,
+      "learning_rate": 5.892551892551893e-05,
+      "loss": 1.1492,
+      "step": 318
+    },
+    {
+      "epoch": 1.1684981684981686,
+      "grad_norm": 65.83439636230469,
+      "learning_rate": 5.89010989010989e-05,
+      "loss": 0.7049,
+      "step": 319
+    },
+    {
+      "epoch": 1.1721611721611722,
+      "grad_norm": 43.5418586730957,
+      "learning_rate": 5.8876678876678875e-05,
+      "loss": 0.4628,
+      "step": 320
+    },
+    {
+      "epoch": 1.1758241758241759,
+      "grad_norm": 137.285400390625,
+      "learning_rate": 5.885225885225885e-05,
+      "loss": 1.4227,
+      "step": 321
+    },
+    {
+      "epoch": 1.1794871794871795,
+      "grad_norm": 42.895565032958984,
+      "learning_rate": 5.8827838827838825e-05,
+      "loss": 0.4264,
+      "step": 322
+    },
+    {
+      "epoch": 1.1831501831501832,
+      "grad_norm": 10.602986335754395,
+      "learning_rate": 5.8803418803418803e-05,
+      "loss": 0.0494,
+      "step": 323
+    },
+    {
+      "epoch": 1.1868131868131868,
+      "grad_norm": 103.92290496826172,
+      "learning_rate": 5.877899877899878e-05,
+      "loss": 2.0111,
+      "step": 324
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 36.497764587402344,
+      "learning_rate": 5.8754578754578754e-05,
+      "loss": 0.4768,
+      "step": 325
+    },
+    {
+      "epoch": 1.1941391941391941,
+      "grad_norm": 45.52228546142578,
+      "learning_rate": 5.873015873015873e-05,
+      "loss": 0.994,
+      "step": 326
+    },
+    {
+      "epoch": 1.1978021978021978,
+      "grad_norm": 24.81894302368164,
+      "learning_rate": 5.870573870573871e-05,
+      "loss": 0.5563,
+      "step": 327
+    },
+    {
+      "epoch": 1.2014652014652014,
+      "grad_norm": 49.82950210571289,
+      "learning_rate": 5.868131868131868e-05,
+      "loss": 1.5448,
+      "step": 328
+    },
+    {
+      "epoch": 1.205128205128205,
+      "grad_norm": 23.945913314819336,
+      "learning_rate": 5.865689865689866e-05,
+      "loss": 0.5256,
+      "step": 329
+    },
+    {
+      "epoch": 1.2087912087912087,
+      "grad_norm": 20.63251304626465,
+      "learning_rate": 5.863247863247864e-05,
+      "loss": 0.3698,
+      "step": 330
+    },
+    {
+      "epoch": 1.2124542124542124,
+      "grad_norm": 32.270328521728516,
+      "learning_rate": 5.860805860805861e-05,
+      "loss": 0.3518,
+      "step": 331
+    },
+    {
+      "epoch": 1.2161172161172162,
+      "grad_norm": 32.445716857910156,
+      "learning_rate": 5.858363858363858e-05,
+      "loss": 0.857,
+      "step": 332
+    },
+    {
+      "epoch": 1.2197802197802199,
+      "grad_norm": 59.69521713256836,
+      "learning_rate": 5.855921855921856e-05,
+      "loss": 1.3786,
+      "step": 333
+    },
+    {
+      "epoch": 1.2234432234432235,
+      "grad_norm": 32.79878234863281,
+      "learning_rate": 5.853479853479853e-05,
+      "loss": 0.7648,
+      "step": 334
+    },
+    {
+      "epoch": 1.2271062271062272,
+      "grad_norm": 26.749393463134766,
+      "learning_rate": 5.851037851037851e-05,
+      "loss": 0.4723,
+      "step": 335
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 40.744102478027344,
+      "learning_rate": 5.848595848595849e-05,
+      "loss": 1.0543,
+      "step": 336
+    },
+    {
+      "epoch": 1.2344322344322345,
+      "grad_norm": 34.2275505065918,
+      "learning_rate": 5.846153846153846e-05,
+      "loss": 0.4533,
+      "step": 337
+    },
+    {
+      "epoch": 1.2380952380952381,
+      "grad_norm": 49.648136138916016,
+      "learning_rate": 5.843711843711844e-05,
+      "loss": 1.2112,
+      "step": 338
+    },
+    {
+      "epoch": 1.2417582417582418,
+      "grad_norm": 64.69720458984375,
+      "learning_rate": 5.841269841269841e-05,
+      "loss": 1.2234,
+      "step": 339
+    },
+    {
+      "epoch": 1.2454212454212454,
+      "grad_norm": 16.81964111328125,
+      "learning_rate": 5.838827838827839e-05,
+      "loss": 0.297,
+      "step": 340
+    },
+    {
+      "epoch": 1.249084249084249,
+      "grad_norm": 17.393678665161133,
+      "learning_rate": 5.836385836385837e-05,
+      "loss": 0.2504,
+      "step": 341
+    },
+    {
+      "epoch": 1.2527472527472527,
+      "grad_norm": 64.2254409790039,
+      "learning_rate": 5.833943833943834e-05,
+      "loss": 1.3656,
+      "step": 342
+    },
+    {
+      "epoch": 1.2564102564102564,
+      "grad_norm": 48.991249084472656,
+      "learning_rate": 5.831501831501832e-05,
+      "loss": 1.0819,
+      "step": 343
+    },
+    {
+      "epoch": 1.26007326007326,
+      "grad_norm": 22.78063201904297,
+      "learning_rate": 5.82905982905983e-05,
+      "loss": 0.1792,
+      "step": 344
+    },
+    {
+      "epoch": 1.2637362637362637,
+      "grad_norm": 35.463233947753906,
+      "learning_rate": 5.826617826617826e-05,
+      "loss": 0.5663,
+      "step": 345
+    },
+    {
+      "epoch": 1.2673992673992673,
+      "grad_norm": 54.528953552246094,
+      "learning_rate": 5.824175824175824e-05,
+      "loss": 1.5814,
+      "step": 346
+    },
+    {
+      "epoch": 1.271062271062271,
+      "grad_norm": 44.60401916503906,
+      "learning_rate": 5.821733821733822e-05,
+      "loss": 0.6471,
+      "step": 347
+    },
+    {
+      "epoch": 1.2747252747252746,
+      "grad_norm": 2.6468827724456787,
+      "learning_rate": 5.819291819291819e-05,
+      "loss": 0.0288,
+      "step": 348
+    },
+    {
+      "epoch": 1.2783882783882783,
+      "grad_norm": 21.465364456176758,
+      "learning_rate": 5.816849816849817e-05,
+      "loss": 0.5259,
+      "step": 349
+    },
+    {
+      "epoch": 1.282051282051282,
+      "grad_norm": 51.20866012573242,
+      "learning_rate": 5.814407814407815e-05,
+      "loss": 0.8054,
+      "step": 350
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 33.52774429321289,
+      "learning_rate": 5.811965811965812e-05,
+      "loss": 0.494,
+      "step": 351
+    },
+    {
+      "epoch": 1.2893772893772895,
+      "grad_norm": 39.15644836425781,
+      "learning_rate": 5.80952380952381e-05,
+      "loss": 1.6315,
+      "step": 352
+    },
+    {
+      "epoch": 1.293040293040293,
+      "grad_norm": 24.35202407836914,
+      "learning_rate": 5.8070818070818076e-05,
+      "loss": 0.6189,
+      "step": 353
+    },
+    {
+      "epoch": 1.2967032967032968,
+      "grad_norm": 39.99496841430664,
+      "learning_rate": 5.804639804639805e-05,
+      "loss": 1.2323,
+      "step": 354
+    },
+    {
+      "epoch": 1.3003663003663004,
+      "grad_norm": 26.282432556152344,
+      "learning_rate": 5.8021978021978026e-05,
+      "loss": 0.5383,
+      "step": 355
+    },
+    {
+      "epoch": 1.304029304029304,
+      "grad_norm": 36.909969329833984,
+      "learning_rate": 5.7997557997558004e-05,
+      "loss": 1.6886,
+      "step": 356
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 18.90056037902832,
+      "learning_rate": 5.7973137973137976e-05,
+      "loss": 0.7226,
+      "step": 357
+    },
+    {
+      "epoch": 1.3113553113553114,
+      "grad_norm": 21.10304832458496,
+      "learning_rate": 5.794871794871795e-05,
+      "loss": 0.8914,
+      "step": 358
+    },
+    {
+      "epoch": 1.315018315018315,
+      "grad_norm": 18.380769729614258,
+      "learning_rate": 5.7924297924297926e-05,
+      "loss": 1.4304,
+      "step": 359
+    },
+    {
+      "epoch": 1.3186813186813187,
+      "grad_norm": 17.992050170898438,
+      "learning_rate": 5.78998778998779e-05,
+      "loss": 1.0023,
+      "step": 360
+    },
+    {
+      "epoch": 1.3223443223443223,
+      "grad_norm": 17.944400787353516,
+      "learning_rate": 5.7875457875457876e-05,
+      "loss": 0.7734,
+      "step": 361
+    },
+    {
+      "epoch": 1.326007326007326,
+      "grad_norm": 19.117143630981445,
+      "learning_rate": 5.7851037851037855e-05,
+      "loss": 0.6923,
+      "step": 362
+    },
+    {
+      "epoch": 1.3296703296703296,
+      "grad_norm": 21.4644718170166,
+      "learning_rate": 5.7826617826617826e-05,
+      "loss": 0.666,
+      "step": 363
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 25.951030731201172,
+      "learning_rate": 5.7802197802197805e-05,
+      "loss": 1.522,
+      "step": 364
+    },
+    {
+      "epoch": 1.3369963369963371,
+      "grad_norm": 32.20412063598633,
+      "learning_rate": 5.7777777777777776e-05,
+      "loss": 1.5771,
+      "step": 365
+    },
+    {
+      "epoch": 1.3406593406593408,
+      "grad_norm": 26.847576141357422,
+      "learning_rate": 5.7753357753357755e-05,
+      "loss": 1.3427,
+      "step": 366
+    },
+    {
+      "epoch": 1.3443223443223444,
+      "grad_norm": 18.596710205078125,
+      "learning_rate": 5.772893772893773e-05,
+      "loss": 0.5533,
+      "step": 367
+    },
+    {
+      "epoch": 1.347985347985348,
+      "grad_norm": 23.6543025970459,
+      "learning_rate": 5.7704517704517705e-05,
+      "loss": 0.581,
+      "step": 368
+    },
+    {
+      "epoch": 1.3516483516483517,
+      "grad_norm": 13.732353210449219,
+      "learning_rate": 5.7680097680097684e-05,
+      "loss": 0.1908,
+      "step": 369
+    },
+    {
+      "epoch": 1.3553113553113554,
+      "grad_norm": 21.231159210205078,
+      "learning_rate": 5.765567765567766e-05,
+      "loss": 0.5858,
+      "step": 370
+    },
+    {
+      "epoch": 1.358974358974359,
+      "grad_norm": 18.647363662719727,
+      "learning_rate": 5.763125763125763e-05,
+      "loss": 0.6205,
+      "step": 371
+    },
+    {
+      "epoch": 1.3626373626373627,
+      "grad_norm": 20.302942276000977,
+      "learning_rate": 5.7606837606837605e-05,
+      "loss": 0.3637,
+      "step": 372
+    },
+    {
+      "epoch": 1.3663003663003663,
+      "grad_norm": 18.72137451171875,
+      "learning_rate": 5.7582417582417584e-05,
+      "loss": 0.2262,
+      "step": 373
+    },
+    {
+      "epoch": 1.36996336996337,
+      "grad_norm": 32.225738525390625,
+      "learning_rate": 5.7557997557997555e-05,
+      "loss": 0.5696,
+      "step": 374
+    },
+    {
+      "epoch": 1.3736263736263736,
+      "grad_norm": 21.453779220581055,
+      "learning_rate": 5.7533577533577534e-05,
+      "loss": 0.3533,
+      "step": 375
+    },
+    {
+      "epoch": 1.3772893772893773,
+      "grad_norm": 26.601511001586914,
+      "learning_rate": 5.750915750915751e-05,
+      "loss": 0.438,
+      "step": 376
+    },
+    {
+      "epoch": 1.380952380952381,
+      "grad_norm": 49.10448455810547,
+      "learning_rate": 5.7484737484737484e-05,
+      "loss": 0.6742,
+      "step": 377
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 51.251136779785156,
+      "learning_rate": 5.746031746031746e-05,
+      "loss": 0.7096,
+      "step": 378
+    },
+    {
+      "epoch": 1.3882783882783882,
+      "grad_norm": 35.14614486694336,
+      "learning_rate": 5.743589743589744e-05,
+      "loss": 1.5348,
+      "step": 379
+    },
+    {
+      "epoch": 1.3919413919413919,
+      "grad_norm": 58.83134078979492,
+      "learning_rate": 5.741147741147741e-05,
+      "loss": 1.303,
+      "step": 380
+    },
+    {
+      "epoch": 1.3956043956043955,
+      "grad_norm": 34.27029800415039,
+      "learning_rate": 5.738705738705739e-05,
+      "loss": 0.3682,
+      "step": 381
+    },
+    {
+      "epoch": 1.3992673992673992,
+      "grad_norm": 59.508628845214844,
+      "learning_rate": 5.736263736263737e-05,
+      "loss": 0.6489,
+      "step": 382
+    },
+    {
+      "epoch": 1.4029304029304028,
+      "grad_norm": 24.804059982299805,
+      "learning_rate": 5.733821733821734e-05,
+      "loss": 0.325,
+      "step": 383
+    },
+    {
+      "epoch": 1.4065934065934065,
+      "grad_norm": 20.69612693786621,
+      "learning_rate": 5.731379731379731e-05,
+      "loss": 0.1529,
+      "step": 384
+    },
+    {
+      "epoch": 1.4102564102564101,
+      "grad_norm": 29.134044647216797,
+      "learning_rate": 5.728937728937729e-05,
+      "loss": 0.8694,
+      "step": 385
+    },
+    {
+      "epoch": 1.4139194139194138,
+      "grad_norm": 37.44430923461914,
+      "learning_rate": 5.726495726495726e-05,
+      "loss": 0.9174,
+      "step": 386
+    },
+    {
+      "epoch": 1.4175824175824177,
+      "grad_norm": 36.84721755981445,
+      "learning_rate": 5.724053724053724e-05,
+      "loss": 0.3522,
+      "step": 387
+    },
+    {
+      "epoch": 1.4212454212454213,
+      "grad_norm": 44.15989685058594,
+      "learning_rate": 5.721611721611722e-05,
+      "loss": 1.4677,
+      "step": 388
+    },
+    {
+      "epoch": 1.424908424908425,
+      "grad_norm": 16.73012351989746,
+      "learning_rate": 5.719169719169719e-05,
+      "loss": 0.1621,
+      "step": 389
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 35.41815185546875,
+      "learning_rate": 5.716727716727717e-05,
+      "loss": 0.6702,
+      "step": 390
+    },
+    {
+      "epoch": 1.4322344322344323,
+      "grad_norm": 19.04936408996582,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 0.1845,
+      "step": 391
+    },
+    {
+      "epoch": 1.435897435897436,
+      "grad_norm": 22.89434242248535,
+      "learning_rate": 5.711843711843712e-05,
+      "loss": 0.5694,
+      "step": 392
+    },
+    {
+      "epoch": 1.4395604395604396,
+      "grad_norm": 22.125951766967773,
+      "learning_rate": 5.70940170940171e-05,
+      "loss": 0.821,
+      "step": 393
+    },
+    {
+      "epoch": 1.4432234432234432,
+      "grad_norm": 37.83376693725586,
+      "learning_rate": 5.706959706959707e-05,
+      "loss": 0.4658,
+      "step": 394
+    },
+    {
+      "epoch": 1.4468864468864469,
+      "grad_norm": 38.37764358520508,
+      "learning_rate": 5.704517704517705e-05,
+      "loss": 0.4146,
+      "step": 395
+    },
+    {
+      "epoch": 1.4505494505494505,
+      "grad_norm": 21.50092315673828,
+      "learning_rate": 5.702075702075703e-05,
+      "loss": 0.5044,
+      "step": 396
+    },
+    {
+      "epoch": 1.4542124542124542,
+      "grad_norm": 20.02173614501953,
+      "learning_rate": 5.699633699633699e-05,
+      "loss": 0.4955,
+      "step": 397
+    },
+    {
+      "epoch": 1.4578754578754578,
+      "grad_norm": 21.474336624145508,
+      "learning_rate": 5.697191697191697e-05,
+      "loss": 0.3818,
+      "step": 398
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 22.903839111328125,
+      "learning_rate": 5.694749694749695e-05,
+      "loss": 0.7603,
+      "step": 399
+    },
+    {
+      "epoch": 1.4652014652014653,
+      "grad_norm": 20.22893524169922,
+      "learning_rate": 5.692307692307692e-05,
+      "loss": 0.5612,
+      "step": 400
+    },
+    {
+      "epoch": 1.468864468864469,
+      "grad_norm": 32.34550857543945,
+      "learning_rate": 5.68986568986569e-05,
+      "loss": 0.4659,
+      "step": 401
+    },
+    {
+      "epoch": 1.4725274725274726,
+      "grad_norm": 49.979034423828125,
+      "learning_rate": 5.687423687423688e-05,
+      "loss": 0.6784,
+      "step": 402
+    },
+    {
+      "epoch": 1.4761904761904763,
+      "grad_norm": 79.79581451416016,
+      "learning_rate": 5.684981684981685e-05,
+      "loss": 0.9404,
+      "step": 403
+    },
+    {
+      "epoch": 1.47985347985348,
+      "grad_norm": 17.678560256958008,
+      "learning_rate": 5.682539682539683e-05,
+      "loss": 0.1675,
+      "step": 404
+    },
+    {
+      "epoch": 1.4835164835164836,
+      "grad_norm": 21.246519088745117,
+      "learning_rate": 5.6800976800976806e-05,
+      "loss": 0.2428,
+      "step": 405
+    },
+    {
+      "epoch": 1.4871794871794872,
+      "grad_norm": 34.815452575683594,
+      "learning_rate": 5.677655677655678e-05,
+      "loss": 0.3925,
+      "step": 406
+    },
+    {
+      "epoch": 1.4908424908424909,
+      "grad_norm": 73.8591079711914,
+      "learning_rate": 5.6752136752136756e-05,
+      "loss": 1.3163,
+      "step": 407
+    },
+    {
+      "epoch": 1.4945054945054945,
+      "grad_norm": 66.63922882080078,
+      "learning_rate": 5.6727716727716735e-05,
+      "loss": 0.9653,
+      "step": 408
+    },
+    {
+      "epoch": 1.4981684981684982,
+      "grad_norm": 52.39488220214844,
+      "learning_rate": 5.6703296703296706e-05,
+      "loss": 0.9322,
+      "step": 409
+    },
+    {
+      "epoch": 1.5018315018315018,
+      "grad_norm": 13.078998565673828,
+      "learning_rate": 5.667887667887668e-05,
+      "loss": 0.1168,
+      "step": 410
+    },
+    {
+      "epoch": 1.5054945054945055,
+      "grad_norm": 41.32448959350586,
+      "learning_rate": 5.6654456654456657e-05,
+      "loss": 0.9296,
+      "step": 411
+    },
+    {
+      "epoch": 1.5091575091575091,
+      "grad_norm": 26.448543548583984,
+      "learning_rate": 5.663003663003663e-05,
+      "loss": 0.5474,
+      "step": 412
+    },
+    {
+      "epoch": 1.5128205128205128,
+      "grad_norm": 29.58432960510254,
+      "learning_rate": 5.660561660561661e-05,
+      "loss": 0.6573,
+      "step": 413
+    },
+    {
+      "epoch": 1.5164835164835164,
+      "grad_norm": 28.568214416503906,
+      "learning_rate": 5.6581196581196585e-05,
+      "loss": 0.9223,
+      "step": 414
+    },
+    {
+      "epoch": 1.52014652014652,
+      "grad_norm": 31.92661476135254,
+      "learning_rate": 5.655677655677656e-05,
+      "loss": 1.0601,
+      "step": 415
+    },
+    {
+      "epoch": 1.5238095238095237,
+      "grad_norm": 31.934263229370117,
+      "learning_rate": 5.6532356532356535e-05,
+      "loss": 0.6288,
+      "step": 416
+    },
+    {
+      "epoch": 1.5274725274725274,
+      "grad_norm": 21.51350975036621,
+      "learning_rate": 5.650793650793651e-05,
+      "loss": 0.7378,
+      "step": 417
+    },
+    {
+      "epoch": 1.531135531135531,
+      "grad_norm": 19.010095596313477,
+      "learning_rate": 5.6483516483516485e-05,
+      "loss": 0.7792,
+      "step": 418
+    },
+    {
+      "epoch": 1.5347985347985347,
+      "grad_norm": 21.7001895904541,
+      "learning_rate": 5.6459096459096464e-05,
+      "loss": 0.7885,
+      "step": 419
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 21.400882720947266,
+      "learning_rate": 5.6434676434676436e-05,
+      "loss": 0.942,
+      "step": 420
+    },
+    {
+      "epoch": 1.542124542124542,
+      "grad_norm": 30.14664649963379,
+      "learning_rate": 5.6410256410256414e-05,
+      "loss": 0.7675,
+      "step": 421
+    },
+    {
+      "epoch": 1.5457875457875456,
+      "grad_norm": 33.25088882446289,
+      "learning_rate": 5.6385836385836386e-05,
+      "loss": 1.1349,
+      "step": 422
+    },
+    {
+      "epoch": 1.5494505494505495,
+      "grad_norm": 22.923208236694336,
+      "learning_rate": 5.636141636141636e-05,
+      "loss": 0.7145,
+      "step": 423
+    },
+    {
+      "epoch": 1.5531135531135531,
+      "grad_norm": 20.00519371032715,
+      "learning_rate": 5.6336996336996336e-05,
+      "loss": 0.5107,
+      "step": 424
+    },
+    {
+      "epoch": 1.5567765567765568,
+      "grad_norm": 21.95383071899414,
+      "learning_rate": 5.6312576312576314e-05,
+      "loss": 0.7836,
+      "step": 425
+    },
+    {
+      "epoch": 1.5604395604395604,
+      "grad_norm": 27.24031639099121,
+      "learning_rate": 5.6288156288156286e-05,
+      "loss": 0.4955,
+      "step": 426
+    },
+    {
+      "epoch": 1.564102564102564,
+      "grad_norm": 45.48428726196289,
+      "learning_rate": 5.6263736263736264e-05,
+      "loss": 1.016,
+      "step": 427
+    },
+    {
+      "epoch": 1.5677655677655677,
+      "grad_norm": 20.055965423583984,
+      "learning_rate": 5.623931623931624e-05,
+      "loss": 0.325,
+      "step": 428
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 22.020767211914062,
+      "learning_rate": 5.6214896214896215e-05,
+      "loss": 0.45,
+      "step": 429
+    },
+    {
+      "epoch": 1.575091575091575,
+      "grad_norm": 32.608741760253906,
+      "learning_rate": 5.619047619047619e-05,
+      "loss": 0.6561,
+      "step": 430
+    },
+    {
+      "epoch": 1.578754578754579,
+      "grad_norm": 38.14396667480469,
+      "learning_rate": 5.616605616605617e-05,
+      "loss": 0.6387,
+      "step": 431
+    },
+    {
+      "epoch": 1.5824175824175826,
+      "grad_norm": 26.266948699951172,
+      "learning_rate": 5.614163614163614e-05,
+      "loss": 0.5593,
+      "step": 432
+    },
+    {
+      "epoch": 1.5860805860805862,
+      "grad_norm": 16.37360954284668,
+      "learning_rate": 5.611721611721612e-05,
+      "loss": 0.1591,
+      "step": 433
+    },
+    {
+      "epoch": 1.5897435897435899,
+      "grad_norm": 21.9448299407959,
+      "learning_rate": 5.60927960927961e-05,
+      "loss": 0.2129,
+      "step": 434
+    },
+    {
+      "epoch": 1.5934065934065935,
+      "grad_norm": 30.096052169799805,
+      "learning_rate": 5.6068376068376065e-05,
+      "loss": 0.3384,
+      "step": 435
+    },
+    {
+      "epoch": 1.5970695970695972,
+      "grad_norm": 40.15864181518555,
+      "learning_rate": 5.604395604395604e-05,
+      "loss": 0.5181,
+      "step": 436
+    },
+    {
+      "epoch": 1.6007326007326008,
+      "grad_norm": 63.40933609008789,
+      "learning_rate": 5.601953601953602e-05,
+      "loss": 0.8834,
+      "step": 437
+    },
+    {
+      "epoch": 1.6043956043956045,
+      "grad_norm": 40.0787353515625,
+      "learning_rate": 5.5995115995115993e-05,
+      "loss": 0.437,
+      "step": 438
+    },
+    {
+      "epoch": 1.6080586080586081,
+      "grad_norm": 40.136863708496094,
+      "learning_rate": 5.597069597069597e-05,
+      "loss": 0.4834,
+      "step": 439
+    },
+    {
+      "epoch": 1.6117216117216118,
+      "grad_norm": 27.898317337036133,
+      "learning_rate": 5.594627594627595e-05,
+      "loss": 0.4862,
+      "step": 440
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 31.5762882232666,
+      "learning_rate": 5.592185592185592e-05,
+      "loss": 0.1878,
+      "step": 441
+    },
+    {
+      "epoch": 1.619047619047619,
+      "grad_norm": 88.90093994140625,
+      "learning_rate": 5.58974358974359e-05,
+      "loss": 1.3343,
+      "step": 442
+    },
+    {
+      "epoch": 1.6227106227106227,
+      "grad_norm": 57.7340202331543,
+      "learning_rate": 5.587301587301587e-05,
+      "loss": 0.3032,
+      "step": 443
+    },
+    {
+      "epoch": 1.6263736263736264,
+      "grad_norm": 57.28425979614258,
+      "learning_rate": 5.584859584859585e-05,
+      "loss": 1.3972,
+      "step": 444
+    },
+    {
+      "epoch": 1.63003663003663,
+      "grad_norm": 39.866302490234375,
+      "learning_rate": 5.582417582417583e-05,
+      "loss": 0.4026,
+      "step": 445
+    },
+    {
+      "epoch": 1.6336996336996337,
+      "grad_norm": 41.72932815551758,
+      "learning_rate": 5.57997557997558e-05,
+      "loss": 0.5407,
+      "step": 446
+    },
+    {
+      "epoch": 1.6373626373626373,
+      "grad_norm": 60.77634811401367,
+      "learning_rate": 5.577533577533578e-05,
+      "loss": 0.8581,
+      "step": 447
+    },
+    {
+      "epoch": 1.641025641025641,
+      "grad_norm": 28.382030487060547,
+      "learning_rate": 5.575091575091575e-05,
+      "loss": 0.3759,
+      "step": 448
+    },
+    {
+      "epoch": 1.6446886446886446,
+      "grad_norm": 62.1085205078125,
+      "learning_rate": 5.572649572649572e-05,
+      "loss": 1.0749,
+      "step": 449
+    },
+    {
+      "epoch": 1.6483516483516483,
+      "grad_norm": 41.8302001953125,
+      "learning_rate": 5.57020757020757e-05,
+      "loss": 0.5884,
+      "step": 450
+    },
+    {
+      "epoch": 1.652014652014652,
+      "grad_norm": 24.128931045532227,
+      "learning_rate": 5.567765567765568e-05,
+      "loss": 0.6113,
+      "step": 451
+    },
+    {
+      "epoch": 1.6556776556776556,
+      "grad_norm": 19.634384155273438,
+      "learning_rate": 5.565323565323565e-05,
+      "loss": 0.3902,
+      "step": 452
+    },
+    {
+      "epoch": 1.6593406593406592,
+      "grad_norm": 18.17875099182129,
+      "learning_rate": 5.562881562881563e-05,
+      "loss": 0.3137,
+      "step": 453
+    },
+    {
+      "epoch": 1.6630036630036629,
+      "grad_norm": 39.68446731567383,
+      "learning_rate": 5.560439560439561e-05,
+      "loss": 0.7587,
+      "step": 454
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 29.387836456298828,
+      "learning_rate": 5.557997557997558e-05,
+      "loss": 0.6397,
+      "step": 455
+    },
+    {
+      "epoch": 1.6703296703296702,
+      "grad_norm": 19.08424949645996,
+      "learning_rate": 5.555555555555556e-05,
+      "loss": 0.2484,
+      "step": 456
+    },
+    {
+      "epoch": 1.673992673992674,
+      "grad_norm": 36.07701873779297,
+      "learning_rate": 5.553113553113554e-05,
+      "loss": 0.8587,
+      "step": 457
+    },
+    {
+      "epoch": 1.6776556776556777,
+      "grad_norm": 52.062339782714844,
+      "learning_rate": 5.550671550671551e-05,
+      "loss": 1.6675,
+      "step": 458
+    },
+    {
+      "epoch": 1.6813186813186813,
+      "grad_norm": 45.415687561035156,
+      "learning_rate": 5.548229548229549e-05,
+      "loss": 1.653,
+      "step": 459
+    },
+    {
+      "epoch": 1.684981684981685,
+      "grad_norm": 31.457420349121094,
+      "learning_rate": 5.5457875457875465e-05,
+      "loss": 0.4578,
+      "step": 460
+    },
+    {
+      "epoch": 1.6886446886446886,
+      "grad_norm": 33.14665603637695,
+      "learning_rate": 5.543345543345543e-05,
+      "loss": 1.3327,
+      "step": 461
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 25.720529556274414,
+      "learning_rate": 5.540903540903541e-05,
+      "loss": 0.5,
+      "step": 462
+    },
+    {
+      "epoch": 1.695970695970696,
+      "grad_norm": 23.71514129638672,
+      "learning_rate": 5.538461538461539e-05,
+      "loss": 0.434,
+      "step": 463
+    },
+    {
+      "epoch": 1.6996336996336996,
+      "grad_norm": 45.231746673583984,
+      "learning_rate": 5.536019536019536e-05,
+      "loss": 0.9448,
+      "step": 464
+    },
+    {
+      "epoch": 1.7032967032967035,
+      "grad_norm": 17.44647789001465,
+      "learning_rate": 5.533577533577534e-05,
+      "loss": 0.3183,
+      "step": 465
+    },
+    {
+      "epoch": 1.7069597069597071,
+      "grad_norm": 18.627901077270508,
+      "learning_rate": 5.531135531135531e-05,
+      "loss": 0.4137,
+      "step": 466
+    },
+    {
+      "epoch": 1.7106227106227108,
+      "grad_norm": 45.57220458984375,
+      "learning_rate": 5.528693528693529e-05,
+      "loss": 1.0096,
+      "step": 467
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 27.329822540283203,
+      "learning_rate": 5.5262515262515266e-05,
+      "loss": 0.5416,
+      "step": 468
+    },
+    {
+      "epoch": 1.717948717948718,
+      "grad_norm": 46.70027160644531,
+      "learning_rate": 5.523809523809524e-05,
+      "loss": 0.983,
+      "step": 469
+    },
+    {
+      "epoch": 1.7216117216117217,
+      "grad_norm": 32.47868728637695,
+      "learning_rate": 5.5213675213675216e-05,
+      "loss": 1.5687,
+      "step": 470
+    },
+    {
+      "epoch": 1.7252747252747254,
+      "grad_norm": 16.49342155456543,
+      "learning_rate": 5.5189255189255194e-05,
+      "loss": 0.3101,
+      "step": 471
+    },
+    {
+      "epoch": 1.728937728937729,
+      "grad_norm": 26.58381462097168,
+      "learning_rate": 5.5164835164835166e-05,
+      "loss": 0.7027,
+      "step": 472
+    },
+    {
+      "epoch": 1.7326007326007327,
+      "grad_norm": 17.435213088989258,
+      "learning_rate": 5.5140415140415144e-05,
+      "loss": 0.3958,
+      "step": 473
+    },
+    {
+      "epoch": 1.7362637362637363,
+      "grad_norm": 19.37874412536621,
+      "learning_rate": 5.5115995115995116e-05,
+      "loss": 0.3979,
+      "step": 474
+    },
+    {
+      "epoch": 1.73992673992674,
+      "grad_norm": 16.509248733520508,
+      "learning_rate": 5.509157509157509e-05,
+      "loss": 0.5121,
+      "step": 475
+    },
+    {
+      "epoch": 1.7435897435897436,
+      "grad_norm": 9.653852462768555,
+      "learning_rate": 5.5067155067155066e-05,
+      "loss": 0.1386,
+      "step": 476
+    },
+    {
+      "epoch": 1.7472527472527473,
+      "grad_norm": 26.486963272094727,
+      "learning_rate": 5.5042735042735045e-05,
+      "loss": 1.0307,
+      "step": 477
+    },
+    {
+      "epoch": 1.750915750915751,
+      "grad_norm": 17.766828536987305,
+      "learning_rate": 5.5018315018315016e-05,
+      "loss": 0.278,
+      "step": 478
+    },
+    {
+      "epoch": 1.7545787545787546,
+      "grad_norm": 12.930633544921875,
+      "learning_rate": 5.4993894993894995e-05,
+      "loss": 0.1487,
+      "step": 479
+    },
+    {
+      "epoch": 1.7582417582417582,
+      "grad_norm": 44.64267349243164,
+      "learning_rate": 5.496947496947497e-05,
+      "loss": 0.7036,
+      "step": 480
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 17.474651336669922,
+      "learning_rate": 5.4945054945054945e-05,
+      "loss": 0.1666,
+      "step": 481
+    },
+    {
+      "epoch": 1.7655677655677655,
+      "grad_norm": 48.3519401550293,
+      "learning_rate": 5.4920634920634923e-05,
+      "loss": 0.6157,
+      "step": 482
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 18.429521560668945,
+      "learning_rate": 5.48962148962149e-05,
+      "loss": 0.2588,
+      "step": 483
+    },
+    {
+      "epoch": 1.7728937728937728,
+      "grad_norm": 66.73760986328125,
+      "learning_rate": 5.4871794871794874e-05,
+      "loss": 0.654,
+      "step": 484
+    },
+    {
+      "epoch": 1.7765567765567765,
+      "grad_norm": 53.831539154052734,
+      "learning_rate": 5.484737484737485e-05,
+      "loss": 0.7538,
+      "step": 485
+    },
+    {
+      "epoch": 1.7802197802197801,
+      "grad_norm": 52.023895263671875,
+      "learning_rate": 5.482295482295483e-05,
+      "loss": 1.6623,
+      "step": 486
+    },
+    {
+      "epoch": 1.7838827838827838,
+      "grad_norm": 38.4475212097168,
+      "learning_rate": 5.4798534798534795e-05,
+      "loss": 0.5079,
+      "step": 487
+    },
+    {
+      "epoch": 1.7875457875457874,
+      "grad_norm": 25.642650604248047,
+      "learning_rate": 5.4774114774114774e-05,
+      "loss": 0.3825,
+      "step": 488
+    },
+    {
+      "epoch": 1.791208791208791,
+      "grad_norm": 57.916900634765625,
+      "learning_rate": 5.474969474969475e-05,
+      "loss": 0.9583,
+      "step": 489
+    },
+    {
+      "epoch": 1.7948717948717947,
+      "grad_norm": 39.23340606689453,
+      "learning_rate": 5.4725274725274724e-05,
+      "loss": 0.4724,
+      "step": 490
+    },
+    {
+      "epoch": 1.7985347985347986,
+      "grad_norm": 24.188661575317383,
+      "learning_rate": 5.47008547008547e-05,
+      "loss": 0.4471,
+      "step": 491
+    },
+    {
+      "epoch": 1.8021978021978022,
+      "grad_norm": 68.73822021484375,
+      "learning_rate": 5.4676434676434674e-05,
+      "loss": 0.6618,
+      "step": 492
+    },
+    {
+      "epoch": 1.8058608058608059,
+      "grad_norm": 26.382184982299805,
+      "learning_rate": 5.465201465201465e-05,
+      "loss": 0.5835,
+      "step": 493
+    },
+    {
+      "epoch": 1.8095238095238095,
+      "grad_norm": 31.758886337280273,
+      "learning_rate": 5.462759462759463e-05,
+      "loss": 0.622,
+      "step": 494
+    },
+    {
+      "epoch": 1.8131868131868132,
+      "grad_norm": 26.657405853271484,
+      "learning_rate": 5.46031746031746e-05,
+      "loss": 0.6003,
+      "step": 495
+    },
+    {
+      "epoch": 1.8168498168498168,
+      "grad_norm": 31.248491287231445,
+      "learning_rate": 5.457875457875458e-05,
+      "loss": 0.4929,
+      "step": 496
+    },
+    {
+      "epoch": 1.8205128205128205,
+      "grad_norm": 53.82766342163086,
+      "learning_rate": 5.455433455433456e-05,
+      "loss": 2.0716,
+      "step": 497
+    },
+    {
+      "epoch": 1.8241758241758241,
+      "grad_norm": 46.39777374267578,
+      "learning_rate": 5.452991452991453e-05,
+      "loss": 1.6767,
+      "step": 498
+    },
+    {
+      "epoch": 1.8278388278388278,
+      "grad_norm": 39.58620071411133,
+      "learning_rate": 5.450549450549451e-05,
+      "loss": 0.8274,
+      "step": 499
+    },
+    {
+      "epoch": 1.8315018315018317,
+      "grad_norm": 29.395286560058594,
+      "learning_rate": 5.448107448107448e-05,
+      "loss": 1.1441,
+      "step": 500
+    },
+    {
+      "epoch": 1.8351648351648353,
+      "grad_norm": 26.250751495361328,
+      "learning_rate": 5.445665445665445e-05,
+      "loss": 0.7496,
+      "step": 501
+    },
+    {
+      "epoch": 1.838827838827839,
+      "grad_norm": 19.820999145507812,
+      "learning_rate": 5.443223443223443e-05,
+      "loss": 0.4367,
+      "step": 502
+    },
+    {
+      "epoch": 1.8424908424908426,
+      "grad_norm": 25.09316062927246,
+      "learning_rate": 5.440781440781441e-05,
+      "loss": 0.8584,
+      "step": 503
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 17.808509826660156,
+      "learning_rate": 5.438339438339438e-05,
+      "loss": 0.3869,
+      "step": 504
+    },
+    {
+      "epoch": 1.84981684981685,
+      "grad_norm": 28.342119216918945,
+      "learning_rate": 5.435897435897436e-05,
+      "loss": 0.8881,
+      "step": 505
+    },
+    {
+      "epoch": 1.8534798534798536,
+      "grad_norm": 33.80287551879883,
+      "learning_rate": 5.433455433455434e-05,
+      "loss": 1.2911,
+      "step": 506
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 55.428138732910156,
+      "learning_rate": 5.431013431013431e-05,
+      "loss": 0.8934,
+      "step": 507
+    },
+    {
+      "epoch": 1.8608058608058609,
+      "grad_norm": 27.962610244750977,
+      "learning_rate": 5.428571428571429e-05,
+      "loss": 0.662,
+      "step": 508
+    },
+    {
+      "epoch": 1.8644688644688645,
+      "grad_norm": 62.84252166748047,
+      "learning_rate": 5.426129426129427e-05,
+      "loss": 1.9216,
+      "step": 509
+    },
+    {
+      "epoch": 1.8681318681318682,
+      "grad_norm": 24.26439666748047,
+      "learning_rate": 5.423687423687424e-05,
+      "loss": 0.2164,
+      "step": 510
+    },
+    {
+      "epoch": 1.8717948717948718,
+      "grad_norm": 50.95674133300781,
+      "learning_rate": 5.421245421245422e-05,
+      "loss": 0.7023,
+      "step": 511
+    },
+    {
+      "epoch": 1.8754578754578755,
+      "grad_norm": 41.17847442626953,
+      "learning_rate": 5.418803418803419e-05,
+      "loss": 1.1081,
+      "step": 512
+    },
+    {
+      "epoch": 1.879120879120879,
+      "grad_norm": 28.701988220214844,
+      "learning_rate": 5.416361416361416e-05,
+      "loss": 0.6519,
+      "step": 513
+    },
+    {
+      "epoch": 1.8827838827838828,
+      "grad_norm": 48.42552947998047,
+      "learning_rate": 5.413919413919414e-05,
+      "loss": 1.5215,
+      "step": 514
+    },
+    {
+      "epoch": 1.8864468864468864,
+      "grad_norm": 19.71268653869629,
+      "learning_rate": 5.411477411477412e-05,
+      "loss": 0.4731,
+      "step": 515
+    },
+    {
+      "epoch": 1.89010989010989,
+      "grad_norm": 68.88224792480469,
+      "learning_rate": 5.409035409035409e-05,
+      "loss": 3.0071,
+      "step": 516
+    },
+    {
+      "epoch": 1.8937728937728937,
+      "grad_norm": 34.33188247680664,
+      "learning_rate": 5.406593406593407e-05,
+      "loss": 0.7014,
+      "step": 517
+    },
+    {
+      "epoch": 1.8974358974358974,
+      "grad_norm": 18.214942932128906,
+      "learning_rate": 5.404151404151404e-05,
+      "loss": 0.2362,
+      "step": 518
+    },
+    {
+      "epoch": 1.901098901098901,
+      "grad_norm": 31.553678512573242,
+      "learning_rate": 5.401709401709402e-05,
+      "loss": 0.5839,
+      "step": 519
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 15.681426048278809,
+      "learning_rate": 5.3992673992673996e-05,
+      "loss": 0.6039,
+      "step": 520
+    },
+    {
+      "epoch": 1.9084249084249083,
+      "grad_norm": 18.462688446044922,
+      "learning_rate": 5.396825396825397e-05,
+      "loss": 0.5773,
+      "step": 521
+    },
+    {
+      "epoch": 1.912087912087912,
+      "grad_norm": 10.23849105834961,
+      "learning_rate": 5.3943833943833946e-05,
+      "loss": 0.3801,
+      "step": 522
+    },
+    {
+      "epoch": 1.9157509157509156,
+      "grad_norm": 35.680973052978516,
+      "learning_rate": 5.3919413919413925e-05,
+      "loss": 1.2559,
+      "step": 523
+    },
+    {
+      "epoch": 1.9194139194139193,
+      "grad_norm": 23.97362518310547,
+      "learning_rate": 5.3894993894993897e-05,
+      "loss": 0.4112,
+      "step": 524
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 25.785356521606445,
+      "learning_rate": 5.387057387057387e-05,
+      "loss": 0.8993,
+      "step": 525
+    },
+    {
+      "epoch": 1.9267399267399268,
+      "grad_norm": 25.246868133544922,
+      "learning_rate": 5.384615384615385e-05,
+      "loss": 0.6534,
+      "step": 526
+    },
+    {
+      "epoch": 1.9304029304029304,
+      "grad_norm": 29.850788116455078,
+      "learning_rate": 5.382173382173382e-05,
+      "loss": 0.52,
+      "step": 527
+    },
+    {
+      "epoch": 1.934065934065934,
+      "grad_norm": 20.702608108520508,
+      "learning_rate": 5.37973137973138e-05,
+      "loss": 0.4093,
+      "step": 528
+    },
+    {
+      "epoch": 1.9377289377289377,
+      "grad_norm": 36.39994812011719,
+      "learning_rate": 5.3772893772893775e-05,
+      "loss": 1.275,
+      "step": 529
+    },
+    {
+      "epoch": 1.9413919413919414,
+      "grad_norm": 27.56822395324707,
+      "learning_rate": 5.374847374847375e-05,
+      "loss": 0.6773,
+      "step": 530
+    },
+    {
+      "epoch": 1.945054945054945,
+      "grad_norm": 26.07769012451172,
+      "learning_rate": 5.3724053724053725e-05,
+      "loss": 0.5373,
+      "step": 531
+    },
+    {
+      "epoch": 1.9487179487179487,
+      "grad_norm": 48.47615051269531,
+      "learning_rate": 5.3699633699633704e-05,
+      "loss": 1.1931,
+      "step": 532
+    },
+    {
+      "epoch": 1.9523809523809523,
+      "grad_norm": 24.416805267333984,
+      "learning_rate": 5.3675213675213675e-05,
+      "loss": 0.4523,
+      "step": 533
+    },
+    {
+      "epoch": 1.9560439560439562,
+      "grad_norm": 56.8088264465332,
+      "learning_rate": 5.3650793650793654e-05,
+      "loss": 1.8992,
+      "step": 534
+    },
+    {
+      "epoch": 1.9597069597069599,
+      "grad_norm": 36.805912017822266,
+      "learning_rate": 5.362637362637363e-05,
+      "loss": 1.0743,
+      "step": 535
+    },
+    {
+      "epoch": 1.9633699633699635,
+      "grad_norm": 17.375244140625,
+      "learning_rate": 5.3601953601953604e-05,
+      "loss": 0.3546,
+      "step": 536
+    },
+    {
+      "epoch": 1.9670329670329672,
+      "grad_norm": 35.297767639160156,
+      "learning_rate": 5.357753357753358e-05,
+      "loss": 1.4903,
+      "step": 537
+    },
+    {
+      "epoch": 1.9706959706959708,
+      "grad_norm": 38.64927673339844,
+      "learning_rate": 5.3553113553113554e-05,
+      "loss": 0.9346,
+      "step": 538
+    },
+    {
+      "epoch": 1.9743589743589745,
+      "grad_norm": 23.494552612304688,
+      "learning_rate": 5.3528693528693526e-05,
+      "loss": 0.3677,
+      "step": 539
+    },
+    {
+      "epoch": 1.978021978021978,
+      "grad_norm": 21.8272647857666,
+      "learning_rate": 5.3504273504273504e-05,
+      "loss": 0.591,
+      "step": 540
+    },
+    {
+      "epoch": 1.9816849816849818,
+      "grad_norm": 15.60590934753418,
+      "learning_rate": 5.347985347985348e-05,
+      "loss": 0.3129,
+      "step": 541
+    },
+    {
+      "epoch": 1.9853479853479854,
+      "grad_norm": 23.846555709838867,
+      "learning_rate": 5.3455433455433454e-05,
+      "loss": 0.6108,
+      "step": 542
+    },
+    {
+      "epoch": 1.989010989010989,
+      "grad_norm": 21.743024826049805,
+      "learning_rate": 5.343101343101343e-05,
+      "loss": 1.0541,
+      "step": 543
+    },
+    {
+      "epoch": 1.9926739926739927,
+      "grad_norm": 29.806121826171875,
+      "learning_rate": 5.3406593406593405e-05,
+      "loss": 0.6088,
+      "step": 544
+    },
+    {
+      "epoch": 1.9963369963369964,
+      "grad_norm": 26.778568267822266,
+      "learning_rate": 5.338217338217338e-05,
+      "loss": 0.5842,
+      "step": 545
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 23.356237411499023,
+      "learning_rate": 5.335775335775336e-05,
+      "loss": 0.4591,
+      "step": 546
+    },
+    {
+      "epoch": 2.0036630036630036,
+      "grad_norm": 17.303443908691406,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.3432,
+      "step": 547
+    },
+    {
+      "epoch": 2.0073260073260073,
+      "grad_norm": 27.082172393798828,
+      "learning_rate": 5.330891330891331e-05,
+      "loss": 0.5156,
+      "step": 548
+    },
+    {
+      "epoch": 2.010989010989011,
+      "grad_norm": 26.520530700683594,
+      "learning_rate": 5.328449328449329e-05,
+      "loss": 0.3989,
+      "step": 549
+    },
+    {
+      "epoch": 2.0146520146520146,
+      "grad_norm": 23.737272262573242,
+      "learning_rate": 5.326007326007326e-05,
+      "loss": 0.5484,
+      "step": 550
+    },
+    {
+      "epoch": 2.0183150183150182,
+      "grad_norm": 24.222341537475586,
+      "learning_rate": 5.3235653235653233e-05,
+      "loss": 0.5365,
+      "step": 551
+    },
+    {
+      "epoch": 2.021978021978022,
+      "grad_norm": 29.081924438476562,
+      "learning_rate": 5.321123321123321e-05,
+      "loss": 0.6694,
+      "step": 552
+    },
+    {
+      "epoch": 2.0256410256410255,
+      "grad_norm": 32.419551849365234,
+      "learning_rate": 5.3186813186813184e-05,
+      "loss": 0.7003,
+      "step": 553
+    },
+    {
+      "epoch": 2.029304029304029,
+      "grad_norm": 42.403709411621094,
+      "learning_rate": 5.316239316239316e-05,
+      "loss": 1.5474,
+      "step": 554
+    },
+    {
+      "epoch": 2.032967032967033,
+      "grad_norm": 17.615140914916992,
+      "learning_rate": 5.313797313797314e-05,
+      "loss": 0.588,
+      "step": 555
+    },
+    {
+      "epoch": 2.0366300366300365,
+      "grad_norm": 14.864067077636719,
+      "learning_rate": 5.311355311355311e-05,
+      "loss": 0.1613,
+      "step": 556
+    },
+    {
+      "epoch": 2.04029304029304,
+      "grad_norm": 20.189815521240234,
+      "learning_rate": 5.308913308913309e-05,
+      "loss": 0.4281,
+      "step": 557
+    },
+    {
+      "epoch": 2.043956043956044,
+      "grad_norm": 28.350017547607422,
+      "learning_rate": 5.306471306471307e-05,
+      "loss": 0.6614,
+      "step": 558
+    },
+    {
+      "epoch": 2.0476190476190474,
+      "grad_norm": 19.987825393676758,
+      "learning_rate": 5.304029304029304e-05,
+      "loss": 0.6906,
+      "step": 559
+    },
+    {
+      "epoch": 2.051282051282051,
+      "grad_norm": 18.6667537689209,
+      "learning_rate": 5.301587301587302e-05,
+      "loss": 0.387,
+      "step": 560
+    },
+    {
+      "epoch": 2.0549450549450547,
+      "grad_norm": 20.930652618408203,
+      "learning_rate": 5.2991452991453e-05,
+      "loss": 0.7157,
+      "step": 561
+    },
+    {
+      "epoch": 2.0586080586080584,
+      "grad_norm": 22.05647087097168,
+      "learning_rate": 5.296703296703297e-05,
+      "loss": 0.3256,
+      "step": 562
+    },
+    {
+      "epoch": 2.062271062271062,
+      "grad_norm": 32.66161346435547,
+      "learning_rate": 5.294261294261295e-05,
+      "loss": 1.3013,
+      "step": 563
+    },
+    {
+      "epoch": 2.065934065934066,
+      "grad_norm": 37.43238067626953,
+      "learning_rate": 5.291819291819292e-05,
+      "loss": 0.186,
+      "step": 564
+    },
+    {
+      "epoch": 2.06959706959707,
+      "grad_norm": 32.39999008178711,
+      "learning_rate": 5.289377289377289e-05,
+      "loss": 0.8047,
+      "step": 565
+    },
+    {
+      "epoch": 2.0732600732600734,
+      "grad_norm": 29.727481842041016,
+      "learning_rate": 5.286935286935287e-05,
+      "loss": 0.662,
+      "step": 566
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 16.536264419555664,
+      "learning_rate": 5.284493284493285e-05,
+      "loss": 0.4,
+      "step": 567
+    },
+    {
+      "epoch": 2.0805860805860807,
+      "grad_norm": 23.41500473022461,
+      "learning_rate": 5.282051282051282e-05,
+      "loss": 0.4945,
+      "step": 568
+    },
+    {
+      "epoch": 2.0842490842490844,
+      "grad_norm": 48.842864990234375,
+      "learning_rate": 5.27960927960928e-05,
+      "loss": 0.7584,
+      "step": 569
+    },
+    {
+      "epoch": 2.087912087912088,
+      "grad_norm": 60.06027603149414,
+      "learning_rate": 5.277167277167277e-05,
+      "loss": 0.7179,
+      "step": 570
+    },
+    {
+      "epoch": 2.0915750915750917,
+      "grad_norm": 59.2591552734375,
+      "learning_rate": 5.274725274725275e-05,
+      "loss": 0.4883,
+      "step": 571
+    },
+    {
+      "epoch": 2.0952380952380953,
+      "grad_norm": 14.527932167053223,
+      "learning_rate": 5.272283272283273e-05,
+      "loss": 0.2811,
+      "step": 572
+    },
+    {
+      "epoch": 2.098901098901099,
+      "grad_norm": 16.2915096282959,
+      "learning_rate": 5.26984126984127e-05,
+      "loss": 0.2524,
+      "step": 573
+    },
+    {
+      "epoch": 2.1025641025641026,
+      "grad_norm": 28.938081741333008,
+      "learning_rate": 5.267399267399268e-05,
+      "loss": 0.5138,
+      "step": 574
+    },
+    {
+      "epoch": 2.1062271062271063,
+      "grad_norm": 27.541440963745117,
+      "learning_rate": 5.2649572649572655e-05,
+      "loss": 0.278,
+      "step": 575
+    },
+    {
+      "epoch": 2.10989010989011,
+      "grad_norm": 23.179025650024414,
+      "learning_rate": 5.262515262515263e-05,
+      "loss": 0.1881,
+      "step": 576
+    },
+    {
+      "epoch": 2.1135531135531136,
+      "grad_norm": 42.55375671386719,
+      "learning_rate": 5.26007326007326e-05,
+      "loss": 0.7882,
+      "step": 577
+    },
+    {
+      "epoch": 2.1172161172161172,
+      "grad_norm": 8.902749061584473,
+      "learning_rate": 5.257631257631258e-05,
+      "loss": 0.0611,
+      "step": 578
+    },
+    {
+      "epoch": 2.120879120879121,
+      "grad_norm": 19.483346939086914,
+      "learning_rate": 5.255189255189255e-05,
+      "loss": 0.0978,
+      "step": 579
+    },
+    {
+      "epoch": 2.1245421245421245,
+      "grad_norm": 13.898221969604492,
+      "learning_rate": 5.252747252747253e-05,
+      "loss": 0.0797,
+      "step": 580
+    },
+    {
+      "epoch": 2.128205128205128,
+      "grad_norm": 53.42538833618164,
+      "learning_rate": 5.2503052503052506e-05,
+      "loss": 0.9066,
+      "step": 581
+    },
+    {
+      "epoch": 2.131868131868132,
+      "grad_norm": 38.467891693115234,
+      "learning_rate": 5.247863247863248e-05,
+      "loss": 0.3272,
+      "step": 582
+    },
+    {
+      "epoch": 2.1355311355311355,
+      "grad_norm": 26.421035766601562,
+      "learning_rate": 5.2454212454212456e-05,
+      "loss": 0.6537,
+      "step": 583
+    },
+    {
+      "epoch": 2.139194139194139,
+      "grad_norm": 32.80412292480469,
+      "learning_rate": 5.2429792429792434e-05,
+      "loss": 1.1225,
+      "step": 584
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 26.87016487121582,
+      "learning_rate": 5.2405372405372406e-05,
+      "loss": 0.5749,
+      "step": 585
+    },
+    {
+      "epoch": 2.1465201465201464,
+      "grad_norm": 34.75699234008789,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 0.6926,
+      "step": 586
+    },
+    {
+      "epoch": 2.15018315018315,
+      "grad_norm": 61.76310348510742,
+      "learning_rate": 5.235653235653236e-05,
+      "loss": 0.9029,
+      "step": 587
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 40.86505126953125,
+      "learning_rate": 5.2332112332112335e-05,
+      "loss": 0.5169,
+      "step": 588
+    },
+    {
+      "epoch": 2.1575091575091574,
+      "grad_norm": 16.05042839050293,
+      "learning_rate": 5.230769230769231e-05,
+      "loss": 0.5211,
+      "step": 589
+    },
+    {
+      "epoch": 2.161172161172161,
+      "grad_norm": 19.56302261352539,
+      "learning_rate": 5.2283272283272285e-05,
+      "loss": 0.5737,
+      "step": 590
+    },
+    {
+      "epoch": 2.1648351648351647,
+      "grad_norm": 22.311508178710938,
+      "learning_rate": 5.2258852258852256e-05,
+      "loss": 0.4223,
+      "step": 591
+    },
+    {
+      "epoch": 2.1684981684981683,
+      "grad_norm": 21.059213638305664,
+      "learning_rate": 5.2234432234432235e-05,
+      "loss": 0.2285,
+      "step": 592
+    },
+    {
+      "epoch": 2.172161172161172,
+      "grad_norm": 28.82351303100586,
+      "learning_rate": 5.221001221001221e-05,
+      "loss": 0.8438,
+      "step": 593
+    },
+    {
+      "epoch": 2.1758241758241756,
+      "grad_norm": 14.425333023071289,
+      "learning_rate": 5.2185592185592185e-05,
+      "loss": 0.1765,
+      "step": 594
+    },
+    {
+      "epoch": 2.1794871794871793,
+      "grad_norm": 16.967479705810547,
+      "learning_rate": 5.2161172161172163e-05,
+      "loss": 0.2465,
+      "step": 595
+    },
+    {
+      "epoch": 2.183150183150183,
+      "grad_norm": 40.79065704345703,
+      "learning_rate": 5.2136752136752135e-05,
+      "loss": 0.6077,
+      "step": 596
+    },
+    {
+      "epoch": 2.186813186813187,
+      "grad_norm": 22.434715270996094,
+      "learning_rate": 5.2112332112332114e-05,
+      "loss": 0.3748,
+      "step": 597
+    },
+    {
+      "epoch": 2.1904761904761907,
+      "grad_norm": 32.18471908569336,
+      "learning_rate": 5.208791208791209e-05,
+      "loss": 0.5163,
+      "step": 598
+    },
+    {
+      "epoch": 2.1941391941391943,
+      "grad_norm": 20.43740463256836,
+      "learning_rate": 5.2063492063492064e-05,
+      "loss": 0.4116,
+      "step": 599
+    },
+    {
+      "epoch": 2.197802197802198,
+      "grad_norm": 6.528069496154785,
+      "learning_rate": 5.203907203907204e-05,
+      "loss": 0.065,
+      "step": 600
+    },
+    {
+      "epoch": 2.2014652014652016,
+      "grad_norm": 35.0635871887207,
+      "learning_rate": 5.201465201465202e-05,
+      "loss": 1.2288,
+      "step": 601
+    },
+    {
+      "epoch": 2.2051282051282053,
+      "grad_norm": 23.499767303466797,
+      "learning_rate": 5.199023199023199e-05,
+      "loss": 0.49,
+      "step": 602
+    },
+    {
+      "epoch": 2.208791208791209,
+      "grad_norm": 20.234952926635742,
+      "learning_rate": 5.1965811965811964e-05,
+      "loss": 0.231,
+      "step": 603
+    },
+    {
+      "epoch": 2.2124542124542126,
+      "grad_norm": 9.268828392028809,
+      "learning_rate": 5.194139194139194e-05,
+      "loss": 0.0732,
+      "step": 604
+    },
+    {
+      "epoch": 2.2161172161172162,
+      "grad_norm": 52.60474395751953,
+      "learning_rate": 5.1916971916971914e-05,
+      "loss": 0.8766,
+      "step": 605
+    },
+    {
+      "epoch": 2.21978021978022,
+      "grad_norm": 41.86642074584961,
+      "learning_rate": 5.189255189255189e-05,
+      "loss": 0.4743,
+      "step": 606
+    },
+    {
+      "epoch": 2.2234432234432235,
+      "grad_norm": 30.304580688476562,
+      "learning_rate": 5.186813186813187e-05,
+      "loss": 0.4412,
+      "step": 607
+    },
+    {
+      "epoch": 2.227106227106227,
+      "grad_norm": 27.26057243347168,
+      "learning_rate": 5.184371184371184e-05,
+      "loss": 0.3496,
+      "step": 608
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 40.55131149291992,
+      "learning_rate": 5.181929181929182e-05,
+      "loss": 0.7097,
+      "step": 609
+    },
+    {
+      "epoch": 2.2344322344322345,
+      "grad_norm": 61.97871017456055,
+      "learning_rate": 5.17948717948718e-05,
+      "loss": 1.3686,
+      "step": 610
+    },
+    {
+      "epoch": 2.238095238095238,
+      "grad_norm": 38.211700439453125,
+      "learning_rate": 5.177045177045177e-05,
+      "loss": 0.565,
+      "step": 611
+    },
+    {
+      "epoch": 2.241758241758242,
+      "grad_norm": 20.10716438293457,
+      "learning_rate": 5.174603174603175e-05,
+      "loss": 0.3468,
+      "step": 612
+    },
+    {
+      "epoch": 2.2454212454212454,
+      "grad_norm": 23.96891975402832,
+      "learning_rate": 5.172161172161173e-05,
+      "loss": 0.2295,
+      "step": 613
+    },
+    {
+      "epoch": 2.249084249084249,
+      "grad_norm": 10.14421272277832,
+      "learning_rate": 5.16971916971917e-05,
+      "loss": 0.0943,
+      "step": 614
+    },
+    {
+      "epoch": 2.2527472527472527,
+      "grad_norm": 15.786056518554688,
+      "learning_rate": 5.167277167277167e-05,
+      "loss": 0.1213,
+      "step": 615
+    },
+    {
+      "epoch": 2.2564102564102564,
+      "grad_norm": 20.907663345336914,
+      "learning_rate": 5.164835164835165e-05,
+      "loss": 0.235,
+      "step": 616
+    },
+    {
+      "epoch": 2.26007326007326,
+      "grad_norm": 32.149600982666016,
+      "learning_rate": 5.162393162393162e-05,
+      "loss": 0.4807,
+      "step": 617
+    },
+    {
+      "epoch": 2.2637362637362637,
+      "grad_norm": 33.965518951416016,
+      "learning_rate": 5.15995115995116e-05,
+      "loss": 0.4517,
+      "step": 618
+    },
+    {
+      "epoch": 2.2673992673992673,
+      "grad_norm": 49.98363494873047,
+      "learning_rate": 5.157509157509158e-05,
+      "loss": 0.6434,
+      "step": 619
+    },
+    {
+      "epoch": 2.271062271062271,
+      "grad_norm": 14.035831451416016,
+      "learning_rate": 5.155067155067155e-05,
+      "loss": 0.1117,
+      "step": 620
+    },
+    {
+      "epoch": 2.2747252747252746,
+      "grad_norm": 28.84484100341797,
+      "learning_rate": 5.152625152625153e-05,
+      "loss": 0.8002,
+      "step": 621
+    },
+    {
+      "epoch": 2.2783882783882783,
+      "grad_norm": 41.59181594848633,
+      "learning_rate": 5.15018315018315e-05,
+      "loss": 0.4465,
+      "step": 622
+    },
+    {
+      "epoch": 2.282051282051282,
+      "grad_norm": 33.10573196411133,
+      "learning_rate": 5.147741147741148e-05,
+      "loss": 0.5795,
+      "step": 623
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 34.79928970336914,
+      "learning_rate": 5.145299145299146e-05,
+      "loss": 0.3135,
+      "step": 624
+    },
+    {
+      "epoch": 2.2893772893772892,
+      "grad_norm": 18.095544815063477,
+      "learning_rate": 5.142857142857143e-05,
+      "loss": 0.0961,
+      "step": 625
+    },
+    {
+      "epoch": 2.293040293040293,
+      "grad_norm": 16.55453872680664,
+      "learning_rate": 5.140415140415141e-05,
+      "loss": 0.0868,
+      "step": 626
+    },
+    {
+      "epoch": 2.2967032967032965,
+      "grad_norm": 42.18946075439453,
+      "learning_rate": 5.1379731379731386e-05,
+      "loss": 0.8892,
+      "step": 627
+    },
+    {
+      "epoch": 2.3003663003663,
+      "grad_norm": 54.753448486328125,
+      "learning_rate": 5.135531135531135e-05,
+      "loss": 0.833,
+      "step": 628
+    },
+    {
+      "epoch": 2.304029304029304,
+      "grad_norm": 27.723228454589844,
+      "learning_rate": 5.133089133089133e-05,
+      "loss": 0.2744,
+      "step": 629
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 28.53034019470215,
+      "learning_rate": 5.130647130647131e-05,
+      "loss": 0.1696,
+      "step": 630
+    },
+    {
+      "epoch": 2.311355311355311,
+      "grad_norm": 65.4127426147461,
+      "learning_rate": 5.128205128205128e-05,
+      "loss": 0.9019,
+      "step": 631
+    },
+    {
+      "epoch": 2.315018315018315,
+      "grad_norm": 22.794870376586914,
+      "learning_rate": 5.125763125763126e-05,
+      "loss": 0.1987,
+      "step": 632
+    },
+    {
+      "epoch": 2.3186813186813184,
+      "grad_norm": 29.870113372802734,
+      "learning_rate": 5.1233211233211236e-05,
+      "loss": 0.4816,
+      "step": 633
+    },
+    {
+      "epoch": 2.3223443223443225,
+      "grad_norm": 38.91164779663086,
+      "learning_rate": 5.120879120879121e-05,
+      "loss": 0.7424,
+      "step": 634
+    },
+    {
+      "epoch": 2.326007326007326,
+      "grad_norm": 36.57811737060547,
+      "learning_rate": 5.1184371184371186e-05,
+      "loss": 1.1365,
+      "step": 635
+    },
+    {
+      "epoch": 2.32967032967033,
+      "grad_norm": 31.59128189086914,
+      "learning_rate": 5.1159951159951165e-05,
+      "loss": 0.6167,
+      "step": 636
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 25.956003189086914,
+      "learning_rate": 5.1135531135531136e-05,
+      "loss": 0.8808,
+      "step": 637
+    },
+    {
+      "epoch": 2.336996336996337,
+      "grad_norm": 38.18582534790039,
+      "learning_rate": 5.1111111111111115e-05,
+      "loss": 0.9417,
+      "step": 638
+    },
+    {
+      "epoch": 2.340659340659341,
+      "grad_norm": 27.436229705810547,
+      "learning_rate": 5.108669108669109e-05,
+      "loss": 0.7539,
+      "step": 639
+    },
+    {
+      "epoch": 2.3443223443223444,
+      "grad_norm": 40.86305618286133,
+      "learning_rate": 5.1062271062271065e-05,
+      "loss": 2.126,
+      "step": 640
+    },
+    {
+      "epoch": 2.347985347985348,
+      "grad_norm": 22.224748611450195,
+      "learning_rate": 5.103785103785104e-05,
+      "loss": 0.9958,
+      "step": 641
+    },
+    {
+      "epoch": 2.3516483516483517,
+      "grad_norm": 19.915552139282227,
+      "learning_rate": 5.1013431013431015e-05,
+      "loss": 1.1045,
+      "step": 642
+    },
+    {
+      "epoch": 2.3553113553113554,
+      "grad_norm": 17.045989990234375,
+      "learning_rate": 5.098901098901099e-05,
+      "loss": 0.8906,
+      "step": 643
+    },
+    {
+      "epoch": 2.358974358974359,
+      "grad_norm": 22.106670379638672,
+      "learning_rate": 5.0964590964590965e-05,
+      "loss": 0.9856,
+      "step": 644
+    },
+    {
+      "epoch": 2.3626373626373627,
+      "grad_norm": 17.583837509155273,
+      "learning_rate": 5.0940170940170944e-05,
+      "loss": 0.8328,
+      "step": 645
+    },
+    {
+      "epoch": 2.3663003663003663,
+      "grad_norm": 57.61167526245117,
+      "learning_rate": 5.0915750915750915e-05,
+      "loss": 0.578,
+      "step": 646
+    },
+    {
+      "epoch": 2.36996336996337,
+      "grad_norm": 13.941128730773926,
+      "learning_rate": 5.0891330891330894e-05,
+      "loss": 0.5892,
+      "step": 647
+    },
+    {
+      "epoch": 2.3736263736263736,
+      "grad_norm": 22.38715171813965,
+      "learning_rate": 5.0866910866910866e-05,
+      "loss": 0.7608,
+      "step": 648
+    },
+    {
+      "epoch": 2.3772893772893773,
+      "grad_norm": 22.42316436767578,
+      "learning_rate": 5.0842490842490844e-05,
+      "loss": 0.7923,
+      "step": 649
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 32.75740432739258,
+      "learning_rate": 5.081807081807082e-05,
+      "loss": 1.0798,
+      "step": 650
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 19.295289993286133,
+      "learning_rate": 5.0793650793650794e-05,
+      "loss": 0.4898,
+      "step": 651
+    },
+    {
+      "epoch": 2.3882783882783882,
+      "grad_norm": 25.849227905273438,
+      "learning_rate": 5.076923076923077e-05,
+      "loss": 0.5557,
+      "step": 652
+    },
+    {
+      "epoch": 2.391941391941392,
+      "grad_norm": 21.321088790893555,
+      "learning_rate": 5.074481074481075e-05,
+      "loss": 0.2743,
+      "step": 653
+    },
+    {
+      "epoch": 2.3956043956043955,
+      "grad_norm": 28.795917510986328,
+      "learning_rate": 5.0720390720390716e-05,
+      "loss": 0.7039,
+      "step": 654
+    },
+    {
+      "epoch": 2.399267399267399,
+      "grad_norm": 19.86751937866211,
+      "learning_rate": 5.0695970695970694e-05,
+      "loss": 0.3155,
+      "step": 655
+    },
+    {
+      "epoch": 2.402930402930403,
+      "grad_norm": 33.3828010559082,
+      "learning_rate": 5.067155067155067e-05,
+      "loss": 1.0696,
+      "step": 656
+    },
+    {
+      "epoch": 2.4065934065934065,
+      "grad_norm": 37.38752746582031,
+      "learning_rate": 5.0647130647130645e-05,
+      "loss": 0.8123,
+      "step": 657
+    },
+    {
+      "epoch": 2.41025641025641,
+      "grad_norm": 29.22795867919922,
+      "learning_rate": 5.062271062271062e-05,
+      "loss": 0.9515,
+      "step": 658
+    },
+    {
+      "epoch": 2.413919413919414,
+      "grad_norm": 41.129981994628906,
+      "learning_rate": 5.05982905982906e-05,
+      "loss": 1.1329,
+      "step": 659
+    },
+    {
+      "epoch": 2.4175824175824174,
+      "grad_norm": 40.985042572021484,
+      "learning_rate": 5.057387057387057e-05,
+      "loss": 0.675,
+      "step": 660
+    },
+    {
+      "epoch": 2.421245421245421,
+      "grad_norm": 33.49393844604492,
+      "learning_rate": 5.054945054945055e-05,
+      "loss": 0.9679,
+      "step": 661
+    },
+    {
+      "epoch": 2.4249084249084247,
+      "grad_norm": 28.741533279418945,
+      "learning_rate": 5.052503052503053e-05,
+      "loss": 0.7928,
+      "step": 662
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 28.89700698852539,
+      "learning_rate": 5.05006105006105e-05,
+      "loss": 0.7594,
+      "step": 663
+    },
+    {
+      "epoch": 2.4322344322344325,
+      "grad_norm": 4.59797477722168,
+      "learning_rate": 5.047619047619048e-05,
+      "loss": 0.0584,
+      "step": 664
+    },
+    {
+      "epoch": 2.435897435897436,
+      "grad_norm": 29.852828979492188,
+      "learning_rate": 5.045177045177046e-05,
+      "loss": 0.614,
+      "step": 665
+    },
+    {
+      "epoch": 2.4395604395604398,
+      "grad_norm": 15.132670402526855,
+      "learning_rate": 5.042735042735043e-05,
+      "loss": 0.2353,
+      "step": 666
+    },
+    {
+      "epoch": 2.4432234432234434,
+      "grad_norm": 23.85403060913086,
+      "learning_rate": 5.04029304029304e-05,
+      "loss": 0.9065,
+      "step": 667
+    },
+    {
+      "epoch": 2.446886446886447,
+      "grad_norm": 12.384196281433105,
+      "learning_rate": 5.037851037851038e-05,
+      "loss": 0.2065,
+      "step": 668
+    },
+    {
+      "epoch": 2.4505494505494507,
+      "grad_norm": 18.347129821777344,
+      "learning_rate": 5.035409035409035e-05,
+      "loss": 0.647,
+      "step": 669
+    },
+    {
+      "epoch": 2.4542124542124544,
+      "grad_norm": 18.645936965942383,
+      "learning_rate": 5.032967032967033e-05,
+      "loss": 0.2072,
+      "step": 670
+    },
+    {
+      "epoch": 2.457875457875458,
+      "grad_norm": 9.493071556091309,
+      "learning_rate": 5.03052503052503e-05,
+      "loss": 0.1805,
+      "step": 671
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 18.552539825439453,
+      "learning_rate": 5.028083028083028e-05,
+      "loss": 0.4078,
+      "step": 672
+    },
+    {
+      "epoch": 2.4652014652014653,
+      "grad_norm": 21.735048294067383,
+      "learning_rate": 5.025641025641026e-05,
+      "loss": 0.4231,
+      "step": 673
+    },
+    {
+      "epoch": 2.468864468864469,
+      "grad_norm": 54.32040023803711,
+      "learning_rate": 5.023199023199023e-05,
+      "loss": 1.3927,
+      "step": 674
+    },
+    {
+      "epoch": 2.4725274725274726,
+      "grad_norm": 26.955970764160156,
+      "learning_rate": 5.020757020757021e-05,
+      "loss": 0.6899,
+      "step": 675
+    },
+    {
+      "epoch": 2.4761904761904763,
+      "grad_norm": 43.423526763916016,
+      "learning_rate": 5.018315018315019e-05,
+      "loss": 1.2084,
+      "step": 676
+    },
+    {
+      "epoch": 2.47985347985348,
+      "grad_norm": 35.98548126220703,
+      "learning_rate": 5.015873015873016e-05,
+      "loss": 1.5047,
+      "step": 677
+    },
+    {
+      "epoch": 2.4835164835164836,
+      "grad_norm": 22.593570709228516,
+      "learning_rate": 5.013431013431014e-05,
+      "loss": 0.6918,
+      "step": 678
+    },
+    {
+      "epoch": 2.4871794871794872,
+      "grad_norm": 21.29257583618164,
+      "learning_rate": 5.0109890109890116e-05,
+      "loss": 0.3578,
+      "step": 679
+    },
+    {
+      "epoch": 2.490842490842491,
+      "grad_norm": 21.672088623046875,
+      "learning_rate": 5.008547008547008e-05,
+      "loss": 0.7757,
+      "step": 680
+    },
+    {
+      "epoch": 2.4945054945054945,
+      "grad_norm": 9.625850677490234,
+      "learning_rate": 5.006105006105006e-05,
+      "loss": 0.1329,
+      "step": 681
+    },
+    {
+      "epoch": 2.498168498168498,
+      "grad_norm": 16.92123794555664,
+      "learning_rate": 5.003663003663004e-05,
+      "loss": 0.5599,
+      "step": 682
+    },
+    {
+      "epoch": 2.501831501831502,
+      "grad_norm": 15.665925025939941,
+      "learning_rate": 5.001221001221001e-05,
+      "loss": 0.3099,
+      "step": 683
+    },
+    {
+      "epoch": 2.5054945054945055,
+      "grad_norm": 21.316635131835938,
+      "learning_rate": 4.998778998778999e-05,
+      "loss": 0.5746,
+      "step": 684
+    },
+    {
+      "epoch": 2.509157509157509,
+      "grad_norm": 24.99594497680664,
+      "learning_rate": 4.996336996336997e-05,
+      "loss": 1.1274,
+      "step": 685
+    },
+    {
+      "epoch": 2.5128205128205128,
+      "grad_norm": 29.795175552368164,
+      "learning_rate": 4.993894993894994e-05,
+      "loss": 0.9991,
+      "step": 686
+    },
+    {
+      "epoch": 2.5164835164835164,
+      "grad_norm": 16.337533950805664,
+      "learning_rate": 4.991452991452992e-05,
+      "loss": 0.4101,
+      "step": 687
+    },
+    {
+      "epoch": 2.52014652014652,
+      "grad_norm": 20.065715789794922,
+      "learning_rate": 4.9890109890109895e-05,
+      "loss": 0.7786,
+      "step": 688
+    },
+    {
+      "epoch": 2.5238095238095237,
+      "grad_norm": 19.341567993164062,
+      "learning_rate": 4.986568986568987e-05,
+      "loss": 0.4989,
+      "step": 689
+    },
+    {
+      "epoch": 2.5274725274725274,
+      "grad_norm": 14.688420295715332,
+      "learning_rate": 4.9841269841269845e-05,
+      "loss": 0.4081,
+      "step": 690
+    },
+    {
+      "epoch": 2.531135531135531,
+      "grad_norm": 39.346012115478516,
+      "learning_rate": 4.9816849816849824e-05,
+      "loss": 1.7919,
+      "step": 691
+    },
+    {
+      "epoch": 2.5347985347985347,
+      "grad_norm": 21.353286743164062,
+      "learning_rate": 4.9792429792429796e-05,
+      "loss": 0.698,
+      "step": 692
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 35.96653366088867,
+      "learning_rate": 4.976800976800977e-05,
+      "loss": 1.6584,
+      "step": 693
+    },
+    {
+      "epoch": 2.542124542124542,
+      "grad_norm": 19.14348793029785,
+      "learning_rate": 4.9743589743589746e-05,
+      "loss": 0.885,
+      "step": 694
+    },
+    {
+      "epoch": 2.5457875457875456,
+      "grad_norm": 9.260897636413574,
+      "learning_rate": 4.971916971916972e-05,
+      "loss": 0.1629,
+      "step": 695
+    },
+    {
+      "epoch": 2.5494505494505493,
+      "grad_norm": 18.497526168823242,
+      "learning_rate": 4.9694749694749696e-05,
+      "loss": 0.7242,
+      "step": 696
+    },
+    {
+      "epoch": 2.553113553113553,
+      "grad_norm": 8.879841804504395,
+      "learning_rate": 4.967032967032967e-05,
+      "loss": 0.1302,
+      "step": 697
+    },
+    {
+      "epoch": 2.5567765567765566,
+      "grad_norm": 26.34065818786621,
+      "learning_rate": 4.9645909645909646e-05,
+      "loss": 0.7333,
+      "step": 698
+    },
+    {
+      "epoch": 2.5604395604395602,
+      "grad_norm": 15.10546588897705,
+      "learning_rate": 4.9621489621489624e-05,
+      "loss": 0.3119,
+      "step": 699
+    },
+    {
+      "epoch": 2.564102564102564,
+      "grad_norm": 10.68095874786377,
+      "learning_rate": 4.9597069597069596e-05,
+      "loss": 0.2505,
+      "step": 700
+    },
+    {
+      "epoch": 2.5677655677655675,
+      "grad_norm": 29.08888053894043,
+      "learning_rate": 4.9572649572649575e-05,
+      "loss": 0.4286,
+      "step": 701
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 29.939416885375977,
+      "learning_rate": 4.954822954822955e-05,
+      "loss": 1.1529,
+      "step": 702
+    },
+    {
+      "epoch": 2.575091575091575,
+      "grad_norm": 32.78864669799805,
+      "learning_rate": 4.9523809523809525e-05,
+      "loss": 0.9834,
+      "step": 703
+    },
+    {
+      "epoch": 2.578754578754579,
+      "grad_norm": 13.99082088470459,
+      "learning_rate": 4.94993894993895e-05,
+      "loss": 0.1934,
+      "step": 704
+    },
+    {
+      "epoch": 2.5824175824175826,
+      "grad_norm": 31.696718215942383,
+      "learning_rate": 4.9474969474969475e-05,
+      "loss": 0.6881,
+      "step": 705
+    },
+    {
+      "epoch": 2.586080586080586,
+      "grad_norm": 39.26205062866211,
+      "learning_rate": 4.9450549450549446e-05,
+      "loss": 0.573,
+      "step": 706
+    },
+    {
+      "epoch": 2.58974358974359,
+      "grad_norm": 42.08647918701172,
+      "learning_rate": 4.9426129426129425e-05,
+      "loss": 1.5935,
+      "step": 707
+    },
+    {
+      "epoch": 2.5934065934065935,
+      "grad_norm": 24.630651473999023,
+      "learning_rate": 4.94017094017094e-05,
+      "loss": 0.7016,
+      "step": 708
+    },
+    {
+      "epoch": 2.597069597069597,
+      "grad_norm": 35.33428192138672,
+      "learning_rate": 4.9377289377289375e-05,
+      "loss": 0.9646,
+      "step": 709
+    },
+    {
+      "epoch": 2.600732600732601,
+      "grad_norm": 21.643918991088867,
+      "learning_rate": 4.9352869352869353e-05,
+      "loss": 0.3679,
+      "step": 710
+    },
+    {
+      "epoch": 2.6043956043956045,
+      "grad_norm": 10.6254301071167,
+      "learning_rate": 4.932844932844933e-05,
+      "loss": 0.1059,
+      "step": 711
+    },
+    {
+      "epoch": 2.608058608058608,
+      "grad_norm": 23.43462562561035,
+      "learning_rate": 4.9304029304029304e-05,
+      "loss": 0.5128,
+      "step": 712
+    },
+    {
+      "epoch": 2.6117216117216118,
+      "grad_norm": 25.748422622680664,
+      "learning_rate": 4.927960927960928e-05,
+      "loss": 0.6154,
+      "step": 713
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 23.163209915161133,
+      "learning_rate": 4.925518925518926e-05,
+      "loss": 0.3978,
+      "step": 714
+    },
+    {
+      "epoch": 2.619047619047619,
+      "grad_norm": 22.306194305419922,
+      "learning_rate": 4.923076923076923e-05,
+      "loss": 0.3984,
+      "step": 715
+    },
+    {
+      "epoch": 2.6227106227106227,
+      "grad_norm": 48.16558074951172,
+      "learning_rate": 4.920634920634921e-05,
+      "loss": 0.9568,
+      "step": 716
+    },
+    {
+      "epoch": 2.6263736263736264,
+      "grad_norm": 48.76753234863281,
+      "learning_rate": 4.918192918192919e-05,
+      "loss": 0.6579,
+      "step": 717
+    },
+    {
+      "epoch": 2.63003663003663,
+      "grad_norm": 57.938720703125,
+      "learning_rate": 4.9157509157509154e-05,
+      "loss": 1.0926,
+      "step": 718
+    },
+    {
+      "epoch": 2.6336996336996337,
+      "grad_norm": 25.495267868041992,
+      "learning_rate": 4.913308913308913e-05,
+      "loss": 0.3717,
+      "step": 719
+    },
+    {
+      "epoch": 2.6373626373626373,
+      "grad_norm": 20.054609298706055,
+      "learning_rate": 4.910866910866911e-05,
+      "loss": 0.4502,
+      "step": 720
+    },
+    {
+      "epoch": 2.641025641025641,
+      "grad_norm": 23.096263885498047,
+      "learning_rate": 4.908424908424908e-05,
+      "loss": 0.2794,
+      "step": 721
+    },
+    {
+      "epoch": 2.6446886446886446,
+      "grad_norm": 6.073278903961182,
+      "learning_rate": 4.905982905982906e-05,
+      "loss": 0.0519,
+      "step": 722
+    },
+    {
+      "epoch": 2.6483516483516483,
+      "grad_norm": 38.562618255615234,
+      "learning_rate": 4.903540903540903e-05,
+      "loss": 0.8839,
+      "step": 723
+    },
+    {
+      "epoch": 2.652014652014652,
+      "grad_norm": 23.544757843017578,
+      "learning_rate": 4.901098901098901e-05,
+      "loss": 0.3935,
+      "step": 724
+    },
+    {
+      "epoch": 2.6556776556776556,
+      "grad_norm": 22.844032287597656,
+      "learning_rate": 4.898656898656899e-05,
+      "loss": 0.2428,
+      "step": 725
+    },
+    {
+      "epoch": 2.659340659340659,
+      "grad_norm": 11.537687301635742,
+      "learning_rate": 4.896214896214896e-05,
+      "loss": 0.1538,
+      "step": 726
+    },
+    {
+      "epoch": 2.663003663003663,
+      "grad_norm": 59.37337112426758,
+      "learning_rate": 4.893772893772894e-05,
+      "loss": 1.181,
+      "step": 727
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 22.206314086914062,
+      "learning_rate": 4.891330891330892e-05,
+      "loss": 0.4044,
+      "step": 728
+    },
+    {
+      "epoch": 2.67032967032967,
+      "grad_norm": 27.44620132446289,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 0.585,
+      "step": 729
+    },
+    {
+      "epoch": 2.6739926739926743,
+      "grad_norm": 35.70675277709961,
+      "learning_rate": 4.886446886446887e-05,
+      "loss": 0.6853,
+      "step": 730
+    },
+    {
+      "epoch": 2.677655677655678,
+      "grad_norm": 25.653356552124023,
+      "learning_rate": 4.884004884004884e-05,
+      "loss": 0.6143,
+      "step": 731
+    },
+    {
+      "epoch": 2.6813186813186816,
+      "grad_norm": 24.242090225219727,
+      "learning_rate": 4.881562881562881e-05,
+      "loss": 0.4365,
+      "step": 732
+    },
+    {
+      "epoch": 2.684981684981685,
+      "grad_norm": 25.621902465820312,
+      "learning_rate": 4.879120879120879e-05,
+      "loss": 0.6644,
+      "step": 733
+    },
+    {
+      "epoch": 2.688644688644689,
+      "grad_norm": 14.14786434173584,
+      "learning_rate": 4.876678876678877e-05,
+      "loss": 0.4117,
+      "step": 734
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 37.98638916015625,
+      "learning_rate": 4.874236874236874e-05,
+      "loss": 1.0452,
+      "step": 735
+    },
+    {
+      "epoch": 2.695970695970696,
+      "grad_norm": 23.186302185058594,
+      "learning_rate": 4.871794871794872e-05,
+      "loss": 0.2642,
+      "step": 736
+    },
+    {
+      "epoch": 2.6996336996337,
+      "grad_norm": 27.23651695251465,
+      "learning_rate": 4.86935286935287e-05,
+      "loss": 0.393,
+      "step": 737
+    },
+    {
+      "epoch": 2.7032967032967035,
+      "grad_norm": 36.44395446777344,
+      "learning_rate": 4.866910866910867e-05,
+      "loss": 1.1309,
+      "step": 738
+    },
+    {
+      "epoch": 2.706959706959707,
+      "grad_norm": 9.733710289001465,
+      "learning_rate": 4.864468864468865e-05,
+      "loss": 0.2466,
+      "step": 739
+    },
+    {
+      "epoch": 2.7106227106227108,
+      "grad_norm": 24.727527618408203,
+      "learning_rate": 4.8620268620268626e-05,
+      "loss": 0.46,
+      "step": 740
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 15.122056007385254,
+      "learning_rate": 4.85958485958486e-05,
+      "loss": 0.3122,
+      "step": 741
+    },
+    {
+      "epoch": 2.717948717948718,
+      "grad_norm": 24.059120178222656,
+      "learning_rate": 4.8571428571428576e-05,
+      "loss": 0.2359,
+      "step": 742
+    },
+    {
+      "epoch": 2.7216117216117217,
+      "grad_norm": 7.659122467041016,
+      "learning_rate": 4.8547008547008554e-05,
+      "loss": 0.1212,
+      "step": 743
+    },
+    {
+      "epoch": 2.7252747252747254,
+      "grad_norm": 27.002117156982422,
+      "learning_rate": 4.852258852258852e-05,
+      "loss": 0.7593,
+      "step": 744
+    },
+    {
+      "epoch": 2.728937728937729,
+      "grad_norm": 6.3852009773254395,
+      "learning_rate": 4.84981684981685e-05,
+      "loss": 0.0644,
+      "step": 745
+    },
+    {
+      "epoch": 2.7326007326007327,
+      "grad_norm": 25.574190139770508,
+      "learning_rate": 4.8473748473748476e-05,
+      "loss": 0.7012,
+      "step": 746
+    },
+    {
+      "epoch": 2.7362637362637363,
+      "grad_norm": 15.720768928527832,
+      "learning_rate": 4.844932844932845e-05,
+      "loss": 0.2692,
+      "step": 747
+    },
+    {
+      "epoch": 2.73992673992674,
+      "grad_norm": 25.527997970581055,
+      "learning_rate": 4.8424908424908426e-05,
+      "loss": 0.2648,
+      "step": 748
+    },
+    {
+      "epoch": 2.7435897435897436,
+      "grad_norm": 27.791011810302734,
+      "learning_rate": 4.84004884004884e-05,
+      "loss": 0.6007,
+      "step": 749
+    },
+    {
+      "epoch": 2.7472527472527473,
+      "grad_norm": 20.487640380859375,
+      "learning_rate": 4.8376068376068376e-05,
+      "loss": 0.5715,
+      "step": 750
+    },
+    {
+      "epoch": 2.750915750915751,
+      "grad_norm": 6.386992454528809,
+      "learning_rate": 4.8351648351648355e-05,
+      "loss": 0.06,
+      "step": 751
+    },
+    {
+      "epoch": 2.7545787545787546,
+      "grad_norm": 13.110812187194824,
+      "learning_rate": 4.8327228327228327e-05,
+      "loss": 0.129,
+      "step": 752
+    },
+    {
+      "epoch": 2.758241758241758,
+      "grad_norm": 26.55845832824707,
+      "learning_rate": 4.8302808302808305e-05,
+      "loss": 0.67,
+      "step": 753
+    },
+    {
+      "epoch": 2.761904761904762,
+      "grad_norm": 38.83135223388672,
+      "learning_rate": 4.8278388278388283e-05,
+      "loss": 1.6656,
+      "step": 754
+    },
+    {
+      "epoch": 2.7655677655677655,
+      "grad_norm": 25.99518585205078,
+      "learning_rate": 4.8253968253968255e-05,
+      "loss": 0.3285,
+      "step": 755
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 17.282081604003906,
+      "learning_rate": 4.8229548229548234e-05,
+      "loss": 0.2217,
+      "step": 756
+    },
+    {
+      "epoch": 2.772893772893773,
+      "grad_norm": 28.849924087524414,
+      "learning_rate": 4.8205128205128205e-05,
+      "loss": 0.7287,
+      "step": 757
+    },
+    {
+      "epoch": 2.7765567765567765,
+      "grad_norm": 45.79567337036133,
+      "learning_rate": 4.818070818070818e-05,
+      "loss": 1.6964,
+      "step": 758
+    },
+    {
+      "epoch": 2.78021978021978,
+      "grad_norm": 15.203421592712402,
+      "learning_rate": 4.8156288156288155e-05,
+      "loss": 0.2351,
+      "step": 759
+    },
+    {
+      "epoch": 2.7838827838827838,
+      "grad_norm": 10.686698913574219,
+      "learning_rate": 4.8131868131868134e-05,
+      "loss": 0.1533,
+      "step": 760
+    },
+    {
+      "epoch": 2.7875457875457874,
+      "grad_norm": 24.186473846435547,
+      "learning_rate": 4.8107448107448106e-05,
+      "loss": 1.0973,
+      "step": 761
+    },
+    {
+      "epoch": 2.791208791208791,
+      "grad_norm": 25.378986358642578,
+      "learning_rate": 4.8083028083028084e-05,
+      "loss": 0.5847,
+      "step": 762
+    },
+    {
+      "epoch": 2.7948717948717947,
+      "grad_norm": 20.066482543945312,
+      "learning_rate": 4.805860805860806e-05,
+      "loss": 0.2643,
+      "step": 763
+    },
+    {
+      "epoch": 2.7985347985347984,
+      "grad_norm": 56.11622619628906,
+      "learning_rate": 4.8034188034188034e-05,
+      "loss": 0.6949,
+      "step": 764
+    },
+    {
+      "epoch": 2.802197802197802,
+      "grad_norm": 27.80112648010254,
+      "learning_rate": 4.800976800976801e-05,
+      "loss": 0.5622,
+      "step": 765
+    },
+    {
+      "epoch": 2.8058608058608057,
+      "grad_norm": 30.947532653808594,
+      "learning_rate": 4.798534798534799e-05,
+      "loss": 0.6276,
+      "step": 766
+    },
+    {
+      "epoch": 2.8095238095238093,
+      "grad_norm": 8.91073226928711,
+      "learning_rate": 4.796092796092796e-05,
+      "loss": 0.1302,
+      "step": 767
+    },
+    {
+      "epoch": 2.813186813186813,
+      "grad_norm": 24.65394401550293,
+      "learning_rate": 4.793650793650794e-05,
+      "loss": 0.6811,
+      "step": 768
+    },
+    {
+      "epoch": 2.8168498168498166,
+      "grad_norm": 18.257539749145508,
+      "learning_rate": 4.791208791208792e-05,
+      "loss": 0.271,
+      "step": 769
+    },
+    {
+      "epoch": 2.8205128205128203,
+      "grad_norm": 41.41588592529297,
+      "learning_rate": 4.7887667887667884e-05,
+      "loss": 1.4149,
+      "step": 770
+    },
+    {
+      "epoch": 2.824175824175824,
+      "grad_norm": 7.753188610076904,
+      "learning_rate": 4.786324786324786e-05,
+      "loss": 0.0825,
+      "step": 771
+    },
+    {
+      "epoch": 2.8278388278388276,
+      "grad_norm": 208.88290405273438,
+      "learning_rate": 4.783882783882784e-05,
+      "loss": 1.032,
+      "step": 772
+    },
+    {
+      "epoch": 2.8315018315018317,
+      "grad_norm": 31.91672706604004,
+      "learning_rate": 4.781440781440781e-05,
+      "loss": 0.9783,
+      "step": 773
+    },
+    {
+      "epoch": 2.8351648351648353,
+      "grad_norm": 5.72416877746582,
+      "learning_rate": 4.778998778998779e-05,
+      "loss": 0.0399,
+      "step": 774
+    },
+    {
+      "epoch": 2.838827838827839,
+      "grad_norm": 30.503149032592773,
+      "learning_rate": 4.776556776556776e-05,
+      "loss": 0.6465,
+      "step": 775
+    },
+    {
+      "epoch": 2.8424908424908426,
+      "grad_norm": 29.615020751953125,
+      "learning_rate": 4.774114774114774e-05,
+      "loss": 0.5823,
+      "step": 776
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 49.922611236572266,
+      "learning_rate": 4.771672771672772e-05,
+      "loss": 1.2045,
+      "step": 777
+    },
+    {
+      "epoch": 2.84981684981685,
+      "grad_norm": 23.30948829650879,
+      "learning_rate": 4.769230769230769e-05,
+      "loss": 0.5962,
+      "step": 778
+    },
+    {
+      "epoch": 2.8534798534798536,
+      "grad_norm": 24.784086227416992,
+      "learning_rate": 4.766788766788767e-05,
+      "loss": 0.5702,
+      "step": 779
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 30.03589630126953,
+      "learning_rate": 4.764346764346765e-05,
+      "loss": 0.8644,
+      "step": 780
+    },
+    {
+      "epoch": 2.860805860805861,
+      "grad_norm": 21.079742431640625,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.2304,
+      "step": 781
+    },
+    {
+      "epoch": 2.8644688644688645,
+      "grad_norm": 18.438365936279297,
+      "learning_rate": 4.75946275946276e-05,
+      "loss": 0.6457,
+      "step": 782
+    },
+    {
+      "epoch": 2.868131868131868,
+      "grad_norm": 16.265140533447266,
+      "learning_rate": 4.757020757020757e-05,
+      "loss": 0.3693,
+      "step": 783
+    },
+    {
+      "epoch": 2.871794871794872,
+      "grad_norm": 17.526954650878906,
+      "learning_rate": 4.754578754578754e-05,
+      "loss": 0.2614,
+      "step": 784
+    },
+    {
+      "epoch": 2.8754578754578755,
+      "grad_norm": 39.94060134887695,
+      "learning_rate": 4.752136752136752e-05,
+      "loss": 0.2829,
+      "step": 785
+    },
+    {
+      "epoch": 2.879120879120879,
+      "grad_norm": 10.09298324584961,
+      "learning_rate": 4.74969474969475e-05,
+      "loss": 0.1489,
+      "step": 786
+    },
+    {
+      "epoch": 2.8827838827838828,
+      "grad_norm": 29.092544555664062,
+      "learning_rate": 4.747252747252747e-05,
+      "loss": 0.6063,
+      "step": 787
+    },
+    {
+      "epoch": 2.8864468864468864,
+      "grad_norm": 30.071422576904297,
+      "learning_rate": 4.744810744810745e-05,
+      "loss": 0.3154,
+      "step": 788
+    },
+    {
+      "epoch": 2.89010989010989,
+      "grad_norm": 26.271251678466797,
+      "learning_rate": 4.742368742368743e-05,
+      "loss": 0.4548,
+      "step": 789
+    },
+    {
+      "epoch": 2.8937728937728937,
+      "grad_norm": 32.386775970458984,
+      "learning_rate": 4.73992673992674e-05,
+      "loss": 0.1872,
+      "step": 790
+    },
+    {
+      "epoch": 2.8974358974358974,
+      "grad_norm": 31.18532943725586,
+      "learning_rate": 4.737484737484738e-05,
+      "loss": 0.847,
+      "step": 791
+    },
+    {
+      "epoch": 2.901098901098901,
+      "grad_norm": 17.924785614013672,
+      "learning_rate": 4.7350427350427356e-05,
+      "loss": 0.1588,
+      "step": 792
+    },
+    {
+      "epoch": 2.9047619047619047,
+      "grad_norm": 16.458614349365234,
+      "learning_rate": 4.732600732600733e-05,
+      "loss": 0.1424,
+      "step": 793
+    },
+    {
+      "epoch": 2.9084249084249083,
+      "grad_norm": 50.29280471801758,
+      "learning_rate": 4.7301587301587306e-05,
+      "loss": 1.5482,
+      "step": 794
+    },
+    {
+      "epoch": 2.912087912087912,
+      "grad_norm": 58.37470245361328,
+      "learning_rate": 4.727716727716728e-05,
+      "loss": 1.8242,
+      "step": 795
+    },
+    {
+      "epoch": 2.9157509157509156,
+      "grad_norm": 32.5267448425293,
+      "learning_rate": 4.725274725274725e-05,
+      "loss": 1.1197,
+      "step": 796
+    },
+    {
+      "epoch": 2.9194139194139193,
+      "grad_norm": 43.77764892578125,
+      "learning_rate": 4.722832722832723e-05,
+      "loss": 0.7322,
+      "step": 797
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 25.303524017333984,
+      "learning_rate": 4.720390720390721e-05,
+      "loss": 0.6557,
+      "step": 798
+    },
+    {
+      "epoch": 2.926739926739927,
+      "grad_norm": 23.90159797668457,
+      "learning_rate": 4.717948717948718e-05,
+      "loss": 0.2669,
+      "step": 799
+    },
+    {
+      "epoch": 2.9304029304029307,
+      "grad_norm": 21.20945930480957,
+      "learning_rate": 4.715506715506716e-05,
+      "loss": 0.3279,
+      "step": 800
+    },
+    {
+      "epoch": 2.9340659340659343,
+      "grad_norm": 28.819482803344727,
+      "learning_rate": 4.713064713064713e-05,
+      "loss": 0.717,
+      "step": 801
+    },
+    {
+      "epoch": 2.937728937728938,
+      "grad_norm": 9.13611125946045,
+      "learning_rate": 4.710622710622711e-05,
+      "loss": 0.1291,
+      "step": 802
+    },
+    {
+      "epoch": 2.9413919413919416,
+      "grad_norm": 22.16252326965332,
+      "learning_rate": 4.7081807081807085e-05,
+      "loss": 0.4406,
+      "step": 803
+    },
+    {
+      "epoch": 2.9450549450549453,
+      "grad_norm": 47.73503112792969,
+      "learning_rate": 4.705738705738706e-05,
+      "loss": 0.6176,
+      "step": 804
+    },
+    {
+      "epoch": 2.948717948717949,
+      "grad_norm": 61.73493576049805,
+      "learning_rate": 4.7032967032967035e-05,
+      "loss": 0.581,
+      "step": 805
+    },
+    {
+      "epoch": 2.9523809523809526,
+      "grad_norm": 22.48004722595215,
+      "learning_rate": 4.7008547008547014e-05,
+      "loss": 0.7404,
+      "step": 806
+    },
+    {
+      "epoch": 2.956043956043956,
+      "grad_norm": 54.2432746887207,
+      "learning_rate": 4.6984126984126986e-05,
+      "loss": 1.1522,
+      "step": 807
+    },
+    {
+      "epoch": 2.95970695970696,
+      "grad_norm": 26.221921920776367,
+      "learning_rate": 4.695970695970696e-05,
+      "loss": 0.4869,
+      "step": 808
+    },
+    {
+      "epoch": 2.9633699633699635,
+      "grad_norm": 21.688526153564453,
+      "learning_rate": 4.6935286935286936e-05,
+      "loss": 0.6639,
+      "step": 809
+    },
+    {
+      "epoch": 2.967032967032967,
+      "grad_norm": 5.81218147277832,
+      "learning_rate": 4.691086691086691e-05,
+      "loss": 0.0824,
+      "step": 810
+    },
+    {
+      "epoch": 2.970695970695971,
+      "grad_norm": 39.09580612182617,
+      "learning_rate": 4.6886446886446886e-05,
+      "loss": 1.5035,
+      "step": 811
+    },
+    {
+      "epoch": 2.9743589743589745,
+      "grad_norm": 24.587574005126953,
+      "learning_rate": 4.6862026862026864e-05,
+      "loss": 1.1107,
+      "step": 812
+    },
+    {
+      "epoch": 2.978021978021978,
+      "grad_norm": 25.25336265563965,
+      "learning_rate": 4.6837606837606836e-05,
+      "loss": 0.7764,
+      "step": 813
+    },
+    {
+      "epoch": 2.9816849816849818,
+      "grad_norm": 16.311378479003906,
+      "learning_rate": 4.6813186813186814e-05,
+      "loss": 0.4079,
+      "step": 814
+    },
+    {
+      "epoch": 2.9853479853479854,
+      "grad_norm": 19.0888729095459,
+      "learning_rate": 4.678876678876679e-05,
+      "loss": 0.5259,
+      "step": 815
+    },
+    {
+      "epoch": 2.989010989010989,
+      "grad_norm": 24.599462509155273,
+      "learning_rate": 4.6764346764346765e-05,
+      "loss": 0.7475,
+      "step": 816
+    },
+    {
+      "epoch": 2.9926739926739927,
+      "grad_norm": 20.4777889251709,
+      "learning_rate": 4.673992673992674e-05,
+      "loss": 0.356,
+      "step": 817
+    },
+    {
+      "epoch": 2.9963369963369964,
+      "grad_norm": 30.4327449798584,
+      "learning_rate": 4.671550671550672e-05,
+      "loss": 0.7958,
+      "step": 818
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 25.57271385192871,
+      "learning_rate": 4.669108669108669e-05,
+      "loss": 0.3918,
+      "step": 819
+    },
+    {
+      "epoch": 3.0036630036630036,
+      "grad_norm": 3.9672563076019287,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.0469,
+      "step": 820
+    },
+    {
+      "epoch": 3.0073260073260073,
+      "grad_norm": 6.657567501068115,
+      "learning_rate": 4.664224664224664e-05,
+      "loss": 0.0939,
+      "step": 821
+    },
+    {
+      "epoch": 3.010989010989011,
+      "grad_norm": 12.558409690856934,
+      "learning_rate": 4.6617826617826615e-05,
+      "loss": 0.1578,
+      "step": 822
+    },
+    {
+      "epoch": 3.0146520146520146,
+      "grad_norm": 18.909244537353516,
+      "learning_rate": 4.6593406593406593e-05,
+      "loss": 0.3209,
+      "step": 823
+    },
+    {
+      "epoch": 3.0183150183150182,
+      "grad_norm": 10.995687484741211,
+      "learning_rate": 4.656898656898657e-05,
+      "loss": 0.1198,
+      "step": 824
+    },
+    {
+      "epoch": 3.021978021978022,
+      "grad_norm": 16.14252471923828,
+      "learning_rate": 4.6544566544566544e-05,
+      "loss": 0.1431,
+      "step": 825
+    },
+    {
+      "epoch": 3.0256410256410255,
+      "grad_norm": 25.924381256103516,
+      "learning_rate": 4.652014652014652e-05,
+      "loss": 0.3989,
+      "step": 826
+    },
+    {
+      "epoch": 3.029304029304029,
+      "grad_norm": 4.87798547744751,
+      "learning_rate": 4.6495726495726494e-05,
+      "loss": 0.0472,
+      "step": 827
+    },
+    {
+      "epoch": 3.032967032967033,
+      "grad_norm": 15.078110694885254,
+      "learning_rate": 4.647130647130647e-05,
+      "loss": 0.1955,
+      "step": 828
+    },
+    {
+      "epoch": 3.0366300366300365,
+      "grad_norm": 19.74415397644043,
+      "learning_rate": 4.644688644688645e-05,
+      "loss": 0.1593,
+      "step": 829
+    },
+    {
+      "epoch": 3.04029304029304,
+      "grad_norm": 43.4788818359375,
+      "learning_rate": 4.642246642246642e-05,
+      "loss": 0.7917,
+      "step": 830
+    },
+    {
+      "epoch": 3.043956043956044,
+      "grad_norm": 27.122041702270508,
+      "learning_rate": 4.63980463980464e-05,
+      "loss": 0.1693,
+      "step": 831
+    },
+    {
+      "epoch": 3.0476190476190474,
+      "grad_norm": 9.51154899597168,
+      "learning_rate": 4.637362637362638e-05,
+      "loss": 0.0806,
+      "step": 832
+    },
+    {
+      "epoch": 3.051282051282051,
+      "grad_norm": 11.48532772064209,
+      "learning_rate": 4.634920634920635e-05,
+      "loss": 0.0815,
+      "step": 833
+    },
+    {
+      "epoch": 3.0549450549450547,
+      "grad_norm": 13.547063827514648,
+      "learning_rate": 4.632478632478632e-05,
+      "loss": 0.0817,
+      "step": 834
+    },
+    {
+      "epoch": 3.0586080586080584,
+      "grad_norm": 24.334409713745117,
+      "learning_rate": 4.63003663003663e-05,
+      "loss": 0.547,
+      "step": 835
+    },
+    {
+      "epoch": 3.062271062271062,
+      "grad_norm": 87.3517837524414,
+      "learning_rate": 4.627594627594627e-05,
+      "loss": 0.6534,
+      "step": 836
+    },
+    {
+      "epoch": 3.065934065934066,
+      "grad_norm": 16.100278854370117,
+      "learning_rate": 4.625152625152625e-05,
+      "loss": 0.2961,
+      "step": 837
+    },
+    {
+      "epoch": 3.06959706959707,
+      "grad_norm": 20.725875854492188,
+      "learning_rate": 4.622710622710623e-05,
+      "loss": 0.1114,
+      "step": 838
+    },
+    {
+      "epoch": 3.0732600732600734,
+      "grad_norm": 53.809722900390625,
+      "learning_rate": 4.62026862026862e-05,
+      "loss": 0.3808,
+      "step": 839
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 3.237959623336792,
+      "learning_rate": 4.617826617826618e-05,
+      "loss": 0.019,
+      "step": 840
+    },
+    {
+      "epoch": 3.0805860805860807,
+      "grad_norm": 69.71659088134766,
+      "learning_rate": 4.615384615384616e-05,
+      "loss": 1.0945,
+      "step": 841
+    },
+    {
+      "epoch": 3.0842490842490844,
+      "grad_norm": 31.005935668945312,
+      "learning_rate": 4.612942612942613e-05,
+      "loss": 0.3241,
+      "step": 842
+    },
+    {
+      "epoch": 3.087912087912088,
+      "grad_norm": 66.98394775390625,
+      "learning_rate": 4.610500610500611e-05,
+      "loss": 1.0213,
+      "step": 843
+    },
+    {
+      "epoch": 3.0915750915750917,
+      "grad_norm": 23.54532814025879,
+      "learning_rate": 4.608058608058609e-05,
+      "loss": 0.2188,
+      "step": 844
+    },
+    {
+      "epoch": 3.0952380952380953,
+      "grad_norm": 25.952709197998047,
+      "learning_rate": 4.605616605616606e-05,
+      "loss": 0.4305,
+      "step": 845
+    },
+    {
+      "epoch": 3.098901098901099,
+      "grad_norm": 36.100746154785156,
+      "learning_rate": 4.603174603174604e-05,
+      "loss": 0.6497,
+      "step": 846
+    },
+    {
+      "epoch": 3.1025641025641026,
+      "grad_norm": 60.34727478027344,
+      "learning_rate": 4.600732600732601e-05,
+      "loss": 0.3083,
+      "step": 847
+    },
+    {
+      "epoch": 3.1062271062271063,
+      "grad_norm": 35.265167236328125,
+      "learning_rate": 4.598290598290598e-05,
+      "loss": 0.3222,
+      "step": 848
+    },
+    {
+      "epoch": 3.10989010989011,
+      "grad_norm": 19.180070877075195,
+      "learning_rate": 4.595848595848596e-05,
+      "loss": 0.4065,
+      "step": 849
+    },
+    {
+      "epoch": 3.1135531135531136,
+      "grad_norm": 22.92152976989746,
+      "learning_rate": 4.593406593406593e-05,
+      "loss": 0.3998,
+      "step": 850
+    },
+    {
+      "epoch": 3.1172161172161172,
+      "grad_norm": 48.91377639770508,
+      "learning_rate": 4.590964590964591e-05,
+      "loss": 0.7035,
+      "step": 851
+    },
+    {
+      "epoch": 3.120879120879121,
+      "grad_norm": 11.615083694458008,
+      "learning_rate": 4.588522588522589e-05,
+      "loss": 0.3102,
+      "step": 852
+    },
+    {
+      "epoch": 3.1245421245421245,
+      "grad_norm": 23.573801040649414,
+      "learning_rate": 4.586080586080586e-05,
+      "loss": 0.3358,
+      "step": 853
+    },
+    {
+      "epoch": 3.128205128205128,
+      "grad_norm": 16.903776168823242,
+      "learning_rate": 4.583638583638584e-05,
+      "loss": 0.2973,
+      "step": 854
+    },
+    {
+      "epoch": 3.131868131868132,
+      "grad_norm": 6.052688121795654,
+      "learning_rate": 4.5811965811965816e-05,
+      "loss": 0.0671,
+      "step": 855
+    },
+    {
+      "epoch": 3.1355311355311355,
+      "grad_norm": 34.40020751953125,
+      "learning_rate": 4.578754578754579e-05,
+      "loss": 0.508,
+      "step": 856
+    },
+    {
+      "epoch": 3.139194139194139,
+      "grad_norm": 21.39589500427246,
+      "learning_rate": 4.5763125763125766e-05,
+      "loss": 0.0805,
+      "step": 857
+    },
+    {
+      "epoch": 3.142857142857143,
+      "grad_norm": 24.03894805908203,
+      "learning_rate": 4.5738705738705744e-05,
+      "loss": 0.1884,
+      "step": 858
+    },
+    {
+      "epoch": 3.1465201465201464,
+      "grad_norm": 66.53777313232422,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.5235,
+      "step": 859
+    },
+    {
+      "epoch": 3.15018315018315,
+      "grad_norm": 33.663490295410156,
+      "learning_rate": 4.568986568986569e-05,
+      "loss": 0.7579,
+      "step": 860
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 30.173309326171875,
+      "learning_rate": 4.5665445665445666e-05,
+      "loss": 0.2263,
+      "step": 861
+    },
+    {
+      "epoch": 3.1575091575091574,
+      "grad_norm": 37.52082824707031,
+      "learning_rate": 4.564102564102564e-05,
+      "loss": 0.5695,
+      "step": 862
+    },
+    {
+      "epoch": 3.161172161172161,
+      "grad_norm": 38.86849594116211,
+      "learning_rate": 4.5616605616605616e-05,
+      "loss": 0.6981,
+      "step": 863
+    },
+    {
+      "epoch": 3.1648351648351647,
+      "grad_norm": 42.702247619628906,
+      "learning_rate": 4.5592185592185595e-05,
+      "loss": 0.9864,
+      "step": 864
+    },
+    {
+      "epoch": 3.1684981684981683,
+      "grad_norm": 16.60870361328125,
+      "learning_rate": 4.5567765567765566e-05,
+      "loss": 0.1595,
+      "step": 865
+    },
+    {
+      "epoch": 3.172161172161172,
+      "grad_norm": 26.309768676757812,
+      "learning_rate": 4.5543345543345545e-05,
+      "loss": 0.4028,
+      "step": 866
+    },
+    {
+      "epoch": 3.1758241758241756,
+      "grad_norm": 45.7955322265625,
+      "learning_rate": 4.551892551892552e-05,
+      "loss": 1.1258,
+      "step": 867
+    },
+    {
+      "epoch": 3.1794871794871793,
+      "grad_norm": 25.780302047729492,
+      "learning_rate": 4.5494505494505495e-05,
+      "loss": 0.4018,
+      "step": 868
+    },
+    {
+      "epoch": 3.183150183150183,
+      "grad_norm": 41.65156555175781,
+      "learning_rate": 4.5470085470085474e-05,
+      "loss": 0.4543,
+      "step": 869
+    },
+    {
+      "epoch": 3.186813186813187,
+      "grad_norm": 56.92537307739258,
+      "learning_rate": 4.544566544566545e-05,
+      "loss": 0.334,
+      "step": 870
+    },
+    {
+      "epoch": 3.1904761904761907,
+      "grad_norm": 19.44786262512207,
+      "learning_rate": 4.5421245421245424e-05,
+      "loss": 0.2855,
+      "step": 871
+    },
+    {
+      "epoch": 3.1941391941391943,
+      "grad_norm": 19.75824546813965,
+      "learning_rate": 4.53968253968254e-05,
+      "loss": 0.2589,
+      "step": 872
+    },
+    {
+      "epoch": 3.197802197802198,
+      "grad_norm": 30.935569763183594,
+      "learning_rate": 4.5372405372405374e-05,
+      "loss": 0.5083,
+      "step": 873
+    },
+    {
+      "epoch": 3.2014652014652016,
+      "grad_norm": 32.59378433227539,
+      "learning_rate": 4.5347985347985345e-05,
+      "loss": 0.6806,
+      "step": 874
+    },
+    {
+      "epoch": 3.2051282051282053,
+      "grad_norm": 32.7809944152832,
+      "learning_rate": 4.5323565323565324e-05,
+      "loss": 0.7094,
+      "step": 875
+    },
+    {
+      "epoch": 3.208791208791209,
+      "grad_norm": 22.95226287841797,
+      "learning_rate": 4.5299145299145296e-05,
+      "loss": 0.3871,
+      "step": 876
+    },
+    {
+      "epoch": 3.2124542124542126,
+      "grad_norm": 13.90613079071045,
+      "learning_rate": 4.5274725274725274e-05,
+      "loss": 0.2049,
+      "step": 877
+    },
+    {
+      "epoch": 3.2161172161172162,
+      "grad_norm": 36.79647445678711,
+      "learning_rate": 4.525030525030525e-05,
+      "loss": 0.959,
+      "step": 878
+    },
+    {
+      "epoch": 3.21978021978022,
+      "grad_norm": 16.770553588867188,
+      "learning_rate": 4.5225885225885224e-05,
+      "loss": 0.3061,
+      "step": 879
+    },
+    {
+      "epoch": 3.2234432234432235,
+      "grad_norm": 22.241527557373047,
+      "learning_rate": 4.52014652014652e-05,
+      "loss": 0.1961,
+      "step": 880
+    },
+    {
+      "epoch": 3.227106227106227,
+      "grad_norm": 51.097957611083984,
+      "learning_rate": 4.517704517704518e-05,
+      "loss": 0.5272,
+      "step": 881
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 43.70039749145508,
+      "learning_rate": 4.515262515262515e-05,
+      "loss": 0.6764,
+      "step": 882
+    },
+    {
+      "epoch": 3.2344322344322345,
+      "grad_norm": 30.666664123535156,
+      "learning_rate": 4.512820512820513e-05,
+      "loss": 0.6524,
+      "step": 883
+    },
+    {
+      "epoch": 3.238095238095238,
+      "grad_norm": 16.787954330444336,
+      "learning_rate": 4.510378510378511e-05,
+      "loss": 0.178,
+      "step": 884
+    },
+    {
+      "epoch": 3.241758241758242,
+      "grad_norm": 32.14992904663086,
+      "learning_rate": 4.507936507936508e-05,
+      "loss": 0.6206,
+      "step": 885
+    },
+    {
+      "epoch": 3.2454212454212454,
+      "grad_norm": 24.926103591918945,
+      "learning_rate": 4.505494505494505e-05,
+      "loss": 0.4696,
+      "step": 886
+    },
+    {
+      "epoch": 3.249084249084249,
+      "grad_norm": 31.044967651367188,
+      "learning_rate": 4.503052503052503e-05,
+      "loss": 0.3021,
+      "step": 887
+    },
+    {
+      "epoch": 3.2527472527472527,
+      "grad_norm": 10.355696678161621,
+      "learning_rate": 4.5006105006105e-05,
+      "loss": 0.0784,
+      "step": 888
+    },
+    {
+      "epoch": 3.2564102564102564,
+      "grad_norm": 28.19644546508789,
+      "learning_rate": 4.498168498168498e-05,
+      "loss": 0.234,
+      "step": 889
+    },
+    {
+      "epoch": 3.26007326007326,
+      "grad_norm": 21.245389938354492,
+      "learning_rate": 4.495726495726496e-05,
+      "loss": 0.2895,
+      "step": 890
+    },
+    {
+      "epoch": 3.2637362637362637,
+      "grad_norm": 27.337587356567383,
+      "learning_rate": 4.493284493284493e-05,
+      "loss": 0.4614,
+      "step": 891
+    },
+    {
+      "epoch": 3.2673992673992673,
+      "grad_norm": 37.06135177612305,
+      "learning_rate": 4.490842490842491e-05,
+      "loss": 0.2717,
+      "step": 892
+    },
+    {
+      "epoch": 3.271062271062271,
+      "grad_norm": 26.85171890258789,
+      "learning_rate": 4.488400488400489e-05,
+      "loss": 0.4965,
+      "step": 893
+    },
+    {
+      "epoch": 3.2747252747252746,
+      "grad_norm": 41.79130935668945,
+      "learning_rate": 4.485958485958486e-05,
+      "loss": 0.4209,
+      "step": 894
+    },
+    {
+      "epoch": 3.2783882783882783,
+      "grad_norm": 32.75770950317383,
+      "learning_rate": 4.483516483516484e-05,
+      "loss": 0.5126,
+      "step": 895
+    },
+    {
+      "epoch": 3.282051282051282,
+      "grad_norm": 67.75275421142578,
+      "learning_rate": 4.481074481074482e-05,
+      "loss": 0.8257,
+      "step": 896
+    },
+    {
+      "epoch": 3.2857142857142856,
+      "grad_norm": 36.773319244384766,
+      "learning_rate": 4.478632478632479e-05,
+      "loss": 1.6113,
+      "step": 897
+    },
+    {
+      "epoch": 3.2893772893772892,
+      "grad_norm": 60.94101333618164,
+      "learning_rate": 4.476190476190476e-05,
+      "loss": 0.7996,
+      "step": 898
+    },
+    {
+      "epoch": 3.293040293040293,
+      "grad_norm": 45.40288162231445,
+      "learning_rate": 4.473748473748474e-05,
+      "loss": 0.7139,
+      "step": 899
+    },
+    {
+      "epoch": 3.2967032967032965,
+      "grad_norm": 27.4019718170166,
+      "learning_rate": 4.471306471306471e-05,
+      "loss": 0.4695,
+      "step": 900
+    },
+    {
+      "epoch": 3.3003663003663,
+      "grad_norm": 20.126493453979492,
+      "learning_rate": 4.468864468864469e-05,
+      "loss": 0.2181,
+      "step": 901
+    },
+    {
+      "epoch": 3.304029304029304,
+      "grad_norm": 37.28034591674805,
+      "learning_rate": 4.466422466422466e-05,
+      "loss": 0.8902,
+      "step": 902
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 15.40217113494873,
+      "learning_rate": 4.463980463980464e-05,
+      "loss": 0.2428,
+      "step": 903
+    },
+    {
+      "epoch": 3.311355311355311,
+      "grad_norm": 21.924699783325195,
+      "learning_rate": 4.461538461538462e-05,
+      "loss": 0.3271,
+      "step": 904
+    },
+    {
+      "epoch": 3.315018315018315,
+      "grad_norm": 29.787410736083984,
+      "learning_rate": 4.459096459096459e-05,
+      "loss": 0.5914,
+      "step": 905
+    },
+    {
+      "epoch": 3.3186813186813184,
+      "grad_norm": 16.91995620727539,
+      "learning_rate": 4.456654456654457e-05,
+      "loss": 0.3442,
+      "step": 906
+    },
+    {
+      "epoch": 3.3223443223443225,
+      "grad_norm": 13.232250213623047,
+      "learning_rate": 4.4542124542124546e-05,
+      "loss": 0.1977,
+      "step": 907
+    },
+    {
+      "epoch": 3.326007326007326,
+      "grad_norm": 25.45724868774414,
+      "learning_rate": 4.451770451770452e-05,
+      "loss": 0.8241,
+      "step": 908
+    },
+    {
+      "epoch": 3.32967032967033,
+      "grad_norm": 20.996292114257812,
+      "learning_rate": 4.4493284493284496e-05,
+      "loss": 0.3154,
+      "step": 909
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 28.150684356689453,
+      "learning_rate": 4.4468864468864475e-05,
+      "loss": 0.4077,
+      "step": 910
+    },
+    {
+      "epoch": 3.336996336996337,
+      "grad_norm": 57.184322357177734,
+      "learning_rate": 4.444444444444444e-05,
+      "loss": 0.5701,
+      "step": 911
+    },
+    {
+      "epoch": 3.340659340659341,
+      "grad_norm": 26.231369018554688,
+      "learning_rate": 4.442002442002442e-05,
+      "loss": 0.4427,
+      "step": 912
+    },
+    {
+      "epoch": 3.3443223443223444,
+      "grad_norm": 32.52253723144531,
+      "learning_rate": 4.43956043956044e-05,
+      "loss": 1.014,
+      "step": 913
+    },
+    {
+      "epoch": 3.347985347985348,
+      "grad_norm": 19.39035987854004,
+      "learning_rate": 4.437118437118437e-05,
+      "loss": 0.1567,
+      "step": 914
+    },
+    {
+      "epoch": 3.3516483516483517,
+      "grad_norm": 24.542327880859375,
+      "learning_rate": 4.434676434676435e-05,
+      "loss": 0.5478,
+      "step": 915
+    },
+    {
+      "epoch": 3.3553113553113554,
+      "grad_norm": 46.6158447265625,
+      "learning_rate": 4.4322344322344325e-05,
+      "loss": 0.5636,
+      "step": 916
+    },
+    {
+      "epoch": 3.358974358974359,
+      "grad_norm": 36.008846282958984,
+      "learning_rate": 4.42979242979243e-05,
+      "loss": 0.4401,
+      "step": 917
+    },
+    {
+      "epoch": 3.3626373626373627,
+      "grad_norm": 6.922544956207275,
+      "learning_rate": 4.4273504273504275e-05,
+      "loss": 0.0885,
+      "step": 918
+    },
+    {
+      "epoch": 3.3663003663003663,
+      "grad_norm": 25.707748413085938,
+      "learning_rate": 4.4249084249084254e-05,
+      "loss": 0.3235,
+      "step": 919
+    },
+    {
+      "epoch": 3.36996336996337,
+      "grad_norm": 47.98778533935547,
+      "learning_rate": 4.4224664224664226e-05,
+      "loss": 1.3738,
+      "step": 920
+    },
+    {
+      "epoch": 3.3736263736263736,
+      "grad_norm": 26.64824104309082,
+      "learning_rate": 4.4200244200244204e-05,
+      "loss": 0.8405,
+      "step": 921
+    },
+    {
+      "epoch": 3.3772893772893773,
+      "grad_norm": 30.66206169128418,
+      "learning_rate": 4.417582417582418e-05,
+      "loss": 1.3021,
+      "step": 922
+    },
+    {
+      "epoch": 3.380952380952381,
+      "grad_norm": 33.15909194946289,
+      "learning_rate": 4.4151404151404154e-05,
+      "loss": 0.3064,
+      "step": 923
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 78.46485137939453,
+      "learning_rate": 4.4126984126984126e-05,
+      "loss": 0.6526,
+      "step": 924
+    },
+    {
+      "epoch": 3.3882783882783882,
+      "grad_norm": 45.584747314453125,
+      "learning_rate": 4.4102564102564104e-05,
+      "loss": 0.9546,
+      "step": 925
+    },
+    {
+      "epoch": 3.391941391941392,
+      "grad_norm": 23.244487762451172,
+      "learning_rate": 4.4078144078144076e-05,
+      "loss": 0.3334,
+      "step": 926
+    },
+    {
+      "epoch": 3.3956043956043955,
+      "grad_norm": 9.296119689941406,
+      "learning_rate": 4.4053724053724054e-05,
+      "loss": 0.1045,
+      "step": 927
+    },
+    {
+      "epoch": 3.399267399267399,
+      "grad_norm": 15.207316398620605,
+      "learning_rate": 4.4029304029304026e-05,
+      "loss": 0.087,
+      "step": 928
+    },
+    {
+      "epoch": 3.402930402930403,
+      "grad_norm": 20.554912567138672,
+      "learning_rate": 4.4004884004884005e-05,
+      "loss": 0.2658,
+      "step": 929
+    },
+    {
+      "epoch": 3.4065934065934065,
+      "grad_norm": 25.304515838623047,
+      "learning_rate": 4.398046398046398e-05,
+      "loss": 0.2862,
+      "step": 930
+    },
+    {
+      "epoch": 3.41025641025641,
+      "grad_norm": 44.320377349853516,
+      "learning_rate": 4.3956043956043955e-05,
+      "loss": 1.1972,
+      "step": 931
+    },
+    {
+      "epoch": 3.413919413919414,
+      "grad_norm": 21.3024845123291,
+      "learning_rate": 4.393162393162393e-05,
+      "loss": 0.2193,
+      "step": 932
+    },
+    {
+      "epoch": 3.4175824175824174,
+      "grad_norm": 12.274759292602539,
+      "learning_rate": 4.390720390720391e-05,
+      "loss": 0.1033,
+      "step": 933
+    },
+    {
+      "epoch": 3.421245421245421,
+      "grad_norm": 29.188446044921875,
+      "learning_rate": 4.388278388278388e-05,
+      "loss": 0.8143,
+      "step": 934
+    },
+    {
+      "epoch": 3.4249084249084247,
+      "grad_norm": 11.880194664001465,
+      "learning_rate": 4.385836385836386e-05,
+      "loss": 0.0932,
+      "step": 935
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 28.859825134277344,
+      "learning_rate": 4.383394383394384e-05,
+      "loss": 0.6026,
+      "step": 936
+    },
+    {
+      "epoch": 3.4322344322344325,
+      "grad_norm": 25.131824493408203,
+      "learning_rate": 4.3809523809523805e-05,
+      "loss": 0.4023,
+      "step": 937
+    },
+    {
+      "epoch": 3.435897435897436,
+      "grad_norm": 35.04637145996094,
+      "learning_rate": 4.3785103785103783e-05,
+      "loss": 0.7765,
+      "step": 938
+    },
+    {
+      "epoch": 3.4395604395604398,
+      "grad_norm": 15.831666946411133,
+      "learning_rate": 4.376068376068376e-05,
+      "loss": 0.1779,
+      "step": 939
+    },
+    {
+      "epoch": 3.4432234432234434,
+      "grad_norm": 26.455148696899414,
+      "learning_rate": 4.3736263736263734e-05,
+      "loss": 0.3165,
+      "step": 940
+    },
+    {
+      "epoch": 3.446886446886447,
+      "grad_norm": 23.840030670166016,
+      "learning_rate": 4.371184371184371e-05,
+      "loss": 0.5363,
+      "step": 941
+    },
+    {
+      "epoch": 3.4505494505494507,
+      "grad_norm": 30.517026901245117,
+      "learning_rate": 4.368742368742369e-05,
+      "loss": 0.422,
+      "step": 942
+    },
+    {
+      "epoch": 3.4542124542124544,
+      "grad_norm": 51.574703216552734,
+      "learning_rate": 4.366300366300366e-05,
+      "loss": 1.5333,
+      "step": 943
+    },
+    {
+      "epoch": 3.457875457875458,
+      "grad_norm": 57.92119216918945,
+      "learning_rate": 4.363858363858364e-05,
+      "loss": 0.5732,
+      "step": 944
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 34.3664436340332,
+      "learning_rate": 4.361416361416362e-05,
+      "loss": 0.5054,
+      "step": 945
+    },
+    {
+      "epoch": 3.4652014652014653,
+      "grad_norm": 14.034111976623535,
+      "learning_rate": 4.358974358974359e-05,
+      "loss": 0.0969,
+      "step": 946
+    },
+    {
+      "epoch": 3.468864468864469,
+      "grad_norm": 15.058267593383789,
+      "learning_rate": 4.356532356532357e-05,
+      "loss": 0.1877,
+      "step": 947
+    },
+    {
+      "epoch": 3.4725274725274726,
+      "grad_norm": 18.598024368286133,
+      "learning_rate": 4.354090354090355e-05,
+      "loss": 0.2378,
+      "step": 948
+    },
+    {
+      "epoch": 3.4761904761904763,
+      "grad_norm": 17.926319122314453,
+      "learning_rate": 4.351648351648352e-05,
+      "loss": 0.2935,
+      "step": 949
+    },
+    {
+      "epoch": 3.47985347985348,
+      "grad_norm": 8.25291633605957,
+      "learning_rate": 4.349206349206349e-05,
+      "loss": 0.0891,
+      "step": 950
+    },
+    {
+      "epoch": 3.4835164835164836,
+      "grad_norm": 26.152061462402344,
+      "learning_rate": 4.346764346764347e-05,
+      "loss": 0.2798,
+      "step": 951
+    },
+    {
+      "epoch": 3.4871794871794872,
+      "grad_norm": 22.669677734375,
+      "learning_rate": 4.344322344322344e-05,
+      "loss": 0.506,
+      "step": 952
+    },
+    {
+      "epoch": 3.490842490842491,
+      "grad_norm": 18.439355850219727,
+      "learning_rate": 4.341880341880342e-05,
+      "loss": 0.3034,
+      "step": 953
+    },
+    {
+      "epoch": 3.4945054945054945,
+      "grad_norm": 30.48084259033203,
+      "learning_rate": 4.339438339438339e-05,
+      "loss": 0.4366,
+      "step": 954
+    },
+    {
+      "epoch": 3.498168498168498,
+      "grad_norm": 51.792381286621094,
+      "learning_rate": 4.336996336996337e-05,
+      "loss": 0.5214,
+      "step": 955
+    },
+    {
+      "epoch": 3.501831501831502,
+      "grad_norm": 44.70718002319336,
+      "learning_rate": 4.334554334554335e-05,
+      "loss": 0.7823,
+      "step": 956
+    },
+    {
+      "epoch": 3.5054945054945055,
+      "grad_norm": 42.00168991088867,
+      "learning_rate": 4.332112332112332e-05,
+      "loss": 0.9207,
+      "step": 957
+    },
+    {
+      "epoch": 3.509157509157509,
+      "grad_norm": 28.97800636291504,
+      "learning_rate": 4.32967032967033e-05,
+      "loss": 0.279,
+      "step": 958
+    },
+    {
+      "epoch": 3.5128205128205128,
+      "grad_norm": 21.902843475341797,
+      "learning_rate": 4.327228327228328e-05,
+      "loss": 0.1969,
+      "step": 959
+    },
+    {
+      "epoch": 3.5164835164835164,
+      "grad_norm": 14.560053825378418,
+      "learning_rate": 4.324786324786325e-05,
+      "loss": 0.0976,
+      "step": 960
+    },
+    {
+      "epoch": 3.52014652014652,
+      "grad_norm": 4.2637104988098145,
+      "learning_rate": 4.322344322344323e-05,
+      "loss": 0.0277,
+      "step": 961
+    },
+    {
+      "epoch": 3.5238095238095237,
+      "grad_norm": 52.4840202331543,
+      "learning_rate": 4.3199023199023205e-05,
+      "loss": 0.2967,
+      "step": 962
+    },
+    {
+      "epoch": 3.5274725274725274,
+      "grad_norm": 48.95661163330078,
+      "learning_rate": 4.317460317460317e-05,
+      "loss": 0.2904,
+      "step": 963
+    },
+    {
+      "epoch": 3.531135531135531,
+      "grad_norm": 79.46379089355469,
+      "learning_rate": 4.315018315018315e-05,
+      "loss": 0.1644,
+      "step": 964
+    },
+    {
+      "epoch": 3.5347985347985347,
+      "grad_norm": 29.678428649902344,
+      "learning_rate": 4.312576312576313e-05,
+      "loss": 0.3498,
+      "step": 965
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 32.71342086791992,
+      "learning_rate": 4.31013431013431e-05,
+      "loss": 0.3509,
+      "step": 966
+    },
+    {
+      "epoch": 3.542124542124542,
+      "grad_norm": 6.679911136627197,
+      "learning_rate": 4.307692307692308e-05,
+      "loss": 0.0658,
+      "step": 967
+    },
+    {
+      "epoch": 3.5457875457875456,
+      "grad_norm": 19.8692626953125,
+      "learning_rate": 4.3052503052503056e-05,
+      "loss": 0.1626,
+      "step": 968
+    },
+    {
+      "epoch": 3.5494505494505493,
+      "grad_norm": 17.69087791442871,
+      "learning_rate": 4.302808302808303e-05,
+      "loss": 0.2592,
+      "step": 969
+    },
+    {
+      "epoch": 3.553113553113553,
+      "grad_norm": 11.734158515930176,
+      "learning_rate": 4.3003663003663006e-05,
+      "loss": 0.1007,
+      "step": 970
+    },
+    {
+      "epoch": 3.5567765567765566,
+      "grad_norm": 34.51172637939453,
+      "learning_rate": 4.2979242979242984e-05,
+      "loss": 0.2823,
+      "step": 971
+    },
+    {
+      "epoch": 3.5604395604395602,
+      "grad_norm": 15.009514808654785,
+      "learning_rate": 4.2954822954822956e-05,
+      "loss": 0.1203,
+      "step": 972
+    },
+    {
+      "epoch": 3.564102564102564,
+      "grad_norm": 67.92166137695312,
+      "learning_rate": 4.2930402930402934e-05,
+      "loss": 0.396,
+      "step": 973
+    },
+    {
+      "epoch": 3.5677655677655675,
+      "grad_norm": 66.84014129638672,
+      "learning_rate": 4.290598290598291e-05,
+      "loss": 0.6545,
+      "step": 974
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 25.811107635498047,
+      "learning_rate": 4.2881562881562885e-05,
+      "loss": 0.1747,
+      "step": 975
+    },
+    {
+      "epoch": 3.575091575091575,
+      "grad_norm": 100.88753509521484,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 0.3991,
+      "step": 976
+    },
+    {
+      "epoch": 3.578754578754579,
+      "grad_norm": 34.51667785644531,
+      "learning_rate": 4.2832722832722835e-05,
+      "loss": 0.1365,
+      "step": 977
+    },
+    {
+      "epoch": 3.5824175824175826,
+      "grad_norm": 26.852561950683594,
+      "learning_rate": 4.2808302808302806e-05,
+      "loss": 0.3627,
+      "step": 978
+    },
+    {
+      "epoch": 3.586080586080586,
+      "grad_norm": 24.968570709228516,
+      "learning_rate": 4.2783882783882785e-05,
+      "loss": 0.2106,
+      "step": 979
+    },
+    {
+      "epoch": 3.58974358974359,
+      "grad_norm": 27.33326530456543,
+      "learning_rate": 4.2759462759462757e-05,
+      "loss": 0.1758,
+      "step": 980
+    },
+    {
+      "epoch": 3.5934065934065935,
+      "grad_norm": 52.63814926147461,
+      "learning_rate": 4.2735042735042735e-05,
+      "loss": 0.601,
+      "step": 981
+    },
+    {
+      "epoch": 3.597069597069597,
+      "grad_norm": 37.77897262573242,
+      "learning_rate": 4.2710622710622713e-05,
+      "loss": 0.5299,
+      "step": 982
+    },
+    {
+      "epoch": 3.600732600732601,
+      "grad_norm": 27.691659927368164,
+      "learning_rate": 4.2686202686202685e-05,
+      "loss": 0.1784,
+      "step": 983
+    },
+    {
+      "epoch": 3.6043956043956045,
+      "grad_norm": 106.33782958984375,
+      "learning_rate": 4.2661782661782664e-05,
+      "loss": 0.8859,
+      "step": 984
+    },
+    {
+      "epoch": 3.608058608058608,
+      "grad_norm": 22.95706558227539,
+      "learning_rate": 4.263736263736264e-05,
+      "loss": 0.1611,
+      "step": 985
+    },
+    {
+      "epoch": 3.6117216117216118,
+      "grad_norm": 22.72148895263672,
+      "learning_rate": 4.2612942612942614e-05,
+      "loss": 0.1561,
+      "step": 986
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 93.37244415283203,
+      "learning_rate": 4.258852258852259e-05,
+      "loss": 0.4287,
+      "step": 987
+    },
+    {
+      "epoch": 3.619047619047619,
+      "grad_norm": 51.54584884643555,
+      "learning_rate": 4.2564102564102564e-05,
+      "loss": 0.6292,
+      "step": 988
+    },
+    {
+      "epoch": 3.6227106227106227,
+      "grad_norm": 61.58243942260742,
+      "learning_rate": 4.2539682539682536e-05,
+      "loss": 1.3205,
+      "step": 989
+    },
+    {
+      "epoch": 3.6263736263736264,
+      "grad_norm": 70.59432220458984,
+      "learning_rate": 4.2515262515262514e-05,
+      "loss": 0.7451,
+      "step": 990
+    },
+    {
+      "epoch": 3.63003663003663,
+      "grad_norm": 76.28730773925781,
+      "learning_rate": 4.249084249084249e-05,
+      "loss": 2.0314,
+      "step": 991
+    },
+    {
+      "epoch": 3.6336996336996337,
+      "grad_norm": 73.5402603149414,
+      "learning_rate": 4.2466422466422464e-05,
+      "loss": 1.6628,
+      "step": 992
+    },
+    {
+      "epoch": 3.6373626373626373,
+      "grad_norm": 75.8978042602539,
+      "learning_rate": 4.244200244200244e-05,
+      "loss": 1.652,
+      "step": 993
+    },
+    {
+      "epoch": 3.641025641025641,
+      "grad_norm": 37.04104232788086,
+      "learning_rate": 4.241758241758242e-05,
+      "loss": 1.5356,
+      "step": 994
+    },
+    {
+      "epoch": 3.6446886446886446,
+      "grad_norm": 34.31178283691406,
+      "learning_rate": 4.239316239316239e-05,
+      "loss": 1.1783,
+      "step": 995
+    },
+    {
+      "epoch": 3.6483516483516483,
+      "grad_norm": 22.934877395629883,
+      "learning_rate": 4.236874236874237e-05,
+      "loss": 1.2995,
+      "step": 996
+    },
+    {
+      "epoch": 3.652014652014652,
+      "grad_norm": 30.25251579284668,
+      "learning_rate": 4.234432234432235e-05,
+      "loss": 1.1304,
+      "step": 997
+    },
+    {
+      "epoch": 3.6556776556776556,
+      "grad_norm": 35.082027435302734,
+      "learning_rate": 4.231990231990232e-05,
+      "loss": 1.0827,
+      "step": 998
+    },
+    {
+      "epoch": 3.659340659340659,
+      "grad_norm": 24.526325225830078,
+      "learning_rate": 4.22954822954823e-05,
+      "loss": 0.8716,
+      "step": 999
+    },
+    {
+      "epoch": 3.663003663003663,
+      "grad_norm": 29.882883071899414,
+      "learning_rate": 4.227106227106228e-05,
+      "loss": 0.5432,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2730,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}