diff --git "a/checkpoint-1199/trainer_state.json" "b/checkpoint-1199/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1199/trainer_state.json"
@@ -0,0 +1,8450 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997915363769022,
+  "eval_steps": 400,
+  "global_step": 1199,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008338544923910778,
+      "grad_norm": 0.35252463817596436,
+      "learning_rate": 2e-08,
+      "loss": 0.8273,
+      "step": 1
+    },
+    {
+      "epoch": 0.0008338544923910778,
+      "eval_loss": 0.861491858959198,
+      "eval_runtime": 319.044,
+      "eval_samples_per_second": 16.872,
+      "eval_steps_per_second": 2.815,
+      "step": 1
+    },
+    {
+      "epoch": 0.0016677089847821555,
+      "grad_norm": 0.3189675807952881,
+      "learning_rate": 4e-08,
+      "loss": 0.8572,
+      "step": 2
+    },
+    {
+      "epoch": 0.0025015634771732333,
+      "grad_norm": 0.3477325141429901,
+      "learning_rate": 6e-08,
+      "loss": 0.9059,
+      "step": 3
+    },
+    {
+      "epoch": 0.003335417969564311,
+      "grad_norm": 0.32423585653305054,
+      "learning_rate": 8e-08,
+      "loss": 0.8386,
+      "step": 4
+    },
+    {
+      "epoch": 0.004169272461955389,
+      "grad_norm": 0.32164493203163147,
+      "learning_rate": 1e-07,
+      "loss": 0.8753,
+      "step": 5
+    },
+    {
+      "epoch": 0.0050031269543464665,
+      "grad_norm": 0.33231157064437866,
+      "learning_rate": 1.2e-07,
+      "loss": 0.8665,
+      "step": 6
+    },
+    {
+      "epoch": 0.005836981446737544,
+      "grad_norm": 0.31154102087020874,
+      "learning_rate": 1.4e-07,
+      "loss": 0.8813,
+      "step": 7
+    },
+    {
+      "epoch": 0.006670835939128622,
+      "grad_norm": 0.3176712393760681,
+      "learning_rate": 1.6e-07,
+      "loss": 0.8683,
+      "step": 8
+    },
+    {
+      "epoch": 0.0075046904315197,
+      "grad_norm": 0.3419969975948334,
+      "learning_rate": 1.8e-07,
+      "loss": 0.849,
+      "step": 9
+    },
+    {
+      "epoch": 0.008338544923910778,
+      "grad_norm": 0.33408281207084656,
+      "learning_rate": 2e-07,
+      "loss": 0.8778,
+      "step": 10
+    },
+    {
+      "epoch": 0.009172399416301855,
+      "grad_norm": 0.3125899136066437,
+      "learning_rate": 2.1999999999999998e-07,
+      "loss": 0.8051,
+      "step": 11
+    },
+    {
+      "epoch": 0.010006253908692933,
+      "grad_norm": 0.32648083567619324,
+      "learning_rate": 2.4e-07,
+      "loss": 0.8297,
+      "step": 12
+    },
+    {
+      "epoch": 0.01084010840108401,
+      "grad_norm": 0.3143678605556488,
+      "learning_rate": 2.6e-07,
+      "loss": 0.8401,
+      "step": 13
+    },
+    {
+      "epoch": 0.011673962893475089,
+      "grad_norm": 0.3291407525539398,
+      "learning_rate": 2.8e-07,
+      "loss": 0.8668,
+      "step": 14
+    },
+    {
+      "epoch": 0.012507817385866166,
+      "grad_norm": 0.3211856484413147,
+      "learning_rate": 3e-07,
+      "loss": 0.84,
+      "step": 15
+    },
+    {
+      "epoch": 0.013341671878257244,
+      "grad_norm": 0.31169289350509644,
+      "learning_rate": 3.2e-07,
+      "loss": 0.8659,
+      "step": 16
+    },
+    {
+      "epoch": 0.014175526370648322,
+      "grad_norm": 0.3279392719268799,
+      "learning_rate": 3.4000000000000003e-07,
+      "loss": 0.8317,
+      "step": 17
+    },
+    {
+      "epoch": 0.0150093808630394,
+      "grad_norm": 0.3028814494609833,
+      "learning_rate": 3.6e-07,
+      "loss": 0.8246,
+      "step": 18
+    },
+    {
+      "epoch": 0.015843235355430477,
+      "grad_norm": 0.307066410779953,
+      "learning_rate": 3.7999999999999996e-07,
+      "loss": 0.8025,
+      "step": 19
+    },
+    {
+      "epoch": 0.016677089847821555,
+      "grad_norm": 0.3293515741825104,
+      "learning_rate": 4e-07,
+      "loss": 0.8674,
+      "step": 20
+    },
+    {
+      "epoch": 0.017510944340212633,
+      "grad_norm": 0.3215535581111908,
+      "learning_rate": 4.1999999999999995e-07,
+      "loss": 0.8533,
+      "step": 21
+    },
+    {
+      "epoch": 0.01834479883260371,
+      "grad_norm": 0.3236245810985565,
+      "learning_rate": 4.3999999999999997e-07,
+      "loss": 0.8957,
+      "step": 22
+    },
+    {
+      "epoch": 0.01917865332499479,
+      "grad_norm": 0.33956846594810486,
+      "learning_rate": 4.6e-07,
+      "loss": 0.8293,
+      "step": 23
+    },
+    {
+      "epoch": 0.020012507817385866,
+      "grad_norm": 0.33766457438468933,
+      "learning_rate": 4.8e-07,
+      "loss": 0.8381,
+      "step": 24
+    },
+    {
+      "epoch": 0.020846362309776944,
+      "grad_norm": 0.3336811065673828,
+      "learning_rate": 5e-07,
+      "loss": 0.8475,
+      "step": 25
+    },
+    {
+      "epoch": 0.02168021680216802,
+      "grad_norm": 0.307309627532959,
+      "learning_rate": 5.2e-07,
+      "loss": 0.852,
+      "step": 26
+    },
+    {
+      "epoch": 0.0225140712945591,
+      "grad_norm": 0.29773110151290894,
+      "learning_rate": 5.4e-07,
+      "loss": 0.8042,
+      "step": 27
+    },
+    {
+      "epoch": 0.023347925786950177,
+      "grad_norm": 0.3210948407649994,
+      "learning_rate": 5.6e-07,
+      "loss": 0.9099,
+      "step": 28
+    },
+    {
+      "epoch": 0.024181780279341255,
+      "grad_norm": 0.3149116337299347,
+      "learning_rate": 5.8e-07,
+      "loss": 0.8,
+      "step": 29
+    },
+    {
+      "epoch": 0.025015634771732333,
+      "grad_norm": 0.29307302832603455,
+      "learning_rate": 6e-07,
+      "loss": 0.8583,
+      "step": 30
+    },
+    {
+      "epoch": 0.02584948926412341,
+      "grad_norm": 0.3131772577762604,
+      "learning_rate": 6.2e-07,
+      "loss": 0.8432,
+      "step": 31
+    },
+    {
+      "epoch": 0.026683343756514488,
+      "grad_norm": 0.28058937191963196,
+      "learning_rate": 6.4e-07,
+      "loss": 0.8099,
+      "step": 32
+    },
+    {
+      "epoch": 0.027517198248905566,
+      "grad_norm": 0.2649301290512085,
+      "learning_rate": 6.6e-07,
+      "loss": 0.8715,
+      "step": 33
+    },
+    {
+      "epoch": 0.028351052741296644,
+      "grad_norm": 0.2659442126750946,
+      "learning_rate": 6.800000000000001e-07,
+      "loss": 0.7432,
+      "step": 34
+    },
+    {
+      "epoch": 0.02918490723368772,
+      "grad_norm": 0.24875527620315552,
+      "learning_rate": 7e-07,
+      "loss": 0.8502,
+      "step": 35
+    },
+    {
+      "epoch": 0.0300187617260788,
+      "grad_norm": 0.21730971336364746,
+      "learning_rate": 7.2e-07,
+      "loss": 0.813,
+      "step": 36
+    },
+    {
+      "epoch": 0.030852616218469877,
+      "grad_norm": 0.19625824689865112,
+      "learning_rate": 7.4e-07,
+      "loss": 0.7839,
+      "step": 37
+    },
+    {
+      "epoch": 0.031686470710860955,
+      "grad_norm": 0.19287347793579102,
+      "learning_rate": 7.599999999999999e-07,
+      "loss": 0.7949,
+      "step": 38
+    },
+    {
+      "epoch": 0.032520325203252036,
+      "grad_norm": 0.1588095873594284,
+      "learning_rate": 7.799999999999999e-07,
+      "loss": 0.7137,
+      "step": 39
+    },
+    {
+      "epoch": 0.03335417969564311,
+      "grad_norm": 0.17520244419574738,
+      "learning_rate": 8e-07,
+      "loss": 0.8112,
+      "step": 40
+    },
+    {
+      "epoch": 0.03418803418803419,
+      "grad_norm": 0.16394659876823425,
+      "learning_rate": 8.199999999999999e-07,
+      "loss": 0.8096,
+      "step": 41
+    },
+    {
+      "epoch": 0.035021888680425266,
+      "grad_norm": 0.1522587239742279,
+      "learning_rate": 8.399999999999999e-07,
+      "loss": 0.708,
+      "step": 42
+    },
+    {
+      "epoch": 0.03585574317281635,
+      "grad_norm": 0.16340652108192444,
+      "learning_rate": 8.599999999999999e-07,
+      "loss": 0.7816,
+      "step": 43
+    },
+    {
+      "epoch": 0.03668959766520742,
+      "grad_norm": 0.15452717244625092,
+      "learning_rate": 8.799999999999999e-07,
+      "loss": 0.7946,
+      "step": 44
+    },
+    {
+      "epoch": 0.0375234521575985,
+      "grad_norm": 0.1545872688293457,
+      "learning_rate": 9e-07,
+      "loss": 0.7908,
+      "step": 45
+    },
+    {
+      "epoch": 0.03835730664998958,
+      "grad_norm": 0.13799017667770386,
+      "learning_rate": 9.2e-07,
+      "loss": 0.7056,
+      "step": 46
+    },
+    {
+      "epoch": 0.03919116114238066,
+      "grad_norm": 0.14599646627902985,
+      "learning_rate": 9.399999999999999e-07,
+      "loss": 0.8098,
+      "step": 47
+    },
+    {
+      "epoch": 0.04002501563477173,
+      "grad_norm": 0.13951805233955383,
+      "learning_rate": 9.6e-07,
+      "loss": 0.7938,
+      "step": 48
+    },
+    {
+      "epoch": 0.040858870127162814,
+      "grad_norm": 0.14058919250965118,
+      "learning_rate": 9.8e-07,
+      "loss": 0.7806,
+      "step": 49
+    },
+    {
+      "epoch": 0.04169272461955389,
+      "grad_norm": 0.14300380647182465,
+      "learning_rate": 1e-06,
+      "loss": 0.8192,
+      "step": 50
+    },
+    {
+      "epoch": 0.04252657911194497,
+      "grad_norm": 0.13719584047794342,
+      "learning_rate": 1.02e-06,
+      "loss": 0.7939,
+      "step": 51
+    },
+    {
+      "epoch": 0.04336043360433604,
+      "grad_norm": 0.14002980291843414,
+      "learning_rate": 1.04e-06,
+      "loss": 0.8163,
+      "step": 52
+    },
+    {
+      "epoch": 0.044194288096727125,
+      "grad_norm": 0.13815677165985107,
+      "learning_rate": 1.06e-06,
+      "loss": 0.8137,
+      "step": 53
+    },
+    {
+      "epoch": 0.0450281425891182,
+      "grad_norm": 0.13772878050804138,
+      "learning_rate": 1.08e-06,
+      "loss": 0.771,
+      "step": 54
+    },
+    {
+      "epoch": 0.04586199708150928,
+      "grad_norm": 0.1414576917886734,
+      "learning_rate": 1.1e-06,
+      "loss": 0.7989,
+      "step": 55
+    },
+    {
+      "epoch": 0.046695851573900354,
+      "grad_norm": 0.11953306943178177,
+      "learning_rate": 1.12e-06,
+      "loss": 0.7074,
+      "step": 56
+    },
+    {
+      "epoch": 0.047529706066291436,
+      "grad_norm": 0.13051201403141022,
+      "learning_rate": 1.1399999999999999e-06,
+      "loss": 0.8034,
+      "step": 57
+    },
+    {
+      "epoch": 0.04836356055868251,
+      "grad_norm": 0.12802985310554504,
+      "learning_rate": 1.16e-06,
+      "loss": 0.7581,
+      "step": 58
+    },
+    {
+      "epoch": 0.04919741505107359,
+      "grad_norm": 0.13593435287475586,
+      "learning_rate": 1.18e-06,
+      "loss": 0.7978,
+      "step": 59
+    },
+    {
+      "epoch": 0.050031269543464665,
+      "grad_norm": 0.12595658004283905,
+      "learning_rate": 1.2e-06,
+      "loss": 0.7997,
+      "step": 60
+    },
+    {
+      "epoch": 0.05086512403585575,
+      "grad_norm": 0.1257220059633255,
+      "learning_rate": 1.22e-06,
+      "loss": 0.7194,
+      "step": 61
+    },
+    {
+      "epoch": 0.05169897852824682,
+      "grad_norm": 0.12169249355792999,
+      "learning_rate": 1.24e-06,
+      "loss": 0.7412,
+      "step": 62
+    },
+    {
+      "epoch": 0.0525328330206379,
+      "grad_norm": 0.12391054630279541,
+      "learning_rate": 1.26e-06,
+      "loss": 0.7849,
+      "step": 63
+    },
+    {
+      "epoch": 0.053366687513028976,
+      "grad_norm": 0.12134575098752975,
+      "learning_rate": 1.28e-06,
+      "loss": 0.6859,
+      "step": 64
+    },
+    {
+      "epoch": 0.05420054200542006,
+      "grad_norm": 0.11605742573738098,
+      "learning_rate": 1.3e-06,
+      "loss": 0.8112,
+      "step": 65
+    },
+    {
+      "epoch": 0.05503439649781113,
+      "grad_norm": 0.11593819409608841,
+      "learning_rate": 1.32e-06,
+      "loss": 0.6993,
+      "step": 66
+    },
+    {
+      "epoch": 0.05586825099020221,
+      "grad_norm": 0.12103109061717987,
+      "learning_rate": 1.34e-06,
+      "loss": 0.7595,
+      "step": 67
+    },
+    {
+      "epoch": 0.05670210548259329,
+      "grad_norm": 0.12391602247953415,
+      "learning_rate": 1.3600000000000001e-06,
+      "loss": 0.7676,
+      "step": 68
+    },
+    {
+      "epoch": 0.05753595997498437,
+      "grad_norm": 0.10895267128944397,
+      "learning_rate": 1.38e-06,
+      "loss": 0.7256,
+      "step": 69
+    },
+    {
+      "epoch": 0.05836981446737544,
+      "grad_norm": 0.1129627376794815,
+      "learning_rate": 1.4e-06,
+      "loss": 0.7097,
+      "step": 70
+    },
+    {
+      "epoch": 0.059203668959766524,
+      "grad_norm": 0.1078682616353035,
+      "learning_rate": 1.42e-06,
+      "loss": 0.7052,
+      "step": 71
+    },
+    {
+      "epoch": 0.0600375234521576,
+      "grad_norm": 0.1140650063753128,
+      "learning_rate": 1.44e-06,
+      "loss": 0.8067,
+      "step": 72
+    },
+    {
+      "epoch": 0.06087137794454868,
+      "grad_norm": 0.10528790205717087,
+      "learning_rate": 1.46e-06,
+      "loss": 0.7241,
+      "step": 73
+    },
+    {
+      "epoch": 0.061705232436939754,
+      "grad_norm": 0.11815836280584335,
+      "learning_rate": 1.48e-06,
+      "loss": 0.7344,
+      "step": 74
+    },
+    {
+      "epoch": 0.06253908692933083,
+      "grad_norm": 0.11140462756156921,
+      "learning_rate": 1.5e-06,
+      "loss": 0.7664,
+      "step": 75
+    },
+    {
+      "epoch": 0.06337294142172191,
+      "grad_norm": 0.11665944010019302,
+      "learning_rate": 1.5199999999999998e-06,
+      "loss": 0.7178,
+      "step": 76
+    },
+    {
+      "epoch": 0.06420679591411299,
+      "grad_norm": 0.10751160234212875,
+      "learning_rate": 1.5399999999999999e-06,
+      "loss": 0.7269,
+      "step": 77
+    },
+    {
+      "epoch": 0.06504065040650407,
+      "grad_norm": 0.09768356382846832,
+      "learning_rate": 1.5599999999999999e-06,
+      "loss": 0.7449,
+      "step": 78
+    },
+    {
+      "epoch": 0.06587450489889514,
+      "grad_norm": 0.09686972200870514,
+      "learning_rate": 1.58e-06,
+      "loss": 0.7097,
+      "step": 79
+    },
+    {
+      "epoch": 0.06670835939128622,
+      "grad_norm": 0.09527275711297989,
+      "learning_rate": 1.6e-06,
+      "loss": 0.7301,
+      "step": 80
+    },
+    {
+      "epoch": 0.0675422138836773,
+      "grad_norm": 0.0931195393204689,
+      "learning_rate": 1.62e-06,
+      "loss": 0.7524,
+      "step": 81
+    },
+    {
+      "epoch": 0.06837606837606838,
+      "grad_norm": 0.0892946794629097,
+      "learning_rate": 1.6399999999999998e-06,
+      "loss": 0.7156,
+      "step": 82
+    },
+    {
+      "epoch": 0.06920992286845945,
+      "grad_norm": 0.08845791220664978,
+      "learning_rate": 1.6599999999999998e-06,
+      "loss": 0.7349,
+      "step": 83
+    },
+    {
+      "epoch": 0.07004377736085053,
+      "grad_norm": 0.0896732360124588,
+      "learning_rate": 1.6799999999999998e-06,
+      "loss": 0.7145,
+      "step": 84
+    },
+    {
+      "epoch": 0.07087763185324161,
+      "grad_norm": 0.09997335076332092,
+      "learning_rate": 1.6999999999999998e-06,
+      "loss": 0.7103,
+      "step": 85
+    },
+    {
+      "epoch": 0.0717114863456327,
+      "grad_norm": 0.08821084350347519,
+      "learning_rate": 1.7199999999999998e-06,
+      "loss": 0.725,
+      "step": 86
+    },
+    {
+      "epoch": 0.07254534083802376,
+      "grad_norm": 0.08937980234622955,
+      "learning_rate": 1.7399999999999999e-06,
+      "loss": 0.7194,
+      "step": 87
+    },
+    {
+      "epoch": 0.07337919533041484,
+      "grad_norm": 0.09307857602834702,
+      "learning_rate": 1.7599999999999999e-06,
+      "loss": 0.7595,
+      "step": 88
+    },
+    {
+      "epoch": 0.07421304982280592,
+      "grad_norm": 0.08865829557180405,
+      "learning_rate": 1.78e-06,
+      "loss": 0.7464,
+      "step": 89
+    },
+    {
+      "epoch": 0.075046904315197,
+      "grad_norm": 0.09038495272397995,
+      "learning_rate": 1.8e-06,
+      "loss": 0.6698,
+      "step": 90
+    },
+    {
+      "epoch": 0.07588075880758807,
+      "grad_norm": 0.08254078030586243,
+      "learning_rate": 1.82e-06,
+      "loss": 0.7426,
+      "step": 91
+    },
+    {
+      "epoch": 0.07671461329997915,
+      "grad_norm": 0.08807505667209625,
+      "learning_rate": 1.84e-06,
+      "loss": 0.6789,
+      "step": 92
+    },
+    {
+      "epoch": 0.07754846779237023,
+      "grad_norm": 0.09523889422416687,
+      "learning_rate": 1.86e-06,
+      "loss": 0.7218,
+      "step": 93
+    },
+    {
+      "epoch": 0.07838232228476132,
+      "grad_norm": 0.08782277256250381,
+      "learning_rate": 1.8799999999999998e-06,
+      "loss": 0.7599,
+      "step": 94
+    },
+    {
+      "epoch": 0.07921617677715238,
+      "grad_norm": 0.08968065679073334,
+      "learning_rate": 1.8999999999999998e-06,
+      "loss": 0.7971,
+      "step": 95
+    },
+    {
+      "epoch": 0.08005003126954346,
+      "grad_norm": 0.08655782043933868,
+      "learning_rate": 1.92e-06,
+      "loss": 0.7479,
+      "step": 96
+    },
+    {
+      "epoch": 0.08088388576193455,
+      "grad_norm": 0.0854155421257019,
+      "learning_rate": 1.94e-06,
+      "loss": 0.7327,
+      "step": 97
+    },
+    {
+      "epoch": 0.08171774025432563,
+      "grad_norm": 0.08121508359909058,
+      "learning_rate": 1.96e-06,
+      "loss": 0.6749,
+      "step": 98
+    },
+    {
+      "epoch": 0.0825515947467167,
+      "grad_norm": 0.08429264277219772,
+      "learning_rate": 1.98e-06,
+      "loss": 0.6402,
+      "step": 99
+    },
+    {
+      "epoch": 0.08338544923910778,
+      "grad_norm": 0.08390086144208908,
+      "learning_rate": 2e-06,
+      "loss": 0.7434,
+      "step": 100
+    },
+    {
+      "epoch": 0.08421930373149886,
+      "grad_norm": 0.08662284910678864,
+      "learning_rate": 1.9999995964675577e-06,
+      "loss": 0.6681,
+      "step": 101
+    },
+    {
+      "epoch": 0.08505315822388994,
+      "grad_norm": 0.08700251579284668,
+      "learning_rate": 1.9999983858705566e-06,
+      "loss": 0.763,
+      "step": 102
+    },
+    {
+      "epoch": 0.085887012716281,
+      "grad_norm": 0.08504343777894974,
+      "learning_rate": 1.9999963682099734e-06,
+      "loss": 0.77,
+      "step": 103
+    },
+    {
+      "epoch": 0.08672086720867209,
+      "grad_norm": 0.08222745358943939,
+      "learning_rate": 1.999993543487437e-06,
+      "loss": 0.7103,
+      "step": 104
+    },
+    {
+      "epoch": 0.08755472170106317,
+      "grad_norm": 0.08124719560146332,
+      "learning_rate": 1.9999899117052263e-06,
+      "loss": 0.6913,
+      "step": 105
+    },
+    {
+      "epoch": 0.08838857619345425,
+      "grad_norm": 0.09033836424350739,
+      "learning_rate": 1.9999854728662734e-06,
+      "loss": 0.7065,
+      "step": 106
+    },
+    {
+      "epoch": 0.08922243068584532,
+      "grad_norm": 0.09045730531215668,
+      "learning_rate": 1.99998022697416e-06,
+      "loss": 0.7624,
+      "step": 107
+    },
+    {
+      "epoch": 0.0900562851782364,
+      "grad_norm": 0.0832991823554039,
+      "learning_rate": 1.9999741740331203e-06,
+      "loss": 0.7399,
+      "step": 108
+    },
+    {
+      "epoch": 0.09089013967062748,
+      "grad_norm": 0.07666011154651642,
+      "learning_rate": 1.9999673140480388e-06,
+      "loss": 0.6477,
+      "step": 109
+    },
+    {
+      "epoch": 0.09172399416301856,
+      "grad_norm": 0.08195007592439651,
+      "learning_rate": 1.9999596470244527e-06,
+      "loss": 0.7068,
+      "step": 110
+    },
+    {
+      "epoch": 0.09255784865540963,
+      "grad_norm": 0.07755164802074432,
+      "learning_rate": 1.99995117296855e-06,
+      "loss": 0.7021,
+      "step": 111
+    },
+    {
+      "epoch": 0.09339170314780071,
+      "grad_norm": 0.0782230794429779,
+      "learning_rate": 1.9999418918871685e-06,
+      "loss": 0.7039,
+      "step": 112
+    },
+    {
+      "epoch": 0.09422555764019179,
+      "grad_norm": 0.08311248570680618,
+      "learning_rate": 1.9999318037877995e-06,
+      "loss": 0.7238,
+      "step": 113
+    },
+    {
+      "epoch": 0.09505941213258287,
+      "grad_norm": 0.08455490320920944,
+      "learning_rate": 1.999920908678585e-06,
+      "loss": 0.7184,
+      "step": 114
+    },
+    {
+      "epoch": 0.09589326662497394,
+      "grad_norm": 0.08463476598262787,
+      "learning_rate": 1.9999092065683174e-06,
+      "loss": 0.6162,
+      "step": 115
+    },
+    {
+      "epoch": 0.09672712111736502,
+      "grad_norm": 0.07902689278125763,
+      "learning_rate": 1.999896697466442e-06,
+      "loss": 0.6649,
+      "step": 116
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 0.08383440971374512,
+      "learning_rate": 1.999883381383053e-06,
+      "loss": 0.703,
+      "step": 117
+    },
+    {
+      "epoch": 0.09839483010214718,
+      "grad_norm": 0.07288578897714615,
+      "learning_rate": 1.999869258328899e-06,
+      "loss": 0.7286,
+      "step": 118
+    },
+    {
+      "epoch": 0.09922868459453825,
+      "grad_norm": 0.08177065849304199,
+      "learning_rate": 1.999854328315377e-06,
+      "loss": 0.6899,
+      "step": 119
+    },
+    {
+      "epoch": 0.10006253908692933,
+      "grad_norm": 0.07529062032699585,
+      "learning_rate": 1.999838591354537e-06,
+      "loss": 0.6781,
+      "step": 120
+    },
+    {
+      "epoch": 0.10089639357932041,
+      "grad_norm": 0.08230678737163544,
+      "learning_rate": 1.9998220474590795e-06,
+      "loss": 0.7059,
+      "step": 121
+    },
+    {
+      "epoch": 0.1017302480717115,
+      "grad_norm": 0.08000528067350388,
+      "learning_rate": 1.9998046966423567e-06,
+      "loss": 0.6561,
+      "step": 122
+    },
+    {
+      "epoch": 0.10256410256410256,
+      "grad_norm": 0.07669328153133392,
+      "learning_rate": 1.9997865389183717e-06,
+      "loss": 0.6452,
+      "step": 123
+    },
+    {
+      "epoch": 0.10339795705649364,
+      "grad_norm": 0.07872146368026733,
+      "learning_rate": 1.9997675743017794e-06,
+      "loss": 0.7337,
+      "step": 124
+    },
+    {
+      "epoch": 0.10423181154888472,
+      "grad_norm": 0.07842138409614563,
+      "learning_rate": 1.9997478028078853e-06,
+      "loss": 0.7126,
+      "step": 125
+    },
+    {
+      "epoch": 0.1050656660412758,
+      "grad_norm": 0.07890679687261581,
+      "learning_rate": 1.9997272244526453e-06,
+      "loss": 0.7049,
+      "step": 126
+    },
+    {
+      "epoch": 0.10589952053366687,
+      "grad_norm": 0.07983887195587158,
+      "learning_rate": 1.999705839252669e-06,
+      "loss": 0.7578,
+      "step": 127
+    },
+    {
+      "epoch": 0.10673337502605795,
+      "grad_norm": 0.07721222192049026,
+      "learning_rate": 1.9996836472252144e-06,
+      "loss": 0.678,
+      "step": 128
+    },
+    {
+      "epoch": 0.10756722951844903,
+      "grad_norm": 0.07825150340795517,
+      "learning_rate": 1.9996606483881927e-06,
+      "loss": 0.6763,
+      "step": 129
+    },
+    {
+      "epoch": 0.10840108401084012,
+      "grad_norm": 0.08063995838165283,
+      "learning_rate": 1.999636842760165e-06,
+      "loss": 0.758,
+      "step": 130
+    },
+    {
+      "epoch": 0.10923493850323118,
+      "grad_norm": 0.0724928081035614,
+      "learning_rate": 1.9996122303603443e-06,
+      "loss": 0.7312,
+      "step": 131
+    },
+    {
+      "epoch": 0.11006879299562226,
+      "grad_norm": 0.08180397748947144,
+      "learning_rate": 1.999586811208594e-06,
+      "loss": 0.6794,
+      "step": 132
+    },
+    {
+      "epoch": 0.11090264748801335,
+      "grad_norm": 0.0856177881360054,
+      "learning_rate": 1.99956058532543e-06,
+      "loss": 0.6885,
+      "step": 133
+    },
+    {
+      "epoch": 0.11173650198040443,
+      "grad_norm": 0.0763324424624443,
+      "learning_rate": 1.999533552732017e-06,
+      "loss": 0.6821,
+      "step": 134
+    },
+    {
+      "epoch": 0.1125703564727955,
+      "grad_norm": 0.07964632660150528,
+      "learning_rate": 1.9995057134501725e-06,
+      "loss": 0.7362,
+      "step": 135
+    },
+    {
+      "epoch": 0.11340421096518657,
+      "grad_norm": 0.07633720338344574,
+      "learning_rate": 1.999477067502365e-06,
+      "loss": 0.7396,
+      "step": 136
+    },
+    {
+      "epoch": 0.11423806545757766,
+      "grad_norm": 0.07655656337738037,
+      "learning_rate": 1.9994476149117132e-06,
+      "loss": 0.728,
+      "step": 137
+    },
+    {
+      "epoch": 0.11507191994996874,
+      "grad_norm": 0.07770884782075882,
+      "learning_rate": 1.9994173557019877e-06,
+      "loss": 0.6878,
+      "step": 138
+    },
+    {
+      "epoch": 0.1159057744423598,
+      "grad_norm": 0.07332731038331985,
+      "learning_rate": 1.999386289897609e-06,
+      "loss": 0.7115,
+      "step": 139
+    },
+    {
+      "epoch": 0.11673962893475089,
+      "grad_norm": 0.0784095972776413,
+      "learning_rate": 1.9993544175236497e-06,
+      "loss": 0.6653,
+      "step": 140
+    },
+    {
+      "epoch": 0.11757348342714197,
+      "grad_norm": 0.08311771601438522,
+      "learning_rate": 1.9993217386058326e-06,
+      "loss": 0.6504,
+      "step": 141
+    },
+    {
+      "epoch": 0.11840733791953305,
+      "grad_norm": 0.07529395818710327,
+      "learning_rate": 1.999288253170532e-06,
+      "loss": 0.7083,
+      "step": 142
+    },
+    {
+      "epoch": 0.11924119241192412,
+      "grad_norm": 0.07608146965503693,
+      "learning_rate": 1.999253961244773e-06,
+      "loss": 0.6856,
+      "step": 143
+    },
+    {
+      "epoch": 0.1200750469043152,
+      "grad_norm": 0.07125360518693924,
+      "learning_rate": 1.9992188628562303e-06,
+      "loss": 0.6095,
+      "step": 144
+    },
+    {
+      "epoch": 0.12090890139670628,
+      "grad_norm": 0.07715660333633423,
+      "learning_rate": 1.999182958033232e-06,
+      "loss": 0.7337,
+      "step": 145
+    },
+    {
+      "epoch": 0.12174275588909736,
+      "grad_norm": 0.07756403833627701,
+      "learning_rate": 1.999146246804755e-06,
+      "loss": 0.6929,
+      "step": 146
+    },
+    {
+      "epoch": 0.12257661038148843,
+      "grad_norm": 0.07667449861764908,
+      "learning_rate": 1.9991087292004273e-06,
+      "loss": 0.67,
+      "step": 147
+    },
+    {
+      "epoch": 0.12341046487387951,
+      "grad_norm": 0.08285968005657196,
+      "learning_rate": 1.9990704052505284e-06,
+      "loss": 0.7483,
+      "step": 148
+    },
+    {
+      "epoch": 0.12424431936627059,
+      "grad_norm": 0.07476554065942764,
+      "learning_rate": 1.9990312749859887e-06,
+      "loss": 0.6846,
+      "step": 149
+    },
+    {
+      "epoch": 0.12507817385866166,
+      "grad_norm": 0.0818118005990982,
+      "learning_rate": 1.998991338438388e-06,
+      "loss": 0.6709,
+      "step": 150
+    },
+    {
+      "epoch": 0.12591202835105275,
+      "grad_norm": 0.08350253850221634,
+      "learning_rate": 1.998950595639958e-06,
+      "loss": 0.7379,
+      "step": 151
+    },
+    {
+      "epoch": 0.12674588284344382,
+      "grad_norm": 0.07916589826345444,
+      "learning_rate": 1.9989090466235806e-06,
+      "loss": 0.7206,
+      "step": 152
+    },
+    {
+      "epoch": 0.1275797373358349,
+      "grad_norm": 0.07860454171895981,
+      "learning_rate": 1.998866691422789e-06,
+      "loss": 0.6777,
+      "step": 153
+    },
+    {
+      "epoch": 0.12841359182822598,
+      "grad_norm": 0.07861131429672241,
+      "learning_rate": 1.998823530071766e-06,
+      "loss": 0.6938,
+      "step": 154
+    },
+    {
+      "epoch": 0.12924744632061705,
+      "grad_norm": 0.07819496840238571,
+      "learning_rate": 1.9987795626053465e-06,
+      "loss": 0.7193,
+      "step": 155
+    },
+    {
+      "epoch": 0.13008130081300814,
+      "grad_norm": 0.07828415185213089,
+      "learning_rate": 1.9987347890590144e-06,
+      "loss": 0.6809,
+      "step": 156
+    },
+    {
+      "epoch": 0.1309151553053992,
+      "grad_norm": 0.07755222916603088,
+      "learning_rate": 1.998689209468905e-06,
+      "loss": 0.7213,
+      "step": 157
+    },
+    {
+      "epoch": 0.13174900979779028,
+      "grad_norm": 0.07672612369060516,
+      "learning_rate": 1.998642823871804e-06,
+      "loss": 0.7196,
+      "step": 158
+    },
+    {
+      "epoch": 0.13258286429018137,
+      "grad_norm": 0.07963719964027405,
+      "learning_rate": 1.9985956323051475e-06,
+      "loss": 0.7465,
+      "step": 159
+    },
+    {
+      "epoch": 0.13341671878257244,
+      "grad_norm": 0.0795946940779686,
+      "learning_rate": 1.9985476348070223e-06,
+      "loss": 0.717,
+      "step": 160
+    },
+    {
+      "epoch": 0.1342505732749635,
+      "grad_norm": 0.07344063371419907,
+      "learning_rate": 1.9984988314161656e-06,
+      "loss": 0.6861,
+      "step": 161
+    },
+    {
+      "epoch": 0.1350844277673546,
+      "grad_norm": 0.07722171396017075,
+      "learning_rate": 1.9984492221719645e-06,
+      "loss": 0.6825,
+      "step": 162
+    },
+    {
+      "epoch": 0.13591828225974567,
+      "grad_norm": 0.0812341496348381,
+      "learning_rate": 1.9983988071144572e-06,
+      "loss": 0.7245,
+      "step": 163
+    },
+    {
+      "epoch": 0.13675213675213677,
+      "grad_norm": 0.08472032845020294,
+      "learning_rate": 1.998347586284332e-06,
+      "loss": 0.7315,
+      "step": 164
+    },
+    {
+      "epoch": 0.13758599124452783,
+      "grad_norm": 0.07785464823246002,
+      "learning_rate": 1.9982955597229275e-06,
+      "loss": 0.6906,
+      "step": 165
+    },
+    {
+      "epoch": 0.1384198457369189,
+      "grad_norm": 0.07382892072200775,
+      "learning_rate": 1.998242727472232e-06,
+      "loss": 0.6488,
+      "step": 166
+    },
+    {
+      "epoch": 0.13925370022931,
+      "grad_norm": 0.07445723563432693,
+      "learning_rate": 1.9981890895748845e-06,
+      "loss": 0.692,
+      "step": 167
+    },
+    {
+      "epoch": 0.14008755472170106,
+      "grad_norm": 0.07889281213283539,
+      "learning_rate": 1.998134646074175e-06,
+      "loss": 0.6438,
+      "step": 168
+    },
+    {
+      "epoch": 0.14092140921409213,
+      "grad_norm": 0.07664698362350464,
+      "learning_rate": 1.9980793970140426e-06,
+      "loss": 0.6993,
+      "step": 169
+    },
+    {
+      "epoch": 0.14175526370648323,
+      "grad_norm": 0.07807335257530212,
+      "learning_rate": 1.9980233424390773e-06,
+      "loss": 0.7334,
+      "step": 170
+    },
+    {
+      "epoch": 0.1425891181988743,
+      "grad_norm": 0.0683249682188034,
+      "learning_rate": 1.9979664823945174e-06,
+      "loss": 0.6325,
+      "step": 171
+    },
+    {
+      "epoch": 0.1434229726912654,
+      "grad_norm": 0.08339341729879379,
+      "learning_rate": 1.997908816926254e-06,
+      "loss": 0.697,
+      "step": 172
+    },
+    {
+      "epoch": 0.14425682718365646,
+      "grad_norm": 0.07781082391738892,
+      "learning_rate": 1.997850346080827e-06,
+      "loss": 0.6748,
+      "step": 173
+    },
+    {
+      "epoch": 0.14509068167604752,
+      "grad_norm": 0.07718750089406967,
+      "learning_rate": 1.9977910699054247e-06,
+      "loss": 0.7002,
+      "step": 174
+    },
+    {
+      "epoch": 0.14592453616843862,
+      "grad_norm": 0.07779830694198608,
+      "learning_rate": 1.997730988447888e-06,
+      "loss": 0.723,
+      "step": 175
+    },
+    {
+      "epoch": 0.14675839066082969,
+      "grad_norm": 0.07958701252937317,
+      "learning_rate": 1.997670101756706e-06,
+      "loss": 0.6651,
+      "step": 176
+    },
+    {
+      "epoch": 0.14759224515322075,
+      "grad_norm": 0.08142640441656113,
+      "learning_rate": 1.997608409881019e-06,
+      "loss": 0.6652,
+      "step": 177
+    },
+    {
+      "epoch": 0.14842609964561185,
+      "grad_norm": 0.07898704707622528,
+      "learning_rate": 1.9975459128706152e-06,
+      "loss": 0.6842,
+      "step": 178
+    },
+    {
+      "epoch": 0.14925995413800291,
+      "grad_norm": 0.07870691269636154,
+      "learning_rate": 1.997482610775935e-06,
+      "loss": 0.7167,
+      "step": 179
+    },
+    {
+      "epoch": 0.150093808630394,
+      "grad_norm": 0.07766727358102798,
+      "learning_rate": 1.9974185036480658e-06,
+      "loss": 0.6145,
+      "step": 180
+    },
+    {
+      "epoch": 0.15092766312278508,
+      "grad_norm": 0.07913653552532196,
+      "learning_rate": 1.997353591538748e-06,
+      "loss": 0.7598,
+      "step": 181
+    },
+    {
+      "epoch": 0.15176151761517614,
+      "grad_norm": 0.07850881665945053,
+      "learning_rate": 1.9972878745003684e-06,
+      "loss": 0.7492,
+      "step": 182
+    },
+    {
+      "epoch": 0.15259537210756724,
+      "grad_norm": 0.07599500566720963,
+      "learning_rate": 1.9972213525859656e-06,
+      "loss": 0.6802,
+      "step": 183
+    },
+    {
+      "epoch": 0.1534292265999583,
+      "grad_norm": 0.08165914565324783,
+      "learning_rate": 1.997154025849227e-06,
+      "loss": 0.6343,
+      "step": 184
+    },
+    {
+      "epoch": 0.15426308109234937,
+      "grad_norm": 0.0763566642999649,
+      "learning_rate": 1.9970858943444893e-06,
+      "loss": 0.7214,
+      "step": 185
+    },
+    {
+      "epoch": 0.15509693558474047,
+      "grad_norm": 0.07993835210800171,
+      "learning_rate": 1.99701695812674e-06,
+      "loss": 0.6975,
+      "step": 186
+    },
+    {
+      "epoch": 0.15593079007713154,
+      "grad_norm": 0.07139826565980911,
+      "learning_rate": 1.996947217251614e-06,
+      "loss": 0.6714,
+      "step": 187
+    },
+    {
+      "epoch": 0.15676464456952263,
+      "grad_norm": 0.0777943804860115,
+      "learning_rate": 1.996876671775397e-06,
+      "loss": 0.6684,
+      "step": 188
+    },
+    {
+      "epoch": 0.1575984990619137,
+      "grad_norm": 0.07822203636169434,
+      "learning_rate": 1.9968053217550242e-06,
+      "loss": 0.7161,
+      "step": 189
+    },
+    {
+      "epoch": 0.15843235355430477,
+      "grad_norm": 0.07645302265882492,
+      "learning_rate": 1.9967331672480794e-06,
+      "loss": 0.647,
+      "step": 190
+    },
+    {
+      "epoch": 0.15926620804669586,
+      "grad_norm": 0.07744666188955307,
+      "learning_rate": 1.996660208312796e-06,
+      "loss": 0.6887,
+      "step": 191
+    },
+    {
+      "epoch": 0.16010006253908693,
+      "grad_norm": 0.07616739720106125,
+      "learning_rate": 1.9965864450080565e-06,
+      "loss": 0.7209,
+      "step": 192
+    },
+    {
+      "epoch": 0.160933917031478,
+      "grad_norm": 0.0808212161064148,
+      "learning_rate": 1.996511877393393e-06,
+      "loss": 0.7229,
+      "step": 193
+    },
+    {
+      "epoch": 0.1617677715238691,
+      "grad_norm": 0.07879424095153809,
+      "learning_rate": 1.996436505528986e-06,
+      "loss": 0.6753,
+      "step": 194
+    },
+    {
+      "epoch": 0.16260162601626016,
+      "grad_norm": 0.07811419665813446,
+      "learning_rate": 1.9963603294756654e-06,
+      "loss": 0.6427,
+      "step": 195
+    },
+    {
+      "epoch": 0.16343548050865125,
+      "grad_norm": 0.08347994834184647,
+      "learning_rate": 1.996283349294911e-06,
+      "loss": 0.6575,
+      "step": 196
+    },
+    {
+      "epoch": 0.16426933500104232,
+      "grad_norm": 0.08415097743272781,
+      "learning_rate": 1.99620556504885e-06,
+      "loss": 0.6859,
+      "step": 197
+    },
+    {
+      "epoch": 0.1651031894934334,
+      "grad_norm": 0.08112502098083496,
+      "learning_rate": 1.9961269768002596e-06,
+      "loss": 0.6601,
+      "step": 198
+    },
+    {
+      "epoch": 0.16593704398582448,
+      "grad_norm": 0.07908611744642258,
+      "learning_rate": 1.9960475846125656e-06,
+      "loss": 0.6697,
+      "step": 199
+    },
+    {
+      "epoch": 0.16677089847821555,
+      "grad_norm": 0.07735829800367355,
+      "learning_rate": 1.995967388549843e-06,
+      "loss": 0.6631,
+      "step": 200
+    },
+    {
+      "epoch": 0.16760475297060662,
+      "grad_norm": 0.07564505934715271,
+      "learning_rate": 1.9958863886768145e-06,
+      "loss": 0.7004,
+      "step": 201
+    },
+    {
+      "epoch": 0.1684386074629977,
+      "grad_norm": 0.08117896318435669,
+      "learning_rate": 1.9958045850588527e-06,
+      "loss": 0.62,
+      "step": 202
+    },
+    {
+      "epoch": 0.16927246195538878,
+      "grad_norm": 0.0761483758687973,
+      "learning_rate": 1.9957219777619784e-06,
+      "loss": 0.7145,
+      "step": 203
+    },
+    {
+      "epoch": 0.17010631644777988,
+      "grad_norm": 0.08249073475599289,
+      "learning_rate": 1.995638566852861e-06,
+      "loss": 0.703,
+      "step": 204
+    },
+    {
+      "epoch": 0.17094017094017094,
+      "grad_norm": 0.08197219669818878,
+      "learning_rate": 1.9955543523988187e-06,
+      "loss": 0.695,
+      "step": 205
+    },
+    {
+      "epoch": 0.171774025432562,
+      "grad_norm": 0.07933253049850464,
+      "learning_rate": 1.995469334467818e-06,
+      "loss": 0.6575,
+      "step": 206
+    },
+    {
+      "epoch": 0.1726078799249531,
+      "grad_norm": 0.07888033986091614,
+      "learning_rate": 1.9953835131284735e-06,
+      "loss": 0.6504,
+      "step": 207
+    },
+    {
+      "epoch": 0.17344173441734417,
+      "grad_norm": 0.080783411860466,
+      "learning_rate": 1.995296888450049e-06,
+      "loss": 0.6586,
+      "step": 208
+    },
+    {
+      "epoch": 0.17427558890973524,
+      "grad_norm": 0.07863239198923111,
+      "learning_rate": 1.995209460502456e-06,
+      "loss": 0.7204,
+      "step": 209
+    },
+    {
+      "epoch": 0.17510944340212634,
+      "grad_norm": 0.07864314317703247,
+      "learning_rate": 1.9951212293562544e-06,
+      "loss": 0.6867,
+      "step": 210
+    },
+    {
+      "epoch": 0.1759432978945174,
+      "grad_norm": 0.07574496418237686,
+      "learning_rate": 1.995032195082653e-06,
+      "loss": 0.6806,
+      "step": 211
+    },
+    {
+      "epoch": 0.1767771523869085,
+      "grad_norm": 0.08079942315816879,
+      "learning_rate": 1.9949423577535078e-06,
+      "loss": 0.6929,
+      "step": 212
+    },
+    {
+      "epoch": 0.17761100687929957,
+      "grad_norm": 0.08361431211233139,
+      "learning_rate": 1.9948517174413235e-06,
+      "loss": 0.6948,
+      "step": 213
+    },
+    {
+      "epoch": 0.17844486137169063,
+      "grad_norm": 0.07974692434072495,
+      "learning_rate": 1.994760274219253e-06,
+      "loss": 0.7269,
+      "step": 214
+    },
+    {
+      "epoch": 0.17927871586408173,
+      "grad_norm": 0.08275672793388367,
+      "learning_rate": 1.9946680281610957e-06,
+      "loss": 0.6268,
+      "step": 215
+    },
+    {
+      "epoch": 0.1801125703564728,
+      "grad_norm": 0.08361789584159851,
+      "learning_rate": 1.9945749793413017e-06,
+      "loss": 0.6145,
+      "step": 216
+    },
+    {
+      "epoch": 0.18094642484886386,
+      "grad_norm": 0.0829552635550499,
+      "learning_rate": 1.9944811278349665e-06,
+      "loss": 0.7203,
+      "step": 217
+    },
+    {
+      "epoch": 0.18178027934125496,
+      "grad_norm": 0.07961665093898773,
+      "learning_rate": 1.994386473717835e-06,
+      "loss": 0.6598,
+      "step": 218
+    },
+    {
+      "epoch": 0.18261413383364603,
+      "grad_norm": 0.08015939593315125,
+      "learning_rate": 1.9942910170662986e-06,
+      "loss": 0.6465,
+      "step": 219
+    },
+    {
+      "epoch": 0.18344798832603712,
+      "grad_norm": 0.08055105805397034,
+      "learning_rate": 1.994194757957397e-06,
+      "loss": 0.6543,
+      "step": 220
+    },
+    {
+      "epoch": 0.1842818428184282,
+      "grad_norm": 0.07763723284006119,
+      "learning_rate": 1.994097696468818e-06,
+      "loss": 0.6984,
+      "step": 221
+    },
+    {
+      "epoch": 0.18511569731081925,
+      "grad_norm": 0.07882793247699738,
+      "learning_rate": 1.9939998326788966e-06,
+      "loss": 0.6672,
+      "step": 222
+    },
+    {
+      "epoch": 0.18594955180321035,
+      "grad_norm": 0.08098764717578888,
+      "learning_rate": 1.993901166666615e-06,
+      "loss": 0.65,
+      "step": 223
+    },
+    {
+      "epoch": 0.18678340629560142,
+      "grad_norm": 0.07791118323802948,
+      "learning_rate": 1.9938016985116027e-06,
+      "loss": 0.6954,
+      "step": 224
+    },
+    {
+      "epoch": 0.18761726078799248,
+      "grad_norm": 0.07814016193151474,
+      "learning_rate": 1.993701428294137e-06,
+      "loss": 0.6788,
+      "step": 225
+    },
+    {
+      "epoch": 0.18845111528038358,
+      "grad_norm": 0.07884248346090317,
+      "learning_rate": 1.9936003560951433e-06,
+      "loss": 0.6577,
+      "step": 226
+    },
+    {
+      "epoch": 0.18928496977277465,
+      "grad_norm": 0.07807621359825134,
+      "learning_rate": 1.9934984819961924e-06,
+      "loss": 0.6424,
+      "step": 227
+    },
+    {
+      "epoch": 0.19011882426516574,
+      "grad_norm": 0.08191139996051788,
+      "learning_rate": 1.9933958060795043e-06,
+      "loss": 0.6809,
+      "step": 228
+    },
+    {
+      "epoch": 0.1909526787575568,
+      "grad_norm": 0.07597807049751282,
+      "learning_rate": 1.9932923284279445e-06,
+      "loss": 0.7109,
+      "step": 229
+    },
+    {
+      "epoch": 0.19178653324994788,
+      "grad_norm": 0.08211328089237213,
+      "learning_rate": 1.9931880491250263e-06,
+      "loss": 0.6916,
+      "step": 230
+    },
+    {
+      "epoch": 0.19262038774233897,
+      "grad_norm": 0.07775567471981049,
+      "learning_rate": 1.993082968254909e-06,
+      "loss": 0.6458,
+      "step": 231
+    },
+    {
+      "epoch": 0.19345424223473004,
+      "grad_norm": 0.08438362181186676,
+      "learning_rate": 1.992977085902402e-06,
+      "loss": 0.6946,
+      "step": 232
+    },
+    {
+      "epoch": 0.1942880967271211,
+      "grad_norm": 0.08631843328475952,
+      "learning_rate": 1.9928704021529564e-06,
+      "loss": 0.6643,
+      "step": 233
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 0.07766929268836975,
+      "learning_rate": 1.992762917092675e-06,
+      "loss": 0.6753,
+      "step": 234
+    },
+    {
+      "epoch": 0.19595580571190327,
+      "grad_norm": 0.07574877887964249,
+      "learning_rate": 1.9926546308083044e-06,
+      "loss": 0.6346,
+      "step": 235
+    },
+    {
+      "epoch": 0.19678966020429436,
+      "grad_norm": 0.07815206050872803,
+      "learning_rate": 1.9925455433872386e-06,
+      "loss": 0.6946,
+      "step": 236
+    },
+    {
+      "epoch": 0.19762351469668543,
+      "grad_norm": 0.07457253336906433,
+      "learning_rate": 1.9924356549175185e-06,
+      "loss": 0.7108,
+      "step": 237
+    },
+    {
+      "epoch": 0.1984573691890765,
+      "grad_norm": 0.07688765227794647,
+      "learning_rate": 1.992324965487831e-06,
+      "loss": 0.6764,
+      "step": 238
+    },
+    {
+      "epoch": 0.1992912236814676,
+      "grad_norm": 0.08050046861171722,
+      "learning_rate": 1.99221347518751e-06,
+      "loss": 0.6263,
+      "step": 239
+    },
+    {
+      "epoch": 0.20012507817385866,
+      "grad_norm": 0.07763144373893738,
+      "learning_rate": 1.9921011841065347e-06,
+      "loss": 0.6618,
+      "step": 240
+    },
+    {
+      "epoch": 0.20095893266624973,
+      "grad_norm": 0.07744456082582474,
+      "learning_rate": 1.991988092335532e-06,
+      "loss": 0.6922,
+      "step": 241
+    },
+    {
+      "epoch": 0.20179278715864082,
+      "grad_norm": 0.07596256583929062,
+      "learning_rate": 1.9918741999657743e-06,
+      "loss": 0.7267,
+      "step": 242
+    },
+    {
+      "epoch": 0.2026266416510319,
+      "grad_norm": 0.08003415167331696,
+      "learning_rate": 1.9917595070891793e-06,
+      "loss": 0.6796,
+      "step": 243
+    },
+    {
+      "epoch": 0.203460496143423,
+      "grad_norm": 0.08418845385313034,
+      "learning_rate": 1.9916440137983125e-06,
+      "loss": 0.6815,
+      "step": 244
+    },
+    {
+      "epoch": 0.20429435063581405,
+      "grad_norm": 0.0794699415564537,
+      "learning_rate": 1.991527720186384e-06,
+      "loss": 0.6962,
+      "step": 245
+    },
+    {
+      "epoch": 0.20512820512820512,
+      "grad_norm": 0.07894127070903778,
+      "learning_rate": 1.991410626347251e-06,
+      "loss": 0.6898,
+      "step": 246
+    },
+    {
+      "epoch": 0.20596205962059622,
+      "grad_norm": 0.0819648951292038,
+      "learning_rate": 1.9912927323754145e-06,
+      "loss": 0.726,
+      "step": 247
+    },
+    {
+      "epoch": 0.20679591411298728,
+      "grad_norm": 0.08042652159929276,
+      "learning_rate": 1.991174038366024e-06,
+      "loss": 0.6398,
+      "step": 248
+    },
+    {
+      "epoch": 0.20762976860537835,
+      "grad_norm": 0.07848809659481049,
+      "learning_rate": 1.9910545444148722e-06,
+      "loss": 0.6562,
+      "step": 249
+    },
+    {
+      "epoch": 0.20846362309776945,
+      "grad_norm": 0.08295590430498123,
+      "learning_rate": 1.9909342506183987e-06,
+      "loss": 0.6767,
+      "step": 250
+    },
+    {
+      "epoch": 0.2092974775901605,
+      "grad_norm": 0.0840587317943573,
+      "learning_rate": 1.9908131570736887e-06,
+      "loss": 0.7274,
+      "step": 251
+    },
+    {
+      "epoch": 0.2101313320825516,
+      "grad_norm": 0.08375398814678192,
+      "learning_rate": 1.9906912638784727e-06,
+      "loss": 0.6716,
+      "step": 252
+    },
+    {
+      "epoch": 0.21096518657494268,
+      "grad_norm": 0.08229872584342957,
+      "learning_rate": 1.990568571131126e-06,
+      "loss": 0.7385,
+      "step": 253
+    },
+    {
+      "epoch": 0.21179904106733374,
+      "grad_norm": 0.07736154645681381,
+      "learning_rate": 1.99044507893067e-06,
+      "loss": 0.71,
+      "step": 254
+    },
+    {
+      "epoch": 0.21263289555972484,
+      "grad_norm": 0.08812221139669418,
+      "learning_rate": 1.9903207873767704e-06,
+      "loss": 0.7045,
+      "step": 255
+    },
+    {
+      "epoch": 0.2134667500521159,
+      "grad_norm": 0.08660853654146194,
+      "learning_rate": 1.9901956965697386e-06,
+      "loss": 0.7218,
+      "step": 256
+    },
+    {
+      "epoch": 0.21430060454450697,
+      "grad_norm": 0.08562801033258438,
+      "learning_rate": 1.9900698066105318e-06,
+      "loss": 0.6376,
+      "step": 257
+    },
+    {
+      "epoch": 0.21513445903689807,
+      "grad_norm": 0.07764414697885513,
+      "learning_rate": 1.9899431176007503e-06,
+      "loss": 0.6629,
+      "step": 258
+    },
+    {
+      "epoch": 0.21596831352928914,
+      "grad_norm": 0.07547292113304138,
+      "learning_rate": 1.989815629642641e-06,
+      "loss": 0.6887,
+      "step": 259
+    },
+    {
+      "epoch": 0.21680216802168023,
+      "grad_norm": 0.08463111519813538,
+      "learning_rate": 1.9896873428390948e-06,
+      "loss": 0.6169,
+      "step": 260
+    },
+    {
+      "epoch": 0.2176360225140713,
+      "grad_norm": 0.08064530789852142,
+      "learning_rate": 1.9895582572936473e-06,
+      "loss": 0.6905,
+      "step": 261
+    },
+    {
+      "epoch": 0.21846987700646237,
+      "grad_norm": 0.07929642498493195,
+      "learning_rate": 1.9894283731104796e-06,
+      "loss": 0.6415,
+      "step": 262
+    },
+    {
+      "epoch": 0.21930373149885346,
+      "grad_norm": 0.07858303934335709,
+      "learning_rate": 1.9892976903944154e-06,
+      "loss": 0.6955,
+      "step": 263
+    },
+    {
+      "epoch": 0.22013758599124453,
+      "grad_norm": 0.07967871427536011,
+      "learning_rate": 1.9891662092509255e-06,
+      "loss": 0.7004,
+      "step": 264
+    },
+    {
+      "epoch": 0.2209714404836356,
+      "grad_norm": 0.08009492605924606,
+      "learning_rate": 1.989033929786123e-06,
+      "loss": 0.6206,
+      "step": 265
+    },
+    {
+      "epoch": 0.2218052949760267,
+      "grad_norm": 0.08245829492807388,
+      "learning_rate": 1.988900852106766e-06,
+      "loss": 0.7182,
+      "step": 266
+    },
+    {
+      "epoch": 0.22263914946841776,
+      "grad_norm": 0.07844138890504837,
+      "learning_rate": 1.9887669763202568e-06,
+      "loss": 0.6372,
+      "step": 267
+    },
+    {
+      "epoch": 0.22347300396080885,
+      "grad_norm": 0.0783441960811615,
+      "learning_rate": 1.9886323025346417e-06,
+      "loss": 0.6182,
+      "step": 268
+    },
+    {
+      "epoch": 0.22430685845319992,
+      "grad_norm": 0.0803346335887909,
+      "learning_rate": 1.9884968308586117e-06,
+      "loss": 0.6436,
+      "step": 269
+    },
+    {
+      "epoch": 0.225140712945591,
+      "grad_norm": 0.0813342034816742,
+      "learning_rate": 1.9883605614015015e-06,
+      "loss": 0.6472,
+      "step": 270
+    },
+    {
+      "epoch": 0.22597456743798208,
+      "grad_norm": 0.07891476154327393,
+      "learning_rate": 1.988223494273288e-06,
+      "loss": 0.6625,
+      "step": 271
+    },
+    {
+      "epoch": 0.22680842193037315,
+      "grad_norm": 0.08087664842605591,
+      "learning_rate": 1.9880856295845945e-06,
+      "loss": 0.7226,
+      "step": 272
+    },
+    {
+      "epoch": 0.22764227642276422,
+      "grad_norm": 0.08254389464855194,
+      "learning_rate": 1.9879469674466865e-06,
+      "loss": 0.6992,
+      "step": 273
+    },
+    {
+      "epoch": 0.2284761309151553,
+      "grad_norm": 0.08207474648952484,
+      "learning_rate": 1.987807507971473e-06,
+      "loss": 0.6827,
+      "step": 274
+    },
+    {
+      "epoch": 0.22930998540754638,
+      "grad_norm": 0.08394856005907059,
+      "learning_rate": 1.9876672512715075e-06,
+      "loss": 0.7445,
+      "step": 275
+    },
+    {
+      "epoch": 0.23014383989993747,
+      "grad_norm": 0.07934897392988205,
+      "learning_rate": 1.9875261974599856e-06,
+      "loss": 0.6988,
+      "step": 276
+    },
+    {
+      "epoch": 0.23097769439232854,
+      "grad_norm": 0.08756977319717407,
+      "learning_rate": 1.987384346650747e-06,
+      "loss": 0.6588,
+      "step": 277
+    },
+    {
+      "epoch": 0.2318115488847196,
+      "grad_norm": 0.08198294788599014,
+      "learning_rate": 1.9872416989582747e-06,
+      "loss": 0.6729,
+      "step": 278
+    },
+    {
+      "epoch": 0.2326454033771107,
+      "grad_norm": 0.07671192288398743,
+      "learning_rate": 1.9870982544976948e-06,
+      "loss": 0.6685,
+      "step": 279
+    },
+    {
+      "epoch": 0.23347925786950177,
+      "grad_norm": 0.07943634688854218,
+      "learning_rate": 1.986954013384776e-06,
+      "loss": 0.7352,
+      "step": 280
+    },
+    {
+      "epoch": 0.23431311236189284,
+      "grad_norm": 0.0829697921872139,
+      "learning_rate": 1.98680897573593e-06,
+      "loss": 0.6645,
+      "step": 281
+    },
+    {
+      "epoch": 0.23514696685428393,
+      "grad_norm": 0.08583538234233856,
+      "learning_rate": 1.986663141668212e-06,
+      "loss": 0.713,
+      "step": 282
+    },
+    {
+      "epoch": 0.235980821346675,
+      "grad_norm": 0.08206543326377869,
+      "learning_rate": 1.9865165112993192e-06,
+      "loss": 0.7211,
+      "step": 283
+    },
+    {
+      "epoch": 0.2368146758390661,
+      "grad_norm": 0.07899665087461472,
+      "learning_rate": 1.9863690847475924e-06,
+      "loss": 0.6757,
+      "step": 284
+    },
+    {
+      "epoch": 0.23764853033145716,
+      "grad_norm": 0.0813852995634079,
+      "learning_rate": 1.986220862132014e-06,
+      "loss": 0.692,
+      "step": 285
+    },
+    {
+      "epoch": 0.23848238482384823,
+      "grad_norm": 0.09079395979642868,
+      "learning_rate": 1.9860718435722093e-06,
+      "loss": 0.6756,
+      "step": 286
+    },
+    {
+      "epoch": 0.23931623931623933,
+      "grad_norm": 0.07518535107374191,
+      "learning_rate": 1.985922029188446e-06,
+      "loss": 0.626,
+      "step": 287
+    },
+    {
+      "epoch": 0.2401500938086304,
+      "grad_norm": 0.08228084444999695,
+      "learning_rate": 1.9857714191016337e-06,
+      "loss": 0.7394,
+      "step": 288
+    },
+    {
+      "epoch": 0.24098394830102146,
+      "grad_norm": 0.08036590367555618,
+      "learning_rate": 1.985620013433325e-06,
+      "loss": 0.6012,
+      "step": 289
+    },
+    {
+      "epoch": 0.24181780279341256,
+      "grad_norm": 0.08429574221372604,
+      "learning_rate": 1.985467812305714e-06,
+      "loss": 0.6881,
+      "step": 290
+    },
+    {
+      "epoch": 0.24265165728580362,
+      "grad_norm": 0.07866602391004562,
+      "learning_rate": 1.9853148158416364e-06,
+      "loss": 0.6708,
+      "step": 291
+    },
+    {
+      "epoch": 0.24348551177819472,
+      "grad_norm": 0.0796622484922409,
+      "learning_rate": 1.9851610241645714e-06,
+      "loss": 0.7129,
+      "step": 292
+    },
+    {
+      "epoch": 0.24431936627058579,
+      "grad_norm": 0.08136799931526184,
+      "learning_rate": 1.9850064373986374e-06,
+      "loss": 0.6806,
+      "step": 293
+    },
+    {
+      "epoch": 0.24515322076297685,
+      "grad_norm": 0.07808342576026917,
+      "learning_rate": 1.9848510556685973e-06,
+      "loss": 0.6217,
+      "step": 294
+    },
+    {
+      "epoch": 0.24598707525536795,
+      "grad_norm": 0.07293733209371567,
+      "learning_rate": 1.984694879099853e-06,
+      "loss": 0.662,
+      "step": 295
+    },
+    {
+      "epoch": 0.24682092974775902,
+      "grad_norm": 0.07730741053819656,
+      "learning_rate": 1.98453790781845e-06,
+      "loss": 0.6386,
+      "step": 296
+    },
+    {
+      "epoch": 0.24765478424015008,
+      "grad_norm": 0.08601437509059906,
+      "learning_rate": 1.9843801419510743e-06,
+      "loss": 0.6639,
+      "step": 297
+    },
+    {
+      "epoch": 0.24848863873254118,
+      "grad_norm": 0.0815718024969101,
+      "learning_rate": 1.9842215816250525e-06,
+      "loss": 0.6891,
+      "step": 298
+    },
+    {
+      "epoch": 0.24932249322493225,
+      "grad_norm": 0.08793067187070847,
+      "learning_rate": 1.9840622269683537e-06,
+      "loss": 0.684,
+      "step": 299
+    },
+    {
+      "epoch": 0.2501563477173233,
+      "grad_norm": 0.07922165095806122,
+      "learning_rate": 1.9839020781095873e-06,
+      "loss": 0.6913,
+      "step": 300
+    },
+    {
+      "epoch": 0.2509902022097144,
+      "grad_norm": 0.0789838656783104,
+      "learning_rate": 1.9837411351780035e-06,
+      "loss": 0.6556,
+      "step": 301
+    },
+    {
+      "epoch": 0.2518240567021055,
+      "grad_norm": 0.07958388328552246,
+      "learning_rate": 1.9835793983034944e-06,
+      "loss": 0.6301,
+      "step": 302
+    },
+    {
+      "epoch": 0.25265791119449654,
+      "grad_norm": 0.08036024868488312,
+      "learning_rate": 1.9834168676165915e-06,
+      "loss": 0.592,
+      "step": 303
+    },
+    {
+      "epoch": 0.25349176568688764,
+      "grad_norm": 0.08134333044290543,
+      "learning_rate": 1.9832535432484677e-06,
+      "loss": 0.7142,
+      "step": 304
+    },
+    {
+      "epoch": 0.25432562017927873,
+      "grad_norm": 0.08333199471235275,
+      "learning_rate": 1.983089425330937e-06,
+      "loss": 0.7021,
+      "step": 305
+    },
+    {
+      "epoch": 0.2551594746716698,
+      "grad_norm": 0.08486991375684738,
+      "learning_rate": 1.982924513996452e-06,
+      "loss": 0.6724,
+      "step": 306
+    },
+    {
+      "epoch": 0.25599332916406087,
+      "grad_norm": 0.08286510407924652,
+      "learning_rate": 1.982758809378108e-06,
+      "loss": 0.6856,
+      "step": 307
+    },
+    {
+      "epoch": 0.25682718365645196,
+      "grad_norm": 0.08175141364336014,
+      "learning_rate": 1.982592311609639e-06,
+      "loss": 0.6605,
+      "step": 308
+    },
+    {
+      "epoch": 0.257661038148843,
+      "grad_norm": 0.08094872534275055,
+      "learning_rate": 1.982425020825419e-06,
+      "loss": 0.7136,
+      "step": 309
+    },
+    {
+      "epoch": 0.2584948926412341,
+      "grad_norm": 0.07797659933567047,
+      "learning_rate": 1.9822569371604632e-06,
+      "loss": 0.6975,
+      "step": 310
+    },
+    {
+      "epoch": 0.2593287471336252,
+      "grad_norm": 0.08523422479629517,
+      "learning_rate": 1.982088060750426e-06,
+      "loss": 0.6487,
+      "step": 311
+    },
+    {
+      "epoch": 0.2601626016260163,
+      "grad_norm": 0.09064790606498718,
+      "learning_rate": 1.981918391731601e-06,
+      "loss": 0.7006,
+      "step": 312
+    },
+    {
+      "epoch": 0.2609964561184073,
+      "grad_norm": 0.08136545121669769,
+      "learning_rate": 1.9817479302409225e-06,
+      "loss": 0.67,
+      "step": 313
+    },
+    {
+      "epoch": 0.2618303106107984,
+      "grad_norm": 0.07847245037555695,
+      "learning_rate": 1.9815766764159642e-06,
+      "loss": 0.6692,
+      "step": 314
+    },
+    {
+      "epoch": 0.2626641651031895,
+      "grad_norm": 0.07690361142158508,
+      "learning_rate": 1.9814046303949387e-06,
+      "loss": 0.5999,
+      "step": 315
+    },
+    {
+      "epoch": 0.26349801959558056,
+      "grad_norm": 0.08166715502738953,
+      "learning_rate": 1.9812317923166987e-06,
+      "loss": 0.6666,
+      "step": 316
+    },
+    {
+      "epoch": 0.26433187408797165,
+      "grad_norm": 0.08074827492237091,
+      "learning_rate": 1.981058162320735e-06,
+      "loss": 0.657,
+      "step": 317
+    },
+    {
+      "epoch": 0.26516572858036275,
+      "grad_norm": 0.08311966806650162,
+      "learning_rate": 1.980883740547179e-06,
+      "loss": 0.6679,
+      "step": 318
+    },
+    {
+      "epoch": 0.2659995830727538,
+      "grad_norm": 0.08272409439086914,
+      "learning_rate": 1.9807085271368e-06,
+      "loss": 0.6467,
+      "step": 319
+    },
+    {
+      "epoch": 0.2668334375651449,
+      "grad_norm": 0.08190209418535233,
+      "learning_rate": 1.980532522231007e-06,
+      "loss": 0.623,
+      "step": 320
+    },
+    {
+      "epoch": 0.267667292057536,
+      "grad_norm": 0.0819978415966034,
+      "learning_rate": 1.980355725971847e-06,
+      "loss": 0.677,
+      "step": 321
+    },
+    {
+      "epoch": 0.268501146549927,
+      "grad_norm": 0.07838036119937897,
+      "learning_rate": 1.980178138502006e-06,
+      "loss": 0.6545,
+      "step": 322
+    },
+    {
+      "epoch": 0.2693350010423181,
+      "grad_norm": 0.08094768226146698,
+      "learning_rate": 1.979999759964809e-06,
+      "loss": 0.6429,
+      "step": 323
+    },
+    {
+      "epoch": 0.2701688555347092,
+      "grad_norm": 0.0796407088637352,
+      "learning_rate": 1.9798205905042184e-06,
+      "loss": 0.7114,
+      "step": 324
+    },
+    {
+      "epoch": 0.27100271002710025,
+      "grad_norm": 0.07667295634746552,
+      "learning_rate": 1.9796406302648366e-06,
+      "loss": 0.7059,
+      "step": 325
+    },
+    {
+      "epoch": 0.27183656451949134,
+      "grad_norm": 0.07456690818071365,
+      "learning_rate": 1.9794598793919023e-06,
+      "loss": 0.5275,
+      "step": 326
+    },
+    {
+      "epoch": 0.27267041901188244,
+      "grad_norm": 0.08084844797849655,
+      "learning_rate": 1.9792783380312935e-06,
+      "loss": 0.6449,
+      "step": 327
+    },
+    {
+      "epoch": 0.27350427350427353,
+      "grad_norm": 0.08692745119333267,
+      "learning_rate": 1.9790960063295254e-06,
+      "loss": 0.6831,
+      "step": 328
+    },
+    {
+      "epoch": 0.27433812799666457,
+      "grad_norm": 0.0852961540222168,
+      "learning_rate": 1.9789128844337524e-06,
+      "loss": 0.6518,
+      "step": 329
+    },
+    {
+      "epoch": 0.27517198248905567,
+      "grad_norm": 0.08140264451503754,
+      "learning_rate": 1.978728972491765e-06,
+      "loss": 0.606,
+      "step": 330
+    },
+    {
+      "epoch": 0.27600583698144676,
+      "grad_norm": 0.08339618146419525,
+      "learning_rate": 1.9785442706519927e-06,
+      "loss": 0.6792,
+      "step": 331
+    },
+    {
+      "epoch": 0.2768396914738378,
+      "grad_norm": 0.08711510896682739,
+      "learning_rate": 1.9783587790635017e-06,
+      "loss": 0.6903,
+      "step": 332
+    },
+    {
+      "epoch": 0.2776735459662289,
+      "grad_norm": 0.08612138777971268,
+      "learning_rate": 1.978172497875995e-06,
+      "loss": 0.6998,
+      "step": 333
+    },
+    {
+      "epoch": 0.27850740045862,
+      "grad_norm": 0.07922092080116272,
+      "learning_rate": 1.9779854272398146e-06,
+      "loss": 0.654,
+      "step": 334
+    },
+    {
+      "epoch": 0.27934125495101103,
+      "grad_norm": 0.08098642528057098,
+      "learning_rate": 1.9777975673059382e-06,
+      "loss": 0.6706,
+      "step": 335
+    },
+    {
+      "epoch": 0.2801751094434021,
+      "grad_norm": 0.0873970165848732,
+      "learning_rate": 1.977608918225981e-06,
+      "loss": 0.6856,
+      "step": 336
+    },
+    {
+      "epoch": 0.2810089639357932,
+      "grad_norm": 0.082435242831707,
+      "learning_rate": 1.9774194801521947e-06,
+      "loss": 0.6375,
+      "step": 337
+    },
+    {
+      "epoch": 0.28184281842818426,
+      "grad_norm": 0.08199653774499893,
+      "learning_rate": 1.9772292532374687e-06,
+      "loss": 0.6785,
+      "step": 338
+    },
+    {
+      "epoch": 0.28267667292057536,
+      "grad_norm": 0.08817990124225616,
+      "learning_rate": 1.9770382376353285e-06,
+      "loss": 0.6642,
+      "step": 339
+    },
+    {
+      "epoch": 0.28351052741296645,
+      "grad_norm": 0.08359472453594208,
+      "learning_rate": 1.976846433499935e-06,
+      "loss": 0.6555,
+      "step": 340
+    },
+    {
+      "epoch": 0.2843443819053575,
+      "grad_norm": 0.08499454706907272,
+      "learning_rate": 1.976653840986088e-06,
+      "loss": 0.687,
+      "step": 341
+    },
+    {
+      "epoch": 0.2851782363977486,
+      "grad_norm": 0.08286130428314209,
+      "learning_rate": 1.976460460249222e-06,
+      "loss": 0.6904,
+      "step": 342
+    },
+    {
+      "epoch": 0.2860120908901397,
+      "grad_norm": 0.08244482427835464,
+      "learning_rate": 1.976266291445406e-06,
+      "loss": 0.6546,
+      "step": 343
+    },
+    {
+      "epoch": 0.2868459453825308,
+      "grad_norm": 0.07950432598590851,
+      "learning_rate": 1.976071334731349e-06,
+      "loss": 0.6683,
+      "step": 344
+    },
+    {
+      "epoch": 0.2876797998749218,
+      "grad_norm": 0.0854944959282875,
+      "learning_rate": 1.975875590264393e-06,
+      "loss": 0.6947,
+      "step": 345
+    },
+    {
+      "epoch": 0.2885136543673129,
+      "grad_norm": 0.08143036812543869,
+      "learning_rate": 1.975679058202516e-06,
+      "loss": 0.6422,
+      "step": 346
+    },
+    {
+      "epoch": 0.289347508859704,
+      "grad_norm": 0.0797310546040535,
+      "learning_rate": 1.9754817387043325e-06,
+      "loss": 0.6939,
+      "step": 347
+    },
+    {
+      "epoch": 0.29018136335209505,
+      "grad_norm": 0.08961952477693558,
+      "learning_rate": 1.9752836319290926e-06,
+      "loss": 0.6848,
+      "step": 348
+    },
+    {
+      "epoch": 0.29101521784448614,
+      "grad_norm": 0.08825525641441345,
+      "learning_rate": 1.9750847380366806e-06,
+      "loss": 0.7088,
+      "step": 349
+    },
+    {
+      "epoch": 0.29184907233687724,
+      "grad_norm": 0.08601095527410507,
+      "learning_rate": 1.9748850571876168e-06,
+      "loss": 0.6283,
+      "step": 350
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 0.08371152728796005,
+      "learning_rate": 1.9746845895430568e-06,
+      "loss": 0.6755,
+      "step": 351
+    },
+    {
+      "epoch": 0.29351678132165937,
+      "grad_norm": 0.08786465227603912,
+      "learning_rate": 1.974483335264791e-06,
+      "loss": 0.6149,
+      "step": 352
+    },
+    {
+      "epoch": 0.29435063581405047,
+      "grad_norm": 0.083786241710186,
+      "learning_rate": 1.974281294515245e-06,
+      "loss": 0.7104,
+      "step": 353
+    },
+    {
+      "epoch": 0.2951844903064415,
+      "grad_norm": 0.0846792683005333,
+      "learning_rate": 1.9740784674574785e-06,
+      "loss": 0.7032,
+      "step": 354
+    },
+    {
+      "epoch": 0.2960183447988326,
+      "grad_norm": 0.08477747440338135,
+      "learning_rate": 1.973874854255186e-06,
+      "loss": 0.6244,
+      "step": 355
+    },
+    {
+      "epoch": 0.2968521992912237,
+      "grad_norm": 0.09119998663663864,
+      "learning_rate": 1.9736704550726965e-06,
+      "loss": 0.6373,
+      "step": 356
+    },
+    {
+      "epoch": 0.29768605378361473,
+      "grad_norm": 0.08474856615066528,
+      "learning_rate": 1.9734652700749733e-06,
+      "loss": 0.736,
+      "step": 357
+    },
+    {
+      "epoch": 0.29851990827600583,
+      "grad_norm": 0.08070901036262512,
+      "learning_rate": 1.9732592994276145e-06,
+      "loss": 0.6442,
+      "step": 358
+    },
+    {
+      "epoch": 0.2993537627683969,
+      "grad_norm": 0.08306030929088593,
+      "learning_rate": 1.9730525432968517e-06,
+      "loss": 0.6863,
+      "step": 359
+    },
+    {
+      "epoch": 0.300187617260788,
+      "grad_norm": 0.08554356545209885,
+      "learning_rate": 1.97284500184955e-06,
+      "loss": 0.6782,
+      "step": 360
+    },
+    {
+      "epoch": 0.30102147175317906,
+      "grad_norm": 0.08350827544927597,
+      "learning_rate": 1.9726366752532096e-06,
+      "loss": 0.6656,
+      "step": 361
+    },
+    {
+      "epoch": 0.30185532624557015,
+      "grad_norm": 0.08501608669757843,
+      "learning_rate": 1.972427563675963e-06,
+      "loss": 0.6651,
+      "step": 362
+    },
+    {
+      "epoch": 0.30268918073796125,
+      "grad_norm": 0.08401606976985931,
+      "learning_rate": 1.972217667286577e-06,
+      "loss": 0.7153,
+      "step": 363
+    },
+    {
+      "epoch": 0.3035230352303523,
+      "grad_norm": 0.08713418990373611,
+      "learning_rate": 1.9720069862544513e-06,
+      "loss": 0.6623,
+      "step": 364
+    },
+    {
+      "epoch": 0.3043568897227434,
+      "grad_norm": 0.0871342197060585,
+      "learning_rate": 1.9717955207496197e-06,
+      "loss": 0.6571,
+      "step": 365
+    },
+    {
+      "epoch": 0.3051907442151345,
+      "grad_norm": 0.08501847088336945,
+      "learning_rate": 1.9715832709427477e-06,
+      "loss": 0.6714,
+      "step": 366
+    },
+    {
+      "epoch": 0.3060245987075255,
+      "grad_norm": 0.07894369214773178,
+      "learning_rate": 1.971370237005136e-06,
+      "loss": 0.6737,
+      "step": 367
+    },
+    {
+      "epoch": 0.3068584531999166,
+      "grad_norm": 0.08392394334077835,
+      "learning_rate": 1.9711564191087157e-06,
+      "loss": 0.6807,
+      "step": 368
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.08396563678979874,
+      "learning_rate": 1.970941817426052e-06,
+      "loss": 0.6766,
+      "step": 369
+    },
+    {
+      "epoch": 0.30852616218469875,
+      "grad_norm": 0.07790801674127579,
+      "learning_rate": 1.9707264321303425e-06,
+      "loss": 0.6293,
+      "step": 370
+    },
+    {
+      "epoch": 0.30936001667708984,
+      "grad_norm": 0.08699779212474823,
+      "learning_rate": 1.970510263395417e-06,
+      "loss": 0.681,
+      "step": 371
+    },
+    {
+      "epoch": 0.31019387116948094,
+      "grad_norm": 0.08176835626363754,
+      "learning_rate": 1.9702933113957384e-06,
+      "loss": 0.6931,
+      "step": 372
+    },
+    {
+      "epoch": 0.311027725661872,
+      "grad_norm": 0.08946527540683746,
+      "learning_rate": 1.9700755763063997e-06,
+      "loss": 0.6062,
+      "step": 373
+    },
+    {
+      "epoch": 0.3118615801542631,
+      "grad_norm": 0.08260657638311386,
+      "learning_rate": 1.9698570583031284e-06,
+      "loss": 0.6466,
+      "step": 374
+    },
+    {
+      "epoch": 0.31269543464665417,
+      "grad_norm": 0.08951833099126816,
+      "learning_rate": 1.969637757562282e-06,
+      "loss": 0.6458,
+      "step": 375
+    },
+    {
+      "epoch": 0.31352928913904526,
+      "grad_norm": 0.08605215698480606,
+      "learning_rate": 1.9694176742608507e-06,
+      "loss": 0.6919,
+      "step": 376
+    },
+    {
+      "epoch": 0.3143631436314363,
+      "grad_norm": 0.08389750868082047,
+      "learning_rate": 1.969196808576456e-06,
+      "loss": 0.6941,
+      "step": 377
+    },
+    {
+      "epoch": 0.3151969981238274,
+      "grad_norm": 0.08429732173681259,
+      "learning_rate": 1.968975160687351e-06,
+      "loss": 0.622,
+      "step": 378
+    },
+    {
+      "epoch": 0.3160308526162185,
+      "grad_norm": 0.08689334988594055,
+      "learning_rate": 1.9687527307724195e-06,
+      "loss": 0.6235,
+      "step": 379
+    },
+    {
+      "epoch": 0.31686470710860953,
+      "grad_norm": 0.08494516462087631,
+      "learning_rate": 1.9685295190111774e-06,
+      "loss": 0.6491,
+      "step": 380
+    },
+    {
+      "epoch": 0.31769856160100063,
+      "grad_norm": 0.0890749841928482,
+      "learning_rate": 1.9683055255837708e-06,
+      "loss": 0.6791,
+      "step": 381
+    },
+    {
+      "epoch": 0.3185324160933917,
+      "grad_norm": 0.07894483208656311,
+      "learning_rate": 1.968080750670977e-06,
+      "loss": 0.7075,
+      "step": 382
+    },
+    {
+      "epoch": 0.31936627058578276,
+      "grad_norm": 0.08575093746185303,
+      "learning_rate": 1.9678551944542033e-06,
+      "loss": 0.7027,
+      "step": 383
+    },
+    {
+      "epoch": 0.32020012507817386,
+      "grad_norm": 0.08360203355550766,
+      "learning_rate": 1.9676288571154895e-06,
+      "loss": 0.5894,
+      "step": 384
+    },
+    {
+      "epoch": 0.32103397957056495,
+      "grad_norm": 0.08341158181428909,
+      "learning_rate": 1.9674017388375036e-06,
+      "loss": 0.658,
+      "step": 385
+    },
+    {
+      "epoch": 0.321867834062956,
+      "grad_norm": 0.0844089537858963,
+      "learning_rate": 1.9671738398035446e-06,
+      "loss": 0.6691,
+      "step": 386
+    },
+    {
+      "epoch": 0.3227016885553471,
+      "grad_norm": 0.08446256816387177,
+      "learning_rate": 1.9669451601975426e-06,
+      "loss": 0.7211,
+      "step": 387
+    },
+    {
+      "epoch": 0.3235355430477382,
+      "grad_norm": 0.08008047193288803,
+      "learning_rate": 1.9667157002040565e-06,
+      "loss": 0.5843,
+      "step": 388
+    },
+    {
+      "epoch": 0.3243693975401292,
+      "grad_norm": 0.08821803331375122,
+      "learning_rate": 1.9664854600082754e-06,
+      "loss": 0.6444,
+      "step": 389
+    },
+    {
+      "epoch": 0.3252032520325203,
+      "grad_norm": 0.08548033237457275,
+      "learning_rate": 1.966254439796018e-06,
+      "loss": 0.6643,
+      "step": 390
+    },
+    {
+      "epoch": 0.3260371065249114,
+      "grad_norm": 0.08261854946613312,
+      "learning_rate": 1.9660226397537322e-06,
+      "loss": 0.6469,
+      "step": 391
+    },
+    {
+      "epoch": 0.3268709610173025,
+      "grad_norm": 0.07960055768489838,
+      "learning_rate": 1.965790060068497e-06,
+      "loss": 0.6149,
+      "step": 392
+    },
+    {
+      "epoch": 0.32770481550969355,
+      "grad_norm": 0.0834561213850975,
+      "learning_rate": 1.9655567009280177e-06,
+      "loss": 0.6798,
+      "step": 393
+    },
+    {
+      "epoch": 0.32853867000208464,
+      "grad_norm": 0.08524155616760254,
+      "learning_rate": 1.9653225625206313e-06,
+      "loss": 0.6999,
+      "step": 394
+    },
+    {
+      "epoch": 0.32937252449447574,
+      "grad_norm": 0.08503536134958267,
+      "learning_rate": 1.965087645035302e-06,
+      "loss": 0.6674,
+      "step": 395
+    },
+    {
+      "epoch": 0.3302063789868668,
+      "grad_norm": 0.09243112802505493,
+      "learning_rate": 1.964851948661624e-06,
+      "loss": 0.6391,
+      "step": 396
+    },
+    {
+      "epoch": 0.3310402334792579,
+      "grad_norm": 0.08015627413988113,
+      "learning_rate": 1.96461547358982e-06,
+      "loss": 0.611,
+      "step": 397
+    },
+    {
+      "epoch": 0.33187408797164897,
+      "grad_norm": 0.08681736886501312,
+      "learning_rate": 1.9643782200107395e-06,
+      "loss": 0.6967,
+      "step": 398
+    },
+    {
+      "epoch": 0.33270794246404,
+      "grad_norm": 0.08576110005378723,
+      "learning_rate": 1.9641401881158623e-06,
+      "loss": 0.6753,
+      "step": 399
+    },
+    {
+      "epoch": 0.3335417969564311,
+      "grad_norm": 0.08432309329509735,
+      "learning_rate": 1.9639013780972954e-06,
+      "loss": 0.6312,
+      "step": 400
+    },
+    {
+      "epoch": 0.3335417969564311,
+      "eval_loss": 0.6676867008209229,
+      "eval_runtime": 321.9965,
+      "eval_samples_per_second": 16.718,
+      "eval_steps_per_second": 2.789,
+      "step": 400
+    },
+    {
+      "epoch": 0.3343756514488222,
+      "grad_norm": 0.08781838417053223,
+      "learning_rate": 1.9636617901477742e-06,
+      "loss": 0.698,
+      "step": 401
+    },
+    {
+      "epoch": 0.33520950594121324,
+      "grad_norm": 0.08723490685224533,
+      "learning_rate": 1.9634214244606613e-06,
+      "loss": 0.6642,
+      "step": 402
+    },
+    {
+      "epoch": 0.33604336043360433,
+      "grad_norm": 0.08249269425868988,
+      "learning_rate": 1.963180281229948e-06,
+      "loss": 0.705,
+      "step": 403
+    },
+    {
+      "epoch": 0.3368772149259954,
+      "grad_norm": 0.08193965256214142,
+      "learning_rate": 1.962938360650252e-06,
+      "loss": 0.6813,
+      "step": 404
+    },
+    {
+      "epoch": 0.33771106941838647,
+      "grad_norm": 0.08125729858875275,
+      "learning_rate": 1.962695662916819e-06,
+      "loss": 0.6571,
+      "step": 405
+    },
+    {
+      "epoch": 0.33854492391077756,
+      "grad_norm": 0.08508846163749695,
+      "learning_rate": 1.962452188225522e-06,
+      "loss": 0.6212,
+      "step": 406
+    },
+    {
+      "epoch": 0.33937877840316866,
+      "grad_norm": 0.0881478488445282,
+      "learning_rate": 1.962207936772861e-06,
+      "loss": 0.6582,
+      "step": 407
+    },
+    {
+      "epoch": 0.34021263289555975,
+      "grad_norm": 0.08948707580566406,
+      "learning_rate": 1.9619629087559622e-06,
+      "loss": 0.6924,
+      "step": 408
+    },
+    {
+      "epoch": 0.3410464873879508,
+      "grad_norm": 0.08214866369962692,
+      "learning_rate": 1.9617171043725797e-06,
+      "loss": 0.6266,
+      "step": 409
+    },
+    {
+      "epoch": 0.3418803418803419,
+      "grad_norm": 0.09406815469264984,
+      "learning_rate": 1.961470523821093e-06,
+      "loss": 0.6754,
+      "step": 410
+    },
+    {
+      "epoch": 0.342714196372733,
+      "grad_norm": 0.08819446712732315,
+      "learning_rate": 1.961223167300509e-06,
+      "loss": 0.6858,
+      "step": 411
+    },
+    {
+      "epoch": 0.343548050865124,
+      "grad_norm": 0.0902935117483139,
+      "learning_rate": 1.9609750350104606e-06,
+      "loss": 0.6654,
+      "step": 412
+    },
+    {
+      "epoch": 0.3443819053575151,
+      "grad_norm": 0.0852714329957962,
+      "learning_rate": 1.9607261271512065e-06,
+      "loss": 0.6604,
+      "step": 413
+    },
+    {
+      "epoch": 0.3452157598499062,
+      "grad_norm": 0.08480080962181091,
+      "learning_rate": 1.9604764439236313e-06,
+      "loss": 0.7252,
+      "step": 414
+    },
+    {
+      "epoch": 0.34604961434229725,
+      "grad_norm": 0.08467400074005127,
+      "learning_rate": 1.960225985529246e-06,
+      "loss": 0.6182,
+      "step": 415
+    },
+    {
+      "epoch": 0.34688346883468835,
+      "grad_norm": 0.09190120548009872,
+      "learning_rate": 1.9599747521701865e-06,
+      "loss": 0.6605,
+      "step": 416
+    },
+    {
+      "epoch": 0.34771732332707944,
+      "grad_norm": 0.08994212001562119,
+      "learning_rate": 1.9597227440492143e-06,
+      "loss": 0.6674,
+      "step": 417
+    },
+    {
+      "epoch": 0.3485511778194705,
+      "grad_norm": 0.08530712872743607,
+      "learning_rate": 1.959469961369716e-06,
+      "loss": 0.673,
+      "step": 418
+    },
+    {
+      "epoch": 0.3493850323118616,
+      "grad_norm": 0.0814647376537323,
+      "learning_rate": 1.9592164043357046e-06,
+      "loss": 0.6224,
+      "step": 419
+    },
+    {
+      "epoch": 0.35021888680425267,
+      "grad_norm": 0.08769946545362473,
+      "learning_rate": 1.9589620731518164e-06,
+      "loss": 0.6446,
+      "step": 420
+    },
+    {
+      "epoch": 0.3510527412966437,
+      "grad_norm": 0.0855315700173378,
+      "learning_rate": 1.958706968023313e-06,
+      "loss": 0.6665,
+      "step": 421
+    },
+    {
+      "epoch": 0.3518865957890348,
+      "grad_norm": 0.08764835447072983,
+      "learning_rate": 1.958451089156082e-06,
+      "loss": 0.6888,
+      "step": 422
+    },
+    {
+      "epoch": 0.3527204502814259,
+      "grad_norm": 0.08674878627061844,
+      "learning_rate": 1.9581944367566324e-06,
+      "loss": 0.6894,
+      "step": 423
+    },
+    {
+      "epoch": 0.353554304773817,
+      "grad_norm": 0.07654520124197006,
+      "learning_rate": 1.9579370110321005e-06,
+      "loss": 0.5972,
+      "step": 424
+    },
+    {
+      "epoch": 0.35438815926620804,
+      "grad_norm": 0.08671235293149948,
+      "learning_rate": 1.9576788121902454e-06,
+      "loss": 0.6705,
+      "step": 425
+    },
+    {
+      "epoch": 0.35522201375859913,
+      "grad_norm": 0.08799296617507935,
+      "learning_rate": 1.9574198404394505e-06,
+      "loss": 0.66,
+      "step": 426
+    },
+    {
+      "epoch": 0.3560558682509902,
+      "grad_norm": 0.08908937871456146,
+      "learning_rate": 1.9571600959887224e-06,
+      "loss": 0.6687,
+      "step": 427
+    },
+    {
+      "epoch": 0.35688972274338127,
+      "grad_norm": 0.08397683501243591,
+      "learning_rate": 1.9568995790476915e-06,
+      "loss": 0.5915,
+      "step": 428
+    },
+    {
+      "epoch": 0.35772357723577236,
+      "grad_norm": 0.08805875480175018,
+      "learning_rate": 1.9566382898266126e-06,
+      "loss": 0.6706,
+      "step": 429
+    },
+    {
+      "epoch": 0.35855743172816346,
+      "grad_norm": 0.08281008154153824,
+      "learning_rate": 1.956376228536363e-06,
+      "loss": 0.5692,
+      "step": 430
+    },
+    {
+      "epoch": 0.3593912862205545,
+      "grad_norm": 0.08514556288719177,
+      "learning_rate": 1.9561133953884427e-06,
+      "loss": 0.5938,
+      "step": 431
+    },
+    {
+      "epoch": 0.3602251407129456,
+      "grad_norm": 0.09112891554832458,
+      "learning_rate": 1.955849790594975e-06,
+      "loss": 0.7044,
+      "step": 432
+    },
+    {
+      "epoch": 0.3610589952053367,
+      "grad_norm": 0.08578234165906906,
+      "learning_rate": 1.9555854143687064e-06,
+      "loss": 0.6901,
+      "step": 433
+    },
+    {
+      "epoch": 0.3618928496977277,
+      "grad_norm": 0.08405105024576187,
+      "learning_rate": 1.955320266923006e-06,
+      "loss": 0.6307,
+      "step": 434
+    },
+    {
+      "epoch": 0.3627267041901188,
+      "grad_norm": 0.085452601313591,
+      "learning_rate": 1.9550543484718644e-06,
+      "loss": 0.6147,
+      "step": 435
+    },
+    {
+      "epoch": 0.3635605586825099,
+      "grad_norm": 0.0871538370847702,
+      "learning_rate": 1.9547876592298955e-06,
+      "loss": 0.6415,
+      "step": 436
+    },
+    {
+      "epoch": 0.36439441317490096,
+      "grad_norm": 0.08855723589658737,
+      "learning_rate": 1.954520199412334e-06,
+      "loss": 0.7026,
+      "step": 437
+    },
+    {
+      "epoch": 0.36522826766729205,
+      "grad_norm": 0.08221881091594696,
+      "learning_rate": 1.9542519692350387e-06,
+      "loss": 0.6813,
+      "step": 438
+    },
+    {
+      "epoch": 0.36606212215968315,
+      "grad_norm": 0.08595414459705353,
+      "learning_rate": 1.9539829689144876e-06,
+      "loss": 0.6809,
+      "step": 439
+    },
+    {
+      "epoch": 0.36689597665207424,
+      "grad_norm": 0.08485814183950424,
+      "learning_rate": 1.9537131986677816e-06,
+      "loss": 0.6352,
+      "step": 440
+    },
+    {
+      "epoch": 0.3677298311444653,
+      "grad_norm": 0.08620929718017578,
+      "learning_rate": 1.9534426587126433e-06,
+      "loss": 0.7072,
+      "step": 441
+    },
+    {
+      "epoch": 0.3685636856368564,
+      "grad_norm": 0.0919066071510315,
+      "learning_rate": 1.9531713492674154e-06,
+      "loss": 0.6809,
+      "step": 442
+    },
+    {
+      "epoch": 0.36939754012924747,
+      "grad_norm": 0.08376951515674591,
+      "learning_rate": 1.9528992705510628e-06,
+      "loss": 0.6486,
+      "step": 443
+    },
+    {
+      "epoch": 0.3702313946216385,
+      "grad_norm": 0.088445745408535,
+      "learning_rate": 1.95262642278317e-06,
+      "loss": 0.709,
+      "step": 444
+    },
+    {
+      "epoch": 0.3710652491140296,
+      "grad_norm": 0.08183503895998001,
+      "learning_rate": 1.952352806183943e-06,
+      "loss": 0.6277,
+      "step": 445
+    },
+    {
+      "epoch": 0.3718991036064207,
+      "grad_norm": 0.08638439327478409,
+      "learning_rate": 1.9520784209742094e-06,
+      "loss": 0.6665,
+      "step": 446
+    },
+    {
+      "epoch": 0.37273295809881174,
+      "grad_norm": 0.08837208896875381,
+      "learning_rate": 1.951803267375414e-06,
+      "loss": 0.6103,
+      "step": 447
+    },
+    {
+      "epoch": 0.37356681259120283,
+      "grad_norm": 0.08010973036289215,
+      "learning_rate": 1.9515273456096247e-06,
+      "loss": 0.662,
+      "step": 448
+    },
+    {
+      "epoch": 0.37440066708359393,
+      "grad_norm": 0.08395659178495407,
+      "learning_rate": 1.9512506558995283e-06,
+      "loss": 0.687,
+      "step": 449
+    },
+    {
+      "epoch": 0.37523452157598497,
+      "grad_norm": 0.08506524562835693,
+      "learning_rate": 1.950973198468431e-06,
+      "loss": 0.6695,
+      "step": 450
+    },
+    {
+      "epoch": 0.37606837606837606,
+      "grad_norm": 0.08150207251310349,
+      "learning_rate": 1.950694973540259e-06,
+      "loss": 0.6258,
+      "step": 451
+    },
+    {
+      "epoch": 0.37690223056076716,
+      "grad_norm": 0.08509568125009537,
+      "learning_rate": 1.9504159813395576e-06,
+      "loss": 0.5774,
+      "step": 452
+    },
+    {
+      "epoch": 0.3777360850531582,
+      "grad_norm": 0.08139500021934509,
+      "learning_rate": 1.950136222091492e-06,
+      "loss": 0.6652,
+      "step": 453
+    },
+    {
+      "epoch": 0.3785699395455493,
+      "grad_norm": 0.08891302347183228,
+      "learning_rate": 1.949855696021846e-06,
+      "loss": 0.6602,
+      "step": 454
+    },
+    {
+      "epoch": 0.3794037940379404,
+      "grad_norm": 0.08729292452335358,
+      "learning_rate": 1.949574403357022e-06,
+      "loss": 0.6232,
+      "step": 455
+    },
+    {
+      "epoch": 0.3802376485303315,
+      "grad_norm": 0.0855722650885582,
+      "learning_rate": 1.9492923443240425e-06,
+      "loss": 0.7249,
+      "step": 456
+    },
+    {
+      "epoch": 0.3810715030227225,
+      "grad_norm": 0.09225167334079742,
+      "learning_rate": 1.949009519150546e-06,
+      "loss": 0.7061,
+      "step": 457
+    },
+    {
+      "epoch": 0.3819053575151136,
+      "grad_norm": 0.08833880722522736,
+      "learning_rate": 1.9487259280647918e-06,
+      "loss": 0.6683,
+      "step": 458
+    },
+    {
+      "epoch": 0.3827392120075047,
+      "grad_norm": 0.08648336678743362,
+      "learning_rate": 1.948441571295656e-06,
+      "loss": 0.6716,
+      "step": 459
+    },
+    {
+      "epoch": 0.38357306649989575,
+      "grad_norm": 0.08678124845027924,
+      "learning_rate": 1.9481564490726324e-06,
+      "loss": 0.6869,
+      "step": 460
+    },
+    {
+      "epoch": 0.38440692099228685,
+      "grad_norm": 0.0866696834564209,
+      "learning_rate": 1.9478705616258344e-06,
+      "loss": 0.6346,
+      "step": 461
+    },
+    {
+      "epoch": 0.38524077548467794,
+      "grad_norm": 0.09009408950805664,
+      "learning_rate": 1.9475839091859904e-06,
+      "loss": 0.6507,
+      "step": 462
+    },
+    {
+      "epoch": 0.386074629977069,
+      "grad_norm": 0.09393912553787231,
+      "learning_rate": 1.9472964919844484e-06,
+      "loss": 0.6624,
+      "step": 463
+    },
+    {
+      "epoch": 0.3869084844694601,
+      "grad_norm": 0.09663153439760208,
+      "learning_rate": 1.9470083102531722e-06,
+      "loss": 0.6378,
+      "step": 464
+    },
+    {
+      "epoch": 0.3877423389618512,
+      "grad_norm": 0.08551555126905441,
+      "learning_rate": 1.9467193642247435e-06,
+      "loss": 0.6717,
+      "step": 465
+    },
+    {
+      "epoch": 0.3885761934542422,
+      "grad_norm": 0.08986209332942963,
+      "learning_rate": 1.9464296541323602e-06,
+      "loss": 0.7313,
+      "step": 466
+    },
+    {
+      "epoch": 0.3894100479466333,
+      "grad_norm": 0.08851300179958344,
+      "learning_rate": 1.9461391802098373e-06,
+      "loss": 0.6783,
+      "step": 467
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 0.09088604897260666,
+      "learning_rate": 1.9458479426916068e-06,
+      "loss": 0.5952,
+      "step": 468
+    },
+    {
+      "epoch": 0.39107775693141544,
+      "grad_norm": 0.08506208658218384,
+      "learning_rate": 1.9455559418127144e-06,
+      "loss": 0.6912,
+      "step": 469
+    },
+    {
+      "epoch": 0.39191161142380654,
+      "grad_norm": 0.087041936814785,
+      "learning_rate": 1.945263177808826e-06,
+      "loss": 0.6722,
+      "step": 470
+    },
+    {
+      "epoch": 0.39274546591619763,
+      "grad_norm": 0.09510099142789841,
+      "learning_rate": 1.9449696509162193e-06,
+      "loss": 0.697,
+      "step": 471
+    },
+    {
+      "epoch": 0.39357932040858873,
+      "grad_norm": 0.0864386260509491,
+      "learning_rate": 1.944675361371791e-06,
+      "loss": 0.6882,
+      "step": 472
+    },
+    {
+      "epoch": 0.39441317490097977,
+      "grad_norm": 0.08964356780052185,
+      "learning_rate": 1.944380309413051e-06,
+      "loss": 0.6403,
+      "step": 473
+    },
+    {
+      "epoch": 0.39524702939337086,
+      "grad_norm": 0.09786627441644669,
+      "learning_rate": 1.9440844952781253e-06,
+      "loss": 0.7037,
+      "step": 474
+    },
+    {
+      "epoch": 0.39608088388576196,
+      "grad_norm": 0.08709276467561722,
+      "learning_rate": 1.9437879192057556e-06,
+      "loss": 0.6619,
+      "step": 475
+    },
+    {
+      "epoch": 0.396914738378153,
+      "grad_norm": 0.08677015453577042,
+      "learning_rate": 1.9434905814352976e-06,
+      "loss": 0.6609,
+      "step": 476
+    },
+    {
+      "epoch": 0.3977485928705441,
+      "grad_norm": 0.08886487782001495,
+      "learning_rate": 1.9431924822067225e-06,
+      "loss": 0.6045,
+      "step": 477
+    },
+    {
+      "epoch": 0.3985824473629352,
+      "grad_norm": 0.08575446903705597,
+      "learning_rate": 1.942893621760616e-06,
+      "loss": 0.6886,
+      "step": 478
+    },
+    {
+      "epoch": 0.39941630185532623,
+      "grad_norm": 0.08386967331171036,
+      "learning_rate": 1.9425940003381767e-06,
+      "loss": 0.6721,
+      "step": 479
+    },
+    {
+      "epoch": 0.4002501563477173,
+      "grad_norm": 0.08702324330806732,
+      "learning_rate": 1.9422936181812197e-06,
+      "loss": 0.6959,
+      "step": 480
+    },
+    {
+      "epoch": 0.4010840108401084,
+      "grad_norm": 0.09090246260166168,
+      "learning_rate": 1.9419924755321727e-06,
+      "loss": 0.7094,
+      "step": 481
+    },
+    {
+      "epoch": 0.40191786533249946,
+      "grad_norm": 0.09213659167289734,
+      "learning_rate": 1.9416905726340767e-06,
+      "loss": 0.6446,
+      "step": 482
+    },
+    {
+      "epoch": 0.40275171982489055,
+      "grad_norm": 0.09240783751010895,
+      "learning_rate": 1.9413879097305878e-06,
+      "loss": 0.6401,
+      "step": 483
+    },
+    {
+      "epoch": 0.40358557431728165,
+      "grad_norm": 0.09058364480733871,
+      "learning_rate": 1.941084487065974e-06,
+      "loss": 0.7035,
+      "step": 484
+    },
+    {
+      "epoch": 0.4044194288096727,
+      "grad_norm": 0.0906047448515892,
+      "learning_rate": 1.940780304885117e-06,
+      "loss": 0.6519,
+      "step": 485
+    },
+    {
+      "epoch": 0.4052532833020638,
+      "grad_norm": 0.09101078659296036,
+      "learning_rate": 1.940475363433512e-06,
+      "loss": 0.6934,
+      "step": 486
+    },
+    {
+      "epoch": 0.4060871377944549,
+      "grad_norm": 0.08729363977909088,
+      "learning_rate": 1.940169662957266e-06,
+      "loss": 0.5822,
+      "step": 487
+    },
+    {
+      "epoch": 0.406920992286846,
+      "grad_norm": 0.0946895033121109,
+      "learning_rate": 1.9398632037031003e-06,
+      "loss": 0.6534,
+      "step": 488
+    },
+    {
+      "epoch": 0.407754846779237,
+      "grad_norm": 0.09069128334522247,
+      "learning_rate": 1.9395559859183463e-06,
+      "loss": 0.6706,
+      "step": 489
+    },
+    {
+      "epoch": 0.4085887012716281,
+      "grad_norm": 0.08318338543176651,
+      "learning_rate": 1.9392480098509483e-06,
+      "loss": 0.6368,
+      "step": 490
+    },
+    {
+      "epoch": 0.4094225557640192,
+      "grad_norm": 0.0826449766755104,
+      "learning_rate": 1.9389392757494645e-06,
+      "loss": 0.667,
+      "step": 491
+    },
+    {
+      "epoch": 0.41025641025641024,
+      "grad_norm": 0.09276958554983139,
+      "learning_rate": 1.938629783863062e-06,
+      "loss": 0.6709,
+      "step": 492
+    },
+    {
+      "epoch": 0.41109026474880134,
+      "grad_norm": 0.08281219005584717,
+      "learning_rate": 1.9383195344415215e-06,
+      "loss": 0.584,
+      "step": 493
+    },
+    {
+      "epoch": 0.41192411924119243,
+      "grad_norm": 0.08632051944732666,
+      "learning_rate": 1.938008527735234e-06,
+      "loss": 0.6743,
+      "step": 494
+    },
+    {
+      "epoch": 0.41275797373358347,
+      "grad_norm": 0.08970591425895691,
+      "learning_rate": 1.937696763995202e-06,
+      "loss": 0.6611,
+      "step": 495
+    },
+    {
+      "epoch": 0.41359182822597457,
+      "grad_norm": 0.0902903825044632,
+      "learning_rate": 1.93738424347304e-06,
+      "loss": 0.6455,
+      "step": 496
+    },
+    {
+      "epoch": 0.41442568271836566,
+      "grad_norm": 0.08964331448078156,
+      "learning_rate": 1.9370709664209715e-06,
+      "loss": 0.6561,
+      "step": 497
+    },
+    {
+      "epoch": 0.4152595372107567,
+      "grad_norm": 0.08478616923093796,
+      "learning_rate": 1.9367569330918317e-06,
+      "loss": 0.595,
+      "step": 498
+    },
+    {
+      "epoch": 0.4160933917031478,
+      "grad_norm": 0.09221872687339783,
+      "learning_rate": 1.9364421437390658e-06,
+      "loss": 0.6363,
+      "step": 499
+    },
+    {
+      "epoch": 0.4169272461955389,
+      "grad_norm": 0.09467694163322449,
+      "learning_rate": 1.936126598616729e-06,
+      "loss": 0.6696,
+      "step": 500
+    },
+    {
+      "epoch": 0.41776110068792993,
+      "grad_norm": 0.09464087337255478,
+      "learning_rate": 1.935810297979487e-06,
+      "loss": 0.6224,
+      "step": 501
+    },
+    {
+      "epoch": 0.418594955180321,
+      "grad_norm": 0.08994987607002258,
+      "learning_rate": 1.9354932420826147e-06,
+      "loss": 0.6383,
+      "step": 502
+    },
+    {
+      "epoch": 0.4194288096727121,
+      "grad_norm": 0.0921832025051117,
+      "learning_rate": 1.9351754311819974e-06,
+      "loss": 0.685,
+      "step": 503
+    },
+    {
+      "epoch": 0.4202626641651032,
+      "grad_norm": 0.08833765983581543,
+      "learning_rate": 1.934856865534129e-06,
+      "loss": 0.6233,
+      "step": 504
+    },
+    {
+      "epoch": 0.42109651865749426,
+      "grad_norm": 0.08863073587417603,
+      "learning_rate": 1.934537545396111e-06,
+      "loss": 0.6756,
+      "step": 505
+    },
+    {
+      "epoch": 0.42193037314988535,
+      "grad_norm": 0.08703204244375229,
+      "learning_rate": 1.934217471025658e-06,
+      "loss": 0.6372,
+      "step": 506
+    },
+    {
+      "epoch": 0.42276422764227645,
+      "grad_norm": 0.09010464698076248,
+      "learning_rate": 1.9338966426810887e-06,
+      "loss": 0.6677,
+      "step": 507
+    },
+    {
+      "epoch": 0.4235980821346675,
+      "grad_norm": 0.09014829248189926,
+      "learning_rate": 1.9335750606213336e-06,
+      "loss": 0.6839,
+      "step": 508
+    },
+    {
+      "epoch": 0.4244319366270586,
+      "grad_norm": 0.08871164172887802,
+      "learning_rate": 1.93325272510593e-06,
+      "loss": 0.7012,
+      "step": 509
+    },
+    {
+      "epoch": 0.4252657911194497,
+      "grad_norm": 0.08788250386714935,
+      "learning_rate": 1.9329296363950236e-06,
+      "loss": 0.6169,
+      "step": 510
+    },
+    {
+      "epoch": 0.4260996456118407,
+      "grad_norm": 0.08845999836921692,
+      "learning_rate": 1.9326057947493675e-06,
+      "loss": 0.6424,
+      "step": 511
+    },
+    {
+      "epoch": 0.4269335001042318,
+      "grad_norm": 0.0858997106552124,
+      "learning_rate": 1.9322812004303238e-06,
+      "loss": 0.6389,
+      "step": 512
+    },
+    {
+      "epoch": 0.4277673545966229,
+      "grad_norm": 0.08850055187940598,
+      "learning_rate": 1.9319558536998603e-06,
+      "loss": 0.6975,
+      "step": 513
+    },
+    {
+      "epoch": 0.42860120908901395,
+      "grad_norm": 0.09360900521278381,
+      "learning_rate": 1.9316297548205534e-06,
+      "loss": 0.686,
+      "step": 514
+    },
+    {
+      "epoch": 0.42943506358140504,
+      "grad_norm": 0.08473866432905197,
+      "learning_rate": 1.931302904055586e-06,
+      "loss": 0.6517,
+      "step": 515
+    },
+    {
+      "epoch": 0.43026891807379614,
+      "grad_norm": 0.08656331151723862,
+      "learning_rate": 1.9309753016687477e-06,
+      "loss": 0.7003,
+      "step": 516
+    },
+    {
+      "epoch": 0.4311027725661872,
+      "grad_norm": 0.08407936990261078,
+      "learning_rate": 1.9306469479244347e-06,
+      "loss": 0.6611,
+      "step": 517
+    },
+    {
+      "epoch": 0.43193662705857827,
+      "grad_norm": 0.09026416391134262,
+      "learning_rate": 1.930317843087651e-06,
+      "loss": 0.6749,
+      "step": 518
+    },
+    {
+      "epoch": 0.43277048155096937,
+      "grad_norm": 0.10031198710203171,
+      "learning_rate": 1.929987987424004e-06,
+      "loss": 0.6403,
+      "step": 519
+    },
+    {
+      "epoch": 0.43360433604336046,
+      "grad_norm": 0.09739411622285843,
+      "learning_rate": 1.9296573811997086e-06,
+      "loss": 0.6893,
+      "step": 520
+    },
+    {
+      "epoch": 0.4344381905357515,
+      "grad_norm": 0.08792299032211304,
+      "learning_rate": 1.929326024681587e-06,
+      "loss": 0.6268,
+      "step": 521
+    },
+    {
+      "epoch": 0.4352720450281426,
+      "grad_norm": 0.0874318853020668,
+      "learning_rate": 1.9289939181370646e-06,
+      "loss": 0.6659,
+      "step": 522
+    },
+    {
+      "epoch": 0.4361058995205337,
+      "grad_norm": 0.09086845815181732,
+      "learning_rate": 1.9286610618341724e-06,
+      "loss": 0.6399,
+      "step": 523
+    },
+    {
+      "epoch": 0.43693975401292473,
+      "grad_norm": 0.09616074711084366,
+      "learning_rate": 1.9283274560415477e-06,
+      "loss": 0.6325,
+      "step": 524
+    },
+    {
+      "epoch": 0.4377736085053158,
+      "grad_norm": 0.08674857765436172,
+      "learning_rate": 1.9279931010284322e-06,
+      "loss": 0.6487,
+      "step": 525
+    },
+    {
+      "epoch": 0.4386074629977069,
+      "grad_norm": 0.09142550826072693,
+      "learning_rate": 1.927657997064671e-06,
+      "loss": 0.6509,
+      "step": 526
+    },
+    {
+      "epoch": 0.43944131749009796,
+      "grad_norm": 0.0838690921664238,
+      "learning_rate": 1.9273221444207158e-06,
+      "loss": 0.6087,
+      "step": 527
+    },
+    {
+      "epoch": 0.44027517198248906,
+      "grad_norm": 0.0922529399394989,
+      "learning_rate": 1.9269855433676213e-06,
+      "loss": 0.6815,
+      "step": 528
+    },
+    {
+      "epoch": 0.44110902647488015,
+      "grad_norm": 0.08945809304714203,
+      "learning_rate": 1.926648194177046e-06,
+      "loss": 0.623,
+      "step": 529
+    },
+    {
+      "epoch": 0.4419428809672712,
+      "grad_norm": 0.09663840383291245,
+      "learning_rate": 1.926310097121253e-06,
+      "loss": 0.6244,
+      "step": 530
+    },
+    {
+      "epoch": 0.4427767354596623,
+      "grad_norm": 0.0872710794210434,
+      "learning_rate": 1.9259712524731083e-06,
+      "loss": 0.6877,
+      "step": 531
+    },
+    {
+      "epoch": 0.4436105899520534,
+      "grad_norm": 0.0860145092010498,
+      "learning_rate": 1.9256316605060813e-06,
+      "loss": 0.6458,
+      "step": 532
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.08691411465406418,
+      "learning_rate": 1.9252913214942456e-06,
+      "loss": 0.6286,
+      "step": 533
+    },
+    {
+      "epoch": 0.4452782989368355,
+      "grad_norm": 0.0899488553404808,
+      "learning_rate": 1.9249502357122757e-06,
+      "loss": 0.6154,
+      "step": 534
+    },
+    {
+      "epoch": 0.4461121534292266,
+      "grad_norm": 0.08670518547296524,
+      "learning_rate": 1.924608403435451e-06,
+      "loss": 0.6189,
+      "step": 535
+    },
+    {
+      "epoch": 0.4469460079216177,
+      "grad_norm": 0.08927959948778152,
+      "learning_rate": 1.9242658249396514e-06,
+      "loss": 0.6591,
+      "step": 536
+    },
+    {
+      "epoch": 0.44777986241400874,
+      "grad_norm": 0.09092804044485092,
+      "learning_rate": 1.9239225005013607e-06,
+      "loss": 0.6756,
+      "step": 537
+    },
+    {
+      "epoch": 0.44861371690639984,
+      "grad_norm": 0.08876843005418777,
+      "learning_rate": 1.923578430397664e-06,
+      "loss": 0.6221,
+      "step": 538
+    },
+    {
+      "epoch": 0.44944757139879093,
+      "grad_norm": 0.09023214131593704,
+      "learning_rate": 1.923233614906248e-06,
+      "loss": 0.6478,
+      "step": 539
+    },
+    {
+      "epoch": 0.450281425891182,
+      "grad_norm": 0.09291035681962967,
+      "learning_rate": 1.9228880543054006e-06,
+      "loss": 0.6568,
+      "step": 540
+    },
+    {
+      "epoch": 0.45111528038357307,
+      "grad_norm": 0.09169748425483704,
+      "learning_rate": 1.9225417488740127e-06,
+      "loss": 0.6966,
+      "step": 541
+    },
+    {
+      "epoch": 0.45194913487596416,
+      "grad_norm": 0.0960812196135521,
+      "learning_rate": 1.922194698891574e-06,
+      "loss": 0.6426,
+      "step": 542
+    },
+    {
+      "epoch": 0.4527829893683552,
+      "grad_norm": 0.09381034225225449,
+      "learning_rate": 1.9218469046381778e-06,
+      "loss": 0.7138,
+      "step": 543
+    },
+    {
+      "epoch": 0.4536168438607463,
+      "grad_norm": 0.09605712443590164,
+      "learning_rate": 1.9214983663945157e-06,
+      "loss": 0.6677,
+      "step": 544
+    },
+    {
+      "epoch": 0.4544506983531374,
+      "grad_norm": 0.09246627241373062,
+      "learning_rate": 1.921149084441881e-06,
+      "loss": 0.5995,
+      "step": 545
+    },
+    {
+      "epoch": 0.45528455284552843,
+      "grad_norm": 0.09615227580070496,
+      "learning_rate": 1.9207990590621663e-06,
+      "loss": 0.6543,
+      "step": 546
+    },
+    {
+      "epoch": 0.45611840733791953,
+      "grad_norm": 0.09620420634746552,
+      "learning_rate": 1.9204482905378654e-06,
+      "loss": 0.5895,
+      "step": 547
+    },
+    {
+      "epoch": 0.4569522618303106,
+      "grad_norm": 0.08878765255212784,
+      "learning_rate": 1.920096779152071e-06,
+      "loss": 0.599,
+      "step": 548
+    },
+    {
+      "epoch": 0.45778611632270166,
+      "grad_norm": 0.09067462384700775,
+      "learning_rate": 1.9197445251884763e-06,
+      "loss": 0.6897,
+      "step": 549
+    },
+    {
+      "epoch": 0.45861997081509276,
+      "grad_norm": 0.09180627018213272,
+      "learning_rate": 1.9193915289313724e-06,
+      "loss": 0.6647,
+      "step": 550
+    },
+    {
+      "epoch": 0.45945382530748385,
+      "grad_norm": 0.08824802190065384,
+      "learning_rate": 1.91903779066565e-06,
+      "loss": 0.6376,
+      "step": 551
+    },
+    {
+      "epoch": 0.46028767979987495,
+      "grad_norm": 0.08835819363594055,
+      "learning_rate": 1.918683310676799e-06,
+      "loss": 0.6922,
+      "step": 552
+    },
+    {
+      "epoch": 0.461121534292266,
+      "grad_norm": 0.09096652269363403,
+      "learning_rate": 1.918328089250908e-06,
+      "loss": 0.7169,
+      "step": 553
+    },
+    {
+      "epoch": 0.4619553887846571,
+      "grad_norm": 0.09415201842784882,
+      "learning_rate": 1.917972126674664e-06,
+      "loss": 0.659,
+      "step": 554
+    },
+    {
+      "epoch": 0.4627892432770482,
+      "grad_norm": 0.08964891731739044,
+      "learning_rate": 1.9176154232353513e-06,
+      "loss": 0.6869,
+      "step": 555
+    },
+    {
+      "epoch": 0.4636230977694392,
+      "grad_norm": 0.08688930422067642,
+      "learning_rate": 1.917257979220853e-06,
+      "loss": 0.627,
+      "step": 556
+    },
+    {
+      "epoch": 0.4644569522618303,
+      "grad_norm": 0.09115028381347656,
+      "learning_rate": 1.9168997949196496e-06,
+      "loss": 0.64,
+      "step": 557
+    },
+    {
+      "epoch": 0.4652908067542214,
+      "grad_norm": 0.09430664777755737,
+      "learning_rate": 1.9165408706208184e-06,
+      "loss": 0.6646,
+      "step": 558
+    },
+    {
+      "epoch": 0.46612466124661245,
+      "grad_norm": 0.09062688797712326,
+      "learning_rate": 1.916181206614036e-06,
+      "loss": 0.6392,
+      "step": 559
+    },
+    {
+      "epoch": 0.46695851573900354,
+      "grad_norm": 0.09099205583333969,
+      "learning_rate": 1.9158208031895737e-06,
+      "loss": 0.7133,
+      "step": 560
+    },
+    {
+      "epoch": 0.46779237023139464,
+      "grad_norm": 0.09944413602352142,
+      "learning_rate": 1.9154596606383002e-06,
+      "loss": 0.6958,
+      "step": 561
+    },
+    {
+      "epoch": 0.4686262247237857,
+      "grad_norm": 0.09643740206956863,
+      "learning_rate": 1.9150977792516816e-06,
+      "loss": 0.6813,
+      "step": 562
+    },
+    {
+      "epoch": 0.4694600792161768,
+      "grad_norm": 0.09532240033149719,
+      "learning_rate": 1.914735159321779e-06,
+      "loss": 0.6528,
+      "step": 563
+    },
+    {
+      "epoch": 0.47029393370856787,
+      "grad_norm": 0.08927876502275467,
+      "learning_rate": 1.914371801141251e-06,
+      "loss": 0.639,
+      "step": 564
+    },
+    {
+      "epoch": 0.4711277882009589,
+      "grad_norm": 0.09561195224523544,
+      "learning_rate": 1.914007705003351e-06,
+      "loss": 0.6795,
+      "step": 565
+    },
+    {
+      "epoch": 0.47196164269335,
+      "grad_norm": 0.0882195308804512,
+      "learning_rate": 1.9136428712019275e-06,
+      "loss": 0.6437,
+      "step": 566
+    },
+    {
+      "epoch": 0.4727954971857411,
+      "grad_norm": 0.08940508961677551,
+      "learning_rate": 1.913277300031426e-06,
+      "loss": 0.6485,
+      "step": 567
+    },
+    {
+      "epoch": 0.4736293516781322,
+      "grad_norm": 0.0933694839477539,
+      "learning_rate": 1.912910991786886e-06,
+      "loss": 0.6876,
+      "step": 568
+    },
+    {
+      "epoch": 0.47446320617052323,
+      "grad_norm": 0.09212841838598251,
+      "learning_rate": 1.9125439467639414e-06,
+      "loss": 0.5669,
+      "step": 569
+    },
+    {
+      "epoch": 0.47529706066291433,
+      "grad_norm": 0.09037458896636963,
+      "learning_rate": 1.9121761652588214e-06,
+      "loss": 0.6489,
+      "step": 570
+    },
+    {
+      "epoch": 0.4761309151553054,
+      "grad_norm": 0.08847213536500931,
+      "learning_rate": 1.9118076475683506e-06,
+      "loss": 0.6456,
+      "step": 571
+    },
+    {
+      "epoch": 0.47696476964769646,
+      "grad_norm": 0.08496098965406418,
+      "learning_rate": 1.9114383939899455e-06,
+      "loss": 0.6174,
+      "step": 572
+    },
+    {
+      "epoch": 0.47779862414008756,
+      "grad_norm": 0.09430284053087234,
+      "learning_rate": 1.911068404821618e-06,
+      "loss": 0.7102,
+      "step": 573
+    },
+    {
+      "epoch": 0.47863247863247865,
+      "grad_norm": 0.09211906790733337,
+      "learning_rate": 1.910697680361974e-06,
+      "loss": 0.6161,
+      "step": 574
+    },
+    {
+      "epoch": 0.4794663331248697,
+      "grad_norm": 0.09092090278863907,
+      "learning_rate": 1.910326220910211e-06,
+      "loss": 0.6595,
+      "step": 575
+    },
+    {
+      "epoch": 0.4803001876172608,
+      "grad_norm": 0.09773701429367065,
+      "learning_rate": 1.909954026766122e-06,
+      "loss": 0.6303,
+      "step": 576
+    },
+    {
+      "epoch": 0.4811340421096519,
+      "grad_norm": 0.09343478083610535,
+      "learning_rate": 1.909581098230091e-06,
+      "loss": 0.6578,
+      "step": 577
+    },
+    {
+      "epoch": 0.4819678966020429,
+      "grad_norm": 0.09686136245727539,
+      "learning_rate": 1.9092074356030966e-06,
+      "loss": 0.6504,
+      "step": 578
+    },
+    {
+      "epoch": 0.482801751094434,
+      "grad_norm": 0.09207551926374435,
+      "learning_rate": 1.9088330391867076e-06,
+      "loss": 0.6092,
+      "step": 579
+    },
+    {
+      "epoch": 0.4836356055868251,
+      "grad_norm": 0.093394935131073,
+      "learning_rate": 1.908457909283087e-06,
+      "loss": 0.645,
+      "step": 580
+    },
+    {
+      "epoch": 0.48446946007921615,
+      "grad_norm": 0.10280878096818924,
+      "learning_rate": 1.9080820461949886e-06,
+      "loss": 0.7288,
+      "step": 581
+    },
+    {
+      "epoch": 0.48530331457160725,
+      "grad_norm": 0.09714383631944656,
+      "learning_rate": 1.9077054502257585e-06,
+      "loss": 0.6213,
+      "step": 582
+    },
+    {
+      "epoch": 0.48613716906399834,
+      "grad_norm": 0.09004565328359604,
+      "learning_rate": 1.9073281216793337e-06,
+      "loss": 0.6459,
+      "step": 583
+    },
+    {
+      "epoch": 0.48697102355638944,
+      "grad_norm": 0.12440581619739532,
+      "learning_rate": 1.906950060860243e-06,
+      "loss": 0.6823,
+      "step": 584
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 0.09314379096031189,
+      "learning_rate": 1.9065712680736066e-06,
+      "loss": 0.6475,
+      "step": 585
+    },
+    {
+      "epoch": 0.48863873254117157,
+      "grad_norm": 0.09512604027986526,
+      "learning_rate": 1.906191743625134e-06,
+      "loss": 0.5932,
+      "step": 586
+    },
+    {
+      "epoch": 0.48947258703356267,
+      "grad_norm": 0.08977750688791275,
+      "learning_rate": 1.9058114878211265e-06,
+      "loss": 0.6546,
+      "step": 587
+    },
+    {
+      "epoch": 0.4903064415259537,
+      "grad_norm": 0.09166161715984344,
+      "learning_rate": 1.905430500968475e-06,
+      "loss": 0.6772,
+      "step": 588
+    },
+    {
+      "epoch": 0.4911402960183448,
+      "grad_norm": 0.10014794021844864,
+      "learning_rate": 1.9050487833746609e-06,
+      "loss": 0.6945,
+      "step": 589
+    },
+    {
+      "epoch": 0.4919741505107359,
+      "grad_norm": 0.09376100450754166,
+      "learning_rate": 1.9046663353477548e-06,
+      "loss": 0.6581,
+      "step": 590
+    },
+    {
+      "epoch": 0.49280800500312694,
+      "grad_norm": 0.09505487233400345,
+      "learning_rate": 1.9042831571964171e-06,
+      "loss": 0.6785,
+      "step": 591
+    },
+    {
+      "epoch": 0.49364185949551803,
+      "grad_norm": 0.09371201694011688,
+      "learning_rate": 1.9038992492298975e-06,
+      "loss": 0.6331,
+      "step": 592
+    },
+    {
+      "epoch": 0.4944757139879091,
+      "grad_norm": 0.09442009776830673,
+      "learning_rate": 1.9035146117580348e-06,
+      "loss": 0.5728,
+      "step": 593
+    },
+    {
+      "epoch": 0.49530956848030017,
+      "grad_norm": 0.09772226214408875,
+      "learning_rate": 1.903129245091256e-06,
+      "loss": 0.6635,
+      "step": 594
+    },
+    {
+      "epoch": 0.49614342297269126,
+      "grad_norm": 0.09484317898750305,
+      "learning_rate": 1.9027431495405776e-06,
+      "loss": 0.6327,
+      "step": 595
+    },
+    {
+      "epoch": 0.49697727746508236,
+      "grad_norm": 0.09658980369567871,
+      "learning_rate": 1.9023563254176032e-06,
+      "loss": 0.6608,
+      "step": 596
+    },
+    {
+      "epoch": 0.4978111319574734,
+      "grad_norm": 0.09717069566249847,
+      "learning_rate": 1.9019687730345251e-06,
+      "loss": 0.6605,
+      "step": 597
+    },
+    {
+      "epoch": 0.4986449864498645,
+      "grad_norm": 0.09019803255796432,
+      "learning_rate": 1.9015804927041233e-06,
+      "loss": 0.6867,
+      "step": 598
+    },
+    {
+      "epoch": 0.4994788409422556,
+      "grad_norm": 0.09377589821815491,
+      "learning_rate": 1.9011914847397653e-06,
+      "loss": 0.6419,
+      "step": 599
+    },
+    {
+      "epoch": 0.5003126954346466,
+      "grad_norm": 0.09151824563741684,
+      "learning_rate": 1.9008017494554055e-06,
+      "loss": 0.7184,
+      "step": 600
+    },
+    {
+      "epoch": 0.5011465499270378,
+      "grad_norm": 0.0963352620601654,
+      "learning_rate": 1.900411287165586e-06,
+      "loss": 0.6897,
+      "step": 601
+    },
+    {
+      "epoch": 0.5019804044194288,
+      "grad_norm": 0.08974531292915344,
+      "learning_rate": 1.9000200981854347e-06,
+      "loss": 0.5847,
+      "step": 602
+    },
+    {
+      "epoch": 0.5028142589118199,
+      "grad_norm": 0.09220883995294571,
+      "learning_rate": 1.8996281828306665e-06,
+      "loss": 0.6652,
+      "step": 603
+    },
+    {
+      "epoch": 0.503648113404211,
+      "grad_norm": 0.09491855651140213,
+      "learning_rate": 1.899235541417583e-06,
+      "loss": 0.595,
+      "step": 604
+    },
+    {
+      "epoch": 0.504481967896602,
+      "grad_norm": 0.09616294503211975,
+      "learning_rate": 1.898842174263071e-06,
+      "loss": 0.6378,
+      "step": 605
+    },
+    {
+      "epoch": 0.5053158223889931,
+      "grad_norm": 0.08942185342311859,
+      "learning_rate": 1.8984480816846035e-06,
+      "loss": 0.6459,
+      "step": 606
+    },
+    {
+      "epoch": 0.5061496768813842,
+      "grad_norm": 0.09059803187847137,
+      "learning_rate": 1.8980532640002388e-06,
+      "loss": 0.6019,
+      "step": 607
+    },
+    {
+      "epoch": 0.5069835313737753,
+      "grad_norm": 0.09913185238838196,
+      "learning_rate": 1.89765772152862e-06,
+      "loss": 0.669,
+      "step": 608
+    },
+    {
+      "epoch": 0.5078173858661663,
+      "grad_norm": 0.10223888605833054,
+      "learning_rate": 1.8972614545889756e-06,
+      "loss": 0.6454,
+      "step": 609
+    },
+    {
+      "epoch": 0.5086512403585575,
+      "grad_norm": 0.09271295368671417,
+      "learning_rate": 1.896864463501119e-06,
+      "loss": 0.6467,
+      "step": 610
+    },
+    {
+      "epoch": 0.5094850948509485,
+      "grad_norm": 0.08862827718257904,
+      "learning_rate": 1.896466748585448e-06,
+      "loss": 0.6343,
+      "step": 611
+    },
+    {
+      "epoch": 0.5103189493433395,
+      "grad_norm": 0.09303930401802063,
+      "learning_rate": 1.8960683101629435e-06,
+      "loss": 0.6389,
+      "step": 612
+    },
+    {
+      "epoch": 0.5111528038357307,
+      "grad_norm": 0.09331949055194855,
+      "learning_rate": 1.8956691485551719e-06,
+      "loss": 0.6507,
+      "step": 613
+    },
+    {
+      "epoch": 0.5119866583281217,
+      "grad_norm": 0.0993635281920433,
+      "learning_rate": 1.8952692640842825e-06,
+      "loss": 0.63,
+      "step": 614
+    },
+    {
+      "epoch": 0.5128205128205128,
+      "grad_norm": 0.09348301589488983,
+      "learning_rate": 1.8948686570730074e-06,
+      "loss": 0.676,
+      "step": 615
+    },
+    {
+      "epoch": 0.5136543673129039,
+      "grad_norm": 0.0950765460729599,
+      "learning_rate": 1.894467327844663e-06,
+      "loss": 0.7035,
+      "step": 616
+    },
+    {
+      "epoch": 0.514488221805295,
+      "grad_norm": 0.08676007390022278,
+      "learning_rate": 1.8940652767231483e-06,
+      "loss": 0.6067,
+      "step": 617
+    },
+    {
+      "epoch": 0.515322076297686,
+      "grad_norm": 0.09285329282283783,
+      "learning_rate": 1.8936625040329438e-06,
+      "loss": 0.6441,
+      "step": 618
+    },
+    {
+      "epoch": 0.5161559307900772,
+      "grad_norm": 0.09361077845096588,
+      "learning_rate": 1.8932590100991136e-06,
+      "loss": 0.6767,
+      "step": 619
+    },
+    {
+      "epoch": 0.5169897852824682,
+      "grad_norm": 0.09173277020454407,
+      "learning_rate": 1.8928547952473033e-06,
+      "loss": 0.6395,
+      "step": 620
+    },
+    {
+      "epoch": 0.5178236397748592,
+      "grad_norm": 0.09198316186666489,
+      "learning_rate": 1.892449859803741e-06,
+      "loss": 0.7006,
+      "step": 621
+    },
+    {
+      "epoch": 0.5186574942672504,
+      "grad_norm": 0.09348779171705246,
+      "learning_rate": 1.8920442040952356e-06,
+      "loss": 0.6641,
+      "step": 622
+    },
+    {
+      "epoch": 0.5194913487596414,
+      "grad_norm": 0.08798382431268692,
+      "learning_rate": 1.8916378284491774e-06,
+      "loss": 0.6478,
+      "step": 623
+    },
+    {
+      "epoch": 0.5203252032520326,
+      "grad_norm": 0.09442038089036942,
+      "learning_rate": 1.8912307331935383e-06,
+      "loss": 0.6201,
+      "step": 624
+    },
+    {
+      "epoch": 0.5211590577444236,
+      "grad_norm": 0.09519796818494797,
+      "learning_rate": 1.8908229186568705e-06,
+      "loss": 0.5917,
+      "step": 625
+    },
+    {
+      "epoch": 0.5219929122368147,
+      "grad_norm": 0.09593921154737473,
+      "learning_rate": 1.8904143851683062e-06,
+      "loss": 0.6859,
+      "step": 626
+    },
+    {
+      "epoch": 0.5228267667292058,
+      "grad_norm": 0.09510339796543121,
+      "learning_rate": 1.890005133057559e-06,
+      "loss": 0.6679,
+      "step": 627
+    },
+    {
+      "epoch": 0.5236606212215968,
+      "grad_norm": 0.09473798424005508,
+      "learning_rate": 1.8895951626549222e-06,
+      "loss": 0.6108,
+      "step": 628
+    },
+    {
+      "epoch": 0.5244944757139879,
+      "grad_norm": 0.10544616729021072,
+      "learning_rate": 1.8891844742912678e-06,
+      "loss": 0.6007,
+      "step": 629
+    },
+    {
+      "epoch": 0.525328330206379,
+      "grad_norm": 0.09652596712112427,
+      "learning_rate": 1.8887730682980482e-06,
+      "loss": 0.6858,
+      "step": 630
+    },
+    {
+      "epoch": 0.5261621846987701,
+      "grad_norm": 0.09385030716657639,
+      "learning_rate": 1.888360945007295e-06,
+      "loss": 0.692,
+      "step": 631
+    },
+    {
+      "epoch": 0.5269960391911611,
+      "grad_norm": 0.10086268931627274,
+      "learning_rate": 1.887948104751618e-06,
+      "loss": 0.7302,
+      "step": 632
+    },
+    {
+      "epoch": 0.5278298936835523,
+      "grad_norm": 0.09164178371429443,
+      "learning_rate": 1.8875345478642065e-06,
+      "loss": 0.5793,
+      "step": 633
+    },
+    {
+      "epoch": 0.5286637481759433,
+      "grad_norm": 0.08686941117048264,
+      "learning_rate": 1.8871202746788274e-06,
+      "loss": 0.6304,
+      "step": 634
+    },
+    {
+      "epoch": 0.5294976026683343,
+      "grad_norm": 0.09579890966415405,
+      "learning_rate": 1.8867052855298264e-06,
+      "loss": 0.6663,
+      "step": 635
+    },
+    {
+      "epoch": 0.5303314571607255,
+      "grad_norm": 0.09054259210824966,
+      "learning_rate": 1.8862895807521264e-06,
+      "loss": 0.6395,
+      "step": 636
+    },
+    {
+      "epoch": 0.5311653116531165,
+      "grad_norm": 0.0913679376244545,
+      "learning_rate": 1.8858731606812284e-06,
+      "loss": 0.6481,
+      "step": 637
+    },
+    {
+      "epoch": 0.5319991661455076,
+      "grad_norm": 0.09573955833911896,
+      "learning_rate": 1.8854560256532098e-06,
+      "loss": 0.6543,
+      "step": 638
+    },
+    {
+      "epoch": 0.5328330206378987,
+      "grad_norm": 0.09667269140481949,
+      "learning_rate": 1.8850381760047262e-06,
+      "loss": 0.6532,
+      "step": 639
+    },
+    {
+      "epoch": 0.5336668751302898,
+      "grad_norm": 0.09536033123731613,
+      "learning_rate": 1.8846196120730093e-06,
+      "loss": 0.6429,
+      "step": 640
+    },
+    {
+      "epoch": 0.5345007296226808,
+      "grad_norm": 0.09058564901351929,
+      "learning_rate": 1.8842003341958673e-06,
+      "loss": 0.579,
+      "step": 641
+    },
+    {
+      "epoch": 0.535334584115072,
+      "grad_norm": 0.09930071234703064,
+      "learning_rate": 1.8837803427116844e-06,
+      "loss": 0.5868,
+      "step": 642
+    },
+    {
+      "epoch": 0.536168438607463,
+      "grad_norm": 0.0978417843580246,
+      "learning_rate": 1.8833596379594211e-06,
+      "loss": 0.6822,
+      "step": 643
+    },
+    {
+      "epoch": 0.537002293099854,
+      "grad_norm": 0.09346287697553635,
+      "learning_rate": 1.882938220278614e-06,
+      "loss": 0.6116,
+      "step": 644
+    },
+    {
+      "epoch": 0.5378361475922452,
+      "grad_norm": 0.09151905030012131,
+      "learning_rate": 1.8825160900093738e-06,
+      "loss": 0.6496,
+      "step": 645
+    },
+    {
+      "epoch": 0.5386700020846362,
+      "grad_norm": 0.0970907062292099,
+      "learning_rate": 1.8820932474923871e-06,
+      "loss": 0.6504,
+      "step": 646
+    },
+    {
+      "epoch": 0.5395038565770273,
+      "grad_norm": 0.09408881515264511,
+      "learning_rate": 1.8816696930689159e-06,
+      "loss": 0.6829,
+      "step": 647
+    },
+    {
+      "epoch": 0.5403377110694184,
+      "grad_norm": 0.09966862946748734,
+      "learning_rate": 1.8812454270807951e-06,
+      "loss": 0.5898,
+      "step": 648
+    },
+    {
+      "epoch": 0.5411715655618095,
+      "grad_norm": 0.0900130420923233,
+      "learning_rate": 1.8808204498704356e-06,
+      "loss": 0.6042,
+      "step": 649
+    },
+    {
+      "epoch": 0.5420054200542005,
+      "grad_norm": 0.09694640338420868,
+      "learning_rate": 1.8803947617808214e-06,
+      "loss": 0.6727,
+      "step": 650
+    },
+    {
+      "epoch": 0.5428392745465916,
+      "grad_norm": 0.09688873589038849,
+      "learning_rate": 1.8799683631555103e-06,
+      "loss": 0.6137,
+      "step": 651
+    },
+    {
+      "epoch": 0.5436731290389827,
+      "grad_norm": 0.09676310420036316,
+      "learning_rate": 1.8795412543386337e-06,
+      "loss": 0.612,
+      "step": 652
+    },
+    {
+      "epoch": 0.5445069835313737,
+      "grad_norm": 0.09581848233938217,
+      "learning_rate": 1.8791134356748962e-06,
+      "loss": 0.6632,
+      "step": 653
+    },
+    {
+      "epoch": 0.5453408380237649,
+      "grad_norm": 0.099350206553936,
+      "learning_rate": 1.8786849075095753e-06,
+      "loss": 0.6652,
+      "step": 654
+    },
+    {
+      "epoch": 0.5461746925161559,
+      "grad_norm": 0.09312117099761963,
+      "learning_rate": 1.878255670188521e-06,
+      "loss": 0.6184,
+      "step": 655
+    },
+    {
+      "epoch": 0.5470085470085471,
+      "grad_norm": 0.09312383830547333,
+      "learning_rate": 1.8778257240581554e-06,
+      "loss": 0.6595,
+      "step": 656
+    },
+    {
+      "epoch": 0.5478424015009381,
+      "grad_norm": 0.09485237300395966,
+      "learning_rate": 1.8773950694654732e-06,
+      "loss": 0.716,
+      "step": 657
+    },
+    {
+      "epoch": 0.5486762559933291,
+      "grad_norm": 0.09585438668727875,
+      "learning_rate": 1.8769637067580405e-06,
+      "loss": 0.7058,
+      "step": 658
+    },
+    {
+      "epoch": 0.5495101104857203,
+      "grad_norm": 0.09024535119533539,
+      "learning_rate": 1.876531636283995e-06,
+      "loss": 0.594,
+      "step": 659
+    },
+    {
+      "epoch": 0.5503439649781113,
+      "grad_norm": 0.09285992383956909,
+      "learning_rate": 1.8760988583920456e-06,
+      "loss": 0.648,
+      "step": 660
+    },
+    {
+      "epoch": 0.5511778194705024,
+      "grad_norm": 0.08939266204833984,
+      "learning_rate": 1.8756653734314722e-06,
+      "loss": 0.5866,
+      "step": 661
+    },
+    {
+      "epoch": 0.5520116739628935,
+      "grad_norm": 0.10212475061416626,
+      "learning_rate": 1.8752311817521254e-06,
+      "loss": 0.5485,
+      "step": 662
+    },
+    {
+      "epoch": 0.5528455284552846,
+      "grad_norm": 0.0938698872923851,
+      "learning_rate": 1.8747962837044256e-06,
+      "loss": 0.692,
+      "step": 663
+    },
+    {
+      "epoch": 0.5536793829476756,
+      "grad_norm": 0.09794861823320389,
+      "learning_rate": 1.8743606796393644e-06,
+      "loss": 0.6449,
+      "step": 664
+    },
+    {
+      "epoch": 0.5545132374400668,
+      "grad_norm": 0.09574401378631592,
+      "learning_rate": 1.8739243699085019e-06,
+      "loss": 0.5878,
+      "step": 665
+    },
+    {
+      "epoch": 0.5553470919324578,
+      "grad_norm": 0.10169863700866699,
+      "learning_rate": 1.8734873548639687e-06,
+      "loss": 0.6968,
+      "step": 666
+    },
+    {
+      "epoch": 0.5561809464248488,
+      "grad_norm": 0.09560302644968033,
+      "learning_rate": 1.8730496348584642e-06,
+      "loss": 0.6448,
+      "step": 667
+    },
+    {
+      "epoch": 0.55701480091724,
+      "grad_norm": 0.10035623610019684,
+      "learning_rate": 1.8726112102452572e-06,
+      "loss": 0.6893,
+      "step": 668
+    },
+    {
+      "epoch": 0.557848655409631,
+      "grad_norm": 0.093927301466465,
+      "learning_rate": 1.872172081378184e-06,
+      "loss": 0.6541,
+      "step": 669
+    },
+    {
+      "epoch": 0.5586825099020221,
+      "grad_norm": 0.0979999378323555,
+      "learning_rate": 1.8717322486116508e-06,
+      "loss": 0.6651,
+      "step": 670
+    },
+    {
+      "epoch": 0.5595163643944132,
+      "grad_norm": 0.10020069032907486,
+      "learning_rate": 1.8712917123006312e-06,
+      "loss": 0.6289,
+      "step": 671
+    },
+    {
+      "epoch": 0.5603502188868043,
+      "grad_norm": 0.10121606290340424,
+      "learning_rate": 1.8708504728006664e-06,
+      "loss": 0.6547,
+      "step": 672
+    },
+    {
+      "epoch": 0.5611840733791953,
+      "grad_norm": 0.09832029789686203,
+      "learning_rate": 1.870408530467865e-06,
+      "loss": 0.651,
+      "step": 673
+    },
+    {
+      "epoch": 0.5620179278715864,
+      "grad_norm": 0.0978146642446518,
+      "learning_rate": 1.8699658856589037e-06,
+      "loss": 0.6453,
+      "step": 674
+    },
+    {
+      "epoch": 0.5628517823639775,
+      "grad_norm": 0.10008279979228973,
+      "learning_rate": 1.8695225387310249e-06,
+      "loss": 0.6945,
+      "step": 675
+    },
+    {
+      "epoch": 0.5636856368563685,
+      "grad_norm": 0.09760496020317078,
+      "learning_rate": 1.8690784900420388e-06,
+      "loss": 0.6222,
+      "step": 676
+    },
+    {
+      "epoch": 0.5645194913487597,
+      "grad_norm": 0.09725689142942429,
+      "learning_rate": 1.8686337399503215e-06,
+      "loss": 0.6464,
+      "step": 677
+    },
+    {
+      "epoch": 0.5653533458411507,
+      "grad_norm": 0.09347565472126007,
+      "learning_rate": 1.8681882888148152e-06,
+      "loss": 0.6282,
+      "step": 678
+    },
+    {
+      "epoch": 0.5661872003335418,
+      "grad_norm": 0.09704628586769104,
+      "learning_rate": 1.867742136995028e-06,
+      "loss": 0.6529,
+      "step": 679
+    },
+    {
+      "epoch": 0.5670210548259329,
+      "grad_norm": 0.09918136149644852,
+      "learning_rate": 1.8672952848510328e-06,
+      "loss": 0.739,
+      "step": 680
+    },
+    {
+      "epoch": 0.5678549093183239,
+      "grad_norm": 0.09772541373968124,
+      "learning_rate": 1.8668477327434686e-06,
+      "loss": 0.6383,
+      "step": 681
+    },
+    {
+      "epoch": 0.568688763810715,
+      "grad_norm": 0.09866555780172348,
+      "learning_rate": 1.866399481033539e-06,
+      "loss": 0.6721,
+      "step": 682
+    },
+    {
+      "epoch": 0.5695226183031061,
+      "grad_norm": 0.09793423861265182,
+      "learning_rate": 1.8659505300830123e-06,
+      "loss": 0.6055,
+      "step": 683
+    },
+    {
+      "epoch": 0.5703564727954972,
+      "grad_norm": 0.10061061382293701,
+      "learning_rate": 1.865500880254221e-06,
+      "loss": 0.6544,
+      "step": 684
+    },
+    {
+      "epoch": 0.5711903272878882,
+      "grad_norm": 0.09208130091428757,
+      "learning_rate": 1.8650505319100617e-06,
+      "loss": 0.6014,
+      "step": 685
+    },
+    {
+      "epoch": 0.5720241817802794,
+      "grad_norm": 0.0995427742600441,
+      "learning_rate": 1.8645994854139948e-06,
+      "loss": 0.6965,
+      "step": 686
+    },
+    {
+      "epoch": 0.5728580362726704,
+      "grad_norm": 0.09244947880506516,
+      "learning_rate": 1.8641477411300438e-06,
+      "loss": 0.6272,
+      "step": 687
+    },
+    {
+      "epoch": 0.5736918907650616,
+      "grad_norm": 0.09937774389982224,
+      "learning_rate": 1.863695299422796e-06,
+      "loss": 0.6648,
+      "step": 688
+    },
+    {
+      "epoch": 0.5745257452574526,
+      "grad_norm": 0.09111293405294418,
+      "learning_rate": 1.8632421606574007e-06,
+      "loss": 0.5476,
+      "step": 689
+    },
+    {
+      "epoch": 0.5753595997498436,
+      "grad_norm": 0.10068730264902115,
+      "learning_rate": 1.8627883251995708e-06,
+      "loss": 0.6535,
+      "step": 690
+    },
+    {
+      "epoch": 0.5761934542422348,
+      "grad_norm": 0.09240923076868057,
+      "learning_rate": 1.862333793415581e-06,
+      "loss": 0.575,
+      "step": 691
+    },
+    {
+      "epoch": 0.5770273087346258,
+      "grad_norm": 0.09765233844518661,
+      "learning_rate": 1.8618785656722676e-06,
+      "loss": 0.6323,
+      "step": 692
+    },
+    {
+      "epoch": 0.5778611632270169,
+      "grad_norm": 0.09996625781059265,
+      "learning_rate": 1.861422642337029e-06,
+      "loss": 0.639,
+      "step": 693
+    },
+    {
+      "epoch": 0.578695017719408,
+      "grad_norm": 0.10148533433675766,
+      "learning_rate": 1.860966023777825e-06,
+      "loss": 0.7196,
+      "step": 694
+    },
+    {
+      "epoch": 0.579528872211799,
+      "grad_norm": 0.09370430558919907,
+      "learning_rate": 1.8605087103631764e-06,
+      "loss": 0.6385,
+      "step": 695
+    },
+    {
+      "epoch": 0.5803627267041901,
+      "grad_norm": 0.09228570014238358,
+      "learning_rate": 1.8600507024621648e-06,
+      "loss": 0.6242,
+      "step": 696
+    },
+    {
+      "epoch": 0.5811965811965812,
+      "grad_norm": 0.09689746052026749,
+      "learning_rate": 1.8595920004444324e-06,
+      "loss": 0.6135,
+      "step": 697
+    },
+    {
+      "epoch": 0.5820304356889723,
+      "grad_norm": 0.09802035987377167,
+      "learning_rate": 1.8591326046801812e-06,
+      "loss": 0.5789,
+      "step": 698
+    },
+    {
+      "epoch": 0.5828642901813633,
+      "grad_norm": 0.10172217339277267,
+      "learning_rate": 1.8586725155401735e-06,
+      "loss": 0.5921,
+      "step": 699
+    },
+    {
+      "epoch": 0.5836981446737545,
+      "grad_norm": 0.10300835222005844,
+      "learning_rate": 1.858211733395731e-06,
+      "loss": 0.6647,
+      "step": 700
+    },
+    {
+      "epoch": 0.5845319991661455,
+      "grad_norm": 0.09017598628997803,
+      "learning_rate": 1.8577502586187353e-06,
+      "loss": 0.6409,
+      "step": 701
+    },
+    {
+      "epoch": 0.5853658536585366,
+      "grad_norm": 0.10058888792991638,
+      "learning_rate": 1.8572880915816259e-06,
+      "loss": 0.5902,
+      "step": 702
+    },
+    {
+      "epoch": 0.5861997081509277,
+      "grad_norm": 0.10125189274549484,
+      "learning_rate": 1.856825232657402e-06,
+      "loss": 0.6392,
+      "step": 703
+    },
+    {
+      "epoch": 0.5870335626433187,
+      "grad_norm": 0.0983993262052536,
+      "learning_rate": 1.8563616822196204e-06,
+      "loss": 0.5915,
+      "step": 704
+    },
+    {
+      "epoch": 0.5878674171357098,
+      "grad_norm": 0.09559126943349838,
+      "learning_rate": 1.8558974406423965e-06,
+      "loss": 0.6386,
+      "step": 705
+    },
+    {
+      "epoch": 0.5887012716281009,
+      "grad_norm": 0.09088745713233948,
+      "learning_rate": 1.8554325083004034e-06,
+      "loss": 0.6483,
+      "step": 706
+    },
+    {
+      "epoch": 0.589535126120492,
+      "grad_norm": 0.09516370296478271,
+      "learning_rate": 1.854966885568872e-06,
+      "loss": 0.6556,
+      "step": 707
+    },
+    {
+      "epoch": 0.590368980612883,
+      "grad_norm": 0.09569456428289413,
+      "learning_rate": 1.8545005728235896e-06,
+      "loss": 0.6394,
+      "step": 708
+    },
+    {
+      "epoch": 0.5912028351052742,
+      "grad_norm": 0.0980261042714119,
+      "learning_rate": 1.854033570440901e-06,
+      "loss": 0.6415,
+      "step": 709
+    },
+    {
+      "epoch": 0.5920366895976652,
+      "grad_norm": 0.09796703606843948,
+      "learning_rate": 1.8535658787977075e-06,
+      "loss": 0.6736,
+      "step": 710
+    },
+    {
+      "epoch": 0.5928705440900562,
+      "grad_norm": 0.09802737832069397,
+      "learning_rate": 1.8530974982714664e-06,
+      "loss": 0.5802,
+      "step": 711
+    },
+    {
+      "epoch": 0.5937043985824474,
+      "grad_norm": 0.09888147562742233,
+      "learning_rate": 1.8526284292401914e-06,
+      "loss": 0.6648,
+      "step": 712
+    },
+    {
+      "epoch": 0.5945382530748384,
+      "grad_norm": 0.10190586000680923,
+      "learning_rate": 1.8521586720824515e-06,
+      "loss": 0.7032,
+      "step": 713
+    },
+    {
+      "epoch": 0.5953721075672295,
+      "grad_norm": 0.09422854334115982,
+      "learning_rate": 1.8516882271773712e-06,
+      "loss": 0.6551,
+      "step": 714
+    },
+    {
+      "epoch": 0.5962059620596206,
+      "grad_norm": 0.09025522321462631,
+      "learning_rate": 1.8512170949046305e-06,
+      "loss": 0.6231,
+      "step": 715
+    },
+    {
+      "epoch": 0.5970398165520117,
+      "grad_norm": 0.102678582072258,
+      "learning_rate": 1.850745275644463e-06,
+      "loss": 0.6924,
+      "step": 716
+    },
+    {
+      "epoch": 0.5978736710444027,
+      "grad_norm": 0.09951562434434891,
+      "learning_rate": 1.8502727697776579e-06,
+      "loss": 0.5763,
+      "step": 717
+    },
+    {
+      "epoch": 0.5987075255367938,
+      "grad_norm": 0.10048934072256088,
+      "learning_rate": 1.8497995776855579e-06,
+      "loss": 0.6369,
+      "step": 718
+    },
+    {
+      "epoch": 0.5995413800291849,
+      "grad_norm": 0.11139318346977234,
+      "learning_rate": 1.84932569975006e-06,
+      "loss": 0.6086,
+      "step": 719
+    },
+    {
+      "epoch": 0.600375234521576,
+      "grad_norm": 0.10424143821001053,
+      "learning_rate": 1.8488511363536138e-06,
+      "loss": 0.6394,
+      "step": 720
+    },
+    {
+      "epoch": 0.6012090890139671,
+      "grad_norm": 0.09213969111442566,
+      "learning_rate": 1.8483758878792236e-06,
+      "loss": 0.6554,
+      "step": 721
+    },
+    {
+      "epoch": 0.6020429435063581,
+      "grad_norm": 0.0998111367225647,
+      "learning_rate": 1.8478999547104452e-06,
+      "loss": 0.6717,
+      "step": 722
+    },
+    {
+      "epoch": 0.6028767979987493,
+      "grad_norm": 0.09211792796850204,
+      "learning_rate": 1.8474233372313877e-06,
+      "loss": 0.5694,
+      "step": 723
+    },
+    {
+      "epoch": 0.6037106524911403,
+      "grad_norm": 0.10054522007703781,
+      "learning_rate": 1.8469460358267127e-06,
+      "loss": 0.6501,
+      "step": 724
+    },
+    {
+      "epoch": 0.6045445069835313,
+      "grad_norm": 0.09979543089866638,
+      "learning_rate": 1.8464680508816323e-06,
+      "loss": 0.6056,
+      "step": 725
+    },
+    {
+      "epoch": 0.6053783614759225,
+      "grad_norm": 0.1000729575753212,
+      "learning_rate": 1.8459893827819126e-06,
+      "loss": 0.6302,
+      "step": 726
+    },
+    {
+      "epoch": 0.6062122159683135,
+      "grad_norm": 0.09924089163541794,
+      "learning_rate": 1.8455100319138692e-06,
+      "loss": 0.6432,
+      "step": 727
+    },
+    {
+      "epoch": 0.6070460704607046,
+      "grad_norm": 0.0920461118221283,
+      "learning_rate": 1.8450299986643695e-06,
+      "loss": 0.6081,
+      "step": 728
+    },
+    {
+      "epoch": 0.6078799249530957,
+      "grad_norm": 0.10430511087179184,
+      "learning_rate": 1.8445492834208316e-06,
+      "loss": 0.6586,
+      "step": 729
+    },
+    {
+      "epoch": 0.6087137794454868,
+      "grad_norm": 0.09923563152551651,
+      "learning_rate": 1.8440678865712236e-06,
+      "loss": 0.6166,
+      "step": 730
+    },
+    {
+      "epoch": 0.6095476339378778,
+      "grad_norm": 0.1012277603149414,
+      "learning_rate": 1.843585808504064e-06,
+      "loss": 0.6843,
+      "step": 731
+    },
+    {
+      "epoch": 0.610381488430269,
+      "grad_norm": 0.09866297990083694,
+      "learning_rate": 1.8431030496084214e-06,
+      "loss": 0.6556,
+      "step": 732
+    },
+    {
+      "epoch": 0.61121534292266,
+      "grad_norm": 0.09874875843524933,
+      "learning_rate": 1.8426196102739132e-06,
+      "loss": 0.5993,
+      "step": 733
+    },
+    {
+      "epoch": 0.612049197415051,
+      "grad_norm": 0.09275837987661362,
+      "learning_rate": 1.8421354908907067e-06,
+      "loss": 0.6326,
+      "step": 734
+    },
+    {
+      "epoch": 0.6128830519074422,
+      "grad_norm": 0.09502032399177551,
+      "learning_rate": 1.8416506918495173e-06,
+      "loss": 0.6077,
+      "step": 735
+    },
+    {
+      "epoch": 0.6137169063998332,
+      "grad_norm": 0.09185368567705154,
+      "learning_rate": 1.8411652135416093e-06,
+      "loss": 0.6064,
+      "step": 736
+    },
+    {
+      "epoch": 0.6145507608922243,
+      "grad_norm": 0.09815992414951324,
+      "learning_rate": 1.8406790563587956e-06,
+      "loss": 0.6313,
+      "step": 737
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.09629546850919724,
+      "learning_rate": 1.8401922206934361e-06,
+      "loss": 0.6106,
+      "step": 738
+    },
+    {
+      "epoch": 0.6162184698770065,
+      "grad_norm": 0.09635171294212341,
+      "learning_rate": 1.839704706938439e-06,
+      "loss": 0.6523,
+      "step": 739
+    },
+    {
+      "epoch": 0.6170523243693975,
+      "grad_norm": 0.10179682075977325,
+      "learning_rate": 1.8392165154872595e-06,
+      "loss": 0.5873,
+      "step": 740
+    },
+    {
+      "epoch": 0.6178861788617886,
+      "grad_norm": 0.10270754247903824,
+      "learning_rate": 1.8387276467338996e-06,
+      "loss": 0.673,
+      "step": 741
+    },
+    {
+      "epoch": 0.6187200333541797,
+      "grad_norm": 0.09746810793876648,
+      "learning_rate": 1.8382381010729086e-06,
+      "loss": 0.6564,
+      "step": 742
+    },
+    {
+      "epoch": 0.6195538878465707,
+      "grad_norm": 0.0962262824177742,
+      "learning_rate": 1.8377478788993813e-06,
+      "loss": 0.5738,
+      "step": 743
+    },
+    {
+      "epoch": 0.6203877423389619,
+      "grad_norm": 0.09518938511610031,
+      "learning_rate": 1.8372569806089586e-06,
+      "loss": 0.6345,
+      "step": 744
+    },
+    {
+      "epoch": 0.6212215968313529,
+      "grad_norm": 0.09188467264175415,
+      "learning_rate": 1.8367654065978276e-06,
+      "loss": 0.6112,
+      "step": 745
+    },
+    {
+      "epoch": 0.622055451323744,
+      "grad_norm": 0.09746599197387695,
+      "learning_rate": 1.8362731572627202e-06,
+      "loss": 0.6349,
+      "step": 746
+    },
+    {
+      "epoch": 0.6228893058161351,
+      "grad_norm": 0.10009641200304031,
+      "learning_rate": 1.8357802330009136e-06,
+      "loss": 0.6502,
+      "step": 747
+    },
+    {
+      "epoch": 0.6237231603085261,
+      "grad_norm": 0.10499203950166702,
+      "learning_rate": 1.8352866342102296e-06,
+      "loss": 0.6812,
+      "step": 748
+    },
+    {
+      "epoch": 0.6245570148009172,
+      "grad_norm": 0.09757328033447266,
+      "learning_rate": 1.8347923612890349e-06,
+      "loss": 0.6025,
+      "step": 749
+    },
+    {
+      "epoch": 0.6253908692933083,
+      "grad_norm": 0.10616346448659897,
+      "learning_rate": 1.8342974146362394e-06,
+      "loss": 0.6638,
+      "step": 750
+    },
+    {
+      "epoch": 0.6262247237856994,
+      "grad_norm": 0.10131774097681046,
+      "learning_rate": 1.8338017946512968e-06,
+      "loss": 0.5456,
+      "step": 751
+    },
+    {
+      "epoch": 0.6270585782780905,
+      "grad_norm": 0.10244927555322647,
+      "learning_rate": 1.8333055017342054e-06,
+      "loss": 0.6751,
+      "step": 752
+    },
+    {
+      "epoch": 0.6278924327704816,
+      "grad_norm": 0.10338141769170761,
+      "learning_rate": 1.8328085362855055e-06,
+      "loss": 0.6647,
+      "step": 753
+    },
+    {
+      "epoch": 0.6287262872628726,
+      "grad_norm": 0.10026909410953522,
+      "learning_rate": 1.8323108987062802e-06,
+      "loss": 0.6543,
+      "step": 754
+    },
+    {
+      "epoch": 0.6295601417552638,
+      "grad_norm": 0.10641954839229584,
+      "learning_rate": 1.8318125893981553e-06,
+      "loss": 0.6425,
+      "step": 755
+    },
+    {
+      "epoch": 0.6303939962476548,
+      "grad_norm": 0.09939148277044296,
+      "learning_rate": 1.8313136087632995e-06,
+      "loss": 0.6531,
+      "step": 756
+    },
+    {
+      "epoch": 0.6312278507400458,
+      "grad_norm": 0.10437260568141937,
+      "learning_rate": 1.8308139572044215e-06,
+      "loss": 0.6783,
+      "step": 757
+    },
+    {
+      "epoch": 0.632061705232437,
+      "grad_norm": 0.09760341793298721,
+      "learning_rate": 1.830313635124773e-06,
+      "loss": 0.6734,
+      "step": 758
+    },
+    {
+      "epoch": 0.632895559724828,
+      "grad_norm": 0.1017068475484848,
+      "learning_rate": 1.8298126429281467e-06,
+      "loss": 0.6171,
+      "step": 759
+    },
+    {
+      "epoch": 0.6337294142172191,
+      "grad_norm": 0.09881062060594559,
+      "learning_rate": 1.8293109810188754e-06,
+      "loss": 0.6309,
+      "step": 760
+    },
+    {
+      "epoch": 0.6345632687096102,
+      "grad_norm": 0.11102797091007233,
+      "learning_rate": 1.8288086498018326e-06,
+      "loss": 0.6361,
+      "step": 761
+    },
+    {
+      "epoch": 0.6353971232020013,
+      "grad_norm": 0.09473263472318649,
+      "learning_rate": 1.828305649682433e-06,
+      "loss": 0.5787,
+      "step": 762
+    },
+    {
+      "epoch": 0.6362309776943923,
+      "grad_norm": 0.10742990672588348,
+      "learning_rate": 1.8278019810666293e-06,
+      "loss": 0.6638,
+      "step": 763
+    },
+    {
+      "epoch": 0.6370648321867834,
+      "grad_norm": 0.09987856447696686,
+      "learning_rate": 1.8272976443609156e-06,
+      "loss": 0.6054,
+      "step": 764
+    },
+    {
+      "epoch": 0.6378986866791745,
+      "grad_norm": 0.09839458018541336,
+      "learning_rate": 1.8267926399723238e-06,
+      "loss": 0.6258,
+      "step": 765
+    },
+    {
+      "epoch": 0.6387325411715655,
+      "grad_norm": 0.09883236885070801,
+      "learning_rate": 1.8262869683084258e-06,
+      "loss": 0.6695,
+      "step": 766
+    },
+    {
+      "epoch": 0.6395663956639567,
+      "grad_norm": 0.10263945162296295,
+      "learning_rate": 1.825780629777331e-06,
+      "loss": 0.6032,
+      "step": 767
+    },
+    {
+      "epoch": 0.6404002501563477,
+      "grad_norm": 0.09623769670724869,
+      "learning_rate": 1.8252736247876874e-06,
+      "loss": 0.6794,
+      "step": 768
+    },
+    {
+      "epoch": 0.6412341046487388,
+      "grad_norm": 0.10218116641044617,
+      "learning_rate": 1.8247659537486811e-06,
+      "loss": 0.5969,
+      "step": 769
+    },
+    {
+      "epoch": 0.6420679591411299,
+      "grad_norm": 0.09405802190303802,
+      "learning_rate": 1.8242576170700352e-06,
+      "loss": 0.6493,
+      "step": 770
+    },
+    {
+      "epoch": 0.642901813633521,
+      "grad_norm": 0.09709781408309937,
+      "learning_rate": 1.8237486151620112e-06,
+      "loss": 0.7054,
+      "step": 771
+    },
+    {
+      "epoch": 0.643735668125912,
+      "grad_norm": 0.09584866464138031,
+      "learning_rate": 1.823238948435406e-06,
+      "loss": 0.6299,
+      "step": 772
+    },
+    {
+      "epoch": 0.6445695226183031,
+      "grad_norm": 0.09731820970773697,
+      "learning_rate": 1.8227286173015538e-06,
+      "loss": 0.6633,
+      "step": 773
+    },
+    {
+      "epoch": 0.6454033771106942,
+      "grad_norm": 0.09664048999547958,
+      "learning_rate": 1.822217622172325e-06,
+      "loss": 0.6075,
+      "step": 774
+    },
+    {
+      "epoch": 0.6462372316030852,
+      "grad_norm": 0.0990557000041008,
+      "learning_rate": 1.8217059634601259e-06,
+      "loss": 0.6755,
+      "step": 775
+    },
+    {
+      "epoch": 0.6470710860954764,
+      "grad_norm": 0.10071469098329544,
+      "learning_rate": 1.8211936415778983e-06,
+      "loss": 0.6197,
+      "step": 776
+    },
+    {
+      "epoch": 0.6479049405878674,
+      "grad_norm": 0.10386556386947632,
+      "learning_rate": 1.820680656939119e-06,
+      "loss": 0.6195,
+      "step": 777
+    },
+    {
+      "epoch": 0.6487387950802584,
+      "grad_norm": 0.09966234862804413,
+      "learning_rate": 1.8201670099578001e-06,
+      "loss": 0.6654,
+      "step": 778
+    },
+    {
+      "epoch": 0.6495726495726496,
+      "grad_norm": 0.09882418066263199,
+      "learning_rate": 1.819652701048488e-06,
+      "loss": 0.6525,
+      "step": 779
+    },
+    {
+      "epoch": 0.6504065040650406,
+      "grad_norm": 0.09504549950361252,
+      "learning_rate": 1.8191377306262633e-06,
+      "loss": 0.6326,
+      "step": 780
+    },
+    {
+      "epoch": 0.6512403585574317,
+      "grad_norm": 0.09834848344326019,
+      "learning_rate": 1.8186220991067404e-06,
+      "loss": 0.6417,
+      "step": 781
+    },
+    {
+      "epoch": 0.6520742130498228,
+      "grad_norm": 0.09583116322755814,
+      "learning_rate": 1.8181058069060676e-06,
+      "loss": 0.6644,
+      "step": 782
+    },
+    {
+      "epoch": 0.6529080675422139,
+      "grad_norm": 0.09287475794553757,
+      "learning_rate": 1.8175888544409264e-06,
+      "loss": 0.6385,
+      "step": 783
+    },
+    {
+      "epoch": 0.653741922034605,
+      "grad_norm": 0.10364864766597748,
+      "learning_rate": 1.8170712421285305e-06,
+      "loss": 0.6996,
+      "step": 784
+    },
+    {
+      "epoch": 0.6545757765269961,
+      "grad_norm": 0.09863609820604324,
+      "learning_rate": 1.816552970386627e-06,
+      "loss": 0.6272,
+      "step": 785
+    },
+    {
+      "epoch": 0.6554096310193871,
+      "grad_norm": 0.10557069629430771,
+      "learning_rate": 1.8160340396334949e-06,
+      "loss": 0.6318,
+      "step": 786
+    },
+    {
+      "epoch": 0.6562434855117782,
+      "grad_norm": 0.10506842285394669,
+      "learning_rate": 1.8155144502879443e-06,
+      "loss": 0.6557,
+      "step": 787
+    },
+    {
+      "epoch": 0.6570773400041693,
+      "grad_norm": 0.09928935766220093,
+      "learning_rate": 1.8149942027693182e-06,
+      "loss": 0.6832,
+      "step": 788
+    },
+    {
+      "epoch": 0.6579111944965603,
+      "grad_norm": 0.09821849316358566,
+      "learning_rate": 1.8144732974974902e-06,
+      "loss": 0.6342,
+      "step": 789
+    },
+    {
+      "epoch": 0.6587450489889515,
+      "grad_norm": 0.09966670721769333,
+      "learning_rate": 1.8139517348928637e-06,
+      "loss": 0.6321,
+      "step": 790
+    },
+    {
+      "epoch": 0.6595789034813425,
+      "grad_norm": 0.10439924150705338,
+      "learning_rate": 1.8134295153763745e-06,
+      "loss": 0.6552,
+      "step": 791
+    },
+    {
+      "epoch": 0.6604127579737336,
+      "grad_norm": 0.09848225861787796,
+      "learning_rate": 1.8129066393694877e-06,
+      "loss": 0.593,
+      "step": 792
+    },
+    {
+      "epoch": 0.6612466124661247,
+      "grad_norm": 0.09793344885110855,
+      "learning_rate": 1.8123831072941974e-06,
+      "loss": 0.6406,
+      "step": 793
+    },
+    {
+      "epoch": 0.6620804669585157,
+      "grad_norm": 0.0945219025015831,
+      "learning_rate": 1.8118589195730285e-06,
+      "loss": 0.6118,
+      "step": 794
+    },
+    {
+      "epoch": 0.6629143214509068,
+      "grad_norm": 0.10230748355388641,
+      "learning_rate": 1.8113340766290343e-06,
+      "loss": 0.65,
+      "step": 795
+    },
+    {
+      "epoch": 0.6637481759432979,
+      "grad_norm": 0.10141695290803909,
+      "learning_rate": 1.8108085788857973e-06,
+      "loss": 0.6109,
+      "step": 796
+    },
+    {
+      "epoch": 0.664582030435689,
+      "grad_norm": 0.11011820286512375,
+      "learning_rate": 1.8102824267674282e-06,
+      "loss": 0.6531,
+      "step": 797
+    },
+    {
+      "epoch": 0.66541588492808,
+      "grad_norm": 0.09687966108322144,
+      "learning_rate": 1.8097556206985658e-06,
+      "loss": 0.649,
+      "step": 798
+    },
+    {
+      "epoch": 0.6662497394204712,
+      "grad_norm": 0.09653454273939133,
+      "learning_rate": 1.8092281611043764e-06,
+      "loss": 0.6487,
+      "step": 799
+    },
+    {
+      "epoch": 0.6670835939128622,
+      "grad_norm": 0.10373052954673767,
+      "learning_rate": 1.808700048410555e-06,
+      "loss": 0.6221,
+      "step": 800
+    },
+    {
+      "epoch": 0.6670835939128622,
+      "eval_loss": 0.6415942907333374,
+      "eval_runtime": 321.6008,
+      "eval_samples_per_second": 16.738,
+      "eval_steps_per_second": 2.792,
+      "step": 800
+    },
+    {
+      "epoch": 0.6679174484052532,
+      "grad_norm": 0.10485277324914932,
+      "learning_rate": 1.8081712830433224e-06,
+      "loss": 0.6176,
+      "step": 801
+    },
+    {
+      "epoch": 0.6687513028976444,
+      "grad_norm": 0.10105489194393158,
+      "learning_rate": 1.8076418654294266e-06,
+      "loss": 0.6708,
+      "step": 802
+    },
+    {
+      "epoch": 0.6695851573900354,
+      "grad_norm": 0.1096423864364624,
+      "learning_rate": 1.8071117959961416e-06,
+      "loss": 0.601,
+      "step": 803
+    },
+    {
+      "epoch": 0.6704190118824265,
+      "grad_norm": 0.10762202739715576,
+      "learning_rate": 1.8065810751712682e-06,
+      "loss": 0.6567,
+      "step": 804
+    },
+    {
+      "epoch": 0.6712528663748176,
+      "grad_norm": 0.1107725277543068,
+      "learning_rate": 1.8060497033831324e-06,
+      "loss": 0.5889,
+      "step": 805
+    },
+    {
+      "epoch": 0.6720867208672087,
+      "grad_norm": 0.10468114167451859,
+      "learning_rate": 1.8055176810605857e-06,
+      "loss": 0.5847,
+      "step": 806
+    },
+    {
+      "epoch": 0.6729205753595997,
+      "grad_norm": 0.11013887077569962,
+      "learning_rate": 1.8049850086330048e-06,
+      "loss": 0.6008,
+      "step": 807
+    },
+    {
+      "epoch": 0.6737544298519909,
+      "grad_norm": 0.1043848916888237,
+      "learning_rate": 1.8044516865302908e-06,
+      "loss": 0.6368,
+      "step": 808
+    },
+    {
+      "epoch": 0.6745882843443819,
+      "grad_norm": 0.09962385892868042,
+      "learning_rate": 1.8039177151828691e-06,
+      "loss": 0.5842,
+      "step": 809
+    },
+    {
+      "epoch": 0.6754221388367729,
+      "grad_norm": 0.1098175197839737,
+      "learning_rate": 1.8033830950216897e-06,
+      "loss": 0.6397,
+      "step": 810
+    },
+    {
+      "epoch": 0.6762559933291641,
+      "grad_norm": 0.10420270264148712,
+      "learning_rate": 1.8028478264782252e-06,
+      "loss": 0.6131,
+      "step": 811
+    },
+    {
+      "epoch": 0.6770898478215551,
+      "grad_norm": 0.10396002978086472,
+      "learning_rate": 1.8023119099844723e-06,
+      "loss": 0.6499,
+      "step": 812
+    },
+    {
+      "epoch": 0.6779237023139462,
+      "grad_norm": 0.09836006909608841,
+      "learning_rate": 1.8017753459729504e-06,
+      "loss": 0.5962,
+      "step": 813
+    },
+    {
+      "epoch": 0.6787575568063373,
+      "grad_norm": 0.11005699634552002,
+      "learning_rate": 1.8012381348767014e-06,
+      "loss": 0.6202,
+      "step": 814
+    },
+    {
+      "epoch": 0.6795914112987284,
+      "grad_norm": 0.10937905311584473,
+      "learning_rate": 1.8007002771292896e-06,
+      "loss": 0.6053,
+      "step": 815
+    },
+    {
+      "epoch": 0.6804252657911195,
+      "grad_norm": 0.10817151516675949,
+      "learning_rate": 1.800161773164801e-06,
+      "loss": 0.6713,
+      "step": 816
+    },
+    {
+      "epoch": 0.6812591202835105,
+      "grad_norm": 0.10142417997121811,
+      "learning_rate": 1.7996226234178434e-06,
+      "loss": 0.6166,
+      "step": 817
+    },
+    {
+      "epoch": 0.6820929747759016,
+      "grad_norm": 0.10415297746658325,
+      "learning_rate": 1.799082828323545e-06,
+      "loss": 0.6912,
+      "step": 818
+    },
+    {
+      "epoch": 0.6829268292682927,
+      "grad_norm": 0.09985463321208954,
+      "learning_rate": 1.7985423883175565e-06,
+      "loss": 0.6694,
+      "step": 819
+    },
+    {
+      "epoch": 0.6837606837606838,
+      "grad_norm": 0.1025400459766388,
+      "learning_rate": 1.7980013038360475e-06,
+      "loss": 0.6464,
+      "step": 820
+    },
+    {
+      "epoch": 0.6845945382530748,
+      "grad_norm": 0.09324987977743149,
+      "learning_rate": 1.7974595753157082e-06,
+      "loss": 0.6251,
+      "step": 821
+    },
+    {
+      "epoch": 0.685428392745466,
+      "grad_norm": 0.1065581887960434,
+      "learning_rate": 1.796917203193749e-06,
+      "loss": 0.6427,
+      "step": 822
+    },
+    {
+      "epoch": 0.686262247237857,
+      "grad_norm": 0.1113988533616066,
+      "learning_rate": 1.7963741879078985e-06,
+      "loss": 0.6652,
+      "step": 823
+    },
+    {
+      "epoch": 0.687096101730248,
+      "grad_norm": 0.09690185636281967,
+      "learning_rate": 1.7958305298964066e-06,
+      "loss": 0.6294,
+      "step": 824
+    },
+    {
+      "epoch": 0.6879299562226392,
+      "grad_norm": 0.10385391116142273,
+      "learning_rate": 1.7952862295980397e-06,
+      "loss": 0.611,
+      "step": 825
+    },
+    {
+      "epoch": 0.6887638107150302,
+      "grad_norm": 0.10155371576547623,
+      "learning_rate": 1.794741287452084e-06,
+      "loss": 0.6268,
+      "step": 826
+    },
+    {
+      "epoch": 0.6895976652074213,
+      "grad_norm": 0.10731323808431625,
+      "learning_rate": 1.7941957038983425e-06,
+      "loss": 0.6294,
+      "step": 827
+    },
+    {
+      "epoch": 0.6904315196998124,
+      "grad_norm": 0.0989241898059845,
+      "learning_rate": 1.7936494793771369e-06,
+      "loss": 0.6739,
+      "step": 828
+    },
+    {
+      "epoch": 0.6912653741922035,
+      "grad_norm": 0.10223259776830673,
+      "learning_rate": 1.7931026143293059e-06,
+      "loss": 0.5857,
+      "step": 829
+    },
+    {
+      "epoch": 0.6920992286845945,
+      "grad_norm": 0.10547658056020737,
+      "learning_rate": 1.792555109196205e-06,
+      "loss": 0.6864,
+      "step": 830
+    },
+    {
+      "epoch": 0.6929330831769857,
+      "grad_norm": 0.1027214452624321,
+      "learning_rate": 1.7920069644197063e-06,
+      "loss": 0.5857,
+      "step": 831
+    },
+    {
+      "epoch": 0.6937669376693767,
+      "grad_norm": 0.10017300397157669,
+      "learning_rate": 1.7914581804421983e-06,
+      "loss": 0.6466,
+      "step": 832
+    },
+    {
+      "epoch": 0.6946007921617677,
+      "grad_norm": 0.09962292015552521,
+      "learning_rate": 1.7909087577065851e-06,
+      "loss": 0.5917,
+      "step": 833
+    },
+    {
+      "epoch": 0.6954346466541589,
+      "grad_norm": 0.10357961803674698,
+      "learning_rate": 1.7903586966562868e-06,
+      "loss": 0.6382,
+      "step": 834
+    },
+    {
+      "epoch": 0.6962685011465499,
+      "grad_norm": 0.1033301129937172,
+      "learning_rate": 1.789807997735238e-06,
+      "loss": 0.624,
+      "step": 835
+    },
+    {
+      "epoch": 0.697102355638941,
+      "grad_norm": 0.09935612231492996,
+      "learning_rate": 1.789256661387889e-06,
+      "loss": 0.6475,
+      "step": 836
+    },
+    {
+      "epoch": 0.6979362101313321,
+      "grad_norm": 0.09738706797361374,
+      "learning_rate": 1.7887046880592033e-06,
+      "loss": 0.652,
+      "step": 837
+    },
+    {
+      "epoch": 0.6987700646237232,
+      "grad_norm": 0.10389488190412521,
+      "learning_rate": 1.7881520781946597e-06,
+      "loss": 0.6755,
+      "step": 838
+    },
+    {
+      "epoch": 0.6996039191161142,
+      "grad_norm": 0.09787558019161224,
+      "learning_rate": 1.7875988322402501e-06,
+      "loss": 0.618,
+      "step": 839
+    },
+    {
+      "epoch": 0.7004377736085053,
+      "grad_norm": 0.09852594137191772,
+      "learning_rate": 1.78704495064248e-06,
+      "loss": 0.618,
+      "step": 840
+    },
+    {
+      "epoch": 0.7012716281008964,
+      "grad_norm": 0.09819158911705017,
+      "learning_rate": 1.7864904338483675e-06,
+      "loss": 0.5994,
+      "step": 841
+    },
+    {
+      "epoch": 0.7021054825932874,
+      "grad_norm": 0.10942944139242172,
+      "learning_rate": 1.7859352823054437e-06,
+      "loss": 0.6405,
+      "step": 842
+    },
+    {
+      "epoch": 0.7029393370856786,
+      "grad_norm": 0.10699477791786194,
+      "learning_rate": 1.7853794964617522e-06,
+      "loss": 0.6444,
+      "step": 843
+    },
+    {
+      "epoch": 0.7037731915780696,
+      "grad_norm": 0.10151588916778564,
+      "learning_rate": 1.784823076765848e-06,
+      "loss": 0.6233,
+      "step": 844
+    },
+    {
+      "epoch": 0.7046070460704607,
+      "grad_norm": 0.10147272795438766,
+      "learning_rate": 1.7842660236667974e-06,
+      "loss": 0.6606,
+      "step": 845
+    },
+    {
+      "epoch": 0.7054409005628518,
+      "grad_norm": 0.1031784638762474,
+      "learning_rate": 1.7837083376141796e-06,
+      "loss": 0.6332,
+      "step": 846
+    },
+    {
+      "epoch": 0.7062747550552428,
+      "grad_norm": 0.10318465530872345,
+      "learning_rate": 1.7831500190580823e-06,
+      "loss": 0.5944,
+      "step": 847
+    },
+    {
+      "epoch": 0.707108609547634,
+      "grad_norm": 0.10107962787151337,
+      "learning_rate": 1.7825910684491054e-06,
+      "loss": 0.686,
+      "step": 848
+    },
+    {
+      "epoch": 0.707942464040025,
+      "grad_norm": 0.1008550301194191,
+      "learning_rate": 1.7820314862383584e-06,
+      "loss": 0.6356,
+      "step": 849
+    },
+    {
+      "epoch": 0.7087763185324161,
+      "grad_norm": 0.1049947738647461,
+      "learning_rate": 1.7814712728774598e-06,
+      "loss": 0.6436,
+      "step": 850
+    },
+    {
+      "epoch": 0.7096101730248072,
+      "grad_norm": 0.10559200495481491,
+      "learning_rate": 1.7809104288185389e-06,
+      "loss": 0.6522,
+      "step": 851
+    },
+    {
+      "epoch": 0.7104440275171983,
+      "grad_norm": 0.09807837009429932,
+      "learning_rate": 1.7803489545142325e-06,
+      "loss": 0.6252,
+      "step": 852
+    },
+    {
+      "epoch": 0.7112778820095893,
+      "grad_norm": 0.1072913408279419,
+      "learning_rate": 1.7797868504176874e-06,
+      "loss": 0.6483,
+      "step": 853
+    },
+    {
+      "epoch": 0.7121117365019805,
+      "grad_norm": 0.09987325221300125,
+      "learning_rate": 1.7792241169825579e-06,
+      "loss": 0.5955,
+      "step": 854
+    },
+    {
+      "epoch": 0.7129455909943715,
+      "grad_norm": 0.10160518437623978,
+      "learning_rate": 1.778660754663006e-06,
+      "loss": 0.6006,
+      "step": 855
+    },
+    {
+      "epoch": 0.7137794454867625,
+      "grad_norm": 0.10053478926420212,
+      "learning_rate": 1.7780967639137025e-06,
+      "loss": 0.6583,
+      "step": 856
+    },
+    {
+      "epoch": 0.7146132999791537,
+      "grad_norm": 0.10520876199007034,
+      "learning_rate": 1.777532145189824e-06,
+      "loss": 0.6061,
+      "step": 857
+    },
+    {
+      "epoch": 0.7154471544715447,
+      "grad_norm": 0.10797824710607529,
+      "learning_rate": 1.776966898947054e-06,
+      "loss": 0.6764,
+      "step": 858
+    },
+    {
+      "epoch": 0.7162810089639358,
+      "grad_norm": 0.10396280139684677,
+      "learning_rate": 1.7764010256415837e-06,
+      "loss": 0.6302,
+      "step": 859
+    },
+    {
+      "epoch": 0.7171148634563269,
+      "grad_norm": 0.10683812946081161,
+      "learning_rate": 1.7758345257301094e-06,
+      "loss": 0.6618,
+      "step": 860
+    },
+    {
+      "epoch": 0.717948717948718,
+      "grad_norm": 0.09601815789937973,
+      "learning_rate": 1.7752673996698326e-06,
+      "loss": 0.6167,
+      "step": 861
+    },
+    {
+      "epoch": 0.718782572441109,
+      "grad_norm": 0.1079709529876709,
+      "learning_rate": 1.774699647918462e-06,
+      "loss": 0.6085,
+      "step": 862
+    },
+    {
+      "epoch": 0.7196164269335001,
+      "grad_norm": 0.10923435539007187,
+      "learning_rate": 1.774131270934209e-06,
+      "loss": 0.6492,
+      "step": 863
+    },
+    {
+      "epoch": 0.7204502814258912,
+      "grad_norm": 0.10494917631149292,
+      "learning_rate": 1.7735622691757912e-06,
+      "loss": 0.6281,
+      "step": 864
+    },
+    {
+      "epoch": 0.7212841359182822,
+      "grad_norm": 0.10305362194776535,
+      "learning_rate": 1.7729926431024301e-06,
+      "loss": 0.6381,
+      "step": 865
+    },
+    {
+      "epoch": 0.7221179904106734,
+      "grad_norm": 0.1005750373005867,
+      "learning_rate": 1.7724223931738505e-06,
+      "loss": 0.667,
+      "step": 866
+    },
+    {
+      "epoch": 0.7229518449030644,
+      "grad_norm": 0.10202402621507645,
+      "learning_rate": 1.7718515198502812e-06,
+      "loss": 0.6685,
+      "step": 867
+    },
+    {
+      "epoch": 0.7237856993954555,
+      "grad_norm": 0.10256760567426682,
+      "learning_rate": 1.7712800235924546e-06,
+      "loss": 0.6728,
+      "step": 868
+    },
+    {
+      "epoch": 0.7246195538878466,
+      "grad_norm": 0.1020650714635849,
+      "learning_rate": 1.7707079048616046e-06,
+      "loss": 0.6144,
+      "step": 869
+    },
+    {
+      "epoch": 0.7254534083802376,
+      "grad_norm": 0.10101509839296341,
+      "learning_rate": 1.770135164119468e-06,
+      "loss": 0.6244,
+      "step": 870
+    },
+    {
+      "epoch": 0.7262872628726287,
+      "grad_norm": 0.09737613797187805,
+      "learning_rate": 1.769561801828284e-06,
+      "loss": 0.6507,
+      "step": 871
+    },
+    {
+      "epoch": 0.7271211173650198,
+      "grad_norm": 0.09866712987422943,
+      "learning_rate": 1.7689878184507937e-06,
+      "loss": 0.5875,
+      "step": 872
+    },
+    {
+      "epoch": 0.7279549718574109,
+      "grad_norm": 0.0963241383433342,
+      "learning_rate": 1.7684132144502382e-06,
+      "loss": 0.6237,
+      "step": 873
+    },
+    {
+      "epoch": 0.7287888263498019,
+      "grad_norm": 0.09706535190343857,
+      "learning_rate": 1.7678379902903603e-06,
+      "loss": 0.5665,
+      "step": 874
+    },
+    {
+      "epoch": 0.7296226808421931,
+      "grad_norm": 0.09960032999515533,
+      "learning_rate": 1.7672621464354034e-06,
+      "loss": 0.6288,
+      "step": 875
+    },
+    {
+      "epoch": 0.7304565353345841,
+      "grad_norm": 0.10458897799253464,
+      "learning_rate": 1.766685683350111e-06,
+      "loss": 0.6148,
+      "step": 876
+    },
+    {
+      "epoch": 0.7312903898269751,
+      "grad_norm": 0.10443190485239029,
+      "learning_rate": 1.766108601499726e-06,
+      "loss": 0.6248,
+      "step": 877
+    },
+    {
+      "epoch": 0.7321242443193663,
+      "grad_norm": 0.10145988315343857,
+      "learning_rate": 1.765530901349991e-06,
+      "loss": 0.6117,
+      "step": 878
+    },
+    {
+      "epoch": 0.7329580988117573,
+      "grad_norm": 0.10847010463476181,
+      "learning_rate": 1.7649525833671474e-06,
+      "loss": 0.7046,
+      "step": 879
+    },
+    {
+      "epoch": 0.7337919533041485,
+      "grad_norm": 0.10503443330526352,
+      "learning_rate": 1.7643736480179352e-06,
+      "loss": 0.6176,
+      "step": 880
+    },
+    {
+      "epoch": 0.7346258077965395,
+      "grad_norm": 0.10929647833108902,
+      "learning_rate": 1.7637940957695934e-06,
+      "loss": 0.6434,
+      "step": 881
+    },
+    {
+      "epoch": 0.7354596622889306,
+      "grad_norm": 0.10153260827064514,
+      "learning_rate": 1.7632139270898576e-06,
+      "loss": 0.638,
+      "step": 882
+    },
+    {
+      "epoch": 0.7362935167813217,
+      "grad_norm": 0.10493889451026917,
+      "learning_rate": 1.7626331424469615e-06,
+      "loss": 0.6517,
+      "step": 883
+    },
+    {
+      "epoch": 0.7371273712737128,
+      "grad_norm": 0.09808988124132156,
+      "learning_rate": 1.7620517423096368e-06,
+      "loss": 0.6179,
+      "step": 884
+    },
+    {
+      "epoch": 0.7379612257661038,
+      "grad_norm": 0.10479886084794998,
+      "learning_rate": 1.7614697271471103e-06,
+      "loss": 0.6419,
+      "step": 885
+    },
+    {
+      "epoch": 0.7387950802584949,
+      "grad_norm": 0.09941962361335754,
+      "learning_rate": 1.7608870974291065e-06,
+      "loss": 0.6128,
+      "step": 886
+    },
+    {
+      "epoch": 0.739628934750886,
+      "grad_norm": 0.10169458389282227,
+      "learning_rate": 1.7603038536258453e-06,
+      "loss": 0.609,
+      "step": 887
+    },
+    {
+      "epoch": 0.740462789243277,
+      "grad_norm": 0.09726240485906601,
+      "learning_rate": 1.7597199962080423e-06,
+      "loss": 0.6371,
+      "step": 888
+    },
+    {
+      "epoch": 0.7412966437356682,
+      "grad_norm": 0.10218486189842224,
+      "learning_rate": 1.759135525646908e-06,
+      "loss": 0.5768,
+      "step": 889
+    },
+    {
+      "epoch": 0.7421304982280592,
+      "grad_norm": 0.10321173816919327,
+      "learning_rate": 1.7585504424141483e-06,
+      "loss": 0.6007,
+      "step": 890
+    },
+    {
+      "epoch": 0.7429643527204502,
+      "grad_norm": 0.10717494040727615,
+      "learning_rate": 1.7579647469819631e-06,
+      "loss": 0.6353,
+      "step": 891
+    },
+    {
+      "epoch": 0.7437982072128414,
+      "grad_norm": 0.1051764264702797,
+      "learning_rate": 1.7573784398230474e-06,
+      "loss": 0.6098,
+      "step": 892
+    },
+    {
+      "epoch": 0.7446320617052324,
+      "grad_norm": 0.10321137309074402,
+      "learning_rate": 1.7567915214105881e-06,
+      "loss": 0.5986,
+      "step": 893
+    },
+    {
+      "epoch": 0.7454659161976235,
+      "grad_norm": 0.10979589819908142,
+      "learning_rate": 1.7562039922182671e-06,
+      "loss": 0.6518,
+      "step": 894
+    },
+    {
+      "epoch": 0.7462997706900146,
+      "grad_norm": 0.10153292864561081,
+      "learning_rate": 1.7556158527202585e-06,
+      "loss": 0.6358,
+      "step": 895
+    },
+    {
+      "epoch": 0.7471336251824057,
+      "grad_norm": 0.10513276606798172,
+      "learning_rate": 1.7550271033912287e-06,
+      "loss": 0.6085,
+      "step": 896
+    },
+    {
+      "epoch": 0.7479674796747967,
+      "grad_norm": 0.10275143384933472,
+      "learning_rate": 1.7544377447063372e-06,
+      "loss": 0.5748,
+      "step": 897
+    },
+    {
+      "epoch": 0.7488013341671879,
+      "grad_norm": 0.10438092797994614,
+      "learning_rate": 1.7538477771412339e-06,
+      "loss": 0.5943,
+      "step": 898
+    },
+    {
+      "epoch": 0.7496351886595789,
+      "grad_norm": 0.1010277271270752,
+      "learning_rate": 1.7532572011720617e-06,
+      "loss": 0.6112,
+      "step": 899
+    },
+    {
+      "epoch": 0.7504690431519699,
+      "grad_norm": 0.10229716449975967,
+      "learning_rate": 1.752666017275453e-06,
+      "loss": 0.5861,
+      "step": 900
+    },
+    {
+      "epoch": 0.7513028976443611,
+      "grad_norm": 0.10199355334043503,
+      "learning_rate": 1.7520742259285323e-06,
+      "loss": 0.6388,
+      "step": 901
+    },
+    {
+      "epoch": 0.7521367521367521,
+      "grad_norm": 0.10268741846084595,
+      "learning_rate": 1.7514818276089128e-06,
+      "loss": 0.6875,
+      "step": 902
+    },
+    {
+      "epoch": 0.7529706066291432,
+      "grad_norm": 0.0992840975522995,
+      "learning_rate": 1.7508888227946992e-06,
+      "loss": 0.6756,
+      "step": 903
+    },
+    {
+      "epoch": 0.7538044611215343,
+      "grad_norm": 0.09615278989076614,
+      "learning_rate": 1.7502952119644845e-06,
+      "loss": 0.6474,
+      "step": 904
+    },
+    {
+      "epoch": 0.7546383156139254,
+      "grad_norm": 0.10315293818712234,
+      "learning_rate": 1.749700995597351e-06,
+      "loss": 0.6127,
+      "step": 905
+    },
+    {
+      "epoch": 0.7554721701063164,
+      "grad_norm": 0.10551444441080093,
+      "learning_rate": 1.7491061741728702e-06,
+      "loss": 0.6357,
+      "step": 906
+    },
+    {
+      "epoch": 0.7563060245987075,
+      "grad_norm": 0.10280577838420868,
+      "learning_rate": 1.748510748171101e-06,
+      "loss": 0.6045,
+      "step": 907
+    },
+    {
+      "epoch": 0.7571398790910986,
+      "grad_norm": 0.09516436606645584,
+      "learning_rate": 1.7479147180725912e-06,
+      "loss": 0.6368,
+      "step": 908
+    },
+    {
+      "epoch": 0.7579737335834896,
+      "grad_norm": 0.10799692571163177,
+      "learning_rate": 1.7473180843583762e-06,
+      "loss": 0.6669,
+      "step": 909
+    },
+    {
+      "epoch": 0.7588075880758808,
+      "grad_norm": 0.1064273938536644,
+      "learning_rate": 1.7467208475099774e-06,
+      "loss": 0.6481,
+      "step": 910
+    },
+    {
+      "epoch": 0.7596414425682718,
+      "grad_norm": 0.1054486557841301,
+      "learning_rate": 1.7461230080094043e-06,
+      "loss": 0.6108,
+      "step": 911
+    },
+    {
+      "epoch": 0.760475297060663,
+      "grad_norm": 0.10060249269008636,
+      "learning_rate": 1.7455245663391516e-06,
+      "loss": 0.5604,
+      "step": 912
+    },
+    {
+      "epoch": 0.761309151553054,
+      "grad_norm": 0.1015784963965416,
+      "learning_rate": 1.7449255229822008e-06,
+      "loss": 0.6248,
+      "step": 913
+    },
+    {
+      "epoch": 0.762143006045445,
+      "grad_norm": 0.10114108771085739,
+      "learning_rate": 1.744325878422019e-06,
+      "loss": 0.6683,
+      "step": 914
+    },
+    {
+      "epoch": 0.7629768605378362,
+      "grad_norm": 0.10236384719610214,
+      "learning_rate": 1.7437256331425576e-06,
+      "loss": 0.657,
+      "step": 915
+    },
+    {
+      "epoch": 0.7638107150302272,
+      "grad_norm": 0.10892149806022644,
+      "learning_rate": 1.743124787628254e-06,
+      "loss": 0.6434,
+      "step": 916
+    },
+    {
+      "epoch": 0.7646445695226183,
+      "grad_norm": 0.10639214515686035,
+      "learning_rate": 1.7425233423640298e-06,
+      "loss": 0.5859,
+      "step": 917
+    },
+    {
+      "epoch": 0.7654784240150094,
+      "grad_norm": 0.10279032588005066,
+      "learning_rate": 1.7419212978352898e-06,
+      "loss": 0.6779,
+      "step": 918
+    },
+    {
+      "epoch": 0.7663122785074005,
+      "grad_norm": 0.11178586632013321,
+      "learning_rate": 1.741318654527923e-06,
+      "loss": 0.6827,
+      "step": 919
+    },
+    {
+      "epoch": 0.7671461329997915,
+      "grad_norm": 0.1018877848982811,
+      "learning_rate": 1.7407154129283019e-06,
+      "loss": 0.6423,
+      "step": 920
+    },
+    {
+      "epoch": 0.7679799874921827,
+      "grad_norm": 0.1052117571234703,
+      "learning_rate": 1.7401115735232817e-06,
+      "loss": 0.6871,
+      "step": 921
+    },
+    {
+      "epoch": 0.7688138419845737,
+      "grad_norm": 0.10656667500734329,
+      "learning_rate": 1.7395071368001997e-06,
+      "loss": 0.6384,
+      "step": 922
+    },
+    {
+      "epoch": 0.7696476964769647,
+      "grad_norm": 0.09973619878292084,
+      "learning_rate": 1.738902103246876e-06,
+      "loss": 0.5659,
+      "step": 923
+    },
+    {
+      "epoch": 0.7704815509693559,
+      "grad_norm": 0.10072916746139526,
+      "learning_rate": 1.7382964733516112e-06,
+      "loss": 0.6315,
+      "step": 924
+    },
+    {
+      "epoch": 0.7713154054617469,
+      "grad_norm": 0.11351827532052994,
+      "learning_rate": 1.7376902476031882e-06,
+      "loss": 0.7005,
+      "step": 925
+    },
+    {
+      "epoch": 0.772149259954138,
+      "grad_norm": 0.1084694042801857,
+      "learning_rate": 1.7370834264908711e-06,
+      "loss": 0.6077,
+      "step": 926
+    },
+    {
+      "epoch": 0.7729831144465291,
+      "grad_norm": 0.10465063899755478,
+      "learning_rate": 1.7364760105044033e-06,
+      "loss": 0.5937,
+      "step": 927
+    },
+    {
+      "epoch": 0.7738169689389202,
+      "grad_norm": 0.09937559068202972,
+      "learning_rate": 1.7358680001340092e-06,
+      "loss": 0.5854,
+      "step": 928
+    },
+    {
+      "epoch": 0.7746508234313112,
+      "grad_norm": 0.11517240107059479,
+      "learning_rate": 1.7352593958703921e-06,
+      "loss": 0.6738,
+      "step": 929
+    },
+    {
+      "epoch": 0.7754846779237023,
+      "grad_norm": 0.10875809192657471,
+      "learning_rate": 1.734650198204736e-06,
+      "loss": 0.6487,
+      "step": 930
+    },
+    {
+      "epoch": 0.7763185324160934,
+      "grad_norm": 0.10209451615810394,
+      "learning_rate": 1.7340404076287021e-06,
+      "loss": 0.6277,
+      "step": 931
+    },
+    {
+      "epoch": 0.7771523869084844,
+      "grad_norm": 0.11190709471702576,
+      "learning_rate": 1.7334300246344316e-06,
+      "loss": 0.6393,
+      "step": 932
+    },
+    {
+      "epoch": 0.7779862414008756,
+      "grad_norm": 0.10622856765985489,
+      "learning_rate": 1.7328190497145427e-06,
+      "loss": 0.6665,
+      "step": 933
+    },
+    {
+      "epoch": 0.7788200958932666,
+      "grad_norm": 0.11038866639137268,
+      "learning_rate": 1.7322074833621318e-06,
+      "loss": 0.6092,
+      "step": 934
+    },
+    {
+      "epoch": 0.7796539503856577,
+      "grad_norm": 0.1060342788696289,
+      "learning_rate": 1.7315953260707735e-06,
+      "loss": 0.6427,
+      "step": 935
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 0.10829740762710571,
+      "learning_rate": 1.7309825783345174e-06,
+      "loss": 0.6851,
+      "step": 936
+    },
+    {
+      "epoch": 0.7813216593704398,
+      "grad_norm": 0.1001635491847992,
+      "learning_rate": 1.7303692406478908e-06,
+      "loss": 0.5645,
+      "step": 937
+    },
+    {
+      "epoch": 0.7821555138628309,
+      "grad_norm": 0.09842494130134583,
+      "learning_rate": 1.7297553135058978e-06,
+      "loss": 0.613,
+      "step": 938
+    },
+    {
+      "epoch": 0.782989368355222,
+      "grad_norm": 0.10335814207792282,
+      "learning_rate": 1.7291407974040167e-06,
+      "loss": 0.6583,
+      "step": 939
+    },
+    {
+      "epoch": 0.7838232228476131,
+      "grad_norm": 0.10193807631731033,
+      "learning_rate": 1.728525692838202e-06,
+      "loss": 0.6242,
+      "step": 940
+    },
+    {
+      "epoch": 0.7846570773400041,
+      "grad_norm": 0.1064397543668747,
+      "learning_rate": 1.7279100003048832e-06,
+      "loss": 0.7043,
+      "step": 941
+    },
+    {
+      "epoch": 0.7854909318323953,
+      "grad_norm": 0.09790313243865967,
+      "learning_rate": 1.7272937203009642e-06,
+      "loss": 0.5869,
+      "step": 942
+    },
+    {
+      "epoch": 0.7863247863247863,
+      "grad_norm": 0.10425697267055511,
+      "learning_rate": 1.7266768533238225e-06,
+      "loss": 0.6282,
+      "step": 943
+    },
+    {
+      "epoch": 0.7871586408171775,
+      "grad_norm": 0.10319948196411133,
+      "learning_rate": 1.72605939987131e-06,
+      "loss": 0.5899,
+      "step": 944
+    },
+    {
+      "epoch": 0.7879924953095685,
+      "grad_norm": 0.10117416828870773,
+      "learning_rate": 1.725441360441752e-06,
+      "loss": 0.5734,
+      "step": 945
+    },
+    {
+      "epoch": 0.7888263498019595,
+      "grad_norm": 0.10920300334692001,
+      "learning_rate": 1.7248227355339458e-06,
+      "loss": 0.612,
+      "step": 946
+    },
+    {
+      "epoch": 0.7896602042943507,
+      "grad_norm": 0.0993722602725029,
+      "learning_rate": 1.7242035256471623e-06,
+      "loss": 0.5778,
+      "step": 947
+    },
+    {
+      "epoch": 0.7904940587867417,
+      "grad_norm": 0.12078487873077393,
+      "learning_rate": 1.723583731281144e-06,
+      "loss": 0.7115,
+      "step": 948
+    },
+    {
+      "epoch": 0.7913279132791328,
+      "grad_norm": 0.10817496478557587,
+      "learning_rate": 1.7229633529361051e-06,
+      "loss": 0.6699,
+      "step": 949
+    },
+    {
+      "epoch": 0.7921617677715239,
+      "grad_norm": 0.10841819643974304,
+      "learning_rate": 1.7223423911127313e-06,
+      "loss": 0.6089,
+      "step": 950
+    },
+    {
+      "epoch": 0.792995622263915,
+      "grad_norm": 0.10630793124437332,
+      "learning_rate": 1.7217208463121788e-06,
+      "loss": 0.6615,
+      "step": 951
+    },
+    {
+      "epoch": 0.793829476756306,
+      "grad_norm": 0.1000799611210823,
+      "learning_rate": 1.721098719036075e-06,
+      "loss": 0.6231,
+      "step": 952
+    },
+    {
+      "epoch": 0.7946633312486971,
+      "grad_norm": 0.11009353399276733,
+      "learning_rate": 1.7204760097865167e-06,
+      "loss": 0.6197,
+      "step": 953
+    },
+    {
+      "epoch": 0.7954971857410882,
+      "grad_norm": 0.10809972882270813,
+      "learning_rate": 1.7198527190660706e-06,
+      "loss": 0.5975,
+      "step": 954
+    },
+    {
+      "epoch": 0.7963310402334792,
+      "grad_norm": 0.10163717716932297,
+      "learning_rate": 1.719228847377773e-06,
+      "loss": 0.7125,
+      "step": 955
+    },
+    {
+      "epoch": 0.7971648947258704,
+      "grad_norm": 0.1121022030711174,
+      "learning_rate": 1.7186043952251286e-06,
+      "loss": 0.6821,
+      "step": 956
+    },
+    {
+      "epoch": 0.7979987492182614,
+      "grad_norm": 0.11300718039274216,
+      "learning_rate": 1.7179793631121106e-06,
+      "loss": 0.6137,
+      "step": 957
+    },
+    {
+      "epoch": 0.7988326037106525,
+      "grad_norm": 0.10873489826917648,
+      "learning_rate": 1.717353751543161e-06,
+      "loss": 0.5621,
+      "step": 958
+    },
+    {
+      "epoch": 0.7996664582030436,
+      "grad_norm": 0.11031791567802429,
+      "learning_rate": 1.7167275610231887e-06,
+      "loss": 0.6363,
+      "step": 959
+    },
+    {
+      "epoch": 0.8005003126954346,
+      "grad_norm": 0.10201506316661835,
+      "learning_rate": 1.7161007920575704e-06,
+      "loss": 0.5696,
+      "step": 960
+    },
+    {
+      "epoch": 0.8013341671878257,
+      "grad_norm": 0.09877403825521469,
+      "learning_rate": 1.7154734451521486e-06,
+      "loss": 0.6027,
+      "step": 961
+    },
+    {
+      "epoch": 0.8021680216802168,
+      "grad_norm": 0.1055438295006752,
+      "learning_rate": 1.7148455208132334e-06,
+      "loss": 0.6204,
+      "step": 962
+    },
+    {
+      "epoch": 0.8030018761726079,
+      "grad_norm": 0.11029175668954849,
+      "learning_rate": 1.7142170195476005e-06,
+      "loss": 0.5826,
+      "step": 963
+    },
+    {
+      "epoch": 0.8038357306649989,
+      "grad_norm": 0.1105756014585495,
+      "learning_rate": 1.7135879418624913e-06,
+      "loss": 0.6526,
+      "step": 964
+    },
+    {
+      "epoch": 0.8046695851573901,
+      "grad_norm": 0.1054423451423645,
+      "learning_rate": 1.712958288265612e-06,
+      "loss": 0.6254,
+      "step": 965
+    },
+    {
+      "epoch": 0.8055034396497811,
+      "grad_norm": 0.10613211989402771,
+      "learning_rate": 1.712328059265134e-06,
+      "loss": 0.5974,
+      "step": 966
+    },
+    {
+      "epoch": 0.8063372941421721,
+      "grad_norm": 0.09556613117456436,
+      "learning_rate": 1.7116972553696932e-06,
+      "loss": 0.6144,
+      "step": 967
+    },
+    {
+      "epoch": 0.8071711486345633,
+      "grad_norm": 0.10485559701919556,
+      "learning_rate": 1.711065877088389e-06,
+      "loss": 0.5998,
+      "step": 968
+    },
+    {
+      "epoch": 0.8080050031269543,
+      "grad_norm": 0.10885662585496902,
+      "learning_rate": 1.7104339249307848e-06,
+      "loss": 0.6298,
+      "step": 969
+    },
+    {
+      "epoch": 0.8088388576193454,
+      "grad_norm": 0.10140252858400345,
+      "learning_rate": 1.709801399406907e-06,
+      "loss": 0.5747,
+      "step": 970
+    },
+    {
+      "epoch": 0.8096727121117365,
+      "grad_norm": 0.09874554723501205,
+      "learning_rate": 1.7091683010272446e-06,
+      "loss": 0.5676,
+      "step": 971
+    },
+    {
+      "epoch": 0.8105065666041276,
+      "grad_norm": 0.10777262598276138,
+      "learning_rate": 1.7085346303027493e-06,
+      "loss": 0.6326,
+      "step": 972
+    },
+    {
+      "epoch": 0.8113404210965186,
+      "grad_norm": 0.09974364936351776,
+      "learning_rate": 1.7079003877448344e-06,
+      "loss": 0.5638,
+      "step": 973
+    },
+    {
+      "epoch": 0.8121742755889098,
+      "grad_norm": 0.10661293566226959,
+      "learning_rate": 1.7072655738653745e-06,
+      "loss": 0.6641,
+      "step": 974
+    },
+    {
+      "epoch": 0.8130081300813008,
+      "grad_norm": 0.10669861733913422,
+      "learning_rate": 1.7066301891767061e-06,
+      "loss": 0.6478,
+      "step": 975
+    },
+    {
+      "epoch": 0.813841984573692,
+      "grad_norm": 0.10368253290653229,
+      "learning_rate": 1.7059942341916256e-06,
+      "loss": 0.6198,
+      "step": 976
+    },
+    {
+      "epoch": 0.814675839066083,
+      "grad_norm": 0.10855797678232193,
+      "learning_rate": 1.7053577094233897e-06,
+      "loss": 0.6453,
+      "step": 977
+    },
+    {
+      "epoch": 0.815509693558474,
+      "grad_norm": 0.11007635295391083,
+      "learning_rate": 1.7047206153857156e-06,
+      "loss": 0.5748,
+      "step": 978
+    },
+    {
+      "epoch": 0.8163435480508652,
+      "grad_norm": 0.10489246994256973,
+      "learning_rate": 1.7040829525927796e-06,
+      "loss": 0.579,
+      "step": 979
+    },
+    {
+      "epoch": 0.8171774025432562,
+      "grad_norm": 0.11224298179149628,
+      "learning_rate": 1.7034447215592164e-06,
+      "loss": 0.6484,
+      "step": 980
+    },
+    {
+      "epoch": 0.8180112570356473,
+      "grad_norm": 0.10526615381240845,
+      "learning_rate": 1.7028059228001204e-06,
+      "loss": 0.6746,
+      "step": 981
+    },
+    {
+      "epoch": 0.8188451115280384,
+      "grad_norm": 0.10187442600727081,
+      "learning_rate": 1.7021665568310435e-06,
+      "loss": 0.6462,
+      "step": 982
+    },
+    {
+      "epoch": 0.8196789660204294,
+      "grad_norm": 0.10846269875764847,
+      "learning_rate": 1.7015266241679952e-06,
+      "loss": 0.6329,
+      "step": 983
+    },
+    {
+      "epoch": 0.8205128205128205,
+      "grad_norm": 0.110533706843853,
+      "learning_rate": 1.7008861253274429e-06,
+      "loss": 0.658,
+      "step": 984
+    },
+    {
+      "epoch": 0.8213466750052116,
+      "grad_norm": 0.10688811540603638,
+      "learning_rate": 1.7002450608263107e-06,
+      "loss": 0.5996,
+      "step": 985
+    },
+    {
+      "epoch": 0.8221805294976027,
+      "grad_norm": 0.10414768755435944,
+      "learning_rate": 1.6996034311819796e-06,
+      "loss": 0.6333,
+      "step": 986
+    },
+    {
+      "epoch": 0.8230143839899937,
+      "grad_norm": 0.0999814048409462,
+      "learning_rate": 1.698961236912286e-06,
+      "loss": 0.6609,
+      "step": 987
+    },
+    {
+      "epoch": 0.8238482384823849,
+      "grad_norm": 0.10203441977500916,
+      "learning_rate": 1.6983184785355222e-06,
+      "loss": 0.6095,
+      "step": 988
+    },
+    {
+      "epoch": 0.8246820929747759,
+      "grad_norm": 0.10571661591529846,
+      "learning_rate": 1.6976751565704362e-06,
+      "loss": 0.6125,
+      "step": 989
+    },
+    {
+      "epoch": 0.8255159474671669,
+      "grad_norm": 0.10715372860431671,
+      "learning_rate": 1.6970312715362304e-06,
+      "loss": 0.689,
+      "step": 990
+    },
+    {
+      "epoch": 0.8263498019595581,
+      "grad_norm": 0.10807687044143677,
+      "learning_rate": 1.696386823952562e-06,
+      "loss": 0.611,
+      "step": 991
+    },
+    {
+      "epoch": 0.8271836564519491,
+      "grad_norm": 0.10625305026769638,
+      "learning_rate": 1.6957418143395418e-06,
+      "loss": 0.6616,
+      "step": 992
+    },
+    {
+      "epoch": 0.8280175109443402,
+      "grad_norm": 0.10512302070856094,
+      "learning_rate": 1.6950962432177348e-06,
+      "loss": 0.6147,
+      "step": 993
+    },
+    {
+      "epoch": 0.8288513654367313,
+      "grad_norm": 0.10524202138185501,
+      "learning_rate": 1.6944501111081579e-06,
+      "loss": 0.6094,
+      "step": 994
+    },
+    {
+      "epoch": 0.8296852199291224,
+      "grad_norm": 0.10839590430259705,
+      "learning_rate": 1.6938034185322828e-06,
+      "loss": 0.5816,
+      "step": 995
+    },
+    {
+      "epoch": 0.8305190744215134,
+      "grad_norm": 0.10930492728948593,
+      "learning_rate": 1.6931561660120312e-06,
+      "loss": 0.6621,
+      "step": 996
+    },
+    {
+      "epoch": 0.8313529289139046,
+      "grad_norm": 0.10111601650714874,
+      "learning_rate": 1.692508354069779e-06,
+      "loss": 0.6083,
+      "step": 997
+    },
+    {
+      "epoch": 0.8321867834062956,
+      "grad_norm": 0.11010827124118805,
+      "learning_rate": 1.691859983228352e-06,
+      "loss": 0.6689,
+      "step": 998
+    },
+    {
+      "epoch": 0.8330206378986866,
+      "grad_norm": 0.10622293502092361,
+      "learning_rate": 1.6912110540110272e-06,
+      "loss": 0.6142,
+      "step": 999
+    },
+    {
+      "epoch": 0.8338544923910778,
+      "grad_norm": 0.10322947055101395,
+      "learning_rate": 1.6905615669415325e-06,
+      "loss": 0.5797,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8346883468834688,
+      "grad_norm": 0.11376485228538513,
+      "learning_rate": 1.689911522544047e-06,
+      "loss": 0.6323,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8355222013758599,
+      "grad_norm": 0.10336299985647202,
+      "learning_rate": 1.6892609213431981e-06,
+      "loss": 0.5907,
+      "step": 1002
+    },
+    {
+      "epoch": 0.836356055868251,
+      "grad_norm": 0.10549134016036987,
+      "learning_rate": 1.6886097638640631e-06,
+      "loss": 0.6483,
+      "step": 1003
+    },
+    {
+      "epoch": 0.837189910360642,
+      "grad_norm": 0.10979744791984558,
+      "learning_rate": 1.6879580506321687e-06,
+      "loss": 0.5829,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8380237648530331,
+      "grad_norm": 0.10647737979888916,
+      "learning_rate": 1.6873057821734894e-06,
+      "loss": 0.5714,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8388576193454242,
+      "grad_norm": 0.11350049823522568,
+      "learning_rate": 1.6866529590144485e-06,
+      "loss": 0.6756,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8396914738378153,
+      "grad_norm": 0.10672125220298767,
+      "learning_rate": 1.6859995816819167e-06,
+      "loss": 0.6459,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8405253283302064,
+      "grad_norm": 0.11561840772628784,
+      "learning_rate": 1.6853456507032112e-06,
+      "loss": 0.6912,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8413591828225975,
+      "grad_norm": 0.11554259806871414,
+      "learning_rate": 1.6846911666060973e-06,
+      "loss": 0.6032,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8421930373149885,
+      "grad_norm": 0.10536365956068039,
+      "learning_rate": 1.6840361299187857e-06,
+      "loss": 0.6182,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8430268918073797,
+      "grad_norm": 0.1125800758600235,
+      "learning_rate": 1.6833805411699344e-06,
+      "loss": 0.6519,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8438607462997707,
+      "grad_norm": 0.10594160109758377,
+      "learning_rate": 1.6827244008886453e-06,
+      "loss": 0.6407,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8446946007921617,
+      "grad_norm": 0.10480530560016632,
+      "learning_rate": 1.6820677096044667e-06,
+      "loss": 0.6398,
+      "step": 1013
+    },
+    {
+      "epoch": 0.8455284552845529,
+      "grad_norm": 0.10666308552026749,
+      "learning_rate": 1.6814104678473905e-06,
+      "loss": 0.612,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8463623097769439,
+      "grad_norm": 0.10556510835886002,
+      "learning_rate": 1.6807526761478533e-06,
+      "loss": 0.5729,
+      "step": 1015
+    },
+    {
+      "epoch": 0.847196164269335,
+      "grad_norm": 0.11039448529481888,
+      "learning_rate": 1.6800943350367368e-06,
+      "loss": 0.6323,
+      "step": 1016
+    },
+    {
+      "epoch": 0.8480300187617261,
+      "grad_norm": 0.10304141789674759,
+      "learning_rate": 1.6794354450453638e-06,
+      "loss": 0.6426,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8488638732541172,
+      "grad_norm": 0.10969959199428558,
+      "learning_rate": 1.678776006705502e-06,
+      "loss": 0.558,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8496977277465082,
+      "grad_norm": 0.10663289576768875,
+      "learning_rate": 1.6781160205493605e-06,
+      "loss": 0.593,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8505315822388994,
+      "grad_norm": 0.10291384905576706,
+      "learning_rate": 1.6774554871095915e-06,
+      "loss": 0.6281,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8513654367312904,
+      "grad_norm": 0.10907971113920212,
+      "learning_rate": 1.6767944069192876e-06,
+      "loss": 0.6158,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8521992912236814,
+      "grad_norm": 0.11564616858959198,
+      "learning_rate": 1.6761327805119838e-06,
+      "loss": 0.6662,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8530331457160726,
+      "grad_norm": 0.10590541362762451,
+      "learning_rate": 1.6754706084216555e-06,
+      "loss": 0.6323,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8538670002084636,
+      "grad_norm": 0.10117157548666,
+      "learning_rate": 1.6748078911827187e-06,
+      "loss": 0.6112,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8547008547008547,
+      "grad_norm": 0.11531368643045425,
+      "learning_rate": 1.674144629330029e-06,
+      "loss": 0.6413,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8555347091932458,
+      "grad_norm": 0.10621378570795059,
+      "learning_rate": 1.673480823398882e-06,
+      "loss": 0.6394,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8563685636856369,
+      "grad_norm": 0.11612808704376221,
+      "learning_rate": 1.672816473925012e-06,
+      "loss": 0.637,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8572024181780279,
+      "grad_norm": 0.10893040150403976,
+      "learning_rate": 1.672151581444592e-06,
+      "loss": 0.5923,
+      "step": 1028
+    },
+    {
+      "epoch": 0.858036272670419,
+      "grad_norm": 0.10822536796331406,
+      "learning_rate": 1.6714861464942333e-06,
+      "loss": 0.6766,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8588701271628101,
+      "grad_norm": 0.10714786499738693,
+      "learning_rate": 1.6708201696109856e-06,
+      "loss": 0.5649,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8597039816552011,
+      "grad_norm": 0.10966484993696213,
+      "learning_rate": 1.6701536513323349e-06,
+      "loss": 0.6219,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8605378361475923,
+      "grad_norm": 0.1037144884467125,
+      "learning_rate": 1.669486592196205e-06,
+      "loss": 0.6352,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8613716906399833,
+      "grad_norm": 0.11299126595258713,
+      "learning_rate": 1.668818992740956e-06,
+      "loss": 0.6734,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8622055451323744,
+      "grad_norm": 0.11029834300279617,
+      "learning_rate": 1.6681508535053834e-06,
+      "loss": 0.6579,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8630393996247655,
+      "grad_norm": 0.10519642382860184,
+      "learning_rate": 1.6674821750287197e-06,
+      "loss": 0.6072,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8638732541171565,
+      "grad_norm": 0.11065202951431274,
+      "learning_rate": 1.6668129578506313e-06,
+      "loss": 0.6562,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8647071086095476,
+      "grad_norm": 0.11024871468544006,
+      "learning_rate": 1.66614320251122e-06,
+      "loss": 0.6,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8655409631019387,
+      "grad_norm": 0.10754359513521194,
+      "learning_rate": 1.6654729095510219e-06,
+      "loss": 0.6028,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8663748175943298,
+      "grad_norm": 0.11345981806516647,
+      "learning_rate": 1.6648020795110069e-06,
+      "loss": 0.7018,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8672086720867209,
+      "grad_norm": 0.10346484929323196,
+      "learning_rate": 1.6641307129325783e-06,
+      "loss": 0.6866,
+      "step": 1040
+    },
+    {
+      "epoch": 0.868042526579112,
+      "grad_norm": 0.10640288889408112,
+      "learning_rate": 1.6634588103575723e-06,
+      "loss": 0.5851,
+      "step": 1041
+    },
+    {
+      "epoch": 0.868876381071503,
+      "grad_norm": 0.11161711812019348,
+      "learning_rate": 1.6627863723282584e-06,
+      "loss": 0.6182,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8697102355638942,
+      "grad_norm": 0.10999694466590881,
+      "learning_rate": 1.662113399387337e-06,
+      "loss": 0.5887,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8705440900562852,
+      "grad_norm": 0.10471642017364502,
+      "learning_rate": 1.6614398920779418e-06,
+      "loss": 0.6854,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8713779445486762,
+      "grad_norm": 0.10635983943939209,
+      "learning_rate": 1.660765850943636e-06,
+      "loss": 0.5945,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8722117990410674,
+      "grad_norm": 0.10250985622406006,
+      "learning_rate": 1.6600912765284153e-06,
+      "loss": 0.6283,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8730456535334584,
+      "grad_norm": 0.10342993587255478,
+      "learning_rate": 1.6594161693767046e-06,
+      "loss": 0.5878,
+      "step": 1047
+    },
+    {
+      "epoch": 0.8738795080258495,
+      "grad_norm": 0.10388782620429993,
+      "learning_rate": 1.6587405300333593e-06,
+      "loss": 0.6197,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8747133625182406,
+      "grad_norm": 0.10757216066122055,
+      "learning_rate": 1.658064359043664e-06,
+      "loss": 0.6643,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8755472170106317,
+      "grad_norm": 0.10185371339321136,
+      "learning_rate": 1.657387656953333e-06,
+      "loss": 0.6129,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8763810715030227,
+      "grad_norm": 0.10625848174095154,
+      "learning_rate": 1.6567104243085081e-06,
+      "loss": 0.6389,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8772149259954138,
+      "grad_norm": 0.1068265438079834,
+      "learning_rate": 1.6560326616557605e-06,
+      "loss": 0.6474,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8780487804878049,
+      "grad_norm": 0.10008926689624786,
+      "learning_rate": 1.6553543695420888e-06,
+      "loss": 0.592,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8788826349801959,
+      "grad_norm": 0.11250849813222885,
+      "learning_rate": 1.6546755485149181e-06,
+      "loss": 0.6612,
+      "step": 1054
+    },
+    {
+      "epoch": 0.8797164894725871,
+      "grad_norm": 0.10780669748783112,
+      "learning_rate": 1.6539961991221017e-06,
+      "loss": 0.5717,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8805503439649781,
+      "grad_norm": 0.10522449761629105,
+      "learning_rate": 1.6533163219119181e-06,
+      "loss": 0.6377,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8813841984573692,
+      "grad_norm": 0.11719299852848053,
+      "learning_rate": 1.6526359174330727e-06,
+      "loss": 0.6148,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8822180529497603,
+      "grad_norm": 0.10418464988470078,
+      "learning_rate": 1.6519549862346959e-06,
+      "loss": 0.547,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8830519074421513,
+      "grad_norm": 0.1149667277932167,
+      "learning_rate": 1.6512735288663433e-06,
+      "loss": 0.5509,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8838857619345424,
+      "grad_norm": 0.11353281885385513,
+      "learning_rate": 1.6505915458779954e-06,
+      "loss": 0.6679,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8847196164269335,
+      "grad_norm": 0.10084139555692673,
+      "learning_rate": 1.6499090378200564e-06,
+      "loss": 0.6315,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8855534709193246,
+      "grad_norm": 0.10753504931926727,
+      "learning_rate": 1.6492260052433551e-06,
+      "loss": 0.6441,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8863873254117156,
+      "grad_norm": 0.10944227129220963,
+      "learning_rate": 1.6485424486991427e-06,
+      "loss": 0.6378,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8872211799041068,
+      "grad_norm": 0.10542403161525726,
+      "learning_rate": 1.6478583687390937e-06,
+      "loss": 0.5715,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8880550343964978,
+      "grad_norm": 0.10192226618528366,
+      "learning_rate": 1.6471737659153054e-06,
+      "loss": 0.6035,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.11085714399814606,
+      "learning_rate": 1.6464886407802958e-06,
+      "loss": 0.6705,
+      "step": 1066
+    },
+    {
+      "epoch": 0.88972274338128,
+      "grad_norm": 0.10521090775728226,
+      "learning_rate": 1.645802993887006e-06,
+      "loss": 0.6487,
+      "step": 1067
+    },
+    {
+      "epoch": 0.890556597873671,
+      "grad_norm": 0.11182847619056702,
+      "learning_rate": 1.645116825788798e-06,
+      "loss": 0.5908,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8913904523660621,
+      "grad_norm": 0.10744032263755798,
+      "learning_rate": 1.6444301370394533e-06,
+      "loss": 0.6978,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8922243068584532,
+      "grad_norm": 0.11645323783159256,
+      "learning_rate": 1.6437429281931742e-06,
+      "loss": 0.6552,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8930581613508443,
+      "grad_norm": 0.10556191205978394,
+      "learning_rate": 1.6430551998045833e-06,
+      "loss": 0.6485,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8938920158432354,
+      "grad_norm": 0.11244357377290726,
+      "learning_rate": 1.6423669524287216e-06,
+      "loss": 0.6352,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8947258703356264,
+      "grad_norm": 0.11015337705612183,
+      "learning_rate": 1.6416781866210494e-06,
+      "loss": 0.5928,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8955597248280175,
+      "grad_norm": 0.11401670426130295,
+      "learning_rate": 1.6409889029374457e-06,
+      "loss": 0.591,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8963935793204086,
+      "grad_norm": 0.10178755223751068,
+      "learning_rate": 1.6402991019342073e-06,
+      "loss": 0.5772,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8972274338127997,
+      "grad_norm": 0.10722572356462479,
+      "learning_rate": 1.6396087841680478e-06,
+      "loss": 0.616,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8980612883051907,
+      "grad_norm": 0.10824041068553925,
+      "learning_rate": 1.6389179501960987e-06,
+      "loss": 0.6289,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8988951427975819,
+      "grad_norm": 0.10898349434137344,
+      "learning_rate": 1.638226600575908e-06,
+      "loss": 0.6197,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8997289972899729,
+      "grad_norm": 0.10828059166669846,
+      "learning_rate": 1.6375347358654397e-06,
+      "loss": 0.6335,
+      "step": 1079
+    },
+    {
+      "epoch": 0.900562851782364,
+      "grad_norm": 0.11353094130754471,
+      "learning_rate": 1.6368423566230728e-06,
+      "loss": 0.6306,
+      "step": 1080
+    },
+    {
+      "epoch": 0.9013967062747551,
+      "grad_norm": 0.1095949038863182,
+      "learning_rate": 1.6361494634076033e-06,
+      "loss": 0.6579,
+      "step": 1081
+    },
+    {
+      "epoch": 0.9022305607671461,
+      "grad_norm": 0.10779423266649246,
+      "learning_rate": 1.6354560567782406e-06,
+      "loss": 0.6268,
+      "step": 1082
+    },
+    {
+      "epoch": 0.9030644152595372,
+      "grad_norm": 0.10539865493774414,
+      "learning_rate": 1.6347621372946088e-06,
+      "loss": 0.6318,
+      "step": 1083
+    },
+    {
+      "epoch": 0.9038982697519283,
+      "grad_norm": 0.11474636197090149,
+      "learning_rate": 1.6340677055167458e-06,
+      "loss": 0.6843,
+      "step": 1084
+    },
+    {
+      "epoch": 0.9047321242443194,
+      "grad_norm": 0.11284147948026657,
+      "learning_rate": 1.633372762005103e-06,
+      "loss": 0.6167,
+      "step": 1085
+    },
+    {
+      "epoch": 0.9055659787367104,
+      "grad_norm": 0.10842804610729218,
+      "learning_rate": 1.632677307320545e-06,
+      "loss": 0.5851,
+      "step": 1086
+    },
+    {
+      "epoch": 0.9063998332291016,
+      "grad_norm": 0.10843487083911896,
+      "learning_rate": 1.6319813420243495e-06,
+      "loss": 0.6337,
+      "step": 1087
+    },
+    {
+      "epoch": 0.9072336877214926,
+      "grad_norm": 0.10792160034179688,
+      "learning_rate": 1.6312848666782048e-06,
+      "loss": 0.6015,
+      "step": 1088
+    },
+    {
+      "epoch": 0.9080675422138836,
+      "grad_norm": 0.10741087794303894,
+      "learning_rate": 1.6305878818442122e-06,
+      "loss": 0.6285,
+      "step": 1089
+    },
+    {
+      "epoch": 0.9089013967062748,
+      "grad_norm": 0.10371655970811844,
+      "learning_rate": 1.6298903880848834e-06,
+      "loss": 0.5982,
+      "step": 1090
+    },
+    {
+      "epoch": 0.9097352511986658,
+      "grad_norm": 0.10612435638904572,
+      "learning_rate": 1.6291923859631415e-06,
+      "loss": 0.5962,
+      "step": 1091
+    },
+    {
+      "epoch": 0.9105691056910569,
+      "grad_norm": 0.10705637186765671,
+      "learning_rate": 1.6284938760423188e-06,
+      "loss": 0.5926,
+      "step": 1092
+    },
+    {
+      "epoch": 0.911402960183448,
+      "grad_norm": 0.10797090083360672,
+      "learning_rate": 1.627794858886159e-06,
+      "loss": 0.632,
+      "step": 1093
+    },
+    {
+      "epoch": 0.9122368146758391,
+      "grad_norm": 0.1132771223783493,
+      "learning_rate": 1.6270953350588137e-06,
+      "loss": 0.647,
+      "step": 1094
+    },
+    {
+      "epoch": 0.9130706691682301,
+      "grad_norm": 0.10793612152338028,
+      "learning_rate": 1.626395305124844e-06,
+      "loss": 0.6367,
+      "step": 1095
+    },
+    {
+      "epoch": 0.9139045236606212,
+      "grad_norm": 0.10706419497728348,
+      "learning_rate": 1.6256947696492196e-06,
+      "loss": 0.6472,
+      "step": 1096
+    },
+    {
+      "epoch": 0.9147383781530123,
+      "grad_norm": 0.10842972993850708,
+      "learning_rate": 1.6249937291973184e-06,
+      "loss": 0.6185,
+      "step": 1097
+    },
+    {
+      "epoch": 0.9155722326454033,
+      "grad_norm": 0.11361444741487503,
+      "learning_rate": 1.6242921843349252e-06,
+      "loss": 0.617,
+      "step": 1098
+    },
+    {
+      "epoch": 0.9164060871377945,
+      "grad_norm": 0.11315753310918808,
+      "learning_rate": 1.6235901356282322e-06,
+      "loss": 0.6478,
+      "step": 1099
+    },
+    {
+      "epoch": 0.9172399416301855,
+      "grad_norm": 0.12495766580104828,
+      "learning_rate": 1.6228875836438385e-06,
+      "loss": 0.6252,
+      "step": 1100
+    },
+    {
+      "epoch": 0.9180737961225766,
+      "grad_norm": 0.11157601326704025,
+      "learning_rate": 1.622184528948749e-06,
+      "loss": 0.6214,
+      "step": 1101
+    },
+    {
+      "epoch": 0.9189076506149677,
+      "grad_norm": 0.10694071650505066,
+      "learning_rate": 1.6214809721103744e-06,
+      "loss": 0.619,
+      "step": 1102
+    },
+    {
+      "epoch": 0.9197415051073587,
+      "grad_norm": 0.1112818494439125,
+      "learning_rate": 1.6207769136965307e-06,
+      "loss": 0.6229,
+      "step": 1103
+    },
+    {
+      "epoch": 0.9205753595997499,
+      "grad_norm": 0.10422814637422562,
+      "learning_rate": 1.6200723542754389e-06,
+      "loss": 0.5837,
+      "step": 1104
+    },
+    {
+      "epoch": 0.9214092140921409,
+      "grad_norm": 0.11240324378013611,
+      "learning_rate": 1.619367294415724e-06,
+      "loss": 0.7071,
+      "step": 1105
+    },
+    {
+      "epoch": 0.922243068584532,
+      "grad_norm": 0.10819468647241592,
+      "learning_rate": 1.6186617346864151e-06,
+      "loss": 0.5728,
+      "step": 1106
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.1053207591176033,
+      "learning_rate": 1.6179556756569448e-06,
+      "loss": 0.5915,
+      "step": 1107
+    },
+    {
+      "epoch": 0.9239107775693142,
+      "grad_norm": 0.1075674444437027,
+      "learning_rate": 1.6172491178971482e-06,
+      "loss": 0.5935,
+      "step": 1108
+    },
+    {
+      "epoch": 0.9247446320617052,
+      "grad_norm": 0.11159633845090866,
+      "learning_rate": 1.6165420619772635e-06,
+      "loss": 0.6493,
+      "step": 1109
+    },
+    {
+      "epoch": 0.9255784865540964,
+      "grad_norm": 0.11671009659767151,
+      "learning_rate": 1.6158345084679307e-06,
+      "loss": 0.6661,
+      "step": 1110
+    },
+    {
+      "epoch": 0.9264123410464874,
+      "grad_norm": 0.11934785544872284,
+      "learning_rate": 1.6151264579401917e-06,
+      "loss": 0.6335,
+      "step": 1111
+    },
+    {
+      "epoch": 0.9272461955388784,
+      "grad_norm": 0.11291314661502838,
+      "learning_rate": 1.6144179109654887e-06,
+      "loss": 0.6413,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9280800500312696,
+      "grad_norm": 0.10902924090623856,
+      "learning_rate": 1.6137088681156654e-06,
+      "loss": 0.6713,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9289139045236606,
+      "grad_norm": 0.10608460754156113,
+      "learning_rate": 1.6129993299629651e-06,
+      "loss": 0.6024,
+      "step": 1114
+    },
+    {
+      "epoch": 0.9297477590160517,
+      "grad_norm": 0.10624940693378448,
+      "learning_rate": 1.6122892970800317e-06,
+      "loss": 0.6128,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9305816135084428,
+      "grad_norm": 0.11591742187738419,
+      "learning_rate": 1.6115787700399071e-06,
+      "loss": 0.6242,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9314154680008339,
+      "grad_norm": 0.11706274002790451,
+      "learning_rate": 1.610867749416033e-06,
+      "loss": 0.6356,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9322493224932249,
+      "grad_norm": 0.10509736090898514,
+      "learning_rate": 1.6101562357822491e-06,
+      "loss": 0.6337,
+      "step": 1118
+    },
+    {
+      "epoch": 0.933083176985616,
+      "grad_norm": 0.10801702737808228,
+      "learning_rate": 1.6094442297127935e-06,
+      "loss": 0.615,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9339170314780071,
+      "grad_norm": 0.11627451330423355,
+      "learning_rate": 1.6087317317823007e-06,
+      "loss": 0.6059,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9347508859703981,
+      "grad_norm": 0.11316727101802826,
+      "learning_rate": 1.6080187425658033e-06,
+      "loss": 0.6092,
+      "step": 1121
+    },
+    {
+      "epoch": 0.9355847404627893,
+      "grad_norm": 0.10899489372968674,
+      "learning_rate": 1.6073052626387296e-06,
+      "loss": 0.5787,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9364185949551803,
+      "grad_norm": 0.10977214574813843,
+      "learning_rate": 1.606591292576904e-06,
+      "loss": 0.6022,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9372524494475714,
+      "grad_norm": 0.10926050692796707,
+      "learning_rate": 1.6058768329565469e-06,
+      "loss": 0.6626,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9380863039399625,
+      "grad_norm": 0.10779014229774475,
+      "learning_rate": 1.6051618843542736e-06,
+      "loss": 0.6381,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9389201584323535,
+      "grad_norm": 0.11509402096271515,
+      "learning_rate": 1.6044464473470939e-06,
+      "loss": 0.5817,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9397540129247446,
+      "grad_norm": 0.10561665147542953,
+      "learning_rate": 1.6037305225124121e-06,
+      "loss": 0.5634,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9405878674171357,
+      "grad_norm": 0.1081278994679451,
+      "learning_rate": 1.6030141104280253e-06,
+      "loss": 0.6437,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9414217219095268,
+      "grad_norm": 0.10470031946897507,
+      "learning_rate": 1.6022972116721256e-06,
+      "loss": 0.5853,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9422555764019178,
+      "grad_norm": 0.1069476306438446,
+      "learning_rate": 1.601579826823296e-06,
+      "loss": 0.5653,
+      "step": 1130
+    },
+    {
+      "epoch": 0.943089430894309,
+      "grad_norm": 0.1095786914229393,
+      "learning_rate": 1.600861956460513e-06,
+      "loss": 0.6191,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9439232853867,
+      "grad_norm": 0.10863108187913895,
+      "learning_rate": 1.600143601163144e-06,
+      "loss": 0.6166,
+      "step": 1132
+    },
+    {
+      "epoch": 0.944757139879091,
+      "grad_norm": 0.110601507127285,
+      "learning_rate": 1.5994247615109498e-06,
+      "loss": 0.5716,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9455909943714822,
+      "grad_norm": 0.108455128967762,
+      "learning_rate": 1.598705438084079e-06,
+      "loss": 0.5888,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9464248488638732,
+      "grad_norm": 0.11156295984983444,
+      "learning_rate": 1.5979856314630728e-06,
+      "loss": 0.6723,
+      "step": 1135
+    },
+    {
+      "epoch": 0.9472587033562644,
+      "grad_norm": 0.11098117381334305,
+      "learning_rate": 1.5972653422288622e-06,
+      "loss": 0.5671,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9480925578486554,
+      "grad_norm": 0.11285511404275894,
+      "learning_rate": 1.5965445709627669e-06,
+      "loss": 0.6437,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9489264123410465,
+      "grad_norm": 0.10554268956184387,
+      "learning_rate": 1.5958233182464965e-06,
+      "loss": 0.6289,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9497602668334376,
+      "grad_norm": 0.10124699026346207,
+      "learning_rate": 1.5951015846621483e-06,
+      "loss": 0.6068,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9505941213258287,
+      "grad_norm": 0.11128734052181244,
+      "learning_rate": 1.5943793707922084e-06,
+      "loss": 0.5907,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9514279758182197,
+      "grad_norm": 0.11676076054573059,
+      "learning_rate": 1.5936566772195503e-06,
+      "loss": 0.602,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9522618303106108,
+      "grad_norm": 0.11664719134569168,
+      "learning_rate": 1.5929335045274343e-06,
+      "loss": 0.6197,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9530956848030019,
+      "grad_norm": 0.1060841903090477,
+      "learning_rate": 1.5922098532995083e-06,
+      "loss": 0.553,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9539295392953929,
+      "grad_norm": 0.11080587655305862,
+      "learning_rate": 1.591485724119805e-06,
+      "loss": 0.6357,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9547633937877841,
+      "grad_norm": 0.10428017377853394,
+      "learning_rate": 1.5907611175727442e-06,
+      "loss": 0.5693,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9555972482801751,
+      "grad_norm": 0.10886859893798828,
+      "learning_rate": 1.59003603424313e-06,
+      "loss": 0.6142,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9564311027725662,
+      "grad_norm": 0.1091897264122963,
+      "learning_rate": 1.5893104747161522e-06,
+      "loss": 0.6234,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9572649572649573,
+      "grad_norm": 0.11705927550792694,
+      "learning_rate": 1.5885844395773841e-06,
+      "loss": 0.6322,
+      "step": 1148
+    },
+    {
+      "epoch": 0.9580988117573483,
+      "grad_norm": 0.11114629358053207,
+      "learning_rate": 1.5878579294127831e-06,
+      "loss": 0.6233,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9589326662497394,
+      "grad_norm": 0.11006072908639908,
+      "learning_rate": 1.5871309448086903e-06,
+      "loss": 0.6271,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9597665207421305,
+      "grad_norm": 0.10859903693199158,
+      "learning_rate": 1.5864034863518292e-06,
+      "loss": 0.6101,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9606003752345216,
+      "grad_norm": 0.11398789286613464,
+      "learning_rate": 1.585675554629306e-06,
+      "loss": 0.6183,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9614342297269126,
+      "grad_norm": 0.10914972424507141,
+      "learning_rate": 1.5849471502286087e-06,
+      "loss": 0.6687,
+      "step": 1153
+    },
+    {
+      "epoch": 0.9622680842193038,
+      "grad_norm": 0.11407602578401566,
+      "learning_rate": 1.5842182737376074e-06,
+      "loss": 0.6513,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9631019387116948,
+      "grad_norm": 0.11537064611911774,
+      "learning_rate": 1.5834889257445525e-06,
+      "loss": 0.6284,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9639357932040858,
+      "grad_norm": 0.10912200808525085,
+      "learning_rate": 1.582759106838075e-06,
+      "loss": 0.634,
+      "step": 1156
+    },
+    {
+      "epoch": 0.964769647696477,
+      "grad_norm": 0.11783526092767715,
+      "learning_rate": 1.5820288176071861e-06,
+      "loss": 0.6435,
+      "step": 1157
+    },
+    {
+      "epoch": 0.965603502188868,
+      "grad_norm": 0.11490530520677567,
+      "learning_rate": 1.5812980586412765e-06,
+      "loss": 0.5872,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9664373566812591,
+      "grad_norm": 0.10283337533473969,
+      "learning_rate": 1.5805668305301167e-06,
+      "loss": 0.565,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9672712111736502,
+      "grad_norm": 0.10747389495372772,
+      "learning_rate": 1.5798351338638548e-06,
+      "loss": 0.5878,
+      "step": 1160
+    },
+    {
+      "epoch": 0.9681050656660413,
+      "grad_norm": 0.11634895950555801,
+      "learning_rate": 1.5791029692330172e-06,
+      "loss": 0.6547,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9689389201584323,
+      "grad_norm": 0.11201413720846176,
+      "learning_rate": 1.5783703372285086e-06,
+      "loss": 0.6039,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9697727746508235,
+      "grad_norm": 0.11117483675479889,
+      "learning_rate": 1.5776372384416105e-06,
+      "loss": 0.6447,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9706066291432145,
+      "grad_norm": 0.11274691671133041,
+      "learning_rate": 1.5769036734639815e-06,
+      "loss": 0.5847,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9714404836356055,
+      "grad_norm": 0.11471108347177505,
+      "learning_rate": 1.5761696428876556e-06,
+      "loss": 0.6523,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9722743381279967,
+      "grad_norm": 0.11089842766523361,
+      "learning_rate": 1.5754351473050434e-06,
+      "loss": 0.5949,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9731081926203877,
+      "grad_norm": 0.10404518991708755,
+      "learning_rate": 1.5747001873089306e-06,
+      "loss": 0.5496,
+      "step": 1167
+    },
+    {
+      "epoch": 0.9739420471127789,
+      "grad_norm": 0.116578109562397,
+      "learning_rate": 1.5739647634924773e-06,
+      "loss": 0.6103,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9747759016051699,
+      "grad_norm": 0.1054786667227745,
+      "learning_rate": 1.5732288764492184e-06,
+      "loss": 0.623,
+      "step": 1169
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 0.11369701474905014,
+      "learning_rate": 1.5724925267730624e-06,
+      "loss": 0.567,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9764436105899521,
+      "grad_norm": 0.1198234111070633,
+      "learning_rate": 1.5717557150582917e-06,
+      "loss": 0.6188,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9772774650823431,
+      "grad_norm": 0.12100456655025482,
+      "learning_rate": 1.5710184418995604e-06,
+      "loss": 0.641,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9781113195747342,
+      "grad_norm": 0.10783829540014267,
+      "learning_rate": 1.5702807078918965e-06,
+      "loss": 0.5889,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9789451740671253,
+      "grad_norm": 0.11204763501882553,
+      "learning_rate": 1.5695425136306987e-06,
+      "loss": 0.6271,
+      "step": 1174
+    },
+    {
+      "epoch": 0.9797790285595164,
+      "grad_norm": 0.11363273113965988,
+      "learning_rate": 1.568803859711738e-06,
+      "loss": 0.6099,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9806128830519074,
+      "grad_norm": 0.10527540743350983,
+      "learning_rate": 1.5680647467311555e-06,
+      "loss": 0.5782,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9814467375442986,
+      "grad_norm": 0.11106487363576889,
+      "learning_rate": 1.5673251752854645e-06,
+      "loss": 0.645,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9822805920366896,
+      "grad_norm": 0.11001728475093842,
+      "learning_rate": 1.5665851459715457e-06,
+      "loss": 0.611,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9831144465290806,
+      "grad_norm": 0.10988467186689377,
+      "learning_rate": 1.5658446593866517e-06,
+      "loss": 0.6229,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9839483010214718,
+      "grad_norm": 0.10910095274448395,
+      "learning_rate": 1.565103716128403e-06,
+      "loss": 0.5964,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9847821555138628,
+      "grad_norm": 0.11653497070074081,
+      "learning_rate": 1.5643623167947891e-06,
+      "loss": 0.6451,
+      "step": 1181
+    },
+    {
+      "epoch": 0.9856160100062539,
+      "grad_norm": 0.11314839869737625,
+      "learning_rate": 1.5636204619841667e-06,
+      "loss": 0.6419,
+      "step": 1182
+    },
+    {
+      "epoch": 0.986449864498645,
+      "grad_norm": 0.10827460885047913,
+      "learning_rate": 1.5628781522952611e-06,
+      "loss": 0.6129,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9872837189910361,
+      "grad_norm": 0.1053231805562973,
+      "learning_rate": 1.5621353883271649e-06,
+      "loss": 0.6455,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9881175734834271,
+      "grad_norm": 0.11166463792324066,
+      "learning_rate": 1.561392170679336e-06,
+      "loss": 0.594,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9889514279758183,
+      "grad_norm": 0.11850026249885559,
+      "learning_rate": 1.5606484999516e-06,
+      "loss": 0.575,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9897852824682093,
+      "grad_norm": 0.11536535620689392,
+      "learning_rate": 1.5599043767441471e-06,
+      "loss": 0.645,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9906191369606003,
+      "grad_norm": 0.1191059872508049,
+      "learning_rate": 1.5591598016575328e-06,
+      "loss": 0.6007,
+      "step": 1188
+    },
+    {
+      "epoch": 0.9914529914529915,
+      "grad_norm": 0.1131712794303894,
+      "learning_rate": 1.558414775292678e-06,
+      "loss": 0.5926,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9922868459453825,
+      "grad_norm": 0.11502361297607422,
+      "learning_rate": 1.5576692982508663e-06,
+      "loss": 0.6223,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9931207004377736,
+      "grad_norm": 0.10861647129058838,
+      "learning_rate": 1.5569233711337474e-06,
+      "loss": 0.5948,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9939545549301647,
+      "grad_norm": 0.11109884828329086,
+      "learning_rate": 1.5561769945433323e-06,
+      "loss": 0.5778,
+      "step": 1192
+    },
+    {
+      "epoch": 0.9947884094225558,
+      "grad_norm": 0.1124456375837326,
+      "learning_rate": 1.5554301690819951e-06,
+      "loss": 0.6732,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9956222639149468,
+      "grad_norm": 0.10982295870780945,
+      "learning_rate": 1.5546828953524727e-06,
+      "loss": 0.6221,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9964561184073379,
+      "grad_norm": 0.10954145342111588,
+      "learning_rate": 1.553935173957863e-06,
+      "loss": 0.6045,
+      "step": 1195
+    },
+    {
+      "epoch": 0.997289972899729,
+      "grad_norm": 0.10329638421535492,
+      "learning_rate": 1.5531870055016265e-06,
+      "loss": 0.5827,
+      "step": 1196
+    },
+    {
+      "epoch": 0.99812382739212,
+      "grad_norm": 0.1147105023264885,
+      "learning_rate": 1.5524383905875828e-06,
+      "loss": 0.6676,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9989576818845112,
+      "grad_norm": 0.10970206558704376,
+      "learning_rate": 1.5516893298199133e-06,
+      "loss": 0.6159,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9997915363769022,
+      "grad_norm": 0.10848627239465714,
+      "learning_rate": 1.5509398238031585e-06,
+      "loss": 0.6106,
+      "step": 1199
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3597,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1199,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.45611918040105e+18,
+  "train_batch_size": 6,
+  "trial_name": null,
+  "trial_params": null
+}