diff --git "a/qwen2_5_math_7b/limo/checkpoint-820/trainer_state.json" "b/qwen2_5_math_7b/limo/checkpoint-820/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/qwen2_5_math_7b/limo/checkpoint-820/trainer_state.json"
@@ -0,0 +1,5774 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 820,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004878048780487805,
+      "grad_norm": 22.138404846191406,
+      "learning_rate": 5e-06,
+      "loss": 2.0514,
+      "step": 1
+    },
+    {
+      "epoch": 0.00975609756097561,
+      "grad_norm": 26.447072982788086,
+      "learning_rate": 4.999997064365715e-06,
+      "loss": 1.4083,
+      "step": 2
+    },
+    {
+      "epoch": 0.014634146341463415,
+      "grad_norm": 15.023693084716797,
+      "learning_rate": 4.999988257469751e-06,
+      "loss": 1.3992,
+      "step": 3
+    },
+    {
+      "epoch": 0.01951219512195122,
+      "grad_norm": 10.351157188415527,
+      "learning_rate": 4.999973579332793e-06,
+      "loss": 1.6141,
+      "step": 4
+    },
+    {
+      "epoch": 0.024390243902439025,
+      "grad_norm": 13.528725624084473,
+      "learning_rate": 4.999953029989312e-06,
+      "loss": 1.7049,
+      "step": 5
+    },
+    {
+      "epoch": 0.02926829268292683,
+      "grad_norm": 6.363333225250244,
+      "learning_rate": 4.999926609487568e-06,
+      "loss": 1.0504,
+      "step": 6
+    },
+    {
+      "epoch": 0.03414634146341464,
+      "grad_norm": 18.042160034179688,
+      "learning_rate": 4.9998943178896106e-06,
+      "loss": 1.7691,
+      "step": 7
+    },
+    {
+      "epoch": 0.03902439024390244,
+      "grad_norm": 18.375417709350586,
+      "learning_rate": 4.999856155271276e-06,
+      "loss": 2.3735,
+      "step": 8
+    },
+    {
+      "epoch": 0.04390243902439024,
+      "grad_norm": 13.300468444824219,
+      "learning_rate": 4.999812121722191e-06,
+      "loss": 1.9056,
+      "step": 9
+    },
+    {
+      "epoch": 0.04878048780487805,
+      "grad_norm": 9.966270446777344,
+      "learning_rate": 4.999762217345766e-06,
+      "loss": 2.5472,
+      "step": 10
+    },
+    {
+      "epoch": 0.05365853658536585,
+      "grad_norm": 9.960493087768555,
+      "learning_rate": 4.999706442259205e-06,
+      "loss": 1.2451,
+      "step": 11
+    },
+    {
+      "epoch": 0.05853658536585366,
+      "grad_norm": 6.841569900512695,
+      "learning_rate": 4.999644796593492e-06,
+      "loss": 1.3519,
+      "step": 12
+    },
+    {
+      "epoch": 0.06341463414634146,
+      "grad_norm": 8.004615783691406,
+      "learning_rate": 4.999577280493407e-06,
+      "loss": 1.2908,
+      "step": 13
+    },
+    {
+      "epoch": 0.06829268292682927,
+      "grad_norm": 8.352128982543945,
+      "learning_rate": 4.99950389411751e-06,
+      "loss": 1.5136,
+      "step": 14
+    },
+    {
+      "epoch": 0.07317073170731707,
+      "grad_norm": 2.6472861766815186,
+      "learning_rate": 4.999424637638148e-06,
+      "loss": 0.6843,
+      "step": 15
+    },
+    {
+      "epoch": 0.07804878048780488,
+      "grad_norm": 74.31636047363281,
+      "learning_rate": 4.999339511241458e-06,
+      "loss": 0.944,
+      "step": 16
+    },
+    {
+      "epoch": 0.08292682926829269,
+      "grad_norm": 65.4996337890625,
+      "learning_rate": 4.9992485151273584e-06,
+      "loss": 2.024,
+      "step": 17
+    },
+    {
+      "epoch": 0.08780487804878048,
+      "grad_norm": 16.723800659179688,
+      "learning_rate": 4.999151649509554e-06,
+      "loss": 1.5804,
+      "step": 18
+    },
+    {
+      "epoch": 0.09268292682926829,
+      "grad_norm": 12.863594055175781,
+      "learning_rate": 4.9990489146155356e-06,
+      "loss": 1.3615,
+      "step": 19
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 12.398324012756348,
+      "learning_rate": 4.9989403106865765e-06,
+      "loss": 1.4979,
+      "step": 20
+    },
+    {
+      "epoch": 0.1024390243902439,
+      "grad_norm": 4.8458123207092285,
+      "learning_rate": 4.9988258379777334e-06,
+      "loss": 0.8265,
+      "step": 21
+    },
+    {
+      "epoch": 0.1073170731707317,
+      "grad_norm": 5.1296844482421875,
+      "learning_rate": 4.998705496757846e-06,
+      "loss": 1.0549,
+      "step": 22
+    },
+    {
+      "epoch": 0.11219512195121951,
+      "grad_norm": 11.290125846862793,
+      "learning_rate": 4.998579287309538e-06,
+      "loss": 1.9388,
+      "step": 23
+    },
+    {
+      "epoch": 0.11707317073170732,
+      "grad_norm": 10.525015830993652,
+      "learning_rate": 4.998447209929211e-06,
+      "loss": 1.4569,
+      "step": 24
+    },
+    {
+      "epoch": 0.12195121951219512,
+      "grad_norm": 1.7889195680618286,
+      "learning_rate": 4.998309264927053e-06,
+      "loss": 0.5703,
+      "step": 25
+    },
+    {
+      "epoch": 0.12682926829268293,
+      "grad_norm": 6.902492523193359,
+      "learning_rate": 4.998165452627025e-06,
+      "loss": 0.8576,
+      "step": 26
+    },
+    {
+      "epoch": 0.13170731707317074,
+      "grad_norm": 11.605420112609863,
+      "learning_rate": 4.998015773366874e-06,
+      "loss": 1.1532,
+      "step": 27
+    },
+    {
+      "epoch": 0.13658536585365855,
+      "grad_norm": 3.192915678024292,
+      "learning_rate": 4.997860227498122e-06,
+      "loss": 0.6788,
+      "step": 28
+    },
+    {
+      "epoch": 0.14146341463414633,
+      "grad_norm": 13.133618354797363,
+      "learning_rate": 4.99769881538607e-06,
+      "loss": 1.5092,
+      "step": 29
+    },
+    {
+      "epoch": 0.14634146341463414,
+      "grad_norm": 12.705998420715332,
+      "learning_rate": 4.997531537409794e-06,
+      "loss": 1.4641,
+      "step": 30
+    },
+    {
+      "epoch": 0.15121951219512195,
+      "grad_norm": 28.609262466430664,
+      "learning_rate": 4.99735839396215e-06,
+      "loss": 0.8727,
+      "step": 31
+    },
+    {
+      "epoch": 0.15609756097560976,
+      "grad_norm": 91.81269836425781,
+      "learning_rate": 4.9971793854497655e-06,
+      "loss": 0.9704,
+      "step": 32
+    },
+    {
+      "epoch": 0.16097560975609757,
+      "grad_norm": 10.507450103759766,
+      "learning_rate": 4.996994512293042e-06,
+      "loss": 1.4177,
+      "step": 33
+    },
+    {
+      "epoch": 0.16585365853658537,
+      "grad_norm": 13.195033073425293,
+      "learning_rate": 4.996803774926157e-06,
+      "loss": 1.1937,
+      "step": 34
+    },
+    {
+      "epoch": 0.17073170731707318,
+      "grad_norm": 2.8658792972564697,
+      "learning_rate": 4.996607173797059e-06,
+      "loss": 1.191,
+      "step": 35
+    },
+    {
+      "epoch": 0.17560975609756097,
+      "grad_norm": 9.105996131896973,
+      "learning_rate": 4.996404709367466e-06,
+      "loss": 1.1722,
+      "step": 36
+    },
+    {
+      "epoch": 0.18048780487804877,
+      "grad_norm": 8.545594215393066,
+      "learning_rate": 4.996196382112868e-06,
+      "loss": 0.8455,
+      "step": 37
+    },
+    {
+      "epoch": 0.18536585365853658,
+      "grad_norm": 14.379878044128418,
+      "learning_rate": 4.9959821925225235e-06,
+      "loss": 1.0889,
+      "step": 38
+    },
+    {
+      "epoch": 0.1902439024390244,
+      "grad_norm": 17.579647064208984,
+      "learning_rate": 4.995762141099456e-06,
+      "loss": 0.9452,
+      "step": 39
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 4.446125507354736,
+      "learning_rate": 4.995536228360461e-06,
+      "loss": 0.9954,
+      "step": 40
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.606940984725952,
+      "learning_rate": 4.995304454836095e-06,
+      "loss": 0.8706,
+      "step": 41
+    },
+    {
+      "epoch": 0.2048780487804878,
+      "grad_norm": 4.837315559387207,
+      "learning_rate": 4.9950668210706795e-06,
+      "loss": 1.0195,
+      "step": 42
+    },
+    {
+      "epoch": 0.2097560975609756,
+      "grad_norm": 7.445003032684326,
+      "learning_rate": 4.994823327622299e-06,
+      "loss": 1.2063,
+      "step": 43
+    },
+    {
+      "epoch": 0.2146341463414634,
+      "grad_norm": 6.544964790344238,
+      "learning_rate": 4.9945739750628e-06,
+      "loss": 0.8807,
+      "step": 44
+    },
+    {
+      "epoch": 0.21951219512195122,
+      "grad_norm": 8.999876022338867,
+      "learning_rate": 4.994318763977789e-06,
+      "loss": 1.0277,
+      "step": 45
+    },
+    {
+      "epoch": 0.22439024390243903,
+      "grad_norm": 9.390464782714844,
+      "learning_rate": 4.994057694966632e-06,
+      "loss": 1.1352,
+      "step": 46
+    },
+    {
+      "epoch": 0.22926829268292684,
+      "grad_norm": 4.313526153564453,
+      "learning_rate": 4.993790768642449e-06,
+      "loss": 1.0033,
+      "step": 47
+    },
+    {
+      "epoch": 0.23414634146341465,
+      "grad_norm": 10.303829193115234,
+      "learning_rate": 4.99351798563212e-06,
+      "loss": 1.6233,
+      "step": 48
+    },
+    {
+      "epoch": 0.23902439024390243,
+      "grad_norm": 4.707674503326416,
+      "learning_rate": 4.993239346576278e-06,
+      "loss": 1.0956,
+      "step": 49
+    },
+    {
+      "epoch": 0.24390243902439024,
+      "grad_norm": 3.8843674659729004,
+      "learning_rate": 4.99295485212931e-06,
+      "loss": 1.236,
+      "step": 50
+    },
+    {
+      "epoch": 0.24878048780487805,
+      "grad_norm": 6.579261779785156,
+      "learning_rate": 4.992664502959351e-06,
+      "loss": 1.1403,
+      "step": 51
+    },
+    {
+      "epoch": 0.25365853658536586,
+      "grad_norm": 1.7781167030334473,
+      "learning_rate": 4.99236829974829e-06,
+      "loss": 0.6883,
+      "step": 52
+    },
+    {
+      "epoch": 0.25853658536585367,
+      "grad_norm": 3.223673105239868,
+      "learning_rate": 4.992066243191762e-06,
+      "loss": 1.0185,
+      "step": 53
+    },
+    {
+      "epoch": 0.2634146341463415,
+      "grad_norm": 2.5291311740875244,
+      "learning_rate": 4.991758333999148e-06,
+      "loss": 0.7823,
+      "step": 54
+    },
+    {
+      "epoch": 0.2682926829268293,
+      "grad_norm": 3.9081308841705322,
+      "learning_rate": 4.991444572893575e-06,
+      "loss": 0.7634,
+      "step": 55
+    },
+    {
+      "epoch": 0.2731707317073171,
+      "grad_norm": 1.7186449766159058,
+      "learning_rate": 4.991124960611916e-06,
+      "loss": 0.5396,
+      "step": 56
+    },
+    {
+      "epoch": 0.2780487804878049,
+      "grad_norm": 3.7625038623809814,
+      "learning_rate": 4.99079949790478e-06,
+      "loss": 0.8568,
+      "step": 57
+    },
+    {
+      "epoch": 0.28292682926829266,
+      "grad_norm": 4.833764553070068,
+      "learning_rate": 4.99046818553652e-06,
+      "loss": 0.9866,
+      "step": 58
+    },
+    {
+      "epoch": 0.28780487804878047,
+      "grad_norm": 17.990612030029297,
+      "learning_rate": 4.9901310242852246e-06,
+      "loss": 1.0425,
+      "step": 59
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 6.425568580627441,
+      "learning_rate": 4.9897880149427206e-06,
+      "loss": 1.085,
+      "step": 60
+    },
+    {
+      "epoch": 0.2975609756097561,
+      "grad_norm": 5.177443027496338,
+      "learning_rate": 4.989439158314566e-06,
+      "loss": 1.0767,
+      "step": 61
+    },
+    {
+      "epoch": 0.3024390243902439,
+      "grad_norm": 2.572352886199951,
+      "learning_rate": 4.989084455220056e-06,
+      "loss": 0.5224,
+      "step": 62
+    },
+    {
+      "epoch": 0.3073170731707317,
+      "grad_norm": 2.277250051498413,
+      "learning_rate": 4.988723906492212e-06,
+      "loss": 0.8736,
+      "step": 63
+    },
+    {
+      "epoch": 0.3121951219512195,
+      "grad_norm": 4.4316935539245605,
+      "learning_rate": 4.988357512977785e-06,
+      "loss": 0.5605,
+      "step": 64
+    },
+    {
+      "epoch": 0.3170731707317073,
+      "grad_norm": 2.4577624797821045,
+      "learning_rate": 4.987985275537252e-06,
+      "loss": 0.5415,
+      "step": 65
+    },
+    {
+      "epoch": 0.32195121951219513,
+      "grad_norm": 7.362552165985107,
+      "learning_rate": 4.9876071950448185e-06,
+      "loss": 0.9167,
+      "step": 66
+    },
+    {
+      "epoch": 0.32682926829268294,
+      "grad_norm": 2.687702178955078,
+      "learning_rate": 4.987223272388407e-06,
+      "loss": 0.6806,
+      "step": 67
+    },
+    {
+      "epoch": 0.33170731707317075,
+      "grad_norm": 4.988495349884033,
+      "learning_rate": 4.986833508469663e-06,
+      "loss": 0.9808,
+      "step": 68
+    },
+    {
+      "epoch": 0.33658536585365856,
+      "grad_norm": 4.317763805389404,
+      "learning_rate": 4.98643790420395e-06,
+      "loss": 0.8506,
+      "step": 69
+    },
+    {
+      "epoch": 0.34146341463414637,
+      "grad_norm": 3.347094774246216,
+      "learning_rate": 4.986036460520348e-06,
+      "loss": 0.8512,
+      "step": 70
+    },
+    {
+      "epoch": 0.3463414634146341,
+      "grad_norm": 6.8587775230407715,
+      "learning_rate": 4.98562917836165e-06,
+      "loss": 1.1533,
+      "step": 71
+    },
+    {
+      "epoch": 0.35121951219512193,
+      "grad_norm": 3.214782238006592,
+      "learning_rate": 4.985216058684362e-06,
+      "loss": 0.7506,
+      "step": 72
+    },
+    {
+      "epoch": 0.35609756097560974,
+      "grad_norm": 2.6237006187438965,
+      "learning_rate": 4.984797102458697e-06,
+      "loss": 1.0345,
+      "step": 73
+    },
+    {
+      "epoch": 0.36097560975609755,
+      "grad_norm": 1.8713479042053223,
+      "learning_rate": 4.984372310668579e-06,
+      "loss": 0.5752,
+      "step": 74
+    },
+    {
+      "epoch": 0.36585365853658536,
+      "grad_norm": 2.518427610397339,
+      "learning_rate": 4.983941684311633e-06,
+      "loss": 1.2944,
+      "step": 75
+    },
+    {
+      "epoch": 0.37073170731707317,
+      "grad_norm": 3.947360038757324,
+      "learning_rate": 4.983505224399188e-06,
+      "loss": 0.8142,
+      "step": 76
+    },
+    {
+      "epoch": 0.375609756097561,
+      "grad_norm": 3.0242512226104736,
+      "learning_rate": 4.983062931956275e-06,
+      "loss": 0.8578,
+      "step": 77
+    },
+    {
+      "epoch": 0.3804878048780488,
+      "grad_norm": 2.711573600769043,
+      "learning_rate": 4.9826148080216195e-06,
+      "loss": 0.9262,
+      "step": 78
+    },
+    {
+      "epoch": 0.3853658536585366,
+      "grad_norm": 2.975578784942627,
+      "learning_rate": 4.9821608536476445e-06,
+      "loss": 1.2846,
+      "step": 79
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 4.4119415283203125,
+      "learning_rate": 4.981701069900465e-06,
+      "loss": 0.942,
+      "step": 80
+    },
+    {
+      "epoch": 0.3951219512195122,
+      "grad_norm": 3.3930721282958984,
+      "learning_rate": 4.9812354578598876e-06,
+      "loss": 0.7599,
+      "step": 81
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.192596435546875,
+      "learning_rate": 4.980764018619405e-06,
+      "loss": 0.7835,
+      "step": 82
+    },
+    {
+      "epoch": 0.40487804878048783,
+      "grad_norm": 3.237403631210327,
+      "learning_rate": 4.980286753286196e-06,
+      "loss": 0.9365,
+      "step": 83
+    },
+    {
+      "epoch": 0.4097560975609756,
+      "grad_norm": 2.1691837310791016,
+      "learning_rate": 4.97980366298112e-06,
+      "loss": 0.782,
+      "step": 84
+    },
+    {
+      "epoch": 0.4146341463414634,
+      "grad_norm": 2.8383233547210693,
+      "learning_rate": 4.97931474883872e-06,
+      "loss": 0.8006,
+      "step": 85
+    },
+    {
+      "epoch": 0.4195121951219512,
+      "grad_norm": 2.01986026763916,
+      "learning_rate": 4.978820012007213e-06,
+      "loss": 0.8591,
+      "step": 86
+    },
+    {
+      "epoch": 0.424390243902439,
+      "grad_norm": 2.3000314235687256,
+      "learning_rate": 4.978319453648495e-06,
+      "loss": 0.9201,
+      "step": 87
+    },
+    {
+      "epoch": 0.4292682926829268,
+      "grad_norm": 2.3446927070617676,
+      "learning_rate": 4.977813074938128e-06,
+      "loss": 0.8661,
+      "step": 88
+    },
+    {
+      "epoch": 0.43414634146341463,
+      "grad_norm": 2.113168239593506,
+      "learning_rate": 4.977300877065347e-06,
+      "loss": 0.8441,
+      "step": 89
+    },
+    {
+      "epoch": 0.43902439024390244,
+      "grad_norm": 2.4370508193969727,
+      "learning_rate": 4.976782861233053e-06,
+      "loss": 0.6732,
+      "step": 90
+    },
+    {
+      "epoch": 0.44390243902439025,
+      "grad_norm": 3.2290725708007812,
+      "learning_rate": 4.976259028657812e-06,
+      "loss": 0.6321,
+      "step": 91
+    },
+    {
+      "epoch": 0.44878048780487806,
+      "grad_norm": 4.33652925491333,
+      "learning_rate": 4.975729380569845e-06,
+      "loss": 0.7268,
+      "step": 92
+    },
+    {
+      "epoch": 0.45365853658536587,
+      "grad_norm": 1.8476589918136597,
+      "learning_rate": 4.975193918213035e-06,
+      "loss": 0.465,
+      "step": 93
+    },
+    {
+      "epoch": 0.4585365853658537,
+      "grad_norm": 1.528070092201233,
+      "learning_rate": 4.974652642844921e-06,
+      "loss": 0.6325,
+      "step": 94
+    },
+    {
+      "epoch": 0.4634146341463415,
+      "grad_norm": 3.360761880874634,
+      "learning_rate": 4.974105555736693e-06,
+      "loss": 1.0014,
+      "step": 95
+    },
+    {
+      "epoch": 0.4682926829268293,
+      "grad_norm": 1.6910600662231445,
+      "learning_rate": 4.973552658173186e-06,
+      "loss": 0.6367,
+      "step": 96
+    },
+    {
+      "epoch": 0.47317073170731705,
+      "grad_norm": 2.9577507972717285,
+      "learning_rate": 4.972993951452887e-06,
+      "loss": 0.8256,
+      "step": 97
+    },
+    {
+      "epoch": 0.47804878048780486,
+      "grad_norm": 2.4482178688049316,
+      "learning_rate": 4.9724294368879214e-06,
+      "loss": 0.7554,
+      "step": 98
+    },
+    {
+      "epoch": 0.48292682926829267,
+      "grad_norm": 2.077953815460205,
+      "learning_rate": 4.971859115804055e-06,
+      "loss": 0.9382,
+      "step": 99
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 2.2485179901123047,
+      "learning_rate": 4.9712829895406935e-06,
+      "loss": 0.7396,
+      "step": 100
+    },
+    {
+      "epoch": 0.4926829268292683,
+      "grad_norm": 2.3033151626586914,
+      "learning_rate": 4.970701059450872e-06,
+      "loss": 0.5185,
+      "step": 101
+    },
+    {
+      "epoch": 0.4975609756097561,
+      "grad_norm": 1.8318403959274292,
+      "learning_rate": 4.970113326901258e-06,
+      "loss": 0.8857,
+      "step": 102
+    },
+    {
+      "epoch": 0.5024390243902439,
+      "grad_norm": 1.961539626121521,
+      "learning_rate": 4.9695197932721455e-06,
+      "loss": 0.8248,
+      "step": 103
+    },
+    {
+      "epoch": 0.5073170731707317,
+      "grad_norm": 2.156985282897949,
+      "learning_rate": 4.968920459957453e-06,
+      "loss": 0.8533,
+      "step": 104
+    },
+    {
+      "epoch": 0.5121951219512195,
+      "grad_norm": 2.9577434062957764,
+      "learning_rate": 4.968315328364719e-06,
+      "loss": 0.9013,
+      "step": 105
+    },
+    {
+      "epoch": 0.5170731707317073,
+      "grad_norm": 1.8476804494857788,
+      "learning_rate": 4.9677043999151e-06,
+      "loss": 1.0077,
+      "step": 106
+    },
+    {
+      "epoch": 0.5219512195121951,
+      "grad_norm": 1.9346246719360352,
+      "learning_rate": 4.967087676043366e-06,
+      "loss": 0.4618,
+      "step": 107
+    },
+    {
+      "epoch": 0.526829268292683,
+      "grad_norm": 2.0548758506774902,
+      "learning_rate": 4.966465158197897e-06,
+      "loss": 0.8305,
+      "step": 108
+    },
+    {
+      "epoch": 0.5317073170731708,
+      "grad_norm": 2.133976697921753,
+      "learning_rate": 4.965836847840681e-06,
+      "loss": 0.6311,
+      "step": 109
+    },
+    {
+      "epoch": 0.5365853658536586,
+      "grad_norm": 2.1547298431396484,
+      "learning_rate": 4.96520274644731e-06,
+      "loss": 0.8104,
+      "step": 110
+    },
+    {
+      "epoch": 0.5414634146341464,
+      "grad_norm": 1.4163529872894287,
+      "learning_rate": 4.964562855506976e-06,
+      "loss": 0.636,
+      "step": 111
+    },
+    {
+      "epoch": 0.5463414634146342,
+      "grad_norm": 1.8655451536178589,
+      "learning_rate": 4.963917176522466e-06,
+      "loss": 0.5861,
+      "step": 112
+    },
+    {
+      "epoch": 0.551219512195122,
+      "grad_norm": 2.7667734622955322,
+      "learning_rate": 4.963265711010164e-06,
+      "loss": 0.9409,
+      "step": 113
+    },
+    {
+      "epoch": 0.5560975609756098,
+      "grad_norm": 3.6238114833831787,
+      "learning_rate": 4.9626084605000395e-06,
+      "loss": 0.771,
+      "step": 114
+    },
+    {
+      "epoch": 0.5609756097560976,
+      "grad_norm": 2.039236545562744,
+      "learning_rate": 4.961945426535652e-06,
+      "loss": 0.5164,
+      "step": 115
+    },
+    {
+      "epoch": 0.5658536585365853,
+      "grad_norm": 2.7303946018218994,
+      "learning_rate": 4.961276610674141e-06,
+      "loss": 0.8206,
+      "step": 116
+    },
+    {
+      "epoch": 0.5707317073170731,
+      "grad_norm": 1.6661245822906494,
+      "learning_rate": 4.960602014486225e-06,
+      "loss": 0.8942,
+      "step": 117
+    },
+    {
+      "epoch": 0.5756097560975609,
+      "grad_norm": 1.6690449714660645,
+      "learning_rate": 4.959921639556199e-06,
+      "loss": 0.6978,
+      "step": 118
+    },
+    {
+      "epoch": 0.5804878048780487,
+      "grad_norm": 1.6131573915481567,
+      "learning_rate": 4.959235487481928e-06,
+      "loss": 0.8862,
+      "step": 119
+    },
+    {
+      "epoch": 0.5853658536585366,
+      "grad_norm": 1.6374850273132324,
+      "learning_rate": 4.958543559874846e-06,
+      "loss": 0.4928,
+      "step": 120
+    },
+    {
+      "epoch": 0.5902439024390244,
+      "grad_norm": 1.8508459329605103,
+      "learning_rate": 4.9578458583599495e-06,
+      "loss": 0.718,
+      "step": 121
+    },
+    {
+      "epoch": 0.5951219512195122,
+      "grad_norm": 1.9640874862670898,
+      "learning_rate": 4.957142384575795e-06,
+      "loss": 0.6702,
+      "step": 122
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5142700672149658,
+      "learning_rate": 4.956433140174498e-06,
+      "loss": 0.7402,
+      "step": 123
+    },
+    {
+      "epoch": 0.6048780487804878,
+      "grad_norm": 1.8278406858444214,
+      "learning_rate": 4.9557181268217225e-06,
+      "loss": 0.7641,
+      "step": 124
+    },
+    {
+      "epoch": 0.6097560975609756,
+      "grad_norm": 1.6352994441986084,
+      "learning_rate": 4.954997346196683e-06,
+      "loss": 1.1429,
+      "step": 125
+    },
+    {
+      "epoch": 0.6146341463414634,
+      "grad_norm": 1.539684772491455,
+      "learning_rate": 4.954270799992138e-06,
+      "loss": 0.6404,
+      "step": 126
+    },
+    {
+      "epoch": 0.6195121951219512,
+      "grad_norm": 1.8513871431350708,
+      "learning_rate": 4.953538489914387e-06,
+      "loss": 0.6615,
+      "step": 127
+    },
+    {
+      "epoch": 0.624390243902439,
+      "grad_norm": 2.163950204849243,
+      "learning_rate": 4.9528004176832654e-06,
+      "loss": 0.5475,
+      "step": 128
+    },
+    {
+      "epoch": 0.6292682926829268,
+      "grad_norm": 2.240516185760498,
+      "learning_rate": 4.952056585032142e-06,
+      "loss": 0.9359,
+      "step": 129
+    },
+    {
+      "epoch": 0.6341463414634146,
+      "grad_norm": 1.5975580215454102,
+      "learning_rate": 4.951306993707913e-06,
+      "loss": 0.7296,
+      "step": 130
+    },
+    {
+      "epoch": 0.6390243902439025,
+      "grad_norm": 1.6257214546203613,
+      "learning_rate": 4.950551645470998e-06,
+      "loss": 0.6884,
+      "step": 131
+    },
+    {
+      "epoch": 0.6439024390243903,
+      "grad_norm": 1.7229961156845093,
+      "learning_rate": 4.9497905420953406e-06,
+      "loss": 0.6359,
+      "step": 132
+    },
+    {
+      "epoch": 0.6487804878048781,
+      "grad_norm": 1.9853991270065308,
+      "learning_rate": 4.949023685368395e-06,
+      "loss": 0.6608,
+      "step": 133
+    },
+    {
+      "epoch": 0.6536585365853659,
+      "grad_norm": 2.1698012351989746,
+      "learning_rate": 4.948251077091131e-06,
+      "loss": 0.9688,
+      "step": 134
+    },
+    {
+      "epoch": 0.6585365853658537,
+      "grad_norm": 2.5448765754699707,
+      "learning_rate": 4.947472719078025e-06,
+      "loss": 0.7493,
+      "step": 135
+    },
+    {
+      "epoch": 0.6634146341463415,
+      "grad_norm": 1.4268618822097778,
+      "learning_rate": 4.9466886131570565e-06,
+      "loss": 0.7836,
+      "step": 136
+    },
+    {
+      "epoch": 0.6682926829268293,
+      "grad_norm": 1.5412968397140503,
+      "learning_rate": 4.945898761169704e-06,
+      "loss": 0.9513,
+      "step": 137
+    },
+    {
+      "epoch": 0.6731707317073171,
+      "grad_norm": 1.6611838340759277,
+      "learning_rate": 4.945103164970941e-06,
+      "loss": 0.548,
+      "step": 138
+    },
+    {
+      "epoch": 0.6780487804878049,
+      "grad_norm": 1.582939863204956,
+      "learning_rate": 4.9443018264292304e-06,
+      "loss": 0.6658,
+      "step": 139
+    },
+    {
+      "epoch": 0.6829268292682927,
+      "grad_norm": 2.266261100769043,
+      "learning_rate": 4.9434947474265225e-06,
+      "loss": 0.827,
+      "step": 140
+    },
+    {
+      "epoch": 0.6878048780487804,
+      "grad_norm": 1.6282507181167603,
+      "learning_rate": 4.942681929858249e-06,
+      "loss": 0.9394,
+      "step": 141
+    },
+    {
+      "epoch": 0.6926829268292682,
+      "grad_norm": 8.833651542663574,
+      "learning_rate": 4.941863375633315e-06,
+      "loss": 0.8216,
+      "step": 142
+    },
+    {
+      "epoch": 0.697560975609756,
+      "grad_norm": 1.7979482412338257,
+      "learning_rate": 4.9410390866741056e-06,
+      "loss": 0.686,
+      "step": 143
+    },
+    {
+      "epoch": 0.7024390243902439,
+      "grad_norm": 1.6304007768630981,
+      "learning_rate": 4.9402090649164655e-06,
+      "loss": 0.6699,
+      "step": 144
+    },
+    {
+      "epoch": 0.7073170731707317,
+      "grad_norm": 2.643566846847534,
+      "learning_rate": 4.9393733123097085e-06,
+      "loss": 0.9212,
+      "step": 145
+    },
+    {
+      "epoch": 0.7121951219512195,
+      "grad_norm": 2.0324487686157227,
+      "learning_rate": 4.9385318308166065e-06,
+      "loss": 0.7961,
+      "step": 146
+    },
+    {
+      "epoch": 0.7170731707317073,
+      "grad_norm": 1.8635185956954956,
+      "learning_rate": 4.937684622413385e-06,
+      "loss": 0.5226,
+      "step": 147
+    },
+    {
+      "epoch": 0.7219512195121951,
+      "grad_norm": 2.0762131214141846,
+      "learning_rate": 4.9368316890897185e-06,
+      "loss": 0.8788,
+      "step": 148
+    },
+    {
+      "epoch": 0.7268292682926829,
+      "grad_norm": 1.3568695783615112,
+      "learning_rate": 4.9359730328487264e-06,
+      "loss": 0.4686,
+      "step": 149
+    },
+    {
+      "epoch": 0.7317073170731707,
+      "grad_norm": 1.3983228206634521,
+      "learning_rate": 4.935108655706972e-06,
+      "loss": 0.7091,
+      "step": 150
+    },
+    {
+      "epoch": 0.7365853658536585,
+      "grad_norm": 1.679435133934021,
+      "learning_rate": 4.934238559694448e-06,
+      "loss": 0.9544,
+      "step": 151
+    },
+    {
+      "epoch": 0.7414634146341463,
+      "grad_norm": 2.03104305267334,
+      "learning_rate": 4.9333627468545845e-06,
+      "loss": 0.6214,
+      "step": 152
+    },
+    {
+      "epoch": 0.7463414634146341,
+      "grad_norm": 1.6827820539474487,
+      "learning_rate": 4.932481219244231e-06,
+      "loss": 0.7024,
+      "step": 153
+    },
+    {
+      "epoch": 0.751219512195122,
+      "grad_norm": 1.92439603805542,
+      "learning_rate": 4.931593978933666e-06,
+      "loss": 0.6718,
+      "step": 154
+    },
+    {
+      "epoch": 0.7560975609756098,
+      "grad_norm": 1.4603160619735718,
+      "learning_rate": 4.930701028006577e-06,
+      "loss": 0.843,
+      "step": 155
+    },
+    {
+      "epoch": 0.7609756097560976,
+      "grad_norm": 2.8143606185913086,
+      "learning_rate": 4.929802368560066e-06,
+      "loss": 0.6528,
+      "step": 156
+    },
+    {
+      "epoch": 0.7658536585365854,
+      "grad_norm": 1.8976019620895386,
+      "learning_rate": 4.928898002704642e-06,
+      "loss": 0.7805,
+      "step": 157
+    },
+    {
+      "epoch": 0.7707317073170732,
+      "grad_norm": 1.811642050743103,
+      "learning_rate": 4.927987932564215e-06,
+      "loss": 0.6496,
+      "step": 158
+    },
+    {
+      "epoch": 0.775609756097561,
+      "grad_norm": 1.4345223903656006,
+      "learning_rate": 4.927072160276092e-06,
+      "loss": 0.6856,
+      "step": 159
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 3.1401326656341553,
+      "learning_rate": 4.926150687990969e-06,
+      "loss": 0.5397,
+      "step": 160
+    },
+    {
+      "epoch": 0.7853658536585366,
+      "grad_norm": 2.784783124923706,
+      "learning_rate": 4.925223517872934e-06,
+      "loss": 0.8164,
+      "step": 161
+    },
+    {
+      "epoch": 0.7902439024390244,
+      "grad_norm": 2.388042449951172,
+      "learning_rate": 4.9242906520994484e-06,
+      "loss": 0.8993,
+      "step": 162
+    },
+    {
+      "epoch": 0.7951219512195122,
+      "grad_norm": 1.573328971862793,
+      "learning_rate": 4.923352092861358e-06,
+      "loss": 0.754,
+      "step": 163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.765645980834961,
+      "learning_rate": 4.922407842362875e-06,
+      "loss": 0.5589,
+      "step": 164
+    },
+    {
+      "epoch": 0.8048780487804879,
+      "grad_norm": 1.603736400604248,
+      "learning_rate": 4.921457902821578e-06,
+      "loss": 0.8421,
+      "step": 165
+    },
+    {
+      "epoch": 0.8097560975609757,
+      "grad_norm": 1.7012734413146973,
+      "learning_rate": 4.920502276468408e-06,
+      "loss": 0.7829,
+      "step": 166
+    },
+    {
+      "epoch": 0.8146341463414634,
+      "grad_norm": 1.4824273586273193,
+      "learning_rate": 4.9195409655476605e-06,
+      "loss": 0.6344,
+      "step": 167
+    },
+    {
+      "epoch": 0.8195121951219512,
+      "grad_norm": 1.7648383378982544,
+      "learning_rate": 4.918573972316982e-06,
+      "loss": 0.8857,
+      "step": 168
+    },
+    {
+      "epoch": 0.824390243902439,
+      "grad_norm": 1.6718090772628784,
+      "learning_rate": 4.917601299047361e-06,
+      "loss": 0.7153,
+      "step": 169
+    },
+    {
+      "epoch": 0.8292682926829268,
+      "grad_norm": 1.5050429105758667,
+      "learning_rate": 4.916622948023129e-06,
+      "loss": 0.6855,
+      "step": 170
+    },
+    {
+      "epoch": 0.8341463414634146,
+      "grad_norm": 1.5966042280197144,
+      "learning_rate": 4.915638921541952e-06,
+      "loss": 0.5559,
+      "step": 171
+    },
+    {
+      "epoch": 0.8390243902439024,
+      "grad_norm": 2.0568573474884033,
+      "learning_rate": 4.914649221914822e-06,
+      "loss": 0.7972,
+      "step": 172
+    },
+    {
+      "epoch": 0.8439024390243902,
+      "grad_norm": 1.8847912549972534,
+      "learning_rate": 4.913653851466057e-06,
+      "loss": 0.572,
+      "step": 173
+    },
+    {
+      "epoch": 0.848780487804878,
+      "grad_norm": 1.6164286136627197,
+      "learning_rate": 4.912652812533291e-06,
+      "loss": 0.7195,
+      "step": 174
+    },
+    {
+      "epoch": 0.8536585365853658,
+      "grad_norm": 1.5945544242858887,
+      "learning_rate": 4.911646107467472e-06,
+      "loss": 0.7432,
+      "step": 175
+    },
+    {
+      "epoch": 0.8585365853658536,
+      "grad_norm": 1.3156569004058838,
+      "learning_rate": 4.9106337386328524e-06,
+      "loss": 0.8753,
+      "step": 176
+    },
+    {
+      "epoch": 0.8634146341463415,
+      "grad_norm": 1.6262340545654297,
+      "learning_rate": 4.909615708406991e-06,
+      "loss": 0.7642,
+      "step": 177
+    },
+    {
+      "epoch": 0.8682926829268293,
+      "grad_norm": 1.4175578355789185,
+      "learning_rate": 4.908592019180738e-06,
+      "loss": 0.5936,
+      "step": 178
+    },
+    {
+      "epoch": 0.8731707317073171,
+      "grad_norm": 1.1894031763076782,
+      "learning_rate": 4.907562673358234e-06,
+      "loss": 0.5751,
+      "step": 179
+    },
+    {
+      "epoch": 0.8780487804878049,
+      "grad_norm": 1.251502275466919,
+      "learning_rate": 4.906527673356907e-06,
+      "loss": 0.5285,
+      "step": 180
+    },
+    {
+      "epoch": 0.8829268292682927,
+      "grad_norm": 1.7352757453918457,
+      "learning_rate": 4.905487021607462e-06,
+      "loss": 0.4379,
+      "step": 181
+    },
+    {
+      "epoch": 0.8878048780487805,
+      "grad_norm": 1.9421669244766235,
+      "learning_rate": 4.904440720553876e-06,
+      "loss": 0.7487,
+      "step": 182
+    },
+    {
+      "epoch": 0.8926829268292683,
+      "grad_norm": 1.370103120803833,
+      "learning_rate": 4.903388772653396e-06,
+      "loss": 0.6868,
+      "step": 183
+    },
+    {
+      "epoch": 0.8975609756097561,
+      "grad_norm": 1.7440688610076904,
+      "learning_rate": 4.902331180376529e-06,
+      "loss": 0.7203,
+      "step": 184
+    },
+    {
+      "epoch": 0.9024390243902439,
+      "grad_norm": 1.4112497568130493,
+      "learning_rate": 4.901267946207038e-06,
+      "loss": 0.8253,
+      "step": 185
+    },
+    {
+      "epoch": 0.9073170731707317,
+      "grad_norm": 1.677984356880188,
+      "learning_rate": 4.900199072641937e-06,
+      "loss": 0.6319,
+      "step": 186
+    },
+    {
+      "epoch": 0.9121951219512195,
+      "grad_norm": 1.313682198524475,
+      "learning_rate": 4.899124562191484e-06,
+      "loss": 0.5418,
+      "step": 187
+    },
+    {
+      "epoch": 0.9170731707317074,
+      "grad_norm": 1.233353853225708,
+      "learning_rate": 4.8980444173791735e-06,
+      "loss": 0.4473,
+      "step": 188
+    },
+    {
+      "epoch": 0.9219512195121952,
+      "grad_norm": 1.7652136087417603,
+      "learning_rate": 4.896958640741735e-06,
+      "loss": 0.765,
+      "step": 189
+    },
+    {
+      "epoch": 0.926829268292683,
+      "grad_norm": 1.6560442447662354,
+      "learning_rate": 4.895867234829121e-06,
+      "loss": 0.8341,
+      "step": 190
+    },
+    {
+      "epoch": 0.9317073170731708,
+      "grad_norm": 1.4490306377410889,
+      "learning_rate": 4.894770202204509e-06,
+      "loss": 0.6267,
+      "step": 191
+    },
+    {
+      "epoch": 0.9365853658536586,
+      "grad_norm": 2.1652960777282715,
+      "learning_rate": 4.893667545444285e-06,
+      "loss": 0.75,
+      "step": 192
+    },
+    {
+      "epoch": 0.9414634146341463,
+      "grad_norm": 1.3940919637680054,
+      "learning_rate": 4.8925592671380495e-06,
+      "loss": 0.599,
+      "step": 193
+    },
+    {
+      "epoch": 0.9463414634146341,
+      "grad_norm": 1.0715932846069336,
+      "learning_rate": 4.891445369888601e-06,
+      "loss": 0.4761,
+      "step": 194
+    },
+    {
+      "epoch": 0.9512195121951219,
+      "grad_norm": 1.3872371912002563,
+      "learning_rate": 4.890325856311936e-06,
+      "loss": 0.7884,
+      "step": 195
+    },
+    {
+      "epoch": 0.9560975609756097,
+      "grad_norm": 1.485743522644043,
+      "learning_rate": 4.889200729037241e-06,
+      "loss": 0.7216,
+      "step": 196
+    },
+    {
+      "epoch": 0.9609756097560975,
+      "grad_norm": 1.3263616561889648,
+      "learning_rate": 4.888069990706884e-06,
+      "loss": 0.5775,
+      "step": 197
+    },
+    {
+      "epoch": 0.9658536585365853,
+      "grad_norm": 3.2764222621917725,
+      "learning_rate": 4.886933643976414e-06,
+      "loss": 0.4755,
+      "step": 198
+    },
+    {
+      "epoch": 0.9707317073170731,
+      "grad_norm": 1.3251742124557495,
+      "learning_rate": 4.885791691514548e-06,
+      "loss": 0.4295,
+      "step": 199
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 1.2534489631652832,
+      "learning_rate": 4.884644136003172e-06,
+      "loss": 0.5451,
+      "step": 200
+    },
+    {
+      "epoch": 0.9804878048780488,
+      "grad_norm": 1.5479427576065063,
+      "learning_rate": 4.883490980137327e-06,
+      "loss": 1.1673,
+      "step": 201
+    },
+    {
+      "epoch": 0.9853658536585366,
+      "grad_norm": 1.5993766784667969,
+      "learning_rate": 4.882332226625208e-06,
+      "loss": 0.6849,
+      "step": 202
+    },
+    {
+      "epoch": 0.9902439024390244,
+      "grad_norm": 1.1724447011947632,
+      "learning_rate": 4.881167878188158e-06,
+      "loss": 0.7465,
+      "step": 203
+    },
+    {
+      "epoch": 0.9951219512195122,
+      "grad_norm": 1.5650769472122192,
+      "learning_rate": 4.8799979375606565e-06,
+      "loss": 0.6518,
+      "step": 204
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4839268922805786,
+      "learning_rate": 4.878822407490319e-06,
+      "loss": 0.6304,
+      "step": 205
+    },
+    {
+      "epoch": 1.0048780487804878,
+      "grad_norm": 2.099705457687378,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 0.5868,
+      "step": 206
+    },
+    {
+      "epoch": 1.0097560975609756,
+      "grad_norm": 1.1467387676239014,
+      "learning_rate": 4.876454590077216e-06,
+      "loss": 0.5253,
+      "step": 207
+    },
+    {
+      "epoch": 1.0146341463414634,
+      "grad_norm": 1.2665221691131592,
+      "learning_rate": 4.875262308295289e-06,
+      "loss": 0.7341,
+      "step": 208
+    },
+    {
+      "epoch": 1.0195121951219512,
+      "grad_norm": 1.6670607328414917,
+      "learning_rate": 4.874064448192185e-06,
+      "loss": 0.6647,
+      "step": 209
+    },
+    {
+      "epoch": 1.024390243902439,
+      "grad_norm": 1.4426093101501465,
+      "learning_rate": 4.872861012581088e-06,
+      "loss": 0.5557,
+      "step": 210
+    },
+    {
+      "epoch": 1.0292682926829269,
+      "grad_norm": 1.6165494918823242,
+      "learning_rate": 4.871652004288275e-06,
+      "loss": 0.6041,
+      "step": 211
+    },
+    {
+      "epoch": 1.0341463414634147,
+      "grad_norm": 1.3780732154846191,
+      "learning_rate": 4.870437426153113e-06,
+      "loss": 0.6103,
+      "step": 212
+    },
+    {
+      "epoch": 1.0390243902439025,
+      "grad_norm": 1.434518814086914,
+      "learning_rate": 4.869217281028045e-06,
+      "loss": 0.7992,
+      "step": 213
+    },
+    {
+      "epoch": 1.0439024390243903,
+      "grad_norm": 2.4002304077148438,
+      "learning_rate": 4.867991571778592e-06,
+      "loss": 0.7997,
+      "step": 214
+    },
+    {
+      "epoch": 1.048780487804878,
+      "grad_norm": 1.2279268503189087,
+      "learning_rate": 4.866760301283342e-06,
+      "loss": 0.4964,
+      "step": 215
+    },
+    {
+      "epoch": 1.053658536585366,
+      "grad_norm": 1.5353184938430786,
+      "learning_rate": 4.865523472433942e-06,
+      "loss": 0.6411,
+      "step": 216
+    },
+    {
+      "epoch": 1.0585365853658537,
+      "grad_norm": 1.1968104839324951,
+      "learning_rate": 4.8642810881350935e-06,
+      "loss": 0.5618,
+      "step": 217
+    },
+    {
+      "epoch": 1.0634146341463415,
+      "grad_norm": 1.1698020696640015,
+      "learning_rate": 4.863033151304546e-06,
+      "loss": 0.5717,
+      "step": 218
+    },
+    {
+      "epoch": 1.0682926829268293,
+      "grad_norm": 2.3843119144439697,
+      "learning_rate": 4.861779664873088e-06,
+      "loss": 0.7808,
+      "step": 219
+    },
+    {
+      "epoch": 1.0731707317073171,
+      "grad_norm": 2.2874321937561035,
+      "learning_rate": 4.8605206317845425e-06,
+      "loss": 0.7433,
+      "step": 220
+    },
+    {
+      "epoch": 1.078048780487805,
+      "grad_norm": 1.626595139503479,
+      "learning_rate": 4.859256054995758e-06,
+      "loss": 0.7767,
+      "step": 221
+    },
+    {
+      "epoch": 1.0829268292682928,
+      "grad_norm": 1.2185815572738647,
+      "learning_rate": 4.8579859374766e-06,
+      "loss": 0.4319,
+      "step": 222
+    },
+    {
+      "epoch": 1.0878048780487806,
+      "grad_norm": 0.9963337779045105,
+      "learning_rate": 4.856710282209952e-06,
+      "loss": 0.3698,
+      "step": 223
+    },
+    {
+      "epoch": 1.0926829268292684,
+      "grad_norm": 1.3038779497146606,
+      "learning_rate": 4.855429092191698e-06,
+      "loss": 0.65,
+      "step": 224
+    },
+    {
+      "epoch": 1.0975609756097562,
+      "grad_norm": 1.1482938528060913,
+      "learning_rate": 4.854142370430725e-06,
+      "loss": 0.5682,
+      "step": 225
+    },
+    {
+      "epoch": 1.102439024390244,
+      "grad_norm": 1.3957362174987793,
+      "learning_rate": 4.8528501199489045e-06,
+      "loss": 0.6119,
+      "step": 226
+    },
+    {
+      "epoch": 1.1073170731707318,
+      "grad_norm": 1.5577399730682373,
+      "learning_rate": 4.851552343781099e-06,
+      "loss": 0.7668,
+      "step": 227
+    },
+    {
+      "epoch": 1.1121951219512196,
+      "grad_norm": 1.3929314613342285,
+      "learning_rate": 4.850249044975145e-06,
+      "loss": 0.7011,
+      "step": 228
+    },
+    {
+      "epoch": 1.1170731707317074,
+      "grad_norm": 1.411049723625183,
+      "learning_rate": 4.848940226591849e-06,
+      "loss": 0.9282,
+      "step": 229
+    },
+    {
+      "epoch": 1.1219512195121952,
+      "grad_norm": 1.1557923555374146,
+      "learning_rate": 4.847625891704982e-06,
+      "loss": 0.5376,
+      "step": 230
+    },
+    {
+      "epoch": 1.126829268292683,
+      "grad_norm": 1.7049798965454102,
+      "learning_rate": 4.846306043401268e-06,
+      "loss": 0.6256,
+      "step": 231
+    },
+    {
+      "epoch": 1.1317073170731708,
+      "grad_norm": 1.57964289188385,
+      "learning_rate": 4.844980684780381e-06,
+      "loss": 0.5689,
+      "step": 232
+    },
+    {
+      "epoch": 1.1365853658536587,
+      "grad_norm": 1.4652674198150635,
+      "learning_rate": 4.8436498189549345e-06,
+      "loss": 0.5314,
+      "step": 233
+    },
+    {
+      "epoch": 1.1414634146341462,
+      "grad_norm": 1.155045747756958,
+      "learning_rate": 4.842313449050477e-06,
+      "loss": 0.523,
+      "step": 234
+    },
+    {
+      "epoch": 1.146341463414634,
+      "grad_norm": 1.0897397994995117,
+      "learning_rate": 4.840971578205486e-06,
+      "loss": 0.462,
+      "step": 235
+    },
+    {
+      "epoch": 1.1512195121951219,
+      "grad_norm": 1.1332272291183472,
+      "learning_rate": 4.839624209571352e-06,
+      "loss": 0.3589,
+      "step": 236
+    },
+    {
+      "epoch": 1.1560975609756097,
+      "grad_norm": 1.314063549041748,
+      "learning_rate": 4.838271346312381e-06,
+      "loss": 0.7708,
+      "step": 237
+    },
+    {
+      "epoch": 1.1609756097560975,
+      "grad_norm": 1.2887887954711914,
+      "learning_rate": 4.836912991605782e-06,
+      "loss": 0.7811,
+      "step": 238
+    },
+    {
+      "epoch": 1.1658536585365853,
+      "grad_norm": 1.3599913120269775,
+      "learning_rate": 4.835549148641663e-06,
+      "loss": 0.5047,
+      "step": 239
+    },
+    {
+      "epoch": 1.170731707317073,
+      "grad_norm": 1.2067794799804688,
+      "learning_rate": 4.834179820623018e-06,
+      "loss": 0.5766,
+      "step": 240
+    },
+    {
+      "epoch": 1.175609756097561,
+      "grad_norm": 1.1860817670822144,
+      "learning_rate": 4.832805010765724e-06,
+      "loss": 0.5397,
+      "step": 241
+    },
+    {
+      "epoch": 1.1804878048780487,
+      "grad_norm": 1.614708662033081,
+      "learning_rate": 4.831424722298531e-06,
+      "loss": 0.6065,
+      "step": 242
+    },
+    {
+      "epoch": 1.1853658536585365,
+      "grad_norm": 1.8012770414352417,
+      "learning_rate": 4.830038958463061e-06,
+      "loss": 0.6891,
+      "step": 243
+    },
+    {
+      "epoch": 1.1902439024390243,
+      "grad_norm": 1.695419192314148,
+      "learning_rate": 4.828647722513785e-06,
+      "loss": 0.8272,
+      "step": 244
+    },
+    {
+      "epoch": 1.1951219512195121,
+      "grad_norm": 1.6885395050048828,
+      "learning_rate": 4.827251017718034e-06,
+      "loss": 0.7352,
+      "step": 245
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.2303481101989746,
+      "learning_rate": 4.8258488473559794e-06,
+      "loss": 0.7576,
+      "step": 246
+    },
+    {
+      "epoch": 1.2048780487804878,
+      "grad_norm": 1.9348554611206055,
+      "learning_rate": 4.824441214720629e-06,
+      "loss": 0.8301,
+      "step": 247
+    },
+    {
+      "epoch": 1.2097560975609756,
+      "grad_norm": 1.3823461532592773,
+      "learning_rate": 4.823028123117818e-06,
+      "loss": 0.3957,
+      "step": 248
+    },
+    {
+      "epoch": 1.2146341463414634,
+      "grad_norm": 1.2902858257293701,
+      "learning_rate": 4.8216095758662015e-06,
+      "loss": 0.7012,
+      "step": 249
+    },
+    {
+      "epoch": 1.2195121951219512,
+      "grad_norm": 1.1826715469360352,
+      "learning_rate": 4.82018557629725e-06,
+      "loss": 0.7745,
+      "step": 250
+    },
+    {
+      "epoch": 1.224390243902439,
+      "grad_norm": 1.2553898096084595,
+      "learning_rate": 4.8187561277552376e-06,
+      "loss": 0.5574,
+      "step": 251
+    },
+    {
+      "epoch": 1.2292682926829268,
+      "grad_norm": 4.769270420074463,
+      "learning_rate": 4.817321233597232e-06,
+      "loss": 0.73,
+      "step": 252
+    },
+    {
+      "epoch": 1.2341463414634146,
+      "grad_norm": 1.3274935483932495,
+      "learning_rate": 4.815880897193095e-06,
+      "loss": 0.4925,
+      "step": 253
+    },
+    {
+      "epoch": 1.2390243902439024,
+      "grad_norm": 1.841747522354126,
+      "learning_rate": 4.814435121925466e-06,
+      "loss": 0.7432,
+      "step": 254
+    },
+    {
+      "epoch": 1.2439024390243902,
+      "grad_norm": 1.6842634677886963,
+      "learning_rate": 4.812983911189761e-06,
+      "loss": 0.6248,
+      "step": 255
+    },
+    {
+      "epoch": 1.248780487804878,
+      "grad_norm": 1.2322099208831787,
+      "learning_rate": 4.811527268394157e-06,
+      "loss": 0.4782,
+      "step": 256
+    },
+    {
+      "epoch": 1.2536585365853659,
+      "grad_norm": 1.27676260471344,
+      "learning_rate": 4.810065196959591e-06,
+      "loss": 0.5567,
+      "step": 257
+    },
+    {
+      "epoch": 1.2585365853658537,
+      "grad_norm": 1.3789864778518677,
+      "learning_rate": 4.8085977003197496e-06,
+      "loss": 0.7759,
+      "step": 258
+    },
+    {
+      "epoch": 1.2634146341463415,
+      "grad_norm": 1.681838870048523,
+      "learning_rate": 4.807124781921059e-06,
+      "loss": 0.9784,
+      "step": 259
+    },
+    {
+      "epoch": 1.2682926829268293,
+      "grad_norm": 1.6011807918548584,
+      "learning_rate": 4.805646445222679e-06,
+      "loss": 0.5745,
+      "step": 260
+    },
+    {
+      "epoch": 1.273170731707317,
+      "grad_norm": 1.1437225341796875,
+      "learning_rate": 4.804162693696494e-06,
+      "loss": 0.5499,
+      "step": 261
+    },
+    {
+      "epoch": 1.278048780487805,
+      "grad_norm": 1.5164761543273926,
+      "learning_rate": 4.802673530827105e-06,
+      "loss": 0.5353,
+      "step": 262
+    },
+    {
+      "epoch": 1.2829268292682927,
+      "grad_norm": 1.4570748805999756,
+      "learning_rate": 4.801178960111823e-06,
+      "loss": 0.5479,
+      "step": 263
+    },
+    {
+      "epoch": 1.2878048780487805,
+      "grad_norm": 1.20429265499115,
+      "learning_rate": 4.799678985060658e-06,
+      "loss": 0.7803,
+      "step": 264
+    },
+    {
+      "epoch": 1.2926829268292683,
+      "grad_norm": 1.3439183235168457,
+      "learning_rate": 4.798173609196314e-06,
+      "loss": 0.7694,
+      "step": 265
+    },
+    {
+      "epoch": 1.2975609756097561,
+      "grad_norm": 1.368338704109192,
+      "learning_rate": 4.796662836054176e-06,
+      "loss": 0.4389,
+      "step": 266
+    },
+    {
+      "epoch": 1.302439024390244,
+      "grad_norm": 1.2450690269470215,
+      "learning_rate": 4.795146669182304e-06,
+      "loss": 0.5893,
+      "step": 267
+    },
+    {
+      "epoch": 1.3073170731707318,
+      "grad_norm": 1.3387326002120972,
+      "learning_rate": 4.793625112141431e-06,
+      "loss": 0.4854,
+      "step": 268
+    },
+    {
+      "epoch": 1.3121951219512196,
+      "grad_norm": 1.0802762508392334,
+      "learning_rate": 4.792098168504943e-06,
+      "loss": 0.5212,
+      "step": 269
+    },
+    {
+      "epoch": 1.3170731707317074,
+      "grad_norm": 1.0994062423706055,
+      "learning_rate": 4.790565841858879e-06,
+      "loss": 0.4369,
+      "step": 270
+    },
+    {
+      "epoch": 1.3219512195121952,
+      "grad_norm": 1.2735989093780518,
+      "learning_rate": 4.789028135801919e-06,
+      "loss": 0.7215,
+      "step": 271
+    },
+    {
+      "epoch": 1.326829268292683,
+      "grad_norm": 1.4099750518798828,
+      "learning_rate": 4.787485053945377e-06,
+      "loss": 0.7693,
+      "step": 272
+    },
+    {
+      "epoch": 1.3317073170731708,
+      "grad_norm": 1.2519317865371704,
+      "learning_rate": 4.785936599913193e-06,
+      "loss": 0.6665,
+      "step": 273
+    },
+    {
+      "epoch": 1.3365853658536586,
+      "grad_norm": 1.6016514301300049,
+      "learning_rate": 4.784382777341922e-06,
+      "loss": 0.6664,
+      "step": 274
+    },
+    {
+      "epoch": 1.3414634146341464,
+      "grad_norm": 1.6531027555465698,
+      "learning_rate": 4.782823589880729e-06,
+      "loss": 1.0051,
+      "step": 275
+    },
+    {
+      "epoch": 1.346341463414634,
+      "grad_norm": 1.294831395149231,
+      "learning_rate": 4.7812590411913755e-06,
+      "loss": 0.7062,
+      "step": 276
+    },
+    {
+      "epoch": 1.3512195121951218,
+      "grad_norm": 1.3379662036895752,
+      "learning_rate": 4.779689134948217e-06,
+      "loss": 0.989,
+      "step": 277
+    },
+    {
+      "epoch": 1.3560975609756096,
+      "grad_norm": 1.415653109550476,
+      "learning_rate": 4.77811387483819e-06,
+      "loss": 0.4874,
+      "step": 278
+    },
+    {
+      "epoch": 1.3609756097560974,
+      "grad_norm": 1.7803961038589478,
+      "learning_rate": 4.776533264560804e-06,
+      "loss": 0.6769,
+      "step": 279
+    },
+    {
+      "epoch": 1.3658536585365852,
+      "grad_norm": 1.6438442468643188,
+      "learning_rate": 4.774947307828134e-06,
+      "loss": 0.8693,
+      "step": 280
+    },
+    {
+      "epoch": 1.370731707317073,
+      "grad_norm": 1.5104045867919922,
+      "learning_rate": 4.773356008364812e-06,
+      "loss": 0.5815,
+      "step": 281
+    },
+    {
+      "epoch": 1.3756097560975609,
+      "grad_norm": 1.1157118082046509,
+      "learning_rate": 4.771759369908017e-06,
+      "loss": 0.3809,
+      "step": 282
+    },
+    {
+      "epoch": 1.3804878048780487,
+      "grad_norm": 1.1048824787139893,
+      "learning_rate": 4.7701573962074635e-06,
+      "loss": 0.5832,
+      "step": 283
+    },
+    {
+      "epoch": 1.3853658536585365,
+      "grad_norm": 1.4487693309783936,
+      "learning_rate": 4.7685500910254015e-06,
+      "loss": 0.5126,
+      "step": 284
+    },
+    {
+      "epoch": 1.3902439024390243,
+      "grad_norm": 2.276735305786133,
+      "learning_rate": 4.766937458136598e-06,
+      "loss": 0.7106,
+      "step": 285
+    },
+    {
+      "epoch": 1.395121951219512,
+      "grad_norm": 1.1994675397872925,
+      "learning_rate": 4.765319501328332e-06,
+      "loss": 0.6759,
+      "step": 286
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.2363086938858032,
+      "learning_rate": 4.763696224400391e-06,
+      "loss": 0.486,
+      "step": 287
+    },
+    {
+      "epoch": 1.4048780487804877,
+      "grad_norm": 1.026413917541504,
+      "learning_rate": 4.762067631165049e-06,
+      "loss": 0.5165,
+      "step": 288
+    },
+    {
+      "epoch": 1.4097560975609755,
+      "grad_norm": 1.7001800537109375,
+      "learning_rate": 4.760433725447071e-06,
+      "loss": 0.724,
+      "step": 289
+    },
+    {
+      "epoch": 1.4146341463414633,
+      "grad_norm": 1.362479567527771,
+      "learning_rate": 4.758794511083697e-06,
+      "loss": 0.7431,
+      "step": 290
+    },
+    {
+      "epoch": 1.4195121951219511,
+      "grad_norm": 1.2791081666946411,
+      "learning_rate": 4.757149991924633e-06,
+      "loss": 0.62,
+      "step": 291
+    },
+    {
+      "epoch": 1.424390243902439,
+      "grad_norm": 19.68299102783203,
+      "learning_rate": 4.755500171832045e-06,
+      "loss": 0.4662,
+      "step": 292
+    },
+    {
+      "epoch": 1.4292682926829268,
+      "grad_norm": 1.8070446252822876,
+      "learning_rate": 4.753845054680548e-06,
+      "loss": 0.6275,
+      "step": 293
+    },
+    {
+      "epoch": 1.4341463414634146,
+      "grad_norm": 1.4444187879562378,
+      "learning_rate": 4.752184644357197e-06,
+      "loss": 0.4678,
+      "step": 294
+    },
+    {
+      "epoch": 1.4390243902439024,
+      "grad_norm": 1.2223567962646484,
+      "learning_rate": 4.750518944761477e-06,
+      "loss": 0.5206,
+      "step": 295
+    },
+    {
+      "epoch": 1.4439024390243902,
+      "grad_norm": 1.3032032251358032,
+      "learning_rate": 4.748847959805297e-06,
+      "loss": 0.5051,
+      "step": 296
+    },
+    {
+      "epoch": 1.448780487804878,
+      "grad_norm": 1.0974922180175781,
+      "learning_rate": 4.7471716934129774e-06,
+      "loss": 0.4994,
+      "step": 297
+    },
+    {
+      "epoch": 1.4536585365853658,
+      "grad_norm": 1.1713820695877075,
+      "learning_rate": 4.745490149521242e-06,
+      "loss": 0.4632,
+      "step": 298
+    },
+    {
+      "epoch": 1.4585365853658536,
+      "grad_norm": 1.1771109104156494,
+      "learning_rate": 4.743803332079209e-06,
+      "loss": 0.532,
+      "step": 299
+    },
+    {
+      "epoch": 1.4634146341463414,
+      "grad_norm": 1.3224542140960693,
+      "learning_rate": 4.742111245048382e-06,
+      "loss": 0.542,
+      "step": 300
+    },
+    {
+      "epoch": 1.4682926829268292,
+      "grad_norm": 1.0940190553665161,
+      "learning_rate": 4.740413892402639e-06,
+      "loss": 0.5109,
+      "step": 301
+    },
+    {
+      "epoch": 1.473170731707317,
+      "grad_norm": 1.4467796087265015,
+      "learning_rate": 4.738711278128228e-06,
+      "loss": 0.6034,
+      "step": 302
+    },
+    {
+      "epoch": 1.4780487804878049,
+      "grad_norm": 1.668028473854065,
+      "learning_rate": 4.7370034062237476e-06,
+      "loss": 0.3694,
+      "step": 303
+    },
+    {
+      "epoch": 1.4829268292682927,
+      "grad_norm": 2.006721258163452,
+      "learning_rate": 4.73529028070015e-06,
+      "loss": 0.5239,
+      "step": 304
+    },
+    {
+      "epoch": 1.4878048780487805,
+      "grad_norm": 1.415723443031311,
+      "learning_rate": 4.733571905580723e-06,
+      "loss": 0.8167,
+      "step": 305
+    },
+    {
+      "epoch": 1.4926829268292683,
+      "grad_norm": 1.1728172302246094,
+      "learning_rate": 4.731848284901082e-06,
+      "loss": 0.6764,
+      "step": 306
+    },
+    {
+      "epoch": 1.497560975609756,
+      "grad_norm": 1.5284311771392822,
+      "learning_rate": 4.730119422709165e-06,
+      "loss": 0.4207,
+      "step": 307
+    },
+    {
+      "epoch": 1.502439024390244,
+      "grad_norm": 1.4171814918518066,
+      "learning_rate": 4.728385323065215e-06,
+      "loss": 0.6778,
+      "step": 308
+    },
+    {
+      "epoch": 1.5073170731707317,
+      "grad_norm": 1.1928322315216064,
+      "learning_rate": 4.7266459900417815e-06,
+      "loss": 0.4901,
+      "step": 309
+    },
+    {
+      "epoch": 1.5121951219512195,
+      "grad_norm": 1.1769434213638306,
+      "learning_rate": 4.724901427723698e-06,
+      "loss": 0.7446,
+      "step": 310
+    },
+    {
+      "epoch": 1.5170731707317073,
+      "grad_norm": 1.1751688718795776,
+      "learning_rate": 4.723151640208084e-06,
+      "loss": 0.3547,
+      "step": 311
+    },
+    {
+      "epoch": 1.5219512195121951,
+      "grad_norm": 1.1819146871566772,
+      "learning_rate": 4.721396631604327e-06,
+      "loss": 0.4299,
+      "step": 312
+    },
+    {
+      "epoch": 1.526829268292683,
+      "grad_norm": 1.9109987020492554,
+      "learning_rate": 4.7196364060340785e-06,
+      "loss": 0.498,
+      "step": 313
+    },
+    {
+      "epoch": 1.5317073170731708,
+      "grad_norm": 1.2052650451660156,
+      "learning_rate": 4.7178709676312416e-06,
+      "loss": 0.7278,
+      "step": 314
+    },
+    {
+      "epoch": 1.5365853658536586,
+      "grad_norm": 1.6961381435394287,
+      "learning_rate": 4.716100320541961e-06,
+      "loss": 0.958,
+      "step": 315
+    },
+    {
+      "epoch": 1.5414634146341464,
+      "grad_norm": 1.3271212577819824,
+      "learning_rate": 4.714324468924614e-06,
+      "loss": 0.6825,
+      "step": 316
+    },
+    {
+      "epoch": 1.5463414634146342,
+      "grad_norm": 1.5681716203689575,
+      "learning_rate": 4.712543416949803e-06,
+      "loss": 0.6835,
+      "step": 317
+    },
+    {
+      "epoch": 1.551219512195122,
+      "grad_norm": 2.1769332885742188,
+      "learning_rate": 4.71075716880034e-06,
+      "loss": 0.6944,
+      "step": 318
+    },
+    {
+      "epoch": 1.5560975609756098,
+      "grad_norm": 1.8306576013565063,
+      "learning_rate": 4.708965728671243e-06,
+      "loss": 0.8235,
+      "step": 319
+    },
+    {
+      "epoch": 1.5609756097560976,
+      "grad_norm": 1.6039628982543945,
+      "learning_rate": 4.7071691007697214e-06,
+      "loss": 0.6311,
+      "step": 320
+    },
+    {
+      "epoch": 1.5658536585365854,
+      "grad_norm": 1.5115036964416504,
+      "learning_rate": 4.705367289315172e-06,
+      "loss": 0.7406,
+      "step": 321
+    },
+    {
+      "epoch": 1.5707317073170732,
+      "grad_norm": 1.2127209901809692,
+      "learning_rate": 4.703560298539158e-06,
+      "loss": 0.4469,
+      "step": 322
+    },
+    {
+      "epoch": 1.575609756097561,
+      "grad_norm": 1.2473901510238647,
+      "learning_rate": 4.701748132685415e-06,
+      "loss": 0.5062,
+      "step": 323
+    },
+    {
+      "epoch": 1.5804878048780489,
+      "grad_norm": 1.0023442506790161,
+      "learning_rate": 4.699930796009825e-06,
+      "loss": 0.5045,
+      "step": 324
+    },
+    {
+      "epoch": 1.5853658536585367,
+      "grad_norm": 1.6661128997802734,
+      "learning_rate": 4.698108292780418e-06,
+      "loss": 0.6671,
+      "step": 325
+    },
+    {
+      "epoch": 1.5902439024390245,
+      "grad_norm": 1.3463183641433716,
+      "learning_rate": 4.696280627277356e-06,
+      "loss": 0.4834,
+      "step": 326
+    },
+    {
+      "epoch": 1.5951219512195123,
+      "grad_norm": 1.0398967266082764,
+      "learning_rate": 4.6944478037929255e-06,
+      "loss": 0.5018,
+      "step": 327
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.3552498817443848,
+      "learning_rate": 4.692609826631525e-06,
+      "loss": 0.6013,
+      "step": 328
+    },
+    {
+      "epoch": 1.604878048780488,
+      "grad_norm": 1.3583056926727295,
+      "learning_rate": 4.690766700109659e-06,
+      "loss": 0.3773,
+      "step": 329
+    },
+    {
+      "epoch": 1.6097560975609757,
+      "grad_norm": 1.050776720046997,
+      "learning_rate": 4.6889184285559234e-06,
+      "loss": 0.4229,
+      "step": 330
+    },
+    {
+      "epoch": 1.6146341463414635,
+      "grad_norm": 1.4871671199798584,
+      "learning_rate": 4.687065016310996e-06,
+      "loss": 0.7061,
+      "step": 331
+    },
+    {
+      "epoch": 1.6195121951219513,
+      "grad_norm": 1.375916600227356,
+      "learning_rate": 4.685206467727631e-06,
+      "loss": 0.5492,
+      "step": 332
+    },
+    {
+      "epoch": 1.6243902439024391,
+      "grad_norm": 1.3035000562667847,
+      "learning_rate": 4.683342787170644e-06,
+      "loss": 0.5454,
+      "step": 333
+    },
+    {
+      "epoch": 1.629268292682927,
+      "grad_norm": 1.3644343614578247,
+      "learning_rate": 4.6814739790169006e-06,
+      "loss": 0.499,
+      "step": 334
+    },
+    {
+      "epoch": 1.6341463414634148,
+      "grad_norm": 1.321886658668518,
+      "learning_rate": 4.679600047655313e-06,
+      "loss": 0.7659,
+      "step": 335
+    },
+    {
+      "epoch": 1.6390243902439026,
+      "grad_norm": 1.8941502571105957,
+      "learning_rate": 4.6777209974868194e-06,
+      "loss": 1.0238,
+      "step": 336
+    },
+    {
+      "epoch": 1.6439024390243904,
+      "grad_norm": 1.2872205972671509,
+      "learning_rate": 4.675836832924387e-06,
+      "loss": 0.5666,
+      "step": 337
+    },
+    {
+      "epoch": 1.6487804878048782,
+      "grad_norm": 1.3836963176727295,
+      "learning_rate": 4.673947558392989e-06,
+      "loss": 0.3711,
+      "step": 338
+    },
+    {
+      "epoch": 1.653658536585366,
+      "grad_norm": 1.0632855892181396,
+      "learning_rate": 4.6720531783296e-06,
+      "loss": 0.5012,
+      "step": 339
+    },
+    {
+      "epoch": 1.6585365853658538,
+      "grad_norm": 1.143263339996338,
+      "learning_rate": 4.670153697183185e-06,
+      "loss": 0.5551,
+      "step": 340
+    },
+    {
+      "epoch": 1.6634146341463416,
+      "grad_norm": 1.2987687587738037,
+      "learning_rate": 4.668249119414692e-06,
+      "loss": 0.5321,
+      "step": 341
+    },
+    {
+      "epoch": 1.6682926829268294,
+      "grad_norm": 1.0663821697235107,
+      "learning_rate": 4.666339449497033e-06,
+      "loss": 0.5874,
+      "step": 342
+    },
+    {
+      "epoch": 1.6731707317073172,
+      "grad_norm": 1.3829140663146973,
+      "learning_rate": 4.664424691915084e-06,
+      "loss": 0.5317,
+      "step": 343
+    },
+    {
+      "epoch": 1.678048780487805,
+      "grad_norm": 1.3959944248199463,
+      "learning_rate": 4.6625048511656675e-06,
+      "loss": 0.5189,
+      "step": 344
+    },
+    {
+      "epoch": 1.6829268292682928,
+      "grad_norm": 1.5844937562942505,
+      "learning_rate": 4.660579931757543e-06,
+      "loss": 0.4425,
+      "step": 345
+    },
+    {
+      "epoch": 1.6878048780487804,
+      "grad_norm": 1.016044020652771,
+      "learning_rate": 4.6586499382113985e-06,
+      "loss": 0.5135,
+      "step": 346
+    },
+    {
+      "epoch": 1.6926829268292682,
+      "grad_norm": 1.3304717540740967,
+      "learning_rate": 4.6567148750598375e-06,
+      "loss": 0.7471,
+      "step": 347
+    },
+    {
+      "epoch": 1.697560975609756,
+      "grad_norm": 1.6659871339797974,
+      "learning_rate": 4.6547747468473705e-06,
+      "loss": 0.8518,
+      "step": 348
+    },
+    {
+      "epoch": 1.7024390243902439,
+      "grad_norm": 1.880418062210083,
+      "learning_rate": 4.652829558130404e-06,
+      "loss": 0.4086,
+      "step": 349
+    },
+    {
+      "epoch": 1.7073170731707317,
+      "grad_norm": 1.4303897619247437,
+      "learning_rate": 4.6508793134772265e-06,
+      "loss": 0.5808,
+      "step": 350
+    },
+    {
+      "epoch": 1.7121951219512195,
+      "grad_norm": 1.1196588277816772,
+      "learning_rate": 4.648924017468003e-06,
+      "loss": 0.5011,
+      "step": 351
+    },
+    {
+      "epoch": 1.7170731707317073,
+      "grad_norm": 1.340223789215088,
+      "learning_rate": 4.646963674694761e-06,
+      "loss": 0.777,
+      "step": 352
+    },
+    {
+      "epoch": 1.721951219512195,
+      "grad_norm": 1.130363941192627,
+      "learning_rate": 4.64499828976138e-06,
+      "loss": 0.3649,
+      "step": 353
+    },
+    {
+      "epoch": 1.726829268292683,
+      "grad_norm": 1.175565481185913,
+      "learning_rate": 4.64302786728358e-06,
+      "loss": 0.4545,
+      "step": 354
+    },
+    {
+      "epoch": 1.7317073170731707,
+      "grad_norm": 1.064070224761963,
+      "learning_rate": 4.641052411888913e-06,
+      "loss": 0.5004,
+      "step": 355
+    },
+    {
+      "epoch": 1.7365853658536585,
+      "grad_norm": 1.2501908540725708,
+      "learning_rate": 4.6390719282167515e-06,
+      "loss": 0.4198,
+      "step": 356
+    },
+    {
+      "epoch": 1.7414634146341463,
+      "grad_norm": 1.4923558235168457,
+      "learning_rate": 4.637086420918276e-06,
+      "loss": 0.739,
+      "step": 357
+    },
+    {
+      "epoch": 1.7463414634146341,
+      "grad_norm": 1.0668526887893677,
+      "learning_rate": 4.635095894656465e-06,
+      "loss": 0.6309,
+      "step": 358
+    },
+    {
+      "epoch": 1.751219512195122,
+      "grad_norm": 1.620019793510437,
+      "learning_rate": 4.633100354106085e-06,
+      "loss": 0.3578,
+      "step": 359
+    },
+    {
+      "epoch": 1.7560975609756098,
+      "grad_norm": 1.3894845247268677,
+      "learning_rate": 4.631099803953677e-06,
+      "loss": 0.7853,
+      "step": 360
+    },
+    {
+      "epoch": 1.7609756097560976,
+      "grad_norm": 1.2985631227493286,
+      "learning_rate": 4.629094248897546e-06,
+      "loss": 0.4396,
+      "step": 361
+    },
+    {
+      "epoch": 1.7658536585365854,
+      "grad_norm": 1.2633106708526611,
+      "learning_rate": 4.627083693647757e-06,
+      "loss": 0.5232,
+      "step": 362
+    },
+    {
+      "epoch": 1.7707317073170732,
+      "grad_norm": 1.3993726968765259,
+      "learning_rate": 4.625068142926111e-06,
+      "loss": 0.778,
+      "step": 363
+    },
+    {
+      "epoch": 1.775609756097561,
+      "grad_norm": 1.4045125246047974,
+      "learning_rate": 4.623047601466144e-06,
+      "loss": 0.6919,
+      "step": 364
+    },
+    {
+      "epoch": 1.7804878048780488,
+      "grad_norm": 1.1647089719772339,
+      "learning_rate": 4.621022074013114e-06,
+      "loss": 0.5897,
+      "step": 365
+    },
+    {
+      "epoch": 1.7853658536585366,
+      "grad_norm": 1.2741727828979492,
+      "learning_rate": 4.618991565323987e-06,
+      "loss": 0.477,
+      "step": 366
+    },
+    {
+      "epoch": 1.7902439024390244,
+      "grad_norm": 1.1093999147415161,
+      "learning_rate": 4.616956080167426e-06,
+      "loss": 0.5143,
+      "step": 367
+    },
+    {
+      "epoch": 1.7951219512195122,
+      "grad_norm": 1.5825276374816895,
+      "learning_rate": 4.614915623323786e-06,
+      "loss": 0.8959,
+      "step": 368
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.071441888809204,
+      "learning_rate": 4.612870199585092e-06,
+      "loss": 0.3929,
+      "step": 369
+    },
+    {
+      "epoch": 1.8048780487804879,
+      "grad_norm": 1.1866506338119507,
+      "learning_rate": 4.610819813755038e-06,
+      "loss": 0.4613,
+      "step": 370
+    },
+    {
+      "epoch": 1.8097560975609757,
+      "grad_norm": 1.4591060876846313,
+      "learning_rate": 4.608764470648971e-06,
+      "loss": 0.4325,
+      "step": 371
+    },
+    {
+      "epoch": 1.8146341463414632,
+      "grad_norm": 1.301404356956482,
+      "learning_rate": 4.606704175093879e-06,
+      "loss": 0.4076,
+      "step": 372
+    },
+    {
+      "epoch": 1.819512195121951,
+      "grad_norm": 1.3091838359832764,
+      "learning_rate": 4.604638931928383e-06,
+      "loss": 0.7532,
+      "step": 373
+    },
+    {
+      "epoch": 1.8243902439024389,
+      "grad_norm": 1.2092056274414062,
+      "learning_rate": 4.602568746002718e-06,
+      "loss": 0.4729,
+      "step": 374
+    },
+    {
+      "epoch": 1.8292682926829267,
+      "grad_norm": 1.3634448051452637,
+      "learning_rate": 4.600493622178734e-06,
+      "loss": 0.7733,
+      "step": 375
+    },
+    {
+      "epoch": 1.8341463414634145,
+      "grad_norm": 1.3015254735946655,
+      "learning_rate": 4.598413565329876e-06,
+      "loss": 0.5129,
+      "step": 376
+    },
+    {
+      "epoch": 1.8390243902439023,
+      "grad_norm": 1.1774072647094727,
+      "learning_rate": 4.596328580341169e-06,
+      "loss": 0.5184,
+      "step": 377
+    },
+    {
+      "epoch": 1.84390243902439,
+      "grad_norm": 1.3183966875076294,
+      "learning_rate": 4.5942386721092195e-06,
+      "loss": 0.6458,
+      "step": 378
+    },
+    {
+      "epoch": 1.848780487804878,
+      "grad_norm": 1.3310905694961548,
+      "learning_rate": 4.592143845542189e-06,
+      "loss": 0.5906,
+      "step": 379
+    },
+    {
+      "epoch": 1.8536585365853657,
+      "grad_norm": 1.223225474357605,
+      "learning_rate": 4.590044105559797e-06,
+      "loss": 0.8205,
+      "step": 380
+    },
+    {
+      "epoch": 1.8585365853658535,
+      "grad_norm": 1.0616337060928345,
+      "learning_rate": 4.587939457093296e-06,
+      "loss": 0.4877,
+      "step": 381
+    },
+    {
+      "epoch": 1.8634146341463413,
+      "grad_norm": 1.877367615699768,
+      "learning_rate": 4.585829905085468e-06,
+      "loss": 0.6037,
+      "step": 382
+    },
+    {
+      "epoch": 1.8682926829268292,
+      "grad_norm": 1.2741544246673584,
+      "learning_rate": 4.5837154544906135e-06,
+      "loss": 0.721,
+      "step": 383
+    },
+    {
+      "epoch": 1.873170731707317,
+      "grad_norm": 1.1264437437057495,
+      "learning_rate": 4.581596110274535e-06,
+      "loss": 0.664,
+      "step": 384
+    },
+    {
+      "epoch": 1.8780487804878048,
+      "grad_norm": 1.1946134567260742,
+      "learning_rate": 4.579471877414527e-06,
+      "loss": 0.8815,
+      "step": 385
+    },
+    {
+      "epoch": 1.8829268292682926,
+      "grad_norm": 1.4959359169006348,
+      "learning_rate": 4.577342760899368e-06,
+      "loss": 0.6356,
+      "step": 386
+    },
+    {
+      "epoch": 1.8878048780487804,
+      "grad_norm": 1.2867447137832642,
+      "learning_rate": 4.575208765729302e-06,
+      "loss": 0.4832,
+      "step": 387
+    },
+    {
+      "epoch": 1.8926829268292682,
+      "grad_norm": 1.5730870962142944,
+      "learning_rate": 4.573069896916035e-06,
+      "loss": 0.7647,
+      "step": 388
+    },
+    {
+      "epoch": 1.897560975609756,
+      "grad_norm": 1.1448895931243896,
+      "learning_rate": 4.5709261594827125e-06,
+      "loss": 0.6053,
+      "step": 389
+    },
+    {
+      "epoch": 1.9024390243902438,
+      "grad_norm": 1.2466707229614258,
+      "learning_rate": 4.568777558463922e-06,
+      "loss": 0.4691,
+      "step": 390
+    },
+    {
+      "epoch": 1.9073170731707316,
+      "grad_norm": 1.2236309051513672,
+      "learning_rate": 4.566624098905665e-06,
+      "loss": 0.6445,
+      "step": 391
+    },
+    {
+      "epoch": 1.9121951219512194,
+      "grad_norm": 1.3577604293823242,
+      "learning_rate": 4.564465785865359e-06,
+      "loss": 0.4848,
+      "step": 392
+    },
+    {
+      "epoch": 1.9170731707317072,
+      "grad_norm": 1.151893973350525,
+      "learning_rate": 4.56230262441182e-06,
+      "loss": 0.3876,
+      "step": 393
+    },
+    {
+      "epoch": 1.921951219512195,
+      "grad_norm": 1.2083380222320557,
+      "learning_rate": 4.560134619625247e-06,
+      "loss": 0.602,
+      "step": 394
+    },
+    {
+      "epoch": 1.9268292682926829,
+      "grad_norm": 1.173264980316162,
+      "learning_rate": 4.5579617765972155e-06,
+      "loss": 0.523,
+      "step": 395
+    },
+    {
+      "epoch": 1.9317073170731707,
+      "grad_norm": 1.1430155038833618,
+      "learning_rate": 4.555784100430662e-06,
+      "loss": 0.468,
+      "step": 396
+    },
+    {
+      "epoch": 1.9365853658536585,
+      "grad_norm": 1.0470151901245117,
+      "learning_rate": 4.553601596239877e-06,
+      "loss": 0.4188,
+      "step": 397
+    },
+    {
+      "epoch": 1.9414634146341463,
+      "grad_norm": 1.1655720472335815,
+      "learning_rate": 4.551414269150489e-06,
+      "loss": 0.499,
+      "step": 398
+    },
+    {
+      "epoch": 1.946341463414634,
+      "grad_norm": 1.142289638519287,
+      "learning_rate": 4.54922212429945e-06,
+      "loss": 0.4927,
+      "step": 399
+    },
+    {
+      "epoch": 1.951219512195122,
+      "grad_norm": 2.494032144546509,
+      "learning_rate": 4.547025166835027e-06,
+      "loss": 0.6699,
+      "step": 400
+    },
+    {
+      "epoch": 1.9560975609756097,
+      "grad_norm": 1.200950264930725,
+      "learning_rate": 4.544823401916794e-06,
+      "loss": 0.7996,
+      "step": 401
+    },
+    {
+      "epoch": 1.9609756097560975,
+      "grad_norm": 1.2763162851333618,
+      "learning_rate": 4.542616834715612e-06,
+      "loss": 0.5751,
+      "step": 402
+    },
+    {
+      "epoch": 1.9658536585365853,
+      "grad_norm": 1.1107186079025269,
+      "learning_rate": 4.540405470413618e-06,
+      "loss": 0.3838,
+      "step": 403
+    },
+    {
+      "epoch": 1.9707317073170731,
+      "grad_norm": 1.218024492263794,
+      "learning_rate": 4.53818931420422e-06,
+      "loss": 0.659,
+      "step": 404
+    },
+    {
+      "epoch": 1.975609756097561,
+      "grad_norm": 1.0650073289871216,
+      "learning_rate": 4.535968371292076e-06,
+      "loss": 0.5387,
+      "step": 405
+    },
+    {
+      "epoch": 1.9804878048780488,
+      "grad_norm": 1.2255375385284424,
+      "learning_rate": 4.533742646893086e-06,
+      "loss": 0.6249,
+      "step": 406
+    },
+    {
+      "epoch": 1.9853658536585366,
+      "grad_norm": 1.0770877599716187,
+      "learning_rate": 4.531512146234383e-06,
+      "loss": 0.618,
+      "step": 407
+    },
+    {
+      "epoch": 1.9902439024390244,
+      "grad_norm": 1.1472445726394653,
+      "learning_rate": 4.529276874554312e-06,
+      "loss": 0.8057,
+      "step": 408
+    },
+    {
+      "epoch": 1.9951219512195122,
+      "grad_norm": 1.6422274112701416,
+      "learning_rate": 4.527036837102426e-06,
+      "loss": 0.4654,
+      "step": 409
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.2691030502319336,
+      "learning_rate": 4.524792039139471e-06,
+      "loss": 0.7222,
+      "step": 410
+    },
+    {
+      "epoch": 2.004878048780488,
+      "grad_norm": 1.0516557693481445,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.3588,
+      "step": 411
+    },
+    {
+      "epoch": 2.0097560975609756,
+      "grad_norm": 1.1649781465530396,
+      "learning_rate": 4.520288182779214e-06,
+      "loss": 0.6146,
+      "step": 412
+    },
+    {
+      "epoch": 2.0146341463414634,
+      "grad_norm": 0.9384630918502808,
+      "learning_rate": 4.518029134959253e-06,
+      "loss": 0.3075,
+      "step": 413
+    },
+    {
+      "epoch": 2.0195121951219512,
+      "grad_norm": 1.1893984079360962,
+      "learning_rate": 4.515765347782878e-06,
+      "loss": 0.6064,
+      "step": 414
+    },
+    {
+      "epoch": 2.024390243902439,
+      "grad_norm": 1.370092749595642,
+      "learning_rate": 4.5134968265666085e-06,
+      "loss": 0.8078,
+      "step": 415
+    },
+    {
+      "epoch": 2.029268292682927,
+      "grad_norm": 1.3880703449249268,
+      "learning_rate": 4.511223576638084e-06,
+      "loss": 0.5884,
+      "step": 416
+    },
+    {
+      "epoch": 2.0341463414634147,
+      "grad_norm": 1.4118834733963013,
+      "learning_rate": 4.508945603336049e-06,
+      "loss": 0.7247,
+      "step": 417
+    },
+    {
+      "epoch": 2.0390243902439025,
+      "grad_norm": 0.9344208836555481,
+      "learning_rate": 4.50666291201034e-06,
+      "loss": 0.436,
+      "step": 418
+    },
+    {
+      "epoch": 2.0439024390243903,
+      "grad_norm": 1.0404282808303833,
+      "learning_rate": 4.504375508021876e-06,
+      "loss": 0.3214,
+      "step": 419
+    },
+    {
+      "epoch": 2.048780487804878,
+      "grad_norm": 1.2519387006759644,
+      "learning_rate": 4.50208339674264e-06,
+      "loss": 0.7228,
+      "step": 420
+    },
+    {
+      "epoch": 2.053658536585366,
+      "grad_norm": 1.4293513298034668,
+      "learning_rate": 4.499786583555675e-06,
+      "loss": 0.7588,
+      "step": 421
+    },
+    {
+      "epoch": 2.0585365853658537,
+      "grad_norm": 1.076704740524292,
+      "learning_rate": 4.497485073855061e-06,
+      "loss": 0.4029,
+      "step": 422
+    },
+    {
+      "epoch": 2.0634146341463415,
+      "grad_norm": 3.5152344703674316,
+      "learning_rate": 4.495178873045913e-06,
+      "loss": 0.4022,
+      "step": 423
+    },
+    {
+      "epoch": 2.0682926829268293,
+      "grad_norm": 1.0236436128616333,
+      "learning_rate": 4.4928679865443605e-06,
+      "loss": 0.4065,
+      "step": 424
+    },
+    {
+      "epoch": 2.073170731707317,
+      "grad_norm": 1.0082788467407227,
+      "learning_rate": 4.4905524197775366e-06,
+      "loss": 0.5286,
+      "step": 425
+    },
+    {
+      "epoch": 2.078048780487805,
+      "grad_norm": 1.0858159065246582,
+      "learning_rate": 4.4882321781835666e-06,
+      "loss": 0.4934,
+      "step": 426
+    },
+    {
+      "epoch": 2.0829268292682928,
+      "grad_norm": 0.9646297097206116,
+      "learning_rate": 4.4859072672115565e-06,
+      "loss": 0.3315,
+      "step": 427
+    },
+    {
+      "epoch": 2.0878048780487806,
+      "grad_norm": 1.3029683828353882,
+      "learning_rate": 4.483577692321577e-06,
+      "loss": 0.7549,
+      "step": 428
+    },
+    {
+      "epoch": 2.0926829268292684,
+      "grad_norm": 1.2982211112976074,
+      "learning_rate": 4.481243458984651e-06,
+      "loss": 0.5255,
+      "step": 429
+    },
+    {
+      "epoch": 2.097560975609756,
+      "grad_norm": 1.0484429597854614,
+      "learning_rate": 4.478904572682743e-06,
+      "loss": 0.5654,
+      "step": 430
+    },
+    {
+      "epoch": 2.102439024390244,
+      "grad_norm": 1.0825645923614502,
+      "learning_rate": 4.476561038908745e-06,
+      "loss": 0.4454,
+      "step": 431
+    },
+    {
+      "epoch": 2.107317073170732,
+      "grad_norm": 1.119730830192566,
+      "learning_rate": 4.474212863166464e-06,
+      "loss": 0.5983,
+      "step": 432
+    },
+    {
+      "epoch": 2.1121951219512196,
+      "grad_norm": 1.542769432067871,
+      "learning_rate": 4.471860050970608e-06,
+      "loss": 0.6498,
+      "step": 433
+    },
+    {
+      "epoch": 2.1170731707317074,
+      "grad_norm": 1.5004665851593018,
+      "learning_rate": 4.469502607846774e-06,
+      "loss": 0.8973,
+      "step": 434
+    },
+    {
+      "epoch": 2.1219512195121952,
+      "grad_norm": 1.2841945886611938,
+      "learning_rate": 4.467140539331434e-06,
+      "loss": 0.401,
+      "step": 435
+    },
+    {
+      "epoch": 2.126829268292683,
+      "grad_norm": 1.2362148761749268,
+      "learning_rate": 4.464773850971924e-06,
+      "loss": 0.2812,
+      "step": 436
+    },
+    {
+      "epoch": 2.131707317073171,
+      "grad_norm": 1.0421311855316162,
+      "learning_rate": 4.46240254832643e-06,
+      "loss": 0.4857,
+      "step": 437
+    },
+    {
+      "epoch": 2.1365853658536587,
+      "grad_norm": 1.265649437904358,
+      "learning_rate": 4.460026636963971e-06,
+      "loss": 0.568,
+      "step": 438
+    },
+    {
+      "epoch": 2.1414634146341465,
+      "grad_norm": 1.3310377597808838,
+      "learning_rate": 4.4576461224643965e-06,
+      "loss": 0.4738,
+      "step": 439
+    },
+    {
+      "epoch": 2.1463414634146343,
+      "grad_norm": 0.9549973607063293,
+      "learning_rate": 4.455261010418359e-06,
+      "loss": 0.3512,
+      "step": 440
+    },
+    {
+      "epoch": 2.151219512195122,
+      "grad_norm": 1.026897668838501,
+      "learning_rate": 4.452871306427314e-06,
+      "loss": 0.6619,
+      "step": 441
+    },
+    {
+      "epoch": 2.15609756097561,
+      "grad_norm": 1.4088711738586426,
+      "learning_rate": 4.450477016103498e-06,
+      "loss": 0.5356,
+      "step": 442
+    },
+    {
+      "epoch": 2.1609756097560977,
+      "grad_norm": 1.3547532558441162,
+      "learning_rate": 4.4480781450699205e-06,
+      "loss": 0.4301,
+      "step": 443
+    },
+    {
+      "epoch": 2.1658536585365855,
+      "grad_norm": 1.1079550981521606,
+      "learning_rate": 4.4456746989603464e-06,
+      "loss": 0.3961,
+      "step": 444
+    },
+    {
+      "epoch": 2.1707317073170733,
+      "grad_norm": 1.4145056009292603,
+      "learning_rate": 4.443266683419289e-06,
+      "loss": 0.5756,
+      "step": 445
+    },
+    {
+      "epoch": 2.175609756097561,
+      "grad_norm": 1.6260194778442383,
+      "learning_rate": 4.440854104101988e-06,
+      "loss": 0.4502,
+      "step": 446
+    },
+    {
+      "epoch": 2.180487804878049,
+      "grad_norm": 1.5400185585021973,
+      "learning_rate": 4.438436966674406e-06,
+      "loss": 0.5924,
+      "step": 447
+    },
+    {
+      "epoch": 2.1853658536585368,
+      "grad_norm": 1.2174339294433594,
+      "learning_rate": 4.436015276813208e-06,
+      "loss": 0.4297,
+      "step": 448
+    },
+    {
+      "epoch": 2.1902439024390246,
+      "grad_norm": 1.039923906326294,
+      "learning_rate": 4.4335890402057505e-06,
+      "loss": 0.5547,
+      "step": 449
+    },
+    {
+      "epoch": 2.1951219512195124,
+      "grad_norm": 1.2806974649429321,
+      "learning_rate": 4.431158262550067e-06,
+      "loss": 0.536,
+      "step": 450
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.1118040084838867,
+      "learning_rate": 4.428722949554858e-06,
+      "loss": 0.3164,
+      "step": 451
+    },
+    {
+      "epoch": 2.204878048780488,
+      "grad_norm": 1.0504515171051025,
+      "learning_rate": 4.426283106939474e-06,
+      "loss": 0.4385,
+      "step": 452
+    },
+    {
+      "epoch": 2.209756097560976,
+      "grad_norm": 1.2834644317626953,
+      "learning_rate": 4.423838740433903e-06,
+      "loss": 0.598,
+      "step": 453
+    },
+    {
+      "epoch": 2.2146341463414636,
+      "grad_norm": 0.9809593558311462,
+      "learning_rate": 4.4213898557787586e-06,
+      "loss": 0.258,
+      "step": 454
+    },
+    {
+      "epoch": 2.2195121951219514,
+      "grad_norm": 1.1549808979034424,
+      "learning_rate": 4.4189364587252636e-06,
+      "loss": 0.7969,
+      "step": 455
+    },
+    {
+      "epoch": 2.2243902439024392,
+      "grad_norm": 1.3864680528640747,
+      "learning_rate": 4.416478555035241e-06,
+      "loss": 0.3594,
+      "step": 456
+    },
+    {
+      "epoch": 2.229268292682927,
+      "grad_norm": 1.177148461341858,
+      "learning_rate": 4.4140161504810935e-06,
+      "loss": 0.4564,
+      "step": 457
+    },
+    {
+      "epoch": 2.234146341463415,
+      "grad_norm": 0.9031911492347717,
+      "learning_rate": 4.4115492508457986e-06,
+      "loss": 0.3143,
+      "step": 458
+    },
+    {
+      "epoch": 2.2390243902439027,
+      "grad_norm": 1.2335823774337769,
+      "learning_rate": 4.409077861922887e-06,
+      "loss": 0.5226,
+      "step": 459
+    },
+    {
+      "epoch": 2.2439024390243905,
+      "grad_norm": 1.0374839305877686,
+      "learning_rate": 4.406601989516435e-06,
+      "loss": 0.4168,
+      "step": 460
+    },
+    {
+      "epoch": 2.2487804878048783,
+      "grad_norm": 0.7933486104011536,
+      "learning_rate": 4.404121639441047e-06,
+      "loss": 0.2772,
+      "step": 461
+    },
+    {
+      "epoch": 2.253658536585366,
+      "grad_norm": 1.3051011562347412,
+      "learning_rate": 4.401636817521843e-06,
+      "loss": 0.4949,
+      "step": 462
+    },
+    {
+      "epoch": 2.258536585365854,
+      "grad_norm": 1.1215354204177856,
+      "learning_rate": 4.399147529594447e-06,
+      "loss": 0.3546,
+      "step": 463
+    },
+    {
+      "epoch": 2.2634146341463417,
+      "grad_norm": 0.9971939325332642,
+      "learning_rate": 4.3966537815049686e-06,
+      "loss": 0.4584,
+      "step": 464
+    },
+    {
+      "epoch": 2.2682926829268295,
+      "grad_norm": 1.3168960809707642,
+      "learning_rate": 4.394155579109994e-06,
+      "loss": 0.5999,
+      "step": 465
+    },
+    {
+      "epoch": 2.2731707317073173,
+      "grad_norm": 1.4978530406951904,
+      "learning_rate": 4.391652928276572e-06,
+      "loss": 0.6378,
+      "step": 466
+    },
+    {
+      "epoch": 2.278048780487805,
+      "grad_norm": 1.1670995950698853,
+      "learning_rate": 4.389145834882195e-06,
+      "loss": 0.5562,
+      "step": 467
+    },
+    {
+      "epoch": 2.2829268292682925,
+      "grad_norm": 1.482450008392334,
+      "learning_rate": 4.386634304814789e-06,
+      "loss": 0.4452,
+      "step": 468
+    },
+    {
+      "epoch": 2.2878048780487803,
+      "grad_norm": 1.140497088432312,
+      "learning_rate": 4.384118343972704e-06,
+      "loss": 0.61,
+      "step": 469
+    },
+    {
+      "epoch": 2.292682926829268,
+      "grad_norm": 1.202193260192871,
+      "learning_rate": 4.381597958264692e-06,
+      "loss": 0.6702,
+      "step": 470
+    },
+    {
+      "epoch": 2.297560975609756,
+      "grad_norm": 1.0919175148010254,
+      "learning_rate": 4.379073153609896e-06,
+      "loss": 0.6909,
+      "step": 471
+    },
+    {
+      "epoch": 2.3024390243902437,
+      "grad_norm": 1.3111231327056885,
+      "learning_rate": 4.37654393593784e-06,
+      "loss": 0.5377,
+      "step": 472
+    },
+    {
+      "epoch": 2.3073170731707315,
+      "grad_norm": 1.1155540943145752,
+      "learning_rate": 4.3740103111884096e-06,
+      "loss": 0.4831,
+      "step": 473
+    },
+    {
+      "epoch": 2.3121951219512193,
+      "grad_norm": 1.2359633445739746,
+      "learning_rate": 4.371472285311842e-06,
+      "loss": 0.4191,
+      "step": 474
+    },
+    {
+      "epoch": 2.317073170731707,
+      "grad_norm": 0.8049008250236511,
+      "learning_rate": 4.368929864268709e-06,
+      "loss": 0.3243,
+      "step": 475
+    },
+    {
+      "epoch": 2.321951219512195,
+      "grad_norm": 1.012024164199829,
+      "learning_rate": 4.366383054029907e-06,
+      "loss": 0.5594,
+      "step": 476
+    },
+    {
+      "epoch": 2.3268292682926828,
+      "grad_norm": 1.1117417812347412,
+      "learning_rate": 4.363831860576638e-06,
+      "loss": 0.6147,
+      "step": 477
+    },
+    {
+      "epoch": 2.3317073170731706,
+      "grad_norm": 1.2145675420761108,
+      "learning_rate": 4.361276289900396e-06,
+      "loss": 0.4982,
+      "step": 478
+    },
+    {
+      "epoch": 2.3365853658536584,
+      "grad_norm": 1.3909268379211426,
+      "learning_rate": 4.358716348002962e-06,
+      "loss": 0.6937,
+      "step": 479
+    },
+    {
+      "epoch": 2.341463414634146,
+      "grad_norm": 1.2233402729034424,
+      "learning_rate": 4.356152040896376e-06,
+      "loss": 0.4374,
+      "step": 480
+    },
+    {
+      "epoch": 2.346341463414634,
+      "grad_norm": 0.8656805753707886,
+      "learning_rate": 4.3535833746029335e-06,
+      "loss": 0.3526,
+      "step": 481
+    },
+    {
+      "epoch": 2.351219512195122,
+      "grad_norm": 1.2934560775756836,
+      "learning_rate": 4.351010355155165e-06,
+      "loss": 0.427,
+      "step": 482
+    },
+    {
+      "epoch": 2.3560975609756096,
+      "grad_norm": 0.9406152367591858,
+      "learning_rate": 4.348432988595828e-06,
+      "loss": 0.346,
+      "step": 483
+    },
+    {
+      "epoch": 2.3609756097560974,
+      "grad_norm": 1.4136567115783691,
+      "learning_rate": 4.345851280977885e-06,
+      "loss": 0.6952,
+      "step": 484
+    },
+    {
+      "epoch": 2.3658536585365852,
+      "grad_norm": 1.0673305988311768,
+      "learning_rate": 4.343265238364496e-06,
+      "loss": 0.38,
+      "step": 485
+    },
+    {
+      "epoch": 2.370731707317073,
+      "grad_norm": 1.392567753791809,
+      "learning_rate": 4.340674866829001e-06,
+      "loss": 0.4756,
+      "step": 486
+    },
+    {
+      "epoch": 2.375609756097561,
+      "grad_norm": 1.3056880235671997,
+      "learning_rate": 4.338080172454908e-06,
+      "loss": 0.7794,
+      "step": 487
+    },
+    {
+      "epoch": 2.3804878048780487,
+      "grad_norm": 0.9703968167304993,
+      "learning_rate": 4.335481161335875e-06,
+      "loss": 0.4966,
+      "step": 488
+    },
+    {
+      "epoch": 2.3853658536585365,
+      "grad_norm": 1.2114007472991943,
+      "learning_rate": 4.332877839575699e-06,
+      "loss": 0.4135,
+      "step": 489
+    },
+    {
+      "epoch": 2.3902439024390243,
+      "grad_norm": 0.988808274269104,
+      "learning_rate": 4.330270213288301e-06,
+      "loss": 0.5718,
+      "step": 490
+    },
+    {
+      "epoch": 2.395121951219512,
+      "grad_norm": 1.1490461826324463,
+      "learning_rate": 4.32765828859771e-06,
+      "loss": 0.8009,
+      "step": 491
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.3094358444213867,
+      "learning_rate": 4.325042071638051e-06,
+      "loss": 0.6113,
+      "step": 492
+    },
+    {
+      "epoch": 2.4048780487804877,
+      "grad_norm": 3.0272164344787598,
+      "learning_rate": 4.322421568553529e-06,
+      "loss": 0.4342,
+      "step": 493
+    },
+    {
+      "epoch": 2.4097560975609755,
+      "grad_norm": 1.0097917318344116,
+      "learning_rate": 4.319796785498416e-06,
+      "loss": 0.391,
+      "step": 494
+    },
+    {
+      "epoch": 2.4146341463414633,
+      "grad_norm": 1.1281390190124512,
+      "learning_rate": 4.317167728637032e-06,
+      "loss": 0.5742,
+      "step": 495
+    },
+    {
+      "epoch": 2.419512195121951,
+      "grad_norm": 1.1757224798202515,
+      "learning_rate": 4.314534404143738e-06,
+      "loss": 0.4793,
+      "step": 496
+    },
+    {
+      "epoch": 2.424390243902439,
+      "grad_norm": 1.0164152383804321,
+      "learning_rate": 4.3118968182029155e-06,
+      "loss": 0.5953,
+      "step": 497
+    },
+    {
+      "epoch": 2.4292682926829268,
+      "grad_norm": 0.9032419323921204,
+      "learning_rate": 4.3092549770089566e-06,
+      "loss": 0.3346,
+      "step": 498
+    },
+    {
+      "epoch": 2.4341463414634146,
+      "grad_norm": 1.3208485841751099,
+      "learning_rate": 4.306608886766243e-06,
+      "loss": 0.498,
+      "step": 499
+    },
+    {
+      "epoch": 2.4390243902439024,
+      "grad_norm": 1.0782032012939453,
+      "learning_rate": 4.303958553689137e-06,
+      "loss": 0.5019,
+      "step": 500
+    },
+    {
+      "epoch": 2.44390243902439,
+      "grad_norm": 1.3496298789978027,
+      "learning_rate": 4.3013039840019675e-06,
+      "loss": 0.6927,
+      "step": 501
+    },
+    {
+      "epoch": 2.448780487804878,
+      "grad_norm": 1.071814775466919,
+      "learning_rate": 4.2986451839390105e-06,
+      "loss": 0.3543,
+      "step": 502
+    },
+    {
+      "epoch": 2.453658536585366,
+      "grad_norm": 1.0938665866851807,
+      "learning_rate": 4.295982159744476e-06,
+      "loss": 0.5101,
+      "step": 503
+    },
+    {
+      "epoch": 2.4585365853658536,
+      "grad_norm": 1.4654054641723633,
+      "learning_rate": 4.293314917672498e-06,
+      "loss": 0.5915,
+      "step": 504
+    },
+    {
+      "epoch": 2.4634146341463414,
+      "grad_norm": 0.8499230146408081,
+      "learning_rate": 4.290643463987114e-06,
+      "loss": 0.3113,
+      "step": 505
+    },
+    {
+      "epoch": 2.4682926829268292,
+      "grad_norm": 1.0232477188110352,
+      "learning_rate": 4.287967804962252e-06,
+      "loss": 0.3895,
+      "step": 506
+    },
+    {
+      "epoch": 2.473170731707317,
+      "grad_norm": 0.9898338317871094,
+      "learning_rate": 4.285287946881718e-06,
+      "loss": 0.2425,
+      "step": 507
+    },
+    {
+      "epoch": 2.478048780487805,
+      "grad_norm": 1.0269073247909546,
+      "learning_rate": 4.282603896039178e-06,
+      "loss": 0.6315,
+      "step": 508
+    },
+    {
+      "epoch": 2.4829268292682927,
+      "grad_norm": 1.055603265762329,
+      "learning_rate": 4.279915658738145e-06,
+      "loss": 0.4708,
+      "step": 509
+    },
+    {
+      "epoch": 2.4878048780487805,
+      "grad_norm": 1.207694172859192,
+      "learning_rate": 4.277223241291966e-06,
+      "loss": 0.6828,
+      "step": 510
+    },
+    {
+      "epoch": 2.4926829268292683,
+      "grad_norm": 1.0360615253448486,
+      "learning_rate": 4.274526650023801e-06,
+      "loss": 0.4955,
+      "step": 511
+    },
+    {
+      "epoch": 2.497560975609756,
+      "grad_norm": 1.61383855342865,
+      "learning_rate": 4.271825891266617e-06,
+      "loss": 0.5325,
+      "step": 512
+    },
+    {
+      "epoch": 2.502439024390244,
+      "grad_norm": 1.7913377285003662,
+      "learning_rate": 4.269120971363164e-06,
+      "loss": 0.7766,
+      "step": 513
+    },
+    {
+      "epoch": 2.5073170731707317,
+      "grad_norm": 1.1869080066680908,
+      "learning_rate": 4.266411896665967e-06,
+      "loss": 0.3936,
+      "step": 514
+    },
+    {
+      "epoch": 2.5121951219512195,
+      "grad_norm": 1.348928689956665,
+      "learning_rate": 4.263698673537309e-06,
+      "loss": 0.4986,
+      "step": 515
+    },
+    {
+      "epoch": 2.5170731707317073,
+      "grad_norm": 1.2809916734695435,
+      "learning_rate": 4.260981308349214e-06,
+      "loss": 0.6386,
+      "step": 516
+    },
+    {
+      "epoch": 2.521951219512195,
+      "grad_norm": 1.1547297239303589,
+      "learning_rate": 4.258259807483434e-06,
+      "loss": 0.5108,
+      "step": 517
+    },
+    {
+      "epoch": 2.526829268292683,
+      "grad_norm": 1.1655329465866089,
+      "learning_rate": 4.255534177331435e-06,
+      "loss": 0.5401,
+      "step": 518
+    },
+    {
+      "epoch": 2.5317073170731708,
+      "grad_norm": 1.127388596534729,
+      "learning_rate": 4.252804424294378e-06,
+      "loss": 0.5112,
+      "step": 519
+    },
+    {
+      "epoch": 2.5365853658536586,
+      "grad_norm": 1.0878446102142334,
+      "learning_rate": 4.25007055478311e-06,
+      "loss": 0.6333,
+      "step": 520
+    },
+    {
+      "epoch": 2.5414634146341464,
+      "grad_norm": 1.1850837469100952,
+      "learning_rate": 4.247332575218144e-06,
+      "loss": 0.4242,
+      "step": 521
+    },
+    {
+      "epoch": 2.546341463414634,
+      "grad_norm": 1.1731857061386108,
+      "learning_rate": 4.244590492029643e-06,
+      "loss": 0.6193,
+      "step": 522
+    },
+    {
+      "epoch": 2.551219512195122,
+      "grad_norm": 1.1234174966812134,
+      "learning_rate": 4.241844311657411e-06,
+      "loss": 0.3962,
+      "step": 523
+    },
+    {
+      "epoch": 2.55609756097561,
+      "grad_norm": 1.1408101320266724,
+      "learning_rate": 4.239094040550875e-06,
+      "loss": 0.326,
+      "step": 524
+    },
+    {
+      "epoch": 2.5609756097560976,
+      "grad_norm": 1.0664328336715698,
+      "learning_rate": 4.236339685169065e-06,
+      "loss": 0.4951,
+      "step": 525
+    },
+    {
+      "epoch": 2.5658536585365854,
+      "grad_norm": 1.0919582843780518,
+      "learning_rate": 4.233581251980604e-06,
+      "loss": 0.2973,
+      "step": 526
+    },
+    {
+      "epoch": 2.5707317073170732,
+      "grad_norm": 1.477919101715088,
+      "learning_rate": 4.230818747463696e-06,
+      "loss": 0.448,
+      "step": 527
+    },
+    {
+      "epoch": 2.575609756097561,
+      "grad_norm": 1.0629746913909912,
+      "learning_rate": 4.228052178106101e-06,
+      "loss": 0.4488,
+      "step": 528
+    },
+    {
+      "epoch": 2.580487804878049,
+      "grad_norm": 0.919862687587738,
+      "learning_rate": 4.2252815504051285e-06,
+      "loss": 0.2738,
+      "step": 529
+    },
+    {
+      "epoch": 2.5853658536585367,
+      "grad_norm": 1.5635737180709839,
+      "learning_rate": 4.222506870867618e-06,
+      "loss": 0.5376,
+      "step": 530
+    },
+    {
+      "epoch": 2.5902439024390245,
+      "grad_norm": 1.1145241260528564,
+      "learning_rate": 4.2197281460099245e-06,
+      "loss": 0.6359,
+      "step": 531
+    },
+    {
+      "epoch": 2.5951219512195123,
+      "grad_norm": 1.2374094724655151,
+      "learning_rate": 4.216945382357905e-06,
+      "loss": 0.5843,
+      "step": 532
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.363054633140564,
+      "learning_rate": 4.214158586446901e-06,
+      "loss": 0.9111,
+      "step": 533
+    },
+    {
+      "epoch": 2.604878048780488,
+      "grad_norm": 1.3563122749328613,
+      "learning_rate": 4.211367764821722e-06,
+      "loss": 0.8462,
+      "step": 534
+    },
+    {
+      "epoch": 2.6097560975609757,
+      "grad_norm": 1.401287317276001,
+      "learning_rate": 4.208572924036634e-06,
+      "loss": 0.4723,
+      "step": 535
+    },
+    {
+      "epoch": 2.6146341463414635,
+      "grad_norm": 1.1179980039596558,
+      "learning_rate": 4.2057740706553415e-06,
+      "loss": 0.4389,
+      "step": 536
+    },
+    {
+      "epoch": 2.6195121951219513,
+      "grad_norm": 1.9154226779937744,
+      "learning_rate": 4.202971211250971e-06,
+      "loss": 0.6297,
+      "step": 537
+    },
+    {
+      "epoch": 2.624390243902439,
+      "grad_norm": 1.1007169485092163,
+      "learning_rate": 4.200164352406061e-06,
+      "loss": 0.3373,
+      "step": 538
+    },
+    {
+      "epoch": 2.629268292682927,
+      "grad_norm": 1.1782920360565186,
+      "learning_rate": 4.197353500712539e-06,
+      "loss": 0.5813,
+      "step": 539
+    },
+    {
+      "epoch": 2.6341463414634148,
+      "grad_norm": 1.1510285139083862,
+      "learning_rate": 4.1945386627717115e-06,
+      "loss": 0.437,
+      "step": 540
+    },
+    {
+      "epoch": 2.6390243902439026,
+      "grad_norm": 2.090263843536377,
+      "learning_rate": 4.191719845194246e-06,
+      "loss": 0.582,
+      "step": 541
+    },
+    {
+      "epoch": 2.6439024390243904,
+      "grad_norm": 1.0950208902359009,
+      "learning_rate": 4.188897054600156e-06,
+      "loss": 0.5838,
+      "step": 542
+    },
+    {
+      "epoch": 2.648780487804878,
+      "grad_norm": 1.1003479957580566,
+      "learning_rate": 4.186070297618787e-06,
+      "loss": 0.6661,
+      "step": 543
+    },
+    {
+      "epoch": 2.653658536585366,
+      "grad_norm": 1.1567481756210327,
+      "learning_rate": 4.183239580888799e-06,
+      "loss": 0.6776,
+      "step": 544
+    },
+    {
+      "epoch": 2.658536585365854,
+      "grad_norm": 1.1681312322616577,
+      "learning_rate": 4.18040491105815e-06,
+      "loss": 0.4609,
+      "step": 545
+    },
+    {
+      "epoch": 2.6634146341463416,
+      "grad_norm": 1.1565169095993042,
+      "learning_rate": 4.177566294784085e-06,
+      "loss": 0.3778,
+      "step": 546
+    },
+    {
+      "epoch": 2.6682926829268294,
+      "grad_norm": 1.4641700983047485,
+      "learning_rate": 4.174723738733114e-06,
+      "loss": 0.7224,
+      "step": 547
+    },
+    {
+      "epoch": 2.6731707317073172,
+      "grad_norm": 1.062207579612732,
+      "learning_rate": 4.171877249581001e-06,
+      "loss": 0.613,
+      "step": 548
+    },
+    {
+      "epoch": 2.678048780487805,
+      "grad_norm": 1.1639560461044312,
+      "learning_rate": 4.169026834012748e-06,
+      "loss": 0.3883,
+      "step": 549
+    },
+    {
+      "epoch": 2.682926829268293,
+      "grad_norm": 1.0708937644958496,
+      "learning_rate": 4.166172498722577e-06,
+      "loss": 0.4024,
+      "step": 550
+    },
+    {
+      "epoch": 2.68780487804878,
+      "grad_norm": 1.7069330215454102,
+      "learning_rate": 4.163314250413913e-06,
+      "loss": 0.7562,
+      "step": 551
+    },
+    {
+      "epoch": 2.692682926829268,
+      "grad_norm": 1.2885770797729492,
+      "learning_rate": 4.160452095799378e-06,
+      "loss": 0.4543,
+      "step": 552
+    },
+    {
+      "epoch": 2.697560975609756,
+      "grad_norm": 0.9496511816978455,
+      "learning_rate": 4.157586041600759e-06,
+      "loss": 0.2383,
+      "step": 553
+    },
+    {
+      "epoch": 2.7024390243902436,
+      "grad_norm": 1.1848583221435547,
+      "learning_rate": 4.154716094549008e-06,
+      "loss": 0.5949,
+      "step": 554
+    },
+    {
+      "epoch": 2.7073170731707314,
+      "grad_norm": 1.0324264764785767,
+      "learning_rate": 4.151842261384217e-06,
+      "loss": 0.3728,
+      "step": 555
+    },
+    {
+      "epoch": 2.7121951219512193,
+      "grad_norm": 1.7933909893035889,
+      "learning_rate": 4.148964548855603e-06,
+      "loss": 0.9089,
+      "step": 556
+    },
+    {
+      "epoch": 2.717073170731707,
+      "grad_norm": 0.9467706084251404,
+      "learning_rate": 4.146082963721496e-06,
+      "loss": 0.3184,
+      "step": 557
+    },
+    {
+      "epoch": 2.721951219512195,
+      "grad_norm": 1.2086960077285767,
+      "learning_rate": 4.143197512749322e-06,
+      "loss": 1.074,
+      "step": 558
+    },
+    {
+      "epoch": 2.7268292682926827,
+      "grad_norm": 1.4787150621414185,
+      "learning_rate": 4.140308202715581e-06,
+      "loss": 0.76,
+      "step": 559
+    },
+    {
+      "epoch": 2.7317073170731705,
+      "grad_norm": 1.043819546699524,
+      "learning_rate": 4.13741504040584e-06,
+      "loss": 0.4029,
+      "step": 560
+    },
+    {
+      "epoch": 2.7365853658536583,
+      "grad_norm": 1.1737489700317383,
+      "learning_rate": 4.134518032614713e-06,
+      "loss": 0.4459,
+      "step": 561
+    },
+    {
+      "epoch": 2.741463414634146,
+      "grad_norm": 1.0608400106430054,
+      "learning_rate": 4.1316171861458445e-06,
+      "loss": 0.3549,
+      "step": 562
+    },
+    {
+      "epoch": 2.746341463414634,
+      "grad_norm": 1.1790850162506104,
+      "learning_rate": 4.128712507811893e-06,
+      "loss": 0.5961,
+      "step": 563
+    },
+    {
+      "epoch": 2.7512195121951217,
+      "grad_norm": 1.207923173904419,
+      "learning_rate": 4.125804004434517e-06,
+      "loss": 0.6095,
+      "step": 564
+    },
+    {
+      "epoch": 2.7560975609756095,
+      "grad_norm": 1.1621302366256714,
+      "learning_rate": 4.12289168284436e-06,
+      "loss": 0.3999,
+      "step": 565
+    },
+    {
+      "epoch": 2.7609756097560973,
+      "grad_norm": 1.0965276956558228,
+      "learning_rate": 4.119975549881029e-06,
+      "loss": 0.5668,
+      "step": 566
+    },
+    {
+      "epoch": 2.765853658536585,
+      "grad_norm": 1.069930076599121,
+      "learning_rate": 4.1170556123930846e-06,
+      "loss": 0.3537,
+      "step": 567
+    },
+    {
+      "epoch": 2.770731707317073,
+      "grad_norm": 1.0472149848937988,
+      "learning_rate": 4.114131877238021e-06,
+      "loss": 0.5583,
+      "step": 568
+    },
+    {
+      "epoch": 2.7756097560975608,
+      "grad_norm": 1.1573572158813477,
+      "learning_rate": 4.111204351282254e-06,
+      "loss": 0.4161,
+      "step": 569
+    },
+    {
+      "epoch": 2.7804878048780486,
+      "grad_norm": 0.8256755471229553,
+      "learning_rate": 4.108273041401098e-06,
+      "loss": 0.3972,
+      "step": 570
+    },
+    {
+      "epoch": 2.7853658536585364,
+      "grad_norm": 1.5649232864379883,
+      "learning_rate": 4.105337954478756e-06,
+      "loss": 0.8073,
+      "step": 571
+    },
+    {
+      "epoch": 2.790243902439024,
+      "grad_norm": 1.195081353187561,
+      "learning_rate": 4.102399097408304e-06,
+      "loss": 0.6545,
+      "step": 572
+    },
+    {
+      "epoch": 2.795121951219512,
+      "grad_norm": 1.0753878355026245,
+      "learning_rate": 4.099456477091667e-06,
+      "loss": 0.2882,
+      "step": 573
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.4838294982910156,
+      "learning_rate": 4.096510100439611e-06,
+      "loss": 0.7649,
+      "step": 574
+    },
+    {
+      "epoch": 2.8048780487804876,
+      "grad_norm": 1.0271984338760376,
+      "learning_rate": 4.093559974371725e-06,
+      "loss": 0.3012,
+      "step": 575
+    },
+    {
+      "epoch": 2.8097560975609754,
+      "grad_norm": 1.218367099761963,
+      "learning_rate": 4.0906061058164e-06,
+      "loss": 0.7672,
+      "step": 576
+    },
+    {
+      "epoch": 2.8146341463414632,
+      "grad_norm": 1.0342233180999756,
+      "learning_rate": 4.087648501710819e-06,
+      "loss": 0.3517,
+      "step": 577
+    },
+    {
+      "epoch": 2.819512195121951,
+      "grad_norm": 1.0327427387237549,
+      "learning_rate": 4.084687169000938e-06,
+      "loss": 0.5373,
+      "step": 578
+    },
+    {
+      "epoch": 2.824390243902439,
+      "grad_norm": 1.0352602005004883,
+      "learning_rate": 4.081722114641469e-06,
+      "loss": 0.4837,
+      "step": 579
+    },
+    {
+      "epoch": 2.8292682926829267,
+      "grad_norm": 1.1615251302719116,
+      "learning_rate": 4.0787533455958626e-06,
+      "loss": 0.3207,
+      "step": 580
+    },
+    {
+      "epoch": 2.8341463414634145,
+      "grad_norm": 1.2180888652801514,
+      "learning_rate": 4.075780868836296e-06,
+      "loss": 0.3952,
+      "step": 581
+    },
+    {
+      "epoch": 2.8390243902439023,
+      "grad_norm": 1.1467193365097046,
+      "learning_rate": 4.072804691343653e-06,
+      "loss": 0.4613,
+      "step": 582
+    },
+    {
+      "epoch": 2.84390243902439,
+      "grad_norm": 1.323662519454956,
+      "learning_rate": 4.069824820107507e-06,
+      "loss": 0.9519,
+      "step": 583
+    },
+    {
+      "epoch": 2.848780487804878,
+      "grad_norm": 1.386763572692871,
+      "learning_rate": 4.06684126212611e-06,
+      "loss": 0.2905,
+      "step": 584
+    },
+    {
+      "epoch": 2.8536585365853657,
+      "grad_norm": 1.1687908172607422,
+      "learning_rate": 4.063854024406369e-06,
+      "loss": 0.5389,
+      "step": 585
+    },
+    {
+      "epoch": 2.8585365853658535,
+      "grad_norm": 1.1061527729034424,
+      "learning_rate": 4.060863113963835e-06,
+      "loss": 0.4356,
+      "step": 586
+    },
+    {
+      "epoch": 2.8634146341463413,
+      "grad_norm": 1.1468473672866821,
+      "learning_rate": 4.057868537822683e-06,
+      "loss": 0.4687,
+      "step": 587
+    },
+    {
+      "epoch": 2.868292682926829,
+      "grad_norm": 1.0252629518508911,
+      "learning_rate": 4.054870303015695e-06,
+      "loss": 0.3387,
+      "step": 588
+    },
+    {
+      "epoch": 2.873170731707317,
+      "grad_norm": 1.4150055646896362,
+      "learning_rate": 4.05186841658425e-06,
+      "loss": 0.4078,
+      "step": 589
+    },
+    {
+      "epoch": 2.8780487804878048,
+      "grad_norm": 1.0989253520965576,
+      "learning_rate": 4.048862885578301e-06,
+      "loss": 0.4832,
+      "step": 590
+    },
+    {
+      "epoch": 2.8829268292682926,
+      "grad_norm": 1.1059105396270752,
+      "learning_rate": 4.045853717056358e-06,
+      "loss": 0.5439,
+      "step": 591
+    },
+    {
+      "epoch": 2.8878048780487804,
+      "grad_norm": 0.9113157391548157,
+      "learning_rate": 4.0428409180854775e-06,
+      "loss": 0.4421,
+      "step": 592
+    },
+    {
+      "epoch": 2.892682926829268,
+      "grad_norm": 1.0300334692001343,
+      "learning_rate": 4.039824495741238e-06,
+      "loss": 0.4173,
+      "step": 593
+    },
+    {
+      "epoch": 2.897560975609756,
+      "grad_norm": 0.995546817779541,
+      "learning_rate": 4.036804457107733e-06,
+      "loss": 0.4941,
+      "step": 594
+    },
+    {
+      "epoch": 2.902439024390244,
+      "grad_norm": 1.863033413887024,
+      "learning_rate": 4.0337808092775435e-06,
+      "loss": 0.6486,
+      "step": 595
+    },
+    {
+      "epoch": 2.9073170731707316,
+      "grad_norm": 0.9955878853797913,
+      "learning_rate": 4.030753559351728e-06,
+      "loss": 0.3511,
+      "step": 596
+    },
+    {
+      "epoch": 2.9121951219512194,
+      "grad_norm": 0.899255633354187,
+      "learning_rate": 4.027722714439808e-06,
+      "loss": 0.3276,
+      "step": 597
+    },
+    {
+      "epoch": 2.9170731707317072,
+      "grad_norm": 1.5171446800231934,
+      "learning_rate": 4.024688281659743e-06,
+      "loss": 0.8516,
+      "step": 598
+    },
+    {
+      "epoch": 2.921951219512195,
+      "grad_norm": 1.933835744857788,
+      "learning_rate": 4.021650268137924e-06,
+      "loss": 0.5823,
+      "step": 599
+    },
+    {
+      "epoch": 2.926829268292683,
+      "grad_norm": 0.9740194082260132,
+      "learning_rate": 4.018608681009143e-06,
+      "loss": 0.4459,
+      "step": 600
+    },
+    {
+      "epoch": 2.9317073170731707,
+      "grad_norm": 1.0849703550338745,
+      "learning_rate": 4.015563527416596e-06,
+      "loss": 0.5311,
+      "step": 601
+    },
+    {
+      "epoch": 2.9365853658536585,
+      "grad_norm": 1.0849350690841675,
+      "learning_rate": 4.012514814511844e-06,
+      "loss": 0.4696,
+      "step": 602
+    },
+    {
+      "epoch": 2.9414634146341463,
+      "grad_norm": 0.9327961802482605,
+      "learning_rate": 4.009462549454816e-06,
+      "loss": 0.5035,
+      "step": 603
+    },
+    {
+      "epoch": 2.946341463414634,
+      "grad_norm": 1.235766053199768,
+      "learning_rate": 4.006406739413775e-06,
+      "loss": 0.5486,
+      "step": 604
+    },
+    {
+      "epoch": 2.951219512195122,
+      "grad_norm": 1.133878231048584,
+      "learning_rate": 4.003347391565317e-06,
+      "loss": 0.4787,
+      "step": 605
+    },
+    {
+      "epoch": 2.9560975609756097,
+      "grad_norm": 1.274250864982605,
+      "learning_rate": 4.000284513094342e-06,
+      "loss": 0.5338,
+      "step": 606
+    },
+    {
+      "epoch": 2.9609756097560975,
+      "grad_norm": 1.0265629291534424,
+      "learning_rate": 3.997218111194042e-06,
+      "loss": 0.4736,
+      "step": 607
+    },
+    {
+      "epoch": 2.9658536585365853,
+      "grad_norm": 1.0662497282028198,
+      "learning_rate": 3.994148193065886e-06,
+      "loss": 0.3592,
+      "step": 608
+    },
+    {
+      "epoch": 2.970731707317073,
+      "grad_norm": 0.9560685753822327,
+      "learning_rate": 3.991074765919598e-06,
+      "loss": 0.3087,
+      "step": 609
+    },
+    {
+      "epoch": 2.975609756097561,
+      "grad_norm": 1.2305241823196411,
+      "learning_rate": 3.987997836973147e-06,
+      "loss": 0.4249,
+      "step": 610
+    },
+    {
+      "epoch": 2.9804878048780488,
+      "grad_norm": 1.2358429431915283,
+      "learning_rate": 3.984917413452721e-06,
+      "loss": 0.4519,
+      "step": 611
+    },
+    {
+      "epoch": 2.9853658536585366,
+      "grad_norm": 1.5506497621536255,
+      "learning_rate": 3.981833502592717e-06,
+      "loss": 0.7231,
+      "step": 612
+    },
+    {
+      "epoch": 2.9902439024390244,
+      "grad_norm": 1.6102908849716187,
+      "learning_rate": 3.978746111635725e-06,
+      "loss": 0.2819,
+      "step": 613
+    },
+    {
+      "epoch": 2.995121951219512,
+      "grad_norm": 1.0251195430755615,
+      "learning_rate": 3.9756552478325045e-06,
+      "loss": 0.4537,
+      "step": 614
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.0950148105621338,
+      "learning_rate": 3.972560918441972e-06,
+      "loss": 0.2831,
+      "step": 615
+    },
+    {
+      "epoch": 3.004878048780488,
+      "grad_norm": 0.8800443410873413,
+      "learning_rate": 3.969463130731183e-06,
+      "loss": 0.3523,
+      "step": 616
+    },
+    {
+      "epoch": 3.0097560975609756,
+      "grad_norm": 1.7478209733963013,
+      "learning_rate": 3.966361891975316e-06,
+      "loss": 0.3588,
+      "step": 617
+    },
+    {
+      "epoch": 3.0146341463414634,
+      "grad_norm": 1.0712541341781616,
+      "learning_rate": 3.963257209457652e-06,
+      "loss": 0.4436,
+      "step": 618
+    },
+    {
+      "epoch": 3.0195121951219512,
+      "grad_norm": 0.8614863753318787,
+      "learning_rate": 3.960149090469561e-06,
+      "loss": 0.2135,
+      "step": 619
+    },
+    {
+      "epoch": 3.024390243902439,
+      "grad_norm": 1.0551412105560303,
+      "learning_rate": 3.957037542310484e-06,
+      "loss": 0.2397,
+      "step": 620
+    },
+    {
+      "epoch": 3.029268292682927,
+      "grad_norm": 1.0886740684509277,
+      "learning_rate": 3.953922572287915e-06,
+      "loss": 0.4147,
+      "step": 621
+    },
+    {
+      "epoch": 3.0341463414634147,
+      "grad_norm": 1.4966646432876587,
+      "learning_rate": 3.950804187717384e-06,
+      "loss": 0.5208,
+      "step": 622
+    },
+    {
+      "epoch": 3.0390243902439025,
+      "grad_norm": 1.1856284141540527,
+      "learning_rate": 3.947682395922439e-06,
+      "loss": 0.6339,
+      "step": 623
+    },
+    {
+      "epoch": 3.0439024390243903,
+      "grad_norm": 0.9861780405044556,
+      "learning_rate": 3.9445572042346346e-06,
+      "loss": 0.1636,
+      "step": 624
+    },
+    {
+      "epoch": 3.048780487804878,
+      "grad_norm": 1.1708422899246216,
+      "learning_rate": 3.941428619993505e-06,
+      "loss": 0.3336,
+      "step": 625
+    },
+    {
+      "epoch": 3.053658536585366,
+      "grad_norm": 1.0953327417373657,
+      "learning_rate": 3.938296650546552e-06,
+      "loss": 0.2564,
+      "step": 626
+    },
+    {
+      "epoch": 3.0585365853658537,
+      "grad_norm": 1.671927571296692,
+      "learning_rate": 3.935161303249231e-06,
+      "loss": 0.4264,
+      "step": 627
+    },
+    {
+      "epoch": 3.0634146341463415,
+      "grad_norm": 2.4444899559020996,
+      "learning_rate": 3.932022585464928e-06,
+      "loss": 0.4284,
+      "step": 628
+    },
+    {
+      "epoch": 3.0682926829268293,
+      "grad_norm": 1.0862361192703247,
+      "learning_rate": 3.928880504564943e-06,
+      "loss": 0.3262,
+      "step": 629
+    },
+    {
+      "epoch": 3.073170731707317,
+      "grad_norm": 0.9641832709312439,
+      "learning_rate": 3.92573506792848e-06,
+      "loss": 0.3027,
+      "step": 630
+    },
+    {
+      "epoch": 3.078048780487805,
+      "grad_norm": 1.1519198417663574,
+      "learning_rate": 3.9225862829426184e-06,
+      "loss": 0.3024,
+      "step": 631
+    },
+    {
+      "epoch": 3.0829268292682928,
+      "grad_norm": 1.1144092082977295,
+      "learning_rate": 3.919434157002303e-06,
+      "loss": 0.4163,
+      "step": 632
+    },
+    {
+      "epoch": 3.0878048780487806,
+      "grad_norm": 0.8596329092979431,
+      "learning_rate": 3.916278697510325e-06,
+      "loss": 0.2329,
+      "step": 633
+    },
+    {
+      "epoch": 3.0926829268292684,
+      "grad_norm": 1.1805473566055298,
+      "learning_rate": 3.913119911877305e-06,
+      "loss": 0.4784,
+      "step": 634
+    },
+    {
+      "epoch": 3.097560975609756,
+      "grad_norm": 1.1723095178604126,
+      "learning_rate": 3.909957807521674e-06,
+      "loss": 0.2264,
+      "step": 635
+    },
+    {
+      "epoch": 3.102439024390244,
+      "grad_norm": 1.0031239986419678,
+      "learning_rate": 3.906792391869657e-06,
+      "loss": 0.3434,
+      "step": 636
+    },
+    {
+      "epoch": 3.107317073170732,
+      "grad_norm": 1.2430922985076904,
+      "learning_rate": 3.903623672355258e-06,
+      "loss": 0.3462,
+      "step": 637
+    },
+    {
+      "epoch": 3.1121951219512196,
+      "grad_norm": 0.9756274223327637,
+      "learning_rate": 3.900451656420237e-06,
+      "loss": 0.3149,
+      "step": 638
+    },
+    {
+      "epoch": 3.1170731707317074,
+      "grad_norm": 0.9727137684822083,
+      "learning_rate": 3.897276351514097e-06,
+      "loss": 0.2312,
+      "step": 639
+    },
+    {
+      "epoch": 3.1219512195121952,
+      "grad_norm": 1.388515830039978,
+      "learning_rate": 3.894097765094065e-06,
+      "loss": 0.5313,
+      "step": 640
+    },
+    {
+      "epoch": 3.126829268292683,
+      "grad_norm": 0.838641345500946,
+      "learning_rate": 3.890915904625075e-06,
+      "loss": 0.2566,
+      "step": 641
+    },
+    {
+      "epoch": 3.131707317073171,
+      "grad_norm": 1.7275861501693726,
+      "learning_rate": 3.887730777579751e-06,
+      "loss": 0.5126,
+      "step": 642
+    },
+    {
+      "epoch": 3.1365853658536587,
+      "grad_norm": 1.4002543687820435,
+      "learning_rate": 3.884542391438387e-06,
+      "loss": 0.6454,
+      "step": 643
+    },
+    {
+      "epoch": 3.1414634146341465,
+      "grad_norm": 1.29892897605896,
+      "learning_rate": 3.88135075368893e-06,
+      "loss": 0.8209,
+      "step": 644
+    },
+    {
+      "epoch": 3.1463414634146343,
+      "grad_norm": 1.2062216997146606,
+      "learning_rate": 3.878155871826968e-06,
+      "loss": 0.375,
+      "step": 645
+    },
+    {
+      "epoch": 3.151219512195122,
+      "grad_norm": 1.0101869106292725,
+      "learning_rate": 3.874957753355701e-06,
+      "loss": 0.3504,
+      "step": 646
+    },
+    {
+      "epoch": 3.15609756097561,
+      "grad_norm": 1.0103880167007446,
+      "learning_rate": 3.8717564057859365e-06,
+      "loss": 0.5863,
+      "step": 647
+    },
+    {
+      "epoch": 3.1609756097560977,
+      "grad_norm": 1.137888789176941,
+      "learning_rate": 3.868551836636063e-06,
+      "loss": 0.526,
+      "step": 648
+    },
+    {
+      "epoch": 3.1658536585365855,
+      "grad_norm": 1.09031343460083,
+      "learning_rate": 3.865344053432035e-06,
+      "loss": 0.282,
+      "step": 649
+    },
+    {
+      "epoch": 3.1707317073170733,
+      "grad_norm": 1.4528731107711792,
+      "learning_rate": 3.862133063707353e-06,
+      "loss": 0.4431,
+      "step": 650
+    },
+    {
+      "epoch": 3.175609756097561,
+      "grad_norm": 1.0537135601043701,
+      "learning_rate": 3.858918875003053e-06,
+      "loss": 0.3105,
+      "step": 651
+    },
+    {
+      "epoch": 3.180487804878049,
+      "grad_norm": 1.1723111867904663,
+      "learning_rate": 3.855701494867679e-06,
+      "loss": 0.3701,
+      "step": 652
+    },
+    {
+      "epoch": 3.1853658536585368,
+      "grad_norm": 0.9918485879898071,
+      "learning_rate": 3.852480930857275e-06,
+      "loss": 0.5181,
+      "step": 653
+    },
+    {
+      "epoch": 3.1902439024390246,
+      "grad_norm": 1.0648022890090942,
+      "learning_rate": 3.849257190535356e-06,
+      "loss": 0.2929,
+      "step": 654
+    },
+    {
+      "epoch": 3.1951219512195124,
+      "grad_norm": 1.5257065296173096,
+      "learning_rate": 3.846030281472902e-06,
+      "loss": 0.6572,
+      "step": 655
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.9846480488777161,
+      "learning_rate": 3.842800211248333e-06,
+      "loss": 0.3101,
+      "step": 656
+    },
+    {
+      "epoch": 3.204878048780488,
+      "grad_norm": 1.5114487409591675,
+      "learning_rate": 3.839566987447492e-06,
+      "loss": 0.4957,
+      "step": 657
+    },
+    {
+      "epoch": 3.209756097560976,
+      "grad_norm": 1.0654270648956299,
+      "learning_rate": 3.8363306176636296e-06,
+      "loss": 0.5607,
+      "step": 658
+    },
+    {
+      "epoch": 3.2146341463414636,
+      "grad_norm": 1.3029534816741943,
+      "learning_rate": 3.833091109497384e-06,
+      "loss": 0.6385,
+      "step": 659
+    },
+    {
+      "epoch": 3.2195121951219514,
+      "grad_norm": 0.9915573000907898,
+      "learning_rate": 3.829848470556765e-06,
+      "loss": 0.2152,
+      "step": 660
+    },
+    {
+      "epoch": 3.2243902439024392,
+      "grad_norm": 1.5952492952346802,
+      "learning_rate": 3.8266027084571335e-06,
+      "loss": 0.426,
+      "step": 661
+    },
+    {
+      "epoch": 3.229268292682927,
+      "grad_norm": 1.1406885385513306,
+      "learning_rate": 3.823353830821187e-06,
+      "loss": 0.1969,
+      "step": 662
+    },
+    {
+      "epoch": 3.234146341463415,
+      "grad_norm": 1.037661075592041,
+      "learning_rate": 3.820101845278937e-06,
+      "loss": 0.3808,
+      "step": 663
+    },
+    {
+      "epoch": 3.2390243902439027,
+      "grad_norm": 1.2333792448043823,
+      "learning_rate": 3.816846759467696e-06,
+      "loss": 0.3618,
+      "step": 664
+    },
+    {
+      "epoch": 3.2439024390243905,
+      "grad_norm": 1.3008856773376465,
+      "learning_rate": 3.8135885810320587e-06,
+      "loss": 0.5053,
+      "step": 665
+    },
+    {
+      "epoch": 3.2487804878048783,
+      "grad_norm": 0.9773457646369934,
+      "learning_rate": 3.810327317623881e-06,
+      "loss": 0.3025,
+      "step": 666
+    },
+    {
+      "epoch": 3.253658536585366,
+      "grad_norm": 1.1375815868377686,
+      "learning_rate": 3.8070629769022628e-06,
+      "loss": 0.4466,
+      "step": 667
+    },
+    {
+      "epoch": 3.258536585365854,
+      "grad_norm": 1.141038417816162,
+      "learning_rate": 3.8037955665335335e-06,
+      "loss": 0.3302,
+      "step": 668
+    },
+    {
+      "epoch": 3.2634146341463417,
+      "grad_norm": 1.2012133598327637,
+      "learning_rate": 3.800525094191231e-06,
+      "loss": 0.4628,
+      "step": 669
+    },
+    {
+      "epoch": 3.2682926829268295,
+      "grad_norm": 1.0996290445327759,
+      "learning_rate": 3.797251567556083e-06,
+      "loss": 0.3041,
+      "step": 670
+    },
+    {
+      "epoch": 3.2731707317073173,
+      "grad_norm": 0.9800164699554443,
+      "learning_rate": 3.793974994315991e-06,
+      "loss": 0.1894,
+      "step": 671
+    },
+    {
+      "epoch": 3.278048780487805,
+      "grad_norm": 1.1430602073669434,
+      "learning_rate": 3.790695382166013e-06,
+      "loss": 0.4311,
+      "step": 672
+    },
+    {
+      "epoch": 3.2829268292682925,
+      "grad_norm": 1.1268576383590698,
+      "learning_rate": 3.7874127388083415e-06,
+      "loss": 0.3273,
+      "step": 673
+    },
+    {
+      "epoch": 3.2878048780487803,
+      "grad_norm": 1.1353093385696411,
+      "learning_rate": 3.7841270719522895e-06,
+      "loss": 0.3813,
+      "step": 674
+    },
+    {
+      "epoch": 3.292682926829268,
+      "grad_norm": 1.353868007659912,
+      "learning_rate": 3.7808383893142692e-06,
+      "loss": 0.1895,
+      "step": 675
+    },
+    {
+      "epoch": 3.297560975609756,
+      "grad_norm": 1.223840355873108,
+      "learning_rate": 3.7775466986177763e-06,
+      "loss": 0.31,
+      "step": 676
+    },
+    {
+      "epoch": 3.3024390243902437,
+      "grad_norm": 1.1622555255889893,
+      "learning_rate": 3.774252007593371e-06,
+      "loss": 0.1965,
+      "step": 677
+    },
+    {
+      "epoch": 3.3073170731707315,
+      "grad_norm": 1.094815969467163,
+      "learning_rate": 3.7709543239786593e-06,
+      "loss": 0.5145,
+      "step": 678
+    },
+    {
+      "epoch": 3.3121951219512193,
+      "grad_norm": 1.3806910514831543,
+      "learning_rate": 3.767653655518277e-06,
+      "loss": 0.3225,
+      "step": 679
+    },
+    {
+      "epoch": 3.317073170731707,
+      "grad_norm": 1.2605657577514648,
+      "learning_rate": 3.7643500099638673e-06,
+      "loss": 0.3046,
+      "step": 680
+    },
+    {
+      "epoch": 3.321951219512195,
+      "grad_norm": 1.0325559377670288,
+      "learning_rate": 3.7610433950740667e-06,
+      "loss": 0.6238,
+      "step": 681
+    },
+    {
+      "epoch": 3.3268292682926828,
+      "grad_norm": 1.5511088371276855,
+      "learning_rate": 3.757733818614485e-06,
+      "loss": 0.4794,
+      "step": 682
+    },
+    {
+      "epoch": 3.3317073170731706,
+      "grad_norm": 0.9807575345039368,
+      "learning_rate": 3.7544212883576856e-06,
+      "loss": 0.4027,
+      "step": 683
+    },
+    {
+      "epoch": 3.3365853658536584,
+      "grad_norm": 1.446093201637268,
+      "learning_rate": 3.751105812083172e-06,
+      "loss": 0.2948,
+      "step": 684
+    },
+    {
+      "epoch": 3.341463414634146,
+      "grad_norm": 1.402915120124817,
+      "learning_rate": 3.7477873975773655e-06,
+      "loss": 0.5001,
+      "step": 685
+    },
+    {
+      "epoch": 3.346341463414634,
+      "grad_norm": 1.3973482847213745,
+      "learning_rate": 3.7444660526335853e-06,
+      "loss": 0.4528,
+      "step": 686
+    },
+    {
+      "epoch": 3.351219512195122,
+      "grad_norm": 2.8907651901245117,
+      "learning_rate": 3.741141785052036e-06,
+      "loss": 0.7388,
+      "step": 687
+    },
+    {
+      "epoch": 3.3560975609756096,
+      "grad_norm": 1.2426471710205078,
+      "learning_rate": 3.737814602639784e-06,
+      "loss": 0.4963,
+      "step": 688
+    },
+    {
+      "epoch": 3.3609756097560974,
+      "grad_norm": 0.959717869758606,
+      "learning_rate": 3.7344845132107427e-06,
+      "loss": 0.4038,
+      "step": 689
+    },
+    {
+      "epoch": 3.3658536585365852,
+      "grad_norm": 1.3278104066848755,
+      "learning_rate": 3.731151524585651e-06,
+      "loss": 0.4465,
+      "step": 690
+    },
+    {
+      "epoch": 3.370731707317073,
+      "grad_norm": 1.3650119304656982,
+      "learning_rate": 3.7278156445920584e-06,
+      "loss": 0.8163,
+      "step": 691
+    },
+    {
+      "epoch": 3.375609756097561,
+      "grad_norm": 0.8338037729263306,
+      "learning_rate": 3.724476881064303e-06,
+      "loss": 0.251,
+      "step": 692
+    },
+    {
+      "epoch": 3.3804878048780487,
+      "grad_norm": 0.9044190049171448,
+      "learning_rate": 3.721135241843496e-06,
+      "loss": 0.3342,
+      "step": 693
+    },
+    {
+      "epoch": 3.3853658536585365,
+      "grad_norm": 1.1551542282104492,
+      "learning_rate": 3.7177907347775016e-06,
+      "loss": 0.452,
+      "step": 694
+    },
+    {
+      "epoch": 3.3902439024390243,
+      "grad_norm": 2.0845346450805664,
+      "learning_rate": 3.71444336772092e-06,
+      "loss": 0.5854,
+      "step": 695
+    },
+    {
+      "epoch": 3.395121951219512,
+      "grad_norm": 1.3493772745132446,
+      "learning_rate": 3.711093148535068e-06,
+      "loss": 0.659,
+      "step": 696
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.1673591136932373,
+      "learning_rate": 3.707740085087959e-06,
+      "loss": 0.2783,
+      "step": 697
+    },
+    {
+      "epoch": 3.4048780487804877,
+      "grad_norm": 0.9679506421089172,
+      "learning_rate": 3.7043841852542884e-06,
+      "loss": 0.4151,
+      "step": 698
+    },
+    {
+      "epoch": 3.4097560975609755,
+      "grad_norm": 0.9405661225318909,
+      "learning_rate": 3.701025456915411e-06,
+      "loss": 0.2932,
+      "step": 699
+    },
+    {
+      "epoch": 3.4146341463414633,
+      "grad_norm": 1.0707734823226929,
+      "learning_rate": 3.697663907959327e-06,
+      "loss": 0.3878,
+      "step": 700
+    },
+    {
+      "epoch": 3.419512195121951,
+      "grad_norm": 1.5344734191894531,
+      "learning_rate": 3.6942995462806574e-06,
+      "loss": 0.6712,
+      "step": 701
+    },
+    {
+      "epoch": 3.424390243902439,
+      "grad_norm": 0.9002655148506165,
+      "learning_rate": 3.6909323797806314e-06,
+      "loss": 0.3109,
+      "step": 702
+    },
+    {
+      "epoch": 3.4292682926829268,
+      "grad_norm": 1.6577272415161133,
+      "learning_rate": 3.6875624163670635e-06,
+      "loss": 0.614,
+      "step": 703
+    },
+    {
+      "epoch": 3.4341463414634146,
+      "grad_norm": 1.2623207569122314,
+      "learning_rate": 3.6841896639543394e-06,
+      "loss": 0.3651,
+      "step": 704
+    },
+    {
+      "epoch": 3.4390243902439024,
+      "grad_norm": 1.0374921560287476,
+      "learning_rate": 3.6808141304633924e-06,
+      "loss": 0.3447,
+      "step": 705
+    },
+    {
+      "epoch": 3.44390243902439,
+      "grad_norm": 1.3352950811386108,
+      "learning_rate": 3.6774358238216878e-06,
+      "loss": 0.3566,
+      "step": 706
+    },
+    {
+      "epoch": 3.448780487804878,
+      "grad_norm": 1.2781732082366943,
+      "learning_rate": 3.6740547519632048e-06,
+      "loss": 0.3003,
+      "step": 707
+    },
+    {
+      "epoch": 3.453658536585366,
+      "grad_norm": 0.8481651544570923,
+      "learning_rate": 3.670670922828414e-06,
+      "loss": 0.3505,
+      "step": 708
+    },
+    {
+      "epoch": 3.4585365853658536,
+      "grad_norm": 1.3136425018310547,
+      "learning_rate": 3.667284344364264e-06,
+      "loss": 0.3415,
+      "step": 709
+    },
+    {
+      "epoch": 3.4634146341463414,
+      "grad_norm": 1.1548678874969482,
+      "learning_rate": 3.6638950245241604e-06,
+      "loss": 0.4958,
+      "step": 710
+    },
+    {
+      "epoch": 3.4682926829268292,
+      "grad_norm": 1.2795634269714355,
+      "learning_rate": 3.660502971267945e-06,
+      "loss": 0.3697,
+      "step": 711
+    },
+    {
+      "epoch": 3.473170731707317,
+      "grad_norm": 0.9530194997787476,
+      "learning_rate": 3.65710819256188e-06,
+      "loss": 0.1465,
+      "step": 712
+    },
+    {
+      "epoch": 3.478048780487805,
+      "grad_norm": 1.041036605834961,
+      "learning_rate": 3.65371069637863e-06,
+      "loss": 0.3663,
+      "step": 713
+    },
+    {
+      "epoch": 3.4829268292682927,
+      "grad_norm": 1.2768229246139526,
+      "learning_rate": 3.650310490697238e-06,
+      "loss": 0.534,
+      "step": 714
+    },
+    {
+      "epoch": 3.4878048780487805,
+      "grad_norm": 1.2672673463821411,
+      "learning_rate": 3.646907583503114e-06,
+      "loss": 0.5489,
+      "step": 715
+    },
+    {
+      "epoch": 3.4926829268292683,
+      "grad_norm": 1.1680879592895508,
+      "learning_rate": 3.6435019827880093e-06,
+      "loss": 0.3485,
+      "step": 716
+    },
+    {
+      "epoch": 3.497560975609756,
+      "grad_norm": 1.1077998876571655,
+      "learning_rate": 3.640093696550003e-06,
+      "loss": 0.4058,
+      "step": 717
+    },
+    {
+      "epoch": 3.502439024390244,
+      "grad_norm": 2.137748956680298,
+      "learning_rate": 3.6366827327934817e-06,
+      "loss": 0.3986,
+      "step": 718
+    },
+    {
+      "epoch": 3.5073170731707317,
+      "grad_norm": 1.1344432830810547,
+      "learning_rate": 3.6332690995291176e-06,
+      "loss": 0.4945,
+      "step": 719
+    },
+    {
+      "epoch": 3.5121951219512195,
+      "grad_norm": 1.4760767221450806,
+      "learning_rate": 3.6298528047738545e-06,
+      "loss": 1.1914,
+      "step": 720
+    },
+    {
+      "epoch": 3.5170731707317073,
+      "grad_norm": 1.1157302856445312,
+      "learning_rate": 3.626433856550886e-06,
+      "loss": 0.5676,
+      "step": 721
+    },
+    {
+      "epoch": 3.521951219512195,
+      "grad_norm": 0.9951807260513306,
+      "learning_rate": 3.623012262889637e-06,
+      "loss": 0.4915,
+      "step": 722
+    },
+    {
+      "epoch": 3.526829268292683,
+      "grad_norm": 1.1941821575164795,
+      "learning_rate": 3.6195880318257465e-06,
+      "loss": 0.5551,
+      "step": 723
+    },
+    {
+      "epoch": 3.5317073170731708,
+      "grad_norm": 1.177983045578003,
+      "learning_rate": 3.616161171401046e-06,
+      "loss": 0.6706,
+      "step": 724
+    },
+    {
+      "epoch": 3.5365853658536586,
+      "grad_norm": 1.0280420780181885,
+      "learning_rate": 3.612731689663542e-06,
+      "loss": 0.3326,
+      "step": 725
+    },
+    {
+      "epoch": 3.5414634146341464,
+      "grad_norm": 0.9995579123497009,
+      "learning_rate": 3.6092995946673996e-06,
+      "loss": 0.5511,
+      "step": 726
+    },
+    {
+      "epoch": 3.546341463414634,
+      "grad_norm": 1.1944303512573242,
+      "learning_rate": 3.605864894472918e-06,
+      "loss": 0.3593,
+      "step": 727
+    },
+    {
+      "epoch": 3.551219512195122,
+      "grad_norm": 1.1579818725585938,
+      "learning_rate": 3.602427597146516e-06,
+      "loss": 0.5585,
+      "step": 728
+    },
+    {
+      "epoch": 3.55609756097561,
+      "grad_norm": 1.5750212669372559,
+      "learning_rate": 3.5989877107607134e-06,
+      "loss": 0.4691,
+      "step": 729
+    },
+    {
+      "epoch": 3.5609756097560976,
+      "grad_norm": 1.23482346534729,
+      "learning_rate": 3.5955452433941075e-06,
+      "loss": 0.5383,
+      "step": 730
+    },
+    {
+      "epoch": 3.5658536585365854,
+      "grad_norm": 1.1154356002807617,
+      "learning_rate": 3.5921002031313586e-06,
+      "loss": 0.2954,
+      "step": 731
+    },
+    {
+      "epoch": 3.5707317073170732,
+      "grad_norm": 1.1274775266647339,
+      "learning_rate": 3.58865259806317e-06,
+      "loss": 0.2895,
+      "step": 732
+    },
+    {
+      "epoch": 3.575609756097561,
+      "grad_norm": 1.207263708114624,
+      "learning_rate": 3.585202436286267e-06,
+      "loss": 0.507,
+      "step": 733
+    },
+    {
+      "epoch": 3.580487804878049,
+      "grad_norm": 1.0866899490356445,
+      "learning_rate": 3.581749725903381e-06,
+      "loss": 0.4687,
+      "step": 734
+    },
+    {
+      "epoch": 3.5853658536585367,
+      "grad_norm": 1.0456942319869995,
+      "learning_rate": 3.5782944750232274e-06,
+      "loss": 0.4353,
+      "step": 735
+    },
+    {
+      "epoch": 3.5902439024390245,
+      "grad_norm": 0.9194226264953613,
+      "learning_rate": 3.574836691760489e-06,
+      "loss": 0.145,
+      "step": 736
+    },
+    {
+      "epoch": 3.5951219512195123,
+      "grad_norm": 1.249432921409607,
+      "learning_rate": 3.571376384235795e-06,
+      "loss": 0.4227,
+      "step": 737
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.0487442016601562,
+      "learning_rate": 3.5679135605757035e-06,
+      "loss": 0.2977,
+      "step": 738
+    },
+    {
+      "epoch": 3.604878048780488,
+      "grad_norm": 1.0600451231002808,
+      "learning_rate": 3.564448228912682e-06,
+      "loss": 0.2691,
+      "step": 739
+    },
+    {
+      "epoch": 3.6097560975609757,
+      "grad_norm": 1.085017442703247,
+      "learning_rate": 3.5609803973850877e-06,
+      "loss": 0.3742,
+      "step": 740
+    },
+    {
+      "epoch": 3.6146341463414635,
+      "grad_norm": 1.1416304111480713,
+      "learning_rate": 3.557510074137147e-06,
+      "loss": 0.4039,
+      "step": 741
+    },
+    {
+      "epoch": 3.6195121951219513,
+      "grad_norm": 0.9345313310623169,
+      "learning_rate": 3.554037267318942e-06,
+      "loss": 0.4048,
+      "step": 742
+    },
+    {
+      "epoch": 3.624390243902439,
+      "grad_norm": 1.1760103702545166,
+      "learning_rate": 3.5505619850863847e-06,
+      "loss": 0.3144,
+      "step": 743
+    },
+    {
+      "epoch": 3.629268292682927,
+      "grad_norm": 1.1537894010543823,
+      "learning_rate": 3.5470842356012007e-06,
+      "loss": 0.1935,
+      "step": 744
+    },
+    {
+      "epoch": 3.6341463414634148,
+      "grad_norm": 1.1496186256408691,
+      "learning_rate": 3.5436040270309113e-06,
+      "loss": 0.4762,
+      "step": 745
+    },
+    {
+      "epoch": 3.6390243902439026,
+      "grad_norm": 0.907806932926178,
+      "learning_rate": 3.540121367548811e-06,
+      "loss": 0.2527,
+      "step": 746
+    },
+    {
+      "epoch": 3.6439024390243904,
+      "grad_norm": 1.072767972946167,
+      "learning_rate": 3.5366362653339524e-06,
+      "loss": 0.6001,
+      "step": 747
+    },
+    {
+      "epoch": 3.648780487804878,
+      "grad_norm": 1.095596194267273,
+      "learning_rate": 3.533148728571124e-06,
+      "loss": 0.2118,
+      "step": 748
+    },
+    {
+      "epoch": 3.653658536585366,
+      "grad_norm": 1.1593589782714844,
+      "learning_rate": 3.5296587654508317e-06,
+      "loss": 0.4225,
+      "step": 749
+    },
+    {
+      "epoch": 3.658536585365854,
+      "grad_norm": 1.5576963424682617,
+      "learning_rate": 3.526166384169279e-06,
+      "loss": 0.6524,
+      "step": 750
+    },
+    {
+      "epoch": 3.6634146341463416,
+      "grad_norm": 1.4875208139419556,
+      "learning_rate": 3.5226715929283507e-06,
+      "loss": 0.3488,
+      "step": 751
+    },
+    {
+      "epoch": 3.6682926829268294,
+      "grad_norm": 0.9630670547485352,
+      "learning_rate": 3.519174399935588e-06,
+      "loss": 0.2533,
+      "step": 752
+    },
+    {
+      "epoch": 3.6731707317073172,
+      "grad_norm": 0.9532560110092163,
+      "learning_rate": 3.5156748134041767e-06,
+      "loss": 0.1669,
+      "step": 753
+    },
+    {
+      "epoch": 3.678048780487805,
+      "grad_norm": 1.3903883695602417,
+      "learning_rate": 3.5121728415529203e-06,
+      "loss": 0.7186,
+      "step": 754
+    },
+    {
+      "epoch": 3.682926829268293,
+      "grad_norm": 1.2028529644012451,
+      "learning_rate": 3.5086684926062266e-06,
+      "loss": 0.3373,
+      "step": 755
+    },
+    {
+      "epoch": 3.68780487804878,
+      "grad_norm": 1.145927906036377,
+      "learning_rate": 3.505161774794085e-06,
+      "loss": 0.4091,
+      "step": 756
+    },
+    {
+      "epoch": 3.692682926829268,
+      "grad_norm": 1.0629132986068726,
+      "learning_rate": 3.5016526963520474e-06,
+      "loss": 0.2756,
+      "step": 757
+    },
+    {
+      "epoch": 3.697560975609756,
+      "grad_norm": 1.2821991443634033,
+      "learning_rate": 3.498141265521212e-06,
+      "loss": 0.7792,
+      "step": 758
+    },
+    {
+      "epoch": 3.7024390243902436,
+      "grad_norm": 1.0347892045974731,
+      "learning_rate": 3.4946274905481997e-06,
+      "loss": 0.2906,
+      "step": 759
+    },
+    {
+      "epoch": 3.7073170731707314,
+      "grad_norm": 1.212327003479004,
+      "learning_rate": 3.4911113796851364e-06,
+      "loss": 0.4449,
+      "step": 760
+    },
+    {
+      "epoch": 3.7121951219512193,
+      "grad_norm": 1.1393001079559326,
+      "learning_rate": 3.487592941189636e-06,
+      "loss": 0.2547,
+      "step": 761
+    },
+    {
+      "epoch": 3.717073170731707,
+      "grad_norm": 1.648065209388733,
+      "learning_rate": 3.484072183324776e-06,
+      "loss": 0.7522,
+      "step": 762
+    },
+    {
+      "epoch": 3.721951219512195,
+      "grad_norm": 1.114680290222168,
+      "learning_rate": 3.4805491143590823e-06,
+      "loss": 0.5238,
+      "step": 763
+    },
+    {
+      "epoch": 3.7268292682926827,
+      "grad_norm": 1.2019354104995728,
+      "learning_rate": 3.4770237425665103e-06,
+      "loss": 0.4918,
+      "step": 764
+    },
+    {
+      "epoch": 3.7317073170731705,
+      "grad_norm": 1.181728720664978,
+      "learning_rate": 3.4734960762264204e-06,
+      "loss": 0.5823,
+      "step": 765
+    },
+    {
+      "epoch": 3.7365853658536583,
+      "grad_norm": 1.0850858688354492,
+      "learning_rate": 3.469966123623563e-06,
+      "loss": 0.5719,
+      "step": 766
+    },
+    {
+      "epoch": 3.741463414634146,
+      "grad_norm": 1.1767489910125732,
+      "learning_rate": 3.46643389304806e-06,
+      "loss": 0.4286,
+      "step": 767
+    },
+    {
+      "epoch": 3.746341463414634,
+      "grad_norm": 1.3543293476104736,
+      "learning_rate": 3.4628993927953786e-06,
+      "loss": 0.9217,
+      "step": 768
+    },
+    {
+      "epoch": 3.7512195121951217,
+      "grad_norm": 0.9874538779258728,
+      "learning_rate": 3.45936263116632e-06,
+      "loss": 0.2432,
+      "step": 769
+    },
+    {
+      "epoch": 3.7560975609756095,
+      "grad_norm": 0.9222186207771301,
+      "learning_rate": 3.4558236164669957e-06,
+      "loss": 0.3024,
+      "step": 770
+    },
+    {
+      "epoch": 3.7609756097560973,
+      "grad_norm": 1.1794410943984985,
+      "learning_rate": 3.4522823570088073e-06,
+      "loss": 0.2452,
+      "step": 771
+    },
+    {
+      "epoch": 3.765853658536585,
+      "grad_norm": 1.1791772842407227,
+      "learning_rate": 3.4487388611084295e-06,
+      "loss": 0.4317,
+      "step": 772
+    },
+    {
+      "epoch": 3.770731707317073,
+      "grad_norm": 1.0632729530334473,
+      "learning_rate": 3.445193137087788e-06,
+      "loss": 0.2299,
+      "step": 773
+    },
+    {
+      "epoch": 3.7756097560975608,
+      "grad_norm": 1.0744452476501465,
+      "learning_rate": 3.4416451932740424e-06,
+      "loss": 0.4244,
+      "step": 774
+    },
+    {
+      "epoch": 3.7804878048780486,
+      "grad_norm": 1.2756675481796265,
+      "learning_rate": 3.4380950379995652e-06,
+      "loss": 0.5733,
+      "step": 775
+    },
+    {
+      "epoch": 3.7853658536585364,
+      "grad_norm": 1.0223393440246582,
+      "learning_rate": 3.434542679601922e-06,
+      "loss": 0.4312,
+      "step": 776
+    },
+    {
+      "epoch": 3.790243902439024,
+      "grad_norm": 1.2650487422943115,
+      "learning_rate": 3.4309881264238538e-06,
+      "loss": 0.1948,
+      "step": 777
+    },
+    {
+      "epoch": 3.795121951219512,
+      "grad_norm": 0.9180625677108765,
+      "learning_rate": 3.4274313868132547e-06,
+      "loss": 0.2998,
+      "step": 778
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.2305192947387695,
+      "learning_rate": 3.4238724691231534e-06,
+      "loss": 0.3367,
+      "step": 779
+    },
+    {
+      "epoch": 3.8048780487804876,
+      "grad_norm": 1.24634850025177,
+      "learning_rate": 3.4203113817116955e-06,
+      "loss": 0.5361,
+      "step": 780
+    },
+    {
+      "epoch": 3.8097560975609754,
+      "grad_norm": 1.145792007446289,
+      "learning_rate": 3.4167481329421204e-06,
+      "loss": 0.308,
+      "step": 781
+    },
+    {
+      "epoch": 3.8146341463414632,
+      "grad_norm": 1.2757189273834229,
+      "learning_rate": 3.4131827311827447e-06,
+      "loss": 0.4851,
+      "step": 782
+    },
+    {
+      "epoch": 3.819512195121951,
+      "grad_norm": 0.9120977520942688,
+      "learning_rate": 3.4096151848069416e-06,
+      "loss": 0.2305,
+      "step": 783
+    },
+    {
+      "epoch": 3.824390243902439,
+      "grad_norm": 1.0253443717956543,
+      "learning_rate": 3.4060455021931195e-06,
+      "loss": 0.3534,
+      "step": 784
+    },
+    {
+      "epoch": 3.8292682926829267,
+      "grad_norm": 1.1518603563308716,
+      "learning_rate": 3.402473691724704e-06,
+      "loss": 0.3636,
+      "step": 785
+    },
+    {
+      "epoch": 3.8341463414634145,
+      "grad_norm": 0.8341196775436401,
+      "learning_rate": 3.39889976179012e-06,
+      "loss": 0.2973,
+      "step": 786
+    },
+    {
+      "epoch": 3.8390243902439023,
+      "grad_norm": 1.025421142578125,
+      "learning_rate": 3.3953237207827673e-06,
+      "loss": 0.4514,
+      "step": 787
+    },
+    {
+      "epoch": 3.84390243902439,
+      "grad_norm": 1.1664432287216187,
+      "learning_rate": 3.391745577101005e-06,
+      "loss": 0.7057,
+      "step": 788
+    },
+    {
+      "epoch": 3.848780487804878,
+      "grad_norm": 1.2613505125045776,
+      "learning_rate": 3.3881653391481306e-06,
+      "loss": 0.382,
+      "step": 789
+    },
+    {
+      "epoch": 3.8536585365853657,
+      "grad_norm": 1.3664671182632446,
+      "learning_rate": 3.384583015332359e-06,
+      "loss": 0.4315,
+      "step": 790
+    },
+    {
+      "epoch": 3.8585365853658535,
+      "grad_norm": 1.0066736936569214,
+      "learning_rate": 3.380998614066805e-06,
+      "loss": 0.2811,
+      "step": 791
+    },
+    {
+      "epoch": 3.8634146341463413,
+      "grad_norm": 1.174258828163147,
+      "learning_rate": 3.3774121437694606e-06,
+      "loss": 0.4093,
+      "step": 792
+    },
+    {
+      "epoch": 3.868292682926829,
+      "grad_norm": 1.1472364664077759,
+      "learning_rate": 3.3738236128631786e-06,
+      "loss": 0.2635,
+      "step": 793
+    },
+    {
+      "epoch": 3.873170731707317,
+      "grad_norm": 1.074864387512207,
+      "learning_rate": 3.3702330297756503e-06,
+      "loss": 0.4941,
+      "step": 794
+    },
+    {
+      "epoch": 3.8780487804878048,
+      "grad_norm": 0.977851390838623,
+      "learning_rate": 3.366640402939387e-06,
+      "loss": 0.1633,
+      "step": 795
+    },
+    {
+      "epoch": 3.8829268292682926,
+      "grad_norm": 1.812105655670166,
+      "learning_rate": 3.363045740791698e-06,
+      "loss": 0.3944,
+      "step": 796
+    },
+    {
+      "epoch": 3.8878048780487804,
+      "grad_norm": 1.1032544374465942,
+      "learning_rate": 3.3594490517746774e-06,
+      "loss": 0.3298,
+      "step": 797
+    },
+    {
+      "epoch": 3.892682926829268,
+      "grad_norm": 1.1241471767425537,
+      "learning_rate": 3.3558503443351733e-06,
+      "loss": 0.4501,
+      "step": 798
+    },
+    {
+      "epoch": 3.897560975609756,
+      "grad_norm": 0.9252558350563049,
+      "learning_rate": 3.352249626924777e-06,
+      "loss": 0.3447,
+      "step": 799
+    },
+    {
+      "epoch": 3.902439024390244,
+      "grad_norm": 1.4837175607681274,
+      "learning_rate": 3.348646907999801e-06,
+      "loss": 0.8945,
+      "step": 800
+    },
+    {
+      "epoch": 3.9073170731707316,
+      "grad_norm": 1.037489891052246,
+      "learning_rate": 3.345042196021257e-06,
+      "loss": 0.4746,
+      "step": 801
+    },
+    {
+      "epoch": 3.9121951219512194,
+      "grad_norm": 1.2995530366897583,
+      "learning_rate": 3.3414354994548385e-06,
+      "loss": 0.7052,
+      "step": 802
+    },
+    {
+      "epoch": 3.9170731707317072,
+      "grad_norm": 1.1786149740219116,
+      "learning_rate": 3.337826826770898e-06,
+      "loss": 0.4128,
+      "step": 803
+    },
+    {
+      "epoch": 3.921951219512195,
+      "grad_norm": 1.0147621631622314,
+      "learning_rate": 3.3342161864444312e-06,
+      "loss": 0.2977,
+      "step": 804
+    },
+    {
+      "epoch": 3.926829268292683,
+      "grad_norm": 0.9299584031105042,
+      "learning_rate": 3.3306035869550534e-06,
+      "loss": 0.2618,
+      "step": 805
+    },
+    {
+      "epoch": 3.9317073170731707,
+      "grad_norm": 1.649917483329773,
+      "learning_rate": 3.326989036786981e-06,
+      "loss": 0.2166,
+      "step": 806
+    },
+    {
+      "epoch": 3.9365853658536585,
+      "grad_norm": 1.2799922227859497,
+      "learning_rate": 3.3233725444290126e-06,
+      "loss": 0.4079,
+      "step": 807
+    },
+    {
+      "epoch": 3.9414634146341463,
+      "grad_norm": 1.1859514713287354,
+      "learning_rate": 3.3197541183745065e-06,
+      "loss": 0.564,
+      "step": 808
+    },
+    {
+      "epoch": 3.946341463414634,
+      "grad_norm": 0.9781062006950378,
+      "learning_rate": 3.3161337671213634e-06,
+      "loss": 0.3557,
+      "step": 809
+    },
+    {
+      "epoch": 3.951219512195122,
+      "grad_norm": 1.1040654182434082,
+      "learning_rate": 3.312511499172006e-06,
+      "loss": 0.2725,
+      "step": 810
+    },
+    {
+      "epoch": 3.9560975609756097,
+      "grad_norm": 1.0453355312347412,
+      "learning_rate": 3.3088873230333562e-06,
+      "loss": 0.4811,
+      "step": 811
+    },
+    {
+      "epoch": 3.9609756097560975,
+      "grad_norm": 1.0773485898971558,
+      "learning_rate": 3.3052612472168193e-06,
+      "loss": 0.2821,
+      "step": 812
+    },
+    {
+      "epoch": 3.9658536585365853,
+      "grad_norm": 0.9467987418174744,
+      "learning_rate": 3.3016332802382618e-06,
+      "loss": 0.3999,
+      "step": 813
+    },
+    {
+      "epoch": 3.970731707317073,
+      "grad_norm": 1.0838000774383545,
+      "learning_rate": 3.2980034306179897e-06,
+      "loss": 0.3441,
+      "step": 814
+    },
+    {
+      "epoch": 3.975609756097561,
+      "grad_norm": 1.0209132432937622,
+      "learning_rate": 3.294371706880733e-06,
+      "loss": 0.4488,
+      "step": 815
+    },
+    {
+      "epoch": 3.9804878048780488,
+      "grad_norm": 1.0593422651290894,
+      "learning_rate": 3.290738117555622e-06,
+      "loss": 0.2754,
+      "step": 816
+    },
+    {
+      "epoch": 3.9853658536585366,
+      "grad_norm": 1.1482516527175903,
+      "learning_rate": 3.2871026711761666e-06,
+      "loss": 0.5598,
+      "step": 817
+    },
+    {
+      "epoch": 3.9902439024390244,
+      "grad_norm": 1.153052568435669,
+      "learning_rate": 3.2834653762802414e-06,
+      "loss": 0.3486,
+      "step": 818
+    },
+    {
+      "epoch": 3.995121951219512,
+      "grad_norm": 1.1609927415847778,
+      "learning_rate": 3.2798262414100594e-06,
+      "loss": 0.3004,
+      "step": 819
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.0488629341125488,
+      "learning_rate": 3.2761852751121566e-06,
+      "loss": 0.2938,
+      "step": 820
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2050,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.4198288618029056e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}