{
  "best_metric": 0.723140299320221,
  "best_model_checkpoint": "./kd_results/microsoft/beit-base-patch16-224_alpha0.7_temp5.0_t3/checkpoint-1148",
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 3280,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1524390243902439,
      "grad_norm": 4.127580642700195,
      "learning_rate": 3.8109756097560976e-06,
      "loss": 0.6029,
      "step": 25
    },
    {
      "epoch": 0.3048780487804878,
      "grad_norm": 4.050529956817627,
      "learning_rate": 7.621951219512195e-06,
      "loss": 0.575,
      "step": 50
    },
    {
      "epoch": 0.4573170731707317,
      "grad_norm": 6.282799243927002,
      "learning_rate": 1.1432926829268294e-05,
      "loss": 0.5801,
      "step": 75
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 2.5533695220947266,
      "learning_rate": 1.524390243902439e-05,
      "loss": 0.5695,
      "step": 100
    },
    {
      "epoch": 0.7621951219512195,
      "grad_norm": 2.692903518676758,
      "learning_rate": 1.9054878048780488e-05,
      "loss": 0.5694,
      "step": 125
    },
    {
      "epoch": 0.9146341463414634,
      "grad_norm": 3.9680466651916504,
      "learning_rate": 2.286585365853659e-05,
      "loss": 0.5663,
      "step": 150
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.25488069414316705,
      "eval_loss": 1.4229953289031982,
      "eval_runtime": 38.8566,
      "eval_samples_per_second": 23.728,
      "eval_steps_per_second": 0.746,
      "step": 164
    },
    {
      "epoch": 1.0670731707317074,
      "grad_norm": 2.8334760665893555,
      "learning_rate": 2.6676829268292686e-05,
      "loss": 0.5725,
      "step": 175
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 4.468701362609863,
      "learning_rate": 3.048780487804878e-05,
      "loss": 0.5711,
      "step": 200
    },
    {
      "epoch": 1.3719512195121952,
      "grad_norm": 5.517786979675293,
      "learning_rate": 3.429878048780488e-05,
      "loss": 0.5559,
      "step": 225
    },
    {
      "epoch": 1.524390243902439,
      "grad_norm": 3.624746084213257,
      "learning_rate": 3.8109756097560976e-05,
      "loss": 0.5653,
      "step": 250
    },
    {
      "epoch": 1.6768292682926829,
      "grad_norm": 2.8956801891326904,
      "learning_rate": 4.1920731707317077e-05,
      "loss": 0.5615,
      "step": 275
    },
    {
      "epoch": 1.8292682926829267,
      "grad_norm": 2.854147434234619,
      "learning_rate": 4.573170731707318e-05,
      "loss": 0.5409,
      "step": 300
    },
    {
      "epoch": 1.9817073170731707,
      "grad_norm": 2.528155565261841,
      "learning_rate": 4.954268292682927e-05,
      "loss": 0.5186,
      "step": 325
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.40238611713665945,
      "eval_loss": 1.2953463792800903,
      "eval_runtime": 38.6956,
      "eval_samples_per_second": 23.827,
      "eval_steps_per_second": 0.749,
      "step": 328
    },
    {
      "epoch": 2.1341463414634148,
      "grad_norm": 6.945356845855713,
      "learning_rate": 4.962737127371274e-05,
      "loss": 0.4894,
      "step": 350
    },
    {
      "epoch": 2.2865853658536586,
      "grad_norm": 4.367215633392334,
      "learning_rate": 4.92039295392954e-05,
      "loss": 0.4628,
      "step": 375
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 4.459686279296875,
      "learning_rate": 4.878048780487805e-05,
      "loss": 0.4369,
      "step": 400
    },
    {
      "epoch": 2.591463414634146,
      "grad_norm": 4.250616550445557,
      "learning_rate": 4.8357046070460704e-05,
      "loss": 0.4069,
      "step": 425
    },
    {
      "epoch": 2.7439024390243905,
      "grad_norm": 3.471693992614746,
      "learning_rate": 4.793360433604336e-05,
      "loss": 0.3965,
      "step": 450
    },
    {
      "epoch": 2.8963414634146343,
      "grad_norm": 2.133843421936035,
      "learning_rate": 4.7510162601626016e-05,
      "loss": 0.3841,
      "step": 475
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7288503253796096,
      "eval_loss": 0.8618374466896057,
      "eval_runtime": 39.7272,
      "eval_samples_per_second": 23.208,
      "eval_steps_per_second": 0.73,
      "step": 492
    },
    {
      "epoch": 3.048780487804878,
      "grad_norm": 2.977363109588623,
      "learning_rate": 4.7086720867208675e-05,
      "loss": 0.3685,
      "step": 500
    },
    {
      "epoch": 3.201219512195122,
      "grad_norm": 1.6001653671264648,
      "learning_rate": 4.6663279132791334e-05,
      "loss": 0.3273,
      "step": 525
    },
    {
      "epoch": 3.3536585365853657,
      "grad_norm": 1.9206469058990479,
      "learning_rate": 4.6239837398373986e-05,
      "loss": 0.3354,
      "step": 550
    },
    {
      "epoch": 3.5060975609756095,
      "grad_norm": 2.6473922729492188,
      "learning_rate": 4.581639566395664e-05,
      "loss": 0.3409,
      "step": 575
    },
    {
      "epoch": 3.658536585365854,
      "grad_norm": 2.05600643157959,
      "learning_rate": 4.53929539295393e-05,
      "loss": 0.3191,
      "step": 600
    },
    {
      "epoch": 3.8109756097560976,
      "grad_norm": 2.0458006858825684,
      "learning_rate": 4.496951219512196e-05,
      "loss": 0.3319,
      "step": 625
    },
    {
      "epoch": 3.9634146341463414,
      "grad_norm": 3.3740410804748535,
      "learning_rate": 4.454607046070461e-05,
      "loss": 0.3209,
      "step": 650
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7505422993492408,
      "eval_loss": 0.8258793354034424,
      "eval_runtime": 41.063,
      "eval_samples_per_second": 22.453,
      "eval_steps_per_second": 0.706,
      "step": 656
    },
    {
      "epoch": 4.115853658536586,
      "grad_norm": 1.513602375984192,
      "learning_rate": 4.412262872628726e-05,
      "loss": 0.306,
      "step": 675
    },
    {
      "epoch": 4.2682926829268295,
      "grad_norm": 1.8460205793380737,
      "learning_rate": 4.369918699186992e-05,
      "loss": 0.2847,
      "step": 700
    },
    {
      "epoch": 4.420731707317073,
      "grad_norm": 1.807004451751709,
      "learning_rate": 4.3275745257452574e-05,
      "loss": 0.2865,
      "step": 725
    },
    {
      "epoch": 4.573170731707317,
      "grad_norm": 1.3295806646347046,
      "learning_rate": 4.285230352303523e-05,
      "loss": 0.2857,
      "step": 750
    },
    {
      "epoch": 4.725609756097561,
      "grad_norm": 1.496420979499817,
      "learning_rate": 4.242886178861789e-05,
      "loss": 0.273,
      "step": 775
    },
    {
      "epoch": 4.878048780487805,
      "grad_norm": 2.111053228378296,
      "learning_rate": 4.2005420054200545e-05,
      "loss": 0.2829,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8036876355748374,
      "eval_loss": 0.7343636155128479,
      "eval_runtime": 41.0787,
      "eval_samples_per_second": 22.445,
      "eval_steps_per_second": 0.706,
      "step": 820
    },
    {
      "epoch": 5.030487804878049,
      "grad_norm": 1.6567422151565552,
      "learning_rate": 4.15819783197832e-05,
      "loss": 0.2783,
      "step": 825
    },
    {
      "epoch": 5.182926829268292,
      "grad_norm": 1.4946651458740234,
      "learning_rate": 4.1158536585365856e-05,
      "loss": 0.2609,
      "step": 850
    },
    {
      "epoch": 5.335365853658536,
      "grad_norm": 1.4999092817306519,
      "learning_rate": 4.0735094850948516e-05,
      "loss": 0.2583,
      "step": 875
    },
    {
      "epoch": 5.487804878048781,
      "grad_norm": 1.725660800933838,
      "learning_rate": 4.031165311653117e-05,
      "loss": 0.2703,
      "step": 900
    },
    {
      "epoch": 5.640243902439025,
      "grad_norm": 1.6573207378387451,
      "learning_rate": 3.988821138211382e-05,
      "loss": 0.2612,
      "step": 925
    },
    {
      "epoch": 5.7926829268292686,
      "grad_norm": 1.7018101215362549,
      "learning_rate": 3.946476964769648e-05,
      "loss": 0.2605,
      "step": 950
    },
    {
      "epoch": 5.945121951219512,
      "grad_norm": 1.4134397506713867,
      "learning_rate": 3.904132791327913e-05,
      "loss": 0.2556,
      "step": 975
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7939262472885033,
      "eval_loss": 0.7868198156356812,
      "eval_runtime": 40.8623,
      "eval_samples_per_second": 22.564,
      "eval_steps_per_second": 0.71,
      "step": 984
    },
    {
      "epoch": 6.097560975609756,
      "grad_norm": 1.5495966672897339,
      "learning_rate": 3.861788617886179e-05,
      "loss": 0.2553,
      "step": 1000
    },
    {
      "epoch": 6.25,
      "grad_norm": 2.302386522293091,
      "learning_rate": 3.8194444444444444e-05,
      "loss": 0.255,
      "step": 1025
    },
    {
      "epoch": 6.402439024390244,
      "grad_norm": 1.0030018091201782,
      "learning_rate": 3.77710027100271e-05,
      "loss": 0.2477,
      "step": 1050
    },
    {
      "epoch": 6.554878048780488,
      "grad_norm": 0.8752070665359497,
      "learning_rate": 3.7347560975609755e-05,
      "loss": 0.2557,
      "step": 1075
    },
    {
      "epoch": 6.7073170731707314,
      "grad_norm": 0.8090361952781677,
      "learning_rate": 3.6924119241192415e-05,
      "loss": 0.2492,
      "step": 1100
    },
    {
      "epoch": 6.859756097560975,
      "grad_norm": 1.297150731086731,
      "learning_rate": 3.6500677506775074e-05,
      "loss": 0.2507,
      "step": 1125
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8080260303687635,
      "eval_loss": 0.723140299320221,
      "eval_runtime": 41.2925,
      "eval_samples_per_second": 22.329,
      "eval_steps_per_second": 0.702,
      "step": 1148
    },
    {
      "epoch": 7.012195121951219,
      "grad_norm": 1.1971936225891113,
      "learning_rate": 3.6077235772357726e-05,
      "loss": 0.2419,
      "step": 1150
    },
    {
      "epoch": 7.164634146341464,
      "grad_norm": 1.0722155570983887,
      "learning_rate": 3.565379403794038e-05,
      "loss": 0.2473,
      "step": 1175
    },
    {
      "epoch": 7.317073170731708,
      "grad_norm": 1.3244762420654297,
      "learning_rate": 3.523035230352303e-05,
      "loss": 0.2438,
      "step": 1200
    },
    {
      "epoch": 7.469512195121951,
      "grad_norm": 1.1973907947540283,
      "learning_rate": 3.48069105691057e-05,
      "loss": 0.2396,
      "step": 1225
    },
    {
      "epoch": 7.621951219512195,
      "grad_norm": 0.8479284644126892,
      "learning_rate": 3.438346883468835e-05,
      "loss": 0.2393,
      "step": 1250
    },
    {
      "epoch": 7.774390243902439,
      "grad_norm": 1.0094653367996216,
      "learning_rate": 3.3960027100271e-05,
      "loss": 0.2496,
      "step": 1275
    },
    {
      "epoch": 7.926829268292683,
      "grad_norm": 1.2563225030899048,
      "learning_rate": 3.353658536585366e-05,
      "loss": 0.2438,
      "step": 1300
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.7841648590021691,
      "eval_loss": 0.763973593711853,
      "eval_runtime": 40.7612,
      "eval_samples_per_second": 22.62,
      "eval_steps_per_second": 0.711,
      "step": 1312
    },
    {
      "epoch": 8.079268292682928,
      "grad_norm": 1.247743844985962,
      "learning_rate": 3.3113143631436314e-05,
      "loss": 0.2382,
      "step": 1325
    },
    {
      "epoch": 8.231707317073171,
      "grad_norm": 1.2395153045654297,
      "learning_rate": 3.268970189701897e-05,
      "loss": 0.2335,
      "step": 1350
    },
    {
      "epoch": 8.384146341463415,
      "grad_norm": 0.90326988697052,
      "learning_rate": 3.226626016260163e-05,
      "loss": 0.2478,
      "step": 1375
    },
    {
      "epoch": 8.536585365853659,
      "grad_norm": 1.507994294166565,
      "learning_rate": 3.1842818428184285e-05,
      "loss": 0.2356,
      "step": 1400
    },
    {
      "epoch": 8.689024390243903,
      "grad_norm": 0.66690593957901,
      "learning_rate": 3.141937669376694e-05,
      "loss": 0.2363,
      "step": 1425
    },
    {
      "epoch": 8.841463414634147,
      "grad_norm": 1.3858404159545898,
      "learning_rate": 3.0995934959349596e-05,
      "loss": 0.2358,
      "step": 1450
    },
    {
      "epoch": 8.99390243902439,
      "grad_norm": 1.3550246953964233,
      "learning_rate": 3.0572493224932255e-05,
      "loss": 0.2438,
      "step": 1475
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8047722342733189,
      "eval_loss": 0.7505329251289368,
      "eval_runtime": 41.101,
      "eval_samples_per_second": 22.433,
      "eval_steps_per_second": 0.706,
      "step": 1476
    },
    {
      "epoch": 9.146341463414634,
      "grad_norm": 0.902326762676239,
      "learning_rate": 3.0149051490514908e-05,
      "loss": 0.2309,
      "step": 1500
    },
    {
      "epoch": 9.298780487804878,
      "grad_norm": 1.2165980339050293,
      "learning_rate": 2.972560975609756e-05,
      "loss": 0.2399,
      "step": 1525
    },
    {
      "epoch": 9.451219512195122,
      "grad_norm": 1.1843523979187012,
      "learning_rate": 2.930216802168022e-05,
      "loss": 0.2353,
      "step": 1550
    },
    {
      "epoch": 9.603658536585366,
      "grad_norm": 1.4470425844192505,
      "learning_rate": 2.8878726287262875e-05,
      "loss": 0.2355,
      "step": 1575
    },
    {
      "epoch": 9.75609756097561,
      "grad_norm": 1.03018057346344,
      "learning_rate": 2.8455284552845528e-05,
      "loss": 0.2358,
      "step": 1600
    },
    {
      "epoch": 9.908536585365853,
      "grad_norm": 1.19709050655365,
      "learning_rate": 2.8031842818428184e-05,
      "loss": 0.2297,
      "step": 1625
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8177874186550976,
      "eval_loss": 0.7407037615776062,
      "eval_runtime": 40.6491,
      "eval_samples_per_second": 22.682,
      "eval_steps_per_second": 0.713,
      "step": 1640
    },
    {
      "epoch": 10.060975609756097,
      "grad_norm": 0.9163194894790649,
      "learning_rate": 2.7608401084010843e-05,
      "loss": 0.2319,
      "step": 1650
    },
    {
      "epoch": 10.213414634146341,
      "grad_norm": 0.8648427724838257,
      "learning_rate": 2.71849593495935e-05,
      "loss": 0.2322,
      "step": 1675
    },
    {
      "epoch": 10.365853658536585,
      "grad_norm": 1.0313268899917603,
      "learning_rate": 2.676151761517615e-05,
      "loss": 0.2347,
      "step": 1700
    },
    {
      "epoch": 10.518292682926829,
      "grad_norm": 1.6096175909042358,
      "learning_rate": 2.633807588075881e-05,
      "loss": 0.232,
      "step": 1725
    },
    {
      "epoch": 10.670731707317072,
      "grad_norm": 0.7083010077476501,
      "learning_rate": 2.5914634146341466e-05,
      "loss": 0.2285,
      "step": 1750
    },
    {
      "epoch": 10.823170731707316,
      "grad_norm": 1.2091063261032104,
      "learning_rate": 2.549119241192412e-05,
      "loss": 0.2328,
      "step": 1775
    },
    {
      "epoch": 10.975609756097562,
      "grad_norm": 0.7183260917663574,
      "learning_rate": 2.5067750677506774e-05,
      "loss": 0.227,
      "step": 1800
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.7993492407809111,
      "eval_loss": 0.7733893990516663,
      "eval_runtime": 40.7015,
      "eval_samples_per_second": 22.653,
      "eval_steps_per_second": 0.713,
      "step": 1804
    },
    {
      "epoch": 11.128048780487806,
      "grad_norm": 1.1267808675765991,
      "learning_rate": 2.464430894308943e-05,
      "loss": 0.2325,
      "step": 1825
    },
    {
      "epoch": 11.28048780487805,
      "grad_norm": 0.41289207339286804,
      "learning_rate": 2.422086720867209e-05,
      "loss": 0.2247,
      "step": 1850
    },
    {
      "epoch": 11.432926829268293,
      "grad_norm": 0.7660936713218689,
      "learning_rate": 2.3797425474254745e-05,
      "loss": 0.2258,
      "step": 1875
    },
    {
      "epoch": 11.585365853658537,
      "grad_norm": 0.826447606086731,
      "learning_rate": 2.3373983739837398e-05,
      "loss": 0.225,
      "step": 1900
    },
    {
      "epoch": 11.737804878048781,
      "grad_norm": 1.080289363861084,
      "learning_rate": 2.2950542005420057e-05,
      "loss": 0.2373,
      "step": 1925
    },
    {
      "epoch": 11.890243902439025,
      "grad_norm": 0.5227652788162231,
      "learning_rate": 2.252710027100271e-05,
      "loss": 0.2303,
      "step": 1950
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7950108459869848,
      "eval_loss": 0.7462745904922485,
      "eval_runtime": 40.7303,
      "eval_samples_per_second": 22.637,
      "eval_steps_per_second": 0.712,
      "step": 1968
    },
    {
      "epoch": 12.042682926829269,
      "grad_norm": 0.6277122497558594,
      "learning_rate": 2.210365853658537e-05,
      "loss": 0.2272,
      "step": 1975
    },
    {
      "epoch": 12.195121951219512,
      "grad_norm": 0.5290517210960388,
      "learning_rate": 2.1680216802168024e-05,
      "loss": 0.226,
      "step": 2000
    },
    {
      "epoch": 12.347560975609756,
      "grad_norm": 0.670170247554779,
      "learning_rate": 2.1256775067750677e-05,
      "loss": 0.232,
      "step": 2025
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.5019763708114624,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.2265,
      "step": 2050
    },
    {
      "epoch": 12.652439024390244,
      "grad_norm": 1.158779501914978,
      "learning_rate": 2.040989159891599e-05,
      "loss": 0.2296,
      "step": 2075
    },
    {
      "epoch": 12.804878048780488,
      "grad_norm": 0.9129577875137329,
      "learning_rate": 1.9986449864498648e-05,
      "loss": 0.2234,
      "step": 2100
    },
    {
      "epoch": 12.957317073170731,
      "grad_norm": 0.9128942489624023,
      "learning_rate": 1.95630081300813e-05,
      "loss": 0.2221,
      "step": 2125
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.7928416485900217,
      "eval_loss": 0.7576633095741272,
      "eval_runtime": 41.4885,
      "eval_samples_per_second": 22.223,
      "eval_steps_per_second": 0.699,
      "step": 2132
    },
    {
      "epoch": 13.109756097560975,
      "grad_norm": 0.34105148911476135,
      "learning_rate": 1.9139566395663956e-05,
      "loss": 0.2273,
      "step": 2150
    },
    {
      "epoch": 13.262195121951219,
      "grad_norm": 0.5225465297698975,
      "learning_rate": 1.8716124661246615e-05,
      "loss": 0.2302,
      "step": 2175
    },
    {
      "epoch": 13.414634146341463,
      "grad_norm": 0.529146134853363,
      "learning_rate": 1.8292682926829268e-05,
      "loss": 0.2242,
      "step": 2200
    },
    {
      "epoch": 13.567073170731707,
      "grad_norm": 0.44471606612205505,
      "learning_rate": 1.7869241192411927e-05,
      "loss": 0.2203,
      "step": 2225
    },
    {
      "epoch": 13.71951219512195,
      "grad_norm": 0.9251234531402588,
      "learning_rate": 1.744579945799458e-05,
      "loss": 0.2205,
      "step": 2250
    },
    {
      "epoch": 13.871951219512194,
      "grad_norm": 0.7494460940361023,
      "learning_rate": 1.702235772357724e-05,
      "loss": 0.2256,
      "step": 2275
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8123644251626898,
      "eval_loss": 0.7402899861335754,
      "eval_runtime": 40.6662,
      "eval_samples_per_second": 22.672,
      "eval_steps_per_second": 0.713,
      "step": 2296
    },
    {
      "epoch": 14.024390243902438,
      "grad_norm": 0.3467264473438263,
      "learning_rate": 1.659891598915989e-05,
      "loss": 0.2246,
      "step": 2300
    },
    {
      "epoch": 14.176829268292684,
      "grad_norm": 0.6693180203437805,
      "learning_rate": 1.6175474254742547e-05,
      "loss": 0.227,
      "step": 2325
    },
    {
      "epoch": 14.329268292682928,
      "grad_norm": 0.7419512867927551,
      "learning_rate": 1.5752032520325206e-05,
      "loss": 0.225,
      "step": 2350
    },
    {
      "epoch": 14.481707317073171,
      "grad_norm": 0.31618252396583557,
      "learning_rate": 1.532859078590786e-05,
      "loss": 0.223,
      "step": 2375
    },
    {
      "epoch": 14.634146341463415,
      "grad_norm": 0.7613115906715393,
      "learning_rate": 1.4905149051490516e-05,
      "loss": 0.2212,
      "step": 2400
    },
    {
      "epoch": 14.786585365853659,
      "grad_norm": 0.5370925068855286,
      "learning_rate": 1.448170731707317e-05,
      "loss": 0.221,
      "step": 2425
    },
    {
      "epoch": 14.939024390243903,
      "grad_norm": 0.3548774719238281,
      "learning_rate": 1.4058265582655828e-05,
      "loss": 0.2267,
      "step": 2450
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.8058568329718004,
      "eval_loss": 0.7454843521118164,
      "eval_runtime": 40.7033,
      "eval_samples_per_second": 22.652,
      "eval_steps_per_second": 0.712,
      "step": 2460
    },
    {
      "epoch": 15.091463414634147,
      "grad_norm": 0.6485327482223511,
      "learning_rate": 1.3634823848238485e-05,
      "loss": 0.2148,
      "step": 2475
    },
    {
      "epoch": 15.24390243902439,
      "grad_norm": 0.74171382188797,
      "learning_rate": 1.321138211382114e-05,
      "loss": 0.2303,
      "step": 2500
    },
    {
      "epoch": 15.396341463414634,
      "grad_norm": 0.48891162872314453,
      "learning_rate": 1.2787940379403795e-05,
      "loss": 0.2245,
      "step": 2525
    },
    {
      "epoch": 15.548780487804878,
      "grad_norm": 0.4822827875614166,
      "learning_rate": 1.2364498644986451e-05,
      "loss": 0.2216,
      "step": 2550
    },
    {
      "epoch": 15.701219512195122,
      "grad_norm": 0.25834500789642334,
      "learning_rate": 1.1941056910569107e-05,
      "loss": 0.2199,
      "step": 2575
    },
    {
      "epoch": 15.853658536585366,
      "grad_norm": 0.3766600489616394,
      "learning_rate": 1.1517615176151763e-05,
      "loss": 0.2222,
      "step": 2600
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.806941431670282,
      "eval_loss": 0.7406081557273865,
      "eval_runtime": 39.32,
      "eval_samples_per_second": 23.449,
      "eval_steps_per_second": 0.738,
      "step": 2624
    },
    {
      "epoch": 16.00609756097561,
      "grad_norm": 0.5023967623710632,
      "learning_rate": 1.1094173441734418e-05,
      "loss": 0.2251,
      "step": 2625
    },
    {
      "epoch": 16.158536585365855,
      "grad_norm": 0.39426037669181824,
      "learning_rate": 1.0670731707317074e-05,
      "loss": 0.2213,
      "step": 2650
    },
    {
      "epoch": 16.3109756097561,
      "grad_norm": 0.9000890851020813,
      "learning_rate": 1.0247289972899728e-05,
      "loss": 0.2203,
      "step": 2675
    },
    {
      "epoch": 16.463414634146343,
      "grad_norm": 0.5663197040557861,
      "learning_rate": 9.823848238482384e-06,
      "loss": 0.2177,
      "step": 2700
    },
    {
      "epoch": 16.615853658536587,
      "grad_norm": 0.5573919415473938,
      "learning_rate": 9.400406504065042e-06,
      "loss": 0.2227,
      "step": 2725
    },
    {
      "epoch": 16.76829268292683,
      "grad_norm": 0.3704644441604614,
      "learning_rate": 8.976964769647697e-06,
      "loss": 0.2172,
      "step": 2750
    },
    {
      "epoch": 16.920731707317074,
      "grad_norm": 0.39436420798301697,
      "learning_rate": 8.553523035230353e-06,
      "loss": 0.2268,
      "step": 2775
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.8101952277657267,
      "eval_loss": 0.7434730529785156,
      "eval_runtime": 40.9446,
      "eval_samples_per_second": 22.518,
      "eval_steps_per_second": 0.708,
      "step": 2788
    },
    {
      "epoch": 17.073170731707318,
      "grad_norm": 0.36405354738235474,
      "learning_rate": 8.130081300813009e-06,
      "loss": 0.2231,
      "step": 2800
    },
    {
      "epoch": 17.225609756097562,
      "grad_norm": 0.3161733150482178,
      "learning_rate": 7.706639566395663e-06,
      "loss": 0.2197,
      "step": 2825
    },
    {
      "epoch": 17.378048780487806,
      "grad_norm": 0.6657469272613525,
      "learning_rate": 7.283197831978319e-06,
      "loss": 0.2264,
      "step": 2850
    },
    {
      "epoch": 17.53048780487805,
      "grad_norm": 0.3313043415546417,
      "learning_rate": 6.859756097560977e-06,
      "loss": 0.2261,
      "step": 2875
    },
    {
      "epoch": 17.682926829268293,
      "grad_norm": 0.49379193782806396,
      "learning_rate": 6.4363143631436324e-06,
      "loss": 0.2188,
      "step": 2900
    },
    {
      "epoch": 17.835365853658537,
      "grad_norm": 0.35201114416122437,
      "learning_rate": 6.0128726287262874e-06,
      "loss": 0.215,
      "step": 2925
    },
    {
      "epoch": 17.98780487804878,
      "grad_norm": 0.2961661219596863,
      "learning_rate": 5.589430894308943e-06,
      "loss": 0.2168,
      "step": 2950
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8058568329718004,
      "eval_loss": 0.7400941848754883,
      "eval_runtime": 40.342,
      "eval_samples_per_second": 22.855,
      "eval_steps_per_second": 0.719,
      "step": 2952
    },
    {
      "epoch": 18.140243902439025,
      "grad_norm": 0.2617061138153076,
      "learning_rate": 5.165989159891599e-06,
      "loss": 0.22,
      "step": 2975
    },
    {
      "epoch": 18.29268292682927,
      "grad_norm": 1.1936415433883667,
      "learning_rate": 4.742547425474255e-06,
      "loss": 0.2202,
      "step": 3000
    },
    {
      "epoch": 18.445121951219512,
      "grad_norm": 0.4254779815673828,
      "learning_rate": 4.319105691056911e-06,
      "loss": 0.2177,
      "step": 3025
    },
    {
      "epoch": 18.597560975609756,
      "grad_norm": 0.38696911931037903,
      "learning_rate": 3.8956639566395666e-06,
      "loss": 0.2209,
      "step": 3050
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.3616865277290344,
      "learning_rate": 3.4722222222222224e-06,
      "loss": 0.2174,
      "step": 3075
    },
    {
      "epoch": 18.902439024390244,
      "grad_norm": 0.2995990514755249,
      "learning_rate": 3.0487804878048782e-06,
      "loss": 0.2224,
      "step": 3100
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8047722342733189,
      "eval_loss": 0.7402113676071167,
      "eval_runtime": 40.7076,
      "eval_samples_per_second": 22.649,
      "eval_steps_per_second": 0.712,
      "step": 3116
    },
    {
      "epoch": 19.054878048780488,
      "grad_norm": 0.24594496190547943,
      "learning_rate": 2.625338753387534e-06,
      "loss": 0.2203,
      "step": 3125
    },
    {
      "epoch": 19.20731707317073,
      "grad_norm": 0.3979138731956482,
      "learning_rate": 2.2018970189701894e-06,
      "loss": 0.2263,
      "step": 3150
    },
    {
      "epoch": 19.359756097560975,
      "grad_norm": 0.30416855216026306,
      "learning_rate": 1.7784552845528455e-06,
      "loss": 0.2221,
      "step": 3175
    },
    {
      "epoch": 19.51219512195122,
      "grad_norm": 0.3640141189098358,
      "learning_rate": 1.3550135501355015e-06,
      "loss": 0.2199,
      "step": 3200
    },
    {
      "epoch": 19.664634146341463,
      "grad_norm": 0.3846640884876251,
      "learning_rate": 9.315718157181572e-07,
      "loss": 0.2208,
      "step": 3225
    },
    {
      "epoch": 19.817073170731707,
      "grad_norm": 0.36082571744918823,
      "learning_rate": 5.081300813008131e-07,
      "loss": 0.2161,
      "step": 3250
    },
    {
      "epoch": 19.96951219512195,
      "grad_norm": 0.2782360017299652,
      "learning_rate": 8.468834688346885e-08,
      "loss": 0.217,
      "step": 3275
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8036876355748374,
      "eval_loss": 0.7440109848976135,
      "eval_runtime": 41.9613,
      "eval_samples_per_second": 21.973,
      "eval_steps_per_second": 0.691,
      "step": 3280
    },
    {
      "epoch": 20.0,
      "step": 3280,
      "total_flos": 0.0,
      "train_loss": 0.2817080278585597,
      "train_runtime": 14759.5588,
      "train_samples_per_second": 7.079,
      "train_steps_per_second": 0.222
    }
  ],
  "logging_steps": 25,
  "max_steps": 3280,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}