{
  "best_metric": 0.8562042713165283,
  "best_model_checkpoint": "./kd_results/apple/mobilevit-small_alpha0.5_temp3.0_t3/checkpoint-2788",
  "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": 7.762810707092285,
      "learning_rate": 3.8109756097560976e-06,
      "loss": 0.9874,
      "step": 25
    },
    {
      "epoch": 0.3048780487804878,
      "grad_norm": 2.2041540145874023,
      "learning_rate": 7.621951219512195e-06,
      "loss": 0.8291,
      "step": 50
    },
    {
      "epoch": 0.4573170731707317,
      "grad_norm": 2.862013101577759,
      "learning_rate": 1.1432926829268294e-05,
      "loss": 0.809,
      "step": 75
    },
    {
      "epoch": 0.6097560975609756,
      "grad_norm": 3.8926451206207275,
      "learning_rate": 1.524390243902439e-05,
      "loss": 0.8056,
      "step": 100
    },
    {
      "epoch": 0.7621951219512195,
      "grad_norm": 3.9026598930358887,
      "learning_rate": 1.9054878048780488e-05,
      "loss": 0.8081,
      "step": 125
    },
    {
      "epoch": 0.9146341463414634,
      "grad_norm": 2.8420770168304443,
      "learning_rate": 2.286585365853659e-05,
      "loss": 0.8034,
      "step": 150
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.28633405639913234,
      "eval_loss": 1.3824383020401,
      "eval_runtime": 40.4477,
      "eval_samples_per_second": 22.795,
      "eval_steps_per_second": 0.717,
      "step": 164
    },
    {
      "epoch": 1.0670731707317074,
      "grad_norm": 2.4602479934692383,
      "learning_rate": 2.6676829268292686e-05,
      "loss": 0.8053,
      "step": 175
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 2.140545606613159,
      "learning_rate": 3.048780487804878e-05,
      "loss": 0.7981,
      "step": 200
    },
    {
      "epoch": 1.3719512195121952,
      "grad_norm": 4.539191246032715,
      "learning_rate": 3.429878048780488e-05,
      "loss": 0.795,
      "step": 225
    },
    {
      "epoch": 1.524390243902439,
      "grad_norm": 3.07138729095459,
      "learning_rate": 3.8109756097560976e-05,
      "loss": 0.8012,
      "step": 250
    },
    {
      "epoch": 1.6768292682926829,
      "grad_norm": 2.5887339115142822,
      "learning_rate": 4.1920731707317077e-05,
      "loss": 0.7953,
      "step": 275
    },
    {
      "epoch": 1.8292682926829267,
      "grad_norm": 4.393621444702148,
      "learning_rate": 4.573170731707318e-05,
      "loss": 0.772,
      "step": 300
    },
    {
      "epoch": 1.9817073170731707,
      "grad_norm": 1.1688978672027588,
      "learning_rate": 4.954268292682927e-05,
      "loss": 0.7605,
      "step": 325
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2917570498915401,
      "eval_loss": 1.3373875617980957,
      "eval_runtime": 40.6346,
      "eval_samples_per_second": 22.69,
      "eval_steps_per_second": 0.714,
      "step": 328
    },
    {
      "epoch": 2.1341463414634148,
      "grad_norm": 1.0945930480957031,
      "learning_rate": 4.962737127371274e-05,
      "loss": 0.7616,
      "step": 350
    },
    {
      "epoch": 2.2865853658536586,
      "grad_norm": 2.745100259780884,
      "learning_rate": 4.92039295392954e-05,
      "loss": 0.7438,
      "step": 375
    },
    {
      "epoch": 2.4390243902439024,
      "grad_norm": 5.273860454559326,
      "learning_rate": 4.878048780487805e-05,
      "loss": 0.7124,
      "step": 400
    },
    {
      "epoch": 2.591463414634146,
      "grad_norm": 5.696925640106201,
      "learning_rate": 4.8357046070460704e-05,
      "loss": 0.7182,
      "step": 425
    },
    {
      "epoch": 2.7439024390243905,
      "grad_norm": 5.35153865814209,
      "learning_rate": 4.793360433604336e-05,
      "loss": 0.7061,
      "step": 450
    },
    {
      "epoch": 2.8963414634146343,
      "grad_norm": 4.60999059677124,
      "learning_rate": 4.7510162601626016e-05,
      "loss": 0.6903,
      "step": 475
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5151843817787418,
      "eval_loss": 1.140594720840454,
      "eval_runtime": 40.2587,
      "eval_samples_per_second": 22.902,
      "eval_steps_per_second": 0.72,
      "step": 492
    },
    {
      "epoch": 3.048780487804878,
      "grad_norm": 5.811767578125,
      "learning_rate": 4.7086720867208675e-05,
      "loss": 0.6817,
      "step": 500
    },
    {
      "epoch": 3.201219512195122,
      "grad_norm": 4.599453449249268,
      "learning_rate": 4.6663279132791334e-05,
      "loss": 0.6679,
      "step": 525
    },
    {
      "epoch": 3.3536585365853657,
      "grad_norm": 6.098621368408203,
      "learning_rate": 4.6239837398373986e-05,
      "loss": 0.6385,
      "step": 550
    },
    {
      "epoch": 3.5060975609756095,
      "grad_norm": 6.299458026885986,
      "learning_rate": 4.581639566395664e-05,
      "loss": 0.6364,
      "step": 575
    },
    {
      "epoch": 3.658536585365854,
      "grad_norm": 3.688998222351074,
      "learning_rate": 4.53929539295393e-05,
      "loss": 0.6396,
      "step": 600
    },
    {
      "epoch": 3.8109756097560976,
      "grad_norm": 5.4493727684021,
      "learning_rate": 4.496951219512196e-05,
      "loss": 0.6175,
      "step": 625
    },
    {
      "epoch": 3.9634146341463414,
      "grad_norm": 6.126657485961914,
      "learning_rate": 4.454607046070461e-05,
      "loss": 0.6045,
      "step": 650
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5629067245119306,
      "eval_loss": 1.0571489334106445,
      "eval_runtime": 40.0524,
      "eval_samples_per_second": 23.02,
      "eval_steps_per_second": 0.724,
      "step": 656
    },
    {
      "epoch": 4.115853658536586,
      "grad_norm": 8.532530784606934,
      "learning_rate": 4.412262872628726e-05,
      "loss": 0.6161,
      "step": 675
    },
    {
      "epoch": 4.2682926829268295,
      "grad_norm": 3.7659671306610107,
      "learning_rate": 4.369918699186992e-05,
      "loss": 0.603,
      "step": 700
    },
    {
      "epoch": 4.420731707317073,
      "grad_norm": 2.876420736312866,
      "learning_rate": 4.3275745257452574e-05,
      "loss": 0.5798,
      "step": 725
    },
    {
      "epoch": 4.573170731707317,
      "grad_norm": 7.7819013595581055,
      "learning_rate": 4.285230352303523e-05,
      "loss": 0.5904,
      "step": 750
    },
    {
      "epoch": 4.725609756097561,
      "grad_norm": 5.8853678703308105,
      "learning_rate": 4.242886178861789e-05,
      "loss": 0.573,
      "step": 775
    },
    {
      "epoch": 4.878048780487805,
      "grad_norm": 6.983067989349365,
      "learning_rate": 4.2005420054200545e-05,
      "loss": 0.5779,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5813449023861171,
      "eval_loss": 1.022658348083496,
      "eval_runtime": 40.4951,
      "eval_samples_per_second": 22.768,
      "eval_steps_per_second": 0.716,
      "step": 820
    },
    {
      "epoch": 5.030487804878049,
      "grad_norm": 7.891158103942871,
      "learning_rate": 4.15819783197832e-05,
      "loss": 0.5842,
      "step": 825
    },
    {
      "epoch": 5.182926829268292,
      "grad_norm": 4.526270389556885,
      "learning_rate": 4.1158536585365856e-05,
      "loss": 0.55,
      "step": 850
    },
    {
      "epoch": 5.335365853658536,
      "grad_norm": 3.818694591522217,
      "learning_rate": 4.0735094850948516e-05,
      "loss": 0.5396,
      "step": 875
    },
    {
      "epoch": 5.487804878048781,
      "grad_norm": 2.55497670173645,
      "learning_rate": 4.031165311653117e-05,
      "loss": 0.554,
      "step": 900
    },
    {
      "epoch": 5.640243902439025,
      "grad_norm": 5.239187717437744,
      "learning_rate": 3.988821138211382e-05,
      "loss": 0.5413,
      "step": 925
    },
    {
      "epoch": 5.7926829268292686,
      "grad_norm": 3.5609655380249023,
      "learning_rate": 3.946476964769648e-05,
      "loss": 0.5388,
      "step": 950
    },
    {
      "epoch": 5.945121951219512,
      "grad_norm": 2.538914918899536,
      "learning_rate": 3.904132791327913e-05,
      "loss": 0.5423,
      "step": 975
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6334056399132321,
      "eval_loss": 0.9594108462333679,
      "eval_runtime": 39.996,
      "eval_samples_per_second": 23.052,
      "eval_steps_per_second": 0.725,
      "step": 984
    },
    {
      "epoch": 6.097560975609756,
      "grad_norm": 9.002790451049805,
      "learning_rate": 3.861788617886179e-05,
      "loss": 0.5333,
      "step": 1000
    },
    {
      "epoch": 6.25,
      "grad_norm": 2.795607805252075,
      "learning_rate": 3.8194444444444444e-05,
      "loss": 0.5304,
      "step": 1025
    },
    {
      "epoch": 6.402439024390244,
      "grad_norm": 3.0136711597442627,
      "learning_rate": 3.77710027100271e-05,
      "loss": 0.5338,
      "step": 1050
    },
    {
      "epoch": 6.554878048780488,
      "grad_norm": 4.359175205230713,
      "learning_rate": 3.7347560975609755e-05,
      "loss": 0.5213,
      "step": 1075
    },
    {
      "epoch": 6.7073170731707314,
      "grad_norm": 5.616451740264893,
      "learning_rate": 3.6924119241192415e-05,
      "loss": 0.5211,
      "step": 1100
    },
    {
      "epoch": 6.859756097560975,
      "grad_norm": 3.337740659713745,
      "learning_rate": 3.6500677506775074e-05,
      "loss": 0.5154,
      "step": 1125
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.6518438177874186,
      "eval_loss": 0.9497526288032532,
      "eval_runtime": 40.4853,
      "eval_samples_per_second": 22.774,
      "eval_steps_per_second": 0.716,
      "step": 1148
    },
    {
      "epoch": 7.012195121951219,
      "grad_norm": 3.785557746887207,
      "learning_rate": 3.6077235772357726e-05,
      "loss": 0.5004,
      "step": 1150
    },
    {
      "epoch": 7.164634146341464,
      "grad_norm": 4.928786277770996,
      "learning_rate": 3.565379403794038e-05,
      "loss": 0.5024,
      "step": 1175
    },
    {
      "epoch": 7.317073170731708,
      "grad_norm": 3.090078353881836,
      "learning_rate": 3.523035230352303e-05,
      "loss": 0.4839,
      "step": 1200
    },
    {
      "epoch": 7.469512195121951,
      "grad_norm": 6.933969020843506,
      "learning_rate": 3.48069105691057e-05,
      "loss": 0.4764,
      "step": 1225
    },
    {
      "epoch": 7.621951219512195,
      "grad_norm": 6.955783367156982,
      "learning_rate": 3.438346883468835e-05,
      "loss": 0.5134,
      "step": 1250
    },
    {
      "epoch": 7.774390243902439,
      "grad_norm": 4.848341941833496,
      "learning_rate": 3.3960027100271e-05,
      "loss": 0.5051,
      "step": 1275
    },
    {
      "epoch": 7.926829268292683,
      "grad_norm": 3.339418888092041,
      "learning_rate": 3.353658536585366e-05,
      "loss": 0.4984,
      "step": 1300
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.675704989154013,
      "eval_loss": 0.9110548496246338,
      "eval_runtime": 40.3793,
      "eval_samples_per_second": 22.833,
      "eval_steps_per_second": 0.718,
      "step": 1312
    },
    {
      "epoch": 8.079268292682928,
      "grad_norm": 5.299203395843506,
      "learning_rate": 3.3113143631436314e-05,
      "loss": 0.4745,
      "step": 1325
    },
    {
      "epoch": 8.231707317073171,
      "grad_norm": 5.381677627563477,
      "learning_rate": 3.268970189701897e-05,
      "loss": 0.4696,
      "step": 1350
    },
    {
      "epoch": 8.384146341463415,
      "grad_norm": 6.133314609527588,
      "learning_rate": 3.226626016260163e-05,
      "loss": 0.4876,
      "step": 1375
    },
    {
      "epoch": 8.536585365853659,
      "grad_norm": 3.0064644813537598,
      "learning_rate": 3.1842818428184285e-05,
      "loss": 0.4581,
      "step": 1400
    },
    {
      "epoch": 8.689024390243903,
      "grad_norm": 3.6373887062072754,
      "learning_rate": 3.141937669376694e-05,
      "loss": 0.4813,
      "step": 1425
    },
    {
      "epoch": 8.841463414634147,
      "grad_norm": 2.668551206588745,
      "learning_rate": 3.0995934959349596e-05,
      "loss": 0.4616,
      "step": 1450
    },
    {
      "epoch": 8.99390243902439,
      "grad_norm": 4.62333345413208,
      "learning_rate": 3.0572493224932255e-05,
      "loss": 0.47,
      "step": 1475
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.6518438177874186,
      "eval_loss": 0.9155306816101074,
      "eval_runtime": 39.9567,
      "eval_samples_per_second": 23.075,
      "eval_steps_per_second": 0.726,
      "step": 1476
    },
    {
      "epoch": 9.146341463414634,
      "grad_norm": 4.4853434562683105,
      "learning_rate": 3.0149051490514908e-05,
      "loss": 0.4618,
      "step": 1500
    },
    {
      "epoch": 9.298780487804878,
      "grad_norm": 3.469381332397461,
      "learning_rate": 2.972560975609756e-05,
      "loss": 0.4687,
      "step": 1525
    },
    {
      "epoch": 9.451219512195122,
      "grad_norm": 5.934073448181152,
      "learning_rate": 2.930216802168022e-05,
      "loss": 0.4563,
      "step": 1550
    },
    {
      "epoch": 9.603658536585366,
      "grad_norm": 3.998873472213745,
      "learning_rate": 2.8878726287262875e-05,
      "loss": 0.4502,
      "step": 1575
    },
    {
      "epoch": 9.75609756097561,
      "grad_norm": 5.720112323760986,
      "learning_rate": 2.8455284552845528e-05,
      "loss": 0.4633,
      "step": 1600
    },
    {
      "epoch": 9.908536585365853,
      "grad_norm": 2.881168842315674,
      "learning_rate": 2.8031842818428184e-05,
      "loss": 0.4474,
      "step": 1625
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6518438177874186,
      "eval_loss": 0.9252634644508362,
      "eval_runtime": 39.6703,
      "eval_samples_per_second": 23.242,
      "eval_steps_per_second": 0.731,
      "step": 1640
    },
    {
      "epoch": 10.060975609756097,
      "grad_norm": 4.207843780517578,
      "learning_rate": 2.7608401084010843e-05,
      "loss": 0.4542,
      "step": 1650
    },
    {
      "epoch": 10.213414634146341,
      "grad_norm": 3.44258713722229,
      "learning_rate": 2.71849593495935e-05,
      "loss": 0.4369,
      "step": 1675
    },
    {
      "epoch": 10.365853658536585,
      "grad_norm": 6.224090099334717,
      "learning_rate": 2.676151761517615e-05,
      "loss": 0.4532,
      "step": 1700
    },
    {
      "epoch": 10.518292682926829,
      "grad_norm": 2.638032913208008,
      "learning_rate": 2.633807588075881e-05,
      "loss": 0.4424,
      "step": 1725
    },
    {
      "epoch": 10.670731707317072,
      "grad_norm": 2.48111629486084,
      "learning_rate": 2.5914634146341466e-05,
      "loss": 0.4279,
      "step": 1750
    },
    {
      "epoch": 10.823170731707316,
      "grad_norm": 4.390626907348633,
      "learning_rate": 2.549119241192412e-05,
      "loss": 0.448,
      "step": 1775
    },
    {
      "epoch": 10.975609756097562,
      "grad_norm": 4.277877330780029,
      "learning_rate": 2.5067750677506774e-05,
      "loss": 0.4547,
      "step": 1800
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.6561822125813449,
      "eval_loss": 0.9355393648147583,
      "eval_runtime": 40.5434,
      "eval_samples_per_second": 22.741,
      "eval_steps_per_second": 0.715,
      "step": 1804
    },
    {
      "epoch": 11.128048780487806,
      "grad_norm": 6.439655303955078,
      "learning_rate": 2.464430894308943e-05,
      "loss": 0.4528,
      "step": 1825
    },
    {
      "epoch": 11.28048780487805,
      "grad_norm": 3.8702235221862793,
      "learning_rate": 2.422086720867209e-05,
      "loss": 0.4277,
      "step": 1850
    },
    {
      "epoch": 11.432926829268293,
      "grad_norm": 6.396579265594482,
      "learning_rate": 2.3797425474254745e-05,
      "loss": 0.4298,
      "step": 1875
    },
    {
      "epoch": 11.585365853658537,
      "grad_norm": 6.784517765045166,
      "learning_rate": 2.3373983739837398e-05,
      "loss": 0.4293,
      "step": 1900
    },
    {
      "epoch": 11.737804878048781,
      "grad_norm": 3.7410504817962646,
      "learning_rate": 2.2950542005420057e-05,
      "loss": 0.4379,
      "step": 1925
    },
    {
      "epoch": 11.890243902439025,
      "grad_norm": 6.854469299316406,
      "learning_rate": 2.252710027100271e-05,
      "loss": 0.4432,
      "step": 1950
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.6648590021691974,
      "eval_loss": 0.8717988729476929,
      "eval_runtime": 40.6359,
      "eval_samples_per_second": 22.689,
      "eval_steps_per_second": 0.714,
      "step": 1968
    },
    {
      "epoch": 12.042682926829269,
      "grad_norm": 4.439425468444824,
      "learning_rate": 2.210365853658537e-05,
      "loss": 0.447,
      "step": 1975
    },
    {
      "epoch": 12.195121951219512,
      "grad_norm": 1.7574331760406494,
      "learning_rate": 2.1680216802168024e-05,
      "loss": 0.4109,
      "step": 2000
    },
    {
      "epoch": 12.347560975609756,
      "grad_norm": 5.339428901672363,
      "learning_rate": 2.1256775067750677e-05,
      "loss": 0.4288,
      "step": 2025
    },
    {
      "epoch": 12.5,
      "grad_norm": 2.4824461936950684,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.4298,
      "step": 2050
    },
    {
      "epoch": 12.652439024390244,
      "grad_norm": 2.4411211013793945,
      "learning_rate": 2.040989159891599e-05,
      "loss": 0.4265,
      "step": 2075
    },
    {
      "epoch": 12.804878048780488,
      "grad_norm": 6.373781204223633,
      "learning_rate": 1.9986449864498648e-05,
      "loss": 0.422,
      "step": 2100
    },
    {
      "epoch": 12.957317073170731,
      "grad_norm": 3.202080011367798,
      "learning_rate": 1.95630081300813e-05,
      "loss": 0.4235,
      "step": 2125
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.7039045553145337,
      "eval_loss": 0.8762956857681274,
      "eval_runtime": 40.5234,
      "eval_samples_per_second": 22.752,
      "eval_steps_per_second": 0.716,
      "step": 2132
    },
    {
      "epoch": 13.109756097560975,
      "grad_norm": 3.008108615875244,
      "learning_rate": 1.9139566395663956e-05,
      "loss": 0.424,
      "step": 2150
    },
    {
      "epoch": 13.262195121951219,
      "grad_norm": 3.0317177772521973,
      "learning_rate": 1.8716124661246615e-05,
      "loss": 0.4246,
      "step": 2175
    },
    {
      "epoch": 13.414634146341463,
      "grad_norm": 2.7585787773132324,
      "learning_rate": 1.8292682926829268e-05,
      "loss": 0.4132,
      "step": 2200
    },
    {
      "epoch": 13.567073170731707,
      "grad_norm": 3.0039117336273193,
      "learning_rate": 1.7869241192411927e-05,
      "loss": 0.4076,
      "step": 2225
    },
    {
      "epoch": 13.71951219512195,
      "grad_norm": 5.297887325286865,
      "learning_rate": 1.744579945799458e-05,
      "loss": 0.4,
      "step": 2250
    },
    {
      "epoch": 13.871951219512194,
      "grad_norm": 3.2223308086395264,
      "learning_rate": 1.702235772357724e-05,
      "loss": 0.419,
      "step": 2275
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.6865509761388287,
      "eval_loss": 0.8877863883972168,
      "eval_runtime": 40.0927,
      "eval_samples_per_second": 22.997,
      "eval_steps_per_second": 0.723,
      "step": 2296
    },
    {
      "epoch": 14.024390243902438,
      "grad_norm": 5.251517295837402,
      "learning_rate": 1.659891598915989e-05,
      "loss": 0.4151,
      "step": 2300
    },
    {
      "epoch": 14.176829268292684,
      "grad_norm": 5.841452598571777,
      "learning_rate": 1.6175474254742547e-05,
      "loss": 0.409,
      "step": 2325
    },
    {
      "epoch": 14.329268292682928,
      "grad_norm": 4.45047664642334,
      "learning_rate": 1.5752032520325206e-05,
      "loss": 0.4073,
      "step": 2350
    },
    {
      "epoch": 14.481707317073171,
      "grad_norm": 2.3922412395477295,
      "learning_rate": 1.532859078590786e-05,
      "loss": 0.4017,
      "step": 2375
    },
    {
      "epoch": 14.634146341463415,
      "grad_norm": 4.807758331298828,
      "learning_rate": 1.4905149051490516e-05,
      "loss": 0.4107,
      "step": 2400
    },
    {
      "epoch": 14.786585365853659,
      "grad_norm": 2.9783949851989746,
      "learning_rate": 1.448170731707317e-05,
      "loss": 0.4122,
      "step": 2425
    },
    {
      "epoch": 14.939024390243903,
      "grad_norm": 2.8491086959838867,
      "learning_rate": 1.4058265582655828e-05,
      "loss": 0.4083,
      "step": 2450
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.6898047722342733,
      "eval_loss": 0.886920154094696,
      "eval_runtime": 39.9853,
      "eval_samples_per_second": 23.058,
      "eval_steps_per_second": 0.725,
      "step": 2460
    },
    {
      "epoch": 15.091463414634147,
      "grad_norm": 5.6329665184021,
      "learning_rate": 1.3634823848238485e-05,
      "loss": 0.393,
      "step": 2475
    },
    {
      "epoch": 15.24390243902439,
      "grad_norm": 4.589029312133789,
      "learning_rate": 1.321138211382114e-05,
      "loss": 0.4068,
      "step": 2500
    },
    {
      "epoch": 15.396341463414634,
      "grad_norm": 1.9374419450759888,
      "learning_rate": 1.2787940379403795e-05,
      "loss": 0.4077,
      "step": 2525
    },
    {
      "epoch": 15.548780487804878,
      "grad_norm": 3.12813401222229,
      "learning_rate": 1.2364498644986451e-05,
      "loss": 0.3932,
      "step": 2550
    },
    {
      "epoch": 15.701219512195122,
      "grad_norm": 4.135628700256348,
      "learning_rate": 1.1941056910569107e-05,
      "loss": 0.4048,
      "step": 2575
    },
    {
      "epoch": 15.853658536585366,
      "grad_norm": 2.268238067626953,
      "learning_rate": 1.1517615176151763e-05,
      "loss": 0.4043,
      "step": 2600
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.6984815618221258,
      "eval_loss": 0.8697348237037659,
      "eval_runtime": 40.2898,
      "eval_samples_per_second": 22.884,
      "eval_steps_per_second": 0.72,
      "step": 2624
    },
    {
      "epoch": 16.00609756097561,
      "grad_norm": 3.1583592891693115,
      "learning_rate": 1.1094173441734418e-05,
      "loss": 0.4083,
      "step": 2625
    },
    {
      "epoch": 16.158536585365855,
      "grad_norm": 3.8587262630462646,
      "learning_rate": 1.0670731707317074e-05,
      "loss": 0.3955,
      "step": 2650
    },
    {
      "epoch": 16.3109756097561,
      "grad_norm": 6.293887615203857,
      "learning_rate": 1.0247289972899728e-05,
      "loss": 0.3932,
      "step": 2675
    },
    {
      "epoch": 16.463414634146343,
      "grad_norm": 4.263759613037109,
      "learning_rate": 9.823848238482384e-06,
      "loss": 0.3929,
      "step": 2700
    },
    {
      "epoch": 16.615853658536587,
      "grad_norm": 2.30064058303833,
      "learning_rate": 9.400406504065042e-06,
      "loss": 0.4003,
      "step": 2725
    },
    {
      "epoch": 16.76829268292683,
      "grad_norm": 4.408405303955078,
      "learning_rate": 8.976964769647697e-06,
      "loss": 0.3919,
      "step": 2750
    },
    {
      "epoch": 16.920731707317074,
      "grad_norm": 4.191670894622803,
      "learning_rate": 8.553523035230353e-06,
      "loss": 0.4061,
      "step": 2775
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.7071583514099783,
      "eval_loss": 0.8562042713165283,
      "eval_runtime": 40.0276,
      "eval_samples_per_second": 23.034,
      "eval_steps_per_second": 0.724,
      "step": 2788
    },
    {
      "epoch": 17.073170731707318,
      "grad_norm": 2.7028963565826416,
      "learning_rate": 8.130081300813009e-06,
      "loss": 0.3994,
      "step": 2800
    },
    {
      "epoch": 17.225609756097562,
      "grad_norm": 2.178621530532837,
      "learning_rate": 7.706639566395663e-06,
      "loss": 0.3833,
      "step": 2825
    },
    {
      "epoch": 17.378048780487806,
      "grad_norm": 1.921392560005188,
      "learning_rate": 7.283197831978319e-06,
      "loss": 0.4023,
      "step": 2850
    },
    {
      "epoch": 17.53048780487805,
      "grad_norm": 3.9117579460144043,
      "learning_rate": 6.859756097560977e-06,
      "loss": 0.3991,
      "step": 2875
    },
    {
      "epoch": 17.682926829268293,
      "grad_norm": 1.6043190956115723,
      "learning_rate": 6.4363143631436324e-06,
      "loss": 0.3924,
      "step": 2900
    },
    {
      "epoch": 17.835365853658537,
      "grad_norm": 2.8831610679626465,
      "learning_rate": 6.0128726287262874e-06,
      "loss": 0.3837,
      "step": 2925
    },
    {
      "epoch": 17.98780487804878,
      "grad_norm": 2.4830398559570312,
      "learning_rate": 5.589430894308943e-06,
      "loss": 0.389,
      "step": 2950
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.7082429501084598,
      "eval_loss": 0.8595220446586609,
      "eval_runtime": 41.6687,
      "eval_samples_per_second": 22.127,
      "eval_steps_per_second": 0.696,
      "step": 2952
    },
    {
      "epoch": 18.140243902439025,
      "grad_norm": 1.9708536863327026,
      "learning_rate": 5.165989159891599e-06,
      "loss": 0.3896,
      "step": 2975
    },
    {
      "epoch": 18.29268292682927,
      "grad_norm": 4.3860673904418945,
      "learning_rate": 4.742547425474255e-06,
      "loss": 0.3939,
      "step": 3000
    },
    {
      "epoch": 18.445121951219512,
      "grad_norm": 2.5189573764801025,
      "learning_rate": 4.319105691056911e-06,
      "loss": 0.3908,
      "step": 3025
    },
    {
      "epoch": 18.597560975609756,
      "grad_norm": 1.8949400186538696,
      "learning_rate": 3.8956639566395666e-06,
      "loss": 0.3947,
      "step": 3050
    },
    {
      "epoch": 18.75,
      "grad_norm": 2.0496764183044434,
      "learning_rate": 3.4722222222222224e-06,
      "loss": 0.3837,
      "step": 3075
    },
    {
      "epoch": 18.902439024390244,
      "grad_norm": 3.2844817638397217,
      "learning_rate": 3.0487804878048782e-06,
      "loss": 0.3982,
      "step": 3100
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.6984815618221258,
      "eval_loss": 0.8660802841186523,
      "eval_runtime": 41.7631,
      "eval_samples_per_second": 22.077,
      "eval_steps_per_second": 0.694,
      "step": 3116
    },
    {
      "epoch": 19.054878048780488,
      "grad_norm": 2.776163339614868,
      "learning_rate": 2.625338753387534e-06,
      "loss": 0.3863,
      "step": 3125
    },
    {
      "epoch": 19.20731707317073,
      "grad_norm": 2.2934467792510986,
      "learning_rate": 2.2018970189701894e-06,
      "loss": 0.394,
      "step": 3150
    },
    {
      "epoch": 19.359756097560975,
      "grad_norm": 1.4411579370498657,
      "learning_rate": 1.7784552845528455e-06,
      "loss": 0.3908,
      "step": 3175
    },
    {
      "epoch": 19.51219512195122,
      "grad_norm": 1.6821469068527222,
      "learning_rate": 1.3550135501355015e-06,
      "loss": 0.3844,
      "step": 3200
    },
    {
      "epoch": 19.664634146341463,
      "grad_norm": 3.2860209941864014,
      "learning_rate": 9.315718157181572e-07,
      "loss": 0.3879,
      "step": 3225
    },
    {
      "epoch": 19.817073170731707,
      "grad_norm": 4.093799591064453,
      "learning_rate": 5.081300813008131e-07,
      "loss": 0.3852,
      "step": 3250
    },
    {
      "epoch": 19.96951219512195,
      "grad_norm": 1.6917803287506104,
      "learning_rate": 8.468834688346885e-08,
      "loss": 0.3823,
      "step": 3275
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.7114967462039046,
      "eval_loss": 0.8688989281654358,
      "eval_runtime": 41.6761,
      "eval_samples_per_second": 22.123,
      "eval_steps_per_second": 0.696,
      "step": 3280
    },
    {
      "epoch": 20.0,
      "step": 3280,
      "total_flos": 0.0,
      "train_loss": 0.5078278935537106,
      "train_runtime": 14435.8806,
      "train_samples_per_second": 7.238,
      "train_steps_per_second": 0.227
    }
  ],
  "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
}