{
  "best_metric": 0.06443430483341217,
  "best_model_checkpoint": "./beans_outputs/checkpoint-520",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 650,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "grad_norm": 2.198350429534912,
      "learning_rate": 1.9692307692307696e-05,
      "loss": 1.0245,
      "step": 10
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.8455473184585571,
      "learning_rate": 1.9384615384615386e-05,
      "loss": 0.9453,
      "step": 20
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.1065616607666016,
      "learning_rate": 1.907692307692308e-05,
      "loss": 0.8403,
      "step": 30
    },
    {
      "epoch": 0.31,
      "grad_norm": 2.614351749420166,
      "learning_rate": 1.876923076923077e-05,
      "loss": 0.6945,
      "step": 40
    },
    {
      "epoch": 0.38,
      "grad_norm": 4.106573104858398,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.6607,
      "step": 50
    },
    {
      "epoch": 0.46,
      "grad_norm": 2.129425048828125,
      "learning_rate": 1.8153846153846155e-05,
      "loss": 0.5811,
      "step": 60
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.8207303285598755,
      "learning_rate": 1.784615384615385e-05,
      "loss": 0.5191,
      "step": 70
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.2188501358032227,
      "learning_rate": 1.753846153846154e-05,
      "loss": 0.3937,
      "step": 80
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.4855401515960693,
      "learning_rate": 1.7230769230769234e-05,
      "loss": 0.3523,
      "step": 90
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.9041554927825928,
      "learning_rate": 1.6923076923076924e-05,
      "loss": 0.3669,
      "step": 100
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.978524923324585,
      "learning_rate": 1.6615384615384618e-05,
      "loss": 0.3324,
      "step": 110
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.6394453048706055,
      "learning_rate": 1.630769230769231e-05,
      "loss": 0.2456,
      "step": 120
    },
    {
      "epoch": 1.0,
      "grad_norm": 11.733540534973145,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.2826,
      "step": 130
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9624060150375939,
      "eval_loss": 0.21461370587348938,
      "eval_runtime": 1.7023,
      "eval_samples_per_second": 78.132,
      "eval_steps_per_second": 9.987,
      "step": 130
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.2605695724487305,
      "learning_rate": 1.5692307692307693e-05,
      "loss": 0.1986,
      "step": 140
    },
    {
      "epoch": 1.15,
      "grad_norm": 3.18729829788208,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.2255,
      "step": 150
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.4626598358154297,
      "learning_rate": 1.5076923076923078e-05,
      "loss": 0.205,
      "step": 160
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.5206894278526306,
      "learning_rate": 1.4769230769230772e-05,
      "loss": 0.2284,
      "step": 170
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.6028104424476624,
      "learning_rate": 1.4461538461538462e-05,
      "loss": 0.2141,
      "step": 180
    },
    {
      "epoch": 1.46,
      "grad_norm": 3.1314830780029297,
      "learning_rate": 1.4153846153846156e-05,
      "loss": 0.1602,
      "step": 190
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.4666638672351837,
      "learning_rate": 1.3846153846153847e-05,
      "loss": 0.259,
      "step": 200
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.2395788431167603,
      "learning_rate": 1.353846153846154e-05,
      "loss": 0.1841,
      "step": 210
    },
    {
      "epoch": 1.69,
      "grad_norm": 2.4894907474517822,
      "learning_rate": 1.3230769230769231e-05,
      "loss": 0.2117,
      "step": 220
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.910448670387268,
      "learning_rate": 1.2923076923076925e-05,
      "loss": 0.1682,
      "step": 230
    },
    {
      "epoch": 1.85,
      "grad_norm": 4.8757429122924805,
      "learning_rate": 1.2615384615384616e-05,
      "loss": 0.2634,
      "step": 240
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.37920522689819336,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.1402,
      "step": 250
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.3483049273490906,
      "learning_rate": 1.2e-05,
      "loss": 0.1306,
      "step": 260
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9699248120300752,
      "eval_loss": 0.12822628021240234,
      "eval_runtime": 1.6876,
      "eval_samples_per_second": 78.81,
      "eval_steps_per_second": 10.073,
      "step": 260
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.4993837773799896,
      "learning_rate": 1.1692307692307694e-05,
      "loss": 0.0933,
      "step": 270
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.33854740858078003,
      "learning_rate": 1.1384615384615385e-05,
      "loss": 0.1407,
      "step": 280
    },
    {
      "epoch": 2.23,
      "grad_norm": 6.487577438354492,
      "learning_rate": 1.1076923076923079e-05,
      "loss": 0.1207,
      "step": 290
    },
    {
      "epoch": 2.31,
      "grad_norm": 7.572403430938721,
      "learning_rate": 1.076923076923077e-05,
      "loss": 0.139,
      "step": 300
    },
    {
      "epoch": 2.38,
      "grad_norm": 4.639190196990967,
      "learning_rate": 1.0461538461538463e-05,
      "loss": 0.1735,
      "step": 310
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.31310051679611206,
      "learning_rate": 1.0153846153846154e-05,
      "loss": 0.1563,
      "step": 320
    },
    {
      "epoch": 2.54,
      "grad_norm": 6.506566047668457,
      "learning_rate": 9.846153846153848e-06,
      "loss": 0.185,
      "step": 330
    },
    {
      "epoch": 2.62,
      "grad_norm": 1.2759567499160767,
      "learning_rate": 9.53846153846154e-06,
      "loss": 0.0761,
      "step": 340
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.34429147839546204,
      "learning_rate": 9.230769230769232e-06,
      "loss": 0.1355,
      "step": 350
    },
    {
      "epoch": 2.77,
      "grad_norm": 9.9702787399292,
      "learning_rate": 8.923076923076925e-06,
      "loss": 0.1176,
      "step": 360
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.3892934024333954,
      "learning_rate": 8.615384615384617e-06,
      "loss": 0.1507,
      "step": 370
    },
    {
      "epoch": 2.92,
      "grad_norm": 1.1851181983947754,
      "learning_rate": 8.307692307692309e-06,
      "loss": 0.1324,
      "step": 380
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.29040664434432983,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.1451,
      "step": 390
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9774436090225563,
      "eval_loss": 0.09662821888923645,
      "eval_runtime": 1.7019,
      "eval_samples_per_second": 78.15,
      "eval_steps_per_second": 9.989,
      "step": 390
    },
    {
      "epoch": 3.08,
      "grad_norm": 2.813366174697876,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.1365,
      "step": 400
    },
    {
      "epoch": 3.15,
      "grad_norm": 6.747802257537842,
      "learning_rate": 7.384615384615386e-06,
      "loss": 0.1604,
      "step": 410
    },
    {
      "epoch": 3.23,
      "grad_norm": 8.719525337219238,
      "learning_rate": 7.076923076923078e-06,
      "loss": 0.1448,
      "step": 420
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.23202432692050934,
      "learning_rate": 6.76923076923077e-06,
      "loss": 0.0717,
      "step": 430
    },
    {
      "epoch": 3.38,
      "grad_norm": 4.157703399658203,
      "learning_rate": 6.461538461538463e-06,
      "loss": 0.1157,
      "step": 440
    },
    {
      "epoch": 3.46,
      "grad_norm": 12.880057334899902,
      "learning_rate": 6.153846153846155e-06,
      "loss": 0.0834,
      "step": 450
    },
    {
      "epoch": 3.54,
      "grad_norm": 5.217643737792969,
      "learning_rate": 5.846153846153847e-06,
      "loss": 0.0972,
      "step": 460
    },
    {
      "epoch": 3.62,
      "grad_norm": 8.088122367858887,
      "learning_rate": 5.538461538461539e-06,
      "loss": 0.0788,
      "step": 470
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.22004778683185577,
      "learning_rate": 5.230769230769232e-06,
      "loss": 0.0884,
      "step": 480
    },
    {
      "epoch": 3.77,
      "grad_norm": 0.21302741765975952,
      "learning_rate": 4.923076923076924e-06,
      "loss": 0.0752,
      "step": 490
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.8432028889656067,
      "learning_rate": 4.615384615384616e-06,
      "loss": 0.1344,
      "step": 500
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.21321173012256622,
      "learning_rate": 4.307692307692308e-06,
      "loss": 0.1274,
      "step": 510
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3039676249027252,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0785,
      "step": 520
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9924812030075187,
      "eval_loss": 0.06443430483341217,
      "eval_runtime": 1.7085,
      "eval_samples_per_second": 77.845,
      "eval_steps_per_second": 9.95,
      "step": 520
    },
    {
      "epoch": 4.08,
      "grad_norm": 8.200630187988281,
      "learning_rate": 3.692307692307693e-06,
      "loss": 0.1003,
      "step": 530
    },
    {
      "epoch": 4.15,
      "grad_norm": 1.6478914022445679,
      "learning_rate": 3.384615384615385e-06,
      "loss": 0.1675,
      "step": 540
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.3463890254497528,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 0.1522,
      "step": 550
    },
    {
      "epoch": 4.31,
      "grad_norm": 1.3014988899230957,
      "learning_rate": 2.7692307692307697e-06,
      "loss": 0.0504,
      "step": 560
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.3259345591068268,
      "learning_rate": 2.461538461538462e-06,
      "loss": 0.1332,
      "step": 570
    },
    {
      "epoch": 4.46,
      "grad_norm": 0.2652953565120697,
      "learning_rate": 2.153846153846154e-06,
      "loss": 0.0824,
      "step": 580
    },
    {
      "epoch": 4.54,
      "grad_norm": 8.841487884521484,
      "learning_rate": 1.8461538461538465e-06,
      "loss": 0.0709,
      "step": 590
    },
    {
      "epoch": 4.62,
      "grad_norm": 1.8583545684814453,
      "learning_rate": 1.5384615384615387e-06,
      "loss": 0.0921,
      "step": 600
    },
    {
      "epoch": 4.69,
      "grad_norm": 5.223345756530762,
      "learning_rate": 1.230769230769231e-06,
      "loss": 0.0869,
      "step": 610
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.26816922426223755,
      "learning_rate": 9.230769230769232e-07,
      "loss": 0.0808,
      "step": 620
    },
    {
      "epoch": 4.85,
      "grad_norm": 0.2151552438735962,
      "learning_rate": 6.153846153846155e-07,
      "loss": 0.0662,
      "step": 630
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.23172561824321747,
      "learning_rate": 3.0769230769230774e-07,
      "loss": 0.0783,
      "step": 640
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5893689393997192,
      "learning_rate": 0.0,
      "loss": 0.1125,
      "step": 650
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9774436090225563,
      "eval_loss": 0.07952458411455154,
      "eval_runtime": 1.7098,
      "eval_samples_per_second": 77.786,
      "eval_steps_per_second": 9.943,
      "step": 650
    },
    {
      "epoch": 5.0,
      "step": 650,
      "total_flos": 4.006371770595533e+17,
      "train_loss": 0.21938713550567626,
      "train_runtime": 203.0056,
      "train_samples_per_second": 25.467,
      "train_steps_per_second": 3.202
    }
  ],
  "logging_steps": 10,
  "max_steps": 650,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 4.006371770595533e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}