{
  "best_metric": 0.9444444179534912,
  "best_model_checkpoint": "/kaggle/working/ckpts/checkpoint-1784",
  "epoch": 11.0,
  "eval_steps": 500,
  "global_step": 2453,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.45,
      "grad_norm": 3.8796565532684326,
      "learning_rate": 9.701046337817639e-06,
      "loss": 1.0979,
      "step": 100
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.6467082500457764,
      "learning_rate": 9.4050822122571e-06,
      "loss": 0.7722,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8434343338012695,
      "eval_loss": 0.4733273386955261,
      "eval_runtime": 9.4201,
      "eval_samples_per_second": 21.019,
      "eval_steps_per_second": 5.308,
      "step": 223
    },
    {
      "epoch": 1.35,
      "grad_norm": 20.371318817138672,
      "learning_rate": 9.106128550074738e-06,
      "loss": 0.5871,
      "step": 300
    },
    {
      "epoch": 1.79,
      "grad_norm": 3.244633674621582,
      "learning_rate": 8.807174887892378e-06,
      "loss": 0.4755,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.868686854839325,
      "eval_loss": 0.42396825551986694,
      "eval_runtime": 9.3161,
      "eval_samples_per_second": 21.254,
      "eval_steps_per_second": 5.367,
      "step": 446
    },
    {
      "epoch": 2.24,
      "grad_norm": 8.883748054504395,
      "learning_rate": 8.51121076233184e-06,
      "loss": 0.377,
      "step": 500
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.6546649932861328,
      "learning_rate": 8.2152466367713e-06,
      "loss": 0.3262,
      "step": 600
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9343434572219849,
      "eval_loss": 0.2939111590385437,
      "eval_runtime": 9.4493,
      "eval_samples_per_second": 20.954,
      "eval_steps_per_second": 5.291,
      "step": 669
    },
    {
      "epoch": 3.14,
      "grad_norm": 7.946984767913818,
      "learning_rate": 7.916292974588939e-06,
      "loss": 0.2347,
      "step": 700
    },
    {
      "epoch": 3.59,
      "grad_norm": 9.960402488708496,
      "learning_rate": 7.617339312406578e-06,
      "loss": 0.2642,
      "step": 800
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9292929172515869,
      "eval_loss": 0.3087417781352997,
      "eval_runtime": 9.224,
      "eval_samples_per_second": 21.466,
      "eval_steps_per_second": 5.421,
      "step": 892
    },
    {
      "epoch": 4.04,
      "grad_norm": 87.66458129882812,
      "learning_rate": 7.318385650224216e-06,
      "loss": 0.2608,
      "step": 900
    },
    {
      "epoch": 4.48,
      "grad_norm": 42.06097412109375,
      "learning_rate": 7.019431988041854e-06,
      "loss": 0.213,
      "step": 1000
    },
    {
      "epoch": 4.93,
      "grad_norm": 21.227588653564453,
      "learning_rate": 6.723467862481315e-06,
      "loss": 0.191,
      "step": 1100
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.939393937587738,
      "eval_loss": 0.30786794424057007,
      "eval_runtime": 9.2259,
      "eval_samples_per_second": 21.461,
      "eval_steps_per_second": 5.42,
      "step": 1115
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.09492979198694229,
      "learning_rate": 6.424514200298954e-06,
      "loss": 0.1891,
      "step": 1200
    },
    {
      "epoch": 5.83,
      "grad_norm": 22.492895126342773,
      "learning_rate": 6.1255605381165925e-06,
      "loss": 0.1534,
      "step": 1300
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.939393937587738,
      "eval_loss": 0.3133719265460968,
      "eval_runtime": 9.3193,
      "eval_samples_per_second": 21.246,
      "eval_steps_per_second": 5.365,
      "step": 1338
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.05382364243268967,
      "learning_rate": 5.826606875934231e-06,
      "loss": 0.1825,
      "step": 1400
    },
    {
      "epoch": 6.73,
      "grad_norm": 5.18447732925415,
      "learning_rate": 5.527653213751869e-06,
      "loss": 0.1571,
      "step": 1500
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9292929172515869,
      "eval_loss": 0.40089717507362366,
      "eval_runtime": 9.2909,
      "eval_samples_per_second": 21.311,
      "eval_steps_per_second": 5.382,
      "step": 1561
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.036003902554512024,
      "learning_rate": 5.228699551569507e-06,
      "loss": 0.1518,
      "step": 1600
    },
    {
      "epoch": 7.62,
      "grad_norm": 0.10409737378358841,
      "learning_rate": 4.929745889387145e-06,
      "loss": 0.1328,
      "step": 1700
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9444444179534912,
      "eval_loss": 0.3189202845096588,
      "eval_runtime": 9.3287,
      "eval_samples_per_second": 21.225,
      "eval_steps_per_second": 5.36,
      "step": 1784
    },
    {
      "epoch": 8.07,
      "grad_norm": 6.580456733703613,
      "learning_rate": 4.630792227204783e-06,
      "loss": 0.1127,
      "step": 1800
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.12464825063943863,
      "learning_rate": 4.3318385650224224e-06,
      "loss": 0.1333,
      "step": 1900
    },
    {
      "epoch": 8.97,
      "grad_norm": 127.78559112548828,
      "learning_rate": 4.03288490284006e-06,
      "loss": 0.1567,
      "step": 2000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9191918969154358,
      "eval_loss": 0.40891000628471375,
      "eval_runtime": 9.2849,
      "eval_samples_per_second": 21.325,
      "eval_steps_per_second": 5.385,
      "step": 2007
    },
    {
      "epoch": 9.42,
      "grad_norm": 0.03460687771439552,
      "learning_rate": 3.7339312406576984e-06,
      "loss": 0.1313,
      "step": 2100
    },
    {
      "epoch": 9.87,
      "grad_norm": 3.1638216972351074,
      "learning_rate": 3.4349775784753366e-06,
      "loss": 0.1043,
      "step": 2200
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9343434572219849,
      "eval_loss": 0.34286314249038696,
      "eval_runtime": 9.3365,
      "eval_samples_per_second": 21.207,
      "eval_steps_per_second": 5.355,
      "step": 2230
    },
    {
      "epoch": 10.31,
      "grad_norm": 0.462053507566452,
      "learning_rate": 3.136023916292975e-06,
      "loss": 0.1551,
      "step": 2300
    },
    {
      "epoch": 10.76,
      "grad_norm": 0.5621947050094604,
      "learning_rate": 2.8370702541106134e-06,
      "loss": 0.1161,
      "step": 2400
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.939393937587738,
      "eval_loss": 0.3534471094608307,
      "eval_runtime": 9.2307,
      "eval_samples_per_second": 21.45,
      "eval_steps_per_second": 5.417,
      "step": 2453
    },
    {
      "epoch": 11.0,
      "step": 2453,
      "total_flos": 6.273609670944864e+17,
      "train_loss": 0.2735439113728912,
      "train_runtime": 1317.5896,
      "train_samples_per_second": 20.276,
      "train_steps_per_second": 2.539
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.9444444179534912,
      "eval_loss": 0.3189202845096588,
      "eval_runtime": 9.0656,
      "eval_samples_per_second": 21.841,
      "eval_steps_per_second": 5.515,
      "step": 2453
    }
  ],
  "logging_steps": 100,
  "max_steps": 3345,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 6.273609670944864e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}