{
  "best_metric": 0.4126315116882324,
  "best_model_checkpoint": "data/Gemma-2-2B_task-3_60-samples_config-2/checkpoint-31",
  "epoch": 17.73913043478261,
  "eval_steps": 500,
  "global_step": 51,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 6.895533084869385,
      "learning_rate": 1e-05,
      "loss": 3.4368,
      "step": 1
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 6.358642101287842,
      "learning_rate": 2e-05,
      "loss": 3.2454,
      "step": 2
    },
    {
      "epoch": 0.6956521739130435,
      "eval_loss": 3.234452486038208,
      "eval_runtime": 4.3298,
      "eval_samples_per_second": 2.772,
      "eval_steps_per_second": 2.772,
      "step": 2
    },
    {
      "epoch": 1.391304347826087,
      "grad_norm": 6.256963729858398,
      "learning_rate": 4e-05,
      "loss": 3.2274,
      "step": 4
    },
    {
      "epoch": 1.7391304347826086,
      "eval_loss": 2.691072702407837,
      "eval_runtime": 4.1437,
      "eval_samples_per_second": 2.896,
      "eval_steps_per_second": 2.896,
      "step": 5
    },
    {
      "epoch": 2.0869565217391304,
      "grad_norm": 5.946145534515381,
      "learning_rate": 6e-05,
      "loss": 2.9386,
      "step": 6
    },
    {
      "epoch": 2.782608695652174,
      "grad_norm": 3.9147167205810547,
      "learning_rate": 8e-05,
      "loss": 2.1741,
      "step": 8
    },
    {
      "epoch": 2.782608695652174,
      "eval_loss": 1.4943164587020874,
      "eval_runtime": 4.1462,
      "eval_samples_per_second": 2.894,
      "eval_steps_per_second": 2.894,
      "step": 8
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 2.797118663787842,
      "learning_rate": 0.0001,
      "loss": 1.3679,
      "step": 10
    },
    {
      "epoch": 3.8260869565217392,
      "eval_loss": 0.8701855540275574,
      "eval_runtime": 4.1432,
      "eval_samples_per_second": 2.896,
      "eval_steps_per_second": 2.896,
      "step": 11
    },
    {
      "epoch": 4.173913043478261,
      "grad_norm": 2.0996110439300537,
      "learning_rate": 9.987820251299122e-05,
      "loss": 0.9062,
      "step": 12
    },
    {
      "epoch": 4.869565217391305,
      "grad_norm": 0.9581406116485596,
      "learning_rate": 9.951340343707852e-05,
      "loss": 0.6621,
      "step": 14
    },
    {
      "epoch": 4.869565217391305,
      "eval_loss": 0.6006700992584229,
      "eval_runtime": 4.1434,
      "eval_samples_per_second": 2.896,
      "eval_steps_per_second": 2.896,
      "step": 14
    },
    {
      "epoch": 5.565217391304348,
      "grad_norm": 0.6334660053253174,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.5808,
      "step": 16
    },
    {
      "epoch": 5.913043478260869,
      "eval_loss": 0.5170578360557556,
      "eval_runtime": 4.1457,
      "eval_samples_per_second": 2.895,
      "eval_steps_per_second": 2.895,
      "step": 17
    },
    {
      "epoch": 6.260869565217392,
      "grad_norm": 0.5526132583618164,
      "learning_rate": 9.806308479691595e-05,
      "loss": 0.3845,
      "step": 18
    },
    {
      "epoch": 6.956521739130435,
      "grad_norm": 0.5105712413787842,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.456,
      "step": 20
    },
    {
      "epoch": 6.956521739130435,
      "eval_loss": 0.4745006561279297,
      "eval_runtime": 4.1408,
      "eval_samples_per_second": 2.898,
      "eval_steps_per_second": 2.898,
      "step": 20
    },
    {
      "epoch": 7.6521739130434785,
      "grad_norm": 0.38354527950286865,
      "learning_rate": 9.567727288213005e-05,
      "loss": 0.3027,
      "step": 22
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.45038852095603943,
      "eval_runtime": 4.142,
      "eval_samples_per_second": 2.897,
      "eval_steps_per_second": 2.897,
      "step": 23
    },
    {
      "epoch": 8.347826086956522,
      "grad_norm": 0.36689186096191406,
      "learning_rate": 9.414737964294636e-05,
      "loss": 0.3998,
      "step": 24
    },
    {
      "epoch": 8.695652173913043,
      "eval_loss": 0.43608060479164124,
      "eval_runtime": 4.1401,
      "eval_samples_per_second": 2.898,
      "eval_steps_per_second": 2.898,
      "step": 25
    },
    {
      "epoch": 9.043478260869565,
      "grad_norm": 0.4304007887840271,
      "learning_rate": 9.24024048078213e-05,
      "loss": 0.3543,
      "step": 26
    },
    {
      "epoch": 9.73913043478261,
      "grad_norm": 0.529371440410614,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.2969,
      "step": 28
    },
    {
      "epoch": 9.73913043478261,
      "eval_loss": 0.42082512378692627,
      "eval_runtime": 4.1406,
      "eval_samples_per_second": 2.898,
      "eval_steps_per_second": 2.898,
      "step": 28
    },
    {
      "epoch": 10.434782608695652,
      "grad_norm": 0.3541157841682434,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.283,
      "step": 30
    },
    {
      "epoch": 10.782608695652174,
      "eval_loss": 0.4126315116882324,
      "eval_runtime": 4.1424,
      "eval_samples_per_second": 2.897,
      "eval_steps_per_second": 2.897,
      "step": 31
    },
    {
      "epoch": 11.130434782608695,
      "grad_norm": 0.42536216974258423,
      "learning_rate": 8.596699001693255e-05,
      "loss": 0.254,
      "step": 32
    },
    {
      "epoch": 11.826086956521738,
      "grad_norm": 0.6485491394996643,
      "learning_rate": 8.345653031794292e-05,
      "loss": 0.1909,
      "step": 34
    },
    {
      "epoch": 11.826086956521738,
      "eval_loss": 0.41434308886528015,
      "eval_runtime": 4.1458,
      "eval_samples_per_second": 2.895,
      "eval_steps_per_second": 2.895,
      "step": 34
    },
    {
      "epoch": 12.521739130434783,
      "grad_norm": 0.8147707581520081,
      "learning_rate": 8.07830737662829e-05,
      "loss": 0.1833,
      "step": 36
    },
    {
      "epoch": 12.869565217391305,
      "eval_loss": 0.43228474259376526,
      "eval_runtime": 4.1454,
      "eval_samples_per_second": 2.895,
      "eval_steps_per_second": 2.895,
      "step": 37
    },
    {
      "epoch": 13.217391304347826,
      "grad_norm": 0.5262936949729919,
      "learning_rate": 7.795964517353735e-05,
      "loss": 0.2061,
      "step": 38
    },
    {
      "epoch": 13.91304347826087,
      "grad_norm": 0.46287646889686584,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.1128,
      "step": 40
    },
    {
      "epoch": 13.91304347826087,
      "eval_loss": 0.454862117767334,
      "eval_runtime": 4.1443,
      "eval_samples_per_second": 2.896,
      "eval_steps_per_second": 2.896,
      "step": 40
    },
    {
      "epoch": 14.608695652173914,
      "grad_norm": 0.6421369314193726,
      "learning_rate": 7.191855733945387e-05,
      "loss": 0.16,
      "step": 42
    },
    {
      "epoch": 14.956521739130435,
      "eval_loss": 0.4736705720424652,
      "eval_runtime": 4.1487,
      "eval_samples_per_second": 2.892,
      "eval_steps_per_second": 2.892,
      "step": 43
    },
    {
      "epoch": 15.304347826086957,
      "grad_norm": 0.7217320203781128,
      "learning_rate": 6.873032967079561e-05,
      "loss": 0.0901,
      "step": 44
    },
    {
      "epoch": 16.0,
      "grad_norm": 0.5217561721801758,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.0817,
      "step": 46
    },
    {
      "epoch": 16.0,
      "eval_loss": 0.5360535979270935,
      "eval_runtime": 4.141,
      "eval_samples_per_second": 2.898,
      "eval_steps_per_second": 2.898,
      "step": 46
    },
    {
      "epoch": 16.695652173913043,
      "grad_norm": 0.7089937329292297,
      "learning_rate": 6.209609477998338e-05,
      "loss": 0.0679,
      "step": 48
    },
    {
      "epoch": 16.695652173913043,
      "eval_loss": 0.5978057980537415,
      "eval_runtime": 4.1483,
      "eval_samples_per_second": 2.893,
      "eval_steps_per_second": 2.893,
      "step": 48
    },
    {
      "epoch": 17.391304347826086,
      "grad_norm": 0.3900970220565796,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.0501,
      "step": 50
    },
    {
      "epoch": 17.73913043478261,
      "eval_loss": 0.6335070729255676,
      "eval_runtime": 4.1479,
      "eval_samples_per_second": 2.893,
      "eval_steps_per_second": 2.893,
      "step": 51
    },
    {
      "epoch": 17.73913043478261,
      "step": 51,
      "total_flos": 2.2594250923311104e+16,
      "train_loss": 0.7488247088941873,
      "train_runtime": 967.2435,
      "train_samples_per_second": 2.378,
      "train_steps_per_second": 0.103
    }
  ],
  "logging_steps": 2,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 25,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 7,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.2594250923311104e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}