{
  "best_metric": 11.062067031860352,
  "best_model_checkpoint": "miner_id_24/checkpoint-30",
  "epoch": 0.003370502485745583,
  "eval_steps": 5,
  "global_step": 30,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00011235008285818611,
      "eval_loss": 11.085953712463379,
      "eval_runtime": 29.6488,
      "eval_samples_per_second": 252.826,
      "eval_steps_per_second": 63.207,
      "step": 1
    },
    {
      "epoch": 0.0003370502485745583,
      "grad_norm": 0.531462550163269,
      "learning_rate": 0.00012,
      "loss": 44.3417,
      "step": 3
    },
    {
      "epoch": 0.0005617504142909305,
      "eval_loss": 11.082646369934082,
      "eval_runtime": 29.604,
      "eval_samples_per_second": 253.209,
      "eval_steps_per_second": 63.302,
      "step": 5
    },
    {
      "epoch": 0.0006741004971491166,
      "grad_norm": 0.5337783098220825,
      "learning_rate": 0.0001992114701314478,
      "loss": 44.3341,
      "step": 6
    },
    {
      "epoch": 0.001011150745723675,
      "grad_norm": 0.5475996136665344,
      "learning_rate": 0.00018763066800438636,
      "loss": 44.3174,
      "step": 9
    },
    {
      "epoch": 0.001123500828581861,
      "eval_loss": 11.075613021850586,
      "eval_runtime": 29.5455,
      "eval_samples_per_second": 253.71,
      "eval_steps_per_second": 63.428,
      "step": 10
    },
    {
      "epoch": 0.0013482009942982333,
      "grad_norm": 0.5746555924415588,
      "learning_rate": 0.000163742398974869,
      "loss": 44.3016,
      "step": 12
    },
    {
      "epoch": 0.0016852512428727916,
      "grad_norm": 0.5716856122016907,
      "learning_rate": 0.00013090169943749476,
      "loss": 44.2886,
      "step": 15
    },
    {
      "epoch": 0.0016852512428727916,
      "eval_loss": 11.069120407104492,
      "eval_runtime": 30.4362,
      "eval_samples_per_second": 246.286,
      "eval_steps_per_second": 61.572,
      "step": 15
    },
    {
      "epoch": 0.00202230149144735,
      "grad_norm": 0.5406591892242432,
      "learning_rate": 9.372094804706867e-05,
      "loss": 44.2775,
      "step": 18
    },
    {
      "epoch": 0.002247001657163722,
      "eval_loss": 11.06486701965332,
      "eval_runtime": 30.1812,
      "eval_samples_per_second": 248.366,
      "eval_steps_per_second": 62.092,
      "step": 20
    },
    {
      "epoch": 0.0023593517400219083,
      "grad_norm": 0.6206434369087219,
      "learning_rate": 5.7422070843492734e-05,
      "loss": 44.2574,
      "step": 21
    },
    {
      "epoch": 0.0026964019885964666,
      "grad_norm": 0.6306174397468567,
      "learning_rate": 2.7103137257858868e-05,
      "loss": 44.2539,
      "step": 24
    },
    {
      "epoch": 0.002808752071454653,
      "eval_loss": 11.06248950958252,
      "eval_runtime": 30.3886,
      "eval_samples_per_second": 246.671,
      "eval_steps_per_second": 61.668,
      "step": 25
    },
    {
      "epoch": 0.003033452237171025,
      "grad_norm": 0.6203663945198059,
      "learning_rate": 7.022351411174866e-06,
      "loss": 44.2556,
      "step": 27
    },
    {
      "epoch": 0.003370502485745583,
      "grad_norm": 0.6384527087211609,
      "learning_rate": 0.0,
      "loss": 44.2531,
      "step": 30
    },
    {
      "epoch": 0.003370502485745583,
      "eval_loss": 11.062067031860352,
      "eval_runtime": 29.7801,
      "eval_samples_per_second": 251.712,
      "eval_steps_per_second": 62.928,
      "step": 30
    }
  ],
  "logging_steps": 3,
  "max_steps": 30,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 10,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 1,
        "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": 224604979200.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}