last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.014035087719298246,
  "eval_steps": 5,
  "global_step": 33,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004253056884635832,
      "eval_loss": 2.213833808898926,
      "eval_runtime": 295.9015,
      "eval_samples_per_second": 3.346,
      "eval_steps_per_second": 1.673,
      "step": 1
    },
    {
      "epoch": 0.0012759170653907496,
      "grad_norm": 5.44368314743042,
      "learning_rate": 3.529411764705883e-05,
      "loss": 8.8189,
      "step": 3
    },
    {
      "epoch": 0.002126528442317916,
      "eval_loss": 2.185218572616577,
      "eval_runtime": 298.0199,
      "eval_samples_per_second": 3.322,
      "eval_steps_per_second": 1.661,
      "step": 5
    },
    {
      "epoch": 0.002551834130781499,
      "grad_norm": 6.1356096267700195,
      "learning_rate": 7.058823529411765e-05,
      "loss": 9.273,
      "step": 6
    },
    {
      "epoch": 0.003827751196172249,
      "grad_norm": 5.341865539550781,
      "learning_rate": 0.00010588235294117647,
      "loss": 8.6346,
      "step": 9
    },
    {
      "epoch": 0.004253056884635832,
      "eval_loss": 2.152097225189209,
      "eval_runtime": 297.9685,
      "eval_samples_per_second": 3.322,
      "eval_steps_per_second": 1.661,
      "step": 10
    },
    {
      "epoch": 0.005103668261562998,
      "grad_norm": 6.478035926818848,
      "learning_rate": 0.0001411764705882353,
      "loss": 8.6021,
      "step": 12
    },
    {
      "epoch": 0.006379585326953748,
      "grad_norm": 5.044421195983887,
      "learning_rate": 0.00017647058823529413,
      "loss": 8.3593,
      "step": 15
    },
    {
      "epoch": 0.006379585326953748,
      "eval_loss": 2.1351099014282227,
      "eval_runtime": 297.5366,
      "eval_samples_per_second": 3.327,
      "eval_steps_per_second": 1.664,
      "step": 15
    },
    {
      "epoch": 0.007655502392344498,
      "grad_norm": 4.361161231994629,
      "learning_rate": 0.00019807852804032305,
      "loss": 8.175,
      "step": 18
    },
    {
      "epoch": 0.008506113769271665,
      "eval_loss": 2.126591205596924,
      "eval_runtime": 297.8187,
      "eval_samples_per_second": 3.324,
      "eval_steps_per_second": 1.662,
      "step": 20
    },
    {
      "epoch": 0.008931419457735247,
      "grad_norm": 4.228249549865723,
      "learning_rate": 0.00017071067811865476,
      "loss": 8.1639,
      "step": 21
    },
    {
      "epoch": 0.010207336523125997,
      "grad_norm": 3.8976006507873535,
      "learning_rate": 0.00011950903220161285,
      "loss": 8.1004,
      "step": 24
    },
    {
      "epoch": 0.01063264221158958,
      "eval_loss": 2.1243057250976562,
      "eval_runtime": 298.0859,
      "eval_samples_per_second": 3.321,
      "eval_steps_per_second": 1.661,
      "step": 25
    },
    {
      "epoch": 0.011483253588516746,
      "grad_norm": 4.909876346588135,
      "learning_rate": 6.173165676349103e-05,
      "loss": 8.4362,
      "step": 27
    },
    {
      "epoch": 0.012759170653907496,
      "grad_norm": 3.950687885284424,
      "learning_rate": 1.6853038769745467e-05,
      "loss": 8.2356,
      "step": 30
    },
    {
      "epoch": 0.012759170653907496,
      "eval_loss": 2.118710994720459,
      "eval_runtime": 298.1844,
      "eval_samples_per_second": 3.32,
      "eval_steps_per_second": 1.66,
      "step": 30
    },
    {
      "epoch": 0.014035087719298246,
      "grad_norm": 4.570465564727783,
      "learning_rate": 0.0,
      "loss": 7.8479,
      "step": 33
    }
  ],
  "logging_steps": 3,
  "max_steps": 33,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 17,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.065442902376448e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}