MusicBot-ft/checkpoint-24/trainer_state.json

{
  "best_metric": 1.7682936191558838,
  "best_model_checkpoint": "MusicBot-ft/checkpoint-24",
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 24,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.1322076320648193,
      "learning_rate": 0.0003,
      "loss": 3.3408,
      "step": 2
    },
    {
      "epoch": 1.0,
      "eval_loss": 3.314174175262451,
      "eval_runtime": 4.3893,
      "eval_samples_per_second": 1.823,
      "eval_steps_per_second": 0.456,
      "step": 2
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.0614089965820312,
      "learning_rate": 0.0002727272727272727,
      "loss": 3.075,
      "step": 4
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.82991886138916,
      "eval_runtime": 4.4015,
      "eval_samples_per_second": 1.818,
      "eval_steps_per_second": 0.454,
      "step": 4
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.8810017108917236,
      "learning_rate": 0.00024545454545454545,
      "loss": 2.618,
      "step": 6
    },
    {
      "epoch": 3.0,
      "eval_loss": 2.520831823348999,
      "eval_runtime": 4.3985,
      "eval_samples_per_second": 1.819,
      "eval_steps_per_second": 0.455,
      "step": 6
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.1504602432250977,
      "learning_rate": 0.00021818181818181816,
      "loss": 2.3422,
      "step": 8
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.325821876525879,
      "eval_runtime": 4.398,
      "eval_samples_per_second": 1.819,
      "eval_steps_per_second": 0.455,
      "step": 8
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.3659000396728516,
      "learning_rate": 0.0001909090909090909,
      "loss": 2.1455,
      "step": 10
    },
    {
      "epoch": 5.0,
      "eval_loss": 2.165768623352051,
      "eval_runtime": 4.3935,
      "eval_samples_per_second": 1.821,
      "eval_steps_per_second": 0.455,
      "step": 10
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.212636470794678,
      "learning_rate": 0.0001636363636363636,
      "loss": 1.9805,
      "step": 12
    },
    {
      "epoch": 6.0,
      "eval_loss": 2.0470268726348877,
      "eval_runtime": 4.4026,
      "eval_samples_per_second": 1.817,
      "eval_steps_per_second": 0.454,
      "step": 12
    },
    {
      "epoch": 7.0,
      "grad_norm": 4.396366596221924,
      "learning_rate": 0.00013636363636363634,
      "loss": 1.8691,
      "step": 14
    },
    {
      "epoch": 7.0,
      "eval_loss": 1.9581711292266846,
      "eval_runtime": 4.3935,
      "eval_samples_per_second": 1.821,
      "eval_steps_per_second": 0.455,
      "step": 14
    },
    {
      "epoch": 8.0,
      "grad_norm": 6.510736465454102,
      "learning_rate": 0.00010909090909090908,
      "loss": 1.7738,
      "step": 16
    },
    {
      "epoch": 8.0,
      "eval_loss": 1.8912162780761719,
      "eval_runtime": 4.4091,
      "eval_samples_per_second": 1.814,
      "eval_steps_per_second": 0.454,
      "step": 16
    },
    {
      "epoch": 9.0,
      "grad_norm": Infinity,
      "learning_rate": 9.545454545454545e-05,
      "loss": 1.7041,
      "step": 18
    },
    {
      "epoch": 9.0,
      "eval_loss": 1.8643579483032227,
      "eval_runtime": 4.408,
      "eval_samples_per_second": 1.815,
      "eval_steps_per_second": 0.454,
      "step": 18
    },
    {
      "epoch": 10.0,
      "grad_norm": 10.343268394470215,
      "learning_rate": 6.818181818181817e-05,
      "loss": 1.6787,
      "step": 20
    },
    {
      "epoch": 10.0,
      "eval_loss": 1.8231159448623657,
      "eval_runtime": 4.4034,
      "eval_samples_per_second": 1.817,
      "eval_steps_per_second": 0.454,
      "step": 20
    },
    {
      "epoch": 11.0,
      "grad_norm": 2.1240291595458984,
      "learning_rate": 4.09090909090909e-05,
      "loss": 1.6322,
      "step": 22
    },
    {
      "epoch": 11.0,
      "eval_loss": 1.7907209396362305,
      "eval_runtime": 4.4035,
      "eval_samples_per_second": 1.817,
      "eval_steps_per_second": 0.454,
      "step": 22
    },
    {
      "epoch": 12.0,
      "grad_norm": 4.139611721038818,
      "learning_rate": 1.3636363636363635e-05,
      "loss": 1.6007,
      "step": 24
    },
    {
      "epoch": 12.0,
      "eval_loss": 1.7682936191558838,
      "eval_runtime": 4.397,
      "eval_samples_per_second": 1.819,
      "eval_steps_per_second": 0.455,
      "step": 24
    }
  ],
  "logging_steps": 500,
  "max_steps": 24,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 12,
  "save_steps": 500,
  "total_flos": 105911406821376.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}