checkpoint-277/trainer_state.json

{
  "best_metric": 0.887685239315033,
  "best_model_checkpoint": "outputs/checkpoint-185",
  "epoch": 5.98110661268556,
  "eval_steps": 500,
  "global_step": 277,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4318488529014845,
      "grad_norm": 0.5924658179283142,
      "learning_rate": 4e-05,
      "loss": 1.9539,
      "step": 20
    },
    {
      "epoch": 0.863697705802969,
      "grad_norm": 0.495579332113266,
      "learning_rate": 8e-05,
      "loss": 1.8186,
      "step": 40
    },
    {
      "epoch": 0.9932523616734144,
      "eval_loss": 1.5013270378112793,
      "eval_runtime": 13.4628,
      "eval_samples_per_second": 27.632,
      "eval_steps_per_second": 3.491,
      "step": 46
    },
    {
      "epoch": 1.2955465587044535,
      "grad_norm": 0.6942113041877747,
      "learning_rate": 0.00012,
      "loss": 1.6882,
      "step": 60
    },
    {
      "epoch": 1.7273954116059378,
      "grad_norm": 0.8643310070037842,
      "learning_rate": 0.00016,
      "loss": 1.4623,
      "step": 80
    },
    {
      "epoch": 1.9865047233468287,
      "eval_loss": 1.1028813123703003,
      "eval_runtime": 13.4743,
      "eval_samples_per_second": 27.608,
      "eval_steps_per_second": 3.488,
      "step": 92
    },
    {
      "epoch": 2.1592442645074224,
      "grad_norm": 1.187238097190857,
      "learning_rate": 0.0002,
      "loss": 1.3448,
      "step": 100
    },
    {
      "epoch": 2.591093117408907,
      "grad_norm": 1.2552473545074463,
      "learning_rate": 0.00019943348002101371,
      "loss": 1.1681,
      "step": 120
    },
    {
      "epoch": 2.979757085020243,
      "eval_loss": 0.9362145066261292,
      "eval_runtime": 13.4657,
      "eval_samples_per_second": 27.626,
      "eval_steps_per_second": 3.49,
      "step": 138
    },
    {
      "epoch": 3.0229419703103915,
      "grad_norm": 1.0589548349380493,
      "learning_rate": 0.00019774033898178667,
      "loss": 1.1426,
      "step": 140
    },
    {
      "epoch": 3.454790823211876,
      "grad_norm": 1.1629831790924072,
      "learning_rate": 0.00019493976084683813,
      "loss": 0.9296,
      "step": 160
    },
    {
      "epoch": 3.8866396761133606,
      "grad_norm": 0.9937772154808044,
      "learning_rate": 0.00019106347728549135,
      "loss": 0.9304,
      "step": 180
    },
    {
      "epoch": 3.9946018893387314,
      "eval_loss": 0.887685239315033,
      "eval_runtime": 13.4691,
      "eval_samples_per_second": 27.619,
      "eval_steps_per_second": 3.489,
      "step": 185
    },
    {
      "epoch": 4.318488529014845,
      "grad_norm": 1.299687385559082,
      "learning_rate": 0.0001861554081393806,
      "loss": 0.7665,
      "step": 200
    },
    {
      "epoch": 4.75033738191633,
      "grad_norm": 1.375931978225708,
      "learning_rate": 0.00018027116379309638,
      "loss": 0.7884,
      "step": 220
    },
    {
      "epoch": 4.987854251012146,
      "eval_loss": 0.8908663392066956,
      "eval_runtime": 13.466,
      "eval_samples_per_second": 27.625,
      "eval_steps_per_second": 3.49,
      "step": 231
    },
    {
      "epoch": 5.182186234817814,
      "grad_norm": 1.134634256362915,
      "learning_rate": 0.00017347741508630672,
      "loss": 0.7193,
      "step": 240
    },
    {
      "epoch": 5.614035087719298,
      "grad_norm": 1.0707848072052002,
      "learning_rate": 0.00016585113790650388,
      "loss": 0.6113,
      "step": 260
    },
    {
      "epoch": 5.98110661268556,
      "eval_loss": 0.9561511874198914,
      "eval_runtime": 13.4701,
      "eval_samples_per_second": 27.617,
      "eval_steps_per_second": 3.489,
      "step": 277
    }
  ],
  "logging_steps": 20,
  "max_steps": 690,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 2.134024227305472e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}