trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 8336,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.2399232245681382,
      "grad_norm": 2.465574026107788,
      "learning_rate": 2.8200575815738963e-05,
      "loss": 1.1212,
      "step": 500
    },
    {
      "epoch": 0.4798464491362764,
      "grad_norm": 3.261763572692871,
      "learning_rate": 2.640115163147793e-05,
      "loss": 0.9721,
      "step": 1000
    },
    {
      "epoch": 0.7197696737044146,
      "grad_norm": 3.3122525215148926,
      "learning_rate": 2.4601727447216888e-05,
      "loss": 0.9282,
      "step": 1500
    },
    {
      "epoch": 0.9596928982725528,
      "grad_norm": 2.439178466796875,
      "learning_rate": 2.2802303262955854e-05,
      "loss": 0.9108,
      "step": 2000
    },
    {
      "epoch": 1.199616122840691,
      "grad_norm": 2.484846830368042,
      "learning_rate": 2.1002879078694817e-05,
      "loss": 0.8742,
      "step": 2500
    },
    {
      "epoch": 1.4395393474088292,
      "grad_norm": 3.605393409729004,
      "learning_rate": 1.9203454894433783e-05,
      "loss": 0.8873,
      "step": 3000
    },
    {
      "epoch": 1.6794625719769674,
      "grad_norm": 2.1433281898498535,
      "learning_rate": 1.7404030710172745e-05,
      "loss": 0.8618,
      "step": 3500
    },
    {
      "epoch": 1.9193857965451055,
      "grad_norm": 3.013890504837036,
      "learning_rate": 1.5604606525911708e-05,
      "loss": 0.8517,
      "step": 4000
    },
    {
      "epoch": 2.159309021113244,
      "grad_norm": 3.020458936691284,
      "learning_rate": 1.3805182341650672e-05,
      "loss": 0.8473,
      "step": 4500
    },
    {
      "epoch": 2.399232245681382,
      "grad_norm": 2.4140872955322266,
      "learning_rate": 1.2005758157389636e-05,
      "loss": 0.8312,
      "step": 5000
    },
    {
      "epoch": 2.63915547024952,
      "grad_norm": 2.538238763809204,
      "learning_rate": 1.0206333973128599e-05,
      "loss": 0.8291,
      "step": 5500
    },
    {
      "epoch": 2.8790786948176583,
      "grad_norm": 2.649840831756592,
      "learning_rate": 8.406909788867563e-06,
      "loss": 0.8218,
      "step": 6000
    },
    {
      "epoch": 3.1190019193857967,
      "grad_norm": 2.9255142211914062,
      "learning_rate": 6.607485604606526e-06,
      "loss": 0.8212,
      "step": 6500
    },
    {
      "epoch": 3.3589251439539347,
      "grad_norm": 2.9272096157073975,
      "learning_rate": 4.8080614203454895e-06,
      "loss": 0.8067,
      "step": 7000
    },
    {
      "epoch": 3.5988483685220727,
      "grad_norm": 3.7314858436584473,
      "learning_rate": 3.008637236084453e-06,
      "loss": 0.818,
      "step": 7500
    },
    {
      "epoch": 3.838771593090211,
      "grad_norm": 2.498305320739746,
      "learning_rate": 1.2092130518234166e-06,
      "loss": 0.8026,
      "step": 8000
    },
    {
      "epoch": 4.0,
      "step": 8336,
      "total_flos": 5.57671514112e+16,
      "train_loss": 0.8718166570974631,
      "train_runtime": 11124.9814,
      "train_samples_per_second": 35.955,
      "train_steps_per_second": 0.749
    }
  ],
  "logging_steps": 500,
  "max_steps": 8336,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.57671514112e+16,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": null
}