bert_mini_squadv2_finetuned/checkpoint-190/trainer_state.json

{
  "best_global_step": 190,
  "best_metric": 3.1472160816192627,
  "best_model_checkpoint": "./bert_mini_squadv2_finetuned/checkpoint-190",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 190,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 4.97021484375,
      "learning_rate": 1.9526315789473688e-05,
      "loss": 5.8799,
      "step": 10
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 5.308949947357178,
      "learning_rate": 1.9e-05,
      "loss": 5.6001,
      "step": 20
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 5.93297004699707,
      "learning_rate": 1.8473684210526317e-05,
      "loss": 5.356,
      "step": 30
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 6.4546613693237305,
      "learning_rate": 1.7947368421052634e-05,
      "loss": 5.0317,
      "step": 40
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 6.0348734855651855,
      "learning_rate": 1.742105263157895e-05,
      "loss": 4.8894,
      "step": 50
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 7.766897201538086,
      "learning_rate": 1.6894736842105263e-05,
      "loss": 4.5613,
      "step": 60
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 6.935646057128906,
      "learning_rate": 1.636842105263158e-05,
      "loss": 4.4883,
      "step": 70
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 6.82500696182251,
      "learning_rate": 1.5842105263157896e-05,
      "loss": 4.2883,
      "step": 80
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 7.159075736999512,
      "learning_rate": 1.5315789473684212e-05,
      "loss": 4.165,
      "step": 90
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 7.179778575897217,
      "learning_rate": 1.4789473684210527e-05,
      "loss": 4.1544,
      "step": 100
    },
    {
      "epoch": 0.5789473684210527,
      "grad_norm": 7.615407943725586,
      "learning_rate": 1.4263157894736843e-05,
      "loss": 3.9639,
      "step": 110
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 8.331777572631836,
      "learning_rate": 1.373684210526316e-05,
      "loss": 3.9313,
      "step": 120
    },
    {
      "epoch": 0.6842105263157895,
      "grad_norm": 8.408103942871094,
      "learning_rate": 1.3210526315789476e-05,
      "loss": 3.8429,
      "step": 130
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 7.173701286315918,
      "learning_rate": 1.268421052631579e-05,
      "loss": 3.8753,
      "step": 140
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 8.077372550964355,
      "learning_rate": 1.2157894736842107e-05,
      "loss": 3.726,
      "step": 150
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 8.297250747680664,
      "learning_rate": 1.1631578947368423e-05,
      "loss": 3.7247,
      "step": 160
    },
    {
      "epoch": 0.8947368421052632,
      "grad_norm": 8.952900886535645,
      "learning_rate": 1.1105263157894736e-05,
      "loss": 3.5831,
      "step": 170
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 7.744213581085205,
      "learning_rate": 1.0578947368421053e-05,
      "loss": 3.6806,
      "step": 180
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.072354316711426,
      "learning_rate": 1.005263157894737e-05,
      "loss": 3.6219,
      "step": 190
    },
    {
      "epoch": 1.0,
      "eval_loss": 3.1472160816192627,
      "eval_runtime": 33.3578,
      "eval_samples_per_second": 15.259,
      "eval_steps_per_second": 0.959,
      "step": 190
    }
  ],
  "logging_steps": 10,
  "max_steps": 380,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 22104484706304.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}