checkpoint-657/trainer_state.json

{
  "best_metric": 0.04255302622914314,
  "best_model_checkpoint": "./teapotllm/checkpoint-657",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 657,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "em_boolean": 0.5294117647058824,
      "em_chat": 0.03783783783783784,
      "em_extraction": 0.5333333333333333,
      "em_qa": 0.5333333333333333,
      "em_summarization": 0.0,
      "em_unanswerable": 0.0,
      "epoch": 1.0,
      "mean_em": 0.12075471698113208,
      "mean_similarity": 0.6624939311014594,
      "mean_word_count_diff": 11.958490566037735,
      "similarity_boolean": 0.649950959993636,
      "similarity_chat": 0.6068178282194846,
      "similarity_extraction": 0.8274670541286469,
      "similarity_qa": 0.8895131280024846,
      "similarity_summarization": 0.7897715005609725,
      "similarity_unanswerable": 0.8186558306217193,
      "word_count_diff_boolean": 2.764705882352941,
      "word_count_diff_chat": 11.275675675675675,
      "word_count_diff_extraction": 1.8666666666666667,
      "word_count_diff_qa": 2.7333333333333334,
      "word_count_diff_summarization": 50.05555555555556,
      "word_count_diff_unanswerable": 4.4
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.29226627945899963,
      "learning_rate": 1.314e-05,
      "loss": 0.0548,
      "step": 219
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.04732182249426842,
      "eval_runtime": 17.2425,
      "eval_samples_per_second": 15.369,
      "eval_steps_per_second": 1.972,
      "step": 219
    },
    {
      "em_boolean": 0.7058823529411765,
      "em_chat": 0.05405405405405406,
      "em_extraction": 0.4,
      "em_qa": 0.4,
      "em_summarization": 0.0,
      "em_unanswerable": 0.0,
      "epoch": 2.0,
      "mean_em": 0.12830188679245283,
      "mean_similarity": 0.6936857629426808,
      "mean_word_count_diff": 11.475471698113207,
      "similarity_boolean": 0.7671171505661571,
      "similarity_chat": 0.6481531772158435,
      "similarity_extraction": 0.7221033732096355,
      "similarity_qa": 0.8645438591639201,
      "similarity_summarization": 0.824706514676412,
      "similarity_unanswerable": 0.8155314723650614,
      "word_count_diff_boolean": 2.7058823529411766,
      "word_count_diff_chat": 10.556756756756757,
      "word_count_diff_extraction": 2.933333333333333,
      "word_count_diff_qa": 3.2,
      "word_count_diff_summarization": 50.166666666666664,
      "word_count_diff_unanswerable": 3.1333333333333333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.30153563618659973,
      "learning_rate": 2.628e-05,
      "loss": 0.0445,
      "step": 438
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.04412226751446724,
      "eval_runtime": 17.2255,
      "eval_samples_per_second": 15.384,
      "eval_steps_per_second": 1.974,
      "step": 438
    },
    {
      "em_boolean": 0.5882352941176471,
      "em_chat": 0.05945945945945946,
      "em_extraction": 0.4666666666666667,
      "em_qa": 0.5333333333333333,
      "em_summarization": 0.0,
      "em_unanswerable": 0.0,
      "epoch": 3.0,
      "mean_em": 0.13584905660377358,
      "mean_similarity": 0.7036299928109039,
      "mean_word_count_diff": 11.230188679245282,
      "similarity_boolean": 0.702288385699777,
      "similarity_chat": 0.6661249467549292,
      "similarity_extraction": 0.7363929619391759,
      "similarity_qa": 0.9095388889312744,
      "similarity_summarization": 0.7749250100718604,
      "similarity_unanswerable": 0.8434868295987447,
      "word_count_diff_boolean": 1.8823529411764706,
      "word_count_diff_chat": 10.41081081081081,
      "word_count_diff_extraction": 1.4,
      "word_count_diff_qa": 2.8666666666666667,
      "word_count_diff_summarization": 50.111111111111114,
      "word_count_diff_unanswerable": 3.466666666666667
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4190770983695984,
      "learning_rate": 2.9365689308796065e-05,
      "loss": 0.0366,
      "step": 657
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.04255302622914314,
      "eval_runtime": 17.2484,
      "eval_samples_per_second": 15.364,
      "eval_steps_per_second": 1.971,
      "step": 657
    }
  ],
  "logging_steps": 500,
  "max_steps": 2190,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "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": 6036187263270912.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}