checkpoint-933/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 10,
  "global_step": 933,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.4457938075065613,
      "learning_rate": 0.0,
      "loss": 0.0601,
      "step": 933
    },
    {
      "epoch": 1.0,
      "eval_custom-arc-semantics-data-jp_cosine_accuracy": 0.9705409748259239,
      "eval_custom-arc-semantics-data-jp_cosine_accuracy_threshold": 0.5118279457092285,
      "eval_custom-arc-semantics-data-jp_cosine_ap": 0.9236294738598163,
      "eval_custom-arc-semantics-data-jp_cosine_f1": 0.9102773246329527,
      "eval_custom-arc-semantics-data-jp_cosine_f1_threshold": 0.45031607151031494,
      "eval_custom-arc-semantics-data-jp_cosine_precision": 0.8857142857142857,
      "eval_custom-arc-semantics-data-jp_cosine_recall": 0.9362416107382551,
      "eval_custom-arc-semantics-data-jp_dot_accuracy": 0.9694697375468666,
      "eval_custom-arc-semantics-data-jp_dot_accuracy_threshold": 251.2455596923828,
      "eval_custom-arc-semantics-data-jp_dot_ap": 0.9196731890884118,
      "eval_custom-arc-semantics-data-jp_dot_f1": 0.9060955518945634,
      "eval_custom-arc-semantics-data-jp_dot_f1_threshold": 246.36648559570312,
      "eval_custom-arc-semantics-data-jp_dot_precision": 0.889967637540453,
      "eval_custom-arc-semantics-data-jp_dot_recall": 0.9228187919463087,
      "eval_custom-arc-semantics-data-jp_euclidean_accuracy": 0.9721478307445099,
      "eval_custom-arc-semantics-data-jp_euclidean_accuracy_threshold": 23.195274353027344,
      "eval_custom-arc-semantics-data-jp_euclidean_ap": 0.9259440018381992,
      "eval_custom-arc-semantics-data-jp_euclidean_f1": 0.9147540983606558,
      "eval_custom-arc-semantics-data-jp_euclidean_f1_threshold": 23.195274353027344,
      "eval_custom-arc-semantics-data-jp_euclidean_precision": 0.8942307692307693,
      "eval_custom-arc-semantics-data-jp_euclidean_recall": 0.9362416107382551,
      "eval_custom-arc-semantics-data-jp_manhattan_accuracy": 0.9716122121049813,
      "eval_custom-arc-semantics-data-jp_manhattan_accuracy_threshold": 514.571533203125,
      "eval_custom-arc-semantics-data-jp_manhattan_ap": 0.9255015709844487,
      "eval_custom-arc-semantics-data-jp_manhattan_f1": 0.9132569558101473,
      "eval_custom-arc-semantics-data-jp_manhattan_f1_threshold": 514.571533203125,
      "eval_custom-arc-semantics-data-jp_manhattan_precision": 0.8913738019169329,
      "eval_custom-arc-semantics-data-jp_manhattan_recall": 0.9362416107382551,
      "eval_custom-arc-semantics-data-jp_max_accuracy": 0.9721478307445099,
      "eval_custom-arc-semantics-data-jp_max_accuracy_threshold": 514.571533203125,
      "eval_custom-arc-semantics-data-jp_max_ap": 0.9259440018381992,
      "eval_custom-arc-semantics-data-jp_max_f1": 0.9147540983606558,
      "eval_custom-arc-semantics-data-jp_max_f1_threshold": 514.571533203125,
      "eval_custom-arc-semantics-data-jp_max_precision": 0.8942307692307693,
      "eval_custom-arc-semantics-data-jp_max_recall": 0.9362416107382551,
      "eval_loss": 0.030283836647868156,
      "eval_runtime": 238.5188,
      "eval_samples_per_second": 7.827,
      "eval_steps_per_second": 0.981,
      "step": 933
    }
  ],
  "logging_steps": 500,
  "max_steps": 933,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}