checkpoint-7089/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9983086680761097,
  "eval_steps": 1000,
  "global_step": 7089,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00042319085907744394,
      "grad_norm": 6.409224510192871,
      "learning_rate": 0.0,
      "loss": 2.8813,
      "step": 1
    },
    {
      "epoch": 0.4231908590774439,
      "grad_norm": 5.12570333480835,
      "learning_rate": 2.8636363636363637e-05,
      "loss": 2.153,
      "step": 1000
    },
    {
      "epoch": 0.4231908590774439,
      "eval_cosine_accuracy": 0.9628772735595703,
      "eval_loss": 0.4968035817146301,
      "eval_runtime": 35.9081,
      "eval_samples_per_second": 264.815,
      "eval_steps_per_second": 1.058,
      "step": 1000
    },
    {
      "epoch": 0.8463817181548878,
      "grad_norm": 7.6122307777404785,
      "learning_rate": 2.3934169278996865e-05,
      "loss": 1.1564,
      "step": 2000
    },
    {
      "epoch": 0.8463817181548878,
      "eval_cosine_accuracy": 0.9654011726379395,
      "eval_loss": 0.45379361510276794,
      "eval_runtime": 33.9596,
      "eval_samples_per_second": 280.009,
      "eval_steps_per_second": 1.119,
      "step": 2000
    },
    {
      "epoch": 1.2693446088794926,
      "grad_norm": 5.791377067565918,
      "learning_rate": 1.9231974921630092e-05,
      "loss": 1.0056,
      "step": 3000
    },
    {
      "epoch": 1.2693446088794926,
      "eval_cosine_accuracy": 0.9682406187057495,
      "eval_loss": 0.43380847573280334,
      "eval_runtime": 34.2533,
      "eval_samples_per_second": 277.608,
      "eval_steps_per_second": 1.109,
      "step": 3000
    },
    {
      "epoch": 1.6921775898520086,
      "grad_norm": 4.054314613342285,
      "learning_rate": 1.4529780564263322e-05,
      "loss": 1.0033,
      "step": 4000
    },
    {
      "epoch": 1.6921775898520086,
      "eval_cosine_accuracy": 0.9717110395431519,
      "eval_loss": 0.42844995856285095,
      "eval_runtime": 34.1625,
      "eval_samples_per_second": 278.347,
      "eval_steps_per_second": 1.112,
      "step": 4000
    },
    {
      "epoch": 2.1150105708245244,
      "grad_norm": 5.430419445037842,
      "learning_rate": 9.832288401253918e-06,
      "loss": 0.9272,
      "step": 5000
    },
    {
      "epoch": 2.1150105708245244,
      "eval_cosine_accuracy": 0.9711852073669434,
      "eval_loss": 0.425130695104599,
      "eval_runtime": 35.1451,
      "eval_samples_per_second": 270.564,
      "eval_steps_per_second": 1.081,
      "step": 5000
    },
    {
      "epoch": 2.53784355179704,
      "grad_norm": 5.48056173324585,
      "learning_rate": 5.1300940438871475e-06,
      "loss": 0.864,
      "step": 6000
    },
    {
      "epoch": 2.53784355179704,
      "eval_cosine_accuracy": 0.9712903499603271,
      "eval_loss": 0.4191531836986542,
      "eval_runtime": 34.5399,
      "eval_samples_per_second": 275.305,
      "eval_steps_per_second": 1.1,
      "step": 6000
    },
    {
      "epoch": 2.960676532769556,
      "grad_norm": 5.734561920166016,
      "learning_rate": 4.3260188087774296e-07,
      "loss": 0.853,
      "step": 7000
    },
    {
      "epoch": 2.960676532769556,
      "eval_cosine_accuracy": 0.9717110395431519,
      "eval_loss": 0.4148150384426117,
      "eval_runtime": 34.4631,
      "eval_samples_per_second": 275.919,
      "eval_steps_per_second": 1.103,
      "step": 7000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 7089,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 0.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}