last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.005925925925925926,
  "eval_steps": 500,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005925925925925926,
      "grad_norm": 10.913461685180664,
      "learning_rate": 5e-07,
      "logits/chosen": -4.444676399230957,
      "logits/rejected": -4.0909342765808105,
      "logps/chosen": -186.1875,
      "logps/rejected": -228.68560791015625,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.0011851851851851852,
      "grad_norm": 7.564138412475586,
      "learning_rate": 4.849231551964771e-07,
      "logits/chosen": -4.530362606048584,
      "logits/rejected": -3.981240749359131,
      "logps/chosen": -192.2100830078125,
      "logps/rejected": -213.73086547851562,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 2
    },
    {
      "epoch": 0.0017777777777777779,
      "grad_norm": 14.393203735351562,
      "learning_rate": 4.415111107797445e-07,
      "logits/chosen": -4.539660930633545,
      "logits/rejected": -4.32346248626709,
      "logps/chosen": -200.69912719726562,
      "logps/rejected": -224.77163696289062,
      "loss": 0.6526,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.04995880275964737,
      "rewards/margins": 0.08375511318445206,
      "rewards/rejected": -0.03379631042480469,
      "step": 3
    },
    {
      "epoch": 0.0023703703703703703,
      "grad_norm": 12.62684440612793,
      "learning_rate": 3.75e-07,
      "logits/chosen": -3.9760360717773438,
      "logits/rejected": -4.4724321365356445,
      "logps/chosen": -239.9991455078125,
      "logps/rejected": -202.63072204589844,
      "loss": 0.7198,
      "rewards/accuracies": 0.0,
      "rewards/chosen": -0.041876986622810364,
      "rewards/margins": -0.052445217967033386,
      "rewards/rejected": 0.01056823693215847,
      "step": 4
    },
    {
      "epoch": 0.002962962962962963,
      "grad_norm": 6.745312213897705,
      "learning_rate": 2.934120444167326e-07,
      "logits/chosen": -4.136646270751953,
      "logits/rejected": -4.701557159423828,
      "logps/chosen": -169.00192260742188,
      "logps/rejected": -149.2642059326172,
      "loss": 0.6818,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.02975158765912056,
      "rewards/margins": 0.023634720593690872,
      "rewards/rejected": 0.006116867531090975,
      "step": 5
    },
    {
      "epoch": 0.0035555555555555557,
      "grad_norm": 13.741002082824707,
      "learning_rate": 2.065879555832674e-07,
      "logits/chosen": -5.057063579559326,
      "logits/rejected": -5.147495746612549,
      "logps/chosen": -193.6820526123047,
      "logps/rejected": -202.20425415039062,
      "loss": 0.6913,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.0083160400390625,
      "rewards/margins": 0.004281995818018913,
      "rewards/rejected": 0.0040340423583984375,
      "step": 6
    },
    {
      "epoch": 0.004148148148148148,
      "grad_norm": 6.920108795166016,
      "learning_rate": 1.2500000000000005e-07,
      "logits/chosen": -4.2448506355285645,
      "logits/rejected": -3.682173490524292,
      "logps/chosen": -176.11734008789062,
      "logps/rejected": -222.982666015625,
      "loss": 0.7008,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.028499603271484375,
      "rewards/margins": -0.015093998052179813,
      "rewards/rejected": 0.043593600392341614,
      "step": 7
    },
    {
      "epoch": 0.004740740740740741,
      "grad_norm": 10.856575965881348,
      "learning_rate": 5.848888922025552e-08,
      "logits/chosen": -4.413946151733398,
      "logits/rejected": -4.419940948486328,
      "logps/chosen": -181.99610900878906,
      "logps/rejected": -196.20811462402344,
      "loss": 0.6967,
      "rewards/accuracies": 0.25,
      "rewards/chosen": -0.02880115620791912,
      "rewards/margins": -0.007002831436693668,
      "rewards/rejected": -0.021798323839902878,
      "step": 8
    },
    {
      "epoch": 0.005333333333333333,
      "grad_norm": 9.559263229370117,
      "learning_rate": 1.507684480352292e-08,
      "logits/chosen": -4.909343242645264,
      "logits/rejected": -4.601991176605225,
      "logps/chosen": -154.6862030029297,
      "logps/rejected": -179.27401733398438,
      "loss": 0.6929,
      "rewards/accuracies": 0.25,
      "rewards/chosen": 0.0032157916575670242,
      "rewards/margins": 0.0011837054044008255,
      "rewards/rejected": 0.0020320871844887733,
      "step": 9
    },
    {
      "epoch": 0.005925925925925926,
      "grad_norm": 10.816996574401855,
      "learning_rate": 0.0,
      "logits/chosen": -3.8037424087524414,
      "logits/rejected": -3.683922529220581,
      "logps/chosen": -232.978515625,
      "logps/rejected": -254.24078369140625,
      "loss": 0.7146,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.05405044183135033,
      "rewards/margins": -0.040559008717536926,
      "rewards/rejected": -0.013491439633071423,
      "step": 10
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 1,
  "trial_name": null,
  "trial_params": null
}