trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 53,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "debug/losses": 0.17192834615707397,
      "debug/policy_weights": 0.24804016947746277,
      "debug/raw_losses": 0.6931471824645996,
      "epoch": 0.018867924528301886,
      "grad_norm": 3.078622153428569,
      "learning_rate": 8.333333333333333e-08,
      "logits/chosen": -2.855412006378174,
      "logits/rejected": -2.8797199726104736,
      "logps/chosen": -320.43853759765625,
      "logps/rejected": -340.07073974609375,
      "loss": 0.2116,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "debug/losses": 0.20365437865257263,
      "debug/policy_weights": 0.2948223948478699,
      "debug/raw_losses": 0.6906725764274597,
      "epoch": 0.18867924528301888,
      "grad_norm": 3.2072554177174166,
      "learning_rate": 4.911172937635942e-07,
      "logits/chosen": -2.8659250736236572,
      "logits/rejected": -2.8918118476867676,
      "logps/chosen": -305.7447509765625,
      "logps/rejected": -332.24249267578125,
      "loss": 0.2033,
      "rewards/accuracies": 0.5416666865348816,
      "rewards/chosen": 0.0007928035338409245,
      "rewards/margins": 0.005046091042459011,
      "rewards/rejected": -0.004253287799656391,
      "step": 10
    },
    {
      "debug/losses": 0.2035388946533203,
      "debug/policy_weights": 0.3013075888156891,
      "debug/raw_losses": 0.6750057935714722,
      "epoch": 0.37735849056603776,
      "grad_norm": 3.0690801879183462,
      "learning_rate": 3.982949361823388e-07,
      "logits/chosen": -2.8558077812194824,
      "logits/rejected": -2.85764479637146,
      "logps/chosen": -324.26263427734375,
      "logps/rejected": -331.02935791015625,
      "loss": 0.2069,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": 0.0035451077856123447,
      "rewards/margins": 0.04073050990700722,
      "rewards/rejected": -0.03718540072441101,
      "step": 20
    },
    {
      "debug/losses": 0.20855382084846497,
      "debug/policy_weights": 0.32070785760879517,
      "debug/raw_losses": 0.6490412354469299,
      "epoch": 0.5660377358490566,
      "grad_norm": 3.2770260980043857,
      "learning_rate": 2.416462557480814e-07,
      "logits/chosen": -2.8393235206604004,
      "logits/rejected": -2.84259033203125,
      "logps/chosen": -298.1471252441406,
      "logps/rejected": -313.35174560546875,
      "loss": 0.2011,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.013330144807696342,
      "rewards/margins": 0.11589495837688446,
      "rewards/rejected": -0.12922510504722595,
      "step": 30
    },
    {
      "debug/losses": 0.18307599425315857,
      "debug/policy_weights": 0.303046315908432,
      "debug/raw_losses": 0.5849612951278687,
      "epoch": 0.7547169811320755,
      "grad_norm": 2.716351513439798,
      "learning_rate": 8.859303711029939e-08,
      "logits/chosen": -2.8420355319976807,
      "logits/rejected": -2.839024543762207,
      "logps/chosen": -290.5810852050781,
      "logps/rejected": -320.720458984375,
      "loss": 0.1967,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": -0.010476941242814064,
      "rewards/margins": 0.30531758069992065,
      "rewards/rejected": -0.31579452753067017,
      "step": 40
    },
    {
      "debug/losses": 0.2313450276851654,
      "debug/policy_weights": 0.35370975732803345,
      "debug/raw_losses": 0.647036612033844,
      "epoch": 0.9433962264150944,
      "grad_norm": 3.007767493180951,
      "learning_rate": 5.009573740853313e-09,
      "logits/chosen": -2.8769078254699707,
      "logits/rejected": -2.8638930320739746,
      "logps/chosen": -282.02764892578125,
      "logps/rejected": -316.0559997558594,
      "loss": 0.1938,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": -0.026972616091370583,
      "rewards/margins": 0.15525773167610168,
      "rewards/rejected": -0.18223035335540771,
      "step": 50
    },
    {
      "epoch": 1.0,
      "step": 53,
      "total_flos": 0.0,
      "train_loss": 0.20106691993632406,
      "train_runtime": 384.3035,
      "train_samples_per_second": 17.564,
      "train_steps_per_second": 0.138
    }
  ],
  "logging_steps": 10,
  "max_steps": 53,
  "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": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}