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": 10.896183013916016,
      "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": 7.562353610992432,
      "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": 13.934762001037598,
      "learning_rate": 4.415111107797445e-07,
      "logits/chosen": -4.541503429412842,
      "logits/rejected": -4.324827194213867,
      "logps/chosen": -200.9736328125,
      "logps/rejected": -224.6981658935547,
      "loss": 0.6691,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.022507095709443092,
      "rewards/margins": 0.04895630106329918,
      "rewards/rejected": -0.026449203491210938,
      "step": 3
    },
    {
      "epoch": 0.0023703703703703703,
      "grad_norm": 13.249699592590332,
      "learning_rate": 3.75e-07,
      "logits/chosen": -3.9642434120178223,
      "logits/rejected": -4.461881160736084,
      "logps/chosen": -239.5960693359375,
      "logps/rejected": -202.10887145996094,
      "loss": 0.7259,
      "rewards/accuracies": 0.0,
      "rewards/chosen": -0.001570509746670723,
      "rewards/margins": -0.06432266533374786,
      "rewards/rejected": 0.06275215744972229,
      "step": 4
    },
    {
      "epoch": 0.002962962962962963,
      "grad_norm": 6.773025989532471,
      "learning_rate": 2.934120444167326e-07,
      "logits/chosen": -4.136728286743164,
      "logits/rejected": -4.700671672821045,
      "logps/chosen": -169.1282958984375,
      "logps/rejected": -149.28570556640625,
      "loss": 0.6866,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.017114639282226562,
      "rewards/margins": 0.013147544115781784,
      "rewards/rejected": 0.003967094700783491,
      "step": 5
    },
    {
      "epoch": 0.0035555555555555557,
      "grad_norm": 13.418476104736328,
      "learning_rate": 2.065879555832674e-07,
      "logits/chosen": -5.0562872886657715,
      "logits/rejected": -5.145793914794922,
      "logps/chosen": -194.10568237304688,
      "logps/rejected": -202.68751525878906,
      "loss": 0.6884,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.034047700464725494,
      "rewards/margins": 0.01024474948644638,
      "rewards/rejected": -0.044292449951171875,
      "step": 6
    },
    {
      "epoch": 0.004148148148148148,
      "grad_norm": 6.973136901855469,
      "learning_rate": 1.2500000000000005e-07,
      "logits/chosen": -4.244935035705566,
      "logits/rejected": -3.681328058242798,
      "logps/chosen": -176.23780822753906,
      "logps/rejected": -222.976318359375,
      "loss": 0.7073,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.016452789306640625,
      "rewards/margins": -0.027773665264248848,
      "rewards/rejected": 0.044226452708244324,
      "step": 7
    },
    {
      "epoch": 0.004740740740740741,
      "grad_norm": 10.38602066040039,
      "learning_rate": 5.848888922025552e-08,
      "logits/chosen": -4.409693241119385,
      "logits/rejected": -4.41237735748291,
      "logps/chosen": -181.64944458007812,
      "logps/rejected": -196.00564575195312,
      "loss": 0.6896,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.005864143371582031,
      "rewards/margins": 0.007416536100208759,
      "rewards/rejected": -0.001552392728626728,
      "step": 8
    },
    {
      "epoch": 0.005333333333333333,
      "grad_norm": 9.169404983520508,
      "learning_rate": 1.507684480352292e-08,
      "logits/chosen": -4.903069972991943,
      "logits/rejected": -4.5977277755737305,
      "logps/chosen": -154.88751220703125,
      "logps/rejected": -179.42715454101562,
      "loss": 0.695,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.016914749518036842,
      "rewards/margins": -0.003633500775322318,
      "rewards/rejected": -0.013281249441206455,
      "step": 9
    },
    {
      "epoch": 0.005925925925925926,
      "grad_norm": 10.337571144104004,
      "learning_rate": 0.0,
      "logits/chosen": -3.817504644393921,
      "logits/rejected": -3.697187900543213,
      "logps/chosen": -233.053466796875,
      "logps/rejected": -254.43130493164062,
      "loss": 0.7085,
      "rewards/accuracies": 0.5,
      "rewards/chosen": -0.06154556944966316,
      "rewards/margins": -0.029002761468291283,
      "rewards/rejected": -0.03254280239343643,
      "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
}