{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.971563981042654,
  "eval_steps": 100,
  "global_step": 104,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.018957345971563982,
      "grad_norm": 2420.545449757295,
      "learning_rate": 4.545454545454545e-08,
      "logits/chosen": 117.53560638427734,
      "logits/rejected": 126.8960952758789,
      "logps/chosen": -1.6956759691238403,
      "logps/rejected": -1.6029529571533203,
      "loss": 100.0,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.1895734597156398,
      "grad_norm": 3136.1061099122267,
      "learning_rate": 4.545454545454545e-07,
      "logits/chosen": 135.21646118164062,
      "logits/rejected": 138.5503387451172,
      "logps/chosen": -1.5280765295028687,
      "logps/rejected": -1.8057875633239746,
      "loss": 99.7805,
      "rewards/accuracies": 0.4305555522441864,
      "rewards/chosen": 0.0016786378109827638,
      "rewards/margins": 0.00015397430979646742,
      "rewards/rejected": 0.0015246635302901268,
      "step": 10
    },
    {
      "epoch": 0.3791469194312796,
      "grad_norm": 3277.7785757484266,
      "learning_rate": 4.885348141000122e-07,
      "logits/chosen": 123.2613296508789,
      "logits/rejected": 127.1202392578125,
      "logps/chosen": -1.318345546722412,
      "logps/rejected": -1.5371055603027344,
      "loss": 100.1618,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.008244398981332779,
      "rewards/margins": -0.0002138310082955286,
      "rewards/rejected": 0.008458229713141918,
      "step": 20
    },
    {
      "epoch": 0.5687203791469194,
      "grad_norm": 2601.9004683236003,
      "learning_rate": 4.5025027361734613e-07,
      "logits/chosen": 143.5557861328125,
      "logits/rejected": 137.32350158691406,
      "logps/chosen": -1.4400651454925537,
      "logps/rejected": -1.815298080444336,
      "loss": 98.0522,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -0.00597388157621026,
      "rewards/margins": 0.005939747206866741,
      "rewards/rejected": -0.011913629248738289,
      "step": 30
    },
    {
      "epoch": 0.7582938388625592,
      "grad_norm": 2796.822456458629,
      "learning_rate": 3.893311157806091e-07,
      "logits/chosen": 124.65242767333984,
      "logits/rejected": 114.0244369506836,
      "logps/chosen": -1.684523344039917,
      "logps/rejected": -2.111952304840088,
      "loss": 96.6206,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.007459363434463739,
      "rewards/margins": 0.01222879346460104,
      "rewards/rejected": -0.019688155502080917,
      "step": 40
    },
    {
      "epoch": 0.9478672985781991,
      "grad_norm": 7140.297491181953,
      "learning_rate": 3.126631330646801e-07,
      "logits/chosen": 132.700439453125,
      "logits/rejected": 137.55661010742188,
      "logps/chosen": -5.176327228546143,
      "logps/rejected": -6.27411413192749,
      "loss": 84.491,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": -0.18510575592517853,
      "rewards/margins": 0.05058497190475464,
      "rewards/rejected": -0.23569071292877197,
      "step": 50
    },
    {
      "epoch": 1.1374407582938388,
      "grad_norm": 7314.820924887574,
      "learning_rate": 2.2891223348923882e-07,
      "logits/chosen": 114.27351379394531,
      "logits/rejected": 123.59075927734375,
      "logps/chosen": -8.415227890014648,
      "logps/rejected": -11.62614917755127,
      "loss": 68.5077,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.34948182106018066,
      "rewards/margins": 0.15308642387390137,
      "rewards/rejected": -0.502568244934082,
      "step": 60
    },
    {
      "epoch": 1.3270142180094786,
      "grad_norm": 6037.368096578428,
      "learning_rate": 1.4754491880085317e-07,
      "logits/chosen": 111.22945404052734,
      "logits/rejected": 117.32829284667969,
      "logps/chosen": -10.453299522399902,
      "logps/rejected": -14.060336112976074,
      "loss": 62.8385,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": -0.4513101577758789,
      "rewards/margins": 0.16897699236869812,
      "rewards/rejected": -0.6202871203422546,
      "step": 70
    },
    {
      "epoch": 1.5165876777251186,
      "grad_norm": 8588.068447257592,
      "learning_rate": 7.775827023107834e-08,
      "logits/chosen": 118.58912658691406,
      "logits/rejected": 138.24412536621094,
      "logps/chosen": -14.044906616210938,
      "logps/rejected": -18.62037467956543,
      "loss": 59.2106,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": -0.6267557144165039,
      "rewards/margins": 0.21960142254829407,
      "rewards/rejected": -0.8463570475578308,
      "step": 80
    },
    {
      "epoch": 1.7061611374407581,
      "grad_norm": 7064.858481699249,
      "learning_rate": 2.7440387297912122e-08,
      "logits/chosen": 108.10774230957031,
      "logits/rejected": 126.4560546875,
      "logps/chosen": -12.219902038574219,
      "logps/rejected": -17.356382369995117,
      "loss": 55.2343,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": -0.5334357619285583,
      "rewards/margins": 0.25061604380607605,
      "rewards/rejected": -0.7840518355369568,
      "step": 90
    },
    {
      "epoch": 1.8957345971563981,
      "grad_norm": 6288.263108573529,
      "learning_rate": 2.27878296044029e-09,
      "logits/chosen": 109.19987487792969,
      "logits/rejected": 115.3917007446289,
      "logps/chosen": -11.196935653686523,
      "logps/rejected": -15.554553031921387,
      "loss": 54.5261,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": -0.48229581117630005,
      "rewards/margins": 0.21144536137580872,
      "rewards/rejected": -0.6937412023544312,
      "step": 100
    },
    {
      "epoch": 1.8957345971563981,
      "eval_logits/chosen": 99.54974365234375,
      "eval_logits/rejected": 99.7590560913086,
      "eval_logps/chosen": -11.464829444885254,
      "eval_logps/rejected": -15.369668006896973,
      "eval_loss": 60.862613677978516,
      "eval_rewards/accuracies": 0.8020833134651184,
      "eval_rewards/chosen": -0.5007163882255554,
      "eval_rewards/margins": 0.18991293013095856,
      "eval_rewards/rejected": -0.6906293034553528,
      "eval_runtime": 121.9617,
      "eval_samples_per_second": 6.149,
      "eval_steps_per_second": 0.197,
      "step": 100
    },
    {
      "epoch": 1.971563981042654,
      "step": 104,
      "total_flos": 0.0,
      "train_loss": 77.37614455589882,
      "train_runtime": 2312.9834,
      "train_samples_per_second": 5.837,
      "train_steps_per_second": 0.045
    }
  ],
  "logging_steps": 10,
  "max_steps": 104,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}