{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9960474308300395,
  "eval_steps": 100,
  "global_step": 126,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 3.6983913280928786,
      "learning_rate": 3.846153846153846e-08,
      "logits/chosen": 0.07181362062692642,
      "logits/rejected": -0.5422722101211548,
      "logps/chosen": -667.6278076171875,
      "logps/rejected": -895.8421020507812,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.097480178742397,
      "learning_rate": 3.8461538461538463e-07,
      "logits/chosen": 0.04356582462787628,
      "logits/rejected": -0.7095460295677185,
      "logps/chosen": -583.3101196289062,
      "logps/rejected": -874.3485107421875,
      "loss": 0.693,
      "rewards/accuracies": 0.4652777910232544,
      "rewards/chosen": 0.0002644425549078733,
      "rewards/margins": 0.0005916350055485964,
      "rewards/rejected": -0.000327192508848384,
      "step": 10
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.056926958975919,
      "learning_rate": 4.952806974561517e-07,
      "logits/chosen": 0.13229021430015564,
      "logits/rejected": -0.5041828751564026,
      "logps/chosen": -534.6011962890625,
      "logps/rejected": -831.9937744140625,
      "loss": 0.691,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": 0.0032804287038743496,
      "rewards/margins": 0.0038477883208543062,
      "rewards/rejected": -0.0005673595005646348,
      "step": 20
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.357718501847275,
      "learning_rate": 4.725936445085709e-07,
      "logits/chosen": -0.0280394796282053,
      "logits/rejected": -0.6336067914962769,
      "logps/chosen": -625.0661010742188,
      "logps/rejected": -936.4635620117188,
      "loss": 0.6823,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.012510508298873901,
      "rewards/margins": 0.018871963024139404,
      "rewards/rejected": -0.006361453328281641,
      "step": 30
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.108185931224436,
      "learning_rate": 4.328120888946271e-07,
      "logits/chosen": 0.0761369839310646,
      "logits/rejected": -0.8145635724067688,
      "logps/chosen": -571.1219482421875,
      "logps/rejected": -935.4173583984375,
      "loss": 0.6735,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.02279135212302208,
      "rewards/margins": 0.04717525467276573,
      "rewards/rejected": -0.024383898824453354,
      "step": 40
    },
    {
      "epoch": 0.4,
      "grad_norm": 3.8656836211850885,
      "learning_rate": 3.789911309071252e-07,
      "logits/chosen": 0.0432172492146492,
      "logits/rejected": -0.8026809692382812,
      "logps/chosen": -599.9905395507812,
      "logps/rejected": -956.2135620117188,
      "loss": 0.6511,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.03818320482969284,
      "rewards/margins": 0.0915202647447586,
      "rewards/rejected": -0.053337059915065765,
      "step": 50
    },
    {
      "epoch": 0.47,
      "grad_norm": 3.683371847061466,
      "learning_rate": 3.152640534699994e-07,
      "logits/chosen": 0.11929800361394882,
      "logits/rejected": -0.7896026372909546,
      "logps/chosen": -527.4035034179688,
      "logps/rejected": -991.9866943359375,
      "loss": 0.6306,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.054078031331300735,
      "rewards/margins": 0.11341840028762817,
      "rewards/rejected": -0.05934036895632744,
      "step": 60
    },
    {
      "epoch": 0.55,
      "grad_norm": 3.8302089344745407,
      "learning_rate": 2.4652489880792125e-07,
      "logits/chosen": 0.16239780187606812,
      "logits/rejected": -0.5915371775627136,
      "logps/chosen": -517.858154296875,
      "logps/rejected": -948.3834228515625,
      "loss": 0.6278,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.07395206391811371,
      "rewards/margins": 0.22345471382141113,
      "rewards/rejected": -0.14950266480445862,
      "step": 70
    },
    {
      "epoch": 0.63,
      "grad_norm": 4.211099057428358,
      "learning_rate": 1.780526211572016e-07,
      "logits/chosen": 0.13101445138454437,
      "logits/rejected": -0.729425847530365,
      "logps/chosen": -607.0712280273438,
      "logps/rejected": -908.4261474609375,
      "loss": 0.6168,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.05716314911842346,
      "rewards/margins": 0.34125471115112305,
      "rewards/rejected": -0.2840915620326996,
      "step": 80
    },
    {
      "epoch": 0.71,
      "grad_norm": 3.8906406186709197,
      "learning_rate": 1.1510567942602889e-07,
      "logits/chosen": 0.047411851584911346,
      "logits/rejected": -0.6215206384658813,
      "logps/chosen": -574.5473022460938,
      "logps/rejected": -883.5465087890625,
      "loss": 0.6002,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 0.07054788619279861,
      "rewards/margins": 0.363060861825943,
      "rewards/rejected": -0.2925129532814026,
      "step": 90
    },
    {
      "epoch": 0.79,
      "grad_norm": 4.2027425160958565,
      "learning_rate": 6.251820383244468e-08,
      "logits/chosen": 8.583068620282575e-07,
      "logits/rejected": -0.6966060400009155,
      "logps/chosen": -600.7443237304688,
      "logps/rejected": -899.9113159179688,
      "loss": 0.5979,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 0.04147876054048538,
      "rewards/margins": 0.5237125754356384,
      "rewards/rejected": -0.48223385214805603,
      "step": 100
    },
    {
      "epoch": 0.79,
      "eval_logits/chosen": 0.4742414057254791,
      "eval_logits/rejected": -0.5929847359657288,
      "eval_logps/chosen": -408.6627197265625,
      "eval_logps/rejected": -1010.139892578125,
      "eval_loss": 0.6596941947937012,
      "eval_rewards/accuracies": 0.804411768913269,
      "eval_rewards/chosen": 0.04729553684592247,
      "eval_rewards/margins": 0.07419417053461075,
      "eval_rewards/rejected": -0.02689863182604313,
      "eval_runtime": 302.2106,
      "eval_samples_per_second": 8.974,
      "eval_steps_per_second": 0.281,
      "step": 100
    },
    {
      "epoch": 0.87,
      "grad_norm": 3.9848703965373873,
      "learning_rate": 2.4328749671846117e-08,
      "logits/chosen": 0.0030116259586066008,
      "logits/rejected": -0.7621723413467407,
      "logps/chosen": -610.0753784179688,
      "logps/rejected": -956.7000122070312,
      "loss": 0.6101,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 0.05626688152551651,
      "rewards/margins": 0.29308319091796875,
      "rewards/rejected": -0.23681628704071045,
      "step": 110
    },
    {
      "epoch": 0.95,
      "grad_norm": 4.008522467528123,
      "learning_rate": 3.4701487751534475e-09,
      "logits/chosen": 0.1057804599404335,
      "logits/rejected": -0.7350226640701294,
      "logps/chosen": -561.1431274414062,
      "logps/rejected": -926.4228515625,
      "loss": 0.599,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 0.06413096934556961,
      "rewards/margins": 0.4213576316833496,
      "rewards/rejected": -0.3572266697883606,
      "step": 120
    },
    {
      "epoch": 1.0,
      "step": 126,
      "total_flos": 0.0,
      "train_loss": 0.6376642821327089,
      "train_runtime": 2213.5067,
      "train_samples_per_second": 3.653,
      "train_steps_per_second": 0.057
    }
  ],
  "logging_steps": 10,
  "max_steps": 126,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}