{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 161,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 4.59084677091884,
      "learning_rate": 2.941176470588235e-08,
      "logits/chosen": 0.16597579419612885,
      "logits/rejected": -0.893264651298523,
      "logps/chosen": -687.0694580078125,
      "logps/rejected": -1150.8621826171875,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.2570451246729855,
      "learning_rate": 2.941176470588235e-07,
      "logits/chosen": 0.3311888873577118,
      "logits/rejected": -0.695980429649353,
      "logps/chosen": -518.3226928710938,
      "logps/rejected": -964.5950317382812,
      "loss": 0.6931,
      "rewards/accuracies": 0.4722222089767456,
      "rewards/chosen": 0.00023680418962612748,
      "rewards/margins": 0.0004964147228747606,
      "rewards/rejected": -0.0002596105041448027,
      "step": 10
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.6335223260874243,
      "learning_rate": 4.994647308096508e-07,
      "logits/chosen": 0.1593589186668396,
      "logits/rejected": -0.6873095631599426,
      "logps/chosen": -594.7459716796875,
      "logps/rejected": -906.0856323242188,
      "loss": 0.6921,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": 0.0020452039316296577,
      "rewards/margins": 0.0012601878261193633,
      "rewards/rejected": 0.0007850162801332772,
      "step": 20
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.767508932992487,
      "learning_rate": 4.900124635964822e-07,
      "logits/chosen": 0.19252721965312958,
      "logits/rejected": -0.6640821099281311,
      "logps/chosen": -542.6539916992188,
      "logps/rejected": -986.5949096679688,
      "loss": 0.6864,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.00898707564920187,
      "rewards/margins": 0.01296447217464447,
      "rewards/rejected": -0.003977396059781313,
      "step": 30
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.787975298295214,
      "learning_rate": 4.691816889268769e-07,
      "logits/chosen": 0.12474404275417328,
      "logits/rejected": -0.6657639741897583,
      "logps/chosen": -521.1658935546875,
      "logps/rejected": -978.1348876953125,
      "loss": 0.6777,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.024259410798549652,
      "rewards/margins": 0.02917664125561714,
      "rewards/rejected": -0.004917231388390064,
      "step": 40
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.7293054585623944,
      "learning_rate": 4.379599518697443e-07,
      "logits/chosen": 0.1471071094274521,
      "logits/rejected": -0.7825916409492493,
      "logps/chosen": -521.8971557617188,
      "logps/rejected": -1019.1888427734375,
      "loss": 0.6663,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": 0.0357883982360363,
      "rewards/margins": 0.05407026410102844,
      "rewards/rejected": -0.01828186586499214,
      "step": 50
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.713443091206878,
      "learning_rate": 3.978274120908956e-07,
      "logits/chosen": 0.2528793215751648,
      "logits/rejected": -0.5950145721435547,
      "logps/chosen": -513.4190673828125,
      "logps/rejected": -923.5573120117188,
      "loss": 0.6465,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": 0.05145885795354843,
      "rewards/margins": 0.11846508085727692,
      "rewards/rejected": -0.06700621545314789,
      "step": 60
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.6423407771754426,
      "learning_rate": 3.506866724646843e-07,
      "logits/chosen": 0.12498810142278671,
      "logits/rejected": -0.5959888100624084,
      "logps/chosen": -588.9588623046875,
      "logps/rejected": -933.2869873046875,
      "loss": 0.6387,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": 0.05268377810716629,
      "rewards/margins": 0.18767260015010834,
      "rewards/rejected": -0.13498882949352264,
      "step": 70
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.012585827479405,
      "learning_rate": 2.987725805040321e-07,
      "logits/chosen": 0.22346019744873047,
      "logits/rejected": -0.7820445895195007,
      "logps/chosen": -509.46636962890625,
      "logps/rejected": -1001.2806396484375,
      "loss": 0.6114,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.06713758409023285,
      "rewards/margins": 0.2159689962863922,
      "rewards/rejected": -0.14883141219615936,
      "step": 80
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.135114682631198,
      "learning_rate": 2.445462787413597e-07,
      "logits/chosen": 0.21334119141101837,
      "logits/rejected": -0.5833973288536072,
      "logps/chosen": -570.7240600585938,
      "logps/rejected": -869.2347412109375,
      "loss": 0.6137,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": 0.06728974729776382,
      "rewards/margins": 0.33051320910453796,
      "rewards/rejected": -0.26322346925735474,
      "step": 90
    },
    {
      "epoch": 0.62,
      "grad_norm": 3.7781884105465315,
      "learning_rate": 1.9057852691845675e-07,
      "logits/chosen": 0.09849376976490021,
      "logits/rejected": -0.5272496938705444,
      "logps/chosen": -549.6043090820312,
      "logps/rejected": -869.72998046875,
      "loss": 0.6083,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.04291980713605881,
      "rewards/margins": 0.2984945476055145,
      "rewards/rejected": -0.2555747330188751,
      "step": 100
    },
    {
      "epoch": 0.62,
      "eval_logits/chosen": 0.483458012342453,
      "eval_logits/rejected": -0.3319720923900604,
      "eval_logps/chosen": -424.87884521484375,
      "eval_logps/rejected": -934.056640625,
      "eval_loss": 0.6257418394088745,
      "eval_rewards/accuracies": 0.7647058963775635,
      "eval_rewards/chosen": 0.04604706913232803,
      "eval_rewards/margins": 0.15231524407863617,
      "eval_rewards/rejected": -0.10626816749572754,
      "eval_runtime": 61.5744,
      "eval_samples_per_second": 8.445,
      "eval_steps_per_second": 0.276,
      "step": 100
    },
    {
      "epoch": 0.68,
      "grad_norm": 4.581420382507564,
      "learning_rate": 1.3942782744524973e-07,
      "logits/chosen": 0.13494840264320374,
      "logits/rejected": -0.6365220546722412,
      "logps/chosen": -582.8129272460938,
      "logps/rejected": -940.0499267578125,
      "loss": 0.5918,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 0.056283652782440186,
      "rewards/margins": 0.4604174494743347,
      "rewards/rejected": -0.40413379669189453,
      "step": 110
    },
    {
      "epoch": 0.75,
      "grad_norm": 4.661011326004169,
      "learning_rate": 9.351913195398522e-08,
      "logits/chosen": 0.19021789729595184,
      "logits/rejected": -0.5815246105194092,
      "logps/chosen": -587.3818359375,
      "logps/rejected": -930.6414184570312,
      "loss": 0.5884,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.04712503030896187,
      "rewards/margins": 0.40363869071006775,
      "rewards/rejected": -0.35651373863220215,
      "step": 120
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.358183039543505,
      "learning_rate": 5.502887922677959e-08,
      "logits/chosen": 0.12769225239753723,
      "logits/rejected": -0.5939640402793884,
      "logps/chosen": -566.1625366210938,
      "logps/rejected": -957.2125854492188,
      "loss": 0.5788,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": 0.04238845035433769,
      "rewards/margins": 0.35202011466026306,
      "rewards/rejected": -0.30963170528411865,
      "step": 130
    },
    {
      "epoch": 0.87,
      "grad_norm": 4.092331782234157,
      "learning_rate": 2.5781814616827933e-08,
      "logits/chosen": 0.15436235070228577,
      "logits/rejected": -0.7000457048416138,
      "logps/chosen": -537.4251708984375,
      "logps/rejected": -1025.563720703125,
      "loss": 0.5739,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 0.042118191719055176,
      "rewards/margins": 0.514610767364502,
      "rewards/rejected": -0.4724925458431244,
      "step": 140
    },
    {
      "epoch": 0.93,
      "grad_norm": 4.210158802402911,
      "learning_rate": 7.1644825466846415e-09,
      "logits/chosen": 0.20159873366355896,
      "logits/rejected": -0.6283941268920898,
      "logps/chosen": -500.183837890625,
      "logps/rejected": -933.0728759765625,
      "loss": 0.5726,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 0.05504927039146423,
      "rewards/margins": 0.4628073275089264,
      "rewards/rejected": -0.40775808691978455,
      "step": 150
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.053548631311742,
      "learning_rate": 5.949323002271688e-11,
      "logits/chosen": 0.21686339378356934,
      "logits/rejected": -0.6202417612075806,
      "logps/chosen": -521.2774047851562,
      "logps/rejected": -943.6966552734375,
      "loss": 0.5866,
      "rewards/accuracies": 0.8500000238418579,
      "rewards/chosen": 0.04219362139701843,
      "rewards/margins": 0.3843230903148651,
      "rewards/rejected": -0.3421294689178467,
      "step": 160
    },
    {
      "epoch": 1.0,
      "step": 161,
      "total_flos": 0.0,
      "train_loss": 0.62646086023461,
      "train_runtime": 2476.5181,
      "train_samples_per_second": 4.151,
      "train_steps_per_second": 0.065
    }
  ],
  "logging_steps": 10,
  "max_steps": 161,
  "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
}