{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9984,
  "eval_steps": 500,
  "global_step": 156,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 1061.4292758130512,
      "learning_rate": 3.125e-09,
      "logits/chosen": -3.9499800205230713,
      "logits/rejected": -4.237819194793701,
      "logps/chosen": -300.693115234375,
      "logps/rejected": -249.96307373046875,
      "loss": 0.6941,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "grad_norm": 1100.0352396547798,
      "learning_rate": 3.125e-08,
      "logits/chosen": -4.129236221313477,
      "logits/rejected": -4.351477146148682,
      "logps/chosen": -351.48150634765625,
      "logps/rejected": -308.8130187988281,
      "loss": 0.7282,
      "rewards/accuracies": 0.3611111044883728,
      "rewards/chosen": -0.027542730793356895,
      "rewards/margins": -0.10096076130867004,
      "rewards/rejected": 0.073418028652668,
      "step": 10
    },
    {
      "epoch": 0.13,
      "grad_norm": 1023.2979189429582,
      "learning_rate": 4.9899357349880975e-08,
      "logits/chosen": -4.195385932922363,
      "logits/rejected": -4.382458686828613,
      "logps/chosen": -334.9761962890625,
      "logps/rejected": -293.8690185546875,
      "loss": 0.6994,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.18794658780097961,
      "rewards/margins": 0.09464438259601593,
      "rewards/rejected": 0.09330219030380249,
      "step": 20
    },
    {
      "epoch": 0.19,
      "grad_norm": 716.4790257172086,
      "learning_rate": 4.877641290737884e-08,
      "logits/chosen": -4.231568813323975,
      "logits/rejected": -4.364924430847168,
      "logps/chosen": -327.92413330078125,
      "logps/rejected": -295.37176513671875,
      "loss": 0.6055,
      "rewards/accuracies": 0.684374988079071,
      "rewards/chosen": 0.8660896420478821,
      "rewards/margins": 0.425116628408432,
      "rewards/rejected": 0.44097304344177246,
      "step": 30
    },
    {
      "epoch": 0.26,
      "grad_norm": 820.682467230606,
      "learning_rate": 4.646121984004665e-08,
      "logits/chosen": -4.1439104080200195,
      "logits/rejected": -4.345718860626221,
      "logps/chosen": -330.7722473144531,
      "logps/rejected": -288.60272216796875,
      "loss": 0.5304,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": 1.6021575927734375,
      "rewards/margins": 0.956769585609436,
      "rewards/rejected": 0.6453880071640015,
      "step": 40
    },
    {
      "epoch": 0.32,
      "grad_norm": 587.6344375884266,
      "learning_rate": 4.3069871595684784e-08,
      "logits/chosen": -4.235655784606934,
      "logits/rejected": -4.4143476486206055,
      "logps/chosen": -330.8318786621094,
      "logps/rejected": -291.75067138671875,
      "loss": 0.525,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 2.010468006134033,
      "rewards/margins": 1.2047243118286133,
      "rewards/rejected": 0.8057435750961304,
      "step": 50
    },
    {
      "epoch": 0.38,
      "grad_norm": 586.1895231882175,
      "learning_rate": 3.8772424536302564e-08,
      "logits/chosen": -4.247714519500732,
      "logits/rejected": -4.417937278747559,
      "logps/chosen": -322.380126953125,
      "logps/rejected": -291.76654052734375,
      "loss": 0.517,
      "rewards/accuracies": 0.8218749761581421,
      "rewards/chosen": 2.19193696975708,
      "rewards/margins": 1.311265230178833,
      "rewards/rejected": 0.8806716203689575,
      "step": 60
    },
    {
      "epoch": 0.45,
      "grad_norm": 528.8947001003231,
      "learning_rate": 3.378437060203357e-08,
      "logits/chosen": -4.170197486877441,
      "logits/rejected": -4.359009742736816,
      "logps/chosen": -322.4336242675781,
      "logps/rejected": -289.197998046875,
      "loss": 0.5096,
      "rewards/accuracies": 0.815625011920929,
      "rewards/chosen": 2.203657865524292,
      "rewards/margins": 1.3729288578033447,
      "rewards/rejected": 0.8307291269302368,
      "step": 70
    },
    {
      "epoch": 0.51,
      "grad_norm": 544.8408089721645,
      "learning_rate": 2.8355831645441387e-08,
      "logits/chosen": -4.038945198059082,
      "logits/rejected": -4.326292991638184,
      "logps/chosen": -348.65472412109375,
      "logps/rejected": -308.1936950683594,
      "loss": 0.5115,
      "rewards/accuracies": 0.875,
      "rewards/chosen": 2.3145015239715576,
      "rewards/margins": 1.5040271282196045,
      "rewards/rejected": 0.8104745149612427,
      "step": 80
    },
    {
      "epoch": 0.58,
      "grad_norm": 526.2097671297138,
      "learning_rate": 2.2759017277414164e-08,
      "logits/chosen": -4.170694351196289,
      "logits/rejected": -4.379502296447754,
      "logps/chosen": -335.43402099609375,
      "logps/rejected": -295.8087463378906,
      "loss": 0.5066,
      "rewards/accuracies": 0.859375,
      "rewards/chosen": 1.9496158361434937,
      "rewards/margins": 1.3254307508468628,
      "rewards/rejected": 0.6241849660873413,
      "step": 90
    },
    {
      "epoch": 0.64,
      "grad_norm": 595.4206000749634,
      "learning_rate": 1.7274575140626317e-08,
      "logits/chosen": -4.159605979919434,
      "logits/rejected": -4.377494812011719,
      "logps/chosen": -333.1622314453125,
      "logps/rejected": -286.3451232910156,
      "loss": 0.5142,
      "rewards/accuracies": 0.871874988079071,
      "rewards/chosen": 2.0654988288879395,
      "rewards/margins": 1.4395055770874023,
      "rewards/rejected": 0.6259931921958923,
      "step": 100
    },
    {
      "epoch": 0.7,
      "grad_norm": 528.6243658840883,
      "learning_rate": 1.217751806485235e-08,
      "logits/chosen": -4.138543128967285,
      "logits/rejected": -4.379323959350586,
      "logps/chosen": -314.89556884765625,
      "logps/rejected": -277.0162658691406,
      "loss": 0.5071,
      "rewards/accuracies": 0.840624988079071,
      "rewards/chosen": 2.0163025856018066,
      "rewards/margins": 1.3896347284317017,
      "rewards/rejected": 0.6266676783561707,
      "step": 110
    },
    {
      "epoch": 0.77,
      "grad_norm": 557.9046270518194,
      "learning_rate": 7.723433775328384e-09,
      "logits/chosen": -4.136676788330078,
      "logits/rejected": -4.345128536224365,
      "logps/chosen": -328.94482421875,
      "logps/rejected": -281.4302673339844,
      "loss": 0.5052,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": 1.9894015789031982,
      "rewards/margins": 1.363441824913025,
      "rewards/rejected": 0.625959575176239,
      "step": 120
    },
    {
      "epoch": 0.83,
      "grad_norm": 523.8022952711281,
      "learning_rate": 4.135668656967433e-09,
      "logits/chosen": -4.222664833068848,
      "logits/rejected": -4.383507251739502,
      "logps/chosen": -334.4266357421875,
      "logps/rejected": -287.63067626953125,
      "loss": 0.5059,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 2.026325225830078,
      "rewards/margins": 1.3963022232055664,
      "rewards/rejected": 0.6300228834152222,
      "step": 130
    },
    {
      "epoch": 0.9,
      "grad_norm": 601.0204054485596,
      "learning_rate": 1.5941282340065698e-09,
      "logits/chosen": -4.176892280578613,
      "logits/rejected": -4.3910746574401855,
      "logps/chosen": -335.8815612792969,
      "logps/rejected": -304.35723876953125,
      "loss": 0.5101,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": 1.9034252166748047,
      "rewards/margins": 1.3076425790786743,
      "rewards/rejected": 0.5957827568054199,
      "step": 140
    },
    {
      "epoch": 0.96,
      "grad_norm": 496.1092972793449,
      "learning_rate": 2.262559558016325e-10,
      "logits/chosen": -4.114329814910889,
      "logits/rejected": -4.342306137084961,
      "logps/chosen": -342.4503479003906,
      "logps/rejected": -296.07086181640625,
      "loss": 0.506,
      "rewards/accuracies": 0.840624988079071,
      "rewards/chosen": 2.0278801918029785,
      "rewards/margins": 1.378021001815796,
      "rewards/rejected": 0.6498591303825378,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 156,
      "total_flos": 0.0,
      "train_loss": 0.5437000210468586,
      "train_runtime": 4622.4032,
      "train_samples_per_second": 8.641,
      "train_steps_per_second": 0.034
    }
  ],
  "logging_steps": 10,
  "max_steps": 156,
  "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
}