{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 396,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007575757575757576,
      "grad_norm": 32.91323715722947,
      "learning_rate": 1.25e-08,
      "logits/chosen": -1.65625,
      "logits/rejected": -1.6640625,
      "logps/chosen": -174.0,
      "logps/rejected": -146.0,
      "loss": 0.6914,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.07575757575757576,
      "grad_norm": 33.44451536419978,
      "learning_rate": 1.25e-07,
      "logits/chosen": -1.59375,
      "logits/rejected": -1.5234375,
      "logps/chosen": -162.0,
      "logps/rejected": -138.0,
      "loss": 0.6866,
      "rewards/accuracies": 0.1666666716337204,
      "rewards/chosen": -0.0111083984375,
      "rewards/margins": -0.00138092041015625,
      "rewards/rejected": -0.00970458984375,
      "step": 10
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 24.18673269022567,
      "learning_rate": 2.5e-07,
      "logits/chosen": -1.5390625,
      "logits/rejected": -1.4296875,
      "logps/chosen": -140.0,
      "logps/rejected": -136.0,
      "loss": 0.581,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 0.01031494140625,
      "rewards/margins": 0.2392578125,
      "rewards/rejected": -0.228515625,
      "step": 20
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 9.48516053714528,
      "learning_rate": 3.75e-07,
      "logits/chosen": -1.5390625,
      "logits/rejected": -1.3828125,
      "logps/chosen": -138.0,
      "logps/rejected": -148.0,
      "loss": 0.29,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.0018768310546875,
      "rewards/margins": 1.2578125,
      "rewards/rejected": -1.2578125,
      "step": 30
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 0.6687099946093109,
      "learning_rate": 5e-07,
      "logits/chosen": -1.53125,
      "logits/rejected": -1.4140625,
      "logps/chosen": -157.0,
      "logps/rejected": -181.0,
      "loss": 0.0402,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.333984375,
      "rewards/margins": 4.21875,
      "rewards/rejected": -4.5625,
      "step": 40
    },
    {
      "epoch": 0.3787878787878788,
      "grad_norm": 0.13276288591607247,
      "learning_rate": 4.859550561797752e-07,
      "logits/chosen": -1.53125,
      "logits/rejected": -1.40625,
      "logps/chosen": -165.0,
      "logps/rejected": -207.0,
      "loss": 0.0018,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.6171875,
      "rewards/margins": 7.0625,
      "rewards/rejected": -7.6875,
      "step": 50
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.22272699774735913,
      "learning_rate": 4.7191011235955054e-07,
      "logits/chosen": -1.578125,
      "logits/rejected": -1.484375,
      "logps/chosen": -170.0,
      "logps/rejected": -227.0,
      "loss": 0.0016,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.6875,
      "rewards/margins": 8.0625,
      "rewards/rejected": -8.75,
      "step": 60
    },
    {
      "epoch": 0.5303030303030303,
      "grad_norm": 0.224870741613382,
      "learning_rate": 4.5786516853932584e-07,
      "logits/chosen": -1.546875,
      "logits/rejected": -1.4921875,
      "logps/chosen": -163.0,
      "logps/rejected": -231.0,
      "loss": 0.0012,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.7734375,
      "rewards/margins": 8.5,
      "rewards/rejected": -9.25,
      "step": 70
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.16097415690521813,
      "learning_rate": 4.438202247191011e-07,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.46875,
      "logps/chosen": -153.0,
      "logps/rejected": -235.0,
      "loss": 0.0027,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.79296875,
      "rewards/margins": 9.125,
      "rewards/rejected": -9.9375,
      "step": 80
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.09475144242280534,
      "learning_rate": 4.297752808988764e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.5546875,
      "logps/chosen": -151.0,
      "logps/rejected": -244.0,
      "loss": 0.0009,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.6640625,
      "rewards/margins": 9.5,
      "rewards/rejected": -10.125,
      "step": 90
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.19544716197287873,
      "learning_rate": 4.157303370786517e-07,
      "logits/chosen": -1.484375,
      "logits/rejected": -1.515625,
      "logps/chosen": -163.0,
      "logps/rejected": -246.0,
      "loss": 0.0009,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.62109375,
      "rewards/margins": 9.75,
      "rewards/rejected": -10.375,
      "step": 100
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.18126368085142058,
      "learning_rate": 4.0168539325842696e-07,
      "logits/chosen": -1.515625,
      "logits/rejected": -1.484375,
      "logps/chosen": -159.0,
      "logps/rejected": -246.0,
      "loss": 0.001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.703125,
      "rewards/margins": 10.0625,
      "rewards/rejected": -10.75,
      "step": 110
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.00566877598086428,
      "learning_rate": 3.876404494382022e-07,
      "logits/chosen": -1.53125,
      "logits/rejected": -1.4453125,
      "logps/chosen": -148.0,
      "logps/rejected": -247.0,
      "loss": 0.0014,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.77734375,
      "rewards/margins": 10.5,
      "rewards/rejected": -11.3125,
      "step": 120
    },
    {
      "epoch": 0.9848484848484849,
      "grad_norm": 0.5918723347747266,
      "learning_rate": 3.735955056179775e-07,
      "logits/chosen": -1.609375,
      "logits/rejected": -1.546875,
      "logps/chosen": -171.0,
      "logps/rejected": -252.0,
      "loss": 0.0007,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.65234375,
      "rewards/margins": 10.9375,
      "rewards/rejected": -11.5625,
      "step": 130
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -1.484375,
      "eval_logits/rejected": -1.4609375,
      "eval_logps/chosen": -117.5,
      "eval_logps/rejected": -245.0,
      "eval_loss": 0.002259350847452879,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.6015625,
      "eval_rewards/margins": 10.375,
      "eval_rewards/rejected": -11.0,
      "eval_runtime": 5.1355,
      "eval_samples_per_second": 16.357,
      "eval_steps_per_second": 0.584,
      "step": 132
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.044137608822413354,
      "learning_rate": 3.5955056179775277e-07,
      "logits/chosen": -1.578125,
      "logits/rejected": -1.5078125,
      "logps/chosen": -155.0,
      "logps/rejected": -260.0,
      "loss": 0.0003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.65625,
      "rewards/margins": 11.3125,
      "rewards/rejected": -12.0,
      "step": 140
    },
    {
      "epoch": 1.1363636363636362,
      "grad_norm": 0.03726527729430989,
      "learning_rate": 3.4550561797752807e-07,
      "logits/chosen": -1.578125,
      "logits/rejected": -1.515625,
      "logps/chosen": -152.0,
      "logps/rejected": -252.0,
      "loss": 0.0017,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.427734375,
      "rewards/margins": 11.0,
      "rewards/rejected": -11.4375,
      "step": 150
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.04249817642262755,
      "learning_rate": 3.314606741573033e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.5703125,
      "logps/chosen": -157.0,
      "logps/rejected": -252.0,
      "loss": 0.0004,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.345703125,
      "rewards/margins": 10.9375,
      "rewards/rejected": -11.25,
      "step": 160
    },
    {
      "epoch": 1.2878787878787878,
      "grad_norm": 0.02636991212367461,
      "learning_rate": 3.1741573033707863e-07,
      "logits/chosen": -1.546875,
      "logits/rejected": -1.5546875,
      "logps/chosen": -153.0,
      "logps/rejected": -253.0,
      "loss": 0.0003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.326171875,
      "rewards/margins": 11.0,
      "rewards/rejected": -11.3125,
      "step": 170
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.002485552309390973,
      "learning_rate": 3.0337078651685393e-07,
      "logits/chosen": -1.59375,
      "logits/rejected": -1.578125,
      "logps/chosen": -170.0,
      "logps/rejected": -251.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.4375,
      "rewards/margins": 11.0625,
      "rewards/rejected": -11.5,
      "step": 180
    },
    {
      "epoch": 1.4393939393939394,
      "grad_norm": 0.06372981794712247,
      "learning_rate": 2.893258426966292e-07,
      "logits/chosen": -1.546875,
      "logits/rejected": -1.53125,
      "logps/chosen": -152.0,
      "logps/rejected": -254.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.384765625,
      "rewards/margins": 11.6875,
      "rewards/rejected": -12.0625,
      "step": 190
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.056568516864774675,
      "learning_rate": 2.752808988764045e-07,
      "logits/chosen": -1.5703125,
      "logits/rejected": -1.5078125,
      "logps/chosen": -158.0,
      "logps/rejected": -256.0,
      "loss": 0.0011,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.53515625,
      "rewards/margins": 11.5,
      "rewards/rejected": -12.0,
      "step": 200
    },
    {
      "epoch": 1.5909090909090908,
      "grad_norm": 0.0017135050317799595,
      "learning_rate": 2.612359550561798e-07,
      "logits/chosen": -1.609375,
      "logits/rejected": -1.53125,
      "logps/chosen": -171.0,
      "logps/rejected": -264.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.6875,
      "rewards/margins": 11.875,
      "rewards/rejected": -12.5,
      "step": 210
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.011438568560823217,
      "learning_rate": 2.4719101123595505e-07,
      "logits/chosen": -1.5,
      "logits/rejected": -1.53125,
      "logps/chosen": -155.0,
      "logps/rejected": -264.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.8125,
      "rewards/margins": 11.6875,
      "rewards/rejected": -12.5,
      "step": 220
    },
    {
      "epoch": 1.7424242424242424,
      "grad_norm": 0.00196532311899783,
      "learning_rate": 2.331460674157303e-07,
      "logits/chosen": -1.59375,
      "logits/rejected": -1.5625,
      "logps/chosen": -160.0,
      "logps/rejected": -264.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.98828125,
      "rewards/margins": 11.75,
      "rewards/rejected": -12.75,
      "step": 230
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.02412516870728502,
      "learning_rate": 2.191011235955056e-07,
      "logits/chosen": -1.578125,
      "logits/rejected": -1.578125,
      "logps/chosen": -163.0,
      "logps/rejected": -272.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.859375,
      "rewards/margins": 12.0625,
      "rewards/rejected": -12.9375,
      "step": 240
    },
    {
      "epoch": 1.893939393939394,
      "grad_norm": 0.008677592300417748,
      "learning_rate": 2.0505617977528089e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.546875,
      "logps/chosen": -164.0,
      "logps/rejected": -268.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.98046875,
      "rewards/margins": 12.0,
      "rewards/rejected": -13.0,
      "step": 250
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.005504070078399264,
      "learning_rate": 1.9101123595505617e-07,
      "logits/chosen": -1.53125,
      "logits/rejected": -1.515625,
      "logps/chosen": -164.0,
      "logps/rejected": -268.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.91796875,
      "rewards/margins": 12.1875,
      "rewards/rejected": -13.0625,
      "step": 260
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -1.4921875,
      "eval_logits/rejected": -1.484375,
      "eval_logps/chosen": -119.5,
      "eval_logps/rejected": -256.0,
      "eval_loss": 0.0008152184309437871,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.79296875,
      "eval_rewards/margins": 11.375,
      "eval_rewards/rejected": -12.125,
      "eval_runtime": 5.7051,
      "eval_samples_per_second": 14.724,
      "eval_steps_per_second": 0.526,
      "step": 264
    },
    {
      "epoch": 2.0454545454545454,
      "grad_norm": 0.00984660780499755,
      "learning_rate": 1.7696629213483144e-07,
      "logits/chosen": -1.5234375,
      "logits/rejected": -1.5390625,
      "logps/chosen": -148.0,
      "logps/rejected": -266.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.7734375,
      "rewards/margins": 12.0,
      "rewards/rejected": -12.75,
      "step": 270
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.004414538457258972,
      "learning_rate": 1.6292134831460675e-07,
      "logits/chosen": -1.5,
      "logits/rejected": -1.515625,
      "logps/chosen": -165.0,
      "logps/rejected": -264.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.75390625,
      "rewards/margins": 12.125,
      "rewards/rejected": -12.875,
      "step": 280
    },
    {
      "epoch": 2.196969696969697,
      "grad_norm": 0.016662720665351936,
      "learning_rate": 1.4887640449438203e-07,
      "logits/chosen": -1.5703125,
      "logits/rejected": -1.546875,
      "logps/chosen": -156.0,
      "logps/rejected": -272.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.79296875,
      "rewards/margins": 12.1875,
      "rewards/rejected": -13.0,
      "step": 290
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.05873531330460493,
      "learning_rate": 1.3483146067415728e-07,
      "logits/chosen": -1.59375,
      "logits/rejected": -1.546875,
      "logps/chosen": -160.0,
      "logps/rejected": -268.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.78515625,
      "rewards/margins": 12.25,
      "rewards/rejected": -13.0,
      "step": 300
    },
    {
      "epoch": 2.3484848484848486,
      "grad_norm": 0.046015511031248915,
      "learning_rate": 1.2078651685393259e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.546875,
      "logps/chosen": -151.0,
      "logps/rejected": -264.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.70703125,
      "rewards/margins": 12.125,
      "rewards/rejected": -12.875,
      "step": 310
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.3021671776155001,
      "learning_rate": 1.0674157303370785e-07,
      "logits/chosen": -1.578125,
      "logits/rejected": -1.578125,
      "logps/chosen": -161.0,
      "logps/rejected": -272.0,
      "loss": 0.0003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.75390625,
      "rewards/margins": 12.3125,
      "rewards/rejected": -13.0625,
      "step": 320
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.0015387455528242642,
      "learning_rate": 9.269662921348314e-08,
      "logits/chosen": -1.6171875,
      "logits/rejected": -1.5703125,
      "logps/chosen": -163.0,
      "logps/rejected": -266.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.74609375,
      "rewards/margins": 12.375,
      "rewards/rejected": -13.125,
      "step": 330
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.001321971934515845,
      "learning_rate": 7.865168539325842e-08,
      "logits/chosen": -1.53125,
      "logits/rejected": -1.609375,
      "logps/chosen": -150.0,
      "logps/rejected": -276.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.7421875,
      "rewards/margins": 12.625,
      "rewards/rejected": -13.375,
      "step": 340
    },
    {
      "epoch": 2.6515151515151514,
      "grad_norm": 0.0057061429256677594,
      "learning_rate": 6.460674157303371e-08,
      "logits/chosen": -1.53125,
      "logits/rejected": -1.5703125,
      "logps/chosen": -159.0,
      "logps/rejected": -276.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.875,
      "rewards/margins": 12.5625,
      "rewards/rejected": -13.4375,
      "step": 350
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.4762341538722265,
      "learning_rate": 5.056179775280899e-08,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.546875,
      "logps/chosen": -143.0,
      "logps/rejected": -268.0,
      "loss": 0.0004,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.68359375,
      "rewards/margins": 12.3125,
      "rewards/rejected": -13.0,
      "step": 360
    },
    {
      "epoch": 2.8030303030303028,
      "grad_norm": 0.2900268557706346,
      "learning_rate": 3.6516853932584266e-08,
      "logits/chosen": -1.53125,
      "logits/rejected": -1.5859375,
      "logps/chosen": -147.0,
      "logps/rejected": -270.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.85546875,
      "rewards/margins": 12.25,
      "rewards/rejected": -13.0625,
      "step": 370
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.005709415780533644,
      "learning_rate": 2.2471910112359548e-08,
      "logits/chosen": -1.59375,
      "logits/rejected": -1.5390625,
      "logps/chosen": -163.0,
      "logps/rejected": -270.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.74609375,
      "rewards/margins": 12.4375,
      "rewards/rejected": -13.1875,
      "step": 380
    },
    {
      "epoch": 2.9545454545454546,
      "grad_norm": 0.009508653721430635,
      "learning_rate": 8.42696629213483e-09,
      "logits/chosen": -1.5703125,
      "logits/rejected": -1.59375,
      "logps/chosen": -160.0,
      "logps/rejected": -266.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.7421875,
      "rewards/margins": 12.125,
      "rewards/rejected": -12.875,
      "step": 390
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -1.5,
      "eval_logits/rejected": -1.484375,
      "eval_logps/chosen": -119.0,
      "eval_logps/rejected": -260.0,
      "eval_loss": 0.0003405184834264219,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.734375,
      "eval_rewards/margins": 11.75,
      "eval_rewards/rejected": -12.5,
      "eval_runtime": 16.5049,
      "eval_samples_per_second": 5.089,
      "eval_steps_per_second": 0.182,
      "step": 396
    },
    {
      "epoch": 3.0,
      "step": 396,
      "total_flos": 0.0,
      "train_loss": 0.04084648387600677,
      "train_runtime": 2792.6222,
      "train_samples_per_second": 4.525,
      "train_steps_per_second": 0.142
    }
  ],
  "logging_steps": 10,
  "max_steps": 396,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 8,
  "trial_name": null,
  "trial_params": null
}