{
  "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": 32.091551837266515,
      "learning_rate": 1.25e-08,
      "logits/chosen": -1.8046875,
      "logits/rejected": -1.5859375,
      "logps/chosen": -163.0,
      "logps/rejected": -142.0,
      "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": 29.307157179340344,
      "learning_rate": 1.25e-07,
      "logits/chosen": -1.59375,
      "logits/rejected": -1.328125,
      "logps/chosen": -152.0,
      "logps/rejected": -132.0,
      "loss": 0.6876,
      "rewards/accuracies": 0.2638888955116272,
      "rewards/chosen": -0.00347900390625,
      "rewards/margins": 0.0069580078125,
      "rewards/rejected": -0.01043701171875,
      "step": 10
    },
    {
      "epoch": 0.15151515151515152,
      "grad_norm": 25.07098632835419,
      "learning_rate": 2.5e-07,
      "logits/chosen": -1.609375,
      "logits/rejected": -1.3828125,
      "logps/chosen": -153.0,
      "logps/rejected": -132.0,
      "loss": 0.5848,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 0.01531982421875,
      "rewards/margins": 0.232421875,
      "rewards/rejected": -0.2177734375,
      "step": 20
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 11.45839328112264,
      "learning_rate": 3.75e-07,
      "logits/chosen": -1.609375,
      "logits/rejected": -1.3359375,
      "logps/chosen": -147.0,
      "logps/rejected": -141.0,
      "loss": 0.292,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 0.032470703125,
      "rewards/margins": 1.21875,
      "rewards/rejected": -1.1796875,
      "step": 30
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 0.46253339983913294,
      "learning_rate": 5e-07,
      "logits/chosen": -1.515625,
      "logits/rejected": -1.3671875,
      "logps/chosen": -132.0,
      "logps/rejected": -174.0,
      "loss": 0.0439,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.255859375,
      "rewards/margins": 4.25,
      "rewards/rejected": -4.5,
      "step": 40
    },
    {
      "epoch": 0.3787878787878788,
      "grad_norm": 0.3395592758682824,
      "learning_rate": 4.859550561797752e-07,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.3671875,
      "logps/chosen": -151.0,
      "logps/rejected": -211.0,
      "loss": 0.0033,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.63671875,
      "rewards/margins": 6.96875,
      "rewards/rejected": -7.625,
      "step": 50
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.5048466762536648,
      "learning_rate": 4.7191011235955054e-07,
      "logits/chosen": -1.515625,
      "logits/rejected": -1.28125,
      "logps/chosen": -145.0,
      "logps/rejected": -214.0,
      "loss": 0.0017,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.71875,
      "rewards/margins": 8.0,
      "rewards/rejected": -8.75,
      "step": 60
    },
    {
      "epoch": 0.5303030303030303,
      "grad_norm": 0.2393323020031099,
      "learning_rate": 4.5786516853932584e-07,
      "logits/chosen": -1.625,
      "logits/rejected": -1.515625,
      "logps/chosen": -169.0,
      "logps/rejected": -224.0,
      "loss": 0.0022,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.8046875,
      "rewards/margins": 8.375,
      "rewards/rejected": -9.1875,
      "step": 70
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 0.4781670788915071,
      "learning_rate": 4.438202247191011e-07,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.359375,
      "logps/chosen": -157.0,
      "logps/rejected": -230.0,
      "loss": 0.0012,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.8046875,
      "rewards/margins": 9.25,
      "rewards/rejected": -10.125,
      "step": 80
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.17816540763501387,
      "learning_rate": 4.297752808988764e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.421875,
      "logps/chosen": -146.0,
      "logps/rejected": -240.0,
      "loss": 0.0005,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.72265625,
      "rewards/margins": 9.8125,
      "rewards/rejected": -10.5,
      "step": 90
    },
    {
      "epoch": 0.7575757575757576,
      "grad_norm": 0.11217914292234316,
      "learning_rate": 4.157303370786517e-07,
      "logits/chosen": -1.6171875,
      "logits/rejected": -1.4453125,
      "logps/chosen": -152.0,
      "logps/rejected": -231.0,
      "loss": 0.0008,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.8125,
      "rewards/margins": 9.5,
      "rewards/rejected": -10.3125,
      "step": 100
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.010563655594026364,
      "learning_rate": 4.0168539325842696e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.4375,
      "logps/chosen": -165.0,
      "logps/rejected": -239.0,
      "loss": 0.0018,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.92578125,
      "rewards/margins": 9.875,
      "rewards/rejected": -10.8125,
      "step": 110
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.023851884708781215,
      "learning_rate": 3.876404494382022e-07,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.4296875,
      "logps/chosen": -152.0,
      "logps/rejected": -244.0,
      "loss": 0.0004,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.9453125,
      "rewards/margins": 10.3125,
      "rewards/rejected": -11.25,
      "step": 120
    },
    {
      "epoch": 0.9848484848484849,
      "grad_norm": 0.09604967540471057,
      "learning_rate": 3.735955056179775e-07,
      "logits/chosen": -1.5546875,
      "logits/rejected": -1.46875,
      "logps/chosen": -153.0,
      "logps/rejected": -244.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.953125,
      "rewards/margins": 10.125,
      "rewards/rejected": -11.0625,
      "step": 130
    },
    {
      "epoch": 1.0,
      "eval_logits/chosen": -1.4921875,
      "eval_logits/rejected": -1.453125,
      "eval_logps/chosen": -147.0,
      "eval_logps/rejected": -244.0,
      "eval_loss": 0.00021177291637286544,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -0.81640625,
      "eval_rewards/margins": 10.4375,
      "eval_rewards/rejected": -11.3125,
      "eval_runtime": 12.2918,
      "eval_samples_per_second": 16.271,
      "eval_steps_per_second": 0.569,
      "step": 132
    },
    {
      "epoch": 1.0606060606060606,
      "grad_norm": 0.025427232520216565,
      "learning_rate": 3.5955056179775277e-07,
      "logits/chosen": -1.5546875,
      "logits/rejected": -1.421875,
      "logps/chosen": -156.0,
      "logps/rejected": -243.0,
      "loss": 0.0003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.9140625,
      "rewards/margins": 10.5625,
      "rewards/rejected": -11.5,
      "step": 140
    },
    {
      "epoch": 1.1363636363636362,
      "grad_norm": 0.01529093932750603,
      "learning_rate": 3.4550561797752807e-07,
      "logits/chosen": -1.53125,
      "logits/rejected": -1.46875,
      "logps/chosen": -140.0,
      "logps/rejected": -250.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.76171875,
      "rewards/margins": 11.125,
      "rewards/rejected": -11.875,
      "step": 150
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.010074912036860673,
      "learning_rate": 3.314606741573033e-07,
      "logits/chosen": -1.5390625,
      "logits/rejected": -1.4453125,
      "logps/chosen": -150.0,
      "logps/rejected": -245.0,
      "loss": 0.0005,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.8984375,
      "rewards/margins": 11.0,
      "rewards/rejected": -11.875,
      "step": 160
    },
    {
      "epoch": 1.2878787878787878,
      "grad_norm": 0.024954993930167084,
      "learning_rate": 3.1741573033707863e-07,
      "logits/chosen": -1.59375,
      "logits/rejected": -1.4921875,
      "logps/chosen": -152.0,
      "logps/rejected": -247.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.76953125,
      "rewards/margins": 10.875,
      "rewards/rejected": -11.625,
      "step": 170
    },
    {
      "epoch": 1.3636363636363638,
      "grad_norm": 0.003532695770478855,
      "learning_rate": 3.0337078651685393e-07,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.46875,
      "logps/chosen": -127.5,
      "logps/rejected": -249.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.9765625,
      "rewards/margins": 10.9375,
      "rewards/rejected": -11.875,
      "step": 180
    },
    {
      "epoch": 1.4393939393939394,
      "grad_norm": 0.0018665433852718737,
      "learning_rate": 2.893258426966292e-07,
      "logits/chosen": -1.6015625,
      "logits/rejected": -1.5078125,
      "logps/chosen": -164.0,
      "logps/rejected": -258.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.91015625,
      "rewards/margins": 11.375,
      "rewards/rejected": -12.25,
      "step": 190
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 0.020725382536818583,
      "learning_rate": 2.752808988764045e-07,
      "logits/chosen": -1.578125,
      "logits/rejected": -1.4765625,
      "logps/chosen": -157.0,
      "logps/rejected": -253.0,
      "loss": 0.0004,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.0078125,
      "rewards/margins": 11.25,
      "rewards/rejected": -12.3125,
      "step": 200
    },
    {
      "epoch": 1.5909090909090908,
      "grad_norm": 0.009333169079013727,
      "learning_rate": 2.612359550561798e-07,
      "logits/chosen": -1.515625,
      "logits/rejected": -1.4765625,
      "logps/chosen": -154.0,
      "logps/rejected": -260.0,
      "loss": 0.0005,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.390625,
      "rewards/margins": 11.4375,
      "rewards/rejected": -12.875,
      "step": 210
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.018292778581923616,
      "learning_rate": 2.4719101123595505e-07,
      "logits/chosen": -1.5703125,
      "logits/rejected": -1.375,
      "logps/chosen": -156.0,
      "logps/rejected": -248.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.2265625,
      "rewards/margins": 11.375,
      "rewards/rejected": -12.625,
      "step": 220
    },
    {
      "epoch": 1.7424242424242424,
      "grad_norm": 0.006108685272396747,
      "learning_rate": 2.331460674157303e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.40625,
      "logps/chosen": -155.0,
      "logps/rejected": -253.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.2578125,
      "rewards/margins": 11.5,
      "rewards/rejected": -12.75,
      "step": 230
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 0.008540878901725096,
      "learning_rate": 2.191011235955056e-07,
      "logits/chosen": -1.5546875,
      "logits/rejected": -1.4921875,
      "logps/chosen": -145.0,
      "logps/rejected": -256.0,
      "loss": 0.0003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.046875,
      "rewards/margins": 11.8125,
      "rewards/rejected": -12.875,
      "step": 240
    },
    {
      "epoch": 1.893939393939394,
      "grad_norm": 0.26338726379953975,
      "learning_rate": 2.0505617977528089e-07,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.5,
      "logps/chosen": -152.0,
      "logps/rejected": -255.0,
      "loss": 0.0003,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.046875,
      "rewards/margins": 11.8125,
      "rewards/rejected": -12.875,
      "step": 250
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 0.0326586506464961,
      "learning_rate": 1.9101123595505617e-07,
      "logits/chosen": -1.5703125,
      "logits/rejected": -1.4453125,
      "logps/chosen": -169.0,
      "logps/rejected": -258.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.1875,
      "rewards/margins": 11.625,
      "rewards/rejected": -12.8125,
      "step": 260
    },
    {
      "epoch": 2.0,
      "eval_logits/chosen": -1.5,
      "eval_logits/rejected": -1.484375,
      "eval_logps/chosen": -149.0,
      "eval_logps/rejected": -260.0,
      "eval_loss": 6.90150263835676e-05,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -1.0078125,
      "eval_rewards/margins": 11.8125,
      "eval_rewards/rejected": -12.8125,
      "eval_runtime": 14.6298,
      "eval_samples_per_second": 13.671,
      "eval_steps_per_second": 0.478,
      "step": 264
    },
    {
      "epoch": 2.0454545454545454,
      "grad_norm": 0.0013074998122167467,
      "learning_rate": 1.7696629213483144e-07,
      "logits/chosen": -1.546875,
      "logits/rejected": -1.4609375,
      "logps/chosen": -155.0,
      "logps/rejected": -258.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.90625,
      "rewards/margins": 11.875,
      "rewards/rejected": -12.8125,
      "step": 270
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 0.018655034734646705,
      "learning_rate": 1.6292134831460675e-07,
      "logits/chosen": -1.59375,
      "logits/rejected": -1.46875,
      "logps/chosen": -166.0,
      "logps/rejected": -258.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.0546875,
      "rewards/margins": 12.125,
      "rewards/rejected": -13.25,
      "step": 280
    },
    {
      "epoch": 2.196969696969697,
      "grad_norm": 0.014607729180860774,
      "learning_rate": 1.4887640449438203e-07,
      "logits/chosen": -1.59375,
      "logits/rejected": -1.484375,
      "logps/chosen": -170.0,
      "logps/rejected": -262.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.96875,
      "rewards/margins": 11.9375,
      "rewards/rejected": -12.9375,
      "step": 290
    },
    {
      "epoch": 2.2727272727272725,
      "grad_norm": 0.009493204077863557,
      "learning_rate": 1.3483146067415728e-07,
      "logits/chosen": -1.609375,
      "logits/rejected": -1.546875,
      "logps/chosen": -160.0,
      "logps/rejected": -260.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.9921875,
      "rewards/margins": 12.125,
      "rewards/rejected": -13.125,
      "step": 300
    },
    {
      "epoch": 2.3484848484848486,
      "grad_norm": 0.003736203163504444,
      "learning_rate": 1.2078651685393259e-07,
      "logits/chosen": -1.578125,
      "logits/rejected": -1.5390625,
      "logps/chosen": -152.0,
      "logps/rejected": -268.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.0234375,
      "rewards/margins": 12.1875,
      "rewards/rejected": -13.25,
      "step": 310
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 0.0053623346535288585,
      "learning_rate": 1.0674157303370785e-07,
      "logits/chosen": -1.5546875,
      "logits/rejected": -1.4453125,
      "logps/chosen": -157.0,
      "logps/rejected": -260.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.1015625,
      "rewards/margins": 12.0,
      "rewards/rejected": -13.125,
      "step": 320
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.001064339902935614,
      "learning_rate": 9.269662921348314e-08,
      "logits/chosen": -1.546875,
      "logits/rejected": -1.515625,
      "logps/chosen": -164.0,
      "logps/rejected": -260.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.1953125,
      "rewards/margins": 12.1875,
      "rewards/rejected": -13.375,
      "step": 330
    },
    {
      "epoch": 2.5757575757575757,
      "grad_norm": 0.04406263550084857,
      "learning_rate": 7.865168539325842e-08,
      "logits/chosen": -1.5390625,
      "logits/rejected": -1.4921875,
      "logps/chosen": -164.0,
      "logps/rejected": -264.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -0.95703125,
      "rewards/margins": 12.6875,
      "rewards/rejected": -13.625,
      "step": 340
    },
    {
      "epoch": 2.6515151515151514,
      "grad_norm": 0.002620164881663117,
      "learning_rate": 6.460674157303371e-08,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.5078125,
      "logps/chosen": -144.0,
      "logps/rejected": -272.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.0390625,
      "rewards/margins": 12.5,
      "rewards/rejected": -13.5,
      "step": 350
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.016885587735988555,
      "learning_rate": 5.056179775280899e-08,
      "logits/chosen": -1.5625,
      "logits/rejected": -1.4765625,
      "logps/chosen": -160.0,
      "logps/rejected": -264.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.1484375,
      "rewards/margins": 12.1875,
      "rewards/rejected": -13.3125,
      "step": 360
    },
    {
      "epoch": 2.8030303030303028,
      "grad_norm": 0.01685236462277552,
      "learning_rate": 3.6516853932584266e-08,
      "logits/chosen": -1.5078125,
      "logits/rejected": -1.453125,
      "logps/chosen": -137.0,
      "logps/rejected": -262.0,
      "loss": 0.0,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.1015625,
      "rewards/margins": 12.4375,
      "rewards/rejected": -13.5,
      "step": 370
    },
    {
      "epoch": 2.878787878787879,
      "grad_norm": 0.0035252409898577407,
      "learning_rate": 2.2471910112359548e-08,
      "logits/chosen": -1.515625,
      "logits/rejected": -1.4609375,
      "logps/chosen": -142.0,
      "logps/rejected": -268.0,
      "loss": 0.0002,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.078125,
      "rewards/margins": 12.625,
      "rewards/rejected": -13.6875,
      "step": 380
    },
    {
      "epoch": 2.9545454545454546,
      "grad_norm": 0.002976821998771009,
      "learning_rate": 8.42696629213483e-09,
      "logits/chosen": -1.546875,
      "logits/rejected": -1.515625,
      "logps/chosen": -158.0,
      "logps/rejected": -268.0,
      "loss": 0.0001,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.1484375,
      "rewards/margins": 12.5625,
      "rewards/rejected": -13.6875,
      "step": 390
    },
    {
      "epoch": 3.0,
      "eval_logits/chosen": -1.5,
      "eval_logits/rejected": -1.4921875,
      "eval_logps/chosen": -149.0,
      "eval_logps/rejected": -264.0,
      "eval_loss": 6.22963925707154e-05,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": -1.046875,
      "eval_rewards/margins": 12.1875,
      "eval_rewards/rejected": -13.25,
      "eval_runtime": 14.9995,
      "eval_samples_per_second": 13.334,
      "eval_steps_per_second": 0.467,
      "step": 396
    },
    {
      "epoch": 3.0,
      "step": 396,
      "total_flos": 0.0,
      "train_loss": 0.041059525631782084,
      "train_runtime": 2403.0973,
      "train_samples_per_second": 5.243,
      "train_steps_per_second": 0.165
    }
  ],
  "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
}