{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.429864253393665,
  "eval_steps": 500,
  "global_step": 150,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.36199095022624433,
      "grad_norm": 0.5034113526344299,
      "learning_rate": 9.259259259259259e-07,
      "logits/chosen": 1.8096959590911865,
      "logits/rejected": 1.7158682346343994,
      "logps/chosen": -106.03636169433594,
      "logps/rejected": -84.16211700439453,
      "loss": 0.695,
      "rewards/accuracies": 0.3499999940395355,
      "rewards/chosen": -0.00430720392614603,
      "rewards/margins": 0.00014101080887485296,
      "rewards/rejected": -0.004448213614523411,
      "step": 10
    },
    {
      "epoch": 0.7239819004524887,
      "grad_norm": 0.4689481854438782,
      "learning_rate": 1.8518518518518519e-06,
      "logits/chosen": 1.6904691457748413,
      "logits/rejected": 1.6791242361068726,
      "logps/chosen": -82.1568603515625,
      "logps/rejected": -81.06039428710938,
      "loss": 0.6942,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.002815738320350647,
      "rewards/margins": 0.005291810259222984,
      "rewards/rejected": -0.008107547648251057,
      "step": 20
    },
    {
      "epoch": 1.085972850678733,
      "grad_norm": 0.5041429996490479,
      "learning_rate": 2.7777777777777783e-06,
      "logits/chosen": 1.877396583557129,
      "logits/rejected": 1.9337213039398193,
      "logps/chosen": -76.86979675292969,
      "logps/rejected": -86.98431396484375,
      "loss": 0.6937,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.002632059855386615,
      "rewards/margins": -0.00260960147716105,
      "rewards/rejected": -2.245856376248412e-05,
      "step": 30
    },
    {
      "epoch": 1.4479638009049773,
      "grad_norm": 0.49786925315856934,
      "learning_rate": 3.7037037037037037e-06,
      "logits/chosen": 1.622865915298462,
      "logits/rejected": 1.649457573890686,
      "logps/chosen": -77.10165405273438,
      "logps/rejected": -75.56939697265625,
      "loss": 0.6927,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": -0.005956621374934912,
      "rewards/margins": -0.0038306519854813814,
      "rewards/rejected": -0.0021259686909615993,
      "step": 40
    },
    {
      "epoch": 1.8099547511312217,
      "grad_norm": 0.6152529120445251,
      "learning_rate": 4.62962962962963e-06,
      "logits/chosen": 1.7763454914093018,
      "logits/rejected": 1.8556989431381226,
      "logps/chosen": -84.42935943603516,
      "logps/rejected": -97.08312225341797,
      "loss": 0.6937,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.008006314747035503,
      "rewards/margins": 0.005445700138807297,
      "rewards/rejected": 0.0025606155395507812,
      "step": 50
    },
    {
      "epoch": 2.171945701357466,
      "grad_norm": 0.539294958114624,
      "learning_rate": 4.998119881260576e-06,
      "logits/chosen": 1.7218296527862549,
      "logits/rejected": 1.759447455406189,
      "logps/chosen": -71.67170715332031,
      "logps/rejected": -86.7840805053711,
      "loss": 0.6935,
      "rewards/accuracies": 0.4625000059604645,
      "rewards/chosen": 0.0036853079218417406,
      "rewards/margins": -0.002186889760196209,
      "rewards/rejected": 0.005872196517884731,
      "step": 60
    },
    {
      "epoch": 2.5339366515837103,
      "grad_norm": 0.5369167327880859,
      "learning_rate": 4.9866405060165044e-06,
      "logits/chosen": 1.7708947658538818,
      "logits/rejected": 1.8614221811294556,
      "logps/chosen": -86.73070526123047,
      "logps/rejected": -83.09151458740234,
      "loss": 0.6909,
      "rewards/accuracies": 0.5,
      "rewards/chosen": 0.008487144485116005,
      "rewards/margins": 0.004935932345688343,
      "rewards/rejected": 0.0035512119065970182,
      "step": 70
    },
    {
      "epoch": 2.8959276018099547,
      "grad_norm": 0.5544038414955139,
      "learning_rate": 4.964774158361991e-06,
      "logits/chosen": 1.772223711013794,
      "logits/rejected": 1.834602952003479,
      "logps/chosen": -69.32654571533203,
      "logps/rejected": -77.55828094482422,
      "loss": 0.6943,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": -0.012612676247954369,
      "rewards/margins": -0.011703734286129475,
      "rewards/rejected": -0.0009089424274861813,
      "step": 80
    },
    {
      "epoch": 3.257918552036199,
      "grad_norm": 0.5990457534790039,
      "learning_rate": 4.93261217644956e-06,
      "logits/chosen": 1.8591846227645874,
      "logits/rejected": 1.8804810047149658,
      "logps/chosen": -79.4209213256836,
      "logps/rejected": -76.4483642578125,
      "loss": 0.6927,
      "rewards/accuracies": 0.4749999940395355,
      "rewards/chosen": -0.0016643519047647715,
      "rewards/margins": 0.006901158951222897,
      "rewards/rejected": -0.008565512485802174,
      "step": 90
    },
    {
      "epoch": 3.6199095022624435,
      "grad_norm": 0.6173596978187561,
      "learning_rate": 4.8902889044347e-06,
      "logits/chosen": 1.777044653892517,
      "logits/rejected": 1.7735064029693604,
      "logps/chosen": -85.6143569946289,
      "logps/rejected": -79.8615493774414,
      "loss": 0.6894,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": -0.0019477702444419265,
      "rewards/margins": 0.002171102212741971,
      "rewards/rejected": -0.004118871875107288,
      "step": 100
    },
    {
      "epoch": 3.981900452488688,
      "grad_norm": 0.6671907901763916,
      "learning_rate": 4.837981131305475e-06,
      "logits/chosen": 1.6935539245605469,
      "logits/rejected": 1.7035915851593018,
      "logps/chosen": -83.69059753417969,
      "logps/rejected": -89.67396545410156,
      "loss": 0.6907,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": -0.005695090629160404,
      "rewards/margins": 0.0027313486207276583,
      "rewards/rejected": -0.00842643715441227,
      "step": 110
    },
    {
      "epoch": 4.343891402714932,
      "grad_norm": 0.9499974250793457,
      "learning_rate": 4.775907352415367e-06,
      "logits/chosen": 1.801300287246704,
      "logits/rejected": 1.8080482482910156,
      "logps/chosen": -92.2596206665039,
      "logps/rejected": -85.3951416015625,
      "loss": 0.6863,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.000912027433514595,
      "rewards/margins": 0.017921410501003265,
      "rewards/rejected": -0.01700938306748867,
      "step": 120
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 0.6524670124053955,
      "learning_rate": 4.70432685680402e-06,
      "logits/chosen": 1.8659807443618774,
      "logits/rejected": 1.8858706951141357,
      "logps/chosen": -78.38592529296875,
      "logps/rejected": -81.4170913696289,
      "loss": 0.6873,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.002340498147532344,
      "rewards/margins": 0.014832529239356518,
      "rewards/rejected": -0.01249203272163868,
      "step": 130
    },
    {
      "epoch": 5.067873303167421,
      "grad_norm": 0.7461792826652527,
      "learning_rate": 4.623538644118244e-06,
      "logits/chosen": 1.7999374866485596,
      "logits/rejected": 1.869183897972107,
      "logps/chosen": -71.37133026123047,
      "logps/rejected": -83.36344909667969,
      "loss": 0.6829,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.013370734639465809,
      "rewards/margins": 0.013819178566336632,
      "rewards/rejected": -0.027189914137125015,
      "step": 140
    },
    {
      "epoch": 5.429864253393665,
      "grad_norm": 0.6743089556694031,
      "learning_rate": 4.533880175657419e-06,
      "logits/chosen": 1.8166606426239014,
      "logits/rejected": 1.7869322299957275,
      "logps/chosen": -87.78181457519531,
      "logps/rejected": -71.32334899902344,
      "loss": 0.6758,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.0059204003773629665,
      "rewards/margins": 0.03691873326897621,
      "rewards/rejected": -0.04283912852406502,
      "step": 150
    }
  ],
  "logging_steps": 10,
  "max_steps": 540,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.137690753362821e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}