{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.984255147355672,
  "eval_steps": 500,
  "global_step": 462,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.12918853451756157,
      "grad_norm": 18.75351905822754,
      "learning_rate": 2.127659574468085e-07,
      "logits/chosen": -1.451206922531128,
      "logits/rejected": -1.453438401222229,
      "logps/chosen": -1178.2191162109375,
      "logps/rejected": -1103.279541015625,
      "loss": 0.6962,
      "rewards/accuracies": 0.4242187440395355,
      "rewards/chosen": -0.0020913523621857166,
      "rewards/margins": -0.008680100552737713,
      "rewards/rejected": 0.006588748190551996,
      "step": 20
    },
    {
      "epoch": 0.25837706903512314,
      "grad_norm": 18.161069869995117,
      "learning_rate": 4.25531914893617e-07,
      "logits/chosen": -1.4470716714859009,
      "logits/rejected": -1.4471380710601807,
      "logps/chosen": -1174.496337890625,
      "logps/rejected": -1101.42333984375,
      "loss": 0.6931,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.012329095974564552,
      "rewards/margins": 0.002261379035189748,
      "rewards/rejected": 0.010067718103528023,
      "step": 40
    },
    {
      "epoch": 0.3875656035526847,
      "grad_norm": 18.111698150634766,
      "learning_rate": 4.987903778327268e-07,
      "logits/chosen": -1.4492727518081665,
      "logits/rejected": -1.454036831855774,
      "logps/chosen": -1144.927490234375,
      "logps/rejected": -1060.2935791015625,
      "loss": 0.6842,
      "rewards/accuracies": 0.5726562738418579,
      "rewards/chosen": 0.04902833700180054,
      "rewards/margins": 0.026971647515892982,
      "rewards/rejected": 0.022056687623262405,
      "step": 60
    },
    {
      "epoch": 0.5167541380702463,
      "grad_norm": 17.73369598388672,
      "learning_rate": 4.922396431162129e-07,
      "logits/chosen": -1.4595801830291748,
      "logits/rejected": -1.4629844427108765,
      "logps/chosen": -1171.3333740234375,
      "logps/rejected": -1109.41455078125,
      "loss": 0.6694,
      "rewards/accuracies": 0.610156238079071,
      "rewards/chosen": 0.1019735336303711,
      "rewards/margins": 0.06121420860290527,
      "rewards/rejected": 0.04075933247804642,
      "step": 80
    },
    {
      "epoch": 0.6459426725878078,
      "grad_norm": 16.904024124145508,
      "learning_rate": 4.801467490723401e-07,
      "logits/chosen": -1.4490680694580078,
      "logits/rejected": -1.4522110223770142,
      "logps/chosen": -1182.560791015625,
      "logps/rejected": -1114.6134033203125,
      "loss": 0.6552,
      "rewards/accuracies": 0.6148437261581421,
      "rewards/chosen": 0.07732206583023071,
      "rewards/margins": 0.11570864915847778,
      "rewards/rejected": -0.03838658332824707,
      "step": 100
    },
    {
      "epoch": 0.7751312071053694,
      "grad_norm": 17.63266372680664,
      "learning_rate": 4.627883669538311e-07,
      "logits/chosen": -1.4325090646743774,
      "logits/rejected": -1.4361627101898193,
      "logps/chosen": -1182.9852294921875,
      "logps/rejected": -1119.953125,
      "loss": 0.647,
      "rewards/accuracies": 0.6351562738418579,
      "rewards/chosen": 0.06345769017934799,
      "rewards/margins": 0.14128440618515015,
      "rewards/rejected": -0.07782672345638275,
      "step": 120
    },
    {
      "epoch": 0.9043197416229309,
      "grad_norm": 17.77622413635254,
      "learning_rate": 4.405616362137017e-07,
      "logits/chosen": -1.4502944946289062,
      "logits/rejected": -1.4528261423110962,
      "logps/chosen": -1160.4075927734375,
      "logps/rejected": -1093.4769287109375,
      "loss": 0.6328,
      "rewards/accuracies": 0.62109375,
      "rewards/chosen": 0.04682188481092453,
      "rewards/margins": 0.19679759442806244,
      "rewards/rejected": -0.1499757319688797,
      "step": 140
    },
    {
      "epoch": 1.0335082761404926,
      "grad_norm": 16.421310424804688,
      "learning_rate": 4.139750784196997e-07,
      "logits/chosen": -1.4416507482528687,
      "logits/rejected": -1.4441019296646118,
      "logps/chosen": -1162.17919921875,
      "logps/rejected": -1097.062744140625,
      "loss": 0.6268,
      "rewards/accuracies": 0.6539062261581421,
      "rewards/chosen": 0.017796631902456284,
      "rewards/margins": 0.2243949919939041,
      "rewards/rejected": -0.20659832656383514,
      "step": 160
    },
    {
      "epoch": 1.1626968106580542,
      "grad_norm": 16.477689743041992,
      "learning_rate": 3.836369628764067e-07,
      "logits/chosen": -1.4492335319519043,
      "logits/rejected": -1.4500490427017212,
      "logps/chosen": -1149.756103515625,
      "logps/rejected": -1079.184326171875,
      "loss": 0.5971,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.020682932808995247,
      "rewards/margins": 0.3120761513710022,
      "rewards/rejected": -0.3327590823173523,
      "step": 180
    },
    {
      "epoch": 1.2918853451756156,
      "grad_norm": 16.82990074157715,
      "learning_rate": 3.5024139013594445e-07,
      "logits/chosen": -1.4431254863739014,
      "logits/rejected": -1.4477680921554565,
      "logps/chosen": -1168.659912109375,
      "logps/rejected": -1104.318115234375,
      "loss": 0.581,
      "rewards/accuracies": 0.7085937261581421,
      "rewards/chosen": -0.008671097457408905,
      "rewards/margins": 0.3281632959842682,
      "rewards/rejected": -0.3368344008922577,
      "step": 200
    },
    {
      "epoch": 1.4210738796931772,
      "grad_norm": 16.532955169677734,
      "learning_rate": 3.1455241179026165e-07,
      "logits/chosen": -1.4374616146087646,
      "logits/rejected": -1.4397255182266235,
      "logps/chosen": -1178.955810546875,
      "logps/rejected": -1098.697021484375,
      "loss": 0.5797,
      "rewards/accuracies": 0.711718738079071,
      "rewards/chosen": -0.046892568469047546,
      "rewards/margins": 0.3680500090122223,
      "rewards/rejected": -0.41494256258010864,
      "step": 220
    },
    {
      "epoch": 1.5502624142107388,
      "grad_norm": 15.883248329162598,
      "learning_rate": 2.7738654986555523e-07,
      "logits/chosen": -1.4424382448196411,
      "logits/rejected": -1.446033239364624,
      "logps/chosen": -1131.3668212890625,
      "logps/rejected": -1066.152099609375,
      "loss": 0.5729,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": -0.07887513935565948,
      "rewards/margins": 0.3466903865337372,
      "rewards/rejected": -0.42556554079055786,
      "step": 240
    },
    {
      "epoch": 1.6794509487283005,
      "grad_norm": 16.528196334838867,
      "learning_rate": 2.3959411575460777e-07,
      "logits/chosen": -1.437410831451416,
      "logits/rejected": -1.4404833316802979,
      "logps/chosen": -1151.679931640625,
      "logps/rejected": -1087.936279296875,
      "loss": 0.5745,
      "rewards/accuracies": 0.7210937738418579,
      "rewards/chosen": -0.05212847515940666,
      "rewards/margins": 0.4079786241054535,
      "rewards/rejected": -0.46010708808898926,
      "step": 260
    },
    {
      "epoch": 1.8086394832458619,
      "grad_norm": 16.318754196166992,
      "learning_rate": 2.02039756087992e-07,
      "logits/chosen": -1.455380916595459,
      "logits/rejected": -1.4586334228515625,
      "logps/chosen": -1133.935791015625,
      "logps/rejected": -1072.302734375,
      "loss": 0.5721,
      "rewards/accuracies": 0.73046875,
      "rewards/chosen": -0.06314127147197723,
      "rewards/margins": 0.40634602308273315,
      "rewards/rejected": -0.4694872796535492,
      "step": 280
    },
    {
      "epoch": 1.9378280177634235,
      "grad_norm": 16.61640739440918,
      "learning_rate": 1.655826706318234e-07,
      "logits/chosen": -1.4276014566421509,
      "logits/rejected": -1.4284788370132446,
      "logps/chosen": -1142.511962890625,
      "logps/rejected": -1087.729248046875,
      "loss": 0.564,
      "rewards/accuracies": 0.7367187738418579,
      "rewards/chosen": -0.07105865329504013,
      "rewards/margins": 0.41813844442367554,
      "rewards/rejected": -0.4891970753669739,
      "step": 300
    },
    {
      "epoch": 2.067016552280985,
      "grad_norm": 14.662156105041504,
      "learning_rate": 1.3105695480339204e-07,
      "logits/chosen": -1.4481165409088135,
      "logits/rejected": -1.4495656490325928,
      "logps/chosen": -1163.4542236328125,
      "logps/rejected": -1094.1142578125,
      "loss": 0.5458,
      "rewards/accuracies": 0.76171875,
      "rewards/chosen": -0.022224558517336845,
      "rewards/margins": 0.4889918267726898,
      "rewards/rejected": -0.5112164616584778,
      "step": 320
    },
    {
      "epoch": 2.1962050867985465,
      "grad_norm": 17.926048278808594,
      "learning_rate": 9.925251654489414e-08,
      "logits/chosen": -1.443882703781128,
      "logits/rejected": -1.4449208974838257,
      "logps/chosen": -1168.8966064453125,
      "logps/rejected": -1111.7706298828125,
      "loss": 0.5392,
      "rewards/accuracies": 0.784375011920929,
      "rewards/chosen": 0.0047746943309903145,
      "rewards/margins": 0.531124472618103,
      "rewards/rejected": -0.5263497829437256,
      "step": 340
    },
    {
      "epoch": 2.3253936213161084,
      "grad_norm": 15.895486831665039,
      "learning_rate": 7.089700415484206e-08,
      "logits/chosen": -1.4399263858795166,
      "logits/rejected": -1.4458813667297363,
      "logps/chosen": -1162.751220703125,
      "logps/rejected": -1100.962158203125,
      "loss": 0.5311,
      "rewards/accuracies": 0.789843738079071,
      "rewards/chosen": 0.017337169498205185,
      "rewards/margins": 0.5304254293441772,
      "rewards/rejected": -0.5130882859230042,
      "step": 360
    },
    {
      "epoch": 2.45458215583367,
      "grad_norm": 14.746881484985352,
      "learning_rate": 4.663915854720396e-08,
      "logits/chosen": -1.4447704553604126,
      "logits/rejected": -1.4479763507843018,
      "logps/chosen": -1171.16748046875,
      "logps/rejected": -1107.5540771484375,
      "loss": 0.5343,
      "rewards/accuracies": 0.754687488079071,
      "rewards/chosen": -0.04202534630894661,
      "rewards/margins": 0.5067408084869385,
      "rewards/rejected": -0.5487660765647888,
      "step": 380
    },
    {
      "epoch": 2.583770690351231,
      "grad_norm": 14.920990943908691,
      "learning_rate": 2.7033970819087397e-08,
      "logits/chosen": -1.4494940042495728,
      "logits/rejected": -1.452332854270935,
      "logps/chosen": -1167.1435546875,
      "logps/rejected": -1098.287841796875,
      "loss": 0.5352,
      "rewards/accuracies": 0.746874988079071,
      "rewards/chosen": -0.007791751530021429,
      "rewards/margins": 0.47063732147216797,
      "rewards/rejected": -0.47842907905578613,
      "step": 400
    },
    {
      "epoch": 2.712959224868793,
      "grad_norm": 15.134760856628418,
      "learning_rate": 1.2529984704433922e-08,
      "logits/chosen": -1.4471887350082397,
      "logits/rejected": -1.452412724494934,
      "logps/chosen": -1167.8065185546875,
      "logps/rejected": -1100.0762939453125,
      "loss": 0.5342,
      "rewards/accuracies": 0.7789062261581421,
      "rewards/chosen": -0.027560725808143616,
      "rewards/margins": 0.5049134492874146,
      "rewards/rejected": -0.532474160194397,
      "step": 420
    },
    {
      "epoch": 2.8421477593863544,
      "grad_norm": 15.015375137329102,
      "learning_rate": 3.4590344187135634e-09,
      "logits/chosen": -1.445507287979126,
      "logits/rejected": -1.4463194608688354,
      "logps/chosen": -1124.656982421875,
      "logps/rejected": -1059.774169921875,
      "loss": 0.5244,
      "rewards/accuracies": 0.7710937261581421,
      "rewards/chosen": -0.0609331913292408,
      "rewards/margins": 0.5137851238250732,
      "rewards/rejected": -0.5747183561325073,
      "step": 440
    },
    {
      "epoch": 2.9713362939039163,
      "grad_norm": 14.278298377990723,
      "learning_rate": 2.8652680713725507e-11,
      "logits/chosen": -1.4286974668502808,
      "logits/rejected": -1.4289166927337646,
      "logps/chosen": -1178.025634765625,
      "logps/rejected": -1113.932861328125,
      "loss": 0.5264,
      "rewards/accuracies": 0.784375011920929,
      "rewards/chosen": -0.011726352386176586,
      "rewards/margins": 0.5253415703773499,
      "rewards/rejected": -0.5370678901672363,
      "step": 460
    }
  ],
  "logging_steps": 20,
  "max_steps": 462,
  "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": 4,
  "trial_name": null,
  "trial_params": null
}