{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.4925373134328357,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 63.20117586062747,
      "learning_rate": 5e-07,
      "logits/chosen": -2.727349281311035,
      "logits/rejected": -2.7335803508758545,
      "logps/chosen": -270.8079833984375,
      "logps/rejected": -222.7043914794922,
      "loss": 0.6897,
      "rewards/accuracies": 0.3187499940395355,
      "rewards/chosen": 0.013827224262058735,
      "rewards/margins": 0.000779601454269141,
      "rewards/rejected": 0.013047623448073864,
      "step": 5
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 50.76880252527562,
      "learning_rate": 1e-06,
      "logits/chosen": -2.686920166015625,
      "logits/rejected": -2.6769723892211914,
      "logps/chosen": -249.0262451171875,
      "logps/rejected": -231.17623901367188,
      "loss": 0.6468,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.47842222452163696,
      "rewards/margins": 0.08185982704162598,
      "rewards/rejected": 0.396562397480011,
      "step": 10
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 44.786597014715575,
      "learning_rate": 9.983100718730718e-07,
      "logits/chosen": -2.4991869926452637,
      "logits/rejected": -2.519619941711426,
      "logps/chosen": -257.5209045410156,
      "logps/rejected": -217.60446166992188,
      "loss": 0.645,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.3132820129394531,
      "rewards/margins": 0.6335776448249817,
      "rewards/rejected": 0.6797044277191162,
      "step": 15
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 88.88932587574585,
      "learning_rate": 9.932517109205849e-07,
      "logits/chosen": -2.3716320991516113,
      "logits/rejected": -2.350722551345825,
      "logps/chosen": -275.037353515625,
      "logps/rejected": -229.4193878173828,
      "loss": 0.639,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.544551134109497,
      "rewards/margins": 0.876867949962616,
      "rewards/rejected": 0.6676831245422363,
      "step": 20
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 42.980070421826184,
      "learning_rate": 9.848591102083375e-07,
      "logits/chosen": -2.2988502979278564,
      "logits/rejected": -2.288430690765381,
      "logps/chosen": -244.6092987060547,
      "logps/rejected": -221.5340118408203,
      "loss": 0.617,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.3750474452972412,
      "rewards/margins": 1.164804220199585,
      "rewards/rejected": 0.21024329960346222,
      "step": 25
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 37.576258990861845,
      "learning_rate": 9.731890013043367e-07,
      "logits/chosen": -2.3538706302642822,
      "logits/rejected": -2.3402347564697266,
      "logps/chosen": -244.2089080810547,
      "logps/rejected": -220.82180786132812,
      "loss": 0.5935,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 1.0666203498840332,
      "rewards/margins": 1.0009348392486572,
      "rewards/rejected": 0.06568538397550583,
      "step": 30
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 48.27486571141399,
      "learning_rate": 9.583202707897073e-07,
      "logits/chosen": -2.4315402507781982,
      "logits/rejected": -2.4107561111450195,
      "logps/chosen": -258.6941223144531,
      "logps/rejected": -200.2947998046875,
      "loss": 0.5815,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.854897677898407,
      "rewards/margins": 1.0769257545471191,
      "rewards/rejected": -0.2220279723405838,
      "step": 35
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 34.10499734328659,
      "learning_rate": 9.403534270080829e-07,
      "logits/chosen": -2.478212833404541,
      "logits/rejected": -2.4579110145568848,
      "logps/chosen": -247.2701416015625,
      "logps/rejected": -204.61563110351562,
      "loss": 0.5553,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.8650096654891968,
      "rewards/margins": 1.3998459577560425,
      "rewards/rejected": -0.5348363518714905,
      "step": 40
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 54.84196205672111,
      "learning_rate": 9.19409920658098e-07,
      "logits/chosen": -2.4562790393829346,
      "logits/rejected": -2.406428575515747,
      "logps/chosen": -251.5320587158203,
      "logps/rejected": -196.53280639648438,
      "loss": 0.6172,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.724992573261261,
      "rewards/margins": 0.909608006477356,
      "rewards/rejected": -0.18461540341377258,
      "step": 45
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 44.60950151362463,
      "learning_rate": 8.956313238215823e-07,
      "logits/chosen": -2.429295063018799,
      "logits/rejected": -2.4014017581939697,
      "logps/chosen": -267.9967041015625,
      "logps/rejected": -231.6461639404297,
      "loss": 0.5811,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.032130241394043,
      "rewards/margins": 1.233904242515564,
      "rewards/rejected": -0.20177388191223145,
      "step": 50
    },
    {
      "epoch": 0.746268656716418,
      "eval_logits/chosen": -2.379634141921997,
      "eval_logits/rejected": -2.3887269496917725,
      "eval_logps/chosen": -262.3047790527344,
      "eval_logps/rejected": -217.30076599121094,
      "eval_loss": 0.5699393153190613,
      "eval_rewards/accuracies": 0.7333333492279053,
      "eval_rewards/chosen": 0.5697149634361267,
      "eval_rewards/margins": 1.181886076927185,
      "eval_rewards/rejected": -0.6121711730957031,
      "eval_runtime": 125.8925,
      "eval_samples_per_second": 15.092,
      "eval_steps_per_second": 0.238,
      "step": 50
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 38.86022913172134,
      "learning_rate": 8.691783729769873e-07,
      "logits/chosen": -2.384519338607788,
      "logits/rejected": -2.369540214538574,
      "logps/chosen": -268.18524169921875,
      "logps/rejected": -256.36614990234375,
      "loss": 0.5328,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.911398708820343,
      "rewards/margins": 1.7182573080062866,
      "rewards/rejected": -0.8068588376045227,
      "step": 55
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 35.804216796075785,
      "learning_rate": 8.402298824670029e-07,
      "logits/chosen": -2.4195339679718018,
      "logits/rejected": -2.3873202800750732,
      "logps/chosen": -259.75775146484375,
      "logps/rejected": -212.31201171875,
      "loss": 0.533,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 1.215578317642212,
      "rewards/margins": 1.4803569316864014,
      "rewards/rejected": -0.2647787034511566,
      "step": 60
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 38.53205216239837,
      "learning_rate": 8.089815357650089e-07,
      "logits/chosen": -2.5004165172576904,
      "logits/rejected": -2.459744930267334,
      "logps/chosen": -245.0717315673828,
      "logps/rejected": -219.3514862060547,
      "loss": 0.5168,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.0596349239349365,
      "rewards/margins": 1.2952834367752075,
      "rewards/rejected": -0.23564846813678741,
      "step": 65
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 18.93894037354211,
      "learning_rate": 7.756445627110522e-07,
      "logits/chosen": -2.5253801345825195,
      "logits/rejected": -2.498584032058716,
      "logps/chosen": -256.0594482421875,
      "logps/rejected": -230.0350799560547,
      "loss": 0.3642,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": 1.1649268865585327,
      "rewards/margins": 2.232837200164795,
      "rewards/rejected": -1.067910075187683,
      "step": 70
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 20.704224755579,
      "learning_rate": 7.404443116588547e-07,
      "logits/chosen": -2.500047206878662,
      "logits/rejected": -2.477970838546753,
      "logps/chosen": -232.2460479736328,
      "logps/rejected": -218.29568481445312,
      "loss": 0.2398,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 1.6713672876358032,
      "rewards/margins": 2.5339531898498535,
      "rewards/rejected": -0.8625856637954712,
      "step": 75
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 18.90792590059534,
      "learning_rate": 7.036187261857288e-07,
      "logits/chosen": -2.4836063385009766,
      "logits/rejected": -2.4460768699645996,
      "logps/chosen": -245.5286407470703,
      "logps/rejected": -243.6932830810547,
      "loss": 0.2302,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.2828805446624756,
      "rewards/margins": 3.4864933490753174,
      "rewards/rejected": -1.2036125659942627,
      "step": 80
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 20.90479576440656,
      "learning_rate": 6.654167366624008e-07,
      "logits/chosen": -2.427332878112793,
      "logits/rejected": -2.425985813140869,
      "logps/chosen": -233.1042938232422,
      "logps/rejected": -215.51437377929688,
      "loss": 0.2332,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.7251182794570923,
      "rewards/margins": 3.2692294120788574,
      "rewards/rejected": -1.5441116094589233,
      "step": 85
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 23.387687179198938,
      "learning_rate": 6.260965775552713e-07,
      "logits/chosen": -2.4304587841033936,
      "logits/rejected": -2.4162423610687256,
      "logps/chosen": -247.0415802001953,
      "logps/rejected": -230.84005737304688,
      "loss": 0.253,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.8513774871826172,
      "rewards/margins": 3.377821445465088,
      "rewards/rejected": -1.5264440774917603,
      "step": 90
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 25.290227374387985,
      "learning_rate": 5.859240418356614e-07,
      "logits/chosen": -2.4067845344543457,
      "logits/rejected": -2.403015375137329,
      "logps/chosen": -229.62283325195312,
      "logps/rejected": -227.48825073242188,
      "loss": 0.2631,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.354150652885437,
      "rewards/margins": 3.5103206634521484,
      "rewards/rejected": -2.156169891357422,
      "step": 95
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 24.465863734128728,
      "learning_rate": 5.451706842957421e-07,
      "logits/chosen": -2.4527347087860107,
      "logits/rejected": -2.454359292984009,
      "logps/chosen": -250.5390625,
      "logps/rejected": -231.2834930419922,
      "loss": 0.2898,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.1465234756469727,
      "rewards/margins": 3.222960948944092,
      "rewards/rejected": -1.076436996459961,
      "step": 100
    },
    {
      "epoch": 1.4925373134328357,
      "eval_logits/chosen": -2.4716811180114746,
      "eval_logits/rejected": -2.4747350215911865,
      "eval_logps/chosen": -255.55557250976562,
      "eval_logps/rejected": -216.7303009033203,
      "eval_loss": 0.5633274912834167,
      "eval_rewards/accuracies": 0.7583333253860474,
      "eval_rewards/chosen": 1.2446327209472656,
      "eval_rewards/margins": 1.799760103225708,
      "eval_rewards/rejected": -0.5551272034645081,
      "eval_runtime": 125.768,
      "eval_samples_per_second": 15.107,
      "eval_steps_per_second": 0.239,
      "step": 100
    }
  ],
  "logging_steps": 5,
  "max_steps": 201,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1178822762299392.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}