{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9345794392523364,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04672897196261682,
      "grad_norm": 68.30027051845477,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7332546710968018,
      "logits/rejected": -2.7345798015594482,
      "logps/chosen": -246.6896209716797,
      "logps/rejected": -212.1261749267578,
      "loss": 0.6905,
      "rewards/accuracies": 0.33125001192092896,
      "rewards/chosen": 0.01393141783773899,
      "rewards/margins": 0.007549271918833256,
      "rewards/rejected": 0.006382145918905735,
      "step": 5
    },
    {
      "epoch": 0.09345794392523364,
      "grad_norm": 47.12543235245713,
      "learning_rate": 1e-06,
      "logits/chosen": -2.690162181854248,
      "logits/rejected": -2.6820242404937744,
      "logps/chosen": -237.89517211914062,
      "logps/rejected": -218.36865234375,
      "loss": 0.6502,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.4055875837802887,
      "rewards/margins": 0.14588704705238342,
      "rewards/rejected": 0.2597005069255829,
      "step": 10
    },
    {
      "epoch": 0.14018691588785046,
      "grad_norm": 49.480284694367455,
      "learning_rate": 9.993623730611148e-07,
      "logits/chosen": -2.56203031539917,
      "logits/rejected": -2.5295283794403076,
      "logps/chosen": -215.8556365966797,
      "logps/rejected": -206.70339965820312,
      "loss": 0.6521,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.2055972814559937,
      "rewards/margins": 0.3723347783088684,
      "rewards/rejected": 0.8332624435424805,
      "step": 15
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 49.42485896222818,
      "learning_rate": 9.97451118516912e-07,
      "logits/chosen": -2.4396374225616455,
      "logits/rejected": -2.4257354736328125,
      "logps/chosen": -233.15951538085938,
      "logps/rejected": -192.3885955810547,
      "loss": 0.6234,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.3042352199554443,
      "rewards/margins": 0.6797620058059692,
      "rewards/rejected": 0.6244731545448303,
      "step": 20
    },
    {
      "epoch": 0.2336448598130841,
      "grad_norm": 43.823143196477226,
      "learning_rate": 9.94271111036929e-07,
      "logits/chosen": -2.318066120147705,
      "logits/rejected": -2.304948091506958,
      "logps/chosen": -241.57180786132812,
      "logps/rejected": -191.6623992919922,
      "loss": 0.6435,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.0954402685165405,
      "rewards/margins": 0.5757324695587158,
      "rewards/rejected": 0.5197077393531799,
      "step": 25
    },
    {
      "epoch": 0.2803738317757009,
      "grad_norm": 39.78735079002501,
      "learning_rate": 9.898304612549066e-07,
      "logits/chosen": -2.3289685249328613,
      "logits/rejected": -2.326596736907959,
      "logps/chosen": -255.0546875,
      "logps/rejected": -215.7193145751953,
      "loss": 0.6003,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 1.374457597732544,
      "rewards/margins": 1.0344688892364502,
      "rewards/rejected": 0.33998847007751465,
      "step": 30
    },
    {
      "epoch": 0.32710280373831774,
      "grad_norm": 43.394927490331185,
      "learning_rate": 9.841404950825536e-07,
      "logits/chosen": -2.3670554161071777,
      "logits/rejected": -2.3704476356506348,
      "logps/chosen": -242.648681640625,
      "logps/rejected": -209.353759765625,
      "loss": 0.5919,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.954569935798645,
      "rewards/margins": 1.0306309461593628,
      "rewards/rejected": -0.07606112957000732,
      "step": 35
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 46.37512759709472,
      "learning_rate": 9.77215724822721e-07,
      "logits/chosen": -2.437983274459839,
      "logits/rejected": -2.465634822845459,
      "logps/chosen": -254.10226440429688,
      "logps/rejected": -226.6571502685547,
      "loss": 0.648,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.6253764033317566,
      "rewards/margins": 0.8199795484542847,
      "rewards/rejected": -0.1946031153202057,
      "step": 40
    },
    {
      "epoch": 0.4205607476635514,
      "grad_norm": 44.696777081131444,
      "learning_rate": 9.69073812155662e-07,
      "logits/chosen": -2.4394752979278564,
      "logits/rejected": -2.4159750938415527,
      "logps/chosen": -246.6658172607422,
      "logps/rejected": -240.363525390625,
      "loss": 0.5918,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.7499775886535645,
      "rewards/margins": 1.0236563682556152,
      "rewards/rejected": -0.2736787497997284,
      "step": 45
    },
    {
      "epoch": 0.4672897196261682,
      "grad_norm": 38.923198317781896,
      "learning_rate": 9.597355230927788e-07,
      "logits/chosen": -2.4589362144470215,
      "logits/rejected": -2.412837505340576,
      "logps/chosen": -264.4033203125,
      "logps/rejected": -213.1973876953125,
      "loss": 0.576,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": 1.1025102138519287,
      "rewards/margins": 1.2783615589141846,
      "rewards/rejected": -0.175851508975029,
      "step": 50
    },
    {
      "epoch": 0.4672897196261682,
      "eval_logits/chosen": -2.496382474899292,
      "eval_logits/rejected": -2.4706027507781982,
      "eval_logps/chosen": -223.9072723388672,
      "eval_logps/rejected": -212.02557373046875,
      "eval_loss": 0.5725350379943848,
      "eval_rewards/accuracies": 0.7317708134651184,
      "eval_rewards/chosen": 0.8729944229125977,
      "eval_rewards/margins": 0.9930557608604431,
      "eval_rewards/rejected": -0.12006133049726486,
      "eval_runtime": 203.179,
      "eval_samples_per_second": 14.962,
      "eval_steps_per_second": 0.236,
      "step": 50
    },
    {
      "epoch": 0.514018691588785,
      "grad_norm": 32.87244621664002,
      "learning_rate": 9.4922467501275e-07,
      "logits/chosen": -2.5147769451141357,
      "logits/rejected": -2.4755141735076904,
      "logps/chosen": -248.0165557861328,
      "logps/rejected": -201.71705627441406,
      "loss": 0.5771,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.9974912405014038,
      "rewards/margins": 1.1693007946014404,
      "rewards/rejected": -0.17180955410003662,
      "step": 55
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 41.4567929981585,
      "learning_rate": 9.375680759151206e-07,
      "logits/chosen": -2.517082452774048,
      "logits/rejected": -2.506866931915283,
      "logps/chosen": -239.5508575439453,
      "logps/rejected": -215.45339965820312,
      "loss": 0.5856,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.7699365615844727,
      "rewards/margins": 1.1056894063949585,
      "rewards/rejected": -0.33575284481048584,
      "step": 60
    },
    {
      "epoch": 0.6074766355140186,
      "grad_norm": 42.95403443814742,
      "learning_rate": 9.247954560462927e-07,
      "logits/chosen": -2.575209140777588,
      "logits/rejected": -2.5410640239715576,
      "logps/chosen": -235.7403106689453,
      "logps/rejected": -221.4304656982422,
      "loss": 0.5815,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.6247920989990234,
      "rewards/margins": 1.079342246055603,
      "rewards/rejected": -0.454550176858902,
      "step": 65
    },
    {
      "epoch": 0.6542056074766355,
      "grad_norm": 39.08891854742411,
      "learning_rate": 9.109393920723001e-07,
      "logits/chosen": -2.584603786468506,
      "logits/rejected": -2.5690274238586426,
      "logps/chosen": -267.28839111328125,
      "logps/rejected": -231.07785034179688,
      "loss": 0.5626,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 0.6171663403511047,
      "rewards/margins": 1.1559643745422363,
      "rewards/rejected": -0.5387979745864868,
      "step": 70
    },
    {
      "epoch": 0.7009345794392523,
      "grad_norm": 44.455775233540244,
      "learning_rate": 8.960352239917699e-07,
      "logits/chosen": -2.566195011138916,
      "logits/rejected": -2.549044132232666,
      "logps/chosen": -246.45858764648438,
      "logps/rejected": -222.82357788085938,
      "loss": 0.5982,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.6170978546142578,
      "rewards/margins": 1.267897367477417,
      "rewards/rejected": -0.6507996916770935,
      "step": 75
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 32.16569720584001,
      "learning_rate": 8.801209650009814e-07,
      "logits/chosen": -2.4980108737945557,
      "logits/rejected": -2.4944815635681152,
      "logps/chosen": -251.861572265625,
      "logps/rejected": -227.2343292236328,
      "loss": 0.4956,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.6364807486534119,
      "rewards/margins": 1.5403988361358643,
      "rewards/rejected": -0.9039181470870972,
      "step": 80
    },
    {
      "epoch": 0.794392523364486,
      "grad_norm": 40.106812622105444,
      "learning_rate": 8.632372045409141e-07,
      "logits/chosen": -2.510608196258545,
      "logits/rejected": -2.50504994392395,
      "logps/chosen": -264.02777099609375,
      "logps/rejected": -228.0398406982422,
      "loss": 0.5527,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.6684669852256775,
      "rewards/margins": 1.134697675704956,
      "rewards/rejected": -0.4662308692932129,
      "step": 85
    },
    {
      "epoch": 0.8411214953271028,
      "grad_norm": 41.44913213337072,
      "learning_rate": 8.454270047735642e-07,
      "logits/chosen": -2.4945034980773926,
      "logits/rejected": -2.4969065189361572,
      "logps/chosen": -266.8227844238281,
      "logps/rejected": -218.293701171875,
      "loss": 0.5291,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 0.4074540138244629,
      "rewards/margins": 1.6132476329803467,
      "rewards/rejected": -1.2057937383651733,
      "step": 90
    },
    {
      "epoch": 0.8878504672897196,
      "grad_norm": 38.304999982614596,
      "learning_rate": 8.267357907515661e-07,
      "logits/chosen": -2.4969687461853027,
      "logits/rejected": -2.4846863746643066,
      "logps/chosen": -253.90396118164062,
      "logps/rejected": -226.99136352539062,
      "loss": 0.531,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.14072401821613312,
      "rewards/margins": 1.3943620920181274,
      "rewards/rejected": -1.5350860357284546,
      "step": 95
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 35.63875623202816,
      "learning_rate": 8.072112345612433e-07,
      "logits/chosen": -2.468038558959961,
      "logits/rejected": -2.446315288543701,
      "logps/chosen": -242.86328125,
      "logps/rejected": -230.5182342529297,
      "loss": 0.508,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.023806383833289146,
      "rewards/margins": 1.7525684833526611,
      "rewards/rejected": -1.7287616729736328,
      "step": 100
    },
    {
      "epoch": 0.9345794392523364,
      "eval_logits/chosen": -2.460448980331421,
      "eval_logits/rejected": -2.4439258575439453,
      "eval_logps/chosen": -234.71864318847656,
      "eval_logps/rejected": -228.7583770751953,
      "eval_loss": 0.5507261753082275,
      "eval_rewards/accuracies": 0.7864583134651184,
      "eval_rewards/chosen": -0.2081437110900879,
      "eval_rewards/margins": 1.5851994752883911,
      "eval_rewards/rejected": -1.793343186378479,
      "eval_runtime": 202.7006,
      "eval_samples_per_second": 14.997,
      "eval_steps_per_second": 0.237,
      "step": 100
    }
  ],
  "logging_steps": 5,
  "max_steps": 321,
  "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
}