{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.955223880597015,
  "eval_steps": 50,
  "global_step": 99,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 57.93605166188967,
      "learning_rate": 5e-07,
      "logits/chosen": -2.716432571411133,
      "logits/rejected": -2.716773748397827,
      "logps/chosen": -277.5178527832031,
      "logps/rejected": -191.4295654296875,
      "loss": 0.6896,
      "rewards/accuracies": 0.3125,
      "rewards/chosen": 0.021530836820602417,
      "rewards/margins": 0.010109614580869675,
      "rewards/rejected": 0.011421223171055317,
      "step": 5
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 48.0567943972191,
      "learning_rate": 1e-06,
      "logits/chosen": -2.681126117706299,
      "logits/rejected": -2.6754660606384277,
      "logps/chosen": -280.4076843261719,
      "logps/rejected": -225.827392578125,
      "loss": 0.6307,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 0.6431528925895691,
      "rewards/margins": 0.25141650438308716,
      "rewards/rejected": 0.39173632860183716,
      "step": 10
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 39.99775130826886,
      "learning_rate": 9.922326639307916e-07,
      "logits/chosen": -2.5069479942321777,
      "logits/rejected": -2.494847059249878,
      "logps/chosen": -244.3173828125,
      "logps/rejected": -199.97879028320312,
      "loss": 0.5514,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.7437444925308228,
      "rewards/margins": 0.8651701211929321,
      "rewards/rejected": 0.8785742521286011,
      "step": 15
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 38.505483452366086,
      "learning_rate": 9.691719817616146e-07,
      "logits/chosen": -2.37453031539917,
      "logits/rejected": -2.383070468902588,
      "logps/chosen": -252.5743865966797,
      "logps/rejected": -214.252685546875,
      "loss": 0.5706,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 2.038425922393799,
      "rewards/margins": 1.3466479778289795,
      "rewards/rejected": 0.6917778253555298,
      "step": 20
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 41.910056196589444,
      "learning_rate": 9.315344337660421e-07,
      "logits/chosen": -2.3432867527008057,
      "logits/rejected": -2.324685573577881,
      "logps/chosen": -247.0692596435547,
      "logps/rejected": -210.7626495361328,
      "loss": 0.5489,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 1.9427118301391602,
      "rewards/margins": 1.2587368488311768,
      "rewards/rejected": 0.683975100517273,
      "step": 25
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 42.12807528715139,
      "learning_rate": 8.804893938804838e-07,
      "logits/chosen": -2.3173532485961914,
      "logits/rejected": -2.3291869163513184,
      "logps/chosen": -257.88848876953125,
      "logps/rejected": -196.9066925048828,
      "loss": 0.5516,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.6921398639678955,
      "rewards/margins": 1.2647020816802979,
      "rewards/rejected": 0.42743778228759766,
      "step": 30
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 24.467782530884058,
      "learning_rate": 8.176227980227692e-07,
      "logits/chosen": -2.3307278156280518,
      "logits/rejected": -2.318765640258789,
      "logps/chosen": -243.4705810546875,
      "logps/rejected": -221.982421875,
      "loss": 0.4937,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": 1.84653639793396,
      "rewards/margins": 2.2111032009124756,
      "rewards/rejected": -0.36456671357154846,
      "step": 35
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 23.95668807919176,
      "learning_rate": 7.448878701031142e-07,
      "logits/chosen": -2.2922379970550537,
      "logits/rejected": -2.2870445251464844,
      "logps/chosen": -248.8621063232422,
      "logps/rejected": -206.43832397460938,
      "loss": 0.2549,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 2.6322567462921143,
      "rewards/margins": 3.199286699295044,
      "rewards/rejected": -0.567030131816864,
      "step": 40
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 22.301472059197497,
      "learning_rate": 6.64544436638005e-07,
      "logits/chosen": -2.2371106147766113,
      "logits/rejected": -2.2345409393310547,
      "logps/chosen": -259.59674072265625,
      "logps/rejected": -229.2697296142578,
      "loss": 0.2511,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 2.4946646690368652,
      "rewards/margins": 3.3896145820617676,
      "rewards/rejected": -0.8949494361877441,
      "step": 45
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 20.520289228538196,
      "learning_rate": 5.790887154221519e-07,
      "logits/chosen": -2.237116575241089,
      "logits/rejected": -2.179481029510498,
      "logps/chosen": -261.8556213378906,
      "logps/rejected": -214.4944610595703,
      "loss": 0.2339,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 2.5970380306243896,
      "rewards/margins": 3.397597551345825,
      "rewards/rejected": -0.8005592226982117,
      "step": 50
    },
    {
      "epoch": 1.4925373134328357,
      "eval_logits/chosen": -2.2193446159362793,
      "eval_logits/rejected": -2.215559720993042,
      "eval_logps/chosen": -267.6477966308594,
      "eval_logps/rejected": -213.25338745117188,
      "eval_loss": 0.5196279883384705,
      "eval_rewards/accuracies": 0.8500000238418579,
      "eval_rewards/chosen": 2.350316047668457,
      "eval_rewards/margins": 2.606759548187256,
      "eval_rewards/rejected": -0.25644367933273315,
      "eval_runtime": 61.8636,
      "eval_samples_per_second": 15.356,
      "eval_steps_per_second": 0.242,
      "step": 50
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 25.665595876079696,
      "learning_rate": 4.911757596784357e-07,
      "logits/chosen": -2.261808395385742,
      "logits/rejected": -2.2048768997192383,
      "logps/chosen": -230.6652374267578,
      "logps/rejected": -215.3101348876953,
      "loss": 0.2951,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 2.6197896003723145,
      "rewards/margins": 3.964939832687378,
      "rewards/rejected": -1.3451504707336426,
      "step": 55
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 26.762136338138212,
      "learning_rate": 4.0353696729525153e-07,
      "logits/chosen": -2.3096041679382324,
      "logits/rejected": -2.287301540374756,
      "logps/chosen": -244.46975708007812,
      "logps/rejected": -219.6262969970703,
      "loss": 0.2662,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 3.1043174266815186,
      "rewards/margins": 4.127373695373535,
      "rewards/rejected": -1.0230560302734375,
      "step": 60
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 22.111621817096417,
      "learning_rate": 3.1889521808515883e-07,
      "logits/chosen": -2.333113431930542,
      "logits/rejected": -2.321504592895508,
      "logps/chosen": -250.5703887939453,
      "logps/rejected": -204.7659454345703,
      "loss": 0.2748,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 2.980164051055908,
      "rewards/margins": 3.9819297790527344,
      "rewards/rejected": -1.0017659664154053,
      "step": 65
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 16.467838593798508,
      "learning_rate": 2.398802756945589e-07,
      "logits/chosen": -2.3565258979797363,
      "logits/rejected": -2.355499744415283,
      "logps/chosen": -233.8147735595703,
      "logps/rejected": -230.2648162841797,
      "loss": 0.2095,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.9649741649627686,
      "rewards/margins": 3.8606581687927246,
      "rewards/rejected": -0.8956834077835083,
      "step": 70
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 13.307360146096983,
      "learning_rate": 1.689470825715998e-07,
      "logits/chosen": -2.3685317039489746,
      "logits/rejected": -2.368412494659424,
      "logps/chosen": -240.4198455810547,
      "logps/rejected": -217.22689819335938,
      "loss": 0.1621,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.9568119049072266,
      "rewards/margins": 3.929192304611206,
      "rewards/rejected": -0.972380518913269,
      "step": 75
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 21.216946699753258,
      "learning_rate": 1.0829948651407372e-07,
      "logits/chosen": -2.4029972553253174,
      "logits/rejected": -2.3749213218688965,
      "logps/chosen": -233.34228515625,
      "logps/rejected": -198.04251098632812,
      "loss": 0.1941,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.807535409927368,
      "rewards/margins": 3.6688785552978516,
      "rewards/rejected": -0.861343264579773,
      "step": 80
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 14.182439461289,
      "learning_rate": 5.982176856345444e-08,
      "logits/chosen": -2.3809475898742676,
      "logits/rejected": -2.366241693496704,
      "logps/chosen": -239.1095733642578,
      "logps/rejected": -207.1281280517578,
      "loss": 0.1338,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 2.8935706615448,
      "rewards/margins": 4.051820755004883,
      "rewards/rejected": -1.158250331878662,
      "step": 85
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 15.711855558491953,
      "learning_rate": 2.5020099628504598e-08,
      "logits/chosen": -2.3501780033111572,
      "logits/rejected": -2.3620352745056152,
      "logps/chosen": -242.01638793945312,
      "logps/rejected": -237.32321166992188,
      "loss": 0.1432,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.8932852745056152,
      "rewards/margins": 4.991795539855957,
      "rewards/rejected": -2.098510265350342,
      "step": 90
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 13.773804276890022,
      "learning_rate": 4.975744742772847e-09,
      "logits/chosen": -2.3822731971740723,
      "logits/rejected": -2.3518662452697754,
      "logps/chosen": -238.9407196044922,
      "logps/rejected": -213.6177215576172,
      "loss": 0.1579,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 2.955489158630371,
      "rewards/margins": 4.035004138946533,
      "rewards/rejected": -1.079514741897583,
      "step": 95
    },
    {
      "epoch": 2.955223880597015,
      "step": 99,
      "total_flos": 1167029855846400.0,
      "train_loss": 0.3387684075519292,
      "train_runtime": 3422.3866,
      "train_samples_per_second": 7.492,
      "train_steps_per_second": 0.029
    }
  ],
  "logging_steps": 5,
  "max_steps": 99,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1167029855846400.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}