{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0695187165775402,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.053475935828877004,
      "grad_norm": 65.08164204257967,
      "learning_rate": 5e-07,
      "logits/chosen": -2.733124017715454,
      "logits/rejected": -2.7089576721191406,
      "logps/chosen": -271.7913818359375,
      "logps/rejected": -220.1679229736328,
      "loss": 0.6907,
      "rewards/accuracies": 0.3499999940395355,
      "rewards/chosen": 0.01848032884299755,
      "rewards/margins": 0.006348843220621347,
      "rewards/rejected": 0.01213148795068264,
      "step": 5
    },
    {
      "epoch": 0.10695187165775401,
      "grad_norm": 50.59833986698045,
      "learning_rate": 1e-06,
      "logits/chosen": -2.66473388671875,
      "logits/rejected": -2.6718990802764893,
      "logps/chosen": -258.00506591796875,
      "logps/rejected": -206.7321319580078,
      "loss": 0.6381,
      "rewards/accuracies": 0.606249988079071,
      "rewards/chosen": 0.5401543378829956,
      "rewards/margins": 0.19901855289936066,
      "rewards/rejected": 0.3411358594894409,
      "step": 10
    },
    {
      "epoch": 0.16042780748663102,
      "grad_norm": 50.14290028396245,
      "learning_rate": 9.991477798614637e-07,
      "logits/chosen": -2.508185863494873,
      "logits/rejected": -2.524341106414795,
      "logps/chosen": -271.5559997558594,
      "logps/rejected": -203.901123046875,
      "loss": 0.6476,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 1.8893938064575195,
      "rewards/margins": 0.9772478342056274,
      "rewards/rejected": 0.9121459126472473,
      "step": 15
    },
    {
      "epoch": 0.21390374331550802,
      "grad_norm": 42.09940353495551,
      "learning_rate": 9.965940245625131e-07,
      "logits/chosen": -2.3450875282287598,
      "logits/rejected": -2.3332550525665283,
      "logps/chosen": -231.1672821044922,
      "logps/rejected": -189.68919372558594,
      "loss": 0.6411,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 1.9618078470230103,
      "rewards/margins": 1.288054347038269,
      "rewards/rejected": 0.6737534403800964,
      "step": 20
    },
    {
      "epoch": 0.26737967914438504,
      "grad_norm": 50.008448633201716,
      "learning_rate": 9.923474395499264e-07,
      "logits/chosen": -2.2368509769439697,
      "logits/rejected": -2.2110252380371094,
      "logps/chosen": -238.6009063720703,
      "logps/rejected": -201.56163024902344,
      "loss": 0.6356,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 1.4835152626037598,
      "rewards/margins": 0.7462359666824341,
      "rewards/rejected": 0.7372792959213257,
      "step": 25
    },
    {
      "epoch": 0.32085561497326204,
      "grad_norm": 37.46999153089867,
      "learning_rate": 9.86422500924775e-07,
      "logits/chosen": -2.279388904571533,
      "logits/rejected": -2.2622551918029785,
      "logps/chosen": -210.3643798828125,
      "logps/rejected": -200.84750366210938,
      "loss": 0.5918,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.0033924579620361,
      "rewards/margins": 0.8053790330886841,
      "rewards/rejected": 0.1980135291814804,
      "step": 30
    },
    {
      "epoch": 0.37433155080213903,
      "grad_norm": 40.36951895662085,
      "learning_rate": 9.788394060951227e-07,
      "logits/chosen": -2.380889654159546,
      "logits/rejected": -2.3877792358398438,
      "logps/chosen": -242.85470581054688,
      "logps/rejected": -207.0241241455078,
      "loss": 0.6045,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.8806816935539246,
      "rewards/margins": 0.9049984812736511,
      "rewards/rejected": -0.024316836148500443,
      "step": 35
    },
    {
      "epoch": 0.42780748663101603,
      "grad_norm": 49.54969179550948,
      "learning_rate": 9.696240049254742e-07,
      "logits/chosen": -2.361086368560791,
      "logits/rejected": -2.358868360519409,
      "logps/chosen": -250.7805633544922,
      "logps/rejected": -223.0339813232422,
      "loss": 0.6251,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.8420997858047485,
      "rewards/margins": 1.0070801973342896,
      "rewards/rejected": -0.16498039662837982,
      "step": 40
    },
    {
      "epoch": 0.48128342245989303,
      "grad_norm": 38.326754460806335,
      "learning_rate": 9.588077116176756e-07,
      "logits/chosen": -2.3445451259613037,
      "logits/rejected": -2.329608917236328,
      "logps/chosen": -244.2638702392578,
      "logps/rejected": -213.9591522216797,
      "loss": 0.6052,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.9911876916885376,
      "rewards/margins": 0.9155967831611633,
      "rewards/rejected": 0.07559084892272949,
      "step": 45
    },
    {
      "epoch": 0.5347593582887701,
      "grad_norm": 41.70114459904792,
      "learning_rate": 9.464273976236516e-07,
      "logits/chosen": -2.3615097999572754,
      "logits/rejected": -2.375725269317627,
      "logps/chosen": -235.63998413085938,
      "logps/rejected": -205.462158203125,
      "loss": 0.5778,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 1.0482008457183838,
      "rewards/margins": 1.3194622993469238,
      "rewards/rejected": -0.27126139402389526,
      "step": 50
    },
    {
      "epoch": 0.5347593582887701,
      "eval_logits/chosen": -2.382444381713867,
      "eval_logits/rejected": -2.353994369506836,
      "eval_logps/chosen": -246.02862548828125,
      "eval_logps/rejected": -197.20895385742188,
      "eval_loss": 0.5711008310317993,
      "eval_rewards/accuracies": 0.7559523582458496,
      "eval_rewards/chosen": 1.21534264087677,
      "eval_rewards/margins": 1.2842590808868408,
      "eval_rewards/rejected": -0.0689164251089096,
      "eval_runtime": 178.2179,
      "eval_samples_per_second": 14.926,
      "eval_steps_per_second": 0.236,
      "step": 50
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 37.61770640976083,
      "learning_rate": 9.325252659550308e-07,
      "logits/chosen": -2.4103198051452637,
      "logits/rejected": -2.375884532928467,
      "logps/chosen": -232.7388458251953,
      "logps/rejected": -197.4444122314453,
      "loss": 0.5385,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.110247015953064,
      "rewards/margins": 1.345210075378418,
      "rewards/rejected": -0.23496298491954803,
      "step": 55
    },
    {
      "epoch": 0.6417112299465241,
      "grad_norm": 36.284977037429975,
      "learning_rate": 9.171487073181197e-07,
      "logits/chosen": -2.45546817779541,
      "logits/rejected": -2.4268698692321777,
      "logps/chosen": -264.7644958496094,
      "logps/rejected": -237.2240753173828,
      "loss": 0.5661,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 1.084460973739624,
      "rewards/margins": 1.5218535661697388,
      "rewards/rejected": -0.4373924136161804,
      "step": 60
    },
    {
      "epoch": 0.6951871657754011,
      "grad_norm": 37.36342990536676,
      "learning_rate": 9.003501385646448e-07,
      "logits/chosen": -2.4519729614257812,
      "logits/rejected": -2.4639739990234375,
      "logps/chosen": -251.09231567382812,
      "logps/rejected": -201.4525909423828,
      "loss": 0.5596,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": 1.1350815296173096,
      "rewards/margins": 1.5527232885360718,
      "rewards/rejected": -0.4176417887210846,
      "step": 65
    },
    {
      "epoch": 0.7486631016042781,
      "grad_norm": 33.50003576842058,
      "learning_rate": 8.821868240089676e-07,
      "logits/chosen": -2.400097608566284,
      "logits/rejected": -2.377209186553955,
      "logps/chosen": -265.53155517578125,
      "logps/rejected": -226.5992431640625,
      "loss": 0.5263,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.7660782337188721,
      "rewards/margins": 1.5463287830352783,
      "rewards/rejected": -0.7802505493164062,
      "step": 70
    },
    {
      "epoch": 0.8021390374331551,
      "grad_norm": 35.415664255337724,
      "learning_rate": 8.62720680220876e-07,
      "logits/chosen": -2.389862060546875,
      "logits/rejected": -2.379544258117676,
      "logps/chosen": -250.21115112304688,
      "logps/rejected": -234.2509765625,
      "loss": 0.5073,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.7700754404067993,
      "rewards/margins": 1.3676471710205078,
      "rewards/rejected": -0.597571849822998,
      "step": 75
    },
    {
      "epoch": 0.8556149732620321,
      "grad_norm": 38.29403188930404,
      "learning_rate": 8.420180649593929e-07,
      "logits/chosen": -2.42349910736084,
      "logits/rejected": -2.4062836170196533,
      "logps/chosen": -261.1897277832031,
      "logps/rejected": -225.38619995117188,
      "loss": 0.5523,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.5706126689910889,
      "rewards/margins": 1.5985792875289917,
      "rewards/rejected": -1.0279667377471924,
      "step": 80
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 34.7463437300499,
      "learning_rate": 8.201495509671036e-07,
      "logits/chosen": -2.4246888160705566,
      "logits/rejected": -2.438143253326416,
      "logps/chosen": -259.6698303222656,
      "logps/rejected": -213.34518432617188,
      "loss": 0.5362,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 1.5000348091125488,
      "rewards/margins": 1.8813749551773071,
      "rewards/rejected": -0.3813403248786926,
      "step": 85
    },
    {
      "epoch": 0.9625668449197861,
      "grad_norm": 34.01709977955445,
      "learning_rate": 7.971896853961042e-07,
      "logits/chosen": -2.4786272048950195,
      "logits/rejected": -2.4621052742004395,
      "logps/chosen": -252.6436004638672,
      "logps/rejected": -250.2777557373047,
      "loss": 0.5064,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 1.4901584386825562,
      "rewards/margins": 1.7402337789535522,
      "rewards/rejected": -0.25007539987564087,
      "step": 90
    },
    {
      "epoch": 1.0160427807486632,
      "grad_norm": 24.058087864441212,
      "learning_rate": 7.732167356856654e-07,
      "logits/chosen": -2.5284178256988525,
      "logits/rejected": -2.500856876373291,
      "logps/chosen": -245.199462890625,
      "logps/rejected": -247.8142852783203,
      "loss": 0.4555,
      "rewards/accuracies": 0.824999988079071,
      "rewards/chosen": 1.14225172996521,
      "rewards/margins": 2.6005594730377197,
      "rewards/rejected": -1.4583076238632202,
      "step": 95
    },
    {
      "epoch": 1.0695187165775402,
      "grad_norm": 34.758664424975294,
      "learning_rate": 7.48312422757881e-07,
      "logits/chosen": -2.516972064971924,
      "logits/rejected": -2.514902114868164,
      "logps/chosen": -236.08767700195312,
      "logps/rejected": -226.1017608642578,
      "loss": 0.232,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.0098340511322021,
      "rewards/margins": 3.1043362617492676,
      "rewards/rejected": -2.0945024490356445,
      "step": 100
    },
    {
      "epoch": 1.0695187165775402,
      "eval_logits/chosen": -2.5262153148651123,
      "eval_logits/rejected": -2.4977195262908936,
      "eval_logps/chosen": -248.79345703125,
      "eval_logps/rejected": -207.30593872070312,
      "eval_loss": 0.49347466230392456,
      "eval_rewards/accuracies": 0.8035714030265808,
      "eval_rewards/chosen": 0.938858687877655,
      "eval_rewards/margins": 2.0174732208251953,
      "eval_rewards/rejected": -1.078614354133606,
      "eval_runtime": 177.1838,
      "eval_samples_per_second": 15.013,
      "eval_steps_per_second": 0.237,
      "step": 100
    }
  ],
  "logging_steps": 5,
  "max_steps": 279,
  "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
}