{
  "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": 53.96980806965151,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7474961280822754,
      "logits/rejected": -2.740154504776001,
      "logps/chosen": -254.6457061767578,
      "logps/rejected": -205.4122314453125,
      "loss": 0.6904,
      "rewards/accuracies": 0.3187499940395355,
      "rewards/chosen": 0.015023870393633842,
      "rewards/margins": 0.000346918503055349,
      "rewards/rejected": 0.014676950871944427,
      "step": 5
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 51.127419187517475,
      "learning_rate": 1e-06,
      "logits/chosen": -2.70166277885437,
      "logits/rejected": -2.6838972568511963,
      "logps/chosen": -239.15487670898438,
      "logps/rejected": -208.0798797607422,
      "loss": 0.6437,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.468679815530777,
      "rewards/margins": 0.14893516898155212,
      "rewards/rejected": 0.31974464654922485,
      "step": 10
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 48.915622995434845,
      "learning_rate": 9.983100718730718e-07,
      "logits/chosen": -2.455461263656616,
      "logits/rejected": -2.436664342880249,
      "logps/chosen": -238.5790557861328,
      "logps/rejected": -190.64874267578125,
      "loss": 0.6643,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 1.17107093334198,
      "rewards/margins": 0.4588069021701813,
      "rewards/rejected": 0.7122639417648315,
      "step": 15
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 42.1310748636241,
      "learning_rate": 9.932517109205849e-07,
      "logits/chosen": -2.3171567916870117,
      "logits/rejected": -2.2968077659606934,
      "logps/chosen": -234.33566284179688,
      "logps/rejected": -199.1301727294922,
      "loss": 0.6155,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 1.4784119129180908,
      "rewards/margins": 0.7344081401824951,
      "rewards/rejected": 0.7440037131309509,
      "step": 20
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 37.58643004926761,
      "learning_rate": 9.848591102083375e-07,
      "logits/chosen": -2.239626884460449,
      "logits/rejected": -2.2184576988220215,
      "logps/chosen": -270.282958984375,
      "logps/rejected": -196.48818969726562,
      "loss": 0.5953,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 1.7982975244522095,
      "rewards/margins": 1.2364143133163452,
      "rewards/rejected": 0.5618831515312195,
      "step": 25
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 40.38923718028625,
      "learning_rate": 9.731890013043367e-07,
      "logits/chosen": -2.2369236946105957,
      "logits/rejected": -2.2441658973693848,
      "logps/chosen": -231.41775512695312,
      "logps/rejected": -221.48721313476562,
      "loss": 0.5606,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 1.3224443197250366,
      "rewards/margins": 1.2497867345809937,
      "rewards/rejected": 0.07265764474868774,
      "step": 30
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 47.56097946401291,
      "learning_rate": 9.583202707897073e-07,
      "logits/chosen": -2.3453598022460938,
      "logits/rejected": -2.339764356613159,
      "logps/chosen": -241.860107421875,
      "logps/rejected": -214.53897094726562,
      "loss": 0.65,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 0.8914967775344849,
      "rewards/margins": 0.727935791015625,
      "rewards/rejected": 0.16356101632118225,
      "step": 35
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 42.0267103133091,
      "learning_rate": 9.403534270080829e-07,
      "logits/chosen": -2.3654093742370605,
      "logits/rejected": -2.353811264038086,
      "logps/chosen": -238.9351043701172,
      "logps/rejected": -209.5872802734375,
      "loss": 0.6079,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 1.131967544555664,
      "rewards/margins": 1.2806609869003296,
      "rewards/rejected": -0.1486934870481491,
      "step": 40
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 39.949353314773276,
      "learning_rate": 9.19409920658098e-07,
      "logits/chosen": -2.212584972381592,
      "logits/rejected": -2.2196860313415527,
      "logps/chosen": -262.6395263671875,
      "logps/rejected": -202.49072265625,
      "loss": 0.6097,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": 1.2208731174468994,
      "rewards/margins": 1.1154453754425049,
      "rewards/rejected": 0.10542778670787811,
      "step": 45
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 43.82292011113308,
      "learning_rate": 8.956313238215823e-07,
      "logits/chosen": -2.107001543045044,
      "logits/rejected": -2.0714831352233887,
      "logps/chosen": -242.4890899658203,
      "logps/rejected": -204.85910034179688,
      "loss": 0.5987,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 1.1507846117019653,
      "rewards/margins": 1.0607472658157349,
      "rewards/rejected": 0.09003732353448868,
      "step": 50
    },
    {
      "epoch": 0.746268656716418,
      "eval_logits/chosen": -2.1097404956817627,
      "eval_logits/rejected": -2.074279546737671,
      "eval_logps/chosen": -264.0938720703125,
      "eval_logps/rejected": -229.32229614257812,
      "eval_loss": 0.6138201951980591,
      "eval_rewards/accuracies": 0.7583333253860474,
      "eval_rewards/chosen": 1.2434982061386108,
      "eval_rewards/margins": 1.4021955728530884,
      "eval_rewards/rejected": -0.15869754552841187,
      "eval_runtime": 129.58,
      "eval_samples_per_second": 14.663,
      "eval_steps_per_second": 0.232,
      "step": 50
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 40.5253665377123,
      "learning_rate": 8.691783729769873e-07,
      "logits/chosen": -2.115365505218506,
      "logits/rejected": -2.0831191539764404,
      "logps/chosen": -245.2086944580078,
      "logps/rejected": -208.19717407226562,
      "loss": 0.5896,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": 0.9533665776252747,
      "rewards/margins": 1.4577382802963257,
      "rewards/rejected": -0.5043715834617615,
      "step": 55
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 34.60112523621595,
      "learning_rate": 8.402298824670029e-07,
      "logits/chosen": -2.101773262023926,
      "logits/rejected": -2.0949249267578125,
      "logps/chosen": -249.70480346679688,
      "logps/rejected": -219.0111083984375,
      "loss": 0.5525,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.742239773273468,
      "rewards/margins": 1.2933545112609863,
      "rewards/rejected": -0.5511147379875183,
      "step": 60
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 33.604589067191306,
      "learning_rate": 8.089815357650089e-07,
      "logits/chosen": -2.090615749359131,
      "logits/rejected": -2.0645270347595215,
      "logps/chosen": -254.70431518554688,
      "logps/rejected": -216.178955078125,
      "loss": 0.5693,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.8447690010070801,
      "rewards/margins": 1.3857734203338623,
      "rewards/rejected": -0.5410043001174927,
      "step": 65
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 21.06317859535433,
      "learning_rate": 7.756445627110522e-07,
      "logits/chosen": -2.06718373298645,
      "logits/rejected": -2.051907539367676,
      "logps/chosen": -242.66635131835938,
      "logps/rejected": -222.59707641601562,
      "loss": 0.4108,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": 1.4726606607437134,
      "rewards/margins": 2.182919979095459,
      "rewards/rejected": -0.710259199142456,
      "step": 70
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 22.259738178667774,
      "learning_rate": 7.404443116588547e-07,
      "logits/chosen": -2.0406711101531982,
      "logits/rejected": -2.018770217895508,
      "logps/chosen": -241.16098022460938,
      "logps/rejected": -227.0569610595703,
      "loss": 0.2542,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": 1.9895944595336914,
      "rewards/margins": 2.734738349914551,
      "rewards/rejected": -0.7451435327529907,
      "step": 75
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 16.79071490181188,
      "learning_rate": 7.036187261857288e-07,
      "logits/chosen": -2.0667386054992676,
      "logits/rejected": -2.020956516265869,
      "logps/chosen": -238.8300018310547,
      "logps/rejected": -230.9183349609375,
      "loss": 0.2169,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.387232542037964,
      "rewards/margins": 3.326624631881714,
      "rewards/rejected": -0.9393925666809082,
      "step": 80
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 26.173042165666224,
      "learning_rate": 6.654167366624008e-07,
      "logits/chosen": -2.1021151542663574,
      "logits/rejected": -2.0837912559509277,
      "logps/chosen": -225.0214080810547,
      "logps/rejected": -230.99365234375,
      "loss": 0.2518,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.4148898124694824,
      "rewards/margins": 3.303870439529419,
      "rewards/rejected": -0.8889807462692261,
      "step": 85
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 22.231137216437176,
      "learning_rate": 6.260965775552713e-07,
      "logits/chosen": -2.164511203765869,
      "logits/rejected": -2.1222901344299316,
      "logps/chosen": -232.6430206298828,
      "logps/rejected": -203.33914184570312,
      "loss": 0.2452,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 2.4132513999938965,
      "rewards/margins": 3.234947919845581,
      "rewards/rejected": -0.8216965794563293,
      "step": 90
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 25.44451135006177,
      "learning_rate": 5.859240418356614e-07,
      "logits/chosen": -2.181170701980591,
      "logits/rejected": -2.1437950134277344,
      "logps/chosen": -227.32174682617188,
      "logps/rejected": -220.1398162841797,
      "loss": 0.3253,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.5023834705352783,
      "rewards/margins": 3.4593710899353027,
      "rewards/rejected": -0.9569879770278931,
      "step": 95
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 21.85215171836367,
      "learning_rate": 5.451706842957421e-07,
      "logits/chosen": -2.190605640411377,
      "logits/rejected": -2.1743760108947754,
      "logps/chosen": -222.10366821289062,
      "logps/rejected": -216.30953979492188,
      "loss": 0.2776,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 2.2073616981506348,
      "rewards/margins": 3.289036989212036,
      "rewards/rejected": -1.08167564868927,
      "step": 100
    },
    {
      "epoch": 1.4925373134328357,
      "eval_logits/chosen": -2.2346103191375732,
      "eval_logits/rejected": -2.197340488433838,
      "eval_logps/chosen": -259.05035400390625,
      "eval_logps/rejected": -231.1164093017578,
      "eval_loss": 0.5806639194488525,
      "eval_rewards/accuracies": 0.75,
      "eval_rewards/chosen": 1.7478455305099487,
      "eval_rewards/margins": 2.0859532356262207,
      "eval_rewards/rejected": -0.33810776472091675,
      "eval_runtime": 127.669,
      "eval_samples_per_second": 14.882,
      "eval_steps_per_second": 0.235,
      "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
}