{
  "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": 65.94647505015266,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7657113075256348,
      "logits/rejected": -2.7525341510772705,
      "logps/chosen": -269.53558349609375,
      "logps/rejected": -194.53842163085938,
      "loss": 0.6891,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 0.020892271772027016,
      "rewards/margins": 0.010709484107792377,
      "rewards/rejected": 0.010182789526879787,
      "step": 5
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 63.96896693841325,
      "learning_rate": 1e-06,
      "logits/chosen": -2.705475330352783,
      "logits/rejected": -2.6866021156311035,
      "logps/chosen": -248.81494140625,
      "logps/rejected": -210.038330078125,
      "loss": 0.6357,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.5479151606559753,
      "rewards/margins": 0.18559947609901428,
      "rewards/rejected": 0.3623156249523163,
      "step": 10
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 40.712613492844916,
      "learning_rate": 9.922326639307916e-07,
      "logits/chosen": -2.508521556854248,
      "logits/rejected": -2.493314743041992,
      "logps/chosen": -236.1773681640625,
      "logps/rejected": -197.7436981201172,
      "loss": 0.6236,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 1.5045735836029053,
      "rewards/margins": 0.5912668108940125,
      "rewards/rejected": 0.9133070111274719,
      "step": 15
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 44.98752011627579,
      "learning_rate": 9.691719817616146e-07,
      "logits/chosen": -2.339185953140259,
      "logits/rejected": -2.3386504650115967,
      "logps/chosen": -265.876220703125,
      "logps/rejected": -234.7918243408203,
      "loss": 0.6062,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 1.8170077800750732,
      "rewards/margins": 1.5930566787719727,
      "rewards/rejected": 0.22395117580890656,
      "step": 20
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 59.48293144738292,
      "learning_rate": 9.315344337660421e-07,
      "logits/chosen": -2.2757716178894043,
      "logits/rejected": -2.2592074871063232,
      "logps/chosen": -252.29898071289062,
      "logps/rejected": -229.03823852539062,
      "loss": 0.5742,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 1.4913432598114014,
      "rewards/margins": 1.7369304895401,
      "rewards/rejected": -0.2455870658159256,
      "step": 25
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 36.615662017204315,
      "learning_rate": 8.804893938804838e-07,
      "logits/chosen": -2.317497730255127,
      "logits/rejected": -2.3040828704833984,
      "logps/chosen": -261.4601135253906,
      "logps/rejected": -244.25497436523438,
      "loss": 0.6156,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 1.3194864988327026,
      "rewards/margins": 1.6472011804580688,
      "rewards/rejected": -0.327714741230011,
      "step": 30
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 21.739970102254446,
      "learning_rate": 8.176227980227692e-07,
      "logits/chosen": -2.390035390853882,
      "logits/rejected": -2.3696439266204834,
      "logps/chosen": -249.12857055664062,
      "logps/rejected": -256.47308349609375,
      "loss": 0.472,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 1.1576669216156006,
      "rewards/margins": 1.7294819355010986,
      "rewards/rejected": -0.5718148350715637,
      "step": 35
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 20.82917677304684,
      "learning_rate": 7.448878701031142e-07,
      "logits/chosen": -2.4545044898986816,
      "logits/rejected": -2.4309377670288086,
      "logps/chosen": -240.4742889404297,
      "logps/rejected": -219.581298828125,
      "loss": 0.2582,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.6987485885620117,
      "rewards/margins": 2.946779251098633,
      "rewards/rejected": -1.2480309009552002,
      "step": 40
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 21.427007652112742,
      "learning_rate": 6.64544436638005e-07,
      "logits/chosen": -2.5179731845855713,
      "logits/rejected": -2.5242679119110107,
      "logps/chosen": -243.6165008544922,
      "logps/rejected": -223.01467895507812,
      "loss": 0.2809,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 2.208566904067993,
      "rewards/margins": 3.001779079437256,
      "rewards/rejected": -0.7932125926017761,
      "step": 45
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 24.490168137290794,
      "learning_rate": 5.790887154221519e-07,
      "logits/chosen": -2.534179449081421,
      "logits/rejected": -2.5223186016082764,
      "logps/chosen": -245.3209228515625,
      "logps/rejected": -230.1135711669922,
      "loss": 0.2704,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.485767364501953,
      "rewards/margins": 3.721315860748291,
      "rewards/rejected": -1.2355483770370483,
      "step": 50
    },
    {
      "epoch": 1.4925373134328357,
      "eval_logits/chosen": -2.4811432361602783,
      "eval_logits/rejected": -2.4862773418426514,
      "eval_logps/chosen": -233.6154022216797,
      "eval_logps/rejected": -202.26788330078125,
      "eval_loss": 0.6698261499404907,
      "eval_rewards/accuracies": 0.7583333253860474,
      "eval_rewards/chosen": 1.3441599607467651,
      "eval_rewards/margins": 1.3986517190933228,
      "eval_rewards/rejected": -0.054491832852363586,
      "eval_runtime": 65.2748,
      "eval_samples_per_second": 14.554,
      "eval_steps_per_second": 0.23,
      "step": 50
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 37.8980453420408,
      "learning_rate": 4.911757596784357e-07,
      "logits/chosen": -2.4741039276123047,
      "logits/rejected": -2.484839916229248,
      "logps/chosen": -238.60690307617188,
      "logps/rejected": -259.42974853515625,
      "loss": 0.3114,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 2.5658390522003174,
      "rewards/margins": 3.876957654953003,
      "rewards/rejected": -1.311118483543396,
      "step": 55
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 25.33622153036637,
      "learning_rate": 4.0353696729525153e-07,
      "logits/chosen": -2.447230815887451,
      "logits/rejected": -2.454420804977417,
      "logps/chosen": -241.8343963623047,
      "logps/rejected": -248.6758575439453,
      "loss": 0.2955,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.8466014862060547,
      "rewards/margins": 3.5575203895568848,
      "rewards/rejected": -0.7109190225601196,
      "step": 60
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 25.54659742532799,
      "learning_rate": 3.1889521808515883e-07,
      "logits/chosen": -2.425412654876709,
      "logits/rejected": -2.4246826171875,
      "logps/chosen": -242.4326934814453,
      "logps/rejected": -209.26889038085938,
      "loss": 0.3105,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": 2.472623348236084,
      "rewards/margins": 2.825917959213257,
      "rewards/rejected": -0.3532942235469818,
      "step": 65
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 16.864437001307557,
      "learning_rate": 2.398802756945589e-07,
      "logits/chosen": -2.3947904109954834,
      "logits/rejected": -2.3913135528564453,
      "logps/chosen": -239.5849151611328,
      "logps/rejected": -242.7691192626953,
      "loss": 0.2356,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.736234188079834,
      "rewards/margins": 3.8313822746276855,
      "rewards/rejected": -1.0951482057571411,
      "step": 70
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 14.451394682026951,
      "learning_rate": 1.689470825715998e-07,
      "logits/chosen": -2.368191957473755,
      "logits/rejected": -2.3499183654785156,
      "logps/chosen": -217.12374877929688,
      "logps/rejected": -230.18588256835938,
      "loss": 0.1513,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 2.5037055015563965,
      "rewards/margins": 3.936789035797119,
      "rewards/rejected": -1.4330837726593018,
      "step": 75
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 25.39911714592628,
      "learning_rate": 1.0829948651407372e-07,
      "logits/chosen": -2.3876171112060547,
      "logits/rejected": -2.3770811557769775,
      "logps/chosen": -242.2554473876953,
      "logps/rejected": -248.1456756591797,
      "loss": 0.1699,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.8140194416046143,
      "rewards/margins": 3.681828022003174,
      "rewards/rejected": -0.86780846118927,
      "step": 80
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 14.791785799598257,
      "learning_rate": 5.982176856345444e-08,
      "logits/chosen": -2.3988237380981445,
      "logits/rejected": -2.385554313659668,
      "logps/chosen": -260.5726623535156,
      "logps/rejected": -248.03329467773438,
      "loss": 0.154,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 3.059401750564575,
      "rewards/margins": 3.9952499866485596,
      "rewards/rejected": -0.9358483552932739,
      "step": 85
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 16.84144820116068,
      "learning_rate": 2.5020099628504598e-08,
      "logits/chosen": -2.3938612937927246,
      "logits/rejected": -2.398956060409546,
      "logps/chosen": -239.680419921875,
      "logps/rejected": -238.61038208007812,
      "loss": 0.1517,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.9095988273620605,
      "rewards/margins": 4.1475067138671875,
      "rewards/rejected": -1.2379077672958374,
      "step": 90
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 15.141407540579351,
      "learning_rate": 4.975744742772847e-09,
      "logits/chosen": -2.40722393989563,
      "logits/rejected": -2.397171974182129,
      "logps/chosen": -232.5000762939453,
      "logps/rejected": -241.65133666992188,
      "loss": 0.1794,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 2.393092632293701,
      "rewards/margins": 3.879788875579834,
      "rewards/rejected": -1.4866969585418701,
      "step": 95
    }
  ],
  "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
}