{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.963855421686747,
  "eval_steps": 500,
  "global_step": 123,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.12048192771084337,
      "grad_norm": 26.10561591835797,
      "learning_rate": 9.998257709344246e-06,
      "logits/chosen": 0.25857219099998474,
      "logits/rejected": 0.2376616895198822,
      "logps/chosen": -50.41423034667969,
      "logps/rejected": -59.181236267089844,
      "loss": 0.9468,
      "rewards/accuracies": 0.29374998807907104,
      "rewards/chosen": -0.02607300505042076,
      "rewards/margins": 0.003030994441360235,
      "rewards/rejected": -0.029103999957442284,
      "step": 5
    },
    {
      "epoch": 0.24096385542168675,
      "grad_norm": 41.86378845239721,
      "learning_rate": 9.937404935870938e-06,
      "logits/chosen": 0.28393980860710144,
      "logits/rejected": 0.27348166704177856,
      "logps/chosen": -85.42002868652344,
      "logps/rejected": -101.35990905761719,
      "loss": 1.3023,
      "rewards/accuracies": 0.59375,
      "rewards/chosen": -3.2973403930664062,
      "rewards/margins": 0.6951224207878113,
      "rewards/rejected": -3.9924628734588623,
      "step": 10
    },
    {
      "epoch": 0.3614457831325301,
      "grad_norm": 23.22750429299514,
      "learning_rate": 9.790648062053341e-06,
      "logits/chosen": 0.4092663824558258,
      "logits/rejected": 0.4109087586402893,
      "logps/chosen": -111.24369812011719,
      "logps/rejected": -125.99955749511719,
      "loss": 1.5654,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -6.065593719482422,
      "rewards/margins": 0.8512781858444214,
      "rewards/rejected": -6.916871547698975,
      "step": 15
    },
    {
      "epoch": 0.4819277108433735,
      "grad_norm": 28.558160672825796,
      "learning_rate": 9.56054045699494e-06,
      "logits/chosen": 0.4347962737083435,
      "logits/rejected": 0.4268385171890259,
      "logps/chosen": -121.10899353027344,
      "logps/rejected": -143.13150024414062,
      "loss": 1.4494,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -6.8496599197387695,
      "rewards/margins": 1.7266939878463745,
      "rewards/rejected": -8.576354026794434,
      "step": 20
    },
    {
      "epoch": 0.6024096385542169,
      "grad_norm": 28.54980921467646,
      "learning_rate": 9.251085678648072e-06,
      "logits/chosen": 0.3079491853713989,
      "logits/rejected": 0.3097611963748932,
      "logps/chosen": -137.2875213623047,
      "logps/rejected": -160.32693481445312,
      "loss": 1.8024,
      "rewards/accuracies": 0.6687500476837158,
      "rewards/chosen": -8.795693397521973,
      "rewards/margins": 1.6571794748306274,
      "rewards/rejected": -10.452873229980469,
      "step": 25
    },
    {
      "epoch": 0.7228915662650602,
      "grad_norm": 34.79226520265609,
      "learning_rate": 8.867667817375266e-06,
      "logits/chosen": 0.4392024874687195,
      "logits/rejected": 0.44088032841682434,
      "logps/chosen": -133.45188903808594,
      "logps/rejected": -157.28854370117188,
      "loss": 1.6427,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": -8.06303596496582,
      "rewards/margins": 1.6960468292236328,
      "rewards/rejected": -9.759082794189453,
      "step": 30
    },
    {
      "epoch": 0.8433734939759037,
      "grad_norm": 14.892114319678067,
      "learning_rate": 8.416957820132743e-06,
      "logits/chosen": 0.615296483039856,
      "logits/rejected": 0.6112682819366455,
      "logps/chosen": -122.12522888183594,
      "logps/rejected": -144.03701782226562,
      "loss": 1.5509,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": -7.20847225189209,
      "rewards/margins": 1.6640231609344482,
      "rewards/rejected": -8.872495651245117,
      "step": 35
    },
    {
      "epoch": 0.963855421686747,
      "grad_norm": 18.023604436984943,
      "learning_rate": 7.9067974251073e-06,
      "logits/chosen": 0.6920608282089233,
      "logits/rejected": 0.694674551486969,
      "logps/chosen": -123.02449035644531,
      "logps/rejected": -142.08978271484375,
      "loss": 1.547,
      "rewards/accuracies": 0.6749999523162842,
      "rewards/chosen": -7.2558465003967285,
      "rewards/margins": 1.388303279876709,
      "rewards/rejected": -8.644149780273438,
      "step": 40
    },
    {
      "epoch": 1.0843373493975903,
      "grad_norm": 8.538276533744181,
      "learning_rate": 7.346062726185332e-06,
      "logits/chosen": 0.7355310916900635,
      "logits/rejected": 0.732780396938324,
      "logps/chosen": -117.85956573486328,
      "logps/rejected": -164.07598876953125,
      "loss": 0.9705,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": -6.436078071594238,
      "rewards/margins": 4.134807586669922,
      "rewards/rejected": -10.57088565826416,
      "step": 45
    },
    {
      "epoch": 1.2048192771084336,
      "grad_norm": 16.51016343701174,
      "learning_rate": 6.744509741045835e-06,
      "logits/chosen": 0.7370636463165283,
      "logits/rejected": 0.7329871654510498,
      "logps/chosen": -99.00725555419922,
      "logps/rejected": -157.13314819335938,
      "loss": 0.71,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": -4.765503883361816,
      "rewards/margins": 5.304986953735352,
      "rewards/rejected": -10.070491790771484,
      "step": 50
    },
    {
      "epoch": 1.3253012048192772,
      "grad_norm": 7.641392545091902,
      "learning_rate": 6.112604669781572e-06,
      "logits/chosen": 0.7394146919250488,
      "logits/rejected": 0.7397529482841492,
      "logps/chosen": -89.04443359375,
      "logps/rejected": -154.7861328125,
      "loss": 0.5588,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -3.863969326019287,
      "rewards/margins": 5.744624614715576,
      "rewards/rejected": -9.60859489440918,
      "step": 55
    },
    {
      "epoch": 1.4457831325301205,
      "grad_norm": 14.23736999395365,
      "learning_rate": 5.46134179731651e-06,
      "logits/chosen": 0.7342737913131714,
      "logits/rejected": 0.7296489477157593,
      "logps/chosen": -92.59645080566406,
      "logps/rejected": -148.73355102539062,
      "loss": 0.6337,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": -3.977640151977539,
      "rewards/margins": 5.177193641662598,
      "rewards/rejected": -9.154833793640137,
      "step": 60
    },
    {
      "epoch": 1.5662650602409638,
      "grad_norm": 3.403281181921107,
      "learning_rate": 4.802052207868654e-06,
      "logits/chosen": 0.7081732749938965,
      "logits/rejected": 0.7035213708877563,
      "logps/chosen": -87.75848388671875,
      "logps/rejected": -155.9546356201172,
      "loss": 0.4989,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": -3.674849271774292,
      "rewards/margins": 6.139934062957764,
      "rewards/rejected": -9.814783096313477,
      "step": 65
    },
    {
      "epoch": 1.6867469879518073,
      "grad_norm": 7.40455077675666,
      "learning_rate": 4.146206639565313e-06,
      "logits/chosen": 0.7068019509315491,
      "logits/rejected": 0.7005646824836731,
      "logps/chosen": -85.49761962890625,
      "logps/rejected": -146.3173828125,
      "loss": 0.5871,
      "rewards/accuracies": 0.9312499761581421,
      "rewards/chosen": -3.4726665019989014,
      "rewards/margins": 5.482043266296387,
      "rewards/rejected": -8.954710006713867,
      "step": 70
    },
    {
      "epoch": 1.8072289156626506,
      "grad_norm": 9.153736023927285,
      "learning_rate": 3.505215909271149e-06,
      "logits/chosen": 0.7240575551986694,
      "logits/rejected": 0.7156036496162415,
      "logps/chosen": -84.72471618652344,
      "logps/rejected": -153.99893188476562,
      "loss": 0.525,
      "rewards/accuracies": 0.9625000357627869,
      "rewards/chosen": -3.215219497680664,
      "rewards/margins": 6.307831764221191,
      "rewards/rejected": -9.523051261901855,
      "step": 75
    },
    {
      "epoch": 1.927710843373494,
      "grad_norm": 11.154887271165487,
      "learning_rate": 2.8902323799643116e-06,
      "logits/chosen": 0.6989457607269287,
      "logits/rejected": 0.6983063220977783,
      "logps/chosen": -82.57940673828125,
      "logps/rejected": -137.20379638671875,
      "loss": 0.5699,
      "rewards/accuracies": 0.9437499642372131,
      "rewards/chosen": -3.1769986152648926,
      "rewards/margins": 4.868185997009277,
      "rewards/rejected": -8.045184135437012,
      "step": 80
    },
    {
      "epoch": 2.0481927710843375,
      "grad_norm": 3.2266526930829826,
      "learning_rate": 2.311955924857113e-06,
      "logits/chosen": 0.6883406639099121,
      "logits/rejected": 0.6766339540481567,
      "logps/chosen": -75.12551879882812,
      "logps/rejected": -136.48965454101562,
      "loss": 0.4908,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": -2.6482861042022705,
      "rewards/margins": 5.347543716430664,
      "rewards/rejected": -7.9958295822143555,
      "step": 85
    },
    {
      "epoch": 2.1686746987951806,
      "grad_norm": 2.0384876528199296,
      "learning_rate": 1.780447764220422e-06,
      "logits/chosen": 0.6653567552566528,
      "logits/rejected": 0.6559253931045532,
      "logps/chosen": -68.8441162109375,
      "logps/rejected": -147.94723510742188,
      "loss": 0.3772,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/chosen": -1.8037909269332886,
      "rewards/margins": 7.034460544586182,
      "rewards/rejected": -8.838251113891602,
      "step": 90
    },
    {
      "epoch": 2.289156626506024,
      "grad_norm": 2.808270224177267,
      "learning_rate": 1.3049554138967052e-06,
      "logits/chosen": 0.6195204257965088,
      "logits/rejected": 0.6127552390098572,
      "logps/chosen": -66.5222396850586,
      "logps/rejected": -146.58905029296875,
      "loss": 0.3616,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.5321145057678223,
      "rewards/margins": 7.234241962432861,
      "rewards/rejected": -8.766357421875,
      "step": 95
    },
    {
      "epoch": 2.4096385542168672,
      "grad_norm": 3.195407301700039,
      "learning_rate": 8.937517911584321e-07,
      "logits/chosen": 0.5877372026443481,
      "logits/rejected": 0.5812422633171082,
      "logps/chosen": -64.000732421875,
      "logps/rejected": -146.1698455810547,
      "loss": 0.3419,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/chosen": -1.3494853973388672,
      "rewards/margins": 7.399579048156738,
      "rewards/rejected": -8.749064445495605,
      "step": 100
    },
    {
      "epoch": 2.5301204819277108,
      "grad_norm": 1.9211402938176323,
      "learning_rate": 5.539912772500943e-07,
      "logits/chosen": 0.570297360420227,
      "logits/rejected": 0.5654271841049194,
      "logps/chosen": -65.35507202148438,
      "logps/rejected": -140.493896484375,
      "loss": 0.3417,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": -1.4222447872161865,
      "rewards/margins": 6.939697265625,
      "rewards/rejected": -8.361942291259766,
      "step": 105
    },
    {
      "epoch": 2.6506024096385543,
      "grad_norm": 2.2064586790956002,
      "learning_rate": 2.915852409289421e-07,
      "logits/chosen": 0.5574687719345093,
      "logits/rejected": 0.549757182598114,
      "logps/chosen": -63.10077667236328,
      "logps/rejected": -142.78936767578125,
      "loss": 0.3327,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.1171743869781494,
      "rewards/margins": 7.488969802856445,
      "rewards/rejected": -8.606144905090332,
      "step": 110
    },
    {
      "epoch": 2.7710843373493974,
      "grad_norm": 2.3617812674765926,
      "learning_rate": 1.1109918872479642e-07,
      "logits/chosen": 0.551267683506012,
      "logits/rejected": 0.5471976399421692,
      "logps/chosen": -62.10367965698242,
      "logps/rejected": -138.47483825683594,
      "loss": 0.3382,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/chosen": -1.2061265707015991,
      "rewards/margins": 6.776681423187256,
      "rewards/rejected": -7.982807636260986,
      "step": 115
    },
    {
      "epoch": 2.891566265060241,
      "grad_norm": 2.7307102397847656,
      "learning_rate": 1.567333136387017e-08,
      "logits/chosen": 0.554203987121582,
      "logits/rejected": 0.5551311373710632,
      "logps/chosen": -64.15137481689453,
      "logps/rejected": -140.28311157226562,
      "loss": 0.3178,
      "rewards/accuracies": 1.0,
      "rewards/chosen": -1.1028921604156494,
      "rewards/margins": 7.438169956207275,
      "rewards/rejected": -8.541062355041504,
      "step": 120
    },
    {
      "epoch": 2.963855421686747,
      "step": 123,
      "total_flos": 8544080363520.0,
      "train_loss": 0.8105673431380978,
      "train_runtime": 814.398,
      "train_samples_per_second": 4.892,
      "train_steps_per_second": 0.151
    }
  ],
  "logging_steps": 5,
  "max_steps": 123,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8544080363520.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}