{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.998272884283247,
  "eval_steps": 100,
  "global_step": 289,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 0.2451171875,
      "learning_rate": 1.7241379310344828e-07,
      "logits/chosen": -2.946108341217041,
      "logits/rejected": -2.9481201171875,
      "logps/chosen": -27.542381286621094,
      "logps/rejected": -28.998493194580078,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.2431640625,
      "learning_rate": 1.724137931034483e-06,
      "logits/chosen": -2.967113494873047,
      "logits/rejected": -2.9693994522094727,
      "logps/chosen": -27.679420471191406,
      "logps/rejected": -30.345211029052734,
      "loss": 0.6931,
      "rewards/accuracies": 0.4166666567325592,
      "rewards/chosen": -0.00010507476690690964,
      "rewards/margins": 6.562391354236752e-05,
      "rewards/rejected": -0.0001706986950011924,
      "step": 10
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.259765625,
      "learning_rate": 3.448275862068966e-06,
      "logits/chosen": -2.9687092304229736,
      "logits/rejected": -2.9725170135498047,
      "logps/chosen": -27.916067123413086,
      "logps/rejected": -32.21611404418945,
      "loss": 0.6923,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": 0.0016102585941553116,
      "rewards/margins": 0.0016238940879702568,
      "rewards/rejected": -1.3635644791065715e-05,
      "step": 20
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.263671875,
      "learning_rate": 4.999817502139027e-06,
      "logits/chosen": -2.991443157196045,
      "logits/rejected": -2.992274761199951,
      "logps/chosen": -27.992717742919922,
      "logps/rejected": -30.998449325561523,
      "loss": 0.6894,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.010343936271965504,
      "rewards/margins": 0.007500241510570049,
      "rewards/rejected": 0.002843693597242236,
      "step": 30
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.359375,
      "learning_rate": 4.977949980164773e-06,
      "logits/chosen": -2.9539809226989746,
      "logits/rejected": -2.9519870281219482,
      "logps/chosen": -26.01015853881836,
      "logps/rejected": -30.401302337646484,
      "loss": 0.6829,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.03149240463972092,
      "rewards/margins": 0.02063887007534504,
      "rewards/rejected": 0.010853536427021027,
      "step": 40
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.34375,
      "learning_rate": 4.919948367622307e-06,
      "logits/chosen": -2.983638048171997,
      "logits/rejected": -2.989013195037842,
      "logps/chosen": -22.443767547607422,
      "logps/rejected": -31.806631088256836,
      "loss": 0.6664,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.06118880584836006,
      "rewards/margins": 0.05425649881362915,
      "rewards/rejected": 0.006932304706424475,
      "step": 50
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.484375,
      "learning_rate": 4.8266584586307555e-06,
      "logits/chosen": -2.988079786300659,
      "logits/rejected": -2.9914956092834473,
      "logps/chosen": -19.77129364013672,
      "logps/rejected": -29.180835723876953,
      "loss": 0.6535,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 0.08603720366954803,
      "rewards/margins": 0.08121357858181,
      "rewards/rejected": 0.004823616705834866,
      "step": 60
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.8515625,
      "learning_rate": 4.699440630133794e-06,
      "logits/chosen": -2.995100736618042,
      "logits/rejected": -2.99615216255188,
      "logps/chosen": -17.9371395111084,
      "logps/rejected": -34.77935028076172,
      "loss": 0.6236,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.11037828773260117,
      "rewards/margins": 0.14534664154052734,
      "rewards/rejected": -0.03496835380792618,
      "step": 70
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.84765625,
      "learning_rate": 4.5401500045405126e-06,
      "logits/chosen": -2.9611220359802246,
      "logits/rejected": -2.963106393814087,
      "logps/chosen": -11.03026008605957,
      "logps/rejected": -39.37724304199219,
      "loss": 0.5669,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.18076485395431519,
      "rewards/margins": 0.27246183156967163,
      "rewards/rejected": -0.09169697761535645,
      "step": 80
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.59765625,
      "learning_rate": 4.351109397863309e-06,
      "logits/chosen": -2.970407009124756,
      "logits/rejected": -2.9756150245666504,
      "logps/chosen": -6.9094672203063965,
      "logps/rejected": -49.88671112060547,
      "loss": 0.5168,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.20805075764656067,
      "rewards/margins": 0.3929779827594757,
      "rewards/rejected": -0.18492721021175385,
      "step": 90
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.828125,
      "learning_rate": 4.135075447829912e-06,
      "logits/chosen": -2.966463327407837,
      "logits/rejected": -2.973740816116333,
      "logps/chosen": -5.979574203491211,
      "logps/rejected": -56.75745391845703,
      "loss": 0.4825,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.23063524067401886,
      "rewards/margins": 0.4814692437648773,
      "rewards/rejected": -0.25083404779434204,
      "step": 100
    },
    {
      "epoch": 0.35,
      "eval_logits/chosen": -2.971499443054199,
      "eval_logits/rejected": -2.975092887878418,
      "eval_logps/chosen": -4.676602840423584,
      "eval_logps/rejected": -60.5069694519043,
      "eval_loss": 0.46322527527809143,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 0.2303335815668106,
      "eval_rewards/margins": 0.5325379371643066,
      "eval_rewards/rejected": -0.3022043704986572,
      "eval_runtime": 59.0284,
      "eval_samples_per_second": 4.896,
      "eval_steps_per_second": 2.456,
      "step": 100
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.15625,
      "learning_rate": 3.895198415897896e-06,
      "logits/chosen": -2.944152355194092,
      "logits/rejected": -2.948521614074707,
      "logps/chosen": -4.758464813232422,
      "logps/rejected": -65.92622375488281,
      "loss": 0.4433,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.23539750277996063,
      "rewards/margins": 0.5871593356132507,
      "rewards/rejected": -0.3517618179321289,
      "step": 110
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.515625,
      "learning_rate": 3.634976249348867e-06,
      "logits/chosen": -2.956526279449463,
      "logits/rejected": -2.9596874713897705,
      "logps/chosen": -3.4839115142822266,
      "logps/rejected": -74.23634338378906,
      "loss": 0.4105,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2543177604675293,
      "rewards/margins": 0.6815550923347473,
      "rewards/rejected": -0.427237331867218,
      "step": 120
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.65625,
      "learning_rate": 3.3582035733403963e-06,
      "logits/chosen": -2.9395928382873535,
      "logits/rejected": -2.9400253295898438,
      "logps/chosen": -3.3433098793029785,
      "logps/rejected": -81.93199920654297,
      "loss": 0.384,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.25131431221961975,
      "rewards/margins": 0.7623029947280884,
      "rewards/rejected": -0.5109885931015015,
      "step": 130
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.56640625,
      "learning_rate": 3.068916356726475e-06,
      "logits/chosen": -2.9392318725585938,
      "logits/rejected": -2.939736843109131,
      "logps/chosen": -2.7690482139587402,
      "logps/rejected": -91.77839660644531,
      "loss": 0.3548,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.24845953285694122,
      "rewards/margins": 0.8564491271972656,
      "rewards/rejected": -0.607989490032196,
      "step": 140
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.52734375,
      "learning_rate": 2.771333058543416e-06,
      "logits/chosen": -2.9237265586853027,
      "logits/rejected": -2.9198527336120605,
      "logps/chosen": -3.30281138420105,
      "logps/rejected": -96.4010009765625,
      "loss": 0.3403,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2496950924396515,
      "rewards/margins": 0.9052786827087402,
      "rewards/rejected": -0.6555835604667664,
      "step": 150
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.50390625,
      "learning_rate": 2.4697931133779566e-06,
      "logits/chosen": -2.9088094234466553,
      "logits/rejected": -2.9039511680603027,
      "logps/chosen": -3.719200849533081,
      "logps/rejected": -103.43408203125,
      "loss": 0.3218,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.24762046337127686,
      "rewards/margins": 0.9735357165336609,
      "rewards/rejected": -0.7259151935577393,
      "step": 160
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.96875,
      "learning_rate": 2.168693652639432e-06,
      "logits/chosen": -2.9292285442352295,
      "logits/rejected": -2.9254345893859863,
      "logps/chosen": -1.5337612628936768,
      "logps/rejected": -105.90274810791016,
      "loss": 0.3082,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2660350203514099,
      "rewards/margins": 1.021663784980774,
      "rewards/rejected": -0.7556287050247192,
      "step": 170
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.5546875,
      "learning_rate": 1.872425384482346e-06,
      "logits/chosen": -2.9200375080108643,
      "logits/rejected": -2.9096457958221436,
      "logps/chosen": -1.4461075067520142,
      "logps/rejected": -111.00732421875,
      "loss": 0.2964,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2637704610824585,
      "rewards/margins": 1.0691814422607422,
      "rewards/rejected": -0.8054108619689941,
      "step": 180
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.6875,
      "learning_rate": 1.5853085673944695e-06,
      "logits/chosen": -2.9019381999969482,
      "logits/rejected": -2.889967441558838,
      "logps/chosen": -3.121753692626953,
      "logps/rejected": -114.66499328613281,
      "loss": 0.2975,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.24539904296398163,
      "rewards/margins": 1.0781965255737305,
      "rewards/rejected": -0.8327974081039429,
      "step": 190
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.0859375,
      "learning_rate": 1.3115300110997097e-06,
      "logits/chosen": -2.867145538330078,
      "logits/rejected": -2.857390880584717,
      "logps/chosen": -4.095256805419922,
      "logps/rejected": -123.1200180053711,
      "loss": 0.2786,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.24890461564064026,
      "rewards/margins": 1.1587778329849243,
      "rewards/rejected": -0.9098731875419617,
      "step": 200
    },
    {
      "epoch": 0.69,
      "eval_logits/chosen": -2.8807153701782227,
      "eval_logits/rejected": -2.8679380416870117,
      "eval_logps/chosen": -6.8403167724609375,
      "eval_logps/rejected": -132.28236389160156,
      "eval_loss": 0.2587815523147583,
      "eval_rewards/accuracies": 1.0,
      "eval_rewards/chosen": 0.2086964100599289,
      "eval_rewards/margins": 1.2286547422409058,
      "eval_rewards/rejected": -1.019958257675171,
      "eval_runtime": 59.0007,
      "eval_samples_per_second": 4.898,
      "eval_steps_per_second": 2.458,
      "step": 200
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.0390625,
      "learning_rate": 1.0550820234444627e-06,
      "logits/chosen": -2.8844265937805176,
      "logits/rejected": -2.8720266819000244,
      "logps/chosen": -7.328047752380371,
      "logps/rejected": -136.2236785888672,
      "loss": 0.2536,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2088628113269806,
      "rewards/margins": 1.2624285221099854,
      "rewards/rejected": -1.0535659790039062,
      "step": 210
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5,
      "learning_rate": 8.197041935593181e-07,
      "logits/chosen": -2.865748643875122,
      "logits/rejected": -2.8470582962036133,
      "logps/chosen": -3.5205326080322266,
      "logps/rejected": -143.3411865234375,
      "loss": 0.2291,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2387482225894928,
      "rewards/margins": 1.365382432937622,
      "rewards/rejected": -1.126634120941162,
      "step": 220
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.97265625,
      "learning_rate": 6.088288602287159e-07,
      "logits/chosen": -2.8658416271209717,
      "logits/rejected": -2.8466126918792725,
      "logps/chosen": -1.298507809638977,
      "logps/rejected": -147.50143432617188,
      "loss": 0.2162,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2748534083366394,
      "rewards/margins": 1.4337130784988403,
      "rewards/rejected": -1.1588596105575562,
      "step": 230
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.34375,
      "learning_rate": 4.255310606625124e-07,
      "logits/chosen": -2.8704545497894287,
      "logits/rejected": -2.8518197536468506,
      "logps/chosen": -2.5422964096069336,
      "logps/rejected": -151.9074249267578,
      "loss": 0.2081,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.25674667954444885,
      "rewards/margins": 1.4753159284591675,
      "rewards/rejected": -1.2185693979263306,
      "step": 240
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.9921875,
      "learning_rate": 2.7248368952908055e-07,
      "logits/chosen": -2.861412525177002,
      "logits/rejected": -2.838165044784546,
      "logps/chosen": -2.5676398277282715,
      "logps/rejected": -152.24969482421875,
      "loss": 0.2105,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2601901888847351,
      "rewards/margins": 1.4616148471832275,
      "rewards/rejected": -1.2014245986938477,
      "step": 250
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.88671875,
      "learning_rate": 1.5191852213221198e-07,
      "logits/chosen": -2.8463876247406006,
      "logits/rejected": -2.817927837371826,
      "logps/chosen": -3.504025936126709,
      "logps/rejected": -156.09738159179688,
      "loss": 0.2054,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2497003823518753,
      "rewards/margins": 1.4921144247055054,
      "rewards/rejected": -1.2424139976501465,
      "step": 260
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.921875,
      "learning_rate": 6.559367010166629e-08,
      "logits/chosen": -2.8501267433166504,
      "logits/rejected": -2.8306031227111816,
      "logps/chosen": -2.110583782196045,
      "logps/rejected": -152.65463256835938,
      "loss": 0.2113,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.26196378469467163,
      "rewards/margins": 1.4719316959381104,
      "rewards/rejected": -1.209967851638794,
      "step": 270
    },
    {
      "epoch": 0.97,
      "grad_norm": 1.453125,
      "learning_rate": 1.4767944166687032e-08,
      "logits/chosen": -2.8537425994873047,
      "logits/rejected": -2.827571392059326,
      "logps/chosen": -2.493337869644165,
      "logps/rejected": -154.96142578125,
      "loss": 0.2011,
      "rewards/accuracies": 1.0,
      "rewards/chosen": 0.2604244649410248,
      "rewards/margins": 1.5092506408691406,
      "rewards/rejected": -1.2488261461257935,
      "step": 280
    },
    {
      "epoch": 1.0,
      "step": 289,
      "total_flos": 0.0,
      "train_loss": 0.4018066387275511,
      "train_runtime": 624.4981,
      "train_samples_per_second": 1.853,
      "train_steps_per_second": 0.463
    }
  ],
  "logging_steps": 10,
  "max_steps": 289,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 200,
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}