{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9972451790633609,
  "eval_steps": 100,
  "global_step": 181,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "grad_norm": 4.4602671481986516,
      "learning_rate": 2.6315789473684208e-08,
      "logits/chosen": -0.7853389978408813,
      "logits/rejected": -0.651086151599884,
      "logps/chosen": -1158.1322021484375,
      "logps/rejected": -1143.569091796875,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.676508624329794,
      "learning_rate": 2.631578947368421e-07,
      "logits/chosen": -0.709392786026001,
      "logits/rejected": -0.8101767897605896,
      "logps/chosen": -995.121826171875,
      "logps/rejected": -1316.8297119140625,
      "loss": 0.693,
      "rewards/accuracies": 0.4305555522441864,
      "rewards/chosen": -0.0007655912195332348,
      "rewards/margins": -0.00027954205870628357,
      "rewards/rejected": -0.0004860491317231208,
      "step": 10
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.611380190345714,
      "learning_rate": 4.999529926121253e-07,
      "logits/chosen": -0.7076992988586426,
      "logits/rejected": -0.826177716255188,
      "logps/chosen": -942.4779052734375,
      "logps/rejected": -1303.2080078125,
      "loss": 0.6922,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": 0.0009432749939151108,
      "rewards/margins": 0.002406059531494975,
      "rewards/rejected": -0.0014627845957875252,
      "step": 20
    },
    {
      "epoch": 0.17,
      "grad_norm": 4.545137388238839,
      "learning_rate": 4.943334645626589e-07,
      "logits/chosen": -0.7218513488769531,
      "logits/rejected": -0.8190325498580933,
      "logps/chosen": -986.5565185546875,
      "logps/rejected": -1347.8033447265625,
      "loss": 0.6883,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.004364156164228916,
      "rewards/margins": 0.008785699494183064,
      "rewards/rejected": -0.004421542398631573,
      "step": 30
    },
    {
      "epoch": 0.22,
      "grad_norm": 4.401192871872299,
      "learning_rate": 4.795540267200686e-07,
      "logits/chosen": -0.7325557470321655,
      "logits/rejected": -0.7795466184616089,
      "logps/chosen": -1022.5738525390625,
      "logps/rejected": -1272.192626953125,
      "loss": 0.6826,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.01184375025331974,
      "rewards/margins": 0.019737884402275085,
      "rewards/rejected": -0.007894134148955345,
      "step": 40
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.67072465163404,
      "learning_rate": 4.561687510272767e-07,
      "logits/chosen": -0.7071916460990906,
      "logits/rejected": -0.7952053546905518,
      "logps/chosen": -973.41259765625,
      "logps/rejected": -1254.411376953125,
      "loss": 0.6759,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.010998593643307686,
      "rewards/margins": 0.031063910573720932,
      "rewards/rejected": -0.020065316930413246,
      "step": 50
    },
    {
      "epoch": 0.33,
      "grad_norm": 4.340731797997725,
      "learning_rate": 4.2505433694179213e-07,
      "logits/chosen": -0.7227746248245239,
      "logits/rejected": -0.8140878677368164,
      "logps/chosen": -1035.677734375,
      "logps/rejected": -1306.7808837890625,
      "loss": 0.6642,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": 0.017496155574917793,
      "rewards/margins": 0.06553898006677628,
      "rewards/rejected": -0.04804282635450363,
      "step": 60
    },
    {
      "epoch": 0.39,
      "grad_norm": 4.496628183792233,
      "learning_rate": 3.873772445177015e-07,
      "logits/chosen": -0.7832368612289429,
      "logits/rejected": -0.7676926851272583,
      "logps/chosen": -1028.62890625,
      "logps/rejected": -1254.6610107421875,
      "loss": 0.6538,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.01262822188436985,
      "rewards/margins": 0.07675839215517044,
      "rewards/rejected": -0.06413016468286514,
      "step": 70
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.939508062028092,
      "learning_rate": 3.445499645429106e-07,
      "logits/chosen": -0.7500189542770386,
      "logits/rejected": -0.8156210780143738,
      "logps/chosen": -1028.065185546875,
      "logps/rejected": -1316.736328125,
      "loss": 0.6433,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": 0.0009801089763641357,
      "rewards/margins": 0.0968375876545906,
      "rewards/rejected": -0.09585747867822647,
      "step": 80
    },
    {
      "epoch": 0.5,
      "grad_norm": 4.7399950008550205,
      "learning_rate": 2.981780651370224e-07,
      "logits/chosen": -0.7468028664588928,
      "logits/rejected": -0.7884877920150757,
      "logps/chosen": -996.8697509765625,
      "logps/rejected": -1192.45654296875,
      "loss": 0.6079,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": 0.004719099495559931,
      "rewards/margins": 0.21119482815265656,
      "rewards/rejected": -0.20647573471069336,
      "step": 90
    },
    {
      "epoch": 0.55,
      "grad_norm": 5.466151335512206,
      "learning_rate": 2.5e-07,
      "logits/chosen": -0.7582017183303833,
      "logits/rejected": -0.8466933369636536,
      "logps/chosen": -979.6242065429688,
      "logps/rejected": -1357.511962890625,
      "loss": 0.592,
      "rewards/accuracies": 0.8187500238418579,
      "rewards/chosen": -0.00921861082315445,
      "rewards/margins": 0.2683759033679962,
      "rewards/rejected": -0.27759450674057007,
      "step": 100
    },
    {
      "epoch": 0.55,
      "eval_logits/chosen": -0.7069126963615417,
      "eval_logits/rejected": -0.8721536993980408,
      "eval_logps/chosen": -955.1961059570312,
      "eval_logps/rejected": -1615.9989013671875,
      "eval_loss": 0.6292956471443176,
      "eval_rewards/accuracies": 0.7796609997749329,
      "eval_rewards/chosen": -0.0145129868760705,
      "eval_rewards/margins": 0.14818498492240906,
      "eval_rewards/rejected": -0.16269797086715698,
      "eval_runtime": 214.9002,
      "eval_samples_per_second": 8.674,
      "eval_steps_per_second": 0.275,
      "step": 100
    },
    {
      "epoch": 0.61,
      "grad_norm": 4.779970412172738,
      "learning_rate": 2.018219348629775e-07,
      "logits/chosen": -0.779070258140564,
      "logits/rejected": -0.8311734199523926,
      "logps/chosen": -992.7100830078125,
      "logps/rejected": -1354.688720703125,
      "loss": 0.5575,
      "rewards/accuracies": 0.856249988079071,
      "rewards/chosen": -0.0526651069521904,
      "rewards/margins": 0.4234324097633362,
      "rewards/rejected": -0.4760975241661072,
      "step": 110
    },
    {
      "epoch": 0.66,
      "grad_norm": 5.9180549707398695,
      "learning_rate": 1.554500354570894e-07,
      "logits/chosen": -0.7683907747268677,
      "logits/rejected": -0.8335941433906555,
      "logps/chosen": -1003.9620971679688,
      "logps/rejected": -1312.5740966796875,
      "loss": 0.5428,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": -0.06494404375553131,
      "rewards/margins": 0.5937026739120483,
      "rewards/rejected": -0.6586467623710632,
      "step": 120
    },
    {
      "epoch": 0.72,
      "grad_norm": 5.7132927229572354,
      "learning_rate": 1.126227554822985e-07,
      "logits/chosen": -0.798209011554718,
      "logits/rejected": -0.8723245859146118,
      "logps/chosen": -1054.9263916015625,
      "logps/rejected": -1349.2750244140625,
      "loss": 0.5312,
      "rewards/accuracies": 0.8125,
      "rewards/chosen": -0.15924784541130066,
      "rewards/margins": 0.5923423767089844,
      "rewards/rejected": -0.7515901923179626,
      "step": 130
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.497632384133095,
      "learning_rate": 7.494566305820788e-08,
      "logits/chosen": -0.8022674322128296,
      "logits/rejected": -0.882843017578125,
      "logps/chosen": -1055.48974609375,
      "logps/rejected": -1417.170654296875,
      "loss": 0.4988,
      "rewards/accuracies": 0.862500011920929,
      "rewards/chosen": -0.17785976827144623,
      "rewards/margins": 1.1309340000152588,
      "rewards/rejected": -1.3087936639785767,
      "step": 140
    },
    {
      "epoch": 0.83,
      "grad_norm": 6.590111680094942,
      "learning_rate": 4.38312489727233e-08,
      "logits/chosen": -0.7941451668739319,
      "logits/rejected": -0.8360635042190552,
      "logps/chosen": -994.5494384765625,
      "logps/rejected": -1374.3560791015625,
      "loss": 0.4917,
      "rewards/accuracies": 0.8687499761581421,
      "rewards/chosen": -0.2062647044658661,
      "rewards/margins": 0.8967850804328918,
      "rewards/rejected": -1.1030497550964355,
      "step": 150
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.361359149599993,
      "learning_rate": 2.044597327993153e-08,
      "logits/chosen": -0.7812503576278687,
      "logits/rejected": -0.8671078681945801,
      "logps/chosen": -1042.375244140625,
      "logps/rejected": -1443.9189453125,
      "loss": 0.4813,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -0.2346261441707611,
      "rewards/margins": 0.8702405095100403,
      "rewards/rejected": -1.104866623878479,
      "step": 160
    },
    {
      "epoch": 0.94,
      "grad_norm": 7.649602408833162,
      "learning_rate": 5.666535437341108e-09,
      "logits/chosen": -0.792350172996521,
      "logits/rejected": -0.8578357696533203,
      "logps/chosen": -1057.263427734375,
      "logps/rejected": -1363.985107421875,
      "loss": 0.4843,
      "rewards/accuracies": 0.8062499761581421,
      "rewards/chosen": -0.292695015668869,
      "rewards/margins": 1.1091253757476807,
      "rewards/rejected": -1.401820421218872,
      "step": 170
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.116244988839705,
      "learning_rate": 4.700738787466463e-11,
      "logits/chosen": -0.7855504751205444,
      "logits/rejected": -0.8505135774612427,
      "logps/chosen": -952.7326049804688,
      "logps/rejected": -1447.79541015625,
      "loss": 0.4771,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.23469653725624084,
      "rewards/margins": 0.9054014086723328,
      "rewards/rejected": -1.140097975730896,
      "step": 180
    },
    {
      "epoch": 1.0,
      "step": 181,
      "total_flos": 0.0,
      "train_loss": 0.5914445311983646,
      "train_runtime": 2922.1625,
      "train_samples_per_second": 3.971,
      "train_steps_per_second": 0.062
    }
  ],
  "logging_steps": 10,
  "max_steps": 181,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}