{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.984,
  "eval_steps": 100,
  "global_step": 124,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 3.846153846153847e-07,
      "logits/chosen": 0.02903342992067337,
      "logits/rejected": 0.16799500584602356,
      "logps/chosen": -204.7097930908203,
      "logps/rejected": -186.28207397460938,
      "loss": 0.0102,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.16,
      "learning_rate": 3.846153846153847e-06,
      "logits/chosen": 0.11909348517656326,
      "logits/rejected": 0.1480591893196106,
      "logps/chosen": -174.38600158691406,
      "logps/rejected": -139.39389038085938,
      "loss": 0.0103,
      "rewards/accuracies": 0.3541666567325592,
      "rewards/chosen": 0.0004580159147735685,
      "rewards/margins": -5.351095023797825e-05,
      "rewards/rejected": 0.0005115267704240978,
      "step": 10
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.951096619903317e-06,
      "logits/chosen": 0.1915196031332016,
      "logits/rejected": 0.26785004138946533,
      "logps/chosen": -186.1024169921875,
      "logps/rejected": -150.16329956054688,
      "loss": 0.0103,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": 8.462425466859713e-05,
      "rewards/margins": 0.00011474495113361627,
      "rewards/rejected": -3.012050910911057e-05,
      "step": 20
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.716164218065246e-06,
      "logits/chosen": 0.09372388571500778,
      "logits/rejected": 0.09187857806682587,
      "logps/chosen": -189.90634155273438,
      "logps/rejected": -176.62911987304688,
      "loss": 0.0102,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": 0.0012555012945085764,
      "rewards/margins": 0.0008926725131459534,
      "rewards/rejected": 0.0003628287522587925,
      "step": 30
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.3048902348863116e-06,
      "logits/chosen": 0.21573364734649658,
      "logits/rejected": 0.1305653154850006,
      "logps/chosen": -178.8751220703125,
      "logps/rejected": -151.7847137451172,
      "loss": 0.0106,
      "rewards/accuracies": 0.35624998807907104,
      "rewards/chosen": 0.0010144033003598452,
      "rewards/margins": -0.0007725629839114845,
      "rewards/rejected": 0.0017869662260636687,
      "step": 40
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.7500000000000005e-06,
      "logits/chosen": 0.10782618820667267,
      "logits/rejected": 0.1629345715045929,
      "logps/chosen": -185.59564208984375,
      "logps/rejected": -174.7571563720703,
      "loss": 0.0106,
      "rewards/accuracies": 0.32499998807907104,
      "rewards/chosen": -0.00029209800413809717,
      "rewards/margins": -0.0012864455347880721,
      "rewards/rejected": 0.0009943475015461445,
      "step": 50
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.0956464785579125e-06,
      "logits/chosen": 0.19225239753723145,
      "logits/rejected": 0.18530502915382385,
      "logps/chosen": -194.72091674804688,
      "logps/rejected": -159.44725036621094,
      "loss": 0.0101,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": 0.002120513701811433,
      "rewards/margins": 0.0013710735365748405,
      "rewards/rejected": 0.0007494401070289314,
      "step": 60
    },
    {
      "epoch": 1.12,
      "learning_rate": 2.39389699200963e-06,
      "logits/chosen": 0.18958896398544312,
      "logits/rejected": 0.14928244054317474,
      "logps/chosen": -196.82078552246094,
      "logps/rejected": -172.227294921875,
      "loss": 0.0107,
      "rewards/accuracies": 0.3687500059604645,
      "rewards/chosen": -0.00045417825458571315,
      "rewards/margins": -0.0015611432027071714,
      "rewards/rejected": 0.0011069647734984756,
      "step": 70
    },
    {
      "epoch": 1.28,
      "learning_rate": 1.700590188571887e-06,
      "logits/chosen": 0.15342679619789124,
      "logits/rejected": 0.13346508145332336,
      "logps/chosen": -176.46441650390625,
      "logps/rejected": -149.63717651367188,
      "loss": 0.0102,
      "rewards/accuracies": 0.39375001192092896,
      "rewards/chosen": 0.0011017677607014775,
      "rewards/margins": 0.0006318273372016847,
      "rewards/rejected": 0.000469940627226606,
      "step": 80
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.0708929268538034e-06,
      "logits/chosen": 0.19051842391490936,
      "logits/rejected": 0.15732800960540771,
      "logps/chosen": -183.8511199951172,
      "logps/rejected": -157.0377197265625,
      "loss": 0.0104,
      "rewards/accuracies": 0.34375,
      "rewards/chosen": 0.000780799426138401,
      "rewards/margins": -0.000656499934848398,
      "rewards/rejected": 0.0014372995356097817,
      "step": 90
    },
    {
      "epoch": 1.6,
      "learning_rate": 5.549106142039018e-07,
      "logits/chosen": 0.16213683784008026,
      "logits/rejected": 0.04842492565512657,
      "logps/chosen": -177.85955810546875,
      "logps/rejected": -152.51058959960938,
      "loss": 0.0102,
      "rewards/accuracies": 0.39375001192092896,
      "rewards/chosen": 0.0025749087799340487,
      "rewards/margins": 0.0006177256000228226,
      "rewards/rejected": 0.001957183238118887,
      "step": 100
    },
    {
      "epoch": 1.6,
      "eval_logits/chosen": -0.007027674000710249,
      "eval_logits/rejected": 0.09070703387260437,
      "eval_logps/chosen": -306.4146728515625,
      "eval_logps/rejected": -278.6007995605469,
      "eval_loss": 0.010791419073939323,
      "eval_rewards/accuracies": 0.4830000102519989,
      "eval_rewards/chosen": 0.001202387735247612,
      "eval_rewards/margins": -0.00015190700651146472,
      "eval_rewards/rejected": 0.001354294829070568,
      "eval_runtime": 432.9902,
      "eval_samples_per_second": 4.619,
      "eval_steps_per_second": 1.155,
      "step": 100
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.937002879188285e-07,
      "logits/chosen": 0.22388437390327454,
      "logits/rejected": 0.19077368080615997,
      "logps/chosen": -205.4290008544922,
      "logps/rejected": -172.20669555664062,
      "loss": 0.0104,
      "rewards/accuracies": 0.3499999940395355,
      "rewards/chosen": 0.00022258120588958263,
      "rewards/margins": -0.00020028270955663174,
      "rewards/rejected": 0.0004228639299981296,
      "step": 110
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.6003680950742728e-08,
      "logits/chosen": 0.17824237048625946,
      "logits/rejected": 0.234628364443779,
      "logps/chosen": -184.6617431640625,
      "logps/rejected": -160.73251342773438,
      "loss": 0.0104,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": 0.0013891400303691626,
      "rewards/margins": 9.483665053267032e-05,
      "rewards/rejected": 0.0012943033361807466,
      "step": 120
    },
    {
      "epoch": 1.98,
      "step": 124,
      "total_flos": 0.0,
      "train_loss": 0.010348274312432735,
      "train_runtime": 1408.9952,
      "train_samples_per_second": 1.419,
      "train_steps_per_second": 0.088
    }
  ],
  "logging_steps": 10,
  "max_steps": 124,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}