{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9968652037617555,
  "eval_steps": 500,
  "global_step": 159,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 3.125e-08,
      "logits/chosen": -2.843719482421875,
      "logits/rejected": -2.8718748092651367,
      "logps/chosen": -340.24505615234375,
      "logps/pi_response": -88.81813049316406,
      "logps/ref_response": -88.81813049316406,
      "logps/rejected": -141.60296630859375,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.06,
      "learning_rate": 3.1249999999999997e-07,
      "logits/chosen": -2.775027275085449,
      "logits/rejected": -2.7582144737243652,
      "logps/chosen": -225.69427490234375,
      "logps/pi_response": -73.32665252685547,
      "logps/ref_response": -73.0557632446289,
      "logps/rejected": -127.79769134521484,
      "loss": 0.6918,
      "rewards/accuracies": 0.5486111044883728,
      "rewards/chosen": 0.0009001567959785461,
      "rewards/margins": 0.0027927933260798454,
      "rewards/rejected": -0.0018926364136859775,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.990353313429303e-07,
      "logits/chosen": -2.7578864097595215,
      "logits/rejected": -2.7449183464050293,
      "logps/chosen": -216.22109985351562,
      "logps/pi_response": -79.10755920410156,
      "logps/ref_response": -73.02125549316406,
      "logps/rejected": -117.88703918457031,
      "loss": 0.6655,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.005766990128904581,
      "rewards/margins": 0.042582910507917404,
      "rewards/rejected": -0.03681592270731926,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.882681251368548e-07,
      "logits/chosen": -2.636462688446045,
      "logits/rejected": -2.6021816730499268,
      "logps/chosen": -240.2003936767578,
      "logps/pi_response": -100.06871032714844,
      "logps/ref_response": -68.24726867675781,
      "logps/rejected": -125.3204345703125,
      "loss": 0.6227,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": -0.08825792372226715,
      "rewards/margins": 0.159133642911911,
      "rewards/rejected": -0.24739158153533936,
      "step": 30
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.6604720940421207e-07,
      "logits/chosen": -2.5370068550109863,
      "logits/rejected": -2.5128865242004395,
      "logps/chosen": -271.98651123046875,
      "logps/pi_response": -144.71682739257812,
      "logps/ref_response": -75.6551742553711,
      "logps/rejected": -181.0801544189453,
      "loss": 0.5839,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": -0.23095576465129852,
      "rewards/margins": 0.38431650400161743,
      "rewards/rejected": -0.6152722239494324,
      "step": 40
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.3344075855595097e-07,
      "logits/chosen": -2.5748300552368164,
      "logits/rejected": -2.5561716556549072,
      "logps/chosen": -276.7567138671875,
      "logps/pi_response": -160.7132110595703,
      "logps/ref_response": -71.23826599121094,
      "logps/rejected": -182.82472229003906,
      "loss": 0.5987,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.3602275252342224,
      "rewards/margins": 0.41353344917297363,
      "rewards/rejected": -0.7737610340118408,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.920161866827889e-07,
      "logits/chosen": -2.5654828548431396,
      "logits/rejected": -2.519744634628296,
      "logps/chosen": -221.7403106689453,
      "logps/pi_response": -161.64913940429688,
      "logps/ref_response": -67.74427795410156,
      "logps/rejected": -205.1297607421875,
      "loss": 0.5637,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": -0.35707658529281616,
      "rewards/margins": 0.4391842782497406,
      "rewards/rejected": -0.7962608337402344,
      "step": 60
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4376480090239047e-07,
      "logits/chosen": -2.5978214740753174,
      "logits/rejected": -2.557990550994873,
      "logps/chosen": -270.705810546875,
      "logps/pi_response": -191.92599487304688,
      "logps/ref_response": -73.56436157226562,
      "logps/rejected": -205.6122589111328,
      "loss": 0.5327,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": -0.4569978713989258,
      "rewards/margins": 0.58502596616745,
      "rewards/rejected": -1.0420238971710205,
      "step": 70
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.910060778827554e-07,
      "logits/chosen": -2.6471352577209473,
      "logits/rejected": -2.6175568103790283,
      "logps/chosen": -273.21337890625,
      "logps/pi_response": -190.89492797851562,
      "logps/ref_response": -77.31620025634766,
      "logps/rejected": -209.2509307861328,
      "loss": 0.5247,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": -0.4046536982059479,
      "rewards/margins": 0.6309649348258972,
      "rewards/rejected": -1.035618543624878,
      "step": 80
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.3627616503391812e-07,
      "logits/chosen": -2.5935592651367188,
      "logits/rejected": -2.5323500633239746,
      "logps/chosen": -294.3185119628906,
      "logps/pi_response": -193.82952880859375,
      "logps/ref_response": -72.20295715332031,
      "logps/rejected": -214.9334716796875,
      "loss": 0.5214,
      "rewards/accuracies": 0.800000011920929,
      "rewards/chosen": -0.4483931064605713,
      "rewards/margins": 0.6660177707672119,
      "rewards/rejected": -1.1144107580184937,
      "step": 90
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8220596619089573e-07,
      "logits/chosen": -2.584444522857666,
      "logits/rejected": -2.538541078567505,
      "logps/chosen": -310.3904113769531,
      "logps/pi_response": -218.18429565429688,
      "logps/ref_response": -74.52151489257812,
      "logps/rejected": -236.4603271484375,
      "loss": 0.524,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": -0.604050874710083,
      "rewards/margins": 0.6805473566055298,
      "rewards/rejected": -1.2845981121063232,
      "step": 100
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.3139467229135998e-07,
      "logits/chosen": -2.508607864379883,
      "logits/rejected": -2.4556093215942383,
      "logps/chosen": -275.2364501953125,
      "logps/pi_response": -222.62771606445312,
      "logps/ref_response": -63.37713623046875,
      "logps/rejected": -252.931640625,
      "loss": 0.4936,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.8243808746337891,
      "rewards/margins": 0.6452677845954895,
      "rewards/rejected": -1.4696485996246338,
      "step": 110
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.628481651367875e-08,
      "logits/chosen": -2.523491621017456,
      "logits/rejected": -2.471564292907715,
      "logps/chosen": -280.9041748046875,
      "logps/pi_response": -221.3182373046875,
      "logps/ref_response": -66.3117446899414,
      "logps/rejected": -240.2959442138672,
      "loss": 0.483,
      "rewards/accuracies": 0.8374999761581421,
      "rewards/chosen": -0.5750901699066162,
      "rewards/margins": 0.953150749206543,
      "rewards/rejected": -1.5282410383224487,
      "step": 120
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.904486005914027e-08,
      "logits/chosen": -2.496025800704956,
      "logits/rejected": -2.4560306072235107,
      "logps/chosen": -316.5722961425781,
      "logps/pi_response": -236.9508056640625,
      "logps/ref_response": -76.07049560546875,
      "logps/rejected": -261.31170654296875,
      "loss": 0.4892,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": -0.6082834601402283,
      "rewards/margins": 0.9130091667175293,
      "rewards/rejected": -1.5212925672531128,
      "step": 130
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.1464952759020856e-08,
      "logits/chosen": -2.502040147781372,
      "logits/rejected": -2.4415221214294434,
      "logps/chosen": -303.0337219238281,
      "logps/pi_response": -231.24246215820312,
      "logps/ref_response": -78.19743347167969,
      "logps/rejected": -247.8349151611328,
      "loss": 0.4841,
      "rewards/accuracies": 0.793749988079071,
      "rewards/chosen": -0.6394041776657104,
      "rewards/margins": 0.7987918257713318,
      "rewards/rejected": -1.4381959438323975,
      "step": 140
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.8708793644441086e-09,
      "logits/chosen": -2.5210301876068115,
      "logits/rejected": -2.4584171772003174,
      "logps/chosen": -301.7514343261719,
      "logps/pi_response": -243.6031494140625,
      "logps/ref_response": -77.64109802246094,
      "logps/rejected": -281.96185302734375,
      "loss": 0.489,
      "rewards/accuracies": 0.75,
      "rewards/chosen": -0.6782695055007935,
      "rewards/margins": 0.8968443870544434,
      "rewards/rejected": -1.5751138925552368,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 159,
      "total_flos": 0.0,
      "train_loss": 0.5467530106598476,
      "train_runtime": 4156.701,
      "train_samples_per_second": 4.902,
      "train_steps_per_second": 0.038
    }
  ],
  "logging_steps": 10,
  "max_steps": 159,
  "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
}