{
  "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": -2.6679039001464844,
      "logits/rejected": -2.599691390991211,
      "logps/chosen": -164.941650390625,
      "logps/pi_response": -79.30451965332031,
      "logps/ref_response": -79.30451965332031,
      "logps/rejected": -111.52906799316406,
      "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": -2.6902599334716797,
      "logits/rejected": -2.705916404724121,
      "logps/chosen": -236.0708770751953,
      "logps/pi_response": -118.64096069335938,
      "logps/ref_response": -118.30906677246094,
      "logps/rejected": -228.03765869140625,
      "loss": 0.6928,
      "rewards/accuracies": 0.4444444477558136,
      "rewards/chosen": -0.0036212028935551643,
      "rewards/margins": -0.0003877294948324561,
      "rewards/rejected": -0.0032334737479686737,
      "step": 10
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.990353313429303e-07,
      "logits/chosen": -2.7183661460876465,
      "logits/rejected": -2.6849148273468018,
      "logps/chosen": -265.54229736328125,
      "logps/pi_response": -142.77395629882812,
      "logps/ref_response": -130.1978302001953,
      "logps/rejected": -228.3546905517578,
      "loss": 0.6905,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": -0.12290030717849731,
      "rewards/margins": 0.00855380017310381,
      "rewards/rejected": -0.13145411014556885,
      "step": 20
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.882681251368548e-07,
      "logits/chosen": -2.555771827697754,
      "logits/rejected": -2.520658254623413,
      "logps/chosen": -249.33755493164062,
      "logps/pi_response": -142.53103637695312,
      "logps/ref_response": -112.13728332519531,
      "logps/rejected": -248.5088653564453,
      "loss": 0.6779,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.22909781336784363,
      "rewards/margins": 0.043774768710136414,
      "rewards/rejected": -0.2728726267814636,
      "step": 30
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.6604720940421207e-07,
      "logits/chosen": -2.379845142364502,
      "logits/rejected": -2.3809399604797363,
      "logps/chosen": -273.18988037109375,
      "logps/pi_response": -181.37307739257812,
      "logps/ref_response": -109.49589538574219,
      "logps/rejected": -302.9486999511719,
      "loss": 0.6705,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": -0.6705134510993958,
      "rewards/margins": 0.07195498049259186,
      "rewards/rejected": -0.7424684762954712,
      "step": 40
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.3344075855595097e-07,
      "logits/chosen": -2.2819459438323975,
      "logits/rejected": -2.274500608444214,
      "logps/chosen": -335.9717102050781,
      "logps/pi_response": -223.96981811523438,
      "logps/ref_response": -117.62721252441406,
      "logps/rejected": -327.4072570800781,
      "loss": 0.6741,
      "rewards/accuracies": 0.53125,
      "rewards/chosen": -1.0777838230133057,
      "rewards/margins": 0.031330838799476624,
      "rewards/rejected": -1.1091147661209106,
      "step": 50
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.920161866827889e-07,
      "logits/chosen": -2.2090117931365967,
      "logits/rejected": -2.1789093017578125,
      "logps/chosen": -308.4948425292969,
      "logps/pi_response": -217.4617156982422,
      "logps/ref_response": -107.9961929321289,
      "logps/rejected": -323.3466796875,
      "loss": 0.662,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.8825389742851257,
      "rewards/margins": 0.26200079917907715,
      "rewards/rejected": -1.1445398330688477,
      "step": 60
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4376480090239047e-07,
      "logits/chosen": -2.2137115001678467,
      "logits/rejected": -2.212620496749878,
      "logps/chosen": -287.8365478515625,
      "logps/pi_response": -201.22882080078125,
      "logps/ref_response": -113.01625061035156,
      "logps/rejected": -314.63323974609375,
      "loss": 0.6625,
      "rewards/accuracies": 0.625,
      "rewards/chosen": -0.7743679285049438,
      "rewards/margins": 0.14890065789222717,
      "rewards/rejected": -0.9232686161994934,
      "step": 70
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.910060778827554e-07,
      "logits/chosen": -2.2056334018707275,
      "logits/rejected": -2.197784900665283,
      "logps/chosen": -296.95391845703125,
      "logps/pi_response": -206.1053924560547,
      "logps/ref_response": -101.80767822265625,
      "logps/rejected": -338.78887939453125,
      "loss": 0.6727,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -0.9507198333740234,
      "rewards/margins": 0.19975939393043518,
      "rewards/rejected": -1.1504793167114258,
      "step": 80
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.3627616503391812e-07,
      "logits/chosen": -2.2584753036499023,
      "logits/rejected": -2.251711130142212,
      "logps/chosen": -307.7080383300781,
      "logps/pi_response": -200.96620178222656,
      "logps/ref_response": -100.00450897216797,
      "logps/rejected": -278.3091735839844,
      "loss": 0.666,
      "rewards/accuracies": 0.581250011920929,
      "rewards/chosen": -0.8939422369003296,
      "rewards/margins": 0.15028740465641022,
      "rewards/rejected": -1.0442297458648682,
      "step": 90
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.8220596619089573e-07,
      "logits/chosen": -2.1924281120300293,
      "logits/rejected": -2.1820199489593506,
      "logps/chosen": -307.92962646484375,
      "logps/pi_response": -219.1331024169922,
      "logps/ref_response": -112.0215072631836,
      "logps/rejected": -334.48773193359375,
      "loss": 0.6639,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": -0.9324432611465454,
      "rewards/margins": 0.1684987097978592,
      "rewards/rejected": -1.1009418964385986,
      "step": 100
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.3139467229135998e-07,
      "logits/chosen": -2.2497620582580566,
      "logits/rejected": -2.256614923477173,
      "logps/chosen": -307.80267333984375,
      "logps/pi_response": -205.81588745117188,
      "logps/ref_response": -100.42203521728516,
      "logps/rejected": -310.6170349121094,
      "loss": 0.6527,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": -0.9657202959060669,
      "rewards/margins": 0.10005203634500504,
      "rewards/rejected": -1.065772294998169,
      "step": 110
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.628481651367875e-08,
      "logits/chosen": -2.239522933959961,
      "logits/rejected": -2.205223321914673,
      "logps/chosen": -338.15972900390625,
      "logps/pi_response": -245.11599731445312,
      "logps/ref_response": -129.8436279296875,
      "logps/rejected": -348.4112854003906,
      "loss": 0.6479,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": -1.022566556930542,
      "rewards/margins": 0.14734260737895966,
      "rewards/rejected": -1.1699092388153076,
      "step": 120
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.904486005914027e-08,
      "logits/chosen": -2.2750391960144043,
      "logits/rejected": -2.278750419616699,
      "logps/chosen": -296.6529846191406,
      "logps/pi_response": -215.0723114013672,
      "logps/ref_response": -111.54500579833984,
      "logps/rejected": -353.0396423339844,
      "loss": 0.6476,
      "rewards/accuracies": 0.550000011920929,
      "rewards/chosen": -0.9499589204788208,
      "rewards/margins": 0.08809442818164825,
      "rewards/rejected": -1.038053274154663,
      "step": 130
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.1464952759020856e-08,
      "logits/chosen": -2.266298770904541,
      "logits/rejected": -2.2696375846862793,
      "logps/chosen": -326.1199951171875,
      "logps/pi_response": -240.04541015625,
      "logps/ref_response": -119.8094711303711,
      "logps/rejected": -347.0991516113281,
      "loss": 0.6424,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": -0.9663877487182617,
      "rewards/margins": 0.2315327674150467,
      "rewards/rejected": -1.197920560836792,
      "step": 140
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.8708793644441086e-09,
      "logits/chosen": -2.2334301471710205,
      "logits/rejected": -2.22126841545105,
      "logps/chosen": -300.8511047363281,
      "logps/pi_response": -202.55142211914062,
      "logps/ref_response": -101.10533142089844,
      "logps/rejected": -299.7745666503906,
      "loss": 0.6286,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": -0.8617112040519714,
      "rewards/margins": 0.21587765216827393,
      "rewards/rejected": -1.0775889158248901,
      "step": 150
    },
    {
      "epoch": 1.0,
      "step": 159,
      "total_flos": 0.0,
      "train_loss": 0.6631228009109977,
      "train_runtime": 4191.0271,
      "train_samples_per_second": 4.862,
      "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
}