{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.2461059190031152,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06230529595015576,
      "grad_norm": 51.38829302969845,
      "learning_rate": 5e-07,
      "logits/chosen": -2.7407402992248535,
      "logits/rejected": -2.726320266723633,
      "logps/chosen": -261.4151916503906,
      "logps/rejected": -221.7285614013672,
      "loss": 0.692,
      "rewards/accuracies": 0.28125,
      "rewards/chosen": 0.0067472741939127445,
      "rewards/margins": -0.002469523111358285,
      "rewards/rejected": 0.009216798469424248,
      "step": 5
    },
    {
      "epoch": 0.12461059190031153,
      "grad_norm": 54.32144262101448,
      "learning_rate": 1e-06,
      "logits/chosen": -2.679516315460205,
      "logits/rejected": -2.6851108074188232,
      "logps/chosen": -244.4600067138672,
      "logps/rejected": -194.29598999023438,
      "loss": 0.6422,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 0.4551638066768646,
      "rewards/margins": 0.20188376307487488,
      "rewards/rejected": 0.2532801032066345,
      "step": 10
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 44.80793063993453,
      "learning_rate": 9.988343845952696e-07,
      "logits/chosen": -2.4403023719787598,
      "logits/rejected": -2.412696361541748,
      "logps/chosen": -237.53103637695312,
      "logps/rejected": -197.9112091064453,
      "loss": 0.6663,
      "rewards/accuracies": 0.65625,
      "rewards/chosen": 1.520341396331787,
      "rewards/margins": 0.5765382051467896,
      "rewards/rejected": 0.9438031315803528,
      "step": 15
    },
    {
      "epoch": 0.24922118380062305,
      "grad_norm": 49.11699733468642,
      "learning_rate": 9.953429730181652e-07,
      "logits/chosen": -2.2657809257507324,
      "logits/rejected": -2.2547507286071777,
      "logps/chosen": -225.1350860595703,
      "logps/rejected": -232.2615203857422,
      "loss": 0.669,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 1.3284717798233032,
      "rewards/margins": 0.6173279285430908,
      "rewards/rejected": 0.7111440896987915,
      "step": 20
    },
    {
      "epoch": 0.3115264797507788,
      "grad_norm": 43.743825005172326,
      "learning_rate": 9.895420438411615e-07,
      "logits/chosen": -2.223740577697754,
      "logits/rejected": -2.239903688430786,
      "logps/chosen": -245.8108673095703,
      "logps/rejected": -246.0238037109375,
      "loss": 0.6389,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 1.2945916652679443,
      "rewards/margins": 0.8869680166244507,
      "rewards/rejected": 0.407623827457428,
      "step": 25
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 38.88842184904586,
      "learning_rate": 9.814586436738997e-07,
      "logits/chosen": -2.302748680114746,
      "logits/rejected": -2.31158185005188,
      "logps/chosen": -246.70578002929688,
      "logps/rejected": -217.460205078125,
      "loss": 0.6192,
      "rewards/accuracies": 0.7562500238418579,
      "rewards/chosen": 1.3854382038116455,
      "rewards/margins": 1.0373018980026245,
      "rewards/rejected": 0.34813636541366577,
      "step": 30
    },
    {
      "epoch": 0.43613707165109034,
      "grad_norm": 51.72443769242076,
      "learning_rate": 9.711304610594102e-07,
      "logits/chosen": -2.3440041542053223,
      "logits/rejected": -2.31417179107666,
      "logps/chosen": -224.3390350341797,
      "logps/rejected": -194.77621459960938,
      "loss": 0.5888,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.972150981426239,
      "rewards/margins": 0.8106788396835327,
      "rewards/rejected": 0.1614721715450287,
      "step": 35
    },
    {
      "epoch": 0.4984423676012461,
      "grad_norm": 41.52935591539919,
      "learning_rate": 9.586056507527264e-07,
      "logits/chosen": -2.383333683013916,
      "logits/rejected": -2.34763765335083,
      "logps/chosen": -246.3532257080078,
      "logps/rejected": -218.7850341796875,
      "loss": 0.6391,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.0201754570007324,
      "rewards/margins": 1.0443416833877563,
      "rewards/rejected": -0.02416619285941124,
      "step": 40
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 37.760052208675,
      "learning_rate": 9.439426092011875e-07,
      "logits/chosen": -2.4347915649414062,
      "logits/rejected": -2.3998470306396484,
      "logps/chosen": -232.26266479492188,
      "logps/rejected": -218.72866821289062,
      "loss": 0.5939,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 1.1044284105300903,
      "rewards/margins": 0.9438824653625488,
      "rewards/rejected": 0.16054585576057434,
      "step": 45
    },
    {
      "epoch": 0.6230529595015576,
      "grad_norm": 37.571273896281745,
      "learning_rate": 9.272097022732443e-07,
      "logits/chosen": -2.45988130569458,
      "logits/rejected": -2.4507269859313965,
      "logps/chosen": -249.6595916748047,
      "logps/rejected": -218.61581420898438,
      "loss": 0.5547,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 1.192905306816101,
      "rewards/margins": 1.339156985282898,
      "rewards/rejected": -0.14625166356563568,
      "step": 50
    },
    {
      "epoch": 0.6230529595015576,
      "eval_logits/chosen": -2.489163637161255,
      "eval_logits/rejected": -2.487579107284546,
      "eval_logps/chosen": -237.33294677734375,
      "eval_logps/rejected": -211.63194274902344,
      "eval_loss": 0.5780755281448364,
      "eval_rewards/accuracies": 0.7222222089767456,
      "eval_rewards/chosen": 0.9537798762321472,
      "eval_rewards/margins": 1.0286411046981812,
      "eval_rewards/rejected": -0.07486122101545334,
      "eval_runtime": 152.1375,
      "eval_samples_per_second": 14.986,
      "eval_steps_per_second": 0.237,
      "step": 50
    },
    {
      "epoch": 0.6853582554517134,
      "grad_norm": 54.80320093683614,
      "learning_rate": 9.084849465052209e-07,
      "logits/chosen": -2.485079765319824,
      "logits/rejected": -2.4753665924072266,
      "logps/chosen": -235.5215301513672,
      "logps/rejected": -220.1900177001953,
      "loss": 0.5909,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 0.7905256152153015,
      "rewards/margins": 1.0766956806182861,
      "rewards/rejected": -0.286170095205307,
      "step": 55
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 39.23815457631208,
      "learning_rate": 8.878556453522099e-07,
      "logits/chosen": -2.4538919925689697,
      "logits/rejected": -2.454207181930542,
      "logps/chosen": -251.35598754882812,
      "logps/rejected": -231.8341522216797,
      "loss": 0.5733,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.9427865743637085,
      "rewards/margins": 1.1441007852554321,
      "rewards/rejected": -0.2013140171766281,
      "step": 60
    },
    {
      "epoch": 0.8099688473520249,
      "grad_norm": 47.13645309746499,
      "learning_rate": 8.654179821390621e-07,
      "logits/chosen": -2.4390835762023926,
      "logits/rejected": -2.449341297149658,
      "logps/chosen": -262.5836486816406,
      "logps/rejected": -209.0801239013672,
      "loss": 0.5441,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 1.3532841205596924,
      "rewards/margins": 1.4332042932510376,
      "rewards/rejected": -0.07991998642683029,
      "step": 65
    },
    {
      "epoch": 0.8722741433021807,
      "grad_norm": 40.00645275405077,
      "learning_rate": 8.41276571609327e-07,
      "logits/chosen": -2.4375357627868652,
      "logits/rejected": -2.4228920936584473,
      "logps/chosen": -246.0580291748047,
      "logps/rejected": -219.20260620117188,
      "loss": 0.5816,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 1.0882443189620972,
      "rewards/margins": 1.3165982961654663,
      "rewards/rejected": -0.22835393249988556,
      "step": 70
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 34.63673701954024,
      "learning_rate": 8.155439721630264e-07,
      "logits/chosen": -2.4634451866149902,
      "logits/rejected": -2.4578163623809814,
      "logps/chosen": -227.31924438476562,
      "logps/rejected": -217.36508178710938,
      "loss": 0.5334,
      "rewards/accuracies": 0.706250011920929,
      "rewards/chosen": 0.8839371800422668,
      "rewards/margins": 1.3334705829620361,
      "rewards/rejected": -0.4495334029197693,
      "step": 75
    },
    {
      "epoch": 0.9968847352024922,
      "grad_norm": 36.51497796461376,
      "learning_rate": 7.883401610574336e-07,
      "logits/chosen": -2.4859111309051514,
      "logits/rejected": -2.465956211090088,
      "logps/chosen": -242.3914794921875,
      "logps/rejected": -210.991943359375,
      "loss": 0.5281,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 0.5210741758346558,
      "rewards/margins": 1.2326009273529053,
      "rewards/rejected": -0.7115266919136047,
      "step": 80
    },
    {
      "epoch": 1.0591900311526479,
      "grad_norm": 20.825789256058748,
      "learning_rate": 7.597919750177168e-07,
      "logits/chosen": -2.437802791595459,
      "logits/rejected": -2.436861515045166,
      "logps/chosen": -260.0554504394531,
      "logps/rejected": -234.86740112304688,
      "loss": 0.2507,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 0.9926292300224304,
      "rewards/margins": 2.598191261291504,
      "rewards/rejected": -1.6055622100830078,
      "step": 85
    },
    {
      "epoch": 1.1214953271028036,
      "grad_norm": 17.02799613983615,
      "learning_rate": 7.30032518865576e-07,
      "logits/chosen": -2.417109489440918,
      "logits/rejected": -2.404921770095825,
      "logps/chosen": -233.2266387939453,
      "logps/rejected": -226.57186889648438,
      "loss": 0.2109,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 1.7625309228897095,
      "rewards/margins": 3.178365707397461,
      "rewards/rejected": -1.4158347845077515,
      "step": 90
    },
    {
      "epoch": 1.1838006230529594,
      "grad_norm": 20.45306812774076,
      "learning_rate": 6.992005449231207e-07,
      "logits/chosen": -2.3889553546905518,
      "logits/rejected": -2.38620924949646,
      "logps/chosen": -251.99356079101562,
      "logps/rejected": -243.1147918701172,
      "loss": 0.2198,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.9874979257583618,
      "rewards/margins": 3.548366069793701,
      "rewards/rejected": -1.5608683824539185,
      "step": 95
    },
    {
      "epoch": 1.2461059190031152,
      "grad_norm": 16.12410058725181,
      "learning_rate": 6.67439806085493e-07,
      "logits/chosen": -2.338783025741577,
      "logits/rejected": -2.325284004211426,
      "logps/chosen": -240.7113800048828,
      "logps/rejected": -222.7356414794922,
      "loss": 0.2103,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.1200156211853027,
      "rewards/margins": 3.5902743339538574,
      "rewards/rejected": -1.4702587127685547,
      "step": 100
    },
    {
      "epoch": 1.2461059190031152,
      "eval_logits/chosen": -2.3138608932495117,
      "eval_logits/rejected": -2.2990024089813232,
      "eval_logps/chosen": -233.84922790527344,
      "eval_logps/rejected": -214.24310302734375,
      "eval_loss": 0.6053553819656372,
      "eval_rewards/accuracies": 0.7777777910232544,
      "eval_rewards/chosen": 1.3021522760391235,
      "eval_rewards/margins": 1.638129472732544,
      "eval_rewards/rejected": -0.33597710728645325,
      "eval_runtime": 151.0351,
      "eval_samples_per_second": 15.096,
      "eval_steps_per_second": 0.238,
      "step": 100
    }
  ],
  "logging_steps": 5,
  "max_steps": 240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1178822762299392.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}