{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.968,
  "eval_steps": 100,
  "global_step": 248,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 2.0000000000000002e-07,
      "logits/chosen": 0.10287265479564667,
      "logits/rejected": 0.011988319456577301,
      "logps/chosen": -192.40402221679688,
      "logps/rejected": -100.80304718017578,
      "loss": 0.001,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.0000000000000003e-06,
      "logits/chosen": 0.16615836322307587,
      "logits/rejected": 0.08426308631896973,
      "logps/chosen": -182.21653747558594,
      "logps/rejected": -161.99514770507812,
      "loss": 0.001,
      "rewards/accuracies": 0.375,
      "rewards/chosen": -0.0011453586630523205,
      "rewards/margins": 0.0010826066136360168,
      "rewards/rejected": -0.0022279650438576937,
      "step": 10
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.000000000000001e-06,
      "logits/chosen": 0.22168031334877014,
      "logits/rejected": 0.2861151099205017,
      "logps/chosen": -176.96937561035156,
      "logps/rejected": -135.0897674560547,
      "loss": 0.0012,
      "rewards/accuracies": 0.36250001192092896,
      "rewards/chosen": -0.0021877787075936794,
      "rewards/margins": -0.000618638179730624,
      "rewards/rejected": -0.0015691407024860382,
      "step": 20
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.993800445762451e-06,
      "logits/chosen": 0.2504645884037018,
      "logits/rejected": 0.2488332986831665,
      "logps/chosen": -169.92291259765625,
      "logps/rejected": -150.56483459472656,
      "loss": 0.0012,
      "rewards/accuracies": 0.39375001192092896,
      "rewards/chosen": -0.0013141350354999304,
      "rewards/margins": -0.00041974737541750073,
      "rewards/rejected": -0.0008943876018747687,
      "step": 30
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.944388344834205e-06,
      "logits/chosen": 0.20393869280815125,
      "logits/rejected": 0.26991668343544006,
      "logps/chosen": -172.50244140625,
      "logps/rejected": -155.56478881835938,
      "loss": 0.0012,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": 0.0012658500345423818,
      "rewards/margins": -0.0002958738768938929,
      "rewards/rejected": 0.0015617238823324442,
      "step": 40
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.8465431931347904e-06,
      "logits/chosen": 0.23398549854755402,
      "logits/rejected": 0.17997679114341736,
      "logps/chosen": -198.8681182861328,
      "logps/rejected": -161.630615234375,
      "loss": 0.0011,
      "rewards/accuracies": 0.4375,
      "rewards/chosen": -0.0012974137207493186,
      "rewards/margins": 0.0008641968597657979,
      "rewards/rejected": -0.0021616104058921337,
      "step": 50
    },
    {
      "epoch": 0.96,
      "learning_rate": 4.702203692102539e-06,
      "logits/chosen": 0.24316315352916718,
      "logits/rejected": 0.29181909561157227,
      "logps/chosen": -187.0422821044922,
      "logps/rejected": -162.91946411132812,
      "loss": 0.0012,
      "rewards/accuracies": 0.3687500059604645,
      "rewards/chosen": 3.171586286043748e-05,
      "rewards/margins": -0.0004161152464803308,
      "rewards/rejected": 0.00044783117482438684,
      "step": 60
    },
    {
      "epoch": 1.12,
      "learning_rate": 4.514229781074239e-06,
      "logits/chosen": 0.23100349307060242,
      "logits/rejected": 0.16395993530750275,
      "logps/chosen": -195.23422241210938,
      "logps/rejected": -167.4615020751953,
      "loss": 0.001,
      "rewards/accuracies": 0.4000000059604645,
      "rewards/chosen": 4.4724576582666487e-05,
      "rewards/margins": 0.0011552829528227448,
      "rewards/rejected": -0.0011105581652373075,
      "step": 70
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.286345970517195e-06,
      "logits/chosen": 0.1470017284154892,
      "logits/rejected": 0.1968422830104828,
      "logps/chosen": -202.30715942382812,
      "logps/rejected": -162.376953125,
      "loss": 0.0011,
      "rewards/accuracies": 0.375,
      "rewards/chosen": 0.0012948224321007729,
      "rewards/margins": 0.0006303158006630838,
      "rewards/rejected": 0.00066450668964535,
      "step": 80
    },
    {
      "epoch": 1.44,
      "learning_rate": 4.023067544670082e-06,
      "logits/chosen": 0.2470308095216751,
      "logits/rejected": 0.2212987244129181,
      "logps/chosen": -173.95779418945312,
      "logps/rejected": -155.6954803466797,
      "loss": 0.0012,
      "rewards/accuracies": 0.3812499940395355,
      "rewards/chosen": -0.0012559869792312384,
      "rewards/margins": -0.000755336950533092,
      "rewards/rejected": -0.0005006499122828245,
      "step": 90
    },
    {
      "epoch": 1.6,
      "learning_rate": 3.7296110958116845e-06,
      "logits/chosen": 0.19465205073356628,
      "logits/rejected": 0.17263346910476685,
      "logps/chosen": -177.3841552734375,
      "logps/rejected": -143.33177185058594,
      "loss": 0.0011,
      "rewards/accuracies": 0.3812499940395355,
      "rewards/chosen": -0.0003903825709130615,
      "rewards/margins": 8.392091694986448e-05,
      "rewards/rejected": -0.0004743034369312227,
      "step": 100
    },
    {
      "epoch": 1.6,
      "eval_logits/chosen": -0.007278905715793371,
      "eval_logits/rejected": 0.09087153524160385,
      "eval_logps/chosen": -306.3416748046875,
      "eval_logps/rejected": -278.65435791015625,
      "eval_loss": 0.0015715558547526598,
      "eval_rewards/accuracies": 0.4959999918937683,
      "eval_rewards/chosen": -0.0007231617928482592,
      "eval_rewards/margins": 0.00012325971329119056,
      "eval_rewards/rejected": -0.0008464214624837041,
      "eval_runtime": 933.3559,
      "eval_samples_per_second": 2.143,
      "eval_steps_per_second": 0.536,
      "step": 100
    },
    {
      "epoch": 1.76,
      "learning_rate": 3.4117911628292944e-06,
      "logits/chosen": 0.29057276248931885,
      "logits/rejected": 0.19747909903526306,
      "logps/chosen": -173.1153564453125,
      "logps/rejected": -151.88233947753906,
      "loss": 0.0011,
      "rewards/accuracies": 0.3187499940395355,
      "rewards/chosen": -0.00022493198048323393,
      "rewards/margins": -0.000448818871518597,
      "rewards/rejected": 0.00022388689103536308,
      "step": 110
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.075905022087675e-06,
      "logits/chosen": 0.1814224272966385,
      "logits/rejected": 0.2010256052017212,
      "logps/chosen": -169.48214721679688,
      "logps/rejected": -147.04067993164062,
      "loss": 0.0011,
      "rewards/accuracies": 0.41874998807907104,
      "rewards/chosen": 0.0006150969420559704,
      "rewards/margins": 0.0005108726327307522,
      "rewards/rejected": 0.00010422446939628571,
      "step": 120
    },
    {
      "epoch": 2.08,
      "learning_rate": 2.728607913349464e-06,
      "logits/chosen": 0.24528518319129944,
      "logits/rejected": 0.2507859468460083,
      "logps/chosen": -177.7463836669922,
      "logps/rejected": -162.4386444091797,
      "loss": 0.0011,
      "rewards/accuracies": 0.38749998807907104,
      "rewards/chosen": -0.0017223177710548043,
      "rewards/margins": 0.0007541574887000024,
      "rewards/rejected": -0.0024764754343777895,
      "step": 130
    },
    {
      "epoch": 2.24,
      "learning_rate": 2.376781173017589e-06,
      "logits/chosen": 0.1746383160352707,
      "logits/rejected": 0.15819591283798218,
      "logps/chosen": -175.5349884033203,
      "logps/rejected": -150.61404418945312,
      "loss": 0.0012,
      "rewards/accuracies": 0.3812499940395355,
      "rewards/chosen": -0.0010503135854378343,
      "rewards/margins": -0.0006477275164797902,
      "rewards/rejected": -0.0004025862435810268,
      "step": 140
    },
    {
      "epoch": 2.4,
      "learning_rate": 2.0273958875043877e-06,
      "logits/chosen": 0.16554135084152222,
      "logits/rejected": 0.20675285160541534,
      "logps/chosen": -198.78173828125,
      "logps/rejected": -161.86558532714844,
      "loss": 0.0011,
      "rewards/accuracies": 0.3812499940395355,
      "rewards/chosen": -0.001666503376327455,
      "rewards/margins": 0.0006793343345634639,
      "rewards/rejected": -0.0023458378855139017,
      "step": 150
    },
    {
      "epoch": 2.56,
      "learning_rate": 1.6873747682962393e-06,
      "logits/chosen": 0.24000254273414612,
      "logits/rejected": 0.27995795011520386,
      "logps/chosen": -190.5437469482422,
      "logps/rejected": -164.19549560546875,
      "loss": 0.001,
      "rewards/accuracies": 0.40625,
      "rewards/chosen": 0.002512627048417926,
      "rewards/margins": 0.00116717757191509,
      "rewards/rejected": 0.001345449360087514,
      "step": 160
    },
    {
      "epoch": 2.72,
      "learning_rate": 1.363454985517803e-06,
      "logits/chosen": 0.2393382042646408,
      "logits/rejected": 0.26565679907798767,
      "logps/chosen": -164.2212677001953,
      "logps/rejected": -140.7093963623047,
      "loss": 0.0011,
      "rewards/accuracies": 0.4312500059604645,
      "rewards/chosen": 0.0010435387957841158,
      "rewards/margins": 0.0007807637448422611,
      "rewards/rejected": 0.0002627749345265329,
      "step": 170
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.062054677808238e-06,
      "logits/chosen": 0.20416012406349182,
      "logits/rejected": 0.1511019915342331,
      "logps/chosen": -183.35943603515625,
      "logps/rejected": -157.43435668945312,
      "loss": 0.001,
      "rewards/accuracies": 0.4124999940395355,
      "rewards/chosen": -7.58793466957286e-05,
      "rewards/margins": 0.0020282561890780926,
      "rewards/rejected": -0.002104135463014245,
      "step": 180
    },
    {
      "epoch": 3.04,
      "learning_rate": 7.891457834794711e-07,
      "logits/chosen": 0.19763195514678955,
      "logits/rejected": 0.2482624500989914,
      "logps/chosen": -169.32882690429688,
      "logps/rejected": -142.48541259765625,
      "loss": 0.001,
      "rewards/accuracies": 0.44999998807907104,
      "rewards/chosen": 0.0012013750383630395,
      "rewards/margins": 0.0014427897986024618,
      "rewards/rejected": -0.00024141438188962638,
      "step": 190
    },
    {
      "epoch": 3.2,
      "learning_rate": 5.501357126768117e-07,
      "logits/chosen": 0.1878470778465271,
      "logits/rejected": 0.18784348666667938,
      "logps/chosen": -188.822998046875,
      "logps/rejected": -153.36880493164062,
      "loss": 0.0011,
      "rewards/accuracies": 0.45625001192092896,
      "rewards/chosen": 0.00035641956492327154,
      "rewards/margins": 0.0015287164133042097,
      "rewards/rejected": -0.001172296586446464,
      "step": 200
    },
    {
      "epoch": 3.2,
      "eval_logits/chosen": -0.004656919743865728,
      "eval_logits/rejected": 0.09210014343261719,
      "eval_logps/chosen": -306.283447265625,
      "eval_logps/rejected": -278.6177062988281,
      "eval_loss": 0.0015858013648539782,
      "eval_rewards/accuracies": 0.4925000071525574,
      "eval_rewards/chosen": -0.00014072553312871605,
      "eval_rewards/margins": 0.00033963273745030165,
      "eval_rewards/rejected": -0.0004803583142347634,
      "eval_runtime": 1595.4244,
      "eval_samples_per_second": 1.254,
      "eval_steps_per_second": 0.313,
      "step": 200
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.4976020508682345e-07,
      "logits/chosen": 0.17026616632938385,
      "logits/rejected": 0.2095976322889328,
      "logps/chosen": -177.265625,
      "logps/rejected": -149.7288055419922,
      "loss": 0.001,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.00014497071970254183,
      "rewards/margins": 0.0017732717096805573,
      "rewards/rejected": -0.0016283008735626936,
      "step": 210
    },
    {
      "epoch": 3.52,
      "learning_rate": 1.9198949610721273e-07,
      "logits/chosen": 0.19490325450897217,
      "logits/rejected": 0.2224169224500656,
      "logps/chosen": -175.52345275878906,
      "logps/rejected": -146.68728637695312,
      "loss": 0.001,
      "rewards/accuracies": 0.45625001192092896,
      "rewards/chosen": 0.002361242426559329,
      "rewards/margins": 0.0017604168970137835,
      "rewards/rejected": 0.0006008257623761892,
      "step": 220
    },
    {
      "epoch": 3.68,
      "learning_rate": 7.994965069994143e-08,
      "logits/chosen": 0.21312932670116425,
      "logits/rejected": 0.18425947427749634,
      "logps/chosen": -187.47511291503906,
      "logps/rejected": -163.0996856689453,
      "loss": 0.001,
      "rewards/accuracies": 0.46875,
      "rewards/chosen": 0.0017060479149222374,
      "rewards/margins": 0.001837022602558136,
      "rewards/rejected": -0.00013097473129164428,
      "step": 230
    },
    {
      "epoch": 3.84,
      "learning_rate": 1.5860623616664183e-08,
      "logits/chosen": 0.1936773955821991,
      "logits/rejected": 0.12241797149181366,
      "logps/chosen": -176.46658325195312,
      "logps/rejected": -148.58331298828125,
      "loss": 0.0011,
      "rewards/accuracies": 0.39375001192092896,
      "rewards/chosen": -0.0003967673401348293,
      "rewards/margins": 0.00025122734950855374,
      "rewards/rejected": -0.0006479948060587049,
      "step": 240
    },
    {
      "epoch": 3.97,
      "step": 248,
      "total_flos": 0.0,
      "train_loss": 0.0010954971686230911,
      "train_runtime": 8408.6931,
      "train_samples_per_second": 0.476,
      "train_steps_per_second": 0.029
    }
  ],
  "logging_steps": 10,
  "max_steps": 248,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}