{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 385,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.282051282051282e-07,
      "logits/chosen": -1.7278180122375488,
      "logits/rejected": -1.7377450466156006,
      "logps/chosen": -29.553977966308594,
      "logps/rejected": -42.813133239746094,
      "loss": 0.6931,
      "rewards/accuracies": 0.0,
      "rewards/chosen": 0.0,
      "rewards/margins": 0.0,
      "rewards/rejected": 0.0,
      "step": 1
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.282051282051282e-06,
      "logits/chosen": -1.866413950920105,
      "logits/rejected": -1.8707411289215088,
      "logps/chosen": -36.98916244506836,
      "logps/rejected": -33.67436981201172,
      "loss": 0.6701,
      "rewards/accuracies": 0.5416666865348816,
      "rewards/chosen": 0.01569323241710663,
      "rewards/margins": 0.05555717274546623,
      "rewards/rejected": -0.039863936603069305,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.564102564102564e-06,
      "logits/chosen": -1.9979650974273682,
      "logits/rejected": -2.0006086826324463,
      "logps/chosen": -29.624820709228516,
      "logps/rejected": -29.0762939453125,
      "loss": 0.6837,
      "rewards/accuracies": 0.5874999761581421,
      "rewards/chosen": 0.01563635841012001,
      "rewards/margins": 0.027204299345612526,
      "rewards/rejected": -0.01156794372946024,
      "step": 20
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.846153846153847e-06,
      "logits/chosen": -1.921021819114685,
      "logits/rejected": -1.9183374643325806,
      "logps/chosen": -31.40532875061035,
      "logps/rejected": -33.23241424560547,
      "loss": 0.6877,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.00968973059207201,
      "rewards/margins": 0.022251319140195847,
      "rewards/rejected": -0.012561586685478687,
      "step": 30
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.999896948438434e-06,
      "logits/chosen": -2.0176353454589844,
      "logits/rejected": -2.008906364440918,
      "logps/chosen": -32.574256896972656,
      "logps/rejected": -32.53368377685547,
      "loss": 0.6874,
      "rewards/accuracies": 0.512499988079071,
      "rewards/chosen": 0.0022967704571783543,
      "rewards/margins": 0.02120940014719963,
      "rewards/rejected": -0.018912632018327713,
      "step": 40
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.987541037542187e-06,
      "logits/chosen": -1.8619186878204346,
      "logits/rejected": -1.85114324092865,
      "logps/chosen": -33.55537414550781,
      "logps/rejected": -35.45675277709961,
      "loss": 0.6957,
      "rewards/accuracies": 0.48750001192092896,
      "rewards/chosen": 0.001892436295747757,
      "rewards/margins": 0.005858602002263069,
      "rewards/rejected": -0.003966164775192738,
      "step": 50
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.954691471941119e-06,
      "logits/chosen": -1.9400945901870728,
      "logits/rejected": -1.9420464038848877,
      "logps/chosen": -32.56509780883789,
      "logps/rejected": -33.2406120300293,
      "loss": 0.6632,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.031578924506902695,
      "rewards/margins": 0.09388783574104309,
      "rewards/rejected": -0.062308914959430695,
      "step": 60
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.901618883413549e-06,
      "logits/chosen": -2.0712790489196777,
      "logits/rejected": -2.0762436389923096,
      "logps/chosen": -33.981910705566406,
      "logps/rejected": -36.62363815307617,
      "loss": 0.6833,
      "rewards/accuracies": 0.5249999761581421,
      "rewards/chosen": 0.005918038543313742,
      "rewards/margins": 0.05520814657211304,
      "rewards/rejected": -0.04929010197520256,
      "step": 70
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.828760511501322e-06,
      "logits/chosen": -1.9327905178070068,
      "logits/rejected": -1.935909628868103,
      "logps/chosen": -34.32685470581055,
      "logps/rejected": -34.65606689453125,
      "loss": 0.639,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.09085920453071594,
      "rewards/margins": 0.14815348386764526,
      "rewards/rejected": -0.057294271886348724,
      "step": 80
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.7367166013034295e-06,
      "logits/chosen": -1.9414918422698975,
      "logits/rejected": -1.946007490158081,
      "logps/chosen": -32.406803131103516,
      "logps/rejected": -32.36021041870117,
      "loss": 0.6792,
      "rewards/accuracies": 0.6000000238418579,
      "rewards/chosen": 0.054556868970394135,
      "rewards/margins": 0.05573350936174393,
      "rewards/rejected": -0.0011766403913497925,
      "step": 90
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.626245458345211e-06,
      "logits/chosen": -2.039034128189087,
      "logits/rejected": -2.0370402336120605,
      "logps/chosen": -32.172786712646484,
      "logps/rejected": -31.333194732666016,
      "loss": 0.6464,
      "rewards/accuracies": 0.625,
      "rewards/chosen": 0.06124376505613327,
      "rewards/margins": 0.12152798473834991,
      "rewards/rejected": -0.06028420478105545,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_logits/chosen": -2.2339773178100586,
      "eval_logits/rejected": -2.229137420654297,
      "eval_logps/chosen": -34.04054641723633,
      "eval_logps/rejected": -37.549957275390625,
      "eval_loss": 0.6902773976325989,
      "eval_rewards/accuracies": 0.5685215592384338,
      "eval_rewards/chosen": -0.005393954925239086,
      "eval_rewards/margins": 0.024608083069324493,
      "eval_rewards/rejected": -0.030002037063241005,
      "eval_runtime": 146.034,
      "eval_samples_per_second": 2.349,
      "eval_steps_per_second": 0.294,
      "step": 100
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.498257201263691e-06,
      "logits/chosen": -1.994192123413086,
      "logits/rejected": -1.9918158054351807,
      "logps/chosen": -33.142940521240234,
      "logps/rejected": -34.01188278198242,
      "loss": 0.6911,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.09078876674175262,
      "rewards/margins": 0.07505009323358536,
      "rewards/rejected": 0.015738680958747864,
      "step": 110
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.353806263777678e-06,
      "logits/chosen": -2.0053954124450684,
      "logits/rejected": -1.997046709060669,
      "logps/chosen": -32.33894348144531,
      "logps/rejected": -32.1308708190918,
      "loss": 0.6746,
      "rewards/accuracies": 0.5625,
      "rewards/chosen": 0.09536493569612503,
      "rewards/margins": 0.06779730319976807,
      "rewards/rejected": 0.027567636221647263,
      "step": 120
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.1940827077152755e-06,
      "logits/chosen": -2.0336387157440186,
      "logits/rejected": -2.025650978088379,
      "logps/chosen": -30.345691680908203,
      "logps/rejected": -32.078697204589844,
      "loss": 0.6527,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.11702337116003036,
      "rewards/margins": 0.14014457166194916,
      "rewards/rejected": -0.023121213540434837,
      "step": 130
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.0204024186666215e-06,
      "logits/chosen": -1.9642337560653687,
      "logits/rejected": -1.9744552373886108,
      "logps/chosen": -31.243911743164062,
      "logps/rejected": -32.590267181396484,
      "loss": 0.6171,
      "rewards/accuracies": 0.6625000238418579,
      "rewards/chosen": 0.1581769436597824,
      "rewards/margins": 0.20802685618400574,
      "rewards/rejected": -0.04984992742538452,
      "step": 140
    },
    {
      "epoch": 0.39,
      "learning_rate": 3.834196265035119e-06,
      "logits/chosen": -1.876604437828064,
      "logits/rejected": -1.8777605295181274,
      "logps/chosen": -33.938690185546875,
      "logps/rejected": -34.807891845703125,
      "loss": 0.6043,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.22860188782215118,
      "rewards/margins": 0.2741745412349701,
      "rewards/rejected": -0.0455726757645607,
      "step": 150
    },
    {
      "epoch": 0.42,
      "learning_rate": 3.636998309800573e-06,
      "logits/chosen": -1.9282041788101196,
      "logits/rejected": -1.9247684478759766,
      "logps/chosen": -36.02125930786133,
      "logps/rejected": -32.71831130981445,
      "loss": 0.6454,
      "rewards/accuracies": 0.612500011920929,
      "rewards/chosen": 0.13537634909152985,
      "rewards/margins": 0.13137592375278473,
      "rewards/rejected": 0.004000450484454632,
      "step": 160
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.4304331721118078e-06,
      "logits/chosen": -2.029125928878784,
      "logits/rejected": -2.0217747688293457,
      "logps/chosen": -33.49839401245117,
      "logps/rejected": -31.400177001953125,
      "loss": 0.5828,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.26951926946640015,
      "rewards/margins": 0.3130132555961609,
      "rewards/rejected": -0.04349397122859955,
      "step": 170
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.2162026428305436e-06,
      "logits/chosen": -2.0355944633483887,
      "logits/rejected": -2.040832042694092,
      "logps/chosen": -32.235923767089844,
      "logps/rejected": -32.460418701171875,
      "loss": 0.5943,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.2779761850833893,
      "rewards/margins": 0.2557251751422882,
      "rewards/rejected": 0.02225096896290779,
      "step": 180
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.996071664294641e-06,
      "logits/chosen": -2.0362112522125244,
      "logits/rejected": -2.0334599018096924,
      "logps/chosen": -31.269250869750977,
      "logps/rejected": -31.325435638427734,
      "loss": 0.6245,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.19773444533348083,
      "rewards/margins": 0.20423230528831482,
      "rewards/rejected": -0.0064978525042533875,
      "step": 190
    },
    {
      "epoch": 0.52,
      "learning_rate": 2.7718537898066833e-06,
      "logits/chosen": -1.9060389995574951,
      "logits/rejected": -1.9106788635253906,
      "logps/chosen": -31.306299209594727,
      "logps/rejected": -32.81407165527344,
      "loss": 0.5931,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.2699825167655945,
      "rewards/margins": 0.2908058166503906,
      "rewards/rejected": -0.02082330361008644,
      "step": 200
    },
    {
      "epoch": 0.52,
      "eval_logits/chosen": -2.231553792953491,
      "eval_logits/rejected": -2.2267112731933594,
      "eval_logps/chosen": -34.07304763793945,
      "eval_logps/rejected": -37.57693862915039,
      "eval_loss": 0.6979728937149048,
      "eval_rewards/accuracies": 0.5157807469367981,
      "eval_rewards/chosen": -0.03464451804757118,
      "eval_rewards/margins": 0.019641490653157234,
      "eval_rewards/rejected": -0.054286014288663864,
      "eval_runtime": 145.8095,
      "eval_samples_per_second": 2.352,
      "eval_steps_per_second": 0.295,
      "step": 200
    },
    {
      "epoch": 0.55,
      "learning_rate": 2.5453962426402006e-06,
      "logits/chosen": -2.018519163131714,
      "logits/rejected": -2.0291810035705566,
      "logps/chosen": -31.742992401123047,
      "logps/rejected": -33.946937561035156,
      "loss": 0.5902,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.2054794579744339,
      "rewards/margins": 0.2812942862510681,
      "rewards/rejected": -0.07581482082605362,
      "step": 210
    },
    {
      "epoch": 0.57,
      "learning_rate": 2.3185646976551794e-06,
      "logits/chosen": -1.911586046218872,
      "logits/rejected": -1.9263393878936768,
      "logps/chosen": -29.84616470336914,
      "logps/rejected": -31.615009307861328,
      "loss": 0.5879,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 0.23883743584156036,
      "rewards/margins": 0.2899848222732544,
      "rewards/rejected": -0.051147449761629105,
      "step": 220
    },
    {
      "epoch": 0.6,
      "learning_rate": 2.0932279108998323e-06,
      "logits/chosen": -1.9677941799163818,
      "logits/rejected": -1.9717823266983032,
      "logps/chosen": -33.100074768066406,
      "logps/rejected": -31.62213134765625,
      "loss": 0.5748,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.28565075993537903,
      "rewards/margins": 0.3511958718299866,
      "rewards/rejected": -0.06554517149925232,
      "step": 230
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8712423238279358e-06,
      "logits/chosen": -1.9661725759506226,
      "logits/rejected": -1.944300651550293,
      "logps/chosen": -33.841453552246094,
      "logps/rejected": -35.11375045776367,
      "loss": 0.5473,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.2810631990432739,
      "rewards/margins": 0.4277234673500061,
      "rewards/rejected": -0.14666026830673218,
      "step": 240
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.6544367689701824e-06,
      "logits/chosen": -2.007416009902954,
      "logits/rejected": -2.0040948390960693,
      "logps/chosen": -32.70330810546875,
      "logps/rejected": -36.29412841796875,
      "loss": 0.5992,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.1995842456817627,
      "rewards/margins": 0.2618715763092041,
      "rewards/rejected": -0.06228730082511902,
      "step": 250
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.4445974030621963e-06,
      "logits/chosen": -1.8749721050262451,
      "logits/rejected": -1.8725513219833374,
      "logps/chosen": -34.00068664550781,
      "logps/rejected": -35.53888702392578,
      "loss": 0.6254,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 0.16894161701202393,
      "rewards/margins": 0.1997825801372528,
      "rewards/rejected": -0.030840963125228882,
      "step": 260
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.243452991757889e-06,
      "logits/chosen": -1.8600317239761353,
      "logits/rejected": -1.8576066493988037,
      "logps/chosen": -34.1875,
      "logps/rejected": -31.8159122467041,
      "loss": 0.616,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.1890900433063507,
      "rewards/margins": 0.22921428084373474,
      "rewards/rejected": -0.04012420028448105,
      "step": 270
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.0526606671603523e-06,
      "logits/chosen": -1.9631398916244507,
      "logits/rejected": -1.9526073932647705,
      "logps/chosen": -35.023719787597656,
      "logps/rejected": -31.869693756103516,
      "loss": 0.5782,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.29963088035583496,
      "rewards/margins": 0.32546472549438477,
      "rewards/rejected": -0.025833839550614357,
      "step": 280
    },
    {
      "epoch": 0.75,
      "learning_rate": 8.737922755071455e-07,
      "logits/chosen": -2.0582926273345947,
      "logits/rejected": -2.0433640480041504,
      "logps/chosen": -30.733753204345703,
      "logps/rejected": -32.67460632324219,
      "loss": 0.6392,
      "rewards/accuracies": 0.637499988079071,
      "rewards/chosen": 0.17133468389511108,
      "rewards/margins": 0.19182677567005157,
      "rewards/rejected": -0.020492086187005043,
      "step": 290
    },
    {
      "epoch": 0.78,
      "learning_rate": 7.08321427484816e-07,
      "logits/chosen": -1.929610013961792,
      "logits/rejected": -1.9270601272583008,
      "logps/chosen": -32.42620086669922,
      "logps/rejected": -30.873455047607422,
      "loss": 0.5301,
      "rewards/accuracies": 0.737500011920929,
      "rewards/chosen": 0.450817346572876,
      "rewards/margins": 0.5018006563186646,
      "rewards/rejected": -0.050983332097530365,
      "step": 300
    },
    {
      "epoch": 0.78,
      "eval_logits/chosen": -2.229154348373413,
      "eval_logits/rejected": -2.2243051528930664,
      "eval_logps/chosen": -34.09621810913086,
      "eval_logps/rejected": -37.59999084472656,
      "eval_loss": 0.6972895860671997,
      "eval_rewards/accuracies": 0.5390365719795227,
      "eval_rewards/chosen": -0.05550166219472885,
      "eval_rewards/margins": 0.019528048112988472,
      "eval_rewards/rejected": -0.07502970844507217,
      "eval_runtime": 145.7792,
      "eval_samples_per_second": 2.353,
      "eval_steps_per_second": 0.295,
      "step": 300
    },
    {
      "epoch": 0.81,
      "learning_rate": 5.576113578589035e-07,
      "logits/chosen": -1.9142345190048218,
      "logits/rejected": -1.9109809398651123,
      "logps/chosen": -31.33791732788086,
      "logps/rejected": -33.82014465332031,
      "loss": 0.5861,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.23888680338859558,
      "rewards/margins": 0.30907896161079407,
      "rewards/rejected": -0.07019217312335968,
      "step": 310
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.229036944380913e-07,
      "logits/chosen": -1.9650068283081055,
      "logits/rejected": -1.9527791738510132,
      "logps/chosen": -34.34791946411133,
      "logps/rejected": -33.650447845458984,
      "loss": 0.5828,
      "rewards/accuracies": 0.6499999761581421,
      "rewards/chosen": 0.22334297001361847,
      "rewards/margins": 0.3237887918949127,
      "rewards/rejected": -0.10044582933187485,
      "step": 320
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.053082288996112e-07,
      "logits/chosen": -2.00040602684021,
      "logits/rejected": -1.9989902973175049,
      "logps/chosen": -33.210105895996094,
      "logps/rejected": -32.56142807006836,
      "loss": 0.5803,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.24812059104442596,
      "rewards/margins": 0.32178014516830444,
      "rewards/rejected": -0.07365953922271729,
      "step": 330
    },
    {
      "epoch": 0.88,
      "learning_rate": 2.0579377374915805e-07,
      "logits/chosen": -2.0870866775512695,
      "logits/rejected": -2.0713772773742676,
      "logps/chosen": -33.80995178222656,
      "logps/rejected": -33.120697021484375,
      "loss": 0.5723,
      "rewards/accuracies": 0.75,
      "rewards/chosen": 0.3324963450431824,
      "rewards/margins": 0.33488941192626953,
      "rewards/rejected": -0.0023930787574499846,
      "step": 340
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.2518018074041684e-07,
      "logits/chosen": -1.959240198135376,
      "logits/rejected": -1.9583876132965088,
      "logps/chosen": -32.863216400146484,
      "logps/rejected": -32.54397201538086,
      "loss": 0.5523,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 0.37699171900749207,
      "rewards/margins": 0.44793614745140076,
      "rewards/rejected": -0.0709443911910057,
      "step": 350
    },
    {
      "epoch": 0.94,
      "learning_rate": 6.41315865106129e-08,
      "logits/chosen": -1.9147189855575562,
      "logits/rejected": -1.9250224828720093,
      "logps/chosen": -31.902795791625977,
      "logps/rejected": -35.3552131652832,
      "loss": 0.5743,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.2717086672782898,
      "rewards/margins": 0.33151620626449585,
      "rewards/rejected": -0.05980752781033516,
      "step": 360
    },
    {
      "epoch": 0.96,
      "learning_rate": 2.3150941078050325e-08,
      "logits/chosen": -2.05413556098938,
      "logits/rejected": -2.047651767730713,
      "logps/chosen": -33.377376556396484,
      "logps/rejected": -29.2799072265625,
      "loss": 0.5801,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.24713313579559326,
      "rewards/margins": 0.2952673137187958,
      "rewards/rejected": -0.04813414067029953,
      "step": 370
    },
    {
      "epoch": 0.99,
      "learning_rate": 2.575864278703266e-09,
      "logits/chosen": -1.9141871929168701,
      "logits/rejected": -1.9163949489593506,
      "logps/chosen": -33.87698745727539,
      "logps/rejected": -30.976858139038086,
      "loss": 0.5489,
      "rewards/accuracies": 0.699999988079071,
      "rewards/chosen": 0.32585546374320984,
      "rewards/margins": 0.4196627140045166,
      "rewards/rejected": -0.09380728751420975,
      "step": 380
    },
    {
      "epoch": 1.0,
      "step": 385,
      "total_flos": 0.0,
      "train_loss": 0.6175476637753573,
      "train_runtime": 3252.7839,
      "train_samples_per_second": 0.947,
      "train_steps_per_second": 0.118
    }
  ],
  "logging_steps": 10,
  "max_steps": 385,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}