{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 201,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 56.38858835713434,
      "learning_rate": 5e-07,
      "logits/chosen": -2.731353282928467,
      "logits/rejected": -2.7158660888671875,
      "logps/chosen": -256.7725524902344,
      "logps/rejected": -205.1663055419922,
      "loss": 0.6903,
      "rewards/accuracies": 0.35624998807907104,
      "rewards/chosen": 0.014972714707255363,
      "rewards/margins": 0.006492167711257935,
      "rewards/rejected": 0.008480546995997429,
      "step": 5
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 48.210654865798084,
      "learning_rate": 1e-06,
      "logits/chosen": -2.6910791397094727,
      "logits/rejected": -2.686784267425537,
      "logps/chosen": -261.28240966796875,
      "logps/rejected": -210.28414916992188,
      "loss": 0.6445,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 0.5361045002937317,
      "rewards/margins": 0.2673659920692444,
      "rewards/rejected": 0.2687385082244873,
      "step": 10
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 48.18867539560807,
      "learning_rate": 9.983100718730718e-07,
      "logits/chosen": -2.5146777629852295,
      "logits/rejected": -2.5069146156311035,
      "logps/chosen": -274.51788330078125,
      "logps/rejected": -212.06613159179688,
      "loss": 0.6397,
      "rewards/accuracies": 0.6812499761581421,
      "rewards/chosen": 1.0824434757232666,
      "rewards/margins": 0.4525600075721741,
      "rewards/rejected": 0.6298834085464478,
      "step": 15
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 56.03620889669535,
      "learning_rate": 9.932517109205849e-07,
      "logits/chosen": -2.346874237060547,
      "logits/rejected": -2.3457634449005127,
      "logps/chosen": -233.57192993164062,
      "logps/rejected": -216.9462890625,
      "loss": 0.6508,
      "rewards/accuracies": 0.675000011920929,
      "rewards/chosen": 1.2721847295761108,
      "rewards/margins": 0.8419507741928101,
      "rewards/rejected": 0.4302339553833008,
      "step": 20
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 44.42398594582461,
      "learning_rate": 9.848591102083375e-07,
      "logits/chosen": -2.321877956390381,
      "logits/rejected": -2.3033697605133057,
      "logps/chosen": -276.81353759765625,
      "logps/rejected": -215.10543823242188,
      "loss": 0.5836,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 1.4672718048095703,
      "rewards/margins": 0.9575842618942261,
      "rewards/rejected": 0.5096874237060547,
      "step": 25
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 47.04899305167527,
      "learning_rate": 9.731890013043367e-07,
      "logits/chosen": -2.3766517639160156,
      "logits/rejected": -2.375422954559326,
      "logps/chosen": -253.60995483398438,
      "logps/rejected": -239.47940063476562,
      "loss": 0.634,
      "rewards/accuracies": 0.6312500238418579,
      "rewards/chosen": 1.2893927097320557,
      "rewards/margins": 0.6847591996192932,
      "rewards/rejected": 0.6046335697174072,
      "step": 30
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 39.80811236547202,
      "learning_rate": 9.583202707897073e-07,
      "logits/chosen": -2.4474780559539795,
      "logits/rejected": -2.450472354888916,
      "logps/chosen": -257.7427673339844,
      "logps/rejected": -207.629638671875,
      "loss": 0.5649,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.217116355895996,
      "rewards/margins": 1.1533949375152588,
      "rewards/rejected": 0.06372135877609253,
      "step": 35
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 42.3182391334653,
      "learning_rate": 9.403534270080829e-07,
      "logits/chosen": -2.4868321418762207,
      "logits/rejected": -2.4954237937927246,
      "logps/chosen": -243.6182403564453,
      "logps/rejected": -235.5696563720703,
      "loss": 0.5926,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 1.1232950687408447,
      "rewards/margins": 1.261273741722107,
      "rewards/rejected": -0.13797876238822937,
      "step": 40
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 38.544673892587376,
      "learning_rate": 9.19409920658098e-07,
      "logits/chosen": -2.4237208366394043,
      "logits/rejected": -2.400038480758667,
      "logps/chosen": -247.0177764892578,
      "logps/rejected": -214.0535888671875,
      "loss": 0.6227,
      "rewards/accuracies": 0.762499988079071,
      "rewards/chosen": 1.1583524942398071,
      "rewards/margins": 1.3776183128356934,
      "rewards/rejected": -0.21926572918891907,
      "step": 45
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 44.95262854721761,
      "learning_rate": 8.956313238215823e-07,
      "logits/chosen": -2.3861770629882812,
      "logits/rejected": -2.361314058303833,
      "logps/chosen": -251.87680053710938,
      "logps/rejected": -241.48794555664062,
      "loss": 0.5946,
      "rewards/accuracies": 0.7124999761581421,
      "rewards/chosen": 1.3533507585525513,
      "rewards/margins": 1.143169641494751,
      "rewards/rejected": 0.2101811170578003,
      "step": 50
    },
    {
      "epoch": 0.746268656716418,
      "eval_logits/chosen": -2.419147253036499,
      "eval_logits/rejected": -2.3880856037139893,
      "eval_logps/chosen": -247.12657165527344,
      "eval_logps/rejected": -217.0163116455078,
      "eval_loss": 0.5850147008895874,
      "eval_rewards/accuracies": 0.7333333492279053,
      "eval_rewards/chosen": 1.2898719310760498,
      "eval_rewards/margins": 1.370118498802185,
      "eval_rewards/rejected": -0.0802464708685875,
      "eval_runtime": 126.7763,
      "eval_samples_per_second": 14.987,
      "eval_steps_per_second": 0.237,
      "step": 50
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 34.637816378109704,
      "learning_rate": 8.691783729769873e-07,
      "logits/chosen": -2.4367661476135254,
      "logits/rejected": -2.4355311393737793,
      "logps/chosen": -237.86105346679688,
      "logps/rejected": -211.58084106445312,
      "loss": 0.555,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 1.1501390933990479,
      "rewards/margins": 1.354534387588501,
      "rewards/rejected": -0.20439541339874268,
      "step": 55
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 39.91245885039146,
      "learning_rate": 8.402298824670029e-07,
      "logits/chosen": -2.5068681240081787,
      "logits/rejected": -2.485455274581909,
      "logps/chosen": -256.1775817871094,
      "logps/rejected": -231.7941436767578,
      "loss": 0.611,
      "rewards/accuracies": 0.78125,
      "rewards/chosen": 1.03065025806427,
      "rewards/margins": 1.9949699640274048,
      "rewards/rejected": -0.9643197059631348,
      "step": 60
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 39.08249314957454,
      "learning_rate": 8.089815357650089e-07,
      "logits/chosen": -2.5430567264556885,
      "logits/rejected": -2.543178081512451,
      "logps/chosen": -232.22262573242188,
      "logps/rejected": -225.10220336914062,
      "loss": 0.5475,
      "rewards/accuracies": 0.731249988079071,
      "rewards/chosen": 0.8175865411758423,
      "rewards/margins": 1.328970193862915,
      "rewards/rejected": -0.5113834738731384,
      "step": 65
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 20.186385290734442,
      "learning_rate": 7.756445627110522e-07,
      "logits/chosen": -2.552123546600342,
      "logits/rejected": -2.5313704013824463,
      "logps/chosen": -234.0012664794922,
      "logps/rejected": -240.6761474609375,
      "loss": 0.3553,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": 1.4629334211349487,
      "rewards/margins": 2.36128568649292,
      "rewards/rejected": -0.8983524441719055,
      "step": 70
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 21.595036305155492,
      "learning_rate": 7.404443116588547e-07,
      "logits/chosen": -2.5418105125427246,
      "logits/rejected": -2.5285840034484863,
      "logps/chosen": -242.90499877929688,
      "logps/rejected": -216.7215576171875,
      "loss": 0.2463,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.360262393951416,
      "rewards/margins": 3.110288381576538,
      "rewards/rejected": -0.7500261068344116,
      "step": 75
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 22.4039757471544,
      "learning_rate": 7.036187261857288e-07,
      "logits/chosen": -2.456780195236206,
      "logits/rejected": -2.439521312713623,
      "logps/chosen": -233.08560180664062,
      "logps/rejected": -244.88961791992188,
      "loss": 0.2423,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 2.1394240856170654,
      "rewards/margins": 3.141387939453125,
      "rewards/rejected": -1.0019636154174805,
      "step": 80
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 25.42866646351196,
      "learning_rate": 6.654167366624008e-07,
      "logits/chosen": -2.3945281505584717,
      "logits/rejected": -2.3779776096343994,
      "logps/chosen": -230.88217163085938,
      "logps/rejected": -222.7549285888672,
      "loss": 0.2769,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.6151447296142578,
      "rewards/margins": 2.812251091003418,
      "rewards/rejected": -1.1971065998077393,
      "step": 85
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 23.51747857481762,
      "learning_rate": 6.260965775552713e-07,
      "logits/chosen": -2.361176013946533,
      "logits/rejected": -2.3344523906707764,
      "logps/chosen": -236.94775390625,
      "logps/rejected": -242.43069458007812,
      "loss": 0.2743,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 1.946207046508789,
      "rewards/margins": 3.319734573364258,
      "rewards/rejected": -1.3735275268554688,
      "step": 90
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 25.346384805721517,
      "learning_rate": 5.859240418356614e-07,
      "logits/chosen": -2.3584346771240234,
      "logits/rejected": -2.3219895362854004,
      "logps/chosen": -246.46530151367188,
      "logps/rejected": -231.7371063232422,
      "loss": 0.2576,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 2.247493267059326,
      "rewards/margins": 3.8193671703338623,
      "rewards/rejected": -1.571873426437378,
      "step": 95
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 17.435014853431674,
      "learning_rate": 5.451706842957421e-07,
      "logits/chosen": -2.4034814834594727,
      "logits/rejected": -2.3993496894836426,
      "logps/chosen": -248.6347198486328,
      "logps/rejected": -231.65682983398438,
      "loss": 0.255,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.421250820159912,
      "rewards/margins": 3.7864270210266113,
      "rewards/rejected": -1.3651763200759888,
      "step": 100
    },
    {
      "epoch": 1.4925373134328357,
      "eval_logits/chosen": -2.4342453479766846,
      "eval_logits/rejected": -2.4099276065826416,
      "eval_logps/chosen": -245.49957275390625,
      "eval_logps/rejected": -221.9740753173828,
      "eval_loss": 0.5856931805610657,
      "eval_rewards/accuracies": 0.7958333492279053,
      "eval_rewards/chosen": 1.4525729417800903,
      "eval_rewards/margins": 2.0285959243774414,
      "eval_rewards/rejected": -0.5760230422019958,
      "eval_runtime": 126.6219,
      "eval_samples_per_second": 15.005,
      "eval_steps_per_second": 0.237,
      "step": 100
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 22.513659156325907,
      "learning_rate": 5.041119859162068e-07,
      "logits/chosen": -2.425330400466919,
      "logits/rejected": -2.4230823516845703,
      "logps/chosen": -244.3297576904297,
      "logps/rejected": -250.95852661132812,
      "loss": 0.2902,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 1.9127695560455322,
      "rewards/margins": 3.8896842002868652,
      "rewards/rejected": -1.9769150018692017,
      "step": 105
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 22.093724425686545,
      "learning_rate": 4.630254916940423e-07,
      "logits/chosen": -2.456946849822998,
      "logits/rejected": -2.423444986343384,
      "logps/chosen": -262.3679504394531,
      "logps/rejected": -220.34878540039062,
      "loss": 0.2802,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.1305854320526123,
      "rewards/margins": 3.4911084175109863,
      "rewards/rejected": -1.3605228662490845,
      "step": 110
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 22.848502540826928,
      "learning_rate": 4.2218893451814e-07,
      "logits/chosen": -2.4240517616271973,
      "logits/rejected": -2.400174379348755,
      "logps/chosen": -250.54122924804688,
      "logps/rejected": -235.89712524414062,
      "loss": 0.2883,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 1.2658907175064087,
      "rewards/margins": 3.4836113452911377,
      "rewards/rejected": -2.2177207469940186,
      "step": 115
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 25.377767671408016,
      "learning_rate": 3.8187835777481375e-07,
      "logits/chosen": -2.4062647819519043,
      "logits/rejected": -2.4059927463531494,
      "logps/chosen": -257.41558837890625,
      "logps/rejected": -225.0140380859375,
      "loss": 0.294,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 1.580168604850769,
      "rewards/margins": 3.4512062072753906,
      "rewards/rejected": -1.871037483215332,
      "step": 120
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 21.398907395127186,
      "learning_rate": 3.423662493738687e-07,
      "logits/chosen": -2.4366660118103027,
      "logits/rejected": -2.4090025424957275,
      "logps/chosen": -270.04345703125,
      "logps/rejected": -228.82913208007812,
      "loss": 0.2818,
      "rewards/accuracies": 0.893750011920929,
      "rewards/chosen": 1.6028369665145874,
      "rewards/margins": 3.6763553619384766,
      "rewards/rejected": -2.0735180377960205,
      "step": 125
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 19.350854009924277,
      "learning_rate": 3.039196998086687e-07,
      "logits/chosen": -2.4322643280029297,
      "logits/rejected": -2.4288134574890137,
      "logps/chosen": -243.41943359375,
      "logps/rejected": -237.18588256835938,
      "loss": 0.2664,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.5144699811935425,
      "rewards/margins": 3.2811291217803955,
      "rewards/rejected": -1.7666590213775635,
      "step": 130
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 12.838839331293379,
      "learning_rate": 2.667985967011878e-07,
      "logits/chosen": -2.4573187828063965,
      "logits/rejected": -2.441751003265381,
      "logps/chosen": -242.3375244140625,
      "logps/rejected": -235.7463836669922,
      "loss": 0.2602,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.0578320026397705,
      "rewards/margins": 3.559199810028076,
      "rewards/rejected": -1.5013679265975952,
      "step": 135
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 13.535639161407488,
      "learning_rate": 2.3125386803640183e-07,
      "logits/chosen": -2.4401307106018066,
      "logits/rejected": -2.417816638946533,
      "logps/chosen": -234.63552856445312,
      "logps/rejected": -217.58316040039062,
      "loss": 0.1614,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 2.3646421432495117,
      "rewards/margins": 3.8635189533233643,
      "rewards/rejected": -1.498876690864563,
      "step": 140
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 12.98503680311861,
      "learning_rate": 1.9752578596124952e-07,
      "logits/chosen": -2.3909850120544434,
      "logits/rejected": -2.3737857341766357,
      "logps/chosen": -237.2566680908203,
      "logps/rejected": -225.0938262939453,
      "loss": 0.1588,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 2.3333511352539062,
      "rewards/margins": 3.83681058883667,
      "rewards/rejected": -1.5034593343734741,
      "step": 145
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 17.53696861401471,
      "learning_rate": 1.6584234261399532e-07,
      "logits/chosen": -2.37007212638855,
      "logits/rejected": -2.352189779281616,
      "logps/chosen": -248.3662567138672,
      "logps/rejected": -211.06494140625,
      "loss": 0.1492,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 2.444425344467163,
      "rewards/margins": 3.893583297729492,
      "rewards/rejected": -1.4491578340530396,
      "step": 150
    },
    {
      "epoch": 2.2388059701492535,
      "eval_logits/chosen": -2.3633620738983154,
      "eval_logits/rejected": -2.328437328338623,
      "eval_logps/chosen": -245.0876922607422,
      "eval_logps/rejected": -221.6795196533203,
      "eval_loss": 0.5705874562263489,
      "eval_rewards/accuracies": 0.7916666865348816,
      "eval_rewards/chosen": 1.4937634468078613,
      "eval_rewards/margins": 2.040332555770874,
      "eval_rewards/rejected": -0.5465689897537231,
      "eval_runtime": 126.3365,
      "eval_samples_per_second": 15.039,
      "eval_steps_per_second": 0.237,
      "step": 150
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 21.317610168262696,
      "learning_rate": 1.3641770896292082e-07,
      "logits/chosen": -2.355787754058838,
      "logits/rejected": -2.3269476890563965,
      "logps/chosen": -244.22164916992188,
      "logps/rejected": -246.47555541992188,
      "loss": 0.1582,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.310300350189209,
      "rewards/margins": 3.7411086559295654,
      "rewards/rejected": -1.4308083057403564,
      "step": 155
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 22.107512419038276,
      "learning_rate": 1.0945078707215221e-07,
      "logits/chosen": -2.354370594024658,
      "logits/rejected": -2.334298610687256,
      "logps/chosen": -240.81668090820312,
      "logps/rejected": -231.38491821289062,
      "loss": 0.1805,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.6548850536346436,
      "rewards/margins": 4.173087120056152,
      "rewards/rejected": -1.5182020664215088,
      "step": 160
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 16.052607396027245,
      "learning_rate": 8.512386558088919e-08,
      "logits/chosen": -2.3480546474456787,
      "logits/rejected": -2.3385627269744873,
      "logps/chosen": -234.52035522460938,
      "logps/rejected": -250.7569580078125,
      "loss": 0.1653,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.110555648803711,
      "rewards/margins": 4.057579040527344,
      "rewards/rejected": -1.9470237493515015,
      "step": 165
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 13.776519776024537,
      "learning_rate": 6.360138748461013e-08,
      "logits/chosen": -2.371123790740967,
      "logits/rejected": -2.344186305999756,
      "logps/chosen": -250.3871307373047,
      "logps/rejected": -216.33401489257812,
      "loss": 0.1378,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.4384984970092773,
      "rewards/margins": 3.937534809112549,
      "rewards/rejected": -1.4990360736846924,
      "step": 170
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 16.423256544059676,
      "learning_rate": 4.5028838547699346e-08,
      "logits/chosen": -2.3799216747283936,
      "logits/rejected": -2.344275951385498,
      "logps/chosen": -246.70285034179688,
      "logps/rejected": -246.2882537841797,
      "loss": 0.136,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.443866729736328,
      "rewards/margins": 4.205219268798828,
      "rewards/rejected": -1.761352777481079,
      "step": 175
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 17.74947523952519,
      "learning_rate": 2.9531763861505964e-08,
      "logits/chosen": -2.356396198272705,
      "logits/rejected": -2.3508291244506836,
      "logps/chosen": -226.5736846923828,
      "logps/rejected": -226.9338836669922,
      "loss": 0.158,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 2.034636974334717,
      "rewards/margins": 3.739074230194092,
      "rewards/rejected": -1.704437017440796,
      "step": 180
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 19.018519807815494,
      "learning_rate": 1.7214919195619125e-08,
      "logits/chosen": -2.3661041259765625,
      "logits/rejected": -2.341630697250366,
      "logps/chosen": -245.50830078125,
      "logps/rejected": -273.1011962890625,
      "loss": 0.1556,
      "rewards/accuracies": 0.96875,
      "rewards/chosen": 2.4174838066101074,
      "rewards/margins": 4.649423599243164,
      "rewards/rejected": -2.2319395542144775,
      "step": 185
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 16.772739857601884,
      "learning_rate": 8.161562878982398e-09,
      "logits/chosen": -2.343977451324463,
      "logits/rejected": -2.338977575302124,
      "logps/chosen": -238.34130859375,
      "logps/rejected": -241.8277130126953,
      "loss": 0.1857,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 2.559854030609131,
      "rewards/margins": 4.380292892456055,
      "rewards/rejected": -1.820439100265503,
      "step": 190
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 23.452590812701438,
      "learning_rate": 2.432892997526026e-09,
      "logits/chosen": -2.3592891693115234,
      "logits/rejected": -2.33678936958313,
      "logps/chosen": -253.15628051757812,
      "logps/rejected": -223.1919403076172,
      "loss": 0.1848,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 2.240382671356201,
      "rewards/margins": 3.8927555084228516,
      "rewards/rejected": -1.6523720026016235,
      "step": 195
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 21.27246528519617,
      "learning_rate": 6.763371270035457e-11,
      "logits/chosen": -2.3560237884521484,
      "logits/rejected": -2.3329052925109863,
      "logps/chosen": -231.1381072998047,
      "logps/rejected": -234.8066864013672,
      "loss": 0.1536,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.2949817180633545,
      "rewards/margins": 4.150731086730957,
      "rewards/rejected": -1.8557497262954712,
      "step": 200
    },
    {
      "epoch": 2.9850746268656714,
      "eval_logits/chosen": -2.3594791889190674,
      "eval_logits/rejected": -2.326153039932251,
      "eval_logps/chosen": -246.07659912109375,
      "eval_logps/rejected": -224.16775512695312,
      "eval_loss": 0.5641939043998718,
      "eval_rewards/accuracies": 0.7916666865348816,
      "eval_rewards/chosen": 1.3948734998703003,
      "eval_rewards/margins": 2.1902639865875244,
      "eval_rewards/rejected": -0.7953903079032898,
      "eval_runtime": 126.3864,
      "eval_samples_per_second": 15.033,
      "eval_steps_per_second": 0.237,
      "step": 200
    },
    {
      "epoch": 3.0,
      "step": 201,
      "total_flos": 2369906314051584.0,
      "train_loss": 0.34582118860524685,
      "train_runtime": 7382.5383,
      "train_samples_per_second": 6.946,
      "train_steps_per_second": 0.027
    }
  ],
  "logging_steps": 5,
  "max_steps": 201,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2369906314051584.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}