{
  "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": 63.20117586062747,
      "learning_rate": 5e-07,
      "logits/chosen": -2.727349281311035,
      "logits/rejected": -2.7335803508758545,
      "logps/chosen": -270.8079833984375,
      "logps/rejected": -222.7043914794922,
      "loss": 0.6897,
      "rewards/accuracies": 0.3187499940395355,
      "rewards/chosen": 0.013827224262058735,
      "rewards/margins": 0.000779601454269141,
      "rewards/rejected": 0.013047623448073864,
      "step": 5
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 50.76880252527562,
      "learning_rate": 1e-06,
      "logits/chosen": -2.686920166015625,
      "logits/rejected": -2.6769723892211914,
      "logps/chosen": -249.0262451171875,
      "logps/rejected": -231.17623901367188,
      "loss": 0.6468,
      "rewards/accuracies": 0.5375000238418579,
      "rewards/chosen": 0.47842222452163696,
      "rewards/margins": 0.08185982704162598,
      "rewards/rejected": 0.396562397480011,
      "step": 10
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 44.786597014715575,
      "learning_rate": 9.983100718730718e-07,
      "logits/chosen": -2.4991869926452637,
      "logits/rejected": -2.519619941711426,
      "logps/chosen": -257.5209045410156,
      "logps/rejected": -217.60446166992188,
      "loss": 0.645,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.3132820129394531,
      "rewards/margins": 0.6335776448249817,
      "rewards/rejected": 0.6797044277191162,
      "step": 15
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 88.88932587574585,
      "learning_rate": 9.932517109205849e-07,
      "logits/chosen": -2.3716320991516113,
      "logits/rejected": -2.350722551345825,
      "logps/chosen": -275.037353515625,
      "logps/rejected": -229.4193878173828,
      "loss": 0.639,
      "rewards/accuracies": 0.668749988079071,
      "rewards/chosen": 1.544551134109497,
      "rewards/margins": 0.876867949962616,
      "rewards/rejected": 0.6676831245422363,
      "step": 20
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 42.980070421826184,
      "learning_rate": 9.848591102083375e-07,
      "logits/chosen": -2.2988502979278564,
      "logits/rejected": -2.288430690765381,
      "logps/chosen": -244.6092987060547,
      "logps/rejected": -221.5340118408203,
      "loss": 0.617,
      "rewards/accuracies": 0.6875,
      "rewards/chosen": 1.3750474452972412,
      "rewards/margins": 1.164804220199585,
      "rewards/rejected": 0.21024329960346222,
      "step": 25
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 37.576258990861845,
      "learning_rate": 9.731890013043367e-07,
      "logits/chosen": -2.3538706302642822,
      "logits/rejected": -2.3402347564697266,
      "logps/chosen": -244.2089080810547,
      "logps/rejected": -220.82180786132812,
      "loss": 0.5935,
      "rewards/accuracies": 0.6937500238418579,
      "rewards/chosen": 1.0666203498840332,
      "rewards/margins": 1.0009348392486572,
      "rewards/rejected": 0.06568538397550583,
      "step": 30
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 48.27486571141399,
      "learning_rate": 9.583202707897073e-07,
      "logits/chosen": -2.4315402507781982,
      "logits/rejected": -2.4107561111450195,
      "logps/chosen": -258.6941223144531,
      "logps/rejected": -200.2947998046875,
      "loss": 0.5815,
      "rewards/accuracies": 0.7749999761581421,
      "rewards/chosen": 0.854897677898407,
      "rewards/margins": 1.0769257545471191,
      "rewards/rejected": -0.2220279723405838,
      "step": 35
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 34.10499734328659,
      "learning_rate": 9.403534270080829e-07,
      "logits/chosen": -2.478212833404541,
      "logits/rejected": -2.4579110145568848,
      "logps/chosen": -247.2701416015625,
      "logps/rejected": -204.61563110351562,
      "loss": 0.5553,
      "rewards/accuracies": 0.7875000238418579,
      "rewards/chosen": 0.8650096654891968,
      "rewards/margins": 1.3998459577560425,
      "rewards/rejected": -0.5348363518714905,
      "step": 40
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 54.84196205672111,
      "learning_rate": 9.19409920658098e-07,
      "logits/chosen": -2.4562790393829346,
      "logits/rejected": -2.406428575515747,
      "logps/chosen": -251.5320587158203,
      "logps/rejected": -196.53280639648438,
      "loss": 0.6172,
      "rewards/accuracies": 0.7250000238418579,
      "rewards/chosen": 0.724992573261261,
      "rewards/margins": 0.909608006477356,
      "rewards/rejected": -0.18461540341377258,
      "step": 45
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 44.60950151362463,
      "learning_rate": 8.956313238215823e-07,
      "logits/chosen": -2.429295063018799,
      "logits/rejected": -2.4014017581939697,
      "logps/chosen": -267.9967041015625,
      "logps/rejected": -231.6461639404297,
      "loss": 0.5811,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.032130241394043,
      "rewards/margins": 1.233904242515564,
      "rewards/rejected": -0.20177388191223145,
      "step": 50
    },
    {
      "epoch": 0.746268656716418,
      "eval_logits/chosen": -2.379634141921997,
      "eval_logits/rejected": -2.3887269496917725,
      "eval_logps/chosen": -262.3047790527344,
      "eval_logps/rejected": -217.30076599121094,
      "eval_loss": 0.5699393153190613,
      "eval_rewards/accuracies": 0.7333333492279053,
      "eval_rewards/chosen": 0.5697149634361267,
      "eval_rewards/margins": 1.181886076927185,
      "eval_rewards/rejected": -0.6121711730957031,
      "eval_runtime": 125.8925,
      "eval_samples_per_second": 15.092,
      "eval_steps_per_second": 0.238,
      "step": 50
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 38.86022913172134,
      "learning_rate": 8.691783729769873e-07,
      "logits/chosen": -2.384519338607788,
      "logits/rejected": -2.369540214538574,
      "logps/chosen": -268.18524169921875,
      "logps/rejected": -256.36614990234375,
      "loss": 0.5328,
      "rewards/accuracies": 0.7437499761581421,
      "rewards/chosen": 0.911398708820343,
      "rewards/margins": 1.7182573080062866,
      "rewards/rejected": -0.8068588376045227,
      "step": 55
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 35.804216796075785,
      "learning_rate": 8.402298824670029e-07,
      "logits/chosen": -2.4195339679718018,
      "logits/rejected": -2.3873202800750732,
      "logps/chosen": -259.75775146484375,
      "logps/rejected": -212.31201171875,
      "loss": 0.533,
      "rewards/accuracies": 0.768750011920929,
      "rewards/chosen": 1.215578317642212,
      "rewards/margins": 1.4803569316864014,
      "rewards/rejected": -0.2647787034511566,
      "step": 60
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 38.53205216239837,
      "learning_rate": 8.089815357650089e-07,
      "logits/chosen": -2.5004165172576904,
      "logits/rejected": -2.459744930267334,
      "logps/chosen": -245.0717315673828,
      "logps/rejected": -219.3514862060547,
      "loss": 0.5168,
      "rewards/accuracies": 0.71875,
      "rewards/chosen": 1.0596349239349365,
      "rewards/margins": 1.2952834367752075,
      "rewards/rejected": -0.23564846813678741,
      "step": 65
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 18.93894037354211,
      "learning_rate": 7.756445627110522e-07,
      "logits/chosen": -2.5253801345825195,
      "logits/rejected": -2.498584032058716,
      "logps/chosen": -256.0594482421875,
      "logps/rejected": -230.0350799560547,
      "loss": 0.3642,
      "rewards/accuracies": 0.831250011920929,
      "rewards/chosen": 1.1649268865585327,
      "rewards/margins": 2.232837200164795,
      "rewards/rejected": -1.067910075187683,
      "step": 70
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 20.704224755579,
      "learning_rate": 7.404443116588547e-07,
      "logits/chosen": -2.500047206878662,
      "logits/rejected": -2.477970838546753,
      "logps/chosen": -232.2460479736328,
      "logps/rejected": -218.29568481445312,
      "loss": 0.2398,
      "rewards/accuracies": 0.8812500238418579,
      "rewards/chosen": 1.6713672876358032,
      "rewards/margins": 2.5339531898498535,
      "rewards/rejected": -0.8625856637954712,
      "step": 75
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 18.90792590059534,
      "learning_rate": 7.036187261857288e-07,
      "logits/chosen": -2.4836063385009766,
      "logits/rejected": -2.4460768699645996,
      "logps/chosen": -245.5286407470703,
      "logps/rejected": -243.6932830810547,
      "loss": 0.2302,
      "rewards/accuracies": 0.925000011920929,
      "rewards/chosen": 2.2828805446624756,
      "rewards/margins": 3.4864933490753174,
      "rewards/rejected": -1.2036125659942627,
      "step": 80
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 20.90479576440656,
      "learning_rate": 6.654167366624008e-07,
      "logits/chosen": -2.427332878112793,
      "logits/rejected": -2.425985813140869,
      "logps/chosen": -233.1042938232422,
      "logps/rejected": -215.51437377929688,
      "loss": 0.2332,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.7251182794570923,
      "rewards/margins": 3.2692294120788574,
      "rewards/rejected": -1.5441116094589233,
      "step": 85
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 23.387687179198938,
      "learning_rate": 6.260965775552713e-07,
      "logits/chosen": -2.4304587841033936,
      "logits/rejected": -2.4162423610687256,
      "logps/chosen": -247.0415802001953,
      "logps/rejected": -230.84005737304688,
      "loss": 0.253,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 1.8513774871826172,
      "rewards/margins": 3.377821445465088,
      "rewards/rejected": -1.5264440774917603,
      "step": 90
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 25.290227374387985,
      "learning_rate": 5.859240418356614e-07,
      "logits/chosen": -2.4067845344543457,
      "logits/rejected": -2.403015375137329,
      "logps/chosen": -229.62283325195312,
      "logps/rejected": -227.48825073242188,
      "loss": 0.2631,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.354150652885437,
      "rewards/margins": 3.5103206634521484,
      "rewards/rejected": -2.156169891357422,
      "step": 95
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 24.465863734128728,
      "learning_rate": 5.451706842957421e-07,
      "logits/chosen": -2.4527347087860107,
      "logits/rejected": -2.454359292984009,
      "logps/chosen": -250.5390625,
      "logps/rejected": -231.2834930419922,
      "loss": 0.2898,
      "rewards/accuracies": 0.918749988079071,
      "rewards/chosen": 2.1465234756469727,
      "rewards/margins": 3.222960948944092,
      "rewards/rejected": -1.076436996459961,
      "step": 100
    },
    {
      "epoch": 1.4925373134328357,
      "eval_logits/chosen": -2.4716811180114746,
      "eval_logits/rejected": -2.4747350215911865,
      "eval_logps/chosen": -255.55557250976562,
      "eval_logps/rejected": -216.7303009033203,
      "eval_loss": 0.5633274912834167,
      "eval_rewards/accuracies": 0.7583333253860474,
      "eval_rewards/chosen": 1.2446327209472656,
      "eval_rewards/margins": 1.799760103225708,
      "eval_rewards/rejected": -0.5551272034645081,
      "eval_runtime": 125.768,
      "eval_samples_per_second": 15.107,
      "eval_steps_per_second": 0.239,
      "step": 100
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 20.464055771718126,
      "learning_rate": 5.041119859162068e-07,
      "logits/chosen": -2.4887359142303467,
      "logits/rejected": -2.4883406162261963,
      "logps/chosen": -241.1067657470703,
      "logps/rejected": -236.4036407470703,
      "loss": 0.2366,
      "rewards/accuracies": 0.9437500238418579,
      "rewards/chosen": 2.339883327484131,
      "rewards/margins": 3.8258018493652344,
      "rewards/rejected": -1.4859185218811035,
      "step": 105
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 26.06752671144799,
      "learning_rate": 4.630254916940423e-07,
      "logits/chosen": -2.5059611797332764,
      "logits/rejected": -2.51178240776062,
      "logps/chosen": -234.89450073242188,
      "logps/rejected": -227.43338012695312,
      "loss": 0.2498,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.0018811225891113,
      "rewards/margins": 3.5609130859375,
      "rewards/rejected": -1.5590317249298096,
      "step": 110
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 22.873206758196236,
      "learning_rate": 4.2218893451814e-07,
      "logits/chosen": -2.5144965648651123,
      "logits/rejected": -2.499683380126953,
      "logps/chosen": -239.4272003173828,
      "logps/rejected": -222.2007293701172,
      "loss": 0.2672,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 1.9761626720428467,
      "rewards/margins": 3.7179388999938965,
      "rewards/rejected": -1.7417774200439453,
      "step": 115
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 24.244439222628877,
      "learning_rate": 3.8187835777481375e-07,
      "logits/chosen": -2.5099034309387207,
      "logits/rejected": -2.512587308883667,
      "logps/chosen": -252.3924560546875,
      "logps/rejected": -243.76400756835938,
      "loss": 0.2711,
      "rewards/accuracies": 0.8999999761581421,
      "rewards/chosen": 2.189035177230835,
      "rewards/margins": 3.753054141998291,
      "rewards/rejected": -1.5640188455581665,
      "step": 120
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 20.99147794265164,
      "learning_rate": 3.423662493738687e-07,
      "logits/chosen": -2.500354290008545,
      "logits/rejected": -2.4761505126953125,
      "logps/chosen": -247.8779296875,
      "logps/rejected": -224.32754516601562,
      "loss": 0.2502,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.132331371307373,
      "rewards/margins": 3.420536756515503,
      "rewards/rejected": -1.288205623626709,
      "step": 125
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 24.00417787159452,
      "learning_rate": 3.039196998086687e-07,
      "logits/chosen": -2.4695286750793457,
      "logits/rejected": -2.4575393199920654,
      "logps/chosen": -234.2464599609375,
      "logps/rejected": -232.9207305908203,
      "loss": 0.281,
      "rewards/accuracies": 0.887499988079071,
      "rewards/chosen": 1.868446707725525,
      "rewards/margins": 3.2393386363983154,
      "rewards/rejected": -1.3708916902542114,
      "step": 130
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 15.645649821850194,
      "learning_rate": 2.667985967011878e-07,
      "logits/chosen": -2.4196391105651855,
      "logits/rejected": -2.4008140563964844,
      "logps/chosen": -224.33154296875,
      "logps/rejected": -219.89468383789062,
      "loss": 0.242,
      "rewards/accuracies": 0.84375,
      "rewards/chosen": 1.9344043731689453,
      "rewards/margins": 3.4023728370666504,
      "rewards/rejected": -1.4679681062698364,
      "step": 135
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 16.08080556848118,
      "learning_rate": 2.3125386803640183e-07,
      "logits/chosen": -2.387631893157959,
      "logits/rejected": -2.358504295349121,
      "logps/chosen": -251.0842742919922,
      "logps/rejected": -223.662841796875,
      "loss": 0.1475,
      "rewards/accuracies": 0.981249988079071,
      "rewards/chosen": 2.8602161407470703,
      "rewards/margins": 4.547825813293457,
      "rewards/rejected": -1.6876094341278076,
      "step": 140
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 11.273909890728381,
      "learning_rate": 1.9752578596124952e-07,
      "logits/chosen": -2.3566510677337646,
      "logits/rejected": -2.3424291610717773,
      "logps/chosen": -231.8817901611328,
      "logps/rejected": -252.8689422607422,
      "loss": 0.1199,
      "rewards/accuracies": 0.9937499761581421,
      "rewards/chosen": 2.450244426727295,
      "rewards/margins": 4.638026237487793,
      "rewards/rejected": -2.187781810760498,
      "step": 145
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 12.684257475594388,
      "learning_rate": 1.6584234261399532e-07,
      "logits/chosen": -2.3457634449005127,
      "logits/rejected": -2.318638563156128,
      "logps/chosen": -238.44412231445312,
      "logps/rejected": -241.29150390625,
      "loss": 0.131,
      "rewards/accuracies": 0.9750000238418579,
      "rewards/chosen": 2.502916097640991,
      "rewards/margins": 4.269684314727783,
      "rewards/rejected": -1.7667678594589233,
      "step": 150
    },
    {
      "epoch": 2.2388059701492535,
      "eval_logits/chosen": -2.324143409729004,
      "eval_logits/rejected": -2.3181183338165283,
      "eval_logps/chosen": -255.06072998046875,
      "eval_logps/rejected": -218.32066345214844,
      "eval_loss": 0.5345308184623718,
      "eval_rewards/accuracies": 0.762499988079071,
      "eval_rewards/chosen": 1.2941176891326904,
      "eval_rewards/margins": 2.008281946182251,
      "eval_rewards/rejected": -0.7141639590263367,
      "eval_runtime": 125.7554,
      "eval_samples_per_second": 15.109,
      "eval_steps_per_second": 0.239,
      "step": 150
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 14.572922876201492,
      "learning_rate": 1.3641770896292082e-07,
      "logits/chosen": -2.3410754203796387,
      "logits/rejected": -2.2934837341308594,
      "logps/chosen": -236.2089385986328,
      "logps/rejected": -228.3870086669922,
      "loss": 0.1338,
      "rewards/accuracies": 0.949999988079071,
      "rewards/chosen": 2.325601100921631,
      "rewards/margins": 4.146761417388916,
      "rewards/rejected": -1.821160078048706,
      "step": 155
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 18.210819683877478,
      "learning_rate": 1.0945078707215221e-07,
      "logits/chosen": -2.299962043762207,
      "logits/rejected": -2.2716877460479736,
      "logps/chosen": -234.1040496826172,
      "logps/rejected": -209.2781982421875,
      "loss": 0.1261,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 2.0497689247131348,
      "rewards/margins": 3.610365629196167,
      "rewards/rejected": -1.5605969429016113,
      "step": 160
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 16.23282621614867,
      "learning_rate": 8.512386558088919e-08,
      "logits/chosen": -2.286097764968872,
      "logits/rejected": -2.2833445072174072,
      "logps/chosen": -255.2034454345703,
      "logps/rejected": -240.033203125,
      "loss": 0.1333,
      "rewards/accuracies": 0.9125000238418579,
      "rewards/chosen": 2.2124927043914795,
      "rewards/margins": 4.191450595855713,
      "rewards/rejected": -1.978957176208496,
      "step": 165
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 20.304425686806827,
      "learning_rate": 6.360138748461013e-08,
      "logits/chosen": -2.2931156158447266,
      "logits/rejected": -2.2632596492767334,
      "logps/chosen": -226.07321166992188,
      "logps/rejected": -235.8397979736328,
      "loss": 0.1608,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.336498498916626,
      "rewards/margins": 4.572902679443359,
      "rewards/rejected": -2.2364044189453125,
      "step": 170
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 18.176538202328782,
      "learning_rate": 4.5028838547699346e-08,
      "logits/chosen": -2.2854504585266113,
      "logits/rejected": -2.2505314350128174,
      "logps/chosen": -228.4530487060547,
      "logps/rejected": -208.823486328125,
      "loss": 0.1623,
      "rewards/accuracies": 0.90625,
      "rewards/chosen": 1.8784024715423584,
      "rewards/margins": 3.9313271045684814,
      "rewards/rejected": -2.052924394607544,
      "step": 175
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 18.209188861736383,
      "learning_rate": 2.9531763861505964e-08,
      "logits/chosen": -2.308990955352783,
      "logits/rejected": -2.2554683685302734,
      "logps/chosen": -231.95565795898438,
      "logps/rejected": -229.5238800048828,
      "loss": 0.1508,
      "rewards/accuracies": 0.9375,
      "rewards/chosen": 2.1670119762420654,
      "rewards/margins": 4.290619850158691,
      "rewards/rejected": -2.1236071586608887,
      "step": 180
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 18.30616144833626,
      "learning_rate": 1.7214919195619125e-08,
      "logits/chosen": -2.3051724433898926,
      "logits/rejected": -2.2820804119110107,
      "logps/chosen": -248.5728759765625,
      "logps/rejected": -230.4145050048828,
      "loss": 0.1535,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.2131433486938477,
      "rewards/margins": 4.307113170623779,
      "rewards/rejected": -2.0939700603485107,
      "step": 185
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 18.174652472233937,
      "learning_rate": 8.161562878982398e-09,
      "logits/chosen": -2.322986125946045,
      "logits/rejected": -2.2646212577819824,
      "logps/chosen": -232.8801727294922,
      "logps/rejected": -228.6397247314453,
      "loss": 0.1079,
      "rewards/accuracies": 0.987500011920929,
      "rewards/chosen": 2.3938283920288086,
      "rewards/margins": 4.4372968673706055,
      "rewards/rejected": -2.043468475341797,
      "step": 190
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 12.452523630475666,
      "learning_rate": 2.432892997526026e-09,
      "logits/chosen": -2.3041179180145264,
      "logits/rejected": -2.2811574935913086,
      "logps/chosen": -239.6481170654297,
      "logps/rejected": -242.7835235595703,
      "loss": 0.1255,
      "rewards/accuracies": 0.956250011920929,
      "rewards/chosen": 2.288182020187378,
      "rewards/margins": 4.6204833984375,
      "rewards/rejected": -2.332301616668701,
      "step": 195
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 14.7942664091877,
      "learning_rate": 6.763371270035457e-11,
      "logits/chosen": -2.315195322036743,
      "logits/rejected": -2.276332139968872,
      "logps/chosen": -239.46133422851562,
      "logps/rejected": -227.2948455810547,
      "loss": 0.1357,
      "rewards/accuracies": 0.9624999761581421,
      "rewards/chosen": 2.137551784515381,
      "rewards/margins": 4.160203456878662,
      "rewards/rejected": -2.0226521492004395,
      "step": 200
    },
    {
      "epoch": 2.9850746268656714,
      "eval_logits/chosen": -2.298764228820801,
      "eval_logits/rejected": -2.289949893951416,
      "eval_logps/chosen": -257.3822021484375,
      "eval_logps/rejected": -221.44564819335938,
      "eval_loss": 0.5440204739570618,
      "eval_rewards/accuracies": 0.7333333492279053,
      "eval_rewards/chosen": 1.0619722604751587,
      "eval_rewards/margins": 2.088634729385376,
      "eval_rewards/rejected": -1.0266624689102173,
      "eval_runtime": 125.8013,
      "eval_samples_per_second": 15.103,
      "eval_steps_per_second": 0.238,
      "step": 200
    },
    {
      "epoch": 3.0,
      "step": 201,
      "total_flos": 2369906314051584.0,
      "train_loss": 0.32917294554894244,
      "train_runtime": 7373.939,
      "train_samples_per_second": 6.954,
      "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
}