zephyr-7b-gpo-v6-i1 / trainer_state.json
lole25's picture
Model save
3b81c80 verified
raw
history blame
121 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.6001919846412287,
"eval_steps": 500,
"global_step": 2501,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 1.9920318725099604e-08,
"logits/chosen": -1.8077198266983032,
"logits/rejected": -1.711557388305664,
"logps/chosen": -187.02471923828125,
"logps/rejected": -122.2266616821289,
"loss": 0.4697,
"rewards/accuracies": 0.0,
"rewards/chosen": 0.0,
"rewards/margins": 0.0,
"rewards/rejected": 0.0,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 1.9920318725099604e-07,
"logits/chosen": -1.91868257522583,
"logits/rejected": -1.382498860359192,
"logps/chosen": -176.21807861328125,
"logps/rejected": -120.50502014160156,
"loss": 0.4051,
"rewards/accuracies": 0.2777777910232544,
"rewards/chosen": -4.647710011340678e-05,
"rewards/margins": -6.936895078979433e-05,
"rewards/rejected": 2.2891843400429934e-05,
"step": 10
},
{
"epoch": 0.0,
"learning_rate": 3.9840637450199207e-07,
"logits/chosen": -1.933166265487671,
"logits/rejected": -1.376651406288147,
"logps/chosen": -183.34518432617188,
"logps/rejected": -109.8729476928711,
"loss": 0.3902,
"rewards/accuracies": 0.574999988079071,
"rewards/chosen": 5.835342381033115e-05,
"rewards/margins": 5.425453491625376e-05,
"rewards/rejected": 4.098887529835338e-06,
"step": 20
},
{
"epoch": 0.01,
"learning_rate": 5.976095617529881e-07,
"logits/chosen": -1.8762012720108032,
"logits/rejected": -1.4956092834472656,
"logps/chosen": -173.73521423339844,
"logps/rejected": -129.2222137451172,
"loss": 0.3999,
"rewards/accuracies": 0.574999988079071,
"rewards/chosen": 0.000283640343695879,
"rewards/margins": 0.0003057140565942973,
"rewards/rejected": -2.2073701984481886e-05,
"step": 30
},
{
"epoch": 0.01,
"learning_rate": 7.968127490039841e-07,
"logits/chosen": -2.074092388153076,
"logits/rejected": -1.6071268320083618,
"logps/chosen": -128.0385284423828,
"logps/rejected": -99.65340423583984,
"loss": 0.3735,
"rewards/accuracies": 0.550000011920929,
"rewards/chosen": 0.0004021901695523411,
"rewards/margins": 0.0006805313169024885,
"rewards/rejected": -0.0002783412055578083,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 9.9601593625498e-07,
"logits/chosen": -2.022669792175293,
"logits/rejected": -1.4573835134506226,
"logps/chosen": -151.2292022705078,
"logps/rejected": -106.66209411621094,
"loss": 0.383,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": 0.0010000061010941863,
"rewards/margins": 0.0013790394878014922,
"rewards/rejected": -0.00037903329939581454,
"step": 50
},
{
"epoch": 0.01,
"learning_rate": 1.1952191235059762e-06,
"logits/chosen": -1.9887306690216064,
"logits/rejected": -1.5061299800872803,
"logps/chosen": -206.5648193359375,
"logps/rejected": -130.1720733642578,
"loss": 0.4111,
"rewards/accuracies": 0.625,
"rewards/chosen": 0.0012637332547456026,
"rewards/margins": 0.0010847109369933605,
"rewards/rejected": 0.00017902204126585275,
"step": 60
},
{
"epoch": 0.02,
"learning_rate": 1.3944223107569721e-06,
"logits/chosen": -2.064723253250122,
"logits/rejected": -1.4734152555465698,
"logps/chosen": -200.28292846679688,
"logps/rejected": -123.68900299072266,
"loss": 0.3915,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.00973748043179512,
"rewards/margins": 0.007851692847907543,
"rewards/rejected": -0.017589174211025238,
"step": 70
},
{
"epoch": 0.02,
"learning_rate": 1.5936254980079683e-06,
"logits/chosen": -1.9502429962158203,
"logits/rejected": -1.3335387706756592,
"logps/chosen": -223.71151733398438,
"logps/rejected": -220.52392578125,
"loss": 0.3923,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.063636913895607,
"rewards/margins": 0.03966347128152847,
"rewards/rejected": -0.10330037772655487,
"step": 80
},
{
"epoch": 0.02,
"learning_rate": 1.7928286852589644e-06,
"logits/chosen": -1.8217380046844482,
"logits/rejected": -1.2573997974395752,
"logps/chosen": -345.07183837890625,
"logps/rejected": -365.8221740722656,
"loss": 0.2872,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.15853700041770935,
"rewards/margins": 0.059460896998643875,
"rewards/rejected": -0.2179979383945465,
"step": 90
},
{
"epoch": 0.02,
"learning_rate": 1.99203187250996e-06,
"logits/chosen": -1.9064254760742188,
"logits/rejected": -1.2212843894958496,
"logps/chosen": -434.5439453125,
"logps/rejected": -479.3141174316406,
"loss": 0.279,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2425861656665802,
"rewards/margins": 0.1260627657175064,
"rewards/rejected": -0.3686489164829254,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 2.1912350597609563e-06,
"logits/chosen": -1.9829845428466797,
"logits/rejected": -1.3048018217086792,
"logps/chosen": -418.07952880859375,
"logps/rejected": -516.9057006835938,
"loss": 0.2641,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.2715674340724945,
"rewards/margins": 0.14631584286689758,
"rewards/rejected": -0.4178832471370697,
"step": 110
},
{
"epoch": 0.03,
"learning_rate": 2.3904382470119524e-06,
"logits/chosen": -1.9264233112335205,
"logits/rejected": -1.4916644096374512,
"logps/chosen": -516.511474609375,
"logps/rejected": -570.0408325195312,
"loss": 0.3119,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.3569049537181854,
"rewards/margins": 0.09669794887304306,
"rewards/rejected": -0.4536028802394867,
"step": 120
},
{
"epoch": 0.03,
"learning_rate": 2.589641434262948e-06,
"logits/chosen": -2.02325177192688,
"logits/rejected": -1.567561388015747,
"logps/chosen": -412.42327880859375,
"logps/rejected": -471.11962890625,
"loss": 0.2955,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.23843295872211456,
"rewards/margins": 0.11186476051807404,
"rewards/rejected": -0.350297749042511,
"step": 130
},
{
"epoch": 0.03,
"learning_rate": 2.7888446215139443e-06,
"logits/chosen": -2.2830417156219482,
"logits/rejected": -1.7622215747833252,
"logps/chosen": -320.74603271484375,
"logps/rejected": -390.9023742675781,
"loss": 0.3239,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.15108834207057953,
"rewards/margins": 0.13633789122104645,
"rewards/rejected": -0.2874262034893036,
"step": 140
},
{
"epoch": 0.04,
"learning_rate": 2.9880478087649404e-06,
"logits/chosen": -2.1028332710266113,
"logits/rejected": -1.609794020652771,
"logps/chosen": -509.71051025390625,
"logps/rejected": -607.0804443359375,
"loss": 0.2944,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.3575127124786377,
"rewards/margins": 0.14793363213539124,
"rewards/rejected": -0.5054463148117065,
"step": 150
},
{
"epoch": 0.04,
"learning_rate": 3.1872509960159366e-06,
"logits/chosen": -2.0964839458465576,
"logits/rejected": -1.6341785192489624,
"logps/chosen": -634.71337890625,
"logps/rejected": -707.5382080078125,
"loss": 0.275,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.45250073075294495,
"rewards/margins": 0.1270226687192917,
"rewards/rejected": -0.5795234441757202,
"step": 160
},
{
"epoch": 0.04,
"learning_rate": 3.3864541832669323e-06,
"logits/chosen": -1.9589221477508545,
"logits/rejected": -1.5987298488616943,
"logps/chosen": -550.2916870117188,
"logps/rejected": -589.5011596679688,
"loss": 0.3143,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.42545145750045776,
"rewards/margins": 0.08496164530515671,
"rewards/rejected": -0.5104131102561951,
"step": 170
},
{
"epoch": 0.04,
"learning_rate": 3.585657370517929e-06,
"logits/chosen": -2.2417616844177246,
"logits/rejected": -1.566329836845398,
"logps/chosen": -465.1465759277344,
"logps/rejected": -618.5088500976562,
"loss": 0.2378,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.28945469856262207,
"rewards/margins": 0.20885801315307617,
"rewards/rejected": -0.49831271171569824,
"step": 180
},
{
"epoch": 0.05,
"learning_rate": 3.7848605577689246e-06,
"logits/chosen": -2.0054221153259277,
"logits/rejected": -1.608533263206482,
"logps/chosen": -474.73468017578125,
"logps/rejected": -527.5888061523438,
"loss": 0.3241,
"rewards/accuracies": 0.5,
"rewards/chosen": -0.3396565020084381,
"rewards/margins": 0.0798039585351944,
"rewards/rejected": -0.4194604754447937,
"step": 190
},
{
"epoch": 0.05,
"learning_rate": 3.98406374501992e-06,
"logits/chosen": -2.1822266578674316,
"logits/rejected": -1.7768001556396484,
"logps/chosen": -381.39166259765625,
"logps/rejected": -489.23114013671875,
"loss": 0.2525,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.2405729591846466,
"rewards/margins": 0.13961976766586304,
"rewards/rejected": -0.38019272685050964,
"step": 200
},
{
"epoch": 0.05,
"learning_rate": 4.183266932270917e-06,
"logits/chosen": -2.0433428287506104,
"logits/rejected": -1.155823826789856,
"logps/chosen": -570.4938354492188,
"logps/rejected": -718.9401245117188,
"loss": 0.2815,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.4118042588233948,
"rewards/margins": 0.2083522379398346,
"rewards/rejected": -0.620156466960907,
"step": 210
},
{
"epoch": 0.05,
"learning_rate": 4.382470119521913e-06,
"logits/chosen": -2.0706868171691895,
"logits/rejected": -1.3509011268615723,
"logps/chosen": -564.2838745117188,
"logps/rejected": -744.248779296875,
"loss": 0.261,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.40354689955711365,
"rewards/margins": 0.22655579447746277,
"rewards/rejected": -0.6301027536392212,
"step": 220
},
{
"epoch": 0.06,
"learning_rate": 4.581673306772908e-06,
"logits/chosen": -2.1797804832458496,
"logits/rejected": -1.5733859539031982,
"logps/chosen": -329.7751159667969,
"logps/rejected": -467.71807861328125,
"loss": 0.2674,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.1671646535396576,
"rewards/margins": 0.19905708730220795,
"rewards/rejected": -0.36622172594070435,
"step": 230
},
{
"epoch": 0.06,
"learning_rate": 4.780876494023905e-06,
"logits/chosen": -2.288848876953125,
"logits/rejected": -1.6748888492584229,
"logps/chosen": -571.8269653320312,
"logps/rejected": -707.6842041015625,
"loss": 0.2649,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.4129388928413391,
"rewards/margins": 0.19371375441551208,
"rewards/rejected": -0.6066526174545288,
"step": 240
},
{
"epoch": 0.06,
"learning_rate": 4.980079681274901e-06,
"logits/chosen": -2.222977876663208,
"logits/rejected": -1.6339671611785889,
"logps/chosen": -457.69000244140625,
"logps/rejected": -580.9161987304688,
"loss": 0.2866,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.3078981339931488,
"rewards/margins": 0.1683819591999054,
"rewards/rejected": -0.4762801229953766,
"step": 250
},
{
"epoch": 0.06,
"learning_rate": 4.999802610509541e-06,
"logits/chosen": -2.17555570602417,
"logits/rejected": -1.641728401184082,
"logps/chosen": -571.9801025390625,
"logps/rejected": -752.2948608398438,
"loss": 0.2416,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.40048933029174805,
"rewards/margins": 0.21740679442882538,
"rewards/rejected": -0.6178960800170898,
"step": 260
},
{
"epoch": 0.06,
"learning_rate": 4.9991203164860365e-06,
"logits/chosen": -2.3382885456085205,
"logits/rejected": -1.802610158920288,
"logps/chosen": -461.989013671875,
"logps/rejected": -589.45166015625,
"loss": 0.2291,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.31522998213768005,
"rewards/margins": 0.16316869854927063,
"rewards/rejected": -0.4783986508846283,
"step": 270
},
{
"epoch": 0.07,
"learning_rate": 4.997950814005098e-06,
"logits/chosen": -2.5300729274749756,
"logits/rejected": -1.6759153604507446,
"logps/chosen": -527.8553466796875,
"logps/rejected": -689.076904296875,
"loss": 0.2481,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.318958580493927,
"rewards/margins": 0.24714262783527374,
"rewards/rejected": -0.5661011934280396,
"step": 280
},
{
"epoch": 0.07,
"learning_rate": 4.99629433106355e-06,
"logits/chosen": -2.233731746673584,
"logits/rejected": -1.6700172424316406,
"logps/chosen": -671.8790283203125,
"logps/rejected": -758.0708618164062,
"loss": 0.2699,
"rewards/accuracies": 0.550000011920929,
"rewards/chosen": -0.4771367013454437,
"rewards/margins": 0.15160521864891052,
"rewards/rejected": -0.628741979598999,
"step": 290
},
{
"epoch": 0.07,
"learning_rate": 4.994151190596025e-06,
"logits/chosen": -2.2926723957061768,
"logits/rejected": -1.6014070510864258,
"logps/chosen": -357.16583251953125,
"logps/rejected": -530.6362915039062,
"loss": 0.2773,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20203897356987,
"rewards/margins": 0.20360472798347473,
"rewards/rejected": -0.4056437015533447,
"step": 300
},
{
"epoch": 0.07,
"learning_rate": 4.9915218104120024e-06,
"logits/chosen": -2.1675281524658203,
"logits/rejected": -1.540875792503357,
"logps/chosen": -517.7607421875,
"logps/rejected": -669.5948486328125,
"loss": 0.2334,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.3637324273586273,
"rewards/margins": 0.20930609107017517,
"rewards/rejected": -0.5730385780334473,
"step": 310
},
{
"epoch": 0.08,
"learning_rate": 4.98840670311436e-06,
"logits/chosen": -2.3536510467529297,
"logits/rejected": -1.794704794883728,
"logps/chosen": -533.6524658203125,
"logps/rejected": -621.9884643554688,
"loss": 0.2721,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3811626136302948,
"rewards/margins": 0.13614344596862793,
"rewards/rejected": -0.5173059701919556,
"step": 320
},
{
"epoch": 0.08,
"learning_rate": 4.984806475999437e-06,
"logits/chosen": -2.2430691719055176,
"logits/rejected": -1.409332036972046,
"logps/chosen": -616.1778564453125,
"logps/rejected": -776.3978271484375,
"loss": 0.2444,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.42925572395324707,
"rewards/margins": 0.21815767884254456,
"rewards/rejected": -0.647413432598114,
"step": 330
},
{
"epoch": 0.08,
"learning_rate": 4.980721830938645e-06,
"logits/chosen": -2.1990833282470703,
"logits/rejected": -1.544798493385315,
"logps/chosen": -605.9384765625,
"logps/rejected": -732.4371337890625,
"loss": 0.2803,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.439105361700058,
"rewards/margins": 0.172859787940979,
"rewards/rejected": -0.6119651198387146,
"step": 340
},
{
"epoch": 0.08,
"learning_rate": 4.9761535642416284e-06,
"logits/chosen": -2.2408649921417236,
"logits/rejected": -1.799768090248108,
"logps/chosen": -489.03533935546875,
"logps/rejected": -665.1949462890625,
"loss": 0.2405,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.37104225158691406,
"rewards/margins": 0.19040416181087494,
"rewards/rejected": -0.5614464282989502,
"step": 350
},
{
"epoch": 0.09,
"learning_rate": 4.9711025665010335e-06,
"logits/chosen": -2.2574844360351562,
"logits/rejected": -1.8442182540893555,
"logps/chosen": -408.14654541015625,
"logps/rejected": -561.5911865234375,
"loss": 0.2777,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.269236296415329,
"rewards/margins": 0.16201291978359222,
"rewards/rejected": -0.4312492311000824,
"step": 360
},
{
"epoch": 0.09,
"learning_rate": 4.965569822418878e-06,
"logits/chosen": -2.0681309700012207,
"logits/rejected": -1.4381722211837769,
"logps/chosen": -559.8223876953125,
"logps/rejected": -734.8892211914062,
"loss": 0.2241,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.43346816301345825,
"rewards/margins": 0.1917620450258255,
"rewards/rejected": -0.6252301931381226,
"step": 370
},
{
"epoch": 0.09,
"learning_rate": 4.9595564106145825e-06,
"logits/chosen": -2.303969383239746,
"logits/rejected": -1.733120322227478,
"logps/chosen": -418.6346740722656,
"logps/rejected": -560.189208984375,
"loss": 0.2358,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.2271452397108078,
"rewards/margins": 0.19738546013832092,
"rewards/rejected": -0.42453068494796753,
"step": 380
},
{
"epoch": 0.09,
"learning_rate": 4.953063503414692e-06,
"logits/chosen": -2.2448253631591797,
"logits/rejected": -1.8518972396850586,
"logps/chosen": -494.76055908203125,
"logps/rejected": -628.48388671875,
"loss": 0.2679,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.335178941488266,
"rewards/margins": 0.17853963375091553,
"rewards/rejected": -0.5137186050415039,
"step": 390
},
{
"epoch": 0.1,
"learning_rate": 4.946092366624333e-06,
"logits/chosen": -2.2507317066192627,
"logits/rejected": -1.619484305381775,
"logps/chosen": -526.4837646484375,
"logps/rejected": -697.7025146484375,
"loss": 0.2529,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3731151223182678,
"rewards/margins": 0.21801939606666565,
"rewards/rejected": -0.5911344885826111,
"step": 400
},
{
"epoch": 0.1,
"learning_rate": 4.938644359280433e-06,
"logits/chosen": -2.375333070755005,
"logits/rejected": -1.6055479049682617,
"logps/chosen": -563.16552734375,
"logps/rejected": -741.4131469726562,
"loss": 0.212,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.36431047320365906,
"rewards/margins": 0.2534615397453308,
"rewards/rejected": -0.6177719831466675,
"step": 410
},
{
"epoch": 0.1,
"learning_rate": 4.930720933386782e-06,
"logits/chosen": -2.233098268508911,
"logits/rejected": -1.6022119522094727,
"logps/chosen": -426.771728515625,
"logps/rejected": -549.2544555664062,
"loss": 0.2503,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.2603939175605774,
"rewards/margins": 0.18492767214775085,
"rewards/rejected": -0.445321649312973,
"step": 420
},
{
"epoch": 0.1,
"learning_rate": 4.922323633630957e-06,
"logits/chosen": -2.1745898723602295,
"logits/rejected": -1.5291332006454468,
"logps/chosen": -516.7222290039062,
"logps/rejected": -726.829345703125,
"loss": 0.2005,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.3355748653411865,
"rewards/margins": 0.27592363953590393,
"rewards/rejected": -0.6114985346794128,
"step": 430
},
{
"epoch": 0.11,
"learning_rate": 4.913454097083185e-06,
"logits/chosen": -2.2320826053619385,
"logits/rejected": -1.747312307357788,
"logps/chosen": -674.8052368164062,
"logps/rejected": -752.8465576171875,
"loss": 0.2891,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.5011820793151855,
"rewards/margins": 0.12799863517284393,
"rewards/rejected": -0.6291807889938354,
"step": 440
},
{
"epoch": 0.11,
"learning_rate": 4.904114052877189e-06,
"logits/chosen": -2.121434211730957,
"logits/rejected": -1.4196488857269287,
"logps/chosen": -554.5970458984375,
"logps/rejected": -728.1629028320312,
"loss": 0.2619,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.39592963457107544,
"rewards/margins": 0.23849153518676758,
"rewards/rejected": -0.6344212293624878,
"step": 450
},
{
"epoch": 0.11,
"learning_rate": 4.894305321873092e-06,
"logits/chosen": -2.1138224601745605,
"logits/rejected": -1.6369972229003906,
"logps/chosen": -714.6239624023438,
"logps/rejected": -836.7276611328125,
"loss": 0.2451,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.5520918965339661,
"rewards/margins": 0.17846426367759705,
"rewards/rejected": -0.7305561304092407,
"step": 460
},
{
"epoch": 0.11,
"learning_rate": 4.884029816302441e-06,
"logits/chosen": -2.2457010746002197,
"logits/rejected": -1.7463334798812866,
"logps/chosen": -603.1580810546875,
"logps/rejected": -735.0208740234375,
"loss": 0.2874,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.4632132947444916,
"rewards/margins": 0.17302510142326355,
"rewards/rejected": -0.6362384557723999,
"step": 470
},
{
"epoch": 0.12,
"learning_rate": 4.873289539395404e-06,
"logits/chosen": -2.1492977142333984,
"logits/rejected": -1.6171681880950928,
"logps/chosen": -582.4373779296875,
"logps/rejected": -740.5318603515625,
"loss": 0.2167,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.42827337980270386,
"rewards/margins": 0.2127343863248825,
"rewards/rejected": -0.6410078406333923,
"step": 480
},
{
"epoch": 0.12,
"learning_rate": 4.862086584990246e-06,
"logits/chosen": -2.2339184284210205,
"logits/rejected": -1.7799112796783447,
"logps/chosen": -582.9187622070312,
"logps/rejected": -674.3992309570312,
"loss": 0.2649,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.3971683382987976,
"rewards/margins": 0.1706150472164154,
"rewards/rejected": -0.5677834153175354,
"step": 490
},
{
"epoch": 0.12,
"learning_rate": 4.850423137125126e-06,
"logits/chosen": -1.98675537109375,
"logits/rejected": -1.467492699623108,
"logps/chosen": -641.2584228515625,
"logps/rejected": -804.790771484375,
"loss": 0.2572,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.4882410168647766,
"rewards/margins": 0.21470656991004944,
"rewards/rejected": -0.7029476165771484,
"step": 500
},
{
"epoch": 0.12,
"learning_rate": 4.838301469612315e-06,
"logits/chosen": -2.24493408203125,
"logits/rejected": -1.498622179031372,
"logps/chosen": -557.9942626953125,
"logps/rejected": -705.88623046875,
"loss": 0.2415,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.3977532982826233,
"rewards/margins": 0.20435115694999695,
"rewards/rejected": -0.6021044850349426,
"step": 510
},
{
"epoch": 0.12,
"learning_rate": 4.825723945594912e-06,
"logits/chosen": -2.1898727416992188,
"logits/rejected": -1.737510323524475,
"logps/chosen": -415.2431640625,
"logps/rejected": -563.3753051757812,
"loss": 0.285,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.27554893493652344,
"rewards/margins": 0.1814078390598297,
"rewards/rejected": -0.4569567143917084,
"step": 520
},
{
"epoch": 0.13,
"learning_rate": 4.812693017086145e-06,
"logits/chosen": -2.2709362506866455,
"logits/rejected": -1.7534446716308594,
"logps/chosen": -521.100341796875,
"logps/rejected": -715.1611938476562,
"loss": 0.2564,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.3482024073600769,
"rewards/margins": 0.23160001635551453,
"rewards/rejected": -0.5798024535179138,
"step": 530
},
{
"epoch": 0.13,
"learning_rate": 4.799211224491348e-06,
"logits/chosen": -2.000866651535034,
"logits/rejected": -1.438919186592102,
"logps/chosen": -606.0504760742188,
"logps/rejected": -769.3602294921875,
"loss": 0.2518,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.5148371458053589,
"rewards/margins": 0.1749372035264969,
"rewards/rejected": -0.6897743344306946,
"step": 540
},
{
"epoch": 0.13,
"learning_rate": 4.7852811961126974e-06,
"logits/chosen": -2.2116503715515137,
"logits/rejected": -1.419684648513794,
"logps/chosen": -566.8517456054688,
"logps/rejected": -797.03076171875,
"loss": 0.2348,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.42510828375816345,
"rewards/margins": 0.28039130568504333,
"rewards/rejected": -0.7054997086524963,
"step": 550
},
{
"epoch": 0.13,
"learning_rate": 4.770905647636828e-06,
"logits/chosen": -2.19708514213562,
"logits/rejected": -1.514418601989746,
"logps/chosen": -372.2975769042969,
"logps/rejected": -572.015625,
"loss": 0.2529,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.22482657432556152,
"rewards/margins": 0.2627830505371094,
"rewards/rejected": -0.4876096844673157,
"step": 560
},
{
"epoch": 0.14,
"learning_rate": 4.756087381605399e-06,
"logits/chosen": -2.212829113006592,
"logits/rejected": -1.5567867755889893,
"logps/chosen": -451.62091064453125,
"logps/rejected": -675.35498046875,
"loss": 0.2243,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.28890347480773926,
"rewards/margins": 0.2642940580844879,
"rewards/rejected": -0.5531975030899048,
"step": 570
},
{
"epoch": 0.14,
"learning_rate": 4.740829286868732e-06,
"logits/chosen": -2.0651936531066895,
"logits/rejected": -1.282173991203308,
"logps/chosen": -608.8034057617188,
"logps/rejected": -814.4363403320312,
"loss": 0.2309,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.4570806920528412,
"rewards/margins": 0.2720951437950134,
"rewards/rejected": -0.729175865650177,
"step": 580
},
{
"epoch": 0.14,
"learning_rate": 4.725134338022631e-06,
"logits/chosen": -1.976910948753357,
"logits/rejected": -1.3392863273620605,
"logps/chosen": -673.251220703125,
"logps/rejected": -844.18408203125,
"loss": 0.2425,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.4962734580039978,
"rewards/margins": 0.2366379052400589,
"rewards/rejected": -0.7329114675521851,
"step": 590
},
{
"epoch": 0.14,
"learning_rate": 4.709005594828471e-06,
"logits/chosen": -2.177516460418701,
"logits/rejected": -1.4188392162322998,
"logps/chosen": -349.5711669921875,
"logps/rejected": -489.1685485839844,
"loss": 0.268,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.18546082079410553,
"rewards/margins": 0.21840114891529083,
"rewards/rejected": -0.4038619101047516,
"step": 600
},
{
"epoch": 0.15,
"learning_rate": 4.692446201616692e-06,
"logits/chosen": -2.171480894088745,
"logits/rejected": -1.486372470855713,
"logps/chosen": -489.7513732910156,
"logps/rejected": -671.7752685546875,
"loss": 0.2265,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.3359985947608948,
"rewards/margins": 0.2379292994737625,
"rewards/rejected": -0.5739278197288513,
"step": 610
},
{
"epoch": 0.15,
"learning_rate": 4.675459386673815e-06,
"logits/chosen": -2.188599109649658,
"logits/rejected": -1.417234182357788,
"logps/chosen": -669.4722290039062,
"logps/rejected": -851.4942626953125,
"loss": 0.2386,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.4645751118659973,
"rewards/margins": 0.2636135220527649,
"rewards/rejected": -0.7281886339187622,
"step": 620
},
{
"epoch": 0.15,
"learning_rate": 4.658048461613068e-06,
"logits/chosen": -2.186264753341675,
"logits/rejected": -1.3317312002182007,
"logps/chosen": -558.1051635742188,
"logps/rejected": -807.9237060546875,
"loss": 0.2096,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.4056881070137024,
"rewards/margins": 0.2980197072029114,
"rewards/rejected": -0.7037078738212585,
"step": 630
},
{
"epoch": 0.15,
"learning_rate": 4.640216820728791e-06,
"logits/chosen": -2.0650200843811035,
"logits/rejected": -1.5434041023254395,
"logps/chosen": -591.6378173828125,
"logps/rejected": -726.2674560546875,
"loss": 0.3041,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.440895140171051,
"rewards/margins": 0.15948018431663513,
"rewards/rejected": -0.6003752946853638,
"step": 640
},
{
"epoch": 0.16,
"learning_rate": 4.621967940334705e-06,
"logits/chosen": -2.1776533126831055,
"logits/rejected": -1.505947470664978,
"logps/chosen": -533.233154296875,
"logps/rejected": -758.1294555664062,
"loss": 0.2585,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.34916678071022034,
"rewards/margins": 0.2798806130886078,
"rewards/rejected": -0.6290473341941833,
"step": 650
},
{
"epoch": 0.16,
"learning_rate": 4.603305378086201e-06,
"logits/chosen": -1.815281629562378,
"logits/rejected": -1.1680887937545776,
"logps/chosen": -623.07275390625,
"logps/rejected": -776.311279296875,
"loss": 0.2435,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.4895709156990051,
"rewards/margins": 0.18619278073310852,
"rewards/rejected": -0.675763726234436,
"step": 660
},
{
"epoch": 0.16,
"learning_rate": 4.584232772286769e-06,
"logits/chosen": -1.9976768493652344,
"logits/rejected": -1.2251824140548706,
"logps/chosen": -644.8419189453125,
"logps/rejected": -831.3948364257812,
"loss": 0.2896,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.4800949990749359,
"rewards/margins": 0.2344970703125,
"rewards/rejected": -0.7145919799804688,
"step": 670
},
{
"epoch": 0.16,
"learning_rate": 4.5647538411786965e-06,
"logits/chosen": -2.1248276233673096,
"logits/rejected": -1.5544915199279785,
"logps/chosen": -530.0264282226562,
"logps/rejected": -631.1822509765625,
"loss": 0.2546,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.36536192893981934,
"rewards/margins": 0.15929196774959564,
"rewards/rejected": -0.5246539115905762,
"step": 680
},
{
"epoch": 0.17,
"learning_rate": 4.544872382218202e-06,
"logits/chosen": -2.3693809509277344,
"logits/rejected": -1.6609468460083008,
"logps/chosen": -424.1357421875,
"logps/rejected": -568.0796508789062,
"loss": 0.2299,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.257046639919281,
"rewards/margins": 0.19673588871955872,
"rewards/rejected": -0.45378249883651733,
"step": 690
},
{
"epoch": 0.17,
"learning_rate": 4.5245922713351e-06,
"logits/chosen": -2.135702610015869,
"logits/rejected": -1.4124939441680908,
"logps/chosen": -638.5548095703125,
"logps/rejected": -831.3570556640625,
"loss": 0.2649,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.46429723501205444,
"rewards/margins": 0.2442464828491211,
"rewards/rejected": -0.7085437774658203,
"step": 700
},
{
"epoch": 0.17,
"learning_rate": 4.503917462177192e-06,
"logits/chosen": -2.142075300216675,
"logits/rejected": -1.5568357706069946,
"logps/chosen": -572.1000366210938,
"logps/rejected": -753.2780151367188,
"loss": 0.292,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.44795316457748413,
"rewards/margins": 0.20818133652210236,
"rewards/rejected": -0.6561344861984253,
"step": 710
},
{
"epoch": 0.17,
"learning_rate": 4.482851985339487e-06,
"logits/chosen": -2.1371376514434814,
"logits/rejected": -1.4664158821105957,
"logps/chosen": -545.5193481445312,
"logps/rejected": -744.4154052734375,
"loss": 0.239,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3687899112701416,
"rewards/margins": 0.2610931992530823,
"rewards/rejected": -0.6298831701278687,
"step": 720
},
{
"epoch": 0.18,
"learning_rate": 4.461399947578434e-06,
"logits/chosen": -2.133291721343994,
"logits/rejected": -1.604524850845337,
"logps/chosen": -517.1326293945312,
"logps/rejected": -649.8372802734375,
"loss": 0.231,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3689945638179779,
"rewards/margins": 0.1885087788105011,
"rewards/rejected": -0.5575034022331238,
"step": 730
},
{
"epoch": 0.18,
"learning_rate": 4.439565531011299e-06,
"logits/chosen": -1.8887875080108643,
"logits/rejected": -1.270674467086792,
"logps/chosen": -613.1268310546875,
"logps/rejected": -801.4429321289062,
"loss": 0.2473,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.4629307687282562,
"rewards/margins": 0.22329919040203094,
"rewards/rejected": -0.6862298846244812,
"step": 740
},
{
"epoch": 0.18,
"learning_rate": 4.417352992300854e-06,
"logits/chosen": -2.3292813301086426,
"logits/rejected": -1.6490875482559204,
"logps/chosen": -425.7843322753906,
"logps/rejected": -597.09814453125,
"loss": 0.2591,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.2748299241065979,
"rewards/margins": 0.2304181158542633,
"rewards/rejected": -0.5052480101585388,
"step": 750
},
{
"epoch": 0.18,
"learning_rate": 4.3947666618255335e-06,
"logits/chosen": -2.14275860786438,
"logits/rejected": -1.552585244178772,
"logps/chosen": -411.56024169921875,
"logps/rejected": -597.5908813476562,
"loss": 0.2509,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2517482042312622,
"rewards/margins": 0.2175951898097992,
"rewards/rejected": -0.4693434238433838,
"step": 760
},
{
"epoch": 0.18,
"learning_rate": 4.3718109428352155e-06,
"logits/chosen": -2.094531536102295,
"logits/rejected": -1.4484500885009766,
"logps/chosen": -561.7736206054688,
"logps/rejected": -734.1261596679688,
"loss": 0.2477,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.4113832116127014,
"rewards/margins": 0.2228744924068451,
"rewards/rejected": -0.6342577338218689,
"step": 770
},
{
"epoch": 0.19,
"learning_rate": 4.348490310592801e-06,
"logits/chosen": -1.8707062005996704,
"logits/rejected": -1.506484866142273,
"logps/chosen": -714.3576049804688,
"logps/rejected": -866.392578125,
"loss": 0.2457,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.5910379886627197,
"rewards/margins": 0.16303986310958862,
"rewards/rejected": -0.7540777921676636,
"step": 780
},
{
"epoch": 0.19,
"learning_rate": 4.3248093115017544e-06,
"logits/chosen": -2.125237464904785,
"logits/rejected": -1.4492751359939575,
"logps/chosen": -582.9632568359375,
"logps/rejected": -715.1638793945312,
"loss": 0.2496,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.3824686110019684,
"rewards/margins": 0.22166451811790466,
"rewards/rejected": -0.6041331887245178,
"step": 790
},
{
"epoch": 0.19,
"learning_rate": 4.3007725622197675e-06,
"logits/chosen": -2.2058329582214355,
"logits/rejected": -1.5816318988800049,
"logps/chosen": -570.4851684570312,
"logps/rejected": -719.0206909179688,
"loss": 0.227,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.38881629705429077,
"rewards/margins": 0.21982832252979279,
"rewards/rejected": -0.6086446046829224,
"step": 800
},
{
"epoch": 0.19,
"learning_rate": 4.27638474875874e-06,
"logits/chosen": -2.1510353088378906,
"logits/rejected": -1.3720935583114624,
"logps/chosen": -377.4300537109375,
"logps/rejected": -588.8758544921875,
"loss": 0.2153,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.23323938250541687,
"rewards/margins": 0.264992356300354,
"rewards/rejected": -0.4982317090034485,
"step": 810
},
{
"epoch": 0.2,
"learning_rate": 4.25165062557123e-06,
"logits/chosen": -2.1816329956054688,
"logits/rejected": -1.862624168395996,
"logps/chosen": -489.541015625,
"logps/rejected": -573.84814453125,
"loss": 0.2876,
"rewards/accuracies": 0.625,
"rewards/chosen": -0.33602797985076904,
"rewards/margins": 0.1257353574037552,
"rewards/rejected": -0.46176332235336304,
"step": 820
},
{
"epoch": 0.2,
"learning_rate": 4.226575014623557e-06,
"logits/chosen": -2.121556520462036,
"logits/rejected": -1.500382661819458,
"logps/chosen": -569.2263793945312,
"logps/rejected": -745.4718017578125,
"loss": 0.2127,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.37383171916007996,
"rewards/margins": 0.24332182109355927,
"rewards/rejected": -0.617153525352478,
"step": 830
},
{
"epoch": 0.2,
"learning_rate": 4.201162804455764e-06,
"logits/chosen": -2.158127784729004,
"logits/rejected": -1.5402179956436157,
"logps/chosen": -477.3233337402344,
"logps/rejected": -729.618408203125,
"loss": 0.2182,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.336378812789917,
"rewards/margins": 0.27327030897140503,
"rewards/rejected": -0.609649121761322,
"step": 840
},
{
"epoch": 0.2,
"learning_rate": 4.175418949228571e-06,
"logits/chosen": -2.0661299228668213,
"logits/rejected": -1.3702499866485596,
"logps/chosen": -421.61676025390625,
"logps/rejected": -608.9027709960938,
"loss": 0.2314,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.28428909182548523,
"rewards/margins": 0.23253202438354492,
"rewards/rejected": -0.5168210864067078,
"step": 850
},
{
"epoch": 0.21,
"learning_rate": 4.149348467757566e-06,
"logits/chosen": -2.2505486011505127,
"logits/rejected": -1.6682631969451904,
"logps/chosen": -519.7850341796875,
"logps/rejected": -675.0706787109375,
"loss": 0.2582,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.40990179777145386,
"rewards/margins": 0.17124588787555695,
"rewards/rejected": -0.5811477303504944,
"step": 860
},
{
"epoch": 0.21,
"learning_rate": 4.122956442534765e-06,
"logits/chosen": -2.070268154144287,
"logits/rejected": -1.589167833328247,
"logps/chosen": -569.9495239257812,
"logps/rejected": -677.8106689453125,
"loss": 0.2566,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.4036158621311188,
"rewards/margins": 0.1584577113389969,
"rewards/rejected": -0.5620735883712769,
"step": 870
},
{
"epoch": 0.21,
"learning_rate": 4.096248018737781e-06,
"logits/chosen": -1.9538896083831787,
"logits/rejected": -1.506744623184204,
"logps/chosen": -598.489990234375,
"logps/rejected": -718.8297729492188,
"loss": 0.2701,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.4573654234409332,
"rewards/margins": 0.15690357983112335,
"rewards/rejected": -0.614268958568573,
"step": 880
},
{
"epoch": 0.21,
"learning_rate": 4.069228403226751e-06,
"logits/chosen": -2.2838289737701416,
"logits/rejected": -1.5676209926605225,
"logps/chosen": -536.0494384765625,
"logps/rejected": -721.7430419921875,
"loss": 0.2196,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.37890490889549255,
"rewards/margins": 0.2322990447282791,
"rewards/rejected": -0.6112040281295776,
"step": 890
},
{
"epoch": 0.22,
"learning_rate": 4.041902863529257e-06,
"logits/chosen": -2.2310588359832764,
"logits/rejected": -1.7672712802886963,
"logps/chosen": -479.8710021972656,
"logps/rejected": -604.5220947265625,
"loss": 0.2552,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.3082687258720398,
"rewards/margins": 0.17145316302776337,
"rewards/rejected": -0.47972187399864197,
"step": 900
},
{
"epoch": 0.22,
"learning_rate": 4.014276726813404e-06,
"logits/chosen": -2.1795742511749268,
"logits/rejected": -1.4126112461090088,
"logps/chosen": -480.098876953125,
"logps/rejected": -701.5238037109375,
"loss": 0.2253,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.32372376322746277,
"rewards/margins": 0.2762491703033447,
"rewards/rejected": -0.5999729037284851,
"step": 910
},
{
"epoch": 0.22,
"learning_rate": 3.986355378849284e-06,
"logits/chosen": -2.3270981311798096,
"logits/rejected": -2.0575973987579346,
"logps/chosen": -562.2911376953125,
"logps/rejected": -678.2664794921875,
"loss": 0.255,
"rewards/accuracies": 0.550000011920929,
"rewards/chosen": -0.39332714676856995,
"rewards/margins": 0.1472180336713791,
"rewards/rejected": -0.5405451655387878,
"step": 920
},
{
"epoch": 0.22,
"learning_rate": 3.958144262959004e-06,
"logits/chosen": -2.204744815826416,
"logits/rejected": -1.6555640697479248,
"logps/chosen": -619.665283203125,
"logps/rejected": -708.1978759765625,
"loss": 0.4841,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.48071688413619995,
"rewards/margins": 0.11553032696247101,
"rewards/rejected": -0.5962471961975098,
"step": 930
},
{
"epoch": 0.23,
"learning_rate": 3.929648878955507e-06,
"logits/chosen": -2.1643776893615723,
"logits/rejected": -1.7133562564849854,
"logps/chosen": -498.3556213378906,
"logps/rejected": -624.0385131835938,
"loss": 0.2892,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.3023318946361542,
"rewards/margins": 0.18938633799552917,
"rewards/rejected": -0.49171820282936096,
"step": 940
},
{
"epoch": 0.23,
"learning_rate": 3.900874782070362e-06,
"logits/chosen": -2.2862777709960938,
"logits/rejected": -1.5697494745254517,
"logps/chosen": -523.8767700195312,
"logps/rejected": -712.3031005859375,
"loss": 0.2906,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.35215550661087036,
"rewards/margins": 0.26922211050987244,
"rewards/rejected": -0.6213775873184204,
"step": 950
},
{
"epoch": 0.23,
"learning_rate": 3.871827581870772e-06,
"logits/chosen": -1.9373928308486938,
"logits/rejected": -1.2519636154174805,
"logps/chosen": -612.5736083984375,
"logps/rejected": -769.370361328125,
"loss": 0.2162,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.44734010100364685,
"rewards/margins": 0.20580251514911652,
"rewards/rejected": -0.6531426906585693,
"step": 960
},
{
"epoch": 0.23,
"learning_rate": 3.842512941165968e-06,
"logits/chosen": -1.9552310705184937,
"logits/rejected": -1.2225010395050049,
"logps/chosen": -595.6505126953125,
"logps/rejected": -807.1351318359375,
"loss": 0.2339,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.4424334168434143,
"rewards/margins": 0.2755126357078552,
"rewards/rejected": -0.7179459929466248,
"step": 970
},
{
"epoch": 0.24,
"learning_rate": 3.8129365749032398e-06,
"logits/chosen": -2.179598331451416,
"logits/rejected": -1.5930414199829102,
"logps/chosen": -470.39666748046875,
"logps/rejected": -708.7127075195312,
"loss": 0.2389,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3327002227306366,
"rewards/margins": 0.247235506772995,
"rewards/rejected": -0.5799357295036316,
"step": 980
},
{
"epoch": 0.24,
"learning_rate": 3.783104249053793e-06,
"logits/chosen": -1.876232385635376,
"logits/rejected": -1.1954014301300049,
"logps/chosen": -482.48809814453125,
"logps/rejected": -653.3897705078125,
"loss": 0.2342,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.3456036448478699,
"rewards/margins": 0.20723696053028107,
"rewards/rejected": -0.5528405904769897,
"step": 990
},
{
"epoch": 0.24,
"learning_rate": 3.7530217794886607e-06,
"logits/chosen": -2.168401002883911,
"logits/rejected": -1.362849473953247,
"logps/chosen": -565.8056030273438,
"logps/rejected": -753.4979248046875,
"loss": 0.2181,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3963169455528259,
"rewards/margins": 0.2578599452972412,
"rewards/rejected": -0.6541768908500671,
"step": 1000
},
{
"epoch": 0.24,
"learning_rate": 3.722695030844891e-06,
"logits/chosen": -1.9112541675567627,
"logits/rejected": -1.4224046468734741,
"logps/chosen": -536.9829711914062,
"logps/rejected": -687.1131591796875,
"loss": 0.2746,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.41709670424461365,
"rewards/margins": 0.18000951409339905,
"rewards/rejected": -0.5971062183380127,
"step": 1010
},
{
"epoch": 0.24,
"learning_rate": 3.6921299153822198e-06,
"logits/chosen": -2.2641046047210693,
"logits/rejected": -1.6229751110076904,
"logps/chosen": -529.2923583984375,
"logps/rejected": -726.6102905273438,
"loss": 0.228,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.34389495849609375,
"rewards/margins": 0.2508707046508789,
"rewards/rejected": -0.5947656035423279,
"step": 1020
},
{
"epoch": 0.25,
"learning_rate": 3.66133239183047e-06,
"logits/chosen": -2.1011762619018555,
"logits/rejected": -1.4804975986480713,
"logps/chosen": -495.906494140625,
"logps/rejected": -683.6201171875,
"loss": 0.2441,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.3397464454174042,
"rewards/margins": 0.23895248770713806,
"rewards/rejected": -0.5786989331245422,
"step": 1030
},
{
"epoch": 0.25,
"learning_rate": 3.630308464227877e-06,
"logits/chosen": -2.093890428543091,
"logits/rejected": -1.4402543306350708,
"logps/chosen": -405.9877014160156,
"logps/rejected": -528.1480712890625,
"loss": 0.2503,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20128004252910614,
"rewards/margins": 0.19632327556610107,
"rewards/rejected": -0.3976033329963684,
"step": 1040
},
{
"epoch": 0.25,
"learning_rate": 3.5990641807506e-06,
"logits/chosen": -2.3380658626556396,
"logits/rejected": -1.6520893573760986,
"logps/chosen": -430.703369140625,
"logps/rejected": -646.5703735351562,
"loss": 0.2577,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.27441808581352234,
"rewards/margins": 0.26531141996383667,
"rewards/rejected": -0.5397294759750366,
"step": 1050
},
{
"epoch": 0.25,
"learning_rate": 3.5676056325336084e-06,
"logits/chosen": -2.1818904876708984,
"logits/rejected": -1.784393310546875,
"logps/chosen": -520.1884155273438,
"logps/rejected": -674.11962890625,
"loss": 0.2316,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.3729000687599182,
"rewards/margins": 0.16966374218463898,
"rewards/rejected": -0.5425638556480408,
"step": 1060
},
{
"epoch": 0.26,
"learning_rate": 3.535938952483211e-06,
"logits/chosen": -2.2213704586029053,
"logits/rejected": -1.418710708618164,
"logps/chosen": -560.0448608398438,
"logps/rejected": -815.6527709960938,
"loss": 0.2124,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.40334635972976685,
"rewards/margins": 0.2945135533809662,
"rewards/rejected": -0.6978598833084106,
"step": 1070
},
{
"epoch": 0.26,
"learning_rate": 3.5040703140814254e-06,
"logits/chosen": -2.221543788909912,
"logits/rejected": -1.6487659215927124,
"logps/chosen": -634.6618041992188,
"logps/rejected": -839.3564453125,
"loss": 0.2324,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.4737940728664398,
"rewards/margins": 0.24561241269111633,
"rewards/rejected": -0.7194064855575562,
"step": 1080
},
{
"epoch": 0.26,
"learning_rate": 3.4720059301824527e-06,
"logits/chosen": -2.189147710800171,
"logits/rejected": -1.7664272785186768,
"logps/chosen": -526.344970703125,
"logps/rejected": -650.4627075195312,
"loss": 0.2229,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.36051493883132935,
"rewards/margins": 0.19068099558353424,
"rewards/rejected": -0.5511959791183472,
"step": 1090
},
{
"epoch": 0.26,
"learning_rate": 3.439752051801467e-06,
"logits/chosen": -2.276291608810425,
"logits/rejected": -1.5065466165542603,
"logps/chosen": -466.7132263183594,
"logps/rejected": -688.0767822265625,
"loss": 0.2285,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.31362882256507874,
"rewards/margins": 0.2555429935455322,
"rewards/rejected": -0.5691717863082886,
"step": 1100
},
{
"epoch": 0.27,
"learning_rate": 3.407314966895966e-06,
"logits/chosen": -2.2229745388031006,
"logits/rejected": -1.6873805522918701,
"logps/chosen": -499.5079650878906,
"logps/rejected": -678.7784423828125,
"loss": 0.248,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.34916746616363525,
"rewards/margins": 0.20373527705669403,
"rewards/rejected": -0.5529027581214905,
"step": 1110
},
{
"epoch": 0.27,
"learning_rate": 3.3747009991399226e-06,
"logits/chosen": -2.0915369987487793,
"logits/rejected": -1.515972375869751,
"logps/chosen": -462.2139587402344,
"logps/rejected": -621.6764526367188,
"loss": 0.2303,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.34132781624794006,
"rewards/margins": 0.19509322941303253,
"rewards/rejected": -0.536421000957489,
"step": 1120
},
{
"epoch": 0.27,
"learning_rate": 3.341916506690971e-06,
"logits/chosen": -2.1414732933044434,
"logits/rejected": -1.5188627243041992,
"logps/chosen": -487.9198303222656,
"logps/rejected": -658.6336059570312,
"loss": 0.2358,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3202676773071289,
"rewards/margins": 0.2106558084487915,
"rewards/rejected": -0.5309234857559204,
"step": 1130
},
{
"epoch": 0.27,
"learning_rate": 3.308967880950874e-06,
"logits/chosen": -2.0529282093048096,
"logits/rejected": -1.4871912002563477,
"logps/chosen": -542.2448120117188,
"logps/rejected": -726.5451049804688,
"loss": 0.2379,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.41947245597839355,
"rewards/margins": 0.209178164601326,
"rewards/rejected": -0.6286506056785583,
"step": 1140
},
{
"epoch": 0.28,
"learning_rate": 3.275861545319504e-06,
"logits/chosen": -2.241400718688965,
"logits/rejected": -1.7179415225982666,
"logps/chosen": -475.1095275878906,
"logps/rejected": -658.5631103515625,
"loss": 0.254,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.31501632928848267,
"rewards/margins": 0.21485765278339386,
"rewards/rejected": -0.5298739671707153,
"step": 1150
},
{
"epoch": 0.28,
"learning_rate": 3.2426039539425875e-06,
"logits/chosen": -2.236320972442627,
"logits/rejected": -1.3771488666534424,
"logps/chosen": -501.77960205078125,
"logps/rejected": -705.9559326171875,
"loss": 0.2485,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.33146998286247253,
"rewards/margins": 0.2820337116718292,
"rewards/rejected": -0.613503634929657,
"step": 1160
},
{
"epoch": 0.28,
"learning_rate": 3.2092015904534614e-06,
"logits/chosen": -2.1324212551116943,
"logits/rejected": -1.3361310958862305,
"logps/chosen": -465.21234130859375,
"logps/rejected": -626.5828857421875,
"loss": 0.222,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.30111175775527954,
"rewards/margins": 0.2380938082933426,
"rewards/rejected": -0.5392054915428162,
"step": 1170
},
{
"epoch": 0.28,
"learning_rate": 3.17566096670907e-06,
"logits/chosen": -2.1174681186676025,
"logits/rejected": -1.8057496547698975,
"logps/chosen": -514.4780883789062,
"logps/rejected": -612.9591064453125,
"loss": 0.2773,
"rewards/accuracies": 0.625,
"rewards/chosen": -0.3825177252292633,
"rewards/margins": 0.13299870491027832,
"rewards/rejected": -0.5155164003372192,
"step": 1180
},
{
"epoch": 0.29,
"learning_rate": 3.14198862152047e-06,
"logits/chosen": -2.3264975547790527,
"logits/rejected": -1.8200445175170898,
"logps/chosen": -509.34332275390625,
"logps/rejected": -662.3193359375,
"loss": 0.2653,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3371141850948334,
"rewards/margins": 0.19747108221054077,
"rewards/rejected": -0.5345852971076965,
"step": 1190
},
{
"epoch": 0.29,
"learning_rate": 3.1081911193780734e-06,
"logits/chosen": -2.1914255619049072,
"logits/rejected": -1.6171748638153076,
"logps/chosen": -607.6774291992188,
"logps/rejected": -785.1390991210938,
"loss": 0.2038,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.4608604907989502,
"rewards/margins": 0.20961742103099823,
"rewards/rejected": -0.6704779267311096,
"step": 1200
},
{
"epoch": 0.29,
"learning_rate": 3.074275049171889e-06,
"logits/chosen": -2.2372031211853027,
"logits/rejected": -1.4800232648849487,
"logps/chosen": -483.6255798339844,
"logps/rejected": -684.0525512695312,
"loss": 0.2085,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3340616226196289,
"rewards/margins": 0.26013877987861633,
"rewards/rejected": -0.5942003726959229,
"step": 1210
},
{
"epoch": 0.29,
"learning_rate": 3.0402470229070057e-06,
"logits/chosen": -2.0542514324188232,
"logits/rejected": -1.2669802904129028,
"logps/chosen": -501.363037109375,
"logps/rejected": -668.2828979492188,
"loss": 0.2264,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.32397252321243286,
"rewards/margins": 0.25711455941200256,
"rewards/rejected": -0.5810869932174683,
"step": 1220
},
{
"epoch": 0.3,
"learning_rate": 3.006113674414565e-06,
"logits/chosen": -2.124558925628662,
"logits/rejected": -1.5115940570831299,
"logps/chosen": -526.3839721679688,
"logps/rejected": -727.1844482421875,
"loss": 0.2239,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.3876585364341736,
"rewards/margins": 0.2264205515384674,
"rewards/rejected": -0.6140791177749634,
"step": 1230
},
{
"epoch": 0.3,
"learning_rate": 2.9718816580584885e-06,
"logits/chosen": -2.330988645553589,
"logits/rejected": -1.7894699573516846,
"logps/chosen": -450.5931701660156,
"logps/rejected": -639.8587036132812,
"loss": 0.2195,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.300833135843277,
"rewards/margins": 0.2428937703371048,
"rewards/rejected": -0.5437268614768982,
"step": 1240
},
{
"epoch": 0.3,
"learning_rate": 2.9375576474381907e-06,
"logits/chosen": -2.09294056892395,
"logits/rejected": -1.5990426540374756,
"logps/chosen": -523.2689208984375,
"logps/rejected": -714.5638427734375,
"loss": 0.2505,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.36846724152565,
"rewards/margins": 0.24714866280555725,
"rewards/rejected": -0.6156159043312073,
"step": 1250
},
{
"epoch": 0.3,
"learning_rate": 2.9031483340875523e-06,
"logits/chosen": -2.161059617996216,
"logits/rejected": -1.649753212928772,
"logps/chosen": -595.0687255859375,
"logps/rejected": -741.806396484375,
"loss": 0.2356,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.42591404914855957,
"rewards/margins": 0.19824771583080292,
"rewards/rejected": -0.6241617202758789,
"step": 1260
},
{
"epoch": 0.3,
"learning_rate": 2.868660426170388e-06,
"logits/chosen": -2.0810506343841553,
"logits/rejected": -1.3984508514404297,
"logps/chosen": -519.595458984375,
"logps/rejected": -709.3812255859375,
"loss": 0.2182,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.37194812297821045,
"rewards/margins": 0.21791160106658936,
"rewards/rejected": -0.5898597836494446,
"step": 1270
},
{
"epoch": 0.31,
"learning_rate": 2.8341006471726817e-06,
"logits/chosen": -1.937578558921814,
"logits/rejected": -1.3747153282165527,
"logps/chosen": -516.6207885742188,
"logps/rejected": -695.8414306640625,
"loss": 0.2206,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3487696647644043,
"rewards/margins": 0.2523016929626465,
"rewards/rejected": -0.6010713577270508,
"step": 1280
},
{
"epoch": 0.31,
"learning_rate": 2.7994757345918244e-06,
"logits/chosen": -2.414841890335083,
"logits/rejected": -1.7823702096939087,
"logps/chosen": -453.10357666015625,
"logps/rejected": -609.0830078125,
"loss": 0.2223,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.29341357946395874,
"rewards/margins": 0.1968296319246292,
"rewards/rejected": -0.49024319648742676,
"step": 1290
},
{
"epoch": 0.31,
"learning_rate": 2.76479243862313e-06,
"logits/chosen": -2.3372766971588135,
"logits/rejected": -1.365880012512207,
"logps/chosen": -455.2708435058594,
"logps/rejected": -712.4964599609375,
"loss": 0.2234,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.29985883831977844,
"rewards/margins": 0.31078290939331055,
"rewards/rejected": -0.6106417775154114,
"step": 1300
},
{
"epoch": 0.31,
"learning_rate": 2.7300575208438684e-06,
"logits/chosen": -2.193761110305786,
"logits/rejected": -1.7236446142196655,
"logps/chosen": -438.6455078125,
"logps/rejected": -565.2931518554688,
"loss": 0.2211,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.28494128584861755,
"rewards/margins": 0.18237076699733734,
"rewards/rejected": -0.4673120379447937,
"step": 1310
},
{
"epoch": 0.32,
"learning_rate": 2.695277752895084e-06,
"logits/chosen": -2.1613688468933105,
"logits/rejected": -1.6745634078979492,
"logps/chosen": -445.5487365722656,
"logps/rejected": -583.8851318359375,
"loss": 0.2158,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.30961447954177856,
"rewards/margins": 0.17475193738937378,
"rewards/rejected": -0.48436641693115234,
"step": 1320
},
{
"epoch": 0.32,
"learning_rate": 2.6604599151614514e-06,
"logits/chosen": -2.3241307735443115,
"logits/rejected": -1.718146562576294,
"logps/chosen": -472.390380859375,
"logps/rejected": -670.6461791992188,
"loss": 0.2484,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.30457672476768494,
"rewards/margins": 0.2522759437561035,
"rewards/rejected": -0.5568526983261108,
"step": 1330
},
{
"epoch": 0.32,
"learning_rate": 2.625610795449424e-06,
"logits/chosen": -2.118846893310547,
"logits/rejected": -1.4407285451889038,
"logps/chosen": -462.6249084472656,
"logps/rejected": -711.793212890625,
"loss": 0.2517,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.32376161217689514,
"rewards/margins": 0.27978068590164185,
"rewards/rejected": -0.6035423278808594,
"step": 1340
},
{
"epoch": 0.32,
"learning_rate": 2.59073718766394e-06,
"logits/chosen": -2.1999683380126953,
"logits/rejected": -1.4900109767913818,
"logps/chosen": -476.880859375,
"logps/rejected": -675.8462524414062,
"loss": 0.2466,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.3132093548774719,
"rewards/margins": 0.25780683755874634,
"rewards/rejected": -0.571016252040863,
"step": 1350
},
{
"epoch": 0.33,
"learning_rate": 2.5558458904839345e-06,
"logits/chosen": -2.192030191421509,
"logits/rejected": -1.7015453577041626,
"logps/chosen": -497.59429931640625,
"logps/rejected": -690.6241455078125,
"loss": 0.2367,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3215945065021515,
"rewards/margins": 0.2261447012424469,
"rewards/rejected": -0.5477392673492432,
"step": 1360
},
{
"epoch": 0.33,
"learning_rate": 2.5209437060369266e-06,
"logits/chosen": -2.1700994968414307,
"logits/rejected": -1.501390814781189,
"logps/chosen": -551.3470458984375,
"logps/rejected": -779.6153564453125,
"loss": 0.2324,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.39861801266670227,
"rewards/margins": 0.272796094417572,
"rewards/rejected": -0.6714141368865967,
"step": 1370
},
{
"epoch": 0.33,
"learning_rate": 2.4860374385729298e-06,
"logits/chosen": -2.3210222721099854,
"logits/rejected": -1.647146224975586,
"logps/chosen": -461.9217224121094,
"logps/rejected": -660.2238159179688,
"loss": 0.2164,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.30266645550727844,
"rewards/margins": 0.2639373540878296,
"rewards/rejected": -0.5666038990020752,
"step": 1380
},
{
"epoch": 0.33,
"learning_rate": 2.4511338931379475e-06,
"logits/chosen": -2.270345687866211,
"logits/rejected": -1.7757459878921509,
"logps/chosen": -534.710693359375,
"logps/rejected": -685.7799072265625,
"loss": 0.1999,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3926665186882019,
"rewards/margins": 0.20033708214759827,
"rewards/rejected": -0.5930036306381226,
"step": 1390
},
{
"epoch": 0.34,
"learning_rate": 2.4162398742473216e-06,
"logits/chosen": -2.436246395111084,
"logits/rejected": -1.9106261730194092,
"logps/chosen": -395.49920654296875,
"logps/rejected": -564.9219970703125,
"loss": 0.2506,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.19471552968025208,
"rewards/margins": 0.22500737011432648,
"rewards/rejected": -0.41972288489341736,
"step": 1400
},
{
"epoch": 0.34,
"learning_rate": 2.381362184559173e-06,
"logits/chosen": -2.3186497688293457,
"logits/rejected": -1.7353124618530273,
"logps/chosen": -545.6030883789062,
"logps/rejected": -732.5777587890625,
"loss": 0.2682,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.37879544496536255,
"rewards/margins": 0.24574975669384003,
"rewards/rejected": -0.6245452165603638,
"step": 1410
},
{
"epoch": 0.34,
"learning_rate": 2.3465076235482117e-06,
"logits/chosen": -2.099091053009033,
"logits/rejected": -1.251534104347229,
"logps/chosen": -591.8385009765625,
"logps/rejected": -823.31201171875,
"loss": 0.2263,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3871614336967468,
"rewards/margins": 0.3113124966621399,
"rewards/rejected": -0.6984738707542419,
"step": 1420
},
{
"epoch": 0.34,
"learning_rate": 2.3116829861801687e-06,
"logits/chosen": -2.2782912254333496,
"logits/rejected": -1.6930453777313232,
"logps/chosen": -549.5288696289062,
"logps/rejected": -727.2321166992188,
"loss": 0.244,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.4014182984828949,
"rewards/margins": 0.21473488211631775,
"rewards/rejected": -0.6161531209945679,
"step": 1430
},
{
"epoch": 0.35,
"learning_rate": 2.276895061587099e-06,
"logits/chosen": -2.3278706073760986,
"logits/rejected": -1.6391801834106445,
"logps/chosen": -570.4884643554688,
"logps/rejected": -818.1508178710938,
"loss": 0.1976,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.40508347749710083,
"rewards/margins": 0.28309813141822815,
"rewards/rejected": -0.6881815791130066,
"step": 1440
},
{
"epoch": 0.35,
"learning_rate": 2.242150631743832e-06,
"logits/chosen": -2.3615341186523438,
"logits/rejected": -1.725358009338379,
"logps/chosen": -478.43603515625,
"logps/rejected": -650.2882080078125,
"loss": 0.2222,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.31791216135025024,
"rewards/margins": 0.2205561101436615,
"rewards/rejected": -0.5384682416915894,
"step": 1450
},
{
"epoch": 0.35,
"learning_rate": 2.207456470145807e-06,
"logits/chosen": -2.35868239402771,
"logits/rejected": -1.982661247253418,
"logps/chosen": -456.50439453125,
"logps/rejected": -552.9801635742188,
"loss": 0.291,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.29267337918281555,
"rewards/margins": 0.14309945702552795,
"rewards/rejected": -0.4357728064060211,
"step": 1460
},
{
"epoch": 0.35,
"learning_rate": 2.17281934048857e-06,
"logits/chosen": -2.1577529907226562,
"logits/rejected": -1.2531640529632568,
"logps/chosen": -470.220947265625,
"logps/rejected": -691.4623413085938,
"loss": 0.2374,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.2939762473106384,
"rewards/margins": 0.2891116142272949,
"rewards/rejected": -0.5830878615379333,
"step": 1470
},
{
"epoch": 0.36,
"learning_rate": 2.1382459953491773e-06,
"logits/chosen": -2.153989315032959,
"logits/rejected": -1.712892770767212,
"logps/chosen": -509.06103515625,
"logps/rejected": -677.8247680664062,
"loss": 0.2359,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.3281121850013733,
"rewards/margins": 0.20805349946022034,
"rewards/rejected": -0.5361656546592712,
"step": 1480
},
{
"epoch": 0.36,
"learning_rate": 2.103743174869769e-06,
"logits/chosen": -1.9424854516983032,
"logits/rejected": -1.422446370124817,
"logps/chosen": -697.8045654296875,
"logps/rejected": -815.5999755859375,
"loss": 0.2559,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.5232383012771606,
"rewards/margins": 0.18479886651039124,
"rewards/rejected": -0.7080371379852295,
"step": 1490
},
{
"epoch": 0.36,
"learning_rate": 2.0693176054435586e-06,
"logits/chosen": -2.266700506210327,
"logits/rejected": -1.5664805173873901,
"logps/chosen": -565.6947021484375,
"logps/rejected": -702.6282348632812,
"loss": 0.2324,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.3801000118255615,
"rewards/margins": 0.20919163525104523,
"rewards/rejected": -0.5892916321754456,
"step": 1500
},
{
"epoch": 0.36,
"learning_rate": 2.034975998403517e-06,
"logits/chosen": -2.202317953109741,
"logits/rejected": -1.178363561630249,
"logps/chosen": -471.96514892578125,
"logps/rejected": -716.80322265625,
"loss": 0.2517,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.29619717597961426,
"rewards/margins": 0.32016804814338684,
"rewards/rejected": -0.6163652539253235,
"step": 1510
},
{
"epoch": 0.36,
"learning_rate": 2.0007250487139827e-06,
"logits/chosen": -2.1976194381713867,
"logits/rejected": -1.7854173183441162,
"logps/chosen": -409.1556396484375,
"logps/rejected": -538.7450561523438,
"loss": 0.2324,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.2653912305831909,
"rewards/margins": 0.16031914949417114,
"rewards/rejected": -0.42571038007736206,
"step": 1520
},
{
"epoch": 0.37,
"learning_rate": 1.9665714336654604e-06,
"logits/chosen": -2.2585816383361816,
"logits/rejected": -1.4220424890518188,
"logps/chosen": -477.648681640625,
"logps/rejected": -702.324462890625,
"loss": 0.2246,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.31048738956451416,
"rewards/margins": 0.29150474071502686,
"rewards/rejected": -0.6019921898841858,
"step": 1530
},
{
"epoch": 0.37,
"learning_rate": 1.9325218115728756e-06,
"logits/chosen": -2.259849786758423,
"logits/rejected": -1.4995836019515991,
"logps/chosen": -520.2024536132812,
"logps/rejected": -719.3076782226562,
"loss": 0.2124,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.3540114760398865,
"rewards/margins": 0.2679198682308197,
"rewards/rejected": -0.6219313740730286,
"step": 1540
},
{
"epoch": 0.37,
"learning_rate": 1.8985828204775206e-06,
"logits/chosen": -2.2383124828338623,
"logits/rejected": -1.5292937755584717,
"logps/chosen": -460.6908264160156,
"logps/rejected": -634.41552734375,
"loss": 0.1988,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.31010597944259644,
"rewards/margins": 0.2322625368833542,
"rewards/rejected": -0.5423685312271118,
"step": 1550
},
{
"epoch": 0.37,
"learning_rate": 1.8647610768529581e-06,
"logits/chosen": -2.3701467514038086,
"logits/rejected": -1.669344186782837,
"logps/chosen": -515.9610595703125,
"logps/rejected": -702.8071899414062,
"loss": 0.2465,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3375261425971985,
"rewards/margins": 0.22942647337913513,
"rewards/rejected": -0.566952645778656,
"step": 1560
},
{
"epoch": 0.38,
"learning_rate": 1.8310631743151187e-06,
"logits/chosen": -2.3348028659820557,
"logits/rejected": -1.6338012218475342,
"logps/chosen": -478.33428955078125,
"logps/rejected": -730.6288452148438,
"loss": 0.2324,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.33109721541404724,
"rewards/margins": 0.2853388786315918,
"rewards/rejected": -0.6164361238479614,
"step": 1570
},
{
"epoch": 0.38,
"learning_rate": 1.7974956823368728e-06,
"logits/chosen": -2.156118631362915,
"logits/rejected": -1.6093488931655884,
"logps/chosen": -582.0015258789062,
"logps/rejected": -735.3878784179688,
"loss": 0.2101,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.4120170474052429,
"rewards/margins": 0.207681804895401,
"rewards/rejected": -0.6196987628936768,
"step": 1580
},
{
"epoch": 0.38,
"learning_rate": 1.7640651449672913e-06,
"logits/chosen": -2.3183345794677734,
"logits/rejected": -1.6045589447021484,
"logps/chosen": -474.8514709472656,
"logps/rejected": -684.0841064453125,
"loss": 0.2061,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.28908270597457886,
"rewards/margins": 0.27025845646858215,
"rewards/rejected": -0.5593411326408386,
"step": 1590
},
{
"epoch": 0.38,
"learning_rate": 1.7307780795558743e-06,
"logits/chosen": -2.1290550231933594,
"logits/rejected": -1.5637315511703491,
"logps/chosen": -484.809814453125,
"logps/rejected": -644.54052734375,
"loss": 0.2349,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3229495882987976,
"rewards/margins": 0.21575050055980682,
"rewards/rejected": -0.5387001633644104,
"step": 1600
},
{
"epoch": 0.39,
"learning_rate": 1.6976409754819767e-06,
"logits/chosen": -2.2557132244110107,
"logits/rejected": -1.787325143814087,
"logps/chosen": -505.43115234375,
"logps/rejected": -684.2761840820312,
"loss": 0.2373,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.3303944170475006,
"rewards/margins": 0.22589488327503204,
"rewards/rejected": -0.5562892556190491,
"step": 1610
},
{
"epoch": 0.39,
"learning_rate": 1.6646602928896962e-06,
"logits/chosen": -2.1125569343566895,
"logits/rejected": -1.5178847312927246,
"logps/chosen": -604.7052612304688,
"logps/rejected": -748.1799926757812,
"loss": 0.2035,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.4366299510002136,
"rewards/margins": 0.2077966034412384,
"rewards/rejected": -0.6444265246391296,
"step": 1620
},
{
"epoch": 0.39,
"learning_rate": 1.6318424614284525e-06,
"logits/chosen": -2.0228872299194336,
"logits/rejected": -1.6469089984893799,
"logps/chosen": -619.0574951171875,
"logps/rejected": -747.111328125,
"loss": 0.2531,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.4670810103416443,
"rewards/margins": 0.15774844586849213,
"rewards/rejected": -0.6248295307159424,
"step": 1630
},
{
"epoch": 0.39,
"learning_rate": 1.5991938789995138e-06,
"logits/chosen": -2.178020477294922,
"logits/rejected": -1.616796851158142,
"logps/chosen": -625.3834838867188,
"logps/rejected": -798.6873168945312,
"loss": 0.2581,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.46300649642944336,
"rewards/margins": 0.23273572325706482,
"rewards/rejected": -0.6957422494888306,
"step": 1640
},
{
"epoch": 0.4,
"learning_rate": 1.5667209105087134e-06,
"logits/chosen": -2.1452012062072754,
"logits/rejected": -1.4882639646530151,
"logps/chosen": -637.6602783203125,
"logps/rejected": -835.0618896484375,
"loss": 0.2314,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.4847962260246277,
"rewards/margins": 0.23071709275245667,
"rewards/rejected": -0.7155133485794067,
"step": 1650
},
{
"epoch": 0.4,
"learning_rate": 1.5344298866256002e-06,
"logits/chosen": -2.1361632347106934,
"logits/rejected": -1.393336296081543,
"logps/chosen": -598.9241333007812,
"logps/rejected": -825.7525634765625,
"loss": 0.2541,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.4569925367832184,
"rewards/margins": 0.26417863368988037,
"rewards/rejected": -0.7211712002754211,
"step": 1660
},
{
"epoch": 0.4,
"learning_rate": 1.502327102549262e-06,
"logits/chosen": -2.1626908779144287,
"logits/rejected": -1.562538743019104,
"logps/chosen": -468.4820251464844,
"logps/rejected": -660.3966064453125,
"loss": 0.204,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.3353859782218933,
"rewards/margins": 0.2233034074306488,
"rewards/rejected": -0.5586894154548645,
"step": 1670
},
{
"epoch": 0.4,
"learning_rate": 1.4704188167810635e-06,
"logits/chosen": -2.2278897762298584,
"logits/rejected": -1.6712696552276611,
"logps/chosen": -513.513916015625,
"logps/rejected": -695.208984375,
"loss": 0.2159,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.362549364566803,
"rewards/margins": 0.23459453880786896,
"rewards/rejected": -0.5971439480781555,
"step": 1680
},
{
"epoch": 0.41,
"learning_rate": 1.438711249904536e-06,
"logits/chosen": -2.2296204566955566,
"logits/rejected": -1.5326087474822998,
"logps/chosen": -495.2484436035156,
"logps/rejected": -728.9078369140625,
"loss": 0.2267,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3783426582813263,
"rewards/margins": 0.26764681935310364,
"rewards/rejected": -0.6459894180297852,
"step": 1690
},
{
"epoch": 0.41,
"learning_rate": 1.4072105833726685e-06,
"logits/chosen": -2.308741569519043,
"logits/rejected": -1.57771897315979,
"logps/chosen": -545.0296630859375,
"logps/rejected": -739.1373291015625,
"loss": 0.2682,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.3485683798789978,
"rewards/margins": 0.25783371925354004,
"rewards/rejected": -0.6064020991325378,
"step": 1700
},
{
"epoch": 0.41,
"learning_rate": 1.375922958302815e-06,
"logits/chosen": -2.177499532699585,
"logits/rejected": -1.6153056621551514,
"logps/chosen": -575.4890747070312,
"logps/rejected": -719.1101684570312,
"loss": 0.257,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.4291258454322815,
"rewards/margins": 0.1913391649723053,
"rewards/rejected": -0.6204649209976196,
"step": 1710
},
{
"epoch": 0.41,
"learning_rate": 1.3448544742794792e-06,
"logits/chosen": -2.359710931777954,
"logits/rejected": -1.8718398809432983,
"logps/chosen": -518.01318359375,
"logps/rejected": -663.861083984375,
"loss": 0.2177,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.39127427339553833,
"rewards/margins": 0.15787231922149658,
"rewards/rejected": -0.5491466522216797,
"step": 1720
},
{
"epoch": 0.42,
"learning_rate": 1.3140111881651773e-06,
"logits/chosen": -1.9541170597076416,
"logits/rejected": -1.3082023859024048,
"logps/chosen": -529.1683349609375,
"logps/rejected": -741.0734252929688,
"loss": 0.219,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.3982730805873871,
"rewards/margins": 0.24865877628326416,
"rewards/rejected": -0.6469318866729736,
"step": 1730
},
{
"epoch": 0.42,
"learning_rate": 1.2833991129196508e-06,
"logits/chosen": -2.241741418838501,
"logits/rejected": -1.4685299396514893,
"logps/chosen": -470.96209716796875,
"logps/rejected": -709.0288696289062,
"loss": 0.216,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.3294115662574768,
"rewards/margins": 0.2778538763523102,
"rewards/rejected": -0.6072654724121094,
"step": 1740
},
{
"epoch": 0.42,
"learning_rate": 1.2530242164276236e-06,
"logits/chosen": -2.0970499515533447,
"logits/rejected": -1.3942879438400269,
"logps/chosen": -486.5403747558594,
"logps/rejected": -718.138671875,
"loss": 0.2207,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.31325727701187134,
"rewards/margins": 0.3117991089820862,
"rewards/rejected": -0.6250563859939575,
"step": 1750
},
{
"epoch": 0.42,
"learning_rate": 1.2228924203353507e-06,
"logits/chosen": -2.068192720413208,
"logits/rejected": -1.585180401802063,
"logps/chosen": -533.9257202148438,
"logps/rejected": -647.5416870117188,
"loss": 0.2262,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.390627920627594,
"rewards/margins": 0.16129513084888458,
"rewards/rejected": -0.5519230365753174,
"step": 1760
},
{
"epoch": 0.42,
"learning_rate": 1.1930095988961837e-06,
"logits/chosen": -2.2689132690429688,
"logits/rejected": -1.6284101009368896,
"logps/chosen": -489.23468017578125,
"logps/rejected": -698.836181640625,
"loss": 0.2103,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.31348201632499695,
"rewards/margins": 0.2569728493690491,
"rewards/rejected": -0.5704549551010132,
"step": 1770
},
{
"epoch": 0.43,
"learning_rate": 1.1633815778253721e-06,
"logits/chosen": -2.223635673522949,
"logits/rejected": -1.5692812204360962,
"logps/chosen": -564.4153442382812,
"logps/rejected": -758.5940551757812,
"loss": 0.2683,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.3882458806037903,
"rewards/margins": 0.24454763531684875,
"rewards/rejected": -0.6327935457229614,
"step": 1780
},
{
"epoch": 0.43,
"learning_rate": 1.1340141331643276e-06,
"logits/chosen": -2.226630926132202,
"logits/rejected": -1.6583023071289062,
"logps/chosen": -499.40155029296875,
"logps/rejected": -738.2061767578125,
"loss": 0.2232,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.3272217810153961,
"rewards/margins": 0.2891238331794739,
"rewards/rejected": -0.6163456439971924,
"step": 1790
},
{
"epoch": 0.43,
"learning_rate": 1.1049129901545756e-06,
"logits/chosen": -2.193066358566284,
"logits/rejected": -1.6954580545425415,
"logps/chosen": -497.5428161621094,
"logps/rejected": -683.050048828125,
"loss": 0.2054,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.3609580397605896,
"rewards/margins": 0.21754273772239685,
"rewards/rejected": -0.5785007476806641,
"step": 1800
},
{
"epoch": 0.43,
"learning_rate": 1.0760838221216065e-06,
"logits/chosen": -2.311552047729492,
"logits/rejected": -1.7207205295562744,
"logps/chosen": -450.60504150390625,
"logps/rejected": -628.2011108398438,
"loss": 0.2457,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.25561460852622986,
"rewards/margins": 0.24946312606334686,
"rewards/rejected": -0.5050776600837708,
"step": 1810
},
{
"epoch": 0.44,
"learning_rate": 1.0475322493688506e-06,
"logits/chosen": -2.1966607570648193,
"logits/rejected": -1.5143920183181763,
"logps/chosen": -420.51885986328125,
"logps/rejected": -698.9932250976562,
"loss": 0.2089,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.26709431409835815,
"rewards/margins": 0.31104663014411926,
"rewards/rejected": -0.578140914440155,
"step": 1820
},
{
"epoch": 0.44,
"learning_rate": 1.0192638380819884e-06,
"logits/chosen": -2.3707401752471924,
"logits/rejected": -1.5773122310638428,
"logps/chosen": -453.8907165527344,
"logps/rejected": -664.8104248046875,
"loss": 0.2373,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2851904630661011,
"rewards/margins": 0.2743096947669983,
"rewards/rejected": -0.5595001578330994,
"step": 1830
},
{
"epoch": 0.44,
"learning_rate": 9.912840992438087e-07,
"logits/chosen": -2.176928758621216,
"logits/rejected": -1.5890274047851562,
"logps/chosen": -573.779541015625,
"logps/rejected": -796.2474975585938,
"loss": 0.2305,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.4253336787223816,
"rewards/margins": 0.26153475046157837,
"rewards/rejected": -0.6868684887886047,
"step": 1840
},
{
"epoch": 0.44,
"learning_rate": 9.63598487559839e-07,
"logits/chosen": -2.2372374534606934,
"logits/rejected": -1.546007752418518,
"logps/chosen": -464.493408203125,
"logps/rejected": -669.3597412109375,
"loss": 0.208,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.30372193455696106,
"rewards/margins": 0.278639018535614,
"rewards/rejected": -0.5823609828948975,
"step": 1850
},
{
"epoch": 0.45,
"learning_rate": 9.362124003949324e-07,
"logits/chosen": -2.1051459312438965,
"logits/rejected": -1.6941722631454468,
"logps/chosen": -516.5181884765625,
"logps/rejected": -681.1585693359375,
"loss": 0.2474,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.36375877261161804,
"rewards/margins": 0.18948772549629211,
"rewards/rejected": -0.5532464981079102,
"step": 1860
},
{
"epoch": 0.45,
"learning_rate": 9.091311767210453e-07,
"logits/chosen": -2.1879124641418457,
"logits/rejected": -1.6842693090438843,
"logps/chosen": -510.8837890625,
"logps/rejected": -665.6920166015625,
"loss": 0.2358,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.36585649847984314,
"rewards/margins": 0.18732208013534546,
"rewards/rejected": -0.553178608417511,
"step": 1870
},
{
"epoch": 0.45,
"learning_rate": 8.823600960763901e-07,
"logits/chosen": -2.130765199661255,
"logits/rejected": -1.5883699655532837,
"logps/chosen": -534.0338745117188,
"logps/rejected": -741.09423828125,
"loss": 0.2077,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.38128194212913513,
"rewards/margins": 0.25697097182273865,
"rewards/rejected": -0.6382529139518738,
"step": 1880
},
{
"epoch": 0.45,
"learning_rate": 8.559043775361816e-07,
"logits/chosen": -2.1295900344848633,
"logits/rejected": -1.5722177028656006,
"logps/chosen": -551.2114868164062,
"logps/rejected": -690.095947265625,
"loss": 0.2276,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3628248870372772,
"rewards/margins": 0.2208392173051834,
"rewards/rejected": -0.5836641192436218,
"step": 1890
},
{
"epoch": 0.46,
"learning_rate": 8.297691786951706e-07,
"logits/chosen": -2.266829013824463,
"logits/rejected": -1.4247468709945679,
"logps/chosen": -547.1033935546875,
"logps/rejected": -776.6325073242188,
"loss": 0.2145,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.35183387994766235,
"rewards/margins": 0.31444767117500305,
"rewards/rejected": -0.666281521320343,
"step": 1900
},
{
"epoch": 0.46,
"learning_rate": 8.039595946621551e-07,
"logits/chosen": -2.2304577827453613,
"logits/rejected": -1.3978160619735718,
"logps/chosen": -538.714599609375,
"logps/rejected": -793.1671142578125,
"loss": 0.2318,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.39072176814079285,
"rewards/margins": 0.2959004342556,
"rewards/rejected": -0.6866222620010376,
"step": 1910
},
{
"epoch": 0.46,
"learning_rate": 7.784806570666795e-07,
"logits/chosen": -2.147185802459717,
"logits/rejected": -1.5393598079681396,
"logps/chosen": -481.76373291015625,
"logps/rejected": -670.560791015625,
"loss": 0.2545,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.3331444263458252,
"rewards/margins": 0.2202250212430954,
"rewards/rejected": -0.553369402885437,
"step": 1920
},
{
"epoch": 0.46,
"learning_rate": 7.533373330781127e-07,
"logits/chosen": -2.355670928955078,
"logits/rejected": -1.5707635879516602,
"logps/chosen": -561.0064697265625,
"logps/rejected": -760.4944458007812,
"loss": 0.2233,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.3677171766757965,
"rewards/margins": 0.2857670485973358,
"rewards/rejected": -0.6534842252731323,
"step": 1930
},
{
"epoch": 0.47,
"learning_rate": 7.285345244372843e-07,
"logits/chosen": -2.2503583431243896,
"logits/rejected": -1.3964335918426514,
"logps/chosen": -503.65777587890625,
"logps/rejected": -756.98828125,
"loss": 0.1685,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.33578115701675415,
"rewards/margins": 0.312565416097641,
"rewards/rejected": -0.6483466029167175,
"step": 1940
},
{
"epoch": 0.47,
"learning_rate": 7.040770665008853e-07,
"logits/chosen": -2.2794625759124756,
"logits/rejected": -1.6694520711898804,
"logps/chosen": -546.9520263671875,
"logps/rejected": -695.8685302734375,
"loss": 0.2262,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.3904178738594055,
"rewards/margins": 0.2141040861606598,
"rewards/rejected": -0.6045219302177429,
"step": 1950
},
{
"epoch": 0.47,
"learning_rate": 6.799697272987976e-07,
"logits/chosen": -2.1750683784484863,
"logits/rejected": -1.3738349676132202,
"logps/chosen": -540.0596923828125,
"logps/rejected": -722.267578125,
"loss": 0.2455,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3699941039085388,
"rewards/margins": 0.2555733621120453,
"rewards/rejected": -0.6255674958229065,
"step": 1960
},
{
"epoch": 0.47,
"learning_rate": 6.562172066045655e-07,
"logits/chosen": -2.167599678039551,
"logits/rejected": -1.6329807043075562,
"logps/chosen": -441.4002380371094,
"logps/rejected": -602.337646484375,
"loss": 0.2318,
"rewards/accuracies": 0.6000000238418579,
"rewards/chosen": -0.31705886125564575,
"rewards/margins": 0.21183231472969055,
"rewards/rejected": -0.5288912057876587,
"step": 1970
},
{
"epoch": 0.48,
"learning_rate": 6.328241350191619e-07,
"logits/chosen": -2.2226500511169434,
"logits/rejected": -1.537512183189392,
"logps/chosen": -480.624755859375,
"logps/rejected": -686.3212890625,
"loss": 0.2092,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.30479687452316284,
"rewards/margins": 0.2621714472770691,
"rewards/rejected": -0.5669684410095215,
"step": 1980
},
{
"epoch": 0.48,
"learning_rate": 6.097950730682426e-07,
"logits/chosen": -2.176600456237793,
"logits/rejected": -1.5454185009002686,
"logps/chosen": -503.0887756347656,
"logps/rejected": -687.1947021484375,
"loss": 0.2479,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3592723608016968,
"rewards/margins": 0.23496529459953308,
"rewards/rejected": -0.5942376255989075,
"step": 1990
},
{
"epoch": 0.48,
"learning_rate": 5.871345103130646e-07,
"logits/chosen": -2.087590217590332,
"logits/rejected": -1.4282548427581787,
"logps/chosen": -597.8372802734375,
"logps/rejected": -783.7901611328125,
"loss": 0.2177,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.44616231322288513,
"rewards/margins": 0.21901002526283264,
"rewards/rejected": -0.665172278881073,
"step": 2000
},
{
"epoch": 0.48,
"learning_rate": 5.64846864475237e-07,
"logits/chosen": -2.1877074241638184,
"logits/rejected": -1.913442611694336,
"logps/chosen": -512.3074951171875,
"logps/rejected": -631.2486572265625,
"loss": 0.2673,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.35744622349739075,
"rewards/margins": 0.16726166009902954,
"rewards/rejected": -0.5247078537940979,
"step": 2010
},
{
"epoch": 0.48,
"learning_rate": 5.429364805754758e-07,
"logits/chosen": -2.0919928550720215,
"logits/rejected": -1.5754165649414062,
"logps/chosen": -502.09405517578125,
"logps/rejected": -650.5390014648438,
"loss": 0.2226,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.3595220446586609,
"rewards/margins": 0.19984325766563416,
"rewards/rejected": -0.5593653321266174,
"step": 2020
},
{
"epoch": 0.49,
"learning_rate": 5.214076300865359e-07,
"logits/chosen": -1.9238027334213257,
"logits/rejected": -1.2277315855026245,
"logps/chosen": -580.2691650390625,
"logps/rejected": -834.8424072265625,
"loss": 0.2114,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.4341210722923279,
"rewards/margins": 0.30377626419067383,
"rewards/rejected": -0.7378972768783569,
"step": 2030
},
{
"epoch": 0.49,
"learning_rate": 5.002645101004766e-07,
"logits/chosen": -2.335980176925659,
"logits/rejected": -1.380081057548523,
"logps/chosen": -504.81024169921875,
"logps/rejected": -794.1038208007812,
"loss": 0.1986,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.3392624855041504,
"rewards/margins": 0.3661222457885742,
"rewards/rejected": -0.7053847908973694,
"step": 2040
},
{
"epoch": 0.49,
"learning_rate": 4.795112425104323e-07,
"logits/chosen": -2.2008166313171387,
"logits/rejected": -1.5865943431854248,
"logps/chosen": -581.0762939453125,
"logps/rejected": -769.32177734375,
"loss": 0.2619,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.4084292948246002,
"rewards/margins": 0.25177350640296936,
"rewards/rejected": -0.6602028012275696,
"step": 2050
},
{
"epoch": 0.49,
"learning_rate": 4.591518732070402e-07,
"logits/chosen": -1.9570486545562744,
"logits/rejected": -1.4179704189300537,
"logps/chosen": -534.230224609375,
"logps/rejected": -729.9030151367188,
"loss": 0.2396,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3860064148902893,
"rewards/margins": 0.24045896530151367,
"rewards/rejected": -0.6264654397964478,
"step": 2060
},
{
"epoch": 0.5,
"learning_rate": 4.391903712896861e-07,
"logits/chosen": -2.1633055210113525,
"logits/rejected": -1.5320649147033691,
"logps/chosen": -577.50439453125,
"logps/rejected": -789.090087890625,
"loss": 0.242,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -0.4344731271266937,
"rewards/margins": 0.2326418161392212,
"rewards/rejected": -0.6671148538589478,
"step": 2070
},
{
"epoch": 0.5,
"learning_rate": 4.196306282927187e-07,
"logits/chosen": -2.198305130004883,
"logits/rejected": -1.6743977069854736,
"logps/chosen": -516.3706665039062,
"logps/rejected": -713.3460693359375,
"loss": 0.2211,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3524255156517029,
"rewards/margins": 0.25238287448883057,
"rewards/rejected": -0.6048084497451782,
"step": 2080
},
{
"epoch": 0.5,
"learning_rate": 4.0047645742679275e-07,
"logits/chosen": -2.15580415725708,
"logits/rejected": -1.6634889841079712,
"logps/chosen": -539.8309326171875,
"logps/rejected": -740.0956420898438,
"loss": 0.218,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.36082297563552856,
"rewards/margins": 0.23072440922260284,
"rewards/rejected": -0.5915473103523254,
"step": 2090
},
{
"epoch": 0.5,
"learning_rate": 3.817315928354695e-07,
"logits/chosen": -2.187629222869873,
"logits/rejected": -1.617875099182129,
"logps/chosen": -528.2720336914062,
"logps/rejected": -703.1224365234375,
"loss": 0.2308,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.36435794830322266,
"rewards/margins": 0.23637576401233673,
"rewards/rejected": -0.6007336378097534,
"step": 2100
},
{
"epoch": 0.51,
"learning_rate": 3.633996888672428e-07,
"logits/chosen": -2.1398205757141113,
"logits/rejected": -1.48526132106781,
"logps/chosen": -582.8170166015625,
"logps/rejected": -740.50390625,
"loss": 0.206,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.4385349154472351,
"rewards/margins": 0.2139698714017868,
"rewards/rejected": -0.6525048017501831,
"step": 2110
},
{
"epoch": 0.51,
"learning_rate": 3.4548431936311275e-07,
"logits/chosen": -2.3196969032287598,
"logits/rejected": -1.9413297176361084,
"logps/chosen": -515.4064331054688,
"logps/rejected": -631.7554931640625,
"loss": 0.24,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3508912920951843,
"rewards/margins": 0.1308077871799469,
"rewards/rejected": -0.4816990792751312,
"step": 2120
},
{
"epoch": 0.51,
"learning_rate": 3.2798897695986155e-07,
"logits/chosen": -2.225336790084839,
"logits/rejected": -1.4455909729003906,
"logps/chosen": -562.00439453125,
"logps/rejected": -792.2058715820312,
"loss": 0.2268,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.3739345073699951,
"rewards/margins": 0.29291829466819763,
"rewards/rejected": -0.6668527722358704,
"step": 2130
},
{
"epoch": 0.51,
"learning_rate": 3.1091707240915704e-07,
"logits/chosen": -2.318101167678833,
"logits/rejected": -1.5058709383010864,
"logps/chosen": -528.3710327148438,
"logps/rejected": -773.945068359375,
"loss": 0.2322,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3821602761745453,
"rewards/margins": 0.2865941524505615,
"rewards/rejected": -0.6687543988227844,
"step": 2140
},
{
"epoch": 0.52,
"learning_rate": 2.942719339126171e-07,
"logits/chosen": -2.114053249359131,
"logits/rejected": -1.4541417360305786,
"logps/chosen": -575.40625,
"logps/rejected": -743.8985595703125,
"loss": 0.2631,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.4056618809700012,
"rewards/margins": 0.24168558418750763,
"rewards/rejected": -0.6473473906517029,
"step": 2150
},
{
"epoch": 0.52,
"learning_rate": 2.780568064729716e-07,
"logits/chosen": -2.1276280879974365,
"logits/rejected": -1.5844396352767944,
"logps/chosen": -556.7276000976562,
"logps/rejected": -722.495849609375,
"loss": 0.232,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.40463319420814514,
"rewards/margins": 0.21296298503875732,
"rewards/rejected": -0.6175961494445801,
"step": 2160
},
{
"epoch": 0.52,
"learning_rate": 2.622748512614437e-07,
"logits/chosen": -2.328648328781128,
"logits/rejected": -1.7766300439834595,
"logps/chosen": -508.13250732421875,
"logps/rejected": -652.4500122070312,
"loss": 0.2499,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.35677647590637207,
"rewards/margins": 0.18978366255760193,
"rewards/rejected": -0.5465601682662964,
"step": 2170
},
{
"epoch": 0.52,
"learning_rate": 2.4692914500147185e-07,
"logits/chosen": -2.1308367252349854,
"logits/rejected": -1.7297455072402954,
"logps/chosen": -539.8509521484375,
"logps/rejected": -701.5316162109375,
"loss": 0.2199,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.37622612714767456,
"rewards/margins": 0.193180650472641,
"rewards/rejected": -0.5694067478179932,
"step": 2180
},
{
"epoch": 0.53,
"learning_rate": 2.320226793688979e-07,
"logits/chosen": -2.176285982131958,
"logits/rejected": -1.585010290145874,
"logps/chosen": -597.8599853515625,
"logps/rejected": -753.0974731445312,
"loss": 0.2317,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.4297494888305664,
"rewards/margins": 0.20314817130565643,
"rewards/rejected": -0.632897675037384,
"step": 2190
},
{
"epoch": 0.53,
"learning_rate": 2.1755836040873197e-07,
"logits/chosen": -2.09965181350708,
"logits/rejected": -1.5018677711486816,
"logps/chosen": -551.79541015625,
"logps/rejected": -750.493408203125,
"loss": 0.2168,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.4179263114929199,
"rewards/margins": 0.248566672205925,
"rewards/rejected": -0.6664929389953613,
"step": 2200
},
{
"epoch": 0.53,
"learning_rate": 2.0353900796861503e-07,
"logits/chosen": -2.2555556297302246,
"logits/rejected": -1.8257776498794556,
"logps/chosen": -505.36663818359375,
"logps/rejected": -631.3004760742188,
"loss": 0.2511,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.35016122460365295,
"rewards/margins": 0.18649618327617645,
"rewards/rejected": -0.5366573929786682,
"step": 2210
},
{
"epoch": 0.53,
"learning_rate": 1.8996735514908327e-07,
"logits/chosen": -2.083270311355591,
"logits/rejected": -1.2722394466400146,
"logps/chosen": -549.7798461914062,
"logps/rejected": -800.9820556640625,
"loss": 0.2384,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3795488774776459,
"rewards/margins": 0.324424147605896,
"rewards/rejected": -0.7039730548858643,
"step": 2220
},
{
"epoch": 0.54,
"learning_rate": 1.7684604777074427e-07,
"logits/chosen": -2.2368104457855225,
"logits/rejected": -1.5557681322097778,
"logps/chosen": -581.6041259765625,
"logps/rejected": -770.4801635742188,
"loss": 0.2392,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.4050823152065277,
"rewards/margins": 0.2399863749742508,
"rewards/rejected": -0.6450687646865845,
"step": 2230
},
{
"epoch": 0.54,
"learning_rate": 1.6417764385846996e-07,
"logits/chosen": -2.2670297622680664,
"logits/rejected": -1.6125434637069702,
"logps/chosen": -544.0134887695312,
"logps/rejected": -734.3864135742188,
"loss": 0.2371,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.38986533880233765,
"rewards/margins": 0.23339907824993134,
"rewards/rejected": -0.6232645511627197,
"step": 2240
},
{
"epoch": 0.54,
"learning_rate": 1.5196461314270438e-07,
"logits/chosen": -2.25602650642395,
"logits/rejected": -1.743583083152771,
"logps/chosen": -564.7320556640625,
"logps/rejected": -716.3707275390625,
"loss": 0.2454,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.42164283990859985,
"rewards/margins": 0.19018793106079102,
"rewards/rejected": -0.6118307709693909,
"step": 2250
},
{
"epoch": 0.54,
"learning_rate": 1.4020933657798385e-07,
"logits/chosen": -2.139263868331909,
"logits/rejected": -1.414298415184021,
"logps/chosen": -474.62567138671875,
"logps/rejected": -724.90234375,
"loss": 0.2306,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3453265130519867,
"rewards/margins": 0.2758641839027405,
"rewards/rejected": -0.6211907267570496,
"step": 2260
},
{
"epoch": 0.54,
"learning_rate": 1.2891410587876714e-07,
"logits/chosen": -2.2542724609375,
"logits/rejected": -1.5322940349578857,
"logps/chosen": -551.5656127929688,
"logps/rejected": -730.9732055664062,
"loss": 0.2194,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.37710070610046387,
"rewards/margins": 0.2374318391084671,
"rewards/rejected": -0.614532470703125,
"step": 2270
},
{
"epoch": 0.55,
"learning_rate": 1.180811230726589e-07,
"logits/chosen": -2.2804083824157715,
"logits/rejected": -1.5980250835418701,
"logps/chosen": -603.2000122070312,
"logps/rejected": -805.7347412109375,
"loss": 0.2254,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.42967596650123596,
"rewards/margins": 0.24143275618553162,
"rewards/rejected": -0.6711087226867676,
"step": 2280
},
{
"epoch": 0.55,
"learning_rate": 1.0771250007112155e-07,
"logits/chosen": -1.982797384262085,
"logits/rejected": -1.3462848663330078,
"logps/chosen": -626.7324829101562,
"logps/rejected": -776.6309814453125,
"loss": 0.2424,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.46235284209251404,
"rewards/margins": 0.2094097137451172,
"rewards/rejected": -0.6717625856399536,
"step": 2290
},
{
"epoch": 0.55,
"learning_rate": 9.781025825775392e-08,
"logits/chosen": -2.231228828430176,
"logits/rejected": -1.511236310005188,
"logps/chosen": -614.570068359375,
"logps/rejected": -825.9508666992188,
"loss": 0.2251,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.45310935378074646,
"rewards/margins": 0.26085203886032104,
"rewards/rejected": -0.7139613628387451,
"step": 2300
},
{
"epoch": 0.55,
"learning_rate": 8.837632809421681e-08,
"logits/chosen": -2.0968010425567627,
"logits/rejected": -1.3801376819610596,
"logps/chosen": -569.6008911132812,
"logps/rejected": -795.4178466796875,
"loss": 0.2706,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.4094913601875305,
"rewards/margins": 0.26325908303260803,
"rewards/rejected": -0.6727504134178162,
"step": 2310
},
{
"epoch": 0.56,
"learning_rate": 7.941254874388904e-08,
"logits/chosen": -2.363359212875366,
"logits/rejected": -1.888514757156372,
"logps/chosen": -576.6600341796875,
"logps/rejected": -702.62060546875,
"loss": 0.2567,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.43606749176979065,
"rewards/margins": 0.16073410212993622,
"rewards/rejected": -0.5968016386032104,
"step": 2320
},
{
"epoch": 0.56,
"learning_rate": 7.092066771331507e-08,
"logits/chosen": -2.1661031246185303,
"logits/rejected": -1.3824275732040405,
"logps/chosen": -557.127685546875,
"logps/rejected": -710.2548828125,
"loss": 0.2119,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.38099274039268494,
"rewards/margins": 0.22092202305793762,
"rewards/rejected": -0.6019147634506226,
"step": 2330
},
{
"epoch": 0.56,
"learning_rate": 6.29023405115281e-08,
"logits/chosen": -2.1738715171813965,
"logits/rejected": -1.355930209159851,
"logps/chosen": -592.2274169921875,
"logps/rejected": -788.6678466796875,
"loss": 0.2163,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.40638837218284607,
"rewards/margins": 0.2868625521659851,
"rewards/rejected": -0.6932509541511536,
"step": 2340
},
{
"epoch": 0.56,
"learning_rate": 5.535913032730295e-08,
"logits/chosen": -2.406480312347412,
"logits/rejected": -1.549068570137024,
"logps/chosen": -548.6936645507812,
"logps/rejected": -776.6002807617188,
"loss": 0.1999,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3767138421535492,
"rewards/margins": 0.2962132692337036,
"rewards/rejected": -0.6729270815849304,
"step": 2350
},
{
"epoch": 0.57,
"learning_rate": 4.829250772441091e-08,
"logits/chosen": -2.178439140319824,
"logits/rejected": -1.8580322265625,
"logps/chosen": -604.1091918945312,
"logps/rejected": -721.529541015625,
"loss": 0.2414,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.45432454347610474,
"rewards/margins": 0.1475105583667755,
"rewards/rejected": -0.6018351316452026,
"step": 2360
},
{
"epoch": 0.57,
"learning_rate": 4.170385035493108e-08,
"logits/chosen": -2.346930503845215,
"logits/rejected": -1.8971471786499023,
"logps/chosen": -603.0961303710938,
"logps/rejected": -768.2886352539062,
"loss": 0.2649,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.43998581171035767,
"rewards/margins": 0.20180657505989075,
"rewards/rejected": -0.6417924165725708,
"step": 2370
},
{
"epoch": 0.57,
"learning_rate": 3.5594442690671806e-08,
"logits/chosen": -2.013385772705078,
"logits/rejected": -1.435459852218628,
"logps/chosen": -645.1131591796875,
"logps/rejected": -806.417724609375,
"loss": 0.2246,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.48335081338882446,
"rewards/margins": 0.2265961617231369,
"rewards/rejected": -0.7099469900131226,
"step": 2380
},
{
"epoch": 0.57,
"learning_rate": 2.9965475772762154e-08,
"logits/chosen": -2.311368465423584,
"logits/rejected": -1.5600301027297974,
"logps/chosen": -517.8285522460938,
"logps/rejected": -704.8583374023438,
"loss": 0.2205,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.3128499984741211,
"rewards/margins": 0.2409205138683319,
"rewards/rejected": -0.5537704825401306,
"step": 2390
},
{
"epoch": 0.58,
"learning_rate": 2.48180469794565e-08,
"logits/chosen": -2.281245470046997,
"logits/rejected": -1.7727069854736328,
"logps/chosen": -474.0367126464844,
"logps/rejected": -638.2462768554688,
"loss": 0.2309,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.33747315406799316,
"rewards/margins": 0.18603594601154327,
"rewards/rejected": -0.5235090851783752,
"step": 2400
},
{
"epoch": 0.58,
"learning_rate": 2.015315981219651e-08,
"logits/chosen": -2.094449281692505,
"logits/rejected": -1.5543745756149292,
"logps/chosen": -585.0030517578125,
"logps/rejected": -759.0601806640625,
"loss": 0.2485,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.4290609359741211,
"rewards/margins": 0.21084634959697723,
"rewards/rejected": -0.6399072408676147,
"step": 2410
},
{
"epoch": 0.58,
"learning_rate": 1.5971723699979015e-08,
"logits/chosen": -2.2383382320404053,
"logits/rejected": -1.490106463432312,
"logps/chosen": -590.9389038085938,
"logps/rejected": -763.5640869140625,
"loss": 0.2426,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.4045529365539551,
"rewards/margins": 0.24901354312896729,
"rewards/rejected": -0.6535664796829224,
"step": 2420
},
{
"epoch": 0.58,
"learning_rate": 1.2274553822058944e-08,
"logits/chosen": -2.32003116607666,
"logits/rejected": -1.4864509105682373,
"logps/chosen": -516.053955078125,
"logps/rejected": -737.1478271484375,
"loss": 0.2119,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.3795922100543976,
"rewards/margins": 0.26055005192756653,
"rewards/rejected": -0.6401422619819641,
"step": 2430
},
{
"epoch": 0.59,
"learning_rate": 9.062370949029231e-09,
"logits/chosen": -2.3300509452819824,
"logits/rejected": -1.7241268157958984,
"logps/chosen": -578.6694946289062,
"logps/rejected": -768.1864013671875,
"loss": 0.2388,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.40254122018814087,
"rewards/margins": 0.2296265810728073,
"rewards/rejected": -0.6321677565574646,
"step": 2440
},
{
"epoch": 0.59,
"learning_rate": 6.3358013023062656e-09,
"logits/chosen": -2.0109941959381104,
"logits/rejected": -1.2714914083480835,
"logps/chosen": -602.9906005859375,
"logps/rejected": -774.6351318359375,
"loss": 0.2411,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.4527587294578552,
"rewards/margins": 0.22364509105682373,
"rewards/rejected": -0.676403820514679,
"step": 2450
},
{
"epoch": 0.59,
"learning_rate": 4.095376432044218e-09,
"logits/chosen": -2.153900623321533,
"logits/rejected": -1.431398630142212,
"logps/chosen": -501.6558532714844,
"logps/rejected": -703.2398681640625,
"loss": 0.2246,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.3516218066215515,
"rewards/margins": 0.26861336827278137,
"rewards/rejected": -0.6202351450920105,
"step": 2460
},
{
"epoch": 0.59,
"learning_rate": 2.3415331135115404e-09,
"logits/chosen": -2.1722538471221924,
"logits/rejected": -1.4859097003936768,
"logps/chosen": -573.3174438476562,
"logps/rejected": -785.3301391601562,
"loss": 0.2381,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.41808199882507324,
"rewards/margins": 0.2717761993408203,
"rewards/rejected": -0.6898581981658936,
"step": 2470
},
{
"epoch": 0.6,
"learning_rate": 1.0746132619374184e-09,
"logits/chosen": -2.1704633235931396,
"logits/rejected": -1.4999759197235107,
"logps/chosen": -571.4072265625,
"logps/rejected": -791.2116088867188,
"loss": 0.2218,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.41102513670921326,
"rewards/margins": 0.2840521037578583,
"rewards/rejected": -0.6950772404670715,
"step": 2480
},
{
"epoch": 0.6,
"learning_rate": 2.9486386585786395e-10,
"logits/chosen": -2.255235433578491,
"logits/rejected": -1.5667657852172852,
"logps/chosen": -458.48248291015625,
"logps/rejected": -655.1217651367188,
"loss": 0.2506,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.30341845750808716,
"rewards/margins": 0.2525646388530731,
"rewards/rejected": -0.5559830665588379,
"step": 2490
},
{
"epoch": 0.6,
"learning_rate": 2.4369389622913575e-12,
"logits/chosen": -2.230522394180298,
"logits/rejected": -1.6053917407989502,
"logps/chosen": -498.0244140625,
"logps/rejected": -680.3582763671875,
"loss": 0.2225,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.36517006158828735,
"rewards/margins": 0.2226562201976776,
"rewards/rejected": -0.5878263115882874,
"step": 2500
},
{
"epoch": 0.6,
"step": 2501,
"total_flos": 0.0,
"train_loss": 0.2318923625944615,
"train_runtime": 76628.4869,
"train_samples_per_second": 0.391,
"train_steps_per_second": 0.033
}
],
"logging_steps": 10,
"max_steps": 2501,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"total_flos": 0.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}