zephyr-7b-gpo-v11-i1 / trainer_state.json
lole25's picture
Model save
6b41881 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 3750,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 1.3333333333333334e-08,
"logits/chosen": -1.9468104839324951,
"logits/rejected": -1.3551281690597534,
"logps/chosen": -418.5311279296875,
"logps/rejected": -228.03335571289062,
"loss": 0.5309,
"rewards/accuracies": 0.0,
"rewards/chosen": 0.0,
"rewards/margins": 0.0,
"rewards/rejected": 0.0,
"step": 1
},
{
"epoch": 0.0,
"learning_rate": 1.3333333333333336e-07,
"logits/chosen": -2.1891419887542725,
"logits/rejected": -1.312096118927002,
"logps/chosen": -321.9786682128906,
"logps/rejected": -262.5549011230469,
"loss": 0.3989,
"rewards/accuracies": 0.3888888955116272,
"rewards/chosen": -0.00040232870378531516,
"rewards/margins": 0.00023253644758369774,
"rewards/rejected": -0.0006348651950247586,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 2.666666666666667e-07,
"logits/chosen": -1.7885940074920654,
"logits/rejected": -1.4744794368743896,
"logps/chosen": -270.1197204589844,
"logps/rejected": -280.1589050292969,
"loss": 0.3157,
"rewards/accuracies": 0.5,
"rewards/chosen": 0.0014780608471482992,
"rewards/margins": 0.002728077583014965,
"rewards/rejected": -0.0012500169686973095,
"step": 20
},
{
"epoch": 0.01,
"learning_rate": 4.0000000000000003e-07,
"logits/chosen": -1.9892597198486328,
"logits/rejected": -1.2171941995620728,
"logps/chosen": -319.71331787109375,
"logps/rejected": -257.2882995605469,
"loss": 0.3278,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": 0.00036752174491994083,
"rewards/margins": 0.0073011466301977634,
"rewards/rejected": -0.006933624390512705,
"step": 30
},
{
"epoch": 0.01,
"learning_rate": 5.333333333333335e-07,
"logits/chosen": -1.955718994140625,
"logits/rejected": -1.1262027025222778,
"logps/chosen": -356.73663330078125,
"logps/rejected": -266.71392822265625,
"loss": 0.2937,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.003951634746044874,
"rewards/margins": 0.028395619243383408,
"rewards/rejected": -0.024443982169032097,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 6.666666666666667e-07,
"logits/chosen": -1.9514992237091064,
"logits/rejected": -1.382391333580017,
"logps/chosen": -284.41790771484375,
"logps/rejected": -230.9414825439453,
"loss": 0.2014,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": 0.0021220450289547443,
"rewards/margins": 0.07282118499279022,
"rewards/rejected": -0.07069914042949677,
"step": 50
},
{
"epoch": 0.02,
"learning_rate": 8.000000000000001e-07,
"logits/chosen": -1.9106193780899048,
"logits/rejected": -1.3348861932754517,
"logps/chosen": -272.3318786621094,
"logps/rejected": -241.1548309326172,
"loss": 0.2992,
"rewards/accuracies": 0.625,
"rewards/chosen": -0.0524127297103405,
"rewards/margins": 0.1212296113371849,
"rewards/rejected": -0.1736423522233963,
"step": 60
},
{
"epoch": 0.02,
"learning_rate": 9.333333333333334e-07,
"logits/chosen": -1.9291963577270508,
"logits/rejected": -1.4450151920318604,
"logps/chosen": -287.9643249511719,
"logps/rejected": -281.98760986328125,
"loss": 0.2153,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.07112433016300201,
"rewards/margins": 0.18519330024719238,
"rewards/rejected": -0.256317675113678,
"step": 70
},
{
"epoch": 0.02,
"learning_rate": 1.066666666666667e-06,
"logits/chosen": -1.8260694742202759,
"logits/rejected": -1.2290842533111572,
"logps/chosen": -311.75787353515625,
"logps/rejected": -303.3421325683594,
"loss": 0.1991,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.050730206072330475,
"rewards/margins": 0.30313217639923096,
"rewards/rejected": -0.35386237502098083,
"step": 80
},
{
"epoch": 0.02,
"learning_rate": 1.2000000000000002e-06,
"logits/chosen": -2.1392667293548584,
"logits/rejected": -1.5895378589630127,
"logps/chosen": -264.8464050292969,
"logps/rejected": -284.0433349609375,
"loss": 0.1789,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.06132856756448746,
"rewards/margins": 0.34737664461135864,
"rewards/rejected": -0.4087051749229431,
"step": 90
},
{
"epoch": 0.03,
"learning_rate": 1.3333333333333334e-06,
"logits/chosen": -2.010281801223755,
"logits/rejected": -1.4457181692123413,
"logps/chosen": -280.8337707519531,
"logps/rejected": -301.07421875,
"loss": 0.1651,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.003755016950890422,
"rewards/margins": 0.3854002356529236,
"rewards/rejected": -0.38164520263671875,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 1.4666666666666669e-06,
"logits/chosen": -1.8176605701446533,
"logits/rejected": -1.4668805599212646,
"logps/chosen": -213.2742919921875,
"logps/rejected": -196.9753875732422,
"loss": 0.2645,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.08552353084087372,
"rewards/margins": 0.1824944019317627,
"rewards/rejected": -0.2680179476737976,
"step": 110
},
{
"epoch": 0.03,
"learning_rate": 1.6000000000000001e-06,
"logits/chosen": -1.8461990356445312,
"logits/rejected": -1.6458969116210938,
"logps/chosen": -206.00204467773438,
"logps/rejected": -255.66989135742188,
"loss": 0.2099,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.07237216085195541,
"rewards/margins": 0.22069768607616425,
"rewards/rejected": -0.29306983947753906,
"step": 120
},
{
"epoch": 0.03,
"learning_rate": 1.7333333333333336e-06,
"logits/chosen": -2.1012439727783203,
"logits/rejected": -1.28009831905365,
"logps/chosen": -324.0368957519531,
"logps/rejected": -254.3141632080078,
"loss": 0.0919,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.01445714384317398,
"rewards/margins": 0.3454706370830536,
"rewards/rejected": -0.3310135304927826,
"step": 130
},
{
"epoch": 0.04,
"learning_rate": 1.8666666666666669e-06,
"logits/chosen": -1.9437087774276733,
"logits/rejected": -1.5214288234710693,
"logps/chosen": -294.4078063964844,
"logps/rejected": -301.8307800292969,
"loss": 0.1193,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.07305511832237244,
"rewards/margins": 0.38559332489967346,
"rewards/rejected": -0.4586483836174011,
"step": 140
},
{
"epoch": 0.04,
"learning_rate": 2.0000000000000003e-06,
"logits/chosen": -1.9032337665557861,
"logits/rejected": -1.50299870967865,
"logps/chosen": -223.5890655517578,
"logps/rejected": -276.28955078125,
"loss": 0.1492,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.035106100142002106,
"rewards/margins": 0.4305610656738281,
"rewards/rejected": -0.46566715836524963,
"step": 150
},
{
"epoch": 0.04,
"learning_rate": 2.133333333333334e-06,
"logits/chosen": -1.9716815948486328,
"logits/rejected": -1.577401876449585,
"logps/chosen": -264.44256591796875,
"logps/rejected": -326.4450988769531,
"loss": 0.1267,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.05759359151124954,
"rewards/margins": 0.33429816365242004,
"rewards/rejected": -0.3918917775154114,
"step": 160
},
{
"epoch": 0.05,
"learning_rate": 2.266666666666667e-06,
"logits/chosen": -1.8987305164337158,
"logits/rejected": -1.3819479942321777,
"logps/chosen": -205.10104370117188,
"logps/rejected": -223.96218872070312,
"loss": 0.1927,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09939566999673843,
"rewards/margins": 0.24186280369758606,
"rewards/rejected": -0.3412584662437439,
"step": 170
},
{
"epoch": 0.05,
"learning_rate": 2.4000000000000003e-06,
"logits/chosen": -2.106255292892456,
"logits/rejected": -1.655613660812378,
"logps/chosen": -264.42547607421875,
"logps/rejected": -308.27130126953125,
"loss": 0.2318,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.09074126183986664,
"rewards/margins": 0.29490557312965393,
"rewards/rejected": -0.385646790266037,
"step": 180
},
{
"epoch": 0.05,
"learning_rate": 2.5333333333333338e-06,
"logits/chosen": -2.1222376823425293,
"logits/rejected": -1.7176287174224854,
"logps/chosen": -266.733154296875,
"logps/rejected": -274.0609436035156,
"loss": 0.1424,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.11357426643371582,
"rewards/margins": 0.35620003938674927,
"rewards/rejected": -0.4697743356227875,
"step": 190
},
{
"epoch": 0.05,
"learning_rate": 2.666666666666667e-06,
"logits/chosen": -1.969417929649353,
"logits/rejected": -1.3014360666275024,
"logps/chosen": -277.68536376953125,
"logps/rejected": -318.40496826171875,
"loss": 0.1694,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.08816908299922943,
"rewards/margins": 0.3808407485485077,
"rewards/rejected": -0.4690098166465759,
"step": 200
},
{
"epoch": 0.06,
"learning_rate": 2.8000000000000003e-06,
"logits/chosen": -2.0031707286834717,
"logits/rejected": -1.5582258701324463,
"logps/chosen": -235.62319946289062,
"logps/rejected": -280.3308410644531,
"loss": 0.1577,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.06687523424625397,
"rewards/margins": 0.30328303575515747,
"rewards/rejected": -0.37015828490257263,
"step": 210
},
{
"epoch": 0.06,
"learning_rate": 2.9333333333333338e-06,
"logits/chosen": -2.026477336883545,
"logits/rejected": -1.461743950843811,
"logps/chosen": -259.7619934082031,
"logps/rejected": -246.75588989257812,
"loss": 0.1748,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11929011344909668,
"rewards/margins": 0.2787408232688904,
"rewards/rejected": -0.39803093671798706,
"step": 220
},
{
"epoch": 0.06,
"learning_rate": 3.066666666666667e-06,
"logits/chosen": -2.0350136756896973,
"logits/rejected": -1.6047760248184204,
"logps/chosen": -291.62371826171875,
"logps/rejected": -326.29083251953125,
"loss": 0.1776,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.035884954035282135,
"rewards/margins": 0.35993653535842896,
"rewards/rejected": -0.3958215117454529,
"step": 230
},
{
"epoch": 0.06,
"learning_rate": 3.2000000000000003e-06,
"logits/chosen": -1.63616144657135,
"logits/rejected": -1.2675909996032715,
"logps/chosen": -295.3503723144531,
"logps/rejected": -272.33544921875,
"loss": 0.1512,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.04819333180785179,
"rewards/margins": 0.3286735713481903,
"rewards/rejected": -0.3768669664859772,
"step": 240
},
{
"epoch": 0.07,
"learning_rate": 3.3333333333333333e-06,
"logits/chosen": -2.153151035308838,
"logits/rejected": -1.5773793458938599,
"logps/chosen": -258.7474670410156,
"logps/rejected": -249.0330047607422,
"loss": 0.2077,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.07630246877670288,
"rewards/margins": 0.2940545678138733,
"rewards/rejected": -0.37035703659057617,
"step": 250
},
{
"epoch": 0.07,
"learning_rate": 3.4666666666666672e-06,
"logits/chosen": -1.8532987833023071,
"logits/rejected": -1.365553855895996,
"logps/chosen": -308.9462585449219,
"logps/rejected": -293.080810546875,
"loss": 0.0988,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.003472552401944995,
"rewards/margins": 0.37654823064804077,
"rewards/rejected": -0.38002076745033264,
"step": 260
},
{
"epoch": 0.07,
"learning_rate": 3.6000000000000003e-06,
"logits/chosen": -1.8927972316741943,
"logits/rejected": -1.3772236108779907,
"logps/chosen": -268.60845947265625,
"logps/rejected": -279.3491516113281,
"loss": 0.1617,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.05976073071360588,
"rewards/margins": 0.3105442523956299,
"rewards/rejected": -0.37030500173568726,
"step": 270
},
{
"epoch": 0.07,
"learning_rate": 3.7333333333333337e-06,
"logits/chosen": -2.0276782512664795,
"logits/rejected": -1.21445894241333,
"logps/chosen": -353.6800842285156,
"logps/rejected": -307.1755676269531,
"loss": 0.1506,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.0015403844881802797,
"rewards/margins": 0.5047177672386169,
"rewards/rejected": -0.5062581300735474,
"step": 280
},
{
"epoch": 0.08,
"learning_rate": 3.866666666666667e-06,
"logits/chosen": -2.00919771194458,
"logits/rejected": -1.3203434944152832,
"logps/chosen": -269.7672119140625,
"logps/rejected": -210.7847900390625,
"loss": 0.1969,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.009751735255122185,
"rewards/margins": 0.2548676133155823,
"rewards/rejected": -0.24511587619781494,
"step": 290
},
{
"epoch": 0.08,
"learning_rate": 4.000000000000001e-06,
"logits/chosen": -1.9286314249038696,
"logits/rejected": -1.4645249843597412,
"logps/chosen": -261.30047607421875,
"logps/rejected": -284.4623718261719,
"loss": 0.2013,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.03132876753807068,
"rewards/margins": 0.20891091227531433,
"rewards/rejected": -0.24023966491222382,
"step": 300
},
{
"epoch": 0.08,
"learning_rate": 4.133333333333333e-06,
"logits/chosen": -1.958105444908142,
"logits/rejected": -1.4698688983917236,
"logps/chosen": -237.3049774169922,
"logps/rejected": -291.87469482421875,
"loss": 0.1677,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07907866686582565,
"rewards/margins": 0.34895187616348267,
"rewards/rejected": -0.4280305802822113,
"step": 310
},
{
"epoch": 0.09,
"learning_rate": 4.266666666666668e-06,
"logits/chosen": -1.717402696609497,
"logits/rejected": -1.14840567111969,
"logps/chosen": -294.26641845703125,
"logps/rejected": -323.1462707519531,
"loss": 0.1992,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.0834791511297226,
"rewards/margins": 0.3541187047958374,
"rewards/rejected": -0.4375979006290436,
"step": 320
},
{
"epoch": 0.09,
"learning_rate": 4.4e-06,
"logits/chosen": -1.8336519002914429,
"logits/rejected": -1.480302333831787,
"logps/chosen": -237.1054229736328,
"logps/rejected": -276.50689697265625,
"loss": 0.0847,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.024688202887773514,
"rewards/margins": 0.40132126212120056,
"rewards/rejected": -0.3766331076622009,
"step": 330
},
{
"epoch": 0.09,
"learning_rate": 4.533333333333334e-06,
"logits/chosen": -1.9516515731811523,
"logits/rejected": -1.399285912513733,
"logps/chosen": -326.23236083984375,
"logps/rejected": -318.9416198730469,
"loss": 0.1551,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.04323751479387283,
"rewards/margins": 0.38205739855766296,
"rewards/rejected": -0.3388199210166931,
"step": 340
},
{
"epoch": 0.09,
"learning_rate": 4.666666666666667e-06,
"logits/chosen": -1.6592636108398438,
"logits/rejected": -1.313191533088684,
"logps/chosen": -246.4650421142578,
"logps/rejected": -316.64849853515625,
"loss": 0.226,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.002355779055505991,
"rewards/margins": 0.34276098012924194,
"rewards/rejected": -0.3404052257537842,
"step": 350
},
{
"epoch": 0.1,
"learning_rate": 4.800000000000001e-06,
"logits/chosen": -1.8236520290374756,
"logits/rejected": -1.3795961141586304,
"logps/chosen": -311.75341796875,
"logps/rejected": -295.3544921875,
"loss": 0.0932,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.029393743723630905,
"rewards/margins": 0.3503008484840393,
"rewards/rejected": -0.3209070861339569,
"step": 360
},
{
"epoch": 0.1,
"learning_rate": 4.933333333333334e-06,
"logits/chosen": -1.7931430339813232,
"logits/rejected": -1.2451846599578857,
"logps/chosen": -292.35650634765625,
"logps/rejected": -278.6131286621094,
"loss": 0.1793,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.10223956406116486,
"rewards/margins": 0.31418344378471375,
"rewards/rejected": -0.41642293334007263,
"step": 370
},
{
"epoch": 0.1,
"learning_rate": 4.999972922944898e-06,
"logits/chosen": -1.858270287513733,
"logits/rejected": -1.498663306236267,
"logps/chosen": -246.0679168701172,
"logps/rejected": -283.5242614746094,
"loss": 0.1688,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.23864665627479553,
"rewards/margins": 0.2593352198600769,
"rewards/rejected": -0.49798187613487244,
"step": 380
},
{
"epoch": 0.1,
"learning_rate": 4.999756310023261e-06,
"logits/chosen": -2.070420265197754,
"logits/rejected": -1.164734959602356,
"logps/chosen": -382.75543212890625,
"logps/rejected": -294.0628662109375,
"loss": 0.112,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1453527957201004,
"rewards/margins": 0.3971993327140808,
"rewards/rejected": -0.54255211353302,
"step": 390
},
{
"epoch": 0.11,
"learning_rate": 4.999323102948655e-06,
"logits/chosen": -1.7827831506729126,
"logits/rejected": -1.5060867071151733,
"logps/chosen": -227.31982421875,
"logps/rejected": -290.09149169921875,
"loss": 0.1737,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.17244981229305267,
"rewards/margins": 0.3016803562641144,
"rewards/rejected": -0.47413015365600586,
"step": 400
},
{
"epoch": 0.11,
"learning_rate": 4.998673339256785e-06,
"logits/chosen": -1.8005807399749756,
"logits/rejected": -1.5056277513504028,
"logps/chosen": -285.5323791503906,
"logps/rejected": -313.2432556152344,
"loss": 0.1204,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.14575102925300598,
"rewards/margins": 0.3333708941936493,
"rewards/rejected": -0.4791219234466553,
"step": 410
},
{
"epoch": 0.11,
"learning_rate": 4.997807075247147e-06,
"logits/chosen": -1.7242355346679688,
"logits/rejected": -1.4590495824813843,
"logps/chosen": -228.1505126953125,
"logps/rejected": -295.41241455078125,
"loss": 0.1442,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.12705926597118378,
"rewards/margins": 0.3108959197998047,
"rewards/rejected": -0.43795520067214966,
"step": 420
},
{
"epoch": 0.11,
"learning_rate": 4.996724385978142e-06,
"logits/chosen": -1.8692153692245483,
"logits/rejected": -1.2414253950119019,
"logps/chosen": -298.9205322265625,
"logps/rejected": -313.76934814453125,
"loss": 0.1191,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.0782383382320404,
"rewards/margins": 0.44491782784461975,
"rewards/rejected": -0.5231561064720154,
"step": 430
},
{
"epoch": 0.12,
"learning_rate": 4.995425365260585e-06,
"logits/chosen": -2.0321779251098633,
"logits/rejected": -1.462066411972046,
"logps/chosen": -301.1550598144531,
"logps/rejected": -312.3240966796875,
"loss": 0.1602,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.031835995614528656,
"rewards/margins": 0.3314369320869446,
"rewards/rejected": -0.3632729649543762,
"step": 440
},
{
"epoch": 0.12,
"learning_rate": 4.993910125649561e-06,
"logits/chosen": -1.8992176055908203,
"logits/rejected": -1.2387077808380127,
"logps/chosen": -250.6095428466797,
"logps/rejected": -202.1986083984375,
"loss": 0.174,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06963483989238739,
"rewards/margins": 0.25463372468948364,
"rewards/rejected": -0.32426854968070984,
"step": 450
},
{
"epoch": 0.12,
"learning_rate": 4.992178798434684e-06,
"logits/chosen": -1.8955596685409546,
"logits/rejected": -1.1744760274887085,
"logps/chosen": -389.3876647949219,
"logps/rejected": -292.69659423828125,
"loss": 0.1316,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.06589002162218094,
"rewards/margins": 0.41696634888648987,
"rewards/rejected": -0.4828563630580902,
"step": 460
},
{
"epoch": 0.13,
"learning_rate": 4.990231533628719e-06,
"logits/chosen": -1.8016172647476196,
"logits/rejected": -1.3248649835586548,
"logps/chosen": -238.6211700439453,
"logps/rejected": -275.3337707519531,
"loss": 0.1501,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.05240337923169136,
"rewards/margins": 0.3930678069591522,
"rewards/rejected": -0.4454711973667145,
"step": 470
},
{
"epoch": 0.13,
"learning_rate": 4.988068499954578e-06,
"logits/chosen": -1.6587965488433838,
"logits/rejected": -1.3655837774276733,
"logps/chosen": -190.276611328125,
"logps/rejected": -268.18341064453125,
"loss": 0.1726,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.02386285737156868,
"rewards/margins": 0.3009086549282074,
"rewards/rejected": -0.3247714936733246,
"step": 480
},
{
"epoch": 0.13,
"learning_rate": 4.985689884830711e-06,
"logits/chosen": -2.0086019039154053,
"logits/rejected": -1.4022401571273804,
"logps/chosen": -269.7921447753906,
"logps/rejected": -315.36761474609375,
"loss": 0.1211,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.0037451249081641436,
"rewards/margins": 0.34425827860832214,
"rewards/rejected": -0.3480033874511719,
"step": 490
},
{
"epoch": 0.13,
"learning_rate": 4.983095894354858e-06,
"logits/chosen": -2.0869174003601074,
"logits/rejected": -1.3568693399429321,
"logps/chosen": -354.6680603027344,
"logps/rejected": -353.8328552246094,
"loss": 0.1258,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07166734337806702,
"rewards/margins": 0.38118380308151245,
"rewards/rejected": -0.45285120606422424,
"step": 500
},
{
"epoch": 0.14,
"learning_rate": 4.980286753286196e-06,
"logits/chosen": -1.717799186706543,
"logits/rejected": -1.27217698097229,
"logps/chosen": -256.9685974121094,
"logps/rejected": -279.153076171875,
"loss": 0.1799,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.14614014327526093,
"rewards/margins": 0.37342625856399536,
"rewards/rejected": -0.5195664167404175,
"step": 510
},
{
"epoch": 0.14,
"learning_rate": 4.97726270502586e-06,
"logits/chosen": -1.9094120264053345,
"logits/rejected": -1.4152483940124512,
"logps/chosen": -214.5406951904297,
"logps/rejected": -217.8648223876953,
"loss": 0.1649,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2031155824661255,
"rewards/margins": 0.3454675078392029,
"rewards/rejected": -0.5485831499099731,
"step": 520
},
{
"epoch": 0.14,
"learning_rate": 4.974024011595864e-06,
"logits/chosen": -1.9784595966339111,
"logits/rejected": -1.203086495399475,
"logps/chosen": -336.6568908691406,
"logps/rejected": -286.4167785644531,
"loss": 0.1357,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.10868784040212631,
"rewards/margins": 0.566789984703064,
"rewards/rejected": -0.6754778623580933,
"step": 530
},
{
"epoch": 0.14,
"learning_rate": 4.970570953616383e-06,
"logits/chosen": -1.8285566568374634,
"logits/rejected": -1.075903296470642,
"logps/chosen": -302.0743103027344,
"logps/rejected": -253.8324432373047,
"loss": 0.124,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.14791302382946014,
"rewards/margins": 0.370217502117157,
"rewards/rejected": -0.5181306004524231,
"step": 540
},
{
"epoch": 0.15,
"learning_rate": 4.966903830281449e-06,
"logits/chosen": -1.774610161781311,
"logits/rejected": -1.4740030765533447,
"logps/chosen": -263.2217712402344,
"logps/rejected": -301.24090576171875,
"loss": 0.1898,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15221451222896576,
"rewards/margins": 0.2885417342185974,
"rewards/rejected": -0.440756231546402,
"step": 550
},
{
"epoch": 0.15,
"learning_rate": 4.9630229593330226e-06,
"logits/chosen": -1.808828592300415,
"logits/rejected": -1.260229468345642,
"logps/chosen": -251.52474975585938,
"logps/rejected": -294.1851806640625,
"loss": 0.1529,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.11935459077358246,
"rewards/margins": 0.34634923934936523,
"rewards/rejected": -0.4657038748264313,
"step": 560
},
{
"epoch": 0.15,
"learning_rate": 4.958928677033465e-06,
"logits/chosen": -1.7420295476913452,
"logits/rejected": -1.058304786682129,
"logps/chosen": -264.7672424316406,
"logps/rejected": -247.16189575195312,
"loss": 0.164,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.18925343453884125,
"rewards/margins": 0.3564419448375702,
"rewards/rejected": -0.545695424079895,
"step": 570
},
{
"epoch": 0.15,
"learning_rate": 4.954621338136399e-06,
"logits/chosen": -1.4567300081253052,
"logits/rejected": -0.9836466908454895,
"logps/chosen": -324.0819396972656,
"logps/rejected": -318.97796630859375,
"loss": 0.1473,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.11459589004516602,
"rewards/margins": 0.3253883421421051,
"rewards/rejected": -0.4399842321872711,
"step": 580
},
{
"epoch": 0.16,
"learning_rate": 4.95010131585597e-06,
"logits/chosen": -1.8161170482635498,
"logits/rejected": -1.4203059673309326,
"logps/chosen": -230.9877166748047,
"logps/rejected": -267.8335266113281,
"loss": 0.1543,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07126758992671967,
"rewards/margins": 0.3637334704399109,
"rewards/rejected": -0.43500104546546936,
"step": 590
},
{
"epoch": 0.16,
"learning_rate": 4.9453690018345144e-06,
"logits/chosen": -1.8937923908233643,
"logits/rejected": -1.3107613325119019,
"logps/chosen": -273.46002197265625,
"logps/rejected": -197.23764038085938,
"loss": 0.175,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.032990988343954086,
"rewards/margins": 0.2917155623435974,
"rewards/rejected": -0.324706494808197,
"step": 600
},
{
"epoch": 0.16,
"learning_rate": 4.940424806108619e-06,
"logits/chosen": -1.6622874736785889,
"logits/rejected": -1.2222545146942139,
"logps/chosen": -258.8514404296875,
"logps/rejected": -311.0167236328125,
"loss": 0.1296,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07337610423564911,
"rewards/margins": 0.39577198028564453,
"rewards/rejected": -0.46914809942245483,
"step": 610
},
{
"epoch": 0.17,
"learning_rate": 4.935269157073597e-06,
"logits/chosen": -1.6483392715454102,
"logits/rejected": -1.3003222942352295,
"logps/chosen": -242.5626983642578,
"logps/rejected": -345.08624267578125,
"loss": 0.1396,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.01702979765832424,
"rewards/margins": 0.4016871452331543,
"rewards/rejected": -0.4187169671058655,
"step": 620
},
{
"epoch": 0.17,
"learning_rate": 4.9299025014463665e-06,
"logits/chosen": -1.6613142490386963,
"logits/rejected": -1.1542552709579468,
"logps/chosen": -288.81915283203125,
"logps/rejected": -262.18365478515625,
"loss": 0.123,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.008388234302401543,
"rewards/margins": 0.43170255422592163,
"rewards/rejected": -0.4400908350944519,
"step": 630
},
{
"epoch": 0.17,
"learning_rate": 4.924325304226745e-06,
"logits/chosen": -1.544345736503601,
"logits/rejected": -1.0032846927642822,
"logps/chosen": -304.28668212890625,
"logps/rejected": -205.69210815429688,
"loss": 0.1941,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.11076553165912628,
"rewards/margins": 0.2719104588031769,
"rewards/rejected": -0.38267600536346436,
"step": 640
},
{
"epoch": 0.17,
"learning_rate": 4.91853804865716e-06,
"logits/chosen": -1.6050885915756226,
"logits/rejected": -1.3113044500350952,
"logps/chosen": -191.19522094726562,
"logps/rejected": -226.1024932861328,
"loss": 0.1883,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.07253624498844147,
"rewards/margins": 0.2440672218799591,
"rewards/rejected": -0.31660348176956177,
"step": 650
},
{
"epoch": 0.18,
"learning_rate": 4.912541236180779e-06,
"logits/chosen": -1.586315393447876,
"logits/rejected": -1.435160756111145,
"logps/chosen": -202.68026733398438,
"logps/rejected": -291.9640197753906,
"loss": 0.1334,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.04099782556295395,
"rewards/margins": 0.39952850341796875,
"rewards/rejected": -0.3585307002067566,
"step": 660
},
{
"epoch": 0.18,
"learning_rate": 4.9063353863980565e-06,
"logits/chosen": -1.5614166259765625,
"logits/rejected": -1.4164257049560547,
"logps/chosen": -227.6790008544922,
"logps/rejected": -290.1402893066406,
"loss": 0.1119,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.13384926319122314,
"rewards/margins": 0.3537839353084564,
"rewards/rejected": -0.48763322830200195,
"step": 670
},
{
"epoch": 0.18,
"learning_rate": 4.899921037021719e-06,
"logits/chosen": -1.6380398273468018,
"logits/rejected": -1.1617525815963745,
"logps/chosen": -242.23193359375,
"logps/rejected": -266.6896057128906,
"loss": 0.1565,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.10104341804981232,
"rewards/margins": 0.41885095834732056,
"rewards/rejected": -0.5198943614959717,
"step": 680
},
{
"epoch": 0.18,
"learning_rate": 4.893298743830168e-06,
"logits/chosen": -1.5904242992401123,
"logits/rejected": -1.138726830482483,
"logps/chosen": -317.62701416015625,
"logps/rejected": -297.424072265625,
"loss": 0.163,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.16763003170490265,
"rewards/margins": 0.35251811146736145,
"rewards/rejected": -0.5201481580734253,
"step": 690
},
{
"epoch": 0.19,
"learning_rate": 4.88646908061933e-06,
"logits/chosen": -1.7436052560806274,
"logits/rejected": -1.1463674306869507,
"logps/chosen": -322.777587890625,
"logps/rejected": -294.9422912597656,
"loss": 0.1452,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.07538704574108124,
"rewards/margins": 0.40292349457740784,
"rewards/rejected": -0.4783105254173279,
"step": 700
},
{
"epoch": 0.19,
"learning_rate": 4.879432639152935e-06,
"logits/chosen": -1.7265838384628296,
"logits/rejected": -1.3548028469085693,
"logps/chosen": -256.05902099609375,
"logps/rejected": -303.35430908203125,
"loss": 0.1221,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07280706614255905,
"rewards/margins": 0.35696297883987427,
"rewards/rejected": -0.4297700524330139,
"step": 710
},
{
"epoch": 0.19,
"learning_rate": 4.8721900291112415e-06,
"logits/chosen": -1.704933524131775,
"logits/rejected": -1.3638499975204468,
"logps/chosen": -240.88784790039062,
"logps/rejected": -307.12860107421875,
"loss": 0.1792,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.04838447645306587,
"rewards/margins": 0.36189574003219604,
"rewards/rejected": -0.41028016805648804,
"step": 720
},
{
"epoch": 0.19,
"learning_rate": 4.864741878038218e-06,
"logits/chosen": -1.490235686302185,
"logits/rejected": -1.2437798976898193,
"logps/chosen": -210.82626342773438,
"logps/rejected": -235.6994171142578,
"loss": 0.1351,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.02063031867146492,
"rewards/margins": 0.4025086462497711,
"rewards/rejected": -0.42313894629478455,
"step": 730
},
{
"epoch": 0.2,
"learning_rate": 4.857088831287158e-06,
"logits/chosen": -1.521240234375,
"logits/rejected": -1.1565624475479126,
"logps/chosen": -254.61605834960938,
"logps/rejected": -277.0546569824219,
"loss": 0.167,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.011497171595692635,
"rewards/margins": 0.3846563696861267,
"rewards/rejected": -0.3731592297554016,
"step": 740
},
{
"epoch": 0.2,
"learning_rate": 4.849231551964771e-06,
"logits/chosen": -1.518618106842041,
"logits/rejected": -0.9731992483139038,
"logps/chosen": -308.2921142578125,
"logps/rejected": -279.6849365234375,
"loss": 0.1304,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.006476040929555893,
"rewards/margins": 0.32605165243148804,
"rewards/rejected": -0.31957560777664185,
"step": 750
},
{
"epoch": 0.2,
"learning_rate": 4.841170720873723e-06,
"logits/chosen": -1.4644962549209595,
"logits/rejected": -0.6850159764289856,
"logps/chosen": -292.0091247558594,
"logps/rejected": -256.04095458984375,
"loss": 0.171,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.06350520253181458,
"rewards/margins": 0.3185378611087799,
"rewards/rejected": -0.3820430338382721,
"step": 760
},
{
"epoch": 0.21,
"learning_rate": 4.832907036453647e-06,
"logits/chosen": -1.5311912298202515,
"logits/rejected": -1.4509427547454834,
"logps/chosen": -149.7740478515625,
"logps/rejected": -253.37808227539062,
"loss": 0.1809,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.014500129036605358,
"rewards/margins": 0.3270939886569977,
"rewards/rejected": -0.3415941298007965,
"step": 770
},
{
"epoch": 0.21,
"learning_rate": 4.824441214720629e-06,
"logits/chosen": -1.3183726072311401,
"logits/rejected": -1.2068135738372803,
"logps/chosen": -227.69140625,
"logps/rejected": -288.60577392578125,
"loss": 0.1621,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.009285451844334602,
"rewards/margins": 0.39103344082832336,
"rewards/rejected": -0.3817480206489563,
"step": 780
},
{
"epoch": 0.21,
"learning_rate": 4.815773989205165e-06,
"logits/chosen": -1.6014219522476196,
"logits/rejected": -1.1616547107696533,
"logps/chosen": -275.4478454589844,
"logps/rejected": -282.85992431640625,
"loss": 0.1306,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.057247720658779144,
"rewards/margins": 0.3137260377407074,
"rewards/rejected": -0.3709737956523895,
"step": 790
},
{
"epoch": 0.21,
"learning_rate": 4.806906110888606e-06,
"logits/chosen": -1.7149804830551147,
"logits/rejected": -1.0609508752822876,
"logps/chosen": -247.8963165283203,
"logps/rejected": -251.2998504638672,
"loss": 0.1311,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.028542649000883102,
"rewards/margins": 0.3868168890476227,
"rewards/rejected": -0.4153594970703125,
"step": 800
},
{
"epoch": 0.22,
"learning_rate": 4.7978383481380865e-06,
"logits/chosen": -1.5295007228851318,
"logits/rejected": -1.3165438175201416,
"logps/chosen": -269.34539794921875,
"logps/rejected": -316.3207702636719,
"loss": 0.1328,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.05052419751882553,
"rewards/margins": 0.3775237798690796,
"rewards/rejected": -0.4280479848384857,
"step": 810
},
{
"epoch": 0.22,
"learning_rate": 4.788571486639948e-06,
"logits/chosen": -1.4131073951721191,
"logits/rejected": -1.1108410358428955,
"logps/chosen": -291.181640625,
"logps/rejected": -362.20416259765625,
"loss": 0.1235,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.13415458798408508,
"rewards/margins": 0.4179055094718933,
"rewards/rejected": -0.5520601272583008,
"step": 820
},
{
"epoch": 0.22,
"learning_rate": 4.779106329331665e-06,
"logits/chosen": -1.445796251296997,
"logits/rejected": -1.2411746978759766,
"logps/chosen": -247.7615203857422,
"logps/rejected": -265.7200927734375,
"loss": 0.1505,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.047227971255779266,
"rewards/margins": 0.3476831614971161,
"rewards/rejected": -0.39491117000579834,
"step": 830
},
{
"epoch": 0.22,
"learning_rate": 4.769443696332272e-06,
"logits/chosen": -1.595717430114746,
"logits/rejected": -1.014550805091858,
"logps/chosen": -277.58428955078125,
"logps/rejected": -247.3020477294922,
"loss": 0.1934,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07608959078788757,
"rewards/margins": 0.350595623254776,
"rewards/rejected": -0.4266851842403412,
"step": 840
},
{
"epoch": 0.23,
"learning_rate": 4.759584424871302e-06,
"logits/chosen": -1.7956939935684204,
"logits/rejected": -1.1622785329818726,
"logps/chosen": -268.0885009765625,
"logps/rejected": -260.65069580078125,
"loss": 0.1546,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11695736646652222,
"rewards/margins": 0.3590616285800934,
"rewards/rejected": -0.4760190546512604,
"step": 850
},
{
"epoch": 0.23,
"learning_rate": 4.749529369216246e-06,
"logits/chosen": -1.682879090309143,
"logits/rejected": -1.2164032459259033,
"logps/chosen": -253.8575439453125,
"logps/rejected": -316.49078369140625,
"loss": 0.1558,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.07984492182731628,
"rewards/margins": 0.36469632387161255,
"rewards/rejected": -0.44454121589660645,
"step": 860
},
{
"epoch": 0.23,
"learning_rate": 4.7392794005985324e-06,
"logits/chosen": -1.8441412448883057,
"logits/rejected": -1.0908098220825195,
"logps/chosen": -320.3082580566406,
"logps/rejected": -322.8711853027344,
"loss": 0.1675,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.027216767892241478,
"rewards/margins": 0.3997232913970947,
"rewards/rejected": -0.42694005370140076,
"step": 870
},
{
"epoch": 0.23,
"learning_rate": 4.7288354071380415e-06,
"logits/chosen": -1.5839688777923584,
"logits/rejected": -1.1818835735321045,
"logps/chosen": -305.91607666015625,
"logps/rejected": -304.3662109375,
"loss": 0.165,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.08447151631116867,
"rewards/margins": 0.3237994313240051,
"rewards/rejected": -0.4082708954811096,
"step": 880
},
{
"epoch": 0.24,
"learning_rate": 4.7181982937661485e-06,
"logits/chosen": -1.5419700145721436,
"logits/rejected": -1.2130048274993896,
"logps/chosen": -266.9795837402344,
"logps/rejected": -272.76708984375,
"loss": 0.1834,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.0005973346414975822,
"rewards/margins": 0.3700031042098999,
"rewards/rejected": -0.3694057762622833,
"step": 890
},
{
"epoch": 0.24,
"learning_rate": 4.707368982147318e-06,
"logits/chosen": -1.478826880455017,
"logits/rejected": -1.1886264085769653,
"logps/chosen": -249.00161743164062,
"logps/rejected": -311.38360595703125,
"loss": 0.1527,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.010296178050339222,
"rewards/margins": 0.27028435468673706,
"rewards/rejected": -0.2599882185459137,
"step": 900
},
{
"epoch": 0.24,
"learning_rate": 4.696348410599244e-06,
"logits/chosen": -1.6099306344985962,
"logits/rejected": -1.0822367668151855,
"logps/chosen": -259.72216796875,
"logps/rejected": -275.6427001953125,
"loss": 0.1892,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.02974315918982029,
"rewards/margins": 0.31014934182167053,
"rewards/rejected": -0.2804061770439148,
"step": 910
},
{
"epoch": 0.25,
"learning_rate": 4.685137534011549e-06,
"logits/chosen": -1.7383455038070679,
"logits/rejected": -1.079681158065796,
"logps/chosen": -249.6421661376953,
"logps/rejected": -230.7484893798828,
"loss": 0.1554,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.0027087554335594177,
"rewards/margins": 0.32712188363075256,
"rewards/rejected": -0.32441315054893494,
"step": 920
},
{
"epoch": 0.25,
"learning_rate": 4.673737323763048e-06,
"logits/chosen": -1.6007206439971924,
"logits/rejected": -1.1181820631027222,
"logps/chosen": -323.9920349121094,
"logps/rejected": -341.07568359375,
"loss": 0.137,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.006621385924518108,
"rewards/margins": 0.36024874448776245,
"rewards/rejected": -0.36687013506889343,
"step": 930
},
{
"epoch": 0.25,
"learning_rate": 4.662148767637578e-06,
"logits/chosen": -1.8025529384613037,
"logits/rejected": -1.0332825183868408,
"logps/chosen": -384.01666259765625,
"logps/rejected": -335.40374755859375,
"loss": 0.1161,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.06002165004611015,
"rewards/margins": 0.43708348274230957,
"rewards/rejected": -0.37706178426742554,
"step": 940
},
{
"epoch": 0.25,
"learning_rate": 4.650372869738415e-06,
"logits/chosen": -1.8271507024765015,
"logits/rejected": -1.211963415145874,
"logps/chosen": -328.3116760253906,
"logps/rejected": -304.9596252441406,
"loss": 0.1885,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.017533788457512856,
"rewards/margins": 0.3359457552433014,
"rewards/rejected": -0.3534795641899109,
"step": 950
},
{
"epoch": 0.26,
"learning_rate": 4.638410650401267e-06,
"logits/chosen": -1.4427958726882935,
"logits/rejected": -1.1035863161087036,
"logps/chosen": -261.233154296875,
"logps/rejected": -296.86358642578125,
"loss": 0.1704,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.11386547982692719,
"rewards/margins": 0.3565741181373596,
"rewards/rejected": -0.470439612865448,
"step": 960
},
{
"epoch": 0.26,
"learning_rate": 4.626263146105875e-06,
"logits/chosen": -1.6715164184570312,
"logits/rejected": -1.2049211263656616,
"logps/chosen": -267.9721374511719,
"logps/rejected": -272.67218017578125,
"loss": 0.1604,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.14121413230895996,
"rewards/margins": 0.35565823316574097,
"rewards/rejected": -0.49687233567237854,
"step": 970
},
{
"epoch": 0.26,
"learning_rate": 4.613931409386196e-06,
"logits/chosen": -1.614467978477478,
"logits/rejected": -1.3309428691864014,
"logps/chosen": -282.2423400878906,
"logps/rejected": -315.1558532714844,
"loss": 0.1396,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09741922467947006,
"rewards/margins": 0.3464723527431488,
"rewards/rejected": -0.44389158487319946,
"step": 980
},
{
"epoch": 0.26,
"learning_rate": 4.601416508739211e-06,
"logits/chosen": -1.6913681030273438,
"logits/rejected": -1.0562645196914673,
"logps/chosen": -260.50457763671875,
"logps/rejected": -236.1143798828125,
"loss": 0.252,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.14730839431285858,
"rewards/margins": 0.2597261965274811,
"rewards/rejected": -0.40703457593917847,
"step": 990
},
{
"epoch": 0.27,
"learning_rate": 4.588719528532342e-06,
"logits/chosen": -1.7020518779754639,
"logits/rejected": -1.1313974857330322,
"logps/chosen": -307.2850646972656,
"logps/rejected": -269.5376892089844,
"loss": 0.1474,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.04724062234163284,
"rewards/margins": 0.38787880539894104,
"rewards/rejected": -0.4351194500923157,
"step": 1000
},
{
"epoch": 0.27,
"learning_rate": 4.575841568909494e-06,
"logits/chosen": -1.3848850727081299,
"logits/rejected": -1.0428290367126465,
"logps/chosen": -240.163330078125,
"logps/rejected": -317.8233337402344,
"loss": 0.2369,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.028999319300055504,
"rewards/margins": 0.30743494629859924,
"rewards/rejected": -0.3364342451095581,
"step": 1010
},
{
"epoch": 0.27,
"learning_rate": 4.562783745695738e-06,
"logits/chosen": -1.3958890438079834,
"logits/rejected": -1.2130458354949951,
"logps/chosen": -203.06283569335938,
"logps/rejected": -265.59600830078125,
"loss": 0.1827,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.020870503038167953,
"rewards/margins": 0.3209065794944763,
"rewards/rejected": -0.34177708625793457,
"step": 1020
},
{
"epoch": 0.27,
"learning_rate": 4.549547190300622e-06,
"logits/chosen": -1.5980967283248901,
"logits/rejected": -1.1070952415466309,
"logps/chosen": -262.9022216796875,
"logps/rejected": -253.2421112060547,
"loss": 0.1491,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.010922988876700401,
"rewards/margins": 0.3387434184551239,
"rewards/rejected": -0.34966641664505005,
"step": 1030
},
{
"epoch": 0.28,
"learning_rate": 4.536133049620143e-06,
"logits/chosen": -1.7377541065216064,
"logits/rejected": -1.0692071914672852,
"logps/chosen": -338.3387145996094,
"logps/rejected": -277.5804443359375,
"loss": 0.1656,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.01836409978568554,
"rewards/margins": 0.3910614550113678,
"rewards/rejected": -0.4094255566596985,
"step": 1040
},
{
"epoch": 0.28,
"learning_rate": 4.522542485937369e-06,
"logits/chosen": -1.4216773509979248,
"logits/rejected": -0.8998391032218933,
"logps/chosen": -255.6497039794922,
"logps/rejected": -270.85980224609375,
"loss": 0.1628,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.049015216529369354,
"rewards/margins": 0.348868191242218,
"rewards/rejected": -0.3978833854198456,
"step": 1050
},
{
"epoch": 0.28,
"learning_rate": 4.508776676821739e-06,
"logits/chosen": -1.501990556716919,
"logits/rejected": -1.2285462617874146,
"logps/chosen": -260.5689392089844,
"logps/rejected": -342.2594299316406,
"loss": 0.1413,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.14666099846363068,
"rewards/margins": 0.32123270630836487,
"rewards/rejected": -0.46789368987083435,
"step": 1060
},
{
"epoch": 0.29,
"learning_rate": 4.494836815027022e-06,
"logits/chosen": -1.86409592628479,
"logits/rejected": -1.2688671350479126,
"logps/chosen": -265.83343505859375,
"logps/rejected": -272.33636474609375,
"loss": 0.1657,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.08181764930486679,
"rewards/margins": 0.3258339464664459,
"rewards/rejected": -0.4076516032218933,
"step": 1070
},
{
"epoch": 0.29,
"learning_rate": 4.4807241083879774e-06,
"logits/chosen": -1.7254140377044678,
"logits/rejected": -1.1994249820709229,
"logps/chosen": -274.38104248046875,
"logps/rejected": -289.0826416015625,
"loss": 0.1598,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.08081559091806412,
"rewards/margins": 0.4157637655735016,
"rewards/rejected": -0.4965793192386627,
"step": 1080
},
{
"epoch": 0.29,
"learning_rate": 4.466439779715696e-06,
"logits/chosen": -1.5560534000396729,
"logits/rejected": -1.0609302520751953,
"logps/chosen": -320.48931884765625,
"logps/rejected": -237.6038055419922,
"loss": 0.1151,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13456735014915466,
"rewards/margins": 0.3497200012207031,
"rewards/rejected": -0.48428741097450256,
"step": 1090
},
{
"epoch": 0.29,
"learning_rate": 4.451985066691649e-06,
"logits/chosen": -1.5969184637069702,
"logits/rejected": -1.100097417831421,
"logps/chosen": -258.00811767578125,
"logps/rejected": -307.08203125,
"loss": 0.1526,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07336065918207169,
"rewards/margins": 0.3625486493110657,
"rewards/rejected": -0.43590933084487915,
"step": 1100
},
{
"epoch": 0.3,
"learning_rate": 4.437361221760449e-06,
"logits/chosen": -1.6261990070343018,
"logits/rejected": -0.9406700134277344,
"logps/chosen": -304.49530029296875,
"logps/rejected": -281.0793151855469,
"loss": 0.1026,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.03398927301168442,
"rewards/margins": 0.4203387200832367,
"rewards/rejected": -0.4543279707431793,
"step": 1110
},
{
"epoch": 0.3,
"learning_rate": 4.422569512021332e-06,
"logits/chosen": -1.5429813861846924,
"logits/rejected": -0.9988912343978882,
"logps/chosen": -248.56948852539062,
"logps/rejected": -265.1681823730469,
"loss": 0.0936,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.09713175892829895,
"rewards/margins": 0.3802811801433563,
"rewards/rejected": -0.4774129390716553,
"step": 1120
},
{
"epoch": 0.3,
"learning_rate": 4.407611219118363e-06,
"logits/chosen": -1.5700933933258057,
"logits/rejected": -1.0617953538894653,
"logps/chosen": -233.2037353515625,
"logps/rejected": -200.68634033203125,
"loss": 0.144,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.11756577342748642,
"rewards/margins": 0.31560632586479187,
"rewards/rejected": -0.4331720769405365,
"step": 1130
},
{
"epoch": 0.3,
"learning_rate": 4.3924876391293915e-06,
"logits/chosen": -1.8913711309432983,
"logits/rejected": -1.039268970489502,
"logps/chosen": -353.1129455566406,
"logps/rejected": -297.39617919921875,
"loss": 0.1334,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.08230503648519516,
"rewards/margins": 0.47759518027305603,
"rewards/rejected": -0.5599002838134766,
"step": 1140
},
{
"epoch": 0.31,
"learning_rate": 4.377200082453748e-06,
"logits/chosen": -1.538246512413025,
"logits/rejected": -1.0102214813232422,
"logps/chosen": -364.6609802246094,
"logps/rejected": -325.3028869628906,
"loss": 0.1223,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1402122676372528,
"rewards/margins": 0.35043373703956604,
"rewards/rejected": -0.49064597487449646,
"step": 1150
},
{
"epoch": 0.31,
"learning_rate": 4.361749873698707e-06,
"logits/chosen": -1.58868408203125,
"logits/rejected": -1.048269271850586,
"logps/chosen": -210.7605743408203,
"logps/rejected": -217.05050659179688,
"loss": 0.1285,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.09829956293106079,
"rewards/margins": 0.3576180338859558,
"rewards/rejected": -0.4559175372123718,
"step": 1160
},
{
"epoch": 0.31,
"learning_rate": 4.346138351564711e-06,
"logits/chosen": -1.5400859117507935,
"logits/rejected": -0.9626834988594055,
"logps/chosen": -281.67779541015625,
"logps/rejected": -280.89837646484375,
"loss": 0.2163,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.11693022400140762,
"rewards/margins": 0.3672144412994385,
"rewards/rejected": -0.4841446876525879,
"step": 1170
},
{
"epoch": 0.31,
"learning_rate": 4.330366868729376e-06,
"logits/chosen": -1.5415042638778687,
"logits/rejected": -1.026490569114685,
"logps/chosen": -233.54293823242188,
"logps/rejected": -307.4654846191406,
"loss": 0.1482,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.061702560633420944,
"rewards/margins": 0.3937299847602844,
"rewards/rejected": -0.45543256402015686,
"step": 1180
},
{
"epoch": 0.32,
"learning_rate": 4.3144367917302964e-06,
"logits/chosen": -1.8853957653045654,
"logits/rejected": -1.1902307271957397,
"logps/chosen": -313.8409118652344,
"logps/rejected": -296.0729064941406,
"loss": 0.1522,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.017522500827908516,
"rewards/margins": 0.3799929618835449,
"rewards/rejected": -0.39751550555229187,
"step": 1190
},
{
"epoch": 0.32,
"learning_rate": 4.2983495008466285e-06,
"logits/chosen": -1.3914777040481567,
"logits/rejected": -0.9230860471725464,
"logps/chosen": -266.98321533203125,
"logps/rejected": -292.6563415527344,
"loss": 0.1513,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.04734548181295395,
"rewards/margins": 0.3820473849773407,
"rewards/rejected": -0.42939287424087524,
"step": 1200
},
{
"epoch": 0.32,
"learning_rate": 4.2821063899795015e-06,
"logits/chosen": -1.335268259048462,
"logits/rejected": -1.4020699262619019,
"logps/chosen": -224.03970336914062,
"logps/rejected": -348.4554138183594,
"loss": 0.1196,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.0435425229370594,
"rewards/margins": 0.4308520257472992,
"rewards/rejected": -0.4743945598602295,
"step": 1210
},
{
"epoch": 0.33,
"learning_rate": 4.265708866531238e-06,
"logits/chosen": -1.5381969213485718,
"logits/rejected": -1.2920863628387451,
"logps/chosen": -273.07086181640625,
"logps/rejected": -312.44000244140625,
"loss": 0.1688,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.03625233843922615,
"rewards/margins": 0.26211121678352356,
"rewards/rejected": -0.29836350679397583,
"step": 1220
},
{
"epoch": 0.33,
"learning_rate": 4.249158351283414e-06,
"logits/chosen": -1.5409828424453735,
"logits/rejected": -1.1099916696548462,
"logps/chosen": -269.79833984375,
"logps/rejected": -264.7692565917969,
"loss": 0.1503,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07565125823020935,
"rewards/margins": 0.32862424850463867,
"rewards/rejected": -0.404275506734848,
"step": 1230
},
{
"epoch": 0.33,
"learning_rate": 4.232456278273743e-06,
"logits/chosen": -1.8242772817611694,
"logits/rejected": -1.0602861642837524,
"logps/chosen": -371.48248291015625,
"logps/rejected": -312.5940856933594,
"loss": 0.1571,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.10061591863632202,
"rewards/margins": 0.3603518605232239,
"rewards/rejected": -0.4609677791595459,
"step": 1240
},
{
"epoch": 0.33,
"learning_rate": 4.215604094671835e-06,
"logits/chosen": -1.5043697357177734,
"logits/rejected": -1.3470711708068848,
"logps/chosen": -348.822021484375,
"logps/rejected": -378.1570739746094,
"loss": 0.1618,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.17215891182422638,
"rewards/margins": 0.3960806727409363,
"rewards/rejected": -0.5682395696640015,
"step": 1250
},
{
"epoch": 0.34,
"learning_rate": 4.198603260653792e-06,
"logits/chosen": -1.1872601509094238,
"logits/rejected": -0.9705106616020203,
"logps/chosen": -272.1139221191406,
"logps/rejected": -295.6619873046875,
"loss": 0.154,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.19514642655849457,
"rewards/margins": 0.3168686032295227,
"rewards/rejected": -0.5120150446891785,
"step": 1260
},
{
"epoch": 0.34,
"learning_rate": 4.181455249275701e-06,
"logits/chosen": -1.668910264968872,
"logits/rejected": -1.1042711734771729,
"logps/chosen": -373.8921813964844,
"logps/rejected": -264.4703369140625,
"loss": 0.2755,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1737692654132843,
"rewards/margins": 0.26853370666503906,
"rewards/rejected": -0.44230300188064575,
"step": 1270
},
{
"epoch": 0.34,
"learning_rate": 4.1641615463459926e-06,
"logits/chosen": -1.7166597843170166,
"logits/rejected": -0.9484345316886902,
"logps/chosen": -329.9338684082031,
"logps/rejected": -262.08343505859375,
"loss": 0.138,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.15403738617897034,
"rewards/margins": 0.3067986071109772,
"rewards/rejected": -0.4608360230922699,
"step": 1280
},
{
"epoch": 0.34,
"learning_rate": 4.146723650296701e-06,
"logits/chosen": -1.5266129970550537,
"logits/rejected": -1.1601712703704834,
"logps/chosen": -353.5336608886719,
"logps/rejected": -334.5635681152344,
"loss": 0.1007,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.12546080350875854,
"rewards/margins": 0.4080546796321869,
"rewards/rejected": -0.5335155129432678,
"step": 1290
},
{
"epoch": 0.35,
"learning_rate": 4.129143072053639e-06,
"logits/chosen": -1.7788454294204712,
"logits/rejected": -1.1070853471755981,
"logps/chosen": -301.1758728027344,
"logps/rejected": -272.003662109375,
"loss": 0.1172,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1457098424434662,
"rewards/margins": 0.41548848152160645,
"rewards/rejected": -0.5611982345581055,
"step": 1300
},
{
"epoch": 0.35,
"learning_rate": 4.111421334905468e-06,
"logits/chosen": -1.445004940032959,
"logits/rejected": -1.095983862876892,
"logps/chosen": -219.327392578125,
"logps/rejected": -277.88983154296875,
"loss": 0.2043,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.19351212680339813,
"rewards/margins": 0.2906314730644226,
"rewards/rejected": -0.48414358496665955,
"step": 1310
},
{
"epoch": 0.35,
"learning_rate": 4.093559974371725e-06,
"logits/chosen": -1.4865139722824097,
"logits/rejected": -1.354561686515808,
"logps/chosen": -248.8349151611328,
"logps/rejected": -252.2455291748047,
"loss": 0.1729,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1259056180715561,
"rewards/margins": 0.31513845920562744,
"rewards/rejected": -0.4410440921783447,
"step": 1320
},
{
"epoch": 0.35,
"learning_rate": 4.075560538069767e-06,
"logits/chosen": -1.8343995809555054,
"logits/rejected": -1.280969500541687,
"logps/chosen": -254.92056274414062,
"logps/rejected": -305.6126708984375,
"loss": 0.1306,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1363029032945633,
"rewards/margins": 0.37759169936180115,
"rewards/rejected": -0.5138946175575256,
"step": 1330
},
{
"epoch": 0.36,
"learning_rate": 4.05742458558068e-06,
"logits/chosen": -1.6347767114639282,
"logits/rejected": -1.1042792797088623,
"logps/chosen": -292.10552978515625,
"logps/rejected": -341.2361145019531,
"loss": 0.1214,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.08132176101207733,
"rewards/margins": 0.5133122801780701,
"rewards/rejected": -0.5946341156959534,
"step": 1340
},
{
"epoch": 0.36,
"learning_rate": 4.039153688314146e-06,
"logits/chosen": -1.6722145080566406,
"logits/rejected": -1.1385935544967651,
"logps/chosen": -247.4423370361328,
"logps/rejected": -231.4825439453125,
"loss": 0.1096,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.11472739279270172,
"rewards/margins": 0.34514716267585754,
"rewards/rejected": -0.45987454056739807,
"step": 1350
},
{
"epoch": 0.36,
"learning_rate": 4.020749429372286e-06,
"logits/chosen": -1.7172428369522095,
"logits/rejected": -1.2585347890853882,
"logps/chosen": -310.28179931640625,
"logps/rejected": -296.54071044921875,
"loss": 0.1346,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.11566674709320068,
"rewards/margins": 0.3406696021556854,
"rewards/rejected": -0.4563364088535309,
"step": 1360
},
{
"epoch": 0.37,
"learning_rate": 4.002213403412492e-06,
"logits/chosen": -1.479115605354309,
"logits/rejected": -1.075224757194519,
"logps/chosen": -250.99026489257812,
"logps/rejected": -262.92755126953125,
"loss": 0.1859,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.14899684488773346,
"rewards/margins": 0.3160502314567566,
"rewards/rejected": -0.46504706144332886,
"step": 1370
},
{
"epoch": 0.37,
"learning_rate": 3.983547216509254e-06,
"logits/chosen": -1.6718261241912842,
"logits/rejected": -1.1382570266723633,
"logps/chosen": -298.2779235839844,
"logps/rejected": -322.4492492675781,
"loss": 0.1162,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.099492147564888,
"rewards/margins": 0.4109002947807312,
"rewards/rejected": -0.510392427444458,
"step": 1380
},
{
"epoch": 0.37,
"learning_rate": 3.964752486015001e-06,
"logits/chosen": -1.741803765296936,
"logits/rejected": -1.154526948928833,
"logps/chosen": -322.1985778808594,
"logps/rejected": -258.2569885253906,
"loss": 0.1312,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.0690280944108963,
"rewards/margins": 0.4529304504394531,
"rewards/rejected": -0.5219585299491882,
"step": 1390
},
{
"epoch": 0.37,
"learning_rate": 3.945830840419966e-06,
"logits/chosen": -1.4246766567230225,
"logits/rejected": -1.1284135580062866,
"logps/chosen": -253.0252227783203,
"logps/rejected": -245.19723510742188,
"loss": 0.1713,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.10711170732975006,
"rewards/margins": 0.3032890558242798,
"rewards/rejected": -0.41040077805519104,
"step": 1400
},
{
"epoch": 0.38,
"learning_rate": 3.92678391921108e-06,
"logits/chosen": -1.580693006515503,
"logits/rejected": -1.0569651126861572,
"logps/chosen": -289.4805603027344,
"logps/rejected": -289.3725280761719,
"loss": 0.1539,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.08762092143297195,
"rewards/margins": 0.33070939779281616,
"rewards/rejected": -0.4183303415775299,
"step": 1410
},
{
"epoch": 0.38,
"learning_rate": 3.907613372729916e-06,
"logits/chosen": -1.6741529703140259,
"logits/rejected": -0.962120532989502,
"logps/chosen": -332.26080322265625,
"logps/rejected": -344.08209228515625,
"loss": 0.1367,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15168026089668274,
"rewards/margins": 0.4351193308830261,
"rewards/rejected": -0.5867995619773865,
"step": 1420
},
{
"epoch": 0.38,
"learning_rate": 3.888320862029699e-06,
"logits/chosen": -1.477236032485962,
"logits/rejected": -1.1961562633514404,
"logps/chosen": -251.10671997070312,
"logps/rejected": -272.6755676269531,
"loss": 0.1711,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.10620205104351044,
"rewards/margins": 0.34451884031295776,
"rewards/rejected": -0.450720876455307,
"step": 1430
},
{
"epoch": 0.38,
"learning_rate": 3.868908058731376e-06,
"logits/chosen": -1.5320178270339966,
"logits/rejected": -1.2863932847976685,
"logps/chosen": -252.96298217773438,
"logps/rejected": -267.1416320800781,
"loss": 0.1757,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.13874351978302002,
"rewards/margins": 0.3073849678039551,
"rewards/rejected": -0.4461284577846527,
"step": 1440
},
{
"epoch": 0.39,
"learning_rate": 3.849376644878783e-06,
"logits/chosen": -1.695948839187622,
"logits/rejected": -1.3465334177017212,
"logps/chosen": -256.7879333496094,
"logps/rejected": -291.6668701171875,
"loss": 0.1414,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.08056484907865524,
"rewards/margins": 0.3060542345046997,
"rewards/rejected": -0.38661906123161316,
"step": 1450
},
{
"epoch": 0.39,
"learning_rate": 3.829728312792895e-06,
"logits/chosen": -1.4604734182357788,
"logits/rejected": -1.1766102313995361,
"logps/chosen": -228.4718780517578,
"logps/rejected": -230.71182250976562,
"loss": 0.2127,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.10572078078985214,
"rewards/margins": 0.21217215061187744,
"rewards/rejected": -0.3178929388523102,
"step": 1460
},
{
"epoch": 0.39,
"learning_rate": 3.8099647649251984e-06,
"logits/chosen": -1.9331858158111572,
"logits/rejected": -1.301509976387024,
"logps/chosen": -338.3063049316406,
"logps/rejected": -301.18365478515625,
"loss": 0.1575,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11515772342681885,
"rewards/margins": 0.33505210280418396,
"rewards/rejected": -0.4502098560333252,
"step": 1470
},
{
"epoch": 0.39,
"learning_rate": 3.790087713710179e-06,
"logits/chosen": -1.7642886638641357,
"logits/rejected": -1.1480679512023926,
"logps/chosen": -296.9232177734375,
"logps/rejected": -282.36151123046875,
"loss": 0.1711,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.017070520669221878,
"rewards/margins": 0.4179867208003998,
"rewards/rejected": -0.43505725264549255,
"step": 1480
},
{
"epoch": 0.4,
"learning_rate": 3.770098881416945e-06,
"logits/chosen": -1.6907027959823608,
"logits/rejected": -1.1980526447296143,
"logps/chosen": -303.41778564453125,
"logps/rejected": -332.55352783203125,
"loss": 0.2434,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.10351963341236115,
"rewards/margins": 0.2676793932914734,
"rewards/rejected": -0.37119898200035095,
"step": 1490
},
{
"epoch": 0.4,
"learning_rate": 3.7500000000000005e-06,
"logits/chosen": -1.7876999378204346,
"logits/rejected": -1.3319778442382812,
"logps/chosen": -264.2928466796875,
"logps/rejected": -257.09503173828125,
"loss": 0.1408,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.0641285628080368,
"rewards/margins": 0.3521239161491394,
"rewards/rejected": -0.4162525236606598,
"step": 1500
},
{
"epoch": 0.4,
"learning_rate": 3.7297928109491765e-06,
"logits/chosen": -1.5949997901916504,
"logits/rejected": -1.1277220249176025,
"logps/chosen": -234.57373046875,
"logps/rejected": -195.37899780273438,
"loss": 0.2328,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.13893568515777588,
"rewards/margins": 0.26205331087112427,
"rewards/rejected": -0.40098896622657776,
"step": 1510
},
{
"epoch": 0.41,
"learning_rate": 3.7094790651387414e-06,
"logits/chosen": -1.6848033666610718,
"logits/rejected": -1.2798172235488892,
"logps/chosen": -236.25436401367188,
"logps/rejected": -269.56182861328125,
"loss": 0.2051,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09305624663829803,
"rewards/margins": 0.26189345121383667,
"rewards/rejected": -0.3549497723579407,
"step": 1520
},
{
"epoch": 0.41,
"learning_rate": 3.689060522675689e-06,
"logits/chosen": -1.6376606225967407,
"logits/rejected": -1.0957934856414795,
"logps/chosen": -259.7792663574219,
"logps/rejected": -254.48489379882812,
"loss": 0.1368,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09541145712137222,
"rewards/margins": 0.31637534499168396,
"rewards/rejected": -0.4117867946624756,
"step": 1530
},
{
"epoch": 0.41,
"learning_rate": 3.668538952747236e-06,
"logits/chosen": -1.534891128540039,
"logits/rejected": -1.1751198768615723,
"logps/chosen": -208.9001007080078,
"logps/rejected": -289.44873046875,
"loss": 0.1984,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.031213950365781784,
"rewards/margins": 0.39292722940444946,
"rewards/rejected": -0.42414116859436035,
"step": 1540
},
{
"epoch": 0.41,
"learning_rate": 3.6479161334675294e-06,
"logits/chosen": -1.7336101531982422,
"logits/rejected": -1.0565919876098633,
"logps/chosen": -326.5076599121094,
"logps/rejected": -337.8768005371094,
"loss": 0.1143,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.03437976911664009,
"rewards/margins": 0.49583953619003296,
"rewards/rejected": -0.46145981550216675,
"step": 1550
},
{
"epoch": 0.42,
"learning_rate": 3.627193851723577e-06,
"logits/chosen": -1.7617276906967163,
"logits/rejected": -0.987285315990448,
"logps/chosen": -256.71514892578125,
"logps/rejected": -230.3134002685547,
"loss": 0.1108,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.02552700974047184,
"rewards/margins": 0.3413180708885193,
"rewards/rejected": -0.3668450117111206,
"step": 1560
},
{
"epoch": 0.42,
"learning_rate": 3.6063739030204226e-06,
"logits/chosen": -1.606143593788147,
"logits/rejected": -1.3145328760147095,
"logps/chosen": -222.5089874267578,
"logps/rejected": -284.4768981933594,
"loss": 0.1263,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.04628473147749901,
"rewards/margins": 0.3009468913078308,
"rewards/rejected": -0.3472316563129425,
"step": 1570
},
{
"epoch": 0.42,
"learning_rate": 3.5854580913255706e-06,
"logits/chosen": -1.6054519414901733,
"logits/rejected": -1.0964720249176025,
"logps/chosen": -284.88446044921875,
"logps/rejected": -234.81417846679688,
"loss": 0.1716,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.049481119960546494,
"rewards/margins": 0.32506829500198364,
"rewards/rejected": -0.37454938888549805,
"step": 1580
},
{
"epoch": 0.42,
"learning_rate": 3.564448228912682e-06,
"logits/chosen": -1.707132339477539,
"logits/rejected": -1.1808102130889893,
"logps/chosen": -215.00680541992188,
"logps/rejected": -242.06338500976562,
"loss": 0.1289,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.016758020967245102,
"rewards/margins": 0.34254926443099976,
"rewards/rejected": -0.35930731892585754,
"step": 1590
},
{
"epoch": 0.43,
"learning_rate": 3.543346136204545e-06,
"logits/chosen": -1.423949122428894,
"logits/rejected": -0.9146944284439087,
"logps/chosen": -251.60986328125,
"logps/rejected": -246.3702850341797,
"loss": 0.1596,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.025834297761321068,
"rewards/margins": 0.4158453941345215,
"rewards/rejected": -0.39001110196113586,
"step": 1600
},
{
"epoch": 0.43,
"learning_rate": 3.522153641615345e-06,
"logits/chosen": -1.6994764804840088,
"logits/rejected": -1.346040964126587,
"logps/chosen": -256.00592041015625,
"logps/rejected": -256.01300048828125,
"loss": 0.1082,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.021267935633659363,
"rewards/margins": 0.4258691370487213,
"rewards/rejected": -0.40460118651390076,
"step": 1610
},
{
"epoch": 0.43,
"learning_rate": 3.5008725813922383e-06,
"logits/chosen": -1.6546297073364258,
"logits/rejected": -1.3536561727523804,
"logps/chosen": -195.32785034179688,
"logps/rejected": -239.7685546875,
"loss": 0.1337,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.07065526396036148,
"rewards/margins": 0.36889463663101196,
"rewards/rejected": -0.2982393801212311,
"step": 1620
},
{
"epoch": 0.43,
"learning_rate": 3.4795047994562463e-06,
"logits/chosen": -1.8300390243530273,
"logits/rejected": -1.308199167251587,
"logps/chosen": -241.9082489013672,
"logps/rejected": -253.09469604492188,
"loss": 0.1305,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.07730422914028168,
"rewards/margins": 0.380765438079834,
"rewards/rejected": -0.3034612536430359,
"step": 1630
},
{
"epoch": 0.44,
"learning_rate": 3.458052147242494e-06,
"logits/chosen": -1.8392903804779053,
"logits/rejected": -1.4302622079849243,
"logps/chosen": -275.21636962890625,
"logps/rejected": -269.1678161621094,
"loss": 0.1645,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": 0.058570653200149536,
"rewards/margins": 0.3049880862236023,
"rewards/rejected": -0.24641743302345276,
"step": 1640
},
{
"epoch": 0.44,
"learning_rate": 3.436516483539781e-06,
"logits/chosen": -1.8582258224487305,
"logits/rejected": -1.2484705448150635,
"logps/chosen": -289.6882019042969,
"logps/rejected": -296.31927490234375,
"loss": 0.1642,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.04015364870429039,
"rewards/margins": 0.35936877131462097,
"rewards/rejected": -0.3192150890827179,
"step": 1650
},
{
"epoch": 0.44,
"learning_rate": 3.4148996743295305e-06,
"logits/chosen": -1.3779845237731934,
"logits/rejected": -0.940921425819397,
"logps/chosen": -277.53009033203125,
"logps/rejected": -309.68157958984375,
"loss": 0.1488,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.04079504683613777,
"rewards/margins": 0.37808963656425476,
"rewards/rejected": -0.3372945785522461,
"step": 1660
},
{
"epoch": 0.45,
"learning_rate": 3.3932035926241103e-06,
"logits/chosen": -1.6403146982192993,
"logits/rejected": -1.2490508556365967,
"logps/chosen": -270.91436767578125,
"logps/rejected": -305.6242370605469,
"loss": 0.121,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.047386664897203445,
"rewards/margins": 0.4330657422542572,
"rewards/rejected": -0.38567906618118286,
"step": 1670
},
{
"epoch": 0.45,
"learning_rate": 3.3714301183045382e-06,
"logits/chosen": -1.4912313222885132,
"logits/rejected": -1.1307358741760254,
"logps/chosen": -237.8841552734375,
"logps/rejected": -263.740966796875,
"loss": 0.15,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": 0.04139203205704689,
"rewards/margins": 0.30756354331970215,
"rewards/rejected": -0.26617151498794556,
"step": 1680
},
{
"epoch": 0.45,
"learning_rate": 3.349581137957604e-06,
"logits/chosen": -1.5840933322906494,
"logits/rejected": -1.2103346586227417,
"logps/chosen": -267.5755920410156,
"logps/rejected": -250.2428741455078,
"loss": 0.1769,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.055591464042663574,
"rewards/margins": 0.32637280225753784,
"rewards/rejected": -0.27078136801719666,
"step": 1690
},
{
"epoch": 0.45,
"learning_rate": 3.3276585447123957e-06,
"logits/chosen": -1.5938819646835327,
"logits/rejected": -1.0726745128631592,
"logps/chosen": -224.1567840576172,
"logps/rejected": -237.98696899414062,
"loss": 0.1735,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.08412063121795654,
"rewards/margins": 0.37582629919052124,
"rewards/rejected": -0.2917056381702423,
"step": 1700
},
{
"epoch": 0.46,
"learning_rate": 3.3056642380762783e-06,
"logits/chosen": -1.77345871925354,
"logits/rejected": -1.088060736656189,
"logps/chosen": -291.2659606933594,
"logps/rejected": -249.79153442382812,
"loss": 0.118,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.06221754476428032,
"rewards/margins": 0.37834566831588745,
"rewards/rejected": -0.31612807512283325,
"step": 1710
},
{
"epoch": 0.46,
"learning_rate": 3.2836001237702993e-06,
"logits/chosen": -1.7594726085662842,
"logits/rejected": -1.1787726879119873,
"logps/chosen": -252.16256713867188,
"logps/rejected": -276.36224365234375,
"loss": 0.1346,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.03996184095740318,
"rewards/margins": 0.4077722430229187,
"rewards/rejected": -0.36781036853790283,
"step": 1720
},
{
"epoch": 0.46,
"learning_rate": 3.2614681135640696e-06,
"logits/chosen": -1.6873159408569336,
"logits/rejected": -1.11794114112854,
"logps/chosen": -331.9217529296875,
"logps/rejected": -301.14129638671875,
"loss": 0.1049,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.008476448245346546,
"rewards/margins": 0.42262354493141174,
"rewards/rejected": -0.4310999810695648,
"step": 1730
},
{
"epoch": 0.46,
"learning_rate": 3.2392701251101172e-06,
"logits/chosen": -1.977207899093628,
"logits/rejected": -1.3080086708068848,
"logps/chosen": -304.4135437011719,
"logps/rejected": -268.8409729003906,
"loss": 0.1299,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.01740439608693123,
"rewards/margins": 0.3875492513179779,
"rewards/rejected": -0.3701448440551758,
"step": 1740
},
{
"epoch": 0.47,
"learning_rate": 3.217008081777726e-06,
"logits/chosen": -1.6914336681365967,
"logits/rejected": -1.2114366292953491,
"logps/chosen": -291.5057067871094,
"logps/rejected": -306.0625,
"loss": 0.1202,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.02021079882979393,
"rewards/margins": 0.4694185256958008,
"rewards/rejected": -0.44920778274536133,
"step": 1750
},
{
"epoch": 0.47,
"learning_rate": 3.1946839124862873e-06,
"logits/chosen": -1.695512056350708,
"logits/rejected": -1.1958303451538086,
"logps/chosen": -275.1326599121094,
"logps/rejected": -306.881591796875,
"loss": 0.1221,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.05711999535560608,
"rewards/margins": 0.4282289147377014,
"rewards/rejected": -0.4853488802909851,
"step": 1760
},
{
"epoch": 0.47,
"learning_rate": 3.1722995515381644e-06,
"logits/chosen": -1.6805217266082764,
"logits/rejected": -1.306983232498169,
"logps/chosen": -227.75039672851562,
"logps/rejected": -306.4040222167969,
"loss": 0.139,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.028795797377824783,
"rewards/margins": 0.3543476462364197,
"rewards/rejected": -0.38314345479011536,
"step": 1770
},
{
"epoch": 0.47,
"learning_rate": 3.149856938451094e-06,
"logits/chosen": -1.7246391773223877,
"logits/rejected": -1.169914960861206,
"logps/chosen": -301.69305419921875,
"logps/rejected": -275.41375732421875,
"loss": 0.1121,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.026128137484192848,
"rewards/margins": 0.3503456711769104,
"rewards/rejected": -0.3242174983024597,
"step": 1780
},
{
"epoch": 0.48,
"learning_rate": 3.127358017790132e-06,
"logits/chosen": -1.6676828861236572,
"logits/rejected": -1.0200117826461792,
"logps/chosen": -245.4638214111328,
"logps/rejected": -210.855224609375,
"loss": 0.1885,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.012035062536597252,
"rewards/margins": 0.33619141578674316,
"rewards/rejected": -0.32415634393692017,
"step": 1790
},
{
"epoch": 0.48,
"learning_rate": 3.1048047389991693e-06,
"logits/chosen": -1.7602665424346924,
"logits/rejected": -1.1496754884719849,
"logps/chosen": -310.8785095214844,
"logps/rejected": -265.19781494140625,
"loss": 0.1373,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.04427367448806763,
"rewards/margins": 0.34181416034698486,
"rewards/rejected": -0.29754048585891724,
"step": 1800
},
{
"epoch": 0.48,
"learning_rate": 3.082199056232015e-06,
"logits/chosen": -1.7544816732406616,
"logits/rejected": -1.1616899967193604,
"logps/chosen": -374.69097900390625,
"logps/rejected": -291.6187744140625,
"loss": 0.1528,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.029999136924743652,
"rewards/margins": 0.37635478377342224,
"rewards/rejected": -0.346355676651001,
"step": 1810
},
{
"epoch": 0.49,
"learning_rate": 3.059542928183079e-06,
"logits/chosen": -1.6221444606781006,
"logits/rejected": -1.0722931623458862,
"logps/chosen": -284.0110168457031,
"logps/rejected": -225.7612762451172,
"loss": 0.1678,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.0024113513063639402,
"rewards/margins": 0.3399294316768646,
"rewards/rejected": -0.34234076738357544,
"step": 1820
},
{
"epoch": 0.49,
"learning_rate": 3.0368383179176584e-06,
"logits/chosen": -1.583720088005066,
"logits/rejected": -0.8212550282478333,
"logps/chosen": -322.15496826171875,
"logps/rejected": -311.7652282714844,
"loss": 0.1152,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.02223331294953823,
"rewards/margins": 0.37121888995170593,
"rewards/rejected": -0.34898558259010315,
"step": 1830
},
{
"epoch": 0.49,
"learning_rate": 3.0140871927018466e-06,
"logits/chosen": -1.6727044582366943,
"logits/rejected": -1.3792378902435303,
"logps/chosen": -273.42938232421875,
"logps/rejected": -241.69912719726562,
"loss": 0.1723,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.006121881306171417,
"rewards/margins": 0.2609279155731201,
"rewards/rejected": -0.2548060119152069,
"step": 1840
},
{
"epoch": 0.49,
"learning_rate": 2.9912915238320755e-06,
"logits/chosen": -1.4808580875396729,
"logits/rejected": -1.312792181968689,
"logps/chosen": -254.2381134033203,
"logps/rejected": -339.28192138671875,
"loss": 0.1501,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.029569197446107864,
"rewards/margins": 0.403090238571167,
"rewards/rejected": -0.43265944719314575,
"step": 1850
},
{
"epoch": 0.5,
"learning_rate": 2.9684532864643123e-06,
"logits/chosen": -1.459582805633545,
"logits/rejected": -1.37647545337677,
"logps/chosen": -223.28524780273438,
"logps/rejected": -268.646484375,
"loss": 0.2247,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.08847669512033463,
"rewards/margins": 0.2632191777229309,
"rewards/rejected": -0.3516958951950073,
"step": 1860
},
{
"epoch": 0.5,
"learning_rate": 2.945574459442917e-06,
"logits/chosen": -1.753136396408081,
"logits/rejected": -1.0855618715286255,
"logps/chosen": -294.66461181640625,
"logps/rejected": -320.9414978027344,
"loss": 0.1462,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.02983088418841362,
"rewards/margins": 0.34618809819221497,
"rewards/rejected": -0.3760189414024353,
"step": 1870
},
{
"epoch": 0.5,
"learning_rate": 2.922657025129185e-06,
"logits/chosen": -1.909767508506775,
"logits/rejected": -1.1421029567718506,
"logps/chosen": -255.0342559814453,
"logps/rejected": -262.11151123046875,
"loss": 0.0831,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.0237131267786026,
"rewards/margins": 0.4205241799354553,
"rewards/rejected": -0.3968110680580139,
"step": 1880
},
{
"epoch": 0.5,
"learning_rate": 2.8997029692295875e-06,
"logits/chosen": -1.6980245113372803,
"logits/rejected": -1.4830033779144287,
"logps/chosen": -216.7667694091797,
"logps/rejected": -314.1409912109375,
"loss": 0.1811,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.03470195457339287,
"rewards/margins": 0.33328303694725037,
"rewards/rejected": -0.3679850101470947,
"step": 1890
},
{
"epoch": 0.51,
"learning_rate": 2.876714280623708e-06,
"logits/chosen": -1.5797358751296997,
"logits/rejected": -1.1642903089523315,
"logps/chosen": -249.17001342773438,
"logps/rejected": -283.16326904296875,
"loss": 0.1434,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.01058603823184967,
"rewards/margins": 0.3307330906391144,
"rewards/rejected": -0.32014700770378113,
"step": 1900
},
{
"epoch": 0.51,
"learning_rate": 2.8536929511919227e-06,
"logits/chosen": -1.848733901977539,
"logits/rejected": -1.0242502689361572,
"logps/chosen": -324.2806091308594,
"logps/rejected": -243.5651092529297,
"loss": 0.118,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.039272844791412354,
"rewards/margins": 0.3580833077430725,
"rewards/rejected": -0.31881046295166016,
"step": 1910
},
{
"epoch": 0.51,
"learning_rate": 2.8306409756428067e-06,
"logits/chosen": -1.5901503562927246,
"logits/rejected": -1.3664997816085815,
"logps/chosen": -298.6604309082031,
"logps/rejected": -299.3940734863281,
"loss": 0.122,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.025997215881943703,
"rewards/margins": 0.39433664083480835,
"rewards/rejected": -0.368339478969574,
"step": 1920
},
{
"epoch": 0.51,
"learning_rate": 2.807560351340302e-06,
"logits/chosen": -1.6624196767807007,
"logits/rejected": -1.1935114860534668,
"logps/chosen": -271.36151123046875,
"logps/rejected": -300.11187744140625,
"loss": 0.134,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.02754376269876957,
"rewards/margins": 0.36226534843444824,
"rewards/rejected": -0.33472156524658203,
"step": 1930
},
{
"epoch": 0.52,
"learning_rate": 2.7844530781306544e-06,
"logits/chosen": -1.7045570611953735,
"logits/rejected": -0.918795108795166,
"logps/chosen": -334.0121154785156,
"logps/rejected": -271.56024169921875,
"loss": 0.134,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.007435244508087635,
"rewards/margins": 0.3523639440536499,
"rewards/rejected": -0.3597991466522217,
"step": 1940
},
{
"epoch": 0.52,
"learning_rate": 2.761321158169134e-06,
"logits/chosen": -1.7817418575286865,
"logits/rejected": -1.3035436868667603,
"logps/chosen": -290.12249755859375,
"logps/rejected": -248.0795135498047,
"loss": 0.1305,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.0643903911113739,
"rewards/margins": 0.3860814571380615,
"rewards/rejected": -0.32169100642204285,
"step": 1950
},
{
"epoch": 0.52,
"learning_rate": 2.738166595746554e-06,
"logits/chosen": -1.5566781759262085,
"logits/rejected": -1.3110841512680054,
"logps/chosen": -178.944091796875,
"logps/rejected": -215.97299194335938,
"loss": 0.1811,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.07068820297718048,
"rewards/margins": 0.3479788601398468,
"rewards/rejected": -0.2772907018661499,
"step": 1960
},
{
"epoch": 0.53,
"learning_rate": 2.7149913971156105e-06,
"logits/chosen": -1.5236642360687256,
"logits/rejected": -1.2001490592956543,
"logps/chosen": -231.1402130126953,
"logps/rejected": -246.59591674804688,
"loss": 0.1994,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.03314649313688278,
"rewards/margins": 0.3443582057952881,
"rewards/rejected": -0.3112117052078247,
"step": 1970
},
{
"epoch": 0.53,
"learning_rate": 2.6917975703170466e-06,
"logits/chosen": -1.6045347452163696,
"logits/rejected": -1.246459722518921,
"logps/chosen": -213.5222625732422,
"logps/rejected": -257.655029296875,
"loss": 0.1179,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.06492890417575836,
"rewards/margins": 0.34639525413513184,
"rewards/rejected": -0.2814663350582123,
"step": 1980
},
{
"epoch": 0.53,
"learning_rate": 2.668587125005663e-06,
"logits/chosen": -1.5607236623764038,
"logits/rejected": -1.0532560348510742,
"logps/chosen": -249.9466094970703,
"logps/rejected": -285.72845458984375,
"loss": 0.1563,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.05708497762680054,
"rewards/margins": 0.33061715960502625,
"rewards/rejected": -0.2735321819782257,
"step": 1990
},
{
"epoch": 0.53,
"learning_rate": 2.6453620722761897e-06,
"logits/chosen": -1.512621521949768,
"logits/rejected": -1.4036608934402466,
"logps/chosen": -217.87484741210938,
"logps/rejected": -261.9187316894531,
"loss": 0.176,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.025976067408919334,
"rewards/margins": 0.3155084252357483,
"rewards/rejected": -0.3414844870567322,
"step": 2000
},
{
"epoch": 0.54,
"learning_rate": 2.6221244244890336e-06,
"logits/chosen": -1.547397494316101,
"logits/rejected": -1.2165327072143555,
"logps/chosen": -319.05810546875,
"logps/rejected": -346.9829406738281,
"loss": 0.1403,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.00498613715171814,
"rewards/margins": 0.37501031160354614,
"rewards/rejected": -0.370024174451828,
"step": 2010
},
{
"epoch": 0.54,
"learning_rate": 2.5988761950959133e-06,
"logits/chosen": -1.7019367218017578,
"logits/rejected": -1.3284026384353638,
"logps/chosen": -230.9080047607422,
"logps/rejected": -242.2611846923828,
"loss": 0.2393,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.054250530898571014,
"rewards/margins": 0.22886402904987335,
"rewards/rejected": -0.28311458230018616,
"step": 2020
},
{
"epoch": 0.54,
"learning_rate": 2.575619398465402e-06,
"logits/chosen": -1.6088831424713135,
"logits/rejected": -1.3118395805358887,
"logps/chosen": -209.9443817138672,
"logps/rejected": -267.1958312988281,
"loss": 0.2087,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.013560935854911804,
"rewards/margins": 0.3352377414703369,
"rewards/rejected": -0.3216767907142639,
"step": 2030
},
{
"epoch": 0.54,
"learning_rate": 2.5523560497083927e-06,
"logits/chosen": -1.6162147521972656,
"logits/rejected": -1.2510854005813599,
"logps/chosen": -231.2008819580078,
"logps/rejected": -321.4208068847656,
"loss": 0.1237,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.03159638121724129,
"rewards/margins": 0.33668285608291626,
"rewards/rejected": -0.30508649349212646,
"step": 2040
},
{
"epoch": 0.55,
"learning_rate": 2.5290881645034932e-06,
"logits/chosen": -1.6034374237060547,
"logits/rejected": -1.2775676250457764,
"logps/chosen": -293.4057312011719,
"logps/rejected": -363.25360107421875,
"loss": 0.1661,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.00653398921713233,
"rewards/margins": 0.32143646478652954,
"rewards/rejected": -0.3279704451560974,
"step": 2050
},
{
"epoch": 0.55,
"learning_rate": 2.5058177589223766e-06,
"logits/chosen": -1.5212050676345825,
"logits/rejected": -1.210669755935669,
"logps/chosen": -240.904296875,
"logps/rejected": -274.26605224609375,
"loss": 0.1595,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.03204118087887764,
"rewards/margins": 0.3350418508052826,
"rewards/rejected": -0.3030007481575012,
"step": 2060
},
{
"epoch": 0.55,
"learning_rate": 2.482546849255096e-06,
"logits/chosen": -1.7197357416152954,
"logits/rejected": -1.0405242443084717,
"logps/chosen": -314.89984130859375,
"logps/rejected": -287.49224853515625,
"loss": 0.0899,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.03998409956693649,
"rewards/margins": 0.43929505348205566,
"rewards/rejected": -0.39931100606918335,
"step": 2070
},
{
"epoch": 0.55,
"learning_rate": 2.4592774518353858e-06,
"logits/chosen": -1.688256025314331,
"logits/rejected": -1.1770254373550415,
"logps/chosen": -238.84921264648438,
"logps/rejected": -256.46063232421875,
"loss": 0.142,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.046583134680986404,
"rewards/margins": 0.42613476514816284,
"rewards/rejected": -0.37955164909362793,
"step": 2080
},
{
"epoch": 0.56,
"learning_rate": 2.436011582865945e-06,
"logits/chosen": -1.4531993865966797,
"logits/rejected": -0.9476866722106934,
"logps/chosen": -272.5257568359375,
"logps/rejected": -301.6541442871094,
"loss": 0.107,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.08881069719791412,
"rewards/margins": 0.4413018226623535,
"rewards/rejected": -0.352491170167923,
"step": 2090
},
{
"epoch": 0.56,
"learning_rate": 2.4127512582437486e-06,
"logits/chosen": -1.719260811805725,
"logits/rejected": -1.2589495182037354,
"logps/chosen": -293.53753662109375,
"logps/rejected": -286.5969543457031,
"loss": 0.1363,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.01633891463279724,
"rewards/margins": 0.3913685381412506,
"rewards/rejected": -0.37502965331077576,
"step": 2100
},
{
"epoch": 0.56,
"learning_rate": 2.3894984933853734e-06,
"logits/chosen": -1.6526820659637451,
"logits/rejected": -1.3534865379333496,
"logps/chosen": -206.388671875,
"logps/rejected": -241.95187377929688,
"loss": 0.1525,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.007084892597049475,
"rewards/margins": 0.36700910329818726,
"rewards/rejected": -0.35992416739463806,
"step": 2110
},
{
"epoch": 0.57,
"learning_rate": 2.366255303052377e-06,
"logits/chosen": -1.6136153936386108,
"logits/rejected": -1.188907504081726,
"logps/chosen": -286.76373291015625,
"logps/rejected": -252.4059600830078,
"loss": 0.1859,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.005851163994520903,
"rewards/margins": 0.316637247800827,
"rewards/rejected": -0.31078606843948364,
"step": 2120
},
{
"epoch": 0.57,
"learning_rate": 2.3430237011767166e-06,
"logits/chosen": -1.654313325881958,
"logits/rejected": -1.2479978799819946,
"logps/chosen": -226.00390625,
"logps/rejected": -232.9287109375,
"loss": 0.1541,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": 0.08312289416790009,
"rewards/margins": 0.3975786864757538,
"rewards/rejected": -0.3144558072090149,
"step": 2130
},
{
"epoch": 0.57,
"learning_rate": 2.319805700686257e-06,
"logits/chosen": -1.6342413425445557,
"logits/rejected": -1.109515905380249,
"logps/chosen": -234.4524383544922,
"logps/rejected": -233.2206573486328,
"loss": 0.1168,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.08855343610048294,
"rewards/margins": 0.38437074422836304,
"rewards/rejected": -0.2958173155784607,
"step": 2140
},
{
"epoch": 0.57,
"learning_rate": 2.296603313330355e-06,
"logits/chosen": -1.8358131647109985,
"logits/rejected": -1.035390019416809,
"logps/chosen": -394.54071044921875,
"logps/rejected": -347.41680908203125,
"loss": 0.2742,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.042184554040431976,
"rewards/margins": 0.3534103333950043,
"rewards/rejected": -0.3112257719039917,
"step": 2150
},
{
"epoch": 0.58,
"learning_rate": 2.2734185495055503e-06,
"logits/chosen": -1.5863606929779053,
"logits/rejected": -1.3380165100097656,
"logps/chosen": -271.6787414550781,
"logps/rejected": -278.33502197265625,
"loss": 0.1506,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.020747099071741104,
"rewards/margins": 0.34211981296539307,
"rewards/rejected": -0.32137271761894226,
"step": 2160
},
{
"epoch": 0.58,
"learning_rate": 2.250253418081373e-06,
"logits/chosen": -1.695081353187561,
"logits/rejected": -1.1732103824615479,
"logps/chosen": -280.98675537109375,
"logps/rejected": -266.2809143066406,
"loss": 0.1702,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.008993232622742653,
"rewards/margins": 0.3143147826194763,
"rewards/rejected": -0.3053215444087982,
"step": 2170
},
{
"epoch": 0.58,
"learning_rate": 2.22710992622628e-06,
"logits/chosen": -1.6708568334579468,
"logits/rejected": -1.061140775680542,
"logps/chosen": -281.99456787109375,
"logps/rejected": -249.1426544189453,
"loss": 0.1246,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.04111826419830322,
"rewards/margins": 0.35372304916381836,
"rewards/rejected": -0.31260478496551514,
"step": 2180
},
{
"epoch": 0.58,
"learning_rate": 2.2039900792337477e-06,
"logits/chosen": -1.6424753665924072,
"logits/rejected": -1.2045528888702393,
"logps/chosen": -228.9128875732422,
"logps/rejected": -266.5892028808594,
"loss": 0.1208,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.0656760185956955,
"rewards/margins": 0.4023992121219635,
"rewards/rejected": -0.3367232382297516,
"step": 2190
},
{
"epoch": 0.59,
"learning_rate": 2.1808958803485134e-06,
"logits/chosen": -1.7162901163101196,
"logits/rejected": -1.1374019384384155,
"logps/chosen": -264.79522705078125,
"logps/rejected": -279.41595458984375,
"loss": 0.1364,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.055478811264038086,
"rewards/margins": 0.4204636216163635,
"rewards/rejected": -0.36498481035232544,
"step": 2200
},
{
"epoch": 0.59,
"learning_rate": 2.157829330593008e-06,
"logits/chosen": -1.537788987159729,
"logits/rejected": -1.129476547241211,
"logps/chosen": -278.3310241699219,
"logps/rejected": -257.82232666015625,
"loss": 0.1514,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.022755134850740433,
"rewards/margins": 0.32495397329330444,
"rewards/rejected": -0.34770917892456055,
"step": 2210
},
{
"epoch": 0.59,
"learning_rate": 2.134792428593971e-06,
"logits/chosen": -1.6505171060562134,
"logits/rejected": -1.1392511129379272,
"logps/chosen": -315.64190673828125,
"logps/rejected": -261.23101806640625,
"loss": 0.1454,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.04185132309794426,
"rewards/margins": 0.35440436005592346,
"rewards/rejected": -0.3125530779361725,
"step": 2220
},
{
"epoch": 0.59,
"learning_rate": 2.1117871704092818e-06,
"logits/chosen": -1.822789192199707,
"logits/rejected": -1.2018978595733643,
"logps/chosen": -249.8803253173828,
"logps/rejected": -278.4112243652344,
"loss": 0.1295,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.07924628257751465,
"rewards/margins": 0.4235721528530121,
"rewards/rejected": -0.3443259298801422,
"step": 2230
},
{
"epoch": 0.6,
"learning_rate": 2.0888155493550027e-06,
"logits/chosen": -1.7266429662704468,
"logits/rejected": -1.0663232803344727,
"logps/chosen": -269.5586853027344,
"logps/rejected": -230.1761474609375,
"loss": 0.1131,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.0925656110048294,
"rewards/margins": 0.38963964581489563,
"rewards/rejected": -0.2970740795135498,
"step": 2240
},
{
"epoch": 0.6,
"learning_rate": 2.0658795558326745e-06,
"logits/chosen": -1.8484570980072021,
"logits/rejected": -1.2969257831573486,
"logps/chosen": -248.79464721679688,
"logps/rejected": -239.7518768310547,
"loss": 0.1618,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.05543617531657219,
"rewards/margins": 0.3949527144432068,
"rewards/rejected": -0.3395165801048279,
"step": 2250
},
{
"epoch": 0.6,
"learning_rate": 2.0429811771568468e-06,
"logits/chosen": -1.5278081893920898,
"logits/rejected": -1.0012762546539307,
"logps/chosen": -280.8786315917969,
"logps/rejected": -276.75689697265625,
"loss": 0.0882,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.07547347992658615,
"rewards/margins": 0.4585336148738861,
"rewards/rejected": -0.38306012749671936,
"step": 2260
},
{
"epoch": 0.61,
"learning_rate": 2.0201223973828917e-06,
"logits/chosen": -1.6746526956558228,
"logits/rejected": -1.205840826034546,
"logps/chosen": -274.74298095703125,
"logps/rejected": -298.1522216796875,
"loss": 0.1112,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.032690681517124176,
"rewards/margins": 0.4341716170310974,
"rewards/rejected": -0.4014809727668762,
"step": 2270
},
{
"epoch": 0.61,
"learning_rate": 1.997305197135089e-06,
"logits/chosen": -1.420925259590149,
"logits/rejected": -1.1142023801803589,
"logps/chosen": -235.92758178710938,
"logps/rejected": -251.2263641357422,
"loss": 0.1505,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.04943353682756424,
"rewards/margins": 0.3853815197944641,
"rewards/rejected": -0.33594799041748047,
"step": 2280
},
{
"epoch": 0.61,
"learning_rate": 1.9745315534350157e-06,
"logits/chosen": -1.5054067373275757,
"logits/rejected": -1.0819201469421387,
"logps/chosen": -313.4949951171875,
"logps/rejected": -295.75164794921875,
"loss": 0.1577,
"rewards/accuracies": 0.875,
"rewards/chosen": -3.1620264053344727e-05,
"rewards/margins": 0.33664727210998535,
"rewards/rejected": -0.3366788923740387,
"step": 2290
},
{
"epoch": 0.61,
"learning_rate": 1.9518034395302413e-06,
"logits/chosen": -1.6651229858398438,
"logits/rejected": -1.2321889400482178,
"logps/chosen": -296.817138671875,
"logps/rejected": -304.33294677734375,
"loss": 0.1157,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.07290974259376526,
"rewards/margins": 0.38127079606056213,
"rewards/rejected": -0.3083610236644745,
"step": 2300
},
{
"epoch": 0.62,
"learning_rate": 1.9291228247233607e-06,
"logits/chosen": -1.7819957733154297,
"logits/rejected": -1.2075005769729614,
"logps/chosen": -259.632080078125,
"logps/rejected": -265.5983581542969,
"loss": 0.1464,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.04883255437016487,
"rewards/margins": 0.39770543575286865,
"rewards/rejected": -0.3488728702068329,
"step": 2310
},
{
"epoch": 0.62,
"learning_rate": 1.9064916742013515e-06,
"logits/chosen": -1.7982776165008545,
"logits/rejected": -1.0426654815673828,
"logps/chosen": -284.5262756347656,
"logps/rejected": -271.22930908203125,
"loss": 0.1141,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.08063653856515884,
"rewards/margins": 0.4655955731868744,
"rewards/rejected": -0.38495901226997375,
"step": 2320
},
{
"epoch": 0.62,
"learning_rate": 1.883911948865306e-06,
"logits/chosen": -1.424214243888855,
"logits/rejected": -1.0795605182647705,
"logps/chosen": -344.02264404296875,
"logps/rejected": -331.96990966796875,
"loss": 0.3091,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": 0.012640709057450294,
"rewards/margins": 0.2547581195831299,
"rewards/rejected": -0.24211737513542175,
"step": 2330
},
{
"epoch": 0.62,
"learning_rate": 1.8613856051605242e-06,
"logits/chosen": -1.6110725402832031,
"logits/rejected": -1.2309287786483765,
"logps/chosen": -276.41448974609375,
"logps/rejected": -281.2205505371094,
"loss": 0.1099,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.046147290617227554,
"rewards/margins": 0.3609221577644348,
"rewards/rejected": -0.31477484107017517,
"step": 2340
},
{
"epoch": 0.63,
"learning_rate": 1.8389145949069953e-06,
"logits/chosen": -1.9351005554199219,
"logits/rejected": -1.308622121810913,
"logps/chosen": -283.0077209472656,
"logps/rejected": -286.6006164550781,
"loss": 0.1261,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.1078963652253151,
"rewards/margins": 0.4045773446559906,
"rewards/rejected": -0.2966809570789337,
"step": 2350
},
{
"epoch": 0.63,
"learning_rate": 1.816500865130279e-06,
"logits/chosen": -1.6536674499511719,
"logits/rejected": -1.3896677494049072,
"logps/chosen": -203.09043884277344,
"logps/rejected": -169.11422729492188,
"loss": 0.1889,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": 0.039239563047885895,
"rewards/margins": 0.27253058552742004,
"rewards/rejected": -0.23329100012779236,
"step": 2360
},
{
"epoch": 0.63,
"learning_rate": 1.7941463578928088e-06,
"logits/chosen": -1.620586633682251,
"logits/rejected": -1.1168583631515503,
"logps/chosen": -280.3619079589844,
"logps/rejected": -252.0579071044922,
"loss": 0.1894,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.06918063014745712,
"rewards/margins": 0.3512100577354431,
"rewards/rejected": -0.2820294499397278,
"step": 2370
},
{
"epoch": 0.63,
"learning_rate": 1.7718530101256115e-06,
"logits/chosen": -1.887487769126892,
"logits/rejected": -1.241081714630127,
"logps/chosen": -284.033935546875,
"logps/rejected": -275.2795104980469,
"loss": 0.1358,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.09624558687210083,
"rewards/margins": 0.3296979069709778,
"rewards/rejected": -0.23345234990119934,
"step": 2380
},
{
"epoch": 0.64,
"learning_rate": 1.7496227534604859e-06,
"logits/chosen": -1.785048246383667,
"logits/rejected": -1.2693895101547241,
"logps/chosen": -263.9631042480469,
"logps/rejected": -324.4400634765625,
"loss": 0.1713,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.08735568076372147,
"rewards/margins": 0.380867063999176,
"rewards/rejected": -0.29351136088371277,
"step": 2390
},
{
"epoch": 0.64,
"learning_rate": 1.7274575140626318e-06,
"logits/chosen": -1.5414637327194214,
"logits/rejected": -0.9286526441574097,
"logps/chosen": -300.5523986816406,
"logps/rejected": -227.65713500976562,
"loss": 0.1229,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.10164747387170792,
"rewards/margins": 0.3701394498348236,
"rewards/rejected": -0.2684919834136963,
"step": 2400
},
{
"epoch": 0.64,
"learning_rate": 1.7053592124637557e-06,
"logits/chosen": -1.7873646020889282,
"logits/rejected": -1.1277539730072021,
"logps/chosen": -322.7608337402344,
"logps/rejected": -267.81787109375,
"loss": 0.148,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.05880016088485718,
"rewards/margins": 0.3506908118724823,
"rewards/rejected": -0.2918906509876251,
"step": 2410
},
{
"epoch": 0.65,
"learning_rate": 1.6833297633956647e-06,
"logits/chosen": -1.532529592514038,
"logits/rejected": -1.4235907793045044,
"logps/chosen": -165.92489624023438,
"logps/rejected": -237.23782348632812,
"loss": 0.1796,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.004235130734741688,
"rewards/margins": 0.31066879630088806,
"rewards/rejected": -0.30643370747566223,
"step": 2420
},
{
"epoch": 0.65,
"learning_rate": 1.661371075624363e-06,
"logits/chosen": -1.508122444152832,
"logits/rejected": -1.1396775245666504,
"logps/chosen": -234.5397186279297,
"logps/rejected": -293.2003479003906,
"loss": 0.1418,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.06259538978338242,
"rewards/margins": 0.40436553955078125,
"rewards/rejected": -0.3417701721191406,
"step": 2430
},
{
"epoch": 0.65,
"learning_rate": 1.6394850517846621e-06,
"logits/chosen": -1.6401771306991577,
"logits/rejected": -0.8912142515182495,
"logps/chosen": -283.13238525390625,
"logps/rejected": -221.9040985107422,
"loss": 0.1623,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.08341696858406067,
"rewards/margins": 0.4074879288673401,
"rewards/rejected": -0.32407090067863464,
"step": 2440
},
{
"epoch": 0.65,
"learning_rate": 1.6176735882153284e-06,
"logits/chosen": -1.9544626474380493,
"logits/rejected": -1.3714215755462646,
"logps/chosen": -302.2862548828125,
"logps/rejected": -252.0078582763672,
"loss": 0.1017,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.09628540277481079,
"rewards/margins": 0.3641790747642517,
"rewards/rejected": -0.2678936719894409,
"step": 2450
},
{
"epoch": 0.66,
"learning_rate": 1.5959385747947697e-06,
"logits/chosen": -1.4549901485443115,
"logits/rejected": -1.019054651260376,
"logps/chosen": -258.96661376953125,
"logps/rejected": -268.4969177246094,
"loss": 0.1197,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.13393627107143402,
"rewards/margins": 0.4325905740261078,
"rewards/rejected": -0.29865431785583496,
"step": 2460
},
{
"epoch": 0.66,
"learning_rate": 1.5742818947772875e-06,
"logits/chosen": -1.6369373798370361,
"logits/rejected": -1.1893460750579834,
"logps/chosen": -270.46185302734375,
"logps/rejected": -253.77633666992188,
"loss": 0.1302,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.04096044600009918,
"rewards/margins": 0.3416442275047302,
"rewards/rejected": -0.30068379640579224,
"step": 2470
},
{
"epoch": 0.66,
"learning_rate": 1.552705424629898e-06,
"logits/chosen": -1.5808465480804443,
"logits/rejected": -1.141884446144104,
"logps/chosen": -285.2702941894531,
"logps/rejected": -277.57745361328125,
"loss": 0.1158,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.059014301747083664,
"rewards/margins": 0.3862994611263275,
"rewards/rejected": -0.32728514075279236,
"step": 2480
},
{
"epoch": 0.66,
"learning_rate": 1.5312110338697427e-06,
"logits/chosen": -1.755192518234253,
"logits/rejected": -1.2506240606307983,
"logps/chosen": -246.61972045898438,
"logps/rejected": -237.236328125,
"loss": 0.1112,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.06155257299542427,
"rewards/margins": 0.430799663066864,
"rewards/rejected": -0.3692471385002136,
"step": 2490
},
{
"epoch": 0.67,
"learning_rate": 1.509800584902108e-06,
"logits/chosen": -1.532773733139038,
"logits/rejected": -1.2026466131210327,
"logps/chosen": -243.93820190429688,
"logps/rejected": -262.2892150878906,
"loss": 0.1839,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.07341442257165909,
"rewards/margins": 0.35419246554374695,
"rewards/rejected": -0.28077805042266846,
"step": 2500
},
{
"epoch": 0.67,
"learning_rate": 1.4884759328590476e-06,
"logits/chosen": -1.8502527475357056,
"logits/rejected": -1.1260955333709717,
"logps/chosen": -276.49945068359375,
"logps/rejected": -261.7967529296875,
"loss": 0.1949,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.07880761474370956,
"rewards/margins": 0.402778685092926,
"rewards/rejected": -0.32397109270095825,
"step": 2510
},
{
"epoch": 0.67,
"learning_rate": 1.467238925438646e-06,
"logits/chosen": -1.6124885082244873,
"logits/rejected": -1.1764719486236572,
"logps/chosen": -242.35806274414062,
"logps/rejected": -281.0080261230469,
"loss": 0.1401,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.05401073768734932,
"rewards/margins": 0.40248528122901917,
"rewards/rejected": -0.34847456216812134,
"step": 2520
},
{
"epoch": 0.67,
"learning_rate": 1.446091402744923e-06,
"logits/chosen": -1.5193445682525635,
"logits/rejected": -0.925061821937561,
"logps/chosen": -275.9288024902344,
"logps/rejected": -256.40155029296875,
"loss": 0.097,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.02620106376707554,
"rewards/margins": 0.3423638343811035,
"rewards/rejected": -0.31616276502609253,
"step": 2530
},
{
"epoch": 0.68,
"learning_rate": 1.4250351971283937e-06,
"logits/chosen": -1.7999191284179688,
"logits/rejected": -1.2691766023635864,
"logps/chosen": -227.58023071289062,
"logps/rejected": -259.0445556640625,
"loss": 0.1104,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.06122167035937309,
"rewards/margins": 0.41422295570373535,
"rewards/rejected": -0.35300129652023315,
"step": 2540
},
{
"epoch": 0.68,
"learning_rate": 1.4040721330273063e-06,
"logits/chosen": -1.9836969375610352,
"logits/rejected": -1.2529933452606201,
"logps/chosen": -243.0897216796875,
"logps/rejected": -234.3355712890625,
"loss": 0.1477,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.0859227403998375,
"rewards/margins": 0.4090031683444977,
"rewards/rejected": -0.323080450296402,
"step": 2550
},
{
"epoch": 0.68,
"learning_rate": 1.3832040268095589e-06,
"logits/chosen": -1.5172902345657349,
"logits/rejected": -1.0578333139419556,
"logps/chosen": -260.90411376953125,
"logps/rejected": -258.8797302246094,
"loss": 0.1353,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.05479772016406059,
"rewards/margins": 0.37679368257522583,
"rewards/rejected": -0.32199597358703613,
"step": 2560
},
{
"epoch": 0.69,
"learning_rate": 1.362432686615316e-06,
"logits/chosen": -1.5748417377471924,
"logits/rejected": -1.0296505689620972,
"logps/chosen": -261.697021484375,
"logps/rejected": -210.9502410888672,
"loss": 0.1437,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": 0.08591978251934052,
"rewards/margins": 0.34443560242652893,
"rewards/rejected": -0.2585158050060272,
"step": 2570
},
{
"epoch": 0.69,
"learning_rate": 1.3417599122003464e-06,
"logits/chosen": -1.7466052770614624,
"logits/rejected": -1.1247678995132446,
"logps/chosen": -253.7733917236328,
"logps/rejected": -230.14089965820312,
"loss": 0.1104,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.10326331853866577,
"rewards/margins": 0.384945809841156,
"rewards/rejected": -0.28168249130249023,
"step": 2580
},
{
"epoch": 0.69,
"learning_rate": 1.3211874947800747e-06,
"logits/chosen": -1.6235640048980713,
"logits/rejected": -1.2449982166290283,
"logps/chosen": -286.81756591796875,
"logps/rejected": -286.7332763671875,
"loss": 0.2038,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.03231514245271683,
"rewards/margins": 0.31858521699905396,
"rewards/rejected": -0.2862700819969177,
"step": 2590
},
{
"epoch": 0.69,
"learning_rate": 1.3007172168743854e-06,
"logits/chosen": -1.9692258834838867,
"logits/rejected": -1.3329085111618042,
"logps/chosen": -256.65008544921875,
"logps/rejected": -220.8751678466797,
"loss": 0.1944,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.06459324806928635,
"rewards/margins": 0.3410438299179077,
"rewards/rejected": -0.27645057439804077,
"step": 2600
},
{
"epoch": 0.7,
"learning_rate": 1.280350852153168e-06,
"logits/chosen": -1.6254962682724,
"logits/rejected": -1.4488201141357422,
"logps/chosen": -212.2681427001953,
"logps/rejected": -273.9403076171875,
"loss": 0.1321,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.09227583557367325,
"rewards/margins": 0.3278459906578064,
"rewards/rejected": -0.23557014763355255,
"step": 2610
},
{
"epoch": 0.7,
"learning_rate": 1.260090165282645e-06,
"logits/chosen": -1.6674009561538696,
"logits/rejected": -1.145491361618042,
"logps/chosen": -273.3908996582031,
"logps/rejected": -231.9703369140625,
"loss": 0.189,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.08204850554466248,
"rewards/margins": 0.3708895742893219,
"rewards/rejected": -0.2888410687446594,
"step": 2620
},
{
"epoch": 0.7,
"learning_rate": 1.2399369117724582e-06,
"logits/chosen": -1.8269517421722412,
"logits/rejected": -1.1365658044815063,
"logps/chosen": -305.4709167480469,
"logps/rejected": -273.973388671875,
"loss": 0.1584,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.08552964776754379,
"rewards/margins": 0.3883412480354309,
"rewards/rejected": -0.30281156301498413,
"step": 2630
},
{
"epoch": 0.7,
"learning_rate": 1.2198928378235717e-06,
"logits/chosen": -1.6987979412078857,
"logits/rejected": -1.3719052076339722,
"logps/chosen": -225.90505981445312,
"logps/rejected": -248.9863739013672,
"loss": 0.1964,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": 0.09808424860239029,
"rewards/margins": 0.2932734787464142,
"rewards/rejected": -0.1951892375946045,
"step": 2640
},
{
"epoch": 0.71,
"learning_rate": 1.1999596801769617e-06,
"logits/chosen": -1.6622822284698486,
"logits/rejected": -1.46449875831604,
"logps/chosen": -194.4739227294922,
"logps/rejected": -270.63922119140625,
"loss": 0.16,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.08343388140201569,
"rewards/margins": 0.3558526933193207,
"rewards/rejected": -0.2724188268184662,
"step": 2650
},
{
"epoch": 0.71,
"learning_rate": 1.1801391659631423e-06,
"logits/chosen": -1.4197065830230713,
"logits/rejected": -0.8729090690612793,
"logps/chosen": -261.99517822265625,
"logps/rejected": -240.2467041015625,
"loss": 0.1424,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.09511663019657135,
"rewards/margins": 0.3631947338581085,
"rewards/rejected": -0.2680780589580536,
"step": 2660
},
{
"epoch": 0.71,
"learning_rate": 1.160433012552508e-06,
"logits/chosen": -1.6888864040374756,
"logits/rejected": -1.1725207567214966,
"logps/chosen": -265.87957763671875,
"logps/rejected": -228.31982421875,
"loss": 0.115,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.11969444900751114,
"rewards/margins": 0.4062970280647278,
"rewards/rejected": -0.28660255670547485,
"step": 2670
},
{
"epoch": 0.71,
"learning_rate": 1.1408429274065418e-06,
"logits/chosen": -1.6261924505233765,
"logits/rejected": -1.3828227519989014,
"logps/chosen": -223.134521484375,
"logps/rejected": -277.44903564453125,
"loss": 0.1311,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.11289075762033463,
"rewards/margins": 0.34422147274017334,
"rewards/rejected": -0.23133070766925812,
"step": 2680
},
{
"epoch": 0.72,
"learning_rate": 1.1213706079298566e-06,
"logits/chosen": -1.8407529592514038,
"logits/rejected": -1.3841874599456787,
"logps/chosen": -282.0005187988281,
"logps/rejected": -308.8989562988281,
"loss": 0.1637,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.08910055458545685,
"rewards/margins": 0.3223797082901001,
"rewards/rejected": -0.23327915370464325,
"step": 2690
},
{
"epoch": 0.72,
"learning_rate": 1.1020177413231334e-06,
"logits/chosen": -1.6514075994491577,
"logits/rejected": -1.1626676321029663,
"logps/chosen": -206.6435546875,
"logps/rejected": -248.57595825195312,
"loss": 0.0849,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.0976477712392807,
"rewards/margins": 0.3547229468822479,
"rewards/rejected": -0.25707516074180603,
"step": 2700
},
{
"epoch": 0.72,
"learning_rate": 1.0827860044369226e-06,
"logits/chosen": -1.8002774715423584,
"logits/rejected": -1.2231605052947998,
"logps/chosen": -263.72991943359375,
"logps/rejected": -265.7281799316406,
"loss": 0.0857,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.09039691835641861,
"rewards/margins": 0.3749600052833557,
"rewards/rejected": -0.2845631241798401,
"step": 2710
},
{
"epoch": 0.73,
"learning_rate": 1.06367706362636e-06,
"logits/chosen": -1.7300952672958374,
"logits/rejected": -1.0147895812988281,
"logps/chosen": -234.8489532470703,
"logps/rejected": -245.09164428710938,
"loss": 0.0854,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.09726179391145706,
"rewards/margins": 0.37331247329711914,
"rewards/rejected": -0.27605074644088745,
"step": 2720
},
{
"epoch": 0.73,
"learning_rate": 1.0446925746067768e-06,
"logits/chosen": -1.8114553689956665,
"logits/rejected": -1.3846733570098877,
"logps/chosen": -270.171142578125,
"logps/rejected": -253.19253540039062,
"loss": 0.1582,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.08086392283439636,
"rewards/margins": 0.3226754665374756,
"rewards/rejected": -0.24181151390075684,
"step": 2730
},
{
"epoch": 0.73,
"learning_rate": 1.0258341823102418e-06,
"logits/chosen": -1.5032013654708862,
"logits/rejected": -1.2384767532348633,
"logps/chosen": -202.64346313476562,
"logps/rejected": -241.52383422851562,
"loss": 0.1571,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.04403448849916458,
"rewards/margins": 0.30034139752388,
"rewards/rejected": -0.25630688667297363,
"step": 2740
},
{
"epoch": 0.73,
"learning_rate": 1.0071035207430352e-06,
"logits/chosen": -1.6960475444793701,
"logits/rejected": -1.202266812324524,
"logps/chosen": -228.84579467773438,
"logps/rejected": -282.767333984375,
"loss": 0.1184,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.08044974505901337,
"rewards/margins": 0.4057890474796295,
"rewards/rejected": -0.32533928751945496,
"step": 2750
},
{
"epoch": 0.74,
"learning_rate": 9.88502212844063e-07,
"logits/chosen": -1.7069820165634155,
"logits/rejected": -1.3219921588897705,
"logps/chosen": -232.90579223632812,
"logps/rejected": -257.5270690917969,
"loss": 0.1422,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.050495147705078125,
"rewards/margins": 0.3361809551715851,
"rewards/rejected": -0.2856857478618622,
"step": 2760
},
{
"epoch": 0.74,
"learning_rate": 9.700318703442437e-07,
"logits/chosen": -1.813307762145996,
"logits/rejected": -1.2688970565795898,
"logps/chosen": -252.4823760986328,
"logps/rejected": -221.725830078125,
"loss": 0.1892,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.03987887501716614,
"rewards/margins": 0.2988082468509674,
"rewards/rejected": -0.2589293420314789,
"step": 2770
},
{
"epoch": 0.74,
"learning_rate": 9.516940936268504e-07,
"logits/chosen": -1.8856050968170166,
"logits/rejected": -1.4962949752807617,
"logps/chosen": -211.00314331054688,
"logps/rejected": -229.956298828125,
"loss": 0.2038,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.026142999529838562,
"rewards/margins": 0.2894168496131897,
"rewards/rejected": -0.2632738947868347,
"step": 2780
},
{
"epoch": 0.74,
"learning_rate": 9.334904715888496e-07,
"logits/chosen": -1.7890255451202393,
"logits/rejected": -1.1871464252471924,
"logps/chosen": -290.72113037109375,
"logps/rejected": -280.96856689453125,
"loss": 0.1023,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.048764027655124664,
"rewards/margins": 0.323917955160141,
"rewards/rejected": -0.2751539349555969,
"step": 2790
},
{
"epoch": 0.75,
"learning_rate": 9.154225815032242e-07,
"logits/chosen": -1.4765609502792358,
"logits/rejected": -1.052433729171753,
"logps/chosen": -270.6768493652344,
"logps/rejected": -251.9602508544922,
"loss": 0.1566,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.08185350149869919,
"rewards/margins": 0.3346666097640991,
"rewards/rejected": -0.2528131604194641,
"step": 2800
},
{
"epoch": 0.75,
"learning_rate": 8.974919888823164e-07,
"logits/chosen": -1.5708179473876953,
"logits/rejected": -1.1366405487060547,
"logps/chosen": -260.322265625,
"logps/rejected": -328.4196472167969,
"loss": 0.1115,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.075380340218544,
"rewards/margins": 0.4507225453853607,
"rewards/rejected": -0.3753421902656555,
"step": 2810
},
{
"epoch": 0.75,
"learning_rate": 8.797002473421729e-07,
"logits/chosen": -1.6871554851531982,
"logits/rejected": -1.015560269355774,
"logps/chosen": -312.00396728515625,
"logps/rejected": -289.4263000488281,
"loss": 0.0967,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.08846127986907959,
"rewards/margins": 0.4753798544406891,
"rewards/rejected": -0.3869186043739319,
"step": 2820
},
{
"epoch": 0.75,
"learning_rate": 8.620488984679378e-07,
"logits/chosen": -1.5905934572219849,
"logits/rejected": -1.0190634727478027,
"logps/chosen": -270.3796081542969,
"logps/rejected": -255.5897216796875,
"loss": 0.1175,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.06361018866300583,
"rewards/margins": 0.41740432381629944,
"rewards/rejected": -0.3537940979003906,
"step": 2830
},
{
"epoch": 0.76,
"learning_rate": 8.445394716802754e-07,
"logits/chosen": -1.716925859451294,
"logits/rejected": -1.2719004154205322,
"logps/chosen": -282.26434326171875,
"logps/rejected": -315.83294677734375,
"loss": 0.1543,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.10373795032501221,
"rewards/margins": 0.4591018557548523,
"rewards/rejected": -0.3553639352321625,
"step": 2840
},
{
"epoch": 0.76,
"learning_rate": 8.271734841028553e-07,
"logits/chosen": -1.6560224294662476,
"logits/rejected": -1.2010629177093506,
"logps/chosen": -273.56549072265625,
"logps/rejected": -295.66180419921875,
"loss": 0.1,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.09685816615819931,
"rewards/margins": 0.42575687170028687,
"rewards/rejected": -0.32889872789382935,
"step": 2850
},
{
"epoch": 0.76,
"learning_rate": 8.099524404308948e-07,
"logits/chosen": -1.5945422649383545,
"logits/rejected": -1.19219172000885,
"logps/chosen": -182.81103515625,
"logps/rejected": -238.70919799804688,
"loss": 0.161,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.06724058091640472,
"rewards/margins": 0.31565195322036743,
"rewards/rejected": -0.24841134250164032,
"step": 2860
},
{
"epoch": 0.77,
"learning_rate": 7.928778328007918e-07,
"logits/chosen": -1.9054105281829834,
"logits/rejected": -1.2900282144546509,
"logps/chosen": -260.6278991699219,
"logps/rejected": -260.5713806152344,
"loss": 0.1954,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.06147562339901924,
"rewards/margins": 0.29727303981781006,
"rewards/rejected": -0.23579740524291992,
"step": 2870
},
{
"epoch": 0.77,
"learning_rate": 7.759511406608255e-07,
"logits/chosen": -1.8338701725006104,
"logits/rejected": -1.3442243337631226,
"logps/chosen": -289.9970703125,
"logps/rejected": -252.2522735595703,
"loss": 0.189,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": 0.06250108033418655,
"rewards/margins": 0.26397019624710083,
"rewards/rejected": -0.20146910846233368,
"step": 2880
},
{
"epoch": 0.77,
"learning_rate": 7.591738306429769e-07,
"logits/chosen": -1.7106062173843384,
"logits/rejected": -0.9574893712997437,
"logps/chosen": -306.48663330078125,
"logps/rejected": -267.07965087890625,
"loss": 0.1403,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.03719034045934677,
"rewards/margins": 0.3732340633869171,
"rewards/rejected": -0.33604371547698975,
"step": 2890
},
{
"epoch": 0.77,
"learning_rate": 7.425473564358457e-07,
"logits/chosen": -1.8127809762954712,
"logits/rejected": -1.1028550863265991,
"logps/chosen": -264.7284240722656,
"logps/rejected": -247.4577178955078,
"loss": 0.1306,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.077170729637146,
"rewards/margins": 0.40461188554763794,
"rewards/rejected": -0.32744115591049194,
"step": 2900
},
{
"epoch": 0.78,
"learning_rate": 7.260731586586983e-07,
"logits/chosen": -1.756270408630371,
"logits/rejected": -1.4134676456451416,
"logps/chosen": -213.25234985351562,
"logps/rejected": -228.63436889648438,
"loss": 0.1606,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.0521668866276741,
"rewards/margins": 0.32329946756362915,
"rewards/rejected": -0.27113252878189087,
"step": 2910
},
{
"epoch": 0.78,
"learning_rate": 7.097526647366379e-07,
"logits/chosen": -1.8526099920272827,
"logits/rejected": -1.2808058261871338,
"logps/chosen": -266.97918701171875,
"logps/rejected": -249.385986328125,
"loss": 0.1041,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.08820641040802002,
"rewards/margins": 0.35536572337150574,
"rewards/rejected": -0.2671593129634857,
"step": 2920
},
{
"epoch": 0.78,
"learning_rate": 6.935872887769299e-07,
"logits/chosen": -1.659076452255249,
"logits/rejected": -1.0968009233474731,
"logps/chosen": -307.5328063964844,
"logps/rejected": -245.99307250976562,
"loss": 0.1618,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.07833143323659897,
"rewards/margins": 0.386299729347229,
"rewards/rejected": -0.3079683184623718,
"step": 2930
},
{
"epoch": 0.78,
"learning_rate": 6.775784314464717e-07,
"logits/chosen": -1.5469852685928345,
"logits/rejected": -1.411586046218872,
"logps/chosen": -225.2750701904297,
"logps/rejected": -242.4783935546875,
"loss": 0.1542,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": 0.04528594762086868,
"rewards/margins": 0.3220139145851135,
"rewards/rejected": -0.27672794461250305,
"step": 2940
},
{
"epoch": 0.79,
"learning_rate": 6.617274798504286e-07,
"logits/chosen": -1.6576216220855713,
"logits/rejected": -1.1561254262924194,
"logps/chosen": -238.072021484375,
"logps/rejected": -259.99688720703125,
"loss": 0.086,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.1014413982629776,
"rewards/margins": 0.43401598930358887,
"rewards/rejected": -0.33257460594177246,
"step": 2950
},
{
"epoch": 0.79,
"learning_rate": 6.460358074120518e-07,
"logits/chosen": -1.9168809652328491,
"logits/rejected": -1.2557293176651,
"logps/chosen": -242.58401489257812,
"logps/rejected": -269.3041687011719,
"loss": 0.152,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.09284786880016327,
"rewards/margins": 0.39133232831954956,
"rewards/rejected": -0.2984844744205475,
"step": 2960
},
{
"epoch": 0.79,
"learning_rate": 6.305047737536707e-07,
"logits/chosen": -1.9695804119110107,
"logits/rejected": -1.1151535511016846,
"logps/chosen": -290.43756103515625,
"logps/rejected": -263.0965881347656,
"loss": 0.1432,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.08852732926607132,
"rewards/margins": 0.39591920375823975,
"rewards/rejected": -0.307391881942749,
"step": 2970
},
{
"epoch": 0.79,
"learning_rate": 6.151357245788917e-07,
"logits/chosen": -1.6923097372055054,
"logits/rejected": -0.981489360332489,
"logps/chosen": -299.1329650878906,
"logps/rejected": -261.29949951171875,
"loss": 0.1661,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.09116804599761963,
"rewards/margins": 0.45771628618240356,
"rewards/rejected": -0.36654824018478394,
"step": 2980
},
{
"epoch": 0.8,
"learning_rate": 5.999299915559956e-07,
"logits/chosen": -1.7366340160369873,
"logits/rejected": -1.307422399520874,
"logps/chosen": -270.86468505859375,
"logps/rejected": -257.4667663574219,
"loss": 0.1394,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.14943814277648926,
"rewards/margins": 0.4425061345100403,
"rewards/rejected": -0.293067991733551,
"step": 2990
},
{
"epoch": 0.8,
"learning_rate": 5.848888922025553e-07,
"logits/chosen": -1.6904795169830322,
"logits/rejected": -1.514775037765503,
"logps/chosen": -245.5784454345703,
"logps/rejected": -294.32110595703125,
"loss": 0.1271,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.1196504607796669,
"rewards/margins": 0.3875979781150818,
"rewards/rejected": -0.2679474949836731,
"step": 3000
},
{
"epoch": 0.8,
"learning_rate": 5.700137297712749e-07,
"logits/chosen": -1.6396605968475342,
"logits/rejected": -1.020140290260315,
"logps/chosen": -293.1066589355469,
"logps/rejected": -233.6679229736328,
"loss": 0.1135,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.13376173377037048,
"rewards/margins": 0.3968765139579773,
"rewards/rejected": -0.26311472058296204,
"step": 3010
},
{
"epoch": 0.81,
"learning_rate": 5.553057931370729e-07,
"logits/chosen": -1.6284799575805664,
"logits/rejected": -1.0204269886016846,
"logps/chosen": -281.06866455078125,
"logps/rejected": -226.4606170654297,
"loss": 0.1341,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.11780212819576263,
"rewards/margins": 0.3741871416568756,
"rewards/rejected": -0.25638502836227417,
"step": 3020
},
{
"epoch": 0.81,
"learning_rate": 5.407663566854008e-07,
"logits/chosen": -1.796983003616333,
"logits/rejected": -1.2661386728286743,
"logps/chosen": -281.26593017578125,
"logps/rejected": -282.1707458496094,
"loss": 0.1166,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.0918949544429779,
"rewards/margins": 0.3467631936073303,
"rewards/rejected": -0.25486817955970764,
"step": 3030
},
{
"epoch": 0.81,
"learning_rate": 5.263966802018275e-07,
"logits/chosen": -1.7147595882415771,
"logits/rejected": -1.1799060106277466,
"logps/chosen": -264.71673583984375,
"logps/rejected": -275.89239501953125,
"loss": 0.1189,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.08099232614040375,
"rewards/margins": 0.3739253282546997,
"rewards/rejected": -0.29293301701545715,
"step": 3040
},
{
"epoch": 0.81,
"learning_rate": 5.121980087628802e-07,
"logits/chosen": -1.8341114521026611,
"logits/rejected": -1.1735069751739502,
"logps/chosen": -296.3938903808594,
"logps/rejected": -258.4015808105469,
"loss": 0.1205,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.06104854494333267,
"rewards/margins": 0.4131019711494446,
"rewards/rejected": -0.35205337405204773,
"step": 3050
},
{
"epoch": 0.82,
"learning_rate": 4.981715726281666e-07,
"logits/chosen": -1.6473156213760376,
"logits/rejected": -1.303798794746399,
"logps/chosen": -253.82205200195312,
"logps/rejected": -311.9048767089844,
"loss": 0.0928,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 0.08462455123662949,
"rewards/margins": 0.483224093914032,
"rewards/rejected": -0.3985995352268219,
"step": 3060
},
{
"epoch": 0.82,
"learning_rate": 4.843185871337722e-07,
"logits/chosen": -1.7317914962768555,
"logits/rejected": -1.2134945392608643,
"logps/chosen": -224.9326629638672,
"logps/rejected": -257.9466857910156,
"loss": 0.1239,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.10292349010705948,
"rewards/margins": 0.4141048789024353,
"rewards/rejected": -0.3111814260482788,
"step": 3070
},
{
"epoch": 0.82,
"learning_rate": 4.706402525869633e-07,
"logits/chosen": -1.887621521949768,
"logits/rejected": -1.0996310710906982,
"logps/chosen": -326.8731384277344,
"logps/rejected": -258.768798828125,
"loss": 0.1689,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.0996357649564743,
"rewards/margins": 0.36071866750717163,
"rewards/rejected": -0.2610829174518585,
"step": 3080
},
{
"epoch": 0.82,
"learning_rate": 4.5713775416217884e-07,
"logits/chosen": -1.3531527519226074,
"logits/rejected": -1.1717339754104614,
"logps/chosen": -224.5545654296875,
"logps/rejected": -308.23358154296875,
"loss": 0.109,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.10629155486822128,
"rewards/margins": 0.40465980768203735,
"rewards/rejected": -0.29836827516555786,
"step": 3090
},
{
"epoch": 0.83,
"learning_rate": 4.438122617983442e-07,
"logits/chosen": -1.9190078973770142,
"logits/rejected": -1.3009909391403198,
"logps/chosen": -285.29534912109375,
"logps/rejected": -245.940673828125,
"loss": 0.1582,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.12030161917209625,
"rewards/margins": 0.36642009019851685,
"rewards/rejected": -0.246118426322937,
"step": 3100
},
{
"epoch": 0.83,
"learning_rate": 4.3066493009749853e-07,
"logits/chosen": -1.6608394384384155,
"logits/rejected": -1.2748819589614868,
"logps/chosen": -249.3994140625,
"logps/rejected": -303.565673828125,
"loss": 0.0973,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.09208109974861145,
"rewards/margins": 0.3958059251308441,
"rewards/rejected": -0.3037247955799103,
"step": 3110
},
{
"epoch": 0.83,
"learning_rate": 4.1769689822475147e-07,
"logits/chosen": -1.7159957885742188,
"logits/rejected": -1.2019530534744263,
"logps/chosen": -243.9926300048828,
"logps/rejected": -216.5254364013672,
"loss": 0.1449,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.0733565241098404,
"rewards/margins": 0.32575708627700806,
"rewards/rejected": -0.25240057706832886,
"step": 3120
},
{
"epoch": 0.83,
"learning_rate": 4.049092898095816e-07,
"logits/chosen": -1.695738434791565,
"logits/rejected": -1.370802402496338,
"logps/chosen": -242.7845458984375,
"logps/rejected": -251.65859985351562,
"loss": 0.1549,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.05762239173054695,
"rewards/margins": 0.2988077998161316,
"rewards/rejected": -0.24118542671203613,
"step": 3130
},
{
"epoch": 0.84,
"learning_rate": 3.9230321284847856e-07,
"logits/chosen": -1.5597546100616455,
"logits/rejected": -1.1920053958892822,
"logps/chosen": -202.78463745117188,
"logps/rejected": -235.42611694335938,
"loss": 0.1828,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": 0.08725061267614365,
"rewards/margins": 0.3124812841415405,
"rewards/rejected": -0.22523066401481628,
"step": 3140
},
{
"epoch": 0.84,
"learning_rate": 3.798797596089351e-07,
"logits/chosen": -1.710565209388733,
"logits/rejected": -1.3830753564834595,
"logps/chosen": -191.14996337890625,
"logps/rejected": -228.13143920898438,
"loss": 0.1745,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": 0.07240664213895798,
"rewards/margins": 0.2955719232559204,
"rewards/rejected": -0.22316527366638184,
"step": 3150
},
{
"epoch": 0.84,
"learning_rate": 3.6764000653481263e-07,
"logits/chosen": -1.6754013299942017,
"logits/rejected": -1.4069788455963135,
"logps/chosen": -212.1529541015625,
"logps/rejected": -288.8616027832031,
"loss": 0.1174,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.01046234555542469,
"rewards/margins": 0.29962268471717834,
"rewards/rejected": -0.2891603410243988,
"step": 3160
},
{
"epoch": 0.85,
"learning_rate": 3.555850141530659e-07,
"logits/chosen": -1.9133336544036865,
"logits/rejected": -1.308585286140442,
"logps/chosen": -272.24310302734375,
"logps/rejected": -284.73089599609375,
"loss": 0.1114,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.06770848482847214,
"rewards/margins": 0.39665132761001587,
"rewards/rejected": -0.3289428651332855,
"step": 3170
},
{
"epoch": 0.85,
"learning_rate": 3.4371582698185636e-07,
"logits/chosen": -1.67145574092865,
"logits/rejected": -1.438444972038269,
"logps/chosen": -208.4847412109375,
"logps/rejected": -290.7266540527344,
"loss": 0.1622,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.05329760164022446,
"rewards/margins": 0.38010022044181824,
"rewards/rejected": -0.3268026113510132,
"step": 3180
},
{
"epoch": 0.85,
"learning_rate": 3.3203347344004737e-07,
"logits/chosen": -1.6708052158355713,
"logits/rejected": -1.0412744283676147,
"logps/chosen": -257.42547607421875,
"logps/rejected": -255.15597534179688,
"loss": 0.099,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.09596081078052521,
"rewards/margins": 0.3767240345478058,
"rewards/rejected": -0.2807632088661194,
"step": 3190
},
{
"epoch": 0.85,
"learning_rate": 3.2053896575809426e-07,
"logits/chosen": -1.5217217206954956,
"logits/rejected": -1.2141330242156982,
"logps/chosen": -183.3319091796875,
"logps/rejected": -227.89114379882812,
"loss": 0.1609,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.0313902273774147,
"rewards/margins": 0.31350037455558777,
"rewards/rejected": -0.28211015462875366,
"step": 3200
},
{
"epoch": 0.86,
"learning_rate": 3.092332998903416e-07,
"logits/chosen": -1.786128282546997,
"logits/rejected": -1.0563210248947144,
"logps/chosen": -275.27703857421875,
"logps/rejected": -250.85952758789062,
"loss": 0.1375,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.09343347698450089,
"rewards/margins": 0.4091528058052063,
"rewards/rejected": -0.3157193064689636,
"step": 3210
},
{
"epoch": 0.86,
"learning_rate": 2.981174554287239e-07,
"logits/chosen": -1.4629366397857666,
"logits/rejected": -1.2211610078811646,
"logps/chosen": -213.59805297851562,
"logps/rejected": -242.63827514648438,
"loss": 0.1274,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.09229601919651031,
"rewards/margins": 0.42800164222717285,
"rewards/rejected": -0.33570563793182373,
"step": 3220
},
{
"epoch": 0.86,
"learning_rate": 2.871923955178918e-07,
"logits/chosen": -1.6027485132217407,
"logits/rejected": -1.358605146408081,
"logps/chosen": -266.48590087890625,
"logps/rejected": -279.26641845703125,
"loss": 0.1804,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.019939150661230087,
"rewards/margins": 0.34129849076271057,
"rewards/rejected": -0.3213593065738678,
"step": 3230
},
{
"epoch": 0.86,
"learning_rate": 2.764590667717562e-07,
"logits/chosen": -1.699567437171936,
"logits/rejected": -1.2492908239364624,
"logps/chosen": -185.8858184814453,
"logps/rejected": -198.7812042236328,
"loss": 0.1461,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.0718788132071495,
"rewards/margins": 0.35736599564552307,
"rewards/rejected": -0.28548720479011536,
"step": 3240
},
{
"epoch": 0.87,
"learning_rate": 2.6591839919146963e-07,
"logits/chosen": -1.7908653020858765,
"logits/rejected": -1.3732439279556274,
"logps/chosen": -259.0335998535156,
"logps/rejected": -245.4866485595703,
"loss": 0.1154,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.03901948407292366,
"rewards/margins": 0.35004353523254395,
"rewards/rejected": -0.3110240697860718,
"step": 3250
},
{
"epoch": 0.87,
"learning_rate": 2.555713060848433e-07,
"logits/chosen": -1.8778269290924072,
"logits/rejected": -1.3474853038787842,
"logps/chosen": -365.189208984375,
"logps/rejected": -387.7873229980469,
"loss": 0.105,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.05286934971809387,
"rewards/margins": 0.4107195734977722,
"rewards/rejected": -0.35785022377967834,
"step": 3260
},
{
"epoch": 0.87,
"learning_rate": 2.454186839872158e-07,
"logits/chosen": -1.6174417734146118,
"logits/rejected": -1.245863914489746,
"logps/chosen": -199.4110107421875,
"logps/rejected": -256.9593200683594,
"loss": 0.0899,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.06935293972492218,
"rewards/margins": 0.3890882432460785,
"rewards/rejected": -0.3197353482246399,
"step": 3270
},
{
"epoch": 0.87,
"learning_rate": 2.3546141258376786e-07,
"logits/chosen": -1.9288465976715088,
"logits/rejected": -1.4281173944473267,
"logps/chosen": -290.6533508300781,
"logps/rejected": -269.7490539550781,
"loss": 0.0973,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.1082816869020462,
"rewards/margins": 0.44768819212913513,
"rewards/rejected": -0.3394065499305725,
"step": 3280
},
{
"epoch": 0.88,
"learning_rate": 2.257003546333042e-07,
"logits/chosen": -2.0008137226104736,
"logits/rejected": -1.320913314819336,
"logps/chosen": -320.672119140625,
"logps/rejected": -271.7630310058594,
"loss": 0.1208,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.07430247217416763,
"rewards/margins": 0.415428102016449,
"rewards/rejected": -0.3411256670951843,
"step": 3290
},
{
"epoch": 0.88,
"learning_rate": 2.1613635589349756e-07,
"logits/chosen": -1.7060868740081787,
"logits/rejected": -1.4465411901474,
"logps/chosen": -220.87380981445312,
"logps/rejected": -298.72625732421875,
"loss": 0.1216,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.09051541984081268,
"rewards/margins": 0.4572044014930725,
"rewards/rejected": -0.366688996553421,
"step": 3300
},
{
"epoch": 0.88,
"learning_rate": 2.0677024504760752e-07,
"logits/chosen": -1.638347864151001,
"logits/rejected": -1.0680015087127686,
"logps/chosen": -267.92608642578125,
"logps/rejected": -270.8176574707031,
"loss": 0.121,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.08300314098596573,
"rewards/margins": 0.42494750022888184,
"rewards/rejected": -0.3419443368911743,
"step": 3310
},
{
"epoch": 0.89,
"learning_rate": 1.9760283363267684e-07,
"logits/chosen": -1.5746757984161377,
"logits/rejected": -1.1135761737823486,
"logps/chosen": -221.92343139648438,
"logps/rejected": -250.6765899658203,
"loss": 0.1311,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.07789792120456696,
"rewards/margins": 0.39919474720954895,
"rewards/rejected": -0.3212968707084656,
"step": 3320
},
{
"epoch": 0.89,
"learning_rate": 1.8863491596921745e-07,
"logits/chosen": -1.5667272806167603,
"logits/rejected": -0.9896243810653687,
"logps/chosen": -296.76202392578125,
"logps/rejected": -303.38665771484375,
"loss": 0.1393,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.0856994241476059,
"rewards/margins": 0.44408607482910156,
"rewards/rejected": -0.35838669538497925,
"step": 3330
},
{
"epoch": 0.89,
"learning_rate": 1.798672690923828e-07,
"logits/chosen": -1.6267836093902588,
"logits/rejected": -1.1321130990982056,
"logps/chosen": -253.453369140625,
"logps/rejected": -270.5718078613281,
"loss": 0.1355,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.03761683404445648,
"rewards/margins": 0.36388128995895386,
"rewards/rejected": -0.3262644410133362,
"step": 3340
},
{
"epoch": 0.89,
"learning_rate": 1.713006526846439e-07,
"logits/chosen": -1.678847312927246,
"logits/rejected": -1.140943169593811,
"logps/chosen": -262.73638916015625,
"logps/rejected": -221.0591583251953,
"loss": 0.12,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.05530963093042374,
"rewards/margins": 0.353360652923584,
"rewards/rejected": -0.29805102944374084,
"step": 3350
},
{
"epoch": 0.9,
"learning_rate": 1.629358090099639e-07,
"logits/chosen": -1.7336199283599854,
"logits/rejected": -1.1567232608795166,
"logps/chosen": -307.455078125,
"logps/rejected": -310.3108825683594,
"loss": 0.1219,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.049434565007686615,
"rewards/margins": 0.42315396666526794,
"rewards/rejected": -0.37371936440467834,
"step": 3360
},
{
"epoch": 0.9,
"learning_rate": 1.5477346284948292e-07,
"logits/chosen": -1.746050238609314,
"logits/rejected": -1.4284813404083252,
"logps/chosen": -237.4105682373047,
"logps/rejected": -288.8939514160156,
"loss": 0.1522,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.028758401051163673,
"rewards/margins": 0.33460500836372375,
"rewards/rejected": -0.30584657192230225,
"step": 3370
},
{
"epoch": 0.9,
"learning_rate": 1.4681432143872133e-07,
"logits/chosen": -1.6106250286102295,
"logits/rejected": -1.148560881614685,
"logps/chosen": -258.1077575683594,
"logps/rejected": -318.8696594238281,
"loss": 0.1528,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.04034858196973801,
"rewards/margins": 0.350983202457428,
"rewards/rejected": -0.3106346130371094,
"step": 3380
},
{
"epoch": 0.9,
"learning_rate": 1.3905907440629752e-07,
"logits/chosen": -1.8278043270111084,
"logits/rejected": -1.0977909564971924,
"logps/chosen": -309.5404357910156,
"logps/rejected": -250.6240692138672,
"loss": 0.1186,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.07560743391513824,
"rewards/margins": 0.4015159606933594,
"rewards/rejected": -0.3259085416793823,
"step": 3390
},
{
"epoch": 0.91,
"learning_rate": 1.31508393714177e-07,
"logits/chosen": -1.6950404644012451,
"logits/rejected": -1.275434136390686,
"logps/chosen": -280.67218017578125,
"logps/rejected": -243.7939910888672,
"loss": 0.1406,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.06180018186569214,
"rewards/margins": 0.3164129853248596,
"rewards/rejected": -0.25461283326148987,
"step": 3400
},
{
"epoch": 0.91,
"learning_rate": 1.241629335994471e-07,
"logits/chosen": -1.6643226146697998,
"logits/rejected": -0.9809878468513489,
"logps/chosen": -246.31515502929688,
"logps/rejected": -255.1239013671875,
"loss": 0.1209,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.06152065843343735,
"rewards/margins": 0.4374101161956787,
"rewards/rejected": -0.37588945031166077,
"step": 3410
},
{
"epoch": 0.91,
"learning_rate": 1.1702333051763271e-07,
"logits/chosen": -1.6431770324707031,
"logits/rejected": -1.1887956857681274,
"logps/chosen": -247.9853515625,
"logps/rejected": -250.0074920654297,
"loss": 0.1573,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.052826426923274994,
"rewards/margins": 0.3623102903366089,
"rewards/rejected": -0.3094838261604309,
"step": 3420
},
{
"epoch": 0.91,
"learning_rate": 1.1009020308754587e-07,
"logits/chosen": -1.7229608297348022,
"logits/rejected": -1.225477933883667,
"logps/chosen": -235.5467529296875,
"logps/rejected": -204.58261108398438,
"loss": 0.1929,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": 0.04221180081367493,
"rewards/margins": 0.2941130995750427,
"rewards/rejected": -0.2519012987613678,
"step": 3430
},
{
"epoch": 0.92,
"learning_rate": 1.0336415203768962e-07,
"logits/chosen": -1.4921534061431885,
"logits/rejected": -1.1619257926940918,
"logps/chosen": -285.34259033203125,
"logps/rejected": -295.04864501953125,
"loss": 0.1181,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.08926016837358475,
"rewards/margins": 0.4191388189792633,
"rewards/rejected": -0.32987862825393677,
"step": 3440
},
{
"epoch": 0.92,
"learning_rate": 9.684576015420277e-08,
"logits/chosen": -1.4940868616104126,
"logits/rejected": -0.863502025604248,
"logps/chosen": -264.92169189453125,
"logps/rejected": -247.3212890625,
"loss": 0.1366,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.027351949363946915,
"rewards/margins": 0.341216504573822,
"rewards/rejected": -0.3138645589351654,
"step": 3450
},
{
"epoch": 0.92,
"learning_rate": 9.053559223036746e-08,
"logits/chosen": -1.7528457641601562,
"logits/rejected": -1.0926567316055298,
"logps/chosen": -313.1197509765625,
"logps/rejected": -255.8910675048828,
"loss": 0.1122,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.09485974907875061,
"rewards/margins": 0.4161204397678375,
"rewards/rejected": -0.3212606906890869,
"step": 3460
},
{
"epoch": 0.93,
"learning_rate": 8.44341950176683e-08,
"logits/chosen": -1.5928373336791992,
"logits/rejected": -1.244089126586914,
"logps/chosen": -265.8317565917969,
"logps/rejected": -325.6632080078125,
"loss": 0.1168,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.02779344841837883,
"rewards/margins": 0.3878537714481354,
"rewards/rejected": -0.36006033420562744,
"step": 3470
},
{
"epoch": 0.93,
"learning_rate": 7.854209717842231e-08,
"logits/chosen": -1.4396581649780273,
"logits/rejected": -1.2138065099716187,
"logps/chosen": -240.5614776611328,
"logps/rejected": -349.2239685058594,
"loss": 0.1016,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": 0.06546439230442047,
"rewards/margins": 0.43350672721862793,
"rewards/rejected": -0.36804237961769104,
"step": 3480
},
{
"epoch": 0.93,
"learning_rate": 7.285980923996989e-08,
"logits/chosen": -1.752174735069275,
"logits/rejected": -1.3645591735839844,
"logps/chosen": -234.2113800048828,
"logps/rejected": -227.34848022460938,
"loss": 0.1145,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.06558214873075485,
"rewards/margins": 0.37207797169685364,
"rewards/rejected": -0.30649590492248535,
"step": 3490
},
{
"epoch": 0.93,
"learning_rate": 6.738782355044048e-08,
"logits/chosen": -1.3806583881378174,
"logits/rejected": -1.1515109539031982,
"logps/chosen": -230.22073364257812,
"logps/rejected": -265.16314697265625,
"loss": 0.1637,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.03167334571480751,
"rewards/margins": 0.4240112900733948,
"rewards/rejected": -0.39233797788619995,
"step": 3500
},
{
"epoch": 0.94,
"learning_rate": 6.212661423609184e-08,
"logits/chosen": -1.608327865600586,
"logits/rejected": -0.9549871683120728,
"logps/chosen": -225.1513214111328,
"logps/rejected": -209.7755584716797,
"loss": 0.134,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.05912930518388748,
"rewards/margins": 0.37384232878685,
"rewards/rejected": -0.3147130608558655,
"step": 3510
},
{
"epoch": 0.94,
"learning_rate": 5.707663716023021e-08,
"logits/chosen": -1.853945016860962,
"logits/rejected": -1.118769884109497,
"logps/chosen": -280.1803283691406,
"logps/rejected": -295.4366760253906,
"loss": 0.1538,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.05890347808599472,
"rewards/margins": 0.4038105010986328,
"rewards/rejected": -0.3449070453643799,
"step": 3520
},
{
"epoch": 0.94,
"learning_rate": 5.22383298837098e-08,
"logits/chosen": -1.6122610569000244,
"logits/rejected": -1.0327246189117432,
"logps/chosen": -242.87600708007812,
"logps/rejected": -208.82144165039062,
"loss": 0.0935,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.06353352963924408,
"rewards/margins": 0.38507527112960815,
"rewards/rejected": -0.3215416669845581,
"step": 3530
},
{
"epoch": 0.94,
"learning_rate": 4.761211162702117e-08,
"logits/chosen": -1.680824875831604,
"logits/rejected": -1.130185604095459,
"logps/chosen": -244.988037109375,
"logps/rejected": -263.0980529785156,
"loss": 0.1587,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.06494718790054321,
"rewards/margins": 0.3642052412033081,
"rewards/rejected": -0.2992580533027649,
"step": 3540
},
{
"epoch": 0.95,
"learning_rate": 4.319838323396691e-08,
"logits/chosen": -1.6353555917739868,
"logits/rejected": -1.2781399488449097,
"logps/chosen": -228.3214569091797,
"logps/rejected": -280.24066162109375,
"loss": 0.1688,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.013865587301552296,
"rewards/margins": 0.3643207550048828,
"rewards/rejected": -0.3781863749027252,
"step": 3550
},
{
"epoch": 0.95,
"learning_rate": 3.8997527136930004e-08,
"logits/chosen": -1.7263469696044922,
"logits/rejected": -1.1389892101287842,
"logps/chosen": -293.1500244140625,
"logps/rejected": -265.1076354980469,
"loss": 0.0911,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.08173110336065292,
"rewards/margins": 0.42599812150001526,
"rewards/rejected": -0.3442670404911041,
"step": 3560
},
{
"epoch": 0.95,
"learning_rate": 3.5009907323737826e-08,
"logits/chosen": -1.561417579650879,
"logits/rejected": -1.165183424949646,
"logps/chosen": -252.6567840576172,
"logps/rejected": -276.4109802246094,
"loss": 0.0987,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.07267307490110397,
"rewards/margins": 0.37683752179145813,
"rewards/rejected": -0.30416446924209595,
"step": 3570
},
{
"epoch": 0.95,
"learning_rate": 3.1235869306123766e-08,
"logits/chosen": -1.6449648141860962,
"logits/rejected": -1.2687057256698608,
"logps/chosen": -221.9775848388672,
"logps/rejected": -275.42449951171875,
"loss": 0.1551,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.01698826439678669,
"rewards/margins": 0.35954660177230835,
"rewards/rejected": -0.3425583243370056,
"step": 3580
},
{
"epoch": 0.96,
"learning_rate": 2.767574008979007e-08,
"logits/chosen": -1.640044927597046,
"logits/rejected": -1.0068800449371338,
"logps/chosen": -286.2217712402344,
"logps/rejected": -278.906494140625,
"loss": 0.0995,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": 0.073039710521698,
"rewards/margins": 0.4568893015384674,
"rewards/rejected": -0.3838495910167694,
"step": 3590
},
{
"epoch": 0.96,
"learning_rate": 2.4329828146074096e-08,
"logits/chosen": -1.9251073598861694,
"logits/rejected": -1.0933765172958374,
"logps/chosen": -352.7594299316406,
"logps/rejected": -274.07269287109375,
"loss": 0.1353,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.07194405049085617,
"rewards/margins": 0.3892292380332947,
"rewards/rejected": -0.3172852396965027,
"step": 3600
},
{
"epoch": 0.96,
"learning_rate": 2.1198423385220822e-08,
"logits/chosen": -1.6453202962875366,
"logits/rejected": -1.2082674503326416,
"logps/chosen": -257.38250732421875,
"logps/rejected": -320.1260986328125,
"loss": 0.0771,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 0.10315804183483124,
"rewards/margins": 0.4763513505458832,
"rewards/rejected": -0.37319326400756836,
"step": 3610
},
{
"epoch": 0.97,
"learning_rate": 1.82817971312621e-08,
"logits/chosen": -1.790305733680725,
"logits/rejected": -1.1114693880081177,
"logps/chosen": -260.489990234375,
"logps/rejected": -238.48779296875,
"loss": 0.1119,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.09379683434963226,
"rewards/margins": 0.4416617453098297,
"rewards/rejected": -0.34786492586135864,
"step": 3620
},
{
"epoch": 0.97,
"learning_rate": 1.5580202098509078e-08,
"logits/chosen": -1.6579450368881226,
"logits/rejected": -1.3401272296905518,
"logps/chosen": -250.71633911132812,
"logps/rejected": -241.178466796875,
"loss": 0.211,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": 0.027279715985059738,
"rewards/margins": 0.31590917706489563,
"rewards/rejected": -0.2886294722557068,
"step": 3630
},
{
"epoch": 0.97,
"learning_rate": 1.3093872369654148e-08,
"logits/chosen": -1.72607421875,
"logits/rejected": -1.3506278991699219,
"logps/chosen": -237.07730102539062,
"logps/rejected": -328.91607666015625,
"loss": 0.1135,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": 0.0632893294095993,
"rewards/margins": 0.4014991819858551,
"rewards/rejected": -0.3382098376750946,
"step": 3640
},
{
"epoch": 0.97,
"learning_rate": 1.0823023375489128e-08,
"logits/chosen": -1.8103011846542358,
"logits/rejected": -1.331343412399292,
"logps/chosen": -218.52322387695312,
"logps/rejected": -215.2735595703125,
"loss": 0.1715,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.05841076374053955,
"rewards/margins": 0.369015634059906,
"rewards/rejected": -0.31060490012168884,
"step": 3650
},
{
"epoch": 0.98,
"learning_rate": 8.767851876239075e-09,
"logits/chosen": -1.3743187189102173,
"logits/rejected": -1.1308249235153198,
"logps/chosen": -215.2615966796875,
"logps/rejected": -300.07080078125,
"loss": 0.0964,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.0790005549788475,
"rewards/margins": 0.41065654158592224,
"rewards/rejected": -0.33165597915649414,
"step": 3660
},
{
"epoch": 0.98,
"learning_rate": 6.9285359445145366e-09,
"logits/chosen": -1.418792486190796,
"logits/rejected": -1.1899298429489136,
"logps/chosen": -232.254638671875,
"logps/rejected": -266.04376220703125,
"loss": 0.153,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.1082618460059166,
"rewards/margins": 0.42941126227378845,
"rewards/rejected": -0.3211493492126465,
"step": 3670
},
{
"epoch": 0.98,
"learning_rate": 5.305234949880001e-09,
"logits/chosen": -1.5959584712982178,
"logits/rejected": -1.2097722291946411,
"logps/chosen": -276.74713134765625,
"logps/rejected": -315.0145568847656,
"loss": 0.1814,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.07109376043081284,
"rewards/margins": 0.38560524582862854,
"rewards/rejected": -0.3145114779472351,
"step": 3680
},
{
"epoch": 0.98,
"learning_rate": 3.8980895450474455e-09,
"logits/chosen": -1.7472782135009766,
"logits/rejected": -1.2426128387451172,
"logps/chosen": -275.06793212890625,
"logps/rejected": -278.3852233886719,
"loss": 0.1241,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.09598256647586823,
"rewards/margins": 0.42074212431907654,
"rewards/rejected": -0.3247596323490143,
"step": 3690
},
{
"epoch": 0.99,
"learning_rate": 2.7072216536885855e-09,
"logits/chosen": -1.481673002243042,
"logits/rejected": -0.9478602409362793,
"logps/chosen": -281.0965576171875,
"logps/rejected": -267.8004150390625,
"loss": 0.1391,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.02585453912615776,
"rewards/margins": 0.3552091717720032,
"rewards/rejected": -0.32935458421707153,
"step": 3700
},
{
"epoch": 0.99,
"learning_rate": 1.7327344598702667e-09,
"logits/chosen": -1.533521056175232,
"logits/rejected": -1.2224196195602417,
"logps/chosen": -232.5213623046875,
"logps/rejected": -339.6204528808594,
"loss": 0.1444,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.004062544088810682,
"rewards/margins": 0.3145105242729187,
"rewards/rejected": -0.3185730576515198,
"step": 3710
},
{
"epoch": 0.99,
"learning_rate": 9.747123991141193e-10,
"logits/chosen": -1.80594801902771,
"logits/rejected": -1.2266263961791992,
"logps/chosen": -253.4310760498047,
"logps/rejected": -281.5027160644531,
"loss": 0.1023,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.06998688727617264,
"rewards/margins": 0.3998931050300598,
"rewards/rejected": -0.3299062252044678,
"step": 3720
},
{
"epoch": 0.99,
"learning_rate": 4.332211510807427e-10,
"logits/chosen": -1.7258926630020142,
"logits/rejected": -1.4480842351913452,
"logps/chosen": -233.64529418945312,
"logps/rejected": -240.29598999023438,
"loss": 0.1289,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.031451232731342316,
"rewards/margins": 0.3203813433647156,
"rewards/rejected": -0.28893011808395386,
"step": 3730
},
{
"epoch": 1.0,
"learning_rate": 1.0830763387897902e-10,
"logits/chosen": -1.8174076080322266,
"logits/rejected": -1.2164661884307861,
"logps/chosen": -286.5456237792969,
"logps/rejected": -223.0185546875,
"loss": 0.1329,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.08067300915718079,
"rewards/margins": 0.36825209856033325,
"rewards/rejected": -0.2875790596008301,
"step": 3740
},
{
"epoch": 1.0,
"learning_rate": 0.0,
"logits/chosen": -1.574351191520691,
"logits/rejected": -1.151698112487793,
"logps/chosen": -267.740234375,
"logps/rejected": -287.2723693847656,
"loss": 0.1084,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.0795169472694397,
"rewards/margins": 0.420736163854599,
"rewards/rejected": -0.3412191867828369,
"step": 3750
},
{
"epoch": 1.0,
"step": 3750,
"total_flos": 0.0,
"train_loss": 0.15019961676597596,
"train_runtime": 15622.399,
"train_samples_per_second": 0.96,
"train_steps_per_second": 0.24
}
],
"logging_steps": 10,
"max_steps": 3750,
"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
}