zephyr-7b-gpo-v2-i3 / trainer_state.json
lole25's picture
Model save
4f3d3bd 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": -1.4607182741165161,
"logits/rejected": -1.0577633380889893,
"logps/chosen": -377.1839599609375,
"logps/rejected": -1292.140625,
"loss": 0.3828,
"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": -1.4249842166900635,
"logits/rejected": -1.0064045190811157,
"logps/chosen": -603.1861572265625,
"logps/rejected": -1303.3701171875,
"loss": 0.3718,
"rewards/accuracies": 0.4444444477558136,
"rewards/chosen": 2.63900601567002e-05,
"rewards/margins": 0.00017314580327365547,
"rewards/rejected": -0.00014675571583211422,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 2.666666666666667e-07,
"logits/chosen": -1.4733855724334717,
"logits/rejected": -1.0931047201156616,
"logps/chosen": -545.4578857421875,
"logps/rejected": -1043.0272216796875,
"loss": 0.4125,
"rewards/accuracies": 0.375,
"rewards/chosen": -0.0002783415839076042,
"rewards/margins": -0.000188146106665954,
"rewards/rejected": -9.019548451760784e-05,
"step": 20
},
{
"epoch": 0.01,
"learning_rate": 4.0000000000000003e-07,
"logits/chosen": -1.6173245906829834,
"logits/rejected": -1.0555126667022705,
"logps/chosen": -654.3507080078125,
"logps/rejected": -1315.276611328125,
"loss": 0.3832,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.0005331189604476094,
"rewards/margins": 0.001845326623879373,
"rewards/rejected": -0.0023784455843269825,
"step": 30
},
{
"epoch": 0.01,
"learning_rate": 5.333333333333335e-07,
"logits/chosen": -1.3433172702789307,
"logits/rejected": -0.6946144700050354,
"logps/chosen": -665.9903564453125,
"logps/rejected": -1348.9879150390625,
"loss": 0.3634,
"rewards/accuracies": 0.625,
"rewards/chosen": -0.0011921643745154142,
"rewards/margins": 0.0034048897214233875,
"rewards/rejected": -0.004597053863108158,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 6.666666666666667e-07,
"logits/chosen": -1.370167851448059,
"logits/rejected": -0.7938201427459717,
"logps/chosen": -664.8009033203125,
"logps/rejected": -1172.046630859375,
"loss": 0.3289,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.0008570374920964241,
"rewards/margins": 0.006419859826564789,
"rewards/rejected": -0.007276897318661213,
"step": 50
},
{
"epoch": 0.02,
"learning_rate": 8.000000000000001e-07,
"logits/chosen": -1.6229826211929321,
"logits/rejected": -0.9315292239189148,
"logps/chosen": -660.7051391601562,
"logps/rejected": -1401.3045654296875,
"loss": 0.3344,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.00019816956773865968,
"rewards/margins": 0.020777523517608643,
"rewards/rejected": -0.020975695922970772,
"step": 60
},
{
"epoch": 0.02,
"learning_rate": 9.333333333333334e-07,
"logits/chosen": -1.5964797735214233,
"logits/rejected": -0.636447548866272,
"logps/chosen": -600.3419189453125,
"logps/rejected": -1338.2294921875,
"loss": 0.3384,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": 0.0026218045968562365,
"rewards/margins": 0.03453027456998825,
"rewards/rejected": -0.03190847486257553,
"step": 70
},
{
"epoch": 0.02,
"learning_rate": 1.066666666666667e-06,
"logits/chosen": -1.6199884414672852,
"logits/rejected": -0.8888322114944458,
"logps/chosen": -589.5464477539062,
"logps/rejected": -1346.308349609375,
"loss": 0.3065,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.005537457764148712,
"rewards/margins": 0.04564369469881058,
"rewards/rejected": -0.05118114873766899,
"step": 80
},
{
"epoch": 0.02,
"learning_rate": 1.2000000000000002e-06,
"logits/chosen": -1.1931430101394653,
"logits/rejected": -0.7813756465911865,
"logps/chosen": -617.9501342773438,
"logps/rejected": -1214.172119140625,
"loss": 0.3072,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.029573017731308937,
"rewards/margins": 0.08381669223308563,
"rewards/rejected": -0.11338971555233002,
"step": 90
},
{
"epoch": 0.03,
"learning_rate": 1.3333333333333334e-06,
"logits/chosen": -1.1626551151275635,
"logits/rejected": -0.36785703897476196,
"logps/chosen": -628.2637329101562,
"logps/rejected": -1500.320556640625,
"loss": 0.2346,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.11836127936840057,
"rewards/margins": 0.16349823772907257,
"rewards/rejected": -0.28185951709747314,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 1.4666666666666669e-06,
"logits/chosen": -1.1234421730041504,
"logits/rejected": -0.16796138882637024,
"logps/chosen": -802.6090087890625,
"logps/rejected": -1697.603759765625,
"loss": 0.1828,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2379436492919922,
"rewards/margins": 0.24783344566822052,
"rewards/rejected": -0.48577699065208435,
"step": 110
},
{
"epoch": 0.03,
"learning_rate": 1.6000000000000001e-06,
"logits/chosen": -1.1458255052566528,
"logits/rejected": -0.4321994185447693,
"logps/chosen": -921.9117431640625,
"logps/rejected": -1853.3466796875,
"loss": 0.2603,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.3034071922302246,
"rewards/margins": 0.25969117879867554,
"rewards/rejected": -0.5630983710289001,
"step": 120
},
{
"epoch": 0.03,
"learning_rate": 1.7333333333333336e-06,
"logits/chosen": -1.2709288597106934,
"logits/rejected": -0.7817854881286621,
"logps/chosen": -811.3619384765625,
"logps/rejected": -1703.904541015625,
"loss": 0.2519,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.21060828864574432,
"rewards/margins": 0.20279932022094727,
"rewards/rejected": -0.413407564163208,
"step": 130
},
{
"epoch": 0.04,
"learning_rate": 1.8666666666666669e-06,
"logits/chosen": -1.4565277099609375,
"logits/rejected": -0.5677322149276733,
"logps/chosen": -727.5758666992188,
"logps/rejected": -1671.2864990234375,
"loss": 0.1793,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2207428514957428,
"rewards/margins": 0.21470816433429718,
"rewards/rejected": -0.43545103073120117,
"step": 140
},
{
"epoch": 0.04,
"learning_rate": 2.0000000000000003e-06,
"logits/chosen": -1.2830668687820435,
"logits/rejected": -0.31070059537887573,
"logps/chosen": -957.83544921875,
"logps/rejected": -1972.1334228515625,
"loss": 0.2268,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.3733420968055725,
"rewards/margins": 0.27919524908065796,
"rewards/rejected": -0.6525374054908752,
"step": 150
},
{
"epoch": 0.04,
"learning_rate": 2.133333333333334e-06,
"logits/chosen": -1.1059176921844482,
"logits/rejected": -0.29462116956710815,
"logps/chosen": -703.9768676757812,
"logps/rejected": -1613.99169921875,
"loss": 0.3007,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.25690004229545593,
"rewards/margins": 0.2521464228630066,
"rewards/rejected": -0.5090464353561401,
"step": 160
},
{
"epoch": 0.05,
"learning_rate": 2.266666666666667e-06,
"logits/chosen": -1.1736003160476685,
"logits/rejected": -0.5167692303657532,
"logps/chosen": -673.1292724609375,
"logps/rejected": -1497.2391357421875,
"loss": 0.2481,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.15533636510372162,
"rewards/margins": 0.13328927755355835,
"rewards/rejected": -0.2886256277561188,
"step": 170
},
{
"epoch": 0.05,
"learning_rate": 2.4000000000000003e-06,
"logits/chosen": -1.507161021232605,
"logits/rejected": -0.1557072103023529,
"logps/chosen": -770.8878784179688,
"logps/rejected": -1511.53857421875,
"loss": 0.1887,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.17462053894996643,
"rewards/margins": 0.15410876274108887,
"rewards/rejected": -0.3287292718887329,
"step": 180
},
{
"epoch": 0.05,
"learning_rate": 2.5333333333333338e-06,
"logits/chosen": -1.4759384393692017,
"logits/rejected": -0.15519209206104279,
"logps/chosen": -887.9100341796875,
"logps/rejected": -1733.956298828125,
"loss": 0.2453,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.2314099818468094,
"rewards/margins": 0.17767982184886932,
"rewards/rejected": -0.4090898036956787,
"step": 190
},
{
"epoch": 0.05,
"learning_rate": 2.666666666666667e-06,
"logits/chosen": -1.5579626560211182,
"logits/rejected": -0.15316779911518097,
"logps/chosen": -958.0621337890625,
"logps/rejected": -1928.722412109375,
"loss": 0.1696,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2732312083244324,
"rewards/margins": 0.27958157658576965,
"rewards/rejected": -0.5528126955032349,
"step": 200
},
{
"epoch": 0.06,
"learning_rate": 2.8000000000000003e-06,
"logits/chosen": -1.1436049938201904,
"logits/rejected": 0.03915581852197647,
"logps/chosen": -968.3499145507812,
"logps/rejected": -1884.228759765625,
"loss": 0.2664,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.2975374460220337,
"rewards/margins": 0.2711247503757477,
"rewards/rejected": -0.568662166595459,
"step": 210
},
{
"epoch": 0.06,
"learning_rate": 2.9333333333333338e-06,
"logits/chosen": -1.4685136079788208,
"logits/rejected": -0.4126014709472656,
"logps/chosen": -667.5554809570312,
"logps/rejected": -1680.7935791015625,
"loss": 0.1758,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.16100670397281647,
"rewards/margins": 0.24677987396717072,
"rewards/rejected": -0.4077865481376648,
"step": 220
},
{
"epoch": 0.06,
"learning_rate": 3.066666666666667e-06,
"logits/chosen": -1.1665582656860352,
"logits/rejected": -0.47384729981422424,
"logps/chosen": -922.4220581054688,
"logps/rejected": -1947.9945068359375,
"loss": 0.1925,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.2523236870765686,
"rewards/margins": 0.24001212418079376,
"rewards/rejected": -0.49233585596084595,
"step": 230
},
{
"epoch": 0.06,
"learning_rate": 3.2000000000000003e-06,
"logits/chosen": -1.6133432388305664,
"logits/rejected": -0.5420857071876526,
"logps/chosen": -857.8092041015625,
"logps/rejected": -1654.6234130859375,
"loss": 0.1908,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.2039283812046051,
"rewards/margins": 0.23253926634788513,
"rewards/rejected": -0.43646764755249023,
"step": 240
},
{
"epoch": 0.07,
"learning_rate": 3.3333333333333333e-06,
"logits/chosen": -1.2248852252960205,
"logits/rejected": 0.04564919322729111,
"logps/chosen": -836.74169921875,
"logps/rejected": -1574.744140625,
"loss": 0.2558,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.167385533452034,
"rewards/margins": 0.18353143334388733,
"rewards/rejected": -0.35091695189476013,
"step": 250
},
{
"epoch": 0.07,
"learning_rate": 3.4666666666666672e-06,
"logits/chosen": -1.4086463451385498,
"logits/rejected": 0.30391791462898254,
"logps/chosen": -697.5711669921875,
"logps/rejected": -1787.910888671875,
"loss": 0.1502,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.12433981895446777,
"rewards/margins": 0.2777388095855713,
"rewards/rejected": -0.40207862854003906,
"step": 260
},
{
"epoch": 0.07,
"learning_rate": 3.6000000000000003e-06,
"logits/chosen": -1.4037072658538818,
"logits/rejected": 0.20726795494556427,
"logps/chosen": -785.6559448242188,
"logps/rejected": -1549.1513671875,
"loss": 0.2138,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.1881810426712036,
"rewards/margins": 0.21742673218250275,
"rewards/rejected": -0.40560778975486755,
"step": 270
},
{
"epoch": 0.07,
"learning_rate": 3.7333333333333337e-06,
"logits/chosen": -1.345931887626648,
"logits/rejected": -0.15625306963920593,
"logps/chosen": -680.088623046875,
"logps/rejected": -1532.015625,
"loss": 0.2158,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.1171560287475586,
"rewards/margins": 0.1917530596256256,
"rewards/rejected": -0.3089090883731842,
"step": 280
},
{
"epoch": 0.08,
"learning_rate": 3.866666666666667e-06,
"logits/chosen": -1.5668309926986694,
"logits/rejected": -0.32349592447280884,
"logps/chosen": -740.4640502929688,
"logps/rejected": -1647.676513671875,
"loss": 0.1976,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.0793481171131134,
"rewards/margins": 0.25623226165771484,
"rewards/rejected": -0.335580438375473,
"step": 290
},
{
"epoch": 0.08,
"learning_rate": 4.000000000000001e-06,
"logits/chosen": -1.4235851764678955,
"logits/rejected": -0.6677430868148804,
"logps/chosen": -766.1461791992188,
"logps/rejected": -1581.3609619140625,
"loss": 0.1918,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.17254561185836792,
"rewards/margins": 0.22318947315216064,
"rewards/rejected": -0.39573508501052856,
"step": 300
},
{
"epoch": 0.08,
"learning_rate": 4.133333333333333e-06,
"logits/chosen": -1.3767132759094238,
"logits/rejected": -0.3363776206970215,
"logps/chosen": -967.8243408203125,
"logps/rejected": -1914.7396240234375,
"loss": 0.1578,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2716659605503082,
"rewards/margins": 0.3169645071029663,
"rewards/rejected": -0.5886305570602417,
"step": 310
},
{
"epoch": 0.09,
"learning_rate": 4.266666666666668e-06,
"logits/chosen": -1.7191673517227173,
"logits/rejected": 0.0708886906504631,
"logps/chosen": -1019.0963134765625,
"logps/rejected": -1888.802001953125,
"loss": 0.185,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.31889739632606506,
"rewards/margins": 0.31718775629997253,
"rewards/rejected": -0.6360851526260376,
"step": 320
},
{
"epoch": 0.09,
"learning_rate": 4.4e-06,
"logits/chosen": -1.2595525979995728,
"logits/rejected": -0.7953510880470276,
"logps/chosen": -697.2740478515625,
"logps/rejected": -1584.4857177734375,
"loss": 0.1932,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.15896865725517273,
"rewards/margins": 0.23230549693107605,
"rewards/rejected": -0.3912741243839264,
"step": 330
},
{
"epoch": 0.09,
"learning_rate": 4.533333333333334e-06,
"logits/chosen": -1.3439395427703857,
"logits/rejected": -0.4132903516292572,
"logps/chosen": -766.2467651367188,
"logps/rejected": -1648.1083984375,
"loss": 1.9196,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.23903754353523254,
"rewards/margins": 0.41279107332229614,
"rewards/rejected": -0.6518285870552063,
"step": 340
},
{
"epoch": 0.09,
"learning_rate": 4.666666666666667e-06,
"logits/chosen": -1.3772680759429932,
"logits/rejected": -0.3717087209224701,
"logps/chosen": -645.0536499023438,
"logps/rejected": -1581.9832763671875,
"loss": 0.2452,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.15307392179965973,
"rewards/margins": 0.14587204158306122,
"rewards/rejected": -0.29894596338272095,
"step": 350
},
{
"epoch": 0.1,
"learning_rate": 4.800000000000001e-06,
"logits/chosen": -1.4199841022491455,
"logits/rejected": -0.7620482444763184,
"logps/chosen": -858.2607421875,
"logps/rejected": -1545.4052734375,
"loss": 0.3538,
"rewards/accuracies": 0.625,
"rewards/chosen": -0.14831410348415375,
"rewards/margins": 0.06550173461437225,
"rewards/rejected": -0.213815838098526,
"step": 360
},
{
"epoch": 0.1,
"learning_rate": 4.933333333333334e-06,
"logits/chosen": -1.6882193088531494,
"logits/rejected": -0.8316219449043274,
"logps/chosen": -846.7376098632812,
"logps/rejected": -1478.4105224609375,
"loss": 0.3138,
"rewards/accuracies": 0.574999988079071,
"rewards/chosen": -0.12148912250995636,
"rewards/margins": 0.06598127633333206,
"rewards/rejected": -0.18747039139270782,
"step": 370
},
{
"epoch": 0.1,
"learning_rate": 4.999972922944898e-06,
"logits/chosen": -1.6390502452850342,
"logits/rejected": -1.078680396080017,
"logps/chosen": -709.0716552734375,
"logps/rejected": -1625.7291259765625,
"loss": 0.1946,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06936102360486984,
"rewards/margins": 0.168907031416893,
"rewards/rejected": -0.23826804757118225,
"step": 380
},
{
"epoch": 0.1,
"learning_rate": 4.999756310023261e-06,
"logits/chosen": -1.5627597570419312,
"logits/rejected": -0.6666626334190369,
"logps/chosen": -657.7476806640625,
"logps/rejected": -1387.281494140625,
"loss": 0.239,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.015483086928725243,
"rewards/margins": 0.20804651081562042,
"rewards/rejected": -0.2235296070575714,
"step": 390
},
{
"epoch": 0.11,
"learning_rate": 4.999323102948655e-06,
"logits/chosen": -1.4536590576171875,
"logits/rejected": -0.26167917251586914,
"logps/chosen": -623.3753051757812,
"logps/rejected": -1424.2047119140625,
"loss": 0.249,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.05082948878407478,
"rewards/margins": 0.19376987218856812,
"rewards/rejected": -0.2445993721485138,
"step": 400
},
{
"epoch": 0.11,
"learning_rate": 4.998673339256785e-06,
"logits/chosen": -1.4819873571395874,
"logits/rejected": -0.8635656237602234,
"logps/chosen": -731.3282470703125,
"logps/rejected": -1732.039794921875,
"loss": 0.2004,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.10082012414932251,
"rewards/margins": 0.21832840144634247,
"rewards/rejected": -0.31914854049682617,
"step": 410
},
{
"epoch": 0.11,
"learning_rate": 4.997807075247147e-06,
"logits/chosen": -1.3139228820800781,
"logits/rejected": -0.579411506652832,
"logps/chosen": -703.5878295898438,
"logps/rejected": -1817.680908203125,
"loss": 0.1782,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.12780095636844635,
"rewards/margins": 0.3425312042236328,
"rewards/rejected": -0.47033214569091797,
"step": 420
},
{
"epoch": 0.11,
"learning_rate": 4.996724385978142e-06,
"logits/chosen": -1.5831680297851562,
"logits/rejected": -0.5023525953292847,
"logps/chosen": -746.4442138671875,
"logps/rejected": -1795.2135009765625,
"loss": 0.1981,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12009434401988983,
"rewards/margins": 0.3284095525741577,
"rewards/rejected": -0.44850391149520874,
"step": 430
},
{
"epoch": 0.12,
"learning_rate": 4.995425365260585e-06,
"logits/chosen": -1.4668796062469482,
"logits/rejected": -0.6527413129806519,
"logps/chosen": -602.0538330078125,
"logps/rejected": -1504.60400390625,
"loss": 0.1676,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.08191387355327606,
"rewards/margins": 0.22698941826820374,
"rewards/rejected": -0.3089032769203186,
"step": 440
},
{
"epoch": 0.12,
"learning_rate": 4.993910125649561e-06,
"logits/chosen": -1.5859527587890625,
"logits/rejected": -0.9323417544364929,
"logps/chosen": -573.4935302734375,
"logps/rejected": -1426.5198974609375,
"loss": 0.1951,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.07085999846458435,
"rewards/margins": 0.21237850189208984,
"rewards/rejected": -0.2832385003566742,
"step": 450
},
{
"epoch": 0.12,
"learning_rate": 4.992178798434684e-06,
"logits/chosen": -1.5028969049453735,
"logits/rejected": -0.5813714265823364,
"logps/chosen": -898.3291015625,
"logps/rejected": -1658.8726806640625,
"loss": 0.2782,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1786281317472458,
"rewards/margins": 0.19002899527549744,
"rewards/rejected": -0.36865711212158203,
"step": 460
},
{
"epoch": 0.13,
"learning_rate": 4.990231533628719e-06,
"logits/chosen": -1.6499645709991455,
"logits/rejected": -0.9359930753707886,
"logps/chosen": -693.5921630859375,
"logps/rejected": -1703.711181640625,
"loss": 0.2171,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13478896021842957,
"rewards/margins": 0.2760683298110962,
"rewards/rejected": -0.41085729002952576,
"step": 470
},
{
"epoch": 0.13,
"learning_rate": 4.988068499954578e-06,
"logits/chosen": -1.5587965250015259,
"logits/rejected": -0.4612099528312683,
"logps/chosen": -695.7940673828125,
"logps/rejected": -1690.544189453125,
"loss": 0.2919,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1203017458319664,
"rewards/margins": 0.3253653943538666,
"rewards/rejected": -0.4456671178340912,
"step": 480
},
{
"epoch": 0.13,
"learning_rate": 4.985689884830711e-06,
"logits/chosen": -1.5129863023757935,
"logits/rejected": -0.6448124647140503,
"logps/chosen": -740.9840698242188,
"logps/rejected": -1641.5972900390625,
"loss": 0.1947,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09521947801113129,
"rewards/margins": 0.17064201831817627,
"rewards/rejected": -0.26586148142814636,
"step": 490
},
{
"epoch": 0.13,
"learning_rate": 4.983095894354858e-06,
"logits/chosen": -1.5538814067840576,
"logits/rejected": -0.5965372323989868,
"logps/chosen": -763.7830200195312,
"logps/rejected": -1575.0704345703125,
"loss": 0.2584,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.10452375560998917,
"rewards/margins": 0.13509635627269745,
"rewards/rejected": -0.2396201193332672,
"step": 500
},
{
"epoch": 0.14,
"learning_rate": 4.980286753286196e-06,
"logits/chosen": -1.489684820175171,
"logits/rejected": -0.7012004852294922,
"logps/chosen": -673.0375366210938,
"logps/rejected": -1428.3021240234375,
"loss": 0.202,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.11526355892419815,
"rewards/margins": 0.1490117609500885,
"rewards/rejected": -0.26427531242370605,
"step": 510
},
{
"epoch": 0.14,
"learning_rate": 4.97726270502586e-06,
"logits/chosen": -1.4848878383636475,
"logits/rejected": -0.7628189921379089,
"logps/chosen": -794.6495971679688,
"logps/rejected": -1725.61328125,
"loss": 0.2026,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.17633184790611267,
"rewards/margins": 0.21604189276695251,
"rewards/rejected": -0.3923737406730652,
"step": 520
},
{
"epoch": 0.14,
"learning_rate": 4.974024011595864e-06,
"logits/chosen": -1.5632755756378174,
"logits/rejected": -0.7222310304641724,
"logps/chosen": -932.9734497070312,
"logps/rejected": -1767.3134765625,
"loss": 0.2952,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.20238515734672546,
"rewards/margins": 0.24491646885871887,
"rewards/rejected": -0.44730162620544434,
"step": 530
},
{
"epoch": 0.14,
"learning_rate": 4.970570953616383e-06,
"logits/chosen": -1.4104833602905273,
"logits/rejected": -0.48003751039505005,
"logps/chosen": -668.6126708984375,
"logps/rejected": -1662.20703125,
"loss": 0.2237,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.11609913408756256,
"rewards/margins": 0.2550623416900635,
"rewards/rejected": -0.37116146087646484,
"step": 540
},
{
"epoch": 0.15,
"learning_rate": 4.966903830281449e-06,
"logits/chosen": -1.7994705438613892,
"logits/rejected": -0.39116740226745605,
"logps/chosen": -684.5289916992188,
"logps/rejected": -1445.06298828125,
"loss": 0.2365,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1520206481218338,
"rewards/margins": 0.17612138390541077,
"rewards/rejected": -0.32814204692840576,
"step": 550
},
{
"epoch": 0.15,
"learning_rate": 4.9630229593330226e-06,
"logits/chosen": -1.5381734371185303,
"logits/rejected": -0.836050808429718,
"logps/chosen": -840.8229370117188,
"logps/rejected": -1694.432861328125,
"loss": 0.2301,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.18990156054496765,
"rewards/margins": 0.2644536793231964,
"rewards/rejected": -0.4543551504611969,
"step": 560
},
{
"epoch": 0.15,
"learning_rate": 4.958928677033465e-06,
"logits/chosen": -1.6244480609893799,
"logits/rejected": -0.5760088562965393,
"logps/chosen": -861.2122192382812,
"logps/rejected": -1686.744873046875,
"loss": 0.2336,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.20879682898521423,
"rewards/margins": 0.24533554911613464,
"rewards/rejected": -0.4541323781013489,
"step": 570
},
{
"epoch": 0.15,
"learning_rate": 4.954621338136399e-06,
"logits/chosen": -1.486128807067871,
"logits/rejected": -0.6559261083602905,
"logps/chosen": -924.2649536132812,
"logps/rejected": -1736.8707275390625,
"loss": 0.2448,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1695316731929779,
"rewards/margins": 0.23115842044353485,
"rewards/rejected": -0.40069007873535156,
"step": 580
},
{
"epoch": 0.16,
"learning_rate": 4.95010131585597e-06,
"logits/chosen": -1.217822790145874,
"logits/rejected": -0.44435954093933105,
"logps/chosen": -779.5258178710938,
"logps/rejected": -1578.96240234375,
"loss": 0.2707,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09426208585500717,
"rewards/margins": 0.23891735076904297,
"rewards/rejected": -0.33317944407463074,
"step": 590
},
{
"epoch": 0.16,
"learning_rate": 4.9453690018345144e-06,
"logits/chosen": -1.4555373191833496,
"logits/rejected": -0.5545033812522888,
"logps/chosen": -509.085693359375,
"logps/rejected": -1402.9798583984375,
"loss": 0.1932,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.0637204647064209,
"rewards/margins": 0.23001065850257874,
"rewards/rejected": -0.29373109340667725,
"step": 600
},
{
"epoch": 0.16,
"learning_rate": 4.940424806108619e-06,
"logits/chosen": -1.6595865488052368,
"logits/rejected": -0.6369374990463257,
"logps/chosen": -777.7699584960938,
"logps/rejected": -1567.2138671875,
"loss": 0.1794,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.08186454325914383,
"rewards/margins": 0.22620078921318054,
"rewards/rejected": -0.30806535482406616,
"step": 610
},
{
"epoch": 0.17,
"learning_rate": 4.935269157073597e-06,
"logits/chosen": -1.598024606704712,
"logits/rejected": -1.1374785900115967,
"logps/chosen": -596.6869506835938,
"logps/rejected": -1656.125244140625,
"loss": 0.2165,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.05304652452468872,
"rewards/margins": 0.3274363577365875,
"rewards/rejected": -0.38048288226127625,
"step": 620
},
{
"epoch": 0.17,
"learning_rate": 4.9299025014463665e-06,
"logits/chosen": -1.399924635887146,
"logits/rejected": -0.20592813193798065,
"logps/chosen": -607.9566040039062,
"logps/rejected": -1577.6639404296875,
"loss": 0.183,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.07505299150943756,
"rewards/margins": 0.2847192883491516,
"rewards/rejected": -0.359772264957428,
"step": 630
},
{
"epoch": 0.17,
"learning_rate": 4.924325304226745e-06,
"logits/chosen": -1.6225976943969727,
"logits/rejected": -0.5168310403823853,
"logps/chosen": -793.9441528320312,
"logps/rejected": -1665.4193115234375,
"loss": 0.1601,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09130421280860901,
"rewards/margins": 0.2901184856891632,
"rewards/rejected": -0.3814226984977722,
"step": 640
},
{
"epoch": 0.17,
"learning_rate": 4.91853804865716e-06,
"logits/chosen": -1.0889428853988647,
"logits/rejected": -0.419971764087677,
"logps/chosen": -714.290771484375,
"logps/rejected": -1576.413330078125,
"loss": 0.1755,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1179242953658104,
"rewards/margins": 0.2568149268627167,
"rewards/rejected": -0.37473922967910767,
"step": 650
},
{
"epoch": 0.18,
"learning_rate": 4.912541236180779e-06,
"logits/chosen": -1.3426125049591064,
"logits/rejected": -0.6203486919403076,
"logps/chosen": -884.5665283203125,
"logps/rejected": -1730.358154296875,
"loss": 0.2716,
"rewards/accuracies": 0.625,
"rewards/chosen": -0.2452804297208786,
"rewards/margins": 0.20476552844047546,
"rewards/rejected": -0.4500458836555481,
"step": 660
},
{
"epoch": 0.18,
"learning_rate": 4.9063353863980565e-06,
"logits/chosen": -1.3583014011383057,
"logits/rejected": -0.6006292104721069,
"logps/chosen": -807.7757568359375,
"logps/rejected": -1728.714599609375,
"loss": 0.1892,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1912967413663864,
"rewards/margins": 0.28994446992874146,
"rewards/rejected": -0.48124128580093384,
"step": 670
},
{
"epoch": 0.18,
"learning_rate": 4.899921037021719e-06,
"logits/chosen": -1.689822793006897,
"logits/rejected": -0.9633957147598267,
"logps/chosen": -701.271240234375,
"logps/rejected": -1522.19921875,
"loss": 0.1606,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1669953167438507,
"rewards/margins": 0.24159979820251465,
"rewards/rejected": -0.40859508514404297,
"step": 680
},
{
"epoch": 0.18,
"learning_rate": 4.893298743830168e-06,
"logits/chosen": -1.3598110675811768,
"logits/rejected": -0.3959788382053375,
"logps/chosen": -761.9188842773438,
"logps/rejected": -1724.9876708984375,
"loss": 0.1722,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.19954383373260498,
"rewards/margins": 0.2616916596889496,
"rewards/rejected": -0.46123552322387695,
"step": 690
},
{
"epoch": 0.19,
"learning_rate": 4.88646908061933e-06,
"logits/chosen": -1.4325469732284546,
"logits/rejected": -0.38342922925949097,
"logps/chosen": -740.5088500976562,
"logps/rejected": -1678.9232177734375,
"loss": 0.2414,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.18760952353477478,
"rewards/margins": 0.2778058648109436,
"rewards/rejected": -0.46541541814804077,
"step": 700
},
{
"epoch": 0.19,
"learning_rate": 4.879432639152935e-06,
"logits/chosen": -1.360878586769104,
"logits/rejected": -0.29857271909713745,
"logps/chosen": -834.4906005859375,
"logps/rejected": -1705.520263671875,
"loss": 0.21,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.11962740123271942,
"rewards/margins": 0.23699507117271423,
"rewards/rejected": -0.35662245750427246,
"step": 710
},
{
"epoch": 0.19,
"learning_rate": 4.8721900291112415e-06,
"logits/chosen": -1.567964792251587,
"logits/rejected": -0.6299314498901367,
"logps/chosen": -786.2874755859375,
"logps/rejected": -1548.3878173828125,
"loss": 0.216,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.11282478272914886,
"rewards/margins": 0.20971831679344177,
"rewards/rejected": -0.32254308462142944,
"step": 720
},
{
"epoch": 0.19,
"learning_rate": 4.864741878038218e-06,
"logits/chosen": -1.4125442504882812,
"logits/rejected": -0.6314720511436462,
"logps/chosen": -745.260498046875,
"logps/rejected": -1698.3111572265625,
"loss": 0.1798,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12032542377710342,
"rewards/margins": 0.2931358516216278,
"rewards/rejected": -0.413461297750473,
"step": 730
},
{
"epoch": 0.2,
"learning_rate": 4.857088831287158e-06,
"logits/chosen": -1.5918242931365967,
"logits/rejected": -0.18947651982307434,
"logps/chosen": -734.33984375,
"logps/rejected": -1762.8245849609375,
"loss": 0.1658,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1603078991174698,
"rewards/margins": 0.3126041293144226,
"rewards/rejected": -0.4729120135307312,
"step": 740
},
{
"epoch": 0.2,
"learning_rate": 4.849231551964771e-06,
"logits/chosen": -1.3384909629821777,
"logits/rejected": -0.19086980819702148,
"logps/chosen": -781.5340576171875,
"logps/rejected": -1614.2730712890625,
"loss": 0.2252,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.16732366383075714,
"rewards/margins": 0.21859999001026154,
"rewards/rejected": -0.38592368364334106,
"step": 750
},
{
"epoch": 0.2,
"learning_rate": 4.841170720873723e-06,
"logits/chosen": -1.4520353078842163,
"logits/rejected": -0.30854731798171997,
"logps/chosen": -704.5045776367188,
"logps/rejected": -1557.046875,
"loss": 0.2166,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.14268702268600464,
"rewards/margins": 0.24501653015613556,
"rewards/rejected": -0.38770347833633423,
"step": 760
},
{
"epoch": 0.21,
"learning_rate": 4.832907036453647e-06,
"logits/chosen": -1.4341394901275635,
"logits/rejected": -0.39417794346809387,
"logps/chosen": -831.6419677734375,
"logps/rejected": -1733.0474853515625,
"loss": 0.1779,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.14101889729499817,
"rewards/margins": 0.255978524684906,
"rewards/rejected": -0.39699748158454895,
"step": 770
},
{
"epoch": 0.21,
"learning_rate": 4.824441214720629e-06,
"logits/chosen": -1.2338621616363525,
"logits/rejected": -0.33901381492614746,
"logps/chosen": -600.5255737304688,
"logps/rejected": -1670.0986328125,
"loss": 0.1492,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11527079343795776,
"rewards/margins": 0.3371114134788513,
"rewards/rejected": -0.45238223671913147,
"step": 780
},
{
"epoch": 0.21,
"learning_rate": 4.815773989205165e-06,
"logits/chosen": -1.5467897653579712,
"logits/rejected": -0.7028628587722778,
"logps/chosen": -760.3792114257812,
"logps/rejected": -1956.612548828125,
"loss": 0.1273,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16627143323421478,
"rewards/margins": 0.44062843918800354,
"rewards/rejected": -0.6068998575210571,
"step": 790
},
{
"epoch": 0.21,
"learning_rate": 4.806906110888606e-06,
"logits/chosen": -1.4377644062042236,
"logits/rejected": -0.6477410197257996,
"logps/chosen": -701.1937255859375,
"logps/rejected": -1622.2232666015625,
"loss": 0.2002,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.14753028750419617,
"rewards/margins": 0.2542650103569031,
"rewards/rejected": -0.40179523825645447,
"step": 800
},
{
"epoch": 0.22,
"learning_rate": 4.7978383481380865e-06,
"logits/chosen": -1.40022873878479,
"logits/rejected": -0.5014703869819641,
"logps/chosen": -643.99609375,
"logps/rejected": -1408.9022216796875,
"loss": 0.1944,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.1288548707962036,
"rewards/margins": 0.18380561470985413,
"rewards/rejected": -0.31266045570373535,
"step": 810
},
{
"epoch": 0.22,
"learning_rate": 4.788571486639948e-06,
"logits/chosen": -1.3129093647003174,
"logits/rejected": -0.592149019241333,
"logps/chosen": -842.0194091796875,
"logps/rejected": -1960.482177734375,
"loss": 0.1644,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14255237579345703,
"rewards/margins": 0.33296501636505127,
"rewards/rejected": -0.4755174219608307,
"step": 820
},
{
"epoch": 0.22,
"learning_rate": 4.779106329331665e-06,
"logits/chosen": -1.5025089979171753,
"logits/rejected": -0.3978004455566406,
"logps/chosen": -727.6754150390625,
"logps/rejected": -1607.7178955078125,
"loss": 0.196,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.14143696427345276,
"rewards/margins": 0.26949542760849,
"rewards/rejected": -0.41093239188194275,
"step": 830
},
{
"epoch": 0.22,
"learning_rate": 4.769443696332272e-06,
"logits/chosen": -1.2928297519683838,
"logits/rejected": -0.20549149811267853,
"logps/chosen": -977.98095703125,
"logps/rejected": -2075.953857421875,
"loss": 0.1975,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.2571466565132141,
"rewards/margins": 0.3657899498939514,
"rewards/rejected": -0.6229366064071655,
"step": 840
},
{
"epoch": 0.23,
"learning_rate": 4.759584424871302e-06,
"logits/chosen": -1.2061738967895508,
"logits/rejected": -0.282677561044693,
"logps/chosen": -832.7267456054688,
"logps/rejected": -1914.269775390625,
"loss": 0.1241,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.23771587014198303,
"rewards/margins": 0.34755539894104004,
"rewards/rejected": -0.5852713584899902,
"step": 850
},
{
"epoch": 0.23,
"learning_rate": 4.749529369216246e-06,
"logits/chosen": -1.4981663227081299,
"logits/rejected": -0.8207576870918274,
"logps/chosen": -818.4663696289062,
"logps/rejected": -1752.6500244140625,
"loss": 0.2125,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.19876405596733093,
"rewards/margins": 0.27873286604881287,
"rewards/rejected": -0.4774969518184662,
"step": 860
},
{
"epoch": 0.23,
"learning_rate": 4.7392794005985324e-06,
"logits/chosen": -1.420318841934204,
"logits/rejected": -0.6080266237258911,
"logps/chosen": -673.9572143554688,
"logps/rejected": -1555.6048583984375,
"loss": 0.1973,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.1334313452243805,
"rewards/margins": 0.25675299763679504,
"rewards/rejected": -0.39018434286117554,
"step": 870
},
{
"epoch": 0.23,
"learning_rate": 4.7288354071380415e-06,
"logits/chosen": -1.2416133880615234,
"logits/rejected": 0.2938464283943176,
"logps/chosen": -641.5850830078125,
"logps/rejected": -1431.835205078125,
"loss": 0.2056,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1211976408958435,
"rewards/margins": 0.21377813816070557,
"rewards/rejected": -0.3349757790565491,
"step": 880
},
{
"epoch": 0.24,
"learning_rate": 4.7181982937661485e-06,
"logits/chosen": -1.2874345779418945,
"logits/rejected": -0.42665576934814453,
"logps/chosen": -844.7786254882812,
"logps/rejected": -1806.428955078125,
"loss": 0.1483,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.2129203826189041,
"rewards/margins": 0.2927423417568207,
"rewards/rejected": -0.505662739276886,
"step": 890
},
{
"epoch": 0.24,
"learning_rate": 4.707368982147318e-06,
"logits/chosen": -1.3794301748275757,
"logits/rejected": -0.777490496635437,
"logps/chosen": -880.0616455078125,
"logps/rejected": -1896.2252197265625,
"loss": 0.1484,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.24488544464111328,
"rewards/margins": 0.32845309376716614,
"rewards/rejected": -0.573338508605957,
"step": 900
},
{
"epoch": 0.24,
"learning_rate": 4.696348410599244e-06,
"logits/chosen": -1.3112980127334595,
"logits/rejected": -0.8322477340698242,
"logps/chosen": -693.696044921875,
"logps/rejected": -1638.334716796875,
"loss": 0.1879,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.21305444836616516,
"rewards/margins": 0.29524970054626465,
"rewards/rejected": -0.5083041787147522,
"step": 910
},
{
"epoch": 0.25,
"learning_rate": 4.685137534011549e-06,
"logits/chosen": -1.5619451999664307,
"logits/rejected": -0.2924344539642334,
"logps/chosen": -848.5318603515625,
"logps/rejected": -1702.1826171875,
"loss": 0.2089,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.2551230788230896,
"rewards/margins": 0.2683793008327484,
"rewards/rejected": -0.5235023498535156,
"step": 920
},
{
"epoch": 0.25,
"learning_rate": 4.673737323763048e-06,
"logits/chosen": -1.2145551443099976,
"logits/rejected": -0.5193670392036438,
"logps/chosen": -665.93017578125,
"logps/rejected": -1560.180419921875,
"loss": 0.1732,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.16463886201381683,
"rewards/margins": 0.2519565224647522,
"rewards/rejected": -0.41659536957740784,
"step": 930
},
{
"epoch": 0.25,
"learning_rate": 4.662148767637578e-06,
"logits/chosen": -1.2429286241531372,
"logits/rejected": -0.5839636325836182,
"logps/chosen": -835.6803588867188,
"logps/rejected": -1747.1480712890625,
"loss": 0.2201,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.18446584045886993,
"rewards/margins": 0.2667251229286194,
"rewards/rejected": -0.4511910080909729,
"step": 940
},
{
"epoch": 0.25,
"learning_rate": 4.650372869738415e-06,
"logits/chosen": -1.551847219467163,
"logits/rejected": -0.7213087677955627,
"logps/chosen": -739.1425170898438,
"logps/rejected": -1562.636962890625,
"loss": 0.2208,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.10864777863025665,
"rewards/margins": 0.25965723395347595,
"rewards/rejected": -0.3683050274848938,
"step": 950
},
{
"epoch": 0.26,
"learning_rate": 4.638410650401267e-06,
"logits/chosen": -1.327823281288147,
"logits/rejected": -0.146169975399971,
"logps/chosen": -627.8694458007812,
"logps/rejected": -1442.491455078125,
"loss": 0.2464,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09383858740329742,
"rewards/margins": 0.18377116322517395,
"rewards/rejected": -0.27760976552963257,
"step": 960
},
{
"epoch": 0.26,
"learning_rate": 4.626263146105875e-06,
"logits/chosen": -1.613373041152954,
"logits/rejected": -0.5505753755569458,
"logps/chosen": -789.1390380859375,
"logps/rejected": -1607.309326171875,
"loss": 0.2438,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.1650734841823578,
"rewards/margins": 0.2032967060804367,
"rewards/rejected": -0.3683702051639557,
"step": 970
},
{
"epoch": 0.26,
"learning_rate": 4.613931409386196e-06,
"logits/chosen": -1.406640648841858,
"logits/rejected": -0.3088572323322296,
"logps/chosen": -764.7514038085938,
"logps/rejected": -1577.6644287109375,
"loss": 0.2139,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.15214017033576965,
"rewards/margins": 0.2213418185710907,
"rewards/rejected": -0.37348201870918274,
"step": 980
},
{
"epoch": 0.26,
"learning_rate": 4.601416508739211e-06,
"logits/chosen": -1.4507434368133545,
"logits/rejected": -0.6369355916976929,
"logps/chosen": -740.4591674804688,
"logps/rejected": -1661.3743896484375,
"loss": 0.1741,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13703958690166473,
"rewards/margins": 0.26543471217155457,
"rewards/rejected": -0.4024743139743805,
"step": 990
},
{
"epoch": 0.27,
"learning_rate": 4.588719528532342e-06,
"logits/chosen": -1.3239177465438843,
"logits/rejected": -0.3033704161643982,
"logps/chosen": -753.6892700195312,
"logps/rejected": -1676.83984375,
"loss": 0.209,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.12344950437545776,
"rewards/margins": 0.274679034948349,
"rewards/rejected": -0.3981285095214844,
"step": 1000
},
{
"epoch": 0.27,
"learning_rate": 4.575841568909494e-06,
"logits/chosen": -1.6556479930877686,
"logits/rejected": -0.39285674691200256,
"logps/chosen": -722.4796142578125,
"logps/rejected": -1681.2886962890625,
"loss": 0.1905,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.09982794523239136,
"rewards/margins": 0.2901255488395691,
"rewards/rejected": -0.38995346426963806,
"step": 1010
},
{
"epoch": 0.27,
"learning_rate": 4.562783745695738e-06,
"logits/chosen": -1.4957317113876343,
"logits/rejected": 0.11952666193246841,
"logps/chosen": -714.2357177734375,
"logps/rejected": -1432.6864013671875,
"loss": 0.2066,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09190934896469116,
"rewards/margins": 0.20793361961841583,
"rewards/rejected": -0.2998429834842682,
"step": 1020
},
{
"epoch": 0.27,
"learning_rate": 4.549547190300622e-06,
"logits/chosen": -1.2969706058502197,
"logits/rejected": -0.45746153593063354,
"logps/chosen": -787.8223266601562,
"logps/rejected": -1744.7171630859375,
"loss": 0.2282,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.161489337682724,
"rewards/margins": 0.31587809324264526,
"rewards/rejected": -0.4773674011230469,
"step": 1030
},
{
"epoch": 0.28,
"learning_rate": 4.536133049620143e-06,
"logits/chosen": -1.13800048828125,
"logits/rejected": -0.3455226421356201,
"logps/chosen": -721.7028198242188,
"logps/rejected": -1831.574951171875,
"loss": 0.1279,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.12353639304637909,
"rewards/margins": 0.3737573027610779,
"rewards/rejected": -0.49729371070861816,
"step": 1040
},
{
"epoch": 0.28,
"learning_rate": 4.522542485937369e-06,
"logits/chosen": -1.4609827995300293,
"logits/rejected": -0.496757835149765,
"logps/chosen": -781.8280029296875,
"logps/rejected": -1767.3248291015625,
"loss": 0.1728,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15113988518714905,
"rewards/margins": 0.3218175768852234,
"rewards/rejected": -0.4729575216770172,
"step": 1050
},
{
"epoch": 0.28,
"learning_rate": 4.508776676821739e-06,
"logits/chosen": -1.4414708614349365,
"logits/rejected": -0.6857733726501465,
"logps/chosen": -832.9376831054688,
"logps/rejected": -1553.129150390625,
"loss": 0.1908,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.17810413241386414,
"rewards/margins": 0.21977660059928894,
"rewards/rejected": -0.3978807330131531,
"step": 1060
},
{
"epoch": 0.29,
"learning_rate": 4.494836815027022e-06,
"logits/chosen": -1.3217737674713135,
"logits/rejected": 0.03468703106045723,
"logps/chosen": -752.7827758789062,
"logps/rejected": -1703.7191162109375,
"loss": 0.2323,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.14807240664958954,
"rewards/margins": 0.2589189410209656,
"rewards/rejected": -0.4069913923740387,
"step": 1070
},
{
"epoch": 0.29,
"learning_rate": 4.4807241083879774e-06,
"logits/chosen": -1.3331490755081177,
"logits/rejected": -0.2640678286552429,
"logps/chosen": -728.1994018554688,
"logps/rejected": -1618.332275390625,
"loss": 0.2073,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1415979564189911,
"rewards/margins": 0.31241506338119507,
"rewards/rejected": -0.4540129601955414,
"step": 1080
},
{
"epoch": 0.29,
"learning_rate": 4.466439779715696e-06,
"logits/chosen": -1.320003867149353,
"logits/rejected": -0.5086170434951782,
"logps/chosen": -710.85009765625,
"logps/rejected": -1577.6478271484375,
"loss": 0.1914,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09440721571445465,
"rewards/margins": 0.2844494879245758,
"rewards/rejected": -0.37885671854019165,
"step": 1090
},
{
"epoch": 0.29,
"learning_rate": 4.451985066691649e-06,
"logits/chosen": -1.3163551092147827,
"logits/rejected": -0.6709269285202026,
"logps/chosen": -674.307373046875,
"logps/rejected": -1554.60498046875,
"loss": 0.1749,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.10322503000497818,
"rewards/margins": 0.28593048453330994,
"rewards/rejected": -0.3891555070877075,
"step": 1100
},
{
"epoch": 0.3,
"learning_rate": 4.437361221760449e-06,
"logits/chosen": -1.5619227886199951,
"logits/rejected": -0.5209615230560303,
"logps/chosen": -693.013916015625,
"logps/rejected": -1884.20703125,
"loss": 0.1298,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.11155110597610474,
"rewards/margins": 0.3766392171382904,
"rewards/rejected": -0.48819035291671753,
"step": 1110
},
{
"epoch": 0.3,
"learning_rate": 4.422569512021332e-06,
"logits/chosen": -1.4108737707138062,
"logits/rejected": 0.11136605590581894,
"logps/chosen": -703.5150146484375,
"logps/rejected": -1624.7801513671875,
"loss": 0.2085,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.10121216624975204,
"rewards/margins": 0.26922935247421265,
"rewards/rejected": -0.3704415559768677,
"step": 1120
},
{
"epoch": 0.3,
"learning_rate": 4.407611219118363e-06,
"logits/chosen": -1.353237509727478,
"logits/rejected": -0.5166940093040466,
"logps/chosen": -706.7504272460938,
"logps/rejected": -1707.227294921875,
"loss": 0.1646,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.144940584897995,
"rewards/margins": 0.2921007573604584,
"rewards/rejected": -0.43704134225845337,
"step": 1130
},
{
"epoch": 0.3,
"learning_rate": 4.3924876391293915e-06,
"logits/chosen": -1.3774515390396118,
"logits/rejected": -0.8835982084274292,
"logps/chosen": -653.8270263671875,
"logps/rejected": -1665.271728515625,
"loss": 0.1742,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.11839635670185089,
"rewards/margins": 0.2974711060523987,
"rewards/rejected": -0.41586747765541077,
"step": 1140
},
{
"epoch": 0.31,
"learning_rate": 4.377200082453748e-06,
"logits/chosen": -1.3856669664382935,
"logits/rejected": -0.6063768863677979,
"logps/chosen": -692.400390625,
"logps/rejected": -1679.411865234375,
"loss": 0.2769,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.12366262823343277,
"rewards/margins": 0.3368341326713562,
"rewards/rejected": -0.4604967534542084,
"step": 1150
},
{
"epoch": 0.31,
"learning_rate": 4.361749873698707e-06,
"logits/chosen": -1.4405038356781006,
"logits/rejected": -0.6485291123390198,
"logps/chosen": -696.2839965820312,
"logps/rejected": -1726.019775390625,
"loss": 0.1222,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.08081836253404617,
"rewards/margins": 0.3042358160018921,
"rewards/rejected": -0.38505417108535767,
"step": 1160
},
{
"epoch": 0.31,
"learning_rate": 4.346138351564711e-06,
"logits/chosen": -1.443703055381775,
"logits/rejected": -0.754281759262085,
"logps/chosen": -663.7955932617188,
"logps/rejected": -1472.655029296875,
"loss": 0.2056,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.05589227005839348,
"rewards/margins": 0.21428577601909637,
"rewards/rejected": -0.27017804980278015,
"step": 1170
},
{
"epoch": 0.31,
"learning_rate": 4.330366868729376e-06,
"logits/chosen": -1.3844349384307861,
"logits/rejected": -0.6004733443260193,
"logps/chosen": -717.2481689453125,
"logps/rejected": -1639.61328125,
"loss": 0.1903,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09332798421382904,
"rewards/margins": 0.2749372124671936,
"rewards/rejected": -0.36826521158218384,
"step": 1180
},
{
"epoch": 0.32,
"learning_rate": 4.3144367917302964e-06,
"logits/chosen": -1.3772212266921997,
"logits/rejected": -1.007070779800415,
"logps/chosen": -626.3247680664062,
"logps/rejected": -1674.417724609375,
"loss": 0.1665,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.09171278029680252,
"rewards/margins": 0.32611554861068726,
"rewards/rejected": -0.4178283214569092,
"step": 1190
},
{
"epoch": 0.32,
"learning_rate": 4.2983495008466285e-06,
"logits/chosen": -1.3759549856185913,
"logits/rejected": -0.7071571350097656,
"logps/chosen": -691.9136962890625,
"logps/rejected": -1873.8238525390625,
"loss": 0.2082,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.13154932856559753,
"rewards/margins": 0.37329337000846863,
"rewards/rejected": -0.5048426389694214,
"step": 1200
},
{
"epoch": 0.32,
"learning_rate": 4.2821063899795015e-06,
"logits/chosen": -1.50946044921875,
"logits/rejected": -0.23106630146503448,
"logps/chosen": -669.9387817382812,
"logps/rejected": -1659.885986328125,
"loss": 0.1434,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07889886945486069,
"rewards/margins": 0.3387802839279175,
"rewards/rejected": -0.4176791310310364,
"step": 1210
},
{
"epoch": 0.33,
"learning_rate": 4.265708866531238e-06,
"logits/chosen": -1.5640531778335571,
"logits/rejected": -0.577285885810852,
"logps/chosen": -672.6451416015625,
"logps/rejected": -1411.490966796875,
"loss": 0.1828,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.0744541734457016,
"rewards/margins": 0.2126346081495285,
"rewards/rejected": -0.2870888113975525,
"step": 1220
},
{
"epoch": 0.33,
"learning_rate": 4.249158351283414e-06,
"logits/chosen": -1.3422248363494873,
"logits/rejected": -0.3964029848575592,
"logps/chosen": -660.489990234375,
"logps/rejected": -1654.5406494140625,
"loss": 0.1949,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.10021636635065079,
"rewards/margins": 0.22853437066078186,
"rewards/rejected": -0.32875072956085205,
"step": 1230
},
{
"epoch": 0.33,
"learning_rate": 4.232456278273743e-06,
"logits/chosen": -1.4052772521972656,
"logits/rejected": -1.0742384195327759,
"logps/chosen": -727.620849609375,
"logps/rejected": -1368.65283203125,
"loss": 0.2202,
"rewards/accuracies": 0.6000000238418579,
"rewards/chosen": -0.13287192583084106,
"rewards/margins": 0.17202439904212952,
"rewards/rejected": -0.3048963248729706,
"step": 1240
},
{
"epoch": 0.33,
"learning_rate": 4.215604094671835e-06,
"logits/chosen": -1.5828149318695068,
"logits/rejected": -0.8305169939994812,
"logps/chosen": -674.4658813476562,
"logps/rejected": -1537.18212890625,
"loss": 0.269,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.13081403076648712,
"rewards/margins": 0.2305031269788742,
"rewards/rejected": -0.3613171875476837,
"step": 1250
},
{
"epoch": 0.34,
"learning_rate": 4.198603260653792e-06,
"logits/chosen": -1.4573118686676025,
"logits/rejected": -0.01602686010301113,
"logps/chosen": -745.9620361328125,
"logps/rejected": -1567.2252197265625,
"loss": 0.16,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.08418073505163193,
"rewards/margins": 0.2803150415420532,
"rewards/rejected": -0.36449578404426575,
"step": 1260
},
{
"epoch": 0.34,
"learning_rate": 4.181455249275701e-06,
"logits/chosen": -1.1945483684539795,
"logits/rejected": -0.8259621858596802,
"logps/chosen": -676.2755126953125,
"logps/rejected": -1655.5445556640625,
"loss": 0.2373,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.09749835729598999,
"rewards/margins": 0.27395281195640564,
"rewards/rejected": -0.371451199054718,
"step": 1270
},
{
"epoch": 0.34,
"learning_rate": 4.1641615463459926e-06,
"logits/chosen": -1.4833742380142212,
"logits/rejected": -0.5753664970397949,
"logps/chosen": -676.084716796875,
"logps/rejected": -1957.914794921875,
"loss": 0.1464,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.045958876609802246,
"rewards/margins": 0.3767011761665344,
"rewards/rejected": -0.42266005277633667,
"step": 1280
},
{
"epoch": 0.34,
"learning_rate": 4.146723650296701e-06,
"logits/chosen": -1.5337693691253662,
"logits/rejected": -0.5462034344673157,
"logps/chosen": -618.5125732421875,
"logps/rejected": -1453.6705322265625,
"loss": 0.1498,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.06487244367599487,
"rewards/margins": 0.21656076610088348,
"rewards/rejected": -0.28143322467803955,
"step": 1290
},
{
"epoch": 0.35,
"learning_rate": 4.129143072053639e-06,
"logits/chosen": -1.6736488342285156,
"logits/rejected": -0.11961223185062408,
"logps/chosen": -786.7601318359375,
"logps/rejected": -1709.4732666015625,
"loss": 0.1415,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08835209906101227,
"rewards/margins": 0.2948789596557617,
"rewards/rejected": -0.3832310736179352,
"step": 1300
},
{
"epoch": 0.35,
"learning_rate": 4.111421334905468e-06,
"logits/chosen": -1.440685510635376,
"logits/rejected": -0.4499839246273041,
"logps/chosen": -866.4661254882812,
"logps/rejected": -1788.9306640625,
"loss": 0.2314,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12372098863124847,
"rewards/margins": 0.28645357489585876,
"rewards/rejected": -0.41017454862594604,
"step": 1310
},
{
"epoch": 0.35,
"learning_rate": 4.093559974371725e-06,
"logits/chosen": -1.2748010158538818,
"logits/rejected": -0.5881192684173584,
"logps/chosen": -938.8273315429688,
"logps/rejected": -1917.021728515625,
"loss": 0.1504,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16988976299762726,
"rewards/margins": 0.32855165004730225,
"rewards/rejected": -0.4984413981437683,
"step": 1320
},
{
"epoch": 0.35,
"learning_rate": 4.075560538069767e-06,
"logits/chosen": -1.4327054023742676,
"logits/rejected": -0.4992315173149109,
"logps/chosen": -743.90234375,
"logps/rejected": -1680.1441650390625,
"loss": 0.1669,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.09745943546295166,
"rewards/margins": 0.39398160576820374,
"rewards/rejected": -0.4914410710334778,
"step": 1330
},
{
"epoch": 0.36,
"learning_rate": 4.05742458558068e-06,
"logits/chosen": -1.5205237865447998,
"logits/rejected": -0.47205105423927307,
"logps/chosen": -656.0806884765625,
"logps/rejected": -1458.2177734375,
"loss": 0.1959,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.08391676843166351,
"rewards/margins": 0.2383715659379959,
"rewards/rejected": -0.32228830456733704,
"step": 1340
},
{
"epoch": 0.36,
"learning_rate": 4.039153688314146e-06,
"logits/chosen": -1.5271790027618408,
"logits/rejected": -0.13848623633384705,
"logps/chosen": -667.71533203125,
"logps/rejected": -1672.637939453125,
"loss": 0.1946,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.04903538152575493,
"rewards/margins": 0.32115495204925537,
"rewards/rejected": -0.3701903223991394,
"step": 1350
},
{
"epoch": 0.36,
"learning_rate": 4.020749429372286e-06,
"logits/chosen": -1.455418348312378,
"logits/rejected": -0.3692135810852051,
"logps/chosen": -778.6569213867188,
"logps/rejected": -1898.542236328125,
"loss": 0.1951,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.10191953182220459,
"rewards/margins": 0.3559776544570923,
"rewards/rejected": -0.45789724588394165,
"step": 1360
},
{
"epoch": 0.37,
"learning_rate": 4.002213403412492e-06,
"logits/chosen": -1.3963444232940674,
"logits/rejected": -0.1295473426580429,
"logps/chosen": -613.0379028320312,
"logps/rejected": -1605.5289306640625,
"loss": 0.2558,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09437403827905655,
"rewards/margins": 0.2921295762062073,
"rewards/rejected": -0.3865036368370056,
"step": 1370
},
{
"epoch": 0.37,
"learning_rate": 3.983547216509254e-06,
"logits/chosen": -1.677065134048462,
"logits/rejected": -0.4172869622707367,
"logps/chosen": -623.8750610351562,
"logps/rejected": -1598.5770263671875,
"loss": 0.1837,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.06079324334859848,
"rewards/margins": 0.25734245777130127,
"rewards/rejected": -0.31813567876815796,
"step": 1380
},
{
"epoch": 0.37,
"learning_rate": 3.964752486015001e-06,
"logits/chosen": -1.5321755409240723,
"logits/rejected": -0.6257954835891724,
"logps/chosen": -681.2603759765625,
"logps/rejected": -1647.580810546875,
"loss": 0.179,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07860198616981506,
"rewards/margins": 0.2607461214065552,
"rewards/rejected": -0.3393481373786926,
"step": 1390
},
{
"epoch": 0.37,
"learning_rate": 3.945830840419966e-06,
"logits/chosen": -1.5395901203155518,
"logits/rejected": -0.23838794231414795,
"logps/chosen": -678.4739990234375,
"logps/rejected": -1783.0814208984375,
"loss": 0.1444,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -0.10125972330570221,
"rewards/margins": 0.36801964044570923,
"rewards/rejected": -0.46927928924560547,
"step": 1400
},
{
"epoch": 0.38,
"learning_rate": 3.92678391921108e-06,
"logits/chosen": -1.4998096227645874,
"logits/rejected": -0.35734957456588745,
"logps/chosen": -606.8898315429688,
"logps/rejected": -1602.2557373046875,
"loss": 0.1684,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09188471734523773,
"rewards/margins": 0.30264589190483093,
"rewards/rejected": -0.39453059434890747,
"step": 1410
},
{
"epoch": 0.38,
"learning_rate": 3.907613372729916e-06,
"logits/chosen": -1.3110413551330566,
"logits/rejected": -0.36875995993614197,
"logps/chosen": -673.6112060546875,
"logps/rejected": -1617.9066162109375,
"loss": 0.1956,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.11483339965343475,
"rewards/margins": 0.2649555504322052,
"rewards/rejected": -0.37978893518447876,
"step": 1420
},
{
"epoch": 0.38,
"learning_rate": 3.888320862029699e-06,
"logits/chosen": -1.726869821548462,
"logits/rejected": -0.4512055814266205,
"logps/chosen": -697.3773193359375,
"logps/rejected": -1701.3092041015625,
"loss": 0.2008,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07687092572450638,
"rewards/margins": 0.31277498602867126,
"rewards/rejected": -0.38964587450027466,
"step": 1430
},
{
"epoch": 0.38,
"learning_rate": 3.868908058731376e-06,
"logits/chosen": -1.5626089572906494,
"logits/rejected": -0.7535050511360168,
"logps/chosen": -776.7896118164062,
"logps/rejected": -1669.3294677734375,
"loss": 0.1674,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.054000310599803925,
"rewards/margins": 0.27108216285705566,
"rewards/rejected": -0.3250824809074402,
"step": 1440
},
{
"epoch": 0.39,
"learning_rate": 3.849376644878783e-06,
"logits/chosen": -1.3612600564956665,
"logits/rejected": -0.8206363916397095,
"logps/chosen": -592.5975341796875,
"logps/rejected": -1729.0345458984375,
"loss": 0.1497,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.03538142889738083,
"rewards/margins": 0.3380415439605713,
"rewards/rejected": -0.3734230101108551,
"step": 1450
},
{
"epoch": 0.39,
"learning_rate": 3.829728312792895e-06,
"logits/chosen": -1.6738402843475342,
"logits/rejected": -0.5649760961532593,
"logps/chosen": -617.51123046875,
"logps/rejected": -1514.9173583984375,
"loss": 0.2153,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.08767777681350708,
"rewards/margins": 0.2846985459327698,
"rewards/rejected": -0.37237635254859924,
"step": 1460
},
{
"epoch": 0.39,
"learning_rate": 3.8099647649251984e-06,
"logits/chosen": -1.2732475996017456,
"logits/rejected": -0.4450169503688812,
"logps/chosen": -827.7120361328125,
"logps/rejected": -1842.5130615234375,
"loss": 0.157,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.15502184629440308,
"rewards/margins": 0.3444690704345703,
"rewards/rejected": -0.4994909167289734,
"step": 1470
},
{
"epoch": 0.39,
"learning_rate": 3.790087713710179e-06,
"logits/chosen": -1.0923655033111572,
"logits/rejected": 0.02032681182026863,
"logps/chosen": -974.3514404296875,
"logps/rejected": -1917.356689453125,
"loss": 0.22,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.23814789950847626,
"rewards/margins": 0.2912440896034241,
"rewards/rejected": -0.5293919444084167,
"step": 1480
},
{
"epoch": 0.4,
"learning_rate": 3.770098881416945e-06,
"logits/chosen": -1.387081265449524,
"logits/rejected": -0.628162145614624,
"logps/chosen": -868.3365478515625,
"logps/rejected": -1810.2318115234375,
"loss": 0.1962,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20213904976844788,
"rewards/margins": 0.28864550590515137,
"rewards/rejected": -0.49078455567359924,
"step": 1490
},
{
"epoch": 0.4,
"learning_rate": 3.7500000000000005e-06,
"logits/chosen": -1.5495609045028687,
"logits/rejected": -0.25171366333961487,
"logps/chosen": -735.914794921875,
"logps/rejected": -1725.832763671875,
"loss": 0.1646,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.14493001997470856,
"rewards/margins": 0.2965463101863861,
"rewards/rejected": -0.44147634506225586,
"step": 1500
},
{
"epoch": 0.4,
"learning_rate": 3.7297928109491765e-06,
"logits/chosen": -1.3633923530578613,
"logits/rejected": -0.3343961536884308,
"logps/chosen": -621.5894165039062,
"logps/rejected": -1620.972412109375,
"loss": 0.145,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.11256246268749237,
"rewards/margins": 0.33683252334594727,
"rewards/rejected": -0.44939494132995605,
"step": 1510
},
{
"epoch": 0.41,
"learning_rate": 3.7094790651387414e-06,
"logits/chosen": -1.618963599205017,
"logits/rejected": -0.7474344968795776,
"logps/chosen": -682.3802490234375,
"logps/rejected": -1560.806396484375,
"loss": 0.1795,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.13379545509815216,
"rewards/margins": 0.25923866033554077,
"rewards/rejected": -0.3930341303348541,
"step": 1520
},
{
"epoch": 0.41,
"learning_rate": 3.689060522675689e-06,
"logits/chosen": -1.407928228378296,
"logits/rejected": -0.4922426640987396,
"logps/chosen": -765.1798706054688,
"logps/rejected": -1726.2021484375,
"loss": 0.2049,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13442331552505493,
"rewards/margins": 0.28671473264694214,
"rewards/rejected": -0.42113804817199707,
"step": 1530
},
{
"epoch": 0.41,
"learning_rate": 3.668538952747236e-06,
"logits/chosen": -1.621731162071228,
"logits/rejected": -0.3162292540073395,
"logps/chosen": -807.1463623046875,
"logps/rejected": -1907.406982421875,
"loss": 0.1259,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.12251107394695282,
"rewards/margins": 0.4035162925720215,
"rewards/rejected": -0.5260274410247803,
"step": 1540
},
{
"epoch": 0.41,
"learning_rate": 3.6479161334675294e-06,
"logits/chosen": -1.415895938873291,
"logits/rejected": 0.0497395396232605,
"logps/chosen": -814.912353515625,
"logps/rejected": -1617.039306640625,
"loss": 0.2032,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.13029535114765167,
"rewards/margins": 0.23458731174468994,
"rewards/rejected": -0.3648826479911804,
"step": 1550
},
{
"epoch": 0.42,
"learning_rate": 3.627193851723577e-06,
"logits/chosen": -1.3018121719360352,
"logits/rejected": -0.9184350967407227,
"logps/chosen": -798.8698120117188,
"logps/rejected": -1686.4742431640625,
"loss": 0.212,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.12557171285152435,
"rewards/margins": 0.2953474819660187,
"rewards/rejected": -0.42091917991638184,
"step": 1560
},
{
"epoch": 0.42,
"learning_rate": 3.6063739030204226e-06,
"logits/chosen": -1.5564700365066528,
"logits/rejected": -0.8456588983535767,
"logps/chosen": -672.8592529296875,
"logps/rejected": -1566.520263671875,
"loss": 0.1958,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.13787488639354706,
"rewards/margins": 0.28301936388015747,
"rewards/rejected": -0.42089423537254333,
"step": 1570
},
{
"epoch": 0.42,
"learning_rate": 3.5854580913255706e-06,
"logits/chosen": -1.3799980878829956,
"logits/rejected": -0.09411342442035675,
"logps/chosen": -759.1764526367188,
"logps/rejected": -1664.320068359375,
"loss": 0.2203,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1847977340221405,
"rewards/margins": 0.2686237096786499,
"rewards/rejected": -0.4534215033054352,
"step": 1580
},
{
"epoch": 0.42,
"learning_rate": 3.564448228912682e-06,
"logits/chosen": -1.5500601530075073,
"logits/rejected": -0.6304312348365784,
"logps/chosen": -741.7132568359375,
"logps/rejected": -1747.054443359375,
"loss": 0.1585,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.11834660917520523,
"rewards/margins": 0.3032262921333313,
"rewards/rejected": -0.4215729236602783,
"step": 1590
},
{
"epoch": 0.43,
"learning_rate": 3.543346136204545e-06,
"logits/chosen": -1.2121784687042236,
"logits/rejected": -0.23208335041999817,
"logps/chosen": -896.2943115234375,
"logps/rejected": -1856.9176025390625,
"loss": 0.1546,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.18068882822990417,
"rewards/margins": 0.29168596863746643,
"rewards/rejected": -0.472374826669693,
"step": 1600
},
{
"epoch": 0.43,
"learning_rate": 3.522153641615345e-06,
"logits/chosen": -1.388629674911499,
"logits/rejected": -0.5730769038200378,
"logps/chosen": -736.3162841796875,
"logps/rejected": -1783.465576171875,
"loss": 0.1359,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.12700147926807404,
"rewards/margins": 0.37471193075180054,
"rewards/rejected": -0.5017133951187134,
"step": 1610
},
{
"epoch": 0.43,
"learning_rate": 3.5008725813922383e-06,
"logits/chosen": -1.4097487926483154,
"logits/rejected": -0.479957640171051,
"logps/chosen": -869.1336669921875,
"logps/rejected": -1629.0390625,
"loss": 0.1687,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1220046877861023,
"rewards/margins": 0.2969478964805603,
"rewards/rejected": -0.4189525544643402,
"step": 1620
},
{
"epoch": 0.43,
"learning_rate": 3.4795047994562463e-06,
"logits/chosen": -1.4321268796920776,
"logits/rejected": -0.3780360221862793,
"logps/chosen": -764.9324340820312,
"logps/rejected": -1601.4052734375,
"loss": 0.2108,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.12771736085414886,
"rewards/margins": 0.2464291751384735,
"rewards/rejected": -0.3741465210914612,
"step": 1630
},
{
"epoch": 0.44,
"learning_rate": 3.458052147242494e-06,
"logits/chosen": -1.6645126342773438,
"logits/rejected": -0.8111234903335571,
"logps/chosen": -817.3748168945312,
"logps/rejected": -1478.569091796875,
"loss": 0.1886,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1138969212770462,
"rewards/margins": 0.18457883596420288,
"rewards/rejected": -0.2984757423400879,
"step": 1640
},
{
"epoch": 0.44,
"learning_rate": 3.436516483539781e-06,
"logits/chosen": -1.6396286487579346,
"logits/rejected": -0.3675927221775055,
"logps/chosen": -737.4990234375,
"logps/rejected": -1537.5166015625,
"loss": 0.187,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09592025727033615,
"rewards/margins": 0.21111655235290527,
"rewards/rejected": -0.307036817073822,
"step": 1650
},
{
"epoch": 0.44,
"learning_rate": 3.4148996743295305e-06,
"logits/chosen": -1.654663324356079,
"logits/rejected": -0.5259418487548828,
"logps/chosen": -859.1363525390625,
"logps/rejected": -1622.9510498046875,
"loss": 0.2101,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.06407856196165085,
"rewards/margins": 0.22986917197704315,
"rewards/rejected": -0.2939477562904358,
"step": 1660
},
{
"epoch": 0.45,
"learning_rate": 3.3932035926241103e-06,
"logits/chosen": -1.4901096820831299,
"logits/rejected": -1.160047173500061,
"logps/chosen": -624.4628295898438,
"logps/rejected": -1744.197021484375,
"loss": 0.1465,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.04981148615479469,
"rewards/margins": 0.3574128746986389,
"rewards/rejected": -0.4072244167327881,
"step": 1670
},
{
"epoch": 0.45,
"learning_rate": 3.3714301183045382e-06,
"logits/chosen": -1.4609191417694092,
"logits/rejected": -0.5895389318466187,
"logps/chosen": -690.073974609375,
"logps/rejected": -1626.9462890625,
"loss": 0.1726,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.08027863502502441,
"rewards/margins": 0.2669757008552551,
"rewards/rejected": -0.34725433588027954,
"step": 1680
},
{
"epoch": 0.45,
"learning_rate": 3.349581137957604e-06,
"logits/chosen": -1.6755574941635132,
"logits/rejected": -0.5948814153671265,
"logps/chosen": -704.0479736328125,
"logps/rejected": -1561.6595458984375,
"loss": 0.2339,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.08259834349155426,
"rewards/margins": 0.2650589346885681,
"rewards/rejected": -0.3476572632789612,
"step": 1690
},
{
"epoch": 0.45,
"learning_rate": 3.3276585447123957e-06,
"logits/chosen": -1.1900542974472046,
"logits/rejected": -0.02419034019112587,
"logps/chosen": -666.2265625,
"logps/rejected": -1475.4140625,
"loss": 0.2067,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.09865899384021759,
"rewards/margins": 0.23837342858314514,
"rewards/rejected": -0.33703240752220154,
"step": 1700
},
{
"epoch": 0.46,
"learning_rate": 3.3056642380762783e-06,
"logits/chosen": -1.4822317361831665,
"logits/rejected": -0.4478825032711029,
"logps/chosen": -719.041015625,
"logps/rejected": -1704.8245849609375,
"loss": 0.1518,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.046217575669288635,
"rewards/margins": 0.35374483466148376,
"rewards/rejected": -0.3999623954296112,
"step": 1710
},
{
"epoch": 0.46,
"learning_rate": 3.2836001237702993e-06,
"logits/chosen": -1.5582091808319092,
"logits/rejected": -0.370614230632782,
"logps/chosen": -738.9437255859375,
"logps/rejected": -1776.6015625,
"loss": 0.1323,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.05881623178720474,
"rewards/margins": 0.310192734003067,
"rewards/rejected": -0.36900898814201355,
"step": 1720
},
{
"epoch": 0.46,
"learning_rate": 3.2614681135640696e-06,
"logits/chosen": -1.4925451278686523,
"logits/rejected": -0.8497873544692993,
"logps/chosen": -724.83935546875,
"logps/rejected": -1761.577392578125,
"loss": 0.1634,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07654894888401031,
"rewards/margins": 0.24484257400035858,
"rewards/rejected": -0.3213915228843689,
"step": 1730
},
{
"epoch": 0.46,
"learning_rate": 3.2392701251101172e-06,
"logits/chosen": -1.396722435951233,
"logits/rejected": -0.7455130815505981,
"logps/chosen": -674.4617309570312,
"logps/rejected": -1584.518798828125,
"loss": 0.1387,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.08803001791238785,
"rewards/margins": 0.26931852102279663,
"rewards/rejected": -0.3573485314846039,
"step": 1740
},
{
"epoch": 0.47,
"learning_rate": 3.217008081777726e-06,
"logits/chosen": -1.504368543624878,
"logits/rejected": -0.6314746141433716,
"logps/chosen": -874.5758056640625,
"logps/rejected": -1906.4967041015625,
"loss": 0.1907,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1991203874349594,
"rewards/margins": 0.36806267499923706,
"rewards/rejected": -0.5671831369400024,
"step": 1750
},
{
"epoch": 0.47,
"learning_rate": 3.1946839124862873e-06,
"logits/chosen": -1.2642552852630615,
"logits/rejected": -0.10561282932758331,
"logps/chosen": -707.1906127929688,
"logps/rejected": -1762.505126953125,
"loss": 0.1346,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12033917009830475,
"rewards/margins": 0.343368262052536,
"rewards/rejected": -0.4637073874473572,
"step": 1760
},
{
"epoch": 0.47,
"learning_rate": 3.1722995515381644e-06,
"logits/chosen": -1.3440335988998413,
"logits/rejected": -0.3857986330986023,
"logps/chosen": -733.5349731445312,
"logps/rejected": -1664.541748046875,
"loss": 0.2086,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.11129789054393768,
"rewards/margins": 0.3099938631057739,
"rewards/rejected": -0.4212917685508728,
"step": 1770
},
{
"epoch": 0.47,
"learning_rate": 3.149856938451094e-06,
"logits/chosen": -1.6117057800292969,
"logits/rejected": 0.15098969638347626,
"logps/chosen": -814.0639038085938,
"logps/rejected": -1576.1326904296875,
"loss": 0.1593,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12756898999214172,
"rewards/margins": 0.25667664408683777,
"rewards/rejected": -0.38424569368362427,
"step": 1780
},
{
"epoch": 0.48,
"learning_rate": 3.127358017790132e-06,
"logits/chosen": -1.3055508136749268,
"logits/rejected": 0.43101415038108826,
"logps/chosen": -844.7630615234375,
"logps/rejected": -1807.894287109375,
"loss": 0.1253,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.14493486285209656,
"rewards/margins": 0.284994512796402,
"rewards/rejected": -0.42992934584617615,
"step": 1790
},
{
"epoch": 0.48,
"learning_rate": 3.1048047389991693e-06,
"logits/chosen": -1.3464422225952148,
"logits/rejected": -0.17827872931957245,
"logps/chosen": -613.8772583007812,
"logps/rejected": -1675.2249755859375,
"loss": 0.1109,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.057198069989681244,
"rewards/margins": 0.3218960464000702,
"rewards/rejected": -0.3790941536426544,
"step": 1800
},
{
"epoch": 0.48,
"learning_rate": 3.082199056232015e-06,
"logits/chosen": -1.4088603258132935,
"logits/rejected": -1.116137981414795,
"logps/chosen": -599.2484130859375,
"logps/rejected": -1451.5693359375,
"loss": 0.2227,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.0676434338092804,
"rewards/margins": 0.21012239158153534,
"rewards/rejected": -0.27776581048965454,
"step": 1810
},
{
"epoch": 0.49,
"learning_rate": 3.059542928183079e-06,
"logits/chosen": -1.0800918340682983,
"logits/rejected": -0.1234823614358902,
"logps/chosen": -755.9251708984375,
"logps/rejected": -1821.8336181640625,
"loss": 0.1684,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.08166182041168213,
"rewards/margins": 0.3705812394618988,
"rewards/rejected": -0.45224303007125854,
"step": 1820
},
{
"epoch": 0.49,
"learning_rate": 3.0368383179176584e-06,
"logits/chosen": -1.2700388431549072,
"logits/rejected": -0.5240074396133423,
"logps/chosen": -747.2010498046875,
"logps/rejected": -1727.675537109375,
"loss": 0.1762,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.0997215062379837,
"rewards/margins": 0.32902830839157104,
"rewards/rejected": -0.4287497401237488,
"step": 1830
},
{
"epoch": 0.49,
"learning_rate": 3.0140871927018466e-06,
"logits/chosen": -1.4908950328826904,
"logits/rejected": -0.6521574854850769,
"logps/chosen": -840.5997924804688,
"logps/rejected": -1888.5599365234375,
"loss": 0.1566,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1681024730205536,
"rewards/margins": 0.35152748227119446,
"rewards/rejected": -0.519629955291748,
"step": 1840
},
{
"epoch": 0.49,
"learning_rate": 2.9912915238320755e-06,
"logits/chosen": -1.2768045663833618,
"logits/rejected": -0.5125023126602173,
"logps/chosen": -649.5341796875,
"logps/rejected": -1694.0074462890625,
"loss": 0.1611,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.10545836389064789,
"rewards/margins": 0.3052206039428711,
"rewards/rejected": -0.41067901253700256,
"step": 1850
},
{
"epoch": 0.5,
"learning_rate": 2.9684532864643123e-06,
"logits/chosen": -1.6448793411254883,
"logits/rejected": -0.8161486387252808,
"logps/chosen": -683.2916870117188,
"logps/rejected": -1537.4920654296875,
"loss": 0.1733,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1456030309200287,
"rewards/margins": 0.2657826840877533,
"rewards/rejected": -0.411385715007782,
"step": 1860
},
{
"epoch": 0.5,
"learning_rate": 2.945574459442917e-06,
"logits/chosen": -1.5162016153335571,
"logits/rejected": -0.9785049557685852,
"logps/chosen": -733.1231079101562,
"logps/rejected": -1667.495849609375,
"loss": 0.1597,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.15169629454612732,
"rewards/margins": 0.29900026321411133,
"rewards/rejected": -0.45069652795791626,
"step": 1870
},
{
"epoch": 0.5,
"learning_rate": 2.922657025129185e-06,
"logits/chosen": -1.4472191333770752,
"logits/rejected": -0.4651806354522705,
"logps/chosen": -756.4955444335938,
"logps/rejected": -1759.114501953125,
"loss": 0.2495,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15917366743087769,
"rewards/margins": 0.2930217683315277,
"rewards/rejected": -0.4521954655647278,
"step": 1880
},
{
"epoch": 0.5,
"learning_rate": 2.8997029692295875e-06,
"logits/chosen": -1.5512139797210693,
"logits/rejected": -0.7475739121437073,
"logps/chosen": -771.3937377929688,
"logps/rejected": -1779.020751953125,
"loss": 0.1492,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.11777035892009735,
"rewards/margins": 0.30730709433555603,
"rewards/rejected": -0.4250774383544922,
"step": 1890
},
{
"epoch": 0.51,
"learning_rate": 2.876714280623708e-06,
"logits/chosen": -1.3046129941940308,
"logits/rejected": -0.2948853671550751,
"logps/chosen": -736.4674682617188,
"logps/rejected": -1671.8529052734375,
"loss": 0.1435,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09072361886501312,
"rewards/margins": 0.3495047688484192,
"rewards/rejected": -0.4402283728122711,
"step": 1900
},
{
"epoch": 0.51,
"learning_rate": 2.8536929511919227e-06,
"logits/chosen": -1.4932053089141846,
"logits/rejected": 0.03270454332232475,
"logps/chosen": -672.2532958984375,
"logps/rejected": -1599.681640625,
"loss": 0.1603,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09094641357660294,
"rewards/margins": 0.29966163635253906,
"rewards/rejected": -0.3906080424785614,
"step": 1910
},
{
"epoch": 0.51,
"learning_rate": 2.8306409756428067e-06,
"logits/chosen": -1.2685102224349976,
"logits/rejected": -0.457380473613739,
"logps/chosen": -708.8721923828125,
"logps/rejected": -1771.715087890625,
"loss": 0.143,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13361448049545288,
"rewards/margins": 0.3276425004005432,
"rewards/rejected": -0.46125689148902893,
"step": 1920
},
{
"epoch": 0.51,
"learning_rate": 2.807560351340302e-06,
"logits/chosen": -1.3997093439102173,
"logits/rejected": -0.37530818581581116,
"logps/chosen": -675.5166015625,
"logps/rejected": -1654.8193359375,
"loss": 0.1566,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.11855722963809967,
"rewards/margins": 0.36131593585014343,
"rewards/rejected": -0.4798731207847595,
"step": 1930
},
{
"epoch": 0.52,
"learning_rate": 2.7844530781306544e-06,
"logits/chosen": -1.329444408416748,
"logits/rejected": -0.4007740616798401,
"logps/chosen": -830.9290161132812,
"logps/rejected": -2067.2666015625,
"loss": 0.1597,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.18545034527778625,
"rewards/margins": 0.4029621481895447,
"rewards/rejected": -0.5884124636650085,
"step": 1940
},
{
"epoch": 0.52,
"learning_rate": 2.761321158169134e-06,
"logits/chosen": -1.5382049083709717,
"logits/rejected": -0.7047163844108582,
"logps/chosen": -823.2682495117188,
"logps/rejected": -1897.6998291015625,
"loss": 0.1462,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1536359339952469,
"rewards/margins": 0.3253306448459625,
"rewards/rejected": -0.4789665639400482,
"step": 1950
},
{
"epoch": 0.52,
"learning_rate": 2.738166595746554e-06,
"logits/chosen": -1.612831711769104,
"logits/rejected": -0.2793089747428894,
"logps/chosen": -831.91357421875,
"logps/rejected": -1897.527099609375,
"loss": 0.1602,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -0.15530729293823242,
"rewards/margins": 0.3191708028316498,
"rewards/rejected": -0.4744780957698822,
"step": 1960
},
{
"epoch": 0.53,
"learning_rate": 2.7149913971156105e-06,
"logits/chosen": -1.5521538257598877,
"logits/rejected": -0.6847071051597595,
"logps/chosen": -752.0145263671875,
"logps/rejected": -1666.796630859375,
"loss": 0.1651,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.14488133788108826,
"rewards/margins": 0.31412333250045776,
"rewards/rejected": -0.4590047001838684,
"step": 1970
},
{
"epoch": 0.53,
"learning_rate": 2.6917975703170466e-06,
"logits/chosen": -1.1928004026412964,
"logits/rejected": -0.8584890365600586,
"logps/chosen": -881.7937622070312,
"logps/rejected": -1752.3238525390625,
"loss": 0.1488,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.21142208576202393,
"rewards/margins": 0.31265494227409363,
"rewards/rejected": -0.5240770578384399,
"step": 1980
},
{
"epoch": 0.53,
"learning_rate": 2.668587125005663e-06,
"logits/chosen": -1.4918617010116577,
"logits/rejected": -0.6560184359550476,
"logps/chosen": -728.0860595703125,
"logps/rejected": -1561.5875244140625,
"loss": 0.1206,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.13730724155902863,
"rewards/margins": 0.3027104437351227,
"rewards/rejected": -0.4400176405906677,
"step": 1990
},
{
"epoch": 0.53,
"learning_rate": 2.6453620722761897e-06,
"logits/chosen": -1.5282548666000366,
"logits/rejected": -0.7147419452667236,
"logps/chosen": -817.76416015625,
"logps/rejected": -1859.003173828125,
"loss": 0.1314,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.21234384179115295,
"rewards/margins": 0.3443707823753357,
"rewards/rejected": -0.556714653968811,
"step": 2000
},
{
"epoch": 0.54,
"learning_rate": 2.6221244244890336e-06,
"logits/chosen": -1.3335682153701782,
"logits/rejected": -0.07726557552814484,
"logps/chosen": -991.7344970703125,
"logps/rejected": -1859.514892578125,
"loss": 0.1715,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.26506227254867554,
"rewards/margins": 0.2776499390602112,
"rewards/rejected": -0.5427122116088867,
"step": 2010
},
{
"epoch": 0.54,
"learning_rate": 2.5988761950959133e-06,
"logits/chosen": -1.4011818170547485,
"logits/rejected": -0.5180048942565918,
"logps/chosen": -809.6094360351562,
"logps/rejected": -1956.6029052734375,
"loss": 0.158,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1943681836128235,
"rewards/margins": 0.3915101885795593,
"rewards/rejected": -0.585878312587738,
"step": 2020
},
{
"epoch": 0.54,
"learning_rate": 2.575619398465402e-06,
"logits/chosen": -1.4724808931350708,
"logits/rejected": -0.7128661274909973,
"logps/chosen": -634.3349609375,
"logps/rejected": -1436.007568359375,
"loss": 0.2107,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.13114015758037567,
"rewards/margins": 0.22343416512012482,
"rewards/rejected": -0.3545742630958557,
"step": 2030
},
{
"epoch": 0.54,
"learning_rate": 2.5523560497083927e-06,
"logits/chosen": -1.1302558183670044,
"logits/rejected": -0.5582550764083862,
"logps/chosen": -715.1376953125,
"logps/rejected": -1611.423828125,
"loss": 0.1614,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11572917550802231,
"rewards/margins": 0.2607465088367462,
"rewards/rejected": -0.3764756917953491,
"step": 2040
},
{
"epoch": 0.55,
"learning_rate": 2.5290881645034932e-06,
"logits/chosen": -1.6198726892471313,
"logits/rejected": -0.6000566482543945,
"logps/chosen": -762.2321166992188,
"logps/rejected": -1829.3511962890625,
"loss": 0.1657,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.16410943865776062,
"rewards/margins": 0.37973588705062866,
"rewards/rejected": -0.5438452959060669,
"step": 2050
},
{
"epoch": 0.55,
"learning_rate": 2.5058177589223766e-06,
"logits/chosen": -1.335745930671692,
"logits/rejected": -0.5967541933059692,
"logps/chosen": -677.7882080078125,
"logps/rejected": -1777.054931640625,
"loss": 0.1599,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.12234095484018326,
"rewards/margins": 0.3689095675945282,
"rewards/rejected": -0.49125057458877563,
"step": 2060
},
{
"epoch": 0.55,
"learning_rate": 2.482546849255096e-06,
"logits/chosen": -1.2376234531402588,
"logits/rejected": -0.5898858308792114,
"logps/chosen": -797.4426879882812,
"logps/rejected": -1635.3062744140625,
"loss": 0.137,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15089258551597595,
"rewards/margins": 0.2774205803871155,
"rewards/rejected": -0.4283131957054138,
"step": 2070
},
{
"epoch": 0.55,
"learning_rate": 2.4592774518353858e-06,
"logits/chosen": -1.448229193687439,
"logits/rejected": -0.610197901725769,
"logps/chosen": -747.7297973632812,
"logps/rejected": -1767.844970703125,
"loss": 0.1608,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16316349804401398,
"rewards/margins": 0.3244298994541168,
"rewards/rejected": -0.4875933527946472,
"step": 2080
},
{
"epoch": 0.56,
"learning_rate": 2.436011582865945e-06,
"logits/chosen": -1.4960334300994873,
"logits/rejected": -0.6341277956962585,
"logps/chosen": -892.7718505859375,
"logps/rejected": -1960.636474609375,
"loss": 0.1577,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.21525397896766663,
"rewards/margins": 0.3495286703109741,
"rewards/rejected": -0.5647826790809631,
"step": 2090
},
{
"epoch": 0.56,
"learning_rate": 2.4127512582437486e-06,
"logits/chosen": -1.0745445489883423,
"logits/rejected": 0.3137953281402588,
"logps/chosen": -765.67919921875,
"logps/rejected": -1701.5640869140625,
"loss": 0.1943,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.164879709482193,
"rewards/margins": 0.2781751751899719,
"rewards/rejected": -0.4430549144744873,
"step": 2100
},
{
"epoch": 0.56,
"learning_rate": 2.3894984933853734e-06,
"logits/chosen": -1.4759852886199951,
"logits/rejected": -0.5236972570419312,
"logps/chosen": -838.9854736328125,
"logps/rejected": -1744.4114990234375,
"loss": 0.1704,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.17717286944389343,
"rewards/margins": 0.257200688123703,
"rewards/rejected": -0.43437355756759644,
"step": 2110
},
{
"epoch": 0.57,
"learning_rate": 2.366255303052377e-06,
"logits/chosen": -1.508411169052124,
"logits/rejected": -0.39122310280799866,
"logps/chosen": -843.5926513671875,
"logps/rejected": -1831.277099609375,
"loss": 0.1721,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.14325647056102753,
"rewards/margins": 0.292641818523407,
"rewards/rejected": -0.4358982443809509,
"step": 2120
},
{
"epoch": 0.57,
"learning_rate": 2.3430237011767166e-06,
"logits/chosen": -1.5056064128875732,
"logits/rejected": 0.36290010809898376,
"logps/chosen": -771.4110717773438,
"logps/rejected": -1977.0465087890625,
"loss": 0.1063,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.13286644220352173,
"rewards/margins": 0.41198697686195374,
"rewards/rejected": -0.5448533892631531,
"step": 2130
},
{
"epoch": 0.57,
"learning_rate": 2.319805700686257e-06,
"logits/chosen": -1.1897435188293457,
"logits/rejected": -0.6799365282058716,
"logps/chosen": -774.5003051757812,
"logps/rejected": -1752.76953125,
"loss": 0.1455,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.10589227825403214,
"rewards/margins": 0.3271291255950928,
"rewards/rejected": -0.4330214560031891,
"step": 2140
},
{
"epoch": 0.57,
"learning_rate": 2.296603313330355e-06,
"logits/chosen": -1.5946743488311768,
"logits/rejected": -0.24514515697956085,
"logps/chosen": -728.986328125,
"logps/rejected": -1501.432373046875,
"loss": 0.1661,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.12233342975378036,
"rewards/margins": 0.2589247226715088,
"rewards/rejected": -0.38125813007354736,
"step": 2150
},
{
"epoch": 0.58,
"learning_rate": 2.2734185495055503e-06,
"logits/chosen": -1.1782925128936768,
"logits/rejected": -0.13661935925483704,
"logps/chosen": -568.3839721679688,
"logps/rejected": -1682.452392578125,
"loss": 0.1338,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12734124064445496,
"rewards/margins": 0.3284556567668915,
"rewards/rejected": -0.4557969570159912,
"step": 2160
},
{
"epoch": 0.58,
"learning_rate": 2.250253418081373e-06,
"logits/chosen": -1.438287615776062,
"logits/rejected": 0.09851478040218353,
"logps/chosen": -888.4573974609375,
"logps/rejected": -1791.545654296875,
"loss": 0.2007,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.20431411266326904,
"rewards/margins": 0.2950129508972168,
"rewards/rejected": -0.49932703375816345,
"step": 2170
},
{
"epoch": 0.58,
"learning_rate": 2.22710992622628e-06,
"logits/chosen": -1.512485384941101,
"logits/rejected": 0.1614830195903778,
"logps/chosen": -803.5087890625,
"logps/rejected": -1817.915771484375,
"loss": 0.1446,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13419024646282196,
"rewards/margins": 0.4049316346645355,
"rewards/rejected": -0.5391219258308411,
"step": 2180
},
{
"epoch": 0.58,
"learning_rate": 2.2039900792337477e-06,
"logits/chosen": -1.3607614040374756,
"logits/rejected": -0.42570924758911133,
"logps/chosen": -765.4649658203125,
"logps/rejected": -1829.791259765625,
"loss": 0.1517,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1280583143234253,
"rewards/margins": 0.36656203866004944,
"rewards/rejected": -0.4946204125881195,
"step": 2190
},
{
"epoch": 0.59,
"learning_rate": 2.1808958803485134e-06,
"logits/chosen": -1.5544694662094116,
"logits/rejected": -0.8007132411003113,
"logps/chosen": -521.413818359375,
"logps/rejected": -1577.453857421875,
"loss": 0.1108,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.07102310657501221,
"rewards/margins": 0.3457227647304535,
"rewards/rejected": -0.4167459011077881,
"step": 2200
},
{
"epoch": 0.59,
"learning_rate": 2.157829330593008e-06,
"logits/chosen": -1.4860761165618896,
"logits/rejected": 0.10385887324810028,
"logps/chosen": -840.1112060546875,
"logps/rejected": -1826.4033203125,
"loss": 0.1199,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.17368373274803162,
"rewards/margins": 0.36537981033325195,
"rewards/rejected": -0.5390635132789612,
"step": 2210
},
{
"epoch": 0.59,
"learning_rate": 2.134792428593971e-06,
"logits/chosen": -1.4830608367919922,
"logits/rejected": -0.2886629104614258,
"logps/chosen": -688.9013671875,
"logps/rejected": -1467.2042236328125,
"loss": 0.196,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.12851116061210632,
"rewards/margins": 0.2524433434009552,
"rewards/rejected": -0.38095444440841675,
"step": 2220
},
{
"epoch": 0.59,
"learning_rate": 2.1117871704092818e-06,
"logits/chosen": -1.511211633682251,
"logits/rejected": -0.5041274428367615,
"logps/chosen": -755.0718994140625,
"logps/rejected": -1825.1058349609375,
"loss": 0.1207,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.10548318922519684,
"rewards/margins": 0.37242189049720764,
"rewards/rejected": -0.4779050946235657,
"step": 2230
},
{
"epoch": 0.6,
"learning_rate": 2.0888155493550027e-06,
"logits/chosen": -1.5032051801681519,
"logits/rejected": -0.2272813320159912,
"logps/chosen": -617.5220947265625,
"logps/rejected": -1630.780517578125,
"loss": 0.1807,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.0706586018204689,
"rewards/margins": 0.3596861958503723,
"rewards/rejected": -0.430344820022583,
"step": 2240
},
{
"epoch": 0.6,
"learning_rate": 2.0658795558326745e-06,
"logits/chosen": -1.4327954053878784,
"logits/rejected": -0.20998969674110413,
"logps/chosen": -752.0433349609375,
"logps/rejected": -1787.226806640625,
"loss": 0.1848,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07883577048778534,
"rewards/margins": 0.3491933047771454,
"rewards/rejected": -0.4280291199684143,
"step": 2250
},
{
"epoch": 0.6,
"learning_rate": 2.0429811771568468e-06,
"logits/chosen": -1.4897148609161377,
"logits/rejected": -0.4423850178718567,
"logps/chosen": -781.8294067382812,
"logps/rejected": -1615.2808837890625,
"loss": 0.1639,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.09271921962499619,
"rewards/margins": 0.27344492077827454,
"rewards/rejected": -0.3661641478538513,
"step": 2260
},
{
"epoch": 0.61,
"learning_rate": 2.0201223973828917e-06,
"logits/chosen": -1.5945818424224854,
"logits/rejected": -0.5884144902229309,
"logps/chosen": -650.7221069335938,
"logps/rejected": -1742.1322021484375,
"loss": 0.1932,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07876542955636978,
"rewards/margins": 0.36041781306266785,
"rewards/rejected": -0.43918323516845703,
"step": 2270
},
{
"epoch": 0.61,
"learning_rate": 1.997305197135089e-06,
"logits/chosen": -1.4727747440338135,
"logits/rejected": -0.5500017404556274,
"logps/chosen": -810.6484985351562,
"logps/rejected": -1713.4674072265625,
"loss": 0.1612,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.0999779924750328,
"rewards/margins": 0.28804337978363037,
"rewards/rejected": -0.3880213797092438,
"step": 2280
},
{
"epoch": 0.61,
"learning_rate": 1.9745315534350157e-06,
"logits/chosen": -1.5009238719940186,
"logits/rejected": -0.7104870080947876,
"logps/chosen": -620.9692993164062,
"logps/rejected": -1662.8277587890625,
"loss": 0.128,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.0903405249118805,
"rewards/margins": 0.30934035778045654,
"rewards/rejected": -0.3996809124946594,
"step": 2290
},
{
"epoch": 0.61,
"learning_rate": 1.9518034395302413e-06,
"logits/chosen": -1.17448091506958,
"logits/rejected": -0.5054240822792053,
"logps/chosen": -598.98046875,
"logps/rejected": -1721.002197265625,
"loss": 0.168,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07766080647706985,
"rewards/margins": 0.37883394956588745,
"rewards/rejected": -0.4564947485923767,
"step": 2300
},
{
"epoch": 0.62,
"learning_rate": 1.9291228247233607e-06,
"logits/chosen": -1.3139159679412842,
"logits/rejected": -0.6714901924133301,
"logps/chosen": -843.5751953125,
"logps/rejected": -1910.8170166015625,
"loss": 0.1071,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16244825720787048,
"rewards/margins": 0.3335781693458557,
"rewards/rejected": -0.4960264265537262,
"step": 2310
},
{
"epoch": 0.62,
"learning_rate": 1.9064916742013515e-06,
"logits/chosen": -1.3740966320037842,
"logits/rejected": -0.3796101212501526,
"logps/chosen": -911.6195068359375,
"logps/rejected": -1917.3980712890625,
"loss": 0.1879,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.19778266549110413,
"rewards/margins": 0.35178643465042114,
"rewards/rejected": -0.5495691895484924,
"step": 2320
},
{
"epoch": 0.62,
"learning_rate": 1.883911948865306e-06,
"logits/chosen": -1.4115228652954102,
"logits/rejected": -0.3268095552921295,
"logps/chosen": -669.3294677734375,
"logps/rejected": -1735.1517333984375,
"loss": 0.1672,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11265420913696289,
"rewards/margins": 0.36892735958099365,
"rewards/rejected": -0.48158153891563416,
"step": 2330
},
{
"epoch": 0.62,
"learning_rate": 1.8613856051605242e-06,
"logits/chosen": -1.6518300771713257,
"logits/rejected": -0.6226130127906799,
"logps/chosen": -585.2733154296875,
"logps/rejected": -1588.8135986328125,
"loss": 0.1303,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.09527496993541718,
"rewards/margins": 0.36038875579833984,
"rewards/rejected": -0.4556637406349182,
"step": 2340
},
{
"epoch": 0.63,
"learning_rate": 1.8389145949069953e-06,
"logits/chosen": -1.3490536212921143,
"logits/rejected": -0.4021245539188385,
"logps/chosen": -672.3787231445312,
"logps/rejected": -1490.810791015625,
"loss": 0.2294,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.09898020327091217,
"rewards/margins": 0.24909377098083496,
"rewards/rejected": -0.34807395935058594,
"step": 2350
},
{
"epoch": 0.63,
"learning_rate": 1.816500865130279e-06,
"logits/chosen": -1.430107831954956,
"logits/rejected": -0.26797226071357727,
"logps/chosen": -658.622314453125,
"logps/rejected": -1494.597900390625,
"loss": 0.1936,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.08045925199985504,
"rewards/margins": 0.2406737506389618,
"rewards/rejected": -0.321133017539978,
"step": 2360
},
{
"epoch": 0.63,
"learning_rate": 1.7941463578928088e-06,
"logits/chosen": -1.6190364360809326,
"logits/rejected": -0.5775830149650574,
"logps/chosen": -652.179443359375,
"logps/rejected": -1785.9398193359375,
"loss": 0.1489,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.0696047991514206,
"rewards/margins": 0.32390493154525757,
"rewards/rejected": -0.39350971579551697,
"step": 2370
},
{
"epoch": 0.63,
"learning_rate": 1.7718530101256115e-06,
"logits/chosen": -1.432355284690857,
"logits/rejected": -0.7160090208053589,
"logps/chosen": -757.6654663085938,
"logps/rejected": -1582.92333984375,
"loss": 0.2207,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.10973949730396271,
"rewards/margins": 0.2269624024629593,
"rewards/rejected": -0.3367019295692444,
"step": 2380
},
{
"epoch": 0.64,
"learning_rate": 1.7496227534604859e-06,
"logits/chosen": -1.6282392740249634,
"logits/rejected": -0.9807308316230774,
"logps/chosen": -681.4638671875,
"logps/rejected": -1636.9273681640625,
"loss": 0.1722,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.10274797677993774,
"rewards/margins": 0.3626464009284973,
"rewards/rejected": -0.46539440751075745,
"step": 2390
},
{
"epoch": 0.64,
"learning_rate": 1.7274575140626318e-06,
"logits/chosen": -1.4671717882156372,
"logits/rejected": -0.4975617825984955,
"logps/chosen": -698.2318725585938,
"logps/rejected": -1724.447509765625,
"loss": 0.1602,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11632432043552399,
"rewards/margins": 0.3676297664642334,
"rewards/rejected": -0.4839541018009186,
"step": 2400
},
{
"epoch": 0.64,
"learning_rate": 1.7053592124637557e-06,
"logits/chosen": -1.5764734745025635,
"logits/rejected": -0.617272675037384,
"logps/chosen": -761.3939819335938,
"logps/rejected": -1650.2503662109375,
"loss": 0.2339,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.15402886271476746,
"rewards/margins": 0.2574668228626251,
"rewards/rejected": -0.4114956855773926,
"step": 2410
},
{
"epoch": 0.65,
"learning_rate": 1.6833297633956647e-06,
"logits/chosen": -1.310483694076538,
"logits/rejected": 0.16814152896404266,
"logps/chosen": -694.213134765625,
"logps/rejected": -1697.9111328125,
"loss": 0.108,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07603000104427338,
"rewards/margins": 0.3201691210269928,
"rewards/rejected": -0.396199107170105,
"step": 2420
},
{
"epoch": 0.65,
"learning_rate": 1.661371075624363e-06,
"logits/chosen": -1.422377347946167,
"logits/rejected": -0.4117739796638489,
"logps/chosen": -695.9147338867188,
"logps/rejected": -1731.9359130859375,
"loss": 0.1581,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.060897957533597946,
"rewards/margins": 0.35658949613571167,
"rewards/rejected": -0.4174874424934387,
"step": 2430
},
{
"epoch": 0.65,
"learning_rate": 1.6394850517846621e-06,
"logits/chosen": -1.6102850437164307,
"logits/rejected": -0.9350086450576782,
"logps/chosen": -763.3278198242188,
"logps/rejected": -1718.2047119140625,
"loss": 0.1575,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1062902957201004,
"rewards/margins": 0.3449605107307434,
"rewards/rejected": -0.45125073194503784,
"step": 2440
},
{
"epoch": 0.65,
"learning_rate": 1.6176735882153284e-06,
"logits/chosen": -1.4336785078048706,
"logits/rejected": -0.12534983456134796,
"logps/chosen": -599.1646728515625,
"logps/rejected": -1452.4405517578125,
"loss": 0.1354,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.06159447878599167,
"rewards/margins": 0.3045565187931061,
"rewards/rejected": -0.36615103483200073,
"step": 2450
},
{
"epoch": 0.66,
"learning_rate": 1.5959385747947697e-06,
"logits/chosen": -1.4207260608673096,
"logits/rejected": -0.777606189250946,
"logps/chosen": -604.3650512695312,
"logps/rejected": -1650.2216796875,
"loss": 0.1083,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06717022508382797,
"rewards/margins": 0.36061280965805054,
"rewards/rejected": -0.4277830123901367,
"step": 2460
},
{
"epoch": 0.66,
"learning_rate": 1.5742818947772875e-06,
"logits/chosen": -1.5350821018218994,
"logits/rejected": 0.10819880664348602,
"logps/chosen": -758.2330322265625,
"logps/rejected": -1617.786376953125,
"loss": 0.1369,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.08228493481874466,
"rewards/margins": 0.2714352011680603,
"rewards/rejected": -0.35372015833854675,
"step": 2470
},
{
"epoch": 0.66,
"learning_rate": 1.552705424629898e-06,
"logits/chosen": -1.4091111421585083,
"logits/rejected": -0.6021371483802795,
"logps/chosen": -744.669189453125,
"logps/rejected": -1612.1370849609375,
"loss": 0.1492,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.08511866629123688,
"rewards/margins": 0.2584208548069,
"rewards/rejected": -0.3435395061969757,
"step": 2480
},
{
"epoch": 0.66,
"learning_rate": 1.5312110338697427e-06,
"logits/chosen": -1.3926702737808228,
"logits/rejected": -1.0168853998184204,
"logps/chosen": -612.21044921875,
"logps/rejected": -1413.4964599609375,
"loss": 0.1469,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.064825639128685,
"rewards/margins": 0.2570902705192566,
"rewards/rejected": -0.3219159245491028,
"step": 2490
},
{
"epoch": 0.67,
"learning_rate": 1.509800584902108e-06,
"logits/chosen": -1.2503700256347656,
"logits/rejected": -0.17403051257133484,
"logps/chosen": -702.2501831054688,
"logps/rejected": -1574.6552734375,
"loss": 0.1228,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.06957344710826874,
"rewards/margins": 0.29189419746398926,
"rewards/rejected": -0.3614676594734192,
"step": 2500
},
{
"epoch": 0.67,
"learning_rate": 1.4884759328590476e-06,
"logits/chosen": -1.7622737884521484,
"logits/rejected": -0.49598488211631775,
"logps/chosen": -769.4957885742188,
"logps/rejected": -1723.9212646484375,
"loss": 0.1593,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.12480386346578598,
"rewards/margins": 0.3833233416080475,
"rewards/rejected": -0.5081272125244141,
"step": 2510
},
{
"epoch": 0.67,
"learning_rate": 1.467238925438646e-06,
"logits/chosen": -1.5906703472137451,
"logits/rejected": 0.46796292066574097,
"logps/chosen": -865.5426025390625,
"logps/rejected": -1904.509765625,
"loss": 0.1402,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1493154764175415,
"rewards/margins": 0.37586697936058044,
"rewards/rejected": -0.5251824259757996,
"step": 2520
},
{
"epoch": 0.67,
"learning_rate": 1.446091402744923e-06,
"logits/chosen": -1.2282660007476807,
"logits/rejected": -0.4859045147895813,
"logps/chosen": -703.8905639648438,
"logps/rejected": -1641.694580078125,
"loss": 0.147,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.10744025558233261,
"rewards/margins": 0.3389386534690857,
"rewards/rejected": -0.4463788866996765,
"step": 2530
},
{
"epoch": 0.68,
"learning_rate": 1.4250351971283937e-06,
"logits/chosen": -1.564168930053711,
"logits/rejected": 0.4942803382873535,
"logps/chosen": -766.39599609375,
"logps/rejected": -1616.6973876953125,
"loss": 0.1396,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.11042846739292145,
"rewards/margins": 0.29587405920028687,
"rewards/rejected": -0.4063025414943695,
"step": 2540
},
{
"epoch": 0.68,
"learning_rate": 1.4040721330273063e-06,
"logits/chosen": -1.4303256273269653,
"logits/rejected": 0.4619303345680237,
"logps/chosen": -751.2760009765625,
"logps/rejected": -1650.994873046875,
"loss": 0.1959,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11452531814575195,
"rewards/margins": 0.2743874192237854,
"rewards/rejected": -0.38891273736953735,
"step": 2550
},
{
"epoch": 0.68,
"learning_rate": 1.3832040268095589e-06,
"logits/chosen": -1.2525991201400757,
"logits/rejected": -0.3175821006298065,
"logps/chosen": -702.5205688476562,
"logps/rejected": -1794.9117431640625,
"loss": 0.1277,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.09264467656612396,
"rewards/margins": 0.3793013393878937,
"rewards/rejected": -0.47194600105285645,
"step": 2560
},
{
"epoch": 0.69,
"learning_rate": 1.362432686615316e-06,
"logits/chosen": -1.4419841766357422,
"logits/rejected": -0.390924870967865,
"logps/chosen": -554.8194580078125,
"logps/rejected": -1714.5797119140625,
"loss": 0.1609,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.052808977663517,
"rewards/margins": 0.37850111722946167,
"rewards/rejected": -0.43131011724472046,
"step": 2570
},
{
"epoch": 0.69,
"learning_rate": 1.3417599122003464e-06,
"logits/chosen": -1.4737399816513062,
"logits/rejected": -0.35947731137275696,
"logps/chosen": -654.5718994140625,
"logps/rejected": -1695.5394287109375,
"loss": 0.1034,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.082417331635952,
"rewards/margins": 0.3580096364021301,
"rewards/rejected": -0.44042691588401794,
"step": 2580
},
{
"epoch": 0.69,
"learning_rate": 1.3211874947800747e-06,
"logits/chosen": -1.5177228450775146,
"logits/rejected": -0.705475926399231,
"logps/chosen": -662.635986328125,
"logps/rejected": -1590.105224609375,
"loss": 0.1596,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.08941694349050522,
"rewards/margins": 0.25798147916793823,
"rewards/rejected": -0.34739845991134644,
"step": 2590
},
{
"epoch": 0.69,
"learning_rate": 1.3007172168743854e-06,
"logits/chosen": -1.3950124979019165,
"logits/rejected": -0.012356835417449474,
"logps/chosen": -656.9428100585938,
"logps/rejected": -1578.4957275390625,
"loss": 0.1501,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07990637421607971,
"rewards/margins": 0.3191567063331604,
"rewards/rejected": -0.3990631103515625,
"step": 2600
},
{
"epoch": 0.7,
"learning_rate": 1.280350852153168e-06,
"logits/chosen": -1.3449013233184814,
"logits/rejected": -0.16039177775382996,
"logps/chosen": -759.0653076171875,
"logps/rejected": -1695.5814208984375,
"loss": 0.1461,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.15088282525539398,
"rewards/margins": 0.29462313652038574,
"rewards/rejected": -0.44550594687461853,
"step": 2610
},
{
"epoch": 0.7,
"learning_rate": 1.260090165282645e-06,
"logits/chosen": -1.3999792337417603,
"logits/rejected": 0.3149392008781433,
"logps/chosen": -689.9563598632812,
"logps/rejected": -1669.5087890625,
"loss": 0.1658,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1789911687374115,
"rewards/margins": 0.27231132984161377,
"rewards/rejected": -0.45130252838134766,
"step": 2620
},
{
"epoch": 0.7,
"learning_rate": 1.2399369117724582e-06,
"logits/chosen": -1.4091784954071045,
"logits/rejected": -0.4129961431026459,
"logps/chosen": -786.0768432617188,
"logps/rejected": -1682.849853515625,
"loss": 0.1508,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.11276306957006454,
"rewards/margins": 0.3028048872947693,
"rewards/rejected": -0.41556796431541443,
"step": 2630
},
{
"epoch": 0.7,
"learning_rate": 1.2198928378235717e-06,
"logits/chosen": -1.5089385509490967,
"logits/rejected": 0.6765385866165161,
"logps/chosen": -682.0382080078125,
"logps/rejected": -1677.535888671875,
"loss": 0.1625,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07076996564865112,
"rewards/margins": 0.3339851498603821,
"rewards/rejected": -0.404755175113678,
"step": 2640
},
{
"epoch": 0.71,
"learning_rate": 1.1999596801769617e-06,
"logits/chosen": -1.625929594039917,
"logits/rejected": -0.4023094177246094,
"logps/chosen": -682.7630615234375,
"logps/rejected": -1596.6417236328125,
"loss": 0.1467,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.07302750647068024,
"rewards/margins": 0.2907186448574066,
"rewards/rejected": -0.36374616622924805,
"step": 2650
},
{
"epoch": 0.71,
"learning_rate": 1.1801391659631423e-06,
"logits/chosen": -1.5305818319320679,
"logits/rejected": 0.7981144785881042,
"logps/chosen": -645.7145385742188,
"logps/rejected": -1584.2552490234375,
"loss": 0.1174,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.05174557492136955,
"rewards/margins": 0.31238168478012085,
"rewards/rejected": -0.3641272485256195,
"step": 2660
},
{
"epoch": 0.71,
"learning_rate": 1.160433012552508e-06,
"logits/chosen": -1.474902629852295,
"logits/rejected": -0.8396472930908203,
"logps/chosen": -683.650146484375,
"logps/rejected": -1486.7801513671875,
"loss": 0.1884,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.06140195578336716,
"rewards/margins": 0.2520787715911865,
"rewards/rejected": -0.3134807050228119,
"step": 2670
},
{
"epoch": 0.71,
"learning_rate": 1.1408429274065418e-06,
"logits/chosen": -1.3169395923614502,
"logits/rejected": -0.6151926517486572,
"logps/chosen": -565.2825927734375,
"logps/rejected": -1585.185791015625,
"loss": 0.1651,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.06299415230751038,
"rewards/margins": 0.2877538800239563,
"rewards/rejected": -0.35074806213378906,
"step": 2680
},
{
"epoch": 0.72,
"learning_rate": 1.1213706079298566e-06,
"logits/chosen": -1.456714391708374,
"logits/rejected": -0.23200377821922302,
"logps/chosen": -555.6742553710938,
"logps/rejected": -1574.927490234375,
"loss": 0.1603,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.04447126016020775,
"rewards/margins": 0.33361220359802246,
"rewards/rejected": -0.3780834376811981,
"step": 2690
},
{
"epoch": 0.72,
"learning_rate": 1.1020177413231334e-06,
"logits/chosen": -1.4803617000579834,
"logits/rejected": -0.2165524661540985,
"logps/chosen": -688.8192749023438,
"logps/rejected": -1610.768310546875,
"loss": 0.1745,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.08206529915332794,
"rewards/margins": 0.2980625331401825,
"rewards/rejected": -0.38012784719467163,
"step": 2700
},
{
"epoch": 0.72,
"learning_rate": 1.0827860044369226e-06,
"logits/chosen": -1.5613023042678833,
"logits/rejected": -0.2233733832836151,
"logps/chosen": -747.8840942382812,
"logps/rejected": -1710.396240234375,
"loss": 0.1608,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1048145517706871,
"rewards/margins": 0.31736674904823303,
"rewards/rejected": -0.42218121886253357,
"step": 2710
},
{
"epoch": 0.73,
"learning_rate": 1.06367706362636e-06,
"logits/chosen": -1.39828360080719,
"logits/rejected": -0.848728358745575,
"logps/chosen": -723.9393920898438,
"logps/rejected": -1687.5172119140625,
"loss": 0.1753,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11115583032369614,
"rewards/margins": 0.2940528988838196,
"rewards/rejected": -0.4052087366580963,
"step": 2720
},
{
"epoch": 0.73,
"learning_rate": 1.0446925746067768e-06,
"logits/chosen": -1.3547604084014893,
"logits/rejected": -0.15690529346466064,
"logps/chosen": -790.1058959960938,
"logps/rejected": -1860.5738525390625,
"loss": 0.1238,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1280745416879654,
"rewards/margins": 0.36689525842666626,
"rewards/rejected": -0.49496984481811523,
"step": 2730
},
{
"epoch": 0.73,
"learning_rate": 1.0258341823102418e-06,
"logits/chosen": -1.540259599685669,
"logits/rejected": -0.36155903339385986,
"logps/chosen": -741.6209106445312,
"logps/rejected": -1832.474609375,
"loss": 0.1492,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13528260588645935,
"rewards/margins": 0.39978405833244324,
"rewards/rejected": -0.5350667238235474,
"step": 2740
},
{
"epoch": 0.73,
"learning_rate": 1.0071035207430352e-06,
"logits/chosen": -1.6379410028457642,
"logits/rejected": -0.6687275171279907,
"logps/chosen": -783.6793212890625,
"logps/rejected": -1837.771728515625,
"loss": 0.1575,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1449609398841858,
"rewards/margins": 0.35221394896507263,
"rewards/rejected": -0.4971748888492584,
"step": 2750
},
{
"epoch": 0.74,
"learning_rate": 9.88502212844063e-07,
"logits/chosen": -1.2955491542816162,
"logits/rejected": -0.5987057685852051,
"logps/chosen": -634.2613525390625,
"logps/rejected": -1641.3394775390625,
"loss": 0.1522,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11327040195465088,
"rewards/margins": 0.3325853943824768,
"rewards/rejected": -0.44585585594177246,
"step": 2760
},
{
"epoch": 0.74,
"learning_rate": 9.700318703442437e-07,
"logits/chosen": -1.466076135635376,
"logits/rejected": -0.3946886658668518,
"logps/chosen": -745.1195068359375,
"logps/rejected": -1823.8948974609375,
"loss": 0.1474,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.09033511579036713,
"rewards/margins": 0.41018062829971313,
"rewards/rejected": -0.5005157589912415,
"step": 2770
},
{
"epoch": 0.74,
"learning_rate": 9.516940936268504e-07,
"logits/chosen": -1.3681772947311401,
"logits/rejected": -0.35859739780426025,
"logps/chosen": -648.45458984375,
"logps/rejected": -1582.552978515625,
"loss": 0.1233,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.09594032913446426,
"rewards/margins": 0.2905716896057129,
"rewards/rejected": -0.38651204109191895,
"step": 2780
},
{
"epoch": 0.74,
"learning_rate": 9.334904715888496e-07,
"logits/chosen": -1.5419670343399048,
"logits/rejected": -0.12930986285209656,
"logps/chosen": -640.37646484375,
"logps/rejected": -1695.059814453125,
"loss": 0.1712,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07272680103778839,
"rewards/margins": 0.3492538332939148,
"rewards/rejected": -0.42198067903518677,
"step": 2790
},
{
"epoch": 0.75,
"learning_rate": 9.154225815032242e-07,
"logits/chosen": -1.4778330326080322,
"logits/rejected": -0.7425899505615234,
"logps/chosen": -604.2279052734375,
"logps/rejected": -1645.4664306640625,
"loss": 0.1527,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.060115180909633636,
"rewards/margins": 0.3599664568901062,
"rewards/rejected": -0.42008161544799805,
"step": 2800
},
{
"epoch": 0.75,
"learning_rate": 8.974919888823164e-07,
"logits/chosen": -1.4436981678009033,
"logits/rejected": -0.3787620961666107,
"logps/chosen": -735.7376708984375,
"logps/rejected": -1799.7640380859375,
"loss": 0.1705,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.08380059152841568,
"rewards/margins": 0.35864678025245667,
"rewards/rejected": -0.4424474239349365,
"step": 2810
},
{
"epoch": 0.75,
"learning_rate": 8.797002473421729e-07,
"logits/chosen": -1.4138429164886475,
"logits/rejected": -0.46681833267211914,
"logps/chosen": -517.6033935546875,
"logps/rejected": -1498.3123779296875,
"loss": 0.1961,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.05182959511876106,
"rewards/margins": 0.31122758984565735,
"rewards/rejected": -0.36305713653564453,
"step": 2820
},
{
"epoch": 0.75,
"learning_rate": 8.620488984679378e-07,
"logits/chosen": -1.698520302772522,
"logits/rejected": -0.209329292178154,
"logps/chosen": -629.7589721679688,
"logps/rejected": -1619.1944580078125,
"loss": 0.1404,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.04417989403009415,
"rewards/margins": 0.3446957468986511,
"rewards/rejected": -0.38887566328048706,
"step": 2830
},
{
"epoch": 0.76,
"learning_rate": 8.445394716802754e-07,
"logits/chosen": -1.4968388080596924,
"logits/rejected": -0.649762749671936,
"logps/chosen": -686.9845581054688,
"logps/rejected": -1680.4127197265625,
"loss": 0.1734,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.07763628661632538,
"rewards/margins": 0.340093731880188,
"rewards/rejected": -0.4177300035953522,
"step": 2840
},
{
"epoch": 0.76,
"learning_rate": 8.271734841028553e-07,
"logits/chosen": -1.3302339315414429,
"logits/rejected": -0.4346889555454254,
"logps/chosen": -617.5177612304688,
"logps/rejected": -1592.0660400390625,
"loss": 0.1696,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.06615222990512848,
"rewards/margins": 0.3168949782848358,
"rewards/rejected": -0.3830471932888031,
"step": 2850
},
{
"epoch": 0.76,
"learning_rate": 8.099524404308948e-07,
"logits/chosen": -1.4116344451904297,
"logits/rejected": -0.4046458601951599,
"logps/chosen": -718.5609130859375,
"logps/rejected": -1903.5677490234375,
"loss": 0.1009,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.0856105238199234,
"rewards/margins": 0.3954885005950928,
"rewards/rejected": -0.48109903931617737,
"step": 2860
},
{
"epoch": 0.77,
"learning_rate": 7.928778328007918e-07,
"logits/chosen": -1.6713218688964844,
"logits/rejected": -0.07077087461948395,
"logps/chosen": -666.365966796875,
"logps/rejected": -1721.5794677734375,
"loss": 0.1288,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.06950782239437103,
"rewards/margins": 0.33922192454338074,
"rewards/rejected": -0.4087297320365906,
"step": 2870
},
{
"epoch": 0.77,
"learning_rate": 7.759511406608255e-07,
"logits/chosen": -1.3845415115356445,
"logits/rejected": -0.046902846544981,
"logps/chosen": -711.4525146484375,
"logps/rejected": -1648.327392578125,
"loss": 0.1565,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.0852561965584755,
"rewards/margins": 0.27144601941108704,
"rewards/rejected": -0.35670217871665955,
"step": 2880
},
{
"epoch": 0.77,
"learning_rate": 7.591738306429769e-07,
"logits/chosen": -1.3664982318878174,
"logits/rejected": -0.31685924530029297,
"logps/chosen": -806.8741455078125,
"logps/rejected": -1734.4677734375,
"loss": 0.1123,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08330997824668884,
"rewards/margins": 0.3360075056552887,
"rewards/rejected": -0.41931748390197754,
"step": 2890
},
{
"epoch": 0.77,
"learning_rate": 7.425473564358457e-07,
"logits/chosen": -1.2000247240066528,
"logits/rejected": -0.19667713344097137,
"logps/chosen": -858.8897705078125,
"logps/rejected": -1800.777587890625,
"loss": 0.2683,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1292639523744583,
"rewards/margins": 0.33466023206710815,
"rewards/rejected": -0.46392422914505005,
"step": 2900
},
{
"epoch": 0.78,
"learning_rate": 7.260731586586983e-07,
"logits/chosen": -1.5677512884140015,
"logits/rejected": -0.5398620367050171,
"logps/chosen": -503.47052001953125,
"logps/rejected": -1553.2418212890625,
"loss": 0.1237,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.04408533498644829,
"rewards/margins": 0.34157511591911316,
"rewards/rejected": -0.38566046953201294,
"step": 2910
},
{
"epoch": 0.78,
"learning_rate": 7.097526647366379e-07,
"logits/chosen": -1.7256578207015991,
"logits/rejected": -0.6446534395217896,
"logps/chosen": -649.2868041992188,
"logps/rejected": -1570.2288818359375,
"loss": 0.1441,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.05645834654569626,
"rewards/margins": 0.304313600063324,
"rewards/rejected": -0.3607719838619232,
"step": 2920
},
{
"epoch": 0.78,
"learning_rate": 6.935872887769299e-07,
"logits/chosen": -1.5906898975372314,
"logits/rejected": -0.8133038282394409,
"logps/chosen": -777.7838134765625,
"logps/rejected": -1589.012451171875,
"loss": 0.2021,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09165602177381516,
"rewards/margins": 0.2887498140335083,
"rewards/rejected": -0.38040581345558167,
"step": 2930
},
{
"epoch": 0.78,
"learning_rate": 6.775784314464717e-07,
"logits/chosen": -1.3574354648590088,
"logits/rejected": -0.9483124613761902,
"logps/chosen": -623.078125,
"logps/rejected": -1650.8558349609375,
"loss": 0.164,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.06772687286138535,
"rewards/margins": 0.29907676577568054,
"rewards/rejected": -0.3668036460876465,
"step": 2940
},
{
"epoch": 0.79,
"learning_rate": 6.617274798504286e-07,
"logits/chosen": -1.6042282581329346,
"logits/rejected": -0.3267243504524231,
"logps/chosen": -772.8688354492188,
"logps/rejected": -1827.7327880859375,
"loss": 0.1381,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07094523310661316,
"rewards/margins": 0.36197200417518616,
"rewards/rejected": -0.43291720747947693,
"step": 2950
},
{
"epoch": 0.79,
"learning_rate": 6.460358074120518e-07,
"logits/chosen": -1.5008442401885986,
"logits/rejected": -0.9061881303787231,
"logps/chosen": -710.754638671875,
"logps/rejected": -1521.227783203125,
"loss": 0.1967,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.06366866081953049,
"rewards/margins": 0.2556148171424866,
"rewards/rejected": -0.31928351521492004,
"step": 2960
},
{
"epoch": 0.79,
"learning_rate": 6.305047737536707e-07,
"logits/chosen": -1.52692449092865,
"logits/rejected": -1.0452547073364258,
"logps/chosen": -648.1273803710938,
"logps/rejected": -1738.8756103515625,
"loss": 0.1429,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.0617615282535553,
"rewards/margins": 0.3802492022514343,
"rewards/rejected": -0.4420107305049896,
"step": 2970
},
{
"epoch": 0.79,
"learning_rate": 6.151357245788917e-07,
"logits/chosen": -1.479436993598938,
"logits/rejected": -0.5046942234039307,
"logps/chosen": -704.0694580078125,
"logps/rejected": -1901.0484619140625,
"loss": 0.134,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.08424389362335205,
"rewards/margins": 0.3907619118690491,
"rewards/rejected": -0.47500577569007874,
"step": 2980
},
{
"epoch": 0.8,
"learning_rate": 5.999299915559956e-07,
"logits/chosen": -1.3773359060287476,
"logits/rejected": -0.7278205156326294,
"logps/chosen": -646.4566650390625,
"logps/rejected": -1756.0074462890625,
"loss": 0.1404,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07592879235744476,
"rewards/margins": 0.3527293801307678,
"rewards/rejected": -0.4286581873893738,
"step": 2990
},
{
"epoch": 0.8,
"learning_rate": 5.848888922025553e-07,
"logits/chosen": -1.4301416873931885,
"logits/rejected": -0.6595948338508606,
"logps/chosen": -633.0948486328125,
"logps/rejected": -1438.056396484375,
"loss": 0.2039,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.0907018780708313,
"rewards/margins": 0.2594781816005707,
"rewards/rejected": -0.35018008947372437,
"step": 3000
},
{
"epoch": 0.8,
"learning_rate": 5.700137297712749e-07,
"logits/chosen": -1.5186036825180054,
"logits/rejected": -0.2126600295305252,
"logps/chosen": -719.3463134765625,
"logps/rejected": -1743.2152099609375,
"loss": 0.136,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.07820829004049301,
"rewards/margins": 0.35745203495025635,
"rewards/rejected": -0.43566030263900757,
"step": 3010
},
{
"epoch": 0.81,
"learning_rate": 5.553057931370729e-07,
"logits/chosen": -1.629778265953064,
"logits/rejected": -0.40432286262512207,
"logps/chosen": -629.0681762695312,
"logps/rejected": -1662.3658447265625,
"loss": 0.1704,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07306727021932602,
"rewards/margins": 0.336677610874176,
"rewards/rejected": -0.40974488854408264,
"step": 3020
},
{
"epoch": 0.81,
"learning_rate": 5.407663566854008e-07,
"logits/chosen": -1.5414226055145264,
"logits/rejected": -0.8694823980331421,
"logps/chosen": -783.411376953125,
"logps/rejected": -1741.0318603515625,
"loss": 0.1431,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.08548710495233536,
"rewards/margins": 0.32277002930641174,
"rewards/rejected": -0.4082570970058441,
"step": 3030
},
{
"epoch": 0.81,
"learning_rate": 5.263966802018275e-07,
"logits/chosen": -1.6246259212493896,
"logits/rejected": -0.6862791180610657,
"logps/chosen": -805.5133056640625,
"logps/rejected": -1652.5374755859375,
"loss": 0.1632,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.08222351223230362,
"rewards/margins": 0.30940836668014526,
"rewards/rejected": -0.3916319012641907,
"step": 3040
},
{
"epoch": 0.81,
"learning_rate": 5.121980087628802e-07,
"logits/chosen": -1.5992133617401123,
"logits/rejected": -0.5684345364570618,
"logps/chosen": -639.3607177734375,
"logps/rejected": -1503.9329833984375,
"loss": 0.1916,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.042655039578676224,
"rewards/margins": 0.3029300570487976,
"rewards/rejected": -0.34558507800102234,
"step": 3050
},
{
"epoch": 0.82,
"learning_rate": 4.981715726281666e-07,
"logits/chosen": -1.3770325183868408,
"logits/rejected": -0.07314357906579971,
"logps/chosen": -626.600341796875,
"logps/rejected": -1564.7174072265625,
"loss": 0.1957,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.061516791582107544,
"rewards/margins": 0.29265180230140686,
"rewards/rejected": -0.3541685938835144,
"step": 3060
},
{
"epoch": 0.82,
"learning_rate": 4.843185871337722e-07,
"logits/chosen": -1.302022933959961,
"logits/rejected": -0.5096440315246582,
"logps/chosen": -616.6043701171875,
"logps/rejected": -1739.4761962890625,
"loss": 0.1331,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.04673675075173378,
"rewards/margins": 0.33783185482025146,
"rewards/rejected": -0.38456863164901733,
"step": 3070
},
{
"epoch": 0.82,
"learning_rate": 4.706402525869633e-07,
"logits/chosen": -1.244085431098938,
"logits/rejected": -0.4011514186859131,
"logps/chosen": -558.00341796875,
"logps/rejected": -1766.441162109375,
"loss": 0.135,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.03591493144631386,
"rewards/margins": 0.3851375877857208,
"rewards/rejected": -0.4210525155067444,
"step": 3080
},
{
"epoch": 0.82,
"learning_rate": 4.5713775416217884e-07,
"logits/chosen": -1.5269744396209717,
"logits/rejected": -0.3547285795211792,
"logps/chosen": -603.1836547851562,
"logps/rejected": -1485.8277587890625,
"loss": 0.1863,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.050064198672771454,
"rewards/margins": 0.30074387788772583,
"rewards/rejected": -0.3508080542087555,
"step": 3090
},
{
"epoch": 0.83,
"learning_rate": 4.438122617983442e-07,
"logits/chosen": -1.3431215286254883,
"logits/rejected": -0.15092921257019043,
"logps/chosen": -600.654296875,
"logps/rejected": -1668.9749755859375,
"loss": 0.1123,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.019012073054909706,
"rewards/margins": 0.3798617422580719,
"rewards/rejected": -0.39887386560440063,
"step": 3100
},
{
"epoch": 0.83,
"learning_rate": 4.3066493009749853e-07,
"logits/chosen": -1.651582956314087,
"logits/rejected": -0.60547935962677,
"logps/chosen": -783.2088623046875,
"logps/rejected": -1752.2379150390625,
"loss": 0.1943,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07545865327119827,
"rewards/margins": 0.28511419892311096,
"rewards/rejected": -0.36057284474372864,
"step": 3110
},
{
"epoch": 0.83,
"learning_rate": 4.1769689822475147e-07,
"logits/chosen": -1.4048668146133423,
"logits/rejected": -0.61931312084198,
"logps/chosen": -734.5963745117188,
"logps/rejected": -1771.947509765625,
"loss": 0.1742,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06757534295320511,
"rewards/margins": 0.3220224380493164,
"rewards/rejected": -0.3895978033542633,
"step": 3120
},
{
"epoch": 0.83,
"learning_rate": 4.049092898095816e-07,
"logits/chosen": -1.2418572902679443,
"logits/rejected": -0.5778509378433228,
"logps/chosen": -649.421142578125,
"logps/rejected": -1639.222412109375,
"loss": 0.1488,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.06102794408798218,
"rewards/margins": 0.2672490179538727,
"rewards/rejected": -0.32827693223953247,
"step": 3130
},
{
"epoch": 0.84,
"learning_rate": 3.9230321284847856e-07,
"logits/chosen": -1.4714672565460205,
"logits/rejected": -0.19613921642303467,
"logps/chosen": -517.8187255859375,
"logps/rejected": -1359.996337890625,
"loss": 0.1738,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.015176740474998951,
"rewards/margins": 0.27705565094947815,
"rewards/rejected": -0.29223236441612244,
"step": 3140
},
{
"epoch": 0.84,
"learning_rate": 3.798797596089351e-07,
"logits/chosen": -1.4680472612380981,
"logits/rejected": -1.0179380178451538,
"logps/chosen": -649.7005615234375,
"logps/rejected": -1706.8095703125,
"loss": 0.1422,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.0649513378739357,
"rewards/margins": 0.37964990735054016,
"rewards/rejected": -0.44460123777389526,
"step": 3150
},
{
"epoch": 0.84,
"learning_rate": 3.6764000653481263e-07,
"logits/chosen": -1.530711054801941,
"logits/rejected": -0.5328912734985352,
"logps/chosen": -678.9803466796875,
"logps/rejected": -1562.3223876953125,
"loss": 0.1803,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07924239337444305,
"rewards/margins": 0.29753822088241577,
"rewards/rejected": -0.37678059935569763,
"step": 3160
},
{
"epoch": 0.85,
"learning_rate": 3.555850141530659e-07,
"logits/chosen": -1.573302149772644,
"logits/rejected": -0.6228225827217102,
"logps/chosen": -720.739990234375,
"logps/rejected": -1550.609619140625,
"loss": 0.1381,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.05328356474637985,
"rewards/margins": 0.3060484826564789,
"rewards/rejected": -0.35933202505111694,
"step": 3170
},
{
"epoch": 0.85,
"learning_rate": 3.4371582698185636e-07,
"logits/chosen": -1.4535822868347168,
"logits/rejected": -0.5261168479919434,
"logps/chosen": -570.4685668945312,
"logps/rejected": -1702.9281005859375,
"loss": 0.1551,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.05110562592744827,
"rewards/margins": 0.36743611097335815,
"rewards/rejected": -0.4185417592525482,
"step": 3180
},
{
"epoch": 0.85,
"learning_rate": 3.3203347344004737e-07,
"logits/chosen": -1.2148401737213135,
"logits/rejected": -0.12717768549919128,
"logps/chosen": -689.377197265625,
"logps/rejected": -1628.9788818359375,
"loss": 0.1693,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1245916485786438,
"rewards/margins": 0.2877344489097595,
"rewards/rejected": -0.41232603788375854,
"step": 3190
},
{
"epoch": 0.85,
"learning_rate": 3.2053896575809426e-07,
"logits/chosen": -1.3511433601379395,
"logits/rejected": -0.6393106579780579,
"logps/chosen": -607.8463745117188,
"logps/rejected": -1700.497802734375,
"loss": 0.1006,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.07283592969179153,
"rewards/margins": 0.3661297857761383,
"rewards/rejected": -0.43896573781967163,
"step": 3200
},
{
"epoch": 0.86,
"learning_rate": 3.092332998903416e-07,
"logits/chosen": -1.3332850933074951,
"logits/rejected": -0.6272661685943604,
"logps/chosen": -599.9690551757812,
"logps/rejected": -1547.851318359375,
"loss": 0.1517,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.061964135617017746,
"rewards/margins": 0.3017534613609314,
"rewards/rejected": -0.36371761560440063,
"step": 3210
},
{
"epoch": 0.86,
"learning_rate": 2.981174554287239e-07,
"logits/chosen": -1.3444700241088867,
"logits/rejected": -0.7092281579971313,
"logps/chosen": -702.991455078125,
"logps/rejected": -1928.198974609375,
"loss": 0.134,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.06982138007879257,
"rewards/margins": 0.3849565088748932,
"rewards/rejected": -0.45477789640426636,
"step": 3220
},
{
"epoch": 0.86,
"learning_rate": 2.871923955178918e-07,
"logits/chosen": -1.1920301914215088,
"logits/rejected": -0.5040346384048462,
"logps/chosen": -689.6361083984375,
"logps/rejected": -1791.2109375,
"loss": 0.1159,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08395170420408249,
"rewards/margins": 0.38336819410324097,
"rewards/rejected": -0.46731987595558167,
"step": 3230
},
{
"epoch": 0.86,
"learning_rate": 2.764590667717562e-07,
"logits/chosen": -1.4373012781143188,
"logits/rejected": -0.292427122592926,
"logps/chosen": -607.8994140625,
"logps/rejected": -1627.1224365234375,
"loss": 0.179,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.0747733861207962,
"rewards/margins": 0.31044524908065796,
"rewards/rejected": -0.38521862030029297,
"step": 3240
},
{
"epoch": 0.87,
"learning_rate": 2.6591839919146963e-07,
"logits/chosen": -1.4667880535125732,
"logits/rejected": 0.36761537194252014,
"logps/chosen": -681.8417358398438,
"logps/rejected": -1700.8343505859375,
"loss": 0.1297,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07614819705486298,
"rewards/margins": 0.3499642312526703,
"rewards/rejected": -0.42611247301101685,
"step": 3250
},
{
"epoch": 0.87,
"learning_rate": 2.555713060848433e-07,
"logits/chosen": -1.3263441324234009,
"logits/rejected": -0.007822990417480469,
"logps/chosen": -645.5946044921875,
"logps/rejected": -1641.087646484375,
"loss": 0.1576,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07578223943710327,
"rewards/margins": 0.32165446877479553,
"rewards/rejected": -0.3974367678165436,
"step": 3260
},
{
"epoch": 0.87,
"learning_rate": 2.454186839872158e-07,
"logits/chosen": -1.598440408706665,
"logits/rejected": -0.6168525815010071,
"logps/chosen": -796.4778442382812,
"logps/rejected": -1850.823974609375,
"loss": 0.1374,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.11286751180887222,
"rewards/margins": 0.3862745761871338,
"rewards/rejected": -0.4991420805454254,
"step": 3270
},
{
"epoch": 0.87,
"learning_rate": 2.3546141258376786e-07,
"logits/chosen": -1.3171125650405884,
"logits/rejected": -0.3912467360496521,
"logps/chosen": -731.4490966796875,
"logps/rejected": -1646.4290771484375,
"loss": 0.1371,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11279473453760147,
"rewards/margins": 0.3294528126716614,
"rewards/rejected": -0.4422474801540375,
"step": 3280
},
{
"epoch": 0.88,
"learning_rate": 2.257003546333042e-07,
"logits/chosen": -1.4132746458053589,
"logits/rejected": -0.49065130949020386,
"logps/chosen": -738.8904418945312,
"logps/rejected": -1917.914306640625,
"loss": 0.1661,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1128961443901062,
"rewards/margins": 0.3549092411994934,
"rewards/rejected": -0.46780532598495483,
"step": 3290
},
{
"epoch": 0.88,
"learning_rate": 2.1613635589349756e-07,
"logits/chosen": -1.6147540807724,
"logits/rejected": -0.7865034341812134,
"logps/chosen": -561.64306640625,
"logps/rejected": -1595.516357421875,
"loss": 0.1442,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.05281323194503784,
"rewards/margins": 0.32325616478919983,
"rewards/rejected": -0.3760693669319153,
"step": 3300
},
{
"epoch": 0.88,
"learning_rate": 2.0677024504760752e-07,
"logits/chosen": -1.467961311340332,
"logits/rejected": -0.4320656359195709,
"logps/chosen": -556.8984375,
"logps/rejected": -1729.6136474609375,
"loss": 0.1361,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.03827287629246712,
"rewards/margins": 0.37058061361312866,
"rewards/rejected": -0.4088534712791443,
"step": 3310
},
{
"epoch": 0.89,
"learning_rate": 1.9760283363267684e-07,
"logits/chosen": -1.6456931829452515,
"logits/rejected": -0.19730210304260254,
"logps/chosen": -729.8663940429688,
"logps/rejected": -1706.380615234375,
"loss": 0.1453,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.07686268538236618,
"rewards/margins": 0.34378868341445923,
"rewards/rejected": -0.4206513464450836,
"step": 3320
},
{
"epoch": 0.89,
"learning_rate": 1.8863491596921745e-07,
"logits/chosen": -1.275883436203003,
"logits/rejected": -0.10841438919305801,
"logps/chosen": -601.3389892578125,
"logps/rejected": -1545.714111328125,
"loss": 0.1123,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06958813965320587,
"rewards/margins": 0.32690125703811646,
"rewards/rejected": -0.3964894115924835,
"step": 3330
},
{
"epoch": 0.89,
"learning_rate": 1.798672690923828e-07,
"logits/chosen": -1.5245708227157593,
"logits/rejected": -0.7873867750167847,
"logps/chosen": -622.5413818359375,
"logps/rejected": -1778.4840087890625,
"loss": 0.0966,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.0721798688173294,
"rewards/margins": 0.36624765396118164,
"rewards/rejected": -0.43842750787734985,
"step": 3340
},
{
"epoch": 0.89,
"learning_rate": 1.713006526846439e-07,
"logits/chosen": -1.507137417793274,
"logits/rejected": -0.2042998969554901,
"logps/chosen": -680.6546630859375,
"logps/rejected": -1971.2806396484375,
"loss": 0.0794,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.06454543769359589,
"rewards/margins": 0.4576357901096344,
"rewards/rejected": -0.5221812129020691,
"step": 3350
},
{
"epoch": 0.9,
"learning_rate": 1.629358090099639e-07,
"logits/chosen": -1.551574468612671,
"logits/rejected": -0.49993395805358887,
"logps/chosen": -704.3600463867188,
"logps/rejected": -1691.518310546875,
"loss": 0.1888,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.08099500089883804,
"rewards/margins": 0.33148878812789917,
"rewards/rejected": -0.4124837815761566,
"step": 3360
},
{
"epoch": 0.9,
"learning_rate": 1.5477346284948292e-07,
"logits/chosen": -1.5370080471038818,
"logits/rejected": -0.8450511693954468,
"logps/chosen": -749.2708740234375,
"logps/rejected": -1634.89453125,
"loss": 0.1592,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09636793285608292,
"rewards/margins": 0.2845004200935364,
"rewards/rejected": -0.3808683753013611,
"step": 3370
},
{
"epoch": 0.9,
"learning_rate": 1.4681432143872133e-07,
"logits/chosen": -1.2466710805892944,
"logits/rejected": -0.42571744322776794,
"logps/chosen": -763.4384155273438,
"logps/rejected": -1733.5638427734375,
"loss": 0.177,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.09244507551193237,
"rewards/margins": 0.330522358417511,
"rewards/rejected": -0.42296743392944336,
"step": 3380
},
{
"epoch": 0.9,
"learning_rate": 1.3905907440629752e-07,
"logits/chosen": -1.5868967771530151,
"logits/rejected": -0.42538732290267944,
"logps/chosen": -704.2333374023438,
"logps/rejected": -1660.2496337890625,
"loss": 0.2384,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09102825075387955,
"rewards/margins": 0.27777737379074097,
"rewards/rejected": -0.3688056468963623,
"step": 3390
},
{
"epoch": 0.91,
"learning_rate": 1.31508393714177e-07,
"logits/chosen": -1.5377957820892334,
"logits/rejected": -0.9308739900588989,
"logps/chosen": -690.0589599609375,
"logps/rejected": -1625.2911376953125,
"loss": 0.1416,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.07044301927089691,
"rewards/margins": 0.33037734031677246,
"rewards/rejected": -0.4008203446865082,
"step": 3400
},
{
"epoch": 0.91,
"learning_rate": 1.241629335994471e-07,
"logits/chosen": -1.2723309993743896,
"logits/rejected": -0.2303469479084015,
"logps/chosen": -564.4718017578125,
"logps/rejected": -1448.8857421875,
"loss": 0.1991,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.06368915736675262,
"rewards/margins": 0.297661155462265,
"rewards/rejected": -0.36135026812553406,
"step": 3410
},
{
"epoch": 0.91,
"learning_rate": 1.1702333051763271e-07,
"logits/chosen": -1.5919114351272583,
"logits/rejected": -0.7112639546394348,
"logps/chosen": -713.6360473632812,
"logps/rejected": -1638.255126953125,
"loss": 0.139,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09833192080259323,
"rewards/margins": 0.317642480134964,
"rewards/rejected": -0.4159744381904602,
"step": 3420
},
{
"epoch": 0.91,
"learning_rate": 1.1009020308754587e-07,
"logits/chosen": -1.3826844692230225,
"logits/rejected": -0.3817708492279053,
"logps/chosen": -922.4486083984375,
"logps/rejected": -1803.8795166015625,
"loss": 0.2071,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1334574669599533,
"rewards/margins": 0.3121677041053772,
"rewards/rejected": -0.4456251263618469,
"step": 3430
},
{
"epoch": 0.92,
"learning_rate": 1.0336415203768962e-07,
"logits/chosen": -1.4255292415618896,
"logits/rejected": 0.35872507095336914,
"logps/chosen": -776.8897705078125,
"logps/rejected": -1673.2662353515625,
"loss": 0.1536,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.0705815777182579,
"rewards/margins": 0.3355250358581543,
"rewards/rejected": -0.4061066210269928,
"step": 3440
},
{
"epoch": 0.92,
"learning_rate": 9.684576015420277e-08,
"logits/chosen": -1.356412649154663,
"logits/rejected": -0.2489086389541626,
"logps/chosen": -556.0355224609375,
"logps/rejected": -1530.9036865234375,
"loss": 0.1951,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07743863016366959,
"rewards/margins": 0.2904726266860962,
"rewards/rejected": -0.36791130900382996,
"step": 3450
},
{
"epoch": 0.92,
"learning_rate": 9.053559223036746e-08,
"logits/chosen": -1.3430372476577759,
"logits/rejected": -0.30119800567626953,
"logps/chosen": -748.9754638671875,
"logps/rejected": -1545.953125,
"loss": 0.1491,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.07480500638484955,
"rewards/margins": 0.2932848632335663,
"rewards/rejected": -0.368089884519577,
"step": 3460
},
{
"epoch": 0.93,
"learning_rate": 8.44341950176683e-08,
"logits/chosen": -1.5457831621170044,
"logits/rejected": 0.028019333258271217,
"logps/chosen": -562.9949340820312,
"logps/rejected": -1467.437744140625,
"loss": 0.1406,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.03476322442293167,
"rewards/margins": 0.30484968423843384,
"rewards/rejected": -0.3396129012107849,
"step": 3470
},
{
"epoch": 0.93,
"learning_rate": 7.854209717842231e-08,
"logits/chosen": -1.5118751525878906,
"logits/rejected": -0.8683287501335144,
"logps/chosen": -623.8234252929688,
"logps/rejected": -1705.7734375,
"loss": 0.0768,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.04193173721432686,
"rewards/margins": 0.3931104838848114,
"rewards/rejected": -0.43504220247268677,
"step": 3480
},
{
"epoch": 0.93,
"learning_rate": 7.285980923996989e-08,
"logits/chosen": -1.451690435409546,
"logits/rejected": -0.4384433627128601,
"logps/chosen": -644.318603515625,
"logps/rejected": -1636.7640380859375,
"loss": 0.1159,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.048310182988643646,
"rewards/margins": 0.3499411344528198,
"rewards/rejected": -0.39825132489204407,
"step": 3490
},
{
"epoch": 0.93,
"learning_rate": 6.738782355044048e-08,
"logits/chosen": -1.5255537033081055,
"logits/rejected": -0.24880623817443848,
"logps/chosen": -594.3384399414062,
"logps/rejected": -1581.1204833984375,
"loss": 0.1277,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.042657919228076935,
"rewards/margins": 0.32654836773872375,
"rewards/rejected": -0.3692063093185425,
"step": 3500
},
{
"epoch": 0.94,
"learning_rate": 6.212661423609184e-08,
"logits/chosen": -1.2601430416107178,
"logits/rejected": -0.05472012236714363,
"logps/chosen": -635.506103515625,
"logps/rejected": -1533.6993408203125,
"loss": 0.1826,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.04515351355075836,
"rewards/margins": 0.28451424837112427,
"rewards/rejected": -0.3296678066253662,
"step": 3510
},
{
"epoch": 0.94,
"learning_rate": 5.707663716023021e-08,
"logits/chosen": -1.5181065797805786,
"logits/rejected": -0.2227887660264969,
"logps/chosen": -702.5686645507812,
"logps/rejected": -1479.894775390625,
"loss": 0.1602,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06307029724121094,
"rewards/margins": 0.28211596608161926,
"rewards/rejected": -0.3451862931251526,
"step": 3520
},
{
"epoch": 0.94,
"learning_rate": 5.22383298837098e-08,
"logits/chosen": -1.5388044118881226,
"logits/rejected": 0.21133394539356232,
"logps/chosen": -800.4866943359375,
"logps/rejected": -1580.938232421875,
"loss": 0.0983,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.08334775269031525,
"rewards/margins": 0.29393917322158813,
"rewards/rejected": -0.3772869110107422,
"step": 3530
},
{
"epoch": 0.94,
"learning_rate": 4.761211162702117e-08,
"logits/chosen": -1.441207766532898,
"logits/rejected": -0.5496786236763,
"logps/chosen": -713.4179077148438,
"logps/rejected": -1426.01171875,
"loss": 0.188,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.10053922981023788,
"rewards/margins": 0.22569382190704346,
"rewards/rejected": -0.32623302936553955,
"step": 3540
},
{
"epoch": 0.95,
"learning_rate": 4.319838323396691e-08,
"logits/chosen": -1.4728825092315674,
"logits/rejected": -0.8214455842971802,
"logps/chosen": -647.0958862304688,
"logps/rejected": -1621.828369140625,
"loss": 0.1727,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.06476866453886032,
"rewards/margins": 0.32229509949684143,
"rewards/rejected": -0.38706374168395996,
"step": 3550
},
{
"epoch": 0.95,
"learning_rate": 3.8997527136930004e-08,
"logits/chosen": -1.3555724620819092,
"logits/rejected": -0.9047378301620483,
"logps/chosen": -717.7271118164062,
"logps/rejected": -1752.408203125,
"loss": 0.1412,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.11167830228805542,
"rewards/margins": 0.3086579442024231,
"rewards/rejected": -0.4203362464904785,
"step": 3560
},
{
"epoch": 0.95,
"learning_rate": 3.5009907323737826e-08,
"logits/chosen": -1.3440874814987183,
"logits/rejected": -0.38815242052078247,
"logps/chosen": -572.4334716796875,
"logps/rejected": -1522.956787109375,
"loss": 0.1838,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.04414026811718941,
"rewards/margins": 0.2903124988079071,
"rewards/rejected": -0.3344528079032898,
"step": 3570
},
{
"epoch": 0.95,
"learning_rate": 3.1235869306123766e-08,
"logits/chosen": -1.7093461751937866,
"logits/rejected": -0.7086327075958252,
"logps/chosen": -623.2698364257812,
"logps/rejected": -1764.114013671875,
"loss": 0.109,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.05702175572514534,
"rewards/margins": 0.3699186444282532,
"rewards/rejected": -0.426940381526947,
"step": 3580
},
{
"epoch": 0.96,
"learning_rate": 2.767574008979007e-08,
"logits/chosen": -1.5890508890151978,
"logits/rejected": -0.45929059386253357,
"logps/chosen": -709.0275268554688,
"logps/rejected": -1698.400146484375,
"loss": 0.1108,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.06950731575489044,
"rewards/margins": 0.3168235123157501,
"rewards/rejected": -0.386330783367157,
"step": 3590
},
{
"epoch": 0.96,
"learning_rate": 2.4329828146074096e-08,
"logits/chosen": -1.4923101663589478,
"logits/rejected": -0.3146892189979553,
"logps/chosen": -752.1887817382812,
"logps/rejected": -1786.089111328125,
"loss": 0.1122,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.06624925881624222,
"rewards/margins": 0.3419404923915863,
"rewards/rejected": -0.4081897735595703,
"step": 3600
},
{
"epoch": 0.96,
"learning_rate": 2.1198423385220822e-08,
"logits/chosen": -1.4802709817886353,
"logits/rejected": -0.7376705408096313,
"logps/chosen": -644.2679443359375,
"logps/rejected": -1632.2701416015625,
"loss": 0.1783,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.042442698031663895,
"rewards/margins": 0.3448942005634308,
"rewards/rejected": -0.3873369097709656,
"step": 3610
},
{
"epoch": 0.97,
"learning_rate": 1.82817971312621e-08,
"logits/chosen": -1.607860803604126,
"logits/rejected": -0.5319154858589172,
"logps/chosen": -662.9993896484375,
"logps/rejected": -1706.1949462890625,
"loss": 0.1668,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.06334526836872101,
"rewards/margins": 0.3336828351020813,
"rewards/rejected": -0.3970281183719635,
"step": 3620
},
{
"epoch": 0.97,
"learning_rate": 1.5580202098509078e-08,
"logits/chosen": -1.5640172958374023,
"logits/rejected": -0.2071477621793747,
"logps/chosen": -634.8540649414062,
"logps/rejected": -1508.478515625,
"loss": 0.1485,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.06261152774095535,
"rewards/margins": 0.30838096141815186,
"rewards/rejected": -0.3709924817085266,
"step": 3630
},
{
"epoch": 0.97,
"learning_rate": 1.3093872369654148e-08,
"logits/chosen": -1.425930380821228,
"logits/rejected": 0.06674868613481522,
"logps/chosen": -671.8314819335938,
"logps/rejected": -1613.8541259765625,
"loss": 0.2111,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.0743933692574501,
"rewards/margins": 0.2954915761947632,
"rewards/rejected": -0.3698849081993103,
"step": 3640
},
{
"epoch": 0.97,
"learning_rate": 1.0823023375489128e-08,
"logits/chosen": -1.6556713581085205,
"logits/rejected": -0.5192909836769104,
"logps/chosen": -614.2623291015625,
"logps/rejected": -1719.060791015625,
"loss": 0.1069,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.052135247737169266,
"rewards/margins": 0.3784480690956116,
"rewards/rejected": -0.43058329820632935,
"step": 3650
},
{
"epoch": 0.98,
"learning_rate": 8.767851876239075e-09,
"logits/chosen": -1.2060736417770386,
"logits/rejected": -0.42587900161743164,
"logps/chosen": -745.9373168945312,
"logps/rejected": -1563.926513671875,
"loss": 0.1715,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.0904444083571434,
"rewards/margins": 0.29221174120903015,
"rewards/rejected": -0.38265615701675415,
"step": 3660
},
{
"epoch": 0.98,
"learning_rate": 6.9285359445145366e-09,
"logits/chosen": -1.5781863927841187,
"logits/rejected": -0.46057629585266113,
"logps/chosen": -630.7301025390625,
"logps/rejected": -1631.5906982421875,
"loss": 0.1287,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.030789542943239212,
"rewards/margins": 0.3398864269256592,
"rewards/rejected": -0.3706759810447693,
"step": 3670
},
{
"epoch": 0.98,
"learning_rate": 5.305234949880001e-09,
"logits/chosen": -1.4181653261184692,
"logits/rejected": -0.7784129977226257,
"logps/chosen": -598.7530517578125,
"logps/rejected": -1537.0283203125,
"loss": 0.2052,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.052745379507541656,
"rewards/margins": 0.27565133571624756,
"rewards/rejected": -0.328396737575531,
"step": 3680
},
{
"epoch": 0.98,
"learning_rate": 3.8980895450474455e-09,
"logits/chosen": -1.4237000942230225,
"logits/rejected": -0.025372039526700974,
"logps/chosen": -612.72998046875,
"logps/rejected": -1505.1551513671875,
"loss": 0.1556,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.048105791211128235,
"rewards/margins": 0.317216157913208,
"rewards/rejected": -0.36532193422317505,
"step": 3690
},
{
"epoch": 0.99,
"learning_rate": 2.7072216536885855e-09,
"logits/chosen": -1.537367582321167,
"logits/rejected": -0.594673752784729,
"logps/chosen": -647.0428466796875,
"logps/rejected": -1649.4918212890625,
"loss": 0.0992,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.06700156629085541,
"rewards/margins": 0.32940348982810974,
"rewards/rejected": -0.39640507102012634,
"step": 3700
},
{
"epoch": 0.99,
"learning_rate": 1.7327344598702667e-09,
"logits/chosen": -1.3739880323410034,
"logits/rejected": -0.8361124992370605,
"logps/chosen": -633.2244262695312,
"logps/rejected": -1885.47265625,
"loss": 0.0874,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.052168119698762894,
"rewards/margins": 0.41237759590148926,
"rewards/rejected": -0.46454566717147827,
"step": 3710
},
{
"epoch": 0.99,
"learning_rate": 9.747123991141193e-10,
"logits/chosen": -1.3567665815353394,
"logits/rejected": -0.42118167877197266,
"logps/chosen": -696.8218994140625,
"logps/rejected": -1684.748779296875,
"loss": 0.1024,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07300444692373276,
"rewards/margins": 0.34567874670028687,
"rewards/rejected": -0.4186832010746002,
"step": 3720
},
{
"epoch": 0.99,
"learning_rate": 4.332211510807427e-10,
"logits/chosen": -1.3411390781402588,
"logits/rejected": -0.13048940896987915,
"logps/chosen": -543.6580810546875,
"logps/rejected": -1574.061279296875,
"loss": 0.1457,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.04192924499511719,
"rewards/margins": 0.32620683312416077,
"rewards/rejected": -0.36813604831695557,
"step": 3730
},
{
"epoch": 1.0,
"learning_rate": 1.0830763387897902e-10,
"logits/chosen": -1.376807451248169,
"logits/rejected": 0.7155303955078125,
"logps/chosen": -710.9319458007812,
"logps/rejected": -1784.8050537109375,
"loss": 0.0957,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.06954208761453629,
"rewards/margins": 0.3634827136993408,
"rewards/rejected": -0.4330248236656189,
"step": 3740
},
{
"epoch": 1.0,
"learning_rate": 0.0,
"logits/chosen": -1.697113275527954,
"logits/rejected": -0.9965224266052246,
"logps/chosen": -698.7452392578125,
"logps/rejected": -1657.3492431640625,
"loss": 0.1618,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07053602486848831,
"rewards/margins": 0.3382786512374878,
"rewards/rejected": -0.4088146686553955,
"step": 3750
},
{
"epoch": 1.0,
"step": 3750,
"total_flos": 0.0,
"train_loss": 0.1831410455385844,
"train_runtime": 17876.9578,
"train_samples_per_second": 0.839,
"train_steps_per_second": 0.21
}
],
"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
}