llama-3.1-8b-kahaani-dpo-2 / trainer_state.json
sameearif's picture
Upload folder using huggingface_hub
5fa0bde verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 192,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.010416666666666666,
"grad_norm": 16.651927947998047,
"learning_rate": 2e-05,
"logits/chosen": -0.780666708946228,
"logits/rejected": -1.1369192600250244,
"logps/chosen": -3061.823486328125,
"logps/rejected": -220.772705078125,
"loss": 0.6931,
"rewards/accuracies": 0.0,
"rewards/chosen": 0.0,
"rewards/margins": 0.0,
"rewards/rejected": 0.0,
"step": 1
},
{
"epoch": 0.020833333333333332,
"grad_norm": 13.908960342407227,
"learning_rate": 4e-05,
"logits/chosen": -0.8081845641136169,
"logits/rejected": -1.2085561752319336,
"logps/chosen": -4412.2578125,
"logps/rejected": -267.08087158203125,
"loss": 0.6931,
"rewards/accuracies": 0.0,
"rewards/chosen": 0.0,
"rewards/margins": 0.0,
"rewards/rejected": 0.0,
"step": 2
},
{
"epoch": 0.03125,
"grad_norm": 15.761119842529297,
"learning_rate": 6e-05,
"logits/chosen": -0.9457322359085083,
"logits/rejected": -1.1844913959503174,
"logps/chosen": -3298.71435546875,
"logps/rejected": -223.52560424804688,
"loss": 0.6763,
"rewards/accuracies": 0.5,
"rewards/chosen": -0.01404876634478569,
"rewards/margins": -0.040423206984996796,
"rewards/rejected": 0.026374435052275658,
"step": 3
},
{
"epoch": 0.041666666666666664,
"grad_norm": 9.980835914611816,
"learning_rate": 8e-05,
"logits/chosen": -0.8384261131286621,
"logits/rejected": -1.271468162536621,
"logps/chosen": -3999.277587890625,
"logps/rejected": -225.41534423828125,
"loss": 0.4997,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.3802902400493622,
"rewards/margins": 0.5509151816368103,
"rewards/rejected": -0.17062492668628693,
"step": 4
},
{
"epoch": 0.052083333333333336,
"grad_norm": 6.138638973236084,
"learning_rate": 0.0001,
"logits/chosen": -0.8365007638931274,
"logits/rejected": -1.362606406211853,
"logps/chosen": -3391.94677734375,
"logps/rejected": -245.4417724609375,
"loss": 0.2954,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.9162139892578125,
"rewards/margins": 1.1493107080459595,
"rewards/rejected": -0.23309671878814697,
"step": 5
},
{
"epoch": 0.0625,
"grad_norm": 1.2922340631484985,
"learning_rate": 0.00012,
"logits/chosen": -0.8477813005447388,
"logits/rejected": -1.4737606048583984,
"logps/chosen": -7458.923828125,
"logps/rejected": -282.75335693359375,
"loss": 0.0518,
"rewards/accuracies": 1.0,
"rewards/chosen": 3.14288330078125,
"rewards/margins": 5.705089569091797,
"rewards/rejected": -2.562206268310547,
"step": 6
},
{
"epoch": 0.07291666666666667,
"grad_norm": 0.03415762260556221,
"learning_rate": 0.00014,
"logits/chosen": -0.7428842186927795,
"logits/rejected": -1.3752648830413818,
"logps/chosen": -6296.4091796875,
"logps/rejected": -276.24859619140625,
"loss": 0.001,
"rewards/accuracies": 1.0,
"rewards/chosen": 6.91751766204834,
"rewards/margins": 11.728842735290527,
"rewards/rejected": -4.8113250732421875,
"step": 7
},
{
"epoch": 0.08333333333333333,
"grad_norm": 0.0005445057176984847,
"learning_rate": 0.00016,
"logits/chosen": -0.8007906079292297,
"logits/rejected": -1.3979616165161133,
"logps/chosen": -6775.5029296875,
"logps/rejected": -370.7599182128906,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 11.398236274719238,
"rewards/margins": 20.943986892700195,
"rewards/rejected": -9.545750617980957,
"step": 8
},
{
"epoch": 0.09375,
"grad_norm": 0.005390194710344076,
"learning_rate": 0.00018,
"logits/chosen": -0.8591928482055664,
"logits/rejected": -0.895159125328064,
"logps/chosen": -2164.9541015625,
"logps/rejected": -472.39715576171875,
"loss": 0.0001,
"rewards/accuracies": 1.0,
"rewards/chosen": 5.005324363708496,
"rewards/margins": 25.770355224609375,
"rewards/rejected": -20.765029907226562,
"step": 9
},
{
"epoch": 0.10416666666666667,
"grad_norm": 3.3636655416557915e-07,
"learning_rate": 0.0002,
"logits/chosen": -0.8266401886940002,
"logits/rejected": -0.8240926265716553,
"logps/chosen": -2985.67236328125,
"logps/rejected": -462.8504638671875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 3.8522109985351562,
"rewards/margins": 27.728797912597656,
"rewards/rejected": -23.876588821411133,
"step": 10
},
{
"epoch": 0.11458333333333333,
"grad_norm": 5.287571491763288e-10,
"learning_rate": 0.00019998510240408496,
"logits/chosen": -0.9868767261505127,
"logits/rejected": -0.8054253458976746,
"logps/chosen": -1482.732177734375,
"logps/rejected": -553.533935546875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.7153975963592529,
"rewards/margins": 32.48991394042969,
"rewards/rejected": -33.20531463623047,
"step": 11
},
{
"epoch": 0.125,
"grad_norm": 26.858783721923828,
"learning_rate": 0.00019994041405510705,
"logits/chosen": -1.03706955909729,
"logits/rejected": -0.9741817712783813,
"logps/chosen": -4695.4384765625,
"logps/rejected": -721.7908935546875,
"loss": 0.0838,
"rewards/accuracies": 1.0,
"rewards/chosen": -9.903982162475586,
"rewards/margins": 36.569175720214844,
"rewards/rejected": -46.47315979003906,
"step": 12
},
{
"epoch": 0.13541666666666666,
"grad_norm": 9.055187155659894e-10,
"learning_rate": 0.0001998659482680456,
"logits/chosen": -0.962079644203186,
"logits/rejected": -1.3349699974060059,
"logps/chosen": -9016.650390625,
"logps/rejected": -793.662109375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.13396000862121582,
"rewards/margins": 55.1308708190918,
"rewards/rejected": -54.99691390991211,
"step": 13
},
{
"epoch": 0.14583333333333334,
"grad_norm": 4.094690808642554e-09,
"learning_rate": 0.0001997617272301248,
"logits/chosen": -0.7595808506011963,
"logits/rejected": -1.0764431953430176,
"logps/chosen": -2220.482421875,
"logps/rejected": -798.6175537109375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.7651824951171875,
"rewards/margins": 52.960113525390625,
"rewards/rejected": -55.72529602050781,
"step": 14
},
{
"epoch": 0.15625,
"grad_norm": 9.434321379675481e-12,
"learning_rate": 0.00019962778199420265,
"logits/chosen": -0.8709142804145813,
"logits/rejected": -1.1885780096054077,
"logps/chosen": -4517.1044921875,
"logps/rejected": -827.2481079101562,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -12.643299102783203,
"rewards/margins": 46.70207977294922,
"rewards/rejected": -59.34538269042969,
"step": 15
},
{
"epoch": 0.16666666666666666,
"grad_norm": 2.213202831359747e-20,
"learning_rate": 0.0001994641524695193,
"logits/chosen": -0.8794220685958862,
"logits/rejected": -0.8576048612594604,
"logps/chosen": -2103.159912109375,
"logps/rejected": -1043.563720703125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.4151291847229004,
"rewards/margins": 73.81220245361328,
"rewards/rejected": -76.22732543945312,
"step": 16
},
{
"epoch": 0.17708333333333334,
"grad_norm": 3.0938494821608487e-11,
"learning_rate": 0.0001992708874098054,
"logits/chosen": -0.8563255071640015,
"logits/rejected": -1.1423197984695435,
"logps/chosen": -5297.2294921875,
"logps/rejected": -875.4127197265625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.190893173217773,
"rewards/margins": 58.889068603515625,
"rewards/rejected": -64.0799560546875,
"step": 17
},
{
"epoch": 0.1875,
"grad_norm": 1.6154518123142973e-19,
"learning_rate": 0.00019904804439875633,
"logits/chosen": -0.7584964632987976,
"logits/rejected": -1.1960705518722534,
"logps/chosen": -4858.4482421875,
"logps/rejected": -979.1630249023438,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.413223743438721,
"rewards/margins": 65.54483795166016,
"rewards/rejected": -72.95806121826172,
"step": 18
},
{
"epoch": 0.19791666666666666,
"grad_norm": 1.5190748665707273e-20,
"learning_rate": 0.00019879568983287467,
"logits/chosen": -0.9190940856933594,
"logits/rejected": -1.1393502950668335,
"logps/chosen": -2892.24267578125,
"logps/rejected": -967.30615234375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.6857757568359375,
"rewards/margins": 66.6280288696289,
"rewards/rejected": -72.31380462646484,
"step": 19
},
{
"epoch": 0.20833333333333334,
"grad_norm": 5.615921756608865e-20,
"learning_rate": 0.0001985138989016874,
"logits/chosen": -0.8962702751159668,
"logits/rejected": -1.3306204080581665,
"logps/chosen": -6762.5302734375,
"logps/rejected": -995.3267211914062,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -4.49383544921875,
"rewards/margins": 71.0030746459961,
"rewards/rejected": -75.49691009521484,
"step": 20
},
{
"epoch": 0.21875,
"grad_norm": 1.7029158998823055e-17,
"learning_rate": 0.00019820275556534304,
"logits/chosen": -0.8105654716491699,
"logits/rejected": -1.3740750551223755,
"logps/chosen": -9757.419921875,
"logps/rejected": -954.4122314453125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.01361083984375,
"rewards/margins": 78.48196411132812,
"rewards/rejected": -71.46835327148438,
"step": 21
},
{
"epoch": 0.22916666666666666,
"grad_norm": 0.0,
"learning_rate": 0.00019786235252959553,
"logits/chosen": -0.8967666625976562,
"logits/rejected": -1.1008408069610596,
"logps/chosen": -2355.2626953125,
"logps/rejected": -1078.12060546875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -13.126736640930176,
"rewards/margins": 69.51663970947266,
"rewards/rejected": -82.64337921142578,
"step": 22
},
{
"epoch": 0.23958333333333334,
"grad_norm": 0.0,
"learning_rate": 0.00019749279121818235,
"logits/chosen": -0.8905834555625916,
"logits/rejected": -1.0848970413208008,
"logps/chosen": -2228.7802734375,
"logps/rejected": -996.7811279296875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -3.612191677093506,
"rewards/margins": 73.11354064941406,
"rewards/rejected": -76.7257308959961,
"step": 23
},
{
"epoch": 0.25,
"grad_norm": 7.647393330508138e-19,
"learning_rate": 0.0001970941817426052,
"logits/chosen": -0.8828914165496826,
"logits/rejected": -1.166571855545044,
"logps/chosen": -3267.3076171875,
"logps/rejected": -1007.6031494140625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.523654460906982,
"rewards/margins": 67.767822265625,
"rewards/rejected": -75.29147338867188,
"step": 24
},
{
"epoch": 0.2604166666666667,
"grad_norm": 1.2719628243941199e-17,
"learning_rate": 0.00019666664286932198,
"logits/chosen": -0.9088311791419983,
"logits/rejected": -1.0525307655334473,
"logps/chosen": -1816.2103271484375,
"logps/rejected": -972.173583984375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -12.783884048461914,
"rewards/margins": 61.763031005859375,
"rewards/rejected": -74.54691314697266,
"step": 25
},
{
"epoch": 0.2708333333333333,
"grad_norm": 5.43750881430877e-17,
"learning_rate": 0.00019621030198436006,
"logits/chosen": -1.0071444511413574,
"logits/rejected": -0.8765456676483154,
"logps/chosen": -1344.3419189453125,
"logps/rejected": -992.904052734375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -15.93281364440918,
"rewards/margins": 58.60974884033203,
"rewards/rejected": -74.54256439208984,
"step": 26
},
{
"epoch": 0.28125,
"grad_norm": 1.87694066018651e-12,
"learning_rate": 0.0001957252950553616,
"logits/chosen": -0.9608248472213745,
"logits/rejected": -1.1069602966308594,
"logps/chosen": -2613.587158203125,
"logps/rejected": -1058.5823974609375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.87086296081543,
"rewards/margins": 73.99343872070312,
"rewards/rejected": -79.86430358886719,
"step": 27
},
{
"epoch": 0.2916666666666667,
"grad_norm": 4.830028054873832e-19,
"learning_rate": 0.00019521176659107142,
"logits/chosen": -0.9120213985443115,
"logits/rejected": -1.3496168851852417,
"logps/chosen": -7237.09912109375,
"logps/rejected": -1196.45458984375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 2.345461368560791,
"rewards/margins": 95.51947784423828,
"rewards/rejected": -93.17402648925781,
"step": 28
},
{
"epoch": 0.3020833333333333,
"grad_norm": 4.793881215301673e-22,
"learning_rate": 0.0001946698695982806,
"logits/chosen": -0.8377887010574341,
"logits/rejected": -1.4051532745361328,
"logps/chosen": -4109.2421875,
"logps/rejected": -1000.89111328125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.656280517578125,
"rewards/margins": 71.58284759521484,
"rewards/rejected": -77.23912811279297,
"step": 29
},
{
"epoch": 0.3125,
"grad_norm": 6.545510560170225e-20,
"learning_rate": 0.00019409976553623766,
"logits/chosen": -1.0247584581375122,
"logits/rejected": -1.1513103246688843,
"logps/chosen": -3428.35400390625,
"logps/rejected": -1014.9542236328125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.2859251499176025,
"rewards/margins": 76.68400573730469,
"rewards/rejected": -77.96993255615234,
"step": 30
},
{
"epoch": 0.3229166666666667,
"grad_norm": 4.597397983254012e-19,
"learning_rate": 0.0001935016242685415,
"logits/chosen": -0.8390676975250244,
"logits/rejected": -1.2956117391586304,
"logps/chosen": -9027.2763671875,
"logps/rejected": -916.6038818359375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 7.76643705368042,
"rewards/margins": 78.03052520751953,
"rewards/rejected": -70.26408386230469,
"step": 31
},
{
"epoch": 0.3333333333333333,
"grad_norm": 1.3613232990958984e-13,
"learning_rate": 0.00019287562401253022,
"logits/chosen": -0.9091357588768005,
"logits/rejected": -1.3022887706756592,
"logps/chosen": -4318.7275390625,
"logps/rejected": -1064.7750244140625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -12.792261123657227,
"rewards/margins": 69.23944854736328,
"rewards/rejected": -82.03170776367188,
"step": 32
},
{
"epoch": 0.34375,
"grad_norm": 1.1914185441577697e-15,
"learning_rate": 0.00019222195128618106,
"logits/chosen": -0.8301032781600952,
"logits/rejected": -1.1454460620880127,
"logps/chosen": -3503.22802734375,
"logps/rejected": -1033.736083984375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 5.304028511047363,
"rewards/margins": 86.28971099853516,
"rewards/rejected": -80.98567962646484,
"step": 33
},
{
"epoch": 0.3541666666666667,
"grad_norm": 1.2077592159173465e-17,
"learning_rate": 0.00019154080085253666,
"logits/chosen": -0.9025095105171204,
"logits/rejected": -1.0029141902923584,
"logps/chosen": -2160.233642578125,
"logps/rejected": -988.8320922851562,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.006901741027832,
"rewards/margins": 67.99139404296875,
"rewards/rejected": -74.99829864501953,
"step": 34
},
{
"epoch": 0.3645833333333333,
"grad_norm": 0.0,
"learning_rate": 0.0001908323756616754,
"logits/chosen": -0.9620730876922607,
"logits/rejected": -1.303593635559082,
"logps/chosen": -4102.265625,
"logps/rejected": -1039.8701171875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.7988373041152954,
"rewards/margins": 78.1258544921875,
"rewards/rejected": -79.92469024658203,
"step": 35
},
{
"epoch": 0.375,
"grad_norm": 2.327858659183149e-19,
"learning_rate": 0.0001900968867902419,
"logits/chosen": -0.9782366156578064,
"logits/rejected": -1.1345422267913818,
"logps/chosen": -5296.07373046875,
"logps/rejected": -1034.09130859375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.9609848260879517,
"rewards/margins": 76.84639739990234,
"rewards/rejected": -75.88541412353516,
"step": 36
},
{
"epoch": 0.3854166666666667,
"grad_norm": 5.993950300697435e-16,
"learning_rate": 0.00018933455337855632,
"logits/chosen": -0.9719871282577515,
"logits/rejected": -1.2671374082565308,
"logps/chosen": -3619.00830078125,
"logps/rejected": -960.418212890625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.8363280296325684,
"rewards/margins": 73.91217041015625,
"rewards/rejected": -73.07584381103516,
"step": 37
},
{
"epoch": 0.3958333333333333,
"grad_norm": 0.0,
"learning_rate": 0.000188545602565321,
"logits/chosen": -0.8795933127403259,
"logits/rejected": -1.1734505891799927,
"logps/chosen": -2961.7783203125,
"logps/rejected": -1115.31494140625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -14.957999229431152,
"rewards/margins": 72.54431915283203,
"rewards/rejected": -87.5023193359375,
"step": 38
},
{
"epoch": 0.40625,
"grad_norm": 9.981808812592569e-17,
"learning_rate": 0.0001877302694199442,
"logits/chosen": -0.9492607116699219,
"logits/rejected": -1.0690467357635498,
"logps/chosen": -3375.642578125,
"logps/rejected": -970.4015502929688,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -11.076546669006348,
"rewards/margins": 63.78575897216797,
"rewards/rejected": -74.8623046875,
"step": 39
},
{
"epoch": 0.4166666666666667,
"grad_norm": 1.677683582476988e-19,
"learning_rate": 0.00018688879687250067,
"logits/chosen": -0.7364388108253479,
"logits/rejected": -1.2299830913543701,
"logps/chosen": -4095.498046875,
"logps/rejected": -1125.764892578125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -15.576887130737305,
"rewards/margins": 71.2161865234375,
"rewards/rejected": -86.79308319091797,
"step": 40
},
{
"epoch": 0.4270833333333333,
"grad_norm": 0.0,
"learning_rate": 0.0001860214356413501,
"logits/chosen": -0.8531750440597534,
"logits/rejected": -1.2205849885940552,
"logps/chosen": -4229.4951171875,
"logps/rejected": -1032.5693359375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 1.7727296352386475,
"rewards/margins": 79.98316192626953,
"rewards/rejected": -78.2104263305664,
"step": 41
},
{
"epoch": 0.4375,
"grad_norm": 7.346881715894921e-19,
"learning_rate": 0.00018512844415843514,
"logits/chosen": -0.8872064352035522,
"logits/rejected": -1.347999095916748,
"logps/chosen": -6776.28759765625,
"logps/rejected": -1047.91064453125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.6724517345428467,
"rewards/margins": 77.55692291259766,
"rewards/rejected": -78.22937774658203,
"step": 42
},
{
"epoch": 0.4479166666666667,
"grad_norm": 0.0,
"learning_rate": 0.00018421008849228118,
"logits/chosen": -0.9359738230705261,
"logits/rejected": -1.132539987564087,
"logps/chosen": -1939.6717529296875,
"logps/rejected": -968.7603759765625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -10.06414794921875,
"rewards/margins": 64.12657928466797,
"rewards/rejected": -74.19072723388672,
"step": 43
},
{
"epoch": 0.4583333333333333,
"grad_norm": 1.1837645481569055e-22,
"learning_rate": 0.00018326664226872065,
"logits/chosen": -1.025484561920166,
"logits/rejected": -1.1521519422531128,
"logps/chosen": -3046.67138671875,
"logps/rejected": -1010.868896484375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 2.433685302734375,
"rewards/margins": 79.09088134765625,
"rewards/rejected": -76.65719604492188,
"step": 44
},
{
"epoch": 0.46875,
"grad_norm": 1.0841111503063809e-20,
"learning_rate": 0.00018229838658936564,
"logits/chosen": -0.8892084956169128,
"logits/rejected": -1.0169939994812012,
"logps/chosen": -3305.859130859375,
"logps/rejected": -1010.6842041015625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -6.042553901672363,
"rewards/margins": 71.8106689453125,
"rewards/rejected": -77.85322570800781,
"step": 45
},
{
"epoch": 0.4791666666666667,
"grad_norm": 4.2357202397516193e-19,
"learning_rate": 0.00018130560994785325,
"logits/chosen": -0.8572372198104858,
"logits/rejected": -1.3382017612457275,
"logps/chosen": -3960.66796875,
"logps/rejected": -1066.4102783203125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -14.808450698852539,
"rewards/margins": 67.266357421875,
"rewards/rejected": -82.07479858398438,
"step": 46
},
{
"epoch": 0.4895833333333333,
"grad_norm": 0.0,
"learning_rate": 0.00018028860814388827,
"logits/chosen": -0.9053932428359985,
"logits/rejected": -1.245862603187561,
"logps/chosen": -4537.33935546875,
"logps/rejected": -1066.2734375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -6.734985828399658,
"rewards/margins": 74.92082977294922,
"rewards/rejected": -81.65581512451172,
"step": 47
},
{
"epoch": 0.5,
"grad_norm": 0.0,
"learning_rate": 0.00017924768419510904,
"logits/chosen": -1.034196138381958,
"logits/rejected": -1.2119903564453125,
"logps/chosen": -4595.2041015625,
"logps/rejected": -964.75146484375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.3183517456054688,
"rewards/margins": 72.99729919433594,
"rewards/rejected": -74.3156509399414,
"step": 48
},
{
"epoch": 0.5104166666666666,
"grad_norm": 1.567794059815996e-19,
"learning_rate": 0.000178183148246803,
"logits/chosen": -0.9009298086166382,
"logits/rejected": -1.2393324375152588,
"logps/chosen": -2687.751953125,
"logps/rejected": -1079.8603515625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.771054267883301,
"rewards/margins": 76.5450210571289,
"rewards/rejected": -82.31607818603516,
"step": 49
},
{
"epoch": 0.5208333333333334,
"grad_norm": 0.0,
"learning_rate": 0.00017709531747949796,
"logits/chosen": -0.7311493158340454,
"logits/rejected": -0.9473284482955933,
"logps/chosen": -1437.1895751953125,
"logps/rejected": -1016.4439086914062,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.1630539894104004,
"rewards/margins": 73.96327209472656,
"rewards/rejected": -75.1263198852539,
"step": 50
},
{
"epoch": 0.53125,
"grad_norm": 7.311199956683466e-17,
"learning_rate": 0.0001759845160144579,
"logits/chosen": -0.9461476802825928,
"logits/rejected": -1.2513340711593628,
"logps/chosen": -9740.1708984375,
"logps/rejected": -998.367431640625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -8.608297348022461,
"rewards/margins": 67.94587707519531,
"rewards/rejected": -76.5541763305664,
"step": 51
},
{
"epoch": 0.5416666666666666,
"grad_norm": 0.0,
"learning_rate": 0.00017485107481711012,
"logits/chosen": -1.0065760612487793,
"logits/rejected": -1.176657795906067,
"logps/chosen": -3810.3505859375,
"logps/rejected": -1045.2769775390625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.958648681640625,
"rewards/margins": 79.65708923339844,
"rewards/rejected": -80.61573791503906,
"step": 52
},
{
"epoch": 0.5520833333333334,
"grad_norm": 0.0,
"learning_rate": 0.00017369533159843369,
"logits/chosen": -0.7354742288589478,
"logits/rejected": -1.1330161094665527,
"logps/chosen": -4675.5478515625,
"logps/rejected": -1095.03857421875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -11.42770004272461,
"rewards/margins": 73.42713928222656,
"rewards/rejected": -84.8548355102539,
"step": 53
},
{
"epoch": 0.5625,
"grad_norm": 0.0,
"learning_rate": 0.00017251763071433765,
"logits/chosen": -0.846391499042511,
"logits/rejected": -1.4115591049194336,
"logps/chosen": -6438.18359375,
"logps/rejected": -1123.985107421875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -10.262456893920898,
"rewards/margins": 74.60464477539062,
"rewards/rejected": -84.86710357666016,
"step": 54
},
{
"epoch": 0.5729166666666666,
"grad_norm": 1.4700509215416664e-14,
"learning_rate": 0.00017131832306305965,
"logits/chosen": -0.960889458656311,
"logits/rejected": -1.234442114830017,
"logps/chosen": -3840.143310546875,
"logps/rejected": -906.8041381835938,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -3.883606433868408,
"rewards/margins": 67.3929443359375,
"rewards/rejected": -71.27655029296875,
"step": 55
},
{
"epoch": 0.5833333333333334,
"grad_norm": 0.0,
"learning_rate": 0.00017009776598061495,
"logits/chosen": -0.9445286393165588,
"logits/rejected": -0.9528659582138062,
"logps/chosen": -1339.93603515625,
"logps/rejected": -1048.4464111328125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.9421162605285645,
"rewards/margins": 72.82998657226562,
"rewards/rejected": -78.77210998535156,
"step": 56
},
{
"epoch": 0.59375,
"grad_norm": 0.0,
"learning_rate": 0.0001688563231343277,
"logits/chosen": -0.921560525894165,
"logits/rejected": -1.464065432548523,
"logps/chosen": -6107.14794921875,
"logps/rejected": -1177.58154296875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.16182255744934082,
"rewards/margins": 89.7105941772461,
"rewards/rejected": -89.54876708984375,
"step": 57
},
{
"epoch": 0.6041666666666666,
"grad_norm": 2.2782913188817383e-17,
"learning_rate": 0.00016759436441447545,
"logits/chosen": -0.9432457685470581,
"logits/rejected": -0.9813876748085022,
"logps/chosen": -2878.421875,
"logps/rejected": -973.9136962890625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -8.945228576660156,
"rewards/margins": 64.66465759277344,
"rewards/rejected": -73.6098861694336,
"step": 58
},
{
"epoch": 0.6145833333333334,
"grad_norm": 3.532171644213204e-18,
"learning_rate": 0.00016631226582407952,
"logits/chosen": -1.011328101158142,
"logits/rejected": -1.1443777084350586,
"logps/chosen": -4537.185546875,
"logps/rejected": -1023.2095336914062,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.814834713935852,
"rewards/margins": 78.44985961914062,
"rewards/rejected": -77.63502502441406,
"step": 59
},
{
"epoch": 0.625,
"grad_norm": 1.0837672078321995e-16,
"learning_rate": 0.00016501040936687443,
"logits/chosen": -0.7958677411079407,
"logits/rejected": -1.2076987028121948,
"logps/chosen": -3676.333984375,
"logps/rejected": -991.25146484375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -14.135693550109863,
"rewards/margins": 61.884666442871094,
"rewards/rejected": -76.0203628540039,
"step": 60
},
{
"epoch": 0.6354166666666666,
"grad_norm": 5.757463345409198e-20,
"learning_rate": 0.00016368918293348892,
"logits/chosen": -0.9800617694854736,
"logits/rejected": -1.1600340604782104,
"logps/chosen": -3713.349853515625,
"logps/rejected": -1017.238037109375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -4.153411865234375,
"rewards/margins": 75.27999114990234,
"rewards/rejected": -79.43340301513672,
"step": 61
},
{
"epoch": 0.6458333333333334,
"grad_norm": 5.311725612781588e-16,
"learning_rate": 0.00016234898018587337,
"logits/chosen": -0.8364681005477905,
"logits/rejected": -1.16648530960083,
"logps/chosen": -2505.11962890625,
"logps/rejected": -1125.0926513671875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -18.13332176208496,
"rewards/margins": 66.983642578125,
"rewards/rejected": -85.11697387695312,
"step": 62
},
{
"epoch": 0.65625,
"grad_norm": 9.133200582579723e-18,
"learning_rate": 0.00016099020044000727,
"logits/chosen": -1.0037099123001099,
"logits/rejected": -1.054494857788086,
"logps/chosen": -2708.493408203125,
"logps/rejected": -1137.5875244140625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -6.967913627624512,
"rewards/margins": 81.5879898071289,
"rewards/rejected": -88.555908203125,
"step": 63
},
{
"epoch": 0.6666666666666666,
"grad_norm": 2.593275012522582e-16,
"learning_rate": 0.00015961324854692254,
"logits/chosen": -0.8983689546585083,
"logits/rejected": -1.127020001411438,
"logps/chosen": -2932.719482421875,
"logps/rejected": -1062.8060302734375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.111117362976074,
"rewards/margins": 79.9205551147461,
"rewards/rejected": -82.03167724609375,
"step": 64
},
{
"epoch": 0.6770833333333334,
"grad_norm": 0.0,
"learning_rate": 0.00015821853477207708,
"logits/chosen": -0.9519010186195374,
"logits/rejected": -1.1288293600082397,
"logps/chosen": -2201.759521484375,
"logps/rejected": -898.9304809570312,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -6.13864278793335,
"rewards/margins": 63.858665466308594,
"rewards/rejected": -69.997314453125,
"step": 65
},
{
"epoch": 0.6875,
"grad_norm": 3.743392066509216e-23,
"learning_rate": 0.00015680647467311557,
"logits/chosen": -0.8885273337364197,
"logits/rejected": -1.2975856065750122,
"logps/chosen": -3025.942138671875,
"logps/rejected": -998.9232177734375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -14.540674209594727,
"rewards/margins": 62.277652740478516,
"rewards/rejected": -76.81832885742188,
"step": 66
},
{
"epoch": 0.6979166666666666,
"grad_norm": 1.5581050296373422e-17,
"learning_rate": 0.0001553774889760533,
"logits/chosen": -0.922539472579956,
"logits/rejected": -1.1917293071746826,
"logps/chosen": -2637.00439453125,
"logps/rejected": -1103.3564453125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.390442848205566,
"rewards/margins": 80.02389526367188,
"rewards/rejected": -85.41433715820312,
"step": 67
},
{
"epoch": 0.7083333333333334,
"grad_norm": 1.5840932732213986e-11,
"learning_rate": 0.00015393200344991995,
"logits/chosen": -0.9601885080337524,
"logits/rejected": -1.379237174987793,
"logps/chosen": -7276.7919921875,
"logps/rejected": -1131.1331787109375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.817126274108887,
"rewards/margins": 79.90594482421875,
"rewards/rejected": -85.72306823730469,
"step": 68
},
{
"epoch": 0.71875,
"grad_norm": 3.447390015483765e-17,
"learning_rate": 0.0001524704487799008,
"logits/chosen": -0.9075456857681274,
"logits/rejected": -1.2865498065948486,
"logps/chosen": -4736.16015625,
"logps/rejected": -1066.030029296875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.429553270339966,
"rewards/margins": 80.66062927246094,
"rewards/rejected": -83.0901870727539,
"step": 69
},
{
"epoch": 0.7291666666666666,
"grad_norm": 0.0,
"learning_rate": 0.0001509932604390136,
"logits/chosen": -0.9576442837715149,
"logits/rejected": -1.1714496612548828,
"logps/chosen": -2777.48486328125,
"logps/rejected": -981.0249633789062,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.5866700410842896,
"rewards/margins": 74.16746520996094,
"rewards/rejected": -74.75413513183594,
"step": 70
},
{
"epoch": 0.7395833333333334,
"grad_norm": 0.0,
"learning_rate": 0.00014950087855835815,
"logits/chosen": -0.957038164138794,
"logits/rejected": -1.309131383895874,
"logps/chosen": -3568.40625,
"logps/rejected": -1100.7939453125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.437561511993408,
"rewards/margins": 76.48944854736328,
"rewards/rejected": -83.92700958251953,
"step": 71
},
{
"epoch": 0.75,
"grad_norm": 2.023447353270838e-18,
"learning_rate": 0.00014799374779597867,
"logits/chosen": -0.879957377910614,
"logits/rejected": -1.1743906736373901,
"logps/chosen": -2927.203369140625,
"logps/rejected": -942.46728515625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.356558322906494,
"rewards/margins": 65.15306854248047,
"rewards/rejected": -72.50962829589844,
"step": 72
},
{
"epoch": 0.7604166666666666,
"grad_norm": 0.0,
"learning_rate": 0.00014647231720437686,
"logits/chosen": -0.9975137710571289,
"logits/rejected": -1.096327543258667,
"logps/chosen": -2194.671875,
"logps/rejected": -1060.455322265625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.339242458343506,
"rewards/margins": 73.82234191894531,
"rewards/rejected": -81.16158294677734,
"step": 73
},
{
"epoch": 0.7708333333333334,
"grad_norm": 0.0,
"learning_rate": 0.00014493704009671613,
"logits/chosen": -0.8398927450180054,
"logits/rejected": -1.05116868019104,
"logps/chosen": -2852.46875,
"logps/rejected": -1053.2982177734375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -3.2871978282928467,
"rewards/margins": 77.74524688720703,
"rewards/rejected": -81.0324478149414,
"step": 74
},
{
"epoch": 0.78125,
"grad_norm": 2.823407507790421e-09,
"learning_rate": 0.00014338837391175582,
"logits/chosen": -0.8749938607215881,
"logits/rejected": -1.3326060771942139,
"logps/chosen": -6237.5947265625,
"logps/rejected": -1008.5339965820312,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -12.8448486328125,
"rewards/margins": 65.41060638427734,
"rewards/rejected": -78.25544738769531,
"step": 75
},
{
"epoch": 0.7916666666666666,
"grad_norm": 2.146854222062471e-20,
"learning_rate": 0.0001418267800775565,
"logits/chosen": -0.9367476105690002,
"logits/rejected": -1.0628546476364136,
"logps/chosen": -3435.0185546875,
"logps/rejected": -1105.599609375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -12.55868911743164,
"rewards/margins": 72.16712951660156,
"rewards/rejected": -84.72582244873047,
"step": 76
},
{
"epoch": 0.8020833333333334,
"grad_norm": 4.4584097976070534e-21,
"learning_rate": 0.00014025272387399674,
"logits/chosen": -1.013418436050415,
"logits/rejected": -1.210724115371704,
"logps/chosen": -4156.4990234375,
"logps/rejected": -1065.85009765625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -16.16033363342285,
"rewards/margins": 65.87461853027344,
"rewards/rejected": -82.03495025634766,
"step": 77
},
{
"epoch": 0.8125,
"grad_norm": 5.350105004942962e-20,
"learning_rate": 0.0001386666742941419,
"logits/chosen": -0.9383605718612671,
"logits/rejected": -1.2753703594207764,
"logps/chosen": -5303.810546875,
"logps/rejected": -1006.8165893554688,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -4.698285102844238,
"rewards/margins": 73.23123168945312,
"rewards/rejected": -77.92951965332031,
"step": 78
},
{
"epoch": 0.8229166666666666,
"grad_norm": 2.6276267314835005e-17,
"learning_rate": 0.00013706910390450677,
"logits/chosen": -0.928679347038269,
"logits/rejected": -1.1945797204971313,
"logps/chosen": -5853.28369140625,
"logps/rejected": -996.289306640625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -11.84226131439209,
"rewards/margins": 66.27059173583984,
"rewards/rejected": -78.11285400390625,
"step": 79
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.0,
"learning_rate": 0.00013546048870425356,
"logits/chosen": -0.90166175365448,
"logits/rejected": -1.432413935661316,
"logps/chosen": -5923.66845703125,
"logps/rejected": -1065.8446044921875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -9.516674041748047,
"rewards/margins": 72.67932891845703,
"rewards/rejected": -82.19600677490234,
"step": 80
},
{
"epoch": 0.84375,
"grad_norm": 0.0,
"learning_rate": 0.00013384130798336705,
"logits/chosen": -0.9168898463249207,
"logits/rejected": -1.3556921482086182,
"logps/chosen": -6369.79296875,
"logps/rejected": -1242.0037841796875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -11.644824028015137,
"rewards/margins": 84.6693344116211,
"rewards/rejected": -96.31415557861328,
"step": 81
},
{
"epoch": 0.8541666666666666,
"grad_norm": 9.279878730339463e-19,
"learning_rate": 0.00013221204417984908,
"logits/chosen": -0.8945307731628418,
"logits/rejected": -1.1446576118469238,
"logps/chosen": -4128.4306640625,
"logps/rejected": -958.488037109375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.115054190158844,
"rewards/margins": 73.7982177734375,
"rewards/rejected": -73.91326904296875,
"step": 82
},
{
"epoch": 0.8645833333333334,
"grad_norm": 2.394110249810547e-15,
"learning_rate": 0.0001305731827359753,
"logits/chosen": -0.9232196807861328,
"logits/rejected": -1.1997020244598389,
"logps/chosen": -4594.69287109375,
"logps/rejected": -1065.5748291015625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.030016899108887,
"rewards/margins": 74.59611511230469,
"rewards/rejected": -81.62613677978516,
"step": 83
},
{
"epoch": 0.875,
"grad_norm": 1.8590894193296648e-17,
"learning_rate": 0.00012892521195365678,
"logits/chosen": -1.0041048526763916,
"logits/rejected": -1.1207554340362549,
"logps/chosen": -3463.850341796875,
"logps/rejected": -1104.9521484375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -3.943849563598633,
"rewards/margins": 80.47566223144531,
"rewards/rejected": -84.41950988769531,
"step": 84
},
{
"epoch": 0.8854166666666666,
"grad_norm": 9.305548438880714e-21,
"learning_rate": 0.00012726862284894938,
"logits/chosen": -0.9365824460983276,
"logits/rejected": -1.378727674484253,
"logps/chosen": -5006.0732421875,
"logps/rejected": -1108.4298095703125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 11.526544570922852,
"rewards/margins": 95.75658416748047,
"rewards/rejected": -84.23004913330078,
"step": 85
},
{
"epoch": 0.8958333333333334,
"grad_norm": 2.6055184079070218e-14,
"learning_rate": 0.0001256039090057547,
"logits/chosen": -0.8847850561141968,
"logits/rejected": -1.2132179737091064,
"logps/chosen": -3019.03466796875,
"logps/rejected": -1116.0423583984375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -17.331069946289062,
"rewards/margins": 68.22811889648438,
"rewards/rejected": -85.55919647216797,
"step": 86
},
{
"epoch": 0.90625,
"grad_norm": 1.337394415048057e-13,
"learning_rate": 0.0001239315664287558,
"logits/chosen": -0.9080939292907715,
"logits/rejected": -1.1880122423171997,
"logps/chosen": -6956.1767578125,
"logps/rejected": -1072.7021484375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.3681154251098633,
"rewards/margins": 78.97460174560547,
"rewards/rejected": -81.34272003173828,
"step": 87
},
{
"epoch": 0.9166666666666666,
"grad_norm": 1.1238145078521735e-18,
"learning_rate": 0.00012225209339563145,
"logits/chosen": -0.871356189250946,
"logits/rejected": -1.2935049533843994,
"logps/chosen": -7261.5849609375,
"logps/rejected": -934.6922607421875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.1817047595977783,
"rewards/margins": 71.47894287109375,
"rewards/rejected": -72.66064453125,
"step": 88
},
{
"epoch": 0.9270833333333334,
"grad_norm": 3.7998657856768726e-21,
"learning_rate": 0.00012056599030859366,
"logits/chosen": -0.8709771633148193,
"logits/rejected": -1.3064707517623901,
"logps/chosen": -4282.35302734375,
"logps/rejected": -1147.09765625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.661895751953125,
"rewards/margins": 80.48426818847656,
"rewards/rejected": -88.14615631103516,
"step": 89
},
{
"epoch": 0.9375,
"grad_norm": 3.9913009729067834e-17,
"learning_rate": 0.00011887375954529168,
"logits/chosen": -1.0120631456375122,
"logits/rejected": -1.4202260971069336,
"logps/chosen": -4564.931640625,
"logps/rejected": -1029.2471923828125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 25.81535530090332,
"rewards/margins": 104.50590515136719,
"rewards/rejected": -78.6905517578125,
"step": 90
},
{
"epoch": 0.9479166666666666,
"grad_norm": 4.153984713967665e-21,
"learning_rate": 0.00011717590530912763,
"logits/chosen": -0.8913455009460449,
"logits/rejected": -1.2823923826217651,
"logps/chosen": -3196.804443359375,
"logps/rejected": -1069.353271484375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -14.811059951782227,
"rewards/margins": 67.35678100585938,
"rewards/rejected": -82.16783142089844,
"step": 91
},
{
"epoch": 0.9583333333333334,
"grad_norm": 4.296434017478194e-15,
"learning_rate": 0.00011547293347902812,
"logits/chosen": -1.0650385618209839,
"logits/rejected": -1.330976963043213,
"logps/chosen": -5423.759765625,
"logps/rejected": -1142.623779296875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -11.46921157836914,
"rewards/margins": 78.08798217773438,
"rewards/rejected": -89.55718994140625,
"step": 92
},
{
"epoch": 0.96875,
"grad_norm": 6.566554551941457e-20,
"learning_rate": 0.00011376535145871684,
"logits/chosen": -0.9406051635742188,
"logits/rejected": -1.2664682865142822,
"logps/chosen": -3839.88671875,
"logps/rejected": -1140.073486328125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -17.9644775390625,
"rewards/margins": 70.13662719726562,
"rewards/rejected": -88.10110473632812,
"step": 93
},
{
"epoch": 0.9791666666666666,
"grad_norm": 0.0,
"learning_rate": 0.0001120536680255323,
"logits/chosen": -0.9597834348678589,
"logits/rejected": -1.1578893661499023,
"logps/chosen": -2502.4541015625,
"logps/rejected": -1016.5823974609375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.05165565013885498,
"rewards/margins": 78.03656005859375,
"rewards/rejected": -78.08821105957031,
"step": 94
},
{
"epoch": 0.9895833333333334,
"grad_norm": 2.4043631071810274e-16,
"learning_rate": 0.00011033839317883701,
"logits/chosen": -0.876549482345581,
"logits/rejected": -1.2577821016311646,
"logps/chosen": -3624.76611328125,
"logps/rejected": -1066.66015625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 13.061856269836426,
"rewards/margins": 95.77525329589844,
"rewards/rejected": -82.7134017944336,
"step": 95
},
{
"epoch": 1.0,
"grad_norm": 0.0,
"learning_rate": 0.00010862003798806196,
"logits/chosen": -0.9283385276794434,
"logits/rejected": -1.1488637924194336,
"logps/chosen": -4157.3427734375,
"logps/rejected": -1044.188232421875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -10.248306274414062,
"rewards/margins": 71.64842224121094,
"rewards/rejected": -81.896728515625,
"step": 96
},
{
"epoch": 1.0104166666666667,
"grad_norm": 5.512825971716348e-20,
"learning_rate": 0.00010689911444043248,
"logits/chosen": -0.7668694257736206,
"logits/rejected": -1.1733273267745972,
"logps/chosen": -3267.61962890625,
"logps/rejected": -1233.9364013671875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -16.425668716430664,
"rewards/margins": 77.82489013671875,
"rewards/rejected": -94.25055694580078,
"step": 97
},
{
"epoch": 1.0208333333333333,
"grad_norm": 0.0,
"learning_rate": 0.00010517613528842097,
"logits/chosen": -1.0368558168411255,
"logits/rejected": -1.1599735021591187,
"logps/chosen": -3525.150634765625,
"logps/rejected": -1166.176025390625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.8862044811248779,
"rewards/margins": 90.17179107666016,
"rewards/rejected": -89.28559112548828,
"step": 98
},
{
"epoch": 1.03125,
"grad_norm": 0.0,
"learning_rate": 0.00010345161389697082,
"logits/chosen": -0.7527876496315002,
"logits/rejected": -1.1752800941467285,
"logps/chosen": -4636.1845703125,
"logps/rejected": -945.1699829101562,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 3.4130585193634033,
"rewards/margins": 77.43528747558594,
"rewards/rejected": -74.02222442626953,
"step": 99
},
{
"epoch": 1.0416666666666667,
"grad_norm": 0.0,
"learning_rate": 0.00010172606409053886,
"logits/chosen": -0.9287230968475342,
"logits/rejected": -1.0137070417404175,
"logps/chosen": -2631.903564453125,
"logps/rejected": -1053.97802734375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -12.378585815429688,
"rewards/margins": 67.58805847167969,
"rewards/rejected": -79.96664428710938,
"step": 100
},
{
"epoch": 1.0520833333333333,
"grad_norm": 2.335907492570711e-16,
"learning_rate": 0.0001,
"logits/chosen": -0.9676191806793213,
"logits/rejected": -1.2866928577423096,
"logps/chosen": -3778.99462890625,
"logps/rejected": -1169.9747314453125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -19.198566436767578,
"rewards/margins": 72.15303802490234,
"rewards/rejected": -91.35160064697266,
"step": 101
},
{
"epoch": 1.0625,
"grad_norm": 3.0693231460494636e-17,
"learning_rate": 9.827393590946116e-05,
"logits/chosen": -0.8066754341125488,
"logits/rejected": -1.3259482383728027,
"logps/chosen": -4285.404296875,
"logps/rejected": -1049.9791259765625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -3.3176026344299316,
"rewards/margins": 77.5743179321289,
"rewards/rejected": -80.89192199707031,
"step": 102
},
{
"epoch": 1.0729166666666667,
"grad_norm": 7.908512019412167e-19,
"learning_rate": 9.654838610302923e-05,
"logits/chosen": -0.8500939607620239,
"logits/rejected": -1.3845233917236328,
"logps/chosen": -4154.744140625,
"logps/rejected": -998.893798828125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 1.2993621826171875,
"rewards/margins": 79.06625366210938,
"rewards/rejected": -77.76689147949219,
"step": 103
},
{
"epoch": 1.0833333333333333,
"grad_norm": 1.884301279524458e-19,
"learning_rate": 9.482386471157904e-05,
"logits/chosen": -0.803484320640564,
"logits/rejected": -1.1875181198120117,
"logps/chosen": -2875.72509765625,
"logps/rejected": -1111.62060546875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.236685037612915,
"rewards/margins": 81.31465148925781,
"rewards/rejected": -83.55133819580078,
"step": 104
},
{
"epoch": 1.09375,
"grad_norm": 0.0,
"learning_rate": 9.31008855595675e-05,
"logits/chosen": -0.9018905162811279,
"logits/rejected": -1.1758604049682617,
"logps/chosen": -2899.448486328125,
"logps/rejected": -1177.0833740234375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -6.149624824523926,
"rewards/margins": 84.1700668334961,
"rewards/rejected": -90.31968688964844,
"step": 105
},
{
"epoch": 1.1041666666666667,
"grad_norm": 4.704146143799659e-13,
"learning_rate": 9.137996201193805e-05,
"logits/chosen": -0.9577762484550476,
"logits/rejected": -1.2123889923095703,
"logps/chosen": -3999.40087890625,
"logps/rejected": -1089.39453125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -3.861725091934204,
"rewards/margins": 79.9576187133789,
"rewards/rejected": -83.81934356689453,
"step": 106
},
{
"epoch": 1.1145833333333333,
"grad_norm": 3.220187770403815e-22,
"learning_rate": 8.9661606821163e-05,
"logits/chosen": -0.9275646209716797,
"logits/rejected": -1.140000343322754,
"logps/chosen": -6079.86474609375,
"logps/rejected": -1045.149658203125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.459826946258545,
"rewards/margins": 80.20182800292969,
"rewards/rejected": -81.66165161132812,
"step": 107
},
{
"epoch": 1.125,
"grad_norm": 0.0,
"learning_rate": 8.79463319744677e-05,
"logits/chosen": -0.9333359003067017,
"logits/rejected": -1.2084262371063232,
"logps/chosen": -3093.892578125,
"logps/rejected": -1011.2752685546875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -14.587095260620117,
"rewards/margins": 64.42304992675781,
"rewards/rejected": -79.01013946533203,
"step": 108
},
{
"epoch": 1.1354166666666667,
"grad_norm": 4.516256854982578e-16,
"learning_rate": 8.62346485412832e-05,
"logits/chosen": -0.9335057735443115,
"logits/rejected": -1.2948254346847534,
"logps/chosen": -3485.91015625,
"logps/rejected": -943.7158203125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -15.117718696594238,
"rewards/margins": 58.55492401123047,
"rewards/rejected": -73.67264556884766,
"step": 109
},
{
"epoch": 1.1458333333333333,
"grad_norm": 1.7443403811759252e-20,
"learning_rate": 8.452706652097186e-05,
"logits/chosen": -0.8658100962638855,
"logits/rejected": -1.2765027284622192,
"logps/chosen": -7044.2666015625,
"logps/rejected": -1111.2900390625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -14.557941436767578,
"rewards/margins": 70.0662841796875,
"rewards/rejected": -84.62422943115234,
"step": 110
},
{
"epoch": 1.15625,
"grad_norm": 8.975433514902403e-14,
"learning_rate": 8.282409469087239e-05,
"logits/chosen": -0.8567153215408325,
"logits/rejected": -1.1527928113937378,
"logps/chosen": -2753.281005859375,
"logps/rejected": -1106.07861328125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -6.908112049102783,
"rewards/margins": 77.96578979492188,
"rewards/rejected": -84.8739013671875,
"step": 111
},
{
"epoch": 1.1666666666666667,
"grad_norm": 1.8399236127730756e-13,
"learning_rate": 8.112624045470835e-05,
"logits/chosen": -0.8963369727134705,
"logits/rejected": -1.1266080141067505,
"logps/chosen": -1911.738037109375,
"logps/rejected": -935.976806640625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -10.368061065673828,
"rewards/margins": 60.68651580810547,
"rewards/rejected": -71.05458068847656,
"step": 112
},
{
"epoch": 1.1770833333333333,
"grad_norm": 4.411308711811606e-19,
"learning_rate": 7.943400969140635e-05,
"logits/chosen": -0.9802509546279907,
"logits/rejected": -1.2022995948791504,
"logps/chosen": -4026.29296875,
"logps/rejected": -1028.704345703125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -9.705430030822754,
"rewards/margins": 69.18346405029297,
"rewards/rejected": -78.8888931274414,
"step": 113
},
{
"epoch": 1.1875,
"grad_norm": 6.856654159841134e-19,
"learning_rate": 7.774790660436858e-05,
"logits/chosen": -0.7806614637374878,
"logits/rejected": -1.2099708318710327,
"logps/chosen": -3049.276123046875,
"logps/rejected": -1083.7523193359375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.252618432044983,
"rewards/margins": 82.09521484375,
"rewards/rejected": -83.34783935546875,
"step": 114
},
{
"epoch": 1.1979166666666667,
"grad_norm": 0.0,
"learning_rate": 7.606843357124426e-05,
"logits/chosen": -0.8244500160217285,
"logits/rejected": -1.3403112888336182,
"logps/chosen": -2625.35205078125,
"logps/rejected": -1107.463134765625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -9.240277290344238,
"rewards/margins": 75.68666076660156,
"rewards/rejected": -84.92694091796875,
"step": 115
},
{
"epoch": 1.2083333333333333,
"grad_norm": 6.782594740894701e-20,
"learning_rate": 7.43960909942453e-05,
"logits/chosen": -0.9496563673019409,
"logits/rejected": -1.0177761316299438,
"logps/chosen": -1753.7301025390625,
"logps/rejected": -994.8018798828125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -13.330022811889648,
"rewards/margins": 62.15534210205078,
"rewards/rejected": -75.48535919189453,
"step": 116
},
{
"epoch": 1.21875,
"grad_norm": 1.650893022233743e-17,
"learning_rate": 7.273137715105063e-05,
"logits/chosen": -0.8159215450286865,
"logits/rejected": -1.11468505859375,
"logps/chosen": -2901.86376953125,
"logps/rejected": -1028.982666015625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -4.9665374755859375,
"rewards/margins": 75.73015594482422,
"rewards/rejected": -80.69668579101562,
"step": 117
},
{
"epoch": 1.2291666666666667,
"grad_norm": 9.338665629292994e-18,
"learning_rate": 7.107478804634325e-05,
"logits/chosen": -0.9511147737503052,
"logits/rejected": -1.131158471107483,
"logps/chosen": -2839.02099609375,
"logps/rejected": -916.0258178710938,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 4.969090938568115,
"rewards/margins": 76.42955780029297,
"rewards/rejected": -71.46046447753906,
"step": 118
},
{
"epoch": 1.2395833333333333,
"grad_norm": 8.370479371910218e-23,
"learning_rate": 6.942681726402473e-05,
"logits/chosen": -0.945068895816803,
"logits/rejected": -1.2830828428268433,
"logps/chosen": -3800.778076171875,
"logps/rejected": -905.9716796875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 1.0239747762680054,
"rewards/margins": 69.88013458251953,
"rewards/rejected": -68.85615539550781,
"step": 119
},
{
"epoch": 1.25,
"grad_norm": 0.0,
"learning_rate": 6.778795582015097e-05,
"logits/chosen": -0.9145641326904297,
"logits/rejected": -1.1153881549835205,
"logps/chosen": -2149.349365234375,
"logps/rejected": -1061.110107421875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.1018218994140625,
"rewards/margins": 79.47099304199219,
"rewards/rejected": -81.57280731201172,
"step": 120
},
{
"epoch": 1.2604166666666667,
"grad_norm": 3.655288958431413e-17,
"learning_rate": 6.615869201663296e-05,
"logits/chosen": -0.8868393898010254,
"logits/rejected": -1.3098794221878052,
"logps/chosen": -5764.04736328125,
"logps/rejected": -1124.540283203125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.0704987049102783,
"rewards/margins": 85.03953552246094,
"rewards/rejected": -86.11003112792969,
"step": 121
},
{
"epoch": 1.2708333333333333,
"grad_norm": 5.505299111008798e-16,
"learning_rate": 6.453951129574644e-05,
"logits/chosen": -0.9295135140419006,
"logits/rejected": -1.1804428100585938,
"logps/chosen": -5864.943359375,
"logps/rejected": -1078.9697265625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -10.316208839416504,
"rewards/margins": 71.87071228027344,
"rewards/rejected": -82.18692016601562,
"step": 122
},
{
"epoch": 1.28125,
"grad_norm": 0.0,
"learning_rate": 6.293089609549325e-05,
"logits/chosen": -0.7975695133209229,
"logits/rejected": -1.3155672550201416,
"logps/chosen": -3814.4150390625,
"logps/rejected": -1204.483154296875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 1.016815185546875,
"rewards/margins": 94.49725341796875,
"rewards/rejected": -93.48043823242188,
"step": 123
},
{
"epoch": 1.2916666666666667,
"grad_norm": 1.4148333719063387e-19,
"learning_rate": 6.133332570585812e-05,
"logits/chosen": -1.063108205795288,
"logits/rejected": -1.2205861806869507,
"logps/chosen": -2892.2685546875,
"logps/rejected": -1125.844970703125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -13.142558097839355,
"rewards/margins": 72.0556411743164,
"rewards/rejected": -85.19820404052734,
"step": 124
},
{
"epoch": 1.3020833333333333,
"grad_norm": 5.8489230255285674e-21,
"learning_rate": 5.9747276126003257e-05,
"logits/chosen": -0.7630377411842346,
"logits/rejected": -1.316693902015686,
"logps/chosen": -3374.70263671875,
"logps/rejected": -1142.8436279296875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -8.068140029907227,
"rewards/margins": 79.92533874511719,
"rewards/rejected": -87.99348449707031,
"step": 125
},
{
"epoch": 1.3125,
"grad_norm": 1.6185688098081561e-18,
"learning_rate": 5.817321992244351e-05,
"logits/chosen": -0.9196380376815796,
"logits/rejected": -1.2248481512069702,
"logps/chosen": -3206.56591796875,
"logps/rejected": -1012.30908203125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -11.104272842407227,
"rewards/margins": 65.4676284790039,
"rewards/rejected": -76.5718994140625,
"step": 126
},
{
"epoch": 1.3229166666666667,
"grad_norm": 2.5687962656379646e-21,
"learning_rate": 5.6611626088244194e-05,
"logits/chosen": -0.9707508087158203,
"logits/rejected": -1.2447017431259155,
"logps/chosen": -4089.699462890625,
"logps/rejected": -1120.4154052734375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -9.00418758392334,
"rewards/margins": 76.06697082519531,
"rewards/rejected": -85.07115936279297,
"step": 127
},
{
"epoch": 1.3333333333333333,
"grad_norm": 2.0503403694456911e-22,
"learning_rate": 5.506295990328385e-05,
"logits/chosen": -0.9009501934051514,
"logits/rejected": -1.353840708732605,
"logps/chosen": -7221.36279296875,
"logps/rejected": -1135.1197509765625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.12373661994934082,
"rewards/margins": 88.40489959716797,
"rewards/rejected": -88.28115844726562,
"step": 128
},
{
"epoch": 1.34375,
"grad_norm": 1.344498075173984e-21,
"learning_rate": 5.3527682795623146e-05,
"logits/chosen": -0.8090143799781799,
"logits/rejected": -1.0533530712127686,
"logps/chosen": -2006.4832763671875,
"logps/rejected": -1043.121826171875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -9.14123249053955,
"rewards/margins": 70.5691146850586,
"rewards/rejected": -79.71034240722656,
"step": 129
},
{
"epoch": 1.3541666666666667,
"grad_norm": 6.440810335695039e-21,
"learning_rate": 5.200625220402139e-05,
"logits/chosen": -0.9704806208610535,
"logits/rejected": -1.0929075479507446,
"logps/chosen": -2996.92822265625,
"logps/rejected": -1068.793212890625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -9.935484886169434,
"rewards/margins": 71.94236755371094,
"rewards/rejected": -81.87785339355469,
"step": 130
},
{
"epoch": 1.3645833333333333,
"grad_norm": 0.0,
"learning_rate": 5.0499121441641864e-05,
"logits/chosen": -0.9403969049453735,
"logits/rejected": -1.3582621812820435,
"logps/chosen": -8281.2607421875,
"logps/rejected": -1090.318115234375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 1.1239867210388184,
"rewards/margins": 84.06963348388672,
"rewards/rejected": -82.94564819335938,
"step": 131
},
{
"epoch": 1.375,
"grad_norm": 1.1185308141413086e-19,
"learning_rate": 4.900673956098644e-05,
"logits/chosen": -1.0328657627105713,
"logits/rejected": -1.1909589767456055,
"logps/chosen": -1817.351318359375,
"logps/rejected": -1014.7122802734375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.688718318939209,
"rewards/margins": 72.41348266601562,
"rewards/rejected": -78.1021957397461,
"step": 132
},
{
"epoch": 1.3854166666666667,
"grad_norm": 3.2959890386576225e-18,
"learning_rate": 4.75295512200992e-05,
"logits/chosen": -0.9048435091972351,
"logits/rejected": -1.2165101766586304,
"logps/chosen": -3945.4169921875,
"logps/rejected": -1073.447265625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.457312107086182,
"rewards/margins": 76.04530334472656,
"rewards/rejected": -81.50261688232422,
"step": 133
},
{
"epoch": 1.3958333333333333,
"grad_norm": 0.0,
"learning_rate": 4.606799655008009e-05,
"logits/chosen": -0.903038740158081,
"logits/rejected": -1.4270763397216797,
"logps/chosen": -5543.68017578125,
"logps/rejected": -1106.446533203125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -3.1213319301605225,
"rewards/margins": 83.86710357666016,
"rewards/rejected": -86.98844146728516,
"step": 134
},
{
"epoch": 1.40625,
"grad_norm": 0.0,
"learning_rate": 4.462251102394669e-05,
"logits/chosen": -0.8994408845901489,
"logits/rejected": -1.3292877674102783,
"logps/chosen": -3318.34814453125,
"logps/rejected": -1010.1569213867188,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.89552640914917,
"rewards/margins": 67.90574645996094,
"rewards/rejected": -75.80128479003906,
"step": 135
},
{
"epoch": 1.4166666666666667,
"grad_norm": 1.062754219307314e-21,
"learning_rate": 4.3193525326884435e-05,
"logits/chosen": -0.9541503190994263,
"logits/rejected": -1.0579588413238525,
"logps/chosen": -4469.5009765625,
"logps/rejected": -955.7963256835938,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -12.544214248657227,
"rewards/margins": 58.10245132446289,
"rewards/rejected": -70.64666748046875,
"step": 136
},
{
"epoch": 1.4270833333333333,
"grad_norm": 4.9187133763885766e-17,
"learning_rate": 4.1781465227922957e-05,
"logits/chosen": -1.0173542499542236,
"logits/rejected": -1.2005999088287354,
"logps/chosen": -2528.12060546875,
"logps/rejected": -1051.4796142578125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.2586699724197388,
"rewards/margins": 77.34017944335938,
"rewards/rejected": -78.59884643554688,
"step": 137
},
{
"epoch": 1.4375,
"grad_norm": 0.0,
"learning_rate": 4.038675145307747e-05,
"logits/chosen": -0.967241644859314,
"logits/rejected": -1.1873809099197388,
"logps/chosen": -3900.778564453125,
"logps/rejected": -1132.687744140625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.1816892623901367,
"rewards/margins": 83.33070373535156,
"rewards/rejected": -84.51239013671875,
"step": 138
},
{
"epoch": 1.4479166666666667,
"grad_norm": 2.2958917370833966e-17,
"learning_rate": 3.900979955999271e-05,
"logits/chosen": -0.8518582582473755,
"logits/rejected": -0.9221400618553162,
"logps/chosen": -1183.7716064453125,
"logps/rejected": -911.7036743164062,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -6.865546226501465,
"rewards/margins": 64.08380126953125,
"rewards/rejected": -70.9493408203125,
"step": 139
},
{
"epoch": 1.4583333333333333,
"grad_norm": 0.0,
"learning_rate": 3.7651019814126654e-05,
"logits/chosen": -0.9010441899299622,
"logits/rejected": -1.190494179725647,
"logps/chosen": -5493.0009765625,
"logps/rejected": -1155.7314453125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.8245346546173096,
"rewards/margins": 85.78129577636719,
"rewards/rejected": -88.6058349609375,
"step": 140
},
{
"epoch": 1.46875,
"grad_norm": 0.0,
"learning_rate": 3.6310817066511105e-05,
"logits/chosen": -1.0124684572219849,
"logits/rejected": -1.103775978088379,
"logps/chosen": -1913.3040771484375,
"logps/rejected": -1010.6784057617188,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.425868511199951,
"rewards/margins": 71.89915466308594,
"rewards/rejected": -77.32502746582031,
"step": 141
},
{
"epoch": 1.4791666666666667,
"grad_norm": 0.0,
"learning_rate": 3.498959063312558e-05,
"logits/chosen": -0.8120708465576172,
"logits/rejected": -1.4091525077819824,
"logps/chosen": -4661.7578125,
"logps/rejected": -888.3680419921875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.055895805358887,
"rewards/margins": 62.75235366821289,
"rewards/rejected": -67.8082504272461,
"step": 142
},
{
"epoch": 1.4895833333333333,
"grad_norm": 1.9747184031200515e-15,
"learning_rate": 3.36877341759205e-05,
"logits/chosen": -0.8257265090942383,
"logits/rejected": -1.121315836906433,
"logps/chosen": -1288.9190673828125,
"logps/rejected": -972.7769775390625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.115662097930908,
"rewards/margins": 67.79209899902344,
"rewards/rejected": -74.90776824951172,
"step": 143
},
{
"epoch": 1.5,
"grad_norm": 0.0,
"learning_rate": 3.2405635585524565e-05,
"logits/chosen": -1.0017499923706055,
"logits/rejected": -1.1351391077041626,
"logps/chosen": -3693.326171875,
"logps/rejected": -1070.761962890625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -13.161089897155762,
"rewards/margins": 69.91401672363281,
"rewards/rejected": -83.07510375976562,
"step": 144
},
{
"epoch": 1.5104166666666665,
"grad_norm": 0.0,
"learning_rate": 3.114367686567228e-05,
"logits/chosen": -0.848055899143219,
"logits/rejected": -1.0882829427719116,
"logps/chosen": -2654.097412109375,
"logps/rejected": -1041.62060546875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -14.7376070022583,
"rewards/margins": 65.37726593017578,
"rewards/rejected": -80.11487579345703,
"step": 145
},
{
"epoch": 1.5208333333333335,
"grad_norm": 3.5242391416192426e-14,
"learning_rate": 2.9902234019385057e-05,
"logits/chosen": -0.8896538019180298,
"logits/rejected": -1.258103370666504,
"logps/chosen": -3455.951171875,
"logps/rejected": -998.981201171875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -11.142499923706055,
"rewards/margins": 64.65351104736328,
"rewards/rejected": -75.79601287841797,
"step": 146
},
{
"epoch": 1.53125,
"grad_norm": 4.5005137312183197e-13,
"learning_rate": 2.8681676936940393e-05,
"logits/chosen": -0.9204589128494263,
"logits/rejected": -1.0349059104919434,
"logps/chosen": -1558.69287109375,
"logps/rejected": -1021.4735107421875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -11.636395454406738,
"rewards/margins": 64.9701919555664,
"rewards/rejected": -76.6065902709961,
"step": 147
},
{
"epoch": 1.5416666666666665,
"grad_norm": 3.658716396254952e-13,
"learning_rate": 2.7482369285662378e-05,
"logits/chosen": -0.8321296572685242,
"logits/rejected": -1.1634087562561035,
"logps/chosen": -2103.5849609375,
"logps/rejected": -974.1217041015625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -28.24515724182129,
"rewards/margins": 45.907371520996094,
"rewards/rejected": -74.15252685546875,
"step": 148
},
{
"epoch": 1.5520833333333335,
"grad_norm": 3.7246280947998844e-22,
"learning_rate": 2.6304668401566335e-05,
"logits/chosen": -0.9244425296783447,
"logits/rejected": -1.3096859455108643,
"logps/chosen": -5354.45458984375,
"logps/rejected": -1064.54443359375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.046282768249512,
"rewards/margins": 76.85649871826172,
"rewards/rejected": -81.90278625488281,
"step": 149
},
{
"epoch": 1.5625,
"grad_norm": 1.2950027054019793e-20,
"learning_rate": 2.514892518288988e-05,
"logits/chosen": -0.8687339425086975,
"logits/rejected": -1.2188998460769653,
"logps/chosen": -4577.24462890625,
"logps/rejected": -929.9271240234375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -4.647838115692139,
"rewards/margins": 67.19004821777344,
"rewards/rejected": -71.83788299560547,
"step": 150
},
{
"epoch": 1.5729166666666665,
"grad_norm": 3.180693751106868e-17,
"learning_rate": 2.401548398554213e-05,
"logits/chosen": -0.8746656775474548,
"logits/rejected": -1.1335382461547852,
"logps/chosen": -2621.76708984375,
"logps/rejected": -1069.098876953125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -4.797000408172607,
"rewards/margins": 76.91392517089844,
"rewards/rejected": -81.71092987060547,
"step": 151
},
{
"epoch": 1.5833333333333335,
"grad_norm": 0.0,
"learning_rate": 2.290468252050204e-05,
"logits/chosen": -0.7061495780944824,
"logits/rejected": -1.2625129222869873,
"logps/chosen": -5062.939453125,
"logps/rejected": -1201.242919921875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -11.179224014282227,
"rewards/margins": 82.08878326416016,
"rewards/rejected": -93.26800537109375,
"step": 152
},
{
"epoch": 1.59375,
"grad_norm": 0.0,
"learning_rate": 2.181685175319702e-05,
"logits/chosen": -0.7203177213668823,
"logits/rejected": -1.3191170692443848,
"logps/chosen": -3958.314453125,
"logps/rejected": -1011.8955078125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.7906357049942017,
"rewards/margins": 78.65202331542969,
"rewards/rejected": -77.86138916015625,
"step": 153
},
{
"epoch": 1.6041666666666665,
"grad_norm": 0.0,
"learning_rate": 2.0752315804890977e-05,
"logits/chosen": -0.9195300340652466,
"logits/rejected": -1.2774739265441895,
"logps/chosen": -8186.68408203125,
"logps/rejected": -1072.425048828125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -1.4632796049118042,
"rewards/margins": 80.83226013183594,
"rewards/rejected": -82.29553985595703,
"step": 154
},
{
"epoch": 1.6145833333333335,
"grad_norm": 4.0147753867538186e-20,
"learning_rate": 1.971139185611176e-05,
"logits/chosen": -0.9163135886192322,
"logits/rejected": -1.3071476221084595,
"logps/chosen": -4530.13427734375,
"logps/rejected": -1092.8966064453125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -4.696630954742432,
"rewards/margins": 80.32160949707031,
"rewards/rejected": -85.01823425292969,
"step": 155
},
{
"epoch": 1.625,
"grad_norm": 1.5609576705471867e-19,
"learning_rate": 1.8694390052146737e-05,
"logits/chosen": -0.8604260683059692,
"logits/rejected": -1.270695686340332,
"logps/chosen": -3943.85595703125,
"logps/rejected": -1087.4296875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -8.993682861328125,
"rewards/margins": 75.90512084960938,
"rewards/rejected": -84.8988037109375,
"step": 156
},
{
"epoch": 1.6354166666666665,
"grad_norm": 2.811402606227264e-17,
"learning_rate": 1.7701613410634365e-05,
"logits/chosen": -0.8890691995620728,
"logits/rejected": -1.3079321384429932,
"logps/chosen": -7185.2529296875,
"logps/rejected": -1080.0333251953125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -8.33111572265625,
"rewards/margins": 74.3923568725586,
"rewards/rejected": -82.72347259521484,
"step": 157
},
{
"epoch": 1.6458333333333335,
"grad_norm": 1.0592323188268055e-13,
"learning_rate": 1.6733357731279377e-05,
"logits/chosen": -0.8480991721153259,
"logits/rejected": -1.1103326082229614,
"logps/chosen": -3468.0048828125,
"logps/rejected": -1067.985595703125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.42734670639038086,
"rewards/margins": 82.94938659667969,
"rewards/rejected": -83.3767318725586,
"step": 158
},
{
"epoch": 1.65625,
"grad_norm": 0.0,
"learning_rate": 1.5789911507718826e-05,
"logits/chosen": -0.7679794430732727,
"logits/rejected": -1.1826531887054443,
"logps/chosen": -2692.7138671875,
"logps/rejected": -1040.9814453125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -13.65127182006836,
"rewards/margins": 67.80027770996094,
"rewards/rejected": -81.45155334472656,
"step": 159
},
{
"epoch": 1.6666666666666665,
"grad_norm": 2.101126106157376e-17,
"learning_rate": 1.4871555841564887e-05,
"logits/chosen": -0.7971072196960449,
"logits/rejected": -1.2701507806777954,
"logps/chosen": -2931.4375,
"logps/rejected": -1022.1775512695312,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.8612060546875,
"rewards/margins": 74.42487335205078,
"rewards/rejected": -80.28607940673828,
"step": 160
},
{
"epoch": 1.6770833333333335,
"grad_norm": 1.693179950985332e-16,
"learning_rate": 1.3978564358649927e-05,
"logits/chosen": -1.000674843788147,
"logits/rejected": -1.1274532079696655,
"logps/chosen": -2341.431640625,
"logps/rejected": -1146.2425537109375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -3.8021240234375,
"rewards/margins": 82.45307922363281,
"rewards/rejected": -86.25519561767578,
"step": 161
},
{
"epoch": 1.6875,
"grad_norm": 1.248795501349495e-18,
"learning_rate": 1.311120312749935e-05,
"logits/chosen": -0.9954484701156616,
"logits/rejected": -1.0421305894851685,
"logps/chosen": -1413.7686767578125,
"logps/rejected": -919.8433837890625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -10.1903657913208,
"rewards/margins": 60.226654052734375,
"rewards/rejected": -70.4170150756836,
"step": 162
},
{
"epoch": 1.6979166666666665,
"grad_norm": 7.506130692859752e-17,
"learning_rate": 1.2269730580055805e-05,
"logits/chosen": -0.7554247975349426,
"logits/rejected": -1.1726787090301514,
"logps/chosen": -2398.90234375,
"logps/rejected": -1041.2353515625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -10.836740493774414,
"rewards/margins": 68.93004608154297,
"rewards/rejected": -79.76679229736328,
"step": 163
},
{
"epoch": 1.7083333333333335,
"grad_norm": 3.144996540161571e-16,
"learning_rate": 1.1454397434679021e-05,
"logits/chosen": -0.8778572082519531,
"logits/rejected": -1.1429264545440674,
"logps/chosen": -2819.60009765625,
"logps/rejected": -1024.1851806640625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -3.5772266387939453,
"rewards/margins": 75.2021713256836,
"rewards/rejected": -78.7793960571289,
"step": 164
},
{
"epoch": 1.71875,
"grad_norm": 4.54889006982497e-15,
"learning_rate": 1.0665446621443708e-05,
"logits/chosen": -1.0242542028427124,
"logits/rejected": -1.2949650287628174,
"logps/chosen": -6388.978515625,
"logps/rejected": -1101.523193359375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 23.088947296142578,
"rewards/margins": 107.18841552734375,
"rewards/rejected": -84.09947204589844,
"step": 165
},
{
"epoch": 1.7291666666666665,
"grad_norm": 1.2591869387457155e-11,
"learning_rate": 9.903113209758096e-06,
"logits/chosen": -1.0302661657333374,
"logits/rejected": -1.0212035179138184,
"logps/chosen": -1567.7783203125,
"logps/rejected": -1051.6270751953125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -6.1559648513793945,
"rewards/margins": 72.15129089355469,
"rewards/rejected": -78.30725860595703,
"step": 166
},
{
"epoch": 1.7395833333333335,
"grad_norm": 0.0,
"learning_rate": 9.1676243383246e-06,
"logits/chosen": -0.9699329137802124,
"logits/rejected": -1.217832088470459,
"logps/chosen": -3370.20361328125,
"logps/rejected": -1056.4344482421875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 1.4962918758392334,
"rewards/margins": 82.08119201660156,
"rewards/rejected": -80.58489990234375,
"step": 167
},
{
"epoch": 1.75,
"grad_norm": 6.902272136647668e-19,
"learning_rate": 8.45919914746337e-06,
"logits/chosen": -0.8314164876937866,
"logits/rejected": -1.1248615980148315,
"logps/chosen": -2003.40185546875,
"logps/rejected": -976.593017578125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.036044120788574,
"rewards/margins": 70.12460327148438,
"rewards/rejected": -75.16064453125,
"step": 168
},
{
"epoch": 1.7604166666666665,
"grad_norm": 0.0,
"learning_rate": 7.778048713818975e-06,
"logits/chosen": -1.0096337795257568,
"logits/rejected": -1.224050521850586,
"logps/chosen": -3408.56396484375,
"logps/rejected": -1120.5277099609375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.887763977050781,
"rewards/margins": 79.56890869140625,
"rewards/rejected": -85.45668029785156,
"step": 169
},
{
"epoch": 1.7708333333333335,
"grad_norm": 0.0,
"learning_rate": 7.124375987469767e-06,
"logits/chosen": -0.7806891798973083,
"logits/rejected": -1.4077966213226318,
"logps/chosen": -4579.115234375,
"logps/rejected": -1092.3642578125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -5.048169136047363,
"rewards/margins": 79.57754516601562,
"rewards/rejected": -84.6257095336914,
"step": 170
},
{
"epoch": 1.78125,
"grad_norm": 8.991780414846024e-21,
"learning_rate": 6.498375731458528e-06,
"logits/chosen": -0.8369664549827576,
"logits/rejected": -1.2401611804962158,
"logps/chosen": -4373.06640625,
"logps/rejected": -1185.349365234375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -10.959744453430176,
"rewards/margins": 80.5737533569336,
"rewards/rejected": -91.53350067138672,
"step": 171
},
{
"epoch": 1.7916666666666665,
"grad_norm": 3.254465541259741e-20,
"learning_rate": 5.900234463762366e-06,
"logits/chosen": -0.9217289090156555,
"logits/rejected": -1.3291540145874023,
"logps/chosen": -4704.79736328125,
"logps/rejected": -1218.4345703125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -18.816030502319336,
"rewards/margins": 75.90402221679688,
"rewards/rejected": -94.72004699707031,
"step": 172
},
{
"epoch": 1.8020833333333335,
"grad_norm": 6.992826570984174e-20,
"learning_rate": 5.3301304017194135e-06,
"logits/chosen": -0.8925175666809082,
"logits/rejected": -1.3186287879943848,
"logps/chosen": -5796.84375,
"logps/rejected": -1010.6710815429688,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -7.582611560821533,
"rewards/margins": 71.03655242919922,
"rewards/rejected": -78.61917114257812,
"step": 173
},
{
"epoch": 1.8125,
"grad_norm": 0.0,
"learning_rate": 4.788233408928589e-06,
"logits/chosen": -0.9071134328842163,
"logits/rejected": -1.1578150987625122,
"logps/chosen": -3065.755615234375,
"logps/rejected": -1122.5201416015625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -14.236775398254395,
"rewards/margins": 73.01652526855469,
"rewards/rejected": -87.2532958984375,
"step": 174
},
{
"epoch": 1.8229166666666665,
"grad_norm": 5.8420251147017e-20,
"learning_rate": 4.27470494463843e-06,
"logits/chosen": -0.8542720079421997,
"logits/rejected": -1.22062087059021,
"logps/chosen": -4755.65087890625,
"logps/rejected": -1087.69970703125,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 1.7637419700622559,
"rewards/margins": 85.65948486328125,
"rewards/rejected": -83.89574432373047,
"step": 175
},
{
"epoch": 1.8333333333333335,
"grad_norm": 4.5275721134567166e-20,
"learning_rate": 3.789698015639953e-06,
"logits/chosen": -0.8436378836631775,
"logits/rejected": -1.2432456016540527,
"logps/chosen": -7707.376953125,
"logps/rejected": -1006.131103515625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.132122755050659,
"rewards/margins": 74.2339096069336,
"rewards/rejected": -76.36602783203125,
"step": 176
},
{
"epoch": 1.84375,
"grad_norm": 0.0,
"learning_rate": 3.3333571306780497e-06,
"logits/chosen": -1.1021344661712646,
"logits/rejected": -1.146433711051941,
"logps/chosen": -2144.27734375,
"logps/rejected": -1221.460693359375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -13.952573776245117,
"rewards/margins": 81.55621337890625,
"rewards/rejected": -95.5087890625,
"step": 177
},
{
"epoch": 1.8541666666666665,
"grad_norm": 3.1447959791819358e-09,
"learning_rate": 2.905818257394799e-06,
"logits/chosen": -0.9445458650588989,
"logits/rejected": -1.1954491138458252,
"logps/chosen": -3473.337158203125,
"logps/rejected": -898.44140625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -6.790580749511719,
"rewards/margins": 61.649497985839844,
"rewards/rejected": -68.44007873535156,
"step": 178
},
{
"epoch": 1.8645833333333335,
"grad_norm": 3.4308754787927946e-22,
"learning_rate": 2.5072087818176382e-06,
"logits/chosen": -0.9306582808494568,
"logits/rejected": -1.2371917963027954,
"logps/chosen": -3138.842041015625,
"logps/rejected": -1054.2342529296875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -13.166943550109863,
"rewards/margins": 68.14605712890625,
"rewards/rejected": -81.31299591064453,
"step": 179
},
{
"epoch": 1.875,
"grad_norm": 0.0,
"learning_rate": 2.137647470404469e-06,
"logits/chosen": -1.0094091892242432,
"logits/rejected": -1.3734456300735474,
"logps/chosen": -6787.31982421875,
"logps/rejected": -909.1167602539062,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.7543637752532959,
"rewards/margins": 70.34402465820312,
"rewards/rejected": -71.09838104248047,
"step": 180
},
{
"epoch": 1.8854166666666665,
"grad_norm": 6.411009053864542e-20,
"learning_rate": 1.797244434656975e-06,
"logits/chosen": -1.0234758853912354,
"logits/rejected": -1.2576205730438232,
"logps/chosen": -2733.744384765625,
"logps/rejected": -1061.349609375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 2.5823726654052734,
"rewards/margins": 84.70896911621094,
"rewards/rejected": -82.12660217285156,
"step": 181
},
{
"epoch": 1.8958333333333335,
"grad_norm": 0.0,
"learning_rate": 1.48610109831262e-06,
"logits/chosen": -0.8497348427772522,
"logits/rejected": -1.0572631359100342,
"logps/chosen": -2718.9326171875,
"logps/rejected": -961.6370849609375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": 0.9914793968200684,
"rewards/margins": 75.01031494140625,
"rewards/rejected": -74.01884460449219,
"step": 182
},
{
"epoch": 1.90625,
"grad_norm": 2.2827734383898875e-18,
"learning_rate": 1.2043101671253554e-06,
"logits/chosen": -0.9114844799041748,
"logits/rejected": -1.1452364921569824,
"logps/chosen": -1678.3172607421875,
"logps/rejected": -1151.85400390625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -14.129800796508789,
"rewards/margins": 74.070556640625,
"rewards/rejected": -88.20036315917969,
"step": 183
},
{
"epoch": 1.9166666666666665,
"grad_norm": 0.0,
"learning_rate": 9.519556012436815e-07,
"logits/chosen": -0.9810994267463684,
"logits/rejected": -1.067641019821167,
"logps/chosen": -2078.021728515625,
"logps/rejected": -1072.85498046875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -13.242856979370117,
"rewards/margins": 66.07984161376953,
"rewards/rejected": -79.32270050048828,
"step": 184
},
{
"epoch": 1.9270833333333335,
"grad_norm": 1.8167316177431924e-16,
"learning_rate": 7.291125901946027e-07,
"logits/chosen": -0.8321889042854309,
"logits/rejected": -1.1509894132614136,
"logps/chosen": -3819.90966796875,
"logps/rejected": -1095.14306640625,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -10.86672592163086,
"rewards/margins": 71.09647369384766,
"rewards/rejected": -81.96320343017578,
"step": 185
},
{
"epoch": 1.9375,
"grad_norm": 1.0820013757477458e-17,
"learning_rate": 5.358475304807375e-07,
"logits/chosen": -1.0654444694519043,
"logits/rejected": -1.2142765522003174,
"logps/chosen": -7236.927734375,
"logps/rejected": -949.28857421875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.7587647438049316,
"rewards/margins": 70.48192596435547,
"rewards/rejected": -73.24069213867188,
"step": 186
},
{
"epoch": 1.9479166666666665,
"grad_norm": 2.45197261230215e-17,
"learning_rate": 3.7221800579735346e-07,
"logits/chosen": -0.8213499188423157,
"logits/rejected": -1.2857908010482788,
"logps/chosen": -6349.48046875,
"logps/rejected": -1107.80615234375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -6.317949295043945,
"rewards/margins": 78.53043365478516,
"rewards/rejected": -84.84838104248047,
"step": 187
},
{
"epoch": 1.9583333333333335,
"grad_norm": 2.0503403694456911e-22,
"learning_rate": 2.382727698752474e-07,
"logits/chosen": -0.8970575332641602,
"logits/rejected": -1.145188570022583,
"logps/chosen": -2160.99365234375,
"logps/rejected": -1063.6136474609375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -8.6992769241333,
"rewards/margins": 72.37908935546875,
"rewards/rejected": -81.07836151123047,
"step": 188
},
{
"epoch": 1.96875,
"grad_norm": 1.2547314776381705e-16,
"learning_rate": 1.340517319543877e-07,
"logits/chosen": -0.8810423612594604,
"logits/rejected": -1.099363088607788,
"logps/chosen": -2089.138671875,
"logps/rejected": -1003.574951171875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -11.649038314819336,
"rewards/margins": 64.49651336669922,
"rewards/rejected": -76.14555358886719,
"step": 189
},
{
"epoch": 1.9791666666666665,
"grad_norm": 0.0,
"learning_rate": 5.958594489295921e-08,
"logits/chosen": -0.9419525861740112,
"logits/rejected": -1.2568933963775635,
"logps/chosen": -3585.5712890625,
"logps/rejected": -979.148193359375,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -2.2342803478240967,
"rewards/margins": 71.78946685791016,
"rewards/rejected": -74.02375030517578,
"step": 190
},
{
"epoch": 1.9895833333333335,
"grad_norm": 0.0,
"learning_rate": 1.4897595915053242e-08,
"logits/chosen": -0.9514296054840088,
"logits/rejected": -1.1773942708969116,
"logps/chosen": -3752.94091796875,
"logps/rejected": -1072.39013671875,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -3.917599678039551,
"rewards/margins": 77.99908447265625,
"rewards/rejected": -81.91668701171875,
"step": 191
},
{
"epoch": 2.0,
"grad_norm": 5.172685426425421e-20,
"learning_rate": 0.0,
"logits/chosen": -0.8919022083282471,
"logits/rejected": -1.0698691606521606,
"logps/chosen": -1968.748046875,
"logps/rejected": -921.0995483398438,
"loss": 0.0,
"rewards/accuracies": 1.0,
"rewards/chosen": -8.440521240234375,
"rewards/margins": 60.94007110595703,
"rewards/rejected": -69.3805923461914,
"step": 192
}
],
"logging_steps": 1,
"max_steps": 192,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 96,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}