zephyr-7b-gpo-v3-2-i2 / trainer_state.json
lole25's picture
Model save
0ee6071 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": -1.5177760124206543,
"logits/rejected": -1.1611042022705078,
"logps/chosen": -309.02911376953125,
"logps/rejected": -848.8409423828125,
"loss": 0.2593,
"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": -1.5949721336364746,
"logits/rejected": -1.165027379989624,
"logps/chosen": -451.9952697753906,
"logps/rejected": -786.9351806640625,
"loss": 0.2269,
"rewards/accuracies": 0.2777777910232544,
"rewards/chosen": -0.0005774286109954119,
"rewards/margins": -0.000440336880274117,
"rewards/rejected": -0.00013709173072129488,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 2.666666666666667e-07,
"logits/chosen": -1.557755470275879,
"logits/rejected": -1.188490629196167,
"logps/chosen": -457.94342041015625,
"logps/rejected": -653.2659912109375,
"loss": 0.2143,
"rewards/accuracies": 0.5249999761581421,
"rewards/chosen": 0.0007440468179993331,
"rewards/margins": 0.000692694156896323,
"rewards/rejected": 5.135267929290421e-05,
"step": 20
},
{
"epoch": 0.01,
"learning_rate": 4.0000000000000003e-07,
"logits/chosen": -1.8446645736694336,
"logits/rejected": -1.1848350763320923,
"logps/chosen": -542.8465576171875,
"logps/rejected": -840.0565185546875,
"loss": 0.1931,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": 0.001833869144320488,
"rewards/margins": 0.002176427748054266,
"rewards/rejected": -0.000342558283591643,
"step": 30
},
{
"epoch": 0.01,
"learning_rate": 5.333333333333335e-07,
"logits/chosen": -1.3947551250457764,
"logits/rejected": -1.0338518619537354,
"logps/chosen": -513.7941284179688,
"logps/rejected": -851.7515869140625,
"loss": 0.1978,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.0025823800824582577,
"rewards/margins": 0.007740010507404804,
"rewards/rejected": -0.005157629959285259,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 6.666666666666667e-07,
"logits/chosen": -1.544039249420166,
"logits/rejected": -1.1640592813491821,
"logps/chosen": -444.1375427246094,
"logps/rejected": -751.6806640625,
"loss": 0.1943,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.006201503798365593,
"rewards/margins": 0.012399530969560146,
"rewards/rejected": -0.006198027171194553,
"step": 50
},
{
"epoch": 0.02,
"learning_rate": 8.000000000000001e-07,
"logits/chosen": -1.7549737691879272,
"logits/rejected": -1.134901523590088,
"logps/chosen": -524.1077880859375,
"logps/rejected": -897.7693481445312,
"loss": 0.1816,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.010963965207338333,
"rewards/margins": 0.028727427124977112,
"rewards/rejected": -0.01776346191763878,
"step": 60
},
{
"epoch": 0.02,
"learning_rate": 9.333333333333334e-07,
"logits/chosen": -1.707073450088501,
"logits/rejected": -0.9422414898872375,
"logps/chosen": -418.4791564941406,
"logps/rejected": -887.1856689453125,
"loss": 0.166,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.011874416843056679,
"rewards/margins": 0.05441901832818985,
"rewards/rejected": -0.042544592171907425,
"step": 70
},
{
"epoch": 0.02,
"learning_rate": 1.066666666666667e-06,
"logits/chosen": -1.7464405298233032,
"logits/rejected": -1.3240439891815186,
"logps/chosen": -393.67059326171875,
"logps/rejected": -797.724609375,
"loss": 0.1765,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": 0.014196820557117462,
"rewards/margins": 0.05819591134786606,
"rewards/rejected": -0.043999094516038895,
"step": 80
},
{
"epoch": 0.02,
"learning_rate": 1.2000000000000002e-06,
"logits/chosen": -1.47357976436615,
"logits/rejected": -0.9985636472702026,
"logps/chosen": -445.72979736328125,
"logps/rejected": -712.3211059570312,
"loss": 0.1603,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": 0.003458250779658556,
"rewards/margins": 0.06728541851043701,
"rewards/rejected": -0.06382717192173004,
"step": 90
},
{
"epoch": 0.03,
"learning_rate": 1.3333333333333334e-06,
"logits/chosen": -1.5050207376480103,
"logits/rejected": -1.0137097835540771,
"logps/chosen": -409.77252197265625,
"logps/rejected": -882.01025390625,
"loss": 0.1755,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.011769925244152546,
"rewards/margins": 0.09516488015651703,
"rewards/rejected": -0.1069348007440567,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 1.4666666666666669e-06,
"logits/chosen": -1.508277177810669,
"logits/rejected": -0.9909588098526001,
"logps/chosen": -446.08514404296875,
"logps/rejected": -970.3818359375,
"loss": 0.1234,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.039314642548561096,
"rewards/margins": 0.12522205710411072,
"rewards/rejected": -0.16453669965267181,
"step": 110
},
{
"epoch": 0.03,
"learning_rate": 1.6000000000000001e-06,
"logits/chosen": -1.5001076459884644,
"logits/rejected": -1.1417639255523682,
"logps/chosen": -595.5568237304688,
"logps/rejected": -1075.8310546875,
"loss": 0.1178,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.11702040582895279,
"rewards/margins": 0.1300191581249237,
"rewards/rejected": -0.2470395863056183,
"step": 120
},
{
"epoch": 0.03,
"learning_rate": 1.7333333333333336e-06,
"logits/chosen": -1.6354900598526,
"logits/rejected": -1.1173183917999268,
"logps/chosen": -596.342041015625,
"logps/rejected": -1051.9910888671875,
"loss": 0.122,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1303187757730484,
"rewards/margins": 0.14471343159675598,
"rewards/rejected": -0.2750321924686432,
"step": 130
},
{
"epoch": 0.04,
"learning_rate": 1.8666666666666669e-06,
"logits/chosen": -1.857448935508728,
"logits/rejected": -1.1995853185653687,
"logps/chosen": -488.5594177246094,
"logps/rejected": -1067.145263671875,
"loss": 0.1382,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.11596964299678802,
"rewards/margins": 0.1565937101840973,
"rewards/rejected": -0.2725633978843689,
"step": 140
},
{
"epoch": 0.04,
"learning_rate": 2.0000000000000003e-06,
"logits/chosen": -1.6261777877807617,
"logits/rejected": -1.1077944040298462,
"logps/chosen": -564.3380737304688,
"logps/rejected": -1317.8497314453125,
"loss": 0.0805,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.14445793628692627,
"rewards/margins": 0.2764219641685486,
"rewards/rejected": -0.42087993025779724,
"step": 150
},
{
"epoch": 0.04,
"learning_rate": 2.133333333333334e-06,
"logits/chosen": -1.4915621280670166,
"logits/rejected": -1.1237232685089111,
"logps/chosen": -487.80487060546875,
"logps/rejected": -1048.151611328125,
"loss": 0.0999,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.14958631992340088,
"rewards/margins": 0.19907937943935394,
"rewards/rejected": -0.3486657440662384,
"step": 160
},
{
"epoch": 0.05,
"learning_rate": 2.266666666666667e-06,
"logits/chosen": -1.462720274925232,
"logits/rejected": -1.1023520231246948,
"logps/chosen": -552.3968505859375,
"logps/rejected": -1198.3912353515625,
"loss": 0.0924,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1677219718694687,
"rewards/margins": 0.21208930015563965,
"rewards/rejected": -0.37981128692626953,
"step": 170
},
{
"epoch": 0.05,
"learning_rate": 2.4000000000000003e-06,
"logits/chosen": -1.6897376775741577,
"logits/rejected": -0.9965440630912781,
"logps/chosen": -626.109619140625,
"logps/rejected": -1155.2620849609375,
"loss": 0.0705,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.15819820761680603,
"rewards/margins": 0.23869426548480988,
"rewards/rejected": -0.3968924880027771,
"step": 180
},
{
"epoch": 0.05,
"learning_rate": 2.5333333333333338e-06,
"logits/chosen": -1.8232837915420532,
"logits/rejected": -0.823337197303772,
"logps/chosen": -675.5474853515625,
"logps/rejected": -1195.9923095703125,
"loss": 0.1007,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12152433395385742,
"rewards/margins": 0.2081596404314041,
"rewards/rejected": -0.32968395948410034,
"step": 190
},
{
"epoch": 0.05,
"learning_rate": 2.666666666666667e-06,
"logits/chosen": -1.9476888179779053,
"logits/rejected": -1.0242823362350464,
"logps/chosen": -624.99951171875,
"logps/rejected": -1197.134521484375,
"loss": 0.0905,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12548482418060303,
"rewards/margins": 0.23341400921344757,
"rewards/rejected": -0.3588988482952118,
"step": 200
},
{
"epoch": 0.06,
"learning_rate": 2.8000000000000003e-06,
"logits/chosen": -1.518812894821167,
"logits/rejected": -1.046945571899414,
"logps/chosen": -764.122314453125,
"logps/rejected": -1280.217041015625,
"loss": 0.1437,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.23460201919078827,
"rewards/margins": 0.19938690960407257,
"rewards/rejected": -0.4339889585971832,
"step": 210
},
{
"epoch": 0.06,
"learning_rate": 2.9333333333333338e-06,
"logits/chosen": -1.6284335851669312,
"logits/rejected": -1.1508177518844604,
"logps/chosen": -523.3719482421875,
"logps/rejected": -1222.5562744140625,
"loss": 0.0616,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.14848622679710388,
"rewards/margins": 0.24298810958862305,
"rewards/rejected": -0.39147430658340454,
"step": 220
},
{
"epoch": 0.06,
"learning_rate": 3.066666666666667e-06,
"logits/chosen": -1.6719776391983032,
"logits/rejected": -0.9572904706001282,
"logps/chosen": -747.065185546875,
"logps/rejected": -1382.8427734375,
"loss": 0.0855,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.23779627680778503,
"rewards/margins": 0.2256799042224884,
"rewards/rejected": -0.4634762406349182,
"step": 230
},
{
"epoch": 0.06,
"learning_rate": 3.2000000000000003e-06,
"logits/chosen": -1.840018630027771,
"logits/rejected": -1.1092523336410522,
"logps/chosen": -612.7644653320312,
"logps/rejected": -1217.589111328125,
"loss": 0.0733,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.16975006461143494,
"rewards/margins": 0.2756669223308563,
"rewards/rejected": -0.44541701674461365,
"step": 240
},
{
"epoch": 0.07,
"learning_rate": 3.3333333333333333e-06,
"logits/chosen": -1.574760913848877,
"logits/rejected": -1.1205968856811523,
"logps/chosen": -629.0726318359375,
"logps/rejected": -1156.187255859375,
"loss": 0.0998,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13886727392673492,
"rewards/margins": 0.202668234705925,
"rewards/rejected": -0.3415355086326599,
"step": 250
},
{
"epoch": 0.07,
"learning_rate": 3.4666666666666672e-06,
"logits/chosen": -1.7936065196990967,
"logits/rejected": -0.9479697942733765,
"logps/chosen": -538.048583984375,
"logps/rejected": -1271.905029296875,
"loss": 0.063,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.10844652354717255,
"rewards/margins": 0.2883257269859314,
"rewards/rejected": -0.39677220582962036,
"step": 260
},
{
"epoch": 0.07,
"learning_rate": 3.6000000000000003e-06,
"logits/chosen": -1.7110416889190674,
"logits/rejected": -1.0976063013076782,
"logps/chosen": -601.929443359375,
"logps/rejected": -1087.264404296875,
"loss": 0.074,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.13733436167240143,
"rewards/margins": 0.2208215445280075,
"rewards/rejected": -0.35815590620040894,
"step": 270
},
{
"epoch": 0.07,
"learning_rate": 3.7333333333333337e-06,
"logits/chosen": -1.684920310974121,
"logits/rejected": -1.0355182886123657,
"logps/chosen": -502.43438720703125,
"logps/rejected": -1141.5850830078125,
"loss": 0.0975,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.10886510461568832,
"rewards/margins": 0.2565527558326721,
"rewards/rejected": -0.3654178977012634,
"step": 280
},
{
"epoch": 0.08,
"learning_rate": 3.866666666666667e-06,
"logits/chosen": -1.7269914150238037,
"logits/rejected": -0.9788557887077332,
"logps/chosen": -717.2500610351562,
"logps/rejected": -1336.142333984375,
"loss": 0.0838,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.23452234268188477,
"rewards/margins": 0.29266995191574097,
"rewards/rejected": -0.527192234992981,
"step": 290
},
{
"epoch": 0.08,
"learning_rate": 4.000000000000001e-06,
"logits/chosen": -1.7060441970825195,
"logits/rejected": -1.1475598812103271,
"logps/chosen": -730.9443969726562,
"logps/rejected": -1244.9676513671875,
"loss": 0.1341,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -0.255375474691391,
"rewards/margins": 0.19869598746299744,
"rewards/rejected": -0.4540714621543884,
"step": 300
},
{
"epoch": 0.08,
"learning_rate": 4.133333333333333e-06,
"logits/chosen": -1.7061704397201538,
"logits/rejected": -1.1434520483016968,
"logps/chosen": -705.079833984375,
"logps/rejected": -1187.599365234375,
"loss": 0.1019,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.13966450095176697,
"rewards/margins": 0.22368240356445312,
"rewards/rejected": -0.3633468747138977,
"step": 310
},
{
"epoch": 0.09,
"learning_rate": 4.266666666666668e-06,
"logits/chosen": -2.0864200592041016,
"logits/rejected": -1.0598593950271606,
"logps/chosen": -621.1461181640625,
"logps/rejected": -1218.3201904296875,
"loss": 0.076,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12446784973144531,
"rewards/margins": 0.285147488117218,
"rewards/rejected": -0.40961527824401855,
"step": 320
},
{
"epoch": 0.09,
"learning_rate": 4.4e-06,
"logits/chosen": -1.633683204650879,
"logits/rejected": -1.259135127067566,
"logps/chosen": -594.0384521484375,
"logps/rejected": -1040.244873046875,
"loss": 0.1407,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.13469278812408447,
"rewards/margins": 0.18073201179504395,
"rewards/rejected": -0.3154247999191284,
"step": 330
},
{
"epoch": 0.09,
"learning_rate": 4.533333333333334e-06,
"logits/chosen": -1.5713467597961426,
"logits/rejected": -1.1714986562728882,
"logps/chosen": -552.28515625,
"logps/rejected": -954.9786376953125,
"loss": 0.102,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.16208061575889587,
"rewards/margins": 0.17426642775535583,
"rewards/rejected": -0.3363470435142517,
"step": 340
},
{
"epoch": 0.09,
"learning_rate": 4.666666666666667e-06,
"logits/chosen": -1.383143424987793,
"logits/rejected": -1.1191000938415527,
"logps/chosen": -495.08245849609375,
"logps/rejected": -1157.8853759765625,
"loss": 0.0813,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1548539251089096,
"rewards/margins": 0.23339009284973145,
"rewards/rejected": -0.38824400305747986,
"step": 350
},
{
"epoch": 0.1,
"learning_rate": 4.800000000000001e-06,
"logits/chosen": -1.7334524393081665,
"logits/rejected": -1.072506070137024,
"logps/chosen": -742.59814453125,
"logps/rejected": -1228.1810302734375,
"loss": 0.128,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.15262000262737274,
"rewards/margins": 0.23166091740131378,
"rewards/rejected": -0.3842809200286865,
"step": 360
},
{
"epoch": 0.1,
"learning_rate": 4.933333333333334e-06,
"logits/chosen": -1.8477518558502197,
"logits/rejected": -1.3124161958694458,
"logps/chosen": -677.6192626953125,
"logps/rejected": -1154.7989501953125,
"loss": 0.1155,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.12368879467248917,
"rewards/margins": 0.20320534706115723,
"rewards/rejected": -0.32689422369003296,
"step": 370
},
{
"epoch": 0.1,
"learning_rate": 4.999972922944898e-06,
"logits/chosen": -1.8957774639129639,
"logits/rejected": -1.3899834156036377,
"logps/chosen": -547.0092163085938,
"logps/rejected": -1174.024658203125,
"loss": 0.0877,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.06287173926830292,
"rewards/margins": 0.2412436455488205,
"rewards/rejected": -0.304115355014801,
"step": 380
},
{
"epoch": 0.1,
"learning_rate": 4.999756310023261e-06,
"logits/chosen": -1.791933298110962,
"logits/rejected": -1.1157985925674438,
"logps/chosen": -516.9517211914062,
"logps/rejected": -1007.5950317382812,
"loss": 0.0993,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.06450649350881577,
"rewards/margins": 0.22859111428260803,
"rewards/rejected": -0.293097585439682,
"step": 390
},
{
"epoch": 0.11,
"learning_rate": 4.999323102948655e-06,
"logits/chosen": -1.6634111404418945,
"logits/rejected": -0.9339207410812378,
"logps/chosen": -541.6902465820312,
"logps/rejected": -1222.6890869140625,
"loss": 0.0786,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.11366814374923706,
"rewards/margins": 0.3344195783138275,
"rewards/rejected": -0.44808775186538696,
"step": 400
},
{
"epoch": 0.11,
"learning_rate": 4.998673339256785e-06,
"logits/chosen": -1.7227275371551514,
"logits/rejected": -1.4248206615447998,
"logps/chosen": -612.4100341796875,
"logps/rejected": -1344.530517578125,
"loss": 0.0903,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.16698965430259705,
"rewards/margins": 0.24890287220478058,
"rewards/rejected": -0.41589251160621643,
"step": 410
},
{
"epoch": 0.11,
"learning_rate": 4.997807075247147e-06,
"logits/chosen": -1.7536413669586182,
"logits/rejected": -1.3619048595428467,
"logps/chosen": -464.01678466796875,
"logps/rejected": -1163.01953125,
"loss": 0.0692,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.021548744291067123,
"rewards/margins": 0.24685168266296387,
"rewards/rejected": -0.2684004306793213,
"step": 420
},
{
"epoch": 0.11,
"learning_rate": 4.996724385978142e-06,
"logits/chosen": -2.028895139694214,
"logits/rejected": -1.2840532064437866,
"logps/chosen": -478.65533447265625,
"logps/rejected": -1112.1922607421875,
"loss": 0.0923,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.00986658688634634,
"rewards/margins": 0.23762169480323792,
"rewards/rejected": -0.24748826026916504,
"step": 430
},
{
"epoch": 0.12,
"learning_rate": 4.995425365260585e-06,
"logits/chosen": -1.9320755004882812,
"logits/rejected": -1.3463362455368042,
"logps/chosen": -468.8946838378906,
"logps/rejected": -1079.24951171875,
"loss": 0.0844,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.05752667784690857,
"rewards/margins": 0.25381767749786377,
"rewards/rejected": -0.31134432554244995,
"step": 440
},
{
"epoch": 0.12,
"learning_rate": 4.993910125649561e-06,
"logits/chosen": -1.8947498798370361,
"logits/rejected": -1.4494032859802246,
"logps/chosen": -612.285400390625,
"logps/rejected": -1175.610595703125,
"loss": 0.0961,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.20282471179962158,
"rewards/margins": 0.24255582690238953,
"rewards/rejected": -0.4453805387020111,
"step": 450
},
{
"epoch": 0.12,
"learning_rate": 4.992178798434684e-06,
"logits/chosen": -1.8909610509872437,
"logits/rejected": -1.2547855377197266,
"logps/chosen": -703.7716064453125,
"logps/rejected": -1249.4072265625,
"loss": 0.0749,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.18841485679149628,
"rewards/margins": 0.24599456787109375,
"rewards/rejected": -0.4344094693660736,
"step": 460
},
{
"epoch": 0.13,
"learning_rate": 4.990231533628719e-06,
"logits/chosen": -1.9798517227172852,
"logits/rejected": -1.4687498807907104,
"logps/chosen": -451.41192626953125,
"logps/rejected": -1110.53466796875,
"loss": 0.0835,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.05847315117716789,
"rewards/margins": 0.2581016421318054,
"rewards/rejected": -0.3165748119354248,
"step": 470
},
{
"epoch": 0.13,
"learning_rate": 4.988068499954578e-06,
"logits/chosen": -1.9682140350341797,
"logits/rejected": -1.161273717880249,
"logps/chosen": -534.8321533203125,
"logps/rejected": -1069.1895751953125,
"loss": 0.0861,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.08595879375934601,
"rewards/margins": 0.2487233430147171,
"rewards/rejected": -0.3346821367740631,
"step": 480
},
{
"epoch": 0.13,
"learning_rate": 4.985689884830711e-06,
"logits/chosen": -1.800855040550232,
"logits/rejected": -1.0105341672897339,
"logps/chosen": -645.858154296875,
"logps/rejected": -1335.704833984375,
"loss": 0.06,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.17263874411582947,
"rewards/margins": 0.28426748514175415,
"rewards/rejected": -0.4569062292575836,
"step": 490
},
{
"epoch": 0.13,
"learning_rate": 4.983095894354858e-06,
"logits/chosen": -1.845642328262329,
"logits/rejected": -1.124245285987854,
"logps/chosen": -684.5682373046875,
"logps/rejected": -1378.302490234375,
"loss": 0.0689,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19359903037548065,
"rewards/margins": 0.3165056109428406,
"rewards/rejected": -0.5101046562194824,
"step": 500
},
{
"epoch": 0.14,
"learning_rate": 4.980286753286196e-06,
"logits/chosen": -1.7543941736221313,
"logits/rejected": -1.139478087425232,
"logps/chosen": -591.703857421875,
"logps/rejected": -1196.9752197265625,
"loss": 0.0752,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.17276380956172943,
"rewards/margins": 0.274630606174469,
"rewards/rejected": -0.44739437103271484,
"step": 510
},
{
"epoch": 0.14,
"learning_rate": 4.97726270502586e-06,
"logits/chosen": -1.7981364727020264,
"logits/rejected": -1.2439225912094116,
"logps/chosen": -622.7135620117188,
"logps/rejected": -1289.652587890625,
"loss": 0.0642,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.16895990073680878,
"rewards/margins": 0.27986788749694824,
"rewards/rejected": -0.4488278329372406,
"step": 520
},
{
"epoch": 0.14,
"learning_rate": 4.974024011595864e-06,
"logits/chosen": -1.7756052017211914,
"logits/rejected": -1.273600697517395,
"logps/chosen": -779.8721923828125,
"logps/rejected": -1319.716064453125,
"loss": 0.0889,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.19681531190872192,
"rewards/margins": 0.24762988090515137,
"rewards/rejected": -0.4444451928138733,
"step": 530
},
{
"epoch": 0.14,
"learning_rate": 4.970570953616383e-06,
"logits/chosen": -1.7580140829086304,
"logits/rejected": -1.2909038066864014,
"logps/chosen": -611.9743041992188,
"logps/rejected": -1292.6343994140625,
"loss": 0.0655,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14458617568016052,
"rewards/margins": 0.3040325939655304,
"rewards/rejected": -0.44861873984336853,
"step": 540
},
{
"epoch": 0.15,
"learning_rate": 4.966903830281449e-06,
"logits/chosen": -2.1057114601135254,
"logits/rejected": -1.096806287765503,
"logps/chosen": -525.5113525390625,
"logps/rejected": -1102.981689453125,
"loss": 0.0727,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.09377844631671906,
"rewards/margins": 0.2827639877796173,
"rewards/rejected": -0.3765423893928528,
"step": 550
},
{
"epoch": 0.15,
"learning_rate": 4.9630229593330226e-06,
"logits/chosen": -1.8463417291641235,
"logits/rejected": -1.2035672664642334,
"logps/chosen": -658.9556884765625,
"logps/rejected": -1261.522705078125,
"loss": 0.0586,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1989162266254425,
"rewards/margins": 0.25805556774139404,
"rewards/rejected": -0.45697179436683655,
"step": 560
},
{
"epoch": 0.15,
"learning_rate": 4.958928677033465e-06,
"logits/chosen": -1.9065221548080444,
"logits/rejected": -1.3213990926742554,
"logps/chosen": -802.2817993164062,
"logps/rejected": -1280.823486328125,
"loss": 0.1215,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.2689030170440674,
"rewards/margins": 0.23530542850494385,
"rewards/rejected": -0.5042084455490112,
"step": 570
},
{
"epoch": 0.15,
"learning_rate": 4.954621338136399e-06,
"logits/chosen": -1.796940803527832,
"logits/rejected": -1.1448333263397217,
"logps/chosen": -800.59619140625,
"logps/rejected": -1366.584716796875,
"loss": 0.0668,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.21505789458751678,
"rewards/margins": 0.30636119842529297,
"rewards/rejected": -0.5214190483093262,
"step": 580
},
{
"epoch": 0.16,
"learning_rate": 4.95010131585597e-06,
"logits/chosen": -1.6272817850112915,
"logits/rejected": -0.8004865646362305,
"logps/chosen": -684.5340576171875,
"logps/rejected": -1158.3621826171875,
"loss": 0.1052,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.14372439682483673,
"rewards/margins": 0.24541731178760529,
"rewards/rejected": -0.389141708612442,
"step": 590
},
{
"epoch": 0.16,
"learning_rate": 4.9453690018345144e-06,
"logits/chosen": -1.791548490524292,
"logits/rejected": -1.2827670574188232,
"logps/chosen": -447.36956787109375,
"logps/rejected": -1077.491943359375,
"loss": 0.0668,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.08956549316644669,
"rewards/margins": 0.26021477580070496,
"rewards/rejected": -0.34978026151657104,
"step": 600
},
{
"epoch": 0.16,
"learning_rate": 4.940424806108619e-06,
"logits/chosen": -1.989205002784729,
"logits/rejected": -1.3596338033676147,
"logps/chosen": -683.2862548828125,
"logps/rejected": -1159.7490234375,
"loss": 0.1118,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.177694171667099,
"rewards/margins": 0.19292449951171875,
"rewards/rejected": -0.37061864137649536,
"step": 610
},
{
"epoch": 0.17,
"learning_rate": 4.935269157073597e-06,
"logits/chosen": -1.8252586126327515,
"logits/rejected": -1.4199360609054565,
"logps/chosen": -499.2095642089844,
"logps/rejected": -1207.395263671875,
"loss": 0.0594,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09857722371816635,
"rewards/margins": 0.3041486144065857,
"rewards/rejected": -0.40272584557533264,
"step": 620
},
{
"epoch": 0.17,
"learning_rate": 4.9299025014463665e-06,
"logits/chosen": -1.646740198135376,
"logits/rejected": -0.9128702878952026,
"logps/chosen": -546.8827514648438,
"logps/rejected": -1260.8980712890625,
"loss": 0.0639,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.10494896024465561,
"rewards/margins": 0.3301486670970917,
"rewards/rejected": -0.4350976347923279,
"step": 630
},
{
"epoch": 0.17,
"learning_rate": 4.924325304226745e-06,
"logits/chosen": -1.8604828119277954,
"logits/rejected": -1.0761983394622803,
"logps/chosen": -693.2246704101562,
"logps/rejected": -1271.934326171875,
"loss": 0.0518,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19305351376533508,
"rewards/margins": 0.2916993200778961,
"rewards/rejected": -0.4847528040409088,
"step": 640
},
{
"epoch": 0.17,
"learning_rate": 4.91853804865716e-06,
"logits/chosen": -1.38680100440979,
"logits/rejected": -0.9548083543777466,
"logps/chosen": -604.3922729492188,
"logps/rejected": -1208.889404296875,
"loss": 0.09,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.18794074654579163,
"rewards/margins": 0.24451354146003723,
"rewards/rejected": -0.43245425820350647,
"step": 650
},
{
"epoch": 0.18,
"learning_rate": 4.912541236180779e-06,
"logits/chosen": -1.663114309310913,
"logits/rejected": -1.3176881074905396,
"logps/chosen": -644.1502685546875,
"logps/rejected": -1255.1209716796875,
"loss": 0.107,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.18079988658428192,
"rewards/margins": 0.23088447749614716,
"rewards/rejected": -0.4116843640804291,
"step": 660
},
{
"epoch": 0.18,
"learning_rate": 4.9063353863980565e-06,
"logits/chosen": -1.651368498802185,
"logits/rejected": -1.0333479642868042,
"logps/chosen": -622.9571533203125,
"logps/rejected": -1196.442138671875,
"loss": 0.1069,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.13679654896259308,
"rewards/margins": 0.2643422484397888,
"rewards/rejected": -0.4011387825012207,
"step": 670
},
{
"epoch": 0.18,
"learning_rate": 4.899921037021719e-06,
"logits/chosen": -1.8946233987808228,
"logits/rejected": -1.4026496410369873,
"logps/chosen": -608.56005859375,
"logps/rejected": -1102.85107421875,
"loss": 0.1145,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.18631377816200256,
"rewards/margins": 0.20509126782417297,
"rewards/rejected": -0.39140504598617554,
"step": 680
},
{
"epoch": 0.18,
"learning_rate": 4.893298743830168e-06,
"logits/chosen": -1.664089560508728,
"logits/rejected": -1.1638383865356445,
"logps/chosen": -681.1607055664062,
"logps/rejected": -1325.0721435546875,
"loss": 0.0936,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.213485985994339,
"rewards/margins": 0.2793218493461609,
"rewards/rejected": -0.4928078055381775,
"step": 690
},
{
"epoch": 0.19,
"learning_rate": 4.88646908061933e-06,
"logits/chosen": -1.7145121097564697,
"logits/rejected": -1.0264990329742432,
"logps/chosen": -640.181396484375,
"logps/rejected": -1224.787841796875,
"loss": 0.0784,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20638792216777802,
"rewards/margins": 0.2602555751800537,
"rewards/rejected": -0.46664348244667053,
"step": 700
},
{
"epoch": 0.19,
"learning_rate": 4.879432639152935e-06,
"logits/chosen": -1.5630238056182861,
"logits/rejected": -0.8563889265060425,
"logps/chosen": -688.0330810546875,
"logps/rejected": -1333.184326171875,
"loss": 0.087,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20561043918132782,
"rewards/margins": 0.285735547542572,
"rewards/rejected": -0.4913460314273834,
"step": 710
},
{
"epoch": 0.19,
"learning_rate": 4.8721900291112415e-06,
"logits/chosen": -1.7675155401229858,
"logits/rejected": -1.3769733905792236,
"logps/chosen": -606.588134765625,
"logps/rejected": -1178.3841552734375,
"loss": 0.0998,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.10641028732061386,
"rewards/margins": 0.26885437965393066,
"rewards/rejected": -0.3752647042274475,
"step": 720
},
{
"epoch": 0.19,
"learning_rate": 4.864741878038218e-06,
"logits/chosen": -1.5133328437805176,
"logits/rejected": -1.1089714765548706,
"logps/chosen": -560.0144653320312,
"logps/rejected": -1235.3402099609375,
"loss": 0.085,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.12044046074151993,
"rewards/margins": 0.26784801483154297,
"rewards/rejected": -0.3882884979248047,
"step": 730
},
{
"epoch": 0.2,
"learning_rate": 4.857088831287158e-06,
"logits/chosen": -1.8833305835723877,
"logits/rejected": -1.1213592290878296,
"logps/chosen": -614.7459716796875,
"logps/rejected": -1231.1275634765625,
"loss": 0.0612,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.14084835350513458,
"rewards/margins": 0.25714007019996643,
"rewards/rejected": -0.3979884088039398,
"step": 740
},
{
"epoch": 0.2,
"learning_rate": 4.849231551964771e-06,
"logits/chosen": -1.6487013101577759,
"logits/rejected": -1.0574976205825806,
"logps/chosen": -638.1517333984375,
"logps/rejected": -1221.0208740234375,
"loss": 0.0773,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.15978939831256866,
"rewards/margins": 0.2666565477848053,
"rewards/rejected": -0.4264459013938904,
"step": 750
},
{
"epoch": 0.2,
"learning_rate": 4.841170720873723e-06,
"logits/chosen": -1.6006218194961548,
"logits/rejected": -1.0765124559402466,
"logps/chosen": -648.8009643554688,
"logps/rejected": -1193.5694580078125,
"loss": 0.0903,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20380504429340363,
"rewards/margins": 0.2502138018608093,
"rewards/rejected": -0.45401889085769653,
"step": 760
},
{
"epoch": 0.21,
"learning_rate": 4.832907036453647e-06,
"logits/chosen": -1.6672321557998657,
"logits/rejected": -1.2181921005249023,
"logps/chosen": -757.1592407226562,
"logps/rejected": -1409.3095703125,
"loss": 0.0809,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.24005849659442902,
"rewards/margins": 0.2847335636615753,
"rewards/rejected": -0.5247920155525208,
"step": 770
},
{
"epoch": 0.21,
"learning_rate": 4.824441214720629e-06,
"logits/chosen": -1.3765209913253784,
"logits/rejected": -1.0029934644699097,
"logps/chosen": -498.0054626464844,
"logps/rejected": -1238.66943359375,
"loss": 0.0667,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15500691533088684,
"rewards/margins": 0.3080350160598755,
"rewards/rejected": -0.4630419611930847,
"step": 780
},
{
"epoch": 0.21,
"learning_rate": 4.815773989205165e-06,
"logits/chosen": -1.7891244888305664,
"logits/rejected": -1.2491934299468994,
"logps/chosen": -556.8863525390625,
"logps/rejected": -1332.5364990234375,
"loss": 0.0617,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13201333582401276,
"rewards/margins": 0.3241254687309265,
"rewards/rejected": -0.4561387896537781,
"step": 790
},
{
"epoch": 0.21,
"learning_rate": 4.806906110888606e-06,
"logits/chosen": -1.6729551553726196,
"logits/rejected": -1.187744379043579,
"logps/chosen": -529.7684936523438,
"logps/rejected": -1193.1453857421875,
"loss": 0.0791,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1251310557126999,
"rewards/margins": 0.2839392423629761,
"rewards/rejected": -0.4090702533721924,
"step": 800
},
{
"epoch": 0.22,
"learning_rate": 4.7978383481380865e-06,
"logits/chosen": -1.6085048913955688,
"logits/rejected": -1.2299137115478516,
"logps/chosen": -579.7472534179688,
"logps/rejected": -1075.484130859375,
"loss": 0.0994,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.15328553318977356,
"rewards/margins": 0.2056526243686676,
"rewards/rejected": -0.3589381277561188,
"step": 810
},
{
"epoch": 0.22,
"learning_rate": 4.788571486639948e-06,
"logits/chosen": -1.4437693357467651,
"logits/rejected": -0.9531752467155457,
"logps/chosen": -721.705078125,
"logps/rejected": -1422.717529296875,
"loss": 0.0707,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.19354596734046936,
"rewards/margins": 0.2650890648365021,
"rewards/rejected": -0.45863503217697144,
"step": 820
},
{
"epoch": 0.22,
"learning_rate": 4.779106329331665e-06,
"logits/chosen": -1.7850446701049805,
"logits/rejected": -1.1975328922271729,
"logps/chosen": -639.4754028320312,
"logps/rejected": -1183.6280517578125,
"loss": 0.1136,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.18970054388046265,
"rewards/margins": 0.22527900338172913,
"rewards/rejected": -0.4149795472621918,
"step": 830
},
{
"epoch": 0.22,
"learning_rate": 4.769443696332272e-06,
"logits/chosen": -1.6451295614242554,
"logits/rejected": -0.9056906700134277,
"logps/chosen": -704.8148193359375,
"logps/rejected": -1420.8524169921875,
"loss": 0.0473,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.17515410482883453,
"rewards/margins": 0.3173523545265198,
"rewards/rejected": -0.4925064444541931,
"step": 840
},
{
"epoch": 0.23,
"learning_rate": 4.759584424871302e-06,
"logits/chosen": -1.5979700088500977,
"logits/rejected": -1.1091766357421875,
"logps/chosen": -686.29052734375,
"logps/rejected": -1212.19580078125,
"loss": 0.1128,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.15797309577465057,
"rewards/margins": 0.21523161232471466,
"rewards/rejected": -0.3732047379016876,
"step": 850
},
{
"epoch": 0.23,
"learning_rate": 4.749529369216246e-06,
"logits/chosen": -1.7736440896987915,
"logits/rejected": -1.172586441040039,
"logps/chosen": -660.5985107421875,
"logps/rejected": -1275.8603515625,
"loss": 0.0675,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.12020470947027206,
"rewards/margins": 0.2714688181877136,
"rewards/rejected": -0.39167362451553345,
"step": 860
},
{
"epoch": 0.23,
"learning_rate": 4.7392794005985324e-06,
"logits/chosen": -1.8436565399169922,
"logits/rejected": -1.348578929901123,
"logps/chosen": -471.02703857421875,
"logps/rejected": -1074.3665771484375,
"loss": 0.0923,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.061982929706573486,
"rewards/margins": 0.23231768608093262,
"rewards/rejected": -0.2943006157875061,
"step": 870
},
{
"epoch": 0.23,
"learning_rate": 4.7288354071380415e-06,
"logits/chosen": -1.6346375942230225,
"logits/rejected": -1.053733468055725,
"logps/chosen": -518.6777954101562,
"logps/rejected": -1128.2366943359375,
"loss": 0.0928,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.11276010423898697,
"rewards/margins": 0.24085621535778046,
"rewards/rejected": -0.35361629724502563,
"step": 880
},
{
"epoch": 0.24,
"learning_rate": 4.7181982937661485e-06,
"logits/chosen": -1.5796483755111694,
"logits/rejected": -0.9548759460449219,
"logps/chosen": -694.7703857421875,
"logps/rejected": -1379.165771484375,
"loss": 0.0694,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.21314816176891327,
"rewards/margins": 0.28819718956947327,
"rewards/rejected": -0.5013453364372253,
"step": 890
},
{
"epoch": 0.24,
"learning_rate": 4.707368982147318e-06,
"logits/chosen": -1.6242942810058594,
"logits/rejected": -1.100235104560852,
"logps/chosen": -656.1912841796875,
"logps/rejected": -1320.0205078125,
"loss": 0.0818,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.20513398945331573,
"rewards/margins": 0.2904512286186218,
"rewards/rejected": -0.49558526277542114,
"step": 900
},
{
"epoch": 0.24,
"learning_rate": 4.696348410599244e-06,
"logits/chosen": -1.6266180276870728,
"logits/rejected": -1.112475872039795,
"logps/chosen": -615.5531005859375,
"logps/rejected": -1176.023681640625,
"loss": 0.1074,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.21413663029670715,
"rewards/margins": 0.24315333366394043,
"rewards/rejected": -0.45728999376296997,
"step": 910
},
{
"epoch": 0.25,
"learning_rate": 4.685137534011549e-06,
"logits/chosen": -1.8083276748657227,
"logits/rejected": -0.9349889755249023,
"logps/chosen": -659.9114990234375,
"logps/rejected": -1210.7021484375,
"loss": 0.0749,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1867416799068451,
"rewards/margins": 0.27049291133880615,
"rewards/rejected": -0.45723456144332886,
"step": 920
},
{
"epoch": 0.25,
"learning_rate": 4.673737323763048e-06,
"logits/chosen": -1.5211584568023682,
"logits/rejected": -0.8713824152946472,
"logps/chosen": -557.5753784179688,
"logps/rejected": -1095.0653076171875,
"loss": 0.1147,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1781322956085205,
"rewards/margins": 0.23243267834186554,
"rewards/rejected": -0.41056495904922485,
"step": 930
},
{
"epoch": 0.25,
"learning_rate": 4.662148767637578e-06,
"logits/chosen": -1.6907579898834229,
"logits/rejected": -0.92974454164505,
"logps/chosen": -672.3154907226562,
"logps/rejected": -1263.9849853515625,
"loss": 0.0513,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.14491698145866394,
"rewards/margins": 0.2988959848880768,
"rewards/rejected": -0.44381293654441833,
"step": 940
},
{
"epoch": 0.25,
"learning_rate": 4.650372869738415e-06,
"logits/chosen": -1.713449239730835,
"logits/rejected": -1.046890139579773,
"logps/chosen": -656.3117065429688,
"logps/rejected": -1259.781982421875,
"loss": 0.0727,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1550595909357071,
"rewards/margins": 0.27939510345458984,
"rewards/rejected": -0.43445467948913574,
"step": 950
},
{
"epoch": 0.26,
"learning_rate": 4.638410650401267e-06,
"logits/chosen": -1.6114223003387451,
"logits/rejected": -0.8841923475265503,
"logps/chosen": -619.8568725585938,
"logps/rejected": -1200.001953125,
"loss": 0.092,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.18295657634735107,
"rewards/margins": 0.27934443950653076,
"rewards/rejected": -0.46230101585388184,
"step": 960
},
{
"epoch": 0.26,
"learning_rate": 4.626263146105875e-06,
"logits/chosen": -1.8829456567764282,
"logits/rejected": -1.1874229907989502,
"logps/chosen": -612.0319213867188,
"logps/rejected": -1310.8118896484375,
"loss": 0.0626,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15665313601493835,
"rewards/margins": 0.32128262519836426,
"rewards/rejected": -0.4779357314109802,
"step": 970
},
{
"epoch": 0.26,
"learning_rate": 4.613931409386196e-06,
"logits/chosen": -1.7148971557617188,
"logits/rejected": -1.1706236600875854,
"logps/chosen": -651.0546875,
"logps/rejected": -1184.420654296875,
"loss": 0.0975,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.15911589562892914,
"rewards/margins": 0.26245635747909546,
"rewards/rejected": -0.4215722680091858,
"step": 980
},
{
"epoch": 0.26,
"learning_rate": 4.601416508739211e-06,
"logits/chosen": -1.7349941730499268,
"logits/rejected": -1.0320522785186768,
"logps/chosen": -641.8563842773438,
"logps/rejected": -1191.799072265625,
"loss": 0.0864,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12740136682987213,
"rewards/margins": 0.27434709668159485,
"rewards/rejected": -0.4017484784126282,
"step": 990
},
{
"epoch": 0.27,
"learning_rate": 4.588719528532342e-06,
"logits/chosen": -1.7186208963394165,
"logits/rejected": -1.139203429222107,
"logps/chosen": -608.3776245117188,
"logps/rejected": -1194.766845703125,
"loss": 0.0803,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12015128135681152,
"rewards/margins": 0.28770390152931213,
"rewards/rejected": -0.4078551232814789,
"step": 1000
},
{
"epoch": 0.27,
"learning_rate": 4.575841568909494e-06,
"logits/chosen": -1.9281619787216187,
"logits/rejected": -1.021177887916565,
"logps/chosen": -606.2479248046875,
"logps/rejected": -1277.761474609375,
"loss": 0.0629,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12076227366924286,
"rewards/margins": 0.31744498014450073,
"rewards/rejected": -0.4382072389125824,
"step": 1010
},
{
"epoch": 0.27,
"learning_rate": 4.562783745695738e-06,
"logits/chosen": -1.772539734840393,
"logits/rejected": -1.1384470462799072,
"logps/chosen": -584.2857666015625,
"logps/rejected": -1120.206787109375,
"loss": 0.0924,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.10691050440073013,
"rewards/margins": 0.25287091732025146,
"rewards/rejected": -0.359781414270401,
"step": 1020
},
{
"epoch": 0.27,
"learning_rate": 4.549547190300622e-06,
"logits/chosen": -1.6101019382476807,
"logits/rejected": -0.995293915271759,
"logps/chosen": -663.1490478515625,
"logps/rejected": -1269.482666015625,
"loss": 0.0903,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.19160196185112,
"rewards/margins": 0.2655636966228485,
"rewards/rejected": -0.4571656584739685,
"step": 1030
},
{
"epoch": 0.28,
"learning_rate": 4.536133049620143e-06,
"logits/chosen": -1.6123756170272827,
"logits/rejected": -0.8887365460395813,
"logps/chosen": -645.6787719726562,
"logps/rejected": -1266.5006103515625,
"loss": 0.0808,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.13368944823741913,
"rewards/margins": 0.28590840101242065,
"rewards/rejected": -0.4195978045463562,
"step": 1040
},
{
"epoch": 0.28,
"learning_rate": 4.522542485937369e-06,
"logits/chosen": -1.7597821950912476,
"logits/rejected": -1.0739284753799438,
"logps/chosen": -670.746337890625,
"logps/rejected": -1303.915771484375,
"loss": 0.0671,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15525345504283905,
"rewards/margins": 0.28055456280708313,
"rewards/rejected": -0.435808002948761,
"step": 1050
},
{
"epoch": 0.28,
"learning_rate": 4.508776676821739e-06,
"logits/chosen": -1.6723902225494385,
"logits/rejected": -1.1365679502487183,
"logps/chosen": -659.1893310546875,
"logps/rejected": -1257.1162109375,
"loss": 0.0961,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.18045826256275177,
"rewards/margins": 0.28194642066955566,
"rewards/rejected": -0.46240463852882385,
"step": 1060
},
{
"epoch": 0.29,
"learning_rate": 4.494836815027022e-06,
"logits/chosen": -1.674168348312378,
"logits/rejected": -0.8119763135910034,
"logps/chosen": -675.2777099609375,
"logps/rejected": -1269.997802734375,
"loss": 0.1022,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2055848389863968,
"rewards/margins": 0.2812694013118744,
"rewards/rejected": -0.48685422539711,
"step": 1070
},
{
"epoch": 0.29,
"learning_rate": 4.4807241083879774e-06,
"logits/chosen": -1.5376824140548706,
"logits/rejected": -0.9127294421195984,
"logps/chosen": -633.2178344726562,
"logps/rejected": -1178.2454833984375,
"loss": 0.0838,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.18428584933280945,
"rewards/margins": 0.24417026340961456,
"rewards/rejected": -0.4284561276435852,
"step": 1080
},
{
"epoch": 0.29,
"learning_rate": 4.466439779715696e-06,
"logits/chosen": -1.4871512651443481,
"logits/rejected": -0.8935056924819946,
"logps/chosen": -606.7117309570312,
"logps/rejected": -1183.592041015625,
"loss": 0.0976,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.1557384431362152,
"rewards/margins": 0.2602311670780182,
"rewards/rejected": -0.4159695506095886,
"step": 1090
},
{
"epoch": 0.29,
"learning_rate": 4.451985066691649e-06,
"logits/chosen": -1.5563310384750366,
"logits/rejected": -1.0307669639587402,
"logps/chosen": -558.4649047851562,
"logps/rejected": -1085.895751953125,
"loss": 0.0918,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1063220277428627,
"rewards/margins": 0.26596084237098694,
"rewards/rejected": -0.37228289246559143,
"step": 1100
},
{
"epoch": 0.3,
"learning_rate": 4.437361221760449e-06,
"logits/chosen": -1.692957878112793,
"logits/rejected": -1.1781022548675537,
"logps/chosen": -514.2180786132812,
"logps/rejected": -1231.850341796875,
"loss": 0.0484,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.07757623493671417,
"rewards/margins": 0.30450260639190674,
"rewards/rejected": -0.3820788860321045,
"step": 1110
},
{
"epoch": 0.3,
"learning_rate": 4.422569512021332e-06,
"logits/chosen": -1.6798826456069946,
"logits/rejected": -1.1389153003692627,
"logps/chosen": -585.5572509765625,
"logps/rejected": -1191.422607421875,
"loss": 0.083,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.06024783104658127,
"rewards/margins": 0.2825511693954468,
"rewards/rejected": -0.34279894828796387,
"step": 1120
},
{
"epoch": 0.3,
"learning_rate": 4.407611219118363e-06,
"logits/chosen": -1.574668526649475,
"logits/rejected": -1.1745796203613281,
"logps/chosen": -573.3482666015625,
"logps/rejected": -1237.1324462890625,
"loss": 0.0851,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.14945417642593384,
"rewards/margins": 0.26941633224487305,
"rewards/rejected": -0.4188705384731293,
"step": 1130
},
{
"epoch": 0.3,
"learning_rate": 4.3924876391293915e-06,
"logits/chosen": -1.5106923580169678,
"logits/rejected": -1.0384470224380493,
"logps/chosen": -653.4393920898438,
"logps/rejected": -1249.9561767578125,
"loss": 0.0827,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.21140387654304504,
"rewards/margins": 0.26517191529273987,
"rewards/rejected": -0.4765757620334625,
"step": 1140
},
{
"epoch": 0.31,
"learning_rate": 4.377200082453748e-06,
"logits/chosen": -1.553951621055603,
"logits/rejected": -1.1727259159088135,
"logps/chosen": -584.9124145507812,
"logps/rejected": -1274.189453125,
"loss": 0.0708,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.17560866475105286,
"rewards/margins": 0.3020893931388855,
"rewards/rejected": -0.47769802808761597,
"step": 1150
},
{
"epoch": 0.31,
"learning_rate": 4.361749873698707e-06,
"logits/chosen": -1.6106698513031006,
"logits/rejected": -0.9710136651992798,
"logps/chosen": -683.4093017578125,
"logps/rejected": -1351.5721435546875,
"loss": 0.0537,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1945325881242752,
"rewards/margins": 0.3134706914424896,
"rewards/rejected": -0.5080032348632812,
"step": 1160
},
{
"epoch": 0.31,
"learning_rate": 4.346138351564711e-06,
"logits/chosen": -1.4298365116119385,
"logits/rejected": -0.9186006784439087,
"logps/chosen": -629.0538940429688,
"logps/rejected": -1190.7537841796875,
"loss": 0.1065,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1838223785161972,
"rewards/margins": 0.25950607657432556,
"rewards/rejected": -0.44332846999168396,
"step": 1170
},
{
"epoch": 0.31,
"learning_rate": 4.330366868729376e-06,
"logits/chosen": -1.5331902503967285,
"logits/rejected": -0.8202164769172668,
"logps/chosen": -719.1092529296875,
"logps/rejected": -1225.5712890625,
"loss": 0.1239,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.23043549060821533,
"rewards/margins": 0.24460101127624512,
"rewards/rejected": -0.47503647208213806,
"step": 1180
},
{
"epoch": 0.32,
"learning_rate": 4.3144367917302964e-06,
"logits/chosen": -1.5329627990722656,
"logits/rejected": -1.0226593017578125,
"logps/chosen": -558.7479248046875,
"logps/rejected": -1248.3746337890625,
"loss": 0.0732,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.17867226898670197,
"rewards/margins": 0.27305805683135986,
"rewards/rejected": -0.451730340719223,
"step": 1190
},
{
"epoch": 0.32,
"learning_rate": 4.2983495008466285e-06,
"logits/chosen": -1.4446375370025635,
"logits/rejected": -1.2019789218902588,
"logps/chosen": -629.9727172851562,
"logps/rejected": -1310.1802978515625,
"loss": 0.0701,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.19203418493270874,
"rewards/margins": 0.26955386996269226,
"rewards/rejected": -0.461588054895401,
"step": 1200
},
{
"epoch": 0.32,
"learning_rate": 4.2821063899795015e-06,
"logits/chosen": -1.5882418155670166,
"logits/rejected": -0.9619570970535278,
"logps/chosen": -633.544921875,
"logps/rejected": -1342.188720703125,
"loss": 0.0702,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15442287921905518,
"rewards/margins": 0.3408200740814209,
"rewards/rejected": -0.49524298310279846,
"step": 1210
},
{
"epoch": 0.33,
"learning_rate": 4.265708866531238e-06,
"logits/chosen": -1.6361687183380127,
"logits/rejected": -1.160875916481018,
"logps/chosen": -604.972412109375,
"logps/rejected": -1180.0421142578125,
"loss": 0.0912,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15669772028923035,
"rewards/margins": 0.2758365571498871,
"rewards/rejected": -0.43253427743911743,
"step": 1220
},
{
"epoch": 0.33,
"learning_rate": 4.249158351283414e-06,
"logits/chosen": -1.4963274002075195,
"logits/rejected": -1.0508579015731812,
"logps/chosen": -670.7274169921875,
"logps/rejected": -1422.38037109375,
"loss": 0.0547,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.18853971362113953,
"rewards/margins": 0.3481212556362152,
"rewards/rejected": -0.5366610288619995,
"step": 1230
},
{
"epoch": 0.33,
"learning_rate": 4.232456278273743e-06,
"logits/chosen": -1.5074360370635986,
"logits/rejected": -1.1942028999328613,
"logps/chosen": -621.916015625,
"logps/rejected": -1097.22705078125,
"loss": 0.1119,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.20438826084136963,
"rewards/margins": 0.19910171627998352,
"rewards/rejected": -0.40348997712135315,
"step": 1240
},
{
"epoch": 0.33,
"learning_rate": 4.215604094671835e-06,
"logits/chosen": -1.7932571172714233,
"logits/rejected": -1.156964659690857,
"logps/chosen": -605.7149047851562,
"logps/rejected": -1170.5318603515625,
"loss": 0.0823,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.13441742956638336,
"rewards/margins": 0.26707369089126587,
"rewards/rejected": -0.4014911651611328,
"step": 1250
},
{
"epoch": 0.34,
"learning_rate": 4.198603260653792e-06,
"logits/chosen": -1.5639320611953735,
"logits/rejected": -1.011678695678711,
"logps/chosen": -570.6861572265625,
"logps/rejected": -1016.75390625,
"loss": 0.1338,
"rewards/accuracies": 0.6000000238418579,
"rewards/chosen": -0.09839320182800293,
"rewards/margins": 0.16072291135787964,
"rewards/rejected": -0.25911611318588257,
"step": 1260
},
{
"epoch": 0.34,
"learning_rate": 4.181455249275701e-06,
"logits/chosen": -1.3490540981292725,
"logits/rejected": -0.8460888862609863,
"logps/chosen": -604.7997436523438,
"logps/rejected": -1211.717529296875,
"loss": 0.0948,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.16084125638008118,
"rewards/margins": 0.24433521926403046,
"rewards/rejected": -0.40517646074295044,
"step": 1270
},
{
"epoch": 0.34,
"learning_rate": 4.1641615463459926e-06,
"logits/chosen": -1.4548208713531494,
"logits/rejected": -0.9700274467468262,
"logps/chosen": -662.6041259765625,
"logps/rejected": -1517.4892578125,
"loss": 0.043,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.16144345700740814,
"rewards/margins": 0.34403008222579956,
"rewards/rejected": -0.5054734945297241,
"step": 1280
},
{
"epoch": 0.34,
"learning_rate": 4.146723650296701e-06,
"logits/chosen": -1.5642893314361572,
"logits/rejected": -1.012499213218689,
"logps/chosen": -522.3748168945312,
"logps/rejected": -1114.373779296875,
"loss": 0.0893,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.11404868215322495,
"rewards/margins": 0.27995288372039795,
"rewards/rejected": -0.3940015733242035,
"step": 1290
},
{
"epoch": 0.35,
"learning_rate": 4.129143072053639e-06,
"logits/chosen": -1.8208191394805908,
"logits/rejected": -1.0344860553741455,
"logps/chosen": -717.9315185546875,
"logps/rejected": -1288.3831787109375,
"loss": 0.0814,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1353410929441452,
"rewards/margins": 0.2894688546657562,
"rewards/rejected": -0.42480993270874023,
"step": 1300
},
{
"epoch": 0.35,
"learning_rate": 4.111421334905468e-06,
"logits/chosen": -1.67376708984375,
"logits/rejected": -0.7968643307685852,
"logps/chosen": -653.73193359375,
"logps/rejected": -1237.788330078125,
"loss": 0.0799,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.10021106898784637,
"rewards/margins": 0.2965288758277893,
"rewards/rejected": -0.3967399299144745,
"step": 1310
},
{
"epoch": 0.35,
"learning_rate": 4.093559974371725e-06,
"logits/chosen": -1.5515210628509521,
"logits/rejected": -1.0218132734298706,
"logps/chosen": -771.3948974609375,
"logps/rejected": -1335.147705078125,
"loss": 0.0912,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.16722288727760315,
"rewards/margins": 0.2642812132835388,
"rewards/rejected": -0.4315040707588196,
"step": 1320
},
{
"epoch": 0.35,
"learning_rate": 4.075560538069767e-06,
"logits/chosen": -1.6814053058624268,
"logits/rejected": -1.175160527229309,
"logps/chosen": -561.907470703125,
"logps/rejected": -1060.85498046875,
"loss": 0.1004,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.06503216922283173,
"rewards/margins": 0.25370270013809204,
"rewards/rejected": -0.3187348544597626,
"step": 1330
},
{
"epoch": 0.36,
"learning_rate": 4.05742458558068e-06,
"logits/chosen": -1.8086637258529663,
"logits/rejected": -1.3191715478897095,
"logps/chosen": -529.0259399414062,
"logps/rejected": -1079.9012451171875,
"loss": 0.0936,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.0950629860162735,
"rewards/margins": 0.24811288714408875,
"rewards/rejected": -0.34317582845687866,
"step": 1340
},
{
"epoch": 0.36,
"learning_rate": 4.039153688314146e-06,
"logits/chosen": -1.6986970901489258,
"logits/rejected": -1.07126784324646,
"logps/chosen": -537.1072387695312,
"logps/rejected": -1259.658935546875,
"loss": 0.0665,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.07069256901741028,
"rewards/margins": 0.353458434343338,
"rewards/rejected": -0.4241510331630707,
"step": 1350
},
{
"epoch": 0.36,
"learning_rate": 4.020749429372286e-06,
"logits/chosen": -1.6649013757705688,
"logits/rejected": -1.1139470338821411,
"logps/chosen": -582.5765991210938,
"logps/rejected": -1291.3966064453125,
"loss": 0.0791,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08450852334499359,
"rewards/margins": 0.30345186591148376,
"rewards/rejected": -0.38796037435531616,
"step": 1360
},
{
"epoch": 0.37,
"learning_rate": 4.002213403412492e-06,
"logits/chosen": -1.6457083225250244,
"logits/rejected": -1.1551355123519897,
"logps/chosen": -443.95770263671875,
"logps/rejected": -1125.097412109375,
"loss": 0.0942,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.04620728641748428,
"rewards/margins": 0.29895836114883423,
"rewards/rejected": -0.3451656699180603,
"step": 1370
},
{
"epoch": 0.37,
"learning_rate": 3.983547216509254e-06,
"logits/chosen": -1.8475234508514404,
"logits/rejected": -0.9538863897323608,
"logps/chosen": -532.26171875,
"logps/rejected": -1173.6441650390625,
"loss": 0.0721,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07093639671802521,
"rewards/margins": 0.30083587765693665,
"rewards/rejected": -0.37177228927612305,
"step": 1380
},
{
"epoch": 0.37,
"learning_rate": 3.964752486015001e-06,
"logits/chosen": -1.702235460281372,
"logits/rejected": -1.0087345838546753,
"logps/chosen": -570.5701904296875,
"logps/rejected": -1237.5167236328125,
"loss": 0.0554,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.10062988102436066,
"rewards/margins": 0.29185742139816284,
"rewards/rejected": -0.3924873471260071,
"step": 1390
},
{
"epoch": 0.37,
"learning_rate": 3.945830840419966e-06,
"logits/chosen": -1.8062200546264648,
"logits/rejected": -1.2370173931121826,
"logps/chosen": -563.0610961914062,
"logps/rejected": -1202.4017333984375,
"loss": 0.1014,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.08197510987520218,
"rewards/margins": 0.284410685300827,
"rewards/rejected": -0.3663857579231262,
"step": 1400
},
{
"epoch": 0.38,
"learning_rate": 3.92678391921108e-06,
"logits/chosen": -1.685450792312622,
"logits/rejected": -1.268028974533081,
"logps/chosen": -474.2938537597656,
"logps/rejected": -1090.5147705078125,
"loss": 0.0983,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.0882890596985817,
"rewards/margins": 0.26188915967941284,
"rewards/rejected": -0.35017821192741394,
"step": 1410
},
{
"epoch": 0.38,
"learning_rate": 3.907613372729916e-06,
"logits/chosen": -1.5756769180297852,
"logits/rejected": -1.1073580980300903,
"logps/chosen": -508.64654541015625,
"logps/rejected": -1192.768798828125,
"loss": 0.0905,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11147113889455795,
"rewards/margins": 0.2684895396232605,
"rewards/rejected": -0.37996068596839905,
"step": 1420
},
{
"epoch": 0.38,
"learning_rate": 3.888320862029699e-06,
"logits/chosen": -1.970336675643921,
"logits/rejected": -1.1683611869812012,
"logps/chosen": -608.2377319335938,
"logps/rejected": -1272.260498046875,
"loss": 0.091,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.08651524782180786,
"rewards/margins": 0.3126547038555145,
"rewards/rejected": -0.399169921875,
"step": 1430
},
{
"epoch": 0.38,
"learning_rate": 3.868908058731376e-06,
"logits/chosen": -1.7017894983291626,
"logits/rejected": -0.9617505073547363,
"logps/chosen": -684.0576782226562,
"logps/rejected": -1245.509765625,
"loss": 0.0996,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.09601768106222153,
"rewards/margins": 0.29448553919792175,
"rewards/rejected": -0.3905032277107239,
"step": 1440
},
{
"epoch": 0.39,
"learning_rate": 3.849376644878783e-06,
"logits/chosen": -1.536409616470337,
"logits/rejected": -1.0880365371704102,
"logps/chosen": -533.0728149414062,
"logps/rejected": -1243.9488525390625,
"loss": 0.0928,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.07541215419769287,
"rewards/margins": 0.28547203540802,
"rewards/rejected": -0.3608841896057129,
"step": 1450
},
{
"epoch": 0.39,
"learning_rate": 3.829728312792895e-06,
"logits/chosen": -1.8492475748062134,
"logits/rejected": -1.256415605545044,
"logps/chosen": -434.9085388183594,
"logps/rejected": -1020.18798828125,
"loss": 0.0619,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.04833076149225235,
"rewards/margins": 0.2557259202003479,
"rewards/rejected": -0.30405664443969727,
"step": 1460
},
{
"epoch": 0.39,
"learning_rate": 3.8099647649251984e-06,
"logits/chosen": -1.5395265817642212,
"logits/rejected": -1.087846040725708,
"logps/chosen": -591.6812744140625,
"logps/rejected": -1221.6868896484375,
"loss": 0.0932,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.10671044886112213,
"rewards/margins": 0.2588750422000885,
"rewards/rejected": -0.36558544635772705,
"step": 1470
},
{
"epoch": 0.39,
"learning_rate": 3.790087713710179e-06,
"logits/chosen": -1.5241081714630127,
"logits/rejected": -0.8990601301193237,
"logps/chosen": -703.971923828125,
"logps/rejected": -1318.2982177734375,
"loss": 0.0708,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.16658341884613037,
"rewards/margins": 0.270012766122818,
"rewards/rejected": -0.436596155166626,
"step": 1480
},
{
"epoch": 0.4,
"learning_rate": 3.770098881416945e-06,
"logits/chosen": -1.489585041999817,
"logits/rejected": -1.3004378080368042,
"logps/chosen": -654.8056030273438,
"logps/rejected": -1286.8641357421875,
"loss": 0.0956,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.15169434249401093,
"rewards/margins": 0.25045520067214966,
"rewards/rejected": -0.40214958786964417,
"step": 1490
},
{
"epoch": 0.4,
"learning_rate": 3.7500000000000005e-06,
"logits/chosen": -1.6937404870986938,
"logits/rejected": -1.0904152393341064,
"logps/chosen": -521.1519775390625,
"logps/rejected": -1150.9736328125,
"loss": 0.0922,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.09178796410560608,
"rewards/margins": 0.2572135031223297,
"rewards/rejected": -0.3490014672279358,
"step": 1500
},
{
"epoch": 0.4,
"learning_rate": 3.7297928109491765e-06,
"logits/chosen": -1.5384725332260132,
"logits/rejected": -1.1151206493377686,
"logps/chosen": -500.30340576171875,
"logps/rejected": -1171.871337890625,
"loss": 0.0648,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.09594957530498505,
"rewards/margins": 0.31415387988090515,
"rewards/rejected": -0.4101034700870514,
"step": 1510
},
{
"epoch": 0.41,
"learning_rate": 3.7094790651387414e-06,
"logits/chosen": -1.750771164894104,
"logits/rejected": -1.0950592756271362,
"logps/chosen": -557.678466796875,
"logps/rejected": -1188.5867919921875,
"loss": 0.0778,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.14062079787254333,
"rewards/margins": 0.282297819852829,
"rewards/rejected": -0.4229187071323395,
"step": 1520
},
{
"epoch": 0.41,
"learning_rate": 3.689060522675689e-06,
"logits/chosen": -1.5061167478561401,
"logits/rejected": -1.0993683338165283,
"logps/chosen": -719.0147705078125,
"logps/rejected": -1315.07421875,
"loss": 0.0885,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.22822144627571106,
"rewards/margins": 0.24384041130542755,
"rewards/rejected": -0.47206181287765503,
"step": 1530
},
{
"epoch": 0.41,
"learning_rate": 3.668538952747236e-06,
"logits/chosen": -1.7315492630004883,
"logits/rejected": -1.098783254623413,
"logps/chosen": -702.6656494140625,
"logps/rejected": -1301.657958984375,
"loss": 0.0741,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.16757100820541382,
"rewards/margins": 0.2810649871826172,
"rewards/rejected": -0.4486359655857086,
"step": 1540
},
{
"epoch": 0.41,
"learning_rate": 3.6479161334675294e-06,
"logits/chosen": -1.6738449335098267,
"logits/rejected": -1.0924396514892578,
"logps/chosen": -584.9110107421875,
"logps/rejected": -1123.946533203125,
"loss": 0.0867,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.09195319563150406,
"rewards/margins": 0.2546834349632263,
"rewards/rejected": -0.3466365933418274,
"step": 1550
},
{
"epoch": 0.42,
"learning_rate": 3.627193851723577e-06,
"logits/chosen": -1.5191009044647217,
"logits/rejected": -1.2671594619750977,
"logps/chosen": -644.867431640625,
"logps/rejected": -1194.0931396484375,
"loss": 0.1019,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.133761465549469,
"rewards/margins": 0.24754054844379425,
"rewards/rejected": -0.38130199909210205,
"step": 1560
},
{
"epoch": 0.42,
"learning_rate": 3.6063739030204226e-06,
"logits/chosen": -1.7906509637832642,
"logits/rejected": -1.3401691913604736,
"logps/chosen": -475.7640075683594,
"logps/rejected": -1105.63623046875,
"loss": 0.0816,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.06556513905525208,
"rewards/margins": 0.2934816777706146,
"rewards/rejected": -0.3590467870235443,
"step": 1570
},
{
"epoch": 0.42,
"learning_rate": 3.5854580913255706e-06,
"logits/chosen": -1.6345583200454712,
"logits/rejected": -0.9686171412467957,
"logps/chosen": -546.1632080078125,
"logps/rejected": -1137.38232421875,
"loss": 0.0774,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.09245268255472183,
"rewards/margins": 0.2836568355560303,
"rewards/rejected": -0.3761095106601715,
"step": 1580
},
{
"epoch": 0.42,
"learning_rate": 3.564448228912682e-06,
"logits/chosen": -1.774083137512207,
"logits/rejected": -1.0874212980270386,
"logps/chosen": -627.0970458984375,
"logps/rejected": -1292.3609619140625,
"loss": 0.0548,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.14228971302509308,
"rewards/margins": 0.29837566614151,
"rewards/rejected": -0.4406653940677643,
"step": 1590
},
{
"epoch": 0.43,
"learning_rate": 3.543346136204545e-06,
"logits/chosen": -1.492363691329956,
"logits/rejected": -0.9753513336181641,
"logps/chosen": -715.6846923828125,
"logps/rejected": -1311.7313232421875,
"loss": 0.1043,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.19200663268566132,
"rewards/margins": 0.2355610430240631,
"rewards/rejected": -0.427567720413208,
"step": 1600
},
{
"epoch": 0.43,
"learning_rate": 3.522153641615345e-06,
"logits/chosen": -1.6796743869781494,
"logits/rejected": -1.0669097900390625,
"logps/chosen": -651.3643798828125,
"logps/rejected": -1211.4693603515625,
"loss": 0.0636,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12537550926208496,
"rewards/margins": 0.29044246673583984,
"rewards/rejected": -0.4158180356025696,
"step": 1610
},
{
"epoch": 0.43,
"learning_rate": 3.5008725813922383e-06,
"logits/chosen": -1.53739333152771,
"logits/rejected": -1.009767770767212,
"logps/chosen": -771.9571533203125,
"logps/rejected": -1199.3785400390625,
"loss": 0.0922,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.17814821004867554,
"rewards/margins": 0.24210956692695618,
"rewards/rejected": -0.4202577471733093,
"step": 1620
},
{
"epoch": 0.43,
"learning_rate": 3.4795047994562463e-06,
"logits/chosen": -1.4524638652801514,
"logits/rejected": -1.051928162574768,
"logps/chosen": -690.6948852539062,
"logps/rejected": -1226.161865234375,
"loss": 0.0876,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.18693287670612335,
"rewards/margins": 0.25937145948410034,
"rewards/rejected": -0.4463043212890625,
"step": 1630
},
{
"epoch": 0.44,
"learning_rate": 3.458052147242494e-06,
"logits/chosen": -1.8221120834350586,
"logits/rejected": -1.1258487701416016,
"logps/chosen": -630.0181274414062,
"logps/rejected": -1044.5679931640625,
"loss": 0.1046,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09280923753976822,
"rewards/margins": 0.23150058090686798,
"rewards/rejected": -0.3243098556995392,
"step": 1640
},
{
"epoch": 0.44,
"learning_rate": 3.436516483539781e-06,
"logits/chosen": -1.8554388284683228,
"logits/rejected": -1.3803541660308838,
"logps/chosen": -633.9047241210938,
"logps/rejected": -1229.6116943359375,
"loss": 0.0778,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13664793968200684,
"rewards/margins": 0.2661344110965729,
"rewards/rejected": -0.4027823805809021,
"step": 1650
},
{
"epoch": 0.44,
"learning_rate": 3.4148996743295305e-06,
"logits/chosen": -1.9048486948013306,
"logits/rejected": -1.058411955833435,
"logps/chosen": -720.1047973632812,
"logps/rejected": -1232.9039306640625,
"loss": 0.0732,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12227736413478851,
"rewards/margins": 0.2787812352180481,
"rewards/rejected": -0.4010585844516754,
"step": 1660
},
{
"epoch": 0.45,
"learning_rate": 3.3932035926241103e-06,
"logits/chosen": -1.6380399465560913,
"logits/rejected": -1.262209177017212,
"logps/chosen": -482.85662841796875,
"logps/rejected": -1295.7718505859375,
"loss": 0.0462,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08021150529384613,
"rewards/margins": 0.3560211658477783,
"rewards/rejected": -0.43623265624046326,
"step": 1670
},
{
"epoch": 0.45,
"learning_rate": 3.3714301183045382e-06,
"logits/chosen": -1.595003366470337,
"logits/rejected": -1.2377384901046753,
"logps/chosen": -547.7657470703125,
"logps/rejected": -1363.28125,
"loss": 0.0617,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12761729955673218,
"rewards/margins": 0.29965031147003174,
"rewards/rejected": -0.4272676110267639,
"step": 1680
},
{
"epoch": 0.45,
"learning_rate": 3.349581137957604e-06,
"logits/chosen": -1.8540780544281006,
"logits/rejected": -1.0956978797912598,
"logps/chosen": -570.6678466796875,
"logps/rejected": -1214.4659423828125,
"loss": 0.0458,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08135120570659637,
"rewards/margins": 0.32362592220306396,
"rewards/rejected": -0.40497714281082153,
"step": 1690
},
{
"epoch": 0.45,
"learning_rate": 3.3276585447123957e-06,
"logits/chosen": -1.4734654426574707,
"logits/rejected": -1.1689575910568237,
"logps/chosen": -471.4612731933594,
"logps/rejected": -1102.623046875,
"loss": 0.0775,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.06378956139087677,
"rewards/margins": 0.2605707049369812,
"rewards/rejected": -0.32436028122901917,
"step": 1700
},
{
"epoch": 0.46,
"learning_rate": 3.3056642380762783e-06,
"logits/chosen": -1.6254606246948242,
"logits/rejected": -0.9264505505561829,
"logps/chosen": -598.3116455078125,
"logps/rejected": -1273.1226806640625,
"loss": 0.0645,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.10875244438648224,
"rewards/margins": 0.3382043242454529,
"rewards/rejected": -0.4469567835330963,
"step": 1710
},
{
"epoch": 0.46,
"learning_rate": 3.2836001237702993e-06,
"logits/chosen": -1.7050098180770874,
"logits/rejected": -1.1736676692962646,
"logps/chosen": -605.455078125,
"logps/rejected": -1361.8411865234375,
"loss": 0.0707,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.09871726483106613,
"rewards/margins": 0.318486750125885,
"rewards/rejected": -0.41720399260520935,
"step": 1720
},
{
"epoch": 0.46,
"learning_rate": 3.2614681135640696e-06,
"logits/chosen": -1.666524887084961,
"logits/rejected": -1.0706026554107666,
"logps/chosen": -633.6300048828125,
"logps/rejected": -1287.6302490234375,
"loss": 0.0617,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.08346323668956757,
"rewards/margins": 0.29316291213035583,
"rewards/rejected": -0.376626193523407,
"step": 1730
},
{
"epoch": 0.46,
"learning_rate": 3.2392701251101172e-06,
"logits/chosen": -1.6395193338394165,
"logits/rejected": -1.1502461433410645,
"logps/chosen": -596.9259033203125,
"logps/rejected": -1209.7242431640625,
"loss": 0.0985,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.13397231698036194,
"rewards/margins": 0.2410469949245453,
"rewards/rejected": -0.3750193417072296,
"step": 1740
},
{
"epoch": 0.47,
"learning_rate": 3.217008081777726e-06,
"logits/chosen": -1.6936429738998413,
"logits/rejected": -1.1541160345077515,
"logps/chosen": -687.9554443359375,
"logps/rejected": -1373.260986328125,
"loss": 0.0775,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.19113728404045105,
"rewards/margins": 0.3089086413383484,
"rewards/rejected": -0.500045895576477,
"step": 1750
},
{
"epoch": 0.47,
"learning_rate": 3.1946839124862873e-06,
"logits/chosen": -1.499732255935669,
"logits/rejected": -1.073246955871582,
"logps/chosen": -632.1316528320312,
"logps/rejected": -1340.867919921875,
"loss": 0.055,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.14432887732982635,
"rewards/margins": 0.3110666275024414,
"rewards/rejected": -0.45539551973342896,
"step": 1760
},
{
"epoch": 0.47,
"learning_rate": 3.1722995515381644e-06,
"logits/chosen": -1.6296335458755493,
"logits/rejected": -0.9428736567497253,
"logps/chosen": -613.3892822265625,
"logps/rejected": -1241.071533203125,
"loss": 0.0922,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1409432291984558,
"rewards/margins": 0.28491806983947754,
"rewards/rejected": -0.42586126923561096,
"step": 1770
},
{
"epoch": 0.47,
"learning_rate": 3.149856938451094e-06,
"logits/chosen": -1.9275051355361938,
"logits/rejected": -0.9073979258537292,
"logps/chosen": -685.4188842773438,
"logps/rejected": -1194.86767578125,
"loss": 0.0869,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15064950287342072,
"rewards/margins": 0.25674349069595337,
"rewards/rejected": -0.4073929786682129,
"step": 1780
},
{
"epoch": 0.48,
"learning_rate": 3.127358017790132e-06,
"logits/chosen": -1.5247657299041748,
"logits/rejected": -0.8251350522041321,
"logps/chosen": -658.7485961914062,
"logps/rejected": -1302.989013671875,
"loss": 0.0565,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.15536119043827057,
"rewards/margins": 0.306736022233963,
"rewards/rejected": -0.46209725737571716,
"step": 1790
},
{
"epoch": 0.48,
"learning_rate": 3.1048047389991693e-06,
"logits/chosen": -1.5092096328735352,
"logits/rejected": -0.8953585624694824,
"logps/chosen": -563.54638671875,
"logps/rejected": -1277.382080078125,
"loss": 0.0635,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12220227718353271,
"rewards/margins": 0.32350245118141174,
"rewards/rejected": -0.44570469856262207,
"step": 1800
},
{
"epoch": 0.48,
"learning_rate": 3.082199056232015e-06,
"logits/chosen": -1.555259108543396,
"logits/rejected": -1.2513010501861572,
"logps/chosen": -576.5824584960938,
"logps/rejected": -1193.6871337890625,
"loss": 0.1014,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14716719090938568,
"rewards/margins": 0.2701808214187622,
"rewards/rejected": -0.4173479974269867,
"step": 1810
},
{
"epoch": 0.49,
"learning_rate": 3.059542928183079e-06,
"logits/chosen": -1.2370166778564453,
"logits/rejected": -0.862767219543457,
"logps/chosen": -630.3663330078125,
"logps/rejected": -1336.5428466796875,
"loss": 0.0544,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1622113287448883,
"rewards/margins": 0.32105860114097595,
"rewards/rejected": -0.48326998949050903,
"step": 1820
},
{
"epoch": 0.49,
"learning_rate": 3.0368383179176584e-06,
"logits/chosen": -1.385969877243042,
"logits/rejected": -1.0176479816436768,
"logps/chosen": -719.9691772460938,
"logps/rejected": -1240.3895263671875,
"loss": 0.1078,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.19016364216804504,
"rewards/margins": 0.2406831681728363,
"rewards/rejected": -0.43084684014320374,
"step": 1830
},
{
"epoch": 0.49,
"learning_rate": 3.0140871927018466e-06,
"logits/chosen": -1.6181665658950806,
"logits/rejected": -1.2087422609329224,
"logps/chosen": -704.0010986328125,
"logps/rejected": -1382.859619140625,
"loss": 0.0552,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19687633216381073,
"rewards/margins": 0.28999894857406616,
"rewards/rejected": -0.4868752360343933,
"step": 1840
},
{
"epoch": 0.49,
"learning_rate": 2.9912915238320755e-06,
"logits/chosen": -1.4043359756469727,
"logits/rejected": -1.1318188905715942,
"logps/chosen": -555.2379760742188,
"logps/rejected": -1354.341796875,
"loss": 0.0577,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.16941991448402405,
"rewards/margins": 0.33636412024497986,
"rewards/rejected": -0.5057840347290039,
"step": 1850
},
{
"epoch": 0.5,
"learning_rate": 2.9684532864643123e-06,
"logits/chosen": -1.7377849817276,
"logits/rejected": -1.16031014919281,
"logps/chosen": -689.5167236328125,
"logps/rejected": -1164.914794921875,
"loss": 0.1023,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.21689483523368835,
"rewards/margins": 0.22367532551288605,
"rewards/rejected": -0.440570205450058,
"step": 1860
},
{
"epoch": 0.5,
"learning_rate": 2.945574459442917e-06,
"logits/chosen": -1.6468284130096436,
"logits/rejected": -1.190332055091858,
"logps/chosen": -639.2093505859375,
"logps/rejected": -1291.997802734375,
"loss": 0.0592,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1794385313987732,
"rewards/margins": 0.31004798412323,
"rewards/rejected": -0.4894865155220032,
"step": 1870
},
{
"epoch": 0.5,
"learning_rate": 2.922657025129185e-06,
"logits/chosen": -1.5556018352508545,
"logits/rejected": -1.0828845500946045,
"logps/chosen": -718.0322875976562,
"logps/rejected": -1318.2115478515625,
"loss": 0.0983,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.27457764744758606,
"rewards/margins": 0.23540663719177246,
"rewards/rejected": -0.5099843144416809,
"step": 1880
},
{
"epoch": 0.5,
"learning_rate": 2.8997029692295875e-06,
"logits/chosen": -1.66329026222229,
"logits/rejected": -1.0540707111358643,
"logps/chosen": -647.1304931640625,
"logps/rejected": -1394.345703125,
"loss": 0.0644,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16953504085540771,
"rewards/margins": 0.3324953019618988,
"rewards/rejected": -0.5020303726196289,
"step": 1890
},
{
"epoch": 0.51,
"learning_rate": 2.876714280623708e-06,
"logits/chosen": -1.3675628900527954,
"logits/rejected": -0.910226047039032,
"logps/chosen": -656.5228881835938,
"logps/rejected": -1325.244140625,
"loss": 0.0585,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.18668699264526367,
"rewards/margins": 0.3124103248119354,
"rewards/rejected": -0.4990972876548767,
"step": 1900
},
{
"epoch": 0.51,
"learning_rate": 2.8536929511919227e-06,
"logits/chosen": -1.5113346576690674,
"logits/rejected": -0.917544960975647,
"logps/chosen": -621.9835815429688,
"logps/rejected": -1159.2508544921875,
"loss": 0.0598,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.17902755737304688,
"rewards/margins": 0.25234436988830566,
"rewards/rejected": -0.43137192726135254,
"step": 1910
},
{
"epoch": 0.51,
"learning_rate": 2.8306409756428067e-06,
"logits/chosen": -1.56842041015625,
"logits/rejected": -0.9856363534927368,
"logps/chosen": -565.5784301757812,
"logps/rejected": -1276.6402587890625,
"loss": 0.0583,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.158840149641037,
"rewards/margins": 0.29064539074897766,
"rewards/rejected": -0.4494854807853699,
"step": 1920
},
{
"epoch": 0.51,
"learning_rate": 2.807560351340302e-06,
"logits/chosen": -1.5603129863739014,
"logits/rejected": -1.0442321300506592,
"logps/chosen": -594.1895751953125,
"logps/rejected": -1224.7371826171875,
"loss": 0.0977,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.17820130288600922,
"rewards/margins": 0.2835688889026642,
"rewards/rejected": -0.4617701470851898,
"step": 1930
},
{
"epoch": 0.52,
"learning_rate": 2.7844530781306544e-06,
"logits/chosen": -1.436647653579712,
"logits/rejected": -0.8475536108016968,
"logps/chosen": -703.9013061523438,
"logps/rejected": -1465.6171875,
"loss": 0.0612,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.22263050079345703,
"rewards/margins": 0.3488808274269104,
"rewards/rejected": -0.5715113282203674,
"step": 1940
},
{
"epoch": 0.52,
"learning_rate": 2.761321158169134e-06,
"logits/chosen": -1.6537196636199951,
"logits/rejected": -1.1116868257522583,
"logps/chosen": -717.2203369140625,
"logps/rejected": -1459.016357421875,
"loss": 0.0578,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.19066976010799408,
"rewards/margins": 0.35232046246528625,
"rewards/rejected": -0.5429901480674744,
"step": 1950
},
{
"epoch": 0.52,
"learning_rate": 2.738166595746554e-06,
"logits/chosen": -1.709814429283142,
"logits/rejected": -0.9134047627449036,
"logps/chosen": -705.0501708984375,
"logps/rejected": -1441.9495849609375,
"loss": 0.0426,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -0.172596737742424,
"rewards/margins": 0.34574776887893677,
"rewards/rejected": -0.518344521522522,
"step": 1960
},
{
"epoch": 0.53,
"learning_rate": 2.7149913971156105e-06,
"logits/chosen": -1.6743762493133545,
"logits/rejected": -1.134342908859253,
"logps/chosen": -682.0380859375,
"logps/rejected": -1288.8563232421875,
"loss": 0.0783,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21863889694213867,
"rewards/margins": 0.30050721764564514,
"rewards/rejected": -0.5191460847854614,
"step": 1970
},
{
"epoch": 0.53,
"learning_rate": 2.6917975703170466e-06,
"logits/chosen": -1.384412407875061,
"logits/rejected": -0.9328360557556152,
"logps/chosen": -756.0421752929688,
"logps/rejected": -1321.971923828125,
"loss": 0.0844,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2494826316833496,
"rewards/margins": 0.2922648787498474,
"rewards/rejected": -0.541747510433197,
"step": 1980
},
{
"epoch": 0.53,
"learning_rate": 2.668587125005663e-06,
"logits/chosen": -1.573412299156189,
"logits/rejected": -0.9849382638931274,
"logps/chosen": -653.5593872070312,
"logps/rejected": -1219.503173828125,
"loss": 0.0674,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.22859041392803192,
"rewards/margins": 0.2729692757129669,
"rewards/rejected": -0.5015596747398376,
"step": 1990
},
{
"epoch": 0.53,
"learning_rate": 2.6453620722761897e-06,
"logits/chosen": -1.6217581033706665,
"logits/rejected": -0.829363226890564,
"logps/chosen": -687.14794921875,
"logps/rejected": -1367.03515625,
"loss": 0.0515,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.2579112648963928,
"rewards/margins": 0.3186108469963074,
"rewards/rejected": -0.5765220522880554,
"step": 2000
},
{
"epoch": 0.54,
"learning_rate": 2.6221244244890336e-06,
"logits/chosen": -1.4822559356689453,
"logits/rejected": -0.8854954838752747,
"logps/chosen": -803.3583984375,
"logps/rejected": -1373.9749755859375,
"loss": 0.0823,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2859230637550354,
"rewards/margins": 0.2797744870185852,
"rewards/rejected": -0.5656975507736206,
"step": 2010
},
{
"epoch": 0.54,
"learning_rate": 2.5988761950959133e-06,
"logits/chosen": -1.405379056930542,
"logits/rejected": -0.980889618396759,
"logps/chosen": -653.2147216796875,
"logps/rejected": -1412.9344482421875,
"loss": 0.0489,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19988027215003967,
"rewards/margins": 0.3394158184528351,
"rewards/rejected": -0.5392960906028748,
"step": 2020
},
{
"epoch": 0.54,
"learning_rate": 2.575619398465402e-06,
"logits/chosen": -1.5412838459014893,
"logits/rejected": -0.9163180589675903,
"logps/chosen": -539.6043090820312,
"logps/rejected": -1107.312255859375,
"loss": 0.0771,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1628035455942154,
"rewards/margins": 0.2565108835697174,
"rewards/rejected": -0.419314444065094,
"step": 2030
},
{
"epoch": 0.54,
"learning_rate": 2.5523560497083927e-06,
"logits/chosen": -1.482742428779602,
"logits/rejected": -1.0331344604492188,
"logps/chosen": -751.7644653320312,
"logps/rejected": -1281.9117431640625,
"loss": 0.0847,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.21723918616771698,
"rewards/margins": 0.26914697885513306,
"rewards/rejected": -0.48638615012168884,
"step": 2040
},
{
"epoch": 0.55,
"learning_rate": 2.5290881645034932e-06,
"logits/chosen": -1.6871871948242188,
"logits/rejected": -0.9969019889831543,
"logps/chosen": -712.149169921875,
"logps/rejected": -1383.6920166015625,
"loss": 0.0813,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2281813621520996,
"rewards/margins": 0.3213092088699341,
"rewards/rejected": -0.5494905710220337,
"step": 2050
},
{
"epoch": 0.55,
"learning_rate": 2.5058177589223766e-06,
"logits/chosen": -1.4727680683135986,
"logits/rejected": -1.048156976699829,
"logps/chosen": -657.0037231445312,
"logps/rejected": -1363.59814453125,
"loss": 0.0745,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.2256421595811844,
"rewards/margins": 0.31263765692710876,
"rewards/rejected": -0.5382798314094543,
"step": 2060
},
{
"epoch": 0.55,
"learning_rate": 2.482546849255096e-06,
"logits/chosen": -1.5077242851257324,
"logits/rejected": -0.8984912633895874,
"logps/chosen": -722.7063598632812,
"logps/rejected": -1281.8148193359375,
"loss": 0.0741,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.23482950031757355,
"rewards/margins": 0.291878879070282,
"rewards/rejected": -0.5267083644866943,
"step": 2070
},
{
"epoch": 0.55,
"learning_rate": 2.4592774518353858e-06,
"logits/chosen": -1.6423757076263428,
"logits/rejected": -0.9732748866081238,
"logps/chosen": -625.0535888671875,
"logps/rejected": -1348.394287109375,
"loss": 0.0562,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19351795315742493,
"rewards/margins": 0.3293379247188568,
"rewards/rejected": -0.5228558778762817,
"step": 2080
},
{
"epoch": 0.56,
"learning_rate": 2.436011582865945e-06,
"logits/chosen": -1.7477773427963257,
"logits/rejected": -1.0844796895980835,
"logps/chosen": -786.804443359375,
"logps/rejected": -1510.250732421875,
"loss": 0.0565,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.24655666947364807,
"rewards/margins": 0.34261250495910645,
"rewards/rejected": -0.5891691446304321,
"step": 2090
},
{
"epoch": 0.56,
"learning_rate": 2.4127512582437486e-06,
"logits/chosen": -1.1730453968048096,
"logits/rejected": -1.0297441482543945,
"logps/chosen": -610.2826538085938,
"logps/rejected": -1334.4935302734375,
"loss": 0.075,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.21944193542003632,
"rewards/margins": 0.28680866956710815,
"rewards/rejected": -0.5062506198883057,
"step": 2100
},
{
"epoch": 0.56,
"learning_rate": 2.3894984933853734e-06,
"logits/chosen": -1.6909167766571045,
"logits/rejected": -1.2207863330841064,
"logps/chosen": -736.7841796875,
"logps/rejected": -1436.5059814453125,
"loss": 0.0715,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2114332616329193,
"rewards/margins": 0.30715304613113403,
"rewards/rejected": -0.518586277961731,
"step": 2110
},
{
"epoch": 0.57,
"learning_rate": 2.366255303052377e-06,
"logits/chosen": -1.7539796829223633,
"logits/rejected": -1.0083894729614258,
"logps/chosen": -702.0449829101562,
"logps/rejected": -1368.424072265625,
"loss": 0.0588,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.17488941550254822,
"rewards/margins": 0.3086225390434265,
"rewards/rejected": -0.4835119843482971,
"step": 2120
},
{
"epoch": 0.57,
"learning_rate": 2.3430237011767166e-06,
"logits/chosen": -1.783517837524414,
"logits/rejected": -0.9859122037887573,
"logps/chosen": -631.6397705078125,
"logps/rejected": -1418.53857421875,
"loss": 0.0386,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.13998395204544067,
"rewards/margins": 0.3510037064552307,
"rewards/rejected": -0.490987628698349,
"step": 2130
},
{
"epoch": 0.57,
"learning_rate": 2.319805700686257e-06,
"logits/chosen": -1.5874156951904297,
"logits/rejected": -1.0402486324310303,
"logps/chosen": -647.2991943359375,
"logps/rejected": -1264.6134033203125,
"loss": 0.0663,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15432874858379364,
"rewards/margins": 0.29131144285202026,
"rewards/rejected": -0.4456402361392975,
"step": 2140
},
{
"epoch": 0.57,
"learning_rate": 2.296603313330355e-06,
"logits/chosen": -1.6740115880966187,
"logits/rejected": -1.049090027809143,
"logps/chosen": -650.9720458984375,
"logps/rejected": -1069.4693603515625,
"loss": 0.1228,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.1620035022497177,
"rewards/margins": 0.2206643521785736,
"rewards/rejected": -0.3826678991317749,
"step": 2150
},
{
"epoch": 0.58,
"learning_rate": 2.2734185495055503e-06,
"logits/chosen": -1.4526993036270142,
"logits/rejected": -1.0033073425292969,
"logps/chosen": -464.20257568359375,
"logps/rejected": -1249.659423828125,
"loss": 0.0672,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.11725373566150665,
"rewards/margins": 0.32836082577705383,
"rewards/rejected": -0.4456145763397217,
"step": 2160
},
{
"epoch": 0.58,
"learning_rate": 2.250253418081373e-06,
"logits/chosen": -1.6637146472930908,
"logits/rejected": -1.0717463493347168,
"logps/chosen": -662.2470703125,
"logps/rejected": -1223.6536865234375,
"loss": 0.0843,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1466962993144989,
"rewards/margins": 0.2738208770751953,
"rewards/rejected": -0.4205172061920166,
"step": 2170
},
{
"epoch": 0.58,
"learning_rate": 2.22710992622628e-06,
"logits/chosen": -1.7047134637832642,
"logits/rejected": -0.7803869843482971,
"logps/chosen": -681.6057739257812,
"logps/rejected": -1309.5865478515625,
"loss": 0.057,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.15952114760875702,
"rewards/margins": 0.309850811958313,
"rewards/rejected": -0.4693719744682312,
"step": 2180
},
{
"epoch": 0.58,
"learning_rate": 2.2039900792337477e-06,
"logits/chosen": -1.5044130086898804,
"logits/rejected": -0.787899374961853,
"logps/chosen": -632.465576171875,
"logps/rejected": -1316.4517822265625,
"loss": 0.0515,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.18237130343914032,
"rewards/margins": 0.32328280806541443,
"rewards/rejected": -0.5056540369987488,
"step": 2190
},
{
"epoch": 0.59,
"learning_rate": 2.1808958803485134e-06,
"logits/chosen": -1.6702648401260376,
"logits/rejected": -0.990136444568634,
"logps/chosen": -455.564208984375,
"logps/rejected": -1168.5498046875,
"loss": 0.0599,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.12168209254741669,
"rewards/margins": 0.3131280541419983,
"rewards/rejected": -0.434810072183609,
"step": 2200
},
{
"epoch": 0.59,
"learning_rate": 2.157829330593008e-06,
"logits/chosen": -1.6868867874145508,
"logits/rejected": -0.9523127675056458,
"logps/chosen": -733.1408081054688,
"logps/rejected": -1337.63818359375,
"loss": 0.0715,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.21179482340812683,
"rewards/margins": 0.30666384100914,
"rewards/rejected": -0.5184586644172668,
"step": 2210
},
{
"epoch": 0.59,
"learning_rate": 2.134792428593971e-06,
"logits/chosen": -1.6914653778076172,
"logits/rejected": -1.2282737493515015,
"logps/chosen": -614.1090087890625,
"logps/rejected": -1088.0286865234375,
"loss": 0.1059,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1738576740026474,
"rewards/margins": 0.21695072948932648,
"rewards/rejected": -0.3908084034919739,
"step": 2220
},
{
"epoch": 0.59,
"learning_rate": 2.1117871704092818e-06,
"logits/chosen": -1.6489229202270508,
"logits/rejected": -0.9123425483703613,
"logps/chosen": -679.4093017578125,
"logps/rejected": -1336.1690673828125,
"loss": 0.0813,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1528438925743103,
"rewards/margins": 0.3031119108200073,
"rewards/rejected": -0.4559558033943176,
"step": 2230
},
{
"epoch": 0.6,
"learning_rate": 2.0888155493550027e-06,
"logits/chosen": -1.658093810081482,
"logits/rejected": -1.2635515928268433,
"logps/chosen": -563.3775634765625,
"logps/rejected": -1217.8135986328125,
"loss": 0.0686,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.12988874316215515,
"rewards/margins": 0.31268182396888733,
"rewards/rejected": -0.44257062673568726,
"step": 2240
},
{
"epoch": 0.6,
"learning_rate": 2.0658795558326745e-06,
"logits/chosen": -1.5542079210281372,
"logits/rejected": -0.9705120921134949,
"logps/chosen": -641.7821655273438,
"logps/rejected": -1335.944091796875,
"loss": 0.0586,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1482844054698944,
"rewards/margins": 0.3350493013858795,
"rewards/rejected": -0.48333367705345154,
"step": 2250
},
{
"epoch": 0.6,
"learning_rate": 2.0429811771568468e-06,
"logits/chosen": -1.6625282764434814,
"logits/rejected": -0.9575098752975464,
"logps/chosen": -673.6990966796875,
"logps/rejected": -1288.328369140625,
"loss": 0.0408,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1419149488210678,
"rewards/margins": 0.2989000678062439,
"rewards/rejected": -0.4408150315284729,
"step": 2260
},
{
"epoch": 0.61,
"learning_rate": 2.0201223973828917e-06,
"logits/chosen": -1.7629448175430298,
"logits/rejected": -1.1014915704727173,
"logps/chosen": -609.2691650390625,
"logps/rejected": -1323.014404296875,
"loss": 0.0751,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12834301590919495,
"rewards/margins": 0.31924059987068176,
"rewards/rejected": -0.4475835859775543,
"step": 2270
},
{
"epoch": 0.61,
"learning_rate": 1.997305197135089e-06,
"logits/chosen": -1.5687427520751953,
"logits/rejected": -0.9063301086425781,
"logps/chosen": -674.9013671875,
"logps/rejected": -1229.5550537109375,
"loss": 0.0873,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12829624116420746,
"rewards/margins": 0.24914589524269104,
"rewards/rejected": -0.3774421215057373,
"step": 2280
},
{
"epoch": 0.61,
"learning_rate": 1.9745315534350157e-06,
"logits/chosen": -1.6142327785491943,
"logits/rejected": -1.2041738033294678,
"logps/chosen": -527.4874877929688,
"logps/rejected": -1276.542724609375,
"loss": 0.0724,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.11953765153884888,
"rewards/margins": 0.3008989989757538,
"rewards/rejected": -0.42043668031692505,
"step": 2290
},
{
"epoch": 0.61,
"learning_rate": 1.9518034395302413e-06,
"logits/chosen": -1.584478497505188,
"logits/rejected": -0.8746352195739746,
"logps/chosen": -507.75128173828125,
"logps/rejected": -1205.164794921875,
"loss": 0.0654,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.12713788449764252,
"rewards/margins": 0.3278200626373291,
"rewards/rejected": -0.4549580216407776,
"step": 2300
},
{
"epoch": 0.62,
"learning_rate": 1.9291228247233607e-06,
"logits/chosen": -1.452108383178711,
"logits/rejected": -0.9603986740112305,
"logps/chosen": -750.004638671875,
"logps/rejected": -1415.546875,
"loss": 0.0642,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.22353100776672363,
"rewards/margins": 0.3011419475078583,
"rewards/rejected": -0.5246729254722595,
"step": 2310
},
{
"epoch": 0.62,
"learning_rate": 1.9064916742013515e-06,
"logits/chosen": -1.5349475145339966,
"logits/rejected": -0.8534983396530151,
"logps/chosen": -691.6678466796875,
"logps/rejected": -1337.88671875,
"loss": 0.0564,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.19090059399604797,
"rewards/margins": 0.3062039911746979,
"rewards/rejected": -0.49710458517074585,
"step": 2320
},
{
"epoch": 0.62,
"learning_rate": 1.883911948865306e-06,
"logits/chosen": -1.6528412103652954,
"logits/rejected": -1.132430076599121,
"logps/chosen": -626.3675537109375,
"logps/rejected": -1352.5123291015625,
"loss": 0.0702,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1741064339876175,
"rewards/margins": 0.31939199566841125,
"rewards/rejected": -0.49349841475486755,
"step": 2330
},
{
"epoch": 0.62,
"learning_rate": 1.8613856051605242e-06,
"logits/chosen": -1.8713703155517578,
"logits/rejected": -1.2219023704528809,
"logps/chosen": -479.8798828125,
"logps/rejected": -1113.4814453125,
"loss": 0.0658,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.10413724184036255,
"rewards/margins": 0.322685569524765,
"rewards/rejected": -0.42682284116744995,
"step": 2340
},
{
"epoch": 0.63,
"learning_rate": 1.8389145949069953e-06,
"logits/chosen": -1.5313884019851685,
"logits/rejected": -1.042482614517212,
"logps/chosen": -540.1700439453125,
"logps/rejected": -1220.364990234375,
"loss": 0.0499,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.13704395294189453,
"rewards/margins": 0.29186248779296875,
"rewards/rejected": -0.4289064407348633,
"step": 2350
},
{
"epoch": 0.63,
"learning_rate": 1.816500865130279e-06,
"logits/chosen": -1.495516300201416,
"logits/rejected": -1.306896448135376,
"logps/chosen": -562.7477416992188,
"logps/rejected": -1185.534912109375,
"loss": 0.1051,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.14714586734771729,
"rewards/margins": 0.26410627365112305,
"rewards/rejected": -0.4112521708011627,
"step": 2360
},
{
"epoch": 0.63,
"learning_rate": 1.7941463578928088e-06,
"logits/chosen": -1.6695563793182373,
"logits/rejected": -0.9721006155014038,
"logps/chosen": -601.9846801757812,
"logps/rejected": -1367.9365234375,
"loss": 0.0603,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1521584689617157,
"rewards/margins": 0.35412856936454773,
"rewards/rejected": -0.5062869787216187,
"step": 2370
},
{
"epoch": 0.63,
"learning_rate": 1.7718530101256115e-06,
"logits/chosen": -1.6784439086914062,
"logits/rejected": -1.0285909175872803,
"logps/chosen": -690.5152587890625,
"logps/rejected": -1287.234130859375,
"loss": 0.083,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.17937234044075012,
"rewards/margins": 0.27991387248039246,
"rewards/rejected": -0.4592861533164978,
"step": 2380
},
{
"epoch": 0.64,
"learning_rate": 1.7496227534604859e-06,
"logits/chosen": -1.7998254299163818,
"logits/rejected": -1.1072345972061157,
"logps/chosen": -648.5956420898438,
"logps/rejected": -1205.2398681640625,
"loss": 0.0657,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1724121868610382,
"rewards/margins": 0.29253289103507996,
"rewards/rejected": -0.46494507789611816,
"step": 2390
},
{
"epoch": 0.64,
"learning_rate": 1.7274575140626318e-06,
"logits/chosen": -1.6273527145385742,
"logits/rejected": -1.183037519454956,
"logps/chosen": -620.6856689453125,
"logps/rejected": -1215.266845703125,
"loss": 0.0901,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.17143599689006805,
"rewards/margins": 0.2781030535697937,
"rewards/rejected": -0.44953903555870056,
"step": 2400
},
{
"epoch": 0.64,
"learning_rate": 1.7053592124637557e-06,
"logits/chosen": -1.8067405223846436,
"logits/rejected": -1.2858269214630127,
"logps/chosen": -659.517822265625,
"logps/rejected": -1250.8367919921875,
"loss": 0.0949,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.17256858944892883,
"rewards/margins": 0.2582642436027527,
"rewards/rejected": -0.43083280324935913,
"step": 2410
},
{
"epoch": 0.65,
"learning_rate": 1.6833297633956647e-06,
"logits/chosen": -1.443641185760498,
"logits/rejected": -1.0768052339553833,
"logps/chosen": -559.7144775390625,
"logps/rejected": -1292.55859375,
"loss": 0.0641,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.11608362197875977,
"rewards/margins": 0.34168118238449097,
"rewards/rejected": -0.45776480436325073,
"step": 2420
},
{
"epoch": 0.65,
"learning_rate": 1.661371075624363e-06,
"logits/chosen": -1.6469463109970093,
"logits/rejected": -1.0461533069610596,
"logps/chosen": -603.0146484375,
"logps/rejected": -1207.7100830078125,
"loss": 0.07,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.10989055782556534,
"rewards/margins": 0.2882770299911499,
"rewards/rejected": -0.39816758036613464,
"step": 2430
},
{
"epoch": 0.65,
"learning_rate": 1.6394850517846621e-06,
"logits/chosen": -1.6739981174468994,
"logits/rejected": -1.2458163499832153,
"logps/chosen": -687.8931884765625,
"logps/rejected": -1236.5269775390625,
"loss": 0.0977,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.148878276348114,
"rewards/margins": 0.2945893704891205,
"rewards/rejected": -0.4434676170349121,
"step": 2440
},
{
"epoch": 0.65,
"learning_rate": 1.6176735882153284e-06,
"logits/chosen": -1.6518714427947998,
"logits/rejected": -0.9769365191459656,
"logps/chosen": -496.71661376953125,
"logps/rejected": -1090.7801513671875,
"loss": 0.081,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.12074653804302216,
"rewards/margins": 0.26993122696876526,
"rewards/rejected": -0.3906777799129486,
"step": 2450
},
{
"epoch": 0.66,
"learning_rate": 1.5959385747947697e-06,
"logits/chosen": -1.5575945377349854,
"logits/rejected": -1.1600332260131836,
"logps/chosen": -541.9180908203125,
"logps/rejected": -1182.4320068359375,
"loss": 0.0797,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.16873383522033691,
"rewards/margins": 0.2602773606777191,
"rewards/rejected": -0.42901119589805603,
"step": 2460
},
{
"epoch": 0.66,
"learning_rate": 1.5742818947772875e-06,
"logits/chosen": -1.6753427982330322,
"logits/rejected": -0.9124045372009277,
"logps/chosen": -678.1805419921875,
"logps/rejected": -1311.1553955078125,
"loss": 0.0578,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.17256096005439758,
"rewards/margins": 0.3086041510105133,
"rewards/rejected": -0.4811651110649109,
"step": 2470
},
{
"epoch": 0.66,
"learning_rate": 1.552705424629898e-06,
"logits/chosen": -1.5884169340133667,
"logits/rejected": -0.9173520803451538,
"logps/chosen": -705.35009765625,
"logps/rejected": -1284.0096435546875,
"loss": 0.0683,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16716155409812927,
"rewards/margins": 0.28985482454299927,
"rewards/rejected": -0.4570164084434509,
"step": 2480
},
{
"epoch": 0.66,
"learning_rate": 1.5312110338697427e-06,
"logits/chosen": -1.533384084701538,
"logits/rejected": -1.1220409870147705,
"logps/chosen": -613.6298828125,
"logps/rejected": -1082.766357421875,
"loss": 0.1039,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.17001187801361084,
"rewards/margins": 0.20055198669433594,
"rewards/rejected": -0.3705638647079468,
"step": 2490
},
{
"epoch": 0.67,
"learning_rate": 1.509800584902108e-06,
"logits/chosen": -1.4955171346664429,
"logits/rejected": -0.8390592336654663,
"logps/chosen": -611.5145263671875,
"logps/rejected": -1197.4530029296875,
"loss": 0.0597,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12041016668081284,
"rewards/margins": 0.2896363139152527,
"rewards/rejected": -0.4100464880466461,
"step": 2500
},
{
"epoch": 0.67,
"learning_rate": 1.4884759328590476e-06,
"logits/chosen": -1.914384126663208,
"logits/rejected": -1.2168171405792236,
"logps/chosen": -600.9666137695312,
"logps/rejected": -1124.942626953125,
"loss": 0.0888,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.11489248275756836,
"rewards/margins": 0.26221710443496704,
"rewards/rejected": -0.3771095871925354,
"step": 2510
},
{
"epoch": 0.67,
"learning_rate": 1.467238925438646e-06,
"logits/chosen": -1.738040566444397,
"logits/rejected": -0.8857895731925964,
"logps/chosen": -679.0386962890625,
"logps/rejected": -1410.933837890625,
"loss": 0.0558,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.16262464225292206,
"rewards/margins": 0.339643657207489,
"rewards/rejected": -0.5022683143615723,
"step": 2520
},
{
"epoch": 0.67,
"learning_rate": 1.446091402744923e-06,
"logits/chosen": -1.3981173038482666,
"logits/rejected": -0.8543848991394043,
"logps/chosen": -590.76513671875,
"logps/rejected": -1206.71240234375,
"loss": 0.0583,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.15895266830921173,
"rewards/margins": 0.2688734829425812,
"rewards/rejected": -0.4278261661529541,
"step": 2530
},
{
"epoch": 0.68,
"learning_rate": 1.4250351971283937e-06,
"logits/chosen": -1.6803276538848877,
"logits/rejected": -1.0602971315383911,
"logps/chosen": -679.5074462890625,
"logps/rejected": -1229.011962890625,
"loss": 0.0871,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19001373648643494,
"rewards/margins": 0.26307451725006104,
"rewards/rejected": -0.45308828353881836,
"step": 2540
},
{
"epoch": 0.68,
"learning_rate": 1.4040721330273063e-06,
"logits/chosen": -1.5601686239242554,
"logits/rejected": -0.9407947659492493,
"logps/chosen": -616.7720336914062,
"logps/rejected": -1353.0308837890625,
"loss": 0.0501,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16183273494243622,
"rewards/margins": 0.3334823250770569,
"rewards/rejected": -0.4953150749206543,
"step": 2550
},
{
"epoch": 0.68,
"learning_rate": 1.3832040268095589e-06,
"logits/chosen": -1.52444589138031,
"logits/rejected": -0.7717920541763306,
"logps/chosen": -666.0655517578125,
"logps/rejected": -1330.985595703125,
"loss": 0.0645,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1790360063314438,
"rewards/margins": 0.3166579306125641,
"rewards/rejected": -0.49569398164749146,
"step": 2560
},
{
"epoch": 0.69,
"learning_rate": 1.362432686615316e-06,
"logits/chosen": -1.511685848236084,
"logits/rejected": -1.2630524635314941,
"logps/chosen": -488.4208068847656,
"logps/rejected": -1233.045166015625,
"loss": 0.0768,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12492994964122772,
"rewards/margins": 0.30061617493629456,
"rewards/rejected": -0.4255460798740387,
"step": 2570
},
{
"epoch": 0.69,
"learning_rate": 1.3417599122003464e-06,
"logits/chosen": -1.624969720840454,
"logits/rejected": -0.8495733141899109,
"logps/chosen": -633.3519897460938,
"logps/rejected": -1253.824951171875,
"loss": 0.0632,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.18349340558052063,
"rewards/margins": 0.2837851643562317,
"rewards/rejected": -0.4672785699367523,
"step": 2580
},
{
"epoch": 0.69,
"learning_rate": 1.3211874947800747e-06,
"logits/chosen": -1.4943301677703857,
"logits/rejected": -1.0596059560775757,
"logps/chosen": -581.8253784179688,
"logps/rejected": -1226.0418701171875,
"loss": 0.061,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1564798802137375,
"rewards/margins": 0.30731362104415894,
"rewards/rejected": -0.4637935161590576,
"step": 2590
},
{
"epoch": 0.69,
"learning_rate": 1.3007172168743854e-06,
"logits/chosen": -1.4282827377319336,
"logits/rejected": -1.0970618724822998,
"logps/chosen": -622.9385375976562,
"logps/rejected": -1321.9234619140625,
"loss": 0.0718,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.18767477571964264,
"rewards/margins": 0.32170000672340393,
"rewards/rejected": -0.5093748569488525,
"step": 2600
},
{
"epoch": 0.7,
"learning_rate": 1.280350852153168e-06,
"logits/chosen": -1.4260807037353516,
"logits/rejected": -0.8716268539428711,
"logps/chosen": -704.5125732421875,
"logps/rejected": -1340.787353515625,
"loss": 0.0737,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.22893400490283966,
"rewards/margins": 0.28585508465766907,
"rewards/rejected": -0.5147891044616699,
"step": 2610
},
{
"epoch": 0.7,
"learning_rate": 1.260090165282645e-06,
"logits/chosen": -1.5481798648834229,
"logits/rejected": -1.048405647277832,
"logps/chosen": -643.6202392578125,
"logps/rejected": -1242.826416015625,
"loss": 0.0806,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2178986817598343,
"rewards/margins": 0.2719099521636963,
"rewards/rejected": -0.4898086488246918,
"step": 2620
},
{
"epoch": 0.7,
"learning_rate": 1.2399369117724582e-06,
"logits/chosen": -1.650472640991211,
"logits/rejected": -0.9782799482345581,
"logps/chosen": -707.5630493164062,
"logps/rejected": -1298.345458984375,
"loss": 0.0753,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16964152455329895,
"rewards/margins": 0.2897980809211731,
"rewards/rejected": -0.45943960547447205,
"step": 2630
},
{
"epoch": 0.7,
"learning_rate": 1.2198928378235717e-06,
"logits/chosen": -1.6851444244384766,
"logits/rejected": -0.92662113904953,
"logps/chosen": -619.037841796875,
"logps/rejected": -1287.4412841796875,
"loss": 0.0522,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13626527786254883,
"rewards/margins": 0.3253551125526428,
"rewards/rejected": -0.46162039041519165,
"step": 2640
},
{
"epoch": 0.71,
"learning_rate": 1.1999596801769617e-06,
"logits/chosen": -1.8352587223052979,
"logits/rejected": -1.171582818031311,
"logps/chosen": -616.7615966796875,
"logps/rejected": -1335.4404296875,
"loss": 0.049,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1257292926311493,
"rewards/margins": 0.32309678196907043,
"rewards/rejected": -0.4488261342048645,
"step": 2650
},
{
"epoch": 0.71,
"learning_rate": 1.1801391659631423e-06,
"logits/chosen": -1.6399614810943604,
"logits/rejected": -1.0923852920532227,
"logps/chosen": -635.6760864257812,
"logps/rejected": -1325.152099609375,
"loss": 0.0653,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.16147668659687042,
"rewards/margins": 0.3237590193748474,
"rewards/rejected": -0.48523569107055664,
"step": 2660
},
{
"epoch": 0.71,
"learning_rate": 1.160433012552508e-06,
"logits/chosen": -1.5931923389434814,
"logits/rejected": -1.2261526584625244,
"logps/chosen": -603.7941284179688,
"logps/rejected": -1107.156982421875,
"loss": 0.1027,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.14636698365211487,
"rewards/margins": 0.22241589426994324,
"rewards/rejected": -0.3687829077243805,
"step": 2670
},
{
"epoch": 0.71,
"learning_rate": 1.1408429274065418e-06,
"logits/chosen": -1.481740117073059,
"logits/rejected": -1.1304103136062622,
"logps/chosen": -606.6870727539062,
"logps/rejected": -1188.147216796875,
"loss": 0.0763,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1330951750278473,
"rewards/margins": 0.2732751667499542,
"rewards/rejected": -0.4063703119754791,
"step": 2680
},
{
"epoch": 0.72,
"learning_rate": 1.1213706079298566e-06,
"logits/chosen": -1.6204341650009155,
"logits/rejected": -1.0541805028915405,
"logps/chosen": -554.864990234375,
"logps/rejected": -1167.746826171875,
"loss": 0.0617,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11585495620965958,
"rewards/margins": 0.30027204751968384,
"rewards/rejected": -0.4161270260810852,
"step": 2690
},
{
"epoch": 0.72,
"learning_rate": 1.1020177413231334e-06,
"logits/chosen": -1.6089589595794678,
"logits/rejected": -1.0122666358947754,
"logps/chosen": -685.2811889648438,
"logps/rejected": -1211.408935546875,
"loss": 0.0745,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16823996603488922,
"rewards/margins": 0.2715032994747162,
"rewards/rejected": -0.4397433400154114,
"step": 2700
},
{
"epoch": 0.72,
"learning_rate": 1.0827860044369226e-06,
"logits/chosen": -1.7601515054702759,
"logits/rejected": -1.007900595664978,
"logps/chosen": -704.5645751953125,
"logps/rejected": -1274.189453125,
"loss": 0.0737,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.17606690526008606,
"rewards/margins": 0.30847105383872986,
"rewards/rejected": -0.48453792929649353,
"step": 2710
},
{
"epoch": 0.73,
"learning_rate": 1.06367706362636e-06,
"logits/chosen": -1.5721882581710815,
"logits/rejected": -1.0723769664764404,
"logps/chosen": -665.5977783203125,
"logps/rejected": -1278.2569580078125,
"loss": 0.0799,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.18703334033489227,
"rewards/margins": 0.2840210795402527,
"rewards/rejected": -0.47105446457862854,
"step": 2720
},
{
"epoch": 0.73,
"learning_rate": 1.0446925746067768e-06,
"logits/chosen": -1.6236509084701538,
"logits/rejected": -1.0335582494735718,
"logps/chosen": -735.1541137695312,
"logps/rejected": -1341.9544677734375,
"loss": 0.0686,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1772700995206833,
"rewards/margins": 0.31488290429115295,
"rewards/rejected": -0.49215301871299744,
"step": 2730
},
{
"epoch": 0.73,
"learning_rate": 1.0258341823102418e-06,
"logits/chosen": -1.6834526062011719,
"logits/rejected": -1.0599212646484375,
"logps/chosen": -650.5093994140625,
"logps/rejected": -1288.187744140625,
"loss": 0.0623,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14554272592067719,
"rewards/margins": 0.3072082996368408,
"rewards/rejected": -0.4527510106563568,
"step": 2740
},
{
"epoch": 0.73,
"learning_rate": 1.0071035207430352e-06,
"logits/chosen": -1.7902591228485107,
"logits/rejected": -1.289333462715149,
"logps/chosen": -654.4462890625,
"logps/rejected": -1320.8380126953125,
"loss": 0.0673,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.16027647256851196,
"rewards/margins": 0.28841906785964966,
"rewards/rejected": -0.44869551062583923,
"step": 2750
},
{
"epoch": 0.74,
"learning_rate": 9.88502212844063e-07,
"logits/chosen": -1.528241515159607,
"logits/rejected": -1.140417218208313,
"logps/chosen": -584.2600708007812,
"logps/rejected": -1204.6785888671875,
"loss": 0.0822,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.12145284563302994,
"rewards/margins": 0.2952510714530945,
"rewards/rejected": -0.4167039394378662,
"step": 2760
},
{
"epoch": 0.74,
"learning_rate": 9.700318703442437e-07,
"logits/chosen": -1.6165825128555298,
"logits/rejected": -1.2629040479660034,
"logps/chosen": -692.536865234375,
"logps/rejected": -1356.869384765625,
"loss": 0.0861,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16812773048877716,
"rewards/margins": 0.2822396159172058,
"rewards/rejected": -0.4503673017024994,
"step": 2770
},
{
"epoch": 0.74,
"learning_rate": 9.516940936268504e-07,
"logits/chosen": -1.4886250495910645,
"logits/rejected": -0.8492704629898071,
"logps/chosen": -567.9114379882812,
"logps/rejected": -1223.29931640625,
"loss": 0.0907,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15036030113697052,
"rewards/margins": 0.3028646409511566,
"rewards/rejected": -0.45322495698928833,
"step": 2780
},
{
"epoch": 0.74,
"learning_rate": 9.334904715888496e-07,
"logits/chosen": -1.6458749771118164,
"logits/rejected": -1.2635785341262817,
"logps/chosen": -567.9539794921875,
"logps/rejected": -1301.9921875,
"loss": 0.0872,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1548917442560196,
"rewards/margins": 0.31059738993644714,
"rewards/rejected": -0.46548905968666077,
"step": 2790
},
{
"epoch": 0.75,
"learning_rate": 9.154225815032242e-07,
"logits/chosen": -1.5953037738800049,
"logits/rejected": -1.0339276790618896,
"logps/chosen": -624.7017822265625,
"logps/rejected": -1236.400634765625,
"loss": 0.0716,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.17240996658802032,
"rewards/margins": 0.28697705268859863,
"rewards/rejected": -0.45938700437545776,
"step": 2800
},
{
"epoch": 0.75,
"learning_rate": 8.974919888823164e-07,
"logits/chosen": -1.4576761722564697,
"logits/rejected": -1.1351208686828613,
"logps/chosen": -624.9636840820312,
"logps/rejected": -1355.374755859375,
"loss": 0.0514,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.19172172248363495,
"rewards/margins": 0.3108896315097809,
"rewards/rejected": -0.5026113390922546,
"step": 2810
},
{
"epoch": 0.75,
"learning_rate": 8.797002473421729e-07,
"logits/chosen": -1.5604914426803589,
"logits/rejected": -0.9569327235221863,
"logps/chosen": -538.5881958007812,
"logps/rejected": -1219.4466552734375,
"loss": 0.0683,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.15165778994560242,
"rewards/margins": 0.28516024351119995,
"rewards/rejected": -0.4368179738521576,
"step": 2820
},
{
"epoch": 0.75,
"learning_rate": 8.620488984679378e-07,
"logits/chosen": -1.852301836013794,
"logits/rejected": -1.1759949922561646,
"logps/chosen": -659.7407836914062,
"logps/rejected": -1306.58203125,
"loss": 0.066,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16961313784122467,
"rewards/margins": 0.3043574392795563,
"rewards/rejected": -0.47397056221961975,
"step": 2830
},
{
"epoch": 0.76,
"learning_rate": 8.445394716802754e-07,
"logits/chosen": -1.6400655508041382,
"logits/rejected": -1.1373300552368164,
"logps/chosen": -735.5645751953125,
"logps/rejected": -1331.341064453125,
"loss": 0.0866,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1878480315208435,
"rewards/margins": 0.27262082695961,
"rewards/rejected": -0.4604688286781311,
"step": 2840
},
{
"epoch": 0.76,
"learning_rate": 8.271734841028553e-07,
"logits/chosen": -1.4932782649993896,
"logits/rejected": -0.9225826263427734,
"logps/chosen": -594.31982421875,
"logps/rejected": -1214.7799072265625,
"loss": 0.0686,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1466977298259735,
"rewards/margins": 0.30178767442703247,
"rewards/rejected": -0.4484853744506836,
"step": 2850
},
{
"epoch": 0.76,
"learning_rate": 8.099524404308948e-07,
"logits/chosen": -1.5658118724822998,
"logits/rejected": -0.9562853574752808,
"logps/chosen": -645.6381225585938,
"logps/rejected": -1404.0450439453125,
"loss": 0.0439,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.15565729141235352,
"rewards/margins": 0.3485789895057678,
"rewards/rejected": -0.5042362809181213,
"step": 2860
},
{
"epoch": 0.77,
"learning_rate": 7.928778328007918e-07,
"logits/chosen": -1.867717981338501,
"logits/rejected": -0.9512616991996765,
"logps/chosen": -674.2090454101562,
"logps/rejected": -1350.9642333984375,
"loss": 0.0573,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.13902226090431213,
"rewards/margins": 0.3449219763278961,
"rewards/rejected": -0.48394423723220825,
"step": 2870
},
{
"epoch": 0.77,
"learning_rate": 7.759511406608255e-07,
"logits/chosen": -1.5874018669128418,
"logits/rejected": -1.1192893981933594,
"logps/chosen": -586.0441284179688,
"logps/rejected": -1302.760986328125,
"loss": 0.0788,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13455680012702942,
"rewards/margins": 0.2815057635307312,
"rewards/rejected": -0.416062593460083,
"step": 2880
},
{
"epoch": 0.77,
"learning_rate": 7.591738306429769e-07,
"logits/chosen": -1.5606791973114014,
"logits/rejected": -1.0470914840698242,
"logps/chosen": -744.95263671875,
"logps/rejected": -1285.8818359375,
"loss": 0.0792,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.18279020488262177,
"rewards/margins": 0.2564167082309723,
"rewards/rejected": -0.43920689821243286,
"step": 2890
},
{
"epoch": 0.77,
"learning_rate": 7.425473564358457e-07,
"logits/chosen": -1.6006968021392822,
"logits/rejected": -1.0099936723709106,
"logps/chosen": -720.3036499023438,
"logps/rejected": -1374.542236328125,
"loss": 0.0453,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.18247629702091217,
"rewards/margins": 0.3243308663368225,
"rewards/rejected": -0.5068072080612183,
"step": 2900
},
{
"epoch": 0.78,
"learning_rate": 7.260731586586983e-07,
"logits/chosen": -1.6800225973129272,
"logits/rejected": -0.9249428510665894,
"logps/chosen": -472.9820251464844,
"logps/rejected": -1168.8466796875,
"loss": 0.0482,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1155017763376236,
"rewards/margins": 0.33621880412101746,
"rewards/rejected": -0.45172062516212463,
"step": 2910
},
{
"epoch": 0.78,
"learning_rate": 7.097526647366379e-07,
"logits/chosen": -1.7830703258514404,
"logits/rejected": -1.3318331241607666,
"logps/chosen": -588.818359375,
"logps/rejected": -1168.9097900390625,
"loss": 0.0765,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1334531605243683,
"rewards/margins": 0.28530946373939514,
"rewards/rejected": -0.41876259446144104,
"step": 2920
},
{
"epoch": 0.78,
"learning_rate": 6.935872887769299e-07,
"logits/chosen": -1.593977928161621,
"logits/rejected": -1.0323081016540527,
"logps/chosen": -722.8409423828125,
"logps/rejected": -1217.73681640625,
"loss": 0.0888,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1773270219564438,
"rewards/margins": 0.2691357731819153,
"rewards/rejected": -0.4464627802371979,
"step": 2930
},
{
"epoch": 0.78,
"learning_rate": 6.775784314464717e-07,
"logits/chosen": -1.6172115802764893,
"logits/rejected": -1.0769383907318115,
"logps/chosen": -590.12939453125,
"logps/rejected": -1277.718505859375,
"loss": 0.0699,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1619381308555603,
"rewards/margins": 0.30093225836753845,
"rewards/rejected": -0.46287041902542114,
"step": 2940
},
{
"epoch": 0.79,
"learning_rate": 6.617274798504286e-07,
"logits/chosen": -1.748478651046753,
"logits/rejected": -0.9003183245658875,
"logps/chosen": -748.8468017578125,
"logps/rejected": -1387.6318359375,
"loss": 0.0683,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.17095907032489777,
"rewards/margins": 0.3251902461051941,
"rewards/rejected": -0.49614930152893066,
"step": 2950
},
{
"epoch": 0.79,
"learning_rate": 6.460358074120518e-07,
"logits/chosen": -1.5600354671478271,
"logits/rejected": -1.1617937088012695,
"logps/chosen": -635.3890380859375,
"logps/rejected": -1242.238525390625,
"loss": 0.0767,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16152323782444,
"rewards/margins": 0.26770225167274475,
"rewards/rejected": -0.42922544479370117,
"step": 2960
},
{
"epoch": 0.79,
"learning_rate": 6.305047737536707e-07,
"logits/chosen": -1.6008501052856445,
"logits/rejected": -1.1396461725234985,
"logps/chosen": -604.7008666992188,
"logps/rejected": -1217.7623291015625,
"loss": 0.0703,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1483077108860016,
"rewards/margins": 0.29934391379356384,
"rewards/rejected": -0.44765162467956543,
"step": 2970
},
{
"epoch": 0.79,
"learning_rate": 6.151357245788917e-07,
"logits/chosen": -1.6701332330703735,
"logits/rejected": -0.9330341219902039,
"logps/chosen": -628.5372924804688,
"logps/rejected": -1414.3253173828125,
"loss": 0.0577,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.154875710606575,
"rewards/margins": 0.35905200242996216,
"rewards/rejected": -0.513927698135376,
"step": 2980
},
{
"epoch": 0.8,
"learning_rate": 5.999299915559956e-07,
"logits/chosen": -1.4873206615447998,
"logits/rejected": -0.9139396548271179,
"logps/chosen": -534.3379516601562,
"logps/rejected": -1303.0247802734375,
"loss": 0.0602,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1485694944858551,
"rewards/margins": 0.3279011845588684,
"rewards/rejected": -0.4764706492424011,
"step": 2990
},
{
"epoch": 0.8,
"learning_rate": 5.848888922025553e-07,
"logits/chosen": -1.5426976680755615,
"logits/rejected": -1.0285556316375732,
"logps/chosen": -576.2730102539062,
"logps/rejected": -1104.698974609375,
"loss": 0.0913,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.1463908702135086,
"rewards/margins": 0.2579698860645294,
"rewards/rejected": -0.40436071157455444,
"step": 3000
},
{
"epoch": 0.8,
"learning_rate": 5.700137297712749e-07,
"logits/chosen": -1.5507534742355347,
"logits/rejected": -1.1646429300308228,
"logps/chosen": -678.5272216796875,
"logps/rejected": -1307.9720458984375,
"loss": 0.0558,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.19729594886302948,
"rewards/margins": 0.2910037338733673,
"rewards/rejected": -0.488299697637558,
"step": 3010
},
{
"epoch": 0.81,
"learning_rate": 5.553057931370729e-07,
"logits/chosen": -1.7067492008209229,
"logits/rejected": -1.3142516613006592,
"logps/chosen": -630.6482543945312,
"logps/rejected": -1377.3018798828125,
"loss": 0.053,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1730308085680008,
"rewards/margins": 0.33525392413139343,
"rewards/rejected": -0.5082847476005554,
"step": 3020
},
{
"epoch": 0.81,
"learning_rate": 5.407663566854008e-07,
"logits/chosen": -1.5402835607528687,
"logits/rejected": -0.9547150731086731,
"logps/chosen": -717.29833984375,
"logps/rejected": -1315.943115234375,
"loss": 0.0595,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.18772569298744202,
"rewards/margins": 0.29532501101493835,
"rewards/rejected": -0.48305076360702515,
"step": 3030
},
{
"epoch": 0.81,
"learning_rate": 5.263966802018275e-07,
"logits/chosen": -1.8520797491073608,
"logits/rejected": -1.0401822328567505,
"logps/chosen": -717.2581176757812,
"logps/rejected": -1253.8289794921875,
"loss": 0.0639,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16455858945846558,
"rewards/margins": 0.29715651273727417,
"rewards/rejected": -0.4617151618003845,
"step": 3040
},
{
"epoch": 0.81,
"learning_rate": 5.121980087628802e-07,
"logits/chosen": -1.7092409133911133,
"logits/rejected": -1.2182010412216187,
"logps/chosen": -586.7152099609375,
"logps/rejected": -1214.601318359375,
"loss": 0.051,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.137488454580307,
"rewards/margins": 0.3142798840999603,
"rewards/rejected": -0.45176833868026733,
"step": 3050
},
{
"epoch": 0.82,
"learning_rate": 4.981715726281666e-07,
"logits/chosen": -1.5101208686828613,
"logits/rejected": -0.8277866244316101,
"logps/chosen": -569.005615234375,
"logps/rejected": -1231.107177734375,
"loss": 0.0616,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13659389317035675,
"rewards/margins": 0.32126671075820923,
"rewards/rejected": -0.45786052942276,
"step": 3060
},
{
"epoch": 0.82,
"learning_rate": 4.843185871337722e-07,
"logits/chosen": -1.4731261730194092,
"logits/rejected": -1.047473669052124,
"logps/chosen": -693.8050537109375,
"logps/rejected": -1432.60009765625,
"loss": 0.0881,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19450101256370544,
"rewards/margins": 0.3186071217060089,
"rewards/rejected": -0.5131081342697144,
"step": 3070
},
{
"epoch": 0.82,
"learning_rate": 4.706402525869633e-07,
"logits/chosen": -1.3473070859909058,
"logits/rejected": -1.2000024318695068,
"logps/chosen": -589.5697021484375,
"logps/rejected": -1317.3638916015625,
"loss": 0.072,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16462837159633636,
"rewards/margins": 0.2750544846057892,
"rewards/rejected": -0.43968287110328674,
"step": 3080
},
{
"epoch": 0.82,
"learning_rate": 4.5713775416217884e-07,
"logits/chosen": -1.6565711498260498,
"logits/rejected": -1.1259280443191528,
"logps/chosen": -551.4010009765625,
"logps/rejected": -1178.7850341796875,
"loss": 0.0716,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.13680130243301392,
"rewards/margins": 0.30534160137176514,
"rewards/rejected": -0.44214290380477905,
"step": 3090
},
{
"epoch": 0.83,
"learning_rate": 4.438122617983442e-07,
"logits/chosen": -1.5101524591445923,
"logits/rejected": -0.8424856066703796,
"logps/chosen": -591.2506103515625,
"logps/rejected": -1273.046142578125,
"loss": 0.0566,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.13511498272418976,
"rewards/margins": 0.33536994457244873,
"rewards/rejected": -0.4704849123954773,
"step": 3100
},
{
"epoch": 0.83,
"learning_rate": 4.3066493009749853e-07,
"logits/chosen": -1.7471688985824585,
"logits/rejected": -1.1047070026397705,
"logps/chosen": -767.1864013671875,
"logps/rejected": -1380.8409423828125,
"loss": 0.053,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.20385906100273132,
"rewards/margins": 0.27676570415496826,
"rewards/rejected": -0.4806247651576996,
"step": 3110
},
{
"epoch": 0.83,
"learning_rate": 4.1769689822475147e-07,
"logits/chosen": -1.4887802600860596,
"logits/rejected": -1.02260422706604,
"logps/chosen": -691.14990234375,
"logps/rejected": -1373.417724609375,
"loss": 0.0965,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.20843443274497986,
"rewards/margins": 0.2790736258029938,
"rewards/rejected": -0.4875081181526184,
"step": 3120
},
{
"epoch": 0.83,
"learning_rate": 4.049092898095816e-07,
"logits/chosen": -1.5980995893478394,
"logits/rejected": -1.023189902305603,
"logps/chosen": -595.4110107421875,
"logps/rejected": -1312.7945556640625,
"loss": 0.0653,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16371554136276245,
"rewards/margins": 0.2928038537502289,
"rewards/rejected": -0.4565194249153137,
"step": 3130
},
{
"epoch": 0.84,
"learning_rate": 3.9230321284847856e-07,
"logits/chosen": -1.5363506078720093,
"logits/rejected": -1.0504465103149414,
"logps/chosen": -489.7978515625,
"logps/rejected": -1119.5257568359375,
"loss": 0.0633,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.11015214771032333,
"rewards/margins": 0.2800517678260803,
"rewards/rejected": -0.39020389318466187,
"step": 3140
},
{
"epoch": 0.84,
"learning_rate": 3.798797596089351e-07,
"logits/chosen": -1.6714694499969482,
"logits/rejected": -1.125106692314148,
"logps/chosen": -604.4114379882812,
"logps/rejected": -1297.808349609375,
"loss": 0.0516,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.12939824163913727,
"rewards/margins": 0.35474586486816406,
"rewards/rejected": -0.4841441214084625,
"step": 3150
},
{
"epoch": 0.84,
"learning_rate": 3.6764000653481263e-07,
"logits/chosen": -1.7083511352539062,
"logits/rejected": -1.0139703750610352,
"logps/chosen": -604.1871337890625,
"logps/rejected": -1156.0162353515625,
"loss": 0.0919,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1465051770210266,
"rewards/margins": 0.27671653032302856,
"rewards/rejected": -0.4232216775417328,
"step": 3160
},
{
"epoch": 0.85,
"learning_rate": 3.555850141530659e-07,
"logits/chosen": -1.7717567682266235,
"logits/rejected": -1.1824371814727783,
"logps/chosen": -566.564453125,
"logps/rejected": -1101.685302734375,
"loss": 0.0733,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.14669661223888397,
"rewards/margins": 0.26966187357902527,
"rewards/rejected": -0.41635847091674805,
"step": 3170
},
{
"epoch": 0.85,
"learning_rate": 3.4371582698185636e-07,
"logits/chosen": -1.4069181680679321,
"logits/rejected": -1.0164799690246582,
"logps/chosen": -516.7618408203125,
"logps/rejected": -1247.140380859375,
"loss": 0.0714,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1548655778169632,
"rewards/margins": 0.2954062819480896,
"rewards/rejected": -0.4502718448638916,
"step": 3180
},
{
"epoch": 0.85,
"learning_rate": 3.3203347344004737e-07,
"logits/chosen": -1.332287073135376,
"logits/rejected": -1.1523934602737427,
"logps/chosen": -562.6694946289062,
"logps/rejected": -1197.0635986328125,
"loss": 0.0921,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.19367225468158722,
"rewards/margins": 0.24201972782611847,
"rewards/rejected": -0.4356919825077057,
"step": 3190
},
{
"epoch": 0.85,
"learning_rate": 3.2053896575809426e-07,
"logits/chosen": -1.4262222051620483,
"logits/rejected": -1.026829719543457,
"logps/chosen": -569.9625854492188,
"logps/rejected": -1329.806884765625,
"loss": 0.0553,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.15987573564052582,
"rewards/margins": 0.3234714865684509,
"rewards/rejected": -0.48334717750549316,
"step": 3200
},
{
"epoch": 0.86,
"learning_rate": 3.092332998903416e-07,
"logits/chosen": -1.403305172920227,
"logits/rejected": -1.228492021560669,
"logps/chosen": -531.6251220703125,
"logps/rejected": -1268.7862548828125,
"loss": 0.0586,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14155253767967224,
"rewards/margins": 0.29640236496925354,
"rewards/rejected": -0.4379549026489258,
"step": 3210
},
{
"epoch": 0.86,
"learning_rate": 2.981174554287239e-07,
"logits/chosen": -1.4185243844985962,
"logits/rejected": -1.2006936073303223,
"logps/chosen": -635.041259765625,
"logps/rejected": -1473.925537109375,
"loss": 0.0452,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -0.16828233003616333,
"rewards/margins": 0.3489801287651062,
"rewards/rejected": -0.5172623991966248,
"step": 3220
},
{
"epoch": 0.86,
"learning_rate": 2.871923955178918e-07,
"logits/chosen": -1.4334993362426758,
"logits/rejected": -0.9838508367538452,
"logps/chosen": -607.728271484375,
"logps/rejected": -1310.9564208984375,
"loss": 0.0531,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.18386729061603546,
"rewards/margins": 0.3068189322948456,
"rewards/rejected": -0.49068623781204224,
"step": 3230
},
{
"epoch": 0.86,
"learning_rate": 2.764590667717562e-07,
"logits/chosen": -1.6019694805145264,
"logits/rejected": -1.1419106721878052,
"logps/chosen": -588.9683837890625,
"logps/rejected": -1284.370849609375,
"loss": 0.0586,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.15091760456562042,
"rewards/margins": 0.30733171105384827,
"rewards/rejected": -0.4582493305206299,
"step": 3240
},
{
"epoch": 0.87,
"learning_rate": 2.6591839919146963e-07,
"logits/chosen": -1.6173986196517944,
"logits/rejected": -1.1335291862487793,
"logps/chosen": -654.0939331054688,
"logps/rejected": -1284.885986328125,
"loss": 0.0572,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1872612088918686,
"rewards/margins": 0.3091946244239807,
"rewards/rejected": -0.4964558482170105,
"step": 3250
},
{
"epoch": 0.87,
"learning_rate": 2.555713060848433e-07,
"logits/chosen": -1.4421133995056152,
"logits/rejected": -1.1747627258300781,
"logps/chosen": -585.655517578125,
"logps/rejected": -1284.943115234375,
"loss": 0.0449,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1650674194097519,
"rewards/margins": 0.31287819147109985,
"rewards/rejected": -0.47794562578201294,
"step": 3260
},
{
"epoch": 0.87,
"learning_rate": 2.454186839872158e-07,
"logits/chosen": -1.7081034183502197,
"logits/rejected": -0.94395911693573,
"logps/chosen": -646.4754028320312,
"logps/rejected": -1342.545166015625,
"loss": 0.0709,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.18191269040107727,
"rewards/margins": 0.326556921005249,
"rewards/rejected": -0.5084696412086487,
"step": 3270
},
{
"epoch": 0.87,
"learning_rate": 2.3546141258376786e-07,
"logits/chosen": -1.3628816604614258,
"logits/rejected": -0.9335969686508179,
"logps/chosen": -603.8056640625,
"logps/rejected": -1245.411376953125,
"loss": 0.0594,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1723943054676056,
"rewards/margins": 0.3125647008419037,
"rewards/rejected": -0.48495903611183167,
"step": 3280
},
{
"epoch": 0.88,
"learning_rate": 2.257003546333042e-07,
"logits/chosen": -1.6881519556045532,
"logits/rejected": -1.116393804550171,
"logps/chosen": -680.4428100585938,
"logps/rejected": -1417.369384765625,
"loss": 0.0578,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.21765998005867004,
"rewards/margins": 0.3092319667339325,
"rewards/rejected": -0.5268920063972473,
"step": 3290
},
{
"epoch": 0.88,
"learning_rate": 2.1613635589349756e-07,
"logits/chosen": -1.7809503078460693,
"logits/rejected": -1.235079050064087,
"logps/chosen": -521.1368408203125,
"logps/rejected": -1202.243408203125,
"loss": 0.0625,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.13693402707576752,
"rewards/margins": 0.2993922233581543,
"rewards/rejected": -0.43632620573043823,
"step": 3300
},
{
"epoch": 0.88,
"learning_rate": 2.0677024504760752e-07,
"logits/chosen": -1.5252535343170166,
"logits/rejected": -1.2412452697753906,
"logps/chosen": -526.3180541992188,
"logps/rejected": -1315.9249267578125,
"loss": 0.0633,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13880392909049988,
"rewards/margins": 0.35248517990112305,
"rewards/rejected": -0.4912891387939453,
"step": 3310
},
{
"epoch": 0.89,
"learning_rate": 1.9760283363267684e-07,
"logits/chosen": -1.778957724571228,
"logits/rejected": -1.0381158590316772,
"logps/chosen": -641.0303955078125,
"logps/rejected": -1229.374267578125,
"loss": 0.0791,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.13701531291007996,
"rewards/margins": 0.3019945025444031,
"rewards/rejected": -0.43900981545448303,
"step": 3320
},
{
"epoch": 0.89,
"learning_rate": 1.8863491596921745e-07,
"logits/chosen": -1.4902263879776,
"logits/rejected": -0.8488144874572754,
"logps/chosen": -537.8624267578125,
"logps/rejected": -1138.487060546875,
"loss": 0.0651,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.12021678686141968,
"rewards/margins": 0.3042137622833252,
"rewards/rejected": -0.4244305491447449,
"step": 3330
},
{
"epoch": 0.89,
"learning_rate": 1.798672690923828e-07,
"logits/chosen": -1.604090929031372,
"logits/rejected": -1.0566972494125366,
"logps/chosen": -546.621826171875,
"logps/rejected": -1305.0101318359375,
"loss": 0.048,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1309491991996765,
"rewards/margins": 0.3338525891304016,
"rewards/rejected": -0.4648017883300781,
"step": 3340
},
{
"epoch": 0.89,
"learning_rate": 1.713006526846439e-07,
"logits/chosen": -1.6890567541122437,
"logits/rejected": -1.0597703456878662,
"logps/chosen": -638.17626953125,
"logps/rejected": -1403.3868408203125,
"loss": 0.0422,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.14803606271743774,
"rewards/margins": 0.36922168731689453,
"rewards/rejected": -0.517257809638977,
"step": 3350
},
{
"epoch": 0.9,
"learning_rate": 1.629358090099639e-07,
"logits/chosen": -1.772783637046814,
"logits/rejected": -1.145141363143921,
"logps/chosen": -645.2138671875,
"logps/rejected": -1247.31005859375,
"loss": 0.0995,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.16153082251548767,
"rewards/margins": 0.27567344903945923,
"rewards/rejected": -0.4372042715549469,
"step": 3360
},
{
"epoch": 0.9,
"learning_rate": 1.5477346284948292e-07,
"logits/chosen": -1.6137062311172485,
"logits/rejected": -1.2323498725891113,
"logps/chosen": -652.45068359375,
"logps/rejected": -1297.9696044921875,
"loss": 0.1,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.18204265832901,
"rewards/margins": 0.2792537212371826,
"rewards/rejected": -0.4612963795661926,
"step": 3370
},
{
"epoch": 0.9,
"learning_rate": 1.4681432143872133e-07,
"logits/chosen": -1.5417966842651367,
"logits/rejected": -0.92780601978302,
"logps/chosen": -655.7986450195312,
"logps/rejected": -1352.8719482421875,
"loss": 0.0452,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1652592271566391,
"rewards/margins": 0.32993632555007935,
"rewards/rejected": -0.49519556760787964,
"step": 3380
},
{
"epoch": 0.9,
"learning_rate": 1.3905907440629752e-07,
"logits/chosen": -1.7428719997406006,
"logits/rejected": -1.234071969985962,
"logps/chosen": -683.9173583984375,
"logps/rejected": -1301.064208984375,
"loss": 0.0749,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1897701770067215,
"rewards/margins": 0.26576271653175354,
"rewards/rejected": -0.45553287863731384,
"step": 3390
},
{
"epoch": 0.91,
"learning_rate": 1.31508393714177e-07,
"logits/chosen": -1.6525071859359741,
"logits/rejected": -1.269195795059204,
"logps/chosen": -573.1322631835938,
"logps/rejected": -1224.781005859375,
"loss": 0.0717,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1454993188381195,
"rewards/margins": 0.29666373133659363,
"rewards/rejected": -0.44216299057006836,
"step": 3400
},
{
"epoch": 0.91,
"learning_rate": 1.241629335994471e-07,
"logits/chosen": -1.4832892417907715,
"logits/rejected": -0.7595690488815308,
"logps/chosen": -546.4700927734375,
"logps/rejected": -1105.534912109375,
"loss": 0.0961,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.12580379843711853,
"rewards/margins": 0.2655871510505676,
"rewards/rejected": -0.39139097929000854,
"step": 3410
},
{
"epoch": 0.91,
"learning_rate": 1.1702333051763271e-07,
"logits/chosen": -1.729821801185608,
"logits/rejected": -1.004158854484558,
"logps/chosen": -648.3139038085938,
"logps/rejected": -1181.37939453125,
"loss": 0.0907,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.17357492446899414,
"rewards/margins": 0.2692447900772095,
"rewards/rejected": -0.4428196847438812,
"step": 3420
},
{
"epoch": 0.91,
"learning_rate": 1.1009020308754587e-07,
"logits/chosen": -1.6045188903808594,
"logits/rejected": -1.1483290195465088,
"logps/chosen": -791.281494140625,
"logps/rejected": -1432.56591796875,
"loss": 0.0588,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.20312626659870148,
"rewards/margins": 0.30023887753486633,
"rewards/rejected": -0.503365159034729,
"step": 3430
},
{
"epoch": 0.92,
"learning_rate": 1.0336415203768962e-07,
"logits/chosen": -1.4623007774353027,
"logits/rejected": -0.9743086695671082,
"logps/chosen": -684.1002807617188,
"logps/rejected": -1355.83935546875,
"loss": 0.0854,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15205714106559753,
"rewards/margins": 0.3153363764286041,
"rewards/rejected": -0.4673934876918793,
"step": 3440
},
{
"epoch": 0.92,
"learning_rate": 9.684576015420277e-08,
"logits/chosen": -1.5630353689193726,
"logits/rejected": -1.1269890069961548,
"logps/chosen": -500.76934814453125,
"logps/rejected": -1174.6031494140625,
"loss": 0.0748,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13523253798484802,
"rewards/margins": 0.2928512990474701,
"rewards/rejected": -0.4280838370323181,
"step": 3450
},
{
"epoch": 0.92,
"learning_rate": 9.053559223036746e-08,
"logits/chosen": -1.4134055376052856,
"logits/rejected": -0.9375116229057312,
"logps/chosen": -684.4464111328125,
"logps/rejected": -1184.0179443359375,
"loss": 0.097,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1784881353378296,
"rewards/margins": 0.25100329518318176,
"rewards/rejected": -0.42949143052101135,
"step": 3460
},
{
"epoch": 0.93,
"learning_rate": 8.44341950176683e-08,
"logits/chosen": -1.7058902978897095,
"logits/rejected": -1.1464042663574219,
"logps/chosen": -518.6105346679688,
"logps/rejected": -1139.312744140625,
"loss": 0.0706,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11180339008569717,
"rewards/margins": 0.3046559691429138,
"rewards/rejected": -0.4164593815803528,
"step": 3470
},
{
"epoch": 0.93,
"learning_rate": 7.854209717842231e-08,
"logits/chosen": -1.7066657543182373,
"logits/rejected": -1.1052082777023315,
"logps/chosen": -634.6749267578125,
"logps/rejected": -1268.662353515625,
"loss": 0.0549,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1490454524755478,
"rewards/margins": 0.32269707322120667,
"rewards/rejected": -0.4717424809932709,
"step": 3480
},
{
"epoch": 0.93,
"learning_rate": 7.285980923996989e-08,
"logits/chosen": -1.583821415901184,
"logits/rejected": -0.9543322324752808,
"logps/chosen": -620.9078979492188,
"logps/rejected": -1256.222900390625,
"loss": 0.0499,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.14574182033538818,
"rewards/margins": 0.3196600079536438,
"rewards/rejected": -0.465401828289032,
"step": 3490
},
{
"epoch": 0.93,
"learning_rate": 6.738782355044048e-08,
"logits/chosen": -1.5047067403793335,
"logits/rejected": -1.0737859010696411,
"logps/chosen": -520.77294921875,
"logps/rejected": -1155.83740234375,
"loss": 0.0764,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1430043876171112,
"rewards/margins": 0.28466781973838806,
"rewards/rejected": -0.42767223715782166,
"step": 3500
},
{
"epoch": 0.94,
"learning_rate": 6.212661423609184e-08,
"logits/chosen": -1.3737636804580688,
"logits/rejected": -1.1867458820343018,
"logps/chosen": -623.9158935546875,
"logps/rejected": -1239.9605712890625,
"loss": 0.078,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1589566022157669,
"rewards/margins": 0.27907633781433105,
"rewards/rejected": -0.43803295493125916,
"step": 3510
},
{
"epoch": 0.94,
"learning_rate": 5.707663716023021e-08,
"logits/chosen": -1.5748332738876343,
"logits/rejected": -1.063622236251831,
"logps/chosen": -616.1275634765625,
"logps/rejected": -1155.8143310546875,
"loss": 0.0835,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.15737947821617126,
"rewards/margins": 0.2580162584781647,
"rewards/rejected": -0.4153957962989807,
"step": 3520
},
{
"epoch": 0.94,
"learning_rate": 5.22383298837098e-08,
"logits/chosen": -1.6702196598052979,
"logits/rejected": -0.9411935806274414,
"logps/chosen": -713.735107421875,
"logps/rejected": -1223.2811279296875,
"loss": 0.0824,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.17610540986061096,
"rewards/margins": 0.29836469888687134,
"rewards/rejected": -0.4744700491428375,
"step": 3530
},
{
"epoch": 0.94,
"learning_rate": 4.761211162702117e-08,
"logits/chosen": -1.500201940536499,
"logits/rejected": -1.2098219394683838,
"logps/chosen": -633.7230224609375,
"logps/rejected": -1120.678466796875,
"loss": 0.0984,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.17905206978321075,
"rewards/margins": 0.2320644110441208,
"rewards/rejected": -0.41111645102500916,
"step": 3540
},
{
"epoch": 0.95,
"learning_rate": 4.319838323396691e-08,
"logits/chosen": -1.6483008861541748,
"logits/rejected": -1.2190577983856201,
"logps/chosen": -614.4102783203125,
"logps/rejected": -1260.265625,
"loss": 0.0958,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.16128340363502502,
"rewards/margins": 0.26488471031188965,
"rewards/rejected": -0.4261681139469147,
"step": 3550
},
{
"epoch": 0.95,
"learning_rate": 3.8997527136930004e-08,
"logits/chosen": -1.429495930671692,
"logits/rejected": -1.0348641872406006,
"logps/chosen": -632.0252685546875,
"logps/rejected": -1319.2584228515625,
"loss": 0.0798,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.17518481612205505,
"rewards/margins": 0.2713126540184021,
"rewards/rejected": -0.44649749994277954,
"step": 3560
},
{
"epoch": 0.95,
"learning_rate": 3.5009907323737826e-08,
"logits/chosen": -1.551414132118225,
"logits/rejected": -1.2644479274749756,
"logps/chosen": -570.0910034179688,
"logps/rejected": -1210.5382080078125,
"loss": 0.0803,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1442852020263672,
"rewards/margins": 0.26128098368644714,
"rewards/rejected": -0.4055662155151367,
"step": 3570
},
{
"epoch": 0.95,
"learning_rate": 3.1235869306123766e-08,
"logits/chosen": -1.8781211376190186,
"logits/rejected": -1.116159439086914,
"logps/chosen": -654.27685546875,
"logps/rejected": -1360.783447265625,
"loss": 0.0586,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.15850645303726196,
"rewards/margins": 0.32171380519866943,
"rewards/rejected": -0.480220228433609,
"step": 3580
},
{
"epoch": 0.96,
"learning_rate": 2.767574008979007e-08,
"logits/chosen": -1.7552858591079712,
"logits/rejected": -1.1240081787109375,
"logps/chosen": -649.2583618164062,
"logps/rejected": -1285.60986328125,
"loss": 0.0746,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1459151804447174,
"rewards/margins": 0.29276043176651,
"rewards/rejected": -0.43867558240890503,
"step": 3590
},
{
"epoch": 0.96,
"learning_rate": 2.4329828146074096e-08,
"logits/chosen": -1.680267333984375,
"logits/rejected": -0.990991473197937,
"logps/chosen": -702.8885498046875,
"logps/rejected": -1430.7728271484375,
"loss": 0.0422,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.16736450791358948,
"rewards/margins": 0.3372777998447418,
"rewards/rejected": -0.5046423673629761,
"step": 3600
},
{
"epoch": 0.96,
"learning_rate": 2.1198423385220822e-08,
"logits/chosen": -1.5231233835220337,
"logits/rejected": -0.983518123626709,
"logps/chosen": -574.9891967773438,
"logps/rejected": -1192.521728515625,
"loss": 0.091,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.13129422068595886,
"rewards/margins": 0.2925337255001068,
"rewards/rejected": -0.42382797598838806,
"step": 3610
},
{
"epoch": 0.97,
"learning_rate": 1.82817971312621e-08,
"logits/chosen": -1.689100980758667,
"logits/rejected": -1.1573827266693115,
"logps/chosen": -602.2514038085938,
"logps/rejected": -1330.6248779296875,
"loss": 0.0539,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1489819586277008,
"rewards/margins": 0.32134801149368286,
"rewards/rejected": -0.47032999992370605,
"step": 3620
},
{
"epoch": 0.97,
"learning_rate": 1.5580202098509078e-08,
"logits/chosen": -1.7418603897094727,
"logits/rejected": -1.2287579774856567,
"logps/chosen": -539.5288696289062,
"logps/rejected": -1092.644775390625,
"loss": 0.084,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.1297844648361206,
"rewards/margins": 0.25757455825805664,
"rewards/rejected": -0.38735905289649963,
"step": 3630
},
{
"epoch": 0.97,
"learning_rate": 1.3093872369654148e-08,
"logits/chosen": -1.6315152645111084,
"logits/rejected": -0.7913056015968323,
"logps/chosen": -566.9249267578125,
"logps/rejected": -1232.651611328125,
"loss": 0.0596,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.15806104242801666,
"rewards/margins": 0.31945645809173584,
"rewards/rejected": -0.4775174558162689,
"step": 3640
},
{
"epoch": 0.97,
"learning_rate": 1.0823023375489128e-08,
"logits/chosen": -1.8176482915878296,
"logits/rejected": -1.307217001914978,
"logps/chosen": -596.6251220703125,
"logps/rejected": -1259.1907958984375,
"loss": 0.0741,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13308800756931305,
"rewards/margins": 0.29899168014526367,
"rewards/rejected": -0.4320797026157379,
"step": 3650
},
{
"epoch": 0.98,
"learning_rate": 8.767851876239075e-09,
"logits/chosen": -1.3138879537582397,
"logits/rejected": -1.0358647108078003,
"logps/chosen": -606.9632568359375,
"logps/rejected": -1166.726806640625,
"loss": 0.126,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.17711606621742249,
"rewards/margins": 0.24026331305503845,
"rewards/rejected": -0.41737937927246094,
"step": 3660
},
{
"epoch": 0.98,
"learning_rate": 6.9285359445145366e-09,
"logits/chosen": -1.6356595754623413,
"logits/rejected": -1.0119235515594482,
"logps/chosen": -606.1302490234375,
"logps/rejected": -1273.985595703125,
"loss": 0.0883,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15847013890743256,
"rewards/margins": 0.29371243715286255,
"rewards/rejected": -0.4521825909614563,
"step": 3670
},
{
"epoch": 0.98,
"learning_rate": 5.305234949880001e-09,
"logits/chosen": -1.659597396850586,
"logits/rejected": -1.0647908449172974,
"logps/chosen": -593.581787109375,
"logps/rejected": -1240.505615234375,
"loss": 0.0669,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15421947836875916,
"rewards/margins": 0.2849588096141815,
"rewards/rejected": -0.4391782879829407,
"step": 3680
},
{
"epoch": 0.98,
"learning_rate": 3.8980895450474455e-09,
"logits/chosen": -1.4947640895843506,
"logits/rejected": -1.1128860712051392,
"logps/chosen": -551.9004516601562,
"logps/rejected": -1167.104736328125,
"loss": 0.053,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13484862446784973,
"rewards/margins": 0.2813241481781006,
"rewards/rejected": -0.4161728024482727,
"step": 3690
},
{
"epoch": 0.99,
"learning_rate": 2.7072216536885855e-09,
"logits/chosen": -1.6360046863555908,
"logits/rejected": -1.196569800376892,
"logps/chosen": -659.7826538085938,
"logps/rejected": -1305.6195068359375,
"loss": 0.0578,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.17655882239341736,
"rewards/margins": 0.3209526538848877,
"rewards/rejected": -0.49751147627830505,
"step": 3700
},
{
"epoch": 0.99,
"learning_rate": 1.7327344598702667e-09,
"logits/chosen": -1.4936860799789429,
"logits/rejected": -1.0485936403274536,
"logps/chosen": -584.678955078125,
"logps/rejected": -1314.91650390625,
"loss": 0.071,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14625641703605652,
"rewards/margins": 0.3208056390285492,
"rewards/rejected": -0.4670620858669281,
"step": 3710
},
{
"epoch": 0.99,
"learning_rate": 9.747123991141193e-10,
"logits/chosen": -1.4201769828796387,
"logits/rejected": -1.045179009437561,
"logps/chosen": -620.8831787109375,
"logps/rejected": -1206.128662109375,
"loss": 0.0691,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15646150708198547,
"rewards/margins": 0.29528263211250305,
"rewards/rejected": -0.4517441391944885,
"step": 3720
},
{
"epoch": 0.99,
"learning_rate": 4.332211510807427e-10,
"logits/chosen": -1.530029296875,
"logits/rejected": -1.3592134714126587,
"logps/chosen": -523.2725219726562,
"logps/rejected": -1284.657958984375,
"loss": 0.0589,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.13241766393184662,
"rewards/margins": 0.30461427569389343,
"rewards/rejected": -0.43703192472457886,
"step": 3730
},
{
"epoch": 1.0,
"learning_rate": 1.0830763387897902e-10,
"logits/chosen": -1.5114389657974243,
"logits/rejected": -0.964741587638855,
"logps/chosen": -626.4848022460938,
"logps/rejected": -1381.5963134765625,
"loss": 0.0456,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.16040828824043274,
"rewards/margins": 0.3570694923400879,
"rewards/rejected": -0.517477810382843,
"step": 3740
},
{
"epoch": 1.0,
"learning_rate": 0.0,
"logits/chosen": -1.877239465713501,
"logits/rejected": -1.3499794006347656,
"logps/chosen": -648.46875,
"logps/rejected": -1231.8519287109375,
"loss": 0.0717,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.18231454491615295,
"rewards/margins": 0.2813524603843689,
"rewards/rejected": -0.46366700530052185,
"step": 3750
},
{
"epoch": 1.0,
"step": 3750,
"total_flos": 0.0,
"train_loss": 0.0800992045879364,
"train_runtime": 15706.9626,
"train_samples_per_second": 0.955,
"train_steps_per_second": 0.239
}
],
"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
}