zephyr-7b-gpo-v3-3-i2 / trainer_state.json
lole25's picture
Model save
82b4cfc 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": -2.2416834831237793,
"logits/rejected": -2.1367297172546387,
"logps/chosen": -309.5174865722656,
"logps/rejected": -533.4085693359375,
"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": -1.8251832723617554,
"logits/rejected": -1.0621190071105957,
"logps/chosen": -543.5095825195312,
"logps/rejected": -825.732177734375,
"loss": 0.1824,
"rewards/accuracies": 0.4722222089767456,
"rewards/chosen": 0.00025084687513299286,
"rewards/margins": 0.0002259216271340847,
"rewards/rejected": 2.4925222533056512e-05,
"step": 10
},
{
"epoch": 0.01,
"learning_rate": 2.666666666666667e-07,
"logits/chosen": -1.6684499979019165,
"logits/rejected": -1.289470911026001,
"logps/chosen": -504.35736083984375,
"logps/rejected": -862.5172729492188,
"loss": 0.214,
"rewards/accuracies": 0.625,
"rewards/chosen": 0.00039741364889778197,
"rewards/margins": 0.0008992180228233337,
"rewards/rejected": -0.0005018044030293822,
"step": 20
},
{
"epoch": 0.01,
"learning_rate": 4.0000000000000003e-07,
"logits/chosen": -1.448547601699829,
"logits/rejected": -1.2029615640640259,
"logps/chosen": -427.83740234375,
"logps/rejected": -854.1618041992188,
"loss": 0.2082,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.0018184988293796778,
"rewards/margins": 0.002095351228490472,
"rewards/rejected": -0.000276852457318455,
"step": 30
},
{
"epoch": 0.01,
"learning_rate": 5.333333333333335e-07,
"logits/chosen": -1.6658061742782593,
"logits/rejected": -1.1971313953399658,
"logps/chosen": -429.181884765625,
"logps/rejected": -865.9625854492188,
"loss": 0.193,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": 0.005454375874251127,
"rewards/margins": 0.008814454078674316,
"rewards/rejected": -0.0033600772731006145,
"step": 40
},
{
"epoch": 0.01,
"learning_rate": 6.666666666666667e-07,
"logits/chosen": -1.4225283861160278,
"logits/rejected": -1.2158631086349487,
"logps/chosen": -430.7893981933594,
"logps/rejected": -787.4990234375,
"loss": 0.1899,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": 0.007644618395715952,
"rewards/margins": 0.01243924256414175,
"rewards/rejected": -0.004794624168425798,
"step": 50
},
{
"epoch": 0.02,
"learning_rate": 8.000000000000001e-07,
"logits/chosen": -1.47100830078125,
"logits/rejected": -0.9200455546379089,
"logps/chosen": -466.3695373535156,
"logps/rejected": -906.1533203125,
"loss": 0.1733,
"rewards/accuracies": 0.875,
"rewards/chosen": 0.008568339981138706,
"rewards/margins": 0.03259299322962761,
"rewards/rejected": -0.024024656042456627,
"step": 60
},
{
"epoch": 0.02,
"learning_rate": 9.333333333333334e-07,
"logits/chosen": -1.579919457435608,
"logits/rejected": -0.9830185174942017,
"logps/chosen": -454.715087890625,
"logps/rejected": -852.3912353515625,
"loss": 0.1825,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.01390122901648283,
"rewards/margins": 0.05136318877339363,
"rewards/rejected": -0.037461958825588226,
"step": 70
},
{
"epoch": 0.02,
"learning_rate": 1.066666666666667e-06,
"logits/chosen": -1.6900907754898071,
"logits/rejected": -1.3661834001541138,
"logps/chosen": -492.48748779296875,
"logps/rejected": -974.8717651367188,
"loss": 0.1312,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": 0.010218140669167042,
"rewards/margins": 0.07319202274084091,
"rewards/rejected": -0.06297388672828674,
"step": 80
},
{
"epoch": 0.02,
"learning_rate": 1.2000000000000002e-06,
"logits/chosen": -1.6498056650161743,
"logits/rejected": -1.0119965076446533,
"logps/chosen": -440.96490478515625,
"logps/rejected": -855.2099609375,
"loss": 0.1564,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": 0.006226236931979656,
"rewards/margins": 0.08165968954563141,
"rewards/rejected": -0.07543345540761948,
"step": 90
},
{
"epoch": 0.03,
"learning_rate": 1.3333333333333334e-06,
"logits/chosen": -1.9902693033218384,
"logits/rejected": -1.1803243160247803,
"logps/chosen": -578.6461791992188,
"logps/rejected": -930.4148559570312,
"loss": 0.1433,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.012515301816165447,
"rewards/margins": 0.10045032203197479,
"rewards/rejected": -0.11296562105417252,
"step": 100
},
{
"epoch": 0.03,
"learning_rate": 1.4666666666666669e-06,
"logits/chosen": -1.6414697170257568,
"logits/rejected": -0.9264998435974121,
"logps/chosen": -572.1052856445312,
"logps/rejected": -975.2772216796875,
"loss": 0.117,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.08218317478895187,
"rewards/margins": 0.11511914432048798,
"rewards/rejected": -0.19730232656002045,
"step": 110
},
{
"epoch": 0.03,
"learning_rate": 1.6000000000000001e-06,
"logits/chosen": -1.7165985107421875,
"logits/rejected": -1.1280840635299683,
"logps/chosen": -655.0675048828125,
"logps/rejected": -1007.7223510742188,
"loss": 0.1456,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.11872265487909317,
"rewards/margins": 0.10497407615184784,
"rewards/rejected": -0.2236967384815216,
"step": 120
},
{
"epoch": 0.03,
"learning_rate": 1.7333333333333336e-06,
"logits/chosen": -1.721414566040039,
"logits/rejected": -1.0421576499938965,
"logps/chosen": -637.6653442382812,
"logps/rejected": -1049.671630859375,
"loss": 0.1214,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.1268424093723297,
"rewards/margins": 0.13156402111053467,
"rewards/rejected": -0.2584064304828644,
"step": 130
},
{
"epoch": 0.04,
"learning_rate": 1.8666666666666669e-06,
"logits/chosen": -1.8063952922821045,
"logits/rejected": -1.1648313999176025,
"logps/chosen": -533.8194580078125,
"logps/rejected": -1127.229248046875,
"loss": 0.0913,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.06718692928552628,
"rewards/margins": 0.20027296245098114,
"rewards/rejected": -0.2674598693847656,
"step": 140
},
{
"epoch": 0.04,
"learning_rate": 2.0000000000000003e-06,
"logits/chosen": -1.8062896728515625,
"logits/rejected": -1.1342815160751343,
"logps/chosen": -531.7640380859375,
"logps/rejected": -1056.5418701171875,
"loss": 0.1041,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09493765234947205,
"rewards/margins": 0.18374750018119812,
"rewards/rejected": -0.27868515253067017,
"step": 150
},
{
"epoch": 0.04,
"learning_rate": 2.133333333333334e-06,
"logits/chosen": -1.5977694988250732,
"logits/rejected": -1.2945902347564697,
"logps/chosen": -567.9004516601562,
"logps/rejected": -1177.0767822265625,
"loss": 0.0735,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.10163917392492294,
"rewards/margins": 0.22462336719036102,
"rewards/rejected": -0.32626253366470337,
"step": 160
},
{
"epoch": 0.05,
"learning_rate": 2.266666666666667e-06,
"logits/chosen": -1.5195014476776123,
"logits/rejected": -0.7815272212028503,
"logps/chosen": -565.3408203125,
"logps/rejected": -1178.865478515625,
"loss": 0.0674,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.09116648882627487,
"rewards/margins": 0.2727690637111664,
"rewards/rejected": -0.36393555998802185,
"step": 170
},
{
"epoch": 0.05,
"learning_rate": 2.4000000000000003e-06,
"logits/chosen": -1.8510783910751343,
"logits/rejected": -1.1616629362106323,
"logps/chosen": -542.5670166015625,
"logps/rejected": -1118.016357421875,
"loss": 0.0656,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.09122536331415176,
"rewards/margins": 0.22348365187644958,
"rewards/rejected": -0.3147090673446655,
"step": 180
},
{
"epoch": 0.05,
"learning_rate": 2.5333333333333338e-06,
"logits/chosen": -1.9452266693115234,
"logits/rejected": -1.231264352798462,
"logps/chosen": -637.9483032226562,
"logps/rejected": -1175.119140625,
"loss": 0.1008,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.12329447269439697,
"rewards/margins": 0.2058517038822174,
"rewards/rejected": -0.3291461765766144,
"step": 190
},
{
"epoch": 0.05,
"learning_rate": 2.666666666666667e-06,
"logits/chosen": -1.4857038259506226,
"logits/rejected": -0.7614498138427734,
"logps/chosen": -659.7203979492188,
"logps/rejected": -1250.89892578125,
"loss": 0.1015,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16732044517993927,
"rewards/margins": 0.22920766472816467,
"rewards/rejected": -0.39652806520462036,
"step": 200
},
{
"epoch": 0.06,
"learning_rate": 2.8000000000000003e-06,
"logits/chosen": -1.6152054071426392,
"logits/rejected": -1.2735153436660767,
"logps/chosen": -557.020751953125,
"logps/rejected": -1197.393798828125,
"loss": 0.0721,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.14088398218154907,
"rewards/margins": 0.26476138830184937,
"rewards/rejected": -0.40564537048339844,
"step": 210
},
{
"epoch": 0.06,
"learning_rate": 2.9333333333333338e-06,
"logits/chosen": -1.4580497741699219,
"logits/rejected": -0.8967218399047852,
"logps/chosen": -538.6234130859375,
"logps/rejected": -1041.23291015625,
"loss": 0.1479,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.08591003715991974,
"rewards/margins": 0.20331135392189026,
"rewards/rejected": -0.2892213761806488,
"step": 220
},
{
"epoch": 0.06,
"learning_rate": 3.066666666666667e-06,
"logits/chosen": -1.6280606985092163,
"logits/rejected": -1.1948456764221191,
"logps/chosen": -464.53472900390625,
"logps/rejected": -1004.7527465820312,
"loss": 0.0994,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.04852370172739029,
"rewards/margins": 0.19602426886558533,
"rewards/rejected": -0.24454793334007263,
"step": 230
},
{
"epoch": 0.06,
"learning_rate": 3.2000000000000003e-06,
"logits/chosen": -2.0036234855651855,
"logits/rejected": -1.0634952783584595,
"logps/chosen": -654.42626953125,
"logps/rejected": -1130.1927490234375,
"loss": 0.0947,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.14522799849510193,
"rewards/margins": 0.20732179284095764,
"rewards/rejected": -0.35254979133605957,
"step": 240
},
{
"epoch": 0.07,
"learning_rate": 3.3333333333333333e-06,
"logits/chosen": -1.987884759902954,
"logits/rejected": -1.166095495223999,
"logps/chosen": -544.5479736328125,
"logps/rejected": -1122.8720703125,
"loss": 0.0906,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.12993717193603516,
"rewards/margins": 0.21453383564949036,
"rewards/rejected": -0.3444710373878479,
"step": 250
},
{
"epoch": 0.07,
"learning_rate": 3.4666666666666672e-06,
"logits/chosen": -1.532130479812622,
"logits/rejected": -0.9931814074516296,
"logps/chosen": -557.2303466796875,
"logps/rejected": -1242.306396484375,
"loss": 0.0831,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.10561565309762955,
"rewards/margins": 0.23905417323112488,
"rewards/rejected": -0.3446698486804962,
"step": 260
},
{
"epoch": 0.07,
"learning_rate": 3.6000000000000003e-06,
"logits/chosen": -1.8610786199569702,
"logits/rejected": -1.1762195825576782,
"logps/chosen": -624.46728515625,
"logps/rejected": -1214.777099609375,
"loss": 0.0777,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.10222941637039185,
"rewards/margins": 0.23841162025928497,
"rewards/rejected": -0.3406410217285156,
"step": 270
},
{
"epoch": 0.07,
"learning_rate": 3.7333333333333337e-06,
"logits/chosen": -1.6525980234146118,
"logits/rejected": -1.1402348279953003,
"logps/chosen": -689.7327880859375,
"logps/rejected": -1228.2315673828125,
"loss": 0.0904,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1671145260334015,
"rewards/margins": 0.22532522678375244,
"rewards/rejected": -0.39243969321250916,
"step": 280
},
{
"epoch": 0.08,
"learning_rate": 3.866666666666667e-06,
"logits/chosen": -1.7951831817626953,
"logits/rejected": -1.0651658773422241,
"logps/chosen": -521.4989013671875,
"logps/rejected": -1288.53466796875,
"loss": 0.0653,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.18049229681491852,
"rewards/margins": 0.28950944542884827,
"rewards/rejected": -0.4700016975402832,
"step": 290
},
{
"epoch": 0.08,
"learning_rate": 4.000000000000001e-06,
"logits/chosen": -1.5762369632720947,
"logits/rejected": -1.2923007011413574,
"logps/chosen": -643.3726806640625,
"logps/rejected": -1266.006591796875,
"loss": 0.0959,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.2561195194721222,
"rewards/margins": 0.247961163520813,
"rewards/rejected": -0.5040806531906128,
"step": 300
},
{
"epoch": 0.08,
"learning_rate": 4.133333333333333e-06,
"logits/chosen": -1.8708035945892334,
"logits/rejected": -0.9332900047302246,
"logps/chosen": -730.3931274414062,
"logps/rejected": -1277.650390625,
"loss": 0.1103,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.2379586398601532,
"rewards/margins": 0.25126343965530396,
"rewards/rejected": -0.48922213912010193,
"step": 310
},
{
"epoch": 0.09,
"learning_rate": 4.266666666666668e-06,
"logits/chosen": -1.7914050817489624,
"logits/rejected": -1.1263148784637451,
"logps/chosen": -617.783203125,
"logps/rejected": -1344.7216796875,
"loss": 0.0582,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.18687450885772705,
"rewards/margins": 0.3094720244407654,
"rewards/rejected": -0.49634653329849243,
"step": 320
},
{
"epoch": 0.09,
"learning_rate": 4.4e-06,
"logits/chosen": -1.8126357793807983,
"logits/rejected": -1.3105145692825317,
"logps/chosen": -567.7543334960938,
"logps/rejected": -1125.550048828125,
"loss": 0.117,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.17715778946876526,
"rewards/margins": 0.23828192055225372,
"rewards/rejected": -0.4154396951198578,
"step": 330
},
{
"epoch": 0.09,
"learning_rate": 4.533333333333334e-06,
"logits/chosen": -1.7919952869415283,
"logits/rejected": -1.0027551651000977,
"logps/chosen": -792.9563598632812,
"logps/rejected": -1416.2335205078125,
"loss": 0.0952,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.27687594294548035,
"rewards/margins": 0.27830666303634644,
"rewards/rejected": -0.5551826357841492,
"step": 340
},
{
"epoch": 0.09,
"learning_rate": 4.666666666666667e-06,
"logits/chosen": -1.93537175655365,
"logits/rejected": -1.2520567178726196,
"logps/chosen": -673.220947265625,
"logps/rejected": -1093.1484375,
"loss": 0.1006,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.20018038153648376,
"rewards/margins": 0.18651911616325378,
"rewards/rejected": -0.38669952750205994,
"step": 350
},
{
"epoch": 0.1,
"learning_rate": 4.800000000000001e-06,
"logits/chosen": -1.802390456199646,
"logits/rejected": -1.3325846195220947,
"logps/chosen": -649.5408935546875,
"logps/rejected": -1129.62890625,
"loss": 0.1359,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.18187561631202698,
"rewards/margins": 0.17299222946166992,
"rewards/rejected": -0.3548678159713745,
"step": 360
},
{
"epoch": 0.1,
"learning_rate": 4.933333333333334e-06,
"logits/chosen": -1.8715474605560303,
"logits/rejected": -1.3402589559555054,
"logps/chosen": -546.8035888671875,
"logps/rejected": -1010.1611328125,
"loss": 0.1034,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.05026025325059891,
"rewards/margins": 0.22379866242408752,
"rewards/rejected": -0.27405890822410583,
"step": 370
},
{
"epoch": 0.1,
"learning_rate": 4.999972922944898e-06,
"logits/chosen": -1.9571716785430908,
"logits/rejected": -1.1372935771942139,
"logps/chosen": -733.9254150390625,
"logps/rejected": -1304.736328125,
"loss": 0.0984,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15234608948230743,
"rewards/margins": 0.2319423407316208,
"rewards/rejected": -0.3842884600162506,
"step": 380
},
{
"epoch": 0.1,
"learning_rate": 4.999756310023261e-06,
"logits/chosen": -1.851304292678833,
"logits/rejected": -1.3916990756988525,
"logps/chosen": -548.0775146484375,
"logps/rejected": -1033.42578125,
"loss": 0.0848,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.11450278759002686,
"rewards/margins": 0.22328679263591766,
"rewards/rejected": -0.3377895951271057,
"step": 390
},
{
"epoch": 0.11,
"learning_rate": 4.999323102948655e-06,
"logits/chosen": -2.0076701641082764,
"logits/rejected": -1.2286336421966553,
"logps/chosen": -721.4017333984375,
"logps/rejected": -1361.4373779296875,
"loss": 0.077,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.20651724934577942,
"rewards/margins": 0.26063308119773865,
"rewards/rejected": -0.46715036034584045,
"step": 400
},
{
"epoch": 0.11,
"learning_rate": 4.998673339256785e-06,
"logits/chosen": -2.0088391304016113,
"logits/rejected": -1.2929044961929321,
"logps/chosen": -666.7005615234375,
"logps/rejected": -1329.128173828125,
"loss": 0.0905,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.21113541722297668,
"rewards/margins": 0.27807140350341797,
"rewards/rejected": -0.48920679092407227,
"step": 410
},
{
"epoch": 0.11,
"learning_rate": 4.997807075247147e-06,
"logits/chosen": -2.1807756423950195,
"logits/rejected": -1.3019901514053345,
"logps/chosen": -720.1719360351562,
"logps/rejected": -1275.56982421875,
"loss": 0.0888,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1774640679359436,
"rewards/margins": 0.26608842611312866,
"rewards/rejected": -0.44355249404907227,
"step": 420
},
{
"epoch": 0.11,
"learning_rate": 4.996724385978142e-06,
"logits/chosen": -2.1555614471435547,
"logits/rejected": -1.3447341918945312,
"logps/chosen": -620.5142822265625,
"logps/rejected": -1299.186279296875,
"loss": 0.0505,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.08281116932630539,
"rewards/margins": 0.3319624364376068,
"rewards/rejected": -0.4147736430168152,
"step": 430
},
{
"epoch": 0.12,
"learning_rate": 4.995425365260585e-06,
"logits/chosen": -2.018123149871826,
"logits/rejected": -1.2482590675354004,
"logps/chosen": -659.6309814453125,
"logps/rejected": -1221.164794921875,
"loss": 0.0884,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.15910136699676514,
"rewards/margins": 0.2621740400791168,
"rewards/rejected": -0.42127543687820435,
"step": 440
},
{
"epoch": 0.12,
"learning_rate": 4.993910125649561e-06,
"logits/chosen": -1.6636863946914673,
"logits/rejected": -0.9911340475082397,
"logps/chosen": -693.9474487304688,
"logps/rejected": -1334.8740234375,
"loss": 0.0902,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.23787899315357208,
"rewards/margins": 0.24970480799674988,
"rewards/rejected": -0.48758387565612793,
"step": 450
},
{
"epoch": 0.12,
"learning_rate": 4.992178798434684e-06,
"logits/chosen": -2.022722005844116,
"logits/rejected": -1.2033276557922363,
"logps/chosen": -636.1743774414062,
"logps/rejected": -1268.32568359375,
"loss": 0.0852,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.19134658575057983,
"rewards/margins": 0.29803937673568726,
"rewards/rejected": -0.4893859326839447,
"step": 460
},
{
"epoch": 0.13,
"learning_rate": 4.990231533628719e-06,
"logits/chosen": -1.8332548141479492,
"logits/rejected": -1.0606578588485718,
"logps/chosen": -568.590576171875,
"logps/rejected": -1158.369873046875,
"loss": 0.0795,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.09971268475055695,
"rewards/margins": 0.267979234457016,
"rewards/rejected": -0.36769190430641174,
"step": 470
},
{
"epoch": 0.13,
"learning_rate": 4.988068499954578e-06,
"logits/chosen": -2.086530923843384,
"logits/rejected": -1.4134924411773682,
"logps/chosen": -491.255126953125,
"logps/rejected": -967.7403564453125,
"loss": 0.0949,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.09277771413326263,
"rewards/margins": 0.21773691475391388,
"rewards/rejected": -0.3105146288871765,
"step": 480
},
{
"epoch": 0.13,
"learning_rate": 4.985689884830711e-06,
"logits/chosen": -1.7672725915908813,
"logits/rejected": -1.2025644779205322,
"logps/chosen": -747.850341796875,
"logps/rejected": -1293.3529052734375,
"loss": 0.1002,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.22476276755332947,
"rewards/margins": 0.23183217644691467,
"rewards/rejected": -0.45659494400024414,
"step": 490
},
{
"epoch": 0.13,
"learning_rate": 4.983095894354858e-06,
"logits/chosen": -1.6667953729629517,
"logits/rejected": -1.2720074653625488,
"logps/chosen": -679.2232666015625,
"logps/rejected": -1342.873291015625,
"loss": 0.1095,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.20331433415412903,
"rewards/margins": 0.3078593611717224,
"rewards/rejected": -0.511173665523529,
"step": 500
},
{
"epoch": 0.14,
"learning_rate": 4.980286753286196e-06,
"logits/chosen": -1.801032304763794,
"logits/rejected": -1.2592867612838745,
"logps/chosen": -528.0743408203125,
"logps/rejected": -1059.638427734375,
"loss": 0.0847,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.10437086969614029,
"rewards/margins": 0.2568288743495941,
"rewards/rejected": -0.3611997663974762,
"step": 510
},
{
"epoch": 0.14,
"learning_rate": 4.97726270502586e-06,
"logits/chosen": -1.8762857913970947,
"logits/rejected": -1.14815354347229,
"logps/chosen": -567.5826416015625,
"logps/rejected": -1179.177490234375,
"loss": 0.0764,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.09893319755792618,
"rewards/margins": 0.27982866764068604,
"rewards/rejected": -0.3787618577480316,
"step": 520
},
{
"epoch": 0.14,
"learning_rate": 4.974024011595864e-06,
"logits/chosen": -1.9012178182601929,
"logits/rejected": -1.498219609260559,
"logps/chosen": -575.654541015625,
"logps/rejected": -1156.733642578125,
"loss": 0.0812,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1486634910106659,
"rewards/margins": 0.2521246075630188,
"rewards/rejected": -0.40078815817832947,
"step": 530
},
{
"epoch": 0.14,
"learning_rate": 4.970570953616383e-06,
"logits/chosen": -1.5046392679214478,
"logits/rejected": -1.2080037593841553,
"logps/chosen": -609.6295166015625,
"logps/rejected": -1234.922607421875,
"loss": 0.0969,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.19593383371829987,
"rewards/margins": 0.2632136940956116,
"rewards/rejected": -0.45914751291275024,
"step": 540
},
{
"epoch": 0.15,
"learning_rate": 4.966903830281449e-06,
"logits/chosen": -1.8756754398345947,
"logits/rejected": -1.3740646839141846,
"logps/chosen": -685.7990112304688,
"logps/rejected": -1199.182373046875,
"loss": 0.1186,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.20507201552391052,
"rewards/margins": 0.22510738670825958,
"rewards/rejected": -0.4301794171333313,
"step": 550
},
{
"epoch": 0.15,
"learning_rate": 4.9630229593330226e-06,
"logits/chosen": -2.1258938312530518,
"logits/rejected": -1.3714039325714111,
"logps/chosen": -674.0800170898438,
"logps/rejected": -1057.264892578125,
"loss": 0.0883,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.14157623052597046,
"rewards/margins": 0.20917348563671112,
"rewards/rejected": -0.35074976086616516,
"step": 560
},
{
"epoch": 0.15,
"learning_rate": 4.958928677033465e-06,
"logits/chosen": -1.9999202489852905,
"logits/rejected": -1.2103183269500732,
"logps/chosen": -599.47607421875,
"logps/rejected": -1122.523681640625,
"loss": 0.0788,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1143900528550148,
"rewards/margins": 0.2687007486820221,
"rewards/rejected": -0.3830908238887787,
"step": 570
},
{
"epoch": 0.15,
"learning_rate": 4.954621338136399e-06,
"logits/chosen": -1.8329057693481445,
"logits/rejected": -1.1794686317443848,
"logps/chosen": -716.9681396484375,
"logps/rejected": -1392.547607421875,
"loss": 0.1106,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2248903512954712,
"rewards/margins": 0.31769150495529175,
"rewards/rejected": -0.5425819158554077,
"step": 580
},
{
"epoch": 0.16,
"learning_rate": 4.95010131585597e-06,
"logits/chosen": -1.878209114074707,
"logits/rejected": -1.0662002563476562,
"logps/chosen": -822.8294677734375,
"logps/rejected": -1354.05810546875,
"loss": 0.0824,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.24139878153800964,
"rewards/margins": 0.24748913943767548,
"rewards/rejected": -0.48888787627220154,
"step": 590
},
{
"epoch": 0.16,
"learning_rate": 4.9453690018345144e-06,
"logits/chosen": -1.870165228843689,
"logits/rejected": -1.3282101154327393,
"logps/chosen": -545.804931640625,
"logps/rejected": -1157.7216796875,
"loss": 0.0724,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.13343581557273865,
"rewards/margins": 0.23632793128490448,
"rewards/rejected": -0.3697637617588043,
"step": 600
},
{
"epoch": 0.16,
"learning_rate": 4.940424806108619e-06,
"logits/chosen": -1.7240148782730103,
"logits/rejected": -1.2849372625350952,
"logps/chosen": -562.0394897460938,
"logps/rejected": -1218.740478515625,
"loss": 0.0821,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.11796098947525024,
"rewards/margins": 0.26410388946533203,
"rewards/rejected": -0.38206490874290466,
"step": 610
},
{
"epoch": 0.17,
"learning_rate": 4.935269157073597e-06,
"logits/chosen": -1.7999210357666016,
"logits/rejected": -1.4692919254302979,
"logps/chosen": -535.56396484375,
"logps/rejected": -1111.559814453125,
"loss": 0.1083,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1319592297077179,
"rewards/margins": 0.24687273800373077,
"rewards/rejected": -0.3788319528102875,
"step": 620
},
{
"epoch": 0.17,
"learning_rate": 4.9299025014463665e-06,
"logits/chosen": -1.8443183898925781,
"logits/rejected": -1.105686902999878,
"logps/chosen": -678.8682250976562,
"logps/rejected": -1240.9027099609375,
"loss": 0.0645,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14497962594032288,
"rewards/margins": 0.2707517147064209,
"rewards/rejected": -0.415731281042099,
"step": 630
},
{
"epoch": 0.17,
"learning_rate": 4.924325304226745e-06,
"logits/chosen": -1.597224235534668,
"logits/rejected": -1.1997989416122437,
"logps/chosen": -519.6995239257812,
"logps/rejected": -1027.214599609375,
"loss": 0.1139,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.1587638556957245,
"rewards/margins": 0.2005012482404709,
"rewards/rejected": -0.3592650890350342,
"step": 640
},
{
"epoch": 0.17,
"learning_rate": 4.91853804865716e-06,
"logits/chosen": -1.8366111516952515,
"logits/rejected": -1.1595910787582397,
"logps/chosen": -564.30615234375,
"logps/rejected": -1044.134521484375,
"loss": 0.1318,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.09679403156042099,
"rewards/margins": 0.2052040547132492,
"rewards/rejected": -0.3019980788230896,
"step": 650
},
{
"epoch": 0.18,
"learning_rate": 4.912541236180779e-06,
"logits/chosen": -1.5248780250549316,
"logits/rejected": -0.9694509506225586,
"logps/chosen": -725.3447265625,
"logps/rejected": -1283.6407470703125,
"loss": 0.077,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.18108052015304565,
"rewards/margins": 0.2481127232313156,
"rewards/rejected": -0.42919325828552246,
"step": 660
},
{
"epoch": 0.18,
"learning_rate": 4.9063353863980565e-06,
"logits/chosen": -1.7408733367919922,
"logits/rejected": -1.133147954940796,
"logps/chosen": -621.7525024414062,
"logps/rejected": -1341.664306640625,
"loss": 0.0597,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.18977642059326172,
"rewards/margins": 0.3114860951900482,
"rewards/rejected": -0.5012625455856323,
"step": 670
},
{
"epoch": 0.18,
"learning_rate": 4.899921037021719e-06,
"logits/chosen": -1.7900583744049072,
"logits/rejected": -1.2211220264434814,
"logps/chosen": -654.8604736328125,
"logps/rejected": -1338.195068359375,
"loss": 0.0749,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.2140086591243744,
"rewards/margins": 0.278323233127594,
"rewards/rejected": -0.4923318922519684,
"step": 680
},
{
"epoch": 0.18,
"learning_rate": 4.893298743830168e-06,
"logits/chosen": -1.8528478145599365,
"logits/rejected": -1.1774396896362305,
"logps/chosen": -580.0806884765625,
"logps/rejected": -1167.764892578125,
"loss": 0.0921,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.12002040445804596,
"rewards/margins": 0.2513282895088196,
"rewards/rejected": -0.37134867906570435,
"step": 690
},
{
"epoch": 0.19,
"learning_rate": 4.88646908061933e-06,
"logits/chosen": -1.7654540538787842,
"logits/rejected": -1.2601468563079834,
"logps/chosen": -561.857177734375,
"logps/rejected": -1196.891357421875,
"loss": 0.0861,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.06450385600328445,
"rewards/margins": 0.24975624680519104,
"rewards/rejected": -0.3142600357532501,
"step": 700
},
{
"epoch": 0.19,
"learning_rate": 4.879432639152935e-06,
"logits/chosen": -1.7489614486694336,
"logits/rejected": -1.418021559715271,
"logps/chosen": -545.885498046875,
"logps/rejected": -1274.466552734375,
"loss": 0.0804,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.11761901527643204,
"rewards/margins": 0.29888078570365906,
"rewards/rejected": -0.4164997935295105,
"step": 710
},
{
"epoch": 0.19,
"learning_rate": 4.8721900291112415e-06,
"logits/chosen": -1.6508852243423462,
"logits/rejected": -1.3089258670806885,
"logps/chosen": -873.3531494140625,
"logps/rejected": -1446.792724609375,
"loss": 0.0813,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3387971818447113,
"rewards/margins": 0.23475828766822815,
"rewards/rejected": -0.5735554695129395,
"step": 720
},
{
"epoch": 0.19,
"learning_rate": 4.864741878038218e-06,
"logits/chosen": -1.5085474252700806,
"logits/rejected": -0.964281439781189,
"logps/chosen": -629.9556274414062,
"logps/rejected": -1125.6422119140625,
"loss": 0.1154,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.21656934916973114,
"rewards/margins": 0.25458166003227234,
"rewards/rejected": -0.47115105390548706,
"step": 730
},
{
"epoch": 0.2,
"learning_rate": 4.857088831287158e-06,
"logits/chosen": -1.6169427633285522,
"logits/rejected": -1.1799370050430298,
"logps/chosen": -644.1641235351562,
"logps/rejected": -1156.3399658203125,
"loss": 0.1243,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.20531603693962097,
"rewards/margins": 0.21790286898612976,
"rewards/rejected": -0.4232189655303955,
"step": 740
},
{
"epoch": 0.2,
"learning_rate": 4.849231551964771e-06,
"logits/chosen": -1.7949107885360718,
"logits/rejected": -0.9273164868354797,
"logps/chosen": -694.5661010742188,
"logps/rejected": -1367.0455322265625,
"loss": 0.0694,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.15553930401802063,
"rewards/margins": 0.3378245532512665,
"rewards/rejected": -0.4933638572692871,
"step": 750
},
{
"epoch": 0.2,
"learning_rate": 4.841170720873723e-06,
"logits/chosen": -1.671508550643921,
"logits/rejected": -1.0756802558898926,
"logps/chosen": -755.0061645507812,
"logps/rejected": -1223.784423828125,
"loss": 0.0906,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.28831586241722107,
"rewards/margins": 0.23300468921661377,
"rewards/rejected": -0.521320641040802,
"step": 760
},
{
"epoch": 0.21,
"learning_rate": 4.832907036453647e-06,
"logits/chosen": -1.6868644952774048,
"logits/rejected": -1.017174482345581,
"logps/chosen": -857.76904296875,
"logps/rejected": -1345.37841796875,
"loss": 0.1096,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.31581613421440125,
"rewards/margins": 0.24452456831932068,
"rewards/rejected": -0.5603407621383667,
"step": 770
},
{
"epoch": 0.21,
"learning_rate": 4.824441214720629e-06,
"logits/chosen": -1.5865137577056885,
"logits/rejected": -0.9322856068611145,
"logps/chosen": -661.6602783203125,
"logps/rejected": -1249.6767578125,
"loss": 0.0862,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2079945057630539,
"rewards/margins": 0.2843298017978668,
"rewards/rejected": -0.4923242926597595,
"step": 780
},
{
"epoch": 0.21,
"learning_rate": 4.815773989205165e-06,
"logits/chosen": -1.4541298151016235,
"logits/rejected": -1.0230770111083984,
"logps/chosen": -536.982421875,
"logps/rejected": -1293.7628173828125,
"loss": 0.0569,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1561233103275299,
"rewards/margins": 0.27445393800735474,
"rewards/rejected": -0.43057721853256226,
"step": 790
},
{
"epoch": 0.21,
"learning_rate": 4.806906110888606e-06,
"logits/chosen": -1.5984938144683838,
"logits/rejected": -1.1344892978668213,
"logps/chosen": -560.5667114257812,
"logps/rejected": -1207.4886474609375,
"loss": 0.0874,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1664171814918518,
"rewards/margins": 0.2653278708457947,
"rewards/rejected": -0.4317450523376465,
"step": 800
},
{
"epoch": 0.22,
"learning_rate": 4.7978383481380865e-06,
"logits/chosen": -1.690313696861267,
"logits/rejected": -1.2708024978637695,
"logps/chosen": -635.1510009765625,
"logps/rejected": -1258.609619140625,
"loss": 0.0729,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.19007566571235657,
"rewards/margins": 0.27011603116989136,
"rewards/rejected": -0.4601917266845703,
"step": 810
},
{
"epoch": 0.22,
"learning_rate": 4.788571486639948e-06,
"logits/chosen": -1.736071228981018,
"logits/rejected": -1.4028418064117432,
"logps/chosen": -652.5921630859375,
"logps/rejected": -1381.756591796875,
"loss": 0.0832,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.20194599032402039,
"rewards/margins": 0.2887413501739502,
"rewards/rejected": -0.49068737030029297,
"step": 820
},
{
"epoch": 0.22,
"learning_rate": 4.779106329331665e-06,
"logits/chosen": -1.542184829711914,
"logits/rejected": -0.9706169962882996,
"logps/chosen": -619.546875,
"logps/rejected": -1174.707763671875,
"loss": 0.0978,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.16596826910972595,
"rewards/margins": 0.2696087956428528,
"rewards/rejected": -0.43557706475257874,
"step": 830
},
{
"epoch": 0.22,
"learning_rate": 4.769443696332272e-06,
"logits/chosen": -1.812591552734375,
"logits/rejected": -1.1759769916534424,
"logps/chosen": -690.1990966796875,
"logps/rejected": -1262.889892578125,
"loss": 0.0952,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21801619231700897,
"rewards/margins": 0.2809485197067261,
"rewards/rejected": -0.49896472692489624,
"step": 840
},
{
"epoch": 0.23,
"learning_rate": 4.759584424871302e-06,
"logits/chosen": -1.5376830101013184,
"logits/rejected": -0.7281585931777954,
"logps/chosen": -729.5716552734375,
"logps/rejected": -1302.743408203125,
"loss": 0.0844,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.26298269629478455,
"rewards/margins": 0.28789180517196655,
"rewards/rejected": -0.5508745908737183,
"step": 850
},
{
"epoch": 0.23,
"learning_rate": 4.749529369216246e-06,
"logits/chosen": -1.5037004947662354,
"logits/rejected": -0.9839785695075989,
"logps/chosen": -823.6188354492188,
"logps/rejected": -1403.5616455078125,
"loss": 0.0976,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2846857011318207,
"rewards/margins": 0.2837271988391876,
"rewards/rejected": -0.5684128403663635,
"step": 860
},
{
"epoch": 0.23,
"learning_rate": 4.7392794005985324e-06,
"logits/chosen": -1.379028081893921,
"logits/rejected": -1.1316661834716797,
"logps/chosen": -523.5051879882812,
"logps/rejected": -1148.9351806640625,
"loss": 0.0857,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.17531812191009521,
"rewards/margins": 0.21767202019691467,
"rewards/rejected": -0.3929901719093323,
"step": 870
},
{
"epoch": 0.23,
"learning_rate": 4.7288354071380415e-06,
"logits/chosen": -1.7908474206924438,
"logits/rejected": -1.2674061059951782,
"logps/chosen": -596.2501831054688,
"logps/rejected": -1295.179931640625,
"loss": 0.0668,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1408284455537796,
"rewards/margins": 0.2789645195007324,
"rewards/rejected": -0.4197929799556732,
"step": 880
},
{
"epoch": 0.24,
"learning_rate": 4.7181982937661485e-06,
"logits/chosen": -1.5081275701522827,
"logits/rejected": -1.0480270385742188,
"logps/chosen": -599.3560791015625,
"logps/rejected": -1368.0018310546875,
"loss": 0.0594,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.17826338112354279,
"rewards/margins": 0.3195931613445282,
"rewards/rejected": -0.497856467962265,
"step": 890
},
{
"epoch": 0.24,
"learning_rate": 4.707368982147318e-06,
"logits/chosen": -1.6069328784942627,
"logits/rejected": -0.8594322204589844,
"logps/chosen": -767.88427734375,
"logps/rejected": -1367.8349609375,
"loss": 0.0836,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.22526466846466064,
"rewards/margins": 0.2738983929157257,
"rewards/rejected": -0.49916306138038635,
"step": 900
},
{
"epoch": 0.24,
"learning_rate": 4.696348410599244e-06,
"logits/chosen": -1.690263032913208,
"logits/rejected": -1.122013807296753,
"logps/chosen": -767.4993896484375,
"logps/rejected": -1443.0506591796875,
"loss": 0.08,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.24120143055915833,
"rewards/margins": 0.29539158940315247,
"rewards/rejected": -0.5365930199623108,
"step": 910
},
{
"epoch": 0.25,
"learning_rate": 4.685137534011549e-06,
"logits/chosen": -1.711660385131836,
"logits/rejected": -1.2863471508026123,
"logps/chosen": -588.4500732421875,
"logps/rejected": -1289.240966796875,
"loss": 0.0833,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.14869460463523865,
"rewards/margins": 0.28291845321655273,
"rewards/rejected": -0.4316130578517914,
"step": 920
},
{
"epoch": 0.25,
"learning_rate": 4.673737323763048e-06,
"logits/chosen": -1.8320610523223877,
"logits/rejected": -1.292765736579895,
"logps/chosen": -668.5866088867188,
"logps/rejected": -1182.257080078125,
"loss": 0.1009,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.18477411568164825,
"rewards/margins": 0.22763219475746155,
"rewards/rejected": -0.4124062955379486,
"step": 930
},
{
"epoch": 0.25,
"learning_rate": 4.662148767637578e-06,
"logits/chosen": -1.5611059665679932,
"logits/rejected": -1.1849793195724487,
"logps/chosen": -463.08355712890625,
"logps/rejected": -1084.614501953125,
"loss": 0.1011,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.08577910810709,
"rewards/margins": 0.24440805613994598,
"rewards/rejected": -0.3301871418952942,
"step": 940
},
{
"epoch": 0.25,
"learning_rate": 4.650372869738415e-06,
"logits/chosen": -1.511456847190857,
"logits/rejected": -1.0551631450653076,
"logps/chosen": -556.8165283203125,
"logps/rejected": -1228.0035400390625,
"loss": 0.0704,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1265622079372406,
"rewards/margins": 0.3069656193256378,
"rewards/rejected": -0.4335278570652008,
"step": 950
},
{
"epoch": 0.26,
"learning_rate": 4.638410650401267e-06,
"logits/chosen": -1.5442140102386475,
"logits/rejected": -0.9868549108505249,
"logps/chosen": -613.2291259765625,
"logps/rejected": -1180.44384765625,
"loss": 0.0921,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.14740554988384247,
"rewards/margins": 0.27208468317985535,
"rewards/rejected": -0.419490247964859,
"step": 960
},
{
"epoch": 0.26,
"learning_rate": 4.626263146105875e-06,
"logits/chosen": -1.7330728769302368,
"logits/rejected": -1.2073631286621094,
"logps/chosen": -700.1824340820312,
"logps/rejected": -1393.9658203125,
"loss": 0.0826,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21020905673503876,
"rewards/margins": 0.29664525389671326,
"rewards/rejected": -0.5068542957305908,
"step": 970
},
{
"epoch": 0.26,
"learning_rate": 4.613931409386196e-06,
"logits/chosen": -1.6415268182754517,
"logits/rejected": -1.1623541116714478,
"logps/chosen": -577.8115844726562,
"logps/rejected": -1281.671875,
"loss": 0.062,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.13329057395458221,
"rewards/margins": 0.31404128670692444,
"rewards/rejected": -0.44733184576034546,
"step": 980
},
{
"epoch": 0.26,
"learning_rate": 4.601416508739211e-06,
"logits/chosen": -1.5495585203170776,
"logits/rejected": -1.219543695449829,
"logps/chosen": -611.6688842773438,
"logps/rejected": -1210.188720703125,
"loss": 0.0883,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.16439898312091827,
"rewards/margins": 0.22104132175445557,
"rewards/rejected": -0.38544028997421265,
"step": 990
},
{
"epoch": 0.27,
"learning_rate": 4.588719528532342e-06,
"logits/chosen": -1.6590620279312134,
"logits/rejected": -1.1156672239303589,
"logps/chosen": -558.9346923828125,
"logps/rejected": -1250.8326416015625,
"loss": 0.0613,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.11309070885181427,
"rewards/margins": 0.27457505464553833,
"rewards/rejected": -0.3876657485961914,
"step": 1000
},
{
"epoch": 0.27,
"learning_rate": 4.575841568909494e-06,
"logits/chosen": -1.72207772731781,
"logits/rejected": -0.9709069132804871,
"logps/chosen": -574.8037719726562,
"logps/rejected": -1254.651123046875,
"loss": 0.0493,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.08526522666215897,
"rewards/margins": 0.31907790899276733,
"rewards/rejected": -0.4043431282043457,
"step": 1010
},
{
"epoch": 0.27,
"learning_rate": 4.562783745695738e-06,
"logits/chosen": -1.6100561618804932,
"logits/rejected": -1.0836453437805176,
"logps/chosen": -605.432373046875,
"logps/rejected": -1389.6484375,
"loss": 0.0562,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13646581768989563,
"rewards/margins": 0.3396046459674835,
"rewards/rejected": -0.47607049345970154,
"step": 1020
},
{
"epoch": 0.27,
"learning_rate": 4.549547190300622e-06,
"logits/chosen": -1.5626428127288818,
"logits/rejected": -1.1030738353729248,
"logps/chosen": -767.6092529296875,
"logps/rejected": -1497.570068359375,
"loss": 0.0747,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.20658251643180847,
"rewards/margins": 0.3094042241573334,
"rewards/rejected": -0.5159868001937866,
"step": 1030
},
{
"epoch": 0.28,
"learning_rate": 4.536133049620143e-06,
"logits/chosen": -1.67703378200531,
"logits/rejected": -1.1302894353866577,
"logps/chosen": -506.2029724121094,
"logps/rejected": -1050.456298828125,
"loss": 0.1074,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.10386884212493896,
"rewards/margins": 0.24720358848571777,
"rewards/rejected": -0.35107240080833435,
"step": 1040
},
{
"epoch": 0.28,
"learning_rate": 4.522542485937369e-06,
"logits/chosen": -1.510943055152893,
"logits/rejected": -1.2439063787460327,
"logps/chosen": -546.3123168945312,
"logps/rejected": -1226.2244873046875,
"loss": 0.0807,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.15128368139266968,
"rewards/margins": 0.2694355547428131,
"rewards/rejected": -0.4207192063331604,
"step": 1050
},
{
"epoch": 0.28,
"learning_rate": 4.508776676821739e-06,
"logits/chosen": -1.4208545684814453,
"logits/rejected": -0.8101575970649719,
"logps/chosen": -645.05126953125,
"logps/rejected": -1315.7305908203125,
"loss": 0.0778,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21077755093574524,
"rewards/margins": 0.2958206832408905,
"rewards/rejected": -0.5065982937812805,
"step": 1060
},
{
"epoch": 0.29,
"learning_rate": 4.494836815027022e-06,
"logits/chosen": -1.7293838262557983,
"logits/rejected": -1.1680035591125488,
"logps/chosen": -792.54443359375,
"logps/rejected": -1469.5382080078125,
"loss": 0.0713,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2889425456523895,
"rewards/margins": 0.28303924202919006,
"rewards/rejected": -0.5719817876815796,
"step": 1070
},
{
"epoch": 0.29,
"learning_rate": 4.4807241083879774e-06,
"logits/chosen": -1.5097590684890747,
"logits/rejected": -0.8481209874153137,
"logps/chosen": -801.5206298828125,
"logps/rejected": -1431.248291015625,
"loss": 0.0753,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.29523470997810364,
"rewards/margins": 0.2885613441467285,
"rewards/rejected": -0.5837960243225098,
"step": 1080
},
{
"epoch": 0.29,
"learning_rate": 4.466439779715696e-06,
"logits/chosen": -1.5163007974624634,
"logits/rejected": -0.9659198522567749,
"logps/chosen": -658.9117431640625,
"logps/rejected": -1309.037841796875,
"loss": 0.0852,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21787042915821075,
"rewards/margins": 0.2800406515598297,
"rewards/rejected": -0.49791112542152405,
"step": 1090
},
{
"epoch": 0.29,
"learning_rate": 4.451985066691649e-06,
"logits/chosen": -1.6100610494613647,
"logits/rejected": -1.1183403730392456,
"logps/chosen": -612.2716064453125,
"logps/rejected": -1260.2725830078125,
"loss": 0.0752,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21446070075035095,
"rewards/margins": 0.3018341362476349,
"rewards/rejected": -0.5162948369979858,
"step": 1100
},
{
"epoch": 0.3,
"learning_rate": 4.437361221760449e-06,
"logits/chosen": -1.4113812446594238,
"logits/rejected": -0.932741641998291,
"logps/chosen": -581.6453247070312,
"logps/rejected": -1137.978515625,
"loss": 0.1072,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.20348218083381653,
"rewards/margins": 0.25568509101867676,
"rewards/rejected": -0.4591673016548157,
"step": 1110
},
{
"epoch": 0.3,
"learning_rate": 4.422569512021332e-06,
"logits/chosen": -1.6003286838531494,
"logits/rejected": -0.9033063054084778,
"logps/chosen": -878.2449340820312,
"logps/rejected": -1371.249267578125,
"loss": 0.0921,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.29766175150871277,
"rewards/margins": 0.27674609422683716,
"rewards/rejected": -0.5744079351425171,
"step": 1120
},
{
"epoch": 0.3,
"learning_rate": 4.407611219118363e-06,
"logits/chosen": -1.6345365047454834,
"logits/rejected": -1.1004862785339355,
"logps/chosen": -627.5486450195312,
"logps/rejected": -1306.81640625,
"loss": 0.0544,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.20319166779518127,
"rewards/margins": 0.3195189833641052,
"rewards/rejected": -0.5227106809616089,
"step": 1130
},
{
"epoch": 0.3,
"learning_rate": 4.3924876391293915e-06,
"logits/chosen": -1.518014669418335,
"logits/rejected": -0.9323236346244812,
"logps/chosen": -658.4776611328125,
"logps/rejected": -1171.27587890625,
"loss": 0.1018,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.22144591808319092,
"rewards/margins": 0.2691202759742737,
"rewards/rejected": -0.4905661642551422,
"step": 1140
},
{
"epoch": 0.31,
"learning_rate": 4.377200082453748e-06,
"logits/chosen": -1.5906239748001099,
"logits/rejected": -0.9715790748596191,
"logps/chosen": -688.2633056640625,
"logps/rejected": -1256.56494140625,
"loss": 0.0857,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2001594752073288,
"rewards/margins": 0.25535720586776733,
"rewards/rejected": -0.45551663637161255,
"step": 1150
},
{
"epoch": 0.31,
"learning_rate": 4.361749873698707e-06,
"logits/chosen": -1.8183555603027344,
"logits/rejected": -1.1526719331741333,
"logps/chosen": -603.915771484375,
"logps/rejected": -1199.3839111328125,
"loss": 0.0831,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1569216549396515,
"rewards/margins": 0.24373742938041687,
"rewards/rejected": -0.40065908432006836,
"step": 1160
},
{
"epoch": 0.31,
"learning_rate": 4.346138351564711e-06,
"logits/chosen": -1.5570250749588013,
"logits/rejected": -1.029444932937622,
"logps/chosen": -720.4706420898438,
"logps/rejected": -1287.0992431640625,
"loss": 0.0828,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.22045788168907166,
"rewards/margins": 0.23490801453590393,
"rewards/rejected": -0.4553658962249756,
"step": 1170
},
{
"epoch": 0.31,
"learning_rate": 4.330366868729376e-06,
"logits/chosen": -1.524060845375061,
"logits/rejected": -1.06089186668396,
"logps/chosen": -683.9163208007812,
"logps/rejected": -1294.5540771484375,
"loss": 0.0878,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.21835967898368835,
"rewards/margins": 0.2567465901374817,
"rewards/rejected": -0.4751063287258148,
"step": 1180
},
{
"epoch": 0.32,
"learning_rate": 4.3144367917302964e-06,
"logits/chosen": -1.5951565504074097,
"logits/rejected": -1.2337514162063599,
"logps/chosen": -591.2457885742188,
"logps/rejected": -1147.8511962890625,
"loss": 0.0876,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.21049340069293976,
"rewards/margins": 0.20914044976234436,
"rewards/rejected": -0.41963380575180054,
"step": 1190
},
{
"epoch": 0.32,
"learning_rate": 4.2983495008466285e-06,
"logits/chosen": -1.3378808498382568,
"logits/rejected": -0.8554127812385559,
"logps/chosen": -704.482177734375,
"logps/rejected": -1244.0657958984375,
"loss": 0.0835,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2313729077577591,
"rewards/margins": 0.24450743198394775,
"rewards/rejected": -0.47588032484054565,
"step": 1200
},
{
"epoch": 0.32,
"learning_rate": 4.2821063899795015e-06,
"logits/chosen": -1.5431578159332275,
"logits/rejected": -1.1451054811477661,
"logps/chosen": -551.98486328125,
"logps/rejected": -1086.984130859375,
"loss": 0.1082,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.17247965931892395,
"rewards/margins": 0.2379414588212967,
"rewards/rejected": -0.41042113304138184,
"step": 1210
},
{
"epoch": 0.33,
"learning_rate": 4.265708866531238e-06,
"logits/chosen": -1.3170068264007568,
"logits/rejected": -1.3953243494033813,
"logps/chosen": -555.5161743164062,
"logps/rejected": -1217.95556640625,
"loss": 0.0874,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1994590014219284,
"rewards/margins": 0.24560394883155823,
"rewards/rejected": -0.4450629651546478,
"step": 1220
},
{
"epoch": 0.33,
"learning_rate": 4.249158351283414e-06,
"logits/chosen": -1.5873005390167236,
"logits/rejected": -0.9497630000114441,
"logps/chosen": -762.7379150390625,
"logps/rejected": -1492.514404296875,
"loss": 0.0591,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.26952821016311646,
"rewards/margins": 0.34387946128845215,
"rewards/rejected": -0.6134077310562134,
"step": 1230
},
{
"epoch": 0.33,
"learning_rate": 4.232456278273743e-06,
"logits/chosen": -1.6160411834716797,
"logits/rejected": -0.9002545475959778,
"logps/chosen": -827.9026489257812,
"logps/rejected": -1476.774169921875,
"loss": 0.0731,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.2839387059211731,
"rewards/margins": 0.3188818097114563,
"rewards/rejected": -0.6028205156326294,
"step": 1240
},
{
"epoch": 0.33,
"learning_rate": 4.215604094671835e-06,
"logits/chosen": -1.7747631072998047,
"logits/rejected": -1.4146007299423218,
"logps/chosen": -544.9703369140625,
"logps/rejected": -1178.328369140625,
"loss": 0.0946,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.151317298412323,
"rewards/margins": 0.26069843769073486,
"rewards/rejected": -0.41201576590538025,
"step": 1250
},
{
"epoch": 0.34,
"learning_rate": 4.198603260653792e-06,
"logits/chosen": -1.6157176494598389,
"logits/rejected": -0.8367627859115601,
"logps/chosen": -626.7408447265625,
"logps/rejected": -1291.121826171875,
"loss": 0.0867,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1371547281742096,
"rewards/margins": 0.3223341405391693,
"rewards/rejected": -0.45948880910873413,
"step": 1260
},
{
"epoch": 0.34,
"learning_rate": 4.181455249275701e-06,
"logits/chosen": -1.7481143474578857,
"logits/rejected": -1.12859308719635,
"logps/chosen": -634.04638671875,
"logps/rejected": -1124.959716796875,
"loss": 0.0954,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.16867594420909882,
"rewards/margins": 0.2416352778673172,
"rewards/rejected": -0.410311222076416,
"step": 1270
},
{
"epoch": 0.34,
"learning_rate": 4.1641615463459926e-06,
"logits/chosen": -1.5910767316818237,
"logits/rejected": -1.0931599140167236,
"logps/chosen": -631.27392578125,
"logps/rejected": -1231.7950439453125,
"loss": 0.0834,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16265997290611267,
"rewards/margins": 0.24569034576416016,
"rewards/rejected": -0.4083503186702728,
"step": 1280
},
{
"epoch": 0.34,
"learning_rate": 4.146723650296701e-06,
"logits/chosen": -1.6471401453018188,
"logits/rejected": -1.2424445152282715,
"logps/chosen": -468.88739013671875,
"logps/rejected": -1020.91259765625,
"loss": 0.102,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.10470066964626312,
"rewards/margins": 0.23985891044139862,
"rewards/rejected": -0.34455958008766174,
"step": 1290
},
{
"epoch": 0.35,
"learning_rate": 4.129143072053639e-06,
"logits/chosen": -1.608716607093811,
"logits/rejected": -0.9606212377548218,
"logps/chosen": -669.7889404296875,
"logps/rejected": -1358.1666259765625,
"loss": 0.0559,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.17117540538311005,
"rewards/margins": 0.31088918447494507,
"rewards/rejected": -0.4820645749568939,
"step": 1300
},
{
"epoch": 0.35,
"learning_rate": 4.111421334905468e-06,
"logits/chosen": -1.5793676376342773,
"logits/rejected": -1.0094425678253174,
"logps/chosen": -710.6974487304688,
"logps/rejected": -1337.674560546875,
"loss": 0.0837,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2793545424938202,
"rewards/margins": 0.28449904918670654,
"rewards/rejected": -0.5638536214828491,
"step": 1310
},
{
"epoch": 0.35,
"learning_rate": 4.093559974371725e-06,
"logits/chosen": -1.3306772708892822,
"logits/rejected": -1.1824491024017334,
"logps/chosen": -800.3571166992188,
"logps/rejected": -1588.775146484375,
"loss": 0.088,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.32887041568756104,
"rewards/margins": 0.32484978437423706,
"rewards/rejected": -0.6537202596664429,
"step": 1320
},
{
"epoch": 0.35,
"learning_rate": 4.075560538069767e-06,
"logits/chosen": -1.5261657238006592,
"logits/rejected": -1.0766921043395996,
"logps/chosen": -690.2559814453125,
"logps/rejected": -1296.5557861328125,
"loss": 0.0804,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21848633885383606,
"rewards/margins": 0.2872334122657776,
"rewards/rejected": -0.505719780921936,
"step": 1330
},
{
"epoch": 0.36,
"learning_rate": 4.05742458558068e-06,
"logits/chosen": -1.416123628616333,
"logits/rejected": -1.1054461002349854,
"logps/chosen": -764.1392822265625,
"logps/rejected": -1332.025634765625,
"loss": 0.1048,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.22151347994804382,
"rewards/margins": 0.24781036376953125,
"rewards/rejected": -0.4693238139152527,
"step": 1340
},
{
"epoch": 0.36,
"learning_rate": 4.039153688314146e-06,
"logits/chosen": -1.4527040719985962,
"logits/rejected": -0.6857194900512695,
"logps/chosen": -812.2103271484375,
"logps/rejected": -1516.066650390625,
"loss": 0.0609,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3405809998512268,
"rewards/margins": 0.31844016909599304,
"rewards/rejected": -0.6590211987495422,
"step": 1350
},
{
"epoch": 0.36,
"learning_rate": 4.020749429372286e-06,
"logits/chosen": -1.446703314781189,
"logits/rejected": -0.9012134671211243,
"logps/chosen": -742.4305419921875,
"logps/rejected": -1499.476806640625,
"loss": 0.069,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.3392742872238159,
"rewards/margins": 0.3239519000053406,
"rewards/rejected": -0.6632262468338013,
"step": 1360
},
{
"epoch": 0.37,
"learning_rate": 4.002213403412492e-06,
"logits/chosen": -1.3960011005401611,
"logits/rejected": -0.9817187190055847,
"logps/chosen": -778.759033203125,
"logps/rejected": -1387.777099609375,
"loss": 0.0727,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.3149294853210449,
"rewards/margins": 0.27625852823257446,
"rewards/rejected": -0.5911880731582642,
"step": 1370
},
{
"epoch": 0.37,
"learning_rate": 3.983547216509254e-06,
"logits/chosen": -1.5987809896469116,
"logits/rejected": -0.7307044863700867,
"logps/chosen": -858.3521728515625,
"logps/rejected": -1550.783935546875,
"loss": 0.0618,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -0.2980232834815979,
"rewards/margins": 0.3726673722267151,
"rewards/rejected": -0.670690655708313,
"step": 1380
},
{
"epoch": 0.37,
"learning_rate": 3.964752486015001e-06,
"logits/chosen": -1.7145130634307861,
"logits/rejected": -1.0826901197433472,
"logps/chosen": -551.7413330078125,
"logps/rejected": -973.1448364257812,
"loss": 0.1367,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.17761428654193878,
"rewards/margins": 0.18822605907917023,
"rewards/rejected": -0.365840345621109,
"step": 1390
},
{
"epoch": 0.37,
"learning_rate": 3.945830840419966e-06,
"logits/chosen": -1.7292487621307373,
"logits/rejected": -1.03867506980896,
"logps/chosen": -610.347900390625,
"logps/rejected": -1196.255126953125,
"loss": 0.0801,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.19123509526252747,
"rewards/margins": 0.27820879220962524,
"rewards/rejected": -0.4694438874721527,
"step": 1400
},
{
"epoch": 0.38,
"learning_rate": 3.92678391921108e-06,
"logits/chosen": -1.7660369873046875,
"logits/rejected": -0.8002158403396606,
"logps/chosen": -790.7001342773438,
"logps/rejected": -1427.8797607421875,
"loss": 0.0402,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.24115972220897675,
"rewards/margins": 0.34107840061187744,
"rewards/rejected": -0.5822380781173706,
"step": 1410
},
{
"epoch": 0.38,
"learning_rate": 3.907613372729916e-06,
"logits/chosen": -1.5305942296981812,
"logits/rejected": -0.9681123495101929,
"logps/chosen": -683.839599609375,
"logps/rejected": -1229.334228515625,
"loss": 0.1139,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.26777878403663635,
"rewards/margins": 0.2556554675102234,
"rewards/rejected": -0.5234342813491821,
"step": 1420
},
{
"epoch": 0.38,
"learning_rate": 3.888320862029699e-06,
"logits/chosen": -1.5409247875213623,
"logits/rejected": -1.1824085712432861,
"logps/chosen": -565.3055419921875,
"logps/rejected": -1225.6197509765625,
"loss": 0.0719,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.21753540635108948,
"rewards/margins": 0.262325644493103,
"rewards/rejected": -0.4798610210418701,
"step": 1430
},
{
"epoch": 0.38,
"learning_rate": 3.868908058731376e-06,
"logits/chosen": -1.6002075672149658,
"logits/rejected": -0.9055458903312683,
"logps/chosen": -626.1796264648438,
"logps/rejected": -1374.696044921875,
"loss": 0.045,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.20963218808174133,
"rewards/margins": 0.36331892013549805,
"rewards/rejected": -0.5729510188102722,
"step": 1440
},
{
"epoch": 0.39,
"learning_rate": 3.849376644878783e-06,
"logits/chosen": -1.5136663913726807,
"logits/rejected": -1.2368497848510742,
"logps/chosen": -744.4407958984375,
"logps/rejected": -1368.80126953125,
"loss": 0.0722,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.24185235798358917,
"rewards/margins": 0.2773440182209015,
"rewards/rejected": -0.5191963911056519,
"step": 1450
},
{
"epoch": 0.39,
"learning_rate": 3.829728312792895e-06,
"logits/chosen": -1.6058063507080078,
"logits/rejected": -1.3173209428787231,
"logps/chosen": -614.1032104492188,
"logps/rejected": -1252.304443359375,
"loss": 0.0718,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1905815303325653,
"rewards/margins": 0.2669087052345276,
"rewards/rejected": -0.45749014616012573,
"step": 1460
},
{
"epoch": 0.39,
"learning_rate": 3.8099647649251984e-06,
"logits/chosen": -1.5442336797714233,
"logits/rejected": -1.108355164527893,
"logps/chosen": -711.9778442382812,
"logps/rejected": -1236.2525634765625,
"loss": 0.1105,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.17289568483829498,
"rewards/margins": 0.2372523844242096,
"rewards/rejected": -0.41014808416366577,
"step": 1470
},
{
"epoch": 0.39,
"learning_rate": 3.790087713710179e-06,
"logits/chosen": -1.7366338968276978,
"logits/rejected": -1.3570505380630493,
"logps/chosen": -617.22607421875,
"logps/rejected": -1209.1883544921875,
"loss": 0.072,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.14606155455112457,
"rewards/margins": 0.2611353397369385,
"rewards/rejected": -0.40719684958457947,
"step": 1480
},
{
"epoch": 0.4,
"learning_rate": 3.770098881416945e-06,
"logits/chosen": -1.6808090209960938,
"logits/rejected": -0.7971758842468262,
"logps/chosen": -862.0647583007812,
"logps/rejected": -1464.6036376953125,
"loss": 0.0635,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -0.23483994603157043,
"rewards/margins": 0.32541361451148987,
"rewards/rejected": -0.5602535009384155,
"step": 1490
},
{
"epoch": 0.4,
"learning_rate": 3.7500000000000005e-06,
"logits/chosen": -1.6327816247940063,
"logits/rejected": -0.9372482299804688,
"logps/chosen": -734.8026123046875,
"logps/rejected": -1427.211181640625,
"loss": 0.0614,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2201705276966095,
"rewards/margins": 0.31630367040634155,
"rewards/rejected": -0.5364742279052734,
"step": 1500
},
{
"epoch": 0.4,
"learning_rate": 3.7297928109491765e-06,
"logits/chosen": -1.4493783712387085,
"logits/rejected": -1.1883541345596313,
"logps/chosen": -693.4669189453125,
"logps/rejected": -1344.7279052734375,
"loss": 0.0707,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2357136309146881,
"rewards/margins": 0.28132572770118713,
"rewards/rejected": -0.5170393586158752,
"step": 1510
},
{
"epoch": 0.41,
"learning_rate": 3.7094790651387414e-06,
"logits/chosen": -1.6289390325546265,
"logits/rejected": -1.1786553859710693,
"logps/chosen": -669.4554443359375,
"logps/rejected": -1307.296875,
"loss": 0.0715,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.2002047747373581,
"rewards/margins": 0.3151922821998596,
"rewards/rejected": -0.5153970122337341,
"step": 1520
},
{
"epoch": 0.41,
"learning_rate": 3.689060522675689e-06,
"logits/chosen": -1.5846166610717773,
"logits/rejected": -0.9855213165283203,
"logps/chosen": -595.7760620117188,
"logps/rejected": -1270.208740234375,
"loss": 0.0617,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1446508765220642,
"rewards/margins": 0.3013971149921417,
"rewards/rejected": -0.4460480213165283,
"step": 1530
},
{
"epoch": 0.41,
"learning_rate": 3.668538952747236e-06,
"logits/chosen": -1.6458574533462524,
"logits/rejected": -1.1102992296218872,
"logps/chosen": -588.1439208984375,
"logps/rejected": -1164.717529296875,
"loss": 0.0999,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1480761170387268,
"rewards/margins": 0.22489504516124725,
"rewards/rejected": -0.37297114729881287,
"step": 1540
},
{
"epoch": 0.41,
"learning_rate": 3.6479161334675294e-06,
"logits/chosen": -1.7304718494415283,
"logits/rejected": -1.0173327922821045,
"logps/chosen": -598.8756713867188,
"logps/rejected": -1213.0777587890625,
"loss": 0.064,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.13348576426506042,
"rewards/margins": 0.30749550461769104,
"rewards/rejected": -0.44098129868507385,
"step": 1550
},
{
"epoch": 0.42,
"learning_rate": 3.627193851723577e-06,
"logits/chosen": -1.7607488632202148,
"logits/rejected": -1.0212706327438354,
"logps/chosen": -677.742431640625,
"logps/rejected": -1323.628173828125,
"loss": 0.0535,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1735551357269287,
"rewards/margins": 0.3078446090221405,
"rewards/rejected": -0.4813997745513916,
"step": 1560
},
{
"epoch": 0.42,
"learning_rate": 3.6063739030204226e-06,
"logits/chosen": -1.6025089025497437,
"logits/rejected": -0.9767768979072571,
"logps/chosen": -597.2946166992188,
"logps/rejected": -1132.9073486328125,
"loss": 0.1038,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.16409102082252502,
"rewards/margins": 0.2722373902797699,
"rewards/rejected": -0.4363284111022949,
"step": 1570
},
{
"epoch": 0.42,
"learning_rate": 3.5854580913255706e-06,
"logits/chosen": -1.616591215133667,
"logits/rejected": -1.1780548095703125,
"logps/chosen": -675.4401245117188,
"logps/rejected": -1367.3082275390625,
"loss": 0.0635,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.22119805216789246,
"rewards/margins": 0.2958839535713196,
"rewards/rejected": -0.5170820951461792,
"step": 1580
},
{
"epoch": 0.42,
"learning_rate": 3.564448228912682e-06,
"logits/chosen": -1.5432292222976685,
"logits/rejected": -0.871782660484314,
"logps/chosen": -773.683837890625,
"logps/rejected": -1428.926025390625,
"loss": 0.0403,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.21216030418872833,
"rewards/margins": 0.35337987542152405,
"rewards/rejected": -0.5655401349067688,
"step": 1590
},
{
"epoch": 0.43,
"learning_rate": 3.543346136204545e-06,
"logits/chosen": -1.3973724842071533,
"logits/rejected": -1.1439921855926514,
"logps/chosen": -646.6536254882812,
"logps/rejected": -1378.410400390625,
"loss": 0.0582,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.18542486429214478,
"rewards/margins": 0.3139794170856476,
"rewards/rejected": -0.49940428137779236,
"step": 1600
},
{
"epoch": 0.43,
"learning_rate": 3.522153641615345e-06,
"logits/chosen": -1.7498506307601929,
"logits/rejected": -0.853812038898468,
"logps/chosen": -691.226806640625,
"logps/rejected": -1190.421630859375,
"loss": 0.0798,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1312675178050995,
"rewards/margins": 0.2993599772453308,
"rewards/rejected": -0.4306275248527527,
"step": 1610
},
{
"epoch": 0.43,
"learning_rate": 3.5008725813922383e-06,
"logits/chosen": -1.6955890655517578,
"logits/rejected": -1.0088529586791992,
"logps/chosen": -625.9053955078125,
"logps/rejected": -1268.4305419921875,
"loss": 0.0572,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.15290668606758118,
"rewards/margins": 0.2966083288192749,
"rewards/rejected": -0.44951504468917847,
"step": 1620
},
{
"epoch": 0.43,
"learning_rate": 3.4795047994562463e-06,
"logits/chosen": -1.787287950515747,
"logits/rejected": -0.991382896900177,
"logps/chosen": -736.6607666015625,
"logps/rejected": -1425.2071533203125,
"loss": 0.041,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.2185276746749878,
"rewards/margins": 0.3424827456474304,
"rewards/rejected": -0.561010479927063,
"step": 1630
},
{
"epoch": 0.44,
"learning_rate": 3.458052147242494e-06,
"logits/chosen": -1.5612847805023193,
"logits/rejected": -0.9777131080627441,
"logps/chosen": -677.7218627929688,
"logps/rejected": -1298.963623046875,
"loss": 0.0625,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.19684460759162903,
"rewards/margins": 0.29719001054763794,
"rewards/rejected": -0.49403461813926697,
"step": 1640
},
{
"epoch": 0.44,
"learning_rate": 3.436516483539781e-06,
"logits/chosen": -1.522048830986023,
"logits/rejected": -0.9393981099128723,
"logps/chosen": -657.0848388671875,
"logps/rejected": -1337.720458984375,
"loss": 0.0442,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.20744872093200684,
"rewards/margins": 0.3454239070415497,
"rewards/rejected": -0.5528727173805237,
"step": 1650
},
{
"epoch": 0.44,
"learning_rate": 3.4148996743295305e-06,
"logits/chosen": -1.6736905574798584,
"logits/rejected": -1.3306677341461182,
"logps/chosen": -594.0120849609375,
"logps/rejected": -1347.886474609375,
"loss": 0.0643,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16157297790050507,
"rewards/margins": 0.33852246403694153,
"rewards/rejected": -0.5000954270362854,
"step": 1660
},
{
"epoch": 0.45,
"learning_rate": 3.3932035926241103e-06,
"logits/chosen": -1.5001100301742554,
"logits/rejected": -1.099002718925476,
"logps/chosen": -574.8411254882812,
"logps/rejected": -1262.277099609375,
"loss": 0.071,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.17917487025260925,
"rewards/margins": 0.3023154139518738,
"rewards/rejected": -0.4814903140068054,
"step": 1670
},
{
"epoch": 0.45,
"learning_rate": 3.3714301183045382e-06,
"logits/chosen": -1.5629947185516357,
"logits/rejected": -1.0339363813400269,
"logps/chosen": -695.541748046875,
"logps/rejected": -1296.75830078125,
"loss": 0.0871,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.19980263710021973,
"rewards/margins": 0.25737708806991577,
"rewards/rejected": -0.4571797847747803,
"step": 1680
},
{
"epoch": 0.45,
"learning_rate": 3.349581137957604e-06,
"logits/chosen": -1.7300357818603516,
"logits/rejected": -0.9519279599189758,
"logps/chosen": -630.9297485351562,
"logps/rejected": -1248.482421875,
"loss": 0.0576,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1405705362558365,
"rewards/margins": 0.3395325243473053,
"rewards/rejected": -0.4801030158996582,
"step": 1690
},
{
"epoch": 0.45,
"learning_rate": 3.3276585447123957e-06,
"logits/chosen": -1.7029545307159424,
"logits/rejected": -0.8141438364982605,
"logps/chosen": -717.7041625976562,
"logps/rejected": -1194.781005859375,
"loss": 0.0961,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1431819498538971,
"rewards/margins": 0.2582007944583893,
"rewards/rejected": -0.40138277411460876,
"step": 1700
},
{
"epoch": 0.46,
"learning_rate": 3.3056642380762783e-06,
"logits/chosen": -1.7564830780029297,
"logits/rejected": -1.1706587076187134,
"logps/chosen": -668.6366577148438,
"logps/rejected": -1349.9146728515625,
"loss": 0.0828,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.15937338769435883,
"rewards/margins": 0.3159894347190857,
"rewards/rejected": -0.4753628373146057,
"step": 1710
},
{
"epoch": 0.46,
"learning_rate": 3.2836001237702993e-06,
"logits/chosen": -1.8231847286224365,
"logits/rejected": -1.20169198513031,
"logps/chosen": -600.72265625,
"logps/rejected": -1295.0966796875,
"loss": 0.0506,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.13170823454856873,
"rewards/margins": 0.3157060444355011,
"rewards/rejected": -0.4474143087863922,
"step": 1720
},
{
"epoch": 0.46,
"learning_rate": 3.2614681135640696e-06,
"logits/chosen": -1.424872636795044,
"logits/rejected": -1.0434539318084717,
"logps/chosen": -765.4076538085938,
"logps/rejected": -1424.029052734375,
"loss": 0.088,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.22880463302135468,
"rewards/margins": 0.25481894612312317,
"rewards/rejected": -0.48362359404563904,
"step": 1730
},
{
"epoch": 0.46,
"learning_rate": 3.2392701251101172e-06,
"logits/chosen": -1.6727033853530884,
"logits/rejected": -0.9850581884384155,
"logps/chosen": -603.7737426757812,
"logps/rejected": -1283.9576416015625,
"loss": 0.0851,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1411423683166504,
"rewards/margins": 0.2962447702884674,
"rewards/rejected": -0.4373871386051178,
"step": 1740
},
{
"epoch": 0.47,
"learning_rate": 3.217008081777726e-06,
"logits/chosen": -1.4920157194137573,
"logits/rejected": -1.0782114267349243,
"logps/chosen": -520.2071533203125,
"logps/rejected": -1167.503662109375,
"loss": 0.0687,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.12038830667734146,
"rewards/margins": 0.26983264088630676,
"rewards/rejected": -0.3902209401130676,
"step": 1750
},
{
"epoch": 0.47,
"learning_rate": 3.1946839124862873e-06,
"logits/chosen": -1.59346342086792,
"logits/rejected": -0.8561047315597534,
"logps/chosen": -620.8773803710938,
"logps/rejected": -1345.0489501953125,
"loss": 0.0611,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1517111212015152,
"rewards/margins": 0.29772210121154785,
"rewards/rejected": -0.44943323731422424,
"step": 1760
},
{
"epoch": 0.47,
"learning_rate": 3.1722995515381644e-06,
"logits/chosen": -1.590381383895874,
"logits/rejected": -1.1091078519821167,
"logps/chosen": -734.8489379882812,
"logps/rejected": -1222.6331787109375,
"loss": 0.0966,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.19709806144237518,
"rewards/margins": 0.24125704169273376,
"rewards/rejected": -0.43835514783859253,
"step": 1770
},
{
"epoch": 0.47,
"learning_rate": 3.149856938451094e-06,
"logits/chosen": -1.6151930093765259,
"logits/rejected": -0.9260737299919128,
"logps/chosen": -644.0718383789062,
"logps/rejected": -1129.24267578125,
"loss": 0.0756,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.14687219262123108,
"rewards/margins": 0.2772686779499054,
"rewards/rejected": -0.4241408407688141,
"step": 1780
},
{
"epoch": 0.48,
"learning_rate": 3.127358017790132e-06,
"logits/chosen": -1.8206260204315186,
"logits/rejected": -1.036929965019226,
"logps/chosen": -636.3726806640625,
"logps/rejected": -1280.237060546875,
"loss": 0.0717,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1703665554523468,
"rewards/margins": 0.33507412672042847,
"rewards/rejected": -0.5054406523704529,
"step": 1790
},
{
"epoch": 0.48,
"learning_rate": 3.1048047389991693e-06,
"logits/chosen": -1.7902650833129883,
"logits/rejected": -1.1972464323043823,
"logps/chosen": -671.567138671875,
"logps/rejected": -1139.710205078125,
"loss": 0.096,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.17933420836925507,
"rewards/margins": 0.23233267664909363,
"rewards/rejected": -0.4116668701171875,
"step": 1800
},
{
"epoch": 0.48,
"learning_rate": 3.082199056232015e-06,
"logits/chosen": -1.74044668674469,
"logits/rejected": -1.0097100734710693,
"logps/chosen": -597.4201049804688,
"logps/rejected": -1311.656494140625,
"loss": 0.0724,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1604226678609848,
"rewards/margins": 0.33031123876571655,
"rewards/rejected": -0.4907340109348297,
"step": 1810
},
{
"epoch": 0.49,
"learning_rate": 3.059542928183079e-06,
"logits/chosen": -1.7327849864959717,
"logits/rejected": -1.105128526687622,
"logps/chosen": -666.2789306640625,
"logps/rejected": -1354.0238037109375,
"loss": 0.0725,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1905040591955185,
"rewards/margins": 0.3310181498527527,
"rewards/rejected": -0.5215222239494324,
"step": 1820
},
{
"epoch": 0.49,
"learning_rate": 3.0368383179176584e-06,
"logits/chosen": -1.4599992036819458,
"logits/rejected": -0.9781301617622375,
"logps/chosen": -612.1408081054688,
"logps/rejected": -1311.3360595703125,
"loss": 0.0769,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.21572145819664001,
"rewards/margins": 0.2840631902217865,
"rewards/rejected": -0.4997846484184265,
"step": 1830
},
{
"epoch": 0.49,
"learning_rate": 3.0140871927018466e-06,
"logits/chosen": -1.6299057006835938,
"logits/rejected": -1.1118695735931396,
"logps/chosen": -627.4758911132812,
"logps/rejected": -1280.96630859375,
"loss": 0.0673,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1827511489391327,
"rewards/margins": 0.30515769124031067,
"rewards/rejected": -0.4879087805747986,
"step": 1840
},
{
"epoch": 0.49,
"learning_rate": 2.9912915238320755e-06,
"logits/chosen": -1.7462234497070312,
"logits/rejected": -1.0023224353790283,
"logps/chosen": -662.7495727539062,
"logps/rejected": -1253.028076171875,
"loss": 0.0537,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1878022700548172,
"rewards/margins": 0.30388498306274414,
"rewards/rejected": -0.4916872978210449,
"step": 1850
},
{
"epoch": 0.5,
"learning_rate": 2.9684532864643123e-06,
"logits/chosen": -1.6837208271026611,
"logits/rejected": -1.037295937538147,
"logps/chosen": -592.4058837890625,
"logps/rejected": -1251.274658203125,
"loss": 0.065,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.164472758769989,
"rewards/margins": 0.3051846921443939,
"rewards/rejected": -0.4696574807167053,
"step": 1860
},
{
"epoch": 0.5,
"learning_rate": 2.945574459442917e-06,
"logits/chosen": -1.7075077295303345,
"logits/rejected": -1.0973665714263916,
"logps/chosen": -623.5327758789062,
"logps/rejected": -1248.406494140625,
"loss": 0.0519,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.14970310032367706,
"rewards/margins": 0.3357471525669098,
"rewards/rejected": -0.48545026779174805,
"step": 1870
},
{
"epoch": 0.5,
"learning_rate": 2.922657025129185e-06,
"logits/chosen": -1.659570336341858,
"logits/rejected": -1.0617914199829102,
"logps/chosen": -783.8493041992188,
"logps/rejected": -1338.663818359375,
"loss": 0.1171,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.24820463359355927,
"rewards/margins": 0.2960302531719208,
"rewards/rejected": -0.5442348718643188,
"step": 1880
},
{
"epoch": 0.5,
"learning_rate": 2.8997029692295875e-06,
"logits/chosen": -1.685340166091919,
"logits/rejected": -1.1272087097167969,
"logps/chosen": -648.3568115234375,
"logps/rejected": -1290.0946044921875,
"loss": 0.0795,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2036975622177124,
"rewards/margins": 0.29076558351516724,
"rewards/rejected": -0.49446314573287964,
"step": 1890
},
{
"epoch": 0.51,
"learning_rate": 2.876714280623708e-06,
"logits/chosen": -1.5509741306304932,
"logits/rejected": -1.353191614151001,
"logps/chosen": -530.915771484375,
"logps/rejected": -1135.705322265625,
"loss": 0.1021,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.18661265075206757,
"rewards/margins": 0.2390277087688446,
"rewards/rejected": -0.425640344619751,
"step": 1900
},
{
"epoch": 0.51,
"learning_rate": 2.8536929511919227e-06,
"logits/chosen": -1.5140665769577026,
"logits/rejected": -0.9606078267097473,
"logps/chosen": -781.3890380859375,
"logps/rejected": -1375.6937255859375,
"loss": 0.0884,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.2563932538032532,
"rewards/margins": 0.2651718854904175,
"rewards/rejected": -0.5215650796890259,
"step": 1910
},
{
"epoch": 0.51,
"learning_rate": 2.8306409756428067e-06,
"logits/chosen": -1.6241651773452759,
"logits/rejected": -0.8508566617965698,
"logps/chosen": -628.0801391601562,
"logps/rejected": -1286.9000244140625,
"loss": 0.0711,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.14530445635318756,
"rewards/margins": 0.3342815041542053,
"rewards/rejected": -0.4795859754085541,
"step": 1920
},
{
"epoch": 0.51,
"learning_rate": 2.807560351340302e-06,
"logits/chosen": -1.8166322708129883,
"logits/rejected": -1.1172343492507935,
"logps/chosen": -561.2113647460938,
"logps/rejected": -1348.037841796875,
"loss": 0.0472,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.12384496629238129,
"rewards/margins": 0.3702481985092163,
"rewards/rejected": -0.4940931797027588,
"step": 1930
},
{
"epoch": 0.52,
"learning_rate": 2.7844530781306544e-06,
"logits/chosen": -1.6891119480133057,
"logits/rejected": -1.0480057001113892,
"logps/chosen": -686.9342041015625,
"logps/rejected": -1289.268310546875,
"loss": 0.0736,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.17461568117141724,
"rewards/margins": 0.27256545424461365,
"rewards/rejected": -0.44718116521835327,
"step": 1940
},
{
"epoch": 0.52,
"learning_rate": 2.761321158169134e-06,
"logits/chosen": -1.7340322732925415,
"logits/rejected": -1.1662895679473877,
"logps/chosen": -579.2034912109375,
"logps/rejected": -1151.33154296875,
"loss": 0.1019,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1479378640651703,
"rewards/margins": 0.24325743317604065,
"rewards/rejected": -0.3911953270435333,
"step": 1950
},
{
"epoch": 0.52,
"learning_rate": 2.738166595746554e-06,
"logits/chosen": -1.6452268362045288,
"logits/rejected": -1.2167978286743164,
"logps/chosen": -568.6005859375,
"logps/rejected": -1262.2685546875,
"loss": 0.0716,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.15343192219734192,
"rewards/margins": 0.26729413866996765,
"rewards/rejected": -0.42072606086730957,
"step": 1960
},
{
"epoch": 0.53,
"learning_rate": 2.7149913971156105e-06,
"logits/chosen": -1.818169355392456,
"logits/rejected": -1.1683781147003174,
"logps/chosen": -652.8570556640625,
"logps/rejected": -1269.952392578125,
"loss": 0.0727,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1778111755847931,
"rewards/margins": 0.29952138662338257,
"rewards/rejected": -0.47733253240585327,
"step": 1970
},
{
"epoch": 0.53,
"learning_rate": 2.6917975703170466e-06,
"logits/chosen": -1.3900476694107056,
"logits/rejected": -0.9688823819160461,
"logps/chosen": -659.0379028320312,
"logps/rejected": -1403.9931640625,
"loss": 0.0587,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.21347947418689728,
"rewards/margins": 0.3426669239997864,
"rewards/rejected": -0.5561463832855225,
"step": 1980
},
{
"epoch": 0.53,
"learning_rate": 2.668587125005663e-06,
"logits/chosen": -1.7847397327423096,
"logits/rejected": -1.250427007675171,
"logps/chosen": -552.1531982421875,
"logps/rejected": -1304.26123046875,
"loss": 0.0691,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15698403120040894,
"rewards/margins": 0.30882400274276733,
"rewards/rejected": -0.46580806374549866,
"step": 1990
},
{
"epoch": 0.53,
"learning_rate": 2.6453620722761897e-06,
"logits/chosen": -1.4962866306304932,
"logits/rejected": -1.1884623765945435,
"logps/chosen": -508.5406188964844,
"logps/rejected": -1144.9775390625,
"loss": 0.0703,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.14970794320106506,
"rewards/margins": 0.26972299814224243,
"rewards/rejected": -0.41943103075027466,
"step": 2000
},
{
"epoch": 0.54,
"learning_rate": 2.6221244244890336e-06,
"logits/chosen": -1.4763376712799072,
"logits/rejected": -1.208345890045166,
"logps/chosen": -659.405029296875,
"logps/rejected": -1311.456298828125,
"loss": 0.0641,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.21015486121177673,
"rewards/margins": 0.261810302734375,
"rewards/rejected": -0.47196516394615173,
"step": 2010
},
{
"epoch": 0.54,
"learning_rate": 2.5988761950959133e-06,
"logits/chosen": -1.6231094598770142,
"logits/rejected": -1.3810958862304688,
"logps/chosen": -578.0490112304688,
"logps/rejected": -1218.637451171875,
"loss": 0.0877,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.15429143607616425,
"rewards/margins": 0.29225456714630127,
"rewards/rejected": -0.4465459883213043,
"step": 2020
},
{
"epoch": 0.54,
"learning_rate": 2.575619398465402e-06,
"logits/chosen": -1.6102511882781982,
"logits/rejected": -0.9045922160148621,
"logps/chosen": -626.107666015625,
"logps/rejected": -1319.302978515625,
"loss": 0.0697,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.14732882380485535,
"rewards/margins": 0.3220018446445465,
"rewards/rejected": -0.46933069825172424,
"step": 2030
},
{
"epoch": 0.54,
"learning_rate": 2.5523560497083927e-06,
"logits/chosen": -1.7386415004730225,
"logits/rejected": -1.0501739978790283,
"logps/chosen": -698.34765625,
"logps/rejected": -1343.9432373046875,
"loss": 0.0424,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.20332041382789612,
"rewards/margins": 0.3136293292045593,
"rewards/rejected": -0.5169497728347778,
"step": 2040
},
{
"epoch": 0.55,
"learning_rate": 2.5290881645034932e-06,
"logits/chosen": -1.7993240356445312,
"logits/rejected": -0.9968295097351074,
"logps/chosen": -737.5123901367188,
"logps/rejected": -1283.9423828125,
"loss": 0.0839,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.2413882464170456,
"rewards/margins": 0.25109177827835083,
"rewards/rejected": -0.4924800395965576,
"step": 2050
},
{
"epoch": 0.55,
"learning_rate": 2.5058177589223766e-06,
"logits/chosen": -1.6581776142120361,
"logits/rejected": -1.2274023294448853,
"logps/chosen": -621.2349853515625,
"logps/rejected": -1304.399169921875,
"loss": 0.0534,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15700002014636993,
"rewards/margins": 0.3114224672317505,
"rewards/rejected": -0.4684225022792816,
"step": 2060
},
{
"epoch": 0.55,
"learning_rate": 2.482546849255096e-06,
"logits/chosen": -1.822373628616333,
"logits/rejected": -0.9462094306945801,
"logps/chosen": -783.3508911132812,
"logps/rejected": -1461.9649658203125,
"loss": 0.0723,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2161974459886551,
"rewards/margins": 0.3368082642555237,
"rewards/rejected": -0.5530056953430176,
"step": 2070
},
{
"epoch": 0.55,
"learning_rate": 2.4592774518353858e-06,
"logits/chosen": -1.8345363140106201,
"logits/rejected": -1.388167381286621,
"logps/chosen": -656.7705688476562,
"logps/rejected": -1437.629638671875,
"loss": 0.0601,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1848996877670288,
"rewards/margins": 0.3406218886375427,
"rewards/rejected": -0.5255215764045715,
"step": 2080
},
{
"epoch": 0.56,
"learning_rate": 2.436011582865945e-06,
"logits/chosen": -1.5363919734954834,
"logits/rejected": -1.1726293563842773,
"logps/chosen": -633.4727172851562,
"logps/rejected": -1333.555908203125,
"loss": 0.0652,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.17989441752433777,
"rewards/margins": 0.3212878704071045,
"rewards/rejected": -0.5011822581291199,
"step": 2090
},
{
"epoch": 0.56,
"learning_rate": 2.4127512582437486e-06,
"logits/chosen": -1.5197073221206665,
"logits/rejected": -0.9411822557449341,
"logps/chosen": -578.3282470703125,
"logps/rejected": -1179.3984375,
"loss": 0.0888,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.18851898610591888,
"rewards/margins": 0.27369505167007446,
"rewards/rejected": -0.46221405267715454,
"step": 2100
},
{
"epoch": 0.56,
"learning_rate": 2.3894984933853734e-06,
"logits/chosen": -1.599889874458313,
"logits/rejected": -0.9709933996200562,
"logps/chosen": -711.7459716796875,
"logps/rejected": -1445.8603515625,
"loss": 0.0469,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.21319825947284698,
"rewards/margins": 0.34705930948257446,
"rewards/rejected": -0.5602575540542603,
"step": 2110
},
{
"epoch": 0.57,
"learning_rate": 2.366255303052377e-06,
"logits/chosen": -1.5162818431854248,
"logits/rejected": -0.9374685287475586,
"logps/chosen": -663.5987548828125,
"logps/rejected": -1244.1162109375,
"loss": 0.0774,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.20607724785804749,
"rewards/margins": 0.28292593359947205,
"rewards/rejected": -0.48900318145751953,
"step": 2120
},
{
"epoch": 0.57,
"learning_rate": 2.3430237011767166e-06,
"logits/chosen": -1.9861704111099243,
"logits/rejected": -1.1126511096954346,
"logps/chosen": -685.7115478515625,
"logps/rejected": -1307.9677734375,
"loss": 0.0711,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.15337470173835754,
"rewards/margins": 0.3153078854084015,
"rewards/rejected": -0.46868258714675903,
"step": 2130
},
{
"epoch": 0.57,
"learning_rate": 2.319805700686257e-06,
"logits/chosen": -1.466512680053711,
"logits/rejected": -0.9953567385673523,
"logps/chosen": -656.9564819335938,
"logps/rejected": -1330.86572265625,
"loss": 0.0865,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.20789778232574463,
"rewards/margins": 0.27107080817222595,
"rewards/rejected": -0.4789685606956482,
"step": 2140
},
{
"epoch": 0.57,
"learning_rate": 2.296603313330355e-06,
"logits/chosen": -1.7993675470352173,
"logits/rejected": -0.795965313911438,
"logps/chosen": -716.7975463867188,
"logps/rejected": -1341.012451171875,
"loss": 0.0413,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -0.16811136901378632,
"rewards/margins": 0.3270968794822693,
"rewards/rejected": -0.4952082633972168,
"step": 2150
},
{
"epoch": 0.58,
"learning_rate": 2.2734185495055503e-06,
"logits/chosen": -1.7564256191253662,
"logits/rejected": -1.0065138339996338,
"logps/chosen": -605.0843505859375,
"logps/rejected": -1259.658447265625,
"loss": 0.0639,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.13863322138786316,
"rewards/margins": 0.3194582462310791,
"rewards/rejected": -0.4580914378166199,
"step": 2160
},
{
"epoch": 0.58,
"learning_rate": 2.250253418081373e-06,
"logits/chosen": -1.7674392461776733,
"logits/rejected": -1.2567976713180542,
"logps/chosen": -696.3572998046875,
"logps/rejected": -1424.5084228515625,
"loss": 0.0523,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.20887522399425507,
"rewards/margins": 0.35076838731765747,
"rewards/rejected": -0.5596436262130737,
"step": 2170
},
{
"epoch": 0.58,
"learning_rate": 2.22710992622628e-06,
"logits/chosen": -1.7430893182754517,
"logits/rejected": -1.2086124420166016,
"logps/chosen": -602.26513671875,
"logps/rejected": -1213.3702392578125,
"loss": 0.0766,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.17254558205604553,
"rewards/margins": 0.29195210337638855,
"rewards/rejected": -0.46449774503707886,
"step": 2180
},
{
"epoch": 0.58,
"learning_rate": 2.2039900792337477e-06,
"logits/chosen": -1.5437920093536377,
"logits/rejected": -1.2289955615997314,
"logps/chosen": -678.7044677734375,
"logps/rejected": -1379.6451416015625,
"loss": 0.0519,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.19320425391197205,
"rewards/margins": 0.32549750804901123,
"rewards/rejected": -0.5187016725540161,
"step": 2190
},
{
"epoch": 0.59,
"learning_rate": 2.1808958803485134e-06,
"logits/chosen": -1.5962891578674316,
"logits/rejected": -1.3936455249786377,
"logps/chosen": -606.307373046875,
"logps/rejected": -1305.8978271484375,
"loss": 0.0898,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2137691229581833,
"rewards/margins": 0.2590574324131012,
"rewards/rejected": -0.4728265702724457,
"step": 2200
},
{
"epoch": 0.59,
"learning_rate": 2.157829330593008e-06,
"logits/chosen": -1.7511285543441772,
"logits/rejected": -0.974310576915741,
"logps/chosen": -633.572509765625,
"logps/rejected": -1414.367431640625,
"loss": 0.0364,
"rewards/accuracies": 1.0,
"rewards/chosen": -0.18712595105171204,
"rewards/margins": 0.37235841155052185,
"rewards/rejected": -0.5594843626022339,
"step": 2210
},
{
"epoch": 0.59,
"learning_rate": 2.134792428593971e-06,
"logits/chosen": -1.644934058189392,
"logits/rejected": -1.11379075050354,
"logps/chosen": -639.0684814453125,
"logps/rejected": -1331.093505859375,
"loss": 0.0665,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2170993536710739,
"rewards/margins": 0.3065374493598938,
"rewards/rejected": -0.5236367583274841,
"step": 2220
},
{
"epoch": 0.59,
"learning_rate": 2.1117871704092818e-06,
"logits/chosen": -1.7506024837493896,
"logits/rejected": -0.7958993911743164,
"logps/chosen": -705.9612426757812,
"logps/rejected": -1230.204833984375,
"loss": 0.074,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.18289552628993988,
"rewards/margins": 0.2924391031265259,
"rewards/rejected": -0.47533464431762695,
"step": 2230
},
{
"epoch": 0.6,
"learning_rate": 2.0888155493550027e-06,
"logits/chosen": -1.74947190284729,
"logits/rejected": -1.175241231918335,
"logps/chosen": -695.1935424804688,
"logps/rejected": -1281.6820068359375,
"loss": 0.0659,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.227014422416687,
"rewards/margins": 0.3012952506542206,
"rewards/rejected": -0.5283096432685852,
"step": 2240
},
{
"epoch": 0.6,
"learning_rate": 2.0658795558326745e-06,
"logits/chosen": -1.3126758337020874,
"logits/rejected": -0.9601501226425171,
"logps/chosen": -722.9898681640625,
"logps/rejected": -1265.3829345703125,
"loss": 0.0976,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.26479509472846985,
"rewards/margins": 0.2494693100452423,
"rewards/rejected": -0.5142643451690674,
"step": 2250
},
{
"epoch": 0.6,
"learning_rate": 2.0429811771568468e-06,
"logits/chosen": -1.9693552255630493,
"logits/rejected": -1.2665364742279053,
"logps/chosen": -614.6461181640625,
"logps/rejected": -1254.004150390625,
"loss": 0.0682,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.18582487106323242,
"rewards/margins": 0.31714367866516113,
"rewards/rejected": -0.5029684901237488,
"step": 2260
},
{
"epoch": 0.61,
"learning_rate": 2.0201223973828917e-06,
"logits/chosen": -1.4698238372802734,
"logits/rejected": -0.9227995872497559,
"logps/chosen": -671.4839477539062,
"logps/rejected": -1292.746826171875,
"loss": 0.089,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.26453524827957153,
"rewards/margins": 0.270897775888443,
"rewards/rejected": -0.5354331135749817,
"step": 2270
},
{
"epoch": 0.61,
"learning_rate": 1.997305197135089e-06,
"logits/chosen": -1.694977045059204,
"logits/rejected": -1.0786057710647583,
"logps/chosen": -716.204833984375,
"logps/rejected": -1146.710205078125,
"loss": 0.0988,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.23150472342967987,
"rewards/margins": 0.21505391597747803,
"rewards/rejected": -0.4465586245059967,
"step": 2280
},
{
"epoch": 0.61,
"learning_rate": 1.9745315534350157e-06,
"logits/chosen": -1.5376017093658447,
"logits/rejected": -1.0097434520721436,
"logps/chosen": -613.5394287109375,
"logps/rejected": -1213.8609619140625,
"loss": 0.0891,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1808188557624817,
"rewards/margins": 0.28589263558387756,
"rewards/rejected": -0.46671146154403687,
"step": 2290
},
{
"epoch": 0.61,
"learning_rate": 1.9518034395302413e-06,
"logits/chosen": -1.6548999547958374,
"logits/rejected": -1.14958655834198,
"logps/chosen": -715.3568115234375,
"logps/rejected": -1187.842529296875,
"loss": 0.0824,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.22143769264221191,
"rewards/margins": 0.2518152594566345,
"rewards/rejected": -0.47325292229652405,
"step": 2300
},
{
"epoch": 0.62,
"learning_rate": 1.9291228247233607e-06,
"logits/chosen": -1.728780746459961,
"logits/rejected": -1.0694595575332642,
"logps/chosen": -720.1282958984375,
"logps/rejected": -1408.6763916015625,
"loss": 0.1004,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.22856228053569794,
"rewards/margins": 0.35453858971595764,
"rewards/rejected": -0.5831009149551392,
"step": 2310
},
{
"epoch": 0.62,
"learning_rate": 1.9064916742013515e-06,
"logits/chosen": -1.483896017074585,
"logits/rejected": -0.8495844602584839,
"logps/chosen": -686.5977783203125,
"logps/rejected": -1382.664794921875,
"loss": 0.0613,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.24766767024993896,
"rewards/margins": 0.31919145584106445,
"rewards/rejected": -0.5668591260910034,
"step": 2320
},
{
"epoch": 0.62,
"learning_rate": 1.883911948865306e-06,
"logits/chosen": -1.490492582321167,
"logits/rejected": -0.8722866177558899,
"logps/chosen": -710.7138671875,
"logps/rejected": -1393.343994140625,
"loss": 0.1017,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2688906192779541,
"rewards/margins": 0.2782156765460968,
"rewards/rejected": -0.5471062660217285,
"step": 2330
},
{
"epoch": 0.62,
"learning_rate": 1.8613856051605242e-06,
"logits/chosen": -1.5532639026641846,
"logits/rejected": -1.1383737325668335,
"logps/chosen": -775.1503295898438,
"logps/rejected": -1363.999267578125,
"loss": 0.094,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.2418668568134308,
"rewards/margins": 0.28089746832847595,
"rewards/rejected": -0.5227643251419067,
"step": 2340
},
{
"epoch": 0.63,
"learning_rate": 1.8389145949069953e-06,
"logits/chosen": -1.7076427936553955,
"logits/rejected": -1.0624377727508545,
"logps/chosen": -605.94287109375,
"logps/rejected": -1299.9039306640625,
"loss": 0.0606,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1737835705280304,
"rewards/margins": 0.30216971039772034,
"rewards/rejected": -0.47595319151878357,
"step": 2350
},
{
"epoch": 0.63,
"learning_rate": 1.816500865130279e-06,
"logits/chosen": -1.704677939414978,
"logits/rejected": -1.0432106256484985,
"logps/chosen": -584.9574584960938,
"logps/rejected": -1062.80078125,
"loss": 0.1013,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.17347513139247894,
"rewards/margins": 0.24399061501026154,
"rewards/rejected": -0.4174656867980957,
"step": 2360
},
{
"epoch": 0.63,
"learning_rate": 1.7941463578928088e-06,
"logits/chosen": -1.4847257137298584,
"logits/rejected": -1.036319375038147,
"logps/chosen": -565.1887817382812,
"logps/rejected": -1178.5665283203125,
"loss": 0.0822,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.18391203880310059,
"rewards/margins": 0.2579399347305298,
"rewards/rejected": -0.44185200333595276,
"step": 2370
},
{
"epoch": 0.63,
"learning_rate": 1.7718530101256115e-06,
"logits/chosen": -1.7321197986602783,
"logits/rejected": -1.0332145690917969,
"logps/chosen": -719.0687866210938,
"logps/rejected": -1425.6676025390625,
"loss": 0.052,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.23235881328582764,
"rewards/margins": 0.32954907417297363,
"rewards/rejected": -0.5619078874588013,
"step": 2380
},
{
"epoch": 0.64,
"learning_rate": 1.7496227534604859e-06,
"logits/chosen": -1.789006233215332,
"logits/rejected": -1.0403592586517334,
"logps/chosen": -728.0673828125,
"logps/rejected": -1402.5009765625,
"loss": 0.0629,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19960185885429382,
"rewards/margins": 0.30768799781799316,
"rewards/rejected": -0.5072898864746094,
"step": 2390
},
{
"epoch": 0.64,
"learning_rate": 1.7274575140626318e-06,
"logits/chosen": -1.9194923639297485,
"logits/rejected": -1.3303556442260742,
"logps/chosen": -637.9463500976562,
"logps/rejected": -1268.9755859375,
"loss": 0.0732,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1636119931936264,
"rewards/margins": 0.2945595383644104,
"rewards/rejected": -0.458171546459198,
"step": 2400
},
{
"epoch": 0.64,
"learning_rate": 1.7053592124637557e-06,
"logits/chosen": -1.7786033153533936,
"logits/rejected": -1.103823184967041,
"logps/chosen": -694.8350219726562,
"logps/rejected": -1348.7509765625,
"loss": 0.0788,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.20321345329284668,
"rewards/margins": 0.2868782877922058,
"rewards/rejected": -0.49009180068969727,
"step": 2410
},
{
"epoch": 0.65,
"learning_rate": 1.6833297633956647e-06,
"logits/chosen": -1.6945937871932983,
"logits/rejected": -1.1909369230270386,
"logps/chosen": -574.365966796875,
"logps/rejected": -1239.800537109375,
"loss": 0.0705,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1648532599210739,
"rewards/margins": 0.2992625832557678,
"rewards/rejected": -0.46411579847335815,
"step": 2420
},
{
"epoch": 0.65,
"learning_rate": 1.661371075624363e-06,
"logits/chosen": -1.5424587726593018,
"logits/rejected": -1.3184611797332764,
"logps/chosen": -526.9766845703125,
"logps/rejected": -1158.5274658203125,
"loss": 0.1027,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": -0.16265107691287994,
"rewards/margins": 0.26259398460388184,
"rewards/rejected": -0.42524510622024536,
"step": 2430
},
{
"epoch": 0.65,
"learning_rate": 1.6394850517846621e-06,
"logits/chosen": -1.6765334606170654,
"logits/rejected": -0.9357019662857056,
"logps/chosen": -637.3072509765625,
"logps/rejected": -1350.0201416015625,
"loss": 0.0483,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1886545568704605,
"rewards/margins": 0.32397735118865967,
"rewards/rejected": -0.5126319527626038,
"step": 2440
},
{
"epoch": 0.65,
"learning_rate": 1.6176735882153284e-06,
"logits/chosen": -1.4673728942871094,
"logits/rejected": -0.7847996950149536,
"logps/chosen": -721.7418212890625,
"logps/rejected": -1300.6024169921875,
"loss": 0.07,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.20094910264015198,
"rewards/margins": 0.28207510709762573,
"rewards/rejected": -0.4830242097377777,
"step": 2450
},
{
"epoch": 0.66,
"learning_rate": 1.5959385747947697e-06,
"logits/chosen": -1.8030465841293335,
"logits/rejected": -1.1619970798492432,
"logps/chosen": -672.4512329101562,
"logps/rejected": -1321.999267578125,
"loss": 0.0729,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20753267407417297,
"rewards/margins": 0.27218765020370483,
"rewards/rejected": -0.4797203540802002,
"step": 2460
},
{
"epoch": 0.66,
"learning_rate": 1.5742818947772875e-06,
"logits/chosen": -1.4533730745315552,
"logits/rejected": -0.9818112254142761,
"logps/chosen": -737.9328002929688,
"logps/rejected": -1423.608154296875,
"loss": 0.0509,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.23550765216350555,
"rewards/margins": 0.3242724537849426,
"rewards/rejected": -0.5597800612449646,
"step": 2470
},
{
"epoch": 0.66,
"learning_rate": 1.552705424629898e-06,
"logits/chosen": -1.8896198272705078,
"logits/rejected": -1.0289218425750732,
"logps/chosen": -772.728515625,
"logps/rejected": -1228.538818359375,
"loss": 0.0765,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.2038073092699051,
"rewards/margins": 0.26559290289878845,
"rewards/rejected": -0.4694002568721771,
"step": 2480
},
{
"epoch": 0.66,
"learning_rate": 1.5312110338697427e-06,
"logits/chosen": -1.698685646057129,
"logits/rejected": -1.106320858001709,
"logps/chosen": -706.8604736328125,
"logps/rejected": -1375.6136474609375,
"loss": 0.0583,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.21833567321300507,
"rewards/margins": 0.31459516286849976,
"rewards/rejected": -0.5329307317733765,
"step": 2490
},
{
"epoch": 0.67,
"learning_rate": 1.509800584902108e-06,
"logits/chosen": -1.5073596239089966,
"logits/rejected": -0.9609388113021851,
"logps/chosen": -698.6984252929688,
"logps/rejected": -1362.0128173828125,
"loss": 0.0613,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.20453910529613495,
"rewards/margins": 0.3083449900150299,
"rewards/rejected": -0.5128840208053589,
"step": 2500
},
{
"epoch": 0.67,
"learning_rate": 1.4884759328590476e-06,
"logits/chosen": -1.6541383266448975,
"logits/rejected": -1.2222260236740112,
"logps/chosen": -686.1484375,
"logps/rejected": -1445.005615234375,
"loss": 0.0586,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.20519807934761047,
"rewards/margins": 0.32793715596199036,
"rewards/rejected": -0.5331352353096008,
"step": 2510
},
{
"epoch": 0.67,
"learning_rate": 1.467238925438646e-06,
"logits/chosen": -1.5963326692581177,
"logits/rejected": -1.0982022285461426,
"logps/chosen": -682.7689208984375,
"logps/rejected": -1145.491943359375,
"loss": 0.1334,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.21956536173820496,
"rewards/margins": 0.209082692861557,
"rewards/rejected": -0.42864808440208435,
"step": 2520
},
{
"epoch": 0.67,
"learning_rate": 1.446091402744923e-06,
"logits/chosen": -1.4184305667877197,
"logits/rejected": -0.9969242215156555,
"logps/chosen": -631.0663452148438,
"logps/rejected": -1298.8023681640625,
"loss": 0.0789,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2030848264694214,
"rewards/margins": 0.2959844470024109,
"rewards/rejected": -0.4990692138671875,
"step": 2530
},
{
"epoch": 0.68,
"learning_rate": 1.4250351971283937e-06,
"logits/chosen": -1.6301358938217163,
"logits/rejected": -1.1058179140090942,
"logps/chosen": -733.7390747070312,
"logps/rejected": -1312.249267578125,
"loss": 0.0789,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.2173445224761963,
"rewards/margins": 0.2981737554073334,
"rewards/rejected": -0.515518307685852,
"step": 2540
},
{
"epoch": 0.68,
"learning_rate": 1.4040721330273063e-06,
"logits/chosen": -1.737184762954712,
"logits/rejected": -0.8671928644180298,
"logps/chosen": -759.1162719726562,
"logps/rejected": -1388.3258056640625,
"loss": 0.0713,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.2206466943025589,
"rewards/margins": 0.301429808139801,
"rewards/rejected": -0.5220764875411987,
"step": 2550
},
{
"epoch": 0.68,
"learning_rate": 1.3832040268095589e-06,
"logits/chosen": -1.6039609909057617,
"logits/rejected": -0.8329612612724304,
"logps/chosen": -676.49755859375,
"logps/rejected": -1311.5169677734375,
"loss": 0.0776,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1940799504518509,
"rewards/margins": 0.3204619288444519,
"rewards/rejected": -0.5145418643951416,
"step": 2560
},
{
"epoch": 0.69,
"learning_rate": 1.362432686615316e-06,
"logits/chosen": -1.7901986837387085,
"logits/rejected": -1.0667396783828735,
"logps/chosen": -656.2354736328125,
"logps/rejected": -1316.062744140625,
"loss": 0.0758,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1804797351360321,
"rewards/margins": 0.30354127287864685,
"rewards/rejected": -0.48402100801467896,
"step": 2570
},
{
"epoch": 0.69,
"learning_rate": 1.3417599122003464e-06,
"logits/chosen": -1.5734020471572876,
"logits/rejected": -1.0475807189941406,
"logps/chosen": -566.615966796875,
"logps/rejected": -1251.5736083984375,
"loss": 0.0652,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.16305221617221832,
"rewards/margins": 0.3017520308494568,
"rewards/rejected": -0.4648042619228363,
"step": 2580
},
{
"epoch": 0.69,
"learning_rate": 1.3211874947800747e-06,
"logits/chosen": -1.821344017982483,
"logits/rejected": -1.1782556772232056,
"logps/chosen": -690.4012451171875,
"logps/rejected": -1318.4063720703125,
"loss": 0.0693,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21744892001152039,
"rewards/margins": 0.29652294516563416,
"rewards/rejected": -0.5139719247817993,
"step": 2590
},
{
"epoch": 0.69,
"learning_rate": 1.3007172168743854e-06,
"logits/chosen": -1.438684105873108,
"logits/rejected": -0.9294763803482056,
"logps/chosen": -742.8309326171875,
"logps/rejected": -1415.3411865234375,
"loss": 0.0694,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2644263803958893,
"rewards/margins": 0.3054746091365814,
"rewards/rejected": -0.5699009895324707,
"step": 2600
},
{
"epoch": 0.7,
"learning_rate": 1.280350852153168e-06,
"logits/chosen": -1.6277799606323242,
"logits/rejected": -1.0877095460891724,
"logps/chosen": -703.9315795898438,
"logps/rejected": -1340.6427001953125,
"loss": 0.0651,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.23966872692108154,
"rewards/margins": 0.2736600339412689,
"rewards/rejected": -0.5133287310600281,
"step": 2610
},
{
"epoch": 0.7,
"learning_rate": 1.260090165282645e-06,
"logits/chosen": -1.6082645654678345,
"logits/rejected": -0.9585866928100586,
"logps/chosen": -733.3088989257812,
"logps/rejected": -1261.9248046875,
"loss": 0.0997,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.23271843791007996,
"rewards/margins": 0.248284250497818,
"rewards/rejected": -0.48100265860557556,
"step": 2620
},
{
"epoch": 0.7,
"learning_rate": 1.2399369117724582e-06,
"logits/chosen": -1.703330397605896,
"logits/rejected": -0.9837586283683777,
"logps/chosen": -662.8977661132812,
"logps/rejected": -1302.231201171875,
"loss": 0.0872,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.17935553193092346,
"rewards/margins": 0.30029359459877014,
"rewards/rejected": -0.4796491265296936,
"step": 2630
},
{
"epoch": 0.7,
"learning_rate": 1.2198928378235717e-06,
"logits/chosen": -1.6038057804107666,
"logits/rejected": -1.2413532733917236,
"logps/chosen": -518.2135009765625,
"logps/rejected": -1082.8677978515625,
"loss": 0.1043,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16320912539958954,
"rewards/margins": 0.2511638402938843,
"rewards/rejected": -0.414372980594635,
"step": 2640
},
{
"epoch": 0.71,
"learning_rate": 1.1999596801769617e-06,
"logits/chosen": -1.6547168493270874,
"logits/rejected": -1.0749003887176514,
"logps/chosen": -674.54541015625,
"logps/rejected": -1409.031982421875,
"loss": 0.0616,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.18622934818267822,
"rewards/margins": 0.3446107506752014,
"rewards/rejected": -0.5308400988578796,
"step": 2650
},
{
"epoch": 0.71,
"learning_rate": 1.1801391659631423e-06,
"logits/chosen": -1.5979634523391724,
"logits/rejected": -1.0661733150482178,
"logps/chosen": -460.7833557128906,
"logps/rejected": -1178.478759765625,
"loss": 0.055,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1438596248626709,
"rewards/margins": 0.30475491285324097,
"rewards/rejected": -0.4486145079135895,
"step": 2660
},
{
"epoch": 0.71,
"learning_rate": 1.160433012552508e-06,
"logits/chosen": -1.8192787170410156,
"logits/rejected": -1.0415807962417603,
"logps/chosen": -610.6654052734375,
"logps/rejected": -1276.691162109375,
"loss": 0.0536,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.15597601234912872,
"rewards/margins": 0.3177304267883301,
"rewards/rejected": -0.4737063944339752,
"step": 2670
},
{
"epoch": 0.71,
"learning_rate": 1.1408429274065418e-06,
"logits/chosen": -1.4305180311203003,
"logits/rejected": -1.164957880973816,
"logps/chosen": -624.7916870117188,
"logps/rejected": -1304.414306640625,
"loss": 0.0804,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.18183110654354095,
"rewards/margins": 0.29839175939559937,
"rewards/rejected": -0.4802228808403015,
"step": 2680
},
{
"epoch": 0.72,
"learning_rate": 1.1213706079298566e-06,
"logits/chosen": -1.6688525676727295,
"logits/rejected": -1.1245152950286865,
"logps/chosen": -636.1717529296875,
"logps/rejected": -1271.97802734375,
"loss": 0.0642,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19803071022033691,
"rewards/margins": 0.29812103509902954,
"rewards/rejected": -0.49615174531936646,
"step": 2690
},
{
"epoch": 0.72,
"learning_rate": 1.1020177413231334e-06,
"logits/chosen": -1.4946445226669312,
"logits/rejected": -0.8901990056037903,
"logps/chosen": -699.8790893554688,
"logps/rejected": -1338.4324951171875,
"loss": 0.0797,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.2173963338136673,
"rewards/margins": 0.31975775957107544,
"rewards/rejected": -0.5371540784835815,
"step": 2700
},
{
"epoch": 0.72,
"learning_rate": 1.0827860044369226e-06,
"logits/chosen": -1.8231594562530518,
"logits/rejected": -0.8806201815605164,
"logps/chosen": -766.8648071289062,
"logps/rejected": -1401.1787109375,
"loss": 0.05,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.21393120288848877,
"rewards/margins": 0.3332614004611969,
"rewards/rejected": -0.5471926331520081,
"step": 2710
},
{
"epoch": 0.73,
"learning_rate": 1.06367706362636e-06,
"logits/chosen": -1.6795213222503662,
"logits/rejected": -1.0750644207000732,
"logps/chosen": -631.3731689453125,
"logps/rejected": -1225.7191162109375,
"loss": 0.0755,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.18206752836704254,
"rewards/margins": 0.2682720720767975,
"rewards/rejected": -0.45033949613571167,
"step": 2720
},
{
"epoch": 0.73,
"learning_rate": 1.0446925746067768e-06,
"logits/chosen": -1.4441897869110107,
"logits/rejected": -0.9866140484809875,
"logps/chosen": -579.9114379882812,
"logps/rejected": -1187.9873046875,
"loss": 0.0804,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.16464915871620178,
"rewards/margins": 0.2846793234348297,
"rewards/rejected": -0.4493285119533539,
"step": 2730
},
{
"epoch": 0.73,
"learning_rate": 1.0258341823102418e-06,
"logits/chosen": -1.6576883792877197,
"logits/rejected": -1.0028798580169678,
"logps/chosen": -635.6063232421875,
"logps/rejected": -1112.1954345703125,
"loss": 0.0729,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.16143657267093658,
"rewards/margins": 0.22869448363780975,
"rewards/rejected": -0.39013105630874634,
"step": 2740
},
{
"epoch": 0.73,
"learning_rate": 1.0071035207430352e-06,
"logits/chosen": -1.6806875467300415,
"logits/rejected": -0.9319907426834106,
"logps/chosen": -598.1236572265625,
"logps/rejected": -1265.469970703125,
"loss": 0.0684,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.14364568889141083,
"rewards/margins": 0.3332282602787018,
"rewards/rejected": -0.4768740236759186,
"step": 2750
},
{
"epoch": 0.74,
"learning_rate": 9.88502212844063e-07,
"logits/chosen": -1.6131244897842407,
"logits/rejected": -1.3068736791610718,
"logps/chosen": -547.2989501953125,
"logps/rejected": -1120.500244140625,
"loss": 0.0879,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.1456841677427292,
"rewards/margins": 0.27742549777030945,
"rewards/rejected": -0.42310968041419983,
"step": 2760
},
{
"epoch": 0.74,
"learning_rate": 9.700318703442437e-07,
"logits/chosen": -1.7162357568740845,
"logits/rejected": -1.1048786640167236,
"logps/chosen": -623.9015502929688,
"logps/rejected": -1287.5198974609375,
"loss": 0.0766,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15075866878032684,
"rewards/margins": 0.31179898977279663,
"rewards/rejected": -0.4625576436519623,
"step": 2770
},
{
"epoch": 0.74,
"learning_rate": 9.516940936268504e-07,
"logits/chosen": -1.7830989360809326,
"logits/rejected": -1.1697231531143188,
"logps/chosen": -662.9132080078125,
"logps/rejected": -1347.837646484375,
"loss": 0.0518,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.21208259463310242,
"rewards/margins": 0.3107925057411194,
"rewards/rejected": -0.5228751301765442,
"step": 2780
},
{
"epoch": 0.74,
"learning_rate": 9.334904715888496e-07,
"logits/chosen": -1.7367998361587524,
"logits/rejected": -0.8941909670829773,
"logps/chosen": -727.5262451171875,
"logps/rejected": -1364.0263671875,
"loss": 0.072,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.1890576183795929,
"rewards/margins": 0.31730249524116516,
"rewards/rejected": -0.5063600540161133,
"step": 2790
},
{
"epoch": 0.75,
"learning_rate": 9.154225815032242e-07,
"logits/chosen": -1.7351534366607666,
"logits/rejected": -0.9410954713821411,
"logps/chosen": -572.9583740234375,
"logps/rejected": -1397.9078369140625,
"loss": 0.0414,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.14058896899223328,
"rewards/margins": 0.37250643968582153,
"rewards/rejected": -0.5130953788757324,
"step": 2800
},
{
"epoch": 0.75,
"learning_rate": 8.974919888823164e-07,
"logits/chosen": -1.6442314386367798,
"logits/rejected": -1.245091438293457,
"logps/chosen": -625.5496215820312,
"logps/rejected": -1145.61865234375,
"loss": 0.1189,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.1761692762374878,
"rewards/margins": 0.21311041712760925,
"rewards/rejected": -0.38927969336509705,
"step": 2810
},
{
"epoch": 0.75,
"learning_rate": 8.797002473421729e-07,
"logits/chosen": -1.5787712335586548,
"logits/rejected": -1.1393609046936035,
"logps/chosen": -564.6102905273438,
"logps/rejected": -1168.3687744140625,
"loss": 0.0749,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.15366871654987335,
"rewards/margins": 0.2722090780735016,
"rewards/rejected": -0.42587780952453613,
"step": 2820
},
{
"epoch": 0.75,
"learning_rate": 8.620488984679378e-07,
"logits/chosen": -1.747091293334961,
"logits/rejected": -1.1368491649627686,
"logps/chosen": -637.8484497070312,
"logps/rejected": -1327.956298828125,
"loss": 0.0344,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.18866987526416779,
"rewards/margins": 0.32480621337890625,
"rewards/rejected": -0.5134760737419128,
"step": 2830
},
{
"epoch": 0.76,
"learning_rate": 8.445394716802754e-07,
"logits/chosen": -1.7131919860839844,
"logits/rejected": -1.1550548076629639,
"logps/chosen": -685.2410888671875,
"logps/rejected": -1205.682861328125,
"loss": 0.097,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20656362175941467,
"rewards/margins": 0.2589021325111389,
"rewards/rejected": -0.4654656946659088,
"step": 2840
},
{
"epoch": 0.76,
"learning_rate": 8.271734841028553e-07,
"logits/chosen": -1.637920618057251,
"logits/rejected": -1.3879964351654053,
"logps/chosen": -607.6511840820312,
"logps/rejected": -1199.6259765625,
"loss": 0.0784,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1954393833875656,
"rewards/margins": 0.25878021121025085,
"rewards/rejected": -0.4542195796966553,
"step": 2850
},
{
"epoch": 0.76,
"learning_rate": 8.099524404308948e-07,
"logits/chosen": -1.4263249635696411,
"logits/rejected": -1.002600908279419,
"logps/chosen": -636.8242797851562,
"logps/rejected": -1328.989501953125,
"loss": 0.0539,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2097894698381424,
"rewards/margins": 0.3159825801849365,
"rewards/rejected": -0.5257720351219177,
"step": 2860
},
{
"epoch": 0.77,
"learning_rate": 7.928778328007918e-07,
"logits/chosen": -1.884615182876587,
"logits/rejected": -0.9678564071655273,
"logps/chosen": -658.4131469726562,
"logps/rejected": -1356.146240234375,
"loss": 0.0639,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.20153768360614777,
"rewards/margins": 0.3334410786628723,
"rewards/rejected": -0.5349787473678589,
"step": 2870
},
{
"epoch": 0.77,
"learning_rate": 7.759511406608255e-07,
"logits/chosen": -1.3170682191848755,
"logits/rejected": -0.7453212141990662,
"logps/chosen": -650.6473999023438,
"logps/rejected": -1283.6512451171875,
"loss": 0.0787,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.20364892482757568,
"rewards/margins": 0.28083834052085876,
"rewards/rejected": -0.48448723554611206,
"step": 2880
},
{
"epoch": 0.77,
"learning_rate": 7.591738306429769e-07,
"logits/chosen": -1.6163628101348877,
"logits/rejected": -0.9143250584602356,
"logps/chosen": -624.8411865234375,
"logps/rejected": -1358.2152099609375,
"loss": 0.05,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.17616188526153564,
"rewards/margins": 0.3604372441768646,
"rewards/rejected": -0.5365991592407227,
"step": 2890
},
{
"epoch": 0.77,
"learning_rate": 7.425473564358457e-07,
"logits/chosen": -1.734678030014038,
"logits/rejected": -1.1536650657653809,
"logps/chosen": -641.8426513671875,
"logps/rejected": -1439.745849609375,
"loss": 0.0529,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.19275933504104614,
"rewards/margins": 0.3525656759738922,
"rewards/rejected": -0.5453251004219055,
"step": 2900
},
{
"epoch": 0.78,
"learning_rate": 7.260731586586983e-07,
"logits/chosen": -1.6836907863616943,
"logits/rejected": -0.9171245694160461,
"logps/chosen": -621.8990478515625,
"logps/rejected": -1194.991943359375,
"loss": 0.0742,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15242193639278412,
"rewards/margins": 0.2880454659461975,
"rewards/rejected": -0.4404674172401428,
"step": 2910
},
{
"epoch": 0.78,
"learning_rate": 7.097526647366379e-07,
"logits/chosen": -1.5863702297210693,
"logits/rejected": -1.0138695240020752,
"logps/chosen": -717.5474853515625,
"logps/rejected": -1396.6239013671875,
"loss": 0.0631,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.19577325880527496,
"rewards/margins": 0.3056796193122864,
"rewards/rejected": -0.5014528632164001,
"step": 2920
},
{
"epoch": 0.78,
"learning_rate": 6.935872887769299e-07,
"logits/chosen": -1.5059714317321777,
"logits/rejected": -0.9991394281387329,
"logps/chosen": -704.1346435546875,
"logps/rejected": -1267.742919921875,
"loss": 0.0853,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.2067212611436844,
"rewards/margins": 0.26990193128585815,
"rewards/rejected": -0.4766232371330261,
"step": 2930
},
{
"epoch": 0.78,
"learning_rate": 6.775784314464717e-07,
"logits/chosen": -1.5541185140609741,
"logits/rejected": -1.0399705171585083,
"logps/chosen": -592.094482421875,
"logps/rejected": -1375.5118408203125,
"loss": 0.0491,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.19767038524150848,
"rewards/margins": 0.3308340013027191,
"rewards/rejected": -0.5285044312477112,
"step": 2940
},
{
"epoch": 0.79,
"learning_rate": 6.617274798504286e-07,
"logits/chosen": -1.4966031312942505,
"logits/rejected": -0.8529815673828125,
"logps/chosen": -752.8843994140625,
"logps/rejected": -1323.464599609375,
"loss": 0.0798,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.24442438781261444,
"rewards/margins": 0.27123716473579407,
"rewards/rejected": -0.5156615972518921,
"step": 2950
},
{
"epoch": 0.79,
"learning_rate": 6.460358074120518e-07,
"logits/chosen": -1.9015638828277588,
"logits/rejected": -1.1131356954574585,
"logps/chosen": -598.8160400390625,
"logps/rejected": -1302.127197265625,
"loss": 0.0637,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15112538635730743,
"rewards/margins": 0.34708380699157715,
"rewards/rejected": -0.4982091784477234,
"step": 2960
},
{
"epoch": 0.79,
"learning_rate": 6.305047737536707e-07,
"logits/chosen": -1.577651858329773,
"logits/rejected": -1.0997803211212158,
"logps/chosen": -706.3680419921875,
"logps/rejected": -1251.1690673828125,
"loss": 0.0795,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.22735171020030975,
"rewards/margins": 0.2633035480976105,
"rewards/rejected": -0.4906553328037262,
"step": 2970
},
{
"epoch": 0.79,
"learning_rate": 6.151357245788917e-07,
"logits/chosen": -1.5803357362747192,
"logits/rejected": -1.0134937763214111,
"logps/chosen": -787.1359252929688,
"logps/rejected": -1268.341064453125,
"loss": 0.0776,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.2360551357269287,
"rewards/margins": 0.2515987157821655,
"rewards/rejected": -0.4876538813114166,
"step": 2980
},
{
"epoch": 0.8,
"learning_rate": 5.999299915559956e-07,
"logits/chosen": -1.863478422164917,
"logits/rejected": -1.186255693435669,
"logps/chosen": -670.4616088867188,
"logps/rejected": -1253.962646484375,
"loss": 0.0807,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.18092487752437592,
"rewards/margins": 0.2903370261192322,
"rewards/rejected": -0.4712619185447693,
"step": 2990
},
{
"epoch": 0.8,
"learning_rate": 5.848888922025553e-07,
"logits/chosen": -1.4436417818069458,
"logits/rejected": -1.0620644092559814,
"logps/chosen": -752.1005859375,
"logps/rejected": -1337.766357421875,
"loss": 0.0742,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.252055823802948,
"rewards/margins": 0.2509405016899109,
"rewards/rejected": -0.5029963254928589,
"step": 3000
},
{
"epoch": 0.8,
"learning_rate": 5.700137297712749e-07,
"logits/chosen": -1.4030582904815674,
"logits/rejected": -1.0030864477157593,
"logps/chosen": -657.2237548828125,
"logps/rejected": -1292.746826171875,
"loss": 0.0651,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.227573424577713,
"rewards/margins": 0.301101952791214,
"rewards/rejected": -0.5286754369735718,
"step": 3010
},
{
"epoch": 0.81,
"learning_rate": 5.553057931370729e-07,
"logits/chosen": -1.7350914478302002,
"logits/rejected": -1.2158474922180176,
"logps/chosen": -638.9827270507812,
"logps/rejected": -1461.6263427734375,
"loss": 0.0521,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19272522628307343,
"rewards/margins": 0.3386714458465576,
"rewards/rejected": -0.5313966274261475,
"step": 3020
},
{
"epoch": 0.81,
"learning_rate": 5.407663566854008e-07,
"logits/chosen": -1.6435270309448242,
"logits/rejected": -1.0143921375274658,
"logps/chosen": -617.9176635742188,
"logps/rejected": -1159.302734375,
"loss": 0.0887,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.19033944606781006,
"rewards/margins": 0.2596290409564972,
"rewards/rejected": -0.44996848702430725,
"step": 3030
},
{
"epoch": 0.81,
"learning_rate": 5.263966802018275e-07,
"logits/chosen": -1.5586180686950684,
"logits/rejected": -0.9451137781143188,
"logps/chosen": -577.8426513671875,
"logps/rejected": -1369.449951171875,
"loss": 0.0498,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.17990294098854065,
"rewards/margins": 0.33183303475379944,
"rewards/rejected": -0.5117359757423401,
"step": 3040
},
{
"epoch": 0.81,
"learning_rate": 5.121980087628802e-07,
"logits/chosen": -1.7278461456298828,
"logits/rejected": -1.0558230876922607,
"logps/chosen": -643.5891723632812,
"logps/rejected": -1279.1890869140625,
"loss": 0.0769,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.19444116950035095,
"rewards/margins": 0.27763885259628296,
"rewards/rejected": -0.4720799922943115,
"step": 3050
},
{
"epoch": 0.82,
"learning_rate": 4.981715726281666e-07,
"logits/chosen": -1.3910518884658813,
"logits/rejected": -0.9968549609184265,
"logps/chosen": -648.2385864257812,
"logps/rejected": -1314.0189208984375,
"loss": 0.0556,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.20168296992778778,
"rewards/margins": 0.27340278029441833,
"rewards/rejected": -0.4750857949256897,
"step": 3060
},
{
"epoch": 0.82,
"learning_rate": 4.843185871337722e-07,
"logits/chosen": -1.7175220251083374,
"logits/rejected": -1.0445207357406616,
"logps/chosen": -652.8839111328125,
"logps/rejected": -1321.618896484375,
"loss": 0.0565,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.16258391737937927,
"rewards/margins": 0.3201899528503418,
"rewards/rejected": -0.48277387022972107,
"step": 3070
},
{
"epoch": 0.82,
"learning_rate": 4.706402525869633e-07,
"logits/chosen": -1.6909739971160889,
"logits/rejected": -1.2913880348205566,
"logps/chosen": -679.966796875,
"logps/rejected": -1319.14794921875,
"loss": 0.0825,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1967785656452179,
"rewards/margins": 0.2842392921447754,
"rewards/rejected": -0.4810178279876709,
"step": 3080
},
{
"epoch": 0.82,
"learning_rate": 4.5713775416217884e-07,
"logits/chosen": -1.450941801071167,
"logits/rejected": -1.0710570812225342,
"logps/chosen": -729.031494140625,
"logps/rejected": -1396.8394775390625,
"loss": 0.0777,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.23177771270275116,
"rewards/margins": 0.28798457980155945,
"rewards/rejected": -0.5197622776031494,
"step": 3090
},
{
"epoch": 0.83,
"learning_rate": 4.438122617983442e-07,
"logits/chosen": -1.2658835649490356,
"logits/rejected": -0.6509383916854858,
"logps/chosen": -678.7276611328125,
"logps/rejected": -1277.664794921875,
"loss": 0.1047,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1934930384159088,
"rewards/margins": 0.27904012799263,
"rewards/rejected": -0.47253313660621643,
"step": 3100
},
{
"epoch": 0.83,
"learning_rate": 4.3066493009749853e-07,
"logits/chosen": -1.8485597372055054,
"logits/rejected": -1.3042134046554565,
"logps/chosen": -598.8069458007812,
"logps/rejected": -1223.86767578125,
"loss": 0.0587,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.16556212306022644,
"rewards/margins": 0.29780706763267517,
"rewards/rejected": -0.4633691906929016,
"step": 3110
},
{
"epoch": 0.83,
"learning_rate": 4.1769689822475147e-07,
"logits/chosen": -1.780917763710022,
"logits/rejected": -1.0629148483276367,
"logps/chosen": -598.9677734375,
"logps/rejected": -1247.2830810546875,
"loss": 0.0715,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.17943796515464783,
"rewards/margins": 0.29289960861206055,
"rewards/rejected": -0.472337543964386,
"step": 3120
},
{
"epoch": 0.83,
"learning_rate": 4.049092898095816e-07,
"logits/chosen": -1.6170413494110107,
"logits/rejected": -1.0848805904388428,
"logps/chosen": -687.8775634765625,
"logps/rejected": -1320.7880859375,
"loss": 0.0641,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.16832710802555084,
"rewards/margins": 0.3181244730949402,
"rewards/rejected": -0.4864516258239746,
"step": 3130
},
{
"epoch": 0.84,
"learning_rate": 3.9230321284847856e-07,
"logits/chosen": -1.835370659828186,
"logits/rejected": -0.9373771548271179,
"logps/chosen": -725.0065307617188,
"logps/rejected": -1372.621337890625,
"loss": 0.0804,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.18989510834217072,
"rewards/margins": 0.32323604822158813,
"rewards/rejected": -0.5131311416625977,
"step": 3140
},
{
"epoch": 0.84,
"learning_rate": 3.798797596089351e-07,
"logits/chosen": -1.5447750091552734,
"logits/rejected": -0.8559864163398743,
"logps/chosen": -728.7130126953125,
"logps/rejected": -1214.0345458984375,
"loss": 0.0969,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2228674590587616,
"rewards/margins": 0.24048006534576416,
"rewards/rejected": -0.46334752440452576,
"step": 3150
},
{
"epoch": 0.84,
"learning_rate": 3.6764000653481263e-07,
"logits/chosen": -1.5830670595169067,
"logits/rejected": -1.0614241361618042,
"logps/chosen": -623.8796997070312,
"logps/rejected": -1382.993408203125,
"loss": 0.0584,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.17642198503017426,
"rewards/margins": 0.3370228409767151,
"rewards/rejected": -0.5134447813034058,
"step": 3160
},
{
"epoch": 0.85,
"learning_rate": 3.555850141530659e-07,
"logits/chosen": -1.6500043869018555,
"logits/rejected": -0.9364662170410156,
"logps/chosen": -665.5390014648438,
"logps/rejected": -1273.1171875,
"loss": 0.054,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.14827661216259003,
"rewards/margins": 0.31309622526168823,
"rewards/rejected": -0.46137291193008423,
"step": 3170
},
{
"epoch": 0.85,
"learning_rate": 3.4371582698185636e-07,
"logits/chosen": -1.5176864862442017,
"logits/rejected": -1.098508596420288,
"logps/chosen": -559.7102661132812,
"logps/rejected": -1274.8511962890625,
"loss": 0.0613,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.16253405809402466,
"rewards/margins": 0.30766361951828003,
"rewards/rejected": -0.4701976180076599,
"step": 3180
},
{
"epoch": 0.85,
"learning_rate": 3.3203347344004737e-07,
"logits/chosen": -1.6515905857086182,
"logits/rejected": -1.023119568824768,
"logps/chosen": -841.9296875,
"logps/rejected": -1383.3951416015625,
"loss": 0.0841,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.2267489731311798,
"rewards/margins": 0.2902681827545166,
"rewards/rejected": -0.517017126083374,
"step": 3190
},
{
"epoch": 0.85,
"learning_rate": 3.2053896575809426e-07,
"logits/chosen": -1.6936334371566772,
"logits/rejected": -1.1252939701080322,
"logps/chosen": -680.880859375,
"logps/rejected": -1214.6029052734375,
"loss": 0.0888,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.20804938673973083,
"rewards/margins": 0.2518993020057678,
"rewards/rejected": -0.45994871854782104,
"step": 3200
},
{
"epoch": 0.86,
"learning_rate": 3.092332998903416e-07,
"logits/chosen": -1.7081215381622314,
"logits/rejected": -0.7063247561454773,
"logps/chosen": -839.2431640625,
"logps/rejected": -1423.2156982421875,
"loss": 0.053,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.2406737506389618,
"rewards/margins": 0.3139367997646332,
"rewards/rejected": -0.5546106100082397,
"step": 3210
},
{
"epoch": 0.86,
"learning_rate": 2.981174554287239e-07,
"logits/chosen": -1.6795532703399658,
"logits/rejected": -1.0844824314117432,
"logps/chosen": -483.9825134277344,
"logps/rejected": -1096.248779296875,
"loss": 0.0943,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1309995800256729,
"rewards/margins": 0.2878738045692444,
"rewards/rejected": -0.4188733994960785,
"step": 3220
},
{
"epoch": 0.86,
"learning_rate": 2.871923955178918e-07,
"logits/chosen": -1.651178002357483,
"logits/rejected": -0.8967401385307312,
"logps/chosen": -714.80810546875,
"logps/rejected": -1329.046630859375,
"loss": 0.0553,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.21358323097229004,
"rewards/margins": 0.28250735998153687,
"rewards/rejected": -0.4960905909538269,
"step": 3230
},
{
"epoch": 0.86,
"learning_rate": 2.764590667717562e-07,
"logits/chosen": -1.6797590255737305,
"logits/rejected": -1.2603545188903809,
"logps/chosen": -557.3433227539062,
"logps/rejected": -1234.2099609375,
"loss": 0.0775,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.14634376764297485,
"rewards/margins": 0.3086283206939697,
"rewards/rejected": -0.4549720883369446,
"step": 3240
},
{
"epoch": 0.87,
"learning_rate": 2.6591839919146963e-07,
"logits/chosen": -1.477281928062439,
"logits/rejected": -0.9903861880302429,
"logps/chosen": -665.1192626953125,
"logps/rejected": -1360.6173095703125,
"loss": 0.0868,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.20300845801830292,
"rewards/margins": 0.2980322539806366,
"rewards/rejected": -0.5010407567024231,
"step": 3250
},
{
"epoch": 0.87,
"learning_rate": 2.555713060848433e-07,
"logits/chosen": -1.2929660081863403,
"logits/rejected": -1.0018800497055054,
"logps/chosen": -685.843017578125,
"logps/rejected": -1353.837158203125,
"loss": 0.0817,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21414248645305634,
"rewards/margins": 0.30294984579086304,
"rewards/rejected": -0.5170923471450806,
"step": 3260
},
{
"epoch": 0.87,
"learning_rate": 2.454186839872158e-07,
"logits/chosen": -1.6706600189208984,
"logits/rejected": -1.0511361360549927,
"logps/chosen": -549.2362670898438,
"logps/rejected": -1177.3765869140625,
"loss": 0.0754,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1762518584728241,
"rewards/margins": 0.275063693523407,
"rewards/rejected": -0.45131558179855347,
"step": 3270
},
{
"epoch": 0.87,
"learning_rate": 2.3546141258376786e-07,
"logits/chosen": -1.6843605041503906,
"logits/rejected": -1.0689982175827026,
"logps/chosen": -718.2612915039062,
"logps/rejected": -1315.400634765625,
"loss": 0.1008,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20562107861042023,
"rewards/margins": 0.2962108850479126,
"rewards/rejected": -0.5018320083618164,
"step": 3280
},
{
"epoch": 0.88,
"learning_rate": 2.257003546333042e-07,
"logits/chosen": -1.6436039209365845,
"logits/rejected": -1.0740540027618408,
"logps/chosen": -530.2700805664062,
"logps/rejected": -1222.441650390625,
"loss": 0.0701,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.17040416598320007,
"rewards/margins": 0.3154822885990143,
"rewards/rejected": -0.48588642477989197,
"step": 3290
},
{
"epoch": 0.88,
"learning_rate": 2.1613635589349756e-07,
"logits/chosen": -1.7491300106048584,
"logits/rejected": -0.9703881144523621,
"logps/chosen": -634.4661865234375,
"logps/rejected": -1327.0771484375,
"loss": 0.0478,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.15670377016067505,
"rewards/margins": 0.3363407254219055,
"rewards/rejected": -0.49304452538490295,
"step": 3300
},
{
"epoch": 0.88,
"learning_rate": 2.0677024504760752e-07,
"logits/chosen": -1.5196778774261475,
"logits/rejected": -1.040668249130249,
"logps/chosen": -678.7161865234375,
"logps/rejected": -1322.600830078125,
"loss": 0.0634,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.19689127802848816,
"rewards/margins": 0.32300588488578796,
"rewards/rejected": -0.5198971033096313,
"step": 3310
},
{
"epoch": 0.89,
"learning_rate": 1.9760283363267684e-07,
"logits/chosen": -1.5344351530075073,
"logits/rejected": -1.066190481185913,
"logps/chosen": -577.383056640625,
"logps/rejected": -1262.2083740234375,
"loss": 0.0653,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.14838716387748718,
"rewards/margins": 0.31335335969924927,
"rewards/rejected": -0.4617405831813812,
"step": 3320
},
{
"epoch": 0.89,
"learning_rate": 1.8863491596921745e-07,
"logits/chosen": -1.5965193510055542,
"logits/rejected": -1.0383847951889038,
"logps/chosen": -608.2281494140625,
"logps/rejected": -1189.1614990234375,
"loss": 0.0814,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.1726662814617157,
"rewards/margins": 0.26928216218948364,
"rewards/rejected": -0.44194841384887695,
"step": 3330
},
{
"epoch": 0.89,
"learning_rate": 1.798672690923828e-07,
"logits/chosen": -1.3265117406845093,
"logits/rejected": -0.8752411007881165,
"logps/chosen": -684.5750732421875,
"logps/rejected": -1294.5103759765625,
"loss": 0.0898,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21347585320472717,
"rewards/margins": 0.2741771936416626,
"rewards/rejected": -0.4876530170440674,
"step": 3340
},
{
"epoch": 0.89,
"learning_rate": 1.713006526846439e-07,
"logits/chosen": -1.5635359287261963,
"logits/rejected": -1.1060292720794678,
"logps/chosen": -639.0496826171875,
"logps/rejected": -1266.63720703125,
"loss": 0.0836,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2038009911775589,
"rewards/margins": 0.26461145281791687,
"rewards/rejected": -0.4684123992919922,
"step": 3350
},
{
"epoch": 0.9,
"learning_rate": 1.629358090099639e-07,
"logits/chosen": -1.5050264596939087,
"logits/rejected": -1.0238497257232666,
"logps/chosen": -667.7071533203125,
"logps/rejected": -1466.681884765625,
"loss": 0.0488,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.21334633231163025,
"rewards/margins": 0.3355127274990082,
"rewards/rejected": -0.5488591194152832,
"step": 3360
},
{
"epoch": 0.9,
"learning_rate": 1.5477346284948292e-07,
"logits/chosen": -1.8201059103012085,
"logits/rejected": -1.0349600315093994,
"logps/chosen": -649.684326171875,
"logps/rejected": -1250.92236328125,
"loss": 0.053,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.1781463921070099,
"rewards/margins": 0.3080436587333679,
"rewards/rejected": -0.4861900210380554,
"step": 3370
},
{
"epoch": 0.9,
"learning_rate": 1.4681432143872133e-07,
"logits/chosen": -1.6552870273590088,
"logits/rejected": -1.2580300569534302,
"logps/chosen": -656.0045166015625,
"logps/rejected": -1372.251953125,
"loss": 0.0669,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.19065485894680023,
"rewards/margins": 0.3371616005897522,
"rewards/rejected": -0.5278164744377136,
"step": 3380
},
{
"epoch": 0.9,
"learning_rate": 1.3905907440629752e-07,
"logits/chosen": -1.679107666015625,
"logits/rejected": -1.0372817516326904,
"logps/chosen": -586.6466674804688,
"logps/rejected": -1267.5791015625,
"loss": 0.0689,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.18233391642570496,
"rewards/margins": 0.279081255197525,
"rewards/rejected": -0.4614151418209076,
"step": 3390
},
{
"epoch": 0.91,
"learning_rate": 1.31508393714177e-07,
"logits/chosen": -1.4867794513702393,
"logits/rejected": -0.8477977514266968,
"logps/chosen": -554.5364990234375,
"logps/rejected": -1226.228271484375,
"loss": 0.0688,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16267123818397522,
"rewards/margins": 0.3219824433326721,
"rewards/rejected": -0.4846537113189697,
"step": 3400
},
{
"epoch": 0.91,
"learning_rate": 1.241629335994471e-07,
"logits/chosen": -1.8708140850067139,
"logits/rejected": -1.0227950811386108,
"logps/chosen": -799.6752319335938,
"logps/rejected": -1433.6239013671875,
"loss": 0.0827,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2366192787885666,
"rewards/margins": 0.31682151556015015,
"rewards/rejected": -0.5534407496452332,
"step": 3410
},
{
"epoch": 0.91,
"learning_rate": 1.1702333051763271e-07,
"logits/chosen": -1.6258529424667358,
"logits/rejected": -1.1475646495819092,
"logps/chosen": -691.2940063476562,
"logps/rejected": -1324.8748779296875,
"loss": 0.0872,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20850765705108643,
"rewards/margins": 0.28260093927383423,
"rewards/rejected": -0.49110865592956543,
"step": 3420
},
{
"epoch": 0.91,
"learning_rate": 1.1009020308754587e-07,
"logits/chosen": -1.414668083190918,
"logits/rejected": -1.0075281858444214,
"logps/chosen": -676.2430419921875,
"logps/rejected": -1345.2977294921875,
"loss": 0.0937,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.22368064522743225,
"rewards/margins": 0.27526089549064636,
"rewards/rejected": -0.498941570520401,
"step": 3430
},
{
"epoch": 0.92,
"learning_rate": 1.0336415203768962e-07,
"logits/chosen": -1.6306703090667725,
"logits/rejected": -1.1734403371810913,
"logps/chosen": -728.4344482421875,
"logps/rejected": -1319.7030029296875,
"loss": 0.068,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.21984486281871796,
"rewards/margins": 0.26837268471717834,
"rewards/rejected": -0.4882175326347351,
"step": 3440
},
{
"epoch": 0.92,
"learning_rate": 9.684576015420277e-08,
"logits/chosen": -1.685302734375,
"logits/rejected": -1.3321112394332886,
"logps/chosen": -745.2166748046875,
"logps/rejected": -1243.215087890625,
"loss": 0.0751,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.1889590322971344,
"rewards/margins": 0.25796008110046387,
"rewards/rejected": -0.44691914319992065,
"step": 3450
},
{
"epoch": 0.92,
"learning_rate": 9.053559223036746e-08,
"logits/chosen": -1.6785064935684204,
"logits/rejected": -1.05990731716156,
"logps/chosen": -662.5062255859375,
"logps/rejected": -1327.50390625,
"loss": 0.0552,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.21094217896461487,
"rewards/margins": 0.28824371099472046,
"rewards/rejected": -0.49918586015701294,
"step": 3460
},
{
"epoch": 0.93,
"learning_rate": 8.44341950176683e-08,
"logits/chosen": -1.6093642711639404,
"logits/rejected": -0.9099162817001343,
"logps/chosen": -605.3573608398438,
"logps/rejected": -1294.749755859375,
"loss": 0.0693,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.16440364718437195,
"rewards/margins": 0.3216980993747711,
"rewards/rejected": -0.4861017167568207,
"step": 3470
},
{
"epoch": 0.93,
"learning_rate": 7.854209717842231e-08,
"logits/chosen": -1.6602537631988525,
"logits/rejected": -1.277896761894226,
"logps/chosen": -698.0342407226562,
"logps/rejected": -1383.7066650390625,
"loss": 0.0846,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2171173393726349,
"rewards/margins": 0.3016865849494934,
"rewards/rejected": -0.5188038945198059,
"step": 3480
},
{
"epoch": 0.93,
"learning_rate": 7.285980923996989e-08,
"logits/chosen": -1.7102686166763306,
"logits/rejected": -1.154813528060913,
"logps/chosen": -705.587158203125,
"logps/rejected": -1409.553955078125,
"loss": 0.088,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.23871013522148132,
"rewards/margins": 0.29305535554885864,
"rewards/rejected": -0.5317655205726624,
"step": 3490
},
{
"epoch": 0.93,
"learning_rate": 6.738782355044048e-08,
"logits/chosen": -1.7222769260406494,
"logits/rejected": -1.0485639572143555,
"logps/chosen": -677.2478637695312,
"logps/rejected": -1264.8682861328125,
"loss": 0.0694,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.21438893675804138,
"rewards/margins": 0.27774578332901,
"rewards/rejected": -0.4921347498893738,
"step": 3500
},
{
"epoch": 0.94,
"learning_rate": 6.212661423609184e-08,
"logits/chosen": -1.6767486333847046,
"logits/rejected": -0.995303750038147,
"logps/chosen": -749.8323974609375,
"logps/rejected": -1487.445068359375,
"loss": 0.0432,
"rewards/accuracies": 0.9750000238418579,
"rewards/chosen": -0.21066558361053467,
"rewards/margins": 0.3550676703453064,
"rewards/rejected": -0.5657332539558411,
"step": 3510
},
{
"epoch": 0.94,
"learning_rate": 5.707663716023021e-08,
"logits/chosen": -1.4784590005874634,
"logits/rejected": -0.6975718140602112,
"logps/chosen": -666.94189453125,
"logps/rejected": -1402.3831787109375,
"loss": 0.0401,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.20545247197151184,
"rewards/margins": 0.34110763669013977,
"rewards/rejected": -0.5465600490570068,
"step": 3520
},
{
"epoch": 0.94,
"learning_rate": 5.22383298837098e-08,
"logits/chosen": -1.527578353881836,
"logits/rejected": -0.8852685689926147,
"logps/chosen": -677.0675659179688,
"logps/rejected": -1252.460205078125,
"loss": 0.0794,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.20508520305156708,
"rewards/margins": 0.2905604839324951,
"rewards/rejected": -0.4956456124782562,
"step": 3530
},
{
"epoch": 0.94,
"learning_rate": 4.761211162702117e-08,
"logits/chosen": -1.4756338596343994,
"logits/rejected": -0.9907282590866089,
"logps/chosen": -628.2028198242188,
"logps/rejected": -1525.0224609375,
"loss": 0.0322,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.21087181568145752,
"rewards/margins": 0.3604298532009125,
"rewards/rejected": -0.5713016390800476,
"step": 3540
},
{
"epoch": 0.95,
"learning_rate": 4.319838323396691e-08,
"logits/chosen": -1.6325428485870361,
"logits/rejected": -1.2529757022857666,
"logps/chosen": -640.9552612304688,
"logps/rejected": -1428.263916015625,
"loss": 0.0422,
"rewards/accuracies": 0.8999999761581421,
"rewards/chosen": -0.1934538334608078,
"rewards/margins": 0.32024726271629333,
"rewards/rejected": -0.5137011408805847,
"step": 3550
},
{
"epoch": 0.95,
"learning_rate": 3.8997527136930004e-08,
"logits/chosen": -1.505150556564331,
"logits/rejected": -0.8225961923599243,
"logps/chosen": -707.0218505859375,
"logps/rejected": -1327.9573974609375,
"loss": 0.0923,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.2203945368528366,
"rewards/margins": 0.2966047525405884,
"rewards/rejected": -0.5169993042945862,
"step": 3560
},
{
"epoch": 0.95,
"learning_rate": 3.5009907323737826e-08,
"logits/chosen": -1.5440739393234253,
"logits/rejected": -0.8097125887870789,
"logps/chosen": -710.9814453125,
"logps/rejected": -1306.008056640625,
"loss": 0.0667,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.21110442280769348,
"rewards/margins": 0.3028753995895386,
"rewards/rejected": -0.5139797925949097,
"step": 3570
},
{
"epoch": 0.95,
"learning_rate": 3.1235869306123766e-08,
"logits/chosen": -1.7210094928741455,
"logits/rejected": -1.13853120803833,
"logps/chosen": -723.6201782226562,
"logps/rejected": -1426.6231689453125,
"loss": 0.062,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.2129667103290558,
"rewards/margins": 0.31347301602363586,
"rewards/rejected": -0.5264397263526917,
"step": 3580
},
{
"epoch": 0.96,
"learning_rate": 2.767574008979007e-08,
"logits/chosen": -1.492082953453064,
"logits/rejected": -0.8750694990158081,
"logps/chosen": -754.5972290039062,
"logps/rejected": -1386.701904296875,
"loss": 0.0783,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.23436252772808075,
"rewards/margins": 0.2830619215965271,
"rewards/rejected": -0.517424464225769,
"step": 3590
},
{
"epoch": 0.96,
"learning_rate": 2.4329828146074096e-08,
"logits/chosen": -1.768261194229126,
"logits/rejected": -1.1734797954559326,
"logps/chosen": -634.6853637695312,
"logps/rejected": -1281.8953857421875,
"loss": 0.0754,
"rewards/accuracies": 0.8500000238418579,
"rewards/chosen": -0.19491219520568848,
"rewards/margins": 0.29776549339294434,
"rewards/rejected": -0.4926777482032776,
"step": 3600
},
{
"epoch": 0.96,
"learning_rate": 2.1198423385220822e-08,
"logits/chosen": -1.518048882484436,
"logits/rejected": -1.1481419801712036,
"logps/chosen": -578.7421875,
"logps/rejected": -1115.6094970703125,
"loss": 0.1095,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -0.18905951082706451,
"rewards/margins": 0.2177170068025589,
"rewards/rejected": -0.4067765176296234,
"step": 3610
},
{
"epoch": 0.97,
"learning_rate": 1.82817971312621e-08,
"logits/chosen": -1.632956862449646,
"logits/rejected": -1.2727057933807373,
"logps/chosen": -638.421875,
"logps/rejected": -1247.95751953125,
"loss": 0.0849,
"rewards/accuracies": 0.75,
"rewards/chosen": -0.20175893604755402,
"rewards/margins": 0.25679153203964233,
"rewards/rejected": -0.45855045318603516,
"step": 3620
},
{
"epoch": 0.97,
"learning_rate": 1.5580202098509078e-08,
"logits/chosen": -1.6741609573364258,
"logits/rejected": -1.1588261127471924,
"logps/chosen": -625.3444213867188,
"logps/rejected": -1219.2210693359375,
"loss": 0.0667,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.15933595597743988,
"rewards/margins": 0.29343315958976746,
"rewards/rejected": -0.4527691900730133,
"step": 3630
},
{
"epoch": 0.97,
"learning_rate": 1.3093872369654148e-08,
"logits/chosen": -1.5549055337905884,
"logits/rejected": -1.1253769397735596,
"logps/chosen": -622.0947265625,
"logps/rejected": -1193.876220703125,
"loss": 0.1058,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -0.21242782473564148,
"rewards/margins": 0.2279575765132904,
"rewards/rejected": -0.4403854012489319,
"step": 3640
},
{
"epoch": 0.97,
"learning_rate": 1.0823023375489128e-08,
"logits/chosen": -1.463275671005249,
"logits/rejected": -0.7773112058639526,
"logps/chosen": -759.4178466796875,
"logps/rejected": -1502.75732421875,
"loss": 0.0479,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.24037644267082214,
"rewards/margins": 0.34176188707351685,
"rewards/rejected": -0.5821383595466614,
"step": 3650
},
{
"epoch": 0.98,
"learning_rate": 8.767851876239075e-09,
"logits/chosen": -1.8317701816558838,
"logits/rejected": -0.975128173828125,
"logps/chosen": -716.7714233398438,
"logps/rejected": -1325.99072265625,
"loss": 0.0745,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.19285240769386292,
"rewards/margins": 0.2963181734085083,
"rewards/rejected": -0.48917055130004883,
"step": 3660
},
{
"epoch": 0.98,
"learning_rate": 6.9285359445145366e-09,
"logits/chosen": -1.6910631656646729,
"logits/rejected": -1.0203049182891846,
"logps/chosen": -723.9457397460938,
"logps/rejected": -1435.2896728515625,
"loss": 0.0543,
"rewards/accuracies": 0.949999988079071,
"rewards/chosen": -0.20491650700569153,
"rewards/margins": 0.3260103464126587,
"rewards/rejected": -0.5309268832206726,
"step": 3670
},
{
"epoch": 0.98,
"learning_rate": 5.305234949880001e-09,
"logits/chosen": -1.5100289583206177,
"logits/rejected": -0.9951759576797485,
"logps/chosen": -733.7612915039062,
"logps/rejected": -1351.5802001953125,
"loss": 0.0599,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.2286415845155716,
"rewards/margins": 0.2934814393520355,
"rewards/rejected": -0.5221229791641235,
"step": 3680
},
{
"epoch": 0.98,
"learning_rate": 3.8980895450474455e-09,
"logits/chosen": -1.6468843221664429,
"logits/rejected": -0.8552689552307129,
"logps/chosen": -657.6549072265625,
"logps/rejected": -1334.9576416015625,
"loss": 0.0688,
"rewards/accuracies": 0.875,
"rewards/chosen": -0.1897544115781784,
"rewards/margins": 0.32274651527404785,
"rewards/rejected": -0.5125009417533875,
"step": 3690
},
{
"epoch": 0.99,
"learning_rate": 2.7072216536885855e-09,
"logits/chosen": -1.424248218536377,
"logits/rejected": -0.8001385927200317,
"logps/chosen": -590.4542846679688,
"logps/rejected": -1241.249267578125,
"loss": 0.0646,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.16355466842651367,
"rewards/margins": 0.31166428327560425,
"rewards/rejected": -0.47521892189979553,
"step": 3700
},
{
"epoch": 0.99,
"learning_rate": 1.7327344598702667e-09,
"logits/chosen": -1.6445577144622803,
"logits/rejected": -0.9098326563835144,
"logps/chosen": -725.5911865234375,
"logps/rejected": -1269.808349609375,
"loss": 0.0797,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": -0.19787754118442535,
"rewards/margins": 0.27285638451576233,
"rewards/rejected": -0.47073397040367126,
"step": 3710
},
{
"epoch": 0.99,
"learning_rate": 9.747123991141193e-10,
"logits/chosen": -1.4280940294265747,
"logits/rejected": -0.8096100687980652,
"logps/chosen": -702.1490478515625,
"logps/rejected": -1242.6480712890625,
"loss": 0.0495,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.1968768984079361,
"rewards/margins": 0.2870542109012604,
"rewards/rejected": -0.48393112421035767,
"step": 3720
},
{
"epoch": 0.99,
"learning_rate": 4.332211510807427e-10,
"logits/chosen": -1.5423029661178589,
"logits/rejected": -1.002629041671753,
"logps/chosen": -668.7158813476562,
"logps/rejected": -1202.1529541015625,
"loss": 0.0746,
"rewards/accuracies": 0.800000011920929,
"rewards/chosen": -0.20696516335010529,
"rewards/margins": 0.26598578691482544,
"rewards/rejected": -0.47295087575912476,
"step": 3730
},
{
"epoch": 1.0,
"learning_rate": 1.0830763387897902e-10,
"logits/chosen": -1.3063008785247803,
"logits/rejected": -1.1998459100723267,
"logps/chosen": -585.5635375976562,
"logps/rejected": -1489.9176025390625,
"loss": 0.0559,
"rewards/accuracies": 0.925000011920929,
"rewards/chosen": -0.20169933140277863,
"rewards/margins": 0.3584030270576477,
"rewards/rejected": -0.5601023435592651,
"step": 3740
},
{
"epoch": 1.0,
"learning_rate": 0.0,
"logits/chosen": -1.7910579442977905,
"logits/rejected": -1.1715670824050903,
"logps/chosen": -708.5948486328125,
"logps/rejected": -1328.570556640625,
"loss": 0.0773,
"rewards/accuracies": 0.824999988079071,
"rewards/chosen": -0.21266476809978485,
"rewards/margins": 0.2909637689590454,
"rewards/rejected": -0.5036285519599915,
"step": 3750
},
{
"epoch": 1.0,
"step": 3750,
"total_flos": 0.0,
"train_loss": 0.0801518168369929,
"train_runtime": 15780.2418,
"train_samples_per_second": 0.951,
"train_steps_per_second": 0.238
}
],
"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
}