| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9982631930527722, | |
| "eval_steps": 400, | |
| "global_step": 467, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01068804275217101, | |
| "grad_norm": 47.923506570215594, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -1.0110366344451904, | |
| "logits/rejected": -0.9818881750106812, | |
| "logps/chosen": -0.27409863471984863, | |
| "logps/rejected": -0.27151164412498474, | |
| "loss": 3.0607, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": -2.7409865856170654, | |
| "rewards/margins": -0.025869915261864662, | |
| "rewards/rejected": -2.715116500854492, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02137608550434202, | |
| "grad_norm": 39.987585891736785, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -1.0418651103973389, | |
| "logits/rejected": -0.9748126864433289, | |
| "logps/chosen": -0.2945522964000702, | |
| "logps/rejected": -0.29994362592697144, | |
| "loss": 3.0104, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -2.9455230236053467, | |
| "rewards/margins": 0.05391312763094902, | |
| "rewards/rejected": -2.999436378479004, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03206412825651302, | |
| "grad_norm": 52.07278122268582, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -0.963701069355011, | |
| "logits/rejected": -0.9835487604141235, | |
| "logps/chosen": -0.2644619345664978, | |
| "logps/rejected": -0.3007102608680725, | |
| "loss": 3.0162, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -2.6446194648742676, | |
| "rewards/margins": 0.362483412027359, | |
| "rewards/rejected": -3.0071024894714355, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04275217100868404, | |
| "grad_norm": 93.33861075914483, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -0.9671205282211304, | |
| "logits/rejected": -0.9406957626342773, | |
| "logps/chosen": -0.27761051058769226, | |
| "logps/rejected": -0.2907746732234955, | |
| "loss": 2.9342, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -2.7761049270629883, | |
| "rewards/margins": 0.13164177536964417, | |
| "rewards/rejected": -2.9077467918395996, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053440213760855046, | |
| "grad_norm": 52.349708457694014, | |
| "learning_rate": 5.319148936170212e-07, | |
| "logits/chosen": -1.015834093093872, | |
| "logits/rejected": -0.9864752888679504, | |
| "logps/chosen": -0.2717323899269104, | |
| "logps/rejected": -0.27839282155036926, | |
| "loss": 3.1216, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -2.7173242568969727, | |
| "rewards/margins": 0.06660404056310654, | |
| "rewards/rejected": -2.783928394317627, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06412825651302605, | |
| "grad_norm": 45.104515251326376, | |
| "learning_rate": 6.382978723404255e-07, | |
| "logits/chosen": -0.9981824159622192, | |
| "logits/rejected": -0.9536676406860352, | |
| "logps/chosen": -0.2733208239078522, | |
| "logps/rejected": -0.2788906693458557, | |
| "loss": 2.9453, | |
| "rewards/accuracies": 0.4437499940395355, | |
| "rewards/chosen": -2.733208179473877, | |
| "rewards/margins": 0.055698495358228683, | |
| "rewards/rejected": -2.7889065742492676, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07481629926519706, | |
| "grad_norm": 61.54928932943931, | |
| "learning_rate": 7.446808510638297e-07, | |
| "logits/chosen": -1.051733136177063, | |
| "logits/rejected": -0.9763606190681458, | |
| "logps/chosen": -0.2938762605190277, | |
| "logps/rejected": -0.3207188844680786, | |
| "loss": 2.9156, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -2.938762664794922, | |
| "rewards/margins": 0.26842620968818665, | |
| "rewards/rejected": -3.2071890830993652, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08550434201736808, | |
| "grad_norm": 55.913783341396325, | |
| "learning_rate": 8.51063829787234e-07, | |
| "logits/chosen": -1.0160491466522217, | |
| "logits/rejected": -0.9717121124267578, | |
| "logps/chosen": -0.27992749214172363, | |
| "logps/rejected": -0.32374969124794006, | |
| "loss": 2.9079, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -2.7992749214172363, | |
| "rewards/margins": 0.43822187185287476, | |
| "rewards/rejected": -3.237496852874756, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09619238476953908, | |
| "grad_norm": 38.79733201252679, | |
| "learning_rate": 9.574468085106384e-07, | |
| "logits/chosen": -1.0506359338760376, | |
| "logits/rejected": -1.0073621273040771, | |
| "logps/chosen": -0.3326144218444824, | |
| "logps/rejected": -0.38409319519996643, | |
| "loss": 2.9658, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -3.3261444568634033, | |
| "rewards/margins": 0.5147874355316162, | |
| "rewards/rejected": -3.8409321308135986, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10688042752171009, | |
| "grad_norm": 101.77454221179983, | |
| "learning_rate": 9.998741174712533e-07, | |
| "logits/chosen": -1.028257131576538, | |
| "logits/rejected": -0.9783049821853638, | |
| "logps/chosen": -0.3342127203941345, | |
| "logps/rejected": -0.3756522536277771, | |
| "loss": 2.9987, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -3.342127561569214, | |
| "rewards/margins": 0.4143945574760437, | |
| "rewards/rejected": -3.7565224170684814, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11756847027388109, | |
| "grad_norm": 70.06029649060484, | |
| "learning_rate": 9.991050648838675e-07, | |
| "logits/chosen": -1.0614262819290161, | |
| "logits/rejected": -1.025525689125061, | |
| "logps/chosen": -0.2905944287776947, | |
| "logps/rejected": -0.35211512446403503, | |
| "loss": 2.7815, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -2.905944347381592, | |
| "rewards/margins": 0.6152070164680481, | |
| "rewards/rejected": -3.521151065826416, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1282565130260521, | |
| "grad_norm": 49.123079394299815, | |
| "learning_rate": 9.97637968732563e-07, | |
| "logits/chosen": -1.0964637994766235, | |
| "logits/rejected": -1.061679720878601, | |
| "logps/chosen": -0.3209289014339447, | |
| "logps/rejected": -0.3418692350387573, | |
| "loss": 2.8596, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -3.209289073944092, | |
| "rewards/margins": 0.2094031274318695, | |
| "rewards/rejected": -3.4186923503875732, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13894455577822312, | |
| "grad_norm": 53.59523574650431, | |
| "learning_rate": 9.954748808839674e-07, | |
| "logits/chosen": -1.0083563327789307, | |
| "logits/rejected": -0.9795120358467102, | |
| "logps/chosen": -0.3694208264350891, | |
| "logps/rejected": -0.4273703694343567, | |
| "loss": 2.7899, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -3.6942081451416016, | |
| "rewards/margins": 0.5794947743415833, | |
| "rewards/rejected": -4.273703098297119, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14963259853039412, | |
| "grad_norm": 39.11989937521066, | |
| "learning_rate": 9.926188266120295e-07, | |
| "logits/chosen": -1.02675461769104, | |
| "logits/rejected": -1.0018466711044312, | |
| "logps/chosen": -0.35180264711380005, | |
| "logps/rejected": -0.4284419119358063, | |
| "loss": 2.8671, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -3.518026828765869, | |
| "rewards/margins": 0.7663925290107727, | |
| "rewards/rejected": -4.284419059753418, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16032064128256512, | |
| "grad_norm": 51.11281867224414, | |
| "learning_rate": 9.890738003669027e-07, | |
| "logits/chosen": -0.9933602213859558, | |
| "logits/rejected": -0.9224111437797546, | |
| "logps/chosen": -0.3594875931739807, | |
| "logps/rejected": -0.40996867418289185, | |
| "loss": 2.7704, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -3.5948760509490967, | |
| "rewards/margins": 0.5048106908798218, | |
| "rewards/rejected": -4.099686622619629, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17100868403473615, | |
| "grad_norm": 48.022103189017436, | |
| "learning_rate": 9.848447601883433e-07, | |
| "logits/chosen": -0.9609634280204773, | |
| "logits/rejected": -0.9471040964126587, | |
| "logps/chosen": -0.35821908712387085, | |
| "logps/rejected": -0.45667845010757446, | |
| "loss": 2.6966, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -3.582190990447998, | |
| "rewards/margins": 0.984593391418457, | |
| "rewards/rejected": -4.566784858703613, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18169672678690715, | |
| "grad_norm": 54.03450562178558, | |
| "learning_rate": 9.799376207714444e-07, | |
| "logits/chosen": -0.9785356521606445, | |
| "logits/rejected": -0.9566847085952759, | |
| "logps/chosen": -0.3405897319316864, | |
| "logps/rejected": -0.4017128050327301, | |
| "loss": 2.6144, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -3.405897617340088, | |
| "rewards/margins": 0.6112309098243713, | |
| "rewards/rejected": -4.0171284675598145, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19238476953907815, | |
| "grad_norm": 62.750052897303675, | |
| "learning_rate": 9.743592451943998e-07, | |
| "logits/chosen": -1.0191900730133057, | |
| "logits/rejected": -0.9845901727676392, | |
| "logps/chosen": -0.4232923090457916, | |
| "logps/rejected": -0.5109944939613342, | |
| "loss": 2.866, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -4.23292350769043, | |
| "rewards/margins": 0.8770216107368469, | |
| "rewards/rejected": -5.109944820404053, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20307281229124916, | |
| "grad_norm": 55.61240306403997, | |
| "learning_rate": 9.681174353198686e-07, | |
| "logits/chosen": -1.1014890670776367, | |
| "logits/rejected": -1.0177241563796997, | |
| "logps/chosen": -0.4533822536468506, | |
| "logps/rejected": -0.4995104670524597, | |
| "loss": 2.7432, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -4.533822536468506, | |
| "rewards/margins": 0.4612821042537689, | |
| "rewards/rejected": -4.995104789733887, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21376085504342018, | |
| "grad_norm": 80.5027346612393, | |
| "learning_rate": 9.612209208833646e-07, | |
| "logits/chosen": -0.9957372546195984, | |
| "logits/rejected": -0.9701834917068481, | |
| "logps/chosen": -0.43816161155700684, | |
| "logps/rejected": -0.5128804445266724, | |
| "loss": 2.7813, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -4.381616592407227, | |
| "rewards/margins": 0.7471875548362732, | |
| "rewards/rejected": -5.128803253173828, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22444889779559118, | |
| "grad_norm": 66.31806821536476, | |
| "learning_rate": 9.536793472839324e-07, | |
| "logits/chosen": -0.9997787475585938, | |
| "logits/rejected": -0.947482705116272, | |
| "logps/chosen": -0.4254922866821289, | |
| "logps/rejected": -0.5347083806991577, | |
| "loss": 2.7046, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -4.254923343658447, | |
| "rewards/margins": 1.092160701751709, | |
| "rewards/rejected": -5.347084045410156, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23513694054776219, | |
| "grad_norm": 61.1266120827584, | |
| "learning_rate": 9.455032620941839e-07, | |
| "logits/chosen": -0.9583929181098938, | |
| "logits/rejected": -0.8993922472000122, | |
| "logps/chosen": -0.4909549355506897, | |
| "logps/rejected": -0.620493471622467, | |
| "loss": 2.6559, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -4.909549713134766, | |
| "rewards/margins": 1.2953848838806152, | |
| "rewards/rejected": -6.204934120178223, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2458249832999332, | |
| "grad_norm": 66.56145340935555, | |
| "learning_rate": 9.367041003085648e-07, | |
| "logits/chosen": -1.019431471824646, | |
| "logits/rejected": -0.9595627784729004, | |
| "logps/chosen": -0.5270282030105591, | |
| "logps/rejected": -0.600238025188446, | |
| "loss": 2.4928, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -5.27028226852417, | |
| "rewards/margins": 0.7320979833602905, | |
| "rewards/rejected": -6.00238037109375, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2565130260521042, | |
| "grad_norm": 66.16205862286387, | |
| "learning_rate": 9.272941683504808e-07, | |
| "logits/chosen": -0.9745362997055054, | |
| "logits/rejected": -0.8843653798103333, | |
| "logps/chosen": -0.5472803115844727, | |
| "logps/rejected": -0.7492850422859192, | |
| "loss": 2.3982, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -5.472803115844727, | |
| "rewards/margins": 2.020047187805176, | |
| "rewards/rejected": -7.492850303649902, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26720106880427524, | |
| "grad_norm": 70.88843943146098, | |
| "learning_rate": 9.172866268606513e-07, | |
| "logits/chosen": -1.0500959157943726, | |
| "logits/rejected": -1.007611632347107, | |
| "logps/chosen": -0.6212247610092163, | |
| "logps/rejected": -0.7247714996337891, | |
| "loss": 2.3233, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -6.212247371673584, | |
| "rewards/margins": 1.0354671478271484, | |
| "rewards/rejected": -7.247714042663574, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27788911155644624, | |
| "grad_norm": 94.40161191780366, | |
| "learning_rate": 9.066954722907638e-07, | |
| "logits/chosen": -1.0675666332244873, | |
| "logits/rejected": -1.0614221096038818, | |
| "logps/chosen": -0.6142371892929077, | |
| "logps/rejected": -0.8813148736953735, | |
| "loss": 2.1102, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -6.1423726081848145, | |
| "rewards/margins": 2.670776844024658, | |
| "rewards/rejected": -8.813148498535156, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.28857715430861725, | |
| "grad_norm": 71.42739738901432, | |
| "learning_rate": 8.955355173281707e-07, | |
| "logits/chosen": -1.0529481172561646, | |
| "logits/rejected": -1.0047996044158936, | |
| "logps/chosen": -0.7235802412033081, | |
| "logps/rejected": -0.8823626637458801, | |
| "loss": 2.1377, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -7.235803127288818, | |
| "rewards/margins": 1.5878244638442993, | |
| "rewards/rejected": -8.823626518249512, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29926519706078825, | |
| "grad_norm": 87.9759333714625, | |
| "learning_rate": 8.838223701790055e-07, | |
| "logits/chosen": -1.1124293804168701, | |
| "logits/rejected": -1.0896517038345337, | |
| "logps/chosen": -0.862978458404541, | |
| "logps/rejected": -1.0037717819213867, | |
| "loss": 2.1017, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -8.62978458404541, | |
| "rewards/margins": 1.4079326391220093, | |
| "rewards/rejected": -10.037717819213867, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30995323981295925, | |
| "grad_norm": 78.07225371686874, | |
| "learning_rate": 8.71572412738697e-07, | |
| "logits/chosen": -1.030829906463623, | |
| "logits/rejected": -1.0042556524276733, | |
| "logps/chosen": -0.8588500022888184, | |
| "logps/rejected": -1.1039783954620361, | |
| "loss": 2.0002, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -8.588499069213867, | |
| "rewards/margins": 2.4512839317321777, | |
| "rewards/rejected": -11.039785385131836, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32064128256513025, | |
| "grad_norm": 70.30730129459549, | |
| "learning_rate": 8.588027776804058e-07, | |
| "logits/chosen": -1.060490369796753, | |
| "logits/rejected": -1.0404036045074463, | |
| "logps/chosen": -0.9423840641975403, | |
| "logps/rejected": -1.1874125003814697, | |
| "loss": 1.9455, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -9.42384147644043, | |
| "rewards/margins": 2.4502837657928467, | |
| "rewards/rejected": -11.874125480651855, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33132932531730125, | |
| "grad_norm": 81.54625041986957, | |
| "learning_rate": 8.455313244934324e-07, | |
| "logits/chosen": -1.0910407304763794, | |
| "logits/rejected": -1.0684020519256592, | |
| "logps/chosen": -0.9991434812545776, | |
| "logps/rejected": -1.3156726360321045, | |
| "loss": 2.0451, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -9.991434097290039, | |
| "rewards/margins": 3.165290355682373, | |
| "rewards/rejected": -13.15672492980957, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3420173680694723, | |
| "grad_norm": 78.5490421908409, | |
| "learning_rate": 8.317766145051057e-07, | |
| "logits/chosen": -1.109403371810913, | |
| "logits/rejected": -1.090001106262207, | |
| "logps/chosen": -1.1215949058532715, | |
| "logps/rejected": -1.5121821165084839, | |
| "loss": 1.9436, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -11.215949058532715, | |
| "rewards/margins": 3.9058711528778076, | |
| "rewards/rejected": -15.121821403503418, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3527054108216433, | |
| "grad_norm": 59.08371857927558, | |
| "learning_rate": 8.175578849210894e-07, | |
| "logits/chosen": -1.1232795715332031, | |
| "logits/rejected": -1.0980435609817505, | |
| "logps/chosen": -1.0903780460357666, | |
| "logps/rejected": -1.459205150604248, | |
| "loss": 1.8384, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -10.903780937194824, | |
| "rewards/margins": 3.688269853591919, | |
| "rewards/rejected": -14.59205150604248, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3633934535738143, | |
| "grad_norm": 85.71218468828272, | |
| "learning_rate": 8.028950219204099e-07, | |
| "logits/chosen": -1.1307361125946045, | |
| "logits/rejected": -1.1074953079223633, | |
| "logps/chosen": -1.0654685497283936, | |
| "logps/rejected": -1.4472792148590088, | |
| "loss": 1.7884, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -10.654685020446777, | |
| "rewards/margins": 3.818106174468994, | |
| "rewards/rejected": -14.47279167175293, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3740814963259853, | |
| "grad_norm": 92.85957749639208, | |
| "learning_rate": 7.878085328428368e-07, | |
| "logits/chosen": -1.1518357992172241, | |
| "logits/rejected": -1.102372407913208, | |
| "logps/chosen": -1.1460392475128174, | |
| "logps/rejected": -1.4155685901641846, | |
| "loss": 1.6771, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -11.460393905639648, | |
| "rewards/margins": 2.695291757583618, | |
| "rewards/rejected": -14.155685424804688, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3847695390781563, | |
| "grad_norm": 75.98858315922392, | |
| "learning_rate": 7.723195175075135e-07, | |
| "logits/chosen": -1.0996112823486328, | |
| "logits/rejected": -1.0788969993591309, | |
| "logps/chosen": -1.1098445653915405, | |
| "logps/rejected": -1.476881504058838, | |
| "loss": 1.6011, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -11.098443984985352, | |
| "rewards/margins": 3.6703686714172363, | |
| "rewards/rejected": -14.76881217956543, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3954575818303273, | |
| "grad_norm": 88.81502196023631, | |
| "learning_rate": 7.564496387029531e-07, | |
| "logits/chosen": -1.1378796100616455, | |
| "logits/rejected": -1.0828906297683716, | |
| "logps/chosen": -1.1474685668945312, | |
| "logps/rejected": -1.5796287059783936, | |
| "loss": 1.6663, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -11.474684715270996, | |
| "rewards/margins": 4.321602821350098, | |
| "rewards/rejected": -15.796287536621094, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4061456245824983, | |
| "grad_norm": 90.03714203036446, | |
| "learning_rate": 7.402210918896689e-07, | |
| "logits/chosen": -1.1278326511383057, | |
| "logits/rejected": -1.1358839273452759, | |
| "logps/chosen": -1.2729408740997314, | |
| "logps/rejected": -1.7558482885360718, | |
| "loss": 1.5442, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -12.729410171508789, | |
| "rewards/margins": 4.829073905944824, | |
| "rewards/rejected": -17.558483123779297, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4168336673346693, | |
| "grad_norm": 72.89600233357321, | |
| "learning_rate": 7.236565741578162e-07, | |
| "logits/chosen": -1.0958189964294434, | |
| "logits/rejected": -1.076554775238037, | |
| "logps/chosen": -1.2896816730499268, | |
| "logps/rejected": -1.6636635065078735, | |
| "loss": 1.6021, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -12.896817207336426, | |
| "rewards/margins": 3.7398200035095215, | |
| "rewards/rejected": -16.636634826660156, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42752171008684037, | |
| "grad_norm": 93.9340667463585, | |
| "learning_rate": 7.067792524832603e-07, | |
| "logits/chosen": -1.0816549062728882, | |
| "logits/rejected": -1.0706536769866943, | |
| "logps/chosen": -1.3197344541549683, | |
| "logps/rejected": -1.7450058460235596, | |
| "loss": 1.5092, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -13.197346687316895, | |
| "rewards/margins": 4.252710819244385, | |
| "rewards/rejected": -17.450056076049805, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43820975283901137, | |
| "grad_norm": 91.15403821743105, | |
| "learning_rate": 6.896127313264642e-07, | |
| "logits/chosen": -1.1295057535171509, | |
| "logits/rejected": -1.0786478519439697, | |
| "logps/chosen": -1.3944091796875, | |
| "logps/rejected": -1.8417927026748657, | |
| "loss": 1.7223, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -13.944093704223633, | |
| "rewards/margins": 4.473834037780762, | |
| "rewards/rejected": -18.417926788330078, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44889779559118237, | |
| "grad_norm": 99.88773415242756, | |
| "learning_rate": 6.721810196195174e-07, | |
| "logits/chosen": -1.1591789722442627, | |
| "logits/rejected": -1.147062063217163, | |
| "logps/chosen": -1.3990533351898193, | |
| "logps/rejected": -1.8112404346466064, | |
| "loss": 1.6082, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -13.990533828735352, | |
| "rewards/margins": 4.121870040893555, | |
| "rewards/rejected": -18.112403869628906, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45958583834335337, | |
| "grad_norm": 108.24791172325133, | |
| "learning_rate": 6.545084971874736e-07, | |
| "logits/chosen": -1.1198530197143555, | |
| "logits/rejected": -1.101109504699707, | |
| "logps/chosen": -1.390649437904358, | |
| "logps/rejected": -1.8630450963974, | |
| "loss": 1.4791, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -13.906494140625, | |
| "rewards/margins": 4.7239580154418945, | |
| "rewards/rejected": -18.630451202392578, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.47027388109552437, | |
| "grad_norm": 95.83911690989143, | |
| "learning_rate": 6.3661988065096e-07, | |
| "logits/chosen": -1.1780140399932861, | |
| "logits/rejected": -1.1579878330230713, | |
| "logps/chosen": -1.4568861722946167, | |
| "logps/rejected": -1.9470503330230713, | |
| "loss": 1.4586, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -14.56886100769043, | |
| "rewards/margins": 4.901640892028809, | |
| "rewards/rejected": -19.470502853393555, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48096192384769537, | |
| "grad_norm": 69.10471107204022, | |
| "learning_rate": 6.185401888577487e-07, | |
| "logits/chosen": -1.1474467515945435, | |
| "logits/rejected": -1.1124647855758667, | |
| "logps/chosen": -1.48002028465271, | |
| "logps/rejected": -1.9400886297225952, | |
| "loss": 1.4409, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -14.800203323364258, | |
| "rewards/margins": 4.600685119628906, | |
| "rewards/rejected": -19.400888442993164, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4916499665998664, | |
| "grad_norm": 80.62484140193865, | |
| "learning_rate": 6.002947078916364e-07, | |
| "logits/chosen": -1.2124546766281128, | |
| "logits/rejected": -1.161115050315857, | |
| "logps/chosen": -1.4423153400421143, | |
| "logps/rejected": -1.9036369323730469, | |
| "loss": 1.3817, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -14.423154830932617, | |
| "rewards/margins": 4.613214015960693, | |
| "rewards/rejected": -19.03636932373047, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5023380093520374, | |
| "grad_norm": 84.05912123531321, | |
| "learning_rate": 5.819089557075688e-07, | |
| "logits/chosen": -1.2311934232711792, | |
| "logits/rejected": -1.2020883560180664, | |
| "logps/chosen": -1.4844694137573242, | |
| "logps/rejected": -1.9821853637695312, | |
| "loss": 1.4172, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -14.844694137573242, | |
| "rewards/margins": 4.97715950012207, | |
| "rewards/rejected": -19.821855545043945, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5130260521042084, | |
| "grad_norm": 84.00316536161533, | |
| "learning_rate": 5.634086464424742e-07, | |
| "logits/chosen": -1.198540449142456, | |
| "logits/rejected": -1.1992590427398682, | |
| "logps/chosen": -1.3957428932189941, | |
| "logps/rejected": -1.8944737911224365, | |
| "loss": 1.4343, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -13.957429885864258, | |
| "rewards/margins": 4.987309455871582, | |
| "rewards/rejected": -18.944738388061523, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5237140948563794, | |
| "grad_norm": 137.49078119090206, | |
| "learning_rate": 5.448196544517167e-07, | |
| "logits/chosen": -1.2955886125564575, | |
| "logits/rejected": -1.23685622215271, | |
| "logps/chosen": -1.460442066192627, | |
| "logps/rejected": -2.0612359046936035, | |
| "loss": 1.3532, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -14.60442066192627, | |
| "rewards/margins": 6.007939338684082, | |
| "rewards/rejected": -20.612361907958984, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5344021376085505, | |
| "grad_norm": 129.54289500612722, | |
| "learning_rate": 5.26167978121472e-07, | |
| "logits/chosen": -1.2238231897354126, | |
| "logits/rejected": -1.2080833911895752, | |
| "logps/chosen": -1.5243932008743286, | |
| "logps/rejected": -2.1077561378479004, | |
| "loss": 1.3459, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.243929862976074, | |
| "rewards/margins": 5.833629608154297, | |
| "rewards/rejected": -21.077558517456055, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5450901803607214, | |
| "grad_norm": 102.89768684384153, | |
| "learning_rate": 5.074797035076318e-07, | |
| "logits/chosen": -1.267345905303955, | |
| "logits/rejected": -1.2413192987442017, | |
| "logps/chosen": -1.6330616474151611, | |
| "logps/rejected": -2.104926824569702, | |
| "loss": 1.4434, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -16.330615997314453, | |
| "rewards/margins": 4.718654155731201, | |
| "rewards/rejected": -21.049266815185547, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5557782231128925, | |
| "grad_norm": 89.20095630673174, | |
| "learning_rate": 4.887809678520975e-07, | |
| "logits/chosen": -1.245793104171753, | |
| "logits/rejected": -1.214970350265503, | |
| "logps/chosen": -1.554158091545105, | |
| "logps/rejected": -2.0427088737487793, | |
| "loss": 1.4276, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -15.541582107543945, | |
| "rewards/margins": 4.8855085372924805, | |
| "rewards/rejected": -20.42708969116211, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5664662658650634, | |
| "grad_norm": 76.055827552827, | |
| "learning_rate": 4.700979230274829e-07, | |
| "logits/chosen": -1.2102077007293701, | |
| "logits/rejected": -1.1913068294525146, | |
| "logps/chosen": -1.6448841094970703, | |
| "logps/rejected": -2.155822277069092, | |
| "loss": 1.3609, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -16.448841094970703, | |
| "rewards/margins": 5.109385967254639, | |
| "rewards/rejected": -21.5582275390625, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5771543086172345, | |
| "grad_norm": 129.8909118017969, | |
| "learning_rate": 4.514566989613559e-07, | |
| "logits/chosen": -1.2157796621322632, | |
| "logits/rejected": -1.186073899269104, | |
| "logps/chosen": -1.4407769441604614, | |
| "logps/rejected": -1.9774402379989624, | |
| "loss": 1.2996, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -14.407770156860352, | |
| "rewards/margins": 5.366633415222168, | |
| "rewards/rejected": -19.774402618408203, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5878423513694054, | |
| "grad_norm": 74.47995587471961, | |
| "learning_rate": 4.328833670911724e-07, | |
| "logits/chosen": -1.177504301071167, | |
| "logits/rejected": -1.1408427953720093, | |
| "logps/chosen": -1.4323005676269531, | |
| "logps/rejected": -1.8869625329971313, | |
| "loss": 1.4405, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -14.323003768920898, | |
| "rewards/margins": 4.546619892120361, | |
| "rewards/rejected": -18.869625091552734, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5985303941215765, | |
| "grad_norm": 96.7891504750656, | |
| "learning_rate": 4.144039039010124e-07, | |
| "logits/chosen": -1.262804627418518, | |
| "logits/rejected": -1.2378443479537964, | |
| "logps/chosen": -1.4722181558609009, | |
| "logps/rejected": -2.023758888244629, | |
| "loss": 1.3202, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -14.72218132019043, | |
| "rewards/margins": 5.515408515930176, | |
| "rewards/rejected": -20.23758888244629, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6092184368737475, | |
| "grad_norm": 98.51578082175142, | |
| "learning_rate": 3.960441545911204e-07, | |
| "logits/chosen": -1.2408018112182617, | |
| "logits/rejected": -1.2075875997543335, | |
| "logps/chosen": -1.5188751220703125, | |
| "logps/rejected": -2.0878236293792725, | |
| "loss": 1.0977, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -15.188751220703125, | |
| "rewards/margins": 5.689483642578125, | |
| "rewards/rejected": -20.878236770629883, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6199064796259185, | |
| "grad_norm": 77.81254701258105, | |
| "learning_rate": 3.778297969310529e-07, | |
| "logits/chosen": -1.2707680463790894, | |
| "logits/rejected": -1.2261282205581665, | |
| "logps/chosen": -1.5314843654632568, | |
| "logps/rejected": -2.0101191997528076, | |
| "loss": 1.3577, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -15.314845085144043, | |
| "rewards/margins": 4.786349296569824, | |
| "rewards/rejected": -20.101192474365234, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6305945223780896, | |
| "grad_norm": 87.64994632483507, | |
| "learning_rate": 3.5978630534699865e-07, | |
| "logits/chosen": -1.199864387512207, | |
| "logits/rejected": -1.1842243671417236, | |
| "logps/chosen": -1.545689344406128, | |
| "logps/rejected": -2.0575714111328125, | |
| "loss": 1.1919, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -15.456893920898438, | |
| "rewards/margins": 5.118819713592529, | |
| "rewards/rejected": -20.575714111328125, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6412825651302605, | |
| "grad_norm": 84.61392252215398, | |
| "learning_rate": 3.4193891529348795e-07, | |
| "logits/chosen": -1.1328258514404297, | |
| "logits/rejected": -1.1063092947006226, | |
| "logps/chosen": -1.631317138671875, | |
| "logps/rejected": -2.079132556915283, | |
| "loss": 1.6288, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -16.31317138671875, | |
| "rewards/margins": 4.478152275085449, | |
| "rewards/rejected": -20.791322708129883, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6519706078824316, | |
| "grad_norm": 84.46013666927763, | |
| "learning_rate": 3.243125879593286e-07, | |
| "logits/chosen": -1.2454413175582886, | |
| "logits/rejected": -1.1997601985931396, | |
| "logps/chosen": -1.6037687063217163, | |
| "logps/rejected": -2.0645315647125244, | |
| "loss": 1.3155, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -16.03768539428711, | |
| "rewards/margins": 4.6076273918151855, | |
| "rewards/rejected": -20.645313262939453, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6626586506346025, | |
| "grad_norm": 91.59670184758677, | |
| "learning_rate": 3.069319753571269e-07, | |
| "logits/chosen": -1.2738150358200073, | |
| "logits/rejected": -1.253278136253357, | |
| "logps/chosen": -1.6317838430404663, | |
| "logps/rejected": -2.138291835784912, | |
| "loss": 1.372, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -16.31783676147461, | |
| "rewards/margins": 5.065082550048828, | |
| "rewards/rejected": -21.382923126220703, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6733466933867736, | |
| "grad_norm": 87.02614481244046, | |
| "learning_rate": 2.898213858452173e-07, | |
| "logits/chosen": -1.284517526626587, | |
| "logits/rejected": -1.2254732847213745, | |
| "logps/chosen": -1.618208885192871, | |
| "logps/rejected": -2.1373062133789062, | |
| "loss": 1.3415, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -16.182092666625977, | |
| "rewards/margins": 5.190975189208984, | |
| "rewards/rejected": -21.373065948486328, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6840347361389446, | |
| "grad_norm": 104.27219685818618, | |
| "learning_rate": 2.730047501302266e-07, | |
| "logits/chosen": -1.2704033851623535, | |
| "logits/rejected": -1.2657862901687622, | |
| "logps/chosen": -1.6442874670028687, | |
| "logps/rejected": -2.2715744972229004, | |
| "loss": 1.2614, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -16.4428768157959, | |
| "rewards/margins": 6.2728681564331055, | |
| "rewards/rejected": -22.715742111206055, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6947227788911156, | |
| "grad_norm": 78.77445808060149, | |
| "learning_rate": 2.5650558779781635e-07, | |
| "logits/chosen": -1.2807691097259521, | |
| "logits/rejected": -1.2303869724273682, | |
| "logps/chosen": -1.6990268230438232, | |
| "logps/rejected": -2.368220329284668, | |
| "loss": 1.3078, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -16.99026870727539, | |
| "rewards/margins": 6.6919355392456055, | |
| "rewards/rejected": -23.682205200195312, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7054108216432866, | |
| "grad_norm": 70.04351714156043, | |
| "learning_rate": 2.403469744184154e-07, | |
| "logits/chosen": -1.1783530712127686, | |
| "logits/rejected": -1.136584758758545, | |
| "logps/chosen": -1.6521613597869873, | |
| "logps/rejected": -2.1305041313171387, | |
| "loss": 1.3592, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -16.5216121673584, | |
| "rewards/margins": 4.783430099487305, | |
| "rewards/rejected": -21.305042266845703, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7160988643954576, | |
| "grad_norm": 75.03379354143011, | |
| "learning_rate": 2.2455150927394878e-07, | |
| "logits/chosen": -1.2156535387039185, | |
| "logits/rejected": -1.1975212097167969, | |
| "logps/chosen": -1.6360639333724976, | |
| "logps/rejected": -2.187391757965088, | |
| "loss": 1.1952, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -16.360637664794922, | |
| "rewards/margins": 5.513278484344482, | |
| "rewards/rejected": -21.873918533325195, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7267869071476286, | |
| "grad_norm": 103.30210442360509, | |
| "learning_rate": 2.0914128375069722e-07, | |
| "logits/chosen": -1.2379086017608643, | |
| "logits/rejected": -1.2029554843902588, | |
| "logps/chosen": -1.5814708471298218, | |
| "logps/rejected": -2.1416497230529785, | |
| "loss": 1.3219, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -15.814706802368164, | |
| "rewards/margins": 5.601790428161621, | |
| "rewards/rejected": -21.41649627685547, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7374749498997996, | |
| "grad_norm": 87.41940209533863, | |
| "learning_rate": 1.9413785044249676e-07, | |
| "logits/chosen": -1.254070520401001, | |
| "logits/rejected": -1.2306808233261108, | |
| "logps/chosen": -1.665123701095581, | |
| "logps/rejected": -2.303457021713257, | |
| "loss": 1.3788, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -16.65123748779297, | |
| "rewards/margins": 6.3833327293396, | |
| "rewards/rejected": -23.034570693969727, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7481629926519706, | |
| "grad_norm": 147.74903637059256, | |
| "learning_rate": 1.7956219300748792e-07, | |
| "logits/chosen": -1.2324841022491455, | |
| "logits/rejected": -1.2356057167053223, | |
| "logps/chosen": -1.5469470024108887, | |
| "logps/rejected": -2.0821375846862793, | |
| "loss": 1.2883, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -15.469470024108887, | |
| "rewards/margins": 5.351906776428223, | |
| "rewards/rejected": -20.82137680053711, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7588510354041417, | |
| "grad_norm": 66.59688038674247, | |
| "learning_rate": 1.6543469682057104e-07, | |
| "logits/chosen": -1.1590429544448853, | |
| "logits/rejected": -1.1739274263381958, | |
| "logps/chosen": -1.5381479263305664, | |
| "logps/rejected": -2.0882415771484375, | |
| "loss": 1.181, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.381479263305664, | |
| "rewards/margins": 5.500934600830078, | |
| "rewards/rejected": -20.882413864135742, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7695390781563126, | |
| "grad_norm": 79.30848988409956, | |
| "learning_rate": 1.5177512046261666e-07, | |
| "logits/chosen": -1.2186603546142578, | |
| "logits/rejected": -1.2177612781524658, | |
| "logps/chosen": -1.5483216047286987, | |
| "logps/rejected": -2.190535306930542, | |
| "loss": 1.3007, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -15.48321533203125, | |
| "rewards/margins": 6.422137260437012, | |
| "rewards/rejected": -21.905353546142578, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7802271209084837, | |
| "grad_norm": 71.1995833686848, | |
| "learning_rate": 1.3860256808630427e-07, | |
| "logits/chosen": -1.2522964477539062, | |
| "logits/rejected": -1.1880736351013184, | |
| "logps/chosen": -1.599200963973999, | |
| "logps/rejected": -2.2274394035339355, | |
| "loss": 1.294, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -15.992010116577148, | |
| "rewards/margins": 6.28238582611084, | |
| "rewards/rejected": -22.274394989013672, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7909151636606546, | |
| "grad_norm": 97.88071644925103, | |
| "learning_rate": 1.2593546269723647e-07, | |
| "logits/chosen": -1.1726973056793213, | |
| "logits/rejected": -1.1615909337997437, | |
| "logps/chosen": -1.5849040746688843, | |
| "logps/rejected": -2.063690662384033, | |
| "loss": 1.2653, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.849041938781738, | |
| "rewards/margins": 4.787867546081543, | |
| "rewards/rejected": -20.63690948486328, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8016032064128257, | |
| "grad_norm": 93.22521656476633, | |
| "learning_rate": 1.1379152038770029e-07, | |
| "logits/chosen": -1.2195771932601929, | |
| "logits/rejected": -1.223771095275879, | |
| "logps/chosen": -1.7087455987930298, | |
| "logps/rejected": -2.2848830223083496, | |
| "loss": 1.2583, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -17.087453842163086, | |
| "rewards/margins": 5.76137638092041, | |
| "rewards/rejected": -22.848833084106445, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8122912491649966, | |
| "grad_norm": 122.65282224004734, | |
| "learning_rate": 1.0218772555910954e-07, | |
| "logits/chosen": -1.2245140075683594, | |
| "logits/rejected": -1.2064614295959473, | |
| "logps/chosen": -1.5752016305923462, | |
| "logps/rejected": -2.1021199226379395, | |
| "loss": 1.4127, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.752016067504883, | |
| "rewards/margins": 5.2691850662231445, | |
| "rewards/rejected": -21.021198272705078, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8229792919171677, | |
| "grad_norm": 79.05677313510314, | |
| "learning_rate": 9.114030716778432e-08, | |
| "logits/chosen": -1.2155396938323975, | |
| "logits/rejected": -1.194136142730713, | |
| "logps/chosen": -1.5979677438735962, | |
| "logps/rejected": -2.291325330734253, | |
| "loss": 1.0803, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -15.9796781539917, | |
| "rewards/margins": 6.9335784912109375, | |
| "rewards/rejected": -22.91325569152832, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8336673346693386, | |
| "grad_norm": 83.81957692142457, | |
| "learning_rate": 8.066471602728803e-08, | |
| "logits/chosen": -1.223331093788147, | |
| "logits/rejected": -1.209160327911377, | |
| "logps/chosen": -1.668593406677246, | |
| "logps/rejected": -2.259793519973755, | |
| "loss": 1.2346, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -16.68593406677246, | |
| "rewards/margins": 5.911999702453613, | |
| "rewards/rejected": -22.59793472290039, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8443553774215097, | |
| "grad_norm": 76.36054598990746, | |
| "learning_rate": 7.077560319906694e-08, | |
| "logits/chosen": -1.2313239574432373, | |
| "logits/rejected": -1.211395502090454, | |
| "logps/chosen": -1.5763094425201416, | |
| "logps/rejected": -2.1317121982574463, | |
| "loss": 1.2849, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -15.763093948364258, | |
| "rewards/margins": 5.554028511047363, | |
| "rewards/rejected": -21.317119598388672, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "grad_norm": 62.72495111163961, | |
| "learning_rate": 6.148679950161672e-08, | |
| "logits/chosen": -1.2403868436813354, | |
| "logits/rejected": -1.2243949174880981, | |
| "logps/chosen": -1.6198228597640991, | |
| "logps/rejected": -2.1331706047058105, | |
| "loss": 1.2445, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -16.19822883605957, | |
| "rewards/margins": 5.133477687835693, | |
| "rewards/rejected": -21.331707000732422, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "eval_logits/chosen": -1.4456316232681274, | |
| "eval_logits/rejected": -1.4547516107559204, | |
| "eval_logps/chosen": -1.623605728149414, | |
| "eval_logps/rejected": -2.176786422729492, | |
| "eval_loss": 1.3307912349700928, | |
| "eval_rewards/accuracies": 0.8353658318519592, | |
| "eval_rewards/chosen": -16.23605728149414, | |
| "eval_rewards/margins": 5.531808376312256, | |
| "eval_rewards/rejected": -21.767864227294922, | |
| "eval_runtime": 94.8719, | |
| "eval_samples_per_second": 20.67, | |
| "eval_steps_per_second": 1.296, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8657314629258517, | |
| "grad_norm": 99.36460974571031, | |
| "learning_rate": 5.2811296166831666e-08, | |
| "logits/chosen": -1.2008370161056519, | |
| "logits/rejected": -1.2194417715072632, | |
| "logps/chosen": -1.7075388431549072, | |
| "logps/rejected": -2.2549824714660645, | |
| "loss": 1.2048, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -17.075387954711914, | |
| "rewards/margins": 5.474437713623047, | |
| "rewards/rejected": -22.54982566833496, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8764195056780227, | |
| "grad_norm": 138.34780783301264, | |
| "learning_rate": 4.4761226670592066e-08, | |
| "logits/chosen": -1.216778039932251, | |
| "logits/rejected": -1.2035914659500122, | |
| "logps/chosen": -1.6350791454315186, | |
| "logps/rejected": -2.172778844833374, | |
| "loss": 1.3538, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -16.350793838500977, | |
| "rewards/margins": 5.376997947692871, | |
| "rewards/rejected": -21.727787017822266, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8871075484301937, | |
| "grad_norm": 72.98512679113071, | |
| "learning_rate": 3.734784976300165e-08, | |
| "logits/chosen": -1.2200865745544434, | |
| "logits/rejected": -1.1671762466430664, | |
| "logps/chosen": -1.5793800354003906, | |
| "logps/rejected": -2.232057571411133, | |
| "loss": 1.3478, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.793802261352539, | |
| "rewards/margins": 6.5267744064331055, | |
| "rewards/rejected": -22.32057762145996, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8977955911823647, | |
| "grad_norm": 98.07615613251582, | |
| "learning_rate": 3.058153372200695e-08, | |
| "logits/chosen": -1.2454715967178345, | |
| "logits/rejected": -1.195953607559204, | |
| "logps/chosen": -1.5281785726547241, | |
| "logps/rejected": -2.1256656646728516, | |
| "loss": 1.2056, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.28178596496582, | |
| "rewards/margins": 5.974873065948486, | |
| "rewards/rejected": -21.25665855407715, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9084836339345357, | |
| "grad_norm": 106.16711447135498, | |
| "learning_rate": 2.4471741852423233e-08, | |
| "logits/chosen": -1.2416163682937622, | |
| "logits/rejected": -1.231783390045166, | |
| "logps/chosen": -1.7100231647491455, | |
| "logps/rejected": -2.2363858222961426, | |
| "loss": 1.4487, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -17.100229263305664, | |
| "rewards/margins": 5.263625621795654, | |
| "rewards/rejected": -22.36385726928711, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9191716766867067, | |
| "grad_norm": 104.40358802274892, | |
| "learning_rate": 1.9027019250647036e-08, | |
| "logits/chosen": -1.2276403903961182, | |
| "logits/rejected": -1.211700439453125, | |
| "logps/chosen": -1.6992714405059814, | |
| "logps/rejected": -2.2929625511169434, | |
| "loss": 1.2603, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -16.992717742919922, | |
| "rewards/margins": 5.9369096755981445, | |
| "rewards/rejected": -22.92962646484375, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9298597194388778, | |
| "grad_norm": 76.71410236428167, | |
| "learning_rate": 1.4254980853566246e-08, | |
| "logits/chosen": -1.1829754114151, | |
| "logits/rejected": -1.1444637775421143, | |
| "logps/chosen": -1.5611233711242676, | |
| "logps/rejected": -2.1490211486816406, | |
| "loss": 1.2066, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -15.611233711242676, | |
| "rewards/margins": 5.8789777755737305, | |
| "rewards/rejected": -21.490211486816406, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9405477621910487, | |
| "grad_norm": 89.37849635838704, | |
| "learning_rate": 1.016230078838226e-08, | |
| "logits/chosen": -1.2062740325927734, | |
| "logits/rejected": -1.148478627204895, | |
| "logps/chosen": -1.6622101068496704, | |
| "logps/rejected": -2.1926727294921875, | |
| "loss": 1.253, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -16.622098922729492, | |
| "rewards/margins": 5.304628849029541, | |
| "rewards/rejected": -21.926727294921875, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9512358049432198, | |
| "grad_norm": 78.38263983661439, | |
| "learning_rate": 6.754703038239329e-09, | |
| "logits/chosen": -1.1661369800567627, | |
| "logits/rejected": -1.1493126153945923, | |
| "logps/chosen": -1.6586837768554688, | |
| "logps/rejected": -2.293992280960083, | |
| "loss": 1.0815, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -16.58683967590332, | |
| "rewards/margins": 6.353082180023193, | |
| "rewards/rejected": -22.939918518066406, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9619238476953907, | |
| "grad_norm": 85.39108439896182, | |
| "learning_rate": 4.036953436716895e-09, | |
| "logits/chosen": -1.2724522352218628, | |
| "logits/rejected": -1.2523143291473389, | |
| "logps/chosen": -1.607690453529358, | |
| "logps/rejected": -2.168273448944092, | |
| "loss": 1.3282, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -16.076906204223633, | |
| "rewards/margins": 5.605828285217285, | |
| "rewards/rejected": -21.682735443115234, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9726118904475618, | |
| "grad_norm": 102.09796631455698, | |
| "learning_rate": 2.0128530023804656e-09, | |
| "logits/chosen": -1.2248659133911133, | |
| "logits/rejected": -1.1913433074951172, | |
| "logps/chosen": -1.6112314462661743, | |
| "logps/rejected": -2.2650115489959717, | |
| "loss": 1.0445, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": -16.112314224243164, | |
| "rewards/margins": 6.537802696228027, | |
| "rewards/rejected": -22.650117874145508, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9832999331997327, | |
| "grad_norm": 90.8560495778911, | |
| "learning_rate": 6.852326227130833e-10, | |
| "logits/chosen": -1.2340514659881592, | |
| "logits/rejected": -1.2225282192230225, | |
| "logps/chosen": -1.6787292957305908, | |
| "logps/rejected": -2.2820496559143066, | |
| "loss": 1.1894, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -16.787290573120117, | |
| "rewards/margins": 6.033202648162842, | |
| "rewards/rejected": -22.82049560546875, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9939879759519038, | |
| "grad_norm": 86.47798441930765, | |
| "learning_rate": 5.594909486328348e-11, | |
| "logits/chosen": -1.2089763879776, | |
| "logits/rejected": -1.2146103382110596, | |
| "logps/chosen": -1.703181266784668, | |
| "logps/rejected": -2.318962335586548, | |
| "loss": 1.3219, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -17.03181266784668, | |
| "rewards/margins": 6.157810688018799, | |
| "rewards/rejected": -23.189624786376953, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9982631930527722, | |
| "step": 467, | |
| "total_flos": 0.0, | |
| "train_loss": 1.8032665589636858, | |
| "train_runtime": 11474.0462, | |
| "train_samples_per_second": 5.218, | |
| "train_steps_per_second": 0.041 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 467, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |