| { | |
| "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": 68.85501439409333, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -1.007319450378418, | |
| "logits/rejected": -0.9782959222793579, | |
| "logps/chosen": -0.2740143835544586, | |
| "logps/rejected": -0.27172547578811646, | |
| "loss": 3.0362, | |
| "rewards/accuracies": 0.4437499940395355, | |
| "rewards/chosen": -2.7401437759399414, | |
| "rewards/margins": -0.02288895845413208, | |
| "rewards/rejected": -2.717254638671875, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02137608550434202, | |
| "grad_norm": 43.462307413510835, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -1.0488415956497192, | |
| "logits/rejected": -0.9815706014633179, | |
| "logps/chosen": -0.29419511556625366, | |
| "logps/rejected": -0.2995131313800812, | |
| "loss": 3.1785, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -2.941951274871826, | |
| "rewards/margins": 0.05318016931414604, | |
| "rewards/rejected": -2.995131015777588, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03206412825651302, | |
| "grad_norm": 59.3077193791241, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -0.9633978605270386, | |
| "logits/rejected": -0.9824289083480835, | |
| "logps/chosen": -0.2639780044555664, | |
| "logps/rejected": -0.30077746510505676, | |
| "loss": 3.0982, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -2.639780044555664, | |
| "rewards/margins": 0.36799436807632446, | |
| "rewards/rejected": -3.0077743530273438, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04275217100868404, | |
| "grad_norm": 100.89613137154764, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -0.9631160497665405, | |
| "logits/rejected": -0.9371601343154907, | |
| "logps/chosen": -0.27738872170448303, | |
| "logps/rejected": -0.29155653715133667, | |
| "loss": 3.0229, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -2.7738869190216064, | |
| "rewards/margins": 0.1416785567998886, | |
| "rewards/rejected": -2.9155654907226562, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053440213760855046, | |
| "grad_norm": 58.79708288817627, | |
| "learning_rate": 5.319148936170212e-07, | |
| "logits/chosen": -1.013113260269165, | |
| "logits/rejected": -0.9839662313461304, | |
| "logps/chosen": -0.27208274602890015, | |
| "logps/rejected": -0.2786521315574646, | |
| "loss": 3.2744, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -2.720827341079712, | |
| "rewards/margins": 0.06569431722164154, | |
| "rewards/rejected": -2.7865219116210938, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06412825651302605, | |
| "grad_norm": 58.05968359177723, | |
| "learning_rate": 6.382978723404255e-07, | |
| "logits/chosen": -1.0021781921386719, | |
| "logits/rejected": -0.9577069282531738, | |
| "logps/chosen": -0.274009644985199, | |
| "logps/rejected": -0.2803052067756653, | |
| "loss": 3.0005, | |
| "rewards/accuracies": 0.42500001192092896, | |
| "rewards/chosen": -2.7400965690612793, | |
| "rewards/margins": 0.06295552104711533, | |
| "rewards/rejected": -2.8030521869659424, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07481629926519706, | |
| "grad_norm": 73.78592278395888, | |
| "learning_rate": 7.446808510638297e-07, | |
| "logits/chosen": -1.0400080680847168, | |
| "logits/rejected": -0.9649320840835571, | |
| "logps/chosen": -0.2968520522117615, | |
| "logps/rejected": -0.32594868540763855, | |
| "loss": 3.0023, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -2.968520164489746, | |
| "rewards/margins": 0.29096680879592896, | |
| "rewards/rejected": -3.259486675262451, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08550434201736808, | |
| "grad_norm": 56.93905382011994, | |
| "learning_rate": 8.51063829787234e-07, | |
| "logits/chosen": -0.9891254305839539, | |
| "logits/rejected": -0.9455845952033997, | |
| "logps/chosen": -0.282151460647583, | |
| "logps/rejected": -0.3367989659309387, | |
| "loss": 2.9458, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -2.82151460647583, | |
| "rewards/margins": 0.5464746952056885, | |
| "rewards/rejected": -3.3679893016815186, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09619238476953908, | |
| "grad_norm": 79.32682008333322, | |
| "learning_rate": 9.574468085106384e-07, | |
| "logits/chosen": -1.0468851327896118, | |
| "logits/rejected": -1.0032992362976074, | |
| "logps/chosen": -0.36789190769195557, | |
| "logps/rejected": -0.42787107825279236, | |
| "loss": 3.0263, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -3.6789193153381348, | |
| "rewards/margins": 0.5997918248176575, | |
| "rewards/rejected": -4.278710842132568, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10688042752171009, | |
| "grad_norm": 91.62537725815142, | |
| "learning_rate": 9.998741174712533e-07, | |
| "logits/chosen": -1.0219460725784302, | |
| "logits/rejected": -0.9719876050949097, | |
| "logps/chosen": -0.38097304105758667, | |
| "logps/rejected": -0.43333888053894043, | |
| "loss": 3.0722, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -3.8097305297851562, | |
| "rewards/margins": 0.5236578583717346, | |
| "rewards/rejected": -4.333388328552246, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11756847027388109, | |
| "grad_norm": 133.17377723672328, | |
| "learning_rate": 9.991050648838675e-07, | |
| "logits/chosen": -1.0433048009872437, | |
| "logits/rejected": -1.0081396102905273, | |
| "logps/chosen": -0.31340470910072327, | |
| "logps/rejected": -0.3743906617164612, | |
| "loss": 2.8041, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -3.134047031402588, | |
| "rewards/margins": 0.6098597645759583, | |
| "rewards/rejected": -3.7439064979553223, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1282565130260521, | |
| "grad_norm": 89.75230108032684, | |
| "learning_rate": 9.97637968732563e-07, | |
| "logits/chosen": -1.0939303636550903, | |
| "logits/rejected": -1.0607116222381592, | |
| "logps/chosen": -0.32906678318977356, | |
| "logps/rejected": -0.3560132384300232, | |
| "loss": 2.8477, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -3.290667772293091, | |
| "rewards/margins": 0.2694646716117859, | |
| "rewards/rejected": -3.5601325035095215, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13894455577822312, | |
| "grad_norm": 73.19084024449225, | |
| "learning_rate": 9.954748808839674e-07, | |
| "logits/chosen": -1.0200589895248413, | |
| "logits/rejected": -0.9912623167037964, | |
| "logps/chosen": -0.38689833879470825, | |
| "logps/rejected": -0.4406129717826843, | |
| "loss": 2.8473, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -3.868983030319214, | |
| "rewards/margins": 0.5371465682983398, | |
| "rewards/rejected": -4.406129360198975, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14963259853039412, | |
| "grad_norm": 46.36345887436274, | |
| "learning_rate": 9.926188266120295e-07, | |
| "logits/chosen": -1.0427110195159912, | |
| "logits/rejected": -1.017996907234192, | |
| "logps/chosen": -0.34935158491134644, | |
| "logps/rejected": -0.43221673369407654, | |
| "loss": 2.8927, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -3.493516206741333, | |
| "rewards/margins": 0.8286512494087219, | |
| "rewards/rejected": -4.32216739654541, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16032064128256512, | |
| "grad_norm": 86.0599966129396, | |
| "learning_rate": 9.890738003669027e-07, | |
| "logits/chosen": -0.9983121752738953, | |
| "logits/rejected": -0.9283790588378906, | |
| "logps/chosen": -0.3482792377471924, | |
| "logps/rejected": -0.41999250650405884, | |
| "loss": 2.7867, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -3.482792615890503, | |
| "rewards/margins": 0.7171324491500854, | |
| "rewards/rejected": -4.199924468994141, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17100868403473615, | |
| "grad_norm": 47.31795000764903, | |
| "learning_rate": 9.848447601883433e-07, | |
| "logits/chosen": -0.974830150604248, | |
| "logits/rejected": -0.9607669115066528, | |
| "logps/chosen": -0.3474265933036804, | |
| "logps/rejected": -0.45176610350608826, | |
| "loss": 2.8224, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -3.4742660522460938, | |
| "rewards/margins": 1.0433948040008545, | |
| "rewards/rejected": -4.517661094665527, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18169672678690715, | |
| "grad_norm": 73.04066752438544, | |
| "learning_rate": 9.799376207714444e-07, | |
| "logits/chosen": -0.9964420199394226, | |
| "logits/rejected": -0.9738152623176575, | |
| "logps/chosen": -0.3353765606880188, | |
| "logps/rejected": -0.4040314555168152, | |
| "loss": 2.7119, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -3.3537659645080566, | |
| "rewards/margins": 0.6865488886833191, | |
| "rewards/rejected": -4.040314674377441, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19238476953907815, | |
| "grad_norm": 76.692510819006, | |
| "learning_rate": 9.743592451943998e-07, | |
| "logits/chosen": -1.0402719974517822, | |
| "logits/rejected": -1.0053231716156006, | |
| "logps/chosen": -0.4286496043205261, | |
| "logps/rejected": -0.519912838935852, | |
| "loss": 2.88, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -4.286496162414551, | |
| "rewards/margins": 0.9126325845718384, | |
| "rewards/rejected": -5.1991286277771, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20307281229124916, | |
| "grad_norm": 46.68556671183106, | |
| "learning_rate": 9.681174353198686e-07, | |
| "logits/chosen": -1.123992681503296, | |
| "logits/rejected": -1.0409624576568604, | |
| "logps/chosen": -0.4465855658054352, | |
| "logps/rejected": -0.4793754518032074, | |
| "loss": 2.83, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -4.465855598449707, | |
| "rewards/margins": 0.3278988003730774, | |
| "rewards/rejected": -4.793754577636719, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21376085504342018, | |
| "grad_norm": 89.65351102525449, | |
| "learning_rate": 9.612209208833646e-07, | |
| "logits/chosen": -1.0020520687103271, | |
| "logits/rejected": -0.9762676358222961, | |
| "logps/chosen": -0.44748955965042114, | |
| "logps/rejected": -0.4820302128791809, | |
| "loss": 3.012, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -4.47489595413208, | |
| "rewards/margins": 0.34540656208992004, | |
| "rewards/rejected": -4.8203020095825195, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22444889779559118, | |
| "grad_norm": 74.62602523369812, | |
| "learning_rate": 9.536793472839324e-07, | |
| "logits/chosen": -1.025708794593811, | |
| "logits/rejected": -0.9719223976135254, | |
| "logps/chosen": -0.3887873888015747, | |
| "logps/rejected": -0.4932475686073303, | |
| "loss": 2.8263, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -3.887873888015747, | |
| "rewards/margins": 1.0446012020111084, | |
| "rewards/rejected": -4.9324750900268555, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23513694054776219, | |
| "grad_norm": 68.98493486526344, | |
| "learning_rate": 9.455032620941839e-07, | |
| "logits/chosen": -0.9832603335380554, | |
| "logits/rejected": -0.921442985534668, | |
| "logps/chosen": -0.43601909279823303, | |
| "logps/rejected": -0.5775080323219299, | |
| "loss": 2.7095, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -4.360190391540527, | |
| "rewards/margins": 1.4148895740509033, | |
| "rewards/rejected": -5.775080680847168, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2458249832999332, | |
| "grad_norm": 104.15189984616235, | |
| "learning_rate": 9.367041003085648e-07, | |
| "logits/chosen": -1.0532798767089844, | |
| "logits/rejected": -0.991862416267395, | |
| "logps/chosen": -0.47051066160202026, | |
| "logps/rejected": -0.5247665047645569, | |
| "loss": 2.7215, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -4.705106258392334, | |
| "rewards/margins": 0.5425585508346558, | |
| "rewards/rejected": -5.247664451599121, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2565130260521042, | |
| "grad_norm": 58.826951355159096, | |
| "learning_rate": 9.272941683504808e-07, | |
| "logits/chosen": -1.003598928451538, | |
| "logits/rejected": -0.9121431112289429, | |
| "logps/chosen": -0.47750264406204224, | |
| "logps/rejected": -0.6715680956840515, | |
| "loss": 2.5484, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -4.775026798248291, | |
| "rewards/margins": 1.9406547546386719, | |
| "rewards/rejected": -6.715681552886963, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26720106880427524, | |
| "grad_norm": 66.98580734997083, | |
| "learning_rate": 9.172866268606513e-07, | |
| "logits/chosen": -1.0721389055252075, | |
| "logits/rejected": -1.0289056301116943, | |
| "logps/chosen": -0.5379669666290283, | |
| "logps/rejected": -0.6219452023506165, | |
| "loss": 2.3715, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -5.379669189453125, | |
| "rewards/margins": 0.8397828340530396, | |
| "rewards/rejected": -6.219452381134033, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27788911155644624, | |
| "grad_norm": 84.77905548021701, | |
| "learning_rate": 9.066954722907638e-07, | |
| "logits/chosen": -1.100857138633728, | |
| "logits/rejected": -1.0902975797653198, | |
| "logps/chosen": -0.5108389258384705, | |
| "logps/rejected": -0.7690914869308472, | |
| "loss": 2.3622, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -5.108389377593994, | |
| "rewards/margins": 2.5825250148773193, | |
| "rewards/rejected": -7.690914154052734, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.28857715430861725, | |
| "grad_norm": 83.79514155786627, | |
| "learning_rate": 8.955355173281707e-07, | |
| "logits/chosen": -1.0757827758789062, | |
| "logits/rejected": -1.024726152420044, | |
| "logps/chosen": -0.5399268269538879, | |
| "logps/rejected": -0.6515442728996277, | |
| "loss": 2.4957, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -5.399269104003906, | |
| "rewards/margins": 1.1161742210388184, | |
| "rewards/rejected": -6.515442848205566, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29926519706078825, | |
| "grad_norm": 84.91769818698891, | |
| "learning_rate": 8.838223701790055e-07, | |
| "logits/chosen": -1.1459643840789795, | |
| "logits/rejected": -1.1201996803283691, | |
| "logps/chosen": -0.611223578453064, | |
| "logps/rejected": -0.7296082973480225, | |
| "loss": 2.4149, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -6.112236022949219, | |
| "rewards/margins": 1.183847188949585, | |
| "rewards/rejected": -7.296082973480225, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30995323981295925, | |
| "grad_norm": 102.01113634181742, | |
| "learning_rate": 8.71572412738697e-07, | |
| "logits/chosen": -1.0526931285858154, | |
| "logits/rejected": -1.0242080688476562, | |
| "logps/chosen": -0.6204456090927124, | |
| "logps/rejected": -0.8014122843742371, | |
| "loss": 2.176, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -6.204455375671387, | |
| "rewards/margins": 1.8096668720245361, | |
| "rewards/rejected": -8.01412296295166, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32064128256513025, | |
| "grad_norm": 74.15980695565004, | |
| "learning_rate": 8.588027776804058e-07, | |
| "logits/chosen": -1.093958854675293, | |
| "logits/rejected": -1.0704580545425415, | |
| "logps/chosen": -0.649553656578064, | |
| "logps/rejected": -0.8072296977043152, | |
| "loss": 2.2659, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -6.495536804199219, | |
| "rewards/margins": 1.5767608880996704, | |
| "rewards/rejected": -8.072297096252441, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33132932531730125, | |
| "grad_norm": 72.62800724311847, | |
| "learning_rate": 8.455313244934324e-07, | |
| "logits/chosen": -1.1008553504943848, | |
| "logits/rejected": -1.0772287845611572, | |
| "logps/chosen": -0.6936475038528442, | |
| "logps/rejected": -0.9287503957748413, | |
| "loss": 2.1291, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -6.9364752769470215, | |
| "rewards/margins": 2.3510289192199707, | |
| "rewards/rejected": -9.287505149841309, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3420173680694723, | |
| "grad_norm": 79.70729288803042, | |
| "learning_rate": 8.317766145051057e-07, | |
| "logits/chosen": -1.1146085262298584, | |
| "logits/rejected": -1.0959725379943848, | |
| "logps/chosen": -0.7885189056396484, | |
| "logps/rejected": -1.1322988271713257, | |
| "loss": 2.2489, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -7.885189056396484, | |
| "rewards/margins": 3.4377999305725098, | |
| "rewards/rejected": -11.322988510131836, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3527054108216433, | |
| "grad_norm": 69.7876072320057, | |
| "learning_rate": 8.175578849210894e-07, | |
| "logits/chosen": -1.1170759201049805, | |
| "logits/rejected": -1.0901795625686646, | |
| "logps/chosen": -0.8508628606796265, | |
| "logps/rejected": -1.1688997745513916, | |
| "loss": 2.149, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -8.508628845214844, | |
| "rewards/margins": 3.180368661880493, | |
| "rewards/rejected": -11.688997268676758, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3633934535738143, | |
| "grad_norm": 81.33338167428913, | |
| "learning_rate": 8.028950219204099e-07, | |
| "logits/chosen": -1.1123058795928955, | |
| "logits/rejected": -1.089231014251709, | |
| "logps/chosen": -0.8155210614204407, | |
| "logps/rejected": -1.1522643566131592, | |
| "loss": 1.9615, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -8.1552095413208, | |
| "rewards/margins": 3.367433547973633, | |
| "rewards/rejected": -11.52264404296875, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3740814963259853, | |
| "grad_norm": 148.27715900401293, | |
| "learning_rate": 7.878085328428368e-07, | |
| "logits/chosen": -1.1275430917739868, | |
| "logits/rejected": -1.0795894861221313, | |
| "logps/chosen": -0.9230238199234009, | |
| "logps/rejected": -1.1233540773391724, | |
| "loss": 1.8921, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -9.23023796081543, | |
| "rewards/margins": 2.0033037662506104, | |
| "rewards/rejected": -11.233541488647461, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3847695390781563, | |
| "grad_norm": 106.59570449253785, | |
| "learning_rate": 7.723195175075135e-07, | |
| "logits/chosen": -1.0845016241073608, | |
| "logits/rejected": -1.0633010864257812, | |
| "logps/chosen": -0.8883851170539856, | |
| "logps/rejected": -1.1803685426712036, | |
| "loss": 1.8854, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -8.883851051330566, | |
| "rewards/margins": 2.9198341369628906, | |
| "rewards/rejected": -11.803686141967773, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3954575818303273, | |
| "grad_norm": 121.9704404283126, | |
| "learning_rate": 7.564496387029531e-07, | |
| "logits/chosen": -1.1263529062271118, | |
| "logits/rejected": -1.0700018405914307, | |
| "logps/chosen": -0.9720249176025391, | |
| "logps/rejected": -1.2823493480682373, | |
| "loss": 1.8685, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -9.720248222351074, | |
| "rewards/margins": 3.1032447814941406, | |
| "rewards/rejected": -12.823492050170898, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4061456245824983, | |
| "grad_norm": 76.82499833409383, | |
| "learning_rate": 7.402210918896689e-07, | |
| "logits/chosen": -1.109298586845398, | |
| "logits/rejected": -1.1157281398773193, | |
| "logps/chosen": -1.0440528392791748, | |
| "logps/rejected": -1.4969879388809204, | |
| "loss": 1.726, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -10.440529823303223, | |
| "rewards/margins": 4.529348850250244, | |
| "rewards/rejected": -14.969879150390625, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4168336673346693, | |
| "grad_norm": 98.31254492588728, | |
| "learning_rate": 7.236565741578162e-07, | |
| "logits/chosen": -1.0706043243408203, | |
| "logits/rejected": -1.049782633781433, | |
| "logps/chosen": -1.0430076122283936, | |
| "logps/rejected": -1.383687973022461, | |
| "loss": 1.7355, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -10.430075645446777, | |
| "rewards/margins": 3.4068026542663574, | |
| "rewards/rejected": -13.836878776550293, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42752171008684037, | |
| "grad_norm": 121.71126629656227, | |
| "learning_rate": 7.067792524832603e-07, | |
| "logits/chosen": -1.0543218851089478, | |
| "logits/rejected": -1.0419845581054688, | |
| "logps/chosen": -1.0992997884750366, | |
| "logps/rejected": -1.4704090356826782, | |
| "loss": 1.8859, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -10.992999076843262, | |
| "rewards/margins": 3.711090564727783, | |
| "rewards/rejected": -14.704089164733887, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43820975283901137, | |
| "grad_norm": 106.07902118865707, | |
| "learning_rate": 6.896127313264642e-07, | |
| "logits/chosen": -1.1024644374847412, | |
| "logits/rejected": -1.052335500717163, | |
| "logps/chosen": -1.1648457050323486, | |
| "logps/rejected": -1.5202152729034424, | |
| "loss": 1.8443, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -11.648456573486328, | |
| "rewards/margins": 3.553696870803833, | |
| "rewards/rejected": -15.202153205871582, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44889779559118237, | |
| "grad_norm": 119.52331235635606, | |
| "learning_rate": 6.721810196195174e-07, | |
| "logits/chosen": -1.120891809463501, | |
| "logits/rejected": -1.1110631227493286, | |
| "logps/chosen": -1.2501453161239624, | |
| "logps/rejected": -1.6519298553466797, | |
| "loss": 1.7963, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -12.50145435333252, | |
| "rewards/margins": 4.017845153808594, | |
| "rewards/rejected": -16.519298553466797, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45958583834335337, | |
| "grad_norm": 98.00238843896823, | |
| "learning_rate": 6.545084971874736e-07, | |
| "logits/chosen": -1.063614845275879, | |
| "logits/rejected": -1.0467475652694702, | |
| "logps/chosen": -1.3223060369491577, | |
| "logps/rejected": -1.7558891773223877, | |
| "loss": 1.7131, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -13.223058700561523, | |
| "rewards/margins": 4.335831642150879, | |
| "rewards/rejected": -17.55889320373535, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.47027388109552437, | |
| "grad_norm": 114.86693788333515, | |
| "learning_rate": 6.3661988065096e-07, | |
| "logits/chosen": -1.1257785558700562, | |
| "logits/rejected": -1.1067885160446167, | |
| "logps/chosen": -1.3776687383651733, | |
| "logps/rejected": -1.8501077890396118, | |
| "loss": 1.7429, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -13.776687622070312, | |
| "rewards/margins": 4.724389553070068, | |
| "rewards/rejected": -18.50107765197754, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48096192384769537, | |
| "grad_norm": 129.99247004695826, | |
| "learning_rate": 6.185401888577487e-07, | |
| "logits/chosen": -1.105886697769165, | |
| "logits/rejected": -1.0728118419647217, | |
| "logps/chosen": -1.3728220462799072, | |
| "logps/rejected": -1.8042919635772705, | |
| "loss": 1.7141, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -13.728219985961914, | |
| "rewards/margins": 4.314699649810791, | |
| "rewards/rejected": -18.042917251586914, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4916499665998664, | |
| "grad_norm": 98.65669520922688, | |
| "learning_rate": 6.002947078916364e-07, | |
| "logits/chosen": -1.1883631944656372, | |
| "logits/rejected": -1.1372156143188477, | |
| "logps/chosen": -1.363006353378296, | |
| "logps/rejected": -1.7407630681991577, | |
| "loss": 1.6921, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -13.630064010620117, | |
| "rewards/margins": 3.7775673866271973, | |
| "rewards/rejected": -17.407630920410156, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5023380093520374, | |
| "grad_norm": 94.32058594396345, | |
| "learning_rate": 5.819089557075688e-07, | |
| "logits/chosen": -1.222760796546936, | |
| "logits/rejected": -1.1934791803359985, | |
| "logps/chosen": -1.3279025554656982, | |
| "logps/rejected": -1.7879940271377563, | |
| "loss": 1.5978, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -13.279026985168457, | |
| "rewards/margins": 4.600913047790527, | |
| "rewards/rejected": -17.879940032958984, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5130260521042084, | |
| "grad_norm": 97.68263555717222, | |
| "learning_rate": 5.634086464424742e-07, | |
| "logits/chosen": -1.2020962238311768, | |
| "logits/rejected": -1.2033917903900146, | |
| "logps/chosen": -1.2446677684783936, | |
| "logps/rejected": -1.7023578882217407, | |
| "loss": 1.6344, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -12.446677207946777, | |
| "rewards/margins": 4.576901435852051, | |
| "rewards/rejected": -17.02358055114746, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5237140948563794, | |
| "grad_norm": 202.05423816354624, | |
| "learning_rate": 5.448196544517167e-07, | |
| "logits/chosen": -1.2868934869766235, | |
| "logits/rejected": -1.225835919380188, | |
| "logps/chosen": -1.2727655172348022, | |
| "logps/rejected": -1.7838712930679321, | |
| "loss": 1.4597, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -12.727655410766602, | |
| "rewards/margins": 5.111058235168457, | |
| "rewards/rejected": -17.838712692260742, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5344021376085505, | |
| "grad_norm": 120.34400804479175, | |
| "learning_rate": 5.26167978121472e-07, | |
| "logits/chosen": -1.2153074741363525, | |
| "logits/rejected": -1.1993039846420288, | |
| "logps/chosen": -1.342670202255249, | |
| "logps/rejected": -1.8750245571136475, | |
| "loss": 1.5212, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -13.426701545715332, | |
| "rewards/margins": 5.323545932769775, | |
| "rewards/rejected": -18.750247955322266, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5450901803607214, | |
| "grad_norm": 95.36132768916778, | |
| "learning_rate": 5.074797035076318e-07, | |
| "logits/chosen": -1.265678882598877, | |
| "logits/rejected": -1.2382146120071411, | |
| "logps/chosen": -1.475931167602539, | |
| "logps/rejected": -1.921068549156189, | |
| "loss": 1.6491, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -14.759312629699707, | |
| "rewards/margins": 4.4513750076293945, | |
| "rewards/rejected": -19.2106876373291, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5557782231128925, | |
| "grad_norm": 117.86214570617697, | |
| "learning_rate": 4.887809678520975e-07, | |
| "logits/chosen": -1.2461557388305664, | |
| "logits/rejected": -1.2149869203567505, | |
| "logps/chosen": -1.3435603380203247, | |
| "logps/rejected": -1.779606580734253, | |
| "loss": 1.5681, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -13.435603141784668, | |
| "rewards/margins": 4.360462188720703, | |
| "rewards/rejected": -17.796064376831055, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5664662658650634, | |
| "grad_norm": 91.7079115328607, | |
| "learning_rate": 4.700979230274829e-07, | |
| "logits/chosen": -1.2090892791748047, | |
| "logits/rejected": -1.1903409957885742, | |
| "logps/chosen": -1.4390499591827393, | |
| "logps/rejected": -1.8992010354995728, | |
| "loss": 1.6488, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -14.390498161315918, | |
| "rewards/margins": 4.6015119552612305, | |
| "rewards/rejected": -18.99201011657715, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5771543086172345, | |
| "grad_norm": 220.87890749833002, | |
| "learning_rate": 4.514566989613559e-07, | |
| "logits/chosen": -1.207067847251892, | |
| "logits/rejected": -1.1773267984390259, | |
| "logps/chosen": -1.3406851291656494, | |
| "logps/rejected": -1.8256925344467163, | |
| "loss": 1.6271, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -13.406852722167969, | |
| "rewards/margins": 4.850072860717773, | |
| "rewards/rejected": -18.25692367553711, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5878423513694054, | |
| "grad_norm": 87.3849854148476, | |
| "learning_rate": 4.328833670911724e-07, | |
| "logits/chosen": -1.1608107089996338, | |
| "logits/rejected": -1.1226792335510254, | |
| "logps/chosen": -1.3398956060409546, | |
| "logps/rejected": -1.7405239343643188, | |
| "loss": 1.7113, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -13.398956298828125, | |
| "rewards/margins": 4.006283760070801, | |
| "rewards/rejected": -17.40523910522461, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5985303941215765, | |
| "grad_norm": 148.2686835393089, | |
| "learning_rate": 4.144039039010124e-07, | |
| "logits/chosen": -1.2560105323791504, | |
| "logits/rejected": -1.2310097217559814, | |
| "logps/chosen": -1.3750091791152954, | |
| "logps/rejected": -1.8952747583389282, | |
| "loss": 1.5337, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -13.750091552734375, | |
| "rewards/margins": 5.202655792236328, | |
| "rewards/rejected": -18.952749252319336, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6092184368737475, | |
| "grad_norm": 98.69515577428142, | |
| "learning_rate": 3.960441545911204e-07, | |
| "logits/chosen": -1.224327564239502, | |
| "logits/rejected": -1.1907469034194946, | |
| "logps/chosen": -1.431443691253662, | |
| "logps/rejected": -1.9771114587783813, | |
| "loss": 1.3013, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -14.314435958862305, | |
| "rewards/margins": 5.4566779136657715, | |
| "rewards/rejected": -19.7711124420166, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6199064796259185, | |
| "grad_norm": 141.5275248058356, | |
| "learning_rate": 3.778297969310529e-07, | |
| "logits/chosen": -1.257261872291565, | |
| "logits/rejected": -1.213803768157959, | |
| "logps/chosen": -1.40240478515625, | |
| "logps/rejected": -1.8522781133651733, | |
| "loss": 1.5508, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -14.0240478515625, | |
| "rewards/margins": 4.498733997344971, | |
| "rewards/rejected": -18.52277946472168, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6305945223780896, | |
| "grad_norm": 114.51767800087, | |
| "learning_rate": 3.5978630534699865e-07, | |
| "logits/chosen": -1.2009317874908447, | |
| "logits/rejected": -1.1843944787979126, | |
| "logps/chosen": -1.482761025428772, | |
| "logps/rejected": -1.9243135452270508, | |
| "loss": 1.4133, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -14.827609062194824, | |
| "rewards/margins": 4.415524482727051, | |
| "rewards/rejected": -19.243135452270508, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6412825651302605, | |
| "grad_norm": 133.03386433388292, | |
| "learning_rate": 3.4193891529348795e-07, | |
| "logits/chosen": -1.1327836513519287, | |
| "logits/rejected": -1.1056478023529053, | |
| "logps/chosen": -1.485954999923706, | |
| "logps/rejected": -1.863865852355957, | |
| "loss": 1.8363, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -14.859548568725586, | |
| "rewards/margins": 3.779109239578247, | |
| "rewards/rejected": -18.638660430908203, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6519706078824316, | |
| "grad_norm": 91.68870349271927, | |
| "learning_rate": 3.243125879593286e-07, | |
| "logits/chosen": -1.239027738571167, | |
| "logits/rejected": -1.1926300525665283, | |
| "logps/chosen": -1.4515793323516846, | |
| "logps/rejected": -1.8309190273284912, | |
| "loss": 1.7083, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -14.515792846679688, | |
| "rewards/margins": 3.7933971881866455, | |
| "rewards/rejected": -18.30919075012207, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6626586506346025, | |
| "grad_norm": 135.3549980063555, | |
| "learning_rate": 3.069319753571269e-07, | |
| "logits/chosen": -1.2658618688583374, | |
| "logits/rejected": -1.2451133728027344, | |
| "logps/chosen": -1.528917670249939, | |
| "logps/rejected": -1.9810943603515625, | |
| "loss": 1.6482, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -15.289176940917969, | |
| "rewards/margins": 4.521766662597656, | |
| "rewards/rejected": -19.810943603515625, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6733466933867736, | |
| "grad_norm": 103.62793682266408, | |
| "learning_rate": 2.898213858452173e-07, | |
| "logits/chosen": -1.2625854015350342, | |
| "logits/rejected": -1.2041596174240112, | |
| "logps/chosen": -1.4363172054290771, | |
| "logps/rejected": -1.9117801189422607, | |
| "loss": 1.5096, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -14.363171577453613, | |
| "rewards/margins": 4.754631042480469, | |
| "rewards/rejected": -19.117801666259766, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6840347361389446, | |
| "grad_norm": 124.93488530926095, | |
| "learning_rate": 2.730047501302266e-07, | |
| "logits/chosen": -1.2458070516586304, | |
| "logits/rejected": -1.2413748502731323, | |
| "logps/chosen": -1.4891000986099243, | |
| "logps/rejected": -2.0556411743164062, | |
| "loss": 1.431, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -14.891000747680664, | |
| "rewards/margins": 5.665411472320557, | |
| "rewards/rejected": -20.556411743164062, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6947227788911156, | |
| "grad_norm": 132.66286552485195, | |
| "learning_rate": 2.5650558779781635e-07, | |
| "logits/chosen": -1.262664556503296, | |
| "logits/rejected": -1.211101770401001, | |
| "logps/chosen": -1.6323843002319336, | |
| "logps/rejected": -2.2400825023651123, | |
| "loss": 1.4653, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -16.323843002319336, | |
| "rewards/margins": 6.076982498168945, | |
| "rewards/rejected": -22.40082550048828, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7054108216432866, | |
| "grad_norm": 77.86519610350517, | |
| "learning_rate": 2.403469744184154e-07, | |
| "logits/chosen": -1.1737123727798462, | |
| "logits/rejected": -1.1300561428070068, | |
| "logps/chosen": -1.5188909769058228, | |
| "logps/rejected": -1.950216293334961, | |
| "loss": 1.5313, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -15.188909530639648, | |
| "rewards/margins": 4.313254356384277, | |
| "rewards/rejected": -19.50216293334961, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7160988643954576, | |
| "grad_norm": 116.83352825706731, | |
| "learning_rate": 2.2455150927394878e-07, | |
| "logits/chosen": -1.202048420906067, | |
| "logits/rejected": -1.1833049058914185, | |
| "logps/chosen": -1.4710850715637207, | |
| "logps/rejected": -1.9782819747924805, | |
| "loss": 1.3351, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -14.710851669311523, | |
| "rewards/margins": 5.071969032287598, | |
| "rewards/rejected": -19.782819747924805, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7267869071476286, | |
| "grad_norm": 117.85874090140514, | |
| "learning_rate": 2.0914128375069722e-07, | |
| "logits/chosen": -1.2336275577545166, | |
| "logits/rejected": -1.1968821287155151, | |
| "logps/chosen": -1.4882738590240479, | |
| "logps/rejected": -2.008549928665161, | |
| "loss": 1.5616, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -14.88273811340332, | |
| "rewards/margins": 5.202761650085449, | |
| "rewards/rejected": -20.085498809814453, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7374749498997996, | |
| "grad_norm": 111.48616582145915, | |
| "learning_rate": 1.9413785044249676e-07, | |
| "logits/chosen": -1.2667595148086548, | |
| "logits/rejected": -1.2406939268112183, | |
| "logps/chosen": -1.5598745346069336, | |
| "logps/rejected": -2.16188383102417, | |
| "loss": 1.5993, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.59874439239502, | |
| "rewards/margins": 6.020094394683838, | |
| "rewards/rejected": -21.618839263916016, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7481629926519706, | |
| "grad_norm": 133.44639116296503, | |
| "learning_rate": 1.7956219300748792e-07, | |
| "logits/chosen": -1.244403600692749, | |
| "logits/rejected": -1.24759042263031, | |
| "logps/chosen": -1.5018080472946167, | |
| "logps/rejected": -2.005552291870117, | |
| "loss": 1.5981, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -15.01807975769043, | |
| "rewards/margins": 5.037442207336426, | |
| "rewards/rejected": -20.055522918701172, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7588510354041417, | |
| "grad_norm": 145.88647972080093, | |
| "learning_rate": 1.6543469682057104e-07, | |
| "logits/chosen": -1.1816396713256836, | |
| "logits/rejected": -1.195244550704956, | |
| "logps/chosen": -1.4624321460723877, | |
| "logps/rejected": -1.9389760494232178, | |
| "loss": 1.3978, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -14.624320983886719, | |
| "rewards/margins": 4.765438079833984, | |
| "rewards/rejected": -19.389759063720703, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7695390781563126, | |
| "grad_norm": 89.39232956423915, | |
| "learning_rate": 1.5177512046261666e-07, | |
| "logits/chosen": -1.247115135192871, | |
| "logits/rejected": -1.2440454959869385, | |
| "logps/chosen": -1.4620821475982666, | |
| "logps/rejected": -2.070996046066284, | |
| "loss": 1.464, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -14.620820999145508, | |
| "rewards/margins": 6.08914041519165, | |
| "rewards/rejected": -20.7099609375, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7802271209084837, | |
| "grad_norm": 97.94137628606371, | |
| "learning_rate": 1.3860256808630427e-07, | |
| "logits/chosen": -1.2937978506088257, | |
| "logits/rejected": -1.2186272144317627, | |
| "logps/chosen": -1.5094481706619263, | |
| "logps/rejected": -2.0971579551696777, | |
| "loss": 1.4916, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -15.094482421875, | |
| "rewards/margins": 5.87709903717041, | |
| "rewards/rejected": -20.97157859802246, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7909151636606546, | |
| "grad_norm": 122.16117083194588, | |
| "learning_rate": 1.2593546269723647e-07, | |
| "logits/chosen": -1.2087019681930542, | |
| "logits/rejected": -1.195441484451294, | |
| "logps/chosen": -1.4670366048812866, | |
| "logps/rejected": -1.9255189895629883, | |
| "loss": 1.5354, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -14.670367240905762, | |
| "rewards/margins": 4.5848236083984375, | |
| "rewards/rejected": -19.255189895629883, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8016032064128257, | |
| "grad_norm": 118.93381227295681, | |
| "learning_rate": 1.1379152038770029e-07, | |
| "logits/chosen": -1.241072416305542, | |
| "logits/rejected": -1.2429205179214478, | |
| "logps/chosen": -1.5817298889160156, | |
| "logps/rejected": -2.1309893131256104, | |
| "loss": 1.5269, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -15.817300796508789, | |
| "rewards/margins": 5.492591381072998, | |
| "rewards/rejected": -21.309892654418945, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8122912491649966, | |
| "grad_norm": 149.16584822794186, | |
| "learning_rate": 1.0218772555910954e-07, | |
| "logits/chosen": -1.2636191844940186, | |
| "logits/rejected": -1.2404122352600098, | |
| "logps/chosen": -1.4529675245285034, | |
| "logps/rejected": -1.929680585861206, | |
| "loss": 1.5703, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -14.52967643737793, | |
| "rewards/margins": 4.767130374908447, | |
| "rewards/rejected": -19.29680824279785, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8229792919171677, | |
| "grad_norm": 134.81427667533356, | |
| "learning_rate": 9.114030716778432e-08, | |
| "logits/chosen": -1.2461379766464233, | |
| "logits/rejected": -1.2212632894515991, | |
| "logps/chosen": -1.488556146621704, | |
| "logps/rejected": -2.128378391265869, | |
| "loss": 1.3302, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -14.8855619430542, | |
| "rewards/margins": 6.398221015930176, | |
| "rewards/rejected": -21.283782958984375, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8336673346693386, | |
| "grad_norm": 106.31823929279197, | |
| "learning_rate": 8.066471602728803e-08, | |
| "logits/chosen": -1.262342095375061, | |
| "logits/rejected": -1.2427619695663452, | |
| "logps/chosen": -1.546651840209961, | |
| "logps/rejected": -2.105900526046753, | |
| "loss": 1.5107, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -15.466519355773926, | |
| "rewards/margins": 5.5924882888793945, | |
| "rewards/rejected": -21.05900764465332, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8443553774215097, | |
| "grad_norm": 82.49774941818607, | |
| "learning_rate": 7.077560319906694e-08, | |
| "logits/chosen": -1.2678358554840088, | |
| "logits/rejected": -1.2441434860229492, | |
| "logps/chosen": -1.4695792198181152, | |
| "logps/rejected": -2.003164768218994, | |
| "loss": 1.4509, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -14.695793151855469, | |
| "rewards/margins": 5.335854530334473, | |
| "rewards/rejected": -20.031646728515625, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "grad_norm": 117.67752136496273, | |
| "learning_rate": 6.148679950161672e-08, | |
| "logits/chosen": -1.2668910026550293, | |
| "logits/rejected": -1.2486127614974976, | |
| "logps/chosen": -1.4717615842819214, | |
| "logps/rejected": -1.9521501064300537, | |
| "loss": 1.4439, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -14.717615127563477, | |
| "rewards/margins": 4.803887367248535, | |
| "rewards/rejected": -19.521503448486328, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "eval_logits/chosen": -1.484498143196106, | |
| "eval_logits/rejected": -1.4926748275756836, | |
| "eval_logps/chosen": -1.490028738975525, | |
| "eval_logps/rejected": -2.012206554412842, | |
| "eval_loss": 1.3685932159423828, | |
| "eval_rewards/accuracies": 0.8353658318519592, | |
| "eval_rewards/chosen": -14.900286674499512, | |
| "eval_rewards/margins": 5.221778392791748, | |
| "eval_rewards/rejected": -20.122066497802734, | |
| "eval_runtime": 95.1291, | |
| "eval_samples_per_second": 20.614, | |
| "eval_steps_per_second": 1.293, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8657314629258517, | |
| "grad_norm": 120.99808262565784, | |
| "learning_rate": 5.2811296166831666e-08, | |
| "logits/chosen": -1.2351343631744385, | |
| "logits/rejected": -1.2513123750686646, | |
| "logps/chosen": -1.5402753353118896, | |
| "logps/rejected": -1.9941785335540771, | |
| "loss": 1.4541, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.402752876281738, | |
| "rewards/margins": 4.539034366607666, | |
| "rewards/rejected": -19.94178581237793, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8764195056780227, | |
| "grad_norm": 184.47321309186552, | |
| "learning_rate": 4.4761226670592066e-08, | |
| "logits/chosen": -1.247280478477478, | |
| "logits/rejected": -1.234876036643982, | |
| "logps/chosen": -1.4927871227264404, | |
| "logps/rejected": -1.9950872659683228, | |
| "loss": 1.6499, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -14.927871704101562, | |
| "rewards/margins": 5.023002624511719, | |
| "rewards/rejected": -19.95087242126465, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8871075484301937, | |
| "grad_norm": 94.29238243376733, | |
| "learning_rate": 3.734784976300165e-08, | |
| "logits/chosen": -1.2503280639648438, | |
| "logits/rejected": -1.1914690732955933, | |
| "logps/chosen": -1.4084547758102417, | |
| "logps/rejected": -1.9929052591323853, | |
| "loss": 1.6712, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -14.084548950195312, | |
| "rewards/margins": 5.844505310058594, | |
| "rewards/rejected": -19.929052352905273, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8977955911823647, | |
| "grad_norm": 129.92639778174433, | |
| "learning_rate": 3.058153372200695e-08, | |
| "logits/chosen": -1.280215859413147, | |
| "logits/rejected": -1.2256780862808228, | |
| "logps/chosen": -1.3968563079833984, | |
| "logps/rejected": -1.92953622341156, | |
| "loss": 1.514, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -13.968562126159668, | |
| "rewards/margins": 5.326799392700195, | |
| "rewards/rejected": -19.295360565185547, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9084836339345357, | |
| "grad_norm": 134.3230870539629, | |
| "learning_rate": 2.4471741852423233e-08, | |
| "logits/chosen": -1.282026767730713, | |
| "logits/rejected": -1.2692543268203735, | |
| "logps/chosen": -1.5369470119476318, | |
| "logps/rejected": -2.0062360763549805, | |
| "loss": 1.6752, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -15.369470596313477, | |
| "rewards/margins": 4.6928911209106445, | |
| "rewards/rejected": -20.062358856201172, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9191716766867067, | |
| "grad_norm": 234.26732939024393, | |
| "learning_rate": 1.9027019250647036e-08, | |
| "logits/chosen": -1.2548713684082031, | |
| "logits/rejected": -1.235939383506775, | |
| "logps/chosen": -1.565554141998291, | |
| "logps/rejected": -2.0692152976989746, | |
| "loss": 1.597, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -15.655542373657227, | |
| "rewards/margins": 5.036610126495361, | |
| "rewards/rejected": -20.692150115966797, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9298597194388778, | |
| "grad_norm": 86.1878830813907, | |
| "learning_rate": 1.4254980853566246e-08, | |
| "logits/chosen": -1.2161545753479004, | |
| "logits/rejected": -1.1717766523361206, | |
| "logps/chosen": -1.4371172189712524, | |
| "logps/rejected": -1.9690968990325928, | |
| "loss": 1.4224, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -14.371172904968262, | |
| "rewards/margins": 5.319798469543457, | |
| "rewards/rejected": -19.69097328186035, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9405477621910487, | |
| "grad_norm": 124.03718973399216, | |
| "learning_rate": 1.016230078838226e-08, | |
| "logits/chosen": -1.2341668605804443, | |
| "logits/rejected": -1.1728591918945312, | |
| "logps/chosen": -1.4924213886260986, | |
| "logps/rejected": -1.9679025411605835, | |
| "loss": 1.4177, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -14.924214363098145, | |
| "rewards/margins": 4.754812240600586, | |
| "rewards/rejected": -19.679025650024414, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9512358049432198, | |
| "grad_norm": 94.74857172171315, | |
| "learning_rate": 6.754703038239329e-09, | |
| "logits/chosen": -1.1919021606445312, | |
| "logits/rejected": -1.1725019216537476, | |
| "logps/chosen": -1.5140306949615479, | |
| "logps/rejected": -2.1036159992218018, | |
| "loss": 1.2401, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -15.140307426452637, | |
| "rewards/margins": 5.895852088928223, | |
| "rewards/rejected": -21.03615951538086, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9619238476953907, | |
| "grad_norm": 125.47450644355813, | |
| "learning_rate": 4.036953436716895e-09, | |
| "logits/chosen": -1.3033939599990845, | |
| "logits/rejected": -1.2800980806350708, | |
| "logps/chosen": -1.4365876913070679, | |
| "logps/rejected": -1.9429874420166016, | |
| "loss": 1.4917, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -14.365875244140625, | |
| "rewards/margins": 5.063995361328125, | |
| "rewards/rejected": -19.42987060546875, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9726118904475618, | |
| "grad_norm": 107.55948121005738, | |
| "learning_rate": 2.0128530023804656e-09, | |
| "logits/chosen": -1.2646621465682983, | |
| "logits/rejected": -1.2269269227981567, | |
| "logps/chosen": -1.468353271484375, | |
| "logps/rejected": -2.0639400482177734, | |
| "loss": 1.1516, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -14.68353271484375, | |
| "rewards/margins": 5.955868244171143, | |
| "rewards/rejected": -20.639400482177734, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9832999331997327, | |
| "grad_norm": 93.15728602919113, | |
| "learning_rate": 6.852326227130833e-10, | |
| "logits/chosen": -1.2586472034454346, | |
| "logits/rejected": -1.2455804347991943, | |
| "logps/chosen": -1.5446629524230957, | |
| "logps/rejected": -2.104782819747925, | |
| "loss": 1.4082, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -15.446629524230957, | |
| "rewards/margins": 5.601197719573975, | |
| "rewards/rejected": -21.047826766967773, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9939879759519038, | |
| "grad_norm": 101.92365413278564, | |
| "learning_rate": 5.594909486328348e-11, | |
| "logits/chosen": -1.2412211894989014, | |
| "logits/rejected": -1.244315505027771, | |
| "logps/chosen": -1.5213388204574585, | |
| "logps/rejected": -2.075791835784912, | |
| "loss": 1.5253, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -15.213388442993164, | |
| "rewards/margins": 5.544531345367432, | |
| "rewards/rejected": -20.757919311523438, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9982631930527722, | |
| "step": 467, | |
| "total_flos": 0.0, | |
| "train_loss": 1.9910774169693157, | |
| "train_runtime": 11451.8288, | |
| "train_samples_per_second": 5.229, | |
| "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 | |
| } | |