| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.998691442030882, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010468463752944255, | |
| "grad_norm": 4.346810230562558, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": -2.770826816558838, | |
| "logits/rejected": -2.7283427715301514, | |
| "logps/chosen": -1.0278762578964233, | |
| "logps/rejected": -1.1730964183807373, | |
| "loss": -0.0311, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.0278762578964233, | |
| "rewards/margins": 0.14522016048431396, | |
| "rewards/rejected": -1.1730964183807373, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 3.9588019875332385, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.75107741355896, | |
| "logits/rejected": -2.7253947257995605, | |
| "logps/chosen": -0.9945151209831238, | |
| "logps/rejected": -1.0702844858169556, | |
| "loss": -0.0261, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.9945151209831238, | |
| "rewards/margins": 0.07576939463615417, | |
| "rewards/rejected": -1.0702844858169556, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 5.603537593720323, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": -2.73160982131958, | |
| "logits/rejected": -2.6982245445251465, | |
| "logps/chosen": -0.9771434664726257, | |
| "logps/rejected": -1.1377661228179932, | |
| "loss": -0.0428, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.9771434664726257, | |
| "rewards/margins": 0.16062253713607788, | |
| "rewards/rejected": -1.1377661228179932, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 4.548032560740917, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.8064258098602295, | |
| "logits/rejected": -2.713654041290283, | |
| "logps/chosen": -0.985003650188446, | |
| "logps/rejected": -1.0847834348678589, | |
| "loss": -0.0332, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.985003650188446, | |
| "rewards/margins": 0.09977984428405762, | |
| "rewards/rejected": -1.0847834348678589, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 4.5597803893373, | |
| "learning_rate": 2.604166666666667e-07, | |
| "logits/chosen": -2.8061375617980957, | |
| "logits/rejected": -2.783907890319824, | |
| "logps/chosen": -0.9467870593070984, | |
| "logps/rejected": -1.0098259449005127, | |
| "loss": -0.0334, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.9467870593070984, | |
| "rewards/margins": 0.06303904205560684, | |
| "rewards/rejected": -1.0098259449005127, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 5.10489068806009, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.7864301204681396, | |
| "logits/rejected": -2.7476391792297363, | |
| "logps/chosen": -0.8423538208007812, | |
| "logps/rejected": -0.9428925514221191, | |
| "loss": -0.0302, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.8423538208007812, | |
| "rewards/margins": 0.10053881257772446, | |
| "rewards/rejected": -0.9428925514221191, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 4.722421889688127, | |
| "learning_rate": 3.645833333333333e-07, | |
| "logits/chosen": -2.799630641937256, | |
| "logits/rejected": -2.7175850868225098, | |
| "logps/chosen": -0.8731309771537781, | |
| "logps/rejected": -0.9851009249687195, | |
| "loss": -0.0404, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.8731309771537781, | |
| "rewards/margins": 0.11196992546319962, | |
| "rewards/rejected": -0.9851009249687195, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 4.248923283073742, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.8324596881866455, | |
| "logits/rejected": -2.7458882331848145, | |
| "logps/chosen": -0.8420968055725098, | |
| "logps/rejected": -1.0914112329483032, | |
| "loss": -0.0618, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.8420968055725098, | |
| "rewards/margins": 0.24931427836418152, | |
| "rewards/rejected": -1.0914112329483032, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 5.0405672369883785, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": -2.849945068359375, | |
| "logits/rejected": -2.8102831840515137, | |
| "logps/chosen": -0.8449975848197937, | |
| "logps/rejected": -0.9582854509353638, | |
| "loss": -0.0548, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.8449975848197937, | |
| "rewards/margins": 0.11328791081905365, | |
| "rewards/rejected": -0.9582854509353638, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 5.64617461868394, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -2.811375379562378, | |
| "logits/rejected": -2.7602672576904297, | |
| "logps/chosen": -0.8518880009651184, | |
| "logps/rejected": -1.0725266933441162, | |
| "loss": -0.0542, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.8518880009651184, | |
| "rewards/margins": 0.22063866257667542, | |
| "rewards/rejected": -1.0725266933441162, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 5.384675273611872, | |
| "learning_rate": 4.996716052911017e-07, | |
| "logits/chosen": -2.808159828186035, | |
| "logits/rejected": -2.7851834297180176, | |
| "logps/chosen": -0.848084568977356, | |
| "logps/rejected": -1.0285918712615967, | |
| "loss": -0.0547, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.848084568977356, | |
| "rewards/margins": 0.18050721287727356, | |
| "rewards/rejected": -1.0285918712615967, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 5.059534659236238, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -2.8628323078155518, | |
| "logits/rejected": -2.819815158843994, | |
| "logps/chosen": -0.8899663686752319, | |
| "logps/rejected": -1.0574657917022705, | |
| "loss": -0.0786, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.8899663686752319, | |
| "rewards/margins": 0.167499378323555, | |
| "rewards/rejected": -1.0574657917022705, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 5.268510766313717, | |
| "learning_rate": 4.980652179769217e-07, | |
| "logits/chosen": -2.8412885665893555, | |
| "logits/rejected": -2.7448229789733887, | |
| "logps/chosen": -0.9110374450683594, | |
| "logps/rejected": -1.2632114887237549, | |
| "loss": -0.0663, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.9110374450683594, | |
| "rewards/margins": 0.3521740436553955, | |
| "rewards/rejected": -1.2632114887237549, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 6.376428469132036, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -2.765308380126953, | |
| "logits/rejected": -2.715372085571289, | |
| "logps/chosen": -0.9485328793525696, | |
| "logps/rejected": -1.1706184148788452, | |
| "loss": -0.0717, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.9485328793525696, | |
| "rewards/margins": 0.2220855951309204, | |
| "rewards/rejected": -1.1706184148788452, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 6.941584657656803, | |
| "learning_rate": 4.951291206355559e-07, | |
| "logits/chosen": -2.835057020187378, | |
| "logits/rejected": -2.782494306564331, | |
| "logps/chosen": -0.9551437497138977, | |
| "logps/rejected": -1.2665637731552124, | |
| "loss": -0.0801, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.9551437497138977, | |
| "rewards/margins": 0.3114200532436371, | |
| "rewards/rejected": -1.2665637731552124, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 8.05840757727182, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -2.825330972671509, | |
| "logits/rejected": -2.775848388671875, | |
| "logps/chosen": -1.0829139947891235, | |
| "logps/rejected": -1.4249590635299683, | |
| "loss": -0.0669, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0829139947891235, | |
| "rewards/margins": 0.34204500913619995, | |
| "rewards/rejected": -1.4249590635299683, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 8.045610703659296, | |
| "learning_rate": 4.908790517010636e-07, | |
| "logits/chosen": -2.845106363296509, | |
| "logits/rejected": -2.8046956062316895, | |
| "logps/chosen": -0.9732005000114441, | |
| "logps/rejected": -1.3423811197280884, | |
| "loss": -0.0729, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.9732005000114441, | |
| "rewards/margins": 0.3691806197166443, | |
| "rewards/rejected": -1.3423811197280884, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 7.045936037540457, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -2.7924728393554688, | |
| "logits/rejected": -2.77547550201416, | |
| "logps/chosen": -1.0358796119689941, | |
| "logps/rejected": -1.3562135696411133, | |
| "loss": -0.0767, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0358796119689941, | |
| "rewards/margins": 0.32033389806747437, | |
| "rewards/rejected": -1.3562135696411133, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 8.250297307046639, | |
| "learning_rate": 4.853377929214243e-07, | |
| "logits/chosen": -2.8639283180236816, | |
| "logits/rejected": -2.774474859237671, | |
| "logps/chosen": -1.0332310199737549, | |
| "logps/rejected": -1.3502418994903564, | |
| "loss": -0.076, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.0332310199737549, | |
| "rewards/margins": 0.31701093912124634, | |
| "rewards/rejected": -1.3502418994903564, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 7.621542538359542, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -2.8361611366271973, | |
| "logits/rejected": -2.7507224082946777, | |
| "logps/chosen": -1.0262795686721802, | |
| "logps/rejected": -1.4525015354156494, | |
| "loss": -0.0706, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0262795686721802, | |
| "rewards/margins": 0.42622217535972595, | |
| "rewards/rejected": -1.4525015354156494, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 8.475279965821194, | |
| "learning_rate": 4.785350472409791e-07, | |
| "logits/chosen": -2.843799591064453, | |
| "logits/rejected": -2.832231044769287, | |
| "logps/chosen": -1.0371147394180298, | |
| "logps/rejected": -1.5802838802337646, | |
| "loss": -0.0941, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0371147394180298, | |
| "rewards/margins": 0.5431692600250244, | |
| "rewards/rejected": -1.5802838802337646, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 10.643477587048372, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -2.9215893745422363, | |
| "logits/rejected": -2.878948211669922, | |
| "logps/chosen": -1.1720930337905884, | |
| "logps/rejected": -1.6131317615509033, | |
| "loss": -0.0876, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.1720930337905884, | |
| "rewards/margins": 0.44103869795799255, | |
| "rewards/rejected": -1.6131317615509033, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 12.930215586611359, | |
| "learning_rate": 4.70507279583015e-07, | |
| "logits/chosen": -2.877657175064087, | |
| "logits/rejected": -2.841059684753418, | |
| "logps/chosen": -1.1004445552825928, | |
| "logps/rejected": -1.676473617553711, | |
| "loss": -0.1099, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.1004445552825928, | |
| "rewards/margins": 0.5760290622711182, | |
| "rewards/rejected": -1.676473617553711, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 10.033868511228727, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -2.8931796550750732, | |
| "logits/rejected": -2.865452289581299, | |
| "logps/chosen": -1.1294819116592407, | |
| "logps/rejected": -1.5640286207199097, | |
| "loss": -0.0865, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1294819116592407, | |
| "rewards/margins": 0.4345467984676361, | |
| "rewards/rejected": -1.5640286207199097, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 6.830768956424789, | |
| "learning_rate": 4.612975213859487e-07, | |
| "logits/chosen": -2.9503254890441895, | |
| "logits/rejected": -2.9335081577301025, | |
| "logps/chosen": -1.0866938829421997, | |
| "logps/rejected": -1.4504375457763672, | |
| "loss": -0.0869, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0866938829421997, | |
| "rewards/margins": 0.36374372243881226, | |
| "rewards/rejected": -1.4504375457763672, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 8.139451436965633, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -2.9254631996154785, | |
| "logits/rejected": -2.9086062908172607, | |
| "logps/chosen": -1.1849825382232666, | |
| "logps/rejected": -1.4210944175720215, | |
| "loss": -0.0901, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1849825382232666, | |
| "rewards/margins": 0.23611190915107727, | |
| "rewards/rejected": -1.4210944175720215, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 9.401861364391584, | |
| "learning_rate": 4.5095513994085974e-07, | |
| "logits/chosen": -2.9616475105285645, | |
| "logits/rejected": -2.905653476715088, | |
| "logps/chosen": -1.2673760652542114, | |
| "logps/rejected": -2.075326681137085, | |
| "loss": -0.092, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.2673760652542114, | |
| "rewards/margins": 0.8079506754875183, | |
| "rewards/rejected": -2.075326681137085, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 9.803219381697673, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -2.973576068878174, | |
| "logits/rejected": -2.9448180198669434, | |
| "logps/chosen": -1.2507134675979614, | |
| "logps/rejected": -1.8247950077056885, | |
| "loss": -0.102, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.2507134675979614, | |
| "rewards/margins": 0.574081301689148, | |
| "rewards/rejected": -1.8247950077056885, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 10.267181544033937, | |
| "learning_rate": 4.395355737667985e-07, | |
| "logits/chosen": -2.9244115352630615, | |
| "logits/rejected": -2.9088027477264404, | |
| "logps/chosen": -1.187650203704834, | |
| "logps/rejected": -1.6933050155639648, | |
| "loss": -0.0981, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.187650203704834, | |
| "rewards/margins": 0.50565505027771, | |
| "rewards/rejected": -1.6933050155639648, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 7.725997612608294, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -2.9658923149108887, | |
| "logits/rejected": -2.961773157119751, | |
| "logps/chosen": -1.0811924934387207, | |
| "logps/rejected": -1.5781309604644775, | |
| "loss": -0.1127, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.0811924934387207, | |
| "rewards/margins": 0.4969385266304016, | |
| "rewards/rejected": -1.5781309604644775, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 12.94256989389059, | |
| "learning_rate": 4.271000354423425e-07, | |
| "logits/chosen": -2.912257671356201, | |
| "logits/rejected": -2.9081103801727295, | |
| "logps/chosen": -1.1538851261138916, | |
| "logps/rejected": -1.6617982387542725, | |
| "loss": -0.1026, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.1538851261138916, | |
| "rewards/margins": 0.5079129934310913, | |
| "rewards/rejected": -1.6617982387542725, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 13.90171157708302, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -2.9360742568969727, | |
| "logits/rejected": -2.9269962310791016, | |
| "logps/chosen": -1.2122118473052979, | |
| "logps/rejected": -1.764357328414917, | |
| "loss": -0.1052, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.2122118473052979, | |
| "rewards/margins": 0.5521454811096191, | |
| "rewards/rejected": -1.764357328414917, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 15.393753334767192, | |
| "learning_rate": 4.137151834863213e-07, | |
| "logits/chosen": -2.9690206050872803, | |
| "logits/rejected": -2.9170870780944824, | |
| "logps/chosen": -1.3134431838989258, | |
| "logps/rejected": -1.9632434844970703, | |
| "loss": -0.1079, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.3134431838989258, | |
| "rewards/margins": 0.649800181388855, | |
| "rewards/rejected": -1.9632434844970703, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 10.80120115966252, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -2.985761880874634, | |
| "logits/rejected": -2.958829164505005, | |
| "logps/chosen": -1.4966654777526855, | |
| "logps/rejected": -2.1491572856903076, | |
| "loss": -0.0867, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.4966654777526855, | |
| "rewards/margins": 0.6524916291236877, | |
| "rewards/rejected": -2.1491572856903076, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 28.803806860841444, | |
| "learning_rate": 3.994527650465352e-07, | |
| "logits/chosen": -2.9681310653686523, | |
| "logits/rejected": -2.9533145427703857, | |
| "logps/chosen": -1.4731115102767944, | |
| "logps/rejected": -2.046133518218994, | |
| "loss": -0.0868, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.4731115102767944, | |
| "rewards/margins": 0.5730219483375549, | |
| "rewards/rejected": -2.046133518218994, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 11.655911123379951, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -2.9637484550476074, | |
| "logits/rejected": -2.967674493789673, | |
| "logps/chosen": -1.3848495483398438, | |
| "logps/rejected": -1.9893649816513062, | |
| "loss": -0.0993, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.3848495483398438, | |
| "rewards/margins": 0.6045154929161072, | |
| "rewards/rejected": -1.9893649816513062, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 6.9952388318039915, | |
| "learning_rate": 3.8438923131177237e-07, | |
| "logits/chosen": -2.9558112621307373, | |
| "logits/rejected": -2.926105499267578, | |
| "logps/chosen": -1.3123891353607178, | |
| "logps/rejected": -2.181980609893799, | |
| "loss": -0.1144, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3123891353607178, | |
| "rewards/margins": 0.8695914149284363, | |
| "rewards/rejected": -2.181980609893799, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 9.48550641226516, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": -2.989470958709717, | |
| "logits/rejected": -2.9828832149505615, | |
| "logps/chosen": -1.266231894493103, | |
| "logps/rejected": -1.7893718481063843, | |
| "loss": -0.0895, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.266231894493103, | |
| "rewards/margins": 0.5231398940086365, | |
| "rewards/rejected": -1.7893718481063843, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 7.2796641852323365, | |
| "learning_rate": 3.6860532770864005e-07, | |
| "logits/chosen": -2.914616346359253, | |
| "logits/rejected": -2.923229217529297, | |
| "logps/chosen": -1.3346220254898071, | |
| "logps/rejected": -1.9331144094467163, | |
| "loss": -0.099, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.3346220254898071, | |
| "rewards/margins": 0.5984923839569092, | |
| "rewards/rejected": -1.9331144094467163, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 8.415181065944099, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": -2.9686055183410645, | |
| "logits/rejected": -2.97672438621521, | |
| "logps/chosen": -1.2434272766113281, | |
| "logps/rejected": -1.8032289743423462, | |
| "loss": -0.1085, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.2434272766113281, | |
| "rewards/margins": 0.5598015785217285, | |
| "rewards/rejected": -1.8032289743423462, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 10.248346197393374, | |
| "learning_rate": 3.5218566107988867e-07, | |
| "logits/chosen": -2.896049976348877, | |
| "logits/rejected": -2.8761019706726074, | |
| "logps/chosen": -1.1519674062728882, | |
| "logps/rejected": -1.8064264059066772, | |
| "loss": -0.1144, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.1519674062728882, | |
| "rewards/margins": 0.6544591188430786, | |
| "rewards/rejected": -1.8064264059066772, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 11.20041465175166, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -2.915235757827759, | |
| "logits/rejected": -2.908862590789795, | |
| "logps/chosen": -1.3097448348999023, | |
| "logps/rejected": -1.8037636280059814, | |
| "loss": -0.098, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3097448348999023, | |
| "rewards/margins": 0.4940188527107239, | |
| "rewards/rejected": -1.8037636280059814, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 8.772446686382459, | |
| "learning_rate": 3.3521824616429284e-07, | |
| "logits/chosen": -2.946486711502075, | |
| "logits/rejected": -2.9635918140411377, | |
| "logps/chosen": -1.1646149158477783, | |
| "logps/rejected": -1.9085626602172852, | |
| "loss": -0.1115, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -1.1646149158477783, | |
| "rewards/margins": 0.7439476847648621, | |
| "rewards/rejected": -1.9085626602172852, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 10.868521388953889, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -2.939148426055908, | |
| "logits/rejected": -2.9385180473327637, | |
| "logps/chosen": -1.1780840158462524, | |
| "logps/rejected": -1.849416732788086, | |
| "loss": -0.091, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.1780840158462524, | |
| "rewards/margins": 0.6713327169418335, | |
| "rewards/rejected": -1.849416732788086, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 8.31945763461472, | |
| "learning_rate": 3.1779403380910425e-07, | |
| "logits/chosen": -2.9442903995513916, | |
| "logits/rejected": -2.9341604709625244, | |
| "logps/chosen": -1.1692168712615967, | |
| "logps/rejected": -1.9466499090194702, | |
| "loss": -0.1079, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.1692168712615967, | |
| "rewards/margins": 0.7774332761764526, | |
| "rewards/rejected": -1.9466499090194702, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 10.106291952518397, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -2.932600736618042, | |
| "logits/rejected": -2.9126811027526855, | |
| "logps/chosen": -1.2057874202728271, | |
| "logps/rejected": -2.0929553508758545, | |
| "loss": -0.1181, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.2057874202728271, | |
| "rewards/margins": 0.8871678113937378, | |
| "rewards/rejected": -2.0929553508758545, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 7.6446782498761365, | |
| "learning_rate": 3.000064234440111e-07, | |
| "logits/chosen": -2.9618754386901855, | |
| "logits/rejected": -2.977595567703247, | |
| "logps/chosen": -1.078565239906311, | |
| "logps/rejected": -1.8398815393447876, | |
| "loss": -0.1114, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.078565239906311, | |
| "rewards/margins": 0.7613162994384766, | |
| "rewards/rejected": -1.8398815393447876, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 8.30952989667027, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -2.9029111862182617, | |
| "logits/rejected": -2.9119153022766113, | |
| "logps/chosen": -1.1839892864227295, | |
| "logps/rejected": -1.901410460472107, | |
| "loss": -0.101, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.1839892864227295, | |
| "rewards/margins": 0.7174211740493774, | |
| "rewards/rejected": -1.901410460472107, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 10.720317971753719, | |
| "learning_rate": 2.8195076242990116e-07, | |
| "logits/chosen": -2.977177143096924, | |
| "logits/rejected": -2.937913417816162, | |
| "logps/chosen": -1.388236403465271, | |
| "logps/rejected": -1.9425328969955444, | |
| "loss": -0.1149, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.388236403465271, | |
| "rewards/margins": 0.5542966723442078, | |
| "rewards/rejected": -1.9425328969955444, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 11.599028045358311, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -2.8707821369171143, | |
| "logits/rejected": -2.8736767768859863, | |
| "logps/chosen": -1.2170326709747314, | |
| "logps/rejected": -2.0088775157928467, | |
| "loss": -0.1132, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.2170326709747314, | |
| "rewards/margins": 0.7918451428413391, | |
| "rewards/rejected": -2.0088775157928467, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 6.767222970874098, | |
| "learning_rate": 2.6372383496608186e-07, | |
| "logits/chosen": -2.9382269382476807, | |
| "logits/rejected": -2.917163133621216, | |
| "logps/chosen": -1.354477882385254, | |
| "logps/rejected": -2.248412609100342, | |
| "loss": -0.1329, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.354477882385254, | |
| "rewards/margins": 0.8939343690872192, | |
| "rewards/rejected": -2.248412609100342, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 9.238458870241827, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -2.8694167137145996, | |
| "logits/rejected": -2.845099687576294, | |
| "logps/chosen": -1.115337610244751, | |
| "logps/rejected": -2.19693660736084, | |
| "loss": -0.1178, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.115337610244751, | |
| "rewards/margins": 1.0815991163253784, | |
| "rewards/rejected": -2.19693660736084, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 13.4888484700747, | |
| "learning_rate": 2.454233432955807e-07, | |
| "logits/chosen": -2.9096131324768066, | |
| "logits/rejected": -2.910264492034912, | |
| "logps/chosen": -1.3030673265457153, | |
| "logps/rejected": -1.8178091049194336, | |
| "loss": -0.1209, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.3030673265457153, | |
| "rewards/margins": 0.5147419571876526, | |
| "rewards/rejected": -1.8178091049194336, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 8.160151674440302, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": -2.918236017227173, | |
| "logits/rejected": -2.903280019760132, | |
| "logps/chosen": -1.3952782154083252, | |
| "logps/rejected": -1.8908274173736572, | |
| "loss": -0.0979, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.3952782154083252, | |
| "rewards/margins": 0.4955490529537201, | |
| "rewards/rejected": -1.8908274173736572, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 6.322203620140199, | |
| "learning_rate": 2.2714738398943308e-07, | |
| "logits/chosen": -2.9722788333892822, | |
| "logits/rejected": -2.9391684532165527, | |
| "logps/chosen": -1.161834716796875, | |
| "logps/rejected": -2.083416700363159, | |
| "loss": -0.1238, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.161834716796875, | |
| "rewards/margins": 0.9215821027755737, | |
| "rewards/rejected": -2.083416700363159, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 14.761253087712326, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": -2.990624189376831, | |
| "logits/rejected": -3.004688024520874, | |
| "logps/chosen": -1.219523549079895, | |
| "logps/rejected": -2.0012402534484863, | |
| "loss": -0.11, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.219523549079895, | |
| "rewards/margins": 0.7817164659500122, | |
| "rewards/rejected": -2.0012402534484863, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 7.781855024216495, | |
| "learning_rate": 2.089939221172446e-07, | |
| "logits/chosen": -2.9097352027893066, | |
| "logits/rejected": -2.9420933723449707, | |
| "logps/chosen": -1.2899097204208374, | |
| "logps/rejected": -2.0939292907714844, | |
| "loss": -0.1177, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.2899097204208374, | |
| "rewards/margins": 0.8040194511413574, | |
| "rewards/rejected": -2.0939292907714844, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 12.655201514814673, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": -2.9253571033477783, | |
| "logits/rejected": -2.953551769256592, | |
| "logps/chosen": -1.3859082460403442, | |
| "logps/rejected": -2.3072381019592285, | |
| "loss": -0.1299, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.3859082460403442, | |
| "rewards/margins": 0.9213299751281738, | |
| "rewards/rejected": -2.3072381019592285, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 11.433352540875402, | |
| "learning_rate": 1.9106026612264315e-07, | |
| "logits/chosen": -2.966884136199951, | |
| "logits/rejected": -2.9579386711120605, | |
| "logps/chosen": -1.3356444835662842, | |
| "logps/rejected": -2.194772481918335, | |
| "loss": -0.1222, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.3356444835662842, | |
| "rewards/margins": 0.8591278791427612, | |
| "rewards/rejected": -2.194772481918335, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 13.654932341283944, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": -2.896115779876709, | |
| "logits/rejected": -2.908766508102417, | |
| "logps/chosen": -1.2968409061431885, | |
| "logps/rejected": -2.0684666633605957, | |
| "loss": -0.1182, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.2968409061431885, | |
| "rewards/margins": 0.771625816822052, | |
| "rewards/rejected": -2.0684666633605957, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 22.023298266121166, | |
| "learning_rate": 1.7344254621846017e-07, | |
| "logits/chosen": -2.8941891193389893, | |
| "logits/rejected": -2.9071710109710693, | |
| "logps/chosen": -1.4331861734390259, | |
| "logps/rejected": -2.3704934120178223, | |
| "loss": -0.1296, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.4331861734390259, | |
| "rewards/margins": 0.9373072385787964, | |
| "rewards/rejected": -2.3704934120178223, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 27.97440416354687, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": -2.9143924713134766, | |
| "logits/rejected": -2.9330332279205322, | |
| "logps/chosen": -1.33648681640625, | |
| "logps/rejected": -2.2400500774383545, | |
| "loss": -0.123, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.33648681640625, | |
| "rewards/margins": 0.9035631418228149, | |
| "rewards/rejected": -2.2400500774383545, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 15.481739945463392, | |
| "learning_rate": 1.562351990976095e-07, | |
| "logits/chosen": -2.910585641860962, | |
| "logits/rejected": -2.9256134033203125, | |
| "logps/chosen": -1.3004146814346313, | |
| "logps/rejected": -2.403611660003662, | |
| "loss": -0.129, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.3004146814346313, | |
| "rewards/margins": 1.1031970977783203, | |
| "rewards/rejected": -2.403611660003662, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 18.909069627054414, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": -2.923509359359741, | |
| "logits/rejected": -2.921705961227417, | |
| "logps/chosen": -1.3893711566925049, | |
| "logps/rejected": -2.168813467025757, | |
| "loss": -0.1115, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.3893711566925049, | |
| "rewards/margins": 0.7794424295425415, | |
| "rewards/rejected": -2.168813467025757, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 11.847316888557176, | |
| "learning_rate": 1.3953046172178413e-07, | |
| "logits/chosen": -2.913099765777588, | |
| "logits/rejected": -2.9084200859069824, | |
| "logps/chosen": -1.4098222255706787, | |
| "logps/rejected": -2.1918272972106934, | |
| "loss": -0.1317, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.4098222255706787, | |
| "rewards/margins": 0.7820051908493042, | |
| "rewards/rejected": -2.1918272972106934, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 9.579695107141715, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": -2.878218173980713, | |
| "logits/rejected": -2.89180326461792, | |
| "logps/chosen": -1.301692247390747, | |
| "logps/rejected": -1.9922940731048584, | |
| "loss": -0.1099, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.301692247390747, | |
| "rewards/margins": 0.69060218334198, | |
| "rewards/rejected": -1.9922940731048584, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 8.934316314851877, | |
| "learning_rate": 1.2341787690142435e-07, | |
| "logits/chosen": -2.977961301803589, | |
| "logits/rejected": -2.9634194374084473, | |
| "logps/chosen": -1.2622421979904175, | |
| "logps/rejected": -2.0908942222595215, | |
| "loss": -0.1114, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2622421979904175, | |
| "rewards/margins": 0.8286520838737488, | |
| "rewards/rejected": -2.0908942222595215, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 11.133085955243153, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": -2.903007984161377, | |
| "logits/rejected": -2.8985226154327393, | |
| "logps/chosen": -1.5628604888916016, | |
| "logps/rejected": -2.234027147293091, | |
| "loss": -0.1115, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.5628604888916016, | |
| "rewards/margins": 0.671166718006134, | |
| "rewards/rejected": -2.234027147293091, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 27.176229278437233, | |
| "learning_rate": 1.0798381331721107e-07, | |
| "logits/chosen": -2.941560983657837, | |
| "logits/rejected": -2.95886492729187, | |
| "logps/chosen": -1.335837483406067, | |
| "logps/rejected": -2.2514700889587402, | |
| "loss": -0.1117, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.335837483406067, | |
| "rewards/margins": 0.915632426738739, | |
| "rewards/rejected": -2.2514700889587402, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 11.273733320111317, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": -2.922356605529785, | |
| "logits/rejected": -2.9291598796844482, | |
| "logps/chosen": -1.3571909666061401, | |
| "logps/rejected": -2.2206640243530273, | |
| "loss": -0.1083, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.3571909666061401, | |
| "rewards/margins": 0.8634729385375977, | |
| "rewards/rejected": -2.2206640243530273, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 8.643223960419931, | |
| "learning_rate": 9.331100255592436e-08, | |
| "logits/chosen": -2.921978712081909, | |
| "logits/rejected": -2.928036689758301, | |
| "logps/chosen": -1.165281057357788, | |
| "logps/rejected": -2.1931347846984863, | |
| "loss": -0.1233, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.165281057357788, | |
| "rewards/margins": 1.0278538465499878, | |
| "rewards/rejected": -2.1931347846984863, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 7.265083364287583, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -2.9317374229431152, | |
| "logits/rejected": -2.9432830810546875, | |
| "logps/chosen": -1.1699204444885254, | |
| "logps/rejected": -1.848118543624878, | |
| "loss": -0.1182, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.1699204444885254, | |
| "rewards/margins": 0.6781980395317078, | |
| "rewards/rejected": -1.848118543624878, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 10.732814114542823, | |
| "learning_rate": 7.947809564230445e-08, | |
| "logits/chosen": -2.8792500495910645, | |
| "logits/rejected": -2.859135866165161, | |
| "logps/chosen": -1.356421709060669, | |
| "logps/rejected": -2.247119903564453, | |
| "loss": -0.1246, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.356421709060669, | |
| "rewards/margins": 0.8906981348991394, | |
| "rewards/rejected": -2.247119903564453, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 23.67344548337751, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": -2.941410541534424, | |
| "logits/rejected": -2.9354825019836426, | |
| "logps/chosen": -1.300153374671936, | |
| "logps/rejected": -2.333343982696533, | |
| "loss": -0.1296, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.300153374671936, | |
| "rewards/margins": 1.0331909656524658, | |
| "rewards/rejected": -2.333343982696533, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 16.50894336718241, | |
| "learning_rate": 6.655924144404906e-08, | |
| "logits/chosen": -2.947685956954956, | |
| "logits/rejected": -2.9544997215270996, | |
| "logps/chosen": -1.372989535331726, | |
| "logps/rejected": -2.1334853172302246, | |
| "loss": -0.1197, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.372989535331726, | |
| "rewards/margins": 0.7604957818984985, | |
| "rewards/rejected": -2.1334853172302246, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 7.038795457682623, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": -2.9626219272613525, | |
| "logits/rejected": -2.918588161468506, | |
| "logps/chosen": -1.3632184267044067, | |
| "logps/rejected": -2.0963335037231445, | |
| "loss": -0.1351, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.3632184267044067, | |
| "rewards/margins": 0.7331152558326721, | |
| "rewards/rejected": -2.0963335037231445, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 61.45840490249221, | |
| "learning_rate": 5.4623689209832484e-08, | |
| "logits/chosen": -2.9400558471679688, | |
| "logits/rejected": -2.9316763877868652, | |
| "logps/chosen": -1.2261004447937012, | |
| "logps/rejected": -1.9760372638702393, | |
| "loss": -0.1311, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.2261004447937012, | |
| "rewards/margins": 0.7499366998672485, | |
| "rewards/rejected": -1.9760372638702393, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 11.104870225126671, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": -2.9536826610565186, | |
| "logits/rejected": -2.9647164344787598, | |
| "logps/chosen": -1.1496851444244385, | |
| "logps/rejected": -2.2672224044799805, | |
| "loss": -0.1224, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1496851444244385, | |
| "rewards/margins": 1.1175371408462524, | |
| "rewards/rejected": -2.2672224044799805, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 10.67768391888189, | |
| "learning_rate": 4.373541737087263e-08, | |
| "logits/chosen": -2.8797359466552734, | |
| "logits/rejected": -2.900519847869873, | |
| "logps/chosen": -1.1874008178710938, | |
| "logps/rejected": -2.000943183898926, | |
| "loss": -0.1076, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.1874008178710938, | |
| "rewards/margins": 0.8135424852371216, | |
| "rewards/rejected": -2.000943183898926, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 9.9535498994722, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": -2.876695156097412, | |
| "logits/rejected": -2.8730883598327637, | |
| "logps/chosen": -1.2059592008590698, | |
| "logps/rejected": -1.958923101425171, | |
| "loss": -0.1152, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.2059592008590698, | |
| "rewards/margins": 0.7529637813568115, | |
| "rewards/rejected": -1.958923101425171, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 14.041238838857938, | |
| "learning_rate": 3.3952790595787986e-08, | |
| "logits/chosen": -2.9533257484436035, | |
| "logits/rejected": -2.954756498336792, | |
| "logps/chosen": -1.3264501094818115, | |
| "logps/rejected": -2.21103572845459, | |
| "loss": -0.1076, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.3264501094818115, | |
| "rewards/margins": 0.8845857381820679, | |
| "rewards/rejected": -2.21103572845459, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 9.61660903668533, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": -2.957660675048828, | |
| "logits/rejected": -2.9421029090881348, | |
| "logps/chosen": -1.1808087825775146, | |
| "logps/rejected": -1.8946870565414429, | |
| "loss": -0.1219, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.1808087825775146, | |
| "rewards/margins": 0.713878333568573, | |
| "rewards/rejected": -1.8946870565414429, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 13.35171463730949, | |
| "learning_rate": 2.5328246937043525e-08, | |
| "logits/chosen": -3.0037460327148438, | |
| "logits/rejected": -3.0106940269470215, | |
| "logps/chosen": -1.2479138374328613, | |
| "logps/rejected": -2.1456570625305176, | |
| "loss": -0.1291, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.2479138374328613, | |
| "rewards/margins": 0.8977430462837219, | |
| "rewards/rejected": -2.1456570625305176, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 8.237171381765927, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": -2.9479126930236816, | |
| "logits/rejected": -2.939549207687378, | |
| "logps/chosen": -1.2511965036392212, | |
| "logps/rejected": -2.1256661415100098, | |
| "loss": -0.1115, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.2511965036392212, | |
| "rewards/margins": 0.8744699358940125, | |
| "rewards/rejected": -2.1256661415100098, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 14.905412422914672, | |
| "learning_rate": 1.7908016745981856e-08, | |
| "logits/chosen": -2.9071784019470215, | |
| "logits/rejected": -2.9462032318115234, | |
| "logps/chosen": -1.3677804470062256, | |
| "logps/rejected": -2.20474910736084, | |
| "loss": -0.129, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.3677804470062256, | |
| "rewards/margins": 0.8369685411453247, | |
| "rewards/rejected": -2.20474910736084, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 12.518203010975816, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": -2.886582612991333, | |
| "logits/rejected": -2.894714832305908, | |
| "logps/chosen": -1.333262324333191, | |
| "logps/rejected": -2.2248027324676514, | |
| "loss": -0.1169, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.333262324333191, | |
| "rewards/margins": 0.8915404081344604, | |
| "rewards/rejected": -2.2248027324676514, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 19.29597280415641, | |
| "learning_rate": 1.1731874863145142e-08, | |
| "logits/chosen": -2.9716782569885254, | |
| "logits/rejected": -2.9714698791503906, | |
| "logps/chosen": -1.3335239887237549, | |
| "logps/rejected": -2.1658711433410645, | |
| "loss": -0.1292, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.3335239887237549, | |
| "rewards/margins": 0.8323470950126648, | |
| "rewards/rejected": -2.1658711433410645, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 12.085590006123791, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": -2.91271710395813, | |
| "logits/rejected": -2.922774076461792, | |
| "logps/chosen": -1.2182632684707642, | |
| "logps/rejected": -2.144477367401123, | |
| "loss": -0.1285, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.2182632684707642, | |
| "rewards/margins": 0.9262140393257141, | |
| "rewards/rejected": -2.144477367401123, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 12.169327202340588, | |
| "learning_rate": 6.832927412229017e-09, | |
| "logits/chosen": -2.9808132648468018, | |
| "logits/rejected": -2.9507627487182617, | |
| "logps/chosen": -1.3045051097869873, | |
| "logps/rejected": -2.0467162132263184, | |
| "loss": -0.1058, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3045051097869873, | |
| "rewards/margins": 0.7422109246253967, | |
| "rewards/rejected": -2.0467162132263184, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 7.578781116399002, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": -2.861370086669922, | |
| "logits/rejected": -2.8586971759796143, | |
| "logps/chosen": -1.26282799243927, | |
| "logps/rejected": -2.12613844871521, | |
| "loss": -0.1329, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.26282799243927, | |
| "rewards/margins": 0.8633106350898743, | |
| "rewards/rejected": -2.12613844871521, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 15.590655526363935, | |
| "learning_rate": 3.2374343405217884e-09, | |
| "logits/chosen": -2.943765163421631, | |
| "logits/rejected": -2.9257071018218994, | |
| "logps/chosen": -1.4191820621490479, | |
| "logps/rejected": -2.1226634979248047, | |
| "loss": -0.1276, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.4191820621490479, | |
| "rewards/margins": 0.7034814953804016, | |
| "rewards/rejected": -2.1226634979248047, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 7.099005793414053, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": -2.9480669498443604, | |
| "logits/rejected": -2.95448637008667, | |
| "logps/chosen": -1.2225438356399536, | |
| "logps/rejected": -2.0639634132385254, | |
| "loss": -0.1163, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.2225438356399536, | |
| "rewards/margins": 0.841419517993927, | |
| "rewards/rejected": -2.0639634132385254, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 9.4638677347299, | |
| "learning_rate": 9.64668657069706e-10, | |
| "logits/chosen": -2.9739177227020264, | |
| "logits/rejected": -2.964212656021118, | |
| "logps/chosen": -1.31159245967865, | |
| "logps/rejected": -2.171808958053589, | |
| "loss": -0.1218, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.31159245967865, | |
| "rewards/margins": 0.860216498374939, | |
| "rewards/rejected": -2.171808958053589, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 25.34970183074257, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": -2.919668674468994, | |
| "logits/rejected": -2.907841920852661, | |
| "logps/chosen": -1.2385847568511963, | |
| "logps/rejected": -1.978467583656311, | |
| "loss": -0.1103, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.2385847568511963, | |
| "rewards/margins": 0.7398825883865356, | |
| "rewards/rejected": -1.978467583656311, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 15.471089020268318, | |
| "learning_rate": 2.6813123097352287e-11, | |
| "logits/chosen": -2.9793272018432617, | |
| "logits/rejected": -2.9733455181121826, | |
| "logps/chosen": -1.2928240299224854, | |
| "logps/rejected": -2.0769708156585693, | |
| "loss": -0.1079, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.2928240299224854, | |
| "rewards/margins": 0.7841470241546631, | |
| "rewards/rejected": -2.0769708156585693, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": -0.10079137468612918, | |
| "train_runtime": 7720.3379, | |
| "train_samples_per_second": 7.919, | |
| "train_steps_per_second": 0.062 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |