| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9952, | |
| "eval_steps": 500, | |
| "global_step": 351, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.042666666666666665, | |
| "grad_norm": 23808.0, | |
| "learning_rate": 8.333333333333334e-06, | |
| "log_odds_chosen": 4.913786888122559, | |
| "log_odds_ratio": -7.455605983734131, | |
| "logits/chosen": 104.62542724609375, | |
| "logits/rejected": 103.22361755371094, | |
| "logps/chosen": -24.285247802734375, | |
| "logps/rejected": -29.19942283630371, | |
| "loss": 999.1373, | |
| "nll_loss": 9.384310722351074, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -12.142623901367188, | |
| "rewards/margins": 2.4570868015289307, | |
| "rewards/rejected": -14.599711418151855, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.08533333333333333, | |
| "grad_norm": 20480.0, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "log_odds_chosen": 3.161984443664551, | |
| "log_odds_ratio": -4.3201093673706055, | |
| "logits/chosen": 116.70096588134766, | |
| "logits/rejected": 107.84611511230469, | |
| "logps/chosen": -22.356525421142578, | |
| "logps/rejected": -25.517377853393555, | |
| "loss": 1471.3339, | |
| "nll_loss": 7.478154182434082, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -11.178262710571289, | |
| "rewards/margins": 1.580425500869751, | |
| "rewards/rejected": -12.758688926696777, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 9792.0, | |
| "learning_rate": 2.5e-05, | |
| "log_odds_chosen": 5.405481338500977, | |
| "log_odds_ratio": -3.742039442062378, | |
| "logits/chosen": 101.9460220336914, | |
| "logits/rejected": 144.56015014648438, | |
| "logps/chosen": -16.96074867248535, | |
| "logps/rejected": -22.36530303955078, | |
| "loss": 2500.4584, | |
| "nll_loss": 9.246469497680664, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -8.480374336242676, | |
| "rewards/margins": 2.7022786140441895, | |
| "rewards/rejected": -11.18265151977539, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17066666666666666, | |
| "grad_norm": 23424.0, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "log_odds_chosen": 1.755802869796753, | |
| "log_odds_ratio": -8.513590812683105, | |
| "logits/chosen": 138.0966033935547, | |
| "logits/rejected": 115.8319320678711, | |
| "logps/chosen": -21.58835220336914, | |
| "logps/rejected": -23.347524642944336, | |
| "loss": -909.4568, | |
| "nll_loss": 7.946342468261719, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -10.79417610168457, | |
| "rewards/margins": 0.879586398601532, | |
| "rewards/rejected": -11.673762321472168, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 81920.0, | |
| "learning_rate": 4.1666666666666665e-05, | |
| "log_odds_chosen": -3.714871883392334, | |
| "log_odds_ratio": -11.706196784973145, | |
| "logits/chosen": 131.1918182373047, | |
| "logits/rejected": 112.025146484375, | |
| "logps/chosen": -27.263925552368164, | |
| "logps/rejected": -23.55154037475586, | |
| "loss": -1289.5892, | |
| "nll_loss": 11.121248245239258, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -13.631962776184082, | |
| "rewards/margins": -1.8561919927597046, | |
| "rewards/rejected": -11.77577018737793, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 23424.0, | |
| "learning_rate": 5e-05, | |
| "log_odds_chosen": 5.615313529968262, | |
| "log_odds_ratio": -6.0788164138793945, | |
| "logits/chosen": 120.47991943359375, | |
| "logits/rejected": 131.72543334960938, | |
| "logps/chosen": -20.34619903564453, | |
| "logps/rejected": -25.962305068969727, | |
| "loss": 1989.467, | |
| "nll_loss": 11.813726425170898, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -10.173099517822266, | |
| "rewards/margins": 2.808053731918335, | |
| "rewards/rejected": -12.981152534484863, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2986666666666667, | |
| "grad_norm": 12288.0, | |
| "learning_rate": 5.833333333333333e-05, | |
| "log_odds_chosen": 6.8634352684021, | |
| "log_odds_ratio": -4.409341335296631, | |
| "logits/chosen": 106.82928466796875, | |
| "logits/rejected": 135.06765747070312, | |
| "logps/chosen": -17.419185638427734, | |
| "logps/rejected": -24.284481048583984, | |
| "loss": 2134.8127, | |
| "nll_loss": 8.908151626586914, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -8.709592819213867, | |
| "rewards/margins": 3.4326491355895996, | |
| "rewards/rejected": -12.142240524291992, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3413333333333333, | |
| "grad_norm": 20096.0, | |
| "learning_rate": 5.997613110678538e-05, | |
| "log_odds_chosen": 10.421220779418945, | |
| "log_odds_ratio": -4.025184154510498, | |
| "logits/chosen": 97.60896301269531, | |
| "logits/rejected": 131.75054931640625, | |
| "logps/chosen": -20.480510711669922, | |
| "logps/rejected": -30.901927947998047, | |
| "loss": 2004.0334, | |
| "nll_loss": 10.660150527954102, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -10.240255355834961, | |
| "rewards/margins": 5.2107110023498535, | |
| "rewards/rejected": -15.450963973999023, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 29952.0, | |
| "learning_rate": 5.987922881985718e-05, | |
| "log_odds_chosen": 2.5370476245880127, | |
| "log_odds_ratio": -9.996942520141602, | |
| "logits/chosen": 125.96684265136719, | |
| "logits/rejected": 126.08040618896484, | |
| "logps/chosen": -22.960010528564453, | |
| "logps/rejected": -25.498239517211914, | |
| "loss": 504.6511, | |
| "nll_loss": 9.340021133422852, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -11.480005264282227, | |
| "rewards/margins": 1.2691147327423096, | |
| "rewards/rejected": -12.749119758605957, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 13312.0, | |
| "learning_rate": 5.970804206224711e-05, | |
| "log_odds_chosen": 9.48165512084961, | |
| "log_odds_ratio": -6.348289489746094, | |
| "logits/chosen": 115.5280532836914, | |
| "logits/rejected": 133.51206970214844, | |
| "logps/chosen": -18.29220962524414, | |
| "logps/rejected": -27.774459838867188, | |
| "loss": 357.9646, | |
| "nll_loss": 8.295930862426758, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -9.14610481262207, | |
| "rewards/margins": 4.74112606048584, | |
| "rewards/rejected": -13.887229919433594, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4693333333333333, | |
| "grad_norm": 36864.0, | |
| "learning_rate": 5.9462996431207166e-05, | |
| "log_odds_chosen": -1.4241477251052856, | |
| "log_odds_ratio": -7.040617942810059, | |
| "logits/chosen": 90.20933532714844, | |
| "logits/rejected": 77.6080322265625, | |
| "logps/chosen": -21.791763305664062, | |
| "logps/rejected": -20.36836051940918, | |
| "loss": 710.7211, | |
| "nll_loss": 8.630704879760742, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -10.895881652832031, | |
| "rewards/margins": -0.7117019891738892, | |
| "rewards/rejected": -10.18418025970459, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 22912.0, | |
| "learning_rate": 5.914470114878602e-05, | |
| "log_odds_chosen": 0.059395600110292435, | |
| "log_odds_ratio": -7.545324802398682, | |
| "logits/chosen": 74.50141906738281, | |
| "logits/rejected": 72.20657348632812, | |
| "logps/chosen": -27.5406551361084, | |
| "logps/rejected": -27.600433349609375, | |
| "loss": 835.8969, | |
| "nll_loss": 11.118535995483398, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -13.7703275680542, | |
| "rewards/margins": 0.029887771233916283, | |
| "rewards/rejected": -13.800216674804688, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5546666666666666, | |
| "grad_norm": 59136.0, | |
| "learning_rate": 5.875394754720707e-05, | |
| "log_odds_chosen": 0.004063797183334827, | |
| "log_odds_ratio": -8.829879760742188, | |
| "logits/chosen": 107.4288101196289, | |
| "logits/rejected": 100.03871154785156, | |
| "logps/chosen": -26.154687881469727, | |
| "logps/rejected": -26.159423828125, | |
| "loss": 1129.1766, | |
| "nll_loss": 8.619396209716797, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -13.077343940734863, | |
| "rewards/margins": 0.0023682594764977694, | |
| "rewards/rejected": -13.0797119140625, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5973333333333334, | |
| "grad_norm": 84992.0, | |
| "learning_rate": 5.8291707101491815e-05, | |
| "log_odds_chosen": -9.859933853149414, | |
| "log_odds_ratio": -13.291154861450195, | |
| "logits/chosen": 130.4120635986328, | |
| "logits/rejected": 107.79060363769531, | |
| "logps/chosen": -33.194881439208984, | |
| "logps/rejected": -23.33577537536621, | |
| "loss": -1070.0952, | |
| "nll_loss": 9.995885848999023, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -16.597440719604492, | |
| "rewards/margins": -4.929553031921387, | |
| "rewards/rejected": -11.667887687683105, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 27264.0, | |
| "learning_rate": 5.77591290142199e-05, | |
| "log_odds_chosen": -1.0986392498016357, | |
| "log_odds_ratio": -6.415988922119141, | |
| "logits/chosen": 202.5902862548828, | |
| "logits/rejected": 170.36766052246094, | |
| "logps/chosen": -28.743408203125, | |
| "logps/rejected": -27.644739151000977, | |
| "loss": 3134.2316, | |
| "nll_loss": 16.13515853881836, | |
| "rewards/accuracies": 0.4124999940395355, | |
| "rewards/chosen": -14.3717041015625, | |
| "rewards/margins": -0.5493333339691162, | |
| "rewards/rejected": -13.822369575500488, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6826666666666666, | |
| "grad_norm": 5920.0, | |
| "learning_rate": 5.7157537358430446e-05, | |
| "log_odds_chosen": 5.658118724822998, | |
| "log_odds_ratio": -7.236645698547363, | |
| "logits/chosen": 79.78996276855469, | |
| "logits/rejected": 120.5929946899414, | |
| "logps/chosen": -39.893455505371094, | |
| "logps/rejected": -45.551578521728516, | |
| "loss": 1430.3527, | |
| "nll_loss": 26.075185775756836, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -19.946727752685547, | |
| "rewards/margins": 2.829059362411499, | |
| "rewards/rejected": -22.775789260864258, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7253333333333334, | |
| "grad_norm": 7264.0, | |
| "learning_rate": 5.648842778576781e-05, | |
| "log_odds_chosen": -2.8542323112487793, | |
| "log_odds_ratio": -11.537806510925293, | |
| "logits/chosen": 43.76961135864258, | |
| "logits/rejected": 40.082550048828125, | |
| "logps/chosen": -41.100486755371094, | |
| "logps/rejected": -38.24618911743164, | |
| "loss": 27.785, | |
| "nll_loss": 18.119293212890625, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -20.550243377685547, | |
| "rewards/margins": -1.4271516799926758, | |
| "rewards/rejected": -19.12309455871582, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 8160.0, | |
| "learning_rate": 5.575346380805599e-05, | |
| "log_odds_chosen": 6.750527858734131, | |
| "log_odds_ratio": -4.975089073181152, | |
| "logits/chosen": 142.46315002441406, | |
| "logits/rejected": 194.29443359375, | |
| "logps/chosen": -27.23212242126465, | |
| "logps/rejected": -33.98273849487305, | |
| "loss": 410.7923, | |
| "nll_loss": 13.293853759765625, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -13.616061210632324, | |
| "rewards/margins": 3.3753085136413574, | |
| "rewards/rejected": -16.991369247436523, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.8106666666666666, | |
| "grad_norm": 6752.0, | |
| "learning_rate": 5.4954472661546075e-05, | |
| "log_odds_chosen": -1.6101436614990234, | |
| "log_odds_ratio": -5.586986064910889, | |
| "logits/chosen": 265.1175537109375, | |
| "logits/rejected": 220.3394012451172, | |
| "logps/chosen": -16.311241149902344, | |
| "logps/rejected": -14.702871322631836, | |
| "loss": 697.2691, | |
| "nll_loss": 8.886590957641602, | |
| "rewards/accuracies": 0.38749998807907104, | |
| "rewards/chosen": -8.155620574951172, | |
| "rewards/margins": -0.8041850924491882, | |
| "rewards/rejected": -7.351435661315918, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 1976.0, | |
| "learning_rate": 5.4093440764119056e-05, | |
| "log_odds_chosen": 0.8467995524406433, | |
| "log_odds_ratio": -3.7493503093719482, | |
| "logits/chosen": 214.7117156982422, | |
| "logits/rejected": 216.2617950439453, | |
| "logps/chosen": -10.674482345581055, | |
| "logps/rejected": -11.517151832580566, | |
| "loss": 671.6329, | |
| "nll_loss": 5.953970909118652, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -5.337241172790527, | |
| "rewards/margins": 0.4213342070579529, | |
| "rewards/rejected": -5.758575916290283, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 233.0, | |
| "learning_rate": 5.317250877673799e-05, | |
| "log_odds_chosen": 0.0025218098890036345, | |
| "log_odds_ratio": -1.1279939413070679, | |
| "logits/chosen": 274.94171142578125, | |
| "logits/rejected": 293.26910400390625, | |
| "logps/chosen": -3.1382219791412354, | |
| "logps/rejected": -3.117029905319214, | |
| "loss": 59.8813, | |
| "nll_loss": 2.759488105773926, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.5691109895706177, | |
| "rewards/margins": -0.010595941916108131, | |
| "rewards/rejected": -1.558514952659607, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.9386666666666666, | |
| "grad_norm": 93.0, | |
| "learning_rate": 5.219396628142752e-05, | |
| "log_odds_chosen": 0.3583167493343353, | |
| "log_odds_ratio": -0.7731421589851379, | |
| "logits/chosen": 287.4047546386719, | |
| "logits/rejected": 319.8270263671875, | |
| "logps/chosen": -2.0119025707244873, | |
| "logps/rejected": -2.3544743061065674, | |
| "loss": 41.0939, | |
| "nll_loss": 2.1219072341918945, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.0059512853622437, | |
| "rewards/margins": 0.17128589749336243, | |
| "rewards/rejected": -1.1772371530532837, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9813333333333333, | |
| "grad_norm": 75.5, | |
| "learning_rate": 5.1160246089012264e-05, | |
| "log_odds_chosen": -0.07454674690961838, | |
| "log_odds_ratio": -0.8472925424575806, | |
| "logits/chosen": 304.05816650390625, | |
| "logits/rejected": 292.1544494628906, | |
| "logps/chosen": -1.64755117893219, | |
| "logps/rejected": -1.558643102645874, | |
| "loss": 35.4153, | |
| "nll_loss": 1.8526198863983154, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.823775589466095, | |
| "rewards/margins": -0.04445408657193184, | |
| "rewards/rejected": -0.779321551322937, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 106.5, | |
| "learning_rate": 5.007391819076575e-05, | |
| "log_odds_chosen": 0.20053406059741974, | |
| "log_odds_ratio": -0.6979594826698303, | |
| "logits/chosen": 296.3975524902344, | |
| "logits/rejected": 293.24871826171875, | |
| "logps/chosen": -1.3861749172210693, | |
| "logps/rejected": -1.5457828044891357, | |
| "loss": 32.7872, | |
| "nll_loss": 1.7506237030029297, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.6930874586105347, | |
| "rewards/margins": 0.07980402559041977, | |
| "rewards/rejected": -0.7728914022445679, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 41.0, | |
| "learning_rate": 4.893768336900717e-05, | |
| "log_odds_chosen": 0.20815667510032654, | |
| "log_odds_ratio": -0.654870331287384, | |
| "logits/chosen": 285.3860168457031, | |
| "logits/rejected": 291.6961975097656, | |
| "logps/chosen": -1.3074676990509033, | |
| "logps/rejected": -1.4680944681167603, | |
| "loss": 30.4078, | |
| "nll_loss": 1.648654580116272, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.6537338495254517, | |
| "rewards/margins": 0.08031338453292847, | |
| "rewards/rejected": -0.7340472340583801, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.1093333333333333, | |
| "grad_norm": 83.5, | |
| "learning_rate": 4.775436648253103e-05, | |
| "log_odds_chosen": 0.010318088345229626, | |
| "log_odds_ratio": -0.7326194643974304, | |
| "logits/chosen": 272.5470886230469, | |
| "logits/rejected": 290.8238830566406, | |
| "logps/chosen": -1.2763969898223877, | |
| "logps/rejected": -1.282832384109497, | |
| "loss": 29.9423, | |
| "nll_loss": 1.5382884740829468, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.6381984949111938, | |
| "rewards/margins": 0.0032175942324101925, | |
| "rewards/rejected": -0.6414161920547485, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 52.25, | |
| "learning_rate": 4.6526909443563075e-05, | |
| "log_odds_chosen": 0.05014984682202339, | |
| "log_odds_ratio": -0.7415339350700378, | |
| "logits/chosen": 283.76141357421875, | |
| "logits/rejected": 269.16754150390625, | |
| "logps/chosen": -1.1821494102478027, | |
| "logps/rejected": -1.2269926071166992, | |
| "loss": 29.0019, | |
| "nll_loss": 1.5523165464401245, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.5910747051239014, | |
| "rewards/margins": 0.02242158353328705, | |
| "rewards/rejected": -0.6134963035583496, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1946666666666665, | |
| "grad_norm": 55.75, | |
| "learning_rate": 4.5258363903702954e-05, | |
| "log_odds_chosen": 0.25334566831588745, | |
| "log_odds_ratio": -0.6387948989868164, | |
| "logits/chosen": 279.69866943359375, | |
| "logits/rejected": 307.73309326171875, | |
| "logps/chosen": -1.0961264371871948, | |
| "logps/rejected": -1.2705694437026978, | |
| "loss": 27.7407, | |
| "nll_loss": 1.4683058261871338, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.5480632185935974, | |
| "rewards/margins": 0.08722147345542908, | |
| "rewards/rejected": -0.6352847218513489, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.2373333333333334, | |
| "grad_norm": 63.75, | |
| "learning_rate": 4.395188366703752e-05, | |
| "log_odds_chosen": 0.27662745118141174, | |
| "log_odds_ratio": -0.6522295475006104, | |
| "logits/chosen": 272.47137451171875, | |
| "logits/rejected": 291.1870422363281, | |
| "logps/chosen": -1.1764074563980103, | |
| "logps/rejected": -1.3605537414550781, | |
| "loss": 27.2205, | |
| "nll_loss": 1.449241042137146, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5882037281990051, | |
| "rewards/margins": 0.09207318723201752, | |
| "rewards/rejected": -0.6802768707275391, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 84.5, | |
| "learning_rate": 4.261071684928697e-05, | |
| "log_odds_chosen": 0.08433417975902557, | |
| "log_odds_ratio": -0.7295799255371094, | |
| "logits/chosen": 283.64739990234375, | |
| "logits/rejected": 284.56048583984375, | |
| "logps/chosen": -1.1405603885650635, | |
| "logps/rejected": -1.2061361074447632, | |
| "loss": 27.0297, | |
| "nll_loss": 1.4862051010131836, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5702801942825317, | |
| "rewards/margins": 0.03278781846165657, | |
| "rewards/rejected": -0.6030680537223816, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.3226666666666667, | |
| "grad_norm": 45.75, | |
| "learning_rate": 4.123819780247737e-05, | |
| "log_odds_chosen": 0.19811879098415375, | |
| "log_odds_ratio": -0.6645184755325317, | |
| "logits/chosen": 271.0818176269531, | |
| "logits/rejected": 282.7620544433594, | |
| "logps/chosen": -1.0194677114486694, | |
| "logps/rejected": -1.1592345237731934, | |
| "loss": 26.6033, | |
| "nll_loss": 1.4380306005477905, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5097338557243347, | |
| "rewards/margins": 0.06988338381052017, | |
| "rewards/rejected": -0.5796172618865967, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.3653333333333333, | |
| "grad_norm": 49.25, | |
| "learning_rate": 3.9837738825216133e-05, | |
| "log_odds_chosen": 0.20502634346485138, | |
| "log_odds_ratio": -0.6395789384841919, | |
| "logits/chosen": 265.66180419921875, | |
| "logits/rejected": 298.80450439453125, | |
| "logps/chosen": -1.0500866174697876, | |
| "logps/rejected": -1.1844158172607422, | |
| "loss": 26.1919, | |
| "nll_loss": 1.4191492795944214, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.5250433087348938, | |
| "rewards/margins": 0.06716466695070267, | |
| "rewards/rejected": -0.5922079086303711, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 26.25, | |
| "learning_rate": 3.8412821679180084e-05, | |
| "log_odds_chosen": 0.17818713188171387, | |
| "log_odds_ratio": -0.6783817410469055, | |
| "logits/chosen": 276.02899169921875, | |
| "logits/rejected": 289.51385498046875, | |
| "logps/chosen": -1.0516808032989502, | |
| "logps/rejected": -1.1585873365402222, | |
| "loss": 26.0691, | |
| "nll_loss": 1.353134274482727, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.5258404016494751, | |
| "rewards/margins": 0.053453266620635986, | |
| "rewards/rejected": -0.5792936682701111, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.4506666666666668, | |
| "grad_norm": 23.125, | |
| "learning_rate": 3.6966988932907276e-05, | |
| "log_odds_chosen": 0.16093948483467102, | |
| "log_odds_ratio": -0.6871160268783569, | |
| "logits/chosen": 278.6529541015625, | |
| "logits/rejected": 298.80657958984375, | |
| "logps/chosen": -1.080251693725586, | |
| "logps/rejected": -1.1996749639511108, | |
| "loss": 26.6724, | |
| "nll_loss": 1.4616249799728394, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.540125846862793, | |
| "rewards/margins": 0.059711672365665436, | |
| "rewards/rejected": -0.5998374819755554, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4933333333333334, | |
| "grad_norm": 27.875, | |
| "learning_rate": 3.5503835154413476e-05, | |
| "log_odds_chosen": 0.286944180727005, | |
| "log_odds_ratio": -0.6341909766197205, | |
| "logits/chosen": 274.6024169921875, | |
| "logits/rejected": 295.36651611328125, | |
| "logps/chosen": -1.0407021045684814, | |
| "logps/rejected": -1.2441030740737915, | |
| "loss": 25.7035, | |
| "nll_loss": 1.411714792251587, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.5203510522842407, | |
| "rewards/margins": 0.10170049965381622, | |
| "rewards/rejected": -0.6220515370368958, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 47.0, | |
| "learning_rate": 3.4026997974529664e-05, | |
| "log_odds_chosen": 0.16404980421066284, | |
| "log_odds_ratio": -0.6638838052749634, | |
| "logits/chosen": 290.327880859375, | |
| "logits/rejected": 284.926513671875, | |
| "logps/chosen": -1.093461275100708, | |
| "logps/rejected": -1.1879903078079224, | |
| "loss": 25.5271, | |
| "nll_loss": 1.4601901769638062, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.546730637550354, | |
| "rewards/margins": 0.047264464199543, | |
| "rewards/rejected": -0.5939951539039612, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.5786666666666667, | |
| "grad_norm": 40.75, | |
| "learning_rate": 3.25401490431787e-05, | |
| "log_odds_chosen": 0.28145521879196167, | |
| "log_odds_ratio": -0.6211186647415161, | |
| "logits/chosen": 297.9057922363281, | |
| "logits/rejected": 286.4951477050781, | |
| "logps/chosen": -1.0374724864959717, | |
| "logps/rejected": -1.2180078029632568, | |
| "loss": 25.5928, | |
| "nll_loss": 1.3627592325210571, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.5187362432479858, | |
| "rewards/margins": 0.09026758372783661, | |
| "rewards/rejected": -0.6090039014816284, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.6213333333333333, | |
| "grad_norm": 56.0, | |
| "learning_rate": 3.104698490107504e-05, | |
| "log_odds_chosen": 0.09670724719762802, | |
| "log_odds_ratio": -0.6962383985519409, | |
| "logits/chosen": 296.83575439453125, | |
| "logits/rejected": 268.6614685058594, | |
| "logps/chosen": -1.0726783275604248, | |
| "logps/rejected": -1.1235979795455933, | |
| "loss": 25.6578, | |
| "nll_loss": 1.3929274082183838, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5363391637802124, | |
| "rewards/margins": 0.025459837168455124, | |
| "rewards/rejected": -0.5617989897727966, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 27.125, | |
| "learning_rate": 2.9551217789542096e-05, | |
| "log_odds_chosen": 0.08583483099937439, | |
| "log_odds_ratio": -0.7124528884887695, | |
| "logits/chosen": 291.0882568359375, | |
| "logits/rejected": 282.97711181640625, | |
| "logps/chosen": -1.0238406658172607, | |
| "logps/rejected": -1.0693080425262451, | |
| "loss": 25.3697, | |
| "nll_loss": 1.4079334735870361, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.5119203329086304, | |
| "rewards/margins": 0.02273363620042801, | |
| "rewards/rejected": -0.5346540212631226, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.7066666666666666, | |
| "grad_norm": 29.875, | |
| "learning_rate": 2.8056566421295443e-05, | |
| "log_odds_chosen": 0.027378028258681297, | |
| "log_odds_ratio": -0.7563061714172363, | |
| "logits/chosen": 280.4498291015625, | |
| "logits/rejected": 268.0575256347656, | |
| "logps/chosen": -1.0595781803131104, | |
| "logps/rejected": -1.0435364246368408, | |
| "loss": 24.6268, | |
| "nll_loss": 1.3725634813308716, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.5297890901565552, | |
| "rewards/margins": -0.008020809851586819, | |
| "rewards/rejected": -0.5217682123184204, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.7493333333333334, | |
| "grad_norm": 40.75, | |
| "learning_rate": 2.656674673513705e-05, | |
| "log_odds_chosen": 0.09508597105741501, | |
| "log_odds_ratio": -0.7279762625694275, | |
| "logits/chosen": 284.59503173828125, | |
| "logits/rejected": 292.72509765625, | |
| "logps/chosen": -1.1167399883270264, | |
| "logps/rejected": -1.2022297382354736, | |
| "loss": 25.7565, | |
| "nll_loss": 1.4108952283859253, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.5583699941635132, | |
| "rewards/margins": 0.04274484142661095, | |
| "rewards/rejected": -0.6011148691177368, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 68.5, | |
| "learning_rate": 2.508546265754587e-05, | |
| "log_odds_chosen": 0.14177300035953522, | |
| "log_odds_ratio": -0.6866236925125122, | |
| "logits/chosen": 271.2067565917969, | |
| "logits/rejected": 289.2135009765625, | |
| "logps/chosen": -0.9864645004272461, | |
| "logps/rejected": -1.0824763774871826, | |
| "loss": 24.89, | |
| "nll_loss": 1.4051529169082642, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.49323225021362305, | |
| "rewards/margins": 0.048005927354097366, | |
| "rewards/rejected": -0.5412381887435913, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.8346666666666667, | |
| "grad_norm": 29.5, | |
| "learning_rate": 2.3616396894133145e-05, | |
| "log_odds_chosen": 0.07325839251279831, | |
| "log_odds_ratio": -0.7426969408988953, | |
| "logits/chosen": 277.15673828125, | |
| "logits/rejected": 272.313232421875, | |
| "logps/chosen": -1.044854760169983, | |
| "logps/rejected": -1.0880095958709717, | |
| "loss": 24.9372, | |
| "nll_loss": 1.420508623123169, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.5224273800849915, | |
| "rewards/margins": 0.02157733403146267, | |
| "rewards/rejected": -0.5440047979354858, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.8773333333333333, | |
| "grad_norm": 18.75, | |
| "learning_rate": 2.216320177385585e-05, | |
| "log_odds_chosen": 0.16561657190322876, | |
| "log_odds_ratio": -0.6900728940963745, | |
| "logits/chosen": 269.1635437011719, | |
| "logits/rejected": 290.73931884765625, | |
| "logps/chosen": -0.9672033190727234, | |
| "logps/rejected": -1.0741941928863525, | |
| "loss": 24.5158, | |
| "nll_loss": 1.3439892530441284, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.4836016595363617, | |
| "rewards/margins": 0.05349547788500786, | |
| "rewards/rejected": -0.5370970964431763, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 51.25, | |
| "learning_rate": 2.072949016875158e-05, | |
| "log_odds_chosen": 0.254954069852829, | |
| "log_odds_ratio": -0.6858216524124146, | |
| "logits/chosen": 277.99786376953125, | |
| "logits/rejected": 272.8140563964844, | |
| "logps/chosen": -1.023193597793579, | |
| "logps/rejected": -1.2254831790924072, | |
| "loss": 25.0004, | |
| "nll_loss": 1.3852344751358032, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5115967988967896, | |
| "rewards/margins": 0.10114479064941406, | |
| "rewards/rejected": -0.6127415895462036, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.9626666666666668, | |
| "grad_norm": 42.0, | |
| "learning_rate": 1.9318826511769297e-05, | |
| "log_odds_chosen": -0.024613792076706886, | |
| "log_odds_ratio": -0.7803007364273071, | |
| "logits/chosen": 278.9836730957031, | |
| "logits/rejected": 275.0539855957031, | |
| "logps/chosen": -1.0930712223052979, | |
| "logps/rejected": -1.0882163047790527, | |
| "loss": 25.3575, | |
| "nll_loss": 1.3715641498565674, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.5465356111526489, | |
| "rewards/margins": -0.0024274878669530153, | |
| "rewards/rejected": -0.5441081523895264, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.005333333333333, | |
| "grad_norm": 27.875, | |
| "learning_rate": 1.793471793502748e-05, | |
| "log_odds_chosen": 0.18588228523731232, | |
| "log_odds_ratio": -0.6733505129814148, | |
| "logits/chosen": 266.9411315917969, | |
| "logits/rejected": 268.85430908203125, | |
| "logps/chosen": -0.966964066028595, | |
| "logps/rejected": -1.075714349746704, | |
| "loss": 24.4463, | |
| "nll_loss": 1.2817761898040771, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.4834820330142975, | |
| "rewards/margins": 0.05437516048550606, | |
| "rewards/rejected": -0.537857174873352, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 26.625, | |
| "learning_rate": 1.6580605550531018e-05, | |
| "log_odds_chosen": 0.12490881979465485, | |
| "log_odds_ratio": -0.6833258271217346, | |
| "logits/chosen": 267.01580810546875, | |
| "logits/rejected": 280.65118408203125, | |
| "logps/chosen": -0.866047739982605, | |
| "logps/rejected": -0.9292898178100586, | |
| "loss": 21.5112, | |
| "nll_loss": 1.1620063781738281, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.4330238699913025, | |
| "rewards/margins": 0.03162097930908203, | |
| "rewards/rejected": -0.4646449089050293, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.0906666666666665, | |
| "grad_norm": 25.0, | |
| "learning_rate": 1.525985589502466e-05, | |
| "log_odds_chosen": 0.490588515996933, | |
| "log_odds_ratio": -0.5364745259284973, | |
| "logits/chosen": 271.6618957519531, | |
| "logits/rejected": 270.92242431640625, | |
| "logps/chosen": -0.8149029016494751, | |
| "logps/rejected": -1.1035759449005127, | |
| "loss": 21.2697, | |
| "nll_loss": 1.204815149307251, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.40745145082473755, | |
| "rewards/margins": 0.1443365067243576, | |
| "rewards/rejected": -0.5517879724502563, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 25.75, | |
| "learning_rate": 1.3975752560252138e-05, | |
| "log_odds_chosen": 0.43111294507980347, | |
| "log_odds_ratio": -0.610099196434021, | |
| "logits/chosen": 256.387939453125, | |
| "logits/rejected": 280.3132629394531, | |
| "logps/chosen": -0.8138604164123535, | |
| "logps/rejected": -1.0904266834259033, | |
| "loss": 20.7732, | |
| "nll_loss": 1.0765711069107056, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.40693020820617676, | |
| "rewards/margins": 0.13828308880329132, | |
| "rewards/rejected": -0.5452133417129517, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 24.125, | |
| "learning_rate": 1.27314880294298e-05, | |
| "log_odds_chosen": 0.3809678852558136, | |
| "log_odds_ratio": -0.6075100898742676, | |
| "logits/chosen": 266.83233642578125, | |
| "logits/rejected": 259.447265625, | |
| "logps/chosen": -0.8320444822311401, | |
| "logps/rejected": -1.043336033821106, | |
| "loss": 20.9562, | |
| "nll_loss": 1.1581926345825195, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.41602224111557007, | |
| "rewards/margins": 0.1056457981467247, | |
| "rewards/rejected": -0.521668016910553, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.2186666666666666, | |
| "grad_norm": 23.5, | |
| "learning_rate": 1.1530155740230252e-05, | |
| "log_odds_chosen": 0.4367187023162842, | |
| "log_odds_ratio": -0.5616321563720703, | |
| "logits/chosen": 255.9156036376953, | |
| "logits/rejected": 277.2770080566406, | |
| "logps/chosen": -0.8188761472702026, | |
| "logps/rejected": -1.048285722732544, | |
| "loss": 20.5943, | |
| "nll_loss": 1.0858075618743896, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.4094380736351013, | |
| "rewards/margins": 0.11470470577478409, | |
| "rewards/rejected": -0.524142861366272, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.2613333333333334, | |
| "grad_norm": 19.0, | |
| "learning_rate": 1.0374742394008972e-05, | |
| "log_odds_chosen": 0.2701203525066376, | |
| "log_odds_ratio": -0.6517602205276489, | |
| "logits/chosen": 260.30401611328125, | |
| "logits/rejected": 264.9652404785156, | |
| "logps/chosen": -0.8311630487442017, | |
| "logps/rejected": -0.991308867931366, | |
| "loss": 20.3899, | |
| "nll_loss": 1.069040298461914, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.41558152437210083, | |
| "rewards/margins": 0.08007291704416275, | |
| "rewards/rejected": -0.495654433965683, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 26.125, | |
| "learning_rate": 9.268120530394061e-06, | |
| "log_odds_chosen": 0.31922250986099243, | |
| "log_odds_ratio": -0.6070750951766968, | |
| "logits/chosen": 260.91009521484375, | |
| "logits/rejected": 261.7254333496094, | |
| "logps/chosen": -0.7734104990959167, | |
| "logps/rejected": -0.9511035084724426, | |
| "loss": 20.48, | |
| "nll_loss": 1.0357019901275635, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.3867052495479584, | |
| "rewards/margins": 0.08884649723768234, | |
| "rewards/rejected": -0.4755517542362213, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.3466666666666667, | |
| "grad_norm": 30.625, | |
| "learning_rate": 8.213041385700211e-06, | |
| "log_odds_chosen": 0.3881288170814514, | |
| "log_odds_ratio": -0.5812792181968689, | |
| "logits/chosen": 275.2894287109375, | |
| "logits/rejected": 252.8758087158203, | |
| "logps/chosen": -0.8068667650222778, | |
| "logps/rejected": -1.0178911685943604, | |
| "loss": 20.1315, | |
| "nll_loss": 1.0528508424758911, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.4034333825111389, | |
| "rewards/margins": 0.10551220178604126, | |
| "rewards/rejected": -0.5089455842971802, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.389333333333333, | |
| "grad_norm": 20.625, | |
| "learning_rate": 7.212128052921661e-06, | |
| "log_odds_chosen": 0.43442073464393616, | |
| "log_odds_ratio": -0.5687755346298218, | |
| "logits/chosen": 260.5086975097656, | |
| "logits/rejected": 262.8999328613281, | |
| "logps/chosen": -0.7444295287132263, | |
| "logps/rejected": -0.9539780616760254, | |
| "loss": 19.7724, | |
| "nll_loss": 1.065710425376892, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.37221476435661316, | |
| "rewards/margins": 0.10477427393198013, | |
| "rewards/rejected": -0.4769890308380127, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 18.875, | |
| "learning_rate": 6.267868960309771e-06, | |
| "log_odds_chosen": 0.3951905071735382, | |
| "log_odds_ratio": -0.5774310231208801, | |
| "logits/chosen": 264.29644775390625, | |
| "logits/rejected": 254.9677276611328, | |
| "logps/chosen": -0.787185788154602, | |
| "logps/rejected": -1.0126456022262573, | |
| "loss": 19.8597, | |
| "nll_loss": 1.018532633781433, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.393592894077301, | |
| "rewards/margins": 0.11272994428873062, | |
| "rewards/rejected": -0.5063228011131287, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.474666666666667, | |
| "grad_norm": 21.875, | |
| "learning_rate": 5.382611684748257e-06, | |
| "log_odds_chosen": 0.35993748903274536, | |
| "log_odds_ratio": -0.5955245494842529, | |
| "logits/chosen": 249.52297973632812, | |
| "logits/rejected": 276.64947509765625, | |
| "logps/chosen": -0.7756280303001404, | |
| "logps/rejected": -0.9831158518791199, | |
| "loss": 20.2687, | |
| "nll_loss": 1.0755739212036133, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.3878140151500702, | |
| "rewards/margins": 0.10374389588832855, | |
| "rewards/rejected": -0.49155792593955994, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.517333333333333, | |
| "grad_norm": 18.375, | |
| "learning_rate": 4.558557115307222e-06, | |
| "log_odds_chosen": 0.3779729902744293, | |
| "log_odds_ratio": -0.6153554320335388, | |
| "logits/chosen": 262.4172668457031, | |
| "logits/rejected": 273.03375244140625, | |
| "logps/chosen": -0.7413235902786255, | |
| "logps/rejected": -0.9600993990898132, | |
| "loss": 19.9948, | |
| "nll_loss": 1.108370304107666, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.37066179513931274, | |
| "rewards/margins": 0.10938791930675507, | |
| "rewards/rejected": -0.4800496995449066, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 23.0, | |
| "learning_rate": 3.7977539814861106e-06, | |
| "log_odds_chosen": 0.326369047164917, | |
| "log_odds_ratio": -0.6079571843147278, | |
| "logits/chosen": 257.0772705078125, | |
| "logits/rejected": 261.3754577636719, | |
| "logps/chosen": -0.792614221572876, | |
| "logps/rejected": -0.960332989692688, | |
| "loss": 20.7289, | |
| "nll_loss": 1.0533356666564941, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.396307110786438, | |
| "rewards/margins": 0.0838593915104866, | |
| "rewards/rejected": -0.480166494846344, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.602666666666667, | |
| "grad_norm": 20.375, | |
| "learning_rate": 3.102093759749376e-06, | |
| "log_odds_chosen": 0.260172963142395, | |
| "log_odds_ratio": -0.6579862833023071, | |
| "logits/chosen": 261.3594055175781, | |
| "logits/rejected": 266.4259033203125, | |
| "logps/chosen": -0.7964383959770203, | |
| "logps/rejected": -0.9274827241897583, | |
| "loss": 20.2045, | |
| "nll_loss": 1.1274524927139282, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.39821919798851013, | |
| "rewards/margins": 0.06552214920520782, | |
| "rewards/rejected": -0.46374136209487915, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.6453333333333333, | |
| "grad_norm": 20.5, | |
| "learning_rate": 2.4733059710179828e-06, | |
| "log_odds_chosen": 0.4331514239311218, | |
| "log_odds_ratio": -0.5700831413269043, | |
| "logits/chosen": 265.8288879394531, | |
| "logits/rejected": 274.1119689941406, | |
| "logps/chosen": -0.7710822820663452, | |
| "logps/rejected": -0.9825912714004517, | |
| "loss": 20.6386, | |
| "nll_loss": 1.094036340713501, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.3855411410331726, | |
| "rewards/margins": 0.10575449466705322, | |
| "rewards/rejected": -0.49129563570022583, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 19.875, | |
| "learning_rate": 1.912953880807884e-06, | |
| "log_odds_chosen": 0.3509272634983063, | |
| "log_odds_ratio": -0.6254650354385376, | |
| "logits/chosen": 269.66375732421875, | |
| "logits/rejected": 276.72247314453125, | |
| "logps/chosen": -0.8052287101745605, | |
| "logps/rejected": -1.0146431922912598, | |
| "loss": 19.5416, | |
| "nll_loss": 1.0689141750335693, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.4026143550872803, | |
| "rewards/margins": 0.10470722615718842, | |
| "rewards/rejected": -0.5073215961456299, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.7306666666666666, | |
| "grad_norm": 22.375, | |
| "learning_rate": 1.422430612705613e-06, | |
| "log_odds_chosen": 0.2932564318180084, | |
| "log_odds_ratio": -0.6279724836349487, | |
| "logits/chosen": 263.7853088378906, | |
| "logits/rejected": 254.1911163330078, | |
| "logps/chosen": -0.8304440379142761, | |
| "logps/rejected": -0.9680387377738953, | |
| "loss": 20.6751, | |
| "nll_loss": 1.1476246118545532, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.41522201895713806, | |
| "rewards/margins": 0.06879737973213196, | |
| "rewards/rejected": -0.48401936888694763, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.7733333333333334, | |
| "grad_norm": 20.375, | |
| "learning_rate": 1.002955684843585e-06, | |
| "log_odds_chosen": 0.4094099998474121, | |
| "log_odds_ratio": -0.6279257535934448, | |
| "logits/chosen": 263.61419677734375, | |
| "logits/rejected": 278.83197021484375, | |
| "logps/chosen": -0.7544084787368774, | |
| "logps/rejected": -0.9667471051216125, | |
| "loss": 20.3327, | |
| "nll_loss": 1.0606472492218018, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.3772042393684387, | |
| "rewards/margins": 0.10616934299468994, | |
| "rewards/rejected": -0.4833735525608063, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 22.25, | |
| "learning_rate": 6.555719779858294e-07, | |
| "log_odds_chosen": 0.20543567836284637, | |
| "log_odds_ratio": -0.6976035833358765, | |
| "logits/chosen": 265.85736083984375, | |
| "logits/rejected": 258.0235900878906, | |
| "logps/chosen": -0.8123346567153931, | |
| "logps/rejected": -0.9113311767578125, | |
| "loss": 20.2874, | |
| "nll_loss": 1.0404599905014038, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.40616732835769653, | |
| "rewards/margins": 0.04949823394417763, | |
| "rewards/rejected": -0.45566558837890625, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.8586666666666667, | |
| "grad_norm": 23.5, | |
| "learning_rate": 3.8114314276213145e-07, | |
| "log_odds_chosen": 0.2348608523607254, | |
| "log_odds_ratio": -0.6606994867324829, | |
| "logits/chosen": 264.12615966796875, | |
| "logits/rejected": 273.71734619140625, | |
| "logps/chosen": -0.7940512895584106, | |
| "logps/rejected": -0.925014317035675, | |
| "loss": 20.1626, | |
| "nll_loss": 1.1582380533218384, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.3970256447792053, | |
| "rewards/margins": 0.0654815211892128, | |
| "rewards/rejected": -0.4625071585178375, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.9013333333333335, | |
| "grad_norm": 22.125, | |
| "learning_rate": 1.8035145249644225e-07, | |
| "log_odds_chosen": 0.23013488948345184, | |
| "log_odds_ratio": -0.6557679772377014, | |
| "logits/chosen": 261.1979064941406, | |
| "logits/rejected": 262.1890563964844, | |
| "logps/chosen": -0.8094332814216614, | |
| "logps/rejected": -0.9246999621391296, | |
| "loss": 19.9063, | |
| "nll_loss": 1.1390663385391235, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.4047166407108307, | |
| "rewards/margins": 0.05763337016105652, | |
| "rewards/rejected": -0.4623499810695648, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 21.875, | |
| "learning_rate": 5.369610696794536e-08, | |
| "log_odds_chosen": 0.22427129745483398, | |
| "log_odds_ratio": -0.6425634622573853, | |
| "logits/chosen": 272.1688232421875, | |
| "logits/rejected": 257.414306640625, | |
| "logps/chosen": -0.8942912817001343, | |
| "logps/rejected": -1.0066581964492798, | |
| "loss": 19.9408, | |
| "nll_loss": 1.096421241760254, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.44714564085006714, | |
| "rewards/margins": 0.056183360517024994, | |
| "rewards/rejected": -0.5033290982246399, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.986666666666667, | |
| "grad_norm": 26.5, | |
| "learning_rate": 1.4919913217092962e-09, | |
| "log_odds_chosen": 0.5936909914016724, | |
| "log_odds_ratio": -0.538439154624939, | |
| "logits/chosen": 274.3494567871094, | |
| "logits/rejected": 245.5052947998047, | |
| "logps/chosen": -0.7209577560424805, | |
| "logps/rejected": -1.0420339107513428, | |
| "loss": 19.7242, | |
| "nll_loss": 1.0608009099960327, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.36047887802124023, | |
| "rewards/margins": 0.16053801774978638, | |
| "rewards/rejected": -0.5210169553756714, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "step": 351, | |
| "total_flos": 0.0, | |
| "train_loss": 270.4568550620663, | |
| "train_runtime": 4053.6602, | |
| "train_samples_per_second": 5.551, | |
| "train_steps_per_second": 0.087 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 351, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100000, | |
| "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": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |