| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9703504043126685, | |
| "eval_steps": 500, | |
| "global_step": 276, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05390835579514825, | |
| "grad_norm": 502.0, | |
| "learning_rate": 8.92857142857143e-06, | |
| "log_odds_chosen": 5.912805557250977, | |
| "log_odds_ratio": -7.710684299468994, | |
| "logps/chosen": -24.25197982788086, | |
| "logps/rejected": -30.16664695739746, | |
| "loss": 186.4347, | |
| "nll_loss": 11.652168273925781, | |
| "rewards/accuracies": 0.484375, | |
| "rewards/chosen": -1.2125989198684692, | |
| "rewards/margins": 0.29573339223861694, | |
| "rewards/rejected": -1.5083322525024414, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.1078167115902965, | |
| "grad_norm": 164.0, | |
| "learning_rate": 1.785714285714286e-05, | |
| "log_odds_chosen": 4.08050537109375, | |
| "log_odds_ratio": -5.399485111236572, | |
| "logps/chosen": -20.047760009765625, | |
| "logps/rejected": -24.1300106048584, | |
| "loss": 155.9258, | |
| "nll_loss": 9.745362281799316, | |
| "rewards/accuracies": 0.4781250059604645, | |
| "rewards/chosen": -1.0023880004882812, | |
| "rewards/margins": 0.20411260426044464, | |
| "rewards/rejected": -1.2065006494522095, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16172506738544473, | |
| "grad_norm": 304.0, | |
| "learning_rate": 2.6785714285714288e-05, | |
| "log_odds_chosen": 3.733511447906494, | |
| "log_odds_ratio": -6.636049747467041, | |
| "logps/chosen": -21.219013214111328, | |
| "logps/rejected": -24.95370864868164, | |
| "loss": 155.6621, | |
| "nll_loss": 9.728882789611816, | |
| "rewards/accuracies": 0.49687498807907104, | |
| "rewards/chosen": -1.060950517654419, | |
| "rewards/margins": 0.18673481047153473, | |
| "rewards/rejected": -1.2476855516433716, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.215633423180593, | |
| "grad_norm": 540.0, | |
| "learning_rate": 3.571428571428572e-05, | |
| "log_odds_chosen": 0.06586956977844238, | |
| "log_odds_ratio": -5.591992378234863, | |
| "logps/chosen": -15.408183097839355, | |
| "logps/rejected": -15.473505020141602, | |
| "loss": 106.968, | |
| "nll_loss": 6.6854963302612305, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.7704092264175415, | |
| "rewards/margins": 0.0032660537399351597, | |
| "rewards/rejected": -0.7736751437187195, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2695417789757412, | |
| "grad_norm": 114.5, | |
| "learning_rate": 4.464285714285715e-05, | |
| "log_odds_chosen": 0.1850651204586029, | |
| "log_odds_ratio": -0.870222270488739, | |
| "logps/chosen": -2.379927158355713, | |
| "logps/rejected": -2.544480562210083, | |
| "loss": 38.8752, | |
| "nll_loss": 2.429699420928955, | |
| "rewards/accuracies": 0.5718749761581421, | |
| "rewards/chosen": -0.11899634450674057, | |
| "rewards/margins": 0.00822767335921526, | |
| "rewards/rejected": -0.12722402811050415, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.32345013477088946, | |
| "grad_norm": 119.0, | |
| "learning_rate": 4.999197688241076e-05, | |
| "log_odds_chosen": 0.2413506954908371, | |
| "log_odds_ratio": -0.7524750232696533, | |
| "logps/chosen": -1.8451862335205078, | |
| "logps/rejected": -2.054591655731201, | |
| "loss": 33.8684, | |
| "nll_loss": 2.1167733669281006, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": -0.09225932508707047, | |
| "rewards/margins": 0.010470272973179817, | |
| "rewards/rejected": -0.10272959619760513, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.37735849056603776, | |
| "grad_norm": 65.0, | |
| "learning_rate": 4.9901775939413026e-05, | |
| "log_odds_chosen": 0.25174349546432495, | |
| "log_odds_ratio": -0.7253037691116333, | |
| "logps/chosen": -1.6166225671768188, | |
| "logps/rejected": -1.8298925161361694, | |
| "loss": 29.1245, | |
| "nll_loss": 1.820279836654663, | |
| "rewards/accuracies": 0.5531250238418579, | |
| "rewards/chosen": -0.0808311253786087, | |
| "rewards/margins": 0.010663499124348164, | |
| "rewards/rejected": -0.091494619846344, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.431266846361186, | |
| "grad_norm": 101.5, | |
| "learning_rate": 4.971170810820279e-05, | |
| "log_odds_chosen": 0.2364540547132492, | |
| "log_odds_ratio": -0.7084470391273499, | |
| "logps/chosen": -1.5713794231414795, | |
| "logps/rejected": -1.7505134344100952, | |
| "loss": 29.3433, | |
| "nll_loss": 1.8339534997940063, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.07856898009777069, | |
| "rewards/margins": 0.008956688456237316, | |
| "rewards/rejected": -0.08752566576004028, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.48517520215633425, | |
| "grad_norm": 43.75, | |
| "learning_rate": 4.942253564296218e-05, | |
| "log_odds_chosen": 0.18266446888446808, | |
| "log_odds_ratio": -0.7139922976493835, | |
| "logps/chosen": -1.3910434246063232, | |
| "logps/rejected": -1.5377957820892334, | |
| "loss": 27.6051, | |
| "nll_loss": 1.725320816040039, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.06955216825008392, | |
| "rewards/margins": 0.007337613496929407, | |
| "rewards/rejected": -0.07688979059457779, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5390835579514824, | |
| "grad_norm": 35.75, | |
| "learning_rate": 4.9035418250305314e-05, | |
| "log_odds_chosen": 0.15111494064331055, | |
| "log_odds_ratio": -0.7071037292480469, | |
| "logps/chosen": -1.3401615619659424, | |
| "logps/rejected": -1.4416334629058838, | |
| "loss": 26.5156, | |
| "nll_loss": 1.6572233438491821, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.06700807809829712, | |
| "rewards/margins": 0.005073595326393843, | |
| "rewards/rejected": -0.07208167761564255, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5929919137466307, | |
| "grad_norm": 117.0, | |
| "learning_rate": 4.8551908438353374e-05, | |
| "log_odds_chosen": 0.20908907055854797, | |
| "log_odds_ratio": -0.6856271028518677, | |
| "logps/chosen": -1.2825366258621216, | |
| "logps/rejected": -1.4346181154251099, | |
| "loss": 25.8045, | |
| "nll_loss": 1.6127817630767822, | |
| "rewards/accuracies": 0.590624988079071, | |
| "rewards/chosen": -0.06412683427333832, | |
| "rewards/margins": 0.007604071404784918, | |
| "rewards/rejected": -0.07173089683055878, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6469002695417789, | |
| "grad_norm": 55.25, | |
| "learning_rate": 4.7973945290505766e-05, | |
| "log_odds_chosen": 0.13437321782112122, | |
| "log_odds_ratio": -0.7068942785263062, | |
| "logps/chosen": -1.2431588172912598, | |
| "logps/rejected": -1.3494141101837158, | |
| "loss": 24.9189, | |
| "nll_loss": 1.5574296712875366, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -0.06215794011950493, | |
| "rewards/margins": 0.005312758963555098, | |
| "rewards/rejected": -0.06747071444988251, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7008086253369272, | |
| "grad_norm": 115.5, | |
| "learning_rate": 4.73038466888773e-05, | |
| "log_odds_chosen": 0.11516331136226654, | |
| "log_odds_ratio": -0.7177757024765015, | |
| "logps/chosen": -1.2244114875793457, | |
| "logps/rejected": -1.3131043910980225, | |
| "loss": 24.9998, | |
| "nll_loss": 1.5624865293502808, | |
| "rewards/accuracies": 0.515625, | |
| "rewards/chosen": -0.061220575124025345, | |
| "rewards/margins": 0.004434647969901562, | |
| "rewards/rejected": -0.06565522402524948, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7547169811320755, | |
| "grad_norm": 85.5, | |
| "learning_rate": 4.654430001858874e-05, | |
| "log_odds_chosen": 0.13228605687618256, | |
| "log_odds_ratio": -0.7034366726875305, | |
| "logps/chosen": -1.1982498168945312, | |
| "logps/rejected": -1.2874435186386108, | |
| "loss": 24.5123, | |
| "nll_loss": 1.532017469406128, | |
| "rewards/accuracies": 0.565625011920929, | |
| "rewards/chosen": -0.05991249158978462, | |
| "rewards/margins": 0.004459693096578121, | |
| "rewards/rejected": -0.06437218189239502, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.8086253369272237, | |
| "grad_norm": 68.0, | |
| "learning_rate": 4.569835139019054e-05, | |
| "log_odds_chosen": 0.22792398929595947, | |
| "log_odds_ratio": -0.6663814783096313, | |
| "logps/chosen": -1.1594752073287964, | |
| "logps/rejected": -1.320555329322815, | |
| "loss": 24.2124, | |
| "nll_loss": 1.5132750272750854, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.05797375366091728, | |
| "rewards/margins": 0.00805400125682354, | |
| "rewards/rejected": -0.06602776050567627, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.862533692722372, | |
| "grad_norm": 70.5, | |
| "learning_rate": 4.476939342344246e-05, | |
| "log_odds_chosen": 0.21846911311149597, | |
| "log_odds_ratio": -0.6512280702590942, | |
| "logps/chosen": -1.0879645347595215, | |
| "logps/rejected": -1.2357169389724731, | |
| "loss": 23.5682, | |
| "nll_loss": 1.4730117321014404, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.054398227483034134, | |
| "rewards/margins": 0.007387618534266949, | |
| "rewards/rejected": -0.06178584694862366, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.9164420485175202, | |
| "grad_norm": 38.0, | |
| "learning_rate": 4.376115164144157e-05, | |
| "log_odds_chosen": 0.15174248814582825, | |
| "log_odds_ratio": -0.6834980845451355, | |
| "logps/chosen": -1.0666790008544922, | |
| "logps/rejected": -1.1634466648101807, | |
| "loss": 22.9808, | |
| "nll_loss": 1.4362987279891968, | |
| "rewards/accuracies": 0.578125, | |
| "rewards/chosen": -0.05333394929766655, | |
| "rewards/margins": 0.004838378168642521, | |
| "rewards/rejected": -0.058172326534986496, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9703504043126685, | |
| "grad_norm": 64.5, | |
| "learning_rate": 4.267766952966369e-05, | |
| "log_odds_chosen": 0.1273517906665802, | |
| "log_odds_ratio": -0.6930577158927917, | |
| "logps/chosen": -1.040971279144287, | |
| "logps/rejected": -1.1280105113983154, | |
| "loss": 22.4486, | |
| "nll_loss": 1.4030355215072632, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.05204857140779495, | |
| "rewards/margins": 0.004351964220404625, | |
| "rewards/rejected": -0.05640053004026413, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.0215633423180592, | |
| "grad_norm": 25.875, | |
| "learning_rate": 4.1523292319838524e-05, | |
| "log_odds_chosen": 0.21998043358325958, | |
| "log_odds_ratio": -0.6607629656791687, | |
| "logps/chosen": -0.9751634001731873, | |
| "logps/rejected": -1.1191428899765015, | |
| "loss": 20.5047, | |
| "nll_loss": 1.348992943763733, | |
| "rewards/accuracies": 0.5953947305679321, | |
| "rewards/chosen": -0.04875817149877548, | |
| "rewards/margins": 0.007198969833552837, | |
| "rewards/rejected": -0.0559571348130703, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.0754716981132075, | |
| "grad_norm": 39.0, | |
| "learning_rate": 4.030264956369157e-05, | |
| "log_odds_chosen": 0.4434036314487457, | |
| "log_odds_ratio": -0.6344673037528992, | |
| "logps/chosen": -0.9448977708816528, | |
| "logps/rejected": -1.1891727447509766, | |
| "loss": 21.1138, | |
| "nll_loss": 1.319615125656128, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.04724489524960518, | |
| "rewards/margins": 0.012213751673698425, | |
| "rewards/rejected": -0.059458643198013306, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.1293800539083558, | |
| "grad_norm": 29.625, | |
| "learning_rate": 3.902063656644012e-05, | |
| "log_odds_chosen": 0.5363696813583374, | |
| "log_odds_ratio": -0.5325912833213806, | |
| "logps/chosen": -0.8145742416381836, | |
| "logps/rejected": -1.1247040033340454, | |
| "loss": 18.4662, | |
| "nll_loss": 1.1541385650634766, | |
| "rewards/accuracies": 0.7281249761581421, | |
| "rewards/chosen": -0.04072871431708336, | |
| "rewards/margins": 0.015506483614444733, | |
| "rewards/rejected": -0.05623519420623779, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.1832884097035041, | |
| "grad_norm": 25.625, | |
| "learning_rate": 3.768239475450269e-05, | |
| "log_odds_chosen": 0.5174719095230103, | |
| "log_odds_ratio": -0.5490429401397705, | |
| "logps/chosen": -0.8520506620407104, | |
| "logps/rejected": -1.1713144779205322, | |
| "loss": 19.2575, | |
| "nll_loss": 1.20359206199646, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.0426025353372097, | |
| "rewards/margins": 0.01596318557858467, | |
| "rewards/rejected": -0.05856572464108467, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.2371967654986522, | |
| "grad_norm": 44.0, | |
| "learning_rate": 3.629329105615617e-05, | |
| "log_odds_chosen": 0.5610077977180481, | |
| "log_odds_ratio": -0.5220470428466797, | |
| "logps/chosen": -0.8756101727485657, | |
| "logps/rejected": -1.2160618305206299, | |
| "loss": 19.0513, | |
| "nll_loss": 1.1907049417495728, | |
| "rewards/accuracies": 0.765625, | |
| "rewards/chosen": -0.04378051310777664, | |
| "rewards/margins": 0.01702258363366127, | |
| "rewards/rejected": -0.060803093016147614, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.2911051212938005, | |
| "grad_norm": 39.5, | |
| "learning_rate": 3.4858896377832966e-05, | |
| "log_odds_chosen": 0.5247588157653809, | |
| "log_odds_ratio": -0.5250480771064758, | |
| "logps/chosen": -0.8190716505050659, | |
| "logps/rejected": -1.128647804260254, | |
| "loss": 17.8761, | |
| "nll_loss": 1.117258906364441, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.040953584015369415, | |
| "rewards/margins": 0.015478810295462608, | |
| "rewards/rejected": -0.05643239617347717, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.3450134770889488, | |
| "grad_norm": 45.25, | |
| "learning_rate": 3.338496326237743e-05, | |
| "log_odds_chosen": 0.4806355834007263, | |
| "log_odds_ratio": -0.550317108631134, | |
| "logps/chosen": -0.8094690442085266, | |
| "logps/rejected": -1.100089430809021, | |
| "loss": 18.3711, | |
| "nll_loss": 1.148194432258606, | |
| "rewards/accuracies": 0.746874988079071, | |
| "rewards/chosen": -0.04047344997525215, | |
| "rewards/margins": 0.01453101821243763, | |
| "rewards/rejected": -0.05500447005033493, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.398921832884097, | |
| "grad_norm": 35.0, | |
| "learning_rate": 3.187740281886195e-05, | |
| "log_odds_chosen": 0.6415280699729919, | |
| "log_odds_ratio": -0.48781052231788635, | |
| "logps/chosen": -0.8225423693656921, | |
| "logps/rejected": -1.2189260721206665, | |
| "loss": 18.4471, | |
| "nll_loss": 1.152944564819336, | |
| "rewards/accuracies": 0.784375011920929, | |
| "rewards/chosen": -0.04112711548805237, | |
| "rewards/margins": 0.01981918141245842, | |
| "rewards/rejected": -0.06094629690051079, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.4528301886792452, | |
| "grad_norm": 21.5, | |
| "learning_rate": 3.034226101648377e-05, | |
| "log_odds_chosen": 0.6234865784645081, | |
| "log_odds_ratio": -0.4981662333011627, | |
| "logps/chosen": -0.8078993558883667, | |
| "logps/rejected": -1.180347204208374, | |
| "loss": 18.1349, | |
| "nll_loss": 1.1334304809570312, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.040394969284534454, | |
| "rewards/margins": 0.01862238720059395, | |
| "rewards/rejected": -0.059017352759838104, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.5067385444743935, | |
| "grad_norm": 54.0, | |
| "learning_rate": 2.878569443761442e-05, | |
| "log_odds_chosen": 0.5453085899353027, | |
| "log_odds_ratio": -0.520926833152771, | |
| "logps/chosen": -0.8231021165847778, | |
| "logps/rejected": -1.1491215229034424, | |
| "loss": 18.0104, | |
| "nll_loss": 1.1256530284881592, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.04115510731935501, | |
| "rewards/margins": 0.016300970688462257, | |
| "rewards/rejected": -0.057456083595752716, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.5606469002695418, | |
| "grad_norm": 35.5, | |
| "learning_rate": 2.7213945587242508e-05, | |
| "log_odds_chosen": 0.4767599105834961, | |
| "log_odds_ratio": -0.5450612902641296, | |
| "logps/chosen": -0.8506708145141602, | |
| "logps/rejected": -1.1368257999420166, | |
| "loss": 19.6345, | |
| "nll_loss": 1.2271578311920166, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.04253353923559189, | |
| "rewards/margins": 0.014307747595012188, | |
| "rewards/rejected": -0.05684129148721695, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.61455525606469, | |
| "grad_norm": 24.75, | |
| "learning_rate": 2.5633317857829697e-05, | |
| "log_odds_chosen": 0.5109966993331909, | |
| "log_odds_ratio": -0.5330369472503662, | |
| "logps/chosen": -0.7959780693054199, | |
| "logps/rejected": -1.100124716758728, | |
| "loss": 17.9396, | |
| "nll_loss": 1.1212230920791626, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.03979891166090965, | |
| "rewards/margins": 0.01520733255892992, | |
| "rewards/rejected": -0.0550062358379364, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.6684636118598384, | |
| "grad_norm": 28.125, | |
| "learning_rate": 2.4050150249981522e-05, | |
| "log_odds_chosen": 0.5154935717582703, | |
| "log_odds_ratio": -0.5551471710205078, | |
| "logps/chosen": -0.7858830094337463, | |
| "logps/rejected": -1.0894118547439575, | |
| "loss": 17.763, | |
| "nll_loss": 1.1101869344711304, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.039294153451919556, | |
| "rewards/margins": 0.015176435932517052, | |
| "rewards/rejected": -0.05447059124708176, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.7223719676549867, | |
| "grad_norm": 25.625, | |
| "learning_rate": 2.24707919503142e-05, | |
| "log_odds_chosen": 0.612939715385437, | |
| "log_odds_ratio": -0.5060396790504456, | |
| "logps/chosen": -0.8042556643486023, | |
| "logps/rejected": -1.1593209505081177, | |
| "loss": 17.9033, | |
| "nll_loss": 1.1189591884613037, | |
| "rewards/accuracies": 0.765625, | |
| "rewards/chosen": -0.04021278768777847, | |
| "rewards/margins": 0.01775326207280159, | |
| "rewards/rejected": -0.057966046035289764, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.7762803234501348, | |
| "grad_norm": 30.5, | |
| "learning_rate": 2.0901576868471125e-05, | |
| "log_odds_chosen": 0.7125197649002075, | |
| "log_odds_ratio": -0.4848386347293854, | |
| "logps/chosen": -0.7776973843574524, | |
| "logps/rejected": -1.216399908065796, | |
| "loss": 17.609, | |
| "nll_loss": 1.1005606651306152, | |
| "rewards/accuracies": 0.8031250238418579, | |
| "rewards/chosen": -0.03888486698269844, | |
| "rewards/margins": 0.02193513885140419, | |
| "rewards/rejected": -0.06082000210881233, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.830188679245283, | |
| "grad_norm": 25.5, | |
| "learning_rate": 1.934879823540663e-05, | |
| "log_odds_chosen": 0.6370295286178589, | |
| "log_odds_ratio": -0.5132786631584167, | |
| "logps/chosen": -0.7767125368118286, | |
| "logps/rejected": -1.1403437852859497, | |
| "loss": 18.0793, | |
| "nll_loss": 1.1299545764923096, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.03883562982082367, | |
| "rewards/margins": 0.018181564286351204, | |
| "rewards/rejected": -0.05701719596982002, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.8840970350404311, | |
| "grad_norm": 24.125, | |
| "learning_rate": 1.7818683364808884e-05, | |
| "log_odds_chosen": 0.5283645391464233, | |
| "log_odds_ratio": -0.5327858328819275, | |
| "logps/chosen": -0.8068645596504211, | |
| "logps/rejected": -1.107076644897461, | |
| "loss": 17.5256, | |
| "nll_loss": 1.0953474044799805, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.04034322127699852, | |
| "rewards/margins": 0.015010610222816467, | |
| "rewards/rejected": -0.055353838950395584, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.9380053908355794, | |
| "grad_norm": 30.375, | |
| "learning_rate": 1.6317368678879495e-05, | |
| "log_odds_chosen": 0.4843871593475342, | |
| "log_odds_ratio": -0.5502706170082092, | |
| "logps/chosen": -0.8421553373336792, | |
| "logps/rejected": -1.1288360357284546, | |
| "loss": 18.9738, | |
| "nll_loss": 1.1858609914779663, | |
| "rewards/accuracies": 0.734375, | |
| "rewards/chosen": -0.04210776835680008, | |
| "rewards/margins": 0.014334036037325859, | |
| "rewards/rejected": -0.05644180253148079, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.9919137466307277, | |
| "grad_norm": 28.375, | |
| "learning_rate": 1.4850875098627326e-05, | |
| "log_odds_chosen": 0.5836633443832397, | |
| "log_odds_ratio": -0.5226461291313171, | |
| "logps/chosen": -0.799677312374115, | |
| "logps/rejected": -1.1371490955352783, | |
| "loss": 18.0322, | |
| "nll_loss": 1.1270129680633545, | |
| "rewards/accuracies": 0.753125011920929, | |
| "rewards/chosen": -0.03998386859893799, | |
| "rewards/margins": 0.01687358133494854, | |
| "rewards/rejected": -0.056857455521821976, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.0431266846361185, | |
| "grad_norm": 21.125, | |
| "learning_rate": 1.3425083897371981e-05, | |
| "log_odds_chosen": 0.8875333666801453, | |
| "log_odds_ratio": -0.4172805845737457, | |
| "logps/chosen": -0.6325610876083374, | |
| "logps/rejected": -1.107661247253418, | |
| "loss": 13.71, | |
| "nll_loss": 0.901972234249115, | |
| "rewards/accuracies": 0.8585526347160339, | |
| "rewards/chosen": -0.03162805363535881, | |
| "rewards/margins": 0.023755012080073357, | |
| "rewards/rejected": -0.055383067578077316, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.0970350404312668, | |
| "grad_norm": 29.5, | |
| "learning_rate": 1.204571311429496e-05, | |
| "log_odds_chosen": 1.1873928308486938, | |
| "log_odds_ratio": -0.3502134680747986, | |
| "logps/chosen": -0.6132036447525024, | |
| "logps/rejected": -1.2639284133911133, | |
| "loss": 14.8674, | |
| "nll_loss": 0.9292107820510864, | |
| "rewards/accuracies": 0.8843749761581421, | |
| "rewards/chosen": -0.030660182237625122, | |
| "rewards/margins": 0.032536230981349945, | |
| "rewards/rejected": -0.06319641321897507, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.150943396226415, | |
| "grad_norm": 27.25, | |
| "learning_rate": 1.0718294622630188e-05, | |
| "log_odds_chosen": 1.1184864044189453, | |
| "log_odds_ratio": -0.3575947880744934, | |
| "logps/chosen": -0.6175593137741089, | |
| "logps/rejected": -1.2147563695907593, | |
| "loss": 13.9253, | |
| "nll_loss": 0.8703301548957825, | |
| "rewards/accuracies": 0.8968750238418579, | |
| "rewards/chosen": -0.030877966433763504, | |
| "rewards/margins": 0.02985985204577446, | |
| "rewards/rejected": -0.06073781102895737, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.2048517520215634, | |
| "grad_norm": 25.125, | |
| "learning_rate": 9.448151944460657e-06, | |
| "log_odds_chosen": 1.1300182342529297, | |
| "log_odds_ratio": -0.3609935939311981, | |
| "logps/chosen": -0.5816246271133423, | |
| "logps/rejected": -1.1945774555206299, | |
| "loss": 13.5606, | |
| "nll_loss": 0.8475350141525269, | |
| "rewards/accuracies": 0.8968750238418579, | |
| "rewards/chosen": -0.029081230983138084, | |
| "rewards/margins": 0.03064764477312565, | |
| "rewards/rejected": -0.059728872030973434, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.2587601078167117, | |
| "grad_norm": 30.25, | |
| "learning_rate": 8.240378901093034e-06, | |
| "log_odds_chosen": 1.1773656606674194, | |
| "log_odds_ratio": -0.36282655596733093, | |
| "logps/chosen": -0.6218483448028564, | |
| "logps/rejected": -1.2632155418395996, | |
| "loss": 14.1748, | |
| "nll_loss": 0.8859266042709351, | |
| "rewards/accuracies": 0.8968750238418579, | |
| "rewards/chosen": -0.03109242022037506, | |
| "rewards/margins": 0.03206836059689522, | |
| "rewards/rejected": -0.06316077709197998, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.31266846361186, | |
| "grad_norm": 27.75, | |
| "learning_rate": 7.099819184631928e-06, | |
| "log_odds_chosen": 1.1673331260681152, | |
| "log_odds_ratio": -0.35059279203414917, | |
| "logps/chosen": -0.5780085325241089, | |
| "logps/rejected": -1.1937094926834106, | |
| "loss": 13.6113, | |
| "nll_loss": 0.8507078289985657, | |
| "rewards/accuracies": 0.890625, | |
| "rewards/chosen": -0.028900425881147385, | |
| "rewards/margins": 0.030785048380494118, | |
| "rewards/rejected": -0.05968547612428665, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.3665768194070083, | |
| "grad_norm": 21.75, | |
| "learning_rate": 6.031046932680229e-06, | |
| "log_odds_chosen": 1.1187890768051147, | |
| "log_odds_ratio": -0.36328989267349243, | |
| "logps/chosen": -0.5981144309043884, | |
| "logps/rejected": -1.2099401950836182, | |
| "loss": 13.4124, | |
| "nll_loss": 0.838273823261261, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.029905717819929123, | |
| "rewards/margins": 0.030591288581490517, | |
| "rewards/rejected": -0.06049700453877449, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.420485175202156, | |
| "grad_norm": 23.25, | |
| "learning_rate": 5.038348384069663e-06, | |
| "log_odds_chosen": 1.0922346115112305, | |
| "log_odds_ratio": -0.36317089200019836, | |
| "logps/chosen": -0.6168917417526245, | |
| "logps/rejected": -1.194858193397522, | |
| "loss": 13.8242, | |
| "nll_loss": 0.8640131950378418, | |
| "rewards/accuracies": 0.909375011920929, | |
| "rewards/chosen": -0.030844587832689285, | |
| "rewards/margins": 0.028898322954773903, | |
| "rewards/rejected": -0.059742916375398636, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.4743935309973044, | |
| "grad_norm": 24.875, | |
| "learning_rate": 4.125704689189819e-06, | |
| "log_odds_chosen": 1.099705696105957, | |
| "log_odds_ratio": -0.3621538579463959, | |
| "logps/chosen": -0.5992009043693542, | |
| "logps/rejected": -1.1859813928604126, | |
| "loss": 14.0204, | |
| "nll_loss": 0.8762725591659546, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.029960045590996742, | |
| "rewards/margins": 0.02933902107179165, | |
| "rewards/rejected": -0.05929907411336899, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.5283018867924527, | |
| "grad_norm": 30.75, | |
| "learning_rate": 3.296775943853789e-06, | |
| "log_odds_chosen": 1.1012507677078247, | |
| "log_odds_ratio": -0.37290987372398376, | |
| "logps/chosen": -0.600740909576416, | |
| "logps/rejected": -1.1732831001281738, | |
| "loss": 13.7312, | |
| "nll_loss": 0.8582011461257935, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.0300370454788208, | |
| "rewards/margins": 0.02862711250782013, | |
| "rewards/rejected": -0.05866416543722153, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 2.582210242587601, | |
| "grad_norm": 21.875, | |
| "learning_rate": 2.5548865107314607e-06, | |
| "log_odds_chosen": 1.1265591382980347, | |
| "log_odds_ratio": -0.36651021242141724, | |
| "logps/chosen": -0.6239336729049683, | |
| "logps/rejected": -1.243729591369629, | |
| "loss": 14.1878, | |
| "nll_loss": 0.8867388963699341, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -0.031196683645248413, | |
| "rewards/margins": 0.030989795923233032, | |
| "rewards/rejected": -0.06218648701906204, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.6361185983827493, | |
| "grad_norm": 24.0, | |
| "learning_rate": 1.9030116872178316e-06, | |
| "log_odds_chosen": 1.1169674396514893, | |
| "log_odds_ratio": -0.3630684018135071, | |
| "logps/chosen": -0.5876578092575073, | |
| "logps/rejected": -1.1631311178207397, | |
| "loss": 14.3677, | |
| "nll_loss": 0.8979824185371399, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -0.029382890090346336, | |
| "rewards/margins": 0.02877367101609707, | |
| "rewards/rejected": -0.058156561106443405, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.6900269541778976, | |
| "grad_norm": 30.625, | |
| "learning_rate": 1.3437657732040782e-06, | |
| "log_odds_chosen": 1.2451064586639404, | |
| "log_odds_ratio": -0.33962780237197876, | |
| "logps/chosen": -0.559921145439148, | |
| "logps/rejected": -1.2268016338348389, | |
| "loss": 13.2893, | |
| "nll_loss": 0.83058100938797, | |
| "rewards/accuracies": 0.921875, | |
| "rewards/chosen": -0.027996059507131577, | |
| "rewards/margins": 0.033344022929668427, | |
| "rewards/rejected": -0.061340074986219406, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.743935309973046, | |
| "grad_norm": 21.625, | |
| "learning_rate": 8.793915866046359e-07, | |
| "log_odds_chosen": 1.1345646381378174, | |
| "log_odds_ratio": -0.36994147300720215, | |
| "logps/chosen": -0.5653634071350098, | |
| "logps/rejected": -1.1617649793624878, | |
| "loss": 14.1254, | |
| "nll_loss": 0.8828363418579102, | |
| "rewards/accuracies": 0.871874988079071, | |
| "rewards/chosen": -0.02826816774904728, | |
| "rewards/margins": 0.02982008457183838, | |
| "rewards/rejected": -0.05808825045824051, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.797843665768194, | |
| "grad_norm": 20.0, | |
| "learning_rate": 5.117514686876379e-07, | |
| "log_odds_chosen": 0.9944526553153992, | |
| "log_odds_ratio": -0.3954170346260071, | |
| "logps/chosen": -0.6244685649871826, | |
| "logps/rejected": -1.1500699520111084, | |
| "loss": 14.1807, | |
| "nll_loss": 0.8862916231155396, | |
| "rewards/accuracies": 0.8968750238418579, | |
| "rewards/chosen": -0.03122343122959137, | |
| "rewards/margins": 0.02628006599843502, | |
| "rewards/rejected": -0.05750349164009094, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.8517520215633425, | |
| "grad_norm": 25.375, | |
| "learning_rate": 2.423198152812306e-07, | |
| "log_odds_chosen": 0.9933083653450012, | |
| "log_odds_ratio": -0.3899889588356018, | |
| "logps/chosen": -0.6133357286453247, | |
| "logps/rejected": -1.120845913887024, | |
| "loss": 13.8159, | |
| "nll_loss": 0.8634947538375854, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -0.030666787177324295, | |
| "rewards/margins": 0.02537550963461399, | |
| "rewards/rejected": -0.05604229494929314, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.9056603773584904, | |
| "grad_norm": 25.0, | |
| "learning_rate": 7.217716380881479e-08, | |
| "log_odds_chosen": 1.142899751663208, | |
| "log_odds_ratio": -0.3680952191352844, | |
| "logps/chosen": -0.5903482437133789, | |
| "logps/rejected": -1.1948912143707275, | |
| "loss": 13.7773, | |
| "nll_loss": 0.8610836863517761, | |
| "rewards/accuracies": 0.8843749761581421, | |
| "rewards/chosen": -0.029517415910959244, | |
| "rewards/margins": 0.03022714890539646, | |
| "rewards/rejected": -0.05974455922842026, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.9595687331536387, | |
| "grad_norm": 25.75, | |
| "learning_rate": 2.0058598667854756e-09, | |
| "log_odds_chosen": 1.1750845909118652, | |
| "log_odds_ratio": -0.3487832844257355, | |
| "logps/chosen": -0.562160313129425, | |
| "logps/rejected": -1.1947548389434814, | |
| "loss": 13.3099, | |
| "nll_loss": 0.8318702578544617, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.028108015656471252, | |
| "rewards/margins": 0.03162972629070282, | |
| "rewards/rejected": -0.05973774194717407, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.9703504043126685, | |
| "step": 276, | |
| "total_flos": 0.0, | |
| "train_loss": 28.78522514605868, | |
| "train_runtime": 3021.362, | |
| "train_samples_per_second": 5.894, | |
| "train_steps_per_second": 0.091 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 276, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "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 | |
| } | |