| { | |
| "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": 33.21111681571131, | |
| "learning_rate": 2.0833333333333334e-06, | |
| "logits/chosen": -0.4980102479457855, | |
| "logits/rejected": -0.5135027170181274, | |
| "logps/chosen": -1.1746745109558105, | |
| "logps/rejected": -1.3606590032577515, | |
| "loss": 2.1734, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.1746745109558105, | |
| "rewards/margins": 0.1859845519065857, | |
| "rewards/rejected": -1.3606590032577515, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 21.16742924169967, | |
| "learning_rate": 4.166666666666667e-06, | |
| "logits/chosen": -0.5296765565872192, | |
| "logits/rejected": -0.5027884244918823, | |
| "logps/chosen": -1.1314122676849365, | |
| "logps/rejected": -1.2633330821990967, | |
| "loss": 2.1306, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -1.1314122676849365, | |
| "rewards/margins": 0.13192060589790344, | |
| "rewards/rejected": -1.2633330821990967, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 18.622273155389507, | |
| "learning_rate": 6.25e-06, | |
| "logits/chosen": -0.45581430196762085, | |
| "logits/rejected": -0.42932063341140747, | |
| "logps/chosen": -1.1560032367706299, | |
| "logps/rejected": -1.4923290014266968, | |
| "loss": 2.0523, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.1560032367706299, | |
| "rewards/margins": 0.3363257944583893, | |
| "rewards/rejected": -1.4923290014266968, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 25.84825355543498, | |
| "learning_rate": 8.333333333333334e-06, | |
| "logits/chosen": -0.6032270789146423, | |
| "logits/rejected": -0.5604568719863892, | |
| "logps/chosen": -1.2145692110061646, | |
| "logps/rejected": -1.5209157466888428, | |
| "loss": 2.101, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.2145692110061646, | |
| "rewards/margins": 0.30634641647338867, | |
| "rewards/rejected": -1.5209157466888428, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 10.051572353875851, | |
| "learning_rate": 1.0416666666666668e-05, | |
| "logits/chosen": -0.7330023646354675, | |
| "logits/rejected": -0.6652411222457886, | |
| "logps/chosen": -1.3188468217849731, | |
| "logps/rejected": -1.643450140953064, | |
| "loss": 2.0473, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.3188468217849731, | |
| "rewards/margins": 0.32460346817970276, | |
| "rewards/rejected": -1.643450140953064, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 12.836119680084701, | |
| "learning_rate": 1.25e-05, | |
| "logits/chosen": -0.7389785051345825, | |
| "logits/rejected": -0.7157658338546753, | |
| "logps/chosen": -1.2610353231430054, | |
| "logps/rejected": -1.5368638038635254, | |
| "loss": 2.1476, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.2610353231430054, | |
| "rewards/margins": 0.27582842111587524, | |
| "rewards/rejected": -1.5368638038635254, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 10.916144118237353, | |
| "learning_rate": 1.4583333333333333e-05, | |
| "logits/chosen": -0.6624680757522583, | |
| "logits/rejected": -0.5841827392578125, | |
| "logps/chosen": -1.3438886404037476, | |
| "logps/rejected": -1.5585218667984009, | |
| "loss": 2.1252, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.3438886404037476, | |
| "rewards/margins": 0.2146332710981369, | |
| "rewards/rejected": -1.5585218667984009, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 7.904056592473059, | |
| "learning_rate": 1.6666666666666667e-05, | |
| "logits/chosen": -0.8896552920341492, | |
| "logits/rejected": -0.7669180631637573, | |
| "logps/chosen": -1.3083586692810059, | |
| "logps/rejected": -1.7862266302108765, | |
| "loss": 2.0664, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.3083586692810059, | |
| "rewards/margins": 0.47786790132522583, | |
| "rewards/rejected": -1.7862266302108765, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 13.316362762434997, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "logits/chosen": -0.7929601669311523, | |
| "logits/rejected": -0.752467930316925, | |
| "logps/chosen": -1.2723389863967896, | |
| "logps/rejected": -1.6567331552505493, | |
| "loss": 2.0997, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.2723389863967896, | |
| "rewards/margins": 0.38439422845840454, | |
| "rewards/rejected": -1.6567331552505493, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 16.98219621825263, | |
| "learning_rate": 1.9998927475076107e-05, | |
| "logits/chosen": -0.3519185483455658, | |
| "logits/rejected": -0.30840247869491577, | |
| "logps/chosen": -1.275742769241333, | |
| "logps/rejected": -1.7419742345809937, | |
| "loss": 2.1089, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.275742769241333, | |
| "rewards/margins": 0.4662315845489502, | |
| "rewards/rejected": -1.7419742345809937, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 7.79923910311647, | |
| "learning_rate": 1.998686421164407e-05, | |
| "logits/chosen": -0.13412383198738098, | |
| "logits/rejected": -0.06430118530988693, | |
| "logps/chosen": -1.3077303171157837, | |
| "logps/rejected": -1.7474453449249268, | |
| "loss": 2.0751, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.3077303171157837, | |
| "rewards/margins": 0.43971508741378784, | |
| "rewards/rejected": -1.7474453449249268, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 7.005978339387122, | |
| "learning_rate": 1.9961413253717214e-05, | |
| "logits/chosen": -0.4779301583766937, | |
| "logits/rejected": -0.4137405455112457, | |
| "logps/chosen": -1.3986326456069946, | |
| "logps/rejected": -1.6036043167114258, | |
| "loss": 2.1009, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.3986326456069946, | |
| "rewards/margins": 0.20497193932533264, | |
| "rewards/rejected": -1.6036043167114258, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 6.982350763810334, | |
| "learning_rate": 1.9922608719076874e-05, | |
| "logits/chosen": -0.267805278301239, | |
| "logits/rejected": -0.1766107976436615, | |
| "logps/chosen": -1.2244327068328857, | |
| "logps/rejected": -2.0722804069519043, | |
| "loss": 2.0512, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.2244327068328857, | |
| "rewards/margins": 0.8478477597236633, | |
| "rewards/rejected": -2.0722804069519043, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 10.132344121190984, | |
| "learning_rate": 1.9870502626379127e-05, | |
| "logits/chosen": -0.35906368494033813, | |
| "logits/rejected": -0.33368802070617676, | |
| "logps/chosen": -1.450307846069336, | |
| "logps/rejected": -1.7906442880630493, | |
| "loss": 2.1396, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.450307846069336, | |
| "rewards/margins": 0.34033653140068054, | |
| "rewards/rejected": -1.7906442880630493, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 13.919238096447465, | |
| "learning_rate": 1.980516482542224e-05, | |
| "logits/chosen": -0.6731249094009399, | |
| "logits/rejected": -0.6837888956069946, | |
| "logps/chosen": -1.2502187490463257, | |
| "logps/rejected": -1.7363303899765015, | |
| "loss": 2.078, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.2502187490463257, | |
| "rewards/margins": 0.4861116409301758, | |
| "rewards/rejected": -1.7363303899765015, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 8.714846434947708, | |
| "learning_rate": 1.972668290351084e-05, | |
| "logits/chosen": -0.8093172311782837, | |
| "logits/rejected": -0.8910678029060364, | |
| "logps/chosen": -1.3465213775634766, | |
| "logps/rejected": -1.8265445232391357, | |
| "loss": 2.1277, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.3465213775634766, | |
| "rewards/margins": 0.480023056268692, | |
| "rewards/rejected": -1.8265445232391357, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 34.11378786664511, | |
| "learning_rate": 1.9635162068042547e-05, | |
| "logits/chosen": -0.6499379873275757, | |
| "logits/rejected": -0.6738103628158569, | |
| "logps/chosen": -1.2838003635406494, | |
| "logps/rejected": -1.6559721231460571, | |
| "loss": 2.1205, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.2838003635406494, | |
| "rewards/margins": 0.37217170000076294, | |
| "rewards/rejected": -1.6559721231460571, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 6.6135900227176, | |
| "learning_rate": 1.9530725005474195e-05, | |
| "logits/chosen": -0.1341579109430313, | |
| "logits/rejected": -0.1497870236635208, | |
| "logps/chosen": -1.3539865016937256, | |
| "logps/rejected": -1.7489871978759766, | |
| "loss": 2.0639, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.3539865016937256, | |
| "rewards/margins": 0.39500072598457336, | |
| "rewards/rejected": -1.7489871978759766, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 7.360183689725487, | |
| "learning_rate": 1.9413511716856973e-05, | |
| "logits/chosen": -0.12092798948287964, | |
| "logits/rejected": -0.07471726834774017, | |
| "logps/chosen": -1.3030513525009155, | |
| "logps/rejected": -1.8159534931182861, | |
| "loss": 2.0725, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.3030513525009155, | |
| "rewards/margins": 0.512902021408081, | |
| "rewards/rejected": -1.8159534931182861, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 7.942085572142913, | |
| "learning_rate": 1.9283679330160726e-05, | |
| "logits/chosen": 0.026644444093108177, | |
| "logits/rejected": 0.05488858371973038, | |
| "logps/chosen": -1.3510897159576416, | |
| "logps/rejected": -1.8336998224258423, | |
| "loss": 2.0911, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.3510897159576416, | |
| "rewards/margins": 0.482610285282135, | |
| "rewards/rejected": -1.8336998224258423, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 10.38696834889441, | |
| "learning_rate": 1.9141401889639167e-05, | |
| "logits/chosen": 0.12454743683338165, | |
| "logits/rejected": 0.1521395593881607, | |
| "logps/chosen": -1.308062195777893, | |
| "logps/rejected": -1.873884916305542, | |
| "loss": 2.029, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.308062195777893, | |
| "rewards/margins": 0.5658227205276489, | |
| "rewards/rejected": -1.873884916305542, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 14.10578954487523, | |
| "learning_rate": 1.898687012251826e-05, | |
| "logits/chosen": -0.14296935498714447, | |
| "logits/rejected": -0.08335347473621368, | |
| "logps/chosen": -1.3113409280776978, | |
| "logps/rejected": -1.7755804061889648, | |
| "loss": 2.0509, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.3113409280776978, | |
| "rewards/margins": 0.4642394483089447, | |
| "rewards/rejected": -1.7755804061889648, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 7.096253964138347, | |
| "learning_rate": 1.8820291183320602e-05, | |
| "logits/chosen": -0.20576635003089905, | |
| "logits/rejected": -0.1285274177789688, | |
| "logps/chosen": -1.2730509042739868, | |
| "logps/rejected": -1.8806079626083374, | |
| "loss": 2.0506, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.2730509042739868, | |
| "rewards/margins": 0.6075571179389954, | |
| "rewards/rejected": -1.8806079626083374, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 8.976667698418499, | |
| "learning_rate": 1.8641888376168483e-05, | |
| "logits/chosen": -0.10974551737308502, | |
| "logits/rejected": -0.07689039409160614, | |
| "logps/chosen": -1.442338466644287, | |
| "logps/rejected": -1.90249764919281, | |
| "loss": 2.1387, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.442338466644287, | |
| "rewards/margins": 0.46015921235084534, | |
| "rewards/rejected": -1.90249764919281, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 11.84902364416852, | |
| "learning_rate": 1.845190085543795e-05, | |
| "logits/chosen": 0.1279464215040207, | |
| "logits/rejected": 0.1599569022655487, | |
| "logps/chosen": -1.29521906375885, | |
| "logps/rejected": -1.5512622594833374, | |
| "loss": 2.0874, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.29521906375885, | |
| "rewards/margins": 0.2560431957244873, | |
| "rewards/rejected": -1.5512622594833374, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 12.665702263586864, | |
| "learning_rate": 1.8250583305165098e-05, | |
| "logits/chosen": 0.10071973502635956, | |
| "logits/rejected": 0.114678755402565, | |
| "logps/chosen": -1.3293492794036865, | |
| "logps/rejected": -1.6235164403915405, | |
| "loss": 2.105, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.3293492794036865, | |
| "rewards/margins": 0.29416733980178833, | |
| "rewards/rejected": -1.6235164403915405, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 11.168921888078422, | |
| "learning_rate": 1.8038205597634392e-05, | |
| "logits/chosen": -0.2312246859073639, | |
| "logits/rejected": -0.13947580754756927, | |
| "logps/chosen": -1.3103423118591309, | |
| "logps/rejected": -1.973184585571289, | |
| "loss": 2.0983, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -1.3103423118591309, | |
| "rewards/margins": 0.662842333316803, | |
| "rewards/rejected": -1.973184585571289, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 15.647952191048011, | |
| "learning_rate": 1.7815052431606702e-05, | |
| "logits/chosen": -0.27144142985343933, | |
| "logits/rejected": -0.2118106335401535, | |
| "logps/chosen": -1.3751564025878906, | |
| "logps/rejected": -2.03005051612854, | |
| "loss": 2.0429, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.3751564025878906, | |
| "rewards/margins": 0.6548939943313599, | |
| "rewards/rejected": -2.03005051612854, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 7.242840211884775, | |
| "learning_rate": 1.7581422950671942e-05, | |
| "logits/chosen": -0.19757069647312164, | |
| "logits/rejected": -0.1668623834848404, | |
| "logps/chosen": -1.3345425128936768, | |
| "logps/rejected": -1.8127644062042236, | |
| "loss": 2.0876, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3345425128936768, | |
| "rewards/margins": 0.4782216548919678, | |
| "rewards/rejected": -1.8127644062042236, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 8.615419856166682, | |
| "learning_rate": 1.733763034223804e-05, | |
| "logits/chosen": -0.21767687797546387, | |
| "logits/rejected": -0.21838533878326416, | |
| "logps/chosen": -1.2229845523834229, | |
| "logps/rejected": -1.660559058189392, | |
| "loss": 2.0294, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.2229845523834229, | |
| "rewards/margins": 0.43757471442222595, | |
| "rewards/rejected": -1.660559058189392, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 11.4467193943517, | |
| "learning_rate": 1.7084001417693702e-05, | |
| "logits/chosen": -0.17819705605506897, | |
| "logits/rejected": -0.1267833411693573, | |
| "logps/chosen": -1.389460563659668, | |
| "logps/rejected": -1.8192943334579468, | |
| "loss": 2.084, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.389460563659668, | |
| "rewards/margins": 0.4298337399959564, | |
| "rewards/rejected": -1.8192943334579468, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 8.12988482306829, | |
| "learning_rate": 1.682087617430782e-05, | |
| "logits/chosen": -0.12651406228542328, | |
| "logits/rejected": -0.04694231227040291, | |
| "logps/chosen": -1.318313479423523, | |
| "logps/rejected": -1.8056846857070923, | |
| "loss": 2.0818, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.318313479423523, | |
| "rewards/margins": 0.4873710572719574, | |
| "rewards/rejected": -1.8056846857070923, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 6.762373951814964, | |
| "learning_rate": 1.6548607339452853e-05, | |
| "logits/chosen": -0.10803677886724472, | |
| "logits/rejected": -0.048906028270721436, | |
| "logps/chosen": -1.25996994972229, | |
| "logps/rejected": -1.8231735229492188, | |
| "loss": 2.0354, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.25996994972229, | |
| "rewards/margins": 0.5632035732269287, | |
| "rewards/rejected": -1.8231735229492188, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 11.050433248891123, | |
| "learning_rate": 1.626755989776303e-05, | |
| "logits/chosen": -0.1651381254196167, | |
| "logits/rejected": -0.04633602499961853, | |
| "logps/chosen": -1.4237867593765259, | |
| "logps/rejected": -2.101548671722412, | |
| "loss": 2.0616, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.4237867593765259, | |
| "rewards/margins": 0.6777619123458862, | |
| "rewards/rejected": -2.101548671722412, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 7.0268134591588245, | |
| "learning_rate": 1.5978110601861408e-05, | |
| "logits/chosen": -0.12373347580432892, | |
| "logits/rejected": -0.0877654105424881, | |
| "logps/chosen": -1.3757156133651733, | |
| "logps/rejected": -1.7569023370742798, | |
| "loss": 2.072, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.3757156133651733, | |
| "rewards/margins": 0.3811867833137512, | |
| "rewards/rejected": -1.7569023370742798, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 12.761833101713918, | |
| "learning_rate": 1.568064746731156e-05, | |
| "logits/chosen": -0.11597935855388641, | |
| "logits/rejected": -0.1455441117286682, | |
| "logps/chosen": -1.374710202217102, | |
| "logps/rejected": -1.7882392406463623, | |
| "loss": 2.0783, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.374710202217102, | |
| "rewards/margins": 0.4135288596153259, | |
| "rewards/rejected": -1.7882392406463623, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 7.06376095255915, | |
| "learning_rate": 1.5375569252470897e-05, | |
| "logits/chosen": -0.16596433520317078, | |
| "logits/rejected": -0.026778871193528175, | |
| "logps/chosen": -1.3514513969421387, | |
| "logps/rejected": -2.0585455894470215, | |
| "loss": 2.0175, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.3514513969421387, | |
| "rewards/margins": 0.7070940732955933, | |
| "rewards/rejected": -2.0585455894470215, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 8.565174444883986, | |
| "learning_rate": 1.506328492394303e-05, | |
| "logits/chosen": -0.1680208444595337, | |
| "logits/rejected": -0.10585353523492813, | |
| "logps/chosen": -1.3384554386138916, | |
| "logps/rejected": -1.7696669101715088, | |
| "loss": 2.1269, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.3384554386138916, | |
| "rewards/margins": 0.43121138215065, | |
| "rewards/rejected": -1.7696669101715088, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 8.66730844602624, | |
| "learning_rate": 1.4744213108345605e-05, | |
| "logits/chosen": -0.18466773629188538, | |
| "logits/rejected": -0.03730706498026848, | |
| "logps/chosen": -1.3560011386871338, | |
| "logps/rejected": -1.740012764930725, | |
| "loss": 2.0877, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.3560011386871338, | |
| "rewards/margins": 0.3840116560459137, | |
| "rewards/rejected": -1.740012764930725, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 7.228970778278195, | |
| "learning_rate": 1.4418781531128636e-05, | |
| "logits/chosen": -0.0062202452681958675, | |
| "logits/rejected": 0.13902577757835388, | |
| "logps/chosen": -1.3838578462600708, | |
| "logps/rejected": -1.9292205572128296, | |
| "loss": 2.0565, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.3838578462600708, | |
| "rewards/margins": 0.5453627705574036, | |
| "rewards/rejected": -1.9292205572128296, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 9.000136464867888, | |
| "learning_rate": 1.4087426443195549e-05, | |
| "logits/chosen": 0.13322147727012634, | |
| "logits/rejected": 0.31764692068099976, | |
| "logps/chosen": -1.2240257263183594, | |
| "logps/rejected": -1.729107141494751, | |
| "loss": 2.0412, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.2240257263183594, | |
| "rewards/margins": 0.5050811171531677, | |
| "rewards/rejected": -1.729107141494751, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 9.154719134241198, | |
| "learning_rate": 1.375059203609562e-05, | |
| "logits/chosen": 0.16319788992404938, | |
| "logits/rejected": 0.3346864581108093, | |
| "logps/chosen": -1.4042994976043701, | |
| "logps/rejected": -1.8507611751556396, | |
| "loss": 2.1446, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.4042994976043701, | |
| "rewards/margins": 0.44646158814430237, | |
| "rewards/rejected": -1.8507611751556396, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 6.0840861971289115, | |
| "learning_rate": 1.3408729846571716e-05, | |
| "logits/chosen": 0.09617350250482559, | |
| "logits/rejected": 0.3308163285255432, | |
| "logps/chosen": -1.284582495689392, | |
| "logps/rejected": -1.9165366888046265, | |
| "loss": 2.0609, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.284582495689392, | |
| "rewards/margins": 0.6319543123245239, | |
| "rewards/rejected": -1.9165366888046265, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 8.00813078143392, | |
| "learning_rate": 1.3062298151261592e-05, | |
| "logits/chosen": 0.044529713690280914, | |
| "logits/rejected": 0.3042605519294739, | |
| "logps/chosen": -1.358139157295227, | |
| "logps/rejected": -1.929091215133667, | |
| "loss": 2.0609, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.358139157295227, | |
| "rewards/margins": 0.5709521770477295, | |
| "rewards/rejected": -1.929091215133667, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 7.246210126735172, | |
| "learning_rate": 1.2711761352364172e-05, | |
| "logits/chosen": 0.03733745217323303, | |
| "logits/rejected": 0.2569599449634552, | |
| "logps/chosen": -1.2875401973724365, | |
| "logps/rejected": -2.0033631324768066, | |
| "loss": 1.9734, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2875401973724365, | |
| "rewards/margins": 0.7158228754997253, | |
| "rewards/rejected": -2.0033631324768066, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 9.113546871891296, | |
| "learning_rate": 1.2357589355094275e-05, | |
| "logits/chosen": 0.014212149195373058, | |
| "logits/rejected": 0.32842034101486206, | |
| "logps/chosen": -1.3232357501983643, | |
| "logps/rejected": -2.133357524871826, | |
| "loss": 2.0056, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.3232357501983643, | |
| "rewards/margins": 0.8101218342781067, | |
| "rewards/rejected": -2.133357524871826, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 8.416270817600122, | |
| "learning_rate": 1.2000256937760446e-05, | |
| "logits/chosen": 0.1557755172252655, | |
| "logits/rejected": 0.4363393187522888, | |
| "logps/chosen": -1.2627068758010864, | |
| "logps/rejected": -1.9027000665664673, | |
| "loss": 2.0341, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.2627068758010864, | |
| "rewards/margins": 0.6399933099746704, | |
| "rewards/rejected": -1.9027000665664673, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 8.509607776673544, | |
| "learning_rate": 1.1640243115310219e-05, | |
| "logits/chosen": 0.1399160474538803, | |
| "logits/rejected": 0.41544660925865173, | |
| "logps/chosen": -1.2238709926605225, | |
| "logps/rejected": -1.8552753925323486, | |
| "loss": 2.043, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.2238709926605225, | |
| "rewards/margins": 0.6314042210578918, | |
| "rewards/rejected": -1.8552753925323486, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 7.675820182938368, | |
| "learning_rate": 1.127803049719605e-05, | |
| "logits/chosen": 0.1161346435546875, | |
| "logits/rejected": 0.3014758825302124, | |
| "logps/chosen": -1.3739269971847534, | |
| "logps/rejected": -1.896535873413086, | |
| "loss": 2.0516, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3739269971847534, | |
| "rewards/margins": 0.5226086378097534, | |
| "rewards/rejected": -1.896535873413086, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 8.105295156815169, | |
| "learning_rate": 1.091410464042268e-05, | |
| "logits/chosen": 0.13467064499855042, | |
| "logits/rejected": 0.20409516990184784, | |
| "logps/chosen": -1.283080816268921, | |
| "logps/rejected": -1.9848954677581787, | |
| "loss": 2.0297, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.283080816268921, | |
| "rewards/margins": 0.701814591884613, | |
| "rewards/rejected": -1.9848954677581787, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 8.059542189088054, | |
| "learning_rate": 1.0548953398643276e-05, | |
| "logits/chosen": 0.16154329478740692, | |
| "logits/rejected": 0.32910025119781494, | |
| "logps/chosen": -1.3794437646865845, | |
| "logps/rejected": -2.083132743835449, | |
| "loss": 2.0164, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.3794437646865845, | |
| "rewards/margins": 0.70368891954422, | |
| "rewards/rejected": -2.083132743835449, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 6.4999384094249955, | |
| "learning_rate": 1.0183066268176775e-05, | |
| "logits/chosen": 0.6510103940963745, | |
| "logits/rejected": 1.0612311363220215, | |
| "logps/chosen": -1.3178019523620605, | |
| "logps/rejected": -2.1374964714050293, | |
| "loss": 2.0503, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.3178019523620605, | |
| "rewards/margins": 0.8196946978569031, | |
| "rewards/rejected": -2.1374964714050293, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 7.313206322781419, | |
| "learning_rate": 9.81693373182323e-06, | |
| "logits/chosen": 0.484092652797699, | |
| "logits/rejected": 0.6402750015258789, | |
| "logps/chosen": -1.3769603967666626, | |
| "logps/rejected": -1.7423893213272095, | |
| "loss": 2.0132, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.3769603967666626, | |
| "rewards/margins": 0.365428626537323, | |
| "rewards/rejected": -1.7423893213272095, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 8.756069730601565, | |
| "learning_rate": 9.451046601356725e-06, | |
| "logits/chosen": 0.13520203530788422, | |
| "logits/rejected": 0.3411861062049866, | |
| "logps/chosen": -1.3921834230422974, | |
| "logps/rejected": -1.8440046310424805, | |
| "loss": 2.0618, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.3921834230422974, | |
| "rewards/margins": 0.4518211781978607, | |
| "rewards/rejected": -1.8440046310424805, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 6.597441645495227, | |
| "learning_rate": 9.085895359577324e-06, | |
| "logits/chosen": -0.17431692779064178, | |
| "logits/rejected": 0.03041163645684719, | |
| "logps/chosen": -1.317625641822815, | |
| "logps/rejected": -2.10213041305542, | |
| "loss": 2.1302, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.317625641822815, | |
| "rewards/margins": 0.7845045328140259, | |
| "rewards/rejected": -2.10213041305542, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 7.014450023153696, | |
| "learning_rate": 8.721969502803954e-06, | |
| "logits/chosen": -0.2185564786195755, | |
| "logits/rejected": -0.061094462871551514, | |
| "logps/chosen": -1.2835057973861694, | |
| "logps/rejected": -1.7011182308197021, | |
| "loss": 2.0526, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2835057973861694, | |
| "rewards/margins": 0.41761231422424316, | |
| "rewards/rejected": -1.7011182308197021, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 8.176491560294703, | |
| "learning_rate": 8.359756884689785e-06, | |
| "logits/chosen": -0.24773511290550232, | |
| "logits/rejected": -0.19381779432296753, | |
| "logps/chosen": -1.4067161083221436, | |
| "logps/rejected": -2.0575575828552246, | |
| "loss": 2.0134, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.4067161083221436, | |
| "rewards/margins": 0.6508415341377258, | |
| "rewards/rejected": -2.0575575828552246, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 7.876329412436082, | |
| "learning_rate": 7.999743062239557e-06, | |
| "logits/chosen": -0.4009264409542084, | |
| "logits/rejected": -0.22962765395641327, | |
| "logps/chosen": -1.3704339265823364, | |
| "logps/rejected": -2.099834680557251, | |
| "loss": 2.0007, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.3704339265823364, | |
| "rewards/margins": 0.7294005155563354, | |
| "rewards/rejected": -2.099834680557251, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 7.662219802711994, | |
| "learning_rate": 7.642410644905726e-06, | |
| "logits/chosen": -0.4107929766178131, | |
| "logits/rejected": -0.22412636876106262, | |
| "logps/chosen": -1.3308082818984985, | |
| "logps/rejected": -2.054591655731201, | |
| "loss": 2.0538, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.3308082818984985, | |
| "rewards/margins": 0.7237831354141235, | |
| "rewards/rejected": -2.054591655731201, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 8.408469077357054, | |
| "learning_rate": 7.2882386476358304e-06, | |
| "logits/chosen": -0.4748775064945221, | |
| "logits/rejected": -0.3870747983455658, | |
| "logps/chosen": -1.3256374597549438, | |
| "logps/rejected": -1.8613688945770264, | |
| "loss": 2.0223, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -1.3256374597549438, | |
| "rewards/margins": 0.5357314348220825, | |
| "rewards/rejected": -1.8613688945770264, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 9.20767722080512, | |
| "learning_rate": 6.937701848738407e-06, | |
| "logits/chosen": -0.5532702803611755, | |
| "logits/rejected": -0.517833411693573, | |
| "logps/chosen": -1.3089849948883057, | |
| "logps/rejected": -1.9914848804473877, | |
| "loss": 2.0097, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.3089849948883057, | |
| "rewards/margins": 0.6825000047683716, | |
| "rewards/rejected": -1.9914848804473877, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 9.168807609346384, | |
| "learning_rate": 6.591270153428288e-06, | |
| "logits/chosen": -0.5921626687049866, | |
| "logits/rejected": -0.5593982934951782, | |
| "logps/chosen": -1.233320951461792, | |
| "logps/rejected": -1.824541449546814, | |
| "loss": 2.0114, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.233320951461792, | |
| "rewards/margins": 0.5912207365036011, | |
| "rewards/rejected": -1.824541449546814, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 11.74928248778955, | |
| "learning_rate": 6.249407963904381e-06, | |
| "logits/chosen": -0.5820972323417664, | |
| "logits/rejected": -0.37617072463035583, | |
| "logps/chosen": -1.308586835861206, | |
| "logps/rejected": -2.1290369033813477, | |
| "loss": 2.0377, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.308586835861206, | |
| "rewards/margins": 0.8204501271247864, | |
| "rewards/rejected": -2.1290369033813477, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 7.446497900716036, | |
| "learning_rate": 5.912573556804453e-06, | |
| "logits/chosen": -0.4366278648376465, | |
| "logits/rejected": -0.34356969594955444, | |
| "logps/chosen": -1.3265436887741089, | |
| "logps/rejected": -1.9333693981170654, | |
| "loss": 1.9563, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.3265436887741089, | |
| "rewards/margins": 0.6068258285522461, | |
| "rewards/rejected": -1.9333693981170654, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 7.834867022615392, | |
| "learning_rate": 5.581218468871365e-06, | |
| "logits/chosen": -0.42373937368392944, | |
| "logits/rejected": -0.1534721851348877, | |
| "logps/chosen": -1.178399682044983, | |
| "logps/rejected": -1.9840328693389893, | |
| "loss": 1.9479, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.178399682044983, | |
| "rewards/margins": 0.8056330680847168, | |
| "rewards/rejected": -1.9840328693389893, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 7.323036569812089, | |
| "learning_rate": 5.2557868916543996e-06, | |
| "logits/chosen": -0.24924680590629578, | |
| "logits/rejected": 0.03496779128909111, | |
| "logps/chosen": -1.226240873336792, | |
| "logps/rejected": -1.8618109226226807, | |
| "loss": 1.9899, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.226240873336792, | |
| "rewards/margins": 0.6355697512626648, | |
| "rewards/rejected": -1.8618109226226807, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 7.704436038290053, | |
| "learning_rate": 4.9367150760569746e-06, | |
| "logits/chosen": -0.289539635181427, | |
| "logits/rejected": 0.08722052723169327, | |
| "logps/chosen": -1.2469284534454346, | |
| "logps/rejected": -2.0381367206573486, | |
| "loss": 1.9837, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.2469284534454346, | |
| "rewards/margins": 0.7912081480026245, | |
| "rewards/rejected": -2.0381367206573486, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 8.947529662445875, | |
| "learning_rate": 4.6244307475291025e-06, | |
| "logits/chosen": -0.18107546865940094, | |
| "logits/rejected": 0.22904996573925018, | |
| "logps/chosen": -1.446345329284668, | |
| "logps/rejected": -2.1773736476898193, | |
| "loss": 2.0338, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.446345329284668, | |
| "rewards/margins": 0.7310282588005066, | |
| "rewards/rejected": -2.1773736476898193, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 9.583549007245818, | |
| "learning_rate": 4.319352532688444e-06, | |
| "logits/chosen": -0.29274436831474304, | |
| "logits/rejected": 0.00033287107362411916, | |
| "logps/chosen": -1.2861610651016235, | |
| "logps/rejected": -1.9985120296478271, | |
| "loss": 2.0307, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.2861610651016235, | |
| "rewards/margins": 0.7123511433601379, | |
| "rewards/rejected": -1.9985120296478271, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 6.266721042983619, | |
| "learning_rate": 4.0218893981385935e-06, | |
| "logits/chosen": -0.28006237745285034, | |
| "logits/rejected": -0.14952346682548523, | |
| "logps/chosen": -1.2466567754745483, | |
| "logps/rejected": -1.7666349411010742, | |
| "loss": 2.0448, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.2466567754745483, | |
| "rewards/margins": 0.5199781656265259, | |
| "rewards/rejected": -1.7666349411010742, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 8.529189010191509, | |
| "learning_rate": 3.732440102236975e-06, | |
| "logits/chosen": -0.38612329959869385, | |
| "logits/rejected": -0.1318252980709076, | |
| "logps/chosen": -1.1369296312332153, | |
| "logps/rejected": -1.9028323888778687, | |
| "loss": 1.9434, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -1.1369296312332153, | |
| "rewards/margins": 0.7659028172492981, | |
| "rewards/rejected": -1.9028323888778687, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 8.98691100719935, | |
| "learning_rate": 3.4513926605471504e-06, | |
| "logits/chosen": -0.2708672881126404, | |
| "logits/rejected": -0.002604148583486676, | |
| "logps/chosen": -1.2128788232803345, | |
| "logps/rejected": -1.881731629371643, | |
| "loss": 1.9188, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.2128788232803345, | |
| "rewards/margins": 0.6688528060913086, | |
| "rewards/rejected": -1.881731629371643, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 8.427363605011601, | |
| "learning_rate": 3.1791238256921785e-06, | |
| "logits/chosen": -0.2245834320783615, | |
| "logits/rejected": 0.03710466995835304, | |
| "logps/chosen": -1.3998098373413086, | |
| "logps/rejected": -2.1002440452575684, | |
| "loss": 2.0448, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3998098373413086, | |
| "rewards/margins": 0.7004340887069702, | |
| "rewards/rejected": -2.1002440452575684, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 8.84465057372313, | |
| "learning_rate": 2.9159985823062997e-06, | |
| "logits/chosen": -0.35639292001724243, | |
| "logits/rejected": -0.17423222959041595, | |
| "logps/chosen": -1.3097021579742432, | |
| "logps/rejected": -2.0959980487823486, | |
| "loss": 1.9799, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.3097021579742432, | |
| "rewards/margins": 0.7862957715988159, | |
| "rewards/rejected": -2.0959980487823486, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 7.000990509721392, | |
| "learning_rate": 2.662369657761963e-06, | |
| "logits/chosen": -0.32897457480430603, | |
| "logits/rejected": -0.3297235369682312, | |
| "logps/chosen": -1.2792903184890747, | |
| "logps/rejected": -1.8416297435760498, | |
| "loss": 2.0065, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.2792903184890747, | |
| "rewards/margins": 0.5623396635055542, | |
| "rewards/rejected": -1.8416297435760498, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 8.146523706243913, | |
| "learning_rate": 2.418577049328058e-06, | |
| "logits/chosen": -0.3611428439617157, | |
| "logits/rejected": -0.2454133927822113, | |
| "logps/chosen": -1.266564130783081, | |
| "logps/rejected": -1.7843055725097656, | |
| "loss": 1.9381, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.266564130783081, | |
| "rewards/margins": 0.517741322517395, | |
| "rewards/rejected": -1.7843055725097656, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 8.167524524758202, | |
| "learning_rate": 2.1849475683932996e-06, | |
| "logits/chosen": -0.3831802010536194, | |
| "logits/rejected": -0.2638740539550781, | |
| "logps/chosen": -1.2620943784713745, | |
| "logps/rejected": -1.9156465530395508, | |
| "loss": 1.9611, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.2620943784713745, | |
| "rewards/margins": 0.6535523533821106, | |
| "rewards/rejected": -1.9156465530395508, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 8.30770659075244, | |
| "learning_rate": 1.961794402365611e-06, | |
| "logits/chosen": -0.3085532486438751, | |
| "logits/rejected": -0.14951160550117493, | |
| "logps/chosen": -1.2462884187698364, | |
| "logps/rejected": -2.1334927082061768, | |
| "loss": 1.9555, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.2462884187698364, | |
| "rewards/margins": 0.8872040510177612, | |
| "rewards/rejected": -2.1334927082061768, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 7.7825890292689905, | |
| "learning_rate": 1.7494166948349057e-06, | |
| "logits/chosen": -0.267643541097641, | |
| "logits/rejected": 0.02370324358344078, | |
| "logps/chosen": -1.3041934967041016, | |
| "logps/rejected": -2.0585289001464844, | |
| "loss": 1.9722, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.3041934967041016, | |
| "rewards/margins": 0.7543356418609619, | |
| "rewards/rejected": -2.0585289001464844, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 11.47946040263076, | |
| "learning_rate": 1.5480991445620541e-06, | |
| "logits/chosen": -0.2226092368364334, | |
| "logits/rejected": 0.03486952185630798, | |
| "logps/chosen": -1.237866759300232, | |
| "logps/rejected": -1.9466907978057861, | |
| "loss": 1.9589, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.237866759300232, | |
| "rewards/margins": 0.7088239192962646, | |
| "rewards/rejected": -1.9466907978057861, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 8.663615848996892, | |
| "learning_rate": 1.3581116238315194e-06, | |
| "logits/chosen": -0.2045535147190094, | |
| "logits/rejected": 0.006712320260703564, | |
| "logps/chosen": -1.3987605571746826, | |
| "logps/rejected": -2.1295289993286133, | |
| "loss": 2.0021, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.3987605571746826, | |
| "rewards/margins": 0.7307685017585754, | |
| "rewards/rejected": -2.1295289993286133, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 8.186995321334237, | |
| "learning_rate": 1.1797088166794002e-06, | |
| "logits/chosen": -0.20332176983356476, | |
| "logits/rejected": 0.029474016278982162, | |
| "logps/chosen": -1.172918677330017, | |
| "logps/rejected": -1.7614654302597046, | |
| "loss": 1.9558, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.172918677330017, | |
| "rewards/margins": 0.5885466933250427, | |
| "rewards/rejected": -1.7614654302597046, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 6.525083767762586, | |
| "learning_rate": 1.013129877481741e-06, | |
| "logits/chosen": -0.2240767925977707, | |
| "logits/rejected": 0.07820748537778854, | |
| "logps/chosen": -1.20893132686615, | |
| "logps/rejected": -1.91985285282135, | |
| "loss": 1.9579, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.20893132686615, | |
| "rewards/margins": 0.7109212875366211, | |
| "rewards/rejected": -1.91985285282135, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 9.083137976581597, | |
| "learning_rate": 8.585981103608343e-07, | |
| "logits/chosen": -0.11082730442285538, | |
| "logits/rejected": 0.09507735818624496, | |
| "logps/chosen": -1.1996517181396484, | |
| "logps/rejected": -1.8896563053131104, | |
| "loss": 2.0042, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1996517181396484, | |
| "rewards/margins": 0.6900044679641724, | |
| "rewards/rejected": -1.8896563053131104, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 6.873945114472838, | |
| "learning_rate": 7.163206698392744e-07, | |
| "logits/chosen": -0.10862954705953598, | |
| "logits/rejected": 0.1946602761745453, | |
| "logps/chosen": -1.3608647584915161, | |
| "logps/rejected": -1.970487356185913, | |
| "loss": 1.9911, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.3608647584915161, | |
| "rewards/margins": 0.6096227169036865, | |
| "rewards/rejected": -1.970487356185913, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 7.542928376234922, | |
| "learning_rate": 5.864882831430274e-07, | |
| "logits/chosen": -0.16213567554950714, | |
| "logits/rejected": 0.21448484063148499, | |
| "logps/chosen": -1.3169506788253784, | |
| "logps/rejected": -2.048978328704834, | |
| "loss": 1.956, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.3169506788253784, | |
| "rewards/margins": 0.7320275902748108, | |
| "rewards/rejected": -2.048978328704834, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 10.100813855786358, | |
| "learning_rate": 4.6927499452580574e-07, | |
| "logits/chosen": -0.12442419677972794, | |
| "logits/rejected": 0.06503897905349731, | |
| "logps/chosen": -1.288496732711792, | |
| "logps/rejected": -2.066117286682129, | |
| "loss": 1.9804, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.288496732711792, | |
| "rewards/margins": 0.777620792388916, | |
| "rewards/rejected": -2.066117286682129, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 10.273579675957365, | |
| "learning_rate": 3.6483793195745686e-07, | |
| "logits/chosen": -0.04678087681531906, | |
| "logits/rejected": 0.3355256915092468, | |
| "logps/chosen": -1.2764372825622559, | |
| "logps/rejected": -2.03460955619812, | |
| "loss": 1.9929, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.2764372825622559, | |
| "rewards/margins": 0.7581723928451538, | |
| "rewards/rejected": -2.03460955619812, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 8.126486085847516, | |
| "learning_rate": 2.733170964891607e-07, | |
| "logits/chosen": -0.17456679046154022, | |
| "logits/rejected": 0.09745622426271439, | |
| "logps/chosen": -1.2472385168075562, | |
| "logps/rejected": -1.892249345779419, | |
| "loss": 1.9962, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.2472385168075562, | |
| "rewards/margins": 0.6450108289718628, | |
| "rewards/rejected": -1.892249345779419, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 8.866488191256463, | |
| "learning_rate": 1.9483517457776436e-07, | |
| "logits/chosen": -0.05351231247186661, | |
| "logits/rejected": 0.1783636510372162, | |
| "logps/chosen": -1.2550338506698608, | |
| "logps/rejected": -1.8777573108673096, | |
| "loss": 1.9585, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.2550338506698608, | |
| "rewards/margins": 0.6227231621742249, | |
| "rewards/rejected": -1.8777573108673096, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 8.43900399314022, | |
| "learning_rate": 1.2949737362087156e-07, | |
| "logits/chosen": -0.09251005947589874, | |
| "logits/rejected": 0.2792736291885376, | |
| "logps/chosen": -1.381317377090454, | |
| "logps/rejected": -1.868950605392456, | |
| "loss": 1.9791, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.381317377090454, | |
| "rewards/margins": 0.48763322830200195, | |
| "rewards/rejected": -1.868950605392456, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 7.3135362806868365, | |
| "learning_rate": 7.73912809231292e-08, | |
| "logits/chosen": -0.15108491480350494, | |
| "logits/rejected": 0.16171926259994507, | |
| "logps/chosen": -1.2007992267608643, | |
| "logps/rejected": -2.045020341873169, | |
| "loss": 1.9082, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.2007992267608643, | |
| "rewards/margins": 0.8442209362983704, | |
| "rewards/rejected": -2.045020341873169, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 8.496120629902867, | |
| "learning_rate": 3.858674628278825e-08, | |
| "logits/chosen": -0.16963747143745422, | |
| "logits/rejected": 0.33284881711006165, | |
| "logps/chosen": -1.3178845643997192, | |
| "logps/rejected": -2.077279806137085, | |
| "loss": 1.9421, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.3178845643997192, | |
| "rewards/margins": 0.7593953013420105, | |
| "rewards/rejected": -2.077279806137085, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 8.29474797261466, | |
| "learning_rate": 1.3135788355934652e-08, | |
| "logits/chosen": -0.18520286679267883, | |
| "logits/rejected": 0.14646300673484802, | |
| "logps/chosen": -1.2585632801055908, | |
| "logps/rejected": -1.9155442714691162, | |
| "loss": 1.9995, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.2585632801055908, | |
| "rewards/margins": 0.6569809317588806, | |
| "rewards/rejected": -1.9155442714691162, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 8.421060788269113, | |
| "learning_rate": 1.0725249238940916e-09, | |
| "logits/chosen": -0.2087993174791336, | |
| "logits/rejected": 0.3205938935279846, | |
| "logps/chosen": -1.2470612525939941, | |
| "logps/rejected": -1.9542922973632812, | |
| "loss": 1.985, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.2470612525939941, | |
| "rewards/margins": 0.7072311639785767, | |
| "rewards/rejected": -1.9542922973632812, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 0.0, | |
| "train_runtime": 4.0963, | |
| "train_samples_per_second": 14924.323, | |
| "train_steps_per_second": 116.446 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |