| { |
| "best_global_step": 300, |
| "best_metric": 0.77076083, |
| "best_model_checkpoint": "/data2/kdd/crag/cjz/output/task3_dpo_gt6_task2_gt60.5+task3gt5/v1-20250609-010021/checkpoint-300", |
| "epoch": 0.5980030968017513, |
| "eval_steps": 300, |
| "global_step": 350, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0017085802765764322, |
| "grad_norm": 9.203307151794434, |
| "learning_rate": 3.3898305084745763e-06, |
| "logits/chosen": -0.6409764885902405, |
| "logits/rejected": -0.6413162350654602, |
| "logps/chosen": -8.591513633728027, |
| "logps/rejected": -9.922354698181152, |
| "loss": 1.2514135837554932, |
| "memory(GiB)": 45.3, |
| "nll_loss": 0.5582665801048279, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 1, |
| "train_speed(iter/s)": 0.005423 |
| }, |
| { |
| "epoch": 0.0034171605531528645, |
| "grad_norm": 4.572653293609619, |
| "learning_rate": 6.779661016949153e-06, |
| "logits/chosen": -0.6253237724304199, |
| "logits/rejected": -0.6232100129127502, |
| "logps/chosen": -9.215431213378906, |
| "logps/rejected": -6.302733898162842, |
| "loss": 1.1804862022399902, |
| "memory(GiB)": 45.3, |
| "nll_loss": 0.4873391091823578, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 2, |
| "train_speed(iter/s)": 0.005426 |
| }, |
| { |
| "epoch": 0.005125740829729297, |
| "grad_norm": 10.135663986206055, |
| "learning_rate": 1.016949152542373e-05, |
| "logits/chosen": -0.5830379724502563, |
| "logits/rejected": -0.5847949981689453, |
| "logps/chosen": -6.387916088104248, |
| "logps/rejected": -12.526434898376465, |
| "loss": 1.1388403177261353, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.44991743564605713, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.0018886500038206577, |
| "rewards/margins": 0.008566470816731453, |
| "rewards/rejected": -0.006677820347249508, |
| "step": 3, |
| "train_speed(iter/s)": 0.005372 |
| }, |
| { |
| "epoch": 0.006834321106305729, |
| "grad_norm": 8.09815788269043, |
| "learning_rate": 1.3559322033898305e-05, |
| "logits/chosen": -0.5831664800643921, |
| "logits/rejected": -0.5881690382957458, |
| "logps/chosen": -9.77461051940918, |
| "logps/rejected": -15.103912353515625, |
| "loss": 1.2644962072372437, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.5699671506881714, |
| "rewards/accuracies": 0.4375, |
| "rewards/chosen": -0.0013185496209189296, |
| "rewards/margins": -0.0025646828580647707, |
| "rewards/rejected": 0.0012461334699764848, |
| "step": 4, |
| "train_speed(iter/s)": 0.005433 |
| }, |
| { |
| "epoch": 0.008542901382882162, |
| "grad_norm": 10.546202659606934, |
| "learning_rate": 1.694915254237288e-05, |
| "logits/chosen": -0.6210501194000244, |
| "logits/rejected": -0.6261105537414551, |
| "logps/chosen": -9.743502616882324, |
| "logps/rejected": -13.954730987548828, |
| "loss": 1.2041635513305664, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.5165743231773376, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.014782106503844261, |
| "rewards/margins": 0.011779396794736385, |
| "rewards/rejected": 0.0030027113389223814, |
| "step": 5, |
| "train_speed(iter/s)": 0.005434 |
| }, |
| { |
| "epoch": 0.010251481659458593, |
| "grad_norm": 5.953731060028076, |
| "learning_rate": 2.033898305084746e-05, |
| "logits/chosen": -0.6194843649864197, |
| "logits/rejected": -0.6225060224533081, |
| "logps/chosen": -8.993276596069336, |
| "logps/rejected": -11.800742149353027, |
| "loss": 1.2383989095687866, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.5413038730621338, |
| "rewards/accuracies": 0.46875, |
| "rewards/chosen": 0.02151685394346714, |
| "rewards/margins": -0.004275719169527292, |
| "rewards/rejected": 0.025792576372623444, |
| "step": 6, |
| "train_speed(iter/s)": 0.005421 |
| }, |
| { |
| "epoch": 0.011960061936035027, |
| "grad_norm": 3.4447383880615234, |
| "learning_rate": 2.3728813559322036e-05, |
| "logits/chosen": -0.5979399681091309, |
| "logits/rejected": -0.597865104675293, |
| "logps/chosen": -11.88587760925293, |
| "logps/rejected": -12.513296127319336, |
| "loss": 1.1787316799163818, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.4771195948123932, |
| "rewards/accuracies": 0.46875, |
| "rewards/chosen": 0.001614744309335947, |
| "rewards/margins": -0.006370083428919315, |
| "rewards/rejected": 0.007984823547303677, |
| "step": 7, |
| "train_speed(iter/s)": 0.005421 |
| }, |
| { |
| "epoch": 0.013668642212611458, |
| "grad_norm": 4.3422627449035645, |
| "learning_rate": 2.711864406779661e-05, |
| "logits/chosen": -0.5695681571960449, |
| "logits/rejected": -0.5736703276634216, |
| "logps/chosen": -8.691235542297363, |
| "logps/rejected": -17.510940551757812, |
| "loss": 1.085353970527649, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.4569101631641388, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.061018284410238266, |
| "rewards/margins": 0.16366331279277802, |
| "rewards/rejected": -0.10264502465724945, |
| "step": 8, |
| "train_speed(iter/s)": 0.005421 |
| }, |
| { |
| "epoch": 0.015377222489187891, |
| "grad_norm": 6.347766399383545, |
| "learning_rate": 3.050847457627119e-05, |
| "logits/chosen": -0.6150330305099487, |
| "logits/rejected": -0.6163787841796875, |
| "logps/chosen": -14.26973819732666, |
| "logps/rejected": -12.656587600708008, |
| "loss": 1.3125040531158447, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.6051884293556213, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": -0.06490564346313477, |
| "rewards/margins": 0.03234899416565895, |
| "rewards/rejected": -0.09725463390350342, |
| "step": 9, |
| "train_speed(iter/s)": 0.005427 |
| }, |
| { |
| "epoch": 0.017085802765764324, |
| "grad_norm": 3.609342575073242, |
| "learning_rate": 3.389830508474576e-05, |
| "logits/chosen": -0.5823943018913269, |
| "logits/rejected": -0.5893073678016663, |
| "logps/chosen": -12.539135932922363, |
| "logps/rejected": -19.231651306152344, |
| "loss": 1.154615044593811, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.5292682647705078, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.0102485166862607, |
| "rewards/margins": 0.24539220333099365, |
| "rewards/rejected": -0.2556407153606415, |
| "step": 10, |
| "train_speed(iter/s)": 0.00544 |
| }, |
| { |
| "epoch": 0.018794383042340754, |
| "grad_norm": 4.803543567657471, |
| "learning_rate": 3.728813559322034e-05, |
| "logits/chosen": -0.6298958659172058, |
| "logits/rejected": -0.6301961541175842, |
| "logps/chosen": -10.286799430847168, |
| "logps/rejected": -11.364641189575195, |
| "loss": 1.143830418586731, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.43475160002708435, |
| "rewards/accuracies": 0.46875, |
| "rewards/chosen": -0.06166127696633339, |
| "rewards/margins": 0.1280723363161087, |
| "rewards/rejected": -0.189733624458313, |
| "step": 11, |
| "train_speed(iter/s)": 0.005459 |
| }, |
| { |
| "epoch": 0.020502963318917187, |
| "grad_norm": 6.843598365783691, |
| "learning_rate": 4.067796610169492e-05, |
| "logits/chosen": -0.6147834062576294, |
| "logits/rejected": -0.6201154589653015, |
| "logps/chosen": -11.104015350341797, |
| "logps/rejected": -14.71792221069336, |
| "loss": 1.2689580917358398, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.4583723843097687, |
| "rewards/accuracies": 0.375, |
| "rewards/chosen": -0.1446399688720703, |
| "rewards/margins": -0.06440749764442444, |
| "rewards/rejected": -0.08023246377706528, |
| "step": 12, |
| "train_speed(iter/s)": 0.005454 |
| }, |
| { |
| "epoch": 0.02221154359549362, |
| "grad_norm": 4.361201286315918, |
| "learning_rate": 4.4067796610169495e-05, |
| "logits/chosen": -0.6593450307846069, |
| "logits/rejected": -0.6623726487159729, |
| "logps/chosen": -8.788830757141113, |
| "logps/rejected": -12.09660530090332, |
| "loss": 1.091460108757019, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.38157719373703003, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": -0.03397805988788605, |
| "rewards/margins": 0.08494514226913452, |
| "rewards/rejected": -0.11892320215702057, |
| "step": 13, |
| "train_speed(iter/s)": 0.005465 |
| }, |
| { |
| "epoch": 0.023920123872070053, |
| "grad_norm": 4.36707067489624, |
| "learning_rate": 4.745762711864407e-05, |
| "logits/chosen": -0.6003404259681702, |
| "logits/rejected": -0.6018444895744324, |
| "logps/chosen": -9.472915649414062, |
| "logps/rejected": -12.008777618408203, |
| "loss": 1.2156641483306885, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.42055219411849976, |
| "rewards/accuracies": 0.40625, |
| "rewards/chosen": -0.07689498364925385, |
| "rewards/margins": -0.1264859437942505, |
| "rewards/rejected": 0.049590952694416046, |
| "step": 14, |
| "train_speed(iter/s)": 0.00548 |
| }, |
| { |
| "epoch": 0.025628704148646483, |
| "grad_norm": 3.2864296436309814, |
| "learning_rate": 5.0847457627118643e-05, |
| "logits/chosen": -0.6285042762756348, |
| "logits/rejected": -0.6272621154785156, |
| "logps/chosen": -14.201019287109375, |
| "logps/rejected": -10.998590469360352, |
| "loss": 1.2311502695083618, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.5152750015258789, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.06618006527423859, |
| "rewards/margins": 0.029741406440734863, |
| "rewards/rejected": 0.036438651382923126, |
| "step": 15, |
| "train_speed(iter/s)": 0.005474 |
| }, |
| { |
| "epoch": 0.027337284425222916, |
| "grad_norm": 2.4144492149353027, |
| "learning_rate": 5.423728813559322e-05, |
| "logits/chosen": -0.6098039746284485, |
| "logits/rejected": -0.6130300760269165, |
| "logps/chosen": -3.4820291996002197, |
| "logps/rejected": -16.196063995361328, |
| "loss": 0.8393359780311584, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.214947909116745, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.0826798751950264, |
| "rewards/margins": 0.17492221295833588, |
| "rewards/rejected": -0.09224233776330948, |
| "step": 16, |
| "train_speed(iter/s)": 0.005477 |
| }, |
| { |
| "epoch": 0.02904586470179935, |
| "grad_norm": 2.3232851028442383, |
| "learning_rate": 5.76271186440678e-05, |
| "logits/chosen": -0.5909145474433899, |
| "logits/rejected": -0.593756377696991, |
| "logps/chosen": -7.929438591003418, |
| "logps/rejected": -14.363272666931152, |
| "loss": 0.9911059737205505, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.34508591890335083, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.10390710830688477, |
| "rewards/margins": 0.13225752115249634, |
| "rewards/rejected": -0.028350409120321274, |
| "step": 17, |
| "train_speed(iter/s)": 0.005482 |
| }, |
| { |
| "epoch": 0.030754444978375782, |
| "grad_norm": 2.344477891921997, |
| "learning_rate": 6.101694915254238e-05, |
| "logits/chosen": -0.5697246789932251, |
| "logits/rejected": -0.5692474246025085, |
| "logps/chosen": -6.287097930908203, |
| "logps/rejected": -10.676097869873047, |
| "loss": 0.9974570274353027, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.3142527639865875, |
| "rewards/accuracies": 0.5625, |
| "rewards/chosen": 0.09520465135574341, |
| "rewards/margins": 0.04853628948330879, |
| "rewards/rejected": 0.046668365597724915, |
| "step": 18, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.03246302525495221, |
| "grad_norm": 2.361417531967163, |
| "learning_rate": 6.440677966101695e-05, |
| "logits/chosen": -0.5909584164619446, |
| "logits/rejected": -0.596694827079773, |
| "logps/chosen": -4.706894874572754, |
| "logps/rejected": -12.18591022491455, |
| "loss": 0.8993738889694214, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.2403678447008133, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.1180216372013092, |
| "rewards/margins": 0.09891890734434128, |
| "rewards/rejected": 0.01910274103283882, |
| "step": 19, |
| "train_speed(iter/s)": 0.005497 |
| }, |
| { |
| "epoch": 0.03417160553152865, |
| "grad_norm": 2.1051840782165527, |
| "learning_rate": 6.779661016949152e-05, |
| "logits/chosen": -0.5925375819206238, |
| "logits/rejected": -0.5939764380455017, |
| "logps/chosen": -7.690135478973389, |
| "logps/rejected": -11.360199928283691, |
| "loss": 0.9811086058616638, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.27671271562576294, |
| "rewards/accuracies": 0.53125, |
| "rewards/chosen": 0.15354080498218536, |
| "rewards/margins": 0.02018781006336212, |
| "rewards/rejected": 0.13335299491882324, |
| "step": 20, |
| "train_speed(iter/s)": 0.005507 |
| }, |
| { |
| "epoch": 0.03588018580810508, |
| "grad_norm": 2.453577995300293, |
| "learning_rate": 7.11864406779661e-05, |
| "logits/chosen": -0.589929461479187, |
| "logits/rejected": -0.5913950800895691, |
| "logps/chosen": -6.006174564361572, |
| "logps/rejected": -12.356660842895508, |
| "loss": 0.8762722015380859, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.26102662086486816, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.15477147698402405, |
| "rewards/margins": 0.20186343789100647, |
| "rewards/rejected": -0.047091953456401825, |
| "step": 21, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.03758876608468151, |
| "grad_norm": 2.334592819213867, |
| "learning_rate": 7.457627118644068e-05, |
| "logits/chosen": -0.6286447048187256, |
| "logits/rejected": -0.6310911178588867, |
| "logps/chosen": -9.828102111816406, |
| "logps/rejected": -16.171802520751953, |
| "loss": 0.9827708601951599, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.40016523003578186, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.22255462408065796, |
| "rewards/margins": 0.2890108525753021, |
| "rewards/rejected": -0.06645623594522476, |
| "step": 22, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.039297346361257944, |
| "grad_norm": 2.763946533203125, |
| "learning_rate": 7.796610169491526e-05, |
| "logits/chosen": -0.6423724889755249, |
| "logits/rejected": -0.6475900411605835, |
| "logps/chosen": -7.048077583312988, |
| "logps/rejected": -17.518115997314453, |
| "loss": 1.029776692390442, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.4258265495300293, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.13568007946014404, |
| "rewards/margins": 0.24192118644714355, |
| "rewards/rejected": -0.10624107718467712, |
| "step": 23, |
| "train_speed(iter/s)": 0.005498 |
| }, |
| { |
| "epoch": 0.041005926637834374, |
| "grad_norm": 5.438773155212402, |
| "learning_rate": 8.135593220338983e-05, |
| "logits/chosen": -0.611075758934021, |
| "logits/rejected": -0.6109172701835632, |
| "logps/chosen": -8.579011917114258, |
| "logps/rejected": -12.74751091003418, |
| "loss": 1.2176915407180786, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.5213994979858398, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": -0.010570855811238289, |
| "rewards/margins": 0.06740333139896393, |
| "rewards/rejected": -0.07797417044639587, |
| "step": 24, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.0427145069144108, |
| "grad_norm": 3.8305866718292236, |
| "learning_rate": 8.474576271186441e-05, |
| "logits/chosen": -0.6644467711448669, |
| "logits/rejected": -0.6674079298973083, |
| "logps/chosen": -6.975298881530762, |
| "logps/rejected": -15.503819465637207, |
| "loss": 0.922063410282135, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.3326256275177002, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.06792093813419342, |
| "rewards/margins": 0.324211061000824, |
| "rewards/rejected": -0.25629013776779175, |
| "step": 25, |
| "train_speed(iter/s)": 0.00549 |
| }, |
| { |
| "epoch": 0.04442308719098724, |
| "grad_norm": 4.446774959564209, |
| "learning_rate": 8.813559322033899e-05, |
| "logits/chosen": -0.6409133076667786, |
| "logits/rejected": -0.6452510356903076, |
| "logps/chosen": -5.22960090637207, |
| "logps/rejected": -16.918365478515625, |
| "loss": 0.7774365544319153, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.2448062300682068, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.2400483787059784, |
| "rewards/margins": 0.39764925837516785, |
| "rewards/rejected": -0.15760089457035065, |
| "step": 26, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.04613166746756367, |
| "grad_norm": 2.5869643688201904, |
| "learning_rate": 9.152542372881357e-05, |
| "logits/chosen": -0.6091906428337097, |
| "logits/rejected": -0.6120930910110474, |
| "logps/chosen": -7.03964376449585, |
| "logps/rejected": -18.091930389404297, |
| "loss": 0.8498358726501465, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.3106813132762909, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.1571023166179657, |
| "rewards/margins": 0.4190843403339386, |
| "rewards/rejected": -0.2619820237159729, |
| "step": 27, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.047840247744140106, |
| "grad_norm": 3.0630600452423096, |
| "learning_rate": 9.491525423728815e-05, |
| "logits/chosen": -0.6014366745948792, |
| "logits/rejected": -0.6016579270362854, |
| "logps/chosen": -9.281977653503418, |
| "logps/rejected": -13.255182266235352, |
| "loss": 0.9999822974205017, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.3916056454181671, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.0838853195309639, |
| "rewards/margins": 0.22357328236103058, |
| "rewards/rejected": -0.13968798518180847, |
| "step": 28, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.049548828020716536, |
| "grad_norm": 4.956308364868164, |
| "learning_rate": 9.830508474576272e-05, |
| "logits/chosen": -0.6361557841300964, |
| "logits/rejected": -0.6375819444656372, |
| "logps/chosen": -5.407459735870361, |
| "logps/rejected": -12.960749626159668, |
| "loss": 0.9380594491958618, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.39800193905830383, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.14713114500045776, |
| "rewards/margins": 0.4169777035713196, |
| "rewards/rejected": -0.2698465585708618, |
| "step": 29, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.051257408297292965, |
| "grad_norm": 6.679542541503906, |
| "learning_rate": 0.00010169491525423729, |
| "logits/chosen": -0.6229231357574463, |
| "logits/rejected": -0.6250395178794861, |
| "logps/chosen": -11.77493667602539, |
| "logps/rejected": -18.192317962646484, |
| "loss": 0.7899165749549866, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.2795858085155487, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.09651458263397217, |
| "rewards/margins": 0.49087047576904297, |
| "rewards/rejected": -0.3943559229373932, |
| "step": 30, |
| "train_speed(iter/s)": 0.005497 |
| }, |
| { |
| "epoch": 0.0529659885738694, |
| "grad_norm": 4.011873245239258, |
| "learning_rate": 0.00010508474576271188, |
| "logits/chosen": -0.6524882912635803, |
| "logits/rejected": -0.6541539430618286, |
| "logps/chosen": -7.8713202476501465, |
| "logps/rejected": -11.641451835632324, |
| "loss": 0.9657379984855652, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.40233710408210754, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.11578395962715149, |
| "rewards/margins": 0.36746829748153687, |
| "rewards/rejected": -0.25168436765670776, |
| "step": 31, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.05467456885044583, |
| "grad_norm": 4.144499778747559, |
| "learning_rate": 0.00010847457627118644, |
| "logits/chosen": -0.6458109617233276, |
| "logits/rejected": -0.6519441604614258, |
| "logps/chosen": -8.17459487915039, |
| "logps/rejected": -19.17450714111328, |
| "loss": 0.9243566989898682, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.3649522066116333, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.17762523889541626, |
| "rewards/margins": 0.41990581154823303, |
| "rewards/rejected": -0.24228057265281677, |
| "step": 32, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.05638314912702227, |
| "grad_norm": 2.5846283435821533, |
| "learning_rate": 0.00011186440677966102, |
| "logits/chosen": -0.6422857046127319, |
| "logits/rejected": -0.64764004945755, |
| "logps/chosen": -7.8391828536987305, |
| "logps/rejected": -16.905941009521484, |
| "loss": 0.8121039867401123, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.3524104952812195, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.21112650632858276, |
| "rewards/margins": 0.6391255259513855, |
| "rewards/rejected": -0.42799898982048035, |
| "step": 33, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.0580917294035987, |
| "grad_norm": 2.831099271774292, |
| "learning_rate": 0.0001152542372881356, |
| "logits/chosen": -0.6351624727249146, |
| "logits/rejected": -0.6340872645378113, |
| "logps/chosen": -8.873239517211914, |
| "logps/rejected": -14.13478946685791, |
| "loss": 0.9939954280853271, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.5031149983406067, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.15203362703323364, |
| "rewards/margins": 0.5661364793777466, |
| "rewards/rejected": -0.41410285234451294, |
| "step": 34, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.05980030968017513, |
| "grad_norm": 6.754673004150391, |
| "learning_rate": 0.00011864406779661017, |
| "logits/chosen": -0.6350009441375732, |
| "logits/rejected": -0.6345161199569702, |
| "logps/chosen": -9.049372673034668, |
| "logps/rejected": -13.872842788696289, |
| "loss": 0.9677298069000244, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.4823642373085022, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.18822304904460907, |
| "rewards/margins": 0.5690968632698059, |
| "rewards/rejected": -0.38087382912635803, |
| "step": 35, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.061508889956751564, |
| "grad_norm": 5.09066915512085, |
| "learning_rate": 0.00012203389830508477, |
| "logits/chosen": -0.5940477252006531, |
| "logits/rejected": -0.5995637774467468, |
| "logps/chosen": -6.552436351776123, |
| "logps/rejected": -22.02668571472168, |
| "loss": 1.0198695659637451, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.5385282635688782, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.09731997549533844, |
| "rewards/margins": 0.7183221578598022, |
| "rewards/rejected": -0.621002197265625, |
| "step": 36, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.063217470233328, |
| "grad_norm": 6.366957664489746, |
| "learning_rate": 0.00012542372881355933, |
| "logits/chosen": -0.5671308040618896, |
| "logits/rejected": -0.5699396133422852, |
| "logps/chosen": -8.690061569213867, |
| "logps/rejected": -18.227741241455078, |
| "loss": 0.9119342565536499, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.4206831753253937, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.008592324331402779, |
| "rewards/margins": 0.7837947010993958, |
| "rewards/rejected": -0.7752023935317993, |
| "step": 37, |
| "train_speed(iter/s)": 0.005499 |
| }, |
| { |
| "epoch": 0.06492605050990442, |
| "grad_norm": 8.224360466003418, |
| "learning_rate": 0.0001288135593220339, |
| "logits/chosen": -0.6006415486335754, |
| "logits/rejected": -0.6039485931396484, |
| "logps/chosen": -7.98246955871582, |
| "logps/rejected": -19.214567184448242, |
| "loss": 1.020494818687439, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.47989213466644287, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": -0.017101695761084557, |
| "rewards/margins": 0.7686665654182434, |
| "rewards/rejected": -0.7857682704925537, |
| "step": 38, |
| "train_speed(iter/s)": 0.005499 |
| }, |
| { |
| "epoch": 0.06663463078648085, |
| "grad_norm": 5.284773349761963, |
| "learning_rate": 0.00013220338983050849, |
| "logits/chosen": -0.5746678709983826, |
| "logits/rejected": -0.5742544531822205, |
| "logps/chosen": -10.300456047058105, |
| "logps/rejected": -14.731122970581055, |
| "loss": 1.1951102018356323, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.6024460196495056, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.005457160994410515, |
| "rewards/margins": 0.43640822172164917, |
| "rewards/rejected": -0.44186532497406006, |
| "step": 39, |
| "train_speed(iter/s)": 0.005499 |
| }, |
| { |
| "epoch": 0.0683432110630573, |
| "grad_norm": 4.581506252288818, |
| "learning_rate": 0.00013559322033898305, |
| "logits/chosen": -0.5436846613883972, |
| "logits/rejected": -0.5412294268608093, |
| "logps/chosen": -11.306229591369629, |
| "logps/rejected": -15.808780670166016, |
| "loss": 0.9979494214057922, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.4069577753543854, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.10538371652364731, |
| "rewards/margins": 0.4707936942577362, |
| "rewards/rejected": -0.5761774778366089, |
| "step": 40, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.07005179133963373, |
| "grad_norm": 4.280210018157959, |
| "learning_rate": 0.00013898305084745764, |
| "logits/chosen": -0.5919702649116516, |
| "logits/rejected": -0.5944460034370422, |
| "logps/chosen": -4.601211071014404, |
| "logps/rejected": -18.39055061340332, |
| "loss": 0.7289348244667053, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.2620496451854706, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.1577654778957367, |
| "rewards/margins": 0.7140093445777893, |
| "rewards/rejected": -0.5562438368797302, |
| "step": 41, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.07176037161621016, |
| "grad_norm": 2.8965229988098145, |
| "learning_rate": 0.0001423728813559322, |
| "logits/chosen": -0.5226963758468628, |
| "logits/rejected": -0.5210604667663574, |
| "logps/chosen": -8.042407035827637, |
| "logps/rejected": -13.117356300354004, |
| "loss": 0.8382841944694519, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.3229953646659851, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.15337654948234558, |
| "rewards/margins": 0.5258168578147888, |
| "rewards/rejected": -0.3724403381347656, |
| "step": 42, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.07346895189278659, |
| "grad_norm": 4.7869791984558105, |
| "learning_rate": 0.00014576271186440677, |
| "logits/chosen": -0.5595600605010986, |
| "logits/rejected": -0.5619646310806274, |
| "logps/chosen": -5.741379261016846, |
| "logps/rejected": -14.867729187011719, |
| "loss": 0.9325981140136719, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.34299078583717346, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.06792251765727997, |
| "rewards/margins": 0.3079211115837097, |
| "rewards/rejected": -0.23999860882759094, |
| "step": 43, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.07517753216936301, |
| "grad_norm": 3.8402724266052246, |
| "learning_rate": 0.00014915254237288136, |
| "logits/chosen": -0.5551996231079102, |
| "logits/rejected": -0.5583171248435974, |
| "logps/chosen": -7.729850769042969, |
| "logps/rejected": -14.535001754760742, |
| "loss": 0.985005795955658, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.37224525213241577, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.0475621297955513, |
| "rewards/margins": 0.2842516601085663, |
| "rewards/rejected": -0.23668958246707916, |
| "step": 44, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.07688611244593946, |
| "grad_norm": 3.136073112487793, |
| "learning_rate": 0.00015254237288135592, |
| "logits/chosen": -0.5538796186447144, |
| "logits/rejected": -0.5563740730285645, |
| "logps/chosen": -8.61514663696289, |
| "logps/rejected": -20.434301376342773, |
| "loss": 0.8447250723838806, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.3852572441101074, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.17051486670970917, |
| "rewards/margins": 0.695674479007721, |
| "rewards/rejected": -0.5251596570014954, |
| "step": 45, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.07859469272251589, |
| "grad_norm": 2.950076103210449, |
| "learning_rate": 0.00015593220338983051, |
| "logits/chosen": -0.588573694229126, |
| "logits/rejected": -0.5898380875587463, |
| "logps/chosen": -8.49618911743164, |
| "logps/rejected": -16.47633171081543, |
| "loss": 0.8776077032089233, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.3434315621852875, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.1189945787191391, |
| "rewards/margins": 0.5393712520599365, |
| "rewards/rejected": -0.42037665843963623, |
| "step": 46, |
| "train_speed(iter/s)": 0.005498 |
| }, |
| { |
| "epoch": 0.08030327299909232, |
| "grad_norm": 2.918926954269409, |
| "learning_rate": 0.00015932203389830508, |
| "logits/chosen": -0.5735058188438416, |
| "logits/rejected": -0.5742441415786743, |
| "logps/chosen": -10.007829666137695, |
| "logps/rejected": -15.225828170776367, |
| "loss": 0.9683640003204346, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.4518345892429352, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.29521822929382324, |
| "rewards/margins": 0.5080306529998779, |
| "rewards/rejected": -0.2128123939037323, |
| "step": 47, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.08201185327566875, |
| "grad_norm": 2.885613441467285, |
| "learning_rate": 0.00016271186440677967, |
| "logits/chosen": -0.5690568089485168, |
| "logits/rejected": -0.570725679397583, |
| "logps/chosen": -6.9214887619018555, |
| "logps/rejected": -17.073692321777344, |
| "loss": 0.8013411164283752, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.2747730612754822, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.1803244799375534, |
| "rewards/margins": 0.5454539060592651, |
| "rewards/rejected": -0.36512941122055054, |
| "step": 48, |
| "train_speed(iter/s)": 0.005502 |
| }, |
| { |
| "epoch": 0.08372043355224518, |
| "grad_norm": 2.8386495113372803, |
| "learning_rate": 0.00016610169491525423, |
| "logits/chosen": -0.5691187381744385, |
| "logits/rejected": -0.5723721981048584, |
| "logps/chosen": -6.647555828094482, |
| "logps/rejected": -18.68927764892578, |
| "loss": 0.731940746307373, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.31857162714004517, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.15964265167713165, |
| "rewards/margins": 0.8172457218170166, |
| "rewards/rejected": -0.6576029062271118, |
| "step": 49, |
| "train_speed(iter/s)": 0.005503 |
| }, |
| { |
| "epoch": 0.0854290138288216, |
| "grad_norm": 6.678228855133057, |
| "learning_rate": 0.00016949152542372882, |
| "logits/chosen": -0.569486141204834, |
| "logits/rejected": -0.5724078416824341, |
| "logps/chosen": -7.978659629821777, |
| "logps/rejected": -18.0915584564209, |
| "loss": 0.8816509246826172, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.41086509823799133, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.16925573348999023, |
| "rewards/margins": 0.6901466846466064, |
| "rewards/rejected": -0.520891010761261, |
| "step": 50, |
| "train_speed(iter/s)": 0.0055 |
| }, |
| { |
| "epoch": 0.08713759410539805, |
| "grad_norm": 3.4269354343414307, |
| "learning_rate": 0.00017288135593220342, |
| "logits/chosen": -0.6229482889175415, |
| "logits/rejected": -0.624045193195343, |
| "logps/chosen": -8.981767654418945, |
| "logps/rejected": -18.87827491760254, |
| "loss": 0.902818500995636, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.4753139615058899, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.12949834764003754, |
| "rewards/margins": 0.91382896900177, |
| "rewards/rejected": -0.7843306064605713, |
| "step": 51, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.08884617438197448, |
| "grad_norm": 3.730417013168335, |
| "learning_rate": 0.00017627118644067798, |
| "logits/chosen": -0.5907954573631287, |
| "logits/rejected": -0.5928065776824951, |
| "logps/chosen": -8.826777458190918, |
| "logps/rejected": -25.743032455444336, |
| "loss": 0.8241711854934692, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.4183700382709503, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.15070247650146484, |
| "rewards/margins": 1.0849279165267944, |
| "rewards/rejected": -0.9342254400253296, |
| "step": 52, |
| "train_speed(iter/s)": 0.00549 |
| }, |
| { |
| "epoch": 0.09055475465855091, |
| "grad_norm": 6.543994426727295, |
| "learning_rate": 0.00017966101694915257, |
| "logits/chosen": -0.5657305717468262, |
| "logits/rejected": -0.5669429898262024, |
| "logps/chosen": -11.976017951965332, |
| "logps/rejected": -18.324581146240234, |
| "loss": 1.1098811626434326, |
| "memory(GiB)": 46.08, |
| "nll_loss": 0.5012481808662415, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": -0.04995343089103699, |
| "rewards/margins": 0.5761421918869019, |
| "rewards/rejected": -0.6260955929756165, |
| "step": 53, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.09226333493512734, |
| "grad_norm": 7.451931476593018, |
| "learning_rate": 0.00018305084745762714, |
| "logits/chosen": -0.5736872553825378, |
| "logits/rejected": -0.5727863907814026, |
| "logps/chosen": -9.956096649169922, |
| "logps/rejected": -17.354209899902344, |
| "loss": 1.0036059617996216, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.35273733735084534, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": -0.07980930060148239, |
| "rewards/margins": 0.5828754901885986, |
| "rewards/rejected": -0.6626847982406616, |
| "step": 54, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.09397191521170377, |
| "grad_norm": 4.901066303253174, |
| "learning_rate": 0.0001864406779661017, |
| "logits/chosen": -0.5865222811698914, |
| "logits/rejected": -0.5857770442962646, |
| "logps/chosen": -6.749457359313965, |
| "logps/rejected": -19.966379165649414, |
| "loss": 0.836351215839386, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.30449604988098145, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.11927387118339539, |
| "rewards/margins": 0.7463572025299072, |
| "rewards/rejected": -0.6270833611488342, |
| "step": 55, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.09568049548828021, |
| "grad_norm": 3.0312962532043457, |
| "learning_rate": 0.0001898305084745763, |
| "logits/chosen": -0.5644817352294922, |
| "logits/rejected": -0.5671055912971497, |
| "logps/chosen": -6.525354862213135, |
| "logps/rejected": -18.556304931640625, |
| "loss": 0.7299084663391113, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2566152811050415, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.051953963935375214, |
| "rewards/margins": 0.8131421804428101, |
| "rewards/rejected": -0.7611882090568542, |
| "step": 56, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.09738907576485664, |
| "grad_norm": 6.161406517028809, |
| "learning_rate": 0.00019322033898305085, |
| "logits/chosen": -0.5638913512229919, |
| "logits/rejected": -0.5684786438941956, |
| "logps/chosen": -8.483434677124023, |
| "logps/rejected": -20.397701263427734, |
| "loss": 0.988010585308075, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4928280711174011, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.17585664987564087, |
| "rewards/margins": 0.6749651432037354, |
| "rewards/rejected": -0.4991084933280945, |
| "step": 57, |
| "train_speed(iter/s)": 0.00549 |
| }, |
| { |
| "epoch": 0.09909765604143307, |
| "grad_norm": 4.03005313873291, |
| "learning_rate": 0.00019661016949152545, |
| "logits/chosen": -0.5518775582313538, |
| "logits/rejected": -0.552839994430542, |
| "logps/chosen": -6.419961929321289, |
| "logps/rejected": -20.10483169555664, |
| "loss": 0.7270359396934509, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2987552881240845, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.23015271127223969, |
| "rewards/margins": 0.8427435159683228, |
| "rewards/rejected": -0.6125907897949219, |
| "step": 58, |
| "train_speed(iter/s)": 0.00549 |
| }, |
| { |
| "epoch": 0.1008062363180095, |
| "grad_norm": 2.9903881549835205, |
| "learning_rate": 0.0002, |
| "logits/chosen": -0.558972954750061, |
| "logits/rejected": -0.559365451335907, |
| "logps/chosen": -7.266337871551514, |
| "logps/rejected": -19.700899124145508, |
| "loss": 0.8035504221916199, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.35604721307754517, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.23076069355010986, |
| "rewards/margins": 0.6581640243530273, |
| "rewards/rejected": -0.4274032711982727, |
| "step": 59, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.10251481659458593, |
| "grad_norm": 2.308953046798706, |
| "learning_rate": 0.00019999960020133237, |
| "logits/chosen": -0.5721818804740906, |
| "logits/rejected": -0.5736841559410095, |
| "logps/chosen": -5.192386150360107, |
| "logps/rejected": -16.873666763305664, |
| "loss": 0.6971458196640015, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.25499823689460754, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.14774081110954285, |
| "rewards/margins": 0.7478816509246826, |
| "rewards/rejected": -0.6001408696174622, |
| "step": 60, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.10422339687116237, |
| "grad_norm": 6.15336275100708, |
| "learning_rate": 0.00019999840080852627, |
| "logits/chosen": -0.6088159680366516, |
| "logits/rejected": -0.6091991662979126, |
| "logps/chosen": -10.085458755493164, |
| "logps/rejected": -16.714794158935547, |
| "loss": 1.2082210779190063, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6319302320480347, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": -0.015833575278520584, |
| "rewards/margins": 0.46795743703842163, |
| "rewards/rejected": -0.4837909936904907, |
| "step": 61, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.1059319771477388, |
| "grad_norm": 4.543377876281738, |
| "learning_rate": 0.00019999640183117196, |
| "logits/chosen": -0.6160375475883484, |
| "logits/rejected": -0.6183785796165466, |
| "logps/chosen": -7.354756832122803, |
| "logps/rejected": -17.332752227783203, |
| "loss": 0.9280235767364502, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.39269304275512695, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.00012398138642311096, |
| "rewards/margins": 0.7065204381942749, |
| "rewards/rejected": -0.7063965797424316, |
| "step": 62, |
| "train_speed(iter/s)": 0.005497 |
| }, |
| { |
| "epoch": 0.10764055742431523, |
| "grad_norm": 3.700834035873413, |
| "learning_rate": 0.00019999360328525325, |
| "logits/chosen": -0.5955262780189514, |
| "logits/rejected": -0.5960131287574768, |
| "logps/chosen": -7.038027763366699, |
| "logps/rejected": -19.62297248840332, |
| "loss": 0.8426406383514404, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.304790735244751, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.04702156037092209, |
| "rewards/margins": 0.6756925582885742, |
| "rewards/rejected": -0.6286709904670715, |
| "step": 63, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.10934913770089166, |
| "grad_norm": 3.616783857345581, |
| "learning_rate": 0.00019999000519314722, |
| "logits/chosen": -0.5933761596679688, |
| "logits/rejected": -0.5956672430038452, |
| "logps/chosen": -10.029403686523438, |
| "logps/rejected": -19.743268966674805, |
| "loss": 0.8910276889801025, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4178580045700073, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.059932250529527664, |
| "rewards/margins": 0.7701122164726257, |
| "rewards/rejected": -0.7101800441741943, |
| "step": 64, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.11105771797746809, |
| "grad_norm": 2.8669040203094482, |
| "learning_rate": 0.00019998560758362416, |
| "logits/chosen": -0.6144078373908997, |
| "logits/rejected": -0.6167560815811157, |
| "logps/chosen": -10.685945510864258, |
| "logps/rejected": -23.724315643310547, |
| "loss": 0.7696695923805237, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3780427873134613, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1446903944015503, |
| "rewards/margins": 1.0186232328414917, |
| "rewards/rejected": -0.873932957649231, |
| "step": 65, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.11276629825404454, |
| "grad_norm": 4.312869548797607, |
| "learning_rate": 0.00019998041049184719, |
| "logits/chosen": -0.620762050151825, |
| "logits/rejected": -0.6244264245033264, |
| "logps/chosen": -11.728214263916016, |
| "logps/rejected": -26.204395294189453, |
| "loss": 0.7992648482322693, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4216463565826416, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.14800474047660828, |
| "rewards/margins": 1.0791853666305542, |
| "rewards/rejected": -0.9311806559562683, |
| "step": 66, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.11447487853062097, |
| "grad_norm": 7.865874290466309, |
| "learning_rate": 0.00019997441395937213, |
| "logits/chosen": -0.6298569440841675, |
| "logits/rejected": -0.6332789659500122, |
| "logps/chosen": -7.128231525421143, |
| "logps/rejected": -22.970563888549805, |
| "loss": 0.7514320015907288, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.33478716015815735, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.05204010754823685, |
| "rewards/margins": 0.9836788773536682, |
| "rewards/rejected": -0.931638777256012, |
| "step": 67, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.1161834588071974, |
| "grad_norm": 6.099157810211182, |
| "learning_rate": 0.0001999676180341471, |
| "logits/chosen": -0.6536255478858948, |
| "logits/rejected": -0.6564731597900391, |
| "logps/chosen": -10.524028778076172, |
| "logps/rejected": -19.91427993774414, |
| "loss": 1.1307811737060547, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.7356545925140381, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.028077581897377968, |
| "rewards/margins": 0.9777989387512207, |
| "rewards/rejected": -0.9497213363647461, |
| "step": 68, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.11789203908377383, |
| "grad_norm": 5.27453088760376, |
| "learning_rate": 0.00019996002277051218, |
| "logits/chosen": -0.6637449264526367, |
| "logits/rejected": -0.6651371717453003, |
| "logps/chosen": -9.458850860595703, |
| "logps/rejected": -16.127042770385742, |
| "loss": 1.4500970840454102, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.8368474841117859, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.15293718874454498, |
| "rewards/margins": 0.500515878200531, |
| "rewards/rejected": -0.6534531116485596, |
| "step": 69, |
| "train_speed(iter/s)": 0.005497 |
| }, |
| { |
| "epoch": 0.11960061936035025, |
| "grad_norm": 4.469899654388428, |
| "learning_rate": 0.00019995162822919883, |
| "logits/chosen": -0.6642380356788635, |
| "logits/rejected": -0.664824366569519, |
| "logps/chosen": -8.725358963012695, |
| "logps/rejected": -16.164390563964844, |
| "loss": 0.8428195714950562, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3803461194038391, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.08946336805820465, |
| "rewards/margins": 0.7338303327560425, |
| "rewards/rejected": -0.644366979598999, |
| "step": 70, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.12130919963692668, |
| "grad_norm": 4.995606422424316, |
| "learning_rate": 0.00019994243447732957, |
| "logits/chosen": -0.6431492567062378, |
| "logits/rejected": -0.6473235487937927, |
| "logps/chosen": -8.270989418029785, |
| "logps/rejected": -24.483354568481445, |
| "loss": 0.8698587417602539, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4563717544078827, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.07262411713600159, |
| "rewards/margins": 1.0772123336791992, |
| "rewards/rejected": -1.0045881271362305, |
| "step": 71, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.12301777991350313, |
| "grad_norm": 3.5250916481018066, |
| "learning_rate": 0.00019993244158841747, |
| "logits/chosen": -0.6772405505180359, |
| "logits/rejected": -0.6789288520812988, |
| "logps/chosen": -9.19353199005127, |
| "logps/rejected": -27.746652603149414, |
| "loss": 0.8238008618354797, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.39661213755607605, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.09302112460136414, |
| "rewards/margins": 1.367937684059143, |
| "rewards/rejected": -1.274916410446167, |
| "step": 72, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.12472636019007956, |
| "grad_norm": 3.276311159133911, |
| "learning_rate": 0.00019992164964236533, |
| "logits/chosen": -0.6857064962387085, |
| "logits/rejected": -0.685559868812561, |
| "logps/chosen": -11.672329902648926, |
| "logps/rejected": -21.510480880737305, |
| "loss": 0.9018468856811523, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.41466090083122253, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.07735104858875275, |
| "rewards/margins": 0.9590628147125244, |
| "rewards/rejected": -0.8817118406295776, |
| "step": 73, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.126434940466656, |
| "grad_norm": 3.369281768798828, |
| "learning_rate": 0.00019991005872546527, |
| "logits/chosen": -0.6608355641365051, |
| "logits/rejected": -0.6626392602920532, |
| "logps/chosen": -7.696371078491211, |
| "logps/rejected": -16.016326904296875, |
| "loss": 0.8347861766815186, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3552168607711792, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.13171803951263428, |
| "rewards/margins": 0.7066640853881836, |
| "rewards/rejected": -0.5749460458755493, |
| "step": 74, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.12814352074323243, |
| "grad_norm": 2.894788980484009, |
| "learning_rate": 0.00019989766893039804, |
| "logits/chosen": -0.6222079992294312, |
| "logits/rejected": -0.6232355833053589, |
| "logps/chosen": -5.741198539733887, |
| "logps/rejected": -21.31194305419922, |
| "loss": 0.7077224850654602, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2760768234729767, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.08473706245422363, |
| "rewards/margins": 0.9282934665679932, |
| "rewards/rejected": -0.8435564041137695, |
| "step": 75, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.12985210101980885, |
| "grad_norm": 2.1586077213287354, |
| "learning_rate": 0.00019988448035623203, |
| "logits/chosen": -0.6409361958503723, |
| "logits/rejected": -0.6418861150741577, |
| "logps/chosen": -7.263876438140869, |
| "logps/rejected": -17.95524024963379, |
| "loss": 0.8056262731552124, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2723458409309387, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.10251796990633011, |
| "rewards/margins": 0.5812361240386963, |
| "rewards/rejected": -0.4787181317806244, |
| "step": 76, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.1315606812963853, |
| "grad_norm": 3.0591204166412354, |
| "learning_rate": 0.00019987049310842272, |
| "logits/chosen": -0.6324824690818787, |
| "logits/rejected": -0.6336201429367065, |
| "logps/chosen": -10.325702667236328, |
| "logps/rejected": -15.849100112915039, |
| "loss": 0.8897943496704102, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3855420649051666, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.16009055078029633, |
| "rewards/margins": 0.6010256409645081, |
| "rewards/rejected": -0.4409351050853729, |
| "step": 77, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.1332692615729617, |
| "grad_norm": 2.668419122695923, |
| "learning_rate": 0.00019985570729881184, |
| "logits/chosen": -0.6462000012397766, |
| "logits/rejected": -0.6490585207939148, |
| "logps/chosen": -5.475035667419434, |
| "logps/rejected": -20.422216415405273, |
| "loss": 0.7187508344650269, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.26306530833244324, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.13646253943443298, |
| "rewards/margins": 0.6754783391952515, |
| "rewards/rejected": -0.5390157699584961, |
| "step": 78, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.13497784184953815, |
| "grad_norm": 4.360518932342529, |
| "learning_rate": 0.00019984012304562622, |
| "logits/chosen": -0.6780133247375488, |
| "logits/rejected": -0.6758167147636414, |
| "logps/chosen": -11.310664176940918, |
| "logps/rejected": -15.12096881866455, |
| "loss": 0.9152082800865173, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.45073550939559937, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.15347006916999817, |
| "rewards/margins": 0.7282993197441101, |
| "rewards/rejected": -0.5748292207717896, |
| "step": 79, |
| "train_speed(iter/s)": 0.00549 |
| }, |
| { |
| "epoch": 0.1366864221261146, |
| "grad_norm": 2.7597768306732178, |
| "learning_rate": 0.00019982374047347723, |
| "logits/chosen": -0.6909894943237305, |
| "logits/rejected": -0.694047212600708, |
| "logps/chosen": -6.004232406616211, |
| "logps/rejected": -23.221824645996094, |
| "loss": 0.6792643666267395, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.31354260444641113, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.21520498394966125, |
| "rewards/margins": 1.1965129375457764, |
| "rewards/rejected": -0.9813080430030823, |
| "step": 80, |
| "train_speed(iter/s)": 0.00549 |
| }, |
| { |
| "epoch": 0.138395002402691, |
| "grad_norm": 4.00593900680542, |
| "learning_rate": 0.00019980655971335945, |
| "logits/chosen": -0.666761577129364, |
| "logits/rejected": -0.6698122620582581, |
| "logps/chosen": -11.200495719909668, |
| "logps/rejected": -20.89803123474121, |
| "loss": 1.0373343229293823, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5717941522598267, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.03891691192984581, |
| "rewards/margins": 0.8691489696502686, |
| "rewards/rejected": -0.9080657958984375, |
| "step": 81, |
| "train_speed(iter/s)": 0.00549 |
| }, |
| { |
| "epoch": 0.14010358267926745, |
| "grad_norm": 3.7606523036956787, |
| "learning_rate": 0.00019978858090264975, |
| "logits/chosen": -0.6808147430419922, |
| "logits/rejected": -0.683088481426239, |
| "logps/chosen": -8.414874076843262, |
| "logps/rejected": -28.505592346191406, |
| "loss": 0.7084274888038635, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3828541338443756, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1215677559375763, |
| "rewards/margins": 1.7334877252578735, |
| "rewards/rejected": -1.6119199991226196, |
| "step": 82, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.14181216295584387, |
| "grad_norm": 4.108179569244385, |
| "learning_rate": 0.0001997698041851063, |
| "logits/chosen": -0.6897909045219421, |
| "logits/rejected": -0.6913143396377563, |
| "logps/chosen": -4.813051223754883, |
| "logps/rejected": -26.06230926513672, |
| "loss": 0.572483241558075, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.24444976449012756, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.15832601487636566, |
| "rewards/margins": 1.6045297384262085, |
| "rewards/rejected": -1.4462038278579712, |
| "step": 83, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.1435207432324203, |
| "grad_norm": 5.49106502532959, |
| "learning_rate": 0.00019975022971086714, |
| "logits/chosen": -0.6466387510299683, |
| "logits/rejected": -0.6520243883132935, |
| "logps/chosen": -7.595500946044922, |
| "logps/rejected": -33.218482971191406, |
| "loss": 0.9090702533721924, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5250993371009827, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.13341139256954193, |
| "rewards/margins": 1.5776695013046265, |
| "rewards/rejected": -1.444258213043213, |
| "step": 84, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.14522932350899675, |
| "grad_norm": 4.539761543273926, |
| "learning_rate": 0.00019972985763644932, |
| "logits/chosen": -0.6421160101890564, |
| "logits/rejected": -0.6411360502243042, |
| "logps/chosen": -7.4049482345581055, |
| "logps/rejected": -21.37047576904297, |
| "loss": 0.8407540917396545, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4470757246017456, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.1697394847869873, |
| "rewards/margins": 1.345821738243103, |
| "rewards/rejected": -1.1760823726654053, |
| "step": 85, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.14693790378557317, |
| "grad_norm": 4.877723217010498, |
| "learning_rate": 0.00019970868812474736, |
| "logits/chosen": -0.6163697242736816, |
| "logits/rejected": -0.6189360022544861, |
| "logps/chosen": -12.07491683959961, |
| "logps/rejected": -29.372955322265625, |
| "loss": 0.848598301410675, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.49112260341644287, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.1017676591873169, |
| "rewards/margins": 1.7578887939453125, |
| "rewards/rejected": -1.6561211347579956, |
| "step": 86, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.14864648406214961, |
| "grad_norm": 5.431716442108154, |
| "learning_rate": 0.00019968672134503213, |
| "logits/chosen": -0.6225081086158752, |
| "logits/rejected": -0.6252766251564026, |
| "logps/chosen": -7.893375873565674, |
| "logps/rejected": -24.980260848999023, |
| "loss": 0.7993619441986084, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3846016228199005, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.16642117500305176, |
| "rewards/margins": 1.3823862075805664, |
| "rewards/rejected": -1.215964913368225, |
| "step": 87, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.15035506433872603, |
| "grad_norm": 3.7740750312805176, |
| "learning_rate": 0.00019966395747294938, |
| "logits/chosen": -0.5984699130058289, |
| "logits/rejected": -0.6030604839324951, |
| "logps/chosen": -5.385066032409668, |
| "logps/rejected": -25.168312072753906, |
| "loss": 0.6654453873634338, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3348524868488312, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.16900494694709778, |
| "rewards/margins": 1.3018345832824707, |
| "rewards/rejected": -1.1328295469284058, |
| "step": 88, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.15206364461530247, |
| "grad_norm": 3.6669890880584717, |
| "learning_rate": 0.0001996403966905185, |
| "logits/chosen": -0.6115779280662537, |
| "logits/rejected": -0.615218997001648, |
| "logps/chosen": -8.385238647460938, |
| "logps/rejected": -20.905569076538086, |
| "loss": 0.8084090948104858, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.39410629868507385, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.07257324457168579, |
| "rewards/margins": 1.0608566999435425, |
| "rewards/rejected": -0.9882834553718567, |
| "step": 89, |
| "train_speed(iter/s)": 0.00549 |
| }, |
| { |
| "epoch": 0.15377222489187892, |
| "grad_norm": 8.954024314880371, |
| "learning_rate": 0.00019961603918613077, |
| "logits/chosen": -0.6033557653427124, |
| "logits/rejected": -0.6063033938407898, |
| "logps/chosen": -7.85408878326416, |
| "logps/rejected": -24.28641700744629, |
| "loss": 0.8988556861877441, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4972013235092163, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.1499406397342682, |
| "rewards/margins": 1.2390244007110596, |
| "rewards/rejected": -1.0890836715698242, |
| "step": 90, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.15548080516845533, |
| "grad_norm": 2.68121075630188, |
| "learning_rate": 0.00019959088515454827, |
| "logits/chosen": -0.6473177075386047, |
| "logits/rejected": -0.6510601043701172, |
| "logps/chosen": -6.986569404602051, |
| "logps/rejected": -28.0955810546875, |
| "loss": 0.6454746723175049, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3593181073665619, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.310781866312027, |
| "rewards/margins": 1.7279632091522217, |
| "rewards/rejected": -1.417181372642517, |
| "step": 91, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.15718938544503178, |
| "grad_norm": 7.000541687011719, |
| "learning_rate": 0.0001995649347969019, |
| "logits/chosen": -0.7174542546272278, |
| "logits/rejected": -0.7201290726661682, |
| "logps/chosen": -11.954474449157715, |
| "logps/rejected": -24.34893226623535, |
| "loss": 1.0948961973190308, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6069883704185486, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.005067221820354462, |
| "rewards/margins": 1.2273277044296265, |
| "rewards/rejected": -1.232394814491272, |
| "step": 92, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.1588979657216082, |
| "grad_norm": 4.3811211585998535, |
| "learning_rate": 0.00019953818832069, |
| "logits/chosen": -0.6739280223846436, |
| "logits/rejected": -0.6795160174369812, |
| "logps/chosen": -6.881966590881348, |
| "logps/rejected": -37.4478645324707, |
| "loss": 0.6469016075134277, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.40633073449134827, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.20596878230571747, |
| "rewards/margins": 2.5138516426086426, |
| "rewards/rejected": -2.3078830242156982, |
| "step": 93, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.16060654599818464, |
| "grad_norm": 12.26343822479248, |
| "learning_rate": 0.00019951064593977668, |
| "logits/chosen": -0.6859129667282104, |
| "logits/rejected": -0.6899083852767944, |
| "logps/chosen": -13.38476848602295, |
| "logps/rejected": -30.595182418823242, |
| "loss": 1.2918721437454224, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6223256587982178, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": -0.30435577034950256, |
| "rewards/margins": 1.605468988418579, |
| "rewards/rejected": -1.9098248481750488, |
| "step": 94, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.16231512627476108, |
| "grad_norm": 6.570938587188721, |
| "learning_rate": 0.00019948230787439017, |
| "logits/chosen": -0.7028453946113586, |
| "logits/rejected": -0.7050482034683228, |
| "logps/chosen": -6.761815071105957, |
| "logps/rejected": -27.314367294311523, |
| "loss": 0.9538392424583435, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.39429235458374023, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.012213893234729767, |
| "rewards/margins": 1.6957728862762451, |
| "rewards/rejected": -1.6835590600967407, |
| "step": 95, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.1640237065513375, |
| "grad_norm": 2.8915793895721436, |
| "learning_rate": 0.0001994531743511208, |
| "logits/chosen": -0.6953604817390442, |
| "logits/rejected": -0.6982015371322632, |
| "logps/chosen": -7.192444801330566, |
| "logps/rejected": -26.632644653320312, |
| "loss": 0.6771315336227417, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3128400444984436, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.21012118458747864, |
| "rewards/margins": 1.6979337930679321, |
| "rewards/rejected": -1.4878127574920654, |
| "step": 96, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.16573228682791394, |
| "grad_norm": 2.7815051078796387, |
| "learning_rate": 0.00019942324560291952, |
| "logits/chosen": -0.6696122884750366, |
| "logits/rejected": -0.6713677644729614, |
| "logps/chosen": -4.250362396240234, |
| "logps/rejected": -22.2177791595459, |
| "loss": 0.6821787357330322, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.26163250207901, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.09910638630390167, |
| "rewards/margins": 1.2715519666671753, |
| "rewards/rejected": -1.1724457740783691, |
| "step": 97, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.16744086710449035, |
| "grad_norm": 2.4311976432800293, |
| "learning_rate": 0.00019939252186909574, |
| "logits/chosen": -0.6763203740119934, |
| "logits/rejected": -0.6752580404281616, |
| "logps/chosen": -6.972185134887695, |
| "logps/rejected": -21.0667667388916, |
| "loss": 0.5966892838478088, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2420358657836914, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.14732374250888824, |
| "rewards/margins": 1.0790168046951294, |
| "rewards/rejected": -0.9316931366920471, |
| "step": 98, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.1691494473810668, |
| "grad_norm": 5.822193622589111, |
| "learning_rate": 0.00019936100339531564, |
| "logits/chosen": -0.6894733905792236, |
| "logits/rejected": -0.6909453272819519, |
| "logps/chosen": -6.009974479675293, |
| "logps/rejected": -21.0426025390625, |
| "loss": 0.7569796442985535, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4093407690525055, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.26579749584198, |
| "rewards/margins": 1.3290454149246216, |
| "rewards/rejected": -1.0632479190826416, |
| "step": 99, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.1708580276576432, |
| "grad_norm": 3.2212934494018555, |
| "learning_rate": 0.0001993286904336001, |
| "logits/chosen": -0.7202757000923157, |
| "logits/rejected": -0.7197086215019226, |
| "logps/chosen": -6.131860733032227, |
| "logps/rejected": -21.27735137939453, |
| "loss": 0.622944712638855, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.28273510932922363, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.26464149355888367, |
| "rewards/margins": 1.1921066045761108, |
| "rewards/rejected": -0.9274651408195496, |
| "step": 100, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.17256660793421966, |
| "grad_norm": 5.42195463180542, |
| "learning_rate": 0.00019929558324232268, |
| "logits/chosen": -0.7189818620681763, |
| "logits/rejected": -0.720359742641449, |
| "logps/chosen": -8.62668228149414, |
| "logps/rejected": -19.183561325073242, |
| "loss": 0.9257504940032959, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5092519521713257, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.09434813261032104, |
| "rewards/margins": 0.8864970803260803, |
| "rewards/rejected": -0.7921488881111145, |
| "step": 101, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.1742751882107961, |
| "grad_norm": 6.533127307891846, |
| "learning_rate": 0.0001992616820862076, |
| "logits/chosen": -0.6889673471450806, |
| "logits/rejected": -0.6902090907096863, |
| "logps/chosen": -9.093452453613281, |
| "logps/rejected": -19.451143264770508, |
| "loss": 0.9490825533866882, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.45314493775367737, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.03901132568717003, |
| "rewards/margins": 0.7304368019104004, |
| "rewards/rejected": -0.6914253830909729, |
| "step": 102, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.17598376848737252, |
| "grad_norm": 3.991544485092163, |
| "learning_rate": 0.00019922698723632767, |
| "logits/chosen": -0.7347920536994934, |
| "logits/rejected": -0.7329245209693909, |
| "logps/chosen": -10.086681365966797, |
| "logps/rejected": -24.030868530273438, |
| "loss": 0.8895858526229858, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5397864580154419, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.12226971983909607, |
| "rewards/margins": 1.3651487827301025, |
| "rewards/rejected": -1.2428791522979736, |
| "step": 103, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.17769234876394896, |
| "grad_norm": 4.0011162757873535, |
| "learning_rate": 0.0001991914989701019, |
| "logits/chosen": -0.7434298992156982, |
| "logits/rejected": -0.7457666993141174, |
| "logps/chosen": -8.058287620544434, |
| "logps/rejected": -20.040809631347656, |
| "loss": 0.7681674957275391, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3557548522949219, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.12300814688205719, |
| "rewards/margins": 1.063488483428955, |
| "rewards/rejected": -0.9404802918434143, |
| "step": 104, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.17940092904052538, |
| "grad_norm": 3.329845905303955, |
| "learning_rate": 0.00019915521757129354, |
| "logits/chosen": -0.7461959719657898, |
| "logits/rejected": -0.7488771677017212, |
| "logps/chosen": -7.418200969696045, |
| "logps/rejected": -25.79469108581543, |
| "loss": 0.6970200538635254, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.36587315797805786, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.12871423363685608, |
| "rewards/margins": 1.5802878141403198, |
| "rewards/rejected": -1.4515737295150757, |
| "step": 105, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.18110950931710182, |
| "grad_norm": 4.27202033996582, |
| "learning_rate": 0.00019911814333000772, |
| "logits/chosen": -0.7183383703231812, |
| "logits/rejected": -0.7201208472251892, |
| "logps/chosen": -9.424932479858398, |
| "logps/rejected": -25.1419677734375, |
| "loss": 0.8681972622871399, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4449956715106964, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.04691670835018158, |
| "rewards/margins": 1.2326714992523193, |
| "rewards/rejected": -1.1857548952102661, |
| "step": 106, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.18281808959367826, |
| "grad_norm": 4.578512191772461, |
| "learning_rate": 0.00019908027654268903, |
| "logits/chosen": -0.7675899267196655, |
| "logits/rejected": -0.7682605385780334, |
| "logps/chosen": -8.889389038085938, |
| "logps/rejected": -24.62353515625, |
| "loss": 0.8598517179489136, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.40520450472831726, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.13606908917427063, |
| "rewards/margins": 1.3364840745925903, |
| "rewards/rejected": -1.200415015220642, |
| "step": 107, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.18452666987025468, |
| "grad_norm": 3.067852020263672, |
| "learning_rate": 0.00019904161751211938, |
| "logits/chosen": -0.7350285649299622, |
| "logits/rejected": -0.7360118627548218, |
| "logps/chosen": -12.491768836975098, |
| "logps/rejected": -23.78077507019043, |
| "loss": 0.8337110280990601, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4792723059654236, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0716579332947731, |
| "rewards/margins": 1.2602936029434204, |
| "rewards/rejected": -1.1886355876922607, |
| "step": 108, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.18623525014683112, |
| "grad_norm": 3.2055752277374268, |
| "learning_rate": 0.00019900216654741528, |
| "logits/chosen": -0.7312414646148682, |
| "logits/rejected": -0.7316892147064209, |
| "logps/chosen": -8.460771560668945, |
| "logps/rejected": -21.30421257019043, |
| "loss": 0.7292563319206238, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3923528492450714, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.2296271026134491, |
| "rewards/margins": 1.3531551361083984, |
| "rewards/rejected": -1.1235278844833374, |
| "step": 109, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.18794383042340754, |
| "grad_norm": 4.93084716796875, |
| "learning_rate": 0.0001989619239640256, |
| "logits/chosen": -0.7337685823440552, |
| "logits/rejected": -0.7344739437103271, |
| "logps/chosen": -11.766962051391602, |
| "logps/rejected": -22.844139099121094, |
| "loss": 0.8851971626281738, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4795154333114624, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.06328105181455612, |
| "rewards/margins": 1.133680820465088, |
| "rewards/rejected": -1.0703996419906616, |
| "step": 110, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.18965241069998398, |
| "grad_norm": 10.518948554992676, |
| "learning_rate": 0.00019892089008372897, |
| "logits/chosen": -0.709618330001831, |
| "logits/rejected": -0.7111095190048218, |
| "logps/chosen": -8.326457977294922, |
| "logps/rejected": -27.638351440429688, |
| "loss": 0.9548771977424622, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5345268249511719, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.045853935182094574, |
| "rewards/margins": 1.551784634590149, |
| "rewards/rejected": -1.5059306621551514, |
| "step": 111, |
| "train_speed(iter/s)": 0.005497 |
| }, |
| { |
| "epoch": 0.19136099097656042, |
| "grad_norm": 7.465288162231445, |
| "learning_rate": 0.0001988790652346312, |
| "logits/chosen": -0.7286843657493591, |
| "logits/rejected": -0.729621946811676, |
| "logps/chosen": -8.483978271484375, |
| "logps/rejected": -28.15561294555664, |
| "loss": 0.922063410282135, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5344014167785645, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.13854525983333588, |
| "rewards/margins": 1.4421772956848145, |
| "rewards/rejected": -1.3036320209503174, |
| "step": 112, |
| "train_speed(iter/s)": 0.005498 |
| }, |
| { |
| "epoch": 0.19306957125313684, |
| "grad_norm": 8.20088005065918, |
| "learning_rate": 0.0001988364497511627, |
| "logits/chosen": -0.740546703338623, |
| "logits/rejected": -0.7427754402160645, |
| "logps/chosen": -8.096268653869629, |
| "logps/rejected": -22.899229049682617, |
| "loss": 0.9872632622718811, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5750774145126343, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.18856163322925568, |
| "rewards/margins": 1.2329988479614258, |
| "rewards/rejected": -1.0444371700286865, |
| "step": 113, |
| "train_speed(iter/s)": 0.005497 |
| }, |
| { |
| "epoch": 0.19477815152971328, |
| "grad_norm": 3.4554946422576904, |
| "learning_rate": 0.0001987930439740757, |
| "logits/chosen": -0.699242115020752, |
| "logits/rejected": -0.7016286849975586, |
| "logps/chosen": -11.758835792541504, |
| "logps/rejected": -22.287220001220703, |
| "loss": 0.7449758052825928, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4075641632080078, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1979522556066513, |
| "rewards/margins": 1.2950595617294312, |
| "rewards/rejected": -1.097107172012329, |
| "step": 114, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.1964867318062897, |
| "grad_norm": 2.8801109790802, |
| "learning_rate": 0.00019874884825044165, |
| "logits/chosen": -0.673065185546875, |
| "logits/rejected": -0.6760713458061218, |
| "logps/chosen": -6.409548759460449, |
| "logps/rejected": -16.89990234375, |
| "loss": 0.7451252937316895, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.31410443782806396, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.19196031987667084, |
| "rewards/margins": 0.9307445287704468, |
| "rewards/rejected": -0.7387841939926147, |
| "step": 115, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.19819531208286614, |
| "grad_norm": 2.597715377807617, |
| "learning_rate": 0.00019870386293364836, |
| "logits/chosen": -0.642209529876709, |
| "logits/rejected": -0.6442127227783203, |
| "logps/chosen": -6.406985282897949, |
| "logps/rejected": -19.21172523498535, |
| "loss": 0.7881473898887634, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3131154477596283, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.14077159762382507, |
| "rewards/margins": 0.9926955699920654, |
| "rewards/rejected": -0.851923942565918, |
| "step": 116, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.1999038923594426, |
| "grad_norm": 2.9220879077911377, |
| "learning_rate": 0.00019865808838339727, |
| "logits/chosen": -0.6920750141143799, |
| "logits/rejected": -0.6941624283790588, |
| "logps/chosen": -7.081650733947754, |
| "logps/rejected": -24.63327407836914, |
| "loss": 0.6955921649932861, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.29941731691360474, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.014344778843224049, |
| "rewards/margins": 1.21220862865448, |
| "rewards/rejected": -1.1978638172149658, |
| "step": 117, |
| "train_speed(iter/s)": 0.005499 |
| }, |
| { |
| "epoch": 0.201612472636019, |
| "grad_norm": 2.535001277923584, |
| "learning_rate": 0.00019861152496570043, |
| "logits/chosen": -0.6474766135215759, |
| "logits/rejected": -0.6474269032478333, |
| "logps/chosen": -4.845932483673096, |
| "logps/rejected": -30.43771743774414, |
| "loss": 0.549997866153717, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2411210685968399, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.22833456099033356, |
| "rewards/margins": 1.7972755432128906, |
| "rewards/rejected": -1.5689408779144287, |
| "step": 118, |
| "train_speed(iter/s)": 0.005498 |
| }, |
| { |
| "epoch": 0.20332105291259545, |
| "grad_norm": 2.592416286468506, |
| "learning_rate": 0.00019856417305287772, |
| "logits/chosen": -0.6813372373580933, |
| "logits/rejected": -0.684921383857727, |
| "logps/chosen": -8.64324951171875, |
| "logps/rejected": -25.918020248413086, |
| "loss": 0.6950156092643738, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.34918904304504395, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.27043017745018005, |
| "rewards/margins": 1.5927720069885254, |
| "rewards/rejected": -1.322341799736023, |
| "step": 119, |
| "train_speed(iter/s)": 0.005498 |
| }, |
| { |
| "epoch": 0.20502963318917186, |
| "grad_norm": 3.0879509449005127, |
| "learning_rate": 0.00019851603302355373, |
| "logits/chosen": -0.7103255987167358, |
| "logits/rejected": -0.7141609191894531, |
| "logps/chosen": -7.374659061431885, |
| "logps/rejected": -27.09424591064453, |
| "loss": 0.7110822200775146, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3552224040031433, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.15158146619796753, |
| "rewards/margins": 1.6895453929901123, |
| "rewards/rejected": -1.5379639863967896, |
| "step": 120, |
| "train_speed(iter/s)": 0.005499 |
| }, |
| { |
| "epoch": 0.2067382134657483, |
| "grad_norm": 3.703174591064453, |
| "learning_rate": 0.00019846710526265487, |
| "logits/chosen": -0.7088285088539124, |
| "logits/rejected": -0.7127594947814941, |
| "logps/chosen": -6.05606746673584, |
| "logps/rejected": -27.269733428955078, |
| "loss": 0.619713306427002, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.333889901638031, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2516608238220215, |
| "rewards/margins": 1.8369193077087402, |
| "rewards/rejected": -1.5852584838867188, |
| "step": 121, |
| "train_speed(iter/s)": 0.0055 |
| }, |
| { |
| "epoch": 0.20844679374232475, |
| "grad_norm": 4.171956539154053, |
| "learning_rate": 0.00019841739016140624, |
| "logits/chosen": -0.7836604118347168, |
| "logits/rejected": -0.7841136455535889, |
| "logps/chosen": -7.950913429260254, |
| "logps/rejected": -26.353294372558594, |
| "loss": 0.7199419736862183, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3016027510166168, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.03286279737949371, |
| "rewards/margins": 1.4103941917419434, |
| "rewards/rejected": -1.3775315284729004, |
| "step": 122, |
| "train_speed(iter/s)": 0.0055 |
| }, |
| { |
| "epoch": 0.21015537401890116, |
| "grad_norm": 5.59217643737793, |
| "learning_rate": 0.0001983668881173284, |
| "logits/chosen": -0.7123353481292725, |
| "logits/rejected": -0.7176805734634399, |
| "logps/chosen": -6.9763007164001465, |
| "logps/rejected": -40.705894470214844, |
| "loss": 0.5055665969848633, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2759905159473419, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.13582277297973633, |
| "rewards/margins": 2.353527069091797, |
| "rewards/rejected": -2.2177045345306396, |
| "step": 123, |
| "train_speed(iter/s)": 0.0055 |
| }, |
| { |
| "epoch": 0.2118639542954776, |
| "grad_norm": 7.052844524383545, |
| "learning_rate": 0.00019831559953423442, |
| "logits/chosen": -0.7205808758735657, |
| "logits/rejected": -0.7224647402763367, |
| "logps/chosen": -8.208627700805664, |
| "logps/rejected": -23.56304359436035, |
| "loss": 0.8278493285179138, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3575640022754669, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0022782105952501297, |
| "rewards/margins": 1.129638910293579, |
| "rewards/rejected": -1.1273607015609741, |
| "step": 124, |
| "train_speed(iter/s)": 0.0055 |
| }, |
| { |
| "epoch": 0.21357253457205402, |
| "grad_norm": 2.853487491607666, |
| "learning_rate": 0.00019826352482222638, |
| "logits/chosen": -0.7876097559928894, |
| "logits/rejected": -0.7900496125221252, |
| "logps/chosen": -6.370059013366699, |
| "logps/rejected": -29.205093383789062, |
| "loss": 0.553210437297821, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.24043366312980652, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.03330150991678238, |
| "rewards/margins": 1.9821114540100098, |
| "rewards/rejected": -1.9488096237182617, |
| "step": 125, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.21528111484863047, |
| "grad_norm": 3.3915181159973145, |
| "learning_rate": 0.00019821066439769234, |
| "logits/chosen": -0.7244046926498413, |
| "logits/rejected": -0.7288062572479248, |
| "logps/chosen": -7.876413822174072, |
| "logps/rejected": -35.36836242675781, |
| "loss": 0.5778641104698181, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.34048542380332947, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.027921676635742188, |
| "rewards/margins": 2.1973161697387695, |
| "rewards/rejected": -2.225238084793091, |
| "step": 126, |
| "train_speed(iter/s)": 0.0055 |
| }, |
| { |
| "epoch": 0.2169896951252069, |
| "grad_norm": 2.667902946472168, |
| "learning_rate": 0.00019815701868330284, |
| "logits/chosen": -0.7135522365570068, |
| "logits/rejected": -0.7176554799079895, |
| "logps/chosen": -11.242607116699219, |
| "logps/rejected": -33.319515228271484, |
| "loss": 0.7801873087882996, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.49927234649658203, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.23649244010448456, |
| "rewards/margins": 2.445638656616211, |
| "rewards/rejected": -2.209146022796631, |
| "step": 127, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.21869827540178333, |
| "grad_norm": 9.078632354736328, |
| "learning_rate": 0.00019810258810800752, |
| "logits/chosen": -0.7386202812194824, |
| "logits/rejected": -0.7396747469902039, |
| "logps/chosen": -9.633708000183105, |
| "logps/rejected": -30.443008422851562, |
| "loss": 0.7859803438186646, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.49833089113235474, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.04507865756750107, |
| "rewards/margins": 1.9784715175628662, |
| "rewards/rejected": -1.933392882347107, |
| "step": 128, |
| "train_speed(iter/s)": 0.005503 |
| }, |
| { |
| "epoch": 0.22040685567835977, |
| "grad_norm": 13.282059669494629, |
| "learning_rate": 0.0001980473731070319, |
| "logits/chosen": -0.7503476142883301, |
| "logits/rejected": -0.755096971988678, |
| "logps/chosen": -7.456392765045166, |
| "logps/rejected": -29.754714965820312, |
| "loss": 0.8027007579803467, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4415196180343628, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.14411740005016327, |
| "rewards/margins": 2.093838691711426, |
| "rewards/rejected": -1.949721336364746, |
| "step": 129, |
| "train_speed(iter/s)": 0.005503 |
| }, |
| { |
| "epoch": 0.22211543595493619, |
| "grad_norm": 5.136019706726074, |
| "learning_rate": 0.0001979913741218736, |
| "logits/chosen": -0.7441588640213013, |
| "logits/rejected": -0.7470377683639526, |
| "logps/chosen": -11.379145622253418, |
| "logps/rejected": -36.20026397705078, |
| "loss": 0.9255682229995728, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5846636891365051, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.014721579849720001, |
| "rewards/margins": 2.245670795440674, |
| "rewards/rejected": -2.2309491634368896, |
| "step": 130, |
| "train_speed(iter/s)": 0.005502 |
| }, |
| { |
| "epoch": 0.22382401623151263, |
| "grad_norm": 5.494217395782471, |
| "learning_rate": 0.00019793459160029903, |
| "logits/chosen": -0.7520757913589478, |
| "logits/rejected": -0.7574396133422852, |
| "logps/chosen": -6.406137466430664, |
| "logps/rejected": -42.744041442871094, |
| "loss": 0.6142405867576599, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3836834132671356, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.19515469670295715, |
| "rewards/margins": 3.062283992767334, |
| "rewards/rejected": -2.867129325866699, |
| "step": 131, |
| "train_speed(iter/s)": 0.005503 |
| }, |
| { |
| "epoch": 0.22553259650808907, |
| "grad_norm": 4.463568687438965, |
| "learning_rate": 0.00019787702599633971, |
| "logits/chosen": -0.7327495217323303, |
| "logits/rejected": -0.738735556602478, |
| "logps/chosen": -6.503672122955322, |
| "logps/rejected": -39.14170837402344, |
| "loss": 0.7878039479255676, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.37348490953445435, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.034972380846738815, |
| "rewards/margins": 2.3170716762542725, |
| "rewards/rejected": -2.352044105529785, |
| "step": 132, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.2272411767846655, |
| "grad_norm": 7.909404754638672, |
| "learning_rate": 0.0001978186777702887, |
| "logits/chosen": -0.6934617757797241, |
| "logits/rejected": -0.6948121786117554, |
| "logps/chosen": -7.257785797119141, |
| "logps/rejected": -27.908708572387695, |
| "loss": 0.7475972175598145, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3585331439971924, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.036370277404785156, |
| "rewards/margins": 1.666322946548462, |
| "rewards/rejected": -1.6299527883529663, |
| "step": 133, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.22894975706124193, |
| "grad_norm": 14.575544357299805, |
| "learning_rate": 0.00019775954738869683, |
| "logits/chosen": -0.6997110843658447, |
| "logits/rejected": -0.7026513814926147, |
| "logps/chosen": -7.946556091308594, |
| "logps/rejected": -20.762996673583984, |
| "loss": 1.056870460510254, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6803016066551208, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.16680209338665009, |
| "rewards/margins": 1.3951897621154785, |
| "rewards/rejected": -1.2283875942230225, |
| "step": 134, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.23065833733781835, |
| "grad_norm": 6.814887046813965, |
| "learning_rate": 0.00019769963532436911, |
| "logits/chosen": -0.661209762096405, |
| "logits/rejected": -0.6659980416297913, |
| "logps/chosen": -8.108284950256348, |
| "logps/rejected": -26.44216537475586, |
| "loss": 1.0938985347747803, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.7105523347854614, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.010204421356320381, |
| "rewards/margins": 1.3176696300506592, |
| "rewards/rejected": -1.3074650764465332, |
| "step": 135, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.2323669176143948, |
| "grad_norm": 5.490352153778076, |
| "learning_rate": 0.00019763894205636072, |
| "logits/chosen": -0.671430230140686, |
| "logits/rejected": -0.673378586769104, |
| "logps/chosen": -9.945783615112305, |
| "logps/rejected": -22.1285343170166, |
| "loss": 0.976132333278656, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.543598473072052, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.1351984441280365, |
| "rewards/margins": 1.1337149143218994, |
| "rewards/rejected": -0.9985164403915405, |
| "step": 136, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.23407549789097123, |
| "grad_norm": 5.260707378387451, |
| "learning_rate": 0.00019757746806997348, |
| "logits/chosen": -0.6116827130317688, |
| "logits/rejected": -0.6123840808868408, |
| "logps/chosen": -8.495226860046387, |
| "logps/rejected": -17.80582046508789, |
| "loss": 0.9422957897186279, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3756696879863739, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": -0.009481308050453663, |
| "rewards/margins": 0.6523244380950928, |
| "rewards/rejected": -0.6618058085441589, |
| "step": 137, |
| "train_speed(iter/s)": 0.005502 |
| }, |
| { |
| "epoch": 0.23578407816754765, |
| "grad_norm": 2.1559102535247803, |
| "learning_rate": 0.00019751521385675172, |
| "logits/chosen": -0.6613444089889526, |
| "logits/rejected": -0.6647210121154785, |
| "logps/chosen": -4.250008583068848, |
| "logps/rejected": -25.66912078857422, |
| "loss": 0.5518596172332764, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.1721974015235901, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.17000286281108856, |
| "rewards/margins": 1.491791009902954, |
| "rewards/rejected": -1.3217883110046387, |
| "step": 138, |
| "train_speed(iter/s)": 0.005502 |
| }, |
| { |
| "epoch": 0.2374926584441241, |
| "grad_norm": 2.9551846981048584, |
| "learning_rate": 0.00019745217991447848, |
| "logits/chosen": -0.6719025373458862, |
| "logits/rejected": -0.6741494536399841, |
| "logps/chosen": -8.146540641784668, |
| "logps/rejected": -28.308319091796875, |
| "loss": 0.7604531049728394, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.34055855870246887, |
| "rewards/accuracies": 0.71875, |
| "rewards/chosen": 0.1295825093984604, |
| "rewards/margins": 1.5078970193862915, |
| "rewards/rejected": -1.37831449508667, |
| "step": 139, |
| "train_speed(iter/s)": 0.005501 |
| }, |
| { |
| "epoch": 0.2392012387207005, |
| "grad_norm": 5.525161266326904, |
| "learning_rate": 0.0001973883667471715, |
| "logits/chosen": -0.6694546937942505, |
| "logits/rejected": -0.6737231612205505, |
| "logps/chosen": -10.553884506225586, |
| "logps/rejected": -32.512393951416016, |
| "loss": 0.822769284248352, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3502211570739746, |
| "rewards/accuracies": 0.6875, |
| "rewards/chosen": 0.01343974657356739, |
| "rewards/margins": 1.7657444477081299, |
| "rewards/rejected": -1.7523046731948853, |
| "step": 140, |
| "train_speed(iter/s)": 0.005502 |
| }, |
| { |
| "epoch": 0.24090981899727695, |
| "grad_norm": 2.4956538677215576, |
| "learning_rate": 0.00019732377486507915, |
| "logits/chosen": -0.6616899371147156, |
| "logits/rejected": -0.6654747724533081, |
| "logps/chosen": -6.116077423095703, |
| "logps/rejected": -22.98920440673828, |
| "loss": 0.7530378103256226, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.33646857738494873, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.09931784868240356, |
| "rewards/margins": 1.2228065729141235, |
| "rewards/rejected": -1.1234887838363647, |
| "step": 141, |
| "train_speed(iter/s)": 0.005502 |
| }, |
| { |
| "epoch": 0.24261839927385337, |
| "grad_norm": 3.0544512271881104, |
| "learning_rate": 0.00019725840478467636, |
| "logits/chosen": -0.6879534125328064, |
| "logits/rejected": -0.6862825155258179, |
| "logps/chosen": -8.299098014831543, |
| "logps/rejected": -18.718904495239258, |
| "loss": 0.7917457818984985, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.31696975231170654, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.11313430219888687, |
| "rewards/margins": 0.9731096029281616, |
| "rewards/rejected": -0.859975278377533, |
| "step": 142, |
| "train_speed(iter/s)": 0.005502 |
| }, |
| { |
| "epoch": 0.2443269795504298, |
| "grad_norm": 2.7635226249694824, |
| "learning_rate": 0.00019719225702866062, |
| "logits/chosen": -0.7383114099502563, |
| "logits/rejected": -0.736755907535553, |
| "logps/chosen": -5.972444534301758, |
| "logps/rejected": -27.483884811401367, |
| "loss": 0.6677893400192261, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3221934735774994, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.15736974775791168, |
| "rewards/margins": 1.7971253395080566, |
| "rewards/rejected": -1.639755368232727, |
| "step": 143, |
| "train_speed(iter/s)": 0.005503 |
| }, |
| { |
| "epoch": 0.24603555982700626, |
| "grad_norm": 5.206491470336914, |
| "learning_rate": 0.0001971253321259476, |
| "logits/chosen": -0.7135074138641357, |
| "logits/rejected": -0.7155020833015442, |
| "logps/chosen": -9.462960243225098, |
| "logps/rejected": -22.44765853881836, |
| "loss": 0.8617446422576904, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4634811282157898, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.1327909231185913, |
| "rewards/margins": 1.3640327453613281, |
| "rewards/rejected": -1.2312417030334473, |
| "step": 144, |
| "train_speed(iter/s)": 0.005502 |
| }, |
| { |
| "epoch": 0.24774414010358267, |
| "grad_norm": 2.5136232376098633, |
| "learning_rate": 0.00019705763061166696, |
| "logits/chosen": -0.7309613227844238, |
| "logits/rejected": -0.7339869141578674, |
| "logps/chosen": -8.707137107849121, |
| "logps/rejected": -30.160343170166016, |
| "loss": 0.6045644879341125, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.31675195693969727, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1917949914932251, |
| "rewards/margins": 2.2541422843933105, |
| "rewards/rejected": -2.062347412109375, |
| "step": 145, |
| "train_speed(iter/s)": 0.005503 |
| }, |
| { |
| "epoch": 0.24945272038015912, |
| "grad_norm": 3.6773223876953125, |
| "learning_rate": 0.00019698915302715836, |
| "logits/chosen": -0.7318214774131775, |
| "logits/rejected": -0.7343602180480957, |
| "logps/chosen": -6.429262161254883, |
| "logps/rejected": -31.925209045410156, |
| "loss": 0.6257779598236084, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.31483885645866394, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1830574870109558, |
| "rewards/margins": 1.988149642944336, |
| "rewards/rejected": -1.8050918579101562, |
| "step": 146, |
| "train_speed(iter/s)": 0.005505 |
| }, |
| { |
| "epoch": 0.25116130065673553, |
| "grad_norm": 9.478280067443848, |
| "learning_rate": 0.00019691989991996663, |
| "logits/chosen": -0.7461144924163818, |
| "logits/rejected": -0.7457807660102844, |
| "logps/chosen": -10.189486503601074, |
| "logps/rejected": -34.494956970214844, |
| "loss": 0.7869279980659485, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.47246837615966797, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1757950335741043, |
| "rewards/margins": 2.238186836242676, |
| "rewards/rejected": -2.062391996383667, |
| "step": 147, |
| "train_speed(iter/s)": 0.005504 |
| }, |
| { |
| "epoch": 0.252869880933312, |
| "grad_norm": 7.141703128814697, |
| "learning_rate": 0.00019684987184383776, |
| "logits/chosen": -0.7700251340866089, |
| "logits/rejected": -0.7725012302398682, |
| "logps/chosen": -6.46027946472168, |
| "logps/rejected": -34.821231842041016, |
| "loss": 0.692828357219696, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3844330608844757, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.12604108452796936, |
| "rewards/margins": 2.6629996299743652, |
| "rewards/rejected": -2.5369582176208496, |
| "step": 148, |
| "train_speed(iter/s)": 0.005505 |
| }, |
| { |
| "epoch": 0.2545784612098884, |
| "grad_norm": 14.823970794677734, |
| "learning_rate": 0.00019677906935871447, |
| "logits/chosen": -0.7374511361122131, |
| "logits/rejected": -0.7389614582061768, |
| "logps/chosen": -10.582510948181152, |
| "logps/rejected": -26.03261375427246, |
| "loss": 1.33296537399292, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6253257989883423, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.18944962322711945, |
| "rewards/margins": 1.5741108655929565, |
| "rewards/rejected": -1.7635605335235596, |
| "step": 149, |
| "train_speed(iter/s)": 0.005504 |
| }, |
| { |
| "epoch": 0.25628704148646486, |
| "grad_norm": 13.657255172729492, |
| "learning_rate": 0.00019670749303073146, |
| "logits/chosen": -0.7172877192497253, |
| "logits/rejected": -0.7211941480636597, |
| "logps/chosen": -6.75715970993042, |
| "logps/rejected": -41.68476867675781, |
| "loss": 0.7531163692474365, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4120977818965912, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.10011129081249237, |
| "rewards/margins": 3.0644404888153076, |
| "rewards/rejected": -2.964329481124878, |
| "step": 150, |
| "train_speed(iter/s)": 0.005503 |
| }, |
| { |
| "epoch": 0.25799562176304125, |
| "grad_norm": 4.298210144042969, |
| "learning_rate": 0.00019663514343221117, |
| "logits/chosen": -0.6899698972702026, |
| "logits/rejected": -0.6891142725944519, |
| "logps/chosen": -9.789582252502441, |
| "logps/rejected": -32.59271240234375, |
| "loss": 0.6402732729911804, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.335668683052063, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.15986008942127228, |
| "rewards/margins": 2.269547939300537, |
| "rewards/rejected": -2.1096878051757812, |
| "step": 151, |
| "train_speed(iter/s)": 0.005498 |
| }, |
| { |
| "epoch": 0.2597042020396177, |
| "grad_norm": 4.9224324226379395, |
| "learning_rate": 0.00019656202114165904, |
| "logits/chosen": -0.717603862285614, |
| "logits/rejected": -0.7218050956726074, |
| "logps/chosen": -8.468698501586914, |
| "logps/rejected": -42.157318115234375, |
| "loss": 0.6935219168663025, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4874541759490967, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.11759095638990402, |
| "rewards/margins": 2.7493767738342285, |
| "rewards/rejected": -2.8669681549072266, |
| "step": 152, |
| "train_speed(iter/s)": 0.005497 |
| }, |
| { |
| "epoch": 0.26141278231619414, |
| "grad_norm": 5.180264472961426, |
| "learning_rate": 0.00019648812674375902, |
| "logits/chosen": -0.6897304058074951, |
| "logits/rejected": -0.6921338438987732, |
| "logps/chosen": -6.188149929046631, |
| "logps/rejected": -35.15382385253906, |
| "loss": 0.6033867597579956, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2803235948085785, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.18143904209136963, |
| "rewards/margins": 2.3970108032226562, |
| "rewards/rejected": -2.215571641921997, |
| "step": 153, |
| "train_speed(iter/s)": 0.005497 |
| }, |
| { |
| "epoch": 0.2631213625927706, |
| "grad_norm": 4.898785591125488, |
| "learning_rate": 0.00019641346082936872, |
| "logits/chosen": -0.6851431131362915, |
| "logits/rejected": -0.6863937973976135, |
| "logps/chosen": -14.464362144470215, |
| "logps/rejected": -23.472118377685547, |
| "loss": 0.783013105392456, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4414927363395691, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.01941562071442604, |
| "rewards/margins": 1.5958939790725708, |
| "rewards/rejected": -1.576478362083435, |
| "step": 154, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.264829942869347, |
| "grad_norm": 5.930595874786377, |
| "learning_rate": 0.00019633802399551473, |
| "logits/chosen": -0.7089685797691345, |
| "logits/rejected": -0.7100527882575989, |
| "logps/chosen": -9.57446575164795, |
| "logps/rejected": -26.301237106323242, |
| "loss": 0.8949770927429199, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4963584840297699, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.03665952757000923, |
| "rewards/margins": 1.5139617919921875, |
| "rewards/rejected": -1.5506211519241333, |
| "step": 155, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.2665385231459234, |
| "grad_norm": 4.985401630401611, |
| "learning_rate": 0.00019626181684538808, |
| "logits/chosen": -0.6955317854881287, |
| "logits/rejected": -0.7012707591056824, |
| "logps/chosen": -6.044739723205566, |
| "logps/rejected": -24.839508056640625, |
| "loss": 0.9819079637527466, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.47508248686790466, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.06785830110311508, |
| "rewards/margins": 0.7281482219696045, |
| "rewards/rejected": -0.660290002822876, |
| "step": 156, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.26824710342249986, |
| "grad_norm": 6.6396403312683105, |
| "learning_rate": 0.000196184839988339, |
| "logits/chosen": -0.6605405211448669, |
| "logits/rejected": -0.6651955842971802, |
| "logps/chosen": -5.8500542640686035, |
| "logps/rejected": -27.22307586669922, |
| "loss": 0.7894259095191956, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.44944238662719727, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.20848321914672852, |
| "rewards/margins": 1.5052435398101807, |
| "rewards/rejected": -1.2967603206634521, |
| "step": 157, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.2699556836990763, |
| "grad_norm": 6.331238269805908, |
| "learning_rate": 0.00019610709403987246, |
| "logits/chosen": -0.6915969848632812, |
| "logits/rejected": -0.6922619938850403, |
| "logps/chosen": -6.449377059936523, |
| "logps/rejected": -22.124189376831055, |
| "loss": 0.7714009881019592, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4102553725242615, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.15198560059070587, |
| "rewards/margins": 1.139552116394043, |
| "rewards/rejected": -0.9875664710998535, |
| "step": 158, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.27166426397565274, |
| "grad_norm": 4.062625408172607, |
| "learning_rate": 0.00019602857962164297, |
| "logits/chosen": -0.7070209383964539, |
| "logits/rejected": -0.7094908952713013, |
| "logps/chosen": -8.760738372802734, |
| "logps/rejected": -21.83628273010254, |
| "loss": 0.6808652877807617, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.30313965678215027, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.16821521520614624, |
| "rewards/margins": 1.2700836658477783, |
| "rewards/rejected": -1.1018683910369873, |
| "step": 159, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.2733728442522292, |
| "grad_norm": 3.5334150791168213, |
| "learning_rate": 0.00019594929736144976, |
| "logits/chosen": -0.6827138066291809, |
| "logits/rejected": -0.6853998303413391, |
| "logps/chosen": -8.673670768737793, |
| "logps/rejected": -22.576623916625977, |
| "loss": 0.8358533978462219, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4210669696331024, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.1305871605873108, |
| "rewards/margins": 1.2085453271865845, |
| "rewards/rejected": -1.077958106994629, |
| "step": 160, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.2750814245288056, |
| "grad_norm": 4.968493938446045, |
| "learning_rate": 0.0001958692478932316, |
| "logits/chosen": -0.6895708441734314, |
| "logits/rejected": -0.6913452744483948, |
| "logps/chosen": -8.595847129821777, |
| "logps/rejected": -15.964265823364258, |
| "loss": 0.9963687658309937, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4303150177001953, |
| "rewards/accuracies": 0.59375, |
| "rewards/chosen": 0.039102740585803986, |
| "rewards/margins": 0.7386963367462158, |
| "rewards/rejected": -0.6995935440063477, |
| "step": 161, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.276790004805382, |
| "grad_norm": 3.1100776195526123, |
| "learning_rate": 0.00019578843185706198, |
| "logits/chosen": -0.6581685543060303, |
| "logits/rejected": -0.6606261730194092, |
| "logps/chosen": -8.823217391967773, |
| "logps/rejected": -20.12518310546875, |
| "loss": 0.7960190773010254, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3306158185005188, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.1894218921661377, |
| "rewards/margins": 0.9922627210617065, |
| "rewards/rejected": -0.8028407096862793, |
| "step": 162, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.27849858508195846, |
| "grad_norm": 2.9061076641082764, |
| "learning_rate": 0.00019570684989914368, |
| "logits/chosen": -0.7005666494369507, |
| "logits/rejected": -0.7033515572547913, |
| "logps/chosen": -6.614690780639648, |
| "logps/rejected": -23.320758819580078, |
| "loss": 0.7204984426498413, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3370620608329773, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.20666998624801636, |
| "rewards/margins": 1.309868335723877, |
| "rewards/rejected": -1.1031984090805054, |
| "step": 163, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.2802071653585349, |
| "grad_norm": 2.7546348571777344, |
| "learning_rate": 0.00019562450267180396, |
| "logits/chosen": -0.713110089302063, |
| "logits/rejected": -0.7162569165229797, |
| "logps/chosen": -11.556283950805664, |
| "logps/rejected": -24.4801025390625, |
| "loss": 0.600546658039093, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.29867106676101685, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.32357320189476013, |
| "rewards/margins": 1.2928061485290527, |
| "rewards/rejected": -0.969232976436615, |
| "step": 164, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.28191574563511135, |
| "grad_norm": 4.433908462524414, |
| "learning_rate": 0.000195541390833489, |
| "logits/chosen": -0.7190208435058594, |
| "logits/rejected": -0.7217738628387451, |
| "logps/chosen": -8.067594528198242, |
| "logps/rejected": -28.02341651916504, |
| "loss": 0.6388089656829834, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3460094630718231, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.26204991340637207, |
| "rewards/margins": 1.4149954319000244, |
| "rewards/rejected": -1.152945637702942, |
| "step": 165, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.28362432591168774, |
| "grad_norm": 2.649824380874634, |
| "learning_rate": 0.0001954575150487589, |
| "logits/chosen": -0.749485969543457, |
| "logits/rejected": -0.7498363852500916, |
| "logps/chosen": -8.293320655822754, |
| "logps/rejected": -18.8927001953125, |
| "loss": 0.6769586205482483, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2774519920349121, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.14119945466518402, |
| "rewards/margins": 1.063216209411621, |
| "rewards/rejected": -0.9220167398452759, |
| "step": 166, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.2853329061882642, |
| "grad_norm": 5.596877574920654, |
| "learning_rate": 0.0001953728759882821, |
| "logits/chosen": -0.7324296832084656, |
| "logits/rejected": -0.7348483204841614, |
| "logps/chosen": -10.596734046936035, |
| "logps/rejected": -26.180767059326172, |
| "loss": 0.6137089729309082, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.338160902261734, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.13076083362102509, |
| "rewards/margins": 1.6062695980072021, |
| "rewards/rejected": -1.4755089282989502, |
| "step": 167, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.2870414864648406, |
| "grad_norm": 3.771359443664551, |
| "learning_rate": 0.00019528747432883035, |
| "logits/chosen": -0.790438711643219, |
| "logits/rejected": -0.794373631477356, |
| "logps/chosen": -10.444026947021484, |
| "logps/rejected": -26.96920394897461, |
| "loss": 0.9520989060401917, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5127608776092529, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.06649212539196014, |
| "rewards/margins": 1.7065045833587646, |
| "rewards/rejected": -1.6400126218795776, |
| "step": 168, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.28875006674141707, |
| "grad_norm": 5.279207706451416, |
| "learning_rate": 0.00019520131075327298, |
| "logits/chosen": -0.811764657497406, |
| "logits/rejected": -0.8173067569732666, |
| "logps/chosen": -5.237765312194824, |
| "logps/rejected": -34.11195755004883, |
| "loss": 0.5845170617103577, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.320607453584671, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.16728658974170685, |
| "rewards/margins": 2.271721363067627, |
| "rewards/rejected": -2.1044344902038574, |
| "step": 169, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.2904586470179935, |
| "grad_norm": 3.968066453933716, |
| "learning_rate": 0.00019511438595057174, |
| "logits/chosen": -0.8485198020935059, |
| "logits/rejected": -0.8553695678710938, |
| "logps/chosen": -2.880854606628418, |
| "logps/rejected": -36.118629455566406, |
| "loss": 0.4797823131084442, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.23597285151481628, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.19838106632232666, |
| "rewards/margins": 2.7422218322753906, |
| "rewards/rejected": -2.5438406467437744, |
| "step": 170, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.2921672272945699, |
| "grad_norm": 5.908665180206299, |
| "learning_rate": 0.00019502670061577498, |
| "logits/chosen": -0.8754770159721375, |
| "logits/rejected": -0.8793540000915527, |
| "logps/chosen": -6.855566501617432, |
| "logps/rejected": -41.512691497802734, |
| "loss": 0.5422110557556152, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.29031190276145935, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1825065314769745, |
| "rewards/margins": 3.2673420906066895, |
| "rewards/rejected": -3.0848352909088135, |
| "step": 171, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.29387580757114634, |
| "grad_norm": 9.865046501159668, |
| "learning_rate": 0.00019493825545001227, |
| "logits/chosen": -0.8389215469360352, |
| "logits/rejected": -0.8400964736938477, |
| "logps/chosen": -5.75340461730957, |
| "logps/rejected": -28.154647827148438, |
| "loss": 0.5484858751296997, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.24028873443603516, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.0924752876162529, |
| "rewards/margins": 2.016855001449585, |
| "rewards/rejected": -1.9243797063827515, |
| "step": 172, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.2955843878477228, |
| "grad_norm": 5.110999584197998, |
| "learning_rate": 0.00019484905116048883, |
| "logits/chosen": -0.8424903750419617, |
| "logits/rejected": -0.8479667901992798, |
| "logps/chosen": -10.439705848693848, |
| "logps/rejected": -34.95732116699219, |
| "loss": 0.9751264452934265, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6046102046966553, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.047889530658721924, |
| "rewards/margins": 2.1707029342651367, |
| "rewards/rejected": -2.1228134632110596, |
| "step": 173, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.29729296812429923, |
| "grad_norm": 5.233139514923096, |
| "learning_rate": 0.00019475908846047982, |
| "logits/chosen": -0.8389162421226501, |
| "logits/rejected": -0.8423767685890198, |
| "logps/chosen": -6.015560626983643, |
| "logps/rejected": -33.967063903808594, |
| "loss": 0.5880640745162964, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.36994248628616333, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.29298272728919983, |
| "rewards/margins": 2.3524224758148193, |
| "rewards/rejected": -2.0594396591186523, |
| "step": 174, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.29900154840087567, |
| "grad_norm": 3.566279649734497, |
| "learning_rate": 0.00019466836806932453, |
| "logits/chosen": -0.8825660943984985, |
| "logits/rejected": -0.8887041807174683, |
| "logps/chosen": -6.019082546234131, |
| "logps/rejected": -35.091529846191406, |
| "loss": 0.6474583148956299, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.41612136363983154, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2304912507534027, |
| "rewards/margins": 2.494393825531006, |
| "rewards/rejected": -2.2639026641845703, |
| "step": 175, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.30071012867745206, |
| "grad_norm": 6.7856950759887695, |
| "learning_rate": 0.00019457689071242085, |
| "logits/chosen": -0.8579093813896179, |
| "logits/rejected": -0.8634753227233887, |
| "logps/chosen": -9.888270378112793, |
| "logps/rejected": -41.228416442871094, |
| "loss": 0.6879767179489136, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.423296183347702, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.31314119696617126, |
| "rewards/margins": 2.737194538116455, |
| "rewards/rejected": -2.424053192138672, |
| "step": 176, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3024187089540285, |
| "grad_norm": 4.894569396972656, |
| "learning_rate": 0.0001944846571212192, |
| "logits/chosen": -0.8246264457702637, |
| "logits/rejected": -0.8278872966766357, |
| "logps/chosen": -9.956787109375, |
| "logps/rejected": -38.914066314697266, |
| "loss": 0.7229227423667908, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3963947594165802, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.03048780933022499, |
| "rewards/margins": 2.575648307800293, |
| "rewards/rejected": -2.5451607704162598, |
| "step": 177, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.30412728923060495, |
| "grad_norm": 5.153270721435547, |
| "learning_rate": 0.000194391668033217, |
| "logits/chosen": -0.8741589784622192, |
| "logits/rejected": -0.879849374294281, |
| "logps/chosen": -5.598626613616943, |
| "logps/rejected": -33.6767463684082, |
| "loss": 0.6376072764396667, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.31503573060035706, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.20653334259986877, |
| "rewards/margins": 2.248800754547119, |
| "rewards/rejected": -2.0422675609588623, |
| "step": 178, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.3058358695071814, |
| "grad_norm": 6.50536584854126, |
| "learning_rate": 0.00019429792419195247, |
| "logits/chosen": -0.83283931016922, |
| "logits/rejected": -0.8357385993003845, |
| "logps/chosen": -5.801810264587402, |
| "logps/rejected": -26.60141372680664, |
| "loss": 0.7859883308410645, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.40051573514938354, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.022562842816114426, |
| "rewards/margins": 1.451593041419983, |
| "rewards/rejected": -1.429030179977417, |
| "step": 179, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.30754444978375783, |
| "grad_norm": 2.6014249324798584, |
| "learning_rate": 0.0001942034263469989, |
| "logits/chosen": -0.8176060914993286, |
| "logits/rejected": -0.8231288194656372, |
| "logps/chosen": -8.629841804504395, |
| "logps/rejected": -33.58430480957031, |
| "loss": 0.6443658471107483, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.34101858735084534, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.10939686745405197, |
| "rewards/margins": 2.1205523014068604, |
| "rewards/rejected": -2.011155605316162, |
| "step": 180, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.3092530300603342, |
| "grad_norm": 1.8803960084915161, |
| "learning_rate": 0.00019410817525395855, |
| "logits/chosen": -0.8076679706573486, |
| "logits/rejected": -0.8171229958534241, |
| "logps/chosen": -3.606616735458374, |
| "logps/rejected": -41.894737243652344, |
| "loss": 0.37792402505874634, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.1496608406305313, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.18168124556541443, |
| "rewards/margins": 2.5405263900756836, |
| "rewards/rejected": -2.3588449954986572, |
| "step": 181, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.31096161033691067, |
| "grad_norm": 1.7322460412979126, |
| "learning_rate": 0.0001940121716744566, |
| "logits/chosen": -0.8273515701293945, |
| "logits/rejected": -0.8358028531074524, |
| "logps/chosen": -4.195955753326416, |
| "logps/rejected": -42.863243103027344, |
| "loss": 0.40653136372566223, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.1898367702960968, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.2309657633304596, |
| "rewards/margins": 2.9126315116882324, |
| "rewards/rejected": -2.6816658973693848, |
| "step": 182, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.3126701906134871, |
| "grad_norm": 7.253870010375977, |
| "learning_rate": 0.00019391541637613508, |
| "logits/chosen": -0.818758487701416, |
| "logits/rejected": -0.8240275979042053, |
| "logps/chosen": -6.234726428985596, |
| "logps/rejected": -38.455230712890625, |
| "loss": 0.6432409882545471, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3129611313343048, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.03605951368808746, |
| "rewards/margins": 2.6540074348449707, |
| "rewards/rejected": -2.617948055267334, |
| "step": 183, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.31437877089006355, |
| "grad_norm": 3.745917320251465, |
| "learning_rate": 0.0001938179101326468, |
| "logits/chosen": -0.8772048354148865, |
| "logits/rejected": -0.8792482614517212, |
| "logps/chosen": -10.402073860168457, |
| "logps/rejected": -31.16716766357422, |
| "loss": 0.7786051630973816, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4461228549480438, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.22984227538108826, |
| "rewards/margins": 2.267395496368408, |
| "rewards/rejected": -2.037553310394287, |
| "step": 184, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.31608735116664, |
| "grad_norm": 3.6105992794036865, |
| "learning_rate": 0.00019371965372364915, |
| "logits/chosen": -0.8841488361358643, |
| "logits/rejected": -0.8877873420715332, |
| "logps/chosen": -7.848278045654297, |
| "logps/rejected": -28.794742584228516, |
| "loss": 0.7686839699745178, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4148586094379425, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.2514728009700775, |
| "rewards/margins": 2.0990750789642334, |
| "rewards/rejected": -1.8476024866104126, |
| "step": 185, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.3177959314432164, |
| "grad_norm": 4.601129531860352, |
| "learning_rate": 0.00019362064793479767, |
| "logits/chosen": -0.890434980392456, |
| "logits/rejected": -0.8937606811523438, |
| "logps/chosen": -6.078680515289307, |
| "logps/rejected": -33.81269454956055, |
| "loss": 0.6938453316688538, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.35300731658935547, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.06243661791086197, |
| "rewards/margins": 2.5140228271484375, |
| "rewards/rejected": -2.4515862464904785, |
| "step": 186, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.3195045117197928, |
| "grad_norm": 3.495896339416504, |
| "learning_rate": 0.00019352089355774004, |
| "logits/chosen": -0.8553616404533386, |
| "logits/rejected": -0.8608652949333191, |
| "logps/chosen": -5.048914432525635, |
| "logps/rejected": -32.950260162353516, |
| "loss": 0.5130528211593628, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.27008938789367676, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2167266607284546, |
| "rewards/margins": 2.171088457107544, |
| "rewards/rejected": -1.9543616771697998, |
| "step": 187, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.32121309199636927, |
| "grad_norm": 2.0062403678894043, |
| "learning_rate": 0.0001934203913901096, |
| "logits/chosen": -0.8811948299407959, |
| "logits/rejected": -0.8872733116149902, |
| "logps/chosen": -6.987203598022461, |
| "logps/rejected": -31.409507751464844, |
| "loss": 0.6117798686027527, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3332153260707855, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.2551400661468506, |
| "rewards/margins": 2.197721004486084, |
| "rewards/rejected": -1.942581057548523, |
| "step": 188, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3229216722729457, |
| "grad_norm": 2.5417840480804443, |
| "learning_rate": 0.000193319142235519, |
| "logits/chosen": -0.8587099313735962, |
| "logits/rejected": -0.860413670539856, |
| "logps/chosen": -10.022211074829102, |
| "logps/rejected": -29.925012588500977, |
| "loss": 0.6710554361343384, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4114764630794525, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2854654788970947, |
| "rewards/margins": 2.179234743118286, |
| "rewards/rejected": -1.893769383430481, |
| "step": 189, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.32463025254952216, |
| "grad_norm": 2.4238228797912598, |
| "learning_rate": 0.0001932171469035538, |
| "logits/chosen": -0.8661004900932312, |
| "logits/rejected": -0.8710408806800842, |
| "logps/chosen": -5.72200870513916, |
| "logps/rejected": -31.211780548095703, |
| "loss": 0.4626935124397278, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.24239738285541534, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.23706884682178497, |
| "rewards/margins": 2.2727468013763428, |
| "rewards/rejected": -2.0356781482696533, |
| "step": 190, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.32633883282609855, |
| "grad_norm": 4.148837089538574, |
| "learning_rate": 0.00019311440620976597, |
| "logits/chosen": -0.9142186045646667, |
| "logits/rejected": -0.9197511076927185, |
| "logps/chosen": -6.843296051025391, |
| "logps/rejected": -34.4539680480957, |
| "loss": 0.5014334321022034, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3269428014755249, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.398705393075943, |
| "rewards/margins": 2.5132031440734863, |
| "rewards/rejected": -2.1144979000091553, |
| "step": 191, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.328047413102675, |
| "grad_norm": 6.930333614349365, |
| "learning_rate": 0.0001930109209756673, |
| "logits/chosen": -0.9393826723098755, |
| "logits/rejected": -0.9431489109992981, |
| "logps/chosen": -12.365182876586914, |
| "logps/rejected": -36.20299530029297, |
| "loss": 0.9010649919509888, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5080785155296326, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.09569638967514038, |
| "rewards/margins": 2.128075361251831, |
| "rewards/rejected": -2.223771810531616, |
| "step": 192, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.32975599337925143, |
| "grad_norm": 1.8063973188400269, |
| "learning_rate": 0.00019290669202872302, |
| "logits/chosen": -0.9646225571632385, |
| "logits/rejected": -0.9725656509399414, |
| "logps/chosen": -4.754866123199463, |
| "logps/rejected": -49.72484588623047, |
| "loss": 0.3587495982646942, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2268112450838089, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.3622514307498932, |
| "rewards/margins": 4.094050884246826, |
| "rewards/rejected": -3.731799602508545, |
| "step": 193, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.3314645736558279, |
| "grad_norm": 9.937451362609863, |
| "learning_rate": 0.00019280172020234496, |
| "logits/chosen": -1.0251308679580688, |
| "logits/rejected": -1.030665636062622, |
| "logps/chosen": -10.551065444946289, |
| "logps/rejected": -53.82728576660156, |
| "loss": 1.1305742263793945, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5491915345191956, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.27872776985168457, |
| "rewards/margins": 3.9652843475341797, |
| "rewards/rejected": -4.244011878967285, |
| "step": 194, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3331731539324043, |
| "grad_norm": 7.0363922119140625, |
| "learning_rate": 0.00019269600633588514, |
| "logits/chosen": -0.98694908618927, |
| "logits/rejected": -0.9900444746017456, |
| "logps/chosen": -11.45719051361084, |
| "logps/rejected": -51.027496337890625, |
| "loss": 0.6249094009399414, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3946501910686493, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.056097134947776794, |
| "rewards/margins": 3.8215420246124268, |
| "rewards/rejected": -3.8776392936706543, |
| "step": 195, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3348817342089807, |
| "grad_norm": 7.936800003051758, |
| "learning_rate": 0.0001925895512746287, |
| "logits/chosen": -0.9765729904174805, |
| "logits/rejected": -0.9803340435028076, |
| "logps/chosen": -7.451618194580078, |
| "logps/rejected": -42.50349426269531, |
| "loss": 0.8021814227104187, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.49508342146873474, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.0933699905872345, |
| "rewards/margins": 3.030968189239502, |
| "rewards/rejected": -3.124338150024414, |
| "step": 196, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.33659031448555715, |
| "grad_norm": 7.247754096984863, |
| "learning_rate": 0.00019248235586978755, |
| "logits/chosen": -0.9704450964927673, |
| "logits/rejected": -0.9796913862228394, |
| "logps/chosen": -5.781133651733398, |
| "logps/rejected": -43.08665084838867, |
| "loss": 0.5009433627128601, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3090832829475403, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.17386135458946228, |
| "rewards/margins": 3.0037147998809814, |
| "rewards/rejected": -2.8298535346984863, |
| "step": 197, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3382988947621336, |
| "grad_norm": 4.252370357513428, |
| "learning_rate": 0.0001923744209784933, |
| "logits/chosen": -0.9834634065628052, |
| "logits/rejected": -0.9903834462165833, |
| "logps/chosen": -10.431640625, |
| "logps/rejected": -41.34126663208008, |
| "loss": 0.892498791217804, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5820133686065674, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.0897313803434372, |
| "rewards/margins": 2.829003095626831, |
| "rewards/rejected": -2.739271640777588, |
| "step": 198, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.34000747503871004, |
| "grad_norm": 7.74161958694458, |
| "learning_rate": 0.00019226574746379038, |
| "logits/chosen": -0.9264886975288391, |
| "logits/rejected": -0.9303056597709656, |
| "logps/chosen": -18.308433532714844, |
| "logps/rejected": -34.25849914550781, |
| "loss": 1.4104645252227783, |
| "memory(GiB)": 46.38, |
| "nll_loss": 1.1379212141036987, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.23661470413208008, |
| "rewards/margins": 2.288109540939331, |
| "rewards/rejected": -2.051494598388672, |
| "step": 199, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.3417160553152864, |
| "grad_norm": 19.06857681274414, |
| "learning_rate": 0.00019215633619462942, |
| "logits/chosen": -0.9472784399986267, |
| "logits/rejected": -0.9554142951965332, |
| "logps/chosen": -9.5065336227417, |
| "logps/rejected": -40.612945556640625, |
| "loss": 1.1060166358947754, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.7597510814666748, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.030025824904441833, |
| "rewards/margins": 2.3827638626098633, |
| "rewards/rejected": -2.35273814201355, |
| "step": 200, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.34342463559186287, |
| "grad_norm": 3.042719841003418, |
| "learning_rate": 0.00019204618804585994, |
| "logits/chosen": -0.8972377181053162, |
| "logits/rejected": -0.9024269580841064, |
| "logps/chosen": -6.98057222366333, |
| "logps/rejected": -34.55181884765625, |
| "loss": 0.5098893642425537, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2953639328479767, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.31456321477890015, |
| "rewards/margins": 2.592310905456543, |
| "rewards/rejected": -2.277747392654419, |
| "step": 201, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.3451332158684393, |
| "grad_norm": 3.191251516342163, |
| "learning_rate": 0.00019193530389822363, |
| "logits/chosen": -0.9289453625679016, |
| "logits/rejected": -0.9359845519065857, |
| "logps/chosen": -9.659342765808105, |
| "logps/rejected": -37.61637878417969, |
| "loss": 1.0019766092300415, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6679946780204773, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1197037547826767, |
| "rewards/margins": 2.4200377464294434, |
| "rewards/rejected": -2.3003342151641846, |
| "step": 202, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.34684179614501576, |
| "grad_norm": 4.930737018585205, |
| "learning_rate": 0.00019182368463834722, |
| "logits/chosen": -0.8630688190460205, |
| "logits/rejected": -0.8667917847633362, |
| "logps/chosen": -9.344999313354492, |
| "logps/rejected": -36.20460510253906, |
| "loss": 0.5554157495498657, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.36027616262435913, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1459665298461914, |
| "rewards/margins": 2.5479657649993896, |
| "rewards/rejected": -2.4019992351531982, |
| "step": 203, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3485503764215922, |
| "grad_norm": 2.625263214111328, |
| "learning_rate": 0.00019171133115873526, |
| "logits/chosen": -0.8702833652496338, |
| "logits/rejected": -0.8725969195365906, |
| "logps/chosen": -7.262643814086914, |
| "logps/rejected": -34.7918701171875, |
| "loss": 0.4979276657104492, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3470659554004669, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.23693254590034485, |
| "rewards/margins": 2.7716078758239746, |
| "rewards/rejected": -2.534675359725952, |
| "step": 204, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3502589566981686, |
| "grad_norm": 4.522634029388428, |
| "learning_rate": 0.0001915982443577632, |
| "logits/chosen": -0.8535888195037842, |
| "logits/rejected": -0.859410285949707, |
| "logps/chosen": -6.171905517578125, |
| "logps/rejected": -37.568458557128906, |
| "loss": 0.5628905296325684, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3058054447174072, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.09114877134561539, |
| "rewards/margins": 2.7372331619262695, |
| "rewards/rejected": -2.6460840702056885, |
| "step": 205, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.35196753697474503, |
| "grad_norm": 3.1796517372131348, |
| "learning_rate": 0.00019148442513967014, |
| "logits/chosen": -0.8468549251556396, |
| "logits/rejected": -0.8539971709251404, |
| "logps/chosen": -9.5682954788208, |
| "logps/rejected": -36.71331024169922, |
| "loss": 0.6788620948791504, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4944525957107544, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.10999859869480133, |
| "rewards/margins": 2.5533294677734375, |
| "rewards/rejected": -2.443330764770508, |
| "step": 206, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.3536761172513215, |
| "grad_norm": 3.1952710151672363, |
| "learning_rate": 0.00019136987441455144, |
| "logits/chosen": -0.8667152523994446, |
| "logits/rejected": -0.8705443143844604, |
| "logps/chosen": -9.689000129699707, |
| "logps/rejected": -30.918254852294922, |
| "loss": 0.6780470013618469, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.41608238220214844, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.2547697126865387, |
| "rewards/margins": 2.3299858570098877, |
| "rewards/rejected": -2.075216293334961, |
| "step": 207, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.3553846975278979, |
| "grad_norm": 2.4894015789031982, |
| "learning_rate": 0.00019125459309835173, |
| "logits/chosen": -0.8714242577552795, |
| "logits/rejected": -0.8752931356430054, |
| "logps/chosen": -7.453229904174805, |
| "logps/rejected": -31.325904846191406, |
| "loss": 0.6907439827919006, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4350113868713379, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.17801815271377563, |
| "rewards/margins": 2.264295816421509, |
| "rewards/rejected": -2.086277484893799, |
| "step": 208, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.35709327780447436, |
| "grad_norm": 3.2602062225341797, |
| "learning_rate": 0.00019113858211285725, |
| "logits/chosen": -0.872046709060669, |
| "logits/rejected": -0.8767064213752747, |
| "logps/chosen": -4.748498439788818, |
| "logps/rejected": -36.4671745300293, |
| "loss": 0.47542524337768555, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.24451857805252075, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.20409636199474335, |
| "rewards/margins": 2.852433443069458, |
| "rewards/rejected": -2.6483371257781982, |
| "step": 209, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.35880185808105075, |
| "grad_norm": 2.723602533340454, |
| "learning_rate": 0.00019102184238568878, |
| "logits/chosen": -0.872526228427887, |
| "logits/rejected": -0.8720419406890869, |
| "logps/chosen": -9.495769500732422, |
| "logps/rejected": -35.4581184387207, |
| "loss": 0.6998460292816162, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4141753911972046, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.08976475894451141, |
| "rewards/margins": 2.3428900241851807, |
| "rewards/rejected": -2.2531251907348633, |
| "step": 210, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3605104383576272, |
| "grad_norm": 7.090069770812988, |
| "learning_rate": 0.00019090437485029413, |
| "logits/chosen": -0.8751546740531921, |
| "logits/rejected": -0.8804129362106323, |
| "logps/chosen": -5.539629936218262, |
| "logps/rejected": -41.93748092651367, |
| "loss": 0.414511114358902, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.24404364824295044, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2657608985900879, |
| "rewards/margins": 3.283001184463501, |
| "rewards/rejected": -3.017240524291992, |
| "step": 211, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.36221901863420364, |
| "grad_norm": 3.3576340675354004, |
| "learning_rate": 0.00019078618044594052, |
| "logits/chosen": -0.9006087779998779, |
| "logits/rejected": -0.9030022621154785, |
| "logps/chosen": -7.570390224456787, |
| "logps/rejected": -34.19873046875, |
| "loss": 0.4811999499797821, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.29778629541397095, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.24354203045368195, |
| "rewards/margins": 2.750795364379883, |
| "rewards/rejected": -2.5072531700134277, |
| "step": 212, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3639275989107801, |
| "grad_norm": 4.36439323425293, |
| "learning_rate": 0.00019066726011770726, |
| "logits/chosen": -0.8826490640640259, |
| "logits/rejected": -0.8854631185531616, |
| "logps/chosen": -7.964271068572998, |
| "logps/rejected": -36.53171920776367, |
| "loss": 0.7800939679145813, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4539424479007721, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1163502037525177, |
| "rewards/margins": 2.651106357574463, |
| "rewards/rejected": -2.5347559452056885, |
| "step": 213, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.3656361791873565, |
| "grad_norm": 3.075863838195801, |
| "learning_rate": 0.00019054761481647817, |
| "logits/chosen": -0.9236173033714294, |
| "logits/rejected": -0.9311164021492004, |
| "logps/chosen": -5.548548698425293, |
| "logps/rejected": -39.62729263305664, |
| "loss": 0.5424690842628479, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.37887126207351685, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.2379705011844635, |
| "rewards/margins": 2.8928284645080566, |
| "rewards/rejected": -2.654857873916626, |
| "step": 214, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.3673447594639329, |
| "grad_norm": 4.081761360168457, |
| "learning_rate": 0.0001904272454989338, |
| "logits/chosen": -0.9173122644424438, |
| "logits/rejected": -0.9237719774246216, |
| "logps/chosen": -7.586677551269531, |
| "logps/rejected": -39.44381332397461, |
| "loss": 0.7125780582427979, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4088558554649353, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.10065207630395889, |
| "rewards/margins": 2.89681077003479, |
| "rewards/rejected": -2.796158790588379, |
| "step": 215, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.36905333974050936, |
| "grad_norm": 6.38205099105835, |
| "learning_rate": 0.00019030615312754412, |
| "logits/chosen": -0.8755264282226562, |
| "logits/rejected": -0.8772911429405212, |
| "logps/chosen": -10.058317184448242, |
| "logps/rejected": -29.399351119995117, |
| "loss": 0.6085036993026733, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.37356507778167725, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.257190465927124, |
| "rewards/margins": 2.3912007808685303, |
| "rewards/rejected": -2.1340105533599854, |
| "step": 216, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3707619200170858, |
| "grad_norm": 3.965923547744751, |
| "learning_rate": 0.00019018433867056043, |
| "logits/chosen": -0.8582707047462463, |
| "logits/rejected": -0.8643903136253357, |
| "logps/chosen": -7.274492263793945, |
| "logps/rejected": -38.09757995605469, |
| "loss": 0.7568970918655396, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5492669343948364, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.09097186475992203, |
| "rewards/margins": 2.650477170944214, |
| "rewards/rejected": -2.5595054626464844, |
| "step": 217, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.37247050029366224, |
| "grad_norm": 6.908701419830322, |
| "learning_rate": 0.00019006180310200788, |
| "logits/chosen": -0.8987607955932617, |
| "logits/rejected": -0.9025367498397827, |
| "logps/chosen": -9.353863716125488, |
| "logps/rejected": -41.96833801269531, |
| "loss": 0.5579907298088074, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.32021838426589966, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.06166917830705643, |
| "rewards/margins": 3.1308600902557373, |
| "rewards/rejected": -3.069190740585327, |
| "step": 218, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3741790805702387, |
| "grad_norm": 11.417703628540039, |
| "learning_rate": 0.00018993854740167768, |
| "logits/chosen": -0.8815793991088867, |
| "logits/rejected": -0.8841702342033386, |
| "logps/chosen": -8.630674362182617, |
| "logps/rejected": -32.739219665527344, |
| "loss": 0.7540084719657898, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.42135101556777954, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.017735840752720833, |
| "rewards/margins": 2.403879165649414, |
| "rewards/rejected": -2.386143445968628, |
| "step": 219, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.3758876608468151, |
| "grad_norm": 7.359007358551025, |
| "learning_rate": 0.00018981457255511902, |
| "logits/chosen": -0.8066080212593079, |
| "logits/rejected": -0.8113219141960144, |
| "logps/chosen": -10.087733268737793, |
| "logps/rejected": -34.230953216552734, |
| "loss": 0.7611066102981567, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.38534992933273315, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.18794448673725128, |
| "rewards/margins": 2.276632785797119, |
| "rewards/rejected": -2.088688373565674, |
| "step": 220, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.3775962411233915, |
| "grad_norm": 2.5772690773010254, |
| "learning_rate": 0.00018968987955363153, |
| "logits/chosen": -0.7865528464317322, |
| "logits/rejected": -0.7891740202903748, |
| "logps/chosen": -13.03406810760498, |
| "logps/rejected": -25.763931274414062, |
| "loss": 0.8478187918663025, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5038392543792725, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.3180733621120453, |
| "rewards/margins": 1.6984466314315796, |
| "rewards/rejected": -1.3803733587265015, |
| "step": 221, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.37930482139996796, |
| "grad_norm": 3.7778072357177734, |
| "learning_rate": 0.00018956446939425716, |
| "logits/chosen": -0.826096773147583, |
| "logits/rejected": -0.8322123885154724, |
| "logps/chosen": -9.263298988342285, |
| "logps/rejected": -25.442668914794922, |
| "loss": 0.6595274209976196, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3389233946800232, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.24548465013504028, |
| "rewards/margins": 1.5120353698730469, |
| "rewards/rejected": -1.2665507793426514, |
| "step": 222, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.3810134016765444, |
| "grad_norm": 4.857659339904785, |
| "learning_rate": 0.00018943834307977216, |
| "logits/chosen": -0.8075899481773376, |
| "logits/rejected": -0.8120123744010925, |
| "logps/chosen": -4.7845778465271, |
| "logps/rejected": -21.904151916503906, |
| "loss": 0.6341399550437927, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2501222789287567, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.19457918405532837, |
| "rewards/margins": 1.1304917335510254, |
| "rewards/rejected": -0.9359126091003418, |
| "step": 223, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.38272198195312085, |
| "grad_norm": 2.3448665142059326, |
| "learning_rate": 0.00018931150161867916, |
| "logits/chosen": -0.882806658744812, |
| "logits/rejected": -0.8879884481430054, |
| "logps/chosen": -9.03802490234375, |
| "logps/rejected": -24.53693199157715, |
| "loss": 0.7897974848747253, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.38777855038642883, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.1946476548910141, |
| "rewards/margins": 1.5893993377685547, |
| "rewards/rejected": -1.3947515487670898, |
| "step": 224, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.38443056222969724, |
| "grad_norm": 11.419641494750977, |
| "learning_rate": 0.00018918394602519914, |
| "logits/chosen": -0.8994475603103638, |
| "logits/rejected": -0.9015949964523315, |
| "logps/chosen": -7.98261022567749, |
| "logps/rejected": -27.965211868286133, |
| "loss": 0.7503970265388489, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4178537130355835, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.22919557988643646, |
| "rewards/margins": 1.8219597339630127, |
| "rewards/rejected": -1.592764139175415, |
| "step": 225, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.3861391425062737, |
| "grad_norm": 5.716240406036377, |
| "learning_rate": 0.00018905567731926324, |
| "logits/chosen": -0.8648837208747864, |
| "logits/rejected": -0.8741611838340759, |
| "logps/chosen": -6.251923084259033, |
| "logps/rejected": -36.683223724365234, |
| "loss": 0.5255903601646423, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3399491310119629, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2363051176071167, |
| "rewards/margins": 2.3525753021240234, |
| "rewards/rejected": -2.1162703037261963, |
| "step": 226, |
| "train_speed(iter/s)": 0.005497 |
| }, |
| { |
| "epoch": 0.3878477227828501, |
| "grad_norm": 2.4492273330688477, |
| "learning_rate": 0.00018892669652650456, |
| "logits/chosen": -0.8707509636878967, |
| "logits/rejected": -0.8740925192832947, |
| "logps/chosen": -8.421378135681152, |
| "logps/rejected": -34.801605224609375, |
| "loss": 0.6341971158981323, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4148554801940918, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3119761645793915, |
| "rewards/margins": 2.5685765743255615, |
| "rewards/rejected": -2.2566003799438477, |
| "step": 227, |
| "train_speed(iter/s)": 0.005496 |
| }, |
| { |
| "epoch": 0.38955630305942657, |
| "grad_norm": 7.605607986450195, |
| "learning_rate": 0.00018879700467825013, |
| "logits/chosen": -0.8900420069694519, |
| "logits/rejected": -0.8933430910110474, |
| "logps/chosen": -5.360377311706543, |
| "logps/rejected": -33.13306427001953, |
| "loss": 0.4900038540363312, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.28295254707336426, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3359947204589844, |
| "rewards/margins": 2.5908119678497314, |
| "rewards/rejected": -2.254817008972168, |
| "step": 228, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.391264883336003, |
| "grad_norm": 8.061660766601562, |
| "learning_rate": 0.00018866660281151247, |
| "logits/chosen": -0.9369041919708252, |
| "logits/rejected": -0.9415433406829834, |
| "logps/chosen": -4.601605415344238, |
| "logps/rejected": -35.486732482910156, |
| "loss": 0.5184635519981384, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2989267408847809, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.22662585973739624, |
| "rewards/margins": 2.7763917446136475, |
| "rewards/rejected": -2.5497660636901855, |
| "step": 229, |
| "train_speed(iter/s)": 0.005495 |
| }, |
| { |
| "epoch": 0.3929734636125794, |
| "grad_norm": 3.7317209243774414, |
| "learning_rate": 0.0001885354919689815, |
| "logits/chosen": -0.9507468342781067, |
| "logits/rejected": -0.9519726634025574, |
| "logps/chosen": -6.97619104385376, |
| "logps/rejected": -38.649261474609375, |
| "loss": 0.4864623546600342, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2420811951160431, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.16122400760650635, |
| "rewards/margins": 2.8739140033721924, |
| "rewards/rejected": -2.7126896381378174, |
| "step": 230, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.39468204388915584, |
| "grad_norm": 2.446930408477783, |
| "learning_rate": 0.00018840367319901592, |
| "logits/chosen": -0.9463286399841309, |
| "logits/rejected": -0.9530838131904602, |
| "logps/chosen": -4.321158409118652, |
| "logps/rejected": -45.260074615478516, |
| "loss": 0.2898697853088379, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.19011972844600677, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.2383752316236496, |
| "rewards/margins": 3.6623451709747314, |
| "rewards/rejected": -3.4239697456359863, |
| "step": 231, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.3963906241657323, |
| "grad_norm": 10.94664478302002, |
| "learning_rate": 0.0001882711475556352, |
| "logits/chosen": -0.9894860982894897, |
| "logits/rejected": -0.9986059665679932, |
| "logps/chosen": -7.955844879150391, |
| "logps/rejected": -57.02360534667969, |
| "loss": 1.0545183420181274, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.7007216215133667, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.0017464309930801392, |
| "rewards/margins": 4.065786361694336, |
| "rewards/rejected": -4.067533016204834, |
| "step": 232, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.39809920444230873, |
| "grad_norm": 5.375171661376953, |
| "learning_rate": 0.0001881379160985108, |
| "logits/chosen": -1.0208549499511719, |
| "logits/rejected": -1.0248205661773682, |
| "logps/chosen": -7.056022644042969, |
| "logps/rejected": -52.94482421875, |
| "loss": 0.6524523496627808, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.44853687286376953, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.08947555720806122, |
| "rewards/margins": 4.471720218658447, |
| "rewards/rejected": -4.382244110107422, |
| "step": 233, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.3998077847188852, |
| "grad_norm": 12.273642539978027, |
| "learning_rate": 0.00018800397989295794, |
| "logits/chosen": -0.9972729682922363, |
| "logits/rejected": -1.0022565126419067, |
| "logps/chosen": -10.15302848815918, |
| "logps/rejected": -46.114967346191406, |
| "loss": 0.7069646716117859, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4566623270511627, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1436801254749298, |
| "rewards/margins": 3.605038642883301, |
| "rewards/rejected": -3.4613585472106934, |
| "step": 234, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.40151636499546156, |
| "grad_norm": 8.246232986450195, |
| "learning_rate": 0.00018786934000992688, |
| "logits/chosen": -1.0225094556808472, |
| "logits/rejected": -1.019821286201477, |
| "logps/chosen": -11.058813095092773, |
| "logps/rejected": -53.49971008300781, |
| "loss": 0.811690092086792, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.42450016736984253, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.12926799058914185, |
| "rewards/margins": 4.236933708190918, |
| "rewards/rejected": -4.366201877593994, |
| "step": 235, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.403224945272038, |
| "grad_norm": 13.076598167419434, |
| "learning_rate": 0.00018773399752599463, |
| "logits/chosen": -1.001189112663269, |
| "logits/rejected": -1.0065340995788574, |
| "logps/chosen": -8.285541534423828, |
| "logps/rejected": -48.76884460449219, |
| "loss": 0.6642578840255737, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3762781023979187, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.013312354683876038, |
| "rewards/margins": 3.655507802963257, |
| "rewards/rejected": -3.642195701599121, |
| "step": 236, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.40493352554861445, |
| "grad_norm": 3.6171457767486572, |
| "learning_rate": 0.000187597953523356, |
| "logits/chosen": -0.9681710600852966, |
| "logits/rejected": -0.9759187698364258, |
| "logps/chosen": -8.311320304870605, |
| "logps/rejected": -54.762943267822266, |
| "loss": 0.5605093240737915, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.39821937680244446, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.021861745044589043, |
| "rewards/margins": 4.216716289520264, |
| "rewards/rejected": -4.194854736328125, |
| "step": 237, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.4066421058251909, |
| "grad_norm": 15.052586555480957, |
| "learning_rate": 0.00018746120908981527, |
| "logits/chosen": -0.9332698583602905, |
| "logits/rejected": -0.9356809258460999, |
| "logps/chosen": -6.916642189025879, |
| "logps/rejected": -39.75300979614258, |
| "loss": 0.5308848023414612, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.33774426579475403, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.11101798713207245, |
| "rewards/margins": 3.2900469303131104, |
| "rewards/rejected": -3.1790287494659424, |
| "step": 238, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.40835068610176734, |
| "grad_norm": 1.4588333368301392, |
| "learning_rate": 0.00018732376531877725, |
| "logits/chosen": -0.8663907647132874, |
| "logits/rejected": -0.8769683241844177, |
| "logps/chosen": -7.994008541107178, |
| "logps/rejected": -53.477638244628906, |
| "loss": 0.3602023124694824, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.27253228425979614, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.39557039737701416, |
| "rewards/margins": 4.195072174072266, |
| "rewards/rejected": -3.799501895904541, |
| "step": 239, |
| "train_speed(iter/s)": 0.005494 |
| }, |
| { |
| "epoch": 0.4100592663783437, |
| "grad_norm": 2.3394670486450195, |
| "learning_rate": 0.0001871856233092387, |
| "logits/chosen": -0.8316817283630371, |
| "logits/rejected": -0.8333123326301575, |
| "logps/chosen": -9.785093307495117, |
| "logps/rejected": -40.74665069580078, |
| "loss": 0.4092681407928467, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.260936975479126, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.18386238813400269, |
| "rewards/margins": 2.943046808242798, |
| "rewards/rejected": -2.7591843605041504, |
| "step": 240, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.41176784665492017, |
| "grad_norm": 15.97940731048584, |
| "learning_rate": 0.00018704678416577941, |
| "logits/chosen": -0.8120774030685425, |
| "logits/rejected": -0.8134472370147705, |
| "logps/chosen": -11.521053314208984, |
| "logps/rejected": -35.660865783691406, |
| "loss": 0.8933948874473572, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5571330785751343, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.0759764090180397, |
| "rewards/margins": 2.2313168048858643, |
| "rewards/rejected": -2.30729341506958, |
| "step": 241, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.4134764269314966, |
| "grad_norm": 4.722672462463379, |
| "learning_rate": 0.00018690724899855348, |
| "logits/chosen": -0.7959482669830322, |
| "logits/rejected": -0.7984044551849365, |
| "logps/chosen": -5.855531692504883, |
| "logps/rejected": -40.35407257080078, |
| "loss": 0.49105608463287354, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2971791923046112, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.221318781375885, |
| "rewards/margins": 3.0078468322753906, |
| "rewards/rejected": -2.7865281105041504, |
| "step": 242, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.41518500720807305, |
| "grad_norm": 6.576427936553955, |
| "learning_rate": 0.00018676701892328046, |
| "logits/chosen": -0.7806111574172974, |
| "logits/rejected": -0.783092200756073, |
| "logps/chosen": -4.328828811645508, |
| "logps/rejected": -40.24433517456055, |
| "loss": 0.45146334171295166, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.30311349034309387, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.22997835278511047, |
| "rewards/margins": 3.151667833328247, |
| "rewards/rejected": -2.921689510345459, |
| "step": 243, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.4168935874846495, |
| "grad_norm": 3.0237584114074707, |
| "learning_rate": 0.0001866260950612362, |
| "logits/chosen": -0.768136203289032, |
| "logits/rejected": -0.7717478275299072, |
| "logps/chosen": -4.951301097869873, |
| "logps/rejected": -33.46363830566406, |
| "loss": 0.4757350981235504, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.20984753966331482, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.28256160020828247, |
| "rewards/margins": 2.4234559535980225, |
| "rewards/rejected": -2.140894651412964, |
| "step": 244, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.4186021677612259, |
| "grad_norm": 6.6580657958984375, |
| "learning_rate": 0.0001864844785392442, |
| "logits/chosen": -0.745742917060852, |
| "logits/rejected": -0.7464924454689026, |
| "logps/chosen": -10.95466136932373, |
| "logps/rejected": -44.5806999206543, |
| "loss": 0.5750457048416138, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3861500322818756, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.1691206991672516, |
| "rewards/margins": 3.562136650085449, |
| "rewards/rejected": -3.3930160999298096, |
| "step": 245, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.42031074803780233, |
| "grad_norm": 6.422461032867432, |
| "learning_rate": 0.00018634217048966637, |
| "logits/chosen": -0.7587261199951172, |
| "logits/rejected": -0.7620642185211182, |
| "logps/chosen": -6.3410139083862305, |
| "logps/rejected": -33.94182586669922, |
| "loss": 0.6545197367668152, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.30830273032188416, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.06438297033309937, |
| "rewards/margins": 2.2670321464538574, |
| "rewards/rejected": -2.3314151763916016, |
| "step": 246, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.42201932831437877, |
| "grad_norm": 4.497966289520264, |
| "learning_rate": 0.00018619917205039407, |
| "logits/chosen": -0.7642217874526978, |
| "logits/rejected": -0.7698625326156616, |
| "logps/chosen": -6.284250259399414, |
| "logps/rejected": -46.83761215209961, |
| "loss": 0.4105835258960724, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2564704418182373, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.22518175840377808, |
| "rewards/margins": 3.5758605003356934, |
| "rewards/rejected": -3.3506786823272705, |
| "step": 247, |
| "train_speed(iter/s)": 0.005493 |
| }, |
| { |
| "epoch": 0.4237279085909552, |
| "grad_norm": 3.6431024074554443, |
| "learning_rate": 0.00018605548436483906, |
| "logits/chosen": -0.7721464037895203, |
| "logits/rejected": -0.7743059396743774, |
| "logps/chosen": -8.004281044006348, |
| "logps/rejected": -34.3867073059082, |
| "loss": 0.6020833849906921, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3395420014858246, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.2702520787715912, |
| "rewards/margins": 2.563030242919922, |
| "rewards/rejected": -2.292778491973877, |
| "step": 248, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.42543648886753166, |
| "grad_norm": 2.3539841175079346, |
| "learning_rate": 0.0001859111085819242, |
| "logits/chosen": -0.7793521285057068, |
| "logits/rejected": -0.7824581265449524, |
| "logps/chosen": -12.44356918334961, |
| "logps/rejected": -37.551902770996094, |
| "loss": 0.569586455821991, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4044492542743683, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.30807700753211975, |
| "rewards/margins": 2.927927017211914, |
| "rewards/rejected": -2.6198501586914062, |
| "step": 249, |
| "train_speed(iter/s)": 0.005492 |
| }, |
| { |
| "epoch": 0.42714506914410805, |
| "grad_norm": 3.378519058227539, |
| "learning_rate": 0.00018576604585607441, |
| "logits/chosen": -0.7945507168769836, |
| "logits/rejected": -0.7992746233940125, |
| "logps/chosen": -7.014500141143799, |
| "logps/rejected": -46.093238830566406, |
| "loss": 0.47747644782066345, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2733898162841797, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.10495103895664215, |
| "rewards/margins": 3.2655720710754395, |
| "rewards/rejected": -3.160621166229248, |
| "step": 250, |
| "train_speed(iter/s)": 0.005491 |
| }, |
| { |
| "epoch": 0.4288536494206845, |
| "grad_norm": 4.231056213378906, |
| "learning_rate": 0.00018562029734720743, |
| "logits/chosen": -0.8051905632019043, |
| "logits/rejected": -0.8092507123947144, |
| "logps/chosen": -6.725405693054199, |
| "logps/rejected": -37.768829345703125, |
| "loss": 0.5110489726066589, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.31649249792099, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.19485878944396973, |
| "rewards/margins": 2.8023769855499268, |
| "rewards/rejected": -2.607518196105957, |
| "step": 251, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.43056222969726093, |
| "grad_norm": 11.981834411621094, |
| "learning_rate": 0.00018547386422072438, |
| "logits/chosen": -0.7716567516326904, |
| "logits/rejected": -0.773201584815979, |
| "logps/chosen": -9.067695617675781, |
| "logps/rejected": -36.645240783691406, |
| "loss": 0.8504749536514282, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.514085054397583, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.07366381585597992, |
| "rewards/margins": 2.7288200855255127, |
| "rewards/rejected": -2.8024840354919434, |
| "step": 252, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.4322708099738374, |
| "grad_norm": 8.43531608581543, |
| "learning_rate": 0.0001853267476475007, |
| "logits/chosen": -0.8169345855712891, |
| "logits/rejected": -0.8159929513931274, |
| "logps/chosen": -11.076251029968262, |
| "logps/rejected": -42.98237991333008, |
| "loss": 0.7847392559051514, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5229505300521851, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1810646504163742, |
| "rewards/margins": 3.4308125972747803, |
| "rewards/rejected": -3.2497482299804688, |
| "step": 253, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.4339793902504138, |
| "grad_norm": 6.133362770080566, |
| "learning_rate": 0.0001851789488038765, |
| "logits/chosen": -0.7999134063720703, |
| "logits/rejected": -0.8003981709480286, |
| "logps/chosen": -6.108333587646484, |
| "logps/rejected": -41.47050857543945, |
| "loss": 0.6103618741035461, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3681967258453369, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1454550325870514, |
| "rewards/margins": 3.3938965797424316, |
| "rewards/rejected": -3.248441457748413, |
| "step": 254, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.4356879705269902, |
| "grad_norm": 27.34607696533203, |
| "learning_rate": 0.0001850304688716475, |
| "logits/chosen": -0.7655246257781982, |
| "logits/rejected": -0.7673658132553101, |
| "logps/chosen": -9.124423027038574, |
| "logps/rejected": -40.40342330932617, |
| "loss": 0.9848135709762573, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6617517471313477, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.10712965577840805, |
| "rewards/margins": 2.936744213104248, |
| "rewards/rejected": -2.8296148777008057, |
| "step": 255, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.43739655080356665, |
| "grad_norm": 3.467927932739258, |
| "learning_rate": 0.00018488130903805523, |
| "logits/chosen": -0.7815067768096924, |
| "logits/rejected": -0.7830615639686584, |
| "logps/chosen": -6.11367654800415, |
| "logps/rejected": -43.792381286621094, |
| "loss": 0.3926101326942444, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2346831113100052, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.3161233365535736, |
| "rewards/margins": 3.5978031158447266, |
| "rewards/rejected": -3.281679630279541, |
| "step": 256, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.4391051310801431, |
| "grad_norm": 3.6716325283050537, |
| "learning_rate": 0.00018473147049577774, |
| "logits/chosen": -0.8119379878044128, |
| "logits/rejected": -0.8167931437492371, |
| "logps/chosen": -5.306201934814453, |
| "logps/rejected": -48.548152923583984, |
| "loss": 0.3491962254047394, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.214007169008255, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.19937750697135925, |
| "rewards/margins": 3.6415648460388184, |
| "rewards/rejected": -3.442187547683716, |
| "step": 257, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.44081371135671954, |
| "grad_norm": 25.686939239501953, |
| "learning_rate": 0.00018458095444292004, |
| "logits/chosen": -0.825552761554718, |
| "logits/rejected": -0.8325546383857727, |
| "logps/chosen": -5.261894702911377, |
| "logps/rejected": -62.226497650146484, |
| "loss": 0.434018075466156, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.287321001291275, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.32799047231674194, |
| "rewards/margins": 4.790035247802734, |
| "rewards/rejected": -4.462045192718506, |
| "step": 258, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.442522291633296, |
| "grad_norm": 6.754378318786621, |
| "learning_rate": 0.00018442976208300448, |
| "logits/chosen": -0.8270843029022217, |
| "logits/rejected": -0.8281224966049194, |
| "logps/chosen": -8.492132186889648, |
| "logps/rejected": -36.23853302001953, |
| "loss": 0.6711763739585876, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3068382143974304, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.05488736927509308, |
| "rewards/margins": 2.777843952178955, |
| "rewards/rejected": -2.722956418991089, |
| "step": 259, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.44423087190987237, |
| "grad_norm": 4.042886257171631, |
| "learning_rate": 0.00018427789462496113, |
| "logits/chosen": -0.8525445461273193, |
| "logits/rejected": -0.8591835498809814, |
| "logps/chosen": -8.231825828552246, |
| "logps/rejected": -47.20307540893555, |
| "loss": 0.8233700394630432, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5154776573181152, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.00042735040187835693, |
| "rewards/margins": 3.4719269275665283, |
| "rewards/rejected": -3.4723546504974365, |
| "step": 260, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.4459394521864488, |
| "grad_norm": 12.385782241821289, |
| "learning_rate": 0.00018412535328311814, |
| "logits/chosen": -0.822756290435791, |
| "logits/rejected": -0.8276916742324829, |
| "logps/chosen": -7.529094219207764, |
| "logps/rejected": -39.26826477050781, |
| "loss": 0.7842540740966797, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.41948723793029785, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.09604156017303467, |
| "rewards/margins": 2.7830724716186523, |
| "rewards/rejected": -2.8791141510009766, |
| "step": 261, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.44764803246302526, |
| "grad_norm": 2.720581531524658, |
| "learning_rate": 0.00018397213927719199, |
| "logits/chosen": -0.7854418158531189, |
| "logits/rejected": -0.7872219085693359, |
| "logps/chosen": -7.754611968994141, |
| "logps/rejected": -47.40762710571289, |
| "loss": 0.36977341771125793, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.23527176678180695, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.1266416311264038, |
| "rewards/margins": 3.6666605472564697, |
| "rewards/rejected": -3.5400190353393555, |
| "step": 262, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.4493566127396017, |
| "grad_norm": 3.862173557281494, |
| "learning_rate": 0.00018381825383227782, |
| "logits/chosen": -0.8271799087524414, |
| "logits/rejected": -0.831858217716217, |
| "logps/chosen": -6.855618476867676, |
| "logps/rejected": -46.51610565185547, |
| "loss": 0.4688463509082794, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3772401511669159, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.2653267979621887, |
| "rewards/margins": 3.657702922821045, |
| "rewards/rejected": -3.392376184463501, |
| "step": 263, |
| "train_speed(iter/s)": 0.00549 |
| }, |
| { |
| "epoch": 0.45106519301617815, |
| "grad_norm": 2.1108744144439697, |
| "learning_rate": 0.00018366369817883954, |
| "logits/chosen": -0.8001513481140137, |
| "logits/rejected": -0.8060557842254639, |
| "logps/chosen": -5.957093715667725, |
| "logps/rejected": -49.00318908691406, |
| "loss": 0.42761629819869995, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2863171100616455, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.46388381719589233, |
| "rewards/margins": 3.825634002685547, |
| "rewards/rejected": -3.3617501258850098, |
| "step": 264, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.45277377329275453, |
| "grad_norm": 4.1208696365356445, |
| "learning_rate": 0.00018350847355270004, |
| "logits/chosen": -0.7904451489448547, |
| "logits/rejected": -0.7935886979103088, |
| "logps/chosen": -12.136712074279785, |
| "logps/rejected": -48.772483825683594, |
| "loss": 0.6188664436340332, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4114742577075958, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.2880480885505676, |
| "rewards/margins": 3.680865526199341, |
| "rewards/rejected": -3.392817258834839, |
| "step": 265, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.454482353569331, |
| "grad_norm": 5.501693248748779, |
| "learning_rate": 0.00018335258119503127, |
| "logits/chosen": -0.8123242259025574, |
| "logits/rejected": -0.8159137964248657, |
| "logps/chosen": -6.165188789367676, |
| "logps/rejected": -47.05010986328125, |
| "loss": 0.45226797461509705, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.35374370217323303, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.20915599167346954, |
| "rewards/margins": 3.658740997314453, |
| "rewards/rejected": -3.4495849609375, |
| "step": 266, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.4561909338459074, |
| "grad_norm": 11.674309730529785, |
| "learning_rate": 0.00018319602235234442, |
| "logits/chosen": -0.7954471111297607, |
| "logits/rejected": -0.7979443073272705, |
| "logps/chosen": -9.799958229064941, |
| "logps/rejected": -51.7642822265625, |
| "loss": 0.7218329310417175, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.44642627239227295, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.12691040337085724, |
| "rewards/margins": 3.8264598846435547, |
| "rewards/rejected": -3.9533700942993164, |
| "step": 267, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.45789951412248386, |
| "grad_norm": 21.236757278442383, |
| "learning_rate": 0.00018303879827647975, |
| "logits/chosen": -0.7900252938270569, |
| "logits/rejected": -0.7937359809875488, |
| "logps/chosen": -9.592187881469727, |
| "logps/rejected": -43.07557678222656, |
| "loss": 0.6332522034645081, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5019268989562988, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1493360996246338, |
| "rewards/margins": 3.395822763442993, |
| "rewards/rejected": -3.2464866638183594, |
| "step": 268, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.4596080943990603, |
| "grad_norm": 13.447307586669922, |
| "learning_rate": 0.00018288091022459685, |
| "logits/chosen": -0.7903649210929871, |
| "logits/rejected": -0.792717695236206, |
| "logps/chosen": -8.880897521972656, |
| "logps/rejected": -48.42265701293945, |
| "loss": 1.129546046257019, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.9457821249961853, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": -0.0744566023349762, |
| "rewards/margins": 3.6476986408233643, |
| "rewards/rejected": -3.7221555709838867, |
| "step": 269, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.4613166746756367, |
| "grad_norm": 15.565564155578613, |
| "learning_rate": 0.00018272235945916436, |
| "logits/chosen": -0.7472554445266724, |
| "logits/rejected": -0.7549644708633423, |
| "logps/chosen": -13.44924259185791, |
| "logps/rejected": -50.098731994628906, |
| "loss": 1.1531708240509033, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.8039681911468506, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.14791038632392883, |
| "rewards/margins": 3.323791980743408, |
| "rewards/rejected": -3.4717020988464355, |
| "step": 270, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.46302525495221314, |
| "grad_norm": 12.007596015930176, |
| "learning_rate": 0.00018256314724794992, |
| "logits/chosen": -0.8131598234176636, |
| "logits/rejected": -0.8223381638526917, |
| "logps/chosen": -5.585975646972656, |
| "logps/rejected": -48.965389251708984, |
| "loss": 0.4150363802909851, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2882520854473114, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.21873319149017334, |
| "rewards/margins": 3.403482675552368, |
| "rewards/rejected": -3.1847496032714844, |
| "step": 271, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.4647338352287896, |
| "grad_norm": 16.59933853149414, |
| "learning_rate": 0.00018240327486401022, |
| "logits/chosen": -0.8030964136123657, |
| "logits/rejected": -0.8079983592033386, |
| "logps/chosen": -8.950094223022461, |
| "logps/rejected": -47.7308349609375, |
| "loss": 0.9466665387153625, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6910676956176758, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.06938548386096954, |
| "rewards/margins": 3.2243666648864746, |
| "rewards/rejected": -3.293752431869507, |
| "step": 272, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.466442415505366, |
| "grad_norm": 4.702775955200195, |
| "learning_rate": 0.00018224274358568052, |
| "logits/chosen": -0.8188191652297974, |
| "logits/rejected": -0.8212082982063293, |
| "logps/chosen": -7.591650485992432, |
| "logps/rejected": -37.239356994628906, |
| "loss": 0.5474280714988708, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3606598973274231, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.29892435669898987, |
| "rewards/margins": 3.0769505500793457, |
| "rewards/rejected": -2.7780261039733887, |
| "step": 273, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.46815099578194247, |
| "grad_norm": 27.876619338989258, |
| "learning_rate": 0.00018208155469656462, |
| "logits/chosen": -0.8266133666038513, |
| "logits/rejected": -0.8309741020202637, |
| "logps/chosen": -6.996228218078613, |
| "logps/rejected": -38.674766540527344, |
| "loss": 0.8267521858215332, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5320640206336975, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.10675782710313797, |
| "rewards/margins": 2.87463116645813, |
| "rewards/rejected": -2.7678730487823486, |
| "step": 274, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.46985957605851886, |
| "grad_norm": 7.991383075714111, |
| "learning_rate": 0.00018191970948552464, |
| "logits/chosen": -0.8153300285339355, |
| "logits/rejected": -0.8158446550369263, |
| "logps/chosen": -8.539220809936523, |
| "logps/rejected": -34.5740966796875, |
| "loss": 0.5116520524024963, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2977459728717804, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.13776344060897827, |
| "rewards/margins": 2.662919282913208, |
| "rewards/rejected": -2.525156021118164, |
| "step": 275, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.4715681563350953, |
| "grad_norm": 4.298219203948975, |
| "learning_rate": 0.00018175720924667056, |
| "logits/chosen": -0.853663444519043, |
| "logits/rejected": -0.8555605411529541, |
| "logps/chosen": -7.882385730743408, |
| "logps/rejected": -27.5541934967041, |
| "loss": 0.5915999412536621, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.284732460975647, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.12615913152694702, |
| "rewards/margins": 1.954941749572754, |
| "rewards/rejected": -1.8287824392318726, |
| "step": 276, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.47327673661167174, |
| "grad_norm": 3.9946365356445312, |
| "learning_rate": 0.00018159405527934996, |
| "logits/chosen": -0.8487964868545532, |
| "logits/rejected": -0.8562285304069519, |
| "logps/chosen": -7.136427879333496, |
| "logps/rejected": -37.46531677246094, |
| "loss": 0.5510011315345764, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2794538140296936, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.141952246427536, |
| "rewards/margins": 2.4826676845550537, |
| "rewards/rejected": -2.3407156467437744, |
| "step": 277, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.4749853168882482, |
| "grad_norm": 1.9724056720733643, |
| "learning_rate": 0.0001814302488881376, |
| "logits/chosen": -0.8201677799224854, |
| "logits/rejected": -0.8276962041854858, |
| "logps/chosen": -4.926413536071777, |
| "logps/rejected": -38.361087799072266, |
| "loss": 0.45584461092948914, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2575777769088745, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.190385103225708, |
| "rewards/margins": 2.507413387298584, |
| "rewards/rejected": -2.317028522491455, |
| "step": 278, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.4766938971648246, |
| "grad_norm": 8.133130073547363, |
| "learning_rate": 0.00018126579138282503, |
| "logits/chosen": -0.8125935792922974, |
| "logits/rejected": -0.8118517398834229, |
| "logps/chosen": -9.689276695251465, |
| "logps/rejected": -29.489994049072266, |
| "loss": 0.7013036608695984, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3926777243614197, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.06979987025260925, |
| "rewards/margins": 2.248946189880371, |
| "rewards/rejected": -2.1791462898254395, |
| "step": 279, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.478402477441401, |
| "grad_norm": 5.878256797790527, |
| "learning_rate": 0.00018110068407841009, |
| "logits/chosen": -0.8344768285751343, |
| "logits/rejected": -0.8355271816253662, |
| "logps/chosen": -7.216890335083008, |
| "logps/rejected": -31.038660049438477, |
| "loss": 0.6482636332511902, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3379790782928467, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.15157388150691986, |
| "rewards/margins": 2.1786890029907227, |
| "rewards/rejected": -2.0271151065826416, |
| "step": 280, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.48011105771797746, |
| "grad_norm": 3.594700336456299, |
| "learning_rate": 0.00018093492829508636, |
| "logits/chosen": -0.809922456741333, |
| "logits/rejected": -0.8155569434165955, |
| "logps/chosen": -8.301997184753418, |
| "logps/rejected": -35.9427604675293, |
| "loss": 0.7471737265586853, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.412964791059494, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.21606266498565674, |
| "rewards/margins": 2.184842824935913, |
| "rewards/rejected": -1.9687800407409668, |
| "step": 281, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.4818196379945539, |
| "grad_norm": 4.715336799621582, |
| "learning_rate": 0.00018076852535823268, |
| "logits/chosen": -0.8024042844772339, |
| "logits/rejected": -0.8047822713851929, |
| "logps/chosen": -9.50023365020752, |
| "logps/rejected": -26.15119743347168, |
| "loss": 0.887703001499176, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4939606785774231, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.09452123939990997, |
| "rewards/margins": 1.60733962059021, |
| "rewards/rejected": -1.512818455696106, |
| "step": 282, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.48352821827113035, |
| "grad_norm": 2.6357319355010986, |
| "learning_rate": 0.0001806014765984025, |
| "logits/chosen": -0.7931514978408813, |
| "logits/rejected": -0.7975669503211975, |
| "logps/chosen": -4.370926856994629, |
| "logps/rejected": -35.80842590332031, |
| "loss": 0.4094904065132141, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.17323070764541626, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.25759178400039673, |
| "rewards/margins": 2.2846627235412598, |
| "rewards/rejected": -2.027071237564087, |
| "step": 283, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.48523679854770674, |
| "grad_norm": 4.521765232086182, |
| "learning_rate": 0.00018043378335131326, |
| "logits/chosen": -0.8225452899932861, |
| "logits/rejected": -0.8211522698402405, |
| "logps/chosen": -7.510035037994385, |
| "logps/rejected": -36.8507194519043, |
| "loss": 0.5709434747695923, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.35384440422058105, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.24606043100357056, |
| "rewards/margins": 2.5631608963012695, |
| "rewards/rejected": -2.3171002864837646, |
| "step": 284, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.4869453788242832, |
| "grad_norm": 2.36458683013916, |
| "learning_rate": 0.00018026544695783566, |
| "logits/chosen": -0.8055735230445862, |
| "logits/rejected": -0.8072599768638611, |
| "logps/chosen": -8.788959503173828, |
| "logps/rejected": -40.345008850097656, |
| "loss": 0.5267823338508606, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.34875571727752686, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.25422000885009766, |
| "rewards/margins": 2.772775650024414, |
| "rewards/rejected": -2.5185558795928955, |
| "step": 285, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.4886539591008596, |
| "grad_norm": 10.764819145202637, |
| "learning_rate": 0.00018009646876398307, |
| "logits/chosen": -0.8059338331222534, |
| "logits/rejected": -0.8106356263160706, |
| "logps/chosen": -4.930308818817139, |
| "logps/rejected": -31.040836334228516, |
| "loss": 0.5658331513404846, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.28465500473976135, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1854298710823059, |
| "rewards/margins": 1.930139422416687, |
| "rewards/rejected": -1.7447094917297363, |
| "step": 286, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.49036253937743607, |
| "grad_norm": 3.613628387451172, |
| "learning_rate": 0.0001799268501209006, |
| "logits/chosen": -0.7612501382827759, |
| "logits/rejected": -0.7627686858177185, |
| "logps/chosen": -10.456869125366211, |
| "logps/rejected": -28.227718353271484, |
| "loss": 0.8326478004455566, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4999573230743408, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.05842627212405205, |
| "rewards/margins": 1.9171324968338013, |
| "rewards/rejected": -1.8587062358856201, |
| "step": 287, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.4920711196540125, |
| "grad_norm": 4.662235736846924, |
| "learning_rate": 0.0001797565923848544, |
| "logits/chosen": -0.8021095991134644, |
| "logits/rejected": -0.8046351671218872, |
| "logps/chosen": -6.780048370361328, |
| "logps/rejected": -31.340526580810547, |
| "loss": 0.5166000723838806, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.29529324173927307, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.24387100338935852, |
| "rewards/margins": 2.386127233505249, |
| "rewards/rejected": -2.142256259918213, |
| "step": 288, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.4937796999305889, |
| "grad_norm": 3.747556447982788, |
| "learning_rate": 0.00017958569691722077, |
| "logits/chosen": -0.7904728055000305, |
| "logits/rejected": -0.7943806052207947, |
| "logps/chosen": -4.052952766418457, |
| "logps/rejected": -39.81676483154297, |
| "loss": 0.444693922996521, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.26461470127105713, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.24267469346523285, |
| "rewards/margins": 3.1456563472747803, |
| "rewards/rejected": -2.9029815196990967, |
| "step": 289, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.49548828020716534, |
| "grad_norm": 3.8240482807159424, |
| "learning_rate": 0.00017941416508447536, |
| "logits/chosen": -0.8258907794952393, |
| "logits/rejected": -0.8271006345748901, |
| "logps/chosen": -12.335704803466797, |
| "logps/rejected": -35.24673843383789, |
| "loss": 0.8288266062736511, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5496087670326233, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.26559069752693176, |
| "rewards/margins": 2.4179635047912598, |
| "rewards/rejected": -2.1523728370666504, |
| "step": 290, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.4971968604837418, |
| "grad_norm": 5.618929386138916, |
| "learning_rate": 0.0001792419982581821, |
| "logits/chosen": -0.8527386784553528, |
| "logits/rejected": -0.8554395437240601, |
| "logps/chosen": -4.423219203948975, |
| "logps/rejected": -37.118656158447266, |
| "loss": 0.4001955986022949, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2208787500858307, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2664373517036438, |
| "rewards/margins": 2.81235933303833, |
| "rewards/rejected": -2.545922040939331, |
| "step": 291, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.49890544076031823, |
| "grad_norm": 4.9184112548828125, |
| "learning_rate": 0.00017906919781498235, |
| "logits/chosen": -0.8180443048477173, |
| "logits/rejected": -0.8243182897567749, |
| "logps/chosen": -7.195798873901367, |
| "logps/rejected": -46.805763244628906, |
| "loss": 0.50282222032547, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.24749858677387238, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.17263692617416382, |
| "rewards/margins": 3.1018624305725098, |
| "rewards/rejected": -2.929225444793701, |
| "step": 292, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.5006140210368947, |
| "grad_norm": 6.550334930419922, |
| "learning_rate": 0.00017889576513658385, |
| "logits/chosen": -0.8828959465026855, |
| "logits/rejected": -0.8903377652168274, |
| "logps/chosen": -5.666334629058838, |
| "logps/rejected": -41.20561218261719, |
| "loss": 0.5666110515594482, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.27025288343429565, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.14388155937194824, |
| "rewards/margins": 2.9238851070404053, |
| "rewards/rejected": -2.780003547668457, |
| "step": 293, |
| "train_speed(iter/s)": 0.005487 |
| }, |
| { |
| "epoch": 0.5023226013134711, |
| "grad_norm": 5.214411735534668, |
| "learning_rate": 0.0001787217016097497, |
| "logits/chosen": -0.8843887448310852, |
| "logits/rejected": -0.8880648612976074, |
| "logps/chosen": -8.593646049499512, |
| "logps/rejected": -31.845436096191406, |
| "loss": 0.7616337537765503, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4269709587097168, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.03095342591404915, |
| "rewards/margins": 2.152078628540039, |
| "rewards/rejected": -2.1211252212524414, |
| "step": 294, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.5040311815900476, |
| "grad_norm": 3.419295310974121, |
| "learning_rate": 0.00017854700862628717, |
| "logits/chosen": -0.8697606325149536, |
| "logits/rejected": -0.8711093068122864, |
| "logps/chosen": -12.32712459564209, |
| "logps/rejected": -35.99604415893555, |
| "loss": 0.7168628573417664, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4990694522857666, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": -0.00020820647478103638, |
| "rewards/margins": 2.4814581871032715, |
| "rewards/rejected": -2.481666088104248, |
| "step": 295, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.505739761866624, |
| "grad_norm": 5.47044038772583, |
| "learning_rate": 0.00017837168758303673, |
| "logits/chosen": -0.9019865989685059, |
| "logits/rejected": -0.9105886220932007, |
| "logps/chosen": -6.227141380310059, |
| "logps/rejected": -53.5411376953125, |
| "loss": 0.4358929395675659, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2536502182483673, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.18560552597045898, |
| "rewards/margins": 4.006980895996094, |
| "rewards/rejected": -3.8213753700256348, |
| "step": 296, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.5074483421432003, |
| "grad_norm": 2.2059149742126465, |
| "learning_rate": 0.00017819573988186077, |
| "logits/chosen": -0.8752774000167847, |
| "logits/rejected": -0.8806784749031067, |
| "logps/chosen": -6.102680206298828, |
| "logps/rejected": -40.7698860168457, |
| "loss": 0.5294958353042603, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.32509148120880127, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.20541568100452423, |
| "rewards/margins": 2.9198334217071533, |
| "rewards/rejected": -2.7144179344177246, |
| "step": 297, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.5091569224197768, |
| "grad_norm": 1.8315078020095825, |
| "learning_rate": 0.00017801916692963246, |
| "logits/chosen": -0.888721764087677, |
| "logits/rejected": -0.8927589654922485, |
| "logps/chosen": -5.959261417388916, |
| "logps/rejected": -39.12718963623047, |
| "loss": 0.41302210092544556, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.25391918420791626, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.36076462268829346, |
| "rewards/margins": 2.8598427772521973, |
| "rewards/rejected": -2.4990782737731934, |
| "step": 298, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.5108655026963532, |
| "grad_norm": 2.749384641647339, |
| "learning_rate": 0.00017784197013822435, |
| "logits/chosen": -0.8662399053573608, |
| "logits/rejected": -0.8707250356674194, |
| "logps/chosen": -9.147662162780762, |
| "logps/rejected": -45.20575714111328, |
| "loss": 0.6315680146217346, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.45470985770225525, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.07495459169149399, |
| "rewards/margins": 3.035630226135254, |
| "rewards/rejected": -2.9606757164001465, |
| "step": 299, |
| "train_speed(iter/s)": 0.005489 |
| }, |
| { |
| "epoch": 0.5125740829729297, |
| "grad_norm": 8.072540283203125, |
| "learning_rate": 0.0001776641509244973, |
| "logits/chosen": -0.8409525156021118, |
| "logits/rejected": -0.8496344685554504, |
| "logps/chosen": -6.282418251037598, |
| "logps/rejected": -41.63984680175781, |
| "loss": 0.871893584728241, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.603092610836029, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.04116540029644966, |
| "rewards/margins": 2.8330185413360596, |
| "rewards/rejected": -2.7918529510498047, |
| "step": 300, |
| "train_speed(iter/s)": 0.005488 |
| }, |
| { |
| "epoch": 0.5125740829729297, |
| "eval_logits/chosen": -0.8405543565750122, |
| "eval_logits/rejected": -0.8445731401443481, |
| "eval_logps/chosen": -8.006686210632324, |
| "eval_logps/rejected": -40.67301559448242, |
| "eval_loss": 0.7707608342170715, |
| "eval_nll_loss": 0.5364512801170349, |
| "eval_rewards/accuracies": 0.8994709253311157, |
| "eval_rewards/chosen": 0.06882528215646744, |
| "eval_rewards/margins": 2.7935357093811035, |
| "eval_rewards/rejected": -2.724710702896118, |
| "eval_runtime": 442.8747, |
| "eval_samples_per_second": 0.427, |
| "eval_steps_per_second": 0.427, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.5142826632495061, |
| "grad_norm": 11.069442749023438, |
| "learning_rate": 0.000177485710710289, |
| "logits/chosen": -0.8602814674377441, |
| "logits/rejected": -0.8637278079986572, |
| "logps/chosen": -7.823677062988281, |
| "logps/rejected": -39.174312591552734, |
| "loss": 0.45352113246917725, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.32754451036453247, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.15461236238479614, |
| "rewards/margins": 3.001373291015625, |
| "rewards/rejected": -2.8467607498168945, |
| "step": 301, |
| "train_speed(iter/s)": 0.005443 |
| }, |
| { |
| "epoch": 0.5159912435260825, |
| "grad_norm": 4.37824821472168, |
| "learning_rate": 0.0001773066509224026, |
| "logits/chosen": -0.7942736744880676, |
| "logits/rejected": -0.7966884970664978, |
| "logps/chosen": -7.395168304443359, |
| "logps/rejected": -37.037654876708984, |
| "loss": 0.6269878149032593, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4148014783859253, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.27507781982421875, |
| "rewards/margins": 2.9940710067749023, |
| "rewards/rejected": -2.7189929485321045, |
| "step": 302, |
| "train_speed(iter/s)": 0.005443 |
| }, |
| { |
| "epoch": 0.517699823802659, |
| "grad_norm": 1.2747763395309448, |
| "learning_rate": 0.0001771269729925955, |
| "logits/chosen": -0.8845440745353699, |
| "logits/rejected": -0.8931821584701538, |
| "logps/chosen": -4.624814510345459, |
| "logps/rejected": -54.01844787597656, |
| "loss": 0.3171394467353821, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2531016170978546, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.3055804371833801, |
| "rewards/margins": 4.609305381774902, |
| "rewards/rejected": -4.303724765777588, |
| "step": 303, |
| "train_speed(iter/s)": 0.005444 |
| }, |
| { |
| "epoch": 0.5194084040792354, |
| "grad_norm": 19.4792423248291, |
| "learning_rate": 0.00017694667835756758, |
| "logits/chosen": -0.8769544363021851, |
| "logits/rejected": -0.8785020112991333, |
| "logps/chosen": -13.249614715576172, |
| "logps/rejected": -43.8223991394043, |
| "loss": 1.2327057123184204, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.8601534962654114, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.30870580673217773, |
| "rewards/margins": 2.7113828659057617, |
| "rewards/rejected": -3.0200884342193604, |
| "step": 304, |
| "train_speed(iter/s)": 0.005444 |
| }, |
| { |
| "epoch": 0.5211169843558119, |
| "grad_norm": 9.179790496826172, |
| "learning_rate": 0.0001767657684589499, |
| "logits/chosen": -0.8855112195014954, |
| "logits/rejected": -0.8912838101387024, |
| "logps/chosen": -7.211820602416992, |
| "logps/rejected": -50.378562927246094, |
| "loss": 0.7304961085319519, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.42998042702674866, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.09179924428462982, |
| "rewards/margins": 4.0141072273254395, |
| "rewards/rejected": -3.9223079681396484, |
| "step": 305, |
| "train_speed(iter/s)": 0.005444 |
| }, |
| { |
| "epoch": 0.5228255646323883, |
| "grad_norm": 1.2901352643966675, |
| "learning_rate": 0.00017658424474329326, |
| "logits/chosen": -0.8675682544708252, |
| "logits/rejected": -0.8787212371826172, |
| "logps/chosen": -6.3633317947387695, |
| "logps/rejected": -68.60183715820312, |
| "loss": 0.31616559624671936, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.265287309885025, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.21000903844833374, |
| "rewards/margins": 5.580310344696045, |
| "rewards/rejected": -5.370301723480225, |
| "step": 306, |
| "train_speed(iter/s)": 0.005444 |
| }, |
| { |
| "epoch": 0.5245341449089647, |
| "grad_norm": 4.6658782958984375, |
| "learning_rate": 0.00017640210866205639, |
| "logits/chosen": -0.906684935092926, |
| "logits/rejected": -0.9144970774650574, |
| "logps/chosen": -9.675024032592773, |
| "logps/rejected": -51.09449768066406, |
| "loss": 0.6057628989219666, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4204673767089844, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.10616739839315414, |
| "rewards/margins": 3.975714921951294, |
| "rewards/rejected": -4.081882476806641, |
| "step": 307, |
| "train_speed(iter/s)": 0.005445 |
| }, |
| { |
| "epoch": 0.5262427251855412, |
| "grad_norm": 5.465089321136475, |
| "learning_rate": 0.00017621936167159458, |
| "logits/chosen": -0.9267672300338745, |
| "logits/rejected": -0.9362570643424988, |
| "logps/chosen": -7.424238681793213, |
| "logps/rejected": -65.96898651123047, |
| "loss": 0.47528597712516785, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3133973479270935, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.09580647945404053, |
| "rewards/margins": 5.268470287322998, |
| "rewards/rejected": -5.364277362823486, |
| "step": 308, |
| "train_speed(iter/s)": 0.005445 |
| }, |
| { |
| "epoch": 0.5279513054621175, |
| "grad_norm": 5.624989986419678, |
| "learning_rate": 0.00017603600523314787, |
| "logits/chosen": -0.898685097694397, |
| "logits/rejected": -0.8996263146400452, |
| "logps/chosen": -12.03923225402832, |
| "logps/rejected": -41.64332962036133, |
| "loss": 0.7952063083648682, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4172392785549164, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.023442525416612625, |
| "rewards/margins": 3.3135008811950684, |
| "rewards/rejected": -3.2900586128234863, |
| "step": 309, |
| "train_speed(iter/s)": 0.005446 |
| }, |
| { |
| "epoch": 0.529659885738694, |
| "grad_norm": 3.880650043487549, |
| "learning_rate": 0.0001758520408128295, |
| "logits/chosen": -0.8696607947349548, |
| "logits/rejected": -0.8726595640182495, |
| "logps/chosen": -7.42938232421875, |
| "logps/rejected": -56.99799346923828, |
| "loss": 0.5451213717460632, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3175729513168335, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.16648581624031067, |
| "rewards/margins": 4.477063179016113, |
| "rewards/rejected": -4.310576915740967, |
| "step": 310, |
| "train_speed(iter/s)": 0.005445 |
| }, |
| { |
| "epoch": 0.5313684660152704, |
| "grad_norm": 2.883660316467285, |
| "learning_rate": 0.00017566746988161402, |
| "logits/chosen": -0.8406087756156921, |
| "logits/rejected": -0.8471619486808777, |
| "logps/chosen": -9.634123802185059, |
| "logps/rejected": -38.94046401977539, |
| "loss": 0.6529473662376404, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.38416606187820435, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.13221141695976257, |
| "rewards/margins": 3.1708903312683105, |
| "rewards/rejected": -3.0386788845062256, |
| "step": 311, |
| "train_speed(iter/s)": 0.005446 |
| }, |
| { |
| "epoch": 0.5330770462918468, |
| "grad_norm": 11.171285629272461, |
| "learning_rate": 0.00017548229391532572, |
| "logits/chosen": -0.8072686791419983, |
| "logits/rejected": -0.8102389574050903, |
| "logps/chosen": -6.779449462890625, |
| "logps/rejected": -49.09239959716797, |
| "loss": 0.5393321514129639, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3658963441848755, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.2095126211643219, |
| "rewards/margins": 3.693910598754883, |
| "rewards/rejected": -3.4843978881835938, |
| "step": 312, |
| "train_speed(iter/s)": 0.005446 |
| }, |
| { |
| "epoch": 0.5347856265684233, |
| "grad_norm": 1.954391360282898, |
| "learning_rate": 0.00017529651439462664, |
| "logits/chosen": -0.7809416055679321, |
| "logits/rejected": -0.7857435345649719, |
| "logps/chosen": -6.740390777587891, |
| "logps/rejected": -48.86231994628906, |
| "loss": 0.36391139030456543, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2839314639568329, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.31254059076309204, |
| "rewards/margins": 3.944000482559204, |
| "rewards/rejected": -3.6314597129821777, |
| "step": 313, |
| "train_speed(iter/s)": 0.005446 |
| }, |
| { |
| "epoch": 0.5364942068449997, |
| "grad_norm": 6.087454319000244, |
| "learning_rate": 0.00017511013280500494, |
| "logits/chosen": -0.7839592695236206, |
| "logits/rejected": -0.7883790731430054, |
| "logps/chosen": -5.397429943084717, |
| "logps/rejected": -38.226078033447266, |
| "loss": 0.47547078132629395, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2724950909614563, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.16703476011753082, |
| "rewards/margins": 2.7412147521972656, |
| "rewards/rejected": -2.5741801261901855, |
| "step": 314, |
| "train_speed(iter/s)": 0.005445 |
| }, |
| { |
| "epoch": 0.5382027871215762, |
| "grad_norm": 7.710384845733643, |
| "learning_rate": 0.0001749231506367628, |
| "logits/chosen": -0.7206589579582214, |
| "logits/rejected": -0.7282582521438599, |
| "logps/chosen": -8.630600929260254, |
| "logps/rejected": -40.66322708129883, |
| "loss": 0.7258225679397583, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4553995132446289, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.14358378946781158, |
| "rewards/margins": 2.626176595687866, |
| "rewards/rejected": -2.482593059539795, |
| "step": 315, |
| "train_speed(iter/s)": 0.005446 |
| }, |
| { |
| "epoch": 0.5399113673981526, |
| "grad_norm": 1.8784159421920776, |
| "learning_rate": 0.0001747355693850047, |
| "logits/chosen": -0.7431631684303284, |
| "logits/rejected": -0.748499870300293, |
| "logps/chosen": -11.859859466552734, |
| "logps/rejected": -43.0447883605957, |
| "loss": 0.6120511293411255, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4628870189189911, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.1693616509437561, |
| "rewards/margins": 2.6937308311462402, |
| "rewards/rejected": -2.524369239807129, |
| "step": 316, |
| "train_speed(iter/s)": 0.005445 |
| }, |
| { |
| "epoch": 0.541619947674729, |
| "grad_norm": 2.4822375774383545, |
| "learning_rate": 0.00017454739054962527, |
| "logits/chosen": -0.7296395301818848, |
| "logits/rejected": -0.7350925207138062, |
| "logps/chosen": -7.4225616455078125, |
| "logps/rejected": -44.191368103027344, |
| "loss": 0.5170390605926514, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3234650194644928, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.14995595812797546, |
| "rewards/margins": 3.103259563446045, |
| "rewards/rejected": -2.953303813934326, |
| "step": 317, |
| "train_speed(iter/s)": 0.005445 |
| }, |
| { |
| "epoch": 0.5433285279513055, |
| "grad_norm": 1.8158119916915894, |
| "learning_rate": 0.00017435861563529746, |
| "logits/chosen": -0.7801295518875122, |
| "logits/rejected": -0.7813860177993774, |
| "logps/chosen": -5.427134990692139, |
| "logps/rejected": -36.239803314208984, |
| "loss": 0.36503538489341736, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.19436712563037872, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.22437262535095215, |
| "rewards/margins": 2.704486131668091, |
| "rewards/rejected": -2.4801137447357178, |
| "step": 318, |
| "train_speed(iter/s)": 0.005446 |
| }, |
| { |
| "epoch": 0.5450371082278819, |
| "grad_norm": 1.2651731967926025, |
| "learning_rate": 0.00017416924615146055, |
| "logits/chosen": -0.7636649012565613, |
| "logits/rejected": -0.7674313187599182, |
| "logps/chosen": -4.830596923828125, |
| "logps/rejected": -49.42289733886719, |
| "loss": 0.2704404294490814, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.17981436848640442, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2880362272262573, |
| "rewards/margins": 3.9220433235168457, |
| "rewards/rejected": -3.634007215499878, |
| "step": 319, |
| "train_speed(iter/s)": 0.005446 |
| }, |
| { |
| "epoch": 0.5467456885044584, |
| "grad_norm": 2.813702344894409, |
| "learning_rate": 0.0001739792836123078, |
| "logits/chosen": -0.763135552406311, |
| "logits/rejected": -0.7641803622245789, |
| "logps/chosen": -8.197504043579102, |
| "logps/rejected": -46.546730041503906, |
| "loss": 0.43457722663879395, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3029811978340149, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.21454845368862152, |
| "rewards/margins": 3.952043056488037, |
| "rewards/rejected": -3.737494707107544, |
| "step": 320, |
| "train_speed(iter/s)": 0.005446 |
| }, |
| { |
| "epoch": 0.5484542687810348, |
| "grad_norm": 10.072509765625, |
| "learning_rate": 0.00017378872953677463, |
| "logits/chosen": -0.7955141663551331, |
| "logits/rejected": -0.8014116883277893, |
| "logps/chosen": -6.334041595458984, |
| "logps/rejected": -47.14308166503906, |
| "loss": 0.4273318648338318, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.26563501358032227, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.30760833621025085, |
| "rewards/margins": 3.725619316101074, |
| "rewards/rejected": -3.418011426925659, |
| "step": 321, |
| "train_speed(iter/s)": 0.005446 |
| }, |
| { |
| "epoch": 0.5501628490576111, |
| "grad_norm": 11.081689834594727, |
| "learning_rate": 0.00017359758544852635, |
| "logits/chosen": -0.8159039616584778, |
| "logits/rejected": -0.8194667100906372, |
| "logps/chosen": -9.277294158935547, |
| "logps/rejected": -51.40448760986328, |
| "loss": 0.8425000309944153, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.717254102230072, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.03732386231422424, |
| "rewards/margins": 4.070914268493652, |
| "rewards/rejected": -4.033590316772461, |
| "step": 322, |
| "train_speed(iter/s)": 0.005447 |
| }, |
| { |
| "epoch": 0.5518714293341876, |
| "grad_norm": 2.0417747497558594, |
| "learning_rate": 0.00017340585287594604, |
| "logits/chosen": -0.8340936899185181, |
| "logits/rejected": -0.8355550169944763, |
| "logps/chosen": -8.740560531616211, |
| "logps/rejected": -48.48387908935547, |
| "loss": 0.5082376599311829, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3596939742565155, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3152270019054413, |
| "rewards/margins": 3.8847768306732178, |
| "rewards/rejected": -3.569550037384033, |
| "step": 323, |
| "train_speed(iter/s)": 0.005447 |
| }, |
| { |
| "epoch": 0.553580009610764, |
| "grad_norm": 11.896222114562988, |
| "learning_rate": 0.00017321353335212215, |
| "logits/chosen": -0.8562039732933044, |
| "logits/rejected": -0.8574480414390564, |
| "logps/chosen": -9.903059005737305, |
| "logps/rejected": -48.68834686279297, |
| "loss": 0.5927642583847046, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3851415812969208, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.1212349608540535, |
| "rewards/margins": 3.81803560256958, |
| "rewards/rejected": -3.696800708770752, |
| "step": 324, |
| "train_speed(iter/s)": 0.005447 |
| }, |
| { |
| "epoch": 0.5552885898873405, |
| "grad_norm": 3.4423718452453613, |
| "learning_rate": 0.00017302062841483655, |
| "logits/chosen": -0.8601418733596802, |
| "logits/rejected": -0.8596406579017639, |
| "logps/chosen": -9.926023483276367, |
| "logps/rejected": -53.83884811401367, |
| "loss": 0.45005056262016296, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.34139543771743774, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.20918819308280945, |
| "rewards/margins": 4.317967891693115, |
| "rewards/rejected": -4.108779430389404, |
| "step": 325, |
| "train_speed(iter/s)": 0.005447 |
| }, |
| { |
| "epoch": 0.5569971701639169, |
| "grad_norm": 1.2915583848953247, |
| "learning_rate": 0.00017282713960655195, |
| "logits/chosen": -0.9310311079025269, |
| "logits/rejected": -0.9432608485221863, |
| "logps/chosen": -3.527310848236084, |
| "logps/rejected": -56.88557434082031, |
| "loss": 0.257373183965683, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.18794924020767212, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.31664133071899414, |
| "rewards/margins": 4.63478422164917, |
| "rewards/rejected": -4.318142890930176, |
| "step": 326, |
| "train_speed(iter/s)": 0.005447 |
| }, |
| { |
| "epoch": 0.5587057504404933, |
| "grad_norm": 12.496362686157227, |
| "learning_rate": 0.00017263306847439968, |
| "logits/chosen": -0.8998693227767944, |
| "logits/rejected": -0.9072844386100769, |
| "logps/chosen": -7.473057270050049, |
| "logps/rejected": -62.43585968017578, |
| "loss": 0.4956499934196472, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.36785978078842163, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2565673589706421, |
| "rewards/margins": 4.963648796081543, |
| "rewards/rejected": -4.707081317901611, |
| "step": 327, |
| "train_speed(iter/s)": 0.005447 |
| }, |
| { |
| "epoch": 0.5604143307170698, |
| "grad_norm": 4.157985210418701, |
| "learning_rate": 0.0001724384165701674, |
| "logits/chosen": -0.9274346828460693, |
| "logits/rejected": -0.9325964450836182, |
| "logps/chosen": -7.024556636810303, |
| "logps/rejected": -62.318359375, |
| "loss": 0.35737261176109314, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.25133123993873596, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.11644118279218674, |
| "rewards/margins": 4.924075126647949, |
| "rewards/rejected": -4.807634353637695, |
| "step": 328, |
| "train_speed(iter/s)": 0.005447 |
| }, |
| { |
| "epoch": 0.5621229109936462, |
| "grad_norm": 4.035698413848877, |
| "learning_rate": 0.00017224318545028649, |
| "logits/chosen": -0.9244407415390015, |
| "logits/rejected": -0.9326272010803223, |
| "logps/chosen": -6.016265392303467, |
| "logps/rejected": -56.14576721191406, |
| "loss": 0.3880583643913269, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3103344738483429, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.07886901497840881, |
| "rewards/margins": 4.746619701385498, |
| "rewards/rejected": -4.667750358581543, |
| "step": 329, |
| "train_speed(iter/s)": 0.005448 |
| }, |
| { |
| "epoch": 0.5638314912702227, |
| "grad_norm": 9.743223190307617, |
| "learning_rate": 0.0001720473766758198, |
| "logits/chosen": -0.926537275314331, |
| "logits/rejected": -0.9304601550102234, |
| "logps/chosen": -13.121172904968262, |
| "logps/rejected": -58.38237380981445, |
| "loss": 0.7646811008453369, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5820285677909851, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.09494726359844208, |
| "rewards/margins": 4.842239856719971, |
| "rewards/rejected": -4.9371867179870605, |
| "step": 330, |
| "train_speed(iter/s)": 0.005448 |
| }, |
| { |
| "epoch": 0.5655400715467991, |
| "grad_norm": 18.880859375, |
| "learning_rate": 0.00017185099181244907, |
| "logits/chosen": -0.9021615386009216, |
| "logits/rejected": -0.9024704694747925, |
| "logps/chosen": -9.637939453125, |
| "logps/rejected": -58.98625183105469, |
| "loss": 0.739081859588623, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6550078392028809, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": -0.00911126658320427, |
| "rewards/margins": 5.053913116455078, |
| "rewards/rejected": -5.063024997711182, |
| "step": 331, |
| "train_speed(iter/s)": 0.005448 |
| }, |
| { |
| "epoch": 0.5672486518233755, |
| "grad_norm": 2.728888750076294, |
| "learning_rate": 0.00017165403243046248, |
| "logits/chosen": -0.9438092112541199, |
| "logits/rejected": -0.9487511515617371, |
| "logps/chosen": -10.97130012512207, |
| "logps/rejected": -60.31357955932617, |
| "loss": 0.48772644996643066, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4556671679019928, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.31157898902893066, |
| "rewards/margins": 5.334751129150391, |
| "rewards/rejected": -5.023171901702881, |
| "step": 332, |
| "train_speed(iter/s)": 0.005448 |
| }, |
| { |
| "epoch": 0.568957232099952, |
| "grad_norm": 17.81595802307129, |
| "learning_rate": 0.00017145650010474194, |
| "logits/chosen": -0.9099326133728027, |
| "logits/rejected": -0.913872241973877, |
| "logps/chosen": -7.330003261566162, |
| "logps/rejected": -56.975791931152344, |
| "loss": 0.44156697392463684, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3481760025024414, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.13420407474040985, |
| "rewards/margins": 4.564350128173828, |
| "rewards/rejected": -4.430145740509033, |
| "step": 333, |
| "train_speed(iter/s)": 0.005449 |
| }, |
| { |
| "epoch": 0.5706658123765284, |
| "grad_norm": 4.562387466430664, |
| "learning_rate": 0.00017125839641475072, |
| "logits/chosen": -0.93560391664505, |
| "logits/rejected": -0.9402634501457214, |
| "logps/chosen": -8.35469913482666, |
| "logps/rejected": -48.29425048828125, |
| "loss": 0.8423429727554321, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5765202045440674, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.20604895055294037, |
| "rewards/margins": 4.131405353546143, |
| "rewards/rejected": -3.925356149673462, |
| "step": 334, |
| "train_speed(iter/s)": 0.005448 |
| }, |
| { |
| "epoch": 0.5723743926531049, |
| "grad_norm": 6.9514899253845215, |
| "learning_rate": 0.00017105972294452056, |
| "logits/chosen": -0.9457940459251404, |
| "logits/rejected": -0.9509940147399902, |
| "logps/chosen": -8.684478759765625, |
| "logps/rejected": -42.20777893066406, |
| "loss": 0.6009758114814758, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.2808820307254791, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.007849723100662231, |
| "rewards/margins": 3.273137092590332, |
| "rewards/rejected": -3.265286922454834, |
| "step": 335, |
| "train_speed(iter/s)": 0.005449 |
| }, |
| { |
| "epoch": 0.5740829729296812, |
| "grad_norm": 41.98228454589844, |
| "learning_rate": 0.0001708604812826393, |
| "logits/chosen": -0.9295578598976135, |
| "logits/rejected": -0.935192346572876, |
| "logps/chosen": -9.62477970123291, |
| "logps/rejected": -54.46697235107422, |
| "loss": 0.598048746585846, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.43297702074050903, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.07474887371063232, |
| "rewards/margins": 4.1404008865356445, |
| "rewards/rejected": -4.065651893615723, |
| "step": 336, |
| "train_speed(iter/s)": 0.005449 |
| }, |
| { |
| "epoch": 0.5757915532062576, |
| "grad_norm": 2.598214626312256, |
| "learning_rate": 0.00017066067302223798, |
| "logits/chosen": -0.8595199584960938, |
| "logits/rejected": -0.8645521402359009, |
| "logps/chosen": -5.353518962860107, |
| "logps/rejected": -45.4082145690918, |
| "loss": 0.3949076235294342, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3393931984901428, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.17312097549438477, |
| "rewards/margins": 3.7592127323150635, |
| "rewards/rejected": -3.5860915184020996, |
| "step": 337, |
| "train_speed(iter/s)": 0.005448 |
| }, |
| { |
| "epoch": 0.5775001334828341, |
| "grad_norm": 6.433867454528809, |
| "learning_rate": 0.00017046029976097805, |
| "logits/chosen": -0.8666653633117676, |
| "logits/rejected": -0.8727781176567078, |
| "logps/chosen": -7.716047286987305, |
| "logps/rejected": -44.40462112426758, |
| "loss": 0.7068074345588684, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.4196939468383789, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.018908580765128136, |
| "rewards/margins": 3.310499429702759, |
| "rewards/rejected": -3.291590929031372, |
| "step": 338, |
| "train_speed(iter/s)": 0.005448 |
| }, |
| { |
| "epoch": 0.5792087137594105, |
| "grad_norm": 4.851775169372559, |
| "learning_rate": 0.00017025936310103882, |
| "logits/chosen": -0.8393141627311707, |
| "logits/rejected": -0.8419257402420044, |
| "logps/chosen": -10.38674259185791, |
| "logps/rejected": -41.55541229248047, |
| "loss": 0.5500025749206543, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.35587984323501587, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.21891042590141296, |
| "rewards/margins": 3.442535400390625, |
| "rewards/rejected": -3.2236249446868896, |
| "step": 339, |
| "train_speed(iter/s)": 0.005448 |
| }, |
| { |
| "epoch": 0.580917294035987, |
| "grad_norm": 5.706538200378418, |
| "learning_rate": 0.0001700578646491045, |
| "logits/chosen": -0.849159300327301, |
| "logits/rejected": -0.8539773225784302, |
| "logps/chosen": -5.776857852935791, |
| "logps/rejected": -41.82091522216797, |
| "loss": 0.4652997851371765, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.1927475780248642, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.1367129385471344, |
| "rewards/margins": 2.951298952102661, |
| "rewards/rejected": -2.8145859241485596, |
| "step": 340, |
| "train_speed(iter/s)": 0.005448 |
| }, |
| { |
| "epoch": 0.5826258743125634, |
| "grad_norm": 15.33443832397461, |
| "learning_rate": 0.00016985580601635124, |
| "logits/chosen": -0.8022753000259399, |
| "logits/rejected": -0.809974193572998, |
| "logps/chosen": -10.902517318725586, |
| "logps/rejected": -38.2662467956543, |
| "loss": 0.7963184118270874, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5153663158416748, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.12418508529663086, |
| "rewards/margins": 2.705582618713379, |
| "rewards/rejected": -2.58139705657959, |
| "step": 341, |
| "train_speed(iter/s)": 0.005449 |
| }, |
| { |
| "epoch": 0.5843344545891398, |
| "grad_norm": 2.2032649517059326, |
| "learning_rate": 0.00016965318881843457, |
| "logits/chosen": -0.7673312425613403, |
| "logits/rejected": -0.7743449211120605, |
| "logps/chosen": -8.663345336914062, |
| "logps/rejected": -45.02417755126953, |
| "loss": 0.5525592565536499, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.42993825674057007, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.231624573469162, |
| "rewards/margins": 3.4039034843444824, |
| "rewards/rejected": -3.172279119491577, |
| "step": 342, |
| "train_speed(iter/s)": 0.00545 |
| }, |
| { |
| "epoch": 0.5860430348657163, |
| "grad_norm": 2.8073010444641113, |
| "learning_rate": 0.0001694500146754762, |
| "logits/chosen": -0.8122238516807556, |
| "logits/rejected": -0.8119744658470154, |
| "logps/chosen": -8.633512496948242, |
| "logps/rejected": -35.36651611328125, |
| "loss": 0.5415773391723633, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.31399714946746826, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.15843312442302704, |
| "rewards/margins": 2.734858989715576, |
| "rewards/rejected": -2.576425790786743, |
| "step": 343, |
| "train_speed(iter/s)": 0.005449 |
| }, |
| { |
| "epoch": 0.5877516151422927, |
| "grad_norm": 2.3773810863494873, |
| "learning_rate": 0.0001692462852120511, |
| "logits/chosen": -0.7741044759750366, |
| "logits/rejected": -0.7800744771957397, |
| "logps/chosen": -9.291975021362305, |
| "logps/rejected": -40.77639389038086, |
| "loss": 0.5208418965339661, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3624192774295807, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2715660035610199, |
| "rewards/margins": 2.76344633102417, |
| "rewards/rejected": -2.491880416870117, |
| "step": 344, |
| "train_speed(iter/s)": 0.005449 |
| }, |
| { |
| "epoch": 0.5894601954188692, |
| "grad_norm": 5.773314952850342, |
| "learning_rate": 0.0001690420020571747, |
| "logits/chosen": -0.7945303916931152, |
| "logits/rejected": -0.8053185343742371, |
| "logps/chosen": -7.3475542068481445, |
| "logps/rejected": -50.37881851196289, |
| "loss": 0.42357194423675537, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.32000547647476196, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3487846553325653, |
| "rewards/margins": 3.4395201206207275, |
| "rewards/rejected": -3.090735673904419, |
| "step": 345, |
| "train_speed(iter/s)": 0.005449 |
| }, |
| { |
| "epoch": 0.5911687756954456, |
| "grad_norm": 6.197271347045898, |
| "learning_rate": 0.0001688371668442896, |
| "logits/chosen": -0.805290937423706, |
| "logits/rejected": -0.8079847097396851, |
| "logps/chosen": -7.421670436859131, |
| "logps/rejected": -37.118797302246094, |
| "loss": 0.5294437408447266, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3508721590042114, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.17643693089485168, |
| "rewards/margins": 2.624368667602539, |
| "rewards/rejected": -2.4479315280914307, |
| "step": 346, |
| "train_speed(iter/s)": 0.00545 |
| }, |
| { |
| "epoch": 0.592877355972022, |
| "grad_norm": 3.835904836654663, |
| "learning_rate": 0.00016863178121125276, |
| "logits/chosen": -0.8487049341201782, |
| "logits/rejected": -0.8516339063644409, |
| "logps/chosen": -8.74366283416748, |
| "logps/rejected": -34.4682502746582, |
| "loss": 0.5114731788635254, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.3230685293674469, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.0005252091214060783, |
| "rewards/margins": 2.4223427772521973, |
| "rewards/rejected": -2.4218177795410156, |
| "step": 347, |
| "train_speed(iter/s)": 0.00545 |
| }, |
| { |
| "epoch": 0.5945859362485985, |
| "grad_norm": 14.421294212341309, |
| "learning_rate": 0.00016842584680032223, |
| "logits/chosen": -0.8961611986160278, |
| "logits/rejected": -0.9048024415969849, |
| "logps/chosen": -5.175715923309326, |
| "logps/rejected": -48.69798278808594, |
| "loss": 0.7653312683105469, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.5099776387214661, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.08455780148506165, |
| "rewards/margins": 3.585397720336914, |
| "rewards/rejected": -3.5008397102355957, |
| "step": 348, |
| "train_speed(iter/s)": 0.005449 |
| }, |
| { |
| "epoch": 0.5962945165251748, |
| "grad_norm": 3.0503435134887695, |
| "learning_rate": 0.000168219365258144, |
| "logits/chosen": -0.9129156470298767, |
| "logits/rejected": -0.9179389476776123, |
| "logps/chosen": -10.671258926391602, |
| "logps/rejected": -43.36078643798828, |
| "loss": 0.37005069851875305, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.25679123401641846, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.24289637804031372, |
| "rewards/margins": 3.3170793056488037, |
| "rewards/rejected": -3.0741827487945557, |
| "step": 349, |
| "train_speed(iter/s)": 0.005449 |
| }, |
| { |
| "epoch": 0.5980030968017513, |
| "grad_norm": 21.91331672668457, |
| "learning_rate": 0.00016801233823573908, |
| "logits/chosen": -0.8929505944252014, |
| "logits/rejected": -0.8917251229286194, |
| "logps/chosen": -14.047558784484863, |
| "logps/rejected": -31.77324676513672, |
| "loss": 0.976524829864502, |
| "memory(GiB)": 46.38, |
| "nll_loss": 0.6302264928817749, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.11739279329776764, |
| "rewards/margins": 2.250826120376587, |
| "rewards/rejected": -2.1334333419799805, |
| "step": 350, |
| "train_speed(iter/s)": 0.005449 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 1170, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "save_steps": 50, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 2.5475296526336e+18, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
