| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.25628704148646486, |
| "eval_steps": 300, |
| "global_step": 150, |
| "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 |
| } |
| ], |
| "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": 1.087227642815447e+18, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
