| { | |
| "best_metric": 0.03543492406606674, | |
| "best_model_checkpoint": "saves/psy-course/Llama-3.1-8B-Instruct/train/fold8/checkpoint-1200", | |
| "epoch": 4.996682778883518, | |
| "eval_steps": 50, | |
| "global_step": 3295, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.015164439389631315, | |
| "grad_norm": 4.908313751220703, | |
| "learning_rate": 3.0303030303030305e-06, | |
| "loss": 1.59, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03032887877926263, | |
| "grad_norm": 4.702552318572998, | |
| "learning_rate": 6.060606060606061e-06, | |
| "loss": 1.5951, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.045493318168893945, | |
| "grad_norm": 7.187352657318115, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 1.2503, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06065775755852526, | |
| "grad_norm": 2.202908515930176, | |
| "learning_rate": 1.2121212121212122e-05, | |
| "loss": 0.9145, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07582219694815658, | |
| "grad_norm": 1.6536071300506592, | |
| "learning_rate": 1.5151515151515153e-05, | |
| "loss": 0.6782, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07582219694815658, | |
| "eval_loss": 0.43230628967285156, | |
| "eval_runtime": 159.6252, | |
| "eval_samples_per_second": 7.348, | |
| "eval_steps_per_second": 7.348, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09098663633778789, | |
| "grad_norm": 1.2327362298965454, | |
| "learning_rate": 1.8181818181818182e-05, | |
| "loss": 0.3929, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1061510757274192, | |
| "grad_norm": 0.9732389450073242, | |
| "learning_rate": 2.1212121212121215e-05, | |
| "loss": 0.2232, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.12131551511705052, | |
| "grad_norm": 0.822551429271698, | |
| "learning_rate": 2.4242424242424244e-05, | |
| "loss": 0.2301, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.13647995450668182, | |
| "grad_norm": 0.8734151124954224, | |
| "learning_rate": 2.7272727272727273e-05, | |
| "loss": 0.127, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.15164439389631315, | |
| "grad_norm": 0.9926938414573669, | |
| "learning_rate": 3.0303030303030306e-05, | |
| "loss": 0.1528, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15164439389631315, | |
| "eval_loss": 0.0901423990726471, | |
| "eval_runtime": 159.655, | |
| "eval_samples_per_second": 7.347, | |
| "eval_steps_per_second": 7.347, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16680883328594445, | |
| "grad_norm": 0.5548447966575623, | |
| "learning_rate": 3.3333333333333335e-05, | |
| "loss": 0.0945, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.18197327267557578, | |
| "grad_norm": 1.0051425695419312, | |
| "learning_rate": 3.6363636363636364e-05, | |
| "loss": 0.0798, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.19713771206520708, | |
| "grad_norm": 1.2151364088058472, | |
| "learning_rate": 3.939393939393939e-05, | |
| "loss": 0.0769, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2123021514548384, | |
| "grad_norm": 0.7133122086524963, | |
| "learning_rate": 4.242424242424243e-05, | |
| "loss": 0.0684, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2274665908444697, | |
| "grad_norm": 1.0978891849517822, | |
| "learning_rate": 4.545454545454546e-05, | |
| "loss": 0.0741, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2274665908444697, | |
| "eval_loss": 0.06654910743236542, | |
| "eval_runtime": 159.7218, | |
| "eval_samples_per_second": 7.344, | |
| "eval_steps_per_second": 7.344, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.24263103023410104, | |
| "grad_norm": 0.9477673768997192, | |
| "learning_rate": 4.848484848484849e-05, | |
| "loss": 0.0795, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.25779546962373234, | |
| "grad_norm": 1.3763097524642944, | |
| "learning_rate": 5.151515151515152e-05, | |
| "loss": 0.0702, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.27295990901336364, | |
| "grad_norm": 0.5879313945770264, | |
| "learning_rate": 5.4545454545454546e-05, | |
| "loss": 0.0818, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.288124348402995, | |
| "grad_norm": 0.590936541557312, | |
| "learning_rate": 5.757575757575758e-05, | |
| "loss": 0.0682, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3032887877926263, | |
| "grad_norm": 1.0090534687042236, | |
| "learning_rate": 6.060606060606061e-05, | |
| "loss": 0.0544, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3032887877926263, | |
| "eval_loss": 0.05628184229135513, | |
| "eval_runtime": 159.7671, | |
| "eval_samples_per_second": 7.342, | |
| "eval_steps_per_second": 7.342, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3184532271822576, | |
| "grad_norm": 0.6095383167266846, | |
| "learning_rate": 6.363636363636364e-05, | |
| "loss": 0.0538, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3336176665718889, | |
| "grad_norm": 0.9077891707420349, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 0.0571, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.34878210596152026, | |
| "grad_norm": 0.4802258014678955, | |
| "learning_rate": 6.96969696969697e-05, | |
| "loss": 0.0628, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.36394654535115156, | |
| "grad_norm": 0.7471550107002258, | |
| "learning_rate": 7.272727272727273e-05, | |
| "loss": 0.0476, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.37911098474078286, | |
| "grad_norm": 0.8440163731575012, | |
| "learning_rate": 7.575757575757576e-05, | |
| "loss": 0.053, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.37911098474078286, | |
| "eval_loss": 0.0572020523250103, | |
| "eval_runtime": 159.824, | |
| "eval_samples_per_second": 7.339, | |
| "eval_steps_per_second": 7.339, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.39427542413041416, | |
| "grad_norm": 0.8062713146209717, | |
| "learning_rate": 7.878787878787879e-05, | |
| "loss": 0.0516, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.40943986352004547, | |
| "grad_norm": 0.33338773250579834, | |
| "learning_rate": 8.181818181818183e-05, | |
| "loss": 0.0712, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4246043029096768, | |
| "grad_norm": 0.550988495349884, | |
| "learning_rate": 8.484848484848486e-05, | |
| "loss": 0.0531, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4397687422993081, | |
| "grad_norm": 0.37053975462913513, | |
| "learning_rate": 8.787878787878789e-05, | |
| "loss": 0.0466, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4549331816889394, | |
| "grad_norm": 0.7273926734924316, | |
| "learning_rate": 9.090909090909092e-05, | |
| "loss": 0.0626, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4549331816889394, | |
| "eval_loss": 0.05890120193362236, | |
| "eval_runtime": 159.7681, | |
| "eval_samples_per_second": 7.342, | |
| "eval_steps_per_second": 7.342, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4700976210785707, | |
| "grad_norm": 0.6740069389343262, | |
| "learning_rate": 9.393939393939395e-05, | |
| "loss": 0.0501, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4852620604682021, | |
| "grad_norm": 0.5992850065231323, | |
| "learning_rate": 9.696969696969698e-05, | |
| "loss": 0.0443, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5004264998578334, | |
| "grad_norm": 0.6561434268951416, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0479, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5155909392474647, | |
| "grad_norm": 0.614966630935669, | |
| "learning_rate": 9.999719336268101e-05, | |
| "loss": 0.0507, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.530755378637096, | |
| "grad_norm": 0.4214603304862976, | |
| "learning_rate": 9.998877376581251e-05, | |
| "loss": 0.0646, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.530755378637096, | |
| "eval_loss": 0.04616093635559082, | |
| "eval_runtime": 159.8294, | |
| "eval_samples_per_second": 7.339, | |
| "eval_steps_per_second": 7.339, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5459198180267273, | |
| "grad_norm": 0.5347225069999695, | |
| "learning_rate": 9.997474215462472e-05, | |
| "loss": 0.0549, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5610842574163586, | |
| "grad_norm": 0.6282948851585388, | |
| "learning_rate": 9.995510010438337e-05, | |
| "loss": 0.0442, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.57624869680599, | |
| "grad_norm": 0.8127485513687134, | |
| "learning_rate": 9.992984982021295e-05, | |
| "loss": 0.0607, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5914131361956213, | |
| "grad_norm": 0.7069970965385437, | |
| "learning_rate": 9.9898994136849e-05, | |
| "loss": 0.0526, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6065775755852526, | |
| "grad_norm": 0.3750600516796112, | |
| "learning_rate": 9.986253651832005e-05, | |
| "loss": 0.0336, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6065775755852526, | |
| "eval_loss": 0.04486996307969093, | |
| "eval_runtime": 159.7479, | |
| "eval_samples_per_second": 7.343, | |
| "eval_steps_per_second": 7.343, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6217420149748839, | |
| "grad_norm": 0.2943829298019409, | |
| "learning_rate": 9.982048105755859e-05, | |
| "loss": 0.0409, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6369064543645152, | |
| "grad_norm": 0.3517925441265106, | |
| "learning_rate": 9.977283247594166e-05, | |
| "loss": 0.052, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6520708937541465, | |
| "grad_norm": 0.47345930337905884, | |
| "learning_rate": 9.971959612276076e-05, | |
| "loss": 0.047, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6672353331437778, | |
| "grad_norm": 0.5188047289848328, | |
| "learning_rate": 9.966077797462129e-05, | |
| "loss": 0.0456, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6823997725334091, | |
| "grad_norm": 0.3099527060985565, | |
| "learning_rate": 9.959638463477165e-05, | |
| "loss": 0.0426, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6823997725334091, | |
| "eval_loss": 0.043455954641103745, | |
| "eval_runtime": 159.7181, | |
| "eval_samples_per_second": 7.344, | |
| "eval_steps_per_second": 7.344, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6975642119230405, | |
| "grad_norm": 0.4291819632053375, | |
| "learning_rate": 9.952642333236186e-05, | |
| "loss": 0.0471, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7127286513126718, | |
| "grad_norm": 0.47742322087287903, | |
| "learning_rate": 9.945090192163202e-05, | |
| "loss": 0.0514, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7278930907023031, | |
| "grad_norm": 0.5357955098152161, | |
| "learning_rate": 9.936982888103051e-05, | |
| "loss": 0.0372, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7430575300919344, | |
| "grad_norm": 0.3578835725784302, | |
| "learning_rate": 9.928321331226219e-05, | |
| "loss": 0.036, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7582219694815657, | |
| "grad_norm": 1.1390984058380127, | |
| "learning_rate": 9.919106493926655e-05, | |
| "loss": 0.0354, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7582219694815657, | |
| "eval_loss": 0.04678919166326523, | |
| "eval_runtime": 159.8624, | |
| "eval_samples_per_second": 7.338, | |
| "eval_steps_per_second": 7.338, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.773386408871197, | |
| "grad_norm": 0.2849079966545105, | |
| "learning_rate": 9.909339410712612e-05, | |
| "loss": 0.0463, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7885508482608283, | |
| "grad_norm": 0.43958020210266113, | |
| "learning_rate": 9.8990211780905e-05, | |
| "loss": 0.0388, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8037152876504596, | |
| "grad_norm": 0.4281414747238159, | |
| "learning_rate": 9.888152954441785e-05, | |
| "loss": 0.0368, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8188797270400909, | |
| "grad_norm": 0.20032383501529694, | |
| "learning_rate": 9.876735959892953e-05, | |
| "loss": 0.039, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8340441664297223, | |
| "grad_norm": 0.697373628616333, | |
| "learning_rate": 9.864771476178522e-05, | |
| "loss": 0.0506, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8340441664297223, | |
| "eval_loss": 0.04438818246126175, | |
| "eval_runtime": 159.8887, | |
| "eval_samples_per_second": 7.336, | |
| "eval_steps_per_second": 7.336, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8492086058193536, | |
| "grad_norm": 0.4126269817352295, | |
| "learning_rate": 9.852260846497153e-05, | |
| "loss": 0.0515, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.864373045208985, | |
| "grad_norm": 0.31782063841819763, | |
| "learning_rate": 9.839205475360851e-05, | |
| "loss": 0.0537, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8795374845986162, | |
| "grad_norm": 0.28636643290519714, | |
| "learning_rate": 9.825606828437291e-05, | |
| "loss": 0.0497, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8947019239882475, | |
| "grad_norm": 0.3322339951992035, | |
| "learning_rate": 9.811466432385267e-05, | |
| "loss": 0.0392, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.9098663633778789, | |
| "grad_norm": 0.40966206789016724, | |
| "learning_rate": 9.796785874683314e-05, | |
| "loss": 0.0346, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9098663633778789, | |
| "eval_loss": 0.04112881049513817, | |
| "eval_runtime": 159.8597, | |
| "eval_samples_per_second": 7.338, | |
| "eval_steps_per_second": 7.338, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9250308027675102, | |
| "grad_norm": 0.34210318326950073, | |
| "learning_rate": 9.781566803451475e-05, | |
| "loss": 0.05, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9401952421571415, | |
| "grad_norm": 0.45713552832603455, | |
| "learning_rate": 9.765810927266281e-05, | |
| "loss": 0.0414, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9553596815467729, | |
| "grad_norm": 0.320637047290802, | |
| "learning_rate": 9.749520014968934e-05, | |
| "loss": 0.0389, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9705241209364042, | |
| "grad_norm": 0.3496937155723572, | |
| "learning_rate": 9.732695895466735e-05, | |
| "loss": 0.0449, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.9856885603260355, | |
| "grad_norm": 0.3174608647823334, | |
| "learning_rate": 9.715340457527746e-05, | |
| "loss": 0.0514, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9856885603260355, | |
| "eval_loss": 0.04426902160048485, | |
| "eval_runtime": 159.8605, | |
| "eval_samples_per_second": 7.338, | |
| "eval_steps_per_second": 7.338, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.0008529997156668, | |
| "grad_norm": 0.26187363266944885, | |
| "learning_rate": 9.697455649568761e-05, | |
| "loss": 0.04, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.016017439105298, | |
| "grad_norm": 0.6161110997200012, | |
| "learning_rate": 9.679043479436556e-05, | |
| "loss": 0.0294, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0311818784949294, | |
| "grad_norm": 0.18970751762390137, | |
| "learning_rate": 9.660106014182489e-05, | |
| "loss": 0.0277, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.0463463178845607, | |
| "grad_norm": 0.5361956357955933, | |
| "learning_rate": 9.640645379830424e-05, | |
| "loss": 0.0444, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.061510757274192, | |
| "grad_norm": 0.288272500038147, | |
| "learning_rate": 9.620663761138067e-05, | |
| "loss": 0.0323, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.061510757274192, | |
| "eval_loss": 0.041203178465366364, | |
| "eval_runtime": 159.9229, | |
| "eval_samples_per_second": 7.335, | |
| "eval_steps_per_second": 7.335, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0766751966638233, | |
| "grad_norm": 0.11579044908285141, | |
| "learning_rate": 9.600163401351688e-05, | |
| "loss": 0.0261, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.0918396360534546, | |
| "grad_norm": 0.3901958167552948, | |
| "learning_rate": 9.579146601954276e-05, | |
| "loss": 0.0453, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.1070040754430859, | |
| "grad_norm": 0.15443949401378632, | |
| "learning_rate": 9.557615722407177e-05, | |
| "loss": 0.0364, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.1221685148327172, | |
| "grad_norm": 0.1345423012971878, | |
| "learning_rate": 9.535573179885191e-05, | |
| "loss": 0.0279, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.1373329542223485, | |
| "grad_norm": 0.17373351752758026, | |
| "learning_rate": 9.513021449005214e-05, | |
| "loss": 0.0261, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.1373329542223485, | |
| "eval_loss": 0.04307902976870537, | |
| "eval_runtime": 159.8102, | |
| "eval_samples_per_second": 7.34, | |
| "eval_steps_per_second": 7.34, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.15249739361198, | |
| "grad_norm": 0.1945541799068451, | |
| "learning_rate": 9.489963061548428e-05, | |
| "loss": 0.0326, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.1676618330016113, | |
| "grad_norm": 0.14704962074756622, | |
| "learning_rate": 9.466400606176062e-05, | |
| "loss": 0.0232, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.1828262723912426, | |
| "grad_norm": 0.2738613486289978, | |
| "learning_rate": 9.442336728138779e-05, | |
| "loss": 0.0235, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.197990711780874, | |
| "grad_norm": 0.14102770388126373, | |
| "learning_rate": 9.417774128979706e-05, | |
| "loss": 0.0364, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.2131551511705052, | |
| "grad_norm": 0.3271743953227997, | |
| "learning_rate": 9.39271556623114e-05, | |
| "loss": 0.0307, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2131551511705052, | |
| "eval_loss": 0.03978438675403595, | |
| "eval_runtime": 159.8781, | |
| "eval_samples_per_second": 7.337, | |
| "eval_steps_per_second": 7.337, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2283195905601365, | |
| "grad_norm": 0.4989463686943054, | |
| "learning_rate": 9.367163853104975e-05, | |
| "loss": 0.039, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.2434840299497678, | |
| "grad_norm": 0.19276122748851776, | |
| "learning_rate": 9.341121858176876e-05, | |
| "loss": 0.0413, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.2586484693393991, | |
| "grad_norm": 0.29775118827819824, | |
| "learning_rate": 9.314592505064239e-05, | |
| "loss": 0.036, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.2738129087290304, | |
| "grad_norm": 0.23416060209274292, | |
| "learning_rate": 9.28757877209796e-05, | |
| "loss": 0.0258, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.2889773481186617, | |
| "grad_norm": 0.14653538167476654, | |
| "learning_rate": 9.260083691988084e-05, | |
| "loss": 0.0284, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.2889773481186617, | |
| "eval_loss": 0.04156646877527237, | |
| "eval_runtime": 159.8802, | |
| "eval_samples_per_second": 7.337, | |
| "eval_steps_per_second": 7.337, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.304141787508293, | |
| "grad_norm": 0.4026676416397095, | |
| "learning_rate": 9.232110351483327e-05, | |
| "loss": 0.0368, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.3193062268979243, | |
| "grad_norm": 0.4935552179813385, | |
| "learning_rate": 9.203661891024547e-05, | |
| "loss": 0.0369, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.3344706662875556, | |
| "grad_norm": 0.3299759328365326, | |
| "learning_rate": 9.174741504392173e-05, | |
| "loss": 0.0345, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.349635105677187, | |
| "grad_norm": 0.24818608164787292, | |
| "learning_rate": 9.145352438347662e-05, | |
| "loss": 0.0413, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.3647995450668184, | |
| "grad_norm": 0.2329672873020172, | |
| "learning_rate": 9.115497992268995e-05, | |
| "loss": 0.0316, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.3647995450668184, | |
| "eval_loss": 0.03982315585017204, | |
| "eval_runtime": 159.8004, | |
| "eval_samples_per_second": 7.34, | |
| "eval_steps_per_second": 7.34, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.3799639844564497, | |
| "grad_norm": 0.07506946474313736, | |
| "learning_rate": 9.085181517780273e-05, | |
| "loss": 0.0309, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.395128423846081, | |
| "grad_norm": 0.19676469266414642, | |
| "learning_rate": 9.054406418375443e-05, | |
| "loss": 0.0355, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.4102928632357123, | |
| "grad_norm": 0.2670162320137024, | |
| "learning_rate": 9.023176149036203e-05, | |
| "loss": 0.0299, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.4254573026253436, | |
| "grad_norm": 0.24655964970588684, | |
| "learning_rate": 8.991494215844132e-05, | |
| "loss": 0.0416, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.440621742014975, | |
| "grad_norm": 0.42746463418006897, | |
| "learning_rate": 8.959364175587069e-05, | |
| "loss": 0.0288, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.440621742014975, | |
| "eval_loss": 0.039692338556051254, | |
| "eval_runtime": 159.8479, | |
| "eval_samples_per_second": 7.338, | |
| "eval_steps_per_second": 7.338, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.4557861814046062, | |
| "grad_norm": 0.20224446058273315, | |
| "learning_rate": 8.926789635359817e-05, | |
| "loss": 0.0448, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.4709506207942376, | |
| "grad_norm": 0.16787214577198029, | |
| "learning_rate": 8.893774252159187e-05, | |
| "loss": 0.0294, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.4861150601838689, | |
| "grad_norm": 0.20546148717403412, | |
| "learning_rate": 8.860321732473439e-05, | |
| "loss": 0.0396, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.5012794995735002, | |
| "grad_norm": 0.4057621657848358, | |
| "learning_rate": 8.826435831866184e-05, | |
| "loss": 0.0389, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.5164439389631315, | |
| "grad_norm": 0.09442304074764252, | |
| "learning_rate": 8.79212035455475e-05, | |
| "loss": 0.0306, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5164439389631315, | |
| "eval_loss": 0.041055887937545776, | |
| "eval_runtime": 159.778, | |
| "eval_samples_per_second": 7.341, | |
| "eval_steps_per_second": 7.341, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5316083783527628, | |
| "grad_norm": 0.2399837076663971, | |
| "learning_rate": 8.757379152983103e-05, | |
| "loss": 0.03, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.546772817742394, | |
| "grad_norm": 0.38911059498786926, | |
| "learning_rate": 8.722216127389363e-05, | |
| "loss": 0.0335, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.5619372571320254, | |
| "grad_norm": 0.6509933471679688, | |
| "learning_rate": 8.686635225367919e-05, | |
| "loss": 0.0361, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.5771016965216567, | |
| "grad_norm": 0.20148853957653046, | |
| "learning_rate": 8.650640441426274e-05, | |
| "loss": 0.038, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.592266135911288, | |
| "grad_norm": 0.3823687434196472, | |
| "learning_rate": 8.614235816536582e-05, | |
| "loss": 0.0449, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.592266135911288, | |
| "eval_loss": 0.03836758807301521, | |
| "eval_runtime": 159.8354, | |
| "eval_samples_per_second": 7.339, | |
| "eval_steps_per_second": 7.339, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.6074305753009193, | |
| "grad_norm": 0.3107754588127136, | |
| "learning_rate": 8.577425437681994e-05, | |
| "loss": 0.0377, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.6225950146905506, | |
| "grad_norm": 0.17179957032203674, | |
| "learning_rate": 8.540213437397833e-05, | |
| "loss": 0.0322, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.6377594540801819, | |
| "grad_norm": 0.10634970664978027, | |
| "learning_rate": 8.502603993307647e-05, | |
| "loss": 0.0273, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.6529238934698132, | |
| "grad_norm": 0.32147282361984253, | |
| "learning_rate": 8.464601327654207e-05, | |
| "loss": 0.0251, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.6680883328594445, | |
| "grad_norm": 0.29508742690086365, | |
| "learning_rate": 8.4262097068255e-05, | |
| "loss": 0.0281, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6680883328594445, | |
| "eval_loss": 0.038478951901197433, | |
| "eval_runtime": 159.8811, | |
| "eval_samples_per_second": 7.337, | |
| "eval_steps_per_second": 7.337, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6832527722490758, | |
| "grad_norm": 0.5426458716392517, | |
| "learning_rate": 8.387433440875758e-05, | |
| "loss": 0.0331, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.698417211638707, | |
| "grad_norm": 0.2759105861186981, | |
| "learning_rate": 8.348276883041583e-05, | |
| "loss": 0.0336, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.7135816510283386, | |
| "grad_norm": 0.4834611415863037, | |
| "learning_rate": 8.308744429253238e-05, | |
| "loss": 0.0464, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.72874609041797, | |
| "grad_norm": 0.3538722097873688, | |
| "learning_rate": 8.268840517641123e-05, | |
| "loss": 0.0345, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.7439105298076012, | |
| "grad_norm": 0.41168689727783203, | |
| "learning_rate": 8.228569628037543e-05, | |
| "loss": 0.038, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7439105298076012, | |
| "eval_loss": 0.03758051246404648, | |
| "eval_runtime": 159.8738, | |
| "eval_samples_per_second": 7.337, | |
| "eval_steps_per_second": 7.337, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.7590749691972325, | |
| "grad_norm": 0.3065544068813324, | |
| "learning_rate": 8.187936281473758e-05, | |
| "loss": 0.0325, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.7742394085868638, | |
| "grad_norm": 0.24398308992385864, | |
| "learning_rate": 8.146945039672446e-05, | |
| "loss": 0.0276, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.789403847976495, | |
| "grad_norm": 0.29391804337501526, | |
| "learning_rate": 8.105600504535567e-05, | |
| "loss": 0.0306, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.8045682873661264, | |
| "grad_norm": 0.33613964915275574, | |
| "learning_rate": 8.06390731762773e-05, | |
| "loss": 0.0416, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.8197327267557577, | |
| "grad_norm": 0.6730492115020752, | |
| "learning_rate": 8.021870159655109e-05, | |
| "loss": 0.0277, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.8197327267557577, | |
| "eval_loss": 0.03543492406606674, | |
| "eval_runtime": 159.8678, | |
| "eval_samples_per_second": 7.337, | |
| "eval_steps_per_second": 7.337, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.8348971661453892, | |
| "grad_norm": 0.31159326434135437, | |
| "learning_rate": 7.979493749939955e-05, | |
| "loss": 0.0335, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.8500616055350205, | |
| "grad_norm": 0.2657136917114258, | |
| "learning_rate": 7.93678284589079e-05, | |
| "loss": 0.0421, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.8652260449246518, | |
| "grad_norm": 0.1698162853717804, | |
| "learning_rate": 7.893742242468301e-05, | |
| "loss": 0.0282, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.8803904843142831, | |
| "grad_norm": 0.5252974629402161, | |
| "learning_rate": 7.850376771647038e-05, | |
| "loss": 0.0386, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.8955549237039144, | |
| "grad_norm": 0.31378066539764404, | |
| "learning_rate": 7.806691301872958e-05, | |
| "loss": 0.0381, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.8955549237039144, | |
| "eval_loss": 0.037748388946056366, | |
| "eval_runtime": 159.8804, | |
| "eval_samples_per_second": 7.337, | |
| "eval_steps_per_second": 7.337, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.9107193630935457, | |
| "grad_norm": 0.16329045593738556, | |
| "learning_rate": 7.762690737516846e-05, | |
| "loss": 0.0359, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.925883802483177, | |
| "grad_norm": 0.23127301037311554, | |
| "learning_rate": 7.718380018323742e-05, | |
| "loss": 0.0326, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.9410482418728083, | |
| "grad_norm": 0.29971396923065186, | |
| "learning_rate": 7.673764118858371e-05, | |
| "loss": 0.0279, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.9562126812624396, | |
| "grad_norm": 0.14393766224384308, | |
| "learning_rate": 7.628848047946675e-05, | |
| "loss": 0.0408, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.971377120652071, | |
| "grad_norm": 0.24297180771827698, | |
| "learning_rate": 7.583636848113483e-05, | |
| "loss": 0.0278, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.971377120652071, | |
| "eval_loss": 0.03594334051012993, | |
| "eval_runtime": 159.9049, | |
| "eval_samples_per_second": 7.336, | |
| "eval_steps_per_second": 7.336, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.9865415600417022, | |
| "grad_norm": 0.33698412775993347, | |
| "learning_rate": 7.538135595016423e-05, | |
| "loss": 0.0322, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.0017059994313335, | |
| "grad_norm": 0.28321677446365356, | |
| "learning_rate": 7.492349396876096e-05, | |
| "loss": 0.0292, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.016870438820965, | |
| "grad_norm": 0.11086804419755936, | |
| "learning_rate": 7.4462833939026e-05, | |
| "loss": 0.0237, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.032034878210596, | |
| "grad_norm": 0.3019621670246124, | |
| "learning_rate": 7.399942757718455e-05, | |
| "loss": 0.0219, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.0471993176002274, | |
| "grad_norm": 0.26927098631858826, | |
| "learning_rate": 7.35333269077802e-05, | |
| "loss": 0.0194, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0471993176002274, | |
| "eval_loss": 0.03932846337556839, | |
| "eval_runtime": 159.9101, | |
| "eval_samples_per_second": 7.335, | |
| "eval_steps_per_second": 7.335, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.0623637569898587, | |
| "grad_norm": 0.1983146220445633, | |
| "learning_rate": 7.306458425783426e-05, | |
| "loss": 0.0287, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.07752819637949, | |
| "grad_norm": 0.2132590264081955, | |
| "learning_rate": 7.25932522509713e-05, | |
| "loss": 0.02, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.0926926357691213, | |
| "grad_norm": 0.4830004572868347, | |
| "learning_rate": 7.211938380151133e-05, | |
| "loss": 0.0244, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.1078570751587526, | |
| "grad_norm": 0.3054353594779968, | |
| "learning_rate": 7.164303210852934e-05, | |
| "loss": 0.0309, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.123021514548384, | |
| "grad_norm": 0.22349397838115692, | |
| "learning_rate": 7.116425064988286e-05, | |
| "loss": 0.0149, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.123021514548384, | |
| "eval_loss": 0.03746328130364418, | |
| "eval_runtime": 159.9069, | |
| "eval_samples_per_second": 7.336, | |
| "eval_steps_per_second": 7.336, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1381859539380152, | |
| "grad_norm": 0.23747169971466064, | |
| "learning_rate": 7.068309317620827e-05, | |
| "loss": 0.0244, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.1533503933276466, | |
| "grad_norm": 0.2875257730484009, | |
| "learning_rate": 7.019961370488645e-05, | |
| "loss": 0.0211, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.168514832717278, | |
| "grad_norm": 0.24418668448925018, | |
| "learning_rate": 6.971386651397849e-05, | |
| "loss": 0.02, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.183679272106909, | |
| "grad_norm": 0.5265827178955078, | |
| "learning_rate": 6.922590613613211e-05, | |
| "loss": 0.0169, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.1988437114965405, | |
| "grad_norm": 0.2406657487154007, | |
| "learning_rate": 6.873578735245961e-05, | |
| "loss": 0.0243, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.1988437114965405, | |
| "eval_loss": 0.04021035507321358, | |
| "eval_runtime": 159.985, | |
| "eval_samples_per_second": 7.332, | |
| "eval_steps_per_second": 7.332, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.2140081508861718, | |
| "grad_norm": 0.14891931414604187, | |
| "learning_rate": 6.824356518638775e-05, | |
| "loss": 0.021, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.229172590275803, | |
| "grad_norm": 0.32252541184425354, | |
| "learning_rate": 6.774929489748052e-05, | |
| "loss": 0.0198, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.2443370296654344, | |
| "grad_norm": 1.1908115148544312, | |
| "learning_rate": 6.725303197523548e-05, | |
| "loss": 0.0302, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.259501469055066, | |
| "grad_norm": 0.19598431885242462, | |
| "learning_rate": 6.675483213285412e-05, | |
| "loss": 0.0225, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.274665908444697, | |
| "grad_norm": 0.37645435333251953, | |
| "learning_rate": 6.625475130098728e-05, | |
| "loss": 0.0234, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.274665908444697, | |
| "eval_loss": 0.039622507989406586, | |
| "eval_runtime": 160.0131, | |
| "eval_samples_per_second": 7.331, | |
| "eval_steps_per_second": 7.331, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.2898303478343287, | |
| "grad_norm": 0.2178558111190796, | |
| "learning_rate": 6.575284562145593e-05, | |
| "loss": 0.0213, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.30499478722396, | |
| "grad_norm": 0.4462236762046814, | |
| "learning_rate": 6.524917144094851e-05, | |
| "loss": 0.0258, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.3201592266135913, | |
| "grad_norm": 0.0876719132065773, | |
| "learning_rate": 6.474378530469509e-05, | |
| "loss": 0.025, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.3353236660032226, | |
| "grad_norm": 0.16359026730060577, | |
| "learning_rate": 6.42367439501193e-05, | |
| "loss": 0.0205, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.350488105392854, | |
| "grad_norm": 0.25489160418510437, | |
| "learning_rate": 6.372810430046862e-05, | |
| "loss": 0.0189, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.350488105392854, | |
| "eval_loss": 0.03922748565673828, | |
| "eval_runtime": 159.9514, | |
| "eval_samples_per_second": 7.333, | |
| "eval_steps_per_second": 7.333, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.365652544782485, | |
| "grad_norm": 0.11957678943872452, | |
| "learning_rate": 6.321792345842402e-05, | |
| "loss": 0.0228, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.3808169841721165, | |
| "grad_norm": 0.23994044959545135, | |
| "learning_rate": 6.270625869968906e-05, | |
| "loss": 0.0323, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.395981423561748, | |
| "grad_norm": 0.1954767256975174, | |
| "learning_rate": 6.219316746656007e-05, | |
| "loss": 0.0284, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.411145862951379, | |
| "grad_norm": 0.25645068287849426, | |
| "learning_rate": 6.167870736147713e-05, | |
| "loss": 0.0234, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.4263103023410104, | |
| "grad_norm": 0.7860223054885864, | |
| "learning_rate": 6.116293614055744e-05, | |
| "loss": 0.0195, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4263103023410104, | |
| "eval_loss": 0.03961315006017685, | |
| "eval_runtime": 159.8725, | |
| "eval_samples_per_second": 7.337, | |
| "eval_steps_per_second": 7.337, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4414747417306417, | |
| "grad_norm": 0.2385544329881668, | |
| "learning_rate": 6.06459117071113e-05, | |
| "loss": 0.0186, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.456639181120273, | |
| "grad_norm": 0.5610677599906921, | |
| "learning_rate": 6.012769210514146e-05, | |
| "loss": 0.0213, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.4718036205099043, | |
| "grad_norm": 0.1168915182352066, | |
| "learning_rate": 5.9608335512826915e-05, | |
| "loss": 0.0207, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.4869680598995356, | |
| "grad_norm": 0.08927187323570251, | |
| "learning_rate": 5.908790023599144e-05, | |
| "loss": 0.0143, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.502132499289167, | |
| "grad_norm": 0.2827674448490143, | |
| "learning_rate": 5.856644470155781e-05, | |
| "loss": 0.0221, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.502132499289167, | |
| "eval_loss": 0.04753277450799942, | |
| "eval_runtime": 159.8134, | |
| "eval_samples_per_second": 7.34, | |
| "eval_steps_per_second": 7.34, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.5172969386787982, | |
| "grad_norm": 0.2750401496887207, | |
| "learning_rate": 5.8044027450988546e-05, | |
| "loss": 0.028, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.5324613780684295, | |
| "grad_norm": 0.19203156232833862, | |
| "learning_rate": 5.752070713371371e-05, | |
| "loss": 0.0161, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.547625817458061, | |
| "grad_norm": 0.1918914020061493, | |
| "learning_rate": 5.699654250054662e-05, | |
| "loss": 0.0165, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.562790256847692, | |
| "grad_norm": 0.1869034469127655, | |
| "learning_rate": 5.647159239708809e-05, | |
| "loss": 0.014, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.5779546962373234, | |
| "grad_norm": 0.18233343958854675, | |
| "learning_rate": 5.5945915757120146e-05, | |
| "loss": 0.0179, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.5779546962373234, | |
| "eval_loss": 0.04084445536136627, | |
| "eval_runtime": 159.9315, | |
| "eval_samples_per_second": 7.334, | |
| "eval_steps_per_second": 7.334, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.5931191356269547, | |
| "grad_norm": 0.4812876284122467, | |
| "learning_rate": 5.5419571595989825e-05, | |
| "loss": 0.0272, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.608283575016586, | |
| "grad_norm": 0.3959842920303345, | |
| "learning_rate": 5.4892619003983734e-05, | |
| "loss": 0.0248, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.6234480144062173, | |
| "grad_norm": 0.14155489206314087, | |
| "learning_rate": 5.436511713969428e-05, | |
| "loss": 0.0157, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.6386124537958486, | |
| "grad_norm": 0.07145866006612778, | |
| "learning_rate": 5.383712522337817e-05, | |
| "loss": 0.024, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.65377689318548, | |
| "grad_norm": 0.26969072222709656, | |
| "learning_rate": 5.3308702530308076e-05, | |
| "loss": 0.0225, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.65377689318548, | |
| "eval_loss": 0.038576629012823105, | |
| "eval_runtime": 159.8548, | |
| "eval_samples_per_second": 7.338, | |
| "eval_steps_per_second": 7.338, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.6689413325751112, | |
| "grad_norm": 0.3706766366958618, | |
| "learning_rate": 5.2779908384118025e-05, | |
| "loss": 0.0258, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.6841057719647425, | |
| "grad_norm": 0.36738407611846924, | |
| "learning_rate": 5.22508021501434e-05, | |
| "loss": 0.0215, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.699270211354374, | |
| "grad_norm": 0.34079742431640625, | |
| "learning_rate": 5.1721443228756284e-05, | |
| "loss": 0.0163, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.714434650744005, | |
| "grad_norm": 0.33863401412963867, | |
| "learning_rate": 5.119189104869683e-05, | |
| "loss": 0.0219, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.729599090133637, | |
| "grad_norm": 0.2650861442089081, | |
| "learning_rate": 5.066220506040148e-05, | |
| "loss": 0.0207, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.729599090133637, | |
| "eval_loss": 0.03975927457213402, | |
| "eval_runtime": 159.781, | |
| "eval_samples_per_second": 7.341, | |
| "eval_steps_per_second": 7.341, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.7447635295232677, | |
| "grad_norm": 0.20335064828395844, | |
| "learning_rate": 5.013244472932872e-05, | |
| "loss": 0.0248, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.7599279689128995, | |
| "grad_norm": 0.29428455233573914, | |
| "learning_rate": 4.960266952928316e-05, | |
| "loss": 0.0215, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.7750924083025303, | |
| "grad_norm": 0.3635200262069702, | |
| "learning_rate": 4.907293893573867e-05, | |
| "loss": 0.0225, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.790256847692162, | |
| "grad_norm": 0.32583731412887573, | |
| "learning_rate": 4.8543312419161396e-05, | |
| "loss": 0.0197, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.805421287081793, | |
| "grad_norm": 0.20632241666316986, | |
| "learning_rate": 4.8013849438333165e-05, | |
| "loss": 0.015, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.805421287081793, | |
| "eval_loss": 0.041027262806892395, | |
| "eval_runtime": 159.7438, | |
| "eval_samples_per_second": 7.343, | |
| "eval_steps_per_second": 7.343, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.8205857264714247, | |
| "grad_norm": 0.8117703795433044, | |
| "learning_rate": 4.748460943367643e-05, | |
| "loss": 0.0212, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.8357501658610555, | |
| "grad_norm": 0.374080628156662, | |
| "learning_rate": 4.695565182058113e-05, | |
| "loss": 0.0312, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 2.8509146052506873, | |
| "grad_norm": 0.36783552169799805, | |
| "learning_rate": 4.642703598273431e-05, | |
| "loss": 0.0249, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 2.8660790446403186, | |
| "grad_norm": 0.2705755829811096, | |
| "learning_rate": 4.589882126545352e-05, | |
| "loss": 0.0234, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 2.88124348402995, | |
| "grad_norm": 0.25217315554618835, | |
| "learning_rate": 4.537106696902425e-05, | |
| "loss": 0.0295, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.88124348402995, | |
| "eval_loss": 0.03719955310225487, | |
| "eval_runtime": 159.7214, | |
| "eval_samples_per_second": 7.344, | |
| "eval_steps_per_second": 7.344, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.896407923419581, | |
| "grad_norm": 0.40652021765708923, | |
| "learning_rate": 4.484383234204266e-05, | |
| "loss": 0.0224, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 2.9115723628092125, | |
| "grad_norm": 0.3361893892288208, | |
| "learning_rate": 4.4317176574763935e-05, | |
| "loss": 0.0247, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 2.926736802198844, | |
| "grad_norm": 0.35932040214538574, | |
| "learning_rate": 4.3791158792457334e-05, | |
| "loss": 0.024, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.941901241588475, | |
| "grad_norm": 0.15191532671451569, | |
| "learning_rate": 4.3265838048768334e-05, | |
| "loss": 0.0216, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 2.9570656809781064, | |
| "grad_norm": 0.5233368873596191, | |
| "learning_rate": 4.274127331908915e-05, | |
| "loss": 0.0227, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.9570656809781064, | |
| "eval_loss": 0.03795257583260536, | |
| "eval_runtime": 159.8645, | |
| "eval_samples_per_second": 7.337, | |
| "eval_steps_per_second": 7.337, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.9722301203677377, | |
| "grad_norm": 0.05902265012264252, | |
| "learning_rate": 4.221752349393758e-05, | |
| "loss": 0.0173, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 2.987394559757369, | |
| "grad_norm": 0.1508518010377884, | |
| "learning_rate": 4.1694647372345886e-05, | |
| "loss": 0.0249, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.0025589991470003, | |
| "grad_norm": 0.13050998747348785, | |
| "learning_rate": 4.117270365525946e-05, | |
| "loss": 0.0182, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.0177234385366316, | |
| "grad_norm": 0.06321557611227036, | |
| "learning_rate": 4.065175093894694e-05, | |
| "loss": 0.0132, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.032887877926263, | |
| "grad_norm": 0.155404731631279, | |
| "learning_rate": 4.013184770842167e-05, | |
| "loss": 0.0128, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.032887877926263, | |
| "eval_loss": 0.03980863094329834, | |
| "eval_runtime": 159.767, | |
| "eval_samples_per_second": 7.342, | |
| "eval_steps_per_second": 7.342, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.048052317315894, | |
| "grad_norm": 0.18662160634994507, | |
| "learning_rate": 3.9613052330876064e-05, | |
| "loss": 0.0166, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.0632167567055255, | |
| "grad_norm": 0.5268431901931763, | |
| "learning_rate": 3.909542304912881e-05, | |
| "loss": 0.0136, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.078381196095157, | |
| "grad_norm": 0.18333646655082703, | |
| "learning_rate": 3.857901797508628e-05, | |
| "loss": 0.0162, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.093545635484788, | |
| "grad_norm": 0.2035679817199707, | |
| "learning_rate": 3.80638950832186e-05, | |
| "loss": 0.0174, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.1087100748744194, | |
| "grad_norm": 0.200400710105896, | |
| "learning_rate": 3.7550112204051014e-05, | |
| "loss": 0.0096, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.1087100748744194, | |
| "eval_loss": 0.04321278631687164, | |
| "eval_runtime": 159.8384, | |
| "eval_samples_per_second": 7.339, | |
| "eval_steps_per_second": 7.339, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.1238745142640507, | |
| "grad_norm": 0.10098811984062195, | |
| "learning_rate": 3.703772701767167e-05, | |
| "loss": 0.0098, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.139038953653682, | |
| "grad_norm": 0.3599345088005066, | |
| "learning_rate": 3.652679704725596e-05, | |
| "loss": 0.0095, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.1542033930433133, | |
| "grad_norm": 0.17654354870319366, | |
| "learning_rate": 3.601737965260882e-05, | |
| "loss": 0.0111, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.1693678324329446, | |
| "grad_norm": 0.3085581660270691, | |
| "learning_rate": 3.550953202372503e-05, | |
| "loss": 0.0072, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.184532271822576, | |
| "grad_norm": 0.5332884192466736, | |
| "learning_rate": 3.500331117436895e-05, | |
| "loss": 0.0076, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.184532271822576, | |
| "eval_loss": 0.048784155398607254, | |
| "eval_runtime": 159.7604, | |
| "eval_samples_per_second": 7.342, | |
| "eval_steps_per_second": 7.342, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.1996967112122072, | |
| "grad_norm": 0.20715317130088806, | |
| "learning_rate": 3.4498773935673626e-05, | |
| "loss": 0.0089, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.2148611506018385, | |
| "grad_norm": 0.1660224199295044, | |
| "learning_rate": 3.399597694976081e-05, | |
| "loss": 0.0116, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.23002558999147, | |
| "grad_norm": 0.19592803716659546, | |
| "learning_rate": 3.349497666338187e-05, | |
| "loss": 0.0152, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.245190029381101, | |
| "grad_norm": 0.19497735798358917, | |
| "learning_rate": 3.299582932158085e-05, | |
| "loss": 0.0118, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.260354468770733, | |
| "grad_norm": 0.23641538619995117, | |
| "learning_rate": 3.2498590961379996e-05, | |
| "loss": 0.0122, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.260354468770733, | |
| "eval_loss": 0.04506111145019531, | |
| "eval_runtime": 159.8166, | |
| "eval_samples_per_second": 7.34, | |
| "eval_steps_per_second": 7.34, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.2755189081603637, | |
| "grad_norm": 0.037290170788764954, | |
| "learning_rate": 3.200331740548887e-05, | |
| "loss": 0.0105, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.2906833475499955, | |
| "grad_norm": 0.43852153420448303, | |
| "learning_rate": 3.1510064256037274e-05, | |
| "loss": 0.0136, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.3058477869396263, | |
| "grad_norm": 0.11988549679517746, | |
| "learning_rate": 3.1018886888333065e-05, | |
| "loss": 0.0089, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.321012226329258, | |
| "grad_norm": 0.32281386852264404, | |
| "learning_rate": 3.052984044464548e-05, | |
| "loss": 0.0159, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.3361766657188894, | |
| "grad_norm": 0.1835431307554245, | |
| "learning_rate": 3.0042979828014496e-05, | |
| "loss": 0.0101, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.3361766657188894, | |
| "eval_loss": 0.046676941215991974, | |
| "eval_runtime": 159.7855, | |
| "eval_samples_per_second": 7.341, | |
| "eval_steps_per_second": 7.341, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.3513411051085207, | |
| "grad_norm": 0.20971719920635223, | |
| "learning_rate": 2.9558359696087152e-05, | |
| "loss": 0.015, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.366505544498152, | |
| "grad_norm": 0.30802661180496216, | |
| "learning_rate": 2.9076034454981332e-05, | |
| "loss": 0.0143, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.3816699838877833, | |
| "grad_norm": 0.24673651158809662, | |
| "learning_rate": 2.8596058253177932e-05, | |
| "loss": 0.0124, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.3968344232774146, | |
| "grad_norm": 0.09349372982978821, | |
| "learning_rate": 2.811848497544175e-05, | |
| "loss": 0.013, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.411998862667046, | |
| "grad_norm": 0.14247673749923706, | |
| "learning_rate": 2.764336823677216e-05, | |
| "loss": 0.0124, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.411998862667046, | |
| "eval_loss": 0.04635697975754738, | |
| "eval_runtime": 159.7764, | |
| "eval_samples_per_second": 7.342, | |
| "eval_steps_per_second": 7.342, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.427163302056677, | |
| "grad_norm": 0.2467641681432724, | |
| "learning_rate": 2.717076137638388e-05, | |
| "loss": 0.0096, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.4423277414463085, | |
| "grad_norm": 0.32236963510513306, | |
| "learning_rate": 2.6700717451719066e-05, | |
| "loss": 0.0153, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.45749218083594, | |
| "grad_norm": 0.33867254853248596, | |
| "learning_rate": 2.6233289232490447e-05, | |
| "loss": 0.0156, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.472656620225571, | |
| "grad_norm": 0.17547796666622162, | |
| "learning_rate": 2.5768529194757474e-05, | |
| "loss": 0.0071, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.4878210596152024, | |
| "grad_norm": 0.22590090334415436, | |
| "learning_rate": 2.5306489515034713e-05, | |
| "loss": 0.0113, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.4878210596152024, | |
| "eval_loss": 0.047961652278900146, | |
| "eval_runtime": 159.815, | |
| "eval_samples_per_second": 7.34, | |
| "eval_steps_per_second": 7.34, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.5029854990048337, | |
| "grad_norm": 0.1563863903284073, | |
| "learning_rate": 2.484722206443455e-05, | |
| "loss": 0.017, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.518149938394465, | |
| "grad_norm": 0.09933234751224518, | |
| "learning_rate": 2.4390778402843605e-05, | |
| "loss": 0.0129, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.5333143777840963, | |
| "grad_norm": 0.2538220286369324, | |
| "learning_rate": 2.3937209773134472e-05, | |
| "loss": 0.0094, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.5484788171737276, | |
| "grad_norm": 0.32708045840263367, | |
| "learning_rate": 2.3486567095412864e-05, | |
| "loss": 0.0098, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.563643256563359, | |
| "grad_norm": 0.3891366720199585, | |
| "learning_rate": 2.3038900961301053e-05, | |
| "loss": 0.0126, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.563643256563359, | |
| "eval_loss": 0.047991689294576645, | |
| "eval_runtime": 159.8161, | |
| "eval_samples_per_second": 7.34, | |
| "eval_steps_per_second": 7.34, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.57880769595299, | |
| "grad_norm": 0.06344733387231827, | |
| "learning_rate": 2.259426162825819e-05, | |
| "loss": 0.0099, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 3.5939721353426215, | |
| "grad_norm": 0.5568035244941711, | |
| "learning_rate": 2.215269901393805e-05, | |
| "loss": 0.0164, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 3.609136574732253, | |
| "grad_norm": 0.06175050139427185, | |
| "learning_rate": 2.1714262690585192e-05, | |
| "loss": 0.0094, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 3.624301014121884, | |
| "grad_norm": 0.6271355152130127, | |
| "learning_rate": 2.1279001879469424e-05, | |
| "loss": 0.0126, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 3.6394654535115154, | |
| "grad_norm": 0.14509186148643494, | |
| "learning_rate": 2.084696544536025e-05, | |
| "loss": 0.0114, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.6394654535115154, | |
| "eval_loss": 0.04358929395675659, | |
| "eval_runtime": 159.7768, | |
| "eval_samples_per_second": 7.341, | |
| "eval_steps_per_second": 7.341, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.6546298929011467, | |
| "grad_norm": 0.04602226987481117, | |
| "learning_rate": 2.0418201891040778e-05, | |
| "loss": 0.0102, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 3.669794332290778, | |
| "grad_norm": 0.4110815227031708, | |
| "learning_rate": 1.9992759351862772e-05, | |
| "loss": 0.0201, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 3.6849587716804093, | |
| "grad_norm": 0.3334294557571411, | |
| "learning_rate": 1.9570685590342463e-05, | |
| "loss": 0.0098, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 3.7001232110700406, | |
| "grad_norm": 0.19647225737571716, | |
| "learning_rate": 1.9152027990798748e-05, | |
| "loss": 0.0109, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 3.715287650459672, | |
| "grad_norm": 0.06987540423870087, | |
| "learning_rate": 1.8736833554033263e-05, | |
| "loss": 0.0071, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.715287650459672, | |
| "eval_loss": 0.047177620232105255, | |
| "eval_runtime": 159.8168, | |
| "eval_samples_per_second": 7.34, | |
| "eval_steps_per_second": 7.34, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.7304520898493037, | |
| "grad_norm": 0.3399925231933594, | |
| "learning_rate": 1.8325148892054093e-05, | |
| "loss": 0.0125, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 3.7456165292389345, | |
| "grad_norm": 0.1921069622039795, | |
| "learning_rate": 1.7917020222842662e-05, | |
| "loss": 0.0089, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 3.7607809686285663, | |
| "grad_norm": 0.10305212438106537, | |
| "learning_rate": 1.751249336516513e-05, | |
| "loss": 0.0106, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 3.775945408018197, | |
| "grad_norm": 0.07067303359508514, | |
| "learning_rate": 1.7111613733428522e-05, | |
| "loss": 0.0125, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 3.791109847407829, | |
| "grad_norm": 0.1127607673406601, | |
| "learning_rate": 1.671442633258218e-05, | |
| "loss": 0.0117, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.791109847407829, | |
| "eval_loss": 0.04904229938983917, | |
| "eval_runtime": 159.7619, | |
| "eval_samples_per_second": 7.342, | |
| "eval_steps_per_second": 7.342, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8062742867974597, | |
| "grad_norm": 0.47790926694869995, | |
| "learning_rate": 1.6320975753065453e-05, | |
| "loss": 0.0095, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 3.8214387261870915, | |
| "grad_norm": 0.6272779703140259, | |
| "learning_rate": 1.5931306165801452e-05, | |
| "loss": 0.01, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 3.8366031655767223, | |
| "grad_norm": 0.11979278177022934, | |
| "learning_rate": 1.554546131723848e-05, | |
| "loss": 0.0112, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 3.851767604966354, | |
| "grad_norm": 0.4187261760234833, | |
| "learning_rate": 1.5163484524438516e-05, | |
| "loss": 0.0075, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 3.8669320443559854, | |
| "grad_norm": 0.5277063846588135, | |
| "learning_rate": 1.4785418670214496e-05, | |
| "loss": 0.0145, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.8669320443559854, | |
| "eval_loss": 0.047451041638851166, | |
| "eval_runtime": 159.8306, | |
| "eval_samples_per_second": 7.339, | |
| "eval_steps_per_second": 7.339, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 3.8820964837456167, | |
| "grad_norm": 0.4268508553504944, | |
| "learning_rate": 1.4411306198315777e-05, | |
| "loss": 0.0123, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 3.897260923135248, | |
| "grad_norm": 0.10897725075483322, | |
| "learning_rate": 1.4041189108663421e-05, | |
| "loss": 0.008, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 3.9124253625248793, | |
| "grad_norm": 0.24288566410541534, | |
| "learning_rate": 1.3675108952634824e-05, | |
| "loss": 0.007, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 3.9275898019145106, | |
| "grad_norm": 0.3925674259662628, | |
| "learning_rate": 1.3313106828399147e-05, | |
| "loss": 0.0147, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 3.942754241304142, | |
| "grad_norm": 0.07586287707090378, | |
| "learning_rate": 1.2955223376303205e-05, | |
| "loss": 0.008, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.942754241304142, | |
| "eval_loss": 0.04594454541802406, | |
| "eval_runtime": 159.8774, | |
| "eval_samples_per_second": 7.337, | |
| "eval_steps_per_second": 7.337, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.957918680693773, | |
| "grad_norm": 0.19912473857402802, | |
| "learning_rate": 1.2601498774309112e-05, | |
| "loss": 0.0104, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 3.9730831200834045, | |
| "grad_norm": 0.40546008944511414, | |
| "learning_rate": 1.2251972733483612e-05, | |
| "loss": 0.0117, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 3.9882475594730358, | |
| "grad_norm": 0.6201793551445007, | |
| "learning_rate": 1.19066844935399e-05, | |
| "loss": 0.0106, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.003411998862667, | |
| "grad_norm": 0.12441912293434143, | |
| "learning_rate": 1.156567281843241e-05, | |
| "loss": 0.0095, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.018576438252298, | |
| "grad_norm": 0.24595439434051514, | |
| "learning_rate": 1.1228975992004842e-05, | |
| "loss": 0.0044, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.018576438252298, | |
| "eval_loss": 0.048001233488321304, | |
| "eval_runtime": 159.8103, | |
| "eval_samples_per_second": 7.34, | |
| "eval_steps_per_second": 7.34, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.03374087764193, | |
| "grad_norm": 0.33981209993362427, | |
| "learning_rate": 1.0896631813692377e-05, | |
| "loss": 0.0076, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.048905317031561, | |
| "grad_norm": 0.2191678285598755, | |
| "learning_rate": 1.0568677594277899e-05, | |
| "loss": 0.0043, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.064069756421192, | |
| "grad_norm": 0.06934363394975662, | |
| "learning_rate": 1.0245150151703509e-05, | |
| "loss": 0.003, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.079234195810824, | |
| "grad_norm": 0.19347621500492096, | |
| "learning_rate": 9.926085806936918e-06, | |
| "loss": 0.0057, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.094398635200455, | |
| "grad_norm": 0.28373149037361145, | |
| "learning_rate": 9.611520379894068e-06, | |
| "loss": 0.0061, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.094398635200455, | |
| "eval_loss": 0.05424731224775314, | |
| "eval_runtime": 159.7665, | |
| "eval_samples_per_second": 7.342, | |
| "eval_steps_per_second": 7.342, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.109563074590087, | |
| "grad_norm": 0.10718700289726257, | |
| "learning_rate": 9.301489185417578e-06, | |
| "loss": 0.0043, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.1247275139797175, | |
| "grad_norm": 0.06889745593070984, | |
| "learning_rate": 8.996027029312314e-06, | |
| "loss": 0.0038, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.139891953369349, | |
| "grad_norm": 0.4347403645515442, | |
| "learning_rate": 8.695168204437721e-06, | |
| "loss": 0.0102, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.15505639275898, | |
| "grad_norm": 1.3129316568374634, | |
| "learning_rate": 8.398946486858029e-06, | |
| "loss": 0.0066, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.170220832148612, | |
| "grad_norm": 0.044139549136161804, | |
| "learning_rate": 8.107395132050333e-06, | |
| "loss": 0.0048, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.170220832148612, | |
| "eval_loss": 0.05559329688549042, | |
| "eval_runtime": 159.8213, | |
| "eval_samples_per_second": 7.339, | |
| "eval_steps_per_second": 7.339, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.185385271538243, | |
| "grad_norm": 0.0535513274371624, | |
| "learning_rate": 7.820546871171142e-06, | |
| "loss": 0.0067, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.200549710927874, | |
| "grad_norm": 0.37897995114326477, | |
| "learning_rate": 7.5384339073818145e-06, | |
| "loss": 0.0052, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.215714150317505, | |
| "grad_norm": 0.18212661147117615, | |
| "learning_rate": 7.261087912233228e-06, | |
| "loss": 0.0057, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.230878589707137, | |
| "grad_norm": 0.09377432614564896, | |
| "learning_rate": 6.988540022110235e-06, | |
| "loss": 0.0062, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.246043029096768, | |
| "grad_norm": 0.1912451833486557, | |
| "learning_rate": 6.720820834735969e-06, | |
| "loss": 0.0043, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.246043029096768, | |
| "eval_loss": 0.05572189763188362, | |
| "eval_runtime": 159.8918, | |
| "eval_samples_per_second": 7.336, | |
| "eval_steps_per_second": 7.336, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.2612074684864, | |
| "grad_norm": 0.32161590456962585, | |
| "learning_rate": 6.457960405736968e-06, | |
| "loss": 0.0077, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.2763719078760305, | |
| "grad_norm": 0.07208878546953201, | |
| "learning_rate": 6.199988245268778e-06, | |
| "loss": 0.0043, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.291536347265662, | |
| "grad_norm": 0.09758219122886658, | |
| "learning_rate": 5.946933314703157e-06, | |
| "loss": 0.003, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.306700786655293, | |
| "grad_norm": 0.22281204164028168, | |
| "learning_rate": 5.698824023376531e-06, | |
| "loss": 0.0051, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.321865226044925, | |
| "grad_norm": 0.4457966983318329, | |
| "learning_rate": 5.455688225400802e-06, | |
| "loss": 0.0049, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.321865226044925, | |
| "eval_loss": 0.05698266625404358, | |
| "eval_runtime": 159.8196, | |
| "eval_samples_per_second": 7.34, | |
| "eval_steps_per_second": 7.34, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.337029665434556, | |
| "grad_norm": 0.22873935103416443, | |
| "learning_rate": 5.217553216536098e-06, | |
| "loss": 0.0069, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.3521941048241874, | |
| "grad_norm": 0.13292750716209412, | |
| "learning_rate": 4.984445731126542e-06, | |
| "loss": 0.0059, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.367358544213818, | |
| "grad_norm": 0.021369215101003647, | |
| "learning_rate": 4.756391939098853e-06, | |
| "loss": 0.0036, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.38252298360345, | |
| "grad_norm": 0.06876254081726074, | |
| "learning_rate": 4.533417443024374e-06, | |
| "loss": 0.0077, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 4.397687422993081, | |
| "grad_norm": 0.1053222119808197, | |
| "learning_rate": 4.315547275244769e-06, | |
| "loss": 0.0019, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.397687422993081, | |
| "eval_loss": 0.05874261260032654, | |
| "eval_runtime": 159.7069, | |
| "eval_samples_per_second": 7.345, | |
| "eval_steps_per_second": 7.345, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.412851862382713, | |
| "grad_norm": 2.4315688610076904, | |
| "learning_rate": 4.10280589506179e-06, | |
| "loss": 0.0069, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 4.4280163017723435, | |
| "grad_norm": 0.21709750592708588, | |
| "learning_rate": 3.895217185991312e-06, | |
| "loss": 0.0039, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.443180741161975, | |
| "grad_norm": 0.3738054037094116, | |
| "learning_rate": 3.692804453082038e-06, | |
| "loss": 0.0046, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.458345180551606, | |
| "grad_norm": 0.07227662950754166, | |
| "learning_rate": 3.495590420299194e-06, | |
| "loss": 0.0051, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.473509619941238, | |
| "grad_norm": 0.25434616208076477, | |
| "learning_rate": 3.3035972279733207e-06, | |
| "loss": 0.0053, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.473509619941238, | |
| "eval_loss": 0.05949469655752182, | |
| "eval_runtime": 159.7116, | |
| "eval_samples_per_second": 7.344, | |
| "eval_steps_per_second": 7.344, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 4.488674059330869, | |
| "grad_norm": 0.5970948934555054, | |
| "learning_rate": 3.116846430314796e-06, | |
| "loss": 0.0053, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 4.5038384987205005, | |
| "grad_norm": 0.07033645361661911, | |
| "learning_rate": 2.9353589929939087e-06, | |
| "loss": 0.0025, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 4.519002938110132, | |
| "grad_norm": 0.0930805653333664, | |
| "learning_rate": 2.7591552907872574e-06, | |
| "loss": 0.0043, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 4.534167377499763, | |
| "grad_norm": 0.10770905762910843, | |
| "learning_rate": 2.5882551052902883e-06, | |
| "loss": 0.0049, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 4.549331816889394, | |
| "grad_norm": 0.7434157133102417, | |
| "learning_rate": 2.4226776226965453e-06, | |
| "loss": 0.0072, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.549331816889394, | |
| "eval_loss": 0.05965949594974518, | |
| "eval_runtime": 159.644, | |
| "eval_samples_per_second": 7.348, | |
| "eval_steps_per_second": 7.348, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.564496256279026, | |
| "grad_norm": 0.01642761193215847, | |
| "learning_rate": 2.262441431643697e-06, | |
| "loss": 0.0062, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 4.579660695668657, | |
| "grad_norm": 0.06564271450042725, | |
| "learning_rate": 2.1075645211266927e-06, | |
| "loss": 0.0017, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 4.594825135058288, | |
| "grad_norm": 0.38251250982284546, | |
| "learning_rate": 1.9580642784782053e-06, | |
| "loss": 0.0046, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 4.60998957444792, | |
| "grad_norm": 0.10563746094703674, | |
| "learning_rate": 1.813957487416651e-06, | |
| "loss": 0.0023, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 4.625154013837551, | |
| "grad_norm": 0.1952509880065918, | |
| "learning_rate": 1.6752603261619315e-06, | |
| "loss": 0.0047, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.625154013837551, | |
| "eval_loss": 0.05990082770586014, | |
| "eval_runtime": 159.6308, | |
| "eval_samples_per_second": 7.348, | |
| "eval_steps_per_second": 7.348, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 4.640318453227183, | |
| "grad_norm": 0.2965155243873596, | |
| "learning_rate": 1.541988365619207e-06, | |
| "loss": 0.0044, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 4.6554828926168135, | |
| "grad_norm": 0.28507962822914124, | |
| "learning_rate": 1.4141565676307865e-06, | |
| "loss": 0.0052, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 4.670647332006445, | |
| "grad_norm": 0.09388308972120285, | |
| "learning_rate": 1.291779283296468e-06, | |
| "loss": 0.0029, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 4.685811771396076, | |
| "grad_norm": 0.3553768992424011, | |
| "learning_rate": 1.1748702513623922e-06, | |
| "loss": 0.0064, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 4.700976210785708, | |
| "grad_norm": 0.01811501383781433, | |
| "learning_rate": 1.0634425966786155e-06, | |
| "loss": 0.003, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.700976210785708, | |
| "eval_loss": 0.06024787947535515, | |
| "eval_runtime": 159.5976, | |
| "eval_samples_per_second": 7.35, | |
| "eval_steps_per_second": 7.35, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.716140650175339, | |
| "grad_norm": 0.17447508871555328, | |
| "learning_rate": 9.575088287257118e-07, | |
| "loss": 0.005, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 4.73130508956497, | |
| "grad_norm": 2.3331735134124756, | |
| "learning_rate": 8.570808402103392e-07, | |
| "loss": 0.0059, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 4.746469528954601, | |
| "grad_norm": 0.20535585284233093, | |
| "learning_rate": 7.62169905730109e-07, | |
| "loss": 0.0048, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 4.761633968344233, | |
| "grad_norm": 0.5628078579902649, | |
| "learning_rate": 6.727866805078531e-07, | |
| "loss": 0.0029, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 4.776798407733864, | |
| "grad_norm": 0.09494868665933609, | |
| "learning_rate": 5.889411991953975e-07, | |
| "loss": 0.0025, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.776798407733864, | |
| "eval_loss": 0.06030402332544327, | |
| "eval_runtime": 159.5412, | |
| "eval_samples_per_second": 7.352, | |
| "eval_steps_per_second": 7.352, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 4.791962847123496, | |
| "grad_norm": 0.21833842992782593, | |
| "learning_rate": 5.106428747470137e-07, | |
| "loss": 0.0033, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 4.8071272865131265, | |
| "grad_norm": 0.24270612001419067, | |
| "learning_rate": 4.37900497362681e-07, | |
| "loss": 0.0061, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 4.822291725902758, | |
| "grad_norm": 0.10475039482116699, | |
| "learning_rate": 3.7072223350124167e-07, | |
| "loss": 0.0028, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 4.837456165292389, | |
| "grad_norm": 0.19077710807323456, | |
| "learning_rate": 3.0911562496358517e-07, | |
| "loss": 0.0062, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 4.852620604682021, | |
| "grad_norm": 0.36398255825042725, | |
| "learning_rate": 2.53087588045986e-07, | |
| "loss": 0.0027, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.852620604682021, | |
| "eval_loss": 0.06035413593053818, | |
| "eval_runtime": 159.4969, | |
| "eval_samples_per_second": 7.354, | |
| "eval_steps_per_second": 7.354, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.867785044071652, | |
| "grad_norm": 0.1687215268611908, | |
| "learning_rate": 2.026444127636029e-07, | |
| "loss": 0.0034, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 4.882949483461283, | |
| "grad_norm": 0.03258746862411499, | |
| "learning_rate": 1.577917621443825e-07, | |
| "loss": 0.0051, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 4.898113922850914, | |
| "grad_norm": 0.1656346619129181, | |
| "learning_rate": 1.185346715932345e-07, | |
| "loss": 0.0083, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 4.913278362240546, | |
| "grad_norm": 0.45706111192703247, | |
| "learning_rate": 8.48775483267783e-08, | |
| "loss": 0.0055, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 4.928442801630177, | |
| "grad_norm": 0.07887681573629379, | |
| "learning_rate": 5.6824170878544414e-08, | |
| "loss": 0.0022, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 4.928442801630177, | |
| "eval_loss": 0.060390591621398926, | |
| "eval_runtime": 159.4858, | |
| "eval_samples_per_second": 7.355, | |
| "eval_steps_per_second": 7.355, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 4.943607241019809, | |
| "grad_norm": 0.4127300977706909, | |
| "learning_rate": 3.4377688674774866e-08, | |
| "loss": 0.0053, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 4.9587716804094395, | |
| "grad_norm": 0.03591546416282654, | |
| "learning_rate": 1.7540621680850466e-08, | |
| "loss": 0.0029, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 4.973936119799071, | |
| "grad_norm": 0.06558314710855484, | |
| "learning_rate": 6.314860118400434e-09, | |
| "loss": 0.0064, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 4.989100559188703, | |
| "grad_norm": 0.06449679285287857, | |
| "learning_rate": 7.016642530777162e-10, | |
| "loss": 0.0032, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 4.996682778883518, | |
| "step": 3295, | |
| "total_flos": 8.269502901505229e+17, | |
| "train_loss": 0.04518177697310657, | |
| "train_runtime": 36119.4942, | |
| "train_samples_per_second": 1.461, | |
| "train_steps_per_second": 0.091 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3295, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.269502901505229e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |