| { | |
| "best_metric": 4.347445964813232, | |
| "best_model_checkpoint": "/Users/bbunzeck/Documents/lexdec-llamas/medium-bpe/checkpoint-8136", | |
| "epoch": 0.9999723214038584, | |
| "eval_steps": 904, | |
| "global_step": 9032, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001107143845664148, | |
| "grad_norm": 1.6811453104019165, | |
| "learning_rate": 1.4999999999999999e-05, | |
| "loss": 8.2236, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.002214287691328296, | |
| "grad_norm": 1.668249487876892, | |
| "learning_rate": 2.9999999999999997e-05, | |
| "loss": 8.2247, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0033214315369924437, | |
| "grad_norm": 2.0669946670532227, | |
| "learning_rate": 4.4999999999999996e-05, | |
| "loss": 8.0856, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.004428575382656592, | |
| "grad_norm": 2.384383201599121, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 7.9062, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.00553571922832074, | |
| "grad_norm": 1.7666205167770386, | |
| "learning_rate": 7.5e-05, | |
| "loss": 7.6013, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.006642863073984887, | |
| "grad_norm": 1.6850837469100952, | |
| "learning_rate": 8.999999999999999e-05, | |
| "loss": 7.4349, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.007750006919649035, | |
| "grad_norm": 1.6724164485931396, | |
| "learning_rate": 0.00010499999999999999, | |
| "loss": 7.3511, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.008857150765313184, | |
| "grad_norm": 1.2229112386703491, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 7.0364, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.009964294610977332, | |
| "grad_norm": 1.2853370904922485, | |
| "learning_rate": 0.000135, | |
| "loss": 6.7934, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.01107143845664148, | |
| "grad_norm": 1.1248363256454468, | |
| "learning_rate": 0.00015, | |
| "loss": 6.5628, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.012178582302305627, | |
| "grad_norm": 1.1091172695159912, | |
| "learning_rate": 0.000165, | |
| "loss": 6.3637, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.013285726147969775, | |
| "grad_norm": 0.8531168699264526, | |
| "learning_rate": 0.00017999999999999998, | |
| "loss": 6.1565, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.014392869993633923, | |
| "grad_norm": 0.6321163773536682, | |
| "learning_rate": 0.000195, | |
| "loss": 6.1479, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.01550001383929807, | |
| "grad_norm": 0.6747925877571106, | |
| "learning_rate": 0.00020999999999999998, | |
| "loss": 6.0843, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.01660715768496222, | |
| "grad_norm": 0.6815639734268188, | |
| "learning_rate": 0.000225, | |
| "loss": 6.0027, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.017714301530626368, | |
| "grad_norm": 0.6445911526679993, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 6.1389, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.018821445376290515, | |
| "grad_norm": 0.7667441368103027, | |
| "learning_rate": 0.00025499999999999996, | |
| "loss": 5.9253, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.019928589221954663, | |
| "grad_norm": 0.7087548971176147, | |
| "learning_rate": 0.00027, | |
| "loss": 5.9363, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.02103573306761881, | |
| "grad_norm": 0.8394642472267151, | |
| "learning_rate": 0.000285, | |
| "loss": 5.8794, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.02214287691328296, | |
| "grad_norm": 0.7698648571968079, | |
| "learning_rate": 0.0003, | |
| "loss": 5.7686, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.023250020758947106, | |
| "grad_norm": 1.0190988779067993, | |
| "learning_rate": 0.0002999990510519666, | |
| "loss": 5.79, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.024357164604611254, | |
| "grad_norm": 0.9215103387832642, | |
| "learning_rate": 0.0002999962042198732, | |
| "loss": 5.7914, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.025464308450275402, | |
| "grad_norm": 0.7936483025550842, | |
| "learning_rate": 0.0002999914595397396, | |
| "loss": 5.7724, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.02657145229593955, | |
| "grad_norm": 0.7805729508399963, | |
| "learning_rate": 0.00029998481707159863, | |
| "loss": 5.6098, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.027678596141603697, | |
| "grad_norm": 0.8455389142036438, | |
| "learning_rate": 0.000299976276899495, | |
| "loss": 5.6673, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.028785739987267845, | |
| "grad_norm": 0.970041811466217, | |
| "learning_rate": 0.00029996583913148453, | |
| "loss": 5.4241, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.029892883832931993, | |
| "grad_norm": 0.8078948855400085, | |
| "learning_rate": 0.00029995350389963246, | |
| "loss": 5.5628, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.03100002767859614, | |
| "grad_norm": 0.7417298555374146, | |
| "learning_rate": 0.0002999392713600121, | |
| "loss": 5.5283, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.03210717152426029, | |
| "grad_norm": 0.931107759475708, | |
| "learning_rate": 0.00029992314169270256, | |
| "loss": 5.5349, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.03321431536992444, | |
| "grad_norm": 1.020939826965332, | |
| "learning_rate": 0.0002999051151017868, | |
| "loss": 5.4123, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.03432145921558859, | |
| "grad_norm": 0.9386515617370605, | |
| "learning_rate": 0.0002998851918153488, | |
| "loss": 5.4886, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.035428603061252735, | |
| "grad_norm": 0.9938029050827026, | |
| "learning_rate": 0.0002998633720854707, | |
| "loss": 5.4602, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.03653574690691688, | |
| "grad_norm": 0.9082671403884888, | |
| "learning_rate": 0.0002998396561882297, | |
| "loss": 5.3659, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.03764289075258103, | |
| "grad_norm": 0.9395098090171814, | |
| "learning_rate": 0.0002998140444236946, | |
| "loss": 5.3126, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.03875003459824518, | |
| "grad_norm": 0.9639041423797607, | |
| "learning_rate": 0.0002997865371159217, | |
| "loss": 5.3202, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.039857178443909326, | |
| "grad_norm": 0.8728797435760498, | |
| "learning_rate": 0.0002997571346129512, | |
| "loss": 5.2688, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.040964322289573474, | |
| "grad_norm": 0.8130624890327454, | |
| "learning_rate": 0.00029972583728680237, | |
| "loss": 5.2099, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.04207146613523762, | |
| "grad_norm": 0.7736504673957825, | |
| "learning_rate": 0.000299692645533469, | |
| "loss": 5.3048, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.04317860998090177, | |
| "grad_norm": 1.0188201665878296, | |
| "learning_rate": 0.00029965755977291443, | |
| "loss": 5.2316, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.04428575382656592, | |
| "grad_norm": 0.8598982691764832, | |
| "learning_rate": 0.0002996205804490661, | |
| "loss": 5.2301, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.045392897672230065, | |
| "grad_norm": 0.8491718173027039, | |
| "learning_rate": 0.00029958170802981023, | |
| "loss": 5.2577, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.04650004151789421, | |
| "grad_norm": 0.950014591217041, | |
| "learning_rate": 0.0002995409430069855, | |
| "loss": 5.1328, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.04760718536355836, | |
| "grad_norm": 1.0754649639129639, | |
| "learning_rate": 0.00029949828589637703, | |
| "loss": 5.2568, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.04871432920922251, | |
| "grad_norm": 0.9353715181350708, | |
| "learning_rate": 0.00029945373723771, | |
| "loss": 5.1069, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.049821473054886656, | |
| "grad_norm": 1.0789395570755005, | |
| "learning_rate": 0.00029940729759464247, | |
| "loss": 5.09, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.050928616900550804, | |
| "grad_norm": 0.9786452054977417, | |
| "learning_rate": 0.00029935896755475865, | |
| "loss": 5.1341, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.05203576074621495, | |
| "grad_norm": 0.9605515599250793, | |
| "learning_rate": 0.00029930874772956113, | |
| "loss": 5.1059, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.0531429045918791, | |
| "grad_norm": 1.0041091442108154, | |
| "learning_rate": 0.00029925663875446325, | |
| "loss": 4.9431, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.05425004843754325, | |
| "grad_norm": 0.9178552627563477, | |
| "learning_rate": 0.0002992026412887812, | |
| "loss": 5.0927, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.055357192283207395, | |
| "grad_norm": 0.8648872375488281, | |
| "learning_rate": 0.0002991467560157254, | |
| "loss": 5.0499, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.05646433612887154, | |
| "grad_norm": 1.0442804098129272, | |
| "learning_rate": 0.0002990889836423922, | |
| "loss": 4.9932, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.05757147997453569, | |
| "grad_norm": 0.8907873034477234, | |
| "learning_rate": 0.00029902932489975464, | |
| "loss": 5.0501, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.05867862382019984, | |
| "grad_norm": 1.0349349975585938, | |
| "learning_rate": 0.0002989677805426534, | |
| "loss": 5.0168, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.059785767665863986, | |
| "grad_norm": 1.0255941152572632, | |
| "learning_rate": 0.000298904351349787, | |
| "loss": 5.0159, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.06089291151152813, | |
| "grad_norm": 0.9427258968353271, | |
| "learning_rate": 0.0002988390381237023, | |
| "loss": 5.0139, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.06200005535719228, | |
| "grad_norm": 0.8957319855690002, | |
| "learning_rate": 0.000298771841690784, | |
| "loss": 4.934, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.06310719920285643, | |
| "grad_norm": 0.9209539890289307, | |
| "learning_rate": 0.00029870276290124443, | |
| "loss": 5.0307, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.06421434304852058, | |
| "grad_norm": 0.956788957118988, | |
| "learning_rate": 0.0002986318026291127, | |
| "loss": 4.9112, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.06532148689418472, | |
| "grad_norm": 0.9777573943138123, | |
| "learning_rate": 0.0002985589617722236, | |
| "loss": 4.8687, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.06642863073984888, | |
| "grad_norm": 0.8883692622184753, | |
| "learning_rate": 0.0002984842412522063, | |
| "loss": 5.0865, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.06753577458551302, | |
| "grad_norm": 0.9959750771522522, | |
| "learning_rate": 0.00029840764201447266, | |
| "loss": 4.9503, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.06864291843117717, | |
| "grad_norm": 0.9859285950660706, | |
| "learning_rate": 0.0002983291650282053, | |
| "loss": 4.8994, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.06975006227684132, | |
| "grad_norm": 0.881946861743927, | |
| "learning_rate": 0.0002982488112863453, | |
| "loss": 4.9115, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.07085720612250547, | |
| "grad_norm": 1.0792934894561768, | |
| "learning_rate": 0.0002981665818055797, | |
| "loss": 4.819, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.07196434996816961, | |
| "grad_norm": 0.9275200963020325, | |
| "learning_rate": 0.0002980824776263286, | |
| "loss": 4.8829, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.07307149381383377, | |
| "grad_norm": 0.9469479322433472, | |
| "learning_rate": 0.00029799649981273186, | |
| "loss": 4.9309, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.0741786376594979, | |
| "grad_norm": 1.1017831563949585, | |
| "learning_rate": 0.0002979086494526358, | |
| "loss": 4.9706, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.07528578150516206, | |
| "grad_norm": 0.9711742401123047, | |
| "learning_rate": 0.0002978189276575795, | |
| "loss": 4.9493, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.0763929253508262, | |
| "grad_norm": 0.9821876287460327, | |
| "learning_rate": 0.0002977273355627806, | |
| "loss": 4.9942, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.07750006919649036, | |
| "grad_norm": 0.9190080165863037, | |
| "learning_rate": 0.0002976338743271209, | |
| "loss": 4.8791, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.0786072130421545, | |
| "grad_norm": 0.8951785564422607, | |
| "learning_rate": 0.0002975385451331318, | |
| "loss": 4.7981, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.07971435688781865, | |
| "grad_norm": 1.0104964971542358, | |
| "learning_rate": 0.00029744134918697937, | |
| "loss": 4.8694, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.08082150073348279, | |
| "grad_norm": 0.9021739959716797, | |
| "learning_rate": 0.0002973422877184489, | |
| "loss": 4.9033, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.08192864457914695, | |
| "grad_norm": 0.8843702077865601, | |
| "learning_rate": 0.00029724136198092967, | |
| "loss": 4.9234, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.08303578842481109, | |
| "grad_norm": 0.9365085363388062, | |
| "learning_rate": 0.00029713857325139857, | |
| "loss": 4.9289, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.08414293227047524, | |
| "grad_norm": 0.9353443384170532, | |
| "learning_rate": 0.00029703392283040453, | |
| "loss": 4.8665, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.08525007611613938, | |
| "grad_norm": 0.9934764504432678, | |
| "learning_rate": 0.0002969274120420517, | |
| "loss": 4.8546, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.08635721996180354, | |
| "grad_norm": 0.9653151035308838, | |
| "learning_rate": 0.00029681904223398273, | |
| "loss": 4.8277, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.08746436380746768, | |
| "grad_norm": 1.0258365869522095, | |
| "learning_rate": 0.0002967088147773619, | |
| "loss": 4.8412, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.08857150765313183, | |
| "grad_norm": 0.9386922717094421, | |
| "learning_rate": 0.0002965967310668576, | |
| "loss": 4.6317, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.08967865149879597, | |
| "grad_norm": 0.9319560527801514, | |
| "learning_rate": 0.00029648279252062463, | |
| "loss": 4.8415, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.09078579534446013, | |
| "grad_norm": 0.8735382556915283, | |
| "learning_rate": 0.00029636700058028654, | |
| "loss": 4.7614, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.09189293919012427, | |
| "grad_norm": 0.8965985178947449, | |
| "learning_rate": 0.0002962493567109171, | |
| "loss": 4.845, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.09300008303578843, | |
| "grad_norm": 0.8760326504707336, | |
| "learning_rate": 0.00029612986240102186, | |
| "loss": 4.8252, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.09410722688145257, | |
| "grad_norm": 1.004482388496399, | |
| "learning_rate": 0.0002960085191625193, | |
| "loss": 4.7128, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.09521437072711672, | |
| "grad_norm": 1.0535037517547607, | |
| "learning_rate": 0.00029588532853072196, | |
| "loss": 4.818, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.09632151457278088, | |
| "grad_norm": 0.9634362459182739, | |
| "learning_rate": 0.0002957602920643164, | |
| "loss": 4.7606, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.09742865841844502, | |
| "grad_norm": 0.8974128365516663, | |
| "learning_rate": 0.00029563341134534424, | |
| "loss": 4.7981, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.09853580226410917, | |
| "grad_norm": 0.8976161479949951, | |
| "learning_rate": 0.0002955046879791816, | |
| "loss": 4.8849, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.09964294610977331, | |
| "grad_norm": 0.9893444180488586, | |
| "learning_rate": 0.0002953741235945188, | |
| "loss": 4.7963, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.10008580364803897, | |
| "eval_loss": 5.10983419418335, | |
| "eval_runtime": 78.9124, | |
| "eval_samples_per_second": 633.614, | |
| "eval_steps_per_second": 79.202, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.10075008995543747, | |
| "grad_norm": 0.9392728805541992, | |
| "learning_rate": 0.0002952417198433401, | |
| "loss": 4.8371, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.10185723380110161, | |
| "grad_norm": 0.9585736393928528, | |
| "learning_rate": 0.00029510747840090273, | |
| "loss": 4.6876, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.10296437764676576, | |
| "grad_norm": 0.8645158410072327, | |
| "learning_rate": 0.00029497140096571516, | |
| "loss": 4.7819, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.1040715214924299, | |
| "grad_norm": 0.9079055786132812, | |
| "learning_rate": 0.00029483348925951635, | |
| "loss": 4.8303, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.10517866533809406, | |
| "grad_norm": 0.9772407412528992, | |
| "learning_rate": 0.00029469374502725355, | |
| "loss": 4.6956, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.1062858091837582, | |
| "grad_norm": 0.9333701729774475, | |
| "learning_rate": 0.0002945521700370603, | |
| "loss": 4.6928, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.10739295302942235, | |
| "grad_norm": 0.8925185203552246, | |
| "learning_rate": 0.00029440876608023394, | |
| "loss": 4.6874, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.1085000968750865, | |
| "grad_norm": 0.933597981929779, | |
| "learning_rate": 0.00029426353497121334, | |
| "loss": 4.7759, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.10960724072075065, | |
| "grad_norm": 0.9932674169540405, | |
| "learning_rate": 0.00029411647854755534, | |
| "loss": 4.6764, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.11071438456641479, | |
| "grad_norm": 0.9036180377006531, | |
| "learning_rate": 0.0002939675986699121, | |
| "loss": 4.6848, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.11182152841207894, | |
| "grad_norm": 0.8861162066459656, | |
| "learning_rate": 0.0002938168972220071, | |
| "loss": 4.7565, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.11292867225774308, | |
| "grad_norm": 0.9766280651092529, | |
| "learning_rate": 0.0002936643761106117, | |
| "loss": 4.7522, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.11403581610340724, | |
| "grad_norm": 0.921286940574646, | |
| "learning_rate": 0.00029351003726552056, | |
| "loss": 4.6959, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.11514295994907138, | |
| "grad_norm": 0.9068124294281006, | |
| "learning_rate": 0.0002933538826395276, | |
| "loss": 4.8112, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.11625010379473553, | |
| "grad_norm": 1.0276354551315308, | |
| "learning_rate": 0.00029319591420840126, | |
| "loss": 4.631, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.11735724764039968, | |
| "grad_norm": 1.092550277709961, | |
| "learning_rate": 0.0002930361339708592, | |
| "loss": 4.682, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.11846439148606383, | |
| "grad_norm": 1.0077763795852661, | |
| "learning_rate": 0.00029287454394854337, | |
| "loss": 4.8113, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.11957153533172797, | |
| "grad_norm": 0.9678068161010742, | |
| "learning_rate": 0.00029271114618599416, | |
| "loss": 4.5612, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.12067867917739213, | |
| "grad_norm": 0.9793437719345093, | |
| "learning_rate": 0.0002925459427506248, | |
| "loss": 4.5687, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.12178582302305627, | |
| "grad_norm": 0.9270321130752563, | |
| "learning_rate": 0.00029237893573269487, | |
| "loss": 4.659, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.12289296686872042, | |
| "grad_norm": 1.0510565042495728, | |
| "learning_rate": 0.0002922101272452842, | |
| "loss": 4.6692, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.12400011071438456, | |
| "grad_norm": 0.9983336329460144, | |
| "learning_rate": 0.0002920395194242658, | |
| "loss": 4.6254, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.1251072545600487, | |
| "grad_norm": 1.0607930421829224, | |
| "learning_rate": 0.00029186711442827915, | |
| "loss": 4.6428, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.12621439840571286, | |
| "grad_norm": 1.0545765161514282, | |
| "learning_rate": 0.0002916929144387027, | |
| "loss": 4.5578, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.127321542251377, | |
| "grad_norm": 0.9109506607055664, | |
| "learning_rate": 0.0002915169216596262, | |
| "loss": 4.6607, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.12842868609704117, | |
| "grad_norm": 0.9867314696311951, | |
| "learning_rate": 0.00029133913831782307, | |
| "loss": 4.5686, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.1295358299427053, | |
| "grad_norm": 1.0707261562347412, | |
| "learning_rate": 0.00029115956666272194, | |
| "loss": 4.5862, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.13064297378836945, | |
| "grad_norm": 0.9916772246360779, | |
| "learning_rate": 0.0002909782089663785, | |
| "loss": 4.608, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.1317501176340336, | |
| "grad_norm": 1.0324965715408325, | |
| "learning_rate": 0.00029079506752344636, | |
| "loss": 4.4986, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.13285726147969776, | |
| "grad_norm": 0.8864001631736755, | |
| "learning_rate": 0.0002906101446511484, | |
| "loss": 4.6429, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.13396440532536188, | |
| "grad_norm": 0.8967823386192322, | |
| "learning_rate": 0.0002904234426892473, | |
| "loss": 4.5248, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.13507154917102604, | |
| "grad_norm": 1.0033129453659058, | |
| "learning_rate": 0.00029023496400001577, | |
| "loss": 4.548, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.1361786930166902, | |
| "grad_norm": 0.9091253876686096, | |
| "learning_rate": 0.0002900447109682069, | |
| "loss": 4.5643, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.13728583686235435, | |
| "grad_norm": 0.9159227013587952, | |
| "learning_rate": 0.0002898526860010239, | |
| "loss": 4.5561, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.13839298070801848, | |
| "grad_norm": 1.0186384916305542, | |
| "learning_rate": 0.00028965889152808967, | |
| "loss": 4.598, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.13950012455368263, | |
| "grad_norm": 0.9537397623062134, | |
| "learning_rate": 0.00028946333000141595, | |
| "loss": 4.5415, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.14060726839934679, | |
| "grad_norm": 0.9938469529151917, | |
| "learning_rate": 0.00028926600389537246, | |
| "loss": 4.4747, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.14171441224501094, | |
| "grad_norm": 1.0113725662231445, | |
| "learning_rate": 0.00028906691570665545, | |
| "loss": 4.6552, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.1428215560906751, | |
| "grad_norm": 0.9132862091064453, | |
| "learning_rate": 0.0002888660679542562, | |
| "loss": 4.6069, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.14392869993633922, | |
| "grad_norm": 1.1167629957199097, | |
| "learning_rate": 0.00028866346317942915, | |
| "loss": 4.4369, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.14503584378200338, | |
| "grad_norm": 1.068723201751709, | |
| "learning_rate": 0.00028845910394565956, | |
| "loss": 4.6153, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.14614298762766753, | |
| "grad_norm": 1.0072897672653198, | |
| "learning_rate": 0.00028825299283863137, | |
| "loss": 4.5664, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.14725013147333169, | |
| "grad_norm": 1.003943920135498, | |
| "learning_rate": 0.0002880451324661944, | |
| "loss": 4.4993, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.1483572753189958, | |
| "grad_norm": 1.0345332622528076, | |
| "learning_rate": 0.00028783552545833115, | |
| "loss": 4.5299, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.14946441916465997, | |
| "grad_norm": 1.0990606546401978, | |
| "learning_rate": 0.00028762417446712363, | |
| "loss": 4.5322, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.15057156301032412, | |
| "grad_norm": 0.9106954336166382, | |
| "learning_rate": 0.00028741108216672015, | |
| "loss": 4.6655, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.15167870685598828, | |
| "grad_norm": 0.9903298020362854, | |
| "learning_rate": 0.00028719625125330075, | |
| "loss": 4.5705, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.1527858507016524, | |
| "grad_norm": 0.9139270782470703, | |
| "learning_rate": 0.00028697968444504386, | |
| "loss": 4.4956, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.15389299454731656, | |
| "grad_norm": 1.0022941827774048, | |
| "learning_rate": 0.0002867613844820914, | |
| "loss": 4.4806, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.1550001383929807, | |
| "grad_norm": 1.0560743808746338, | |
| "learning_rate": 0.0002865413541265143, | |
| "loss": 4.5748, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.15610728223864487, | |
| "grad_norm": 0.951771080493927, | |
| "learning_rate": 0.0002863195961622776, | |
| "loss": 4.528, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.157214426084309, | |
| "grad_norm": 1.0707118511199951, | |
| "learning_rate": 0.00028609611339520496, | |
| "loss": 4.508, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.15832156992997315, | |
| "grad_norm": 0.9690203666687012, | |
| "learning_rate": 0.0002858709086529436, | |
| "loss": 4.4299, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.1594287137756373, | |
| "grad_norm": 0.9330080151557922, | |
| "learning_rate": 0.0002856439847849281, | |
| "loss": 4.4642, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.16053585762130146, | |
| "grad_norm": 0.9159601926803589, | |
| "learning_rate": 0.00028541534466234463, | |
| "loss": 4.546, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.16164300146696559, | |
| "grad_norm": 1.0101563930511475, | |
| "learning_rate": 0.0002851849911780943, | |
| "loss": 4.5763, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.16275014531262974, | |
| "grad_norm": 1.0901501178741455, | |
| "learning_rate": 0.00028495292724675707, | |
| "loss": 4.5013, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.1638572891582939, | |
| "grad_norm": 1.0354019403457642, | |
| "learning_rate": 0.0002847191558045544, | |
| "loss": 4.4684, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.16496443300395805, | |
| "grad_norm": 0.9800440669059753, | |
| "learning_rate": 0.00028448367980931223, | |
| "loss": 4.4123, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.16607157684962218, | |
| "grad_norm": 0.9751323461532593, | |
| "learning_rate": 0.0002842465022404237, | |
| "loss": 4.5517, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.16717872069528633, | |
| "grad_norm": 1.0189952850341797, | |
| "learning_rate": 0.0002840076260988113, | |
| "loss": 4.3959, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.16828586454095049, | |
| "grad_norm": 0.9599214196205139, | |
| "learning_rate": 0.0002837670544068891, | |
| "loss": 4.5229, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.16939300838661464, | |
| "grad_norm": 0.9783859252929688, | |
| "learning_rate": 0.00028352479020852395, | |
| "loss": 4.4955, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.17050015223227877, | |
| "grad_norm": 0.8636813163757324, | |
| "learning_rate": 0.00028328083656899784, | |
| "loss": 4.4853, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.17160729607794292, | |
| "grad_norm": 0.9867982864379883, | |
| "learning_rate": 0.0002830351965749684, | |
| "loss": 4.4407, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.17271443992360708, | |
| "grad_norm": 1.002221941947937, | |
| "learning_rate": 0.0002827878733344301, | |
| "loss": 4.505, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.17382158376927123, | |
| "grad_norm": 0.9240091443061829, | |
| "learning_rate": 0.00028253886997667505, | |
| "loss": 4.4321, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.17492872761493536, | |
| "grad_norm": 0.989526093006134, | |
| "learning_rate": 0.0002822881896522532, | |
| "loss": 4.383, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.1760358714605995, | |
| "grad_norm": 0.8748844861984253, | |
| "learning_rate": 0.0002820358355329326, | |
| "loss": 4.4833, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.17714301530626367, | |
| "grad_norm": 0.9923803806304932, | |
| "learning_rate": 0.00028178181081165907, | |
| "loss": 4.412, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.17825015915192782, | |
| "grad_norm": 1.0207901000976562, | |
| "learning_rate": 0.0002815261187025162, | |
| "loss": 4.4337, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.17935730299759195, | |
| "grad_norm": 1.0174667835235596, | |
| "learning_rate": 0.00028126876244068415, | |
| "loss": 4.4756, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.1804644468432561, | |
| "grad_norm": 0.956628680229187, | |
| "learning_rate": 0.00028100974528239936, | |
| "loss": 4.3981, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.18157159068892026, | |
| "grad_norm": 0.9558259844779968, | |
| "learning_rate": 0.0002807490705049127, | |
| "loss": 4.4169, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.1826787345345844, | |
| "grad_norm": 0.994884729385376, | |
| "learning_rate": 0.0002804867414064483, | |
| "loss": 4.4233, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.18378587838024854, | |
| "grad_norm": 0.95069819688797, | |
| "learning_rate": 0.00028022276130616216, | |
| "loss": 4.4682, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.1848930222259127, | |
| "grad_norm": 1.0090645551681519, | |
| "learning_rate": 0.0002799571335440994, | |
| "loss": 4.3889, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.18600016607157685, | |
| "grad_norm": 1.0222293138504028, | |
| "learning_rate": 0.00027968986148115265, | |
| "loss": 4.4165, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.187107309917241, | |
| "grad_norm": 0.9743198156356812, | |
| "learning_rate": 0.0002794209484990192, | |
| "loss": 4.4207, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.18821445376290513, | |
| "grad_norm": 0.9410222768783569, | |
| "learning_rate": 0.0002791503980001583, | |
| "loss": 4.3689, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.1893215976085693, | |
| "grad_norm": 1.0146762132644653, | |
| "learning_rate": 0.00027887821340774816, | |
| "loss": 4.3927, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.19042874145423344, | |
| "grad_norm": 1.0493577718734741, | |
| "learning_rate": 0.0002786043981656426, | |
| "loss": 4.4216, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.1915358852998976, | |
| "grad_norm": 0.9138984680175781, | |
| "learning_rate": 0.00027832895573832734, | |
| "loss": 4.4269, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.19264302914556175, | |
| "grad_norm": 0.8888731598854065, | |
| "learning_rate": 0.0002780518896108765, | |
| "loss": 4.4865, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.19375017299122588, | |
| "grad_norm": 0.9734951257705688, | |
| "learning_rate": 0.0002777732032889081, | |
| "loss": 4.3014, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.19485731683689003, | |
| "grad_norm": 0.9728529453277588, | |
| "learning_rate": 0.00027749290029853993, | |
| "loss": 4.3601, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.1959644606825542, | |
| "grad_norm": 0.9178133606910706, | |
| "learning_rate": 0.00027721098418634495, | |
| "loss": 4.4898, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.19707160452821834, | |
| "grad_norm": 1.069711685180664, | |
| "learning_rate": 0.0002769274585193065, | |
| "loss": 4.4106, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.19817874837388247, | |
| "grad_norm": 1.125223994255066, | |
| "learning_rate": 0.0002766423268847727, | |
| "loss": 4.4101, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.19928589221954662, | |
| "grad_norm": 1.0172550678253174, | |
| "learning_rate": 0.0002763555928904117, | |
| "loss": 4.4104, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.20017160729607794, | |
| "eval_loss": 4.8134379386901855, | |
| "eval_runtime": 79.2808, | |
| "eval_samples_per_second": 630.67, | |
| "eval_steps_per_second": 78.834, | |
| "step": 1808 | |
| }, | |
| { | |
| "epoch": 0.20039303606521078, | |
| "grad_norm": 0.9390134811401367, | |
| "learning_rate": 0.00027606726016416567, | |
| "loss": 4.3504, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.20150017991087493, | |
| "grad_norm": 1.0286601781845093, | |
| "learning_rate": 0.00027577733235420487, | |
| "loss": 4.3912, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.20260732375653906, | |
| "grad_norm": 1.0030829906463623, | |
| "learning_rate": 0.0002754858131288817, | |
| "loss": 4.3799, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.20371446760220321, | |
| "grad_norm": 1.0928406715393066, | |
| "learning_rate": 0.000275192706176684, | |
| "loss": 4.4249, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.20482161144786737, | |
| "grad_norm": 1.0378659963607788, | |
| "learning_rate": 0.00027489801520618867, | |
| "loss": 4.3748, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.20592875529353152, | |
| "grad_norm": 1.006199836730957, | |
| "learning_rate": 0.0002746017439460146, | |
| "loss": 4.415, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.20703589913919565, | |
| "grad_norm": 1.0454730987548828, | |
| "learning_rate": 0.0002743038961447757, | |
| "loss": 4.4568, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.2081430429848598, | |
| "grad_norm": 1.004185676574707, | |
| "learning_rate": 0.0002740044755710328, | |
| "loss": 4.3259, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.20925018683052396, | |
| "grad_norm": 1.018157958984375, | |
| "learning_rate": 0.000273703486013247, | |
| "loss": 4.4445, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.21035733067618811, | |
| "grad_norm": 1.0541191101074219, | |
| "learning_rate": 0.0002734009312797309, | |
| "loss": 4.3904, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.21146447452185224, | |
| "grad_norm": 1.093253493309021, | |
| "learning_rate": 0.0002730968151986007, | |
| "loss": 4.3681, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.2125716183675164, | |
| "grad_norm": 0.9687184691429138, | |
| "learning_rate": 0.0002727911416177278, | |
| "loss": 4.3366, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.21367876221318055, | |
| "grad_norm": 0.9801533818244934, | |
| "learning_rate": 0.0002724839144046902, | |
| "loss": 4.409, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.2147859060588447, | |
| "grad_norm": 0.9940269589424133, | |
| "learning_rate": 0.00027217513744672337, | |
| "loss": 4.443, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.21589304990450883, | |
| "grad_norm": 1.037682294845581, | |
| "learning_rate": 0.000271864814650671, | |
| "loss": 4.3273, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.217000193750173, | |
| "grad_norm": 0.9736511707305908, | |
| "learning_rate": 0.00027155294994293604, | |
| "loss": 4.4257, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.21810733759583714, | |
| "grad_norm": 0.9707434773445129, | |
| "learning_rate": 0.0002712395472694304, | |
| "loss": 4.3029, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.2192144814415013, | |
| "grad_norm": 1.0243725776672363, | |
| "learning_rate": 0.0002709246105955254, | |
| "loss": 4.3413, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.22032162528716542, | |
| "grad_norm": 0.9983444213867188, | |
| "learning_rate": 0.0002706081439060016, | |
| "loss": 4.3196, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.22142876913282958, | |
| "grad_norm": 1.1372456550598145, | |
| "learning_rate": 0.00027029015120499825, | |
| "loss": 4.3786, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.22253591297849373, | |
| "grad_norm": 1.010695219039917, | |
| "learning_rate": 0.00026997063651596257, | |
| "loss": 4.4482, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.2236430568241579, | |
| "grad_norm": 0.9691513180732727, | |
| "learning_rate": 0.00026964960388159903, | |
| "loss": 4.4275, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.22475020066982201, | |
| "grad_norm": 1.045020341873169, | |
| "learning_rate": 0.00026932705736381826, | |
| "loss": 4.3429, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.22585734451548617, | |
| "grad_norm": 1.0445832014083862, | |
| "learning_rate": 0.00026900300104368524, | |
| "loss": 4.4422, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.22696448836115032, | |
| "grad_norm": 1.1480467319488525, | |
| "learning_rate": 0.0002686774390213682, | |
| "loss": 4.3145, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.22807163220681448, | |
| "grad_norm": 0.9440922141075134, | |
| "learning_rate": 0.00026835037541608623, | |
| "loss": 4.3922, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.2291787760524786, | |
| "grad_norm": 1.06472647190094, | |
| "learning_rate": 0.00026802181436605757, | |
| "loss": 4.3396, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.23028591989814276, | |
| "grad_norm": 1.0416735410690308, | |
| "learning_rate": 0.00026769176002844705, | |
| "loss": 4.286, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.23139306374380691, | |
| "grad_norm": 1.050659418106079, | |
| "learning_rate": 0.0002673602165793136, | |
| "loss": 4.3494, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.23250020758947107, | |
| "grad_norm": 1.0259265899658203, | |
| "learning_rate": 0.00026702718821355713, | |
| "loss": 4.2568, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.2336073514351352, | |
| "grad_norm": 0.9743848443031311, | |
| "learning_rate": 0.00026669267914486596, | |
| "loss": 4.3431, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.23471449528079935, | |
| "grad_norm": 1.2621181011199951, | |
| "learning_rate": 0.00026635669360566296, | |
| "loss": 4.3081, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.2358216391264635, | |
| "grad_norm": 1.0135219097137451, | |
| "learning_rate": 0.0002660192358470524, | |
| "loss": 4.2929, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.23692878297212766, | |
| "grad_norm": 1.118515968322754, | |
| "learning_rate": 0.000265680310138766, | |
| "loss": 4.3137, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.23803592681779182, | |
| "grad_norm": 1.064631700515747, | |
| "learning_rate": 0.00026533992076910876, | |
| "loss": 4.3147, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.23914307066345594, | |
| "grad_norm": 0.9858683347702026, | |
| "learning_rate": 0.00026499807204490505, | |
| "loss": 4.2987, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.2402502145091201, | |
| "grad_norm": 0.9662603139877319, | |
| "learning_rate": 0.00026465476829144394, | |
| "loss": 4.3254, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.24135735835478425, | |
| "grad_norm": 1.0694128274917603, | |
| "learning_rate": 0.00026431001385242433, | |
| "loss": 4.3021, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.2424645022004484, | |
| "grad_norm": 0.9891943335533142, | |
| "learning_rate": 0.00026396381308990013, | |
| "loss": 4.2128, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.24357164604611253, | |
| "grad_norm": 0.9906103610992432, | |
| "learning_rate": 0.00026361617038422517, | |
| "loss": 4.3037, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.2446787898917767, | |
| "grad_norm": 1.0046894550323486, | |
| "learning_rate": 0.0002632670901339976, | |
| "loss": 4.3087, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.24578593373744084, | |
| "grad_norm": 0.9986013174057007, | |
| "learning_rate": 0.00026291657675600426, | |
| "loss": 4.2126, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.246893077583105, | |
| "grad_norm": 1.0473418235778809, | |
| "learning_rate": 0.00026256463468516495, | |
| "loss": 4.2568, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.24800022142876912, | |
| "grad_norm": 1.1471214294433594, | |
| "learning_rate": 0.00026221126837447617, | |
| "loss": 4.2784, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.24910736527443328, | |
| "grad_norm": 1.0758048295974731, | |
| "learning_rate": 0.00026185648229495477, | |
| "loss": 4.3663, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.2502145091200974, | |
| "grad_norm": 1.0431517362594604, | |
| "learning_rate": 0.0002615002809355814, | |
| "loss": 4.2634, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.2513216529657616, | |
| "grad_norm": 1.030040979385376, | |
| "learning_rate": 0.00026114266880324387, | |
| "loss": 4.3618, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.2524287968114257, | |
| "grad_norm": 1.0849761962890625, | |
| "learning_rate": 0.00026078365042267985, | |
| "loss": 4.2401, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.2535359406570899, | |
| "grad_norm": 1.042960524559021, | |
| "learning_rate": 0.0002604232303364198, | |
| "loss": 4.2827, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.254643084502754, | |
| "grad_norm": 1.0438108444213867, | |
| "learning_rate": 0.00026006141310472953, | |
| "loss": 4.3032, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.25575022834841815, | |
| "grad_norm": 0.9929292798042297, | |
| "learning_rate": 0.00025969820330555234, | |
| "loss": 4.2951, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.25685737219408233, | |
| "grad_norm": 1.2115192413330078, | |
| "learning_rate": 0.0002593336055344513, | |
| "loss": 4.3076, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.25796451603974646, | |
| "grad_norm": 1.09332275390625, | |
| "learning_rate": 0.0002589676244045509, | |
| "loss": 4.2926, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.2590716598854106, | |
| "grad_norm": 0.904438853263855, | |
| "learning_rate": 0.0002586002645464787, | |
| "loss": 4.3919, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.26017880373107477, | |
| "grad_norm": 1.117883324623108, | |
| "learning_rate": 0.00025823153060830695, | |
| "loss": 4.3141, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.2612859475767389, | |
| "grad_norm": 1.028071403503418, | |
| "learning_rate": 0.0002578614272554936, | |
| "loss": 4.2078, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.2623930914224031, | |
| "grad_norm": 1.0506621599197388, | |
| "learning_rate": 0.0002574899591708233, | |
| "loss": 4.2699, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.2635002352680672, | |
| "grad_norm": 1.1326364278793335, | |
| "learning_rate": 0.0002571171310543481, | |
| "loss": 4.3271, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.26460737911373133, | |
| "grad_norm": 1.0729281902313232, | |
| "learning_rate": 0.00025674294762332824, | |
| "loss": 4.2616, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.2657145229593955, | |
| "grad_norm": 1.078567385673523, | |
| "learning_rate": 0.00025636741361217204, | |
| "loss": 4.2071, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.26682166680505964, | |
| "grad_norm": 0.965688943862915, | |
| "learning_rate": 0.0002559905337723763, | |
| "loss": 4.2973, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.26792881065072377, | |
| "grad_norm": 0.9861277341842651, | |
| "learning_rate": 0.00025561231287246617, | |
| "loss": 4.358, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.26903595449638795, | |
| "grad_norm": 0.992848813533783, | |
| "learning_rate": 0.00025523275569793466, | |
| "loss": 4.2873, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.2701430983420521, | |
| "grad_norm": 1.036658525466919, | |
| "learning_rate": 0.00025485186705118227, | |
| "loss": 4.3265, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.27125024218771626, | |
| "grad_norm": 1.0744857788085938, | |
| "learning_rate": 0.00025446965175145614, | |
| "loss": 4.2722, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.2723573860333804, | |
| "grad_norm": 1.0158337354660034, | |
| "learning_rate": 0.00025408611463478896, | |
| "loss": 4.2373, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.2734645298790445, | |
| "grad_norm": 1.1085355281829834, | |
| "learning_rate": 0.0002537012605539379, | |
| "loss": 4.2036, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.2745716737247087, | |
| "grad_norm": 1.0522576570510864, | |
| "learning_rate": 0.0002533150943783234, | |
| "loss": 4.1614, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.2756788175703728, | |
| "grad_norm": 1.1285879611968994, | |
| "learning_rate": 0.0002529276209939672, | |
| "loss": 4.2591, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.27678596141603695, | |
| "grad_norm": 1.0693973302841187, | |
| "learning_rate": 0.0002525388453034307, | |
| "loss": 4.1703, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.27789310526170113, | |
| "grad_norm": 1.051142692565918, | |
| "learning_rate": 0.00025214877222575285, | |
| "loss": 4.2305, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.27900024910736526, | |
| "grad_norm": 1.010872721672058, | |
| "learning_rate": 0.0002517574066963882, | |
| "loss": 4.3454, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.28010739295302944, | |
| "grad_norm": 1.0949872732162476, | |
| "learning_rate": 0.00025136475366714395, | |
| "loss": 4.2176, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.28121453679869357, | |
| "grad_norm": 1.147640347480774, | |
| "learning_rate": 0.0002509708181061178, | |
| "loss": 4.2035, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.2823216806443577, | |
| "grad_norm": 0.9978836178779602, | |
| "learning_rate": 0.0002505756049976346, | |
| "loss": 4.294, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.2834288244900219, | |
| "grad_norm": 1.0316861867904663, | |
| "learning_rate": 0.00025017911934218387, | |
| "loss": 4.1768, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.284535968335686, | |
| "grad_norm": 1.078104019165039, | |
| "learning_rate": 0.000249781366156356, | |
| "loss": 4.1717, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.2856431121813502, | |
| "grad_norm": 0.9834052324295044, | |
| "learning_rate": 0.0002493823504727791, | |
| "loss": 4.2255, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.2867502560270143, | |
| "grad_norm": 1.1098978519439697, | |
| "learning_rate": 0.00024898207734005506, | |
| "loss": 4.2572, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.28785739987267844, | |
| "grad_norm": 0.9931405186653137, | |
| "learning_rate": 0.00024858055182269586, | |
| "loss": 4.1798, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.2889645437183426, | |
| "grad_norm": 1.2387120723724365, | |
| "learning_rate": 0.0002481777790010596, | |
| "loss": 4.1369, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.29007168756400675, | |
| "grad_norm": 1.0250961780548096, | |
| "learning_rate": 0.00024777376397128593, | |
| "loss": 4.2373, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.2911788314096709, | |
| "grad_norm": 1.058564305305481, | |
| "learning_rate": 0.00024736851184523176, | |
| "loss": 4.1806, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.29228597525533506, | |
| "grad_norm": 1.0357334613800049, | |
| "learning_rate": 0.0002469620277504066, | |
| "loss": 4.2104, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.2933931191009992, | |
| "grad_norm": 1.132127285003662, | |
| "learning_rate": 0.00024655431682990743, | |
| "loss": 4.0946, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.29450026294666337, | |
| "grad_norm": 0.98349928855896, | |
| "learning_rate": 0.000246145384242354, | |
| "loss": 4.2092, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.2956074067923275, | |
| "grad_norm": 1.141265869140625, | |
| "learning_rate": 0.0002457352351618233, | |
| "loss": 4.2161, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.2967145506379916, | |
| "grad_norm": 0.9681409597396851, | |
| "learning_rate": 0.00024532387477778415, | |
| "loss": 4.2555, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.2978216944836558, | |
| "grad_norm": 0.9907973408699036, | |
| "learning_rate": 0.00024491130829503165, | |
| "loss": 4.2346, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.29892883832931993, | |
| "grad_norm": 1.1116936206817627, | |
| "learning_rate": 0.0002444975409336211, | |
| "loss": 4.2145, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.30003598217498406, | |
| "grad_norm": 1.0614186525344849, | |
| "learning_rate": 0.00024408257792880216, | |
| "loss": 4.1742, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.3002574109441169, | |
| "eval_loss": 4.660221099853516, | |
| "eval_runtime": 78.5739, | |
| "eval_samples_per_second": 636.344, | |
| "eval_steps_per_second": 79.543, | |
| "step": 2712 | |
| }, | |
| { | |
| "epoch": 0.30114312602064824, | |
| "grad_norm": 1.0894559621810913, | |
| "learning_rate": 0.00024366642453095254, | |
| "loss": 4.2724, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.30225026986631237, | |
| "grad_norm": 1.0665363073349, | |
| "learning_rate": 0.00024324908600551162, | |
| "loss": 4.1882, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.30335741371197655, | |
| "grad_norm": 1.1777842044830322, | |
| "learning_rate": 0.00024283056763291355, | |
| "loss": 4.1859, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.3044645575576407, | |
| "grad_norm": 1.0209153890609741, | |
| "learning_rate": 0.00024241087470852095, | |
| "loss": 4.0898, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.3055717014033048, | |
| "grad_norm": 1.0154958963394165, | |
| "learning_rate": 0.00024199001254255744, | |
| "loss": 4.2616, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.306678845248969, | |
| "grad_norm": 1.0368013381958008, | |
| "learning_rate": 0.00024156798646004065, | |
| "loss": 4.1792, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.3077859890946331, | |
| "grad_norm": 1.12043035030365, | |
| "learning_rate": 0.00024114480180071495, | |
| "loss": 4.1669, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.30889313294029724, | |
| "grad_norm": 1.0686219930648804, | |
| "learning_rate": 0.00024072046391898362, | |
| "loss": 4.2474, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.3100002767859614, | |
| "grad_norm": 1.0914517641067505, | |
| "learning_rate": 0.0002402949781838413, | |
| "loss": 4.2248, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.31110742063162555, | |
| "grad_norm": 1.1432217359542847, | |
| "learning_rate": 0.000239868349978806, | |
| "loss": 4.2573, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.31221456447728974, | |
| "grad_norm": 1.023133397102356, | |
| "learning_rate": 0.00023944058470185106, | |
| "loss": 4.1781, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.31332170832295386, | |
| "grad_norm": 1.1117956638336182, | |
| "learning_rate": 0.0002390116877653367, | |
| "loss": 4.2335, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.314428852168618, | |
| "grad_norm": 1.1213032007217407, | |
| "learning_rate": 0.0002385816645959416, | |
| "loss": 4.1265, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.31553599601428217, | |
| "grad_norm": 1.0825856924057007, | |
| "learning_rate": 0.00023815052063459437, | |
| "loss": 4.1973, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.3166431398599463, | |
| "grad_norm": 0.9986578822135925, | |
| "learning_rate": 0.0002377182613364045, | |
| "loss": 4.1846, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.3177502837056104, | |
| "grad_norm": 1.1070269346237183, | |
| "learning_rate": 0.0002372848921705934, | |
| "loss": 4.1048, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.3188574275512746, | |
| "grad_norm": 0.9696531295776367, | |
| "learning_rate": 0.00023685041862042547, | |
| "loss": 4.2459, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.31996457139693874, | |
| "grad_norm": 1.0587655305862427, | |
| "learning_rate": 0.00023641484618313818, | |
| "loss": 4.1294, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.3210717152426029, | |
| "grad_norm": 1.0469372272491455, | |
| "learning_rate": 0.00023597818036987306, | |
| "loss": 4.2292, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.32217885908826704, | |
| "grad_norm": 1.104966402053833, | |
| "learning_rate": 0.00023554042670560554, | |
| "loss": 4.2957, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.32328600293393117, | |
| "grad_norm": 1.0973048210144043, | |
| "learning_rate": 0.00023510159072907545, | |
| "loss": 4.1428, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.32439314677959535, | |
| "grad_norm": 0.9804009795188904, | |
| "learning_rate": 0.0002346616779927165, | |
| "loss": 4.2025, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.3255002906252595, | |
| "grad_norm": 1.082687497138977, | |
| "learning_rate": 0.00023422069406258647, | |
| "loss": 4.1881, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.3266074344709236, | |
| "grad_norm": 1.0501227378845215, | |
| "learning_rate": 0.00023377864451829642, | |
| "loss": 4.1357, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.3277145783165878, | |
| "grad_norm": 1.0640146732330322, | |
| "learning_rate": 0.0002333355349529403, | |
| "loss": 4.1446, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.3288217221622519, | |
| "grad_norm": 1.0626839399337769, | |
| "learning_rate": 0.00023289137097302415, | |
| "loss": 4.214, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.3299288660079161, | |
| "grad_norm": 1.1319100856781006, | |
| "learning_rate": 0.00023244615819839506, | |
| "loss": 4.1368, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.3310360098535802, | |
| "grad_norm": 1.0692615509033203, | |
| "learning_rate": 0.0002319999022621702, | |
| "loss": 4.133, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.33214315369924435, | |
| "grad_norm": 1.0490069389343262, | |
| "learning_rate": 0.00023155260881066548, | |
| "loss": 4.2116, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.33325029754490854, | |
| "grad_norm": 1.135221242904663, | |
| "learning_rate": 0.0002311042835033241, | |
| "loss": 4.1324, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.33435744139057266, | |
| "grad_norm": 1.108528733253479, | |
| "learning_rate": 0.00023065493201264504, | |
| "loss": 4.1497, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.33546458523623685, | |
| "grad_norm": 1.0444023609161377, | |
| "learning_rate": 0.0002302045600241111, | |
| "loss": 4.2573, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.33657172908190097, | |
| "grad_norm": 1.0995585918426514, | |
| "learning_rate": 0.00022975317323611707, | |
| "loss": 4.1674, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.3376788729275651, | |
| "grad_norm": 1.1554394960403442, | |
| "learning_rate": 0.00022930077735989777, | |
| "loss": 4.161, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.3387860167732293, | |
| "grad_norm": 1.0832637548446655, | |
| "learning_rate": 0.0002288473781194555, | |
| "loss": 4.1295, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.3398931606188934, | |
| "grad_norm": 1.208899974822998, | |
| "learning_rate": 0.0002283929812514879, | |
| "loss": 4.2094, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.34100030446455754, | |
| "grad_norm": 1.114460825920105, | |
| "learning_rate": 0.00022793759250531502, | |
| "loss": 4.0485, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.3421074483102217, | |
| "grad_norm": 1.1555848121643066, | |
| "learning_rate": 0.00022748121764280707, | |
| "loss": 4.0879, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.34321459215588584, | |
| "grad_norm": 1.0038038492202759, | |
| "learning_rate": 0.000227023862438311, | |
| "loss": 4.1866, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.34432173600155, | |
| "grad_norm": 1.0777366161346436, | |
| "learning_rate": 0.0002265655326785778, | |
| "loss": 4.2149, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.34542887984721415, | |
| "grad_norm": 1.0569167137145996, | |
| "learning_rate": 0.0002261062341626892, | |
| "loss": 4.1043, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.3465360236928783, | |
| "grad_norm": 1.0322402715682983, | |
| "learning_rate": 0.00022564597270198408, | |
| "loss": 4.1319, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.34764316753854246, | |
| "grad_norm": 1.0698827505111694, | |
| "learning_rate": 0.0002251847541199852, | |
| "loss": 4.1375, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.3487503113842066, | |
| "grad_norm": 1.1204584836959839, | |
| "learning_rate": 0.00022472258425232554, | |
| "loss": 4.139, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.3498574552298707, | |
| "grad_norm": 1.0228657722473145, | |
| "learning_rate": 0.00022425946894667417, | |
| "loss": 4.259, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.3509645990755349, | |
| "grad_norm": 1.0952000617980957, | |
| "learning_rate": 0.0002237954140626626, | |
| "loss": 4.1238, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.352071742921199, | |
| "grad_norm": 1.086155652999878, | |
| "learning_rate": 0.0002233304254718104, | |
| "loss": 4.0527, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.3531788867668632, | |
| "grad_norm": 1.0240938663482666, | |
| "learning_rate": 0.000222864509057451, | |
| "loss": 4.1267, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.35428603061252734, | |
| "grad_norm": 1.153633713722229, | |
| "learning_rate": 0.00022239767071465728, | |
| "loss": 4.183, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.35539317445819146, | |
| "grad_norm": 1.1948312520980835, | |
| "learning_rate": 0.00022192991635016697, | |
| "loss": 4.1148, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.35650031830385565, | |
| "grad_norm": 1.132027268409729, | |
| "learning_rate": 0.00022146125188230782, | |
| "loss": 4.2102, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.3576074621495198, | |
| "grad_norm": 1.161388635635376, | |
| "learning_rate": 0.00022099168324092288, | |
| "loss": 4.0985, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.3587146059951839, | |
| "grad_norm": 1.1411139965057373, | |
| "learning_rate": 0.00022052121636729525, | |
| "loss": 4.1281, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.3598217498408481, | |
| "grad_norm": 1.1698862314224243, | |
| "learning_rate": 0.0002200498572140732, | |
| "loss": 4.0102, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.3609288936865122, | |
| "grad_norm": 1.1300621032714844, | |
| "learning_rate": 0.00021957761174519458, | |
| "loss": 4.0765, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.3620360375321764, | |
| "grad_norm": 1.3213725090026855, | |
| "learning_rate": 0.00021910448593581155, | |
| "loss": 4.1429, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.3631431813778405, | |
| "grad_norm": 1.0809500217437744, | |
| "learning_rate": 0.0002186304857722148, | |
| "loss": 4.1057, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.36425032522350465, | |
| "grad_norm": 1.1164082288742065, | |
| "learning_rate": 0.00021815561725175802, | |
| "loss": 4.1541, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.3653574690691688, | |
| "grad_norm": 1.1995413303375244, | |
| "learning_rate": 0.00021767988638278182, | |
| "loss": 4.1409, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.36646461291483295, | |
| "grad_norm": 1.2960383892059326, | |
| "learning_rate": 0.00021720329918453786, | |
| "loss": 4.1596, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.3675717567604971, | |
| "grad_norm": 1.015963077545166, | |
| "learning_rate": 0.0002167258616871126, | |
| "loss": 4.1271, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.36867890060616126, | |
| "grad_norm": 1.0631139278411865, | |
| "learning_rate": 0.00021624757993135107, | |
| "loss": 4.0834, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.3697860444518254, | |
| "grad_norm": 1.2399165630340576, | |
| "learning_rate": 0.00021576845996878024, | |
| "loss": 4.0455, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.3708931882974896, | |
| "grad_norm": 1.1509207487106323, | |
| "learning_rate": 0.00021528850786153283, | |
| "loss": 4.093, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.3720003321431537, | |
| "grad_norm": 1.1760320663452148, | |
| "learning_rate": 0.00021480772968227025, | |
| "loss": 4.0607, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.3731074759888178, | |
| "grad_norm": 1.160736083984375, | |
| "learning_rate": 0.00021432613151410593, | |
| "loss": 4.1104, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.374214619834482, | |
| "grad_norm": 1.0643970966339111, | |
| "learning_rate": 0.00021384371945052836, | |
| "loss": 4.0252, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.37532176368014614, | |
| "grad_norm": 1.0459327697753906, | |
| "learning_rate": 0.00021336049959532388, | |
| "loss": 4.0747, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.37642890752581026, | |
| "grad_norm": 1.114637017250061, | |
| "learning_rate": 0.0002128764780624996, | |
| "loss": 4.1002, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.37753605137147445, | |
| "grad_norm": 1.1435799598693848, | |
| "learning_rate": 0.00021239166097620596, | |
| "loss": 4.026, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.3786431952171386, | |
| "grad_norm": 1.1939269304275513, | |
| "learning_rate": 0.00021190605447065917, | |
| "loss": 4.0469, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.37975033906280276, | |
| "grad_norm": 1.1145963668823242, | |
| "learning_rate": 0.00021141966469006386, | |
| "loss": 4.1348, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.3808574829084669, | |
| "grad_norm": 1.1215327978134155, | |
| "learning_rate": 0.00021093249778853494, | |
| "loss": 4.0656, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.381964626754131, | |
| "grad_norm": 1.0755778551101685, | |
| "learning_rate": 0.00021044455993002003, | |
| "loss": 4.0587, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.3830717705997952, | |
| "grad_norm": 1.077532172203064, | |
| "learning_rate": 0.0002099558572882215, | |
| "loss": 4.1518, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.3841789144454593, | |
| "grad_norm": 0.9992897510528564, | |
| "learning_rate": 0.00020946639604651815, | |
| "loss": 4.047, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.3852860582911235, | |
| "grad_norm": 1.1796753406524658, | |
| "learning_rate": 0.00020897618239788704, | |
| "loss": 4.0725, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.38639320213678763, | |
| "grad_norm": 1.09968101978302, | |
| "learning_rate": 0.00020848522254482527, | |
| "loss": 4.1054, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.38750034598245175, | |
| "grad_norm": 1.0939786434173584, | |
| "learning_rate": 0.00020799352269927122, | |
| "loss": 4.1339, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.38860748982811594, | |
| "grad_norm": 1.266184687614441, | |
| "learning_rate": 0.0002075010890825263, | |
| "loss": 4.0547, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.38971463367378006, | |
| "grad_norm": 1.1003026962280273, | |
| "learning_rate": 0.0002070079279251761, | |
| "loss": 4.0736, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.3908217775194442, | |
| "grad_norm": 1.2582813501358032, | |
| "learning_rate": 0.00020651404546701127, | |
| "loss": 4.0929, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.3919289213651084, | |
| "grad_norm": 1.2194759845733643, | |
| "learning_rate": 0.00020601944795694906, | |
| "loss": 4.0803, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.3930360652107725, | |
| "grad_norm": 1.0450422763824463, | |
| "learning_rate": 0.000205524141652954, | |
| "loss": 4.1673, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.3941432090564367, | |
| "grad_norm": 1.1397161483764648, | |
| "learning_rate": 0.00020502813282195846, | |
| "loss": 4.1549, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.3952503529021008, | |
| "grad_norm": 1.1790965795516968, | |
| "learning_rate": 0.000204531427739784, | |
| "loss": 4.0352, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.39635749674776494, | |
| "grad_norm": 1.0304241180419922, | |
| "learning_rate": 0.0002040340326910614, | |
| "loss": 4.0493, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.3974646405934291, | |
| "grad_norm": 1.2007451057434082, | |
| "learning_rate": 0.0002035359539691514, | |
| "loss": 4.0324, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.39857178443909325, | |
| "grad_norm": 1.1203726530075073, | |
| "learning_rate": 0.0002030371978760649, | |
| "loss": 4.0389, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.3996789282847574, | |
| "grad_norm": 1.191495656967163, | |
| "learning_rate": 0.00020253777072238349, | |
| "loss": 4.1485, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.4003432145921559, | |
| "eval_loss": 4.550379276275635, | |
| "eval_runtime": 78.8783, | |
| "eval_samples_per_second": 633.888, | |
| "eval_steps_per_second": 79.236, | |
| "step": 3616 | |
| }, | |
| { | |
| "epoch": 0.40078607213042156, | |
| "grad_norm": 1.0996659994125366, | |
| "learning_rate": 0.0002020376788271793, | |
| "loss": 4.0977, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.4018932159760857, | |
| "grad_norm": 1.1475967168807983, | |
| "learning_rate": 0.00020153692851793535, | |
| "loss": 4.0419, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.40300035982174987, | |
| "grad_norm": 1.133640170097351, | |
| "learning_rate": 0.0002010355261304652, | |
| "loss": 4.0434, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.404107503667414, | |
| "grad_norm": 1.0792150497436523, | |
| "learning_rate": 0.00020053347800883298, | |
| "loss": 3.9828, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.4052146475130781, | |
| "grad_norm": 1.033315896987915, | |
| "learning_rate": 0.0002000307905052731, | |
| "loss": 4.152, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.4063217913587423, | |
| "grad_norm": 1.0719643831253052, | |
| "learning_rate": 0.0001995274699801098, | |
| "loss": 4.0721, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.40742893520440643, | |
| "grad_norm": 1.2880775928497314, | |
| "learning_rate": 0.00019902352280167665, | |
| "loss": 4.0708, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.40853607905007056, | |
| "grad_norm": 1.1417231559753418, | |
| "learning_rate": 0.0001985189553462361, | |
| "loss": 4.0588, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.40964322289573474, | |
| "grad_norm": 1.2042443752288818, | |
| "learning_rate": 0.00019801377399789882, | |
| "loss": 4.061, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.41075036674139886, | |
| "grad_norm": 1.0673474073410034, | |
| "learning_rate": 0.00019750798514854274, | |
| "loss": 4.1215, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.41185751058706305, | |
| "grad_norm": 1.1126561164855957, | |
| "learning_rate": 0.0001970015951977323, | |
| "loss": 4.1259, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.4129646544327272, | |
| "grad_norm": 1.1779885292053223, | |
| "learning_rate": 0.0001964946105526375, | |
| "loss": 4.1334, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.4140717982783913, | |
| "grad_norm": 1.0562344789505005, | |
| "learning_rate": 0.00019598703762795273, | |
| "loss": 4.1164, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.4151789421240555, | |
| "grad_norm": 1.2249536514282227, | |
| "learning_rate": 0.0001954788828458157, | |
| "loss": 4.1002, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.4162860859697196, | |
| "grad_norm": 1.2630491256713867, | |
| "learning_rate": 0.00019497015263572617, | |
| "loss": 4.0244, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.41739322981538374, | |
| "grad_norm": 1.214250087738037, | |
| "learning_rate": 0.00019446085343446462, | |
| "loss": 4.1181, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.4185003736610479, | |
| "grad_norm": 1.111898422241211, | |
| "learning_rate": 0.00019395099168601065, | |
| "loss": 4.0602, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.41960751750671205, | |
| "grad_norm": 1.0913716554641724, | |
| "learning_rate": 0.00019344057384146167, | |
| "loss": 4.1376, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.42071466135237623, | |
| "grad_norm": 1.2286676168441772, | |
| "learning_rate": 0.0001929296063589512, | |
| "loss": 4.0739, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.42182180519804036, | |
| "grad_norm": 1.0938048362731934, | |
| "learning_rate": 0.00019241809570356698, | |
| "loss": 4.0552, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.4229289490437045, | |
| "grad_norm": 1.1894221305847168, | |
| "learning_rate": 0.00019190604834726943, | |
| "loss": 4.0808, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.42403609288936867, | |
| "grad_norm": 1.1462655067443848, | |
| "learning_rate": 0.00019139347076880972, | |
| "loss": 4.078, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.4251432367350328, | |
| "grad_norm": 1.196509838104248, | |
| "learning_rate": 0.00019088036945364754, | |
| "loss": 4.0419, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.426250380580697, | |
| "grad_norm": 1.1123558282852173, | |
| "learning_rate": 0.00019036675089386944, | |
| "loss": 4.0761, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.4273575244263611, | |
| "grad_norm": 1.1525100469589233, | |
| "learning_rate": 0.00018985262158810636, | |
| "loss": 4.0491, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.42846466827202523, | |
| "grad_norm": 1.169020414352417, | |
| "learning_rate": 0.00018933798804145148, | |
| "loss": 4.0859, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.4295718121176894, | |
| "grad_norm": 1.0914548635482788, | |
| "learning_rate": 0.0001888228567653781, | |
| "loss": 4.0633, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.43067895596335354, | |
| "grad_norm": 1.165087342262268, | |
| "learning_rate": 0.000188307234277657, | |
| "loss": 3.9989, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.43178609980901766, | |
| "grad_norm": 1.323397159576416, | |
| "learning_rate": 0.00018779112710227416, | |
| "loss": 4.0071, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.43289324365468185, | |
| "grad_norm": 1.1434069871902466, | |
| "learning_rate": 0.00018727454176934808, | |
| "loss": 4.0965, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.434000387500346, | |
| "grad_norm": 1.0918630361557007, | |
| "learning_rate": 0.00018675748481504728, | |
| "loss": 4.0013, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.43510753134601016, | |
| "grad_norm": 1.0577778816223145, | |
| "learning_rate": 0.0001862399627815074, | |
| "loss": 3.9447, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 0.4362146751916743, | |
| "grad_norm": 1.2388492822647095, | |
| "learning_rate": 0.00018572198221674868, | |
| "loss": 4.0225, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.4373218190373384, | |
| "grad_norm": 1.0684038400650024, | |
| "learning_rate": 0.00018520354967459303, | |
| "loss": 4.0854, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.4384289628830026, | |
| "grad_norm": 1.1683038473129272, | |
| "learning_rate": 0.00018468467171458094, | |
| "loss": 4.0022, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.4395361067286667, | |
| "grad_norm": 1.271133542060852, | |
| "learning_rate": 0.00018416535490188866, | |
| "loss": 3.9859, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 0.44064325057433085, | |
| "grad_norm": 1.12361741065979, | |
| "learning_rate": 0.0001836456058072452, | |
| "loss": 3.9782, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.44175039441999503, | |
| "grad_norm": 1.1791516542434692, | |
| "learning_rate": 0.00018312543100684888, | |
| "loss": 4.0026, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 0.44285753826565916, | |
| "grad_norm": 1.2462031841278076, | |
| "learning_rate": 0.00018260483708228438, | |
| "loss": 3.9726, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.44396468211132334, | |
| "grad_norm": 1.0854827165603638, | |
| "learning_rate": 0.00018208383062043957, | |
| "loss": 4.0443, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 0.44507182595698747, | |
| "grad_norm": 1.2484006881713867, | |
| "learning_rate": 0.0001815624182134218, | |
| "loss": 3.9681, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.4461789698026516, | |
| "grad_norm": 1.2864941358566284, | |
| "learning_rate": 0.00018104060645847476, | |
| "loss": 3.9699, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 0.4472861136483158, | |
| "grad_norm": 1.16134512424469, | |
| "learning_rate": 0.00018051840195789506, | |
| "loss": 4.0391, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.4483932574939799, | |
| "grad_norm": 1.1202607154846191, | |
| "learning_rate": 0.00017999581131894843, | |
| "loss": 4.0649, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.44950040133964403, | |
| "grad_norm": 1.1849491596221924, | |
| "learning_rate": 0.00017947284115378626, | |
| "loss": 4.0279, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.4506075451853082, | |
| "grad_norm": 1.0735421180725098, | |
| "learning_rate": 0.00017894949807936212, | |
| "loss": 4.0537, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 0.45171468903097234, | |
| "grad_norm": 1.1313494443893433, | |
| "learning_rate": 0.0001784257887173477, | |
| "loss": 4.0912, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.4528218328766365, | |
| "grad_norm": 1.3376647233963013, | |
| "learning_rate": 0.0001779017196940493, | |
| "loss": 4.0682, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 0.45392897672230065, | |
| "grad_norm": 1.2430145740509033, | |
| "learning_rate": 0.00017737729764032384, | |
| "loss": 3.9409, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.4550361205679648, | |
| "grad_norm": 1.1707900762557983, | |
| "learning_rate": 0.00017685252919149493, | |
| "loss": 3.9651, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 0.45614326441362896, | |
| "grad_norm": 1.272524356842041, | |
| "learning_rate": 0.0001763274209872691, | |
| "loss": 4.0076, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.4572504082592931, | |
| "grad_norm": 1.0461047887802124, | |
| "learning_rate": 0.00017580197967165167, | |
| "loss": 3.9586, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 0.4583575521049572, | |
| "grad_norm": 1.2831692695617676, | |
| "learning_rate": 0.0001752762118928627, | |
| "loss": 3.9654, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.4594646959506214, | |
| "grad_norm": 1.181222677230835, | |
| "learning_rate": 0.00017475012430325285, | |
| "loss": 3.9983, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.4605718397962855, | |
| "grad_norm": 1.089134931564331, | |
| "learning_rate": 0.00017422372355921923, | |
| "loss": 3.9948, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.4616789836419497, | |
| "grad_norm": 1.172930121421814, | |
| "learning_rate": 0.00017369701632112119, | |
| "loss": 3.9738, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 0.46278612748761383, | |
| "grad_norm": 1.1669162511825562, | |
| "learning_rate": 0.00017317000925319597, | |
| "loss": 4.0157, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.46389327133327796, | |
| "grad_norm": 1.1006979942321777, | |
| "learning_rate": 0.00017264270902347454, | |
| "loss": 4.0526, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 0.46500041517894214, | |
| "grad_norm": 1.261461615562439, | |
| "learning_rate": 0.00017211512230369715, | |
| "loss": 3.9332, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.46610755902460627, | |
| "grad_norm": 1.1611567735671997, | |
| "learning_rate": 0.00017158725576922888, | |
| "loss": 3.9698, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 0.4672147028702704, | |
| "grad_norm": 1.1162567138671875, | |
| "learning_rate": 0.0001710591160989752, | |
| "loss": 3.9986, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 0.4683218467159346, | |
| "grad_norm": 1.2045059204101562, | |
| "learning_rate": 0.00017053070997529737, | |
| "loss": 4.0257, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 0.4694289905615987, | |
| "grad_norm": 1.178435206413269, | |
| "learning_rate": 0.00017000204408392815, | |
| "loss": 4.0011, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 0.4705361344072629, | |
| "grad_norm": 1.2001043558120728, | |
| "learning_rate": 0.00016947312511388695, | |
| "loss": 3.973, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.471643278252927, | |
| "grad_norm": 1.152808666229248, | |
| "learning_rate": 0.00016894395975739534, | |
| "loss": 3.9191, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 0.47275042209859114, | |
| "grad_norm": 1.0874874591827393, | |
| "learning_rate": 0.00016841455470979235, | |
| "loss": 4.0415, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 0.4738575659442553, | |
| "grad_norm": 1.120444655418396, | |
| "learning_rate": 0.00016788491666944966, | |
| "loss": 3.955, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 0.47496470978991945, | |
| "grad_norm": 1.2311396598815918, | |
| "learning_rate": 0.00016735505233768695, | |
| "loss": 4.0224, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 0.47607185363558363, | |
| "grad_norm": 1.1801338195800781, | |
| "learning_rate": 0.0001668249684186871, | |
| "loss": 3.96, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.47717899748124776, | |
| "grad_norm": 1.0829944610595703, | |
| "learning_rate": 0.00016629467161941133, | |
| "loss": 3.9588, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 0.4782861413269119, | |
| "grad_norm": 1.242344617843628, | |
| "learning_rate": 0.00016576416864951443, | |
| "loss": 3.9906, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.47939328517257607, | |
| "grad_norm": 1.1387934684753418, | |
| "learning_rate": 0.00016523346622125966, | |
| "loss": 3.9735, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 0.4805004290182402, | |
| "grad_norm": 1.1672329902648926, | |
| "learning_rate": 0.0001647025710494341, | |
| "loss": 3.9583, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 0.4816075728639043, | |
| "grad_norm": 1.197190761566162, | |
| "learning_rate": 0.0001641714898512634, | |
| "loss": 3.848, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.4827147167095685, | |
| "grad_norm": 1.1592804193496704, | |
| "learning_rate": 0.00016364022934632706, | |
| "loss": 4.0387, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 0.48382186055523263, | |
| "grad_norm": 1.2098631858825684, | |
| "learning_rate": 0.0001631087962564732, | |
| "loss": 3.9723, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 0.4849290044008968, | |
| "grad_norm": 1.1750229597091675, | |
| "learning_rate": 0.00016257719730573359, | |
| "loss": 4.0462, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 0.48603614824656094, | |
| "grad_norm": 1.1587116718292236, | |
| "learning_rate": 0.0001620454392202387, | |
| "loss": 4.015, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 0.48714329209222507, | |
| "grad_norm": 1.0906429290771484, | |
| "learning_rate": 0.00016151352872813239, | |
| "loss": 4.0024, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.48825043593788925, | |
| "grad_norm": 1.1398390531539917, | |
| "learning_rate": 0.00016098147255948683, | |
| "loss": 4.0048, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 0.4893575797835534, | |
| "grad_norm": 1.1513657569885254, | |
| "learning_rate": 0.0001604492774462174, | |
| "loss": 3.9599, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 0.4904647236292175, | |
| "grad_norm": 1.1190105676651, | |
| "learning_rate": 0.0001599169501219975, | |
| "loss": 3.9602, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 0.4915718674748817, | |
| "grad_norm": 1.1455727815628052, | |
| "learning_rate": 0.0001593844973221734, | |
| "loss": 4.0079, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 0.4926790113205458, | |
| "grad_norm": 1.0649514198303223, | |
| "learning_rate": 0.00015885192578367894, | |
| "loss": 3.985, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.49378615516621, | |
| "grad_norm": 1.1714825630187988, | |
| "learning_rate": 0.00015831924224495028, | |
| "loss": 3.933, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 0.4948932990118741, | |
| "grad_norm": 1.1414313316345215, | |
| "learning_rate": 0.00015778645344584078, | |
| "loss": 4.0725, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 0.49600044285753825, | |
| "grad_norm": 1.1334409713745117, | |
| "learning_rate": 0.00015725356612753541, | |
| "loss": 3.9749, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.49710758670320243, | |
| "grad_norm": 1.2870898246765137, | |
| "learning_rate": 0.00015672058703246596, | |
| "loss": 3.9704, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 0.49821473054886656, | |
| "grad_norm": 1.2545424699783325, | |
| "learning_rate": 0.00015618752290422512, | |
| "loss": 4.0397, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.4993218743945307, | |
| "grad_norm": 1.1725001335144043, | |
| "learning_rate": 0.00015565438048748174, | |
| "loss": 4.0018, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 0.5004290182401948, | |
| "grad_norm": 1.138156533241272, | |
| "learning_rate": 0.00015512116652789516, | |
| "loss": 3.9533, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.5004290182401948, | |
| "eval_loss": 4.483375072479248, | |
| "eval_runtime": 78.7402, | |
| "eval_samples_per_second": 635.0, | |
| "eval_steps_per_second": 79.375, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.501536162085859, | |
| "grad_norm": 1.32381272315979, | |
| "learning_rate": 0.00015458788777202986, | |
| "loss": 4.0086, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 0.5026433059315232, | |
| "grad_norm": 1.2566626071929932, | |
| "learning_rate": 0.00015405455096727014, | |
| "loss": 4.0203, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 0.5037504497771873, | |
| "grad_norm": 1.2790629863739014, | |
| "learning_rate": 0.00015352116286173486, | |
| "loss": 3.9438, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.5048575936228514, | |
| "grad_norm": 1.1521967649459839, | |
| "learning_rate": 0.00015298773020419192, | |
| "loss": 3.9641, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.5059647374685156, | |
| "grad_norm": 1.218268871307373, | |
| "learning_rate": 0.000152454259743973, | |
| "loss": 3.9514, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 0.5070718813141798, | |
| "grad_norm": 1.3280975818634033, | |
| "learning_rate": 0.00015192075823088796, | |
| "loss": 3.9655, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 0.5081790251598439, | |
| "grad_norm": 1.1647611856460571, | |
| "learning_rate": 0.00015138723241513968, | |
| "loss": 4.0506, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 0.509286169005508, | |
| "grad_norm": 1.2784322500228882, | |
| "learning_rate": 0.0001508536890472384, | |
| "loss": 4.0122, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.5103933128511722, | |
| "grad_norm": 1.1873068809509277, | |
| "learning_rate": 0.0001503201348779166, | |
| "loss": 4.0547, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 0.5115004566968363, | |
| "grad_norm": 1.1526970863342285, | |
| "learning_rate": 0.00014978657665804323, | |
| "loss": 3.9192, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 0.5126076005425004, | |
| "grad_norm": 1.2054905891418457, | |
| "learning_rate": 0.00014925302113853866, | |
| "loss": 3.9217, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 0.5137147443881647, | |
| "grad_norm": 1.2948694229125977, | |
| "learning_rate": 0.00014871947507028913, | |
| "loss": 3.9556, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.5148218882338288, | |
| "grad_norm": 1.1605454683303833, | |
| "learning_rate": 0.00014818594520406105, | |
| "loss": 3.9476, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.5159290320794929, | |
| "grad_norm": 1.1079860925674438, | |
| "learning_rate": 0.00014765243829041605, | |
| "loss": 3.9542, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 0.517036175925157, | |
| "grad_norm": 1.09770667552948, | |
| "learning_rate": 0.00014711896107962534, | |
| "loss": 3.9398, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 0.5181433197708212, | |
| "grad_norm": 1.4634814262390137, | |
| "learning_rate": 0.00014658552032158412, | |
| "loss": 3.9305, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 0.5192504636164854, | |
| "grad_norm": 1.1489685773849487, | |
| "learning_rate": 0.00014605212276572665, | |
| "loss": 4.114, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 0.5203576074621495, | |
| "grad_norm": 1.0958575010299683, | |
| "learning_rate": 0.0001455187751609403, | |
| "loss": 3.8907, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.5214647513078137, | |
| "grad_norm": 1.1471234560012817, | |
| "learning_rate": 0.0001449854842554806, | |
| "loss": 3.9564, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 0.5225718951534778, | |
| "grad_norm": 1.1879585981369019, | |
| "learning_rate": 0.00014445225679688564, | |
| "loss": 3.9575, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 0.5236790389991419, | |
| "grad_norm": 1.201126217842102, | |
| "learning_rate": 0.00014391909953189068, | |
| "loss": 3.9839, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 0.5247861828448062, | |
| "grad_norm": 1.2734222412109375, | |
| "learning_rate": 0.00014338601920634296, | |
| "loss": 3.9027, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 0.5258933266904703, | |
| "grad_norm": 1.204102873802185, | |
| "learning_rate": 0.0001428530225651161, | |
| "loss": 3.9184, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.5270004705361344, | |
| "grad_norm": 1.1919819116592407, | |
| "learning_rate": 0.00014232011635202505, | |
| "loss": 3.9172, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 0.5281076143817985, | |
| "grad_norm": 1.2805883884429932, | |
| "learning_rate": 0.00014178730730974042, | |
| "loss": 3.8941, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 0.5292147582274627, | |
| "grad_norm": 1.1164813041687012, | |
| "learning_rate": 0.00014125460217970345, | |
| "loss": 3.9599, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 0.5303219020731269, | |
| "grad_norm": 1.2935210466384888, | |
| "learning_rate": 0.00014072200770204072, | |
| "loss": 3.9961, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 0.531429045918791, | |
| "grad_norm": 1.2029496431350708, | |
| "learning_rate": 0.0001401895306154785, | |
| "loss": 3.8823, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.5325361897644552, | |
| "grad_norm": 1.2709356546401978, | |
| "learning_rate": 0.0001396571776572581, | |
| "loss": 4.0073, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 0.5336433336101193, | |
| "grad_norm": 1.220680594444275, | |
| "learning_rate": 0.00013912495556304993, | |
| "loss": 3.8967, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 0.5347504774557834, | |
| "grad_norm": 1.1263922452926636, | |
| "learning_rate": 0.00013859287106686888, | |
| "loss": 3.9997, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 0.5358576213014475, | |
| "grad_norm": 1.2142183780670166, | |
| "learning_rate": 0.0001380609309009887, | |
| "loss": 4.0184, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 0.5369647651471118, | |
| "grad_norm": 1.2874531745910645, | |
| "learning_rate": 0.00013752914179585707, | |
| "loss": 3.9983, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.5380719089927759, | |
| "grad_norm": 1.2166811227798462, | |
| "learning_rate": 0.00013699751048001037, | |
| "loss": 3.9616, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 0.53917905283844, | |
| "grad_norm": 1.2182731628417969, | |
| "learning_rate": 0.0001364660436799884, | |
| "loss": 4.0648, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 0.5402861966841042, | |
| "grad_norm": 1.2633212804794312, | |
| "learning_rate": 0.00013593474812024966, | |
| "loss": 3.8975, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 0.5413933405297683, | |
| "grad_norm": 1.1508208513259888, | |
| "learning_rate": 0.00013540363052308562, | |
| "loss": 3.9876, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 0.5425004843754325, | |
| "grad_norm": 1.1188697814941406, | |
| "learning_rate": 0.0001348726976085364, | |
| "loss": 3.914, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.5436076282210967, | |
| "grad_norm": 1.200383186340332, | |
| "learning_rate": 0.00013434195609430527, | |
| "loss": 3.9302, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 0.5447147720667608, | |
| "grad_norm": 1.240676999092102, | |
| "learning_rate": 0.0001338114126956737, | |
| "loss": 3.9685, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 0.5458219159124249, | |
| "grad_norm": 1.2484670877456665, | |
| "learning_rate": 0.0001332810741254167, | |
| "loss": 3.9845, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 0.546929059758089, | |
| "grad_norm": 1.2180975675582886, | |
| "learning_rate": 0.0001327509470937173, | |
| "loss": 3.9045, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 0.5480362036037533, | |
| "grad_norm": 1.1109565496444702, | |
| "learning_rate": 0.00013222103830808245, | |
| "loss": 3.9727, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.5491433474494174, | |
| "grad_norm": 1.3606740236282349, | |
| "learning_rate": 0.00013169135447325733, | |
| "loss": 3.9007, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.5502504912950815, | |
| "grad_norm": 1.179724097251892, | |
| "learning_rate": 0.00013116190229114105, | |
| "loss": 3.9765, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 0.5513576351407456, | |
| "grad_norm": 1.1777377128601074, | |
| "learning_rate": 0.00013063268846070182, | |
| "loss": 3.9602, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 0.5524647789864098, | |
| "grad_norm": 1.1586858034133911, | |
| "learning_rate": 0.00013010371967789184, | |
| "loss": 3.9964, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 0.5535719228320739, | |
| "grad_norm": 1.1349024772644043, | |
| "learning_rate": 0.00012957500263556304, | |
| "loss": 3.9686, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5546790666777381, | |
| "grad_norm": 1.228308081626892, | |
| "learning_rate": 0.00012904654402338188, | |
| "loss": 3.9584, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 0.5557862105234023, | |
| "grad_norm": 1.1629983186721802, | |
| "learning_rate": 0.00012851835052774536, | |
| "loss": 3.8847, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 0.5568933543690664, | |
| "grad_norm": 1.1682049036026, | |
| "learning_rate": 0.00012799042883169574, | |
| "loss": 3.9101, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 0.5580004982147305, | |
| "grad_norm": 1.1060129404067993, | |
| "learning_rate": 0.00012746278561483643, | |
| "loss": 3.9531, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 0.5591076420603946, | |
| "grad_norm": 1.0978598594665527, | |
| "learning_rate": 0.00012693542755324745, | |
| "loss": 3.9229, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.5602147859060589, | |
| "grad_norm": 1.1360820531845093, | |
| "learning_rate": 0.0001264083613194006, | |
| "loss": 3.8605, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 0.561321929751723, | |
| "grad_norm": 1.2562825679779053, | |
| "learning_rate": 0.00012588159358207552, | |
| "loss": 4.0045, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 0.5624290735973871, | |
| "grad_norm": 1.1977635622024536, | |
| "learning_rate": 0.00012535513100627493, | |
| "loss": 3.9795, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 0.5635362174430513, | |
| "grad_norm": 1.2226331233978271, | |
| "learning_rate": 0.0001248289802531406, | |
| "loss": 3.9787, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 0.5646433612887154, | |
| "grad_norm": 1.8836126327514648, | |
| "learning_rate": 0.00012430314797986868, | |
| "loss": 3.9252, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.5657505051343796, | |
| "grad_norm": 1.071244478225708, | |
| "learning_rate": 0.00012377764083962593, | |
| "loss": 4.0156, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 0.5668576489800438, | |
| "grad_norm": 1.1860228776931763, | |
| "learning_rate": 0.0001232524654814653, | |
| "loss": 3.9259, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 0.5679647928257079, | |
| "grad_norm": 1.1561336517333984, | |
| "learning_rate": 0.0001227276285502417, | |
| "loss": 3.9085, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 0.569071936671372, | |
| "grad_norm": 1.231423258781433, | |
| "learning_rate": 0.0001222031366865282, | |
| "loss": 3.914, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 0.5701790805170361, | |
| "grad_norm": 1.1998087167739868, | |
| "learning_rate": 0.00012167899652653166, | |
| "loss": 4.0063, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.5712862243627004, | |
| "grad_norm": 1.2834868431091309, | |
| "learning_rate": 0.00012115521470200909, | |
| "loss": 3.8973, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 0.5723933682083645, | |
| "grad_norm": 1.121349811553955, | |
| "learning_rate": 0.00012063179784018367, | |
| "loss": 3.9243, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 0.5735005120540286, | |
| "grad_norm": 1.151056170463562, | |
| "learning_rate": 0.00012010875256366065, | |
| "loss": 3.9357, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 0.5746076558996928, | |
| "grad_norm": 1.1169615983963013, | |
| "learning_rate": 0.00011958608549034393, | |
| "loss": 3.9047, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 0.5757147997453569, | |
| "grad_norm": 1.368613362312317, | |
| "learning_rate": 0.00011906380323335202, | |
| "loss": 3.9465, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.576821943591021, | |
| "grad_norm": 1.2103781700134277, | |
| "learning_rate": 0.00011854191240093454, | |
| "loss": 3.882, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 0.5779290874366853, | |
| "grad_norm": 1.2073278427124023, | |
| "learning_rate": 0.00011802041959638852, | |
| "loss": 3.954, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 0.5790362312823494, | |
| "grad_norm": 1.0920655727386475, | |
| "learning_rate": 0.00011749933141797491, | |
| "loss": 3.9638, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 0.5801433751280135, | |
| "grad_norm": 1.2782762050628662, | |
| "learning_rate": 0.00011697865445883518, | |
| "loss": 3.9574, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 0.5812505189736776, | |
| "grad_norm": 1.1822972297668457, | |
| "learning_rate": 0.0001164583953069075, | |
| "loss": 3.9676, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.5823576628193418, | |
| "grad_norm": 1.26366126537323, | |
| "learning_rate": 0.00011593856054484402, | |
| "loss": 3.9439, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 0.583464806665006, | |
| "grad_norm": 1.1330398321151733, | |
| "learning_rate": 0.00011541915674992698, | |
| "loss": 3.9556, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 0.5845719505106701, | |
| "grad_norm": 1.1555025577545166, | |
| "learning_rate": 0.00011490019049398592, | |
| "loss": 3.9128, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.5856790943563343, | |
| "grad_norm": 1.2722302675247192, | |
| "learning_rate": 0.00011438166834331415, | |
| "loss": 3.869, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 0.5867862382019984, | |
| "grad_norm": 1.2975918054580688, | |
| "learning_rate": 0.00011386359685858608, | |
| "loss": 3.9495, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.5878933820476625, | |
| "grad_norm": 1.1809889078140259, | |
| "learning_rate": 0.00011334598259477403, | |
| "loss": 4.0583, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 0.5890005258933267, | |
| "grad_norm": 1.221969485282898, | |
| "learning_rate": 0.00011282883210106502, | |
| "loss": 3.9248, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 0.5901076697389909, | |
| "grad_norm": 1.307375431060791, | |
| "learning_rate": 0.00011231215192077837, | |
| "loss": 3.9219, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 0.591214813584655, | |
| "grad_norm": 1.206119179725647, | |
| "learning_rate": 0.00011179594859128261, | |
| "loss": 3.884, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 0.5923219574303191, | |
| "grad_norm": 1.241197109222412, | |
| "learning_rate": 0.00011128022864391282, | |
| "loss": 3.934, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.5934291012759833, | |
| "grad_norm": 1.1164604425430298, | |
| "learning_rate": 0.00011076499860388822, | |
| "loss": 3.9202, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 0.5945362451216474, | |
| "grad_norm": 1.2918251752853394, | |
| "learning_rate": 0.00011025026499022907, | |
| "loss": 4.0321, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 0.5956433889673116, | |
| "grad_norm": 1.3590047359466553, | |
| "learning_rate": 0.00010973603431567483, | |
| "loss": 3.9664, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 0.5967505328129757, | |
| "grad_norm": 1.290653109550476, | |
| "learning_rate": 0.00010922231308660126, | |
| "loss": 3.8759, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 0.5978576766586399, | |
| "grad_norm": 1.2495226860046387, | |
| "learning_rate": 0.00010870910780293843, | |
| "loss": 3.9344, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.598964820504304, | |
| "grad_norm": 1.2402859926223755, | |
| "learning_rate": 0.00010819642495808816, | |
| "loss": 3.8884, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 0.6000719643499681, | |
| "grad_norm": 1.1546714305877686, | |
| "learning_rate": 0.00010768427103884222, | |
| "loss": 3.8935, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 0.6005148218882338, | |
| "eval_loss": 4.42229700088501, | |
| "eval_runtime": 79.4049, | |
| "eval_samples_per_second": 629.684, | |
| "eval_steps_per_second": 78.711, | |
| "step": 5424 | |
| }, | |
| { | |
| "epoch": 0.6011791081956324, | |
| "grad_norm": 1.1268699169158936, | |
| "learning_rate": 0.00010717265252530005, | |
| "loss": 3.9351, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 0.6022862520412965, | |
| "grad_norm": 1.2566118240356445, | |
| "learning_rate": 0.00010666157589078666, | |
| "loss": 3.825, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 0.6033933958869606, | |
| "grad_norm": 1.4127118587493896, | |
| "learning_rate": 0.00010615104760177099, | |
| "loss": 3.9359, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.6045005397326247, | |
| "grad_norm": 1.3040101528167725, | |
| "learning_rate": 0.00010564107411778389, | |
| "loss": 3.9789, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 0.6056076835782889, | |
| "grad_norm": 1.3017617464065552, | |
| "learning_rate": 0.00010513166189133652, | |
| "loss": 3.906, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 0.6067148274239531, | |
| "grad_norm": 1.2195632457733154, | |
| "learning_rate": 0.00010462281736783863, | |
| "loss": 3.8786, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 0.6078219712696172, | |
| "grad_norm": 1.244666337966919, | |
| "learning_rate": 0.00010411454698551695, | |
| "loss": 3.979, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 0.6089291151152814, | |
| "grad_norm": 1.1431467533111572, | |
| "learning_rate": 0.00010360685717533399, | |
| "loss": 3.8793, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6100362589609455, | |
| "grad_norm": 1.1680443286895752, | |
| "learning_rate": 0.00010309975436090626, | |
| "loss": 3.9007, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 0.6111434028066096, | |
| "grad_norm": 1.214645504951477, | |
| "learning_rate": 0.00010259324495842343, | |
| "loss": 3.8913, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 0.6122505466522737, | |
| "grad_norm": 1.15861177444458, | |
| "learning_rate": 0.00010208733537656674, | |
| "loss": 3.9379, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 0.613357690497938, | |
| "grad_norm": 1.3604415655136108, | |
| "learning_rate": 0.00010158203201642832, | |
| "loss": 3.9016, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 0.6144648343436021, | |
| "grad_norm": 1.1846257448196411, | |
| "learning_rate": 0.0001010773412714299, | |
| "loss": 3.8572, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.6155719781892662, | |
| "grad_norm": 1.1964309215545654, | |
| "learning_rate": 0.00010057326952724198, | |
| "loss": 3.9061, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 0.6166791220349304, | |
| "grad_norm": 1.3186078071594238, | |
| "learning_rate": 0.00010006982316170313, | |
| "loss": 3.9322, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 0.6177862658805945, | |
| "grad_norm": 1.276946783065796, | |
| "learning_rate": 9.956700854473915e-05, | |
| "loss": 3.9155, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 0.6188934097262587, | |
| "grad_norm": 1.2113755941390991, | |
| "learning_rate": 9.906483203828267e-05, | |
| "loss": 3.9004, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 0.6200005535719229, | |
| "grad_norm": 1.154305338859558, | |
| "learning_rate": 9.856329999619234e-05, | |
| "loss": 3.9567, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.621107697417587, | |
| "grad_norm": 1.185226559638977, | |
| "learning_rate": 9.806241876417283e-05, | |
| "loss": 3.8733, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 0.6222148412632511, | |
| "grad_norm": 1.277397632598877, | |
| "learning_rate": 9.756219467969439e-05, | |
| "loss": 3.8314, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 0.6233219851089152, | |
| "grad_norm": 1.2497365474700928, | |
| "learning_rate": 9.706263407191233e-05, | |
| "loss": 3.8579, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 0.6244291289545795, | |
| "grad_norm": 1.2919963598251343, | |
| "learning_rate": 9.656374326158751e-05, | |
| "loss": 3.8682, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 0.6255362728002436, | |
| "grad_norm": 1.3063186407089233, | |
| "learning_rate": 9.606552856100587e-05, | |
| "loss": 3.8481, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.6266434166459077, | |
| "grad_norm": 1.2712632417678833, | |
| "learning_rate": 9.556799627389895e-05, | |
| "loss": 3.8627, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 0.6277505604915719, | |
| "grad_norm": 1.2737300395965576, | |
| "learning_rate": 9.507115269536393e-05, | |
| "loss": 3.8744, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 0.628857704337236, | |
| "grad_norm": 1.3533867597579956, | |
| "learning_rate": 9.457500411178387e-05, | |
| "loss": 3.8352, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 0.6299648481829002, | |
| "grad_norm": 1.2895146608352661, | |
| "learning_rate": 9.407955680074851e-05, | |
| "loss": 3.964, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 0.6310719920285643, | |
| "grad_norm": 1.2696341276168823, | |
| "learning_rate": 9.358481703097444e-05, | |
| "loss": 3.8574, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.6321791358742285, | |
| "grad_norm": 1.1142823696136475, | |
| "learning_rate": 9.309079106222617e-05, | |
| "loss": 3.8189, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 0.6332862797198926, | |
| "grad_norm": 1.1882737874984741, | |
| "learning_rate": 9.259748514523653e-05, | |
| "loss": 3.9385, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 0.6343934235655567, | |
| "grad_norm": 1.3217726945877075, | |
| "learning_rate": 9.210490552162802e-05, | |
| "loss": 3.8771, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 0.6355005674112209, | |
| "grad_norm": 1.3164798021316528, | |
| "learning_rate": 9.16130584238335e-05, | |
| "loss": 3.9765, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 0.6366077112568851, | |
| "grad_norm": 1.3711764812469482, | |
| "learning_rate": 9.112195007501741e-05, | |
| "loss": 3.9088, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.6377148551025492, | |
| "grad_norm": 1.145696759223938, | |
| "learning_rate": 9.063158668899718e-05, | |
| "loss": 3.8768, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.6388219989482133, | |
| "grad_norm": 1.249787449836731, | |
| "learning_rate": 9.014197447016429e-05, | |
| "loss": 3.9084, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 0.6399291427938775, | |
| "grad_norm": 1.3100545406341553, | |
| "learning_rate": 8.965311961340625e-05, | |
| "loss": 3.9052, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 0.6410362866395416, | |
| "grad_norm": 1.1464250087738037, | |
| "learning_rate": 8.916502830402766e-05, | |
| "loss": 3.9723, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 0.6421434304852058, | |
| "grad_norm": 1.1405073404312134, | |
| "learning_rate": 8.867770671767245e-05, | |
| "loss": 3.8943, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.64325057433087, | |
| "grad_norm": 1.4155075550079346, | |
| "learning_rate": 8.819116102024543e-05, | |
| "loss": 3.8744, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 0.6443577181765341, | |
| "grad_norm": 1.218760371208191, | |
| "learning_rate": 8.770539736783434e-05, | |
| "loss": 3.9435, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 0.6454648620221982, | |
| "grad_norm": 1.218724250793457, | |
| "learning_rate": 8.722042190663202e-05, | |
| "loss": 3.9357, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 0.6465720058678623, | |
| "grad_norm": 1.2589612007141113, | |
| "learning_rate": 8.673624077285862e-05, | |
| "loss": 3.8755, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 0.6476791497135266, | |
| "grad_norm": 1.267483115196228, | |
| "learning_rate": 8.625286009268392e-05, | |
| "loss": 4.0072, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.6487862935591907, | |
| "grad_norm": 1.2953178882598877, | |
| "learning_rate": 8.577028598214987e-05, | |
| "loss": 3.8651, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 0.6498934374048548, | |
| "grad_norm": 1.1896841526031494, | |
| "learning_rate": 8.528852454709317e-05, | |
| "loss": 3.9083, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 0.651000581250519, | |
| "grad_norm": 1.2403925657272339, | |
| "learning_rate": 8.480758188306806e-05, | |
| "loss": 3.9844, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 0.6521077250961831, | |
| "grad_norm": 1.2764791250228882, | |
| "learning_rate": 8.432746407526912e-05, | |
| "loss": 3.8907, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 0.6532148689418472, | |
| "grad_norm": 1.2560244798660278, | |
| "learning_rate": 8.384817719845437e-05, | |
| "loss": 3.8562, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.6543220127875115, | |
| "grad_norm": 1.242156982421875, | |
| "learning_rate": 8.336972731686829e-05, | |
| "loss": 3.9085, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 0.6554291566331756, | |
| "grad_norm": 1.2683637142181396, | |
| "learning_rate": 8.289212048416525e-05, | |
| "loss": 3.8655, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 0.6565363004788397, | |
| "grad_norm": 1.1777539253234863, | |
| "learning_rate": 8.241536274333274e-05, | |
| "loss": 3.9407, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 0.6576434443245038, | |
| "grad_norm": 1.3885868787765503, | |
| "learning_rate": 8.193946012661506e-05, | |
| "loss": 3.8508, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 0.658750588170168, | |
| "grad_norm": 1.146349549293518, | |
| "learning_rate": 8.146441865543689e-05, | |
| "loss": 3.9098, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.6598577320158322, | |
| "grad_norm": 1.2075341939926147, | |
| "learning_rate": 8.099024434032717e-05, | |
| "loss": 3.8844, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 0.6609648758614963, | |
| "grad_norm": 1.2260401248931885, | |
| "learning_rate": 8.051694318084302e-05, | |
| "loss": 3.9025, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 0.6620720197071605, | |
| "grad_norm": 1.1611764430999756, | |
| "learning_rate": 8.004452116549383e-05, | |
| "loss": 3.922, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 0.6631791635528246, | |
| "grad_norm": 1.2772732973098755, | |
| "learning_rate": 7.95729842716655e-05, | |
| "loss": 3.8677, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 0.6642863073984887, | |
| "grad_norm": 1.2778719663619995, | |
| "learning_rate": 7.910233846554479e-05, | |
| "loss": 3.831, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.665393451244153, | |
| "grad_norm": 1.345259189605713, | |
| "learning_rate": 7.863258970204388e-05, | |
| "loss": 3.837, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 0.6665005950898171, | |
| "grad_norm": 1.194235920906067, | |
| "learning_rate": 7.816374392472497e-05, | |
| "loss": 3.8996, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 0.6676077389354812, | |
| "grad_norm": 1.2076237201690674, | |
| "learning_rate": 7.76958070657251e-05, | |
| "loss": 3.8747, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 0.6687148827811453, | |
| "grad_norm": 1.2254306077957153, | |
| "learning_rate": 7.722878504568109e-05, | |
| "loss": 3.8166, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 0.6698220266268095, | |
| "grad_norm": 1.2513906955718994, | |
| "learning_rate": 7.676268377365466e-05, | |
| "loss": 3.9011, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.6709291704724737, | |
| "grad_norm": 1.3574330806732178, | |
| "learning_rate": 7.629750914705761e-05, | |
| "loss": 3.8706, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 0.6720363143181378, | |
| "grad_norm": 1.2563990354537964, | |
| "learning_rate": 7.583326705157721e-05, | |
| "loss": 3.9134, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 0.6731434581638019, | |
| "grad_norm": 1.2403935194015503, | |
| "learning_rate": 7.53699633611018e-05, | |
| "loss": 3.8881, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 0.6742506020094661, | |
| "grad_norm": 1.2332943677902222, | |
| "learning_rate": 7.49076039376464e-05, | |
| "loss": 3.8897, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 0.6753577458551302, | |
| "grad_norm": 1.292163372039795, | |
| "learning_rate": 7.444619463127853e-05, | |
| "loss": 4.0093, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.6764648897007943, | |
| "grad_norm": 1.3399368524551392, | |
| "learning_rate": 7.398574128004426e-05, | |
| "loss": 3.8901, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 0.6775720335464586, | |
| "grad_norm": 1.2575509548187256, | |
| "learning_rate": 7.352624970989424e-05, | |
| "loss": 3.8975, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 0.6786791773921227, | |
| "grad_norm": 1.2246592044830322, | |
| "learning_rate": 7.306772573461017e-05, | |
| "loss": 3.8286, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 0.6797863212377868, | |
| "grad_norm": 1.3602088689804077, | |
| "learning_rate": 7.261017515573098e-05, | |
| "loss": 3.8489, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 0.6808934650834509, | |
| "grad_norm": 1.2406338453292847, | |
| "learning_rate": 7.215360376247964e-05, | |
| "loss": 3.9004, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.6820006089291151, | |
| "grad_norm": 1.306460976600647, | |
| "learning_rate": 7.169801733168987e-05, | |
| "loss": 3.954, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 0.6831077527747793, | |
| "grad_norm": 1.1845752000808716, | |
| "learning_rate": 7.124342162773293e-05, | |
| "loss": 3.909, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 0.6842148966204434, | |
| "grad_norm": 1.398349642753601, | |
| "learning_rate": 7.07898224024448e-05, | |
| "loss": 3.9076, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 0.6853220404661076, | |
| "grad_norm": 1.201772689819336, | |
| "learning_rate": 7.033722539505341e-05, | |
| "loss": 3.8861, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 0.6864291843117717, | |
| "grad_norm": 1.2474838495254517, | |
| "learning_rate": 6.988563633210597e-05, | |
| "loss": 3.8107, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.6875363281574358, | |
| "grad_norm": 1.2475601434707642, | |
| "learning_rate": 6.94350609273965e-05, | |
| "loss": 3.8719, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 0.6886434720031, | |
| "grad_norm": 1.2588428258895874, | |
| "learning_rate": 6.89855048818936e-05, | |
| "loss": 3.9381, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 0.6897506158487642, | |
| "grad_norm": 1.1882851123809814, | |
| "learning_rate": 6.853697388366826e-05, | |
| "loss": 3.9121, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 0.6908577596944283, | |
| "grad_norm": 1.2219841480255127, | |
| "learning_rate": 6.808947360782194e-05, | |
| "loss": 3.9465, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 0.6919649035400924, | |
| "grad_norm": 1.1857879161834717, | |
| "learning_rate": 6.764300971641474e-05, | |
| "loss": 3.9217, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.6930720473857566, | |
| "grad_norm": 1.4175400733947754, | |
| "learning_rate": 6.719758785839373e-05, | |
| "loss": 3.8987, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 0.6941791912314207, | |
| "grad_norm": 1.240195393562317, | |
| "learning_rate": 6.675321366952153e-05, | |
| "loss": 3.9209, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 0.6952863350770849, | |
| "grad_norm": 1.2016427516937256, | |
| "learning_rate": 6.6309892772305e-05, | |
| "loss": 3.9553, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 0.696393478922749, | |
| "grad_norm": 1.2848678827285767, | |
| "learning_rate": 6.5867630775924e-05, | |
| "loss": 3.8291, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 0.6975006227684132, | |
| "grad_norm": 1.3034757375717163, | |
| "learning_rate": 6.542643327616062e-05, | |
| "loss": 3.939, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.6986077666140773, | |
| "grad_norm": 1.3155343532562256, | |
| "learning_rate": 6.498630585532813e-05, | |
| "loss": 3.8159, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 0.6997149104597414, | |
| "grad_norm": 1.1082866191864014, | |
| "learning_rate": 6.454725408220057e-05, | |
| "loss": 3.8826, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 0.7006006255362728, | |
| "eval_loss": 4.3824357986450195, | |
| "eval_runtime": 78.7022, | |
| "eval_samples_per_second": 635.306, | |
| "eval_steps_per_second": 79.413, | |
| "step": 6328 | |
| }, | |
| { | |
| "epoch": 0.7008220543054057, | |
| "grad_norm": 1.290253758430481, | |
| "learning_rate": 6.410928351194216e-05, | |
| "loss": 3.7839, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 0.7019291981510698, | |
| "grad_norm": 1.3300949335098267, | |
| "learning_rate": 6.367239968603704e-05, | |
| "loss": 3.8714, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 0.7030363419967339, | |
| "grad_norm": 1.41908860206604, | |
| "learning_rate": 6.323660813221919e-05, | |
| "loss": 3.8855, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.704143485842398, | |
| "grad_norm": 1.2172080278396606, | |
| "learning_rate": 6.280191436440243e-05, | |
| "loss": 3.9296, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 0.7052506296880622, | |
| "grad_norm": 1.3784725666046143, | |
| "learning_rate": 6.236832388261071e-05, | |
| "loss": 3.8657, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 0.7063577735337264, | |
| "grad_norm": 1.3897366523742676, | |
| "learning_rate": 6.19358421729085e-05, | |
| "loss": 3.8988, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 0.7074649173793905, | |
| "grad_norm": 1.227491855621338, | |
| "learning_rate": 6.150447470733139e-05, | |
| "loss": 3.8212, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 0.7085720612250547, | |
| "grad_norm": 1.3251690864562988, | |
| "learning_rate": 6.10742269438168e-05, | |
| "loss": 3.8815, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.7096792050707188, | |
| "grad_norm": 1.3054172992706299, | |
| "learning_rate": 6.064510432613499e-05, | |
| "loss": 3.8909, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 0.7107863489163829, | |
| "grad_norm": 1.3053603172302246, | |
| "learning_rate": 6.0217112283820164e-05, | |
| "loss": 3.9107, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 0.7118934927620472, | |
| "grad_norm": 1.2271584272384644, | |
| "learning_rate": 5.979025623210173e-05, | |
| "loss": 3.8813, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 0.7130006366077113, | |
| "grad_norm": 1.3368256092071533, | |
| "learning_rate": 5.936454157183582e-05, | |
| "loss": 3.8803, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 0.7141077804533754, | |
| "grad_norm": 1.304469347000122, | |
| "learning_rate": 5.8939973689436986e-05, | |
| "loss": 3.8414, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.7152149242990395, | |
| "grad_norm": 1.2510364055633545, | |
| "learning_rate": 5.851655795680997e-05, | |
| "loss": 3.8427, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 0.7163220681447037, | |
| "grad_norm": 1.2723710536956787, | |
| "learning_rate": 5.809429973128188e-05, | |
| "loss": 3.7954, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 0.7174292119903678, | |
| "grad_norm": 1.336011290550232, | |
| "learning_rate": 5.767320435553402e-05, | |
| "loss": 3.8473, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 0.718536355836032, | |
| "grad_norm": 1.4587254524230957, | |
| "learning_rate": 5.725327715753496e-05, | |
| "loss": 3.7964, | |
| "step": 6490 | |
| }, | |
| { | |
| "epoch": 0.7196434996816962, | |
| "grad_norm": 1.2736002206802368, | |
| "learning_rate": 5.683452345047249e-05, | |
| "loss": 3.8949, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7207506435273603, | |
| "grad_norm": 1.4224684238433838, | |
| "learning_rate": 5.641694853268668e-05, | |
| "loss": 3.8171, | |
| "step": 6510 | |
| }, | |
| { | |
| "epoch": 0.7218577873730244, | |
| "grad_norm": 1.2361469268798828, | |
| "learning_rate": 5.600055768760286e-05, | |
| "loss": 3.8774, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 0.7229649312186885, | |
| "grad_norm": 1.3639686107635498, | |
| "learning_rate": 5.5585356183664645e-05, | |
| "loss": 3.8847, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 0.7240720750643528, | |
| "grad_norm": 1.134554386138916, | |
| "learning_rate": 5.51713492742674e-05, | |
| "loss": 3.8157, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 0.7251792189100169, | |
| "grad_norm": 1.2591432332992554, | |
| "learning_rate": 5.475854219769166e-05, | |
| "loss": 3.9335, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.726286362755681, | |
| "grad_norm": 1.3043485879898071, | |
| "learning_rate": 5.434694017703697e-05, | |
| "loss": 3.9075, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 0.7273935066013452, | |
| "grad_norm": 1.4282703399658203, | |
| "learning_rate": 5.3936548420155697e-05, | |
| "loss": 3.9239, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 0.7285006504470093, | |
| "grad_norm": 1.2308979034423828, | |
| "learning_rate": 5.352737211958718e-05, | |
| "loss": 3.8555, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 0.7296077942926735, | |
| "grad_norm": 1.2003191709518433, | |
| "learning_rate": 5.311941645249202e-05, | |
| "loss": 3.841, | |
| "step": 6590 | |
| }, | |
| { | |
| "epoch": 0.7307149381383377, | |
| "grad_norm": 1.2548750638961792, | |
| "learning_rate": 5.27126865805866e-05, | |
| "loss": 3.7766, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.7318220819840018, | |
| "grad_norm": 1.3334059715270996, | |
| "learning_rate": 5.2307187650077754e-05, | |
| "loss": 3.8145, | |
| "step": 6610 | |
| }, | |
| { | |
| "epoch": 0.7329292258296659, | |
| "grad_norm": 1.353700876235962, | |
| "learning_rate": 5.190292479159762e-05, | |
| "loss": 3.8613, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 0.73403636967533, | |
| "grad_norm": 1.4207923412322998, | |
| "learning_rate": 5.1499903120138836e-05, | |
| "loss": 3.8978, | |
| "step": 6630 | |
| }, | |
| { | |
| "epoch": 0.7351435135209942, | |
| "grad_norm": 1.3334434032440186, | |
| "learning_rate": 5.109812773498967e-05, | |
| "loss": 3.7772, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 0.7362506573666584, | |
| "grad_norm": 1.342227816581726, | |
| "learning_rate": 5.069760371966961e-05, | |
| "loss": 3.9639, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.7373578012123225, | |
| "grad_norm": 1.3207051753997803, | |
| "learning_rate": 5.029833614186508e-05, | |
| "loss": 3.8492, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 0.7384649450579867, | |
| "grad_norm": 1.3107365369796753, | |
| "learning_rate": 4.9900330053365016e-05, | |
| "loss": 3.7909, | |
| "step": 6670 | |
| }, | |
| { | |
| "epoch": 0.7395720889036508, | |
| "grad_norm": 1.295186161994934, | |
| "learning_rate": 4.9503590489997526e-05, | |
| "loss": 3.8721, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 0.7406792327493149, | |
| "grad_norm": 1.28228759765625, | |
| "learning_rate": 4.9108122471565575e-05, | |
| "loss": 3.8568, | |
| "step": 6690 | |
| }, | |
| { | |
| "epoch": 0.7417863765949791, | |
| "grad_norm": 1.1960582733154297, | |
| "learning_rate": 4.871393100178383e-05, | |
| "loss": 3.7653, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.7428935204406433, | |
| "grad_norm": 1.1884863376617432, | |
| "learning_rate": 4.8321021068215213e-05, | |
| "loss": 3.7915, | |
| "step": 6710 | |
| }, | |
| { | |
| "epoch": 0.7440006642863074, | |
| "grad_norm": 1.2912013530731201, | |
| "learning_rate": 4.792939764220783e-05, | |
| "loss": 3.8594, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 0.7451078081319715, | |
| "grad_norm": 1.238106608390808, | |
| "learning_rate": 4.7539065678832124e-05, | |
| "loss": 3.8881, | |
| "step": 6730 | |
| }, | |
| { | |
| "epoch": 0.7462149519776357, | |
| "grad_norm": 1.2224785089492798, | |
| "learning_rate": 4.715003011681802e-05, | |
| "loss": 3.834, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 0.7473220958232999, | |
| "grad_norm": 1.28524649143219, | |
| "learning_rate": 4.676229587849266e-05, | |
| "loss": 3.8105, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.748429239668964, | |
| "grad_norm": 1.2872871160507202, | |
| "learning_rate": 4.6375867869717907e-05, | |
| "loss": 3.8112, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 0.7495363835146281, | |
| "grad_norm": 1.2735174894332886, | |
| "learning_rate": 4.599075097982845e-05, | |
| "loss": 3.9655, | |
| "step": 6770 | |
| }, | |
| { | |
| "epoch": 0.7506435273602923, | |
| "grad_norm": 1.3700590133666992, | |
| "learning_rate": 4.5606950081569844e-05, | |
| "loss": 3.808, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 0.7517506712059564, | |
| "grad_norm": 1.2019731998443604, | |
| "learning_rate": 4.522447003103673e-05, | |
| "loss": 3.7932, | |
| "step": 6790 | |
| }, | |
| { | |
| "epoch": 0.7528578150516205, | |
| "grad_norm": 1.3262470960617065, | |
| "learning_rate": 4.4843315667611805e-05, | |
| "loss": 3.8936, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.7539649588972848, | |
| "grad_norm": 1.2831884622573853, | |
| "learning_rate": 4.446349181390415e-05, | |
| "loss": 3.8978, | |
| "step": 6810 | |
| }, | |
| { | |
| "epoch": 0.7550721027429489, | |
| "grad_norm": 1.2284660339355469, | |
| "learning_rate": 4.408500327568839e-05, | |
| "loss": 3.8265, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 0.756179246588613, | |
| "grad_norm": 1.3516490459442139, | |
| "learning_rate": 4.370785484184392e-05, | |
| "loss": 3.8313, | |
| "step": 6830 | |
| }, | |
| { | |
| "epoch": 0.7572863904342771, | |
| "grad_norm": 1.3729974031448364, | |
| "learning_rate": 4.3332051284294294e-05, | |
| "loss": 3.8592, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 0.7583935342799413, | |
| "grad_norm": 1.2988899946212769, | |
| "learning_rate": 4.295759735794677e-05, | |
| "loss": 3.9323, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.7595006781256055, | |
| "grad_norm": 1.3927830457687378, | |
| "learning_rate": 4.258449780063224e-05, | |
| "loss": 3.8242, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 0.7606078219712696, | |
| "grad_norm": 1.3057420253753662, | |
| "learning_rate": 4.2212757333045283e-05, | |
| "loss": 3.7698, | |
| "step": 6870 | |
| }, | |
| { | |
| "epoch": 0.7617149658169338, | |
| "grad_norm": 1.2033535242080688, | |
| "learning_rate": 4.184238065868437e-05, | |
| "loss": 3.8656, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 0.7628221096625979, | |
| "grad_norm": 1.2414050102233887, | |
| "learning_rate": 4.147337246379235e-05, | |
| "loss": 3.8633, | |
| "step": 6890 | |
| }, | |
| { | |
| "epoch": 0.763929253508262, | |
| "grad_norm": 1.2692502737045288, | |
| "learning_rate": 4.1105737417297344e-05, | |
| "loss": 3.8992, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.7650363973539263, | |
| "grad_norm": 1.4052729606628418, | |
| "learning_rate": 4.0739480170753206e-05, | |
| "loss": 3.847, | |
| "step": 6910 | |
| }, | |
| { | |
| "epoch": 0.7661435411995904, | |
| "grad_norm": 1.373024821281433, | |
| "learning_rate": 4.0374605358281376e-05, | |
| "loss": 3.865, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 0.7672506850452545, | |
| "grad_norm": 1.307616114616394, | |
| "learning_rate": 4.00111175965116e-05, | |
| "loss": 3.9321, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 0.7683578288909186, | |
| "grad_norm": 1.4080592393875122, | |
| "learning_rate": 3.964902148452383e-05, | |
| "loss": 3.8153, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 0.7694649727365828, | |
| "grad_norm": 1.322647213935852, | |
| "learning_rate": 3.9288321603789986e-05, | |
| "loss": 3.8073, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.770572116582247, | |
| "grad_norm": 1.4065742492675781, | |
| "learning_rate": 3.892902251811595e-05, | |
| "loss": 3.9123, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 0.7716792604279111, | |
| "grad_norm": 1.2831237316131592, | |
| "learning_rate": 3.857112877358391e-05, | |
| "loss": 3.8878, | |
| "step": 6970 | |
| }, | |
| { | |
| "epoch": 0.7727864042735753, | |
| "grad_norm": 1.2117375135421753, | |
| "learning_rate": 3.8214644898494636e-05, | |
| "loss": 3.9187, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 0.7738935481192394, | |
| "grad_norm": 1.2771859169006348, | |
| "learning_rate": 3.7859575403310534e-05, | |
| "loss": 3.8607, | |
| "step": 6990 | |
| }, | |
| { | |
| "epoch": 0.7750006919649035, | |
| "grad_norm": 1.396920919418335, | |
| "learning_rate": 3.75059247805982e-05, | |
| "loss": 3.8288, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7761078358105676, | |
| "grad_norm": 1.277076244354248, | |
| "learning_rate": 3.715369750497182e-05, | |
| "loss": 3.8414, | |
| "step": 7010 | |
| }, | |
| { | |
| "epoch": 0.7772149796562319, | |
| "grad_norm": 1.464040994644165, | |
| "learning_rate": 3.68028980330365e-05, | |
| "loss": 3.8072, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 0.778322123501896, | |
| "grad_norm": 1.2639997005462646, | |
| "learning_rate": 3.645353080333174e-05, | |
| "loss": 3.8769, | |
| "step": 7030 | |
| }, | |
| { | |
| "epoch": 0.7794292673475601, | |
| "grad_norm": 1.3652697801589966, | |
| "learning_rate": 3.610560023627555e-05, | |
| "loss": 3.883, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 0.7805364111932243, | |
| "grad_norm": 1.3596171140670776, | |
| "learning_rate": 3.57591107341083e-05, | |
| "loss": 3.8382, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.7816435550388884, | |
| "grad_norm": 1.3012511730194092, | |
| "learning_rate": 3.541406668083704e-05, | |
| "loss": 3.9435, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 0.7827506988845526, | |
| "grad_norm": 1.3602337837219238, | |
| "learning_rate": 3.507047244218014e-05, | |
| "loss": 3.813, | |
| "step": 7070 | |
| }, | |
| { | |
| "epoch": 0.7838578427302167, | |
| "grad_norm": 1.1615537405014038, | |
| "learning_rate": 3.4728332365511954e-05, | |
| "loss": 3.9047, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 0.7849649865758809, | |
| "grad_norm": 1.2543020248413086, | |
| "learning_rate": 3.43876507798079e-05, | |
| "loss": 3.854, | |
| "step": 7090 | |
| }, | |
| { | |
| "epoch": 0.786072130421545, | |
| "grad_norm": 1.230818748474121, | |
| "learning_rate": 3.404843199558945e-05, | |
| "loss": 3.8495, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.7871792742672091, | |
| "grad_norm": 1.423280954360962, | |
| "learning_rate": 3.3710680304870056e-05, | |
| "loss": 3.8872, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 0.7882864181128734, | |
| "grad_norm": 1.2478442192077637, | |
| "learning_rate": 3.3374399981100354e-05, | |
| "loss": 3.8242, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 0.7893935619585375, | |
| "grad_norm": 1.226762056350708, | |
| "learning_rate": 3.30395952791144e-05, | |
| "loss": 3.8557, | |
| "step": 7130 | |
| }, | |
| { | |
| "epoch": 0.7905007058042016, | |
| "grad_norm": 1.3779078722000122, | |
| "learning_rate": 3.2706270435075716e-05, | |
| "loss": 3.9277, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 0.7916078496498657, | |
| "grad_norm": 1.2733426094055176, | |
| "learning_rate": 3.2374429666423614e-05, | |
| "loss": 3.9097, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.7927149934955299, | |
| "grad_norm": 1.245455265045166, | |
| "learning_rate": 3.2044077171820156e-05, | |
| "loss": 3.8863, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 0.793822137341194, | |
| "grad_norm": 1.3173507452011108, | |
| "learning_rate": 3.1715217131096554e-05, | |
| "loss": 3.849, | |
| "step": 7170 | |
| }, | |
| { | |
| "epoch": 0.7949292811868582, | |
| "grad_norm": 1.3037158250808716, | |
| "learning_rate": 3.138785370520073e-05, | |
| "loss": 3.7975, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 0.7960364250325224, | |
| "grad_norm": 1.3901885747909546, | |
| "learning_rate": 3.106199103614441e-05, | |
| "loss": 3.7659, | |
| "step": 7190 | |
| }, | |
| { | |
| "epoch": 0.7971435688781865, | |
| "grad_norm": 1.2567756175994873, | |
| "learning_rate": 3.073763324695078e-05, | |
| "loss": 3.8241, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.7982507127238506, | |
| "grad_norm": 1.307466745376587, | |
| "learning_rate": 3.041478444160233e-05, | |
| "loss": 3.8134, | |
| "step": 7210 | |
| }, | |
| { | |
| "epoch": 0.7993578565695147, | |
| "grad_norm": 1.4410961866378784, | |
| "learning_rate": 3.0093448704988815e-05, | |
| "loss": 3.8067, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 0.800465000415179, | |
| "grad_norm": 1.376894474029541, | |
| "learning_rate": 2.977363010285588e-05, | |
| "loss": 3.9238, | |
| "step": 7230 | |
| }, | |
| { | |
| "epoch": 0.8006864291843118, | |
| "eval_loss": 4.36316442489624, | |
| "eval_runtime": 79.0884, | |
| "eval_samples_per_second": 632.204, | |
| "eval_steps_per_second": 79.025, | |
| "step": 7232 | |
| }, | |
| { | |
| "epoch": 0.8015721442608431, | |
| "grad_norm": 1.3287566900253296, | |
| "learning_rate": 2.945533268175326e-05, | |
| "loss": 3.8127, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 0.8026792881065072, | |
| "grad_norm": 1.3364062309265137, | |
| "learning_rate": 2.9138560468983813e-05, | |
| "loss": 3.9328, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 0.8037864319521714, | |
| "grad_norm": 1.3750771284103394, | |
| "learning_rate": 2.882331747255242e-05, | |
| "loss": 3.7963, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 0.8048935757978355, | |
| "grad_norm": 1.3068097829818726, | |
| "learning_rate": 2.8509607681115376e-05, | |
| "loss": 3.8571, | |
| "step": 7270 | |
| }, | |
| { | |
| "epoch": 0.8060007196434997, | |
| "grad_norm": 1.3587127923965454, | |
| "learning_rate": 2.8197435063929918e-05, | |
| "loss": 3.9128, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 0.8071078634891639, | |
| "grad_norm": 1.2562081813812256, | |
| "learning_rate": 2.7886803570803794e-05, | |
| "loss": 3.9159, | |
| "step": 7290 | |
| }, | |
| { | |
| "epoch": 0.808215007334828, | |
| "grad_norm": 1.2627272605895996, | |
| "learning_rate": 2.757771713204575e-05, | |
| "loss": 3.8688, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.8093221511804921, | |
| "grad_norm": 1.3324167728424072, | |
| "learning_rate": 2.7270179658415286e-05, | |
| "loss": 3.7928, | |
| "step": 7310 | |
| }, | |
| { | |
| "epoch": 0.8104292950261562, | |
| "grad_norm": 1.3950971364974976, | |
| "learning_rate": 2.6964195041073493e-05, | |
| "loss": 3.8723, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 0.8115364388718205, | |
| "grad_norm": 1.352571725845337, | |
| "learning_rate": 2.665976715153377e-05, | |
| "loss": 3.8208, | |
| "step": 7330 | |
| }, | |
| { | |
| "epoch": 0.8126435827174846, | |
| "grad_norm": 1.3980189561843872, | |
| "learning_rate": 2.6356899841612606e-05, | |
| "loss": 3.8424, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 0.8137507265631487, | |
| "grad_norm": 1.328074336051941, | |
| "learning_rate": 2.6055596943381336e-05, | |
| "loss": 3.8523, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 0.8148578704088129, | |
| "grad_norm": 1.3194916248321533, | |
| "learning_rate": 2.575586226911705e-05, | |
| "loss": 3.7971, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 0.815965014254477, | |
| "grad_norm": 1.3236674070358276, | |
| "learning_rate": 2.5457699611254932e-05, | |
| "loss": 3.8363, | |
| "step": 7370 | |
| }, | |
| { | |
| "epoch": 0.8170721581001411, | |
| "grad_norm": 1.23525869846344, | |
| "learning_rate": 2.516111274233983e-05, | |
| "loss": 3.9244, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 0.8181793019458053, | |
| "grad_norm": 1.347618579864502, | |
| "learning_rate": 2.486610541497876e-05, | |
| "loss": 3.9152, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 0.8192864457914695, | |
| "grad_norm": 1.382225751876831, | |
| "learning_rate": 2.4572681361793407e-05, | |
| "loss": 3.7993, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.8203935896371336, | |
| "grad_norm": 1.1705516576766968, | |
| "learning_rate": 2.4280844295372688e-05, | |
| "loss": 3.8679, | |
| "step": 7410 | |
| }, | |
| { | |
| "epoch": 0.8215007334827977, | |
| "grad_norm": 1.3987349271774292, | |
| "learning_rate": 2.3990597908226218e-05, | |
| "loss": 3.7693, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 0.8226078773284619, | |
| "grad_norm": 1.2838268280029297, | |
| "learning_rate": 2.370194587273711e-05, | |
| "loss": 3.8835, | |
| "step": 7430 | |
| }, | |
| { | |
| "epoch": 0.8237150211741261, | |
| "grad_norm": 1.2486863136291504, | |
| "learning_rate": 2.3414891841115786e-05, | |
| "loss": 3.8035, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 0.8248221650197902, | |
| "grad_norm": 1.3080980777740479, | |
| "learning_rate": 2.312943944535373e-05, | |
| "loss": 3.7948, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 0.8259293088654543, | |
| "grad_norm": 1.3226797580718994, | |
| "learning_rate": 2.2845592297177385e-05, | |
| "loss": 3.8812, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 0.8270364527111185, | |
| "grad_norm": 1.4011107683181763, | |
| "learning_rate": 2.2563353988002756e-05, | |
| "loss": 3.8073, | |
| "step": 7470 | |
| }, | |
| { | |
| "epoch": 0.8281435965567826, | |
| "grad_norm": 1.2449629306793213, | |
| "learning_rate": 2.2282728088889602e-05, | |
| "loss": 3.7904, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 0.8292507404024468, | |
| "grad_norm": 1.2937129735946655, | |
| "learning_rate": 2.2003718150496596e-05, | |
| "loss": 3.8566, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 0.830357884248111, | |
| "grad_norm": 1.3584239482879639, | |
| "learning_rate": 2.1726327703036133e-05, | |
| "loss": 3.8677, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8314650280937751, | |
| "grad_norm": 1.1813411712646484, | |
| "learning_rate": 2.145056025622981e-05, | |
| "loss": 3.9732, | |
| "step": 7510 | |
| }, | |
| { | |
| "epoch": 0.8325721719394392, | |
| "grad_norm": 1.2757023572921753, | |
| "learning_rate": 2.117641929926403e-05, | |
| "loss": 3.8457, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 0.8336793157851033, | |
| "grad_norm": 1.2757357358932495, | |
| "learning_rate": 2.090390830074563e-05, | |
| "loss": 3.9597, | |
| "step": 7530 | |
| }, | |
| { | |
| "epoch": 0.8347864596307675, | |
| "grad_norm": 1.3658595085144043, | |
| "learning_rate": 2.0633030708658395e-05, | |
| "loss": 3.7943, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 0.8358936034764317, | |
| "grad_norm": 1.462045669555664, | |
| "learning_rate": 2.036378995031908e-05, | |
| "loss": 3.7708, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 0.8370007473220958, | |
| "grad_norm": 1.3909573554992676, | |
| "learning_rate": 2.009618943233419e-05, | |
| "loss": 3.8159, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 0.83810789116776, | |
| "grad_norm": 1.353805422782898, | |
| "learning_rate": 1.9830232540556907e-05, | |
| "loss": 3.9027, | |
| "step": 7570 | |
| }, | |
| { | |
| "epoch": 0.8392150350134241, | |
| "grad_norm": 1.5484203100204468, | |
| "learning_rate": 1.9565922640044023e-05, | |
| "loss": 3.7995, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 0.8403221788590882, | |
| "grad_norm": 1.4042919874191284, | |
| "learning_rate": 1.930326307501383e-05, | |
| "loss": 3.8311, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 0.8414293227047525, | |
| "grad_norm": 1.2047467231750488, | |
| "learning_rate": 1.9042257168803232e-05, | |
| "loss": 3.8063, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.8425364665504166, | |
| "grad_norm": 1.3214939832687378, | |
| "learning_rate": 1.878290822382622e-05, | |
| "loss": 3.8101, | |
| "step": 7610 | |
| }, | |
| { | |
| "epoch": 0.8436436103960807, | |
| "grad_norm": 1.330228567123413, | |
| "learning_rate": 1.8525219521531693e-05, | |
| "loss": 3.881, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 0.8447507542417448, | |
| "grad_norm": 1.302899956703186, | |
| "learning_rate": 1.8269194322362164e-05, | |
| "loss": 3.8532, | |
| "step": 7630 | |
| }, | |
| { | |
| "epoch": 0.845857898087409, | |
| "grad_norm": 1.3165912628173828, | |
| "learning_rate": 1.8014835865712434e-05, | |
| "loss": 3.9101, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 0.8469650419330732, | |
| "grad_norm": 1.365644931793213, | |
| "learning_rate": 1.776214736988853e-05, | |
| "loss": 3.722, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 0.8480721857787373, | |
| "grad_norm": 1.2290687561035156, | |
| "learning_rate": 1.7511132032067238e-05, | |
| "loss": 3.8496, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 0.8491793296244015, | |
| "grad_norm": 1.4127010107040405, | |
| "learning_rate": 1.7261793028255266e-05, | |
| "loss": 3.7735, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 0.8502864734700656, | |
| "grad_norm": 1.33592689037323, | |
| "learning_rate": 1.7014133513249474e-05, | |
| "loss": 3.8462, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 0.8513936173157297, | |
| "grad_norm": 1.4080361127853394, | |
| "learning_rate": 1.676815662059663e-05, | |
| "loss": 3.7812, | |
| "step": 7690 | |
| }, | |
| { | |
| "epoch": 0.852500761161394, | |
| "grad_norm": 1.3327304124832153, | |
| "learning_rate": 1.6523865462553854e-05, | |
| "loss": 3.8345, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.8536079050070581, | |
| "grad_norm": 1.2660259008407593, | |
| "learning_rate": 1.6281263130049442e-05, | |
| "loss": 3.8347, | |
| "step": 7710 | |
| }, | |
| { | |
| "epoch": 0.8547150488527222, | |
| "grad_norm": 1.3123475313186646, | |
| "learning_rate": 1.6040352692643356e-05, | |
| "loss": 3.8205, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 0.8558221926983863, | |
| "grad_norm": 1.342924952507019, | |
| "learning_rate": 1.5801137198488817e-05, | |
| "loss": 3.9076, | |
| "step": 7730 | |
| }, | |
| { | |
| "epoch": 0.8569293365440505, | |
| "grad_norm": 1.366937518119812, | |
| "learning_rate": 1.5563619674293466e-05, | |
| "loss": 3.8997, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 0.8580364803897146, | |
| "grad_norm": 1.3002761602401733, | |
| "learning_rate": 1.5327803125281103e-05, | |
| "loss": 3.869, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 0.8591436242353788, | |
| "grad_norm": 1.2442548274993896, | |
| "learning_rate": 1.509369053515378e-05, | |
| "loss": 3.7878, | |
| "step": 7760 | |
| }, | |
| { | |
| "epoch": 0.860250768081043, | |
| "grad_norm": 1.3569366931915283, | |
| "learning_rate": 1.486128486605383e-05, | |
| "loss": 3.8049, | |
| "step": 7770 | |
| }, | |
| { | |
| "epoch": 0.8613579119267071, | |
| "grad_norm": 1.3670191764831543, | |
| "learning_rate": 1.4630589058526743e-05, | |
| "loss": 3.7823, | |
| "step": 7780 | |
| }, | |
| { | |
| "epoch": 0.8624650557723712, | |
| "grad_norm": 1.2652491331100464, | |
| "learning_rate": 1.4401606031483497e-05, | |
| "loss": 3.9722, | |
| "step": 7790 | |
| }, | |
| { | |
| "epoch": 0.8635721996180353, | |
| "grad_norm": 1.253395438194275, | |
| "learning_rate": 1.4174338682164094e-05, | |
| "loss": 3.8504, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.8646793434636996, | |
| "grad_norm": 1.4219810962677002, | |
| "learning_rate": 1.3948789886100591e-05, | |
| "loss": 3.879, | |
| "step": 7810 | |
| }, | |
| { | |
| "epoch": 0.8657864873093637, | |
| "grad_norm": 1.3197954893112183, | |
| "learning_rate": 1.3724962497080732e-05, | |
| "loss": 3.8282, | |
| "step": 7820 | |
| }, | |
| { | |
| "epoch": 0.8668936311550278, | |
| "grad_norm": 1.3688158988952637, | |
| "learning_rate": 1.3502859347112095e-05, | |
| "loss": 3.8105, | |
| "step": 7830 | |
| }, | |
| { | |
| "epoch": 0.868000775000692, | |
| "grad_norm": 1.358634352684021, | |
| "learning_rate": 1.3282483246385856e-05, | |
| "loss": 3.8335, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 0.8691079188463561, | |
| "grad_norm": 1.3314993381500244, | |
| "learning_rate": 1.306383698324171e-05, | |
| "loss": 3.8114, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 0.8702150626920203, | |
| "grad_norm": 1.3824890851974487, | |
| "learning_rate": 1.284692332413209e-05, | |
| "loss": 3.751, | |
| "step": 7860 | |
| }, | |
| { | |
| "epoch": 0.8713222065376844, | |
| "grad_norm": 1.3824279308319092, | |
| "learning_rate": 1.2631745013587602e-05, | |
| "loss": 3.8927, | |
| "step": 7870 | |
| }, | |
| { | |
| "epoch": 0.8724293503833486, | |
| "grad_norm": 2.055382490158081, | |
| "learning_rate": 1.2418304774182075e-05, | |
| "loss": 3.8666, | |
| "step": 7880 | |
| }, | |
| { | |
| "epoch": 0.8735364942290127, | |
| "grad_norm": 1.3714137077331543, | |
| "learning_rate": 1.2206605306498003e-05, | |
| "loss": 3.8378, | |
| "step": 7890 | |
| }, | |
| { | |
| "epoch": 0.8746436380746768, | |
| "grad_norm": 1.2260453701019287, | |
| "learning_rate": 1.1996649289092742e-05, | |
| "loss": 3.921, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.875750781920341, | |
| "grad_norm": 1.2452430725097656, | |
| "learning_rate": 1.1788439378464198e-05, | |
| "loss": 3.8037, | |
| "step": 7910 | |
| }, | |
| { | |
| "epoch": 0.8768579257660052, | |
| "grad_norm": 1.3326976299285889, | |
| "learning_rate": 1.1581978209017567e-05, | |
| "loss": 3.8228, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 0.8779650696116693, | |
| "grad_norm": 1.2535789012908936, | |
| "learning_rate": 1.1377268393031791e-05, | |
| "loss": 3.7823, | |
| "step": 7930 | |
| }, | |
| { | |
| "epoch": 0.8790722134573334, | |
| "grad_norm": 1.4074896574020386, | |
| "learning_rate": 1.1174312520626555e-05, | |
| "loss": 3.7897, | |
| "step": 7940 | |
| }, | |
| { | |
| "epoch": 0.8801793573029976, | |
| "grad_norm": 1.308258295059204, | |
| "learning_rate": 1.0973113159729562e-05, | |
| "loss": 3.8746, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 0.8812865011486617, | |
| "grad_norm": 1.5126914978027344, | |
| "learning_rate": 1.0773672856043914e-05, | |
| "loss": 3.8781, | |
| "step": 7960 | |
| }, | |
| { | |
| "epoch": 0.8823936449943259, | |
| "grad_norm": 1.3011531829833984, | |
| "learning_rate": 1.0575994133016142e-05, | |
| "loss": 3.8433, | |
| "step": 7970 | |
| }, | |
| { | |
| "epoch": 0.8835007888399901, | |
| "grad_norm": 1.3286775350570679, | |
| "learning_rate": 1.0380079491803994e-05, | |
| "loss": 3.8648, | |
| "step": 7980 | |
| }, | |
| { | |
| "epoch": 0.8846079326856542, | |
| "grad_norm": 1.3164076805114746, | |
| "learning_rate": 1.0185931411245008e-05, | |
| "loss": 3.8174, | |
| "step": 7990 | |
| }, | |
| { | |
| "epoch": 0.8857150765313183, | |
| "grad_norm": 1.3122352361679077, | |
| "learning_rate": 9.993552347825078e-06, | |
| "loss": 3.805, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8868222203769824, | |
| "grad_norm": 1.3544143438339233, | |
| "learning_rate": 9.802944735647245e-06, | |
| "loss": 3.8302, | |
| "step": 8010 | |
| }, | |
| { | |
| "epoch": 0.8879293642226467, | |
| "grad_norm": 1.2577146291732788, | |
| "learning_rate": 9.614110986401169e-06, | |
| "loss": 3.8657, | |
| "step": 8020 | |
| }, | |
| { | |
| "epoch": 0.8890365080683108, | |
| "grad_norm": 1.3847260475158691, | |
| "learning_rate": 9.427053489332304e-06, | |
| "loss": 3.7758, | |
| "step": 8030 | |
| }, | |
| { | |
| "epoch": 0.8901436519139749, | |
| "grad_norm": 1.1531234979629517, | |
| "learning_rate": 9.241774611211977e-06, | |
| "loss": 3.8208, | |
| "step": 8040 | |
| }, | |
| { | |
| "epoch": 0.8912507957596391, | |
| "grad_norm": 1.2929434776306152, | |
| "learning_rate": 9.058276696307155e-06, | |
| "loss": 3.8414, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 0.8923579396053032, | |
| "grad_norm": 1.3089600801467896, | |
| "learning_rate": 8.876562066351023e-06, | |
| "loss": 3.787, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 0.8934650834509673, | |
| "grad_norm": 1.297970175743103, | |
| "learning_rate": 8.696633020513455e-06, | |
| "loss": 3.8252, | |
| "step": 8070 | |
| }, | |
| { | |
| "epoch": 0.8945722272966316, | |
| "grad_norm": 1.4533628225326538, | |
| "learning_rate": 8.518491835371926e-06, | |
| "loss": 3.7595, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 0.8956793711422957, | |
| "grad_norm": 1.3536813259124756, | |
| "learning_rate": 8.342140764882876e-06, | |
| "loss": 3.8415, | |
| "step": 8090 | |
| }, | |
| { | |
| "epoch": 0.8967865149879598, | |
| "grad_norm": 1.4366772174835205, | |
| "learning_rate": 8.16758204035291e-06, | |
| "loss": 3.8125, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.8978936588336239, | |
| "grad_norm": 1.2198227643966675, | |
| "learning_rate": 7.994817870410858e-06, | |
| "loss": 3.9182, | |
| "step": 8110 | |
| }, | |
| { | |
| "epoch": 0.8990008026792881, | |
| "grad_norm": 1.3517742156982422, | |
| "learning_rate": 7.823850440979645e-06, | |
| "loss": 3.8429, | |
| "step": 8120 | |
| }, | |
| { | |
| "epoch": 0.9001079465249523, | |
| "grad_norm": 1.455114483833313, | |
| "learning_rate": 7.65468191524865e-06, | |
| "loss": 3.8055, | |
| "step": 8130 | |
| }, | |
| { | |
| "epoch": 0.9007722328323507, | |
| "eval_loss": 4.347445964813232, | |
| "eval_runtime": 78.833, | |
| "eval_samples_per_second": 634.252, | |
| "eval_steps_per_second": 79.281, | |
| "step": 8136 | |
| }, | |
| { | |
| "epoch": 0.9012150903706164, | |
| "grad_norm": 1.3386210203170776, | |
| "learning_rate": 7.487314433646441e-06, | |
| "loss": 3.7847, | |
| "step": 8140 | |
| }, | |
| { | |
| "epoch": 0.9023222342162805, | |
| "grad_norm": 1.2345080375671387, | |
| "learning_rate": 7.321750113813546e-06, | |
| "loss": 3.8519, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 0.9034293780619447, | |
| "grad_norm": 1.3259464502334595, | |
| "learning_rate": 7.1579910505758285e-06, | |
| "loss": 3.7809, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 0.9045365219076088, | |
| "grad_norm": 1.327785611152649, | |
| "learning_rate": 6.996039315917756e-06, | |
| "loss": 3.9018, | |
| "step": 8170 | |
| }, | |
| { | |
| "epoch": 0.905643665753273, | |
| "grad_norm": 1.285530924797058, | |
| "learning_rate": 6.8358969589564715e-06, | |
| "loss": 3.8406, | |
| "step": 8180 | |
| }, | |
| { | |
| "epoch": 0.9067508095989372, | |
| "grad_norm": 1.2374411821365356, | |
| "learning_rate": 6.6775660059156295e-06, | |
| "loss": 3.8301, | |
| "step": 8190 | |
| }, | |
| { | |
| "epoch": 0.9078579534446013, | |
| "grad_norm": 1.3395283222198486, | |
| "learning_rate": 6.521048460099787e-06, | |
| "loss": 3.8325, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.9089650972902654, | |
| "grad_norm": 1.3900843858718872, | |
| "learning_rate": 6.366346301869252e-06, | |
| "loss": 3.79, | |
| "step": 8210 | |
| }, | |
| { | |
| "epoch": 0.9100722411359295, | |
| "grad_norm": 1.29372239112854, | |
| "learning_rate": 6.213461488614741e-06, | |
| "loss": 3.8059, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 0.9111793849815938, | |
| "grad_norm": 1.457944631576538, | |
| "learning_rate": 6.062395954732879e-06, | |
| "loss": 3.8301, | |
| "step": 8230 | |
| }, | |
| { | |
| "epoch": 0.9122865288272579, | |
| "grad_norm": 1.3149160146713257, | |
| "learning_rate": 5.913151611601541e-06, | |
| "loss": 3.8115, | |
| "step": 8240 | |
| }, | |
| { | |
| "epoch": 0.913393672672922, | |
| "grad_norm": 1.4059027433395386, | |
| "learning_rate": 5.7657303475556974e-06, | |
| "loss": 3.8083, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 0.9145008165185862, | |
| "grad_norm": 1.28351891040802, | |
| "learning_rate": 5.620134027863671e-06, | |
| "loss": 3.8029, | |
| "step": 8260 | |
| }, | |
| { | |
| "epoch": 0.9156079603642503, | |
| "grad_norm": 1.2808688879013062, | |
| "learning_rate": 5.476364494703256e-06, | |
| "loss": 3.7857, | |
| "step": 8270 | |
| }, | |
| { | |
| "epoch": 0.9167151042099144, | |
| "grad_norm": 1.2623027563095093, | |
| "learning_rate": 5.334423567138734e-06, | |
| "loss": 3.7994, | |
| "step": 8280 | |
| }, | |
| { | |
| "epoch": 0.9178222480555787, | |
| "grad_norm": 1.5176137685775757, | |
| "learning_rate": 5.1943130410975805e-06, | |
| "loss": 3.7781, | |
| "step": 8290 | |
| }, | |
| { | |
| "epoch": 0.9189293919012428, | |
| "grad_norm": 1.3034307956695557, | |
| "learning_rate": 5.056034689347927e-06, | |
| "loss": 3.8951, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.9200365357469069, | |
| "grad_norm": 1.3277850151062012, | |
| "learning_rate": 4.919590261476047e-06, | |
| "loss": 3.8128, | |
| "step": 8310 | |
| }, | |
| { | |
| "epoch": 0.921143679592571, | |
| "grad_norm": 1.2219781875610352, | |
| "learning_rate": 4.784981483864198e-06, | |
| "loss": 3.8202, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 0.9222508234382352, | |
| "grad_norm": 1.509503960609436, | |
| "learning_rate": 4.652210059668893e-06, | |
| "loss": 3.8175, | |
| "step": 8330 | |
| }, | |
| { | |
| "epoch": 0.9233579672838994, | |
| "grad_norm": 1.287468433380127, | |
| "learning_rate": 4.521277668799144e-06, | |
| "loss": 3.8433, | |
| "step": 8340 | |
| }, | |
| { | |
| "epoch": 0.9244651111295635, | |
| "grad_norm": 1.403244972229004, | |
| "learning_rate": 4.392185967895495e-06, | |
| "loss": 3.8225, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 0.9255722549752277, | |
| "grad_norm": 1.3234807252883911, | |
| "learning_rate": 4.264936590308782e-06, | |
| "loss": 3.794, | |
| "step": 8360 | |
| }, | |
| { | |
| "epoch": 0.9266793988208918, | |
| "grad_norm": 1.3365390300750732, | |
| "learning_rate": 4.139531146079622e-06, | |
| "loss": 3.8385, | |
| "step": 8370 | |
| }, | |
| { | |
| "epoch": 0.9277865426665559, | |
| "grad_norm": 1.4297044277191162, | |
| "learning_rate": 4.015971221918046e-06, | |
| "loss": 3.7669, | |
| "step": 8380 | |
| }, | |
| { | |
| "epoch": 0.9288936865122202, | |
| "grad_norm": 1.2327905893325806, | |
| "learning_rate": 3.894258381183313e-06, | |
| "loss": 3.9167, | |
| "step": 8390 | |
| }, | |
| { | |
| "epoch": 0.9300008303578843, | |
| "grad_norm": 1.5783439874649048, | |
| "learning_rate": 3.774394163864258e-06, | |
| "loss": 3.8481, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.9311079742035484, | |
| "grad_norm": 1.3087785243988037, | |
| "learning_rate": 3.6563800865596794e-06, | |
| "loss": 3.8525, | |
| "step": 8410 | |
| }, | |
| { | |
| "epoch": 0.9322151180492125, | |
| "grad_norm": 1.4142072200775146, | |
| "learning_rate": 3.540217642459331e-06, | |
| "loss": 3.8305, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 0.9333222618948767, | |
| "grad_norm": 1.367326259613037, | |
| "learning_rate": 3.4259083013248257e-06, | |
| "loss": 3.7999, | |
| "step": 8430 | |
| }, | |
| { | |
| "epoch": 0.9344294057405408, | |
| "grad_norm": 1.4351729154586792, | |
| "learning_rate": 3.3134535094711146e-06, | |
| "loss": 3.815, | |
| "step": 8440 | |
| }, | |
| { | |
| "epoch": 0.935536549586205, | |
| "grad_norm": 1.2734133005142212, | |
| "learning_rate": 3.2028546897483186e-06, | |
| "loss": 3.8686, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 0.9366436934318692, | |
| "grad_norm": 1.307189702987671, | |
| "learning_rate": 3.0941132415234602e-06, | |
| "loss": 3.8612, | |
| "step": 8460 | |
| }, | |
| { | |
| "epoch": 0.9377508372775333, | |
| "grad_norm": 1.2423137426376343, | |
| "learning_rate": 2.98723054066306e-06, | |
| "loss": 3.8491, | |
| "step": 8470 | |
| }, | |
| { | |
| "epoch": 0.9388579811231974, | |
| "grad_norm": 1.4321695566177368, | |
| "learning_rate": 2.882207939515435e-06, | |
| "loss": 3.7613, | |
| "step": 8480 | |
| }, | |
| { | |
| "epoch": 0.9399651249688615, | |
| "grad_norm": 1.368023157119751, | |
| "learning_rate": 2.7790467668938277e-06, | |
| "loss": 3.8696, | |
| "step": 8490 | |
| }, | |
| { | |
| "epoch": 0.9410722688145258, | |
| "grad_norm": 1.4758479595184326, | |
| "learning_rate": 2.677748328059487e-06, | |
| "loss": 3.8977, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9421794126601899, | |
| "grad_norm": 1.1756988763809204, | |
| "learning_rate": 2.578313904705098e-06, | |
| "loss": 3.8936, | |
| "step": 8510 | |
| }, | |
| { | |
| "epoch": 0.943286556505854, | |
| "grad_norm": 1.2310662269592285, | |
| "learning_rate": 2.480744754938696e-06, | |
| "loss": 3.7503, | |
| "step": 8520 | |
| }, | |
| { | |
| "epoch": 0.9443937003515182, | |
| "grad_norm": 1.3179534673690796, | |
| "learning_rate": 2.38504211326761e-06, | |
| "loss": 3.8942, | |
| "step": 8530 | |
| }, | |
| { | |
| "epoch": 0.9455008441971823, | |
| "grad_norm": 1.2465641498565674, | |
| "learning_rate": 2.291207190583011e-06, | |
| "loss": 3.8337, | |
| "step": 8540 | |
| }, | |
| { | |
| "epoch": 0.9466079880428465, | |
| "grad_norm": 1.3716423511505127, | |
| "learning_rate": 2.1992411741443896e-06, | |
| "loss": 3.8565, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 0.9477151318885106, | |
| "grad_norm": 1.291727066040039, | |
| "learning_rate": 2.1091452275647013e-06, | |
| "loss": 3.8635, | |
| "step": 8560 | |
| }, | |
| { | |
| "epoch": 0.9488222757341748, | |
| "grad_norm": 1.4095420837402344, | |
| "learning_rate": 2.02092049079558e-06, | |
| "loss": 3.931, | |
| "step": 8570 | |
| }, | |
| { | |
| "epoch": 0.9499294195798389, | |
| "grad_norm": 1.3469890356063843, | |
| "learning_rate": 1.9345680801128806e-06, | |
| "loss": 3.8031, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 0.951036563425503, | |
| "grad_norm": 1.2289035320281982, | |
| "learning_rate": 1.8500890881026076e-06, | |
| "loss": 3.8872, | |
| "step": 8590 | |
| }, | |
| { | |
| "epoch": 0.9521437072711673, | |
| "grad_norm": 1.2923479080200195, | |
| "learning_rate": 1.7674845836470609e-06, | |
| "loss": 3.9204, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.9532508511168314, | |
| "grad_norm": 1.3823212385177612, | |
| "learning_rate": 1.6867556119113612e-06, | |
| "loss": 3.8461, | |
| "step": 8610 | |
| }, | |
| { | |
| "epoch": 0.9543579949624955, | |
| "grad_norm": 1.2156734466552734, | |
| "learning_rate": 1.6079031943301623e-06, | |
| "loss": 3.7474, | |
| "step": 8620 | |
| }, | |
| { | |
| "epoch": 0.9554651388081596, | |
| "grad_norm": 1.316729187965393, | |
| "learning_rate": 1.5309283285947271e-06, | |
| "loss": 3.8394, | |
| "step": 8630 | |
| }, | |
| { | |
| "epoch": 0.9565722826538238, | |
| "grad_norm": 1.2862942218780518, | |
| "learning_rate": 1.4558319886403714e-06, | |
| "loss": 3.751, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 0.9576794264994879, | |
| "grad_norm": 1.2597215175628662, | |
| "learning_rate": 1.3826151246340566e-06, | |
| "loss": 3.8598, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 0.9587865703451521, | |
| "grad_norm": 1.4564849138259888, | |
| "learning_rate": 1.311278662962484e-06, | |
| "loss": 3.8726, | |
| "step": 8660 | |
| }, | |
| { | |
| "epoch": 0.9598937141908163, | |
| "grad_norm": 1.3483980894088745, | |
| "learning_rate": 1.2418235062202187e-06, | |
| "loss": 3.8471, | |
| "step": 8670 | |
| }, | |
| { | |
| "epoch": 0.9610008580364804, | |
| "grad_norm": 1.3388372659683228, | |
| "learning_rate": 1.174250533198401e-06, | |
| "loss": 3.8267, | |
| "step": 8680 | |
| }, | |
| { | |
| "epoch": 0.9621080018821445, | |
| "grad_norm": 1.333998441696167, | |
| "learning_rate": 1.1085605988735536e-06, | |
| "loss": 3.8232, | |
| "step": 8690 | |
| }, | |
| { | |
| "epoch": 0.9632151457278086, | |
| "grad_norm": 1.4082062244415283, | |
| "learning_rate": 1.0447545343968078e-06, | |
| "loss": 3.838, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.9643222895734729, | |
| "grad_norm": 1.3201833963394165, | |
| "learning_rate": 9.82833147083345e-07, | |
| "loss": 3.8155, | |
| "step": 8710 | |
| }, | |
| { | |
| "epoch": 0.965429433419137, | |
| "grad_norm": 1.21062171459198, | |
| "learning_rate": 9.22797220402205e-07, | |
| "loss": 3.8986, | |
| "step": 8720 | |
| }, | |
| { | |
| "epoch": 0.9665365772648011, | |
| "grad_norm": 1.3460724353790283, | |
| "learning_rate": 8.646475139664266e-07, | |
| "loss": 3.8592, | |
| "step": 8730 | |
| }, | |
| { | |
| "epoch": 0.9676437211104653, | |
| "grad_norm": 1.3785293102264404, | |
| "learning_rate": 8.083847635232732e-07, | |
| "loss": 3.8057, | |
| "step": 8740 | |
| }, | |
| { | |
| "epoch": 0.9687508649561294, | |
| "grad_norm": 1.3589550256729126, | |
| "learning_rate": 7.54009680945139e-07, | |
| "loss": 3.7868, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 0.9698580088017936, | |
| "grad_norm": 1.3476109504699707, | |
| "learning_rate": 7.015229542203904e-07, | |
| "loss": 3.8143, | |
| "step": 8760 | |
| }, | |
| { | |
| "epoch": 0.9709651526474578, | |
| "grad_norm": 1.4007353782653809, | |
| "learning_rate": 6.509252474446891e-07, | |
| "loss": 3.8802, | |
| "step": 8770 | |
| }, | |
| { | |
| "epoch": 0.9720722964931219, | |
| "grad_norm": 1.2617483139038086, | |
| "learning_rate": 6.022172008126658e-07, | |
| "loss": 3.8853, | |
| "step": 8780 | |
| }, | |
| { | |
| "epoch": 0.973179440338786, | |
| "grad_norm": 1.2954703569412231, | |
| "learning_rate": 5.553994306096765e-07, | |
| "loss": 3.8283, | |
| "step": 8790 | |
| }, | |
| { | |
| "epoch": 0.9742865841844501, | |
| "grad_norm": 1.3063156604766846, | |
| "learning_rate": 5.104725292041423e-07, | |
| "loss": 3.8089, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.9753937280301143, | |
| "grad_norm": 1.3450706005096436, | |
| "learning_rate": 4.674370650400217e-07, | |
| "loss": 3.9179, | |
| "step": 8810 | |
| }, | |
| { | |
| "epoch": 0.9765008718757785, | |
| "grad_norm": 1.2739299535751343, | |
| "learning_rate": 4.2629358262953327e-07, | |
| "loss": 3.8104, | |
| "step": 8820 | |
| }, | |
| { | |
| "epoch": 0.9776080157214426, | |
| "grad_norm": 1.334546446800232, | |
| "learning_rate": 3.870426025463946e-07, | |
| "loss": 3.8589, | |
| "step": 8830 | |
| }, | |
| { | |
| "epoch": 0.9787151595671068, | |
| "grad_norm": 1.26683509349823, | |
| "learning_rate": 3.496846214191107e-07, | |
| "loss": 3.8631, | |
| "step": 8840 | |
| }, | |
| { | |
| "epoch": 0.9798223034127709, | |
| "grad_norm": 1.293189287185669, | |
| "learning_rate": 3.142201119248289e-07, | |
| "loss": 3.8844, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 0.980929447258435, | |
| "grad_norm": 1.3638145923614502, | |
| "learning_rate": 2.8064952278321063e-07, | |
| "loss": 3.7635, | |
| "step": 8860 | |
| }, | |
| { | |
| "epoch": 0.9820365911040992, | |
| "grad_norm": 1.270646333694458, | |
| "learning_rate": 2.489732787508525e-07, | |
| "loss": 3.7922, | |
| "step": 8870 | |
| }, | |
| { | |
| "epoch": 0.9831437349497634, | |
| "grad_norm": 1.4044461250305176, | |
| "learning_rate": 2.1919178061589048e-07, | |
| "loss": 3.8613, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 0.9842508787954275, | |
| "grad_norm": 1.216270923614502, | |
| "learning_rate": 1.913054051928875e-07, | |
| "loss": 3.8759, | |
| "step": 8890 | |
| }, | |
| { | |
| "epoch": 0.9853580226410916, | |
| "grad_norm": 1.3241215944290161, | |
| "learning_rate": 1.653145053181537e-07, | |
| "loss": 3.8011, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.9864651664867558, | |
| "grad_norm": 1.3769909143447876, | |
| "learning_rate": 1.412194098451669e-07, | |
| "loss": 3.8194, | |
| "step": 8910 | |
| }, | |
| { | |
| "epoch": 0.98757231033242, | |
| "grad_norm": 1.3283618688583374, | |
| "learning_rate": 1.1902042364054232e-07, | |
| "loss": 3.8752, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 0.9886794541780841, | |
| "grad_norm": 1.2784342765808105, | |
| "learning_rate": 9.871782758005264e-08, | |
| "loss": 3.8847, | |
| "step": 8930 | |
| }, | |
| { | |
| "epoch": 0.9897865980237482, | |
| "grad_norm": 1.40086829662323, | |
| "learning_rate": 8.031187854514731e-08, | |
| "loss": 3.8774, | |
| "step": 8940 | |
| }, | |
| { | |
| "epoch": 0.9908937418694124, | |
| "grad_norm": 1.3735054731369019, | |
| "learning_rate": 6.380280941968852e-08, | |
| "loss": 3.8302, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 0.9920008857150765, | |
| "grad_norm": 1.4871920347213745, | |
| "learning_rate": 4.9190829086986904e-08, | |
| "loss": 3.8496, | |
| "step": 8960 | |
| }, | |
| { | |
| "epoch": 0.9931080295607406, | |
| "grad_norm": 1.3167036771774292, | |
| "learning_rate": 3.647612242717035e-08, | |
| "loss": 3.8671, | |
| "step": 8970 | |
| }, | |
| { | |
| "epoch": 0.9942151734064049, | |
| "grad_norm": 1.39736008644104, | |
| "learning_rate": 2.5658850314852486e-08, | |
| "loss": 3.9115, | |
| "step": 8980 | |
| }, | |
| { | |
| "epoch": 0.995322317252069, | |
| "grad_norm": 1.3444159030914307, | |
| "learning_rate": 1.673914961710099e-08, | |
| "loss": 3.8201, | |
| "step": 8990 | |
| }, | |
| { | |
| "epoch": 0.9964294610977331, | |
| "grad_norm": 1.2445727586746216, | |
| "learning_rate": 9.717133191672332e-09, | |
| "loss": 3.8123, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.9975366049433972, | |
| "grad_norm": 1.2063875198364258, | |
| "learning_rate": 4.592889885596252e-09, | |
| "loss": 3.8155, | |
| "step": 9010 | |
| }, | |
| { | |
| "epoch": 0.9986437487890614, | |
| "grad_norm": 1.5283468961715698, | |
| "learning_rate": 1.366484534109924e-09, | |
| "loss": 3.6151, | |
| "step": 9020 | |
| }, | |
| { | |
| "epoch": 0.9997508926347256, | |
| "grad_norm": 1.2638022899627686, | |
| "learning_rate": 3.795795975869431e-11, | |
| "loss": 3.7867, | |
| "step": 9030 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 9032, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 904, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4891741101465600.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |