| { | |
| "best_metric": 1.1406679153442383, | |
| "best_model_checkpoint": "/Users/bbunzeck/Documents/lexdec-llamas/large-char/checkpoint-9063", | |
| "epoch": 0.9999130726340234, | |
| "eval_steps": 1007, | |
| "global_step": 10065, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0009934556111614737, | |
| "grad_norm": 5.29207706451416, | |
| "learning_rate": 1.4999999999999999e-05, | |
| "loss": 4.4002, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0019869112223229474, | |
| "grad_norm": 2.1869823932647705, | |
| "learning_rate": 2.9999999999999997e-05, | |
| "loss": 3.6401, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0029803668334844216, | |
| "grad_norm": 0.9984068870544434, | |
| "learning_rate": 4.4999999999999996e-05, | |
| "loss": 3.2167, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.003973822444645895, | |
| "grad_norm": 0.700920820236206, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 3.1034, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.004967278055807369, | |
| "grad_norm": 2.1026079654693604, | |
| "learning_rate": 7.5e-05, | |
| "loss": 3.0336, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.005960733666968843, | |
| "grad_norm": 1.2076551914215088, | |
| "learning_rate": 8.999999999999999e-05, | |
| "loss": 2.8545, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.006954189278130317, | |
| "grad_norm": 1.447243332862854, | |
| "learning_rate": 0.00010499999999999999, | |
| "loss": 2.6664, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.00794764488929179, | |
| "grad_norm": 1.2562823295593262, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 2.5406, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.008941100500453264, | |
| "grad_norm": 1.3105968236923218, | |
| "learning_rate": 0.000135, | |
| "loss": 2.4273, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.009934556111614737, | |
| "grad_norm": 1.9283251762390137, | |
| "learning_rate": 0.00015, | |
| "loss": 2.3703, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.010928011722776212, | |
| "grad_norm": 1.5582048892974854, | |
| "learning_rate": 0.000165, | |
| "loss": 2.3019, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.011921467333937686, | |
| "grad_norm": 1.2038098573684692, | |
| "learning_rate": 0.00017999999999999998, | |
| "loss": 2.2016, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.01291492294509916, | |
| "grad_norm": 1.2954462766647339, | |
| "learning_rate": 0.000195, | |
| "loss": 2.1344, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.013908378556260634, | |
| "grad_norm": 1.1450471878051758, | |
| "learning_rate": 0.00020999999999999998, | |
| "loss": 2.0852, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.014901834167422107, | |
| "grad_norm": 1.3502197265625, | |
| "learning_rate": 0.000225, | |
| "loss": 2.03, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.01589528977858358, | |
| "grad_norm": 1.293748140335083, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 1.9633, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.016888745389745056, | |
| "grad_norm": 1.2430918216705322, | |
| "learning_rate": 0.00025499999999999996, | |
| "loss": 1.9261, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.01788220100090653, | |
| "grad_norm": 1.6550723314285278, | |
| "learning_rate": 0.00027, | |
| "loss": 1.8994, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.018875656612068, | |
| "grad_norm": 1.2588621377944946, | |
| "learning_rate": 0.000285, | |
| "loss": 1.8689, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.019869112223229474, | |
| "grad_norm": 1.072903037071228, | |
| "learning_rate": 0.0003, | |
| "loss": 1.8291, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.02086256783439095, | |
| "grad_norm": 1.5793360471725464, | |
| "learning_rate": 0.0002999992393822385, | |
| "loss": 1.8128, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.021856023445552424, | |
| "grad_norm": 0.9704577922821045, | |
| "learning_rate": 0.000299996957536668, | |
| "loss": 1.7878, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.022849479056713896, | |
| "grad_norm": 0.8479828238487244, | |
| "learning_rate": 0.0002999931544864298, | |
| "loss": 1.763, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.023842934667875373, | |
| "grad_norm": 0.8488192558288574, | |
| "learning_rate": 0.00029998783027009295, | |
| "loss": 1.7213, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.024836390279036846, | |
| "grad_norm": 0.7466472387313843, | |
| "learning_rate": 0.0002999809849416534, | |
| "loss": 1.7309, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.02582984589019832, | |
| "grad_norm": 0.9153957366943359, | |
| "learning_rate": 0.0002999726185705334, | |
| "loss": 1.6989, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.02682330150135979, | |
| "grad_norm": 0.9594451189041138, | |
| "learning_rate": 0.00029996273124158115, | |
| "loss": 1.7, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.027816757112521268, | |
| "grad_norm": 0.8583829998970032, | |
| "learning_rate": 0.0002999513230550697, | |
| "loss": 1.7048, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.02881021272368274, | |
| "grad_norm": 0.8427692651748657, | |
| "learning_rate": 0.00029993839412669593, | |
| "loss": 1.6674, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.029803668334844213, | |
| "grad_norm": 0.8587358593940735, | |
| "learning_rate": 0.0002999239445875795, | |
| "loss": 1.6609, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.030797123946005686, | |
| "grad_norm": 0.7885432839393616, | |
| "learning_rate": 0.00029990797458426146, | |
| "loss": 1.6317, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.03179057955716716, | |
| "grad_norm": 0.8649609684944153, | |
| "learning_rate": 0.0002998904842787027, | |
| "loss": 1.6259, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.03278403516832863, | |
| "grad_norm": 0.757290244102478, | |
| "learning_rate": 0.00029987147384828234, | |
| "loss": 1.6038, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.03377749077949011, | |
| "grad_norm": 0.8383753299713135, | |
| "learning_rate": 0.000299850943485796, | |
| "loss": 1.5934, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.034770946390651584, | |
| "grad_norm": 0.723807692527771, | |
| "learning_rate": 0.00029982889339945384, | |
| "loss": 1.5714, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.03576440200181306, | |
| "grad_norm": 0.7404069304466248, | |
| "learning_rate": 0.00029980532381287836, | |
| "loss": 1.576, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.03675785761297453, | |
| "grad_norm": 0.7834609150886536, | |
| "learning_rate": 0.0002997802349651021, | |
| "loss": 1.5961, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.037751313224136, | |
| "grad_norm": 0.681057870388031, | |
| "learning_rate": 0.00029975362711056545, | |
| "loss": 1.5537, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.038744768835297476, | |
| "grad_norm": 0.6808239221572876, | |
| "learning_rate": 0.00029972550051911377, | |
| "loss": 1.565, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.03973822444645895, | |
| "grad_norm": 0.7648847699165344, | |
| "learning_rate": 0.00029969585547599494, | |
| "loss": 1.5384, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.04073168005762043, | |
| "grad_norm": 0.6745840311050415, | |
| "learning_rate": 0.00029966469228185613, | |
| "loss": 1.536, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.0417251356687819, | |
| "grad_norm": 0.7665774822235107, | |
| "learning_rate": 0.0002996320112527412, | |
| "loss": 1.5349, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.042718591279943374, | |
| "grad_norm": 0.6684510111808777, | |
| "learning_rate": 0.00029959781272008695, | |
| "loss": 1.5285, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.04371204689110485, | |
| "grad_norm": 0.6902784705162048, | |
| "learning_rate": 0.0002995620970307203, | |
| "loss": 1.526, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.04470550250226632, | |
| "grad_norm": 0.7477157115936279, | |
| "learning_rate": 0.00029952486454685437, | |
| "loss": 1.5277, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.04569895811342779, | |
| "grad_norm": 0.8565313220024109, | |
| "learning_rate": 0.00029948611564608506, | |
| "loss": 1.486, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.046692413724589266, | |
| "grad_norm": 0.6772172451019287, | |
| "learning_rate": 0.000299445850721387, | |
| "loss": 1.5142, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.047685869335750745, | |
| "grad_norm": 0.6629778146743774, | |
| "learning_rate": 0.0002994040701811098, | |
| "loss": 1.4998, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.04867932494691222, | |
| "grad_norm": 0.6849949955940247, | |
| "learning_rate": 0.00029936077444897373, | |
| "loss": 1.5, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.04967278055807369, | |
| "grad_norm": 0.6776939630508423, | |
| "learning_rate": 0.0002993159639640655, | |
| "loss": 1.5044, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.050666236169235164, | |
| "grad_norm": 0.7044647336006165, | |
| "learning_rate": 0.00029926963918083376, | |
| "loss": 1.4885, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.05165969178039664, | |
| "grad_norm": 0.8003182411193848, | |
| "learning_rate": 0.00029922180056908456, | |
| "loss": 1.4668, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.05265314739155811, | |
| "grad_norm": 0.6357293725013733, | |
| "learning_rate": 0.00029917244861397653, | |
| "loss": 1.4843, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.05364660300271958, | |
| "grad_norm": 0.6657519936561584, | |
| "learning_rate": 0.000299121583816016, | |
| "loss": 1.4871, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.054640058613881055, | |
| "grad_norm": 0.6451624035835266, | |
| "learning_rate": 0.0002990692066910519, | |
| "loss": 1.4561, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.055633514225042535, | |
| "grad_norm": 0.669553279876709, | |
| "learning_rate": 0.00029901531777027046, | |
| "loss": 1.4723, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.05662696983620401, | |
| "grad_norm": 0.6627147197723389, | |
| "learning_rate": 0.00029895991760019007, | |
| "loss": 1.4591, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.05762042544736548, | |
| "grad_norm": 0.7017959952354431, | |
| "learning_rate": 0.00029890300674265526, | |
| "loss": 1.4747, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.058613881058526954, | |
| "grad_norm": 0.6519443392753601, | |
| "learning_rate": 0.0002988445857748316, | |
| "loss": 1.4642, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.05960733666968843, | |
| "grad_norm": 0.6378771066665649, | |
| "learning_rate": 0.0002987846552891995, | |
| "loss": 1.4427, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.0606007922808499, | |
| "grad_norm": 0.6792002320289612, | |
| "learning_rate": 0.000298723215893548, | |
| "loss": 1.4413, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.06159424789201137, | |
| "grad_norm": 0.6617796421051025, | |
| "learning_rate": 0.00029866026821096924, | |
| "loss": 1.4423, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.06258770350317285, | |
| "grad_norm": 0.6430485844612122, | |
| "learning_rate": 0.0002985958128798514, | |
| "loss": 1.426, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.06358115911433432, | |
| "grad_norm": 0.6020046472549438, | |
| "learning_rate": 0.00029852985055387284, | |
| "loss": 1.4353, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.06457461472549579, | |
| "grad_norm": 0.6419572234153748, | |
| "learning_rate": 0.00029846238190199504, | |
| "loss": 1.4392, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.06556807033665726, | |
| "grad_norm": 0.6080176830291748, | |
| "learning_rate": 0.00029839340760845614, | |
| "loss": 1.3977, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.06656152594781875, | |
| "grad_norm": 0.6436944603919983, | |
| "learning_rate": 0.00029832292837276376, | |
| "loss": 1.4243, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.06755498155898022, | |
| "grad_norm": 0.602425754070282, | |
| "learning_rate": 0.00029825094490968794, | |
| "loss": 1.4259, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.0685484371701417, | |
| "grad_norm": 0.6178025603294373, | |
| "learning_rate": 0.00029817745794925415, | |
| "loss": 1.4354, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.06954189278130317, | |
| "grad_norm": 0.63409423828125, | |
| "learning_rate": 0.00029810246823673537, | |
| "loss": 1.4314, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.07053534839246464, | |
| "grad_norm": 0.6901246905326843, | |
| "learning_rate": 0.0002980259765326452, | |
| "loss": 1.4197, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.07152880400362611, | |
| "grad_norm": 0.6050143837928772, | |
| "learning_rate": 0.00029794798361272964, | |
| "loss": 1.4255, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.07252225961478759, | |
| "grad_norm": 0.59127277135849, | |
| "learning_rate": 0.0002978684902679592, | |
| "loss": 1.4323, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.07351571522594906, | |
| "grad_norm": 0.5614879131317139, | |
| "learning_rate": 0.00029778749730452133, | |
| "loss": 1.3949, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.07450917083711053, | |
| "grad_norm": 0.5925185680389404, | |
| "learning_rate": 0.0002977050055438118, | |
| "loss": 1.4169, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.075502626448272, | |
| "grad_norm": 0.6346535682678223, | |
| "learning_rate": 0.0002976210158224266, | |
| "loss": 1.4183, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.07649608205943348, | |
| "grad_norm": 0.6336431503295898, | |
| "learning_rate": 0.00029753552899215344, | |
| "loss": 1.3799, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.07748953767059495, | |
| "grad_norm": 0.6319008469581604, | |
| "learning_rate": 0.0002974485459199629, | |
| "loss": 1.3991, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.07848299328175642, | |
| "grad_norm": 0.5974861979484558, | |
| "learning_rate": 0.000297360067488, | |
| "loss": 1.4196, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.0794764488929179, | |
| "grad_norm": 0.6162097454071045, | |
| "learning_rate": 0.00029727009459357487, | |
| "loss": 1.3995, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.08046990450407937, | |
| "grad_norm": 0.5667980313301086, | |
| "learning_rate": 0.00029717862814915405, | |
| "loss": 1.3789, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.08146336011524086, | |
| "grad_norm": 0.6212806105613708, | |
| "learning_rate": 0.00029708566908235077, | |
| "loss": 1.3819, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.08245681572640233, | |
| "grad_norm": 0.5819615721702576, | |
| "learning_rate": 0.00029699121833591605, | |
| "loss": 1.3975, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.0834502713375638, | |
| "grad_norm": 0.5672417879104614, | |
| "learning_rate": 0.0002968952768677287, | |
| "loss": 1.3884, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.08444372694872528, | |
| "grad_norm": 0.567499041557312, | |
| "learning_rate": 0.00029679784565078584, | |
| "loss": 1.3972, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.08543718255988675, | |
| "grad_norm": 0.7018792629241943, | |
| "learning_rate": 0.00029669892567319307, | |
| "loss": 1.3806, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.08643063817104822, | |
| "grad_norm": 0.593984067440033, | |
| "learning_rate": 0.0002965985179381542, | |
| "loss": 1.3832, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.0874240937822097, | |
| "grad_norm": 0.568536102771759, | |
| "learning_rate": 0.00029649662346396133, | |
| "loss": 1.3846, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.08841754939337117, | |
| "grad_norm": 0.5485777258872986, | |
| "learning_rate": 0.0002963932432839844, | |
| "loss": 1.3981, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.08941100500453264, | |
| "grad_norm": 0.6345034241676331, | |
| "learning_rate": 0.0002962883784466608, | |
| "loss": 1.3647, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.09040446061569411, | |
| "grad_norm": 0.6001213192939758, | |
| "learning_rate": 0.00029618203001548465, | |
| "loss": 1.3908, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.09139791622685559, | |
| "grad_norm": 0.5769600868225098, | |
| "learning_rate": 0.000296074199068996, | |
| "loss": 1.388, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.09239137183801706, | |
| "grad_norm": 0.5454570651054382, | |
| "learning_rate": 0.0002959648867007699, | |
| "loss": 1.3648, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.09338482744917853, | |
| "grad_norm": 0.5193580389022827, | |
| "learning_rate": 0.0002958540940194055, | |
| "loss": 1.3651, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.09437828306034, | |
| "grad_norm": 0.6213797926902771, | |
| "learning_rate": 0.0002957418221485145, | |
| "loss": 1.3795, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.09537173867150149, | |
| "grad_norm": 0.5813576579093933, | |
| "learning_rate": 0.00029562807222671, | |
| "loss": 1.3668, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.09636519428266296, | |
| "grad_norm": 0.6071997880935669, | |
| "learning_rate": 0.00029551284540759473, | |
| "loss": 1.3474, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.09735864989382444, | |
| "grad_norm": 0.6366572380065918, | |
| "learning_rate": 0.0002953961428597496, | |
| "loss": 1.3742, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.09835210550498591, | |
| "grad_norm": 0.5855928659439087, | |
| "learning_rate": 0.00029527796576672173, | |
| "loss": 1.3807, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.09934556111614738, | |
| "grad_norm": 0.5801404714584351, | |
| "learning_rate": 0.00029515831532701234, | |
| "loss": 1.364, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1000409800439604, | |
| "eval_loss": 1.337802767753601, | |
| "eval_runtime": 217.2125, | |
| "eval_samples_per_second": 230.189, | |
| "eval_steps_per_second": 28.774, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.10033901672730886, | |
| "grad_norm": 0.6000391244888306, | |
| "learning_rate": 0.00029503719275406477, | |
| "loss": 1.3675, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.10133247233847033, | |
| "grad_norm": 0.6123185753822327, | |
| "learning_rate": 0.0002949145992762521, | |
| "loss": 1.3735, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.1023259279496318, | |
| "grad_norm": 0.610657811164856, | |
| "learning_rate": 0.0002947905361368647, | |
| "loss": 1.3546, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.10331938356079327, | |
| "grad_norm": 0.6239680051803589, | |
| "learning_rate": 0.00029466500459409756, | |
| "loss": 1.3486, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.10431283917195475, | |
| "grad_norm": 0.5398300886154175, | |
| "learning_rate": 0.0002945380059210377, | |
| "loss": 1.3575, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.10530629478311622, | |
| "grad_norm": 0.6671786904335022, | |
| "learning_rate": 0.000294409541405651, | |
| "loss": 1.3435, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.10629975039427769, | |
| "grad_norm": 0.6025234460830688, | |
| "learning_rate": 0.0002942796123507693, | |
| "loss": 1.3498, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.10729320600543917, | |
| "grad_norm": 0.5941970348358154, | |
| "learning_rate": 0.0002941482200740774, | |
| "loss": 1.355, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.10828666161660064, | |
| "grad_norm": 0.5551139712333679, | |
| "learning_rate": 0.00029401536590809913, | |
| "loss": 1.3345, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.10928011722776211, | |
| "grad_norm": 0.5619410276412964, | |
| "learning_rate": 0.00029388105120018446, | |
| "loss": 1.3361, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.1102735728389236, | |
| "grad_norm": 0.5853624939918518, | |
| "learning_rate": 0.00029374527731249533, | |
| "loss": 1.3438, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.11126702845008507, | |
| "grad_norm": 0.5613996386528015, | |
| "learning_rate": 0.00029360804562199224, | |
| "loss": 1.3356, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.11226048406124654, | |
| "grad_norm": 0.539417028427124, | |
| "learning_rate": 0.0002934693575204199, | |
| "loss": 1.3412, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.11325393967240802, | |
| "grad_norm": 0.570012092590332, | |
| "learning_rate": 0.00029332921441429345, | |
| "loss": 1.3558, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.11424739528356949, | |
| "grad_norm": 0.5662525296211243, | |
| "learning_rate": 0.0002931876177248841, | |
| "loss": 1.3532, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.11524085089473096, | |
| "grad_norm": 0.6034231781959534, | |
| "learning_rate": 0.0002930445688882045, | |
| "loss": 1.3469, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.11623430650589243, | |
| "grad_norm": 0.5469877123832703, | |
| "learning_rate": 0.00029290006935499453, | |
| "loss": 1.3246, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.11722776211705391, | |
| "grad_norm": 0.5636057257652283, | |
| "learning_rate": 0.00029275412059070636, | |
| "loss": 1.3527, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.11822121772821538, | |
| "grad_norm": 0.5818510055541992, | |
| "learning_rate": 0.0002926067240754896, | |
| "loss": 1.308, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.11921467333937685, | |
| "grad_norm": 0.5315884947776794, | |
| "learning_rate": 0.0002924578813041764, | |
| "loss": 1.341, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.12020812895053833, | |
| "grad_norm": 0.5662302374839783, | |
| "learning_rate": 0.000292307593786266, | |
| "loss": 1.3364, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.1212015845616998, | |
| "grad_norm": 0.597668468952179, | |
| "learning_rate": 0.00029215586304590996, | |
| "loss": 1.339, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.12219504017286127, | |
| "grad_norm": 0.5625566840171814, | |
| "learning_rate": 0.00029200269062189623, | |
| "loss": 1.3503, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.12318849578402274, | |
| "grad_norm": 0.5155960321426392, | |
| "learning_rate": 0.0002918480780676336, | |
| "loss": 1.3166, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.12418195139518422, | |
| "grad_norm": 0.5308235883712769, | |
| "learning_rate": 0.00029169202695113617, | |
| "loss": 1.3372, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.1251754070063457, | |
| "grad_norm": 0.5910816192626953, | |
| "learning_rate": 0.00029153453885500736, | |
| "loss": 1.3414, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.12616886261750718, | |
| "grad_norm": 0.5780258178710938, | |
| "learning_rate": 0.00029137561537642367, | |
| "loss": 1.3277, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.12716231822866864, | |
| "grad_norm": 0.6031139492988586, | |
| "learning_rate": 0.00029121525812711874, | |
| "loss": 1.3304, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.12815577383983012, | |
| "grad_norm": 0.5696874856948853, | |
| "learning_rate": 0.00029105346873336684, | |
| "loss": 1.3399, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.12914922945099158, | |
| "grad_norm": 0.5428373217582703, | |
| "learning_rate": 0.0002908902488359664, | |
| "loss": 1.3514, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.13014268506215307, | |
| "grad_norm": 0.5869764089584351, | |
| "learning_rate": 0.0002907256000902236, | |
| "loss": 1.3248, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.13113614067331453, | |
| "grad_norm": 0.5231671929359436, | |
| "learning_rate": 0.0002905595241659351, | |
| "loss": 1.3241, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.13212959628447601, | |
| "grad_norm": 0.5357515215873718, | |
| "learning_rate": 0.0002903920227473717, | |
| "loss": 1.3163, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.1331230518956375, | |
| "grad_norm": 0.5605716705322266, | |
| "learning_rate": 0.00029022309753326065, | |
| "loss": 1.3366, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.13411650750679896, | |
| "grad_norm": 0.5430563688278198, | |
| "learning_rate": 0.0002900527502367689, | |
| "loss": 1.3148, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.13510996311796045, | |
| "grad_norm": 0.571535050868988, | |
| "learning_rate": 0.0002898809825854855, | |
| "loss": 1.3227, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.1361034187291219, | |
| "grad_norm": 0.5518633723258972, | |
| "learning_rate": 0.00028970779632140416, | |
| "loss": 1.3149, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.1370968743402834, | |
| "grad_norm": 0.5351448059082031, | |
| "learning_rate": 0.0002895331932009055, | |
| "loss": 1.3057, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.13809032995144485, | |
| "grad_norm": 0.5526089668273926, | |
| "learning_rate": 0.0002893571749947393, | |
| "loss": 1.3167, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.13908378556260634, | |
| "grad_norm": 0.5857049226760864, | |
| "learning_rate": 0.0002891797434880066, | |
| "loss": 1.3246, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.1400772411737678, | |
| "grad_norm": 0.5625345706939697, | |
| "learning_rate": 0.0002890009004801415, | |
| "loss": 1.3059, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.14107069678492928, | |
| "grad_norm": 0.5215427279472351, | |
| "learning_rate": 0.0002888206477848928, | |
| "loss": 1.3116, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.14206415239609074, | |
| "grad_norm": 0.5295782685279846, | |
| "learning_rate": 0.00028863898723030594, | |
| "loss": 1.3154, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.14305760800725223, | |
| "grad_norm": 0.573486328125, | |
| "learning_rate": 0.00028845592065870413, | |
| "loss": 1.3134, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.1440510636184137, | |
| "grad_norm": 0.5500824451446533, | |
| "learning_rate": 0.00028827144992666986, | |
| "loss": 1.3151, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.14504451922957518, | |
| "grad_norm": 0.5774082541465759, | |
| "learning_rate": 0.000288085576905026, | |
| "loss": 1.307, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.14603797484073663, | |
| "grad_norm": 0.5409136414527893, | |
| "learning_rate": 0.0002878983034788169, | |
| "loss": 1.3201, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.14703143045189812, | |
| "grad_norm": 0.5557056069374084, | |
| "learning_rate": 0.00028770963154728905, | |
| "loss": 1.3092, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.1480248860630596, | |
| "grad_norm": 0.5294632911682129, | |
| "learning_rate": 0.0002875195630238721, | |
| "loss": 1.2847, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.14901834167422107, | |
| "grad_norm": 0.5896183848381042, | |
| "learning_rate": 0.0002873280998361594, | |
| "loss": 1.2891, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.15001179728538255, | |
| "grad_norm": 0.5323734283447266, | |
| "learning_rate": 0.00028713524392588827, | |
| "loss": 1.2986, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.151005252896544, | |
| "grad_norm": 0.5725658535957336, | |
| "learning_rate": 0.0002869409972489204, | |
| "loss": 1.309, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.1519987085077055, | |
| "grad_norm": 0.5617366433143616, | |
| "learning_rate": 0.00028674536177522206, | |
| "loss": 1.297, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.15299216411886696, | |
| "grad_norm": 0.5333068370819092, | |
| "learning_rate": 0.00028654833948884423, | |
| "loss": 1.302, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.15398561973002844, | |
| "grad_norm": 0.5795490145683289, | |
| "learning_rate": 0.0002863499323879022, | |
| "loss": 1.29, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.1549790753411899, | |
| "grad_norm": 0.5586980581283569, | |
| "learning_rate": 0.0002861501424845555, | |
| "loss": 1.2961, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.1559725309523514, | |
| "grad_norm": 0.593062698841095, | |
| "learning_rate": 0.00028594897180498745, | |
| "loss": 1.3074, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.15696598656351285, | |
| "grad_norm": 0.5487144589424133, | |
| "learning_rate": 0.00028574642238938467, | |
| "loss": 1.2938, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.15795944217467434, | |
| "grad_norm": 0.49600857496261597, | |
| "learning_rate": 0.00028554249629191616, | |
| "loss": 1.3077, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.1589528977858358, | |
| "grad_norm": 0.5532371997833252, | |
| "learning_rate": 0.00028533719558071285, | |
| "loss": 1.2892, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.15994635339699728, | |
| "grad_norm": 0.5463429689407349, | |
| "learning_rate": 0.0002851305223378462, | |
| "loss": 1.3017, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.16093980900815874, | |
| "grad_norm": 0.7598294019699097, | |
| "learning_rate": 0.0002849224786593075, | |
| "loss": 1.2844, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.16193326461932023, | |
| "grad_norm": 0.589923083782196, | |
| "learning_rate": 0.00028471306665498624, | |
| "loss": 1.304, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.16292672023048171, | |
| "grad_norm": 0.5281336307525635, | |
| "learning_rate": 0.000284502288448649, | |
| "loss": 1.2912, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.16392017584164317, | |
| "grad_norm": 0.5142702460289001, | |
| "learning_rate": 0.00028429014617791773, | |
| "loss": 1.2966, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.16491363145280466, | |
| "grad_norm": 0.5056828856468201, | |
| "learning_rate": 0.0002840766419942481, | |
| "loss": 1.2957, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.16590708706396612, | |
| "grad_norm": 0.5111185312271118, | |
| "learning_rate": 0.0002838617780629079, | |
| "loss": 1.2908, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.1669005426751276, | |
| "grad_norm": 0.5428747534751892, | |
| "learning_rate": 0.0002836455565629546, | |
| "loss": 1.2999, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.16789399828628906, | |
| "grad_norm": 0.5203731060028076, | |
| "learning_rate": 0.0002834279796872138, | |
| "loss": 1.2998, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.16888745389745055, | |
| "grad_norm": 0.5578112006187439, | |
| "learning_rate": 0.00028320904964225665, | |
| "loss": 1.2774, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.169880909508612, | |
| "grad_norm": 0.5637850165367126, | |
| "learning_rate": 0.0002829887686483775, | |
| "loss": 1.2941, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.1708743651197735, | |
| "grad_norm": 0.5894926190376282, | |
| "learning_rate": 0.0002827671389395716, | |
| "loss": 1.2879, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.17186782073093496, | |
| "grad_norm": 1.018946647644043, | |
| "learning_rate": 0.0002825441627635121, | |
| "loss": 1.2683, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.17286127634209644, | |
| "grad_norm": 0.5260887742042542, | |
| "learning_rate": 0.00028231984238152766, | |
| "loss": 1.2786, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.1738547319532579, | |
| "grad_norm": 0.5494965314865112, | |
| "learning_rate": 0.000282094180068579, | |
| "loss": 1.3083, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.1748481875644194, | |
| "grad_norm": 0.5133727788925171, | |
| "learning_rate": 0.0002818671781132364, | |
| "loss": 1.279, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.17584164317558085, | |
| "grad_norm": 0.5044395923614502, | |
| "learning_rate": 0.0002816388388176561, | |
| "loss": 1.2786, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.17683509878674233, | |
| "grad_norm": 0.5667337775230408, | |
| "learning_rate": 0.00028140916449755706, | |
| "loss": 1.2694, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.17782855439790382, | |
| "grad_norm": 0.5417622923851013, | |
| "learning_rate": 0.0002811781574821975, | |
| "loss": 1.2952, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.17882201000906528, | |
| "grad_norm": 0.5346735119819641, | |
| "learning_rate": 0.0002809458201143513, | |
| "loss": 1.2869, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.17981546562022677, | |
| "grad_norm": 0.475115567445755, | |
| "learning_rate": 0.0002807121547502843, | |
| "loss": 1.2775, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.18080892123138823, | |
| "grad_norm": 0.5102688670158386, | |
| "learning_rate": 0.00028047716375972996, | |
| "loss": 1.3103, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.1818023768425497, | |
| "grad_norm": 0.5632643103599548, | |
| "learning_rate": 0.000280240849525866, | |
| "loss": 1.2923, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.18279583245371117, | |
| "grad_norm": 0.5179861187934875, | |
| "learning_rate": 0.0002800032144452898, | |
| "loss": 1.2797, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.18378928806487266, | |
| "grad_norm": 0.5114406943321228, | |
| "learning_rate": 0.00027976426092799416, | |
| "loss": 1.28, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.18478274367603412, | |
| "grad_norm": 0.5344343781471252, | |
| "learning_rate": 0.000279523991397343, | |
| "loss": 1.2766, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.1857761992871956, | |
| "grad_norm": 0.4935680329799652, | |
| "learning_rate": 0.00027928240829004663, | |
| "loss": 1.2927, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.18676965489835706, | |
| "grad_norm": 0.5560677647590637, | |
| "learning_rate": 0.000279039514056137, | |
| "loss": 1.2671, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.18776311050951855, | |
| "grad_norm": 0.4876042902469635, | |
| "learning_rate": 0.0002787953111589431, | |
| "loss": 1.2711, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.18875656612068, | |
| "grad_norm": 0.5092322826385498, | |
| "learning_rate": 0.00027854980207506566, | |
| "loss": 1.2919, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.1897500217318415, | |
| "grad_norm": 0.5449466109275818, | |
| "learning_rate": 0.00027830298929435235, | |
| "loss": 1.2756, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.19074347734300298, | |
| "grad_norm": 0.5372341871261597, | |
| "learning_rate": 0.0002780548753198723, | |
| "loss": 1.2878, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.19173693295416444, | |
| "grad_norm": 0.5671026706695557, | |
| "learning_rate": 0.00027780546266789076, | |
| "loss": 1.2665, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.19273038856532593, | |
| "grad_norm": 0.4836546778678894, | |
| "learning_rate": 0.00027755475386784364, | |
| "loss": 1.2862, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.1937238441764874, | |
| "grad_norm": 0.5355504751205444, | |
| "learning_rate": 0.00027730275146231186, | |
| "loss": 1.276, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.19471729978764887, | |
| "grad_norm": 0.4880260229110718, | |
| "learning_rate": 0.0002770494580069954, | |
| "loss": 1.2778, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.19571075539881033, | |
| "grad_norm": 0.49935540556907654, | |
| "learning_rate": 0.0002767948760706877, | |
| "loss": 1.2787, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.19670421100997182, | |
| "grad_norm": 0.5408512353897095, | |
| "learning_rate": 0.00027653900823524925, | |
| "loss": 1.2841, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.19769766662113328, | |
| "grad_norm": 0.531233549118042, | |
| "learning_rate": 0.00027628185709558176, | |
| "loss": 1.2799, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.19869112223229476, | |
| "grad_norm": 0.5099895000457764, | |
| "learning_rate": 0.0002760234252596015, | |
| "loss": 1.2842, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.19968457784345622, | |
| "grad_norm": 0.5348741412162781, | |
| "learning_rate": 0.000275763715348213, | |
| "loss": 1.2491, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.2000819600879208, | |
| "eval_loss": 1.2746073007583618, | |
| "eval_runtime": 216.9749, | |
| "eval_samples_per_second": 230.441, | |
| "eval_steps_per_second": 28.805, | |
| "step": 2014 | |
| }, | |
| { | |
| "epoch": 0.2006780334546177, | |
| "grad_norm": 0.5270881056785583, | |
| "learning_rate": 0.00027550272999528265, | |
| "loss": 1.2673, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.20167148906577917, | |
| "grad_norm": 0.49110147356987, | |
| "learning_rate": 0.0002752404718476116, | |
| "loss": 1.2561, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.20266494467694066, | |
| "grad_norm": 0.4829249680042267, | |
| "learning_rate": 0.0002749769435649094, | |
| "loss": 1.268, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.20365840028810211, | |
| "grad_norm": 0.5357676148414612, | |
| "learning_rate": 0.00027471214781976645, | |
| "loss": 1.2626, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.2046518558992636, | |
| "grad_norm": 0.5558575391769409, | |
| "learning_rate": 0.00027444608729762754, | |
| "loss": 1.2562, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.2056453115104251, | |
| "grad_norm": 0.5351165533065796, | |
| "learning_rate": 0.000274178764696764, | |
| "loss": 1.2741, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.20663876712158655, | |
| "grad_norm": 0.5279297232627869, | |
| "learning_rate": 0.0002739101827282468, | |
| "loss": 1.2431, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.20763222273274803, | |
| "grad_norm": 0.48787686228752136, | |
| "learning_rate": 0.0002736403441159188, | |
| "loss": 1.2672, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.2086256783439095, | |
| "grad_norm": 0.519707977771759, | |
| "learning_rate": 0.0002733692515963672, | |
| "loss": 1.264, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.20961913395507098, | |
| "grad_norm": 0.48795512318611145, | |
| "learning_rate": 0.0002730969079188958, | |
| "loss": 1.2605, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.21061258956623244, | |
| "grad_norm": 0.5582920908927917, | |
| "learning_rate": 0.00027282331584549716, | |
| "loss": 1.2645, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.21160604517739393, | |
| "grad_norm": 0.5244210958480835, | |
| "learning_rate": 0.00027254847815082444, | |
| "loss": 1.2577, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.21259950078855538, | |
| "grad_norm": 0.510850191116333, | |
| "learning_rate": 0.0002722723976221634, | |
| "loss": 1.2462, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.21359295639971687, | |
| "grad_norm": 0.49397844076156616, | |
| "learning_rate": 0.00027199507705940413, | |
| "loss": 1.2725, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.21458641201087833, | |
| "grad_norm": 0.53153395652771, | |
| "learning_rate": 0.0002717165192750125, | |
| "loss": 1.2727, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.21557986762203982, | |
| "grad_norm": 0.5746108293533325, | |
| "learning_rate": 0.000271436727094002, | |
| "loss": 1.2576, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.21657332323320128, | |
| "grad_norm": 0.5098176002502441, | |
| "learning_rate": 0.0002711557033539044, | |
| "loss": 1.2503, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.21756677884436276, | |
| "grad_norm": 0.5269674062728882, | |
| "learning_rate": 0.0002708734509047419, | |
| "loss": 1.269, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.21856023445552422, | |
| "grad_norm": 0.5084241032600403, | |
| "learning_rate": 0.0002705899726089973, | |
| "loss": 1.249, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.2195536900666857, | |
| "grad_norm": 0.49269863963127136, | |
| "learning_rate": 0.00027030527134158585, | |
| "loss": 1.2597, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.2205471456778472, | |
| "grad_norm": 0.5180286169052124, | |
| "learning_rate": 0.00027001934998982524, | |
| "loss": 1.2289, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.22154060128900865, | |
| "grad_norm": 0.5060125589370728, | |
| "learning_rate": 0.0002697322114534071, | |
| "loss": 1.2557, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.22253405690017014, | |
| "grad_norm": 0.4967941641807556, | |
| "learning_rate": 0.0002694438586443669, | |
| "loss": 1.2596, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.2235275125113316, | |
| "grad_norm": 0.48990583419799805, | |
| "learning_rate": 0.0002691542944870549, | |
| "loss": 1.2463, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.2245209681224931, | |
| "grad_norm": 0.5117115378379822, | |
| "learning_rate": 0.00026886352191810627, | |
| "loss": 1.2314, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.22551442373365455, | |
| "grad_norm": 0.5112183690071106, | |
| "learning_rate": 0.0002685715438864116, | |
| "loss": 1.2411, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.22650787934481603, | |
| "grad_norm": 0.5363621711730957, | |
| "learning_rate": 0.00026827836335308644, | |
| "loss": 1.2472, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.2275013349559775, | |
| "grad_norm": 0.4729025661945343, | |
| "learning_rate": 0.00026798398329144166, | |
| "loss": 1.236, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.22849479056713898, | |
| "grad_norm": 0.5384593605995178, | |
| "learning_rate": 0.0002676884066869535, | |
| "loss": 1.2667, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.22948824617830044, | |
| "grad_norm": 0.48534199595451355, | |
| "learning_rate": 0.0002673916365372327, | |
| "loss": 1.2261, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.23048170178946192, | |
| "grad_norm": 0.4830830991268158, | |
| "learning_rate": 0.0002670936758519946, | |
| "loss": 1.2634, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.23147515740062338, | |
| "grad_norm": 0.5320212244987488, | |
| "learning_rate": 0.0002667945276530284, | |
| "loss": 1.2364, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.23246861301178487, | |
| "grad_norm": 0.4748252034187317, | |
| "learning_rate": 0.0002664941949741665, | |
| "loss": 1.2546, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.23346206862294633, | |
| "grad_norm": 0.5805224776268005, | |
| "learning_rate": 0.0002661926808612539, | |
| "loss": 1.2569, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.23445552423410781, | |
| "grad_norm": 0.4791530966758728, | |
| "learning_rate": 0.000265889988372117, | |
| "loss": 1.2639, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.2354489798452693, | |
| "grad_norm": 0.45609885454177856, | |
| "learning_rate": 0.0002655861205765331, | |
| "loss": 1.2589, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.23644243545643076, | |
| "grad_norm": 0.5291593670845032, | |
| "learning_rate": 0.00026528108055619856, | |
| "loss": 1.2485, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.23743589106759225, | |
| "grad_norm": 0.4872785806655884, | |
| "learning_rate": 0.0002649748714046983, | |
| "loss": 1.2436, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.2384293466787537, | |
| "grad_norm": 0.4877389371395111, | |
| "learning_rate": 0.0002646674962274739, | |
| "loss": 1.2463, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.2394228022899152, | |
| "grad_norm": 0.5236742496490479, | |
| "learning_rate": 0.0002643589581417922, | |
| "loss": 1.2537, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.24041625790107665, | |
| "grad_norm": 0.5331168174743652, | |
| "learning_rate": 0.0002640492602767139, | |
| "loss": 1.2531, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.24140971351223814, | |
| "grad_norm": 0.5398425459861755, | |
| "learning_rate": 0.00026373840577306165, | |
| "loss": 1.2413, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.2424031691233996, | |
| "grad_norm": 0.5574949979782104, | |
| "learning_rate": 0.00026342639778338813, | |
| "loss": 1.244, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.24339662473456108, | |
| "grad_norm": 0.4968748390674591, | |
| "learning_rate": 0.0002631132394719443, | |
| "loss": 1.2412, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.24439008034572254, | |
| "grad_norm": 0.5098783373832703, | |
| "learning_rate": 0.0002627989340146471, | |
| "loss": 1.2671, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.24538353595688403, | |
| "grad_norm": 0.4852953255176544, | |
| "learning_rate": 0.0002624834845990475, | |
| "loss": 1.2457, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.2463769915680455, | |
| "grad_norm": 0.5067562460899353, | |
| "learning_rate": 0.00026216689442429774, | |
| "loss": 1.242, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.24737044717920698, | |
| "grad_norm": 0.46974197030067444, | |
| "learning_rate": 0.0002618491667011193, | |
| "loss": 1.2425, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.24836390279036843, | |
| "grad_norm": 0.48239001631736755, | |
| "learning_rate": 0.0002615303046517703, | |
| "loss": 1.2468, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.24935735840152992, | |
| "grad_norm": 0.46368446946144104, | |
| "learning_rate": 0.00026121031151001245, | |
| "loss": 1.2401, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.2503508140126914, | |
| "grad_norm": 0.4512680470943451, | |
| "learning_rate": 0.00026088919052107867, | |
| "loss": 1.2537, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.25134426962385287, | |
| "grad_norm": 0.5069119334220886, | |
| "learning_rate": 0.00026056694494164, | |
| "loss": 1.2514, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.25233772523501435, | |
| "grad_norm": 0.48463037610054016, | |
| "learning_rate": 0.00026024357803977263, | |
| "loss": 1.238, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.25333118084617584, | |
| "grad_norm": 0.5035899877548218, | |
| "learning_rate": 0.0002599190930949247, | |
| "loss": 1.2601, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.25432463645733727, | |
| "grad_norm": 0.5174955725669861, | |
| "learning_rate": 0.00025959349339788295, | |
| "loss": 1.251, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.25531809206849876, | |
| "grad_norm": 0.48764464259147644, | |
| "learning_rate": 0.00025926678225073963, | |
| "loss": 1.2406, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.25631154767966025, | |
| "grad_norm": 0.48704472184181213, | |
| "learning_rate": 0.00025893896296685866, | |
| "loss": 1.2393, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.25730500329082173, | |
| "grad_norm": 0.5210717916488647, | |
| "learning_rate": 0.00025861003887084246, | |
| "loss": 1.2467, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.25829845890198316, | |
| "grad_norm": 0.49823907017707825, | |
| "learning_rate": 0.0002582800132984977, | |
| "loss": 1.243, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.25929191451314465, | |
| "grad_norm": 0.5364810228347778, | |
| "learning_rate": 0.0002579488895968019, | |
| "loss": 1.2313, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.26028537012430614, | |
| "grad_norm": 0.45561811327934265, | |
| "learning_rate": 0.0002576166711238694, | |
| "loss": 1.2194, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.2612788257354676, | |
| "grad_norm": 0.4552469849586487, | |
| "learning_rate": 0.0002572833612489169, | |
| "loss": 1.2601, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.26227228134662905, | |
| "grad_norm": 0.542340099811554, | |
| "learning_rate": 0.00025694896335223024, | |
| "loss": 1.2461, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.26326573695779054, | |
| "grad_norm": 0.5542205572128296, | |
| "learning_rate": 0.00025661348082512884, | |
| "loss": 1.2374, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.26425919256895203, | |
| "grad_norm": 0.4834555685520172, | |
| "learning_rate": 0.0002562769170699324, | |
| "loss": 1.2396, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.2652526481801135, | |
| "grad_norm": 0.5201283693313599, | |
| "learning_rate": 0.00025593927549992594, | |
| "loss": 1.2332, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.266246103791275, | |
| "grad_norm": 0.5460206270217896, | |
| "learning_rate": 0.000255600559539325, | |
| "loss": 1.2342, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.26723955940243643, | |
| "grad_norm": 0.46429523825645447, | |
| "learning_rate": 0.0002552607726232413, | |
| "loss": 1.2316, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.2682330150135979, | |
| "grad_norm": 0.5969310998916626, | |
| "learning_rate": 0.0002549199181976477, | |
| "loss": 1.2501, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.2692264706247594, | |
| "grad_norm": 0.48196089267730713, | |
| "learning_rate": 0.00025457799971934325, | |
| "loss": 1.2389, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.2702199262359209, | |
| "grad_norm": 0.5023167133331299, | |
| "learning_rate": 0.0002542350206559182, | |
| "loss": 1.2322, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.2712133818470823, | |
| "grad_norm": 0.47286665439605713, | |
| "learning_rate": 0.0002538909844857187, | |
| "loss": 1.2618, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.2722068374582438, | |
| "grad_norm": 0.5069117546081543, | |
| "learning_rate": 0.00025354589469781186, | |
| "loss": 1.244, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.2732002930694053, | |
| "grad_norm": 0.4682158827781677, | |
| "learning_rate": 0.0002531997547919499, | |
| "loss": 1.2455, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.2741937486805668, | |
| "grad_norm": 0.5212481617927551, | |
| "learning_rate": 0.0002528525682785349, | |
| "loss": 1.2362, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.2751872042917282, | |
| "grad_norm": 0.5417184233665466, | |
| "learning_rate": 0.0002525043386785833, | |
| "loss": 1.2382, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.2761806599028897, | |
| "grad_norm": 0.46628710627555847, | |
| "learning_rate": 0.00025215506952369003, | |
| "loss": 1.2395, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.2771741155140512, | |
| "grad_norm": 0.48039111495018005, | |
| "learning_rate": 0.0002518047643559927, | |
| "loss": 1.2459, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.2781675711252127, | |
| "grad_norm": 0.4942474961280823, | |
| "learning_rate": 0.00025145342672813574, | |
| "loss": 1.2441, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.27916102673637416, | |
| "grad_norm": 0.4979240894317627, | |
| "learning_rate": 0.00025110106020323437, | |
| "loss": 1.2372, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.2801544823475356, | |
| "grad_norm": 0.4905546009540558, | |
| "learning_rate": 0.0002507476683548384, | |
| "loss": 1.2351, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.2811479379586971, | |
| "grad_norm": 0.4817952513694763, | |
| "learning_rate": 0.00025039325476689607, | |
| "loss": 1.2204, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.28214139356985857, | |
| "grad_norm": 0.4718853533267975, | |
| "learning_rate": 0.00025003782303371767, | |
| "loss": 1.2331, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.28313484918102005, | |
| "grad_norm": 0.48936033248901367, | |
| "learning_rate": 0.00024968137675993897, | |
| "loss": 1.2367, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.2841283047921815, | |
| "grad_norm": 0.503688633441925, | |
| "learning_rate": 0.00024932391956048497, | |
| "loss": 1.2189, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.28512176040334297, | |
| "grad_norm": 0.4779193699359894, | |
| "learning_rate": 0.00024896545506053285, | |
| "loss": 1.2485, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.28611521601450446, | |
| "grad_norm": 0.4970162808895111, | |
| "learning_rate": 0.00024860598689547555, | |
| "loss": 1.2167, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.28710867162566595, | |
| "grad_norm": 0.5190041065216064, | |
| "learning_rate": 0.0002482455187108846, | |
| "loss": 1.196, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.2881021272368274, | |
| "grad_norm": 0.5404960513114929, | |
| "learning_rate": 0.00024788405416247347, | |
| "loss": 1.2235, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.28909558284798886, | |
| "grad_norm": 0.4441899359226227, | |
| "learning_rate": 0.00024752159691606015, | |
| "loss": 1.2471, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.29008903845915035, | |
| "grad_norm": 0.48505038022994995, | |
| "learning_rate": 0.0002471581506475303, | |
| "loss": 1.2312, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.29108249407031184, | |
| "grad_norm": 0.49762651324272156, | |
| "learning_rate": 0.00024679371904279973, | |
| "loss": 1.2261, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.29207594968147327, | |
| "grad_norm": 0.5253481268882751, | |
| "learning_rate": 0.0002464283057977771, | |
| "loss": 1.2405, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.29306940529263475, | |
| "grad_norm": 0.4968172311782837, | |
| "learning_rate": 0.00024606191461832655, | |
| "loss": 1.2225, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.29406286090379624, | |
| "grad_norm": 0.4646175503730774, | |
| "learning_rate": 0.00024569454922022987, | |
| "loss": 1.2109, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.29505631651495773, | |
| "grad_norm": 0.49472692608833313, | |
| "learning_rate": 0.00024532621332914897, | |
| "loss": 1.2191, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.2960497721261192, | |
| "grad_norm": 0.49351736903190613, | |
| "learning_rate": 0.0002449569106805882, | |
| "loss": 1.216, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.29704322773728065, | |
| "grad_norm": 0.49852266907691956, | |
| "learning_rate": 0.00024458664501985624, | |
| "loss": 1.2301, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.29803668334844213, | |
| "grad_norm": 0.5016636848449707, | |
| "learning_rate": 0.0002442154201020283, | |
| "loss": 1.2289, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.2990301389596036, | |
| "grad_norm": 0.4916824996471405, | |
| "learning_rate": 0.00024384323969190789, | |
| "loss": 1.2274, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.3000235945707651, | |
| "grad_norm": 0.46829256415367126, | |
| "learning_rate": 0.0002434701075639888, | |
| "loss": 1.2118, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.30012294013188123, | |
| "eval_loss": 1.2400528192520142, | |
| "eval_runtime": 217.1452, | |
| "eval_samples_per_second": 230.261, | |
| "eval_steps_per_second": 28.783, | |
| "step": 3021 | |
| }, | |
| { | |
| "epoch": 0.30101705018192654, | |
| "grad_norm": 0.44890084862709045, | |
| "learning_rate": 0.0002430960275024166, | |
| "loss": 1.2103, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.302010505793088, | |
| "grad_norm": 0.4859289526939392, | |
| "learning_rate": 0.00024272100330095055, | |
| "loss": 1.2367, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.3030039614042495, | |
| "grad_norm": 0.4968155026435852, | |
| "learning_rate": 0.00024234503876292485, | |
| "loss": 1.235, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.303997417015411, | |
| "grad_norm": 0.5458197593688965, | |
| "learning_rate": 0.00024196813770121026, | |
| "loss": 1.2294, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.30499087262657243, | |
| "grad_norm": 0.4604020118713379, | |
| "learning_rate": 0.00024159030393817538, | |
| "loss": 1.222, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.3059843282377339, | |
| "grad_norm": 0.5177088379859924, | |
| "learning_rate": 0.0002412115413056477, | |
| "loss": 1.2068, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.3069777838488954, | |
| "grad_norm": 0.47766199707984924, | |
| "learning_rate": 0.00024083185364487522, | |
| "loss": 1.2256, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.3079712394600569, | |
| "grad_norm": 0.5014012455940247, | |
| "learning_rate": 0.0002404512448064868, | |
| "loss": 1.1924, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.3089646950712184, | |
| "grad_norm": 0.4730978310108185, | |
| "learning_rate": 0.00024006971865045377, | |
| "loss": 1.2349, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.3099581506823798, | |
| "grad_norm": 0.45509788393974304, | |
| "learning_rate": 0.0002396872790460504, | |
| "loss": 1.2177, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.3109516062935413, | |
| "grad_norm": 0.4922787845134735, | |
| "learning_rate": 0.00023930392987181473, | |
| "loss": 1.2151, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.3119450619047028, | |
| "grad_norm": 0.4871610403060913, | |
| "learning_rate": 0.00023891967501550937, | |
| "loss": 1.2133, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.31293851751586427, | |
| "grad_norm": 0.4459008276462555, | |
| "learning_rate": 0.00023853451837408185, | |
| "loss": 1.2097, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.3139319731270257, | |
| "grad_norm": 0.41748374700546265, | |
| "learning_rate": 0.00023814846385362525, | |
| "loss": 1.2236, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.3149254287381872, | |
| "grad_norm": 0.5209752321243286, | |
| "learning_rate": 0.00023776151536933864, | |
| "loss": 1.2015, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.31591888434934867, | |
| "grad_norm": 0.4768379032611847, | |
| "learning_rate": 0.00023737367684548716, | |
| "loss": 1.2233, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.31691233996051016, | |
| "grad_norm": 0.46641460061073303, | |
| "learning_rate": 0.00023698495221536245, | |
| "loss": 1.2144, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.3179057955716716, | |
| "grad_norm": 0.49821555614471436, | |
| "learning_rate": 0.00023659534542124261, | |
| "loss": 1.2003, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.3188992511828331, | |
| "grad_norm": 0.5253980159759521, | |
| "learning_rate": 0.0002362048604143522, | |
| "loss": 1.2165, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.31989270679399456, | |
| "grad_norm": 0.5121644735336304, | |
| "learning_rate": 0.00023581350115482245, | |
| "loss": 1.2076, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.32088616240515605, | |
| "grad_norm": 0.499426007270813, | |
| "learning_rate": 0.00023542127161165063, | |
| "loss": 1.2294, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.3218796180163175, | |
| "grad_norm": 0.48408859968185425, | |
| "learning_rate": 0.00023502817576266023, | |
| "loss": 1.2156, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.32287307362747897, | |
| "grad_norm": 0.4970872402191162, | |
| "learning_rate": 0.00023463421759446033, | |
| "loss": 1.2057, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.32386652923864045, | |
| "grad_norm": 0.5192016363143921, | |
| "learning_rate": 0.00023423940110240538, | |
| "loss": 1.234, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.32485998484980194, | |
| "grad_norm": 0.47482579946517944, | |
| "learning_rate": 0.00023384373029055445, | |
| "loss": 1.2087, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.32585344046096343, | |
| "grad_norm": 0.4967539608478546, | |
| "learning_rate": 0.00023344720917163093, | |
| "loss": 1.203, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.32684689607212486, | |
| "grad_norm": 0.46486181020736694, | |
| "learning_rate": 0.00023304984176698157, | |
| "loss": 1.2044, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.32784035168328635, | |
| "grad_norm": 0.45617178082466125, | |
| "learning_rate": 0.00023265163210653569, | |
| "loss": 1.2189, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.32883380729444783, | |
| "grad_norm": 0.46393707394599915, | |
| "learning_rate": 0.00023225258422876455, | |
| "loss": 1.2091, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.3298272629056093, | |
| "grad_norm": 0.4585002064704895, | |
| "learning_rate": 0.00023185270218064021, | |
| "loss": 1.2031, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.33082071851677075, | |
| "grad_norm": 0.4875582158565521, | |
| "learning_rate": 0.00023145199001759452, | |
| "loss": 1.2075, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.33181417412793224, | |
| "grad_norm": 0.483347624540329, | |
| "learning_rate": 0.00023105045180347788, | |
| "loss": 1.2256, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.3328076297390937, | |
| "grad_norm": 0.5154065489768982, | |
| "learning_rate": 0.00023064809161051842, | |
| "loss": 1.2059, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.3338010853502552, | |
| "grad_norm": 0.5070950984954834, | |
| "learning_rate": 0.00023024491351928019, | |
| "loss": 1.2033, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.33479454096141664, | |
| "grad_norm": 0.5440912842750549, | |
| "learning_rate": 0.00022984092161862206, | |
| "loss": 1.2197, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.33578799657257813, | |
| "grad_norm": 0.5029067993164062, | |
| "learning_rate": 0.00022943612000565625, | |
| "loss": 1.1932, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.3367814521837396, | |
| "grad_norm": 0.5002272129058838, | |
| "learning_rate": 0.00022903051278570677, | |
| "loss": 1.2088, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.3377749077949011, | |
| "grad_norm": 0.46260136365890503, | |
| "learning_rate": 0.00022862410407226762, | |
| "loss": 1.2287, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.3387683634060626, | |
| "grad_norm": 0.558641791343689, | |
| "learning_rate": 0.00022821689798696132, | |
| "loss": 1.2198, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.339761819017224, | |
| "grad_norm": 0.4847974181175232, | |
| "learning_rate": 0.00022780889865949689, | |
| "loss": 1.2059, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.3407552746283855, | |
| "grad_norm": 0.4504203796386719, | |
| "learning_rate": 0.00022740011022762827, | |
| "loss": 1.2089, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.341748730239547, | |
| "grad_norm": 0.5143907070159912, | |
| "learning_rate": 0.00022699053683711187, | |
| "loss": 1.2274, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.3427421858507085, | |
| "grad_norm": 0.4598551094532013, | |
| "learning_rate": 0.00022658018264166502, | |
| "loss": 1.2136, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.3437356414618699, | |
| "grad_norm": 0.45654499530792236, | |
| "learning_rate": 0.00022616905180292363, | |
| "loss": 1.1994, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.3447290970730314, | |
| "grad_norm": 0.48295536637306213, | |
| "learning_rate": 0.00022575714849039984, | |
| "loss": 1.21, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.3457225526841929, | |
| "grad_norm": 0.5002532005310059, | |
| "learning_rate": 0.00022534447688144006, | |
| "loss": 1.2229, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.34671600829535437, | |
| "grad_norm": 0.45252901315689087, | |
| "learning_rate": 0.0002249310411611824, | |
| "loss": 1.1866, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.3477094639065158, | |
| "grad_norm": 0.46162939071655273, | |
| "learning_rate": 0.00022451684552251414, | |
| "loss": 1.2159, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.3487029195176773, | |
| "grad_norm": 0.46276578307151794, | |
| "learning_rate": 0.00022410189416602943, | |
| "loss": 1.2128, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.3496963751288388, | |
| "grad_norm": 0.5181559920310974, | |
| "learning_rate": 0.00022368619129998655, | |
| "loss": 1.2188, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.35068983074000026, | |
| "grad_norm": 0.4349026381969452, | |
| "learning_rate": 0.00022326974114026535, | |
| "loss": 1.2045, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.3516832863511617, | |
| "grad_norm": 0.45742759108543396, | |
| "learning_rate": 0.00022285254791032423, | |
| "loss": 1.2024, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.3526767419623232, | |
| "grad_norm": 0.48607686161994934, | |
| "learning_rate": 0.0002224346158411577, | |
| "loss": 1.201, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.35367019757348467, | |
| "grad_norm": 0.4831947982311249, | |
| "learning_rate": 0.00022201594917125306, | |
| "loss": 1.1895, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.35466365318464615, | |
| "grad_norm": 0.5034534335136414, | |
| "learning_rate": 0.00022159655214654787, | |
| "loss": 1.2008, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.35565710879580764, | |
| "grad_norm": 0.47319284081459045, | |
| "learning_rate": 0.00022117642902038629, | |
| "loss": 1.1967, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.3566505644069691, | |
| "grad_norm": 0.5058645009994507, | |
| "learning_rate": 0.00022075558405347659, | |
| "loss": 1.2149, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.35764402001813056, | |
| "grad_norm": 0.4877967834472656, | |
| "learning_rate": 0.0002203340215138475, | |
| "loss": 1.2252, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.35863747562929205, | |
| "grad_norm": 0.4703255593776703, | |
| "learning_rate": 0.00021991174567680512, | |
| "loss": 1.195, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.35963093124045353, | |
| "grad_norm": 0.4860426187515259, | |
| "learning_rate": 0.00021948876082488936, | |
| "loss": 1.1959, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.36062438685161496, | |
| "grad_norm": 0.46194490790367126, | |
| "learning_rate": 0.0002190650712478309, | |
| "loss": 1.2074, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.36161784246277645, | |
| "grad_norm": 0.44126129150390625, | |
| "learning_rate": 0.0002186406812425073, | |
| "loss": 1.1912, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.36261129807393794, | |
| "grad_norm": 0.48812511563301086, | |
| "learning_rate": 0.00021821559511289948, | |
| "loss": 1.2068, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.3636047536850994, | |
| "grad_norm": 0.4943760931491852, | |
| "learning_rate": 0.0002177898171700483, | |
| "loss": 1.1923, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.36459820929626086, | |
| "grad_norm": 0.4492897391319275, | |
| "learning_rate": 0.00021736335173201066, | |
| "loss": 1.1859, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.36559166490742234, | |
| "grad_norm": 0.45745688676834106, | |
| "learning_rate": 0.00021693620312381568, | |
| "loss": 1.1978, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.36658512051858383, | |
| "grad_norm": 0.5050104856491089, | |
| "learning_rate": 0.00021650837567742094, | |
| "loss": 1.2172, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.3675785761297453, | |
| "grad_norm": 0.5013807415962219, | |
| "learning_rate": 0.00021607987373166855, | |
| "loss": 1.2039, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.3685720317409068, | |
| "grad_norm": 0.4632636308670044, | |
| "learning_rate": 0.00021565070163224095, | |
| "loss": 1.2005, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.36956548735206823, | |
| "grad_norm": 0.46821290254592896, | |
| "learning_rate": 0.0002152208637316172, | |
| "loss": 1.2227, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.3705589429632297, | |
| "grad_norm": 0.4609948396682739, | |
| "learning_rate": 0.00021479036438902847, | |
| "loss": 1.2018, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.3715523985743912, | |
| "grad_norm": 0.5113796591758728, | |
| "learning_rate": 0.00021435920797041404, | |
| "loss": 1.1925, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.3725458541855527, | |
| "grad_norm": 0.47701483964920044, | |
| "learning_rate": 0.00021392739884837697, | |
| "loss": 1.1992, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.3735393097967141, | |
| "grad_norm": 0.4520336389541626, | |
| "learning_rate": 0.00021349494140213986, | |
| "loss": 1.202, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.3745327654078756, | |
| "grad_norm": 0.4652255177497864, | |
| "learning_rate": 0.00021306184001750012, | |
| "loss": 1.1829, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.3755262210190371, | |
| "grad_norm": 0.4625130295753479, | |
| "learning_rate": 0.00021262809908678584, | |
| "loss": 1.1927, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.3765196766301986, | |
| "grad_norm": 0.44085580110549927, | |
| "learning_rate": 0.00021219372300881117, | |
| "loss": 1.2191, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.37751313224136, | |
| "grad_norm": 0.47911298274993896, | |
| "learning_rate": 0.0002117587161888315, | |
| "loss": 1.1941, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.3785065878525215, | |
| "grad_norm": 0.49052807688713074, | |
| "learning_rate": 0.00021132308303849906, | |
| "loss": 1.214, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.379500043463683, | |
| "grad_norm": 0.48691338300704956, | |
| "learning_rate": 0.000210886827975818, | |
| "loss": 1.2108, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.3804934990748445, | |
| "grad_norm": 0.44401276111602783, | |
| "learning_rate": 0.0002104499554250996, | |
| "loss": 1.1792, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.38148695468600596, | |
| "grad_norm": 0.4931083917617798, | |
| "learning_rate": 0.00021001246981691758, | |
| "loss": 1.2012, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.3824804102971674, | |
| "grad_norm": 0.472159206867218, | |
| "learning_rate": 0.0002095743755880628, | |
| "loss": 1.2112, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.3834738659083289, | |
| "grad_norm": 0.50935298204422, | |
| "learning_rate": 0.0002091356771814987, | |
| "loss": 1.206, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.38446732151949037, | |
| "grad_norm": 0.5085238814353943, | |
| "learning_rate": 0.0002086963790463159, | |
| "loss": 1.2066, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.38546077713065185, | |
| "grad_norm": 0.47772541642189026, | |
| "learning_rate": 0.0002082564856376873, | |
| "loss": 1.1908, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.3864542327418133, | |
| "grad_norm": 0.5096874833106995, | |
| "learning_rate": 0.0002078160014168227, | |
| "loss": 1.197, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.3874476883529748, | |
| "grad_norm": 0.5250852704048157, | |
| "learning_rate": 0.00020737493085092382, | |
| "loss": 1.1967, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.38844114396413626, | |
| "grad_norm": 0.45944342017173767, | |
| "learning_rate": 0.00020693327841313867, | |
| "loss": 1.2116, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.38943459957529775, | |
| "grad_norm": 0.4569138288497925, | |
| "learning_rate": 0.00020649104858251648, | |
| "loss": 1.2057, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.3904280551864592, | |
| "grad_norm": 0.4284946322441101, | |
| "learning_rate": 0.00020604824584396204, | |
| "loss": 1.1949, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 0.39142151079762066, | |
| "grad_norm": 0.4908120036125183, | |
| "learning_rate": 0.0002056048746881905, | |
| "loss": 1.1973, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.39241496640878215, | |
| "grad_norm": 0.49430763721466064, | |
| "learning_rate": 0.00020516093961168137, | |
| "loss": 1.2042, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.39340842201994364, | |
| "grad_norm": 0.45789289474487305, | |
| "learning_rate": 0.00020471644511663352, | |
| "loss": 1.1794, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.39440187763110507, | |
| "grad_norm": 0.5000009536743164, | |
| "learning_rate": 0.00020427139571091893, | |
| "loss": 1.1859, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 0.39539533324226656, | |
| "grad_norm": 0.4725204110145569, | |
| "learning_rate": 0.00020382579590803748, | |
| "loss": 1.1818, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.39638878885342804, | |
| "grad_norm": 0.4737090766429901, | |
| "learning_rate": 0.00020337965022707084, | |
| "loss": 1.1922, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 0.39738224446458953, | |
| "grad_norm": 0.44001784920692444, | |
| "learning_rate": 0.00020293296319263664, | |
| "loss": 1.1988, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.398375700075751, | |
| "grad_norm": 0.48931002616882324, | |
| "learning_rate": 0.00020248573933484286, | |
| "loss": 1.1952, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 0.39936915568691245, | |
| "grad_norm": 0.48461276292800903, | |
| "learning_rate": 0.00020203798318924152, | |
| "loss": 1.1918, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.4001639201758416, | |
| "eval_loss": 1.209017038345337, | |
| "eval_runtime": 217.2474, | |
| "eval_samples_per_second": 230.152, | |
| "eval_steps_per_second": 28.769, | |
| "step": 4028 | |
| }, | |
| { | |
| "epoch": 0.40036261129807393, | |
| "grad_norm": 0.4797476530075073, | |
| "learning_rate": 0.00020158969929678306, | |
| "loss": 1.1786, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 0.4013560669092354, | |
| "grad_norm": 0.46265390515327454, | |
| "learning_rate": 0.00020114089220377005, | |
| "loss": 1.2022, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.4023495225203969, | |
| "grad_norm": 0.4856548607349396, | |
| "learning_rate": 0.00020069156646181107, | |
| "loss": 1.2066, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.40334297813155834, | |
| "grad_norm": 0.5205573439598083, | |
| "learning_rate": 0.0002002417266277746, | |
| "loss": 1.2003, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.4043364337427198, | |
| "grad_norm": 0.484475314617157, | |
| "learning_rate": 0.00019979137726374295, | |
| "loss": 1.1849, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 0.4053298893538813, | |
| "grad_norm": 0.5231068730354309, | |
| "learning_rate": 0.0001993405229369658, | |
| "loss": 1.1948, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.4063233449650428, | |
| "grad_norm": 0.46499723196029663, | |
| "learning_rate": 0.0001988891682198139, | |
| "loss": 1.1968, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 0.40731680057620423, | |
| "grad_norm": 0.4803023934364319, | |
| "learning_rate": 0.00019843731768973274, | |
| "loss": 1.1899, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.4083102561873657, | |
| "grad_norm": 0.43668457865715027, | |
| "learning_rate": 0.00019798497592919624, | |
| "loss": 1.1801, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 0.4093037117985272, | |
| "grad_norm": 0.45928719639778137, | |
| "learning_rate": 0.00019753214752566006, | |
| "loss": 1.1942, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.4102971674096887, | |
| "grad_norm": 0.4647465646266937, | |
| "learning_rate": 0.0001970788370715152, | |
| "loss": 1.1839, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 0.4112906230208502, | |
| "grad_norm": 0.43589547276496887, | |
| "learning_rate": 0.0001966250491640415, | |
| "loss": 1.1944, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.4122840786320116, | |
| "grad_norm": 0.5093525052070618, | |
| "learning_rate": 0.0001961707884053608, | |
| "loss": 1.1657, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.4132775342431731, | |
| "grad_norm": 0.4544735848903656, | |
| "learning_rate": 0.0001957160594023905, | |
| "loss": 1.1793, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.4142709898543346, | |
| "grad_norm": 0.43822965025901794, | |
| "learning_rate": 0.00019526086676679663, | |
| "loss": 1.1679, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 0.41526444546549607, | |
| "grad_norm": 0.43124911189079285, | |
| "learning_rate": 0.00019480521511494727, | |
| "loss": 1.1797, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.4162579010766575, | |
| "grad_norm": 0.44309306144714355, | |
| "learning_rate": 0.0001943491090678656, | |
| "loss": 1.1674, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 0.417251356687819, | |
| "grad_norm": 0.4919864237308502, | |
| "learning_rate": 0.00019389255325118305, | |
| "loss": 1.1823, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.4182448122989805, | |
| "grad_norm": 0.4437265396118164, | |
| "learning_rate": 0.0001934355522950926, | |
| "loss": 1.1908, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 0.41923826791014196, | |
| "grad_norm": 0.5036196112632751, | |
| "learning_rate": 0.00019297811083430138, | |
| "loss": 1.1932, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 0.4202317235213034, | |
| "grad_norm": 0.4650508463382721, | |
| "learning_rate": 0.0001925202335079841, | |
| "loss": 1.1797, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 0.4212251791324649, | |
| "grad_norm": 0.44443807005882263, | |
| "learning_rate": 0.0001920619249597357, | |
| "loss": 1.2067, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 0.42221863474362636, | |
| "grad_norm": 0.49494221806526184, | |
| "learning_rate": 0.00019160318983752465, | |
| "loss": 1.2038, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.42321209035478785, | |
| "grad_norm": 0.48809465765953064, | |
| "learning_rate": 0.00019114403279364524, | |
| "loss": 1.1701, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 0.4242055459659493, | |
| "grad_norm": 0.4628167748451233, | |
| "learning_rate": 0.00019068445848467088, | |
| "loss": 1.1821, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 0.42519900157711077, | |
| "grad_norm": 0.5052282214164734, | |
| "learning_rate": 0.00019022447157140667, | |
| "loss": 1.166, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 0.42619245718827226, | |
| "grad_norm": 0.45449918508529663, | |
| "learning_rate": 0.00018976407671884217, | |
| "loss": 1.172, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 0.42718591279943374, | |
| "grad_norm": 0.463618665933609, | |
| "learning_rate": 0.000189303278596104, | |
| "loss": 1.1616, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.42817936841059523, | |
| "grad_norm": 0.48953506350517273, | |
| "learning_rate": 0.00018884208187640875, | |
| "loss": 1.2022, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 0.42917282402175666, | |
| "grad_norm": 0.5115376114845276, | |
| "learning_rate": 0.00018838049123701524, | |
| "loss": 1.18, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.43016627963291815, | |
| "grad_norm": 0.44180265069007874, | |
| "learning_rate": 0.00018791851135917739, | |
| "loss": 1.1851, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 0.43115973524407963, | |
| "grad_norm": 0.46057870984077454, | |
| "learning_rate": 0.00018745614692809644, | |
| "loss": 1.1856, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 0.4321531908552411, | |
| "grad_norm": 0.4646226167678833, | |
| "learning_rate": 0.0001869934026328738, | |
| "loss": 1.1756, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.43314664646640255, | |
| "grad_norm": 0.4839836657047272, | |
| "learning_rate": 0.00018653028316646317, | |
| "loss": 1.203, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 0.43414010207756404, | |
| "grad_norm": 0.44381341338157654, | |
| "learning_rate": 0.00018606679322562307, | |
| "loss": 1.1633, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 0.4351335576887255, | |
| "grad_norm": 0.48917341232299805, | |
| "learning_rate": 0.00018560293751086927, | |
| "loss": 1.1794, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 0.436127013299887, | |
| "grad_norm": 0.46939659118652344, | |
| "learning_rate": 0.00018513872072642708, | |
| "loss": 1.1754, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 0.43712046891104844, | |
| "grad_norm": 0.5408668518066406, | |
| "learning_rate": 0.00018467414758018358, | |
| "loss": 1.1794, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.43811392452220993, | |
| "grad_norm": 0.4510866105556488, | |
| "learning_rate": 0.0001842092227836398, | |
| "loss": 1.1839, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 0.4391073801333714, | |
| "grad_norm": 0.45456522703170776, | |
| "learning_rate": 0.00018374395105186342, | |
| "loss": 1.2092, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 0.4401008357445329, | |
| "grad_norm": 0.49070245027542114, | |
| "learning_rate": 0.00018327833710344012, | |
| "loss": 1.1782, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 0.4410942913556944, | |
| "grad_norm": 0.49562186002731323, | |
| "learning_rate": 0.00018281238566042655, | |
| "loss": 1.1865, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 0.4420877469668558, | |
| "grad_norm": 0.4716269373893738, | |
| "learning_rate": 0.00018234610144830189, | |
| "loss": 1.1852, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.4430812025780173, | |
| "grad_norm": 0.4782949388027191, | |
| "learning_rate": 0.00018187948919592023, | |
| "loss": 1.17, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 0.4440746581891788, | |
| "grad_norm": 0.6574309468269348, | |
| "learning_rate": 0.00018141255363546244, | |
| "loss": 1.1777, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 0.4450681138003403, | |
| "grad_norm": 0.4680328965187073, | |
| "learning_rate": 0.00018094529950238829, | |
| "loss": 1.1595, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.4460615694115017, | |
| "grad_norm": 0.44611451029777527, | |
| "learning_rate": 0.00018047773153538835, | |
| "loss": 1.1956, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 0.4470550250226632, | |
| "grad_norm": 0.4821056127548218, | |
| "learning_rate": 0.00018000985447633592, | |
| "loss": 1.1744, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.4480484806338247, | |
| "grad_norm": 0.4614666700363159, | |
| "learning_rate": 0.00017954167307023902, | |
| "loss": 1.1776, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 0.4490419362449862, | |
| "grad_norm": 0.4979506731033325, | |
| "learning_rate": 0.00017907319206519228, | |
| "loss": 1.1666, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.4500353918561476, | |
| "grad_norm": 0.4544410705566406, | |
| "learning_rate": 0.00017860441621232864, | |
| "loss": 1.189, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 0.4510288474673091, | |
| "grad_norm": 0.5314710736274719, | |
| "learning_rate": 0.00017813535026577128, | |
| "loss": 1.1726, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 0.4520223030784706, | |
| "grad_norm": 0.4279493987560272, | |
| "learning_rate": 0.0001776659989825854, | |
| "loss": 1.1875, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.45301575868963206, | |
| "grad_norm": 0.4388754963874817, | |
| "learning_rate": 0.00017719636712273005, | |
| "loss": 1.1734, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.4540092143007935, | |
| "grad_norm": 0.490273118019104, | |
| "learning_rate": 0.0001767264594490095, | |
| "loss": 1.1733, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 0.455002669911955, | |
| "grad_norm": 0.40806838870048523, | |
| "learning_rate": 0.00017625628072702562, | |
| "loss": 1.1559, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 0.45599612552311647, | |
| "grad_norm": 0.4387802183628082, | |
| "learning_rate": 0.00017578583572512877, | |
| "loss": 1.1799, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 0.45698958113427796, | |
| "grad_norm": 0.44947245717048645, | |
| "learning_rate": 0.00017531512921436997, | |
| "loss": 1.1874, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.45798303674543944, | |
| "grad_norm": 0.5053061842918396, | |
| "learning_rate": 0.00017484416596845226, | |
| "loss": 1.1804, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 0.4589764923566009, | |
| "grad_norm": 0.4740687310695648, | |
| "learning_rate": 0.00017437295076368252, | |
| "loss": 1.1537, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 0.45996994796776236, | |
| "grad_norm": 0.4325508177280426, | |
| "learning_rate": 0.0001739014883789228, | |
| "loss": 1.1573, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 0.46096340357892385, | |
| "grad_norm": 0.478100061416626, | |
| "learning_rate": 0.00017342978359554187, | |
| "loss": 1.1819, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.46195685919008533, | |
| "grad_norm": 0.4926578104496002, | |
| "learning_rate": 0.00017295784119736698, | |
| "loss": 1.1886, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.46295031480124677, | |
| "grad_norm": 0.5400477051734924, | |
| "learning_rate": 0.000172485665970635, | |
| "loss": 1.1698, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 0.46394377041240825, | |
| "grad_norm": 0.4423806965351105, | |
| "learning_rate": 0.00017201326270394418, | |
| "loss": 1.1771, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 0.46493722602356974, | |
| "grad_norm": 0.4588819742202759, | |
| "learning_rate": 0.00017154063618820533, | |
| "loss": 1.1817, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 0.4659306816347312, | |
| "grad_norm": 0.5755475759506226, | |
| "learning_rate": 0.00017106779121659344, | |
| "loss": 1.1763, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 0.46692413724589266, | |
| "grad_norm": 0.45726439356803894, | |
| "learning_rate": 0.00017059473258449894, | |
| "loss": 1.1706, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.46791759285705414, | |
| "grad_norm": 0.4874899089336395, | |
| "learning_rate": 0.00017012146508947916, | |
| "loss": 1.1842, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 0.46891104846821563, | |
| "grad_norm": 0.4870886206626892, | |
| "learning_rate": 0.00016964799353120963, | |
| "loss": 1.1556, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 0.4699045040793771, | |
| "grad_norm": 0.4287513494491577, | |
| "learning_rate": 0.00016917432271143534, | |
| "loss": 1.1887, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 0.4708979596905386, | |
| "grad_norm": 0.47464653849601746, | |
| "learning_rate": 0.00016870045743392207, | |
| "loss": 1.1768, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 0.47189141530170003, | |
| "grad_norm": 0.48040926456451416, | |
| "learning_rate": 0.0001682264025044079, | |
| "loss": 1.1398, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.4728848709128615, | |
| "grad_norm": 0.4692235589027405, | |
| "learning_rate": 0.00016775216273055406, | |
| "loss": 1.1681, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 0.473878326524023, | |
| "grad_norm": 0.4798252582550049, | |
| "learning_rate": 0.00016727774292189657, | |
| "loss": 1.1564, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 0.4748717821351845, | |
| "grad_norm": 0.4889027774333954, | |
| "learning_rate": 0.0001668031478897971, | |
| "loss": 1.1723, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 0.4758652377463459, | |
| "grad_norm": 0.498568058013916, | |
| "learning_rate": 0.00016632838244739454, | |
| "loss": 1.1648, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 0.4768586933575074, | |
| "grad_norm": 0.46477189660072327, | |
| "learning_rate": 0.00016585345140955593, | |
| "loss": 1.1548, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.4778521489686689, | |
| "grad_norm": 0.46384197473526, | |
| "learning_rate": 0.00016537835959282767, | |
| "loss": 1.1611, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 0.4788456045798304, | |
| "grad_norm": 0.4606354832649231, | |
| "learning_rate": 0.00016490311181538672, | |
| "loss": 1.1592, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 0.4798390601909918, | |
| "grad_norm": 0.46360450983047485, | |
| "learning_rate": 0.00016442771289699185, | |
| "loss": 1.1653, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 0.4808325158021533, | |
| "grad_norm": 0.47855210304260254, | |
| "learning_rate": 0.00016395216765893443, | |
| "loss": 1.1784, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 0.4818259714133148, | |
| "grad_norm": 0.4936244487762451, | |
| "learning_rate": 0.00016347648092398997, | |
| "loss": 1.1697, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.4828194270244763, | |
| "grad_norm": 0.5215084552764893, | |
| "learning_rate": 0.00016300065751636878, | |
| "loss": 1.1623, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 0.4838128826356377, | |
| "grad_norm": 0.49761301279067993, | |
| "learning_rate": 0.00016252470226166735, | |
| "loss": 1.1845, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 0.4848063382467992, | |
| "grad_norm": 0.47753623127937317, | |
| "learning_rate": 0.0001620486199868193, | |
| "loss": 1.1761, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 0.4857997938579607, | |
| "grad_norm": 0.4606369137763977, | |
| "learning_rate": 0.0001615724155200464, | |
| "loss": 1.1824, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 0.48679324946912217, | |
| "grad_norm": 0.4668729305267334, | |
| "learning_rate": 0.00016109609369080964, | |
| "loss": 1.1553, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.48778670508028366, | |
| "grad_norm": 0.47536182403564453, | |
| "learning_rate": 0.00016061965932976033, | |
| "loss": 1.1663, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 0.4887801606914451, | |
| "grad_norm": 0.4314727783203125, | |
| "learning_rate": 0.00016014311726869092, | |
| "loss": 1.1698, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 0.4897736163026066, | |
| "grad_norm": 0.44343844056129456, | |
| "learning_rate": 0.00015966647234048618, | |
| "loss": 1.1678, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 0.49076707191376806, | |
| "grad_norm": 0.45859286189079285, | |
| "learning_rate": 0.000159189729379074, | |
| "loss": 1.1873, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 0.49176052752492955, | |
| "grad_norm": 0.43602892756462097, | |
| "learning_rate": 0.0001587128932193767, | |
| "loss": 1.157, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.492753983136091, | |
| "grad_norm": 0.48735079169273376, | |
| "learning_rate": 0.00015823596869726152, | |
| "loss": 1.1682, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.49374743874725247, | |
| "grad_norm": 0.4638786017894745, | |
| "learning_rate": 0.0001577589606494921, | |
| "loss": 1.166, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 0.49474089435841395, | |
| "grad_norm": 0.44135597348213196, | |
| "learning_rate": 0.0001572818739136789, | |
| "loss": 1.1742, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 0.49573434996957544, | |
| "grad_norm": 0.4518062174320221, | |
| "learning_rate": 0.00015680471332823055, | |
| "loss": 1.1699, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 0.49672780558073687, | |
| "grad_norm": 0.4678098261356354, | |
| "learning_rate": 0.00015632748373230462, | |
| "loss": 1.1634, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.49772126119189836, | |
| "grad_norm": 0.4754001498222351, | |
| "learning_rate": 0.00015585018996575853, | |
| "loss": 1.1536, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 0.49871471680305984, | |
| "grad_norm": 0.5086048245429993, | |
| "learning_rate": 0.00015537283686910047, | |
| "loss": 1.1584, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 0.49970817241422133, | |
| "grad_norm": 0.4387681484222412, | |
| "learning_rate": 0.00015489542928344037, | |
| "loss": 1.1691, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 0.500204900219802, | |
| "eval_loss": 1.1872472763061523, | |
| "eval_runtime": 217.2564, | |
| "eval_samples_per_second": 230.143, | |
| "eval_steps_per_second": 28.768, | |
| "step": 5035 | |
| }, | |
| { | |
| "epoch": 0.5007016280253828, | |
| "grad_norm": 0.5286032557487488, | |
| "learning_rate": 0.00015441797205044083, | |
| "loss": 1.1789, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 0.5016950836365442, | |
| "grad_norm": 0.4659021198749542, | |
| "learning_rate": 0.00015394047001226775, | |
| "loss": 1.1713, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.5026885392477057, | |
| "grad_norm": 0.442771852016449, | |
| "learning_rate": 0.0001534629280115416, | |
| "loss": 1.1679, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 0.5036819948588672, | |
| "grad_norm": 0.5038478374481201, | |
| "learning_rate": 0.00015298535089128813, | |
| "loss": 1.1657, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 0.5046754504700287, | |
| "grad_norm": 0.4490121304988861, | |
| "learning_rate": 0.00015250774349488916, | |
| "loss": 1.1552, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 0.5056689060811902, | |
| "grad_norm": 0.41231900453567505, | |
| "learning_rate": 0.00015203011066603357, | |
| "loss": 1.1688, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 0.5066623616923517, | |
| "grad_norm": 0.42963293194770813, | |
| "learning_rate": 0.0001515524572486683, | |
| "loss": 1.1657, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.5076558173035131, | |
| "grad_norm": 0.4359903931617737, | |
| "learning_rate": 0.00015107478808694887, | |
| "loss": 1.1612, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 0.5086492729146745, | |
| "grad_norm": 0.48204314708709717, | |
| "learning_rate": 0.0001505971080251907, | |
| "loss": 1.1583, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 0.509642728525836, | |
| "grad_norm": 0.4897457957267761, | |
| "learning_rate": 0.00015011942190781954, | |
| "loss": 1.1604, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 0.5106361841369975, | |
| "grad_norm": 0.5152519345283508, | |
| "learning_rate": 0.0001496417345793228, | |
| "loss": 1.16, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 0.511629639748159, | |
| "grad_norm": 0.48295149207115173, | |
| "learning_rate": 0.00014916405088419998, | |
| "loss": 1.1748, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.5126230953593205, | |
| "grad_norm": 0.46634209156036377, | |
| "learning_rate": 0.0001486863756669137, | |
| "loss": 1.1541, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 0.513616550970482, | |
| "grad_norm": 0.4615635871887207, | |
| "learning_rate": 0.00014820871377184075, | |
| "loss": 1.1675, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 0.5146100065816435, | |
| "grad_norm": 0.4817815124988556, | |
| "learning_rate": 0.00014773107004322278, | |
| "loss": 1.1546, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 0.515603462192805, | |
| "grad_norm": 0.4719342887401581, | |
| "learning_rate": 0.0001472534493251171, | |
| "loss": 1.156, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 0.5165969178039663, | |
| "grad_norm": 0.4903907775878906, | |
| "learning_rate": 0.00014677585646134778, | |
| "loss": 1.1622, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.5175903734151278, | |
| "grad_norm": 0.4558243155479431, | |
| "learning_rate": 0.00014629829629545632, | |
| "loss": 1.1388, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 0.5185838290262893, | |
| "grad_norm": 0.48934248089790344, | |
| "learning_rate": 0.00014582077367065278, | |
| "loss": 1.1584, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 0.5195772846374508, | |
| "grad_norm": 0.4721735417842865, | |
| "learning_rate": 0.00014534329342976612, | |
| "loss": 1.1593, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 0.5205707402486123, | |
| "grad_norm": 0.45554596185684204, | |
| "learning_rate": 0.00014486586041519587, | |
| "loss": 1.1659, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 0.5215641958597738, | |
| "grad_norm": 0.4855104684829712, | |
| "learning_rate": 0.00014438847946886236, | |
| "loss": 1.1833, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.5225576514709352, | |
| "grad_norm": 0.45857903361320496, | |
| "learning_rate": 0.00014391115543215801, | |
| "loss": 1.1389, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 0.5235511070820967, | |
| "grad_norm": 0.4469977021217346, | |
| "learning_rate": 0.00014343389314589797, | |
| "loss": 1.1607, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 0.5245445626932581, | |
| "grad_norm": 0.5102470517158508, | |
| "learning_rate": 0.00014295669745027125, | |
| "loss": 1.1543, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.5255380183044196, | |
| "grad_norm": 0.44836586713790894, | |
| "learning_rate": 0.00014247957318479148, | |
| "loss": 1.1553, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 0.5265314739155811, | |
| "grad_norm": 0.47462692856788635, | |
| "learning_rate": 0.0001420025251882478, | |
| "loss": 1.1671, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.5275249295267426, | |
| "grad_norm": 0.4809962511062622, | |
| "learning_rate": 0.00014152555829865593, | |
| "loss": 1.1536, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 0.5285183851379041, | |
| "grad_norm": 0.49753010272979736, | |
| "learning_rate": 0.00014104867735320912, | |
| "loss": 1.1645, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 0.5295118407490655, | |
| "grad_norm": 0.4475755989551544, | |
| "learning_rate": 0.0001405718871882289, | |
| "loss": 1.1403, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 0.530505296360227, | |
| "grad_norm": 0.45260560512542725, | |
| "learning_rate": 0.00014009519263911616, | |
| "loss": 1.1516, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 0.5314987519713885, | |
| "grad_norm": 0.4384668469429016, | |
| "learning_rate": 0.00013961859854030208, | |
| "loss": 1.1622, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.53249220758255, | |
| "grad_norm": 0.45300155878067017, | |
| "learning_rate": 0.00013914210972519928, | |
| "loss": 1.167, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 0.5334856631937114, | |
| "grad_norm": 0.42206814885139465, | |
| "learning_rate": 0.00013866573102615237, | |
| "loss": 1.1656, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 0.5344791188048729, | |
| "grad_norm": 0.428107887506485, | |
| "learning_rate": 0.0001381894672743894, | |
| "loss": 1.1582, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 0.5354725744160344, | |
| "grad_norm": 0.452441930770874, | |
| "learning_rate": 0.0001377133232999726, | |
| "loss": 1.1541, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 0.5364660300271958, | |
| "grad_norm": 0.4755808711051941, | |
| "learning_rate": 0.0001372373039317495, | |
| "loss": 1.1601, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.5374594856383573, | |
| "grad_norm": 0.4153245687484741, | |
| "learning_rate": 0.0001367614139973039, | |
| "loss": 1.1545, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 0.5384529412495188, | |
| "grad_norm": 0.545102059841156, | |
| "learning_rate": 0.000136285658322907, | |
| "loss": 1.162, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 0.5394463968606803, | |
| "grad_norm": 0.41430729627609253, | |
| "learning_rate": 0.00013581004173346834, | |
| "loss": 1.1573, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 0.5404398524718418, | |
| "grad_norm": 0.43141236901283264, | |
| "learning_rate": 0.00013533456905248683, | |
| "loss": 1.148, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 0.5414333080830032, | |
| "grad_norm": 0.5155332088470459, | |
| "learning_rate": 0.00013485924510200215, | |
| "loss": 1.1574, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.5424267636941646, | |
| "grad_norm": 0.45795363187789917, | |
| "learning_rate": 0.00013438407470254543, | |
| "loss": 1.1485, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 0.5434202193053261, | |
| "grad_norm": 0.42748311161994934, | |
| "learning_rate": 0.0001339090626730906, | |
| "loss": 1.1608, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 0.5444136749164876, | |
| "grad_norm": 0.47087302803993225, | |
| "learning_rate": 0.00013343421383100554, | |
| "loss": 1.1383, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 0.5454071305276491, | |
| "grad_norm": 0.44623327255249023, | |
| "learning_rate": 0.000132959532992003, | |
| "loss": 1.1434, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 0.5464005861388106, | |
| "grad_norm": 0.4788418114185333, | |
| "learning_rate": 0.0001324850249700921, | |
| "loss": 1.162, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.5473940417499721, | |
| "grad_norm": 0.4729274809360504, | |
| "learning_rate": 0.0001320106945775292, | |
| "loss": 1.1502, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 0.5483874973611336, | |
| "grad_norm": 0.46386897563934326, | |
| "learning_rate": 0.0001315365466247692, | |
| "loss": 1.1205, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 0.5493809529722951, | |
| "grad_norm": 0.481486976146698, | |
| "learning_rate": 0.0001310625859204169, | |
| "loss": 1.1629, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 0.5503744085834564, | |
| "grad_norm": 0.47394269704818726, | |
| "learning_rate": 0.00013058881727117798, | |
| "loss": 1.1538, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 0.5513678641946179, | |
| "grad_norm": 0.49581393599510193, | |
| "learning_rate": 0.00013011524548181048, | |
| "loss": 1.1592, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.5523613198057794, | |
| "grad_norm": 0.4401605725288391, | |
| "learning_rate": 0.00012964187535507594, | |
| "loss": 1.1313, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 0.5533547754169409, | |
| "grad_norm": 0.4444689154624939, | |
| "learning_rate": 0.00012916871169169061, | |
| "loss": 1.1543, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 0.5543482310281024, | |
| "grad_norm": 0.45363694429397583, | |
| "learning_rate": 0.00012869575929027704, | |
| "loss": 1.1467, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 0.5553416866392639, | |
| "grad_norm": 0.444588303565979, | |
| "learning_rate": 0.0001282230229473152, | |
| "loss": 1.1617, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 0.5563351422504254, | |
| "grad_norm": 0.43078240752220154, | |
| "learning_rate": 0.00012775050745709383, | |
| "loss": 1.1503, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.5573285978615868, | |
| "grad_norm": 0.48163166642189026, | |
| "learning_rate": 0.00012727821761166196, | |
| "loss": 1.1515, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 0.5583220534727483, | |
| "grad_norm": 0.5182563662528992, | |
| "learning_rate": 0.00012680615820078017, | |
| "loss": 1.1501, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 0.5593155090839097, | |
| "grad_norm": 0.4423312842845917, | |
| "learning_rate": 0.0001263343340118722, | |
| "loss": 1.1455, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 0.5603089646950712, | |
| "grad_norm": 0.46226614713668823, | |
| "learning_rate": 0.00012586274982997596, | |
| "loss": 1.1618, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 0.5613024203062327, | |
| "grad_norm": 0.4315178096294403, | |
| "learning_rate": 0.00012539141043769566, | |
| "loss": 1.1513, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.5622958759173942, | |
| "grad_norm": 0.4481825530529022, | |
| "learning_rate": 0.00012492032061515274, | |
| "loss": 1.1332, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 0.5632893315285556, | |
| "grad_norm": 0.47609102725982666, | |
| "learning_rate": 0.00012444948513993774, | |
| "loss": 1.1482, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 0.5642827871397171, | |
| "grad_norm": 0.45403629541397095, | |
| "learning_rate": 0.00012397890878706162, | |
| "loss": 1.1515, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 0.5652762427508786, | |
| "grad_norm": 0.49668949842453003, | |
| "learning_rate": 0.00012350859632890752, | |
| "loss": 1.1629, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 0.5662696983620401, | |
| "grad_norm": 0.4693937301635742, | |
| "learning_rate": 0.00012303855253518222, | |
| "loss": 1.1386, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.5672631539732015, | |
| "grad_norm": 0.5180835127830505, | |
| "learning_rate": 0.00012256878217286774, | |
| "loss": 1.1454, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 0.568256609584363, | |
| "grad_norm": 0.4567181169986725, | |
| "learning_rate": 0.0001220992900061733, | |
| "loss": 1.1535, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 0.5692500651955245, | |
| "grad_norm": 0.45975708961486816, | |
| "learning_rate": 0.00012163008079648655, | |
| "loss": 1.158, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 0.5702435208066859, | |
| "grad_norm": 0.4463796615600586, | |
| "learning_rate": 0.0001211611593023256, | |
| "loss": 1.1439, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 0.5712369764178474, | |
| "grad_norm": 0.46401357650756836, | |
| "learning_rate": 0.00012069253027929076, | |
| "loss": 1.1325, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.5722304320290089, | |
| "grad_norm": 0.4709807336330414, | |
| "learning_rate": 0.00012022419848001607, | |
| "loss": 1.1458, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.5732238876401704, | |
| "grad_norm": 0.43508750200271606, | |
| "learning_rate": 0.0001197561686541214, | |
| "loss": 1.1312, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 0.5742173432513319, | |
| "grad_norm": 0.47728458046913147, | |
| "learning_rate": 0.00011928844554816401, | |
| "loss": 1.1529, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 0.5752107988624934, | |
| "grad_norm": 0.5009467601776123, | |
| "learning_rate": 0.00011882103390559055, | |
| "loss": 1.144, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 0.5762042544736548, | |
| "grad_norm": 0.41657334566116333, | |
| "learning_rate": 0.00011835393846668907, | |
| "loss": 1.1423, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.5771977100848162, | |
| "grad_norm": 0.4635260999202728, | |
| "learning_rate": 0.00011788716396854067, | |
| "loss": 1.162, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 0.5781911656959777, | |
| "grad_norm": 0.4591808319091797, | |
| "learning_rate": 0.00011742071514497172, | |
| "loss": 1.1526, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 0.5791846213071392, | |
| "grad_norm": 0.4596636891365051, | |
| "learning_rate": 0.00011695459672650566, | |
| "loss": 1.16, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 0.5801780769183007, | |
| "grad_norm": 0.49629148840904236, | |
| "learning_rate": 0.00011648881344031516, | |
| "loss": 1.1517, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 0.5811715325294622, | |
| "grad_norm": 0.5193811655044556, | |
| "learning_rate": 0.00011602337001017399, | |
| "loss": 1.1521, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.5821649881406237, | |
| "grad_norm": 0.5228062272071838, | |
| "learning_rate": 0.00011555827115640947, | |
| "loss": 1.1403, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 0.5831584437517852, | |
| "grad_norm": 0.47855010628700256, | |
| "learning_rate": 0.00011509352159585417, | |
| "loss": 1.1249, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 0.5841518993629465, | |
| "grad_norm": 0.4856327474117279, | |
| "learning_rate": 0.00011462912604179843, | |
| "loss": 1.1467, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 0.585145354974108, | |
| "grad_norm": 0.46515578031539917, | |
| "learning_rate": 0.00011416508920394228, | |
| "loss": 1.1362, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 0.5861388105852695, | |
| "grad_norm": 0.43773916363716125, | |
| "learning_rate": 0.00011370141578834791, | |
| "loss": 1.1276, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.587132266196431, | |
| "grad_norm": 0.4612313210964203, | |
| "learning_rate": 0.00011323811049739184, | |
| "loss": 1.1434, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 0.5881257218075925, | |
| "grad_norm": 0.4420901834964752, | |
| "learning_rate": 0.00011277517802971704, | |
| "loss": 1.1363, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 0.589119177418754, | |
| "grad_norm": 0.47640514373779297, | |
| "learning_rate": 0.00011231262308018568, | |
| "loss": 1.1415, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 0.5901126330299155, | |
| "grad_norm": 0.5014554262161255, | |
| "learning_rate": 0.00011185045033983126, | |
| "loss": 1.1485, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 0.5911060886410769, | |
| "grad_norm": 0.4807797968387604, | |
| "learning_rate": 0.00011138866449581098, | |
| "loss": 1.1464, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.5920995442522384, | |
| "grad_norm": 0.4298631250858307, | |
| "learning_rate": 0.00011092727023135847, | |
| "loss": 1.141, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 0.5930929998633998, | |
| "grad_norm": 0.4730483889579773, | |
| "learning_rate": 0.00011046627222573598, | |
| "loss": 1.135, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 0.5940864554745613, | |
| "grad_norm": 0.5459967851638794, | |
| "learning_rate": 0.00011000567515418709, | |
| "loss": 1.1385, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 0.5950799110857228, | |
| "grad_norm": 0.4976285696029663, | |
| "learning_rate": 0.00010954548368788932, | |
| "loss": 1.1202, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 0.5960733666968843, | |
| "grad_norm": 0.46126067638397217, | |
| "learning_rate": 0.00010908570249390679, | |
| "loss": 1.1218, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.5970668223080458, | |
| "grad_norm": 0.44945207238197327, | |
| "learning_rate": 0.0001086263362351426, | |
| "loss": 1.1402, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 0.5980602779192072, | |
| "grad_norm": 0.4707368314266205, | |
| "learning_rate": 0.00010816738957029202, | |
| "loss": 1.1474, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 0.5990537335303687, | |
| "grad_norm": 0.45081543922424316, | |
| "learning_rate": 0.00010770886715379475, | |
| "loss": 1.1361, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 0.6000471891415302, | |
| "grad_norm": 0.4335690140724182, | |
| "learning_rate": 0.00010725077363578817, | |
| "loss": 1.1266, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 0.6002458802637625, | |
| "eval_loss": 1.1717205047607422, | |
| "eval_runtime": 217.1349, | |
| "eval_samples_per_second": 230.272, | |
| "eval_steps_per_second": 28.784, | |
| "step": 6042 | |
| }, | |
| { | |
| "epoch": 0.6010406447526916, | |
| "grad_norm": 0.4617256224155426, | |
| "learning_rate": 0.00010679311366205971, | |
| "loss": 1.1539, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.6020341003638531, | |
| "grad_norm": 0.4457249045372009, | |
| "learning_rate": 0.00010633589187400009, | |
| "loss": 1.1466, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 0.6030275559750146, | |
| "grad_norm": 0.5100376009941101, | |
| "learning_rate": 0.00010587911290855625, | |
| "loss": 1.1374, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 0.604021011586176, | |
| "grad_norm": 0.5062493681907654, | |
| "learning_rate": 0.00010542278139818404, | |
| "loss": 1.1294, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 0.6050144671973375, | |
| "grad_norm": 0.46003487706184387, | |
| "learning_rate": 0.00010496690197080147, | |
| "loss": 1.1456, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 0.606007922808499, | |
| "grad_norm": 0.46181225776672363, | |
| "learning_rate": 0.00010451147924974182, | |
| "loss": 1.1337, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.6070013784196605, | |
| "grad_norm": 0.441353440284729, | |
| "learning_rate": 0.00010405651785370654, | |
| "loss": 1.1297, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 0.607994834030822, | |
| "grad_norm": 0.4799842834472656, | |
| "learning_rate": 0.00010360202239671843, | |
| "loss": 1.1326, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 0.6089882896419835, | |
| "grad_norm": 0.4211976230144501, | |
| "learning_rate": 0.0001031479974880752, | |
| "loss": 1.1404, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 0.6099817452531449, | |
| "grad_norm": 0.4226428270339966, | |
| "learning_rate": 0.0001026944477323022, | |
| "loss": 1.1398, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 0.6109752008643063, | |
| "grad_norm": 0.47160401940345764, | |
| "learning_rate": 0.00010224137772910618, | |
| "loss": 1.1356, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.6119686564754678, | |
| "grad_norm": 0.46349361538887024, | |
| "learning_rate": 0.00010178879207332834, | |
| "loss": 1.1294, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 0.6129621120866293, | |
| "grad_norm": 0.4886729121208191, | |
| "learning_rate": 0.0001013366953548978, | |
| "loss": 1.142, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 0.6139555676977908, | |
| "grad_norm": 0.4758968949317932, | |
| "learning_rate": 0.00010088509215878528, | |
| "loss": 1.1339, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 0.6149490233089523, | |
| "grad_norm": 0.47533220052719116, | |
| "learning_rate": 0.00010043398706495613, | |
| "loss": 1.1345, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 0.6159424789201138, | |
| "grad_norm": 0.5099872946739197, | |
| "learning_rate": 9.998338464832434e-05, | |
| "loss": 1.1543, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.6169359345312753, | |
| "grad_norm": 0.4612841010093689, | |
| "learning_rate": 9.953328947870598e-05, | |
| "loss": 1.1223, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 0.6179293901424368, | |
| "grad_norm": 0.4441303610801697, | |
| "learning_rate": 9.908370612077272e-05, | |
| "loss": 1.1365, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 0.6189228457535981, | |
| "grad_norm": 0.4773942232131958, | |
| "learning_rate": 9.863463913400582e-05, | |
| "loss": 1.1127, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 0.6199163013647596, | |
| "grad_norm": 0.4970172643661499, | |
| "learning_rate": 9.818609307264953e-05, | |
| "loss": 1.1174, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 0.6209097569759211, | |
| "grad_norm": 0.4488612413406372, | |
| "learning_rate": 9.773807248566534e-05, | |
| "loss": 1.135, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.6219032125870826, | |
| "grad_norm": 0.4947340488433838, | |
| "learning_rate": 9.729058191668527e-05, | |
| "loss": 1.1314, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 0.6228966681982441, | |
| "grad_norm": 0.45969218015670776, | |
| "learning_rate": 9.684362590396643e-05, | |
| "loss": 1.1388, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 0.6238901238094056, | |
| "grad_norm": 0.4401233494281769, | |
| "learning_rate": 9.639720898034454e-05, | |
| "loss": 1.1507, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 0.624883579420567, | |
| "grad_norm": 0.4139381945133209, | |
| "learning_rate": 9.59513356731882e-05, | |
| "loss": 1.1186, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 0.6258770350317285, | |
| "grad_norm": 0.47768634557724, | |
| "learning_rate": 9.550601050435279e-05, | |
| "loss": 1.1327, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.6268704906428899, | |
| "grad_norm": 0.4693679213523865, | |
| "learning_rate": 9.506123799013476e-05, | |
| "loss": 1.1194, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 0.6278639462540514, | |
| "grad_norm": 0.47899654507637024, | |
| "learning_rate": 9.461702264122584e-05, | |
| "loss": 1.1265, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 0.6288574018652129, | |
| "grad_norm": 0.4823016822338104, | |
| "learning_rate": 9.417336896266701e-05, | |
| "loss": 1.1275, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 0.6298508574763744, | |
| "grad_norm": 0.4641190469264984, | |
| "learning_rate": 9.373028145380326e-05, | |
| "loss": 1.1441, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 0.6308443130875359, | |
| "grad_norm": 0.4628022313117981, | |
| "learning_rate": 9.328776460823764e-05, | |
| "loss": 1.1281, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.6318377686986973, | |
| "grad_norm": 0.4776950776576996, | |
| "learning_rate": 9.284582291378573e-05, | |
| "loss": 1.1374, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 0.6328312243098588, | |
| "grad_norm": 0.4577087163925171, | |
| "learning_rate": 9.240446085243033e-05, | |
| "loss": 1.1427, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 0.6338246799210203, | |
| "grad_norm": 0.44643932580947876, | |
| "learning_rate": 9.196368290027572e-05, | |
| "loss": 1.1274, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 0.6348181355321818, | |
| "grad_norm": 0.46955859661102295, | |
| "learning_rate": 9.152349352750232e-05, | |
| "loss": 1.1528, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 0.6358115911433432, | |
| "grad_norm": 0.4443439841270447, | |
| "learning_rate": 9.108389719832165e-05, | |
| "loss": 1.1308, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.6368050467545047, | |
| "grad_norm": 0.482073575258255, | |
| "learning_rate": 9.064489837093069e-05, | |
| "loss": 1.1355, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 0.6377985023656662, | |
| "grad_norm": 0.49913904070854187, | |
| "learning_rate": 9.020650149746679e-05, | |
| "loss": 1.1257, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 0.6387919579768276, | |
| "grad_norm": 0.468218594789505, | |
| "learning_rate": 8.976871102396269e-05, | |
| "loss": 1.1463, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 0.6397854135879891, | |
| "grad_norm": 0.45414066314697266, | |
| "learning_rate": 8.93315313903011e-05, | |
| "loss": 1.1282, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 0.6407788691991506, | |
| "grad_norm": 0.5159599781036377, | |
| "learning_rate": 8.889496703017006e-05, | |
| "loss": 1.1465, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.6417723248103121, | |
| "grad_norm": 0.4611871838569641, | |
| "learning_rate": 8.845902237101754e-05, | |
| "loss": 1.1271, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 0.6427657804214736, | |
| "grad_norm": 0.4411086440086365, | |
| "learning_rate": 8.80237018340069e-05, | |
| "loss": 1.1407, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 0.643759236032635, | |
| "grad_norm": 0.5023905634880066, | |
| "learning_rate": 8.758900983397195e-05, | |
| "loss": 1.1351, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 0.6447526916437964, | |
| "grad_norm": 0.47759824991226196, | |
| "learning_rate": 8.71549507793721e-05, | |
| "loss": 1.1267, | |
| "step": 6490 | |
| }, | |
| { | |
| "epoch": 0.6457461472549579, | |
| "grad_norm": 0.46164053678512573, | |
| "learning_rate": 8.672152907224765e-05, | |
| "loss": 1.136, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.6467396028661194, | |
| "grad_norm": 0.4555439352989197, | |
| "learning_rate": 8.628874910817528e-05, | |
| "loss": 1.131, | |
| "step": 6510 | |
| }, | |
| { | |
| "epoch": 0.6477330584772809, | |
| "grad_norm": 0.49815115332603455, | |
| "learning_rate": 8.585661527622343e-05, | |
| "loss": 1.1535, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 0.6487265140884424, | |
| "grad_norm": 0.4290812909603119, | |
| "learning_rate": 8.542513195890758e-05, | |
| "loss": 1.1386, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 0.6497199696996039, | |
| "grad_norm": 0.4509336054325104, | |
| "learning_rate": 8.499430353214605e-05, | |
| "loss": 1.147, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 0.6507134253107654, | |
| "grad_norm": 0.4503090977668762, | |
| "learning_rate": 8.456413436521558e-05, | |
| "loss": 1.1416, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.6517068809219269, | |
| "grad_norm": 0.5119098424911499, | |
| "learning_rate": 8.413462882070699e-05, | |
| "loss": 1.1187, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 0.6527003365330882, | |
| "grad_norm": 0.4846774637699127, | |
| "learning_rate": 8.370579125448091e-05, | |
| "loss": 1.1266, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 0.6536937921442497, | |
| "grad_norm": 0.46956613659858704, | |
| "learning_rate": 8.327762601562346e-05, | |
| "loss": 1.1507, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 0.6546872477554112, | |
| "grad_norm": 0.4863440990447998, | |
| "learning_rate": 8.285013744640266e-05, | |
| "loss": 1.1214, | |
| "step": 6590 | |
| }, | |
| { | |
| "epoch": 0.6556807033665727, | |
| "grad_norm": 0.45145437121391296, | |
| "learning_rate": 8.24233298822236e-05, | |
| "loss": 1.1216, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.6566741589777342, | |
| "grad_norm": 0.46321019530296326, | |
| "learning_rate": 8.199720765158533e-05, | |
| "loss": 1.1395, | |
| "step": 6610 | |
| }, | |
| { | |
| "epoch": 0.6576676145888957, | |
| "grad_norm": 0.4763803780078888, | |
| "learning_rate": 8.15717750760362e-05, | |
| "loss": 1.1231, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 0.6586610702000572, | |
| "grad_norm": 0.4704056680202484, | |
| "learning_rate": 8.114703647013056e-05, | |
| "loss": 1.1189, | |
| "step": 6630 | |
| }, | |
| { | |
| "epoch": 0.6596545258112186, | |
| "grad_norm": 0.42659151554107666, | |
| "learning_rate": 8.072299614138482e-05, | |
| "loss": 1.127, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 0.6606479814223801, | |
| "grad_norm": 0.4490351676940918, | |
| "learning_rate": 8.029965839023373e-05, | |
| "loss": 1.13, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.6616414370335415, | |
| "grad_norm": 0.4944393038749695, | |
| "learning_rate": 7.98770275099868e-05, | |
| "loss": 1.104, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 0.662634892644703, | |
| "grad_norm": 0.4969693720340729, | |
| "learning_rate": 7.94551077867846e-05, | |
| "loss": 1.1124, | |
| "step": 6670 | |
| }, | |
| { | |
| "epoch": 0.6636283482558645, | |
| "grad_norm": 0.6347571015357971, | |
| "learning_rate": 7.903390349955574e-05, | |
| "loss": 1.1362, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 0.664621803867026, | |
| "grad_norm": 0.48183801770210266, | |
| "learning_rate": 7.861341891997308e-05, | |
| "loss": 1.1368, | |
| "step": 6690 | |
| }, | |
| { | |
| "epoch": 0.6656152594781874, | |
| "grad_norm": 0.44424450397491455, | |
| "learning_rate": 7.819365831241036e-05, | |
| "loss": 1.1216, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.6666087150893489, | |
| "grad_norm": 0.5084063410758972, | |
| "learning_rate": 7.77746259338993e-05, | |
| "loss": 1.1184, | |
| "step": 6710 | |
| }, | |
| { | |
| "epoch": 0.6676021707005104, | |
| "grad_norm": 0.4749411940574646, | |
| "learning_rate": 7.73563260340862e-05, | |
| "loss": 1.1196, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 0.6685956263116719, | |
| "grad_norm": 0.4693940579891205, | |
| "learning_rate": 7.693876285518873e-05, | |
| "loss": 1.135, | |
| "step": 6730 | |
| }, | |
| { | |
| "epoch": 0.6695890819228333, | |
| "grad_norm": 0.46934443712234497, | |
| "learning_rate": 7.652194063195329e-05, | |
| "loss": 1.1349, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 0.6705825375339948, | |
| "grad_norm": 0.4727706015110016, | |
| "learning_rate": 7.61058635916116e-05, | |
| "loss": 1.1158, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.6715759931451563, | |
| "grad_norm": 0.45518192648887634, | |
| "learning_rate": 7.569053595383814e-05, | |
| "loss": 1.1143, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 0.6725694487563177, | |
| "grad_norm": 0.5114362835884094, | |
| "learning_rate": 7.527596193070742e-05, | |
| "loss": 1.1223, | |
| "step": 6770 | |
| }, | |
| { | |
| "epoch": 0.6735629043674792, | |
| "grad_norm": 0.4581073224544525, | |
| "learning_rate": 7.486214572665084e-05, | |
| "loss": 1.1209, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 0.6745563599786407, | |
| "grad_norm": 0.4839348793029785, | |
| "learning_rate": 7.444909153841455e-05, | |
| "loss": 1.1201, | |
| "step": 6790 | |
| }, | |
| { | |
| "epoch": 0.6755498155898022, | |
| "grad_norm": 0.4677003026008606, | |
| "learning_rate": 7.40368035550165e-05, | |
| "loss": 1.111, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.6765432712009637, | |
| "grad_norm": 0.47851434350013733, | |
| "learning_rate": 7.362528595770423e-05, | |
| "loss": 1.1365, | |
| "step": 6810 | |
| }, | |
| { | |
| "epoch": 0.6775367268121252, | |
| "grad_norm": 0.5189026594161987, | |
| "learning_rate": 7.321454291991239e-05, | |
| "loss": 1.1314, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 0.6785301824232866, | |
| "grad_norm": 0.6284201741218567, | |
| "learning_rate": 7.28045786072202e-05, | |
| "loss": 1.1249, | |
| "step": 6830 | |
| }, | |
| { | |
| "epoch": 0.679523638034448, | |
| "grad_norm": 0.43699899315834045, | |
| "learning_rate": 7.239539717730956e-05, | |
| "loss": 1.1302, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 0.6805170936456095, | |
| "grad_norm": 0.47337090969085693, | |
| "learning_rate": 7.198700277992268e-05, | |
| "loss": 1.1182, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.681510549256771, | |
| "grad_norm": 0.47022855281829834, | |
| "learning_rate": 7.157939955681996e-05, | |
| "loss": 1.1297, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 0.6825040048679325, | |
| "grad_norm": 0.4596823453903198, | |
| "learning_rate": 7.117259164173799e-05, | |
| "loss": 1.1275, | |
| "step": 6870 | |
| }, | |
| { | |
| "epoch": 0.683497460479094, | |
| "grad_norm": 0.44979989528656006, | |
| "learning_rate": 7.07665831603479e-05, | |
| "loss": 1.1286, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 0.6844909160902555, | |
| "grad_norm": 0.5144652128219604, | |
| "learning_rate": 7.036137823021311e-05, | |
| "loss": 1.1467, | |
| "step": 6890 | |
| }, | |
| { | |
| "epoch": 0.685484371701417, | |
| "grad_norm": 0.48372334241867065, | |
| "learning_rate": 6.995698096074797e-05, | |
| "loss": 1.1158, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.6864778273125783, | |
| "grad_norm": 0.47437506914138794, | |
| "learning_rate": 6.95533954531756e-05, | |
| "loss": 1.1125, | |
| "step": 6910 | |
| }, | |
| { | |
| "epoch": 0.6874712829237398, | |
| "grad_norm": 0.4924662709236145, | |
| "learning_rate": 6.91506258004868e-05, | |
| "loss": 1.1178, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 0.6884647385349013, | |
| "grad_norm": 0.46547582745552063, | |
| "learning_rate": 6.874867608739833e-05, | |
| "loss": 1.1199, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 0.6894581941460628, | |
| "grad_norm": 0.4472568929195404, | |
| "learning_rate": 6.834755039031133e-05, | |
| "loss": 1.1346, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 0.6904516497572243, | |
| "grad_norm": 0.45158204436302185, | |
| "learning_rate": 6.794725277727016e-05, | |
| "loss": 1.1231, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.6914451053683858, | |
| "grad_norm": 0.45651930570602417, | |
| "learning_rate": 6.75477873079212e-05, | |
| "loss": 1.1179, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 0.6924385609795473, | |
| "grad_norm": 0.43802908062934875, | |
| "learning_rate": 6.714915803347151e-05, | |
| "loss": 1.1132, | |
| "step": 6970 | |
| }, | |
| { | |
| "epoch": 0.6934320165907087, | |
| "grad_norm": 0.5048275589942932, | |
| "learning_rate": 6.675136899664792e-05, | |
| "loss": 1.1225, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 0.6944254722018702, | |
| "grad_norm": 0.4675825238227844, | |
| "learning_rate": 6.635442423165571e-05, | |
| "loss": 1.1294, | |
| "step": 6990 | |
| }, | |
| { | |
| "epoch": 0.6954189278130316, | |
| "grad_norm": 0.4738437235355377, | |
| "learning_rate": 6.595832776413823e-05, | |
| "loss": 1.1366, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.6964123834241931, | |
| "grad_norm": 0.44783490896224976, | |
| "learning_rate": 6.55630836111354e-05, | |
| "loss": 1.1166, | |
| "step": 7010 | |
| }, | |
| { | |
| "epoch": 0.6974058390353546, | |
| "grad_norm": 0.4776122570037842, | |
| "learning_rate": 6.516869578104371e-05, | |
| "loss": 1.1091, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 0.6983992946465161, | |
| "grad_norm": 0.4984320402145386, | |
| "learning_rate": 6.477516827357486e-05, | |
| "loss": 1.1122, | |
| "step": 7030 | |
| }, | |
| { | |
| "epoch": 0.6993927502576776, | |
| "grad_norm": 0.4734194874763489, | |
| "learning_rate": 6.438250507971574e-05, | |
| "loss": 1.1184, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 0.7002868603077229, | |
| "eval_loss": 1.158501148223877, | |
| "eval_runtime": 217.2336, | |
| "eval_samples_per_second": 230.167, | |
| "eval_steps_per_second": 28.771, | |
| "step": 7049 | |
| }, | |
| { | |
| "epoch": 0.700386205868839, | |
| "grad_norm": 0.44331488013267517, | |
| "learning_rate": 6.399071018168768e-05, | |
| "loss": 1.131, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.7013796614800005, | |
| "grad_norm": 0.4725908041000366, | |
| "learning_rate": 6.359978755290616e-05, | |
| "loss": 1.1043, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 0.702373117091162, | |
| "grad_norm": 0.49042266607284546, | |
| "learning_rate": 6.320974115794042e-05, | |
| "loss": 1.1078, | |
| "step": 7070 | |
| }, | |
| { | |
| "epoch": 0.7033665727023234, | |
| "grad_norm": 0.48375070095062256, | |
| "learning_rate": 6.282057495247326e-05, | |
| "loss": 1.1141, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 0.7043600283134849, | |
| "grad_norm": 0.5035314559936523, | |
| "learning_rate": 6.24322928832611e-05, | |
| "loss": 1.1205, | |
| "step": 7090 | |
| }, | |
| { | |
| "epoch": 0.7053534839246464, | |
| "grad_norm": 0.5112337470054626, | |
| "learning_rate": 6.204489888809386e-05, | |
| "loss": 1.1102, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.7063469395358078, | |
| "grad_norm": 0.4766629934310913, | |
| "learning_rate": 6.16583968957548e-05, | |
| "loss": 1.1269, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 0.7073403951469693, | |
| "grad_norm": 0.4522871673107147, | |
| "learning_rate": 6.127279082598104e-05, | |
| "loss": 1.1266, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 0.7083338507581308, | |
| "grad_norm": 0.454632043838501, | |
| "learning_rate": 6.088808458942365e-05, | |
| "loss": 1.1302, | |
| "step": 7130 | |
| }, | |
| { | |
| "epoch": 0.7093273063692923, | |
| "grad_norm": 0.4837658703327179, | |
| "learning_rate": 6.050428208760781e-05, | |
| "loss": 1.1296, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 0.7103207619804538, | |
| "grad_norm": 0.49942177534103394, | |
| "learning_rate": 6.012138721289361e-05, | |
| "loss": 1.1091, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.7113142175916153, | |
| "grad_norm": 0.5515566468238831, | |
| "learning_rate": 5.97394038484362e-05, | |
| "loss": 1.1259, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 0.7123076732027767, | |
| "grad_norm": 0.4293889105319977, | |
| "learning_rate": 5.93583358681467e-05, | |
| "loss": 1.107, | |
| "step": 7170 | |
| }, | |
| { | |
| "epoch": 0.7133011288139381, | |
| "grad_norm": 0.41633716225624084, | |
| "learning_rate": 5.897818713665282e-05, | |
| "loss": 1.0986, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 0.7142945844250996, | |
| "grad_norm": 0.5082070231437683, | |
| "learning_rate": 5.8598961509259504e-05, | |
| "loss": 1.119, | |
| "step": 7190 | |
| }, | |
| { | |
| "epoch": 0.7152880400362611, | |
| "grad_norm": 0.47421905398368835, | |
| "learning_rate": 5.8220662831910135e-05, | |
| "loss": 1.1287, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.7162814956474226, | |
| "grad_norm": 0.4765404462814331, | |
| "learning_rate": 5.784329494114721e-05, | |
| "loss": 1.1085, | |
| "step": 7210 | |
| }, | |
| { | |
| "epoch": 0.7172749512585841, | |
| "grad_norm": 0.46578580141067505, | |
| "learning_rate": 5.746686166407369e-05, | |
| "loss": 1.1377, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 0.7182684068697456, | |
| "grad_norm": 0.47334933280944824, | |
| "learning_rate": 5.709136681831413e-05, | |
| "loss": 1.1177, | |
| "step": 7230 | |
| }, | |
| { | |
| "epoch": 0.7192618624809071, | |
| "grad_norm": 0.49011778831481934, | |
| "learning_rate": 5.671681421197576e-05, | |
| "loss": 1.1225, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 0.7202553180920686, | |
| "grad_norm": 0.43727976083755493, | |
| "learning_rate": 5.634320764361013e-05, | |
| "loss": 1.1127, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 0.7212487737032299, | |
| "grad_norm": 0.4456111490726471, | |
| "learning_rate": 5.5970550902174486e-05, | |
| "loss": 1.1053, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 0.7222422293143914, | |
| "grad_norm": 0.45397263765335083, | |
| "learning_rate": 5.559884776699341e-05, | |
| "loss": 1.1067, | |
| "step": 7270 | |
| }, | |
| { | |
| "epoch": 0.7232356849255529, | |
| "grad_norm": 0.4732288122177124, | |
| "learning_rate": 5.522810200772013e-05, | |
| "loss": 1.1174, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 0.7242291405367144, | |
| "grad_norm": 0.4691105782985687, | |
| "learning_rate": 5.485831738429888e-05, | |
| "loss": 1.115, | |
| "step": 7290 | |
| }, | |
| { | |
| "epoch": 0.7252225961478759, | |
| "grad_norm": 0.42165741324424744, | |
| "learning_rate": 5.448949764692632e-05, | |
| "loss": 1.1222, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.7262160517590374, | |
| "grad_norm": 0.44083935022354126, | |
| "learning_rate": 5.412164653601378e-05, | |
| "loss": 1.129, | |
| "step": 7310 | |
| }, | |
| { | |
| "epoch": 0.7272095073701988, | |
| "grad_norm": 0.46778151392936707, | |
| "learning_rate": 5.375476778214899e-05, | |
| "loss": 1.1169, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 0.7282029629813603, | |
| "grad_norm": 0.4633641541004181, | |
| "learning_rate": 5.338886510605862e-05, | |
| "loss": 1.1047, | |
| "step": 7330 | |
| }, | |
| { | |
| "epoch": 0.7291964185925217, | |
| "grad_norm": 0.4749923050403595, | |
| "learning_rate": 5.30239422185704e-05, | |
| "loss": 1.1272, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 0.7301898742036832, | |
| "grad_norm": 0.47855105996131897, | |
| "learning_rate": 5.266000282057532e-05, | |
| "loss": 1.1113, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 0.7311833298148447, | |
| "grad_norm": 0.4963724911212921, | |
| "learning_rate": 5.229705060299026e-05, | |
| "loss": 1.1158, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 0.7321767854260062, | |
| "grad_norm": 0.4988853633403778, | |
| "learning_rate": 5.193508924672067e-05, | |
| "loss": 1.0987, | |
| "step": 7370 | |
| }, | |
| { | |
| "epoch": 0.7331702410371677, | |
| "grad_norm": 0.4893016219139099, | |
| "learning_rate": 5.157412242262303e-05, | |
| "loss": 1.1102, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 0.7341636966483291, | |
| "grad_norm": 0.4634610712528229, | |
| "learning_rate": 5.1214153791467755e-05, | |
| "loss": 1.1077, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 0.7351571522594906, | |
| "grad_norm": 0.4903322458267212, | |
| "learning_rate": 5.085518700390187e-05, | |
| "loss": 1.1157, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.7361506078706521, | |
| "grad_norm": 0.48547953367233276, | |
| "learning_rate": 5.0497225700412334e-05, | |
| "loss": 1.1186, | |
| "step": 7410 | |
| }, | |
| { | |
| "epoch": 0.7371440634818136, | |
| "grad_norm": 0.4877443313598633, | |
| "learning_rate": 5.014027351128872e-05, | |
| "loss": 1.1081, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 0.738137519092975, | |
| "grad_norm": 0.4813385307788849, | |
| "learning_rate": 4.978433405658679e-05, | |
| "loss": 1.1058, | |
| "step": 7430 | |
| }, | |
| { | |
| "epoch": 0.7391309747041365, | |
| "grad_norm": 0.48355165123939514, | |
| "learning_rate": 4.9429410946091416e-05, | |
| "loss": 1.111, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 0.740124430315298, | |
| "grad_norm": 0.4709435701370239, | |
| "learning_rate": 4.9075507779280246e-05, | |
| "loss": 1.1368, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 0.7411178859264594, | |
| "grad_norm": 0.4199182093143463, | |
| "learning_rate": 4.872262814528707e-05, | |
| "loss": 1.1167, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 0.7421113415376209, | |
| "grad_norm": 0.4348940849304199, | |
| "learning_rate": 4.8370775622865535e-05, | |
| "loss": 1.1209, | |
| "step": 7470 | |
| }, | |
| { | |
| "epoch": 0.7431047971487824, | |
| "grad_norm": 0.48564040660858154, | |
| "learning_rate": 4.801995378035259e-05, | |
| "loss": 1.1043, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 0.7440982527599439, | |
| "grad_norm": 0.45493337512016296, | |
| "learning_rate": 4.767016617563254e-05, | |
| "loss": 1.1222, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 0.7450917083711054, | |
| "grad_norm": 0.4598395824432373, | |
| "learning_rate": 4.732141635610098e-05, | |
| "loss": 1.1182, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.7460851639822668, | |
| "grad_norm": 0.49085426330566406, | |
| "learning_rate": 4.697370785862863e-05, | |
| "loss": 1.1018, | |
| "step": 7510 | |
| }, | |
| { | |
| "epoch": 0.7470786195934283, | |
| "grad_norm": 0.453783243894577, | |
| "learning_rate": 4.6627044209525674e-05, | |
| "loss": 1.1188, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 0.7480720752045897, | |
| "grad_norm": 0.4396882653236389, | |
| "learning_rate": 4.628142892450575e-05, | |
| "loss": 1.1036, | |
| "step": 7530 | |
| }, | |
| { | |
| "epoch": 0.7490655308157512, | |
| "grad_norm": 0.4846864342689514, | |
| "learning_rate": 4.593686550865059e-05, | |
| "loss": 1.1215, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 0.7500589864269127, | |
| "grad_norm": 0.5176936388015747, | |
| "learning_rate": 4.5593357456374175e-05, | |
| "loss": 1.1189, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 0.7510524420380742, | |
| "grad_norm": 0.444042444229126, | |
| "learning_rate": 4.5250908251387594e-05, | |
| "loss": 1.1057, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 0.7520458976492357, | |
| "grad_norm": 0.4722656011581421, | |
| "learning_rate": 4.490952136666338e-05, | |
| "loss": 1.1209, | |
| "step": 7570 | |
| }, | |
| { | |
| "epoch": 0.7530393532603972, | |
| "grad_norm": 0.5398408770561218, | |
| "learning_rate": 4.456920026440064e-05, | |
| "loss": 1.1152, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 0.7540328088715587, | |
| "grad_norm": 0.4860037565231323, | |
| "learning_rate": 4.422994839598969e-05, | |
| "loss": 1.1058, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 0.75502626448272, | |
| "grad_norm": 0.4634205102920532, | |
| "learning_rate": 4.389176920197723e-05, | |
| "loss": 1.0958, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.7560197200938815, | |
| "grad_norm": 0.48629575967788696, | |
| "learning_rate": 4.355466611203119e-05, | |
| "loss": 1.0981, | |
| "step": 7610 | |
| }, | |
| { | |
| "epoch": 0.757013175705043, | |
| "grad_norm": 0.4982178509235382, | |
| "learning_rate": 4.32186425449062e-05, | |
| "loss": 1.1066, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 0.7580066313162045, | |
| "grad_norm": 0.4328030049800873, | |
| "learning_rate": 4.288370190840887e-05, | |
| "loss": 1.1062, | |
| "step": 7630 | |
| }, | |
| { | |
| "epoch": 0.759000086927366, | |
| "grad_norm": 0.4465351700782776, | |
| "learning_rate": 4.2549847599363196e-05, | |
| "loss": 1.1038, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 0.7599935425385275, | |
| "grad_norm": 0.48122888803482056, | |
| "learning_rate": 4.221708300357598e-05, | |
| "loss": 1.0967, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 0.760986998149689, | |
| "grad_norm": 0.4321027100086212, | |
| "learning_rate": 4.1885411495802794e-05, | |
| "loss": 1.1123, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 0.7619804537608504, | |
| "grad_norm": 0.47689056396484375, | |
| "learning_rate": 4.155483643971349e-05, | |
| "loss": 1.1147, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 0.7629739093720119, | |
| "grad_norm": 0.4920623004436493, | |
| "learning_rate": 4.122536118785824e-05, | |
| "loss": 1.112, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 0.7639673649831733, | |
| "grad_norm": 0.4757070243358612, | |
| "learning_rate": 4.089698908163329e-05, | |
| "loss": 1.1338, | |
| "step": 7690 | |
| }, | |
| { | |
| "epoch": 0.7649608205943348, | |
| "grad_norm": 0.4824003279209137, | |
| "learning_rate": 4.056972345124746e-05, | |
| "loss": 1.1306, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.7659542762054963, | |
| "grad_norm": 0.4163155257701874, | |
| "learning_rate": 4.0243567615688116e-05, | |
| "loss": 1.107, | |
| "step": 7710 | |
| }, | |
| { | |
| "epoch": 0.7669477318166578, | |
| "grad_norm": 0.47464796900749207, | |
| "learning_rate": 3.9918524882687566e-05, | |
| "loss": 1.1041, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 0.7679411874278192, | |
| "grad_norm": 0.5027470588684082, | |
| "learning_rate": 3.959459854868942e-05, | |
| "loss": 1.115, | |
| "step": 7730 | |
| }, | |
| { | |
| "epoch": 0.7689346430389807, | |
| "grad_norm": 0.4467754662036896, | |
| "learning_rate": 3.9271791898815366e-05, | |
| "loss": 1.1251, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 0.7699280986501422, | |
| "grad_norm": 0.4438583254814148, | |
| "learning_rate": 3.895010820683172e-05, | |
| "loss": 1.0996, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 0.7709215542613037, | |
| "grad_norm": 0.47995617985725403, | |
| "learning_rate": 3.8629550735116114e-05, | |
| "loss": 1.1178, | |
| "step": 7760 | |
| }, | |
| { | |
| "epoch": 0.7719150098724651, | |
| "grad_norm": 0.4706804156303406, | |
| "learning_rate": 3.8310122734624754e-05, | |
| "loss": 1.1237, | |
| "step": 7770 | |
| }, | |
| { | |
| "epoch": 0.7729084654836266, | |
| "grad_norm": 0.448822021484375, | |
| "learning_rate": 3.7991827444859e-05, | |
| "loss": 1.1222, | |
| "step": 7780 | |
| }, | |
| { | |
| "epoch": 0.7739019210947881, | |
| "grad_norm": 0.5071418881416321, | |
| "learning_rate": 3.767466809383293e-05, | |
| "loss": 1.129, | |
| "step": 7790 | |
| }, | |
| { | |
| "epoch": 0.7748953767059495, | |
| "grad_norm": 0.49738389253616333, | |
| "learning_rate": 3.7358647898040364e-05, | |
| "loss": 1.1227, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.775888832317111, | |
| "grad_norm": 0.4692402184009552, | |
| "learning_rate": 3.7043770062422246e-05, | |
| "loss": 1.1085, | |
| "step": 7810 | |
| }, | |
| { | |
| "epoch": 0.7768822879282725, | |
| "grad_norm": 0.5283324718475342, | |
| "learning_rate": 3.6730037780334277e-05, | |
| "loss": 1.1146, | |
| "step": 7820 | |
| }, | |
| { | |
| "epoch": 0.777875743539434, | |
| "grad_norm": 0.4743114411830902, | |
| "learning_rate": 3.6417454233514335e-05, | |
| "loss": 1.1042, | |
| "step": 7830 | |
| }, | |
| { | |
| "epoch": 0.7788691991505955, | |
| "grad_norm": 0.4658038914203644, | |
| "learning_rate": 3.610602259205044e-05, | |
| "loss": 1.1038, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 0.779862654761757, | |
| "grad_norm": 0.4278145432472229, | |
| "learning_rate": 3.5795746014348466e-05, | |
| "loss": 1.0945, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 0.7808561103729184, | |
| "grad_norm": 0.44836124777793884, | |
| "learning_rate": 3.54866276471e-05, | |
| "loss": 1.1145, | |
| "step": 7860 | |
| }, | |
| { | |
| "epoch": 0.7818495659840798, | |
| "grad_norm": 0.5174553990364075, | |
| "learning_rate": 3.517867062525072e-05, | |
| "loss": 1.0976, | |
| "step": 7870 | |
| }, | |
| { | |
| "epoch": 0.7828430215952413, | |
| "grad_norm": 0.4808596968650818, | |
| "learning_rate": 3.487187807196838e-05, | |
| "loss": 1.1224, | |
| "step": 7880 | |
| }, | |
| { | |
| "epoch": 0.7838364772064028, | |
| "grad_norm": 0.4512399435043335, | |
| "learning_rate": 3.456625309861116e-05, | |
| "loss": 1.1017, | |
| "step": 7890 | |
| }, | |
| { | |
| "epoch": 0.7848299328175643, | |
| "grad_norm": 0.5241881012916565, | |
| "learning_rate": 3.426179880469612e-05, | |
| "loss": 1.1113, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.7858233884287258, | |
| "grad_norm": 0.5214290022850037, | |
| "learning_rate": 3.3958518277867864e-05, | |
| "loss": 1.1136, | |
| "step": 7910 | |
| }, | |
| { | |
| "epoch": 0.7868168440398873, | |
| "grad_norm": 0.47837206721305847, | |
| "learning_rate": 3.365641459386715e-05, | |
| "loss": 1.1177, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 0.7878102996510488, | |
| "grad_norm": 0.48158422112464905, | |
| "learning_rate": 3.3355490816499716e-05, | |
| "loss": 1.099, | |
| "step": 7930 | |
| }, | |
| { | |
| "epoch": 0.7888037552622101, | |
| "grad_norm": 0.49961307644844055, | |
| "learning_rate": 3.305574999760507e-05, | |
| "loss": 1.1128, | |
| "step": 7940 | |
| }, | |
| { | |
| "epoch": 0.7897972108733716, | |
| "grad_norm": 0.4905932545661926, | |
| "learning_rate": 3.275719517702587e-05, | |
| "loss": 1.1126, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 0.7907906664845331, | |
| "grad_norm": 0.5185438394546509, | |
| "learning_rate": 3.2459829382576664e-05, | |
| "loss": 1.098, | |
| "step": 7960 | |
| }, | |
| { | |
| "epoch": 0.7917841220956946, | |
| "grad_norm": 0.7136569023132324, | |
| "learning_rate": 3.21636556300136e-05, | |
| "loss": 1.0941, | |
| "step": 7970 | |
| }, | |
| { | |
| "epoch": 0.7927775777068561, | |
| "grad_norm": 0.5284348130226135, | |
| "learning_rate": 3.186867692300349e-05, | |
| "loss": 1.115, | |
| "step": 7980 | |
| }, | |
| { | |
| "epoch": 0.7937710333180176, | |
| "grad_norm": 0.46627479791641235, | |
| "learning_rate": 3.157489625309363e-05, | |
| "loss": 1.1049, | |
| "step": 7990 | |
| }, | |
| { | |
| "epoch": 0.7947644889291791, | |
| "grad_norm": 0.44333553314208984, | |
| "learning_rate": 3.12823165996813e-05, | |
| "loss": 1.1043, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.7957579445403405, | |
| "grad_norm": 0.45862966775894165, | |
| "learning_rate": 3.099094092998359e-05, | |
| "loss": 1.0948, | |
| "step": 8010 | |
| }, | |
| { | |
| "epoch": 0.796751400151502, | |
| "grad_norm": 0.4916597902774811, | |
| "learning_rate": 3.070077219900729e-05, | |
| "loss": 1.1166, | |
| "step": 8020 | |
| }, | |
| { | |
| "epoch": 0.7977448557626634, | |
| "grad_norm": 0.49191707372665405, | |
| "learning_rate": 3.041181334951887e-05, | |
| "loss": 1.1113, | |
| "step": 8030 | |
| }, | |
| { | |
| "epoch": 0.7987383113738249, | |
| "grad_norm": 0.4618023931980133, | |
| "learning_rate": 3.012406731201485e-05, | |
| "loss": 1.0965, | |
| "step": 8040 | |
| }, | |
| { | |
| "epoch": 0.7997317669849864, | |
| "grad_norm": 0.4719119369983673, | |
| "learning_rate": 2.9837537004691852e-05, | |
| "loss": 1.11, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 0.8003278403516833, | |
| "eval_loss": 1.1472482681274414, | |
| "eval_runtime": 217.1534, | |
| "eval_samples_per_second": 230.252, | |
| "eval_steps_per_second": 28.781, | |
| "step": 8056 | |
| }, | |
| { | |
| "epoch": 0.8007252225961479, | |
| "grad_norm": 0.47482019662857056, | |
| "learning_rate": 2.9552225333417047e-05, | |
| "loss": 1.1232, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 0.8017186782073094, | |
| "grad_norm": 0.5259802937507629, | |
| "learning_rate": 2.9268135191698776e-05, | |
| "loss": 1.1066, | |
| "step": 8070 | |
| }, | |
| { | |
| "epoch": 0.8027121338184708, | |
| "grad_norm": 0.4933863580226898, | |
| "learning_rate": 2.8985269460657175e-05, | |
| "loss": 1.1193, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 0.8037055894296323, | |
| "grad_norm": 0.512607991695404, | |
| "learning_rate": 2.870363100899491e-05, | |
| "loss": 1.1055, | |
| "step": 8090 | |
| }, | |
| { | |
| "epoch": 0.8046990450407938, | |
| "grad_norm": 0.4950838088989258, | |
| "learning_rate": 2.8423222692968052e-05, | |
| "loss": 1.1113, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.8056925006519552, | |
| "grad_norm": 0.5034070014953613, | |
| "learning_rate": 2.814404735635721e-05, | |
| "loss": 1.1177, | |
| "step": 8110 | |
| }, | |
| { | |
| "epoch": 0.8066859562631167, | |
| "grad_norm": 0.463691771030426, | |
| "learning_rate": 2.7866107830438626e-05, | |
| "loss": 1.1178, | |
| "step": 8120 | |
| }, | |
| { | |
| "epoch": 0.8076794118742782, | |
| "grad_norm": 0.4671690762042999, | |
| "learning_rate": 2.7589406933955597e-05, | |
| "loss": 1.0893, | |
| "step": 8130 | |
| }, | |
| { | |
| "epoch": 0.8086728674854397, | |
| "grad_norm": 0.4814799726009369, | |
| "learning_rate": 2.731394747308956e-05, | |
| "loss": 1.1296, | |
| "step": 8140 | |
| }, | |
| { | |
| "epoch": 0.8096663230966011, | |
| "grad_norm": 0.5085082650184631, | |
| "learning_rate": 2.7039732241432016e-05, | |
| "loss": 1.1144, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 0.8106597787077626, | |
| "grad_norm": 0.5503705143928528, | |
| "learning_rate": 2.6766764019956006e-05, | |
| "loss": 1.0936, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 0.8116532343189241, | |
| "grad_norm": 0.5416733622550964, | |
| "learning_rate": 2.6495045576987828e-05, | |
| "loss": 1.0877, | |
| "step": 8170 | |
| }, | |
| { | |
| "epoch": 0.8126466899300856, | |
| "grad_norm": 0.5096998810768127, | |
| "learning_rate": 2.622457966817921e-05, | |
| "loss": 1.1084, | |
| "step": 8180 | |
| }, | |
| { | |
| "epoch": 0.8136401455412471, | |
| "grad_norm": 0.5110299587249756, | |
| "learning_rate": 2.5955369036479047e-05, | |
| "loss": 1.1143, | |
| "step": 8190 | |
| }, | |
| { | |
| "epoch": 0.8146336011524085, | |
| "grad_norm": 0.49047693610191345, | |
| "learning_rate": 2.568741641210591e-05, | |
| "loss": 1.1078, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.81562705676357, | |
| "grad_norm": 0.6099783778190613, | |
| "learning_rate": 2.5420724512520142e-05, | |
| "loss": 1.104, | |
| "step": 8210 | |
| }, | |
| { | |
| "epoch": 0.8166205123747314, | |
| "grad_norm": 0.42165687680244446, | |
| "learning_rate": 2.5155296042396305e-05, | |
| "loss": 1.1026, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 0.8176139679858929, | |
| "grad_norm": 0.5082433223724365, | |
| "learning_rate": 2.489113369359592e-05, | |
| "loss": 1.1073, | |
| "step": 8230 | |
| }, | |
| { | |
| "epoch": 0.8186074235970544, | |
| "grad_norm": 0.522196888923645, | |
| "learning_rate": 2.462824014513991e-05, | |
| "loss": 1.0992, | |
| "step": 8240 | |
| }, | |
| { | |
| "epoch": 0.8196008792082159, | |
| "grad_norm": 0.4945440888404846, | |
| "learning_rate": 2.4366618063181652e-05, | |
| "loss": 1.1087, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 0.8205943348193774, | |
| "grad_norm": 0.4586648643016815, | |
| "learning_rate": 2.4106270100979898e-05, | |
| "loss": 1.0966, | |
| "step": 8260 | |
| }, | |
| { | |
| "epoch": 0.8215877904305389, | |
| "grad_norm": 0.49676162004470825, | |
| "learning_rate": 2.3847198898871688e-05, | |
| "loss": 1.1034, | |
| "step": 8270 | |
| }, | |
| { | |
| "epoch": 0.8225812460417004, | |
| "grad_norm": 0.4228706657886505, | |
| "learning_rate": 2.3589407084245854e-05, | |
| "loss": 1.11, | |
| "step": 8280 | |
| }, | |
| { | |
| "epoch": 0.8235747016528617, | |
| "grad_norm": 0.43764999508857727, | |
| "learning_rate": 2.3332897271516177e-05, | |
| "loss": 1.0958, | |
| "step": 8290 | |
| }, | |
| { | |
| "epoch": 0.8245681572640232, | |
| "grad_norm": 0.4700354337692261, | |
| "learning_rate": 2.3077672062094892e-05, | |
| "loss": 1.1059, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.8255616128751847, | |
| "grad_norm": 0.44920220971107483, | |
| "learning_rate": 2.2823734044366325e-05, | |
| "loss": 1.1033, | |
| "step": 8310 | |
| }, | |
| { | |
| "epoch": 0.8265550684863462, | |
| "grad_norm": 0.519624650478363, | |
| "learning_rate": 2.257108579366074e-05, | |
| "loss": 1.0999, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 0.8275485240975077, | |
| "grad_norm": 0.5019779205322266, | |
| "learning_rate": 2.2319729872228086e-05, | |
| "loss": 1.0867, | |
| "step": 8330 | |
| }, | |
| { | |
| "epoch": 0.8285419797086692, | |
| "grad_norm": 0.4180760383605957, | |
| "learning_rate": 2.2069668829212122e-05, | |
| "loss": 1.1194, | |
| "step": 8340 | |
| }, | |
| { | |
| "epoch": 0.8295354353198306, | |
| "grad_norm": 0.47347915172576904, | |
| "learning_rate": 2.1820905200624382e-05, | |
| "loss": 1.1144, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 0.8305288909309921, | |
| "grad_norm": 0.4253215789794922, | |
| "learning_rate": 2.1573441509318727e-05, | |
| "loss": 1.0972, | |
| "step": 8360 | |
| }, | |
| { | |
| "epoch": 0.8315223465421535, | |
| "grad_norm": 0.4493556618690491, | |
| "learning_rate": 2.1327280264965462e-05, | |
| "loss": 1.0999, | |
| "step": 8370 | |
| }, | |
| { | |
| "epoch": 0.832515802153315, | |
| "grad_norm": 0.486507773399353, | |
| "learning_rate": 2.1082423964026208e-05, | |
| "loss": 1.0979, | |
| "step": 8380 | |
| }, | |
| { | |
| "epoch": 0.8335092577644765, | |
| "grad_norm": 0.5007907152175903, | |
| "learning_rate": 2.0838875089728236e-05, | |
| "loss": 1.1062, | |
| "step": 8390 | |
| }, | |
| { | |
| "epoch": 0.834502713375638, | |
| "grad_norm": 0.6081424951553345, | |
| "learning_rate": 2.059663611203956e-05, | |
| "loss": 1.1103, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.8354961689867995, | |
| "grad_norm": 0.4662105441093445, | |
| "learning_rate": 2.0355709487643807e-05, | |
| "loss": 1.0989, | |
| "step": 8410 | |
| }, | |
| { | |
| "epoch": 0.836489624597961, | |
| "grad_norm": 0.48797285556793213, | |
| "learning_rate": 2.0116097659915225e-05, | |
| "loss": 1.0989, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 0.8374830802091224, | |
| "grad_norm": 0.4916144609451294, | |
| "learning_rate": 1.987780305889397e-05, | |
| "loss": 1.0876, | |
| "step": 8430 | |
| }, | |
| { | |
| "epoch": 0.8384765358202839, | |
| "grad_norm": 0.4921468198299408, | |
| "learning_rate": 1.9640828101261413e-05, | |
| "loss": 1.1062, | |
| "step": 8440 | |
| }, | |
| { | |
| "epoch": 0.8394699914314454, | |
| "grad_norm": 0.48090696334838867, | |
| "learning_rate": 1.9405175190315747e-05, | |
| "loss": 1.1114, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 0.8404634470426068, | |
| "grad_norm": 0.46354711055755615, | |
| "learning_rate": 1.917084671594754e-05, | |
| "loss": 1.1038, | |
| "step": 8460 | |
| }, | |
| { | |
| "epoch": 0.8414569026537683, | |
| "grad_norm": 0.5036799907684326, | |
| "learning_rate": 1.8937845054615373e-05, | |
| "loss": 1.1213, | |
| "step": 8470 | |
| }, | |
| { | |
| "epoch": 0.8424503582649298, | |
| "grad_norm": 0.509793221950531, | |
| "learning_rate": 1.870617256932199e-05, | |
| "loss": 1.1071, | |
| "step": 8480 | |
| }, | |
| { | |
| "epoch": 0.8434438138760912, | |
| "grad_norm": 0.46539217233657837, | |
| "learning_rate": 1.8475831609590135e-05, | |
| "loss": 1.11, | |
| "step": 8490 | |
| }, | |
| { | |
| "epoch": 0.8444372694872527, | |
| "grad_norm": 0.4852668046951294, | |
| "learning_rate": 1.8246824511438868e-05, | |
| "loss": 1.1038, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.8454307250984142, | |
| "grad_norm": 0.5432482361793518, | |
| "learning_rate": 1.801915359735967e-05, | |
| "loss": 1.0954, | |
| "step": 8510 | |
| }, | |
| { | |
| "epoch": 0.8464241807095757, | |
| "grad_norm": 0.4479030668735504, | |
| "learning_rate": 1.7792821176293097e-05, | |
| "loss": 1.1083, | |
| "step": 8520 | |
| }, | |
| { | |
| "epoch": 0.8474176363207372, | |
| "grad_norm": 0.49862831830978394, | |
| "learning_rate": 1.7567829543605282e-05, | |
| "loss": 1.1043, | |
| "step": 8530 | |
| }, | |
| { | |
| "epoch": 0.8484110919318986, | |
| "grad_norm": 0.47839561104774475, | |
| "learning_rate": 1.734418098106468e-05, | |
| "loss": 1.1001, | |
| "step": 8540 | |
| }, | |
| { | |
| "epoch": 0.84940454754306, | |
| "grad_norm": 0.49436119198799133, | |
| "learning_rate": 1.712187775681883e-05, | |
| "loss": 1.0945, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 0.8503980031542215, | |
| "grad_norm": 0.496619313955307, | |
| "learning_rate": 1.6900922125371498e-05, | |
| "loss": 1.0862, | |
| "step": 8560 | |
| }, | |
| { | |
| "epoch": 0.851391458765383, | |
| "grad_norm": 0.4753607511520386, | |
| "learning_rate": 1.6681316327559763e-05, | |
| "loss": 1.1149, | |
| "step": 8570 | |
| }, | |
| { | |
| "epoch": 0.8523849143765445, | |
| "grad_norm": 0.5014109015464783, | |
| "learning_rate": 1.646306259053116e-05, | |
| "loss": 1.0854, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 0.853378369987706, | |
| "grad_norm": 0.48267966508865356, | |
| "learning_rate": 1.624616312772133e-05, | |
| "loss": 1.0944, | |
| "step": 8590 | |
| }, | |
| { | |
| "epoch": 0.8543718255988675, | |
| "grad_norm": 0.49954667687416077, | |
| "learning_rate": 1.6030620138831336e-05, | |
| "loss": 1.0941, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.855365281210029, | |
| "grad_norm": 0.494306355714798, | |
| "learning_rate": 1.5816435809805557e-05, | |
| "loss": 1.1007, | |
| "step": 8610 | |
| }, | |
| { | |
| "epoch": 0.8563587368211905, | |
| "grad_norm": 0.4919048845767975, | |
| "learning_rate": 1.5603612312809373e-05, | |
| "loss": 1.0906, | |
| "step": 8620 | |
| }, | |
| { | |
| "epoch": 0.8573521924323518, | |
| "grad_norm": 0.501146674156189, | |
| "learning_rate": 1.5392151806207254e-05, | |
| "loss": 1.1094, | |
| "step": 8630 | |
| }, | |
| { | |
| "epoch": 0.8583456480435133, | |
| "grad_norm": 0.49974754452705383, | |
| "learning_rate": 1.5182056434540729e-05, | |
| "loss": 1.1068, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 0.8593391036546748, | |
| "grad_norm": 0.4796450436115265, | |
| "learning_rate": 1.4973328328506734e-05, | |
| "loss": 1.0967, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 0.8603325592658363, | |
| "grad_norm": 0.4462718963623047, | |
| "learning_rate": 1.4765969604935995e-05, | |
| "loss": 1.1199, | |
| "step": 8660 | |
| }, | |
| { | |
| "epoch": 0.8613260148769978, | |
| "grad_norm": 0.48056918382644653, | |
| "learning_rate": 1.4559982366771616e-05, | |
| "loss": 1.0958, | |
| "step": 8670 | |
| }, | |
| { | |
| "epoch": 0.8623194704881593, | |
| "grad_norm": 0.49090835452079773, | |
| "learning_rate": 1.4355368703047543e-05, | |
| "loss": 1.1091, | |
| "step": 8680 | |
| }, | |
| { | |
| "epoch": 0.8633129260993208, | |
| "grad_norm": 0.4484087824821472, | |
| "learning_rate": 1.4152130688867647e-05, | |
| "loss": 1.0883, | |
| "step": 8690 | |
| }, | |
| { | |
| "epoch": 0.8643063817104822, | |
| "grad_norm": 0.47980430722236633, | |
| "learning_rate": 1.3950270385384526e-05, | |
| "loss": 1.1294, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.8652998373216437, | |
| "grad_norm": 0.44564327597618103, | |
| "learning_rate": 1.3749789839778574e-05, | |
| "loss": 1.0989, | |
| "step": 8710 | |
| }, | |
| { | |
| "epoch": 0.8662932929328051, | |
| "grad_norm": 0.5069140791893005, | |
| "learning_rate": 1.3550691085237292e-05, | |
| "loss": 1.0914, | |
| "step": 8720 | |
| }, | |
| { | |
| "epoch": 0.8672867485439666, | |
| "grad_norm": 0.46581125259399414, | |
| "learning_rate": 1.335297614093469e-05, | |
| "loss": 1.0894, | |
| "step": 8730 | |
| }, | |
| { | |
| "epoch": 0.8682802041551281, | |
| "grad_norm": 0.47419315576553345, | |
| "learning_rate": 1.315664701201074e-05, | |
| "loss": 1.0856, | |
| "step": 8740 | |
| }, | |
| { | |
| "epoch": 0.8692736597662896, | |
| "grad_norm": 0.4483776092529297, | |
| "learning_rate": 1.2961705689551122e-05, | |
| "loss": 1.1067, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 0.870267115377451, | |
| "grad_norm": 0.49877870082855225, | |
| "learning_rate": 1.2768154150566856e-05, | |
| "loss": 1.1183, | |
| "step": 8760 | |
| }, | |
| { | |
| "epoch": 0.8712605709886125, | |
| "grad_norm": 0.4901898205280304, | |
| "learning_rate": 1.2575994357974523e-05, | |
| "loss": 1.0934, | |
| "step": 8770 | |
| }, | |
| { | |
| "epoch": 0.872254026599774, | |
| "grad_norm": 0.454645574092865, | |
| "learning_rate": 1.2385228260576091e-05, | |
| "loss": 1.1092, | |
| "step": 8780 | |
| }, | |
| { | |
| "epoch": 0.8732474822109355, | |
| "grad_norm": 0.46396562457084656, | |
| "learning_rate": 1.2195857793039337e-05, | |
| "loss": 1.1136, | |
| "step": 8790 | |
| }, | |
| { | |
| "epoch": 0.8742409378220969, | |
| "grad_norm": 0.45802921056747437, | |
| "learning_rate": 1.2007884875878143e-05, | |
| "loss": 1.0872, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.8752343934332584, | |
| "grad_norm": 0.487572580575943, | |
| "learning_rate": 1.182131141543301e-05, | |
| "loss": 1.1161, | |
| "step": 8810 | |
| }, | |
| { | |
| "epoch": 0.8762278490444199, | |
| "grad_norm": 0.46634870767593384, | |
| "learning_rate": 1.163613930385181e-05, | |
| "loss": 1.095, | |
| "step": 8820 | |
| }, | |
| { | |
| "epoch": 0.8772213046555813, | |
| "grad_norm": 0.4800409972667694, | |
| "learning_rate": 1.1452370419070517e-05, | |
| "loss": 1.0905, | |
| "step": 8830 | |
| }, | |
| { | |
| "epoch": 0.8782147602667428, | |
| "grad_norm": 0.4725741446018219, | |
| "learning_rate": 1.1270006624794154e-05, | |
| "loss": 1.088, | |
| "step": 8840 | |
| }, | |
| { | |
| "epoch": 0.8792082158779043, | |
| "grad_norm": 0.44453057646751404, | |
| "learning_rate": 1.1089049770477892e-05, | |
| "loss": 1.1069, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 0.8802016714890658, | |
| "grad_norm": 0.47712013125419617, | |
| "learning_rate": 1.0909501691308387e-05, | |
| "loss": 1.1027, | |
| "step": 8860 | |
| }, | |
| { | |
| "epoch": 0.8811951271002273, | |
| "grad_norm": 0.4958110451698303, | |
| "learning_rate": 1.07313642081851e-05, | |
| "loss": 1.094, | |
| "step": 8870 | |
| }, | |
| { | |
| "epoch": 0.8821885827113888, | |
| "grad_norm": 0.4982723593711853, | |
| "learning_rate": 1.0554639127701808e-05, | |
| "loss": 1.0986, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 0.8831820383225502, | |
| "grad_norm": 0.4847375154495239, | |
| "learning_rate": 1.0379328242128294e-05, | |
| "loss": 1.095, | |
| "step": 8890 | |
| }, | |
| { | |
| "epoch": 0.8841754939337116, | |
| "grad_norm": 0.5036672949790955, | |
| "learning_rate": 1.020543332939221e-05, | |
| "loss": 1.1015, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.8851689495448731, | |
| "grad_norm": 0.4520683288574219, | |
| "learning_rate": 1.003295615306105e-05, | |
| "loss": 1.1039, | |
| "step": 8910 | |
| }, | |
| { | |
| "epoch": 0.8861624051560346, | |
| "grad_norm": 0.45851820707321167, | |
| "learning_rate": 9.861898462324158e-06, | |
| "loss": 1.0939, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 0.8871558607671961, | |
| "grad_norm": 0.6486339569091797, | |
| "learning_rate": 9.692261991975076e-06, | |
| "loss": 1.0876, | |
| "step": 8930 | |
| }, | |
| { | |
| "epoch": 0.8881493163783576, | |
| "grad_norm": 0.46632036566734314, | |
| "learning_rate": 9.524048462394024e-06, | |
| "loss": 1.1131, | |
| "step": 8940 | |
| }, | |
| { | |
| "epoch": 0.8891427719895191, | |
| "grad_norm": 0.45053717494010925, | |
| "learning_rate": 9.357259579530285e-06, | |
| "loss": 1.1084, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 0.8901362276006806, | |
| "grad_norm": 0.4486932158470154, | |
| "learning_rate": 9.191897034885033e-06, | |
| "loss": 1.0832, | |
| "step": 8960 | |
| }, | |
| { | |
| "epoch": 0.8911296832118419, | |
| "grad_norm": 0.4375132620334625, | |
| "learning_rate": 9.027962505494108e-06, | |
| "loss": 1.0946, | |
| "step": 8970 | |
| }, | |
| { | |
| "epoch": 0.8921231388230034, | |
| "grad_norm": 0.49327731132507324, | |
| "learning_rate": 8.86545765391104e-06, | |
| "loss": 1.1064, | |
| "step": 8980 | |
| }, | |
| { | |
| "epoch": 0.8931165944341649, | |
| "grad_norm": 0.4718957245349884, | |
| "learning_rate": 8.704384128190156e-06, | |
| "loss": 1.0874, | |
| "step": 8990 | |
| }, | |
| { | |
| "epoch": 0.8941100500453264, | |
| "grad_norm": 0.4812639653682709, | |
| "learning_rate": 8.544743561869977e-06, | |
| "loss": 1.0994, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.8951035056564879, | |
| "grad_norm": 0.5206982493400574, | |
| "learning_rate": 8.386537573956431e-06, | |
| "loss": 1.1021, | |
| "step": 9010 | |
| }, | |
| { | |
| "epoch": 0.8960969612676494, | |
| "grad_norm": 0.49871826171875, | |
| "learning_rate": 8.229767768906687e-06, | |
| "loss": 1.1138, | |
| "step": 9020 | |
| }, | |
| { | |
| "epoch": 0.8970904168788109, | |
| "grad_norm": 0.4538607895374298, | |
| "learning_rate": 8.074435736612712e-06, | |
| "loss": 1.0826, | |
| "step": 9030 | |
| }, | |
| { | |
| "epoch": 0.8980838724899723, | |
| "grad_norm": 0.4945856034755707, | |
| "learning_rate": 7.920543052385236e-06, | |
| "loss": 1.1204, | |
| "step": 9040 | |
| }, | |
| { | |
| "epoch": 0.8990773281011338, | |
| "grad_norm": 0.5100029110908508, | |
| "learning_rate": 7.768091276937688e-06, | |
| "loss": 1.1011, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 0.9000707837122952, | |
| "grad_norm": 0.4894271194934845, | |
| "learning_rate": 7.617081956370413e-06, | |
| "loss": 1.0853, | |
| "step": 9060 | |
| }, | |
| { | |
| "epoch": 0.9003688203956437, | |
| "eval_loss": 1.1406679153442383, | |
| "eval_runtime": 216.5409, | |
| "eval_samples_per_second": 230.903, | |
| "eval_steps_per_second": 28.863, | |
| "step": 9063 | |
| }, | |
| { | |
| "epoch": 0.9010642393234567, | |
| "grad_norm": 0.511590838432312, | |
| "learning_rate": 7.467516622155079e-06, | |
| "loss": 1.1079, | |
| "step": 9070 | |
| }, | |
| { | |
| "epoch": 0.9020576949346182, | |
| "grad_norm": 0.45951762795448303, | |
| "learning_rate": 7.319396791118998e-06, | |
| "loss": 1.0963, | |
| "step": 9080 | |
| }, | |
| { | |
| "epoch": 0.9030511505457797, | |
| "grad_norm": 0.4679279625415802, | |
| "learning_rate": 7.172723965429811e-06, | |
| "loss": 1.0922, | |
| "step": 9090 | |
| }, | |
| { | |
| "epoch": 0.9040446061569412, | |
| "grad_norm": 0.5013454556465149, | |
| "learning_rate": 7.027499632580264e-06, | |
| "loss": 1.0782, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.9050380617681026, | |
| "grad_norm": 0.49527832865715027, | |
| "learning_rate": 6.883725265373185e-06, | |
| "loss": 1.09, | |
| "step": 9110 | |
| }, | |
| { | |
| "epoch": 0.9060315173792641, | |
| "grad_norm": 0.4694949984550476, | |
| "learning_rate": 6.74140232190632e-06, | |
| "loss": 1.083, | |
| "step": 9120 | |
| }, | |
| { | |
| "epoch": 0.9070249729904256, | |
| "grad_norm": 0.4897943139076233, | |
| "learning_rate": 6.600532245557833e-06, | |
| "loss": 1.0954, | |
| "step": 9130 | |
| }, | |
| { | |
| "epoch": 0.908018428601587, | |
| "grad_norm": 0.48320844769477844, | |
| "learning_rate": 6.461116464971461e-06, | |
| "loss": 1.1021, | |
| "step": 9140 | |
| }, | |
| { | |
| "epoch": 0.9090118842127485, | |
| "grad_norm": 0.498017817735672, | |
| "learning_rate": 6.323156394042139e-06, | |
| "loss": 1.1142, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 0.91000533982391, | |
| "grad_norm": 0.4676865041255951, | |
| "learning_rate": 6.186653431901628e-06, | |
| "loss": 1.0964, | |
| "step": 9160 | |
| }, | |
| { | |
| "epoch": 0.9109987954350715, | |
| "grad_norm": 0.5362632870674133, | |
| "learning_rate": 6.051608962904225e-06, | |
| "loss": 1.1146, | |
| "step": 9170 | |
| }, | |
| { | |
| "epoch": 0.9119922510462329, | |
| "grad_norm": 0.5107495188713074, | |
| "learning_rate": 5.91802435661296e-06, | |
| "loss": 1.1102, | |
| "step": 9180 | |
| }, | |
| { | |
| "epoch": 0.9129857066573944, | |
| "grad_norm": 0.47366631031036377, | |
| "learning_rate": 5.78590096778544e-06, | |
| "loss": 1.0898, | |
| "step": 9190 | |
| }, | |
| { | |
| "epoch": 0.9139791622685559, | |
| "grad_norm": 0.44392383098602295, | |
| "learning_rate": 5.655240136360306e-06, | |
| "loss": 1.0917, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.9149726178797174, | |
| "grad_norm": 0.5138424038887024, | |
| "learning_rate": 5.5260431874435515e-06, | |
| "loss": 1.0922, | |
| "step": 9210 | |
| }, | |
| { | |
| "epoch": 0.9159660734908789, | |
| "grad_norm": 0.49133333563804626, | |
| "learning_rate": 5.398311431295061e-06, | |
| "loss": 1.103, | |
| "step": 9220 | |
| }, | |
| { | |
| "epoch": 0.9169595291020403, | |
| "grad_norm": 0.4637450873851776, | |
| "learning_rate": 5.2720461633154354e-06, | |
| "loss": 1.1077, | |
| "step": 9230 | |
| }, | |
| { | |
| "epoch": 0.9179529847132017, | |
| "grad_norm": 0.47091439366340637, | |
| "learning_rate": 5.147248664032744e-06, | |
| "loss": 1.0943, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 0.9189464403243632, | |
| "grad_norm": 0.4741147756576538, | |
| "learning_rate": 5.023920199089576e-06, | |
| "loss": 1.0887, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 0.9199398959355247, | |
| "grad_norm": 0.48109114170074463, | |
| "learning_rate": 4.902062019230191e-06, | |
| "loss": 1.0887, | |
| "step": 9260 | |
| }, | |
| { | |
| "epoch": 0.9209333515466862, | |
| "grad_norm": 0.5040728449821472, | |
| "learning_rate": 4.78167536028789e-06, | |
| "loss": 1.0824, | |
| "step": 9270 | |
| }, | |
| { | |
| "epoch": 0.9219268071578477, | |
| "grad_norm": 0.5055690407752991, | |
| "learning_rate": 4.662761443172414e-06, | |
| "loss": 1.1001, | |
| "step": 9280 | |
| }, | |
| { | |
| "epoch": 0.9229202627690092, | |
| "grad_norm": 0.4466319978237152, | |
| "learning_rate": 4.545321473857616e-06, | |
| "loss": 1.0991, | |
| "step": 9290 | |
| }, | |
| { | |
| "epoch": 0.9239137183801707, | |
| "grad_norm": 0.5149044394493103, | |
| "learning_rate": 4.429356643369159e-06, | |
| "loss": 1.0929, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.9249071739913322, | |
| "grad_norm": 0.4559183418750763, | |
| "learning_rate": 4.314868127772503e-06, | |
| "loss": 1.0871, | |
| "step": 9310 | |
| }, | |
| { | |
| "epoch": 0.9259006296024935, | |
| "grad_norm": 0.4630989134311676, | |
| "learning_rate": 4.201857088160987e-06, | |
| "loss": 1.0856, | |
| "step": 9320 | |
| }, | |
| { | |
| "epoch": 0.926894085213655, | |
| "grad_norm": 0.4689970016479492, | |
| "learning_rate": 4.090324670643968e-06, | |
| "loss": 1.1021, | |
| "step": 9330 | |
| }, | |
| { | |
| "epoch": 0.9278875408248165, | |
| "grad_norm": 0.4830831289291382, | |
| "learning_rate": 3.9802720063353e-06, | |
| "loss": 1.1006, | |
| "step": 9340 | |
| }, | |
| { | |
| "epoch": 0.928880996435978, | |
| "grad_norm": 0.44650521874427795, | |
| "learning_rate": 3.8717002113417885e-06, | |
| "loss": 1.1021, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 0.9298744520471395, | |
| "grad_norm": 0.5089627504348755, | |
| "learning_rate": 3.764610386751904e-06, | |
| "loss": 1.0884, | |
| "step": 9360 | |
| }, | |
| { | |
| "epoch": 0.930867907658301, | |
| "grad_norm": 0.5018206238746643, | |
| "learning_rate": 3.659003618624656e-06, | |
| "loss": 1.087, | |
| "step": 9370 | |
| }, | |
| { | |
| "epoch": 0.9318613632694625, | |
| "grad_norm": 0.6836445331573486, | |
| "learning_rate": 3.554880977978436e-06, | |
| "loss": 1.1166, | |
| "step": 9380 | |
| }, | |
| { | |
| "epoch": 0.9328548188806239, | |
| "grad_norm": 0.4803834855556488, | |
| "learning_rate": 3.4522435207803567e-06, | |
| "loss": 1.0942, | |
| "step": 9390 | |
| }, | |
| { | |
| "epoch": 0.9338482744917853, | |
| "grad_norm": 0.52688068151474, | |
| "learning_rate": 3.3510922879353476e-06, | |
| "loss": 1.0967, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.9348417301029468, | |
| "grad_norm": 0.4448314309120178, | |
| "learning_rate": 3.2514283052757605e-06, | |
| "loss": 1.1083, | |
| "step": 9410 | |
| }, | |
| { | |
| "epoch": 0.9358351857141083, | |
| "grad_norm": 0.5127049088478088, | |
| "learning_rate": 3.1532525835508636e-06, | |
| "loss": 1.1008, | |
| "step": 9420 | |
| }, | |
| { | |
| "epoch": 0.9368286413252698, | |
| "grad_norm": 0.4845719635486603, | |
| "learning_rate": 3.0565661184166146e-06, | |
| "loss": 1.0877, | |
| "step": 9430 | |
| }, | |
| { | |
| "epoch": 0.9378220969364313, | |
| "grad_norm": 0.4783158302307129, | |
| "learning_rate": 2.9613698904255857e-06, | |
| "loss": 1.0992, | |
| "step": 9440 | |
| }, | |
| { | |
| "epoch": 0.9388155525475927, | |
| "grad_norm": 0.44096460938453674, | |
| "learning_rate": 2.867664865017022e-06, | |
| "loss": 1.1033, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 0.9398090081587542, | |
| "grad_norm": 0.47711020708084106, | |
| "learning_rate": 2.7754519925069995e-06, | |
| "loss": 1.0719, | |
| "step": 9460 | |
| }, | |
| { | |
| "epoch": 0.9408024637699157, | |
| "grad_norm": 0.46728119254112244, | |
| "learning_rate": 2.684732208078816e-06, | |
| "loss": 1.0953, | |
| "step": 9470 | |
| }, | |
| { | |
| "epoch": 0.9417959193810772, | |
| "grad_norm": 0.48205727338790894, | |
| "learning_rate": 2.5955064317735474e-06, | |
| "loss": 1.1177, | |
| "step": 9480 | |
| }, | |
| { | |
| "epoch": 0.9427893749922386, | |
| "grad_norm": 0.4576367139816284, | |
| "learning_rate": 2.5077755684806745e-06, | |
| "loss": 1.11, | |
| "step": 9490 | |
| }, | |
| { | |
| "epoch": 0.9437828306034001, | |
| "grad_norm": 0.4937973618507385, | |
| "learning_rate": 2.421540507928871e-06, | |
| "loss": 1.0879, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.9447762862145616, | |
| "grad_norm": 0.4714371860027313, | |
| "learning_rate": 2.3368021246770786e-06, | |
| "loss": 1.0791, | |
| "step": 9510 | |
| }, | |
| { | |
| "epoch": 0.945769741825723, | |
| "grad_norm": 0.4823916554450989, | |
| "learning_rate": 2.2535612781055478e-06, | |
| "loss": 1.0933, | |
| "step": 9520 | |
| }, | |
| { | |
| "epoch": 0.9467631974368845, | |
| "grad_norm": 0.48076769709587097, | |
| "learning_rate": 2.1718188124071613e-06, | |
| "loss": 1.0987, | |
| "step": 9530 | |
| }, | |
| { | |
| "epoch": 0.947756653048046, | |
| "grad_norm": 0.48411691188812256, | |
| "learning_rate": 2.0915755565788728e-06, | |
| "loss": 1.1189, | |
| "step": 9540 | |
| }, | |
| { | |
| "epoch": 0.9487501086592075, | |
| "grad_norm": 0.47723978757858276, | |
| "learning_rate": 2.0128323244132826e-06, | |
| "loss": 1.11, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 0.949743564270369, | |
| "grad_norm": 0.4722275137901306, | |
| "learning_rate": 1.9355899144904263e-06, | |
| "loss": 1.1024, | |
| "step": 9560 | |
| }, | |
| { | |
| "epoch": 0.9507370198815304, | |
| "grad_norm": 0.4917920231819153, | |
| "learning_rate": 1.8598491101695978e-06, | |
| "loss": 1.1054, | |
| "step": 9570 | |
| }, | |
| { | |
| "epoch": 0.9517304754926919, | |
| "grad_norm": 0.47444161772727966, | |
| "learning_rate": 1.7856106795814729e-06, | |
| "loss": 1.1151, | |
| "step": 9580 | |
| }, | |
| { | |
| "epoch": 0.9527239311038533, | |
| "grad_norm": 0.4620038866996765, | |
| "learning_rate": 1.7128753756203484e-06, | |
| "loss": 1.0862, | |
| "step": 9590 | |
| }, | |
| { | |
| "epoch": 0.9537173867150148, | |
| "grad_norm": 0.45393404364585876, | |
| "learning_rate": 1.6416439359363487e-06, | |
| "loss": 1.0934, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.9547108423261763, | |
| "grad_norm": 0.5473287105560303, | |
| "learning_rate": 1.5719170829281646e-06, | |
| "loss": 1.0973, | |
| "step": 9610 | |
| }, | |
| { | |
| "epoch": 0.9557042979373378, | |
| "grad_norm": 0.4900316298007965, | |
| "learning_rate": 1.503695523735543e-06, | |
| "loss": 1.095, | |
| "step": 9620 | |
| }, | |
| { | |
| "epoch": 0.9566977535484993, | |
| "grad_norm": 0.49104568362236023, | |
| "learning_rate": 1.4369799502322254e-06, | |
| "loss": 1.0882, | |
| "step": 9630 | |
| }, | |
| { | |
| "epoch": 0.9576912091596608, | |
| "grad_norm": 0.4944591820240021, | |
| "learning_rate": 1.3717710390188709e-06, | |
| "loss": 1.0907, | |
| "step": 9640 | |
| }, | |
| { | |
| "epoch": 0.9586846647708223, | |
| "grad_norm": 0.46487265825271606, | |
| "learning_rate": 1.3080694514162437e-06, | |
| "loss": 1.1026, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 0.9596781203819836, | |
| "grad_norm": 0.49333804845809937, | |
| "learning_rate": 1.245875833458454e-06, | |
| "loss": 1.0832, | |
| "step": 9660 | |
| }, | |
| { | |
| "epoch": 0.9606715759931451, | |
| "grad_norm": 0.48843681812286377, | |
| "learning_rate": 1.1851908158864443e-06, | |
| "loss": 1.0941, | |
| "step": 9670 | |
| }, | |
| { | |
| "epoch": 0.9616650316043066, | |
| "grad_norm": 0.4628733694553375, | |
| "learning_rate": 1.126015014141546e-06, | |
| "loss": 1.0861, | |
| "step": 9680 | |
| }, | |
| { | |
| "epoch": 0.9626584872154681, | |
| "grad_norm": 0.5206993818283081, | |
| "learning_rate": 1.0683490283593499e-06, | |
| "loss": 1.0882, | |
| "step": 9690 | |
| }, | |
| { | |
| "epoch": 0.9636519428266296, | |
| "grad_norm": 0.4243178367614746, | |
| "learning_rate": 1.0121934433634793e-06, | |
| "loss": 1.0828, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.9646453984377911, | |
| "grad_norm": 0.4830005466938019, | |
| "learning_rate": 9.575488286597267e-07, | |
| "loss": 1.0858, | |
| "step": 9710 | |
| }, | |
| { | |
| "epoch": 0.9656388540489526, | |
| "grad_norm": 0.475365549325943, | |
| "learning_rate": 9.044157384302919e-07, | |
| "loss": 1.0938, | |
| "step": 9720 | |
| }, | |
| { | |
| "epoch": 0.966632309660114, | |
| "grad_norm": 0.5186288356781006, | |
| "learning_rate": 8.527947115281542e-07, | |
| "loss": 1.1088, | |
| "step": 9730 | |
| }, | |
| { | |
| "epoch": 0.9676257652712754, | |
| "grad_norm": 0.4813586175441742, | |
| "learning_rate": 8.026862714715587e-07, | |
| "loss": 1.1051, | |
| "step": 9740 | |
| }, | |
| { | |
| "epoch": 0.9686192208824369, | |
| "grad_norm": 0.4743388295173645, | |
| "learning_rate": 7.540909264387718e-07, | |
| "loss": 1.0817, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 0.9696126764935984, | |
| "grad_norm": 0.5275624394416809, | |
| "learning_rate": 7.07009169262901e-07, | |
| "loss": 1.0867, | |
| "step": 9760 | |
| }, | |
| { | |
| "epoch": 0.9706061321047599, | |
| "grad_norm": 0.5017628073692322, | |
| "learning_rate": 6.614414774268661e-07, | |
| "loss": 1.0886, | |
| "step": 9770 | |
| }, | |
| { | |
| "epoch": 0.9715995877159214, | |
| "grad_norm": 0.5396377444267273, | |
| "learning_rate": 6.173883130586532e-07, | |
| "loss": 1.0984, | |
| "step": 9780 | |
| }, | |
| { | |
| "epoch": 0.9725930433270829, | |
| "grad_norm": 0.4866466522216797, | |
| "learning_rate": 5.748501229264679e-07, | |
| "loss": 1.0913, | |
| "step": 9790 | |
| }, | |
| { | |
| "epoch": 0.9735864989382443, | |
| "grad_norm": 0.4783855080604553, | |
| "learning_rate": 5.338273384343895e-07, | |
| "loss": 1.0894, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.9745799545494058, | |
| "grad_norm": 0.4767221510410309, | |
| "learning_rate": 4.94320375617857e-07, | |
| "loss": 1.0908, | |
| "step": 9810 | |
| }, | |
| { | |
| "epoch": 0.9755734101605673, | |
| "grad_norm": 0.506241500377655, | |
| "learning_rate": 4.563296351394907e-07, | |
| "loss": 1.0976, | |
| "step": 9820 | |
| }, | |
| { | |
| "epoch": 0.9765668657717287, | |
| "grad_norm": 0.46907883882522583, | |
| "learning_rate": 4.1985550228506025e-07, | |
| "loss": 1.0968, | |
| "step": 9830 | |
| }, | |
| { | |
| "epoch": 0.9775603213828902, | |
| "grad_norm": 0.4644452929496765, | |
| "learning_rate": 3.8489834695953903e-07, | |
| "loss": 1.0773, | |
| "step": 9840 | |
| }, | |
| { | |
| "epoch": 0.9785537769940517, | |
| "grad_norm": 0.5320615768432617, | |
| "learning_rate": 3.5145852368337336e-07, | |
| "loss": 1.1059, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 0.9795472326052131, | |
| "grad_norm": 0.4989818036556244, | |
| "learning_rate": 3.195363715888688e-07, | |
| "loss": 1.0912, | |
| "step": 9860 | |
| }, | |
| { | |
| "epoch": 0.9805406882163746, | |
| "grad_norm": 0.4837660789489746, | |
| "learning_rate": 2.891322144167929e-07, | |
| "loss": 1.097, | |
| "step": 9870 | |
| }, | |
| { | |
| "epoch": 0.9815341438275361, | |
| "grad_norm": 0.4884801208972931, | |
| "learning_rate": 2.6024636051301116e-07, | |
| "loss": 1.1015, | |
| "step": 9880 | |
| }, | |
| { | |
| "epoch": 0.9825275994386976, | |
| "grad_norm": 0.48822057247161865, | |
| "learning_rate": 2.328791028254562e-07, | |
| "loss": 1.0956, | |
| "step": 9890 | |
| }, | |
| { | |
| "epoch": 0.9835210550498591, | |
| "grad_norm": 0.49632689356803894, | |
| "learning_rate": 2.0703071890108003e-07, | |
| "loss": 1.1018, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.9845145106610206, | |
| "grad_norm": 0.5012553334236145, | |
| "learning_rate": 1.8270147088308984e-07, | |
| "loss": 1.0819, | |
| "step": 9910 | |
| }, | |
| { | |
| "epoch": 0.985507966272182, | |
| "grad_norm": 0.4779779613018036, | |
| "learning_rate": 1.5989160550826663e-07, | |
| "loss": 1.1054, | |
| "step": 9920 | |
| }, | |
| { | |
| "epoch": 0.9865014218833434, | |
| "grad_norm": 0.4931352138519287, | |
| "learning_rate": 1.3860135410443395e-07, | |
| "loss": 1.1017, | |
| "step": 9930 | |
| }, | |
| { | |
| "epoch": 0.9874948774945049, | |
| "grad_norm": 0.44229504466056824, | |
| "learning_rate": 1.1883093258819309e-07, | |
| "loss": 1.0976, | |
| "step": 9940 | |
| }, | |
| { | |
| "epoch": 0.9884883331056664, | |
| "grad_norm": 0.462110310792923, | |
| "learning_rate": 1.0058054146265815e-07, | |
| "loss": 1.0927, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 0.9894817887168279, | |
| "grad_norm": 0.491909921169281, | |
| "learning_rate": 8.385036581545768e-08, | |
| "loss": 1.0874, | |
| "step": 9960 | |
| }, | |
| { | |
| "epoch": 0.9904752443279894, | |
| "grad_norm": 0.46828725934028625, | |
| "learning_rate": 6.864057531681955e-08, | |
| "loss": 1.0923, | |
| "step": 9970 | |
| }, | |
| { | |
| "epoch": 0.9914686999391509, | |
| "grad_norm": 0.4887199401855469, | |
| "learning_rate": 5.49513242179056e-08, | |
| "loss": 1.1085, | |
| "step": 9980 | |
| }, | |
| { | |
| "epoch": 0.9924621555503124, | |
| "grad_norm": 0.4563218057155609, | |
| "learning_rate": 4.2782751349246246e-08, | |
| "loss": 1.0857, | |
| "step": 9990 | |
| }, | |
| { | |
| "epoch": 0.9934556111614737, | |
| "grad_norm": 0.4627005159854889, | |
| "learning_rate": 3.213498011924165e-08, | |
| "loss": 1.1072, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.9944490667726352, | |
| "grad_norm": 0.48577913641929626, | |
| "learning_rate": 2.3008118513012653e-08, | |
| "loss": 1.0907, | |
| "step": 10010 | |
| }, | |
| { | |
| "epoch": 0.9954425223837967, | |
| "grad_norm": 0.45495930314064026, | |
| "learning_rate": 1.540225909130166e-08, | |
| "loss": 1.0997, | |
| "step": 10020 | |
| }, | |
| { | |
| "epoch": 0.9964359779949582, | |
| "grad_norm": 0.4800977408885956, | |
| "learning_rate": 9.317478989423477e-09, | |
| "loss": 1.0939, | |
| "step": 10030 | |
| }, | |
| { | |
| "epoch": 0.9974294336061197, | |
| "grad_norm": 0.5329833030700684, | |
| "learning_rate": 4.753839916632474e-09, | |
| "loss": 1.1044, | |
| "step": 10040 | |
| }, | |
| { | |
| "epoch": 0.9984228892172812, | |
| "grad_norm": 0.4718695282936096, | |
| "learning_rate": 1.7113881553731945e-09, | |
| "loss": 1.1051, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 0.9994163448284427, | |
| "grad_norm": 0.5231459140777588, | |
| "learning_rate": 1.901545608973265e-10, | |
| "loss": 1.1028, | |
| "step": 10060 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 10065, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1007, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.165714517491712e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |