| { | |
| "best_metric": 1.188559889793396, | |
| "best_model_checkpoint": "/Users/bbunzeck/Documents/lexdec-llamas/medium-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": 5.559729099273682, | |
| "learning_rate": 1.4999999999999999e-05, | |
| "loss": 4.4437, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0019869112223229474, | |
| "grad_norm": 3.1059770584106445, | |
| "learning_rate": 2.9999999999999997e-05, | |
| "loss": 4.0692, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0029803668334844216, | |
| "grad_norm": 2.097958564758301, | |
| "learning_rate": 4.4999999999999996e-05, | |
| "loss": 3.6821, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.003973822444645895, | |
| "grad_norm": 1.6627309322357178, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 3.4722, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.004967278055807369, | |
| "grad_norm": 1.3022949695587158, | |
| "learning_rate": 7.5e-05, | |
| "loss": 3.3631, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.005960733666968843, | |
| "grad_norm": 1.0009351968765259, | |
| "learning_rate": 8.999999999999999e-05, | |
| "loss": 3.2379, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.006954189278130317, | |
| "grad_norm": 0.9346084594726562, | |
| "learning_rate": 0.00010499999999999999, | |
| "loss": 3.1241, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.00794764488929179, | |
| "grad_norm": 1.258599042892456, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 2.9734, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.008941100500453264, | |
| "grad_norm": 1.1829655170440674, | |
| "learning_rate": 0.000135, | |
| "loss": 2.7926, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.009934556111614737, | |
| "grad_norm": 0.586632251739502, | |
| "learning_rate": 0.00015, | |
| "loss": 2.687, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.010928011722776212, | |
| "grad_norm": 1.2552680969238281, | |
| "learning_rate": 0.000165, | |
| "loss": 2.5883, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.011921467333937686, | |
| "grad_norm": 1.2504466772079468, | |
| "learning_rate": 0.00017999999999999998, | |
| "loss": 2.4944, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.01291492294509916, | |
| "grad_norm": 1.8723715543746948, | |
| "learning_rate": 0.000195, | |
| "loss": 2.4432, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.013908378556260634, | |
| "grad_norm": 1.228500485420227, | |
| "learning_rate": 0.00020999999999999998, | |
| "loss": 2.4051, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.014901834167422107, | |
| "grad_norm": 3.232206344604492, | |
| "learning_rate": 0.000225, | |
| "loss": 2.3568, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.01589528977858358, | |
| "grad_norm": 1.2030946016311646, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 2.2926, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.016888745389745056, | |
| "grad_norm": 1.6830676794052124, | |
| "learning_rate": 0.00025499999999999996, | |
| "loss": 2.2398, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.01788220100090653, | |
| "grad_norm": 1.5019638538360596, | |
| "learning_rate": 0.00027, | |
| "loss": 2.1882, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.018875656612068, | |
| "grad_norm": 1.9123952388763428, | |
| "learning_rate": 0.000285, | |
| "loss": 2.1395, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.019869112223229474, | |
| "grad_norm": 1.6992309093475342, | |
| "learning_rate": 0.0003, | |
| "loss": 2.0927, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.02086256783439095, | |
| "grad_norm": 2.0331718921661377, | |
| "learning_rate": 0.0002999992393822385, | |
| "loss": 2.0621, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.021856023445552424, | |
| "grad_norm": 1.2475999593734741, | |
| "learning_rate": 0.000299996957536668, | |
| "loss": 2.0215, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.022849479056713896, | |
| "grad_norm": 1.3022820949554443, | |
| "learning_rate": 0.0002999931544864298, | |
| "loss": 1.988, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.023842934667875373, | |
| "grad_norm": 1.366831660270691, | |
| "learning_rate": 0.00029998783027009295, | |
| "loss": 1.9385, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.024836390279036846, | |
| "grad_norm": 1.264056921005249, | |
| "learning_rate": 0.0002999809849416534, | |
| "loss": 1.9355, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.02582984589019832, | |
| "grad_norm": 1.2465442419052124, | |
| "learning_rate": 0.0002999726185705334, | |
| "loss": 1.8965, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.02682330150135979, | |
| "grad_norm": 1.3731709718704224, | |
| "learning_rate": 0.00029996273124158115, | |
| "loss": 1.8949, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.027816757112521268, | |
| "grad_norm": 1.4700868129730225, | |
| "learning_rate": 0.0002999513230550697, | |
| "loss": 1.8948, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.02881021272368274, | |
| "grad_norm": 1.144715428352356, | |
| "learning_rate": 0.00029993839412669593, | |
| "loss": 1.8495, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.029803668334844213, | |
| "grad_norm": 1.3280311822891235, | |
| "learning_rate": 0.0002999239445875795, | |
| "loss": 1.8427, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.030797123946005686, | |
| "grad_norm": 1.0667450428009033, | |
| "learning_rate": 0.00029990797458426146, | |
| "loss": 1.8068, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.03179057955716716, | |
| "grad_norm": 1.1140128374099731, | |
| "learning_rate": 0.0002998904842787027, | |
| "loss": 1.798, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.03278403516832863, | |
| "grad_norm": 1.1570714712142944, | |
| "learning_rate": 0.00029987147384828234, | |
| "loss": 1.7624, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.03377749077949011, | |
| "grad_norm": 1.1484074592590332, | |
| "learning_rate": 0.000299850943485796, | |
| "loss": 1.7552, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.034770946390651584, | |
| "grad_norm": 1.1275447607040405, | |
| "learning_rate": 0.00029982889339945384, | |
| "loss": 1.7342, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.03576440200181306, | |
| "grad_norm": 1.0428403615951538, | |
| "learning_rate": 0.00029980532381287836, | |
| "loss": 1.7309, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.03675785761297453, | |
| "grad_norm": 1.0170860290527344, | |
| "learning_rate": 0.0002997802349651021, | |
| "loss": 1.7488, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.037751313224136, | |
| "grad_norm": 1.0605427026748657, | |
| "learning_rate": 0.00029975362711056545, | |
| "loss": 1.7014, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.038744768835297476, | |
| "grad_norm": 1.165959358215332, | |
| "learning_rate": 0.00029972550051911377, | |
| "loss": 1.7096, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.03973822444645895, | |
| "grad_norm": 1.0883318185806274, | |
| "learning_rate": 0.00029969585547599494, | |
| "loss": 1.6848, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.04073168005762043, | |
| "grad_norm": 1.061632752418518, | |
| "learning_rate": 0.00029966469228185613, | |
| "loss": 1.6736, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.0417251356687819, | |
| "grad_norm": 1.1726093292236328, | |
| "learning_rate": 0.0002996320112527412, | |
| "loss": 1.6722, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.042718591279943374, | |
| "grad_norm": 1.00032639503479, | |
| "learning_rate": 0.00029959781272008695, | |
| "loss": 1.6657, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.04371204689110485, | |
| "grad_norm": 1.0932872295379639, | |
| "learning_rate": 0.0002995620970307203, | |
| "loss": 1.6602, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.04470550250226632, | |
| "grad_norm": 1.1422666311264038, | |
| "learning_rate": 0.00029952486454685437, | |
| "loss": 1.6651, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.04569895811342779, | |
| "grad_norm": 1.184408187866211, | |
| "learning_rate": 0.00029948611564608506, | |
| "loss": 1.6225, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.046692413724589266, | |
| "grad_norm": 1.0788524150848389, | |
| "learning_rate": 0.000299445850721387, | |
| "loss": 1.6422, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.047685869335750745, | |
| "grad_norm": 1.0767419338226318, | |
| "learning_rate": 0.0002994040701811098, | |
| "loss": 1.6228, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.04867932494691222, | |
| "grad_norm": 1.0282295942306519, | |
| "learning_rate": 0.00029936077444897373, | |
| "loss": 1.6293, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.04967278055807369, | |
| "grad_norm": 1.0670958757400513, | |
| "learning_rate": 0.0002993159639640655, | |
| "loss": 1.6306, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.050666236169235164, | |
| "grad_norm": 1.0718344449996948, | |
| "learning_rate": 0.00029926963918083376, | |
| "loss": 1.6101, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.05165969178039664, | |
| "grad_norm": 1.2627387046813965, | |
| "learning_rate": 0.00029922180056908456, | |
| "loss": 1.5933, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.05265314739155811, | |
| "grad_norm": 1.1007704734802246, | |
| "learning_rate": 0.00029917244861397653, | |
| "loss": 1.6004, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.05364660300271958, | |
| "grad_norm": 1.0734246969223022, | |
| "learning_rate": 0.000299121583816016, | |
| "loss": 1.6069, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.054640058613881055, | |
| "grad_norm": 1.0851387977600098, | |
| "learning_rate": 0.0002990692066910519, | |
| "loss": 1.5759, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.055633514225042535, | |
| "grad_norm": 1.089416265487671, | |
| "learning_rate": 0.00029901531777027046, | |
| "loss": 1.5894, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.05662696983620401, | |
| "grad_norm": 1.0396111011505127, | |
| "learning_rate": 0.00029895991760019007, | |
| "loss": 1.5774, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.05762042544736548, | |
| "grad_norm": 0.9830198884010315, | |
| "learning_rate": 0.00029890300674265526, | |
| "loss": 1.5824, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.058613881058526954, | |
| "grad_norm": 0.9663365483283997, | |
| "learning_rate": 0.0002988445857748316, | |
| "loss": 1.5726, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.05960733666968843, | |
| "grad_norm": 1.0020092725753784, | |
| "learning_rate": 0.0002987846552891995, | |
| "loss": 1.5493, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.0606007922808499, | |
| "grad_norm": 1.198016881942749, | |
| "learning_rate": 0.000298723215893548, | |
| "loss": 1.5481, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.06159424789201137, | |
| "grad_norm": 1.059265375137329, | |
| "learning_rate": 0.00029866026821096924, | |
| "loss": 1.5474, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.06258770350317285, | |
| "grad_norm": 1.0283076763153076, | |
| "learning_rate": 0.0002985958128798514, | |
| "loss": 1.5343, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.06358115911433432, | |
| "grad_norm": 1.0506069660186768, | |
| "learning_rate": 0.00029852985055387284, | |
| "loss": 1.545, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.06457461472549579, | |
| "grad_norm": 0.9348044991493225, | |
| "learning_rate": 0.00029846238190199504, | |
| "loss": 1.544, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.06556807033665726, | |
| "grad_norm": 0.9143947958946228, | |
| "learning_rate": 0.00029839340760845614, | |
| "loss": 1.5034, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.06656152594781875, | |
| "grad_norm": 0.9924721717834473, | |
| "learning_rate": 0.00029832292837276376, | |
| "loss": 1.5291, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.06755498155898022, | |
| "grad_norm": 0.9926716685295105, | |
| "learning_rate": 0.00029825094490968794, | |
| "loss": 1.5276, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.0685484371701417, | |
| "grad_norm": 1.0773115158081055, | |
| "learning_rate": 0.00029817745794925415, | |
| "loss": 1.5364, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.06954189278130317, | |
| "grad_norm": 0.9482336044311523, | |
| "learning_rate": 0.00029810246823673537, | |
| "loss": 1.532, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.07053534839246464, | |
| "grad_norm": 1.0166770219802856, | |
| "learning_rate": 0.0002980259765326452, | |
| "loss": 1.5199, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.07152880400362611, | |
| "grad_norm": 0.9191911220550537, | |
| "learning_rate": 0.00029794798361272964, | |
| "loss": 1.5219, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.07252225961478759, | |
| "grad_norm": 0.9159291982650757, | |
| "learning_rate": 0.0002978684902679592, | |
| "loss": 1.527, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.07351571522594906, | |
| "grad_norm": 0.8621293902397156, | |
| "learning_rate": 0.00029778749730452133, | |
| "loss": 1.4955, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.07450917083711053, | |
| "grad_norm": 0.9404488205909729, | |
| "learning_rate": 0.0002977050055438118, | |
| "loss": 1.5091, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.075502626448272, | |
| "grad_norm": 1.0706228017807007, | |
| "learning_rate": 0.0002976210158224266, | |
| "loss": 1.5201, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.07649608205943348, | |
| "grad_norm": 0.9456526041030884, | |
| "learning_rate": 0.00029753552899215344, | |
| "loss": 1.4711, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.07748953767059495, | |
| "grad_norm": 0.965735673904419, | |
| "learning_rate": 0.0002974485459199629, | |
| "loss": 1.4904, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.07848299328175642, | |
| "grad_norm": 0.9361475110054016, | |
| "learning_rate": 0.000297360067488, | |
| "loss": 1.5124, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.0794764488929179, | |
| "grad_norm": 0.9289664626121521, | |
| "learning_rate": 0.00029727009459357487, | |
| "loss": 1.4897, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.08046990450407937, | |
| "grad_norm": 0.9380794167518616, | |
| "learning_rate": 0.00029717862814915405, | |
| "loss": 1.4693, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.08146336011524086, | |
| "grad_norm": 0.885932981967926, | |
| "learning_rate": 0.00029708566908235077, | |
| "loss": 1.4781, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.08245681572640233, | |
| "grad_norm": 0.8831870555877686, | |
| "learning_rate": 0.00029699121833591605, | |
| "loss": 1.4876, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.0834502713375638, | |
| "grad_norm": 0.9156155586242676, | |
| "learning_rate": 0.0002968952768677287, | |
| "loss": 1.4791, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.08444372694872528, | |
| "grad_norm": 0.8863077163696289, | |
| "learning_rate": 0.00029679784565078584, | |
| "loss": 1.4914, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.08543718255988675, | |
| "grad_norm": 0.9516109228134155, | |
| "learning_rate": 0.00029669892567319307, | |
| "loss": 1.4757, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.08643063817104822, | |
| "grad_norm": 0.8887268900871277, | |
| "learning_rate": 0.0002965985179381542, | |
| "loss": 1.474, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.0874240937822097, | |
| "grad_norm": 0.9214118719100952, | |
| "learning_rate": 0.00029649662346396133, | |
| "loss": 1.4741, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.08841754939337117, | |
| "grad_norm": 0.9164912104606628, | |
| "learning_rate": 0.0002963932432839844, | |
| "loss": 1.4851, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.08941100500453264, | |
| "grad_norm": 0.9365724325180054, | |
| "learning_rate": 0.0002962883784466608, | |
| "loss": 1.4528, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.09040446061569411, | |
| "grad_norm": 0.9028154611587524, | |
| "learning_rate": 0.00029618203001548465, | |
| "loss": 1.4743, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.09139791622685559, | |
| "grad_norm": 0.849539577960968, | |
| "learning_rate": 0.000296074199068996, | |
| "loss": 1.4704, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.09239137183801706, | |
| "grad_norm": 0.8442996740341187, | |
| "learning_rate": 0.0002959648867007699, | |
| "loss": 1.4475, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.09338482744917853, | |
| "grad_norm": 0.7994653582572937, | |
| "learning_rate": 0.0002958540940194055, | |
| "loss": 1.4537, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.09437828306034, | |
| "grad_norm": 0.9640315175056458, | |
| "learning_rate": 0.0002957418221485145, | |
| "loss": 1.4627, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.09537173867150149, | |
| "grad_norm": 0.905960738658905, | |
| "learning_rate": 0.00029562807222671, | |
| "loss": 1.4532, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.09636519428266296, | |
| "grad_norm": 0.8957730531692505, | |
| "learning_rate": 0.00029551284540759473, | |
| "loss": 1.427, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.09735864989382444, | |
| "grad_norm": 0.9604114890098572, | |
| "learning_rate": 0.0002953961428597496, | |
| "loss": 1.456, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.09835210550498591, | |
| "grad_norm": 0.8389785289764404, | |
| "learning_rate": 0.00029527796576672173, | |
| "loss": 1.4633, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.09934556111614738, | |
| "grad_norm": 0.8791437149047852, | |
| "learning_rate": 0.00029515831532701234, | |
| "loss": 1.4458, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1000409800439604, | |
| "eval_loss": 1.4042127132415771, | |
| "eval_runtime": 69.0886, | |
| "eval_samples_per_second": 723.708, | |
| "eval_steps_per_second": 90.464, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.10033901672730886, | |
| "grad_norm": 1.017812967300415, | |
| "learning_rate": 0.00029503719275406477, | |
| "loss": 1.4523, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.10133247233847033, | |
| "grad_norm": 0.8620976209640503, | |
| "learning_rate": 0.0002949145992762521, | |
| "loss": 1.4533, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.1023259279496318, | |
| "grad_norm": 0.910183846950531, | |
| "learning_rate": 0.0002947905361368647, | |
| "loss": 1.4353, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.10331938356079327, | |
| "grad_norm": 0.9588209390640259, | |
| "learning_rate": 0.00029466500459409756, | |
| "loss": 1.4231, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.10431283917195475, | |
| "grad_norm": 0.8078910708427429, | |
| "learning_rate": 0.0002945380059210377, | |
| "loss": 1.4384, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.10530629478311622, | |
| "grad_norm": 0.9615573883056641, | |
| "learning_rate": 0.000294409541405651, | |
| "loss": 1.4252, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.10629975039427769, | |
| "grad_norm": 0.9259393811225891, | |
| "learning_rate": 0.0002942796123507693, | |
| "loss": 1.4301, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.10729320600543917, | |
| "grad_norm": 0.8774005770683289, | |
| "learning_rate": 0.0002941482200740774, | |
| "loss": 1.4322, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.10828666161660064, | |
| "grad_norm": 0.8224411606788635, | |
| "learning_rate": 0.00029401536590809913, | |
| "loss": 1.4147, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.10928011722776211, | |
| "grad_norm": 0.8637117147445679, | |
| "learning_rate": 0.00029388105120018446, | |
| "loss": 1.4186, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.1102735728389236, | |
| "grad_norm": 0.9045201539993286, | |
| "learning_rate": 0.00029374527731249533, | |
| "loss": 1.4234, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.11126702845008507, | |
| "grad_norm": 0.9177843928337097, | |
| "learning_rate": 0.00029360804562199224, | |
| "loss": 1.4105, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.11226048406124654, | |
| "grad_norm": 0.7671967148780823, | |
| "learning_rate": 0.0002934693575204199, | |
| "loss": 1.4142, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.11325393967240802, | |
| "grad_norm": 0.9050709009170532, | |
| "learning_rate": 0.00029332921441429345, | |
| "loss": 1.4373, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.11424739528356949, | |
| "grad_norm": 0.899939239025116, | |
| "learning_rate": 0.0002931876177248841, | |
| "loss": 1.4309, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.11524085089473096, | |
| "grad_norm": 0.8968805074691772, | |
| "learning_rate": 0.0002930445688882045, | |
| "loss": 1.4222, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.11623430650589243, | |
| "grad_norm": 0.8722752928733826, | |
| "learning_rate": 0.00029290006935499453, | |
| "loss": 1.4009, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.11722776211705391, | |
| "grad_norm": 0.865422785282135, | |
| "learning_rate": 0.00029275412059070636, | |
| "loss": 1.4268, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.11822121772821538, | |
| "grad_norm": 0.9007077813148499, | |
| "learning_rate": 0.0002926067240754896, | |
| "loss": 1.3844, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.11921467333937685, | |
| "grad_norm": 0.9152818322181702, | |
| "learning_rate": 0.0002924578813041764, | |
| "loss": 1.4153, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.12020812895053833, | |
| "grad_norm": 0.897662878036499, | |
| "learning_rate": 0.000292307593786266, | |
| "loss": 1.4116, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.1212015845616998, | |
| "grad_norm": 0.8988394737243652, | |
| "learning_rate": 0.00029215586304590996, | |
| "loss": 1.4117, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.12219504017286127, | |
| "grad_norm": 0.8096486330032349, | |
| "learning_rate": 0.00029200269062189623, | |
| "loss": 1.4234, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.12318849578402274, | |
| "grad_norm": 0.8231295347213745, | |
| "learning_rate": 0.0002918480780676336, | |
| "loss": 1.3923, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.12418195139518422, | |
| "grad_norm": 0.8011823296546936, | |
| "learning_rate": 0.00029169202695113617, | |
| "loss": 1.4154, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.1251754070063457, | |
| "grad_norm": 0.911997377872467, | |
| "learning_rate": 0.00029153453885500736, | |
| "loss": 1.4121, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.12616886261750718, | |
| "grad_norm": 0.9170190691947937, | |
| "learning_rate": 0.00029137561537642367, | |
| "loss": 1.3965, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.12716231822866864, | |
| "grad_norm": 0.9410304427146912, | |
| "learning_rate": 0.00029121525812711874, | |
| "loss": 1.4084, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.12815577383983012, | |
| "grad_norm": 0.850074052810669, | |
| "learning_rate": 0.00029105346873336684, | |
| "loss": 1.4064, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.12914922945099158, | |
| "grad_norm": 0.9568073749542236, | |
| "learning_rate": 0.0002908902488359664, | |
| "loss": 1.4243, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.13014268506215307, | |
| "grad_norm": 0.8900890350341797, | |
| "learning_rate": 0.0002907256000902236, | |
| "loss": 1.4018, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.13113614067331453, | |
| "grad_norm": 0.7969645857810974, | |
| "learning_rate": 0.0002905595241659351, | |
| "loss": 1.402, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.13212959628447601, | |
| "grad_norm": 0.846467137336731, | |
| "learning_rate": 0.0002903920227473717, | |
| "loss": 1.3875, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.1331230518956375, | |
| "grad_norm": 0.8248132467269897, | |
| "learning_rate": 0.00029022309753326065, | |
| "loss": 1.4073, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.13411650750679896, | |
| "grad_norm": 0.801365315914154, | |
| "learning_rate": 0.0002900527502367689, | |
| "loss": 1.3879, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.13510996311796045, | |
| "grad_norm": 0.8489567637443542, | |
| "learning_rate": 0.0002898809825854855, | |
| "loss": 1.3947, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.1361034187291219, | |
| "grad_norm": 0.8375482559204102, | |
| "learning_rate": 0.00028970779632140416, | |
| "loss": 1.3818, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.1370968743402834, | |
| "grad_norm": 0.81903475522995, | |
| "learning_rate": 0.0002895331932009055, | |
| "loss": 1.3724, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.13809032995144485, | |
| "grad_norm": 0.8887081146240234, | |
| "learning_rate": 0.0002893571749947393, | |
| "loss": 1.3881, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.13908378556260634, | |
| "grad_norm": 0.8574780821800232, | |
| "learning_rate": 0.0002891797434880066, | |
| "loss": 1.3944, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.1400772411737678, | |
| "grad_norm": 0.9162040948867798, | |
| "learning_rate": 0.0002890009004801415, | |
| "loss": 1.3753, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.14107069678492928, | |
| "grad_norm": 0.8440521359443665, | |
| "learning_rate": 0.0002888206477848928, | |
| "loss": 1.3841, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.14206415239609074, | |
| "grad_norm": 0.7782073616981506, | |
| "learning_rate": 0.00028863898723030594, | |
| "loss": 1.3865, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.14305760800725223, | |
| "grad_norm": 0.8943415880203247, | |
| "learning_rate": 0.00028845592065870413, | |
| "loss": 1.3872, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.1440510636184137, | |
| "grad_norm": 0.8131309151649475, | |
| "learning_rate": 0.00028827144992666986, | |
| "loss": 1.3913, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.14504451922957518, | |
| "grad_norm": 0.9032795429229736, | |
| "learning_rate": 0.000288085576905026, | |
| "loss": 1.3764, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.14603797484073663, | |
| "grad_norm": 0.8517566323280334, | |
| "learning_rate": 0.0002878983034788169, | |
| "loss": 1.3931, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.14703143045189812, | |
| "grad_norm": 0.8652574419975281, | |
| "learning_rate": 0.00028770963154728905, | |
| "loss": 1.3784, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.1480248860630596, | |
| "grad_norm": 0.8200552463531494, | |
| "learning_rate": 0.0002875195630238721, | |
| "loss": 1.3574, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.14901834167422107, | |
| "grad_norm": 0.8909544944763184, | |
| "learning_rate": 0.0002873280998361594, | |
| "loss": 1.353, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.15001179728538255, | |
| "grad_norm": 0.7865379452705383, | |
| "learning_rate": 0.00028713524392588827, | |
| "loss": 1.3698, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.151005252896544, | |
| "grad_norm": 0.8479179739952087, | |
| "learning_rate": 0.0002869409972489204, | |
| "loss": 1.3797, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.1519987085077055, | |
| "grad_norm": 0.8300036787986755, | |
| "learning_rate": 0.00028674536177522206, | |
| "loss": 1.3658, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.15299216411886696, | |
| "grad_norm": 0.7997909188270569, | |
| "learning_rate": 0.00028654833948884423, | |
| "loss": 1.3727, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.15398561973002844, | |
| "grad_norm": 0.8874923586845398, | |
| "learning_rate": 0.0002863499323879022, | |
| "loss": 1.3608, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.1549790753411899, | |
| "grad_norm": 0.8596087694168091, | |
| "learning_rate": 0.0002861501424845555, | |
| "loss": 1.3642, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.1559725309523514, | |
| "grad_norm": 0.8910351991653442, | |
| "learning_rate": 0.00028594897180498745, | |
| "loss": 1.3748, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.15696598656351285, | |
| "grad_norm": 0.8197429776191711, | |
| "learning_rate": 0.00028574642238938467, | |
| "loss": 1.3609, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.15795944217467434, | |
| "grad_norm": 0.7360444068908691, | |
| "learning_rate": 0.00028554249629191616, | |
| "loss": 1.3756, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.1589528977858358, | |
| "grad_norm": 0.8304336667060852, | |
| "learning_rate": 0.00028533719558071285, | |
| "loss": 1.3565, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.15994635339699728, | |
| "grad_norm": 0.8104565143585205, | |
| "learning_rate": 0.0002851305223378462, | |
| "loss": 1.3695, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.16093980900815874, | |
| "grad_norm": 0.9834653735160828, | |
| "learning_rate": 0.0002849224786593075, | |
| "loss": 1.3596, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.16193326461932023, | |
| "grad_norm": 0.8340282440185547, | |
| "learning_rate": 0.00028471306665498624, | |
| "loss": 1.3764, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.16292672023048171, | |
| "grad_norm": 0.7866162657737732, | |
| "learning_rate": 0.000284502288448649, | |
| "loss": 1.3577, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.16392017584164317, | |
| "grad_norm": 0.8105541467666626, | |
| "learning_rate": 0.00028429014617791773, | |
| "loss": 1.3625, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.16491363145280466, | |
| "grad_norm": 0.8018678426742554, | |
| "learning_rate": 0.0002840766419942481, | |
| "loss": 1.3623, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.16590708706396612, | |
| "grad_norm": 0.8143903017044067, | |
| "learning_rate": 0.0002838617780629079, | |
| "loss": 1.3574, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.1669005426751276, | |
| "grad_norm": 0.8252373337745667, | |
| "learning_rate": 0.0002836455565629546, | |
| "loss": 1.3697, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.16789399828628906, | |
| "grad_norm": 0.8147935271263123, | |
| "learning_rate": 0.0002834279796872138, | |
| "loss": 1.37, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.16888745389745055, | |
| "grad_norm": 0.8154494762420654, | |
| "learning_rate": 0.00028320904964225665, | |
| "loss": 1.3397, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.169880909508612, | |
| "grad_norm": 0.8471342325210571, | |
| "learning_rate": 0.0002829887686483775, | |
| "loss": 1.363, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.1708743651197735, | |
| "grad_norm": 0.8784846067428589, | |
| "learning_rate": 0.0002827671389395716, | |
| "loss": 1.3551, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.17186782073093496, | |
| "grad_norm": 1.6446071863174438, | |
| "learning_rate": 0.0002825441627635121, | |
| "loss": 1.3348, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.17286127634209644, | |
| "grad_norm": 0.8208043575286865, | |
| "learning_rate": 0.00028231984238152766, | |
| "loss": 1.3439, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.1738547319532579, | |
| "grad_norm": 0.8499208688735962, | |
| "learning_rate": 0.000282094180068579, | |
| "loss": 1.3722, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.1748481875644194, | |
| "grad_norm": 0.8070714473724365, | |
| "learning_rate": 0.0002818671781132364, | |
| "loss": 1.3411, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.17584164317558085, | |
| "grad_norm": 0.7955201864242554, | |
| "learning_rate": 0.0002816388388176561, | |
| "loss": 1.3451, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.17683509878674233, | |
| "grad_norm": 0.9026922583580017, | |
| "learning_rate": 0.00028140916449755706, | |
| "loss": 1.3408, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.17782855439790382, | |
| "grad_norm": 0.8375670313835144, | |
| "learning_rate": 0.0002811781574821975, | |
| "loss": 1.3623, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.17882201000906528, | |
| "grad_norm": 0.8459949493408203, | |
| "learning_rate": 0.0002809458201143513, | |
| "loss": 1.3508, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.17981546562022677, | |
| "grad_norm": 0.7782659530639648, | |
| "learning_rate": 0.0002807121547502843, | |
| "loss": 1.3457, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.18080892123138823, | |
| "grad_norm": 0.7607855200767517, | |
| "learning_rate": 0.00028047716375972996, | |
| "loss": 1.3772, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.1818023768425497, | |
| "grad_norm": 0.8945542573928833, | |
| "learning_rate": 0.000280240849525866, | |
| "loss": 1.3603, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.18279583245371117, | |
| "grad_norm": 0.855777382850647, | |
| "learning_rate": 0.0002800032144452898, | |
| "loss": 1.349, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.18378928806487266, | |
| "grad_norm": 0.7937926054000854, | |
| "learning_rate": 0.00027976426092799416, | |
| "loss": 1.35, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.18478274367603412, | |
| "grad_norm": 0.7763211727142334, | |
| "learning_rate": 0.000279523991397343, | |
| "loss": 1.3415, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.1857761992871956, | |
| "grad_norm": 0.7721043825149536, | |
| "learning_rate": 0.00027928240829004663, | |
| "loss": 1.3606, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.18676965489835706, | |
| "grad_norm": 0.8609164357185364, | |
| "learning_rate": 0.000279039514056137, | |
| "loss": 1.3358, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.18776311050951855, | |
| "grad_norm": 0.7426526546478271, | |
| "learning_rate": 0.0002787953111589431, | |
| "loss": 1.3366, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.18875656612068, | |
| "grad_norm": 0.7921430468559265, | |
| "learning_rate": 0.00027854980207506566, | |
| "loss": 1.3567, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.1897500217318415, | |
| "grad_norm": 0.7708673477172852, | |
| "learning_rate": 0.00027830298929435235, | |
| "loss": 1.3425, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.19074347734300298, | |
| "grad_norm": 0.790612518787384, | |
| "learning_rate": 0.0002780548753198723, | |
| "loss": 1.3533, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.19173693295416444, | |
| "grad_norm": 0.8253899812698364, | |
| "learning_rate": 0.00027780546266789076, | |
| "loss": 1.3372, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.19273038856532593, | |
| "grad_norm": 0.7579362392425537, | |
| "learning_rate": 0.00027755475386784364, | |
| "loss": 1.3563, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.1937238441764874, | |
| "grad_norm": 0.7975558042526245, | |
| "learning_rate": 0.00027730275146231186, | |
| "loss": 1.3424, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.19471729978764887, | |
| "grad_norm": 0.7709769010543823, | |
| "learning_rate": 0.0002770494580069954, | |
| "loss": 1.3438, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.19571075539881033, | |
| "grad_norm": 0.7712469100952148, | |
| "learning_rate": 0.0002767948760706877, | |
| "loss": 1.3448, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.19670421100997182, | |
| "grad_norm": 0.8867691159248352, | |
| "learning_rate": 0.00027653900823524925, | |
| "loss": 1.3506, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.19769766662113328, | |
| "grad_norm": 0.8083359599113464, | |
| "learning_rate": 0.00027628185709558176, | |
| "loss": 1.3494, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.19869112223229476, | |
| "grad_norm": 0.7772343158721924, | |
| "learning_rate": 0.0002760234252596015, | |
| "loss": 1.3503, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.19968457784345622, | |
| "grad_norm": 0.8192439079284668, | |
| "learning_rate": 0.000275763715348213, | |
| "loss": 1.3149, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.2000819600879208, | |
| "eval_loss": 1.318973183631897, | |
| "eval_runtime": 69.2267, | |
| "eval_samples_per_second": 722.265, | |
| "eval_steps_per_second": 90.283, | |
| "step": 2014 | |
| }, | |
| { | |
| "epoch": 0.2006780334546177, | |
| "grad_norm": 0.8178575038909912, | |
| "learning_rate": 0.00027550272999528265, | |
| "loss": 1.3337, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.20167148906577917, | |
| "grad_norm": 0.7740596532821655, | |
| "learning_rate": 0.0002752404718476116, | |
| "loss": 1.3257, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.20266494467694066, | |
| "grad_norm": 0.7558030486106873, | |
| "learning_rate": 0.0002749769435649094, | |
| "loss": 1.3365, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.20365840028810211, | |
| "grad_norm": 0.8379716277122498, | |
| "learning_rate": 0.00027471214781976645, | |
| "loss": 1.3342, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.2046518558992636, | |
| "grad_norm": 0.8267189264297485, | |
| "learning_rate": 0.00027444608729762754, | |
| "loss": 1.3241, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.2056453115104251, | |
| "grad_norm": 0.8100239634513855, | |
| "learning_rate": 0.000274178764696764, | |
| "loss": 1.3428, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.20663876712158655, | |
| "grad_norm": 0.8235071301460266, | |
| "learning_rate": 0.0002739101827282468, | |
| "loss": 1.3101, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.20763222273274803, | |
| "grad_norm": 0.7558978199958801, | |
| "learning_rate": 0.0002736403441159188, | |
| "loss": 1.3343, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.2086256783439095, | |
| "grad_norm": 0.8406065106391907, | |
| "learning_rate": 0.0002733692515963672, | |
| "loss": 1.3317, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.20961913395507098, | |
| "grad_norm": 0.7444151639938354, | |
| "learning_rate": 0.0002730969079188958, | |
| "loss": 1.3287, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.21061258956623244, | |
| "grad_norm": 0.8489364385604858, | |
| "learning_rate": 0.00027282331584549716, | |
| "loss": 1.3258, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.21160604517739393, | |
| "grad_norm": 0.7956468462944031, | |
| "learning_rate": 0.00027254847815082444, | |
| "loss": 1.326, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.21259950078855538, | |
| "grad_norm": 0.781928539276123, | |
| "learning_rate": 0.0002722723976221634, | |
| "loss": 1.3119, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.21359295639971687, | |
| "grad_norm": 0.7586515545845032, | |
| "learning_rate": 0.00027199507705940413, | |
| "loss": 1.3395, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.21458641201087833, | |
| "grad_norm": 0.8354089856147766, | |
| "learning_rate": 0.0002717165192750125, | |
| "loss": 1.3357, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.21557986762203982, | |
| "grad_norm": 0.8384989500045776, | |
| "learning_rate": 0.000271436727094002, | |
| "loss": 1.3287, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.21657332323320128, | |
| "grad_norm": 0.7756235003471375, | |
| "learning_rate": 0.0002711557033539044, | |
| "loss": 1.3166, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.21756677884436276, | |
| "grad_norm": 0.8618668913841248, | |
| "learning_rate": 0.0002708734509047419, | |
| "loss": 1.3362, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.21856023445552422, | |
| "grad_norm": 0.8257716298103333, | |
| "learning_rate": 0.0002705899726089973, | |
| "loss": 1.3124, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.2195536900666857, | |
| "grad_norm": 0.7277160286903381, | |
| "learning_rate": 0.00027030527134158585, | |
| "loss": 1.3273, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.2205471456778472, | |
| "grad_norm": 0.8157845735549927, | |
| "learning_rate": 0.00027001934998982524, | |
| "loss": 1.2907, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.22154060128900865, | |
| "grad_norm": 0.7717984318733215, | |
| "learning_rate": 0.0002697322114534071, | |
| "loss": 1.3261, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.22253405690017014, | |
| "grad_norm": 0.8100425601005554, | |
| "learning_rate": 0.0002694438586443669, | |
| "loss": 1.3264, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.2235275125113316, | |
| "grad_norm": 0.7536413669586182, | |
| "learning_rate": 0.0002691542944870549, | |
| "loss": 1.3111, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.2245209681224931, | |
| "grad_norm": 0.8206626772880554, | |
| "learning_rate": 0.00026886352191810627, | |
| "loss": 1.2977, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.22551442373365455, | |
| "grad_norm": 0.8038385510444641, | |
| "learning_rate": 0.0002685715438864116, | |
| "loss": 1.3063, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.22650787934481603, | |
| "grad_norm": 0.8154683709144592, | |
| "learning_rate": 0.00026827836335308644, | |
| "loss": 1.3096, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.2275013349559775, | |
| "grad_norm": 0.7548006772994995, | |
| "learning_rate": 0.00026798398329144166, | |
| "loss": 1.3053, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.22849479056713898, | |
| "grad_norm": 0.8106780648231506, | |
| "learning_rate": 0.0002676884066869535, | |
| "loss": 1.3318, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.22948824617830044, | |
| "grad_norm": 0.786957859992981, | |
| "learning_rate": 0.0002673916365372327, | |
| "loss": 1.2902, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.23048170178946192, | |
| "grad_norm": 0.754601776599884, | |
| "learning_rate": 0.0002670936758519946, | |
| "loss": 1.3268, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.23147515740062338, | |
| "grad_norm": 0.8212381601333618, | |
| "learning_rate": 0.0002667945276530284, | |
| "loss": 1.2995, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.23246861301178487, | |
| "grad_norm": 0.7691715955734253, | |
| "learning_rate": 0.0002664941949741665, | |
| "loss": 1.3189, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.23346206862294633, | |
| "grad_norm": 0.873024046421051, | |
| "learning_rate": 0.0002661926808612539, | |
| "loss": 1.32, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.23445552423410781, | |
| "grad_norm": 0.752914309501648, | |
| "learning_rate": 0.000265889988372117, | |
| "loss": 1.3279, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.2354489798452693, | |
| "grad_norm": 0.74873286485672, | |
| "learning_rate": 0.0002655861205765331, | |
| "loss": 1.3221, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.23644243545643076, | |
| "grad_norm": 0.8336202502250671, | |
| "learning_rate": 0.00026528108055619856, | |
| "loss": 1.3098, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.23743589106759225, | |
| "grad_norm": 0.7261322736740112, | |
| "learning_rate": 0.0002649748714046983, | |
| "loss": 1.3052, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.2384293466787537, | |
| "grad_norm": 0.7554280757904053, | |
| "learning_rate": 0.0002646674962274739, | |
| "loss": 1.3127, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.2394228022899152, | |
| "grad_norm": 0.7936358451843262, | |
| "learning_rate": 0.0002643589581417922, | |
| "loss": 1.3136, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.24041625790107665, | |
| "grad_norm": 0.8146955966949463, | |
| "learning_rate": 0.0002640492602767139, | |
| "loss": 1.321, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.24140971351223814, | |
| "grad_norm": 0.8108673691749573, | |
| "learning_rate": 0.00026373840577306165, | |
| "loss": 1.304, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.2424031691233996, | |
| "grad_norm": 0.8439114689826965, | |
| "learning_rate": 0.00026342639778338813, | |
| "loss": 1.3037, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.24339662473456108, | |
| "grad_norm": 0.8041278719902039, | |
| "learning_rate": 0.0002631132394719443, | |
| "loss": 1.3014, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.24439008034572254, | |
| "grad_norm": 0.8038398027420044, | |
| "learning_rate": 0.0002627989340146471, | |
| "loss": 1.3366, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.24538353595688403, | |
| "grad_norm": 0.7703317403793335, | |
| "learning_rate": 0.0002624834845990475, | |
| "loss": 1.3088, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.2463769915680455, | |
| "grad_norm": 0.8210429549217224, | |
| "learning_rate": 0.00026216689442429774, | |
| "loss": 1.3074, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.24737044717920698, | |
| "grad_norm": 0.7259739637374878, | |
| "learning_rate": 0.0002618491667011193, | |
| "loss": 1.305, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.24836390279036843, | |
| "grad_norm": 0.7319886684417725, | |
| "learning_rate": 0.0002615303046517703, | |
| "loss": 1.3087, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.24935735840152992, | |
| "grad_norm": 0.7114730477333069, | |
| "learning_rate": 0.00026121031151001245, | |
| "loss": 1.3051, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.2503508140126914, | |
| "grad_norm": 0.7243214845657349, | |
| "learning_rate": 0.00026088919052107867, | |
| "loss": 1.3204, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.25134426962385287, | |
| "grad_norm": 0.7545714974403381, | |
| "learning_rate": 0.00026056694494164, | |
| "loss": 1.311, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.25233772523501435, | |
| "grad_norm": 0.7784477472305298, | |
| "learning_rate": 0.00026024357803977263, | |
| "loss": 1.2978, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.25333118084617584, | |
| "grad_norm": 0.758330762386322, | |
| "learning_rate": 0.0002599190930949247, | |
| "loss": 1.3217, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.25432463645733727, | |
| "grad_norm": 0.7913810610771179, | |
| "learning_rate": 0.00025959349339788295, | |
| "loss": 1.3168, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.25531809206849876, | |
| "grad_norm": 0.7649120688438416, | |
| "learning_rate": 0.00025926678225073963, | |
| "loss": 1.3048, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.25631154767966025, | |
| "grad_norm": 0.7658051252365112, | |
| "learning_rate": 0.00025893896296685866, | |
| "loss": 1.3037, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.25730500329082173, | |
| "grad_norm": 0.8071420788764954, | |
| "learning_rate": 0.00025861003887084246, | |
| "loss": 1.3125, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.25829845890198316, | |
| "grad_norm": 0.7597982883453369, | |
| "learning_rate": 0.0002582800132984977, | |
| "loss": 1.3048, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.25929191451314465, | |
| "grad_norm": 0.8165181875228882, | |
| "learning_rate": 0.0002579488895968019, | |
| "loss": 1.2943, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.26028537012430614, | |
| "grad_norm": 0.7120139598846436, | |
| "learning_rate": 0.0002576166711238694, | |
| "loss": 1.2827, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.2612788257354676, | |
| "grad_norm": 0.6952322721481323, | |
| "learning_rate": 0.0002572833612489169, | |
| "loss": 1.326, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.26227228134662905, | |
| "grad_norm": 0.8077282309532166, | |
| "learning_rate": 0.00025694896335223024, | |
| "loss": 1.3102, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.26326573695779054, | |
| "grad_norm": 0.8499687910079956, | |
| "learning_rate": 0.00025661348082512884, | |
| "loss": 1.303, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.26425919256895203, | |
| "grad_norm": 0.7348212599754333, | |
| "learning_rate": 0.0002562769170699324, | |
| "loss": 1.2988, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.2652526481801135, | |
| "grad_norm": 0.8220813274383545, | |
| "learning_rate": 0.00025593927549992594, | |
| "loss": 1.2973, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.266246103791275, | |
| "grad_norm": 0.8074153661727905, | |
| "learning_rate": 0.000255600559539325, | |
| "loss": 1.2978, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.26723955940243643, | |
| "grad_norm": 0.7362489104270935, | |
| "learning_rate": 0.0002552607726232413, | |
| "loss": 1.2929, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.2682330150135979, | |
| "grad_norm": 0.7918626070022583, | |
| "learning_rate": 0.0002549199181976477, | |
| "loss": 1.3131, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.2692264706247594, | |
| "grad_norm": 0.7445553541183472, | |
| "learning_rate": 0.00025457799971934325, | |
| "loss": 1.3053, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.2702199262359209, | |
| "grad_norm": 0.7787147760391235, | |
| "learning_rate": 0.0002542350206559182, | |
| "loss": 1.3, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.2712133818470823, | |
| "grad_norm": 0.7520889639854431, | |
| "learning_rate": 0.0002538909844857187, | |
| "loss": 1.3229, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.2722068374582438, | |
| "grad_norm": 0.7582774758338928, | |
| "learning_rate": 0.00025354589469781186, | |
| "loss": 1.3068, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.2732002930694053, | |
| "grad_norm": 0.7228103876113892, | |
| "learning_rate": 0.0002531997547919499, | |
| "loss": 1.3073, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.2741937486805668, | |
| "grad_norm": 0.8127725720405579, | |
| "learning_rate": 0.0002528525682785349, | |
| "loss": 1.2972, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.2751872042917282, | |
| "grad_norm": 0.8237895369529724, | |
| "learning_rate": 0.0002525043386785833, | |
| "loss": 1.3037, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.2761806599028897, | |
| "grad_norm": 0.760171115398407, | |
| "learning_rate": 0.00025215506952369003, | |
| "loss": 1.2987, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.2771741155140512, | |
| "grad_norm": 0.7730082273483276, | |
| "learning_rate": 0.0002518047643559927, | |
| "loss": 1.3104, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.2781675711252127, | |
| "grad_norm": 0.7766639590263367, | |
| "learning_rate": 0.00025145342672813574, | |
| "loss": 1.3071, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.27916102673637416, | |
| "grad_norm": 0.7805323600769043, | |
| "learning_rate": 0.00025110106020323437, | |
| "loss": 1.3013, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.2801544823475356, | |
| "grad_norm": 0.768355131149292, | |
| "learning_rate": 0.0002507476683548384, | |
| "loss": 1.302, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.2811479379586971, | |
| "grad_norm": 0.7298022508621216, | |
| "learning_rate": 0.00025039325476689607, | |
| "loss": 1.2862, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.28214139356985857, | |
| "grad_norm": 0.7600789070129395, | |
| "learning_rate": 0.00025003782303371767, | |
| "loss": 1.2975, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.28313484918102005, | |
| "grad_norm": 0.7599009275436401, | |
| "learning_rate": 0.00024968137675993897, | |
| "loss": 1.2996, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.2841283047921815, | |
| "grad_norm": 0.7398258447647095, | |
| "learning_rate": 0.00024932391956048497, | |
| "loss": 1.2822, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.28512176040334297, | |
| "grad_norm": 0.749475359916687, | |
| "learning_rate": 0.00024896545506053285, | |
| "loss": 1.3159, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.28611521601450446, | |
| "grad_norm": 0.8502193093299866, | |
| "learning_rate": 0.00024860598689547555, | |
| "loss": 1.2759, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.28710867162566595, | |
| "grad_norm": 0.789801299571991, | |
| "learning_rate": 0.0002482455187108846, | |
| "loss": 1.2588, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.2881021272368274, | |
| "grad_norm": 0.8479325175285339, | |
| "learning_rate": 0.00024788405416247347, | |
| "loss": 1.2888, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.28909558284798886, | |
| "grad_norm": 0.6943687200546265, | |
| "learning_rate": 0.00024752159691606015, | |
| "loss": 1.3073, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.29008903845915035, | |
| "grad_norm": 0.7406265735626221, | |
| "learning_rate": 0.0002471581506475303, | |
| "loss": 1.2962, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.29108249407031184, | |
| "grad_norm": 0.7715753316879272, | |
| "learning_rate": 0.00024679371904279973, | |
| "loss": 1.2865, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.29207594968147327, | |
| "grad_norm": 0.8168973326683044, | |
| "learning_rate": 0.0002464283057977771, | |
| "loss": 1.3033, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.29306940529263475, | |
| "grad_norm": 0.7826045155525208, | |
| "learning_rate": 0.00024606191461832655, | |
| "loss": 1.2871, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.29406286090379624, | |
| "grad_norm": 0.7249663472175598, | |
| "learning_rate": 0.00024569454922022987, | |
| "loss": 1.272, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.29505631651495773, | |
| "grad_norm": 0.743850827217102, | |
| "learning_rate": 0.00024532621332914897, | |
| "loss": 1.2845, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.2960497721261192, | |
| "grad_norm": 0.7418140769004822, | |
| "learning_rate": 0.0002449569106805882, | |
| "loss": 1.2804, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.29704322773728065, | |
| "grad_norm": 0.7663955688476562, | |
| "learning_rate": 0.00024458664501985624, | |
| "loss": 1.2935, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.29803668334844213, | |
| "grad_norm": 0.767795741558075, | |
| "learning_rate": 0.0002442154201020283, | |
| "loss": 1.2942, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.2990301389596036, | |
| "grad_norm": 0.7638338208198547, | |
| "learning_rate": 0.00024384323969190789, | |
| "loss": 1.2948, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.3000235945707651, | |
| "grad_norm": 0.7255727648735046, | |
| "learning_rate": 0.0002434701075639888, | |
| "loss": 1.2757, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.30012294013188123, | |
| "eval_loss": 1.2853599786758423, | |
| "eval_runtime": 69.1882, | |
| "eval_samples_per_second": 722.667, | |
| "eval_steps_per_second": 90.333, | |
| "step": 3021 | |
| }, | |
| { | |
| "epoch": 0.30101705018192654, | |
| "grad_norm": 0.721829891204834, | |
| "learning_rate": 0.0002430960275024166, | |
| "loss": 1.277, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.302010505793088, | |
| "grad_norm": 0.7433609962463379, | |
| "learning_rate": 0.00024272100330095055, | |
| "loss": 1.2969, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.3030039614042495, | |
| "grad_norm": 0.751215398311615, | |
| "learning_rate": 0.00024234503876292485, | |
| "loss": 1.2982, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.303997417015411, | |
| "grad_norm": 0.8018587827682495, | |
| "learning_rate": 0.00024196813770121026, | |
| "loss": 1.292, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.30499087262657243, | |
| "grad_norm": 0.7160184383392334, | |
| "learning_rate": 0.00024159030393817538, | |
| "loss": 1.2897, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.3059843282377339, | |
| "grad_norm": 0.8007465600967407, | |
| "learning_rate": 0.0002412115413056477, | |
| "loss": 1.2677, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.3069777838488954, | |
| "grad_norm": 0.7749207019805908, | |
| "learning_rate": 0.00024083185364487522, | |
| "loss": 1.289, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.3079712394600569, | |
| "grad_norm": 0.794239342212677, | |
| "learning_rate": 0.0002404512448064868, | |
| "loss": 1.2561, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.3089646950712184, | |
| "grad_norm": 0.7076348662376404, | |
| "learning_rate": 0.00024006971865045377, | |
| "loss": 1.3002, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.3099581506823798, | |
| "grad_norm": 0.6787956953048706, | |
| "learning_rate": 0.0002396872790460504, | |
| "loss": 1.2813, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.3109516062935413, | |
| "grad_norm": 0.7118437886238098, | |
| "learning_rate": 0.00023930392987181473, | |
| "loss": 1.2766, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.3119450619047028, | |
| "grad_norm": 0.7802293300628662, | |
| "learning_rate": 0.00023891967501550937, | |
| "loss": 1.2734, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.31293851751586427, | |
| "grad_norm": 0.7025743722915649, | |
| "learning_rate": 0.00023853451837408185, | |
| "loss": 1.269, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.3139319731270257, | |
| "grad_norm": 0.6797525882720947, | |
| "learning_rate": 0.00023814846385362525, | |
| "loss": 1.2858, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.3149254287381872, | |
| "grad_norm": 0.8364133238792419, | |
| "learning_rate": 0.00023776151536933864, | |
| "loss": 1.264, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.31591888434934867, | |
| "grad_norm": 0.7367081046104431, | |
| "learning_rate": 0.00023737367684548716, | |
| "loss": 1.2892, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.31691233996051016, | |
| "grad_norm": 0.7518606781959534, | |
| "learning_rate": 0.00023698495221536245, | |
| "loss": 1.2765, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.3179057955716716, | |
| "grad_norm": 0.8132895827293396, | |
| "learning_rate": 0.00023659534542124261, | |
| "loss": 1.2619, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.3188992511828331, | |
| "grad_norm": 0.8604214191436768, | |
| "learning_rate": 0.0002362048604143522, | |
| "loss": 1.2829, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.31989270679399456, | |
| "grad_norm": 0.7896199822425842, | |
| "learning_rate": 0.00023581350115482245, | |
| "loss": 1.2721, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.32088616240515605, | |
| "grad_norm": 0.7754125595092773, | |
| "learning_rate": 0.00023542127161165063, | |
| "loss": 1.2925, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.3218796180163175, | |
| "grad_norm": 0.7678704261779785, | |
| "learning_rate": 0.00023502817576266023, | |
| "loss": 1.2781, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.32287307362747897, | |
| "grad_norm": 0.8086171746253967, | |
| "learning_rate": 0.00023463421759446033, | |
| "loss": 1.2685, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.32386652923864045, | |
| "grad_norm": 0.7817544341087341, | |
| "learning_rate": 0.00023423940110240538, | |
| "loss": 1.2965, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.32485998484980194, | |
| "grad_norm": 0.7475219368934631, | |
| "learning_rate": 0.00023384373029055445, | |
| "loss": 1.2697, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.32585344046096343, | |
| "grad_norm": 0.7677096724510193, | |
| "learning_rate": 0.00023344720917163093, | |
| "loss": 1.2632, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.32684689607212486, | |
| "grad_norm": 0.7176823616027832, | |
| "learning_rate": 0.00023304984176698157, | |
| "loss": 1.266, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.32784035168328635, | |
| "grad_norm": 0.7511374950408936, | |
| "learning_rate": 0.00023265163210653569, | |
| "loss": 1.2843, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.32883380729444783, | |
| "grad_norm": 0.7132532000541687, | |
| "learning_rate": 0.00023225258422876455, | |
| "loss": 1.273, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.3298272629056093, | |
| "grad_norm": 0.7219734191894531, | |
| "learning_rate": 0.00023185270218064021, | |
| "loss": 1.2628, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.33082071851677075, | |
| "grad_norm": 0.7662040591239929, | |
| "learning_rate": 0.00023145199001759452, | |
| "loss": 1.2691, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.33181417412793224, | |
| "grad_norm": 0.7833425998687744, | |
| "learning_rate": 0.00023105045180347788, | |
| "loss": 1.2875, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.3328076297390937, | |
| "grad_norm": 0.78608238697052, | |
| "learning_rate": 0.00023064809161051842, | |
| "loss": 1.2696, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.3338010853502552, | |
| "grad_norm": 0.8074819445610046, | |
| "learning_rate": 0.00023024491351928019, | |
| "loss": 1.2654, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.33479454096141664, | |
| "grad_norm": 0.8331537246704102, | |
| "learning_rate": 0.00022984092161862206, | |
| "loss": 1.2829, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.33578799657257813, | |
| "grad_norm": 0.7963647246360779, | |
| "learning_rate": 0.00022943612000565625, | |
| "loss": 1.2555, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.3367814521837396, | |
| "grad_norm": 0.7512292265892029, | |
| "learning_rate": 0.00022903051278570677, | |
| "loss": 1.2697, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.3377749077949011, | |
| "grad_norm": 0.7556365132331848, | |
| "learning_rate": 0.00022862410407226762, | |
| "loss": 1.2895, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.3387683634060626, | |
| "grad_norm": 0.8226991891860962, | |
| "learning_rate": 0.00022821689798696132, | |
| "loss": 1.2827, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.339761819017224, | |
| "grad_norm": 0.7264415621757507, | |
| "learning_rate": 0.00022780889865949689, | |
| "loss": 1.2698, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.3407552746283855, | |
| "grad_norm": 0.705476701259613, | |
| "learning_rate": 0.00022740011022762827, | |
| "loss": 1.274, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.341748730239547, | |
| "grad_norm": 0.7758600115776062, | |
| "learning_rate": 0.00022699053683711187, | |
| "loss": 1.2898, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.3427421858507085, | |
| "grad_norm": 0.7096195220947266, | |
| "learning_rate": 0.00022658018264166502, | |
| "loss": 1.2754, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.3437356414618699, | |
| "grad_norm": 0.7327998876571655, | |
| "learning_rate": 0.00022616905180292363, | |
| "loss": 1.2638, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.3447290970730314, | |
| "grad_norm": 0.7593440413475037, | |
| "learning_rate": 0.00022575714849039984, | |
| "loss": 1.2682, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.3457225526841929, | |
| "grad_norm": 0.7941192388534546, | |
| "learning_rate": 0.00022534447688144006, | |
| "loss": 1.2853, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.34671600829535437, | |
| "grad_norm": 0.681774914264679, | |
| "learning_rate": 0.0002249310411611824, | |
| "loss": 1.2451, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.3477094639065158, | |
| "grad_norm": 0.7430617809295654, | |
| "learning_rate": 0.00022451684552251414, | |
| "loss": 1.2779, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.3487029195176773, | |
| "grad_norm": 0.7101045846939087, | |
| "learning_rate": 0.00022410189416602943, | |
| "loss": 1.2774, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.3496963751288388, | |
| "grad_norm": 0.8233627080917358, | |
| "learning_rate": 0.00022368619129998655, | |
| "loss": 1.2802, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.35068983074000026, | |
| "grad_norm": 0.6701515316963196, | |
| "learning_rate": 0.00022326974114026535, | |
| "loss": 1.266, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.3516832863511617, | |
| "grad_norm": 0.7553005218505859, | |
| "learning_rate": 0.00022285254791032423, | |
| "loss": 1.2663, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.3526767419623232, | |
| "grad_norm": 0.7621037364006042, | |
| "learning_rate": 0.0002224346158411577, | |
| "loss": 1.2617, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.35367019757348467, | |
| "grad_norm": 0.7574499845504761, | |
| "learning_rate": 0.00022201594917125306, | |
| "loss": 1.2502, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.35466365318464615, | |
| "grad_norm": 0.7560155987739563, | |
| "learning_rate": 0.00022159655214654787, | |
| "loss": 1.2612, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.35565710879580764, | |
| "grad_norm": 0.7298286557197571, | |
| "learning_rate": 0.00022117642902038629, | |
| "loss": 1.2576, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.3566505644069691, | |
| "grad_norm": 0.7721870541572571, | |
| "learning_rate": 0.00022075558405347659, | |
| "loss": 1.2801, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.35764402001813056, | |
| "grad_norm": 0.7403223514556885, | |
| "learning_rate": 0.0002203340215138475, | |
| "loss": 1.287, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.35863747562929205, | |
| "grad_norm": 0.7545426487922668, | |
| "learning_rate": 0.00021991174567680512, | |
| "loss": 1.2597, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.35963093124045353, | |
| "grad_norm": 0.7739009857177734, | |
| "learning_rate": 0.00021948876082488936, | |
| "loss": 1.2585, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.36062438685161496, | |
| "grad_norm": 0.6917877793312073, | |
| "learning_rate": 0.0002190650712478309, | |
| "loss": 1.2685, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.36161784246277645, | |
| "grad_norm": 0.6634495258331299, | |
| "learning_rate": 0.0002186406812425073, | |
| "loss": 1.2531, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.36261129807393794, | |
| "grad_norm": 0.7418808937072754, | |
| "learning_rate": 0.00021821559511289948, | |
| "loss": 1.2711, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.3636047536850994, | |
| "grad_norm": 0.7590010762214661, | |
| "learning_rate": 0.0002177898171700483, | |
| "loss": 1.2544, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.36459820929626086, | |
| "grad_norm": 0.7070173025131226, | |
| "learning_rate": 0.00021736335173201066, | |
| "loss": 1.2448, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.36559166490742234, | |
| "grad_norm": 0.7456412315368652, | |
| "learning_rate": 0.00021693620312381568, | |
| "loss": 1.2596, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.36658512051858383, | |
| "grad_norm": 0.8234733939170837, | |
| "learning_rate": 0.00021650837567742094, | |
| "loss": 1.2806, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.3675785761297453, | |
| "grad_norm": 0.8216111660003662, | |
| "learning_rate": 0.00021607987373166855, | |
| "loss": 1.268, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.3685720317409068, | |
| "grad_norm": 0.7473270297050476, | |
| "learning_rate": 0.00021565070163224095, | |
| "loss": 1.2604, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.36956548735206823, | |
| "grad_norm": 0.68471759557724, | |
| "learning_rate": 0.0002152208637316172, | |
| "loss": 1.2857, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.3705589429632297, | |
| "grad_norm": 0.7492942810058594, | |
| "learning_rate": 0.00021479036438902847, | |
| "loss": 1.2658, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.3715523985743912, | |
| "grad_norm": 0.8198530673980713, | |
| "learning_rate": 0.00021435920797041404, | |
| "loss": 1.2513, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.3725458541855527, | |
| "grad_norm": 0.7183761596679688, | |
| "learning_rate": 0.00021392739884837697, | |
| "loss": 1.2581, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.3735393097967141, | |
| "grad_norm": 0.7246140837669373, | |
| "learning_rate": 0.00021349494140213986, | |
| "loss": 1.2668, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.3745327654078756, | |
| "grad_norm": 0.7589888572692871, | |
| "learning_rate": 0.00021306184001750012, | |
| "loss": 1.2448, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.3755262210190371, | |
| "grad_norm": 0.7513408660888672, | |
| "learning_rate": 0.00021262809908678584, | |
| "loss": 1.2527, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.3765196766301986, | |
| "grad_norm": 0.698563814163208, | |
| "learning_rate": 0.00021219372300881117, | |
| "loss": 1.2796, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.37751313224136, | |
| "grad_norm": 0.7656630873680115, | |
| "learning_rate": 0.0002117587161888315, | |
| "loss": 1.2551, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.3785065878525215, | |
| "grad_norm": 0.7977383136749268, | |
| "learning_rate": 0.00021132308303849906, | |
| "loss": 1.2768, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.379500043463683, | |
| "grad_norm": 0.7182868123054504, | |
| "learning_rate": 0.000210886827975818, | |
| "loss": 1.272, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.3804934990748445, | |
| "grad_norm": 0.7245107293128967, | |
| "learning_rate": 0.0002104499554250996, | |
| "loss": 1.2431, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.38148695468600596, | |
| "grad_norm": 0.7931720614433289, | |
| "learning_rate": 0.00021001246981691758, | |
| "loss": 1.2603, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.3824804102971674, | |
| "grad_norm": 0.7385146021842957, | |
| "learning_rate": 0.0002095743755880628, | |
| "loss": 1.2768, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.3834738659083289, | |
| "grad_norm": 0.795901358127594, | |
| "learning_rate": 0.0002091356771814987, | |
| "loss": 1.2691, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.38446732151949037, | |
| "grad_norm": 0.794334352016449, | |
| "learning_rate": 0.0002086963790463159, | |
| "loss": 1.2753, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.38546077713065185, | |
| "grad_norm": 0.7204954028129578, | |
| "learning_rate": 0.0002082564856376873, | |
| "loss": 1.2546, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.3864542327418133, | |
| "grad_norm": 0.7791348099708557, | |
| "learning_rate": 0.0002078160014168227, | |
| "loss": 1.2683, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.3874476883529748, | |
| "grad_norm": 0.81683748960495, | |
| "learning_rate": 0.00020737493085092382, | |
| "loss": 1.2589, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.38844114396413626, | |
| "grad_norm": 0.7451496124267578, | |
| "learning_rate": 0.00020693327841313867, | |
| "loss": 1.2776, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.38943459957529775, | |
| "grad_norm": 0.7492960095405579, | |
| "learning_rate": 0.00020649104858251648, | |
| "loss": 1.2701, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.3904280551864592, | |
| "grad_norm": 0.6795782446861267, | |
| "learning_rate": 0.00020604824584396204, | |
| "loss": 1.2569, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 0.39142151079762066, | |
| "grad_norm": 0.7614774107933044, | |
| "learning_rate": 0.0002056048746881905, | |
| "loss": 1.2607, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.39241496640878215, | |
| "grad_norm": 0.7439120411872864, | |
| "learning_rate": 0.00020516093961168137, | |
| "loss": 1.2634, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.39340842201994364, | |
| "grad_norm": 0.7193949818611145, | |
| "learning_rate": 0.00020471644511663352, | |
| "loss": 1.2426, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.39440187763110507, | |
| "grad_norm": 0.794485867023468, | |
| "learning_rate": 0.00020427139571091893, | |
| "loss": 1.2463, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 0.39539533324226656, | |
| "grad_norm": 0.7421975135803223, | |
| "learning_rate": 0.00020382579590803748, | |
| "loss": 1.2437, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.39638878885342804, | |
| "grad_norm": 0.7745868563652039, | |
| "learning_rate": 0.00020337965022707084, | |
| "loss": 1.2521, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 0.39738224446458953, | |
| "grad_norm": 0.6921913623809814, | |
| "learning_rate": 0.00020293296319263664, | |
| "loss": 1.2608, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.398375700075751, | |
| "grad_norm": 0.754993200302124, | |
| "learning_rate": 0.00020248573933484286, | |
| "loss": 1.2569, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 0.39936915568691245, | |
| "grad_norm": 0.7566757798194885, | |
| "learning_rate": 0.00020203798318924152, | |
| "loss": 1.2528, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.4001639201758416, | |
| "eval_loss": 1.2557162046432495, | |
| "eval_runtime": 69.4049, | |
| "eval_samples_per_second": 720.41, | |
| "eval_steps_per_second": 90.051, | |
| "step": 4028 | |
| }, | |
| { | |
| "epoch": 0.40036261129807393, | |
| "grad_norm": 0.7448116540908813, | |
| "learning_rate": 0.00020158969929678306, | |
| "loss": 1.2405, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 0.4013560669092354, | |
| "grad_norm": 0.7187281250953674, | |
| "learning_rate": 0.00020114089220377005, | |
| "loss": 1.2611, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.4023495225203969, | |
| "grad_norm": 0.7817150354385376, | |
| "learning_rate": 0.00020069156646181107, | |
| "loss": 1.2694, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.40334297813155834, | |
| "grad_norm": 0.7916039228439331, | |
| "learning_rate": 0.0002002417266277746, | |
| "loss": 1.2618, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.4043364337427198, | |
| "grad_norm": 0.7549243569374084, | |
| "learning_rate": 0.00019979137726374295, | |
| "loss": 1.2469, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 0.4053298893538813, | |
| "grad_norm": 0.8509347438812256, | |
| "learning_rate": 0.0001993405229369658, | |
| "loss": 1.2616, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.4063233449650428, | |
| "grad_norm": 0.7418258786201477, | |
| "learning_rate": 0.0001988891682198139, | |
| "loss": 1.2621, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 0.40731680057620423, | |
| "grad_norm": 0.7590004801750183, | |
| "learning_rate": 0.00019843731768973274, | |
| "loss": 1.2518, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.4083102561873657, | |
| "grad_norm": 0.6706249117851257, | |
| "learning_rate": 0.00019798497592919624, | |
| "loss": 1.2422, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 0.4093037117985272, | |
| "grad_norm": 0.7325282096862793, | |
| "learning_rate": 0.00019753214752566006, | |
| "loss": 1.2604, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.4102971674096887, | |
| "grad_norm": 0.7502084374427795, | |
| "learning_rate": 0.0001970788370715152, | |
| "loss": 1.2398, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 0.4112906230208502, | |
| "grad_norm": 0.6897179484367371, | |
| "learning_rate": 0.0001966250491640415, | |
| "loss": 1.2542, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.4122840786320116, | |
| "grad_norm": 0.8166541457176208, | |
| "learning_rate": 0.0001961707884053608, | |
| "loss": 1.2282, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.4132775342431731, | |
| "grad_norm": 0.7377915382385254, | |
| "learning_rate": 0.0001957160594023905, | |
| "loss": 1.2418, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.4142709898543346, | |
| "grad_norm": 0.7019013166427612, | |
| "learning_rate": 0.00019526086676679663, | |
| "loss": 1.2327, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 0.41526444546549607, | |
| "grad_norm": 0.7024330496788025, | |
| "learning_rate": 0.00019480521511494727, | |
| "loss": 1.241, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.4162579010766575, | |
| "grad_norm": 0.6933190822601318, | |
| "learning_rate": 0.0001943491090678656, | |
| "loss": 1.2261, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 0.417251356687819, | |
| "grad_norm": 0.7336214780807495, | |
| "learning_rate": 0.00019389255325118305, | |
| "loss": 1.2463, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.4182448122989805, | |
| "grad_norm": 0.6893949508666992, | |
| "learning_rate": 0.0001934355522950926, | |
| "loss": 1.2561, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 0.41923826791014196, | |
| "grad_norm": 0.7990818023681641, | |
| "learning_rate": 0.00019297811083430138, | |
| "loss": 1.2569, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 0.4202317235213034, | |
| "grad_norm": 0.751339316368103, | |
| "learning_rate": 0.0001925202335079841, | |
| "loss": 1.2413, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 0.4212251791324649, | |
| "grad_norm": 0.7313910722732544, | |
| "learning_rate": 0.0001920619249597357, | |
| "loss": 1.2712, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 0.42221863474362636, | |
| "grad_norm": 0.7901277542114258, | |
| "learning_rate": 0.00019160318983752465, | |
| "loss": 1.2633, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.42321209035478785, | |
| "grad_norm": 0.7726847529411316, | |
| "learning_rate": 0.00019114403279364524, | |
| "loss": 1.2344, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 0.4242055459659493, | |
| "grad_norm": 0.7513970136642456, | |
| "learning_rate": 0.00019068445848467088, | |
| "loss": 1.2465, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 0.42519900157711077, | |
| "grad_norm": 0.8085342645645142, | |
| "learning_rate": 0.00019022447157140667, | |
| "loss": 1.2225, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 0.42619245718827226, | |
| "grad_norm": 0.7322817444801331, | |
| "learning_rate": 0.00018976407671884217, | |
| "loss": 1.237, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 0.42718591279943374, | |
| "grad_norm": 0.6895975470542908, | |
| "learning_rate": 0.000189303278596104, | |
| "loss": 1.2241, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.42817936841059523, | |
| "grad_norm": 0.8121458292007446, | |
| "learning_rate": 0.00018884208187640875, | |
| "loss": 1.2669, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 0.42917282402175666, | |
| "grad_norm": 0.7596007585525513, | |
| "learning_rate": 0.00018838049123701524, | |
| "loss": 1.2455, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.43016627963291815, | |
| "grad_norm": 0.707015335559845, | |
| "learning_rate": 0.00018791851135917739, | |
| "loss": 1.249, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 0.43115973524407963, | |
| "grad_norm": 0.709096372127533, | |
| "learning_rate": 0.00018745614692809644, | |
| "loss": 1.2474, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 0.4321531908552411, | |
| "grad_norm": 0.7107598781585693, | |
| "learning_rate": 0.0001869934026328738, | |
| "loss": 1.2382, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.43314664646640255, | |
| "grad_norm": 0.7625442147254944, | |
| "learning_rate": 0.00018653028316646317, | |
| "loss": 1.2663, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 0.43414010207756404, | |
| "grad_norm": 0.6910580396652222, | |
| "learning_rate": 0.00018606679322562307, | |
| "loss": 1.2244, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 0.4351335576887255, | |
| "grad_norm": 0.7820401787757874, | |
| "learning_rate": 0.00018560293751086927, | |
| "loss": 1.2429, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 0.436127013299887, | |
| "grad_norm": 0.7634170651435852, | |
| "learning_rate": 0.00018513872072642708, | |
| "loss": 1.2342, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 0.43712046891104844, | |
| "grad_norm": 0.8119933605194092, | |
| "learning_rate": 0.00018467414758018358, | |
| "loss": 1.2408, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.43811392452220993, | |
| "grad_norm": 0.6820570230484009, | |
| "learning_rate": 0.0001842092227836398, | |
| "loss": 1.2449, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 0.4391073801333714, | |
| "grad_norm": 0.7118420004844666, | |
| "learning_rate": 0.00018374395105186342, | |
| "loss": 1.2712, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 0.4401008357445329, | |
| "grad_norm": 0.7938652634620667, | |
| "learning_rate": 0.00018327833710344012, | |
| "loss": 1.2425, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 0.4410942913556944, | |
| "grad_norm": 0.8038454651832581, | |
| "learning_rate": 0.00018281238566042655, | |
| "loss": 1.2493, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 0.4420877469668558, | |
| "grad_norm": 0.7482326626777649, | |
| "learning_rate": 0.00018234610144830189, | |
| "loss": 1.2453, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.4430812025780173, | |
| "grad_norm": 0.7408166527748108, | |
| "learning_rate": 0.00018187948919592023, | |
| "loss": 1.2317, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 0.4440746581891788, | |
| "grad_norm": 0.8955593705177307, | |
| "learning_rate": 0.00018141255363546244, | |
| "loss": 1.2397, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 0.4450681138003403, | |
| "grad_norm": 0.7213795781135559, | |
| "learning_rate": 0.00018094529950238829, | |
| "loss": 1.2232, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.4460615694115017, | |
| "grad_norm": 0.7478641271591187, | |
| "learning_rate": 0.00018047773153538835, | |
| "loss": 1.2582, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 0.4470550250226632, | |
| "grad_norm": 0.743283212184906, | |
| "learning_rate": 0.00018000985447633592, | |
| "loss": 1.2376, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.4480484806338247, | |
| "grad_norm": 0.7658199071884155, | |
| "learning_rate": 0.00017954167307023902, | |
| "loss": 1.2414, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 0.4490419362449862, | |
| "grad_norm": 0.8095099329948425, | |
| "learning_rate": 0.00017907319206519228, | |
| "loss": 1.2323, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.4500353918561476, | |
| "grad_norm": 0.7070140242576599, | |
| "learning_rate": 0.00017860441621232864, | |
| "loss": 1.2507, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 0.4510288474673091, | |
| "grad_norm": 0.8216035962104797, | |
| "learning_rate": 0.00017813535026577128, | |
| "loss": 1.2358, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 0.4520223030784706, | |
| "grad_norm": 0.6905801296234131, | |
| "learning_rate": 0.0001776659989825854, | |
| "loss": 1.2522, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.45301575868963206, | |
| "grad_norm": 0.6887661218643188, | |
| "learning_rate": 0.00017719636712273005, | |
| "loss": 1.2347, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.4540092143007935, | |
| "grad_norm": 0.7682887315750122, | |
| "learning_rate": 0.0001767264594490095, | |
| "loss": 1.2367, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 0.455002669911955, | |
| "grad_norm": 0.6445176005363464, | |
| "learning_rate": 0.00017625628072702562, | |
| "loss": 1.2195, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 0.45599612552311647, | |
| "grad_norm": 0.6851564049720764, | |
| "learning_rate": 0.00017578583572512877, | |
| "loss": 1.2392, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 0.45698958113427796, | |
| "grad_norm": 0.6970175504684448, | |
| "learning_rate": 0.00017531512921436997, | |
| "loss": 1.2491, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.45798303674543944, | |
| "grad_norm": 0.7864800095558167, | |
| "learning_rate": 0.00017484416596845226, | |
| "loss": 1.2431, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 0.4589764923566009, | |
| "grad_norm": 0.7439005374908447, | |
| "learning_rate": 0.00017437295076368252, | |
| "loss": 1.2196, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 0.45996994796776236, | |
| "grad_norm": 0.6971604228019714, | |
| "learning_rate": 0.0001739014883789228, | |
| "loss": 1.2214, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 0.46096340357892385, | |
| "grad_norm": 0.7208828330039978, | |
| "learning_rate": 0.00017342978359554187, | |
| "loss": 1.2412, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.46195685919008533, | |
| "grad_norm": 0.8010265231132507, | |
| "learning_rate": 0.00017295784119736698, | |
| "loss": 1.2535, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.46295031480124677, | |
| "grad_norm": 0.8064112663269043, | |
| "learning_rate": 0.000172485665970635, | |
| "loss": 1.2271, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 0.46394377041240825, | |
| "grad_norm": 0.7055312395095825, | |
| "learning_rate": 0.00017201326270394418, | |
| "loss": 1.2423, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 0.46493722602356974, | |
| "grad_norm": 0.7537856698036194, | |
| "learning_rate": 0.00017154063618820533, | |
| "loss": 1.2431, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 0.4659306816347312, | |
| "grad_norm": 0.7738401889801025, | |
| "learning_rate": 0.00017106779121659344, | |
| "loss": 1.2398, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 0.46692413724589266, | |
| "grad_norm": 0.7333301901817322, | |
| "learning_rate": 0.00017059473258449894, | |
| "loss": 1.2322, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.46791759285705414, | |
| "grad_norm": 0.7468922138214111, | |
| "learning_rate": 0.00017012146508947916, | |
| "loss": 1.2501, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 0.46891104846821563, | |
| "grad_norm": 0.784343421459198, | |
| "learning_rate": 0.00016964799353120963, | |
| "loss": 1.2211, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 0.4699045040793771, | |
| "grad_norm": 0.6925473809242249, | |
| "learning_rate": 0.00016917432271143534, | |
| "loss": 1.2485, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 0.4708979596905386, | |
| "grad_norm": 0.790807843208313, | |
| "learning_rate": 0.00016870045743392207, | |
| "loss": 1.2378, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 0.47189141530170003, | |
| "grad_norm": 0.7804043889045715, | |
| "learning_rate": 0.0001682264025044079, | |
| "loss": 1.201, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.4728848709128615, | |
| "grad_norm": 0.7349110245704651, | |
| "learning_rate": 0.00016775216273055406, | |
| "loss": 1.2291, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 0.473878326524023, | |
| "grad_norm": 0.7869210839271545, | |
| "learning_rate": 0.00016727774292189657, | |
| "loss": 1.2158, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 0.4748717821351845, | |
| "grad_norm": 0.801625669002533, | |
| "learning_rate": 0.0001668031478897971, | |
| "loss": 1.2345, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 0.4758652377463459, | |
| "grad_norm": 0.7812928557395935, | |
| "learning_rate": 0.00016632838244739454, | |
| "loss": 1.2276, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 0.4768586933575074, | |
| "grad_norm": 0.7344529628753662, | |
| "learning_rate": 0.00016585345140955593, | |
| "loss": 1.2161, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.4778521489686689, | |
| "grad_norm": 0.7317351698875427, | |
| "learning_rate": 0.00016537835959282767, | |
| "loss": 1.229, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 0.4788456045798304, | |
| "grad_norm": 0.7621145248413086, | |
| "learning_rate": 0.00016490311181538672, | |
| "loss": 1.2241, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 0.4798390601909918, | |
| "grad_norm": 0.7385951280593872, | |
| "learning_rate": 0.00016442771289699185, | |
| "loss": 1.2337, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 0.4808325158021533, | |
| "grad_norm": 0.8051323294639587, | |
| "learning_rate": 0.00016395216765893443, | |
| "loss": 1.2416, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 0.4818259714133148, | |
| "grad_norm": 0.7685057520866394, | |
| "learning_rate": 0.00016347648092398997, | |
| "loss": 1.2318, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.4828194270244763, | |
| "grad_norm": 0.8106998205184937, | |
| "learning_rate": 0.00016300065751636878, | |
| "loss": 1.2282, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 0.4838128826356377, | |
| "grad_norm": 0.7715012431144714, | |
| "learning_rate": 0.00016252470226166735, | |
| "loss": 1.2476, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 0.4848063382467992, | |
| "grad_norm": 0.7523658871650696, | |
| "learning_rate": 0.0001620486199868193, | |
| "loss": 1.2429, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 0.4857997938579607, | |
| "grad_norm": 0.7648922801017761, | |
| "learning_rate": 0.0001615724155200464, | |
| "loss": 1.2417, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 0.48679324946912217, | |
| "grad_norm": 0.7386729121208191, | |
| "learning_rate": 0.00016109609369080964, | |
| "loss": 1.2219, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.48778670508028366, | |
| "grad_norm": 0.7744386792182922, | |
| "learning_rate": 0.00016061965932976033, | |
| "loss": 1.2273, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 0.4887801606914451, | |
| "grad_norm": 0.6591721773147583, | |
| "learning_rate": 0.00016014311726869092, | |
| "loss": 1.2331, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 0.4897736163026066, | |
| "grad_norm": 0.739093542098999, | |
| "learning_rate": 0.00015966647234048618, | |
| "loss": 1.2329, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 0.49076707191376806, | |
| "grad_norm": 0.7013362646102905, | |
| "learning_rate": 0.000159189729379074, | |
| "loss": 1.2478, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 0.49176052752492955, | |
| "grad_norm": 0.7066161632537842, | |
| "learning_rate": 0.0001587128932193767, | |
| "loss": 1.2163, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.492753983136091, | |
| "grad_norm": 0.7728776335716248, | |
| "learning_rate": 0.00015823596869726152, | |
| "loss": 1.234, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.49374743874725247, | |
| "grad_norm": 0.788206934928894, | |
| "learning_rate": 0.0001577589606494921, | |
| "loss": 1.2268, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 0.49474089435841395, | |
| "grad_norm": 0.70174640417099, | |
| "learning_rate": 0.0001572818739136789, | |
| "loss": 1.2391, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 0.49573434996957544, | |
| "grad_norm": 0.7398496866226196, | |
| "learning_rate": 0.00015680471332823055, | |
| "loss": 1.2358, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 0.49672780558073687, | |
| "grad_norm": 0.7275456786155701, | |
| "learning_rate": 0.00015632748373230462, | |
| "loss": 1.2275, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.49772126119189836, | |
| "grad_norm": 0.7550251483917236, | |
| "learning_rate": 0.00015585018996575853, | |
| "loss": 1.2158, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 0.49871471680305984, | |
| "grad_norm": 0.7849989533424377, | |
| "learning_rate": 0.00015537283686910047, | |
| "loss": 1.2188, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 0.49970817241422133, | |
| "grad_norm": 0.696049153804779, | |
| "learning_rate": 0.00015489542928344037, | |
| "loss": 1.2311, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 0.500204900219802, | |
| "eval_loss": 1.2343899011611938, | |
| "eval_runtime": 69.1827, | |
| "eval_samples_per_second": 722.724, | |
| "eval_steps_per_second": 90.34, | |
| "step": 5035 | |
| }, | |
| { | |
| "epoch": 0.5007016280253828, | |
| "grad_norm": 0.8199466466903687, | |
| "learning_rate": 0.00015441797205044083, | |
| "loss": 1.2398, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 0.5016950836365442, | |
| "grad_norm": 0.7249656319618225, | |
| "learning_rate": 0.00015394047001226775, | |
| "loss": 1.2378, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.5026885392477057, | |
| "grad_norm": 0.7430068850517273, | |
| "learning_rate": 0.0001534629280115416, | |
| "loss": 1.2324, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 0.5036819948588672, | |
| "grad_norm": 0.8324121236801147, | |
| "learning_rate": 0.00015298535089128813, | |
| "loss": 1.2263, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 0.5046754504700287, | |
| "grad_norm": 0.7186987996101379, | |
| "learning_rate": 0.00015250774349488916, | |
| "loss": 1.2149, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 0.5056689060811902, | |
| "grad_norm": 0.6761648654937744, | |
| "learning_rate": 0.00015203011066603357, | |
| "loss": 1.232, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 0.5066623616923517, | |
| "grad_norm": 0.7147372961044312, | |
| "learning_rate": 0.0001515524572486683, | |
| "loss": 1.2278, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.5076558173035131, | |
| "grad_norm": 0.7228516340255737, | |
| "learning_rate": 0.00015107478808694887, | |
| "loss": 1.2228, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 0.5086492729146745, | |
| "grad_norm": 0.7766802310943604, | |
| "learning_rate": 0.0001505971080251907, | |
| "loss": 1.2195, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 0.509642728525836, | |
| "grad_norm": 0.7892508506774902, | |
| "learning_rate": 0.00015011942190781954, | |
| "loss": 1.2231, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 0.5106361841369975, | |
| "grad_norm": 0.797983705997467, | |
| "learning_rate": 0.0001496417345793228, | |
| "loss": 1.2225, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 0.511629639748159, | |
| "grad_norm": 0.7285056710243225, | |
| "learning_rate": 0.00014916405088419998, | |
| "loss": 1.2389, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.5126230953593205, | |
| "grad_norm": 0.7284179329872131, | |
| "learning_rate": 0.0001486863756669137, | |
| "loss": 1.2162, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 0.513616550970482, | |
| "grad_norm": 0.7014544606208801, | |
| "learning_rate": 0.00014820871377184075, | |
| "loss": 1.2271, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 0.5146100065816435, | |
| "grad_norm": 0.7812590003013611, | |
| "learning_rate": 0.00014773107004322278, | |
| "loss": 1.2154, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 0.515603462192805, | |
| "grad_norm": 0.7643766403198242, | |
| "learning_rate": 0.0001472534493251171, | |
| "loss": 1.2182, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 0.5165969178039663, | |
| "grad_norm": 0.7680448293685913, | |
| "learning_rate": 0.00014677585646134778, | |
| "loss": 1.2251, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.5175903734151278, | |
| "grad_norm": 0.7179694771766663, | |
| "learning_rate": 0.00014629829629545632, | |
| "loss": 1.2028, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 0.5185838290262893, | |
| "grad_norm": 0.7599826455116272, | |
| "learning_rate": 0.00014582077367065278, | |
| "loss": 1.2199, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 0.5195772846374508, | |
| "grad_norm": 0.7526030540466309, | |
| "learning_rate": 0.00014534329342976612, | |
| "loss": 1.226, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 0.5205707402486123, | |
| "grad_norm": 0.7358463406562805, | |
| "learning_rate": 0.00014486586041519587, | |
| "loss": 1.2333, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 0.5215641958597738, | |
| "grad_norm": 0.8013938069343567, | |
| "learning_rate": 0.00014438847946886236, | |
| "loss": 1.2465, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.5225576514709352, | |
| "grad_norm": 0.7533153295516968, | |
| "learning_rate": 0.00014391115543215801, | |
| "loss": 1.2028, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 0.5235511070820967, | |
| "grad_norm": 0.7143856883049011, | |
| "learning_rate": 0.00014343389314589797, | |
| "loss": 1.2194, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 0.5245445626932581, | |
| "grad_norm": 0.7963373064994812, | |
| "learning_rate": 0.00014295669745027125, | |
| "loss": 1.2156, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.5255380183044196, | |
| "grad_norm": 0.7041774988174438, | |
| "learning_rate": 0.00014247957318479148, | |
| "loss": 1.2184, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 0.5265314739155811, | |
| "grad_norm": 0.7579498887062073, | |
| "learning_rate": 0.0001420025251882478, | |
| "loss": 1.2328, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.5275249295267426, | |
| "grad_norm": 0.753058135509491, | |
| "learning_rate": 0.00014152555829865593, | |
| "loss": 1.2161, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 0.5285183851379041, | |
| "grad_norm": 0.7893427610397339, | |
| "learning_rate": 0.00014104867735320912, | |
| "loss": 1.2248, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 0.5295118407490655, | |
| "grad_norm": 0.7220308184623718, | |
| "learning_rate": 0.0001405718871882289, | |
| "loss": 1.2059, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 0.530505296360227, | |
| "grad_norm": 0.735222578048706, | |
| "learning_rate": 0.00014009519263911616, | |
| "loss": 1.2205, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 0.5314987519713885, | |
| "grad_norm": 0.7031397223472595, | |
| "learning_rate": 0.00013961859854030208, | |
| "loss": 1.227, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.53249220758255, | |
| "grad_norm": 0.7258872985839844, | |
| "learning_rate": 0.00013914210972519928, | |
| "loss": 1.2349, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 0.5334856631937114, | |
| "grad_norm": 0.7001321315765381, | |
| "learning_rate": 0.00013866573102615237, | |
| "loss": 1.2241, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 0.5344791188048729, | |
| "grad_norm": 0.7036392092704773, | |
| "learning_rate": 0.0001381894672743894, | |
| "loss": 1.2235, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 0.5354725744160344, | |
| "grad_norm": 0.7783767580986023, | |
| "learning_rate": 0.0001377133232999726, | |
| "loss": 1.2119, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 0.5364660300271958, | |
| "grad_norm": 0.7359232902526855, | |
| "learning_rate": 0.0001372373039317495, | |
| "loss": 1.2234, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.5374594856383573, | |
| "grad_norm": 0.6736981868743896, | |
| "learning_rate": 0.0001367614139973039, | |
| "loss": 1.2162, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 0.5384529412495188, | |
| "grad_norm": 0.828008234500885, | |
| "learning_rate": 0.000136285658322907, | |
| "loss": 1.2267, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 0.5394463968606803, | |
| "grad_norm": 0.7058656215667725, | |
| "learning_rate": 0.00013581004173346834, | |
| "loss": 1.2217, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 0.5404398524718418, | |
| "grad_norm": 0.6887176036834717, | |
| "learning_rate": 0.00013533456905248683, | |
| "loss": 1.2133, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 0.5414333080830032, | |
| "grad_norm": 0.8097847104072571, | |
| "learning_rate": 0.00013485924510200215, | |
| "loss": 1.2228, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.5424267636941646, | |
| "grad_norm": 0.732067883014679, | |
| "learning_rate": 0.00013438407470254543, | |
| "loss": 1.214, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 0.5434202193053261, | |
| "grad_norm": 0.7297837138175964, | |
| "learning_rate": 0.0001339090626730906, | |
| "loss": 1.2191, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 0.5444136749164876, | |
| "grad_norm": 0.7470762133598328, | |
| "learning_rate": 0.00013343421383100554, | |
| "loss": 1.2028, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 0.5454071305276491, | |
| "grad_norm": 0.6716703772544861, | |
| "learning_rate": 0.000132959532992003, | |
| "loss": 1.2084, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 0.5464005861388106, | |
| "grad_norm": 0.7758064866065979, | |
| "learning_rate": 0.0001324850249700921, | |
| "loss": 1.2238, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.5473940417499721, | |
| "grad_norm": 0.7496920228004456, | |
| "learning_rate": 0.0001320106945775292, | |
| "loss": 1.2171, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 0.5483874973611336, | |
| "grad_norm": 0.7656776309013367, | |
| "learning_rate": 0.0001315365466247692, | |
| "loss": 1.1829, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 0.5493809529722951, | |
| "grad_norm": 0.7667086720466614, | |
| "learning_rate": 0.0001310625859204169, | |
| "loss": 1.2266, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 0.5503744085834564, | |
| "grad_norm": 0.7529829740524292, | |
| "learning_rate": 0.00013058881727117798, | |
| "loss": 1.2135, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 0.5513678641946179, | |
| "grad_norm": 0.7866660952568054, | |
| "learning_rate": 0.00013011524548181048, | |
| "loss": 1.2245, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.5523613198057794, | |
| "grad_norm": 0.7378416657447815, | |
| "learning_rate": 0.00012964187535507594, | |
| "loss": 1.1949, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 0.5533547754169409, | |
| "grad_norm": 0.6917827129364014, | |
| "learning_rate": 0.00012916871169169061, | |
| "loss": 1.2217, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 0.5543482310281024, | |
| "grad_norm": 0.7293698191642761, | |
| "learning_rate": 0.00012869575929027704, | |
| "loss": 1.2143, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 0.5553416866392639, | |
| "grad_norm": 0.7007254958152771, | |
| "learning_rate": 0.0001282230229473152, | |
| "loss": 1.226, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 0.5563351422504254, | |
| "grad_norm": 0.6827494502067566, | |
| "learning_rate": 0.00012775050745709383, | |
| "loss": 1.2163, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.5573285978615868, | |
| "grad_norm": 0.7645403146743774, | |
| "learning_rate": 0.00012727821761166196, | |
| "loss": 1.2138, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 0.5583220534727483, | |
| "grad_norm": 0.8425547480583191, | |
| "learning_rate": 0.00012680615820078017, | |
| "loss": 1.2146, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 0.5593155090839097, | |
| "grad_norm": 0.6900782585144043, | |
| "learning_rate": 0.0001263343340118722, | |
| "loss": 1.2088, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 0.5603089646950712, | |
| "grad_norm": 0.7670438289642334, | |
| "learning_rate": 0.00012586274982997596, | |
| "loss": 1.2275, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 0.5613024203062327, | |
| "grad_norm": 0.6959342956542969, | |
| "learning_rate": 0.00012539141043769566, | |
| "loss": 1.2153, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.5622958759173942, | |
| "grad_norm": 0.7077504992485046, | |
| "learning_rate": 0.00012492032061515274, | |
| "loss": 1.1948, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 0.5632893315285556, | |
| "grad_norm": 0.7480520009994507, | |
| "learning_rate": 0.00012444948513993774, | |
| "loss": 1.2114, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 0.5642827871397171, | |
| "grad_norm": 0.7556664943695068, | |
| "learning_rate": 0.00012397890878706162, | |
| "loss": 1.2154, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 0.5652762427508786, | |
| "grad_norm": 0.8014240860939026, | |
| "learning_rate": 0.00012350859632890752, | |
| "loss": 1.2225, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 0.5662696983620401, | |
| "grad_norm": 0.7471588850021362, | |
| "learning_rate": 0.00012303855253518222, | |
| "loss": 1.1975, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.5672631539732015, | |
| "grad_norm": 0.7830742597579956, | |
| "learning_rate": 0.00012256878217286774, | |
| "loss": 1.2124, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 0.568256609584363, | |
| "grad_norm": 0.7449409365653992, | |
| "learning_rate": 0.0001220992900061733, | |
| "loss": 1.2226, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 0.5692500651955245, | |
| "grad_norm": 0.7470856308937073, | |
| "learning_rate": 0.00012163008079648655, | |
| "loss": 1.2201, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 0.5702435208066859, | |
| "grad_norm": 0.7292488813400269, | |
| "learning_rate": 0.0001211611593023256, | |
| "loss": 1.2043, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 0.5712369764178474, | |
| "grad_norm": 0.7818967700004578, | |
| "learning_rate": 0.00012069253027929076, | |
| "loss": 1.1942, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.5722304320290089, | |
| "grad_norm": 0.7638885378837585, | |
| "learning_rate": 0.00012022419848001607, | |
| "loss": 1.209, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.5732238876401704, | |
| "grad_norm": 0.7104228138923645, | |
| "learning_rate": 0.0001197561686541214, | |
| "loss": 1.1939, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 0.5742173432513319, | |
| "grad_norm": 0.7472487688064575, | |
| "learning_rate": 0.00011928844554816401, | |
| "loss": 1.2187, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 0.5752107988624934, | |
| "grad_norm": 0.7629554867744446, | |
| "learning_rate": 0.00011882103390559055, | |
| "loss": 1.2106, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 0.5762042544736548, | |
| "grad_norm": 0.6695338487625122, | |
| "learning_rate": 0.00011835393846668907, | |
| "loss": 1.2111, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.5771977100848162, | |
| "grad_norm": 0.7315328121185303, | |
| "learning_rate": 0.00011788716396854067, | |
| "loss": 1.2248, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 0.5781911656959777, | |
| "grad_norm": 0.7161918878555298, | |
| "learning_rate": 0.00011742071514497172, | |
| "loss": 1.215, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 0.5791846213071392, | |
| "grad_norm": 0.7948908805847168, | |
| "learning_rate": 0.00011695459672650566, | |
| "loss": 1.2248, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 0.5801780769183007, | |
| "grad_norm": 0.8110846281051636, | |
| "learning_rate": 0.00011648881344031516, | |
| "loss": 1.2163, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 0.5811715325294622, | |
| "grad_norm": 0.8162423372268677, | |
| "learning_rate": 0.00011602337001017399, | |
| "loss": 1.2172, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.5821649881406237, | |
| "grad_norm": 0.8222430348396301, | |
| "learning_rate": 0.00011555827115640947, | |
| "loss": 1.2058, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 0.5831584437517852, | |
| "grad_norm": 0.7712266445159912, | |
| "learning_rate": 0.00011509352159585417, | |
| "loss": 1.1862, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 0.5841518993629465, | |
| "grad_norm": 0.7557414770126343, | |
| "learning_rate": 0.00011462912604179843, | |
| "loss": 1.2103, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 0.585145354974108, | |
| "grad_norm": 0.7453897595405579, | |
| "learning_rate": 0.00011416508920394228, | |
| "loss": 1.2, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 0.5861388105852695, | |
| "grad_norm": 0.7189779281616211, | |
| "learning_rate": 0.00011370141578834791, | |
| "loss": 1.193, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.587132266196431, | |
| "grad_norm": 0.7440248131752014, | |
| "learning_rate": 0.00011323811049739184, | |
| "loss": 1.2091, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 0.5881257218075925, | |
| "grad_norm": 0.7051219940185547, | |
| "learning_rate": 0.00011277517802971704, | |
| "loss": 1.202, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 0.589119177418754, | |
| "grad_norm": 0.7295770049095154, | |
| "learning_rate": 0.00011231262308018568, | |
| "loss": 1.2049, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 0.5901126330299155, | |
| "grad_norm": 0.8079803586006165, | |
| "learning_rate": 0.00011185045033983126, | |
| "loss": 1.2156, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 0.5911060886410769, | |
| "grad_norm": 0.768571138381958, | |
| "learning_rate": 0.00011138866449581098, | |
| "loss": 1.2072, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 0.5920995442522384, | |
| "grad_norm": 0.7064054608345032, | |
| "learning_rate": 0.00011092727023135847, | |
| "loss": 1.2074, | |
| "step": 5960 | |
| }, | |
| { | |
| "epoch": 0.5930929998633998, | |
| "grad_norm": 0.7940462231636047, | |
| "learning_rate": 0.00011046627222573598, | |
| "loss": 1.1943, | |
| "step": 5970 | |
| }, | |
| { | |
| "epoch": 0.5940864554745613, | |
| "grad_norm": 0.7651082873344421, | |
| "learning_rate": 0.00011000567515418709, | |
| "loss": 1.2021, | |
| "step": 5980 | |
| }, | |
| { | |
| "epoch": 0.5950799110857228, | |
| "grad_norm": 0.8014243841171265, | |
| "learning_rate": 0.00010954548368788932, | |
| "loss": 1.1826, | |
| "step": 5990 | |
| }, | |
| { | |
| "epoch": 0.5960733666968843, | |
| "grad_norm": 0.7298262119293213, | |
| "learning_rate": 0.00010908570249390679, | |
| "loss": 1.1855, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.5970668223080458, | |
| "grad_norm": 0.7572104334831238, | |
| "learning_rate": 0.0001086263362351426, | |
| "loss": 1.2054, | |
| "step": 6010 | |
| }, | |
| { | |
| "epoch": 0.5980602779192072, | |
| "grad_norm": 0.8020755052566528, | |
| "learning_rate": 0.00010816738957029202, | |
| "loss": 1.2148, | |
| "step": 6020 | |
| }, | |
| { | |
| "epoch": 0.5990537335303687, | |
| "grad_norm": 0.7517848014831543, | |
| "learning_rate": 0.00010770886715379475, | |
| "loss": 1.1962, | |
| "step": 6030 | |
| }, | |
| { | |
| "epoch": 0.6000471891415302, | |
| "grad_norm": 0.7051617503166199, | |
| "learning_rate": 0.00010725077363578817, | |
| "loss": 1.1908, | |
| "step": 6040 | |
| }, | |
| { | |
| "epoch": 0.6002458802637625, | |
| "eval_loss": 1.2166225910186768, | |
| "eval_runtime": 69.3414, | |
| "eval_samples_per_second": 721.07, | |
| "eval_steps_per_second": 90.134, | |
| "step": 6042 | |
| }, | |
| { | |
| "epoch": 0.6010406447526916, | |
| "grad_norm": 0.7492873072624207, | |
| "learning_rate": 0.00010679311366205971, | |
| "loss": 1.216, | |
| "step": 6050 | |
| }, | |
| { | |
| "epoch": 0.6020341003638531, | |
| "grad_norm": 0.7082617282867432, | |
| "learning_rate": 0.00010633589187400009, | |
| "loss": 1.2065, | |
| "step": 6060 | |
| }, | |
| { | |
| "epoch": 0.6030275559750146, | |
| "grad_norm": 0.7885176539421082, | |
| "learning_rate": 0.00010587911290855625, | |
| "loss": 1.2004, | |
| "step": 6070 | |
| }, | |
| { | |
| "epoch": 0.604021011586176, | |
| "grad_norm": 0.802290141582489, | |
| "learning_rate": 0.00010542278139818404, | |
| "loss": 1.1995, | |
| "step": 6080 | |
| }, | |
| { | |
| "epoch": 0.6050144671973375, | |
| "grad_norm": 0.7638839483261108, | |
| "learning_rate": 0.00010496690197080147, | |
| "loss": 1.2111, | |
| "step": 6090 | |
| }, | |
| { | |
| "epoch": 0.606007922808499, | |
| "grad_norm": 0.7744988203048706, | |
| "learning_rate": 0.00010451147924974182, | |
| "loss": 1.1956, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.6070013784196605, | |
| "grad_norm": 0.7348924875259399, | |
| "learning_rate": 0.00010405651785370654, | |
| "loss": 1.1958, | |
| "step": 6110 | |
| }, | |
| { | |
| "epoch": 0.607994834030822, | |
| "grad_norm": 0.7749565839767456, | |
| "learning_rate": 0.00010360202239671843, | |
| "loss": 1.1998, | |
| "step": 6120 | |
| }, | |
| { | |
| "epoch": 0.6089882896419835, | |
| "grad_norm": 0.6604474186897278, | |
| "learning_rate": 0.0001031479974880752, | |
| "loss": 1.2076, | |
| "step": 6130 | |
| }, | |
| { | |
| "epoch": 0.6099817452531449, | |
| "grad_norm": 0.6728931069374084, | |
| "learning_rate": 0.0001026944477323022, | |
| "loss": 1.203, | |
| "step": 6140 | |
| }, | |
| { | |
| "epoch": 0.6109752008643063, | |
| "grad_norm": 0.774555504322052, | |
| "learning_rate": 0.00010224137772910618, | |
| "loss": 1.1991, | |
| "step": 6150 | |
| }, | |
| { | |
| "epoch": 0.6119686564754678, | |
| "grad_norm": 0.7287729382514954, | |
| "learning_rate": 0.00010178879207332834, | |
| "loss": 1.1934, | |
| "step": 6160 | |
| }, | |
| { | |
| "epoch": 0.6129621120866293, | |
| "grad_norm": 0.7626068592071533, | |
| "learning_rate": 0.0001013366953548978, | |
| "loss": 1.2079, | |
| "step": 6170 | |
| }, | |
| { | |
| "epoch": 0.6139555676977908, | |
| "grad_norm": 0.7775630950927734, | |
| "learning_rate": 0.00010088509215878528, | |
| "loss": 1.201, | |
| "step": 6180 | |
| }, | |
| { | |
| "epoch": 0.6149490233089523, | |
| "grad_norm": 0.7765801548957825, | |
| "learning_rate": 0.00010043398706495613, | |
| "loss": 1.203, | |
| "step": 6190 | |
| }, | |
| { | |
| "epoch": 0.6159424789201138, | |
| "grad_norm": 0.8291947245597839, | |
| "learning_rate": 9.998338464832434e-05, | |
| "loss": 1.2218, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.6169359345312753, | |
| "grad_norm": 0.7486442923545837, | |
| "learning_rate": 9.953328947870598e-05, | |
| "loss": 1.1889, | |
| "step": 6210 | |
| }, | |
| { | |
| "epoch": 0.6179293901424368, | |
| "grad_norm": 0.7109442949295044, | |
| "learning_rate": 9.908370612077272e-05, | |
| "loss": 1.1958, | |
| "step": 6220 | |
| }, | |
| { | |
| "epoch": 0.6189228457535981, | |
| "grad_norm": 0.8005870580673218, | |
| "learning_rate": 9.863463913400582e-05, | |
| "loss": 1.1805, | |
| "step": 6230 | |
| }, | |
| { | |
| "epoch": 0.6199163013647596, | |
| "grad_norm": 0.7783021926879883, | |
| "learning_rate": 9.818609307264953e-05, | |
| "loss": 1.1875, | |
| "step": 6240 | |
| }, | |
| { | |
| "epoch": 0.6209097569759211, | |
| "grad_norm": 0.737714946269989, | |
| "learning_rate": 9.773807248566534e-05, | |
| "loss": 1.2003, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.6219032125870826, | |
| "grad_norm": 0.7947556376457214, | |
| "learning_rate": 9.729058191668527e-05, | |
| "loss": 1.1961, | |
| "step": 6260 | |
| }, | |
| { | |
| "epoch": 0.6228966681982441, | |
| "grad_norm": 0.7534834742546082, | |
| "learning_rate": 9.684362590396643e-05, | |
| "loss": 1.2063, | |
| "step": 6270 | |
| }, | |
| { | |
| "epoch": 0.6238901238094056, | |
| "grad_norm": 0.7267855405807495, | |
| "learning_rate": 9.639720898034454e-05, | |
| "loss": 1.2206, | |
| "step": 6280 | |
| }, | |
| { | |
| "epoch": 0.624883579420567, | |
| "grad_norm": 0.6868566870689392, | |
| "learning_rate": 9.59513356731882e-05, | |
| "loss": 1.1868, | |
| "step": 6290 | |
| }, | |
| { | |
| "epoch": 0.6258770350317285, | |
| "grad_norm": 0.7699499130249023, | |
| "learning_rate": 9.550601050435279e-05, | |
| "loss": 1.201, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.6268704906428899, | |
| "grad_norm": 0.7472879886627197, | |
| "learning_rate": 9.506123799013476e-05, | |
| "loss": 1.1853, | |
| "step": 6310 | |
| }, | |
| { | |
| "epoch": 0.6278639462540514, | |
| "grad_norm": 0.7912560105323792, | |
| "learning_rate": 9.461702264122584e-05, | |
| "loss": 1.1877, | |
| "step": 6320 | |
| }, | |
| { | |
| "epoch": 0.6288574018652129, | |
| "grad_norm": 0.7609166502952576, | |
| "learning_rate": 9.417336896266701e-05, | |
| "loss": 1.194, | |
| "step": 6330 | |
| }, | |
| { | |
| "epoch": 0.6298508574763744, | |
| "grad_norm": 0.781128466129303, | |
| "learning_rate": 9.373028145380326e-05, | |
| "loss": 1.2085, | |
| "step": 6340 | |
| }, | |
| { | |
| "epoch": 0.6308443130875359, | |
| "grad_norm": 0.7291694283485413, | |
| "learning_rate": 9.328776460823764e-05, | |
| "loss": 1.1957, | |
| "step": 6350 | |
| }, | |
| { | |
| "epoch": 0.6318377686986973, | |
| "grad_norm": 0.7578558921813965, | |
| "learning_rate": 9.284582291378573e-05, | |
| "loss": 1.2025, | |
| "step": 6360 | |
| }, | |
| { | |
| "epoch": 0.6328312243098588, | |
| "grad_norm": 0.7515532970428467, | |
| "learning_rate": 9.240446085243033e-05, | |
| "loss": 1.207, | |
| "step": 6370 | |
| }, | |
| { | |
| "epoch": 0.6338246799210203, | |
| "grad_norm": 0.7200992107391357, | |
| "learning_rate": 9.196368290027572e-05, | |
| "loss": 1.1909, | |
| "step": 6380 | |
| }, | |
| { | |
| "epoch": 0.6348181355321818, | |
| "grad_norm": 0.7989151477813721, | |
| "learning_rate": 9.152349352750232e-05, | |
| "loss": 1.2204, | |
| "step": 6390 | |
| }, | |
| { | |
| "epoch": 0.6358115911433432, | |
| "grad_norm": 0.724198043346405, | |
| "learning_rate": 9.108389719832165e-05, | |
| "loss": 1.1959, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.6368050467545047, | |
| "grad_norm": 0.8064570426940918, | |
| "learning_rate": 9.064489837093069e-05, | |
| "loss": 1.1962, | |
| "step": 6410 | |
| }, | |
| { | |
| "epoch": 0.6377985023656662, | |
| "grad_norm": 0.8495492339134216, | |
| "learning_rate": 9.020650149746679e-05, | |
| "loss": 1.1937, | |
| "step": 6420 | |
| }, | |
| { | |
| "epoch": 0.6387919579768276, | |
| "grad_norm": 0.7575849890708923, | |
| "learning_rate": 8.976871102396269e-05, | |
| "loss": 1.2105, | |
| "step": 6430 | |
| }, | |
| { | |
| "epoch": 0.6397854135879891, | |
| "grad_norm": 0.7532674074172974, | |
| "learning_rate": 8.93315313903011e-05, | |
| "loss": 1.1936, | |
| "step": 6440 | |
| }, | |
| { | |
| "epoch": 0.6407788691991506, | |
| "grad_norm": 0.8248761892318726, | |
| "learning_rate": 8.889496703017006e-05, | |
| "loss": 1.2107, | |
| "step": 6450 | |
| }, | |
| { | |
| "epoch": 0.6417723248103121, | |
| "grad_norm": 0.7390906810760498, | |
| "learning_rate": 8.845902237101754e-05, | |
| "loss": 1.197, | |
| "step": 6460 | |
| }, | |
| { | |
| "epoch": 0.6427657804214736, | |
| "grad_norm": 0.7399100661277771, | |
| "learning_rate": 8.80237018340069e-05, | |
| "loss": 1.2062, | |
| "step": 6470 | |
| }, | |
| { | |
| "epoch": 0.643759236032635, | |
| "grad_norm": 0.8141552209854126, | |
| "learning_rate": 8.758900983397195e-05, | |
| "loss": 1.199, | |
| "step": 6480 | |
| }, | |
| { | |
| "epoch": 0.6447526916437964, | |
| "grad_norm": 0.7984867095947266, | |
| "learning_rate": 8.71549507793721e-05, | |
| "loss": 1.191, | |
| "step": 6490 | |
| }, | |
| { | |
| "epoch": 0.6457461472549579, | |
| "grad_norm": 0.7291640043258667, | |
| "learning_rate": 8.672152907224765e-05, | |
| "loss": 1.1994, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.6467396028661194, | |
| "grad_norm": 0.7574365735054016, | |
| "learning_rate": 8.628874910817528e-05, | |
| "loss": 1.1906, | |
| "step": 6510 | |
| }, | |
| { | |
| "epoch": 0.6477330584772809, | |
| "grad_norm": 0.8019115924835205, | |
| "learning_rate": 8.585661527622343e-05, | |
| "loss": 1.2178, | |
| "step": 6520 | |
| }, | |
| { | |
| "epoch": 0.6487265140884424, | |
| "grad_norm": 0.7128658890724182, | |
| "learning_rate": 8.542513195890758e-05, | |
| "loss": 1.2077, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 0.6497199696996039, | |
| "grad_norm": 0.7478747963905334, | |
| "learning_rate": 8.499430353214605e-05, | |
| "loss": 1.2125, | |
| "step": 6540 | |
| }, | |
| { | |
| "epoch": 0.6507134253107654, | |
| "grad_norm": 0.7416719794273376, | |
| "learning_rate": 8.456413436521558e-05, | |
| "loss": 1.2043, | |
| "step": 6550 | |
| }, | |
| { | |
| "epoch": 0.6517068809219269, | |
| "grad_norm": 0.827017068862915, | |
| "learning_rate": 8.413462882070699e-05, | |
| "loss": 1.1865, | |
| "step": 6560 | |
| }, | |
| { | |
| "epoch": 0.6527003365330882, | |
| "grad_norm": 0.7773194313049316, | |
| "learning_rate": 8.370579125448091e-05, | |
| "loss": 1.1912, | |
| "step": 6570 | |
| }, | |
| { | |
| "epoch": 0.6536937921442497, | |
| "grad_norm": 0.7700802683830261, | |
| "learning_rate": 8.327762601562346e-05, | |
| "loss": 1.2157, | |
| "step": 6580 | |
| }, | |
| { | |
| "epoch": 0.6546872477554112, | |
| "grad_norm": 0.7788603901863098, | |
| "learning_rate": 8.285013744640266e-05, | |
| "loss": 1.1885, | |
| "step": 6590 | |
| }, | |
| { | |
| "epoch": 0.6556807033665727, | |
| "grad_norm": 0.7349107265472412, | |
| "learning_rate": 8.24233298822236e-05, | |
| "loss": 1.1859, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.6566741589777342, | |
| "grad_norm": 0.7807127833366394, | |
| "learning_rate": 8.199720765158533e-05, | |
| "loss": 1.2021, | |
| "step": 6610 | |
| }, | |
| { | |
| "epoch": 0.6576676145888957, | |
| "grad_norm": 0.7519397139549255, | |
| "learning_rate": 8.15717750760362e-05, | |
| "loss": 1.1865, | |
| "step": 6620 | |
| }, | |
| { | |
| "epoch": 0.6586610702000572, | |
| "grad_norm": 0.7665129899978638, | |
| "learning_rate": 8.114703647013056e-05, | |
| "loss": 1.1847, | |
| "step": 6630 | |
| }, | |
| { | |
| "epoch": 0.6596545258112186, | |
| "grad_norm": 0.7215622067451477, | |
| "learning_rate": 8.072299614138482e-05, | |
| "loss": 1.1907, | |
| "step": 6640 | |
| }, | |
| { | |
| "epoch": 0.6606479814223801, | |
| "grad_norm": 0.7784072160720825, | |
| "learning_rate": 8.029965839023373e-05, | |
| "loss": 1.1946, | |
| "step": 6650 | |
| }, | |
| { | |
| "epoch": 0.6616414370335415, | |
| "grad_norm": 0.8263142704963684, | |
| "learning_rate": 7.98770275099868e-05, | |
| "loss": 1.1675, | |
| "step": 6660 | |
| }, | |
| { | |
| "epoch": 0.662634892644703, | |
| "grad_norm": 0.7922942042350769, | |
| "learning_rate": 7.94551077867846e-05, | |
| "loss": 1.1792, | |
| "step": 6670 | |
| }, | |
| { | |
| "epoch": 0.6636283482558645, | |
| "grad_norm": 0.8771615028381348, | |
| "learning_rate": 7.903390349955574e-05, | |
| "loss": 1.2025, | |
| "step": 6680 | |
| }, | |
| { | |
| "epoch": 0.664621803867026, | |
| "grad_norm": 0.7998060584068298, | |
| "learning_rate": 7.861341891997308e-05, | |
| "loss": 1.2092, | |
| "step": 6690 | |
| }, | |
| { | |
| "epoch": 0.6656152594781874, | |
| "grad_norm": 0.7180555462837219, | |
| "learning_rate": 7.819365831241036e-05, | |
| "loss": 1.1892, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.6666087150893489, | |
| "grad_norm": 0.7950117588043213, | |
| "learning_rate": 7.77746259338993e-05, | |
| "loss": 1.1799, | |
| "step": 6710 | |
| }, | |
| { | |
| "epoch": 0.6676021707005104, | |
| "grad_norm": 0.7974181771278381, | |
| "learning_rate": 7.73563260340862e-05, | |
| "loss": 1.1834, | |
| "step": 6720 | |
| }, | |
| { | |
| "epoch": 0.6685956263116719, | |
| "grad_norm": 0.775913655757904, | |
| "learning_rate": 7.693876285518873e-05, | |
| "loss": 1.1979, | |
| "step": 6730 | |
| }, | |
| { | |
| "epoch": 0.6695890819228333, | |
| "grad_norm": 0.8048379421234131, | |
| "learning_rate": 7.652194063195329e-05, | |
| "loss": 1.2006, | |
| "step": 6740 | |
| }, | |
| { | |
| "epoch": 0.6705825375339948, | |
| "grad_norm": 0.7799326777458191, | |
| "learning_rate": 7.61058635916116e-05, | |
| "loss": 1.1849, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.6715759931451563, | |
| "grad_norm": 0.7183064818382263, | |
| "learning_rate": 7.569053595383814e-05, | |
| "loss": 1.1761, | |
| "step": 6760 | |
| }, | |
| { | |
| "epoch": 0.6725694487563177, | |
| "grad_norm": 0.8436702489852905, | |
| "learning_rate": 7.527596193070742e-05, | |
| "loss": 1.1882, | |
| "step": 6770 | |
| }, | |
| { | |
| "epoch": 0.6735629043674792, | |
| "grad_norm": 0.748023271560669, | |
| "learning_rate": 7.486214572665084e-05, | |
| "loss": 1.1869, | |
| "step": 6780 | |
| }, | |
| { | |
| "epoch": 0.6745563599786407, | |
| "grad_norm": 0.7698994874954224, | |
| "learning_rate": 7.444909153841455e-05, | |
| "loss": 1.1877, | |
| "step": 6790 | |
| }, | |
| { | |
| "epoch": 0.6755498155898022, | |
| "grad_norm": 0.7901074886322021, | |
| "learning_rate": 7.40368035550165e-05, | |
| "loss": 1.1727, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.6765432712009637, | |
| "grad_norm": 0.7774555087089539, | |
| "learning_rate": 7.362528595770423e-05, | |
| "loss": 1.2012, | |
| "step": 6810 | |
| }, | |
| { | |
| "epoch": 0.6775367268121252, | |
| "grad_norm": 0.8479429483413696, | |
| "learning_rate": 7.321454291991239e-05, | |
| "loss": 1.2, | |
| "step": 6820 | |
| }, | |
| { | |
| "epoch": 0.6785301824232866, | |
| "grad_norm": 0.8364279270172119, | |
| "learning_rate": 7.28045786072202e-05, | |
| "loss": 1.1875, | |
| "step": 6830 | |
| }, | |
| { | |
| "epoch": 0.679523638034448, | |
| "grad_norm": 0.7133791446685791, | |
| "learning_rate": 7.239539717730956e-05, | |
| "loss": 1.1957, | |
| "step": 6840 | |
| }, | |
| { | |
| "epoch": 0.6805170936456095, | |
| "grad_norm": 0.7974032163619995, | |
| "learning_rate": 7.198700277992268e-05, | |
| "loss": 1.1854, | |
| "step": 6850 | |
| }, | |
| { | |
| "epoch": 0.681510549256771, | |
| "grad_norm": 0.793781578540802, | |
| "learning_rate": 7.157939955681996e-05, | |
| "loss": 1.1982, | |
| "step": 6860 | |
| }, | |
| { | |
| "epoch": 0.6825040048679325, | |
| "grad_norm": 0.7522931694984436, | |
| "learning_rate": 7.117259164173799e-05, | |
| "loss": 1.1891, | |
| "step": 6870 | |
| }, | |
| { | |
| "epoch": 0.683497460479094, | |
| "grad_norm": 0.7739388942718506, | |
| "learning_rate": 7.07665831603479e-05, | |
| "loss": 1.1905, | |
| "step": 6880 | |
| }, | |
| { | |
| "epoch": 0.6844909160902555, | |
| "grad_norm": 0.8487018346786499, | |
| "learning_rate": 7.036137823021311e-05, | |
| "loss": 1.2147, | |
| "step": 6890 | |
| }, | |
| { | |
| "epoch": 0.685484371701417, | |
| "grad_norm": 0.8022236227989197, | |
| "learning_rate": 6.995698096074797e-05, | |
| "loss": 1.1806, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.6864778273125783, | |
| "grad_norm": 0.7998080849647522, | |
| "learning_rate": 6.95533954531756e-05, | |
| "loss": 1.175, | |
| "step": 6910 | |
| }, | |
| { | |
| "epoch": 0.6874712829237398, | |
| "grad_norm": 0.7794836163520813, | |
| "learning_rate": 6.91506258004868e-05, | |
| "loss": 1.1814, | |
| "step": 6920 | |
| }, | |
| { | |
| "epoch": 0.6884647385349013, | |
| "grad_norm": 0.7790430784225464, | |
| "learning_rate": 6.874867608739833e-05, | |
| "loss": 1.1858, | |
| "step": 6930 | |
| }, | |
| { | |
| "epoch": 0.6894581941460628, | |
| "grad_norm": 0.7373054027557373, | |
| "learning_rate": 6.834755039031133e-05, | |
| "loss": 1.2058, | |
| "step": 6940 | |
| }, | |
| { | |
| "epoch": 0.6904516497572243, | |
| "grad_norm": 0.7512431740760803, | |
| "learning_rate": 6.794725277727016e-05, | |
| "loss": 1.19, | |
| "step": 6950 | |
| }, | |
| { | |
| "epoch": 0.6914451053683858, | |
| "grad_norm": 0.7354750633239746, | |
| "learning_rate": 6.75477873079212e-05, | |
| "loss": 1.1872, | |
| "step": 6960 | |
| }, | |
| { | |
| "epoch": 0.6924385609795473, | |
| "grad_norm": 0.7236737012863159, | |
| "learning_rate": 6.714915803347151e-05, | |
| "loss": 1.1782, | |
| "step": 6970 | |
| }, | |
| { | |
| "epoch": 0.6934320165907087, | |
| "grad_norm": 0.8318518996238708, | |
| "learning_rate": 6.675136899664792e-05, | |
| "loss": 1.1899, | |
| "step": 6980 | |
| }, | |
| { | |
| "epoch": 0.6944254722018702, | |
| "grad_norm": 0.7331385016441345, | |
| "learning_rate": 6.635442423165571e-05, | |
| "loss": 1.1939, | |
| "step": 6990 | |
| }, | |
| { | |
| "epoch": 0.6954189278130316, | |
| "grad_norm": 0.7837140560150146, | |
| "learning_rate": 6.595832776413823e-05, | |
| "loss": 1.2028, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.6964123834241931, | |
| "grad_norm": 0.745232343673706, | |
| "learning_rate": 6.55630836111354e-05, | |
| "loss": 1.1782, | |
| "step": 7010 | |
| }, | |
| { | |
| "epoch": 0.6974058390353546, | |
| "grad_norm": 0.8045364618301392, | |
| "learning_rate": 6.516869578104371e-05, | |
| "loss": 1.1775, | |
| "step": 7020 | |
| }, | |
| { | |
| "epoch": 0.6983992946465161, | |
| "grad_norm": 0.82541823387146, | |
| "learning_rate": 6.477516827357486e-05, | |
| "loss": 1.1748, | |
| "step": 7030 | |
| }, | |
| { | |
| "epoch": 0.6993927502576776, | |
| "grad_norm": 0.8108699917793274, | |
| "learning_rate": 6.438250507971574e-05, | |
| "loss": 1.1888, | |
| "step": 7040 | |
| }, | |
| { | |
| "epoch": 0.7002868603077229, | |
| "eval_loss": 1.204665184020996, | |
| "eval_runtime": 69.1809, | |
| "eval_samples_per_second": 722.743, | |
| "eval_steps_per_second": 90.343, | |
| "step": 7049 | |
| }, | |
| { | |
| "epoch": 0.700386205868839, | |
| "grad_norm": 0.7349569201469421, | |
| "learning_rate": 6.399071018168768e-05, | |
| "loss": 1.1968, | |
| "step": 7050 | |
| }, | |
| { | |
| "epoch": 0.7013796614800005, | |
| "grad_norm": 0.755907416343689, | |
| "learning_rate": 6.359978755290616e-05, | |
| "loss": 1.1673, | |
| "step": 7060 | |
| }, | |
| { | |
| "epoch": 0.702373117091162, | |
| "grad_norm": 0.7702360153198242, | |
| "learning_rate": 6.320974115794042e-05, | |
| "loss": 1.178, | |
| "step": 7070 | |
| }, | |
| { | |
| "epoch": 0.7033665727023234, | |
| "grad_norm": 0.8145601153373718, | |
| "learning_rate": 6.282057495247326e-05, | |
| "loss": 1.1819, | |
| "step": 7080 | |
| }, | |
| { | |
| "epoch": 0.7043600283134849, | |
| "grad_norm": 0.8232987523078918, | |
| "learning_rate": 6.24322928832611e-05, | |
| "loss": 1.1861, | |
| "step": 7090 | |
| }, | |
| { | |
| "epoch": 0.7053534839246464, | |
| "grad_norm": 0.8213473558425903, | |
| "learning_rate": 6.204489888809386e-05, | |
| "loss": 1.1794, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.7063469395358078, | |
| "grad_norm": 0.7550660371780396, | |
| "learning_rate": 6.16583968957548e-05, | |
| "loss": 1.1974, | |
| "step": 7110 | |
| }, | |
| { | |
| "epoch": 0.7073403951469693, | |
| "grad_norm": 0.7469960451126099, | |
| "learning_rate": 6.127279082598104e-05, | |
| "loss": 1.1936, | |
| "step": 7120 | |
| }, | |
| { | |
| "epoch": 0.7083338507581308, | |
| "grad_norm": 0.7609216570854187, | |
| "learning_rate": 6.088808458942365e-05, | |
| "loss": 1.1961, | |
| "step": 7130 | |
| }, | |
| { | |
| "epoch": 0.7093273063692923, | |
| "grad_norm": 0.812157154083252, | |
| "learning_rate": 6.050428208760781e-05, | |
| "loss": 1.1913, | |
| "step": 7140 | |
| }, | |
| { | |
| "epoch": 0.7103207619804538, | |
| "grad_norm": 0.7892830967903137, | |
| "learning_rate": 6.012138721289361e-05, | |
| "loss": 1.1733, | |
| "step": 7150 | |
| }, | |
| { | |
| "epoch": 0.7113142175916153, | |
| "grad_norm": 0.8952081203460693, | |
| "learning_rate": 5.97394038484362e-05, | |
| "loss": 1.1922, | |
| "step": 7160 | |
| }, | |
| { | |
| "epoch": 0.7123076732027767, | |
| "grad_norm": 0.704018235206604, | |
| "learning_rate": 5.93583358681467e-05, | |
| "loss": 1.1721, | |
| "step": 7170 | |
| }, | |
| { | |
| "epoch": 0.7133011288139381, | |
| "grad_norm": 0.7272495627403259, | |
| "learning_rate": 5.897818713665282e-05, | |
| "loss": 1.1664, | |
| "step": 7180 | |
| }, | |
| { | |
| "epoch": 0.7142945844250996, | |
| "grad_norm": 0.8413404822349548, | |
| "learning_rate": 5.8598961509259504e-05, | |
| "loss": 1.1866, | |
| "step": 7190 | |
| }, | |
| { | |
| "epoch": 0.7152880400362611, | |
| "grad_norm": 0.7830241918563843, | |
| "learning_rate": 5.8220662831910135e-05, | |
| "loss": 1.1959, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.7162814956474226, | |
| "grad_norm": 0.7813401222229004, | |
| "learning_rate": 5.784329494114721e-05, | |
| "loss": 1.1786, | |
| "step": 7210 | |
| }, | |
| { | |
| "epoch": 0.7172749512585841, | |
| "grad_norm": 0.7667372822761536, | |
| "learning_rate": 5.746686166407369e-05, | |
| "loss": 1.209, | |
| "step": 7220 | |
| }, | |
| { | |
| "epoch": 0.7182684068697456, | |
| "grad_norm": 0.7826696634292603, | |
| "learning_rate": 5.709136681831413e-05, | |
| "loss": 1.1849, | |
| "step": 7230 | |
| }, | |
| { | |
| "epoch": 0.7192618624809071, | |
| "grad_norm": 0.8333719968795776, | |
| "learning_rate": 5.671681421197576e-05, | |
| "loss": 1.1931, | |
| "step": 7240 | |
| }, | |
| { | |
| "epoch": 0.7202553180920686, | |
| "grad_norm": 0.7203525304794312, | |
| "learning_rate": 5.634320764361013e-05, | |
| "loss": 1.1755, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 0.7212487737032299, | |
| "grad_norm": 0.7497394680976868, | |
| "learning_rate": 5.5970550902174486e-05, | |
| "loss": 1.1784, | |
| "step": 7260 | |
| }, | |
| { | |
| "epoch": 0.7222422293143914, | |
| "grad_norm": 0.7904976606369019, | |
| "learning_rate": 5.559884776699341e-05, | |
| "loss": 1.1756, | |
| "step": 7270 | |
| }, | |
| { | |
| "epoch": 0.7232356849255529, | |
| "grad_norm": 0.8021249771118164, | |
| "learning_rate": 5.522810200772013e-05, | |
| "loss": 1.1824, | |
| "step": 7280 | |
| }, | |
| { | |
| "epoch": 0.7242291405367144, | |
| "grad_norm": 0.777125358581543, | |
| "learning_rate": 5.485831738429888e-05, | |
| "loss": 1.185, | |
| "step": 7290 | |
| }, | |
| { | |
| "epoch": 0.7252225961478759, | |
| "grad_norm": 0.6873647570610046, | |
| "learning_rate": 5.448949764692632e-05, | |
| "loss": 1.1891, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.7262160517590374, | |
| "grad_norm": 0.7390309572219849, | |
| "learning_rate": 5.412164653601378e-05, | |
| "loss": 1.1955, | |
| "step": 7310 | |
| }, | |
| { | |
| "epoch": 0.7272095073701988, | |
| "grad_norm": 0.7592355608940125, | |
| "learning_rate": 5.375476778214899e-05, | |
| "loss": 1.1843, | |
| "step": 7320 | |
| }, | |
| { | |
| "epoch": 0.7282029629813603, | |
| "grad_norm": 0.7683237791061401, | |
| "learning_rate": 5.338886510605862e-05, | |
| "loss": 1.1728, | |
| "step": 7330 | |
| }, | |
| { | |
| "epoch": 0.7291964185925217, | |
| "grad_norm": 0.779436469078064, | |
| "learning_rate": 5.30239422185704e-05, | |
| "loss": 1.1964, | |
| "step": 7340 | |
| }, | |
| { | |
| "epoch": 0.7301898742036832, | |
| "grad_norm": 0.8311836123466492, | |
| "learning_rate": 5.266000282057532e-05, | |
| "loss": 1.1812, | |
| "step": 7350 | |
| }, | |
| { | |
| "epoch": 0.7311833298148447, | |
| "grad_norm": 0.8085299730300903, | |
| "learning_rate": 5.229705060299026e-05, | |
| "loss": 1.1834, | |
| "step": 7360 | |
| }, | |
| { | |
| "epoch": 0.7321767854260062, | |
| "grad_norm": 0.7968699336051941, | |
| "learning_rate": 5.193508924672067e-05, | |
| "loss": 1.1671, | |
| "step": 7370 | |
| }, | |
| { | |
| "epoch": 0.7331702410371677, | |
| "grad_norm": 0.7955072522163391, | |
| "learning_rate": 5.157412242262303e-05, | |
| "loss": 1.177, | |
| "step": 7380 | |
| }, | |
| { | |
| "epoch": 0.7341636966483291, | |
| "grad_norm": 0.7670128345489502, | |
| "learning_rate": 5.1214153791467755e-05, | |
| "loss": 1.1798, | |
| "step": 7390 | |
| }, | |
| { | |
| "epoch": 0.7351571522594906, | |
| "grad_norm": 0.7874760031700134, | |
| "learning_rate": 5.085518700390187e-05, | |
| "loss": 1.182, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.7361506078706521, | |
| "grad_norm": 0.8138370513916016, | |
| "learning_rate": 5.0497225700412334e-05, | |
| "loss": 1.1776, | |
| "step": 7410 | |
| }, | |
| { | |
| "epoch": 0.7371440634818136, | |
| "grad_norm": 0.8087670207023621, | |
| "learning_rate": 5.014027351128872e-05, | |
| "loss": 1.1768, | |
| "step": 7420 | |
| }, | |
| { | |
| "epoch": 0.738137519092975, | |
| "grad_norm": 0.8033069372177124, | |
| "learning_rate": 4.978433405658679e-05, | |
| "loss": 1.1716, | |
| "step": 7430 | |
| }, | |
| { | |
| "epoch": 0.7391309747041365, | |
| "grad_norm": 0.8186310529708862, | |
| "learning_rate": 4.9429410946091416e-05, | |
| "loss": 1.1778, | |
| "step": 7440 | |
| }, | |
| { | |
| "epoch": 0.740124430315298, | |
| "grad_norm": 0.7752142548561096, | |
| "learning_rate": 4.9075507779280246e-05, | |
| "loss": 1.2033, | |
| "step": 7450 | |
| }, | |
| { | |
| "epoch": 0.7411178859264594, | |
| "grad_norm": 0.6990722417831421, | |
| "learning_rate": 4.872262814528707e-05, | |
| "loss": 1.182, | |
| "step": 7460 | |
| }, | |
| { | |
| "epoch": 0.7421113415376209, | |
| "grad_norm": 0.7314838767051697, | |
| "learning_rate": 4.8370775622865535e-05, | |
| "loss": 1.1879, | |
| "step": 7470 | |
| }, | |
| { | |
| "epoch": 0.7431047971487824, | |
| "grad_norm": 0.8103840351104736, | |
| "learning_rate": 4.801995378035259e-05, | |
| "loss": 1.1734, | |
| "step": 7480 | |
| }, | |
| { | |
| "epoch": 0.7440982527599439, | |
| "grad_norm": 0.7305147051811218, | |
| "learning_rate": 4.767016617563254e-05, | |
| "loss": 1.1883, | |
| "step": 7490 | |
| }, | |
| { | |
| "epoch": 0.7450917083711054, | |
| "grad_norm": 0.781702995300293, | |
| "learning_rate": 4.732141635610098e-05, | |
| "loss": 1.1826, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.7460851639822668, | |
| "grad_norm": 0.8269298076629639, | |
| "learning_rate": 4.697370785862863e-05, | |
| "loss": 1.1701, | |
| "step": 7510 | |
| }, | |
| { | |
| "epoch": 0.7470786195934283, | |
| "grad_norm": 0.7583678364753723, | |
| "learning_rate": 4.6627044209525674e-05, | |
| "loss": 1.1834, | |
| "step": 7520 | |
| }, | |
| { | |
| "epoch": 0.7480720752045897, | |
| "grad_norm": 0.7382724285125732, | |
| "learning_rate": 4.628142892450575e-05, | |
| "loss": 1.1696, | |
| "step": 7530 | |
| }, | |
| { | |
| "epoch": 0.7490655308157512, | |
| "grad_norm": 0.8067662715911865, | |
| "learning_rate": 4.593686550865059e-05, | |
| "loss": 1.1863, | |
| "step": 7540 | |
| }, | |
| { | |
| "epoch": 0.7500589864269127, | |
| "grad_norm": 0.8716863989830017, | |
| "learning_rate": 4.5593357456374175e-05, | |
| "loss": 1.1944, | |
| "step": 7550 | |
| }, | |
| { | |
| "epoch": 0.7510524420380742, | |
| "grad_norm": 0.7286023497581482, | |
| "learning_rate": 4.5250908251387594e-05, | |
| "loss": 1.1767, | |
| "step": 7560 | |
| }, | |
| { | |
| "epoch": 0.7520458976492357, | |
| "grad_norm": 0.7735511064529419, | |
| "learning_rate": 4.490952136666338e-05, | |
| "loss": 1.185, | |
| "step": 7570 | |
| }, | |
| { | |
| "epoch": 0.7530393532603972, | |
| "grad_norm": 0.877860963344574, | |
| "learning_rate": 4.456920026440064e-05, | |
| "loss": 1.1813, | |
| "step": 7580 | |
| }, | |
| { | |
| "epoch": 0.7540328088715587, | |
| "grad_norm": 0.8026107549667358, | |
| "learning_rate": 4.422994839598969e-05, | |
| "loss": 1.1695, | |
| "step": 7590 | |
| }, | |
| { | |
| "epoch": 0.75502626448272, | |
| "grad_norm": 0.7550562024116516, | |
| "learning_rate": 4.389176920197723e-05, | |
| "loss": 1.1637, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.7560197200938815, | |
| "grad_norm": 0.8293710350990295, | |
| "learning_rate": 4.355466611203119e-05, | |
| "loss": 1.1645, | |
| "step": 7610 | |
| }, | |
| { | |
| "epoch": 0.757013175705043, | |
| "grad_norm": 0.8430331945419312, | |
| "learning_rate": 4.32186425449062e-05, | |
| "loss": 1.1743, | |
| "step": 7620 | |
| }, | |
| { | |
| "epoch": 0.7580066313162045, | |
| "grad_norm": 0.7460552453994751, | |
| "learning_rate": 4.288370190840887e-05, | |
| "loss": 1.1744, | |
| "step": 7630 | |
| }, | |
| { | |
| "epoch": 0.759000086927366, | |
| "grad_norm": 0.7314460277557373, | |
| "learning_rate": 4.2549847599363196e-05, | |
| "loss": 1.1713, | |
| "step": 7640 | |
| }, | |
| { | |
| "epoch": 0.7599935425385275, | |
| "grad_norm": 0.8186727166175842, | |
| "learning_rate": 4.221708300357598e-05, | |
| "loss": 1.1628, | |
| "step": 7650 | |
| }, | |
| { | |
| "epoch": 0.760986998149689, | |
| "grad_norm": 0.7350887060165405, | |
| "learning_rate": 4.1885411495802794e-05, | |
| "loss": 1.1794, | |
| "step": 7660 | |
| }, | |
| { | |
| "epoch": 0.7619804537608504, | |
| "grad_norm": 0.7874166369438171, | |
| "learning_rate": 4.155483643971349e-05, | |
| "loss": 1.1836, | |
| "step": 7670 | |
| }, | |
| { | |
| "epoch": 0.7629739093720119, | |
| "grad_norm": 0.7989648580551147, | |
| "learning_rate": 4.122536118785824e-05, | |
| "loss": 1.1799, | |
| "step": 7680 | |
| }, | |
| { | |
| "epoch": 0.7639673649831733, | |
| "grad_norm": 0.7851313352584839, | |
| "learning_rate": 4.089698908163329e-05, | |
| "loss": 1.2067, | |
| "step": 7690 | |
| }, | |
| { | |
| "epoch": 0.7649608205943348, | |
| "grad_norm": 0.8075825572013855, | |
| "learning_rate": 4.056972345124746e-05, | |
| "loss": 1.1983, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.7659542762054963, | |
| "grad_norm": 0.689725399017334, | |
| "learning_rate": 4.0243567615688116e-05, | |
| "loss": 1.1737, | |
| "step": 7710 | |
| }, | |
| { | |
| "epoch": 0.7669477318166578, | |
| "grad_norm": 0.7839321494102478, | |
| "learning_rate": 3.9918524882687566e-05, | |
| "loss": 1.1701, | |
| "step": 7720 | |
| }, | |
| { | |
| "epoch": 0.7679411874278192, | |
| "grad_norm": 0.8125682473182678, | |
| "learning_rate": 3.959459854868942e-05, | |
| "loss": 1.1846, | |
| "step": 7730 | |
| }, | |
| { | |
| "epoch": 0.7689346430389807, | |
| "grad_norm": 0.7505900263786316, | |
| "learning_rate": 3.9271791898815366e-05, | |
| "loss": 1.1908, | |
| "step": 7740 | |
| }, | |
| { | |
| "epoch": 0.7699280986501422, | |
| "grad_norm": 0.7755024433135986, | |
| "learning_rate": 3.895010820683172e-05, | |
| "loss": 1.1665, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 0.7709215542613037, | |
| "grad_norm": 0.8277163505554199, | |
| "learning_rate": 3.8629550735116114e-05, | |
| "loss": 1.1896, | |
| "step": 7760 | |
| }, | |
| { | |
| "epoch": 0.7719150098724651, | |
| "grad_norm": 0.7777776122093201, | |
| "learning_rate": 3.8310122734624754e-05, | |
| "loss": 1.1871, | |
| "step": 7770 | |
| }, | |
| { | |
| "epoch": 0.7729084654836266, | |
| "grad_norm": 0.747857391834259, | |
| "learning_rate": 3.7991827444859e-05, | |
| "loss": 1.1869, | |
| "step": 7780 | |
| }, | |
| { | |
| "epoch": 0.7739019210947881, | |
| "grad_norm": 0.8087073564529419, | |
| "learning_rate": 3.767466809383293e-05, | |
| "loss": 1.1975, | |
| "step": 7790 | |
| }, | |
| { | |
| "epoch": 0.7748953767059495, | |
| "grad_norm": 0.806193470954895, | |
| "learning_rate": 3.7358647898040364e-05, | |
| "loss": 1.1906, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.775888832317111, | |
| "grad_norm": 0.7572741508483887, | |
| "learning_rate": 3.7043770062422246e-05, | |
| "loss": 1.1764, | |
| "step": 7810 | |
| }, | |
| { | |
| "epoch": 0.7768822879282725, | |
| "grad_norm": 0.8623882532119751, | |
| "learning_rate": 3.6730037780334277e-05, | |
| "loss": 1.1813, | |
| "step": 7820 | |
| }, | |
| { | |
| "epoch": 0.777875743539434, | |
| "grad_norm": 0.7911874055862427, | |
| "learning_rate": 3.6417454233514335e-05, | |
| "loss": 1.1721, | |
| "step": 7830 | |
| }, | |
| { | |
| "epoch": 0.7788691991505955, | |
| "grad_norm": 0.7923675775527954, | |
| "learning_rate": 3.610602259205044e-05, | |
| "loss": 1.1732, | |
| "step": 7840 | |
| }, | |
| { | |
| "epoch": 0.779862654761757, | |
| "grad_norm": 0.752544641494751, | |
| "learning_rate": 3.5795746014348466e-05, | |
| "loss": 1.1639, | |
| "step": 7850 | |
| }, | |
| { | |
| "epoch": 0.7808561103729184, | |
| "grad_norm": 0.7295422554016113, | |
| "learning_rate": 3.54866276471e-05, | |
| "loss": 1.1822, | |
| "step": 7860 | |
| }, | |
| { | |
| "epoch": 0.7818495659840798, | |
| "grad_norm": 0.8824591040611267, | |
| "learning_rate": 3.517867062525072e-05, | |
| "loss": 1.1598, | |
| "step": 7870 | |
| }, | |
| { | |
| "epoch": 0.7828430215952413, | |
| "grad_norm": 0.8134926557540894, | |
| "learning_rate": 3.487187807196838e-05, | |
| "loss": 1.1874, | |
| "step": 7880 | |
| }, | |
| { | |
| "epoch": 0.7838364772064028, | |
| "grad_norm": 0.7414039373397827, | |
| "learning_rate": 3.456625309861116e-05, | |
| "loss": 1.1684, | |
| "step": 7890 | |
| }, | |
| { | |
| "epoch": 0.7848299328175643, | |
| "grad_norm": 0.853863537311554, | |
| "learning_rate": 3.426179880469612e-05, | |
| "loss": 1.1753, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.7858233884287258, | |
| "grad_norm": 0.8354621529579163, | |
| "learning_rate": 3.3958518277867864e-05, | |
| "loss": 1.1836, | |
| "step": 7910 | |
| }, | |
| { | |
| "epoch": 0.7868168440398873, | |
| "grad_norm": 0.7733498811721802, | |
| "learning_rate": 3.365641459386715e-05, | |
| "loss": 1.1845, | |
| "step": 7920 | |
| }, | |
| { | |
| "epoch": 0.7878102996510488, | |
| "grad_norm": 0.7986984848976135, | |
| "learning_rate": 3.3355490816499716e-05, | |
| "loss": 1.1655, | |
| "step": 7930 | |
| }, | |
| { | |
| "epoch": 0.7888037552622101, | |
| "grad_norm": 0.8243693709373474, | |
| "learning_rate": 3.305574999760507e-05, | |
| "loss": 1.1809, | |
| "step": 7940 | |
| }, | |
| { | |
| "epoch": 0.7897972108733716, | |
| "grad_norm": 0.7955079674720764, | |
| "learning_rate": 3.275719517702587e-05, | |
| "loss": 1.1823, | |
| "step": 7950 | |
| }, | |
| { | |
| "epoch": 0.7907906664845331, | |
| "grad_norm": 0.8276157975196838, | |
| "learning_rate": 3.2459829382576664e-05, | |
| "loss": 1.1676, | |
| "step": 7960 | |
| }, | |
| { | |
| "epoch": 0.7917841220956946, | |
| "grad_norm": 0.9546493887901306, | |
| "learning_rate": 3.21636556300136e-05, | |
| "loss": 1.1659, | |
| "step": 7970 | |
| }, | |
| { | |
| "epoch": 0.7927775777068561, | |
| "grad_norm": 0.8397723436355591, | |
| "learning_rate": 3.186867692300349e-05, | |
| "loss": 1.181, | |
| "step": 7980 | |
| }, | |
| { | |
| "epoch": 0.7937710333180176, | |
| "grad_norm": 0.7565129995346069, | |
| "learning_rate": 3.157489625309363e-05, | |
| "loss": 1.1724, | |
| "step": 7990 | |
| }, | |
| { | |
| "epoch": 0.7947644889291791, | |
| "grad_norm": 0.7505949139595032, | |
| "learning_rate": 3.12823165996813e-05, | |
| "loss": 1.1767, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.7957579445403405, | |
| "grad_norm": 0.7717160582542419, | |
| "learning_rate": 3.099094092998359e-05, | |
| "loss": 1.1594, | |
| "step": 8010 | |
| }, | |
| { | |
| "epoch": 0.796751400151502, | |
| "grad_norm": 0.7836345434188843, | |
| "learning_rate": 3.070077219900729e-05, | |
| "loss": 1.1822, | |
| "step": 8020 | |
| }, | |
| { | |
| "epoch": 0.7977448557626634, | |
| "grad_norm": 0.8176319003105164, | |
| "learning_rate": 3.041181334951887e-05, | |
| "loss": 1.1818, | |
| "step": 8030 | |
| }, | |
| { | |
| "epoch": 0.7987383113738249, | |
| "grad_norm": 0.7628329396247864, | |
| "learning_rate": 3.012406731201485e-05, | |
| "loss": 1.1633, | |
| "step": 8040 | |
| }, | |
| { | |
| "epoch": 0.7997317669849864, | |
| "grad_norm": 0.7567328214645386, | |
| "learning_rate": 2.9837537004691852e-05, | |
| "loss": 1.174, | |
| "step": 8050 | |
| }, | |
| { | |
| "epoch": 0.8003278403516833, | |
| "eval_loss": 1.1950321197509766, | |
| "eval_runtime": 68.8887, | |
| "eval_samples_per_second": 725.808, | |
| "eval_steps_per_second": 90.726, | |
| "step": 8056 | |
| }, | |
| { | |
| "epoch": 0.8007252225961479, | |
| "grad_norm": 0.7983657717704773, | |
| "learning_rate": 2.9552225333417047e-05, | |
| "loss": 1.1869, | |
| "step": 8060 | |
| }, | |
| { | |
| "epoch": 0.8017186782073094, | |
| "grad_norm": 0.8555404543876648, | |
| "learning_rate": 2.9268135191698776e-05, | |
| "loss": 1.1711, | |
| "step": 8070 | |
| }, | |
| { | |
| "epoch": 0.8027121338184708, | |
| "grad_norm": 0.8287683725357056, | |
| "learning_rate": 2.8985269460657175e-05, | |
| "loss": 1.1876, | |
| "step": 8080 | |
| }, | |
| { | |
| "epoch": 0.8037055894296323, | |
| "grad_norm": 0.8173229694366455, | |
| "learning_rate": 2.870363100899491e-05, | |
| "loss": 1.1715, | |
| "step": 8090 | |
| }, | |
| { | |
| "epoch": 0.8046990450407938, | |
| "grad_norm": 0.8117800354957581, | |
| "learning_rate": 2.8423222692968052e-05, | |
| "loss": 1.1778, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.8056925006519552, | |
| "grad_norm": 0.8124157786369324, | |
| "learning_rate": 2.814404735635721e-05, | |
| "loss": 1.1878, | |
| "step": 8110 | |
| }, | |
| { | |
| "epoch": 0.8066859562631167, | |
| "grad_norm": 0.7694131135940552, | |
| "learning_rate": 2.7866107830438626e-05, | |
| "loss": 1.1852, | |
| "step": 8120 | |
| }, | |
| { | |
| "epoch": 0.8076794118742782, | |
| "grad_norm": 0.7774087190628052, | |
| "learning_rate": 2.7589406933955597e-05, | |
| "loss": 1.155, | |
| "step": 8130 | |
| }, | |
| { | |
| "epoch": 0.8086728674854397, | |
| "grad_norm": 0.8443126082420349, | |
| "learning_rate": 2.731394747308956e-05, | |
| "loss": 1.1987, | |
| "step": 8140 | |
| }, | |
| { | |
| "epoch": 0.8096663230966011, | |
| "grad_norm": 0.8412149548530579, | |
| "learning_rate": 2.7039732241432016e-05, | |
| "loss": 1.1836, | |
| "step": 8150 | |
| }, | |
| { | |
| "epoch": 0.8106597787077626, | |
| "grad_norm": 0.8976776003837585, | |
| "learning_rate": 2.6766764019956006e-05, | |
| "loss": 1.1656, | |
| "step": 8160 | |
| }, | |
| { | |
| "epoch": 0.8116532343189241, | |
| "grad_norm": 0.9143022894859314, | |
| "learning_rate": 2.6495045576987828e-05, | |
| "loss": 1.154, | |
| "step": 8170 | |
| }, | |
| { | |
| "epoch": 0.8126466899300856, | |
| "grad_norm": 0.836409330368042, | |
| "learning_rate": 2.622457966817921e-05, | |
| "loss": 1.1801, | |
| "step": 8180 | |
| }, | |
| { | |
| "epoch": 0.8136401455412471, | |
| "grad_norm": 0.8520244359970093, | |
| "learning_rate": 2.5955369036479047e-05, | |
| "loss": 1.1819, | |
| "step": 8190 | |
| }, | |
| { | |
| "epoch": 0.8146336011524085, | |
| "grad_norm": 0.808027982711792, | |
| "learning_rate": 2.568741641210591e-05, | |
| "loss": 1.1731, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.81562705676357, | |
| "grad_norm": 0.8576006293296814, | |
| "learning_rate": 2.5420724512520142e-05, | |
| "loss": 1.1728, | |
| "step": 8210 | |
| }, | |
| { | |
| "epoch": 0.8166205123747314, | |
| "grad_norm": 0.7371129393577576, | |
| "learning_rate": 2.5155296042396305e-05, | |
| "loss": 1.1696, | |
| "step": 8220 | |
| }, | |
| { | |
| "epoch": 0.8176139679858929, | |
| "grad_norm": 0.8558706045150757, | |
| "learning_rate": 2.489113369359592e-05, | |
| "loss": 1.1738, | |
| "step": 8230 | |
| }, | |
| { | |
| "epoch": 0.8186074235970544, | |
| "grad_norm": 0.8826967477798462, | |
| "learning_rate": 2.462824014513991e-05, | |
| "loss": 1.1706, | |
| "step": 8240 | |
| }, | |
| { | |
| "epoch": 0.8196008792082159, | |
| "grad_norm": 0.8093799948692322, | |
| "learning_rate": 2.4366618063181652e-05, | |
| "loss": 1.1755, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 0.8205943348193774, | |
| "grad_norm": 0.7722868323326111, | |
| "learning_rate": 2.4106270100979898e-05, | |
| "loss": 1.1652, | |
| "step": 8260 | |
| }, | |
| { | |
| "epoch": 0.8215877904305389, | |
| "grad_norm": 0.7924707531929016, | |
| "learning_rate": 2.3847198898871688e-05, | |
| "loss": 1.1743, | |
| "step": 8270 | |
| }, | |
| { | |
| "epoch": 0.8225812460417004, | |
| "grad_norm": 0.7128357887268066, | |
| "learning_rate": 2.3589407084245854e-05, | |
| "loss": 1.1795, | |
| "step": 8280 | |
| }, | |
| { | |
| "epoch": 0.8235747016528617, | |
| "grad_norm": 0.7344933152198792, | |
| "learning_rate": 2.3332897271516177e-05, | |
| "loss": 1.1619, | |
| "step": 8290 | |
| }, | |
| { | |
| "epoch": 0.8245681572640232, | |
| "grad_norm": 0.8097250461578369, | |
| "learning_rate": 2.3077672062094892e-05, | |
| "loss": 1.1717, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.8255616128751847, | |
| "grad_norm": 0.7288644313812256, | |
| "learning_rate": 2.2823734044366325e-05, | |
| "loss": 1.1694, | |
| "step": 8310 | |
| }, | |
| { | |
| "epoch": 0.8265550684863462, | |
| "grad_norm": 0.857816755771637, | |
| "learning_rate": 2.257108579366074e-05, | |
| "loss": 1.1663, | |
| "step": 8320 | |
| }, | |
| { | |
| "epoch": 0.8275485240975077, | |
| "grad_norm": 0.8221887350082397, | |
| "learning_rate": 2.2319729872228086e-05, | |
| "loss": 1.155, | |
| "step": 8330 | |
| }, | |
| { | |
| "epoch": 0.8285419797086692, | |
| "grad_norm": 0.6939984560012817, | |
| "learning_rate": 2.2069668829212122e-05, | |
| "loss": 1.1872, | |
| "step": 8340 | |
| }, | |
| { | |
| "epoch": 0.8295354353198306, | |
| "grad_norm": 0.7757712602615356, | |
| "learning_rate": 2.1820905200624382e-05, | |
| "loss": 1.1858, | |
| "step": 8350 | |
| }, | |
| { | |
| "epoch": 0.8305288909309921, | |
| "grad_norm": 0.7371575236320496, | |
| "learning_rate": 2.1573441509318727e-05, | |
| "loss": 1.1663, | |
| "step": 8360 | |
| }, | |
| { | |
| "epoch": 0.8315223465421535, | |
| "grad_norm": 0.7536593675613403, | |
| "learning_rate": 2.1327280264965462e-05, | |
| "loss": 1.1678, | |
| "step": 8370 | |
| }, | |
| { | |
| "epoch": 0.832515802153315, | |
| "grad_norm": 0.8135145902633667, | |
| "learning_rate": 2.1082423964026208e-05, | |
| "loss": 1.1662, | |
| "step": 8380 | |
| }, | |
| { | |
| "epoch": 0.8335092577644765, | |
| "grad_norm": 0.8527625203132629, | |
| "learning_rate": 2.0838875089728236e-05, | |
| "loss": 1.1738, | |
| "step": 8390 | |
| }, | |
| { | |
| "epoch": 0.834502713375638, | |
| "grad_norm": 0.8822425007820129, | |
| "learning_rate": 2.059663611203956e-05, | |
| "loss": 1.1786, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.8354961689867995, | |
| "grad_norm": 0.7784630656242371, | |
| "learning_rate": 2.0355709487643807e-05, | |
| "loss": 1.1707, | |
| "step": 8410 | |
| }, | |
| { | |
| "epoch": 0.836489624597961, | |
| "grad_norm": 0.8175725936889648, | |
| "learning_rate": 2.0116097659915225e-05, | |
| "loss": 1.1651, | |
| "step": 8420 | |
| }, | |
| { | |
| "epoch": 0.8374830802091224, | |
| "grad_norm": 0.8268553614616394, | |
| "learning_rate": 1.987780305889397e-05, | |
| "loss": 1.1558, | |
| "step": 8430 | |
| }, | |
| { | |
| "epoch": 0.8384765358202839, | |
| "grad_norm": 0.843661904335022, | |
| "learning_rate": 1.9640828101261413e-05, | |
| "loss": 1.1748, | |
| "step": 8440 | |
| }, | |
| { | |
| "epoch": 0.8394699914314454, | |
| "grad_norm": 0.799462616443634, | |
| "learning_rate": 1.9405175190315747e-05, | |
| "loss": 1.1803, | |
| "step": 8450 | |
| }, | |
| { | |
| "epoch": 0.8404634470426068, | |
| "grad_norm": 0.7566912174224854, | |
| "learning_rate": 1.917084671594754e-05, | |
| "loss": 1.1688, | |
| "step": 8460 | |
| }, | |
| { | |
| "epoch": 0.8414569026537683, | |
| "grad_norm": 0.8587673306465149, | |
| "learning_rate": 1.8937845054615373e-05, | |
| "loss": 1.1858, | |
| "step": 8470 | |
| }, | |
| { | |
| "epoch": 0.8424503582649298, | |
| "grad_norm": 0.8087132573127747, | |
| "learning_rate": 1.870617256932199e-05, | |
| "loss": 1.1817, | |
| "step": 8480 | |
| }, | |
| { | |
| "epoch": 0.8434438138760912, | |
| "grad_norm": 0.8132085204124451, | |
| "learning_rate": 1.8475831609590135e-05, | |
| "loss": 1.1777, | |
| "step": 8490 | |
| }, | |
| { | |
| "epoch": 0.8444372694872527, | |
| "grad_norm": 0.7978829145431519, | |
| "learning_rate": 1.8246824511438868e-05, | |
| "loss": 1.1709, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.8454307250984142, | |
| "grad_norm": 0.9162365794181824, | |
| "learning_rate": 1.801915359735967e-05, | |
| "loss": 1.16, | |
| "step": 8510 | |
| }, | |
| { | |
| "epoch": 0.8464241807095757, | |
| "grad_norm": 0.7341846823692322, | |
| "learning_rate": 1.7792821176293097e-05, | |
| "loss": 1.1752, | |
| "step": 8520 | |
| }, | |
| { | |
| "epoch": 0.8474176363207372, | |
| "grad_norm": 0.8641170859336853, | |
| "learning_rate": 1.7567829543605282e-05, | |
| "loss": 1.1738, | |
| "step": 8530 | |
| }, | |
| { | |
| "epoch": 0.8484110919318986, | |
| "grad_norm": 0.7875086069107056, | |
| "learning_rate": 1.734418098106468e-05, | |
| "loss": 1.1672, | |
| "step": 8540 | |
| }, | |
| { | |
| "epoch": 0.84940454754306, | |
| "grad_norm": 0.79355788230896, | |
| "learning_rate": 1.712187775681883e-05, | |
| "loss": 1.1644, | |
| "step": 8550 | |
| }, | |
| { | |
| "epoch": 0.8503980031542215, | |
| "grad_norm": 0.8003886938095093, | |
| "learning_rate": 1.6900922125371498e-05, | |
| "loss": 1.1545, | |
| "step": 8560 | |
| }, | |
| { | |
| "epoch": 0.851391458765383, | |
| "grad_norm": 0.7897474765777588, | |
| "learning_rate": 1.6681316327559763e-05, | |
| "loss": 1.1803, | |
| "step": 8570 | |
| }, | |
| { | |
| "epoch": 0.8523849143765445, | |
| "grad_norm": 0.8580754399299622, | |
| "learning_rate": 1.646306259053116e-05, | |
| "loss": 1.1554, | |
| "step": 8580 | |
| }, | |
| { | |
| "epoch": 0.853378369987706, | |
| "grad_norm": 0.7858597040176392, | |
| "learning_rate": 1.624616312772133e-05, | |
| "loss": 1.1608, | |
| "step": 8590 | |
| }, | |
| { | |
| "epoch": 0.8543718255988675, | |
| "grad_norm": 0.8318164348602295, | |
| "learning_rate": 1.6030620138831336e-05, | |
| "loss": 1.1613, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.855365281210029, | |
| "grad_norm": 0.8344900608062744, | |
| "learning_rate": 1.5816435809805557e-05, | |
| "loss": 1.1685, | |
| "step": 8610 | |
| }, | |
| { | |
| "epoch": 0.8563587368211905, | |
| "grad_norm": 0.8422057628631592, | |
| "learning_rate": 1.5603612312809373e-05, | |
| "loss": 1.1558, | |
| "step": 8620 | |
| }, | |
| { | |
| "epoch": 0.8573521924323518, | |
| "grad_norm": 0.8174178600311279, | |
| "learning_rate": 1.5392151806207254e-05, | |
| "loss": 1.1784, | |
| "step": 8630 | |
| }, | |
| { | |
| "epoch": 0.8583456480435133, | |
| "grad_norm": 0.8342280983924866, | |
| "learning_rate": 1.5182056434540729e-05, | |
| "loss": 1.1774, | |
| "step": 8640 | |
| }, | |
| { | |
| "epoch": 0.8593391036546748, | |
| "grad_norm": 0.7873639464378357, | |
| "learning_rate": 1.4973328328506734e-05, | |
| "loss": 1.1618, | |
| "step": 8650 | |
| }, | |
| { | |
| "epoch": 0.8603325592658363, | |
| "grad_norm": 0.7351046204566956, | |
| "learning_rate": 1.4765969604935995e-05, | |
| "loss": 1.191, | |
| "step": 8660 | |
| }, | |
| { | |
| "epoch": 0.8613260148769978, | |
| "grad_norm": 0.7845040559768677, | |
| "learning_rate": 1.4559982366771616e-05, | |
| "loss": 1.1645, | |
| "step": 8670 | |
| }, | |
| { | |
| "epoch": 0.8623194704881593, | |
| "grad_norm": 0.8081471920013428, | |
| "learning_rate": 1.4355368703047543e-05, | |
| "loss": 1.1765, | |
| "step": 8680 | |
| }, | |
| { | |
| "epoch": 0.8633129260993208, | |
| "grad_norm": 0.7547688484191895, | |
| "learning_rate": 1.4152130688867647e-05, | |
| "loss": 1.1518, | |
| "step": 8690 | |
| }, | |
| { | |
| "epoch": 0.8643063817104822, | |
| "grad_norm": 0.7919139862060547, | |
| "learning_rate": 1.3950270385384526e-05, | |
| "loss": 1.197, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.8652998373216437, | |
| "grad_norm": 0.739045262336731, | |
| "learning_rate": 1.3749789839778574e-05, | |
| "loss": 1.1626, | |
| "step": 8710 | |
| }, | |
| { | |
| "epoch": 0.8662932929328051, | |
| "grad_norm": 0.826343834400177, | |
| "learning_rate": 1.3550691085237292e-05, | |
| "loss": 1.1568, | |
| "step": 8720 | |
| }, | |
| { | |
| "epoch": 0.8672867485439666, | |
| "grad_norm": 0.7689382433891296, | |
| "learning_rate": 1.335297614093469e-05, | |
| "loss": 1.1513, | |
| "step": 8730 | |
| }, | |
| { | |
| "epoch": 0.8682802041551281, | |
| "grad_norm": 0.8238071203231812, | |
| "learning_rate": 1.315664701201074e-05, | |
| "loss": 1.1565, | |
| "step": 8740 | |
| }, | |
| { | |
| "epoch": 0.8692736597662896, | |
| "grad_norm": 0.7542545795440674, | |
| "learning_rate": 1.2961705689551122e-05, | |
| "loss": 1.1765, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 0.870267115377451, | |
| "grad_norm": 0.8268214464187622, | |
| "learning_rate": 1.2768154150566856e-05, | |
| "loss": 1.1852, | |
| "step": 8760 | |
| }, | |
| { | |
| "epoch": 0.8712605709886125, | |
| "grad_norm": 0.824836015701294, | |
| "learning_rate": 1.2575994357974523e-05, | |
| "loss": 1.1688, | |
| "step": 8770 | |
| }, | |
| { | |
| "epoch": 0.872254026599774, | |
| "grad_norm": 0.7631413340568542, | |
| "learning_rate": 1.2385228260576091e-05, | |
| "loss": 1.1749, | |
| "step": 8780 | |
| }, | |
| { | |
| "epoch": 0.8732474822109355, | |
| "grad_norm": 0.7756265997886658, | |
| "learning_rate": 1.2195857793039337e-05, | |
| "loss": 1.1888, | |
| "step": 8790 | |
| }, | |
| { | |
| "epoch": 0.8742409378220969, | |
| "grad_norm": 0.7800768613815308, | |
| "learning_rate": 1.2007884875878143e-05, | |
| "loss": 1.1522, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.8752343934332584, | |
| "grad_norm": 0.8278046250343323, | |
| "learning_rate": 1.182131141543301e-05, | |
| "loss": 1.185, | |
| "step": 8810 | |
| }, | |
| { | |
| "epoch": 0.8762278490444199, | |
| "grad_norm": 0.7902718782424927, | |
| "learning_rate": 1.163613930385181e-05, | |
| "loss": 1.1609, | |
| "step": 8820 | |
| }, | |
| { | |
| "epoch": 0.8772213046555813, | |
| "grad_norm": 0.7979821562767029, | |
| "learning_rate": 1.1452370419070517e-05, | |
| "loss": 1.158, | |
| "step": 8830 | |
| }, | |
| { | |
| "epoch": 0.8782147602667428, | |
| "grad_norm": 0.8079224824905396, | |
| "learning_rate": 1.1270006624794154e-05, | |
| "loss": 1.1587, | |
| "step": 8840 | |
| }, | |
| { | |
| "epoch": 0.8792082158779043, | |
| "grad_norm": 0.7798774242401123, | |
| "learning_rate": 1.1089049770477892e-05, | |
| "loss": 1.1785, | |
| "step": 8850 | |
| }, | |
| { | |
| "epoch": 0.8802016714890658, | |
| "grad_norm": 0.7864110469818115, | |
| "learning_rate": 1.0909501691308387e-05, | |
| "loss": 1.1724, | |
| "step": 8860 | |
| }, | |
| { | |
| "epoch": 0.8811951271002273, | |
| "grad_norm": 0.85590660572052, | |
| "learning_rate": 1.07313642081851e-05, | |
| "loss": 1.1655, | |
| "step": 8870 | |
| }, | |
| { | |
| "epoch": 0.8821885827113888, | |
| "grad_norm": 0.8220794200897217, | |
| "learning_rate": 1.0554639127701808e-05, | |
| "loss": 1.1715, | |
| "step": 8880 | |
| }, | |
| { | |
| "epoch": 0.8831820383225502, | |
| "grad_norm": 0.7962018251419067, | |
| "learning_rate": 1.0379328242128294e-05, | |
| "loss": 1.1624, | |
| "step": 8890 | |
| }, | |
| { | |
| "epoch": 0.8841754939337116, | |
| "grad_norm": 0.8360487222671509, | |
| "learning_rate": 1.020543332939221e-05, | |
| "loss": 1.1717, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.8851689495448731, | |
| "grad_norm": 0.766708254814148, | |
| "learning_rate": 1.003295615306105e-05, | |
| "loss": 1.1723, | |
| "step": 8910 | |
| }, | |
| { | |
| "epoch": 0.8861624051560346, | |
| "grad_norm": 0.7551305890083313, | |
| "learning_rate": 9.861898462324158e-06, | |
| "loss": 1.1592, | |
| "step": 8920 | |
| }, | |
| { | |
| "epoch": 0.8871558607671961, | |
| "grad_norm": 0.918136477470398, | |
| "learning_rate": 9.692261991975076e-06, | |
| "loss": 1.1513, | |
| "step": 8930 | |
| }, | |
| { | |
| "epoch": 0.8881493163783576, | |
| "grad_norm": 0.7590110898017883, | |
| "learning_rate": 9.524048462394024e-06, | |
| "loss": 1.1785, | |
| "step": 8940 | |
| }, | |
| { | |
| "epoch": 0.8891427719895191, | |
| "grad_norm": 0.7427127957344055, | |
| "learning_rate": 9.357259579530285e-06, | |
| "loss": 1.1805, | |
| "step": 8950 | |
| }, | |
| { | |
| "epoch": 0.8901362276006806, | |
| "grad_norm": 0.7382677793502808, | |
| "learning_rate": 9.191897034885033e-06, | |
| "loss": 1.1467, | |
| "step": 8960 | |
| }, | |
| { | |
| "epoch": 0.8911296832118419, | |
| "grad_norm": 0.7082747220993042, | |
| "learning_rate": 9.027962505494108e-06, | |
| "loss": 1.1633, | |
| "step": 8970 | |
| }, | |
| { | |
| "epoch": 0.8921231388230034, | |
| "grad_norm": 0.8186103701591492, | |
| "learning_rate": 8.86545765391104e-06, | |
| "loss": 1.1761, | |
| "step": 8980 | |
| }, | |
| { | |
| "epoch": 0.8931165944341649, | |
| "grad_norm": 0.8058136701583862, | |
| "learning_rate": 8.704384128190156e-06, | |
| "loss": 1.157, | |
| "step": 8990 | |
| }, | |
| { | |
| "epoch": 0.8941100500453264, | |
| "grad_norm": 0.8163788914680481, | |
| "learning_rate": 8.544743561869977e-06, | |
| "loss": 1.1704, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.8951035056564879, | |
| "grad_norm": 0.8774188756942749, | |
| "learning_rate": 8.386537573956431e-06, | |
| "loss": 1.1691, | |
| "step": 9010 | |
| }, | |
| { | |
| "epoch": 0.8960969612676494, | |
| "grad_norm": 0.8346332907676697, | |
| "learning_rate": 8.229767768906687e-06, | |
| "loss": 1.1819, | |
| "step": 9020 | |
| }, | |
| { | |
| "epoch": 0.8970904168788109, | |
| "grad_norm": 0.7622361183166504, | |
| "learning_rate": 8.074435736612712e-06, | |
| "loss": 1.1507, | |
| "step": 9030 | |
| }, | |
| { | |
| "epoch": 0.8980838724899723, | |
| "grad_norm": 0.8322452306747437, | |
| "learning_rate": 7.920543052385236e-06, | |
| "loss": 1.1888, | |
| "step": 9040 | |
| }, | |
| { | |
| "epoch": 0.8990773281011338, | |
| "grad_norm": 0.8201948404312134, | |
| "learning_rate": 7.768091276937688e-06, | |
| "loss": 1.1689, | |
| "step": 9050 | |
| }, | |
| { | |
| "epoch": 0.9000707837122952, | |
| "grad_norm": 0.8077680468559265, | |
| "learning_rate": 7.617081956370413e-06, | |
| "loss": 1.149, | |
| "step": 9060 | |
| }, | |
| { | |
| "epoch": 0.9003688203956437, | |
| "eval_loss": 1.188559889793396, | |
| "eval_runtime": 69.534, | |
| "eval_samples_per_second": 719.072, | |
| "eval_steps_per_second": 89.884, | |
| "step": 9063 | |
| }, | |
| { | |
| "epoch": 0.9010642393234567, | |
| "grad_norm": 0.834770679473877, | |
| "learning_rate": 7.467516622155079e-06, | |
| "loss": 1.1752, | |
| "step": 9070 | |
| }, | |
| { | |
| "epoch": 0.9020576949346182, | |
| "grad_norm": 0.7466756105422974, | |
| "learning_rate": 7.319396791118998e-06, | |
| "loss": 1.1644, | |
| "step": 9080 | |
| }, | |
| { | |
| "epoch": 0.9030511505457797, | |
| "grad_norm": 0.7706833481788635, | |
| "learning_rate": 7.172723965429811e-06, | |
| "loss": 1.1632, | |
| "step": 9090 | |
| }, | |
| { | |
| "epoch": 0.9040446061569412, | |
| "grad_norm": 0.8348069787025452, | |
| "learning_rate": 7.027499632580264e-06, | |
| "loss": 1.1452, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 0.9050380617681026, | |
| "grad_norm": 0.7877137660980225, | |
| "learning_rate": 6.883725265373185e-06, | |
| "loss": 1.1604, | |
| "step": 9110 | |
| }, | |
| { | |
| "epoch": 0.9060315173792641, | |
| "grad_norm": 0.7920989990234375, | |
| "learning_rate": 6.74140232190632e-06, | |
| "loss": 1.1508, | |
| "step": 9120 | |
| }, | |
| { | |
| "epoch": 0.9070249729904256, | |
| "grad_norm": 0.8129798769950867, | |
| "learning_rate": 6.600532245557833e-06, | |
| "loss": 1.1642, | |
| "step": 9130 | |
| }, | |
| { | |
| "epoch": 0.908018428601587, | |
| "grad_norm": 0.8059536218643188, | |
| "learning_rate": 6.461116464971461e-06, | |
| "loss": 1.1709, | |
| "step": 9140 | |
| }, | |
| { | |
| "epoch": 0.9090118842127485, | |
| "grad_norm": 0.8482730388641357, | |
| "learning_rate": 6.323156394042139e-06, | |
| "loss": 1.1842, | |
| "step": 9150 | |
| }, | |
| { | |
| "epoch": 0.91000533982391, | |
| "grad_norm": 0.7679184675216675, | |
| "learning_rate": 6.186653431901628e-06, | |
| "loss": 1.1646, | |
| "step": 9160 | |
| }, | |
| { | |
| "epoch": 0.9109987954350715, | |
| "grad_norm": 0.8613263964653015, | |
| "learning_rate": 6.051608962904225e-06, | |
| "loss": 1.1815, | |
| "step": 9170 | |
| }, | |
| { | |
| "epoch": 0.9119922510462329, | |
| "grad_norm": 0.869649350643158, | |
| "learning_rate": 5.91802435661296e-06, | |
| "loss": 1.1809, | |
| "step": 9180 | |
| }, | |
| { | |
| "epoch": 0.9129857066573944, | |
| "grad_norm": 0.8057877421379089, | |
| "learning_rate": 5.78590096778544e-06, | |
| "loss": 1.1587, | |
| "step": 9190 | |
| }, | |
| { | |
| "epoch": 0.9139791622685559, | |
| "grad_norm": 0.7845506072044373, | |
| "learning_rate": 5.655240136360306e-06, | |
| "loss": 1.1636, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 0.9149726178797174, | |
| "grad_norm": 0.8760326504707336, | |
| "learning_rate": 5.5260431874435515e-06, | |
| "loss": 1.1608, | |
| "step": 9210 | |
| }, | |
| { | |
| "epoch": 0.9159660734908789, | |
| "grad_norm": 0.8257114291191101, | |
| "learning_rate": 5.398311431295061e-06, | |
| "loss": 1.17, | |
| "step": 9220 | |
| }, | |
| { | |
| "epoch": 0.9169595291020403, | |
| "grad_norm": 0.7520872354507446, | |
| "learning_rate": 5.2720461633154354e-06, | |
| "loss": 1.1806, | |
| "step": 9230 | |
| }, | |
| { | |
| "epoch": 0.9179529847132017, | |
| "grad_norm": 0.7858303785324097, | |
| "learning_rate": 5.147248664032744e-06, | |
| "loss": 1.1662, | |
| "step": 9240 | |
| }, | |
| { | |
| "epoch": 0.9189464403243632, | |
| "grad_norm": 0.8171572685241699, | |
| "learning_rate": 5.023920199089576e-06, | |
| "loss": 1.1593, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 0.9199398959355247, | |
| "grad_norm": 0.8115087747573853, | |
| "learning_rate": 4.902062019230191e-06, | |
| "loss": 1.1581, | |
| "step": 9260 | |
| }, | |
| { | |
| "epoch": 0.9209333515466862, | |
| "grad_norm": 0.8597251176834106, | |
| "learning_rate": 4.78167536028789e-06, | |
| "loss": 1.1497, | |
| "step": 9270 | |
| }, | |
| { | |
| "epoch": 0.9219268071578477, | |
| "grad_norm": 0.870826780796051, | |
| "learning_rate": 4.662761443172414e-06, | |
| "loss": 1.1718, | |
| "step": 9280 | |
| }, | |
| { | |
| "epoch": 0.9229202627690092, | |
| "grad_norm": 0.7543158531188965, | |
| "learning_rate": 4.545321473857616e-06, | |
| "loss": 1.1673, | |
| "step": 9290 | |
| }, | |
| { | |
| "epoch": 0.9239137183801707, | |
| "grad_norm": 0.8586430549621582, | |
| "learning_rate": 4.429356643369159e-06, | |
| "loss": 1.1612, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.9249071739913322, | |
| "grad_norm": 0.7621310353279114, | |
| "learning_rate": 4.314868127772503e-06, | |
| "loss": 1.1572, | |
| "step": 9310 | |
| }, | |
| { | |
| "epoch": 0.9259006296024935, | |
| "grad_norm": 0.7759812474250793, | |
| "learning_rate": 4.201857088160987e-06, | |
| "loss": 1.1596, | |
| "step": 9320 | |
| }, | |
| { | |
| "epoch": 0.926894085213655, | |
| "grad_norm": 0.8084028959274292, | |
| "learning_rate": 4.090324670643968e-06, | |
| "loss": 1.1666, | |
| "step": 9330 | |
| }, | |
| { | |
| "epoch": 0.9278875408248165, | |
| "grad_norm": 0.783535361289978, | |
| "learning_rate": 3.9802720063353e-06, | |
| "loss": 1.1731, | |
| "step": 9340 | |
| }, | |
| { | |
| "epoch": 0.928880996435978, | |
| "grad_norm": 0.7652338743209839, | |
| "learning_rate": 3.8717002113417885e-06, | |
| "loss": 1.1694, | |
| "step": 9350 | |
| }, | |
| { | |
| "epoch": 0.9298744520471395, | |
| "grad_norm": 0.8363282084465027, | |
| "learning_rate": 3.764610386751904e-06, | |
| "loss": 1.161, | |
| "step": 9360 | |
| }, | |
| { | |
| "epoch": 0.930867907658301, | |
| "grad_norm": 0.8643134832382202, | |
| "learning_rate": 3.659003618624656e-06, | |
| "loss": 1.1585, | |
| "step": 9370 | |
| }, | |
| { | |
| "epoch": 0.9318613632694625, | |
| "grad_norm": 0.9236533641815186, | |
| "learning_rate": 3.554880977978436e-06, | |
| "loss": 1.1805, | |
| "step": 9380 | |
| }, | |
| { | |
| "epoch": 0.9328548188806239, | |
| "grad_norm": 0.8215625286102295, | |
| "learning_rate": 3.4522435207803567e-06, | |
| "loss": 1.1658, | |
| "step": 9390 | |
| }, | |
| { | |
| "epoch": 0.9338482744917853, | |
| "grad_norm": 0.8430384993553162, | |
| "learning_rate": 3.3510922879353476e-06, | |
| "loss": 1.1682, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 0.9348417301029468, | |
| "grad_norm": 0.758561372756958, | |
| "learning_rate": 3.2514283052757605e-06, | |
| "loss": 1.1798, | |
| "step": 9410 | |
| }, | |
| { | |
| "epoch": 0.9358351857141083, | |
| "grad_norm": 0.8656542301177979, | |
| "learning_rate": 3.1532525835508636e-06, | |
| "loss": 1.1689, | |
| "step": 9420 | |
| }, | |
| { | |
| "epoch": 0.9368286413252698, | |
| "grad_norm": 0.8123307228088379, | |
| "learning_rate": 3.0565661184166146e-06, | |
| "loss": 1.1573, | |
| "step": 9430 | |
| }, | |
| { | |
| "epoch": 0.9378220969364313, | |
| "grad_norm": 0.8003418445587158, | |
| "learning_rate": 2.9613698904255857e-06, | |
| "loss": 1.1707, | |
| "step": 9440 | |
| }, | |
| { | |
| "epoch": 0.9388155525475927, | |
| "grad_norm": 0.7613969445228577, | |
| "learning_rate": 2.867664865017022e-06, | |
| "loss": 1.1752, | |
| "step": 9450 | |
| }, | |
| { | |
| "epoch": 0.9398090081587542, | |
| "grad_norm": 0.7638654112815857, | |
| "learning_rate": 2.7754519925069995e-06, | |
| "loss": 1.1435, | |
| "step": 9460 | |
| }, | |
| { | |
| "epoch": 0.9408024637699157, | |
| "grad_norm": 0.8079109191894531, | |
| "learning_rate": 2.684732208078816e-06, | |
| "loss": 1.1662, | |
| "step": 9470 | |
| }, | |
| { | |
| "epoch": 0.9417959193810772, | |
| "grad_norm": 0.7946303486824036, | |
| "learning_rate": 2.5955064317735474e-06, | |
| "loss": 1.1885, | |
| "step": 9480 | |
| }, | |
| { | |
| "epoch": 0.9427893749922386, | |
| "grad_norm": 0.7448862195014954, | |
| "learning_rate": 2.5077755684806745e-06, | |
| "loss": 1.178, | |
| "step": 9490 | |
| }, | |
| { | |
| "epoch": 0.9437828306034001, | |
| "grad_norm": 0.8106252551078796, | |
| "learning_rate": 2.421540507928871e-06, | |
| "loss": 1.1519, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.9447762862145616, | |
| "grad_norm": 0.7957506775856018, | |
| "learning_rate": 2.3368021246770786e-06, | |
| "loss": 1.1427, | |
| "step": 9510 | |
| }, | |
| { | |
| "epoch": 0.945769741825723, | |
| "grad_norm": 0.8124579787254333, | |
| "learning_rate": 2.2535612781055478e-06, | |
| "loss": 1.1639, | |
| "step": 9520 | |
| }, | |
| { | |
| "epoch": 0.9467631974368845, | |
| "grad_norm": 0.8010926842689514, | |
| "learning_rate": 2.1718188124071613e-06, | |
| "loss": 1.164, | |
| "step": 9530 | |
| }, | |
| { | |
| "epoch": 0.947756653048046, | |
| "grad_norm": 0.7903379797935486, | |
| "learning_rate": 2.0915755565788728e-06, | |
| "loss": 1.1864, | |
| "step": 9540 | |
| }, | |
| { | |
| "epoch": 0.9487501086592075, | |
| "grad_norm": 0.8351110219955444, | |
| "learning_rate": 2.0128323244132826e-06, | |
| "loss": 1.1852, | |
| "step": 9550 | |
| }, | |
| { | |
| "epoch": 0.949743564270369, | |
| "grad_norm": 0.795259952545166, | |
| "learning_rate": 1.9355899144904263e-06, | |
| "loss": 1.1736, | |
| "step": 9560 | |
| }, | |
| { | |
| "epoch": 0.9507370198815304, | |
| "grad_norm": 0.8469606637954712, | |
| "learning_rate": 1.8598491101695978e-06, | |
| "loss": 1.1738, | |
| "step": 9570 | |
| }, | |
| { | |
| "epoch": 0.9517304754926919, | |
| "grad_norm": 0.7984156012535095, | |
| "learning_rate": 1.7856106795814729e-06, | |
| "loss": 1.1785, | |
| "step": 9580 | |
| }, | |
| { | |
| "epoch": 0.9527239311038533, | |
| "grad_norm": 0.7923161387443542, | |
| "learning_rate": 1.7128753756203484e-06, | |
| "loss": 1.1546, | |
| "step": 9590 | |
| }, | |
| { | |
| "epoch": 0.9537173867150148, | |
| "grad_norm": 0.7685096859931946, | |
| "learning_rate": 1.6416439359363487e-06, | |
| "loss": 1.1614, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 0.9547108423261763, | |
| "grad_norm": 0.9111791849136353, | |
| "learning_rate": 1.5719170829281646e-06, | |
| "loss": 1.1668, | |
| "step": 9610 | |
| }, | |
| { | |
| "epoch": 0.9557042979373378, | |
| "grad_norm": 0.831987738609314, | |
| "learning_rate": 1.503695523735543e-06, | |
| "loss": 1.163, | |
| "step": 9620 | |
| }, | |
| { | |
| "epoch": 0.9566977535484993, | |
| "grad_norm": 0.8165103793144226, | |
| "learning_rate": 1.4369799502322254e-06, | |
| "loss": 1.1578, | |
| "step": 9630 | |
| }, | |
| { | |
| "epoch": 0.9576912091596608, | |
| "grad_norm": 0.8375231623649597, | |
| "learning_rate": 1.3717710390188709e-06, | |
| "loss": 1.1581, | |
| "step": 9640 | |
| }, | |
| { | |
| "epoch": 0.9586846647708223, | |
| "grad_norm": 0.7733319401741028, | |
| "learning_rate": 1.3080694514162437e-06, | |
| "loss": 1.1683, | |
| "step": 9650 | |
| }, | |
| { | |
| "epoch": 0.9596781203819836, | |
| "grad_norm": 0.8246985077857971, | |
| "learning_rate": 1.245875833458454e-06, | |
| "loss": 1.1503, | |
| "step": 9660 | |
| }, | |
| { | |
| "epoch": 0.9606715759931451, | |
| "grad_norm": 0.820965051651001, | |
| "learning_rate": 1.1851908158864443e-06, | |
| "loss": 1.1609, | |
| "step": 9670 | |
| }, | |
| { | |
| "epoch": 0.9616650316043066, | |
| "grad_norm": 0.7744933366775513, | |
| "learning_rate": 1.126015014141546e-06, | |
| "loss": 1.1528, | |
| "step": 9680 | |
| }, | |
| { | |
| "epoch": 0.9626584872154681, | |
| "grad_norm": 0.8470078706741333, | |
| "learning_rate": 1.0683490283593499e-06, | |
| "loss": 1.1573, | |
| "step": 9690 | |
| }, | |
| { | |
| "epoch": 0.9636519428266296, | |
| "grad_norm": 0.7112888693809509, | |
| "learning_rate": 1.0121934433634793e-06, | |
| "loss": 1.1525, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 0.9646453984377911, | |
| "grad_norm": 0.7978183031082153, | |
| "learning_rate": 9.575488286597267e-07, | |
| "loss": 1.1537, | |
| "step": 9710 | |
| }, | |
| { | |
| "epoch": 0.9656388540489526, | |
| "grad_norm": 0.8091950416564941, | |
| "learning_rate": 9.044157384302919e-07, | |
| "loss": 1.1595, | |
| "step": 9720 | |
| }, | |
| { | |
| "epoch": 0.966632309660114, | |
| "grad_norm": 0.8772552609443665, | |
| "learning_rate": 8.527947115281542e-07, | |
| "loss": 1.1775, | |
| "step": 9730 | |
| }, | |
| { | |
| "epoch": 0.9676257652712754, | |
| "grad_norm": 0.8062159419059753, | |
| "learning_rate": 8.026862714715587e-07, | |
| "loss": 1.1764, | |
| "step": 9740 | |
| }, | |
| { | |
| "epoch": 0.9686192208824369, | |
| "grad_norm": 0.8115324974060059, | |
| "learning_rate": 7.540909264387718e-07, | |
| "loss": 1.1508, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 0.9696126764935984, | |
| "grad_norm": 0.86089688539505, | |
| "learning_rate": 7.07009169262901e-07, | |
| "loss": 1.1555, | |
| "step": 9760 | |
| }, | |
| { | |
| "epoch": 0.9706061321047599, | |
| "grad_norm": 0.8365864753723145, | |
| "learning_rate": 6.614414774268661e-07, | |
| "loss": 1.1578, | |
| "step": 9770 | |
| }, | |
| { | |
| "epoch": 0.9715995877159214, | |
| "grad_norm": 0.8947882652282715, | |
| "learning_rate": 6.173883130586532e-07, | |
| "loss": 1.1672, | |
| "step": 9780 | |
| }, | |
| { | |
| "epoch": 0.9725930433270829, | |
| "grad_norm": 0.8084526658058167, | |
| "learning_rate": 5.748501229264679e-07, | |
| "loss": 1.1638, | |
| "step": 9790 | |
| }, | |
| { | |
| "epoch": 0.9735864989382443, | |
| "grad_norm": 0.7722569108009338, | |
| "learning_rate": 5.338273384343895e-07, | |
| "loss": 1.1612, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 0.9745799545494058, | |
| "grad_norm": 0.8078683614730835, | |
| "learning_rate": 4.94320375617857e-07, | |
| "loss": 1.1594, | |
| "step": 9810 | |
| }, | |
| { | |
| "epoch": 0.9755734101605673, | |
| "grad_norm": 0.8281704783439636, | |
| "learning_rate": 4.563296351394907e-07, | |
| "loss": 1.1648, | |
| "step": 9820 | |
| }, | |
| { | |
| "epoch": 0.9765668657717287, | |
| "grad_norm": 0.7503883838653564, | |
| "learning_rate": 4.1985550228506025e-07, | |
| "loss": 1.1648, | |
| "step": 9830 | |
| }, | |
| { | |
| "epoch": 0.9775603213828902, | |
| "grad_norm": 0.7680873870849609, | |
| "learning_rate": 3.8489834695953903e-07, | |
| "loss": 1.1459, | |
| "step": 9840 | |
| }, | |
| { | |
| "epoch": 0.9785537769940517, | |
| "grad_norm": 0.934911847114563, | |
| "learning_rate": 3.5145852368337336e-07, | |
| "loss": 1.1768, | |
| "step": 9850 | |
| }, | |
| { | |
| "epoch": 0.9795472326052131, | |
| "grad_norm": 0.8128591179847717, | |
| "learning_rate": 3.195363715888688e-07, | |
| "loss": 1.1567, | |
| "step": 9860 | |
| }, | |
| { | |
| "epoch": 0.9805406882163746, | |
| "grad_norm": 0.816461980342865, | |
| "learning_rate": 2.891322144167929e-07, | |
| "loss": 1.1673, | |
| "step": 9870 | |
| }, | |
| { | |
| "epoch": 0.9815341438275361, | |
| "grad_norm": 0.8307123780250549, | |
| "learning_rate": 2.6024636051301116e-07, | |
| "loss": 1.1731, | |
| "step": 9880 | |
| }, | |
| { | |
| "epoch": 0.9825275994386976, | |
| "grad_norm": 0.8082227110862732, | |
| "learning_rate": 2.328791028254562e-07, | |
| "loss": 1.1666, | |
| "step": 9890 | |
| }, | |
| { | |
| "epoch": 0.9835210550498591, | |
| "grad_norm": 0.8234909176826477, | |
| "learning_rate": 2.0703071890108003e-07, | |
| "loss": 1.171, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 0.9845145106610206, | |
| "grad_norm": 0.8709157705307007, | |
| "learning_rate": 1.8270147088308984e-07, | |
| "loss": 1.1469, | |
| "step": 9910 | |
| }, | |
| { | |
| "epoch": 0.985507966272182, | |
| "grad_norm": 0.7917748093605042, | |
| "learning_rate": 1.5989160550826663e-07, | |
| "loss": 1.1719, | |
| "step": 9920 | |
| }, | |
| { | |
| "epoch": 0.9865014218833434, | |
| "grad_norm": 0.8313570022583008, | |
| "learning_rate": 1.3860135410443395e-07, | |
| "loss": 1.1747, | |
| "step": 9930 | |
| }, | |
| { | |
| "epoch": 0.9874948774945049, | |
| "grad_norm": 0.7465665936470032, | |
| "learning_rate": 1.1883093258819309e-07, | |
| "loss": 1.1626, | |
| "step": 9940 | |
| }, | |
| { | |
| "epoch": 0.9884883331056664, | |
| "grad_norm": 0.7561696767807007, | |
| "learning_rate": 1.0058054146265815e-07, | |
| "loss": 1.1603, | |
| "step": 9950 | |
| }, | |
| { | |
| "epoch": 0.9894817887168279, | |
| "grad_norm": 0.8261615037918091, | |
| "learning_rate": 8.385036581545768e-08, | |
| "loss": 1.1504, | |
| "step": 9960 | |
| }, | |
| { | |
| "epoch": 0.9904752443279894, | |
| "grad_norm": 0.7834208011627197, | |
| "learning_rate": 6.864057531681955e-08, | |
| "loss": 1.1643, | |
| "step": 9970 | |
| }, | |
| { | |
| "epoch": 0.9914686999391509, | |
| "grad_norm": 0.8045084476470947, | |
| "learning_rate": 5.49513242179056e-08, | |
| "loss": 1.1801, | |
| "step": 9980 | |
| }, | |
| { | |
| "epoch": 0.9924621555503124, | |
| "grad_norm": 0.7740246057510376, | |
| "learning_rate": 4.2782751349246246e-08, | |
| "loss": 1.1537, | |
| "step": 9990 | |
| }, | |
| { | |
| "epoch": 0.9934556111614737, | |
| "grad_norm": 0.780992329120636, | |
| "learning_rate": 3.213498011924165e-08, | |
| "loss": 1.1764, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.9944490667726352, | |
| "grad_norm": 0.8399405479431152, | |
| "learning_rate": 2.3008118513012653e-08, | |
| "loss": 1.164, | |
| "step": 10010 | |
| }, | |
| { | |
| "epoch": 0.9954425223837967, | |
| "grad_norm": 0.7551798224449158, | |
| "learning_rate": 1.540225909130166e-08, | |
| "loss": 1.1673, | |
| "step": 10020 | |
| }, | |
| { | |
| "epoch": 0.9964359779949582, | |
| "grad_norm": 0.8198529481887817, | |
| "learning_rate": 9.317478989423477e-09, | |
| "loss": 1.1585, | |
| "step": 10030 | |
| }, | |
| { | |
| "epoch": 0.9974294336061197, | |
| "grad_norm": 1.6382498741149902, | |
| "learning_rate": 4.753839916632474e-09, | |
| "loss": 1.1752, | |
| "step": 10040 | |
| }, | |
| { | |
| "epoch": 0.9984228892172812, | |
| "grad_norm": 0.801209568977356, | |
| "learning_rate": 1.7113881553731945e-09, | |
| "loss": 1.1774, | |
| "step": 10050 | |
| }, | |
| { | |
| "epoch": 0.9994163448284427, | |
| "grad_norm": 0.8735531568527222, | |
| "learning_rate": 1.901545608973265e-10, | |
| "loss": 1.1713, | |
| "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": 3661365038284800.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |