| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 5949, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0016809547823163557, | |
| "grad_norm": 79552.359375, | |
| "learning_rate": 4.9915952260884186e-05, | |
| "loss": 1.0656, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0033619095646327114, | |
| "grad_norm": 79056.6640625, | |
| "learning_rate": 4.983190452176837e-05, | |
| "loss": 1.0215, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.005042864346949067, | |
| "grad_norm": 85323.828125, | |
| "learning_rate": 4.9747856782652546e-05, | |
| "loss": 0.9578, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.006723819129265423, | |
| "grad_norm": 71872.125, | |
| "learning_rate": 4.966380904353673e-05, | |
| "loss": 0.9151, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.008404773911581778, | |
| "grad_norm": 71638.46875, | |
| "learning_rate": 4.957976130442091e-05, | |
| "loss": 0.8803, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.010085728693898134, | |
| "grad_norm": 74304.5234375, | |
| "learning_rate": 4.9495713565305096e-05, | |
| "loss": 0.9141, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.01176668347621449, | |
| "grad_norm": 73676.1953125, | |
| "learning_rate": 4.941166582618928e-05, | |
| "loss": 0.9949, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.013447638258530846, | |
| "grad_norm": 76955.7890625, | |
| "learning_rate": 4.9327618087073463e-05, | |
| "loss": 0.8852, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.015128593040847202, | |
| "grad_norm": 69415.2578125, | |
| "learning_rate": 4.924357034795764e-05, | |
| "loss": 0.8891, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.016809547823163556, | |
| "grad_norm": 84991.7734375, | |
| "learning_rate": 4.9159522608841824e-05, | |
| "loss": 0.939, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01849050260547991, | |
| "grad_norm": 72803.203125, | |
| "learning_rate": 4.907547486972601e-05, | |
| "loss": 0.8879, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.020171457387796268, | |
| "grad_norm": 73581.6640625, | |
| "learning_rate": 4.899142713061019e-05, | |
| "loss": 0.923, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.021852412170112624, | |
| "grad_norm": 80610.7421875, | |
| "learning_rate": 4.8907379391494374e-05, | |
| "loss": 0.8933, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.02353336695242898, | |
| "grad_norm": 80112.125, | |
| "learning_rate": 4.882333165237855e-05, | |
| "loss": 0.8761, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.025214321734745335, | |
| "grad_norm": 73625.0546875, | |
| "learning_rate": 4.8739283913262734e-05, | |
| "loss": 0.8442, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.02689527651706169, | |
| "grad_norm": 79199.546875, | |
| "learning_rate": 4.865523617414692e-05, | |
| "loss": 0.9312, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.028576231299378047, | |
| "grad_norm": 69096.6484375, | |
| "learning_rate": 4.85711884350311e-05, | |
| "loss": 0.9435, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.030257186081694403, | |
| "grad_norm": 66993.53125, | |
| "learning_rate": 4.8487140695915285e-05, | |
| "loss": 0.8933, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.031938140864010756, | |
| "grad_norm": 72366.515625, | |
| "learning_rate": 4.840309295679946e-05, | |
| "loss": 0.8513, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.03361909564632711, | |
| "grad_norm": 70981.46875, | |
| "learning_rate": 4.8319045217683645e-05, | |
| "loss": 0.9126, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03530005042864347, | |
| "grad_norm": 73961.296875, | |
| "learning_rate": 4.823499747856783e-05, | |
| "loss": 0.908, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.03698100521095982, | |
| "grad_norm": 71402.046875, | |
| "learning_rate": 4.815094973945201e-05, | |
| "loss": 0.915, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.03866195999327618, | |
| "grad_norm": 69597.8984375, | |
| "learning_rate": 4.8066902000336195e-05, | |
| "loss": 0.892, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.040342914775592535, | |
| "grad_norm": 77749.3359375, | |
| "learning_rate": 4.798285426122038e-05, | |
| "loss": 0.9004, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.04202386955790889, | |
| "grad_norm": 69176.328125, | |
| "learning_rate": 4.7898806522104555e-05, | |
| "loss": 0.9878, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.04370482434022525, | |
| "grad_norm": 76940.0625, | |
| "learning_rate": 4.781475878298874e-05, | |
| "loss": 0.8873, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.0453857791225416, | |
| "grad_norm": 81460.0546875, | |
| "learning_rate": 4.773071104387292e-05, | |
| "loss": 0.9022, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.04706673390485796, | |
| "grad_norm": 69520.1953125, | |
| "learning_rate": 4.7646663304757106e-05, | |
| "loss": 0.8729, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.048747688687174315, | |
| "grad_norm": 76301.21875, | |
| "learning_rate": 4.756261556564129e-05, | |
| "loss": 0.9464, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.05042864346949067, | |
| "grad_norm": 72831.2265625, | |
| "learning_rate": 4.7478567826525466e-05, | |
| "loss": 0.8346, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.05210959825180703, | |
| "grad_norm": 73389.6328125, | |
| "learning_rate": 4.739452008740965e-05, | |
| "loss": 0.8794, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.05379055303412338, | |
| "grad_norm": 78976.9453125, | |
| "learning_rate": 4.731047234829383e-05, | |
| "loss": 0.9318, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.05547150781643974, | |
| "grad_norm": 72745.15625, | |
| "learning_rate": 4.7226424609178016e-05, | |
| "loss": 0.923, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.057152462598756094, | |
| "grad_norm": 65032.38671875, | |
| "learning_rate": 4.71423768700622e-05, | |
| "loss": 0.9431, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.05883341738107245, | |
| "grad_norm": 67867.875, | |
| "learning_rate": 4.705832913094638e-05, | |
| "loss": 0.853, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.060514372163388806, | |
| "grad_norm": 132520.328125, | |
| "learning_rate": 4.697428139183056e-05, | |
| "loss": 0.8772, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.06219532694570516, | |
| "grad_norm": 71777.140625, | |
| "learning_rate": 4.6890233652714743e-05, | |
| "loss": 0.841, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.06387628172802151, | |
| "grad_norm": 71626.65625, | |
| "learning_rate": 4.680618591359893e-05, | |
| "loss": 0.8281, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.06555723651033787, | |
| "grad_norm": 66127.0703125, | |
| "learning_rate": 4.672213817448311e-05, | |
| "loss": 0.9221, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.06723819129265422, | |
| "grad_norm": 66148.8203125, | |
| "learning_rate": 4.6638090435367294e-05, | |
| "loss": 0.8607, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.06891914607497059, | |
| "grad_norm": 75503.5078125, | |
| "learning_rate": 4.655404269625147e-05, | |
| "loss": 0.9062, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.07060010085728693, | |
| "grad_norm": 61276.5625, | |
| "learning_rate": 4.6469994957135654e-05, | |
| "loss": 0.8669, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.0722810556396033, | |
| "grad_norm": 71196.9453125, | |
| "learning_rate": 4.638594721801984e-05, | |
| "loss": 0.8686, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.07396201042191965, | |
| "grad_norm": 66918.8671875, | |
| "learning_rate": 4.630189947890402e-05, | |
| "loss": 0.9121, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.07564296520423601, | |
| "grad_norm": 64304.73828125, | |
| "learning_rate": 4.6217851739788204e-05, | |
| "loss": 0.8663, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.07732391998655236, | |
| "grad_norm": 62031.97265625, | |
| "learning_rate": 4.613380400067238e-05, | |
| "loss": 0.8258, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.07900487476886872, | |
| "grad_norm": 63083.9140625, | |
| "learning_rate": 4.6049756261556565e-05, | |
| "loss": 0.8657, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.08068582955118507, | |
| "grad_norm": 67486.75, | |
| "learning_rate": 4.596570852244075e-05, | |
| "loss": 0.9225, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.08236678433350143, | |
| "grad_norm": 68711.7890625, | |
| "learning_rate": 4.588166078332493e-05, | |
| "loss": 0.8172, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.08404773911581778, | |
| "grad_norm": 69319.46875, | |
| "learning_rate": 4.5797613044209115e-05, | |
| "loss": 0.8034, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.08572869389813415, | |
| "grad_norm": 67749.515625, | |
| "learning_rate": 4.57135653050933e-05, | |
| "loss": 0.9285, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.0874096486804505, | |
| "grad_norm": 78357.3203125, | |
| "learning_rate": 4.5629517565977475e-05, | |
| "loss": 0.8566, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.08909060346276686, | |
| "grad_norm": 59291.59375, | |
| "learning_rate": 4.554546982686166e-05, | |
| "loss": 0.8088, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.0907715582450832, | |
| "grad_norm": 72898.515625, | |
| "learning_rate": 4.546142208774584e-05, | |
| "loss": 0.8411, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.09245251302739957, | |
| "grad_norm": 65205.4296875, | |
| "learning_rate": 4.5377374348630026e-05, | |
| "loss": 0.9546, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.09413346780971592, | |
| "grad_norm": 104009.6953125, | |
| "learning_rate": 4.529332660951421e-05, | |
| "loss": 0.8406, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.09581442259203228, | |
| "grad_norm": 68204.453125, | |
| "learning_rate": 4.5209278870398386e-05, | |
| "loss": 0.897, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.09749537737434863, | |
| "grad_norm": 70454.1484375, | |
| "learning_rate": 4.512523113128257e-05, | |
| "loss": 0.839, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.09917633215666499, | |
| "grad_norm": 67070.96875, | |
| "learning_rate": 4.504118339216675e-05, | |
| "loss": 0.9008, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.10085728693898134, | |
| "grad_norm": 70355.5234375, | |
| "learning_rate": 4.4957135653050936e-05, | |
| "loss": 0.9318, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.10253824172129769, | |
| "grad_norm": 68497.7109375, | |
| "learning_rate": 4.487308791393512e-05, | |
| "loss": 0.9127, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.10421919650361405, | |
| "grad_norm": 81705.2578125, | |
| "learning_rate": 4.47890401748193e-05, | |
| "loss": 0.8575, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.1059001512859304, | |
| "grad_norm": 72572.75, | |
| "learning_rate": 4.470499243570348e-05, | |
| "loss": 0.8387, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.10758110606824677, | |
| "grad_norm": 66429.328125, | |
| "learning_rate": 4.462094469658766e-05, | |
| "loss": 0.8322, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.10926206085056311, | |
| "grad_norm": 73876.7265625, | |
| "learning_rate": 4.453689695747185e-05, | |
| "loss": 0.8788, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.11094301563287948, | |
| "grad_norm": 73238.40625, | |
| "learning_rate": 4.445284921835603e-05, | |
| "loss": 0.8983, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.11262397041519583, | |
| "grad_norm": 65609.2578125, | |
| "learning_rate": 4.4368801479240214e-05, | |
| "loss": 0.8406, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.11430492519751219, | |
| "grad_norm": 79022.8515625, | |
| "learning_rate": 4.428475374012439e-05, | |
| "loss": 0.8204, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.11598587997982854, | |
| "grad_norm": 63224.625, | |
| "learning_rate": 4.4200706001008574e-05, | |
| "loss": 0.8252, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.1176668347621449, | |
| "grad_norm": 61681.51171875, | |
| "learning_rate": 4.411665826189276e-05, | |
| "loss": 0.9041, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.11934778954446125, | |
| "grad_norm": 65631.578125, | |
| "learning_rate": 4.403261052277694e-05, | |
| "loss": 0.8541, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.12102874432677761, | |
| "grad_norm": 70977.875, | |
| "learning_rate": 4.3948562783661124e-05, | |
| "loss": 0.9705, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.12270969910909396, | |
| "grad_norm": 67375.078125, | |
| "learning_rate": 4.38645150445453e-05, | |
| "loss": 0.8542, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.12439065389141032, | |
| "grad_norm": 60750.97265625, | |
| "learning_rate": 4.3780467305429484e-05, | |
| "loss": 0.7832, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.1260716086737267, | |
| "grad_norm": 62710.984375, | |
| "learning_rate": 4.369641956631367e-05, | |
| "loss": 0.855, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.12775256345604302, | |
| "grad_norm": 56731.50390625, | |
| "learning_rate": 4.361237182719785e-05, | |
| "loss": 0.8591, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.12943351823835939, | |
| "grad_norm": 67084.78125, | |
| "learning_rate": 4.3528324088082035e-05, | |
| "loss": 0.8406, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.13111447302067575, | |
| "grad_norm": 66152.15625, | |
| "learning_rate": 4.344427634896622e-05, | |
| "loss": 0.9062, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.1327954278029921, | |
| "grad_norm": 71438.4765625, | |
| "learning_rate": 4.3360228609850395e-05, | |
| "loss": 0.8206, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.13447638258530845, | |
| "grad_norm": 60924.875, | |
| "learning_rate": 4.327618087073458e-05, | |
| "loss": 0.8164, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.1361573373676248, | |
| "grad_norm": 70178.5390625, | |
| "learning_rate": 4.319213313161876e-05, | |
| "loss": 0.8303, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.13783829214994117, | |
| "grad_norm": 64838.84765625, | |
| "learning_rate": 4.3108085392502945e-05, | |
| "loss": 0.8503, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.13951924693225753, | |
| "grad_norm": 72861.828125, | |
| "learning_rate": 4.302403765338713e-05, | |
| "loss": 0.9165, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.14120020171457387, | |
| "grad_norm": 74359.4140625, | |
| "learning_rate": 4.2939989914271306e-05, | |
| "loss": 0.9102, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.14288115649689023, | |
| "grad_norm": 70118.2109375, | |
| "learning_rate": 4.285594217515549e-05, | |
| "loss": 0.8703, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.1445621112792066, | |
| "grad_norm": 63720.28125, | |
| "learning_rate": 4.277189443603967e-05, | |
| "loss": 0.893, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.14624306606152296, | |
| "grad_norm": 77761.90625, | |
| "learning_rate": 4.2687846696923856e-05, | |
| "loss": 0.9613, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.1479240208438393, | |
| "grad_norm": 64969.37890625, | |
| "learning_rate": 4.260379895780804e-05, | |
| "loss": 0.8691, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.14960497562615566, | |
| "grad_norm": 70299.8203125, | |
| "learning_rate": 4.2519751218692216e-05, | |
| "loss": 0.8719, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.15128593040847202, | |
| "grad_norm": 65577.1171875, | |
| "learning_rate": 4.24357034795764e-05, | |
| "loss": 0.8554, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.15296688519078835, | |
| "grad_norm": 71740.8671875, | |
| "learning_rate": 4.235165574046058e-05, | |
| "loss": 0.9948, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.15464783997310472, | |
| "grad_norm": 59819.0234375, | |
| "learning_rate": 4.226760800134477e-05, | |
| "loss": 0.8605, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.15632879475542108, | |
| "grad_norm": 57238.7109375, | |
| "learning_rate": 4.218356026222895e-05, | |
| "loss": 0.8281, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.15800974953773744, | |
| "grad_norm": 70315.765625, | |
| "learning_rate": 4.2099512523113134e-05, | |
| "loss": 0.9126, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.15969070432005378, | |
| "grad_norm": 63374.6796875, | |
| "learning_rate": 4.201546478399731e-05, | |
| "loss": 0.8744, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.16137165910237014, | |
| "grad_norm": 75860.09375, | |
| "learning_rate": 4.1931417044881494e-05, | |
| "loss": 0.8078, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.1630526138846865, | |
| "grad_norm": 58454.1484375, | |
| "learning_rate": 4.184736930576568e-05, | |
| "loss": 0.8084, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.16473356866700287, | |
| "grad_norm": 68843.625, | |
| "learning_rate": 4.176332156664986e-05, | |
| "loss": 0.8458, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.1664145234493192, | |
| "grad_norm": 58082.43359375, | |
| "learning_rate": 4.1679273827534044e-05, | |
| "loss": 0.8148, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.16809547823163556, | |
| "grad_norm": 58417.78125, | |
| "learning_rate": 4.159522608841822e-05, | |
| "loss": 0.8373, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.16977643301395193, | |
| "grad_norm": 61807.9140625, | |
| "learning_rate": 4.1511178349302404e-05, | |
| "loss": 0.8633, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.1714573877962683, | |
| "grad_norm": 61544.55078125, | |
| "learning_rate": 4.142713061018659e-05, | |
| "loss": 0.7921, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.17313834257858463, | |
| "grad_norm": 78909.9921875, | |
| "learning_rate": 4.134308287107077e-05, | |
| "loss": 0.8422, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.174819297360901, | |
| "grad_norm": 75892.40625, | |
| "learning_rate": 4.1259035131954955e-05, | |
| "loss": 0.8307, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.17650025214321735, | |
| "grad_norm": 62162.703125, | |
| "learning_rate": 4.117498739283914e-05, | |
| "loss": 0.8243, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.1781812069255337, | |
| "grad_norm": 66981.4609375, | |
| "learning_rate": 4.1090939653723315e-05, | |
| "loss": 0.819, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.17986216170785005, | |
| "grad_norm": 69593.6796875, | |
| "learning_rate": 4.10068919146075e-05, | |
| "loss": 0.8817, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.1815431164901664, | |
| "grad_norm": 67435.6640625, | |
| "learning_rate": 4.092284417549168e-05, | |
| "loss": 0.8077, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.18322407127248277, | |
| "grad_norm": 66041.1640625, | |
| "learning_rate": 4.0838796436375865e-05, | |
| "loss": 0.841, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.18490502605479914, | |
| "grad_norm": 65792.2890625, | |
| "learning_rate": 4.075474869726005e-05, | |
| "loss": 0.8742, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.18658598083711547, | |
| "grad_norm": 61461.2265625, | |
| "learning_rate": 4.0670700958144226e-05, | |
| "loss": 0.8381, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.18826693561943184, | |
| "grad_norm": 62316.87109375, | |
| "learning_rate": 4.058665321902841e-05, | |
| "loss": 0.8629, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.1899478904017482, | |
| "grad_norm": 60252.7421875, | |
| "learning_rate": 4.050260547991259e-05, | |
| "loss": 0.8003, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.19162884518406456, | |
| "grad_norm": 133503.328125, | |
| "learning_rate": 4.0418557740796776e-05, | |
| "loss": 0.8469, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.1933097999663809, | |
| "grad_norm": 64984.43359375, | |
| "learning_rate": 4.033451000168096e-05, | |
| "loss": 0.8024, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.19499075474869726, | |
| "grad_norm": 64012.15234375, | |
| "learning_rate": 4.0250462262565136e-05, | |
| "loss": 0.8655, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.19667170953101362, | |
| "grad_norm": 80887.5859375, | |
| "learning_rate": 4.016641452344932e-05, | |
| "loss": 0.8273, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.19835266431332998, | |
| "grad_norm": 67626.34375, | |
| "learning_rate": 4.00823667843335e-05, | |
| "loss": 0.7986, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.20003361909564632, | |
| "grad_norm": 71404.109375, | |
| "learning_rate": 3.9998319045217687e-05, | |
| "loss": 0.796, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.20171457387796268, | |
| "grad_norm": 69063.609375, | |
| "learning_rate": 3.991427130610187e-05, | |
| "loss": 0.8819, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.20339552866027905, | |
| "grad_norm": 64813.16796875, | |
| "learning_rate": 3.9830223566986053e-05, | |
| "loss": 0.8451, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.20507648344259538, | |
| "grad_norm": 76470.9765625, | |
| "learning_rate": 3.974617582787023e-05, | |
| "loss": 0.8079, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.20675743822491174, | |
| "grad_norm": 57304.13671875, | |
| "learning_rate": 3.9662128088754414e-05, | |
| "loss": 0.7686, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.2084383930072281, | |
| "grad_norm": 58822.703125, | |
| "learning_rate": 3.95780803496386e-05, | |
| "loss": 0.8457, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.21011934778954447, | |
| "grad_norm": 61174.65234375, | |
| "learning_rate": 3.949403261052278e-05, | |
| "loss": 0.752, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.2118003025718608, | |
| "grad_norm": 72221.9296875, | |
| "learning_rate": 3.9409984871406964e-05, | |
| "loss": 0.8431, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.21348125735417717, | |
| "grad_norm": 64120.45703125, | |
| "learning_rate": 3.932593713229114e-05, | |
| "loss": 0.8505, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.21516221213649353, | |
| "grad_norm": 67320.4140625, | |
| "learning_rate": 3.9241889393175324e-05, | |
| "loss": 0.9042, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.2168431669188099, | |
| "grad_norm": 64738.58203125, | |
| "learning_rate": 3.915784165405951e-05, | |
| "loss": 0.8681, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.21852412170112623, | |
| "grad_norm": 87432.8515625, | |
| "learning_rate": 3.907379391494369e-05, | |
| "loss": 0.8661, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.2202050764834426, | |
| "grad_norm": 56789.4453125, | |
| "learning_rate": 3.8989746175827875e-05, | |
| "loss": 0.8058, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.22188603126575895, | |
| "grad_norm": 55766.0234375, | |
| "learning_rate": 3.890569843671206e-05, | |
| "loss": 0.8627, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.22356698604807532, | |
| "grad_norm": 68595.6640625, | |
| "learning_rate": 3.8821650697596235e-05, | |
| "loss": 0.8299, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.22524794083039165, | |
| "grad_norm": 67234.25, | |
| "learning_rate": 3.873760295848042e-05, | |
| "loss": 0.8106, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.22692889561270801, | |
| "grad_norm": 67707.8828125, | |
| "learning_rate": 3.86535552193646e-05, | |
| "loss": 0.8399, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.22860985039502438, | |
| "grad_norm": 60259.1015625, | |
| "learning_rate": 3.8569507480248785e-05, | |
| "loss": 0.8484, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.23029080517734074, | |
| "grad_norm": 68358.1484375, | |
| "learning_rate": 3.848545974113297e-05, | |
| "loss": 0.8264, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.23197175995965708, | |
| "grad_norm": 62014.7109375, | |
| "learning_rate": 3.8401412002017145e-05, | |
| "loss": 0.8507, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.23365271474197344, | |
| "grad_norm": 61060.28125, | |
| "learning_rate": 3.831736426290133e-05, | |
| "loss": 0.8586, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.2353336695242898, | |
| "grad_norm": 57990.08984375, | |
| "learning_rate": 3.823331652378551e-05, | |
| "loss": 0.8011, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.23701462430660616, | |
| "grad_norm": 69242.015625, | |
| "learning_rate": 3.8149268784669696e-05, | |
| "loss": 0.7943, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.2386955790889225, | |
| "grad_norm": 85857.8046875, | |
| "learning_rate": 3.806522104555388e-05, | |
| "loss": 0.864, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.24037653387123886, | |
| "grad_norm": 69873.5234375, | |
| "learning_rate": 3.7981173306438056e-05, | |
| "loss": 0.8578, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.24205748865355523, | |
| "grad_norm": 59830.73046875, | |
| "learning_rate": 3.789712556732224e-05, | |
| "loss": 0.8517, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.2437384434358716, | |
| "grad_norm": 59296.34765625, | |
| "learning_rate": 3.781307782820642e-05, | |
| "loss": 0.864, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.24541939821818792, | |
| "grad_norm": 63888.60546875, | |
| "learning_rate": 3.7729030089090606e-05, | |
| "loss": 0.8161, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.24710035300050429, | |
| "grad_norm": 63231.421875, | |
| "learning_rate": 3.764498234997479e-05, | |
| "loss": 0.8901, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.24878130778282065, | |
| "grad_norm": 62661.12890625, | |
| "learning_rate": 3.756093461085897e-05, | |
| "loss": 0.8638, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.250462262565137, | |
| "grad_norm": 69614.8984375, | |
| "learning_rate": 3.747688687174315e-05, | |
| "loss": 0.8644, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.2521432173474534, | |
| "grad_norm": 67442.9296875, | |
| "learning_rate": 3.7392839132627334e-05, | |
| "loss": 0.7659, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2538241721297697, | |
| "grad_norm": 67846.0390625, | |
| "learning_rate": 3.730879139351152e-05, | |
| "loss": 0.8243, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.25550512691208604, | |
| "grad_norm": 65967.703125, | |
| "learning_rate": 3.72247436543957e-05, | |
| "loss": 0.8573, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.25718608169440244, | |
| "grad_norm": 66188.1484375, | |
| "learning_rate": 3.7140695915279884e-05, | |
| "loss": 0.8491, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.25886703647671877, | |
| "grad_norm": 55857.7890625, | |
| "learning_rate": 3.705664817616406e-05, | |
| "loss": 0.8314, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.2605479912590351, | |
| "grad_norm": 78828.6328125, | |
| "learning_rate": 3.6972600437048244e-05, | |
| "loss": 0.7589, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.2622289460413515, | |
| "grad_norm": 64696.65234375, | |
| "learning_rate": 3.688855269793243e-05, | |
| "loss": 0.774, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.26390990082366783, | |
| "grad_norm": 71397.7265625, | |
| "learning_rate": 3.680450495881661e-05, | |
| "loss": 0.8351, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.2655908556059842, | |
| "grad_norm": 66192.203125, | |
| "learning_rate": 3.6720457219700795e-05, | |
| "loss": 0.8099, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.26727181038830056, | |
| "grad_norm": 70641.421875, | |
| "learning_rate": 3.663640948058498e-05, | |
| "loss": 0.8571, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.2689527651706169, | |
| "grad_norm": 75064.75, | |
| "learning_rate": 3.6552361741469155e-05, | |
| "loss": 0.8681, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.2706337199529333, | |
| "grad_norm": 75823.125, | |
| "learning_rate": 3.646831400235334e-05, | |
| "loss": 0.8258, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.2723146747352496, | |
| "grad_norm": 72296.765625, | |
| "learning_rate": 3.638426626323752e-05, | |
| "loss": 0.8561, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.27399562951756595, | |
| "grad_norm": 64309.3125, | |
| "learning_rate": 3.6300218524121705e-05, | |
| "loss": 0.856, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.27567658429988234, | |
| "grad_norm": 73902.8359375, | |
| "learning_rate": 3.621617078500589e-05, | |
| "loss": 0.8424, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.2773575390821987, | |
| "grad_norm": 63552.06640625, | |
| "learning_rate": 3.6132123045890065e-05, | |
| "loss": 0.861, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.27903849386451507, | |
| "grad_norm": 74067.7265625, | |
| "learning_rate": 3.604807530677425e-05, | |
| "loss": 0.7832, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.2807194486468314, | |
| "grad_norm": 68881.3046875, | |
| "learning_rate": 3.596402756765843e-05, | |
| "loss": 0.8443, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.28240040342914774, | |
| "grad_norm": 62996.19921875, | |
| "learning_rate": 3.5879979828542616e-05, | |
| "loss": 0.8075, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.28408135821146413, | |
| "grad_norm": 73364.8125, | |
| "learning_rate": 3.57959320894268e-05, | |
| "loss": 0.8425, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.28576231299378047, | |
| "grad_norm": 67509.296875, | |
| "learning_rate": 3.5711884350310976e-05, | |
| "loss": 0.755, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.2874432677760968, | |
| "grad_norm": 66616.984375, | |
| "learning_rate": 3.562783661119516e-05, | |
| "loss": 0.8679, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.2891242225584132, | |
| "grad_norm": 74004.359375, | |
| "learning_rate": 3.554378887207934e-05, | |
| "loss": 0.8793, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.2908051773407295, | |
| "grad_norm": 59084.9609375, | |
| "learning_rate": 3.5459741132963526e-05, | |
| "loss": 0.8322, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.2924861321230459, | |
| "grad_norm": 74027.28125, | |
| "learning_rate": 3.537569339384771e-05, | |
| "loss": 0.8661, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.29416708690536225, | |
| "grad_norm": 64524.97265625, | |
| "learning_rate": 3.529164565473189e-05, | |
| "loss": 0.7334, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.2958480416876786, | |
| "grad_norm": 74809.4921875, | |
| "learning_rate": 3.520759791561607e-05, | |
| "loss": 0.8091, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.297528996469995, | |
| "grad_norm": 66084.609375, | |
| "learning_rate": 3.5123550176500253e-05, | |
| "loss": 0.8715, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.2992099512523113, | |
| "grad_norm": 62919.05859375, | |
| "learning_rate": 3.503950243738444e-05, | |
| "loss": 0.8326, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.30089090603462765, | |
| "grad_norm": 68219.046875, | |
| "learning_rate": 3.495545469826862e-05, | |
| "loss": 0.8387, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.30257186081694404, | |
| "grad_norm": 61066.58203125, | |
| "learning_rate": 3.4871406959152804e-05, | |
| "loss": 0.8113, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.3042528155992604, | |
| "grad_norm": 64199.75390625, | |
| "learning_rate": 3.478735922003698e-05, | |
| "loss": 0.8359, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.3059337703815767, | |
| "grad_norm": 64780.81640625, | |
| "learning_rate": 3.4703311480921164e-05, | |
| "loss": 0.8528, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.3076147251638931, | |
| "grad_norm": 66866.6875, | |
| "learning_rate": 3.461926374180535e-05, | |
| "loss": 0.7795, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.30929567994620943, | |
| "grad_norm": 74536.5078125, | |
| "learning_rate": 3.453521600268953e-05, | |
| "loss": 0.8432, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.3109766347285258, | |
| "grad_norm": 63969.90234375, | |
| "learning_rate": 3.4451168263573714e-05, | |
| "loss": 0.8459, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.31265758951084216, | |
| "grad_norm": 68175.203125, | |
| "learning_rate": 3.43671205244579e-05, | |
| "loss": 0.8437, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.3143385442931585, | |
| "grad_norm": 61941.46484375, | |
| "learning_rate": 3.4283072785342075e-05, | |
| "loss": 0.8182, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.3160194990754749, | |
| "grad_norm": 67861.203125, | |
| "learning_rate": 3.419902504622626e-05, | |
| "loss": 0.8742, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.3177004538577912, | |
| "grad_norm": 66184.25, | |
| "learning_rate": 3.411497730711044e-05, | |
| "loss": 0.8312, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.31938140864010756, | |
| "grad_norm": 63603.37109375, | |
| "learning_rate": 3.4030929567994625e-05, | |
| "loss": 0.8162, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.32106236342242395, | |
| "grad_norm": 76040.2421875, | |
| "learning_rate": 3.394688182887881e-05, | |
| "loss": 0.8555, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.3227433182047403, | |
| "grad_norm": 62280.6328125, | |
| "learning_rate": 3.3862834089762985e-05, | |
| "loss": 0.8086, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.32442427298705667, | |
| "grad_norm": 68005.015625, | |
| "learning_rate": 3.377878635064717e-05, | |
| "loss": 0.7433, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.326105227769373, | |
| "grad_norm": 58576.92578125, | |
| "learning_rate": 3.369473861153135e-05, | |
| "loss": 0.7728, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.32778618255168934, | |
| "grad_norm": 64847.859375, | |
| "learning_rate": 3.3610690872415536e-05, | |
| "loss": 0.7596, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.32946713733400573, | |
| "grad_norm": 63781.38671875, | |
| "learning_rate": 3.352664313329972e-05, | |
| "loss": 0.7308, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.33114809211632207, | |
| "grad_norm": 66446.5859375, | |
| "learning_rate": 3.3442595394183896e-05, | |
| "loss": 0.8058, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.3328290468986384, | |
| "grad_norm": 61722.046875, | |
| "learning_rate": 3.335854765506808e-05, | |
| "loss": 0.869, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.3345100016809548, | |
| "grad_norm": 66245.3359375, | |
| "learning_rate": 3.327449991595226e-05, | |
| "loss": 0.8153, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.33619095646327113, | |
| "grad_norm": 60609.90625, | |
| "learning_rate": 3.3190452176836446e-05, | |
| "loss": 0.8484, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.3378719112455875, | |
| "grad_norm": 70234.8828125, | |
| "learning_rate": 3.310640443772063e-05, | |
| "loss": 0.8249, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.33955286602790385, | |
| "grad_norm": 52655.56640625, | |
| "learning_rate": 3.302235669860481e-05, | |
| "loss": 0.7768, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.3412338208102202, | |
| "grad_norm": 73065.375, | |
| "learning_rate": 3.293830895948899e-05, | |
| "loss": 0.8911, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.3429147755925366, | |
| "grad_norm": 59607.45703125, | |
| "learning_rate": 3.285426122037317e-05, | |
| "loss": 0.7773, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.3445957303748529, | |
| "grad_norm": 64399.9375, | |
| "learning_rate": 3.277021348125736e-05, | |
| "loss": 0.8161, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.34627668515716925, | |
| "grad_norm": 75411.3359375, | |
| "learning_rate": 3.268616574214154e-05, | |
| "loss": 0.7615, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.34795763993948564, | |
| "grad_norm": 69012.03125, | |
| "learning_rate": 3.2602118003025724e-05, | |
| "loss": 0.8245, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.349638594721802, | |
| "grad_norm": 69290.5625, | |
| "learning_rate": 3.25180702639099e-05, | |
| "loss": 0.786, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.35131954950411837, | |
| "grad_norm": 68525.0859375, | |
| "learning_rate": 3.2434022524794084e-05, | |
| "loss": 0.7572, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.3530005042864347, | |
| "grad_norm": 61886.20703125, | |
| "learning_rate": 3.234997478567827e-05, | |
| "loss": 0.8115, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.35468145906875104, | |
| "grad_norm": 73001.8828125, | |
| "learning_rate": 3.226592704656245e-05, | |
| "loss": 0.8336, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.3563624138510674, | |
| "grad_norm": 65161.6484375, | |
| "learning_rate": 3.2181879307446634e-05, | |
| "loss": 0.8124, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.35804336863338376, | |
| "grad_norm": 61001.9140625, | |
| "learning_rate": 3.209783156833082e-05, | |
| "loss": 0.8067, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.3597243234157001, | |
| "grad_norm": 56932.25, | |
| "learning_rate": 3.2013783829214994e-05, | |
| "loss": 0.77, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.3614052781980165, | |
| "grad_norm": 68792.859375, | |
| "learning_rate": 3.192973609009918e-05, | |
| "loss": 0.8217, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.3630862329803328, | |
| "grad_norm": 61329.98828125, | |
| "learning_rate": 3.184568835098336e-05, | |
| "loss": 0.7714, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.36476718776264916, | |
| "grad_norm": 66838.0078125, | |
| "learning_rate": 3.1761640611867545e-05, | |
| "loss": 0.8314, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.36644814254496555, | |
| "grad_norm": 73817.578125, | |
| "learning_rate": 3.167759287275173e-05, | |
| "loss": 0.8413, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.3681290973272819, | |
| "grad_norm": 67156.03125, | |
| "learning_rate": 3.1593545133635905e-05, | |
| "loss": 0.7996, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.3698100521095983, | |
| "grad_norm": 83176.359375, | |
| "learning_rate": 3.150949739452009e-05, | |
| "loss": 0.7875, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.3714910068919146, | |
| "grad_norm": 68843.5859375, | |
| "learning_rate": 3.142544965540427e-05, | |
| "loss": 0.8156, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.37317196167423095, | |
| "grad_norm": 61444.25, | |
| "learning_rate": 3.1341401916288455e-05, | |
| "loss": 0.7747, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.37485291645654734, | |
| "grad_norm": 70228.59375, | |
| "learning_rate": 3.125735417717264e-05, | |
| "loss": 0.8088, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.37653387123886367, | |
| "grad_norm": 56036.578125, | |
| "learning_rate": 3.1173306438056816e-05, | |
| "loss": 0.7834, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.37821482602118, | |
| "grad_norm": 62951.3828125, | |
| "learning_rate": 3.1089258698941e-05, | |
| "loss": 0.7642, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.3798957808034964, | |
| "grad_norm": 66556.96875, | |
| "learning_rate": 3.100521095982518e-05, | |
| "loss": 0.8115, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.38157673558581273, | |
| "grad_norm": 83066.890625, | |
| "learning_rate": 3.0921163220709366e-05, | |
| "loss": 0.8311, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.3832576903681291, | |
| "grad_norm": 68849.5078125, | |
| "learning_rate": 3.083711548159355e-05, | |
| "loss": 0.7935, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.38493864515044546, | |
| "grad_norm": 68215.6953125, | |
| "learning_rate": 3.075306774247773e-05, | |
| "loss": 0.8071, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.3866195999327618, | |
| "grad_norm": 65878.6484375, | |
| "learning_rate": 3.066902000336191e-05, | |
| "loss": 0.9158, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.3883005547150782, | |
| "grad_norm": 59912.58203125, | |
| "learning_rate": 3.058497226424609e-05, | |
| "loss": 0.8217, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.3899815094973945, | |
| "grad_norm": 65707.3828125, | |
| "learning_rate": 3.0500924525130277e-05, | |
| "loss": 0.7693, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.39166246427971085, | |
| "grad_norm": 75819.90625, | |
| "learning_rate": 3.041687678601446e-05, | |
| "loss": 0.8007, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.39334341906202724, | |
| "grad_norm": 66755.734375, | |
| "learning_rate": 3.0332829046898644e-05, | |
| "loss": 0.8416, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.3950243738443436, | |
| "grad_norm": 68420.984375, | |
| "learning_rate": 3.024878130778282e-05, | |
| "loss": 0.8119, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.39670532862665997, | |
| "grad_norm": 70407.578125, | |
| "learning_rate": 3.0164733568667004e-05, | |
| "loss": 0.7907, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.3983862834089763, | |
| "grad_norm": 70415.9453125, | |
| "learning_rate": 3.0080685829551187e-05, | |
| "loss": 0.7508, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.40006723819129264, | |
| "grad_norm": 64331.296875, | |
| "learning_rate": 2.999663809043537e-05, | |
| "loss": 0.7812, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.40174819297360903, | |
| "grad_norm": 55358.15625, | |
| "learning_rate": 2.9912590351319554e-05, | |
| "loss": 0.8044, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.40342914775592537, | |
| "grad_norm": 74381.0078125, | |
| "learning_rate": 2.9828542612203734e-05, | |
| "loss": 0.7862, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.4051101025382417, | |
| "grad_norm": 69503.421875, | |
| "learning_rate": 2.9744494873087914e-05, | |
| "loss": 0.8414, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.4067910573205581, | |
| "grad_norm": 63817.671875, | |
| "learning_rate": 2.9660447133972098e-05, | |
| "loss": 0.8239, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.4084720121028744, | |
| "grad_norm": 64829.07421875, | |
| "learning_rate": 2.957639939485628e-05, | |
| "loss": 0.8409, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.41015296688519076, | |
| "grad_norm": 90130.75, | |
| "learning_rate": 2.9492351655740465e-05, | |
| "loss": 0.7979, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.41183392166750715, | |
| "grad_norm": 65037.46484375, | |
| "learning_rate": 2.9408303916624648e-05, | |
| "loss": 0.7724, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.4135148764498235, | |
| "grad_norm": 74054.9765625, | |
| "learning_rate": 2.9324256177508825e-05, | |
| "loss": 0.9302, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.4151958312321399, | |
| "grad_norm": 65572.390625, | |
| "learning_rate": 2.924020843839301e-05, | |
| "loss": 0.8096, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.4168767860144562, | |
| "grad_norm": 70729.125, | |
| "learning_rate": 2.9156160699277192e-05, | |
| "loss": 0.7749, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.41855774079677255, | |
| "grad_norm": 60806.2265625, | |
| "learning_rate": 2.9072112960161375e-05, | |
| "loss": 0.7964, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.42023869557908894, | |
| "grad_norm": 66842.796875, | |
| "learning_rate": 2.898806522104556e-05, | |
| "loss": 0.7956, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.4219196503614053, | |
| "grad_norm": 65062.66015625, | |
| "learning_rate": 2.8904017481929735e-05, | |
| "loss": 0.765, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.4236006051437216, | |
| "grad_norm": 70468.4921875, | |
| "learning_rate": 2.881996974281392e-05, | |
| "loss": 0.8029, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.425281559926038, | |
| "grad_norm": 63679.37109375, | |
| "learning_rate": 2.8735922003698102e-05, | |
| "loss": 0.8116, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.42696251470835433, | |
| "grad_norm": 79301.3984375, | |
| "learning_rate": 2.8651874264582286e-05, | |
| "loss": 0.8561, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.4286434694906707, | |
| "grad_norm": 66358.953125, | |
| "learning_rate": 2.856782652546647e-05, | |
| "loss": 0.8753, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.43032442427298706, | |
| "grad_norm": 64068.6875, | |
| "learning_rate": 2.848377878635065e-05, | |
| "loss": 0.7874, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.4320053790553034, | |
| "grad_norm": 70842.203125, | |
| "learning_rate": 2.839973104723483e-05, | |
| "loss": 0.8036, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.4336863338376198, | |
| "grad_norm": 69636.0390625, | |
| "learning_rate": 2.8315683308119013e-05, | |
| "loss": 0.8069, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.4353672886199361, | |
| "grad_norm": 89295.2421875, | |
| "learning_rate": 2.8231635569003196e-05, | |
| "loss": 0.8322, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.43704824340225246, | |
| "grad_norm": 70937.7265625, | |
| "learning_rate": 2.814758782988738e-05, | |
| "loss": 0.9161, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.43872919818456885, | |
| "grad_norm": 88488.953125, | |
| "learning_rate": 2.8063540090771563e-05, | |
| "loss": 0.7753, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.4404101529668852, | |
| "grad_norm": 68679.078125, | |
| "learning_rate": 2.797949235165574e-05, | |
| "loss": 0.8489, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.4420911077492016, | |
| "grad_norm": 62188.109375, | |
| "learning_rate": 2.7895444612539924e-05, | |
| "loss": 0.7826, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.4437720625315179, | |
| "grad_norm": 72500.828125, | |
| "learning_rate": 2.7811396873424107e-05, | |
| "loss": 0.8122, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.44545301731383424, | |
| "grad_norm": 61758.7421875, | |
| "learning_rate": 2.772734913430829e-05, | |
| "loss": 0.7645, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.44713397209615063, | |
| "grad_norm": 59873.4375, | |
| "learning_rate": 2.7643301395192474e-05, | |
| "loss": 0.7859, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.44881492687846697, | |
| "grad_norm": 78891.734375, | |
| "learning_rate": 2.755925365607665e-05, | |
| "loss": 0.7696, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.4504958816607833, | |
| "grad_norm": 69256.2578125, | |
| "learning_rate": 2.7475205916960834e-05, | |
| "loss": 0.8161, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.4521768364430997, | |
| "grad_norm": 68757.21875, | |
| "learning_rate": 2.7391158177845018e-05, | |
| "loss": 0.7662, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.45385779122541603, | |
| "grad_norm": 80412.984375, | |
| "learning_rate": 2.73071104387292e-05, | |
| "loss": 0.7875, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.45553874600773236, | |
| "grad_norm": 74670.53125, | |
| "learning_rate": 2.7223062699613385e-05, | |
| "loss": 0.7664, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.45721970079004876, | |
| "grad_norm": 69425.640625, | |
| "learning_rate": 2.7139014960497565e-05, | |
| "loss": 0.8521, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.4589006555723651, | |
| "grad_norm": 70010.1484375, | |
| "learning_rate": 2.7054967221381745e-05, | |
| "loss": 0.7819, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.4605816103546815, | |
| "grad_norm": 76622.5078125, | |
| "learning_rate": 2.6970919482265928e-05, | |
| "loss": 0.8632, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.4622625651369978, | |
| "grad_norm": 63262.62890625, | |
| "learning_rate": 2.6886871743150112e-05, | |
| "loss": 0.7303, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.46394351991931415, | |
| "grad_norm": 63295.68359375, | |
| "learning_rate": 2.6802824004034295e-05, | |
| "loss": 0.773, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.46562447470163054, | |
| "grad_norm": 61289.41015625, | |
| "learning_rate": 2.6718776264918475e-05, | |
| "loss": 0.7776, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.4673054294839469, | |
| "grad_norm": 75931.3671875, | |
| "learning_rate": 2.6634728525802655e-05, | |
| "loss": 0.826, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.4689863842662632, | |
| "grad_norm": 69288.890625, | |
| "learning_rate": 2.655068078668684e-05, | |
| "loss": 0.7978, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.4706673390485796, | |
| "grad_norm": 61528.40234375, | |
| "learning_rate": 2.6466633047571022e-05, | |
| "loss": 0.7891, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.47234829383089594, | |
| "grad_norm": 67145.6796875, | |
| "learning_rate": 2.6382585308455206e-05, | |
| "loss": 0.7483, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.47402924861321233, | |
| "grad_norm": 62958.0703125, | |
| "learning_rate": 2.629853756933939e-05, | |
| "loss": 0.8407, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.47571020339552866, | |
| "grad_norm": 71565.0078125, | |
| "learning_rate": 2.621448983022357e-05, | |
| "loss": 0.8093, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.477391158177845, | |
| "grad_norm": 70295.3125, | |
| "learning_rate": 2.613044209110775e-05, | |
| "loss": 0.8328, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.4790721129601614, | |
| "grad_norm": 66831.65625, | |
| "learning_rate": 2.6046394351991933e-05, | |
| "loss": 0.7844, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.4807530677424777, | |
| "grad_norm": 62661.7421875, | |
| "learning_rate": 2.5962346612876116e-05, | |
| "loss": 0.7925, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.48243402252479406, | |
| "grad_norm": 61856.66796875, | |
| "learning_rate": 2.58782988737603e-05, | |
| "loss": 0.7488, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.48411497730711045, | |
| "grad_norm": 73633.4765625, | |
| "learning_rate": 2.579425113464448e-05, | |
| "loss": 0.7881, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.4857959320894268, | |
| "grad_norm": 69115.28125, | |
| "learning_rate": 2.571020339552866e-05, | |
| "loss": 0.8175, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.4874768868717432, | |
| "grad_norm": 61891.3828125, | |
| "learning_rate": 2.5626155656412843e-05, | |
| "loss": 0.8784, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.4891578416540595, | |
| "grad_norm": 77769.9921875, | |
| "learning_rate": 2.5542107917297027e-05, | |
| "loss": 0.8684, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.49083879643637585, | |
| "grad_norm": 68048.734375, | |
| "learning_rate": 2.545806017818121e-05, | |
| "loss": 0.7853, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.49251975121869224, | |
| "grad_norm": 61355.94140625, | |
| "learning_rate": 2.537401243906539e-05, | |
| "loss": 0.7845, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.49420070600100857, | |
| "grad_norm": 69287.953125, | |
| "learning_rate": 2.528996469994957e-05, | |
| "loss": 0.8007, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.4958816607833249, | |
| "grad_norm": 68851.8359375, | |
| "learning_rate": 2.5205916960833754e-05, | |
| "loss": 0.8159, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.4975626155656413, | |
| "grad_norm": 69794.859375, | |
| "learning_rate": 2.5121869221717938e-05, | |
| "loss": 0.7958, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.49924357034795763, | |
| "grad_norm": 73403.1953125, | |
| "learning_rate": 2.503782148260212e-05, | |
| "loss": 0.7855, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.500924525130274, | |
| "grad_norm": 74224.1328125, | |
| "learning_rate": 2.49537737434863e-05, | |
| "loss": 0.7728, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.5026054799125903, | |
| "grad_norm": 63414.859375, | |
| "learning_rate": 2.4869726004370485e-05, | |
| "loss": 0.8105, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.5042864346949067, | |
| "grad_norm": 77378.109375, | |
| "learning_rate": 2.4785678265254668e-05, | |
| "loss": 0.8636, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5059673894772231, | |
| "grad_norm": 70468.2734375, | |
| "learning_rate": 2.4701630526138848e-05, | |
| "loss": 0.7242, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.5076483442595394, | |
| "grad_norm": 60809.28515625, | |
| "learning_rate": 2.461758278702303e-05, | |
| "loss": 0.8078, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.5093292990418558, | |
| "grad_norm": 51373.625, | |
| "learning_rate": 2.4533535047907215e-05, | |
| "loss": 0.8366, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.5110102538241721, | |
| "grad_norm": 59719.765625, | |
| "learning_rate": 2.4449487308791395e-05, | |
| "loss": 0.8665, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.5126912086064885, | |
| "grad_norm": 64678.41015625, | |
| "learning_rate": 2.436543956967558e-05, | |
| "loss": 0.7776, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.5143721633888049, | |
| "grad_norm": 74790.40625, | |
| "learning_rate": 2.4281391830559762e-05, | |
| "loss": 0.7729, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.5160531181711212, | |
| "grad_norm": 71427.921875, | |
| "learning_rate": 2.4197344091443942e-05, | |
| "loss": 0.7956, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.5177340729534375, | |
| "grad_norm": 66004.3671875, | |
| "learning_rate": 2.4113296352328126e-05, | |
| "loss": 0.7588, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.5194150277357539, | |
| "grad_norm": 64083.765625, | |
| "learning_rate": 2.4029248613212306e-05, | |
| "loss": 0.7931, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.5210959825180702, | |
| "grad_norm": 54726.55078125, | |
| "learning_rate": 2.394520087409649e-05, | |
| "loss": 0.7163, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.5227769373003867, | |
| "grad_norm": 73129.3125, | |
| "learning_rate": 2.3861153134980673e-05, | |
| "loss": 0.8364, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.524457892082703, | |
| "grad_norm": 62834.08984375, | |
| "learning_rate": 2.3777105395864853e-05, | |
| "loss": 0.7431, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.5261388468650193, | |
| "grad_norm": 61612.23828125, | |
| "learning_rate": 2.3693057656749036e-05, | |
| "loss": 0.7975, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.5278198016473357, | |
| "grad_norm": 71137.6640625, | |
| "learning_rate": 2.3609009917633216e-05, | |
| "loss": 0.783, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.529500756429652, | |
| "grad_norm": 63406.10546875, | |
| "learning_rate": 2.35249621785174e-05, | |
| "loss": 0.824, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.5311817112119684, | |
| "grad_norm": 64795.73828125, | |
| "learning_rate": 2.3440914439401583e-05, | |
| "loss": 0.8715, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.5328626659942848, | |
| "grad_norm": 59525.79296875, | |
| "learning_rate": 2.3356866700285763e-05, | |
| "loss": 0.7934, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.5345436207766011, | |
| "grad_norm": 72914.4609375, | |
| "learning_rate": 2.3272818961169947e-05, | |
| "loss": 0.7243, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.5362245755589174, | |
| "grad_norm": 62966.078125, | |
| "learning_rate": 2.318877122205413e-05, | |
| "loss": 0.712, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.5379055303412338, | |
| "grad_norm": 66647.0546875, | |
| "learning_rate": 2.310472348293831e-05, | |
| "loss": 0.7674, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.5395864851235502, | |
| "grad_norm": 65908.0703125, | |
| "learning_rate": 2.3020675743822494e-05, | |
| "loss": 0.8484, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.5412674399058666, | |
| "grad_norm": 60565.56640625, | |
| "learning_rate": 2.2936628004706674e-05, | |
| "loss": 0.7844, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.5429483946881829, | |
| "grad_norm": 67560.3828125, | |
| "learning_rate": 2.2852580265590857e-05, | |
| "loss": 0.9071, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.5446293494704992, | |
| "grad_norm": 66084.265625, | |
| "learning_rate": 2.276853252647504e-05, | |
| "loss": 0.7323, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.5463103042528156, | |
| "grad_norm": 78780.078125, | |
| "learning_rate": 2.268448478735922e-05, | |
| "loss": 0.9259, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.5479912590351319, | |
| "grad_norm": 67810.703125, | |
| "learning_rate": 2.2600437048243404e-05, | |
| "loss": 0.7813, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.5496722138174484, | |
| "grad_norm": 62699.97265625, | |
| "learning_rate": 2.2516389309127588e-05, | |
| "loss": 0.7831, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.5513531685997647, | |
| "grad_norm": 66553.1640625, | |
| "learning_rate": 2.2432341570011768e-05, | |
| "loss": 0.8185, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.553034123382081, | |
| "grad_norm": 60711.96875, | |
| "learning_rate": 2.234829383089595e-05, | |
| "loss": 0.7551, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.5547150781643974, | |
| "grad_norm": 64588.28515625, | |
| "learning_rate": 2.226424609178013e-05, | |
| "loss": 0.6919, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.5563960329467137, | |
| "grad_norm": 73545.9765625, | |
| "learning_rate": 2.2180198352664315e-05, | |
| "loss": 0.7834, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.5580769877290301, | |
| "grad_norm": 66515.1796875, | |
| "learning_rate": 2.20961506135485e-05, | |
| "loss": 0.8135, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.5597579425113465, | |
| "grad_norm": 66021.4140625, | |
| "learning_rate": 2.201210287443268e-05, | |
| "loss": 0.7767, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.5614388972936628, | |
| "grad_norm": 72027.734375, | |
| "learning_rate": 2.1928055135316862e-05, | |
| "loss": 0.7623, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.5631198520759791, | |
| "grad_norm": 69724.7109375, | |
| "learning_rate": 2.1844007396201046e-05, | |
| "loss": 0.8539, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.5648008068582955, | |
| "grad_norm": 70565.2890625, | |
| "learning_rate": 2.1759959657085226e-05, | |
| "loss": 0.8081, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.5664817616406118, | |
| "grad_norm": 69982.8515625, | |
| "learning_rate": 2.167591191796941e-05, | |
| "loss": 0.766, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.5681627164229283, | |
| "grad_norm": 73730.5859375, | |
| "learning_rate": 2.159186417885359e-05, | |
| "loss": 0.7858, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.5698436712052446, | |
| "grad_norm": 68504.3359375, | |
| "learning_rate": 2.1507816439737773e-05, | |
| "loss": 0.7774, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.5715246259875609, | |
| "grad_norm": 68371.71875, | |
| "learning_rate": 2.1423768700621956e-05, | |
| "loss": 0.7397, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.5732055807698773, | |
| "grad_norm": 62352.87890625, | |
| "learning_rate": 2.1339720961506136e-05, | |
| "loss": 0.7727, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.5748865355521936, | |
| "grad_norm": 67821.3671875, | |
| "learning_rate": 2.125567322239032e-05, | |
| "loss": 0.7494, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.57656749033451, | |
| "grad_norm": 75150.0703125, | |
| "learning_rate": 2.1171625483274503e-05, | |
| "loss": 0.738, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.5782484451168264, | |
| "grad_norm": 71489.2109375, | |
| "learning_rate": 2.1087577744158683e-05, | |
| "loss": 0.7666, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.5799293998991427, | |
| "grad_norm": 77000.265625, | |
| "learning_rate": 2.1003530005042867e-05, | |
| "loss": 0.8051, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.581610354681459, | |
| "grad_norm": 63612.04296875, | |
| "learning_rate": 2.0919482265927047e-05, | |
| "loss": 0.7408, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.5832913094637754, | |
| "grad_norm": 65412.390625, | |
| "learning_rate": 2.083543452681123e-05, | |
| "loss": 0.7585, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.5849722642460918, | |
| "grad_norm": 63992.859375, | |
| "learning_rate": 2.0751386787695414e-05, | |
| "loss": 0.8036, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.5866532190284082, | |
| "grad_norm": 67541.6015625, | |
| "learning_rate": 2.0667339048579594e-05, | |
| "loss": 0.7835, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.5883341738107245, | |
| "grad_norm": 87275.59375, | |
| "learning_rate": 2.0583291309463777e-05, | |
| "loss": 0.843, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5900151285930408, | |
| "grad_norm": 62353.66015625, | |
| "learning_rate": 2.0499243570347957e-05, | |
| "loss": 0.8229, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.5916960833753572, | |
| "grad_norm": 76160.890625, | |
| "learning_rate": 2.041519583123214e-05, | |
| "loss": 0.839, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.5933770381576735, | |
| "grad_norm": 71393.7109375, | |
| "learning_rate": 2.0331148092116324e-05, | |
| "loss": 0.7136, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.59505799293999, | |
| "grad_norm": 70572.75, | |
| "learning_rate": 2.0247100353000504e-05, | |
| "loss": 0.7987, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.5967389477223063, | |
| "grad_norm": 66998.0703125, | |
| "learning_rate": 2.0163052613884688e-05, | |
| "loss": 0.7874, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.5984199025046226, | |
| "grad_norm": 53749.26953125, | |
| "learning_rate": 2.007900487476887e-05, | |
| "loss": 0.767, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.600100857286939, | |
| "grad_norm": 71350.6328125, | |
| "learning_rate": 1.999495713565305e-05, | |
| "loss": 0.8136, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.6017818120692553, | |
| "grad_norm": 64163.25390625, | |
| "learning_rate": 1.9910909396537235e-05, | |
| "loss": 0.7565, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.6034627668515717, | |
| "grad_norm": 61004.05078125, | |
| "learning_rate": 1.9826861657421415e-05, | |
| "loss": 0.7502, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.6051437216338881, | |
| "grad_norm": 66495.4375, | |
| "learning_rate": 1.97428139183056e-05, | |
| "loss": 0.7485, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.6068246764162044, | |
| "grad_norm": 62442.1015625, | |
| "learning_rate": 1.9658766179189782e-05, | |
| "loss": 0.7484, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.6085056311985207, | |
| "grad_norm": 66803.453125, | |
| "learning_rate": 1.9574718440073962e-05, | |
| "loss": 0.8459, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.6101865859808371, | |
| "grad_norm": 71394.6875, | |
| "learning_rate": 1.9490670700958145e-05, | |
| "loss": 0.8108, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.6118675407631534, | |
| "grad_norm": 72386.0859375, | |
| "learning_rate": 1.940662296184233e-05, | |
| "loss": 0.8039, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.6135484955454699, | |
| "grad_norm": 69155.015625, | |
| "learning_rate": 1.932257522272651e-05, | |
| "loss": 0.823, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.6152294503277862, | |
| "grad_norm": 60343.54296875, | |
| "learning_rate": 1.9238527483610693e-05, | |
| "loss": 0.7284, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.6169104051101025, | |
| "grad_norm": 70258.7734375, | |
| "learning_rate": 1.9154479744494873e-05, | |
| "loss": 0.7925, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.6185913598924189, | |
| "grad_norm": 68638.21875, | |
| "learning_rate": 1.9070432005379056e-05, | |
| "loss": 0.7802, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.6202723146747352, | |
| "grad_norm": 67208.890625, | |
| "learning_rate": 1.898638426626324e-05, | |
| "loss": 0.7921, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.6219532694570516, | |
| "grad_norm": 56271.5703125, | |
| "learning_rate": 1.890233652714742e-05, | |
| "loss": 0.7752, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.623634224239368, | |
| "grad_norm": 84835.21875, | |
| "learning_rate": 1.8818288788031603e-05, | |
| "loss": 0.8012, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.6253151790216843, | |
| "grad_norm": 58683.70703125, | |
| "learning_rate": 1.8734241048915787e-05, | |
| "loss": 0.7547, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.6269961338040007, | |
| "grad_norm": 66791.8828125, | |
| "learning_rate": 1.8650193309799967e-05, | |
| "loss": 0.782, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.628677088586317, | |
| "grad_norm": 60903.69140625, | |
| "learning_rate": 1.856614557068415e-05, | |
| "loss": 0.7304, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.6303580433686334, | |
| "grad_norm": 63727.53515625, | |
| "learning_rate": 1.848209783156833e-05, | |
| "loss": 0.7591, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.6320389981509498, | |
| "grad_norm": 65615.4140625, | |
| "learning_rate": 1.8398050092452514e-05, | |
| "loss": 0.7483, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.6337199529332661, | |
| "grad_norm": 72326.921875, | |
| "learning_rate": 1.8314002353336697e-05, | |
| "loss": 0.7795, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.6354009077155824, | |
| "grad_norm": 65443.81640625, | |
| "learning_rate": 1.8229954614220877e-05, | |
| "loss": 0.8154, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.6370818624978988, | |
| "grad_norm": 66016.125, | |
| "learning_rate": 1.814590687510506e-05, | |
| "loss": 0.8109, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.6387628172802151, | |
| "grad_norm": 70201.703125, | |
| "learning_rate": 1.8061859135989244e-05, | |
| "loss": 0.8039, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.6404437720625316, | |
| "grad_norm": 72543.0859375, | |
| "learning_rate": 1.7977811396873424e-05, | |
| "loss": 0.8201, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.6421247268448479, | |
| "grad_norm": 65741.5703125, | |
| "learning_rate": 1.7893763657757608e-05, | |
| "loss": 0.7919, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.6438056816271642, | |
| "grad_norm": 67635.5078125, | |
| "learning_rate": 1.7809715918641788e-05, | |
| "loss": 0.8174, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.6454866364094806, | |
| "grad_norm": 64830.48046875, | |
| "learning_rate": 1.772566817952597e-05, | |
| "loss": 0.7375, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.6471675911917969, | |
| "grad_norm": 66888.5234375, | |
| "learning_rate": 1.7641620440410155e-05, | |
| "loss": 0.7547, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.6488485459741133, | |
| "grad_norm": 67550.5234375, | |
| "learning_rate": 1.7557572701294335e-05, | |
| "loss": 0.7532, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.6505295007564297, | |
| "grad_norm": 66258.5859375, | |
| "learning_rate": 1.747352496217852e-05, | |
| "loss": 0.7623, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.652210455538746, | |
| "grad_norm": 70212.1875, | |
| "learning_rate": 1.7389477223062702e-05, | |
| "loss": 0.8022, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.6538914103210623, | |
| "grad_norm": 69388.6171875, | |
| "learning_rate": 1.7305429483946882e-05, | |
| "loss": 0.7672, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.6555723651033787, | |
| "grad_norm": 61498.984375, | |
| "learning_rate": 1.7221381744831065e-05, | |
| "loss": 0.7243, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.657253319885695, | |
| "grad_norm": 73348.640625, | |
| "learning_rate": 1.7137334005715245e-05, | |
| "loss": 0.7699, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.6589342746680115, | |
| "grad_norm": 76186.703125, | |
| "learning_rate": 1.705328626659943e-05, | |
| "loss": 0.8293, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 0.6606152294503278, | |
| "grad_norm": 72204.1171875, | |
| "learning_rate": 1.6969238527483612e-05, | |
| "loss": 0.8325, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 0.6622961842326441, | |
| "grad_norm": 68690.1328125, | |
| "learning_rate": 1.6885190788367792e-05, | |
| "loss": 0.8174, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 0.6639771390149605, | |
| "grad_norm": 69304.6015625, | |
| "learning_rate": 1.6801143049251976e-05, | |
| "loss": 0.7395, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.6656580937972768, | |
| "grad_norm": 67933.7890625, | |
| "learning_rate": 1.6717095310136156e-05, | |
| "loss": 0.7976, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 0.6673390485795933, | |
| "grad_norm": 74347.6953125, | |
| "learning_rate": 1.663304757102034e-05, | |
| "loss": 0.8003, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 0.6690200033619096, | |
| "grad_norm": 74672.625, | |
| "learning_rate": 1.6548999831904523e-05, | |
| "loss": 0.7809, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 0.6707009581442259, | |
| "grad_norm": 69611.8984375, | |
| "learning_rate": 1.6464952092788703e-05, | |
| "loss": 0.7884, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 0.6723819129265423, | |
| "grad_norm": 61687.546875, | |
| "learning_rate": 1.6380904353672887e-05, | |
| "loss": 0.7726, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.6740628677088586, | |
| "grad_norm": 57029.6953125, | |
| "learning_rate": 1.629685661455707e-05, | |
| "loss": 0.7537, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 0.675743822491175, | |
| "grad_norm": 62528.2265625, | |
| "learning_rate": 1.621280887544125e-05, | |
| "loss": 0.7542, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 0.6774247772734914, | |
| "grad_norm": 69657.46875, | |
| "learning_rate": 1.6128761136325434e-05, | |
| "loss": 0.7846, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 0.6791057320558077, | |
| "grad_norm": 61733.50390625, | |
| "learning_rate": 1.6044713397209614e-05, | |
| "loss": 0.7342, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 0.680786686838124, | |
| "grad_norm": 63571.60546875, | |
| "learning_rate": 1.5960665658093797e-05, | |
| "loss": 0.7046, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.6824676416204404, | |
| "grad_norm": 80437.828125, | |
| "learning_rate": 1.587661791897798e-05, | |
| "loss": 0.8413, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 0.6841485964027567, | |
| "grad_norm": 107205.921875, | |
| "learning_rate": 1.579257017986216e-05, | |
| "loss": 0.8052, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 0.6858295511850732, | |
| "grad_norm": 71237.375, | |
| "learning_rate": 1.5708522440746344e-05, | |
| "loss": 0.7729, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 0.6875105059673895, | |
| "grad_norm": 61227.4296875, | |
| "learning_rate": 1.5624474701630528e-05, | |
| "loss": 0.7972, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 0.6891914607497058, | |
| "grad_norm": 81108.9296875, | |
| "learning_rate": 1.5540426962514708e-05, | |
| "loss": 0.7173, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.6908724155320222, | |
| "grad_norm": 58387.5703125, | |
| "learning_rate": 1.545637922339889e-05, | |
| "loss": 0.8341, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 0.6925533703143385, | |
| "grad_norm": 77154.796875, | |
| "learning_rate": 1.537233148428307e-05, | |
| "loss": 0.7459, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 0.694234325096655, | |
| "grad_norm": 66626.140625, | |
| "learning_rate": 1.5288283745167255e-05, | |
| "loss": 0.7965, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 0.6959152798789713, | |
| "grad_norm": 65993.8671875, | |
| "learning_rate": 1.5204236006051437e-05, | |
| "loss": 0.8077, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 0.6975962346612876, | |
| "grad_norm": 65909.390625, | |
| "learning_rate": 1.512018826693562e-05, | |
| "loss": 0.7933, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.699277189443604, | |
| "grad_norm": 73940.3671875, | |
| "learning_rate": 1.5036140527819803e-05, | |
| "loss": 0.7937, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 0.7009581442259203, | |
| "grad_norm": 80516.8203125, | |
| "learning_rate": 1.4952092788703984e-05, | |
| "loss": 0.7921, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 0.7026390990082367, | |
| "grad_norm": 62047.8046875, | |
| "learning_rate": 1.4868045049588167e-05, | |
| "loss": 0.7405, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 0.7043200537905531, | |
| "grad_norm": 68648.3828125, | |
| "learning_rate": 1.478399731047235e-05, | |
| "loss": 0.7956, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 0.7060010085728694, | |
| "grad_norm": 74798.203125, | |
| "learning_rate": 1.469994957135653e-05, | |
| "loss": 0.7026, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.7076819633551857, | |
| "grad_norm": 65563.265625, | |
| "learning_rate": 1.4615901832240714e-05, | |
| "loss": 0.7846, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 0.7093629181375021, | |
| "grad_norm": 76866.96875, | |
| "learning_rate": 1.4531854093124898e-05, | |
| "loss": 0.829, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 0.7110438729198184, | |
| "grad_norm": 64904.796875, | |
| "learning_rate": 1.4447806354009078e-05, | |
| "loss": 0.8053, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 0.7127248277021349, | |
| "grad_norm": 72151.6484375, | |
| "learning_rate": 1.4363758614893261e-05, | |
| "loss": 0.7583, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 0.7144057824844512, | |
| "grad_norm": 57045.8203125, | |
| "learning_rate": 1.4279710875777441e-05, | |
| "loss": 0.7584, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.7160867372667675, | |
| "grad_norm": 74145.6953125, | |
| "learning_rate": 1.4195663136661625e-05, | |
| "loss": 0.7637, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 0.7177676920490839, | |
| "grad_norm": 63434.96875, | |
| "learning_rate": 1.4111615397545808e-05, | |
| "loss": 0.8278, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 0.7194486468314002, | |
| "grad_norm": 70511.5546875, | |
| "learning_rate": 1.4027567658429988e-05, | |
| "loss": 0.7512, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 0.7211296016137166, | |
| "grad_norm": 65691.6640625, | |
| "learning_rate": 1.3943519919314172e-05, | |
| "loss": 0.7509, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 0.722810556396033, | |
| "grad_norm": 72238.515625, | |
| "learning_rate": 1.3859472180198355e-05, | |
| "loss": 0.7977, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.7244915111783493, | |
| "grad_norm": 79115.5625, | |
| "learning_rate": 1.3775424441082535e-05, | |
| "loss": 0.8284, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 0.7261724659606656, | |
| "grad_norm": 64790.1484375, | |
| "learning_rate": 1.3691376701966719e-05, | |
| "loss": 0.7124, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 0.727853420742982, | |
| "grad_norm": 75220.65625, | |
| "learning_rate": 1.3607328962850899e-05, | |
| "loss": 0.7687, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 0.7295343755252983, | |
| "grad_norm": 60745.61328125, | |
| "learning_rate": 1.3523281223735082e-05, | |
| "loss": 0.7283, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 0.7312153303076148, | |
| "grad_norm": 54197.08203125, | |
| "learning_rate": 1.3439233484619266e-05, | |
| "loss": 0.768, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.7328962850899311, | |
| "grad_norm": 61422.91015625, | |
| "learning_rate": 1.3355185745503446e-05, | |
| "loss": 0.7324, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 0.7345772398722474, | |
| "grad_norm": 68666.9765625, | |
| "learning_rate": 1.327113800638763e-05, | |
| "loss": 0.8131, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 0.7362581946545638, | |
| "grad_norm": 54904.734375, | |
| "learning_rate": 1.3187090267271813e-05, | |
| "loss": 0.751, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 0.7379391494368801, | |
| "grad_norm": 74403.234375, | |
| "learning_rate": 1.3103042528155993e-05, | |
| "loss": 0.7564, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 0.7396201042191966, | |
| "grad_norm": 83832.09375, | |
| "learning_rate": 1.3018994789040176e-05, | |
| "loss": 0.7588, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.7413010590015129, | |
| "grad_norm": 65015.13671875, | |
| "learning_rate": 1.2934947049924356e-05, | |
| "loss": 0.7333, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 0.7429820137838292, | |
| "grad_norm": 70224.0234375, | |
| "learning_rate": 1.285089931080854e-05, | |
| "loss": 0.7437, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 0.7446629685661456, | |
| "grad_norm": 65735.6015625, | |
| "learning_rate": 1.2766851571692723e-05, | |
| "loss": 0.7311, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 0.7463439233484619, | |
| "grad_norm": 64739.11328125, | |
| "learning_rate": 1.2682803832576903e-05, | |
| "loss": 0.752, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 0.7480248781307783, | |
| "grad_norm": 84499.734375, | |
| "learning_rate": 1.2598756093461087e-05, | |
| "loss": 0.7085, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.7497058329130947, | |
| "grad_norm": 58072.46484375, | |
| "learning_rate": 1.251470835434527e-05, | |
| "loss": 0.7869, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 0.751386787695411, | |
| "grad_norm": 60343.64453125, | |
| "learning_rate": 1.243066061522945e-05, | |
| "loss": 0.748, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 0.7530677424777273, | |
| "grad_norm": 73976.90625, | |
| "learning_rate": 1.2346612876113634e-05, | |
| "loss": 0.8125, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 0.7547486972600437, | |
| "grad_norm": 83017.234375, | |
| "learning_rate": 1.2262565136997816e-05, | |
| "loss": 0.7647, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 0.75642965204236, | |
| "grad_norm": 63707.15625, | |
| "learning_rate": 1.2178517397881997e-05, | |
| "loss": 0.689, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.7581106068246765, | |
| "grad_norm": 62471.03125, | |
| "learning_rate": 1.2094469658766181e-05, | |
| "loss": 0.7642, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 0.7597915616069928, | |
| "grad_norm": 71938.2421875, | |
| "learning_rate": 1.2010421919650363e-05, | |
| "loss": 0.7018, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 0.7614725163893091, | |
| "grad_norm": 62137.34375, | |
| "learning_rate": 1.1926374180534544e-05, | |
| "loss": 0.7742, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 0.7631534711716255, | |
| "grad_norm": 74779.3515625, | |
| "learning_rate": 1.1842326441418726e-05, | |
| "loss": 0.76, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 0.7648344259539418, | |
| "grad_norm": 65167.984375, | |
| "learning_rate": 1.175827870230291e-05, | |
| "loss": 0.7814, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.7665153807362582, | |
| "grad_norm": 64530.6015625, | |
| "learning_rate": 1.1674230963187092e-05, | |
| "loss": 0.7496, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 0.7681963355185746, | |
| "grad_norm": 69663.7109375, | |
| "learning_rate": 1.1590183224071273e-05, | |
| "loss": 0.794, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 0.7698772903008909, | |
| "grad_norm": 58575.296875, | |
| "learning_rate": 1.1506135484955455e-05, | |
| "loss": 0.7438, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 0.7715582450832073, | |
| "grad_norm": 60153.59375, | |
| "learning_rate": 1.1422087745839639e-05, | |
| "loss": 0.7246, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 0.7732391998655236, | |
| "grad_norm": 63029.4140625, | |
| "learning_rate": 1.133804000672382e-05, | |
| "loss": 0.7449, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.7749201546478399, | |
| "grad_norm": 70431.859375, | |
| "learning_rate": 1.1253992267608002e-05, | |
| "loss": 0.8162, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 0.7766011094301564, | |
| "grad_norm": 56848.18359375, | |
| "learning_rate": 1.1169944528492184e-05, | |
| "loss": 0.8272, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 0.7782820642124727, | |
| "grad_norm": 63638.44921875, | |
| "learning_rate": 1.1085896789376367e-05, | |
| "loss": 0.6735, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 0.779963018994789, | |
| "grad_norm": 81583.1328125, | |
| "learning_rate": 1.1001849050260549e-05, | |
| "loss": 0.7195, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 0.7816439737771054, | |
| "grad_norm": 65090.41796875, | |
| "learning_rate": 1.0917801311144731e-05, | |
| "loss": 0.7348, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.7833249285594217, | |
| "grad_norm": 81298.7265625, | |
| "learning_rate": 1.0833753572028913e-05, | |
| "loss": 0.7036, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 0.7850058833417382, | |
| "grad_norm": 72416.609375, | |
| "learning_rate": 1.0749705832913096e-05, | |
| "loss": 0.759, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 0.7866868381240545, | |
| "grad_norm": 62808.859375, | |
| "learning_rate": 1.0665658093797278e-05, | |
| "loss": 0.8043, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 0.7883677929063708, | |
| "grad_norm": 57125.68359375, | |
| "learning_rate": 1.058161035468146e-05, | |
| "loss": 0.695, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 0.7900487476886872, | |
| "grad_norm": 70024.90625, | |
| "learning_rate": 1.0497562615565642e-05, | |
| "loss": 0.8026, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.7917297024710035, | |
| "grad_norm": 67909.5, | |
| "learning_rate": 1.0413514876449825e-05, | |
| "loss": 0.7139, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 0.7934106572533199, | |
| "grad_norm": 63104.14453125, | |
| "learning_rate": 1.0329467137334007e-05, | |
| "loss": 0.7578, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 0.7950916120356363, | |
| "grad_norm": 59612.7578125, | |
| "learning_rate": 1.0245419398218189e-05, | |
| "loss": 0.8141, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 0.7967725668179526, | |
| "grad_norm": 61766.25, | |
| "learning_rate": 1.016137165910237e-05, | |
| "loss": 0.7066, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 0.798453521600269, | |
| "grad_norm": 61020.91015625, | |
| "learning_rate": 1.0077323919986554e-05, | |
| "loss": 0.7129, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.8001344763825853, | |
| "grad_norm": 73282.953125, | |
| "learning_rate": 9.993276180870736e-06, | |
| "loss": 0.7443, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 0.8018154311649016, | |
| "grad_norm": 61251.37890625, | |
| "learning_rate": 9.909228441754917e-06, | |
| "loss": 0.7202, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 0.8034963859472181, | |
| "grad_norm": 57658.30859375, | |
| "learning_rate": 9.825180702639099e-06, | |
| "loss": 0.7261, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 0.8051773407295344, | |
| "grad_norm": 58032.66796875, | |
| "learning_rate": 9.741132963523283e-06, | |
| "loss": 0.7772, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 0.8068582955118507, | |
| "grad_norm": 67425.0703125, | |
| "learning_rate": 9.657085224407464e-06, | |
| "loss": 0.75, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.8085392502941671, | |
| "grad_norm": 61734.875, | |
| "learning_rate": 9.573037485291646e-06, | |
| "loss": 0.7046, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 0.8102202050764834, | |
| "grad_norm": 68259.4453125, | |
| "learning_rate": 9.48898974617583e-06, | |
| "loss": 0.7702, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 0.8119011598587998, | |
| "grad_norm": 62056.20703125, | |
| "learning_rate": 9.404942007060011e-06, | |
| "loss": 0.7043, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 0.8135821146411162, | |
| "grad_norm": 65011.296875, | |
| "learning_rate": 9.320894267944193e-06, | |
| "loss": 0.7163, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 0.8152630694234325, | |
| "grad_norm": 74655.8203125, | |
| "learning_rate": 9.236846528828375e-06, | |
| "loss": 0.8321, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.8169440242057489, | |
| "grad_norm": 75672.0703125, | |
| "learning_rate": 9.152798789712558e-06, | |
| "loss": 0.7922, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 0.8186249789880652, | |
| "grad_norm": 61180.29296875, | |
| "learning_rate": 9.06875105059674e-06, | |
| "loss": 0.7159, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 0.8203059337703815, | |
| "grad_norm": 87931.515625, | |
| "learning_rate": 8.984703311480922e-06, | |
| "loss": 0.6781, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 0.821986888552698, | |
| "grad_norm": 73175.0390625, | |
| "learning_rate": 8.900655572365104e-06, | |
| "loss": 0.7326, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 0.8236678433350143, | |
| "grad_norm": 63801.890625, | |
| "learning_rate": 8.816607833249287e-06, | |
| "loss": 0.7975, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.8253487981173306, | |
| "grad_norm": 80039.4375, | |
| "learning_rate": 8.732560094133469e-06, | |
| "loss": 0.6984, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 0.827029752899647, | |
| "grad_norm": 71785.921875, | |
| "learning_rate": 8.64851235501765e-06, | |
| "loss": 0.7481, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 0.8287107076819633, | |
| "grad_norm": 72605.5234375, | |
| "learning_rate": 8.564464615901833e-06, | |
| "loss": 0.7317, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 0.8303916624642798, | |
| "grad_norm": 72289.1171875, | |
| "learning_rate": 8.480416876786016e-06, | |
| "loss": 0.752, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 0.8320726172465961, | |
| "grad_norm": 69316.015625, | |
| "learning_rate": 8.396369137670198e-06, | |
| "loss": 0.722, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.8337535720289124, | |
| "grad_norm": 68252.515625, | |
| "learning_rate": 8.31232139855438e-06, | |
| "loss": 0.7674, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 0.8354345268112288, | |
| "grad_norm": 66645.7421875, | |
| "learning_rate": 8.228273659438561e-06, | |
| "loss": 0.7386, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 0.8371154815935451, | |
| "grad_norm": 64145.96875, | |
| "learning_rate": 8.144225920322745e-06, | |
| "loss": 0.7213, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 0.8387964363758615, | |
| "grad_norm": 62728.8125, | |
| "learning_rate": 8.060178181206927e-06, | |
| "loss": 0.6535, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 0.8404773911581779, | |
| "grad_norm": 72365.625, | |
| "learning_rate": 7.976130442091108e-06, | |
| "loss": 0.6833, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.8421583459404942, | |
| "grad_norm": 62805.015625, | |
| "learning_rate": 7.89208270297529e-06, | |
| "loss": 0.8574, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 0.8438393007228105, | |
| "grad_norm": 65521.5703125, | |
| "learning_rate": 7.808034963859474e-06, | |
| "loss": 0.7408, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 0.8455202555051269, | |
| "grad_norm": 67055.6171875, | |
| "learning_rate": 7.723987224743655e-06, | |
| "loss": 0.6515, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 0.8472012102874432, | |
| "grad_norm": 56498.43359375, | |
| "learning_rate": 7.639939485627837e-06, | |
| "loss": 0.8291, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 0.8488821650697597, | |
| "grad_norm": 68116.96875, | |
| "learning_rate": 7.555891746512018e-06, | |
| "loss": 0.7886, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.850563119852076, | |
| "grad_norm": 64178.546875, | |
| "learning_rate": 7.471844007396202e-06, | |
| "loss": 0.7746, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 0.8522440746343923, | |
| "grad_norm": 63573.09375, | |
| "learning_rate": 7.387796268280383e-06, | |
| "loss": 0.7769, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 0.8539250294167087, | |
| "grad_norm": 57395.2421875, | |
| "learning_rate": 7.303748529164565e-06, | |
| "loss": 0.7334, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 0.855605984199025, | |
| "grad_norm": 74868.375, | |
| "learning_rate": 7.219700790048749e-06, | |
| "loss": 0.7575, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 0.8572869389813415, | |
| "grad_norm": 69290.953125, | |
| "learning_rate": 7.1356530509329304e-06, | |
| "loss": 0.6967, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.8589678937636578, | |
| "grad_norm": 59573.5234375, | |
| "learning_rate": 7.051605311817112e-06, | |
| "loss": 0.8442, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 0.8606488485459741, | |
| "grad_norm": 61021.09375, | |
| "learning_rate": 6.967557572701294e-06, | |
| "loss": 0.7069, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 0.8623298033282905, | |
| "grad_norm": 71908.3359375, | |
| "learning_rate": 6.8835098335854775e-06, | |
| "loss": 0.6972, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 0.8640107581106068, | |
| "grad_norm": 71261.375, | |
| "learning_rate": 6.799462094469659e-06, | |
| "loss": 0.7308, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 0.8656917128929231, | |
| "grad_norm": 76563.328125, | |
| "learning_rate": 6.715414355353841e-06, | |
| "loss": 0.8391, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.8673726676752396, | |
| "grad_norm": 73876.8515625, | |
| "learning_rate": 6.631366616238023e-06, | |
| "loss": 0.6593, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 0.8690536224575559, | |
| "grad_norm": 58473.9140625, | |
| "learning_rate": 6.547318877122206e-06, | |
| "loss": 0.7766, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 0.8707345772398722, | |
| "grad_norm": 61153.5390625, | |
| "learning_rate": 6.463271138006388e-06, | |
| "loss": 0.7679, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 0.8724155320221886, | |
| "grad_norm": 58491.39453125, | |
| "learning_rate": 6.37922339889057e-06, | |
| "loss": 0.7115, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 0.8740964868045049, | |
| "grad_norm": 64786.53125, | |
| "learning_rate": 6.295175659774752e-06, | |
| "loss": 0.6976, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.8757774415868214, | |
| "grad_norm": 68189.5546875, | |
| "learning_rate": 6.211127920658934e-06, | |
| "loss": 0.8054, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 0.8774583963691377, | |
| "grad_norm": 66427.9921875, | |
| "learning_rate": 6.127080181543117e-06, | |
| "loss": 0.7512, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 0.879139351151454, | |
| "grad_norm": 57290.70703125, | |
| "learning_rate": 6.043032442427299e-06, | |
| "loss": 0.6927, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 0.8808203059337704, | |
| "grad_norm": 72431.390625, | |
| "learning_rate": 5.958984703311481e-06, | |
| "loss": 0.8137, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 0.8825012607160867, | |
| "grad_norm": 70580.7265625, | |
| "learning_rate": 5.874936964195663e-06, | |
| "loss": 0.7712, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.8841822154984031, | |
| "grad_norm": 73078.3046875, | |
| "learning_rate": 5.790889225079846e-06, | |
| "loss": 0.6761, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 0.8858631702807195, | |
| "grad_norm": 78010.0703125, | |
| "learning_rate": 5.7068414859640274e-06, | |
| "loss": 0.7715, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 0.8875441250630358, | |
| "grad_norm": 66805.8671875, | |
| "learning_rate": 5.62279374684821e-06, | |
| "loss": 0.7067, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 0.8892250798453522, | |
| "grad_norm": 68583.328125, | |
| "learning_rate": 5.538746007732392e-06, | |
| "loss": 0.7403, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 0.8909060346276685, | |
| "grad_norm": 80909.6796875, | |
| "learning_rate": 5.4546982686165745e-06, | |
| "loss": 0.7469, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.8925869894099848, | |
| "grad_norm": 62168.92578125, | |
| "learning_rate": 5.370650529500756e-06, | |
| "loss": 0.7284, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 0.8942679441923013, | |
| "grad_norm": 70689.0546875, | |
| "learning_rate": 5.286602790384939e-06, | |
| "loss": 0.7804, | |
| "step": 5320 | |
| }, | |
| { | |
| "epoch": 0.8959488989746176, | |
| "grad_norm": 70834.4609375, | |
| "learning_rate": 5.202555051269121e-06, | |
| "loss": 0.7504, | |
| "step": 5330 | |
| }, | |
| { | |
| "epoch": 0.8976298537569339, | |
| "grad_norm": 72244.8828125, | |
| "learning_rate": 5.118507312153303e-06, | |
| "loss": 0.6909, | |
| "step": 5340 | |
| }, | |
| { | |
| "epoch": 0.8993108085392503, | |
| "grad_norm": 68406.7734375, | |
| "learning_rate": 5.034459573037485e-06, | |
| "loss": 0.7554, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.9009917633215666, | |
| "grad_norm": 59602.8515625, | |
| "learning_rate": 4.950411833921668e-06, | |
| "loss": 0.7673, | |
| "step": 5360 | |
| }, | |
| { | |
| "epoch": 0.902672718103883, | |
| "grad_norm": 61461.9375, | |
| "learning_rate": 4.8663640948058495e-06, | |
| "loss": 0.7082, | |
| "step": 5370 | |
| }, | |
| { | |
| "epoch": 0.9043536728861994, | |
| "grad_norm": 64041.15625, | |
| "learning_rate": 4.782316355690032e-06, | |
| "loss": 0.682, | |
| "step": 5380 | |
| }, | |
| { | |
| "epoch": 0.9060346276685157, | |
| "grad_norm": 67531.3046875, | |
| "learning_rate": 4.698268616574215e-06, | |
| "loss": 0.8184, | |
| "step": 5390 | |
| }, | |
| { | |
| "epoch": 0.9077155824508321, | |
| "grad_norm": 64709.15625, | |
| "learning_rate": 4.6142208774583965e-06, | |
| "loss": 0.7433, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.9093965372331484, | |
| "grad_norm": 73473.203125, | |
| "learning_rate": 4.530173138342579e-06, | |
| "loss": 0.7386, | |
| "step": 5410 | |
| }, | |
| { | |
| "epoch": 0.9110774920154647, | |
| "grad_norm": 66414.84375, | |
| "learning_rate": 4.446125399226761e-06, | |
| "loss": 0.7839, | |
| "step": 5420 | |
| }, | |
| { | |
| "epoch": 0.9127584467977812, | |
| "grad_norm": 73395.6015625, | |
| "learning_rate": 4.3620776601109435e-06, | |
| "loss": 0.7699, | |
| "step": 5430 | |
| }, | |
| { | |
| "epoch": 0.9144394015800975, | |
| "grad_norm": 60129.15625, | |
| "learning_rate": 4.278029920995125e-06, | |
| "loss": 0.7111, | |
| "step": 5440 | |
| }, | |
| { | |
| "epoch": 0.9161203563624138, | |
| "grad_norm": 81137.3203125, | |
| "learning_rate": 4.193982181879308e-06, | |
| "loss": 0.7419, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 0.9178013111447302, | |
| "grad_norm": 69321.53125, | |
| "learning_rate": 4.10993444276349e-06, | |
| "loss": 0.6546, | |
| "step": 5460 | |
| }, | |
| { | |
| "epoch": 0.9194822659270465, | |
| "grad_norm": 73445.609375, | |
| "learning_rate": 4.025886703647672e-06, | |
| "loss": 0.7571, | |
| "step": 5470 | |
| }, | |
| { | |
| "epoch": 0.921163220709363, | |
| "grad_norm": 55211.56640625, | |
| "learning_rate": 3.941838964531854e-06, | |
| "loss": 0.7361, | |
| "step": 5480 | |
| }, | |
| { | |
| "epoch": 0.9228441754916793, | |
| "grad_norm": 60575.71484375, | |
| "learning_rate": 3.857791225416037e-06, | |
| "loss": 0.7854, | |
| "step": 5490 | |
| }, | |
| { | |
| "epoch": 0.9245251302739956, | |
| "grad_norm": 70186.140625, | |
| "learning_rate": 3.7737434863002185e-06, | |
| "loss": 0.8255, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.926206085056312, | |
| "grad_norm": 73100.078125, | |
| "learning_rate": 3.689695747184401e-06, | |
| "loss": 0.7469, | |
| "step": 5510 | |
| }, | |
| { | |
| "epoch": 0.9278870398386283, | |
| "grad_norm": 75805.078125, | |
| "learning_rate": 3.605648008068583e-06, | |
| "loss": 0.7616, | |
| "step": 5520 | |
| }, | |
| { | |
| "epoch": 0.9295679946209447, | |
| "grad_norm": 77849.140625, | |
| "learning_rate": 3.5216002689527655e-06, | |
| "loss": 0.7562, | |
| "step": 5530 | |
| }, | |
| { | |
| "epoch": 0.9312489494032611, | |
| "grad_norm": 68207.265625, | |
| "learning_rate": 3.4375525298369473e-06, | |
| "loss": 0.6813, | |
| "step": 5540 | |
| }, | |
| { | |
| "epoch": 0.9329299041855774, | |
| "grad_norm": 71734.1640625, | |
| "learning_rate": 3.35350479072113e-06, | |
| "loss": 0.6668, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 0.9346108589678938, | |
| "grad_norm": 80375.4375, | |
| "learning_rate": 3.2694570516053117e-06, | |
| "loss": 0.7542, | |
| "step": 5560 | |
| }, | |
| { | |
| "epoch": 0.9362918137502101, | |
| "grad_norm": 76297.75, | |
| "learning_rate": 3.1854093124894943e-06, | |
| "loss": 0.7539, | |
| "step": 5570 | |
| }, | |
| { | |
| "epoch": 0.9379727685325264, | |
| "grad_norm": 59915.84765625, | |
| "learning_rate": 3.1013615733736765e-06, | |
| "loss": 0.6965, | |
| "step": 5580 | |
| }, | |
| { | |
| "epoch": 0.9396537233148429, | |
| "grad_norm": 66482.078125, | |
| "learning_rate": 3.0173138342578587e-06, | |
| "loss": 0.7416, | |
| "step": 5590 | |
| }, | |
| { | |
| "epoch": 0.9413346780971592, | |
| "grad_norm": 73498.7578125, | |
| "learning_rate": 2.933266095142041e-06, | |
| "loss": 0.7404, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.9430156328794755, | |
| "grad_norm": 56677.17578125, | |
| "learning_rate": 2.849218356026223e-06, | |
| "loss": 0.7199, | |
| "step": 5610 | |
| }, | |
| { | |
| "epoch": 0.9446965876617919, | |
| "grad_norm": 63067.734375, | |
| "learning_rate": 2.7651706169104053e-06, | |
| "loss": 0.738, | |
| "step": 5620 | |
| }, | |
| { | |
| "epoch": 0.9463775424441082, | |
| "grad_norm": 65398.22265625, | |
| "learning_rate": 2.6811228777945875e-06, | |
| "loss": 0.7177, | |
| "step": 5630 | |
| }, | |
| { | |
| "epoch": 0.9480584972264247, | |
| "grad_norm": 56372.125, | |
| "learning_rate": 2.5970751386787698e-06, | |
| "loss": 0.7594, | |
| "step": 5640 | |
| }, | |
| { | |
| "epoch": 0.949739452008741, | |
| "grad_norm": 66133.2109375, | |
| "learning_rate": 2.513027399562952e-06, | |
| "loss": 0.755, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 0.9514204067910573, | |
| "grad_norm": 73855.5859375, | |
| "learning_rate": 2.428979660447134e-06, | |
| "loss": 0.7286, | |
| "step": 5660 | |
| }, | |
| { | |
| "epoch": 0.9531013615733737, | |
| "grad_norm": 80979.5703125, | |
| "learning_rate": 2.3449319213313164e-06, | |
| "loss": 0.7508, | |
| "step": 5670 | |
| }, | |
| { | |
| "epoch": 0.95478231635569, | |
| "grad_norm": 59096.21484375, | |
| "learning_rate": 2.2608841822154986e-06, | |
| "loss": 0.7273, | |
| "step": 5680 | |
| }, | |
| { | |
| "epoch": 0.9564632711380063, | |
| "grad_norm": 73199.34375, | |
| "learning_rate": 2.1768364430996808e-06, | |
| "loss": 0.8463, | |
| "step": 5690 | |
| }, | |
| { | |
| "epoch": 0.9581442259203228, | |
| "grad_norm": 70557.265625, | |
| "learning_rate": 2.092788703983863e-06, | |
| "loss": 0.7518, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.9598251807026391, | |
| "grad_norm": 64123.15234375, | |
| "learning_rate": 2.008740964868045e-06, | |
| "loss": 0.6743, | |
| "step": 5710 | |
| }, | |
| { | |
| "epoch": 0.9615061354849554, | |
| "grad_norm": 70495.03125, | |
| "learning_rate": 1.9246932257522274e-06, | |
| "loss": 0.7179, | |
| "step": 5720 | |
| }, | |
| { | |
| "epoch": 0.9631870902672718, | |
| "grad_norm": 62604.77734375, | |
| "learning_rate": 1.8406454866364096e-06, | |
| "loss": 0.7073, | |
| "step": 5730 | |
| }, | |
| { | |
| "epoch": 0.9648680450495881, | |
| "grad_norm": 69528.75, | |
| "learning_rate": 1.7565977475205918e-06, | |
| "loss": 0.7738, | |
| "step": 5740 | |
| }, | |
| { | |
| "epoch": 0.9665489998319046, | |
| "grad_norm": 64379.51953125, | |
| "learning_rate": 1.672550008404774e-06, | |
| "loss": 0.774, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.9682299546142209, | |
| "grad_norm": 70070.359375, | |
| "learning_rate": 1.5885022692889562e-06, | |
| "loss": 0.7902, | |
| "step": 5760 | |
| }, | |
| { | |
| "epoch": 0.9699109093965372, | |
| "grad_norm": 75525.7265625, | |
| "learning_rate": 1.5044545301731386e-06, | |
| "loss": 0.754, | |
| "step": 5770 | |
| }, | |
| { | |
| "epoch": 0.9715918641788536, | |
| "grad_norm": 72437.4375, | |
| "learning_rate": 1.4204067910573208e-06, | |
| "loss": 0.7122, | |
| "step": 5780 | |
| }, | |
| { | |
| "epoch": 0.9732728189611699, | |
| "grad_norm": 55177.8125, | |
| "learning_rate": 1.3363590519415028e-06, | |
| "loss": 0.7083, | |
| "step": 5790 | |
| }, | |
| { | |
| "epoch": 0.9749537737434864, | |
| "grad_norm": 74118.140625, | |
| "learning_rate": 1.252311312825685e-06, | |
| "loss": 0.7124, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.9766347285258027, | |
| "grad_norm": 75796.015625, | |
| "learning_rate": 1.1682635737098672e-06, | |
| "loss": 0.7563, | |
| "step": 5810 | |
| }, | |
| { | |
| "epoch": 0.978315683308119, | |
| "grad_norm": 66329.46875, | |
| "learning_rate": 1.0842158345940494e-06, | |
| "loss": 0.7379, | |
| "step": 5820 | |
| }, | |
| { | |
| "epoch": 0.9799966380904354, | |
| "grad_norm": 71971.1796875, | |
| "learning_rate": 1.0001680954782316e-06, | |
| "loss": 0.7237, | |
| "step": 5830 | |
| }, | |
| { | |
| "epoch": 0.9816775928727517, | |
| "grad_norm": 70406.546875, | |
| "learning_rate": 9.161203563624139e-07, | |
| "loss": 0.7518, | |
| "step": 5840 | |
| }, | |
| { | |
| "epoch": 0.983358547655068, | |
| "grad_norm": 65358.0, | |
| "learning_rate": 8.320726172465961e-07, | |
| "loss": 0.7263, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 0.9850395024373845, | |
| "grad_norm": 56430.7265625, | |
| "learning_rate": 7.480248781307783e-07, | |
| "loss": 0.7174, | |
| "step": 5860 | |
| }, | |
| { | |
| "epoch": 0.9867204572197008, | |
| "grad_norm": 69264.21875, | |
| "learning_rate": 6.639771390149606e-07, | |
| "loss": 0.7782, | |
| "step": 5870 | |
| }, | |
| { | |
| "epoch": 0.9884014120020171, | |
| "grad_norm": 60301.59765625, | |
| "learning_rate": 5.799293998991428e-07, | |
| "loss": 0.7487, | |
| "step": 5880 | |
| }, | |
| { | |
| "epoch": 0.9900823667843335, | |
| "grad_norm": 66026.0078125, | |
| "learning_rate": 4.95881660783325e-07, | |
| "loss": 0.7565, | |
| "step": 5890 | |
| }, | |
| { | |
| "epoch": 0.9917633215666498, | |
| "grad_norm": 63975.43359375, | |
| "learning_rate": 4.1183392166750716e-07, | |
| "loss": 0.7349, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.9934442763489663, | |
| "grad_norm": 72585.8046875, | |
| "learning_rate": 3.2778618255168936e-07, | |
| "loss": 0.7754, | |
| "step": 5910 | |
| }, | |
| { | |
| "epoch": 0.9951252311312826, | |
| "grad_norm": 67646.4296875, | |
| "learning_rate": 2.4373844343587156e-07, | |
| "loss": 0.7622, | |
| "step": 5920 | |
| }, | |
| { | |
| "epoch": 0.9968061859135989, | |
| "grad_norm": 74414.9375, | |
| "learning_rate": 1.596907043200538e-07, | |
| "loss": 0.7312, | |
| "step": 5930 | |
| }, | |
| { | |
| "epoch": 0.9984871406959153, | |
| "grad_norm": 69075.203125, | |
| "learning_rate": 7.5642965204236e-08, | |
| "loss": 0.7293, | |
| "step": 5940 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 5949, | |
| "total_flos": 9.074524143432499e+17, | |
| "train_loss": 0.8014113598860578, | |
| "train_runtime": 82603.7739, | |
| "train_samples_per_second": 0.864, | |
| "train_steps_per_second": 0.072 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 5949, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 5949, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.074524143432499e+17, | |
| "train_batch_size": 12, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |