{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 36246, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0008276775368316504, "grad_norm": 46.14799118041992, "learning_rate": 2.4827586206896555e-08, "loss": 1.5047, "step": 10 }, { "epoch": 0.0016553550736633008, "grad_norm": 22.840229034423828, "learning_rate": 5.2413793103448285e-08, "loss": 1.3711, "step": 20 }, { "epoch": 0.002483032610494951, "grad_norm": 45.4863395690918, "learning_rate": 8e-08, "loss": 1.5035, "step": 30 }, { "epoch": 0.0033107101473266016, "grad_norm": 30.114967346191406, "learning_rate": 1.0758620689655174e-07, "loss": 1.4755, "step": 40 }, { "epoch": 0.004138387684158252, "grad_norm": 33.25872802734375, "learning_rate": 1.3517241379310345e-07, "loss": 1.3472, "step": 50 }, { "epoch": 0.004966065220989902, "grad_norm": 14.948714256286621, "learning_rate": 1.6275862068965518e-07, "loss": 1.5429, "step": 60 }, { "epoch": 0.005793742757821553, "grad_norm": 23.4176082611084, "learning_rate": 1.903448275862069e-07, "loss": 1.3873, "step": 70 }, { "epoch": 0.006621420294653203, "grad_norm": 12.504230499267578, "learning_rate": 2.1793103448275863e-07, "loss": 1.4693, "step": 80 }, { "epoch": 0.007449097831484853, "grad_norm": 30.348501205444336, "learning_rate": 2.455172413793104e-07, "loss": 1.4577, "step": 90 }, { "epoch": 0.008276775368316504, "grad_norm": 12.569012641906738, "learning_rate": 2.731034482758621e-07, "loss": 1.4487, "step": 100 }, { "epoch": 0.009104452905148155, "grad_norm": 22.93501091003418, "learning_rate": 3.006896551724138e-07, "loss": 1.424, "step": 110 }, { "epoch": 0.009932130441979804, "grad_norm": 18.028295516967773, "learning_rate": 3.282758620689656e-07, "loss": 1.2583, "step": 120 }, { "epoch": 0.010759807978811455, "grad_norm": 13.857125282287598, "learning_rate": 3.558620689655173e-07, "loss": 1.2549, "step": 130 }, { "epoch": 0.011587485515643106, "grad_norm": 9.900038719177246, "learning_rate": 3.8344827586206895e-07, "loss": 1.4044, "step": 140 }, { "epoch": 0.012415163052474755, "grad_norm": 20.69013214111328, "learning_rate": 4.110344827586207e-07, "loss": 1.2974, "step": 150 }, { "epoch": 0.013242840589306406, "grad_norm": 21.640762329101562, "learning_rate": 4.386206896551724e-07, "loss": 1.242, "step": 160 }, { "epoch": 0.014070518126138057, "grad_norm": 13.101654052734375, "learning_rate": 4.662068965517242e-07, "loss": 1.2213, "step": 170 }, { "epoch": 0.014898195662969706, "grad_norm": 15.014242172241211, "learning_rate": 4.937931034482759e-07, "loss": 1.1303, "step": 180 }, { "epoch": 0.015725873199801357, "grad_norm": 10.959694862365723, "learning_rate": 5.213793103448275e-07, "loss": 1.2214, "step": 190 }, { "epoch": 0.01655355073663301, "grad_norm": 9.056197166442871, "learning_rate": 5.489655172413793e-07, "loss": 1.2172, "step": 200 }, { "epoch": 0.01738122827346466, "grad_norm": 22.233816146850586, "learning_rate": 5.765517241379311e-07, "loss": 1.1887, "step": 210 }, { "epoch": 0.01820890581029631, "grad_norm": 8.905389785766602, "learning_rate": 6.041379310344828e-07, "loss": 1.2706, "step": 220 }, { "epoch": 0.019036583347127958, "grad_norm": 8.299859046936035, "learning_rate": 6.317241379310346e-07, "loss": 1.2712, "step": 230 }, { "epoch": 0.01986426088395961, "grad_norm": 17.16954231262207, "learning_rate": 6.593103448275863e-07, "loss": 1.1182, "step": 240 }, { "epoch": 0.02069193842079126, "grad_norm": 8.578152656555176, "learning_rate": 6.86896551724138e-07, "loss": 1.2233, "step": 250 }, { "epoch": 0.02151961595762291, "grad_norm": 8.546585083007812, "learning_rate": 7.144827586206897e-07, "loss": 1.1661, "step": 260 }, { "epoch": 0.02234729349445456, "grad_norm": 14.001766204833984, "learning_rate": 7.420689655172414e-07, "loss": 1.1541, "step": 270 }, { "epoch": 0.023174971031286212, "grad_norm": 9.023934364318848, "learning_rate": 7.696551724137932e-07, "loss": 1.13, "step": 280 }, { "epoch": 0.02400264856811786, "grad_norm": 17.447298049926758, "learning_rate": 7.97241379310345e-07, "loss": 1.0678, "step": 290 }, { "epoch": 0.02483032610494951, "grad_norm": 20.659399032592773, "learning_rate": 8.248275862068966e-07, "loss": 1.1353, "step": 300 }, { "epoch": 0.02565800364178116, "grad_norm": 7.847487449645996, "learning_rate": 8.524137931034483e-07, "loss": 1.1126, "step": 310 }, { "epoch": 0.026485681178612813, "grad_norm": 19.57134437561035, "learning_rate": 8.8e-07, "loss": 1.1319, "step": 320 }, { "epoch": 0.027313358715444463, "grad_norm": 7.312618732452393, "learning_rate": 9.075862068965518e-07, "loss": 1.0699, "step": 330 }, { "epoch": 0.028141036252276114, "grad_norm": 8.64587688446045, "learning_rate": 9.351724137931036e-07, "loss": 1.1577, "step": 340 }, { "epoch": 0.028968713789107765, "grad_norm": 7.389246463775635, "learning_rate": 9.627586206896552e-07, "loss": 1.1216, "step": 350 }, { "epoch": 0.029796391325939413, "grad_norm": 12.635276794433594, "learning_rate": 9.903448275862069e-07, "loss": 1.0112, "step": 360 }, { "epoch": 0.030624068862771064, "grad_norm": 11.252161026000977, "learning_rate": 1.0179310344827588e-06, "loss": 1.0233, "step": 370 }, { "epoch": 0.031451746399602715, "grad_norm": 9.154858589172363, "learning_rate": 1.0455172413793104e-06, "loss": 1.1328, "step": 380 }, { "epoch": 0.03227942393643436, "grad_norm": 13.546175003051758, "learning_rate": 1.0731034482758623e-06, "loss": 1.0774, "step": 390 }, { "epoch": 0.03310710147326602, "grad_norm": 13.653828620910645, "learning_rate": 1.100689655172414e-06, "loss": 0.9911, "step": 400 }, { "epoch": 0.033934779010097664, "grad_norm": 13.791787147521973, "learning_rate": 1.1282758620689656e-06, "loss": 1.0003, "step": 410 }, { "epoch": 0.03476245654692932, "grad_norm": 14.163640975952148, "learning_rate": 1.1558620689655173e-06, "loss": 0.8495, "step": 420 }, { "epoch": 0.035590134083760966, "grad_norm": 7.913918972015381, "learning_rate": 1.1834482758620691e-06, "loss": 0.9708, "step": 430 }, { "epoch": 0.03641781162059262, "grad_norm": 9.651756286621094, "learning_rate": 1.2110344827586208e-06, "loss": 0.9748, "step": 440 }, { "epoch": 0.03724548915742427, "grad_norm": 9.967205047607422, "learning_rate": 1.2386206896551724e-06, "loss": 0.9856, "step": 450 }, { "epoch": 0.038073166694255915, "grad_norm": 9.666905403137207, "learning_rate": 1.2662068965517243e-06, "loss": 0.9186, "step": 460 }, { "epoch": 0.03890084423108757, "grad_norm": 8.290987014770508, "learning_rate": 1.293793103448276e-06, "loss": 0.8474, "step": 470 }, { "epoch": 0.03972852176791922, "grad_norm": 6.986677169799805, "learning_rate": 1.3213793103448278e-06, "loss": 0.9851, "step": 480 }, { "epoch": 0.04055619930475087, "grad_norm": 8.501372337341309, "learning_rate": 1.3489655172413793e-06, "loss": 1.0601, "step": 490 }, { "epoch": 0.04138387684158252, "grad_norm": 14.146003723144531, "learning_rate": 1.3765517241379311e-06, "loss": 1.0291, "step": 500 }, { "epoch": 0.042211554378414166, "grad_norm": 7.238130569458008, "learning_rate": 1.404137931034483e-06, "loss": 1.0196, "step": 510 }, { "epoch": 0.04303923191524582, "grad_norm": 9.992795944213867, "learning_rate": 1.4317241379310345e-06, "loss": 1.0543, "step": 520 }, { "epoch": 0.04386690945207747, "grad_norm": 12.808954238891602, "learning_rate": 1.4593103448275863e-06, "loss": 1.0102, "step": 530 }, { "epoch": 0.04469458698890912, "grad_norm": 7.4756693840026855, "learning_rate": 1.486896551724138e-06, "loss": 0.9486, "step": 540 }, { "epoch": 0.04552226452574077, "grad_norm": 7.882749557495117, "learning_rate": 1.5144827586206899e-06, "loss": 1.0513, "step": 550 }, { "epoch": 0.046349942062572425, "grad_norm": 7.399966239929199, "learning_rate": 1.5420689655172415e-06, "loss": 1.0466, "step": 560 }, { "epoch": 0.04717761959940407, "grad_norm": 7.492778301239014, "learning_rate": 1.5696551724137932e-06, "loss": 0.8725, "step": 570 }, { "epoch": 0.04800529713623572, "grad_norm": 8.505664825439453, "learning_rate": 1.597241379310345e-06, "loss": 0.8351, "step": 580 }, { "epoch": 0.048832974673067374, "grad_norm": 7.636898994445801, "learning_rate": 1.6248275862068965e-06, "loss": 0.948, "step": 590 }, { "epoch": 0.04966065220989902, "grad_norm": 7.210017681121826, "learning_rate": 1.6524137931034484e-06, "loss": 1.0359, "step": 600 }, { "epoch": 0.050488329746730676, "grad_norm": 6.870676040649414, "learning_rate": 1.6800000000000002e-06, "loss": 1.0095, "step": 610 }, { "epoch": 0.05131600728356232, "grad_norm": 7.416876792907715, "learning_rate": 1.7075862068965519e-06, "loss": 0.9256, "step": 620 }, { "epoch": 0.05214368482039398, "grad_norm": 6.512162208557129, "learning_rate": 1.7351724137931035e-06, "loss": 0.9974, "step": 630 }, { "epoch": 0.052971362357225625, "grad_norm": 13.495314598083496, "learning_rate": 1.7627586206896552e-06, "loss": 0.8688, "step": 640 }, { "epoch": 0.05379903989405727, "grad_norm": 9.676804542541504, "learning_rate": 1.790344827586207e-06, "loss": 0.8644, "step": 650 }, { "epoch": 0.05462671743088893, "grad_norm": 17.121936798095703, "learning_rate": 1.817931034482759e-06, "loss": 0.9485, "step": 660 }, { "epoch": 0.055454394967720574, "grad_norm": 11.457856178283691, "learning_rate": 1.8455172413793104e-06, "loss": 0.9568, "step": 670 }, { "epoch": 0.05628207250455223, "grad_norm": 8.24860954284668, "learning_rate": 1.8731034482758622e-06, "loss": 0.9834, "step": 680 }, { "epoch": 0.057109750041383876, "grad_norm": 12.339548110961914, "learning_rate": 1.900689655172414e-06, "loss": 0.885, "step": 690 }, { "epoch": 0.05793742757821553, "grad_norm": 9.935588836669922, "learning_rate": 1.9282758620689656e-06, "loss": 0.9331, "step": 700 }, { "epoch": 0.05876510511504718, "grad_norm": 11.100628852844238, "learning_rate": 1.9558620689655174e-06, "loss": 0.8285, "step": 710 }, { "epoch": 0.059592782651878826, "grad_norm": 18.9801025390625, "learning_rate": 1.983448275862069e-06, "loss": 0.8181, "step": 720 }, { "epoch": 0.06042046018871048, "grad_norm": 7.354071140289307, "learning_rate": 2.0110344827586207e-06, "loss": 0.8242, "step": 730 }, { "epoch": 0.06124813772554213, "grad_norm": 6.890077114105225, "learning_rate": 2.0386206896551726e-06, "loss": 0.935, "step": 740 }, { "epoch": 0.06207581526237378, "grad_norm": 8.745278358459473, "learning_rate": 2.0662068965517245e-06, "loss": 0.913, "step": 750 }, { "epoch": 0.06290349279920543, "grad_norm": 14.381141662597656, "learning_rate": 2.093793103448276e-06, "loss": 0.9398, "step": 760 }, { "epoch": 0.06373117033603708, "grad_norm": 8.542009353637695, "learning_rate": 2.121379310344828e-06, "loss": 0.838, "step": 770 }, { "epoch": 0.06455884787286872, "grad_norm": 8.157075881958008, "learning_rate": 2.1489655172413797e-06, "loss": 0.7613, "step": 780 }, { "epoch": 0.06538652540970039, "grad_norm": 7.913341522216797, "learning_rate": 2.176551724137931e-06, "loss": 0.768, "step": 790 }, { "epoch": 0.06621420294653203, "grad_norm": 7.685206890106201, "learning_rate": 2.204137931034483e-06, "loss": 0.967, "step": 800 }, { "epoch": 0.06704188048336368, "grad_norm": 13.355488777160645, "learning_rate": 2.231724137931035e-06, "loss": 0.7974, "step": 810 }, { "epoch": 0.06786955802019533, "grad_norm": 11.190116882324219, "learning_rate": 2.2593103448275863e-06, "loss": 0.7706, "step": 820 }, { "epoch": 0.06869723555702698, "grad_norm": 17.367341995239258, "learning_rate": 2.286896551724138e-06, "loss": 0.8575, "step": 830 }, { "epoch": 0.06952491309385864, "grad_norm": 8.927794456481934, "learning_rate": 2.3144827586206896e-06, "loss": 0.9357, "step": 840 }, { "epoch": 0.07035259063069028, "grad_norm": 7.562526226043701, "learning_rate": 2.3420689655172415e-06, "loss": 0.8572, "step": 850 }, { "epoch": 0.07118026816752193, "grad_norm": 9.388384819030762, "learning_rate": 2.3696551724137933e-06, "loss": 0.8834, "step": 860 }, { "epoch": 0.07200794570435358, "grad_norm": 8.42138957977295, "learning_rate": 2.397241379310345e-06, "loss": 0.8682, "step": 870 }, { "epoch": 0.07283562324118524, "grad_norm": 7.1926960945129395, "learning_rate": 2.4248275862068967e-06, "loss": 0.9079, "step": 880 }, { "epoch": 0.07366330077801689, "grad_norm": 11.91021728515625, "learning_rate": 2.4524137931034485e-06, "loss": 0.751, "step": 890 }, { "epoch": 0.07449097831484854, "grad_norm": 6.640119552612305, "learning_rate": 2.4800000000000004e-06, "loss": 0.93, "step": 900 }, { "epoch": 0.07531865585168018, "grad_norm": 6.442651271820068, "learning_rate": 2.507586206896552e-06, "loss": 0.8505, "step": 910 }, { "epoch": 0.07614633338851183, "grad_norm": 7.823519706726074, "learning_rate": 2.5351724137931037e-06, "loss": 0.9394, "step": 920 }, { "epoch": 0.07697401092534349, "grad_norm": 9.239322662353516, "learning_rate": 2.5627586206896556e-06, "loss": 0.7493, "step": 930 }, { "epoch": 0.07780168846217514, "grad_norm": 8.130383491516113, "learning_rate": 2.590344827586207e-06, "loss": 0.932, "step": 940 }, { "epoch": 0.07862936599900679, "grad_norm": 15.20906925201416, "learning_rate": 2.617931034482759e-06, "loss": 0.8832, "step": 950 }, { "epoch": 0.07945704353583843, "grad_norm": 10.066596984863281, "learning_rate": 2.6455172413793108e-06, "loss": 0.8958, "step": 960 }, { "epoch": 0.08028472107267008, "grad_norm": 7.820455074310303, "learning_rate": 2.673103448275862e-06, "loss": 0.9423, "step": 970 }, { "epoch": 0.08111239860950174, "grad_norm": 7.842801094055176, "learning_rate": 2.7006896551724137e-06, "loss": 0.8284, "step": 980 }, { "epoch": 0.08194007614633339, "grad_norm": 10.371053695678711, "learning_rate": 2.728275862068966e-06, "loss": 0.9373, "step": 990 }, { "epoch": 0.08276775368316504, "grad_norm": 6.259485244750977, "learning_rate": 2.7558620689655174e-06, "loss": 0.8678, "step": 1000 }, { "epoch": 0.08359543121999669, "grad_norm": 7.114963531494141, "learning_rate": 2.783448275862069e-06, "loss": 0.787, "step": 1010 }, { "epoch": 0.08442310875682833, "grad_norm": 11.261567115783691, "learning_rate": 2.811034482758621e-06, "loss": 0.926, "step": 1020 }, { "epoch": 0.08525078629366, "grad_norm": 7.193630218505859, "learning_rate": 2.8386206896551726e-06, "loss": 0.8513, "step": 1030 }, { "epoch": 0.08607846383049164, "grad_norm": 7.794323444366455, "learning_rate": 2.8662068965517244e-06, "loss": 0.762, "step": 1040 }, { "epoch": 0.08690614136732329, "grad_norm": 9.509486198425293, "learning_rate": 2.8937931034482763e-06, "loss": 0.9228, "step": 1050 }, { "epoch": 0.08773381890415494, "grad_norm": 6.6723785400390625, "learning_rate": 2.9213793103448278e-06, "loss": 0.7977, "step": 1060 }, { "epoch": 0.0885614964409866, "grad_norm": 9.796627044677734, "learning_rate": 2.9489655172413796e-06, "loss": 0.8301, "step": 1070 }, { "epoch": 0.08938917397781825, "grad_norm": 9.115704536437988, "learning_rate": 2.976551724137931e-06, "loss": 0.7999, "step": 1080 }, { "epoch": 0.09021685151464989, "grad_norm": 9.4700288772583, "learning_rate": 3.004137931034483e-06, "loss": 0.8432, "step": 1090 }, { "epoch": 0.09104452905148154, "grad_norm": 8.827855110168457, "learning_rate": 3.031724137931035e-06, "loss": 0.8557, "step": 1100 }, { "epoch": 0.09187220658831319, "grad_norm": 8.324784278869629, "learning_rate": 3.0593103448275863e-06, "loss": 0.8462, "step": 1110 }, { "epoch": 0.09269988412514485, "grad_norm": 7.058176040649414, "learning_rate": 3.0868965517241385e-06, "loss": 0.8119, "step": 1120 }, { "epoch": 0.0935275616619765, "grad_norm": 7.699660301208496, "learning_rate": 3.11448275862069e-06, "loss": 0.9005, "step": 1130 }, { "epoch": 0.09435523919880814, "grad_norm": 7.573885917663574, "learning_rate": 3.1420689655172414e-06, "loss": 0.8208, "step": 1140 }, { "epoch": 0.09518291673563979, "grad_norm": 11.120659828186035, "learning_rate": 3.1696551724137937e-06, "loss": 0.8158, "step": 1150 }, { "epoch": 0.09601059427247144, "grad_norm": 10.248250007629395, "learning_rate": 3.197241379310345e-06, "loss": 0.8485, "step": 1160 }, { "epoch": 0.0968382718093031, "grad_norm": 6.736319541931152, "learning_rate": 3.2248275862068966e-06, "loss": 1.025, "step": 1170 }, { "epoch": 0.09766594934613475, "grad_norm": 7.685535430908203, "learning_rate": 3.2524137931034485e-06, "loss": 0.8314, "step": 1180 }, { "epoch": 0.0984936268829664, "grad_norm": 8.43617057800293, "learning_rate": 3.2800000000000004e-06, "loss": 0.8783, "step": 1190 }, { "epoch": 0.09932130441979804, "grad_norm": 7.262084484100342, "learning_rate": 3.307586206896552e-06, "loss": 0.8855, "step": 1200 }, { "epoch": 0.10014898195662969, "grad_norm": 11.033655166625977, "learning_rate": 3.3351724137931037e-06, "loss": 0.7862, "step": 1210 }, { "epoch": 0.10097665949346135, "grad_norm": 10.095381736755371, "learning_rate": 3.3627586206896555e-06, "loss": 0.8322, "step": 1220 }, { "epoch": 0.101804337030293, "grad_norm": 11.873652458190918, "learning_rate": 3.390344827586207e-06, "loss": 0.865, "step": 1230 }, { "epoch": 0.10263201456712465, "grad_norm": 9.190414428710938, "learning_rate": 3.417931034482759e-06, "loss": 0.9041, "step": 1240 }, { "epoch": 0.1034596921039563, "grad_norm": 7.542434215545654, "learning_rate": 3.4455172413793107e-06, "loss": 0.9174, "step": 1250 }, { "epoch": 0.10428736964078796, "grad_norm": 8.151114463806152, "learning_rate": 3.4731034482758626e-06, "loss": 0.8806, "step": 1260 }, { "epoch": 0.1051150471776196, "grad_norm": 7.657033920288086, "learning_rate": 3.500689655172414e-06, "loss": 0.9218, "step": 1270 }, { "epoch": 0.10594272471445125, "grad_norm": 7.770766258239746, "learning_rate": 3.5282758620689655e-06, "loss": 0.8086, "step": 1280 }, { "epoch": 0.1067704022512829, "grad_norm": 7.381650924682617, "learning_rate": 3.5558620689655178e-06, "loss": 0.905, "step": 1290 }, { "epoch": 0.10759807978811455, "grad_norm": 6.547089099884033, "learning_rate": 3.5834482758620692e-06, "loss": 0.8068, "step": 1300 }, { "epoch": 0.1084257573249462, "grad_norm": 6.162235260009766, "learning_rate": 3.6110344827586207e-06, "loss": 0.8127, "step": 1310 }, { "epoch": 0.10925343486177785, "grad_norm": 7.459283351898193, "learning_rate": 3.638620689655173e-06, "loss": 0.8296, "step": 1320 }, { "epoch": 0.1100811123986095, "grad_norm": 7.895771503448486, "learning_rate": 3.6662068965517244e-06, "loss": 0.9099, "step": 1330 }, { "epoch": 0.11090878993544115, "grad_norm": 8.573246955871582, "learning_rate": 3.693793103448276e-06, "loss": 0.7521, "step": 1340 }, { "epoch": 0.1117364674722728, "grad_norm": 6.840447902679443, "learning_rate": 3.721379310344828e-06, "loss": 0.9412, "step": 1350 }, { "epoch": 0.11256414500910446, "grad_norm": 7.455838680267334, "learning_rate": 3.7489655172413796e-06, "loss": 0.7776, "step": 1360 }, { "epoch": 0.1133918225459361, "grad_norm": 8.462231636047363, "learning_rate": 3.7765517241379315e-06, "loss": 0.8133, "step": 1370 }, { "epoch": 0.11421950008276775, "grad_norm": 7.468634605407715, "learning_rate": 3.804137931034483e-06, "loss": 0.8568, "step": 1380 }, { "epoch": 0.1150471776195994, "grad_norm": 7.882059574127197, "learning_rate": 3.831724137931035e-06, "loss": 0.8832, "step": 1390 }, { "epoch": 0.11587485515643106, "grad_norm": 8.732821464538574, "learning_rate": 3.859310344827587e-06, "loss": 0.8816, "step": 1400 }, { "epoch": 0.11670253269326271, "grad_norm": 8.944989204406738, "learning_rate": 3.886896551724138e-06, "loss": 0.8278, "step": 1410 }, { "epoch": 0.11753021023009436, "grad_norm": 14.057497024536133, "learning_rate": 3.91448275862069e-06, "loss": 0.7833, "step": 1420 }, { "epoch": 0.118357887766926, "grad_norm": 7.274802207946777, "learning_rate": 3.942068965517241e-06, "loss": 0.8131, "step": 1430 }, { "epoch": 0.11918556530375765, "grad_norm": 8.092809677124023, "learning_rate": 3.969655172413793e-06, "loss": 0.8064, "step": 1440 }, { "epoch": 0.12001324284058931, "grad_norm": 6.9187211990356445, "learning_rate": 3.997241379310345e-06, "loss": 0.6907, "step": 1450 }, { "epoch": 0.12084092037742096, "grad_norm": 8.519347190856934, "learning_rate": 4.024827586206897e-06, "loss": 0.7279, "step": 1460 }, { "epoch": 0.12166859791425261, "grad_norm": 6.592753887176514, "learning_rate": 4.052413793103449e-06, "loss": 0.8769, "step": 1470 }, { "epoch": 0.12249627545108425, "grad_norm": 9.070652961730957, "learning_rate": 4.08e-06, "loss": 0.7353, "step": 1480 }, { "epoch": 0.1233239529879159, "grad_norm": 9.220254898071289, "learning_rate": 4.107586206896552e-06, "loss": 0.743, "step": 1490 }, { "epoch": 0.12415163052474756, "grad_norm": 8.688329696655273, "learning_rate": 4.135172413793104e-06, "loss": 0.8239, "step": 1500 }, { "epoch": 0.12497930806157921, "grad_norm": 8.735703468322754, "learning_rate": 4.1627586206896555e-06, "loss": 0.7639, "step": 1510 }, { "epoch": 0.12580698559841086, "grad_norm": 12.208942413330078, "learning_rate": 4.190344827586207e-06, "loss": 0.9293, "step": 1520 }, { "epoch": 0.12663466313524252, "grad_norm": 9.241416931152344, "learning_rate": 4.217931034482759e-06, "loss": 0.7874, "step": 1530 }, { "epoch": 0.12746234067207415, "grad_norm": 7.683334827423096, "learning_rate": 4.24551724137931e-06, "loss": 0.838, "step": 1540 }, { "epoch": 0.12829001820890581, "grad_norm": 9.873690605163574, "learning_rate": 4.273103448275863e-06, "loss": 0.8501, "step": 1550 }, { "epoch": 0.12911769574573745, "grad_norm": 13.48336410522461, "learning_rate": 4.300689655172414e-06, "loss": 0.8162, "step": 1560 }, { "epoch": 0.1299453732825691, "grad_norm": 8.262998580932617, "learning_rate": 4.328275862068966e-06, "loss": 0.789, "step": 1570 }, { "epoch": 0.13077305081940077, "grad_norm": 6.916970729827881, "learning_rate": 4.355862068965518e-06, "loss": 0.7051, "step": 1580 }, { "epoch": 0.1316007283562324, "grad_norm": 7.220826148986816, "learning_rate": 4.38344827586207e-06, "loss": 0.7926, "step": 1590 }, { "epoch": 0.13242840589306407, "grad_norm": 8.303420066833496, "learning_rate": 4.411034482758621e-06, "loss": 0.7459, "step": 1600 }, { "epoch": 0.1332560834298957, "grad_norm": 6.866578578948975, "learning_rate": 4.4386206896551725e-06, "loss": 0.84, "step": 1610 }, { "epoch": 0.13408376096672736, "grad_norm": 7.686966896057129, "learning_rate": 4.466206896551724e-06, "loss": 0.8161, "step": 1620 }, { "epoch": 0.13491143850355902, "grad_norm": 8.104334831237793, "learning_rate": 4.493793103448276e-06, "loss": 0.7725, "step": 1630 }, { "epoch": 0.13573911604039066, "grad_norm": 7.13726806640625, "learning_rate": 4.521379310344828e-06, "loss": 0.7424, "step": 1640 }, { "epoch": 0.13656679357722232, "grad_norm": 5.701939105987549, "learning_rate": 4.54896551724138e-06, "loss": 0.8674, "step": 1650 }, { "epoch": 0.13739447111405395, "grad_norm": 10.230535507202148, "learning_rate": 4.576551724137932e-06, "loss": 0.8472, "step": 1660 }, { "epoch": 0.1382221486508856, "grad_norm": 6.748244285583496, "learning_rate": 4.604137931034483e-06, "loss": 0.8022, "step": 1670 }, { "epoch": 0.13904982618771727, "grad_norm": 7.462981224060059, "learning_rate": 4.631724137931035e-06, "loss": 0.7754, "step": 1680 }, { "epoch": 0.1398775037245489, "grad_norm": 8.685107231140137, "learning_rate": 4.659310344827587e-06, "loss": 0.7896, "step": 1690 }, { "epoch": 0.14070518126138057, "grad_norm": 13.128334045410156, "learning_rate": 4.6868965517241385e-06, "loss": 0.7578, "step": 1700 }, { "epoch": 0.14153285879821223, "grad_norm": 7.565915107727051, "learning_rate": 4.7144827586206895e-06, "loss": 0.8021, "step": 1710 }, { "epoch": 0.14236053633504386, "grad_norm": 7.316013336181641, "learning_rate": 4.742068965517242e-06, "loss": 0.7871, "step": 1720 }, { "epoch": 0.14318821387187552, "grad_norm": 10.263166427612305, "learning_rate": 4.769655172413793e-06, "loss": 0.7365, "step": 1730 }, { "epoch": 0.14401589140870716, "grad_norm": 10.295011520385742, "learning_rate": 4.797241379310345e-06, "loss": 0.7964, "step": 1740 }, { "epoch": 0.14484356894553882, "grad_norm": 6.592663764953613, "learning_rate": 4.824827586206897e-06, "loss": 0.7944, "step": 1750 }, { "epoch": 0.14567124648237048, "grad_norm": 12.505395889282227, "learning_rate": 4.852413793103449e-06, "loss": 0.7724, "step": 1760 }, { "epoch": 0.14649892401920211, "grad_norm": 7.767071723937988, "learning_rate": 4.880000000000001e-06, "loss": 0.8758, "step": 1770 }, { "epoch": 0.14732660155603378, "grad_norm": 11.239216804504395, "learning_rate": 4.907586206896552e-06, "loss": 0.7398, "step": 1780 }, { "epoch": 0.1481542790928654, "grad_norm": 8.762969970703125, "learning_rate": 4.935172413793104e-06, "loss": 0.7556, "step": 1790 }, { "epoch": 0.14898195662969707, "grad_norm": 9.991865158081055, "learning_rate": 4.9627586206896555e-06, "loss": 0.8757, "step": 1800 }, { "epoch": 0.14980963416652873, "grad_norm": 8.235747337341309, "learning_rate": 4.990344827586207e-06, "loss": 0.7061, "step": 1810 }, { "epoch": 0.15063731170336037, "grad_norm": 8.77885913848877, "learning_rate": 5.017931034482759e-06, "loss": 0.7139, "step": 1820 }, { "epoch": 0.15146498924019203, "grad_norm": 6.980468273162842, "learning_rate": 5.045517241379311e-06, "loss": 0.8598, "step": 1830 }, { "epoch": 0.15229266677702366, "grad_norm": 7.68463134765625, "learning_rate": 5.073103448275862e-06, "loss": 0.8162, "step": 1840 }, { "epoch": 0.15312034431385532, "grad_norm": 12.931082725524902, "learning_rate": 5.100689655172414e-06, "loss": 0.7576, "step": 1850 }, { "epoch": 0.15394802185068698, "grad_norm": 8.319313049316406, "learning_rate": 5.128275862068966e-06, "loss": 0.7761, "step": 1860 }, { "epoch": 0.15477569938751862, "grad_norm": 7.096038818359375, "learning_rate": 5.155862068965518e-06, "loss": 0.7825, "step": 1870 }, { "epoch": 0.15560337692435028, "grad_norm": 7.043363094329834, "learning_rate": 5.1834482758620696e-06, "loss": 0.7957, "step": 1880 }, { "epoch": 0.1564310544611819, "grad_norm": 12.084305763244629, "learning_rate": 5.2110344827586214e-06, "loss": 0.7443, "step": 1890 }, { "epoch": 0.15725873199801357, "grad_norm": 7.071404457092285, "learning_rate": 5.2386206896551725e-06, "loss": 0.7428, "step": 1900 }, { "epoch": 0.15808640953484523, "grad_norm": 6.8817830085754395, "learning_rate": 5.266206896551724e-06, "loss": 0.8485, "step": 1910 }, { "epoch": 0.15891408707167687, "grad_norm": 9.035568237304688, "learning_rate": 5.293793103448276e-06, "loss": 0.836, "step": 1920 }, { "epoch": 0.15974176460850853, "grad_norm": 9.296754837036133, "learning_rate": 5.321379310344829e-06, "loss": 0.7687, "step": 1930 }, { "epoch": 0.16056944214534016, "grad_norm": 8.886366844177246, "learning_rate": 5.34896551724138e-06, "loss": 0.7084, "step": 1940 }, { "epoch": 0.16139711968217182, "grad_norm": 8.598686218261719, "learning_rate": 5.376551724137932e-06, "loss": 0.8033, "step": 1950 }, { "epoch": 0.16222479721900349, "grad_norm": 6.800096035003662, "learning_rate": 5.404137931034483e-06, "loss": 0.7969, "step": 1960 }, { "epoch": 0.16305247475583512, "grad_norm": 7.165079593658447, "learning_rate": 5.431724137931035e-06, "loss": 0.7143, "step": 1970 }, { "epoch": 0.16388015229266678, "grad_norm": 12.248144149780273, "learning_rate": 5.4593103448275866e-06, "loss": 0.7992, "step": 1980 }, { "epoch": 0.16470782982949841, "grad_norm": 11.9479341506958, "learning_rate": 5.486896551724138e-06, "loss": 0.7536, "step": 1990 }, { "epoch": 0.16553550736633008, "grad_norm": 11.062726020812988, "learning_rate": 5.51448275862069e-06, "loss": 0.7577, "step": 2000 }, { "epoch": 0.16636318490316174, "grad_norm": 6.469613075256348, "learning_rate": 5.542068965517242e-06, "loss": 0.7517, "step": 2010 }, { "epoch": 0.16719086243999337, "grad_norm": 8.283201217651367, "learning_rate": 5.569655172413794e-06, "loss": 0.8068, "step": 2020 }, { "epoch": 0.16801853997682503, "grad_norm": 8.749649047851562, "learning_rate": 5.597241379310345e-06, "loss": 0.789, "step": 2030 }, { "epoch": 0.16884621751365667, "grad_norm": 6.738229274749756, "learning_rate": 5.624827586206897e-06, "loss": 0.7082, "step": 2040 }, { "epoch": 0.16967389505048833, "grad_norm": 15.499320983886719, "learning_rate": 5.652413793103449e-06, "loss": 0.8842, "step": 2050 }, { "epoch": 0.17050157258732, "grad_norm": 5.899293899536133, "learning_rate": 5.68e-06, "loss": 0.7926, "step": 2060 }, { "epoch": 0.17132925012415162, "grad_norm": 7.767611503601074, "learning_rate": 5.7075862068965525e-06, "loss": 0.7788, "step": 2070 }, { "epoch": 0.17215692766098328, "grad_norm": 7.4701619148254395, "learning_rate": 5.735172413793104e-06, "loss": 0.6726, "step": 2080 }, { "epoch": 0.17298460519781494, "grad_norm": 5.770844459533691, "learning_rate": 5.7627586206896554e-06, "loss": 0.8054, "step": 2090 }, { "epoch": 0.17381228273464658, "grad_norm": 13.475322723388672, "learning_rate": 5.790344827586207e-06, "loss": 0.7582, "step": 2100 }, { "epoch": 0.17463996027147824, "grad_norm": 7.453348159790039, "learning_rate": 5.817931034482759e-06, "loss": 0.7005, "step": 2110 }, { "epoch": 0.17546763780830987, "grad_norm": 7.863591194152832, "learning_rate": 5.84551724137931e-06, "loss": 0.8148, "step": 2120 }, { "epoch": 0.17629531534514153, "grad_norm": 8.45164966583252, "learning_rate": 5.873103448275863e-06, "loss": 0.7674, "step": 2130 }, { "epoch": 0.1771229928819732, "grad_norm": 11.415011405944824, "learning_rate": 5.900689655172415e-06, "loss": 0.7831, "step": 2140 }, { "epoch": 0.17795067041880483, "grad_norm": 7.594445705413818, "learning_rate": 5.928275862068966e-06, "loss": 0.8264, "step": 2150 }, { "epoch": 0.1787783479556365, "grad_norm": 7.368844032287598, "learning_rate": 5.955862068965518e-06, "loss": 0.9232, "step": 2160 }, { "epoch": 0.17960602549246812, "grad_norm": 8.73694133758545, "learning_rate": 5.9834482758620695e-06, "loss": 0.8108, "step": 2170 }, { "epoch": 0.18043370302929979, "grad_norm": 11.164834022521973, "learning_rate": 6.0110344827586206e-06, "loss": 0.8051, "step": 2180 }, { "epoch": 0.18126138056613145, "grad_norm": 8.533145904541016, "learning_rate": 6.0386206896551724e-06, "loss": 0.7254, "step": 2190 }, { "epoch": 0.18208905810296308, "grad_norm": 7.987492561340332, "learning_rate": 6.066206896551725e-06, "loss": 0.7664, "step": 2200 }, { "epoch": 0.18291673563979474, "grad_norm": 6.07111120223999, "learning_rate": 6.093793103448277e-06, "loss": 0.7615, "step": 2210 }, { "epoch": 0.18374441317662638, "grad_norm": 7.719367504119873, "learning_rate": 6.121379310344828e-06, "loss": 0.8209, "step": 2220 }, { "epoch": 0.18457209071345804, "grad_norm": 13.052729606628418, "learning_rate": 6.14896551724138e-06, "loss": 0.8058, "step": 2230 }, { "epoch": 0.1853997682502897, "grad_norm": 7.551183700561523, "learning_rate": 6.176551724137932e-06, "loss": 0.8662, "step": 2240 }, { "epoch": 0.18622744578712133, "grad_norm": 10.130929946899414, "learning_rate": 6.204137931034483e-06, "loss": 0.6247, "step": 2250 }, { "epoch": 0.187055123323953, "grad_norm": 6.396650314331055, "learning_rate": 6.231724137931035e-06, "loss": 0.7603, "step": 2260 }, { "epoch": 0.18788280086078463, "grad_norm": 5.974127769470215, "learning_rate": 6.259310344827587e-06, "loss": 0.7039, "step": 2270 }, { "epoch": 0.1887104783976163, "grad_norm": 12.845807075500488, "learning_rate": 6.286896551724138e-06, "loss": 0.8063, "step": 2280 }, { "epoch": 0.18953815593444795, "grad_norm": 8.013260841369629, "learning_rate": 6.31448275862069e-06, "loss": 0.8106, "step": 2290 }, { "epoch": 0.19036583347127958, "grad_norm": 7.192122459411621, "learning_rate": 6.342068965517242e-06, "loss": 0.729, "step": 2300 }, { "epoch": 0.19119351100811124, "grad_norm": 9.58069896697998, "learning_rate": 6.369655172413793e-06, "loss": 0.6877, "step": 2310 }, { "epoch": 0.19202118854494288, "grad_norm": 7.573521137237549, "learning_rate": 6.397241379310345e-06, "loss": 0.8292, "step": 2320 }, { "epoch": 0.19284886608177454, "grad_norm": 6.273680210113525, "learning_rate": 6.424827586206898e-06, "loss": 0.8216, "step": 2330 }, { "epoch": 0.1936765436186062, "grad_norm": 5.834144115447998, "learning_rate": 6.452413793103449e-06, "loss": 0.828, "step": 2340 }, { "epoch": 0.19450422115543783, "grad_norm": 7.580963611602783, "learning_rate": 6.480000000000001e-06, "loss": 0.7376, "step": 2350 }, { "epoch": 0.1953318986922695, "grad_norm": 6.9146928787231445, "learning_rate": 6.5075862068965525e-06, "loss": 0.7694, "step": 2360 }, { "epoch": 0.19615957622910113, "grad_norm": 8.36328125, "learning_rate": 6.5351724137931035e-06, "loss": 0.8395, "step": 2370 }, { "epoch": 0.1969872537659328, "grad_norm": 7.232540607452393, "learning_rate": 6.562758620689655e-06, "loss": 0.8177, "step": 2380 }, { "epoch": 0.19781493130276445, "grad_norm": 7.676825046539307, "learning_rate": 6.590344827586207e-06, "loss": 0.7233, "step": 2390 }, { "epoch": 0.19864260883959609, "grad_norm": 7.215404510498047, "learning_rate": 6.61793103448276e-06, "loss": 0.8078, "step": 2400 }, { "epoch": 0.19947028637642775, "grad_norm": 7.93449592590332, "learning_rate": 6.645517241379311e-06, "loss": 0.6672, "step": 2410 }, { "epoch": 0.20029796391325938, "grad_norm": 8.673205375671387, "learning_rate": 6.673103448275863e-06, "loss": 0.7495, "step": 2420 }, { "epoch": 0.20112564145009104, "grad_norm": 7.223012447357178, "learning_rate": 6.700689655172415e-06, "loss": 0.8733, "step": 2430 }, { "epoch": 0.2019533189869227, "grad_norm": 6.771484851837158, "learning_rate": 6.728275862068966e-06, "loss": 0.7454, "step": 2440 }, { "epoch": 0.20278099652375434, "grad_norm": 6.906106948852539, "learning_rate": 6.755862068965518e-06, "loss": 0.6594, "step": 2450 }, { "epoch": 0.203608674060586, "grad_norm": 8.127269744873047, "learning_rate": 6.783448275862069e-06, "loss": 0.8, "step": 2460 }, { "epoch": 0.20443635159741766, "grad_norm": 8.180729866027832, "learning_rate": 6.811034482758621e-06, "loss": 0.776, "step": 2470 }, { "epoch": 0.2052640291342493, "grad_norm": 7.284351348876953, "learning_rate": 6.838620689655173e-06, "loss": 0.7012, "step": 2480 }, { "epoch": 0.20609170667108095, "grad_norm": 8.30833625793457, "learning_rate": 6.866206896551725e-06, "loss": 0.7877, "step": 2490 }, { "epoch": 0.2069193842079126, "grad_norm": 8.156807899475098, "learning_rate": 6.893793103448276e-06, "loss": 0.7551, "step": 2500 }, { "epoch": 0.20774706174474425, "grad_norm": 9.105708122253418, "learning_rate": 6.921379310344828e-06, "loss": 0.7525, "step": 2510 }, { "epoch": 0.2085747392815759, "grad_norm": 17.45189666748047, "learning_rate": 6.94896551724138e-06, "loss": 0.7164, "step": 2520 }, { "epoch": 0.20940241681840754, "grad_norm": 7.532208442687988, "learning_rate": 6.976551724137932e-06, "loss": 0.8148, "step": 2530 }, { "epoch": 0.2102300943552392, "grad_norm": 11.10726261138916, "learning_rate": 7.004137931034484e-06, "loss": 0.6312, "step": 2540 }, { "epoch": 0.21105777189207084, "grad_norm": 6.374979019165039, "learning_rate": 7.0317241379310355e-06, "loss": 0.7354, "step": 2550 }, { "epoch": 0.2118854494289025, "grad_norm": 6.365556716918945, "learning_rate": 7.0593103448275865e-06, "loss": 0.8223, "step": 2560 }, { "epoch": 0.21271312696573416, "grad_norm": 7.2601237297058105, "learning_rate": 7.086896551724138e-06, "loss": 0.7424, "step": 2570 }, { "epoch": 0.2135408045025658, "grad_norm": 6.324973106384277, "learning_rate": 7.11448275862069e-06, "loss": 0.7263, "step": 2580 }, { "epoch": 0.21436848203939746, "grad_norm": 4.7687883377075195, "learning_rate": 7.142068965517241e-06, "loss": 0.6771, "step": 2590 }, { "epoch": 0.2151961595762291, "grad_norm": 8.233292579650879, "learning_rate": 7.169655172413794e-06, "loss": 0.768, "step": 2600 }, { "epoch": 0.21602383711306075, "grad_norm": 6.167933940887451, "learning_rate": 7.197241379310346e-06, "loss": 0.7873, "step": 2610 }, { "epoch": 0.2168515146498924, "grad_norm": 11.615789413452148, "learning_rate": 7.224827586206897e-06, "loss": 0.7692, "step": 2620 }, { "epoch": 0.21767919218672405, "grad_norm": 5.612741470336914, "learning_rate": 7.252413793103449e-06, "loss": 0.7195, "step": 2630 }, { "epoch": 0.2185068697235557, "grad_norm": 6.363499641418457, "learning_rate": 7.280000000000001e-06, "loss": 0.8135, "step": 2640 }, { "epoch": 0.21933454726038734, "grad_norm": 6.777540683746338, "learning_rate": 7.307586206896552e-06, "loss": 0.7693, "step": 2650 }, { "epoch": 0.220162224797219, "grad_norm": 6.548894882202148, "learning_rate": 7.3351724137931035e-06, "loss": 0.8058, "step": 2660 }, { "epoch": 0.22098990233405066, "grad_norm": 8.381720542907715, "learning_rate": 7.362758620689656e-06, "loss": 0.78, "step": 2670 }, { "epoch": 0.2218175798708823, "grad_norm": 7.393632888793945, "learning_rate": 7.390344827586208e-06, "loss": 0.8168, "step": 2680 }, { "epoch": 0.22264525740771396, "grad_norm": 7.97804594039917, "learning_rate": 7.417931034482759e-06, "loss": 0.7898, "step": 2690 }, { "epoch": 0.2234729349445456, "grad_norm": 6.426300525665283, "learning_rate": 7.445517241379311e-06, "loss": 0.7072, "step": 2700 }, { "epoch": 0.22430061248137725, "grad_norm": 6.097227573394775, "learning_rate": 7.473103448275863e-06, "loss": 0.7443, "step": 2710 }, { "epoch": 0.22512829001820892, "grad_norm": 7.343870639801025, "learning_rate": 7.500689655172414e-06, "loss": 0.6726, "step": 2720 }, { "epoch": 0.22595596755504055, "grad_norm": 7.795609951019287, "learning_rate": 7.5282758620689666e-06, "loss": 0.7399, "step": 2730 }, { "epoch": 0.2267836450918722, "grad_norm": 7.497182369232178, "learning_rate": 7.5558620689655184e-06, "loss": 0.8465, "step": 2740 }, { "epoch": 0.22761132262870384, "grad_norm": 7.815415382385254, "learning_rate": 7.5834482758620695e-06, "loss": 0.7677, "step": 2750 }, { "epoch": 0.2284390001655355, "grad_norm": 6.56506872177124, "learning_rate": 7.611034482758621e-06, "loss": 0.8499, "step": 2760 }, { "epoch": 0.22926667770236717, "grad_norm": 7.4583916664123535, "learning_rate": 7.638620689655172e-06, "loss": 0.7791, "step": 2770 }, { "epoch": 0.2300943552391988, "grad_norm": 7.336313724517822, "learning_rate": 7.666206896551724e-06, "loss": 0.8238, "step": 2780 }, { "epoch": 0.23092203277603046, "grad_norm": 14.851055145263672, "learning_rate": 7.693793103448276e-06, "loss": 0.7239, "step": 2790 }, { "epoch": 0.23174971031286212, "grad_norm": 7.575233459472656, "learning_rate": 7.721379310344828e-06, "loss": 0.8797, "step": 2800 }, { "epoch": 0.23257738784969376, "grad_norm": 6.171555995941162, "learning_rate": 7.74896551724138e-06, "loss": 0.7986, "step": 2810 }, { "epoch": 0.23340506538652542, "grad_norm": 6.3385910987854, "learning_rate": 7.776551724137932e-06, "loss": 0.7665, "step": 2820 }, { "epoch": 0.23423274292335705, "grad_norm": 9.962767601013184, "learning_rate": 7.804137931034484e-06, "loss": 0.7266, "step": 2830 }, { "epoch": 0.2350604204601887, "grad_norm": 8.821725845336914, "learning_rate": 7.831724137931035e-06, "loss": 0.7182, "step": 2840 }, { "epoch": 0.23588809799702037, "grad_norm": 6.448456764221191, "learning_rate": 7.859310344827587e-06, "loss": 0.8134, "step": 2850 }, { "epoch": 0.236715775533852, "grad_norm": 7.002468109130859, "learning_rate": 7.886896551724137e-06, "loss": 0.7731, "step": 2860 }, { "epoch": 0.23754345307068367, "grad_norm": 6.797628402709961, "learning_rate": 7.914482758620691e-06, "loss": 0.7012, "step": 2870 }, { "epoch": 0.2383711306075153, "grad_norm": 6.411037921905518, "learning_rate": 7.942068965517243e-06, "loss": 0.6651, "step": 2880 }, { "epoch": 0.23919880814434696, "grad_norm": 7.10167932510376, "learning_rate": 7.969655172413793e-06, "loss": 0.684, "step": 2890 }, { "epoch": 0.24002648568117863, "grad_norm": 7.986268520355225, "learning_rate": 7.997241379310345e-06, "loss": 0.6872, "step": 2900 }, { "epoch": 0.24085416321801026, "grad_norm": 5.454895496368408, "learning_rate": 8.024827586206897e-06, "loss": 0.9023, "step": 2910 }, { "epoch": 0.24168184075484192, "grad_norm": 6.361053943634033, "learning_rate": 8.052413793103449e-06, "loss": 0.7306, "step": 2920 }, { "epoch": 0.24250951829167355, "grad_norm": 6.19024658203125, "learning_rate": 8.08e-06, "loss": 0.6985, "step": 2930 }, { "epoch": 0.24333719582850522, "grad_norm": 7.439888954162598, "learning_rate": 8.107586206896552e-06, "loss": 0.7526, "step": 2940 }, { "epoch": 0.24416487336533688, "grad_norm": 6.509820938110352, "learning_rate": 8.135172413793104e-06, "loss": 0.7379, "step": 2950 }, { "epoch": 0.2449925509021685, "grad_norm": 7.4113993644714355, "learning_rate": 8.162758620689656e-06, "loss": 0.669, "step": 2960 }, { "epoch": 0.24582022843900017, "grad_norm": 5.187702655792236, "learning_rate": 8.190344827586208e-06, "loss": 0.7367, "step": 2970 }, { "epoch": 0.2466479059758318, "grad_norm": 6.528170585632324, "learning_rate": 8.217931034482758e-06, "loss": 0.7504, "step": 2980 }, { "epoch": 0.24747558351266347, "grad_norm": 14.141435623168945, "learning_rate": 8.24551724137931e-06, "loss": 0.7803, "step": 2990 }, { "epoch": 0.24830326104949513, "grad_norm": 6.171212673187256, "learning_rate": 8.273103448275864e-06, "loss": 0.8076, "step": 3000 }, { "epoch": 0.24913093858632676, "grad_norm": 8.228569030761719, "learning_rate": 8.300689655172416e-06, "loss": 0.6893, "step": 3010 }, { "epoch": 0.24995861612315842, "grad_norm": 6.737586975097656, "learning_rate": 8.328275862068966e-06, "loss": 0.7402, "step": 3020 }, { "epoch": 0.25078629365999006, "grad_norm": 13.37900161743164, "learning_rate": 8.355862068965518e-06, "loss": 0.7437, "step": 3030 }, { "epoch": 0.2516139711968217, "grad_norm": 6.189072608947754, "learning_rate": 8.38344827586207e-06, "loss": 0.7793, "step": 3040 }, { "epoch": 0.2524416487336534, "grad_norm": 6.708379745483398, "learning_rate": 8.411034482758621e-06, "loss": 0.7555, "step": 3050 }, { "epoch": 0.25326932627048504, "grad_norm": 6.613540172576904, "learning_rate": 8.438620689655173e-06, "loss": 0.8063, "step": 3060 }, { "epoch": 0.25409700380731665, "grad_norm": 8.459820747375488, "learning_rate": 8.466206896551725e-06, "loss": 0.7102, "step": 3070 }, { "epoch": 0.2549246813441483, "grad_norm": 6.833925247192383, "learning_rate": 8.493793103448277e-06, "loss": 0.7631, "step": 3080 }, { "epoch": 0.25575235888097997, "grad_norm": 5.531505107879639, "learning_rate": 8.521379310344829e-06, "loss": 0.7266, "step": 3090 }, { "epoch": 0.25658003641781163, "grad_norm": 8.19724178314209, "learning_rate": 8.54896551724138e-06, "loss": 0.7502, "step": 3100 }, { "epoch": 0.2574077139546433, "grad_norm": 7.0897908210754395, "learning_rate": 8.57655172413793e-06, "loss": 0.6943, "step": 3110 }, { "epoch": 0.2582353914914749, "grad_norm": 6.397073268890381, "learning_rate": 8.604137931034483e-06, "loss": 0.7452, "step": 3120 }, { "epoch": 0.25906306902830656, "grad_norm": 10.09765625, "learning_rate": 8.631724137931036e-06, "loss": 0.7427, "step": 3130 }, { "epoch": 0.2598907465651382, "grad_norm": 5.98114538192749, "learning_rate": 8.659310344827586e-06, "loss": 0.7405, "step": 3140 }, { "epoch": 0.2607184241019699, "grad_norm": 9.228670120239258, "learning_rate": 8.686896551724138e-06, "loss": 0.8088, "step": 3150 }, { "epoch": 0.26154610163880154, "grad_norm": 7.793185710906982, "learning_rate": 8.71448275862069e-06, "loss": 0.8388, "step": 3160 }, { "epoch": 0.26237377917563315, "grad_norm": 14.522049903869629, "learning_rate": 8.742068965517242e-06, "loss": 0.7144, "step": 3170 }, { "epoch": 0.2632014567124648, "grad_norm": 5.113579273223877, "learning_rate": 8.769655172413794e-06, "loss": 0.6766, "step": 3180 }, { "epoch": 0.26402913424929647, "grad_norm": 5.0188775062561035, "learning_rate": 8.797241379310346e-06, "loss": 0.7216, "step": 3190 }, { "epoch": 0.26485681178612813, "grad_norm": 6.824369430541992, "learning_rate": 8.824827586206898e-06, "loss": 0.8085, "step": 3200 }, { "epoch": 0.2656844893229598, "grad_norm": 6.079699993133545, "learning_rate": 8.85241379310345e-06, "loss": 0.6944, "step": 3210 }, { "epoch": 0.2665121668597914, "grad_norm": 7.916310787200928, "learning_rate": 8.880000000000001e-06, "loss": 0.7319, "step": 3220 }, { "epoch": 0.26733984439662306, "grad_norm": 8.215185165405273, "learning_rate": 8.907586206896553e-06, "loss": 0.6646, "step": 3230 }, { "epoch": 0.2681675219334547, "grad_norm": 6.852560997009277, "learning_rate": 8.935172413793103e-06, "loss": 0.7393, "step": 3240 }, { "epoch": 0.2689951994702864, "grad_norm": 8.733128547668457, "learning_rate": 8.962758620689655e-06, "loss": 0.823, "step": 3250 }, { "epoch": 0.26982287700711804, "grad_norm": 5.837035655975342, "learning_rate": 8.990344827586207e-06, "loss": 0.7958, "step": 3260 }, { "epoch": 0.27065055454394965, "grad_norm": 7.896422863006592, "learning_rate": 9.017931034482759e-06, "loss": 0.5961, "step": 3270 }, { "epoch": 0.2714782320807813, "grad_norm": 7.217525959014893, "learning_rate": 9.045517241379311e-06, "loss": 0.6598, "step": 3280 }, { "epoch": 0.272305909617613, "grad_norm": 9.825339317321777, "learning_rate": 9.073103448275863e-06, "loss": 0.7673, "step": 3290 }, { "epoch": 0.27313358715444463, "grad_norm": 11.47857666015625, "learning_rate": 9.100689655172415e-06, "loss": 0.7566, "step": 3300 }, { "epoch": 0.2739612646912763, "grad_norm": 6.565371036529541, "learning_rate": 9.128275862068967e-06, "loss": 0.8588, "step": 3310 }, { "epoch": 0.2747889422281079, "grad_norm": 5.687349319458008, "learning_rate": 9.155862068965518e-06, "loss": 0.7442, "step": 3320 }, { "epoch": 0.27561661976493956, "grad_norm": 10.703747749328613, "learning_rate": 9.18344827586207e-06, "loss": 0.7418, "step": 3330 }, { "epoch": 0.2764442973017712, "grad_norm": 7.3640947341918945, "learning_rate": 9.211034482758622e-06, "loss": 0.7524, "step": 3340 }, { "epoch": 0.2772719748386029, "grad_norm": 6.118717193603516, "learning_rate": 9.238620689655174e-06, "loss": 0.7716, "step": 3350 }, { "epoch": 0.27809965237543455, "grad_norm": 5.455191612243652, "learning_rate": 9.266206896551724e-06, "loss": 0.6274, "step": 3360 }, { "epoch": 0.2789273299122662, "grad_norm": 4.7933855056762695, "learning_rate": 9.293793103448276e-06, "loss": 0.7429, "step": 3370 }, { "epoch": 0.2797550074490978, "grad_norm": 8.966490745544434, "learning_rate": 9.321379310344828e-06, "loss": 0.7736, "step": 3380 }, { "epoch": 0.2805826849859295, "grad_norm": 5.160833358764648, "learning_rate": 9.34896551724138e-06, "loss": 0.7123, "step": 3390 }, { "epoch": 0.28141036252276114, "grad_norm": 6.411797046661377, "learning_rate": 9.376551724137932e-06, "loss": 0.7823, "step": 3400 }, { "epoch": 0.2822380400595928, "grad_norm": 6.302147388458252, "learning_rate": 9.404137931034483e-06, "loss": 0.7764, "step": 3410 }, { "epoch": 0.28306571759642446, "grad_norm": 6.269154071807861, "learning_rate": 9.431724137931035e-06, "loss": 0.7971, "step": 3420 }, { "epoch": 0.28389339513325607, "grad_norm": 5.679346561431885, "learning_rate": 9.459310344827587e-06, "loss": 0.7366, "step": 3430 }, { "epoch": 0.2847210726700877, "grad_norm": 6.946877479553223, "learning_rate": 9.486896551724139e-06, "loss": 0.8055, "step": 3440 }, { "epoch": 0.2855487502069194, "grad_norm": 5.445899486541748, "learning_rate": 9.51448275862069e-06, "loss": 0.7868, "step": 3450 }, { "epoch": 0.28637642774375105, "grad_norm": 6.944708347320557, "learning_rate": 9.542068965517243e-06, "loss": 0.8238, "step": 3460 }, { "epoch": 0.2872041052805827, "grad_norm": 5.885404586791992, "learning_rate": 9.569655172413795e-06, "loss": 0.769, "step": 3470 }, { "epoch": 0.2880317828174143, "grad_norm": 7.060633182525635, "learning_rate": 9.597241379310347e-06, "loss": 0.6971, "step": 3480 }, { "epoch": 0.288859460354246, "grad_norm": 6.893513202667236, "learning_rate": 9.624827586206897e-06, "loss": 0.7057, "step": 3490 }, { "epoch": 0.28968713789107764, "grad_norm": 6.315013885498047, "learning_rate": 9.652413793103449e-06, "loss": 0.7072, "step": 3500 }, { "epoch": 0.2905148154279093, "grad_norm": 6.517123699188232, "learning_rate": 9.68e-06, "loss": 0.6808, "step": 3510 }, { "epoch": 0.29134249296474096, "grad_norm": 9.190183639526367, "learning_rate": 9.707586206896552e-06, "loss": 0.6397, "step": 3520 }, { "epoch": 0.29217017050157257, "grad_norm": 8.152873039245605, "learning_rate": 9.735172413793104e-06, "loss": 0.7868, "step": 3530 }, { "epoch": 0.29299784803840423, "grad_norm": 6.544162750244141, "learning_rate": 9.762758620689656e-06, "loss": 0.7763, "step": 3540 }, { "epoch": 0.2938255255752359, "grad_norm": 6.587504863739014, "learning_rate": 9.790344827586208e-06, "loss": 0.7333, "step": 3550 }, { "epoch": 0.29465320311206755, "grad_norm": 7.048343658447266, "learning_rate": 9.81793103448276e-06, "loss": 0.7624, "step": 3560 }, { "epoch": 0.2954808806488992, "grad_norm": 10.886177062988281, "learning_rate": 9.845517241379312e-06, "loss": 0.6741, "step": 3570 }, { "epoch": 0.2963085581857308, "grad_norm": 8.286664009094238, "learning_rate": 9.873103448275862e-06, "loss": 0.7237, "step": 3580 }, { "epoch": 0.2971362357225625, "grad_norm": 12.517027854919434, "learning_rate": 9.900689655172414e-06, "loss": 0.8612, "step": 3590 }, { "epoch": 0.29796391325939414, "grad_norm": 6.545004367828369, "learning_rate": 9.928275862068967e-06, "loss": 0.6624, "step": 3600 }, { "epoch": 0.2987915907962258, "grad_norm": 11.047503471374512, "learning_rate": 9.955862068965517e-06, "loss": 0.6232, "step": 3610 }, { "epoch": 0.29961926833305746, "grad_norm": 5.795104503631592, "learning_rate": 9.98344827586207e-06, "loss": 0.703, "step": 3620 }, { "epoch": 0.30044694586988907, "grad_norm": 6.570605754852295, "learning_rate": 9.999999629007449e-06, "loss": 0.8003, "step": 3630 }, { "epoch": 0.30127462340672073, "grad_norm": 6.392898082733154, "learning_rate": 9.99999545534187e-06, "loss": 0.7025, "step": 3640 }, { "epoch": 0.3021023009435524, "grad_norm": 6.936322212219238, "learning_rate": 9.99998664427391e-06, "loss": 0.7907, "step": 3650 }, { "epoch": 0.30292997848038405, "grad_norm": 9.38542366027832, "learning_rate": 9.999973195811736e-06, "loss": 0.7517, "step": 3660 }, { "epoch": 0.3037576560172157, "grad_norm": 14.276922225952148, "learning_rate": 9.999955109967825e-06, "loss": 0.8097, "step": 3670 }, { "epoch": 0.3045853335540473, "grad_norm": 10.349431991577148, "learning_rate": 9.99993238675895e-06, "loss": 0.7336, "step": 3680 }, { "epoch": 0.305413011090879, "grad_norm": 9.479475021362305, "learning_rate": 9.999905026206186e-06, "loss": 0.6715, "step": 3690 }, { "epoch": 0.30624068862771064, "grad_norm": 6.936245918273926, "learning_rate": 9.99987302833491e-06, "loss": 0.7724, "step": 3700 }, { "epoch": 0.3070683661645423, "grad_norm": 6.638550758361816, "learning_rate": 9.999836393174799e-06, "loss": 0.7842, "step": 3710 }, { "epoch": 0.30789604370137397, "grad_norm": 5.233384132385254, "learning_rate": 9.99979512075983e-06, "loss": 0.801, "step": 3720 }, { "epoch": 0.3087237212382056, "grad_norm": 10.094964981079102, "learning_rate": 9.999749211128284e-06, "loss": 0.6382, "step": 3730 }, { "epoch": 0.30955139877503723, "grad_norm": 6.864792823791504, "learning_rate": 9.999698664322743e-06, "loss": 0.7082, "step": 3740 }, { "epoch": 0.3103790763118689, "grad_norm": 6.34977388381958, "learning_rate": 9.999643480390085e-06, "loss": 0.7077, "step": 3750 }, { "epoch": 0.31120675384870056, "grad_norm": 5.912789821624756, "learning_rate": 9.999583659381495e-06, "loss": 0.6796, "step": 3760 }, { "epoch": 0.3120344313855322, "grad_norm": 5.825126647949219, "learning_rate": 9.999519201352452e-06, "loss": 0.7373, "step": 3770 }, { "epoch": 0.3128621089223638, "grad_norm": 5.921746253967285, "learning_rate": 9.999450106362745e-06, "loss": 0.6724, "step": 3780 }, { "epoch": 0.3136897864591955, "grad_norm": 6.277993202209473, "learning_rate": 9.999376374476453e-06, "loss": 0.7449, "step": 3790 }, { "epoch": 0.31451746399602715, "grad_norm": 7.837622165679932, "learning_rate": 9.999298005761964e-06, "loss": 0.7363, "step": 3800 }, { "epoch": 0.3153451415328588, "grad_norm": 5.344274520874023, "learning_rate": 9.999215000291962e-06, "loss": 0.7577, "step": 3810 }, { "epoch": 0.31617281906969047, "grad_norm": 6.780192852020264, "learning_rate": 9.999127358143434e-06, "loss": 0.646, "step": 3820 }, { "epoch": 0.3170004966065221, "grad_norm": 6.09605073928833, "learning_rate": 9.999035079397668e-06, "loss": 0.7676, "step": 3830 }, { "epoch": 0.31782817414335374, "grad_norm": 5.220916748046875, "learning_rate": 9.998938164140248e-06, "loss": 0.6421, "step": 3840 }, { "epoch": 0.3186558516801854, "grad_norm": 4.940789222717285, "learning_rate": 9.998836612461062e-06, "loss": 0.721, "step": 3850 }, { "epoch": 0.31948352921701706, "grad_norm": 5.293781280517578, "learning_rate": 9.998730424454297e-06, "loss": 0.7476, "step": 3860 }, { "epoch": 0.3203112067538487, "grad_norm": 5.640490531921387, "learning_rate": 9.99861960021844e-06, "loss": 0.7017, "step": 3870 }, { "epoch": 0.3211388842906803, "grad_norm": 5.843111991882324, "learning_rate": 9.998504139856282e-06, "loss": 0.7463, "step": 3880 }, { "epoch": 0.321966561827512, "grad_norm": 5.692841529846191, "learning_rate": 9.998384043474905e-06, "loss": 0.7111, "step": 3890 }, { "epoch": 0.32279423936434365, "grad_norm": 8.223881721496582, "learning_rate": 9.9982593111857e-06, "loss": 0.6625, "step": 3900 }, { "epoch": 0.3236219169011753, "grad_norm": 5.052455425262451, "learning_rate": 9.998129943104352e-06, "loss": 0.7336, "step": 3910 }, { "epoch": 0.32444959443800697, "grad_norm": 5.500833988189697, "learning_rate": 9.997995939350849e-06, "loss": 0.6798, "step": 3920 }, { "epoch": 0.3252772719748386, "grad_norm": 5.494566440582275, "learning_rate": 9.997857300049474e-06, "loss": 0.802, "step": 3930 }, { "epoch": 0.32610494951167024, "grad_norm": 5.324145793914795, "learning_rate": 9.997714025328817e-06, "loss": 0.7229, "step": 3940 }, { "epoch": 0.3269326270485019, "grad_norm": 6.409938335418701, "learning_rate": 9.997566115321758e-06, "loss": 0.8189, "step": 3950 }, { "epoch": 0.32776030458533356, "grad_norm": 10.51091194152832, "learning_rate": 9.997413570165484e-06, "loss": 0.7043, "step": 3960 }, { "epoch": 0.3285879821221652, "grad_norm": 5.197306156158447, "learning_rate": 9.997256390001475e-06, "loss": 0.8014, "step": 3970 }, { "epoch": 0.32941565965899683, "grad_norm": 5.635978698730469, "learning_rate": 9.997094574975517e-06, "loss": 0.7673, "step": 3980 }, { "epoch": 0.3302433371958285, "grad_norm": 5.672449588775635, "learning_rate": 9.996928125237684e-06, "loss": 0.7412, "step": 3990 }, { "epoch": 0.33107101473266015, "grad_norm": 4.873480319976807, "learning_rate": 9.996757040942362e-06, "loss": 0.694, "step": 4000 }, { "epoch": 0.3318986922694918, "grad_norm": 6.706229209899902, "learning_rate": 9.996581322248223e-06, "loss": 0.7277, "step": 4010 }, { "epoch": 0.3327263698063235, "grad_norm": 8.831635475158691, "learning_rate": 9.996400969318247e-06, "loss": 0.7636, "step": 4020 }, { "epoch": 0.3335540473431551, "grad_norm": 8.166945457458496, "learning_rate": 9.996215982319703e-06, "loss": 0.7444, "step": 4030 }, { "epoch": 0.33438172487998674, "grad_norm": 5.820005893707275, "learning_rate": 9.99602636142417e-06, "loss": 0.7325, "step": 4040 }, { "epoch": 0.3352094024168184, "grad_norm": 7.974793434143066, "learning_rate": 9.99583210680751e-06, "loss": 0.6221, "step": 4050 }, { "epoch": 0.33603707995365006, "grad_norm": 6.16264533996582, "learning_rate": 9.995633218649897e-06, "loss": 0.7212, "step": 4060 }, { "epoch": 0.3368647574904817, "grad_norm": 7.6961517333984375, "learning_rate": 9.99542969713579e-06, "loss": 0.7967, "step": 4070 }, { "epoch": 0.33769243502731333, "grad_norm": 5.429426670074463, "learning_rate": 9.995221542453958e-06, "loss": 0.615, "step": 4080 }, { "epoch": 0.338520112564145, "grad_norm": 7.401728630065918, "learning_rate": 9.995008754797455e-06, "loss": 0.6678, "step": 4090 }, { "epoch": 0.33934779010097665, "grad_norm": 5.658791542053223, "learning_rate": 9.994791334363642e-06, "loss": 0.7634, "step": 4100 }, { "epoch": 0.3401754676378083, "grad_norm": 7.103274822235107, "learning_rate": 9.99456928135417e-06, "loss": 0.7266, "step": 4110 }, { "epoch": 0.34100314517464, "grad_norm": 6.673758506774902, "learning_rate": 9.994342595974989e-06, "loss": 0.6051, "step": 4120 }, { "epoch": 0.34183082271147164, "grad_norm": 4.926333427429199, "learning_rate": 9.994111278436345e-06, "loss": 0.6602, "step": 4130 }, { "epoch": 0.34265850024830324, "grad_norm": 6.293793678283691, "learning_rate": 9.993875328952781e-06, "loss": 0.7702, "step": 4140 }, { "epoch": 0.3434861777851349, "grad_norm": 8.074064254760742, "learning_rate": 9.993634747743137e-06, "loss": 0.6997, "step": 4150 }, { "epoch": 0.34431385532196657, "grad_norm": 6.637765884399414, "learning_rate": 9.993389535030546e-06, "loss": 0.6603, "step": 4160 }, { "epoch": 0.34514153285879823, "grad_norm": 5.805063724517822, "learning_rate": 9.993139691042442e-06, "loss": 0.7853, "step": 4170 }, { "epoch": 0.3459692103956299, "grad_norm": 5.482685565948486, "learning_rate": 9.992885216010543e-06, "loss": 0.7896, "step": 4180 }, { "epoch": 0.3467968879324615, "grad_norm": 6.267633438110352, "learning_rate": 9.99262611017088e-06, "loss": 0.7893, "step": 4190 }, { "epoch": 0.34762456546929316, "grad_norm": 5.804652214050293, "learning_rate": 9.99236237376376e-06, "loss": 0.754, "step": 4200 }, { "epoch": 0.3484522430061248, "grad_norm": 4.803692817687988, "learning_rate": 9.992094007033796e-06, "loss": 0.7777, "step": 4210 }, { "epoch": 0.3492799205429565, "grad_norm": 5.312870025634766, "learning_rate": 9.991821010229895e-06, "loss": 0.747, "step": 4220 }, { "epoch": 0.35010759807978814, "grad_norm": 6.852916240692139, "learning_rate": 9.991543383605255e-06, "loss": 0.7287, "step": 4230 }, { "epoch": 0.35093527561661975, "grad_norm": 5.714303493499756, "learning_rate": 9.991261127417371e-06, "loss": 0.6959, "step": 4240 }, { "epoch": 0.3517629531534514, "grad_norm": 6.861367702484131, "learning_rate": 9.99097424192803e-06, "loss": 0.7218, "step": 4250 }, { "epoch": 0.35259063069028307, "grad_norm": 6.675659656524658, "learning_rate": 9.990682727403311e-06, "loss": 0.727, "step": 4260 }, { "epoch": 0.35341830822711473, "grad_norm": 6.21479606628418, "learning_rate": 9.99038658411359e-06, "loss": 0.7258, "step": 4270 }, { "epoch": 0.3542459857639464, "grad_norm": 5.704850196838379, "learning_rate": 9.990085812333533e-06, "loss": 0.7083, "step": 4280 }, { "epoch": 0.355073663300778, "grad_norm": 4.860738754272461, "learning_rate": 9.9897804123421e-06, "loss": 0.7264, "step": 4290 }, { "epoch": 0.35590134083760966, "grad_norm": 5.862810134887695, "learning_rate": 9.989470384422547e-06, "loss": 0.762, "step": 4300 }, { "epoch": 0.3567290183744413, "grad_norm": 5.577769756317139, "learning_rate": 9.989155728862416e-06, "loss": 0.7228, "step": 4310 }, { "epoch": 0.357556695911273, "grad_norm": 6.874015808105469, "learning_rate": 9.988836445953546e-06, "loss": 0.7959, "step": 4320 }, { "epoch": 0.35838437344810464, "grad_norm": 7.093836784362793, "learning_rate": 9.988512535992065e-06, "loss": 0.7216, "step": 4330 }, { "epoch": 0.35921205098493625, "grad_norm": 5.978187561035156, "learning_rate": 9.988183999278392e-06, "loss": 0.6723, "step": 4340 }, { "epoch": 0.3600397285217679, "grad_norm": 7.46330451965332, "learning_rate": 9.987850836117242e-06, "loss": 0.7186, "step": 4350 }, { "epoch": 0.36086740605859957, "grad_norm": 7.534542083740234, "learning_rate": 9.987513046817614e-06, "loss": 0.7281, "step": 4360 }, { "epoch": 0.36169508359543123, "grad_norm": 7.389641284942627, "learning_rate": 9.987170631692804e-06, "loss": 0.764, "step": 4370 }, { "epoch": 0.3625227611322629, "grad_norm": 4.504805088043213, "learning_rate": 9.986823591060396e-06, "loss": 0.6577, "step": 4380 }, { "epoch": 0.3633504386690945, "grad_norm": 4.492609024047852, "learning_rate": 9.98647192524226e-06, "loss": 0.7604, "step": 4390 }, { "epoch": 0.36417811620592616, "grad_norm": 7.624553203582764, "learning_rate": 9.986115634564563e-06, "loss": 0.7731, "step": 4400 }, { "epoch": 0.3650057937427578, "grad_norm": 6.383185863494873, "learning_rate": 9.985754719357756e-06, "loss": 0.6792, "step": 4410 }, { "epoch": 0.3658334712795895, "grad_norm": 6.292774677276611, "learning_rate": 9.985389179956584e-06, "loss": 0.7471, "step": 4420 }, { "epoch": 0.36666114881642115, "grad_norm": 10.477404594421387, "learning_rate": 9.985019016700073e-06, "loss": 0.6645, "step": 4430 }, { "epoch": 0.36748882635325275, "grad_norm": 6.952937602996826, "learning_rate": 9.984644229931549e-06, "loss": 0.744, "step": 4440 }, { "epoch": 0.3683165038900844, "grad_norm": 6.131562232971191, "learning_rate": 9.984264819998612e-06, "loss": 0.6548, "step": 4450 }, { "epoch": 0.3691441814269161, "grad_norm": 9.081440925598145, "learning_rate": 9.983880787253165e-06, "loss": 0.6641, "step": 4460 }, { "epoch": 0.36997185896374774, "grad_norm": 5.294651508331299, "learning_rate": 9.983492132051386e-06, "loss": 0.7558, "step": 4470 }, { "epoch": 0.3707995365005794, "grad_norm": 5.512598991394043, "learning_rate": 9.983098854753748e-06, "loss": 0.7349, "step": 4480 }, { "epoch": 0.371627214037411, "grad_norm": 6.285693645477295, "learning_rate": 9.982700955725008e-06, "loss": 0.7047, "step": 4490 }, { "epoch": 0.37245489157424266, "grad_norm": 4.733898162841797, "learning_rate": 9.982298435334208e-06, "loss": 0.744, "step": 4500 }, { "epoch": 0.3732825691110743, "grad_norm": 6.339005947113037, "learning_rate": 9.981891293954682e-06, "loss": 0.5781, "step": 4510 }, { "epoch": 0.374110246647906, "grad_norm": 7.253324508666992, "learning_rate": 9.981479531964042e-06, "loss": 0.6846, "step": 4520 }, { "epoch": 0.37493792418473765, "grad_norm": 6.09147310256958, "learning_rate": 9.981063149744192e-06, "loss": 0.7472, "step": 4530 }, { "epoch": 0.37576560172156925, "grad_norm": 5.909043312072754, "learning_rate": 9.980642147681318e-06, "loss": 0.8522, "step": 4540 }, { "epoch": 0.3765932792584009, "grad_norm": 4.500476837158203, "learning_rate": 9.980216526165893e-06, "loss": 0.66, "step": 4550 }, { "epoch": 0.3774209567952326, "grad_norm": 5.849154949188232, "learning_rate": 9.979786285592671e-06, "loss": 0.8166, "step": 4560 }, { "epoch": 0.37824863433206424, "grad_norm": 3.882854700088501, "learning_rate": 9.979351426360692e-06, "loss": 0.7157, "step": 4570 }, { "epoch": 0.3790763118688959, "grad_norm": 4.510802745819092, "learning_rate": 9.97891194887328e-06, "loss": 0.7884, "step": 4580 }, { "epoch": 0.3799039894057275, "grad_norm": 8.001437187194824, "learning_rate": 9.978467853538046e-06, "loss": 0.5479, "step": 4590 }, { "epoch": 0.38073166694255917, "grad_norm": 6.834483623504639, "learning_rate": 9.978019140766873e-06, "loss": 0.754, "step": 4600 }, { "epoch": 0.3815593444793908, "grad_norm": 6.045119762420654, "learning_rate": 9.97756581097594e-06, "loss": 0.6905, "step": 4610 }, { "epoch": 0.3823870220162225, "grad_norm": 9.374484062194824, "learning_rate": 9.977107864585698e-06, "loss": 0.7692, "step": 4620 }, { "epoch": 0.38321469955305415, "grad_norm": 7.437722682952881, "learning_rate": 9.976645302020885e-06, "loss": 0.629, "step": 4630 }, { "epoch": 0.38404237708988576, "grad_norm": 6.0828351974487305, "learning_rate": 9.97617812371052e-06, "loss": 0.6515, "step": 4640 }, { "epoch": 0.3848700546267174, "grad_norm": 6.885842323303223, "learning_rate": 9.975706330087902e-06, "loss": 0.764, "step": 4650 }, { "epoch": 0.3856977321635491, "grad_norm": 7.088222503662109, "learning_rate": 9.97522992159061e-06, "loss": 0.7269, "step": 4660 }, { "epoch": 0.38652540970038074, "grad_norm": 3.233670234680176, "learning_rate": 9.974748898660504e-06, "loss": 0.6803, "step": 4670 }, { "epoch": 0.3873530872372124, "grad_norm": 6.678426265716553, "learning_rate": 9.974263261743724e-06, "loss": 0.6656, "step": 4680 }, { "epoch": 0.388180764774044, "grad_norm": 5.749527454376221, "learning_rate": 9.973773011290689e-06, "loss": 0.7997, "step": 4690 }, { "epoch": 0.38900844231087567, "grad_norm": 7.757101535797119, "learning_rate": 9.973278147756097e-06, "loss": 0.6678, "step": 4700 }, { "epoch": 0.38983611984770733, "grad_norm": 5.96592378616333, "learning_rate": 9.972778671598924e-06, "loss": 0.7844, "step": 4710 }, { "epoch": 0.390663797384539, "grad_norm": 5.997199535369873, "learning_rate": 9.972274583282429e-06, "loss": 0.769, "step": 4720 }, { "epoch": 0.39149147492137065, "grad_norm": 4.279834270477295, "learning_rate": 9.971765883274139e-06, "loss": 0.6823, "step": 4730 }, { "epoch": 0.39231915245820226, "grad_norm": 5.608923435211182, "learning_rate": 9.971252572045867e-06, "loss": 0.6729, "step": 4740 }, { "epoch": 0.3931468299950339, "grad_norm": 5.442038536071777, "learning_rate": 9.970734650073698e-06, "loss": 0.7235, "step": 4750 }, { "epoch": 0.3939745075318656, "grad_norm": 6.961414337158203, "learning_rate": 9.970212117837996e-06, "loss": 0.6645, "step": 4760 }, { "epoch": 0.39480218506869724, "grad_norm": 4.804913520812988, "learning_rate": 9.969684975823399e-06, "loss": 0.6664, "step": 4770 }, { "epoch": 0.3956298626055289, "grad_norm": 3.72979736328125, "learning_rate": 9.969153224518821e-06, "loss": 0.6351, "step": 4780 }, { "epoch": 0.3964575401423605, "grad_norm": 5.864390850067139, "learning_rate": 9.968616864417453e-06, "loss": 0.8023, "step": 4790 }, { "epoch": 0.39728521767919217, "grad_norm": 5.433023929595947, "learning_rate": 9.96807589601676e-06, "loss": 0.6544, "step": 4800 }, { "epoch": 0.39811289521602383, "grad_norm": 4.795073986053467, "learning_rate": 9.967530319818476e-06, "loss": 0.7944, "step": 4810 }, { "epoch": 0.3989405727528555, "grad_norm": 6.124776840209961, "learning_rate": 9.966980136328616e-06, "loss": 0.7689, "step": 4820 }, { "epoch": 0.39976825028968715, "grad_norm": 6.205166339874268, "learning_rate": 9.966425346057464e-06, "loss": 0.5698, "step": 4830 }, { "epoch": 0.40059592782651876, "grad_norm": 7.460726737976074, "learning_rate": 9.965865949519578e-06, "loss": 0.7116, "step": 4840 }, { "epoch": 0.4014236053633504, "grad_norm": 6.210309982299805, "learning_rate": 9.965301947233788e-06, "loss": 0.709, "step": 4850 }, { "epoch": 0.4022512829001821, "grad_norm": 5.577939987182617, "learning_rate": 9.964733339723194e-06, "loss": 0.6873, "step": 4860 }, { "epoch": 0.40307896043701374, "grad_norm": 4.457425594329834, "learning_rate": 9.964160127515168e-06, "loss": 0.6759, "step": 4870 }, { "epoch": 0.4039066379738454, "grad_norm": 5.732440948486328, "learning_rate": 9.963582311141358e-06, "loss": 0.6679, "step": 4880 }, { "epoch": 0.40473431551067707, "grad_norm": 5.150137424468994, "learning_rate": 9.962999891137675e-06, "loss": 0.7305, "step": 4890 }, { "epoch": 0.4055619930475087, "grad_norm": 5.604026794433594, "learning_rate": 9.962412868044302e-06, "loss": 0.7039, "step": 4900 }, { "epoch": 0.40638967058434033, "grad_norm": 5.2866291999816895, "learning_rate": 9.961821242405694e-06, "loss": 0.6253, "step": 4910 }, { "epoch": 0.407217348121172, "grad_norm": 9.58226490020752, "learning_rate": 9.96122501477057e-06, "loss": 0.6488, "step": 4920 }, { "epoch": 0.40804502565800366, "grad_norm": 5.246994495391846, "learning_rate": 9.960624185691924e-06, "loss": 0.6898, "step": 4930 }, { "epoch": 0.4088727031948353, "grad_norm": 6.933422565460205, "learning_rate": 9.96001875572701e-06, "loss": 0.6136, "step": 4940 }, { "epoch": 0.4097003807316669, "grad_norm": 6.661437511444092, "learning_rate": 9.959408725437355e-06, "loss": 0.6694, "step": 4950 }, { "epoch": 0.4105280582684986, "grad_norm": 5.600121974945068, "learning_rate": 9.95879409538875e-06, "loss": 0.671, "step": 4960 }, { "epoch": 0.41135573580533025, "grad_norm": 6.544322490692139, "learning_rate": 9.958174866151253e-06, "loss": 0.6361, "step": 4970 }, { "epoch": 0.4121834133421619, "grad_norm": 5.7647385597229, "learning_rate": 9.957551038299186e-06, "loss": 0.7424, "step": 4980 }, { "epoch": 0.41301109087899357, "grad_norm": 4.860726833343506, "learning_rate": 9.956922612411141e-06, "loss": 0.7183, "step": 4990 }, { "epoch": 0.4138387684158252, "grad_norm": 5.92608642578125, "learning_rate": 9.95628958906997e-06, "loss": 0.7606, "step": 5000 }, { "epoch": 0.41466644595265684, "grad_norm": 10.975728988647461, "learning_rate": 9.95565196886279e-06, "loss": 0.7561, "step": 5010 }, { "epoch": 0.4154941234894885, "grad_norm": 6.202707767486572, "learning_rate": 9.95500975238098e-06, "loss": 0.7272, "step": 5020 }, { "epoch": 0.41632180102632016, "grad_norm": 5.936666488647461, "learning_rate": 9.954362940220185e-06, "loss": 0.6159, "step": 5030 }, { "epoch": 0.4171494785631518, "grad_norm": 10.496636390686035, "learning_rate": 9.953711532980313e-06, "loss": 0.6799, "step": 5040 }, { "epoch": 0.4179771560999834, "grad_norm": 5.334098815917969, "learning_rate": 9.953055531265532e-06, "loss": 0.7072, "step": 5050 }, { "epoch": 0.4188048336368151, "grad_norm": 5.6064066886901855, "learning_rate": 9.952394935684268e-06, "loss": 0.7161, "step": 5060 }, { "epoch": 0.41963251117364675, "grad_norm": 6.289244651794434, "learning_rate": 9.951729746849213e-06, "loss": 0.5707, "step": 5070 }, { "epoch": 0.4204601887104784, "grad_norm": 6.245842933654785, "learning_rate": 9.951059965377319e-06, "loss": 0.6655, "step": 5080 }, { "epoch": 0.42128786624731007, "grad_norm": 5.4839067459106445, "learning_rate": 9.950385591889792e-06, "loss": 0.6738, "step": 5090 }, { "epoch": 0.4221155437841417, "grad_norm": 6.144285202026367, "learning_rate": 9.949706627012104e-06, "loss": 0.6234, "step": 5100 }, { "epoch": 0.42294322132097334, "grad_norm": 5.049564838409424, "learning_rate": 9.949023071373981e-06, "loss": 0.6669, "step": 5110 }, { "epoch": 0.423770898857805, "grad_norm": 5.704695224761963, "learning_rate": 9.94833492560941e-06, "loss": 0.6709, "step": 5120 }, { "epoch": 0.42459857639463666, "grad_norm": 6.166596412658691, "learning_rate": 9.947642190356629e-06, "loss": 0.6273, "step": 5130 }, { "epoch": 0.4254262539314683, "grad_norm": 6.706107139587402, "learning_rate": 9.946944866258142e-06, "loss": 0.7055, "step": 5140 }, { "epoch": 0.42625393146829993, "grad_norm": 4.903844833374023, "learning_rate": 9.946242953960699e-06, "loss": 0.7541, "step": 5150 }, { "epoch": 0.4270816090051316, "grad_norm": 5.526392459869385, "learning_rate": 9.945536454115314e-06, "loss": 0.6393, "step": 5160 }, { "epoch": 0.42790928654196325, "grad_norm": 6.188036918640137, "learning_rate": 9.944825367377253e-06, "loss": 0.5986, "step": 5170 }, { "epoch": 0.4287369640787949, "grad_norm": 6.459216594696045, "learning_rate": 9.944109694406033e-06, "loss": 0.6954, "step": 5180 }, { "epoch": 0.4295646416156266, "grad_norm": 5.400817394256592, "learning_rate": 9.94338943586543e-06, "loss": 0.6806, "step": 5190 }, { "epoch": 0.4303923191524582, "grad_norm": 4.606400489807129, "learning_rate": 9.942664592423468e-06, "loss": 0.7708, "step": 5200 }, { "epoch": 0.43121999668928984, "grad_norm": 8.056431770324707, "learning_rate": 9.941935164752426e-06, "loss": 0.7615, "step": 5210 }, { "epoch": 0.4320476742261215, "grad_norm": 6.622241497039795, "learning_rate": 9.941201153528837e-06, "loss": 0.677, "step": 5220 }, { "epoch": 0.43287535176295316, "grad_norm": 7.080117702484131, "learning_rate": 9.94046255943348e-06, "loss": 0.6348, "step": 5230 }, { "epoch": 0.4337030292997848, "grad_norm": 6.572682857513428, "learning_rate": 9.939719383151387e-06, "loss": 0.7575, "step": 5240 }, { "epoch": 0.43453070683661643, "grad_norm": 6.5756378173828125, "learning_rate": 9.938971625371843e-06, "loss": 0.7285, "step": 5250 }, { "epoch": 0.4353583843734481, "grad_norm": 5.9012131690979, "learning_rate": 9.938219286788376e-06, "loss": 0.6963, "step": 5260 }, { "epoch": 0.43618606191027975, "grad_norm": 4.027134895324707, "learning_rate": 9.937462368098769e-06, "loss": 0.6593, "step": 5270 }, { "epoch": 0.4370137394471114, "grad_norm": 5.749174118041992, "learning_rate": 9.936700870005048e-06, "loss": 0.7638, "step": 5280 }, { "epoch": 0.4378414169839431, "grad_norm": 6.383908271789551, "learning_rate": 9.935934793213489e-06, "loss": 0.707, "step": 5290 }, { "epoch": 0.4386690945207747, "grad_norm": 6.214728832244873, "learning_rate": 9.935164138434614e-06, "loss": 0.7306, "step": 5300 }, { "epoch": 0.43949677205760634, "grad_norm": 4.899090766906738, "learning_rate": 9.93438890638319e-06, "loss": 0.6846, "step": 5310 }, { "epoch": 0.440324449594438, "grad_norm": 4.123104572296143, "learning_rate": 9.933609097778231e-06, "loss": 0.7412, "step": 5320 }, { "epoch": 0.44115212713126967, "grad_norm": 5.618980407714844, "learning_rate": 9.932824713342996e-06, "loss": 0.634, "step": 5330 }, { "epoch": 0.44197980466810133, "grad_norm": 5.218521595001221, "learning_rate": 9.932035753804986e-06, "loss": 0.6788, "step": 5340 }, { "epoch": 0.44280748220493293, "grad_norm": 5.863386631011963, "learning_rate": 9.931242219895944e-06, "loss": 0.707, "step": 5350 }, { "epoch": 0.4436351597417646, "grad_norm": 3.221190929412842, "learning_rate": 9.930444112351862e-06, "loss": 0.5858, "step": 5360 }, { "epoch": 0.44446283727859626, "grad_norm": 7.146638870239258, "learning_rate": 9.929641431912967e-06, "loss": 0.7103, "step": 5370 }, { "epoch": 0.4452905148154279, "grad_norm": 4.422207355499268, "learning_rate": 9.92883417932373e-06, "loss": 0.6865, "step": 5380 }, { "epoch": 0.4461181923522596, "grad_norm": 6.6922383308410645, "learning_rate": 9.928022355332863e-06, "loss": 0.6457, "step": 5390 }, { "epoch": 0.4469458698890912, "grad_norm": 8.881153106689453, "learning_rate": 9.92720596069332e-06, "loss": 0.8087, "step": 5400 }, { "epoch": 0.44777354742592285, "grad_norm": 4.743436813354492, "learning_rate": 9.926384996162288e-06, "loss": 0.7111, "step": 5410 }, { "epoch": 0.4486012249627545, "grad_norm": 3.605637311935425, "learning_rate": 9.9255594625012e-06, "loss": 0.6214, "step": 5420 }, { "epoch": 0.44942890249958617, "grad_norm": 5.08591890335083, "learning_rate": 9.92472936047572e-06, "loss": 0.6059, "step": 5430 }, { "epoch": 0.45025658003641783, "grad_norm": 5.223016738891602, "learning_rate": 9.923894690855754e-06, "loss": 0.7179, "step": 5440 }, { "epoch": 0.45108425757324944, "grad_norm": 6.13178825378418, "learning_rate": 9.92305545441544e-06, "loss": 0.6887, "step": 5450 }, { "epoch": 0.4519119351100811, "grad_norm": 4.896916389465332, "learning_rate": 9.922211651933156e-06, "loss": 0.6891, "step": 5460 }, { "epoch": 0.45273961264691276, "grad_norm": 6.236217021942139, "learning_rate": 9.921363284191516e-06, "loss": 0.6783, "step": 5470 }, { "epoch": 0.4535672901837444, "grad_norm": 5.643880367279053, "learning_rate": 9.92051035197736e-06, "loss": 0.6322, "step": 5480 }, { "epoch": 0.4543949677205761, "grad_norm": 5.628150939941406, "learning_rate": 9.919652856081768e-06, "loss": 0.6136, "step": 5490 }, { "epoch": 0.4552226452574077, "grad_norm": 4.235380172729492, "learning_rate": 9.918790797300053e-06, "loss": 0.6197, "step": 5500 }, { "epoch": 0.45605032279423935, "grad_norm": 7.789366245269775, "learning_rate": 9.91792417643176e-06, "loss": 0.5841, "step": 5510 }, { "epoch": 0.456878000331071, "grad_norm": 4.543807029724121, "learning_rate": 9.917052994280657e-06, "loss": 0.7574, "step": 5520 }, { "epoch": 0.45770567786790267, "grad_norm": 5.39536714553833, "learning_rate": 9.916177251654758e-06, "loss": 0.605, "step": 5530 }, { "epoch": 0.45853335540473433, "grad_norm": 4.011848449707031, "learning_rate": 9.915296949366291e-06, "loss": 0.7906, "step": 5540 }, { "epoch": 0.45936103294156594, "grad_norm": 4.532012939453125, "learning_rate": 9.914412088231723e-06, "loss": 0.6526, "step": 5550 }, { "epoch": 0.4601887104783976, "grad_norm": 6.959229469299316, "learning_rate": 9.913522669071746e-06, "loss": 0.7189, "step": 5560 }, { "epoch": 0.46101638801522926, "grad_norm": 5.632550239562988, "learning_rate": 9.912628692711278e-06, "loss": 0.6649, "step": 5570 }, { "epoch": 0.4618440655520609, "grad_norm": 5.071171283721924, "learning_rate": 9.911730159979468e-06, "loss": 0.652, "step": 5580 }, { "epoch": 0.4626717430888926, "grad_norm": 6.410516738891602, "learning_rate": 9.910827071709686e-06, "loss": 0.7791, "step": 5590 }, { "epoch": 0.46349942062572425, "grad_norm": 8.320860862731934, "learning_rate": 9.909919428739531e-06, "loss": 0.7112, "step": 5600 }, { "epoch": 0.46432709816255585, "grad_norm": 18.287456512451172, "learning_rate": 9.909007231910823e-06, "loss": 0.6883, "step": 5610 }, { "epoch": 0.4651547756993875, "grad_norm": 6.0761284828186035, "learning_rate": 9.90809048206961e-06, "loss": 0.7838, "step": 5620 }, { "epoch": 0.4659824532362192, "grad_norm": 5.656871318817139, "learning_rate": 9.90716918006616e-06, "loss": 0.6585, "step": 5630 }, { "epoch": 0.46681013077305084, "grad_norm": 6.063356399536133, "learning_rate": 9.906243326754961e-06, "loss": 0.6986, "step": 5640 }, { "epoch": 0.4676378083098825, "grad_norm": 9.047858238220215, "learning_rate": 9.905312922994727e-06, "loss": 0.6946, "step": 5650 }, { "epoch": 0.4684654858467141, "grad_norm": 5.150297164916992, "learning_rate": 9.90437796964839e-06, "loss": 0.6017, "step": 5660 }, { "epoch": 0.46929316338354576, "grad_norm": 5.0125322341918945, "learning_rate": 9.9034384675831e-06, "loss": 0.6563, "step": 5670 }, { "epoch": 0.4701208409203774, "grad_norm": 5.321043968200684, "learning_rate": 9.90249441767023e-06, "loss": 0.5484, "step": 5680 }, { "epoch": 0.4709485184572091, "grad_norm": 10.758487701416016, "learning_rate": 9.901545820785366e-06, "loss": 0.7379, "step": 5690 }, { "epoch": 0.47177619599404075, "grad_norm": 5.563775062561035, "learning_rate": 9.900592677808317e-06, "loss": 0.6725, "step": 5700 }, { "epoch": 0.47260387353087235, "grad_norm": 4.343594074249268, "learning_rate": 9.899634989623101e-06, "loss": 0.6522, "step": 5710 }, { "epoch": 0.473431551067704, "grad_norm": 4.106492042541504, "learning_rate": 9.89867275711796e-06, "loss": 0.6271, "step": 5720 }, { "epoch": 0.4742592286045357, "grad_norm": 4.387215614318848, "learning_rate": 9.897705981185347e-06, "loss": 0.6813, "step": 5730 }, { "epoch": 0.47508690614136734, "grad_norm": 6.3960185050964355, "learning_rate": 9.896734662721923e-06, "loss": 0.688, "step": 5740 }, { "epoch": 0.475914583678199, "grad_norm": 5.252898693084717, "learning_rate": 9.895758802628574e-06, "loss": 0.6855, "step": 5750 }, { "epoch": 0.4767422612150306, "grad_norm": 5.79520845413208, "learning_rate": 9.894778401810389e-06, "loss": 0.7192, "step": 5760 }, { "epoch": 0.47756993875186227, "grad_norm": 5.353193759918213, "learning_rate": 9.893793461176672e-06, "loss": 0.6352, "step": 5770 }, { "epoch": 0.4783976162886939, "grad_norm": 10.204635620117188, "learning_rate": 9.892803981640937e-06, "loss": 0.6743, "step": 5780 }, { "epoch": 0.4792252938255256, "grad_norm": 7.3138885498046875, "learning_rate": 9.891809964120909e-06, "loss": 0.6518, "step": 5790 }, { "epoch": 0.48005297136235725, "grad_norm": 4.75705623626709, "learning_rate": 9.890811409538516e-06, "loss": 0.7214, "step": 5800 }, { "epoch": 0.48088064889918886, "grad_norm": 6.143803119659424, "learning_rate": 9.889808318819906e-06, "loss": 0.6597, "step": 5810 }, { "epoch": 0.4817083264360205, "grad_norm": 5.962342739105225, "learning_rate": 9.888800692895422e-06, "loss": 0.6564, "step": 5820 }, { "epoch": 0.4825360039728522, "grad_norm": 5.2933735847473145, "learning_rate": 9.887788532699618e-06, "loss": 0.6211, "step": 5830 }, { "epoch": 0.48336368150968384, "grad_norm": 6.206865310668945, "learning_rate": 9.886771839171256e-06, "loss": 0.6513, "step": 5840 }, { "epoch": 0.4841913590465155, "grad_norm": 5.648596286773682, "learning_rate": 9.8857506132533e-06, "loss": 0.6189, "step": 5850 }, { "epoch": 0.4850190365833471, "grad_norm": 6.334989070892334, "learning_rate": 9.884724855892915e-06, "loss": 0.6881, "step": 5860 }, { "epoch": 0.48584671412017877, "grad_norm": 4.489048480987549, "learning_rate": 9.883694568041476e-06, "loss": 0.7137, "step": 5870 }, { "epoch": 0.48667439165701043, "grad_norm": 6.012289524078369, "learning_rate": 9.882659750654552e-06, "loss": 0.6127, "step": 5880 }, { "epoch": 0.4875020691938421, "grad_norm": 6.266094207763672, "learning_rate": 9.88162040469192e-06, "loss": 0.6599, "step": 5890 }, { "epoch": 0.48832974673067375, "grad_norm": 4.165154933929443, "learning_rate": 9.880576531117553e-06, "loss": 0.6113, "step": 5900 }, { "epoch": 0.48915742426750536, "grad_norm": 5.808906555175781, "learning_rate": 9.879528130899622e-06, "loss": 0.6549, "step": 5910 }, { "epoch": 0.489985101804337, "grad_norm": 4.528868675231934, "learning_rate": 9.878475205010503e-06, "loss": 0.6022, "step": 5920 }, { "epoch": 0.4908127793411687, "grad_norm": 5.325387001037598, "learning_rate": 9.87741775442676e-06, "loss": 0.6598, "step": 5930 }, { "epoch": 0.49164045687800034, "grad_norm": 6.4886016845703125, "learning_rate": 9.876355780129162e-06, "loss": 0.6558, "step": 5940 }, { "epoch": 0.492468134414832, "grad_norm": 15.435273170471191, "learning_rate": 9.87528928310267e-06, "loss": 0.6841, "step": 5950 }, { "epoch": 0.4932958119516636, "grad_norm": 5.9535322189331055, "learning_rate": 9.874218264336439e-06, "loss": 0.5762, "step": 5960 }, { "epoch": 0.49412348948849527, "grad_norm": 6.465355396270752, "learning_rate": 9.873142724823818e-06, "loss": 0.7525, "step": 5970 }, { "epoch": 0.49495116702532693, "grad_norm": 5.203250408172607, "learning_rate": 9.872062665562353e-06, "loss": 0.7227, "step": 5980 }, { "epoch": 0.4957788445621586, "grad_norm": 6.257732391357422, "learning_rate": 9.870978087553776e-06, "loss": 0.7561, "step": 5990 }, { "epoch": 0.49660652209899026, "grad_norm": 5.868444919586182, "learning_rate": 9.869888991804016e-06, "loss": 0.7649, "step": 6000 }, { "epoch": 0.49743419963582186, "grad_norm": 5.705538272857666, "learning_rate": 9.868795379323184e-06, "loss": 0.6257, "step": 6010 }, { "epoch": 0.4982618771726535, "grad_norm": 5.158013820648193, "learning_rate": 9.86769725112559e-06, "loss": 0.5543, "step": 6020 }, { "epoch": 0.4990895547094852, "grad_norm": 4.2921833992004395, "learning_rate": 9.866594608229723e-06, "loss": 0.6472, "step": 6030 }, { "epoch": 0.49991723224631684, "grad_norm": 5.217587471008301, "learning_rate": 9.865487451658268e-06, "loss": 0.7544, "step": 6040 }, { "epoch": 0.5007449097831485, "grad_norm": 6.176877498626709, "learning_rate": 9.86437578243809e-06, "loss": 0.6268, "step": 6050 }, { "epoch": 0.5015725873199801, "grad_norm": 5.464317798614502, "learning_rate": 9.863259601600242e-06, "loss": 0.674, "step": 6060 }, { "epoch": 0.5024002648568118, "grad_norm": 4.816394805908203, "learning_rate": 9.862138910179959e-06, "loss": 0.7372, "step": 6070 }, { "epoch": 0.5032279423936434, "grad_norm": 6.461970806121826, "learning_rate": 9.861013709216663e-06, "loss": 0.7469, "step": 6080 }, { "epoch": 0.5040556199304751, "grad_norm": 8.292777061462402, "learning_rate": 9.859883999753957e-06, "loss": 0.6711, "step": 6090 }, { "epoch": 0.5048832974673068, "grad_norm": 6.818353176116943, "learning_rate": 9.858749782839624e-06, "loss": 0.6265, "step": 6100 }, { "epoch": 0.5057109750041384, "grad_norm": 5.117384433746338, "learning_rate": 9.857611059525632e-06, "loss": 0.6615, "step": 6110 }, { "epoch": 0.5065386525409701, "grad_norm": 5.265927314758301, "learning_rate": 9.856467830868122e-06, "loss": 0.6189, "step": 6120 }, { "epoch": 0.5073663300778017, "grad_norm": 5.611670970916748, "learning_rate": 9.855320097927419e-06, "loss": 0.7894, "step": 6130 }, { "epoch": 0.5081940076146333, "grad_norm": 5.511541366577148, "learning_rate": 9.854167861768025e-06, "loss": 0.6137, "step": 6140 }, { "epoch": 0.509021685151465, "grad_norm": 5.142088413238525, "learning_rate": 9.853011123458615e-06, "loss": 0.7202, "step": 6150 }, { "epoch": 0.5098493626882966, "grad_norm": 5.705070972442627, "learning_rate": 9.851849884072043e-06, "loss": 0.6878, "step": 6160 }, { "epoch": 0.5106770402251283, "grad_norm": 4.986788749694824, "learning_rate": 9.850684144685338e-06, "loss": 0.6777, "step": 6170 }, { "epoch": 0.5115047177619599, "grad_norm": 5.947421073913574, "learning_rate": 9.849513906379703e-06, "loss": 0.7806, "step": 6180 }, { "epoch": 0.5123323952987916, "grad_norm": 5.927761554718018, "learning_rate": 9.848339170240505e-06, "loss": 0.5766, "step": 6190 }, { "epoch": 0.5131600728356233, "grad_norm": 3.962414026260376, "learning_rate": 9.847159937357299e-06, "loss": 0.6622, "step": 6200 }, { "epoch": 0.5139877503724549, "grad_norm": 5.385538101196289, "learning_rate": 9.845976208823796e-06, "loss": 0.676, "step": 6210 }, { "epoch": 0.5148154279092866, "grad_norm": 4.0557966232299805, "learning_rate": 9.844787985737883e-06, "loss": 0.6191, "step": 6220 }, { "epoch": 0.5156431054461182, "grad_norm": 5.51990270614624, "learning_rate": 9.843595269201616e-06, "loss": 0.7428, "step": 6230 }, { "epoch": 0.5164707829829498, "grad_norm": 4.687502861022949, "learning_rate": 9.842398060321215e-06, "loss": 0.6364, "step": 6240 }, { "epoch": 0.5172984605197815, "grad_norm": 4.069591522216797, "learning_rate": 9.84119636020707e-06, "loss": 0.6734, "step": 6250 }, { "epoch": 0.5181261380566131, "grad_norm": 5.85957670211792, "learning_rate": 9.839990169973737e-06, "loss": 0.6478, "step": 6260 }, { "epoch": 0.5189538155934448, "grad_norm": 6.435286045074463, "learning_rate": 9.838779490739933e-06, "loss": 0.6422, "step": 6270 }, { "epoch": 0.5197814931302764, "grad_norm": 5.5521626472473145, "learning_rate": 9.837564323628542e-06, "loss": 0.753, "step": 6280 }, { "epoch": 0.5206091706671081, "grad_norm": 4.386659622192383, "learning_rate": 9.836344669766605e-06, "loss": 0.7484, "step": 6290 }, { "epoch": 0.5214368482039398, "grad_norm": 4.996462345123291, "learning_rate": 9.835120530285332e-06, "loss": 0.6656, "step": 6300 }, { "epoch": 0.5222645257407714, "grad_norm": 10.207401275634766, "learning_rate": 9.83389190632009e-06, "loss": 0.6667, "step": 6310 }, { "epoch": 0.5230922032776031, "grad_norm": 5.229557991027832, "learning_rate": 9.832658799010402e-06, "loss": 0.5891, "step": 6320 }, { "epoch": 0.5239198808144347, "grad_norm": 5.585395812988281, "learning_rate": 9.831421209499952e-06, "loss": 0.702, "step": 6330 }, { "epoch": 0.5247475583512663, "grad_norm": 5.793268203735352, "learning_rate": 9.830179138936584e-06, "loss": 0.676, "step": 6340 }, { "epoch": 0.525575235888098, "grad_norm": 4.831668853759766, "learning_rate": 9.828932588472294e-06, "loss": 0.5911, "step": 6350 }, { "epoch": 0.5264029134249296, "grad_norm": 5.089508533477783, "learning_rate": 9.827681559263231e-06, "loss": 0.6267, "step": 6360 }, { "epoch": 0.5272305909617613, "grad_norm": 6.1107964515686035, "learning_rate": 9.826426052469706e-06, "loss": 0.6595, "step": 6370 }, { "epoch": 0.5280582684985929, "grad_norm": 5.154699802398682, "learning_rate": 9.825166069256178e-06, "loss": 0.6308, "step": 6380 }, { "epoch": 0.5288859460354246, "grad_norm": 7.942720413208008, "learning_rate": 9.823901610791254e-06, "loss": 0.5962, "step": 6390 }, { "epoch": 0.5297136235722563, "grad_norm": 9.431669235229492, "learning_rate": 9.822632678247696e-06, "loss": 0.6195, "step": 6400 }, { "epoch": 0.5305413011090879, "grad_norm": 4.2293243408203125, "learning_rate": 9.821359272802419e-06, "loss": 0.675, "step": 6410 }, { "epoch": 0.5313689786459196, "grad_norm": 5.265641689300537, "learning_rate": 9.82008139563648e-06, "loss": 0.7195, "step": 6420 }, { "epoch": 0.5321966561827512, "grad_norm": 6.055838584899902, "learning_rate": 9.818799047935087e-06, "loss": 0.7053, "step": 6430 }, { "epoch": 0.5330243337195828, "grad_norm": 5.7174577713012695, "learning_rate": 9.817512230887593e-06, "loss": 0.6832, "step": 6440 }, { "epoch": 0.5338520112564145, "grad_norm": 5.204899787902832, "learning_rate": 9.816220945687499e-06, "loss": 0.6979, "step": 6450 }, { "epoch": 0.5346796887932461, "grad_norm": 4.752451419830322, "learning_rate": 9.814925193532446e-06, "loss": 0.6161, "step": 6460 }, { "epoch": 0.5355073663300778, "grad_norm": 6.230308532714844, "learning_rate": 9.813624975624217e-06, "loss": 0.6753, "step": 6470 }, { "epoch": 0.5363350438669094, "grad_norm": 3.307098865509033, "learning_rate": 9.812320293168744e-06, "loss": 0.6255, "step": 6480 }, { "epoch": 0.5371627214037411, "grad_norm": 8.26621150970459, "learning_rate": 9.811011147376096e-06, "loss": 0.688, "step": 6490 }, { "epoch": 0.5379903989405728, "grad_norm": 6.156355381011963, "learning_rate": 9.809697539460478e-06, "loss": 0.6654, "step": 6500 }, { "epoch": 0.5388180764774044, "grad_norm": 7.16554594039917, "learning_rate": 9.808379470640239e-06, "loss": 0.6385, "step": 6510 }, { "epoch": 0.5396457540142361, "grad_norm": 8.444708824157715, "learning_rate": 9.807056942137862e-06, "loss": 0.6388, "step": 6520 }, { "epoch": 0.5404734315510678, "grad_norm": 5.617023944854736, "learning_rate": 9.80572995517997e-06, "loss": 0.6915, "step": 6530 }, { "epoch": 0.5413011090878993, "grad_norm": 7.381171226501465, "learning_rate": 9.804398510997315e-06, "loss": 0.718, "step": 6540 }, { "epoch": 0.542128786624731, "grad_norm": 6.1859893798828125, "learning_rate": 9.80306261082479e-06, "loss": 0.7276, "step": 6550 }, { "epoch": 0.5429564641615626, "grad_norm": 8.374130249023438, "learning_rate": 9.801722255901415e-06, "loss": 0.6179, "step": 6560 }, { "epoch": 0.5437841416983943, "grad_norm": 5.813504219055176, "learning_rate": 9.800377447470344e-06, "loss": 0.6902, "step": 6570 }, { "epoch": 0.544611819235226, "grad_norm": 5.622889041900635, "learning_rate": 9.799028186778865e-06, "loss": 0.7136, "step": 6580 }, { "epoch": 0.5454394967720576, "grad_norm": 7.071735858917236, "learning_rate": 9.797674475078386e-06, "loss": 0.6295, "step": 6590 }, { "epoch": 0.5462671743088893, "grad_norm": 6.300906181335449, "learning_rate": 9.796316313624457e-06, "loss": 0.7518, "step": 6600 }, { "epoch": 0.5470948518457209, "grad_norm": 5.2096967697143555, "learning_rate": 9.794953703676741e-06, "loss": 0.6989, "step": 6610 }, { "epoch": 0.5479225293825526, "grad_norm": 4.588347911834717, "learning_rate": 9.793586646499038e-06, "loss": 0.6224, "step": 6620 }, { "epoch": 0.5487502069193843, "grad_norm": 6.719950199127197, "learning_rate": 9.792215143359264e-06, "loss": 0.6904, "step": 6630 }, { "epoch": 0.5495778844562158, "grad_norm": 4.808568477630615, "learning_rate": 9.790839195529465e-06, "loss": 0.6929, "step": 6640 }, { "epoch": 0.5504055619930475, "grad_norm": 3.853245496749878, "learning_rate": 9.789458804285805e-06, "loss": 0.6304, "step": 6650 }, { "epoch": 0.5512332395298791, "grad_norm": 6.886984825134277, "learning_rate": 9.788073970908573e-06, "loss": 0.714, "step": 6660 }, { "epoch": 0.5520609170667108, "grad_norm": 5.107656955718994, "learning_rate": 9.786684696682177e-06, "loss": 0.6842, "step": 6670 }, { "epoch": 0.5528885946035424, "grad_norm": 6.035403251647949, "learning_rate": 9.78529098289514e-06, "loss": 0.6837, "step": 6680 }, { "epoch": 0.5537162721403741, "grad_norm": 5.778775691986084, "learning_rate": 9.783892830840104e-06, "loss": 0.6372, "step": 6690 }, { "epoch": 0.5545439496772058, "grad_norm": 4.911495208740234, "learning_rate": 9.782490241813836e-06, "loss": 0.6382, "step": 6700 }, { "epoch": 0.5553716272140374, "grad_norm": 7.44188117980957, "learning_rate": 9.781083217117205e-06, "loss": 0.65, "step": 6710 }, { "epoch": 0.5561993047508691, "grad_norm": 6.902495384216309, "learning_rate": 9.779671758055202e-06, "loss": 0.6222, "step": 6720 }, { "epoch": 0.5570269822877008, "grad_norm": 5.617130279541016, "learning_rate": 9.778255865936928e-06, "loss": 0.6506, "step": 6730 }, { "epoch": 0.5578546598245324, "grad_norm": 6.424291133880615, "learning_rate": 9.776835542075597e-06, "loss": 0.7162, "step": 6740 }, { "epoch": 0.558682337361364, "grad_norm": 6.432615280151367, "learning_rate": 9.775410787788536e-06, "loss": 0.5583, "step": 6750 }, { "epoch": 0.5595100148981956, "grad_norm": 5.39361572265625, "learning_rate": 9.773981604397174e-06, "loss": 0.7175, "step": 6760 }, { "epoch": 0.5603376924350273, "grad_norm": 5.559186935424805, "learning_rate": 9.772547993227051e-06, "loss": 0.7692, "step": 6770 }, { "epoch": 0.561165369971859, "grad_norm": 4.8183369636535645, "learning_rate": 9.771109955607817e-06, "loss": 0.7106, "step": 6780 }, { "epoch": 0.5619930475086906, "grad_norm": 4.539303302764893, "learning_rate": 9.769667492873228e-06, "loss": 0.6684, "step": 6790 }, { "epoch": 0.5628207250455223, "grad_norm": 6.130037784576416, "learning_rate": 9.768220606361135e-06, "loss": 0.7311, "step": 6800 }, { "epoch": 0.5636484025823539, "grad_norm": 4.962726593017578, "learning_rate": 9.766769297413501e-06, "loss": 0.7188, "step": 6810 }, { "epoch": 0.5644760801191856, "grad_norm": 4.951356410980225, "learning_rate": 9.76531356737639e-06, "loss": 0.676, "step": 6820 }, { "epoch": 0.5653037576560173, "grad_norm": 6.1915283203125, "learning_rate": 9.763853417599962e-06, "loss": 0.6893, "step": 6830 }, { "epoch": 0.5661314351928489, "grad_norm": 5.163766860961914, "learning_rate": 9.762388849438477e-06, "loss": 0.5016, "step": 6840 }, { "epoch": 0.5669591127296805, "grad_norm": 4.581547737121582, "learning_rate": 9.760919864250298e-06, "loss": 0.6425, "step": 6850 }, { "epoch": 0.5677867902665121, "grad_norm": 6.274184703826904, "learning_rate": 9.75944646339788e-06, "loss": 0.7665, "step": 6860 }, { "epoch": 0.5686144678033438, "grad_norm": 5.6550374031066895, "learning_rate": 9.757968648247777e-06, "loss": 0.6545, "step": 6870 }, { "epoch": 0.5694421453401755, "grad_norm": 7.340173244476318, "learning_rate": 9.75648642017063e-06, "loss": 0.6802, "step": 6880 }, { "epoch": 0.5702698228770071, "grad_norm": 4.7062273025512695, "learning_rate": 9.75499978054118e-06, "loss": 0.6193, "step": 6890 }, { "epoch": 0.5710975004138388, "grad_norm": 7.71263313293457, "learning_rate": 9.753508730738262e-06, "loss": 0.6855, "step": 6900 }, { "epoch": 0.5719251779506704, "grad_norm": 5.460494518280029, "learning_rate": 9.752013272144789e-06, "loss": 0.6678, "step": 6910 }, { "epoch": 0.5727528554875021, "grad_norm": 7.356581211090088, "learning_rate": 9.750513406147777e-06, "loss": 0.6257, "step": 6920 }, { "epoch": 0.5735805330243338, "grad_norm": 5.356894493103027, "learning_rate": 9.74900913413832e-06, "loss": 0.6452, "step": 6930 }, { "epoch": 0.5744082105611654, "grad_norm": 6.302246570587158, "learning_rate": 9.747500457511605e-06, "loss": 0.5535, "step": 6940 }, { "epoch": 0.575235888097997, "grad_norm": 5.575039863586426, "learning_rate": 9.745987377666901e-06, "loss": 0.6463, "step": 6950 }, { "epoch": 0.5760635656348286, "grad_norm": 5.664886951446533, "learning_rate": 9.744469896007559e-06, "loss": 0.7096, "step": 6960 }, { "epoch": 0.5768912431716603, "grad_norm": 4.111787796020508, "learning_rate": 9.742948013941016e-06, "loss": 0.5393, "step": 6970 }, { "epoch": 0.577718920708492, "grad_norm": 6.1224517822265625, "learning_rate": 9.74142173287879e-06, "loss": 0.6762, "step": 6980 }, { "epoch": 0.5785465982453236, "grad_norm": 6.469292163848877, "learning_rate": 9.739891054236479e-06, "loss": 0.6873, "step": 6990 }, { "epoch": 0.5793742757821553, "grad_norm": 5.696608543395996, "learning_rate": 9.738355979433755e-06, "loss": 0.6714, "step": 7000 }, { "epoch": 0.5802019533189869, "grad_norm": 6.1136860847473145, "learning_rate": 9.736816509894375e-06, "loss": 0.6647, "step": 7010 }, { "epoch": 0.5810296308558186, "grad_norm": 5.361754894256592, "learning_rate": 9.735272647046166e-06, "loss": 0.5753, "step": 7020 }, { "epoch": 0.5818573083926503, "grad_norm": 4.675689220428467, "learning_rate": 9.733724392321035e-06, "loss": 0.6511, "step": 7030 }, { "epoch": 0.5826849859294819, "grad_norm": 5.269612789154053, "learning_rate": 9.732171747154954e-06, "loss": 0.7415, "step": 7040 }, { "epoch": 0.5835126634663135, "grad_norm": 4.850591659545898, "learning_rate": 9.730614712987979e-06, "loss": 0.651, "step": 7050 }, { "epoch": 0.5843403410031451, "grad_norm": 5.067111492156982, "learning_rate": 9.729053291264225e-06, "loss": 0.7313, "step": 7060 }, { "epoch": 0.5851680185399768, "grad_norm": 5.280792236328125, "learning_rate": 9.727487483431884e-06, "loss": 0.6982, "step": 7070 }, { "epoch": 0.5859956960768085, "grad_norm": 6.380840301513672, "learning_rate": 9.725917290943213e-06, "loss": 0.686, "step": 7080 }, { "epoch": 0.5868233736136401, "grad_norm": 3.977672815322876, "learning_rate": 9.724342715254534e-06, "loss": 0.5783, "step": 7090 }, { "epoch": 0.5876510511504718, "grad_norm": 7.27426815032959, "learning_rate": 9.722763757826241e-06, "loss": 0.6152, "step": 7100 }, { "epoch": 0.5884787286873034, "grad_norm": 5.565025329589844, "learning_rate": 9.721180420122782e-06, "loss": 0.5747, "step": 7110 }, { "epoch": 0.5893064062241351, "grad_norm": 3.9998064041137695, "learning_rate": 9.719592703612679e-06, "loss": 0.5575, "step": 7120 }, { "epoch": 0.5901340837609668, "grad_norm": 5.093729019165039, "learning_rate": 9.718000609768506e-06, "loss": 0.6458, "step": 7130 }, { "epoch": 0.5909617612977984, "grad_norm": 4.977616310119629, "learning_rate": 9.7164041400669e-06, "loss": 0.7339, "step": 7140 }, { "epoch": 0.59178943883463, "grad_norm": 6.840735912322998, "learning_rate": 9.714803295988557e-06, "loss": 0.7348, "step": 7150 }, { "epoch": 0.5926171163714616, "grad_norm": 5.367594242095947, "learning_rate": 9.713198079018231e-06, "loss": 0.6852, "step": 7160 }, { "epoch": 0.5934447939082933, "grad_norm": 5.409182548522949, "learning_rate": 9.711588490644732e-06, "loss": 0.6117, "step": 7170 }, { "epoch": 0.594272471445125, "grad_norm": 5.51194953918457, "learning_rate": 9.70997453236092e-06, "loss": 0.6335, "step": 7180 }, { "epoch": 0.5951001489819566, "grad_norm": 5.015728950500488, "learning_rate": 9.708356205663714e-06, "loss": 0.6591, "step": 7190 }, { "epoch": 0.5959278265187883, "grad_norm": 4.786473274230957, "learning_rate": 9.706733512054078e-06, "loss": 0.6265, "step": 7200 }, { "epoch": 0.59675550405562, "grad_norm": 5.138513565063477, "learning_rate": 9.705106453037035e-06, "loss": 0.6425, "step": 7210 }, { "epoch": 0.5975831815924516, "grad_norm": 6.516335487365723, "learning_rate": 9.70347503012165e-06, "loss": 0.6361, "step": 7220 }, { "epoch": 0.5984108591292833, "grad_norm": 5.47883415222168, "learning_rate": 9.701839244821034e-06, "loss": 0.6945, "step": 7230 }, { "epoch": 0.5992385366661149, "grad_norm": 4.867238998413086, "learning_rate": 9.700199098652352e-06, "loss": 0.596, "step": 7240 }, { "epoch": 0.6000662142029465, "grad_norm": 5.063808441162109, "learning_rate": 9.698554593136805e-06, "loss": 0.8114, "step": 7250 }, { "epoch": 0.6008938917397781, "grad_norm": 5.4276123046875, "learning_rate": 9.696905729799644e-06, "loss": 0.6292, "step": 7260 }, { "epoch": 0.6017215692766098, "grad_norm": 4.5989251136779785, "learning_rate": 9.695252510170157e-06, "loss": 0.57, "step": 7270 }, { "epoch": 0.6025492468134415, "grad_norm": 6.404171466827393, "learning_rate": 9.693594935781676e-06, "loss": 0.6758, "step": 7280 }, { "epoch": 0.6033769243502731, "grad_norm": 4.350455284118652, "learning_rate": 9.69193300817157e-06, "loss": 0.6792, "step": 7290 }, { "epoch": 0.6042046018871048, "grad_norm": 5.992627143859863, "learning_rate": 9.690266728881246e-06, "loss": 0.6096, "step": 7300 }, { "epoch": 0.6050322794239364, "grad_norm": 3.699326276779175, "learning_rate": 9.688596099456145e-06, "loss": 0.6168, "step": 7310 }, { "epoch": 0.6058599569607681, "grad_norm": 4.657351016998291, "learning_rate": 9.686921121445747e-06, "loss": 0.551, "step": 7320 }, { "epoch": 0.6066876344975998, "grad_norm": 4.114646911621094, "learning_rate": 9.68524179640356e-06, "loss": 0.6973, "step": 7330 }, { "epoch": 0.6075153120344314, "grad_norm": 4.447441101074219, "learning_rate": 9.683558125887131e-06, "loss": 0.6221, "step": 7340 }, { "epoch": 0.608342989571263, "grad_norm": 6.133334636688232, "learning_rate": 9.681870111458027e-06, "loss": 0.7659, "step": 7350 }, { "epoch": 0.6091706671080946, "grad_norm": 4.048940181732178, "learning_rate": 9.680177754681854e-06, "loss": 0.6069, "step": 7360 }, { "epoch": 0.6099983446449263, "grad_norm": 6.409754276275635, "learning_rate": 9.678481057128242e-06, "loss": 0.5963, "step": 7370 }, { "epoch": 0.610826022181758, "grad_norm": 5.653744697570801, "learning_rate": 9.676780020370845e-06, "loss": 0.6555, "step": 7380 }, { "epoch": 0.6116536997185896, "grad_norm": 4.338168144226074, "learning_rate": 9.67507464598734e-06, "loss": 0.5589, "step": 7390 }, { "epoch": 0.6124813772554213, "grad_norm": 4.669665813446045, "learning_rate": 9.673364935559433e-06, "loss": 0.6491, "step": 7400 }, { "epoch": 0.613309054792253, "grad_norm": 5.505494594573975, "learning_rate": 9.67165089067285e-06, "loss": 0.6822, "step": 7410 }, { "epoch": 0.6141367323290846, "grad_norm": 4.319378852844238, "learning_rate": 9.669932512917331e-06, "loss": 0.6578, "step": 7420 }, { "epoch": 0.6149644098659163, "grad_norm": 3.995344400405884, "learning_rate": 9.668209803886642e-06, "loss": 0.6289, "step": 7430 }, { "epoch": 0.6157920874027479, "grad_norm": 5.960126876831055, "learning_rate": 9.666482765178564e-06, "loss": 0.6871, "step": 7440 }, { "epoch": 0.6166197649395795, "grad_norm": 6.423346519470215, "learning_rate": 9.664751398394893e-06, "loss": 0.7615, "step": 7450 }, { "epoch": 0.6174474424764111, "grad_norm": 5.5318145751953125, "learning_rate": 9.663015705141437e-06, "loss": 0.5922, "step": 7460 }, { "epoch": 0.6182751200132428, "grad_norm": 5.26185941696167, "learning_rate": 9.661275687028022e-06, "loss": 0.5741, "step": 7470 }, { "epoch": 0.6191027975500745, "grad_norm": 6.655054569244385, "learning_rate": 9.65953134566848e-06, "loss": 0.6544, "step": 7480 }, { "epoch": 0.6199304750869061, "grad_norm": 5.880560874938965, "learning_rate": 9.657782682680656e-06, "loss": 0.6634, "step": 7490 }, { "epoch": 0.6207581526237378, "grad_norm": 7.350770473480225, "learning_rate": 9.656029699686404e-06, "loss": 0.753, "step": 7500 }, { "epoch": 0.6215858301605695, "grad_norm": 5.856102466583252, "learning_rate": 9.65427239831158e-06, "loss": 0.6681, "step": 7510 }, { "epoch": 0.6224135076974011, "grad_norm": 4.241453170776367, "learning_rate": 9.65251078018605e-06, "loss": 0.686, "step": 7520 }, { "epoch": 0.6232411852342328, "grad_norm": 6.360185623168945, "learning_rate": 9.650744846943681e-06, "loss": 0.6936, "step": 7530 }, { "epoch": 0.6240688627710644, "grad_norm": 6.331421375274658, "learning_rate": 9.648974600222345e-06, "loss": 0.6317, "step": 7540 }, { "epoch": 0.6248965403078961, "grad_norm": 5.513025283813477, "learning_rate": 9.647200041663909e-06, "loss": 0.6833, "step": 7550 }, { "epoch": 0.6257242178447276, "grad_norm": 4.819066047668457, "learning_rate": 9.645421172914246e-06, "loss": 0.6953, "step": 7560 }, { "epoch": 0.6265518953815593, "grad_norm": 4.754744052886963, "learning_rate": 9.643637995623224e-06, "loss": 0.6283, "step": 7570 }, { "epoch": 0.627379572918391, "grad_norm": 5.06771183013916, "learning_rate": 9.641850511444704e-06, "loss": 0.4819, "step": 7580 }, { "epoch": 0.6282072504552226, "grad_norm": 4.523754596710205, "learning_rate": 9.640058722036545e-06, "loss": 0.596, "step": 7590 }, { "epoch": 0.6290349279920543, "grad_norm": 5.100466728210449, "learning_rate": 9.6382626290606e-06, "loss": 0.6473, "step": 7600 }, { "epoch": 0.629862605528886, "grad_norm": 4.914906978607178, "learning_rate": 9.63646223418271e-06, "loss": 0.5828, "step": 7610 }, { "epoch": 0.6306902830657176, "grad_norm": 5.268558979034424, "learning_rate": 9.634657539072708e-06, "loss": 0.7255, "step": 7620 }, { "epoch": 0.6315179606025493, "grad_norm": 5.140072822570801, "learning_rate": 9.632848545404416e-06, "loss": 0.627, "step": 7630 }, { "epoch": 0.6323456381393809, "grad_norm": 5.770638465881348, "learning_rate": 9.63103525485564e-06, "loss": 0.674, "step": 7640 }, { "epoch": 0.6331733156762126, "grad_norm": 5.481020927429199, "learning_rate": 9.629217669108173e-06, "loss": 0.6574, "step": 7650 }, { "epoch": 0.6340009932130442, "grad_norm": 5.604450702667236, "learning_rate": 9.627395789847795e-06, "loss": 0.6686, "step": 7660 }, { "epoch": 0.6348286707498758, "grad_norm": 5.704145908355713, "learning_rate": 9.625569618764261e-06, "loss": 0.6133, "step": 7670 }, { "epoch": 0.6356563482867075, "grad_norm": 6.528250217437744, "learning_rate": 9.623739157551312e-06, "loss": 0.5265, "step": 7680 }, { "epoch": 0.6364840258235391, "grad_norm": 5.382601261138916, "learning_rate": 9.621904407906669e-06, "loss": 0.6297, "step": 7690 }, { "epoch": 0.6373117033603708, "grad_norm": 5.793369293212891, "learning_rate": 9.620065371532024e-06, "loss": 0.5974, "step": 7700 }, { "epoch": 0.6381393808972025, "grad_norm": 6.288873672485352, "learning_rate": 9.618222050133052e-06, "loss": 0.5767, "step": 7710 }, { "epoch": 0.6389670584340341, "grad_norm": 6.283613681793213, "learning_rate": 9.6163744454194e-06, "loss": 0.5684, "step": 7720 }, { "epoch": 0.6397947359708658, "grad_norm": 3.393005847930908, "learning_rate": 9.614522559104679e-06, "loss": 0.5879, "step": 7730 }, { "epoch": 0.6406224135076974, "grad_norm": 4.732958793640137, "learning_rate": 9.612666392906489e-06, "loss": 0.6571, "step": 7740 }, { "epoch": 0.6414500910445291, "grad_norm": 4.492164134979248, "learning_rate": 9.610805948546385e-06, "loss": 0.7098, "step": 7750 }, { "epoch": 0.6422777685813607, "grad_norm": 6.13973331451416, "learning_rate": 9.608941227749895e-06, "loss": 0.621, "step": 7760 }, { "epoch": 0.6431054461181923, "grad_norm": 5.078057765960693, "learning_rate": 9.607072232246512e-06, "loss": 0.6681, "step": 7770 }, { "epoch": 0.643933123655024, "grad_norm": 5.39637565612793, "learning_rate": 9.605198963769693e-06, "loss": 0.733, "step": 7780 }, { "epoch": 0.6447608011918556, "grad_norm": 5.181097030639648, "learning_rate": 9.603321424056865e-06, "loss": 0.6083, "step": 7790 }, { "epoch": 0.6455884787286873, "grad_norm": 5.932875156402588, "learning_rate": 9.601439614849403e-06, "loss": 0.6886, "step": 7800 }, { "epoch": 0.646416156265519, "grad_norm": 6.571812152862549, "learning_rate": 9.599553537892659e-06, "loss": 0.5722, "step": 7810 }, { "epoch": 0.6472438338023506, "grad_norm": 5.604794979095459, "learning_rate": 9.597663194935927e-06, "loss": 0.6955, "step": 7820 }, { "epoch": 0.6480715113391823, "grad_norm": 4.326800346374512, "learning_rate": 9.595768587732465e-06, "loss": 0.6176, "step": 7830 }, { "epoch": 0.6488991888760139, "grad_norm": 4.263263702392578, "learning_rate": 9.59386971803949e-06, "loss": 0.6198, "step": 7840 }, { "epoch": 0.6497268664128456, "grad_norm": 7.278620719909668, "learning_rate": 9.591966587618166e-06, "loss": 0.5812, "step": 7850 }, { "epoch": 0.6505545439496772, "grad_norm": 4.418271064758301, "learning_rate": 9.590059198233612e-06, "loss": 0.7537, "step": 7860 }, { "epoch": 0.6513822214865088, "grad_norm": 3.83166241645813, "learning_rate": 9.588147551654895e-06, "loss": 0.6041, "step": 7870 }, { "epoch": 0.6522098990233405, "grad_norm": 5.515861511230469, "learning_rate": 9.58623164965503e-06, "loss": 0.7154, "step": 7880 }, { "epoch": 0.6530375765601721, "grad_norm": 4.48759651184082, "learning_rate": 9.584311494010985e-06, "loss": 0.6602, "step": 7890 }, { "epoch": 0.6538652540970038, "grad_norm": 6.287261486053467, "learning_rate": 9.582387086503661e-06, "loss": 0.5793, "step": 7900 }, { "epoch": 0.6546929316338355, "grad_norm": 5.651237487792969, "learning_rate": 9.580458428917917e-06, "loss": 0.6469, "step": 7910 }, { "epoch": 0.6555206091706671, "grad_norm": 6.236342430114746, "learning_rate": 9.578525523042543e-06, "loss": 0.6799, "step": 7920 }, { "epoch": 0.6563482867074988, "grad_norm": 6.384500503540039, "learning_rate": 9.576588370670276e-06, "loss": 0.739, "step": 7930 }, { "epoch": 0.6571759642443304, "grad_norm": 5.727816104888916, "learning_rate": 9.574646973597785e-06, "loss": 0.6171, "step": 7940 }, { "epoch": 0.6580036417811621, "grad_norm": 4.847169876098633, "learning_rate": 9.572701333625683e-06, "loss": 0.6659, "step": 7950 }, { "epoch": 0.6588313193179937, "grad_norm": 4.9109930992126465, "learning_rate": 9.570751452558514e-06, "loss": 0.6067, "step": 7960 }, { "epoch": 0.6596589968548253, "grad_norm": 6.125060081481934, "learning_rate": 9.568797332204754e-06, "loss": 0.6642, "step": 7970 }, { "epoch": 0.660486674391657, "grad_norm": 5.10365629196167, "learning_rate": 9.566838974376816e-06, "loss": 0.6419, "step": 7980 }, { "epoch": 0.6613143519284886, "grad_norm": 4.759987831115723, "learning_rate": 9.564876380891036e-06, "loss": 0.7096, "step": 7990 }, { "epoch": 0.6621420294653203, "grad_norm": 4.152451515197754, "learning_rate": 9.562909553567687e-06, "loss": 0.608, "step": 8000 }, { "epoch": 0.662969707002152, "grad_norm": 12.125520706176758, "learning_rate": 9.560938494230964e-06, "loss": 0.7111, "step": 8010 }, { "epoch": 0.6637973845389836, "grad_norm": 5.63936710357666, "learning_rate": 9.558963204708988e-06, "loss": 0.6712, "step": 8020 }, { "epoch": 0.6646250620758153, "grad_norm": 4.4002838134765625, "learning_rate": 9.556983686833801e-06, "loss": 0.6892, "step": 8030 }, { "epoch": 0.665452739612647, "grad_norm": 5.103034019470215, "learning_rate": 9.55499994244137e-06, "loss": 0.6793, "step": 8040 }, { "epoch": 0.6662804171494786, "grad_norm": 3.7829928398132324, "learning_rate": 9.55301197337158e-06, "loss": 0.5457, "step": 8050 }, { "epoch": 0.6671080946863102, "grad_norm": 4.988119602203369, "learning_rate": 9.551019781468237e-06, "loss": 0.6868, "step": 8060 }, { "epoch": 0.6679357722231418, "grad_norm": 8.042692184448242, "learning_rate": 9.549023368579063e-06, "loss": 0.6605, "step": 8070 }, { "epoch": 0.6687634497599735, "grad_norm": 5.496834754943848, "learning_rate": 9.547022736555689e-06, "loss": 0.6741, "step": 8080 }, { "epoch": 0.6695911272968051, "grad_norm": 5.695679187774658, "learning_rate": 9.545017887253667e-06, "loss": 0.5947, "step": 8090 }, { "epoch": 0.6704188048336368, "grad_norm": 4.6930389404296875, "learning_rate": 9.543008822532455e-06, "loss": 0.6514, "step": 8100 }, { "epoch": 0.6712464823704685, "grad_norm": 5.946260929107666, "learning_rate": 9.540995544255427e-06, "loss": 0.6511, "step": 8110 }, { "epoch": 0.6720741599073001, "grad_norm": 8.2454833984375, "learning_rate": 9.538978054289858e-06, "loss": 0.7313, "step": 8120 }, { "epoch": 0.6729018374441318, "grad_norm": 3.8761649131774902, "learning_rate": 9.536956354506932e-06, "loss": 0.6579, "step": 8130 }, { "epoch": 0.6737295149809635, "grad_norm": 4.517693996429443, "learning_rate": 9.534930446781739e-06, "loss": 0.6182, "step": 8140 }, { "epoch": 0.6745571925177951, "grad_norm": 4.291830062866211, "learning_rate": 9.532900332993271e-06, "loss": 0.5681, "step": 8150 }, { "epoch": 0.6753848700546267, "grad_norm": 5.030625820159912, "learning_rate": 9.530866015024419e-06, "loss": 0.5823, "step": 8160 }, { "epoch": 0.6762125475914583, "grad_norm": 5.484288215637207, "learning_rate": 9.528827494761977e-06, "loss": 0.5945, "step": 8170 }, { "epoch": 0.67704022512829, "grad_norm": 6.025727272033691, "learning_rate": 9.526784774096632e-06, "loss": 0.6124, "step": 8180 }, { "epoch": 0.6778679026651216, "grad_norm": 4.599993705749512, "learning_rate": 9.52473785492297e-06, "loss": 0.5715, "step": 8190 }, { "epoch": 0.6786955802019533, "grad_norm": 5.511344909667969, "learning_rate": 9.52268673913947e-06, "loss": 0.6586, "step": 8200 }, { "epoch": 0.679523257738785, "grad_norm": 5.013758659362793, "learning_rate": 9.520631428648507e-06, "loss": 0.688, "step": 8210 }, { "epoch": 0.6803509352756166, "grad_norm": 7.3988213539123535, "learning_rate": 9.518571925356339e-06, "loss": 0.5697, "step": 8220 }, { "epoch": 0.6811786128124483, "grad_norm": 5.941640377044678, "learning_rate": 9.516508231173118e-06, "loss": 0.6879, "step": 8230 }, { "epoch": 0.68200629034928, "grad_norm": 5.02189826965332, "learning_rate": 9.514440348012883e-06, "loss": 0.648, "step": 8240 }, { "epoch": 0.6828339678861116, "grad_norm": 6.1207451820373535, "learning_rate": 9.512368277793555e-06, "loss": 0.6029, "step": 8250 }, { "epoch": 0.6836616454229433, "grad_norm": 5.76517391204834, "learning_rate": 9.510292022436941e-06, "loss": 0.6881, "step": 8260 }, { "epoch": 0.6844893229597748, "grad_norm": 3.9579432010650635, "learning_rate": 9.508211583868731e-06, "loss": 0.5228, "step": 8270 }, { "epoch": 0.6853170004966065, "grad_norm": 5.904852390289307, "learning_rate": 9.506126964018492e-06, "loss": 0.5387, "step": 8280 }, { "epoch": 0.6861446780334381, "grad_norm": 5.313442230224609, "learning_rate": 9.504038164819669e-06, "loss": 0.5333, "step": 8290 }, { "epoch": 0.6869723555702698, "grad_norm": 4.384395122528076, "learning_rate": 9.501945188209582e-06, "loss": 0.5346, "step": 8300 }, { "epoch": 0.6878000331071015, "grad_norm": 5.140133857727051, "learning_rate": 9.499848036129433e-06, "loss": 0.5738, "step": 8310 }, { "epoch": 0.6886277106439331, "grad_norm": 4.416533946990967, "learning_rate": 9.497746710524289e-06, "loss": 0.5953, "step": 8320 }, { "epoch": 0.6894553881807648, "grad_norm": 4.817104339599609, "learning_rate": 9.49564121334309e-06, "loss": 0.5829, "step": 8330 }, { "epoch": 0.6902830657175965, "grad_norm": 9.510987281799316, "learning_rate": 9.493531546538644e-06, "loss": 0.6542, "step": 8340 }, { "epoch": 0.6911107432544281, "grad_norm": 5.337761402130127, "learning_rate": 9.491417712067629e-06, "loss": 0.7058, "step": 8350 }, { "epoch": 0.6919384207912598, "grad_norm": 5.41985559463501, "learning_rate": 9.489299711890587e-06, "loss": 0.6695, "step": 8360 }, { "epoch": 0.6927660983280913, "grad_norm": 5.715612411499023, "learning_rate": 9.487177547971924e-06, "loss": 0.6429, "step": 8370 }, { "epoch": 0.693593775864923, "grad_norm": 3.7750930786132812, "learning_rate": 9.485051222279907e-06, "loss": 0.5834, "step": 8380 }, { "epoch": 0.6944214534017547, "grad_norm": 4.803799152374268, "learning_rate": 9.48292073678666e-06, "loss": 0.6116, "step": 8390 }, { "epoch": 0.6952491309385863, "grad_norm": 5.199375152587891, "learning_rate": 9.480786093468176e-06, "loss": 0.5804, "step": 8400 }, { "epoch": 0.696076808475418, "grad_norm": 7.1734747886657715, "learning_rate": 9.47864729430429e-06, "loss": 0.5675, "step": 8410 }, { "epoch": 0.6969044860122496, "grad_norm": 3.0254194736480713, "learning_rate": 9.476504341278703e-06, "loss": 0.6485, "step": 8420 }, { "epoch": 0.6977321635490813, "grad_norm": 4.259583473205566, "learning_rate": 9.47435723637896e-06, "loss": 0.6287, "step": 8430 }, { "epoch": 0.698559841085913, "grad_norm": 5.717179775238037, "learning_rate": 9.472205981596463e-06, "loss": 0.5781, "step": 8440 }, { "epoch": 0.6993875186227446, "grad_norm": 6.725274085998535, "learning_rate": 9.470050578926458e-06, "loss": 0.6096, "step": 8450 }, { "epoch": 0.7002151961595763, "grad_norm": 5.758815288543701, "learning_rate": 9.467891030368045e-06, "loss": 0.6557, "step": 8460 }, { "epoch": 0.7010428736964078, "grad_norm": 5.0641961097717285, "learning_rate": 9.465727337924162e-06, "loss": 0.5404, "step": 8470 }, { "epoch": 0.7018705512332395, "grad_norm": 4.139156818389893, "learning_rate": 9.463559503601594e-06, "loss": 0.6401, "step": 8480 }, { "epoch": 0.7026982287700712, "grad_norm": 7.095492839813232, "learning_rate": 9.461387529410966e-06, "loss": 0.5725, "step": 8490 }, { "epoch": 0.7035259063069028, "grad_norm": 4.679490566253662, "learning_rate": 9.459211417366744e-06, "loss": 0.6021, "step": 8500 }, { "epoch": 0.7043535838437345, "grad_norm": 5.608545780181885, "learning_rate": 9.457031169487233e-06, "loss": 0.5799, "step": 8510 }, { "epoch": 0.7051812613805661, "grad_norm": 6.848021507263184, "learning_rate": 9.45484678779457e-06, "loss": 0.6685, "step": 8520 }, { "epoch": 0.7060089389173978, "grad_norm": 7.54874849319458, "learning_rate": 9.45265827431473e-06, "loss": 0.6109, "step": 8530 }, { "epoch": 0.7068366164542295, "grad_norm": 5.265584468841553, "learning_rate": 9.450465631077516e-06, "loss": 0.63, "step": 8540 }, { "epoch": 0.7076642939910611, "grad_norm": 3.3235714435577393, "learning_rate": 9.448268860116565e-06, "loss": 0.6171, "step": 8550 }, { "epoch": 0.7084919715278928, "grad_norm": 6.858134746551514, "learning_rate": 9.446067963469341e-06, "loss": 0.6671, "step": 8560 }, { "epoch": 0.7093196490647243, "grad_norm": 5.54447603225708, "learning_rate": 9.443862943177135e-06, "loss": 0.6714, "step": 8570 }, { "epoch": 0.710147326601556, "grad_norm": 4.113467693328857, "learning_rate": 9.441653801285062e-06, "loss": 0.5686, "step": 8580 }, { "epoch": 0.7109750041383877, "grad_norm": 7.966207027435303, "learning_rate": 9.439440539842059e-06, "loss": 0.6178, "step": 8590 }, { "epoch": 0.7118026816752193, "grad_norm": 5.764838695526123, "learning_rate": 9.437223160900884e-06, "loss": 0.6599, "step": 8600 }, { "epoch": 0.712630359212051, "grad_norm": 7.645096778869629, "learning_rate": 9.435001666518118e-06, "loss": 0.6448, "step": 8610 }, { "epoch": 0.7134580367488826, "grad_norm": 4.181859970092773, "learning_rate": 9.432776058754152e-06, "loss": 0.6518, "step": 8620 }, { "epoch": 0.7142857142857143, "grad_norm": 5.847019672393799, "learning_rate": 9.430546339673197e-06, "loss": 0.5936, "step": 8630 }, { "epoch": 0.715113391822546, "grad_norm": 4.446036338806152, "learning_rate": 9.428312511343274e-06, "loss": 0.7234, "step": 8640 }, { "epoch": 0.7159410693593776, "grad_norm": 5.708004951477051, "learning_rate": 9.42607457583622e-06, "loss": 0.6757, "step": 8650 }, { "epoch": 0.7167687468962093, "grad_norm": 5.345437049865723, "learning_rate": 9.423832535227677e-06, "loss": 0.6677, "step": 8660 }, { "epoch": 0.7175964244330408, "grad_norm": 5.480767726898193, "learning_rate": 9.421586391597097e-06, "loss": 0.6298, "step": 8670 }, { "epoch": 0.7184241019698725, "grad_norm": 4.26676607131958, "learning_rate": 9.419336147027733e-06, "loss": 0.6225, "step": 8680 }, { "epoch": 0.7192517795067042, "grad_norm": 5.671989440917969, "learning_rate": 9.417081803606649e-06, "loss": 0.6458, "step": 8690 }, { "epoch": 0.7200794570435358, "grad_norm": 6.458143711090088, "learning_rate": 9.414823363424703e-06, "loss": 0.6435, "step": 8700 }, { "epoch": 0.7209071345803675, "grad_norm": 5.306424617767334, "learning_rate": 9.412560828576558e-06, "loss": 0.5873, "step": 8710 }, { "epoch": 0.7217348121171991, "grad_norm": 6.95552396774292, "learning_rate": 9.410294201160673e-06, "loss": 0.6085, "step": 8720 }, { "epoch": 0.7225624896540308, "grad_norm": 5.4268035888671875, "learning_rate": 9.4080234832793e-06, "loss": 0.6458, "step": 8730 }, { "epoch": 0.7233901671908625, "grad_norm": 2.6680908203125, "learning_rate": 9.40574867703849e-06, "loss": 0.6596, "step": 8740 }, { "epoch": 0.7242178447276941, "grad_norm": 6.534355640411377, "learning_rate": 9.403469784548083e-06, "loss": 0.6566, "step": 8750 }, { "epoch": 0.7250455222645258, "grad_norm": 5.280076026916504, "learning_rate": 9.401186807921707e-06, "loss": 0.5925, "step": 8760 }, { "epoch": 0.7258731998013573, "grad_norm": 9.512300491333008, "learning_rate": 9.398899749276785e-06, "loss": 0.6354, "step": 8770 }, { "epoch": 0.726700877338189, "grad_norm": 4.644611358642578, "learning_rate": 9.396608610734515e-06, "loss": 0.6676, "step": 8780 }, { "epoch": 0.7275285548750207, "grad_norm": 5.151789665222168, "learning_rate": 9.394313394419889e-06, "loss": 0.5816, "step": 8790 }, { "epoch": 0.7283562324118523, "grad_norm": 4.709674835205078, "learning_rate": 9.392014102461677e-06, "loss": 0.5786, "step": 8800 }, { "epoch": 0.729183909948684, "grad_norm": 4.660391330718994, "learning_rate": 9.389710736992427e-06, "loss": 0.5563, "step": 8810 }, { "epoch": 0.7300115874855156, "grad_norm": 6.696812152862549, "learning_rate": 9.387403300148472e-06, "loss": 0.7307, "step": 8820 }, { "epoch": 0.7308392650223473, "grad_norm": 4.249696254730225, "learning_rate": 9.385091794069913e-06, "loss": 0.6237, "step": 8830 }, { "epoch": 0.731666942559179, "grad_norm": 3.452157735824585, "learning_rate": 9.382776220900631e-06, "loss": 0.5942, "step": 8840 }, { "epoch": 0.7324946200960106, "grad_norm": 5.5069708824157715, "learning_rate": 9.380456582788274e-06, "loss": 0.6241, "step": 8850 }, { "epoch": 0.7333222976328423, "grad_norm": 4.244978904724121, "learning_rate": 9.378132881884264e-06, "loss": 0.6167, "step": 8860 }, { "epoch": 0.7341499751696738, "grad_norm": 5.150333404541016, "learning_rate": 9.375805120343791e-06, "loss": 0.6691, "step": 8870 }, { "epoch": 0.7349776527065055, "grad_norm": 4.343332290649414, "learning_rate": 9.37347330032581e-06, "loss": 0.641, "step": 8880 }, { "epoch": 0.7358053302433372, "grad_norm": 4.822870254516602, "learning_rate": 9.37113742399304e-06, "loss": 0.5879, "step": 8890 }, { "epoch": 0.7366330077801688, "grad_norm": 4.078579902648926, "learning_rate": 9.368797493511964e-06, "loss": 0.4908, "step": 8900 }, { "epoch": 0.7374606853170005, "grad_norm": 4.472438812255859, "learning_rate": 9.366453511052823e-06, "loss": 0.663, "step": 8910 }, { "epoch": 0.7382883628538321, "grad_norm": 3.1786911487579346, "learning_rate": 9.364105478789616e-06, "loss": 0.611, "step": 8920 }, { "epoch": 0.7391160403906638, "grad_norm": 3.592332601547241, "learning_rate": 9.361753398900103e-06, "loss": 0.6725, "step": 8930 }, { "epoch": 0.7399437179274955, "grad_norm": 4.288097858428955, "learning_rate": 9.359397273565786e-06, "loss": 0.6541, "step": 8940 }, { "epoch": 0.7407713954643271, "grad_norm": 4.274669170379639, "learning_rate": 9.357037104971936e-06, "loss": 0.6829, "step": 8950 }, { "epoch": 0.7415990730011588, "grad_norm": 5.2250237464904785, "learning_rate": 9.354672895307561e-06, "loss": 0.574, "step": 8960 }, { "epoch": 0.7424267505379905, "grad_norm": 5.827564239501953, "learning_rate": 9.352304646765422e-06, "loss": 0.5615, "step": 8970 }, { "epoch": 0.743254428074822, "grad_norm": 6.202907562255859, "learning_rate": 9.349932361542025e-06, "loss": 0.6497, "step": 8980 }, { "epoch": 0.7440821056116537, "grad_norm": 5.4073405265808105, "learning_rate": 9.34755604183762e-06, "loss": 0.6546, "step": 8990 }, { "epoch": 0.7449097831484853, "grad_norm": 5.011789321899414, "learning_rate": 9.3451756898562e-06, "loss": 0.6783, "step": 9000 }, { "epoch": 0.745737460685317, "grad_norm": 5.958730697631836, "learning_rate": 9.342791307805497e-06, "loss": 0.607, "step": 9010 }, { "epoch": 0.7465651382221486, "grad_norm": 6.131625175476074, "learning_rate": 9.34040289789698e-06, "loss": 0.6119, "step": 9020 }, { "epoch": 0.7473928157589803, "grad_norm": 6.155272006988525, "learning_rate": 9.338010462345855e-06, "loss": 0.6277, "step": 9030 }, { "epoch": 0.748220493295812, "grad_norm": 3.6516366004943848, "learning_rate": 9.335614003371062e-06, "loss": 0.5645, "step": 9040 }, { "epoch": 0.7490481708326436, "grad_norm": 5.593658924102783, "learning_rate": 9.33321352319527e-06, "loss": 0.5805, "step": 9050 }, { "epoch": 0.7498758483694753, "grad_norm": 5.568034648895264, "learning_rate": 9.33080902404488e-06, "loss": 0.5567, "step": 9060 }, { "epoch": 0.750703525906307, "grad_norm": 5.087985992431641, "learning_rate": 9.328400508150022e-06, "loss": 0.5613, "step": 9070 }, { "epoch": 0.7515312034431385, "grad_norm": 4.900256156921387, "learning_rate": 9.325987977744549e-06, "loss": 0.6466, "step": 9080 }, { "epoch": 0.7523588809799702, "grad_norm": 4.500089168548584, "learning_rate": 9.323571435066036e-06, "loss": 0.6269, "step": 9090 }, { "epoch": 0.7531865585168018, "grad_norm": 5.740048408508301, "learning_rate": 9.321150882355783e-06, "loss": 0.5536, "step": 9100 }, { "epoch": 0.7540142360536335, "grad_norm": 5.465550899505615, "learning_rate": 9.318726321858807e-06, "loss": 0.6155, "step": 9110 }, { "epoch": 0.7548419135904652, "grad_norm": 9.747673034667969, "learning_rate": 9.31629775582384e-06, "loss": 0.692, "step": 9120 }, { "epoch": 0.7556695911272968, "grad_norm": 5.115851879119873, "learning_rate": 9.313865186503336e-06, "loss": 0.6364, "step": 9130 }, { "epoch": 0.7564972686641285, "grad_norm": 4.4185309410095215, "learning_rate": 9.311428616153457e-06, "loss": 0.6429, "step": 9140 }, { "epoch": 0.7573249462009601, "grad_norm": 5.2742438316345215, "learning_rate": 9.308988047034073e-06, "loss": 0.6625, "step": 9150 }, { "epoch": 0.7581526237377918, "grad_norm": 4.925389289855957, "learning_rate": 9.306543481408767e-06, "loss": 0.5732, "step": 9160 }, { "epoch": 0.7589803012746235, "grad_norm": 4.831460475921631, "learning_rate": 9.304094921544833e-06, "loss": 0.6239, "step": 9170 }, { "epoch": 0.759807978811455, "grad_norm": 4.988065242767334, "learning_rate": 9.301642369713257e-06, "loss": 0.6815, "step": 9180 }, { "epoch": 0.7606356563482867, "grad_norm": 5.020726680755615, "learning_rate": 9.299185828188742e-06, "loss": 0.5904, "step": 9190 }, { "epoch": 0.7614633338851183, "grad_norm": 4.344644069671631, "learning_rate": 9.29672529924968e-06, "loss": 0.6253, "step": 9200 }, { "epoch": 0.76229101142195, "grad_norm": 6.128726482391357, "learning_rate": 9.294260785178167e-06, "loss": 0.5648, "step": 9210 }, { "epoch": 0.7631186889587817, "grad_norm": 4.861498832702637, "learning_rate": 9.291792288259994e-06, "loss": 0.6813, "step": 9220 }, { "epoch": 0.7639463664956133, "grad_norm": 4.046803951263428, "learning_rate": 9.289319810784643e-06, "loss": 0.6436, "step": 9230 }, { "epoch": 0.764774044032445, "grad_norm": 7.723761558532715, "learning_rate": 9.286843355045295e-06, "loss": 0.6455, "step": 9240 }, { "epoch": 0.7656017215692766, "grad_norm": 5.379072189331055, "learning_rate": 9.284362923338815e-06, "loss": 0.654, "step": 9250 }, { "epoch": 0.7664293991061083, "grad_norm": 5.126608371734619, "learning_rate": 9.281878517965755e-06, "loss": 0.6976, "step": 9260 }, { "epoch": 0.76725707664294, "grad_norm": 7.456500053405762, "learning_rate": 9.279390141230356e-06, "loss": 0.6786, "step": 9270 }, { "epoch": 0.7680847541797715, "grad_norm": 5.284501552581787, "learning_rate": 9.276897795440542e-06, "loss": 0.7244, "step": 9280 }, { "epoch": 0.7689124317166032, "grad_norm": 4.664309501647949, "learning_rate": 9.274401482907915e-06, "loss": 0.6654, "step": 9290 }, { "epoch": 0.7697401092534348, "grad_norm": 4.8233771324157715, "learning_rate": 9.271901205947761e-06, "loss": 0.5974, "step": 9300 }, { "epoch": 0.7705677867902665, "grad_norm": 5.095113277435303, "learning_rate": 9.269396966879038e-06, "loss": 0.6778, "step": 9310 }, { "epoch": 0.7713954643270982, "grad_norm": 5.031292915344238, "learning_rate": 9.266888768024382e-06, "loss": 0.6038, "step": 9320 }, { "epoch": 0.7722231418639298, "grad_norm": 5.960371971130371, "learning_rate": 9.2643766117101e-06, "loss": 0.6797, "step": 9330 }, { "epoch": 0.7730508194007615, "grad_norm": 4.599192142486572, "learning_rate": 9.26186050026617e-06, "loss": 0.5807, "step": 9340 }, { "epoch": 0.7738784969375931, "grad_norm": 5.6515960693359375, "learning_rate": 9.259340436026238e-06, "loss": 0.7014, "step": 9350 }, { "epoch": 0.7747061744744248, "grad_norm": 4.409613132476807, "learning_rate": 9.25681642132762e-06, "loss": 0.6926, "step": 9360 }, { "epoch": 0.7755338520112565, "grad_norm": 3.916626214981079, "learning_rate": 9.254288458511287e-06, "loss": 0.6184, "step": 9370 }, { "epoch": 0.776361529548088, "grad_norm": 4.2588372230529785, "learning_rate": 9.25175654992188e-06, "loss": 0.6647, "step": 9380 }, { "epoch": 0.7771892070849197, "grad_norm": 5.050411224365234, "learning_rate": 9.249220697907697e-06, "loss": 0.599, "step": 9390 }, { "epoch": 0.7780168846217513, "grad_norm": 4.382516860961914, "learning_rate": 9.246680904820694e-06, "loss": 0.6213, "step": 9400 }, { "epoch": 0.778844562158583, "grad_norm": 4.899589538574219, "learning_rate": 9.24413717301648e-06, "loss": 0.5463, "step": 9410 }, { "epoch": 0.7796722396954147, "grad_norm": 7.138366222381592, "learning_rate": 9.24158950485432e-06, "loss": 0.5944, "step": 9420 }, { "epoch": 0.7804999172322463, "grad_norm": 4.455055236816406, "learning_rate": 9.239037902697127e-06, "loss": 0.6917, "step": 9430 }, { "epoch": 0.781327594769078, "grad_norm": 4.407454967498779, "learning_rate": 9.236482368911466e-06, "loss": 0.5931, "step": 9440 }, { "epoch": 0.7821552723059096, "grad_norm": 5.514267444610596, "learning_rate": 9.233922905867547e-06, "loss": 0.5691, "step": 9450 }, { "epoch": 0.7829829498427413, "grad_norm": 6.1955342292785645, "learning_rate": 9.231359515939222e-06, "loss": 0.6173, "step": 9460 }, { "epoch": 0.783810627379573, "grad_norm": 7.648247241973877, "learning_rate": 9.228792201503992e-06, "loss": 0.6272, "step": 9470 }, { "epoch": 0.7846383049164045, "grad_norm": 4.228515148162842, "learning_rate": 9.226220964942987e-06, "loss": 0.6451, "step": 9480 }, { "epoch": 0.7854659824532362, "grad_norm": 5.015971660614014, "learning_rate": 9.223645808640985e-06, "loss": 0.6366, "step": 9490 }, { "epoch": 0.7862936599900678, "grad_norm": 5.373770713806152, "learning_rate": 9.221066734986393e-06, "loss": 0.7007, "step": 9500 }, { "epoch": 0.7871213375268995, "grad_norm": 4.676547527313232, "learning_rate": 9.218483746371255e-06, "loss": 0.6673, "step": 9510 }, { "epoch": 0.7879490150637312, "grad_norm": 6.42137336730957, "learning_rate": 9.215896845191249e-06, "loss": 0.6468, "step": 9520 }, { "epoch": 0.7887766926005628, "grad_norm": 5.271933555603027, "learning_rate": 9.213306033845668e-06, "loss": 0.6518, "step": 9530 }, { "epoch": 0.7896043701373945, "grad_norm": 5.100630283355713, "learning_rate": 9.210711314737448e-06, "loss": 0.6521, "step": 9540 }, { "epoch": 0.7904320476742261, "grad_norm": 4.400863170623779, "learning_rate": 9.20811269027314e-06, "loss": 0.5368, "step": 9550 }, { "epoch": 0.7912597252110578, "grad_norm": 6.290362358093262, "learning_rate": 9.20551016286292e-06, "loss": 0.5971, "step": 9560 }, { "epoch": 0.7920874027478895, "grad_norm": 4.676914215087891, "learning_rate": 9.202903734920586e-06, "loss": 0.6131, "step": 9570 }, { "epoch": 0.792915080284721, "grad_norm": 5.518727779388428, "learning_rate": 9.200293408863547e-06, "loss": 0.5674, "step": 9580 }, { "epoch": 0.7937427578215527, "grad_norm": 11.917981147766113, "learning_rate": 9.197679187112833e-06, "loss": 0.6085, "step": 9590 }, { "epoch": 0.7945704353583843, "grad_norm": 4.717650413513184, "learning_rate": 9.19506107209309e-06, "loss": 0.6128, "step": 9600 }, { "epoch": 0.795398112895216, "grad_norm": 4.6282501220703125, "learning_rate": 9.192439066232564e-06, "loss": 0.6018, "step": 9610 }, { "epoch": 0.7962257904320477, "grad_norm": 5.656414985656738, "learning_rate": 9.189813171963121e-06, "loss": 0.6194, "step": 9620 }, { "epoch": 0.7970534679688793, "grad_norm": 4.991113185882568, "learning_rate": 9.18718339172023e-06, "loss": 0.6049, "step": 9630 }, { "epoch": 0.797881145505711, "grad_norm": 7.138369083404541, "learning_rate": 9.184549727942958e-06, "loss": 0.6646, "step": 9640 }, { "epoch": 0.7987088230425426, "grad_norm": 4.843081951141357, "learning_rate": 9.181912183073982e-06, "loss": 0.5978, "step": 9650 }, { "epoch": 0.7995365005793743, "grad_norm": 3.986147403717041, "learning_rate": 9.179270759559577e-06, "loss": 0.661, "step": 9660 }, { "epoch": 0.800364178116206, "grad_norm": 7.07649564743042, "learning_rate": 9.176625459849611e-06, "loss": 0.571, "step": 9670 }, { "epoch": 0.8011918556530375, "grad_norm": 6.846652030944824, "learning_rate": 9.173976286397552e-06, "loss": 0.6219, "step": 9680 }, { "epoch": 0.8020195331898692, "grad_norm": 5.500576496124268, "learning_rate": 9.171323241660458e-06, "loss": 0.6351, "step": 9690 }, { "epoch": 0.8028472107267008, "grad_norm": 6.127823829650879, "learning_rate": 9.168666328098979e-06, "loss": 0.6788, "step": 9700 }, { "epoch": 0.8036748882635325, "grad_norm": 5.048381805419922, "learning_rate": 9.166005548177352e-06, "loss": 0.6618, "step": 9710 }, { "epoch": 0.8045025658003642, "grad_norm": 6.012566089630127, "learning_rate": 9.1633409043634e-06, "loss": 0.5976, "step": 9720 }, { "epoch": 0.8053302433371958, "grad_norm": 4.90334415435791, "learning_rate": 9.160672399128534e-06, "loss": 0.6334, "step": 9730 }, { "epoch": 0.8061579208740275, "grad_norm": 6.516371726989746, "learning_rate": 9.158000034947739e-06, "loss": 0.6058, "step": 9740 }, { "epoch": 0.8069855984108592, "grad_norm": 4.045821189880371, "learning_rate": 9.15532381429958e-06, "loss": 0.5993, "step": 9750 }, { "epoch": 0.8078132759476908, "grad_norm": 5.548466205596924, "learning_rate": 9.152643739666212e-06, "loss": 0.5505, "step": 9760 }, { "epoch": 0.8086409534845225, "grad_norm": 4.745909690856934, "learning_rate": 9.149959813533345e-06, "loss": 0.6555, "step": 9770 }, { "epoch": 0.8094686310213541, "grad_norm": 5.833310127258301, "learning_rate": 9.147272038390274e-06, "loss": 0.6071, "step": 9780 }, { "epoch": 0.8102963085581857, "grad_norm": 4.267097473144531, "learning_rate": 9.144580416729858e-06, "loss": 0.7202, "step": 9790 }, { "epoch": 0.8111239860950173, "grad_norm": 5.388729572296143, "learning_rate": 9.14188495104853e-06, "loss": 0.5776, "step": 9800 }, { "epoch": 0.811951663631849, "grad_norm": 5.602564334869385, "learning_rate": 9.139185643846278e-06, "loss": 0.6917, "step": 9810 }, { "epoch": 0.8127793411686807, "grad_norm": 4.925511360168457, "learning_rate": 9.136482497626665e-06, "loss": 0.6003, "step": 9820 }, { "epoch": 0.8136070187055123, "grad_norm": 3.7184879779815674, "learning_rate": 9.133775514896803e-06, "loss": 0.5525, "step": 9830 }, { "epoch": 0.814434696242344, "grad_norm": 6.180205821990967, "learning_rate": 9.131064698167374e-06, "loss": 0.6581, "step": 9840 }, { "epoch": 0.8152623737791757, "grad_norm": 5.54541015625, "learning_rate": 9.128350049952606e-06, "loss": 0.6928, "step": 9850 }, { "epoch": 0.8160900513160073, "grad_norm": 6.336387634277344, "learning_rate": 9.125631572770283e-06, "loss": 0.573, "step": 9860 }, { "epoch": 0.816917728852839, "grad_norm": 3.332813262939453, "learning_rate": 9.122909269141746e-06, "loss": 0.6178, "step": 9870 }, { "epoch": 0.8177454063896706, "grad_norm": 5.389678955078125, "learning_rate": 9.120183141591879e-06, "loss": 0.6106, "step": 9880 }, { "epoch": 0.8185730839265022, "grad_norm": 4.314904689788818, "learning_rate": 9.117453192649113e-06, "loss": 0.5947, "step": 9890 }, { "epoch": 0.8194007614633338, "grad_norm": 4.276676654815674, "learning_rate": 9.114719424845426e-06, "loss": 0.6367, "step": 9900 }, { "epoch": 0.8202284390001655, "grad_norm": 4.69083833694458, "learning_rate": 9.111981840716338e-06, "loss": 0.555, "step": 9910 }, { "epoch": 0.8210561165369972, "grad_norm": 5.330567359924316, "learning_rate": 9.109240442800903e-06, "loss": 0.5493, "step": 9920 }, { "epoch": 0.8218837940738288, "grad_norm": 4.562192440032959, "learning_rate": 9.106495233641722e-06, "loss": 0.5847, "step": 9930 }, { "epoch": 0.8227114716106605, "grad_norm": 4.289856910705566, "learning_rate": 9.103746215784921e-06, "loss": 0.6935, "step": 9940 }, { "epoch": 0.8235391491474922, "grad_norm": 3.167022228240967, "learning_rate": 9.100993391780162e-06, "loss": 0.5681, "step": 9950 }, { "epoch": 0.8243668266843238, "grad_norm": 6.391303062438965, "learning_rate": 9.098236764180642e-06, "loss": 0.6031, "step": 9960 }, { "epoch": 0.8251945042211555, "grad_norm": 4.232006072998047, "learning_rate": 9.095476335543078e-06, "loss": 0.5907, "step": 9970 }, { "epoch": 0.8260221817579871, "grad_norm": 9.437912940979004, "learning_rate": 9.092712108427718e-06, "loss": 0.5578, "step": 9980 }, { "epoch": 0.8268498592948187, "grad_norm": 4.943689346313477, "learning_rate": 9.08994408539833e-06, "loss": 0.6437, "step": 9990 }, { "epoch": 0.8276775368316504, "grad_norm": 5.271090507507324, "learning_rate": 9.087172269022204e-06, "loss": 0.5302, "step": 10000 }, { "epoch": 0.828505214368482, "grad_norm": 4.090700149536133, "learning_rate": 9.084396661870147e-06, "loss": 0.534, "step": 10010 }, { "epoch": 0.8293328919053137, "grad_norm": 3.9861342906951904, "learning_rate": 9.081617266516485e-06, "loss": 0.577, "step": 10020 }, { "epoch": 0.8301605694421453, "grad_norm": 4.639739036560059, "learning_rate": 9.078834085539054e-06, "loss": 0.5682, "step": 10030 }, { "epoch": 0.830988246978977, "grad_norm": 4.792266845703125, "learning_rate": 9.076047121519202e-06, "loss": 0.605, "step": 10040 }, { "epoch": 0.8318159245158087, "grad_norm": 4.811140060424805, "learning_rate": 9.073256377041785e-06, "loss": 0.4878, "step": 10050 }, { "epoch": 0.8326436020526403, "grad_norm": 4.930051803588867, "learning_rate": 9.070461854695169e-06, "loss": 0.7001, "step": 10060 }, { "epoch": 0.833471279589472, "grad_norm": 5.721372127532959, "learning_rate": 9.06766355707122e-06, "loss": 0.7271, "step": 10070 }, { "epoch": 0.8342989571263036, "grad_norm": 5.659189224243164, "learning_rate": 9.064861486765307e-06, "loss": 0.6569, "step": 10080 }, { "epoch": 0.8351266346631352, "grad_norm": 5.9102396965026855, "learning_rate": 9.062055646376295e-06, "loss": 0.6668, "step": 10090 }, { "epoch": 0.8359543121999669, "grad_norm": 4.943755626678467, "learning_rate": 9.059246038506555e-06, "loss": 0.7298, "step": 10100 }, { "epoch": 0.8367819897367985, "grad_norm": 7.072639465332031, "learning_rate": 9.056432665761938e-06, "loss": 0.6257, "step": 10110 }, { "epoch": 0.8376096672736302, "grad_norm": 4.805159568786621, "learning_rate": 9.053615530751802e-06, "loss": 0.5656, "step": 10120 }, { "epoch": 0.8384373448104618, "grad_norm": 3.86311674118042, "learning_rate": 9.050794636088982e-06, "loss": 0.6979, "step": 10130 }, { "epoch": 0.8392650223472935, "grad_norm": 2.8739194869995117, "learning_rate": 9.047969984389808e-06, "loss": 0.6412, "step": 10140 }, { "epoch": 0.8400926998841252, "grad_norm": 4.404178142547607, "learning_rate": 9.04514157827409e-06, "loss": 0.6418, "step": 10150 }, { "epoch": 0.8409203774209568, "grad_norm": 3.914870262145996, "learning_rate": 9.042309420365123e-06, "loss": 0.5428, "step": 10160 }, { "epoch": 0.8417480549577885, "grad_norm": 8.01590633392334, "learning_rate": 9.039473513289679e-06, "loss": 0.6205, "step": 10170 }, { "epoch": 0.8425757324946201, "grad_norm": 6.804129600524902, "learning_rate": 9.03663385967801e-06, "loss": 0.5774, "step": 10180 }, { "epoch": 0.8434034100314517, "grad_norm": 4.953549385070801, "learning_rate": 9.033790462163842e-06, "loss": 0.6779, "step": 10190 }, { "epoch": 0.8442310875682834, "grad_norm": 6.291667461395264, "learning_rate": 9.030943323384372e-06, "loss": 0.6666, "step": 10200 }, { "epoch": 0.845058765105115, "grad_norm": 6.553426742553711, "learning_rate": 9.028092445980267e-06, "loss": 0.6595, "step": 10210 }, { "epoch": 0.8458864426419467, "grad_norm": 4.023377895355225, "learning_rate": 9.025237832595666e-06, "loss": 0.5099, "step": 10220 }, { "epoch": 0.8467141201787783, "grad_norm": 4.922636032104492, "learning_rate": 9.022379485878168e-06, "loss": 0.6651, "step": 10230 }, { "epoch": 0.84754179771561, "grad_norm": 5.434474945068359, "learning_rate": 9.019517408478836e-06, "loss": 0.6702, "step": 10240 }, { "epoch": 0.8483694752524417, "grad_norm": 4.186858177185059, "learning_rate": 9.016651603052192e-06, "loss": 0.5891, "step": 10250 }, { "epoch": 0.8491971527892733, "grad_norm": 3.8597772121429443, "learning_rate": 9.01378207225622e-06, "loss": 0.5657, "step": 10260 }, { "epoch": 0.850024830326105, "grad_norm": 4.298988342285156, "learning_rate": 9.010908818752352e-06, "loss": 0.657, "step": 10270 }, { "epoch": 0.8508525078629366, "grad_norm": 4.433570861816406, "learning_rate": 9.008031845205481e-06, "loss": 0.646, "step": 10280 }, { "epoch": 0.8516801853997682, "grad_norm": 3.6598236560821533, "learning_rate": 9.005151154283943e-06, "loss": 0.5069, "step": 10290 }, { "epoch": 0.8525078629365999, "grad_norm": 2.6936726570129395, "learning_rate": 9.002266748659527e-06, "loss": 0.5216, "step": 10300 }, { "epoch": 0.8533355404734315, "grad_norm": 7.747993469238281, "learning_rate": 8.999378631007467e-06, "loss": 0.6357, "step": 10310 }, { "epoch": 0.8541632180102632, "grad_norm": 4.126842498779297, "learning_rate": 8.996486804006434e-06, "loss": 0.5856, "step": 10320 }, { "epoch": 0.8549908955470948, "grad_norm": 4.915649890899658, "learning_rate": 8.993591270338546e-06, "loss": 0.6092, "step": 10330 }, { "epoch": 0.8558185730839265, "grad_norm": 4.490804195404053, "learning_rate": 8.990692032689356e-06, "loss": 0.6646, "step": 10340 }, { "epoch": 0.8566462506207582, "grad_norm": 4.121708869934082, "learning_rate": 8.987789093747854e-06, "loss": 0.6342, "step": 10350 }, { "epoch": 0.8574739281575898, "grad_norm": 5.0540852546691895, "learning_rate": 8.98488245620646e-06, "loss": 0.6236, "step": 10360 }, { "epoch": 0.8583016056944215, "grad_norm": 5.748679161071777, "learning_rate": 8.981972122761025e-06, "loss": 0.6516, "step": 10370 }, { "epoch": 0.8591292832312531, "grad_norm": 5.278295040130615, "learning_rate": 8.979058096110831e-06, "loss": 0.5607, "step": 10380 }, { "epoch": 0.8599569607680847, "grad_norm": 4.701098918914795, "learning_rate": 8.976140378958585e-06, "loss": 0.6325, "step": 10390 }, { "epoch": 0.8607846383049164, "grad_norm": 4.71337366104126, "learning_rate": 8.97321897401041e-06, "loss": 0.6438, "step": 10400 }, { "epoch": 0.861612315841748, "grad_norm": 3.77280855178833, "learning_rate": 8.970293883975858e-06, "loss": 0.5476, "step": 10410 }, { "epoch": 0.8624399933785797, "grad_norm": 3.865321636199951, "learning_rate": 8.967365111567894e-06, "loss": 0.625, "step": 10420 }, { "epoch": 0.8632676709154113, "grad_norm": 3.8015401363372803, "learning_rate": 8.964432659502901e-06, "loss": 0.6274, "step": 10430 }, { "epoch": 0.864095348452243, "grad_norm": 3.4587810039520264, "learning_rate": 8.961496530500675e-06, "loss": 0.6058, "step": 10440 }, { "epoch": 0.8649230259890747, "grad_norm": 5.080514907836914, "learning_rate": 8.958556727284415e-06, "loss": 0.6853, "step": 10450 }, { "epoch": 0.8657507035259063, "grad_norm": 5.457615852355957, "learning_rate": 8.955613252580739e-06, "loss": 0.6194, "step": 10460 }, { "epoch": 0.866578381062738, "grad_norm": 4.019306659698486, "learning_rate": 8.952666109119663e-06, "loss": 0.6593, "step": 10470 }, { "epoch": 0.8674060585995697, "grad_norm": 5.211251735687256, "learning_rate": 8.949715299634608e-06, "loss": 0.5833, "step": 10480 }, { "epoch": 0.8682337361364013, "grad_norm": 6.740599632263184, "learning_rate": 8.946760826862393e-06, "loss": 0.5474, "step": 10490 }, { "epoch": 0.8690614136732329, "grad_norm": 8.00527286529541, "learning_rate": 8.943802693543238e-06, "loss": 0.6199, "step": 10500 }, { "epoch": 0.8698890912100645, "grad_norm": 7.206969261169434, "learning_rate": 8.940840902420757e-06, "loss": 0.6813, "step": 10510 }, { "epoch": 0.8707167687468962, "grad_norm": 5.747097015380859, "learning_rate": 8.937875456241954e-06, "loss": 0.6251, "step": 10520 }, { "epoch": 0.8715444462837278, "grad_norm": 5.1278910636901855, "learning_rate": 8.934906357757226e-06, "loss": 0.4549, "step": 10530 }, { "epoch": 0.8723721238205595, "grad_norm": 4.264970779418945, "learning_rate": 8.931933609720355e-06, "loss": 0.6464, "step": 10540 }, { "epoch": 0.8731998013573912, "grad_norm": 4.263440132141113, "learning_rate": 8.928957214888513e-06, "loss": 0.5738, "step": 10550 }, { "epoch": 0.8740274788942228, "grad_norm": 7.290867805480957, "learning_rate": 8.925977176022246e-06, "loss": 0.5605, "step": 10560 }, { "epoch": 0.8748551564310545, "grad_norm": 5.177835464477539, "learning_rate": 8.922993495885486e-06, "loss": 0.6328, "step": 10570 }, { "epoch": 0.8756828339678862, "grad_norm": 4.927421569824219, "learning_rate": 8.920006177245543e-06, "loss": 0.6539, "step": 10580 }, { "epoch": 0.8765105115047178, "grad_norm": 5.347707271575928, "learning_rate": 8.917015222873099e-06, "loss": 0.5189, "step": 10590 }, { "epoch": 0.8773381890415494, "grad_norm": 5.15294885635376, "learning_rate": 8.914020635542203e-06, "loss": 0.6395, "step": 10600 }, { "epoch": 0.878165866578381, "grad_norm": 5.221729278564453, "learning_rate": 8.911022418030282e-06, "loss": 0.5489, "step": 10610 }, { "epoch": 0.8789935441152127, "grad_norm": 4.4085211753845215, "learning_rate": 8.90802057311813e-06, "loss": 0.6381, "step": 10620 }, { "epoch": 0.8798212216520443, "grad_norm": 4.1876220703125, "learning_rate": 8.905015103589896e-06, "loss": 0.5596, "step": 10630 }, { "epoch": 0.880648899188876, "grad_norm": 3.8062400817871094, "learning_rate": 8.902006012233102e-06, "loss": 0.5137, "step": 10640 }, { "epoch": 0.8814765767257077, "grad_norm": 4.312417507171631, "learning_rate": 8.898993301838622e-06, "loss": 0.5778, "step": 10650 }, { "epoch": 0.8823042542625393, "grad_norm": 4.872250556945801, "learning_rate": 8.89597697520069e-06, "loss": 0.663, "step": 10660 }, { "epoch": 0.883131931799371, "grad_norm": 4.912514686584473, "learning_rate": 8.892957035116888e-06, "loss": 0.6404, "step": 10670 }, { "epoch": 0.8839596093362027, "grad_norm": 4.441906929016113, "learning_rate": 8.88993348438816e-06, "loss": 0.6718, "step": 10680 }, { "epoch": 0.8847872868730343, "grad_norm": 4.704990386962891, "learning_rate": 8.88690632581879e-06, "loss": 0.5637, "step": 10690 }, { "epoch": 0.8856149644098659, "grad_norm": 4.517215728759766, "learning_rate": 8.883875562216409e-06, "loss": 0.6765, "step": 10700 }, { "epoch": 0.8864426419466975, "grad_norm": 10.296659469604492, "learning_rate": 8.880841196391997e-06, "loss": 0.6168, "step": 10710 }, { "epoch": 0.8872703194835292, "grad_norm": 2.9136416912078857, "learning_rate": 8.877803231159872e-06, "loss": 0.5907, "step": 10720 }, { "epoch": 0.8880979970203609, "grad_norm": 6.426093101501465, "learning_rate": 8.874761669337687e-06, "loss": 0.7066, "step": 10730 }, { "epoch": 0.8889256745571925, "grad_norm": 4.890642166137695, "learning_rate": 8.871716513746434e-06, "loss": 0.4958, "step": 10740 }, { "epoch": 0.8897533520940242, "grad_norm": 4.350751876831055, "learning_rate": 8.868667767210441e-06, "loss": 0.5887, "step": 10750 }, { "epoch": 0.8905810296308558, "grad_norm": 4.148437023162842, "learning_rate": 8.86561543255736e-06, "loss": 0.5431, "step": 10760 }, { "epoch": 0.8914087071676875, "grad_norm": 6.185011386871338, "learning_rate": 8.862559512618177e-06, "loss": 0.5182, "step": 10770 }, { "epoch": 0.8922363847045192, "grad_norm": 6.1957855224609375, "learning_rate": 8.8595000102272e-06, "loss": 0.5908, "step": 10780 }, { "epoch": 0.8930640622413508, "grad_norm": 3.922578811645508, "learning_rate": 8.856436928222059e-06, "loss": 0.5982, "step": 10790 }, { "epoch": 0.8938917397781824, "grad_norm": 4.52108097076416, "learning_rate": 8.853370269443707e-06, "loss": 0.5139, "step": 10800 }, { "epoch": 0.894719417315014, "grad_norm": 4.284722805023193, "learning_rate": 8.85030003673641e-06, "loss": 0.7655, "step": 10810 }, { "epoch": 0.8955470948518457, "grad_norm": 4.164956569671631, "learning_rate": 8.847226232947757e-06, "loss": 0.6321, "step": 10820 }, { "epoch": 0.8963747723886774, "grad_norm": 5.7610063552856445, "learning_rate": 8.844148860928636e-06, "loss": 0.5251, "step": 10830 }, { "epoch": 0.897202449925509, "grad_norm": 7.6746368408203125, "learning_rate": 8.84106792353326e-06, "loss": 0.5941, "step": 10840 }, { "epoch": 0.8980301274623407, "grad_norm": 5.245319843292236, "learning_rate": 8.837983423619135e-06, "loss": 0.6202, "step": 10850 }, { "epoch": 0.8988578049991723, "grad_norm": 5.604116916656494, "learning_rate": 8.834895364047079e-06, "loss": 0.6075, "step": 10860 }, { "epoch": 0.899685482536004, "grad_norm": 3.5348095893859863, "learning_rate": 8.83180374768121e-06, "loss": 0.5777, "step": 10870 }, { "epoch": 0.9005131600728357, "grad_norm": 4.439830303192139, "learning_rate": 8.828708577388944e-06, "loss": 0.6054, "step": 10880 }, { "epoch": 0.9013408376096673, "grad_norm": 5.263004779815674, "learning_rate": 8.825609856040994e-06, "loss": 0.6086, "step": 10890 }, { "epoch": 0.9021685151464989, "grad_norm": 3.6800389289855957, "learning_rate": 8.822507586511366e-06, "loss": 0.5568, "step": 10900 }, { "epoch": 0.9029961926833305, "grad_norm": 4.093077659606934, "learning_rate": 8.819401771677356e-06, "loss": 0.5261, "step": 10910 }, { "epoch": 0.9038238702201622, "grad_norm": 4.503404140472412, "learning_rate": 8.81629241441955e-06, "loss": 0.5432, "step": 10920 }, { "epoch": 0.9046515477569939, "grad_norm": 6.669723987579346, "learning_rate": 8.813179517621819e-06, "loss": 0.5522, "step": 10930 }, { "epoch": 0.9054792252938255, "grad_norm": 4.862995147705078, "learning_rate": 8.810063084171317e-06, "loss": 0.5161, "step": 10940 }, { "epoch": 0.9063069028306572, "grad_norm": 5.506811141967773, "learning_rate": 8.806943116958477e-06, "loss": 0.6283, "step": 10950 }, { "epoch": 0.9071345803674888, "grad_norm": 3.4980666637420654, "learning_rate": 8.803819618877012e-06, "loss": 0.6029, "step": 10960 }, { "epoch": 0.9079622579043205, "grad_norm": 5.281864166259766, "learning_rate": 8.800692592823904e-06, "loss": 0.6093, "step": 10970 }, { "epoch": 0.9087899354411522, "grad_norm": 5.925779342651367, "learning_rate": 8.797562041699416e-06, "loss": 0.5272, "step": 10980 }, { "epoch": 0.9096176129779838, "grad_norm": 10.292342185974121, "learning_rate": 8.794427968407074e-06, "loss": 0.6136, "step": 10990 }, { "epoch": 0.9104452905148154, "grad_norm": 5.113034725189209, "learning_rate": 8.791290375853671e-06, "loss": 0.6679, "step": 11000 }, { "epoch": 0.911272968051647, "grad_norm": 5.229300022125244, "learning_rate": 8.788149266949268e-06, "loss": 0.565, "step": 11010 }, { "epoch": 0.9121006455884787, "grad_norm": 15.635132789611816, "learning_rate": 8.785004644607183e-06, "loss": 0.6265, "step": 11020 }, { "epoch": 0.9129283231253104, "grad_norm": 5.380911350250244, "learning_rate": 8.781856511743994e-06, "loss": 0.6025, "step": 11030 }, { "epoch": 0.913756000662142, "grad_norm": 3.7259576320648193, "learning_rate": 8.778704871279537e-06, "loss": 0.5917, "step": 11040 }, { "epoch": 0.9145836781989737, "grad_norm": 6.049541473388672, "learning_rate": 8.775549726136899e-06, "loss": 0.6269, "step": 11050 }, { "epoch": 0.9154113557358053, "grad_norm": 3.978732109069824, "learning_rate": 8.772391079242419e-06, "loss": 0.6475, "step": 11060 }, { "epoch": 0.916239033272637, "grad_norm": 3.7224273681640625, "learning_rate": 8.76922893352568e-06, "loss": 0.5578, "step": 11070 }, { "epoch": 0.9170667108094687, "grad_norm": 4.457015037536621, "learning_rate": 8.766063291919518e-06, "loss": 0.4832, "step": 11080 }, { "epoch": 0.9178943883463003, "grad_norm": 4.491396427154541, "learning_rate": 8.762894157360001e-06, "loss": 0.6389, "step": 11090 }, { "epoch": 0.9187220658831319, "grad_norm": 5.511111736297607, "learning_rate": 8.759721532786447e-06, "loss": 0.7031, "step": 11100 }, { "epoch": 0.9195497434199635, "grad_norm": 3.489241600036621, "learning_rate": 8.756545421141403e-06, "loss": 0.4944, "step": 11110 }, { "epoch": 0.9203774209567952, "grad_norm": 3.4977121353149414, "learning_rate": 8.753365825370655e-06, "loss": 0.6308, "step": 11120 }, { "epoch": 0.9212050984936269, "grad_norm": 5.87430477142334, "learning_rate": 8.750182748423215e-06, "loss": 0.6137, "step": 11130 }, { "epoch": 0.9220327760304585, "grad_norm": 4.301729202270508, "learning_rate": 8.746996193251333e-06, "loss": 0.5892, "step": 11140 }, { "epoch": 0.9228604535672902, "grad_norm": 4.1149678230285645, "learning_rate": 8.743806162810475e-06, "loss": 0.5841, "step": 11150 }, { "epoch": 0.9236881311041218, "grad_norm": 4.939455032348633, "learning_rate": 8.740612660059337e-06, "loss": 0.6539, "step": 11160 }, { "epoch": 0.9245158086409535, "grad_norm": 5.4220991134643555, "learning_rate": 8.737415687959831e-06, "loss": 0.6933, "step": 11170 }, { "epoch": 0.9253434861777852, "grad_norm": 5.657209396362305, "learning_rate": 8.734215249477091e-06, "loss": 0.6312, "step": 11180 }, { "epoch": 0.9261711637146168, "grad_norm": 4.327859878540039, "learning_rate": 8.731011347579465e-06, "loss": 0.5721, "step": 11190 }, { "epoch": 0.9269988412514485, "grad_norm": 6.563467025756836, "learning_rate": 8.727803985238508e-06, "loss": 0.6576, "step": 11200 }, { "epoch": 0.92782651878828, "grad_norm": 4.424740314483643, "learning_rate": 8.724593165428992e-06, "loss": 0.5849, "step": 11210 }, { "epoch": 0.9286541963251117, "grad_norm": 5.067470550537109, "learning_rate": 8.72137889112889e-06, "loss": 0.6986, "step": 11220 }, { "epoch": 0.9294818738619434, "grad_norm": 4.919747829437256, "learning_rate": 8.718161165319385e-06, "loss": 0.6056, "step": 11230 }, { "epoch": 0.930309551398775, "grad_norm": 4.403709411621094, "learning_rate": 8.714939990984855e-06, "loss": 0.6037, "step": 11240 }, { "epoch": 0.9311372289356067, "grad_norm": 6.076066493988037, "learning_rate": 8.711715371112879e-06, "loss": 0.4969, "step": 11250 }, { "epoch": 0.9319649064724383, "grad_norm": 7.4890594482421875, "learning_rate": 8.708487308694231e-06, "loss": 0.6187, "step": 11260 }, { "epoch": 0.93279258400927, "grad_norm": 3.940394639968872, "learning_rate": 8.705255806722881e-06, "loss": 0.6364, "step": 11270 }, { "epoch": 0.9336202615461017, "grad_norm": 5.4572978019714355, "learning_rate": 8.702020868195985e-06, "loss": 0.6087, "step": 11280 }, { "epoch": 0.9344479390829333, "grad_norm": 5.093771934509277, "learning_rate": 8.698782496113887e-06, "loss": 0.5727, "step": 11290 }, { "epoch": 0.935275616619765, "grad_norm": 4.796142578125, "learning_rate": 8.69554069348012e-06, "loss": 0.6313, "step": 11300 }, { "epoch": 0.9361032941565965, "grad_norm": 3.3686892986297607, "learning_rate": 8.69229546330139e-06, "loss": 0.5821, "step": 11310 }, { "epoch": 0.9369309716934282, "grad_norm": 4.582352161407471, "learning_rate": 8.689046808587594e-06, "loss": 0.6818, "step": 11320 }, { "epoch": 0.9377586492302599, "grad_norm": 4.378573894500732, "learning_rate": 8.685794732351796e-06, "loss": 0.6386, "step": 11330 }, { "epoch": 0.9385863267670915, "grad_norm": 4.302477836608887, "learning_rate": 8.682539237610233e-06, "loss": 0.6289, "step": 11340 }, { "epoch": 0.9394140043039232, "grad_norm": 3.8949368000030518, "learning_rate": 8.679280327382319e-06, "loss": 0.4951, "step": 11350 }, { "epoch": 0.9402416818407549, "grad_norm": 6.117213726043701, "learning_rate": 8.67601800469063e-06, "loss": 0.5928, "step": 11360 }, { "epoch": 0.9410693593775865, "grad_norm": 5.024581432342529, "learning_rate": 8.672752272560911e-06, "loss": 0.6012, "step": 11370 }, { "epoch": 0.9418970369144182, "grad_norm": 6.17067289352417, "learning_rate": 8.669483134022068e-06, "loss": 0.5618, "step": 11380 }, { "epoch": 0.9427247144512498, "grad_norm": 4.449373245239258, "learning_rate": 8.666210592106165e-06, "loss": 0.7609, "step": 11390 }, { "epoch": 0.9435523919880815, "grad_norm": 2.8911995887756348, "learning_rate": 8.662934649848423e-06, "loss": 0.4888, "step": 11400 }, { "epoch": 0.944380069524913, "grad_norm": 4.171323776245117, "learning_rate": 8.659655310287218e-06, "loss": 0.6104, "step": 11410 }, { "epoch": 0.9452077470617447, "grad_norm": 5.321386814117432, "learning_rate": 8.656372576464075e-06, "loss": 0.6183, "step": 11420 }, { "epoch": 0.9460354245985764, "grad_norm": 4.635565757751465, "learning_rate": 8.653086451423672e-06, "loss": 0.5747, "step": 11430 }, { "epoch": 0.946863102135408, "grad_norm": 4.035436630249023, "learning_rate": 8.649796938213824e-06, "loss": 0.634, "step": 11440 }, { "epoch": 0.9476907796722397, "grad_norm": 5.100336074829102, "learning_rate": 8.646504039885494e-06, "loss": 0.6721, "step": 11450 }, { "epoch": 0.9485184572090714, "grad_norm": 5.214798450469971, "learning_rate": 8.643207759492785e-06, "loss": 0.6224, "step": 11460 }, { "epoch": 0.949346134745903, "grad_norm": 4.152312278747559, "learning_rate": 8.639908100092935e-06, "loss": 0.6316, "step": 11470 }, { "epoch": 0.9501738122827347, "grad_norm": 5.562802314758301, "learning_rate": 8.636605064746318e-06, "loss": 0.6323, "step": 11480 }, { "epoch": 0.9510014898195663, "grad_norm": 8.199454307556152, "learning_rate": 8.633298656516435e-06, "loss": 0.6184, "step": 11490 }, { "epoch": 0.951829167356398, "grad_norm": 5.1113715171813965, "learning_rate": 8.629988878469919e-06, "loss": 0.5271, "step": 11500 }, { "epoch": 0.9526568448932295, "grad_norm": 4.901302814483643, "learning_rate": 8.626675733676527e-06, "loss": 0.5244, "step": 11510 }, { "epoch": 0.9534845224300612, "grad_norm": 5.598797798156738, "learning_rate": 8.623359225209141e-06, "loss": 0.5941, "step": 11520 }, { "epoch": 0.9543121999668929, "grad_norm": 2.680325984954834, "learning_rate": 8.620039356143756e-06, "loss": 0.639, "step": 11530 }, { "epoch": 0.9551398775037245, "grad_norm": 4.94930362701416, "learning_rate": 8.616716129559494e-06, "loss": 0.5539, "step": 11540 }, { "epoch": 0.9559675550405562, "grad_norm": 3.9232635498046875, "learning_rate": 8.613389548538582e-06, "loss": 0.6042, "step": 11550 }, { "epoch": 0.9567952325773879, "grad_norm": 4.492664337158203, "learning_rate": 8.610059616166362e-06, "loss": 0.5887, "step": 11560 }, { "epoch": 0.9576229101142195, "grad_norm": 6.194626331329346, "learning_rate": 8.606726335531285e-06, "loss": 0.5741, "step": 11570 }, { "epoch": 0.9584505876510512, "grad_norm": 4.965316295623779, "learning_rate": 8.603389709724908e-06, "loss": 0.6132, "step": 11580 }, { "epoch": 0.9592782651878828, "grad_norm": 5.3093342781066895, "learning_rate": 8.600049741841885e-06, "loss": 0.6472, "step": 11590 }, { "epoch": 0.9601059427247145, "grad_norm": 6.977835655212402, "learning_rate": 8.596706434979979e-06, "loss": 0.6489, "step": 11600 }, { "epoch": 0.960933620261546, "grad_norm": 5.874053478240967, "learning_rate": 8.593359792240039e-06, "loss": 0.5883, "step": 11610 }, { "epoch": 0.9617612977983777, "grad_norm": 4.672454357147217, "learning_rate": 8.590009816726017e-06, "loss": 0.4686, "step": 11620 }, { "epoch": 0.9625889753352094, "grad_norm": 3.8055925369262695, "learning_rate": 8.586656511544954e-06, "loss": 0.5544, "step": 11630 }, { "epoch": 0.963416652872041, "grad_norm": 5.159496307373047, "learning_rate": 8.583299879806975e-06, "loss": 0.7009, "step": 11640 }, { "epoch": 0.9642443304088727, "grad_norm": 4.340429782867432, "learning_rate": 8.579939924625294e-06, "loss": 0.5342, "step": 11650 }, { "epoch": 0.9650720079457044, "grad_norm": 4.149951934814453, "learning_rate": 8.57657664911621e-06, "loss": 0.6678, "step": 11660 }, { "epoch": 0.965899685482536, "grad_norm": 5.489538669586182, "learning_rate": 8.573210056399091e-06, "loss": 0.6185, "step": 11670 }, { "epoch": 0.9667273630193677, "grad_norm": 4.239717960357666, "learning_rate": 8.569840149596396e-06, "loss": 0.5675, "step": 11680 }, { "epoch": 0.9675550405561993, "grad_norm": 4.307322025299072, "learning_rate": 8.566466931833646e-06, "loss": 0.5956, "step": 11690 }, { "epoch": 0.968382718093031, "grad_norm": 4.929356098175049, "learning_rate": 8.56309040623944e-06, "loss": 0.5386, "step": 11700 }, { "epoch": 0.9692103956298626, "grad_norm": 3.9974191188812256, "learning_rate": 8.55971057594544e-06, "loss": 0.6008, "step": 11710 }, { "epoch": 0.9700380731666942, "grad_norm": 4.019601821899414, "learning_rate": 8.556327444086381e-06, "loss": 0.572, "step": 11720 }, { "epoch": 0.9708657507035259, "grad_norm": 7.778622627258301, "learning_rate": 8.552941013800047e-06, "loss": 0.5728, "step": 11730 }, { "epoch": 0.9716934282403575, "grad_norm": 6.832522869110107, "learning_rate": 8.549551288227295e-06, "loss": 0.564, "step": 11740 }, { "epoch": 0.9725211057771892, "grad_norm": 6.1252121925354, "learning_rate": 8.546158270512028e-06, "loss": 0.6479, "step": 11750 }, { "epoch": 0.9733487833140209, "grad_norm": 4.74187707901001, "learning_rate": 8.54276196380121e-06, "loss": 0.6453, "step": 11760 }, { "epoch": 0.9741764608508525, "grad_norm": 4.796623706817627, "learning_rate": 8.539362371244848e-06, "loss": 0.6179, "step": 11770 }, { "epoch": 0.9750041383876842, "grad_norm": 5.4982500076293945, "learning_rate": 8.535959495996004e-06, "loss": 0.5886, "step": 11780 }, { "epoch": 0.9758318159245158, "grad_norm": 3.742030143737793, "learning_rate": 8.53255334121078e-06, "loss": 0.536, "step": 11790 }, { "epoch": 0.9766594934613475, "grad_norm": 4.652359962463379, "learning_rate": 8.529143910048321e-06, "loss": 0.6007, "step": 11800 }, { "epoch": 0.9774871709981791, "grad_norm": 4.561132431030273, "learning_rate": 8.52573120567081e-06, "loss": 0.6247, "step": 11810 }, { "epoch": 0.9783148485350107, "grad_norm": 5.5122294425964355, "learning_rate": 8.522315231243468e-06, "loss": 0.6068, "step": 11820 }, { "epoch": 0.9791425260718424, "grad_norm": 4.09151554107666, "learning_rate": 8.518895989934546e-06, "loss": 0.626, "step": 11830 }, { "epoch": 0.979970203608674, "grad_norm": 4.717523097991943, "learning_rate": 8.515473484915328e-06, "loss": 0.5799, "step": 11840 }, { "epoch": 0.9807978811455057, "grad_norm": 5.209766864776611, "learning_rate": 8.512047719360121e-06, "loss": 0.7522, "step": 11850 }, { "epoch": 0.9816255586823374, "grad_norm": 4.579046249389648, "learning_rate": 8.508618696446262e-06, "loss": 0.6227, "step": 11860 }, { "epoch": 0.982453236219169, "grad_norm": 6.950780391693115, "learning_rate": 8.505186419354104e-06, "loss": 0.5885, "step": 11870 }, { "epoch": 0.9832809137560007, "grad_norm": 4.163252353668213, "learning_rate": 8.50175089126702e-06, "loss": 0.5578, "step": 11880 }, { "epoch": 0.9841085912928323, "grad_norm": 6.352504253387451, "learning_rate": 8.498312115371396e-06, "loss": 0.5421, "step": 11890 }, { "epoch": 0.984936268829664, "grad_norm": 3.8143649101257324, "learning_rate": 8.494870094856635e-06, "loss": 0.539, "step": 11900 }, { "epoch": 0.9857639463664956, "grad_norm": 3.8517417907714844, "learning_rate": 8.491424832915148e-06, "loss": 0.6374, "step": 11910 }, { "epoch": 0.9865916239033272, "grad_norm": 6.069272518157959, "learning_rate": 8.487976332742346e-06, "loss": 0.5975, "step": 11920 }, { "epoch": 0.9874193014401589, "grad_norm": 7.386569976806641, "learning_rate": 8.484524597536654e-06, "loss": 0.5757, "step": 11930 }, { "epoch": 0.9882469789769905, "grad_norm": 5.58911657333374, "learning_rate": 8.481069630499487e-06, "loss": 0.5213, "step": 11940 }, { "epoch": 0.9890746565138222, "grad_norm": 4.932733535766602, "learning_rate": 8.477611434835268e-06, "loss": 0.6088, "step": 11950 }, { "epoch": 0.9899023340506539, "grad_norm": 3.5897300243377686, "learning_rate": 8.474150013751406e-06, "loss": 0.5408, "step": 11960 }, { "epoch": 0.9907300115874855, "grad_norm": 4.77000617980957, "learning_rate": 8.470685370458302e-06, "loss": 0.5824, "step": 11970 }, { "epoch": 0.9915576891243172, "grad_norm": 7.552821159362793, "learning_rate": 8.46721750816935e-06, "loss": 0.5259, "step": 11980 }, { "epoch": 0.9923853666611488, "grad_norm": 5.8146653175354, "learning_rate": 8.463746430100929e-06, "loss": 0.612, "step": 11990 }, { "epoch": 0.9932130441979805, "grad_norm": 4.025126934051514, "learning_rate": 8.460272139472397e-06, "loss": 0.6286, "step": 12000 }, { "epoch": 0.9940407217348122, "grad_norm": 3.9145679473876953, "learning_rate": 8.456794639506094e-06, "loss": 0.5856, "step": 12010 }, { "epoch": 0.9948683992716437, "grad_norm": 4.304585933685303, "learning_rate": 8.453313933427338e-06, "loss": 0.5602, "step": 12020 }, { "epoch": 0.9956960768084754, "grad_norm": 3.302152395248413, "learning_rate": 8.449830024464416e-06, "loss": 0.5922, "step": 12030 }, { "epoch": 0.996523754345307, "grad_norm": 6.1709723472595215, "learning_rate": 8.446342915848591e-06, "loss": 0.5936, "step": 12040 }, { "epoch": 0.9973514318821387, "grad_norm": 4.047056198120117, "learning_rate": 8.44285261081409e-06, "loss": 0.6423, "step": 12050 }, { "epoch": 0.9981791094189704, "grad_norm": 3.9220340251922607, "learning_rate": 8.439359112598106e-06, "loss": 0.589, "step": 12060 }, { "epoch": 0.999006786955802, "grad_norm": 4.662756443023682, "learning_rate": 8.435862424440793e-06, "loss": 0.4961, "step": 12070 }, { "epoch": 0.9998344644926337, "grad_norm": 9.452362060546875, "learning_rate": 8.432362549585263e-06, "loss": 0.5201, "step": 12080 }, { "epoch": 1.0006621420294652, "grad_norm": 4.2579779624938965, "learning_rate": 8.428859491277588e-06, "loss": 0.5903, "step": 12090 }, { "epoch": 1.001489819566297, "grad_norm": 5.5866169929504395, "learning_rate": 8.425353252766783e-06, "loss": 0.6147, "step": 12100 }, { "epoch": 1.0023174971031286, "grad_norm": 4.677128791809082, "learning_rate": 8.421843837304825e-06, "loss": 0.5392, "step": 12110 }, { "epoch": 1.0031451746399602, "grad_norm": 4.055984020233154, "learning_rate": 8.41833124814663e-06, "loss": 0.6005, "step": 12120 }, { "epoch": 1.0039728521767919, "grad_norm": 3.76383113861084, "learning_rate": 8.414815488550054e-06, "loss": 0.4675, "step": 12130 }, { "epoch": 1.0048005297136235, "grad_norm": 4.942571640014648, "learning_rate": 8.411296561775904e-06, "loss": 0.5246, "step": 12140 }, { "epoch": 1.0056282072504552, "grad_norm": 5.23897123336792, "learning_rate": 8.407774471087917e-06, "loss": 0.5498, "step": 12150 }, { "epoch": 1.0064558847872869, "grad_norm": 3.659213066101074, "learning_rate": 8.404249219752767e-06, "loss": 0.5314, "step": 12160 }, { "epoch": 1.0072835623241185, "grad_norm": 4.0269365310668945, "learning_rate": 8.400720811040056e-06, "loss": 0.5279, "step": 12170 }, { "epoch": 1.0081112398609502, "grad_norm": 3.0806832313537598, "learning_rate": 8.397189248222319e-06, "loss": 0.5096, "step": 12180 }, { "epoch": 1.0089389173977819, "grad_norm": 12.004498481750488, "learning_rate": 8.393654534575015e-06, "loss": 0.5262, "step": 12190 }, { "epoch": 1.0097665949346135, "grad_norm": 9.917510032653809, "learning_rate": 8.390116673376527e-06, "loss": 0.5189, "step": 12200 }, { "epoch": 1.0105942724714452, "grad_norm": 4.648590564727783, "learning_rate": 8.38657566790815e-06, "loss": 0.5161, "step": 12210 }, { "epoch": 1.0114219500082768, "grad_norm": 8.644752502441406, "learning_rate": 8.383031521454103e-06, "loss": 0.5508, "step": 12220 }, { "epoch": 1.0122496275451085, "grad_norm": 3.990889310836792, "learning_rate": 8.379484237301518e-06, "loss": 0.4536, "step": 12230 }, { "epoch": 1.0130773050819402, "grad_norm": 6.4288482666015625, "learning_rate": 8.37593381874043e-06, "loss": 0.4871, "step": 12240 }, { "epoch": 1.0139049826187718, "grad_norm": 4.514700412750244, "learning_rate": 8.372380269063788e-06, "loss": 0.5432, "step": 12250 }, { "epoch": 1.0147326601556035, "grad_norm": 5.091573238372803, "learning_rate": 8.368823591567446e-06, "loss": 0.5929, "step": 12260 }, { "epoch": 1.015560337692435, "grad_norm": 4.340421199798584, "learning_rate": 8.36526378955015e-06, "loss": 0.489, "step": 12270 }, { "epoch": 1.0163880152292666, "grad_norm": 5.485507965087891, "learning_rate": 8.361700866313555e-06, "loss": 0.5, "step": 12280 }, { "epoch": 1.0172156927660982, "grad_norm": 4.600914001464844, "learning_rate": 8.358134825162206e-06, "loss": 0.5088, "step": 12290 }, { "epoch": 1.01804337030293, "grad_norm": 3.365206003189087, "learning_rate": 8.354565669403534e-06, "loss": 0.5622, "step": 12300 }, { "epoch": 1.0188710478397616, "grad_norm": 3.6412417888641357, "learning_rate": 8.35099340234787e-06, "loss": 0.6287, "step": 12310 }, { "epoch": 1.0196987253765932, "grad_norm": 3.4910101890563965, "learning_rate": 8.34741802730842e-06, "loss": 0.52, "step": 12320 }, { "epoch": 1.020526402913425, "grad_norm": 4.179900169372559, "learning_rate": 8.343839547601282e-06, "loss": 0.4853, "step": 12330 }, { "epoch": 1.0213540804502566, "grad_norm": 5.627762317657471, "learning_rate": 8.340257966545428e-06, "loss": 0.4819, "step": 12340 }, { "epoch": 1.0221817579870882, "grad_norm": 7.498992443084717, "learning_rate": 8.336673287462706e-06, "loss": 0.526, "step": 12350 }, { "epoch": 1.0230094355239199, "grad_norm": 5.093673229217529, "learning_rate": 8.333085513677839e-06, "loss": 0.5727, "step": 12360 }, { "epoch": 1.0238371130607515, "grad_norm": 4.163850784301758, "learning_rate": 8.329494648518424e-06, "loss": 0.5013, "step": 12370 }, { "epoch": 1.0246647905975832, "grad_norm": 4.539938926696777, "learning_rate": 8.325900695314914e-06, "loss": 0.594, "step": 12380 }, { "epoch": 1.0254924681344149, "grad_norm": 7.386503219604492, "learning_rate": 8.32230365740064e-06, "loss": 0.5523, "step": 12390 }, { "epoch": 1.0263201456712465, "grad_norm": 4.856707572937012, "learning_rate": 8.318703538111787e-06, "loss": 0.5258, "step": 12400 }, { "epoch": 1.0271478232080782, "grad_norm": 6.871192455291748, "learning_rate": 8.315100340787394e-06, "loss": 0.5873, "step": 12410 }, { "epoch": 1.0279755007449098, "grad_norm": 5.226491928100586, "learning_rate": 8.311494068769362e-06, "loss": 0.5052, "step": 12420 }, { "epoch": 1.0288031782817415, "grad_norm": 5.227518558502197, "learning_rate": 8.307884725402442e-06, "loss": 0.5037, "step": 12430 }, { "epoch": 1.0296308558185732, "grad_norm": 4.742794990539551, "learning_rate": 8.304272314034229e-06, "loss": 0.5119, "step": 12440 }, { "epoch": 1.0304585333554048, "grad_norm": 5.129772186279297, "learning_rate": 8.300656838015172e-06, "loss": 0.5546, "step": 12450 }, { "epoch": 1.0312862108922365, "grad_norm": 6.231854438781738, "learning_rate": 8.297038300698554e-06, "loss": 0.547, "step": 12460 }, { "epoch": 1.032113888429068, "grad_norm": 5.2711944580078125, "learning_rate": 8.2934167054405e-06, "loss": 0.5164, "step": 12470 }, { "epoch": 1.0329415659658996, "grad_norm": 3.740936517715454, "learning_rate": 8.289792055599974e-06, "loss": 0.5294, "step": 12480 }, { "epoch": 1.0337692435027312, "grad_norm": 5.128392696380615, "learning_rate": 8.286164354538773e-06, "loss": 0.5481, "step": 12490 }, { "epoch": 1.034596921039563, "grad_norm": 4.475179195404053, "learning_rate": 8.282533605621516e-06, "loss": 0.5383, "step": 12500 }, { "epoch": 1.0354245985763946, "grad_norm": 4.70118522644043, "learning_rate": 8.278899812215662e-06, "loss": 0.5655, "step": 12510 }, { "epoch": 1.0362522761132262, "grad_norm": 5.230587005615234, "learning_rate": 8.27526297769148e-06, "loss": 0.4206, "step": 12520 }, { "epoch": 1.037079953650058, "grad_norm": 9.383631706237793, "learning_rate": 8.271623105422071e-06, "loss": 0.5884, "step": 12530 }, { "epoch": 1.0379076311868896, "grad_norm": 3.2431139945983887, "learning_rate": 8.267980198783344e-06, "loss": 0.4696, "step": 12540 }, { "epoch": 1.0387353087237212, "grad_norm": 4.503329277038574, "learning_rate": 8.26433426115403e-06, "loss": 0.5268, "step": 12550 }, { "epoch": 1.0395629862605529, "grad_norm": 6.688050270080566, "learning_rate": 8.260685295915669e-06, "loss": 0.482, "step": 12560 }, { "epoch": 1.0403906637973845, "grad_norm": 5.718338966369629, "learning_rate": 8.257033306452605e-06, "loss": 0.5029, "step": 12570 }, { "epoch": 1.0412183413342162, "grad_norm": 4.449705123901367, "learning_rate": 8.253378296151991e-06, "loss": 0.4867, "step": 12580 }, { "epoch": 1.0420460188710479, "grad_norm": 4.092711925506592, "learning_rate": 8.249720268403782e-06, "loss": 0.5137, "step": 12590 }, { "epoch": 1.0428736964078795, "grad_norm": 4.637039661407471, "learning_rate": 8.24605922660073e-06, "loss": 0.5113, "step": 12600 }, { "epoch": 1.0437013739447112, "grad_norm": 6.343768119812012, "learning_rate": 8.242395174138382e-06, "loss": 0.5249, "step": 12610 }, { "epoch": 1.0445290514815428, "grad_norm": 5.0237627029418945, "learning_rate": 8.238728114415079e-06, "loss": 0.5253, "step": 12620 }, { "epoch": 1.0453567290183745, "grad_norm": 4.911555290222168, "learning_rate": 8.23505805083195e-06, "loss": 0.4938, "step": 12630 }, { "epoch": 1.0461844065552062, "grad_norm": 5.417574882507324, "learning_rate": 8.23138498679291e-06, "loss": 0.5204, "step": 12640 }, { "epoch": 1.0470120840920378, "grad_norm": 5.859879493713379, "learning_rate": 8.227708925704659e-06, "loss": 0.5272, "step": 12650 }, { "epoch": 1.0478397616288695, "grad_norm": 3.7163021564483643, "learning_rate": 8.224029870976674e-06, "loss": 0.5226, "step": 12660 }, { "epoch": 1.0486674391657012, "grad_norm": 4.7459635734558105, "learning_rate": 8.220347826021206e-06, "loss": 0.5135, "step": 12670 }, { "epoch": 1.0494951167025326, "grad_norm": 6.927330017089844, "learning_rate": 8.216662794253289e-06, "loss": 0.4707, "step": 12680 }, { "epoch": 1.0503227942393643, "grad_norm": 5.710663795471191, "learning_rate": 8.212974779090718e-06, "loss": 0.5512, "step": 12690 }, { "epoch": 1.051150471776196, "grad_norm": 4.8944830894470215, "learning_rate": 8.20928378395406e-06, "loss": 0.565, "step": 12700 }, { "epoch": 1.0519781493130276, "grad_norm": 4.018921375274658, "learning_rate": 8.20558981226664e-06, "loss": 0.5287, "step": 12710 }, { "epoch": 1.0528058268498592, "grad_norm": 5.402132987976074, "learning_rate": 8.201892867454552e-06, "loss": 0.4768, "step": 12720 }, { "epoch": 1.053633504386691, "grad_norm": 4.921926975250244, "learning_rate": 8.19819295294664e-06, "loss": 0.5737, "step": 12730 }, { "epoch": 1.0544611819235226, "grad_norm": 6.424612045288086, "learning_rate": 8.194490072174509e-06, "loss": 0.5219, "step": 12740 }, { "epoch": 1.0552888594603542, "grad_norm": 5.143903732299805, "learning_rate": 8.190784228572507e-06, "loss": 0.5871, "step": 12750 }, { "epoch": 1.0561165369971859, "grad_norm": 6.070103645324707, "learning_rate": 8.187075425577742e-06, "loss": 0.5777, "step": 12760 }, { "epoch": 1.0569442145340175, "grad_norm": 5.262286186218262, "learning_rate": 8.183363666630053e-06, "loss": 0.4752, "step": 12770 }, { "epoch": 1.0577718920708492, "grad_norm": 5.2405781745910645, "learning_rate": 8.179648955172028e-06, "loss": 0.5693, "step": 12780 }, { "epoch": 1.0585995696076809, "grad_norm": 4.608757019042969, "learning_rate": 8.175931294648995e-06, "loss": 0.6234, "step": 12790 }, { "epoch": 1.0594272471445125, "grad_norm": 7.408969879150391, "learning_rate": 8.172210688509015e-06, "loss": 0.568, "step": 12800 }, { "epoch": 1.0602549246813442, "grad_norm": 4.684446334838867, "learning_rate": 8.168487140202876e-06, "loss": 0.4961, "step": 12810 }, { "epoch": 1.0610826022181759, "grad_norm": 4.570870876312256, "learning_rate": 8.164760653184103e-06, "loss": 0.5347, "step": 12820 }, { "epoch": 1.0619102797550075, "grad_norm": 4.582344055175781, "learning_rate": 8.161031230908944e-06, "loss": 0.5497, "step": 12830 }, { "epoch": 1.0627379572918392, "grad_norm": 6.032464504241943, "learning_rate": 8.157298876836366e-06, "loss": 0.4841, "step": 12840 }, { "epoch": 1.0635656348286708, "grad_norm": 5.224885940551758, "learning_rate": 8.15356359442806e-06, "loss": 0.5495, "step": 12850 }, { "epoch": 1.0643933123655025, "grad_norm": 3.9073660373687744, "learning_rate": 8.149825387148429e-06, "loss": 0.4829, "step": 12860 }, { "epoch": 1.065220989902334, "grad_norm": 4.2036452293396, "learning_rate": 8.14608425846459e-06, "loss": 0.5252, "step": 12870 }, { "epoch": 1.0660486674391656, "grad_norm": 4.019886493682861, "learning_rate": 8.142340211846373e-06, "loss": 0.5872, "step": 12880 }, { "epoch": 1.0668763449759973, "grad_norm": 4.345486640930176, "learning_rate": 8.13859325076631e-06, "loss": 0.4733, "step": 12890 }, { "epoch": 1.067704022512829, "grad_norm": 4.186624050140381, "learning_rate": 8.134843378699633e-06, "loss": 0.5373, "step": 12900 }, { "epoch": 1.0685317000496606, "grad_norm": 4.371232032775879, "learning_rate": 8.131090599124283e-06, "loss": 0.5595, "step": 12910 }, { "epoch": 1.0693593775864922, "grad_norm": 5.803108215332031, "learning_rate": 8.127334915520893e-06, "loss": 0.589, "step": 12920 }, { "epoch": 1.070187055123324, "grad_norm": 4.657205104827881, "learning_rate": 8.123576331372786e-06, "loss": 0.6281, "step": 12930 }, { "epoch": 1.0710147326601556, "grad_norm": 4.77728271484375, "learning_rate": 8.119814850165983e-06, "loss": 0.5494, "step": 12940 }, { "epoch": 1.0718424101969872, "grad_norm": 3.3915305137634277, "learning_rate": 8.116050475389185e-06, "loss": 0.4664, "step": 12950 }, { "epoch": 1.0726700877338189, "grad_norm": 6.732929706573486, "learning_rate": 8.112283210533778e-06, "loss": 0.4659, "step": 12960 }, { "epoch": 1.0734977652706506, "grad_norm": 11.031495094299316, "learning_rate": 8.108513059093834e-06, "loss": 0.5252, "step": 12970 }, { "epoch": 1.0743254428074822, "grad_norm": 5.385486602783203, "learning_rate": 8.104740024566093e-06, "loss": 0.4843, "step": 12980 }, { "epoch": 1.0751531203443139, "grad_norm": 5.743884086608887, "learning_rate": 8.10096411044998e-06, "loss": 0.5354, "step": 12990 }, { "epoch": 1.0759807978811455, "grad_norm": 4.8538923263549805, "learning_rate": 8.097185320247579e-06, "loss": 0.5899, "step": 13000 }, { "epoch": 1.0768084754179772, "grad_norm": 4.198935031890869, "learning_rate": 8.09340365746365e-06, "loss": 0.4672, "step": 13010 }, { "epoch": 1.0776361529548089, "grad_norm": 4.631344318389893, "learning_rate": 8.089619125605613e-06, "loss": 0.5249, "step": 13020 }, { "epoch": 1.0784638304916405, "grad_norm": 5.111343860626221, "learning_rate": 8.085831728183553e-06, "loss": 0.6169, "step": 13030 }, { "epoch": 1.0792915080284722, "grad_norm": 4.795958995819092, "learning_rate": 8.082041468710208e-06, "loss": 0.5776, "step": 13040 }, { "epoch": 1.0801191855653038, "grad_norm": 10.299899101257324, "learning_rate": 8.078248350700978e-06, "loss": 0.5705, "step": 13050 }, { "epoch": 1.0809468631021355, "grad_norm": 4.788166046142578, "learning_rate": 8.074452377673901e-06, "loss": 0.5759, "step": 13060 }, { "epoch": 1.0817745406389672, "grad_norm": 4.298780918121338, "learning_rate": 8.070653553149676e-06, "loss": 0.4218, "step": 13070 }, { "epoch": 1.0826022181757988, "grad_norm": 5.270418643951416, "learning_rate": 8.06685188065164e-06, "loss": 0.5121, "step": 13080 }, { "epoch": 1.0834298957126303, "grad_norm": 5.595242977142334, "learning_rate": 8.063047363705777e-06, "loss": 0.4816, "step": 13090 }, { "epoch": 1.084257573249462, "grad_norm": 5.6150898933410645, "learning_rate": 8.0592400058407e-06, "loss": 0.523, "step": 13100 }, { "epoch": 1.0850852507862936, "grad_norm": 4.42987585067749, "learning_rate": 8.055429810587668e-06, "loss": 0.5428, "step": 13110 }, { "epoch": 1.0859129283231252, "grad_norm": 5.027298450469971, "learning_rate": 8.05161678148056e-06, "loss": 0.4933, "step": 13120 }, { "epoch": 1.086740605859957, "grad_norm": 4.4268412590026855, "learning_rate": 8.047800922055895e-06, "loss": 0.62, "step": 13130 }, { "epoch": 1.0875682833967886, "grad_norm": 4.804693698883057, "learning_rate": 8.043982235852806e-06, "loss": 0.4773, "step": 13140 }, { "epoch": 1.0883959609336202, "grad_norm": 4.404482364654541, "learning_rate": 8.040160726413057e-06, "loss": 0.5716, "step": 13150 }, { "epoch": 1.089223638470452, "grad_norm": 4.000824451446533, "learning_rate": 8.036336397281024e-06, "loss": 0.5415, "step": 13160 }, { "epoch": 1.0900513160072836, "grad_norm": 4.276086330413818, "learning_rate": 8.032509252003702e-06, "loss": 0.5225, "step": 13170 }, { "epoch": 1.0908789935441152, "grad_norm": 5.200662136077881, "learning_rate": 8.0286792941307e-06, "loss": 0.5996, "step": 13180 }, { "epoch": 1.0917066710809469, "grad_norm": 3.7599315643310547, "learning_rate": 8.024846527214225e-06, "loss": 0.4291, "step": 13190 }, { "epoch": 1.0925343486177785, "grad_norm": 5.098784923553467, "learning_rate": 8.021010954809102e-06, "loss": 0.5085, "step": 13200 }, { "epoch": 1.0933620261546102, "grad_norm": 4.457065105438232, "learning_rate": 8.017172580472751e-06, "loss": 0.4645, "step": 13210 }, { "epoch": 1.0941897036914419, "grad_norm": 6.021533966064453, "learning_rate": 8.013331407765194e-06, "loss": 0.5448, "step": 13220 }, { "epoch": 1.0950173812282735, "grad_norm": 4.480930805206299, "learning_rate": 8.009487440249047e-06, "loss": 0.5747, "step": 13230 }, { "epoch": 1.0958450587651052, "grad_norm": 4.339868068695068, "learning_rate": 8.005640681489517e-06, "loss": 0.5334, "step": 13240 }, { "epoch": 1.0966727363019368, "grad_norm": 4.2518310546875, "learning_rate": 8.0017911350544e-06, "loss": 0.5862, "step": 13250 }, { "epoch": 1.0975004138387685, "grad_norm": 3.8625926971435547, "learning_rate": 7.99793880451408e-06, "loss": 0.5285, "step": 13260 }, { "epoch": 1.0983280913756002, "grad_norm": 5.532796382904053, "learning_rate": 7.99408369344152e-06, "loss": 0.5752, "step": 13270 }, { "epoch": 1.0991557689124316, "grad_norm": 5.075899600982666, "learning_rate": 7.990225805412266e-06, "loss": 0.5428, "step": 13280 }, { "epoch": 1.0999834464492633, "grad_norm": 5.449062347412109, "learning_rate": 7.986365144004435e-06, "loss": 0.5819, "step": 13290 }, { "epoch": 1.100811123986095, "grad_norm": 4.856030464172363, "learning_rate": 7.98250171279872e-06, "loss": 0.4973, "step": 13300 }, { "epoch": 1.1016388015229266, "grad_norm": 4.686433792114258, "learning_rate": 7.978635515378378e-06, "loss": 0.5783, "step": 13310 }, { "epoch": 1.1024664790597583, "grad_norm": 5.1001362800598145, "learning_rate": 7.97476655532924e-06, "loss": 0.6014, "step": 13320 }, { "epoch": 1.10329415659659, "grad_norm": 3.821927547454834, "learning_rate": 7.970894836239688e-06, "loss": 0.5466, "step": 13330 }, { "epoch": 1.1041218341334216, "grad_norm": 4.915258407592773, "learning_rate": 7.967020361700673e-06, "loss": 0.514, "step": 13340 }, { "epoch": 1.1049495116702532, "grad_norm": 5.438986778259277, "learning_rate": 7.9631431353057e-06, "loss": 0.5692, "step": 13350 }, { "epoch": 1.105777189207085, "grad_norm": 5.483590126037598, "learning_rate": 7.959263160650819e-06, "loss": 0.6163, "step": 13360 }, { "epoch": 1.1066048667439166, "grad_norm": 4.044435501098633, "learning_rate": 7.955380441334635e-06, "loss": 0.4473, "step": 13370 }, { "epoch": 1.1074325442807482, "grad_norm": 4.617773056030273, "learning_rate": 7.951494980958303e-06, "loss": 0.4973, "step": 13380 }, { "epoch": 1.1082602218175799, "grad_norm": 4.903042316436768, "learning_rate": 7.947606783125507e-06, "loss": 0.5455, "step": 13390 }, { "epoch": 1.1090878993544115, "grad_norm": 5.368357181549072, "learning_rate": 7.943715851442483e-06, "loss": 0.53, "step": 13400 }, { "epoch": 1.1099155768912432, "grad_norm": 4.600588798522949, "learning_rate": 7.939822189517996e-06, "loss": 0.6302, "step": 13410 }, { "epoch": 1.1107432544280749, "grad_norm": 4.509477615356445, "learning_rate": 7.935925800963345e-06, "loss": 0.5227, "step": 13420 }, { "epoch": 1.1115709319649065, "grad_norm": 3.799985408782959, "learning_rate": 7.932026689392358e-06, "loss": 0.5529, "step": 13430 }, { "epoch": 1.1123986095017382, "grad_norm": 3.1460869312286377, "learning_rate": 7.928124858421387e-06, "loss": 0.4902, "step": 13440 }, { "epoch": 1.1132262870385699, "grad_norm": 4.26908016204834, "learning_rate": 7.924220311669308e-06, "loss": 0.5113, "step": 13450 }, { "epoch": 1.1140539645754015, "grad_norm": 4.5507354736328125, "learning_rate": 7.920313052757515e-06, "loss": 0.5615, "step": 13460 }, { "epoch": 1.1148816421122332, "grad_norm": 4.5587687492370605, "learning_rate": 7.916403085309917e-06, "loss": 0.5683, "step": 13470 }, { "epoch": 1.1157093196490648, "grad_norm": 3.892963409423828, "learning_rate": 7.912490412952938e-06, "loss": 0.5289, "step": 13480 }, { "epoch": 1.1165369971858963, "grad_norm": 3.5938198566436768, "learning_rate": 7.908575039315504e-06, "loss": 0.5811, "step": 13490 }, { "epoch": 1.117364674722728, "grad_norm": 4.0914082527160645, "learning_rate": 7.904656968029058e-06, "loss": 0.5112, "step": 13500 }, { "epoch": 1.1181923522595596, "grad_norm": 3.496018409729004, "learning_rate": 7.900736202727533e-06, "loss": 0.4566, "step": 13510 }, { "epoch": 1.1190200297963913, "grad_norm": 4.367074012756348, "learning_rate": 7.896812747047364e-06, "loss": 0.5387, "step": 13520 }, { "epoch": 1.119847707333223, "grad_norm": 4.303811073303223, "learning_rate": 7.892886604627487e-06, "loss": 0.6168, "step": 13530 }, { "epoch": 1.1206753848700546, "grad_norm": 4.786275386810303, "learning_rate": 7.888957779109324e-06, "loss": 0.5511, "step": 13540 }, { "epoch": 1.1215030624068862, "grad_norm": 5.858986854553223, "learning_rate": 7.885026274136787e-06, "loss": 0.4186, "step": 13550 }, { "epoch": 1.122330739943718, "grad_norm": 4.38638973236084, "learning_rate": 7.881092093356273e-06, "loss": 0.5806, "step": 13560 }, { "epoch": 1.1231584174805496, "grad_norm": 4.94023323059082, "learning_rate": 7.877155240416663e-06, "loss": 0.5074, "step": 13570 }, { "epoch": 1.1239860950173812, "grad_norm": 6.814149856567383, "learning_rate": 7.873215718969313e-06, "loss": 0.5221, "step": 13580 }, { "epoch": 1.1248137725542129, "grad_norm": 3.3779373168945312, "learning_rate": 7.869273532668055e-06, "loss": 0.544, "step": 13590 }, { "epoch": 1.1256414500910445, "grad_norm": 5.174316883087158, "learning_rate": 7.865328685169195e-06, "loss": 0.5373, "step": 13600 }, { "epoch": 1.1264691276278762, "grad_norm": 9.28717041015625, "learning_rate": 7.861381180131506e-06, "loss": 0.5501, "step": 13610 }, { "epoch": 1.1272968051647079, "grad_norm": 4.616262912750244, "learning_rate": 7.857431021216222e-06, "loss": 0.6062, "step": 13620 }, { "epoch": 1.1281244827015395, "grad_norm": 4.358617782592773, "learning_rate": 7.853478212087045e-06, "loss": 0.5679, "step": 13630 }, { "epoch": 1.1289521602383712, "grad_norm": 4.082508087158203, "learning_rate": 7.849522756410129e-06, "loss": 0.4714, "step": 13640 }, { "epoch": 1.1297798377752029, "grad_norm": 4.988232612609863, "learning_rate": 7.845564657854086e-06, "loss": 0.5297, "step": 13650 }, { "epoch": 1.1306075153120345, "grad_norm": 5.196653842926025, "learning_rate": 7.841603920089979e-06, "loss": 0.4989, "step": 13660 }, { "epoch": 1.1314351928488662, "grad_norm": 3.805586814880371, "learning_rate": 7.837640546791316e-06, "loss": 0.5635, "step": 13670 }, { "epoch": 1.1322628703856976, "grad_norm": 8.56905460357666, "learning_rate": 7.833674541634057e-06, "loss": 0.5928, "step": 13680 }, { "epoch": 1.1330905479225293, "grad_norm": 4.850135326385498, "learning_rate": 7.829705908296592e-06, "loss": 0.6378, "step": 13690 }, { "epoch": 1.133918225459361, "grad_norm": 4.482873439788818, "learning_rate": 7.825734650459758e-06, "loss": 0.5272, "step": 13700 }, { "epoch": 1.1347459029961926, "grad_norm": 4.825905799865723, "learning_rate": 7.82176077180682e-06, "loss": 0.5351, "step": 13710 }, { "epoch": 1.1355735805330243, "grad_norm": 3.6959104537963867, "learning_rate": 7.81778427602348e-06, "loss": 0.5676, "step": 13720 }, { "epoch": 1.136401258069856, "grad_norm": 3.846830129623413, "learning_rate": 7.813805166797857e-06, "loss": 0.5243, "step": 13730 }, { "epoch": 1.1372289356066876, "grad_norm": 3.8342015743255615, "learning_rate": 7.809823447820506e-06, "loss": 0.6026, "step": 13740 }, { "epoch": 1.1380566131435192, "grad_norm": 3.2163431644439697, "learning_rate": 7.805839122784396e-06, "loss": 0.535, "step": 13750 }, { "epoch": 1.138884290680351, "grad_norm": 4.739038944244385, "learning_rate": 7.801852195384913e-06, "loss": 0.5155, "step": 13760 }, { "epoch": 1.1397119682171826, "grad_norm": 4.5617876052856445, "learning_rate": 7.797862669319858e-06, "loss": 0.5168, "step": 13770 }, { "epoch": 1.1405396457540142, "grad_norm": 4.256223678588867, "learning_rate": 7.79387054828944e-06, "loss": 0.5429, "step": 13780 }, { "epoch": 1.141367323290846, "grad_norm": 4.727659702301025, "learning_rate": 7.78987583599628e-06, "loss": 0.507, "step": 13790 }, { "epoch": 1.1421950008276776, "grad_norm": 3.9745349884033203, "learning_rate": 7.785878536145399e-06, "loss": 0.564, "step": 13800 }, { "epoch": 1.1430226783645092, "grad_norm": 3.8055927753448486, "learning_rate": 7.781878652444216e-06, "loss": 0.5535, "step": 13810 }, { "epoch": 1.1438503559013409, "grad_norm": 4.74408483505249, "learning_rate": 7.77787618860255e-06, "loss": 0.4965, "step": 13820 }, { "epoch": 1.1446780334381725, "grad_norm": 4.228391170501709, "learning_rate": 7.773871148332609e-06, "loss": 0.5103, "step": 13830 }, { "epoch": 1.1455057109750042, "grad_norm": 5.780876159667969, "learning_rate": 7.769863535348994e-06, "loss": 0.5539, "step": 13840 }, { "epoch": 1.1463333885118359, "grad_norm": 5.516439437866211, "learning_rate": 7.765853353368692e-06, "loss": 0.5233, "step": 13850 }, { "epoch": 1.1471610660486675, "grad_norm": 5.959690093994141, "learning_rate": 7.761840606111073e-06, "loss": 0.5123, "step": 13860 }, { "epoch": 1.1479887435854992, "grad_norm": 5.51188325881958, "learning_rate": 7.757825297297882e-06, "loss": 0.5306, "step": 13870 }, { "epoch": 1.1488164211223308, "grad_norm": 4.8426055908203125, "learning_rate": 7.753807430653246e-06, "loss": 0.6021, "step": 13880 }, { "epoch": 1.1496440986591625, "grad_norm": 4.403369903564453, "learning_rate": 7.74978700990366e-06, "loss": 0.4582, "step": 13890 }, { "epoch": 1.150471776195994, "grad_norm": 4.511072158813477, "learning_rate": 7.745764038777988e-06, "loss": 0.4523, "step": 13900 }, { "epoch": 1.1512994537328256, "grad_norm": 4.811670780181885, "learning_rate": 7.741738521007462e-06, "loss": 0.5334, "step": 13910 }, { "epoch": 1.1521271312696573, "grad_norm": 4.581150054931641, "learning_rate": 7.737710460325674e-06, "loss": 0.5509, "step": 13920 }, { "epoch": 1.152954808806489, "grad_norm": 2.952906608581543, "learning_rate": 7.733679860468575e-06, "loss": 0.5269, "step": 13930 }, { "epoch": 1.1537824863433206, "grad_norm": 3.484959363937378, "learning_rate": 7.729646725174471e-06, "loss": 0.4496, "step": 13940 }, { "epoch": 1.1546101638801523, "grad_norm": 4.614175319671631, "learning_rate": 7.725611058184019e-06, "loss": 0.556, "step": 13950 }, { "epoch": 1.155437841416984, "grad_norm": 4.163119316101074, "learning_rate": 7.721572863240227e-06, "loss": 0.5343, "step": 13960 }, { "epoch": 1.1562655189538156, "grad_norm": 3.5966482162475586, "learning_rate": 7.717532144088444e-06, "loss": 0.5587, "step": 13970 }, { "epoch": 1.1570931964906472, "grad_norm": 6.855951309204102, "learning_rate": 7.713488904476362e-06, "loss": 0.5061, "step": 13980 }, { "epoch": 1.157920874027479, "grad_norm": 4.27704381942749, "learning_rate": 7.709443148154008e-06, "loss": 0.5641, "step": 13990 }, { "epoch": 1.1587485515643106, "grad_norm": 3.731775999069214, "learning_rate": 7.705394878873749e-06, "loss": 0.5619, "step": 14000 }, { "epoch": 1.1595762291011422, "grad_norm": 5.7600836753845215, "learning_rate": 7.701344100390276e-06, "loss": 0.4762, "step": 14010 }, { "epoch": 1.1604039066379739, "grad_norm": 5.302695274353027, "learning_rate": 7.697290816460612e-06, "loss": 0.4556, "step": 14020 }, { "epoch": 1.1612315841748055, "grad_norm": 4.323911190032959, "learning_rate": 7.693235030844102e-06, "loss": 0.5003, "step": 14030 }, { "epoch": 1.1620592617116372, "grad_norm": 5.102097034454346, "learning_rate": 7.68917674730241e-06, "loss": 0.5631, "step": 14040 }, { "epoch": 1.1628869392484689, "grad_norm": 5.366724491119385, "learning_rate": 7.68511596959952e-06, "loss": 0.5362, "step": 14050 }, { "epoch": 1.1637146167853005, "grad_norm": 4.368926048278809, "learning_rate": 7.681052701501728e-06, "loss": 0.5298, "step": 14060 }, { "epoch": 1.1645422943221322, "grad_norm": 4.982699394226074, "learning_rate": 7.676986946777637e-06, "loss": 0.4987, "step": 14070 }, { "epoch": 1.1653699718589636, "grad_norm": 7.01771879196167, "learning_rate": 7.67291870919816e-06, "loss": 0.5188, "step": 14080 }, { "epoch": 1.1661976493957953, "grad_norm": 4.325961112976074, "learning_rate": 7.668847992536511e-06, "loss": 0.584, "step": 14090 }, { "epoch": 1.167025326932627, "grad_norm": 4.827487945556641, "learning_rate": 7.664774800568205e-06, "loss": 0.5674, "step": 14100 }, { "epoch": 1.1678530044694586, "grad_norm": 3.478350877761841, "learning_rate": 7.660699137071047e-06, "loss": 0.4969, "step": 14110 }, { "epoch": 1.1686806820062903, "grad_norm": 4.445967674255371, "learning_rate": 7.656621005825145e-06, "loss": 0.4848, "step": 14120 }, { "epoch": 1.169508359543122, "grad_norm": 6.423912525177002, "learning_rate": 7.652540410612885e-06, "loss": 0.5428, "step": 14130 }, { "epoch": 1.1703360370799536, "grad_norm": 5.257619380950928, "learning_rate": 7.648457355218944e-06, "loss": 0.5545, "step": 14140 }, { "epoch": 1.1711637146167853, "grad_norm": 4.464993953704834, "learning_rate": 7.644371843430279e-06, "loss": 0.5896, "step": 14150 }, { "epoch": 1.171991392153617, "grad_norm": 4.109048843383789, "learning_rate": 7.640283879036128e-06, "loss": 0.4785, "step": 14160 }, { "epoch": 1.1728190696904486, "grad_norm": 4.619018077850342, "learning_rate": 7.636193465828e-06, "loss": 0.444, "step": 14170 }, { "epoch": 1.1736467472272802, "grad_norm": 4.831028461456299, "learning_rate": 7.632100607599677e-06, "loss": 0.5786, "step": 14180 }, { "epoch": 1.174474424764112, "grad_norm": 4.772920608520508, "learning_rate": 7.628005308147211e-06, "loss": 0.4541, "step": 14190 }, { "epoch": 1.1753021023009436, "grad_norm": 4.710547924041748, "learning_rate": 7.6239075712689105e-06, "loss": 0.4879, "step": 14200 }, { "epoch": 1.1761297798377752, "grad_norm": 5.154600620269775, "learning_rate": 7.619807400765354e-06, "loss": 0.4678, "step": 14210 }, { "epoch": 1.1769574573746069, "grad_norm": 4.161100387573242, "learning_rate": 7.6157048004393715e-06, "loss": 0.468, "step": 14220 }, { "epoch": 1.1777851349114385, "grad_norm": 4.9395365715026855, "learning_rate": 7.611599774096049e-06, "loss": 0.4975, "step": 14230 }, { "epoch": 1.1786128124482702, "grad_norm": 5.822861671447754, "learning_rate": 7.6074923255427216e-06, "loss": 0.5806, "step": 14240 }, { "epoch": 1.1794404899851019, "grad_norm": 4.676958084106445, "learning_rate": 7.60338245858897e-06, "loss": 0.5062, "step": 14250 }, { "epoch": 1.1802681675219335, "grad_norm": 4.6155877113342285, "learning_rate": 7.599270177046621e-06, "loss": 0.4788, "step": 14260 }, { "epoch": 1.1810958450587652, "grad_norm": 6.515411376953125, "learning_rate": 7.595155484729737e-06, "loss": 0.5505, "step": 14270 }, { "epoch": 1.1819235225955969, "grad_norm": 4.268336772918701, "learning_rate": 7.591038385454619e-06, "loss": 0.57, "step": 14280 }, { "epoch": 1.1827512001324285, "grad_norm": 3.152036666870117, "learning_rate": 7.5869188830398e-06, "loss": 0.5179, "step": 14290 }, { "epoch": 1.1835788776692602, "grad_norm": 2.737365245819092, "learning_rate": 7.582796981306042e-06, "loss": 0.5206, "step": 14300 }, { "epoch": 1.1844065552060916, "grad_norm": 4.97829008102417, "learning_rate": 7.578672684076329e-06, "loss": 0.5512, "step": 14310 }, { "epoch": 1.1852342327429233, "grad_norm": 4.136258602142334, "learning_rate": 7.574545995175873e-06, "loss": 0.5367, "step": 14320 }, { "epoch": 1.186061910279755, "grad_norm": 4.557806968688965, "learning_rate": 7.5704169184321e-06, "loss": 0.5299, "step": 14330 }, { "epoch": 1.1868895878165866, "grad_norm": 3.3367574214935303, "learning_rate": 7.5662854576746515e-06, "loss": 0.5123, "step": 14340 }, { "epoch": 1.1877172653534183, "grad_norm": 4.352635383605957, "learning_rate": 7.56215161673538e-06, "loss": 0.5484, "step": 14350 }, { "epoch": 1.18854494289025, "grad_norm": 6.9415717124938965, "learning_rate": 7.558015399448344e-06, "loss": 0.5251, "step": 14360 }, { "epoch": 1.1893726204270816, "grad_norm": 5.521824836730957, "learning_rate": 7.55387680964981e-06, "loss": 0.5211, "step": 14370 }, { "epoch": 1.1902002979639132, "grad_norm": 4.190937042236328, "learning_rate": 7.549735851178242e-06, "loss": 0.4927, "step": 14380 }, { "epoch": 1.191027975500745, "grad_norm": 2.9800562858581543, "learning_rate": 7.545592527874302e-06, "loss": 0.428, "step": 14390 }, { "epoch": 1.1918556530375766, "grad_norm": 3.7488462924957275, "learning_rate": 7.5414468435808445e-06, "loss": 0.5238, "step": 14400 }, { "epoch": 1.1926833305744082, "grad_norm": 2.829303741455078, "learning_rate": 7.537298802142915e-06, "loss": 0.4452, "step": 14410 }, { "epoch": 1.19351100811124, "grad_norm": 3.6135036945343018, "learning_rate": 7.533148407407743e-06, "loss": 0.4722, "step": 14420 }, { "epoch": 1.1943386856480716, "grad_norm": 4.38289213180542, "learning_rate": 7.528995663224743e-06, "loss": 0.6153, "step": 14430 }, { "epoch": 1.1951663631849032, "grad_norm": 4.724255084991455, "learning_rate": 7.524840573445509e-06, "loss": 0.4631, "step": 14440 }, { "epoch": 1.1959940407217349, "grad_norm": 4.386266708374023, "learning_rate": 7.520683141923807e-06, "loss": 0.4692, "step": 14450 }, { "epoch": 1.1968217182585665, "grad_norm": 4.722465991973877, "learning_rate": 7.5165233725155775e-06, "loss": 0.5078, "step": 14460 }, { "epoch": 1.1976493957953982, "grad_norm": 6.413486003875732, "learning_rate": 7.512361269078929e-06, "loss": 0.4573, "step": 14470 }, { "epoch": 1.1984770733322299, "grad_norm": 9.461272239685059, "learning_rate": 7.508196835474135e-06, "loss": 0.519, "step": 14480 }, { "epoch": 1.1993047508690613, "grad_norm": 3.2223453521728516, "learning_rate": 7.5040300755636305e-06, "loss": 0.5234, "step": 14490 }, { "epoch": 1.200132428405893, "grad_norm": 4.183562755584717, "learning_rate": 7.499860993212007e-06, "loss": 0.5112, "step": 14500 }, { "epoch": 1.2009601059427246, "grad_norm": 5.343631744384766, "learning_rate": 7.49568959228601e-06, "loss": 0.5358, "step": 14510 }, { "epoch": 1.2017877834795563, "grad_norm": 4.703043460845947, "learning_rate": 7.491515876654536e-06, "loss": 0.5224, "step": 14520 }, { "epoch": 1.202615461016388, "grad_norm": 5.664011478424072, "learning_rate": 7.487339850188627e-06, "loss": 0.4798, "step": 14530 }, { "epoch": 1.2034431385532196, "grad_norm": 4.5571746826171875, "learning_rate": 7.483161516761474e-06, "loss": 0.6113, "step": 14540 }, { "epoch": 1.2042708160900513, "grad_norm": 3.7921829223632812, "learning_rate": 7.478980880248398e-06, "loss": 0.5102, "step": 14550 }, { "epoch": 1.205098493626883, "grad_norm": 4.284449577331543, "learning_rate": 7.474797944526865e-06, "loss": 0.5441, "step": 14560 }, { "epoch": 1.2059261711637146, "grad_norm": 5.51432991027832, "learning_rate": 7.470612713476469e-06, "loss": 0.4713, "step": 14570 }, { "epoch": 1.2067538487005462, "grad_norm": 3.588151693344116, "learning_rate": 7.46642519097893e-06, "loss": 0.4866, "step": 14580 }, { "epoch": 1.207581526237378, "grad_norm": 4.491145610809326, "learning_rate": 7.462235380918101e-06, "loss": 0.525, "step": 14590 }, { "epoch": 1.2084092037742096, "grad_norm": 4.033792018890381, "learning_rate": 7.458043287179953e-06, "loss": 0.5243, "step": 14600 }, { "epoch": 1.2092368813110412, "grad_norm": 8.704864501953125, "learning_rate": 7.453848913652571e-06, "loss": 0.5303, "step": 14610 }, { "epoch": 1.210064558847873, "grad_norm": 4.799952507019043, "learning_rate": 7.449652264226162e-06, "loss": 0.4538, "step": 14620 }, { "epoch": 1.2108922363847046, "grad_norm": 5.950053691864014, "learning_rate": 7.445453342793037e-06, "loss": 0.5246, "step": 14630 }, { "epoch": 1.2117199139215362, "grad_norm": 3.1047070026397705, "learning_rate": 7.441252153247618e-06, "loss": 0.453, "step": 14640 }, { "epoch": 1.2125475914583679, "grad_norm": 5.296886920928955, "learning_rate": 7.437048699486429e-06, "loss": 0.6074, "step": 14650 }, { "epoch": 1.2133752689951995, "grad_norm": 3.060892343521118, "learning_rate": 7.432842985408098e-06, "loss": 0.4937, "step": 14660 }, { "epoch": 1.2142029465320312, "grad_norm": 5.916923522949219, "learning_rate": 7.4286350149133426e-06, "loss": 0.5428, "step": 14670 }, { "epoch": 1.2150306240688629, "grad_norm": 4.48724889755249, "learning_rate": 7.424424791904979e-06, "loss": 0.5335, "step": 14680 }, { "epoch": 1.2158583016056945, "grad_norm": 3.1871442794799805, "learning_rate": 7.420212320287909e-06, "loss": 0.5361, "step": 14690 }, { "epoch": 1.2166859791425262, "grad_norm": 4.879150390625, "learning_rate": 7.415997603969123e-06, "loss": 0.5375, "step": 14700 }, { "epoch": 1.2175136566793576, "grad_norm": 4.330007076263428, "learning_rate": 7.4117806468576905e-06, "loss": 0.5575, "step": 14710 }, { "epoch": 1.2183413342161893, "grad_norm": 4.503864288330078, "learning_rate": 7.407561452864761e-06, "loss": 0.559, "step": 14720 }, { "epoch": 1.219169011753021, "grad_norm": 3.4837846755981445, "learning_rate": 7.403340025903557e-06, "loss": 0.5192, "step": 14730 }, { "epoch": 1.2199966892898526, "grad_norm": 5.131473541259766, "learning_rate": 7.399116369889376e-06, "loss": 0.6064, "step": 14740 }, { "epoch": 1.2208243668266843, "grad_norm": 10.506852149963379, "learning_rate": 7.3948904887395745e-06, "loss": 0.5934, "step": 14750 }, { "epoch": 1.221652044363516, "grad_norm": 4.027787685394287, "learning_rate": 7.390662386373584e-06, "loss": 0.5218, "step": 14760 }, { "epoch": 1.2224797219003476, "grad_norm": 7.608210563659668, "learning_rate": 7.3864320667128885e-06, "loss": 0.621, "step": 14770 }, { "epoch": 1.2233073994371793, "grad_norm": 5.95069694519043, "learning_rate": 7.382199533681031e-06, "loss": 0.5514, "step": 14780 }, { "epoch": 1.224135076974011, "grad_norm": 4.565195560455322, "learning_rate": 7.377964791203603e-06, "loss": 0.5574, "step": 14790 }, { "epoch": 1.2249627545108426, "grad_norm": 5.668195724487305, "learning_rate": 7.373727843208256e-06, "loss": 0.5944, "step": 14800 }, { "epoch": 1.2257904320476742, "grad_norm": 5.777683258056641, "learning_rate": 7.369488693624673e-06, "loss": 0.4995, "step": 14810 }, { "epoch": 1.226618109584506, "grad_norm": 5.175351619720459, "learning_rate": 7.36524734638459e-06, "loss": 0.4854, "step": 14820 }, { "epoch": 1.2274457871213376, "grad_norm": 4.15785026550293, "learning_rate": 7.361003805421778e-06, "loss": 0.5259, "step": 14830 }, { "epoch": 1.2282734646581692, "grad_norm": 4.758409023284912, "learning_rate": 7.35675807467204e-06, "loss": 0.5279, "step": 14840 }, { "epoch": 1.2291011421950009, "grad_norm": 4.910494327545166, "learning_rate": 7.352510158073212e-06, "loss": 0.5831, "step": 14850 }, { "epoch": 1.2299288197318325, "grad_norm": 4.115296840667725, "learning_rate": 7.348260059565158e-06, "loss": 0.4822, "step": 14860 }, { "epoch": 1.2307564972686642, "grad_norm": 4.5270562171936035, "learning_rate": 7.344007783089765e-06, "loss": 0.551, "step": 14870 }, { "epoch": 1.2315841748054959, "grad_norm": 3.6031360626220703, "learning_rate": 7.3397533325909394e-06, "loss": 0.5567, "step": 14880 }, { "epoch": 1.2324118523423273, "grad_norm": 3.908078670501709, "learning_rate": 7.335496712014606e-06, "loss": 0.5717, "step": 14890 }, { "epoch": 1.233239529879159, "grad_norm": 4.745827674865723, "learning_rate": 7.3312379253087e-06, "loss": 0.4763, "step": 14900 }, { "epoch": 1.2340672074159906, "grad_norm": 3.9471640586853027, "learning_rate": 7.326976976423166e-06, "loss": 0.5383, "step": 14910 }, { "epoch": 1.2348948849528223, "grad_norm": 7.008781433105469, "learning_rate": 7.3227138693099535e-06, "loss": 0.5015, "step": 14920 }, { "epoch": 1.235722562489654, "grad_norm": 3.9592959880828857, "learning_rate": 7.318448607923017e-06, "loss": 0.5972, "step": 14930 }, { "epoch": 1.2365502400264856, "grad_norm": 4.601749420166016, "learning_rate": 7.314181196218306e-06, "loss": 0.4803, "step": 14940 }, { "epoch": 1.2373779175633173, "grad_norm": 5.029962062835693, "learning_rate": 7.309911638153765e-06, "loss": 0.5744, "step": 14950 }, { "epoch": 1.238205595100149, "grad_norm": 3.921144723892212, "learning_rate": 7.3056399376893284e-06, "loss": 0.5402, "step": 14960 }, { "epoch": 1.2390332726369806, "grad_norm": 3.8433289527893066, "learning_rate": 7.30136609878692e-06, "loss": 0.5426, "step": 14970 }, { "epoch": 1.2398609501738123, "grad_norm": 7.477971076965332, "learning_rate": 7.297090125410446e-06, "loss": 0.5155, "step": 14980 }, { "epoch": 1.240688627710644, "grad_norm": 4.884349346160889, "learning_rate": 7.29281202152579e-06, "loss": 0.6034, "step": 14990 }, { "epoch": 1.2415163052474756, "grad_norm": 4.53805685043335, "learning_rate": 7.288531791100814e-06, "loss": 0.6141, "step": 15000 }, { "epoch": 1.2423439827843072, "grad_norm": 4.461956977844238, "learning_rate": 7.2842494381053544e-06, "loss": 0.4884, "step": 15010 }, { "epoch": 1.243171660321139, "grad_norm": 4.376681327819824, "learning_rate": 7.279964966511208e-06, "loss": 0.5888, "step": 15020 }, { "epoch": 1.2439993378579706, "grad_norm": 5.736946105957031, "learning_rate": 7.275678380292147e-06, "loss": 0.5612, "step": 15030 }, { "epoch": 1.2448270153948022, "grad_norm": 5.065586566925049, "learning_rate": 7.2713896834239e-06, "loss": 0.5705, "step": 15040 }, { "epoch": 1.2456546929316339, "grad_norm": 5.161757469177246, "learning_rate": 7.267098879884149e-06, "loss": 0.5358, "step": 15050 }, { "epoch": 1.2464823704684656, "grad_norm": 4.385884761810303, "learning_rate": 7.2628059736525395e-06, "loss": 0.5359, "step": 15060 }, { "epoch": 1.2473100480052972, "grad_norm": 5.21990966796875, "learning_rate": 7.258510968710657e-06, "loss": 0.556, "step": 15070 }, { "epoch": 1.2481377255421289, "grad_norm": 3.686868190765381, "learning_rate": 7.254213869042042e-06, "loss": 0.5832, "step": 15080 }, { "epoch": 1.2489654030789605, "grad_norm": 5.68666410446167, "learning_rate": 7.24991467863217e-06, "loss": 0.5107, "step": 15090 }, { "epoch": 1.2497930806157922, "grad_norm": 4.0952301025390625, "learning_rate": 7.245613401468466e-06, "loss": 0.5107, "step": 15100 }, { "epoch": 1.2506207581526239, "grad_norm": 4.177093982696533, "learning_rate": 7.241310041540279e-06, "loss": 0.4977, "step": 15110 }, { "epoch": 1.2514484356894555, "grad_norm": 4.795938491821289, "learning_rate": 7.237004602838897e-06, "loss": 0.5907, "step": 15120 }, { "epoch": 1.252276113226287, "grad_norm": 4.544720649719238, "learning_rate": 7.2326970893575344e-06, "loss": 0.5195, "step": 15130 }, { "epoch": 1.2531037907631186, "grad_norm": 5.686479568481445, "learning_rate": 7.228387505091328e-06, "loss": 0.5665, "step": 15140 }, { "epoch": 1.2539314682999503, "grad_norm": 4.0465922355651855, "learning_rate": 7.224075854037338e-06, "loss": 0.4751, "step": 15150 }, { "epoch": 1.254759145836782, "grad_norm": 5.339515209197998, "learning_rate": 7.219762140194539e-06, "loss": 0.608, "step": 15160 }, { "epoch": 1.2555868233736136, "grad_norm": 3.924895763397217, "learning_rate": 7.215446367563821e-06, "loss": 0.4676, "step": 15170 }, { "epoch": 1.2564145009104453, "grad_norm": 4.048250675201416, "learning_rate": 7.211128540147983e-06, "loss": 0.5582, "step": 15180 }, { "epoch": 1.257242178447277, "grad_norm": 4.150835990905762, "learning_rate": 7.2068086619517255e-06, "loss": 0.4814, "step": 15190 }, { "epoch": 1.2580698559841086, "grad_norm": 3.839165449142456, "learning_rate": 7.20248673698166e-06, "loss": 0.4677, "step": 15200 }, { "epoch": 1.2588975335209402, "grad_norm": 4.625576496124268, "learning_rate": 7.198162769246288e-06, "loss": 0.5723, "step": 15210 }, { "epoch": 1.259725211057772, "grad_norm": 4.687930107116699, "learning_rate": 7.1938367627560114e-06, "loss": 0.4924, "step": 15220 }, { "epoch": 1.2605528885946036, "grad_norm": 3.772742748260498, "learning_rate": 7.1895087215231175e-06, "loss": 0.5301, "step": 15230 }, { "epoch": 1.2613805661314352, "grad_norm": 4.789167881011963, "learning_rate": 7.1851786495617855e-06, "loss": 0.4632, "step": 15240 }, { "epoch": 1.262208243668267, "grad_norm": 4.7335662841796875, "learning_rate": 7.1808465508880755e-06, "loss": 0.5159, "step": 15250 }, { "epoch": 1.2630359212050986, "grad_norm": 4.9275383949279785, "learning_rate": 7.176512429519927e-06, "loss": 0.5664, "step": 15260 }, { "epoch": 1.2638635987419302, "grad_norm": 6.30328893661499, "learning_rate": 7.1721762894771595e-06, "loss": 0.504, "step": 15270 }, { "epoch": 1.2646912762787617, "grad_norm": 5.6261138916015625, "learning_rate": 7.167838134781461e-06, "loss": 0.4818, "step": 15280 }, { "epoch": 1.2655189538155933, "grad_norm": 4.608573913574219, "learning_rate": 7.163497969456386e-06, "loss": 0.5607, "step": 15290 }, { "epoch": 1.266346631352425, "grad_norm": 5.248178482055664, "learning_rate": 7.15915579752736e-06, "loss": 0.5809, "step": 15300 }, { "epoch": 1.2671743088892566, "grad_norm": 4.23056173324585, "learning_rate": 7.154811623021665e-06, "loss": 0.5241, "step": 15310 }, { "epoch": 1.2680019864260883, "grad_norm": 4.322697162628174, "learning_rate": 7.150465449968443e-06, "loss": 0.4481, "step": 15320 }, { "epoch": 1.26882966396292, "grad_norm": 4.059395790100098, "learning_rate": 7.146117282398688e-06, "loss": 0.4081, "step": 15330 }, { "epoch": 1.2696573414997516, "grad_norm": 4.709310054779053, "learning_rate": 7.141767124345243e-06, "loss": 0.5386, "step": 15340 }, { "epoch": 1.2704850190365833, "grad_norm": 4.049015522003174, "learning_rate": 7.137414979842797e-06, "loss": 0.5197, "step": 15350 }, { "epoch": 1.271312696573415, "grad_norm": 4.057380199432373, "learning_rate": 7.133060852927886e-06, "loss": 0.5195, "step": 15360 }, { "epoch": 1.2721403741102466, "grad_norm": 3.5303122997283936, "learning_rate": 7.12870474763888e-06, "loss": 0.5261, "step": 15370 }, { "epoch": 1.2729680516470783, "grad_norm": 4.534543991088867, "learning_rate": 7.124346668015984e-06, "loss": 0.5357, "step": 15380 }, { "epoch": 1.27379572918391, "grad_norm": 3.9425249099731445, "learning_rate": 7.11998661810124e-06, "loss": 0.5277, "step": 15390 }, { "epoch": 1.2746234067207416, "grad_norm": 4.674860954284668, "learning_rate": 7.115624601938508e-06, "loss": 0.5046, "step": 15400 }, { "epoch": 1.2754510842575733, "grad_norm": 3.7010905742645264, "learning_rate": 7.111260623573476e-06, "loss": 0.392, "step": 15410 }, { "epoch": 1.276278761794405, "grad_norm": 6.164430618286133, "learning_rate": 7.106894687053656e-06, "loss": 0.5063, "step": 15420 }, { "epoch": 1.2771064393312366, "grad_norm": 4.550137996673584, "learning_rate": 7.102526796428373e-06, "loss": 0.5107, "step": 15430 }, { "epoch": 1.2779341168680682, "grad_norm": 5.200800895690918, "learning_rate": 7.098156955748758e-06, "loss": 0.567, "step": 15440 }, { "epoch": 1.2787617944049, "grad_norm": 4.151554584503174, "learning_rate": 7.093785169067764e-06, "loss": 0.499, "step": 15450 }, { "epoch": 1.2795894719417316, "grad_norm": 5.61179256439209, "learning_rate": 7.089411440440137e-06, "loss": 0.5862, "step": 15460 }, { "epoch": 1.2804171494785632, "grad_norm": 5.0833258628845215, "learning_rate": 7.085035773922428e-06, "loss": 0.4649, "step": 15470 }, { "epoch": 1.2812448270153949, "grad_norm": 4.790840148925781, "learning_rate": 7.08065817357299e-06, "loss": 0.499, "step": 15480 }, { "epoch": 1.2820725045522265, "grad_norm": 4.773902416229248, "learning_rate": 7.076278643451963e-06, "loss": 0.5241, "step": 15490 }, { "epoch": 1.2829001820890582, "grad_norm": 4.8187665939331055, "learning_rate": 7.071897187621279e-06, "loss": 0.5433, "step": 15500 }, { "epoch": 1.2837278596258899, "grad_norm": 3.83878755569458, "learning_rate": 7.067513810144658e-06, "loss": 0.5135, "step": 15510 }, { "epoch": 1.2845555371627215, "grad_norm": 4.503657817840576, "learning_rate": 7.063128515087598e-06, "loss": 0.4512, "step": 15520 }, { "epoch": 1.2853832146995532, "grad_norm": 5.919280052185059, "learning_rate": 7.058741306517382e-06, "loss": 0.5018, "step": 15530 }, { "epoch": 1.2862108922363846, "grad_norm": 3.407658338546753, "learning_rate": 7.054352188503061e-06, "loss": 0.5097, "step": 15540 }, { "epoch": 1.2870385697732163, "grad_norm": 4.0305070877075195, "learning_rate": 7.049961165115461e-06, "loss": 0.5874, "step": 15550 }, { "epoch": 1.287866247310048, "grad_norm": 3.842447519302368, "learning_rate": 7.045568240427174e-06, "loss": 0.5058, "step": 15560 }, { "epoch": 1.2886939248468796, "grad_norm": 7.874479293823242, "learning_rate": 7.041173418512554e-06, "loss": 0.4612, "step": 15570 }, { "epoch": 1.2895216023837113, "grad_norm": 4.689648151397705, "learning_rate": 7.0367767034477204e-06, "loss": 0.5413, "step": 15580 }, { "epoch": 1.290349279920543, "grad_norm": 4.647397041320801, "learning_rate": 7.032378099310542e-06, "loss": 0.5324, "step": 15590 }, { "epoch": 1.2911769574573746, "grad_norm": 5.232405185699463, "learning_rate": 7.027977610180641e-06, "loss": 0.5146, "step": 15600 }, { "epoch": 1.2920046349942063, "grad_norm": 9.008837699890137, "learning_rate": 7.023575240139389e-06, "loss": 0.5211, "step": 15610 }, { "epoch": 1.292832312531038, "grad_norm": 3.4537198543548584, "learning_rate": 7.0191709932699034e-06, "loss": 0.5291, "step": 15620 }, { "epoch": 1.2936599900678696, "grad_norm": 4.833574295043945, "learning_rate": 7.014764873657039e-06, "loss": 0.5909, "step": 15630 }, { "epoch": 1.2944876676047012, "grad_norm": 3.9166674613952637, "learning_rate": 7.010356885387393e-06, "loss": 0.5299, "step": 15640 }, { "epoch": 1.295315345141533, "grad_norm": 4.498083591461182, "learning_rate": 7.005947032549287e-06, "loss": 0.4846, "step": 15650 }, { "epoch": 1.2961430226783646, "grad_norm": 5.166796684265137, "learning_rate": 7.001535319232783e-06, "loss": 0.5519, "step": 15660 }, { "epoch": 1.2969707002151962, "grad_norm": 5.114251136779785, "learning_rate": 6.997121749529656e-06, "loss": 0.4732, "step": 15670 }, { "epoch": 1.2977983777520279, "grad_norm": 3.8739941120147705, "learning_rate": 6.992706327533416e-06, "loss": 0.5116, "step": 15680 }, { "epoch": 1.2986260552888593, "grad_norm": 4.259034156799316, "learning_rate": 6.988289057339279e-06, "loss": 0.5081, "step": 15690 }, { "epoch": 1.299453732825691, "grad_norm": 5.837639331817627, "learning_rate": 6.983869943044185e-06, "loss": 0.5486, "step": 15700 }, { "epoch": 1.3002814103625226, "grad_norm": 5.076846599578857, "learning_rate": 6.979448988746777e-06, "loss": 0.5151, "step": 15710 }, { "epoch": 1.3011090878993543, "grad_norm": 4.468459129333496, "learning_rate": 6.97502619854741e-06, "loss": 0.5013, "step": 15720 }, { "epoch": 1.301936765436186, "grad_norm": 4.608712196350098, "learning_rate": 6.970601576548136e-06, "loss": 0.4856, "step": 15730 }, { "epoch": 1.3027644429730176, "grad_norm": 2.9254019260406494, "learning_rate": 6.9661751268527115e-06, "loss": 0.5192, "step": 15740 }, { "epoch": 1.3035921205098493, "grad_norm": 4.668467998504639, "learning_rate": 6.9617468535665865e-06, "loss": 0.5458, "step": 15750 }, { "epoch": 1.304419798046681, "grad_norm": 3.8550522327423096, "learning_rate": 6.9573167607969e-06, "loss": 0.4629, "step": 15760 }, { "epoch": 1.3052474755835126, "grad_norm": 3.0289571285247803, "learning_rate": 6.952884852652481e-06, "loss": 0.4755, "step": 15770 }, { "epoch": 1.3060751531203443, "grad_norm": 4.308413505554199, "learning_rate": 6.948451133243841e-06, "loss": 0.4759, "step": 15780 }, { "epoch": 1.306902830657176, "grad_norm": 4.596653461456299, "learning_rate": 6.9440156066831736e-06, "loss": 0.4661, "step": 15790 }, { "epoch": 1.3077305081940076, "grad_norm": 5.542954444885254, "learning_rate": 6.9395782770843445e-06, "loss": 0.4803, "step": 15800 }, { "epoch": 1.3085581857308393, "grad_norm": 3.8407368659973145, "learning_rate": 6.935139148562896e-06, "loss": 0.4954, "step": 15810 }, { "epoch": 1.309385863267671, "grad_norm": 4.402979373931885, "learning_rate": 6.930698225236034e-06, "loss": 0.4558, "step": 15820 }, { "epoch": 1.3102135408045026, "grad_norm": 5.203643321990967, "learning_rate": 6.926255511222637e-06, "loss": 0.5096, "step": 15830 }, { "epoch": 1.3110412183413342, "grad_norm": 5.968966484069824, "learning_rate": 6.921811010643235e-06, "loss": 0.5619, "step": 15840 }, { "epoch": 1.311868895878166, "grad_norm": 4.6355671882629395, "learning_rate": 6.917364727620019e-06, "loss": 0.6069, "step": 15850 }, { "epoch": 1.3126965734149976, "grad_norm": 5.51416015625, "learning_rate": 6.912916666276839e-06, "loss": 0.5366, "step": 15860 }, { "epoch": 1.3135242509518292, "grad_norm": 2.9898557662963867, "learning_rate": 6.908466830739182e-06, "loss": 0.4143, "step": 15870 }, { "epoch": 1.314351928488661, "grad_norm": 3.734379291534424, "learning_rate": 6.9040152251341894e-06, "loss": 0.5278, "step": 15880 }, { "epoch": 1.3151796060254926, "grad_norm": 4.31004524230957, "learning_rate": 6.899561853590645e-06, "loss": 0.5899, "step": 15890 }, { "epoch": 1.3160072835623242, "grad_norm": 3.258831262588501, "learning_rate": 6.895106720238964e-06, "loss": 0.5325, "step": 15900 }, { "epoch": 1.3168349610991559, "grad_norm": 6.051225185394287, "learning_rate": 6.8906498292112025e-06, "loss": 0.516, "step": 15910 }, { "epoch": 1.3176626386359875, "grad_norm": 4.483006000518799, "learning_rate": 6.886191184641041e-06, "loss": 0.5001, "step": 15920 }, { "epoch": 1.3184903161728192, "grad_norm": 5.81752347946167, "learning_rate": 6.881730790663791e-06, "loss": 0.5207, "step": 15930 }, { "epoch": 1.3193179937096506, "grad_norm": 3.366375684738159, "learning_rate": 6.877268651416384e-06, "loss": 0.4983, "step": 15940 }, { "epoch": 1.3201456712464823, "grad_norm": 5.648759841918945, "learning_rate": 6.87280477103737e-06, "loss": 0.5081, "step": 15950 }, { "epoch": 1.320973348783314, "grad_norm": 5.5440568923950195, "learning_rate": 6.868339153666916e-06, "loss": 0.488, "step": 15960 }, { "epoch": 1.3218010263201456, "grad_norm": 4.067953109741211, "learning_rate": 6.863871803446795e-06, "loss": 0.595, "step": 15970 }, { "epoch": 1.3226287038569773, "grad_norm": 4.461876392364502, "learning_rate": 6.859402724520398e-06, "loss": 0.5182, "step": 15980 }, { "epoch": 1.323456381393809, "grad_norm": 5.78801155090332, "learning_rate": 6.854931921032707e-06, "loss": 0.532, "step": 15990 }, { "epoch": 1.3242840589306406, "grad_norm": 3.6251752376556396, "learning_rate": 6.850459397130309e-06, "loss": 0.5098, "step": 16000 }, { "epoch": 1.3251117364674723, "grad_norm": 5.284784317016602, "learning_rate": 6.845985156961387e-06, "loss": 0.5636, "step": 16010 }, { "epoch": 1.325939414004304, "grad_norm": 5.904472827911377, "learning_rate": 6.841509204675716e-06, "loss": 0.5511, "step": 16020 }, { "epoch": 1.3267670915411356, "grad_norm": 5.009022235870361, "learning_rate": 6.8370315444246574e-06, "loss": 0.5051, "step": 16030 }, { "epoch": 1.3275947690779673, "grad_norm": 4.404051303863525, "learning_rate": 6.832552180361158e-06, "loss": 0.4972, "step": 16040 }, { "epoch": 1.328422446614799, "grad_norm": 4.629133224487305, "learning_rate": 6.828071116639744e-06, "loss": 0.5237, "step": 16050 }, { "epoch": 1.3292501241516306, "grad_norm": 4.040166854858398, "learning_rate": 6.823588357416516e-06, "loss": 0.552, "step": 16060 }, { "epoch": 1.3300778016884622, "grad_norm": 3.865670919418335, "learning_rate": 6.819103906849154e-06, "loss": 0.6224, "step": 16070 }, { "epoch": 1.330905479225294, "grad_norm": 5.230727672576904, "learning_rate": 6.814617769096898e-06, "loss": 0.5251, "step": 16080 }, { "epoch": 1.3317331567621256, "grad_norm": 5.241765975952148, "learning_rate": 6.81012994832056e-06, "loss": 0.5315, "step": 16090 }, { "epoch": 1.332560834298957, "grad_norm": 6.004307270050049, "learning_rate": 6.805640448682508e-06, "loss": 0.4895, "step": 16100 }, { "epoch": 1.3333885118357887, "grad_norm": 4.16942024230957, "learning_rate": 6.80114927434667e-06, "loss": 0.4735, "step": 16110 }, { "epoch": 1.3342161893726203, "grad_norm": 3.997509241104126, "learning_rate": 6.796656429478525e-06, "loss": 0.5424, "step": 16120 }, { "epoch": 1.335043866909452, "grad_norm": 3.928248405456543, "learning_rate": 6.7921619182451035e-06, "loss": 0.5021, "step": 16130 }, { "epoch": 1.3358715444462836, "grad_norm": 4.469057559967041, "learning_rate": 6.7876657448149826e-06, "loss": 0.479, "step": 16140 }, { "epoch": 1.3366992219831153, "grad_norm": 3.2943105697631836, "learning_rate": 6.783167913358276e-06, "loss": 0.5484, "step": 16150 }, { "epoch": 1.337526899519947, "grad_norm": 3.9518074989318848, "learning_rate": 6.77866842804664e-06, "loss": 0.4932, "step": 16160 }, { "epoch": 1.3383545770567786, "grad_norm": 9.36893367767334, "learning_rate": 6.774167293053263e-06, "loss": 0.4933, "step": 16170 }, { "epoch": 1.3391822545936103, "grad_norm": 4.266812801361084, "learning_rate": 6.769664512552864e-06, "loss": 0.4819, "step": 16180 }, { "epoch": 1.340009932130442, "grad_norm": 2.43154239654541, "learning_rate": 6.765160090721687e-06, "loss": 0.4619, "step": 16190 }, { "epoch": 1.3408376096672736, "grad_norm": 4.550715446472168, "learning_rate": 6.7606540317375e-06, "loss": 0.507, "step": 16200 }, { "epoch": 1.3416652872041053, "grad_norm": 4.761877536773682, "learning_rate": 6.756146339779588e-06, "loss": 0.5288, "step": 16210 }, { "epoch": 1.342492964740937, "grad_norm": 5.6451592445373535, "learning_rate": 6.75163701902875e-06, "loss": 0.533, "step": 16220 }, { "epoch": 1.3433206422777686, "grad_norm": 4.433726787567139, "learning_rate": 6.747126073667299e-06, "loss": 0.5762, "step": 16230 }, { "epoch": 1.3441483198146003, "grad_norm": 4.6672139167785645, "learning_rate": 6.74261350787905e-06, "loss": 0.507, "step": 16240 }, { "epoch": 1.344975997351432, "grad_norm": 5.780379772186279, "learning_rate": 6.738099325849327e-06, "loss": 0.5421, "step": 16250 }, { "epoch": 1.3458036748882636, "grad_norm": 5.410336017608643, "learning_rate": 6.733583531764947e-06, "loss": 0.535, "step": 16260 }, { "epoch": 1.3466313524250952, "grad_norm": 7.223818302154541, "learning_rate": 6.729066129814223e-06, "loss": 0.4701, "step": 16270 }, { "epoch": 1.347459029961927, "grad_norm": 3.280339241027832, "learning_rate": 6.724547124186964e-06, "loss": 0.5382, "step": 16280 }, { "epoch": 1.3482867074987586, "grad_norm": 2.9953091144561768, "learning_rate": 6.720026519074463e-06, "loss": 0.4546, "step": 16290 }, { "epoch": 1.3491143850355902, "grad_norm": 5.723715305328369, "learning_rate": 6.715504318669495e-06, "loss": 0.4991, "step": 16300 }, { "epoch": 1.3499420625724219, "grad_norm": 3.526946783065796, "learning_rate": 6.710980527166318e-06, "loss": 0.4987, "step": 16310 }, { "epoch": 1.3507697401092535, "grad_norm": 5.0376434326171875, "learning_rate": 6.706455148760664e-06, "loss": 0.4824, "step": 16320 }, { "epoch": 1.3515974176460852, "grad_norm": 7.554466724395752, "learning_rate": 6.701928187649737e-06, "loss": 0.4235, "step": 16330 }, { "epoch": 1.3524250951829169, "grad_norm": 3.0390775203704834, "learning_rate": 6.697399648032207e-06, "loss": 0.5171, "step": 16340 }, { "epoch": 1.3532527727197483, "grad_norm": 4.7751030921936035, "learning_rate": 6.692869534108211e-06, "loss": 0.5114, "step": 16350 }, { "epoch": 1.35408045025658, "grad_norm": 4.189140796661377, "learning_rate": 6.6883378500793465e-06, "loss": 0.5592, "step": 16360 }, { "epoch": 1.3549081277934116, "grad_norm": 4.376929759979248, "learning_rate": 6.6838046001486645e-06, "loss": 0.5426, "step": 16370 }, { "epoch": 1.3557358053302433, "grad_norm": 5.652745246887207, "learning_rate": 6.679269788520669e-06, "loss": 0.5143, "step": 16380 }, { "epoch": 1.356563482867075, "grad_norm": 4.304300785064697, "learning_rate": 6.674733419401316e-06, "loss": 0.5287, "step": 16390 }, { "epoch": 1.3573911604039066, "grad_norm": 4.9771833419799805, "learning_rate": 6.670195496997998e-06, "loss": 0.4753, "step": 16400 }, { "epoch": 1.3582188379407383, "grad_norm": 5.458761692047119, "learning_rate": 6.665656025519557e-06, "loss": 0.4985, "step": 16410 }, { "epoch": 1.35904651547757, "grad_norm": 4.717292308807373, "learning_rate": 6.661115009176269e-06, "loss": 0.4896, "step": 16420 }, { "epoch": 1.3598741930144016, "grad_norm": 7.8386383056640625, "learning_rate": 6.656572452179838e-06, "loss": 0.5715, "step": 16430 }, { "epoch": 1.3607018705512333, "grad_norm": 3.206807851791382, "learning_rate": 6.652028358743403e-06, "loss": 0.455, "step": 16440 }, { "epoch": 1.361529548088065, "grad_norm": 5.1854658126831055, "learning_rate": 6.647482733081526e-06, "loss": 0.4753, "step": 16450 }, { "epoch": 1.3623572256248966, "grad_norm": 4.486457824707031, "learning_rate": 6.64293557941019e-06, "loss": 0.5434, "step": 16460 }, { "epoch": 1.3631849031617282, "grad_norm": 5.162239074707031, "learning_rate": 6.638386901946794e-06, "loss": 0.5016, "step": 16470 }, { "epoch": 1.36401258069856, "grad_norm": 4.528653144836426, "learning_rate": 6.633836704910151e-06, "loss": 0.5465, "step": 16480 }, { "epoch": 1.3648402582353916, "grad_norm": 5.9014201164245605, "learning_rate": 6.629284992520487e-06, "loss": 0.5294, "step": 16490 }, { "epoch": 1.365667935772223, "grad_norm": 6.426442623138428, "learning_rate": 6.624731768999426e-06, "loss": 0.4901, "step": 16500 }, { "epoch": 1.3664956133090547, "grad_norm": 3.552138328552246, "learning_rate": 6.6201770385699995e-06, "loss": 0.538, "step": 16510 }, { "epoch": 1.3673232908458863, "grad_norm": 4.2509589195251465, "learning_rate": 6.615620805456636e-06, "loss": 0.5991, "step": 16520 }, { "epoch": 1.368150968382718, "grad_norm": 4.161865234375, "learning_rate": 6.611063073885156e-06, "loss": 0.4826, "step": 16530 }, { "epoch": 1.3689786459195497, "grad_norm": 2.87579345703125, "learning_rate": 6.606503848082769e-06, "loss": 0.4627, "step": 16540 }, { "epoch": 1.3698063234563813, "grad_norm": 8.269384384155273, "learning_rate": 6.601943132278074e-06, "loss": 0.5891, "step": 16550 }, { "epoch": 1.370634000993213, "grad_norm": 4.360108375549316, "learning_rate": 6.597380930701046e-06, "loss": 0.5009, "step": 16560 }, { "epoch": 1.3714616785300446, "grad_norm": 4.754298210144043, "learning_rate": 6.592817247583045e-06, "loss": 0.4853, "step": 16570 }, { "epoch": 1.3722893560668763, "grad_norm": 3.9913313388824463, "learning_rate": 6.588252087156803e-06, "loss": 0.4588, "step": 16580 }, { "epoch": 1.373117033603708, "grad_norm": 4.051966190338135, "learning_rate": 6.583685453656417e-06, "loss": 0.5133, "step": 16590 }, { "epoch": 1.3739447111405396, "grad_norm": 5.236918926239014, "learning_rate": 6.579117351317359e-06, "loss": 0.4973, "step": 16600 }, { "epoch": 1.3747723886773713, "grad_norm": 6.156694412231445, "learning_rate": 6.574547784376453e-06, "loss": 0.4777, "step": 16610 }, { "epoch": 1.375600066214203, "grad_norm": 3.866133213043213, "learning_rate": 6.569976757071891e-06, "loss": 0.505, "step": 16620 }, { "epoch": 1.3764277437510346, "grad_norm": 4.722629547119141, "learning_rate": 6.565404273643215e-06, "loss": 0.5068, "step": 16630 }, { "epoch": 1.3772554212878663, "grad_norm": 4.494420528411865, "learning_rate": 6.5608303383313174e-06, "loss": 0.5157, "step": 16640 }, { "epoch": 1.378083098824698, "grad_norm": 3.8187670707702637, "learning_rate": 6.556254955378437e-06, "loss": 0.5116, "step": 16650 }, { "epoch": 1.3789107763615296, "grad_norm": 4.299365997314453, "learning_rate": 6.551678129028157e-06, "loss": 0.4534, "step": 16660 }, { "epoch": 1.3797384538983613, "grad_norm": 4.473401069641113, "learning_rate": 6.547099863525398e-06, "loss": 0.5622, "step": 16670 }, { "epoch": 1.380566131435193, "grad_norm": 4.731281757354736, "learning_rate": 6.5425201631164145e-06, "loss": 0.4727, "step": 16680 }, { "epoch": 1.3813938089720246, "grad_norm": 3.908998727798462, "learning_rate": 6.5379390320487965e-06, "loss": 0.5073, "step": 16690 }, { "epoch": 1.3822214865088562, "grad_norm": 4.96552848815918, "learning_rate": 6.533356474571455e-06, "loss": 0.5654, "step": 16700 }, { "epoch": 1.383049164045688, "grad_norm": 4.804232120513916, "learning_rate": 6.528772494934625e-06, "loss": 0.5186, "step": 16710 }, { "epoch": 1.3838768415825196, "grad_norm": 5.500203609466553, "learning_rate": 6.524187097389867e-06, "loss": 0.5787, "step": 16720 }, { "epoch": 1.3847045191193512, "grad_norm": 4.251430511474609, "learning_rate": 6.519600286190045e-06, "loss": 0.5505, "step": 16730 }, { "epoch": 1.3855321966561829, "grad_norm": 4.478255748748779, "learning_rate": 6.5150120655893455e-06, "loss": 0.5151, "step": 16740 }, { "epoch": 1.3863598741930143, "grad_norm": 5.137304306030273, "learning_rate": 6.510422439843256e-06, "loss": 0.4702, "step": 16750 }, { "epoch": 1.387187551729846, "grad_norm": 4.26886510848999, "learning_rate": 6.50583141320857e-06, "loss": 0.5815, "step": 16760 }, { "epoch": 1.3880152292666776, "grad_norm": 4.753237247467041, "learning_rate": 6.501238989943375e-06, "loss": 0.5054, "step": 16770 }, { "epoch": 1.3888429068035093, "grad_norm": 5.908036231994629, "learning_rate": 6.496645174307062e-06, "loss": 0.5, "step": 16780 }, { "epoch": 1.389670584340341, "grad_norm": 3.776991605758667, "learning_rate": 6.492049970560307e-06, "loss": 0.5353, "step": 16790 }, { "epoch": 1.3904982618771726, "grad_norm": 6.236854076385498, "learning_rate": 6.487453382965075e-06, "loss": 0.4742, "step": 16800 }, { "epoch": 1.3913259394140043, "grad_norm": 3.632115364074707, "learning_rate": 6.482855415784618e-06, "loss": 0.5246, "step": 16810 }, { "epoch": 1.392153616950836, "grad_norm": 2.9482967853546143, "learning_rate": 6.4782560732834635e-06, "loss": 0.5222, "step": 16820 }, { "epoch": 1.3929812944876676, "grad_norm": 4.794979572296143, "learning_rate": 6.473655359727413e-06, "loss": 0.4975, "step": 16830 }, { "epoch": 1.3938089720244993, "grad_norm": 4.532748222351074, "learning_rate": 6.469053279383546e-06, "loss": 0.5785, "step": 16840 }, { "epoch": 1.394636649561331, "grad_norm": 5.398358345031738, "learning_rate": 6.464449836520204e-06, "loss": 0.5425, "step": 16850 }, { "epoch": 1.3954643270981626, "grad_norm": 4.704810619354248, "learning_rate": 6.4598450354069955e-06, "loss": 0.5048, "step": 16860 }, { "epoch": 1.3962920046349943, "grad_norm": 5.823635578155518, "learning_rate": 6.455238880314787e-06, "loss": 0.5936, "step": 16870 }, { "epoch": 1.397119682171826, "grad_norm": 6.192531585693359, "learning_rate": 6.450631375515701e-06, "loss": 0.5575, "step": 16880 }, { "epoch": 1.3979473597086576, "grad_norm": 5.110040664672852, "learning_rate": 6.446022525283111e-06, "loss": 0.5093, "step": 16890 }, { "epoch": 1.3987750372454892, "grad_norm": 4.305551528930664, "learning_rate": 6.441412333891642e-06, "loss": 0.5533, "step": 16900 }, { "epoch": 1.3996027147823207, "grad_norm": 4.771417140960693, "learning_rate": 6.43680080561716e-06, "loss": 0.5228, "step": 16910 }, { "epoch": 1.4004303923191523, "grad_norm": 5.044512748718262, "learning_rate": 6.43218794473677e-06, "loss": 0.5268, "step": 16920 }, { "epoch": 1.401258069855984, "grad_norm": 4.7248854637146, "learning_rate": 6.427573755528815e-06, "loss": 0.5597, "step": 16930 }, { "epoch": 1.4020857473928157, "grad_norm": 4.702849388122559, "learning_rate": 6.422958242272868e-06, "loss": 0.5547, "step": 16940 }, { "epoch": 1.4029134249296473, "grad_norm": 4.459070682525635, "learning_rate": 6.418341409249733e-06, "loss": 0.4444, "step": 16950 }, { "epoch": 1.403741102466479, "grad_norm": 4.318877220153809, "learning_rate": 6.413723260741437e-06, "loss": 0.4762, "step": 16960 }, { "epoch": 1.4045687800033106, "grad_norm": 4.672999858856201, "learning_rate": 6.409103801031226e-06, "loss": 0.5294, "step": 16970 }, { "epoch": 1.4053964575401423, "grad_norm": 4.003495216369629, "learning_rate": 6.404483034403559e-06, "loss": 0.4986, "step": 16980 }, { "epoch": 1.406224135076974, "grad_norm": 4.601378917694092, "learning_rate": 6.399860965144116e-06, "loss": 0.4985, "step": 16990 }, { "epoch": 1.4070518126138056, "grad_norm": 8.104015350341797, "learning_rate": 6.395237597539776e-06, "loss": 0.51, "step": 17000 }, { "epoch": 1.4078794901506373, "grad_norm": 6.262922286987305, "learning_rate": 6.390612935878628e-06, "loss": 0.398, "step": 17010 }, { "epoch": 1.408707167687469, "grad_norm": 5.603402614593506, "learning_rate": 6.385986984449961e-06, "loss": 0.5727, "step": 17020 }, { "epoch": 1.4095348452243006, "grad_norm": 4.247402191162109, "learning_rate": 6.3813597475442545e-06, "loss": 0.4954, "step": 17030 }, { "epoch": 1.4103625227611323, "grad_norm": 4.352746963500977, "learning_rate": 6.376731229453187e-06, "loss": 0.5813, "step": 17040 }, { "epoch": 1.411190200297964, "grad_norm": 4.427673816680908, "learning_rate": 6.372101434469622e-06, "loss": 0.5432, "step": 17050 }, { "epoch": 1.4120178778347956, "grad_norm": 3.541292428970337, "learning_rate": 6.367470366887606e-06, "loss": 0.4768, "step": 17060 }, { "epoch": 1.4128455553716273, "grad_norm": 3.8863332271575928, "learning_rate": 6.3628380310023716e-06, "loss": 0.569, "step": 17070 }, { "epoch": 1.413673232908459, "grad_norm": 3.7950732707977295, "learning_rate": 6.358204431110321e-06, "loss": 0.4615, "step": 17080 }, { "epoch": 1.4145009104452906, "grad_norm": 5.041335582733154, "learning_rate": 6.3535695715090315e-06, "loss": 0.5179, "step": 17090 }, { "epoch": 1.4153285879821222, "grad_norm": 6.467047691345215, "learning_rate": 6.34893345649725e-06, "loss": 0.4592, "step": 17100 }, { "epoch": 1.416156265518954, "grad_norm": 3.9973227977752686, "learning_rate": 6.344296090374886e-06, "loss": 0.4609, "step": 17110 }, { "epoch": 1.4169839430557856, "grad_norm": 4.052887439727783, "learning_rate": 6.339657477443009e-06, "loss": 0.5612, "step": 17120 }, { "epoch": 1.4178116205926172, "grad_norm": 4.657537460327148, "learning_rate": 6.335017622003848e-06, "loss": 0.5344, "step": 17130 }, { "epoch": 1.418639298129449, "grad_norm": 4.766136169433594, "learning_rate": 6.330376528360781e-06, "loss": 0.4665, "step": 17140 }, { "epoch": 1.4194669756662806, "grad_norm": 4.718607425689697, "learning_rate": 6.325734200818335e-06, "loss": 0.4491, "step": 17150 }, { "epoch": 1.420294653203112, "grad_norm": 6.496137619018555, "learning_rate": 6.321090643682184e-06, "loss": 0.5184, "step": 17160 }, { "epoch": 1.4211223307399437, "grad_norm": 5.955876350402832, "learning_rate": 6.3164458612591375e-06, "loss": 0.4832, "step": 17170 }, { "epoch": 1.4219500082767753, "grad_norm": 3.7983646392822266, "learning_rate": 6.3117998578571475e-06, "loss": 0.5322, "step": 17180 }, { "epoch": 1.422777685813607, "grad_norm": 4.215494155883789, "learning_rate": 6.307152637785293e-06, "loss": 0.5103, "step": 17190 }, { "epoch": 1.4236053633504386, "grad_norm": 4.294918537139893, "learning_rate": 6.302504205353786e-06, "loss": 0.5547, "step": 17200 }, { "epoch": 1.4244330408872703, "grad_norm": 4.659933090209961, "learning_rate": 6.2978545648739595e-06, "loss": 0.4518, "step": 17210 }, { "epoch": 1.425260718424102, "grad_norm": 5.056770324707031, "learning_rate": 6.2932037206582675e-06, "loss": 0.4824, "step": 17220 }, { "epoch": 1.4260883959609336, "grad_norm": 2.9267754554748535, "learning_rate": 6.288551677020282e-06, "loss": 0.4507, "step": 17230 }, { "epoch": 1.4269160734977653, "grad_norm": 8.731608390808105, "learning_rate": 6.2838984382746874e-06, "loss": 0.4775, "step": 17240 }, { "epoch": 1.427743751034597, "grad_norm": 4.645627975463867, "learning_rate": 6.279244008737276e-06, "loss": 0.5006, "step": 17250 }, { "epoch": 1.4285714285714286, "grad_norm": 4.20763635635376, "learning_rate": 6.274588392724941e-06, "loss": 0.4647, "step": 17260 }, { "epoch": 1.4293991061082603, "grad_norm": 4.230691909790039, "learning_rate": 6.2699315945556815e-06, "loss": 0.5359, "step": 17270 }, { "epoch": 1.430226783645092, "grad_norm": 5.542954921722412, "learning_rate": 6.265273618548591e-06, "loss": 0.5339, "step": 17280 }, { "epoch": 1.4310544611819236, "grad_norm": 4.68393611907959, "learning_rate": 6.260614469023855e-06, "loss": 0.5865, "step": 17290 }, { "epoch": 1.4318821387187552, "grad_norm": 4.997852802276611, "learning_rate": 6.255954150302746e-06, "loss": 0.5532, "step": 17300 }, { "epoch": 1.4327098162555867, "grad_norm": 4.259016036987305, "learning_rate": 6.251292666707627e-06, "loss": 0.4985, "step": 17310 }, { "epoch": 1.4335374937924183, "grad_norm": 4.844446659088135, "learning_rate": 6.246630022561934e-06, "loss": 0.5381, "step": 17320 }, { "epoch": 1.43436517132925, "grad_norm": 6.028258323669434, "learning_rate": 6.241966222190179e-06, "loss": 0.5133, "step": 17330 }, { "epoch": 1.4351928488660817, "grad_norm": 4.7519049644470215, "learning_rate": 6.237301269917956e-06, "loss": 0.4726, "step": 17340 }, { "epoch": 1.4360205264029133, "grad_norm": 4.81894588470459, "learning_rate": 6.232635170071918e-06, "loss": 0.4822, "step": 17350 }, { "epoch": 1.436848203939745, "grad_norm": 5.024934768676758, "learning_rate": 6.2279679269797844e-06, "loss": 0.5094, "step": 17360 }, { "epoch": 1.4376758814765767, "grad_norm": 4.124948978424072, "learning_rate": 6.223299544970339e-06, "loss": 0.4706, "step": 17370 }, { "epoch": 1.4385035590134083, "grad_norm": 4.801942825317383, "learning_rate": 6.2186300283734155e-06, "loss": 0.4819, "step": 17380 }, { "epoch": 1.43933123655024, "grad_norm": 5.78679084777832, "learning_rate": 6.213959381519905e-06, "loss": 0.5842, "step": 17390 }, { "epoch": 1.4401589140870716, "grad_norm": 4.336836338043213, "learning_rate": 6.2092876087417475e-06, "loss": 0.5498, "step": 17400 }, { "epoch": 1.4409865916239033, "grad_norm": 4.4522318840026855, "learning_rate": 6.204614714371922e-06, "loss": 0.4678, "step": 17410 }, { "epoch": 1.441814269160735, "grad_norm": 3.6464383602142334, "learning_rate": 6.199940702744449e-06, "loss": 0.5271, "step": 17420 }, { "epoch": 1.4426419466975666, "grad_norm": 6.036866188049316, "learning_rate": 6.195265578194392e-06, "loss": 0.5086, "step": 17430 }, { "epoch": 1.4434696242343983, "grad_norm": 8.413274765014648, "learning_rate": 6.1905893450578365e-06, "loss": 0.5778, "step": 17440 }, { "epoch": 1.44429730177123, "grad_norm": 4.075913906097412, "learning_rate": 6.185912007671906e-06, "loss": 0.5775, "step": 17450 }, { "epoch": 1.4451249793080616, "grad_norm": 5.440705299377441, "learning_rate": 6.181233570374742e-06, "loss": 0.5394, "step": 17460 }, { "epoch": 1.4459526568448933, "grad_norm": 5.1512250900268555, "learning_rate": 6.176554037505507e-06, "loss": 0.5264, "step": 17470 }, { "epoch": 1.446780334381725, "grad_norm": 3.7476906776428223, "learning_rate": 6.171873413404378e-06, "loss": 0.4452, "step": 17480 }, { "epoch": 1.4476080119185566, "grad_norm": 4.941946029663086, "learning_rate": 6.1671917024125506e-06, "loss": 0.4879, "step": 17490 }, { "epoch": 1.4484356894553883, "grad_norm": 4.190521717071533, "learning_rate": 6.162508908872223e-06, "loss": 0.5546, "step": 17500 }, { "epoch": 1.44926336699222, "grad_norm": 5.7772674560546875, "learning_rate": 6.157825037126598e-06, "loss": 0.4987, "step": 17510 }, { "epoch": 1.4500910445290516, "grad_norm": 5.5730438232421875, "learning_rate": 6.15314009151988e-06, "loss": 0.5323, "step": 17520 }, { "epoch": 1.4509187220658832, "grad_norm": 4.364959239959717, "learning_rate": 6.148454076397268e-06, "loss": 0.5032, "step": 17530 }, { "epoch": 1.451746399602715, "grad_norm": 3.3635926246643066, "learning_rate": 6.143766996104953e-06, "loss": 0.4987, "step": 17540 }, { "epoch": 1.4525740771395466, "grad_norm": 3.824867010116577, "learning_rate": 6.1390788549901146e-06, "loss": 0.5412, "step": 17550 }, { "epoch": 1.4534017546763782, "grad_norm": 3.7288737297058105, "learning_rate": 6.134389657400918e-06, "loss": 0.5265, "step": 17560 }, { "epoch": 1.4542294322132097, "grad_norm": 4.487273693084717, "learning_rate": 6.1296994076865024e-06, "loss": 0.5358, "step": 17570 }, { "epoch": 1.4550571097500413, "grad_norm": 3.086376667022705, "learning_rate": 6.125008110196992e-06, "loss": 0.4808, "step": 17580 }, { "epoch": 1.455884787286873, "grad_norm": 5.91103982925415, "learning_rate": 6.1203157692834736e-06, "loss": 0.5117, "step": 17590 }, { "epoch": 1.4567124648237046, "grad_norm": 4.592260837554932, "learning_rate": 6.115622389298006e-06, "loss": 0.4679, "step": 17600 }, { "epoch": 1.4575401423605363, "grad_norm": 4.705517292022705, "learning_rate": 6.110927974593614e-06, "loss": 0.5333, "step": 17610 }, { "epoch": 1.458367819897368, "grad_norm": 5.857121467590332, "learning_rate": 6.106232529524277e-06, "loss": 0.5463, "step": 17620 }, { "epoch": 1.4591954974341996, "grad_norm": 4.909053325653076, "learning_rate": 6.101536058444934e-06, "loss": 0.5083, "step": 17630 }, { "epoch": 1.4600231749710313, "grad_norm": 4.526946067810059, "learning_rate": 6.096838565711474e-06, "loss": 0.5217, "step": 17640 }, { "epoch": 1.460850852507863, "grad_norm": 4.766422748565674, "learning_rate": 6.092140055680732e-06, "loss": 0.5674, "step": 17650 }, { "epoch": 1.4616785300446946, "grad_norm": 4.698999404907227, "learning_rate": 6.08744053271049e-06, "loss": 0.4661, "step": 17660 }, { "epoch": 1.4625062075815263, "grad_norm": 2.8881328105926514, "learning_rate": 6.082740001159469e-06, "loss": 0.4171, "step": 17670 }, { "epoch": 1.463333885118358, "grad_norm": 4.529531002044678, "learning_rate": 6.078038465387324e-06, "loss": 0.5421, "step": 17680 }, { "epoch": 1.4641615626551896, "grad_norm": 3.5579888820648193, "learning_rate": 6.073335929754637e-06, "loss": 0.5281, "step": 17690 }, { "epoch": 1.4649892401920213, "grad_norm": 4.325390338897705, "learning_rate": 6.068632398622928e-06, "loss": 0.5804, "step": 17700 }, { "epoch": 1.465816917728853, "grad_norm": 4.548850059509277, "learning_rate": 6.063927876354629e-06, "loss": 0.4993, "step": 17710 }, { "epoch": 1.4666445952656844, "grad_norm": 4.759029865264893, "learning_rate": 6.059222367313101e-06, "loss": 0.5555, "step": 17720 }, { "epoch": 1.467472272802516, "grad_norm": 4.661491870880127, "learning_rate": 6.0545158758626144e-06, "loss": 0.5323, "step": 17730 }, { "epoch": 1.4682999503393477, "grad_norm": 6.387763977050781, "learning_rate": 6.0498084063683495e-06, "loss": 0.5258, "step": 17740 }, { "epoch": 1.4691276278761793, "grad_norm": 4.969717025756836, "learning_rate": 6.045099963196401e-06, "loss": 0.4784, "step": 17750 }, { "epoch": 1.469955305413011, "grad_norm": 4.234992027282715, "learning_rate": 6.040390550713757e-06, "loss": 0.5081, "step": 17760 }, { "epoch": 1.4707829829498427, "grad_norm": 3.8605053424835205, "learning_rate": 6.035680173288312e-06, "loss": 0.5141, "step": 17770 }, { "epoch": 1.4716106604866743, "grad_norm": 5.532476425170898, "learning_rate": 6.030968835288854e-06, "loss": 0.5397, "step": 17780 }, { "epoch": 1.472438338023506, "grad_norm": 3.4699814319610596, "learning_rate": 6.026256541085061e-06, "loss": 0.5429, "step": 17790 }, { "epoch": 1.4732660155603376, "grad_norm": 3.8035755157470703, "learning_rate": 6.021543295047495e-06, "loss": 0.525, "step": 17800 }, { "epoch": 1.4740936930971693, "grad_norm": 4.27392053604126, "learning_rate": 6.016829101547607e-06, "loss": 0.5574, "step": 17810 }, { "epoch": 1.474921370634001, "grad_norm": 6.089301109313965, "learning_rate": 6.0121139649577205e-06, "loss": 0.4679, "step": 17820 }, { "epoch": 1.4757490481708326, "grad_norm": 3.276428699493408, "learning_rate": 6.0073978896510386e-06, "loss": 0.4351, "step": 17830 }, { "epoch": 1.4765767257076643, "grad_norm": 4.008108139038086, "learning_rate": 6.0026808800016325e-06, "loss": 0.5158, "step": 17840 }, { "epoch": 1.477404403244496, "grad_norm": 4.05471658706665, "learning_rate": 5.997962940384442e-06, "loss": 0.5251, "step": 17850 }, { "epoch": 1.4782320807813276, "grad_norm": 4.248988151550293, "learning_rate": 5.993244075175264e-06, "loss": 0.4671, "step": 17860 }, { "epoch": 1.4790597583181593, "grad_norm": 4.340898513793945, "learning_rate": 5.988524288750762e-06, "loss": 0.4895, "step": 17870 }, { "epoch": 1.479887435854991, "grad_norm": 12.698265075683594, "learning_rate": 5.983803585488447e-06, "loss": 0.5506, "step": 17880 }, { "epoch": 1.4807151133918226, "grad_norm": 3.1507463455200195, "learning_rate": 5.9790819697666844e-06, "loss": 0.4478, "step": 17890 }, { "epoch": 1.4815427909286543, "grad_norm": 3.3283352851867676, "learning_rate": 5.974359445964685e-06, "loss": 0.5257, "step": 17900 }, { "epoch": 1.482370468465486, "grad_norm": 5.702081680297852, "learning_rate": 5.9696360184625016e-06, "loss": 0.5634, "step": 17910 }, { "epoch": 1.4831981460023176, "grad_norm": 4.346798419952393, "learning_rate": 5.964911691641022e-06, "loss": 0.5113, "step": 17920 }, { "epoch": 1.4840258235391492, "grad_norm": 8.046238899230957, "learning_rate": 5.960186469881973e-06, "loss": 0.5693, "step": 17930 }, { "epoch": 1.484853501075981, "grad_norm": 4.905395984649658, "learning_rate": 5.955460357567912e-06, "loss": 0.57, "step": 17940 }, { "epoch": 1.4856811786128126, "grad_norm": 4.503864288330078, "learning_rate": 5.950733359082215e-06, "loss": 0.5231, "step": 17950 }, { "epoch": 1.4865088561496442, "grad_norm": 4.176373481750488, "learning_rate": 5.946005478809089e-06, "loss": 0.489, "step": 17960 }, { "epoch": 1.4873365336864757, "grad_norm": 4.4933013916015625, "learning_rate": 5.941276721133553e-06, "loss": 0.4438, "step": 17970 }, { "epoch": 1.4881642112233073, "grad_norm": 4.576328277587891, "learning_rate": 5.9365470904414416e-06, "loss": 0.5366, "step": 17980 }, { "epoch": 1.488991888760139, "grad_norm": 2.6693735122680664, "learning_rate": 5.931816591119397e-06, "loss": 0.4321, "step": 17990 }, { "epoch": 1.4898195662969707, "grad_norm": 3.380887508392334, "learning_rate": 5.927085227554873e-06, "loss": 0.4675, "step": 18000 }, { "epoch": 1.4906472438338023, "grad_norm": 3.0014922618865967, "learning_rate": 5.922353004136117e-06, "loss": 0.5213, "step": 18010 }, { "epoch": 1.491474921370634, "grad_norm": 3.885087013244629, "learning_rate": 5.91761992525218e-06, "loss": 0.4693, "step": 18020 }, { "epoch": 1.4923025989074656, "grad_norm": 5.178191661834717, "learning_rate": 5.9128859952929056e-06, "loss": 0.5324, "step": 18030 }, { "epoch": 1.4931302764442973, "grad_norm": 7.748894691467285, "learning_rate": 5.908151218648921e-06, "loss": 0.486, "step": 18040 }, { "epoch": 1.493957953981129, "grad_norm": 4.475771427154541, "learning_rate": 5.9034155997116445e-06, "loss": 0.4325, "step": 18050 }, { "epoch": 1.4947856315179606, "grad_norm": 3.9688868522644043, "learning_rate": 5.898679142873278e-06, "loss": 0.474, "step": 18060 }, { "epoch": 1.4956133090547923, "grad_norm": 5.14984130859375, "learning_rate": 5.893941852526792e-06, "loss": 0.5588, "step": 18070 }, { "epoch": 1.496440986591624, "grad_norm": 5.150649547576904, "learning_rate": 5.889203733065939e-06, "loss": 0.4793, "step": 18080 }, { "epoch": 1.4972686641284556, "grad_norm": 6.694243431091309, "learning_rate": 5.884464788885232e-06, "loss": 0.5586, "step": 18090 }, { "epoch": 1.4980963416652873, "grad_norm": 4.590006351470947, "learning_rate": 5.879725024379956e-06, "loss": 0.5039, "step": 18100 }, { "epoch": 1.498924019202119, "grad_norm": 4.316522598266602, "learning_rate": 5.8749844439461545e-06, "loss": 0.4992, "step": 18110 }, { "epoch": 1.4997516967389504, "grad_norm": 3.874584436416626, "learning_rate": 5.870243051980627e-06, "loss": 0.5371, "step": 18120 }, { "epoch": 1.500579374275782, "grad_norm": 4.514276504516602, "learning_rate": 5.865500852880922e-06, "loss": 0.5203, "step": 18130 }, { "epoch": 1.5014070518126137, "grad_norm": 5.030210018157959, "learning_rate": 5.8607578510453454e-06, "loss": 0.4976, "step": 18140 }, { "epoch": 1.5022347293494454, "grad_norm": 4.673699378967285, "learning_rate": 5.856014050872941e-06, "loss": 0.5423, "step": 18150 }, { "epoch": 1.503062406886277, "grad_norm": 4.66920280456543, "learning_rate": 5.851269456763494e-06, "loss": 0.522, "step": 18160 }, { "epoch": 1.5038900844231087, "grad_norm": 4.6721086502075195, "learning_rate": 5.84652407311753e-06, "loss": 0.511, "step": 18170 }, { "epoch": 1.5047177619599403, "grad_norm": 3.2627997398376465, "learning_rate": 5.8417779043363e-06, "loss": 0.4758, "step": 18180 }, { "epoch": 1.505545439496772, "grad_norm": 4.797391891479492, "learning_rate": 5.8370309548217884e-06, "loss": 0.5271, "step": 18190 }, { "epoch": 1.5063731170336037, "grad_norm": 5.2328691482543945, "learning_rate": 5.8322832289767024e-06, "loss": 0.4993, "step": 18200 }, { "epoch": 1.5072007945704353, "grad_norm": 4.11328649520874, "learning_rate": 5.827534731204468e-06, "loss": 0.5021, "step": 18210 }, { "epoch": 1.508028472107267, "grad_norm": 5.236258029937744, "learning_rate": 5.8227854659092285e-06, "loss": 0.4893, "step": 18220 }, { "epoch": 1.5088561496440986, "grad_norm": 3.8987386226654053, "learning_rate": 5.818035437495842e-06, "loss": 0.507, "step": 18230 }, { "epoch": 1.5096838271809303, "grad_norm": 6.12503719329834, "learning_rate": 5.813284650369867e-06, "loss": 0.4546, "step": 18240 }, { "epoch": 1.510511504717762, "grad_norm": 4.594759464263916, "learning_rate": 5.8085331089375695e-06, "loss": 0.5415, "step": 18250 }, { "epoch": 1.5113391822545936, "grad_norm": 4.762172222137451, "learning_rate": 5.803780817605919e-06, "loss": 0.4517, "step": 18260 }, { "epoch": 1.5121668597914253, "grad_norm": 5.421925067901611, "learning_rate": 5.799027780782573e-06, "loss": 0.5641, "step": 18270 }, { "epoch": 1.512994537328257, "grad_norm": 5.069186210632324, "learning_rate": 5.794274002875888e-06, "loss": 0.5907, "step": 18280 }, { "epoch": 1.5138222148650886, "grad_norm": 5.921437740325928, "learning_rate": 5.789519488294903e-06, "loss": 0.4878, "step": 18290 }, { "epoch": 1.5146498924019203, "grad_norm": 4.12938117980957, "learning_rate": 5.784764241449341e-06, "loss": 0.5162, "step": 18300 }, { "epoch": 1.515477569938752, "grad_norm": 4.413852691650391, "learning_rate": 5.780008266749603e-06, "loss": 0.5262, "step": 18310 }, { "epoch": 1.5163052474755836, "grad_norm": 4.209931373596191, "learning_rate": 5.775251568606769e-06, "loss": 0.5054, "step": 18320 }, { "epoch": 1.5171329250124153, "grad_norm": 5.419059753417969, "learning_rate": 5.770494151432587e-06, "loss": 0.4724, "step": 18330 }, { "epoch": 1.517960602549247, "grad_norm": 5.0951337814331055, "learning_rate": 5.765736019639473e-06, "loss": 0.4855, "step": 18340 }, { "epoch": 1.5187882800860786, "grad_norm": 4.459385395050049, "learning_rate": 5.760977177640505e-06, "loss": 0.5629, "step": 18350 }, { "epoch": 1.5196159576229102, "grad_norm": 5.837329864501953, "learning_rate": 5.75621762984942e-06, "loss": 0.3771, "step": 18360 }, { "epoch": 1.520443635159742, "grad_norm": 4.591871738433838, "learning_rate": 5.751457380680607e-06, "loss": 0.483, "step": 18370 }, { "epoch": 1.5212713126965736, "grad_norm": 5.599931240081787, "learning_rate": 5.746696434549115e-06, "loss": 0.4579, "step": 18380 }, { "epoch": 1.5220989902334052, "grad_norm": 4.124990463256836, "learning_rate": 5.741934795870626e-06, "loss": 0.5399, "step": 18390 }, { "epoch": 1.5229266677702367, "grad_norm": 4.608062267303467, "learning_rate": 5.737172469061474e-06, "loss": 0.5147, "step": 18400 }, { "epoch": 1.5237543453070683, "grad_norm": 5.422941207885742, "learning_rate": 5.732409458538628e-06, "loss": 0.5706, "step": 18410 }, { "epoch": 1.5245820228439, "grad_norm": 5.0350775718688965, "learning_rate": 5.727645768719689e-06, "loss": 0.4857, "step": 18420 }, { "epoch": 1.5254097003807316, "grad_norm": 4.596884250640869, "learning_rate": 5.722881404022893e-06, "loss": 0.4999, "step": 18430 }, { "epoch": 1.5262373779175633, "grad_norm": 4.010284900665283, "learning_rate": 5.7181163688671e-06, "loss": 0.5654, "step": 18440 }, { "epoch": 1.527065055454395, "grad_norm": 7.092634201049805, "learning_rate": 5.713350667671788e-06, "loss": 0.5078, "step": 18450 }, { "epoch": 1.5278927329912266, "grad_norm": 5.259923458099365, "learning_rate": 5.708584304857058e-06, "loss": 0.516, "step": 18460 }, { "epoch": 1.5287204105280583, "grad_norm": 3.8542299270629883, "learning_rate": 5.703817284843621e-06, "loss": 0.5682, "step": 18470 }, { "epoch": 1.52954808806489, "grad_norm": 4.138514518737793, "learning_rate": 5.6990496120528e-06, "loss": 0.484, "step": 18480 }, { "epoch": 1.5303757656017216, "grad_norm": 4.454927444458008, "learning_rate": 5.694281290906521e-06, "loss": 0.5347, "step": 18490 }, { "epoch": 1.531203443138553, "grad_norm": 5.165355682373047, "learning_rate": 5.689512325827317e-06, "loss": 0.4868, "step": 18500 }, { "epoch": 1.5320311206753847, "grad_norm": 4.237500190734863, "learning_rate": 5.684742721238308e-06, "loss": 0.5075, "step": 18510 }, { "epoch": 1.5328587982122164, "grad_norm": 4.558645248413086, "learning_rate": 5.679972481563219e-06, "loss": 0.5931, "step": 18520 }, { "epoch": 1.533686475749048, "grad_norm": 2.5049455165863037, "learning_rate": 5.675201611226353e-06, "loss": 0.4723, "step": 18530 }, { "epoch": 1.5345141532858797, "grad_norm": 4.695624828338623, "learning_rate": 5.6704301146526056e-06, "loss": 0.4953, "step": 18540 }, { "epoch": 1.5353418308227114, "grad_norm": 4.36061954498291, "learning_rate": 5.665657996267452e-06, "loss": 0.4857, "step": 18550 }, { "epoch": 1.536169508359543, "grad_norm": 6.563568115234375, "learning_rate": 5.660885260496941e-06, "loss": 0.5157, "step": 18560 }, { "epoch": 1.5369971858963747, "grad_norm": 5.811282634735107, "learning_rate": 5.656111911767697e-06, "loss": 0.538, "step": 18570 }, { "epoch": 1.5378248634332063, "grad_norm": 4.2170209884643555, "learning_rate": 5.65133795450691e-06, "loss": 0.5607, "step": 18580 }, { "epoch": 1.538652540970038, "grad_norm": 4.05963659286499, "learning_rate": 5.646563393142338e-06, "loss": 0.4822, "step": 18590 }, { "epoch": 1.5394802185068697, "grad_norm": 4.3824639320373535, "learning_rate": 5.641788232102297e-06, "loss": 0.505, "step": 18600 }, { "epoch": 1.5403078960437013, "grad_norm": 2.6280925273895264, "learning_rate": 5.637012475815659e-06, "loss": 0.4469, "step": 18610 }, { "epoch": 1.541135573580533, "grad_norm": 5.623599529266357, "learning_rate": 5.63223612871185e-06, "loss": 0.5839, "step": 18620 }, { "epoch": 1.5419632511173647, "grad_norm": 4.406469821929932, "learning_rate": 5.627459195220841e-06, "loss": 0.5761, "step": 18630 }, { "epoch": 1.5427909286541963, "grad_norm": 4.002017974853516, "learning_rate": 5.62268167977315e-06, "loss": 0.4592, "step": 18640 }, { "epoch": 1.543618606191028, "grad_norm": 3.679898262023926, "learning_rate": 5.617903586799832e-06, "loss": 0.4579, "step": 18650 }, { "epoch": 1.5444462837278596, "grad_norm": 5.135529518127441, "learning_rate": 5.613124920732481e-06, "loss": 0.5029, "step": 18660 }, { "epoch": 1.5452739612646913, "grad_norm": 5.11453914642334, "learning_rate": 5.60834568600322e-06, "loss": 0.544, "step": 18670 }, { "epoch": 1.546101638801523, "grad_norm": 4.690515518188477, "learning_rate": 5.603565887044698e-06, "loss": 0.4762, "step": 18680 }, { "epoch": 1.5469293163383546, "grad_norm": 3.7463669776916504, "learning_rate": 5.598785528290091e-06, "loss": 0.4428, "step": 18690 }, { "epoch": 1.5477569938751863, "grad_norm": 4.460480213165283, "learning_rate": 5.5940046141730905e-06, "loss": 0.5174, "step": 18700 }, { "epoch": 1.548584671412018, "grad_norm": 5.59206485748291, "learning_rate": 5.589223149127908e-06, "loss": 0.5353, "step": 18710 }, { "epoch": 1.5494123489488496, "grad_norm": 6.926443099975586, "learning_rate": 5.584441137589261e-06, "loss": 0.4927, "step": 18720 }, { "epoch": 1.5502400264856813, "grad_norm": 5.115754127502441, "learning_rate": 5.579658583992378e-06, "loss": 0.4558, "step": 18730 }, { "epoch": 1.551067704022513, "grad_norm": 4.523055076599121, "learning_rate": 5.5748754927729856e-06, "loss": 0.5176, "step": 18740 }, { "epoch": 1.5518953815593446, "grad_norm": 4.360938549041748, "learning_rate": 5.570091868367311e-06, "loss": 0.4703, "step": 18750 }, { "epoch": 1.5527230590961763, "grad_norm": 4.573211193084717, "learning_rate": 5.5653077152120784e-06, "loss": 0.5626, "step": 18760 }, { "epoch": 1.553550736633008, "grad_norm": 3.7773189544677734, "learning_rate": 5.5605230377445004e-06, "loss": 0.5525, "step": 18770 }, { "epoch": 1.5543784141698396, "grad_norm": 3.0475616455078125, "learning_rate": 5.5557378404022755e-06, "loss": 0.4879, "step": 18780 }, { "epoch": 1.5552060917066712, "grad_norm": 4.2254838943481445, "learning_rate": 5.550952127623587e-06, "loss": 0.4884, "step": 18790 }, { "epoch": 1.556033769243503, "grad_norm": 3.603300094604492, "learning_rate": 5.5461659038470906e-06, "loss": 0.5198, "step": 18800 }, { "epoch": 1.5568614467803343, "grad_norm": 3.8294901847839355, "learning_rate": 5.5413791735119185e-06, "loss": 0.4686, "step": 18810 }, { "epoch": 1.557689124317166, "grad_norm": 4.507630825042725, "learning_rate": 5.536591941057679e-06, "loss": 0.5024, "step": 18820 }, { "epoch": 1.5585168018539977, "grad_norm": 5.7595720291137695, "learning_rate": 5.531804210924437e-06, "loss": 0.5136, "step": 18830 }, { "epoch": 1.5593444793908293, "grad_norm": 4.055469512939453, "learning_rate": 5.527015987552723e-06, "loss": 0.4799, "step": 18840 }, { "epoch": 1.560172156927661, "grad_norm": 5.389695167541504, "learning_rate": 5.522227275383527e-06, "loss": 0.5484, "step": 18850 }, { "epoch": 1.5609998344644926, "grad_norm": 5.9087233543396, "learning_rate": 5.517438078858287e-06, "loss": 0.5671, "step": 18860 }, { "epoch": 1.5618275120013243, "grad_norm": 4.049314498901367, "learning_rate": 5.512648402418895e-06, "loss": 0.4467, "step": 18870 }, { "epoch": 1.562655189538156, "grad_norm": 4.185881614685059, "learning_rate": 5.507858250507686e-06, "loss": 0.4679, "step": 18880 }, { "epoch": 1.5634828670749876, "grad_norm": 5.9443039894104, "learning_rate": 5.503067627567438e-06, "loss": 0.4682, "step": 18890 }, { "epoch": 1.5643105446118193, "grad_norm": 4.763498783111572, "learning_rate": 5.49827653804136e-06, "loss": 0.5364, "step": 18900 }, { "epoch": 1.5651382221486507, "grad_norm": 3.6701152324676514, "learning_rate": 5.493484986373103e-06, "loss": 0.4477, "step": 18910 }, { "epoch": 1.5659658996854824, "grad_norm": 4.159217357635498, "learning_rate": 5.488692977006739e-06, "loss": 0.4954, "step": 18920 }, { "epoch": 1.566793577222314, "grad_norm": 3.6371073722839355, "learning_rate": 5.483900514386767e-06, "loss": 0.5196, "step": 18930 }, { "epoch": 1.5676212547591457, "grad_norm": 3.5883243083953857, "learning_rate": 5.4791076029581095e-06, "loss": 0.5136, "step": 18940 }, { "epoch": 1.5684489322959774, "grad_norm": 4.045977592468262, "learning_rate": 5.474314247166099e-06, "loss": 0.5052, "step": 18950 }, { "epoch": 1.569276609832809, "grad_norm": 3.6960721015930176, "learning_rate": 5.4695204514564835e-06, "loss": 0.475, "step": 18960 }, { "epoch": 1.5701042873696407, "grad_norm": 5.477766513824463, "learning_rate": 5.464726220275421e-06, "loss": 0.5413, "step": 18970 }, { "epoch": 1.5709319649064724, "grad_norm": 4.930051326751709, "learning_rate": 5.45993155806947e-06, "loss": 0.5218, "step": 18980 }, { "epoch": 1.571759642443304, "grad_norm": 5.729191303253174, "learning_rate": 5.45513646928559e-06, "loss": 0.5035, "step": 18990 }, { "epoch": 1.5725873199801357, "grad_norm": 2.7616562843322754, "learning_rate": 5.450340958371138e-06, "loss": 0.5015, "step": 19000 }, { "epoch": 1.5734149975169673, "grad_norm": 6.882381916046143, "learning_rate": 5.44554502977386e-06, "loss": 0.4373, "step": 19010 }, { "epoch": 1.574242675053799, "grad_norm": 4.209589004516602, "learning_rate": 5.440748687941888e-06, "loss": 0.5558, "step": 19020 }, { "epoch": 1.5750703525906307, "grad_norm": 5.148728370666504, "learning_rate": 5.4359519373237425e-06, "loss": 0.4908, "step": 19030 }, { "epoch": 1.5758980301274623, "grad_norm": 4.6136088371276855, "learning_rate": 5.431154782368319e-06, "loss": 0.4702, "step": 19040 }, { "epoch": 1.576725707664294, "grad_norm": 4.257224082946777, "learning_rate": 5.426357227524888e-06, "loss": 0.4722, "step": 19050 }, { "epoch": 1.5775533852011256, "grad_norm": 5.0578107833862305, "learning_rate": 5.421559277243094e-06, "loss": 0.4805, "step": 19060 }, { "epoch": 1.5783810627379573, "grad_norm": 3.3327159881591797, "learning_rate": 5.416760935972944e-06, "loss": 0.5297, "step": 19070 }, { "epoch": 1.579208740274789, "grad_norm": 3.1915793418884277, "learning_rate": 5.411962208164812e-06, "loss": 0.5145, "step": 19080 }, { "epoch": 1.5800364178116206, "grad_norm": 4.2144246101379395, "learning_rate": 5.407163098269425e-06, "loss": 0.5005, "step": 19090 }, { "epoch": 1.5808640953484523, "grad_norm": 4.031592845916748, "learning_rate": 5.402363610737873e-06, "loss": 0.4092, "step": 19100 }, { "epoch": 1.581691772885284, "grad_norm": 4.0822529792785645, "learning_rate": 5.397563750021588e-06, "loss": 0.4997, "step": 19110 }, { "epoch": 1.5825194504221156, "grad_norm": 2.5493385791778564, "learning_rate": 5.392763520572351e-06, "loss": 0.4447, "step": 19120 }, { "epoch": 1.5833471279589473, "grad_norm": 4.543835639953613, "learning_rate": 5.387962926842285e-06, "loss": 0.5412, "step": 19130 }, { "epoch": 1.584174805495779, "grad_norm": 6.600100040435791, "learning_rate": 5.383161973283852e-06, "loss": 0.447, "step": 19140 }, { "epoch": 1.5850024830326106, "grad_norm": 3.798158884048462, "learning_rate": 5.378360664349848e-06, "loss": 0.5138, "step": 19150 }, { "epoch": 1.5858301605694423, "grad_norm": 4.134421348571777, "learning_rate": 5.373559004493393e-06, "loss": 0.4835, "step": 19160 }, { "epoch": 1.586657838106274, "grad_norm": 4.358469486236572, "learning_rate": 5.3687569981679415e-06, "loss": 0.4474, "step": 19170 }, { "epoch": 1.5874855156431056, "grad_norm": 4.398070812225342, "learning_rate": 5.363954649827262e-06, "loss": 0.5436, "step": 19180 }, { "epoch": 1.5883131931799372, "grad_norm": 4.892857074737549, "learning_rate": 5.359151963925443e-06, "loss": 0.5248, "step": 19190 }, { "epoch": 1.589140870716769, "grad_norm": 3.6319165229797363, "learning_rate": 5.354348944916886e-06, "loss": 0.4731, "step": 19200 }, { "epoch": 1.5899685482536003, "grad_norm": 5.292303085327148, "learning_rate": 5.349545597256304e-06, "loss": 0.4946, "step": 19210 }, { "epoch": 1.590796225790432, "grad_norm": 4.3955230712890625, "learning_rate": 5.344741925398709e-06, "loss": 0.5194, "step": 19220 }, { "epoch": 1.5916239033272637, "grad_norm": 5.852062702178955, "learning_rate": 5.3399379337994184e-06, "loss": 0.4583, "step": 19230 }, { "epoch": 1.5924515808640953, "grad_norm": 4.714190483093262, "learning_rate": 5.335133626914044e-06, "loss": 0.4111, "step": 19240 }, { "epoch": 1.593279258400927, "grad_norm": 4.809226989746094, "learning_rate": 5.33032900919849e-06, "loss": 0.5184, "step": 19250 }, { "epoch": 1.5941069359377587, "grad_norm": 4.574151515960693, "learning_rate": 5.325524085108951e-06, "loss": 0.4847, "step": 19260 }, { "epoch": 1.5949346134745903, "grad_norm": 3.373408079147339, "learning_rate": 5.320718859101906e-06, "loss": 0.4098, "step": 19270 }, { "epoch": 1.595762291011422, "grad_norm": 5.786484718322754, "learning_rate": 5.315913335634108e-06, "loss": 0.5505, "step": 19280 }, { "epoch": 1.5965899685482536, "grad_norm": 5.836145877838135, "learning_rate": 5.311107519162594e-06, "loss": 0.4844, "step": 19290 }, { "epoch": 1.5974176460850853, "grad_norm": 4.368595123291016, "learning_rate": 5.306301414144667e-06, "loss": 0.5066, "step": 19300 }, { "epoch": 1.5982453236219167, "grad_norm": 4.696157455444336, "learning_rate": 5.3014950250379e-06, "loss": 0.5095, "step": 19310 }, { "epoch": 1.5990730011587484, "grad_norm": 4.586207389831543, "learning_rate": 5.296688356300129e-06, "loss": 0.4804, "step": 19320 }, { "epoch": 1.59990067869558, "grad_norm": 4.671030044555664, "learning_rate": 5.291881412389451e-06, "loss": 0.475, "step": 19330 }, { "epoch": 1.6007283562324117, "grad_norm": 4.667694568634033, "learning_rate": 5.2870741977642135e-06, "loss": 0.4805, "step": 19340 }, { "epoch": 1.6015560337692434, "grad_norm": 4.659377098083496, "learning_rate": 5.282266716883022e-06, "loss": 0.5311, "step": 19350 }, { "epoch": 1.602383711306075, "grad_norm": 4.055227279663086, "learning_rate": 5.277458974204722e-06, "loss": 0.5435, "step": 19360 }, { "epoch": 1.6032113888429067, "grad_norm": 4.676709175109863, "learning_rate": 5.272650974188407e-06, "loss": 0.4399, "step": 19370 }, { "epoch": 1.6040390663797384, "grad_norm": 5.16737174987793, "learning_rate": 5.267842721293407e-06, "loss": 0.512, "step": 19380 }, { "epoch": 1.60486674391657, "grad_norm": 5.186887264251709, "learning_rate": 5.263034219979286e-06, "loss": 0.5032, "step": 19390 }, { "epoch": 1.6056944214534017, "grad_norm": 5.030655860900879, "learning_rate": 5.258225474705839e-06, "loss": 0.4341, "step": 19400 }, { "epoch": 1.6065220989902333, "grad_norm": 4.4976654052734375, "learning_rate": 5.2534164899330885e-06, "loss": 0.5151, "step": 19410 }, { "epoch": 1.607349776527065, "grad_norm": 4.968053817749023, "learning_rate": 5.248607270121278e-06, "loss": 0.4766, "step": 19420 }, { "epoch": 1.6081774540638967, "grad_norm": 3.951568126678467, "learning_rate": 5.243797819730867e-06, "loss": 0.515, "step": 19430 }, { "epoch": 1.6090051316007283, "grad_norm": 4.584232807159424, "learning_rate": 5.238988143222534e-06, "loss": 0.4537, "step": 19440 }, { "epoch": 1.60983280913756, "grad_norm": 4.2197675704956055, "learning_rate": 5.234178245057163e-06, "loss": 0.4548, "step": 19450 }, { "epoch": 1.6106604866743917, "grad_norm": 5.970284461975098, "learning_rate": 5.229368129695842e-06, "loss": 0.567, "step": 19460 }, { "epoch": 1.6114881642112233, "grad_norm": 4.4099931716918945, "learning_rate": 5.224557801599867e-06, "loss": 0.4793, "step": 19470 }, { "epoch": 1.612315841748055, "grad_norm": 3.8999903202056885, "learning_rate": 5.219747265230725e-06, "loss": 0.4438, "step": 19480 }, { "epoch": 1.6131435192848866, "grad_norm": 4.66508150100708, "learning_rate": 5.2149365250500995e-06, "loss": 0.4774, "step": 19490 }, { "epoch": 1.6139711968217183, "grad_norm": 4.452279090881348, "learning_rate": 5.2101255855198625e-06, "loss": 0.4746, "step": 19500 }, { "epoch": 1.61479887435855, "grad_norm": 5.90384578704834, "learning_rate": 5.20531445110207e-06, "loss": 0.4789, "step": 19510 }, { "epoch": 1.6156265518953816, "grad_norm": 4.933854103088379, "learning_rate": 5.200503126258959e-06, "loss": 0.4084, "step": 19520 }, { "epoch": 1.6164542294322133, "grad_norm": 4.482888221740723, "learning_rate": 5.195691615452944e-06, "loss": 0.4995, "step": 19530 }, { "epoch": 1.617281906969045, "grad_norm": 3.440183639526367, "learning_rate": 5.190879923146612e-06, "loss": 0.443, "step": 19540 }, { "epoch": 1.6181095845058766, "grad_norm": 5.6853132247924805, "learning_rate": 5.186068053802716e-06, "loss": 0.5227, "step": 19550 }, { "epoch": 1.6189372620427083, "grad_norm": 4.528489112854004, "learning_rate": 5.181256011884176e-06, "loss": 0.5263, "step": 19560 }, { "epoch": 1.61976493957954, "grad_norm": 4.542843818664551, "learning_rate": 5.176443801854071e-06, "loss": 0.5387, "step": 19570 }, { "epoch": 1.6205926171163716, "grad_norm": 5.118425369262695, "learning_rate": 5.1716314281756355e-06, "loss": 0.4824, "step": 19580 }, { "epoch": 1.6214202946532033, "grad_norm": 5.857172966003418, "learning_rate": 5.166818895312258e-06, "loss": 0.4792, "step": 19590 }, { "epoch": 1.622247972190035, "grad_norm": 5.2976975440979, "learning_rate": 5.16200620772747e-06, "loss": 0.533, "step": 19600 }, { "epoch": 1.6230756497268666, "grad_norm": 3.7605583667755127, "learning_rate": 5.157193369884951e-06, "loss": 0.5266, "step": 19610 }, { "epoch": 1.623903327263698, "grad_norm": 5.4784955978393555, "learning_rate": 5.1523803862485176e-06, "loss": 0.4956, "step": 19620 }, { "epoch": 1.6247310048005297, "grad_norm": 4.448113441467285, "learning_rate": 5.147567261282122e-06, "loss": 0.5146, "step": 19630 }, { "epoch": 1.6255586823373613, "grad_norm": 4.036832332611084, "learning_rate": 5.142753999449849e-06, "loss": 0.4084, "step": 19640 }, { "epoch": 1.626386359874193, "grad_norm": 6.9718499183654785, "learning_rate": 5.137940605215907e-06, "loss": 0.482, "step": 19650 }, { "epoch": 1.6272140374110247, "grad_norm": 4.184277534484863, "learning_rate": 5.133127083044634e-06, "loss": 0.5231, "step": 19660 }, { "epoch": 1.6280417149478563, "grad_norm": 5.068964958190918, "learning_rate": 5.128313437400474e-06, "loss": 0.4815, "step": 19670 }, { "epoch": 1.628869392484688, "grad_norm": 4.297544956207275, "learning_rate": 5.123499672748e-06, "loss": 0.507, "step": 19680 }, { "epoch": 1.6296970700215196, "grad_norm": 4.507935047149658, "learning_rate": 5.118685793551885e-06, "loss": 0.5314, "step": 19690 }, { "epoch": 1.6305247475583513, "grad_norm": 4.970942974090576, "learning_rate": 5.113871804276915e-06, "loss": 0.4821, "step": 19700 }, { "epoch": 1.631352425095183, "grad_norm": 4.167306423187256, "learning_rate": 5.109057709387973e-06, "loss": 0.5014, "step": 19710 }, { "epoch": 1.6321801026320144, "grad_norm": 4.733585357666016, "learning_rate": 5.104243513350043e-06, "loss": 0.5478, "step": 19720 }, { "epoch": 1.633007780168846, "grad_norm": 3.593817949295044, "learning_rate": 5.0994292206282015e-06, "loss": 0.5206, "step": 19730 }, { "epoch": 1.6338354577056777, "grad_norm": 3.9072697162628174, "learning_rate": 5.094614835687615e-06, "loss": 0.4439, "step": 19740 }, { "epoch": 1.6346631352425094, "grad_norm": 5.277987957000732, "learning_rate": 5.089800362993537e-06, "loss": 0.4223, "step": 19750 }, { "epoch": 1.635490812779341, "grad_norm": 4.682607650756836, "learning_rate": 5.084985807011298e-06, "loss": 0.5084, "step": 19760 }, { "epoch": 1.6363184903161727, "grad_norm": 6.800065517425537, "learning_rate": 5.080171172206313e-06, "loss": 0.4349, "step": 19770 }, { "epoch": 1.6371461678530044, "grad_norm": 5.253383636474609, "learning_rate": 5.075356463044061e-06, "loss": 0.5151, "step": 19780 }, { "epoch": 1.637973845389836, "grad_norm": 3.608611822128296, "learning_rate": 5.070541683990099e-06, "loss": 0.4424, "step": 19790 }, { "epoch": 1.6388015229266677, "grad_norm": 5.425894260406494, "learning_rate": 5.065726839510041e-06, "loss": 0.52, "step": 19800 }, { "epoch": 1.6396292004634994, "grad_norm": 5.091259002685547, "learning_rate": 5.060911934069568e-06, "loss": 0.555, "step": 19810 }, { "epoch": 1.640456878000331, "grad_norm": 4.54715633392334, "learning_rate": 5.056096972134414e-06, "loss": 0.4328, "step": 19820 }, { "epoch": 1.6412845555371627, "grad_norm": 5.0009236335754395, "learning_rate": 5.051281958170367e-06, "loss": 0.4537, "step": 19830 }, { "epoch": 1.6421122330739943, "grad_norm": 5.125114440917969, "learning_rate": 5.046466896643261e-06, "loss": 0.5478, "step": 19840 }, { "epoch": 1.642939910610826, "grad_norm": 4.54011344909668, "learning_rate": 5.041651792018975e-06, "loss": 0.5173, "step": 19850 }, { "epoch": 1.6437675881476577, "grad_norm": 4.684621334075928, "learning_rate": 5.036836648763432e-06, "loss": 0.4261, "step": 19860 }, { "epoch": 1.6445952656844893, "grad_norm": 5.55106782913208, "learning_rate": 5.032021471342585e-06, "loss": 0.5311, "step": 19870 }, { "epoch": 1.645422943221321, "grad_norm": 4.60247278213501, "learning_rate": 5.027206264222422e-06, "loss": 0.4762, "step": 19880 }, { "epoch": 1.6462506207581526, "grad_norm": 4.9610819816589355, "learning_rate": 5.022391031868957e-06, "loss": 0.4514, "step": 19890 }, { "epoch": 1.6470782982949843, "grad_norm": 4.293190002441406, "learning_rate": 5.0175757787482275e-06, "loss": 0.4663, "step": 19900 }, { "epoch": 1.647905975831816, "grad_norm": 4.038928985595703, "learning_rate": 5.012760509326291e-06, "loss": 0.4721, "step": 19910 }, { "epoch": 1.6487336533686476, "grad_norm": 3.290175199508667, "learning_rate": 5.007945228069222e-06, "loss": 0.5189, "step": 19920 }, { "epoch": 1.6495613309054793, "grad_norm": 4.585897445678711, "learning_rate": 5.003129939443101e-06, "loss": 0.5227, "step": 19930 }, { "epoch": 1.650389008442311, "grad_norm": 5.171100616455078, "learning_rate": 4.998314647914022e-06, "loss": 0.5439, "step": 19940 }, { "epoch": 1.6512166859791426, "grad_norm": 5.160886764526367, "learning_rate": 4.993499357948072e-06, "loss": 0.5365, "step": 19950 }, { "epoch": 1.6520443635159743, "grad_norm": 9.4364013671875, "learning_rate": 4.988684074011348e-06, "loss": 0.4985, "step": 19960 }, { "epoch": 1.652872041052806, "grad_norm": 4.297188758850098, "learning_rate": 4.983868800569934e-06, "loss": 0.5096, "step": 19970 }, { "epoch": 1.6536997185896376, "grad_norm": 3.56721830368042, "learning_rate": 4.979053542089906e-06, "loss": 0.4955, "step": 19980 }, { "epoch": 1.6545273961264693, "grad_norm": 5.070008754730225, "learning_rate": 4.974238303037327e-06, "loss": 0.4477, "step": 19990 }, { "epoch": 1.655355073663301, "grad_norm": 5.85329008102417, "learning_rate": 4.96942308787824e-06, "loss": 0.4788, "step": 20000 }, { "epoch": 1.6561827512001326, "grad_norm": 4.117453098297119, "learning_rate": 4.964607901078668e-06, "loss": 0.44, "step": 20010 }, { "epoch": 1.657010428736964, "grad_norm": 3.9311776161193848, "learning_rate": 4.959792747104606e-06, "loss": 0.5276, "step": 20020 }, { "epoch": 1.6578381062737957, "grad_norm": 4.029578685760498, "learning_rate": 4.95497763042202e-06, "loss": 0.5004, "step": 20030 }, { "epoch": 1.6586657838106273, "grad_norm": 5.447892189025879, "learning_rate": 4.950162555496843e-06, "loss": 0.5349, "step": 20040 }, { "epoch": 1.659493461347459, "grad_norm": 4.14055871963501, "learning_rate": 4.945347526794963e-06, "loss": 0.459, "step": 20050 }, { "epoch": 1.6603211388842907, "grad_norm": 5.953919887542725, "learning_rate": 4.94053254878223e-06, "loss": 0.496, "step": 20060 }, { "epoch": 1.6611488164211223, "grad_norm": 6.961580753326416, "learning_rate": 4.935717625924448e-06, "loss": 0.5097, "step": 20070 }, { "epoch": 1.661976493957954, "grad_norm": 4.182796478271484, "learning_rate": 4.930902762687367e-06, "loss": 0.5373, "step": 20080 }, { "epoch": 1.6628041714947857, "grad_norm": 4.258849143981934, "learning_rate": 4.926087963536684e-06, "loss": 0.518, "step": 20090 }, { "epoch": 1.6636318490316173, "grad_norm": 5.903066158294678, "learning_rate": 4.921273232938033e-06, "loss": 0.4826, "step": 20100 }, { "epoch": 1.664459526568449, "grad_norm": 3.8233680725097656, "learning_rate": 4.916458575356989e-06, "loss": 0.4983, "step": 20110 }, { "epoch": 1.6652872041052804, "grad_norm": 6.548916816711426, "learning_rate": 4.911643995259054e-06, "loss": 0.4708, "step": 20120 }, { "epoch": 1.666114881642112, "grad_norm": 5.125694274902344, "learning_rate": 4.906829497109663e-06, "loss": 0.4753, "step": 20130 }, { "epoch": 1.6669425591789437, "grad_norm": 3.869962215423584, "learning_rate": 4.9020150853741745e-06, "loss": 0.4993, "step": 20140 }, { "epoch": 1.6677702367157754, "grad_norm": 4.441439151763916, "learning_rate": 4.897200764517863e-06, "loss": 0.4989, "step": 20150 }, { "epoch": 1.668597914252607, "grad_norm": 6.38893461227417, "learning_rate": 4.892386539005924e-06, "loss": 0.5391, "step": 20160 }, { "epoch": 1.6694255917894387, "grad_norm": 5.463329315185547, "learning_rate": 4.8875724133034565e-06, "loss": 0.5492, "step": 20170 }, { "epoch": 1.6702532693262704, "grad_norm": 4.002647399902344, "learning_rate": 4.882758391875478e-06, "loss": 0.4856, "step": 20180 }, { "epoch": 1.671080946863102, "grad_norm": 5.03566837310791, "learning_rate": 4.877944479186898e-06, "loss": 0.4675, "step": 20190 }, { "epoch": 1.6719086243999337, "grad_norm": 5.070801258087158, "learning_rate": 4.873130679702537e-06, "loss": 0.5648, "step": 20200 }, { "epoch": 1.6727363019367654, "grad_norm": 5.488514423370361, "learning_rate": 4.8683169978870995e-06, "loss": 0.4979, "step": 20210 }, { "epoch": 1.673563979473597, "grad_norm": 6.543341159820557, "learning_rate": 4.863503438205187e-06, "loss": 0.5151, "step": 20220 }, { "epoch": 1.6743916570104287, "grad_norm": 3.910025119781494, "learning_rate": 4.858690005121285e-06, "loss": 0.5079, "step": 20230 }, { "epoch": 1.6752193345472604, "grad_norm": 6.105428695678711, "learning_rate": 4.853876703099763e-06, "loss": 0.4507, "step": 20240 }, { "epoch": 1.676047012084092, "grad_norm": 4.521209239959717, "learning_rate": 4.8490635366048685e-06, "loss": 0.5273, "step": 20250 }, { "epoch": 1.6768746896209237, "grad_norm": 3.2880806922912598, "learning_rate": 4.844250510100727e-06, "loss": 0.459, "step": 20260 }, { "epoch": 1.6777023671577553, "grad_norm": 4.235536098480225, "learning_rate": 4.839437628051327e-06, "loss": 0.5287, "step": 20270 }, { "epoch": 1.678530044694587, "grad_norm": 4.751125335693359, "learning_rate": 4.834624894920527e-06, "loss": 0.4852, "step": 20280 }, { "epoch": 1.6793577222314187, "grad_norm": 5.304824352264404, "learning_rate": 4.829812315172048e-06, "loss": 0.4274, "step": 20290 }, { "epoch": 1.6801853997682503, "grad_norm": 4.6887664794921875, "learning_rate": 4.824999893269466e-06, "loss": 0.499, "step": 20300 }, { "epoch": 1.681013077305082, "grad_norm": 4.063558101654053, "learning_rate": 4.8201876336762164e-06, "loss": 0.5365, "step": 20310 }, { "epoch": 1.6818407548419136, "grad_norm": 4.087381362915039, "learning_rate": 4.815375540855577e-06, "loss": 0.4816, "step": 20320 }, { "epoch": 1.6826684323787453, "grad_norm": 4.689122676849365, "learning_rate": 4.810563619270675e-06, "loss": 0.49, "step": 20330 }, { "epoch": 1.683496109915577, "grad_norm": 4.5147175788879395, "learning_rate": 4.805751873384476e-06, "loss": 0.5069, "step": 20340 }, { "epoch": 1.6843237874524086, "grad_norm": 4.037771224975586, "learning_rate": 4.800940307659787e-06, "loss": 0.4713, "step": 20350 }, { "epoch": 1.6851514649892403, "grad_norm": 3.1092946529388428, "learning_rate": 4.796128926559242e-06, "loss": 0.4643, "step": 20360 }, { "epoch": 1.685979142526072, "grad_norm": 3.9456145763397217, "learning_rate": 4.791317734545312e-06, "loss": 0.5198, "step": 20370 }, { "epoch": 1.6868068200629036, "grad_norm": 4.52134370803833, "learning_rate": 4.786506736080284e-06, "loss": 0.565, "step": 20380 }, { "epoch": 1.6876344975997353, "grad_norm": 7.11758279800415, "learning_rate": 4.78169593562627e-06, "loss": 0.553, "step": 20390 }, { "epoch": 1.688462175136567, "grad_norm": 5.547896385192871, "learning_rate": 4.776885337645198e-06, "loss": 0.5557, "step": 20400 }, { "epoch": 1.6892898526733986, "grad_norm": 5.532888412475586, "learning_rate": 4.772074946598808e-06, "loss": 0.4877, "step": 20410 }, { "epoch": 1.6901175302102303, "grad_norm": 3.8221609592437744, "learning_rate": 4.767264766948645e-06, "loss": 0.4745, "step": 20420 }, { "epoch": 1.6909452077470617, "grad_norm": 4.871181964874268, "learning_rate": 4.762454803156065e-06, "loss": 0.4946, "step": 20430 }, { "epoch": 1.6917728852838934, "grad_norm": 4.617056369781494, "learning_rate": 4.757645059682218e-06, "loss": 0.5096, "step": 20440 }, { "epoch": 1.692600562820725, "grad_norm": 4.311750411987305, "learning_rate": 4.75283554098805e-06, "loss": 0.4609, "step": 20450 }, { "epoch": 1.6934282403575567, "grad_norm": 4.406552791595459, "learning_rate": 4.748026251534301e-06, "loss": 0.4055, "step": 20460 }, { "epoch": 1.6942559178943883, "grad_norm": 5.776482105255127, "learning_rate": 4.7432171957814976e-06, "loss": 0.5005, "step": 20470 }, { "epoch": 1.69508359543122, "grad_norm": 4.0988054275512695, "learning_rate": 4.738408378189946e-06, "loss": 0.5472, "step": 20480 }, { "epoch": 1.6959112729680517, "grad_norm": 5.436655044555664, "learning_rate": 4.733599803219739e-06, "loss": 0.4543, "step": 20490 }, { "epoch": 1.6967389505048833, "grad_norm": 3.7724997997283936, "learning_rate": 4.7287914753307385e-06, "loss": 0.4897, "step": 20500 }, { "epoch": 1.697566628041715, "grad_norm": 5.570647716522217, "learning_rate": 4.723983398982577e-06, "loss": 0.4731, "step": 20510 }, { "epoch": 1.6983943055785466, "grad_norm": 5.511480808258057, "learning_rate": 4.719175578634658e-06, "loss": 0.4784, "step": 20520 }, { "epoch": 1.699221983115378, "grad_norm": 5.056786060333252, "learning_rate": 4.714368018746144e-06, "loss": 0.4431, "step": 20530 }, { "epoch": 1.7000496606522097, "grad_norm": 4.658056735992432, "learning_rate": 4.709560723775955e-06, "loss": 0.5164, "step": 20540 }, { "epoch": 1.7008773381890414, "grad_norm": 4.8835930824279785, "learning_rate": 4.70475369818277e-06, "loss": 0.518, "step": 20550 }, { "epoch": 1.701705015725873, "grad_norm": 5.151829719543457, "learning_rate": 4.699946946425016e-06, "loss": 0.5341, "step": 20560 }, { "epoch": 1.7025326932627047, "grad_norm": 5.382249355316162, "learning_rate": 4.695140472960863e-06, "loss": 0.5422, "step": 20570 }, { "epoch": 1.7033603707995364, "grad_norm": 6.22735595703125, "learning_rate": 4.6903342822482275e-06, "loss": 0.4414, "step": 20580 }, { "epoch": 1.704188048336368, "grad_norm": 5.270052909851074, "learning_rate": 4.68552837874476e-06, "loss": 0.567, "step": 20590 }, { "epoch": 1.7050157258731997, "grad_norm": 4.429057598114014, "learning_rate": 4.680722766907845e-06, "loss": 0.499, "step": 20600 }, { "epoch": 1.7058434034100314, "grad_norm": 5.190255641937256, "learning_rate": 4.675917451194602e-06, "loss": 0.4099, "step": 20610 }, { "epoch": 1.706671080946863, "grad_norm": 5.39527702331543, "learning_rate": 4.671112436061868e-06, "loss": 0.4949, "step": 20620 }, { "epoch": 1.7074987584836947, "grad_norm": 4.767218112945557, "learning_rate": 4.666307725966207e-06, "loss": 0.4625, "step": 20630 }, { "epoch": 1.7083264360205264, "grad_norm": 4.040586471557617, "learning_rate": 4.661503325363896e-06, "loss": 0.4332, "step": 20640 }, { "epoch": 1.709154113557358, "grad_norm": 5.337644577026367, "learning_rate": 4.656699238710928e-06, "loss": 0.5331, "step": 20650 }, { "epoch": 1.7099817910941897, "grad_norm": 4.243056774139404, "learning_rate": 4.651895470463e-06, "loss": 0.4818, "step": 20660 }, { "epoch": 1.7108094686310213, "grad_norm": 6.0655717849731445, "learning_rate": 4.647092025075524e-06, "loss": 0.447, "step": 20670 }, { "epoch": 1.711637146167853, "grad_norm": 4.337826251983643, "learning_rate": 4.642288907003601e-06, "loss": 0.5027, "step": 20680 }, { "epoch": 1.7124648237046847, "grad_norm": 5.331536293029785, "learning_rate": 4.637486120702035e-06, "loss": 0.5011, "step": 20690 }, { "epoch": 1.7132925012415163, "grad_norm": 5.420194149017334, "learning_rate": 4.632683670625321e-06, "loss": 0.5363, "step": 20700 }, { "epoch": 1.714120178778348, "grad_norm": 4.55141544342041, "learning_rate": 4.627881561227641e-06, "loss": 0.4903, "step": 20710 }, { "epoch": 1.7149478563151797, "grad_norm": 4.8301310539245605, "learning_rate": 4.623079796962859e-06, "loss": 0.4806, "step": 20720 }, { "epoch": 1.7157755338520113, "grad_norm": 3.1805427074432373, "learning_rate": 4.618278382284525e-06, "loss": 0.5171, "step": 20730 }, { "epoch": 1.716603211388843, "grad_norm": 4.910946369171143, "learning_rate": 4.613477321645863e-06, "loss": 0.5118, "step": 20740 }, { "epoch": 1.7174308889256746, "grad_norm": 3.3921985626220703, "learning_rate": 4.608676619499764e-06, "loss": 0.4615, "step": 20750 }, { "epoch": 1.7182585664625063, "grad_norm": 4.196315765380859, "learning_rate": 4.6038762802987905e-06, "loss": 0.4532, "step": 20760 }, { "epoch": 1.719086243999338, "grad_norm": 4.133450031280518, "learning_rate": 4.599076308495167e-06, "loss": 0.5153, "step": 20770 }, { "epoch": 1.7199139215361696, "grad_norm": 3.9897754192352295, "learning_rate": 4.594276708540777e-06, "loss": 0.5133, "step": 20780 }, { "epoch": 1.7207415990730013, "grad_norm": 3.7448127269744873, "learning_rate": 4.589477484887162e-06, "loss": 0.5642, "step": 20790 }, { "epoch": 1.721569276609833, "grad_norm": 4.749078273773193, "learning_rate": 4.58467864198551e-06, "loss": 0.4764, "step": 20800 }, { "epoch": 1.7223969541466646, "grad_norm": 4.397484302520752, "learning_rate": 4.5798801842866615e-06, "loss": 0.4535, "step": 20810 }, { "epoch": 1.7232246316834963, "grad_norm": 4.019654273986816, "learning_rate": 4.575082116241094e-06, "loss": 0.4589, "step": 20820 }, { "epoch": 1.7240523092203277, "grad_norm": 4.780057907104492, "learning_rate": 4.570284442298923e-06, "loss": 0.4892, "step": 20830 }, { "epoch": 1.7248799867571594, "grad_norm": 3.9628732204437256, "learning_rate": 4.5654871669099065e-06, "loss": 0.4472, "step": 20840 }, { "epoch": 1.725707664293991, "grad_norm": 7.245631217956543, "learning_rate": 4.560690294523426e-06, "loss": 0.5126, "step": 20850 }, { "epoch": 1.7265353418308227, "grad_norm": 4.727609157562256, "learning_rate": 4.555893829588491e-06, "loss": 0.4603, "step": 20860 }, { "epoch": 1.7273630193676544, "grad_norm": 6.146944046020508, "learning_rate": 4.551097776553733e-06, "loss": 0.5038, "step": 20870 }, { "epoch": 1.728190696904486, "grad_norm": 5.590038299560547, "learning_rate": 4.546302139867401e-06, "loss": 0.5455, "step": 20880 }, { "epoch": 1.7290183744413177, "grad_norm": 3.720099687576294, "learning_rate": 4.541506923977359e-06, "loss": 0.5875, "step": 20890 }, { "epoch": 1.7298460519781493, "grad_norm": 5.11767578125, "learning_rate": 4.536712133331081e-06, "loss": 0.5213, "step": 20900 }, { "epoch": 1.730673729514981, "grad_norm": 8.937885284423828, "learning_rate": 4.531917772375645e-06, "loss": 0.5265, "step": 20910 }, { "epoch": 1.7315014070518127, "grad_norm": 3.76680064201355, "learning_rate": 4.527123845557732e-06, "loss": 0.5029, "step": 20920 }, { "epoch": 1.7323290845886443, "grad_norm": 3.5443484783172607, "learning_rate": 4.522330357323618e-06, "loss": 0.537, "step": 20930 }, { "epoch": 1.7331567621254758, "grad_norm": 4.902674674987793, "learning_rate": 4.517537312119174e-06, "loss": 0.462, "step": 20940 }, { "epoch": 1.7339844396623074, "grad_norm": 3.4437811374664307, "learning_rate": 4.512744714389861e-06, "loss": 0.5394, "step": 20950 }, { "epoch": 1.734812117199139, "grad_norm": 5.951440811157227, "learning_rate": 4.507952568580722e-06, "loss": 0.5588, "step": 20960 }, { "epoch": 1.7356397947359707, "grad_norm": 3.853818893432617, "learning_rate": 4.503160879136386e-06, "loss": 0.443, "step": 20970 }, { "epoch": 1.7364674722728024, "grad_norm": 5.718456268310547, "learning_rate": 4.498369650501053e-06, "loss": 0.4227, "step": 20980 }, { "epoch": 1.737295149809634, "grad_norm": 5.046834468841553, "learning_rate": 4.4935788871185e-06, "loss": 0.4987, "step": 20990 }, { "epoch": 1.7381228273464657, "grad_norm": 3.4160773754119873, "learning_rate": 4.488788593432066e-06, "loss": 0.4805, "step": 21000 }, { "epoch": 1.7389505048832974, "grad_norm": 5.2083821296691895, "learning_rate": 4.483998773884663e-06, "loss": 0.5072, "step": 21010 }, { "epoch": 1.739778182420129, "grad_norm": 3.4137027263641357, "learning_rate": 4.4792094329187594e-06, "loss": 0.5134, "step": 21020 }, { "epoch": 1.7406058599569607, "grad_norm": 4.070508003234863, "learning_rate": 4.4744205749763785e-06, "loss": 0.3914, "step": 21030 }, { "epoch": 1.7414335374937924, "grad_norm": 3.878054618835449, "learning_rate": 4.469632204499096e-06, "loss": 0.5027, "step": 21040 }, { "epoch": 1.742261215030624, "grad_norm": 4.685537338256836, "learning_rate": 4.464844325928035e-06, "loss": 0.513, "step": 21050 }, { "epoch": 1.7430888925674557, "grad_norm": 3.950528144836426, "learning_rate": 4.460056943703867e-06, "loss": 0.4667, "step": 21060 }, { "epoch": 1.7439165701042874, "grad_norm": 4.152957916259766, "learning_rate": 4.455270062266795e-06, "loss": 0.4621, "step": 21070 }, { "epoch": 1.744744247641119, "grad_norm": 4.439316749572754, "learning_rate": 4.450483686056565e-06, "loss": 0.5066, "step": 21080 }, { "epoch": 1.7455719251779507, "grad_norm": 3.7215447425842285, "learning_rate": 4.445697819512453e-06, "loss": 0.4954, "step": 21090 }, { "epoch": 1.7463996027147823, "grad_norm": 4.8607306480407715, "learning_rate": 4.440912467073259e-06, "loss": 0.5821, "step": 21100 }, { "epoch": 1.747227280251614, "grad_norm": 4.978580951690674, "learning_rate": 4.4361276331773055e-06, "loss": 0.5589, "step": 21110 }, { "epoch": 1.7480549577884457, "grad_norm": 3.9407875537872314, "learning_rate": 4.431343322262441e-06, "loss": 0.4662, "step": 21120 }, { "epoch": 1.7488826353252773, "grad_norm": 7.012410640716553, "learning_rate": 4.42655953876602e-06, "loss": 0.5306, "step": 21130 }, { "epoch": 1.749710312862109, "grad_norm": 4.114837646484375, "learning_rate": 4.421776287124917e-06, "loss": 0.4701, "step": 21140 }, { "epoch": 1.7505379903989406, "grad_norm": 5.128622055053711, "learning_rate": 4.416993571775506e-06, "loss": 0.4617, "step": 21150 }, { "epoch": 1.7513656679357723, "grad_norm": 4.844719886779785, "learning_rate": 4.412211397153666e-06, "loss": 0.5038, "step": 21160 }, { "epoch": 1.752193345472604, "grad_norm": 2.8411476612091064, "learning_rate": 4.407429767694776e-06, "loss": 0.5179, "step": 21170 }, { "epoch": 1.7530210230094356, "grad_norm": 3.894740104675293, "learning_rate": 4.4026486878337075e-06, "loss": 0.5052, "step": 21180 }, { "epoch": 1.7538487005462673, "grad_norm": 5.88032341003418, "learning_rate": 4.397868162004819e-06, "loss": 0.5177, "step": 21190 }, { "epoch": 1.754676378083099, "grad_norm": 6.528963088989258, "learning_rate": 4.393088194641966e-06, "loss": 0.4903, "step": 21200 }, { "epoch": 1.7555040556199306, "grad_norm": 5.093194484710693, "learning_rate": 4.388308790178474e-06, "loss": 0.504, "step": 21210 }, { "epoch": 1.7563317331567623, "grad_norm": 5.747429847717285, "learning_rate": 4.383529953047151e-06, "loss": 0.5377, "step": 21220 }, { "epoch": 1.757159410693594, "grad_norm": 4.508754730224609, "learning_rate": 4.378751687680284e-06, "loss": 0.4329, "step": 21230 }, { "epoch": 1.7579870882304254, "grad_norm": 5.197535037994385, "learning_rate": 4.37397399850962e-06, "loss": 0.4858, "step": 21240 }, { "epoch": 1.758814765767257, "grad_norm": 4.289979934692383, "learning_rate": 4.369196889966376e-06, "loss": 0.4927, "step": 21250 }, { "epoch": 1.7596424433040887, "grad_norm": 4.2137651443481445, "learning_rate": 4.3644203664812355e-06, "loss": 0.488, "step": 21260 }, { "epoch": 1.7604701208409204, "grad_norm": 3.159083604812622, "learning_rate": 4.359644432484332e-06, "loss": 0.5146, "step": 21270 }, { "epoch": 1.761297798377752, "grad_norm": 4.967798709869385, "learning_rate": 4.354869092405254e-06, "loss": 0.5365, "step": 21280 }, { "epoch": 1.7621254759145837, "grad_norm": 4.2390217781066895, "learning_rate": 4.350094350673044e-06, "loss": 0.4254, "step": 21290 }, { "epoch": 1.7629531534514153, "grad_norm": 3.227137804031372, "learning_rate": 4.345320211716182e-06, "loss": 0.4795, "step": 21300 }, { "epoch": 1.763780830988247, "grad_norm": 5.936793804168701, "learning_rate": 4.340546679962593e-06, "loss": 0.5268, "step": 21310 }, { "epoch": 1.7646085085250787, "grad_norm": 4.080915451049805, "learning_rate": 4.33577375983964e-06, "loss": 0.5389, "step": 21320 }, { "epoch": 1.7654361860619103, "grad_norm": 4.361737251281738, "learning_rate": 4.331001455774117e-06, "loss": 0.4078, "step": 21330 }, { "epoch": 1.7662638635987418, "grad_norm": 4.914424896240234, "learning_rate": 4.326229772192248e-06, "loss": 0.4704, "step": 21340 }, { "epoch": 1.7670915411355734, "grad_norm": 4.573816776275635, "learning_rate": 4.321458713519679e-06, "loss": 0.547, "step": 21350 }, { "epoch": 1.767919218672405, "grad_norm": 5.025706768035889, "learning_rate": 4.316688284181478e-06, "loss": 0.3872, "step": 21360 }, { "epoch": 1.7687468962092368, "grad_norm": 4.043321132659912, "learning_rate": 4.3119184886021255e-06, "loss": 0.469, "step": 21370 }, { "epoch": 1.7695745737460684, "grad_norm": 4.625423431396484, "learning_rate": 4.307149331205524e-06, "loss": 0.451, "step": 21380 }, { "epoch": 1.7704022512829, "grad_norm": 4.754919052124023, "learning_rate": 4.302380816414975e-06, "loss": 0.4928, "step": 21390 }, { "epoch": 1.7712299288197317, "grad_norm": 4.42257833480835, "learning_rate": 4.297612948653188e-06, "loss": 0.4951, "step": 21400 }, { "epoch": 1.7720576063565634, "grad_norm": 4.394595623016357, "learning_rate": 4.29284573234227e-06, "loss": 0.5164, "step": 21410 }, { "epoch": 1.772885283893395, "grad_norm": 4.219302654266357, "learning_rate": 4.288079171903725e-06, "loss": 0.4842, "step": 21420 }, { "epoch": 1.7737129614302267, "grad_norm": 4.481112957000732, "learning_rate": 4.283313271758448e-06, "loss": 0.5065, "step": 21430 }, { "epoch": 1.7745406389670584, "grad_norm": 4.957474708557129, "learning_rate": 4.278548036326724e-06, "loss": 0.5114, "step": 21440 }, { "epoch": 1.77536831650389, "grad_norm": 4.718572616577148, "learning_rate": 4.273783470028221e-06, "loss": 0.4613, "step": 21450 }, { "epoch": 1.7761959940407217, "grad_norm": 4.586964130401611, "learning_rate": 4.2690195772819835e-06, "loss": 0.4437, "step": 21460 }, { "epoch": 1.7770236715775534, "grad_norm": 4.7771830558776855, "learning_rate": 4.264256362506433e-06, "loss": 0.4571, "step": 21470 }, { "epoch": 1.777851349114385, "grad_norm": 9.210153579711914, "learning_rate": 4.259493830119361e-06, "loss": 0.5119, "step": 21480 }, { "epoch": 1.7786790266512167, "grad_norm": 4.850682258605957, "learning_rate": 4.2547319845379295e-06, "loss": 0.4736, "step": 21490 }, { "epoch": 1.7795067041880483, "grad_norm": 3.910918951034546, "learning_rate": 4.24997083017866e-06, "loss": 0.5453, "step": 21500 }, { "epoch": 1.78033438172488, "grad_norm": 4.990531921386719, "learning_rate": 4.245210371457436e-06, "loss": 0.4759, "step": 21510 }, { "epoch": 1.7811620592617117, "grad_norm": 2.962331771850586, "learning_rate": 4.240450612789493e-06, "loss": 0.5149, "step": 21520 }, { "epoch": 1.7819897367985433, "grad_norm": 5.555054187774658, "learning_rate": 4.235691558589417e-06, "loss": 0.4254, "step": 21530 }, { "epoch": 1.782817414335375, "grad_norm": 4.009247779846191, "learning_rate": 4.2309332132711426e-06, "loss": 0.4835, "step": 21540 }, { "epoch": 1.7836450918722067, "grad_norm": 3.5421335697174072, "learning_rate": 4.226175581247947e-06, "loss": 0.4442, "step": 21550 }, { "epoch": 1.7844727694090383, "grad_norm": 9.43945026397705, "learning_rate": 4.221418666932445e-06, "loss": 0.4741, "step": 21560 }, { "epoch": 1.78530044694587, "grad_norm": 4.215597152709961, "learning_rate": 4.216662474736586e-06, "loss": 0.5873, "step": 21570 }, { "epoch": 1.7861281244827016, "grad_norm": 5.212898254394531, "learning_rate": 4.211907009071648e-06, "loss": 0.478, "step": 21580 }, { "epoch": 1.7869558020195333, "grad_norm": 3.3247711658477783, "learning_rate": 4.207152274348238e-06, "loss": 0.4544, "step": 21590 }, { "epoch": 1.787783479556365, "grad_norm": 4.9905104637146, "learning_rate": 4.20239827497628e-06, "loss": 0.4926, "step": 21600 }, { "epoch": 1.7886111570931966, "grad_norm": 5.202200889587402, "learning_rate": 4.1976450153650235e-06, "loss": 0.4709, "step": 21610 }, { "epoch": 1.7894388346300283, "grad_norm": 3.781646251678467, "learning_rate": 4.19289249992303e-06, "loss": 0.4768, "step": 21620 }, { "epoch": 1.79026651216686, "grad_norm": 4.933925151824951, "learning_rate": 4.188140733058164e-06, "loss": 0.4383, "step": 21630 }, { "epoch": 1.7910941897036916, "grad_norm": 5.017529487609863, "learning_rate": 4.183389719177603e-06, "loss": 0.5109, "step": 21640 }, { "epoch": 1.791921867240523, "grad_norm": 5.5895819664001465, "learning_rate": 4.178639462687822e-06, "loss": 0.5363, "step": 21650 }, { "epoch": 1.7927495447773547, "grad_norm": 7.989558219909668, "learning_rate": 4.173889967994596e-06, "loss": 0.4698, "step": 21660 }, { "epoch": 1.7935772223141864, "grad_norm": 5.577826976776123, "learning_rate": 4.1691412395029925e-06, "loss": 0.5057, "step": 21670 }, { "epoch": 1.794404899851018, "grad_norm": 5.889301776885986, "learning_rate": 4.16439328161737e-06, "loss": 0.5311, "step": 21680 }, { "epoch": 1.7952325773878497, "grad_norm": 3.9766714572906494, "learning_rate": 4.15964609874137e-06, "loss": 0.4946, "step": 21690 }, { "epoch": 1.7960602549246814, "grad_norm": 4.746738910675049, "learning_rate": 4.154899695277915e-06, "loss": 0.533, "step": 21700 }, { "epoch": 1.796887932461513, "grad_norm": 4.189581871032715, "learning_rate": 4.150154075629205e-06, "loss": 0.3976, "step": 21710 }, { "epoch": 1.7977156099983447, "grad_norm": 5.659235000610352, "learning_rate": 4.145409244196714e-06, "loss": 0.4724, "step": 21720 }, { "epoch": 1.7985432875351763, "grad_norm": 5.391319274902344, "learning_rate": 4.140665205381186e-06, "loss": 0.5087, "step": 21730 }, { "epoch": 1.799370965072008, "grad_norm": 4.170831680297852, "learning_rate": 4.135921963582627e-06, "loss": 0.4374, "step": 21740 }, { "epoch": 1.8001986426088394, "grad_norm": 7.397477149963379, "learning_rate": 4.131179523200307e-06, "loss": 0.5262, "step": 21750 }, { "epoch": 1.801026320145671, "grad_norm": 4.346487522125244, "learning_rate": 4.126437888632748e-06, "loss": 0.5146, "step": 21760 }, { "epoch": 1.8018539976825028, "grad_norm": 4.316692352294922, "learning_rate": 4.12169706427773e-06, "loss": 0.4696, "step": 21770 }, { "epoch": 1.8026816752193344, "grad_norm": 4.636841297149658, "learning_rate": 4.1169570545322755e-06, "loss": 0.4584, "step": 21780 }, { "epoch": 1.803509352756166, "grad_norm": 3.8209869861602783, "learning_rate": 4.1122178637926605e-06, "loss": 0.544, "step": 21790 }, { "epoch": 1.8043370302929977, "grad_norm": 3.3335301876068115, "learning_rate": 4.107479496454392e-06, "loss": 0.4955, "step": 21800 }, { "epoch": 1.8051647078298294, "grad_norm": 5.409304141998291, "learning_rate": 4.102741956912218e-06, "loss": 0.5518, "step": 21810 }, { "epoch": 1.805992385366661, "grad_norm": 4.313687801361084, "learning_rate": 4.098005249560117e-06, "loss": 0.4997, "step": 21820 }, { "epoch": 1.8068200629034927, "grad_norm": 5.232903957366943, "learning_rate": 4.093269378791299e-06, "loss": 0.5277, "step": 21830 }, { "epoch": 1.8076477404403244, "grad_norm": 5.999606132507324, "learning_rate": 4.0885343489981915e-06, "loss": 0.498, "step": 21840 }, { "epoch": 1.808475417977156, "grad_norm": 5.848676681518555, "learning_rate": 4.083800164572451e-06, "loss": 0.5177, "step": 21850 }, { "epoch": 1.8093030955139877, "grad_norm": 4.817202091217041, "learning_rate": 4.079066829904943e-06, "loss": 0.4738, "step": 21860 }, { "epoch": 1.8101307730508194, "grad_norm": 4.908805847167969, "learning_rate": 4.074334349385748e-06, "loss": 0.4547, "step": 21870 }, { "epoch": 1.810958450587651, "grad_norm": 2.8370201587677, "learning_rate": 4.06960272740415e-06, "loss": 0.4359, "step": 21880 }, { "epoch": 1.8117861281244827, "grad_norm": 3.2654428482055664, "learning_rate": 4.0648719683486445e-06, "loss": 0.4158, "step": 21890 }, { "epoch": 1.8126138056613144, "grad_norm": 3.20102596282959, "learning_rate": 4.060142076606917e-06, "loss": 0.527, "step": 21900 }, { "epoch": 1.813441483198146, "grad_norm": 4.916654586791992, "learning_rate": 4.05541305656586e-06, "loss": 0.5132, "step": 21910 }, { "epoch": 1.8142691607349777, "grad_norm": 3.8432557582855225, "learning_rate": 4.050684912611546e-06, "loss": 0.5047, "step": 21920 }, { "epoch": 1.8150968382718093, "grad_norm": 5.007942199707031, "learning_rate": 4.045957649129243e-06, "loss": 0.4904, "step": 21930 }, { "epoch": 1.815924515808641, "grad_norm": 6.925799369812012, "learning_rate": 4.041231270503398e-06, "loss": 0.4704, "step": 21940 }, { "epoch": 1.8167521933454727, "grad_norm": 5.494277000427246, "learning_rate": 4.036505781117641e-06, "loss": 0.4592, "step": 21950 }, { "epoch": 1.8175798708823043, "grad_norm": 4.014790058135986, "learning_rate": 4.03178118535477e-06, "loss": 0.4457, "step": 21960 }, { "epoch": 1.818407548419136, "grad_norm": 4.778895378112793, "learning_rate": 4.0270574875967665e-06, "loss": 0.4751, "step": 21970 }, { "epoch": 1.8192352259559676, "grad_norm": 5.283944606781006, "learning_rate": 4.022334692224768e-06, "loss": 0.3917, "step": 21980 }, { "epoch": 1.8200629034927993, "grad_norm": 2.6902074813842773, "learning_rate": 4.017612803619079e-06, "loss": 0.5144, "step": 21990 }, { "epoch": 1.820890581029631, "grad_norm": 6.665579319000244, "learning_rate": 4.012891826159163e-06, "loss": 0.4674, "step": 22000 }, { "epoch": 1.8217182585664626, "grad_norm": 3.3130650520324707, "learning_rate": 4.00817176422364e-06, "loss": 0.5158, "step": 22010 }, { "epoch": 1.8225459361032943, "grad_norm": 4.8857550621032715, "learning_rate": 4.003452622190276e-06, "loss": 0.4362, "step": 22020 }, { "epoch": 1.823373613640126, "grad_norm": 5.208150863647461, "learning_rate": 3.998734404435991e-06, "loss": 0.4547, "step": 22030 }, { "epoch": 1.8242012911769576, "grad_norm": 4.226018905639648, "learning_rate": 3.994017115336841e-06, "loss": 0.4894, "step": 22040 }, { "epoch": 1.825028968713789, "grad_norm": 7.035213947296143, "learning_rate": 3.989300759268025e-06, "loss": 0.495, "step": 22050 }, { "epoch": 1.8258566462506207, "grad_norm": 4.6590895652771, "learning_rate": 3.984585340603876e-06, "loss": 0.4726, "step": 22060 }, { "epoch": 1.8266843237874524, "grad_norm": 4.0533928871154785, "learning_rate": 3.979870863717854e-06, "loss": 0.5298, "step": 22070 }, { "epoch": 1.827512001324284, "grad_norm": 4.722029209136963, "learning_rate": 3.9751573329825475e-06, "loss": 0.5099, "step": 22080 }, { "epoch": 1.8283396788611157, "grad_norm": 3.6414806842803955, "learning_rate": 3.970444752769673e-06, "loss": 0.4473, "step": 22090 }, { "epoch": 1.8291673563979474, "grad_norm": 4.947720050811768, "learning_rate": 3.965733127450057e-06, "loss": 0.5502, "step": 22100 }, { "epoch": 1.829995033934779, "grad_norm": 6.760218143463135, "learning_rate": 3.961022461393647e-06, "loss": 0.4835, "step": 22110 }, { "epoch": 1.8308227114716107, "grad_norm": 5.458581447601318, "learning_rate": 3.956312758969494e-06, "loss": 0.4676, "step": 22120 }, { "epoch": 1.8316503890084423, "grad_norm": 5.036542892456055, "learning_rate": 3.951604024545761e-06, "loss": 0.4377, "step": 22130 }, { "epoch": 1.832478066545274, "grad_norm": 3.5829007625579834, "learning_rate": 3.946896262489709e-06, "loss": 0.4524, "step": 22140 }, { "epoch": 1.8333057440821054, "grad_norm": 3.775632381439209, "learning_rate": 3.942189477167704e-06, "loss": 0.4561, "step": 22150 }, { "epoch": 1.834133421618937, "grad_norm": 2.999098539352417, "learning_rate": 3.937483672945197e-06, "loss": 0.4705, "step": 22160 }, { "epoch": 1.8349610991557688, "grad_norm": 4.038907527923584, "learning_rate": 3.932778854186737e-06, "loss": 0.3814, "step": 22170 }, { "epoch": 1.8357887766926004, "grad_norm": 3.239431381225586, "learning_rate": 3.928075025255954e-06, "loss": 0.4596, "step": 22180 }, { "epoch": 1.836616454229432, "grad_norm": 4.6215362548828125, "learning_rate": 3.923372190515558e-06, "loss": 0.4966, "step": 22190 }, { "epoch": 1.8374441317662638, "grad_norm": 4.4355854988098145, "learning_rate": 3.918670354327346e-06, "loss": 0.4742, "step": 22200 }, { "epoch": 1.8382718093030954, "grad_norm": 5.224678039550781, "learning_rate": 3.913969521052181e-06, "loss": 0.4603, "step": 22210 }, { "epoch": 1.839099486839927, "grad_norm": 4.8325700759887695, "learning_rate": 3.909269695049998e-06, "loss": 0.4857, "step": 22220 }, { "epoch": 1.8399271643767587, "grad_norm": 4.115813732147217, "learning_rate": 3.904570880679797e-06, "loss": 0.5491, "step": 22230 }, { "epoch": 1.8407548419135904, "grad_norm": 4.918093204498291, "learning_rate": 3.899873082299642e-06, "loss": 0.5671, "step": 22240 }, { "epoch": 1.841582519450422, "grad_norm": 4.354982852935791, "learning_rate": 3.895176304266651e-06, "loss": 0.4038, "step": 22250 }, { "epoch": 1.8424101969872537, "grad_norm": 4.05272912979126, "learning_rate": 3.890480550937002e-06, "loss": 0.422, "step": 22260 }, { "epoch": 1.8432378745240854, "grad_norm": 3.8030874729156494, "learning_rate": 3.885785826665913e-06, "loss": 0.4513, "step": 22270 }, { "epoch": 1.844065552060917, "grad_norm": 3.212639808654785, "learning_rate": 3.881092135807658e-06, "loss": 0.5259, "step": 22280 }, { "epoch": 1.8448932295977487, "grad_norm": 4.951358318328857, "learning_rate": 3.876399482715546e-06, "loss": 0.473, "step": 22290 }, { "epoch": 1.8457209071345804, "grad_norm": 4.879635810852051, "learning_rate": 3.871707871741924e-06, "loss": 0.4404, "step": 22300 }, { "epoch": 1.846548584671412, "grad_norm": 3.4224483966827393, "learning_rate": 3.867017307238172e-06, "loss": 0.5544, "step": 22310 }, { "epoch": 1.8473762622082437, "grad_norm": 4.774829387664795, "learning_rate": 3.862327793554705e-06, "loss": 0.4124, "step": 22320 }, { "epoch": 1.8482039397450754, "grad_norm": 6.194809436798096, "learning_rate": 3.857639335040958e-06, "loss": 0.4574, "step": 22330 }, { "epoch": 1.849031617281907, "grad_norm": 3.0252485275268555, "learning_rate": 3.852951936045388e-06, "loss": 0.4818, "step": 22340 }, { "epoch": 1.8498592948187387, "grad_norm": 5.300130844116211, "learning_rate": 3.84826560091547e-06, "loss": 0.5029, "step": 22350 }, { "epoch": 1.8506869723555703, "grad_norm": 4.890450477600098, "learning_rate": 3.843580333997692e-06, "loss": 0.5796, "step": 22360 }, { "epoch": 1.851514649892402, "grad_norm": 6.78560733795166, "learning_rate": 3.838896139637551e-06, "loss": 0.5518, "step": 22370 }, { "epoch": 1.8523423274292337, "grad_norm": 4.856921195983887, "learning_rate": 3.834213022179552e-06, "loss": 0.4976, "step": 22380 }, { "epoch": 1.8531700049660653, "grad_norm": 5.178829669952393, "learning_rate": 3.829530985967198e-06, "loss": 0.505, "step": 22390 }, { "epoch": 1.853997682502897, "grad_norm": 4.928654193878174, "learning_rate": 3.8248500353429904e-06, "loss": 0.5071, "step": 22400 }, { "epoch": 1.8548253600397286, "grad_norm": 4.254227161407471, "learning_rate": 3.820170174648423e-06, "loss": 0.5079, "step": 22410 }, { "epoch": 1.8556530375765603, "grad_norm": 3.9955945014953613, "learning_rate": 3.815491408223978e-06, "loss": 0.4517, "step": 22420 }, { "epoch": 1.856480715113392, "grad_norm": 4.606036186218262, "learning_rate": 3.8108137404091244e-06, "loss": 0.5227, "step": 22430 }, { "epoch": 1.8573083926502236, "grad_norm": 5.9690470695495605, "learning_rate": 3.8061371755423126e-06, "loss": 0.4997, "step": 22440 }, { "epoch": 1.8581360701870553, "grad_norm": 3.3051505088806152, "learning_rate": 3.8014617179609698e-06, "loss": 0.504, "step": 22450 }, { "epoch": 1.8589637477238867, "grad_norm": 4.825021743774414, "learning_rate": 3.7967873720014947e-06, "loss": 0.4672, "step": 22460 }, { "epoch": 1.8597914252607184, "grad_norm": 6.07810640335083, "learning_rate": 3.792114141999256e-06, "loss": 0.4909, "step": 22470 }, { "epoch": 1.86061910279755, "grad_norm": 8.225743293762207, "learning_rate": 3.7874420322885858e-06, "loss": 0.5031, "step": 22480 }, { "epoch": 1.8614467803343817, "grad_norm": 4.247690200805664, "learning_rate": 3.7827710472027785e-06, "loss": 0.4434, "step": 22490 }, { "epoch": 1.8622744578712134, "grad_norm": 3.7208316326141357, "learning_rate": 3.7781011910740896e-06, "loss": 0.4971, "step": 22500 }, { "epoch": 1.863102135408045, "grad_norm": 4.844273567199707, "learning_rate": 3.77343246823372e-06, "loss": 0.4142, "step": 22510 }, { "epoch": 1.8639298129448767, "grad_norm": 5.308693885803223, "learning_rate": 3.7687648830118233e-06, "loss": 0.5044, "step": 22520 }, { "epoch": 1.8647574904817084, "grad_norm": 4.850386619567871, "learning_rate": 3.764098439737497e-06, "loss": 0.5479, "step": 22530 }, { "epoch": 1.86558516801854, "grad_norm": 4.675843238830566, "learning_rate": 3.759433142738782e-06, "loss": 0.5659, "step": 22540 }, { "epoch": 1.8664128455553717, "grad_norm": 5.061076641082764, "learning_rate": 3.754768996342651e-06, "loss": 0.4342, "step": 22550 }, { "epoch": 1.8672405230922031, "grad_norm": 5.522410869598389, "learning_rate": 3.7501060048750173e-06, "loss": 0.4904, "step": 22560 }, { "epoch": 1.8680682006290348, "grad_norm": 4.497849464416504, "learning_rate": 3.745444172660715e-06, "loss": 0.5226, "step": 22570 }, { "epoch": 1.8688958781658664, "grad_norm": 4.29834508895874, "learning_rate": 3.740783504023507e-06, "loss": 0.4906, "step": 22580 }, { "epoch": 1.869723555702698, "grad_norm": 4.1297101974487305, "learning_rate": 3.7361240032860757e-06, "loss": 0.5108, "step": 22590 }, { "epoch": 1.8705512332395298, "grad_norm": 4.422542572021484, "learning_rate": 3.731465674770023e-06, "loss": 0.5184, "step": 22600 }, { "epoch": 1.8713789107763614, "grad_norm": 4.838630199432373, "learning_rate": 3.7268085227958584e-06, "loss": 0.5147, "step": 22610 }, { "epoch": 1.872206588313193, "grad_norm": 4.229909420013428, "learning_rate": 3.7221525516830076e-06, "loss": 0.5689, "step": 22620 }, { "epoch": 1.8730342658500247, "grad_norm": 5.009522438049316, "learning_rate": 3.7174977657497947e-06, "loss": 0.5224, "step": 22630 }, { "epoch": 1.8738619433868564, "grad_norm": 6.401365280151367, "learning_rate": 3.712844169313445e-06, "loss": 0.4703, "step": 22640 }, { "epoch": 1.874689620923688, "grad_norm": 4.622521877288818, "learning_rate": 3.7081917666900853e-06, "loss": 0.4186, "step": 22650 }, { "epoch": 1.8755172984605197, "grad_norm": 3.3077144622802734, "learning_rate": 3.7035405621947305e-06, "loss": 0.4708, "step": 22660 }, { "epoch": 1.8763449759973514, "grad_norm": 3.8989028930664062, "learning_rate": 3.6988905601412837e-06, "loss": 0.45, "step": 22670 }, { "epoch": 1.877172653534183, "grad_norm": 3.9935693740844727, "learning_rate": 3.69424176484254e-06, "loss": 0.5125, "step": 22680 }, { "epoch": 1.8780003310710147, "grad_norm": 2.3461532592773438, "learning_rate": 3.6895941806101677e-06, "loss": 0.4623, "step": 22690 }, { "epoch": 1.8788280086078464, "grad_norm": 4.400909423828125, "learning_rate": 3.6849478117547127e-06, "loss": 0.455, "step": 22700 }, { "epoch": 1.879655686144678, "grad_norm": 4.609935283660889, "learning_rate": 3.680302662585598e-06, "loss": 0.4541, "step": 22710 }, { "epoch": 1.8804833636815097, "grad_norm": 5.331910610198975, "learning_rate": 3.675658737411111e-06, "loss": 0.5565, "step": 22720 }, { "epoch": 1.8813110412183414, "grad_norm": 5.201432704925537, "learning_rate": 3.671016040538405e-06, "loss": 0.493, "step": 22730 }, { "epoch": 1.882138718755173, "grad_norm": 3.4024415016174316, "learning_rate": 3.666374576273497e-06, "loss": 0.475, "step": 22740 }, { "epoch": 1.8829663962920047, "grad_norm": 10.61039924621582, "learning_rate": 3.661734348921257e-06, "loss": 0.5349, "step": 22750 }, { "epoch": 1.8837940738288363, "grad_norm": 4.224462985992432, "learning_rate": 3.6570953627854088e-06, "loss": 0.4663, "step": 22760 }, { "epoch": 1.884621751365668, "grad_norm": 4.930361270904541, "learning_rate": 3.652457622168526e-06, "loss": 0.4159, "step": 22770 }, { "epoch": 1.8854494289024997, "grad_norm": 3.8317506313323975, "learning_rate": 3.6478211313720268e-06, "loss": 0.5197, "step": 22780 }, { "epoch": 1.8862771064393313, "grad_norm": 3.4340455532073975, "learning_rate": 3.6431858946961675e-06, "loss": 0.3734, "step": 22790 }, { "epoch": 1.887104783976163, "grad_norm": 5.769198894500732, "learning_rate": 3.638551916440046e-06, "loss": 0.5316, "step": 22800 }, { "epoch": 1.8879324615129947, "grad_norm": 4.866128444671631, "learning_rate": 3.63391920090159e-06, "loss": 0.4562, "step": 22810 }, { "epoch": 1.8887601390498263, "grad_norm": 3.725839376449585, "learning_rate": 3.629287752377557e-06, "loss": 0.468, "step": 22820 }, { "epoch": 1.889587816586658, "grad_norm": 5.070926666259766, "learning_rate": 3.6246575751635288e-06, "loss": 0.5247, "step": 22830 }, { "epoch": 1.8904154941234896, "grad_norm": 4.8087382316589355, "learning_rate": 3.6200286735539077e-06, "loss": 0.5155, "step": 22840 }, { "epoch": 1.8912431716603213, "grad_norm": 3.0604701042175293, "learning_rate": 3.615401051841911e-06, "loss": 0.4827, "step": 22850 }, { "epoch": 1.8920708491971527, "grad_norm": 4.720392227172852, "learning_rate": 3.6107747143195773e-06, "loss": 0.4292, "step": 22860 }, { "epoch": 1.8928985267339844, "grad_norm": 5.32857084274292, "learning_rate": 3.6061496652777427e-06, "loss": 0.4611, "step": 22870 }, { "epoch": 1.893726204270816, "grad_norm": 3.7890965938568115, "learning_rate": 3.6015259090060584e-06, "loss": 0.4366, "step": 22880 }, { "epoch": 1.8945538818076477, "grad_norm": 3.399778127670288, "learning_rate": 3.596903449792969e-06, "loss": 0.4876, "step": 22890 }, { "epoch": 1.8953815593444794, "grad_norm": 4.587395191192627, "learning_rate": 3.5922822919257174e-06, "loss": 0.5316, "step": 22900 }, { "epoch": 1.896209236881311, "grad_norm": 4.589295387268066, "learning_rate": 3.587662439690346e-06, "loss": 0.4559, "step": 22910 }, { "epoch": 1.8970369144181427, "grad_norm": 4.042995452880859, "learning_rate": 3.583043897371678e-06, "loss": 0.4946, "step": 22920 }, { "epoch": 1.8978645919549744, "grad_norm": 5.331206321716309, "learning_rate": 3.578426669253327e-06, "loss": 0.5017, "step": 22930 }, { "epoch": 1.898692269491806, "grad_norm": 6.9334187507629395, "learning_rate": 3.5738107596176854e-06, "loss": 0.4114, "step": 22940 }, { "epoch": 1.8995199470286377, "grad_norm": 4.456981658935547, "learning_rate": 3.569196172745922e-06, "loss": 0.5166, "step": 22950 }, { "epoch": 1.9003476245654691, "grad_norm": 4.464366912841797, "learning_rate": 3.5645829129179787e-06, "loss": 0.4346, "step": 22960 }, { "epoch": 1.9011753021023008, "grad_norm": 4.69502592086792, "learning_rate": 3.559970984412572e-06, "loss": 0.5331, "step": 22970 }, { "epoch": 1.9020029796391325, "grad_norm": 3.8177883625030518, "learning_rate": 3.5553603915071757e-06, "loss": 0.4907, "step": 22980 }, { "epoch": 1.9028306571759641, "grad_norm": 3.8227429389953613, "learning_rate": 3.5507511384780315e-06, "loss": 0.4831, "step": 22990 }, { "epoch": 1.9036583347127958, "grad_norm": 3.117769241333008, "learning_rate": 3.546143229600134e-06, "loss": 0.5004, "step": 23000 }, { "epoch": 1.9044860122496274, "grad_norm": 4.150272369384766, "learning_rate": 3.541536669147232e-06, "loss": 0.4634, "step": 23010 }, { "epoch": 1.905313689786459, "grad_norm": 4.633411884307861, "learning_rate": 3.536931461391824e-06, "loss": 0.4939, "step": 23020 }, { "epoch": 1.9061413673232908, "grad_norm": 4.338712215423584, "learning_rate": 3.5323276106051564e-06, "loss": 0.5459, "step": 23030 }, { "epoch": 1.9069690448601224, "grad_norm": 4.6486358642578125, "learning_rate": 3.527725121057213e-06, "loss": 0.4595, "step": 23040 }, { "epoch": 1.907796722396954, "grad_norm": 3.8518564701080322, "learning_rate": 3.5231239970167174e-06, "loss": 0.4177, "step": 23050 }, { "epoch": 1.9086243999337857, "grad_norm": 4.98037576675415, "learning_rate": 3.5185242427511267e-06, "loss": 0.5381, "step": 23060 }, { "epoch": 1.9094520774706174, "grad_norm": 5.04826545715332, "learning_rate": 3.513925862526627e-06, "loss": 0.4519, "step": 23070 }, { "epoch": 1.910279755007449, "grad_norm": 4.0200371742248535, "learning_rate": 3.5093288606081265e-06, "loss": 0.5088, "step": 23080 }, { "epoch": 1.9111074325442807, "grad_norm": 3.7543885707855225, "learning_rate": 3.504733241259264e-06, "loss": 0.4554, "step": 23090 }, { "epoch": 1.9119351100811124, "grad_norm": 3.3219220638275146, "learning_rate": 3.500139008742388e-06, "loss": 0.5073, "step": 23100 }, { "epoch": 1.912762787617944, "grad_norm": 5.129497528076172, "learning_rate": 3.495546167318565e-06, "loss": 0.4899, "step": 23110 }, { "epoch": 1.9135904651547757, "grad_norm": 3.8424880504608154, "learning_rate": 3.490954721247567e-06, "loss": 0.4906, "step": 23120 }, { "epoch": 1.9144181426916074, "grad_norm": 1.427848219871521, "learning_rate": 3.4863646747878755e-06, "loss": 0.4694, "step": 23130 }, { "epoch": 1.915245820228439, "grad_norm": 3.9323184490203857, "learning_rate": 3.4817760321966727e-06, "loss": 0.5214, "step": 23140 }, { "epoch": 1.9160734977652707, "grad_norm": 4.593691349029541, "learning_rate": 3.477188797729839e-06, "loss": 0.5347, "step": 23150 }, { "epoch": 1.9169011753021024, "grad_norm": 4.528171062469482, "learning_rate": 3.4726029756419505e-06, "loss": 0.572, "step": 23160 }, { "epoch": 1.917728852838934, "grad_norm": 6.689661502838135, "learning_rate": 3.46801857018627e-06, "loss": 0.4487, "step": 23170 }, { "epoch": 1.9185565303757657, "grad_norm": 4.449799060821533, "learning_rate": 3.4634355856147473e-06, "loss": 0.4952, "step": 23180 }, { "epoch": 1.9193842079125973, "grad_norm": 5.861848831176758, "learning_rate": 3.4588540261780145e-06, "loss": 0.4837, "step": 23190 }, { "epoch": 1.920211885449429, "grad_norm": 4.516426086425781, "learning_rate": 3.454273896125383e-06, "loss": 0.468, "step": 23200 }, { "epoch": 1.9210395629862607, "grad_norm": 4.948601722717285, "learning_rate": 3.44969519970484e-06, "loss": 0.5268, "step": 23210 }, { "epoch": 1.9218672405230923, "grad_norm": 4.200933456420898, "learning_rate": 3.445117941163039e-06, "loss": 0.5057, "step": 23220 }, { "epoch": 1.922694918059924, "grad_norm": 20.02766227722168, "learning_rate": 3.4405421247453024e-06, "loss": 0.5285, "step": 23230 }, { "epoch": 1.9235225955967556, "grad_norm": 5.932528018951416, "learning_rate": 3.4359677546956126e-06, "loss": 0.4532, "step": 23240 }, { "epoch": 1.9243502731335873, "grad_norm": 4.912433624267578, "learning_rate": 3.4313948352566153e-06, "loss": 0.4801, "step": 23250 }, { "epoch": 1.925177950670419, "grad_norm": 3.7840917110443115, "learning_rate": 3.426823370669604e-06, "loss": 0.4423, "step": 23260 }, { "epoch": 1.9260056282072504, "grad_norm": 5.122824192047119, "learning_rate": 3.4222533651745315e-06, "loss": 0.4708, "step": 23270 }, { "epoch": 1.926833305744082, "grad_norm": 5.649184703826904, "learning_rate": 3.4176848230099903e-06, "loss": 0.4896, "step": 23280 }, { "epoch": 1.9276609832809137, "grad_norm": 3.5402565002441406, "learning_rate": 3.4131177484132174e-06, "loss": 0.4559, "step": 23290 }, { "epoch": 1.9284886608177454, "grad_norm": 4.258113861083984, "learning_rate": 3.4085521456200876e-06, "loss": 0.4845, "step": 23300 }, { "epoch": 1.929316338354577, "grad_norm": 4.0362467765808105, "learning_rate": 3.4039880188651166e-06, "loss": 0.4119, "step": 23310 }, { "epoch": 1.9301440158914087, "grad_norm": 5.028758525848389, "learning_rate": 3.3994253723814407e-06, "loss": 0.507, "step": 23320 }, { "epoch": 1.9309716934282404, "grad_norm": 4.049137115478516, "learning_rate": 3.3948642104008358e-06, "loss": 0.481, "step": 23330 }, { "epoch": 1.931799370965072, "grad_norm": 3.7800025939941406, "learning_rate": 3.3903045371536914e-06, "loss": 0.4717, "step": 23340 }, { "epoch": 1.9326270485019037, "grad_norm": 3.1977648735046387, "learning_rate": 3.385746356869019e-06, "loss": 0.4245, "step": 23350 }, { "epoch": 1.9334547260387354, "grad_norm": 6.352260589599609, "learning_rate": 3.3811896737744452e-06, "loss": 0.4871, "step": 23360 }, { "epoch": 1.9342824035755668, "grad_norm": 4.456660270690918, "learning_rate": 3.37663449209621e-06, "loss": 0.4921, "step": 23370 }, { "epoch": 1.9351100811123985, "grad_norm": 4.2656989097595215, "learning_rate": 3.3720808160591565e-06, "loss": 0.5065, "step": 23380 }, { "epoch": 1.9359377586492301, "grad_norm": 4.852078437805176, "learning_rate": 3.367528649886738e-06, "loss": 0.5111, "step": 23390 }, { "epoch": 1.9367654361860618, "grad_norm": 5.908005714416504, "learning_rate": 3.3629779978010026e-06, "loss": 0.4548, "step": 23400 }, { "epoch": 1.9375931137228934, "grad_norm": 5.623996257781982, "learning_rate": 3.3584288640225925e-06, "loss": 0.463, "step": 23410 }, { "epoch": 1.938420791259725, "grad_norm": 3.963123083114624, "learning_rate": 3.3538812527707474e-06, "loss": 0.407, "step": 23420 }, { "epoch": 1.9392484687965568, "grad_norm": 4.919665813446045, "learning_rate": 3.3493351682632904e-06, "loss": 0.5078, "step": 23430 }, { "epoch": 1.9400761463333884, "grad_norm": 3.252660036087036, "learning_rate": 3.3447906147166278e-06, "loss": 0.392, "step": 23440 }, { "epoch": 1.94090382387022, "grad_norm": 5.767667293548584, "learning_rate": 3.3402475963457526e-06, "loss": 0.4446, "step": 23450 }, { "epoch": 1.9417315014070518, "grad_norm": 4.845109939575195, "learning_rate": 3.335706117364227e-06, "loss": 0.4518, "step": 23460 }, { "epoch": 1.9425591789438834, "grad_norm": 5.0321550369262695, "learning_rate": 3.3311661819841866e-06, "loss": 0.436, "step": 23470 }, { "epoch": 1.943386856480715, "grad_norm": 4.3272318840026855, "learning_rate": 3.3266277944163407e-06, "loss": 0.4018, "step": 23480 }, { "epoch": 1.9442145340175467, "grad_norm": 5.073896408081055, "learning_rate": 3.3220909588699553e-06, "loss": 0.4944, "step": 23490 }, { "epoch": 1.9450422115543784, "grad_norm": 4.104284763336182, "learning_rate": 3.3175556795528597e-06, "loss": 0.4674, "step": 23500 }, { "epoch": 1.94586988909121, "grad_norm": 6.392711639404297, "learning_rate": 3.3130219606714455e-06, "loss": 0.519, "step": 23510 }, { "epoch": 1.9466975666280417, "grad_norm": 4.639893054962158, "learning_rate": 3.308489806430648e-06, "loss": 0.4771, "step": 23520 }, { "epoch": 1.9475252441648734, "grad_norm": 4.032243251800537, "learning_rate": 3.3039592210339586e-06, "loss": 0.4683, "step": 23530 }, { "epoch": 1.948352921701705, "grad_norm": 5.333529472351074, "learning_rate": 3.2994302086834095e-06, "loss": 0.4696, "step": 23540 }, { "epoch": 1.9491805992385367, "grad_norm": 3.7379071712493896, "learning_rate": 3.2949027735795723e-06, "loss": 0.4515, "step": 23550 }, { "epoch": 1.9500082767753684, "grad_norm": 4.954298973083496, "learning_rate": 3.290376919921563e-06, "loss": 0.4444, "step": 23560 }, { "epoch": 1.9508359543122, "grad_norm": 4.890693187713623, "learning_rate": 3.2858526519070245e-06, "loss": 0.4869, "step": 23570 }, { "epoch": 1.9516636318490317, "grad_norm": 4.121739864349365, "learning_rate": 3.2813299737321296e-06, "loss": 0.4725, "step": 23580 }, { "epoch": 1.9524913093858633, "grad_norm": 4.627659320831299, "learning_rate": 3.27680888959158e-06, "loss": 0.5052, "step": 23590 }, { "epoch": 1.953318986922695, "grad_norm": 4.186211109161377, "learning_rate": 3.2722894036785956e-06, "loss": 0.499, "step": 23600 }, { "epoch": 1.9541466644595267, "grad_norm": 4.375804901123047, "learning_rate": 3.267771520184913e-06, "loss": 0.4549, "step": 23610 }, { "epoch": 1.9549743419963583, "grad_norm": 3.280574083328247, "learning_rate": 3.263255243300789e-06, "loss": 0.462, "step": 23620 }, { "epoch": 1.95580201953319, "grad_norm": 6.21018648147583, "learning_rate": 3.258740577214985e-06, "loss": 0.4801, "step": 23630 }, { "epoch": 1.9566296970700217, "grad_norm": 4.478578090667725, "learning_rate": 3.254227526114766e-06, "loss": 0.512, "step": 23640 }, { "epoch": 1.9574573746068533, "grad_norm": 3.748800754547119, "learning_rate": 3.2497160941859074e-06, "loss": 0.5032, "step": 23650 }, { "epoch": 1.958285052143685, "grad_norm": 4.438237190246582, "learning_rate": 3.2452062856126747e-06, "loss": 0.4612, "step": 23660 }, { "epoch": 1.9591127296805164, "grad_norm": 4.704222202301025, "learning_rate": 3.2406981045778307e-06, "loss": 0.5149, "step": 23670 }, { "epoch": 1.959940407217348, "grad_norm": 4.205870628356934, "learning_rate": 3.2361915552626323e-06, "loss": 0.4974, "step": 23680 }, { "epoch": 1.9607680847541797, "grad_norm": 7.513729572296143, "learning_rate": 3.231686641846817e-06, "loss": 0.4645, "step": 23690 }, { "epoch": 1.9615957622910114, "grad_norm": 4.5418572425842285, "learning_rate": 3.2271833685086094e-06, "loss": 0.4521, "step": 23700 }, { "epoch": 1.962423439827843, "grad_norm": 5.082376003265381, "learning_rate": 3.2226817394247116e-06, "loss": 0.4644, "step": 23710 }, { "epoch": 1.9632511173646747, "grad_norm": 4.976138591766357, "learning_rate": 3.2181817587702994e-06, "loss": 0.4814, "step": 23720 }, { "epoch": 1.9640787949015064, "grad_norm": 5.797845840454102, "learning_rate": 3.2136834307190197e-06, "loss": 0.5787, "step": 23730 }, { "epoch": 1.964906472438338, "grad_norm": 3.6008169651031494, "learning_rate": 3.2091867594429902e-06, "loss": 0.4923, "step": 23740 }, { "epoch": 1.9657341499751697, "grad_norm": 4.048726558685303, "learning_rate": 3.2046917491127884e-06, "loss": 0.5647, "step": 23750 }, { "epoch": 1.9665618275120014, "grad_norm": 6.694769859313965, "learning_rate": 3.200198403897453e-06, "loss": 0.4835, "step": 23760 }, { "epoch": 1.9673895050488328, "grad_norm": 3.41642427444458, "learning_rate": 3.195706727964477e-06, "loss": 0.4247, "step": 23770 }, { "epoch": 1.9682171825856645, "grad_norm": 4.645669460296631, "learning_rate": 3.191216725479807e-06, "loss": 0.4148, "step": 23780 }, { "epoch": 1.9690448601224961, "grad_norm": 3.8225669860839844, "learning_rate": 3.186728400607833e-06, "loss": 0.4472, "step": 23790 }, { "epoch": 1.9698725376593278, "grad_norm": 6.727963447570801, "learning_rate": 3.182241757511396e-06, "loss": 0.4741, "step": 23800 }, { "epoch": 1.9707002151961595, "grad_norm": 5.915018558502197, "learning_rate": 3.1777568003517744e-06, "loss": 0.503, "step": 23810 }, { "epoch": 1.9715278927329911, "grad_norm": 4.628719806671143, "learning_rate": 3.17327353328868e-06, "loss": 0.4651, "step": 23820 }, { "epoch": 1.9723555702698228, "grad_norm": 4.544474124908447, "learning_rate": 3.1687919604802597e-06, "loss": 0.4886, "step": 23830 }, { "epoch": 1.9731832478066544, "grad_norm": 3.8799145221710205, "learning_rate": 3.164312086083088e-06, "loss": 0.5219, "step": 23840 }, { "epoch": 1.974010925343486, "grad_norm": 4.080442905426025, "learning_rate": 3.159833914252165e-06, "loss": 0.4331, "step": 23850 }, { "epoch": 1.9748386028803178, "grad_norm": 4.781988143920898, "learning_rate": 3.1553574491409124e-06, "loss": 0.494, "step": 23860 }, { "epoch": 1.9756662804171494, "grad_norm": 3.357755661010742, "learning_rate": 3.150882694901168e-06, "loss": 0.5092, "step": 23870 }, { "epoch": 1.976493957953981, "grad_norm": 5.072999477386475, "learning_rate": 3.146409655683182e-06, "loss": 0.524, "step": 23880 }, { "epoch": 1.9773216354908127, "grad_norm": 3.176198720932007, "learning_rate": 3.141938335635615e-06, "loss": 0.5135, "step": 23890 }, { "epoch": 1.9781493130276444, "grad_norm": 4.3313727378845215, "learning_rate": 3.1374687389055315e-06, "loss": 0.5105, "step": 23900 }, { "epoch": 1.978976990564476, "grad_norm": 4.122094631195068, "learning_rate": 3.1330008696383994e-06, "loss": 0.4384, "step": 23910 }, { "epoch": 1.9798046681013077, "grad_norm": 5.698936462402344, "learning_rate": 3.1285347319780855e-06, "loss": 0.3625, "step": 23920 }, { "epoch": 1.9806323456381394, "grad_norm": 4.970904350280762, "learning_rate": 3.1240703300668496e-06, "loss": 0.4689, "step": 23930 }, { "epoch": 1.981460023174971, "grad_norm": 5.049126625061035, "learning_rate": 3.119607668045339e-06, "loss": 0.4458, "step": 23940 }, { "epoch": 1.9822877007118027, "grad_norm": 5.677995681762695, "learning_rate": 3.115146750052591e-06, "loss": 0.4302, "step": 23950 }, { "epoch": 1.9831153782486344, "grad_norm": 5.640052795410156, "learning_rate": 3.110687580226021e-06, "loss": 0.5313, "step": 23960 }, { "epoch": 1.983943055785466, "grad_norm": 5.955263614654541, "learning_rate": 3.1062301627014267e-06, "loss": 0.4977, "step": 23970 }, { "epoch": 1.9847707333222977, "grad_norm": 5.615230083465576, "learning_rate": 3.101774501612983e-06, "loss": 0.4936, "step": 23980 }, { "epoch": 1.9855984108591294, "grad_norm": 6.3071208000183105, "learning_rate": 3.0973206010932287e-06, "loss": 0.4444, "step": 23990 }, { "epoch": 1.986426088395961, "grad_norm": 5.383240699768066, "learning_rate": 3.0928684652730757e-06, "loss": 0.5183, "step": 24000 }, { "epoch": 1.9872537659327927, "grad_norm": 5.326765060424805, "learning_rate": 3.0884180982817936e-06, "loss": 0.5175, "step": 24010 }, { "epoch": 1.9880814434696243, "grad_norm": 4.1490159034729, "learning_rate": 3.0839695042470176e-06, "loss": 0.4446, "step": 24020 }, { "epoch": 1.988909121006456, "grad_norm": 4.3488054275512695, "learning_rate": 3.079522687294732e-06, "loss": 0.4257, "step": 24030 }, { "epoch": 1.9897367985432877, "grad_norm": 3.467595338821411, "learning_rate": 3.075077651549282e-06, "loss": 0.4664, "step": 24040 }, { "epoch": 1.9905644760801193, "grad_norm": 4.836792469024658, "learning_rate": 3.070634401133351e-06, "loss": 0.472, "step": 24050 }, { "epoch": 1.991392153616951, "grad_norm": 2.9180548191070557, "learning_rate": 3.066192940167973e-06, "loss": 0.4259, "step": 24060 }, { "epoch": 1.9922198311537826, "grad_norm": 3.8061983585357666, "learning_rate": 3.0617532727725173e-06, "loss": 0.4791, "step": 24070 }, { "epoch": 1.993047508690614, "grad_norm": 4.312309741973877, "learning_rate": 3.0573154030646955e-06, "loss": 0.4872, "step": 24080 }, { "epoch": 1.9938751862274458, "grad_norm": 4.456234931945801, "learning_rate": 3.0528793351605455e-06, "loss": 0.4246, "step": 24090 }, { "epoch": 1.9947028637642774, "grad_norm": 4.520194053649902, "learning_rate": 3.048445073174442e-06, "loss": 0.5136, "step": 24100 }, { "epoch": 1.995530541301109, "grad_norm": 5.626556873321533, "learning_rate": 3.044012621219077e-06, "loss": 0.5285, "step": 24110 }, { "epoch": 1.9963582188379407, "grad_norm": 4.750120162963867, "learning_rate": 3.0395819834054663e-06, "loss": 0.4339, "step": 24120 }, { "epoch": 1.9971858963747724, "grad_norm": 5.290566921234131, "learning_rate": 3.035153163842946e-06, "loss": 0.5059, "step": 24130 }, { "epoch": 1.998013573911604, "grad_norm": 5.857840061187744, "learning_rate": 3.030726166639163e-06, "loss": 0.4965, "step": 24140 }, { "epoch": 1.9988412514484357, "grad_norm": 5.022451400756836, "learning_rate": 3.0263009959000712e-06, "loss": 0.5076, "step": 24150 }, { "epoch": 1.9996689289852674, "grad_norm": 4.579245567321777, "learning_rate": 3.021877655729939e-06, "loss": 0.5256, "step": 24160 }, { "epoch": 2.000496606522099, "grad_norm": 3.3500888347625732, "learning_rate": 3.017456150231329e-06, "loss": 0.3969, "step": 24170 }, { "epoch": 2.0013242840589305, "grad_norm": 4.254011631011963, "learning_rate": 3.013036483505104e-06, "loss": 0.4303, "step": 24180 }, { "epoch": 2.002151961595762, "grad_norm": 3.871917247772217, "learning_rate": 3.008618659650425e-06, "loss": 0.3933, "step": 24190 }, { "epoch": 2.002979639132594, "grad_norm": 4.940983772277832, "learning_rate": 3.0042026827647396e-06, "loss": 0.3782, "step": 24200 }, { "epoch": 2.0038073166694255, "grad_norm": 5.401775360107422, "learning_rate": 2.9997885569437824e-06, "loss": 0.3501, "step": 24210 }, { "epoch": 2.004634994206257, "grad_norm": 2.451137065887451, "learning_rate": 2.995376286281576e-06, "loss": 0.383, "step": 24220 }, { "epoch": 2.005462671743089, "grad_norm": 4.324498176574707, "learning_rate": 2.990965874870417e-06, "loss": 0.4557, "step": 24230 }, { "epoch": 2.0062903492799204, "grad_norm": 4.54441499710083, "learning_rate": 2.9865573268008785e-06, "loss": 0.4173, "step": 24240 }, { "epoch": 2.007118026816752, "grad_norm": 4.354808807373047, "learning_rate": 2.9821506461618095e-06, "loss": 0.4086, "step": 24250 }, { "epoch": 2.0079457043535838, "grad_norm": 4.070717811584473, "learning_rate": 2.97774583704032e-06, "loss": 0.4146, "step": 24260 }, { "epoch": 2.0087733818904154, "grad_norm": 4.818182468414307, "learning_rate": 2.973342903521792e-06, "loss": 0.4069, "step": 24270 }, { "epoch": 2.009601059427247, "grad_norm": 4.40711784362793, "learning_rate": 2.9689418496898627e-06, "loss": 0.4567, "step": 24280 }, { "epoch": 2.0104287369640788, "grad_norm": 4.2970871925354, "learning_rate": 2.9645426796264267e-06, "loss": 0.3817, "step": 24290 }, { "epoch": 2.0112564145009104, "grad_norm": 3.2909488677978516, "learning_rate": 2.9601453974116335e-06, "loss": 0.3706, "step": 24300 }, { "epoch": 2.012084092037742, "grad_norm": 4.302274227142334, "learning_rate": 2.9557500071238787e-06, "loss": 0.4399, "step": 24310 }, { "epoch": 2.0129117695745737, "grad_norm": 3.9358479976654053, "learning_rate": 2.951356512839803e-06, "loss": 0.4026, "step": 24320 }, { "epoch": 2.0137394471114054, "grad_norm": 4.5143561363220215, "learning_rate": 2.946964918634295e-06, "loss": 0.3872, "step": 24330 }, { "epoch": 2.014567124648237, "grad_norm": 3.7798380851745605, "learning_rate": 2.9425752285804733e-06, "loss": 0.3879, "step": 24340 }, { "epoch": 2.0153948021850687, "grad_norm": 4.670631408691406, "learning_rate": 2.9381874467496933e-06, "loss": 0.3977, "step": 24350 }, { "epoch": 2.0162224797219004, "grad_norm": 4.920087814331055, "learning_rate": 2.9338015772115425e-06, "loss": 0.4375, "step": 24360 }, { "epoch": 2.017050157258732, "grad_norm": 3.6136631965637207, "learning_rate": 2.929417624033831e-06, "loss": 0.3924, "step": 24370 }, { "epoch": 2.0178778347955637, "grad_norm": 4.216578006744385, "learning_rate": 2.9250355912825935e-06, "loss": 0.4552, "step": 24380 }, { "epoch": 2.0187055123323954, "grad_norm": 5.152340888977051, "learning_rate": 2.9206554830220845e-06, "loss": 0.4894, "step": 24390 }, { "epoch": 2.019533189869227, "grad_norm": 3.941826343536377, "learning_rate": 2.916277303314773e-06, "loss": 0.4515, "step": 24400 }, { "epoch": 2.0203608674060587, "grad_norm": 4.486544132232666, "learning_rate": 2.911901056221339e-06, "loss": 0.4064, "step": 24410 }, { "epoch": 2.0211885449428904, "grad_norm": 4.138355255126953, "learning_rate": 2.9075267458006677e-06, "loss": 0.4546, "step": 24420 }, { "epoch": 2.022016222479722, "grad_norm": 3.9021685123443604, "learning_rate": 2.903154376109852e-06, "loss": 0.4005, "step": 24430 }, { "epoch": 2.0228439000165537, "grad_norm": 3.8971877098083496, "learning_rate": 2.898783951204183e-06, "loss": 0.398, "step": 24440 }, { "epoch": 2.0236715775533853, "grad_norm": 5.266237258911133, "learning_rate": 2.8944154751371466e-06, "loss": 0.3824, "step": 24450 }, { "epoch": 2.024499255090217, "grad_norm": 4.013658046722412, "learning_rate": 2.8900489519604274e-06, "loss": 0.3845, "step": 24460 }, { "epoch": 2.0253269326270487, "grad_norm": 4.7401885986328125, "learning_rate": 2.8856843857238885e-06, "loss": 0.4797, "step": 24470 }, { "epoch": 2.0261546101638803, "grad_norm": 4.384000301361084, "learning_rate": 2.8813217804755866e-06, "loss": 0.4023, "step": 24480 }, { "epoch": 2.026982287700712, "grad_norm": 4.017022132873535, "learning_rate": 2.876961140261757e-06, "loss": 0.4167, "step": 24490 }, { "epoch": 2.0278099652375436, "grad_norm": 4.124701023101807, "learning_rate": 2.8726024691268074e-06, "loss": 0.4022, "step": 24500 }, { "epoch": 2.0286376427743753, "grad_norm": 5.854796886444092, "learning_rate": 2.868245771113331e-06, "loss": 0.4013, "step": 24510 }, { "epoch": 2.029465320311207, "grad_norm": 4.562315940856934, "learning_rate": 2.8638910502620785e-06, "loss": 0.3996, "step": 24520 }, { "epoch": 2.030292997848038, "grad_norm": 4.465712070465088, "learning_rate": 2.8595383106119755e-06, "loss": 0.3956, "step": 24530 }, { "epoch": 2.03112067538487, "grad_norm": 5.629533290863037, "learning_rate": 2.8551875562001043e-06, "loss": 0.4824, "step": 24540 }, { "epoch": 2.0319483529217015, "grad_norm": 4.194096088409424, "learning_rate": 2.8508387910617073e-06, "loss": 0.4141, "step": 24550 }, { "epoch": 2.032776030458533, "grad_norm": 5.5547919273376465, "learning_rate": 2.846492019230186e-06, "loss": 0.4576, "step": 24560 }, { "epoch": 2.033603707995365, "grad_norm": 5.341403007507324, "learning_rate": 2.8421472447370878e-06, "loss": 0.4227, "step": 24570 }, { "epoch": 2.0344313855321965, "grad_norm": 3.789414644241333, "learning_rate": 2.837804471612114e-06, "loss": 0.3794, "step": 24580 }, { "epoch": 2.035259063069028, "grad_norm": 3.9165499210357666, "learning_rate": 2.8334637038831e-06, "loss": 0.4269, "step": 24590 }, { "epoch": 2.03608674060586, "grad_norm": 4.005009174346924, "learning_rate": 2.829124945576029e-06, "loss": 0.4051, "step": 24600 }, { "epoch": 2.0369144181426915, "grad_norm": 4.589481353759766, "learning_rate": 2.8247882007150208e-06, "loss": 0.4177, "step": 24610 }, { "epoch": 2.037742095679523, "grad_norm": 4.231782913208008, "learning_rate": 2.820453473322319e-06, "loss": 0.4611, "step": 24620 }, { "epoch": 2.038569773216355, "grad_norm": 7.108029842376709, "learning_rate": 2.81612076741831e-06, "loss": 0.4175, "step": 24630 }, { "epoch": 2.0393974507531865, "grad_norm": 3.6660125255584717, "learning_rate": 2.8117900870214914e-06, "loss": 0.4162, "step": 24640 }, { "epoch": 2.040225128290018, "grad_norm": 3.922412872314453, "learning_rate": 2.8074614361484913e-06, "loss": 0.4273, "step": 24650 }, { "epoch": 2.04105280582685, "grad_norm": 3.8914589881896973, "learning_rate": 2.803134818814054e-06, "loss": 0.4441, "step": 24660 }, { "epoch": 2.0418804833636814, "grad_norm": 10.968942642211914, "learning_rate": 2.7988102390310325e-06, "loss": 0.3606, "step": 24670 }, { "epoch": 2.042708160900513, "grad_norm": 3.834630012512207, "learning_rate": 2.7944877008103953e-06, "loss": 0.3554, "step": 24680 }, { "epoch": 2.0435358384373448, "grad_norm": 3.9626221656799316, "learning_rate": 2.7901672081612162e-06, "loss": 0.3967, "step": 24690 }, { "epoch": 2.0443635159741764, "grad_norm": 3.95365571975708, "learning_rate": 2.785848765090671e-06, "loss": 0.4205, "step": 24700 }, { "epoch": 2.045191193511008, "grad_norm": 4.5237531661987305, "learning_rate": 2.7815323756040376e-06, "loss": 0.407, "step": 24710 }, { "epoch": 2.0460188710478397, "grad_norm": 1.9657940864562988, "learning_rate": 2.777218043704682e-06, "loss": 0.4094, "step": 24720 }, { "epoch": 2.0468465485846714, "grad_norm": 5.4810471534729, "learning_rate": 2.7729057733940713e-06, "loss": 0.4056, "step": 24730 }, { "epoch": 2.047674226121503, "grad_norm": 3.6740331649780273, "learning_rate": 2.7685955686717484e-06, "loss": 0.4335, "step": 24740 }, { "epoch": 2.0485019036583347, "grad_norm": 4.41782808303833, "learning_rate": 2.7642874335353566e-06, "loss": 0.3779, "step": 24750 }, { "epoch": 2.0493295811951664, "grad_norm": 5.061495304107666, "learning_rate": 2.7599813719806047e-06, "loss": 0.3842, "step": 24760 }, { "epoch": 2.050157258731998, "grad_norm": 4.553317070007324, "learning_rate": 2.7556773880012865e-06, "loss": 0.4127, "step": 24770 }, { "epoch": 2.0509849362688297, "grad_norm": 6.9860405921936035, "learning_rate": 2.751375485589268e-06, "loss": 0.4364, "step": 24780 }, { "epoch": 2.0518126138056614, "grad_norm": 4.116398334503174, "learning_rate": 2.7470756687344802e-06, "loss": 0.4481, "step": 24790 }, { "epoch": 2.052640291342493, "grad_norm": 3.740372657775879, "learning_rate": 2.742777941424924e-06, "loss": 0.3864, "step": 24800 }, { "epoch": 2.0534679688793247, "grad_norm": 6.213690757751465, "learning_rate": 2.7384823076466614e-06, "loss": 0.4063, "step": 24810 }, { "epoch": 2.0542956464161564, "grad_norm": 4.339824199676514, "learning_rate": 2.7341887713838135e-06, "loss": 0.407, "step": 24820 }, { "epoch": 2.055123323952988, "grad_norm": 4.946688652038574, "learning_rate": 2.7298973366185555e-06, "loss": 0.4363, "step": 24830 }, { "epoch": 2.0559510014898197, "grad_norm": 4.627228736877441, "learning_rate": 2.725608007331111e-06, "loss": 0.434, "step": 24840 }, { "epoch": 2.0567786790266513, "grad_norm": 5.038379192352295, "learning_rate": 2.721320787499754e-06, "loss": 0.3529, "step": 24850 }, { "epoch": 2.057606356563483, "grad_norm": 6.287378311157227, "learning_rate": 2.717035681100797e-06, "loss": 0.4445, "step": 24860 }, { "epoch": 2.0584340341003147, "grad_norm": 4.331061840057373, "learning_rate": 2.712752692108601e-06, "loss": 0.4522, "step": 24870 }, { "epoch": 2.0592617116371463, "grad_norm": 4.241365909576416, "learning_rate": 2.70847182449556e-06, "loss": 0.3969, "step": 24880 }, { "epoch": 2.060089389173978, "grad_norm": 4.001123428344727, "learning_rate": 2.704193082232092e-06, "loss": 0.4312, "step": 24890 }, { "epoch": 2.0609170667108097, "grad_norm": 4.470404624938965, "learning_rate": 2.699916469286657e-06, "loss": 0.3978, "step": 24900 }, { "epoch": 2.0617447442476413, "grad_norm": 4.328632354736328, "learning_rate": 2.695641989625728e-06, "loss": 0.4064, "step": 24910 }, { "epoch": 2.062572421784473, "grad_norm": 5.030815601348877, "learning_rate": 2.691369647213808e-06, "loss": 0.423, "step": 24920 }, { "epoch": 2.0634000993213046, "grad_norm": 4.620389938354492, "learning_rate": 2.687099446013414e-06, "loss": 0.3625, "step": 24930 }, { "epoch": 2.064227776858136, "grad_norm": 4.644993305206299, "learning_rate": 2.6828313899850788e-06, "loss": 0.4133, "step": 24940 }, { "epoch": 2.0650554543949675, "grad_norm": 3.1251823902130127, "learning_rate": 2.678565483087346e-06, "loss": 0.3857, "step": 24950 }, { "epoch": 2.065883131931799, "grad_norm": 4.438612937927246, "learning_rate": 2.67430172927676e-06, "loss": 0.4345, "step": 24960 }, { "epoch": 2.066710809468631, "grad_norm": 3.0464046001434326, "learning_rate": 2.670040132507876e-06, "loss": 0.3402, "step": 24970 }, { "epoch": 2.0675384870054625, "grad_norm": 4.895553112030029, "learning_rate": 2.6657806967332457e-06, "loss": 0.4383, "step": 24980 }, { "epoch": 2.068366164542294, "grad_norm": 3.0672926902770996, "learning_rate": 2.6615234259034144e-06, "loss": 0.3845, "step": 24990 }, { "epoch": 2.069193842079126, "grad_norm": 6.276889324188232, "learning_rate": 2.6572683239669245e-06, "loss": 0.484, "step": 25000 }, { "epoch": 2.0700215196159575, "grad_norm": 4.584026336669922, "learning_rate": 2.6530153948702997e-06, "loss": 0.418, "step": 25010 }, { "epoch": 2.070849197152789, "grad_norm": 5.015149116516113, "learning_rate": 2.648764642558053e-06, "loss": 0.3637, "step": 25020 }, { "epoch": 2.071676874689621, "grad_norm": 4.468961715698242, "learning_rate": 2.644516070972679e-06, "loss": 0.4086, "step": 25030 }, { "epoch": 2.0725045522264525, "grad_norm": 4.727617263793945, "learning_rate": 2.640269684054649e-06, "loss": 0.3969, "step": 25040 }, { "epoch": 2.073332229763284, "grad_norm": 5.124699592590332, "learning_rate": 2.6360254857424077e-06, "loss": 0.3918, "step": 25050 }, { "epoch": 2.074159907300116, "grad_norm": 4.145781993865967, "learning_rate": 2.6317834799723674e-06, "loss": 0.3274, "step": 25060 }, { "epoch": 2.0749875848369475, "grad_norm": 5.10432767868042, "learning_rate": 2.6275436706789126e-06, "loss": 0.4144, "step": 25070 }, { "epoch": 2.075815262373779, "grad_norm": 2.3923914432525635, "learning_rate": 2.623306061794383e-06, "loss": 0.4231, "step": 25080 }, { "epoch": 2.0766429399106108, "grad_norm": 4.757331848144531, "learning_rate": 2.619070657249083e-06, "loss": 0.439, "step": 25090 }, { "epoch": 2.0774706174474424, "grad_norm": 3.976261615753174, "learning_rate": 2.6148374609712735e-06, "loss": 0.274, "step": 25100 }, { "epoch": 2.078298294984274, "grad_norm": 4.369542121887207, "learning_rate": 2.6106064768871626e-06, "loss": 0.4027, "step": 25110 }, { "epoch": 2.0791259725211058, "grad_norm": 4.941155433654785, "learning_rate": 2.6063777089209118e-06, "loss": 0.3704, "step": 25120 }, { "epoch": 2.0799536500579374, "grad_norm": 3.301645278930664, "learning_rate": 2.602151160994621e-06, "loss": 0.4943, "step": 25130 }, { "epoch": 2.080781327594769, "grad_norm": 6.7457685470581055, "learning_rate": 2.597926837028336e-06, "loss": 0.3579, "step": 25140 }, { "epoch": 2.0816090051316007, "grad_norm": 5.0859150886535645, "learning_rate": 2.5937047409400385e-06, "loss": 0.386, "step": 25150 }, { "epoch": 2.0824366826684324, "grad_norm": 4.758631229400635, "learning_rate": 2.5894848766456425e-06, "loss": 0.3627, "step": 25160 }, { "epoch": 2.083264360205264, "grad_norm": 5.840559005737305, "learning_rate": 2.5852672480589978e-06, "loss": 0.3631, "step": 25170 }, { "epoch": 2.0840920377420957, "grad_norm": 4.705920219421387, "learning_rate": 2.5810518590918697e-06, "loss": 0.4006, "step": 25180 }, { "epoch": 2.0849197152789274, "grad_norm": 5.084292411804199, "learning_rate": 2.576838713653956e-06, "loss": 0.4075, "step": 25190 }, { "epoch": 2.085747392815759, "grad_norm": 3.8470757007598877, "learning_rate": 2.5726278156528726e-06, "loss": 0.3328, "step": 25200 }, { "epoch": 2.0865750703525907, "grad_norm": 5.140650749206543, "learning_rate": 2.568419168994142e-06, "loss": 0.37, "step": 25210 }, { "epoch": 2.0874027478894224, "grad_norm": 5.633903980255127, "learning_rate": 2.564212777581214e-06, "loss": 0.4619, "step": 25220 }, { "epoch": 2.088230425426254, "grad_norm": 5.085748672485352, "learning_rate": 2.5600086453154316e-06, "loss": 0.4077, "step": 25230 }, { "epoch": 2.0890581029630857, "grad_norm": 5.004804611206055, "learning_rate": 2.5558067760960524e-06, "loss": 0.3602, "step": 25240 }, { "epoch": 2.0898857804999174, "grad_norm": 3.418703317642212, "learning_rate": 2.5516071738202285e-06, "loss": 0.4324, "step": 25250 }, { "epoch": 2.090713458036749, "grad_norm": 3.077589511871338, "learning_rate": 2.5474098423830146e-06, "loss": 0.4112, "step": 25260 }, { "epoch": 2.0915411355735807, "grad_norm": 4.753369331359863, "learning_rate": 2.5432147856773572e-06, "loss": 0.4223, "step": 25270 }, { "epoch": 2.0923688131104123, "grad_norm": 4.08544397354126, "learning_rate": 2.539022007594092e-06, "loss": 0.4121, "step": 25280 }, { "epoch": 2.093196490647244, "grad_norm": 5.633303165435791, "learning_rate": 2.534831512021945e-06, "loss": 0.4778, "step": 25290 }, { "epoch": 2.0940241681840757, "grad_norm": 5.4304351806640625, "learning_rate": 2.5306433028475197e-06, "loss": 0.3075, "step": 25300 }, { "epoch": 2.0948518457209073, "grad_norm": 5.162240505218506, "learning_rate": 2.526457383955302e-06, "loss": 0.4571, "step": 25310 }, { "epoch": 2.095679523257739, "grad_norm": 4.172022819519043, "learning_rate": 2.522273759227656e-06, "loss": 0.4157, "step": 25320 }, { "epoch": 2.0965072007945706, "grad_norm": 4.472196578979492, "learning_rate": 2.518092432544811e-06, "loss": 0.3961, "step": 25330 }, { "epoch": 2.0973348783314023, "grad_norm": 3.5140254497528076, "learning_rate": 2.513913407784875e-06, "loss": 0.4093, "step": 25340 }, { "epoch": 2.0981625558682335, "grad_norm": 3.8516218662261963, "learning_rate": 2.5097366888238107e-06, "loss": 0.4143, "step": 25350 }, { "epoch": 2.098990233405065, "grad_norm": 5.105112552642822, "learning_rate": 2.5055622795354496e-06, "loss": 0.3873, "step": 25360 }, { "epoch": 2.099817910941897, "grad_norm": 4.999550819396973, "learning_rate": 2.501390183791479e-06, "loss": 0.4275, "step": 25370 }, { "epoch": 2.1006455884787285, "grad_norm": 6.195010662078857, "learning_rate": 2.497220405461436e-06, "loss": 0.348, "step": 25380 }, { "epoch": 2.10147326601556, "grad_norm": 7.071352958679199, "learning_rate": 2.4930529484127146e-06, "loss": 0.4495, "step": 25390 }, { "epoch": 2.102300943552392, "grad_norm": 3.7321019172668457, "learning_rate": 2.4888878165105536e-06, "loss": 0.4009, "step": 25400 }, { "epoch": 2.1031286210892235, "grad_norm": 3.8883514404296875, "learning_rate": 2.484725013618036e-06, "loss": 0.451, "step": 25410 }, { "epoch": 2.103956298626055, "grad_norm": 4.960960388183594, "learning_rate": 2.4805645435960806e-06, "loss": 0.4552, "step": 25420 }, { "epoch": 2.104783976162887, "grad_norm": 4.873196601867676, "learning_rate": 2.4764064103034473e-06, "loss": 0.4634, "step": 25430 }, { "epoch": 2.1056116536997185, "grad_norm": 4.635401725769043, "learning_rate": 2.472250617596729e-06, "loss": 0.4278, "step": 25440 }, { "epoch": 2.10643933123655, "grad_norm": 4.5227508544921875, "learning_rate": 2.4680971693303402e-06, "loss": 0.3887, "step": 25450 }, { "epoch": 2.107267008773382, "grad_norm": 4.251188278198242, "learning_rate": 2.4639460693565343e-06, "loss": 0.3841, "step": 25460 }, { "epoch": 2.1080946863102135, "grad_norm": 4.205645561218262, "learning_rate": 2.4597973215253727e-06, "loss": 0.4412, "step": 25470 }, { "epoch": 2.108922363847045, "grad_norm": 4.134787082672119, "learning_rate": 2.455650929684744e-06, "loss": 0.4389, "step": 25480 }, { "epoch": 2.109750041383877, "grad_norm": 4.357507228851318, "learning_rate": 2.451506897680351e-06, "loss": 0.4159, "step": 25490 }, { "epoch": 2.1105777189207084, "grad_norm": 4.429449081420898, "learning_rate": 2.4473652293557014e-06, "loss": 0.3874, "step": 25500 }, { "epoch": 2.11140539645754, "grad_norm": 4.901861190795898, "learning_rate": 2.4432259285521183e-06, "loss": 0.3884, "step": 25510 }, { "epoch": 2.1122330739943718, "grad_norm": 4.352156162261963, "learning_rate": 2.4390889991087252e-06, "loss": 0.417, "step": 25520 }, { "epoch": 2.1130607515312034, "grad_norm": 5.033200740814209, "learning_rate": 2.4349544448624467e-06, "loss": 0.382, "step": 25530 }, { "epoch": 2.113888429068035, "grad_norm": 3.9242870807647705, "learning_rate": 2.430822269648007e-06, "loss": 0.4304, "step": 25540 }, { "epoch": 2.1147161066048668, "grad_norm": 4.309840202331543, "learning_rate": 2.4266924772979173e-06, "loss": 0.3725, "step": 25550 }, { "epoch": 2.1155437841416984, "grad_norm": 5.294429779052734, "learning_rate": 2.422565071642487e-06, "loss": 0.4767, "step": 25560 }, { "epoch": 2.11637146167853, "grad_norm": 4.9906325340271, "learning_rate": 2.418440056509802e-06, "loss": 0.4539, "step": 25570 }, { "epoch": 2.1171991392153617, "grad_norm": 6.109132289886475, "learning_rate": 2.4143174357257422e-06, "loss": 0.3875, "step": 25580 }, { "epoch": 2.1180268167521934, "grad_norm": 4.528285026550293, "learning_rate": 2.4101972131139613e-06, "loss": 0.4076, "step": 25590 }, { "epoch": 2.118854494289025, "grad_norm": 4.171262741088867, "learning_rate": 2.406079392495885e-06, "loss": 0.3753, "step": 25600 }, { "epoch": 2.1196821718258567, "grad_norm": 4.027106285095215, "learning_rate": 2.4019639776907195e-06, "loss": 0.4538, "step": 25610 }, { "epoch": 2.1205098493626884, "grad_norm": 8.687530517578125, "learning_rate": 2.397850972515429e-06, "loss": 0.4698, "step": 25620 }, { "epoch": 2.12133752689952, "grad_norm": 3.7001123428344727, "learning_rate": 2.3937403807847557e-06, "loss": 0.3639, "step": 25630 }, { "epoch": 2.1221652044363517, "grad_norm": 5.658789157867432, "learning_rate": 2.389632206311193e-06, "loss": 0.5022, "step": 25640 }, { "epoch": 2.1229928819731834, "grad_norm": 5.723340034484863, "learning_rate": 2.3855264529049955e-06, "loss": 0.416, "step": 25650 }, { "epoch": 2.123820559510015, "grad_norm": 5.815126419067383, "learning_rate": 2.3814231243741766e-06, "loss": 0.3929, "step": 25660 }, { "epoch": 2.1246482370468467, "grad_norm": 4.367076396942139, "learning_rate": 2.3773222245244915e-06, "loss": 0.4062, "step": 25670 }, { "epoch": 2.1254759145836783, "grad_norm": 4.15652322769165, "learning_rate": 2.3732237571594515e-06, "loss": 0.4079, "step": 25680 }, { "epoch": 2.12630359212051, "grad_norm": 4.369402885437012, "learning_rate": 2.369127726080308e-06, "loss": 0.3605, "step": 25690 }, { "epoch": 2.1271312696573417, "grad_norm": 3.822674512863159, "learning_rate": 2.3650341350860527e-06, "loss": 0.4184, "step": 25700 }, { "epoch": 2.1279589471941733, "grad_norm": 5.2030816078186035, "learning_rate": 2.3609429879734174e-06, "loss": 0.3859, "step": 25710 }, { "epoch": 2.128786624731005, "grad_norm": 5.142391204833984, "learning_rate": 2.3568542885368614e-06, "loss": 0.3972, "step": 25720 }, { "epoch": 2.1296143022678367, "grad_norm": 4.153464317321777, "learning_rate": 2.3527680405685788e-06, "loss": 0.4073, "step": 25730 }, { "epoch": 2.130441979804668, "grad_norm": 4.695365905761719, "learning_rate": 2.3486842478584846e-06, "loss": 0.4434, "step": 25740 }, { "epoch": 2.1312696573415, "grad_norm": 5.580692768096924, "learning_rate": 2.3446029141942252e-06, "loss": 0.4123, "step": 25750 }, { "epoch": 2.132097334878331, "grad_norm": 4.813806533813477, "learning_rate": 2.340524043361161e-06, "loss": 0.4609, "step": 25760 }, { "epoch": 2.132925012415163, "grad_norm": 5.213347434997559, "learning_rate": 2.336447639142365e-06, "loss": 0.3645, "step": 25770 }, { "epoch": 2.1337526899519945, "grad_norm": 5.932922840118408, "learning_rate": 2.33237370531863e-06, "loss": 0.4534, "step": 25780 }, { "epoch": 2.134580367488826, "grad_norm": 4.42698335647583, "learning_rate": 2.3283022456684494e-06, "loss": 0.3978, "step": 25790 }, { "epoch": 2.135408045025658, "grad_norm": 3.4329817295074463, "learning_rate": 2.324233263968028e-06, "loss": 0.3751, "step": 25800 }, { "epoch": 2.1362357225624895, "grad_norm": 4.2323784828186035, "learning_rate": 2.3201667639912704e-06, "loss": 0.4452, "step": 25810 }, { "epoch": 2.137063400099321, "grad_norm": 4.652828216552734, "learning_rate": 2.3161027495097794e-06, "loss": 0.3877, "step": 25820 }, { "epoch": 2.137891077636153, "grad_norm": 4.3284912109375, "learning_rate": 2.312041224292854e-06, "loss": 0.4206, "step": 25830 }, { "epoch": 2.1387187551729845, "grad_norm": 5.975958824157715, "learning_rate": 2.3079821921074795e-06, "loss": 0.4187, "step": 25840 }, { "epoch": 2.139546432709816, "grad_norm": 2.839904308319092, "learning_rate": 2.3039256567183343e-06, "loss": 0.3947, "step": 25850 }, { "epoch": 2.140374110246648, "grad_norm": 4.506594181060791, "learning_rate": 2.299871621887779e-06, "loss": 0.4087, "step": 25860 }, { "epoch": 2.1412017877834795, "grad_norm": 3.905850410461426, "learning_rate": 2.2958200913758555e-06, "loss": 0.3484, "step": 25870 }, { "epoch": 2.142029465320311, "grad_norm": 4.744638442993164, "learning_rate": 2.2917710689402844e-06, "loss": 0.4452, "step": 25880 }, { "epoch": 2.142857142857143, "grad_norm": 3.897965431213379, "learning_rate": 2.287724558336455e-06, "loss": 0.4208, "step": 25890 }, { "epoch": 2.1436848203939745, "grad_norm": 5.082087993621826, "learning_rate": 2.2836805633174313e-06, "loss": 0.4338, "step": 25900 }, { "epoch": 2.144512497930806, "grad_norm": 3.783059597015381, "learning_rate": 2.279639087633945e-06, "loss": 0.4616, "step": 25910 }, { "epoch": 2.1453401754676378, "grad_norm": 5.290283679962158, "learning_rate": 2.2756001350343853e-06, "loss": 0.4303, "step": 25920 }, { "epoch": 2.1461678530044694, "grad_norm": 6.499427318572998, "learning_rate": 2.271563709264811e-06, "loss": 0.4366, "step": 25930 }, { "epoch": 2.146995530541301, "grad_norm": 5.474023342132568, "learning_rate": 2.2675298140689267e-06, "loss": 0.4194, "step": 25940 }, { "epoch": 2.1478232080781328, "grad_norm": 4.777155876159668, "learning_rate": 2.2634984531880983e-06, "loss": 0.4821, "step": 25950 }, { "epoch": 2.1486508856149644, "grad_norm": 4.983157157897949, "learning_rate": 2.2594696303613344e-06, "loss": 0.4811, "step": 25960 }, { "epoch": 2.149478563151796, "grad_norm": 3.4889073371887207, "learning_rate": 2.2554433493252946e-06, "loss": 0.4563, "step": 25970 }, { "epoch": 2.1503062406886277, "grad_norm": 5.443056583404541, "learning_rate": 2.2514196138142795e-06, "loss": 0.377, "step": 25980 }, { "epoch": 2.1511339182254594, "grad_norm": 3.7812552452087402, "learning_rate": 2.2473984275602278e-06, "loss": 0.4122, "step": 25990 }, { "epoch": 2.151961595762291, "grad_norm": 4.683426380157471, "learning_rate": 2.243379794292718e-06, "loss": 0.4579, "step": 26000 }, { "epoch": 2.1527892732991227, "grad_norm": 4.949398040771484, "learning_rate": 2.239363717738952e-06, "loss": 0.4288, "step": 26010 }, { "epoch": 2.1536169508359544, "grad_norm": 4.7929534912109375, "learning_rate": 2.235350201623769e-06, "loss": 0.3707, "step": 26020 }, { "epoch": 2.154444628372786, "grad_norm": 4.559963226318359, "learning_rate": 2.2313392496696313e-06, "loss": 0.4109, "step": 26030 }, { "epoch": 2.1552723059096177, "grad_norm": 6.811455249786377, "learning_rate": 2.2273308655966164e-06, "loss": 0.3373, "step": 26040 }, { "epoch": 2.1560999834464494, "grad_norm": 4.981826305389404, "learning_rate": 2.2233250531224336e-06, "loss": 0.441, "step": 26050 }, { "epoch": 2.156927660983281, "grad_norm": 4.302591800689697, "learning_rate": 2.219321815962393e-06, "loss": 0.3937, "step": 26060 }, { "epoch": 2.1577553385201127, "grad_norm": 7.017866611480713, "learning_rate": 2.215321157829424e-06, "loss": 0.4026, "step": 26070 }, { "epoch": 2.1585830160569444, "grad_norm": 5.076597213745117, "learning_rate": 2.2113230824340654e-06, "loss": 0.3984, "step": 26080 }, { "epoch": 2.159410693593776, "grad_norm": 2.919853925704956, "learning_rate": 2.2073275934844525e-06, "loss": 0.4809, "step": 26090 }, { "epoch": 2.1602383711306077, "grad_norm": 5.019326210021973, "learning_rate": 2.2033346946863284e-06, "loss": 0.4828, "step": 26100 }, { "epoch": 2.1610660486674393, "grad_norm": 4.998690605163574, "learning_rate": 2.199344389743033e-06, "loss": 0.4528, "step": 26110 }, { "epoch": 2.161893726204271, "grad_norm": 4.153848648071289, "learning_rate": 2.1953566823555005e-06, "loss": 0.4275, "step": 26120 }, { "epoch": 2.1627214037411027, "grad_norm": 3.6960158348083496, "learning_rate": 2.1913715762222522e-06, "loss": 0.3232, "step": 26130 }, { "epoch": 2.1635490812779343, "grad_norm": 4.2864580154418945, "learning_rate": 2.1873890750394013e-06, "loss": 0.4356, "step": 26140 }, { "epoch": 2.1643767588147655, "grad_norm": 5.444047927856445, "learning_rate": 2.183409182500644e-06, "loss": 0.4151, "step": 26150 }, { "epoch": 2.1652044363515976, "grad_norm": 3.9595274925231934, "learning_rate": 2.179431902297253e-06, "loss": 0.4362, "step": 26160 }, { "epoch": 2.166032113888429, "grad_norm": 4.48524808883667, "learning_rate": 2.175457238118087e-06, "loss": 0.3917, "step": 26170 }, { "epoch": 2.1668597914252605, "grad_norm": 4.782451152801514, "learning_rate": 2.171485193649567e-06, "loss": 0.4096, "step": 26180 }, { "epoch": 2.167687468962092, "grad_norm": 5.225121021270752, "learning_rate": 2.1675157725756928e-06, "loss": 0.4497, "step": 26190 }, { "epoch": 2.168515146498924, "grad_norm": 4.732147693634033, "learning_rate": 2.1635489785780302e-06, "loss": 0.4222, "step": 26200 }, { "epoch": 2.1693428240357555, "grad_norm": 4.892897605895996, "learning_rate": 2.1595848153357026e-06, "loss": 0.4524, "step": 26210 }, { "epoch": 2.170170501572587, "grad_norm": 5.644736289978027, "learning_rate": 2.1556232865253986e-06, "loss": 0.4216, "step": 26220 }, { "epoch": 2.170998179109419, "grad_norm": 4.827288627624512, "learning_rate": 2.151664395821363e-06, "loss": 0.383, "step": 26230 }, { "epoch": 2.1718258566462505, "grad_norm": 6.449303150177002, "learning_rate": 2.1477081468953925e-06, "loss": 0.3471, "step": 26240 }, { "epoch": 2.172653534183082, "grad_norm": 5.793910026550293, "learning_rate": 2.143754543416836e-06, "loss": 0.4159, "step": 26250 }, { "epoch": 2.173481211719914, "grad_norm": 4.741961479187012, "learning_rate": 2.139803589052584e-06, "loss": 0.4048, "step": 26260 }, { "epoch": 2.1743088892567455, "grad_norm": 4.4309515953063965, "learning_rate": 2.135855287467075e-06, "loss": 0.4157, "step": 26270 }, { "epoch": 2.175136566793577, "grad_norm": 4.915762424468994, "learning_rate": 2.1319096423222807e-06, "loss": 0.401, "step": 26280 }, { "epoch": 2.175964244330409, "grad_norm": 5.618540287017822, "learning_rate": 2.1279666572777204e-06, "loss": 0.4053, "step": 26290 }, { "epoch": 2.1767919218672405, "grad_norm": 5.186798095703125, "learning_rate": 2.124026335990433e-06, "loss": 0.4519, "step": 26300 }, { "epoch": 2.177619599404072, "grad_norm": 7.20513391494751, "learning_rate": 2.1200886821149945e-06, "loss": 0.4157, "step": 26310 }, { "epoch": 2.178447276940904, "grad_norm": 6.136571884155273, "learning_rate": 2.116153699303509e-06, "loss": 0.3832, "step": 26320 }, { "epoch": 2.1792749544777354, "grad_norm": 5.166226387023926, "learning_rate": 2.1122213912055915e-06, "loss": 0.3978, "step": 26330 }, { "epoch": 2.180102632014567, "grad_norm": 4.445766448974609, "learning_rate": 2.108291761468394e-06, "loss": 0.3615, "step": 26340 }, { "epoch": 2.1809303095513988, "grad_norm": 4.085413455963135, "learning_rate": 2.1043648137365674e-06, "loss": 0.4648, "step": 26350 }, { "epoch": 2.1817579870882304, "grad_norm": 4.2033209800720215, "learning_rate": 2.1004405516522853e-06, "loss": 0.419, "step": 26360 }, { "epoch": 2.182585664625062, "grad_norm": 5.435886859893799, "learning_rate": 2.0965189788552286e-06, "loss": 0.4492, "step": 26370 }, { "epoch": 2.1834133421618938, "grad_norm": 4.972742557525635, "learning_rate": 2.0926000989825796e-06, "loss": 0.4706, "step": 26380 }, { "epoch": 2.1842410196987254, "grad_norm": 4.248584747314453, "learning_rate": 2.088683915669028e-06, "loss": 0.3805, "step": 26390 }, { "epoch": 2.185068697235557, "grad_norm": 4.133742809295654, "learning_rate": 2.0847704325467604e-06, "loss": 0.3578, "step": 26400 }, { "epoch": 2.1858963747723887, "grad_norm": 4.321877479553223, "learning_rate": 2.0808596532454595e-06, "loss": 0.4138, "step": 26410 }, { "epoch": 2.1867240523092204, "grad_norm": 3.857802152633667, "learning_rate": 2.0769515813923014e-06, "loss": 0.4262, "step": 26420 }, { "epoch": 2.187551729846052, "grad_norm": 5.392834186553955, "learning_rate": 2.073046220611947e-06, "loss": 0.469, "step": 26430 }, { "epoch": 2.1883794073828837, "grad_norm": 7.845005512237549, "learning_rate": 2.069143574526548e-06, "loss": 0.4498, "step": 26440 }, { "epoch": 2.1892070849197154, "grad_norm": 5.4609246253967285, "learning_rate": 2.065243646755731e-06, "loss": 0.3851, "step": 26450 }, { "epoch": 2.190034762456547, "grad_norm": 4.708670616149902, "learning_rate": 2.0613464409166124e-06, "loss": 0.3293, "step": 26460 }, { "epoch": 2.1908624399933787, "grad_norm": 3.982907772064209, "learning_rate": 2.057451960623776e-06, "loss": 0.3774, "step": 26470 }, { "epoch": 2.1916901175302104, "grad_norm": 5.379378795623779, "learning_rate": 2.0535602094892772e-06, "loss": 0.4286, "step": 26480 }, { "epoch": 2.192517795067042, "grad_norm": 4.243257522583008, "learning_rate": 2.049671191122645e-06, "loss": 0.3735, "step": 26490 }, { "epoch": 2.1933454726038737, "grad_norm": 5.334815979003906, "learning_rate": 2.045784909130869e-06, "loss": 0.4133, "step": 26500 }, { "epoch": 2.1941731501407054, "grad_norm": 4.115193843841553, "learning_rate": 2.041901367118404e-06, "loss": 0.3983, "step": 26510 }, { "epoch": 2.195000827677537, "grad_norm": 5.7688188552856445, "learning_rate": 2.0380205686871634e-06, "loss": 0.4018, "step": 26520 }, { "epoch": 2.1958285052143687, "grad_norm": 4.091005802154541, "learning_rate": 2.0341425174365147e-06, "loss": 0.4017, "step": 26530 }, { "epoch": 2.1966561827512003, "grad_norm": 3.541567325592041, "learning_rate": 2.0302672169632797e-06, "loss": 0.4145, "step": 26540 }, { "epoch": 2.197483860288032, "grad_norm": 4.2828850746154785, "learning_rate": 2.026394670861724e-06, "loss": 0.4433, "step": 26550 }, { "epoch": 2.198311537824863, "grad_norm": 5.667818546295166, "learning_rate": 2.0225248827235635e-06, "loss": 0.453, "step": 26560 }, { "epoch": 2.199139215361695, "grad_norm": 4.793360710144043, "learning_rate": 2.0186578561379544e-06, "loss": 0.4211, "step": 26570 }, { "epoch": 2.1999668928985265, "grad_norm": 4.7050862312316895, "learning_rate": 2.014793594691491e-06, "loss": 0.4491, "step": 26580 }, { "epoch": 2.200794570435358, "grad_norm": 11.024834632873535, "learning_rate": 2.0109321019682065e-06, "loss": 0.363, "step": 26590 }, { "epoch": 2.20162224797219, "grad_norm": 4.623368263244629, "learning_rate": 2.0070733815495585e-06, "loss": 0.4031, "step": 26600 }, { "epoch": 2.2024499255090215, "grad_norm": 3.4675891399383545, "learning_rate": 2.003217437014443e-06, "loss": 0.3774, "step": 26610 }, { "epoch": 2.203277603045853, "grad_norm": 4.449466705322266, "learning_rate": 1.999364271939172e-06, "loss": 0.3971, "step": 26620 }, { "epoch": 2.204105280582685, "grad_norm": 3.0519137382507324, "learning_rate": 1.995513889897484e-06, "loss": 0.4051, "step": 26630 }, { "epoch": 2.2049329581195165, "grad_norm": 6.210385322570801, "learning_rate": 1.991666294460543e-06, "loss": 0.3825, "step": 26640 }, { "epoch": 2.205760635656348, "grad_norm": 4.634474277496338, "learning_rate": 1.987821489196915e-06, "loss": 0.3655, "step": 26650 }, { "epoch": 2.20658831319318, "grad_norm": 5.597301006317139, "learning_rate": 1.9839794776725894e-06, "loss": 0.4665, "step": 26660 }, { "epoch": 2.2074159907300115, "grad_norm": 5.824253559112549, "learning_rate": 1.9801402634509565e-06, "loss": 0.4733, "step": 26670 }, { "epoch": 2.208243668266843, "grad_norm": 4.413086414337158, "learning_rate": 1.9763038500928177e-06, "loss": 0.4363, "step": 26680 }, { "epoch": 2.209071345803675, "grad_norm": 3.2133877277374268, "learning_rate": 1.9724702411563747e-06, "loss": 0.4323, "step": 26690 }, { "epoch": 2.2098990233405065, "grad_norm": 4.836264610290527, "learning_rate": 1.9686394401972274e-06, "loss": 0.3736, "step": 26700 }, { "epoch": 2.210726700877338, "grad_norm": 4.041816234588623, "learning_rate": 1.9648114507683754e-06, "loss": 0.4026, "step": 26710 }, { "epoch": 2.21155437841417, "grad_norm": 5.108896732330322, "learning_rate": 1.9609862764202026e-06, "loss": 0.3691, "step": 26720 }, { "epoch": 2.2123820559510015, "grad_norm": 4.51970911026001, "learning_rate": 1.9571639207004887e-06, "loss": 0.3584, "step": 26730 }, { "epoch": 2.213209733487833, "grad_norm": 2.976283311843872, "learning_rate": 1.9533443871543995e-06, "loss": 0.4205, "step": 26740 }, { "epoch": 2.214037411024665, "grad_norm": 3.7958662509918213, "learning_rate": 1.949527679324475e-06, "loss": 0.3932, "step": 26750 }, { "epoch": 2.2148650885614964, "grad_norm": 4.639182090759277, "learning_rate": 1.945713800750648e-06, "loss": 0.4397, "step": 26760 }, { "epoch": 2.215692766098328, "grad_norm": 5.3484296798706055, "learning_rate": 1.9419027549702144e-06, "loss": 0.3832, "step": 26770 }, { "epoch": 2.2165204436351598, "grad_norm": 4.711378574371338, "learning_rate": 1.9380945455178494e-06, "loss": 0.467, "step": 26780 }, { "epoch": 2.2173481211719914, "grad_norm": 8.973124504089355, "learning_rate": 1.9342891759255987e-06, "loss": 0.3571, "step": 26790 }, { "epoch": 2.218175798708823, "grad_norm": 4.079860210418701, "learning_rate": 1.930486649722867e-06, "loss": 0.3112, "step": 26800 }, { "epoch": 2.2190034762456547, "grad_norm": 4.582979202270508, "learning_rate": 1.9266869704364283e-06, "loss": 0.3929, "step": 26810 }, { "epoch": 2.2198311537824864, "grad_norm": 5.583993434906006, "learning_rate": 1.9228901415904137e-06, "loss": 0.3515, "step": 26820 }, { "epoch": 2.220658831319318, "grad_norm": 4.185256481170654, "learning_rate": 1.919096166706314e-06, "loss": 0.366, "step": 26830 }, { "epoch": 2.2214865088561497, "grad_norm": 4.427895545959473, "learning_rate": 1.9153050493029664e-06, "loss": 0.4279, "step": 26840 }, { "epoch": 2.2223141863929814, "grad_norm": 5.1793107986450195, "learning_rate": 1.9115167928965617e-06, "loss": 0.3591, "step": 26850 }, { "epoch": 2.223141863929813, "grad_norm": 6.150559902191162, "learning_rate": 1.90773140100064e-06, "loss": 0.3978, "step": 26860 }, { "epoch": 2.2239695414666447, "grad_norm": 4.455411434173584, "learning_rate": 1.9039488771260767e-06, "loss": 0.3814, "step": 26870 }, { "epoch": 2.2247972190034764, "grad_norm": 3.58602237701416, "learning_rate": 1.9001692247810976e-06, "loss": 0.3827, "step": 26880 }, { "epoch": 2.225624896540308, "grad_norm": 5.439874649047852, "learning_rate": 1.8963924474712552e-06, "loss": 0.4308, "step": 26890 }, { "epoch": 2.2264525740771397, "grad_norm": 4.935842514038086, "learning_rate": 1.892618548699442e-06, "loss": 0.4332, "step": 26900 }, { "epoch": 2.2272802516139714, "grad_norm": 3.9980905055999756, "learning_rate": 1.88884753196588e-06, "loss": 0.4737, "step": 26910 }, { "epoch": 2.228107929150803, "grad_norm": 2.93924617767334, "learning_rate": 1.8850794007681134e-06, "loss": 0.3568, "step": 26920 }, { "epoch": 2.2289356066876347, "grad_norm": 4.784413814544678, "learning_rate": 1.8813141586010148e-06, "loss": 0.4314, "step": 26930 }, { "epoch": 2.2297632842244663, "grad_norm": 4.673656463623047, "learning_rate": 1.8775518089567761e-06, "loss": 0.4046, "step": 26940 }, { "epoch": 2.2305909617612976, "grad_norm": 3.3637664318084717, "learning_rate": 1.8737923553249065e-06, "loss": 0.4036, "step": 26950 }, { "epoch": 2.2314186392981297, "grad_norm": 6.0438055992126465, "learning_rate": 1.870035801192231e-06, "loss": 0.4075, "step": 26960 }, { "epoch": 2.232246316834961, "grad_norm": 3.8552932739257812, "learning_rate": 1.8662821500428796e-06, "loss": 0.4079, "step": 26970 }, { "epoch": 2.2330739943717925, "grad_norm": 7.548283576965332, "learning_rate": 1.8625314053582954e-06, "loss": 0.4923, "step": 26980 }, { "epoch": 2.233901671908624, "grad_norm": 4.020804405212402, "learning_rate": 1.8587835706172242e-06, "loss": 0.343, "step": 26990 }, { "epoch": 2.234729349445456, "grad_norm": 3.3734779357910156, "learning_rate": 1.8550386492957146e-06, "loss": 0.4026, "step": 27000 }, { "epoch": 2.2355570269822875, "grad_norm": 4.086536884307861, "learning_rate": 1.8512966448671066e-06, "loss": 0.4834, "step": 27010 }, { "epoch": 2.236384704519119, "grad_norm": 4.944655418395996, "learning_rate": 1.8475575608020425e-06, "loss": 0.4377, "step": 27020 }, { "epoch": 2.237212382055951, "grad_norm": 5.548043251037598, "learning_rate": 1.8438214005684541e-06, "loss": 0.3923, "step": 27030 }, { "epoch": 2.2380400595927825, "grad_norm": 4.717337131500244, "learning_rate": 1.8400881676315547e-06, "loss": 0.4846, "step": 27040 }, { "epoch": 2.238867737129614, "grad_norm": 3.6438515186309814, "learning_rate": 1.8363578654538556e-06, "loss": 0.429, "step": 27050 }, { "epoch": 2.239695414666446, "grad_norm": 4.578632354736328, "learning_rate": 1.832630497495137e-06, "loss": 0.4058, "step": 27060 }, { "epoch": 2.2405230922032775, "grad_norm": 5.336453437805176, "learning_rate": 1.8289060672124647e-06, "loss": 0.3847, "step": 27070 }, { "epoch": 2.241350769740109, "grad_norm": 7.327024459838867, "learning_rate": 1.8251845780601797e-06, "loss": 0.4443, "step": 27080 }, { "epoch": 2.242178447276941, "grad_norm": 6.060422420501709, "learning_rate": 1.821466033489891e-06, "loss": 0.3435, "step": 27090 }, { "epoch": 2.2430061248137725, "grad_norm": 3.215965747833252, "learning_rate": 1.81775043695048e-06, "loss": 0.3831, "step": 27100 }, { "epoch": 2.243833802350604, "grad_norm": 2.823294162750244, "learning_rate": 1.8140377918880936e-06, "loss": 0.4265, "step": 27110 }, { "epoch": 2.244661479887436, "grad_norm": 3.9084115028381348, "learning_rate": 1.810328101746141e-06, "loss": 0.3675, "step": 27120 }, { "epoch": 2.2454891574242675, "grad_norm": 3.736576557159424, "learning_rate": 1.8066213699652913e-06, "loss": 0.4406, "step": 27130 }, { "epoch": 2.246316834961099, "grad_norm": 5.450920581817627, "learning_rate": 1.802917599983467e-06, "loss": 0.3998, "step": 27140 }, { "epoch": 2.247144512497931, "grad_norm": 3.3924849033355713, "learning_rate": 1.7992167952358475e-06, "loss": 0.3971, "step": 27150 }, { "epoch": 2.2479721900347625, "grad_norm": 5.872109889984131, "learning_rate": 1.7955189591548555e-06, "loss": 0.428, "step": 27160 }, { "epoch": 2.248799867571594, "grad_norm": 6.13421630859375, "learning_rate": 1.791824095170172e-06, "loss": 0.4046, "step": 27170 }, { "epoch": 2.2496275451084258, "grad_norm": 3.8173975944519043, "learning_rate": 1.7881322067087093e-06, "loss": 0.4055, "step": 27180 }, { "epoch": 2.2504552226452574, "grad_norm": 5.525921821594238, "learning_rate": 1.7844432971946262e-06, "loss": 0.4239, "step": 27190 }, { "epoch": 2.251282900182089, "grad_norm": 4.060267925262451, "learning_rate": 1.7807573700493192e-06, "loss": 0.3684, "step": 27200 }, { "epoch": 2.2521105777189208, "grad_norm": 5.918928146362305, "learning_rate": 1.7770744286914138e-06, "loss": 0.4119, "step": 27210 }, { "epoch": 2.2529382552557524, "grad_norm": 4.723400115966797, "learning_rate": 1.7733944765367706e-06, "loss": 0.4803, "step": 27220 }, { "epoch": 2.253765932792584, "grad_norm": 3.904430627822876, "learning_rate": 1.7697175169984765e-06, "loss": 0.3781, "step": 27230 }, { "epoch": 2.2545936103294157, "grad_norm": 3.3352696895599365, "learning_rate": 1.766043553486843e-06, "loss": 0.3936, "step": 27240 }, { "epoch": 2.2554212878662474, "grad_norm": 4.199538230895996, "learning_rate": 1.762372589409404e-06, "loss": 0.4615, "step": 27250 }, { "epoch": 2.256248965403079, "grad_norm": 4.583846569061279, "learning_rate": 1.7587046281709076e-06, "loss": 0.4514, "step": 27260 }, { "epoch": 2.2570766429399107, "grad_norm": 5.3214521408081055, "learning_rate": 1.7550396731733198e-06, "loss": 0.388, "step": 27270 }, { "epoch": 2.2579043204767424, "grad_norm": 5.319254398345947, "learning_rate": 1.751377727815819e-06, "loss": 0.3803, "step": 27280 }, { "epoch": 2.258731998013574, "grad_norm": 5.752534866333008, "learning_rate": 1.74771879549479e-06, "loss": 0.3787, "step": 27290 }, { "epoch": 2.2595596755504057, "grad_norm": 5.0199503898620605, "learning_rate": 1.7440628796038268e-06, "loss": 0.373, "step": 27300 }, { "epoch": 2.2603873530872374, "grad_norm": 2.9004056453704834, "learning_rate": 1.740409983533719e-06, "loss": 0.3595, "step": 27310 }, { "epoch": 2.261215030624069, "grad_norm": 3.6701042652130127, "learning_rate": 1.736760110672464e-06, "loss": 0.3848, "step": 27320 }, { "epoch": 2.2620427081609007, "grad_norm": 4.970771789550781, "learning_rate": 1.7331132644052462e-06, "loss": 0.3985, "step": 27330 }, { "epoch": 2.2628703856977324, "grad_norm": 5.387011528015137, "learning_rate": 1.7294694481144475e-06, "loss": 0.4159, "step": 27340 }, { "epoch": 2.263698063234564, "grad_norm": 5.52504301071167, "learning_rate": 1.7258286651796446e-06, "loss": 0.4475, "step": 27350 }, { "epoch": 2.2645257407713952, "grad_norm": 4.010457515716553, "learning_rate": 1.7221909189775909e-06, "loss": 0.4168, "step": 27360 }, { "epoch": 2.2653534183082273, "grad_norm": 4.085269927978516, "learning_rate": 1.7185562128822313e-06, "loss": 0.367, "step": 27370 }, { "epoch": 2.2661810958450586, "grad_norm": 4.601396083831787, "learning_rate": 1.7149245502646849e-06, "loss": 0.4311, "step": 27380 }, { "epoch": 2.2670087733818907, "grad_norm": 4.506036758422852, "learning_rate": 1.7112959344932517e-06, "loss": 0.4075, "step": 27390 }, { "epoch": 2.267836450918722, "grad_norm": 5.177253723144531, "learning_rate": 1.7076703689334056e-06, "loss": 0.4255, "step": 27400 }, { "epoch": 2.2686641284555535, "grad_norm": 5.844703197479248, "learning_rate": 1.7040478569477903e-06, "loss": 0.3784, "step": 27410 }, { "epoch": 2.269491805992385, "grad_norm": 2.546117067337036, "learning_rate": 1.7004284018962208e-06, "loss": 0.3593, "step": 27420 }, { "epoch": 2.270319483529217, "grad_norm": 5.194912910461426, "learning_rate": 1.6968120071356697e-06, "loss": 0.3919, "step": 27430 }, { "epoch": 2.2711471610660485, "grad_norm": 3.4522769451141357, "learning_rate": 1.693198676020278e-06, "loss": 0.3938, "step": 27440 }, { "epoch": 2.27197483860288, "grad_norm": 4.12183952331543, "learning_rate": 1.6895884119013433e-06, "loss": 0.4221, "step": 27450 }, { "epoch": 2.272802516139712, "grad_norm": 3.4051856994628906, "learning_rate": 1.6859812181273144e-06, "loss": 0.4793, "step": 27460 }, { "epoch": 2.2736301936765435, "grad_norm": 3.689699411392212, "learning_rate": 1.6823770980438025e-06, "loss": 0.4064, "step": 27470 }, { "epoch": 2.274457871213375, "grad_norm": 3.6719000339508057, "learning_rate": 1.6787760549935566e-06, "loss": 0.3762, "step": 27480 }, { "epoch": 2.275285548750207, "grad_norm": 4.98137092590332, "learning_rate": 1.6751780923164784e-06, "loss": 0.4883, "step": 27490 }, { "epoch": 2.2761132262870385, "grad_norm": 3.867928981781006, "learning_rate": 1.6715832133496124e-06, "loss": 0.4636, "step": 27500 }, { "epoch": 2.27694090382387, "grad_norm": 6.737900733947754, "learning_rate": 1.6679914214271387e-06, "loss": 0.4303, "step": 27510 }, { "epoch": 2.277768581360702, "grad_norm": 4.427409648895264, "learning_rate": 1.664402719880379e-06, "loss": 0.3894, "step": 27520 }, { "epoch": 2.2785962588975335, "grad_norm": 4.025156497955322, "learning_rate": 1.6608171120377869e-06, "loss": 0.39, "step": 27530 }, { "epoch": 2.279423936434365, "grad_norm": 3.3690478801727295, "learning_rate": 1.6572346012249485e-06, "loss": 0.339, "step": 27540 }, { "epoch": 2.280251613971197, "grad_norm": 2.319610357284546, "learning_rate": 1.653655190764572e-06, "loss": 0.3589, "step": 27550 }, { "epoch": 2.2810792915080285, "grad_norm": 3.434643268585205, "learning_rate": 1.6500788839764952e-06, "loss": 0.4186, "step": 27560 }, { "epoch": 2.28190696904486, "grad_norm": 2.505845785140991, "learning_rate": 1.6465056841776784e-06, "loss": 0.3916, "step": 27570 }, { "epoch": 2.282734646581692, "grad_norm": 4.731109142303467, "learning_rate": 1.6429355946821918e-06, "loss": 0.4129, "step": 27580 }, { "epoch": 2.2835623241185234, "grad_norm": 4.936246395111084, "learning_rate": 1.6393686188012337e-06, "loss": 0.3933, "step": 27590 }, { "epoch": 2.284390001655355, "grad_norm": 5.507744789123535, "learning_rate": 1.6358047598431026e-06, "loss": 0.4246, "step": 27600 }, { "epoch": 2.2852176791921868, "grad_norm": 4.697116374969482, "learning_rate": 1.6322440211132129e-06, "loss": 0.4495, "step": 27610 }, { "epoch": 2.2860453567290184, "grad_norm": 3.328240394592285, "learning_rate": 1.6286864059140844e-06, "loss": 0.4311, "step": 27620 }, { "epoch": 2.28687303426585, "grad_norm": 4.109248161315918, "learning_rate": 1.625131917545336e-06, "loss": 0.3248, "step": 27630 }, { "epoch": 2.2877007118026818, "grad_norm": 5.301110744476318, "learning_rate": 1.621580559303691e-06, "loss": 0.4273, "step": 27640 }, { "epoch": 2.2885283893395134, "grad_norm": 5.720044136047363, "learning_rate": 1.6180323344829674e-06, "loss": 0.2806, "step": 27650 }, { "epoch": 2.289356066876345, "grad_norm": 5.660804748535156, "learning_rate": 1.6144872463740773e-06, "loss": 0.384, "step": 27660 }, { "epoch": 2.2901837444131767, "grad_norm": 6.7290940284729, "learning_rate": 1.610945298265026e-06, "loss": 0.4363, "step": 27670 }, { "epoch": 2.2910114219500084, "grad_norm": 3.508053779602051, "learning_rate": 1.6074064934409001e-06, "loss": 0.4014, "step": 27680 }, { "epoch": 2.29183909948684, "grad_norm": 5.08007287979126, "learning_rate": 1.6038708351838773e-06, "loss": 0.4482, "step": 27690 }, { "epoch": 2.2926667770236717, "grad_norm": 3.9972171783447266, "learning_rate": 1.6003383267732143e-06, "loss": 0.4435, "step": 27700 }, { "epoch": 2.2934944545605034, "grad_norm": 5.6634368896484375, "learning_rate": 1.5968089714852486e-06, "loss": 0.4499, "step": 27710 }, { "epoch": 2.294322132097335, "grad_norm": 3.153798818588257, "learning_rate": 1.5932827725933875e-06, "loss": 0.4372, "step": 27720 }, { "epoch": 2.2951498096341667, "grad_norm": 7.541629314422607, "learning_rate": 1.5897597333681164e-06, "loss": 0.3758, "step": 27730 }, { "epoch": 2.2959774871709984, "grad_norm": 2.6404671669006348, "learning_rate": 1.5862398570769904e-06, "loss": 0.3939, "step": 27740 }, { "epoch": 2.2968051647078296, "grad_norm": 6.577878475189209, "learning_rate": 1.5827231469846237e-06, "loss": 0.4707, "step": 27750 }, { "epoch": 2.2976328422446617, "grad_norm": 4.865602970123291, "learning_rate": 1.5792096063527062e-06, "loss": 0.3934, "step": 27760 }, { "epoch": 2.298460519781493, "grad_norm": 5.568453788757324, "learning_rate": 1.5756992384399756e-06, "loss": 0.3875, "step": 27770 }, { "epoch": 2.299288197318325, "grad_norm": 4.739243030548096, "learning_rate": 1.5721920465022345e-06, "loss": 0.3973, "step": 27780 }, { "epoch": 2.3001158748551562, "grad_norm": 5.616820335388184, "learning_rate": 1.5686880337923398e-06, "loss": 0.3952, "step": 27790 }, { "epoch": 2.300943552391988, "grad_norm": 4.912978649139404, "learning_rate": 1.5651872035601945e-06, "loss": 0.3553, "step": 27800 }, { "epoch": 2.3017712299288196, "grad_norm": 5.634611129760742, "learning_rate": 1.5616895590527548e-06, "loss": 0.3937, "step": 27810 }, { "epoch": 2.302598907465651, "grad_norm": 4.283088684082031, "learning_rate": 1.5581951035140203e-06, "loss": 0.4019, "step": 27820 }, { "epoch": 2.303426585002483, "grad_norm": 4.244801044464111, "learning_rate": 1.5547038401850334e-06, "loss": 0.3839, "step": 27830 }, { "epoch": 2.3042542625393145, "grad_norm": 6.17240047454834, "learning_rate": 1.5512157723038778e-06, "loss": 0.394, "step": 27840 }, { "epoch": 2.305081940076146, "grad_norm": 4.82076358795166, "learning_rate": 1.5477309031056676e-06, "loss": 0.4312, "step": 27850 }, { "epoch": 2.305909617612978, "grad_norm": 3.735229253768921, "learning_rate": 1.5442492358225568e-06, "loss": 0.4486, "step": 27860 }, { "epoch": 2.3067372951498095, "grad_norm": 5.289453983306885, "learning_rate": 1.5407707736837223e-06, "loss": 0.4114, "step": 27870 }, { "epoch": 2.307564972686641, "grad_norm": 3.6590137481689453, "learning_rate": 1.5372955199153799e-06, "loss": 0.4018, "step": 27880 }, { "epoch": 2.308392650223473, "grad_norm": 4.094576358795166, "learning_rate": 1.5338234777407563e-06, "loss": 0.4437, "step": 27890 }, { "epoch": 2.3092203277603045, "grad_norm": 4.762070178985596, "learning_rate": 1.5303546503801082e-06, "loss": 0.3458, "step": 27900 }, { "epoch": 2.310048005297136, "grad_norm": 3.4525187015533447, "learning_rate": 1.5268890410507098e-06, "loss": 0.3499, "step": 27910 }, { "epoch": 2.310875682833968, "grad_norm": 5.951925277709961, "learning_rate": 1.5234266529668458e-06, "loss": 0.4134, "step": 27920 }, { "epoch": 2.3117033603707995, "grad_norm": 5.768953323364258, "learning_rate": 1.5199674893398175e-06, "loss": 0.3644, "step": 27930 }, { "epoch": 2.312531037907631, "grad_norm": 4.956112384796143, "learning_rate": 1.5165115533779356e-06, "loss": 0.4232, "step": 27940 }, { "epoch": 2.313358715444463, "grad_norm": 4.30687952041626, "learning_rate": 1.513058848286515e-06, "loss": 0.4141, "step": 27950 }, { "epoch": 2.3141863929812945, "grad_norm": 5.232196807861328, "learning_rate": 1.509609377267877e-06, "loss": 0.4219, "step": 27960 }, { "epoch": 2.315014070518126, "grad_norm": 4.734195232391357, "learning_rate": 1.5061631435213398e-06, "loss": 0.4343, "step": 27970 }, { "epoch": 2.315841748054958, "grad_norm": 4.412688732147217, "learning_rate": 1.5027201502432203e-06, "loss": 0.407, "step": 27980 }, { "epoch": 2.3166694255917895, "grad_norm": 4.529552459716797, "learning_rate": 1.499280400626832e-06, "loss": 0.4377, "step": 27990 }, { "epoch": 2.317497103128621, "grad_norm": 7.1748504638671875, "learning_rate": 1.4958438978624774e-06, "loss": 0.4136, "step": 28000 }, { "epoch": 2.3183247806654528, "grad_norm": 3.9291234016418457, "learning_rate": 1.4924106451374504e-06, "loss": 0.414, "step": 28010 }, { "epoch": 2.3191524582022844, "grad_norm": 5.822134494781494, "learning_rate": 1.488980645636026e-06, "loss": 0.4353, "step": 28020 }, { "epoch": 2.319980135739116, "grad_norm": 4.478569030761719, "learning_rate": 1.4855539025394677e-06, "loss": 0.453, "step": 28030 }, { "epoch": 2.3208078132759478, "grad_norm": 5.432758331298828, "learning_rate": 1.4821304190260127e-06, "loss": 0.4159, "step": 28040 }, { "epoch": 2.3216354908127794, "grad_norm": 4.721989631652832, "learning_rate": 1.4787101982708795e-06, "loss": 0.4191, "step": 28050 }, { "epoch": 2.322463168349611, "grad_norm": 7.671929359436035, "learning_rate": 1.4752932434462585e-06, "loss": 0.3393, "step": 28060 }, { "epoch": 2.3232908458864427, "grad_norm": 4.571646213531494, "learning_rate": 1.4718795577213124e-06, "loss": 0.3895, "step": 28070 }, { "epoch": 2.3241185234232744, "grad_norm": 4.506448745727539, "learning_rate": 1.4684691442621712e-06, "loss": 0.3742, "step": 28080 }, { "epoch": 2.324946200960106, "grad_norm": 5.199589252471924, "learning_rate": 1.4650620062319276e-06, "loss": 0.3566, "step": 28090 }, { "epoch": 2.3257738784969377, "grad_norm": 5.4506754875183105, "learning_rate": 1.46165814679064e-06, "loss": 0.3895, "step": 28100 }, { "epoch": 2.3266015560337694, "grad_norm": 4.678353786468506, "learning_rate": 1.4582575690953238e-06, "loss": 0.4134, "step": 28110 }, { "epoch": 2.327429233570601, "grad_norm": 5.18079948425293, "learning_rate": 1.4548602762999515e-06, "loss": 0.4345, "step": 28120 }, { "epoch": 2.3282569111074327, "grad_norm": 3.9160988330841064, "learning_rate": 1.4514662715554507e-06, "loss": 0.4271, "step": 28130 }, { "epoch": 2.3290845886442644, "grad_norm": 10.92162036895752, "learning_rate": 1.4480755580096939e-06, "loss": 0.375, "step": 28140 }, { "epoch": 2.329912266181096, "grad_norm": 2.9048848152160645, "learning_rate": 1.4446881388075062e-06, "loss": 0.4, "step": 28150 }, { "epoch": 2.3307399437179273, "grad_norm": 4.205909252166748, "learning_rate": 1.4413040170906572e-06, "loss": 0.4046, "step": 28160 }, { "epoch": 2.3315676212547594, "grad_norm": 3.306919574737549, "learning_rate": 1.4379231959978525e-06, "loss": 0.472, "step": 28170 }, { "epoch": 2.3323952987915906, "grad_norm": 4.327777862548828, "learning_rate": 1.4345456786647454e-06, "loss": 0.3766, "step": 28180 }, { "epoch": 2.3332229763284227, "grad_norm": 5.376691818237305, "learning_rate": 1.4311714682239163e-06, "loss": 0.4104, "step": 28190 }, { "epoch": 2.334050653865254, "grad_norm": 5.020846366882324, "learning_rate": 1.4278005678048856e-06, "loss": 0.4091, "step": 28200 }, { "epoch": 2.3348783314020856, "grad_norm": 4.071484565734863, "learning_rate": 1.4244329805340973e-06, "loss": 0.4145, "step": 28210 }, { "epoch": 2.335706008938917, "grad_norm": 4.433290004730225, "learning_rate": 1.4210687095349269e-06, "loss": 0.4237, "step": 28220 }, { "epoch": 2.336533686475749, "grad_norm": 4.236571788787842, "learning_rate": 1.4177077579276733e-06, "loss": 0.4153, "step": 28230 }, { "epoch": 2.3373613640125805, "grad_norm": 5.76658821105957, "learning_rate": 1.4143501288295558e-06, "loss": 0.4121, "step": 28240 }, { "epoch": 2.338189041549412, "grad_norm": 4.446232795715332, "learning_rate": 1.410995825354715e-06, "loss": 0.3595, "step": 28250 }, { "epoch": 2.339016719086244, "grad_norm": 5.639676570892334, "learning_rate": 1.4076448506142009e-06, "loss": 0.4699, "step": 28260 }, { "epoch": 2.3398443966230755, "grad_norm": 3.763577699661255, "learning_rate": 1.4042972077159817e-06, "loss": 0.3683, "step": 28270 }, { "epoch": 2.340672074159907, "grad_norm": 3.7265727519989014, "learning_rate": 1.4009528997649353e-06, "loss": 0.4112, "step": 28280 }, { "epoch": 2.341499751696739, "grad_norm": 5.299983978271484, "learning_rate": 1.3976119298628404e-06, "loss": 0.3825, "step": 28290 }, { "epoch": 2.3423274292335705, "grad_norm": 5.869339942932129, "learning_rate": 1.3942743011083897e-06, "loss": 0.4036, "step": 28300 }, { "epoch": 2.343155106770402, "grad_norm": 3.674177885055542, "learning_rate": 1.3909400165971675e-06, "loss": 0.3479, "step": 28310 }, { "epoch": 2.343982784307234, "grad_norm": 4.162397861480713, "learning_rate": 1.3876090794216624e-06, "loss": 0.4062, "step": 28320 }, { "epoch": 2.3448104618440655, "grad_norm": 4.867801189422607, "learning_rate": 1.3842814926712567e-06, "loss": 0.4109, "step": 28330 }, { "epoch": 2.345638139380897, "grad_norm": 4.219747543334961, "learning_rate": 1.380957259432223e-06, "loss": 0.4194, "step": 28340 }, { "epoch": 2.346465816917729, "grad_norm": 4.771431922912598, "learning_rate": 1.377636382787726e-06, "loss": 0.413, "step": 28350 }, { "epoch": 2.3472934944545605, "grad_norm": 4.352506637573242, "learning_rate": 1.3743188658178176e-06, "loss": 0.4655, "step": 28360 }, { "epoch": 2.348121171991392, "grad_norm": 4.9182868003845215, "learning_rate": 1.3710047115994324e-06, "loss": 0.4483, "step": 28370 }, { "epoch": 2.348948849528224, "grad_norm": 4.905154705047607, "learning_rate": 1.3676939232063884e-06, "loss": 0.4116, "step": 28380 }, { "epoch": 2.3497765270650555, "grad_norm": 5.069812297821045, "learning_rate": 1.3643865037093768e-06, "loss": 0.4383, "step": 28390 }, { "epoch": 2.350604204601887, "grad_norm": 5.890801906585693, "learning_rate": 1.3610824561759688e-06, "loss": 0.4027, "step": 28400 }, { "epoch": 2.351431882138719, "grad_norm": 5.284999370574951, "learning_rate": 1.3577817836706064e-06, "loss": 0.4689, "step": 28410 }, { "epoch": 2.3522595596755504, "grad_norm": 4.111775875091553, "learning_rate": 1.3544844892546044e-06, "loss": 0.4276, "step": 28420 }, { "epoch": 2.353087237212382, "grad_norm": 4.658370018005371, "learning_rate": 1.3511905759861377e-06, "loss": 0.4385, "step": 28430 }, { "epoch": 2.3539149147492138, "grad_norm": 3.2117526531219482, "learning_rate": 1.3479000469202507e-06, "loss": 0.4181, "step": 28440 }, { "epoch": 2.3547425922860454, "grad_norm": 4.50427770614624, "learning_rate": 1.3446129051088502e-06, "loss": 0.4103, "step": 28450 }, { "epoch": 2.355570269822877, "grad_norm": 5.560866832733154, "learning_rate": 1.3413291536006935e-06, "loss": 0.4473, "step": 28460 }, { "epoch": 2.3563979473597088, "grad_norm": 5.7528157234191895, "learning_rate": 1.3380487954414046e-06, "loss": 0.3856, "step": 28470 }, { "epoch": 2.3572256248965404, "grad_norm": 5.189345836639404, "learning_rate": 1.3347718336734506e-06, "loss": 0.3651, "step": 28480 }, { "epoch": 2.358053302433372, "grad_norm": 5.5674920082092285, "learning_rate": 1.3314982713361536e-06, "loss": 0.3282, "step": 28490 }, { "epoch": 2.3588809799702037, "grad_norm": 5.535891056060791, "learning_rate": 1.3282281114656825e-06, "loss": 0.4534, "step": 28500 }, { "epoch": 2.3597086575070354, "grad_norm": 5.032627105712891, "learning_rate": 1.324961357095047e-06, "loss": 0.4443, "step": 28510 }, { "epoch": 2.360536335043867, "grad_norm": 3.603914737701416, "learning_rate": 1.3216980112541028e-06, "loss": 0.4004, "step": 28520 }, { "epoch": 2.3613640125806987, "grad_norm": 4.954718112945557, "learning_rate": 1.318438076969541e-06, "loss": 0.4359, "step": 28530 }, { "epoch": 2.3621916901175304, "grad_norm": 6.189148426055908, "learning_rate": 1.3151815572648907e-06, "loss": 0.4064, "step": 28540 }, { "epoch": 2.363019367654362, "grad_norm": 3.4975969791412354, "learning_rate": 1.3119284551605139e-06, "loss": 0.3772, "step": 28550 }, { "epoch": 2.3638470451911937, "grad_norm": 4.451476097106934, "learning_rate": 1.3086787736735996e-06, "loss": 0.4033, "step": 28560 }, { "epoch": 2.364674722728025, "grad_norm": 4.867207050323486, "learning_rate": 1.3054325158181692e-06, "loss": 0.4437, "step": 28570 }, { "epoch": 2.365502400264857, "grad_norm": 3.6006336212158203, "learning_rate": 1.302189684605062e-06, "loss": 0.4135, "step": 28580 }, { "epoch": 2.3663300778016882, "grad_norm": 3.591465950012207, "learning_rate": 1.2989502830419499e-06, "loss": 0.4591, "step": 28590 }, { "epoch": 2.3671577553385204, "grad_norm": 4.055998802185059, "learning_rate": 1.2957143141333117e-06, "loss": 0.468, "step": 28600 }, { "epoch": 2.3679854328753516, "grad_norm": 3.7380611896514893, "learning_rate": 1.2924817808804502e-06, "loss": 0.4128, "step": 28610 }, { "epoch": 2.3688131104121832, "grad_norm": 4.334061145782471, "learning_rate": 1.2892526862814809e-06, "loss": 0.4356, "step": 28620 }, { "epoch": 2.369640787949015, "grad_norm": 2.8340775966644287, "learning_rate": 1.2860270333313262e-06, "loss": 0.3636, "step": 28630 }, { "epoch": 2.3704684654858466, "grad_norm": 3.9213757514953613, "learning_rate": 1.28280482502172e-06, "loss": 0.3443, "step": 28640 }, { "epoch": 2.371296143022678, "grad_norm": 5.27379035949707, "learning_rate": 1.2795860643412012e-06, "loss": 0.3894, "step": 28650 }, { "epoch": 2.37212382055951, "grad_norm": 3.273238182067871, "learning_rate": 1.2763707542751086e-06, "loss": 0.4125, "step": 28660 }, { "epoch": 2.3729514980963415, "grad_norm": 4.099873065948486, "learning_rate": 1.2731588978055847e-06, "loss": 0.4182, "step": 28670 }, { "epoch": 2.373779175633173, "grad_norm": 4.524769306182861, "learning_rate": 1.269950497911564e-06, "loss": 0.416, "step": 28680 }, { "epoch": 2.374606853170005, "grad_norm": 4.14206075668335, "learning_rate": 1.2667455575687776e-06, "loss": 0.3824, "step": 28690 }, { "epoch": 2.3754345307068365, "grad_norm": 4.441004276275635, "learning_rate": 1.2635440797497484e-06, "loss": 0.3815, "step": 28700 }, { "epoch": 2.376262208243668, "grad_norm": 6.4068121910095215, "learning_rate": 1.2603460674237877e-06, "loss": 0.415, "step": 28710 }, { "epoch": 2.3770898857805, "grad_norm": 4.661662578582764, "learning_rate": 1.257151523556993e-06, "loss": 0.3372, "step": 28720 }, { "epoch": 2.3779175633173315, "grad_norm": 4.0884294509887695, "learning_rate": 1.2539604511122416e-06, "loss": 0.3799, "step": 28730 }, { "epoch": 2.378745240854163, "grad_norm": 4.499998092651367, "learning_rate": 1.2507728530491964e-06, "loss": 0.3942, "step": 28740 }, { "epoch": 2.379572918390995, "grad_norm": 2.88297963142395, "learning_rate": 1.247588732324292e-06, "loss": 0.3329, "step": 28750 }, { "epoch": 2.3804005959278265, "grad_norm": 5.38878059387207, "learning_rate": 1.2444080918907425e-06, "loss": 0.3967, "step": 28760 }, { "epoch": 2.381228273464658, "grad_norm": 5.5864787101745605, "learning_rate": 1.2412309346985318e-06, "loss": 0.4724, "step": 28770 }, { "epoch": 2.38205595100149, "grad_norm": 5.410919189453125, "learning_rate": 1.238057263694415e-06, "loss": 0.4513, "step": 28780 }, { "epoch": 2.3828836285383215, "grad_norm": 3.8552286624908447, "learning_rate": 1.2348870818219134e-06, "loss": 0.3925, "step": 28790 }, { "epoch": 2.383711306075153, "grad_norm": 5.440114974975586, "learning_rate": 1.2317203920213089e-06, "loss": 0.3237, "step": 28800 }, { "epoch": 2.384538983611985, "grad_norm": 4.429317474365234, "learning_rate": 1.2285571972296485e-06, "loss": 0.4146, "step": 28810 }, { "epoch": 2.3853666611488165, "grad_norm": 4.19567346572876, "learning_rate": 1.2253975003807362e-06, "loss": 0.375, "step": 28820 }, { "epoch": 2.386194338685648, "grad_norm": 5.304452896118164, "learning_rate": 1.2222413044051312e-06, "loss": 0.3977, "step": 28830 }, { "epoch": 2.38702201622248, "grad_norm": 4.674620628356934, "learning_rate": 1.2190886122301487e-06, "loss": 0.4299, "step": 28840 }, { "epoch": 2.3878496937593114, "grad_norm": 4.787268161773682, "learning_rate": 1.2159394267798486e-06, "loss": 0.4617, "step": 28850 }, { "epoch": 2.388677371296143, "grad_norm": 4.056704521179199, "learning_rate": 1.2127937509750432e-06, "loss": 0.4085, "step": 28860 }, { "epoch": 2.3895050488329748, "grad_norm": 4.915304183959961, "learning_rate": 1.2096515877332892e-06, "loss": 0.3843, "step": 28870 }, { "epoch": 2.3903327263698064, "grad_norm": 5.537027835845947, "learning_rate": 1.2065129399688802e-06, "loss": 0.4541, "step": 28880 }, { "epoch": 2.391160403906638, "grad_norm": 5.51710319519043, "learning_rate": 1.2033778105928584e-06, "loss": 0.398, "step": 28890 }, { "epoch": 2.3919880814434697, "grad_norm": 6.353671550750732, "learning_rate": 1.2002462025129946e-06, "loss": 0.3582, "step": 28900 }, { "epoch": 2.3928157589803014, "grad_norm": 4.6771979331970215, "learning_rate": 1.1971181186337988e-06, "loss": 0.4081, "step": 28910 }, { "epoch": 2.393643436517133, "grad_norm": 5.11600399017334, "learning_rate": 1.193993561856508e-06, "loss": 0.4781, "step": 28920 }, { "epoch": 2.3944711140539647, "grad_norm": 3.983680248260498, "learning_rate": 1.1908725350790912e-06, "loss": 0.4412, "step": 28930 }, { "epoch": 2.3952987915907964, "grad_norm": 2.564603567123413, "learning_rate": 1.1877550411962424e-06, "loss": 0.3832, "step": 28940 }, { "epoch": 2.396126469127628, "grad_norm": 3.9275381565093994, "learning_rate": 1.18464108309938e-06, "loss": 0.4033, "step": 28950 }, { "epoch": 2.3969541466644597, "grad_norm": 3.1288907527923584, "learning_rate": 1.181530663676642e-06, "loss": 0.3265, "step": 28960 }, { "epoch": 2.3977818242012914, "grad_norm": 5.327974796295166, "learning_rate": 1.1784237858128823e-06, "loss": 0.4164, "step": 28970 }, { "epoch": 2.3986095017381226, "grad_norm": 4.6366376876831055, "learning_rate": 1.1753204523896733e-06, "loss": 0.3623, "step": 28980 }, { "epoch": 2.3994371792749547, "grad_norm": 5.088975429534912, "learning_rate": 1.1722206662853004e-06, "loss": 0.4036, "step": 28990 }, { "epoch": 2.400264856811786, "grad_norm": 5.123826503753662, "learning_rate": 1.1691244303747524e-06, "loss": 0.3713, "step": 29000 }, { "epoch": 2.401092534348618, "grad_norm": 5.472893238067627, "learning_rate": 1.1660317475297367e-06, "loss": 0.4349, "step": 29010 }, { "epoch": 2.4019202118854492, "grad_norm": 5.829949378967285, "learning_rate": 1.162942620618654e-06, "loss": 0.4349, "step": 29020 }, { "epoch": 2.402747889422281, "grad_norm": 5.478431224822998, "learning_rate": 1.1598570525066138e-06, "loss": 0.347, "step": 29030 }, { "epoch": 2.4035755669591126, "grad_norm": 4.135829925537109, "learning_rate": 1.156775046055424e-06, "loss": 0.3925, "step": 29040 }, { "epoch": 2.4044032444959442, "grad_norm": 4.85306453704834, "learning_rate": 1.153696604123583e-06, "loss": 0.4363, "step": 29050 }, { "epoch": 2.405230922032776, "grad_norm": 4.742678642272949, "learning_rate": 1.1506217295662957e-06, "loss": 0.345, "step": 29060 }, { "epoch": 2.4060585995696075, "grad_norm": 5.0223236083984375, "learning_rate": 1.1475504252354452e-06, "loss": 0.4686, "step": 29070 }, { "epoch": 2.406886277106439, "grad_norm": 5.215400218963623, "learning_rate": 1.1444826939796116e-06, "loss": 0.372, "step": 29080 }, { "epoch": 2.407713954643271, "grad_norm": 4.28311014175415, "learning_rate": 1.1414185386440562e-06, "loss": 0.3649, "step": 29090 }, { "epoch": 2.4085416321801025, "grad_norm": 3.2143168449401855, "learning_rate": 1.138357962070727e-06, "loss": 0.4026, "step": 29100 }, { "epoch": 2.409369309716934, "grad_norm": 3.557802200317383, "learning_rate": 1.1353009670982518e-06, "loss": 0.3966, "step": 29110 }, { "epoch": 2.410196987253766, "grad_norm": 4.6892781257629395, "learning_rate": 1.1322475565619357e-06, "loss": 0.377, "step": 29120 }, { "epoch": 2.4110246647905975, "grad_norm": 4.629303455352783, "learning_rate": 1.1291977332937615e-06, "loss": 0.4151, "step": 29130 }, { "epoch": 2.411852342327429, "grad_norm": 4.320498943328857, "learning_rate": 1.126151500122381e-06, "loss": 0.4139, "step": 29140 }, { "epoch": 2.412680019864261, "grad_norm": 4.676682949066162, "learning_rate": 1.1231088598731198e-06, "loss": 0.4345, "step": 29150 }, { "epoch": 2.4135076974010925, "grad_norm": 3.9916861057281494, "learning_rate": 1.1200698153679712e-06, "loss": 0.3833, "step": 29160 }, { "epoch": 2.414335374937924, "grad_norm": 5.1546735763549805, "learning_rate": 1.117034369425588e-06, "loss": 0.3972, "step": 29170 }, { "epoch": 2.415163052474756, "grad_norm": 4.868451118469238, "learning_rate": 1.1140025248612956e-06, "loss": 0.3916, "step": 29180 }, { "epoch": 2.4159907300115875, "grad_norm": 4.451469421386719, "learning_rate": 1.110974284487068e-06, "loss": 0.4472, "step": 29190 }, { "epoch": 2.416818407548419, "grad_norm": 5.653703212738037, "learning_rate": 1.1079496511115439e-06, "loss": 0.4293, "step": 29200 }, { "epoch": 2.417646085085251, "grad_norm": 4.9112396240234375, "learning_rate": 1.104928627540015e-06, "loss": 0.4423, "step": 29210 }, { "epoch": 2.4184737626220825, "grad_norm": 5.552716255187988, "learning_rate": 1.1019112165744222e-06, "loss": 0.4123, "step": 29220 }, { "epoch": 2.419301440158914, "grad_norm": 3.8118607997894287, "learning_rate": 1.0988974210133586e-06, "loss": 0.4203, "step": 29230 }, { "epoch": 2.420129117695746, "grad_norm": 6.658673286437988, "learning_rate": 1.0958872436520635e-06, "loss": 0.4147, "step": 29240 }, { "epoch": 2.4209567952325775, "grad_norm": 5.1274285316467285, "learning_rate": 1.0928806872824206e-06, "loss": 0.3434, "step": 29250 }, { "epoch": 2.421784472769409, "grad_norm": 3.80479097366333, "learning_rate": 1.089877754692955e-06, "loss": 0.3873, "step": 29260 }, { "epoch": 2.4226121503062408, "grad_norm": 6.672432899475098, "learning_rate": 1.0868784486688283e-06, "loss": 0.4254, "step": 29270 }, { "epoch": 2.4234398278430724, "grad_norm": 5.2967729568481445, "learning_rate": 1.0838827719918444e-06, "loss": 0.4145, "step": 29280 }, { "epoch": 2.424267505379904, "grad_norm": 3.7274181842803955, "learning_rate": 1.0808907274404324e-06, "loss": 0.3855, "step": 29290 }, { "epoch": 2.4250951829167358, "grad_norm": 3.897632598876953, "learning_rate": 1.0779023177896635e-06, "loss": 0.3857, "step": 29300 }, { "epoch": 2.4259228604535674, "grad_norm": 6.3068647384643555, "learning_rate": 1.074917545811228e-06, "loss": 0.4045, "step": 29310 }, { "epoch": 2.426750537990399, "grad_norm": 5.782512664794922, "learning_rate": 1.0719364142734472e-06, "loss": 0.3423, "step": 29320 }, { "epoch": 2.4275782155272307, "grad_norm": 5.18137264251709, "learning_rate": 1.0689589259412665e-06, "loss": 0.3806, "step": 29330 }, { "epoch": 2.4284058930640624, "grad_norm": 5.095381736755371, "learning_rate": 1.0659850835762476e-06, "loss": 0.4386, "step": 29340 }, { "epoch": 2.429233570600894, "grad_norm": 3.7249982357025146, "learning_rate": 1.063014889936575e-06, "loss": 0.4235, "step": 29350 }, { "epoch": 2.4300612481377257, "grad_norm": 3.9944071769714355, "learning_rate": 1.0600483477770485e-06, "loss": 0.4943, "step": 29360 }, { "epoch": 2.430888925674557, "grad_norm": 3.9241442680358887, "learning_rate": 1.0570854598490803e-06, "loss": 0.3724, "step": 29370 }, { "epoch": 2.431716603211389, "grad_norm": 7.393401622772217, "learning_rate": 1.0541262289006942e-06, "loss": 0.3857, "step": 29380 }, { "epoch": 2.4325442807482203, "grad_norm": 5.896998405456543, "learning_rate": 1.05117065767652e-06, "loss": 0.3338, "step": 29390 }, { "epoch": 2.4333719582850524, "grad_norm": 4.8336639404296875, "learning_rate": 1.0482187489177975e-06, "loss": 0.3849, "step": 29400 }, { "epoch": 2.4341996358218836, "grad_norm": 5.6089863777160645, "learning_rate": 1.0452705053623618e-06, "loss": 0.3815, "step": 29410 }, { "epoch": 2.4350273133587153, "grad_norm": 4.182496547698975, "learning_rate": 1.0423259297446585e-06, "loss": 0.4293, "step": 29420 }, { "epoch": 2.435854990895547, "grad_norm": 3.5278141498565674, "learning_rate": 1.0393850247957271e-06, "loss": 0.4182, "step": 29430 }, { "epoch": 2.4366826684323786, "grad_norm": 5.077456474304199, "learning_rate": 1.0364477932431978e-06, "loss": 0.4064, "step": 29440 }, { "epoch": 2.4375103459692102, "grad_norm": 4.215777397155762, "learning_rate": 1.0335142378113017e-06, "loss": 0.3968, "step": 29450 }, { "epoch": 2.438338023506042, "grad_norm": 6.61751651763916, "learning_rate": 1.0305843612208537e-06, "loss": 0.4615, "step": 29460 }, { "epoch": 2.4391657010428736, "grad_norm": 4.5991291999816895, "learning_rate": 1.0276581661892615e-06, "loss": 0.4373, "step": 29470 }, { "epoch": 2.439993378579705, "grad_norm": 3.251699924468994, "learning_rate": 1.0247356554305155e-06, "loss": 0.3887, "step": 29480 }, { "epoch": 2.440821056116537, "grad_norm": 6.756894111633301, "learning_rate": 1.0218168316551902e-06, "loss": 0.3941, "step": 29490 }, { "epoch": 2.4416487336533685, "grad_norm": 4.758742809295654, "learning_rate": 1.0189016975704413e-06, "loss": 0.4412, "step": 29500 }, { "epoch": 2.4424764111902, "grad_norm": 5.780791759490967, "learning_rate": 1.0159902558799988e-06, "loss": 0.4015, "step": 29510 }, { "epoch": 2.443304088727032, "grad_norm": 5.445563793182373, "learning_rate": 1.0130825092841718e-06, "loss": 0.3321, "step": 29520 }, { "epoch": 2.4441317662638635, "grad_norm": 4.327962398529053, "learning_rate": 1.0101784604798404e-06, "loss": 0.4345, "step": 29530 }, { "epoch": 2.444959443800695, "grad_norm": 4.556379795074463, "learning_rate": 1.0072781121604563e-06, "loss": 0.2988, "step": 29540 }, { "epoch": 2.445787121337527, "grad_norm": 4.81104040145874, "learning_rate": 1.0043814670160395e-06, "loss": 0.4156, "step": 29550 }, { "epoch": 2.4466147988743585, "grad_norm": 3.1925878524780273, "learning_rate": 1.001488527733172e-06, "loss": 0.4077, "step": 29560 }, { "epoch": 2.44744247641119, "grad_norm": 4.382953643798828, "learning_rate": 9.985992969950014e-07, "loss": 0.4169, "step": 29570 }, { "epoch": 2.448270153948022, "grad_norm": 3.515422821044922, "learning_rate": 9.95713777481237e-07, "loss": 0.3931, "step": 29580 }, { "epoch": 2.4490978314848535, "grad_norm": 3.645803213119507, "learning_rate": 9.928319718681407e-07, "loss": 0.3567, "step": 29590 }, { "epoch": 2.449925509021685, "grad_norm": 4.392451763153076, "learning_rate": 9.899538828285382e-07, "loss": 0.4343, "step": 29600 }, { "epoch": 2.450753186558517, "grad_norm": 4.955005645751953, "learning_rate": 9.870795130317988e-07, "loss": 0.4251, "step": 29610 }, { "epoch": 2.4515808640953485, "grad_norm": 5.34551477432251, "learning_rate": 9.842088651438509e-07, "loss": 0.4192, "step": 29620 }, { "epoch": 2.45240854163218, "grad_norm": 5.435704708099365, "learning_rate": 9.813419418271624e-07, "loss": 0.3581, "step": 29630 }, { "epoch": 2.453236219169012, "grad_norm": 5.041081428527832, "learning_rate": 9.784787457407541e-07, "loss": 0.3738, "step": 29640 }, { "epoch": 2.4540638967058435, "grad_norm": 4.534113883972168, "learning_rate": 9.75619279540186e-07, "loss": 0.417, "step": 29650 }, { "epoch": 2.454891574242675, "grad_norm": 4.6946868896484375, "learning_rate": 9.727635458775592e-07, "loss": 0.359, "step": 29660 }, { "epoch": 2.455719251779507, "grad_norm": 4.957231521606445, "learning_rate": 9.699115474015157e-07, "loss": 0.4018, "step": 29670 }, { "epoch": 2.4565469293163384, "grad_norm": 4.421084403991699, "learning_rate": 9.670632867572278e-07, "loss": 0.3498, "step": 29680 }, { "epoch": 2.45737460685317, "grad_norm": 4.3222975730896, "learning_rate": 9.642187665864044e-07, "loss": 0.373, "step": 29690 }, { "epoch": 2.4582022843900018, "grad_norm": 5.130385875701904, "learning_rate": 9.613779895272874e-07, "loss": 0.4225, "step": 29700 }, { "epoch": 2.4590299619268334, "grad_norm": 3.658276319503784, "learning_rate": 9.585409582146404e-07, "loss": 0.3744, "step": 29710 }, { "epoch": 2.459857639463665, "grad_norm": 4.543152809143066, "learning_rate": 9.557076752797612e-07, "loss": 0.4114, "step": 29720 }, { "epoch": 2.4606853170004968, "grad_norm": 5.665431022644043, "learning_rate": 9.528781433504636e-07, "loss": 0.4518, "step": 29730 }, { "epoch": 2.4615129945373284, "grad_norm": 4.464978218078613, "learning_rate": 9.500523650510873e-07, "loss": 0.3549, "step": 29740 }, { "epoch": 2.46234067207416, "grad_norm": 3.8190176486968994, "learning_rate": 9.472303430024893e-07, "loss": 0.3489, "step": 29750 }, { "epoch": 2.4631683496109917, "grad_norm": 4.909446716308594, "learning_rate": 9.444120798220391e-07, "loss": 0.3925, "step": 29760 }, { "epoch": 2.4639960271478234, "grad_norm": 3.602609634399414, "learning_rate": 9.415975781236281e-07, "loss": 0.3855, "step": 29770 }, { "epoch": 2.4648237046846546, "grad_norm": 4.099685192108154, "learning_rate": 9.387868405176498e-07, "loss": 0.374, "step": 29780 }, { "epoch": 2.4656513822214867, "grad_norm": 3.793269157409668, "learning_rate": 9.359798696110139e-07, "loss": 0.3662, "step": 29790 }, { "epoch": 2.466479059758318, "grad_norm": 3.3377091884613037, "learning_rate": 9.331766680071308e-07, "loss": 0.4217, "step": 29800 }, { "epoch": 2.46730673729515, "grad_norm": 4.998449325561523, "learning_rate": 9.30377238305919e-07, "loss": 0.4174, "step": 29810 }, { "epoch": 2.4681344148319813, "grad_norm": 4.803717136383057, "learning_rate": 9.275815831037971e-07, "loss": 0.3701, "step": 29820 }, { "epoch": 2.468962092368813, "grad_norm": 4.3982625007629395, "learning_rate": 9.247897049936827e-07, "loss": 0.4401, "step": 29830 }, { "epoch": 2.4697897699056446, "grad_norm": 4.866006851196289, "learning_rate": 9.220016065649928e-07, "loss": 0.3815, "step": 29840 }, { "epoch": 2.4706174474424762, "grad_norm": 4.655172348022461, "learning_rate": 9.192172904036334e-07, "loss": 0.3883, "step": 29850 }, { "epoch": 2.471445124979308, "grad_norm": 4.236408710479736, "learning_rate": 9.16436759092007e-07, "loss": 0.4071, "step": 29860 }, { "epoch": 2.4722728025161396, "grad_norm": 4.916239261627197, "learning_rate": 9.136600152090058e-07, "loss": 0.4522, "step": 29870 }, { "epoch": 2.4731004800529712, "grad_norm": 2.7913360595703125, "learning_rate": 9.108870613300041e-07, "loss": 0.4049, "step": 29880 }, { "epoch": 2.473928157589803, "grad_norm": 4.650424480438232, "learning_rate": 9.081179000268703e-07, "loss": 0.4597, "step": 29890 }, { "epoch": 2.4747558351266346, "grad_norm": 3.27610182762146, "learning_rate": 9.053525338679453e-07, "loss": 0.359, "step": 29900 }, { "epoch": 2.475583512663466, "grad_norm": 3.7831413745880127, "learning_rate": 9.02590965418056e-07, "loss": 0.4, "step": 29910 }, { "epoch": 2.476411190200298, "grad_norm": 3.6940393447875977, "learning_rate": 8.998331972385072e-07, "loss": 0.3634, "step": 29920 }, { "epoch": 2.4772388677371295, "grad_norm": 4.775101661682129, "learning_rate": 8.970792318870747e-07, "loss": 0.3792, "step": 29930 }, { "epoch": 2.478066545273961, "grad_norm": 5.300714492797852, "learning_rate": 8.943290719180109e-07, "loss": 0.4034, "step": 29940 }, { "epoch": 2.478894222810793, "grad_norm": 7.991399765014648, "learning_rate": 8.915827198820376e-07, "loss": 0.4302, "step": 29950 }, { "epoch": 2.4797219003476245, "grad_norm": 3.9490270614624023, "learning_rate": 8.888401783263473e-07, "loss": 0.3225, "step": 29960 }, { "epoch": 2.480549577884456, "grad_norm": 4.312414169311523, "learning_rate": 8.861014497945924e-07, "loss": 0.4916, "step": 29970 }, { "epoch": 2.481377255421288, "grad_norm": 5.109755516052246, "learning_rate": 8.833665368268946e-07, "loss": 0.3682, "step": 29980 }, { "epoch": 2.4822049329581195, "grad_norm": 5.653954029083252, "learning_rate": 8.806354419598351e-07, "loss": 0.3571, "step": 29990 }, { "epoch": 2.483032610494951, "grad_norm": 5.347287654876709, "learning_rate": 8.7790816772645e-07, "loss": 0.4241, "step": 30000 }, { "epoch": 2.483860288031783, "grad_norm": 4.241307735443115, "learning_rate": 8.7518471665624e-07, "loss": 0.3907, "step": 30010 }, { "epoch": 2.4846879655686145, "grad_norm": 5.2339324951171875, "learning_rate": 8.72465091275152e-07, "loss": 0.3444, "step": 30020 }, { "epoch": 2.485515643105446, "grad_norm": 3.16575288772583, "learning_rate": 8.697492941055879e-07, "loss": 0.4043, "step": 30030 }, { "epoch": 2.486343320642278, "grad_norm": 5.273294925689697, "learning_rate": 8.670373276664007e-07, "loss": 0.3637, "step": 30040 }, { "epoch": 2.4871709981791095, "grad_norm": 5.452273368835449, "learning_rate": 8.643291944728866e-07, "loss": 0.4319, "step": 30050 }, { "epoch": 2.487998675715941, "grad_norm": 3.460118532180786, "learning_rate": 8.616248970367891e-07, "loss": 0.4647, "step": 30060 }, { "epoch": 2.488826353252773, "grad_norm": 5.0418524742126465, "learning_rate": 8.589244378662936e-07, "loss": 0.4131, "step": 30070 }, { "epoch": 2.4896540307896045, "grad_norm": 4.844895362854004, "learning_rate": 8.562278194660256e-07, "loss": 0.3642, "step": 30080 }, { "epoch": 2.490481708326436, "grad_norm": 5.085948467254639, "learning_rate": 8.535350443370499e-07, "loss": 0.4464, "step": 30090 }, { "epoch": 2.4913093858632678, "grad_norm": 3.8857643604278564, "learning_rate": 8.508461149768621e-07, "loss": 0.3235, "step": 30100 }, { "epoch": 2.4921370634000994, "grad_norm": 4.636258125305176, "learning_rate": 8.481610338793967e-07, "loss": 0.4627, "step": 30110 }, { "epoch": 2.492964740936931, "grad_norm": 5.40024471282959, "learning_rate": 8.454798035350131e-07, "loss": 0.3442, "step": 30120 }, { "epoch": 2.4937924184737628, "grad_norm": 7.09425687789917, "learning_rate": 8.428024264305046e-07, "loss": 0.402, "step": 30130 }, { "epoch": 2.4946200960105944, "grad_norm": 4.147519588470459, "learning_rate": 8.401289050490902e-07, "loss": 0.3259, "step": 30140 }, { "epoch": 2.495447773547426, "grad_norm": 4.252360820770264, "learning_rate": 8.374592418704069e-07, "loss": 0.4395, "step": 30150 }, { "epoch": 2.4962754510842577, "grad_norm": 4.382754325866699, "learning_rate": 8.347934393705215e-07, "loss": 0.3714, "step": 30160 }, { "epoch": 2.4971031286210894, "grad_norm": 5.618997573852539, "learning_rate": 8.321315000219127e-07, "loss": 0.4056, "step": 30170 }, { "epoch": 2.497930806157921, "grad_norm": 4.167751312255859, "learning_rate": 8.294734262934801e-07, "loss": 0.3979, "step": 30180 }, { "epoch": 2.4987584836947523, "grad_norm": 4.432166576385498, "learning_rate": 8.268192206505388e-07, "loss": 0.3924, "step": 30190 }, { "epoch": 2.4995861612315844, "grad_norm": 5.823348045349121, "learning_rate": 8.241688855548141e-07, "loss": 0.323, "step": 30200 }, { "epoch": 2.5004138387684156, "grad_norm": 5.612451553344727, "learning_rate": 8.215224234644436e-07, "loss": 0.4305, "step": 30210 }, { "epoch": 2.5012415163052477, "grad_norm": 4.6764302253723145, "learning_rate": 8.188798368339695e-07, "loss": 0.431, "step": 30220 }, { "epoch": 2.502069193842079, "grad_norm": 3.4053902626037598, "learning_rate": 8.162411281143423e-07, "loss": 0.3877, "step": 30230 }, { "epoch": 2.502896871378911, "grad_norm": 4.549859046936035, "learning_rate": 8.136062997529148e-07, "loss": 0.4221, "step": 30240 }, { "epoch": 2.5037245489157423, "grad_norm": 5.075228691101074, "learning_rate": 8.109753541934418e-07, "loss": 0.3989, "step": 30250 }, { "epoch": 2.504552226452574, "grad_norm": 6.758948802947998, "learning_rate": 8.083482938760768e-07, "loss": 0.4591, "step": 30260 }, { "epoch": 2.5053799039894056, "grad_norm": 5.4567551612854, "learning_rate": 8.057251212373674e-07, "loss": 0.4533, "step": 30270 }, { "epoch": 2.5062075815262372, "grad_norm": 4.0089874267578125, "learning_rate": 8.031058387102586e-07, "loss": 0.4018, "step": 30280 }, { "epoch": 2.507035259063069, "grad_norm": 6.7961812019348145, "learning_rate": 8.004904487240844e-07, "loss": 0.376, "step": 30290 }, { "epoch": 2.5078629365999006, "grad_norm": 4.31401252746582, "learning_rate": 7.978789537045701e-07, "loss": 0.4395, "step": 30300 }, { "epoch": 2.5086906141367322, "grad_norm": 3.8522799015045166, "learning_rate": 7.952713560738317e-07, "loss": 0.3586, "step": 30310 }, { "epoch": 2.509518291673564, "grad_norm": 5.486454010009766, "learning_rate": 7.926676582503639e-07, "loss": 0.4088, "step": 30320 }, { "epoch": 2.5103459692103955, "grad_norm": 4.603687763214111, "learning_rate": 7.900678626490505e-07, "loss": 0.3568, "step": 30330 }, { "epoch": 2.511173646747227, "grad_norm": 5.146494388580322, "learning_rate": 7.874719716811507e-07, "loss": 0.3672, "step": 30340 }, { "epoch": 2.512001324284059, "grad_norm": 3.956376075744629, "learning_rate": 7.848799877543062e-07, "loss": 0.3953, "step": 30350 }, { "epoch": 2.5128290018208905, "grad_norm": 3.049837350845337, "learning_rate": 7.822919132725337e-07, "loss": 0.3734, "step": 30360 }, { "epoch": 2.513656679357722, "grad_norm": 4.349221706390381, "learning_rate": 7.797077506362238e-07, "loss": 0.3657, "step": 30370 }, { "epoch": 2.514484356894554, "grad_norm": 4.3254547119140625, "learning_rate": 7.771275022421398e-07, "loss": 0.4473, "step": 30380 }, { "epoch": 2.5153120344313855, "grad_norm": 3.684570550918579, "learning_rate": 7.745511704834125e-07, "loss": 0.3649, "step": 30390 }, { "epoch": 2.516139711968217, "grad_norm": 3.9014892578125, "learning_rate": 7.719787577495419e-07, "loss": 0.4052, "step": 30400 }, { "epoch": 2.516967389505049, "grad_norm": 4.6407036781311035, "learning_rate": 7.694102664263931e-07, "loss": 0.4479, "step": 30410 }, { "epoch": 2.5177950670418805, "grad_norm": 5.0361785888671875, "learning_rate": 7.668456988961942e-07, "loss": 0.3816, "step": 30420 }, { "epoch": 2.518622744578712, "grad_norm": 3.2392258644104004, "learning_rate": 7.642850575375338e-07, "loss": 0.4261, "step": 30430 }, { "epoch": 2.519450422115544, "grad_norm": 4.5296220779418945, "learning_rate": 7.61728344725357e-07, "loss": 0.4446, "step": 30440 }, { "epoch": 2.5202780996523755, "grad_norm": 4.950967788696289, "learning_rate": 7.591755628309693e-07, "loss": 0.3971, "step": 30450 }, { "epoch": 2.521105777189207, "grad_norm": 5.266068935394287, "learning_rate": 7.566267142220279e-07, "loss": 0.3959, "step": 30460 }, { "epoch": 2.521933454726039, "grad_norm": 5.000637054443359, "learning_rate": 7.5408180126254e-07, "loss": 0.3789, "step": 30470 }, { "epoch": 2.5227611322628705, "grad_norm": 5.223601818084717, "learning_rate": 7.515408263128688e-07, "loss": 0.3691, "step": 30480 }, { "epoch": 2.523588809799702, "grad_norm": 4.452635765075684, "learning_rate": 7.490037917297178e-07, "loss": 0.3796, "step": 30490 }, { "epoch": 2.524416487336534, "grad_norm": 5.618330955505371, "learning_rate": 7.464706998661414e-07, "loss": 0.3822, "step": 30500 }, { "epoch": 2.5252441648733654, "grad_norm": 5.763148784637451, "learning_rate": 7.439415530715327e-07, "loss": 0.4364, "step": 30510 }, { "epoch": 2.526071842410197, "grad_norm": 5.61473274230957, "learning_rate": 7.414163536916291e-07, "loss": 0.4382, "step": 30520 }, { "epoch": 2.5268995199470288, "grad_norm": 4.865473747253418, "learning_rate": 7.388951040685066e-07, "loss": 0.3407, "step": 30530 }, { "epoch": 2.5277271974838604, "grad_norm": 7.3130011558532715, "learning_rate": 7.363778065405763e-07, "loss": 0.353, "step": 30540 }, { "epoch": 2.528554875020692, "grad_norm": 3.9308934211730957, "learning_rate": 7.338644634425862e-07, "loss": 0.4026, "step": 30550 }, { "epoch": 2.5293825525575233, "grad_norm": 4.30000114440918, "learning_rate": 7.313550771056122e-07, "loss": 0.429, "step": 30560 }, { "epoch": 2.5302102300943554, "grad_norm": 4.638671398162842, "learning_rate": 7.288496498570652e-07, "loss": 0.3752, "step": 30570 }, { "epoch": 2.5310379076311866, "grad_norm": 5.034365653991699, "learning_rate": 7.263481840206832e-07, "loss": 0.4724, "step": 30580 }, { "epoch": 2.5318655851680187, "grad_norm": 4.124532699584961, "learning_rate": 7.238506819165253e-07, "loss": 0.355, "step": 30590 }, { "epoch": 2.53269326270485, "grad_norm": 4.242930889129639, "learning_rate": 7.213571458609825e-07, "loss": 0.4393, "step": 30600 }, { "epoch": 2.533520940241682, "grad_norm": 4.479305744171143, "learning_rate": 7.188675781667598e-07, "loss": 0.4067, "step": 30610 }, { "epoch": 2.5343486177785133, "grad_norm": 5.0444231033325195, "learning_rate": 7.163819811428857e-07, "loss": 0.3616, "step": 30620 }, { "epoch": 2.5351762953153454, "grad_norm": 4.909320831298828, "learning_rate": 7.139003570947062e-07, "loss": 0.4335, "step": 30630 }, { "epoch": 2.5360039728521766, "grad_norm": 3.43121337890625, "learning_rate": 7.114227083238789e-07, "loss": 0.3616, "step": 30640 }, { "epoch": 2.5368316503890087, "grad_norm": 3.9701597690582275, "learning_rate": 7.08949037128378e-07, "loss": 0.3403, "step": 30650 }, { "epoch": 2.53765932792584, "grad_norm": 5.008255481719971, "learning_rate": 7.064793458024877e-07, "loss": 0.398, "step": 30660 }, { "epoch": 2.5384870054626716, "grad_norm": 4.581993579864502, "learning_rate": 7.040136366368005e-07, "loss": 0.4618, "step": 30670 }, { "epoch": 2.5393146829995032, "grad_norm": 5.3673787117004395, "learning_rate": 7.015519119182146e-07, "loss": 0.4002, "step": 30680 }, { "epoch": 2.540142360536335, "grad_norm": 2.9999492168426514, "learning_rate": 6.990941739299345e-07, "loss": 0.378, "step": 30690 }, { "epoch": 2.5409700380731666, "grad_norm": 4.770166397094727, "learning_rate": 6.966404249514668e-07, "loss": 0.3952, "step": 30700 }, { "epoch": 2.5417977156099982, "grad_norm": 4.961124420166016, "learning_rate": 6.941906672586151e-07, "loss": 0.4097, "step": 30710 }, { "epoch": 2.54262539314683, "grad_norm": 4.248905181884766, "learning_rate": 6.917449031234885e-07, "loss": 0.3056, "step": 30720 }, { "epoch": 2.5434530706836616, "grad_norm": 5.165794849395752, "learning_rate": 6.893031348144835e-07, "loss": 0.4874, "step": 30730 }, { "epoch": 2.544280748220493, "grad_norm": 5.4268364906311035, "learning_rate": 6.868653645962963e-07, "loss": 0.3859, "step": 30740 }, { "epoch": 2.545108425757325, "grad_norm": 4.585019588470459, "learning_rate": 6.84431594729914e-07, "loss": 0.4378, "step": 30750 }, { "epoch": 2.5459361032941565, "grad_norm": 4.342596054077148, "learning_rate": 6.820018274726103e-07, "loss": 0.4021, "step": 30760 }, { "epoch": 2.546763780830988, "grad_norm": 3.620152235031128, "learning_rate": 6.79576065077951e-07, "loss": 0.3327, "step": 30770 }, { "epoch": 2.54759145836782, "grad_norm": 5.424134254455566, "learning_rate": 6.771543097957844e-07, "loss": 0.3905, "step": 30780 }, { "epoch": 2.5484191359046515, "grad_norm": 5.996721267700195, "learning_rate": 6.747365638722437e-07, "loss": 0.3741, "step": 30790 }, { "epoch": 2.549246813441483, "grad_norm": 3.389099359512329, "learning_rate": 6.723228295497442e-07, "loss": 0.4022, "step": 30800 }, { "epoch": 2.550074490978315, "grad_norm": 4.43535041809082, "learning_rate": 6.69913109066977e-07, "loss": 0.4259, "step": 30810 }, { "epoch": 2.5509021685151465, "grad_norm": 4.8189496994018555, "learning_rate": 6.675074046589153e-07, "loss": 0.3773, "step": 30820 }, { "epoch": 2.551729846051978, "grad_norm": 3.210176706314087, "learning_rate": 6.651057185568016e-07, "loss": 0.3569, "step": 30830 }, { "epoch": 2.55255752358881, "grad_norm": 4.344388008117676, "learning_rate": 6.627080529881591e-07, "loss": 0.4426, "step": 30840 }, { "epoch": 2.5533852011256415, "grad_norm": 3.749138832092285, "learning_rate": 6.60314410176775e-07, "loss": 0.423, "step": 30850 }, { "epoch": 2.554212878662473, "grad_norm": 6.291297435760498, "learning_rate": 6.57924792342709e-07, "loss": 0.4246, "step": 30860 }, { "epoch": 2.555040556199305, "grad_norm": 5.303653240203857, "learning_rate": 6.555392017022876e-07, "loss": 0.4219, "step": 30870 }, { "epoch": 2.5558682337361365, "grad_norm": 3.891221046447754, "learning_rate": 6.531576404681006e-07, "loss": 0.4116, "step": 30880 }, { "epoch": 2.556695911272968, "grad_norm": 4.171040058135986, "learning_rate": 6.507801108490014e-07, "loss": 0.3737, "step": 30890 }, { "epoch": 2.5575235888098, "grad_norm": 4.169492244720459, "learning_rate": 6.484066150501056e-07, "loss": 0.3481, "step": 30900 }, { "epoch": 2.5583512663466315, "grad_norm": 3.786961317062378, "learning_rate": 6.460371552727851e-07, "loss": 0.4116, "step": 30910 }, { "epoch": 2.559178943883463, "grad_norm": 4.80360746383667, "learning_rate": 6.436717337146714e-07, "loss": 0.4347, "step": 30920 }, { "epoch": 2.560006621420295, "grad_norm": 5.496336936950684, "learning_rate": 6.413103525696457e-07, "loss": 0.3634, "step": 30930 }, { "epoch": 2.5608342989571264, "grad_norm": 4.488288879394531, "learning_rate": 6.389530140278471e-07, "loss": 0.3776, "step": 30940 }, { "epoch": 2.561661976493958, "grad_norm": 5.021377086639404, "learning_rate": 6.365997202756624e-07, "loss": 0.3276, "step": 30950 }, { "epoch": 2.5624896540307898, "grad_norm": 3.7784087657928467, "learning_rate": 6.342504734957278e-07, "loss": 0.3602, "step": 30960 }, { "epoch": 2.563317331567621, "grad_norm": 5.310364723205566, "learning_rate": 6.31905275866927e-07, "loss": 0.3687, "step": 30970 }, { "epoch": 2.564145009104453, "grad_norm": 5.222259998321533, "learning_rate": 6.295641295643845e-07, "loss": 0.3714, "step": 30980 }, { "epoch": 2.5649726866412843, "grad_norm": 4.007057189941406, "learning_rate": 6.272270367594718e-07, "loss": 0.398, "step": 30990 }, { "epoch": 2.5658003641781164, "grad_norm": 5.683523654937744, "learning_rate": 6.248939996197978e-07, "loss": 0.3983, "step": 31000 }, { "epoch": 2.5666280417149476, "grad_norm": 3.4667980670928955, "learning_rate": 6.225650203092093e-07, "loss": 0.3962, "step": 31010 }, { "epoch": 2.5674557192517797, "grad_norm": 6.592273235321045, "learning_rate": 6.202401009877956e-07, "loss": 0.388, "step": 31020 }, { "epoch": 2.568283396788611, "grad_norm": 5.676321029663086, "learning_rate": 6.179192438118714e-07, "loss": 0.3916, "step": 31030 }, { "epoch": 2.569111074325443, "grad_norm": 3.8077549934387207, "learning_rate": 6.15602450933992e-07, "loss": 0.3441, "step": 31040 }, { "epoch": 2.5699387518622743, "grad_norm": 3.8724429607391357, "learning_rate": 6.132897245029362e-07, "loss": 0.3484, "step": 31050 }, { "epoch": 2.5707664293991064, "grad_norm": 3.3436968326568604, "learning_rate": 6.109810666637167e-07, "loss": 0.3599, "step": 31060 }, { "epoch": 2.5715941069359376, "grad_norm": 4.470296859741211, "learning_rate": 6.086764795575695e-07, "loss": 0.3949, "step": 31070 }, { "epoch": 2.5724217844727693, "grad_norm": 5.285587310791016, "learning_rate": 6.063759653219565e-07, "loss": 0.4037, "step": 31080 }, { "epoch": 2.573249462009601, "grad_norm": 3.4945266246795654, "learning_rate": 6.04079526090563e-07, "loss": 0.3764, "step": 31090 }, { "epoch": 2.5740771395464326, "grad_norm": 4.5235490798950195, "learning_rate": 6.01787163993291e-07, "loss": 0.3806, "step": 31100 }, { "epoch": 2.5749048170832642, "grad_norm": 3.7682793140411377, "learning_rate": 5.994988811562647e-07, "loss": 0.332, "step": 31110 }, { "epoch": 2.575732494620096, "grad_norm": 4.048414707183838, "learning_rate": 5.972146797018235e-07, "loss": 0.333, "step": 31120 }, { "epoch": 2.5765601721569276, "grad_norm": 5.788407802581787, "learning_rate": 5.94934561748522e-07, "loss": 0.3737, "step": 31130 }, { "epoch": 2.5773878496937592, "grad_norm": 4.50048828125, "learning_rate": 5.926585294111282e-07, "loss": 0.4361, "step": 31140 }, { "epoch": 2.578215527230591, "grad_norm": 4.807448863983154, "learning_rate": 5.903865848006163e-07, "loss": 0.4043, "step": 31150 }, { "epoch": 2.5790432047674225, "grad_norm": 4.822179794311523, "learning_rate": 5.881187300241758e-07, "loss": 0.4078, "step": 31160 }, { "epoch": 2.579870882304254, "grad_norm": 4.051673412322998, "learning_rate": 5.858549671851976e-07, "loss": 0.3156, "step": 31170 }, { "epoch": 2.580698559841086, "grad_norm": 3.842236280441284, "learning_rate": 5.835952983832782e-07, "loss": 0.3921, "step": 31180 }, { "epoch": 2.5815262373779175, "grad_norm": 6.228193283081055, "learning_rate": 5.813397257142228e-07, "loss": 0.3552, "step": 31190 }, { "epoch": 2.582353914914749, "grad_norm": 2.9241585731506348, "learning_rate": 5.790882512700285e-07, "loss": 0.3581, "step": 31200 }, { "epoch": 2.583181592451581, "grad_norm": 5.41754674911499, "learning_rate": 5.768408771388989e-07, "loss": 0.3576, "step": 31210 }, { "epoch": 2.5840092699884125, "grad_norm": 4.764238357543945, "learning_rate": 5.745976054052288e-07, "loss": 0.433, "step": 31220 }, { "epoch": 2.584836947525244, "grad_norm": 4.028949737548828, "learning_rate": 5.723584381496122e-07, "loss": 0.4671, "step": 31230 }, { "epoch": 2.585664625062076, "grad_norm": 4.913052082061768, "learning_rate": 5.701233774488347e-07, "loss": 0.4049, "step": 31240 }, { "epoch": 2.5864923025989075, "grad_norm": 4.2850823402404785, "learning_rate": 5.678924253758734e-07, "loss": 0.4656, "step": 31250 }, { "epoch": 2.587319980135739, "grad_norm": 4.414880752563477, "learning_rate": 5.65665583999896e-07, "loss": 0.3355, "step": 31260 }, { "epoch": 2.588147657672571, "grad_norm": 5.565047264099121, "learning_rate": 5.634428553862537e-07, "loss": 0.3948, "step": 31270 }, { "epoch": 2.5889753352094025, "grad_norm": 4.7641777992248535, "learning_rate": 5.612242415964869e-07, "loss": 0.4358, "step": 31280 }, { "epoch": 2.589803012746234, "grad_norm": 4.746800899505615, "learning_rate": 5.590097446883203e-07, "loss": 0.427, "step": 31290 }, { "epoch": 2.590630690283066, "grad_norm": 3.040292501449585, "learning_rate": 5.56799366715654e-07, "loss": 0.4239, "step": 31300 }, { "epoch": 2.5914583678198975, "grad_norm": 4.912197113037109, "learning_rate": 5.545931097285772e-07, "loss": 0.4197, "step": 31310 }, { "epoch": 2.592286045356729, "grad_norm": 4.493854522705078, "learning_rate": 5.523909757733492e-07, "loss": 0.4741, "step": 31320 }, { "epoch": 2.593113722893561, "grad_norm": 5.2058491706848145, "learning_rate": 5.501929668924083e-07, "loss": 0.3648, "step": 31330 }, { "epoch": 2.5939414004303925, "grad_norm": 4.516733646392822, "learning_rate": 5.479990851243683e-07, "loss": 0.3639, "step": 31340 }, { "epoch": 2.594769077967224, "grad_norm": 4.7266740798950195, "learning_rate": 5.458093325040115e-07, "loss": 0.3914, "step": 31350 }, { "epoch": 2.5955967555040558, "grad_norm": 5.6341962814331055, "learning_rate": 5.436237110622932e-07, "loss": 0.3763, "step": 31360 }, { "epoch": 2.5964244330408874, "grad_norm": 4.550677299499512, "learning_rate": 5.414422228263371e-07, "loss": 0.4344, "step": 31370 }, { "epoch": 2.5972521105777187, "grad_norm": 3.826220989227295, "learning_rate": 5.39264869819433e-07, "loss": 0.3471, "step": 31380 }, { "epoch": 2.5980797881145508, "grad_norm": 4.57026481628418, "learning_rate": 5.370916540610338e-07, "loss": 0.3182, "step": 31390 }, { "epoch": 2.598907465651382, "grad_norm": 4.7249603271484375, "learning_rate": 5.349225775667572e-07, "loss": 0.417, "step": 31400 }, { "epoch": 2.599735143188214, "grad_norm": 4.737354755401611, "learning_rate": 5.327576423483821e-07, "loss": 0.3939, "step": 31410 }, { "epoch": 2.6005628207250453, "grad_norm": 5.305750846862793, "learning_rate": 5.305968504138426e-07, "loss": 0.3949, "step": 31420 }, { "epoch": 2.6013904982618774, "grad_norm": 4.192201614379883, "learning_rate": 5.284402037672364e-07, "loss": 0.4706, "step": 31430 }, { "epoch": 2.6022181757987086, "grad_norm": 4.691887855529785, "learning_rate": 5.26287704408811e-07, "loss": 0.395, "step": 31440 }, { "epoch": 2.6030458533355407, "grad_norm": 4.39100456237793, "learning_rate": 5.241393543349694e-07, "loss": 0.3538, "step": 31450 }, { "epoch": 2.603873530872372, "grad_norm": 4.340761184692383, "learning_rate": 5.219951555382675e-07, "loss": 0.4515, "step": 31460 }, { "epoch": 2.604701208409204, "grad_norm": 2.29512357711792, "learning_rate": 5.198551100074073e-07, "loss": 0.4022, "step": 31470 }, { "epoch": 2.6055288859460353, "grad_norm": 4.717724800109863, "learning_rate": 5.177192197272429e-07, "loss": 0.3561, "step": 31480 }, { "epoch": 2.606356563482867, "grad_norm": 5.959236145019531, "learning_rate": 5.155874866787719e-07, "loss": 0.3896, "step": 31490 }, { "epoch": 2.6071842410196986, "grad_norm": 5.550564765930176, "learning_rate": 5.134599128391371e-07, "loss": 0.3993, "step": 31500 }, { "epoch": 2.6080119185565303, "grad_norm": 4.355971336364746, "learning_rate": 5.113365001816245e-07, "loss": 0.4053, "step": 31510 }, { "epoch": 2.608839596093362, "grad_norm": 5.343199729919434, "learning_rate": 5.092172506756577e-07, "loss": 0.4232, "step": 31520 }, { "epoch": 2.6096672736301936, "grad_norm": 4.07174825668335, "learning_rate": 5.071021662868025e-07, "loss": 0.4389, "step": 31530 }, { "epoch": 2.6104949511670252, "grad_norm": 3.518461227416992, "learning_rate": 5.049912489767583e-07, "loss": 0.362, "step": 31540 }, { "epoch": 2.611322628703857, "grad_norm": 5.844148635864258, "learning_rate": 5.028845007033645e-07, "loss": 0.4018, "step": 31550 }, { "epoch": 2.6121503062406886, "grad_norm": 4.647653579711914, "learning_rate": 5.007819234205885e-07, "loss": 0.3476, "step": 31560 }, { "epoch": 2.61297798377752, "grad_norm": 3.817786693572998, "learning_rate": 4.986835190785316e-07, "loss": 0.4043, "step": 31570 }, { "epoch": 2.613805661314352, "grad_norm": 3.9685862064361572, "learning_rate": 4.965892896234259e-07, "loss": 0.331, "step": 31580 }, { "epoch": 2.6146333388511835, "grad_norm": 4.146365642547607, "learning_rate": 4.944992369976287e-07, "loss": 0.3563, "step": 31590 }, { "epoch": 2.615461016388015, "grad_norm": 4.786243438720703, "learning_rate": 4.924133631396249e-07, "loss": 0.3655, "step": 31600 }, { "epoch": 2.616288693924847, "grad_norm": 3.4640161991119385, "learning_rate": 4.903316699840244e-07, "loss": 0.3608, "step": 31610 }, { "epoch": 2.6171163714616785, "grad_norm": 4.534228801727295, "learning_rate": 4.882541594615581e-07, "loss": 0.411, "step": 31620 }, { "epoch": 2.61794404899851, "grad_norm": 3.9207332134246826, "learning_rate": 4.861808334990792e-07, "loss": 0.3847, "step": 31630 }, { "epoch": 2.618771726535342, "grad_norm": 7.119131088256836, "learning_rate": 4.841116940195573e-07, "loss": 0.4551, "step": 31640 }, { "epoch": 2.6195994040721735, "grad_norm": 4.494716167449951, "learning_rate": 4.820467429420811e-07, "loss": 0.3782, "step": 31650 }, { "epoch": 2.620427081609005, "grad_norm": 6.040804862976074, "learning_rate": 4.799859821818547e-07, "loss": 0.4601, "step": 31660 }, { "epoch": 2.621254759145837, "grad_norm": 4.3555588722229, "learning_rate": 4.779294136501944e-07, "loss": 0.3741, "step": 31670 }, { "epoch": 2.6220824366826685, "grad_norm": 4.2531418800354, "learning_rate": 4.7587703925453086e-07, "loss": 0.3657, "step": 31680 }, { "epoch": 2.6229101142195, "grad_norm": 4.720728874206543, "learning_rate": 4.73828860898401e-07, "loss": 0.4167, "step": 31690 }, { "epoch": 2.623737791756332, "grad_norm": 6.631955623626709, "learning_rate": 4.7178488048145287e-07, "loss": 0.4316, "step": 31700 }, { "epoch": 2.6245654692931635, "grad_norm": 4.8736891746521, "learning_rate": 4.697450998994396e-07, "loss": 0.4024, "step": 31710 }, { "epoch": 2.625393146829995, "grad_norm": 5.746395111083984, "learning_rate": 4.677095210442195e-07, "loss": 0.4496, "step": 31720 }, { "epoch": 2.626220824366827, "grad_norm": 4.306918144226074, "learning_rate": 4.656781458037546e-07, "loss": 0.388, "step": 31730 }, { "epoch": 2.6270485019036585, "grad_norm": 5.481353759765625, "learning_rate": 4.636509760621072e-07, "loss": 0.3937, "step": 31740 }, { "epoch": 2.62787617944049, "grad_norm": 4.1362385749816895, "learning_rate": 4.6162801369944e-07, "loss": 0.3085, "step": 31750 }, { "epoch": 2.628703856977322, "grad_norm": 4.8346943855285645, "learning_rate": 4.596092605920116e-07, "loss": 0.4312, "step": 31760 }, { "epoch": 2.6295315345141534, "grad_norm": 4.910533905029297, "learning_rate": 4.575947186121782e-07, "loss": 0.39, "step": 31770 }, { "epoch": 2.630359212050985, "grad_norm": 9.095038414001465, "learning_rate": 4.555843896283907e-07, "loss": 0.3653, "step": 31780 }, { "epoch": 2.6311868895878163, "grad_norm": 4.334075927734375, "learning_rate": 4.535782755051904e-07, "loss": 0.3885, "step": 31790 }, { "epoch": 2.6320145671246484, "grad_norm": 4.059500217437744, "learning_rate": 4.515763781032123e-07, "loss": 0.3579, "step": 31800 }, { "epoch": 2.6328422446614796, "grad_norm": 3.905716896057129, "learning_rate": 4.4957869927917674e-07, "loss": 0.3748, "step": 31810 }, { "epoch": 2.6336699221983118, "grad_norm": 5.899052143096924, "learning_rate": 4.4758524088589497e-07, "loss": 0.3835, "step": 31820 }, { "epoch": 2.634497599735143, "grad_norm": 4.210059642791748, "learning_rate": 4.4559600477226183e-07, "loss": 0.3517, "step": 31830 }, { "epoch": 2.635325277271975, "grad_norm": 3.7246906757354736, "learning_rate": 4.436109927832571e-07, "loss": 0.3462, "step": 31840 }, { "epoch": 2.6361529548088063, "grad_norm": 2.805030584335327, "learning_rate": 4.4163020675994316e-07, "loss": 0.4056, "step": 31850 }, { "epoch": 2.6369806323456384, "grad_norm": 4.3997039794921875, "learning_rate": 4.396536485394598e-07, "loss": 0.4596, "step": 31860 }, { "epoch": 2.6378083098824696, "grad_norm": 4.0674543380737305, "learning_rate": 4.3768131995503e-07, "loss": 0.3718, "step": 31870 }, { "epoch": 2.6386359874193013, "grad_norm": 4.5048627853393555, "learning_rate": 4.3571322283594954e-07, "loss": 0.3902, "step": 31880 }, { "epoch": 2.639463664956133, "grad_norm": 5.721669673919678, "learning_rate": 4.337493590075925e-07, "loss": 0.3447, "step": 31890 }, { "epoch": 2.6402913424929646, "grad_norm": 4.531260967254639, "learning_rate": 4.317897302914076e-07, "loss": 0.4565, "step": 31900 }, { "epoch": 2.6411190200297963, "grad_norm": 4.4966840744018555, "learning_rate": 4.2983433850491194e-07, "loss": 0.3969, "step": 31910 }, { "epoch": 2.641946697566628, "grad_norm": 4.069575786590576, "learning_rate": 4.278831854616966e-07, "loss": 0.3971, "step": 31920 }, { "epoch": 2.6427743751034596, "grad_norm": 8.150683403015137, "learning_rate": 4.2593627297141724e-07, "loss": 0.392, "step": 31930 }, { "epoch": 2.6436020526402912, "grad_norm": 5.04276180267334, "learning_rate": 4.239936028398006e-07, "loss": 0.4249, "step": 31940 }, { "epoch": 2.644429730177123, "grad_norm": 3.100745439529419, "learning_rate": 4.2205517686863597e-07, "loss": 0.3709, "step": 31950 }, { "epoch": 2.6452574077139546, "grad_norm": 5.363617897033691, "learning_rate": 4.201209968557779e-07, "loss": 0.4021, "step": 31960 }, { "epoch": 2.6460850852507862, "grad_norm": 3.9461581707000732, "learning_rate": 4.1819106459514256e-07, "loss": 0.4281, "step": 31970 }, { "epoch": 2.646912762787618, "grad_norm": 3.931255578994751, "learning_rate": 4.1626538187670486e-07, "loss": 0.3319, "step": 31980 }, { "epoch": 2.6477404403244496, "grad_norm": 4.1396636962890625, "learning_rate": 4.143439504864999e-07, "loss": 0.3998, "step": 31990 }, { "epoch": 2.648568117861281, "grad_norm": 4.775804042816162, "learning_rate": 4.1242677220662107e-07, "loss": 0.401, "step": 32000 }, { "epoch": 2.649395795398113, "grad_norm": 3.7101492881774902, "learning_rate": 4.1051384881521164e-07, "loss": 0.3734, "step": 32010 }, { "epoch": 2.6502234729349445, "grad_norm": 5.161265850067139, "learning_rate": 4.0860518208647724e-07, "loss": 0.3491, "step": 32020 }, { "epoch": 2.651051150471776, "grad_norm": 5.543527603149414, "learning_rate": 4.067007737906664e-07, "loss": 0.3991, "step": 32030 }, { "epoch": 2.651878828008608, "grad_norm": 5.706044673919678, "learning_rate": 4.048006256940856e-07, "loss": 0.4325, "step": 32040 }, { "epoch": 2.6527065055454395, "grad_norm": 4.780057907104492, "learning_rate": 4.029047395590835e-07, "loss": 0.3524, "step": 32050 }, { "epoch": 2.653534183082271, "grad_norm": 4.779311656951904, "learning_rate": 4.0101311714406066e-07, "loss": 0.3492, "step": 32060 }, { "epoch": 2.654361860619103, "grad_norm": 3.4569971561431885, "learning_rate": 3.9912576020346205e-07, "loss": 0.3408, "step": 32070 }, { "epoch": 2.6551895381559345, "grad_norm": 4.290452003479004, "learning_rate": 3.9724267048777464e-07, "loss": 0.3743, "step": 32080 }, { "epoch": 2.656017215692766, "grad_norm": 3.070537567138672, "learning_rate": 3.9536384974353135e-07, "loss": 0.4341, "step": 32090 }, { "epoch": 2.656844893229598, "grad_norm": 4.9330220222473145, "learning_rate": 3.9348929971330054e-07, "loss": 0.4096, "step": 32100 }, { "epoch": 2.6576725707664295, "grad_norm": 6.035449028015137, "learning_rate": 3.9161902213569305e-07, "loss": 0.3929, "step": 32110 }, { "epoch": 2.658500248303261, "grad_norm": 5.06627082824707, "learning_rate": 3.8975301874535797e-07, "loss": 0.4081, "step": 32120 }, { "epoch": 2.659327925840093, "grad_norm": 5.0348005294799805, "learning_rate": 3.87891291272976e-07, "loss": 0.3301, "step": 32130 }, { "epoch": 2.6601556033769245, "grad_norm": 6.148066997528076, "learning_rate": 3.8603384144526747e-07, "loss": 0.4315, "step": 32140 }, { "epoch": 2.660983280913756, "grad_norm": 5.760768890380859, "learning_rate": 3.8418067098498045e-07, "loss": 0.3527, "step": 32150 }, { "epoch": 2.661810958450588, "grad_norm": 3.6283059120178223, "learning_rate": 3.823317816108962e-07, "loss": 0.4099, "step": 32160 }, { "epoch": 2.6626386359874195, "grad_norm": 5.654618740081787, "learning_rate": 3.804871750378264e-07, "loss": 0.4265, "step": 32170 }, { "epoch": 2.663466313524251, "grad_norm": 3.8522374629974365, "learning_rate": 3.786468529766074e-07, "loss": 0.4736, "step": 32180 }, { "epoch": 2.664293991061083, "grad_norm": 3.4939448833465576, "learning_rate": 3.7681081713410395e-07, "loss": 0.3654, "step": 32190 }, { "epoch": 2.665121668597914, "grad_norm": 4.214698314666748, "learning_rate": 3.749790692132055e-07, "loss": 0.3671, "step": 32200 }, { "epoch": 2.665949346134746, "grad_norm": 3.8274528980255127, "learning_rate": 3.731516109128236e-07, "loss": 0.4217, "step": 32210 }, { "epoch": 2.6667770236715773, "grad_norm": 3.453131914138794, "learning_rate": 3.7132844392789304e-07, "loss": 0.379, "step": 32220 }, { "epoch": 2.6676047012084094, "grad_norm": 5.03052282333374, "learning_rate": 3.695095699493645e-07, "loss": 0.4376, "step": 32230 }, { "epoch": 2.6684323787452406, "grad_norm": 5.548057556152344, "learning_rate": 3.67694990664213e-07, "loss": 0.4226, "step": 32240 }, { "epoch": 2.6692600562820727, "grad_norm": 3.491267442703247, "learning_rate": 3.6588470775542286e-07, "loss": 0.3318, "step": 32250 }, { "epoch": 2.670087733818904, "grad_norm": 3.7329397201538086, "learning_rate": 3.640787229020015e-07, "loss": 0.367, "step": 32260 }, { "epoch": 2.670915411355736, "grad_norm": 4.087145805358887, "learning_rate": 3.622770377789636e-07, "loss": 0.3298, "step": 32270 }, { "epoch": 2.6717430888925673, "grad_norm": 4.028382778167725, "learning_rate": 3.6047965405733966e-07, "loss": 0.3904, "step": 32280 }, { "epoch": 2.672570766429399, "grad_norm": 4.379392147064209, "learning_rate": 3.5868657340417003e-07, "loss": 0.4064, "step": 32290 }, { "epoch": 2.6733984439662306, "grad_norm": 6.5227227210998535, "learning_rate": 3.5689779748250165e-07, "loss": 0.4499, "step": 32300 }, { "epoch": 2.6742261215030623, "grad_norm": 3.252995014190674, "learning_rate": 3.5511332795139186e-07, "loss": 0.4043, "step": 32310 }, { "epoch": 2.675053799039894, "grad_norm": 4.339319229125977, "learning_rate": 3.533331664659029e-07, "loss": 0.4183, "step": 32320 }, { "epoch": 2.6758814765767256, "grad_norm": 3.5538065433502197, "learning_rate": 3.515573146771012e-07, "loss": 0.3622, "step": 32330 }, { "epoch": 2.6767091541135573, "grad_norm": 4.8099822998046875, "learning_rate": 3.4978577423205704e-07, "loss": 0.3898, "step": 32340 }, { "epoch": 2.677536831650389, "grad_norm": 5.357463359832764, "learning_rate": 3.4801854677383896e-07, "loss": 0.4559, "step": 32350 }, { "epoch": 2.6783645091872206, "grad_norm": 4.338512897491455, "learning_rate": 3.462556339415196e-07, "loss": 0.4026, "step": 32360 }, { "epoch": 2.6791921867240522, "grad_norm": 5.338489532470703, "learning_rate": 3.4449703737016504e-07, "loss": 0.3883, "step": 32370 }, { "epoch": 2.680019864260884, "grad_norm": 6.001187324523926, "learning_rate": 3.4274275869084326e-07, "loss": 0.3775, "step": 32380 }, { "epoch": 2.6808475417977156, "grad_norm": 4.487107753753662, "learning_rate": 3.4099279953061503e-07, "loss": 0.4186, "step": 32390 }, { "epoch": 2.6816752193345472, "grad_norm": 13.832345962524414, "learning_rate": 3.3924716151253254e-07, "loss": 0.4507, "step": 32400 }, { "epoch": 2.682502896871379, "grad_norm": 4.7976765632629395, "learning_rate": 3.375058462556452e-07, "loss": 0.4479, "step": 32410 }, { "epoch": 2.6833305744082105, "grad_norm": 4.926270008087158, "learning_rate": 3.357688553749877e-07, "loss": 0.343, "step": 32420 }, { "epoch": 2.684158251945042, "grad_norm": 4.114280700683594, "learning_rate": 3.340361904815881e-07, "loss": 0.3953, "step": 32430 }, { "epoch": 2.684985929481874, "grad_norm": 5.29146671295166, "learning_rate": 3.323078531824603e-07, "loss": 0.4086, "step": 32440 }, { "epoch": 2.6858136070187055, "grad_norm": 5.49489164352417, "learning_rate": 3.3058384508060483e-07, "loss": 0.3928, "step": 32450 }, { "epoch": 2.686641284555537, "grad_norm": 4.928554058074951, "learning_rate": 3.2886416777500875e-07, "loss": 0.3554, "step": 32460 }, { "epoch": 2.687468962092369, "grad_norm": 3.4741992950439453, "learning_rate": 3.2714882286063733e-07, "loss": 0.4024, "step": 32470 }, { "epoch": 2.6882966396292005, "grad_norm": 5.41702127456665, "learning_rate": 3.254378119284435e-07, "loss": 0.4848, "step": 32480 }, { "epoch": 2.689124317166032, "grad_norm": 5.144486427307129, "learning_rate": 3.237311365653567e-07, "loss": 0.3708, "step": 32490 }, { "epoch": 2.689951994702864, "grad_norm": 4.018123149871826, "learning_rate": 3.220287983542875e-07, "loss": 0.3751, "step": 32500 }, { "epoch": 2.6907796722396955, "grad_norm": 6.001951694488525, "learning_rate": 3.203307988741222e-07, "loss": 0.4204, "step": 32510 }, { "epoch": 2.691607349776527, "grad_norm": 6.295750617980957, "learning_rate": 3.1863713969972343e-07, "loss": 0.4083, "step": 32520 }, { "epoch": 2.692435027313359, "grad_norm": 4.167087078094482, "learning_rate": 3.169478224019279e-07, "loss": 0.4291, "step": 32530 }, { "epoch": 2.6932627048501905, "grad_norm": 5.873100280761719, "learning_rate": 3.152628485475473e-07, "loss": 0.4164, "step": 32540 }, { "epoch": 2.694090382387022, "grad_norm": 5.038753032684326, "learning_rate": 3.135822196993621e-07, "loss": 0.3548, "step": 32550 }, { "epoch": 2.694918059923854, "grad_norm": 4.601801872253418, "learning_rate": 3.11905937416126e-07, "loss": 0.3999, "step": 32560 }, { "epoch": 2.6957457374606855, "grad_norm": 4.0485944747924805, "learning_rate": 3.102340032525569e-07, "loss": 0.3894, "step": 32570 }, { "epoch": 2.696573414997517, "grad_norm": 3.9255211353302, "learning_rate": 3.0856641875934454e-07, "loss": 0.4597, "step": 32580 }, { "epoch": 2.6974010925343483, "grad_norm": 4.893420696258545, "learning_rate": 3.069031854831411e-07, "loss": 0.4039, "step": 32590 }, { "epoch": 2.6982287700711804, "grad_norm": 5.569843292236328, "learning_rate": 3.0524430496656485e-07, "loss": 0.4137, "step": 32600 }, { "epoch": 2.6990564476080117, "grad_norm": 4.419327735900879, "learning_rate": 3.035897787481967e-07, "loss": 0.3319, "step": 32610 }, { "epoch": 2.6998841251448438, "grad_norm": 4.86116886138916, "learning_rate": 3.019396083625792e-07, "loss": 0.3814, "step": 32620 }, { "epoch": 2.700711802681675, "grad_norm": 4.786719799041748, "learning_rate": 3.0029379534021407e-07, "loss": 0.3483, "step": 32630 }, { "epoch": 2.701539480218507, "grad_norm": 4.323993682861328, "learning_rate": 2.9865234120756147e-07, "loss": 0.4832, "step": 32640 }, { "epoch": 2.7023671577553383, "grad_norm": 3.098665475845337, "learning_rate": 2.9701524748704025e-07, "loss": 0.4115, "step": 32650 }, { "epoch": 2.7031948352921704, "grad_norm": 4.100757598876953, "learning_rate": 2.953825156970236e-07, "loss": 0.4589, "step": 32660 }, { "epoch": 2.7040225128290016, "grad_norm": 5.408498764038086, "learning_rate": 2.937541473518407e-07, "loss": 0.4368, "step": 32670 }, { "epoch": 2.7048501903658337, "grad_norm": 3.13352108001709, "learning_rate": 2.92130143961773e-07, "loss": 0.4097, "step": 32680 }, { "epoch": 2.705677867902665, "grad_norm": 4.309831619262695, "learning_rate": 2.905105070330522e-07, "loss": 0.3762, "step": 32690 }, { "epoch": 2.7065055454394966, "grad_norm": 5.683094501495361, "learning_rate": 2.8889523806786114e-07, "loss": 0.4352, "step": 32700 }, { "epoch": 2.7073332229763283, "grad_norm": 3.276320695877075, "learning_rate": 2.8728433856433314e-07, "loss": 0.3807, "step": 32710 }, { "epoch": 2.70816090051316, "grad_norm": 5.010665416717529, "learning_rate": 2.856778100165447e-07, "loss": 0.376, "step": 32720 }, { "epoch": 2.7089885780499916, "grad_norm": 4.2375617027282715, "learning_rate": 2.8407565391452453e-07, "loss": 0.3803, "step": 32730 }, { "epoch": 2.7098162555868233, "grad_norm": 3.4677000045776367, "learning_rate": 2.8247787174424e-07, "loss": 0.4769, "step": 32740 }, { "epoch": 2.710643933123655, "grad_norm": 5.172788619995117, "learning_rate": 2.808844649876058e-07, "loss": 0.4456, "step": 32750 }, { "epoch": 2.7114716106604866, "grad_norm": 4.5200958251953125, "learning_rate": 2.792954351224753e-07, "loss": 0.3939, "step": 32760 }, { "epoch": 2.7122992881973182, "grad_norm": 4.168025016784668, "learning_rate": 2.77710783622645e-07, "loss": 0.38, "step": 32770 }, { "epoch": 2.71312696573415, "grad_norm": 4.088207244873047, "learning_rate": 2.761305119578489e-07, "loss": 0.4244, "step": 32780 }, { "epoch": 2.7139546432709816, "grad_norm": 3.011644124984741, "learning_rate": 2.745546215937606e-07, "loss": 0.3718, "step": 32790 }, { "epoch": 2.7147823208078132, "grad_norm": 5.985958099365234, "learning_rate": 2.729831139919886e-07, "loss": 0.3828, "step": 32800 }, { "epoch": 2.715609998344645, "grad_norm": 4.448300838470459, "learning_rate": 2.714159906100755e-07, "loss": 0.4742, "step": 32810 }, { "epoch": 2.7164376758814766, "grad_norm": 4.0071539878845215, "learning_rate": 2.698532529015002e-07, "loss": 0.4098, "step": 32820 }, { "epoch": 2.717265353418308, "grad_norm": 4.380973815917969, "learning_rate": 2.682949023156739e-07, "loss": 0.3441, "step": 32830 }, { "epoch": 2.71809303095514, "grad_norm": 3.3187096118927, "learning_rate": 2.667409402979343e-07, "loss": 0.3483, "step": 32840 }, { "epoch": 2.7189207084919715, "grad_norm": 3.641551971435547, "learning_rate": 2.651913682895557e-07, "loss": 0.4326, "step": 32850 }, { "epoch": 2.719748386028803, "grad_norm": 5.67224645614624, "learning_rate": 2.6364618772773465e-07, "loss": 0.4141, "step": 32860 }, { "epoch": 2.720576063565635, "grad_norm": 5.168071269989014, "learning_rate": 2.621054000455986e-07, "loss": 0.4208, "step": 32870 }, { "epoch": 2.7214037411024665, "grad_norm": 4.626334190368652, "learning_rate": 2.605690066721994e-07, "loss": 0.4392, "step": 32880 }, { "epoch": 2.722231418639298, "grad_norm": 4.135583877563477, "learning_rate": 2.5903700903251227e-07, "loss": 0.3278, "step": 32890 }, { "epoch": 2.72305909617613, "grad_norm": 5.525430202484131, "learning_rate": 2.5750940854743666e-07, "loss": 0.3986, "step": 32900 }, { "epoch": 2.7238867737129615, "grad_norm": 4.489912509918213, "learning_rate": 2.5598620663379426e-07, "loss": 0.3746, "step": 32910 }, { "epoch": 2.724714451249793, "grad_norm": 2.828843116760254, "learning_rate": 2.544674047043261e-07, "loss": 0.4314, "step": 32920 }, { "epoch": 2.725542128786625, "grad_norm": 5.1664719581604, "learning_rate": 2.5295300416769197e-07, "loss": 0.3364, "step": 32930 }, { "epoch": 2.7263698063234565, "grad_norm": 3.03627610206604, "learning_rate": 2.5144300642847006e-07, "loss": 0.4269, "step": 32940 }, { "epoch": 2.727197483860288, "grad_norm": 4.015223026275635, "learning_rate": 2.499374128871568e-07, "loss": 0.4292, "step": 32950 }, { "epoch": 2.72802516139712, "grad_norm": 4.636115074157715, "learning_rate": 2.4843622494015896e-07, "loss": 0.4467, "step": 32960 }, { "epoch": 2.7288528389339515, "grad_norm": 3.405292272567749, "learning_rate": 2.469394439798034e-07, "loss": 0.3709, "step": 32970 }, { "epoch": 2.729680516470783, "grad_norm": 5.986789226531982, "learning_rate": 2.454470713943247e-07, "loss": 0.3669, "step": 32980 }, { "epoch": 2.730508194007615, "grad_norm": 5.0434184074401855, "learning_rate": 2.4395910856787107e-07, "loss": 0.3879, "step": 32990 }, { "epoch": 2.731335871544446, "grad_norm": 3.813908815383911, "learning_rate": 2.42475556880501e-07, "loss": 0.3608, "step": 33000 }, { "epoch": 2.732163549081278, "grad_norm": 5.985851287841797, "learning_rate": 2.409964177081797e-07, "loss": 0.3625, "step": 33010 }, { "epoch": 2.7329912266181093, "grad_norm": 5.3613152503967285, "learning_rate": 2.395216924227811e-07, "loss": 0.3928, "step": 33020 }, { "epoch": 2.7338189041549414, "grad_norm": 4.321949005126953, "learning_rate": 2.3805138239208658e-07, "loss": 0.3911, "step": 33030 }, { "epoch": 2.7346465816917727, "grad_norm": 3.9433469772338867, "learning_rate": 2.3658548897977973e-07, "loss": 0.3967, "step": 33040 }, { "epoch": 2.7354742592286048, "grad_norm": 5.005143165588379, "learning_rate": 2.3512401354545134e-07, "loss": 0.3677, "step": 33050 }, { "epoch": 2.736301936765436, "grad_norm": 4.992791652679443, "learning_rate": 2.3366695744459044e-07, "loss": 0.3816, "step": 33060 }, { "epoch": 2.737129614302268, "grad_norm": 7.758734226226807, "learning_rate": 2.3221432202859106e-07, "loss": 0.3523, "step": 33070 }, { "epoch": 2.7379572918390993, "grad_norm": 5.580657482147217, "learning_rate": 2.3076610864474325e-07, "loss": 0.3902, "step": 33080 }, { "epoch": 2.7387849693759314, "grad_norm": 4.710116863250732, "learning_rate": 2.2932231863623978e-07, "loss": 0.4672, "step": 33090 }, { "epoch": 2.7396126469127626, "grad_norm": 4.4537482261657715, "learning_rate": 2.2788295334216949e-07, "loss": 0.4054, "step": 33100 }, { "epoch": 2.7404403244495943, "grad_norm": 7.477679252624512, "learning_rate": 2.264480140975145e-07, "loss": 0.4264, "step": 33110 }, { "epoch": 2.741268001986426, "grad_norm": 7.285399913787842, "learning_rate": 2.2501750223315578e-07, "loss": 0.3869, "step": 33120 }, { "epoch": 2.7420956795232576, "grad_norm": 5.346725940704346, "learning_rate": 2.235914190758659e-07, "loss": 0.4651, "step": 33130 }, { "epoch": 2.7429233570600893, "grad_norm": 4.916027545928955, "learning_rate": 2.2216976594830964e-07, "loss": 0.4208, "step": 33140 }, { "epoch": 2.743751034596921, "grad_norm": 4.534061431884766, "learning_rate": 2.2075254416904502e-07, "loss": 0.4227, "step": 33150 }, { "epoch": 2.7445787121337526, "grad_norm": 4.961166858673096, "learning_rate": 2.193397550525178e-07, "loss": 0.4721, "step": 33160 }, { "epoch": 2.7454063896705843, "grad_norm": 4.42424201965332, "learning_rate": 2.1793139990906486e-07, "loss": 0.3862, "step": 33170 }, { "epoch": 2.746234067207416, "grad_norm": 5.511905670166016, "learning_rate": 2.1652748004490742e-07, "loss": 0.4421, "step": 33180 }, { "epoch": 2.7470617447442476, "grad_norm": 3.1267266273498535, "learning_rate": 2.1512799676215557e-07, "loss": 0.4274, "step": 33190 }, { "epoch": 2.7478894222810792, "grad_norm": 4.486332893371582, "learning_rate": 2.1373295135880435e-07, "loss": 0.3765, "step": 33200 }, { "epoch": 2.748717099817911, "grad_norm": 3.9114043712615967, "learning_rate": 2.1234234512873264e-07, "loss": 0.3823, "step": 33210 }, { "epoch": 2.7495447773547426, "grad_norm": 5.05128288269043, "learning_rate": 2.1095617936170098e-07, "loss": 0.3191, "step": 33220 }, { "epoch": 2.7503724548915742, "grad_norm": 3.976297378540039, "learning_rate": 2.0957445534335264e-07, "loss": 0.3905, "step": 33230 }, { "epoch": 2.751200132428406, "grad_norm": 5.2983808517456055, "learning_rate": 2.081971743552108e-07, "loss": 0.3909, "step": 33240 }, { "epoch": 2.7520278099652375, "grad_norm": 4.170076847076416, "learning_rate": 2.0682433767467646e-07, "loss": 0.4302, "step": 33250 }, { "epoch": 2.752855487502069, "grad_norm": 5.027781009674072, "learning_rate": 2.0545594657503166e-07, "loss": 0.3151, "step": 33260 }, { "epoch": 2.753683165038901, "grad_norm": 6.220704555511475, "learning_rate": 2.0409200232543446e-07, "loss": 0.4327, "step": 33270 }, { "epoch": 2.7545108425757325, "grad_norm": 5.478437423706055, "learning_rate": 2.0273250619091522e-07, "loss": 0.4168, "step": 33280 }, { "epoch": 2.755338520112564, "grad_norm": 4.632569789886475, "learning_rate": 2.0137745943238307e-07, "loss": 0.3915, "step": 33290 }, { "epoch": 2.756166197649396, "grad_norm": 4.589548110961914, "learning_rate": 2.0002686330661714e-07, "loss": 0.3702, "step": 33300 }, { "epoch": 2.7569938751862275, "grad_norm": 3.282201051712036, "learning_rate": 1.9868071906627095e-07, "loss": 0.3827, "step": 33310 }, { "epoch": 2.757821552723059, "grad_norm": 7.25346040725708, "learning_rate": 1.9733902795986804e-07, "loss": 0.388, "step": 33320 }, { "epoch": 2.758649230259891, "grad_norm": 5.096437931060791, "learning_rate": 1.9600179123180186e-07, "loss": 0.4162, "step": 33330 }, { "epoch": 2.7594769077967225, "grad_norm": 4.088429927825928, "learning_rate": 1.946690101223353e-07, "loss": 0.4103, "step": 33340 }, { "epoch": 2.760304585333554, "grad_norm": 3.8281784057617188, "learning_rate": 1.9334068586759568e-07, "loss": 0.4168, "step": 33350 }, { "epoch": 2.761132262870386, "grad_norm": 4.416464328765869, "learning_rate": 1.9201681969958142e-07, "loss": 0.4216, "step": 33360 }, { "epoch": 2.7619599404072175, "grad_norm": 5.154476642608643, "learning_rate": 1.9069741284615194e-07, "loss": 0.3306, "step": 33370 }, { "epoch": 2.762787617944049, "grad_norm": 6.88673734664917, "learning_rate": 1.8938246653103287e-07, "loss": 0.377, "step": 33380 }, { "epoch": 2.763615295480881, "grad_norm": 4.665334224700928, "learning_rate": 1.880719819738136e-07, "loss": 0.4398, "step": 33390 }, { "epoch": 2.7644429730177125, "grad_norm": 6.106595993041992, "learning_rate": 1.8676596038994242e-07, "loss": 0.3881, "step": 33400 }, { "epoch": 2.7652706505545437, "grad_norm": 5.157016754150391, "learning_rate": 1.8546440299073042e-07, "loss": 0.2907, "step": 33410 }, { "epoch": 2.766098328091376, "grad_norm": 4.569086074829102, "learning_rate": 1.8416731098334916e-07, "loss": 0.4039, "step": 33420 }, { "epoch": 2.766926005628207, "grad_norm": 3.796894073486328, "learning_rate": 1.828746855708241e-07, "loss": 0.4173, "step": 33430 }, { "epoch": 2.767753683165039, "grad_norm": 5.623795986175537, "learning_rate": 1.8158652795204457e-07, "loss": 0.3996, "step": 33440 }, { "epoch": 2.7685813607018703, "grad_norm": 4.310593605041504, "learning_rate": 1.803028393217504e-07, "loss": 0.4031, "step": 33450 }, { "epoch": 2.7694090382387024, "grad_norm": 5.009363174438477, "learning_rate": 1.7902362087054092e-07, "loss": 0.3159, "step": 33460 }, { "epoch": 2.7702367157755337, "grad_norm": 4.147219657897949, "learning_rate": 1.7774887378486483e-07, "loss": 0.3574, "step": 33470 }, { "epoch": 2.7710643933123658, "grad_norm": 4.775441646575928, "learning_rate": 1.7647859924702747e-07, "loss": 0.3327, "step": 33480 }, { "epoch": 2.771892070849197, "grad_norm": 5.077761650085449, "learning_rate": 1.7521279843518536e-07, "loss": 0.302, "step": 33490 }, { "epoch": 2.7727197483860286, "grad_norm": 5.8022356033325195, "learning_rate": 1.7395147252334377e-07, "loss": 0.4202, "step": 33500 }, { "epoch": 2.7735474259228603, "grad_norm": 4.867712020874023, "learning_rate": 1.7269462268136082e-07, "loss": 0.4066, "step": 33510 }, { "epoch": 2.774375103459692, "grad_norm": 4.971965312957764, "learning_rate": 1.7144225007493908e-07, "loss": 0.3808, "step": 33520 }, { "epoch": 2.7752027809965236, "grad_norm": 5.564455986022949, "learning_rate": 1.7019435586563216e-07, "loss": 0.3751, "step": 33530 }, { "epoch": 2.7760304585333553, "grad_norm": 3.8257389068603516, "learning_rate": 1.689509412108381e-07, "loss": 0.5013, "step": 33540 }, { "epoch": 2.776858136070187, "grad_norm": 5.264125347137451, "learning_rate": 1.6771200726380054e-07, "loss": 0.4303, "step": 33550 }, { "epoch": 2.7776858136070186, "grad_norm": 5.745591640472412, "learning_rate": 1.6647755517360863e-07, "loss": 0.3781, "step": 33560 }, { "epoch": 2.7785134911438503, "grad_norm": 4.346833229064941, "learning_rate": 1.6524758608519265e-07, "loss": 0.359, "step": 33570 }, { "epoch": 2.779341168680682, "grad_norm": 4.707183361053467, "learning_rate": 1.640221011393256e-07, "loss": 0.4268, "step": 33580 }, { "epoch": 2.7801688462175136, "grad_norm": 5.710835933685303, "learning_rate": 1.6280110147262384e-07, "loss": 0.4181, "step": 33590 }, { "epoch": 2.7809965237543453, "grad_norm": 4.461061954498291, "learning_rate": 1.615845882175393e-07, "loss": 0.4474, "step": 33600 }, { "epoch": 2.781824201291177, "grad_norm": 4.314586162567139, "learning_rate": 1.6037256250236667e-07, "loss": 0.4036, "step": 33610 }, { "epoch": 2.7826518788280086, "grad_norm": 5.423824787139893, "learning_rate": 1.5916502545123724e-07, "loss": 0.3951, "step": 33620 }, { "epoch": 2.7834795563648402, "grad_norm": 5.419331073760986, "learning_rate": 1.5796197818411851e-07, "loss": 0.4628, "step": 33630 }, { "epoch": 2.784307233901672, "grad_norm": 5.113924026489258, "learning_rate": 1.5676342181681457e-07, "loss": 0.3638, "step": 33640 }, { "epoch": 2.7851349114385036, "grad_norm": 3.785917282104492, "learning_rate": 1.5556935746096402e-07, "loss": 0.3357, "step": 33650 }, { "epoch": 2.785962588975335, "grad_norm": 6.132958889007568, "learning_rate": 1.5437978622403926e-07, "loss": 0.4436, "step": 33660 }, { "epoch": 2.786790266512167, "grad_norm": 4.519553184509277, "learning_rate": 1.5319470920934498e-07, "loss": 0.309, "step": 33670 }, { "epoch": 2.7876179440489985, "grad_norm": 4.002167224884033, "learning_rate": 1.5201412751601918e-07, "loss": 0.4094, "step": 33680 }, { "epoch": 2.78844562158583, "grad_norm": 4.231441974639893, "learning_rate": 1.508380422390282e-07, "loss": 0.4127, "step": 33690 }, { "epoch": 2.789273299122662, "grad_norm": 4.558610916137695, "learning_rate": 1.496664544691695e-07, "loss": 0.408, "step": 33700 }, { "epoch": 2.7901009766594935, "grad_norm": 5.5869574546813965, "learning_rate": 1.4849936529306942e-07, "loss": 0.4348, "step": 33710 }, { "epoch": 2.790928654196325, "grad_norm": 5.159029006958008, "learning_rate": 1.4733677579317983e-07, "loss": 0.369, "step": 33720 }, { "epoch": 2.791756331733157, "grad_norm": 3.849769115447998, "learning_rate": 1.4617868704778217e-07, "loss": 0.4226, "step": 33730 }, { "epoch": 2.7925840092699885, "grad_norm": 5.091282367706299, "learning_rate": 1.4502510013098158e-07, "loss": 0.4008, "step": 33740 }, { "epoch": 2.79341168680682, "grad_norm": 3.4144201278686523, "learning_rate": 1.4387601611270897e-07, "loss": 0.3756, "step": 33750 }, { "epoch": 2.794239364343652, "grad_norm": 4.466291427612305, "learning_rate": 1.4273143605871786e-07, "loss": 0.4268, "step": 33760 }, { "epoch": 2.7950670418804835, "grad_norm": 5.3216047286987305, "learning_rate": 1.4159136103058414e-07, "loss": 0.4473, "step": 33770 }, { "epoch": 2.795894719417315, "grad_norm": 3.700798988342285, "learning_rate": 1.4045579208570802e-07, "loss": 0.4114, "step": 33780 }, { "epoch": 2.796722396954147, "grad_norm": 5.026780605316162, "learning_rate": 1.3932473027730542e-07, "loss": 0.4124, "step": 33790 }, { "epoch": 2.7975500744909785, "grad_norm": 4.910149097442627, "learning_rate": 1.3819817665441883e-07, "loss": 0.4848, "step": 33800 }, { "epoch": 2.79837775202781, "grad_norm": 5.24513578414917, "learning_rate": 1.370761322619024e-07, "loss": 0.4016, "step": 33810 }, { "epoch": 2.7992054295646414, "grad_norm": 4.3488264083862305, "learning_rate": 1.3595859814043322e-07, "loss": 0.3318, "step": 33820 }, { "epoch": 2.8000331071014735, "grad_norm": 6.739727020263672, "learning_rate": 1.3484557532650323e-07, "loss": 0.3958, "step": 33830 }, { "epoch": 2.8008607846383047, "grad_norm": 4.352637767791748, "learning_rate": 1.3373706485241888e-07, "loss": 0.3963, "step": 33840 }, { "epoch": 2.801688462175137, "grad_norm": 2.9602715969085693, "learning_rate": 1.3263306774630448e-07, "loss": 0.402, "step": 33850 }, { "epoch": 2.802516139711968, "grad_norm": 4.556917667388916, "learning_rate": 1.315335850320959e-07, "loss": 0.4342, "step": 33860 }, { "epoch": 2.8033438172488, "grad_norm": 5.121495723724365, "learning_rate": 1.3043861772954303e-07, "loss": 0.3893, "step": 33870 }, { "epoch": 2.8041714947856313, "grad_norm": 4.840038776397705, "learning_rate": 1.2934816685420792e-07, "loss": 0.4446, "step": 33880 }, { "epoch": 2.8049991723224634, "grad_norm": 3.7057182788848877, "learning_rate": 1.2826223341746324e-07, "loss": 0.3004, "step": 33890 }, { "epoch": 2.8058268498592946, "grad_norm": 5.31887149810791, "learning_rate": 1.271808184264911e-07, "loss": 0.391, "step": 33900 }, { "epoch": 2.8066545273961263, "grad_norm": 4.97280740737915, "learning_rate": 1.2610392288428475e-07, "loss": 0.434, "step": 33910 }, { "epoch": 2.807482204932958, "grad_norm": 4.508645534515381, "learning_rate": 1.2503154778964467e-07, "loss": 0.4239, "step": 33920 }, { "epoch": 2.8083098824697896, "grad_norm": 2.787139892578125, "learning_rate": 1.2396369413717856e-07, "loss": 0.3989, "step": 33930 }, { "epoch": 2.8091375600066213, "grad_norm": 4.502056121826172, "learning_rate": 1.2290036291730035e-07, "loss": 0.415, "step": 33940 }, { "epoch": 2.809965237543453, "grad_norm": 4.3798747062683105, "learning_rate": 1.2184155511622998e-07, "loss": 0.3614, "step": 33950 }, { "epoch": 2.8107929150802846, "grad_norm": 5.5043110847473145, "learning_rate": 1.207872717159919e-07, "loss": 0.4099, "step": 33960 }, { "epoch": 2.8116205926171163, "grad_norm": 2.6445658206939697, "learning_rate": 1.1973751369441399e-07, "loss": 0.4023, "step": 33970 }, { "epoch": 2.812448270153948, "grad_norm": 4.339644432067871, "learning_rate": 1.1869228202512905e-07, "loss": 0.3474, "step": 33980 }, { "epoch": 2.8132759476907796, "grad_norm": 4.0086541175842285, "learning_rate": 1.1765157767756774e-07, "loss": 0.3696, "step": 33990 }, { "epoch": 2.8141036252276113, "grad_norm": 3.959331750869751, "learning_rate": 1.1661540161696461e-07, "loss": 0.3761, "step": 34000 }, { "epoch": 2.814931302764443, "grad_norm": 6.2976298332214355, "learning_rate": 1.1558375480435369e-07, "loss": 0.3994, "step": 34010 }, { "epoch": 2.8157589803012746, "grad_norm": 5.0176682472229, "learning_rate": 1.1455663819656793e-07, "loss": 0.3483, "step": 34020 }, { "epoch": 2.8165866578381062, "grad_norm": 7.6254777908325195, "learning_rate": 1.1353405274623918e-07, "loss": 0.3975, "step": 34030 }, { "epoch": 2.817414335374938, "grad_norm": 3.5800650119781494, "learning_rate": 1.1251599940179603e-07, "loss": 0.4037, "step": 34040 }, { "epoch": 2.8182420129117696, "grad_norm": 4.386075496673584, "learning_rate": 1.1150247910746481e-07, "loss": 0.3557, "step": 34050 }, { "epoch": 2.8190696904486012, "grad_norm": 5.1455159187316895, "learning_rate": 1.1049349280326471e-07, "loss": 0.4445, "step": 34060 }, { "epoch": 2.819897367985433, "grad_norm": 4.749353885650635, "learning_rate": 1.0948904142501326e-07, "loss": 0.3294, "step": 34070 }, { "epoch": 2.8207250455222646, "grad_norm": 6.391259670257568, "learning_rate": 1.0848912590432026e-07, "loss": 0.4112, "step": 34080 }, { "epoch": 2.821552723059096, "grad_norm": 3.7254531383514404, "learning_rate": 1.0749374716858774e-07, "loss": 0.4281, "step": 34090 }, { "epoch": 2.822380400595928, "grad_norm": 4.0615553855896, "learning_rate": 1.0650290614101278e-07, "loss": 0.3915, "step": 34100 }, { "epoch": 2.8232080781327595, "grad_norm": 3.8052685260772705, "learning_rate": 1.055166037405797e-07, "loss": 0.3836, "step": 34110 }, { "epoch": 2.824035755669591, "grad_norm": 3.9211692810058594, "learning_rate": 1.0453484088206678e-07, "loss": 0.3676, "step": 34120 }, { "epoch": 2.824863433206423, "grad_norm": 2.4579498767852783, "learning_rate": 1.0355761847604007e-07, "loss": 0.3928, "step": 34130 }, { "epoch": 2.8256911107432545, "grad_norm": 5.316070079803467, "learning_rate": 1.0258493742885511e-07, "loss": 0.4336, "step": 34140 }, { "epoch": 2.826518788280086, "grad_norm": 4.199663162231445, "learning_rate": 1.0161679864265694e-07, "loss": 0.4389, "step": 34150 }, { "epoch": 2.827346465816918, "grad_norm": 3.501479148864746, "learning_rate": 1.0065320301537451e-07, "loss": 0.4141, "step": 34160 }, { "epoch": 2.8281741433537495, "grad_norm": 5.269373893737793, "learning_rate": 9.969415144072569e-08, "loss": 0.4413, "step": 34170 }, { "epoch": 2.829001820890581, "grad_norm": 5.124218940734863, "learning_rate": 9.873964480821175e-08, "loss": 0.3349, "step": 34180 }, { "epoch": 2.829829498427413, "grad_norm": 6.449194431304932, "learning_rate": 9.778968400312117e-08, "loss": 0.4129, "step": 34190 }, { "epoch": 2.8306571759642445, "grad_norm": 3.660088062286377, "learning_rate": 9.684426990652363e-08, "loss": 0.3982, "step": 34200 }, { "epoch": 2.8314848535010757, "grad_norm": 6.106384754180908, "learning_rate": 9.590340339527382e-08, "loss": 0.3831, "step": 34210 }, { "epoch": 2.832312531037908, "grad_norm": 4.34075403213501, "learning_rate": 9.496708534200816e-08, "loss": 0.4048, "step": 34220 }, { "epoch": 2.833140208574739, "grad_norm": 3.3307230472564697, "learning_rate": 9.40353166151431e-08, "loss": 0.3572, "step": 34230 }, { "epoch": 2.833967886111571, "grad_norm": 5.380303859710693, "learning_rate": 9.310809807887678e-08, "loss": 0.3839, "step": 34240 }, { "epoch": 2.8347955636484024, "grad_norm": 5.087488174438477, "learning_rate": 9.218543059318851e-08, "loss": 0.4348, "step": 34250 }, { "epoch": 2.8356232411852345, "grad_norm": 5.621816158294678, "learning_rate": 9.126731501383324e-08, "loss": 0.4179, "step": 34260 }, { "epoch": 2.8364509187220657, "grad_norm": 5.186544895172119, "learning_rate": 9.03537521923481e-08, "loss": 0.3908, "step": 34270 }, { "epoch": 2.837278596258898, "grad_norm": 3.7417140007019043, "learning_rate": 8.944474297604366e-08, "loss": 0.3724, "step": 34280 }, { "epoch": 2.838106273795729, "grad_norm": 2.197061538696289, "learning_rate": 8.854028820800997e-08, "loss": 0.4122, "step": 34290 }, { "epoch": 2.838933951332561, "grad_norm": 3.5063562393188477, "learning_rate": 8.764038872711156e-08, "loss": 0.393, "step": 34300 }, { "epoch": 2.8397616288693923, "grad_norm": 5.814626216888428, "learning_rate": 8.674504536798856e-08, "loss": 0.4128, "step": 34310 }, { "epoch": 2.840589306406224, "grad_norm": 5.684357643127441, "learning_rate": 8.585425896105559e-08, "loss": 0.4304, "step": 34320 }, { "epoch": 2.8414169839430556, "grad_norm": 5.19627046585083, "learning_rate": 8.496803033249957e-08, "loss": 0.4255, "step": 34330 }, { "epoch": 2.8422446614798873, "grad_norm": 4.0316481590271, "learning_rate": 8.40863603042813e-08, "loss": 0.4064, "step": 34340 }, { "epoch": 2.843072339016719, "grad_norm": 3.5759475231170654, "learning_rate": 8.320924969413391e-08, "loss": 0.3454, "step": 34350 }, { "epoch": 2.8439000165535506, "grad_norm": 4.293327808380127, "learning_rate": 8.233669931555999e-08, "loss": 0.404, "step": 34360 }, { "epoch": 2.8447276940903823, "grad_norm": 4.977463245391846, "learning_rate": 8.146870997783495e-08, "loss": 0.4551, "step": 34370 }, { "epoch": 2.845555371627214, "grad_norm": 8.3674898147583, "learning_rate": 8.060528248600097e-08, "loss": 0.4184, "step": 34380 }, { "epoch": 2.8463830491640456, "grad_norm": 4.663473606109619, "learning_rate": 7.974641764087298e-08, "loss": 0.3842, "step": 34390 }, { "epoch": 2.8472107267008773, "grad_norm": 7.245804309844971, "learning_rate": 7.889211623903103e-08, "loss": 0.3867, "step": 34400 }, { "epoch": 2.848038404237709, "grad_norm": 4.302618503570557, "learning_rate": 7.804237907282353e-08, "loss": 0.4529, "step": 34410 }, { "epoch": 2.8488660817745406, "grad_norm": 4.214704990386963, "learning_rate": 7.719720693036615e-08, "loss": 0.4034, "step": 34420 }, { "epoch": 2.8496937593113723, "grad_norm": 3.774834394454956, "learning_rate": 7.635660059554017e-08, "loss": 0.4041, "step": 34430 }, { "epoch": 2.850521436848204, "grad_norm": 4.034228324890137, "learning_rate": 7.552056084799197e-08, "loss": 0.3651, "step": 34440 }, { "epoch": 2.8513491143850356, "grad_norm": 5.094396591186523, "learning_rate": 7.468908846313349e-08, "loss": 0.3386, "step": 34450 }, { "epoch": 2.8521767919218672, "grad_norm": 4.422966957092285, "learning_rate": 7.386218421213897e-08, "loss": 0.3956, "step": 34460 }, { "epoch": 2.853004469458699, "grad_norm": 5.348505020141602, "learning_rate": 7.30398488619477e-08, "loss": 0.4343, "step": 34470 }, { "epoch": 2.8538321469955306, "grad_norm": 4.585025787353516, "learning_rate": 7.222208317525958e-08, "loss": 0.4468, "step": 34480 }, { "epoch": 2.8546598245323622, "grad_norm": 4.774411201477051, "learning_rate": 7.140888791053791e-08, "loss": 0.4376, "step": 34490 }, { "epoch": 2.855487502069194, "grad_norm": 2.6165592670440674, "learning_rate": 7.060026382200435e-08, "loss": 0.4199, "step": 34500 }, { "epoch": 2.8563151796060255, "grad_norm": 4.52363920211792, "learning_rate": 6.9796211659644e-08, "loss": 0.4024, "step": 34510 }, { "epoch": 2.857142857142857, "grad_norm": 4.796764850616455, "learning_rate": 6.899673216920034e-08, "loss": 0.4215, "step": 34520 }, { "epoch": 2.857970534679689, "grad_norm": 6.429803848266602, "learning_rate": 6.820182609217462e-08, "loss": 0.3687, "step": 34530 }, { "epoch": 2.8587982122165205, "grad_norm": 5.891969680786133, "learning_rate": 6.741149416582881e-08, "loss": 0.3778, "step": 34540 }, { "epoch": 2.859625889753352, "grad_norm": 2.8533151149749756, "learning_rate": 6.662573712317932e-08, "loss": 0.3747, "step": 34550 }, { "epoch": 2.860453567290184, "grad_norm": 4.142826557159424, "learning_rate": 6.584455569300263e-08, "loss": 0.4188, "step": 34560 }, { "epoch": 2.8612812448270155, "grad_norm": 4.108867168426514, "learning_rate": 6.506795059982918e-08, "loss": 0.368, "step": 34570 }, { "epoch": 2.862108922363847, "grad_norm": 4.669468402862549, "learning_rate": 6.429592256394557e-08, "loss": 0.4032, "step": 34580 }, { "epoch": 2.862936599900679, "grad_norm": 5.335855007171631, "learning_rate": 6.352847230139403e-08, "loss": 0.3779, "step": 34590 }, { "epoch": 2.8637642774375105, "grad_norm": 3.9285643100738525, "learning_rate": 6.276560052396963e-08, "loss": 0.3656, "step": 34600 }, { "epoch": 2.864591954974342, "grad_norm": 4.058329105377197, "learning_rate": 6.200730793922194e-08, "loss": 0.4569, "step": 34610 }, { "epoch": 2.8654196325111734, "grad_norm": 4.341635704040527, "learning_rate": 6.125359525045338e-08, "loss": 0.3947, "step": 34620 }, { "epoch": 2.8662473100480055, "grad_norm": 3.485604763031006, "learning_rate": 6.05044631567181e-08, "loss": 0.4169, "step": 34630 }, { "epoch": 2.8670749875848367, "grad_norm": 3.61708664894104, "learning_rate": 5.975991235282252e-08, "loss": 0.3618, "step": 34640 }, { "epoch": 2.867902665121669, "grad_norm": 3.6740596294403076, "learning_rate": 5.9019943529323164e-08, "loss": 0.4393, "step": 34650 }, { "epoch": 2.8687303426585, "grad_norm": 4.192541599273682, "learning_rate": 5.8284557372527694e-08, "loss": 0.3152, "step": 34660 }, { "epoch": 2.869558020195332, "grad_norm": 10.934873580932617, "learning_rate": 5.755375456449275e-08, "loss": 0.4344, "step": 34670 }, { "epoch": 2.8703856977321633, "grad_norm": 4.508777618408203, "learning_rate": 5.6827535783024466e-08, "loss": 0.3813, "step": 34680 }, { "epoch": 2.8712133752689954, "grad_norm": 5.16764497756958, "learning_rate": 5.6105901701676844e-08, "loss": 0.4124, "step": 34690 }, { "epoch": 2.8720410528058267, "grad_norm": 6.030416011810303, "learning_rate": 5.538885298975172e-08, "loss": 0.4176, "step": 34700 }, { "epoch": 2.8728687303426588, "grad_norm": 5.433518409729004, "learning_rate": 5.4676390312299874e-08, "loss": 0.3825, "step": 34710 }, { "epoch": 2.87369640787949, "grad_norm": 4.612037658691406, "learning_rate": 5.3968514330115516e-08, "loss": 0.411, "step": 34720 }, { "epoch": 2.8745240854163217, "grad_norm": 4.086051940917969, "learning_rate": 5.326522569974124e-08, "loss": 0.3871, "step": 34730 }, { "epoch": 2.8753517629531533, "grad_norm": 4.321046352386475, "learning_rate": 5.256652507346416e-08, "loss": 0.4796, "step": 34740 }, { "epoch": 2.876179440489985, "grad_norm": 5.720175743103027, "learning_rate": 5.1872413099315346e-08, "loss": 0.3573, "step": 34750 }, { "epoch": 2.8770071180268166, "grad_norm": 2.6809885501861572, "learning_rate": 5.1182890421071494e-08, "loss": 0.3702, "step": 34760 }, { "epoch": 2.8778347955636483, "grad_norm": 3.8667960166931152, "learning_rate": 5.049795767825161e-08, "loss": 0.4001, "step": 34770 }, { "epoch": 2.87866247310048, "grad_norm": 4.202096939086914, "learning_rate": 4.981761550611808e-08, "loss": 0.3524, "step": 34780 }, { "epoch": 2.8794901506373116, "grad_norm": 4.400666236877441, "learning_rate": 4.91418645356756e-08, "loss": 0.3607, "step": 34790 }, { "epoch": 2.8803178281741433, "grad_norm": 5.9544677734375, "learning_rate": 4.847070539367005e-08, "loss": 0.3953, "step": 34800 }, { "epoch": 2.881145505710975, "grad_norm": 2.7166481018066406, "learning_rate": 4.780413870259015e-08, "loss": 0.3688, "step": 34810 }, { "epoch": 2.8819731832478066, "grad_norm": 6.439957141876221, "learning_rate": 4.714216508066305e-08, "loss": 0.3671, "step": 34820 }, { "epoch": 2.8828008607846383, "grad_norm": 4.229747772216797, "learning_rate": 4.6484785141857614e-08, "loss": 0.4158, "step": 34830 }, { "epoch": 2.88362853832147, "grad_norm": 5.447269439697266, "learning_rate": 4.5831999495880576e-08, "loss": 0.4198, "step": 34840 }, { "epoch": 2.8844562158583016, "grad_norm": 3.612624168395996, "learning_rate": 4.518380874817929e-08, "loss": 0.4166, "step": 34850 }, { "epoch": 2.8852838933951332, "grad_norm": 5.952037334442139, "learning_rate": 4.454021349993787e-08, "loss": 0.395, "step": 34860 }, { "epoch": 2.886111570931965, "grad_norm": 4.05803918838501, "learning_rate": 4.390121434807937e-08, "loss": 0.3297, "step": 34870 }, { "epoch": 2.8869392484687966, "grad_norm": 5.370245456695557, "learning_rate": 4.326681188526416e-08, "loss": 0.3959, "step": 34880 }, { "epoch": 2.8877669260056282, "grad_norm": 4.8259806632995605, "learning_rate": 4.2637006699887127e-08, "loss": 0.4239, "step": 34890 }, { "epoch": 2.88859460354246, "grad_norm": 5.106606483459473, "learning_rate": 4.2011799376081574e-08, "loss": 0.3838, "step": 34900 }, { "epoch": 2.8894222810792916, "grad_norm": 5.049558639526367, "learning_rate": 4.139119049371587e-08, "loss": 0.4201, "step": 34910 }, { "epoch": 2.890249958616123, "grad_norm": 5.685618877410889, "learning_rate": 4.0775180628392916e-08, "loss": 0.4081, "step": 34920 }, { "epoch": 2.891077636152955, "grad_norm": 5.3696722984313965, "learning_rate": 4.016377035145069e-08, "loss": 0.3939, "step": 34930 }, { "epoch": 2.8919053136897865, "grad_norm": 4.061400890350342, "learning_rate": 3.955696022996003e-08, "loss": 0.4685, "step": 34940 }, { "epoch": 2.892732991226618, "grad_norm": 3.8671109676361084, "learning_rate": 3.895475082672684e-08, "loss": 0.4138, "step": 34950 }, { "epoch": 2.89356066876345, "grad_norm": 4.248999118804932, "learning_rate": 3.835714270028879e-08, "loss": 0.428, "step": 34960 }, { "epoch": 2.8943883463002815, "grad_norm": 5.584463119506836, "learning_rate": 3.776413640491583e-08, "loss": 0.4275, "step": 34970 }, { "epoch": 2.895216023837113, "grad_norm": 6.575563430786133, "learning_rate": 3.7175732490610796e-08, "loss": 0.3891, "step": 34980 }, { "epoch": 2.896043701373945, "grad_norm": 3.4103360176086426, "learning_rate": 3.659193150310714e-08, "loss": 0.3485, "step": 34990 }, { "epoch": 2.8968713789107765, "grad_norm": 4.881225109100342, "learning_rate": 3.601273398386895e-08, "loss": 0.4316, "step": 35000 }, { "epoch": 2.897699056447608, "grad_norm": 4.609978675842285, "learning_rate": 3.5438140470092083e-08, "loss": 0.3796, "step": 35010 }, { "epoch": 2.89852673398444, "grad_norm": 3.40297794342041, "learning_rate": 3.486815149470024e-08, "loss": 0.3979, "step": 35020 }, { "epoch": 2.899354411521271, "grad_norm": 3.6820528507232666, "learning_rate": 3.4302767586348315e-08, "loss": 0.4141, "step": 35030 }, { "epoch": 2.900182089058103, "grad_norm": 5.406149864196777, "learning_rate": 3.374198926941907e-08, "loss": 0.3968, "step": 35040 }, { "epoch": 2.9010097665949344, "grad_norm": 6.444420337677002, "learning_rate": 3.318581706402368e-08, "loss": 0.4149, "step": 35050 }, { "epoch": 2.9018374441317665, "grad_norm": 3.1498911380767822, "learning_rate": 3.263425148600175e-08, "loss": 0.378, "step": 35060 }, { "epoch": 2.9026651216685977, "grad_norm": 3.970010757446289, "learning_rate": 3.208729304692015e-08, "loss": 0.4021, "step": 35070 }, { "epoch": 2.90349279920543, "grad_norm": 4.329597473144531, "learning_rate": 3.154494225407201e-08, "loss": 0.3905, "step": 35080 }, { "epoch": 2.904320476742261, "grad_norm": 3.524667739868164, "learning_rate": 3.1007199610478266e-08, "loss": 0.3702, "step": 35090 }, { "epoch": 2.905148154279093, "grad_norm": 3.522092342376709, "learning_rate": 3.0474065614884976e-08, "loss": 0.3214, "step": 35100 }, { "epoch": 2.9059758318159243, "grad_norm": 5.15186071395874, "learning_rate": 2.994554076176326e-08, "loss": 0.4011, "step": 35110 }, { "epoch": 2.9068035093527564, "grad_norm": 3.3724968433380127, "learning_rate": 2.9421625541311006e-08, "loss": 0.3829, "step": 35120 }, { "epoch": 2.9076311868895877, "grad_norm": 3.275097608566284, "learning_rate": 2.8902320439450072e-08, "loss": 0.3503, "step": 35130 }, { "epoch": 2.9084588644264193, "grad_norm": 5.420446872711182, "learning_rate": 2.8387625937824624e-08, "loss": 0.3296, "step": 35140 }, { "epoch": 2.909286541963251, "grad_norm": 3.3749096393585205, "learning_rate": 2.787754251380559e-08, "loss": 0.4145, "step": 35150 }, { "epoch": 2.9101142195000826, "grad_norm": 4.404683589935303, "learning_rate": 2.7372070640485638e-08, "loss": 0.382, "step": 35160 }, { "epoch": 2.9109418970369143, "grad_norm": 3.988985776901245, "learning_rate": 2.687121078668031e-08, "loss": 0.3983, "step": 35170 }, { "epoch": 2.911769574573746, "grad_norm": 4.4609527587890625, "learning_rate": 2.6374963416927468e-08, "loss": 0.4033, "step": 35180 }, { "epoch": 2.9125972521105776, "grad_norm": 5.654046535491943, "learning_rate": 2.5883328991487266e-08, "loss": 0.4006, "step": 35190 }, { "epoch": 2.9134249296474093, "grad_norm": 3.299189567565918, "learning_rate": 2.5396307966342736e-08, "loss": 0.3811, "step": 35200 }, { "epoch": 2.914252607184241, "grad_norm": 4.087685585021973, "learning_rate": 2.4913900793194225e-08, "loss": 0.4341, "step": 35210 }, { "epoch": 2.9150802847210726, "grad_norm": 4.526525497436523, "learning_rate": 2.4436107919468266e-08, "loss": 0.3988, "step": 35220 }, { "epoch": 2.9159079622579043, "grad_norm": 4.215282917022705, "learning_rate": 2.3962929788307056e-08, "loss": 0.4554, "step": 35230 }, { "epoch": 2.916735639794736, "grad_norm": 4.59426212310791, "learning_rate": 2.3494366838574535e-08, "loss": 0.4032, "step": 35240 }, { "epoch": 2.9175633173315676, "grad_norm": 5.619063377380371, "learning_rate": 2.303041950485474e-08, "loss": 0.4045, "step": 35250 }, { "epoch": 2.9183909948683993, "grad_norm": 4.793766975402832, "learning_rate": 2.2571088217449576e-08, "loss": 0.4536, "step": 35260 }, { "epoch": 2.919218672405231, "grad_norm": 4.363614082336426, "learning_rate": 2.2116373402380488e-08, "loss": 0.4177, "step": 35270 }, { "epoch": 2.9200463499420626, "grad_norm": 5.158024787902832, "learning_rate": 2.1666275481387335e-08, "loss": 0.4075, "step": 35280 }, { "epoch": 2.9208740274788942, "grad_norm": 4.945051193237305, "learning_rate": 2.1220794871927298e-08, "loss": 0.4098, "step": 35290 }, { "epoch": 2.921701705015726, "grad_norm": 5.293569564819336, "learning_rate": 2.0779931987174872e-08, "loss": 0.361, "step": 35300 }, { "epoch": 2.9225293825525576, "grad_norm": 5.283277988433838, "learning_rate": 2.0343687236022425e-08, "loss": 0.3685, "step": 35310 }, { "epoch": 2.9233570600893892, "grad_norm": 5.713778018951416, "learning_rate": 1.9912061023079076e-08, "loss": 0.3963, "step": 35320 }, { "epoch": 2.924184737626221, "grad_norm": 5.495223045349121, "learning_rate": 1.94850537486696e-08, "loss": 0.3676, "step": 35330 }, { "epoch": 2.9250124151630525, "grad_norm": 4.596635818481445, "learning_rate": 1.9062665808836088e-08, "loss": 0.44, "step": 35340 }, { "epoch": 2.925840092699884, "grad_norm": 4.7999587059021, "learning_rate": 1.8644897595334612e-08, "loss": 0.4529, "step": 35350 }, { "epoch": 2.926667770236716, "grad_norm": 4.563276767730713, "learning_rate": 1.823174949563744e-08, "loss": 0.4085, "step": 35360 }, { "epoch": 2.9274954477735475, "grad_norm": 4.509946346282959, "learning_rate": 1.7823221892932506e-08, "loss": 0.4244, "step": 35370 }, { "epoch": 2.928323125310379, "grad_norm": 4.840301513671875, "learning_rate": 1.741931516612061e-08, "loss": 0.3957, "step": 35380 }, { "epoch": 2.929150802847211, "grad_norm": 4.135120868682861, "learning_rate": 1.7020029689818195e-08, "loss": 0.3616, "step": 35390 }, { "epoch": 2.9299784803840425, "grad_norm": 4.282018661499023, "learning_rate": 1.662536583435459e-08, "loss": 0.3474, "step": 35400 }, { "epoch": 2.930806157920874, "grad_norm": 4.0736985206604, "learning_rate": 1.6235323965774208e-08, "loss": 0.421, "step": 35410 }, { "epoch": 2.931633835457706, "grad_norm": 7.598786354064941, "learning_rate": 1.5849904445832677e-08, "loss": 0.3818, "step": 35420 }, { "epoch": 2.9324615129945375, "grad_norm": 4.413760662078857, "learning_rate": 1.5469107631999048e-08, "loss": 0.4391, "step": 35430 }, { "epoch": 2.9332891905313687, "grad_norm": 6.215459823608398, "learning_rate": 1.5092933877456362e-08, "loss": 0.3909, "step": 35440 }, { "epoch": 2.934116868068201, "grad_norm": 4.132980823516846, "learning_rate": 1.472138353109831e-08, "loss": 0.3926, "step": 35450 }, { "epoch": 2.934944545605032, "grad_norm": 3.8542494773864746, "learning_rate": 1.4354456937530348e-08, "loss": 0.4224, "step": 35460 }, { "epoch": 2.935772223141864, "grad_norm": 5.272983074188232, "learning_rate": 1.399215443707136e-08, "loss": 0.31, "step": 35470 }, { "epoch": 2.9365999006786954, "grad_norm": 4.106378555297852, "learning_rate": 1.363447636574866e-08, "loss": 0.4379, "step": 35480 }, { "epoch": 2.9374275782155275, "grad_norm": 3.3772284984588623, "learning_rate": 1.3281423055302444e-08, "loss": 0.3684, "step": 35490 }, { "epoch": 2.9382552557523587, "grad_norm": 6.086026191711426, "learning_rate": 1.2932994833183554e-08, "loss": 0.417, "step": 35500 }, { "epoch": 2.939082933289191, "grad_norm": 4.900084018707275, "learning_rate": 1.2589192022552377e-08, "loss": 0.4173, "step": 35510 }, { "epoch": 2.939910610826022, "grad_norm": 3.853996753692627, "learning_rate": 1.2250014942279398e-08, "loss": 0.3815, "step": 35520 }, { "epoch": 2.9407382883628537, "grad_norm": 3.264935255050659, "learning_rate": 1.1915463906945202e-08, "loss": 0.3877, "step": 35530 }, { "epoch": 2.9415659658996853, "grad_norm": 3.901629686355591, "learning_rate": 1.158553922683936e-08, "loss": 0.3602, "step": 35540 }, { "epoch": 2.942393643436517, "grad_norm": 4.596108913421631, "learning_rate": 1.1260241207961543e-08, "loss": 0.4298, "step": 35550 }, { "epoch": 2.9432213209733487, "grad_norm": 4.626630783081055, "learning_rate": 1.0939570152018741e-08, "loss": 0.4238, "step": 35560 }, { "epoch": 2.9440489985101803, "grad_norm": 4.507631778717041, "learning_rate": 1.0623526356428049e-08, "loss": 0.3944, "step": 35570 }, { "epoch": 2.944876676047012, "grad_norm": 4.0000433921813965, "learning_rate": 1.0312110114313878e-08, "loss": 0.3976, "step": 35580 }, { "epoch": 2.9457043535838436, "grad_norm": 4.450434684753418, "learning_rate": 1.0005321714509075e-08, "loss": 0.4431, "step": 35590 }, { "epoch": 2.9465320311206753, "grad_norm": 4.1186652183532715, "learning_rate": 9.70316144155381e-09, "loss": 0.4211, "step": 35600 }, { "epoch": 2.947359708657507, "grad_norm": 4.401911735534668, "learning_rate": 9.40562957569613e-09, "loss": 0.4089, "step": 35610 }, { "epoch": 2.9481873861943386, "grad_norm": 3.7215123176574707, "learning_rate": 9.112726392891957e-09, "loss": 0.4002, "step": 35620 }, { "epoch": 2.9490150637311703, "grad_norm": 6.921922206878662, "learning_rate": 8.824452164802876e-09, "loss": 0.4127, "step": 35630 }, { "epoch": 2.949842741268002, "grad_norm": 4.071949481964111, "learning_rate": 8.540807158798347e-09, "loss": 0.3657, "step": 35640 }, { "epoch": 2.9506704188048336, "grad_norm": 4.698385238647461, "learning_rate": 8.261791637952932e-09, "loss": 0.4269, "step": 35650 }, { "epoch": 2.9514980963416653, "grad_norm": 4.9215545654296875, "learning_rate": 7.987405861049068e-09, "loss": 0.3828, "step": 35660 }, { "epoch": 2.952325773878497, "grad_norm": 3.805814266204834, "learning_rate": 7.717650082574302e-09, "loss": 0.383, "step": 35670 }, { "epoch": 2.9531534514153286, "grad_norm": 4.326443672180176, "learning_rate": 7.452524552722384e-09, "loss": 0.4191, "step": 35680 }, { "epoch": 2.9539811289521603, "grad_norm": 6.665375709533691, "learning_rate": 7.192029517391619e-09, "loss": 0.4008, "step": 35690 }, { "epoch": 2.954808806488992, "grad_norm": 5.553435802459717, "learning_rate": 6.936165218186519e-09, "loss": 0.421, "step": 35700 }, { "epoch": 2.9556364840258236, "grad_norm": 5.4914631843566895, "learning_rate": 6.6849318924167016e-09, "loss": 0.4103, "step": 35710 }, { "epoch": 2.9564641615626552, "grad_norm": 4.322508335113525, "learning_rate": 6.43832977309633e-09, "loss": 0.4321, "step": 35720 }, { "epoch": 2.957291839099487, "grad_norm": 3.850646495819092, "learning_rate": 6.196359088944115e-09, "loss": 0.4185, "step": 35730 }, { "epoch": 2.9581195166363186, "grad_norm": 4.911293029785156, "learning_rate": 5.959020064382759e-09, "loss": 0.3855, "step": 35740 }, { "epoch": 2.95894719417315, "grad_norm": 5.626308917999268, "learning_rate": 5.7263129195411774e-09, "loss": 0.4394, "step": 35750 }, { "epoch": 2.959774871709982, "grad_norm": 5.535147666931152, "learning_rate": 5.498237870249501e-09, "loss": 0.4583, "step": 35760 }, { "epoch": 2.9606025492468135, "grad_norm": 5.172118186950684, "learning_rate": 5.274795128044075e-09, "loss": 0.4416, "step": 35770 }, { "epoch": 2.961430226783645, "grad_norm": 3.435422658920288, "learning_rate": 5.055984900163568e-09, "loss": 0.3125, "step": 35780 }, { "epoch": 2.962257904320477, "grad_norm": 4.026068210601807, "learning_rate": 4.84180738955009e-09, "loss": 0.4275, "step": 35790 }, { "epoch": 2.9630855818573085, "grad_norm": 5.661964416503906, "learning_rate": 4.63226279484974e-09, "loss": 0.3891, "step": 35800 }, { "epoch": 2.96391325939414, "grad_norm": 6.334976673126221, "learning_rate": 4.427351310410388e-09, "loss": 0.4354, "step": 35810 }, { "epoch": 2.964740936930972, "grad_norm": 4.783035755157471, "learning_rate": 4.227073126285009e-09, "loss": 0.3582, "step": 35820 }, { "epoch": 2.9655686144678035, "grad_norm": 4.494670867919922, "learning_rate": 4.0314284282261255e-09, "loss": 0.3884, "step": 35830 }, { "epoch": 2.966396292004635, "grad_norm": 4.914855003356934, "learning_rate": 3.840417397692475e-09, "loss": 0.3823, "step": 35840 }, { "epoch": 2.9672239695414664, "grad_norm": 5.23052978515625, "learning_rate": 3.6540402118412365e-09, "loss": 0.3654, "step": 35850 }, { "epoch": 2.9680516470782985, "grad_norm": 5.960564613342285, "learning_rate": 3.4722970435352444e-09, "loss": 0.4113, "step": 35860 }, { "epoch": 2.9688793246151297, "grad_norm": 3.6176891326904297, "learning_rate": 3.29518806133744e-09, "loss": 0.4153, "step": 35870 }, { "epoch": 2.969707002151962, "grad_norm": 3.583420753479004, "learning_rate": 3.1227134295130913e-09, "loss": 0.3862, "step": 35880 }, { "epoch": 2.970534679688793, "grad_norm": 4.0642008781433105, "learning_rate": 2.954873308028683e-09, "loss": 0.3903, "step": 35890 }, { "epoch": 2.971362357225625, "grad_norm": 7.276008605957031, "learning_rate": 2.7916678525535813e-09, "loss": 0.3599, "step": 35900 }, { "epoch": 2.9721900347624564, "grad_norm": 2.117279052734375, "learning_rate": 2.6330972144572587e-09, "loss": 0.3315, "step": 35910 }, { "epoch": 2.9730177122992885, "grad_norm": 4.571266174316406, "learning_rate": 2.4791615408115144e-09, "loss": 0.3793, "step": 35920 }, { "epoch": 2.9738453898361197, "grad_norm": 6.166502475738525, "learning_rate": 2.329860974388809e-09, "loss": 0.3736, "step": 35930 }, { "epoch": 2.9746730673729513, "grad_norm": 6.15018367767334, "learning_rate": 2.185195653661709e-09, "loss": 0.391, "step": 35940 }, { "epoch": 2.975500744909783, "grad_norm": 3.3910906314849854, "learning_rate": 2.0451657128056635e-09, "loss": 0.4285, "step": 35950 }, { "epoch": 2.9763284224466147, "grad_norm": 7.489564418792725, "learning_rate": 1.9097712816951164e-09, "loss": 0.3538, "step": 35960 }, { "epoch": 2.9771560999834463, "grad_norm": 3.326009750366211, "learning_rate": 1.7790124859062842e-09, "loss": 0.4195, "step": 35970 }, { "epoch": 2.977983777520278, "grad_norm": 5.221288204193115, "learning_rate": 1.6528894467154888e-09, "loss": 0.395, "step": 35980 }, { "epoch": 2.9788114550571096, "grad_norm": 4.172491550445557, "learning_rate": 1.5314022810997142e-09, "loss": 0.362, "step": 35990 }, { "epoch": 2.9796391325939413, "grad_norm": 3.5094873905181885, "learning_rate": 1.41455110173494e-09, "loss": 0.3917, "step": 36000 }, { "epoch": 2.980466810130773, "grad_norm": 4.813369274139404, "learning_rate": 1.3023360170000276e-09, "loss": 0.3895, "step": 36010 }, { "epoch": 2.9812944876676046, "grad_norm": 4.023046970367432, "learning_rate": 1.1947571309711691e-09, "loss": 0.4331, "step": 36020 }, { "epoch": 2.9821221652044363, "grad_norm": 4.858694553375244, "learning_rate": 1.0918145434257732e-09, "loss": 0.3733, "step": 36030 }, { "epoch": 2.982949842741268, "grad_norm": 3.9494071006774902, "learning_rate": 9.935083498419096e-10, "loss": 0.3962, "step": 36040 }, { "epoch": 2.9837775202780996, "grad_norm": 5.036151885986328, "learning_rate": 8.998386413966442e-10, "loss": 0.4136, "step": 36050 }, { "epoch": 2.9846051978149313, "grad_norm": 5.024433135986328, "learning_rate": 8.108055049665942e-10, "loss": 0.433, "step": 36060 }, { "epoch": 2.985432875351763, "grad_norm": 2.8054895401000977, "learning_rate": 7.264090231279275e-10, "loss": 0.3333, "step": 36070 }, { "epoch": 2.9862605528885946, "grad_norm": 3.8869380950927734, "learning_rate": 6.466492741574736e-10, "loss": 0.3874, "step": 36080 }, { "epoch": 2.9870882304254263, "grad_norm": 4.749934196472168, "learning_rate": 5.715263320310582e-10, "loss": 0.4068, "step": 36090 }, { "epoch": 2.987915907962258, "grad_norm": 3.9954581260681152, "learning_rate": 5.010402664229474e-10, "loss": 0.4343, "step": 36100 }, { "epoch": 2.9887435854990896, "grad_norm": 2.869035243988037, "learning_rate": 4.3519114270806906e-10, "loss": 0.4021, "step": 36110 }, { "epoch": 2.9895712630359212, "grad_norm": 4.604835510253906, "learning_rate": 3.7397902196090186e-10, "loss": 0.361, "step": 36120 }, { "epoch": 2.990398940572753, "grad_norm": 5.96121072769165, "learning_rate": 3.174039609532553e-10, "loss": 0.3756, "step": 36130 }, { "epoch": 2.9912266181095846, "grad_norm": 3.353064775466919, "learning_rate": 2.654660121587105e-10, "loss": 0.4724, "step": 36140 }, { "epoch": 2.9920542956464162, "grad_norm": 4.507892608642578, "learning_rate": 2.1816522374817904e-10, "loss": 0.4039, "step": 36150 }, { "epoch": 2.992881973183248, "grad_norm": 4.885626316070557, "learning_rate": 1.7550163959156875e-10, "loss": 0.4494, "step": 36160 }, { "epoch": 2.9937096507200796, "grad_norm": 4.624135971069336, "learning_rate": 1.3747529926000368e-10, "loss": 0.41, "step": 36170 }, { "epoch": 2.994537328256911, "grad_norm": 4.493701934814453, "learning_rate": 1.0408623802082851e-10, "loss": 0.3819, "step": 36180 }, { "epoch": 2.995365005793743, "grad_norm": 5.522945404052734, "learning_rate": 7.533448684315936e-11, "loss": 0.4329, "step": 36190 }, { "epoch": 2.9961926833305745, "grad_norm": 3.275085926055908, "learning_rate": 5.1220072392332845e-11, "loss": 0.3894, "step": 36200 }, { "epoch": 2.997020360867406, "grad_norm": 4.0116705894470215, "learning_rate": 3.1743017034902015e-11, "loss": 0.4527, "step": 36210 }, { "epoch": 2.997848038404238, "grad_norm": 4.9803571701049805, "learning_rate": 1.6903338834750594e-11, "loss": 0.3915, "step": 36220 }, { "epoch": 2.9986757159410695, "grad_norm": 5.003103256225586, "learning_rate": 6.701051556423644e-12, "loss": 0.3863, "step": 36230 }, { "epoch": 2.9995033934779007, "grad_norm": 3.8307979106903076, "learning_rate": 1.1361646612417787e-12, "loss": 0.3639, "step": 36240 }, { "epoch": 3.0, "step": 36246, "total_flos": 3.31333070252671e+17, "train_loss": 0.5390885038064857, "train_runtime": 13003.5027, "train_samples_per_second": 22.299, "train_steps_per_second": 2.787 } ], "logging_steps": 10, "max_steps": 36246, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 100000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.31333070252671e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }