diff --git "a/multi_student_6/trainer_state.json" "b/multi_student_6/trainer_state.json" new file mode 100644--- /dev/null +++ "b/multi_student_6/trainer_state.json" @@ -0,0 +1,8842 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 7.0, + "eval_steps": 500, + "global_step": 12572, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00556947925368978, + "grad_norm": 19.463690451045245, + "learning_rate": 7.154213036565979e-07, + "loss": 8.2534, + "step": 10 + }, + { + "epoch": 0.01113895850737956, + "grad_norm": 19.73356911714459, + "learning_rate": 1.51033386327504e-06, + "loss": 8.1778, + "step": 20 + }, + { + "epoch": 0.01670843776106934, + "grad_norm": 10.029986538109684, + "learning_rate": 2.3052464228934817e-06, + "loss": 7.6937, + "step": 30 + }, + { + "epoch": 0.02227791701475912, + "grad_norm": 7.379827099032999, + "learning_rate": 3.100158982511924e-06, + "loss": 6.9205, + "step": 40 + }, + { + "epoch": 0.0278473962684489, + "grad_norm": 3.5728211270030625, + "learning_rate": 3.895071542130366e-06, + "loss": 6.3948, + "step": 50 + }, + { + "epoch": 0.03341687552213868, + "grad_norm": 2.8392085477918747, + "learning_rate": 4.689984101748808e-06, + "loss": 6.0575, + "step": 60 + }, + { + "epoch": 0.03898635477582846, + "grad_norm": 3.3563782303356495, + "learning_rate": 5.48489666136725e-06, + "loss": 5.7294, + "step": 70 + }, + { + "epoch": 0.04455583402951824, + "grad_norm": 3.826365209612832, + "learning_rate": 6.279809220985691e-06, + "loss": 5.2614, + "step": 80 + }, + { + "epoch": 0.05012531328320802, + "grad_norm": 3.965749199888462, + "learning_rate": 7.0747217806041336e-06, + "loss": 4.658, + "step": 90 + }, + { + "epoch": 0.0556947925368978, + "grad_norm": 5.147891128586454, + "learning_rate": 7.869634340222575e-06, + "loss": 4.0477, + "step": 100 + }, + { + "epoch": 0.06126427179058758, + "grad_norm": 3.695764472570234, + "learning_rate": 8.664546899841019e-06, + "loss": 3.5584, + "step": 110 + }, + { + "epoch": 0.06683375104427736, + "grad_norm": 4.026692299224468, + "learning_rate": 9.45945945945946e-06, + "loss": 3.2227, + "step": 120 + }, + { + "epoch": 0.07240323029796714, + "grad_norm": 5.389627945808325, + "learning_rate": 1.02543720190779e-05, + "loss": 3.0255, + "step": 130 + }, + { + "epoch": 0.07797270955165692, + "grad_norm": 5.554941152889833, + "learning_rate": 1.1049284578696344e-05, + "loss": 2.8932, + "step": 140 + }, + { + "epoch": 0.0835421888053467, + "grad_norm": 4.8986035107973995, + "learning_rate": 1.1844197138314786e-05, + "loss": 2.7959, + "step": 150 + }, + { + "epoch": 0.08911166805903648, + "grad_norm": 4.56461997585786, + "learning_rate": 1.2639109697933228e-05, + "loss": 2.7067, + "step": 160 + }, + { + "epoch": 0.09468114731272625, + "grad_norm": 9.192995633454027, + "learning_rate": 1.3434022257551671e-05, + "loss": 2.6194, + "step": 170 + }, + { + "epoch": 0.10025062656641603, + "grad_norm": 8.520097501414543, + "learning_rate": 1.4228934817170111e-05, + "loss": 2.5516, + "step": 180 + }, + { + "epoch": 0.10582010582010581, + "grad_norm": 14.274847317292076, + "learning_rate": 1.5023847376788555e-05, + "loss": 2.4789, + "step": 190 + }, + { + "epoch": 0.1113895850737956, + "grad_norm": 19.214437222876512, + "learning_rate": 1.5818759936406995e-05, + "loss": 2.4363, + "step": 200 + }, + { + "epoch": 0.11695906432748537, + "grad_norm": 16.394738746567327, + "learning_rate": 1.6613672496025438e-05, + "loss": 2.3902, + "step": 210 + }, + { + "epoch": 0.12252854358117515, + "grad_norm": 18.247117883573495, + "learning_rate": 1.740858505564388e-05, + "loss": 2.3452, + "step": 220 + }, + { + "epoch": 0.12809802283486493, + "grad_norm": 15.142493870476338, + "learning_rate": 1.820349761526232e-05, + "loss": 2.3042, + "step": 230 + }, + { + "epoch": 0.1336675020885547, + "grad_norm": 8.204095855953893, + "learning_rate": 1.8998410174880765e-05, + "loss": 2.2554, + "step": 240 + }, + { + "epoch": 0.1392369813422445, + "grad_norm": 12.061502490559853, + "learning_rate": 1.9793322734499205e-05, + "loss": 2.2062, + "step": 250 + }, + { + "epoch": 0.14480646059593427, + "grad_norm": 13.078924954972894, + "learning_rate": 2.058823529411765e-05, + "loss": 2.1414, + "step": 260 + }, + { + "epoch": 0.15037593984962405, + "grad_norm": 13.325114907384508, + "learning_rate": 2.1383147853736092e-05, + "loss": 2.059, + "step": 270 + }, + { + "epoch": 0.15594541910331383, + "grad_norm": 7.4709725407687255, + "learning_rate": 2.2178060413354532e-05, + "loss": 1.9772, + "step": 280 + }, + { + "epoch": 0.1615148983570036, + "grad_norm": 19.25846753081156, + "learning_rate": 2.2972972972972976e-05, + "loss": 1.9011, + "step": 290 + }, + { + "epoch": 0.1670843776106934, + "grad_norm": 13.01475345365111, + "learning_rate": 2.3767885532591416e-05, + "loss": 1.8118, + "step": 300 + }, + { + "epoch": 0.17265385686438317, + "grad_norm": 6.508219489647502, + "learning_rate": 2.4562798092209856e-05, + "loss": 1.7365, + "step": 310 + }, + { + "epoch": 0.17822333611807295, + "grad_norm": 8.638960149014848, + "learning_rate": 2.53577106518283e-05, + "loss": 1.6377, + "step": 320 + }, + { + "epoch": 0.18379281537176273, + "grad_norm": 5.254562295866414, + "learning_rate": 2.6152623211446743e-05, + "loss": 1.5737, + "step": 330 + }, + { + "epoch": 0.1893622946254525, + "grad_norm": 8.547148285512923, + "learning_rate": 2.6947535771065186e-05, + "loss": 1.5303, + "step": 340 + }, + { + "epoch": 0.1949317738791423, + "grad_norm": 6.278158376274084, + "learning_rate": 2.7742448330683623e-05, + "loss": 1.4716, + "step": 350 + }, + { + "epoch": 0.20050125313283207, + "grad_norm": 8.26795777275924, + "learning_rate": 2.8537360890302066e-05, + "loss": 1.4151, + "step": 360 + }, + { + "epoch": 0.20607073238652185, + "grad_norm": 4.425290147294405, + "learning_rate": 2.933227344992051e-05, + "loss": 1.368, + "step": 370 + }, + { + "epoch": 0.21164021164021163, + "grad_norm": 6.848350616045537, + "learning_rate": 3.0127186009538953e-05, + "loss": 1.3383, + "step": 380 + }, + { + "epoch": 0.2172096908939014, + "grad_norm": 4.869312013250087, + "learning_rate": 3.092209856915739e-05, + "loss": 1.2751, + "step": 390 + }, + { + "epoch": 0.2227791701475912, + "grad_norm": 7.264173089824176, + "learning_rate": 3.171701112877583e-05, + "loss": 1.2448, + "step": 400 + }, + { + "epoch": 0.22834864940128097, + "grad_norm": 7.343411468837996, + "learning_rate": 3.251192368839428e-05, + "loss": 1.1956, + "step": 410 + }, + { + "epoch": 0.23391812865497075, + "grad_norm": 8.139260855658039, + "learning_rate": 3.330683624801272e-05, + "loss": 1.1626, + "step": 420 + }, + { + "epoch": 0.23948760790866053, + "grad_norm": 4.441795212879553, + "learning_rate": 3.410174880763116e-05, + "loss": 1.1313, + "step": 430 + }, + { + "epoch": 0.2450570871623503, + "grad_norm": 7.0655482262903115, + "learning_rate": 3.489666136724961e-05, + "loss": 1.1016, + "step": 440 + }, + { + "epoch": 0.2506265664160401, + "grad_norm": 5.343797555692575, + "learning_rate": 3.569157392686804e-05, + "loss": 1.0682, + "step": 450 + }, + { + "epoch": 0.25619604566972987, + "grad_norm": 4.837151670486665, + "learning_rate": 3.648648648648649e-05, + "loss": 1.0388, + "step": 460 + }, + { + "epoch": 0.26176552492341965, + "grad_norm": 3.9144650767846345, + "learning_rate": 3.7281399046104934e-05, + "loss": 1.0, + "step": 470 + }, + { + "epoch": 0.2673350041771094, + "grad_norm": 6.192111002424375, + "learning_rate": 3.8076311605723374e-05, + "loss": 0.9819, + "step": 480 + }, + { + "epoch": 0.2729044834307992, + "grad_norm": 4.664894996941613, + "learning_rate": 3.8871224165341814e-05, + "loss": 0.9512, + "step": 490 + }, + { + "epoch": 0.278473962684489, + "grad_norm": 5.728600939316023, + "learning_rate": 3.9666136724960254e-05, + "loss": 0.9292, + "step": 500 + }, + { + "epoch": 0.28404344193817876, + "grad_norm": 7.358090778625861, + "learning_rate": 4.04610492845787e-05, + "loss": 0.9018, + "step": 510 + }, + { + "epoch": 0.28961292119186854, + "grad_norm": 6.346761431472806, + "learning_rate": 4.125596184419714e-05, + "loss": 0.8805, + "step": 520 + }, + { + "epoch": 0.2951824004455583, + "grad_norm": 3.586718245895629, + "learning_rate": 4.205087440381558e-05, + "loss": 0.8568, + "step": 530 + }, + { + "epoch": 0.3007518796992481, + "grad_norm": 5.430039185536721, + "learning_rate": 4.284578696343403e-05, + "loss": 0.8284, + "step": 540 + }, + { + "epoch": 0.3063213589529379, + "grad_norm": 6.430553831376545, + "learning_rate": 4.364069952305246e-05, + "loss": 0.8128, + "step": 550 + }, + { + "epoch": 0.31189083820662766, + "grad_norm": 5.48500374978629, + "learning_rate": 4.443561208267091e-05, + "loss": 0.7842, + "step": 560 + }, + { + "epoch": 0.31746031746031744, + "grad_norm": 5.0124287229570745, + "learning_rate": 4.523052464228935e-05, + "loss": 0.7803, + "step": 570 + }, + { + "epoch": 0.3230297967140072, + "grad_norm": 4.960089415525465, + "learning_rate": 4.6025437201907795e-05, + "loss": 0.7489, + "step": 580 + }, + { + "epoch": 0.328599275967697, + "grad_norm": 4.429208732066051, + "learning_rate": 4.6820349761526235e-05, + "loss": 0.7243, + "step": 590 + }, + { + "epoch": 0.3341687552213868, + "grad_norm": 5.1268739971720905, + "learning_rate": 4.7615262321144675e-05, + "loss": 0.7044, + "step": 600 + }, + { + "epoch": 0.33973823447507656, + "grad_norm": 4.915480885708807, + "learning_rate": 4.841017488076312e-05, + "loss": 0.6939, + "step": 610 + }, + { + "epoch": 0.34530771372876634, + "grad_norm": 5.102518466488775, + "learning_rate": 4.9205087440381556e-05, + "loss": 0.6767, + "step": 620 + }, + { + "epoch": 0.3508771929824561, + "grad_norm": 4.364437305811626, + "learning_rate": 5e-05, + "loss": 0.6467, + "step": 630 + }, + { + "epoch": 0.3564466722361459, + "grad_norm": 4.372236216369743, + "learning_rate": 5.079491255961845e-05, + "loss": 0.6234, + "step": 640 + }, + { + "epoch": 0.3620161514898357, + "grad_norm": 4.439347899189148, + "learning_rate": 5.158982511923689e-05, + "loss": 0.5961, + "step": 650 + }, + { + "epoch": 0.36758563074352546, + "grad_norm": 3.1594832305447316, + "learning_rate": 5.2384737678855336e-05, + "loss": 0.5786, + "step": 660 + }, + { + "epoch": 0.37315510999721524, + "grad_norm": 4.473968649061542, + "learning_rate": 5.317965023847376e-05, + "loss": 0.5711, + "step": 670 + }, + { + "epoch": 0.378724589250905, + "grad_norm": 3.4873086420651456, + "learning_rate": 5.397456279809221e-05, + "loss": 0.5503, + "step": 680 + }, + { + "epoch": 0.3842940685045948, + "grad_norm": 3.831386918783317, + "learning_rate": 5.4769475357710656e-05, + "loss": 0.529, + "step": 690 + }, + { + "epoch": 0.3898635477582846, + "grad_norm": 3.653243174902386, + "learning_rate": 5.5564387917329096e-05, + "loss": 0.5132, + "step": 700 + }, + { + "epoch": 0.39543302701197436, + "grad_norm": 3.5521167023222047, + "learning_rate": 5.635930047694754e-05, + "loss": 0.4877, + "step": 710 + }, + { + "epoch": 0.40100250626566414, + "grad_norm": 3.8712402877933934, + "learning_rate": 5.715421303656598e-05, + "loss": 0.4833, + "step": 720 + }, + { + "epoch": 0.4065719855193539, + "grad_norm": 3.379905907827214, + "learning_rate": 5.794912559618442e-05, + "loss": 0.4615, + "step": 730 + }, + { + "epoch": 0.4121414647730437, + "grad_norm": 2.5936051096516417, + "learning_rate": 5.8744038155802863e-05, + "loss": 0.4444, + "step": 740 + }, + { + "epoch": 0.4177109440267335, + "grad_norm": 3.322838857318257, + "learning_rate": 5.9538950715421304e-05, + "loss": 0.4318, + "step": 750 + }, + { + "epoch": 0.42328042328042326, + "grad_norm": 3.3843216521985253, + "learning_rate": 6.033386327503975e-05, + "loss": 0.4145, + "step": 760 + }, + { + "epoch": 0.42884990253411304, + "grad_norm": 3.0846350440353434, + "learning_rate": 6.11287758346582e-05, + "loss": 0.3998, + "step": 770 + }, + { + "epoch": 0.4344193817878028, + "grad_norm": 2.7361994149983566, + "learning_rate": 6.192368839427664e-05, + "loss": 0.3863, + "step": 780 + }, + { + "epoch": 0.4399888610414926, + "grad_norm": 3.27515261932524, + "learning_rate": 6.271860095389508e-05, + "loss": 0.375, + "step": 790 + }, + { + "epoch": 0.4455583402951824, + "grad_norm": 3.029805585678618, + "learning_rate": 6.351351351351352e-05, + "loss": 0.364, + "step": 800 + }, + { + "epoch": 0.45112781954887216, + "grad_norm": 2.2590750215054287, + "learning_rate": 6.430842607313196e-05, + "loss": 0.3453, + "step": 810 + }, + { + "epoch": 0.45669729880256194, + "grad_norm": 3.1086383281500805, + "learning_rate": 6.51033386327504e-05, + "loss": 0.3329, + "step": 820 + }, + { + "epoch": 0.4622667780562517, + "grad_norm": 2.312028278866729, + "learning_rate": 6.589825119236884e-05, + "loss": 0.324, + "step": 830 + }, + { + "epoch": 0.4678362573099415, + "grad_norm": 2.8653110069877226, + "learning_rate": 6.669316375198729e-05, + "loss": 0.3113, + "step": 840 + }, + { + "epoch": 0.4734057365636313, + "grad_norm": 2.0346455435163224, + "learning_rate": 6.748807631160573e-05, + "loss": 0.2976, + "step": 850 + }, + { + "epoch": 0.47897521581732105, + "grad_norm": 2.2187482142761854, + "learning_rate": 6.828298887122417e-05, + "loss": 0.2838, + "step": 860 + }, + { + "epoch": 0.48454469507101083, + "grad_norm": 2.5760581999823926, + "learning_rate": 6.907790143084261e-05, + "loss": 0.2712, + "step": 870 + }, + { + "epoch": 0.4901141743247006, + "grad_norm": 2.336167318842384, + "learning_rate": 6.987281399046105e-05, + "loss": 0.2659, + "step": 880 + }, + { + "epoch": 0.4956836535783904, + "grad_norm": 1.942101679442901, + "learning_rate": 7.066772655007949e-05, + "loss": 0.2524, + "step": 890 + }, + { + "epoch": 0.5012531328320802, + "grad_norm": 1.9676135441330864, + "learning_rate": 7.146263910969793e-05, + "loss": 0.242, + "step": 900 + }, + { + "epoch": 0.50682261208577, + "grad_norm": 2.42452293659568, + "learning_rate": 7.225755166931639e-05, + "loss": 0.2303, + "step": 910 + }, + { + "epoch": 0.5123920913394597, + "grad_norm": 2.0689938731425754, + "learning_rate": 7.305246422893483e-05, + "loss": 0.2204, + "step": 920 + }, + { + "epoch": 0.5179615705931495, + "grad_norm": 2.171275305780617, + "learning_rate": 7.384737678855325e-05, + "loss": 0.2108, + "step": 930 + }, + { + "epoch": 0.5235310498468393, + "grad_norm": 2.018182012710717, + "learning_rate": 7.46422893481717e-05, + "loss": 0.1988, + "step": 940 + }, + { + "epoch": 0.5291005291005291, + "grad_norm": 1.8435919023282064, + "learning_rate": 7.543720190779015e-05, + "loss": 0.1951, + "step": 950 + }, + { + "epoch": 0.5346700083542189, + "grad_norm": 1.6966065845617362, + "learning_rate": 7.623211446740859e-05, + "loss": 0.1868, + "step": 960 + }, + { + "epoch": 0.5402394876079086, + "grad_norm": 2.107747858389665, + "learning_rate": 7.702702702702703e-05, + "loss": 0.1752, + "step": 970 + }, + { + "epoch": 0.5458089668615984, + "grad_norm": 1.626232566427141, + "learning_rate": 7.782193958664548e-05, + "loss": 0.1726, + "step": 980 + }, + { + "epoch": 0.5513784461152882, + "grad_norm": 1.710543294805638, + "learning_rate": 7.861685214626392e-05, + "loss": 0.1625, + "step": 990 + }, + { + "epoch": 0.556947925368978, + "grad_norm": 1.568201961955694, + "learning_rate": 7.941176470588235e-05, + "loss": 0.1552, + "step": 1000 + }, + { + "epoch": 0.5625174046226677, + "grad_norm": 1.4211321412129432, + "learning_rate": 8.02066772655008e-05, + "loss": 0.1485, + "step": 1010 + }, + { + "epoch": 0.5680868838763575, + "grad_norm": 1.4970140750834295, + "learning_rate": 8.100158982511924e-05, + "loss": 0.1434, + "step": 1020 + }, + { + "epoch": 0.5736563631300473, + "grad_norm": 1.853797381238809, + "learning_rate": 8.179650238473768e-05, + "loss": 0.1356, + "step": 1030 + }, + { + "epoch": 0.5792258423837371, + "grad_norm": 1.475678872579458, + "learning_rate": 8.259141494435613e-05, + "loss": 0.1298, + "step": 1040 + }, + { + "epoch": 0.5847953216374269, + "grad_norm": 1.690324091049369, + "learning_rate": 8.338632750397457e-05, + "loss": 0.1258, + "step": 1050 + }, + { + "epoch": 0.5903648008911166, + "grad_norm": 1.908861076416452, + "learning_rate": 8.418124006359301e-05, + "loss": 0.1249, + "step": 1060 + }, + { + "epoch": 0.5959342801448064, + "grad_norm": 1.236244782970751, + "learning_rate": 8.497615262321145e-05, + "loss": 0.1171, + "step": 1070 + }, + { + "epoch": 0.6015037593984962, + "grad_norm": 1.2774164058752888, + "learning_rate": 8.57710651828299e-05, + "loss": 0.1114, + "step": 1080 + }, + { + "epoch": 0.607073238652186, + "grad_norm": 0.8951848955749875, + "learning_rate": 8.656597774244833e-05, + "loss": 0.1059, + "step": 1090 + }, + { + "epoch": 0.6126427179058758, + "grad_norm": 1.0572189677651873, + "learning_rate": 8.736089030206677e-05, + "loss": 0.1012, + "step": 1100 + }, + { + "epoch": 0.6182121971595655, + "grad_norm": 0.9583280956588, + "learning_rate": 8.815580286168523e-05, + "loss": 0.0973, + "step": 1110 + }, + { + "epoch": 0.6237816764132553, + "grad_norm": 0.966004358455493, + "learning_rate": 8.895071542130367e-05, + "loss": 0.098, + "step": 1120 + }, + { + "epoch": 0.6293511556669451, + "grad_norm": 0.8285031685005042, + "learning_rate": 8.97456279809221e-05, + "loss": 0.0897, + "step": 1130 + }, + { + "epoch": 0.6349206349206349, + "grad_norm": 1.3954391622958595, + "learning_rate": 9.054054054054055e-05, + "loss": 0.087, + "step": 1140 + }, + { + "epoch": 0.6404901141743247, + "grad_norm": 0.6608729564295608, + "learning_rate": 9.133545310015899e-05, + "loss": 0.0825, + "step": 1150 + }, + { + "epoch": 0.6460595934280144, + "grad_norm": 0.8111382270197721, + "learning_rate": 9.213036565977743e-05, + "loss": 0.0764, + "step": 1160 + }, + { + "epoch": 0.6516290726817042, + "grad_norm": 0.6891511752919594, + "learning_rate": 9.292527821939587e-05, + "loss": 0.0745, + "step": 1170 + }, + { + "epoch": 0.657198551935394, + "grad_norm": 0.8999745942051316, + "learning_rate": 9.372019077901432e-05, + "loss": 0.0718, + "step": 1180 + }, + { + "epoch": 0.6627680311890838, + "grad_norm": 0.7165073695545044, + "learning_rate": 9.451510333863276e-05, + "loss": 0.0699, + "step": 1190 + }, + { + "epoch": 0.6683375104427736, + "grad_norm": 0.5691546082373657, + "learning_rate": 9.531001589825119e-05, + "loss": 0.0663, + "step": 1200 + }, + { + "epoch": 0.6739069896964633, + "grad_norm": 0.618154476760873, + "learning_rate": 9.610492845786964e-05, + "loss": 0.065, + "step": 1210 + }, + { + "epoch": 0.6794764689501531, + "grad_norm": 0.5460710019961567, + "learning_rate": 9.689984101748808e-05, + "loss": 0.0622, + "step": 1220 + }, + { + "epoch": 0.6850459482038429, + "grad_norm": 0.44149906211171397, + "learning_rate": 9.769475357710652e-05, + "loss": 0.0582, + "step": 1230 + }, + { + "epoch": 0.6906154274575327, + "grad_norm": 0.5860441580160569, + "learning_rate": 9.848966613672496e-05, + "loss": 0.0567, + "step": 1240 + }, + { + "epoch": 0.6961849067112225, + "grad_norm": 0.8453684625264357, + "learning_rate": 9.928457869634342e-05, + "loss": 0.0564, + "step": 1250 + }, + { + "epoch": 0.7017543859649122, + "grad_norm": 2.1379268435693555, + "learning_rate": 9.999999807244224e-05, + "loss": 0.0564, + "step": 1260 + }, + { + "epoch": 0.707323865218602, + "grad_norm": 3.3074045726639083, + "learning_rate": 9.999976676568994e-05, + "loss": 0.1234, + "step": 1270 + }, + { + "epoch": 0.7128933444722918, + "grad_norm": 0.6698300987807628, + "learning_rate": 9.99991499494276e-05, + "loss": 0.0776, + "step": 1280 + }, + { + "epoch": 0.7184628237259816, + "grad_norm": 0.40986381929677007, + "learning_rate": 9.999814762841102e-05, + "loss": 0.0563, + "step": 1290 + }, + { + "epoch": 0.7240323029796714, + "grad_norm": 0.5823587617428915, + "learning_rate": 9.999675981036832e-05, + "loss": 0.05, + "step": 1300 + }, + { + "epoch": 0.7296017822333611, + "grad_norm": 0.36554660320089016, + "learning_rate": 9.999498650599987e-05, + "loss": 0.0505, + "step": 1310 + }, + { + "epoch": 0.7351712614870509, + "grad_norm": 0.3490628453059406, + "learning_rate": 9.999282772897828e-05, + "loss": 0.0474, + "step": 1320 + }, + { + "epoch": 0.7407407407407407, + "grad_norm": 0.3963213598340753, + "learning_rate": 9.99902834959482e-05, + "loss": 0.0462, + "step": 1330 + }, + { + "epoch": 0.7463102199944305, + "grad_norm": 0.5565743717844861, + "learning_rate": 9.998735382652624e-05, + "loss": 0.0442, + "step": 1340 + }, + { + "epoch": 0.7518796992481203, + "grad_norm": 0.29567277754356086, + "learning_rate": 9.99840387433008e-05, + "loss": 0.0439, + "step": 1350 + }, + { + "epoch": 0.75744917850181, + "grad_norm": 0.42840154636629013, + "learning_rate": 9.998033827183193e-05, + "loss": 0.0432, + "step": 1360 + }, + { + "epoch": 0.7630186577554998, + "grad_norm": 0.3843192755820257, + "learning_rate": 9.997625244065113e-05, + "loss": 0.0419, + "step": 1370 + }, + { + "epoch": 0.7685881370091896, + "grad_norm": 0.2794549966082612, + "learning_rate": 9.997178128126101e-05, + "loss": 0.04, + "step": 1380 + }, + { + "epoch": 0.7741576162628794, + "grad_norm": 0.3169631435418986, + "learning_rate": 9.99669248281353e-05, + "loss": 0.0395, + "step": 1390 + }, + { + "epoch": 0.7797270955165692, + "grad_norm": 0.2751145242333607, + "learning_rate": 9.996168311871832e-05, + "loss": 0.039, + "step": 1400 + }, + { + "epoch": 0.7852965747702589, + "grad_norm": 0.2873246557323452, + "learning_rate": 9.995605619342484e-05, + "loss": 0.0386, + "step": 1410 + }, + { + "epoch": 0.7908660540239487, + "grad_norm": 0.26158401303739276, + "learning_rate": 9.995004409563969e-05, + "loss": 0.0364, + "step": 1420 + }, + { + "epoch": 0.7964355332776385, + "grad_norm": 0.3512702948952268, + "learning_rate": 9.994364687171756e-05, + "loss": 0.0358, + "step": 1430 + }, + { + "epoch": 0.8020050125313283, + "grad_norm": 0.30233442060518095, + "learning_rate": 9.993686457098246e-05, + "loss": 0.0355, + "step": 1440 + }, + { + "epoch": 0.8075744917850181, + "grad_norm": 0.31354227279951397, + "learning_rate": 9.992969724572746e-05, + "loss": 0.0355, + "step": 1450 + }, + { + "epoch": 0.8131439710387078, + "grad_norm": 0.257194967469868, + "learning_rate": 9.992214495121428e-05, + "loss": 0.0357, + "step": 1460 + }, + { + "epoch": 0.8187134502923976, + "grad_norm": 0.3479392485987129, + "learning_rate": 9.99142077456728e-05, + "loss": 0.0345, + "step": 1470 + }, + { + "epoch": 0.8242829295460874, + "grad_norm": 0.24684398703549518, + "learning_rate": 9.990588569030068e-05, + "loss": 0.0337, + "step": 1480 + }, + { + "epoch": 0.8298524087997772, + "grad_norm": 0.23435105711158805, + "learning_rate": 9.989717884926286e-05, + "loss": 0.0328, + "step": 1490 + }, + { + "epoch": 0.835421888053467, + "grad_norm": 0.226826410374761, + "learning_rate": 9.988808728969104e-05, + "loss": 0.0325, + "step": 1500 + }, + { + "epoch": 0.8409913673071567, + "grad_norm": 0.23919961520195954, + "learning_rate": 9.987861108168317e-05, + "loss": 0.0319, + "step": 1510 + }, + { + "epoch": 0.8465608465608465, + "grad_norm": 0.262516092950579, + "learning_rate": 9.986875029830302e-05, + "loss": 0.0311, + "step": 1520 + }, + { + "epoch": 0.8521303258145363, + "grad_norm": 0.27505170240477905, + "learning_rate": 9.985850501557941e-05, + "loss": 0.0306, + "step": 1530 + }, + { + "epoch": 0.8576998050682261, + "grad_norm": 0.2136353900303048, + "learning_rate": 9.984787531250581e-05, + "loss": 0.0306, + "step": 1540 + }, + { + "epoch": 0.8632692843219159, + "grad_norm": 0.2681884993661502, + "learning_rate": 9.983686127103963e-05, + "loss": 0.0304, + "step": 1550 + }, + { + "epoch": 0.8688387635756056, + "grad_norm": 0.27064206276631336, + "learning_rate": 9.982546297610163e-05, + "loss": 0.0302, + "step": 1560 + }, + { + "epoch": 0.8744082428292954, + "grad_norm": 0.22756357273919053, + "learning_rate": 9.981368051557522e-05, + "loss": 0.0296, + "step": 1570 + }, + { + "epoch": 0.8799777220829852, + "grad_norm": 0.23817203224591024, + "learning_rate": 9.980151398030586e-05, + "loss": 0.029, + "step": 1580 + }, + { + "epoch": 0.885547201336675, + "grad_norm": 0.23612451468006537, + "learning_rate": 9.978896346410025e-05, + "loss": 0.0283, + "step": 1590 + }, + { + "epoch": 0.8911166805903648, + "grad_norm": 0.21633179926653082, + "learning_rate": 9.977602906372577e-05, + "loss": 0.0284, + "step": 1600 + }, + { + "epoch": 0.8966861598440545, + "grad_norm": 0.3007834433966076, + "learning_rate": 9.97627108789095e-05, + "loss": 0.0276, + "step": 1610 + }, + { + "epoch": 0.9022556390977443, + "grad_norm": 0.22874451896032774, + "learning_rate": 9.974900901233772e-05, + "loss": 0.0275, + "step": 1620 + }, + { + "epoch": 0.9078251183514341, + "grad_norm": 0.21229440559613427, + "learning_rate": 9.973492356965488e-05, + "loss": 0.0272, + "step": 1630 + }, + { + "epoch": 0.9133945976051239, + "grad_norm": 0.21316799793517352, + "learning_rate": 9.972045465946295e-05, + "loss": 0.0266, + "step": 1640 + }, + { + "epoch": 0.9189640768588136, + "grad_norm": 0.21840470782406107, + "learning_rate": 9.97056023933205e-05, + "loss": 0.0262, + "step": 1650 + }, + { + "epoch": 0.9245335561125034, + "grad_norm": 0.2296202067200069, + "learning_rate": 9.969036688574184e-05, + "loss": 0.0262, + "step": 1660 + }, + { + "epoch": 0.9301030353661932, + "grad_norm": 0.20035916457466751, + "learning_rate": 9.967474825419619e-05, + "loss": 0.0256, + "step": 1670 + }, + { + "epoch": 0.935672514619883, + "grad_norm": 0.2392722746437584, + "learning_rate": 9.965874661910675e-05, + "loss": 0.0252, + "step": 1680 + }, + { + "epoch": 0.9412419938735728, + "grad_norm": 0.2373750894769379, + "learning_rate": 9.964236210384972e-05, + "loss": 0.025, + "step": 1690 + }, + { + "epoch": 0.9468114731272625, + "grad_norm": 0.23565128407439787, + "learning_rate": 9.962559483475342e-05, + "loss": 0.0245, + "step": 1700 + }, + { + "epoch": 0.9523809523809523, + "grad_norm": 0.17651355447696596, + "learning_rate": 9.960844494109731e-05, + "loss": 0.024, + "step": 1710 + }, + { + "epoch": 0.9579504316346421, + "grad_norm": 0.19685563495858988, + "learning_rate": 9.959091255511093e-05, + "loss": 0.0237, + "step": 1720 + }, + { + "epoch": 0.9635199108883319, + "grad_norm": 0.1904961294977808, + "learning_rate": 9.957299781197295e-05, + "loss": 0.0235, + "step": 1730 + }, + { + "epoch": 0.9690893901420217, + "grad_norm": 0.21714296341973274, + "learning_rate": 9.955470084981009e-05, + "loss": 0.0232, + "step": 1740 + }, + { + "epoch": 0.9746588693957114, + "grad_norm": 0.21953603915286493, + "learning_rate": 9.953602180969606e-05, + "loss": 0.0227, + "step": 1750 + }, + { + "epoch": 0.9802283486494012, + "grad_norm": 0.21497187075207344, + "learning_rate": 9.951696083565048e-05, + "loss": 0.0224, + "step": 1760 + }, + { + "epoch": 0.985797827903091, + "grad_norm": 0.18608055976176674, + "learning_rate": 9.949751807463777e-05, + "loss": 0.0226, + "step": 1770 + }, + { + "epoch": 0.9913673071567808, + "grad_norm": 0.20352070118783105, + "learning_rate": 9.947769367656603e-05, + "loss": 0.0223, + "step": 1780 + }, + { + "epoch": 0.9969367864104706, + "grad_norm": 0.24791255800751252, + "learning_rate": 9.945748779428585e-05, + "loss": 0.0217, + "step": 1790 + }, + { + "epoch": 1.002227791701476, + "grad_norm": 0.22472692732126306, + "learning_rate": 9.943690058358915e-05, + "loss": 0.0217, + "step": 1800 + }, + { + "epoch": 1.0077972709551657, + "grad_norm": 0.2324141522417415, + "learning_rate": 9.941593220320797e-05, + "loss": 0.021, + "step": 1810 + }, + { + "epoch": 1.0133667502088555, + "grad_norm": 0.19233986715335943, + "learning_rate": 9.939458281481327e-05, + "loss": 0.0212, + "step": 1820 + }, + { + "epoch": 1.0189362294625453, + "grad_norm": 0.19005795401708103, + "learning_rate": 9.937285258301368e-05, + "loss": 0.0204, + "step": 1830 + }, + { + "epoch": 1.024505708716235, + "grad_norm": 0.21691243705398805, + "learning_rate": 9.935074167535419e-05, + "loss": 0.0205, + "step": 1840 + }, + { + "epoch": 1.0300751879699248, + "grad_norm": 0.21307344896990474, + "learning_rate": 9.932825026231488e-05, + "loss": 0.0197, + "step": 1850 + }, + { + "epoch": 1.0356446672236146, + "grad_norm": 0.2554320699991988, + "learning_rate": 9.930537851730966e-05, + "loss": 0.0193, + "step": 1860 + }, + { + "epoch": 1.0412141464773044, + "grad_norm": 0.204769243946618, + "learning_rate": 9.928212661668483e-05, + "loss": 0.0197, + "step": 1870 + }, + { + "epoch": 1.0467836257309941, + "grad_norm": 0.19859219088449717, + "learning_rate": 9.925849473971784e-05, + "loss": 0.0196, + "step": 1880 + }, + { + "epoch": 1.052353104984684, + "grad_norm": 0.20983760283366612, + "learning_rate": 9.923448306861577e-05, + "loss": 0.019, + "step": 1890 + }, + { + "epoch": 1.0579225842383737, + "grad_norm": 0.20589867040322582, + "learning_rate": 9.921009178851405e-05, + "loss": 0.0193, + "step": 1900 + }, + { + "epoch": 1.0634920634920635, + "grad_norm": 0.18568807424493425, + "learning_rate": 9.918532108747496e-05, + "loss": 0.0189, + "step": 1910 + }, + { + "epoch": 1.0690615427457533, + "grad_norm": 0.24088752127036372, + "learning_rate": 9.916017115648622e-05, + "loss": 0.0186, + "step": 1920 + }, + { + "epoch": 1.074631021999443, + "grad_norm": 0.1926030304649344, + "learning_rate": 9.913464218945946e-05, + "loss": 0.0185, + "step": 1930 + }, + { + "epoch": 1.0802005012531328, + "grad_norm": 0.8741730355575924, + "learning_rate": 9.910873438322884e-05, + "loss": 0.0199, + "step": 1940 + }, + { + "epoch": 1.0857699805068226, + "grad_norm": 0.2217866848455536, + "learning_rate": 9.908244793754936e-05, + "loss": 0.0193, + "step": 1950 + }, + { + "epoch": 1.0913394597605124, + "grad_norm": 0.21213745327394049, + "learning_rate": 9.905578305509546e-05, + "loss": 0.0183, + "step": 1960 + }, + { + "epoch": 1.0969089390142022, + "grad_norm": 0.18251989541031047, + "learning_rate": 9.902873994145945e-05, + "loss": 0.0178, + "step": 1970 + }, + { + "epoch": 1.102478418267892, + "grad_norm": 0.17043242205288903, + "learning_rate": 9.900131880514982e-05, + "loss": 0.0173, + "step": 1980 + }, + { + "epoch": 1.1080478975215817, + "grad_norm": 0.18442275031618716, + "learning_rate": 9.897351985758974e-05, + "loss": 0.0169, + "step": 1990 + }, + { + "epoch": 1.1136173767752715, + "grad_norm": 0.169313148723845, + "learning_rate": 9.89453433131154e-05, + "loss": 0.0169, + "step": 2000 + }, + { + "epoch": 1.1191868560289613, + "grad_norm": 0.1764961917184802, + "learning_rate": 9.891678938897431e-05, + "loss": 0.0167, + "step": 2010 + }, + { + "epoch": 1.124756335282651, + "grad_norm": 0.17888846393164137, + "learning_rate": 9.888785830532368e-05, + "loss": 0.0165, + "step": 2020 + }, + { + "epoch": 1.1303258145363408, + "grad_norm": 0.19125391501293337, + "learning_rate": 9.885855028522873e-05, + "loss": 0.0167, + "step": 2030 + }, + { + "epoch": 1.1358952937900306, + "grad_norm": 0.19194405473426313, + "learning_rate": 9.882886555466091e-05, + "loss": 0.0161, + "step": 2040 + }, + { + "epoch": 1.1414647730437204, + "grad_norm": 0.1770624411187416, + "learning_rate": 9.879880434249621e-05, + "loss": 0.0161, + "step": 2050 + }, + { + "epoch": 1.1470342522974102, + "grad_norm": 0.16975918892019828, + "learning_rate": 9.876836688051337e-05, + "loss": 0.0161, + "step": 2060 + }, + { + "epoch": 1.1526037315511, + "grad_norm": 0.16854837841564632, + "learning_rate": 9.873755340339211e-05, + "loss": 0.0163, + "step": 2070 + }, + { + "epoch": 1.1581732108047897, + "grad_norm": 0.17269429721308055, + "learning_rate": 9.870636414871131e-05, + "loss": 0.0159, + "step": 2080 + }, + { + "epoch": 1.1637426900584795, + "grad_norm": 0.17625877797958917, + "learning_rate": 9.867479935694717e-05, + "loss": 0.0153, + "step": 2090 + }, + { + "epoch": 1.1693121693121693, + "grad_norm": 0.15960923595525844, + "learning_rate": 9.864285927147138e-05, + "loss": 0.0152, + "step": 2100 + }, + { + "epoch": 1.174881648565859, + "grad_norm": 0.15697254246300588, + "learning_rate": 9.861054413854922e-05, + "loss": 0.015, + "step": 2110 + }, + { + "epoch": 1.1804511278195489, + "grad_norm": 0.16827296886706583, + "learning_rate": 9.857785420733768e-05, + "loss": 0.0149, + "step": 2120 + }, + { + "epoch": 1.1860206070732386, + "grad_norm": 0.15126212242845452, + "learning_rate": 9.854478972988351e-05, + "loss": 0.0146, + "step": 2130 + }, + { + "epoch": 1.1915900863269284, + "grad_norm": 0.18190859691476494, + "learning_rate": 9.85113509611213e-05, + "loss": 0.0147, + "step": 2140 + }, + { + "epoch": 1.1971595655806182, + "grad_norm": 0.1620227168907316, + "learning_rate": 9.847753815887155e-05, + "loss": 0.0144, + "step": 2150 + }, + { + "epoch": 1.202729044834308, + "grad_norm": 0.16437934378165805, + "learning_rate": 9.84433515838386e-05, + "loss": 0.0144, + "step": 2160 + }, + { + "epoch": 1.2082985240879978, + "grad_norm": 0.19971461246923847, + "learning_rate": 9.840879149960866e-05, + "loss": 0.0144, + "step": 2170 + }, + { + "epoch": 1.2138680033416875, + "grad_norm": 0.17545008123882608, + "learning_rate": 9.837385817264783e-05, + "loss": 0.014, + "step": 2180 + }, + { + "epoch": 1.2194374825953773, + "grad_norm": 0.17002175092464197, + "learning_rate": 9.833855187229993e-05, + "loss": 0.0141, + "step": 2190 + }, + { + "epoch": 1.225006961849067, + "grad_norm": 0.1502742530729085, + "learning_rate": 9.830287287078453e-05, + "loss": 0.0138, + "step": 2200 + }, + { + "epoch": 1.2305764411027569, + "grad_norm": 0.1733361600396075, + "learning_rate": 9.826682144319481e-05, + "loss": 0.0136, + "step": 2210 + }, + { + "epoch": 1.2361459203564467, + "grad_norm": 0.1703011609444242, + "learning_rate": 9.823039786749542e-05, + "loss": 0.0135, + "step": 2220 + }, + { + "epoch": 1.2417153996101364, + "grad_norm": 0.25652792710234223, + "learning_rate": 9.819360242452037e-05, + "loss": 0.0137, + "step": 2230 + }, + { + "epoch": 1.2472848788638262, + "grad_norm": 0.18955265882691177, + "learning_rate": 9.815643539797085e-05, + "loss": 0.0135, + "step": 2240 + }, + { + "epoch": 1.252854358117516, + "grad_norm": 0.176670947546437, + "learning_rate": 9.811889707441304e-05, + "loss": 0.0133, + "step": 2250 + }, + { + "epoch": 1.2584238373712058, + "grad_norm": 0.1611843791331948, + "learning_rate": 9.808098774327589e-05, + "loss": 0.0127, + "step": 2260 + }, + { + "epoch": 1.2639933166248956, + "grad_norm": 0.17553791221607762, + "learning_rate": 9.804270769684894e-05, + "loss": 0.0128, + "step": 2270 + }, + { + "epoch": 1.2695627958785853, + "grad_norm": 0.1713258488893283, + "learning_rate": 9.800405723027998e-05, + "loss": 0.0125, + "step": 2280 + }, + { + "epoch": 1.2751322751322751, + "grad_norm": 0.1491188978760988, + "learning_rate": 9.796503664157289e-05, + "loss": 0.0125, + "step": 2290 + }, + { + "epoch": 1.280701754385965, + "grad_norm": 0.16382367249207125, + "learning_rate": 9.792564623158518e-05, + "loss": 0.0123, + "step": 2300 + }, + { + "epoch": 1.2862712336396547, + "grad_norm": 0.14990536572540042, + "learning_rate": 9.788588630402586e-05, + "loss": 0.012, + "step": 2310 + }, + { + "epoch": 1.2918407128933445, + "grad_norm": 0.1612176660850985, + "learning_rate": 9.784575716545294e-05, + "loss": 0.0118, + "step": 2320 + }, + { + "epoch": 1.2974101921470342, + "grad_norm": 0.13562541267893746, + "learning_rate": 9.780525912527118e-05, + "loss": 0.0118, + "step": 2330 + }, + { + "epoch": 1.302979671400724, + "grad_norm": 0.1680412513721527, + "learning_rate": 9.776439249572962e-05, + "loss": 0.0114, + "step": 2340 + }, + { + "epoch": 1.3085491506544138, + "grad_norm": 0.15829618023893888, + "learning_rate": 9.77231575919192e-05, + "loss": 0.0115, + "step": 2350 + }, + { + "epoch": 1.3141186299081036, + "grad_norm": 0.16260390780626038, + "learning_rate": 9.768155473177037e-05, + "loss": 0.0116, + "step": 2360 + }, + { + "epoch": 1.3196881091617934, + "grad_norm": 0.16275068223487438, + "learning_rate": 9.763958423605057e-05, + "loss": 0.0115, + "step": 2370 + }, + { + "epoch": 1.3252575884154831, + "grad_norm": 0.20119872670951966, + "learning_rate": 9.759724642836184e-05, + "loss": 0.0109, + "step": 2380 + }, + { + "epoch": 1.330827067669173, + "grad_norm": 0.172255342957574, + "learning_rate": 9.755454163513824e-05, + "loss": 0.0108, + "step": 2390 + }, + { + "epoch": 1.3363965469228627, + "grad_norm": 0.15223068522445793, + "learning_rate": 9.75114701856434e-05, + "loss": 0.0109, + "step": 2400 + }, + { + "epoch": 1.3419660261765525, + "grad_norm": 0.14759093427536293, + "learning_rate": 9.746803241196791e-05, + "loss": 0.0108, + "step": 2410 + }, + { + "epoch": 1.3475355054302423, + "grad_norm": 0.14365720679126023, + "learning_rate": 9.742422864902684e-05, + "loss": 0.0103, + "step": 2420 + }, + { + "epoch": 1.353104984683932, + "grad_norm": 0.14640480563703434, + "learning_rate": 9.738005923455714e-05, + "loss": 0.0104, + "step": 2430 + }, + { + "epoch": 1.3586744639376218, + "grad_norm": 0.13917727110739087, + "learning_rate": 9.733552450911494e-05, + "loss": 0.0104, + "step": 2440 + }, + { + "epoch": 1.3642439431913116, + "grad_norm": 0.14296325607296143, + "learning_rate": 9.729062481607307e-05, + "loss": 0.01, + "step": 2450 + }, + { + "epoch": 1.3698134224450014, + "grad_norm": 0.1513321418311315, + "learning_rate": 9.72453605016183e-05, + "loss": 0.0099, + "step": 2460 + }, + { + "epoch": 1.3753829016986912, + "grad_norm": 0.13542887753220842, + "learning_rate": 9.719973191474875e-05, + "loss": 0.0097, + "step": 2470 + }, + { + "epoch": 1.380952380952381, + "grad_norm": 0.16834906807878625, + "learning_rate": 9.715373940727112e-05, + "loss": 0.0097, + "step": 2480 + }, + { + "epoch": 1.3865218602060707, + "grad_norm": 0.14493297637350916, + "learning_rate": 9.71073833337981e-05, + "loss": 0.0099, + "step": 2490 + }, + { + "epoch": 1.3920913394597605, + "grad_norm": 0.13388729788015086, + "learning_rate": 9.706066405174543e-05, + "loss": 0.0096, + "step": 2500 + }, + { + "epoch": 1.3976608187134503, + "grad_norm": 0.12948166422313045, + "learning_rate": 9.701358192132939e-05, + "loss": 0.0095, + "step": 2510 + }, + { + "epoch": 1.40323029796714, + "grad_norm": 0.14831134004472454, + "learning_rate": 9.696613730556383e-05, + "loss": 0.009, + "step": 2520 + }, + { + "epoch": 1.4087997772208298, + "grad_norm": 0.13855341547369657, + "learning_rate": 9.691833057025747e-05, + "loss": 0.0092, + "step": 2530 + }, + { + "epoch": 1.4143692564745196, + "grad_norm": 0.1333888893995881, + "learning_rate": 9.687016208401107e-05, + "loss": 0.0093, + "step": 2540 + }, + { + "epoch": 1.4199387357282094, + "grad_norm": 0.13806477992850671, + "learning_rate": 9.682163221821455e-05, + "loss": 0.0091, + "step": 2550 + }, + { + "epoch": 1.4255082149818992, + "grad_norm": 0.1424299673717202, + "learning_rate": 9.677274134704411e-05, + "loss": 0.009, + "step": 2560 + }, + { + "epoch": 1.431077694235589, + "grad_norm": 0.11566085539435643, + "learning_rate": 9.672348984745949e-05, + "loss": 0.0088, + "step": 2570 + }, + { + "epoch": 1.4366471734892787, + "grad_norm": 0.12261270767102363, + "learning_rate": 9.667387809920083e-05, + "loss": 0.0087, + "step": 2580 + }, + { + "epoch": 1.4422166527429685, + "grad_norm": 0.16396158500026034, + "learning_rate": 9.662390648478597e-05, + "loss": 0.0086, + "step": 2590 + }, + { + "epoch": 1.4477861319966583, + "grad_norm": 0.1532929944740895, + "learning_rate": 9.657357538950733e-05, + "loss": 0.0088, + "step": 2600 + }, + { + "epoch": 1.453355611250348, + "grad_norm": 0.14377084311922123, + "learning_rate": 9.652288520142907e-05, + "loss": 0.0087, + "step": 2610 + }, + { + "epoch": 1.4589250905040378, + "grad_norm": 0.13047962164454835, + "learning_rate": 9.647183631138397e-05, + "loss": 0.0084, + "step": 2620 + }, + { + "epoch": 1.4644945697577276, + "grad_norm": 0.13602569186021218, + "learning_rate": 9.642042911297055e-05, + "loss": 0.0078, + "step": 2630 + }, + { + "epoch": 1.4700640490114174, + "grad_norm": 0.11739120328716636, + "learning_rate": 9.636866400254991e-05, + "loss": 0.008, + "step": 2640 + }, + { + "epoch": 1.4756335282651072, + "grad_norm": 0.13171022429436544, + "learning_rate": 9.631654137924278e-05, + "loss": 0.008, + "step": 2650 + }, + { + "epoch": 1.481203007518797, + "grad_norm": 0.12987231380199957, + "learning_rate": 9.626406164492636e-05, + "loss": 0.0081, + "step": 2660 + }, + { + "epoch": 1.4867724867724867, + "grad_norm": 0.11746026303181731, + "learning_rate": 9.621122520423128e-05, + "loss": 0.0079, + "step": 2670 + }, + { + "epoch": 1.4923419660261765, + "grad_norm": 0.11452815904361804, + "learning_rate": 9.615803246453843e-05, + "loss": 0.0078, + "step": 2680 + }, + { + "epoch": 1.4979114452798663, + "grad_norm": 0.111682722051899, + "learning_rate": 9.61044838359759e-05, + "loss": 0.0079, + "step": 2690 + }, + { + "epoch": 1.503480924533556, + "grad_norm": 0.11111224496044826, + "learning_rate": 9.60505797314157e-05, + "loss": 0.0081, + "step": 2700 + }, + { + "epoch": 1.5090504037872459, + "grad_norm": 0.12434768563055779, + "learning_rate": 9.599632056647065e-05, + "loss": 0.0075, + "step": 2710 + }, + { + "epoch": 1.5146198830409356, + "grad_norm": 0.12762658726155382, + "learning_rate": 9.594170675949122e-05, + "loss": 0.0077, + "step": 2720 + }, + { + "epoch": 1.5201893622946254, + "grad_norm": 0.13462630987728993, + "learning_rate": 9.588673873156218e-05, + "loss": 0.0074, + "step": 2730 + }, + { + "epoch": 1.5257588415483152, + "grad_norm": 0.11873990531246101, + "learning_rate": 9.583141690649949e-05, + "loss": 0.0075, + "step": 2740 + }, + { + "epoch": 1.531328320802005, + "grad_norm": 0.11897341907272085, + "learning_rate": 9.577574171084691e-05, + "loss": 0.0072, + "step": 2750 + }, + { + "epoch": 1.5368978000556948, + "grad_norm": 0.1342155424043113, + "learning_rate": 9.571971357387281e-05, + "loss": 0.0073, + "step": 2760 + }, + { + "epoch": 1.5424672793093845, + "grad_norm": 0.11109000135852713, + "learning_rate": 9.566333292756678e-05, + "loss": 0.0073, + "step": 2770 + }, + { + "epoch": 1.5480367585630743, + "grad_norm": 0.10749381139994726, + "learning_rate": 9.560660020663636e-05, + "loss": 0.0071, + "step": 2780 + }, + { + "epoch": 1.553606237816764, + "grad_norm": 0.12678846265983787, + "learning_rate": 9.554951584850366e-05, + "loss": 0.0072, + "step": 2790 + }, + { + "epoch": 1.5591757170704539, + "grad_norm": 0.11887715567893196, + "learning_rate": 9.549208029330198e-05, + "loss": 0.0071, + "step": 2800 + }, + { + "epoch": 1.5647451963241437, + "grad_norm": 0.11528952921949762, + "learning_rate": 9.543429398387246e-05, + "loss": 0.0072, + "step": 2810 + }, + { + "epoch": 1.5703146755778334, + "grad_norm": 0.1130261076594759, + "learning_rate": 9.537615736576062e-05, + "loss": 0.0068, + "step": 2820 + }, + { + "epoch": 1.5758841548315232, + "grad_norm": 0.1150612035172818, + "learning_rate": 9.531767088721293e-05, + "loss": 0.0068, + "step": 2830 + }, + { + "epoch": 1.581453634085213, + "grad_norm": 0.12119751877826486, + "learning_rate": 9.525883499917334e-05, + "loss": 0.0066, + "step": 2840 + }, + { + "epoch": 1.5870231133389028, + "grad_norm": 0.14369604650594714, + "learning_rate": 9.519965015527987e-05, + "loss": 0.0067, + "step": 2850 + }, + { + "epoch": 1.5925925925925926, + "grad_norm": 0.11828049051691453, + "learning_rate": 9.514011681186107e-05, + "loss": 0.0067, + "step": 2860 + }, + { + "epoch": 1.5981620718462823, + "grad_norm": 0.1041622423129153, + "learning_rate": 9.508023542793246e-05, + "loss": 0.0067, + "step": 2870 + }, + { + "epoch": 1.6037315510999721, + "grad_norm": 0.109441034459224, + "learning_rate": 9.502000646519304e-05, + "loss": 0.0064, + "step": 2880 + }, + { + "epoch": 1.609301030353662, + "grad_norm": 0.11407457138432828, + "learning_rate": 9.495943038802178e-05, + "loss": 0.0064, + "step": 2890 + }, + { + "epoch": 1.6148705096073517, + "grad_norm": 0.11169346620933446, + "learning_rate": 9.48985076634739e-05, + "loss": 0.0063, + "step": 2900 + }, + { + "epoch": 1.6204399888610415, + "grad_norm": 0.1068634570105383, + "learning_rate": 9.48372387612774e-05, + "loss": 0.0063, + "step": 2910 + }, + { + "epoch": 1.6260094681147312, + "grad_norm": 0.10960556348842432, + "learning_rate": 9.477562415382937e-05, + "loss": 0.0064, + "step": 2920 + }, + { + "epoch": 1.631578947368421, + "grad_norm": 0.09964438396517761, + "learning_rate": 9.471366431619236e-05, + "loss": 0.0063, + "step": 2930 + }, + { + "epoch": 1.6371484266221108, + "grad_norm": 0.10857713507090318, + "learning_rate": 9.465135972609074e-05, + "loss": 0.0061, + "step": 2940 + }, + { + "epoch": 1.6427179058758006, + "grad_norm": 0.12812497053072483, + "learning_rate": 9.458871086390699e-05, + "loss": 0.0063, + "step": 2950 + }, + { + "epoch": 1.6482873851294904, + "grad_norm": 0.12920573246968883, + "learning_rate": 9.452571821267799e-05, + "loss": 0.0062, + "step": 2960 + }, + { + "epoch": 1.6538568643831801, + "grad_norm": 0.11378467589478443, + "learning_rate": 9.446238225809137e-05, + "loss": 0.0061, + "step": 2970 + }, + { + "epoch": 1.65942634363687, + "grad_norm": 0.1025631830473413, + "learning_rate": 9.439870348848162e-05, + "loss": 0.0062, + "step": 2980 + }, + { + "epoch": 1.6649958228905597, + "grad_norm": 0.09752210801488877, + "learning_rate": 9.433468239482647e-05, + "loss": 0.0063, + "step": 2990 + }, + { + "epoch": 1.6705653021442495, + "grad_norm": 0.10280694509228655, + "learning_rate": 9.427031947074305e-05, + "loss": 0.006, + "step": 3000 + }, + { + "epoch": 1.6761347813979393, + "grad_norm": 0.10219752893301338, + "learning_rate": 9.420561521248403e-05, + "loss": 0.0059, + "step": 3010 + }, + { + "epoch": 1.681704260651629, + "grad_norm": 0.10744460806316994, + "learning_rate": 9.414057011893389e-05, + "loss": 0.0058, + "step": 3020 + }, + { + "epoch": 1.6872737399053188, + "grad_norm": 0.10272875016713308, + "learning_rate": 9.407518469160504e-05, + "loss": 0.0057, + "step": 3030 + }, + { + "epoch": 1.6928432191590086, + "grad_norm": 0.10784343296872123, + "learning_rate": 9.400945943463387e-05, + "loss": 0.0056, + "step": 3040 + }, + { + "epoch": 1.6984126984126984, + "grad_norm": 0.1265310671727756, + "learning_rate": 9.394339485477697e-05, + "loss": 0.0059, + "step": 3050 + }, + { + "epoch": 1.7039821776663882, + "grad_norm": 0.12380016024169994, + "learning_rate": 9.387699146140724e-05, + "loss": 0.0058, + "step": 3060 + }, + { + "epoch": 1.709551656920078, + "grad_norm": 0.1116972787626009, + "learning_rate": 9.381024976650983e-05, + "loss": 0.0055, + "step": 3070 + }, + { + "epoch": 1.7151211361737677, + "grad_norm": 0.10613073795533987, + "learning_rate": 9.37431702846783e-05, + "loss": 0.0055, + "step": 3080 + }, + { + "epoch": 1.7206906154274575, + "grad_norm": 0.11702121991080053, + "learning_rate": 9.367575353311064e-05, + "loss": 0.0058, + "step": 3090 + }, + { + "epoch": 1.7262600946811473, + "grad_norm": 0.10126745527171219, + "learning_rate": 9.360800003160525e-05, + "loss": 0.0057, + "step": 3100 + }, + { + "epoch": 1.731829573934837, + "grad_norm": 0.10028282479858917, + "learning_rate": 9.353991030255695e-05, + "loss": 0.0056, + "step": 3110 + }, + { + "epoch": 1.7373990531885268, + "grad_norm": 0.1102978478866508, + "learning_rate": 9.347148487095294e-05, + "loss": 0.0055, + "step": 3120 + }, + { + "epoch": 1.7429685324422166, + "grad_norm": 0.10831564723664697, + "learning_rate": 9.340272426436877e-05, + "loss": 0.0056, + "step": 3130 + }, + { + "epoch": 1.7485380116959064, + "grad_norm": 0.1012748795569494, + "learning_rate": 9.33336290129643e-05, + "loss": 0.0053, + "step": 3140 + }, + { + "epoch": 1.7541074909495962, + "grad_norm": 0.11575041254493004, + "learning_rate": 9.32641996494795e-05, + "loss": 0.0052, + "step": 3150 + }, + { + "epoch": 1.759676970203286, + "grad_norm": 0.10943195809368053, + "learning_rate": 9.319443670923051e-05, + "loss": 0.0056, + "step": 3160 + }, + { + "epoch": 1.7652464494569757, + "grad_norm": 0.0939129237227479, + "learning_rate": 9.312434073010534e-05, + "loss": 0.0052, + "step": 3170 + }, + { + "epoch": 1.7708159287106655, + "grad_norm": 0.1047650747357604, + "learning_rate": 9.305391225255986e-05, + "loss": 0.0053, + "step": 3180 + }, + { + "epoch": 1.7763854079643553, + "grad_norm": 0.09958975875505627, + "learning_rate": 9.298315181961356e-05, + "loss": 0.0052, + "step": 3190 + }, + { + "epoch": 1.781954887218045, + "grad_norm": 0.10425138001460964, + "learning_rate": 9.291205997684538e-05, + "loss": 0.0053, + "step": 3200 + }, + { + "epoch": 1.7875243664717348, + "grad_norm": 0.09069036788388674, + "learning_rate": 9.284063727238952e-05, + "loss": 0.0051, + "step": 3210 + }, + { + "epoch": 1.7930938457254246, + "grad_norm": 0.09833804691470377, + "learning_rate": 9.276888425693115e-05, + "loss": 0.0049, + "step": 3220 + }, + { + "epoch": 1.7986633249791144, + "grad_norm": 0.08709400377373291, + "learning_rate": 9.269680148370228e-05, + "loss": 0.0048, + "step": 3230 + }, + { + "epoch": 1.8042328042328042, + "grad_norm": 0.11915976087290946, + "learning_rate": 9.262438950847738e-05, + "loss": 0.0048, + "step": 3240 + }, + { + "epoch": 1.809802283486494, + "grad_norm": 0.08988574166722815, + "learning_rate": 9.255164888956918e-05, + "loss": 0.005, + "step": 3250 + }, + { + "epoch": 1.8153717627401837, + "grad_norm": 0.09796027881734362, + "learning_rate": 9.247858018782425e-05, + "loss": 0.0052, + "step": 3260 + }, + { + "epoch": 1.8209412419938735, + "grad_norm": 0.09351986222567911, + "learning_rate": 9.240518396661886e-05, + "loss": 0.0048, + "step": 3270 + }, + { + "epoch": 1.8265107212475633, + "grad_norm": 0.09530159956409921, + "learning_rate": 9.233146079185444e-05, + "loss": 0.0049, + "step": 3280 + }, + { + "epoch": 1.832080200501253, + "grad_norm": 0.09731186711139259, + "learning_rate": 9.225741123195338e-05, + "loss": 0.005, + "step": 3290 + }, + { + "epoch": 1.8376496797549429, + "grad_norm": 0.08614001336056387, + "learning_rate": 9.218303585785448e-05, + "loss": 0.0047, + "step": 3300 + }, + { + "epoch": 1.8432191590086326, + "grad_norm": 0.09126829864578764, + "learning_rate": 9.210833524300873e-05, + "loss": 0.0048, + "step": 3310 + }, + { + "epoch": 1.8487886382623224, + "grad_norm": 0.08448191010827921, + "learning_rate": 9.203330996337477e-05, + "loss": 0.0048, + "step": 3320 + }, + { + "epoch": 1.8543581175160122, + "grad_norm": 0.08979291201416467, + "learning_rate": 9.195796059741447e-05, + "loss": 0.0046, + "step": 3330 + }, + { + "epoch": 1.859927596769702, + "grad_norm": 0.09340579831138453, + "learning_rate": 9.188228772608846e-05, + "loss": 0.0046, + "step": 3340 + }, + { + "epoch": 1.8654970760233918, + "grad_norm": 0.0924248400058176, + "learning_rate": 9.180629193285172e-05, + "loss": 0.0047, + "step": 3350 + }, + { + "epoch": 1.8710665552770815, + "grad_norm": 0.09465367072304114, + "learning_rate": 9.1729973803649e-05, + "loss": 0.0048, + "step": 3360 + }, + { + "epoch": 1.8766360345307713, + "grad_norm": 0.09532898949385789, + "learning_rate": 9.165333392691032e-05, + "loss": 0.0049, + "step": 3370 + }, + { + "epoch": 1.882205513784461, + "grad_norm": 0.22241436561372555, + "learning_rate": 9.157637289354648e-05, + "loss": 0.0046, + "step": 3380 + }, + { + "epoch": 1.8877749930381509, + "grad_norm": 0.09048559280760698, + "learning_rate": 9.149909129694442e-05, + "loss": 0.0051, + "step": 3390 + }, + { + "epoch": 1.8933444722918407, + "grad_norm": 0.09851219221221012, + "learning_rate": 9.142148973296279e-05, + "loss": 0.0047, + "step": 3400 + }, + { + "epoch": 1.8989139515455304, + "grad_norm": 0.11728224731469955, + "learning_rate": 9.134356879992716e-05, + "loss": 0.0046, + "step": 3410 + }, + { + "epoch": 1.9044834307992202, + "grad_norm": 0.09083216143262172, + "learning_rate": 9.126532909862554e-05, + "loss": 0.0045, + "step": 3420 + }, + { + "epoch": 1.91005291005291, + "grad_norm": 0.09298839697175458, + "learning_rate": 9.118677123230372e-05, + "loss": 0.0044, + "step": 3430 + }, + { + "epoch": 1.9156223893065998, + "grad_norm": 0.09340867022639449, + "learning_rate": 9.110789580666066e-05, + "loss": 0.0045, + "step": 3440 + }, + { + "epoch": 1.9211918685602896, + "grad_norm": 0.11685852733746609, + "learning_rate": 9.102870342984368e-05, + "loss": 0.0045, + "step": 3450 + }, + { + "epoch": 1.9267613478139793, + "grad_norm": 0.08462199664128196, + "learning_rate": 9.094919471244394e-05, + "loss": 0.0043, + "step": 3460 + }, + { + "epoch": 1.9323308270676691, + "grad_norm": 0.09235283313911731, + "learning_rate": 9.086937026749163e-05, + "loss": 0.0044, + "step": 3470 + }, + { + "epoch": 1.937900306321359, + "grad_norm": 0.09557316586337207, + "learning_rate": 9.078923071045129e-05, + "loss": 0.0044, + "step": 3480 + }, + { + "epoch": 1.9434697855750487, + "grad_norm": 0.09889989513485295, + "learning_rate": 9.0708776659217e-05, + "loss": 0.0045, + "step": 3490 + }, + { + "epoch": 1.9490392648287385, + "grad_norm": 0.1042326585110314, + "learning_rate": 9.062800873410769e-05, + "loss": 0.0044, + "step": 3500 + }, + { + "epoch": 1.9546087440824282, + "grad_norm": 0.09129522893069832, + "learning_rate": 9.054692755786236e-05, + "loss": 0.0042, + "step": 3510 + }, + { + "epoch": 1.960178223336118, + "grad_norm": 0.10382677134797022, + "learning_rate": 9.04655337556352e-05, + "loss": 0.0042, + "step": 3520 + }, + { + "epoch": 1.9657477025898078, + "grad_norm": 0.08987092514326217, + "learning_rate": 9.038382795499083e-05, + "loss": 0.0043, + "step": 3530 + }, + { + "epoch": 1.9713171818434976, + "grad_norm": 0.0849449608197563, + "learning_rate": 9.030181078589947e-05, + "loss": 0.0042, + "step": 3540 + }, + { + "epoch": 1.9768866610971874, + "grad_norm": 0.09495769793832361, + "learning_rate": 9.021948288073201e-05, + "loss": 0.0042, + "step": 3550 + }, + { + "epoch": 1.9824561403508771, + "grad_norm": 0.11307494496950236, + "learning_rate": 9.013684487425525e-05, + "loss": 0.0043, + "step": 3560 + }, + { + "epoch": 1.988025619604567, + "grad_norm": 0.10184785608835481, + "learning_rate": 9.005389740362691e-05, + "loss": 0.004, + "step": 3570 + }, + { + "epoch": 1.9935950988582567, + "grad_norm": 0.6780573883609672, + "learning_rate": 8.997064110839072e-05, + "loss": 0.0052, + "step": 3580 + }, + { + "epoch": 1.9991645781119465, + "grad_norm": 1.913891302534067, + "learning_rate": 8.988707663047156e-05, + "loss": 0.0382, + "step": 3590 + }, + { + "epoch": 2.004455583402952, + "grad_norm": 0.20491348477515167, + "learning_rate": 8.980320461417046e-05, + "loss": 0.0127, + "step": 3600 + }, + { + "epoch": 2.0100250626566414, + "grad_norm": 0.11495094855882845, + "learning_rate": 8.971902570615964e-05, + "loss": 0.0063, + "step": 3610 + }, + { + "epoch": 2.0155945419103314, + "grad_norm": 0.11775260158307438, + "learning_rate": 8.963454055547751e-05, + "loss": 0.0047, + "step": 3620 + }, + { + "epoch": 2.0211640211640214, + "grad_norm": 0.08730294248283851, + "learning_rate": 8.954974981352368e-05, + "loss": 0.0041, + "step": 3630 + }, + { + "epoch": 2.026733500417711, + "grad_norm": 0.08688062935334084, + "learning_rate": 8.946465413405398e-05, + "loss": 0.0037, + "step": 3640 + }, + { + "epoch": 2.0323029796714005, + "grad_norm": 0.07712628832792173, + "learning_rate": 8.93792541731753e-05, + "loss": 0.0034, + "step": 3650 + }, + { + "epoch": 2.0378724589250905, + "grad_norm": 0.06765602225550173, + "learning_rate": 8.929355058934068e-05, + "loss": 0.0033, + "step": 3660 + }, + { + "epoch": 2.0434419381787805, + "grad_norm": 0.07792927728833504, + "learning_rate": 8.920754404334415e-05, + "loss": 0.0031, + "step": 3670 + }, + { + "epoch": 2.04901141743247, + "grad_norm": 0.06508427058988357, + "learning_rate": 8.912123519831559e-05, + "loss": 0.003, + "step": 3680 + }, + { + "epoch": 2.0545808966861596, + "grad_norm": 0.06602478817547071, + "learning_rate": 8.903462471971573e-05, + "loss": 0.0029, + "step": 3690 + }, + { + "epoch": 2.0601503759398496, + "grad_norm": 0.07073489621873132, + "learning_rate": 8.894771327533098e-05, + "loss": 0.003, + "step": 3700 + }, + { + "epoch": 2.0657198551935396, + "grad_norm": 0.07513785677590291, + "learning_rate": 8.886050153526819e-05, + "loss": 0.003, + "step": 3710 + }, + { + "epoch": 2.071289334447229, + "grad_norm": 0.075200356674605, + "learning_rate": 8.877299017194962e-05, + "loss": 0.0029, + "step": 3720 + }, + { + "epoch": 2.0768588137009187, + "grad_norm": 0.07145305892339188, + "learning_rate": 8.868517986010768e-05, + "loss": 0.003, + "step": 3730 + }, + { + "epoch": 2.0824282929546087, + "grad_norm": 0.079246440019147, + "learning_rate": 8.859707127677972e-05, + "loss": 0.0027, + "step": 3740 + }, + { + "epoch": 2.0879977722082987, + "grad_norm": 0.07053381470878965, + "learning_rate": 8.850866510130285e-05, + "loss": 0.0027, + "step": 3750 + }, + { + "epoch": 2.0935672514619883, + "grad_norm": 0.07151886382294959, + "learning_rate": 8.841996201530869e-05, + "loss": 0.0027, + "step": 3760 + }, + { + "epoch": 2.099136730715678, + "grad_norm": 0.07398921035039588, + "learning_rate": 8.833096270271806e-05, + "loss": 0.0028, + "step": 3770 + }, + { + "epoch": 2.104706209969368, + "grad_norm": 0.06628976492907845, + "learning_rate": 8.824166784973582e-05, + "loss": 0.0027, + "step": 3780 + }, + { + "epoch": 2.110275689223058, + "grad_norm": 0.06967783368415605, + "learning_rate": 8.815207814484547e-05, + "loss": 0.0026, + "step": 3790 + }, + { + "epoch": 2.1158451684767474, + "grad_norm": 0.07367756743304106, + "learning_rate": 8.806219427880391e-05, + "loss": 0.0026, + "step": 3800 + }, + { + "epoch": 2.121414647730437, + "grad_norm": 0.07016041871073152, + "learning_rate": 8.797201694463606e-05, + "loss": 0.0026, + "step": 3810 + }, + { + "epoch": 2.126984126984127, + "grad_norm": 0.06957343148915318, + "learning_rate": 8.788154683762958e-05, + "loss": 0.0024, + "step": 3820 + }, + { + "epoch": 2.132553606237817, + "grad_norm": 0.0698865506500755, + "learning_rate": 8.779078465532942e-05, + "loss": 0.0025, + "step": 3830 + }, + { + "epoch": 2.1381230854915065, + "grad_norm": 0.0638845189601394, + "learning_rate": 8.769973109753257e-05, + "loss": 0.0025, + "step": 3840 + }, + { + "epoch": 2.143692564745196, + "grad_norm": 0.06605357071535665, + "learning_rate": 8.760838686628254e-05, + "loss": 0.0026, + "step": 3850 + }, + { + "epoch": 2.149262043998886, + "grad_norm": 0.06646085746495409, + "learning_rate": 8.751675266586401e-05, + "loss": 0.0027, + "step": 3860 + }, + { + "epoch": 2.154831523252576, + "grad_norm": 0.08537101005349107, + "learning_rate": 8.74248292027974e-05, + "loss": 0.0027, + "step": 3870 + }, + { + "epoch": 2.1604010025062657, + "grad_norm": 0.0674112060918515, + "learning_rate": 8.733261718583339e-05, + "loss": 0.0025, + "step": 3880 + }, + { + "epoch": 2.165970481759955, + "grad_norm": 0.06660804464627262, + "learning_rate": 8.724011732594749e-05, + "loss": 0.0026, + "step": 3890 + }, + { + "epoch": 2.171539961013645, + "grad_norm": 0.06849541239000093, + "learning_rate": 8.714733033633454e-05, + "loss": 0.0026, + "step": 3900 + }, + { + "epoch": 2.177109440267335, + "grad_norm": 0.0720897091900461, + "learning_rate": 8.705425693240319e-05, + "loss": 0.0026, + "step": 3910 + }, + { + "epoch": 2.1826789195210248, + "grad_norm": 0.06824332799424324, + "learning_rate": 8.696089783177049e-05, + "loss": 0.0025, + "step": 3920 + }, + { + "epoch": 2.1882483987747143, + "grad_norm": 0.06717944971109528, + "learning_rate": 8.686725375425617e-05, + "loss": 0.0025, + "step": 3930 + }, + { + "epoch": 2.1938178780284043, + "grad_norm": 0.06136685454453069, + "learning_rate": 8.677332542187727e-05, + "loss": 0.0025, + "step": 3940 + }, + { + "epoch": 2.1993873572820943, + "grad_norm": 0.07922796621477682, + "learning_rate": 8.667911355884248e-05, + "loss": 0.0025, + "step": 3950 + }, + { + "epoch": 2.204956836535784, + "grad_norm": 0.06748085462192983, + "learning_rate": 8.658461889154654e-05, + "loss": 0.0026, + "step": 3960 + }, + { + "epoch": 2.2105263157894735, + "grad_norm": 0.07033503509897161, + "learning_rate": 8.648984214856474e-05, + "loss": 0.0026, + "step": 3970 + }, + { + "epoch": 2.2160957950431635, + "grad_norm": 0.0693408764546056, + "learning_rate": 8.639478406064722e-05, + "loss": 0.0025, + "step": 3980 + }, + { + "epoch": 2.2216652742968535, + "grad_norm": 0.06290458960875848, + "learning_rate": 8.629944536071328e-05, + "loss": 0.0026, + "step": 3990 + }, + { + "epoch": 2.227234753550543, + "grad_norm": 0.06522173944000591, + "learning_rate": 8.620382678384591e-05, + "loss": 0.0025, + "step": 4000 + }, + { + "epoch": 2.2328042328042326, + "grad_norm": 0.0692135716494379, + "learning_rate": 8.610792906728593e-05, + "loss": 0.0025, + "step": 4010 + }, + { + "epoch": 2.2383737120579226, + "grad_norm": 0.06218397041619794, + "learning_rate": 8.601175295042643e-05, + "loss": 0.0025, + "step": 4020 + }, + { + "epoch": 2.2439431913116126, + "grad_norm": 0.07425856965054406, + "learning_rate": 8.591529917480705e-05, + "loss": 0.0026, + "step": 4030 + }, + { + "epoch": 2.249512670565302, + "grad_norm": 0.07106967477843956, + "learning_rate": 8.581856848410816e-05, + "loss": 0.0024, + "step": 4040 + }, + { + "epoch": 2.2550821498189917, + "grad_norm": 0.07511095612847542, + "learning_rate": 8.572156162414533e-05, + "loss": 0.0025, + "step": 4050 + }, + { + "epoch": 2.2606516290726817, + "grad_norm": 0.06682058463634225, + "learning_rate": 8.562427934286334e-05, + "loss": 0.0023, + "step": 4060 + }, + { + "epoch": 2.2662211083263717, + "grad_norm": 0.07894516840214212, + "learning_rate": 8.55267223903306e-05, + "loss": 0.0025, + "step": 4070 + }, + { + "epoch": 2.2717905875800612, + "grad_norm": 0.07442523909487508, + "learning_rate": 8.542889151873328e-05, + "loss": 0.0023, + "step": 4080 + }, + { + "epoch": 2.277360066833751, + "grad_norm": 0.07047966962425675, + "learning_rate": 8.533078748236951e-05, + "loss": 0.0024, + "step": 4090 + }, + { + "epoch": 2.282929546087441, + "grad_norm": 0.0670309181354895, + "learning_rate": 8.52324110376436e-05, + "loss": 0.0024, + "step": 4100 + }, + { + "epoch": 2.288499025341131, + "grad_norm": 0.05841158200173529, + "learning_rate": 8.513376294306022e-05, + "loss": 0.0025, + "step": 4110 + }, + { + "epoch": 2.2940685045948204, + "grad_norm": 0.07235767865059842, + "learning_rate": 8.503484395921847e-05, + "loss": 0.0025, + "step": 4120 + }, + { + "epoch": 2.29963798384851, + "grad_norm": 0.06368681101578703, + "learning_rate": 8.493565484880607e-05, + "loss": 0.0025, + "step": 4130 + }, + { + "epoch": 2.3052074631022, + "grad_norm": 0.06306798916456337, + "learning_rate": 8.483619637659352e-05, + "loss": 0.0023, + "step": 4140 + }, + { + "epoch": 2.31077694235589, + "grad_norm": 0.063978959789188, + "learning_rate": 8.473646930942811e-05, + "loss": 0.0023, + "step": 4150 + }, + { + "epoch": 2.3163464216095795, + "grad_norm": 0.08113440594399092, + "learning_rate": 8.463647441622811e-05, + "loss": 0.0024, + "step": 4160 + }, + { + "epoch": 2.321915900863269, + "grad_norm": 0.06560709654743707, + "learning_rate": 8.453621246797674e-05, + "loss": 0.0023, + "step": 4170 + }, + { + "epoch": 2.327485380116959, + "grad_norm": 0.05878754983194171, + "learning_rate": 8.443568423771632e-05, + "loss": 0.0025, + "step": 4180 + }, + { + "epoch": 2.333054859370649, + "grad_norm": 0.06519491241511483, + "learning_rate": 8.433489050054222e-05, + "loss": 0.0024, + "step": 4190 + }, + { + "epoch": 2.3386243386243386, + "grad_norm": 0.06265875286229199, + "learning_rate": 8.423383203359696e-05, + "loss": 0.0023, + "step": 4200 + }, + { + "epoch": 2.344193817878028, + "grad_norm": 0.07488402161472248, + "learning_rate": 8.413250961606417e-05, + "loss": 0.0023, + "step": 4210 + }, + { + "epoch": 2.349763297131718, + "grad_norm": 0.06477354979976067, + "learning_rate": 8.403092402916261e-05, + "loss": 0.0023, + "step": 4220 + }, + { + "epoch": 2.355332776385408, + "grad_norm": 0.06498384745452408, + "learning_rate": 8.392907605614013e-05, + "loss": 0.0024, + "step": 4230 + }, + { + "epoch": 2.3609022556390977, + "grad_norm": 0.06078330248348828, + "learning_rate": 8.382696648226764e-05, + "loss": 0.0024, + "step": 4240 + }, + { + "epoch": 2.3664717348927873, + "grad_norm": 0.0687320440446581, + "learning_rate": 8.372459609483303e-05, + "loss": 0.0025, + "step": 4250 + }, + { + "epoch": 2.3720412141464773, + "grad_norm": 0.06587983128890958, + "learning_rate": 8.362196568313516e-05, + "loss": 0.0023, + "step": 4260 + }, + { + "epoch": 2.3776106934001673, + "grad_norm": 0.06659212111823104, + "learning_rate": 8.351907603847773e-05, + "loss": 0.0023, + "step": 4270 + }, + { + "epoch": 2.383180172653857, + "grad_norm": 0.06922434253253595, + "learning_rate": 8.341592795416312e-05, + "loss": 0.0024, + "step": 4280 + }, + { + "epoch": 2.3887496519075464, + "grad_norm": 0.06994327293771829, + "learning_rate": 8.331252222548645e-05, + "loss": 0.0024, + "step": 4290 + }, + { + "epoch": 2.3943191311612364, + "grad_norm": 0.06643409266880786, + "learning_rate": 8.320885964972922e-05, + "loss": 0.0025, + "step": 4300 + }, + { + "epoch": 2.3998886104149264, + "grad_norm": 0.0811114126700153, + "learning_rate": 8.310494102615336e-05, + "loss": 0.0025, + "step": 4310 + }, + { + "epoch": 2.405458089668616, + "grad_norm": 0.07211869955643185, + "learning_rate": 8.300076715599496e-05, + "loss": 0.0024, + "step": 4320 + }, + { + "epoch": 2.4110275689223055, + "grad_norm": 0.06687148486771088, + "learning_rate": 8.289633884245811e-05, + "loss": 0.0024, + "step": 4330 + }, + { + "epoch": 2.4165970481759955, + "grad_norm": 0.06431135751569382, + "learning_rate": 8.279165689070875e-05, + "loss": 0.0023, + "step": 4340 + }, + { + "epoch": 2.4221665274296855, + "grad_norm": 0.07773019218992226, + "learning_rate": 8.268672210786839e-05, + "loss": 0.0023, + "step": 4350 + }, + { + "epoch": 2.427736006683375, + "grad_norm": 0.07191534653352975, + "learning_rate": 8.258153530300792e-05, + "loss": 0.0025, + "step": 4360 + }, + { + "epoch": 2.4333054859370646, + "grad_norm": 0.07424856314720017, + "learning_rate": 8.247609728714142e-05, + "loss": 0.0024, + "step": 4370 + }, + { + "epoch": 2.4388749651907546, + "grad_norm": 0.06644817422012073, + "learning_rate": 8.237040887321983e-05, + "loss": 0.0024, + "step": 4380 + }, + { + "epoch": 2.4444444444444446, + "grad_norm": 0.07935382842757754, + "learning_rate": 8.226447087612472e-05, + "loss": 0.0024, + "step": 4390 + }, + { + "epoch": 2.450013923698134, + "grad_norm": 0.07344510881235311, + "learning_rate": 8.215828411266202e-05, + "loss": 0.0023, + "step": 4400 + }, + { + "epoch": 2.4555834029518238, + "grad_norm": 0.06437241925673821, + "learning_rate": 8.205184940155567e-05, + "loss": 0.0023, + "step": 4410 + }, + { + "epoch": 2.4611528822055138, + "grad_norm": 0.06191846693537428, + "learning_rate": 8.19451675634414e-05, + "loss": 0.0022, + "step": 4420 + }, + { + "epoch": 2.4667223614592038, + "grad_norm": 0.0743929168561731, + "learning_rate": 8.183823942086026e-05, + "loss": 0.0022, + "step": 4430 + }, + { + "epoch": 2.4722918407128933, + "grad_norm": 0.07877045476260705, + "learning_rate": 8.173106579825246e-05, + "loss": 0.0024, + "step": 4440 + }, + { + "epoch": 2.477861319966583, + "grad_norm": 0.06063611479460526, + "learning_rate": 8.162364752195084e-05, + "loss": 0.0023, + "step": 4450 + }, + { + "epoch": 2.483430799220273, + "grad_norm": 0.06803918605199016, + "learning_rate": 8.151598542017463e-05, + "loss": 0.0021, + "step": 4460 + }, + { + "epoch": 2.489000278473963, + "grad_norm": 0.06684605150130163, + "learning_rate": 8.140808032302296e-05, + "loss": 0.0022, + "step": 4470 + }, + { + "epoch": 2.4945697577276524, + "grad_norm": 0.07187232516181655, + "learning_rate": 8.129993306246855e-05, + "loss": 0.0022, + "step": 4480 + }, + { + "epoch": 2.500139236981342, + "grad_norm": 0.06494266366065195, + "learning_rate": 8.119154447235123e-05, + "loss": 0.0022, + "step": 4490 + }, + { + "epoch": 2.505708716235032, + "grad_norm": 0.06464060504434475, + "learning_rate": 8.108291538837151e-05, + "loss": 0.0022, + "step": 4500 + }, + { + "epoch": 2.511278195488722, + "grad_norm": 0.06569241269740814, + "learning_rate": 8.097404664808428e-05, + "loss": 0.0022, + "step": 4510 + }, + { + "epoch": 2.5168476747424116, + "grad_norm": 0.055242094122246455, + "learning_rate": 8.086493909089208e-05, + "loss": 0.0021, + "step": 4520 + }, + { + "epoch": 2.522417153996101, + "grad_norm": 0.06777582514746222, + "learning_rate": 8.075559355803887e-05, + "loss": 0.0023, + "step": 4530 + }, + { + "epoch": 2.527986633249791, + "grad_norm": 0.06478881950733577, + "learning_rate": 8.064601089260344e-05, + "loss": 0.0022, + "step": 4540 + }, + { + "epoch": 2.533556112503481, + "grad_norm": 0.06924218417878182, + "learning_rate": 8.053619193949293e-05, + "loss": 0.0022, + "step": 4550 + }, + { + "epoch": 2.5391255917571707, + "grad_norm": 0.06574296301329537, + "learning_rate": 8.042613754543631e-05, + "loss": 0.0022, + "step": 4560 + }, + { + "epoch": 2.5446950710108602, + "grad_norm": 0.0652453535875289, + "learning_rate": 8.031584855897782e-05, + "loss": 0.0022, + "step": 4570 + }, + { + "epoch": 2.5502645502645502, + "grad_norm": 0.0669710260970334, + "learning_rate": 8.020532583047052e-05, + "loss": 0.002, + "step": 4580 + }, + { + "epoch": 2.5558340295182402, + "grad_norm": 0.06195618827884842, + "learning_rate": 8.009457021206964e-05, + "loss": 0.0022, + "step": 4590 + }, + { + "epoch": 2.56140350877193, + "grad_norm": 0.05421163706378184, + "learning_rate": 7.998358255772604e-05, + "loss": 0.0022, + "step": 4600 + }, + { + "epoch": 2.5669729880256194, + "grad_norm": 0.07244346720460094, + "learning_rate": 7.987236372317964e-05, + "loss": 0.0021, + "step": 4610 + }, + { + "epoch": 2.5725424672793094, + "grad_norm": 0.07122830520202149, + "learning_rate": 7.976091456595278e-05, + "loss": 0.0022, + "step": 4620 + }, + { + "epoch": 2.5781119465329994, + "grad_norm": 0.058368117308028544, + "learning_rate": 7.964923594534372e-05, + "loss": 0.002, + "step": 4630 + }, + { + "epoch": 2.583681425786689, + "grad_norm": 0.06617475020782944, + "learning_rate": 7.953732872241984e-05, + "loss": 0.0022, + "step": 4640 + }, + { + "epoch": 2.5892509050403785, + "grad_norm": 0.05757509551408226, + "learning_rate": 7.942519376001116e-05, + "loss": 0.0021, + "step": 4650 + }, + { + "epoch": 2.5948203842940685, + "grad_norm": 0.06085258271174403, + "learning_rate": 7.93128319227036e-05, + "loss": 0.0021, + "step": 4660 + }, + { + "epoch": 2.6003898635477585, + "grad_norm": 0.0653949688851885, + "learning_rate": 7.920024407683234e-05, + "loss": 0.0021, + "step": 4670 + }, + { + "epoch": 2.605959342801448, + "grad_norm": 0.06266065978713677, + "learning_rate": 7.908743109047512e-05, + "loss": 0.0021, + "step": 4680 + }, + { + "epoch": 2.6115288220551376, + "grad_norm": 0.06607862312823722, + "learning_rate": 7.89743938334456e-05, + "loss": 0.0021, + "step": 4690 + }, + { + "epoch": 2.6170983013088276, + "grad_norm": 0.05456069896485288, + "learning_rate": 7.886113317728658e-05, + "loss": 0.002, + "step": 4700 + }, + { + "epoch": 2.6226677805625176, + "grad_norm": 0.07060726920422762, + "learning_rate": 7.874764999526334e-05, + "loss": 0.0021, + "step": 4710 + }, + { + "epoch": 2.628237259816207, + "grad_norm": 0.0628459007652655, + "learning_rate": 7.863394516235687e-05, + "loss": 0.0021, + "step": 4720 + }, + { + "epoch": 2.6338067390698967, + "grad_norm": 0.07184850567413902, + "learning_rate": 7.852001955525717e-05, + "loss": 0.0021, + "step": 4730 + }, + { + "epoch": 2.6393762183235867, + "grad_norm": 0.05931313997400148, + "learning_rate": 7.840587405235642e-05, + "loss": 0.0022, + "step": 4740 + }, + { + "epoch": 2.6449456975772767, + "grad_norm": 0.06266716450193778, + "learning_rate": 7.829150953374228e-05, + "loss": 0.0021, + "step": 4750 + }, + { + "epoch": 2.6505151768309663, + "grad_norm": 0.06018038870616263, + "learning_rate": 7.817692688119103e-05, + "loss": 0.0022, + "step": 4760 + }, + { + "epoch": 2.656084656084656, + "grad_norm": 0.06516106860781504, + "learning_rate": 7.806212697816085e-05, + "loss": 0.0022, + "step": 4770 + }, + { + "epoch": 2.661654135338346, + "grad_norm": 0.07560214888027027, + "learning_rate": 7.794711070978496e-05, + "loss": 0.0024, + "step": 4780 + }, + { + "epoch": 2.667223614592036, + "grad_norm": 0.058287088159306756, + "learning_rate": 7.78318789628648e-05, + "loss": 0.002, + "step": 4790 + }, + { + "epoch": 2.6727930938457254, + "grad_norm": 0.05041812827730329, + "learning_rate": 7.771643262586319e-05, + "loss": 0.0023, + "step": 4800 + }, + { + "epoch": 2.678362573099415, + "grad_norm": 0.06769617798698746, + "learning_rate": 7.76007725888975e-05, + "loss": 0.0023, + "step": 4810 + }, + { + "epoch": 2.683932052353105, + "grad_norm": 0.07237732703780383, + "learning_rate": 7.748489974373277e-05, + "loss": 0.0024, + "step": 4820 + }, + { + "epoch": 2.689501531606795, + "grad_norm": 0.06774331956099881, + "learning_rate": 7.736881498377488e-05, + "loss": 0.0024, + "step": 4830 + }, + { + "epoch": 2.6950710108604845, + "grad_norm": 0.08272693895736795, + "learning_rate": 7.725251920406353e-05, + "loss": 0.0023, + "step": 4840 + }, + { + "epoch": 2.700640490114174, + "grad_norm": 0.0647479021182408, + "learning_rate": 7.713601330126551e-05, + "loss": 0.0024, + "step": 4850 + }, + { + "epoch": 2.706209969367864, + "grad_norm": 0.06747122196352953, + "learning_rate": 7.70192981736677e-05, + "loss": 0.0023, + "step": 4860 + }, + { + "epoch": 2.711779448621554, + "grad_norm": 0.06660918185561457, + "learning_rate": 7.69023747211701e-05, + "loss": 0.0021, + "step": 4870 + }, + { + "epoch": 2.7173489278752436, + "grad_norm": 0.06014673632175607, + "learning_rate": 7.678524384527897e-05, + "loss": 0.0022, + "step": 4880 + }, + { + "epoch": 2.722918407128933, + "grad_norm": 0.060474215458977645, + "learning_rate": 7.666790644909988e-05, + "loss": 0.0021, + "step": 4890 + }, + { + "epoch": 2.728487886382623, + "grad_norm": 0.05475551997626191, + "learning_rate": 7.655036343733068e-05, + "loss": 0.0021, + "step": 4900 + }, + { + "epoch": 2.734057365636313, + "grad_norm": 0.05525281945014114, + "learning_rate": 7.643261571625457e-05, + "loss": 0.0021, + "step": 4910 + }, + { + "epoch": 2.7396268448900027, + "grad_norm": 0.06451171594706764, + "learning_rate": 7.631466419373312e-05, + "loss": 0.0021, + "step": 4920 + }, + { + "epoch": 2.7451963241436923, + "grad_norm": 0.07105940351069889, + "learning_rate": 7.619650977919923e-05, + "loss": 0.0022, + "step": 4930 + }, + { + "epoch": 2.7507658033973823, + "grad_norm": 0.0748324693501243, + "learning_rate": 7.607815338365018e-05, + "loss": 0.002, + "step": 4940 + }, + { + "epoch": 2.7563352826510723, + "grad_norm": 0.0588056102878865, + "learning_rate": 7.595959591964053e-05, + "loss": 0.0021, + "step": 4950 + }, + { + "epoch": 2.761904761904762, + "grad_norm": 0.05974979070561233, + "learning_rate": 7.584083830127515e-05, + "loss": 0.0021, + "step": 4960 + }, + { + "epoch": 2.7674742411584514, + "grad_norm": 0.05841681850185086, + "learning_rate": 7.572188144420214e-05, + "loss": 0.0021, + "step": 4970 + }, + { + "epoch": 2.7730437204121414, + "grad_norm": 0.06753208138557094, + "learning_rate": 7.560272626560575e-05, + "loss": 0.0021, + "step": 4980 + }, + { + "epoch": 2.7786131996658314, + "grad_norm": 0.06433255988165418, + "learning_rate": 7.548337368419937e-05, + "loss": 0.002, + "step": 4990 + }, + { + "epoch": 2.784182678919521, + "grad_norm": 0.06973811409366241, + "learning_rate": 7.536382462021843e-05, + "loss": 0.0022, + "step": 5000 + }, + { + "epoch": 2.7897521581732105, + "grad_norm": 0.06230697352063296, + "learning_rate": 7.52440799954132e-05, + "loss": 0.002, + "step": 5010 + }, + { + "epoch": 2.7953216374269005, + "grad_norm": 0.05784103057124063, + "learning_rate": 7.512414073304183e-05, + "loss": 0.0019, + "step": 5020 + }, + { + "epoch": 2.8008911166805905, + "grad_norm": 0.055170016243759425, + "learning_rate": 7.500400775786319e-05, + "loss": 0.0021, + "step": 5030 + }, + { + "epoch": 2.80646059593428, + "grad_norm": 0.051739538811377245, + "learning_rate": 7.488368199612965e-05, + "loss": 0.0019, + "step": 5040 + }, + { + "epoch": 2.8120300751879697, + "grad_norm": 0.0545065555627121, + "learning_rate": 7.476316437558006e-05, + "loss": 0.0021, + "step": 5050 + }, + { + "epoch": 2.8175995544416597, + "grad_norm": 0.051858173462886396, + "learning_rate": 7.464245582543254e-05, + "loss": 0.002, + "step": 5060 + }, + { + "epoch": 2.8231690336953497, + "grad_norm": 0.05625922810310046, + "learning_rate": 7.452155727637731e-05, + "loss": 0.0019, + "step": 5070 + }, + { + "epoch": 2.828738512949039, + "grad_norm": 0.06451749546980787, + "learning_rate": 7.440046966056954e-05, + "loss": 0.0019, + "step": 5080 + }, + { + "epoch": 2.8343079922027288, + "grad_norm": 0.0592788295080246, + "learning_rate": 7.42791939116221e-05, + "loss": 0.0018, + "step": 5090 + }, + { + "epoch": 2.839877471456419, + "grad_norm": 0.060714792093629934, + "learning_rate": 7.415773096459846e-05, + "loss": 0.0019, + "step": 5100 + }, + { + "epoch": 2.845446950710109, + "grad_norm": 0.054065946398541405, + "learning_rate": 7.40360817560054e-05, + "loss": 0.0018, + "step": 5110 + }, + { + "epoch": 2.8510164299637983, + "grad_norm": 0.06021220447231555, + "learning_rate": 7.391424722378589e-05, + "loss": 0.0019, + "step": 5120 + }, + { + "epoch": 2.856585909217488, + "grad_norm": 0.05567546706138025, + "learning_rate": 7.379222830731164e-05, + "loss": 0.002, + "step": 5130 + }, + { + "epoch": 2.862155388471178, + "grad_norm": 0.056617114000235835, + "learning_rate": 7.367002594737614e-05, + "loss": 0.002, + "step": 5140 + }, + { + "epoch": 2.867724867724868, + "grad_norm": 0.05902717684116007, + "learning_rate": 7.354764108618722e-05, + "loss": 0.0019, + "step": 5150 + }, + { + "epoch": 2.8732943469785575, + "grad_norm": 0.05468033114324629, + "learning_rate": 7.342507466735984e-05, + "loss": 0.002, + "step": 5160 + }, + { + "epoch": 2.878863826232247, + "grad_norm": 0.05355173047966122, + "learning_rate": 7.330232763590882e-05, + "loss": 0.002, + "step": 5170 + }, + { + "epoch": 2.884433305485937, + "grad_norm": 0.047606675945423234, + "learning_rate": 7.317940093824151e-05, + "loss": 0.0018, + "step": 5180 + }, + { + "epoch": 2.890002784739627, + "grad_norm": 0.05710987607858959, + "learning_rate": 7.305629552215055e-05, + "loss": 0.002, + "step": 5190 + }, + { + "epoch": 2.8955722639933166, + "grad_norm": 0.061919641274551386, + "learning_rate": 7.293301233680656e-05, + "loss": 0.0019, + "step": 5200 + }, + { + "epoch": 2.901141743247006, + "grad_norm": 0.06034082866829357, + "learning_rate": 7.280955233275078e-05, + "loss": 0.0019, + "step": 5210 + }, + { + "epoch": 2.906711222500696, + "grad_norm": 0.05055117819606763, + "learning_rate": 7.268591646188776e-05, + "loss": 0.0019, + "step": 5220 + }, + { + "epoch": 2.912280701754386, + "grad_norm": 0.07536575724606918, + "learning_rate": 7.256210567747801e-05, + "loss": 0.0019, + "step": 5230 + }, + { + "epoch": 2.9178501810080757, + "grad_norm": 0.05837113985627112, + "learning_rate": 7.24381209341307e-05, + "loss": 0.0019, + "step": 5240 + }, + { + "epoch": 2.9234196602617653, + "grad_norm": 0.0834784993335112, + "learning_rate": 7.231396318779623e-05, + "loss": 0.0018, + "step": 5250 + }, + { + "epoch": 2.9289891395154553, + "grad_norm": 0.05474102004179984, + "learning_rate": 7.218963339575892e-05, + "loss": 0.0019, + "step": 5260 + }, + { + "epoch": 2.9345586187691453, + "grad_norm": 0.052680493771874885, + "learning_rate": 7.206513251662957e-05, + "loss": 0.0018, + "step": 5270 + }, + { + "epoch": 2.940128098022835, + "grad_norm": 0.057180281513574904, + "learning_rate": 7.194046151033813e-05, + "loss": 0.0019, + "step": 5280 + }, + { + "epoch": 2.9456975772765244, + "grad_norm": 0.05234359452879529, + "learning_rate": 7.181562133812624e-05, + "loss": 0.0018, + "step": 5290 + }, + { + "epoch": 2.9512670565302144, + "grad_norm": 0.05463305761815249, + "learning_rate": 7.169061296253987e-05, + "loss": 0.0018, + "step": 5300 + }, + { + "epoch": 2.9568365357839044, + "grad_norm": 0.06861187936883137, + "learning_rate": 7.156543734742186e-05, + "loss": 0.0019, + "step": 5310 + }, + { + "epoch": 2.962406015037594, + "grad_norm": 0.048975954225027464, + "learning_rate": 7.144009545790454e-05, + "loss": 0.0017, + "step": 5320 + }, + { + "epoch": 2.9679754942912835, + "grad_norm": 0.05348185471944446, + "learning_rate": 7.13145882604022e-05, + "loss": 0.0018, + "step": 5330 + }, + { + "epoch": 2.9735449735449735, + "grad_norm": 0.060040519890833914, + "learning_rate": 7.118891672260372e-05, + "loss": 0.0017, + "step": 5340 + }, + { + "epoch": 2.9791144527986635, + "grad_norm": 0.04694679144483152, + "learning_rate": 7.10630818134651e-05, + "loss": 0.0017, + "step": 5350 + }, + { + "epoch": 2.984683932052353, + "grad_norm": 0.06715588212175277, + "learning_rate": 7.093708450320191e-05, + "loss": 0.0018, + "step": 5360 + }, + { + "epoch": 2.9902534113060426, + "grad_norm": 0.05154669242107701, + "learning_rate": 7.081092576328195e-05, + "loss": 0.0018, + "step": 5370 + }, + { + "epoch": 2.9958228905597326, + "grad_norm": 0.06543458650756455, + "learning_rate": 7.068460656641761e-05, + "loss": 0.0018, + "step": 5380 + }, + { + "epoch": 3.0011138958507377, + "grad_norm": 0.05641584896255111, + "learning_rate": 7.055812788655849e-05, + "loss": 0.0015, + "step": 5390 + }, + { + "epoch": 3.0066833751044277, + "grad_norm": 0.04151708786487322, + "learning_rate": 7.043149069888379e-05, + "loss": 0.0011, + "step": 5400 + }, + { + "epoch": 3.0122528543581177, + "grad_norm": 0.04449099061436636, + "learning_rate": 7.03046959797949e-05, + "loss": 0.0011, + "step": 5410 + }, + { + "epoch": 3.0178223336118073, + "grad_norm": 0.04417184911473767, + "learning_rate": 7.017774470690774e-05, + "loss": 0.0011, + "step": 5420 + }, + { + "epoch": 3.023391812865497, + "grad_norm": 0.048344448951397646, + "learning_rate": 7.005063785904538e-05, + "loss": 0.0009, + "step": 5430 + }, + { + "epoch": 3.028961292119187, + "grad_norm": 0.1412351691134576, + "learning_rate": 6.992337641623034e-05, + "loss": 0.0009, + "step": 5440 + }, + { + "epoch": 3.034530771372877, + "grad_norm": 0.03972914676477569, + "learning_rate": 6.979596135967712e-05, + "loss": 0.0012, + "step": 5450 + }, + { + "epoch": 3.0401002506265664, + "grad_norm": 0.04289250130205762, + "learning_rate": 6.966839367178464e-05, + "loss": 0.001, + "step": 5460 + }, + { + "epoch": 3.045669729880256, + "grad_norm": 0.04564933326024313, + "learning_rate": 6.95406743361286e-05, + "loss": 0.0009, + "step": 5470 + }, + { + "epoch": 3.051239209133946, + "grad_norm": 0.038123777319650005, + "learning_rate": 6.9412804337454e-05, + "loss": 0.0009, + "step": 5480 + }, + { + "epoch": 3.056808688387636, + "grad_norm": 0.04926106513035876, + "learning_rate": 6.928478466166744e-05, + "loss": 0.0009, + "step": 5490 + }, + { + "epoch": 3.0623781676413255, + "grad_norm": 0.047995647182890684, + "learning_rate": 6.915661629582954e-05, + "loss": 0.0009, + "step": 5500 + }, + { + "epoch": 3.067947646895015, + "grad_norm": 0.040171572840021576, + "learning_rate": 6.902830022814743e-05, + "loss": 0.001, + "step": 5510 + }, + { + "epoch": 3.073517126148705, + "grad_norm": 0.055509936260737816, + "learning_rate": 6.8899837447967e-05, + "loss": 0.0011, + "step": 5520 + }, + { + "epoch": 3.079086605402395, + "grad_norm": 0.03523906409046715, + "learning_rate": 6.877122894576534e-05, + "loss": 0.001, + "step": 5530 + }, + { + "epoch": 3.0846560846560847, + "grad_norm": 0.04580610959850388, + "learning_rate": 6.864247571314306e-05, + "loss": 0.001, + "step": 5540 + }, + { + "epoch": 3.090225563909774, + "grad_norm": 0.051886173804192096, + "learning_rate": 6.851357874281674e-05, + "loss": 0.0009, + "step": 5550 + }, + { + "epoch": 3.095795043163464, + "grad_norm": 0.04714143061177957, + "learning_rate": 6.838453902861116e-05, + "loss": 0.0009, + "step": 5560 + }, + { + "epoch": 3.101364522417154, + "grad_norm": 0.049297634678337636, + "learning_rate": 6.825535756545171e-05, + "loss": 0.001, + "step": 5570 + }, + { + "epoch": 3.1069340016708438, + "grad_norm": 0.045642039727666096, + "learning_rate": 6.812603534935668e-05, + "loss": 0.0009, + "step": 5580 + }, + { + "epoch": 3.1125034809245333, + "grad_norm": 0.041516250533900546, + "learning_rate": 6.799657337742959e-05, + "loss": 0.0009, + "step": 5590 + }, + { + "epoch": 3.1180729601782233, + "grad_norm": 0.050389258244108, + "learning_rate": 6.786697264785155e-05, + "loss": 0.0009, + "step": 5600 + }, + { + "epoch": 3.1236424394319133, + "grad_norm": 0.04000069942861406, + "learning_rate": 6.773723415987349e-05, + "loss": 0.0009, + "step": 5610 + }, + { + "epoch": 3.129211918685603, + "grad_norm": 0.04460251124520155, + "learning_rate": 6.760735891380847e-05, + "loss": 0.0008, + "step": 5620 + }, + { + "epoch": 3.1347813979392924, + "grad_norm": 0.050564236575500694, + "learning_rate": 6.747734791102404e-05, + "loss": 0.0009, + "step": 5630 + }, + { + "epoch": 3.1403508771929824, + "grad_norm": 0.03547412384750168, + "learning_rate": 6.734720215393442e-05, + "loss": 0.0009, + "step": 5640 + }, + { + "epoch": 3.1459203564466725, + "grad_norm": 0.044443351345187426, + "learning_rate": 6.721692264599283e-05, + "loss": 0.001, + "step": 5650 + }, + { + "epoch": 3.151489835700362, + "grad_norm": 0.05009658237187927, + "learning_rate": 6.708651039168377e-05, + "loss": 0.0009, + "step": 5660 + }, + { + "epoch": 3.1570593149540516, + "grad_norm": 0.04022813015926799, + "learning_rate": 6.695596639651518e-05, + "loss": 0.0009, + "step": 5670 + }, + { + "epoch": 3.1626287942077416, + "grad_norm": 0.041082780186479186, + "learning_rate": 6.682529166701079e-05, + "loss": 0.0008, + "step": 5680 + }, + { + "epoch": 3.1681982734614316, + "grad_norm": 0.0407370680395219, + "learning_rate": 6.669448721070233e-05, + "loss": 0.0009, + "step": 5690 + }, + { + "epoch": 3.173767752715121, + "grad_norm": 0.04052598689224245, + "learning_rate": 6.656355403612176e-05, + "loss": 0.0009, + "step": 5700 + }, + { + "epoch": 3.1793372319688107, + "grad_norm": 0.046204651640613846, + "learning_rate": 6.643249315279345e-05, + "loss": 0.0009, + "step": 5710 + }, + { + "epoch": 3.1849067112225007, + "grad_norm": 0.03766862316669004, + "learning_rate": 6.630130557122643e-05, + "loss": 0.0009, + "step": 5720 + }, + { + "epoch": 3.1904761904761907, + "grad_norm": 0.04183293682495362, + "learning_rate": 6.616999230290666e-05, + "loss": 0.0009, + "step": 5730 + }, + { + "epoch": 3.1960456697298802, + "grad_norm": 0.0390744908206799, + "learning_rate": 6.603855436028913e-05, + "loss": 0.0009, + "step": 5740 + }, + { + "epoch": 3.20161514898357, + "grad_norm": 0.04575970052068148, + "learning_rate": 6.59069927567901e-05, + "loss": 0.0009, + "step": 5750 + }, + { + "epoch": 3.20718462823726, + "grad_norm": 0.05425010874525675, + "learning_rate": 6.577530850677926e-05, + "loss": 0.001, + "step": 5760 + }, + { + "epoch": 3.21275410749095, + "grad_norm": 0.04582600242314767, + "learning_rate": 6.5643502625572e-05, + "loss": 0.001, + "step": 5770 + }, + { + "epoch": 3.2183235867446394, + "grad_norm": 0.044630932807082725, + "learning_rate": 6.551157612942148e-05, + "loss": 0.0009, + "step": 5780 + }, + { + "epoch": 3.223893065998329, + "grad_norm": 0.037263314461341894, + "learning_rate": 6.537953003551077e-05, + "loss": 0.0009, + "step": 5790 + }, + { + "epoch": 3.229462545252019, + "grad_norm": 0.0331112589072301, + "learning_rate": 6.524736536194517e-05, + "loss": 0.0009, + "step": 5800 + }, + { + "epoch": 3.235032024505709, + "grad_norm": 0.0988917417195699, + "learning_rate": 6.511508312774419e-05, + "loss": 0.001, + "step": 5810 + }, + { + "epoch": 3.2406015037593985, + "grad_norm": 0.05299822261657893, + "learning_rate": 6.498268435283376e-05, + "loss": 0.0009, + "step": 5820 + }, + { + "epoch": 3.246170983013088, + "grad_norm": 0.0413702849720354, + "learning_rate": 6.485017005803839e-05, + "loss": 0.001, + "step": 5830 + }, + { + "epoch": 3.251740462266778, + "grad_norm": 0.045399718037569606, + "learning_rate": 6.471754126507326e-05, + "loss": 0.0009, + "step": 5840 + }, + { + "epoch": 3.257309941520468, + "grad_norm": 0.03916394078603809, + "learning_rate": 6.458479899653635e-05, + "loss": 0.0011, + "step": 5850 + }, + { + "epoch": 3.2628794207741576, + "grad_norm": 0.04693906298858904, + "learning_rate": 6.44519442759006e-05, + "loss": 0.0014, + "step": 5860 + }, + { + "epoch": 3.268448900027847, + "grad_norm": 0.06579717480932165, + "learning_rate": 6.431897812750594e-05, + "loss": 0.0012, + "step": 5870 + }, + { + "epoch": 3.274018379281537, + "grad_norm": 0.044831739432227374, + "learning_rate": 6.418590157655142e-05, + "loss": 0.0011, + "step": 5880 + }, + { + "epoch": 3.279587858535227, + "grad_norm": 0.060975308078475204, + "learning_rate": 6.40527156490874e-05, + "loss": 0.001, + "step": 5890 + }, + { + "epoch": 3.2851573377889167, + "grad_norm": 0.04126107886999729, + "learning_rate": 6.391942137200745e-05, + "loss": 0.001, + "step": 5900 + }, + { + "epoch": 3.2907268170426063, + "grad_norm": 0.04719841074758441, + "learning_rate": 6.378601977304062e-05, + "loss": 0.001, + "step": 5910 + }, + { + "epoch": 3.2962962962962963, + "grad_norm": 0.0416075388272423, + "learning_rate": 6.365251188074339e-05, + "loss": 0.0009, + "step": 5920 + }, + { + "epoch": 3.3018657755499863, + "grad_norm": 0.09489837294527975, + "learning_rate": 6.351889872449182e-05, + "loss": 0.001, + "step": 5930 + }, + { + "epoch": 3.307435254803676, + "grad_norm": 0.049969664361995014, + "learning_rate": 6.338518133447358e-05, + "loss": 0.001, + "step": 5940 + }, + { + "epoch": 3.3130047340573654, + "grad_norm": 1.421807966504872, + "learning_rate": 6.325136074167997e-05, + "loss": 0.0106, + "step": 5950 + }, + { + "epoch": 3.3185742133110554, + "grad_norm": 1.439549073392725, + "learning_rate": 6.311743797789801e-05, + "loss": 0.0742, + "step": 5960 + }, + { + "epoch": 3.3241436925647454, + "grad_norm": 0.15647767908812543, + "learning_rate": 6.298341407570254e-05, + "loss": 0.0113, + "step": 5970 + }, + { + "epoch": 3.329713171818435, + "grad_norm": 0.09054542675523507, + "learning_rate": 6.284929006844811e-05, + "loss": 0.0044, + "step": 5980 + }, + { + "epoch": 3.3352826510721245, + "grad_norm": 0.0712984112272879, + "learning_rate": 6.271506699026117e-05, + "loss": 0.0029, + "step": 5990 + }, + { + "epoch": 3.3408521303258145, + "grad_norm": 0.048910263295018026, + "learning_rate": 6.258074587603203e-05, + "loss": 0.002, + "step": 6000 + }, + { + "epoch": 3.3464216095795045, + "grad_norm": 0.04531779661921402, + "learning_rate": 6.244632776140685e-05, + "loss": 0.0016, + "step": 6010 + }, + { + "epoch": 3.351991088833194, + "grad_norm": 0.05430636843709886, + "learning_rate": 6.231181368277966e-05, + "loss": 0.0013, + "step": 6020 + }, + { + "epoch": 3.3575605680868836, + "grad_norm": 0.052463656235233254, + "learning_rate": 6.217720467728448e-05, + "loss": 0.0012, + "step": 6030 + }, + { + "epoch": 3.3631300473405736, + "grad_norm": 0.04886826201389684, + "learning_rate": 6.204250178278714e-05, + "loss": 0.0012, + "step": 6040 + }, + { + "epoch": 3.3686995265942636, + "grad_norm": 0.052066003676750405, + "learning_rate": 6.190770603787746e-05, + "loss": 0.0011, + "step": 6050 + }, + { + "epoch": 3.374269005847953, + "grad_norm": 0.05360537743015038, + "learning_rate": 6.17728184818611e-05, + "loss": 0.0011, + "step": 6060 + }, + { + "epoch": 3.3798384851016428, + "grad_norm": 0.041695913106923485, + "learning_rate": 6.163784015475163e-05, + "loss": 0.0011, + "step": 6070 + }, + { + "epoch": 3.3854079643553328, + "grad_norm": 0.04532801732698349, + "learning_rate": 6.150277209726246e-05, + "loss": 0.001, + "step": 6080 + }, + { + "epoch": 3.3909774436090228, + "grad_norm": 0.04689447502281178, + "learning_rate": 6.136761535079889e-05, + "loss": 0.0009, + "step": 6090 + }, + { + "epoch": 3.3965469228627123, + "grad_norm": 0.04378654425299924, + "learning_rate": 6.123237095744998e-05, + "loss": 0.001, + "step": 6100 + }, + { + "epoch": 3.402116402116402, + "grad_norm": 0.04589882330754375, + "learning_rate": 6.109703995998062e-05, + "loss": 0.0009, + "step": 6110 + }, + { + "epoch": 3.407685881370092, + "grad_norm": 0.04799144071919005, + "learning_rate": 6.096162340182339e-05, + "loss": 0.0009, + "step": 6120 + }, + { + "epoch": 3.413255360623782, + "grad_norm": 0.04513415794722263, + "learning_rate": 6.082612232707057e-05, + "loss": 0.0009, + "step": 6130 + }, + { + "epoch": 3.4188248398774714, + "grad_norm": 0.04598843966625588, + "learning_rate": 6.06905377804661e-05, + "loss": 0.0009, + "step": 6140 + }, + { + "epoch": 3.424394319131161, + "grad_norm": 0.04074319943511705, + "learning_rate": 6.0554870807397514e-05, + "loss": 0.0009, + "step": 6150 + }, + { + "epoch": 3.429963798384851, + "grad_norm": 0.03684736201846013, + "learning_rate": 6.0419122453887835e-05, + "loss": 0.0008, + "step": 6160 + }, + { + "epoch": 3.435533277638541, + "grad_norm": 0.03882963532663941, + "learning_rate": 6.0283293766587565e-05, + "loss": 0.0008, + "step": 6170 + }, + { + "epoch": 3.4411027568922306, + "grad_norm": 0.0417008510488538, + "learning_rate": 6.014738579276663e-05, + "loss": 0.0008, + "step": 6180 + }, + { + "epoch": 3.44667223614592, + "grad_norm": 0.03821033614198779, + "learning_rate": 6.001139958030624e-05, + "loss": 0.0009, + "step": 6190 + }, + { + "epoch": 3.45224171539961, + "grad_norm": 0.03931741762159325, + "learning_rate": 5.987533617769085e-05, + "loss": 0.0009, + "step": 6200 + }, + { + "epoch": 3.4578111946533, + "grad_norm": 0.03958789501100999, + "learning_rate": 5.9739196634000036e-05, + "loss": 0.0008, + "step": 6210 + }, + { + "epoch": 3.4633806739069897, + "grad_norm": 0.04152058070780572, + "learning_rate": 5.9602981998900506e-05, + "loss": 0.0008, + "step": 6220 + }, + { + "epoch": 3.4689501531606792, + "grad_norm": 0.033696582310678305, + "learning_rate": 5.946669332263789e-05, + "loss": 0.0008, + "step": 6230 + }, + { + "epoch": 3.4745196324143692, + "grad_norm": 0.0409207454609883, + "learning_rate": 5.933033165602871e-05, + "loss": 0.0008, + "step": 6240 + }, + { + "epoch": 3.4800891116680592, + "grad_norm": 0.038742917441849135, + "learning_rate": 5.9193898050452254e-05, + "loss": 0.0008, + "step": 6250 + }, + { + "epoch": 3.485658590921749, + "grad_norm": 0.04258747135195118, + "learning_rate": 5.9057393557842464e-05, + "loss": 0.0009, + "step": 6260 + }, + { + "epoch": 3.4912280701754383, + "grad_norm": 0.04426364572697952, + "learning_rate": 5.892081923067988e-05, + "loss": 0.0008, + "step": 6270 + }, + { + "epoch": 3.4967975494291283, + "grad_norm": 0.03710017622802728, + "learning_rate": 5.8784176121983424e-05, + "loss": 0.0008, + "step": 6280 + }, + { + "epoch": 3.5023670286828184, + "grad_norm": 0.04186562929932779, + "learning_rate": 5.8647465285302396e-05, + "loss": 0.0008, + "step": 6290 + }, + { + "epoch": 3.507936507936508, + "grad_norm": 0.033774563864596964, + "learning_rate": 5.851068777470824e-05, + "loss": 0.0009, + "step": 6300 + }, + { + "epoch": 3.5135059871901975, + "grad_norm": 0.043453099514572775, + "learning_rate": 5.837384464478651e-05, + "loss": 0.0008, + "step": 6310 + }, + { + "epoch": 3.5190754664438875, + "grad_norm": 0.031255192673870356, + "learning_rate": 5.8236936950628685e-05, + "loss": 0.0008, + "step": 6320 + }, + { + "epoch": 3.5246449456975775, + "grad_norm": 0.04177287180613854, + "learning_rate": 5.809996574782406e-05, + "loss": 0.0008, + "step": 6330 + }, + { + "epoch": 3.530214424951267, + "grad_norm": 0.03570891056507373, + "learning_rate": 5.796293209245156e-05, + "loss": 0.0008, + "step": 6340 + }, + { + "epoch": 3.5357839042049566, + "grad_norm": 0.04543490225974056, + "learning_rate": 5.78258370410717e-05, + "loss": 0.0008, + "step": 6350 + }, + { + "epoch": 3.5413533834586466, + "grad_norm": 0.037674589582602676, + "learning_rate": 5.768868165071828e-05, + "loss": 0.0007, + "step": 6360 + }, + { + "epoch": 3.5469228627123366, + "grad_norm": 0.04283672689689677, + "learning_rate": 5.755146697889043e-05, + "loss": 0.0007, + "step": 6370 + }, + { + "epoch": 3.552492341966026, + "grad_norm": 0.04613490963291274, + "learning_rate": 5.741419408354428e-05, + "loss": 0.0007, + "step": 6380 + }, + { + "epoch": 3.5580618212197157, + "grad_norm": 0.03831948184941402, + "learning_rate": 5.727686402308491e-05, + "loss": 0.0008, + "step": 6390 + }, + { + "epoch": 3.5636313004734057, + "grad_norm": 0.04373492805743164, + "learning_rate": 5.713947785635814e-05, + "loss": 0.0007, + "step": 6400 + }, + { + "epoch": 3.5692007797270957, + "grad_norm": 0.03606071794903106, + "learning_rate": 5.7002036642642385e-05, + "loss": 0.0008, + "step": 6410 + }, + { + "epoch": 3.5747702589807853, + "grad_norm": 0.0435467415119081, + "learning_rate": 5.686454144164048e-05, + "loss": 0.0008, + "step": 6420 + }, + { + "epoch": 3.580339738234475, + "grad_norm": 0.04553140953577427, + "learning_rate": 5.6726993313471534e-05, + "loss": 0.0007, + "step": 6430 + }, + { + "epoch": 3.585909217488165, + "grad_norm": 0.03763157999258997, + "learning_rate": 5.658939331866272e-05, + "loss": 0.0008, + "step": 6440 + }, + { + "epoch": 3.591478696741855, + "grad_norm": 0.03125791174922972, + "learning_rate": 5.645174251814113e-05, + "loss": 0.0008, + "step": 6450 + }, + { + "epoch": 3.5970481759955444, + "grad_norm": 0.038362149890796976, + "learning_rate": 5.631404197322555e-05, + "loss": 0.0007, + "step": 6460 + }, + { + "epoch": 3.602617655249234, + "grad_norm": 0.02880089055072718, + "learning_rate": 5.6176292745618334e-05, + "loss": 0.0008, + "step": 6470 + }, + { + "epoch": 3.608187134502924, + "grad_norm": 0.03836015232318297, + "learning_rate": 5.603849589739717e-05, + "loss": 0.0008, + "step": 6480 + }, + { + "epoch": 3.613756613756614, + "grad_norm": 0.0414519817813859, + "learning_rate": 5.5900652491006934e-05, + "loss": 0.0008, + "step": 6490 + }, + { + "epoch": 3.6193260930103035, + "grad_norm": 0.033864081042114254, + "learning_rate": 5.576276358925145e-05, + "loss": 0.0008, + "step": 6500 + }, + { + "epoch": 3.624895572263993, + "grad_norm": 0.04455497998172137, + "learning_rate": 5.5624830255285375e-05, + "loss": 0.0008, + "step": 6510 + }, + { + "epoch": 3.630465051517683, + "grad_norm": 0.04357727523734538, + "learning_rate": 5.548685355260588e-05, + "loss": 0.0007, + "step": 6520 + }, + { + "epoch": 3.636034530771373, + "grad_norm": 0.042652456078840245, + "learning_rate": 5.534883454504454e-05, + "loss": 0.0007, + "step": 6530 + }, + { + "epoch": 3.6416040100250626, + "grad_norm": 0.037426870745509155, + "learning_rate": 5.5210774296759124e-05, + "loss": 0.0007, + "step": 6540 + }, + { + "epoch": 3.647173489278752, + "grad_norm": 0.04532962407686617, + "learning_rate": 5.50726738722254e-05, + "loss": 0.0008, + "step": 6550 + }, + { + "epoch": 3.652742968532442, + "grad_norm": 0.037854949680051396, + "learning_rate": 5.493453433622884e-05, + "loss": 0.0008, + "step": 6560 + }, + { + "epoch": 3.658312447786132, + "grad_norm": 0.04632593363217753, + "learning_rate": 5.479635675385651e-05, + "loss": 0.0008, + "step": 6570 + }, + { + "epoch": 3.6638819270398217, + "grad_norm": 0.03465170646267114, + "learning_rate": 5.465814219048884e-05, + "loss": 0.0008, + "step": 6580 + }, + { + "epoch": 3.6694514062935113, + "grad_norm": 0.03887446506597083, + "learning_rate": 5.451989171179137e-05, + "loss": 0.0007, + "step": 6590 + }, + { + "epoch": 3.6750208855472013, + "grad_norm": 0.04531880959425709, + "learning_rate": 5.438160638370655e-05, + "loss": 0.0007, + "step": 6600 + }, + { + "epoch": 3.6805903648008913, + "grad_norm": 0.044488727708099154, + "learning_rate": 5.424328727244554e-05, + "loss": 0.0008, + "step": 6610 + }, + { + "epoch": 3.686159844054581, + "grad_norm": 0.035419225329793136, + "learning_rate": 5.410493544447998e-05, + "loss": 0.0008, + "step": 6620 + }, + { + "epoch": 3.6917293233082704, + "grad_norm": 0.04187577661691093, + "learning_rate": 5.396655196653373e-05, + "loss": 0.0007, + "step": 6630 + }, + { + "epoch": 3.6972988025619604, + "grad_norm": 0.04459019148048656, + "learning_rate": 5.382813790557473e-05, + "loss": 0.0008, + "step": 6640 + }, + { + "epoch": 3.7028682818156504, + "grad_norm": 0.03548460780780217, + "learning_rate": 5.368969432880666e-05, + "loss": 0.0008, + "step": 6650 + }, + { + "epoch": 3.70843776106934, + "grad_norm": 0.03289781714218229, + "learning_rate": 5.355122230366084e-05, + "loss": 0.0007, + "step": 6660 + }, + { + "epoch": 3.7140072403230295, + "grad_norm": 0.038909546771706056, + "learning_rate": 5.341272289778789e-05, + "loss": 0.0007, + "step": 6670 + }, + { + "epoch": 3.7195767195767195, + "grad_norm": 0.03786515696640156, + "learning_rate": 5.3274197179049555e-05, + "loss": 0.0007, + "step": 6680 + }, + { + "epoch": 3.7251461988304095, + "grad_norm": 0.03524733328906414, + "learning_rate": 5.313564621551047e-05, + "loss": 0.0006, + "step": 6690 + }, + { + "epoch": 3.730715678084099, + "grad_norm": 0.04464320912427213, + "learning_rate": 5.299707107542987e-05, + "loss": 0.0006, + "step": 6700 + }, + { + "epoch": 3.7362851573377887, + "grad_norm": 0.03303522366491739, + "learning_rate": 5.2858472827253446e-05, + "loss": 0.0007, + "step": 6710 + }, + { + "epoch": 3.7418546365914787, + "grad_norm": 0.03311237338465701, + "learning_rate": 5.271985253960503e-05, + "loss": 0.0006, + "step": 6720 + }, + { + "epoch": 3.7474241158451687, + "grad_norm": 0.03837108996999619, + "learning_rate": 5.25812112812784e-05, + "loss": 0.0007, + "step": 6730 + }, + { + "epoch": 3.752993595098858, + "grad_norm": 0.05238733597259272, + "learning_rate": 5.2442550121229004e-05, + "loss": 0.0007, + "step": 6740 + }, + { + "epoch": 3.7585630743525478, + "grad_norm": 0.039429398538116084, + "learning_rate": 5.230387012856575e-05, + "loss": 0.0007, + "step": 6750 + }, + { + "epoch": 3.7641325536062378, + "grad_norm": 0.04249634326988299, + "learning_rate": 5.216517237254274e-05, + "loss": 0.0008, + "step": 6760 + }, + { + "epoch": 3.7697020328599278, + "grad_norm": 0.03815002787017476, + "learning_rate": 5.2026457922551054e-05, + "loss": 0.0007, + "step": 6770 + }, + { + "epoch": 3.7752715121136173, + "grad_norm": 0.0439584610592346, + "learning_rate": 5.188772784811048e-05, + "loss": 0.0008, + "step": 6780 + }, + { + "epoch": 3.780840991367307, + "grad_norm": 0.03737865410521001, + "learning_rate": 5.174898321886123e-05, + "loss": 0.0007, + "step": 6790 + }, + { + "epoch": 3.786410470620997, + "grad_norm": 0.03963441733048993, + "learning_rate": 5.161022510455581e-05, + "loss": 0.0007, + "step": 6800 + }, + { + "epoch": 3.791979949874687, + "grad_norm": 0.04046347514097831, + "learning_rate": 5.147145457505067e-05, + "loss": 0.0007, + "step": 6810 + }, + { + "epoch": 3.7975494291283765, + "grad_norm": 0.045904007964164696, + "learning_rate": 5.133267270029795e-05, + "loss": 0.0007, + "step": 6820 + }, + { + "epoch": 3.803118908382066, + "grad_norm": 0.03663875772520332, + "learning_rate": 5.1193880550337306e-05, + "loss": 0.0007, + "step": 6830 + }, + { + "epoch": 3.808688387635756, + "grad_norm": 0.026644648845485425, + "learning_rate": 5.10550791952876e-05, + "loss": 0.0006, + "step": 6840 + }, + { + "epoch": 3.814257866889446, + "grad_norm": 0.03982278227791823, + "learning_rate": 5.091626970533868e-05, + "loss": 0.0007, + "step": 6850 + }, + { + "epoch": 3.8198273461431356, + "grad_norm": 0.03625913736670255, + "learning_rate": 5.0777453150743104e-05, + "loss": 0.0006, + "step": 6860 + }, + { + "epoch": 3.825396825396825, + "grad_norm": 0.04059785016249621, + "learning_rate": 5.06386306018079e-05, + "loss": 0.0007, + "step": 6870 + }, + { + "epoch": 3.830966304650515, + "grad_norm": 0.03575355672682709, + "learning_rate": 5.049980312888631e-05, + "loss": 0.0007, + "step": 6880 + }, + { + "epoch": 3.836535783904205, + "grad_norm": 0.040692137996130244, + "learning_rate": 5.036097180236957e-05, + "loss": 0.0007, + "step": 6890 + }, + { + "epoch": 3.8421052631578947, + "grad_norm": 0.030776445659514116, + "learning_rate": 5.02221376926786e-05, + "loss": 0.0007, + "step": 6900 + }, + { + "epoch": 3.8476747424115842, + "grad_norm": 0.02786855557090512, + "learning_rate": 5.008330187025577e-05, + "loss": 0.0007, + "step": 6910 + }, + { + "epoch": 3.8532442216652742, + "grad_norm": 0.037399307076388964, + "learning_rate": 4.994446540555669e-05, + "loss": 0.0007, + "step": 6920 + }, + { + "epoch": 3.8588137009189643, + "grad_norm": 0.037103795554716795, + "learning_rate": 4.980562936904187e-05, + "loss": 0.0007, + "step": 6930 + }, + { + "epoch": 3.864383180172654, + "grad_norm": 0.03943454576749188, + "learning_rate": 4.96667948311686e-05, + "loss": 0.0007, + "step": 6940 + }, + { + "epoch": 3.8699526594263434, + "grad_norm": 0.038423947928509325, + "learning_rate": 4.952796286238252e-05, + "loss": 0.0007, + "step": 6950 + }, + { + "epoch": 3.8755221386800334, + "grad_norm": 0.03413635281209976, + "learning_rate": 4.938913453310955e-05, + "loss": 0.0007, + "step": 6960 + }, + { + "epoch": 3.8810916179337234, + "grad_norm": 0.03207580548519312, + "learning_rate": 4.9250310913747464e-05, + "loss": 0.0007, + "step": 6970 + }, + { + "epoch": 3.886661097187413, + "grad_norm": 0.046358932116040454, + "learning_rate": 4.911149307465781e-05, + "loss": 0.0006, + "step": 6980 + }, + { + "epoch": 3.8922305764411025, + "grad_norm": 0.04034266055056202, + "learning_rate": 4.897268208615748e-05, + "loss": 0.0007, + "step": 6990 + }, + { + "epoch": 3.8978000556947925, + "grad_norm": 0.045459825398252815, + "learning_rate": 4.883387901851063e-05, + "loss": 0.0007, + "step": 7000 + }, + { + "epoch": 3.9033695349484825, + "grad_norm": 0.03267704387989124, + "learning_rate": 4.869508494192027e-05, + "loss": 0.0006, + "step": 7010 + }, + { + "epoch": 3.908939014202172, + "grad_norm": 0.04188935061274112, + "learning_rate": 4.855630092652013e-05, + "loss": 0.0007, + "step": 7020 + }, + { + "epoch": 3.9145084934558616, + "grad_norm": 0.041272075094844714, + "learning_rate": 4.8417528042366384e-05, + "loss": 0.0007, + "step": 7030 + }, + { + "epoch": 3.9200779727095516, + "grad_norm": 0.0382784606501393, + "learning_rate": 4.827876735942932e-05, + "loss": 0.0007, + "step": 7040 + }, + { + "epoch": 3.9256474519632416, + "grad_norm": 0.04008544652564574, + "learning_rate": 4.814001994758522e-05, + "loss": 0.0007, + "step": 7050 + }, + { + "epoch": 3.931216931216931, + "grad_norm": 0.038679074945492324, + "learning_rate": 4.800128687660796e-05, + "loss": 0.0006, + "step": 7060 + }, + { + "epoch": 3.9367864104706207, + "grad_norm": 0.041705017679490114, + "learning_rate": 4.7862569216160945e-05, + "loss": 0.0007, + "step": 7070 + }, + { + "epoch": 3.9423558897243107, + "grad_norm": 0.03723297982293688, + "learning_rate": 4.7723868035788685e-05, + "loss": 0.0006, + "step": 7080 + }, + { + "epoch": 3.9479253689780007, + "grad_norm": 0.04150976073193777, + "learning_rate": 4.758518440490865e-05, + "loss": 0.0007, + "step": 7090 + }, + { + "epoch": 3.9534948482316903, + "grad_norm": 0.02783514600583792, + "learning_rate": 4.7446519392803e-05, + "loss": 0.0007, + "step": 7100 + }, + { + "epoch": 3.95906432748538, + "grad_norm": 0.03505648605569088, + "learning_rate": 4.730787406861035e-05, + "loss": 0.0006, + "step": 7110 + }, + { + "epoch": 3.96463380673907, + "grad_norm": 0.042118582112596864, + "learning_rate": 4.7169249501317496e-05, + "loss": 0.0007, + "step": 7120 + }, + { + "epoch": 3.97020328599276, + "grad_norm": 0.03185619472602036, + "learning_rate": 4.703064675975122e-05, + "loss": 0.0007, + "step": 7130 + }, + { + "epoch": 3.9757727652464494, + "grad_norm": 0.03679155697503452, + "learning_rate": 4.689206691256997e-05, + "loss": 0.0007, + "step": 7140 + }, + { + "epoch": 3.981342244500139, + "grad_norm": 0.04325458714225093, + "learning_rate": 4.675351102825576e-05, + "loss": 0.0007, + "step": 7150 + }, + { + "epoch": 3.986911723753829, + "grad_norm": 0.039917026007133004, + "learning_rate": 4.661498017510574e-05, + "loss": 0.0007, + "step": 7160 + }, + { + "epoch": 3.992481203007519, + "grad_norm": 0.0436246236883007, + "learning_rate": 4.647647542122417e-05, + "loss": 0.0007, + "step": 7170 + }, + { + "epoch": 3.9980506822612085, + "grad_norm": 0.03163468168248903, + "learning_rate": 4.6337997834514005e-05, + "loss": 0.0006, + "step": 7180 + }, + { + "epoch": 4.003341687552214, + "grad_norm": 0.03436761663971402, + "learning_rate": 4.6199548482668764e-05, + "loss": 0.0006, + "step": 7190 + }, + { + "epoch": 4.008911166805904, + "grad_norm": 0.03230596602205761, + "learning_rate": 4.606112843316423e-05, + "loss": 0.0005, + "step": 7200 + }, + { + "epoch": 4.014480646059593, + "grad_norm": 0.05420299272802651, + "learning_rate": 4.5922738753250344e-05, + "loss": 0.0004, + "step": 7210 + }, + { + "epoch": 4.020050125313283, + "grad_norm": 0.02208052954000821, + "learning_rate": 4.5784380509942796e-05, + "loss": 0.0004, + "step": 7220 + }, + { + "epoch": 4.025619604566973, + "grad_norm": 0.019864104272280855, + "learning_rate": 4.564605477001497e-05, + "loss": 0.0003, + "step": 7230 + }, + { + "epoch": 4.031189083820663, + "grad_norm": 0.027669750336795477, + "learning_rate": 4.550776259998955e-05, + "loss": 0.0003, + "step": 7240 + }, + { + "epoch": 4.036758563074352, + "grad_norm": 0.02443792386135483, + "learning_rate": 4.5369505066130505e-05, + "loss": 0.0003, + "step": 7250 + }, + { + "epoch": 4.042328042328043, + "grad_norm": 0.015124310693838015, + "learning_rate": 4.523128323443465e-05, + "loss": 0.0003, + "step": 7260 + }, + { + "epoch": 4.047897521581732, + "grad_norm": 0.014594677971953096, + "learning_rate": 4.509309817062359e-05, + "loss": 0.0003, + "step": 7270 + }, + { + "epoch": 4.053467000835422, + "grad_norm": 0.020101645396513233, + "learning_rate": 4.4954950940135396e-05, + "loss": 0.0003, + "step": 7280 + }, + { + "epoch": 4.059036480089111, + "grad_norm": 0.020007595628344848, + "learning_rate": 4.481684260811649e-05, + "loss": 0.0003, + "step": 7290 + }, + { + "epoch": 4.064605959342801, + "grad_norm": 0.017021378897841468, + "learning_rate": 4.467877423941332e-05, + "loss": 0.0003, + "step": 7300 + }, + { + "epoch": 4.0701754385964914, + "grad_norm": 0.029975896537104644, + "learning_rate": 4.454074689856426e-05, + "loss": 0.0003, + "step": 7310 + }, + { + "epoch": 4.075744917850181, + "grad_norm": 0.016759188379980753, + "learning_rate": 4.4402761649791305e-05, + "loss": 0.0003, + "step": 7320 + }, + { + "epoch": 4.081314397103871, + "grad_norm": 0.01560375700274465, + "learning_rate": 4.426481955699195e-05, + "loss": 0.0003, + "step": 7330 + }, + { + "epoch": 4.086883876357561, + "grad_norm": 0.021138502878005942, + "learning_rate": 4.4126921683730894e-05, + "loss": 0.0003, + "step": 7340 + }, + { + "epoch": 4.092453355611251, + "grad_norm": 0.023994105639638687, + "learning_rate": 4.398906909323197e-05, + "loss": 0.0003, + "step": 7350 + }, + { + "epoch": 4.09802283486494, + "grad_norm": 0.02229787949311062, + "learning_rate": 4.38512628483698e-05, + "loss": 0.0003, + "step": 7360 + }, + { + "epoch": 4.10359231411863, + "grad_norm": 0.023856939824296347, + "learning_rate": 4.371350401166172e-05, + "loss": 0.0003, + "step": 7370 + }, + { + "epoch": 4.109161793372319, + "grad_norm": 0.015277959676836197, + "learning_rate": 4.357579364525948e-05, + "loss": 0.0003, + "step": 7380 + }, + { + "epoch": 4.11473127262601, + "grad_norm": 0.01897198700103696, + "learning_rate": 4.343813281094119e-05, + "loss": 0.0003, + "step": 7390 + }, + { + "epoch": 4.120300751879699, + "grad_norm": 0.029599977858735915, + "learning_rate": 4.330052257010301e-05, + "loss": 0.0003, + "step": 7400 + }, + { + "epoch": 4.125870231133389, + "grad_norm": 0.02245476111575353, + "learning_rate": 4.316296398375101e-05, + "loss": 0.0003, + "step": 7410 + }, + { + "epoch": 4.131439710387079, + "grad_norm": 0.020892877872472525, + "learning_rate": 4.3025458112492984e-05, + "loss": 0.0003, + "step": 7420 + }, + { + "epoch": 4.137009189640769, + "grad_norm": 0.022232091610007172, + "learning_rate": 4.288800601653034e-05, + "loss": 0.0003, + "step": 7430 + }, + { + "epoch": 4.142578668894458, + "grad_norm": 0.023342817173689547, + "learning_rate": 4.2750608755649794e-05, + "loss": 0.0002, + "step": 7440 + }, + { + "epoch": 4.148148148148148, + "grad_norm": 0.021609314457823704, + "learning_rate": 4.26132673892153e-05, + "loss": 0.0003, + "step": 7450 + }, + { + "epoch": 4.1537176274018375, + "grad_norm": 0.02524920196519335, + "learning_rate": 4.247598297615989e-05, + "loss": 0.0003, + "step": 7460 + }, + { + "epoch": 4.159287106655528, + "grad_norm": 0.019251961157805003, + "learning_rate": 4.23387565749774e-05, + "loss": 0.0003, + "step": 7470 + }, + { + "epoch": 4.1648565859092175, + "grad_norm": 0.020220265229742084, + "learning_rate": 4.220158924371444e-05, + "loss": 0.0003, + "step": 7480 + }, + { + "epoch": 4.170426065162907, + "grad_norm": 0.024468493324441506, + "learning_rate": 4.206448203996215e-05, + "loss": 0.0002, + "step": 7490 + }, + { + "epoch": 4.1759955444165975, + "grad_norm": 0.011904755189469405, + "learning_rate": 4.192743602084809e-05, + "loss": 0.0002, + "step": 7500 + }, + { + "epoch": 4.181565023670287, + "grad_norm": 0.016958659356378095, + "learning_rate": 4.1790452243028025e-05, + "loss": 0.0002, + "step": 7510 + }, + { + "epoch": 4.187134502923977, + "grad_norm": 0.025602517987447074, + "learning_rate": 4.165353176267792e-05, + "loss": 0.0002, + "step": 7520 + }, + { + "epoch": 4.192703982177666, + "grad_norm": 0.018474401125215874, + "learning_rate": 4.15166756354856e-05, + "loss": 0.0002, + "step": 7530 + }, + { + "epoch": 4.198273461431356, + "grad_norm": 0.021443456401005506, + "learning_rate": 4.137988491664279e-05, + "loss": 0.0002, + "step": 7540 + }, + { + "epoch": 4.203842940685046, + "grad_norm": 0.025575679305059097, + "learning_rate": 4.124316066083683e-05, + "loss": 0.0002, + "step": 7550 + }, + { + "epoch": 4.209412419938736, + "grad_norm": 0.023659157265637706, + "learning_rate": 4.110650392224271e-05, + "loss": 0.0003, + "step": 7560 + }, + { + "epoch": 4.214981899192425, + "grad_norm": 0.019470818718145953, + "learning_rate": 4.096991575451474e-05, + "loss": 0.0002, + "step": 7570 + }, + { + "epoch": 4.220551378446116, + "grad_norm": 0.020649921784656862, + "learning_rate": 4.083339721077862e-05, + "loss": 0.0002, + "step": 7580 + }, + { + "epoch": 4.226120857699805, + "grad_norm": 0.02241734155863576, + "learning_rate": 4.069694934362316e-05, + "loss": 0.0002, + "step": 7590 + }, + { + "epoch": 4.231690336953495, + "grad_norm": 0.033895846389183276, + "learning_rate": 4.056057320509232e-05, + "loss": 0.0003, + "step": 7600 + }, + { + "epoch": 4.237259816207184, + "grad_norm": 0.021868221161894033, + "learning_rate": 4.042426984667694e-05, + "loss": 0.0003, + "step": 7610 + }, + { + "epoch": 4.242829295460874, + "grad_norm": 0.02874027151985194, + "learning_rate": 4.028804031930676e-05, + "loss": 0.0003, + "step": 7620 + }, + { + "epoch": 4.248398774714564, + "grad_norm": 0.023713541150551867, + "learning_rate": 4.0151885673342224e-05, + "loss": 0.0002, + "step": 7630 + }, + { + "epoch": 4.253968253968254, + "grad_norm": 0.02144103561746673, + "learning_rate": 4.001580695856648e-05, + "loss": 0.0002, + "step": 7640 + }, + { + "epoch": 4.2595377332219435, + "grad_norm": 0.013571026066090681, + "learning_rate": 3.987980522417716e-05, + "loss": 0.0002, + "step": 7650 + }, + { + "epoch": 4.265107212475634, + "grad_norm": 0.027387007880172357, + "learning_rate": 3.974388151877842e-05, + "loss": 0.0003, + "step": 7660 + }, + { + "epoch": 4.2706766917293235, + "grad_norm": 0.026816035871890628, + "learning_rate": 3.960803689037276e-05, + "loss": 0.0003, + "step": 7670 + }, + { + "epoch": 4.276246170983013, + "grad_norm": 0.018835268597604964, + "learning_rate": 3.947227238635299e-05, + "loss": 0.0002, + "step": 7680 + }, + { + "epoch": 4.281815650236703, + "grad_norm": 0.02241465372687628, + "learning_rate": 3.933658905349413e-05, + "loss": 0.0002, + "step": 7690 + }, + { + "epoch": 4.287385129490392, + "grad_norm": 0.012914133442070772, + "learning_rate": 3.920098793794539e-05, + "loss": 0.0002, + "step": 7700 + }, + { + "epoch": 4.292954608744083, + "grad_norm": 0.029041701416590862, + "learning_rate": 3.9065470085222016e-05, + "loss": 0.0002, + "step": 7710 + }, + { + "epoch": 4.298524087997772, + "grad_norm": 0.021565298478882302, + "learning_rate": 3.8930036540197315e-05, + "loss": 0.0003, + "step": 7720 + }, + { + "epoch": 4.304093567251462, + "grad_norm": 0.0196454710051085, + "learning_rate": 3.879468834709453e-05, + "loss": 0.0003, + "step": 7730 + }, + { + "epoch": 4.309663046505152, + "grad_norm": 0.04106022926156392, + "learning_rate": 3.8659426549478874e-05, + "loss": 0.0003, + "step": 7740 + }, + { + "epoch": 4.315232525758842, + "grad_norm": 0.02256221787840566, + "learning_rate": 3.852425219024937e-05, + "loss": 0.0003, + "step": 7750 + }, + { + "epoch": 4.320802005012531, + "grad_norm": 0.026396462378366594, + "learning_rate": 3.838916631163091e-05, + "loss": 0.0003, + "step": 7760 + }, + { + "epoch": 4.326371484266221, + "grad_norm": 0.0336704117003004, + "learning_rate": 3.825416995516616e-05, + "loss": 0.0003, + "step": 7770 + }, + { + "epoch": 4.33194096351991, + "grad_norm": 0.022827340577645957, + "learning_rate": 3.811926416170759e-05, + "loss": 0.0002, + "step": 7780 + }, + { + "epoch": 4.337510442773601, + "grad_norm": 0.02437956026905337, + "learning_rate": 3.798444997140933e-05, + "loss": 0.0002, + "step": 7790 + }, + { + "epoch": 4.34307992202729, + "grad_norm": 0.027116486246569028, + "learning_rate": 3.7849728423719336e-05, + "loss": 0.0002, + "step": 7800 + }, + { + "epoch": 4.34864940128098, + "grad_norm": 0.028870409567424055, + "learning_rate": 3.771510055737117e-05, + "loss": 0.0002, + "step": 7810 + }, + { + "epoch": 4.35421888053467, + "grad_norm": 0.026556512704674462, + "learning_rate": 3.7580567410376133e-05, + "loss": 0.0002, + "step": 7820 + }, + { + "epoch": 4.35978835978836, + "grad_norm": 0.022088843902153613, + "learning_rate": 3.7446130020015215e-05, + "loss": 0.0002, + "step": 7830 + }, + { + "epoch": 4.3653578390420495, + "grad_norm": 0.031303991520526965, + "learning_rate": 3.731178942283112e-05, + "loss": 0.0002, + "step": 7840 + }, + { + "epoch": 4.370927318295739, + "grad_norm": 0.036051525501609685, + "learning_rate": 3.7177546654620225e-05, + "loss": 0.0003, + "step": 7850 + }, + { + "epoch": 4.376496797549429, + "grad_norm": 0.055378211817165644, + "learning_rate": 3.7043402750424636e-05, + "loss": 0.0003, + "step": 7860 + }, + { + "epoch": 4.382066276803119, + "grad_norm": 0.023987910434473545, + "learning_rate": 3.690935874452417e-05, + "loss": 0.0003, + "step": 7870 + }, + { + "epoch": 4.387635756056809, + "grad_norm": 0.02605032255225048, + "learning_rate": 3.6775415670428456e-05, + "loss": 0.0003, + "step": 7880 + }, + { + "epoch": 4.393205235310498, + "grad_norm": 0.02410479141489606, + "learning_rate": 3.6641574560868904e-05, + "loss": 0.0003, + "step": 7890 + }, + { + "epoch": 4.398774714564189, + "grad_norm": 0.03278671808228018, + "learning_rate": 3.650783644779069e-05, + "loss": 0.0003, + "step": 7900 + }, + { + "epoch": 4.404344193817878, + "grad_norm": 0.025990571415587593, + "learning_rate": 3.6374202362344954e-05, + "loss": 0.0003, + "step": 7910 + }, + { + "epoch": 4.409913673071568, + "grad_norm": 0.018414445695970897, + "learning_rate": 3.6240673334880696e-05, + "loss": 0.0003, + "step": 7920 + }, + { + "epoch": 4.415483152325257, + "grad_norm": 0.022968321794147085, + "learning_rate": 3.6107250394936934e-05, + "loss": 0.0003, + "step": 7930 + }, + { + "epoch": 4.421052631578947, + "grad_norm": 0.019323665804101168, + "learning_rate": 3.597393457123469e-05, + "loss": 0.0003, + "step": 7940 + }, + { + "epoch": 4.426622110832637, + "grad_norm": 0.0430581209750874, + "learning_rate": 3.584072689166914e-05, + "loss": 0.0003, + "step": 7950 + }, + { + "epoch": 4.432191590086327, + "grad_norm": 0.023622088905074327, + "learning_rate": 3.570762838330159e-05, + "loss": 0.0003, + "step": 7960 + }, + { + "epoch": 4.4377610693400165, + "grad_norm": 0.021491779099359728, + "learning_rate": 3.557464007235168e-05, + "loss": 0.0003, + "step": 7970 + }, + { + "epoch": 4.443330548593707, + "grad_norm": 0.023928076379700115, + "learning_rate": 3.544176298418934e-05, + "loss": 0.0002, + "step": 7980 + }, + { + "epoch": 4.4489000278473965, + "grad_norm": 0.030204980302547018, + "learning_rate": 3.5308998143326985e-05, + "loss": 0.0003, + "step": 7990 + }, + { + "epoch": 4.454469507101086, + "grad_norm": 0.02102160312448877, + "learning_rate": 3.5176346573411534e-05, + "loss": 0.0002, + "step": 8000 + }, + { + "epoch": 4.460038986354776, + "grad_norm": 0.027408675266427065, + "learning_rate": 3.504380929721665e-05, + "loss": 0.0002, + "step": 8010 + }, + { + "epoch": 4.465608465608465, + "grad_norm": 0.01734059975076004, + "learning_rate": 3.491138733663465e-05, + "loss": 0.0003, + "step": 8020 + }, + { + "epoch": 4.471177944862156, + "grad_norm": 0.028648917276859448, + "learning_rate": 3.477908171266884e-05, + "loss": 0.0003, + "step": 8030 + }, + { + "epoch": 4.476747424115845, + "grad_norm": 0.02314351525817084, + "learning_rate": 3.4646893445425465e-05, + "loss": 0.0003, + "step": 8040 + }, + { + "epoch": 4.482316903369535, + "grad_norm": 0.03036672650760454, + "learning_rate": 3.451482355410601e-05, + "loss": 0.0003, + "step": 8050 + }, + { + "epoch": 4.487886382623225, + "grad_norm": 0.021404521834980084, + "learning_rate": 3.438287305699916e-05, + "loss": 0.0002, + "step": 8060 + }, + { + "epoch": 4.493455861876915, + "grad_norm": 0.024267023163270517, + "learning_rate": 3.4251042971473135e-05, + "loss": 0.0002, + "step": 8070 + }, + { + "epoch": 4.499025341130604, + "grad_norm": 0.020483509250194612, + "learning_rate": 3.4119334313967666e-05, + "loss": 0.0002, + "step": 8080 + }, + { + "epoch": 4.504594820384294, + "grad_norm": 0.019967150897930747, + "learning_rate": 3.398774809998633e-05, + "loss": 0.0002, + "step": 8090 + }, + { + "epoch": 4.510164299637983, + "grad_norm": 0.029627594458198295, + "learning_rate": 3.3856285344088556e-05, + "loss": 0.0002, + "step": 8100 + }, + { + "epoch": 4.515733778891674, + "grad_norm": 0.028535191791404346, + "learning_rate": 3.372494705988197e-05, + "loss": 0.0003, + "step": 8110 + }, + { + "epoch": 4.521303258145363, + "grad_norm": 0.02010989192665651, + "learning_rate": 3.3593734260014405e-05, + "loss": 0.0002, + "step": 8120 + }, + { + "epoch": 4.526872737399053, + "grad_norm": 0.020516481817362656, + "learning_rate": 3.346264795616626e-05, + "loss": 0.0003, + "step": 8130 + }, + { + "epoch": 4.532442216652743, + "grad_norm": 0.026664864935030445, + "learning_rate": 3.3331689159042546e-05, + "loss": 0.0003, + "step": 8140 + }, + { + "epoch": 4.538011695906433, + "grad_norm": 0.017416444545264947, + "learning_rate": 3.3200858878365246e-05, + "loss": 0.0003, + "step": 8150 + }, + { + "epoch": 4.5435811751601225, + "grad_norm": 0.021124422854669214, + "learning_rate": 3.30701581228654e-05, + "loss": 0.0003, + "step": 8160 + }, + { + "epoch": 4.549150654413812, + "grad_norm": 0.019670718961026068, + "learning_rate": 3.2939587900275395e-05, + "loss": 0.0003, + "step": 8170 + }, + { + "epoch": 4.554720133667502, + "grad_norm": 0.01895664049190062, + "learning_rate": 3.280914921732115e-05, + "loss": 0.0002, + "step": 8180 + }, + { + "epoch": 4.560289612921192, + "grad_norm": 0.01892167690469078, + "learning_rate": 3.267884307971446e-05, + "loss": 0.0002, + "step": 8190 + }, + { + "epoch": 4.565859092174882, + "grad_norm": 0.03029342708143485, + "learning_rate": 3.2548670492145075e-05, + "loss": 0.0003, + "step": 8200 + }, + { + "epoch": 4.571428571428571, + "grad_norm": 0.020911849580668548, + "learning_rate": 3.241863245827312e-05, + "loss": 0.0003, + "step": 8210 + }, + { + "epoch": 4.576998050682262, + "grad_norm": 0.024780862090682042, + "learning_rate": 3.22887299807212e-05, + "loss": 0.0003, + "step": 8220 + }, + { + "epoch": 4.582567529935951, + "grad_norm": 0.043718626657064816, + "learning_rate": 3.215896406106686e-05, + "loss": 0.0003, + "step": 8230 + }, + { + "epoch": 4.588137009189641, + "grad_norm": 0.019657762700119062, + "learning_rate": 3.2029335699834624e-05, + "loss": 0.0003, + "step": 8240 + }, + { + "epoch": 4.59370648844333, + "grad_norm": 0.02104812097424732, + "learning_rate": 3.189984589648852e-05, + "loss": 0.0003, + "step": 8250 + }, + { + "epoch": 4.59927596769702, + "grad_norm": 0.025773131748369645, + "learning_rate": 3.177049564942419e-05, + "loss": 0.0003, + "step": 8260 + }, + { + "epoch": 4.60484544695071, + "grad_norm": 0.018834428471579606, + "learning_rate": 3.1641285955961334e-05, + "loss": 0.0002, + "step": 8270 + }, + { + "epoch": 4.6104149262044, + "grad_norm": 0.022936079055118547, + "learning_rate": 3.151221781233585e-05, + "loss": 0.0003, + "step": 8280 + }, + { + "epoch": 4.615984405458089, + "grad_norm": 0.03490801003953351, + "learning_rate": 3.138329221369237e-05, + "loss": 0.0003, + "step": 8290 + }, + { + "epoch": 4.62155388471178, + "grad_norm": 0.03400909673674782, + "learning_rate": 3.12545101540764e-05, + "loss": 0.0003, + "step": 8300 + }, + { + "epoch": 4.627123363965469, + "grad_norm": 0.02991740466452301, + "learning_rate": 3.112587262642672e-05, + "loss": 0.0003, + "step": 8310 + }, + { + "epoch": 4.632692843219159, + "grad_norm": 0.026657724785102224, + "learning_rate": 3.0997380622567815e-05, + "loss": 0.0003, + "step": 8320 + }, + { + "epoch": 4.6382623224728485, + "grad_norm": 0.020664202196168057, + "learning_rate": 3.0869035133202065e-05, + "loss": 0.0003, + "step": 8330 + }, + { + "epoch": 4.643831801726538, + "grad_norm": 0.0329415133713272, + "learning_rate": 3.074083714790222e-05, + "loss": 0.0003, + "step": 8340 + }, + { + "epoch": 4.6494012809802285, + "grad_norm": 0.03920090357403892, + "learning_rate": 3.061278765510373e-05, + "loss": 0.0003, + "step": 8350 + }, + { + "epoch": 4.654970760233918, + "grad_norm": 0.0231832195261828, + "learning_rate": 3.048488764209717e-05, + "loss": 0.0003, + "step": 8360 + }, + { + "epoch": 4.660540239487608, + "grad_norm": 0.033371731741804044, + "learning_rate": 3.0357138095020547e-05, + "loss": 0.0003, + "step": 8370 + }, + { + "epoch": 4.666109718741298, + "grad_norm": 0.023306466181162215, + "learning_rate": 3.0229539998851784e-05, + "loss": 0.0003, + "step": 8380 + }, + { + "epoch": 4.671679197994988, + "grad_norm": 0.02613757787170549, + "learning_rate": 3.010209433740102e-05, + "loss": 0.0002, + "step": 8390 + }, + { + "epoch": 4.677248677248677, + "grad_norm": 0.022703352234951433, + "learning_rate": 2.9974802093303178e-05, + "loss": 0.0003, + "step": 8400 + }, + { + "epoch": 4.682818156502367, + "grad_norm": 0.02577174650387999, + "learning_rate": 2.984766424801022e-05, + "loss": 0.0003, + "step": 8410 + }, + { + "epoch": 4.688387635756056, + "grad_norm": 0.01797136177359523, + "learning_rate": 2.97206817817837e-05, + "loss": 0.0002, + "step": 8420 + }, + { + "epoch": 4.693957115009747, + "grad_norm": 0.031685735073726094, + "learning_rate": 2.9593855673687128e-05, + "loss": 0.0003, + "step": 8430 + }, + { + "epoch": 4.699526594263436, + "grad_norm": 0.024635245360987712, + "learning_rate": 2.9467186901578515e-05, + "loss": 0.0002, + "step": 8440 + }, + { + "epoch": 4.705096073517126, + "grad_norm": 0.02964575750320147, + "learning_rate": 2.934067644210271e-05, + "loss": 0.0002, + "step": 8450 + }, + { + "epoch": 4.710665552770816, + "grad_norm": 0.02329123306178194, + "learning_rate": 2.9214325270683996e-05, + "loss": 0.0002, + "step": 8460 + }, + { + "epoch": 4.716235032024506, + "grad_norm": 0.02507822092946219, + "learning_rate": 2.908813436151846e-05, + "loss": 0.0002, + "step": 8470 + }, + { + "epoch": 4.7218045112781954, + "grad_norm": 0.0222917907146546, + "learning_rate": 2.8962104687566584e-05, + "loss": 0.0002, + "step": 8480 + }, + { + "epoch": 4.727373990531885, + "grad_norm": 0.02736685367757904, + "learning_rate": 2.8836237220545615e-05, + "loss": 0.0003, + "step": 8490 + }, + { + "epoch": 4.732943469785575, + "grad_norm": 0.022742144192448254, + "learning_rate": 2.8710532930922206e-05, + "loss": 0.0003, + "step": 8500 + }, + { + "epoch": 4.738512949039265, + "grad_norm": 0.022180282552811976, + "learning_rate": 2.8584992787904874e-05, + "loss": 0.0002, + "step": 8510 + }, + { + "epoch": 4.744082428292955, + "grad_norm": 0.030657833147517678, + "learning_rate": 2.84596177594365e-05, + "loss": 0.0003, + "step": 8520 + }, + { + "epoch": 4.749651907546644, + "grad_norm": 0.02421528075541618, + "learning_rate": 2.833440881218692e-05, + "loss": 0.0002, + "step": 8530 + }, + { + "epoch": 4.755221386800335, + "grad_norm": 0.033635613649564196, + "learning_rate": 2.820936691154544e-05, + "loss": 0.0003, + "step": 8540 + }, + { + "epoch": 4.760790866054024, + "grad_norm": 0.029729370376208944, + "learning_rate": 2.808449302161335e-05, + "loss": 0.0002, + "step": 8550 + }, + { + "epoch": 4.766360345307714, + "grad_norm": 0.023425710360053496, + "learning_rate": 2.795978810519666e-05, + "loss": 0.0002, + "step": 8560 + }, + { + "epoch": 4.771929824561403, + "grad_norm": 0.028247102625589725, + "learning_rate": 2.783525312379841e-05, + "loss": 0.0003, + "step": 8570 + }, + { + "epoch": 4.777499303815093, + "grad_norm": 0.01978652584936416, + "learning_rate": 2.7710889037611508e-05, + "loss": 0.0002, + "step": 8580 + }, + { + "epoch": 4.783068783068783, + "grad_norm": 0.025857640915810193, + "learning_rate": 2.758669680551117e-05, + "loss": 0.0002, + "step": 8590 + }, + { + "epoch": 4.788638262322473, + "grad_norm": 0.0379149114456046, + "learning_rate": 2.7462677385047603e-05, + "loss": 0.0003, + "step": 8600 + }, + { + "epoch": 4.794207741576162, + "grad_norm": 0.022333280872816632, + "learning_rate": 2.7338831732438586e-05, + "loss": 0.0002, + "step": 8610 + }, + { + "epoch": 4.799777220829853, + "grad_norm": 0.022411381379781593, + "learning_rate": 2.7215160802562112e-05, + "loss": 0.0003, + "step": 8620 + }, + { + "epoch": 4.805346700083542, + "grad_norm": 0.027275458282489208, + "learning_rate": 2.7091665548949007e-05, + "loss": 0.0003, + "step": 8630 + }, + { + "epoch": 4.810916179337232, + "grad_norm": 0.023728367905275223, + "learning_rate": 2.696834692377565e-05, + "loss": 0.0002, + "step": 8640 + }, + { + "epoch": 4.8164856585909215, + "grad_norm": 0.020985388088190393, + "learning_rate": 2.684520587785646e-05, + "loss": 0.0002, + "step": 8650 + }, + { + "epoch": 4.822055137844611, + "grad_norm": 0.021320914542750827, + "learning_rate": 2.6722243360636824e-05, + "loss": 0.0002, + "step": 8660 + }, + { + "epoch": 4.8276246170983015, + "grad_norm": 0.01707609645467518, + "learning_rate": 2.659946032018551e-05, + "loss": 0.0002, + "step": 8670 + }, + { + "epoch": 4.833194096351991, + "grad_norm": 0.02272743436595909, + "learning_rate": 2.6476857703187552e-05, + "loss": 0.0002, + "step": 8680 + }, + { + "epoch": 4.838763575605681, + "grad_norm": 0.026188916089218816, + "learning_rate": 2.6354436454936844e-05, + "loss": 0.0003, + "step": 8690 + }, + { + "epoch": 4.844333054859371, + "grad_norm": 0.031215309350085547, + "learning_rate": 2.6232197519328905e-05, + "loss": 0.0003, + "step": 8700 + }, + { + "epoch": 4.849902534113061, + "grad_norm": 0.034589168983563165, + "learning_rate": 2.6110141838853575e-05, + "loss": 0.0003, + "step": 8710 + }, + { + "epoch": 4.85547201336675, + "grad_norm": 0.0281489577450034, + "learning_rate": 2.598827035458777e-05, + "loss": 0.0002, + "step": 8720 + }, + { + "epoch": 4.86104149262044, + "grad_norm": 0.03614573089164002, + "learning_rate": 2.5866584006188145e-05, + "loss": 0.0002, + "step": 8730 + }, + { + "epoch": 4.866610971874129, + "grad_norm": 0.023874343822438974, + "learning_rate": 2.574508373188401e-05, + "loss": 0.0002, + "step": 8740 + }, + { + "epoch": 4.87218045112782, + "grad_norm": 0.026401871822566477, + "learning_rate": 2.5623770468469964e-05, + "loss": 0.0002, + "step": 8750 + }, + { + "epoch": 4.877749930381509, + "grad_norm": 0.019135617952618127, + "learning_rate": 2.5502645151298697e-05, + "loss": 0.0002, + "step": 8760 + }, + { + "epoch": 4.883319409635199, + "grad_norm": 0.026485961984275715, + "learning_rate": 2.538170871427381e-05, + "loss": 0.0002, + "step": 8770 + }, + { + "epoch": 4.888888888888889, + "grad_norm": 0.02497871443840377, + "learning_rate": 2.5260962089842532e-05, + "loss": 0.0002, + "step": 8780 + }, + { + "epoch": 4.894458368142579, + "grad_norm": 0.02564851769279258, + "learning_rate": 2.514040620898872e-05, + "loss": 0.0002, + "step": 8790 + }, + { + "epoch": 4.900027847396268, + "grad_norm": 0.02355372871300342, + "learning_rate": 2.502004200122543e-05, + "loss": 0.0002, + "step": 8800 + }, + { + "epoch": 4.905597326649958, + "grad_norm": 0.022719293776681672, + "learning_rate": 2.4899870394587937e-05, + "loss": 0.0002, + "step": 8810 + }, + { + "epoch": 4.9111668059036475, + "grad_norm": 0.023789832642214247, + "learning_rate": 2.4779892315626514e-05, + "loss": 0.0002, + "step": 8820 + }, + { + "epoch": 4.916736285157338, + "grad_norm": 0.023163412880375815, + "learning_rate": 2.4660108689399282e-05, + "loss": 0.0002, + "step": 8830 + }, + { + "epoch": 4.9223057644110275, + "grad_norm": 0.02039564582191083, + "learning_rate": 2.4540520439465086e-05, + "loss": 0.0002, + "step": 8840 + }, + { + "epoch": 4.927875243664717, + "grad_norm": 0.020586838162390306, + "learning_rate": 2.4421128487876404e-05, + "loss": 0.0002, + "step": 8850 + }, + { + "epoch": 4.9334447229184075, + "grad_norm": 0.025941866252199, + "learning_rate": 2.4301933755172128e-05, + "loss": 0.0002, + "step": 8860 + }, + { + "epoch": 4.939014202172097, + "grad_norm": 0.020521698748630118, + "learning_rate": 2.4182937160370682e-05, + "loss": 0.0002, + "step": 8870 + }, + { + "epoch": 4.944583681425787, + "grad_norm": 0.017620494333411084, + "learning_rate": 2.4064139620962668e-05, + "loss": 0.0002, + "step": 8880 + }, + { + "epoch": 4.950153160679476, + "grad_norm": 0.023124118052035198, + "learning_rate": 2.3945542052903995e-05, + "loss": 0.0002, + "step": 8890 + }, + { + "epoch": 4.955722639933166, + "grad_norm": 0.021163249156264444, + "learning_rate": 2.382714537060874e-05, + "loss": 0.0002, + "step": 8900 + }, + { + "epoch": 4.961292119186856, + "grad_norm": 0.020894802450974588, + "learning_rate": 2.3708950486942107e-05, + "loss": 0.0003, + "step": 8910 + }, + { + "epoch": 4.966861598440546, + "grad_norm": 0.019175010080264482, + "learning_rate": 2.3590958313213367e-05, + "loss": 0.0002, + "step": 8920 + }, + { + "epoch": 4.972431077694235, + "grad_norm": 0.025185626994050473, + "learning_rate": 2.3473169759168877e-05, + "loss": 0.0002, + "step": 8930 + }, + { + "epoch": 4.978000556947926, + "grad_norm": 0.02034678775771328, + "learning_rate": 2.335558573298503e-05, + "loss": 0.0002, + "step": 8940 + }, + { + "epoch": 4.983570036201615, + "grad_norm": 0.022770313086261238, + "learning_rate": 2.323820714126127e-05, + "loss": 0.0002, + "step": 8950 + }, + { + "epoch": 4.989139515455305, + "grad_norm": 0.027755607424792477, + "learning_rate": 2.312103488901304e-05, + "loss": 0.0002, + "step": 8960 + }, + { + "epoch": 4.994708994708994, + "grad_norm": 0.02063457400036376, + "learning_rate": 2.300406987966497e-05, + "loss": 0.0002, + "step": 8970 + }, + { + "epoch": 5.0, + "grad_norm": 0.024874224177501434, + "learning_rate": 2.2887313015043687e-05, + "loss": 0.0002, + "step": 8980 + }, + { + "epoch": 5.00556947925369, + "grad_norm": 0.006217790342697641, + "learning_rate": 2.2770765195371043e-05, + "loss": 0.0001, + "step": 8990 + }, + { + "epoch": 5.011138958507379, + "grad_norm": 0.01587034595166106, + "learning_rate": 2.2654427319257084e-05, + "loss": 0.0001, + "step": 9000 + }, + { + "epoch": 5.01670843776107, + "grad_norm": 0.010849989008813702, + "learning_rate": 2.253830028369314e-05, + "loss": 0.0001, + "step": 9010 + }, + { + "epoch": 5.022277917014759, + "grad_norm": 0.01907741445975058, + "learning_rate": 2.2422384984044935e-05, + "loss": 0.0001, + "step": 9020 + }, + { + "epoch": 5.027847396268449, + "grad_norm": 0.00874202185438713, + "learning_rate": 2.2306682314045653e-05, + "loss": 0.0001, + "step": 9030 + }, + { + "epoch": 5.033416875522139, + "grad_norm": 0.010929981189416935, + "learning_rate": 2.2191193165788986e-05, + "loss": 0.0001, + "step": 9040 + }, + { + "epoch": 5.038986354775829, + "grad_norm": 0.014111398489776139, + "learning_rate": 2.207591842972248e-05, + "loss": 0.0001, + "step": 9050 + }, + { + "epoch": 5.044555834029518, + "grad_norm": 0.008572656876728533, + "learning_rate": 2.1960858994640348e-05, + "loss": 0.0001, + "step": 9060 + }, + { + "epoch": 5.050125313283208, + "grad_norm": 0.011927901272917192, + "learning_rate": 2.184601574767689e-05, + "loss": 0.0001, + "step": 9070 + }, + { + "epoch": 5.055694792536897, + "grad_norm": 0.016702031289223207, + "learning_rate": 2.17313895742995e-05, + "loss": 0.0001, + "step": 9080 + }, + { + "epoch": 5.061264271790588, + "grad_norm": 0.011428971895819283, + "learning_rate": 2.1616981358301907e-05, + "loss": 0.0001, + "step": 9090 + }, + { + "epoch": 5.066833751044277, + "grad_norm": 0.009655702221342169, + "learning_rate": 2.150279198179733e-05, + "loss": 0.0001, + "step": 9100 + }, + { + "epoch": 5.072403230297967, + "grad_norm": 0.018624745737060137, + "learning_rate": 2.1388822325211688e-05, + "loss": 0.0001, + "step": 9110 + }, + { + "epoch": 5.077972709551657, + "grad_norm": 0.007454496312417194, + "learning_rate": 2.1275073267276813e-05, + "loss": 0.0001, + "step": 9120 + }, + { + "epoch": 5.083542188805347, + "grad_norm": 0.01563550577990858, + "learning_rate": 2.116154568502368e-05, + "loss": 0.0001, + "step": 9130 + }, + { + "epoch": 5.0891116680590365, + "grad_norm": 0.00696953615854371, + "learning_rate": 2.1048240453775576e-05, + "loss": 0.0001, + "step": 9140 + }, + { + "epoch": 5.094681147312726, + "grad_norm": 0.013391703050799598, + "learning_rate": 2.0935158447141524e-05, + "loss": 0.0001, + "step": 9150 + }, + { + "epoch": 5.100250626566416, + "grad_norm": 0.01003023054276716, + "learning_rate": 2.082230053700932e-05, + "loss": 0.0001, + "step": 9160 + }, + { + "epoch": 5.105820105820106, + "grad_norm": 0.007199330094277244, + "learning_rate": 2.0709667593538977e-05, + "loss": 0.0001, + "step": 9170 + }, + { + "epoch": 5.111389585073796, + "grad_norm": 0.007386464951870623, + "learning_rate": 2.0597260485156024e-05, + "loss": 0.0001, + "step": 9180 + }, + { + "epoch": 5.116959064327485, + "grad_norm": 0.008860753238961988, + "learning_rate": 2.0485080078544623e-05, + "loss": 0.0001, + "step": 9190 + }, + { + "epoch": 5.122528543581176, + "grad_norm": 0.019006662288993868, + "learning_rate": 2.0373127238641098e-05, + "loss": 0.0001, + "step": 9200 + }, + { + "epoch": 5.128098022834865, + "grad_norm": 0.014577217681286371, + "learning_rate": 2.026140282862718e-05, + "loss": 0.0001, + "step": 9210 + }, + { + "epoch": 5.133667502088555, + "grad_norm": 0.011506364356976575, + "learning_rate": 2.0149907709923322e-05, + "loss": 0.0001, + "step": 9220 + }, + { + "epoch": 5.139236981342244, + "grad_norm": 0.01261004509216635, + "learning_rate": 2.0038642742182116e-05, + "loss": 0.0001, + "step": 9230 + }, + { + "epoch": 5.144806460595934, + "grad_norm": 0.006938738587666887, + "learning_rate": 1.992760878328162e-05, + "loss": 0.0001, + "step": 9240 + }, + { + "epoch": 5.150375939849624, + "grad_norm": 0.013190312017053861, + "learning_rate": 1.981680668931878e-05, + "loss": 0.0001, + "step": 9250 + }, + { + "epoch": 5.155945419103314, + "grad_norm": 0.007551935343800794, + "learning_rate": 1.9706237314602804e-05, + "loss": 0.0001, + "step": 9260 + }, + { + "epoch": 5.161514898357003, + "grad_norm": 0.006315355164397492, + "learning_rate": 1.959590151164854e-05, + "loss": 0.0001, + "step": 9270 + }, + { + "epoch": 5.167084377610694, + "grad_norm": 0.016385981109577168, + "learning_rate": 1.948580013117004e-05, + "loss": 0.0001, + "step": 9280 + }, + { + "epoch": 5.172653856864383, + "grad_norm": 0.006608954495640065, + "learning_rate": 1.9375934022073806e-05, + "loss": 0.0001, + "step": 9290 + }, + { + "epoch": 5.178223336118073, + "grad_norm": 0.007785818700461465, + "learning_rate": 1.9266304031452393e-05, + "loss": 0.0001, + "step": 9300 + }, + { + "epoch": 5.1837928153717625, + "grad_norm": 0.013184729497990542, + "learning_rate": 1.9156911004577837e-05, + "loss": 0.0001, + "step": 9310 + }, + { + "epoch": 5.189362294625452, + "grad_norm": 0.012962757292446296, + "learning_rate": 1.904775578489511e-05, + "loss": 0.0001, + "step": 9320 + }, + { + "epoch": 5.1949317738791425, + "grad_norm": 0.008939438875242749, + "learning_rate": 1.893883921401564e-05, + "loss": 0.0001, + "step": 9330 + }, + { + "epoch": 5.200501253132832, + "grad_norm": 0.007410447099168215, + "learning_rate": 1.883016213171085e-05, + "loss": 0.0001, + "step": 9340 + }, + { + "epoch": 5.206070732386522, + "grad_norm": 0.014467130973199354, + "learning_rate": 1.8721725375905557e-05, + "loss": 0.0001, + "step": 9350 + }, + { + "epoch": 5.211640211640212, + "grad_norm": 0.012830743257718916, + "learning_rate": 1.8613529782671752e-05, + "loss": 0.0001, + "step": 9360 + }, + { + "epoch": 5.217209690893902, + "grad_norm": 0.005920380890793187, + "learning_rate": 1.8505576186221873e-05, + "loss": 0.0001, + "step": 9370 + }, + { + "epoch": 5.222779170147591, + "grad_norm": 0.010838833212016414, + "learning_rate": 1.8397865418902583e-05, + "loss": 0.0001, + "step": 9380 + }, + { + "epoch": 5.228348649401281, + "grad_norm": 0.011949706366452111, + "learning_rate": 1.829039831118826e-05, + "loss": 0.0001, + "step": 9390 + }, + { + "epoch": 5.23391812865497, + "grad_norm": 0.004714570176858442, + "learning_rate": 1.818317569167461e-05, + "loss": 0.0001, + "step": 9400 + }, + { + "epoch": 5.239487607908661, + "grad_norm": 0.009512444111388263, + "learning_rate": 1.8076198387072274e-05, + "loss": 0.0001, + "step": 9410 + }, + { + "epoch": 5.24505708716235, + "grad_norm": 0.008745117469617859, + "learning_rate": 1.796946722220047e-05, + "loss": 0.0001, + "step": 9420 + }, + { + "epoch": 5.25062656641604, + "grad_norm": 0.0077277404511557945, + "learning_rate": 1.7862983019980618e-05, + "loss": 0.0001, + "step": 9430 + }, + { + "epoch": 5.25619604566973, + "grad_norm": 0.0029832537768149313, + "learning_rate": 1.7756746601430013e-05, + "loss": 0.0001, + "step": 9440 + }, + { + "epoch": 5.26176552492342, + "grad_norm": 0.00863960433517615, + "learning_rate": 1.765075878565543e-05, + "loss": 0.0001, + "step": 9450 + }, + { + "epoch": 5.267335004177109, + "grad_norm": 0.0055831379884572685, + "learning_rate": 1.7545020389846955e-05, + "loss": 0.0001, + "step": 9460 + }, + { + "epoch": 5.272904483430799, + "grad_norm": 0.006835526676668753, + "learning_rate": 1.743953222927149e-05, + "loss": 0.0001, + "step": 9470 + }, + { + "epoch": 5.2784739626844885, + "grad_norm": 0.008671394354776962, + "learning_rate": 1.733429511726662e-05, + "loss": 0.0001, + "step": 9480 + }, + { + "epoch": 5.284043441938179, + "grad_norm": 0.012169495460159153, + "learning_rate": 1.722930986523429e-05, + "loss": 0.0001, + "step": 9490 + }, + { + "epoch": 5.2896129211918685, + "grad_norm": 0.006295459956744182, + "learning_rate": 1.7124577282634524e-05, + "loss": 0.0001, + "step": 9500 + }, + { + "epoch": 5.295182400445558, + "grad_norm": 0.006721097993828257, + "learning_rate": 1.7020098176979238e-05, + "loss": 0.0001, + "step": 9510 + }, + { + "epoch": 5.3007518796992485, + "grad_norm": 0.009125509546954735, + "learning_rate": 1.6915873353825972e-05, + "loss": 0.0001, + "step": 9520 + }, + { + "epoch": 5.306321358952938, + "grad_norm": 0.012833766105851099, + "learning_rate": 1.6811903616771634e-05, + "loss": 0.0001, + "step": 9530 + }, + { + "epoch": 5.311890838206628, + "grad_norm": 0.019292866325228763, + "learning_rate": 1.6708189767446487e-05, + "loss": 0.0001, + "step": 9540 + }, + { + "epoch": 5.317460317460317, + "grad_norm": 0.007491944010267753, + "learning_rate": 1.6604732605507707e-05, + "loss": 0.0001, + "step": 9550 + }, + { + "epoch": 5.323029796714007, + "grad_norm": 0.028528157547900085, + "learning_rate": 1.650153292863343e-05, + "loss": 0.0001, + "step": 9560 + }, + { + "epoch": 5.328599275967697, + "grad_norm": 0.014020071488836957, + "learning_rate": 1.63985915325165e-05, + "loss": 0.0001, + "step": 9570 + }, + { + "epoch": 5.334168755221387, + "grad_norm": 0.015842899787616542, + "learning_rate": 1.629590921085837e-05, + "loss": 0.0001, + "step": 9580 + }, + { + "epoch": 5.339738234475076, + "grad_norm": 0.00705419990513781, + "learning_rate": 1.6193486755362968e-05, + "loss": 0.0001, + "step": 9590 + }, + { + "epoch": 5.345307713728767, + "grad_norm": 0.01724263495143505, + "learning_rate": 1.609132495573058e-05, + "loss": 0.0001, + "step": 9600 + }, + { + "epoch": 5.350877192982456, + "grad_norm": 0.01812646832523781, + "learning_rate": 1.5989424599651793e-05, + "loss": 0.0001, + "step": 9610 + }, + { + "epoch": 5.356446672236146, + "grad_norm": 0.014880624430847995, + "learning_rate": 1.5887786472801396e-05, + "loss": 0.0001, + "step": 9620 + }, + { + "epoch": 5.3620161514898355, + "grad_norm": 0.003584074902851557, + "learning_rate": 1.5786411358832337e-05, + "loss": 0.0001, + "step": 9630 + }, + { + "epoch": 5.367585630743525, + "grad_norm": 0.005028556446495052, + "learning_rate": 1.5685300039369665e-05, + "loss": 0.0001, + "step": 9640 + }, + { + "epoch": 5.3731551099972155, + "grad_norm": 0.007112418862433726, + "learning_rate": 1.558445329400454e-05, + "loss": 0.0001, + "step": 9650 + }, + { + "epoch": 5.378724589250905, + "grad_norm": 0.007629712154120982, + "learning_rate": 1.5483871900288138e-05, + "loss": 0.0001, + "step": 9660 + }, + { + "epoch": 5.384294068504595, + "grad_norm": 0.006814983314540237, + "learning_rate": 1.5383556633725808e-05, + "loss": 0.0001, + "step": 9670 + }, + { + "epoch": 5.389863547758285, + "grad_norm": 0.007416247300254645, + "learning_rate": 1.5283508267770895e-05, + "loss": 0.0001, + "step": 9680 + }, + { + "epoch": 5.395433027011975, + "grad_norm": 0.003582939841346356, + "learning_rate": 1.518372757381895e-05, + "loss": 0.0001, + "step": 9690 + }, + { + "epoch": 5.401002506265664, + "grad_norm": 0.007059220269918166, + "learning_rate": 1.5084215321201683e-05, + "loss": 0.0001, + "step": 9700 + }, + { + "epoch": 5.406571985519354, + "grad_norm": 0.010853295468330427, + "learning_rate": 1.498497227718107e-05, + "loss": 0.0001, + "step": 9710 + }, + { + "epoch": 5.412141464773043, + "grad_norm": 0.0078016058229723895, + "learning_rate": 1.4885999206943429e-05, + "loss": 0.0001, + "step": 9720 + }, + { + "epoch": 5.417710944026734, + "grad_norm": 0.009795898595364918, + "learning_rate": 1.4787296873593521e-05, + "loss": 0.0001, + "step": 9730 + }, + { + "epoch": 5.423280423280423, + "grad_norm": 0.013935207176503598, + "learning_rate": 1.468886603814864e-05, + "loss": 0.0001, + "step": 9740 + }, + { + "epoch": 5.428849902534113, + "grad_norm": 0.008574965238175263, + "learning_rate": 1.4590707459532826e-05, + "loss": 0.0001, + "step": 9750 + }, + { + "epoch": 5.434419381787803, + "grad_norm": 0.005470756550457302, + "learning_rate": 1.4492821894570862e-05, + "loss": 0.0001, + "step": 9760 + }, + { + "epoch": 5.439988861041493, + "grad_norm": 0.03229566267255889, + "learning_rate": 1.4395210097982659e-05, + "loss": 0.0001, + "step": 9770 + }, + { + "epoch": 5.445558340295182, + "grad_norm": 0.018241894226829376, + "learning_rate": 1.4297872822377195e-05, + "loss": 0.0001, + "step": 9780 + }, + { + "epoch": 5.451127819548872, + "grad_norm": 0.014306531550890214, + "learning_rate": 1.42008108182469e-05, + "loss": 0.0001, + "step": 9790 + }, + { + "epoch": 5.4566972988025615, + "grad_norm": 0.005305361131643873, + "learning_rate": 1.410402483396177e-05, + "loss": 0.0001, + "step": 9800 + }, + { + "epoch": 5.462266778056252, + "grad_norm": 0.010209083671454882, + "learning_rate": 1.4007515615763639e-05, + "loss": 0.0001, + "step": 9810 + }, + { + "epoch": 5.4678362573099415, + "grad_norm": 0.011061215572991567, + "learning_rate": 1.3911283907760409e-05, + "loss": 0.0001, + "step": 9820 + }, + { + "epoch": 5.473405736563631, + "grad_norm": 0.0125551027838632, + "learning_rate": 1.3815330451920328e-05, + "loss": 0.0001, + "step": 9830 + }, + { + "epoch": 5.4789752158173215, + "grad_norm": 0.008507229324151194, + "learning_rate": 1.3719655988066188e-05, + "loss": 0.0001, + "step": 9840 + }, + { + "epoch": 5.484544695071011, + "grad_norm": 0.00562307569039561, + "learning_rate": 1.3624261253869807e-05, + "loss": 0.0001, + "step": 9850 + }, + { + "epoch": 5.490114174324701, + "grad_norm": 0.0039033663678380475, + "learning_rate": 1.3529146984846126e-05, + "loss": 0.0001, + "step": 9860 + }, + { + "epoch": 5.49568365357839, + "grad_norm": 0.0068028572810532135, + "learning_rate": 1.3434313914347674e-05, + "loss": 0.0001, + "step": 9870 + }, + { + "epoch": 5.50125313283208, + "grad_norm": 0.011208421921777609, + "learning_rate": 1.3339762773558878e-05, + "loss": 0.0001, + "step": 9880 + }, + { + "epoch": 5.50682261208577, + "grad_norm": 0.0034339365806012245, + "learning_rate": 1.3245494291490418e-05, + "loss": 0.0001, + "step": 9890 + }, + { + "epoch": 5.51239209133946, + "grad_norm": 0.011215553294500347, + "learning_rate": 1.3151509194973605e-05, + "loss": 0.0001, + "step": 9900 + }, + { + "epoch": 5.517961570593149, + "grad_norm": 0.005394597086700374, + "learning_rate": 1.3057808208654799e-05, + "loss": 0.0001, + "step": 9910 + }, + { + "epoch": 5.52353104984684, + "grad_norm": 0.004046819257214625, + "learning_rate": 1.2964392054989788e-05, + "loss": 0.0001, + "step": 9920 + }, + { + "epoch": 5.529100529100529, + "grad_norm": 0.004839734905421367, + "learning_rate": 1.2871261454238265e-05, + "loss": 0.0001, + "step": 9930 + }, + { + "epoch": 5.534670008354219, + "grad_norm": 0.005956839775207441, + "learning_rate": 1.277841712445818e-05, + "loss": 0.0001, + "step": 9940 + }, + { + "epoch": 5.540239487607908, + "grad_norm": 0.00536325405970698, + "learning_rate": 1.2685859781500369e-05, + "loss": 0.0001, + "step": 9950 + }, + { + "epoch": 5.545808966861598, + "grad_norm": 0.01820380362336847, + "learning_rate": 1.2593590139002837e-05, + "loss": 0.0001, + "step": 9960 + }, + { + "epoch": 5.551378446115288, + "grad_norm": 0.010096027683705468, + "learning_rate": 1.250160890838541e-05, + "loss": 0.0001, + "step": 9970 + }, + { + "epoch": 5.556947925368978, + "grad_norm": 0.010833360328215708, + "learning_rate": 1.2409916798844184e-05, + "loss": 0.0001, + "step": 9980 + }, + { + "epoch": 5.5625174046226675, + "grad_norm": 0.021372645265980945, + "learning_rate": 1.2318514517346058e-05, + "loss": 0.0001, + "step": 9990 + }, + { + "epoch": 5.568086883876358, + "grad_norm": 0.011023549280459642, + "learning_rate": 1.2227402768623292e-05, + "loss": 0.0001, + "step": 10000 + }, + { + "epoch": 5.5736563631300475, + "grad_norm": 0.012072581317560095, + "learning_rate": 1.2136582255168093e-05, + "loss": 0.0001, + "step": 10010 + }, + { + "epoch": 5.579225842383737, + "grad_norm": 0.011871268710578442, + "learning_rate": 1.2046053677227115e-05, + "loss": 0.0, + "step": 10020 + }, + { + "epoch": 5.584795321637427, + "grad_norm": 0.00389279835256559, + "learning_rate": 1.1955817732796209e-05, + "loss": 0.0001, + "step": 10030 + }, + { + "epoch": 5.590364800891116, + "grad_norm": 0.006322568199496999, + "learning_rate": 1.186587511761491e-05, + "loss": 0.0001, + "step": 10040 + }, + { + "epoch": 5.595934280144807, + "grad_norm": 0.007097833260591808, + "learning_rate": 1.1776226525161115e-05, + "loss": 0.0001, + "step": 10050 + }, + { + "epoch": 5.601503759398496, + "grad_norm": 0.009564826998497755, + "learning_rate": 1.1686872646645763e-05, + "loss": 0.0001, + "step": 10060 + }, + { + "epoch": 5.607073238652186, + "grad_norm": 0.006486553039125118, + "learning_rate": 1.159781417100742e-05, + "loss": 0.0, + "step": 10070 + }, + { + "epoch": 5.612642717905876, + "grad_norm": 0.006726779457844698, + "learning_rate": 1.1509051784907144e-05, + "loss": 0.0, + "step": 10080 + }, + { + "epoch": 5.618212197159566, + "grad_norm": 0.015335009860350034, + "learning_rate": 1.1420586172722953e-05, + "loss": 0.0, + "step": 10090 + }, + { + "epoch": 5.623781676413255, + "grad_norm": 0.0036834158155430122, + "learning_rate": 1.1332418016544743e-05, + "loss": 0.0001, + "step": 10100 + }, + { + "epoch": 5.629351155666945, + "grad_norm": 0.009705253904705452, + "learning_rate": 1.1244547996168936e-05, + "loss": 0.0001, + "step": 10110 + }, + { + "epoch": 5.634920634920634, + "grad_norm": 0.00806610940009477, + "learning_rate": 1.1156976789093266e-05, + "loss": 0.0001, + "step": 10120 + }, + { + "epoch": 5.640490114174325, + "grad_norm": 0.022571131663582007, + "learning_rate": 1.1069705070511543e-05, + "loss": 0.0001, + "step": 10130 + }, + { + "epoch": 5.6460595934280144, + "grad_norm": 0.006173286600948062, + "learning_rate": 1.0982733513308474e-05, + "loss": 0.0001, + "step": 10140 + }, + { + "epoch": 5.651629072681704, + "grad_norm": 0.0036828037455108134, + "learning_rate": 1.089606278805439e-05, + "loss": 0.0, + "step": 10150 + }, + { + "epoch": 5.6571985519353944, + "grad_norm": 0.010480247692275369, + "learning_rate": 1.0809693563000233e-05, + "loss": 0.0001, + "step": 10160 + }, + { + "epoch": 5.662768031189084, + "grad_norm": 0.011703503638958761, + "learning_rate": 1.0723626504072232e-05, + "loss": 0.0001, + "step": 10170 + }, + { + "epoch": 5.668337510442774, + "grad_norm": 0.009697848543839762, + "learning_rate": 1.0637862274866884e-05, + "loss": 0.0001, + "step": 10180 + }, + { + "epoch": 5.673906989696463, + "grad_norm": 0.00630165789251493, + "learning_rate": 1.0552401536645794e-05, + "loss": 0.0, + "step": 10190 + }, + { + "epoch": 5.679476468950153, + "grad_norm": 0.003433101141090687, + "learning_rate": 1.0467244948330584e-05, + "loss": 0.0001, + "step": 10200 + }, + { + "epoch": 5.685045948203843, + "grad_norm": 0.009543863025411714, + "learning_rate": 1.0382393166497812e-05, + "loss": 0.0, + "step": 10210 + }, + { + "epoch": 5.690615427457533, + "grad_norm": 0.015216894875078954, + "learning_rate": 1.0297846845373898e-05, + "loss": 0.0001, + "step": 10220 + }, + { + "epoch": 5.696184906711222, + "grad_norm": 0.002814794063743051, + "learning_rate": 1.0213606636830108e-05, + "loss": 0.0, + "step": 10230 + }, + { + "epoch": 5.701754385964913, + "grad_norm": 0.007305047076188955, + "learning_rate": 1.0129673190377514e-05, + "loss": 0.0001, + "step": 10240 + }, + { + "epoch": 5.707323865218602, + "grad_norm": 0.0036119210667479246, + "learning_rate": 1.0046047153161924e-05, + "loss": 0.0001, + "step": 10250 + }, + { + "epoch": 5.712893344472292, + "grad_norm": 0.011805837280768464, + "learning_rate": 9.962729169959061e-06, + "loss": 0.0001, + "step": 10260 + }, + { + "epoch": 5.718462823725981, + "grad_norm": 0.010292578452244542, + "learning_rate": 9.87971988316937e-06, + "loss": 0.0001, + "step": 10270 + }, + { + "epoch": 5.724032302979671, + "grad_norm": 0.005840046739605118, + "learning_rate": 9.797019932813251e-06, + "loss": 0.0001, + "step": 10280 + }, + { + "epoch": 5.729601782233361, + "grad_norm": 0.018148030771589437, + "learning_rate": 9.71462995652601e-06, + "loss": 0.0001, + "step": 10290 + }, + { + "epoch": 5.735171261487051, + "grad_norm": 0.005467691611123558, + "learning_rate": 9.632550589553014e-06, + "loss": 0.0001, + "step": 10300 + }, + { + "epoch": 5.7407407407407405, + "grad_norm": 0.028316170683896358, + "learning_rate": 9.550782464744734e-06, + "loss": 0.0001, + "step": 10310 + }, + { + "epoch": 5.746310219994431, + "grad_norm": 0.00455545348446607, + "learning_rate": 9.469326212551927e-06, + "loss": 0.0001, + "step": 10320 + }, + { + "epoch": 5.7518796992481205, + "grad_norm": 0.008400577550121191, + "learning_rate": 9.388182461020679e-06, + "loss": 0.0001, + "step": 10330 + }, + { + "epoch": 5.75744917850181, + "grad_norm": 0.006789944321922049, + "learning_rate": 9.307351835787708e-06, + "loss": 0.0, + "step": 10340 + }, + { + "epoch": 5.7630186577555, + "grad_norm": 0.00931971035531967, + "learning_rate": 9.226834960075392e-06, + "loss": 0.0001, + "step": 10350 + }, + { + "epoch": 5.768588137009189, + "grad_norm": 0.004198720386912199, + "learning_rate": 9.14663245468706e-06, + "loss": 0.0, + "step": 10360 + }, + { + "epoch": 5.77415761626288, + "grad_norm": 0.004794256467587593, + "learning_rate": 9.066744938002164e-06, + "loss": 0.0, + "step": 10370 + }, + { + "epoch": 5.779727095516569, + "grad_norm": 0.001972932329993146, + "learning_rate": 8.98717302597153e-06, + "loss": 0.0, + "step": 10380 + }, + { + "epoch": 5.785296574770259, + "grad_norm": 0.008348607392636058, + "learning_rate": 8.907917332112592e-06, + "loss": 0.0, + "step": 10390 + }, + { + "epoch": 5.790866054023949, + "grad_norm": 0.0014946710068592596, + "learning_rate": 8.828978467504678e-06, + "loss": 0.0, + "step": 10400 + }, + { + "epoch": 5.796435533277639, + "grad_norm": 0.00532602836028796, + "learning_rate": 8.750357040784285e-06, + "loss": 0.0, + "step": 10410 + }, + { + "epoch": 5.802005012531328, + "grad_norm": 0.0022713123081523623, + "learning_rate": 8.672053658140412e-06, + "loss": 0.0001, + "step": 10420 + }, + { + "epoch": 5.807574491785018, + "grad_norm": 0.006563055427858778, + "learning_rate": 8.594068923309806e-06, + "loss": 0.0, + "step": 10430 + }, + { + "epoch": 5.813143971038707, + "grad_norm": 0.003583751004845842, + "learning_rate": 8.516403437572462e-06, + "loss": 0.0, + "step": 10440 + }, + { + "epoch": 5.818713450292398, + "grad_norm": 0.008445449853703161, + "learning_rate": 8.439057799746807e-06, + "loss": 0.0001, + "step": 10450 + }, + { + "epoch": 5.824282929546087, + "grad_norm": 0.002697743640737849, + "learning_rate": 8.362032606185193e-06, + "loss": 0.0001, + "step": 10460 + }, + { + "epoch": 5.829852408799777, + "grad_norm": 0.003853963648908576, + "learning_rate": 8.28532845076933e-06, + "loss": 0.0001, + "step": 10470 + }, + { + "epoch": 5.835421888053467, + "grad_norm": 0.010896757966762134, + "learning_rate": 8.208945924905558e-06, + "loss": 0.0, + "step": 10480 + }, + { + "epoch": 5.840991367307157, + "grad_norm": 0.005106282544741418, + "learning_rate": 8.132885617520441e-06, + "loss": 0.0, + "step": 10490 + }, + { + "epoch": 5.8465608465608465, + "grad_norm": 0.0026255497556767937, + "learning_rate": 8.057148115056157e-06, + "loss": 0.0, + "step": 10500 + }, + { + "epoch": 5.852130325814536, + "grad_norm": 0.013783424438370625, + "learning_rate": 7.981734001465973e-06, + "loss": 0.0001, + "step": 10510 + }, + { + "epoch": 5.857699805068226, + "grad_norm": 0.01533270461811719, + "learning_rate": 7.906643858209766e-06, + "loss": 0.0001, + "step": 10520 + }, + { + "epoch": 5.863269284321916, + "grad_norm": 0.00894043569194934, + "learning_rate": 7.831878264249521e-06, + "loss": 0.0, + "step": 10530 + }, + { + "epoch": 5.868838763575606, + "grad_norm": 0.001946262608940776, + "learning_rate": 7.757437796044875e-06, + "loss": 0.0, + "step": 10540 + }, + { + "epoch": 5.874408242829295, + "grad_norm": 0.0070301440562824296, + "learning_rate": 7.683323027548684e-06, + "loss": 0.0001, + "step": 10550 + }, + { + "epoch": 5.879977722082986, + "grad_norm": 0.01227565916392796, + "learning_rate": 7.6095345302025445e-06, + "loss": 0.0001, + "step": 10560 + }, + { + "epoch": 5.885547201336675, + "grad_norm": 0.010534715518330637, + "learning_rate": 7.536072872932509e-06, + "loss": 0.0, + "step": 10570 + }, + { + "epoch": 5.891116680590365, + "grad_norm": 0.003767113610863153, + "learning_rate": 7.462938622144527e-06, + "loss": 0.0, + "step": 10580 + }, + { + "epoch": 5.896686159844054, + "grad_norm": 0.0070938921600654055, + "learning_rate": 7.39013234172023e-06, + "loss": 0.0, + "step": 10590 + }, + { + "epoch": 5.902255639097744, + "grad_norm": 0.006056856849520333, + "learning_rate": 7.3176545930125015e-06, + "loss": 0.0, + "step": 10600 + }, + { + "epoch": 5.907825118351434, + "grad_norm": 0.003816825669340532, + "learning_rate": 7.245505934841179e-06, + "loss": 0.0, + "step": 10610 + }, + { + "epoch": 5.913394597605124, + "grad_norm": 0.015338319047506191, + "learning_rate": 7.173686923488737e-06, + "loss": 0.0, + "step": 10620 + }, + { + "epoch": 5.918964076858813, + "grad_norm": 0.008738663400299647, + "learning_rate": 7.102198112696007e-06, + "loss": 0.0, + "step": 10630 + }, + { + "epoch": 5.924533556112504, + "grad_norm": 0.011688967121241855, + "learning_rate": 7.03104005365785e-06, + "loss": 0.0, + "step": 10640 + }, + { + "epoch": 5.930103035366193, + "grad_norm": 0.0031254926205588596, + "learning_rate": 6.960213295019042e-06, + "loss": 0.0, + "step": 10650 + }, + { + "epoch": 5.935672514619883, + "grad_norm": 0.003923121653662076, + "learning_rate": 6.889718382869875e-06, + "loss": 0.0001, + "step": 10660 + }, + { + "epoch": 5.9412419938735725, + "grad_norm": 0.007112825979153977, + "learning_rate": 6.8195558607420825e-06, + "loss": 0.0, + "step": 10670 + }, + { + "epoch": 5.946811473127262, + "grad_norm": 0.010401217739414471, + "learning_rate": 6.7497262696045785e-06, + "loss": 0.0, + "step": 10680 + }, + { + "epoch": 5.9523809523809526, + "grad_norm": 0.010323543610143683, + "learning_rate": 6.680230147859296e-06, + "loss": 0.0, + "step": 10690 + }, + { + "epoch": 5.957950431634642, + "grad_norm": 0.010217459202813754, + "learning_rate": 6.611068031337053e-06, + "loss": 0.0, + "step": 10700 + }, + { + "epoch": 5.963519910888332, + "grad_norm": 0.0038734832759616156, + "learning_rate": 6.542240453293419e-06, + "loss": 0.0, + "step": 10710 + }, + { + "epoch": 5.969089390142022, + "grad_norm": 0.0037776146881943664, + "learning_rate": 6.473747944404573e-06, + "loss": 0.0, + "step": 10720 + }, + { + "epoch": 5.974658869395712, + "grad_norm": 0.015250229517906553, + "learning_rate": 6.405591032763275e-06, + "loss": 0.0, + "step": 10730 + }, + { + "epoch": 5.980228348649401, + "grad_norm": 0.002884991315515665, + "learning_rate": 6.3377702438746855e-06, + "loss": 0.0, + "step": 10740 + }, + { + "epoch": 5.985797827903091, + "grad_norm": 0.014068109537677587, + "learning_rate": 6.270286100652478e-06, + "loss": 0.0001, + "step": 10750 + }, + { + "epoch": 5.99136730715678, + "grad_norm": 0.002936398892475942, + "learning_rate": 6.203139123414614e-06, + "loss": 0.0, + "step": 10760 + }, + { + "epoch": 5.996936786410471, + "grad_norm": 0.004607219704432529, + "learning_rate": 6.136329829879494e-06, + "loss": 0.0, + "step": 10770 + }, + { + "epoch": 6.0022277917014755, + "grad_norm": 0.002103022463964109, + "learning_rate": 6.069858735161882e-06, + "loss": 0.0, + "step": 10780 + }, + { + "epoch": 6.007797270955166, + "grad_norm": 0.0017200499995359958, + "learning_rate": 6.003726351768951e-06, + "loss": 0.0, + "step": 10790 + }, + { + "epoch": 6.0133667502088555, + "grad_norm": 0.002087103348685246, + "learning_rate": 5.9379331895963344e-06, + "loss": 0.0, + "step": 10800 + }, + { + "epoch": 6.018936229462545, + "grad_norm": 0.006396801828766977, + "learning_rate": 5.872479755924193e-06, + "loss": 0.0, + "step": 10810 + }, + { + "epoch": 6.0245057087162355, + "grad_norm": 0.0019075753802109867, + "learning_rate": 5.807366555413285e-06, + "loss": 0.0, + "step": 10820 + }, + { + "epoch": 6.030075187969925, + "grad_norm": 0.004212997005868181, + "learning_rate": 5.7425940901011555e-06, + "loss": 0.0, + "step": 10830 + }, + { + "epoch": 6.035644667223615, + "grad_norm": 0.0008100040152319433, + "learning_rate": 5.6781628593981065e-06, + "loss": 0.0, + "step": 10840 + }, + { + "epoch": 6.041214146477304, + "grad_norm": 0.0013001845088641201, + "learning_rate": 5.614073360083555e-06, + "loss": 0.0, + "step": 10850 + }, + { + "epoch": 6.046783625730994, + "grad_norm": 0.008330448712097887, + "learning_rate": 5.5503260863019935e-06, + "loss": 0.0, + "step": 10860 + }, + { + "epoch": 6.052353104984684, + "grad_norm": 0.0014126871355353272, + "learning_rate": 5.486921529559336e-06, + "loss": 0.0, + "step": 10870 + }, + { + "epoch": 6.057922584238374, + "grad_norm": 0.001401879434651411, + "learning_rate": 5.423860178719053e-06, + "loss": 0.0, + "step": 10880 + }, + { + "epoch": 6.063492063492063, + "grad_norm": 0.0017927032512884602, + "learning_rate": 5.36114251999843e-06, + "loss": 0.0, + "step": 10890 + }, + { + "epoch": 6.069061542745754, + "grad_norm": 0.0026897329777931574, + "learning_rate": 5.298769036964785e-06, + "loss": 0.0, + "step": 10900 + }, + { + "epoch": 6.074631021999443, + "grad_norm": 0.0011654098698293208, + "learning_rate": 5.236740210531793e-06, + "loss": 0.0, + "step": 10910 + }, + { + "epoch": 6.080200501253133, + "grad_norm": 0.003092681279432609, + "learning_rate": 5.175056518955729e-06, + "loss": 0.0, + "step": 10920 + }, + { + "epoch": 6.085769980506822, + "grad_norm": 0.001110883446541133, + "learning_rate": 5.1137184378318e-06, + "loss": 0.0, + "step": 10930 + }, + { + "epoch": 6.091339459760512, + "grad_norm": 0.0016428020744309274, + "learning_rate": 5.052726440090505e-06, + "loss": 0.0, + "step": 10940 + }, + { + "epoch": 6.096908939014202, + "grad_norm": 0.0009879546552532135, + "learning_rate": 4.992080995993903e-06, + "loss": 0.0, + "step": 10950 + }, + { + "epoch": 6.102478418267892, + "grad_norm": 0.0016778464504424726, + "learning_rate": 4.931782573132132e-06, + "loss": 0.0, + "step": 10960 + }, + { + "epoch": 6.1080478975215815, + "grad_norm": 0.006996404296772489, + "learning_rate": 4.8718316364196325e-06, + "loss": 0.0, + "step": 10970 + }, + { + "epoch": 6.113617376775272, + "grad_norm": 0.001177788633420527, + "learning_rate": 4.812228648091693e-06, + "loss": 0.0, + "step": 10980 + }, + { + "epoch": 6.1191868560289615, + "grad_norm": 0.001263652656721551, + "learning_rate": 4.752974067700844e-06, + "loss": 0.0, + "step": 10990 + }, + { + "epoch": 6.124756335282651, + "grad_norm": 0.0011986872469284266, + "learning_rate": 4.69406835211329e-06, + "loss": 0.0, + "step": 11000 + }, + { + "epoch": 6.130325814536341, + "grad_norm": 0.0011193612960442563, + "learning_rate": 4.635511955505434e-06, + "loss": 0.0, + "step": 11010 + }, + { + "epoch": 6.13589529379003, + "grad_norm": 0.0013739274404055296, + "learning_rate": 4.5773053293603265e-06, + "loss": 0.0, + "step": 11020 + }, + { + "epoch": 6.141464773043721, + "grad_norm": 0.003660756977781613, + "learning_rate": 4.51944892246422e-06, + "loss": 0.0, + "step": 11030 + }, + { + "epoch": 6.14703425229741, + "grad_norm": 0.0014222825756073213, + "learning_rate": 4.461943180903117e-06, + "loss": 0.0, + "step": 11040 + }, + { + "epoch": 6.1526037315511, + "grad_norm": 0.0010445538964330798, + "learning_rate": 4.404788548059247e-06, + "loss": 0.0, + "step": 11050 + }, + { + "epoch": 6.15817321080479, + "grad_norm": 0.0013420796720466936, + "learning_rate": 4.347985464607807e-06, + "loss": 0.0, + "step": 11060 + }, + { + "epoch": 6.16374269005848, + "grad_norm": 0.0006265644860625889, + "learning_rate": 4.291534368513378e-06, + "loss": 0.0, + "step": 11070 + }, + { + "epoch": 6.169312169312169, + "grad_norm": 0.00782394096255494, + "learning_rate": 4.235435695026696e-06, + "loss": 0.0, + "step": 11080 + }, + { + "epoch": 6.174881648565859, + "grad_norm": 0.0011443578416120786, + "learning_rate": 4.1796898766812196e-06, + "loss": 0.0, + "step": 11090 + }, + { + "epoch": 6.180451127819548, + "grad_norm": 0.0007741754839499156, + "learning_rate": 4.124297343289812e-06, + "loss": 0.0, + "step": 11100 + }, + { + "epoch": 6.186020607073239, + "grad_norm": 0.000988381754417583, + "learning_rate": 4.069258521941438e-06, + "loss": 0.0, + "step": 11110 + }, + { + "epoch": 6.191590086326928, + "grad_norm": 0.00102596590385583, + "learning_rate": 4.014573836997865e-06, + "loss": 0.0, + "step": 11120 + }, + { + "epoch": 6.197159565580618, + "grad_norm": 0.0015072310088745937, + "learning_rate": 3.96024371009035e-06, + "loss": 0.0, + "step": 11130 + }, + { + "epoch": 6.202729044834308, + "grad_norm": 0.007788209754665941, + "learning_rate": 3.906268560116505e-06, + "loss": 0.0, + "step": 11140 + }, + { + "epoch": 6.208298524087998, + "grad_norm": 0.0006798782485147815, + "learning_rate": 3.852648803236914e-06, + "loss": 0.0, + "step": 11150 + }, + { + "epoch": 6.2138680033416875, + "grad_norm": 0.0017456458735800276, + "learning_rate": 3.7993848528720345e-06, + "loss": 0.0, + "step": 11160 + }, + { + "epoch": 6.219437482595377, + "grad_norm": 0.001822298385632042, + "learning_rate": 3.7464771196989713e-06, + "loss": 0.0, + "step": 11170 + }, + { + "epoch": 6.225006961849067, + "grad_norm": 0.0019579709911475424, + "learning_rate": 3.693926011648313e-06, + "loss": 0.0, + "step": 11180 + }, + { + "epoch": 6.230576441102757, + "grad_norm": 0.0010500885150315038, + "learning_rate": 3.6417319339009935e-06, + "loss": 0.0, + "step": 11190 + }, + { + "epoch": 6.236145920356447, + "grad_norm": 0.0025425005631188064, + "learning_rate": 3.5898952888851477e-06, + "loss": 0.0, + "step": 11200 + }, + { + "epoch": 6.241715399610136, + "grad_norm": 0.001034735251771114, + "learning_rate": 3.538416476273043e-06, + "loss": 0.0, + "step": 11210 + }, + { + "epoch": 6.247284878863827, + "grad_norm": 0.0009976146389985817, + "learning_rate": 3.487295892977965e-06, + "loss": 0.0, + "step": 11220 + }, + { + "epoch": 6.252854358117516, + "grad_norm": 0.004592652832049199, + "learning_rate": 3.4365339331511537e-06, + "loss": 0.0, + "step": 11230 + }, + { + "epoch": 6.258423837371206, + "grad_norm": 0.0015506711303966776, + "learning_rate": 3.3861309881788318e-06, + "loss": 0.0, + "step": 11240 + }, + { + "epoch": 6.263993316624895, + "grad_norm": 0.002456548964461565, + "learning_rate": 3.3360874466790813e-06, + "loss": 0.0, + "step": 11250 + }, + { + "epoch": 6.269562795878585, + "grad_norm": 0.0012107900283020697, + "learning_rate": 3.2864036944989296e-06, + "loss": 0.0, + "step": 11260 + }, + { + "epoch": 6.275132275132275, + "grad_norm": 0.0009560650139291303, + "learning_rate": 3.2370801147113492e-06, + "loss": 0.0, + "step": 11270 + }, + { + "epoch": 6.280701754385965, + "grad_norm": 0.0006742048590846136, + "learning_rate": 3.1881170876122857e-06, + "loss": 0.0, + "step": 11280 + }, + { + "epoch": 6.2862712336396545, + "grad_norm": 0.0013959014317348106, + "learning_rate": 3.1395149907177623e-06, + "loss": 0.0, + "step": 11290 + }, + { + "epoch": 6.291840712893345, + "grad_norm": 0.0020273196071930633, + "learning_rate": 3.091274198760946e-06, + "loss": 0.0, + "step": 11300 + }, + { + "epoch": 6.2974101921470345, + "grad_norm": 0.0009956372448673046, + "learning_rate": 3.043395083689221e-06, + "loss": 0.0, + "step": 11310 + }, + { + "epoch": 6.302979671400724, + "grad_norm": 0.0009038637019344187, + "learning_rate": 2.995878014661424e-06, + "loss": 0.0, + "step": 11320 + }, + { + "epoch": 6.308549150654414, + "grad_norm": 0.0006001476738215733, + "learning_rate": 2.948723358044886e-06, + "loss": 0.0, + "step": 11330 + }, + { + "epoch": 6.314118629908103, + "grad_norm": 0.0030378201837942672, + "learning_rate": 2.9019314774126836e-06, + "loss": 0.0, + "step": 11340 + }, + { + "epoch": 6.319688109161794, + "grad_norm": 0.0009344175301111891, + "learning_rate": 2.8555027335408103e-06, + "loss": 0.0, + "step": 11350 + }, + { + "epoch": 6.325257588415483, + "grad_norm": 0.000960089468070245, + "learning_rate": 2.8094374844053474e-06, + "loss": 0.0, + "step": 11360 + }, + { + "epoch": 6.330827067669173, + "grad_norm": 0.0010066582006857097, + "learning_rate": 2.7637360851798245e-06, + "loss": 0.0, + "step": 11370 + }, + { + "epoch": 6.336396546922863, + "grad_norm": 0.0008342859922061077, + "learning_rate": 2.7183988882323376e-06, + "loss": 0.0, + "step": 11380 + }, + { + "epoch": 6.341966026176553, + "grad_norm": 0.0019577388481825725, + "learning_rate": 2.6734262431229397e-06, + "loss": 0.0, + "step": 11390 + }, + { + "epoch": 6.347535505430242, + "grad_norm": 0.0013340387179093401, + "learning_rate": 2.628818496600893e-06, + "loss": 0.0, + "step": 11400 + }, + { + "epoch": 6.353104984683932, + "grad_norm": 0.0007239338182924446, + "learning_rate": 2.5845759926020176e-06, + "loss": 0.0, + "step": 11410 + }, + { + "epoch": 6.358674463937621, + "grad_norm": 0.0024160060184378906, + "learning_rate": 2.5406990722460166e-06, + "loss": 0.0, + "step": 11420 + }, + { + "epoch": 6.364243943191312, + "grad_norm": 0.0007271374794436343, + "learning_rate": 2.497188073833884e-06, + "loss": 0.0, + "step": 11430 + }, + { + "epoch": 6.369813422445001, + "grad_norm": 0.0009884290737217547, + "learning_rate": 2.4540433328452294e-06, + "loss": 0.0, + "step": 11440 + }, + { + "epoch": 6.375382901698691, + "grad_norm": 0.0006659894069069587, + "learning_rate": 2.411265181935801e-06, + "loss": 0.0, + "step": 11450 + }, + { + "epoch": 6.380952380952381, + "grad_norm": 0.0008445874003379258, + "learning_rate": 2.3688539509347906e-06, + "loss": 0.0, + "step": 11460 + }, + { + "epoch": 6.386521860206071, + "grad_norm": 0.0019510980368620506, + "learning_rate": 2.3268099668423915e-06, + "loss": 0.0, + "step": 11470 + }, + { + "epoch": 6.3920913394597605, + "grad_norm": 0.0013178760111201844, + "learning_rate": 2.285133553827229e-06, + "loss": 0.0, + "step": 11480 + }, + { + "epoch": 6.39766081871345, + "grad_norm": 0.0017317732963746104, + "learning_rate": 2.2438250332238675e-06, + "loss": 0.0, + "step": 11490 + }, + { + "epoch": 6.40323029796714, + "grad_norm": 0.0008059270007033118, + "learning_rate": 2.2028847235303518e-06, + "loss": 0.0, + "step": 11500 + }, + { + "epoch": 6.40879977722083, + "grad_norm": 0.006832522841772991, + "learning_rate": 2.1623129404057254e-06, + "loss": 0.0, + "step": 11510 + }, + { + "epoch": 6.41436925647452, + "grad_norm": 0.0012547090462844083, + "learning_rate": 2.122109996667615e-06, + "loss": 0.0, + "step": 11520 + }, + { + "epoch": 6.419938735728209, + "grad_norm": 0.0005842039545480173, + "learning_rate": 2.0822762022898066e-06, + "loss": 0.0, + "step": 11530 + }, + { + "epoch": 6.4255082149819, + "grad_norm": 0.006382270173032574, + "learning_rate": 2.042811864399857e-06, + "loss": 0.0, + "step": 11540 + }, + { + "epoch": 6.431077694235589, + "grad_norm": 0.0010172233345372248, + "learning_rate": 2.0037172872767516e-06, + "loss": 0.0, + "step": 11550 + }, + { + "epoch": 6.436647173489279, + "grad_norm": 0.0014075685676727566, + "learning_rate": 1.9649927723485116e-06, + "loss": 0.0, + "step": 11560 + }, + { + "epoch": 6.442216652742968, + "grad_norm": 0.0007951596123314817, + "learning_rate": 1.9266386181899078e-06, + "loss": 0.0, + "step": 11570 + }, + { + "epoch": 6.447786131996658, + "grad_norm": 0.0011799481715930665, + "learning_rate": 1.8886551205201441e-06, + "loss": 0.0, + "step": 11580 + }, + { + "epoch": 6.453355611250348, + "grad_norm": 0.0014513611701954627, + "learning_rate": 1.8510425722005732e-06, + "loss": 0.0, + "step": 11590 + }, + { + "epoch": 6.458925090504038, + "grad_norm": 0.0008342680637432075, + "learning_rate": 1.8138012632324563e-06, + "loss": 0.0, + "step": 11600 + }, + { + "epoch": 6.464494569757727, + "grad_norm": 0.001101722433406307, + "learning_rate": 1.7769314807547122e-06, + "loss": 0.0, + "step": 11610 + }, + { + "epoch": 6.470064049011418, + "grad_norm": 0.001937765218000163, + "learning_rate": 1.7404335090416835e-06, + "loss": 0.0, + "step": 11620 + }, + { + "epoch": 6.475633528265107, + "grad_norm": 0.0005994964745001434, + "learning_rate": 1.7043076295010064e-06, + "loss": 0.0, + "step": 11630 + }, + { + "epoch": 6.481203007518797, + "grad_norm": 0.001012072077127105, + "learning_rate": 1.6685541206713672e-06, + "loss": 0.0, + "step": 11640 + }, + { + "epoch": 6.4867724867724865, + "grad_norm": 0.0006031308599773534, + "learning_rate": 1.6331732582204218e-06, + "loss": 0.0, + "step": 11650 + }, + { + "epoch": 6.492341966026176, + "grad_norm": 0.0009166598342177924, + "learning_rate": 1.5981653149426123e-06, + "loss": 0.0, + "step": 11660 + }, + { + "epoch": 6.4979114452798665, + "grad_norm": 0.0008140978333758799, + "learning_rate": 1.5635305607570982e-06, + "loss": 0.0, + "step": 11670 + }, + { + "epoch": 6.503480924533556, + "grad_norm": 0.0006185110565113945, + "learning_rate": 1.5292692627056682e-06, + "loss": 0.0, + "step": 11680 + }, + { + "epoch": 6.509050403787246, + "grad_norm": 0.0014897169838815466, + "learning_rate": 1.4953816849506752e-06, + "loss": 0.0, + "step": 11690 + }, + { + "epoch": 6.514619883040936, + "grad_norm": 0.0009653017687475486, + "learning_rate": 1.4618680887730217e-06, + "loss": 0.0, + "step": 11700 + }, + { + "epoch": 6.520189362294626, + "grad_norm": 0.0017670732912596749, + "learning_rate": 1.4287287325701004e-06, + "loss": 0.0, + "step": 11710 + }, + { + "epoch": 6.525758841548315, + "grad_norm": 0.0011959954105058694, + "learning_rate": 1.3959638718538448e-06, + "loss": 0.0, + "step": 11720 + }, + { + "epoch": 6.531328320802005, + "grad_norm": 0.0016680580465719568, + "learning_rate": 1.3635737592487485e-06, + "loss": 0.0, + "step": 11730 + }, + { + "epoch": 6.536897800055694, + "grad_norm": 0.0007104137534095261, + "learning_rate": 1.3315586444898997e-06, + "loss": 0.0, + "step": 11740 + }, + { + "epoch": 6.542467279309385, + "grad_norm": 0.0010590387401472434, + "learning_rate": 1.2999187744210717e-06, + "loss": 0.0, + "step": 11750 + }, + { + "epoch": 6.548036758563074, + "grad_norm": 0.0010055718756446727, + "learning_rate": 1.2686543929928185e-06, + "loss": 0.0, + "step": 11760 + }, + { + "epoch": 6.553606237816764, + "grad_norm": 0.0015052946778503146, + "learning_rate": 1.2377657412605992e-06, + "loss": 0.0, + "step": 11770 + }, + { + "epoch": 6.559175717070454, + "grad_norm": 0.0019839250529317606, + "learning_rate": 1.207253057382901e-06, + "loss": 0.0, + "step": 11780 + }, + { + "epoch": 6.564745196324144, + "grad_norm": 0.0005096645801104724, + "learning_rate": 1.1771165766194137e-06, + "loss": 0.0, + "step": 11790 + }, + { + "epoch": 6.570314675577833, + "grad_norm": 0.0008023152879175682, + "learning_rate": 1.1473565313292245e-06, + "loss": 0.0, + "step": 11800 + }, + { + "epoch": 6.575884154831523, + "grad_norm": 0.0009524114032008881, + "learning_rate": 1.1179731509690094e-06, + "loss": 0.0, + "step": 11810 + }, + { + "epoch": 6.581453634085213, + "grad_norm": 0.0007945440990455872, + "learning_rate": 1.0889666620912842e-06, + "loss": 0.0, + "step": 11820 + }, + { + "epoch": 6.587023113338903, + "grad_norm": 0.0010644232531016574, + "learning_rate": 1.0603372883426278e-06, + "loss": 0.0, + "step": 11830 + }, + { + "epoch": 6.592592592592593, + "grad_norm": 0.0014538201886809016, + "learning_rate": 1.0320852504620005e-06, + "loss": 0.0, + "step": 11840 + }, + { + "epoch": 6.598162071846282, + "grad_norm": 0.0010041965913668962, + "learning_rate": 1.0042107662789846e-06, + "loss": 0.0, + "step": 11850 + }, + { + "epoch": 6.603731551099973, + "grad_norm": 0.0007290946306888641, + "learning_rate": 9.767140507121741e-07, + "loss": 0.0, + "step": 11860 + }, + { + "epoch": 6.609301030353662, + "grad_norm": 0.0007955920822686162, + "learning_rate": 9.495953157674597e-07, + "loss": 0.0, + "step": 11870 + }, + { + "epoch": 6.614870509607352, + "grad_norm": 0.0008559162733761905, + "learning_rate": 9.228547705364188e-07, + "loss": 0.0, + "step": 11880 + }, + { + "epoch": 6.620439988861041, + "grad_norm": 0.0008993963913795823, + "learning_rate": 8.964926211947056e-07, + "loss": 0.0, + "step": 11890 + }, + { + "epoch": 6.626009468114731, + "grad_norm": 0.001544982390285978, + "learning_rate": 8.705090710004527e-07, + "loss": 0.0, + "step": 11900 + }, + { + "epoch": 6.631578947368421, + "grad_norm": 0.001934875919969642, + "learning_rate": 8.449043202927109e-07, + "loss": 0.0, + "step": 11910 + }, + { + "epoch": 6.637148426622111, + "grad_norm": 0.0009915224445657509, + "learning_rate": 8.196785664898954e-07, + "loss": 0.0, + "step": 11920 + }, + { + "epoch": 6.6427179058758, + "grad_norm": 0.001209406009171057, + "learning_rate": 7.948320040882751e-07, + "loss": 0.0, + "step": 11930 + }, + { + "epoch": 6.648287385129491, + "grad_norm": 0.0011556284519519885, + "learning_rate": 7.703648246604689e-07, + "loss": 0.0, + "step": 11940 + }, + { + "epoch": 6.65385686438318, + "grad_norm": 0.0006428727223080637, + "learning_rate": 7.462772168539578e-07, + "loss": 0.0, + "step": 11950 + }, + { + "epoch": 6.65942634363687, + "grad_norm": 0.0011911306593847476, + "learning_rate": 7.225693663896471e-07, + "loss": 0.0, + "step": 11960 + }, + { + "epoch": 6.6649958228905595, + "grad_norm": 0.000696057651409639, + "learning_rate": 6.992414560604232e-07, + "loss": 0.0, + "step": 11970 + }, + { + "epoch": 6.670565302144249, + "grad_norm": 0.0010214086618505706, + "learning_rate": 6.762936657297603e-07, + "loss": 0.0, + "step": 11980 + }, + { + "epoch": 6.6761347813979395, + "grad_norm": 0.0005986547419041449, + "learning_rate": 6.537261723303046e-07, + "loss": 0.0, + "step": 11990 + }, + { + "epoch": 6.681704260651629, + "grad_norm": 0.0010176250762334801, + "learning_rate": 6.315391498625367e-07, + "loss": 0.0, + "step": 12000 + }, + { + "epoch": 6.687273739905319, + "grad_norm": 0.0010742503511750707, + "learning_rate": 6.097327693934174e-07, + "loss": 0.0, + "step": 12010 + }, + { + "epoch": 6.692843219159009, + "grad_norm": 0.001137439150217556, + "learning_rate": 5.883071990550771e-07, + "loss": 0.0, + "step": 12020 + }, + { + "epoch": 6.698412698412699, + "grad_norm": 0.0012496457588128294, + "learning_rate": 5.672626040434948e-07, + "loss": 0.0, + "step": 12030 + }, + { + "epoch": 6.703982177666388, + "grad_norm": 0.0021635192230207816, + "learning_rate": 5.465991466172715e-07, + "loss": 0.0, + "step": 12040 + }, + { + "epoch": 6.709551656920078, + "grad_norm": 0.0010427186033879598, + "learning_rate": 5.263169860963257e-07, + "loss": 0.0, + "step": 12050 + }, + { + "epoch": 6.715121136173767, + "grad_norm": 0.0011765223738410928, + "learning_rate": 5.064162788607107e-07, + "loss": 0.0, + "step": 12060 + }, + { + "epoch": 6.720690615427458, + "grad_norm": 0.0012225958614020492, + "learning_rate": 4.868971783493825e-07, + "loss": 0.0, + "step": 12070 + }, + { + "epoch": 6.726260094681147, + "grad_norm": 0.0008472383426439395, + "learning_rate": 4.6775983505901176e-07, + "loss": 0.0, + "step": 12080 + }, + { + "epoch": 6.731829573934837, + "grad_norm": 0.0010663756839943985, + "learning_rate": 4.490043965428514e-07, + "loss": 0.0, + "step": 12090 + }, + { + "epoch": 6.737399053188527, + "grad_norm": 0.0009272759327855216, + "learning_rate": 4.306310074095654e-07, + "loss": 0.0, + "step": 12100 + }, + { + "epoch": 6.742968532442217, + "grad_norm": 0.010308593389942753, + "learning_rate": 4.126398093221462e-07, + "loss": 0.0, + "step": 12110 + }, + { + "epoch": 6.748538011695906, + "grad_norm": 0.014728749597270252, + "learning_rate": 3.950309409968045e-07, + "loss": 0.0, + "step": 12120 + }, + { + "epoch": 6.754107490949596, + "grad_norm": 0.0010040175580927613, + "learning_rate": 3.778045382018813e-07, + "loss": 0.0, + "step": 12130 + }, + { + "epoch": 6.7596769702032855, + "grad_norm": 0.000793603650382175, + "learning_rate": 3.6096073375685436e-07, + "loss": 0.0, + "step": 12140 + }, + { + "epoch": 6.765246449456976, + "grad_norm": 0.0010132203137523063, + "learning_rate": 3.444996575312609e-07, + "loss": 0.0, + "step": 12150 + }, + { + "epoch": 6.7708159287106655, + "grad_norm": 0.0005485842362797631, + "learning_rate": 3.284214364437155e-07, + "loss": 0.0, + "step": 12160 + }, + { + "epoch": 6.776385407964355, + "grad_norm": 0.000887270641868201, + "learning_rate": 3.1272619446094943e-07, + "loss": 0.0, + "step": 12170 + }, + { + "epoch": 6.7819548872180455, + "grad_norm": 0.0011737905932534352, + "learning_rate": 2.974140525968172e-07, + "loss": 0.0, + "step": 12180 + }, + { + "epoch": 6.787524366471735, + "grad_norm": 0.0008072319894312352, + "learning_rate": 2.824851289114028e-07, + "loss": 0.0, + "step": 12190 + }, + { + "epoch": 6.793093845725425, + "grad_norm": 0.0010796225629890238, + "learning_rate": 2.679395385100869e-07, + "loss": 0.0, + "step": 12200 + }, + { + "epoch": 6.798663324979114, + "grad_norm": 0.0009600707644567773, + "learning_rate": 2.537773935426535e-07, + "loss": 0.0, + "step": 12210 + }, + { + "epoch": 6.804232804232804, + "grad_norm": 0.0007425449239002865, + "learning_rate": 2.3999880320244583e-07, + "loss": 0.0, + "step": 12220 + }, + { + "epoch": 6.809802283486494, + "grad_norm": 0.0009082748057632814, + "learning_rate": 2.2660387372551162e-07, + "loss": 0.0, + "step": 12230 + }, + { + "epoch": 6.815371762740184, + "grad_norm": 0.0007248110671612175, + "learning_rate": 2.1359270838978706e-07, + "loss": 0.0, + "step": 12240 + }, + { + "epoch": 6.820941241993873, + "grad_norm": 0.0008861206846062583, + "learning_rate": 2.0096540751430305e-07, + "loss": 0.0, + "step": 12250 + }, + { + "epoch": 6.826510721247564, + "grad_norm": 0.0008521378416569952, + "learning_rate": 1.8872206845840234e-07, + "loss": 0.0, + "step": 12260 + }, + { + "epoch": 6.832080200501253, + "grad_norm": 0.0005542695655730265, + "learning_rate": 1.768627856209959e-07, + "loss": 0.0, + "step": 12270 + }, + { + "epoch": 6.837649679754943, + "grad_norm": 0.0007914433832759296, + "learning_rate": 1.6538765043983552e-07, + "loss": 0.0, + "step": 12280 + }, + { + "epoch": 6.843219159008632, + "grad_norm": 0.0007587024668305786, + "learning_rate": 1.5429675139081446e-07, + "loss": 0.0, + "step": 12290 + }, + { + "epoch": 6.848788638262322, + "grad_norm": 0.0013487299085746336, + "learning_rate": 1.435901739872736e-07, + "loss": 0.0, + "step": 12300 + }, + { + "epoch": 6.854358117516012, + "grad_norm": 0.000586521585011824, + "learning_rate": 1.3326800077933521e-07, + "loss": 0.0, + "step": 12310 + }, + { + "epoch": 6.859927596769702, + "grad_norm": 0.0007312839427339511, + "learning_rate": 1.2333031135330353e-07, + "loss": 0.0, + "step": 12320 + }, + { + "epoch": 6.8654970760233915, + "grad_norm": 0.0005875240502264119, + "learning_rate": 1.1377718233100409e-07, + "loss": 0.0, + "step": 12330 + }, + { + "epoch": 6.871066555277082, + "grad_norm": 0.001497294104984269, + "learning_rate": 1.0460868736921758e-07, + "loss": 0.0, + "step": 12340 + }, + { + "epoch": 6.8766360345307715, + "grad_norm": 0.000711562189962145, + "learning_rate": 9.582489715912468e-08, + "loss": 0.0, + "step": 12350 + }, + { + "epoch": 6.882205513784461, + "grad_norm": 0.006542978117968571, + "learning_rate": 8.742587942572322e-08, + "loss": 0.0, + "step": 12360 + }, + { + "epoch": 6.887774993038151, + "grad_norm": 0.00045215906089428337, + "learning_rate": 7.941169892735078e-08, + "loss": 0.0, + "step": 12370 + }, + { + "epoch": 6.89334447229184, + "grad_norm": 0.0012082823621510032, + "learning_rate": 7.178241745514625e-08, + "loss": 0.0, + "step": 12380 + }, + { + "epoch": 6.898913951545531, + "grad_norm": 0.0031225658201267346, + "learning_rate": 6.453809383258902e-08, + "loss": 0.0, + "step": 12390 + }, + { + "epoch": 6.90448343079922, + "grad_norm": 0.0011543412857343766, + "learning_rate": 5.7678783915066094e-08, + "loss": 0.0, + "step": 12400 + }, + { + "epoch": 6.91005291005291, + "grad_norm": 0.0007313799526435521, + "learning_rate": 5.1204540589394615e-08, + "loss": 0.0, + "step": 12410 + }, + { + "epoch": 6.9156223893066, + "grad_norm": 0.0016778937960582093, + "learning_rate": 4.511541377346107e-08, + "loss": 0.0, + "step": 12420 + }, + { + "epoch": 6.92119186856029, + "grad_norm": 0.001203249315839277, + "learning_rate": 3.941145041581606e-08, + "loss": 0.0, + "step": 12430 + }, + { + "epoch": 6.926761347813979, + "grad_norm": 0.002447353900469344, + "learning_rate": 3.409269449529684e-08, + "loss": 0.0, + "step": 12440 + }, + { + "epoch": 6.932330827067669, + "grad_norm": 0.0005101626339405929, + "learning_rate": 2.9159187020721956e-08, + "loss": 0.0, + "step": 12450 + }, + { + "epoch": 6.9379003063213585, + "grad_norm": 0.000924343958398749, + "learning_rate": 2.4610966030547135e-08, + "loss": 0.0, + "step": 12460 + }, + { + "epoch": 6.943469785575049, + "grad_norm": 0.00718380740299027, + "learning_rate": 2.044806659258214e-08, + "loss": 0.0, + "step": 12470 + }, + { + "epoch": 6.9490392648287385, + "grad_norm": 0.0015048135627257007, + "learning_rate": 1.667052080372433e-08, + "loss": 0.0, + "step": 12480 + }, + { + "epoch": 6.954608744082428, + "grad_norm": 0.008617558993811256, + "learning_rate": 1.3278357789708874e-08, + "loss": 0.0, + "step": 12490 + }, + { + "epoch": 6.9601782233361185, + "grad_norm": 0.0007147672191014178, + "learning_rate": 1.0271603704881117e-08, + "loss": 0.0, + "step": 12500 + }, + { + "epoch": 6.965747702589808, + "grad_norm": 0.0006008867991665653, + "learning_rate": 7.650281731991226e-09, + "loss": 0.0, + "step": 12510 + }, + { + "epoch": 6.971317181843498, + "grad_norm": 0.0018045889087410232, + "learning_rate": 5.414412082022091e-09, + "loss": 0.0, + "step": 12520 + }, + { + "epoch": 6.976886661097187, + "grad_norm": 0.0006798063502179644, + "learning_rate": 3.5640119940449913e-09, + "loss": 0.0, + "step": 12530 + }, + { + "epoch": 6.982456140350877, + "grad_norm": 0.0008635206293557315, + "learning_rate": 2.0990957350475183e-09, + "loss": 0.0, + "step": 12540 + }, + { + "epoch": 6.988025619604567, + "grad_norm": 0.0008417661283192795, + "learning_rate": 1.019674599878062e-09, + "loss": 0.0, + "step": 12550 + }, + { + "epoch": 6.993595098858257, + "grad_norm": 0.0025442627266620125, + "learning_rate": 3.2575691111813577e-10, + "loss": 0.0, + "step": 12560 + }, + { + "epoch": 6.999164578111946, + "grad_norm": 0.0010194478366231856, + "learning_rate": 1.7348019015761552e-11, + "loss": 0.0, + "step": 12570 + }, + { + "epoch": 7.0, + "step": 12572, + "total_flos": 1044493520338944.0, + "train_loss": 0.133531805426867, + "train_runtime": 69701.3289, + "train_samples_per_second": 92.308, + "train_steps_per_second": 0.18 + } + ], + "logging_steps": 10, + "max_steps": 12572, + "num_input_tokens_seen": 0, + "num_train_epochs": 7, + "save_steps": 1800, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1044493520338944.0, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +}