{ "best_metric": null, "best_model_checkpoint": null, "epoch": 20.70393374741201, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.020703933747412008, "grad_norm": 10.954449653625488, "learning_rate": 2.0000000000000003e-06, "loss": 0.9733, "step": 10 }, { "epoch": 0.041407867494824016, "grad_norm": 8.008611679077148, "learning_rate": 4.000000000000001e-06, "loss": 0.9684, "step": 20 }, { "epoch": 0.062111801242236024, "grad_norm": 4.479910373687744, "learning_rate": 6e-06, "loss": 0.8292, "step": 30 }, { "epoch": 0.08281573498964803, "grad_norm": 2.9149158000946045, "learning_rate": 8.000000000000001e-06, "loss": 0.5329, "step": 40 }, { "epoch": 0.10351966873706005, "grad_norm": 3.4621212482452393, "learning_rate": 1e-05, "loss": 0.4276, "step": 50 }, { "epoch": 0.12422360248447205, "grad_norm": 4.219636917114258, "learning_rate": 1.2e-05, "loss": 0.3639, "step": 60 }, { "epoch": 0.14492753623188406, "grad_norm": 1.731255054473877, "learning_rate": 1.4000000000000001e-05, "loss": 0.267, "step": 70 }, { "epoch": 0.16563146997929606, "grad_norm": 1.9635241031646729, "learning_rate": 1.6000000000000003e-05, "loss": 0.2161, "step": 80 }, { "epoch": 0.18633540372670807, "grad_norm": 1.9471702575683594, "learning_rate": 1.8e-05, "loss": 0.1705, "step": 90 }, { "epoch": 0.2070393374741201, "grad_norm": 2.2283477783203125, "learning_rate": 2e-05, "loss": 0.1804, "step": 100 }, { "epoch": 0.2277432712215321, "grad_norm": 2.581425666809082, "learning_rate": 2.2000000000000003e-05, "loss": 0.1546, "step": 110 }, { "epoch": 0.2484472049689441, "grad_norm": 1.49114191532135, "learning_rate": 2.4e-05, "loss": 0.1537, "step": 120 }, { "epoch": 0.2691511387163561, "grad_norm": 1.0912235975265503, "learning_rate": 2.6000000000000002e-05, "loss": 0.1145, "step": 130 }, { "epoch": 0.2898550724637681, "grad_norm": 1.0337021350860596, "learning_rate": 2.8000000000000003e-05, "loss": 0.1099, "step": 140 }, { "epoch": 0.3105590062111801, "grad_norm": 1.2799198627471924, "learning_rate": 3e-05, "loss": 0.1242, "step": 150 }, { "epoch": 0.33126293995859213, "grad_norm": 0.7911274433135986, "learning_rate": 3.2000000000000005e-05, "loss": 0.1081, "step": 160 }, { "epoch": 0.35196687370600416, "grad_norm": 2.2392091751098633, "learning_rate": 3.4000000000000007e-05, "loss": 0.0998, "step": 170 }, { "epoch": 0.37267080745341613, "grad_norm": 1.2489616870880127, "learning_rate": 3.6e-05, "loss": 0.0962, "step": 180 }, { "epoch": 0.39337474120082816, "grad_norm": 0.9761829972267151, "learning_rate": 3.8e-05, "loss": 0.1063, "step": 190 }, { "epoch": 0.4140786749482402, "grad_norm": 2.4069204330444336, "learning_rate": 4e-05, "loss": 0.1264, "step": 200 }, { "epoch": 0.43478260869565216, "grad_norm": 1.5159558057785034, "learning_rate": 4.2e-05, "loss": 0.1092, "step": 210 }, { "epoch": 0.4554865424430642, "grad_norm": 2.228130578994751, "learning_rate": 4.4000000000000006e-05, "loss": 0.1154, "step": 220 }, { "epoch": 0.47619047619047616, "grad_norm": 2.4746170043945312, "learning_rate": 4.600000000000001e-05, "loss": 0.0982, "step": 230 }, { "epoch": 0.4968944099378882, "grad_norm": 1.3343104124069214, "learning_rate": 4.8e-05, "loss": 0.1034, "step": 240 }, { "epoch": 0.5175983436853002, "grad_norm": 1.149359941482544, "learning_rate": 5e-05, "loss": 0.0888, "step": 250 }, { "epoch": 0.5383022774327122, "grad_norm": 1.2964980602264404, "learning_rate": 5.2000000000000004e-05, "loss": 0.0906, "step": 260 }, { "epoch": 0.5590062111801242, "grad_norm": 1.0716509819030762, "learning_rate": 5.4000000000000005e-05, "loss": 0.0809, "step": 270 }, { "epoch": 0.5797101449275363, "grad_norm": 0.7431539297103882, "learning_rate": 5.6000000000000006e-05, "loss": 0.0997, "step": 280 }, { "epoch": 0.6004140786749482, "grad_norm": 1.0073963403701782, "learning_rate": 5.8e-05, "loss": 0.081, "step": 290 }, { "epoch": 0.6211180124223602, "grad_norm": 1.2362390756607056, "learning_rate": 6e-05, "loss": 0.0801, "step": 300 }, { "epoch": 0.6418219461697723, "grad_norm": 1.4281072616577148, "learning_rate": 6.2e-05, "loss": 0.0907, "step": 310 }, { "epoch": 0.6625258799171843, "grad_norm": 0.9254273772239685, "learning_rate": 6.400000000000001e-05, "loss": 0.0815, "step": 320 }, { "epoch": 0.6832298136645962, "grad_norm": 1.467358946800232, "learning_rate": 6.6e-05, "loss": 0.082, "step": 330 }, { "epoch": 0.7039337474120083, "grad_norm": 1.0454292297363281, "learning_rate": 6.800000000000001e-05, "loss": 0.0678, "step": 340 }, { "epoch": 0.7246376811594203, "grad_norm": 0.908035397529602, "learning_rate": 7e-05, "loss": 0.0814, "step": 350 }, { "epoch": 0.7453416149068323, "grad_norm": 1.0740008354187012, "learning_rate": 7.2e-05, "loss": 0.0873, "step": 360 }, { "epoch": 0.7660455486542443, "grad_norm": 0.601429283618927, "learning_rate": 7.4e-05, "loss": 0.0641, "step": 370 }, { "epoch": 0.7867494824016563, "grad_norm": 0.7383983135223389, "learning_rate": 7.6e-05, "loss": 0.0709, "step": 380 }, { "epoch": 0.8074534161490683, "grad_norm": 1.0996742248535156, "learning_rate": 7.800000000000001e-05, "loss": 0.0852, "step": 390 }, { "epoch": 0.8281573498964804, "grad_norm": 1.0498085021972656, "learning_rate": 8e-05, "loss": 0.0688, "step": 400 }, { "epoch": 0.8488612836438924, "grad_norm": 0.7805310487747192, "learning_rate": 8.2e-05, "loss": 0.0696, "step": 410 }, { "epoch": 0.8695652173913043, "grad_norm": 0.9954224228858948, "learning_rate": 8.4e-05, "loss": 0.0793, "step": 420 }, { "epoch": 0.8902691511387164, "grad_norm": 5.625354290008545, "learning_rate": 8.6e-05, "loss": 0.1018, "step": 430 }, { "epoch": 0.9109730848861284, "grad_norm": 1.0061132907867432, "learning_rate": 8.800000000000001e-05, "loss": 0.079, "step": 440 }, { "epoch": 0.9316770186335404, "grad_norm": 0.8849142789840698, "learning_rate": 9e-05, "loss": 0.0855, "step": 450 }, { "epoch": 0.9523809523809523, "grad_norm": 0.9031293392181396, "learning_rate": 9.200000000000001e-05, "loss": 0.0654, "step": 460 }, { "epoch": 0.9730848861283644, "grad_norm": 3.3224282264709473, "learning_rate": 9.4e-05, "loss": 0.0812, "step": 470 }, { "epoch": 0.9937888198757764, "grad_norm": 0.8469148278236389, "learning_rate": 9.6e-05, "loss": 0.0739, "step": 480 }, { "epoch": 1.0144927536231885, "grad_norm": 1.2605608701705933, "learning_rate": 9.8e-05, "loss": 0.0771, "step": 490 }, { "epoch": 1.0351966873706004, "grad_norm": 1.0490443706512451, "learning_rate": 0.0001, "loss": 0.0682, "step": 500 }, { "epoch": 1.0559006211180124, "grad_norm": 1.1014221906661987, "learning_rate": 9.999972660400536e-05, "loss": 0.0729, "step": 510 }, { "epoch": 1.0766045548654244, "grad_norm": 0.6805455088615417, "learning_rate": 9.999890641901125e-05, "loss": 0.0666, "step": 520 }, { "epoch": 1.0973084886128364, "grad_norm": 1.2141882181167603, "learning_rate": 9.999753945398704e-05, "loss": 0.0754, "step": 530 }, { "epoch": 1.1180124223602483, "grad_norm": 0.7302441000938416, "learning_rate": 9.99956257238817e-05, "loss": 0.0834, "step": 540 }, { "epoch": 1.1387163561076605, "grad_norm": 1.407357096672058, "learning_rate": 9.999316524962345e-05, "loss": 0.0637, "step": 550 }, { "epoch": 1.1594202898550725, "grad_norm": 0.548819363117218, "learning_rate": 9.999015805811965e-05, "loss": 0.0679, "step": 560 }, { "epoch": 1.1801242236024845, "grad_norm": 0.7563784122467041, "learning_rate": 9.998660418225645e-05, "loss": 0.0742, "step": 570 }, { "epoch": 1.2008281573498965, "grad_norm": 0.7766153812408447, "learning_rate": 9.998250366089848e-05, "loss": 0.0723, "step": 580 }, { "epoch": 1.2215320910973084, "grad_norm": 0.47164079546928406, "learning_rate": 9.997785653888835e-05, "loss": 0.0598, "step": 590 }, { "epoch": 1.2422360248447206, "grad_norm": 0.5156500339508057, "learning_rate": 9.997266286704631e-05, "loss": 0.0645, "step": 600 }, { "epoch": 1.2629399585921326, "grad_norm": 0.708940863609314, "learning_rate": 9.996692270216947e-05, "loss": 0.0715, "step": 610 }, { "epoch": 1.2836438923395446, "grad_norm": 0.7232051491737366, "learning_rate": 9.996063610703137e-05, "loss": 0.0656, "step": 620 }, { "epoch": 1.3043478260869565, "grad_norm": 0.8126294612884521, "learning_rate": 9.995380315038119e-05, "loss": 0.0583, "step": 630 }, { "epoch": 1.3250517598343685, "grad_norm": 0.8648736476898193, "learning_rate": 9.994642390694308e-05, "loss": 0.0633, "step": 640 }, { "epoch": 1.3457556935817805, "grad_norm": 0.8260406851768494, "learning_rate": 9.993849845741524e-05, "loss": 0.0637, "step": 650 }, { "epoch": 1.3664596273291925, "grad_norm": 0.8227211236953735, "learning_rate": 9.993002688846913e-05, "loss": 0.0637, "step": 660 }, { "epoch": 1.3871635610766044, "grad_norm": 0.5047316551208496, "learning_rate": 9.992100929274846e-05, "loss": 0.0524, "step": 670 }, { "epoch": 1.4078674948240166, "grad_norm": 0.554146945476532, "learning_rate": 9.991144576886823e-05, "loss": 0.068, "step": 680 }, { "epoch": 1.4285714285714286, "grad_norm": 0.6884613633155823, "learning_rate": 9.990133642141359e-05, "loss": 0.0465, "step": 690 }, { "epoch": 1.4492753623188406, "grad_norm": 0.6914176344871521, "learning_rate": 9.989068136093873e-05, "loss": 0.0498, "step": 700 }, { "epoch": 1.4699792960662525, "grad_norm": 0.48163503408432007, "learning_rate": 9.987948070396571e-05, "loss": 0.0524, "step": 710 }, { "epoch": 1.4906832298136645, "grad_norm": 0.7618436813354492, "learning_rate": 9.986773457298311e-05, "loss": 0.0573, "step": 720 }, { "epoch": 1.5113871635610767, "grad_norm": 0.5370245575904846, "learning_rate": 9.985544309644475e-05, "loss": 0.052, "step": 730 }, { "epoch": 1.5320910973084887, "grad_norm": 0.9338485598564148, "learning_rate": 9.984260640876821e-05, "loss": 0.047, "step": 740 }, { "epoch": 1.5527950310559007, "grad_norm": 1.0029613971710205, "learning_rate": 9.98292246503335e-05, "loss": 0.0549, "step": 750 }, { "epoch": 1.5734989648033126, "grad_norm": 0.7760806679725647, "learning_rate": 9.981529796748134e-05, "loss": 0.0552, "step": 760 }, { "epoch": 1.5942028985507246, "grad_norm": 0.642235279083252, "learning_rate": 9.980082651251175e-05, "loss": 0.058, "step": 770 }, { "epoch": 1.6149068322981366, "grad_norm": 0.36283358931541443, "learning_rate": 9.97858104436822e-05, "loss": 0.046, "step": 780 }, { "epoch": 1.6356107660455486, "grad_norm": 0.6888338923454285, "learning_rate": 9.977024992520602e-05, "loss": 0.038, "step": 790 }, { "epoch": 1.6563146997929605, "grad_norm": 0.48172858357429504, "learning_rate": 9.975414512725057e-05, "loss": 0.05, "step": 800 }, { "epoch": 1.6770186335403725, "grad_norm": 0.33963388204574585, "learning_rate": 9.973749622593534e-05, "loss": 0.0487, "step": 810 }, { "epoch": 1.6977225672877847, "grad_norm": 0.5968677401542664, "learning_rate": 9.972030340333001e-05, "loss": 0.0605, "step": 820 }, { "epoch": 1.7184265010351967, "grad_norm": 0.5687774419784546, "learning_rate": 9.970256684745258e-05, "loss": 0.0573, "step": 830 }, { "epoch": 1.7391304347826086, "grad_norm": 0.5432090163230896, "learning_rate": 9.968428675226714e-05, "loss": 0.0502, "step": 840 }, { "epoch": 1.7598343685300208, "grad_norm": 0.8394588232040405, "learning_rate": 9.966546331768191e-05, "loss": 0.0507, "step": 850 }, { "epoch": 1.7805383022774328, "grad_norm": 0.5862396955490112, "learning_rate": 9.964609674954696e-05, "loss": 0.0564, "step": 860 }, { "epoch": 1.8012422360248448, "grad_norm": 0.5196480751037598, "learning_rate": 9.962618725965196e-05, "loss": 0.0545, "step": 870 }, { "epoch": 1.8219461697722568, "grad_norm": 0.7387868165969849, "learning_rate": 9.96057350657239e-05, "loss": 0.0546, "step": 880 }, { "epoch": 1.8426501035196687, "grad_norm": 0.6779050827026367, "learning_rate": 9.95847403914247e-05, "loss": 0.0569, "step": 890 }, { "epoch": 1.8633540372670807, "grad_norm": 0.5078296661376953, "learning_rate": 9.956320346634876e-05, "loss": 0.0522, "step": 900 }, { "epoch": 1.8840579710144927, "grad_norm": 0.5735164284706116, "learning_rate": 9.954112452602045e-05, "loss": 0.0426, "step": 910 }, { "epoch": 1.9047619047619047, "grad_norm": 0.7340726852416992, "learning_rate": 9.95185038118915e-05, "loss": 0.052, "step": 920 }, { "epoch": 1.9254658385093166, "grad_norm": 0.8836645483970642, "learning_rate": 9.949534157133844e-05, "loss": 0.0682, "step": 930 }, { "epoch": 1.9461697722567288, "grad_norm": 0.44937947392463684, "learning_rate": 9.94716380576598e-05, "loss": 0.0447, "step": 940 }, { "epoch": 1.9668737060041408, "grad_norm": 0.3076542019844055, "learning_rate": 9.944739353007344e-05, "loss": 0.0483, "step": 950 }, { "epoch": 1.9875776397515528, "grad_norm": 0.5312771797180176, "learning_rate": 9.942260825371358e-05, "loss": 0.0597, "step": 960 }, { "epoch": 2.008281573498965, "grad_norm": 0.932463526725769, "learning_rate": 9.939728249962807e-05, "loss": 0.0444, "step": 970 }, { "epoch": 2.028985507246377, "grad_norm": 0.636582612991333, "learning_rate": 9.937141654477528e-05, "loss": 0.05, "step": 980 }, { "epoch": 2.049689440993789, "grad_norm": 0.49437057971954346, "learning_rate": 9.934501067202117e-05, "loss": 0.039, "step": 990 }, { "epoch": 2.070393374741201, "grad_norm": 0.6429054737091064, "learning_rate": 9.931806517013612e-05, "loss": 0.0416, "step": 1000 }, { "epoch": 2.091097308488613, "grad_norm": 0.4953542947769165, "learning_rate": 9.929058033379181e-05, "loss": 0.0501, "step": 1010 }, { "epoch": 2.111801242236025, "grad_norm": 0.36293789744377136, "learning_rate": 9.926255646355804e-05, "loss": 0.0377, "step": 1020 }, { "epoch": 2.132505175983437, "grad_norm": 0.39418739080429077, "learning_rate": 9.923399386589933e-05, "loss": 0.0385, "step": 1030 }, { "epoch": 2.153209109730849, "grad_norm": 0.6820284128189087, "learning_rate": 9.92048928531717e-05, "loss": 0.0356, "step": 1040 }, { "epoch": 2.1739130434782608, "grad_norm": 0.35965871810913086, "learning_rate": 9.917525374361912e-05, "loss": 0.0457, "step": 1050 }, { "epoch": 2.1946169772256727, "grad_norm": 0.5950950384140015, "learning_rate": 9.914507686137019e-05, "loss": 0.0488, "step": 1060 }, { "epoch": 2.2153209109730847, "grad_norm": 0.354141503572464, "learning_rate": 9.911436253643445e-05, "loss": 0.0329, "step": 1070 }, { "epoch": 2.2360248447204967, "grad_norm": 0.3924166262149811, "learning_rate": 9.90831111046988e-05, "loss": 0.0393, "step": 1080 }, { "epoch": 2.256728778467909, "grad_norm": 0.6275556683540344, "learning_rate": 9.905132290792394e-05, "loss": 0.0388, "step": 1090 }, { "epoch": 2.277432712215321, "grad_norm": 0.5029153227806091, "learning_rate": 9.901899829374047e-05, "loss": 0.0425, "step": 1100 }, { "epoch": 2.298136645962733, "grad_norm": 0.3522745966911316, "learning_rate": 9.89861376156452e-05, "loss": 0.0362, "step": 1110 }, { "epoch": 2.318840579710145, "grad_norm": 0.4343843460083008, "learning_rate": 9.895274123299723e-05, "loss": 0.0384, "step": 1120 }, { "epoch": 2.339544513457557, "grad_norm": 0.3715870976448059, "learning_rate": 9.891880951101407e-05, "loss": 0.0402, "step": 1130 }, { "epoch": 2.360248447204969, "grad_norm": 0.6320225596427917, "learning_rate": 9.888434282076758e-05, "loss": 0.0453, "step": 1140 }, { "epoch": 2.380952380952381, "grad_norm": 0.5040298104286194, "learning_rate": 9.884934153917997e-05, "loss": 0.0461, "step": 1150 }, { "epoch": 2.401656314699793, "grad_norm": 0.6153097152709961, "learning_rate": 9.881380604901964e-05, "loss": 0.0421, "step": 1160 }, { "epoch": 2.422360248447205, "grad_norm": 0.371401846408844, "learning_rate": 9.877773673889701e-05, "loss": 0.0423, "step": 1170 }, { "epoch": 2.443064182194617, "grad_norm": 0.6483995914459229, "learning_rate": 9.87411340032603e-05, "loss": 0.0455, "step": 1180 }, { "epoch": 2.463768115942029, "grad_norm": 0.5516171455383301, "learning_rate": 9.870399824239117e-05, "loss": 0.0469, "step": 1190 }, { "epoch": 2.4844720496894412, "grad_norm": 0.4739258885383606, "learning_rate": 9.86663298624003e-05, "loss": 0.0407, "step": 1200 }, { "epoch": 2.505175983436853, "grad_norm": 0.5367606282234192, "learning_rate": 9.862812927522309e-05, "loss": 0.0456, "step": 1210 }, { "epoch": 2.525879917184265, "grad_norm": 0.6636613607406616, "learning_rate": 9.858939689861506e-05, "loss": 0.0534, "step": 1220 }, { "epoch": 2.546583850931677, "grad_norm": 0.5070763230323792, "learning_rate": 9.855013315614725e-05, "loss": 0.0395, "step": 1230 }, { "epoch": 2.567287784679089, "grad_norm": 0.622347354888916, "learning_rate": 9.851033847720166e-05, "loss": 0.0445, "step": 1240 }, { "epoch": 2.587991718426501, "grad_norm": 0.23474477231502533, "learning_rate": 9.847001329696653e-05, "loss": 0.0368, "step": 1250 }, { "epoch": 2.608695652173913, "grad_norm": 0.4050194323062897, "learning_rate": 9.842915805643155e-05, "loss": 0.0425, "step": 1260 }, { "epoch": 2.629399585921325, "grad_norm": 0.486103355884552, "learning_rate": 9.838777320238312e-05, "loss": 0.034, "step": 1270 }, { "epoch": 2.650103519668737, "grad_norm": 0.5604904890060425, "learning_rate": 9.834585918739936e-05, "loss": 0.0424, "step": 1280 }, { "epoch": 2.670807453416149, "grad_norm": 0.3005557656288147, "learning_rate": 9.830341646984521e-05, "loss": 0.0377, "step": 1290 }, { "epoch": 2.691511387163561, "grad_norm": 0.38094374537467957, "learning_rate": 9.826044551386744e-05, "loss": 0.0419, "step": 1300 }, { "epoch": 2.712215320910973, "grad_norm": 0.4101620018482208, "learning_rate": 9.821694678938953e-05, "loss": 0.0323, "step": 1310 }, { "epoch": 2.732919254658385, "grad_norm": 0.6201085448265076, "learning_rate": 9.817292077210659e-05, "loss": 0.0385, "step": 1320 }, { "epoch": 2.753623188405797, "grad_norm": 0.42668941617012024, "learning_rate": 9.812836794348004e-05, "loss": 0.0427, "step": 1330 }, { "epoch": 2.774327122153209, "grad_norm": 0.47652679681777954, "learning_rate": 9.808328879073251e-05, "loss": 0.0446, "step": 1340 }, { "epoch": 2.795031055900621, "grad_norm": 0.24209241569042206, "learning_rate": 9.803768380684242e-05, "loss": 0.0362, "step": 1350 }, { "epoch": 2.8157349896480333, "grad_norm": 0.5838305950164795, "learning_rate": 9.799155349053851e-05, "loss": 0.0511, "step": 1360 }, { "epoch": 2.8364389233954452, "grad_norm": 0.46614477038383484, "learning_rate": 9.794489834629455e-05, "loss": 0.0371, "step": 1370 }, { "epoch": 2.857142857142857, "grad_norm": 0.509036123752594, "learning_rate": 9.789771888432375e-05, "loss": 0.0378, "step": 1380 }, { "epoch": 2.877846790890269, "grad_norm": 0.31127938628196716, "learning_rate": 9.785001562057309e-05, "loss": 0.0426, "step": 1390 }, { "epoch": 2.898550724637681, "grad_norm": 0.6117241978645325, "learning_rate": 9.780178907671789e-05, "loss": 0.0439, "step": 1400 }, { "epoch": 2.919254658385093, "grad_norm": 0.5319984555244446, "learning_rate": 9.775303978015585e-05, "loss": 0.0492, "step": 1410 }, { "epoch": 2.939958592132505, "grad_norm": 0.3564855456352234, "learning_rate": 9.77037682640015e-05, "loss": 0.0422, "step": 1420 }, { "epoch": 2.960662525879917, "grad_norm": 0.45328769087791443, "learning_rate": 9.765397506708023e-05, "loss": 0.0465, "step": 1430 }, { "epoch": 2.981366459627329, "grad_norm": 0.38785624504089355, "learning_rate": 9.760366073392246e-05, "loss": 0.033, "step": 1440 }, { "epoch": 3.002070393374741, "grad_norm": 0.43779534101486206, "learning_rate": 9.755282581475769e-05, "loss": 0.0494, "step": 1450 }, { "epoch": 3.022774327122153, "grad_norm": 0.4085356295108795, "learning_rate": 9.750147086550844e-05, "loss": 0.0441, "step": 1460 }, { "epoch": 3.0434782608695654, "grad_norm": 0.5406202077865601, "learning_rate": 9.744959644778422e-05, "loss": 0.0334, "step": 1470 }, { "epoch": 3.0641821946169774, "grad_norm": 0.38816168904304504, "learning_rate": 9.739720312887535e-05, "loss": 0.0365, "step": 1480 }, { "epoch": 3.0848861283643894, "grad_norm": 0.5753049254417419, "learning_rate": 9.734429148174675e-05, "loss": 0.0285, "step": 1490 }, { "epoch": 3.1055900621118013, "grad_norm": 0.2940957844257355, "learning_rate": 9.729086208503174e-05, "loss": 0.0373, "step": 1500 }, { "epoch": 3.1262939958592133, "grad_norm": 0.44986191391944885, "learning_rate": 9.723691552302562e-05, "loss": 0.0277, "step": 1510 }, { "epoch": 3.1469979296066253, "grad_norm": 0.3255261778831482, "learning_rate": 9.718245238567939e-05, "loss": 0.038, "step": 1520 }, { "epoch": 3.1677018633540373, "grad_norm": 0.5038771629333496, "learning_rate": 9.712747326859315e-05, "loss": 0.0331, "step": 1530 }, { "epoch": 3.1884057971014492, "grad_norm": 0.37753045558929443, "learning_rate": 9.707197877300974e-05, "loss": 0.0291, "step": 1540 }, { "epoch": 3.209109730848861, "grad_norm": 0.25833946466445923, "learning_rate": 9.701596950580806e-05, "loss": 0.0416, "step": 1550 }, { "epoch": 3.229813664596273, "grad_norm": 0.31511035561561584, "learning_rate": 9.695944607949649e-05, "loss": 0.0299, "step": 1560 }, { "epoch": 3.250517598343685, "grad_norm": 0.386797159910202, "learning_rate": 9.690240911220618e-05, "loss": 0.032, "step": 1570 }, { "epoch": 3.271221532091097, "grad_norm": 0.5350190997123718, "learning_rate": 9.684485922768422e-05, "loss": 0.0376, "step": 1580 }, { "epoch": 3.291925465838509, "grad_norm": 0.39487507939338684, "learning_rate": 9.6786797055287e-05, "loss": 0.0383, "step": 1590 }, { "epoch": 3.3126293995859215, "grad_norm": 0.4336493909358978, "learning_rate": 9.672822322997305e-05, "loss": 0.0425, "step": 1600 }, { "epoch": 3.3333333333333335, "grad_norm": 0.6210718154907227, "learning_rate": 9.66691383922964e-05, "loss": 0.047, "step": 1610 }, { "epoch": 3.3540372670807455, "grad_norm": 0.5560672283172607, "learning_rate": 9.660954318839933e-05, "loss": 0.039, "step": 1620 }, { "epoch": 3.3747412008281574, "grad_norm": 0.791201651096344, "learning_rate": 9.654943827000548e-05, "loss": 0.0429, "step": 1630 }, { "epoch": 3.3954451345755694, "grad_norm": 0.46332821249961853, "learning_rate": 9.648882429441257e-05, "loss": 0.036, "step": 1640 }, { "epoch": 3.4161490683229814, "grad_norm": 0.45162495970726013, "learning_rate": 9.642770192448536e-05, "loss": 0.0303, "step": 1650 }, { "epoch": 3.4368530020703933, "grad_norm": 0.46616628766059875, "learning_rate": 9.636607182864827e-05, "loss": 0.0503, "step": 1660 }, { "epoch": 3.4575569358178053, "grad_norm": 0.41262561082839966, "learning_rate": 9.630393468087818e-05, "loss": 0.0372, "step": 1670 }, { "epoch": 3.4782608695652173, "grad_norm": 0.41251444816589355, "learning_rate": 9.624129116069694e-05, "loss": 0.0336, "step": 1680 }, { "epoch": 3.4989648033126293, "grad_norm": 0.52425217628479, "learning_rate": 9.617814195316411e-05, "loss": 0.0512, "step": 1690 }, { "epoch": 3.5196687370600412, "grad_norm": 0.4962485134601593, "learning_rate": 9.611448774886924e-05, "loss": 0.034, "step": 1700 }, { "epoch": 3.5403726708074537, "grad_norm": 0.390449196100235, "learning_rate": 9.605032924392457e-05, "loss": 0.0311, "step": 1710 }, { "epoch": 3.5610766045548656, "grad_norm": 0.42751261591911316, "learning_rate": 9.598566713995718e-05, "loss": 0.0372, "step": 1720 }, { "epoch": 3.5817805383022776, "grad_norm": 0.5136762857437134, "learning_rate": 9.59205021441015e-05, "loss": 0.0351, "step": 1730 }, { "epoch": 3.6024844720496896, "grad_norm": 0.44243088364601135, "learning_rate": 9.58548349689915e-05, "loss": 0.0388, "step": 1740 }, { "epoch": 3.6231884057971016, "grad_norm": 0.3298715651035309, "learning_rate": 9.578866633275288e-05, "loss": 0.0321, "step": 1750 }, { "epoch": 3.6438923395445135, "grad_norm": 0.5570884346961975, "learning_rate": 9.572199695899522e-05, "loss": 0.0312, "step": 1760 }, { "epoch": 3.6645962732919255, "grad_norm": 0.6233508586883545, "learning_rate": 9.565482757680415e-05, "loss": 0.0325, "step": 1770 }, { "epoch": 3.6853002070393375, "grad_norm": 0.33611059188842773, "learning_rate": 9.558715892073323e-05, "loss": 0.0291, "step": 1780 }, { "epoch": 3.7060041407867494, "grad_norm": 0.7385571002960205, "learning_rate": 9.551899173079607e-05, "loss": 0.0381, "step": 1790 }, { "epoch": 3.7267080745341614, "grad_norm": 0.35607799887657166, "learning_rate": 9.545032675245813e-05, "loss": 0.0411, "step": 1800 }, { "epoch": 3.7474120082815734, "grad_norm": 0.474151611328125, "learning_rate": 9.538116473662861e-05, "loss": 0.0266, "step": 1810 }, { "epoch": 3.7681159420289854, "grad_norm": 0.25087299942970276, "learning_rate": 9.531150643965223e-05, "loss": 0.0279, "step": 1820 }, { "epoch": 3.7888198757763973, "grad_norm": 0.288636177778244, "learning_rate": 9.524135262330098e-05, "loss": 0.0361, "step": 1830 }, { "epoch": 3.8095238095238093, "grad_norm": 0.4557327926158905, "learning_rate": 9.517070405476575e-05, "loss": 0.0355, "step": 1840 }, { "epoch": 3.8302277432712213, "grad_norm": 0.5964323878288269, "learning_rate": 9.509956150664796e-05, "loss": 0.0322, "step": 1850 }, { "epoch": 3.8509316770186337, "grad_norm": 0.43868887424468994, "learning_rate": 9.502792575695112e-05, "loss": 0.0362, "step": 1860 }, { "epoch": 3.8716356107660457, "grad_norm": 0.361118882894516, "learning_rate": 9.49557975890723e-05, "loss": 0.0327, "step": 1870 }, { "epoch": 3.8923395445134576, "grad_norm": 0.48435598611831665, "learning_rate": 9.488317779179361e-05, "loss": 0.0332, "step": 1880 }, { "epoch": 3.9130434782608696, "grad_norm": 0.2833453118801117, "learning_rate": 9.481006715927351e-05, "loss": 0.0372, "step": 1890 }, { "epoch": 3.9337474120082816, "grad_norm": 0.4986327588558197, "learning_rate": 9.473646649103818e-05, "loss": 0.0352, "step": 1900 }, { "epoch": 3.9544513457556936, "grad_norm": 0.6213241815567017, "learning_rate": 9.46623765919727e-05, "loss": 0.0279, "step": 1910 }, { "epoch": 3.9751552795031055, "grad_norm": 0.3823232352733612, "learning_rate": 9.458779827231237e-05, "loss": 0.0314, "step": 1920 }, { "epoch": 3.9958592132505175, "grad_norm": 0.4441359043121338, "learning_rate": 9.451273234763371e-05, "loss": 0.0326, "step": 1930 }, { "epoch": 4.01656314699793, "grad_norm": 0.21336762607097626, "learning_rate": 9.443717963884569e-05, "loss": 0.0299, "step": 1940 }, { "epoch": 4.037267080745342, "grad_norm": 0.3937424123287201, "learning_rate": 9.43611409721806e-05, "loss": 0.0312, "step": 1950 }, { "epoch": 4.057971014492754, "grad_norm": 0.5097484588623047, "learning_rate": 9.428461717918511e-05, "loss": 0.0252, "step": 1960 }, { "epoch": 4.078674948240166, "grad_norm": 0.42226624488830566, "learning_rate": 9.420760909671118e-05, "loss": 0.0316, "step": 1970 }, { "epoch": 4.099378881987578, "grad_norm": 0.31922483444213867, "learning_rate": 9.413011756690685e-05, "loss": 0.0363, "step": 1980 }, { "epoch": 4.12008281573499, "grad_norm": 0.522516131401062, "learning_rate": 9.405214343720707e-05, "loss": 0.0346, "step": 1990 }, { "epoch": 4.140786749482402, "grad_norm": 0.5185388326644897, "learning_rate": 9.397368756032445e-05, "loss": 0.0333, "step": 2000 }, { "epoch": 4.161490683229814, "grad_norm": 0.348429411649704, "learning_rate": 9.389475079423988e-05, "loss": 0.0272, "step": 2010 }, { "epoch": 4.182194616977226, "grad_norm": 0.5857434272766113, "learning_rate": 9.381533400219318e-05, "loss": 0.0341, "step": 2020 }, { "epoch": 4.202898550724638, "grad_norm": 0.20706452429294586, "learning_rate": 9.373543805267368e-05, "loss": 0.0248, "step": 2030 }, { "epoch": 4.22360248447205, "grad_norm": 0.3589947819709778, "learning_rate": 9.365506381941066e-05, "loss": 0.0305, "step": 2040 }, { "epoch": 4.244306418219462, "grad_norm": 0.23909683525562286, "learning_rate": 9.357421218136386e-05, "loss": 0.0256, "step": 2050 }, { "epoch": 4.265010351966874, "grad_norm": 0.25338152050971985, "learning_rate": 9.349288402271388e-05, "loss": 0.0316, "step": 2060 }, { "epoch": 4.285714285714286, "grad_norm": 0.4180321991443634, "learning_rate": 9.341108023285238e-05, "loss": 0.0339, "step": 2070 }, { "epoch": 4.306418219461698, "grad_norm": 0.48361265659332275, "learning_rate": 9.332880170637252e-05, "loss": 0.0306, "step": 2080 }, { "epoch": 4.3271221532091095, "grad_norm": 0.5146428346633911, "learning_rate": 9.32460493430591e-05, "loss": 0.0252, "step": 2090 }, { "epoch": 4.3478260869565215, "grad_norm": 0.27129676938056946, "learning_rate": 9.316282404787871e-05, "loss": 0.0235, "step": 2100 }, { "epoch": 4.3685300207039335, "grad_norm": 0.5199556946754456, "learning_rate": 9.30791267309698e-05, "loss": 0.0365, "step": 2110 }, { "epoch": 4.3892339544513455, "grad_norm": 0.47860321402549744, "learning_rate": 9.299495830763286e-05, "loss": 0.0422, "step": 2120 }, { "epoch": 4.409937888198757, "grad_norm": 0.47205787897109985, "learning_rate": 9.291031969832026e-05, "loss": 0.0404, "step": 2130 }, { "epoch": 4.430641821946169, "grad_norm": 0.43115857243537903, "learning_rate": 9.282521182862629e-05, "loss": 0.0304, "step": 2140 }, { "epoch": 4.451345755693581, "grad_norm": 0.48119500279426575, "learning_rate": 9.273963562927695e-05, "loss": 0.0292, "step": 2150 }, { "epoch": 4.472049689440993, "grad_norm": 0.40839287638664246, "learning_rate": 9.265359203611987e-05, "loss": 0.0322, "step": 2160 }, { "epoch": 4.492753623188406, "grad_norm": 0.2620416581630707, "learning_rate": 9.256708199011401e-05, "loss": 0.0237, "step": 2170 }, { "epoch": 4.513457556935818, "grad_norm": 0.3477851450443268, "learning_rate": 9.248010643731935e-05, "loss": 0.0231, "step": 2180 }, { "epoch": 4.53416149068323, "grad_norm": 0.4392837584018707, "learning_rate": 9.239266632888659e-05, "loss": 0.0275, "step": 2190 }, { "epoch": 4.554865424430642, "grad_norm": 0.33065155148506165, "learning_rate": 9.230476262104677e-05, "loss": 0.0268, "step": 2200 }, { "epoch": 4.575569358178054, "grad_norm": 0.3106347620487213, "learning_rate": 9.221639627510076e-05, "loss": 0.0308, "step": 2210 }, { "epoch": 4.596273291925466, "grad_norm": 0.4171868860721588, "learning_rate": 9.212756825740873e-05, "loss": 0.0277, "step": 2220 }, { "epoch": 4.616977225672878, "grad_norm": 0.3243824541568756, "learning_rate": 9.20382795393797e-05, "loss": 0.0333, "step": 2230 }, { "epoch": 4.63768115942029, "grad_norm": 0.1873687207698822, "learning_rate": 9.194853109746074e-05, "loss": 0.0311, "step": 2240 }, { "epoch": 4.658385093167702, "grad_norm": 0.4594937562942505, "learning_rate": 9.185832391312644e-05, "loss": 0.0293, "step": 2250 }, { "epoch": 4.679089026915114, "grad_norm": 0.5178239941596985, "learning_rate": 9.176765897286813e-05, "loss": 0.0388, "step": 2260 }, { "epoch": 4.699792960662526, "grad_norm": 0.42131152749061584, "learning_rate": 9.167653726818305e-05, "loss": 0.0332, "step": 2270 }, { "epoch": 4.720496894409938, "grad_norm": 0.42832908034324646, "learning_rate": 9.158495979556358e-05, "loss": 0.0329, "step": 2280 }, { "epoch": 4.74120082815735, "grad_norm": 0.4283015727996826, "learning_rate": 9.14929275564863e-05, "loss": 0.0346, "step": 2290 }, { "epoch": 4.761904761904762, "grad_norm": 0.30669862031936646, "learning_rate": 9.140044155740101e-05, "loss": 0.0255, "step": 2300 }, { "epoch": 4.782608695652174, "grad_norm": 0.4990992248058319, "learning_rate": 9.130750280971978e-05, "loss": 0.0328, "step": 2310 }, { "epoch": 4.803312629399586, "grad_norm": 0.5123105049133301, "learning_rate": 9.121411232980588e-05, "loss": 0.0277, "step": 2320 }, { "epoch": 4.824016563146998, "grad_norm": 0.4550347924232483, "learning_rate": 9.112027113896262e-05, "loss": 0.0273, "step": 2330 }, { "epoch": 4.84472049689441, "grad_norm": 0.4566217362880707, "learning_rate": 9.102598026342222e-05, "loss": 0.0318, "step": 2340 }, { "epoch": 4.865424430641822, "grad_norm": 0.4516465663909912, "learning_rate": 9.093124073433463e-05, "loss": 0.0329, "step": 2350 }, { "epoch": 4.886128364389234, "grad_norm": 0.32221224904060364, "learning_rate": 9.083605358775612e-05, "loss": 0.0258, "step": 2360 }, { "epoch": 4.906832298136646, "grad_norm": 0.5452279448509216, "learning_rate": 9.074041986463808e-05, "loss": 0.0351, "step": 2370 }, { "epoch": 4.927536231884058, "grad_norm": 0.484084814786911, "learning_rate": 9.064434061081562e-05, "loss": 0.0243, "step": 2380 }, { "epoch": 4.94824016563147, "grad_norm": 0.33230844140052795, "learning_rate": 9.0547816876996e-05, "loss": 0.032, "step": 2390 }, { "epoch": 4.9689440993788825, "grad_norm": 0.3503859341144562, "learning_rate": 9.045084971874738e-05, "loss": 0.0268, "step": 2400 }, { "epoch": 4.989648033126294, "grad_norm": 0.282907098531723, "learning_rate": 9.035344019648702e-05, "loss": 0.0301, "step": 2410 }, { "epoch": 5.010351966873706, "grad_norm": 0.4622001051902771, "learning_rate": 9.025558937546988e-05, "loss": 0.0318, "step": 2420 }, { "epoch": 5.031055900621118, "grad_norm": 0.40841957926750183, "learning_rate": 9.015729832577681e-05, "loss": 0.0317, "step": 2430 }, { "epoch": 5.05175983436853, "grad_norm": 0.4344206750392914, "learning_rate": 9.005856812230304e-05, "loss": 0.0281, "step": 2440 }, { "epoch": 5.072463768115942, "grad_norm": 0.44237494468688965, "learning_rate": 8.995939984474624e-05, "loss": 0.0307, "step": 2450 }, { "epoch": 5.093167701863354, "grad_norm": 0.2730436325073242, "learning_rate": 8.98597945775948e-05, "loss": 0.0265, "step": 2460 }, { "epoch": 5.113871635610766, "grad_norm": 0.35804224014282227, "learning_rate": 8.975975341011596e-05, "loss": 0.0256, "step": 2470 }, { "epoch": 5.134575569358178, "grad_norm": 0.4116530418395996, "learning_rate": 8.965927743634391e-05, "loss": 0.0365, "step": 2480 }, { "epoch": 5.15527950310559, "grad_norm": 0.5391436219215393, "learning_rate": 8.955836775506776e-05, "loss": 0.0412, "step": 2490 }, { "epoch": 5.175983436853002, "grad_norm": 0.3000620901584625, "learning_rate": 8.945702546981969e-05, "loss": 0.0321, "step": 2500 }, { "epoch": 5.196687370600414, "grad_norm": 0.5762383341789246, "learning_rate": 8.935525168886262e-05, "loss": 0.0324, "step": 2510 }, { "epoch": 5.217391304347826, "grad_norm": 0.6489707827568054, "learning_rate": 8.92530475251784e-05, "loss": 0.0294, "step": 2520 }, { "epoch": 5.238095238095238, "grad_norm": 0.29640257358551025, "learning_rate": 8.91504140964553e-05, "loss": 0.0287, "step": 2530 }, { "epoch": 5.25879917184265, "grad_norm": 0.48081815242767334, "learning_rate": 8.90473525250761e-05, "loss": 0.0384, "step": 2540 }, { "epoch": 5.279503105590062, "grad_norm": 0.448220819234848, "learning_rate": 8.894386393810563e-05, "loss": 0.0228, "step": 2550 }, { "epoch": 5.300207039337474, "grad_norm": 0.5022518038749695, "learning_rate": 8.883994946727849e-05, "loss": 0.0373, "step": 2560 }, { "epoch": 5.320910973084886, "grad_norm": 0.5916407704353333, "learning_rate": 8.873561024898668e-05, "loss": 0.034, "step": 2570 }, { "epoch": 5.341614906832298, "grad_norm": 0.5583520531654358, "learning_rate": 8.863084742426719e-05, "loss": 0.0351, "step": 2580 }, { "epoch": 5.36231884057971, "grad_norm": 0.3444780707359314, "learning_rate": 8.852566213878947e-05, "loss": 0.0313, "step": 2590 }, { "epoch": 5.383022774327122, "grad_norm": 0.2927059531211853, "learning_rate": 8.842005554284296e-05, "loss": 0.0349, "step": 2600 }, { "epoch": 5.403726708074534, "grad_norm": 0.6348573565483093, "learning_rate": 8.831402879132446e-05, "loss": 0.0271, "step": 2610 }, { "epoch": 5.424430641821946, "grad_norm": 0.3420064449310303, "learning_rate": 8.820758304372557e-05, "loss": 0.0254, "step": 2620 }, { "epoch": 5.445134575569358, "grad_norm": 0.7388267517089844, "learning_rate": 8.810071946411989e-05, "loss": 0.0421, "step": 2630 }, { "epoch": 5.46583850931677, "grad_norm": 0.4998408555984497, "learning_rate": 8.799343922115044e-05, "loss": 0.0287, "step": 2640 }, { "epoch": 5.486542443064182, "grad_norm": 0.24863387644290924, "learning_rate": 8.788574348801675e-05, "loss": 0.0311, "step": 2650 }, { "epoch": 5.507246376811594, "grad_norm": 0.366595983505249, "learning_rate": 8.77776334424621e-05, "loss": 0.0222, "step": 2660 }, { "epoch": 5.527950310559007, "grad_norm": 0.35403600335121155, "learning_rate": 8.766911026676064e-05, "loss": 0.0302, "step": 2670 }, { "epoch": 5.548654244306418, "grad_norm": 0.4470798373222351, "learning_rate": 8.756017514770443e-05, "loss": 0.0283, "step": 2680 }, { "epoch": 5.569358178053831, "grad_norm": 0.13218455016613007, "learning_rate": 8.745082927659047e-05, "loss": 0.0237, "step": 2690 }, { "epoch": 5.590062111801243, "grad_norm": 0.25023823976516724, "learning_rate": 8.73410738492077e-05, "loss": 0.0276, "step": 2700 }, { "epoch": 5.6107660455486545, "grad_norm": 0.4505211114883423, "learning_rate": 8.723091006582389e-05, "loss": 0.026, "step": 2710 }, { "epoch": 5.6314699792960665, "grad_norm": 0.31018003821372986, "learning_rate": 8.71203391311725e-05, "loss": 0.0325, "step": 2720 }, { "epoch": 5.6521739130434785, "grad_norm": 0.35771510004997253, "learning_rate": 8.700936225443959e-05, "loss": 0.0219, "step": 2730 }, { "epoch": 5.6728778467908905, "grad_norm": 0.33808183670043945, "learning_rate": 8.689798064925049e-05, "loss": 0.0269, "step": 2740 }, { "epoch": 5.693581780538302, "grad_norm": 0.3441556692123413, "learning_rate": 8.678619553365659e-05, "loss": 0.0258, "step": 2750 }, { "epoch": 5.714285714285714, "grad_norm": 0.34911125898361206, "learning_rate": 8.6674008130122e-05, "loss": 0.0311, "step": 2760 }, { "epoch": 5.734989648033126, "grad_norm": 0.329429030418396, "learning_rate": 8.656141966551019e-05, "loss": 0.0289, "step": 2770 }, { "epoch": 5.755693581780538, "grad_norm": 0.2914882302284241, "learning_rate": 8.644843137107059e-05, "loss": 0.0261, "step": 2780 }, { "epoch": 5.77639751552795, "grad_norm": 0.4363202750682831, "learning_rate": 8.633504448242505e-05, "loss": 0.0268, "step": 2790 }, { "epoch": 5.797101449275362, "grad_norm": 0.3086094558238983, "learning_rate": 8.622126023955446e-05, "loss": 0.0288, "step": 2800 }, { "epoch": 5.817805383022774, "grad_norm": 0.3617849051952362, "learning_rate": 8.610707988678503e-05, "loss": 0.0303, "step": 2810 }, { "epoch": 5.838509316770186, "grad_norm": 0.45564624667167664, "learning_rate": 8.599250467277483e-05, "loss": 0.0282, "step": 2820 }, { "epoch": 5.859213250517598, "grad_norm": 0.6213960647583008, "learning_rate": 8.587753585050004e-05, "loss": 0.0331, "step": 2830 }, { "epoch": 5.87991718426501, "grad_norm": 0.5800597071647644, "learning_rate": 8.576217467724128e-05, "loss": 0.0343, "step": 2840 }, { "epoch": 5.900621118012422, "grad_norm": 0.33786582946777344, "learning_rate": 8.564642241456986e-05, "loss": 0.0292, "step": 2850 }, { "epoch": 5.921325051759834, "grad_norm": 0.41556116938591003, "learning_rate": 8.553028032833397e-05, "loss": 0.0338, "step": 2860 }, { "epoch": 5.942028985507246, "grad_norm": 0.5863654017448425, "learning_rate": 8.541374968864487e-05, "loss": 0.029, "step": 2870 }, { "epoch": 5.962732919254658, "grad_norm": 0.4155533015727997, "learning_rate": 8.529683176986295e-05, "loss": 0.0313, "step": 2880 }, { "epoch": 5.98343685300207, "grad_norm": 0.5086021423339844, "learning_rate": 8.517952785058385e-05, "loss": 0.0287, "step": 2890 }, { "epoch": 6.004140786749482, "grad_norm": 0.3907274603843689, "learning_rate": 8.506183921362443e-05, "loss": 0.0307, "step": 2900 }, { "epoch": 6.024844720496894, "grad_norm": 0.30351269245147705, "learning_rate": 8.494376714600878e-05, "loss": 0.028, "step": 2910 }, { "epoch": 6.045548654244306, "grad_norm": 0.2281622290611267, "learning_rate": 8.482531293895412e-05, "loss": 0.0261, "step": 2920 }, { "epoch": 6.066252587991719, "grad_norm": 0.35197684168815613, "learning_rate": 8.470647788785665e-05, "loss": 0.0276, "step": 2930 }, { "epoch": 6.086956521739131, "grad_norm": 0.41056281328201294, "learning_rate": 8.458726329227747e-05, "loss": 0.0281, "step": 2940 }, { "epoch": 6.107660455486543, "grad_norm": 0.31276586651802063, "learning_rate": 8.44676704559283e-05, "loss": 0.0352, "step": 2950 }, { "epoch": 6.128364389233955, "grad_norm": 0.48580434918403625, "learning_rate": 8.434770068665723e-05, "loss": 0.0258, "step": 2960 }, { "epoch": 6.149068322981367, "grad_norm": 0.4106568992137909, "learning_rate": 8.422735529643444e-05, "loss": 0.0305, "step": 2970 }, { "epoch": 6.169772256728779, "grad_norm": 0.4538796544075012, "learning_rate": 8.410663560133784e-05, "loss": 0.0237, "step": 2980 }, { "epoch": 6.190476190476191, "grad_norm": 0.3809707462787628, "learning_rate": 8.398554292153866e-05, "loss": 0.0297, "step": 2990 }, { "epoch": 6.211180124223603, "grad_norm": 0.3905803859233856, "learning_rate": 8.386407858128706e-05, "loss": 0.0335, "step": 3000 }, { "epoch": 6.231884057971015, "grad_norm": 0.3672570288181305, "learning_rate": 8.37422439088976e-05, "loss": 0.0234, "step": 3010 }, { "epoch": 6.252587991718427, "grad_norm": 0.34976840019226074, "learning_rate": 8.362004023673474e-05, "loss": 0.0252, "step": 3020 }, { "epoch": 6.273291925465839, "grad_norm": 0.19189304113388062, "learning_rate": 8.349746890119826e-05, "loss": 0.0238, "step": 3030 }, { "epoch": 6.2939958592132506, "grad_norm": 0.3705725073814392, "learning_rate": 8.337453124270863e-05, "loss": 0.0275, "step": 3040 }, { "epoch": 6.3146997929606625, "grad_norm": 0.4177100360393524, "learning_rate": 8.32512286056924e-05, "loss": 0.0299, "step": 3050 }, { "epoch": 6.3354037267080745, "grad_norm": 0.2962421476840973, "learning_rate": 8.31275623385675e-05, "loss": 0.0223, "step": 3060 }, { "epoch": 6.3561076604554865, "grad_norm": 0.3034144341945648, "learning_rate": 8.300353379372834e-05, "loss": 0.0223, "step": 3070 }, { "epoch": 6.3768115942028984, "grad_norm": 0.29205501079559326, "learning_rate": 8.287914432753123e-05, "loss": 0.0399, "step": 3080 }, { "epoch": 6.39751552795031, "grad_norm": 0.47431355714797974, "learning_rate": 8.275439530027948e-05, "loss": 0.0319, "step": 3090 }, { "epoch": 6.418219461697722, "grad_norm": 0.2878510355949402, "learning_rate": 8.262928807620843e-05, "loss": 0.028, "step": 3100 }, { "epoch": 6.438923395445134, "grad_norm": 0.33809617161750793, "learning_rate": 8.250382402347065e-05, "loss": 0.0207, "step": 3110 }, { "epoch": 6.459627329192546, "grad_norm": 0.5072120428085327, "learning_rate": 8.237800451412095e-05, "loss": 0.0314, "step": 3120 }, { "epoch": 6.480331262939958, "grad_norm": 0.3112965524196625, "learning_rate": 8.225183092410128e-05, "loss": 0.0232, "step": 3130 }, { "epoch": 6.50103519668737, "grad_norm": 0.3710940182209015, "learning_rate": 8.212530463322583e-05, "loss": 0.0232, "step": 3140 }, { "epoch": 6.521739130434782, "grad_norm": 0.31945547461509705, "learning_rate": 8.199842702516583e-05, "loss": 0.0325, "step": 3150 }, { "epoch": 6.542443064182194, "grad_norm": 0.29330453276634216, "learning_rate": 8.18711994874345e-05, "loss": 0.0191, "step": 3160 }, { "epoch": 6.563146997929607, "grad_norm": 0.13522489368915558, "learning_rate": 8.174362341137177e-05, "loss": 0.0278, "step": 3170 }, { "epoch": 6.583850931677018, "grad_norm": 0.3913141191005707, "learning_rate": 8.161570019212921e-05, "loss": 0.0287, "step": 3180 }, { "epoch": 6.604554865424431, "grad_norm": 0.2847929894924164, "learning_rate": 8.148743122865463e-05, "loss": 0.0226, "step": 3190 }, { "epoch": 6.625258799171843, "grad_norm": 0.5095137357711792, "learning_rate": 8.135881792367686e-05, "loss": 0.0294, "step": 3200 }, { "epoch": 6.645962732919255, "grad_norm": 0.199754998087883, "learning_rate": 8.12298616836904e-05, "loss": 0.0253, "step": 3210 }, { "epoch": 6.666666666666667, "grad_norm": 0.2732815444469452, "learning_rate": 8.110056391894005e-05, "loss": 0.024, "step": 3220 }, { "epoch": 6.687370600414079, "grad_norm": 0.3035735785961151, "learning_rate": 8.097092604340542e-05, "loss": 0.0197, "step": 3230 }, { "epoch": 6.708074534161491, "grad_norm": 0.2598547637462616, "learning_rate": 8.084094947478556e-05, "loss": 0.0228, "step": 3240 }, { "epoch": 6.728778467908903, "grad_norm": 0.1960446983575821, "learning_rate": 8.07106356344834e-05, "loss": 0.0233, "step": 3250 }, { "epoch": 6.749482401656315, "grad_norm": 0.32702773809432983, "learning_rate": 8.057998594759022e-05, "loss": 0.0183, "step": 3260 }, { "epoch": 6.770186335403727, "grad_norm": 0.265981525182724, "learning_rate": 8.044900184287007e-05, "loss": 0.0268, "step": 3270 }, { "epoch": 6.790890269151139, "grad_norm": 0.24002982676029205, "learning_rate": 8.031768475274413e-05, "loss": 0.0182, "step": 3280 }, { "epoch": 6.811594202898551, "grad_norm": 0.46799013018608093, "learning_rate": 8.018603611327504e-05, "loss": 0.027, "step": 3290 }, { "epoch": 6.832298136645963, "grad_norm": 0.29448822140693665, "learning_rate": 8.005405736415126e-05, "loss": 0.0328, "step": 3300 }, { "epoch": 6.853002070393375, "grad_norm": 0.3686560094356537, "learning_rate": 7.992174994867123e-05, "loss": 0.0178, "step": 3310 }, { "epoch": 6.873706004140787, "grad_norm": 0.28374069929122925, "learning_rate": 7.978911531372765e-05, "loss": 0.022, "step": 3320 }, { "epoch": 6.894409937888199, "grad_norm": 0.3148317337036133, "learning_rate": 7.965615490979163e-05, "loss": 0.0305, "step": 3330 }, { "epoch": 6.915113871635611, "grad_norm": 0.4778421223163605, "learning_rate": 7.952287019089685e-05, "loss": 0.0277, "step": 3340 }, { "epoch": 6.935817805383023, "grad_norm": 0.29899469017982483, "learning_rate": 7.938926261462366e-05, "loss": 0.0257, "step": 3350 }, { "epoch": 6.956521739130435, "grad_norm": 0.3062913715839386, "learning_rate": 7.925533364208309e-05, "loss": 0.0286, "step": 3360 }, { "epoch": 6.977225672877847, "grad_norm": 0.37298765778541565, "learning_rate": 7.912108473790092e-05, "loss": 0.0248, "step": 3370 }, { "epoch": 6.9979296066252585, "grad_norm": 0.29648107290267944, "learning_rate": 7.898651737020166e-05, "loss": 0.0219, "step": 3380 }, { "epoch": 7.0186335403726705, "grad_norm": 0.40795350074768066, "learning_rate": 7.88516330105925e-05, "loss": 0.0216, "step": 3390 }, { "epoch": 7.0393374741200825, "grad_norm": 0.343569278717041, "learning_rate": 7.871643313414718e-05, "loss": 0.0225, "step": 3400 }, { "epoch": 7.0600414078674945, "grad_norm": 0.3091362416744232, "learning_rate": 7.858091921938988e-05, "loss": 0.0205, "step": 3410 }, { "epoch": 7.080745341614906, "grad_norm": 0.18946941196918488, "learning_rate": 7.844509274827907e-05, "loss": 0.0246, "step": 3420 }, { "epoch": 7.101449275362318, "grad_norm": 0.2987332046031952, "learning_rate": 7.830895520619128e-05, "loss": 0.0219, "step": 3430 }, { "epoch": 7.122153209109731, "grad_norm": 0.24055786430835724, "learning_rate": 7.817250808190483e-05, "loss": 0.0308, "step": 3440 }, { "epoch": 7.142857142857143, "grad_norm": 0.34465551376342773, "learning_rate": 7.803575286758364e-05, "loss": 0.0225, "step": 3450 }, { "epoch": 7.163561076604555, "grad_norm": 0.37269163131713867, "learning_rate": 7.789869105876083e-05, "loss": 0.0246, "step": 3460 }, { "epoch": 7.184265010351967, "grad_norm": 0.22413749992847443, "learning_rate": 7.776132415432234e-05, "loss": 0.0191, "step": 3470 }, { "epoch": 7.204968944099379, "grad_norm": 0.269936203956604, "learning_rate": 7.762365365649067e-05, "loss": 0.021, "step": 3480 }, { "epoch": 7.225672877846791, "grad_norm": 0.41103509068489075, "learning_rate": 7.748568107080832e-05, "loss": 0.0283, "step": 3490 }, { "epoch": 7.246376811594203, "grad_norm": 0.2623867690563202, "learning_rate": 7.734740790612136e-05, "loss": 0.0234, "step": 3500 }, { "epoch": 7.267080745341615, "grad_norm": 0.27451378107070923, "learning_rate": 7.720883567456298e-05, "loss": 0.0273, "step": 3510 }, { "epoch": 7.287784679089027, "grad_norm": 0.3108258545398712, "learning_rate": 7.70699658915369e-05, "loss": 0.0273, "step": 3520 }, { "epoch": 7.308488612836439, "grad_norm": 0.37050384283065796, "learning_rate": 7.693080007570084e-05, "loss": 0.0198, "step": 3530 }, { "epoch": 7.329192546583851, "grad_norm": 0.4067702889442444, "learning_rate": 7.679133974894983e-05, "loss": 0.0279, "step": 3540 }, { "epoch": 7.349896480331263, "grad_norm": 0.22218254208564758, "learning_rate": 7.66515864363997e-05, "loss": 0.0212, "step": 3550 }, { "epoch": 7.370600414078675, "grad_norm": 0.26355263590812683, "learning_rate": 7.651154166637025e-05, "loss": 0.0219, "step": 3560 }, { "epoch": 7.391304347826087, "grad_norm": 0.2989826202392578, "learning_rate": 7.637120697036866e-05, "loss": 0.0318, "step": 3570 }, { "epoch": 7.412008281573499, "grad_norm": 0.2896329462528229, "learning_rate": 7.623058388307269e-05, "loss": 0.0249, "step": 3580 }, { "epoch": 7.432712215320911, "grad_norm": 0.28438034653663635, "learning_rate": 7.608967394231387e-05, "loss": 0.0265, "step": 3590 }, { "epoch": 7.453416149068323, "grad_norm": 0.1998050957918167, "learning_rate": 7.594847868906076e-05, "loss": 0.0215, "step": 3600 }, { "epoch": 7.474120082815735, "grad_norm": 0.3249650299549103, "learning_rate": 7.580699966740201e-05, "loss": 0.0255, "step": 3610 }, { "epoch": 7.494824016563147, "grad_norm": 0.41978025436401367, "learning_rate": 7.566523842452958e-05, "loss": 0.0254, "step": 3620 }, { "epoch": 7.515527950310559, "grad_norm": 0.17664960026741028, "learning_rate": 7.552319651072164e-05, "loss": 0.0242, "step": 3630 }, { "epoch": 7.536231884057971, "grad_norm": 0.26141443848609924, "learning_rate": 7.538087547932585e-05, "loss": 0.0224, "step": 3640 }, { "epoch": 7.556935817805383, "grad_norm": 0.2887045741081238, "learning_rate": 7.52382768867422e-05, "loss": 0.0249, "step": 3650 }, { "epoch": 7.577639751552795, "grad_norm": 0.2639414370059967, "learning_rate": 7.509540229240601e-05, "loss": 0.0192, "step": 3660 }, { "epoch": 7.598343685300207, "grad_norm": 0.2664073705673218, "learning_rate": 7.495225325877103e-05, "loss": 0.0214, "step": 3670 }, { "epoch": 7.619047619047619, "grad_norm": 0.18211261928081512, "learning_rate": 7.480883135129211e-05, "loss": 0.0175, "step": 3680 }, { "epoch": 7.6397515527950315, "grad_norm": 0.3653377294540405, "learning_rate": 7.466513813840825e-05, "loss": 0.0244, "step": 3690 }, { "epoch": 7.660455486542443, "grad_norm": 0.46319136023521423, "learning_rate": 7.452117519152542e-05, "loss": 0.0258, "step": 3700 }, { "epoch": 7.681159420289855, "grad_norm": 0.38133862614631653, "learning_rate": 7.437694408499933e-05, "loss": 0.0274, "step": 3710 }, { "epoch": 7.701863354037267, "grad_norm": 0.3771878182888031, "learning_rate": 7.423244639611826e-05, "loss": 0.024, "step": 3720 }, { "epoch": 7.722567287784679, "grad_norm": 0.32930412888526917, "learning_rate": 7.408768370508576e-05, "loss": 0.0213, "step": 3730 }, { "epoch": 7.743271221532091, "grad_norm": 0.49230140447616577, "learning_rate": 7.394265759500348e-05, "loss": 0.0293, "step": 3740 }, { "epoch": 7.763975155279503, "grad_norm": 0.374051034450531, "learning_rate": 7.379736965185368e-05, "loss": 0.0269, "step": 3750 }, { "epoch": 7.784679089026915, "grad_norm": 0.21926304697990417, "learning_rate": 7.365182146448205e-05, "loss": 0.0201, "step": 3760 }, { "epoch": 7.805383022774327, "grad_norm": 0.3494378626346588, "learning_rate": 7.350601462458024e-05, "loss": 0.0201, "step": 3770 }, { "epoch": 7.826086956521739, "grad_norm": 0.2537420988082886, "learning_rate": 7.335995072666848e-05, "loss": 0.02, "step": 3780 }, { "epoch": 7.846790890269151, "grad_norm": 0.3533625304698944, "learning_rate": 7.32136313680782e-05, "loss": 0.0299, "step": 3790 }, { "epoch": 7.867494824016563, "grad_norm": 0.5519553422927856, "learning_rate": 7.30670581489344e-05, "loss": 0.0312, "step": 3800 }, { "epoch": 7.888198757763975, "grad_norm": 0.3124694228172302, "learning_rate": 7.292023267213835e-05, "loss": 0.029, "step": 3810 }, { "epoch": 7.908902691511387, "grad_norm": 0.3334481716156006, "learning_rate": 7.277315654334997e-05, "loss": 0.0209, "step": 3820 }, { "epoch": 7.929606625258799, "grad_norm": 0.17688047885894775, "learning_rate": 7.262583137097018e-05, "loss": 0.0224, "step": 3830 }, { "epoch": 7.950310559006211, "grad_norm": 0.236612007021904, "learning_rate": 7.247825876612353e-05, "loss": 0.0259, "step": 3840 }, { "epoch": 7.971014492753623, "grad_norm": 0.26868879795074463, "learning_rate": 7.233044034264034e-05, "loss": 0.0175, "step": 3850 }, { "epoch": 7.991718426501035, "grad_norm": 0.4900842010974884, "learning_rate": 7.218237771703921e-05, "loss": 0.0327, "step": 3860 }, { "epoch": 8.012422360248447, "grad_norm": 0.28051623702049255, "learning_rate": 7.203407250850928e-05, "loss": 0.0237, "step": 3870 }, { "epoch": 8.03312629399586, "grad_norm": 0.3828851878643036, "learning_rate": 7.188552633889259e-05, "loss": 0.0257, "step": 3880 }, { "epoch": 8.053830227743271, "grad_norm": 0.38572344183921814, "learning_rate": 7.173674083266624e-05, "loss": 0.0247, "step": 3890 }, { "epoch": 8.074534161490684, "grad_norm": 0.3239911198616028, "learning_rate": 7.158771761692464e-05, "loss": 0.0245, "step": 3900 }, { "epoch": 8.095238095238095, "grad_norm": 0.34434774518013, "learning_rate": 7.143845832136188e-05, "loss": 0.0174, "step": 3910 }, { "epoch": 8.115942028985508, "grad_norm": 0.23144909739494324, "learning_rate": 7.128896457825364e-05, "loss": 0.0209, "step": 3920 }, { "epoch": 8.136645962732919, "grad_norm": 0.24774691462516785, "learning_rate": 7.113923802243957e-05, "loss": 0.0207, "step": 3930 }, { "epoch": 8.157349896480332, "grad_norm": 0.24319271743297577, "learning_rate": 7.09892802913053e-05, "loss": 0.0242, "step": 3940 }, { "epoch": 8.178053830227743, "grad_norm": 0.3723747730255127, "learning_rate": 7.083909302476453e-05, "loss": 0.0255, "step": 3950 }, { "epoch": 8.198757763975156, "grad_norm": 0.5450414419174194, "learning_rate": 7.068867786524116e-05, "loss": 0.0356, "step": 3960 }, { "epoch": 8.219461697722567, "grad_norm": 0.4084027409553528, "learning_rate": 7.053803645765128e-05, "loss": 0.0271, "step": 3970 }, { "epoch": 8.24016563146998, "grad_norm": 0.49706506729125977, "learning_rate": 7.038717044938519e-05, "loss": 0.0228, "step": 3980 }, { "epoch": 8.26086956521739, "grad_norm": 0.4912974536418915, "learning_rate": 7.023608149028937e-05, "loss": 0.0201, "step": 3990 }, { "epoch": 8.281573498964804, "grad_norm": 0.3758380711078644, "learning_rate": 7.008477123264848e-05, "loss": 0.0238, "step": 4000 }, { "epoch": 8.302277432712215, "grad_norm": 0.18395407497882843, "learning_rate": 6.993324133116726e-05, "loss": 0.0198, "step": 4010 }, { "epoch": 8.322981366459627, "grad_norm": 0.3328471779823303, "learning_rate": 6.978149344295242e-05, "loss": 0.0213, "step": 4020 }, { "epoch": 8.343685300207039, "grad_norm": 0.22626806795597076, "learning_rate": 6.962952922749457e-05, "loss": 0.0218, "step": 4030 }, { "epoch": 8.364389233954451, "grad_norm": 0.4286823272705078, "learning_rate": 6.947735034665002e-05, "loss": 0.0229, "step": 4040 }, { "epoch": 8.385093167701863, "grad_norm": 0.2956433594226837, "learning_rate": 6.932495846462261e-05, "loss": 0.0272, "step": 4050 }, { "epoch": 8.405797101449275, "grad_norm": 0.4478289783000946, "learning_rate": 6.917235524794558e-05, "loss": 0.0262, "step": 4060 }, { "epoch": 8.426501035196686, "grad_norm": 0.45018693804740906, "learning_rate": 6.901954236546323e-05, "loss": 0.0305, "step": 4070 }, { "epoch": 8.4472049689441, "grad_norm": 0.49448198080062866, "learning_rate": 6.886652148831279e-05, "loss": 0.0229, "step": 4080 }, { "epoch": 8.467908902691512, "grad_norm": 0.2818650007247925, "learning_rate": 6.871329428990602e-05, "loss": 0.0193, "step": 4090 }, { "epoch": 8.488612836438923, "grad_norm": 0.3502029776573181, "learning_rate": 6.855986244591104e-05, "loss": 0.0191, "step": 4100 }, { "epoch": 8.509316770186336, "grad_norm": 0.13417436182498932, "learning_rate": 6.840622763423391e-05, "loss": 0.0183, "step": 4110 }, { "epoch": 8.530020703933747, "grad_norm": 0.24381810426712036, "learning_rate": 6.825239153500029e-05, "loss": 0.0211, "step": 4120 }, { "epoch": 8.55072463768116, "grad_norm": 0.3400433659553528, "learning_rate": 6.809835583053715e-05, "loss": 0.0176, "step": 4130 }, { "epoch": 8.571428571428571, "grad_norm": 0.34003961086273193, "learning_rate": 6.794412220535426e-05, "loss": 0.0181, "step": 4140 }, { "epoch": 8.592132505175984, "grad_norm": 0.48874515295028687, "learning_rate": 6.778969234612584e-05, "loss": 0.0248, "step": 4150 }, { "epoch": 8.612836438923395, "grad_norm": 0.24148191511631012, "learning_rate": 6.763506794167208e-05, "loss": 0.0143, "step": 4160 }, { "epoch": 8.633540372670808, "grad_norm": 0.3760093152523041, "learning_rate": 6.748025068294067e-05, "loss": 0.017, "step": 4170 }, { "epoch": 8.654244306418219, "grad_norm": 0.1769023984670639, "learning_rate": 6.732524226298841e-05, "loss": 0.0179, "step": 4180 }, { "epoch": 8.674948240165632, "grad_norm": 0.14992591738700867, "learning_rate": 6.71700443769625e-05, "loss": 0.0149, "step": 4190 }, { "epoch": 8.695652173913043, "grad_norm": 0.16137127578258514, "learning_rate": 6.701465872208216e-05, "loss": 0.0147, "step": 4200 }, { "epoch": 8.716356107660456, "grad_norm": 0.2600420117378235, "learning_rate": 6.685908699762002e-05, "loss": 0.019, "step": 4210 }, { "epoch": 8.737060041407867, "grad_norm": 0.24808962643146515, "learning_rate": 6.670333090488356e-05, "loss": 0.0209, "step": 4220 }, { "epoch": 8.75776397515528, "grad_norm": 0.47357505559921265, "learning_rate": 6.654739214719641e-05, "loss": 0.0187, "step": 4230 }, { "epoch": 8.778467908902691, "grad_norm": 0.3729642331600189, "learning_rate": 6.639127242987988e-05, "loss": 0.0222, "step": 4240 }, { "epoch": 8.799171842650104, "grad_norm": 0.31445324420928955, "learning_rate": 6.623497346023418e-05, "loss": 0.0205, "step": 4250 }, { "epoch": 8.819875776397515, "grad_norm": 0.15997394919395447, "learning_rate": 6.607849694751977e-05, "loss": 0.0177, "step": 4260 }, { "epoch": 8.840579710144928, "grad_norm": 0.2809275686740875, "learning_rate": 6.592184460293877e-05, "loss": 0.0196, "step": 4270 }, { "epoch": 8.861283643892339, "grad_norm": 0.3261137008666992, "learning_rate": 6.576501813961609e-05, "loss": 0.0204, "step": 4280 }, { "epoch": 8.881987577639752, "grad_norm": 0.3846384584903717, "learning_rate": 6.56080192725808e-05, "loss": 0.0311, "step": 4290 }, { "epoch": 8.902691511387163, "grad_norm": 0.3049221932888031, "learning_rate": 6.545084971874738e-05, "loss": 0.0154, "step": 4300 }, { "epoch": 8.923395445134576, "grad_norm": 0.22341875731945038, "learning_rate": 6.529351119689688e-05, "loss": 0.0187, "step": 4310 }, { "epoch": 8.944099378881987, "grad_norm": 0.28026238083839417, "learning_rate": 6.513600542765817e-05, "loss": 0.0183, "step": 4320 }, { "epoch": 8.9648033126294, "grad_norm": 0.18726959824562073, "learning_rate": 6.497833413348909e-05, "loss": 0.0198, "step": 4330 }, { "epoch": 8.985507246376812, "grad_norm": 0.2277226448059082, "learning_rate": 6.48204990386577e-05, "loss": 0.0155, "step": 4340 }, { "epoch": 9.006211180124224, "grad_norm": 0.20866835117340088, "learning_rate": 6.466250186922325e-05, "loss": 0.0144, "step": 4350 }, { "epoch": 9.026915113871636, "grad_norm": 0.23058342933654785, "learning_rate": 6.450434435301751e-05, "loss": 0.0198, "step": 4360 }, { "epoch": 9.047619047619047, "grad_norm": 0.24547383189201355, "learning_rate": 6.43460282196257e-05, "loss": 0.0192, "step": 4370 }, { "epoch": 9.06832298136646, "grad_norm": 0.27292782068252563, "learning_rate": 6.418755520036775e-05, "loss": 0.0206, "step": 4380 }, { "epoch": 9.089026915113871, "grad_norm": 0.3834291398525238, "learning_rate": 6.402892702827916e-05, "loss": 0.0247, "step": 4390 }, { "epoch": 9.109730848861284, "grad_norm": 0.19833965599536896, "learning_rate": 6.387014543809223e-05, "loss": 0.0154, "step": 4400 }, { "epoch": 9.130434782608695, "grad_norm": 0.22265447676181793, "learning_rate": 6.371121216621698e-05, "loss": 0.0161, "step": 4410 }, { "epoch": 9.151138716356108, "grad_norm": 0.12021901458501816, "learning_rate": 6.355212895072223e-05, "loss": 0.0154, "step": 4420 }, { "epoch": 9.17184265010352, "grad_norm": 0.330624520778656, "learning_rate": 6.339289753131649e-05, "loss": 0.014, "step": 4430 }, { "epoch": 9.192546583850932, "grad_norm": 0.2923831045627594, "learning_rate": 6.323351964932908e-05, "loss": 0.0184, "step": 4440 }, { "epoch": 9.213250517598343, "grad_norm": 0.26423075795173645, "learning_rate": 6.307399704769099e-05, "loss": 0.0224, "step": 4450 }, { "epoch": 9.233954451345756, "grad_norm": 0.5406923294067383, "learning_rate": 6.291433147091583e-05, "loss": 0.0227, "step": 4460 }, { "epoch": 9.254658385093167, "grad_norm": 0.34261617064476013, "learning_rate": 6.275452466508077e-05, "loss": 0.0202, "step": 4470 }, { "epoch": 9.27536231884058, "grad_norm": 0.36320188641548157, "learning_rate": 6.259457837780742e-05, "loss": 0.0144, "step": 4480 }, { "epoch": 9.296066252587991, "grad_norm": 0.3232795298099518, "learning_rate": 6.243449435824276e-05, "loss": 0.0219, "step": 4490 }, { "epoch": 9.316770186335404, "grad_norm": 0.314527690410614, "learning_rate": 6.227427435703997e-05, "loss": 0.0273, "step": 4500 }, { "epoch": 9.337474120082815, "grad_norm": 0.186433807015419, "learning_rate": 6.211392012633932e-05, "loss": 0.017, "step": 4510 }, { "epoch": 9.358178053830228, "grad_norm": 0.17290981113910675, "learning_rate": 6.195343341974899e-05, "loss": 0.026, "step": 4520 }, { "epoch": 9.378881987577639, "grad_norm": 0.31283560395240784, "learning_rate": 6.179281599232591e-05, "loss": 0.0235, "step": 4530 }, { "epoch": 9.399585921325052, "grad_norm": 0.27143594622612, "learning_rate": 6.163206960055651e-05, "loss": 0.0145, "step": 4540 }, { "epoch": 9.420289855072463, "grad_norm": 0.16036300361156464, "learning_rate": 6.147119600233758e-05, "loss": 0.0221, "step": 4550 }, { "epoch": 9.440993788819876, "grad_norm": 0.3918144404888153, "learning_rate": 6.131019695695702e-05, "loss": 0.018, "step": 4560 }, { "epoch": 9.461697722567287, "grad_norm": 0.28290480375289917, "learning_rate": 6.11490742250746e-05, "loss": 0.0191, "step": 4570 }, { "epoch": 9.4824016563147, "grad_norm": 0.3056785762310028, "learning_rate": 6.0987829568702656e-05, "loss": 0.0215, "step": 4580 }, { "epoch": 9.503105590062113, "grad_norm": 0.2111171931028366, "learning_rate": 6.0826464751186994e-05, "loss": 0.0137, "step": 4590 }, { "epoch": 9.523809523809524, "grad_norm": 0.23214763402938843, "learning_rate": 6.066498153718735e-05, "loss": 0.0135, "step": 4600 }, { "epoch": 9.544513457556937, "grad_norm": 0.21107251942157745, "learning_rate": 6.05033816926583e-05, "loss": 0.0218, "step": 4610 }, { "epoch": 9.565217391304348, "grad_norm": 0.27265289425849915, "learning_rate": 6.034166698482984e-05, "loss": 0.0238, "step": 4620 }, { "epoch": 9.58592132505176, "grad_norm": 0.1942167729139328, "learning_rate": 6.017983918218812e-05, "loss": 0.0133, "step": 4630 }, { "epoch": 9.606625258799172, "grad_norm": 0.2295994758605957, "learning_rate": 6.001790005445607e-05, "loss": 0.0179, "step": 4640 }, { "epoch": 9.627329192546584, "grad_norm": 0.23539908230304718, "learning_rate": 5.985585137257401e-05, "loss": 0.0257, "step": 4650 }, { "epoch": 9.648033126293996, "grad_norm": 0.34829631447792053, "learning_rate": 5.969369490868042e-05, "loss": 0.0228, "step": 4660 }, { "epoch": 9.668737060041408, "grad_norm": 0.19475673139095306, "learning_rate": 5.953143243609235e-05, "loss": 0.0159, "step": 4670 }, { "epoch": 9.68944099378882, "grad_norm": 0.31705808639526367, "learning_rate": 5.9369065729286245e-05, "loss": 0.0162, "step": 4680 }, { "epoch": 9.710144927536232, "grad_norm": 0.2687765955924988, "learning_rate": 5.9206596563878357e-05, "loss": 0.0235, "step": 4690 }, { "epoch": 9.730848861283643, "grad_norm": 0.3420402705669403, "learning_rate": 5.90440267166055e-05, "loss": 0.019, "step": 4700 }, { "epoch": 9.751552795031056, "grad_norm": 0.37165331840515137, "learning_rate": 5.888135796530544e-05, "loss": 0.0172, "step": 4710 }, { "epoch": 9.772256728778467, "grad_norm": 0.334695428609848, "learning_rate": 5.871859208889759e-05, "loss": 0.0171, "step": 4720 }, { "epoch": 9.79296066252588, "grad_norm": 0.25561410188674927, "learning_rate": 5.85557308673635e-05, "loss": 0.0172, "step": 4730 }, { "epoch": 9.813664596273291, "grad_norm": 0.3481326103210449, "learning_rate": 5.8392776081727385e-05, "loss": 0.0217, "step": 4740 }, { "epoch": 9.834368530020704, "grad_norm": 0.22137852013111115, "learning_rate": 5.8229729514036705e-05, "loss": 0.0143, "step": 4750 }, { "epoch": 9.855072463768115, "grad_norm": 0.23530082404613495, "learning_rate": 5.8066592947342555e-05, "loss": 0.0206, "step": 4760 }, { "epoch": 9.875776397515528, "grad_norm": 0.27072563767433167, "learning_rate": 5.7903368165680327e-05, "loss": 0.018, "step": 4770 }, { "epoch": 9.89648033126294, "grad_norm": 0.20534668862819672, "learning_rate": 5.7740056954050084e-05, "loss": 0.0225, "step": 4780 }, { "epoch": 9.917184265010352, "grad_norm": 0.28875890374183655, "learning_rate": 5.757666109839702e-05, "loss": 0.0258, "step": 4790 }, { "epoch": 9.937888198757763, "grad_norm": 0.18692868947982788, "learning_rate": 5.74131823855921e-05, "loss": 0.0158, "step": 4800 }, { "epoch": 9.958592132505176, "grad_norm": 0.3396063446998596, "learning_rate": 5.72496226034123e-05, "loss": 0.0192, "step": 4810 }, { "epoch": 9.979296066252587, "grad_norm": 0.3160467743873596, "learning_rate": 5.7085983540521216e-05, "loss": 0.0122, "step": 4820 }, { "epoch": 10.0, "grad_norm": 0.40236353874206543, "learning_rate": 5.692226698644938e-05, "loss": 0.02, "step": 4830 }, { "epoch": 10.020703933747413, "grad_norm": 0.2102344036102295, "learning_rate": 5.675847473157485e-05, "loss": 0.0151, "step": 4840 }, { "epoch": 10.041407867494824, "grad_norm": 0.33124828338623047, "learning_rate": 5.6594608567103456e-05, "loss": 0.0169, "step": 4850 }, { "epoch": 10.062111801242237, "grad_norm": 0.2858668565750122, "learning_rate": 5.6430670285049314e-05, "loss": 0.0142, "step": 4860 }, { "epoch": 10.082815734989648, "grad_norm": 0.2905900776386261, "learning_rate": 5.6266661678215216e-05, "loss": 0.0195, "step": 4870 }, { "epoch": 10.10351966873706, "grad_norm": 0.24875830113887787, "learning_rate": 5.6102584540173006e-05, "loss": 0.0189, "step": 4880 }, { "epoch": 10.124223602484472, "grad_norm": 0.43277251720428467, "learning_rate": 5.5938440665244006e-05, "loss": 0.0186, "step": 4890 }, { "epoch": 10.144927536231885, "grad_norm": 0.2178991585969925, "learning_rate": 5.577423184847932e-05, "loss": 0.0162, "step": 4900 }, { "epoch": 10.165631469979296, "grad_norm": 0.3437841832637787, "learning_rate": 5.560995988564023e-05, "loss": 0.0214, "step": 4910 }, { "epoch": 10.186335403726709, "grad_norm": 0.44986671209335327, "learning_rate": 5.544562657317863e-05, "loss": 0.0217, "step": 4920 }, { "epoch": 10.20703933747412, "grad_norm": 0.36503782868385315, "learning_rate": 5.52812337082173e-05, "loss": 0.0188, "step": 4930 }, { "epoch": 10.227743271221533, "grad_norm": 0.25650519132614136, "learning_rate": 5.511678308853026e-05, "loss": 0.0168, "step": 4940 }, { "epoch": 10.248447204968944, "grad_norm": 0.26455608010292053, "learning_rate": 5.495227651252315e-05, "loss": 0.0147, "step": 4950 }, { "epoch": 10.269151138716357, "grad_norm": 0.2997487783432007, "learning_rate": 5.478771577921351e-05, "loss": 0.0203, "step": 4960 }, { "epoch": 10.289855072463768, "grad_norm": 0.31882262229919434, "learning_rate": 5.462310268821118e-05, "loss": 0.0157, "step": 4970 }, { "epoch": 10.31055900621118, "grad_norm": 0.24016189575195312, "learning_rate": 5.445843903969854e-05, "loss": 0.0125, "step": 4980 }, { "epoch": 10.331262939958592, "grad_norm": 0.19592761993408203, "learning_rate": 5.4293726634410855e-05, "loss": 0.0178, "step": 4990 }, { "epoch": 10.351966873706004, "grad_norm": 0.23570533096790314, "learning_rate": 5.4128967273616625e-05, "loss": 0.0171, "step": 5000 }, { "epoch": 10.372670807453416, "grad_norm": 0.34885501861572266, "learning_rate": 5.396416275909779e-05, "loss": 0.0247, "step": 5010 }, { "epoch": 10.393374741200828, "grad_norm": 0.24810874462127686, "learning_rate": 5.379931489313016e-05, "loss": 0.0187, "step": 5020 }, { "epoch": 10.41407867494824, "grad_norm": 0.21873916685581207, "learning_rate": 5.363442547846356e-05, "loss": 0.0197, "step": 5030 }, { "epoch": 10.434782608695652, "grad_norm": 0.3136371970176697, "learning_rate": 5.3469496318302204e-05, "loss": 0.0163, "step": 5040 }, { "epoch": 10.455486542443063, "grad_norm": 0.3234352171421051, "learning_rate": 5.330452921628497e-05, "loss": 0.0158, "step": 5050 }, { "epoch": 10.476190476190476, "grad_norm": 0.34137850999832153, "learning_rate": 5.313952597646568e-05, "loss": 0.0146, "step": 5060 }, { "epoch": 10.496894409937887, "grad_norm": 0.26483842730522156, "learning_rate": 5.297448840329329e-05, "loss": 0.0205, "step": 5070 }, { "epoch": 10.5175983436853, "grad_norm": 0.3661891520023346, "learning_rate": 5.280941830159227e-05, "loss": 0.0187, "step": 5080 }, { "epoch": 10.538302277432713, "grad_norm": 0.2463710755109787, "learning_rate": 5.264431747654284e-05, "loss": 0.0218, "step": 5090 }, { "epoch": 10.559006211180124, "grad_norm": 0.19206295907497406, "learning_rate": 5.247918773366112e-05, "loss": 0.0234, "step": 5100 }, { "epoch": 10.579710144927537, "grad_norm": 0.37339940667152405, "learning_rate": 5.231403087877955e-05, "loss": 0.0196, "step": 5110 }, { "epoch": 10.600414078674948, "grad_norm": 0.25263726711273193, "learning_rate": 5.214884871802703e-05, "loss": 0.0181, "step": 5120 }, { "epoch": 10.621118012422361, "grad_norm": 0.22332456707954407, "learning_rate": 5.198364305780922e-05, "loss": 0.0121, "step": 5130 }, { "epoch": 10.641821946169772, "grad_norm": 0.3210849165916443, "learning_rate": 5.1818415704788725e-05, "loss": 0.0203, "step": 5140 }, { "epoch": 10.662525879917185, "grad_norm": 0.2539861798286438, "learning_rate": 5.165316846586541e-05, "loss": 0.019, "step": 5150 }, { "epoch": 10.683229813664596, "grad_norm": 0.17158880829811096, "learning_rate": 5.148790314815663e-05, "loss": 0.0167, "step": 5160 }, { "epoch": 10.703933747412009, "grad_norm": 0.3758942484855652, "learning_rate": 5.132262155897739e-05, "loss": 0.0195, "step": 5170 }, { "epoch": 10.72463768115942, "grad_norm": 0.19196736812591553, "learning_rate": 5.1157325505820694e-05, "loss": 0.0165, "step": 5180 }, { "epoch": 10.745341614906833, "grad_norm": 0.26753002405166626, "learning_rate": 5.0992016796337686e-05, "loss": 0.0172, "step": 5190 }, { "epoch": 10.766045548654244, "grad_norm": 0.3296034038066864, "learning_rate": 5.0826697238317935e-05, "loss": 0.0187, "step": 5200 }, { "epoch": 10.786749482401657, "grad_norm": 0.3748403787612915, "learning_rate": 5.066136863966963e-05, "loss": 0.0166, "step": 5210 }, { "epoch": 10.807453416149068, "grad_norm": 0.15547168254852295, "learning_rate": 5.0496032808399815e-05, "loss": 0.013, "step": 5220 }, { "epoch": 10.82815734989648, "grad_norm": 0.32950443029403687, "learning_rate": 5.033069155259471e-05, "loss": 0.0177, "step": 5230 }, { "epoch": 10.848861283643892, "grad_norm": 0.2860148251056671, "learning_rate": 5.016534668039976e-05, "loss": 0.0158, "step": 5240 }, { "epoch": 10.869565217391305, "grad_norm": 0.2898460626602173, "learning_rate": 5e-05, "loss": 0.0124, "step": 5250 }, { "epoch": 10.890269151138716, "grad_norm": 0.15522050857543945, "learning_rate": 4.9834653319600246e-05, "loss": 0.0192, "step": 5260 }, { "epoch": 10.910973084886129, "grad_norm": 0.1646585315465927, "learning_rate": 4.96693084474053e-05, "loss": 0.0157, "step": 5270 }, { "epoch": 10.93167701863354, "grad_norm": 0.20039130747318268, "learning_rate": 4.950396719160018e-05, "loss": 0.0106, "step": 5280 }, { "epoch": 10.952380952380953, "grad_norm": 0.3037121891975403, "learning_rate": 4.93386313603304e-05, "loss": 0.0158, "step": 5290 }, { "epoch": 10.973084886128364, "grad_norm": 0.2536084055900574, "learning_rate": 4.917330276168208e-05, "loss": 0.0193, "step": 5300 }, { "epoch": 10.993788819875776, "grad_norm": 0.1764489859342575, "learning_rate": 4.9007983203662326e-05, "loss": 0.0305, "step": 5310 }, { "epoch": 11.014492753623188, "grad_norm": 0.330178827047348, "learning_rate": 4.884267449417931e-05, "loss": 0.0161, "step": 5320 }, { "epoch": 11.0351966873706, "grad_norm": 0.14894415438175201, "learning_rate": 4.867737844102261e-05, "loss": 0.0178, "step": 5330 }, { "epoch": 11.055900621118013, "grad_norm": 0.10195163637399673, "learning_rate": 4.851209685184338e-05, "loss": 0.0125, "step": 5340 }, { "epoch": 11.076604554865424, "grad_norm": 0.25648781657218933, "learning_rate": 4.834683153413459e-05, "loss": 0.0238, "step": 5350 }, { "epoch": 11.097308488612837, "grad_norm": 0.43927186727523804, "learning_rate": 4.818158429521129e-05, "loss": 0.0165, "step": 5360 }, { "epoch": 11.118012422360248, "grad_norm": 0.1805030107498169, "learning_rate": 4.801635694219079e-05, "loss": 0.0108, "step": 5370 }, { "epoch": 11.138716356107661, "grad_norm": 0.24022994935512543, "learning_rate": 4.785115128197298e-05, "loss": 0.0204, "step": 5380 }, { "epoch": 11.159420289855072, "grad_norm": 0.2912791073322296, "learning_rate": 4.7685969121220456e-05, "loss": 0.0171, "step": 5390 }, { "epoch": 11.180124223602485, "grad_norm": 0.3098582923412323, "learning_rate": 4.7520812266338885e-05, "loss": 0.0175, "step": 5400 }, { "epoch": 11.200828157349896, "grad_norm": 0.3121486008167267, "learning_rate": 4.735568252345718e-05, "loss": 0.0158, "step": 5410 }, { "epoch": 11.221532091097309, "grad_norm": 0.35292255878448486, "learning_rate": 4.7190581698407725e-05, "loss": 0.021, "step": 5420 }, { "epoch": 11.24223602484472, "grad_norm": 0.31248387694358826, "learning_rate": 4.702551159670672e-05, "loss": 0.013, "step": 5430 }, { "epoch": 11.262939958592133, "grad_norm": 0.31839171051979065, "learning_rate": 4.6860474023534335e-05, "loss": 0.0206, "step": 5440 }, { "epoch": 11.283643892339544, "grad_norm": 0.3239287734031677, "learning_rate": 4.669547078371504e-05, "loss": 0.0198, "step": 5450 }, { "epoch": 11.304347826086957, "grad_norm": 0.3033500015735626, "learning_rate": 4.65305036816978e-05, "loss": 0.0174, "step": 5460 }, { "epoch": 11.325051759834368, "grad_norm": 0.28079020977020264, "learning_rate": 4.6365574521536445e-05, "loss": 0.0197, "step": 5470 }, { "epoch": 11.345755693581781, "grad_norm": 0.39810672402381897, "learning_rate": 4.620068510686985e-05, "loss": 0.0166, "step": 5480 }, { "epoch": 11.366459627329192, "grad_norm": 0.3433759808540344, "learning_rate": 4.60358372409022e-05, "loss": 0.0131, "step": 5490 }, { "epoch": 11.387163561076605, "grad_norm": 0.27469465136528015, "learning_rate": 4.5871032726383386e-05, "loss": 0.0124, "step": 5500 }, { "epoch": 11.407867494824016, "grad_norm": 0.28122034668922424, "learning_rate": 4.570627336558915e-05, "loss": 0.0144, "step": 5510 }, { "epoch": 11.428571428571429, "grad_norm": 0.3627743124961853, "learning_rate": 4.554156096030149e-05, "loss": 0.0119, "step": 5520 }, { "epoch": 11.44927536231884, "grad_norm": 0.2289089411497116, "learning_rate": 4.537689731178883e-05, "loss": 0.0163, "step": 5530 }, { "epoch": 11.469979296066253, "grad_norm": 0.3457997739315033, "learning_rate": 4.5212284220786494e-05, "loss": 0.0154, "step": 5540 }, { "epoch": 11.490683229813664, "grad_norm": 0.23602600395679474, "learning_rate": 4.504772348747687e-05, "loss": 0.0156, "step": 5550 }, { "epoch": 11.511387163561077, "grad_norm": 0.295601487159729, "learning_rate": 4.488321691146975e-05, "loss": 0.0156, "step": 5560 }, { "epoch": 11.532091097308488, "grad_norm": 0.21376757323741913, "learning_rate": 4.471876629178273e-05, "loss": 0.0164, "step": 5570 }, { "epoch": 11.5527950310559, "grad_norm": 0.12493230402469635, "learning_rate": 4.4554373426821374e-05, "loss": 0.0115, "step": 5580 }, { "epoch": 11.573498964803314, "grad_norm": 0.2876069247722626, "learning_rate": 4.439004011435979e-05, "loss": 0.0144, "step": 5590 }, { "epoch": 11.594202898550725, "grad_norm": 0.3381740152835846, "learning_rate": 4.4225768151520694e-05, "loss": 0.0156, "step": 5600 }, { "epoch": 11.614906832298137, "grad_norm": 0.3720422685146332, "learning_rate": 4.406155933475599e-05, "loss": 0.0177, "step": 5610 }, { "epoch": 11.635610766045549, "grad_norm": 0.34128573536872864, "learning_rate": 4.3897415459827e-05, "loss": 0.018, "step": 5620 }, { "epoch": 11.656314699792961, "grad_norm": 0.26400160789489746, "learning_rate": 4.373333832178478e-05, "loss": 0.0103, "step": 5630 }, { "epoch": 11.677018633540373, "grad_norm": 0.25610440969467163, "learning_rate": 4.3569329714950704e-05, "loss": 0.0156, "step": 5640 }, { "epoch": 11.697722567287785, "grad_norm": 0.4677472710609436, "learning_rate": 4.3405391432896555e-05, "loss": 0.0198, "step": 5650 }, { "epoch": 11.718426501035196, "grad_norm": 0.25295865535736084, "learning_rate": 4.324152526842517e-05, "loss": 0.0148, "step": 5660 }, { "epoch": 11.73913043478261, "grad_norm": 0.36017468571662903, "learning_rate": 4.307773301355062e-05, "loss": 0.0105, "step": 5670 }, { "epoch": 11.75983436853002, "grad_norm": 0.2575490474700928, "learning_rate": 4.291401645947879e-05, "loss": 0.0205, "step": 5680 }, { "epoch": 11.780538302277433, "grad_norm": 0.3439244329929352, "learning_rate": 4.275037739658771e-05, "loss": 0.0093, "step": 5690 }, { "epoch": 11.801242236024844, "grad_norm": 0.36050695180892944, "learning_rate": 4.2586817614407895e-05, "loss": 0.0184, "step": 5700 }, { "epoch": 11.821946169772257, "grad_norm": 0.21288472414016724, "learning_rate": 4.2423338901602985e-05, "loss": 0.021, "step": 5710 }, { "epoch": 11.842650103519668, "grad_norm": 0.3182883858680725, "learning_rate": 4.2259943045949934e-05, "loss": 0.0135, "step": 5720 }, { "epoch": 11.863354037267081, "grad_norm": 0.2907312512397766, "learning_rate": 4.209663183431969e-05, "loss": 0.0157, "step": 5730 }, { "epoch": 11.884057971014492, "grad_norm": 0.2787594199180603, "learning_rate": 4.1933407052657456e-05, "loss": 0.0106, "step": 5740 }, { "epoch": 11.904761904761905, "grad_norm": 0.2142610400915146, "learning_rate": 4.17702704859633e-05, "loss": 0.0133, "step": 5750 }, { "epoch": 11.925465838509316, "grad_norm": 0.13525636494159698, "learning_rate": 4.160722391827262e-05, "loss": 0.0134, "step": 5760 }, { "epoch": 11.946169772256729, "grad_norm": 0.2539882957935333, "learning_rate": 4.14442691326365e-05, "loss": 0.0128, "step": 5770 }, { "epoch": 11.96687370600414, "grad_norm": 0.23852507770061493, "learning_rate": 4.1281407911102425e-05, "loss": 0.0138, "step": 5780 }, { "epoch": 11.987577639751553, "grad_norm": 0.2654222548007965, "learning_rate": 4.111864203469457e-05, "loss": 0.0158, "step": 5790 }, { "epoch": 12.008281573498964, "grad_norm": 0.2055487334728241, "learning_rate": 4.095597328339452e-05, "loss": 0.0141, "step": 5800 }, { "epoch": 12.028985507246377, "grad_norm": 0.42218226194381714, "learning_rate": 4.079340343612165e-05, "loss": 0.0227, "step": 5810 }, { "epoch": 12.049689440993788, "grad_norm": 0.44103577733039856, "learning_rate": 4.063093427071376e-05, "loss": 0.0277, "step": 5820 }, { "epoch": 12.0703933747412, "grad_norm": 0.3023831248283386, "learning_rate": 4.046856756390767e-05, "loss": 0.0153, "step": 5830 }, { "epoch": 12.091097308488612, "grad_norm": 0.1203257143497467, "learning_rate": 4.0306305091319595e-05, "loss": 0.0114, "step": 5840 }, { "epoch": 12.111801242236025, "grad_norm": 0.25661325454711914, "learning_rate": 4.0144148627425993e-05, "loss": 0.0154, "step": 5850 }, { "epoch": 12.132505175983438, "grad_norm": 0.14174573123455048, "learning_rate": 3.9982099945543945e-05, "loss": 0.0109, "step": 5860 }, { "epoch": 12.153209109730849, "grad_norm": 0.152806356549263, "learning_rate": 3.982016081781189e-05, "loss": 0.0106, "step": 5870 }, { "epoch": 12.173913043478262, "grad_norm": 0.20406414568424225, "learning_rate": 3.965833301517017e-05, "loss": 0.0179, "step": 5880 }, { "epoch": 12.194616977225673, "grad_norm": 0.21962039172649384, "learning_rate": 3.949661830734172e-05, "loss": 0.013, "step": 5890 }, { "epoch": 12.215320910973086, "grad_norm": 0.11857223510742188, "learning_rate": 3.933501846281267e-05, "loss": 0.0141, "step": 5900 }, { "epoch": 12.236024844720497, "grad_norm": 0.3677363693714142, "learning_rate": 3.917353524881302e-05, "loss": 0.0178, "step": 5910 }, { "epoch": 12.25672877846791, "grad_norm": 0.15321829915046692, "learning_rate": 3.901217043129735e-05, "loss": 0.0143, "step": 5920 }, { "epoch": 12.27743271221532, "grad_norm": 0.2945491373538971, "learning_rate": 3.8850925774925425e-05, "loss": 0.016, "step": 5930 }, { "epoch": 12.298136645962733, "grad_norm": 0.22627153992652893, "learning_rate": 3.8689803043043e-05, "loss": 0.0142, "step": 5940 }, { "epoch": 12.318840579710145, "grad_norm": 0.21437634527683258, "learning_rate": 3.852880399766243e-05, "loss": 0.0132, "step": 5950 }, { "epoch": 12.339544513457557, "grad_norm": 0.2832322120666504, "learning_rate": 3.836793039944349e-05, "loss": 0.0136, "step": 5960 }, { "epoch": 12.360248447204969, "grad_norm": 0.3467109799385071, "learning_rate": 3.820718400767409e-05, "loss": 0.0127, "step": 5970 }, { "epoch": 12.380952380952381, "grad_norm": 0.17044273018836975, "learning_rate": 3.8046566580251e-05, "loss": 0.008, "step": 5980 }, { "epoch": 12.401656314699792, "grad_norm": 0.16087700426578522, "learning_rate": 3.788607987366069e-05, "loss": 0.0119, "step": 5990 }, { "epoch": 12.422360248447205, "grad_norm": 0.3066701889038086, "learning_rate": 3.772572564296005e-05, "loss": 0.017, "step": 6000 }, { "epoch": 12.443064182194616, "grad_norm": 0.30797335505485535, "learning_rate": 3.756550564175727e-05, "loss": 0.0145, "step": 6010 }, { "epoch": 12.46376811594203, "grad_norm": 0.33485573530197144, "learning_rate": 3.74054216221926e-05, "loss": 0.0152, "step": 6020 }, { "epoch": 12.48447204968944, "grad_norm": 0.22613269090652466, "learning_rate": 3.7245475334919246e-05, "loss": 0.0136, "step": 6030 }, { "epoch": 12.505175983436853, "grad_norm": 0.31739288568496704, "learning_rate": 3.7085668529084184e-05, "loss": 0.0176, "step": 6040 }, { "epoch": 12.525879917184264, "grad_norm": 0.17527784407138824, "learning_rate": 3.6926002952309016e-05, "loss": 0.014, "step": 6050 }, { "epoch": 12.546583850931677, "grad_norm": 0.25045859813690186, "learning_rate": 3.676648035067093e-05, "loss": 0.012, "step": 6060 }, { "epoch": 12.567287784679088, "grad_norm": 0.26200345158576965, "learning_rate": 3.6607102468683526e-05, "loss": 0.0152, "step": 6070 }, { "epoch": 12.587991718426501, "grad_norm": 0.23731692135334015, "learning_rate": 3.6447871049277796e-05, "loss": 0.0118, "step": 6080 }, { "epoch": 12.608695652173914, "grad_norm": 0.2703661024570465, "learning_rate": 3.628878783378302e-05, "loss": 0.0146, "step": 6090 }, { "epoch": 12.629399585921325, "grad_norm": 0.09868067502975464, "learning_rate": 3.612985456190778e-05, "loss": 0.0165, "step": 6100 }, { "epoch": 12.650103519668736, "grad_norm": 0.19754783809185028, "learning_rate": 3.597107297172084e-05, "loss": 0.0179, "step": 6110 }, { "epoch": 12.670807453416149, "grad_norm": 0.248275488615036, "learning_rate": 3.581244479963225e-05, "loss": 0.0113, "step": 6120 }, { "epoch": 12.691511387163562, "grad_norm": 0.21552643179893494, "learning_rate": 3.5653971780374295e-05, "loss": 0.015, "step": 6130 }, { "epoch": 12.712215320910973, "grad_norm": 0.18627336621284485, "learning_rate": 3.5495655646982505e-05, "loss": 0.0139, "step": 6140 }, { "epoch": 12.732919254658386, "grad_norm": 0.1864493191242218, "learning_rate": 3.533749813077677e-05, "loss": 0.0121, "step": 6150 }, { "epoch": 12.753623188405797, "grad_norm": 0.3911803662776947, "learning_rate": 3.517950096134232e-05, "loss": 0.0125, "step": 6160 }, { "epoch": 12.77432712215321, "grad_norm": 0.3846626579761505, "learning_rate": 3.5021665866510925e-05, "loss": 0.0136, "step": 6170 }, { "epoch": 12.79503105590062, "grad_norm": 0.248709574341774, "learning_rate": 3.4863994572341843e-05, "loss": 0.0145, "step": 6180 }, { "epoch": 12.815734989648034, "grad_norm": 0.2012386918067932, "learning_rate": 3.470648880310313e-05, "loss": 0.0099, "step": 6190 }, { "epoch": 12.836438923395445, "grad_norm": 0.21333587169647217, "learning_rate": 3.4549150281252636e-05, "loss": 0.0147, "step": 6200 }, { "epoch": 12.857142857142858, "grad_norm": 0.3370843231678009, "learning_rate": 3.439198072741921e-05, "loss": 0.0146, "step": 6210 }, { "epoch": 12.877846790890269, "grad_norm": 0.17124126851558685, "learning_rate": 3.423498186038393e-05, "loss": 0.0183, "step": 6220 }, { "epoch": 12.898550724637682, "grad_norm": 0.28277552127838135, "learning_rate": 3.407815539706124e-05, "loss": 0.0139, "step": 6230 }, { "epoch": 12.919254658385093, "grad_norm": 0.21696391701698303, "learning_rate": 3.392150305248024e-05, "loss": 0.0169, "step": 6240 }, { "epoch": 12.939958592132506, "grad_norm": 0.11720360070466995, "learning_rate": 3.3765026539765834e-05, "loss": 0.0131, "step": 6250 }, { "epoch": 12.960662525879917, "grad_norm": 0.16178110241889954, "learning_rate": 3.360872757012011e-05, "loss": 0.0123, "step": 6260 }, { "epoch": 12.98136645962733, "grad_norm": 0.3274001479148865, "learning_rate": 3.3452607852803584e-05, "loss": 0.0149, "step": 6270 }, { "epoch": 13.00207039337474, "grad_norm": 0.1669159233570099, "learning_rate": 3.329666909511645e-05, "loss": 0.0106, "step": 6280 }, { "epoch": 13.022774327122153, "grad_norm": 0.16714906692504883, "learning_rate": 3.3140913002379995e-05, "loss": 0.0107, "step": 6290 }, { "epoch": 13.043478260869565, "grad_norm": 0.3376113176345825, "learning_rate": 3.298534127791785e-05, "loss": 0.0137, "step": 6300 }, { "epoch": 13.064182194616977, "grad_norm": 0.33254274725914, "learning_rate": 3.282995562303754e-05, "loss": 0.0123, "step": 6310 }, { "epoch": 13.084886128364388, "grad_norm": 0.2172705978155136, "learning_rate": 3.267475773701161e-05, "loss": 0.009, "step": 6320 }, { "epoch": 13.105590062111801, "grad_norm": 0.3303101658821106, "learning_rate": 3.251974931705933e-05, "loss": 0.0135, "step": 6330 }, { "epoch": 13.126293995859212, "grad_norm": 0.20792706310749054, "learning_rate": 3.236493205832795e-05, "loss": 0.015, "step": 6340 }, { "epoch": 13.146997929606625, "grad_norm": 0.2666330337524414, "learning_rate": 3.221030765387417e-05, "loss": 0.0112, "step": 6350 }, { "epoch": 13.167701863354038, "grad_norm": 0.14030727744102478, "learning_rate": 3.205587779464576e-05, "loss": 0.0095, "step": 6360 }, { "epoch": 13.18840579710145, "grad_norm": 0.2144712507724762, "learning_rate": 3.190164416946285e-05, "loss": 0.0127, "step": 6370 }, { "epoch": 13.209109730848862, "grad_norm": 0.258699506521225, "learning_rate": 3.1747608464999725e-05, "loss": 0.0113, "step": 6380 }, { "epoch": 13.229813664596273, "grad_norm": 0.19403113424777985, "learning_rate": 3.1593772365766105e-05, "loss": 0.0154, "step": 6390 }, { "epoch": 13.250517598343686, "grad_norm": 0.31140780448913574, "learning_rate": 3.144013755408895e-05, "loss": 0.0143, "step": 6400 }, { "epoch": 13.271221532091097, "grad_norm": 0.170811265707016, "learning_rate": 3.128670571009399e-05, "loss": 0.0164, "step": 6410 }, { "epoch": 13.29192546583851, "grad_norm": 0.15772926807403564, "learning_rate": 3.113347851168721e-05, "loss": 0.0094, "step": 6420 }, { "epoch": 13.312629399585921, "grad_norm": 0.15099747478961945, "learning_rate": 3.098045763453678e-05, "loss": 0.009, "step": 6430 }, { "epoch": 13.333333333333334, "grad_norm": 0.23406678438186646, "learning_rate": 3.082764475205442e-05, "loss": 0.0083, "step": 6440 }, { "epoch": 13.354037267080745, "grad_norm": 0.28785207867622375, "learning_rate": 3.0675041535377405e-05, "loss": 0.0131, "step": 6450 }, { "epoch": 13.374741200828158, "grad_norm": 0.13396427035331726, "learning_rate": 3.052264965335e-05, "loss": 0.0087, "step": 6460 }, { "epoch": 13.395445134575569, "grad_norm": 0.1653931587934494, "learning_rate": 3.0370470772505433e-05, "loss": 0.0128, "step": 6470 }, { "epoch": 13.416149068322982, "grad_norm": 0.1561514437198639, "learning_rate": 3.0218506557047598e-05, "loss": 0.0091, "step": 6480 }, { "epoch": 13.436853002070393, "grad_norm": 0.24145016074180603, "learning_rate": 3.006675866883275e-05, "loss": 0.0087, "step": 6490 }, { "epoch": 13.457556935817806, "grad_norm": 0.09539690613746643, "learning_rate": 2.991522876735154e-05, "loss": 0.0116, "step": 6500 }, { "epoch": 13.478260869565217, "grad_norm": 0.2703956365585327, "learning_rate": 2.976391850971065e-05, "loss": 0.0214, "step": 6510 }, { "epoch": 13.49896480331263, "grad_norm": 0.2412251979112625, "learning_rate": 2.9612829550614836e-05, "loss": 0.0198, "step": 6520 }, { "epoch": 13.51966873706004, "grad_norm": 0.21599094569683075, "learning_rate": 2.9461963542348737e-05, "loss": 0.0072, "step": 6530 }, { "epoch": 13.540372670807454, "grad_norm": 0.25855210423469543, "learning_rate": 2.931132213475884e-05, "loss": 0.0141, "step": 6540 }, { "epoch": 13.561076604554865, "grad_norm": 0.12804104387760162, "learning_rate": 2.916090697523549e-05, "loss": 0.0155, "step": 6550 }, { "epoch": 13.581780538302278, "grad_norm": 0.16140198707580566, "learning_rate": 2.9010719708694722e-05, "loss": 0.0078, "step": 6560 }, { "epoch": 13.602484472049689, "grad_norm": 0.3511630594730377, "learning_rate": 2.8860761977560436e-05, "loss": 0.013, "step": 6570 }, { "epoch": 13.623188405797102, "grad_norm": 0.14886726438999176, "learning_rate": 2.8711035421746367e-05, "loss": 0.0102, "step": 6580 }, { "epoch": 13.643892339544513, "grad_norm": 0.24444444477558136, "learning_rate": 2.8561541678638142e-05, "loss": 0.0104, "step": 6590 }, { "epoch": 13.664596273291925, "grad_norm": 0.24820072948932648, "learning_rate": 2.8412282383075363e-05, "loss": 0.0102, "step": 6600 }, { "epoch": 13.685300207039337, "grad_norm": 0.2242002636194229, "learning_rate": 2.8263259167333777e-05, "loss": 0.007, "step": 6610 }, { "epoch": 13.70600414078675, "grad_norm": 0.16944734752178192, "learning_rate": 2.811447366110741e-05, "loss": 0.0131, "step": 6620 }, { "epoch": 13.726708074534162, "grad_norm": 0.27491500973701477, "learning_rate": 2.7965927491490705e-05, "loss": 0.0128, "step": 6630 }, { "epoch": 13.747412008281573, "grad_norm": 0.23044107854366302, "learning_rate": 2.7817622282960815e-05, "loss": 0.0107, "step": 6640 }, { "epoch": 13.768115942028986, "grad_norm": 0.2772413492202759, "learning_rate": 2.766955965735968e-05, "loss": 0.0086, "step": 6650 }, { "epoch": 13.788819875776397, "grad_norm": 0.30441394448280334, "learning_rate": 2.7521741233876496e-05, "loss": 0.0108, "step": 6660 }, { "epoch": 13.80952380952381, "grad_norm": 0.14362414181232452, "learning_rate": 2.7374168629029813e-05, "loss": 0.0117, "step": 6670 }, { "epoch": 13.830227743271221, "grad_norm": 0.19836893677711487, "learning_rate": 2.7226843456650037e-05, "loss": 0.0077, "step": 6680 }, { "epoch": 13.850931677018634, "grad_norm": 0.16324682533740997, "learning_rate": 2.707976732786166e-05, "loss": 0.0117, "step": 6690 }, { "epoch": 13.871635610766045, "grad_norm": 0.35979026556015015, "learning_rate": 2.693294185106562e-05, "loss": 0.0185, "step": 6700 }, { "epoch": 13.892339544513458, "grad_norm": 0.19380564987659454, "learning_rate": 2.6786368631921836e-05, "loss": 0.0097, "step": 6710 }, { "epoch": 13.91304347826087, "grad_norm": 0.15711010992527008, "learning_rate": 2.6640049273331515e-05, "loss": 0.0087, "step": 6720 }, { "epoch": 13.933747412008282, "grad_norm": 0.26416146755218506, "learning_rate": 2.6493985375419778e-05, "loss": 0.018, "step": 6730 }, { "epoch": 13.954451345755693, "grad_norm": 0.12716226279735565, "learning_rate": 2.6348178535517966e-05, "loss": 0.0123, "step": 6740 }, { "epoch": 13.975155279503106, "grad_norm": 0.2648696303367615, "learning_rate": 2.6202630348146324e-05, "loss": 0.0142, "step": 6750 }, { "epoch": 13.995859213250517, "grad_norm": 0.23273397982120514, "learning_rate": 2.6057342404996522e-05, "loss": 0.021, "step": 6760 }, { "epoch": 14.01656314699793, "grad_norm": 0.22148902714252472, "learning_rate": 2.591231629491423e-05, "loss": 0.0163, "step": 6770 }, { "epoch": 14.037267080745341, "grad_norm": 0.408302903175354, "learning_rate": 2.5767553603881767e-05, "loss": 0.0136, "step": 6780 }, { "epoch": 14.057971014492754, "grad_norm": 0.36084645986557007, "learning_rate": 2.562305591500069e-05, "loss": 0.0091, "step": 6790 }, { "epoch": 14.078674948240165, "grad_norm": 0.2885228991508484, "learning_rate": 2.547882480847461e-05, "loss": 0.0193, "step": 6800 }, { "epoch": 14.099378881987578, "grad_norm": 0.3389575183391571, "learning_rate": 2.5334861861591753e-05, "loss": 0.0139, "step": 6810 }, { "epoch": 14.120082815734989, "grad_norm": 0.4300172030925751, "learning_rate": 2.5191168648707887e-05, "loss": 0.0101, "step": 6820 }, { "epoch": 14.140786749482402, "grad_norm": 0.16732239723205566, "learning_rate": 2.5047746741228978e-05, "loss": 0.0156, "step": 6830 }, { "epoch": 14.161490683229813, "grad_norm": 0.22943831980228424, "learning_rate": 2.490459770759398e-05, "loss": 0.0099, "step": 6840 }, { "epoch": 14.182194616977226, "grad_norm": 0.15673941373825073, "learning_rate": 2.476172311325783e-05, "loss": 0.0096, "step": 6850 }, { "epoch": 14.202898550724637, "grad_norm": 0.28810983896255493, "learning_rate": 2.4619124520674146e-05, "loss": 0.0096, "step": 6860 }, { "epoch": 14.22360248447205, "grad_norm": 0.18461503088474274, "learning_rate": 2.447680348927837e-05, "loss": 0.0087, "step": 6870 }, { "epoch": 14.244306418219463, "grad_norm": 0.4287914037704468, "learning_rate": 2.433476157547044e-05, "loss": 0.0179, "step": 6880 }, { "epoch": 14.265010351966874, "grad_norm": 0.4492188096046448, "learning_rate": 2.419300033259798e-05, "loss": 0.0098, "step": 6890 }, { "epoch": 14.285714285714286, "grad_norm": 0.1756904274225235, "learning_rate": 2.405152131093926e-05, "loss": 0.0103, "step": 6900 }, { "epoch": 14.306418219461698, "grad_norm": 0.2844470143318176, "learning_rate": 2.3910326057686127e-05, "loss": 0.0179, "step": 6910 }, { "epoch": 14.32712215320911, "grad_norm": 0.17838165163993835, "learning_rate": 2.3769416116927335e-05, "loss": 0.0133, "step": 6920 }, { "epoch": 14.347826086956522, "grad_norm": 0.08132132887840271, "learning_rate": 2.362879302963135e-05, "loss": 0.0108, "step": 6930 }, { "epoch": 14.368530020703934, "grad_norm": 0.17442704737186432, "learning_rate": 2.3488458333629777e-05, "loss": 0.0112, "step": 6940 }, { "epoch": 14.389233954451345, "grad_norm": 0.14526265859603882, "learning_rate": 2.3348413563600325e-05, "loss": 0.0122, "step": 6950 }, { "epoch": 14.409937888198758, "grad_norm": 0.1873375028371811, "learning_rate": 2.3208660251050158e-05, "loss": 0.0135, "step": 6960 }, { "epoch": 14.43064182194617, "grad_norm": 0.19545254111289978, "learning_rate": 2.3069199924299174e-05, "loss": 0.0114, "step": 6970 }, { "epoch": 14.451345755693582, "grad_norm": 0.22550365328788757, "learning_rate": 2.29300341084631e-05, "loss": 0.0084, "step": 6980 }, { "epoch": 14.472049689440993, "grad_norm": 0.23783229291439056, "learning_rate": 2.279116432543705e-05, "loss": 0.018, "step": 6990 }, { "epoch": 14.492753623188406, "grad_norm": 0.23706865310668945, "learning_rate": 2.2652592093878666e-05, "loss": 0.0117, "step": 7000 }, { "epoch": 14.513457556935817, "grad_norm": 0.26092207431793213, "learning_rate": 2.251431892919171e-05, "loss": 0.0152, "step": 7010 }, { "epoch": 14.53416149068323, "grad_norm": 0.1474156379699707, "learning_rate": 2.237634634350934e-05, "loss": 0.0109, "step": 7020 }, { "epoch": 14.554865424430641, "grad_norm": 0.17140665650367737, "learning_rate": 2.2238675845677663e-05, "loss": 0.0091, "step": 7030 }, { "epoch": 14.575569358178054, "grad_norm": 0.2452748715877533, "learning_rate": 2.2101308941239203e-05, "loss": 0.011, "step": 7040 }, { "epoch": 14.596273291925465, "grad_norm": 0.17611108720302582, "learning_rate": 2.196424713241637e-05, "loss": 0.0118, "step": 7050 }, { "epoch": 14.616977225672878, "grad_norm": 0.19471800327301025, "learning_rate": 2.182749191809518e-05, "loss": 0.0103, "step": 7060 }, { "epoch": 14.63768115942029, "grad_norm": 0.1733923703432083, "learning_rate": 2.1691044793808734e-05, "loss": 0.0155, "step": 7070 }, { "epoch": 14.658385093167702, "grad_norm": 0.14660657942295074, "learning_rate": 2.1554907251720945e-05, "loss": 0.0135, "step": 7080 }, { "epoch": 14.679089026915113, "grad_norm": 0.16407570242881775, "learning_rate": 2.1419080780610123e-05, "loss": 0.0118, "step": 7090 }, { "epoch": 14.699792960662526, "grad_norm": 0.19673511385917664, "learning_rate": 2.128356686585282e-05, "loss": 0.0146, "step": 7100 }, { "epoch": 14.720496894409937, "grad_norm": 0.11931564658880234, "learning_rate": 2.1148366989407496e-05, "loss": 0.0136, "step": 7110 }, { "epoch": 14.74120082815735, "grad_norm": 0.3042986989021301, "learning_rate": 2.1013482629798333e-05, "loss": 0.0069, "step": 7120 }, { "epoch": 14.761904761904763, "grad_norm": 0.23372496664524078, "learning_rate": 2.0878915262099098e-05, "loss": 0.0112, "step": 7130 }, { "epoch": 14.782608695652174, "grad_norm": 0.09996534138917923, "learning_rate": 2.0744666357916925e-05, "loss": 0.0104, "step": 7140 }, { "epoch": 14.803312629399587, "grad_norm": 0.29712775349617004, "learning_rate": 2.061073738537635e-05, "loss": 0.0112, "step": 7150 }, { "epoch": 14.824016563146998, "grad_norm": 0.18898549675941467, "learning_rate": 2.0477129809103147e-05, "loss": 0.0136, "step": 7160 }, { "epoch": 14.84472049689441, "grad_norm": 0.36737433075904846, "learning_rate": 2.0343845090208368e-05, "loss": 0.0086, "step": 7170 }, { "epoch": 14.865424430641822, "grad_norm": 0.1691131293773651, "learning_rate": 2.0210884686272368e-05, "loss": 0.0094, "step": 7180 }, { "epoch": 14.886128364389235, "grad_norm": 0.4129745662212372, "learning_rate": 2.0078250051328784e-05, "loss": 0.0089, "step": 7190 }, { "epoch": 14.906832298136646, "grad_norm": 0.2072894424200058, "learning_rate": 1.9945942635848748e-05, "loss": 0.0157, "step": 7200 }, { "epoch": 14.927536231884059, "grad_norm": 0.20162107050418854, "learning_rate": 1.981396388672496e-05, "loss": 0.0087, "step": 7210 }, { "epoch": 14.94824016563147, "grad_norm": 0.10208477079868317, "learning_rate": 1.9682315247255894e-05, "loss": 0.0073, "step": 7220 }, { "epoch": 14.968944099378882, "grad_norm": 0.13886311650276184, "learning_rate": 1.9550998157129946e-05, "loss": 0.0094, "step": 7230 }, { "epoch": 14.989648033126294, "grad_norm": 0.13782455027103424, "learning_rate": 1.942001405240979e-05, "loss": 0.0121, "step": 7240 }, { "epoch": 15.010351966873706, "grad_norm": 0.12699589133262634, "learning_rate": 1.928936436551661e-05, "loss": 0.0108, "step": 7250 }, { "epoch": 15.031055900621118, "grad_norm": 0.2035098373889923, "learning_rate": 1.9159050525214452e-05, "loss": 0.0118, "step": 7260 }, { "epoch": 15.05175983436853, "grad_norm": 0.35179731249809265, "learning_rate": 1.9029073956594606e-05, "loss": 0.016, "step": 7270 }, { "epoch": 15.072463768115941, "grad_norm": 0.2005324363708496, "learning_rate": 1.8899436081059975e-05, "loss": 0.0091, "step": 7280 }, { "epoch": 15.093167701863354, "grad_norm": 0.19326607882976532, "learning_rate": 1.877013831630961e-05, "loss": 0.0093, "step": 7290 }, { "epoch": 15.113871635610765, "grad_norm": 0.33832287788391113, "learning_rate": 1.8641182076323148e-05, "loss": 0.0181, "step": 7300 }, { "epoch": 15.134575569358178, "grad_norm": 0.1692480891942978, "learning_rate": 1.851256877134538e-05, "loss": 0.0055, "step": 7310 }, { "epoch": 15.15527950310559, "grad_norm": 0.11265581101179123, "learning_rate": 1.838429980787081e-05, "loss": 0.0087, "step": 7320 }, { "epoch": 15.175983436853002, "grad_norm": 0.21832120418548584, "learning_rate": 1.8256376588628238e-05, "loss": 0.0084, "step": 7330 }, { "epoch": 15.196687370600413, "grad_norm": 0.14270982146263123, "learning_rate": 1.8128800512565513e-05, "loss": 0.0083, "step": 7340 }, { "epoch": 15.217391304347826, "grad_norm": 0.17222732305526733, "learning_rate": 1.800157297483417e-05, "loss": 0.0121, "step": 7350 }, { "epoch": 15.238095238095237, "grad_norm": 0.3032512068748474, "learning_rate": 1.787469536677419e-05, "loss": 0.0186, "step": 7360 }, { "epoch": 15.25879917184265, "grad_norm": 0.22801370918750763, "learning_rate": 1.774816907589873e-05, "loss": 0.0115, "step": 7370 }, { "epoch": 15.279503105590063, "grad_norm": 0.2955957055091858, "learning_rate": 1.7621995485879062e-05, "loss": 0.0121, "step": 7380 }, { "epoch": 15.300207039337474, "grad_norm": 0.3061939775943756, "learning_rate": 1.749617597652934e-05, "loss": 0.0083, "step": 7390 }, { "epoch": 15.320910973084887, "grad_norm": 0.1889140009880066, "learning_rate": 1.7370711923791567e-05, "loss": 0.007, "step": 7400 }, { "epoch": 15.341614906832298, "grad_norm": 0.2401454746723175, "learning_rate": 1.7245604699720535e-05, "loss": 0.0106, "step": 7410 }, { "epoch": 15.36231884057971, "grad_norm": 0.12350483983755112, "learning_rate": 1.712085567246878e-05, "loss": 0.0186, "step": 7420 }, { "epoch": 15.383022774327122, "grad_norm": 0.27292507886886597, "learning_rate": 1.699646620627168e-05, "loss": 0.0129, "step": 7430 }, { "epoch": 15.403726708074535, "grad_norm": 0.17787718772888184, "learning_rate": 1.6872437661432517e-05, "loss": 0.0101, "step": 7440 }, { "epoch": 15.424430641821946, "grad_norm": 0.1515520066022873, "learning_rate": 1.6748771394307585e-05, "loss": 0.0075, "step": 7450 }, { "epoch": 15.445134575569359, "grad_norm": 0.21004188060760498, "learning_rate": 1.662546875729138e-05, "loss": 0.0086, "step": 7460 }, { "epoch": 15.46583850931677, "grad_norm": 0.1831471472978592, "learning_rate": 1.6502531098801753e-05, "loss": 0.0133, "step": 7470 }, { "epoch": 15.486542443064183, "grad_norm": 0.22139641642570496, "learning_rate": 1.637995976326527e-05, "loss": 0.0139, "step": 7480 }, { "epoch": 15.507246376811594, "grad_norm": 0.20815637707710266, "learning_rate": 1.62577560911024e-05, "loss": 0.0118, "step": 7490 }, { "epoch": 15.527950310559007, "grad_norm": 0.21807079017162323, "learning_rate": 1.6135921418712956e-05, "loss": 0.0108, "step": 7500 }, { "epoch": 15.548654244306418, "grad_norm": 0.11604408174753189, "learning_rate": 1.6014457078461353e-05, "loss": 0.0059, "step": 7510 }, { "epoch": 15.56935817805383, "grad_norm": 0.11558694392442703, "learning_rate": 1.5893364398662176e-05, "loss": 0.0136, "step": 7520 }, { "epoch": 15.590062111801242, "grad_norm": 0.12991271913051605, "learning_rate": 1.5772644703565565e-05, "loss": 0.0163, "step": 7530 }, { "epoch": 15.610766045548655, "grad_norm": 0.20291006565093994, "learning_rate": 1.5652299313342773e-05, "loss": 0.0103, "step": 7540 }, { "epoch": 15.631469979296066, "grad_norm": 0.2377055585384369, "learning_rate": 1.553232954407171e-05, "loss": 0.012, "step": 7550 }, { "epoch": 15.652173913043478, "grad_norm": 0.22580869495868683, "learning_rate": 1.5412736707722537e-05, "loss": 0.0103, "step": 7560 }, { "epoch": 15.67287784679089, "grad_norm": 0.1061275526881218, "learning_rate": 1.5293522112143373e-05, "loss": 0.0061, "step": 7570 }, { "epoch": 15.693581780538302, "grad_norm": 0.16939014196395874, "learning_rate": 1.517468706104589e-05, "loss": 0.0086, "step": 7580 }, { "epoch": 15.714285714285714, "grad_norm": 0.3030497431755066, "learning_rate": 1.5056232853991209e-05, "loss": 0.0123, "step": 7590 }, { "epoch": 15.734989648033126, "grad_norm": 0.10592043399810791, "learning_rate": 1.4938160786375572e-05, "loss": 0.0089, "step": 7600 }, { "epoch": 15.755693581780537, "grad_norm": 0.275499552488327, "learning_rate": 1.4820472149416154e-05, "loss": 0.009, "step": 7610 }, { "epoch": 15.77639751552795, "grad_norm": 0.3377675712108612, "learning_rate": 1.470316823013707e-05, "loss": 0.0074, "step": 7620 }, { "epoch": 15.797101449275363, "grad_norm": 0.25571051239967346, "learning_rate": 1.4586250311355132e-05, "loss": 0.017, "step": 7630 }, { "epoch": 15.817805383022774, "grad_norm": 0.0750851258635521, "learning_rate": 1.4469719671666043e-05, "loss": 0.0078, "step": 7640 }, { "epoch": 15.838509316770187, "grad_norm": 0.34919869899749756, "learning_rate": 1.435357758543015e-05, "loss": 0.0144, "step": 7650 }, { "epoch": 15.859213250517598, "grad_norm": 0.12831534445285797, "learning_rate": 1.4237825322758736e-05, "loss": 0.0133, "step": 7660 }, { "epoch": 15.879917184265011, "grad_norm": 0.25880032777786255, "learning_rate": 1.412246414949997e-05, "loss": 0.0093, "step": 7670 }, { "epoch": 15.900621118012422, "grad_norm": 0.16009213030338287, "learning_rate": 1.4007495327225162e-05, "loss": 0.0082, "step": 7680 }, { "epoch": 15.921325051759835, "grad_norm": 0.08201411366462708, "learning_rate": 1.389292011321498e-05, "loss": 0.008, "step": 7690 }, { "epoch": 15.942028985507246, "grad_norm": 0.124351866543293, "learning_rate": 1.3778739760445552e-05, "loss": 0.009, "step": 7700 }, { "epoch": 15.962732919254659, "grad_norm": 0.2569082975387573, "learning_rate": 1.3664955517574968e-05, "loss": 0.0083, "step": 7710 }, { "epoch": 15.98343685300207, "grad_norm": 0.19009342789649963, "learning_rate": 1.3551568628929434e-05, "loss": 0.0155, "step": 7720 }, { "epoch": 16.00414078674948, "grad_norm": 0.21828222274780273, "learning_rate": 1.343858033448982e-05, "loss": 0.011, "step": 7730 }, { "epoch": 16.024844720496894, "grad_norm": 0.37066009640693665, "learning_rate": 1.3325991869878013e-05, "loss": 0.0093, "step": 7740 }, { "epoch": 16.045548654244307, "grad_norm": 0.20857572555541992, "learning_rate": 1.3213804466343421e-05, "loss": 0.0111, "step": 7750 }, { "epoch": 16.06625258799172, "grad_norm": 0.09337040781974792, "learning_rate": 1.3102019350749528e-05, "loss": 0.0083, "step": 7760 }, { "epoch": 16.08695652173913, "grad_norm": 0.19573694467544556, "learning_rate": 1.299063774556042e-05, "loss": 0.0076, "step": 7770 }, { "epoch": 16.107660455486542, "grad_norm": 0.13350056111812592, "learning_rate": 1.2879660868827508e-05, "loss": 0.0075, "step": 7780 }, { "epoch": 16.128364389233955, "grad_norm": 0.25092849135398865, "learning_rate": 1.2769089934176126e-05, "loss": 0.0132, "step": 7790 }, { "epoch": 16.149068322981368, "grad_norm": 0.22119292616844177, "learning_rate": 1.2658926150792322e-05, "loss": 0.0056, "step": 7800 }, { "epoch": 16.169772256728777, "grad_norm": 0.30450862646102905, "learning_rate": 1.2549170723409549e-05, "loss": 0.0086, "step": 7810 }, { "epoch": 16.19047619047619, "grad_norm": 0.13586996495723724, "learning_rate": 1.243982485229559e-05, "loss": 0.007, "step": 7820 }, { "epoch": 16.211180124223603, "grad_norm": 0.2382475584745407, "learning_rate": 1.233088973323937e-05, "loss": 0.0082, "step": 7830 }, { "epoch": 16.231884057971016, "grad_norm": 0.10683726519346237, "learning_rate": 1.2222366557537911e-05, "loss": 0.0098, "step": 7840 }, { "epoch": 16.252587991718425, "grad_norm": 0.27863967418670654, "learning_rate": 1.2114256511983274e-05, "loss": 0.0088, "step": 7850 }, { "epoch": 16.273291925465838, "grad_norm": 0.16211223602294922, "learning_rate": 1.2006560778849578e-05, "loss": 0.0091, "step": 7860 }, { "epoch": 16.29399585921325, "grad_norm": 0.23455710709095, "learning_rate": 1.1899280535880119e-05, "loss": 0.0113, "step": 7870 }, { "epoch": 16.314699792960663, "grad_norm": 0.2368423193693161, "learning_rate": 1.1792416956274444e-05, "loss": 0.0092, "step": 7880 }, { "epoch": 16.335403726708076, "grad_norm": 0.14468133449554443, "learning_rate": 1.1685971208675539e-05, "loss": 0.0089, "step": 7890 }, { "epoch": 16.356107660455486, "grad_norm": 0.23661158978939056, "learning_rate": 1.157994445715706e-05, "loss": 0.0148, "step": 7900 }, { "epoch": 16.3768115942029, "grad_norm": 0.17524808645248413, "learning_rate": 1.1474337861210543e-05, "loss": 0.0069, "step": 7910 }, { "epoch": 16.39751552795031, "grad_norm": 0.16132678091526031, "learning_rate": 1.1369152575732822e-05, "loss": 0.0057, "step": 7920 }, { "epoch": 16.418219461697724, "grad_norm": 0.28519511222839355, "learning_rate": 1.1264389751013326e-05, "loss": 0.0086, "step": 7930 }, { "epoch": 16.438923395445133, "grad_norm": 0.14455687999725342, "learning_rate": 1.1160050532721528e-05, "loss": 0.0125, "step": 7940 }, { "epoch": 16.459627329192546, "grad_norm": 0.1167943924665451, "learning_rate": 1.1056136061894384e-05, "loss": 0.0073, "step": 7950 }, { "epoch": 16.48033126293996, "grad_norm": 0.12794506549835205, "learning_rate": 1.095264747492391e-05, "loss": 0.0077, "step": 7960 }, { "epoch": 16.501035196687372, "grad_norm": 0.2192164659500122, "learning_rate": 1.0849585903544706e-05, "loss": 0.0129, "step": 7970 }, { "epoch": 16.52173913043478, "grad_norm": 0.13658955693244934, "learning_rate": 1.0746952474821614e-05, "loss": 0.0059, "step": 7980 }, { "epoch": 16.542443064182194, "grad_norm": 0.07518316805362701, "learning_rate": 1.0644748311137376e-05, "loss": 0.0078, "step": 7990 }, { "epoch": 16.563146997929607, "grad_norm": 0.21368205547332764, "learning_rate": 1.0542974530180327e-05, "loss": 0.0103, "step": 8000 }, { "epoch": 16.58385093167702, "grad_norm": 0.11877325922250748, "learning_rate": 1.0441632244932237e-05, "loss": 0.0088, "step": 8010 }, { "epoch": 16.60455486542443, "grad_norm": 0.14218921959400177, "learning_rate": 1.0340722563656107e-05, "loss": 0.0081, "step": 8020 }, { "epoch": 16.625258799171842, "grad_norm": 0.14418555796146393, "learning_rate": 1.0240246589884044e-05, "loss": 0.0066, "step": 8030 }, { "epoch": 16.645962732919255, "grad_norm": 0.2092021405696869, "learning_rate": 1.0140205422405214e-05, "loss": 0.0071, "step": 8040 }, { "epoch": 16.666666666666668, "grad_norm": 0.23222719132900238, "learning_rate": 1.0040600155253765e-05, "loss": 0.0093, "step": 8050 }, { "epoch": 16.687370600414077, "grad_norm": 0.22380509972572327, "learning_rate": 9.941431877696955e-06, "loss": 0.0107, "step": 8060 }, { "epoch": 16.70807453416149, "grad_norm": 0.0874108150601387, "learning_rate": 9.842701674223187e-06, "loss": 0.0098, "step": 8070 }, { "epoch": 16.728778467908903, "grad_norm": 0.1355743557214737, "learning_rate": 9.744410624530148e-06, "loss": 0.0106, "step": 8080 }, { "epoch": 16.749482401656316, "grad_norm": 0.2618853747844696, "learning_rate": 9.646559803512994e-06, "loss": 0.0061, "step": 8090 }, { "epoch": 16.770186335403725, "grad_norm": 0.15244142711162567, "learning_rate": 9.549150281252633e-06, "loss": 0.0085, "step": 8100 }, { "epoch": 16.790890269151138, "grad_norm": 0.13537409901618958, "learning_rate": 9.452183123004e-06, "loss": 0.0114, "step": 8110 }, { "epoch": 16.81159420289855, "grad_norm": 0.19039101898670197, "learning_rate": 9.355659389184396e-06, "loss": 0.0129, "step": 8120 }, { "epoch": 16.832298136645964, "grad_norm": 0.2499101459980011, "learning_rate": 9.259580135361929e-06, "loss": 0.0107, "step": 8130 }, { "epoch": 16.853002070393373, "grad_norm": 0.14106379449367523, "learning_rate": 9.163946412243896e-06, "loss": 0.0108, "step": 8140 }, { "epoch": 16.873706004140786, "grad_norm": 0.15071670711040497, "learning_rate": 9.068759265665384e-06, "loss": 0.0083, "step": 8150 }, { "epoch": 16.8944099378882, "grad_norm": 0.1283147633075714, "learning_rate": 8.974019736577777e-06, "loss": 0.0083, "step": 8160 }, { "epoch": 16.91511387163561, "grad_norm": 0.20437172055244446, "learning_rate": 8.879728861037384e-06, "loss": 0.0109, "step": 8170 }, { "epoch": 16.935817805383024, "grad_norm": 0.10465730726718903, "learning_rate": 8.785887670194138e-06, "loss": 0.0069, "step": 8180 }, { "epoch": 16.956521739130434, "grad_norm": 0.14158622920513153, "learning_rate": 8.692497190280224e-06, "loss": 0.0096, "step": 8190 }, { "epoch": 16.977225672877847, "grad_norm": 0.18538232147693634, "learning_rate": 8.599558442598998e-06, "loss": 0.0091, "step": 8200 }, { "epoch": 16.99792960662526, "grad_norm": 0.21249662339687347, "learning_rate": 8.507072443513702e-06, "loss": 0.0091, "step": 8210 }, { "epoch": 17.018633540372672, "grad_norm": 0.15286517143249512, "learning_rate": 8.415040204436426e-06, "loss": 0.0079, "step": 8220 }, { "epoch": 17.03933747412008, "grad_norm": 0.09370838850736618, "learning_rate": 8.323462731816961e-06, "loss": 0.0069, "step": 8230 }, { "epoch": 17.060041407867494, "grad_norm": 0.25686654448509216, "learning_rate": 8.232341027131885e-06, "loss": 0.009, "step": 8240 }, { "epoch": 17.080745341614907, "grad_norm": 0.19205008447170258, "learning_rate": 8.141676086873572e-06, "loss": 0.0101, "step": 8250 }, { "epoch": 17.10144927536232, "grad_norm": 0.2870364785194397, "learning_rate": 8.051468902539272e-06, "loss": 0.009, "step": 8260 }, { "epoch": 17.12215320910973, "grad_norm": 0.3246857821941376, "learning_rate": 7.96172046062032e-06, "loss": 0.0091, "step": 8270 }, { "epoch": 17.142857142857142, "grad_norm": 0.21676860749721527, "learning_rate": 7.872431742591268e-06, "loss": 0.0096, "step": 8280 }, { "epoch": 17.163561076604555, "grad_norm": 0.25541365146636963, "learning_rate": 7.783603724899257e-06, "loss": 0.0144, "step": 8290 }, { "epoch": 17.184265010351968, "grad_norm": 0.2326376587152481, "learning_rate": 7.695237378953223e-06, "loss": 0.0066, "step": 8300 }, { "epoch": 17.204968944099377, "grad_norm": 0.08917392045259476, "learning_rate": 7.607333671113409e-06, "loss": 0.0041, "step": 8310 }, { "epoch": 17.22567287784679, "grad_norm": 0.0948718711733818, "learning_rate": 7.519893562680663e-06, "loss": 0.0074, "step": 8320 }, { "epoch": 17.246376811594203, "grad_norm": 0.21385958790779114, "learning_rate": 7.432918009885997e-06, "loss": 0.012, "step": 8330 }, { "epoch": 17.267080745341616, "grad_norm": 0.12020477652549744, "learning_rate": 7.3464079638801365e-06, "loss": 0.0113, "step": 8340 }, { "epoch": 17.287784679089025, "grad_norm": 0.191249817609787, "learning_rate": 7.260364370723044e-06, "loss": 0.0076, "step": 8350 }, { "epoch": 17.308488612836438, "grad_norm": 0.26434335112571716, "learning_rate": 7.174788171373731e-06, "loss": 0.0118, "step": 8360 }, { "epoch": 17.32919254658385, "grad_norm": 0.24450653791427612, "learning_rate": 7.089680301679752e-06, "loss": 0.014, "step": 8370 }, { "epoch": 17.349896480331264, "grad_norm": 0.1545344591140747, "learning_rate": 7.005041692367154e-06, "loss": 0.0099, "step": 8380 }, { "epoch": 17.370600414078673, "grad_norm": 0.1290704309940338, "learning_rate": 6.92087326903022e-06, "loss": 0.0058, "step": 8390 }, { "epoch": 17.391304347826086, "grad_norm": 0.17769216001033783, "learning_rate": 6.837175952121306e-06, "loss": 0.0065, "step": 8400 }, { "epoch": 17.4120082815735, "grad_norm": 0.09124590456485748, "learning_rate": 6.753950656940905e-06, "loss": 0.0053, "step": 8410 }, { "epoch": 17.43271221532091, "grad_norm": 0.06734632700681686, "learning_rate": 6.671198293627479e-06, "loss": 0.0053, "step": 8420 }, { "epoch": 17.453416149068325, "grad_norm": 0.2577734887599945, "learning_rate": 6.588919767147639e-06, "loss": 0.0084, "step": 8430 }, { "epoch": 17.474120082815734, "grad_norm": 0.22394207119941711, "learning_rate": 6.5071159772861436e-06, "loss": 0.0047, "step": 8440 }, { "epoch": 17.494824016563147, "grad_norm": 0.11470034718513489, "learning_rate": 6.425787818636131e-06, "loss": 0.0061, "step": 8450 }, { "epoch": 17.51552795031056, "grad_norm": 0.17826999723911285, "learning_rate": 6.344936180589351e-06, "loss": 0.0064, "step": 8460 }, { "epoch": 17.536231884057973, "grad_norm": 0.28449898958206177, "learning_rate": 6.264561947326331e-06, "loss": 0.0075, "step": 8470 }, { "epoch": 17.556935817805382, "grad_norm": 0.1425904631614685, "learning_rate": 6.184665997806832e-06, "loss": 0.0103, "step": 8480 }, { "epoch": 17.577639751552795, "grad_norm": 0.24854503571987152, "learning_rate": 6.1052492057601275e-06, "loss": 0.0066, "step": 8490 }, { "epoch": 17.598343685300208, "grad_norm": 0.3306760787963867, "learning_rate": 6.026312439675552e-06, "loss": 0.0088, "step": 8500 }, { "epoch": 17.61904761904762, "grad_norm": 0.344206303358078, "learning_rate": 5.947856562792925e-06, "loss": 0.0145, "step": 8510 }, { "epoch": 17.63975155279503, "grad_norm": 0.16447465121746063, "learning_rate": 5.869882433093155e-06, "loss": 0.0064, "step": 8520 }, { "epoch": 17.660455486542443, "grad_norm": 0.03391391411423683, "learning_rate": 5.79239090328883e-06, "loss": 0.0051, "step": 8530 }, { "epoch": 17.681159420289855, "grad_norm": 0.1371137499809265, "learning_rate": 5.715382820814885e-06, "loss": 0.0113, "step": 8540 }, { "epoch": 17.70186335403727, "grad_norm": 0.14977498352527618, "learning_rate": 5.6388590278194096e-06, "loss": 0.0071, "step": 8550 }, { "epoch": 17.722567287784678, "grad_norm": 0.11440341174602509, "learning_rate": 5.562820361154314e-06, "loss": 0.0066, "step": 8560 }, { "epoch": 17.74327122153209, "grad_norm": 0.08808150142431259, "learning_rate": 5.48726765236629e-06, "loss": 0.0064, "step": 8570 }, { "epoch": 17.763975155279503, "grad_norm": 0.19239559769630432, "learning_rate": 5.412201727687644e-06, "loss": 0.0072, "step": 8580 }, { "epoch": 17.784679089026916, "grad_norm": 0.12104422599077225, "learning_rate": 5.337623408027293e-06, "loss": 0.0091, "step": 8590 }, { "epoch": 17.805383022774325, "grad_norm": 0.12552820146083832, "learning_rate": 5.263533508961827e-06, "loss": 0.0062, "step": 8600 }, { "epoch": 17.82608695652174, "grad_norm": 0.15666736662387848, "learning_rate": 5.1899328407264855e-06, "loss": 0.0072, "step": 8610 }, { "epoch": 17.84679089026915, "grad_norm": 0.053199708461761475, "learning_rate": 5.116822208206396e-06, "loss": 0.0051, "step": 8620 }, { "epoch": 17.867494824016564, "grad_norm": 0.21540696918964386, "learning_rate": 5.044202410927706e-06, "loss": 0.0128, "step": 8630 }, { "epoch": 17.888198757763973, "grad_norm": 0.09231136739253998, "learning_rate": 4.972074243048897e-06, "loss": 0.0044, "step": 8640 }, { "epoch": 17.908902691511386, "grad_norm": 0.11690150946378708, "learning_rate": 4.900438493352055e-06, "loss": 0.0069, "step": 8650 }, { "epoch": 17.9296066252588, "grad_norm": 0.16514591872692108, "learning_rate": 4.829295945234258e-06, "loss": 0.0048, "step": 8660 }, { "epoch": 17.950310559006212, "grad_norm": 0.27732592821121216, "learning_rate": 4.758647376699032e-06, "loss": 0.0082, "step": 8670 }, { "epoch": 17.971014492753625, "grad_norm": 0.15874800086021423, "learning_rate": 4.688493560347773e-06, "loss": 0.0086, "step": 8680 }, { "epoch": 17.991718426501034, "grad_norm": 0.08021140843629837, "learning_rate": 4.618835263371396e-06, "loss": 0.0066, "step": 8690 }, { "epoch": 18.012422360248447, "grad_norm": 0.436303049325943, "learning_rate": 4.549673247541875e-06, "loss": 0.0083, "step": 8700 }, { "epoch": 18.03312629399586, "grad_norm": 0.18909156322479248, "learning_rate": 4.48100826920394e-06, "loss": 0.0056, "step": 8710 }, { "epoch": 18.053830227743273, "grad_norm": 0.29225167632102966, "learning_rate": 4.412841079266777e-06, "loss": 0.0066, "step": 8720 }, { "epoch": 18.074534161490682, "grad_norm": 0.17670607566833496, "learning_rate": 4.3451724231958644e-06, "loss": 0.0062, "step": 8730 }, { "epoch": 18.095238095238095, "grad_norm": 0.19054444134235382, "learning_rate": 4.27800304100478e-06, "loss": 0.0059, "step": 8740 }, { "epoch": 18.115942028985508, "grad_norm": 0.14144396781921387, "learning_rate": 4.2113336672471245e-06, "loss": 0.0057, "step": 8750 }, { "epoch": 18.13664596273292, "grad_norm": 0.19352711737155914, "learning_rate": 4.145165031008508e-06, "loss": 0.0066, "step": 8760 }, { "epoch": 18.15734989648033, "grad_norm": 0.183538556098938, "learning_rate": 4.079497855898501e-06, "loss": 0.0085, "step": 8770 }, { "epoch": 18.178053830227743, "grad_norm": 0.0954415500164032, "learning_rate": 4.01433286004283e-06, "loss": 0.01, "step": 8780 }, { "epoch": 18.198757763975156, "grad_norm": 0.1649128943681717, "learning_rate": 3.949670756075447e-06, "loss": 0.0094, "step": 8790 }, { "epoch": 18.21946169772257, "grad_norm": 0.282094269990921, "learning_rate": 3.885512251130763e-06, "loss": 0.0111, "step": 8800 }, { "epoch": 18.240165631469978, "grad_norm": 0.13689859211444855, "learning_rate": 3.821858046835913e-06, "loss": 0.0099, "step": 8810 }, { "epoch": 18.26086956521739, "grad_norm": 0.1872493177652359, "learning_rate": 3.75870883930306e-06, "loss": 0.004, "step": 8820 }, { "epoch": 18.281573498964804, "grad_norm": 0.12117026746273041, "learning_rate": 3.696065319121833e-06, "loss": 0.0058, "step": 8830 }, { "epoch": 18.302277432712216, "grad_norm": 0.19425128400325775, "learning_rate": 3.6339281713517303e-06, "loss": 0.0112, "step": 8840 }, { "epoch": 18.322981366459626, "grad_norm": 0.10005170851945877, "learning_rate": 3.5722980755146517e-06, "loss": 0.0106, "step": 8850 }, { "epoch": 18.34368530020704, "grad_norm": 0.282600075006485, "learning_rate": 3.511175705587433e-06, "loss": 0.0073, "step": 8860 }, { "epoch": 18.36438923395445, "grad_norm": 0.1280834972858429, "learning_rate": 3.4505617299945336e-06, "loss": 0.008, "step": 8870 }, { "epoch": 18.385093167701864, "grad_norm": 0.3659622073173523, "learning_rate": 3.390456811600673e-06, "loss": 0.013, "step": 8880 }, { "epoch": 18.405797101449274, "grad_norm": 0.29221051931381226, "learning_rate": 3.3308616077036115e-06, "loss": 0.0106, "step": 8890 }, { "epoch": 18.426501035196686, "grad_norm": 0.13315758109092712, "learning_rate": 3.271776770026963e-06, "loss": 0.0051, "step": 8900 }, { "epoch": 18.4472049689441, "grad_norm": 0.27095624804496765, "learning_rate": 3.213202944713023e-06, "loss": 0.0097, "step": 8910 }, { "epoch": 18.467908902691512, "grad_norm": 0.18304917216300964, "learning_rate": 3.155140772315773e-06, "loss": 0.0069, "step": 8920 }, { "epoch": 18.488612836438925, "grad_norm": 0.1304343044757843, "learning_rate": 3.0975908877938277e-06, "loss": 0.0058, "step": 8930 }, { "epoch": 18.509316770186334, "grad_norm": 0.3121682405471802, "learning_rate": 3.040553920503503e-06, "loss": 0.0072, "step": 8940 }, { "epoch": 18.530020703933747, "grad_norm": 0.12300664931535721, "learning_rate": 2.9840304941919415e-06, "loss": 0.0068, "step": 8950 }, { "epoch": 18.55072463768116, "grad_norm": 0.1591895967721939, "learning_rate": 2.928021226990263e-06, "loss": 0.007, "step": 8960 }, { "epoch": 18.571428571428573, "grad_norm": 0.10691188275814056, "learning_rate": 2.8725267314068495e-06, "loss": 0.005, "step": 8970 }, { "epoch": 18.592132505175982, "grad_norm": 0.22670221328735352, "learning_rate": 2.817547614320615e-06, "loss": 0.0095, "step": 8980 }, { "epoch": 18.612836438923395, "grad_norm": 0.1577608436346054, "learning_rate": 2.7630844769743757e-06, "loss": 0.0094, "step": 8990 }, { "epoch": 18.633540372670808, "grad_norm": 0.07361846417188644, "learning_rate": 2.7091379149682685e-06, "loss": 0.0062, "step": 9000 }, { "epoch": 18.65424430641822, "grad_norm": 0.1678360253572464, "learning_rate": 2.6557085182532582e-06, "loss": 0.0092, "step": 9010 }, { "epoch": 18.67494824016563, "grad_norm": 0.15553714334964752, "learning_rate": 2.602796871124663e-06, "loss": 0.0116, "step": 9020 }, { "epoch": 18.695652173913043, "grad_norm": 0.12781013548374176, "learning_rate": 2.5504035522157854e-06, "loss": 0.0069, "step": 9030 }, { "epoch": 18.716356107660456, "grad_norm": 0.10278218239545822, "learning_rate": 2.4985291344915674e-06, "loss": 0.0068, "step": 9040 }, { "epoch": 18.73706004140787, "grad_norm": 0.31690600514411926, "learning_rate": 2.4471741852423237e-06, "loss": 0.0093, "step": 9050 }, { "epoch": 18.757763975155278, "grad_norm": 0.27544263005256653, "learning_rate": 2.3963392660775575e-06, "loss": 0.0075, "step": 9060 }, { "epoch": 18.77846790890269, "grad_norm": 0.26249682903289795, "learning_rate": 2.3460249329197824e-06, "loss": 0.0076, "step": 9070 }, { "epoch": 18.799171842650104, "grad_norm": 0.09675180912017822, "learning_rate": 2.296231735998511e-06, "loss": 0.0059, "step": 9080 }, { "epoch": 18.819875776397517, "grad_norm": 0.08136120438575745, "learning_rate": 2.2469602198441573e-06, "loss": 0.0049, "step": 9090 }, { "epoch": 18.840579710144926, "grad_norm": 0.17611472308635712, "learning_rate": 2.1982109232821178e-06, "loss": 0.0063, "step": 9100 }, { "epoch": 18.86128364389234, "grad_norm": 0.09379903227090836, "learning_rate": 2.149984379426906e-06, "loss": 0.0064, "step": 9110 }, { "epoch": 18.88198757763975, "grad_norm": 0.05829261615872383, "learning_rate": 2.102281115676258e-06, "loss": 0.0042, "step": 9120 }, { "epoch": 18.902691511387165, "grad_norm": 0.05887362360954285, "learning_rate": 2.0551016537054493e-06, "loss": 0.0091, "step": 9130 }, { "epoch": 18.923395445134574, "grad_norm": 0.08341336995363235, "learning_rate": 2.008446509461498e-06, "loss": 0.0075, "step": 9140 }, { "epoch": 18.944099378881987, "grad_norm": 0.13443836569786072, "learning_rate": 1.962316193157593e-06, "loss": 0.0051, "step": 9150 }, { "epoch": 18.9648033126294, "grad_norm": 0.16308577358722687, "learning_rate": 1.91671120926748e-06, "loss": 0.0084, "step": 9160 }, { "epoch": 18.985507246376812, "grad_norm": 0.080412857234478, "learning_rate": 1.8716320565199618e-06, "loss": 0.0094, "step": 9170 }, { "epoch": 19.006211180124225, "grad_norm": 0.29631340503692627, "learning_rate": 1.8270792278934302e-06, "loss": 0.0051, "step": 9180 }, { "epoch": 19.026915113871635, "grad_norm": 0.15507416427135468, "learning_rate": 1.7830532106104747e-06, "loss": 0.0062, "step": 9190 }, { "epoch": 19.047619047619047, "grad_norm": 0.23660752177238464, "learning_rate": 1.7395544861325718e-06, "loss": 0.0073, "step": 9200 }, { "epoch": 19.06832298136646, "grad_norm": 0.2019672840833664, "learning_rate": 1.696583530154794e-06, "loss": 0.0115, "step": 9210 }, { "epoch": 19.089026915113873, "grad_norm": 0.06961517781019211, "learning_rate": 1.6541408126006463e-06, "loss": 0.0061, "step": 9220 }, { "epoch": 19.109730848861282, "grad_norm": 0.2266739010810852, "learning_rate": 1.6122267976168781e-06, "loss": 0.0104, "step": 9230 }, { "epoch": 19.130434782608695, "grad_norm": 0.10511913150548935, "learning_rate": 1.5708419435684462e-06, "loss": 0.0094, "step": 9240 }, { "epoch": 19.151138716356108, "grad_norm": 0.1996021419763565, "learning_rate": 1.5299867030334814e-06, "loss": 0.0093, "step": 9250 }, { "epoch": 19.17184265010352, "grad_norm": 0.12162107229232788, "learning_rate": 1.4896615227983468e-06, "loss": 0.0044, "step": 9260 }, { "epoch": 19.19254658385093, "grad_norm": 0.09749267995357513, "learning_rate": 1.4498668438527597e-06, "loss": 0.0064, "step": 9270 }, { "epoch": 19.213250517598343, "grad_norm": 0.26853621006011963, "learning_rate": 1.4106031013849496e-06, "loss": 0.0141, "step": 9280 }, { "epoch": 19.233954451345756, "grad_norm": 0.08542772382497787, "learning_rate": 1.3718707247769135e-06, "loss": 0.0049, "step": 9290 }, { "epoch": 19.25465838509317, "grad_norm": 0.4365129768848419, "learning_rate": 1.333670137599713e-06, "loss": 0.01, "step": 9300 }, { "epoch": 19.27536231884058, "grad_norm": 0.3198767602443695, "learning_rate": 1.2960017576088446e-06, "loss": 0.0104, "step": 9310 }, { "epoch": 19.29606625258799, "grad_norm": 0.2250336855649948, "learning_rate": 1.2588659967397e-06, "loss": 0.0175, "step": 9320 }, { "epoch": 19.316770186335404, "grad_norm": 0.22645168006420135, "learning_rate": 1.222263261102985e-06, "loss": 0.0079, "step": 9330 }, { "epoch": 19.337474120082817, "grad_norm": 0.18786205351352692, "learning_rate": 1.1861939509803687e-06, "loss": 0.0074, "step": 9340 }, { "epoch": 19.358178053830226, "grad_norm": 0.1437540352344513, "learning_rate": 1.1506584608200367e-06, "loss": 0.0044, "step": 9350 }, { "epoch": 19.37888198757764, "grad_norm": 0.13397444784641266, "learning_rate": 1.1156571792324211e-06, "loss": 0.0055, "step": 9360 }, { "epoch": 19.399585921325052, "grad_norm": 0.1180369108915329, "learning_rate": 1.0811904889859336e-06, "loss": 0.0147, "step": 9370 }, { "epoch": 19.420289855072465, "grad_norm": 0.13278833031654358, "learning_rate": 1.0472587670027678e-06, "loss": 0.0038, "step": 9380 }, { "epoch": 19.440993788819874, "grad_norm": 0.2261534333229065, "learning_rate": 1.0138623843548078e-06, "loss": 0.0145, "step": 9390 }, { "epoch": 19.461697722567287, "grad_norm": 0.16062599420547485, "learning_rate": 9.810017062595322e-07, "loss": 0.0091, "step": 9400 }, { "epoch": 19.4824016563147, "grad_norm": 0.09465906769037247, "learning_rate": 9.486770920760668e-07, "loss": 0.0079, "step": 9410 }, { "epoch": 19.503105590062113, "grad_norm": 0.06441060453653336, "learning_rate": 9.168888953011989e-07, "loss": 0.0056, "step": 9420 }, { "epoch": 19.523809523809526, "grad_norm": 0.3446926176548004, "learning_rate": 8.856374635655695e-07, "loss": 0.0066, "step": 9430 }, { "epoch": 19.544513457556935, "grad_norm": 0.264846533536911, "learning_rate": 8.549231386298151e-07, "loss": 0.008, "step": 9440 }, { "epoch": 19.565217391304348, "grad_norm": 0.23693622648715973, "learning_rate": 8.247462563808817e-07, "loss": 0.0085, "step": 9450 }, { "epoch": 19.58592132505176, "grad_norm": 0.1094195693731308, "learning_rate": 7.951071468283167e-07, "loss": 0.0042, "step": 9460 }, { "epoch": 19.606625258799173, "grad_norm": 0.058433957397937775, "learning_rate": 7.66006134100672e-07, "loss": 0.0063, "step": 9470 }, { "epoch": 19.627329192546583, "grad_norm": 0.28139904141426086, "learning_rate": 7.374435364419674e-07, "loss": 0.0064, "step": 9480 }, { "epoch": 19.648033126293996, "grad_norm": 0.13973474502563477, "learning_rate": 7.094196662081831e-07, "loss": 0.0054, "step": 9490 }, { "epoch": 19.66873706004141, "grad_norm": 0.27205267548561096, "learning_rate": 6.819348298638839e-07, "loss": 0.0077, "step": 9500 }, { "epoch": 19.68944099378882, "grad_norm": 0.20381984114646912, "learning_rate": 6.549893279788277e-07, "loss": 0.0079, "step": 9510 }, { "epoch": 19.71014492753623, "grad_norm": 0.15530520677566528, "learning_rate": 6.285834552247128e-07, "loss": 0.0062, "step": 9520 }, { "epoch": 19.730848861283643, "grad_norm": 0.05255560576915741, "learning_rate": 6.027175003719354e-07, "loss": 0.0047, "step": 9530 }, { "epoch": 19.751552795031056, "grad_norm": 0.06348340958356857, "learning_rate": 5.773917462864264e-07, "loss": 0.0062, "step": 9540 }, { "epoch": 19.77225672877847, "grad_norm": 0.12964807450771332, "learning_rate": 5.526064699265753e-07, "loss": 0.0127, "step": 9550 }, { "epoch": 19.79296066252588, "grad_norm": 0.1979799121618271, "learning_rate": 5.283619423401998e-07, "loss": 0.0048, "step": 9560 }, { "epoch": 19.81366459627329, "grad_norm": 0.14146266877651215, "learning_rate": 5.046584286615697e-07, "loss": 0.0089, "step": 9570 }, { "epoch": 19.834368530020704, "grad_norm": 0.1563977748155594, "learning_rate": 4.814961881085045e-07, "loss": 0.0042, "step": 9580 }, { "epoch": 19.855072463768117, "grad_norm": 0.04129205644130707, "learning_rate": 4.5887547397955864e-07, "loss": 0.0047, "step": 9590 }, { "epoch": 19.875776397515526, "grad_norm": 0.2608759105205536, "learning_rate": 4.367965336512403e-07, "loss": 0.0134, "step": 9600 }, { "epoch": 19.89648033126294, "grad_norm": 0.16297097504138947, "learning_rate": 4.1525960857530243e-07, "loss": 0.0035, "step": 9610 }, { "epoch": 19.917184265010352, "grad_norm": 0.14169001579284668, "learning_rate": 3.9426493427611177e-07, "loss": 0.0076, "step": 9620 }, { "epoch": 19.937888198757765, "grad_norm": 0.141464963555336, "learning_rate": 3.738127403480507e-07, "loss": 0.0052, "step": 9630 }, { "epoch": 19.958592132505174, "grad_norm": 0.08023510873317719, "learning_rate": 3.5390325045304706e-07, "loss": 0.0055, "step": 9640 }, { "epoch": 19.979296066252587, "grad_norm": 0.09788880497217178, "learning_rate": 3.3453668231809286e-07, "loss": 0.0086, "step": 9650 }, { "epoch": 20.0, "grad_norm": 0.2991919219493866, "learning_rate": 3.157132477328628e-07, "loss": 0.0117, "step": 9660 }, { "epoch": 20.020703933747413, "grad_norm": 0.152107372879982, "learning_rate": 2.9743315254743833e-07, "loss": 0.0212, "step": 9670 }, { "epoch": 20.041407867494826, "grad_norm": 0.1049988642334938, "learning_rate": 2.796965966699927e-07, "loss": 0.012, "step": 9680 }, { "epoch": 20.062111801242235, "grad_norm": 0.28902608156204224, "learning_rate": 2.625037740646763e-07, "loss": 0.0103, "step": 9690 }, { "epoch": 20.082815734989648, "grad_norm": 0.2800842821598053, "learning_rate": 2.458548727494292e-07, "loss": 0.0067, "step": 9700 }, { "epoch": 20.10351966873706, "grad_norm": 0.18260431289672852, "learning_rate": 2.2975007479397738e-07, "loss": 0.0049, "step": 9710 }, { "epoch": 20.124223602484474, "grad_norm": 0.19015970826148987, "learning_rate": 2.1418955631781202e-07, "loss": 0.0117, "step": 9720 }, { "epoch": 20.144927536231883, "grad_norm": 0.1346769630908966, "learning_rate": 1.9917348748826335e-07, "loss": 0.0065, "step": 9730 }, { "epoch": 20.165631469979296, "grad_norm": 0.12038490176200867, "learning_rate": 1.847020325186577e-07, "loss": 0.0078, "step": 9740 }, { "epoch": 20.18633540372671, "grad_norm": 0.2004089653491974, "learning_rate": 1.7077534966650766e-07, "loss": 0.0096, "step": 9750 }, { "epoch": 20.20703933747412, "grad_norm": 0.19906413555145264, "learning_rate": 1.5739359123178587e-07, "loss": 0.0078, "step": 9760 }, { "epoch": 20.22774327122153, "grad_norm": 0.15226063132286072, "learning_rate": 1.4455690355525964e-07, "loss": 0.0049, "step": 9770 }, { "epoch": 20.248447204968944, "grad_norm": 0.12389522045850754, "learning_rate": 1.3226542701689215e-07, "loss": 0.0062, "step": 9780 }, { "epoch": 20.269151138716357, "grad_norm": 0.10260294377803802, "learning_rate": 1.2051929603428825e-07, "loss": 0.0131, "step": 9790 }, { "epoch": 20.28985507246377, "grad_norm": 0.059663962572813034, "learning_rate": 1.0931863906127327e-07, "loss": 0.0059, "step": 9800 }, { "epoch": 20.31055900621118, "grad_norm": 0.05520065873861313, "learning_rate": 9.866357858642205e-08, "loss": 0.0074, "step": 9810 }, { "epoch": 20.33126293995859, "grad_norm": 0.21085630357265472, "learning_rate": 8.855423113177664e-08, "loss": 0.0072, "step": 9820 }, { "epoch": 20.351966873706004, "grad_norm": 0.1336776465177536, "learning_rate": 7.899070725153613e-08, "loss": 0.0054, "step": 9830 }, { "epoch": 20.372670807453417, "grad_norm": 0.19630035758018494, "learning_rate": 6.997311153086883e-08, "loss": 0.0046, "step": 9840 }, { "epoch": 20.393374741200827, "grad_norm": 0.20599442720413208, "learning_rate": 6.150154258476315e-08, "loss": 0.0081, "step": 9850 }, { "epoch": 20.41407867494824, "grad_norm": 0.22885958850383759, "learning_rate": 5.3576093056922906e-08, "loss": 0.0076, "step": 9860 }, { "epoch": 20.434782608695652, "grad_norm": 0.07992696017026901, "learning_rate": 4.619684961881254e-08, "loss": 0.0071, "step": 9870 }, { "epoch": 20.455486542443065, "grad_norm": 0.2755813002586365, "learning_rate": 3.936389296864129e-08, "loss": 0.0064, "step": 9880 }, { "epoch": 20.476190476190474, "grad_norm": 0.1255924105644226, "learning_rate": 3.3077297830541584e-08, "loss": 0.0075, "step": 9890 }, { "epoch": 20.496894409937887, "grad_norm": 0.10394856333732605, "learning_rate": 2.7337132953697554e-08, "loss": 0.0072, "step": 9900 }, { "epoch": 20.5175983436853, "grad_norm": 0.11971770226955414, "learning_rate": 2.214346111164556e-08, "loss": 0.007, "step": 9910 }, { "epoch": 20.538302277432713, "grad_norm": 0.20464111864566803, "learning_rate": 1.749633910153592e-08, "loss": 0.0046, "step": 9920 }, { "epoch": 20.559006211180126, "grad_norm": 0.12219670414924622, "learning_rate": 1.3395817743561134e-08, "loss": 0.0082, "step": 9930 }, { "epoch": 20.579710144927535, "grad_norm": 0.2536000907421112, "learning_rate": 9.841941880361916e-09, "loss": 0.0152, "step": 9940 }, { "epoch": 20.600414078674948, "grad_norm": 0.16146942973136902, "learning_rate": 6.834750376549792e-09, "loss": 0.0055, "step": 9950 }, { "epoch": 20.62111801242236, "grad_norm": 0.1295255571603775, "learning_rate": 4.3742761183018784e-09, "loss": 0.0118, "step": 9960 }, { "epoch": 20.641821946169774, "grad_norm": 0.19239592552185059, "learning_rate": 2.4605460129556445e-09, "loss": 0.0065, "step": 9970 }, { "epoch": 20.662525879917183, "grad_norm": 0.21993553638458252, "learning_rate": 1.0935809887702154e-09, "loss": 0.0065, "step": 9980 }, { "epoch": 20.683229813664596, "grad_norm": 0.15056583285331726, "learning_rate": 2.7339599464326627e-10, "loss": 0.0062, "step": 9990 }, { "epoch": 20.70393374741201, "grad_norm": 0.24346260726451874, "learning_rate": 0.0, "loss": 0.0084, "step": 10000 }, { "epoch": 20.70393374741201, "step": 10000, "total_flos": 3.6267190177586125e+17, "train_loss": 0.02824210727363825, "train_runtime": 5841.6813, "train_samples_per_second": 27.389, "train_steps_per_second": 1.712 } ], "logging_steps": 10, "max_steps": 10000, "num_input_tokens_seen": 0, "num_train_epochs": 21, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.6267190177586125e+17, "train_batch_size": 16, "trial_name": null, "trial_params": null }