diff --git "a/multi_teacher_electronic_assembly_without_cot/trainer_state.json" "b/multi_teacher_electronic_assembly_without_cot/trainer_state.json" new file mode 100644--- /dev/null +++ "b/multi_teacher_electronic_assembly_without_cot/trainer_state.json" @@ -0,0 +1,6854 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 8.0, + "eval_steps": 500, + "global_step": 9736, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.008216926869350863, + "grad_norm": 19.974408737204264, + "learning_rate": 9.240246406570842e-07, + "loss": 8.1444, + "step": 10 + }, + { + "epoch": 0.016433853738701727, + "grad_norm": 17.651529960979847, + "learning_rate": 1.9507186858316226e-06, + "loss": 7.9807, + "step": 20 + }, + { + "epoch": 0.024650780608052588, + "grad_norm": 11.386632405383008, + "learning_rate": 2.9774127310061602e-06, + "loss": 7.2669, + "step": 30 + }, + { + "epoch": 0.03286770747740345, + "grad_norm": 7.046114654796961, + "learning_rate": 4.004106776180698e-06, + "loss": 6.2786, + "step": 40 + }, + { + "epoch": 0.041084634346754315, + "grad_norm": 6.151945803126112, + "learning_rate": 5.030800821355237e-06, + "loss": 5.4899, + "step": 50 + }, + { + "epoch": 0.049301561216105176, + "grad_norm": 8.193676876026643, + "learning_rate": 6.057494866529775e-06, + "loss": 4.6739, + "step": 60 + }, + { + "epoch": 0.05751848808545604, + "grad_norm": 6.679224346201021, + "learning_rate": 7.084188911704312e-06, + "loss": 3.4236, + "step": 70 + }, + { + "epoch": 0.0657354149548069, + "grad_norm": 2.6366005239013384, + "learning_rate": 8.11088295687885e-06, + "loss": 2.3599, + "step": 80 + }, + { + "epoch": 0.07395234182415776, + "grad_norm": 1.375969201844041, + "learning_rate": 9.137577002053388e-06, + "loss": 1.9964, + "step": 90 + }, + { + "epoch": 0.08216926869350863, + "grad_norm": 1.7699452853204731, + "learning_rate": 1.0164271047227925e-05, + "loss": 1.8573, + "step": 100 + }, + { + "epoch": 0.09038619556285948, + "grad_norm": 8.30917131347657, + "learning_rate": 1.1190965092402464e-05, + "loss": 1.7505, + "step": 110 + }, + { + "epoch": 0.09860312243221035, + "grad_norm": 6.75443517080473, + "learning_rate": 1.2217659137577003e-05, + "loss": 1.6748, + "step": 120 + }, + { + "epoch": 0.10682004930156122, + "grad_norm": 6.3489758184708975, + "learning_rate": 1.324435318275154e-05, + "loss": 1.6308, + "step": 130 + }, + { + "epoch": 0.11503697617091208, + "grad_norm": 7.53791236524009, + "learning_rate": 1.427104722792608e-05, + "loss": 1.5869, + "step": 140 + }, + { + "epoch": 0.12325390304026294, + "grad_norm": 8.054957976158356, + "learning_rate": 1.5297741273100617e-05, + "loss": 1.5436, + "step": 150 + }, + { + "epoch": 0.1314708299096138, + "grad_norm": 7.410251002629914, + "learning_rate": 1.6324435318275156e-05, + "loss": 1.5198, + "step": 160 + }, + { + "epoch": 0.13968775677896467, + "grad_norm": 7.667346984161794, + "learning_rate": 1.735112936344969e-05, + "loss": 1.4864, + "step": 170 + }, + { + "epoch": 0.14790468364831552, + "grad_norm": 3.427899808429963, + "learning_rate": 1.837782340862423e-05, + "loss": 1.4329, + "step": 180 + }, + { + "epoch": 0.1561216105176664, + "grad_norm": 7.389097005777736, + "learning_rate": 1.940451745379877e-05, + "loss": 1.3795, + "step": 190 + }, + { + "epoch": 0.16433853738701726, + "grad_norm": 3.8578784994944915, + "learning_rate": 2.0431211498973308e-05, + "loss": 1.3356, + "step": 200 + }, + { + "epoch": 0.1725554642563681, + "grad_norm": 6.847593619711181, + "learning_rate": 2.1457905544147843e-05, + "loss": 1.298, + "step": 210 + }, + { + "epoch": 0.18077239112571897, + "grad_norm": 3.9848894805058768, + "learning_rate": 2.2484599589322382e-05, + "loss": 1.2761, + "step": 220 + }, + { + "epoch": 0.18898931799506985, + "grad_norm": 3.42148486562202, + "learning_rate": 2.351129363449692e-05, + "loss": 1.2514, + "step": 230 + }, + { + "epoch": 0.1972062448644207, + "grad_norm": 4.183793781638511, + "learning_rate": 2.453798767967146e-05, + "loss": 1.212, + "step": 240 + }, + { + "epoch": 0.20542317173377156, + "grad_norm": 4.417796453724179, + "learning_rate": 2.5564681724845996e-05, + "loss": 1.1858, + "step": 250 + }, + { + "epoch": 0.21364009860312244, + "grad_norm": 3.1317362646162863, + "learning_rate": 2.6591375770020538e-05, + "loss": 1.1511, + "step": 260 + }, + { + "epoch": 0.2218570254724733, + "grad_norm": 3.6674629817218523, + "learning_rate": 2.7618069815195074e-05, + "loss": 1.1253, + "step": 270 + }, + { + "epoch": 0.23007395234182415, + "grad_norm": 2.4646163322898187, + "learning_rate": 2.8644763860369613e-05, + "loss": 1.0806, + "step": 280 + }, + { + "epoch": 0.23829087921117503, + "grad_norm": 3.79598763836086, + "learning_rate": 2.9671457905544148e-05, + "loss": 1.0437, + "step": 290 + }, + { + "epoch": 0.2465078060805259, + "grad_norm": 2.495574663825656, + "learning_rate": 3.069815195071869e-05, + "loss": 0.9992, + "step": 300 + }, + { + "epoch": 0.25472473294987674, + "grad_norm": 2.708197176131291, + "learning_rate": 3.172484599589323e-05, + "loss": 0.9809, + "step": 310 + }, + { + "epoch": 0.2629416598192276, + "grad_norm": 3.2301936096498336, + "learning_rate": 3.2751540041067765e-05, + "loss": 0.9391, + "step": 320 + }, + { + "epoch": 0.27115858668857845, + "grad_norm": 3.3281900964809314, + "learning_rate": 3.37782340862423e-05, + "loss": 0.9122, + "step": 330 + }, + { + "epoch": 0.27937551355792933, + "grad_norm": 3.7547304430550508, + "learning_rate": 3.480492813141684e-05, + "loss": 0.8853, + "step": 340 + }, + { + "epoch": 0.2875924404272802, + "grad_norm": 2.960454105842976, + "learning_rate": 3.583162217659138e-05, + "loss": 0.8593, + "step": 350 + }, + { + "epoch": 0.29580936729663104, + "grad_norm": 3.4887397124991772, + "learning_rate": 3.6858316221765914e-05, + "loss": 0.8366, + "step": 360 + }, + { + "epoch": 0.3040262941659819, + "grad_norm": 3.759040242435043, + "learning_rate": 3.788501026694045e-05, + "loss": 0.808, + "step": 370 + }, + { + "epoch": 0.3122432210353328, + "grad_norm": 3.5316624341034393, + "learning_rate": 3.891170431211499e-05, + "loss": 0.7784, + "step": 380 + }, + { + "epoch": 0.32046014790468363, + "grad_norm": 2.536712563451594, + "learning_rate": 3.993839835728953e-05, + "loss": 0.7638, + "step": 390 + }, + { + "epoch": 0.3286770747740345, + "grad_norm": 3.1494878741303958, + "learning_rate": 4.096509240246407e-05, + "loss": 0.7436, + "step": 400 + }, + { + "epoch": 0.3368940016433854, + "grad_norm": 3.429889342571355, + "learning_rate": 4.1991786447638605e-05, + "loss": 0.7202, + "step": 410 + }, + { + "epoch": 0.3451109285127362, + "grad_norm": 3.6597537437672423, + "learning_rate": 4.301848049281315e-05, + "loss": 0.7059, + "step": 420 + }, + { + "epoch": 0.3533278553820871, + "grad_norm": 3.3239802253491155, + "learning_rate": 4.404517453798768e-05, + "loss": 0.6915, + "step": 430 + }, + { + "epoch": 0.36154478225143794, + "grad_norm": 3.087880747049263, + "learning_rate": 4.507186858316222e-05, + "loss": 0.668, + "step": 440 + }, + { + "epoch": 0.3697617091207888, + "grad_norm": 3.083327651120632, + "learning_rate": 4.6098562628336754e-05, + "loss": 0.6437, + "step": 450 + }, + { + "epoch": 0.3779786359901397, + "grad_norm": 3.062369129556137, + "learning_rate": 4.7125256673511296e-05, + "loss": 0.6347, + "step": 460 + }, + { + "epoch": 0.3861955628594905, + "grad_norm": 3.9335033286871117, + "learning_rate": 4.815195071868583e-05, + "loss": 0.6111, + "step": 470 + }, + { + "epoch": 0.3944124897288414, + "grad_norm": 2.845179486951232, + "learning_rate": 4.9178644763860374e-05, + "loss": 0.5961, + "step": 480 + }, + { + "epoch": 0.4026294165981923, + "grad_norm": 2.979224322509434, + "learning_rate": 5.020533880903491e-05, + "loss": 0.5813, + "step": 490 + }, + { + "epoch": 0.4108463434675431, + "grad_norm": 3.6005231646388736, + "learning_rate": 5.1232032854209445e-05, + "loss": 0.559, + "step": 500 + }, + { + "epoch": 0.419063270336894, + "grad_norm": 3.781201309516029, + "learning_rate": 5.225872689938398e-05, + "loss": 0.5505, + "step": 510 + }, + { + "epoch": 0.4272801972062449, + "grad_norm": 2.820314146132875, + "learning_rate": 5.328542094455853e-05, + "loss": 0.5206, + "step": 520 + }, + { + "epoch": 0.4354971240755957, + "grad_norm": 3.3247760436321783, + "learning_rate": 5.4312114989733065e-05, + "loss": 0.5191, + "step": 530 + }, + { + "epoch": 0.4437140509449466, + "grad_norm": 3.185990273496067, + "learning_rate": 5.53388090349076e-05, + "loss": 0.4937, + "step": 540 + }, + { + "epoch": 0.4519309778142975, + "grad_norm": 3.0364259829392717, + "learning_rate": 5.6365503080082136e-05, + "loss": 0.4812, + "step": 550 + }, + { + "epoch": 0.4601479046836483, + "grad_norm": 2.981635296387684, + "learning_rate": 5.739219712525667e-05, + "loss": 0.4634, + "step": 560 + }, + { + "epoch": 0.4683648315529992, + "grad_norm": 3.9637927429253934, + "learning_rate": 5.8418891170431214e-05, + "loss": 0.4496, + "step": 570 + }, + { + "epoch": 0.47658175842235007, + "grad_norm": 3.422720069544419, + "learning_rate": 5.944558521560575e-05, + "loss": 0.4345, + "step": 580 + }, + { + "epoch": 0.4847986852917009, + "grad_norm": 2.618638333473933, + "learning_rate": 6.04722792607803e-05, + "loss": 0.4169, + "step": 590 + }, + { + "epoch": 0.4930156121610518, + "grad_norm": 3.4310336539551147, + "learning_rate": 6.149897330595483e-05, + "loss": 0.4069, + "step": 600 + }, + { + "epoch": 0.5012325390304027, + "grad_norm": 2.695323794945264, + "learning_rate": 6.252566735112937e-05, + "loss": 0.3941, + "step": 610 + }, + { + "epoch": 0.5094494658997535, + "grad_norm": 3.2005546191921357, + "learning_rate": 6.35523613963039e-05, + "loss": 0.3844, + "step": 620 + }, + { + "epoch": 0.5176663927691043, + "grad_norm": 2.698949275451768, + "learning_rate": 6.457905544147844e-05, + "loss": 0.3685, + "step": 630 + }, + { + "epoch": 0.5258833196384552, + "grad_norm": 3.0068086367783846, + "learning_rate": 6.560574948665298e-05, + "loss": 0.3654, + "step": 640 + }, + { + "epoch": 0.5341002465078061, + "grad_norm": 2.699112912753321, + "learning_rate": 6.663244353182751e-05, + "loss": 0.3543, + "step": 650 + }, + { + "epoch": 0.5423171733771569, + "grad_norm": 2.2278954979820833, + "learning_rate": 6.765913757700205e-05, + "loss": 0.3413, + "step": 660 + }, + { + "epoch": 0.5505341002465078, + "grad_norm": 2.245661320124028, + "learning_rate": 6.86858316221766e-05, + "loss": 0.3332, + "step": 670 + }, + { + "epoch": 0.5587510271158587, + "grad_norm": 2.0476933338187524, + "learning_rate": 6.971252566735113e-05, + "loss": 0.3261, + "step": 680 + }, + { + "epoch": 0.5669679539852095, + "grad_norm": 2.389837965434361, + "learning_rate": 7.073921971252567e-05, + "loss": 0.3164, + "step": 690 + }, + { + "epoch": 0.5751848808545604, + "grad_norm": 2.1365833016872644, + "learning_rate": 7.17659137577002e-05, + "loss": 0.3038, + "step": 700 + }, + { + "epoch": 0.5834018077239113, + "grad_norm": 2.7541557475997234, + "learning_rate": 7.279260780287475e-05, + "loss": 0.2988, + "step": 710 + }, + { + "epoch": 0.5916187345932621, + "grad_norm": 2.0589479996604774, + "learning_rate": 7.381930184804929e-05, + "loss": 0.2872, + "step": 720 + }, + { + "epoch": 0.599835661462613, + "grad_norm": 2.991548851558623, + "learning_rate": 7.484599589322382e-05, + "loss": 0.2872, + "step": 730 + }, + { + "epoch": 0.6080525883319639, + "grad_norm": 1.6552071864163582, + "learning_rate": 7.587268993839836e-05, + "loss": 0.2762, + "step": 740 + }, + { + "epoch": 0.6162695152013147, + "grad_norm": 1.789564293031145, + "learning_rate": 7.689938398357291e-05, + "loss": 0.2661, + "step": 750 + }, + { + "epoch": 0.6244864420706656, + "grad_norm": 1.9221080215457937, + "learning_rate": 7.792607802874744e-05, + "loss": 0.2633, + "step": 760 + }, + { + "epoch": 0.6327033689400164, + "grad_norm": 1.9268969992622877, + "learning_rate": 7.895277207392198e-05, + "loss": 0.2562, + "step": 770 + }, + { + "epoch": 0.6409202958093673, + "grad_norm": 1.854381700776246, + "learning_rate": 7.997946611909651e-05, + "loss": 0.2486, + "step": 780 + }, + { + "epoch": 0.6491372226787182, + "grad_norm": 2.3180166811886704, + "learning_rate": 8.100616016427105e-05, + "loss": 0.2409, + "step": 790 + }, + { + "epoch": 0.657354149548069, + "grad_norm": 1.7752686819968808, + "learning_rate": 8.203285420944559e-05, + "loss": 0.2323, + "step": 800 + }, + { + "epoch": 0.6655710764174199, + "grad_norm": 1.4459478240777937, + "learning_rate": 8.305954825462012e-05, + "loss": 0.2258, + "step": 810 + }, + { + "epoch": 0.6737880032867708, + "grad_norm": 1.5953025726188175, + "learning_rate": 8.408624229979466e-05, + "loss": 0.2255, + "step": 820 + }, + { + "epoch": 0.6820049301561216, + "grad_norm": 1.6045621144575808, + "learning_rate": 8.51129363449692e-05, + "loss": 0.2169, + "step": 830 + }, + { + "epoch": 0.6902218570254725, + "grad_norm": 1.7224271707297807, + "learning_rate": 8.613963039014374e-05, + "loss": 0.2136, + "step": 840 + }, + { + "epoch": 0.6984387838948234, + "grad_norm": 1.310776192683364, + "learning_rate": 8.716632443531828e-05, + "loss": 0.2056, + "step": 850 + }, + { + "epoch": 0.7066557107641742, + "grad_norm": 1.4088847490761656, + "learning_rate": 8.819301848049281e-05, + "loss": 0.1985, + "step": 860 + }, + { + "epoch": 0.714872637633525, + "grad_norm": 1.60851910808792, + "learning_rate": 8.921971252566735e-05, + "loss": 0.1943, + "step": 870 + }, + { + "epoch": 0.7230895645028759, + "grad_norm": 1.5206100365618338, + "learning_rate": 9.02464065708419e-05, + "loss": 0.1922, + "step": 880 + }, + { + "epoch": 0.7313064913722268, + "grad_norm": 1.3408978875102782, + "learning_rate": 9.127310061601643e-05, + "loss": 0.1851, + "step": 890 + }, + { + "epoch": 0.7395234182415776, + "grad_norm": 1.3916859583046408, + "learning_rate": 9.229979466119097e-05, + "loss": 0.1805, + "step": 900 + }, + { + "epoch": 0.7477403451109285, + "grad_norm": 1.0284073963140599, + "learning_rate": 9.332648870636552e-05, + "loss": 0.1765, + "step": 910 + }, + { + "epoch": 0.7559572719802794, + "grad_norm": 1.312106418490207, + "learning_rate": 9.435318275154005e-05, + "loss": 0.1729, + "step": 920 + }, + { + "epoch": 0.7641741988496302, + "grad_norm": 1.2873051569354634, + "learning_rate": 9.537987679671459e-05, + "loss": 0.1697, + "step": 930 + }, + { + "epoch": 0.772391125718981, + "grad_norm": 1.3089626745248004, + "learning_rate": 9.640657084188912e-05, + "loss": 0.1644, + "step": 940 + }, + { + "epoch": 0.780608052588332, + "grad_norm": 1.1995762867540671, + "learning_rate": 9.743326488706366e-05, + "loss": 0.1605, + "step": 950 + }, + { + "epoch": 0.7888249794576828, + "grad_norm": 1.2716058027661918, + "learning_rate": 9.84599589322382e-05, + "loss": 0.1568, + "step": 960 + }, + { + "epoch": 0.7970419063270336, + "grad_norm": 1.0777715246378874, + "learning_rate": 9.948665297741273e-05, + "loss": 0.1524, + "step": 970 + }, + { + "epoch": 0.8052588331963846, + "grad_norm": 1.2259646342544421, + "learning_rate": 9.999991965235563e-05, + "loss": 0.1501, + "step": 980 + }, + { + "epoch": 0.8134757600657354, + "grad_norm": 1.2693664502136377, + "learning_rate": 9.999927687274992e-05, + "loss": 0.1462, + "step": 990 + }, + { + "epoch": 0.8216926869350862, + "grad_norm": 0.9595367386056208, + "learning_rate": 9.999799132180184e-05, + "loss": 0.1427, + "step": 1000 + }, + { + "epoch": 0.8299096138044372, + "grad_norm": 0.9879321961046859, + "learning_rate": 9.999606301603793e-05, + "loss": 0.1389, + "step": 1010 + }, + { + "epoch": 0.838126540673788, + "grad_norm": 1.0545188211346712, + "learning_rate": 9.999349198024775e-05, + "loss": 0.1374, + "step": 1020 + }, + { + "epoch": 0.8463434675431388, + "grad_norm": 0.9381854736129261, + "learning_rate": 9.999027824748352e-05, + "loss": 0.134, + "step": 1030 + }, + { + "epoch": 0.8545603944124898, + "grad_norm": 1.018456405955215, + "learning_rate": 9.998642185905977e-05, + "loss": 0.1294, + "step": 1040 + }, + { + "epoch": 0.8627773212818406, + "grad_norm": 0.817274757757614, + "learning_rate": 9.998192286455273e-05, + "loss": 0.1277, + "step": 1050 + }, + { + "epoch": 0.8709942481511914, + "grad_norm": 0.877627283927687, + "learning_rate": 9.997678132179969e-05, + "loss": 0.1259, + "step": 1060 + }, + { + "epoch": 0.8792111750205424, + "grad_norm": 0.9719564910761921, + "learning_rate": 9.997099729689839e-05, + "loss": 0.1211, + "step": 1070 + }, + { + "epoch": 0.8874281018898932, + "grad_norm": 1.0155987976096954, + "learning_rate": 9.996457086420598e-05, + "loss": 0.1172, + "step": 1080 + }, + { + "epoch": 0.895645028759244, + "grad_norm": 0.905332195577956, + "learning_rate": 9.995750210633822e-05, + "loss": 0.1151, + "step": 1090 + }, + { + "epoch": 0.903861955628595, + "grad_norm": 0.875108702397198, + "learning_rate": 9.994979111416829e-05, + "loss": 0.1114, + "step": 1100 + }, + { + "epoch": 0.9120788824979458, + "grad_norm": 0.8315209452920128, + "learning_rate": 9.994143798682573e-05, + "loss": 0.1108, + "step": 1110 + }, + { + "epoch": 0.9202958093672966, + "grad_norm": 0.8419546904865063, + "learning_rate": 9.993244283169513e-05, + "loss": 0.1091, + "step": 1120 + }, + { + "epoch": 0.9285127362366475, + "grad_norm": 0.8298844114966769, + "learning_rate": 9.99228057644147e-05, + "loss": 0.1037, + "step": 1130 + }, + { + "epoch": 0.9367296631059984, + "grad_norm": 0.8012814539297765, + "learning_rate": 9.991252690887485e-05, + "loss": 0.1023, + "step": 1140 + }, + { + "epoch": 0.9449465899753492, + "grad_norm": 0.8677793896939191, + "learning_rate": 9.990160639721657e-05, + "loss": 0.1008, + "step": 1150 + }, + { + "epoch": 0.9531635168447001, + "grad_norm": 0.7447389790617776, + "learning_rate": 9.989004436982974e-05, + "loss": 0.0992, + "step": 1160 + }, + { + "epoch": 0.961380443714051, + "grad_norm": 0.6685887225271923, + "learning_rate": 9.987784097535127e-05, + "loss": 0.0961, + "step": 1170 + }, + { + "epoch": 0.9695973705834018, + "grad_norm": 0.7320800211798315, + "learning_rate": 9.986499637066331e-05, + "loss": 0.0915, + "step": 1180 + }, + { + "epoch": 0.9778142974527527, + "grad_norm": 0.7347748013970976, + "learning_rate": 9.985151072089112e-05, + "loss": 0.0922, + "step": 1190 + }, + { + "epoch": 0.9860312243221035, + "grad_norm": 0.7658898129650706, + "learning_rate": 9.983738419940096e-05, + "loss": 0.0887, + "step": 1200 + }, + { + "epoch": 0.9942481511914544, + "grad_norm": 0.6371211246975768, + "learning_rate": 9.982261698779795e-05, + "loss": 0.0879, + "step": 1210 + }, + { + "epoch": 1.0024650780608053, + "grad_norm": 0.6728408512142455, + "learning_rate": 9.980720927592366e-05, + "loss": 0.0862, + "step": 1220 + }, + { + "epoch": 1.0106820049301561, + "grad_norm": 0.7086330539017783, + "learning_rate": 9.979116126185364e-05, + "loss": 0.0784, + "step": 1230 + }, + { + "epoch": 1.018898931799507, + "grad_norm": 0.6735114050298749, + "learning_rate": 9.977447315189493e-05, + "loss": 0.0754, + "step": 1240 + }, + { + "epoch": 1.0271158586688578, + "grad_norm": 0.610617286707555, + "learning_rate": 9.975714516058344e-05, + "loss": 0.0744, + "step": 1250 + }, + { + "epoch": 1.0353327855382086, + "grad_norm": 0.6567215002986171, + "learning_rate": 9.973917751068111e-05, + "loss": 0.0724, + "step": 1260 + }, + { + "epoch": 1.0435497124075597, + "grad_norm": 0.6711754808875117, + "learning_rate": 9.972057043317308e-05, + "loss": 0.0721, + "step": 1270 + }, + { + "epoch": 1.0517666392769105, + "grad_norm": 0.6518939760978527, + "learning_rate": 9.970132416726474e-05, + "loss": 0.0703, + "step": 1280 + }, + { + "epoch": 1.0599835661462613, + "grad_norm": 0.6612508351869697, + "learning_rate": 9.968143896037864e-05, + "loss": 0.0702, + "step": 1290 + }, + { + "epoch": 1.0682004930156122, + "grad_norm": 0.6539738578193367, + "learning_rate": 9.966091506815128e-05, + "loss": 0.068, + "step": 1300 + }, + { + "epoch": 1.076417419884963, + "grad_norm": 0.5743874788903558, + "learning_rate": 9.963975275442989e-05, + "loss": 0.0706, + "step": 1310 + }, + { + "epoch": 1.0846343467543138, + "grad_norm": 0.6303386908851022, + "learning_rate": 9.961795229126896e-05, + "loss": 0.0666, + "step": 1320 + }, + { + "epoch": 1.0928512736236646, + "grad_norm": 0.589250923788632, + "learning_rate": 9.959551395892681e-05, + "loss": 0.0662, + "step": 1330 + }, + { + "epoch": 1.1010682004930157, + "grad_norm": 0.576469666142936, + "learning_rate": 9.957243804586197e-05, + "loss": 0.0648, + "step": 1340 + }, + { + "epoch": 1.1092851273623665, + "grad_norm": 0.6424675739164545, + "learning_rate": 9.954872484872944e-05, + "loss": 0.0635, + "step": 1350 + }, + { + "epoch": 1.1175020542317173, + "grad_norm": 0.6304193275617966, + "learning_rate": 9.952437467237687e-05, + "loss": 0.0632, + "step": 1360 + }, + { + "epoch": 1.1257189811010682, + "grad_norm": 0.60085541104229, + "learning_rate": 9.949938782984072e-05, + "loss": 0.0627, + "step": 1370 + }, + { + "epoch": 1.133935907970419, + "grad_norm": 0.6387032797411912, + "learning_rate": 9.947376464234215e-05, + "loss": 0.0619, + "step": 1380 + }, + { + "epoch": 1.14215283483977, + "grad_norm": 0.590359230196097, + "learning_rate": 9.944750543928297e-05, + "loss": 0.0601, + "step": 1390 + }, + { + "epoch": 1.1503697617091209, + "grad_norm": 0.6016993560357249, + "learning_rate": 9.942061055824127e-05, + "loss": 0.0585, + "step": 1400 + }, + { + "epoch": 1.1585866885784717, + "grad_norm": 0.6227776527484776, + "learning_rate": 9.939308034496728e-05, + "loss": 0.0585, + "step": 1410 + }, + { + "epoch": 1.1668036154478225, + "grad_norm": 0.5840424383749039, + "learning_rate": 9.936491515337874e-05, + "loss": 0.059, + "step": 1420 + }, + { + "epoch": 1.1750205423171733, + "grad_norm": 0.602439104016672, + "learning_rate": 9.933611534555644e-05, + "loss": 0.056, + "step": 1430 + }, + { + "epoch": 1.1832374691865242, + "grad_norm": 0.5131524987589036, + "learning_rate": 9.930668129173956e-05, + "loss": 0.0565, + "step": 1440 + }, + { + "epoch": 1.191454396055875, + "grad_norm": 0.5607165065501415, + "learning_rate": 9.927661337032093e-05, + "loss": 0.0554, + "step": 1450 + }, + { + "epoch": 1.199671322925226, + "grad_norm": 0.5938542954233547, + "learning_rate": 9.924591196784205e-05, + "loss": 0.0547, + "step": 1460 + }, + { + "epoch": 1.2078882497945769, + "grad_norm": 0.5689585616491131, + "learning_rate": 9.92145774789883e-05, + "loss": 0.0544, + "step": 1470 + }, + { + "epoch": 1.2161051766639277, + "grad_norm": 0.5849055658080314, + "learning_rate": 9.918261030658371e-05, + "loss": 0.0519, + "step": 1480 + }, + { + "epoch": 1.2243221035332785, + "grad_norm": 0.5248794740274818, + "learning_rate": 9.915001086158588e-05, + "loss": 0.0514, + "step": 1490 + }, + { + "epoch": 1.2325390304026294, + "grad_norm": 0.5972426102710096, + "learning_rate": 9.911677956308068e-05, + "loss": 0.0525, + "step": 1500 + }, + { + "epoch": 1.2407559572719804, + "grad_norm": 0.5374698363190101, + "learning_rate": 9.908291683827678e-05, + "loss": 0.0505, + "step": 1510 + }, + { + "epoch": 1.2489728841413312, + "grad_norm": 0.5061181252518886, + "learning_rate": 9.904842312250024e-05, + "loss": 0.0503, + "step": 1520 + }, + { + "epoch": 1.257189811010682, + "grad_norm": 0.5022441363216565, + "learning_rate": 9.901329885918898e-05, + "loss": 0.0492, + "step": 1530 + }, + { + "epoch": 1.2654067378800329, + "grad_norm": 0.5451769166028749, + "learning_rate": 9.897754449988687e-05, + "loss": 0.0508, + "step": 1540 + }, + { + "epoch": 1.2736236647493837, + "grad_norm": 0.584898627104412, + "learning_rate": 9.894116050423812e-05, + "loss": 0.0491, + "step": 1550 + }, + { + "epoch": 1.2818405916187345, + "grad_norm": 0.5371972267407287, + "learning_rate": 9.890414733998131e-05, + "loss": 0.0473, + "step": 1560 + }, + { + "epoch": 1.2900575184880854, + "grad_norm": 0.6017034266346077, + "learning_rate": 9.886650548294333e-05, + "loss": 0.0484, + "step": 1570 + }, + { + "epoch": 1.2982744453574364, + "grad_norm": 0.5010695241137435, + "learning_rate": 9.882823541703332e-05, + "loss": 0.0464, + "step": 1580 + }, + { + "epoch": 1.3064913722267872, + "grad_norm": 0.5676853943666109, + "learning_rate": 9.878933763423644e-05, + "loss": 0.0483, + "step": 1590 + }, + { + "epoch": 1.314708299096138, + "grad_norm": 0.5418403716225952, + "learning_rate": 9.874981263460749e-05, + "loss": 0.0461, + "step": 1600 + }, + { + "epoch": 1.322925225965489, + "grad_norm": 0.5584739393482486, + "learning_rate": 9.87096609262646e-05, + "loss": 0.0442, + "step": 1610 + }, + { + "epoch": 1.3311421528348397, + "grad_norm": 0.4890574045825851, + "learning_rate": 9.866888302538255e-05, + "loss": 0.0456, + "step": 1620 + }, + { + "epoch": 1.3393590797041908, + "grad_norm": 0.5311077678345358, + "learning_rate": 9.862747945618629e-05, + "loss": 0.046, + "step": 1630 + }, + { + "epoch": 1.3475760065735414, + "grad_norm": 0.545353294627644, + "learning_rate": 9.858545075094402e-05, + "loss": 0.0443, + "step": 1640 + }, + { + "epoch": 1.3557929334428924, + "grad_norm": 0.5284222343562959, + "learning_rate": 9.854279744996054e-05, + "loss": 0.0435, + "step": 1650 + }, + { + "epoch": 1.3640098603122432, + "grad_norm": 0.46725321717942186, + "learning_rate": 9.849952010157014e-05, + "loss": 0.0438, + "step": 1660 + }, + { + "epoch": 1.372226787181594, + "grad_norm": 0.47587942407600886, + "learning_rate": 9.845561926212967e-05, + "loss": 0.0431, + "step": 1670 + }, + { + "epoch": 1.380443714050945, + "grad_norm": 0.42647419036174755, + "learning_rate": 9.841109549601133e-05, + "loss": 0.0423, + "step": 1680 + }, + { + "epoch": 1.3886606409202957, + "grad_norm": 0.47348266699383407, + "learning_rate": 9.83659493755954e-05, + "loss": 0.0417, + "step": 1690 + }, + { + "epoch": 1.3968775677896468, + "grad_norm": 0.5734279662260182, + "learning_rate": 9.832018148126291e-05, + "loss": 0.0418, + "step": 1700 + }, + { + "epoch": 1.4050944946589976, + "grad_norm": 0.5079390469387594, + "learning_rate": 9.827379240138824e-05, + "loss": 0.0407, + "step": 1710 + }, + { + "epoch": 1.4133114215283484, + "grad_norm": 0.48071880121247074, + "learning_rate": 9.822678273233139e-05, + "loss": 0.0407, + "step": 1720 + }, + { + "epoch": 1.4215283483976993, + "grad_norm": 0.4533213147750345, + "learning_rate": 9.817915307843046e-05, + "loss": 0.0392, + "step": 1730 + }, + { + "epoch": 1.42974527526705, + "grad_norm": 0.45354884644459964, + "learning_rate": 9.813090405199388e-05, + "loss": 0.0386, + "step": 1740 + }, + { + "epoch": 1.437962202136401, + "grad_norm": 0.4060068627853837, + "learning_rate": 9.80820362732924e-05, + "loss": 0.0382, + "step": 1750 + }, + { + "epoch": 1.4461791290057517, + "grad_norm": 0.4507780656354325, + "learning_rate": 9.803255037055128e-05, + "loss": 0.0382, + "step": 1760 + }, + { + "epoch": 1.4543960558751028, + "grad_norm": 0.5120472100440739, + "learning_rate": 9.798244697994214e-05, + "loss": 0.0376, + "step": 1770 + }, + { + "epoch": 1.4626129827444536, + "grad_norm": 0.4669544207241051, + "learning_rate": 9.793172674557475e-05, + "loss": 0.0368, + "step": 1780 + }, + { + "epoch": 1.4708299096138044, + "grad_norm": 0.41686952393568316, + "learning_rate": 9.78803903194888e-05, + "loss": 0.036, + "step": 1790 + }, + { + "epoch": 1.4790468364831553, + "grad_norm": 0.47436080111491863, + "learning_rate": 9.782843836164553e-05, + "loss": 0.0354, + "step": 1800 + }, + { + "epoch": 1.487263763352506, + "grad_norm": 0.47129658653308115, + "learning_rate": 9.777587153991916e-05, + "loss": 0.0352, + "step": 1810 + }, + { + "epoch": 1.4954806902218571, + "grad_norm": 0.4459800537917433, + "learning_rate": 9.77226905300884e-05, + "loss": 0.0361, + "step": 1820 + }, + { + "epoch": 1.5036976170912077, + "grad_norm": 0.43894409691624636, + "learning_rate": 9.766889601582774e-05, + "loss": 0.0341, + "step": 1830 + }, + { + "epoch": 1.5119145439605588, + "grad_norm": 0.4681621495969991, + "learning_rate": 9.761448868869863e-05, + "loss": 0.0365, + "step": 1840 + }, + { + "epoch": 1.5201314708299096, + "grad_norm": 0.44125169083676535, + "learning_rate": 9.755946924814055e-05, + "loss": 0.0349, + "step": 1850 + }, + { + "epoch": 1.5283483976992605, + "grad_norm": 0.42454765007479894, + "learning_rate": 9.750383840146218e-05, + "loss": 0.034, + "step": 1860 + }, + { + "epoch": 1.5365653245686115, + "grad_norm": 0.4312790936546121, + "learning_rate": 9.744759686383211e-05, + "loss": 0.0342, + "step": 1870 + }, + { + "epoch": 1.544782251437962, + "grad_norm": 0.46580009646158177, + "learning_rate": 9.739074535826978e-05, + "loss": 0.0335, + "step": 1880 + }, + { + "epoch": 1.5529991783073132, + "grad_norm": 0.4275717475424306, + "learning_rate": 9.733328461563612e-05, + "loss": 0.0344, + "step": 1890 + }, + { + "epoch": 1.561216105176664, + "grad_norm": 0.4756865714632423, + "learning_rate": 9.72752153746242e-05, + "loss": 0.034, + "step": 1900 + }, + { + "epoch": 1.5694330320460148, + "grad_norm": 0.41615409042671186, + "learning_rate": 9.721653838174972e-05, + "loss": 0.0322, + "step": 1910 + }, + { + "epoch": 1.5776499589153656, + "grad_norm": 0.4261218829495612, + "learning_rate": 9.715725439134132e-05, + "loss": 0.0323, + "step": 1920 + }, + { + "epoch": 1.5858668857847165, + "grad_norm": 0.4103285392171169, + "learning_rate": 9.709736416553106e-05, + "loss": 0.0321, + "step": 1930 + }, + { + "epoch": 1.5940838126540675, + "grad_norm": 0.42730940931321443, + "learning_rate": 9.703686847424449e-05, + "loss": 0.0323, + "step": 1940 + }, + { + "epoch": 1.6023007395234181, + "grad_norm": 0.37284494211308117, + "learning_rate": 9.697576809519079e-05, + "loss": 0.0306, + "step": 1950 + }, + { + "epoch": 1.6105176663927692, + "grad_norm": 0.4347235376642582, + "learning_rate": 9.691406381385278e-05, + "loss": 0.0309, + "step": 1960 + }, + { + "epoch": 1.61873459326212, + "grad_norm": 0.41859821688218857, + "learning_rate": 9.685175642347679e-05, + "loss": 0.0306, + "step": 1970 + }, + { + "epoch": 1.6269515201314708, + "grad_norm": 0.458193281780139, + "learning_rate": 9.678884672506252e-05, + "loss": 0.0304, + "step": 1980 + }, + { + "epoch": 1.6351684470008216, + "grad_norm": 0.4467297131068921, + "learning_rate": 9.672533552735268e-05, + "loss": 0.0307, + "step": 1990 + }, + { + "epoch": 1.6433853738701725, + "grad_norm": 0.4246649874690625, + "learning_rate": 9.666122364682263e-05, + "loss": 0.029, + "step": 2000 + }, + { + "epoch": 1.6516023007395235, + "grad_norm": 0.44552544340004335, + "learning_rate": 9.659651190766992e-05, + "loss": 0.0295, + "step": 2010 + }, + { + "epoch": 1.6598192276088741, + "grad_norm": 0.4164644781602266, + "learning_rate": 9.653120114180356e-05, + "loss": 0.0291, + "step": 2020 + }, + { + "epoch": 1.6680361544782252, + "grad_norm": 0.4579243177302215, + "learning_rate": 9.646529218883349e-05, + "loss": 0.0302, + "step": 2030 + }, + { + "epoch": 1.676253081347576, + "grad_norm": 0.48543566357037554, + "learning_rate": 9.639878589605969e-05, + "loss": 0.0291, + "step": 2040 + }, + { + "epoch": 1.6844700082169268, + "grad_norm": 0.46469944987261236, + "learning_rate": 9.63316831184613e-05, + "loss": 0.0292, + "step": 2050 + }, + { + "epoch": 1.6926869350862779, + "grad_norm": 0.45357303638368995, + "learning_rate": 9.626398471868564e-05, + "loss": 0.0287, + "step": 2060 + }, + { + "epoch": 1.7009038619556285, + "grad_norm": 0.42210490946569273, + "learning_rate": 9.619569156703714e-05, + "loss": 0.0282, + "step": 2070 + }, + { + "epoch": 1.7091207888249795, + "grad_norm": 0.41082451259817443, + "learning_rate": 9.612680454146609e-05, + "loss": 0.0276, + "step": 2080 + }, + { + "epoch": 1.7173377156943304, + "grad_norm": 0.41363141928585667, + "learning_rate": 9.605732452755742e-05, + "loss": 0.0273, + "step": 2090 + }, + { + "epoch": 1.7255546425636812, + "grad_norm": 0.4206911941733154, + "learning_rate": 9.598725241851932e-05, + "loss": 0.0287, + "step": 2100 + }, + { + "epoch": 1.733771569433032, + "grad_norm": 0.3951369498197376, + "learning_rate": 9.591658911517162e-05, + "loss": 0.0275, + "step": 2110 + }, + { + "epoch": 1.7419884963023828, + "grad_norm": 0.4071839799145019, + "learning_rate": 9.584533552593447e-05, + "loss": 0.0263, + "step": 2120 + }, + { + "epoch": 1.7502054231717339, + "grad_norm": 0.35587647953674884, + "learning_rate": 9.577349256681636e-05, + "loss": 0.0274, + "step": 2130 + }, + { + "epoch": 1.7584223500410845, + "grad_norm": 0.43711475000779765, + "learning_rate": 9.570106116140256e-05, + "loss": 0.0268, + "step": 2140 + }, + { + "epoch": 1.7666392769104355, + "grad_norm": 0.3991641744179894, + "learning_rate": 9.56280422408432e-05, + "loss": 0.0268, + "step": 2150 + }, + { + "epoch": 1.7748562037797864, + "grad_norm": 0.4076460125133768, + "learning_rate": 9.555443674384122e-05, + "loss": 0.0255, + "step": 2160 + }, + { + "epoch": 1.7830731306491372, + "grad_norm": 0.4348100956752934, + "learning_rate": 9.548024561664039e-05, + "loss": 0.0277, + "step": 2170 + }, + { + "epoch": 1.7912900575184882, + "grad_norm": 0.4034844759662755, + "learning_rate": 9.540546981301313e-05, + "loss": 0.0254, + "step": 2180 + }, + { + "epoch": 1.7995069843878388, + "grad_norm": 0.4132706716223311, + "learning_rate": 9.53301102942482e-05, + "loss": 0.0253, + "step": 2190 + }, + { + "epoch": 1.80772391125719, + "grad_norm": 0.37701202488605456, + "learning_rate": 9.525416802913841e-05, + "loss": 0.0246, + "step": 2200 + }, + { + "epoch": 1.8159408381265407, + "grad_norm": 0.39051860707046865, + "learning_rate": 9.51776439939681e-05, + "loss": 0.0257, + "step": 2210 + }, + { + "epoch": 1.8241577649958916, + "grad_norm": 0.4106016394020398, + "learning_rate": 9.510053917250061e-05, + "loss": 0.0257, + "step": 2220 + }, + { + "epoch": 1.8323746918652424, + "grad_norm": 0.3866184281474688, + "learning_rate": 9.502285455596572e-05, + "loss": 0.0245, + "step": 2230 + }, + { + "epoch": 1.8405916187345932, + "grad_norm": 0.4021017932429677, + "learning_rate": 9.494459114304672e-05, + "loss": 0.0244, + "step": 2240 + }, + { + "epoch": 1.8488085456039443, + "grad_norm": 0.3920008094917924, + "learning_rate": 9.48657499398678e-05, + "loss": 0.0248, + "step": 2250 + }, + { + "epoch": 1.8570254724732949, + "grad_norm": 0.4004129162133985, + "learning_rate": 9.47863319599809e-05, + "loss": 0.0249, + "step": 2260 + }, + { + "epoch": 1.865242399342646, + "grad_norm": 0.3782452193154733, + "learning_rate": 9.470633822435282e-05, + "loss": 0.0246, + "step": 2270 + }, + { + "epoch": 1.8734593262119967, + "grad_norm": 0.36635346240750793, + "learning_rate": 9.462576976135207e-05, + "loss": 0.0228, + "step": 2280 + }, + { + "epoch": 1.8816762530813476, + "grad_norm": 0.38839169002774604, + "learning_rate": 9.454462760673559e-05, + "loss": 0.0242, + "step": 2290 + }, + { + "epoch": 1.8898931799506984, + "grad_norm": 0.37421172265077624, + "learning_rate": 9.446291280363551e-05, + "loss": 0.0239, + "step": 2300 + }, + { + "epoch": 1.8981101068200492, + "grad_norm": 0.3832912238881792, + "learning_rate": 9.43806264025457e-05, + "loss": 0.023, + "step": 2310 + }, + { + "epoch": 1.9063270336894003, + "grad_norm": 0.3814194428999421, + "learning_rate": 9.429776946130826e-05, + "loss": 0.0241, + "step": 2320 + }, + { + "epoch": 1.9145439605587509, + "grad_norm": 0.39219173513731687, + "learning_rate": 9.421434304509992e-05, + "loss": 0.0228, + "step": 2330 + }, + { + "epoch": 1.922760887428102, + "grad_norm": 0.3543509894192511, + "learning_rate": 9.413034822641844e-05, + "loss": 0.0232, + "step": 2340 + }, + { + "epoch": 1.9309778142974527, + "grad_norm": 0.35733995786931505, + "learning_rate": 9.404578608506865e-05, + "loss": 0.0228, + "step": 2350 + }, + { + "epoch": 1.9391947411668036, + "grad_norm": 0.386740275684357, + "learning_rate": 9.396065770814867e-05, + "loss": 0.0232, + "step": 2360 + }, + { + "epoch": 1.9474116680361546, + "grad_norm": 0.44141491887351664, + "learning_rate": 9.387496419003599e-05, + "loss": 0.0233, + "step": 2370 + }, + { + "epoch": 1.9556285949055052, + "grad_norm": 0.3802346697580586, + "learning_rate": 9.378870663237328e-05, + "loss": 0.022, + "step": 2380 + }, + { + "epoch": 1.9638455217748563, + "grad_norm": 0.3692117355721275, + "learning_rate": 9.37018861440543e-05, + "loss": 0.0223, + "step": 2390 + }, + { + "epoch": 1.972062448644207, + "grad_norm": 0.39846659821583263, + "learning_rate": 9.361450384120963e-05, + "loss": 0.0215, + "step": 2400 + }, + { + "epoch": 1.980279375513558, + "grad_norm": 0.3262153453252359, + "learning_rate": 9.352656084719231e-05, + "loss": 0.0212, + "step": 2410 + }, + { + "epoch": 1.9884963023829088, + "grad_norm": 0.34632528206609053, + "learning_rate": 9.343805829256343e-05, + "loss": 0.0214, + "step": 2420 + }, + { + "epoch": 1.9967132292522596, + "grad_norm": 0.3654934223559427, + "learning_rate": 9.334899731507756e-05, + "loss": 0.021, + "step": 2430 + }, + { + "epoch": 2.0049301561216106, + "grad_norm": 0.26848148846319253, + "learning_rate": 9.325937905966813e-05, + "loss": 0.0164, + "step": 2440 + }, + { + "epoch": 2.0131470829909612, + "grad_norm": 0.29112160593983694, + "learning_rate": 9.316920467843273e-05, + "loss": 0.0126, + "step": 2450 + }, + { + "epoch": 2.0213640098603123, + "grad_norm": 0.25419379869277514, + "learning_rate": 9.307847533061831e-05, + "loss": 0.0126, + "step": 2460 + }, + { + "epoch": 2.029580936729663, + "grad_norm": 0.284347026269926, + "learning_rate": 9.29871921826062e-05, + "loss": 0.0115, + "step": 2470 + }, + { + "epoch": 2.037797863599014, + "grad_norm": 0.2804187708379741, + "learning_rate": 9.289535640789721e-05, + "loss": 0.0109, + "step": 2480 + }, + { + "epoch": 2.046014790468365, + "grad_norm": 0.21561701614022483, + "learning_rate": 9.280296918709653e-05, + "loss": 0.0112, + "step": 2490 + }, + { + "epoch": 2.0542317173377156, + "grad_norm": 0.2770066140071947, + "learning_rate": 9.271003170789846e-05, + "loss": 0.0106, + "step": 2500 + }, + { + "epoch": 2.0624486442070666, + "grad_norm": 0.3110912363010226, + "learning_rate": 9.261654516507126e-05, + "loss": 0.0108, + "step": 2510 + }, + { + "epoch": 2.0706655710764172, + "grad_norm": 0.27296541462941787, + "learning_rate": 9.252251076044172e-05, + "loss": 0.0109, + "step": 2520 + }, + { + "epoch": 2.0788824979457683, + "grad_norm": 0.24002967576989534, + "learning_rate": 9.242792970287972e-05, + "loss": 0.0108, + "step": 2530 + }, + { + "epoch": 2.0870994248151193, + "grad_norm": 0.29764076691462776, + "learning_rate": 9.233280320828274e-05, + "loss": 0.0117, + "step": 2540 + }, + { + "epoch": 2.09531635168447, + "grad_norm": 0.2913901487844858, + "learning_rate": 9.223713249956016e-05, + "loss": 0.0108, + "step": 2550 + }, + { + "epoch": 2.103533278553821, + "grad_norm": 0.24293966139484074, + "learning_rate": 9.214091880661752e-05, + "loss": 0.0104, + "step": 2560 + }, + { + "epoch": 2.1117502054231716, + "grad_norm": 0.24074539605557427, + "learning_rate": 9.204416336634084e-05, + "loss": 0.0099, + "step": 2570 + }, + { + "epoch": 2.1199671322925226, + "grad_norm": 0.2999981940123419, + "learning_rate": 9.194686742258059e-05, + "loss": 0.0104, + "step": 2580 + }, + { + "epoch": 2.1281840591618733, + "grad_norm": 0.25277770668402233, + "learning_rate": 9.184903222613574e-05, + "loss": 0.0102, + "step": 2590 + }, + { + "epoch": 2.1364009860312243, + "grad_norm": 0.26178176275955506, + "learning_rate": 9.175065903473768e-05, + "loss": 0.0103, + "step": 2600 + }, + { + "epoch": 2.1446179129005754, + "grad_norm": 0.26730537093378065, + "learning_rate": 9.165174911303409e-05, + "loss": 0.0103, + "step": 2610 + }, + { + "epoch": 2.152834839769926, + "grad_norm": 0.2931413736706241, + "learning_rate": 9.15523037325726e-05, + "loss": 0.0108, + "step": 2620 + }, + { + "epoch": 2.161051766639277, + "grad_norm": 0.2870887376128566, + "learning_rate": 9.145232417178453e-05, + "loss": 0.0102, + "step": 2630 + }, + { + "epoch": 2.1692686935086276, + "grad_norm": 0.24704281922605292, + "learning_rate": 9.135181171596838e-05, + "loss": 0.0103, + "step": 2640 + }, + { + "epoch": 2.1774856203779787, + "grad_norm": 0.26790802194061564, + "learning_rate": 9.125076765727338e-05, + "loss": 0.011, + "step": 2650 + }, + { + "epoch": 2.1857025472473293, + "grad_norm": 0.3449249477967815, + "learning_rate": 9.114919329468282e-05, + "loss": 0.0116, + "step": 2660 + }, + { + "epoch": 2.1939194741166803, + "grad_norm": 0.30244563420360604, + "learning_rate": 9.104708993399738e-05, + "loss": 0.011, + "step": 2670 + }, + { + "epoch": 2.2021364009860314, + "grad_norm": 0.30736920260370454, + "learning_rate": 9.094445888781829e-05, + "loss": 0.0107, + "step": 2680 + }, + { + "epoch": 2.210353327855382, + "grad_norm": 0.30219679969169827, + "learning_rate": 9.084130147553057e-05, + "loss": 0.0116, + "step": 2690 + }, + { + "epoch": 2.218570254724733, + "grad_norm": 0.3286996402717758, + "learning_rate": 9.073761902328597e-05, + "loss": 0.0102, + "step": 2700 + }, + { + "epoch": 2.2267871815940836, + "grad_norm": 0.30798545297089197, + "learning_rate": 9.063341286398592e-05, + "loss": 0.0106, + "step": 2710 + }, + { + "epoch": 2.2350041084634347, + "grad_norm": 0.33718978166952857, + "learning_rate": 9.052868433726448e-05, + "loss": 0.0105, + "step": 2720 + }, + { + "epoch": 2.2432210353327857, + "grad_norm": 0.2954643606610759, + "learning_rate": 9.042343478947103e-05, + "loss": 0.011, + "step": 2730 + }, + { + "epoch": 2.2514379622021363, + "grad_norm": 0.30272236743742853, + "learning_rate": 9.0317665573653e-05, + "loss": 0.011, + "step": 2740 + }, + { + "epoch": 2.2596548890714874, + "grad_norm": 0.3302443829248607, + "learning_rate": 9.021137804953846e-05, + "loss": 0.0112, + "step": 2750 + }, + { + "epoch": 2.267871815940838, + "grad_norm": 0.29949297685046744, + "learning_rate": 9.010457358351867e-05, + "loss": 0.0109, + "step": 2760 + }, + { + "epoch": 2.276088742810189, + "grad_norm": 0.2724599152940893, + "learning_rate": 8.99972535486305e-05, + "loss": 0.0114, + "step": 2770 + }, + { + "epoch": 2.28430566967954, + "grad_norm": 0.31846220224170235, + "learning_rate": 8.988941932453872e-05, + "loss": 0.0118, + "step": 2780 + }, + { + "epoch": 2.2925225965488907, + "grad_norm": 0.29238177724291164, + "learning_rate": 8.978107229751842e-05, + "loss": 0.0108, + "step": 2790 + }, + { + "epoch": 2.3007395234182417, + "grad_norm": 0.27814751669130666, + "learning_rate": 8.967221386043699e-05, + "loss": 0.0112, + "step": 2800 + }, + { + "epoch": 2.3089564502875923, + "grad_norm": 0.29768803452504616, + "learning_rate": 8.956284541273632e-05, + "loss": 0.0108, + "step": 2810 + }, + { + "epoch": 2.3171733771569434, + "grad_norm": 0.30479438136210424, + "learning_rate": 8.945296836041487e-05, + "loss": 0.011, + "step": 2820 + }, + { + "epoch": 2.325390304026294, + "grad_norm": 0.30400201170563734, + "learning_rate": 8.934258411600945e-05, + "loss": 0.0111, + "step": 2830 + }, + { + "epoch": 2.333607230895645, + "grad_norm": 0.279262780416122, + "learning_rate": 8.923169409857718e-05, + "loss": 0.011, + "step": 2840 + }, + { + "epoch": 2.341824157764996, + "grad_norm": 0.3105378625590925, + "learning_rate": 8.912029973367717e-05, + "loss": 0.0107, + "step": 2850 + }, + { + "epoch": 2.3500410846343467, + "grad_norm": 0.29400831967333496, + "learning_rate": 8.900840245335225e-05, + "loss": 0.0113, + "step": 2860 + }, + { + "epoch": 2.3582580115036977, + "grad_norm": 0.2893100065249251, + "learning_rate": 8.889600369611051e-05, + "loss": 0.0114, + "step": 2870 + }, + { + "epoch": 2.3664749383730483, + "grad_norm": 0.2815383019080243, + "learning_rate": 8.878310490690686e-05, + "loss": 0.011, + "step": 2880 + }, + { + "epoch": 2.3746918652423994, + "grad_norm": 0.3427726387197821, + "learning_rate": 8.866970753712442e-05, + "loss": 0.0113, + "step": 2890 + }, + { + "epoch": 2.38290879211175, + "grad_norm": 0.278157815073135, + "learning_rate": 8.855581304455584e-05, + "loss": 0.0108, + "step": 2900 + }, + { + "epoch": 2.391125718981101, + "grad_norm": 0.29741130920284337, + "learning_rate": 8.844142289338466e-05, + "loss": 0.0104, + "step": 2910 + }, + { + "epoch": 2.399342645850452, + "grad_norm": 0.2729299367333149, + "learning_rate": 8.832653855416634e-05, + "loss": 0.0104, + "step": 2920 + }, + { + "epoch": 2.4075595727198027, + "grad_norm": 0.2865288059261725, + "learning_rate": 8.821116150380946e-05, + "loss": 0.011, + "step": 2930 + }, + { + "epoch": 2.4157764995891537, + "grad_norm": 0.2791177339025258, + "learning_rate": 8.80952932255567e-05, + "loss": 0.0107, + "step": 2940 + }, + { + "epoch": 2.4239934264585044, + "grad_norm": 0.2980724885727162, + "learning_rate": 8.797893520896579e-05, + "loss": 0.011, + "step": 2950 + }, + { + "epoch": 2.4322103533278554, + "grad_norm": 0.31076617373261356, + "learning_rate": 8.786208894989032e-05, + "loss": 0.0108, + "step": 2960 + }, + { + "epoch": 2.4404272801972064, + "grad_norm": 0.30152315696501025, + "learning_rate": 8.774475595046055e-05, + "loss": 0.0111, + "step": 2970 + }, + { + "epoch": 2.448644207066557, + "grad_norm": 0.2609839163777236, + "learning_rate": 8.762693771906409e-05, + "loss": 0.0109, + "step": 2980 + }, + { + "epoch": 2.456861133935908, + "grad_norm": 0.30973305475716906, + "learning_rate": 8.750863577032652e-05, + "loss": 0.011, + "step": 2990 + }, + { + "epoch": 2.4650780608052587, + "grad_norm": 0.3393765462966183, + "learning_rate": 8.738985162509186e-05, + "loss": 0.0112, + "step": 3000 + }, + { + "epoch": 2.4732949876746098, + "grad_norm": 0.3048152571960306, + "learning_rate": 8.727058681040308e-05, + "loss": 0.011, + "step": 3010 + }, + { + "epoch": 2.481511914543961, + "grad_norm": 0.25716354456976137, + "learning_rate": 8.715084285948248e-05, + "loss": 0.0107, + "step": 3020 + }, + { + "epoch": 2.4897288414133114, + "grad_norm": 0.3076118802762377, + "learning_rate": 8.703062131171192e-05, + "loss": 0.0109, + "step": 3030 + }, + { + "epoch": 2.4979457682826625, + "grad_norm": 0.2897555835996962, + "learning_rate": 8.690992371261302e-05, + "loss": 0.0105, + "step": 3040 + }, + { + "epoch": 2.506162695152013, + "grad_norm": 0.2990726423763531, + "learning_rate": 8.678875161382745e-05, + "loss": 0.0114, + "step": 3050 + }, + { + "epoch": 2.514379622021364, + "grad_norm": 0.3308837352116956, + "learning_rate": 8.666710657309672e-05, + "loss": 0.011, + "step": 3060 + }, + { + "epoch": 2.5225965488907147, + "grad_norm": 0.2783330323448899, + "learning_rate": 8.654499015424243e-05, + "loss": 0.0106, + "step": 3070 + }, + { + "epoch": 2.5308134757600658, + "grad_norm": 0.2834321980541149, + "learning_rate": 8.642240392714593e-05, + "loss": 0.0108, + "step": 3080 + }, + { + "epoch": 2.5390304026294164, + "grad_norm": 0.21923295269567925, + "learning_rate": 8.629934946772831e-05, + "loss": 0.0105, + "step": 3090 + }, + { + "epoch": 2.5472473294987674, + "grad_norm": 0.3113319375295575, + "learning_rate": 8.617582835793003e-05, + "loss": 0.011, + "step": 3100 + }, + { + "epoch": 2.5554642563681185, + "grad_norm": 0.22522445036106728, + "learning_rate": 8.605184218569067e-05, + "loss": 0.0108, + "step": 3110 + }, + { + "epoch": 2.563681183237469, + "grad_norm": 0.29129457748014026, + "learning_rate": 8.592739254492844e-05, + "loss": 0.0114, + "step": 3120 + }, + { + "epoch": 2.57189811010682, + "grad_norm": 0.298278233144733, + "learning_rate": 8.580248103551973e-05, + "loss": 0.0104, + "step": 3130 + }, + { + "epoch": 2.5801150369761707, + "grad_norm": 0.28438151731996136, + "learning_rate": 8.567710926327855e-05, + "loss": 0.0109, + "step": 3140 + }, + { + "epoch": 2.5883319638455218, + "grad_norm": 0.29615207997528636, + "learning_rate": 8.555127883993582e-05, + "loss": 0.0111, + "step": 3150 + }, + { + "epoch": 2.596548890714873, + "grad_norm": 0.2668249857546702, + "learning_rate": 8.542499138311877e-05, + "loss": 0.0104, + "step": 3160 + }, + { + "epoch": 2.6047658175842234, + "grad_norm": 0.29885476401362265, + "learning_rate": 8.529824851633004e-05, + "loss": 0.0109, + "step": 3170 + }, + { + "epoch": 2.6129827444535745, + "grad_norm": 0.2964635585154554, + "learning_rate": 8.517105186892683e-05, + "loss": 0.0103, + "step": 3180 + }, + { + "epoch": 2.621199671322925, + "grad_norm": 0.26419372229458143, + "learning_rate": 8.50434030761e-05, + "loss": 0.0102, + "step": 3190 + }, + { + "epoch": 2.629416598192276, + "grad_norm": 0.26957526996934655, + "learning_rate": 8.4915303778853e-05, + "loss": 0.0105, + "step": 3200 + }, + { + "epoch": 2.637633525061627, + "grad_norm": 0.30995887012931245, + "learning_rate": 8.478675562398077e-05, + "loss": 0.0105, + "step": 3210 + }, + { + "epoch": 2.645850451930978, + "grad_norm": 0.2885157857565514, + "learning_rate": 8.465776026404866e-05, + "loss": 0.0112, + "step": 3220 + }, + { + "epoch": 2.654067378800329, + "grad_norm": 0.33415184613805593, + "learning_rate": 8.452831935737102e-05, + "loss": 0.0106, + "step": 3230 + }, + { + "epoch": 2.6622843056696794, + "grad_norm": 0.2978809771443008, + "learning_rate": 8.439843456799005e-05, + "loss": 0.0101, + "step": 3240 + }, + { + "epoch": 2.6705012325390305, + "grad_norm": 0.25393760859554587, + "learning_rate": 8.426810756565429e-05, + "loss": 0.0107, + "step": 3250 + }, + { + "epoch": 2.6787181594083815, + "grad_norm": 0.2839448149328719, + "learning_rate": 8.413734002579723e-05, + "loss": 0.0111, + "step": 3260 + }, + { + "epoch": 2.686935086277732, + "grad_norm": 0.2721771832897388, + "learning_rate": 8.40061336295157e-05, + "loss": 0.0111, + "step": 3270 + }, + { + "epoch": 2.6951520131470827, + "grad_norm": 0.2872702777954717, + "learning_rate": 8.387449006354837e-05, + "loss": 0.0103, + "step": 3280 + }, + { + "epoch": 2.703368940016434, + "grad_norm": 0.29596031948807433, + "learning_rate": 8.374241102025391e-05, + "loss": 0.0104, + "step": 3290 + }, + { + "epoch": 2.711585866885785, + "grad_norm": 0.27424169300037154, + "learning_rate": 8.360989819758935e-05, + "loss": 0.0103, + "step": 3300 + }, + { + "epoch": 2.7198027937551355, + "grad_norm": 0.2699351448831838, + "learning_rate": 8.347695329908827e-05, + "loss": 0.0103, + "step": 3310 + }, + { + "epoch": 2.7280197206244865, + "grad_norm": 0.26678359974892135, + "learning_rate": 8.334357803383875e-05, + "loss": 0.0102, + "step": 3320 + }, + { + "epoch": 2.736236647493837, + "grad_norm": 0.2948773860549256, + "learning_rate": 8.32097741164616e-05, + "loss": 0.0096, + "step": 3330 + }, + { + "epoch": 2.744453574363188, + "grad_norm": 0.3090800658579484, + "learning_rate": 8.307554326708815e-05, + "loss": 0.0102, + "step": 3340 + }, + { + "epoch": 2.752670501232539, + "grad_norm": 0.2735640172287781, + "learning_rate": 8.294088721133821e-05, + "loss": 0.0097, + "step": 3350 + }, + { + "epoch": 2.76088742810189, + "grad_norm": 0.26265127181426384, + "learning_rate": 8.28058076802979e-05, + "loss": 0.0103, + "step": 3360 + }, + { + "epoch": 2.769104354971241, + "grad_norm": 0.3267515866056434, + "learning_rate": 8.267030641049738e-05, + "loss": 0.0105, + "step": 3370 + }, + { + "epoch": 2.7773212818405915, + "grad_norm": 0.2807694986872119, + "learning_rate": 8.25343851438885e-05, + "loss": 0.0101, + "step": 3380 + }, + { + "epoch": 2.7855382087099425, + "grad_norm": 0.2918400047001943, + "learning_rate": 8.239804562782241e-05, + "loss": 0.0103, + "step": 3390 + }, + { + "epoch": 2.7937551355792936, + "grad_norm": 0.3368518827581203, + "learning_rate": 8.226128961502715e-05, + "loss": 0.0099, + "step": 3400 + }, + { + "epoch": 2.801972062448644, + "grad_norm": 0.30338648344652236, + "learning_rate": 8.212411886358505e-05, + "loss": 0.0105, + "step": 3410 + }, + { + "epoch": 2.810188989317995, + "grad_norm": 0.26222245141291084, + "learning_rate": 8.198653513691019e-05, + "loss": 0.0096, + "step": 3420 + }, + { + "epoch": 2.818405916187346, + "grad_norm": 0.31220931761376275, + "learning_rate": 8.184854020372567e-05, + "loss": 0.0107, + "step": 3430 + }, + { + "epoch": 2.826622843056697, + "grad_norm": 0.26292349810527665, + "learning_rate": 8.171013583804095e-05, + "loss": 0.0099, + "step": 3440 + }, + { + "epoch": 2.834839769926048, + "grad_norm": 0.28178421575879453, + "learning_rate": 8.157132381912894e-05, + "loss": 0.01, + "step": 3450 + }, + { + "epoch": 2.8430566967953985, + "grad_norm": 0.2794905489623295, + "learning_rate": 8.143210593150317e-05, + "loss": 0.0109, + "step": 3460 + }, + { + "epoch": 2.851273623664749, + "grad_norm": 0.30819250761179967, + "learning_rate": 8.129248396489493e-05, + "loss": 0.0105, + "step": 3470 + }, + { + "epoch": 2.8594905505341, + "grad_norm": 0.30996445249233995, + "learning_rate": 8.115245971423016e-05, + "loss": 0.0103, + "step": 3480 + }, + { + "epoch": 2.867707477403451, + "grad_norm": 0.2998585643222057, + "learning_rate": 8.101203497960639e-05, + "loss": 0.0103, + "step": 3490 + }, + { + "epoch": 2.875924404272802, + "grad_norm": 0.2886522658424423, + "learning_rate": 8.087121156626965e-05, + "loss": 0.0102, + "step": 3500 + }, + { + "epoch": 2.884141331142153, + "grad_norm": 0.2896678741036796, + "learning_rate": 8.072999128459119e-05, + "loss": 0.0101, + "step": 3510 + }, + { + "epoch": 2.8923582580115035, + "grad_norm": 0.2668371079506528, + "learning_rate": 8.05883759500443e-05, + "loss": 0.0101, + "step": 3520 + }, + { + "epoch": 2.9005751848808545, + "grad_norm": 0.262945737927121, + "learning_rate": 8.044636738318082e-05, + "loss": 0.01, + "step": 3530 + }, + { + "epoch": 2.9087921117502056, + "grad_norm": 0.3051556790714007, + "learning_rate": 8.030396740960795e-05, + "loss": 0.0106, + "step": 3540 + }, + { + "epoch": 2.917009038619556, + "grad_norm": 0.2681398854835005, + "learning_rate": 8.01611778599646e-05, + "loss": 0.0095, + "step": 3550 + }, + { + "epoch": 2.9252259654889072, + "grad_norm": 0.2812885203155075, + "learning_rate": 8.001800056989792e-05, + "loss": 0.0096, + "step": 3560 + }, + { + "epoch": 2.933442892358258, + "grad_norm": 0.28431854280176366, + "learning_rate": 7.987443738003973e-05, + "loss": 0.009, + "step": 3570 + }, + { + "epoch": 2.941659819227609, + "grad_norm": 0.25500549263286926, + "learning_rate": 7.973049013598277e-05, + "loss": 0.0092, + "step": 3580 + }, + { + "epoch": 2.94987674609696, + "grad_norm": 0.2804777447797684, + "learning_rate": 7.958616068825709e-05, + "loss": 0.0099, + "step": 3590 + }, + { + "epoch": 2.9580936729663105, + "grad_norm": 0.2837446971120676, + "learning_rate": 7.944145089230619e-05, + "loss": 0.0098, + "step": 3600 + }, + { + "epoch": 2.9663105998356616, + "grad_norm": 0.3227611704344215, + "learning_rate": 7.929636260846316e-05, + "loss": 0.0094, + "step": 3610 + }, + { + "epoch": 2.974527526705012, + "grad_norm": 0.24511611699828814, + "learning_rate": 7.915089770192679e-05, + "loss": 0.0091, + "step": 3620 + }, + { + "epoch": 2.9827444535743632, + "grad_norm": 0.25170601641570906, + "learning_rate": 7.900505804273759e-05, + "loss": 0.0097, + "step": 3630 + }, + { + "epoch": 2.9909613804437143, + "grad_norm": 0.24467025269265683, + "learning_rate": 7.885884550575376e-05, + "loss": 0.0094, + "step": 3640 + }, + { + "epoch": 2.999178307313065, + "grad_norm": 0.30300613229858187, + "learning_rate": 7.871226197062705e-05, + "loss": 0.0101, + "step": 3650 + }, + { + "epoch": 3.007395234182416, + "grad_norm": 0.2553155644623302, + "learning_rate": 7.856530932177862e-05, + "loss": 0.0057, + "step": 3660 + }, + { + "epoch": 3.0156121610517665, + "grad_norm": 0.161635559138198, + "learning_rate": 7.841798944837483e-05, + "loss": 0.005, + "step": 3670 + }, + { + "epoch": 3.0238290879211176, + "grad_norm": 0.21798459223868738, + "learning_rate": 7.827030424430292e-05, + "loss": 0.0051, + "step": 3680 + }, + { + "epoch": 3.032046014790468, + "grad_norm": 0.19996462325026684, + "learning_rate": 7.812225560814669e-05, + "loss": 0.0047, + "step": 3690 + }, + { + "epoch": 3.0402629416598193, + "grad_norm": 0.21044170035200818, + "learning_rate": 7.797384544316207e-05, + "loss": 0.0044, + "step": 3700 + }, + { + "epoch": 3.0484798685291703, + "grad_norm": 0.17148623896064416, + "learning_rate": 7.782507565725267e-05, + "loss": 0.0043, + "step": 3710 + }, + { + "epoch": 3.056696795398521, + "grad_norm": 0.1593250206284325, + "learning_rate": 7.767594816294528e-05, + "loss": 0.0039, + "step": 3720 + }, + { + "epoch": 3.064913722267872, + "grad_norm": 0.16866106993385668, + "learning_rate": 7.752646487736519e-05, + "loss": 0.0038, + "step": 3730 + }, + { + "epoch": 3.0731306491372226, + "grad_norm": 0.16963910509449706, + "learning_rate": 7.737662772221165e-05, + "loss": 0.0041, + "step": 3740 + }, + { + "epoch": 3.0813475760065736, + "grad_norm": 0.17845171206464283, + "learning_rate": 7.722643862373309e-05, + "loss": 0.0044, + "step": 3750 + }, + { + "epoch": 3.089564502875924, + "grad_norm": 0.186635114157523, + "learning_rate": 7.707589951270242e-05, + "loss": 0.0038, + "step": 3760 + }, + { + "epoch": 3.0977814297452753, + "grad_norm": 0.1703635035429931, + "learning_rate": 7.692501232439214e-05, + "loss": 0.0038, + "step": 3770 + }, + { + "epoch": 3.1059983566146263, + "grad_norm": 0.19056650995512647, + "learning_rate": 7.677377899854954e-05, + "loss": 0.0038, + "step": 3780 + }, + { + "epoch": 3.114215283483977, + "grad_norm": 0.17879847530750556, + "learning_rate": 7.662220147937167e-05, + "loss": 0.004, + "step": 3790 + }, + { + "epoch": 3.122432210353328, + "grad_norm": 0.1563794811454263, + "learning_rate": 7.647028171548043e-05, + "loss": 0.0038, + "step": 3800 + }, + { + "epoch": 3.1306491372226786, + "grad_norm": 0.15522383132353818, + "learning_rate": 7.631802165989749e-05, + "loss": 0.0038, + "step": 3810 + }, + { + "epoch": 3.1388660640920296, + "grad_norm": 0.17065109715005505, + "learning_rate": 7.616542327001915e-05, + "loss": 0.0034, + "step": 3820 + }, + { + "epoch": 3.1470829909613807, + "grad_norm": 0.18964115722808972, + "learning_rate": 7.601248850759122e-05, + "loss": 0.0037, + "step": 3830 + }, + { + "epoch": 3.1552999178307313, + "grad_norm": 0.19369231771575623, + "learning_rate": 7.585921933868379e-05, + "loss": 0.0039, + "step": 3840 + }, + { + "epoch": 3.1635168447000823, + "grad_norm": 0.1708702973124068, + "learning_rate": 7.570561773366596e-05, + "loss": 0.0036, + "step": 3850 + }, + { + "epoch": 3.171733771569433, + "grad_norm": 0.1664287179991504, + "learning_rate": 7.555168566718046e-05, + "loss": 0.0039, + "step": 3860 + }, + { + "epoch": 3.179950698438784, + "grad_norm": 0.1909884799322728, + "learning_rate": 7.539742511811833e-05, + "loss": 0.0034, + "step": 3870 + }, + { + "epoch": 3.1881676253081346, + "grad_norm": 0.13823755818616978, + "learning_rate": 7.52428380695935e-05, + "loss": 0.0036, + "step": 3880 + }, + { + "epoch": 3.1963845521774856, + "grad_norm": 0.19048854976208782, + "learning_rate": 7.508792650891713e-05, + "loss": 0.0035, + "step": 3890 + }, + { + "epoch": 3.2046014790468367, + "grad_norm": 0.18423860891951985, + "learning_rate": 7.493269242757233e-05, + "loss": 0.0035, + "step": 3900 + }, + { + "epoch": 3.2128184059161873, + "grad_norm": 0.16535953588600988, + "learning_rate": 7.477713782118829e-05, + "loss": 0.0038, + "step": 3910 + }, + { + "epoch": 3.2210353327855383, + "grad_norm": 0.16467023552242535, + "learning_rate": 7.462126468951483e-05, + "loss": 0.0037, + "step": 3920 + }, + { + "epoch": 3.229252259654889, + "grad_norm": 0.1622463299797392, + "learning_rate": 7.446507503639654e-05, + "loss": 0.0038, + "step": 3930 + }, + { + "epoch": 3.23746918652424, + "grad_norm": 0.16397494619914454, + "learning_rate": 7.430857086974714e-05, + "loss": 0.004, + "step": 3940 + }, + { + "epoch": 3.2456861133935906, + "grad_norm": 0.20002859806848547, + "learning_rate": 7.415175420152358e-05, + "loss": 0.0035, + "step": 3950 + }, + { + "epoch": 3.2539030402629416, + "grad_norm": 0.19911004433428217, + "learning_rate": 7.399462704770023e-05, + "loss": 0.0039, + "step": 3960 + }, + { + "epoch": 3.2621199671322927, + "grad_norm": 0.2263507029551575, + "learning_rate": 7.383719142824296e-05, + "loss": 0.0043, + "step": 3970 + }, + { + "epoch": 3.2703368940016433, + "grad_norm": 0.2031778434080109, + "learning_rate": 7.367944936708309e-05, + "loss": 0.0041, + "step": 3980 + }, + { + "epoch": 3.2785538208709943, + "grad_norm": 0.13904055858611516, + "learning_rate": 7.352140289209149e-05, + "loss": 0.0039, + "step": 3990 + }, + { + "epoch": 3.286770747740345, + "grad_norm": 0.21115955791987592, + "learning_rate": 7.336305403505246e-05, + "loss": 0.004, + "step": 4000 + }, + { + "epoch": 3.294987674609696, + "grad_norm": 0.23197783980269898, + "learning_rate": 7.320440483163757e-05, + "loss": 0.0044, + "step": 4010 + }, + { + "epoch": 3.303204601479047, + "grad_norm": 0.1784402807858933, + "learning_rate": 7.304545732137955e-05, + "loss": 0.0044, + "step": 4020 + }, + { + "epoch": 3.3114215283483976, + "grad_norm": 0.1875146755894153, + "learning_rate": 7.288621354764604e-05, + "loss": 0.0042, + "step": 4030 + }, + { + "epoch": 3.3196384552177487, + "grad_norm": 0.2165579103041975, + "learning_rate": 7.272667555761332e-05, + "loss": 0.0048, + "step": 4040 + }, + { + "epoch": 3.3278553820870993, + "grad_norm": 0.17776521946377408, + "learning_rate": 7.256684540224006e-05, + "loss": 0.0039, + "step": 4050 + }, + { + "epoch": 3.3360723089564503, + "grad_norm": 0.19326883956856228, + "learning_rate": 7.24067251362408e-05, + "loss": 0.0041, + "step": 4060 + }, + { + "epoch": 3.3442892358258014, + "grad_norm": 0.2030379252919242, + "learning_rate": 7.224631681805969e-05, + "loss": 0.0038, + "step": 4070 + }, + { + "epoch": 3.352506162695152, + "grad_norm": 0.18326685170212664, + "learning_rate": 7.208562250984395e-05, + "loss": 0.0042, + "step": 4080 + }, + { + "epoch": 3.360723089564503, + "grad_norm": 0.2635913039945207, + "learning_rate": 7.19246442774174e-05, + "loss": 0.004, + "step": 4090 + }, + { + "epoch": 3.3689400164338537, + "grad_norm": 0.1738566616275142, + "learning_rate": 7.176338419025384e-05, + "loss": 0.0038, + "step": 4100 + }, + { + "epoch": 3.3771569433032047, + "grad_norm": 0.19701001766195692, + "learning_rate": 7.160184432145049e-05, + "loss": 0.0044, + "step": 4110 + }, + { + "epoch": 3.3853738701725553, + "grad_norm": 0.31205136637531716, + "learning_rate": 7.144002674770142e-05, + "loss": 0.0046, + "step": 4120 + }, + { + "epoch": 3.3935907970419064, + "grad_norm": 0.2662957005247976, + "learning_rate": 7.127793354927062e-05, + "loss": 0.0043, + "step": 4130 + }, + { + "epoch": 3.401807723911257, + "grad_norm": 0.18267371828092857, + "learning_rate": 7.111556680996551e-05, + "loss": 0.0049, + "step": 4140 + }, + { + "epoch": 3.410024650780608, + "grad_norm": 0.24800845600219898, + "learning_rate": 7.095292861711005e-05, + "loss": 0.0047, + "step": 4150 + }, + { + "epoch": 3.418241577649959, + "grad_norm": 0.19856665292006692, + "learning_rate": 7.079002106151782e-05, + "loss": 0.0045, + "step": 4160 + }, + { + "epoch": 3.4264585045193097, + "grad_norm": 0.19905680055740865, + "learning_rate": 7.06268462374653e-05, + "loss": 0.0051, + "step": 4170 + }, + { + "epoch": 3.4346754313886607, + "grad_norm": 0.23109221308173886, + "learning_rate": 7.046340624266483e-05, + "loss": 0.0045, + "step": 4180 + }, + { + "epoch": 3.4428923582580113, + "grad_norm": 0.18540375026657363, + "learning_rate": 7.029970317823771e-05, + "loss": 0.0046, + "step": 4190 + }, + { + "epoch": 3.4511092851273624, + "grad_norm": 0.24126679487230862, + "learning_rate": 7.013573914868711e-05, + "loss": 0.0047, + "step": 4200 + }, + { + "epoch": 3.4593262119967134, + "grad_norm": 0.17443909497201535, + "learning_rate": 6.997151626187116e-05, + "loss": 0.0044, + "step": 4210 + }, + { + "epoch": 3.467543138866064, + "grad_norm": 0.21903898676606628, + "learning_rate": 6.980703662897565e-05, + "loss": 0.0046, + "step": 4220 + }, + { + "epoch": 3.475760065735415, + "grad_norm": 0.47925386398079817, + "learning_rate": 6.96423023644871e-05, + "loss": 0.0051, + "step": 4230 + }, + { + "epoch": 3.4839769926047657, + "grad_norm": 0.21967719971828806, + "learning_rate": 6.947731558616535e-05, + "loss": 0.0046, + "step": 4240 + }, + { + "epoch": 3.4921939194741167, + "grad_norm": 0.24128148900378013, + "learning_rate": 6.93120784150166e-05, + "loss": 0.0046, + "step": 4250 + }, + { + "epoch": 3.5004108463434678, + "grad_norm": 0.23177392756125653, + "learning_rate": 6.914659297526593e-05, + "loss": 0.0047, + "step": 4260 + }, + { + "epoch": 3.5086277732128184, + "grad_norm": 0.22320122276510435, + "learning_rate": 6.898086139433003e-05, + "loss": 0.0048, + "step": 4270 + }, + { + "epoch": 3.5168447000821694, + "grad_norm": 0.22969464925909125, + "learning_rate": 6.881488580278999e-05, + "loss": 0.0046, + "step": 4280 + }, + { + "epoch": 3.52506162695152, + "grad_norm": 0.18070239345410893, + "learning_rate": 6.864866833436368e-05, + "loss": 0.0051, + "step": 4290 + }, + { + "epoch": 3.533278553820871, + "grad_norm": 0.24598867660371512, + "learning_rate": 6.848221112587857e-05, + "loss": 0.0048, + "step": 4300 + }, + { + "epoch": 3.541495480690222, + "grad_norm": 0.2175914034410038, + "learning_rate": 6.831551631724402e-05, + "loss": 0.0048, + "step": 4310 + }, + { + "epoch": 3.5497124075595727, + "grad_norm": 0.20476801309531795, + "learning_rate": 6.8148586051424e-05, + "loss": 0.0042, + "step": 4320 + }, + { + "epoch": 3.5579293344289233, + "grad_norm": 0.15384120639623336, + "learning_rate": 6.798142247440932e-05, + "loss": 0.0043, + "step": 4330 + }, + { + "epoch": 3.5661462612982744, + "grad_norm": 0.22126125082354575, + "learning_rate": 6.781402773519021e-05, + "loss": 0.0046, + "step": 4340 + }, + { + "epoch": 3.5743631881676254, + "grad_norm": 0.1988742245455091, + "learning_rate": 6.764640398572864e-05, + "loss": 0.0042, + "step": 4350 + }, + { + "epoch": 3.582580115036976, + "grad_norm": 0.21433952160921746, + "learning_rate": 6.747855338093058e-05, + "loss": 0.0047, + "step": 4360 + }, + { + "epoch": 3.590797041906327, + "grad_norm": 0.18511148015770507, + "learning_rate": 6.731047807861846e-05, + "loss": 0.0049, + "step": 4370 + }, + { + "epoch": 3.5990139687756777, + "grad_norm": 0.23535685882092502, + "learning_rate": 6.714218023950323e-05, + "loss": 0.0049, + "step": 4380 + }, + { + "epoch": 3.6072308956450287, + "grad_norm": 0.24669656378100588, + "learning_rate": 6.697366202715677e-05, + "loss": 0.0049, + "step": 4390 + }, + { + "epoch": 3.61544782251438, + "grad_norm": 0.1642491691407952, + "learning_rate": 6.680492560798396e-05, + "loss": 0.0042, + "step": 4400 + }, + { + "epoch": 3.6236647493837304, + "grad_norm": 0.18863669344103587, + "learning_rate": 6.663597315119487e-05, + "loss": 0.0044, + "step": 4410 + }, + { + "epoch": 3.6318816762530814, + "grad_norm": 0.2075799857074795, + "learning_rate": 6.646680682877687e-05, + "loss": 0.0041, + "step": 4420 + }, + { + "epoch": 3.640098603122432, + "grad_norm": 0.24038171575001221, + "learning_rate": 6.629742881546666e-05, + "loss": 0.0045, + "step": 4430 + }, + { + "epoch": 3.648315529991783, + "grad_norm": 0.19373821701006716, + "learning_rate": 6.612784128872243e-05, + "loss": 0.0044, + "step": 4440 + }, + { + "epoch": 3.656532456861134, + "grad_norm": 0.26219760000934733, + "learning_rate": 6.595804642869572e-05, + "loss": 0.0049, + "step": 4450 + }, + { + "epoch": 3.6647493837304848, + "grad_norm": 0.1780121668656151, + "learning_rate": 6.578804641820352e-05, + "loss": 0.0044, + "step": 4460 + }, + { + "epoch": 3.672966310599836, + "grad_norm": 0.2722581987570137, + "learning_rate": 6.561784344270014e-05, + "loss": 0.0045, + "step": 4470 + }, + { + "epoch": 3.6811832374691864, + "grad_norm": 0.23442737569165115, + "learning_rate": 6.544743969024912e-05, + "loss": 0.0042, + "step": 4480 + }, + { + "epoch": 3.6894001643385375, + "grad_norm": 0.19312032039255483, + "learning_rate": 6.527683735149512e-05, + "loss": 0.0049, + "step": 4490 + }, + { + "epoch": 3.6976170912078885, + "grad_norm": 0.22104390581407352, + "learning_rate": 6.510603861963575e-05, + "loss": 0.0047, + "step": 4500 + }, + { + "epoch": 3.705834018077239, + "grad_norm": 0.17202214959069365, + "learning_rate": 6.493504569039336e-05, + "loss": 0.0047, + "step": 4510 + }, + { + "epoch": 3.7140509449465897, + "grad_norm": 0.20444235013175366, + "learning_rate": 6.476386076198682e-05, + "loss": 0.0049, + "step": 4520 + }, + { + "epoch": 3.7222678718159408, + "grad_norm": 0.15789841316130268, + "learning_rate": 6.459248603510331e-05, + "loss": 0.0044, + "step": 4530 + }, + { + "epoch": 3.730484798685292, + "grad_norm": 0.19701565858775433, + "learning_rate": 6.442092371286996e-05, + "loss": 0.0045, + "step": 4540 + }, + { + "epoch": 3.7387017255546424, + "grad_norm": 0.21510830791255192, + "learning_rate": 6.424917600082551e-05, + "loss": 0.0043, + "step": 4550 + }, + { + "epoch": 3.7469186524239935, + "grad_norm": 0.19555737555381525, + "learning_rate": 6.407724510689208e-05, + "loss": 0.0044, + "step": 4560 + }, + { + "epoch": 3.755135579293344, + "grad_norm": 0.21753813241750067, + "learning_rate": 6.390513324134666e-05, + "loss": 0.0044, + "step": 4570 + }, + { + "epoch": 3.763352506162695, + "grad_norm": 0.19406253627397735, + "learning_rate": 6.373284261679274e-05, + "loss": 0.0047, + "step": 4580 + }, + { + "epoch": 3.771569433032046, + "grad_norm": 0.2186684097360348, + "learning_rate": 6.356037544813186e-05, + "loss": 0.0048, + "step": 4590 + }, + { + "epoch": 3.7797863599013968, + "grad_norm": 0.20669853247165137, + "learning_rate": 6.338773395253518e-05, + "loss": 0.0045, + "step": 4600 + }, + { + "epoch": 3.788003286770748, + "grad_norm": 0.21063217177957072, + "learning_rate": 6.32149203494149e-05, + "loss": 0.0043, + "step": 4610 + }, + { + "epoch": 3.7962202136400984, + "grad_norm": 0.2081955005121378, + "learning_rate": 6.30419368603958e-05, + "loss": 0.0047, + "step": 4620 + }, + { + "epoch": 3.8044371405094495, + "grad_norm": 0.1906883717173744, + "learning_rate": 6.286878570928661e-05, + "loss": 0.0042, + "step": 4630 + }, + { + "epoch": 3.8126540673788005, + "grad_norm": 0.2650865281010175, + "learning_rate": 6.26954691220515e-05, + "loss": 0.0047, + "step": 4640 + }, + { + "epoch": 3.820870994248151, + "grad_norm": 0.24166721788693682, + "learning_rate": 6.252198932678139e-05, + "loss": 0.0049, + "step": 4650 + }, + { + "epoch": 3.829087921117502, + "grad_norm": 0.20466761246056162, + "learning_rate": 6.234834855366536e-05, + "loss": 0.0046, + "step": 4660 + }, + { + "epoch": 3.837304847986853, + "grad_norm": 0.24945575129445083, + "learning_rate": 6.217454903496194e-05, + "loss": 0.0046, + "step": 4670 + }, + { + "epoch": 3.845521774856204, + "grad_norm": 0.20717926987482888, + "learning_rate": 6.200059300497045e-05, + "loss": 0.0045, + "step": 4680 + }, + { + "epoch": 3.853738701725555, + "grad_norm": 0.18631021429913022, + "learning_rate": 6.182648270000226e-05, + "loss": 0.0044, + "step": 4690 + }, + { + "epoch": 3.8619556285949055, + "grad_norm": 0.18705852062600037, + "learning_rate": 6.1652220358352e-05, + "loss": 0.0046, + "step": 4700 + }, + { + "epoch": 3.870172555464256, + "grad_norm": 0.23067465344752314, + "learning_rate": 6.147780822026888e-05, + "loss": 0.0046, + "step": 4710 + }, + { + "epoch": 3.878389482333607, + "grad_norm": 0.20406187242210666, + "learning_rate": 6.13032485279278e-05, + "loss": 0.0042, + "step": 4720 + }, + { + "epoch": 3.886606409202958, + "grad_norm": 0.20558339476006401, + "learning_rate": 6.112854352540058e-05, + "loss": 0.0045, + "step": 4730 + }, + { + "epoch": 3.894823336072309, + "grad_norm": 0.18954011506041926, + "learning_rate": 6.095369545862708e-05, + "loss": 0.0044, + "step": 4740 + }, + { + "epoch": 3.90304026294166, + "grad_norm": 0.21088780233197682, + "learning_rate": 6.0778706575386335e-05, + "loss": 0.0045, + "step": 4750 + }, + { + "epoch": 3.9112571898110104, + "grad_norm": 0.19182428409113075, + "learning_rate": 6.060357912526766e-05, + "loss": 0.0041, + "step": 4760 + }, + { + "epoch": 3.9194741166803615, + "grad_norm": 0.18379309047296535, + "learning_rate": 6.042831535964177e-05, + "loss": 0.0047, + "step": 4770 + }, + { + "epoch": 3.9276910435497125, + "grad_norm": 0.2066878049916885, + "learning_rate": 6.025291753163174e-05, + "loss": 0.0045, + "step": 4780 + }, + { + "epoch": 3.935907970419063, + "grad_norm": 0.17842087564059625, + "learning_rate": 6.007738789608414e-05, + "loss": 0.0042, + "step": 4790 + }, + { + "epoch": 3.944124897288414, + "grad_norm": 0.19115358832374404, + "learning_rate": 5.990172870953997e-05, + "loss": 0.0047, + "step": 4800 + }, + { + "epoch": 3.952341824157765, + "grad_norm": 0.2304758477307126, + "learning_rate": 5.972594223020576e-05, + "loss": 0.0043, + "step": 4810 + }, + { + "epoch": 3.960558751027116, + "grad_norm": 0.24918337033112942, + "learning_rate": 5.955003071792438e-05, + "loss": 0.0049, + "step": 4820 + }, + { + "epoch": 3.968775677896467, + "grad_norm": 0.22164270544229575, + "learning_rate": 5.937399643414611e-05, + "loss": 0.0046, + "step": 4830 + }, + { + "epoch": 3.9769926047658175, + "grad_norm": 0.16636180064885725, + "learning_rate": 5.9197841641899575e-05, + "loss": 0.0044, + "step": 4840 + }, + { + "epoch": 3.9852095316351686, + "grad_norm": 0.22018121804404778, + "learning_rate": 5.9021568605762534e-05, + "loss": 0.0043, + "step": 4850 + }, + { + "epoch": 3.993426458504519, + "grad_norm": 0.19104764872628785, + "learning_rate": 5.884517959183291e-05, + "loss": 0.0043, + "step": 4860 + }, + { + "epoch": 4.00164338537387, + "grad_norm": 0.16273181089666328, + "learning_rate": 5.8668676867699545e-05, + "loss": 0.0037, + "step": 4870 + }, + { + "epoch": 4.009860312243221, + "grad_norm": 0.14031153030330312, + "learning_rate": 5.849206270241313e-05, + "loss": 0.0026, + "step": 4880 + }, + { + "epoch": 4.018077239112572, + "grad_norm": 0.17810977194615177, + "learning_rate": 5.831533936645697e-05, + "loss": 0.0026, + "step": 4890 + }, + { + "epoch": 4.0262941659819225, + "grad_norm": 0.12070674974372055, + "learning_rate": 5.813850913171782e-05, + "loss": 0.0019, + "step": 4900 + }, + { + "epoch": 4.0345110928512735, + "grad_norm": 0.19101003028048594, + "learning_rate": 5.7961574271456766e-05, + "loss": 0.0019, + "step": 4910 + }, + { + "epoch": 4.042728019720625, + "grad_norm": 0.12167846777672976, + "learning_rate": 5.778453706027979e-05, + "loss": 0.0019, + "step": 4920 + }, + { + "epoch": 4.050944946589976, + "grad_norm": 0.12533316088213667, + "learning_rate": 5.7607399774108775e-05, + "loss": 0.0019, + "step": 4930 + }, + { + "epoch": 4.059161873459326, + "grad_norm": 0.16098040185374668, + "learning_rate": 5.743016469015204e-05, + "loss": 0.0018, + "step": 4940 + }, + { + "epoch": 4.067378800328677, + "grad_norm": 0.12346938777973959, + "learning_rate": 5.725283408687524e-05, + "loss": 0.0018, + "step": 4950 + }, + { + "epoch": 4.075595727198028, + "grad_norm": 0.15087616253725697, + "learning_rate": 5.7075410243971896e-05, + "loss": 0.0018, + "step": 4960 + }, + { + "epoch": 4.083812654067379, + "grad_norm": 0.12390044360621293, + "learning_rate": 5.689789544233425e-05, + "loss": 0.0016, + "step": 4970 + }, + { + "epoch": 4.09202958093673, + "grad_norm": 0.10856631216511352, + "learning_rate": 5.6720291964023866e-05, + "loss": 0.0016, + "step": 4980 + }, + { + "epoch": 4.10024650780608, + "grad_norm": 0.0808681205596615, + "learning_rate": 5.654260209224225e-05, + "loss": 0.0017, + "step": 4990 + }, + { + "epoch": 4.108463434675431, + "grad_norm": 0.11307581793343353, + "learning_rate": 5.6364828111301636e-05, + "loss": 0.0016, + "step": 5000 + }, + { + "epoch": 4.116680361544782, + "grad_norm": 0.12184885404041186, + "learning_rate": 5.618697230659545e-05, + "loss": 0.0016, + "step": 5010 + }, + { + "epoch": 4.124897288414133, + "grad_norm": 0.18146076630413796, + "learning_rate": 5.600903696456907e-05, + "loss": 0.0017, + "step": 5020 + }, + { + "epoch": 4.133114215283484, + "grad_norm": 0.11458975751728653, + "learning_rate": 5.583102437269032e-05, + "loss": 0.0015, + "step": 5030 + }, + { + "epoch": 4.1413311421528345, + "grad_norm": 0.14302857289154566, + "learning_rate": 5.565293681942018e-05, + "loss": 0.0017, + "step": 5040 + }, + { + "epoch": 4.1495480690221855, + "grad_norm": 0.14569636180173395, + "learning_rate": 5.5474776594183266e-05, + "loss": 0.0017, + "step": 5050 + }, + { + "epoch": 4.157764995891537, + "grad_norm": 0.11733001575422573, + "learning_rate": 5.5296545987338423e-05, + "loss": 0.0017, + "step": 5060 + }, + { + "epoch": 4.165981922760888, + "grad_norm": 0.15340626640171287, + "learning_rate": 5.511824729014936e-05, + "loss": 0.0017, + "step": 5070 + }, + { + "epoch": 4.174198849630239, + "grad_norm": 0.1527413611619146, + "learning_rate": 5.493988279475504e-05, + "loss": 0.0018, + "step": 5080 + }, + { + "epoch": 4.182415776499589, + "grad_norm": 0.10072787752708584, + "learning_rate": 5.476145479414039e-05, + "loss": 0.0016, + "step": 5090 + }, + { + "epoch": 4.19063270336894, + "grad_norm": 0.12517136882496385, + "learning_rate": 5.458296558210665e-05, + "loss": 0.0015, + "step": 5100 + }, + { + "epoch": 4.198849630238291, + "grad_norm": 0.11755591393444985, + "learning_rate": 5.440441745324205e-05, + "loss": 0.0017, + "step": 5110 + }, + { + "epoch": 4.207066557107642, + "grad_norm": 0.13088199706970402, + "learning_rate": 5.422581270289219e-05, + "loss": 0.0013, + "step": 5120 + }, + { + "epoch": 4.215283483976993, + "grad_norm": 0.15575466242380118, + "learning_rate": 5.4047153627130554e-05, + "loss": 0.0016, + "step": 5130 + }, + { + "epoch": 4.223500410846343, + "grad_norm": 0.1192032192677757, + "learning_rate": 5.386844252272909e-05, + "loss": 0.0017, + "step": 5140 + }, + { + "epoch": 4.231717337715694, + "grad_norm": 0.11710047975800318, + "learning_rate": 5.368968168712849e-05, + "loss": 0.0017, + "step": 5150 + }, + { + "epoch": 4.239934264585045, + "grad_norm": 0.15845063208946036, + "learning_rate": 5.351087341840888e-05, + "loss": 0.0016, + "step": 5160 + }, + { + "epoch": 4.248151191454396, + "grad_norm": 0.06352579799938522, + "learning_rate": 5.3332020015260096e-05, + "loss": 0.0014, + "step": 5170 + }, + { + "epoch": 4.2563681183237465, + "grad_norm": 0.11963200150675178, + "learning_rate": 5.315312377695226e-05, + "loss": 0.0016, + "step": 5180 + }, + { + "epoch": 4.264585045193098, + "grad_norm": 0.15727980882777126, + "learning_rate": 5.297418700330612e-05, + "loss": 0.0016, + "step": 5190 + }, + { + "epoch": 4.272801972062449, + "grad_norm": 0.12595576022850796, + "learning_rate": 5.279521199466356e-05, + "loss": 0.0015, + "step": 5200 + }, + { + "epoch": 4.2810188989318, + "grad_norm": 0.14198177188942032, + "learning_rate": 5.2616201051857975e-05, + "loss": 0.0017, + "step": 5210 + }, + { + "epoch": 4.289235825801151, + "grad_norm": 0.08571106307850862, + "learning_rate": 5.243715647618473e-05, + "loss": 0.0015, + "step": 5220 + }, + { + "epoch": 4.297452752670501, + "grad_norm": 0.19335321961848947, + "learning_rate": 5.225808056937156e-05, + "loss": 0.0016, + "step": 5230 + }, + { + "epoch": 4.305669679539852, + "grad_norm": 0.17584451002798301, + "learning_rate": 5.2078975633548985e-05, + "loss": 0.002, + "step": 5240 + }, + { + "epoch": 4.313886606409203, + "grad_norm": 0.12362312558566853, + "learning_rate": 5.189984397122071e-05, + "loss": 0.0018, + "step": 5250 + }, + { + "epoch": 4.322103533278554, + "grad_norm": 0.12779208588086832, + "learning_rate": 5.1720687885233996e-05, + "loss": 0.0016, + "step": 5260 + }, + { + "epoch": 4.330320460147905, + "grad_norm": 0.12272917918774859, + "learning_rate": 5.1541509678750146e-05, + "loss": 0.0018, + "step": 5270 + }, + { + "epoch": 4.338537387017255, + "grad_norm": 0.13144809599351778, + "learning_rate": 5.1362311655214765e-05, + "loss": 0.0018, + "step": 5280 + }, + { + "epoch": 4.346754313886606, + "grad_norm": 0.1501360407312155, + "learning_rate": 5.118309611832828e-05, + "loss": 0.0016, + "step": 5290 + }, + { + "epoch": 4.354971240755957, + "grad_norm": 0.1342226222912258, + "learning_rate": 5.100386537201621e-05, + "loss": 0.0016, + "step": 5300 + }, + { + "epoch": 4.363188167625308, + "grad_norm": 0.10332645406125511, + "learning_rate": 5.082462172039966e-05, + "loss": 0.0016, + "step": 5310 + }, + { + "epoch": 4.3714050944946585, + "grad_norm": 0.11600118982950115, + "learning_rate": 5.064536746776558e-05, + "loss": 0.0016, + "step": 5320 + }, + { + "epoch": 4.37962202136401, + "grad_norm": 0.13593042660242863, + "learning_rate": 5.046610491853724e-05, + "loss": 0.0016, + "step": 5330 + }, + { + "epoch": 4.387838948233361, + "grad_norm": 0.1103636180089552, + "learning_rate": 5.028683637724455e-05, + "loss": 0.0014, + "step": 5340 + }, + { + "epoch": 4.396055875102712, + "grad_norm": 0.16189492115470053, + "learning_rate": 5.010756414849448e-05, + "loss": 0.0016, + "step": 5350 + }, + { + "epoch": 4.404272801972063, + "grad_norm": 0.11851076824230075, + "learning_rate": 4.9928290536941355e-05, + "loss": 0.0016, + "step": 5360 + }, + { + "epoch": 4.412489728841413, + "grad_norm": 0.17200078664179622, + "learning_rate": 4.9749017847257324e-05, + "loss": 0.0016, + "step": 5370 + }, + { + "epoch": 4.420706655710764, + "grad_norm": 0.17373679665466385, + "learning_rate": 4.956974838410265e-05, + "loss": 0.0016, + "step": 5380 + }, + { + "epoch": 4.428923582580115, + "grad_norm": 0.1366595073623169, + "learning_rate": 4.939048445209617e-05, + "loss": 0.0019, + "step": 5390 + }, + { + "epoch": 4.437140509449466, + "grad_norm": 0.11088120132876558, + "learning_rate": 4.921122835578554e-05, + "loss": 0.0015, + "step": 5400 + }, + { + "epoch": 4.445357436318817, + "grad_norm": 0.1478951748133635, + "learning_rate": 4.9031982399617715e-05, + "loss": 0.0017, + "step": 5410 + }, + { + "epoch": 4.453574363188167, + "grad_norm": 0.09455361863566832, + "learning_rate": 4.885274888790936e-05, + "loss": 0.0016, + "step": 5420 + }, + { + "epoch": 4.461791290057518, + "grad_norm": 0.13977050922055867, + "learning_rate": 4.8673530124817016e-05, + "loss": 0.0017, + "step": 5430 + }, + { + "epoch": 4.470008216926869, + "grad_norm": 0.12339015560603427, + "learning_rate": 4.849432841430775e-05, + "loss": 0.0016, + "step": 5440 + }, + { + "epoch": 4.47822514379622, + "grad_norm": 0.13195741645423847, + "learning_rate": 4.831514606012935e-05, + "loss": 0.0015, + "step": 5450 + }, + { + "epoch": 4.486442070665571, + "grad_norm": 0.1345402711835336, + "learning_rate": 4.81359853657808e-05, + "loss": 0.0015, + "step": 5460 + }, + { + "epoch": 4.494658997534922, + "grad_norm": 0.1388691545527912, + "learning_rate": 4.795684863448258e-05, + "loss": 0.0015, + "step": 5470 + }, + { + "epoch": 4.502875924404273, + "grad_norm": 0.16117893496282612, + "learning_rate": 4.777773816914713e-05, + "loss": 0.0016, + "step": 5480 + }, + { + "epoch": 4.511092851273624, + "grad_norm": 0.16497389400859977, + "learning_rate": 4.75986562723493e-05, + "loss": 0.0018, + "step": 5490 + }, + { + "epoch": 4.519309778142975, + "grad_norm": 0.1379908663299503, + "learning_rate": 4.741960524629652e-05, + "loss": 0.0014, + "step": 5500 + }, + { + "epoch": 4.527526705012326, + "grad_norm": 0.11596689544083429, + "learning_rate": 4.7240587392799516e-05, + "loss": 0.0015, + "step": 5510 + }, + { + "epoch": 4.535743631881676, + "grad_norm": 0.16758502738868467, + "learning_rate": 4.7061605013242456e-05, + "loss": 0.0016, + "step": 5520 + }, + { + "epoch": 4.543960558751027, + "grad_norm": 0.12226214244435939, + "learning_rate": 4.6882660408553536e-05, + "loss": 0.0016, + "step": 5530 + }, + { + "epoch": 4.552177485620378, + "grad_norm": 0.11914211666989985, + "learning_rate": 4.6703755879175275e-05, + "loss": 0.0015, + "step": 5540 + }, + { + "epoch": 4.560394412489729, + "grad_norm": 0.12948836324623705, + "learning_rate": 4.652489372503502e-05, + "loss": 0.0016, + "step": 5550 + }, + { + "epoch": 4.56861133935908, + "grad_norm": 0.11824128435596361, + "learning_rate": 4.6346076245515394e-05, + "loss": 0.0016, + "step": 5560 + }, + { + "epoch": 4.57682826622843, + "grad_norm": 0.12960557482806048, + "learning_rate": 4.616730573942467e-05, + "loss": 0.0017, + "step": 5570 + }, + { + "epoch": 4.585045193097781, + "grad_norm": 0.15837959768390375, + "learning_rate": 4.598858450496722e-05, + "loss": 0.002, + "step": 5580 + }, + { + "epoch": 4.593262119967132, + "grad_norm": 0.12423889996551822, + "learning_rate": 4.580991483971403e-05, + "loss": 0.0017, + "step": 5590 + }, + { + "epoch": 4.6014790468364835, + "grad_norm": 0.1540031422713507, + "learning_rate": 4.563129904057317e-05, + "loss": 0.0016, + "step": 5600 + }, + { + "epoch": 4.6096959737058345, + "grad_norm": 0.17340454139099684, + "learning_rate": 4.5452739403760155e-05, + "loss": 0.0017, + "step": 5610 + }, + { + "epoch": 4.617912900575185, + "grad_norm": 0.164272522667569, + "learning_rate": 4.527423822476852e-05, + "loss": 0.0019, + "step": 5620 + }, + { + "epoch": 4.626129827444536, + "grad_norm": 0.15219766980908056, + "learning_rate": 4.509579779834034e-05, + "loss": 0.0017, + "step": 5630 + }, + { + "epoch": 4.634346754313887, + "grad_norm": 0.1484745735040649, + "learning_rate": 4.4917420418436627e-05, + "loss": 0.0019, + "step": 5640 + }, + { + "epoch": 4.642563681183238, + "grad_norm": 0.14258387648372567, + "learning_rate": 4.47391083782079e-05, + "loss": 0.0016, + "step": 5650 + }, + { + "epoch": 4.650780608052588, + "grad_norm": 0.1914247851999732, + "learning_rate": 4.456086396996467e-05, + "loss": 0.0019, + "step": 5660 + }, + { + "epoch": 4.658997534921939, + "grad_norm": 0.12631285973774356, + "learning_rate": 4.4382689485148096e-05, + "loss": 0.0017, + "step": 5670 + }, + { + "epoch": 4.66721446179129, + "grad_norm": 0.12608426835092465, + "learning_rate": 4.420458721430031e-05, + "loss": 0.0017, + "step": 5680 + }, + { + "epoch": 4.675431388660641, + "grad_norm": 0.11136231365273414, + "learning_rate": 4.402655944703514e-05, + "loss": 0.0019, + "step": 5690 + }, + { + "epoch": 4.683648315529992, + "grad_norm": 0.09984619232229586, + "learning_rate": 4.384860847200864e-05, + "loss": 0.0017, + "step": 5700 + }, + { + "epoch": 4.691865242399342, + "grad_norm": 0.16576927627505217, + "learning_rate": 4.367073657688966e-05, + "loss": 0.0015, + "step": 5710 + }, + { + "epoch": 4.700082169268693, + "grad_norm": 0.13689350163330438, + "learning_rate": 4.349294604833037e-05, + "loss": 0.0018, + "step": 5720 + }, + { + "epoch": 4.708299096138044, + "grad_norm": 0.1543870168232943, + "learning_rate": 4.3315239171936956e-05, + "loss": 0.0016, + "step": 5730 + }, + { + "epoch": 4.7165160230073955, + "grad_norm": 0.13889936677409087, + "learning_rate": 4.313761823224023e-05, + "loss": 0.0018, + "step": 5740 + }, + { + "epoch": 4.724732949876746, + "grad_norm": 0.13878438883995636, + "learning_rate": 4.2960085512666225e-05, + "loss": 0.0015, + "step": 5750 + }, + { + "epoch": 4.732949876746097, + "grad_norm": 0.13054196113503636, + "learning_rate": 4.2782643295506794e-05, + "loss": 0.0016, + "step": 5760 + }, + { + "epoch": 4.741166803615448, + "grad_norm": 0.15145632262571682, + "learning_rate": 4.260529386189041e-05, + "loss": 0.0017, + "step": 5770 + }, + { + "epoch": 4.749383730484799, + "grad_norm": 0.09848701513820292, + "learning_rate": 4.242803949175271e-05, + "loss": 0.0017, + "step": 5780 + }, + { + "epoch": 4.75760065735415, + "grad_norm": 0.1428008669375196, + "learning_rate": 4.225088246380723e-05, + "loss": 0.0016, + "step": 5790 + }, + { + "epoch": 4.7658175842235, + "grad_norm": 0.14445807146841044, + "learning_rate": 4.20738250555161e-05, + "loss": 0.0016, + "step": 5800 + }, + { + "epoch": 4.774034511092851, + "grad_norm": 0.11199748833763745, + "learning_rate": 4.1896869543060846e-05, + "loss": 0.0012, + "step": 5810 + }, + { + "epoch": 4.782251437962202, + "grad_norm": 0.1472929541162101, + "learning_rate": 4.1720018201312995e-05, + "loss": 0.0015, + "step": 5820 + }, + { + "epoch": 4.790468364831553, + "grad_norm": 0.06984693872760507, + "learning_rate": 4.1543273303804886e-05, + "loss": 0.0014, + "step": 5830 + }, + { + "epoch": 4.798685291700904, + "grad_norm": 0.1449482039819671, + "learning_rate": 4.136663712270052e-05, + "loss": 0.0016, + "step": 5840 + }, + { + "epoch": 4.806902218570254, + "grad_norm": 0.1126925094089255, + "learning_rate": 4.119011192876624e-05, + "loss": 0.0016, + "step": 5850 + }, + { + "epoch": 4.815119145439605, + "grad_norm": 0.14392002915242472, + "learning_rate": 4.101369999134156e-05, + "loss": 0.0016, + "step": 5860 + }, + { + "epoch": 4.8233360723089564, + "grad_norm": 0.09327701563723739, + "learning_rate": 4.083740357831003e-05, + "loss": 0.0017, + "step": 5870 + }, + { + "epoch": 4.8315529991783075, + "grad_norm": 0.17846780582919994, + "learning_rate": 4.0661224956070105e-05, + "loss": 0.0013, + "step": 5880 + }, + { + "epoch": 4.8397699260476585, + "grad_norm": 0.22395076452307713, + "learning_rate": 4.048516638950592e-05, + "loss": 0.0018, + "step": 5890 + }, + { + "epoch": 4.847986852917009, + "grad_norm": 0.13910754969362077, + "learning_rate": 4.030923014195821e-05, + "loss": 0.0017, + "step": 5900 + }, + { + "epoch": 4.85620377978636, + "grad_norm": 0.10786783582854183, + "learning_rate": 4.013341847519527e-05, + "loss": 0.0016, + "step": 5910 + }, + { + "epoch": 4.864420706655711, + "grad_norm": 0.1600479840388043, + "learning_rate": 3.995773364938378e-05, + "loss": 0.0015, + "step": 5920 + }, + { + "epoch": 4.872637633525062, + "grad_norm": 0.14304498482382672, + "learning_rate": 3.978217792305989e-05, + "loss": 0.0014, + "step": 5930 + }, + { + "epoch": 4.880854560394413, + "grad_norm": 0.1495718784942365, + "learning_rate": 3.960675355309997e-05, + "loss": 0.0015, + "step": 5940 + }, + { + "epoch": 4.889071487263763, + "grad_norm": 0.13742562135075806, + "learning_rate": 3.9431462794691827e-05, + "loss": 0.0015, + "step": 5950 + }, + { + "epoch": 4.897288414133114, + "grad_norm": 0.14059143591335166, + "learning_rate": 3.925630790130558e-05, + "loss": 0.0015, + "step": 5960 + }, + { + "epoch": 4.905505341002465, + "grad_norm": 0.1272582897759724, + "learning_rate": 3.908129112466467e-05, + "loss": 0.0015, + "step": 5970 + }, + { + "epoch": 4.913722267871816, + "grad_norm": 0.11346883026276684, + "learning_rate": 3.890641471471705e-05, + "loss": 0.0017, + "step": 5980 + }, + { + "epoch": 4.921939194741167, + "grad_norm": 0.19682961593745832, + "learning_rate": 3.8731680919606106e-05, + "loss": 0.0016, + "step": 5990 + }, + { + "epoch": 4.930156121610517, + "grad_norm": 0.14653774266779215, + "learning_rate": 3.855709198564189e-05, + "loss": 0.0016, + "step": 6000 + }, + { + "epoch": 4.9383730484798685, + "grad_norm": 0.14000295835917995, + "learning_rate": 3.838265015727206e-05, + "loss": 0.0016, + "step": 6010 + }, + { + "epoch": 4.9465899753492195, + "grad_norm": 0.1173618999933057, + "learning_rate": 3.820835767705326e-05, + "loss": 0.0018, + "step": 6020 + }, + { + "epoch": 4.954806902218571, + "grad_norm": 0.13558054845706705, + "learning_rate": 3.803421678562213e-05, + "loss": 0.0015, + "step": 6030 + }, + { + "epoch": 4.963023829087922, + "grad_norm": 0.1229069509854008, + "learning_rate": 3.7860229721666494e-05, + "loss": 0.0014, + "step": 6040 + }, + { + "epoch": 4.971240755957272, + "grad_norm": 0.10998811248862572, + "learning_rate": 3.768639872189672e-05, + "loss": 0.0016, + "step": 6050 + }, + { + "epoch": 4.979457682826623, + "grad_norm": 0.1278331117805423, + "learning_rate": 3.7512726021016794e-05, + "loss": 0.0016, + "step": 6060 + }, + { + "epoch": 4.987674609695974, + "grad_norm": 0.1086796177789489, + "learning_rate": 3.733921385169576e-05, + "loss": 0.0015, + "step": 6070 + }, + { + "epoch": 4.995891536565325, + "grad_norm": 0.1402358894590808, + "learning_rate": 3.7165864444538806e-05, + "loss": 0.0014, + "step": 6080 + }, + { + "epoch": 5.004108463434675, + "grad_norm": 0.05521122917926474, + "learning_rate": 3.6992680028058835e-05, + "loss": 0.0012, + "step": 6090 + }, + { + "epoch": 5.012325390304026, + "grad_norm": 0.1282610737186574, + "learning_rate": 3.681966282864762e-05, + "loss": 0.0009, + "step": 6100 + }, + { + "epoch": 5.020542317173377, + "grad_norm": 0.10130100928103676, + "learning_rate": 3.664681507054731e-05, + "loss": 0.0008, + "step": 6110 + }, + { + "epoch": 5.028759244042728, + "grad_norm": 0.10906663584676196, + "learning_rate": 3.6474138975821726e-05, + "loss": 0.0007, + "step": 6120 + }, + { + "epoch": 5.036976170912079, + "grad_norm": 0.08323865583066783, + "learning_rate": 3.630163676432786e-05, + "loss": 0.0007, + "step": 6130 + }, + { + "epoch": 5.045193097781429, + "grad_norm": 0.0879328373850264, + "learning_rate": 3.612931065368741e-05, + "loss": 0.0007, + "step": 6140 + }, + { + "epoch": 5.0534100246507805, + "grad_norm": 0.09059851354271062, + "learning_rate": 3.595716285925804e-05, + "loss": 0.0005, + "step": 6150 + }, + { + "epoch": 5.0616269515201315, + "grad_norm": 0.09855990642602323, + "learning_rate": 3.57851955941052e-05, + "loss": 0.0006, + "step": 6160 + }, + { + "epoch": 5.069843878389483, + "grad_norm": 0.06236994365325129, + "learning_rate": 3.5613411068973416e-05, + "loss": 0.0005, + "step": 6170 + }, + { + "epoch": 5.078060805258834, + "grad_norm": 0.15430891126738064, + "learning_rate": 3.544181149225807e-05, + "loss": 0.0005, + "step": 6180 + }, + { + "epoch": 5.086277732128184, + "grad_norm": 0.05324163041902492, + "learning_rate": 3.527039906997684e-05, + "loss": 0.0005, + "step": 6190 + }, + { + "epoch": 5.094494658997535, + "grad_norm": 0.08058791512594968, + "learning_rate": 3.5099176005741445e-05, + "loss": 0.0006, + "step": 6200 + }, + { + "epoch": 5.102711585866886, + "grad_norm": 0.06433527838817536, + "learning_rate": 3.492814450072934e-05, + "loss": 0.0005, + "step": 6210 + }, + { + "epoch": 5.110928512736237, + "grad_norm": 0.03911496388371956, + "learning_rate": 3.4757306753655266e-05, + "loss": 0.0005, + "step": 6220 + }, + { + "epoch": 5.119145439605587, + "grad_norm": 0.08820383491598298, + "learning_rate": 3.4586664960743184e-05, + "loss": 0.0006, + "step": 6230 + }, + { + "epoch": 5.127362366474938, + "grad_norm": 0.061049528693663965, + "learning_rate": 3.441622131569789e-05, + "loss": 0.0006, + "step": 6240 + }, + { + "epoch": 5.135579293344289, + "grad_norm": 0.062088926900644076, + "learning_rate": 3.424597800967691e-05, + "loss": 0.0006, + "step": 6250 + }, + { + "epoch": 5.14379622021364, + "grad_norm": 0.0691179049188523, + "learning_rate": 3.407593723126223e-05, + "loss": 0.0005, + "step": 6260 + }, + { + "epoch": 5.152013147082991, + "grad_norm": 0.05212931991146724, + "learning_rate": 3.390610116643227e-05, + "loss": 0.0005, + "step": 6270 + }, + { + "epoch": 5.1602300739523415, + "grad_norm": 0.07493691969962672, + "learning_rate": 3.373647199853371e-05, + "loss": 0.0004, + "step": 6280 + }, + { + "epoch": 5.1684470008216925, + "grad_norm": 0.05760941171264436, + "learning_rate": 3.3567051908253465e-05, + "loss": 0.0004, + "step": 6290 + }, + { + "epoch": 5.1766639276910436, + "grad_norm": 0.06408109870628004, + "learning_rate": 3.3397843073590593e-05, + "loss": 0.0005, + "step": 6300 + }, + { + "epoch": 5.184880854560395, + "grad_norm": 0.0679845976956898, + "learning_rate": 3.322884766982833e-05, + "loss": 0.0004, + "step": 6310 + }, + { + "epoch": 5.193097781429746, + "grad_norm": 0.07127864944959102, + "learning_rate": 3.3060067869506204e-05, + "loss": 0.0005, + "step": 6320 + }, + { + "epoch": 5.201314708299096, + "grad_norm": 0.07905018657014251, + "learning_rate": 3.289150584239192e-05, + "loss": 0.0005, + "step": 6330 + }, + { + "epoch": 5.209531635168447, + "grad_norm": 0.030102819859615534, + "learning_rate": 3.272316375545363e-05, + "loss": 0.0004, + "step": 6340 + }, + { + "epoch": 5.217748562037798, + "grad_norm": 0.030664173090735842, + "learning_rate": 3.255504377283205e-05, + "loss": 0.0005, + "step": 6350 + }, + { + "epoch": 5.225965488907149, + "grad_norm": 0.0661821339917806, + "learning_rate": 3.2387148055812564e-05, + "loss": 0.0005, + "step": 6360 + }, + { + "epoch": 5.2341824157765, + "grad_norm": 0.09251211399157652, + "learning_rate": 3.221947876279749e-05, + "loss": 0.0004, + "step": 6370 + }, + { + "epoch": 5.24239934264585, + "grad_norm": 0.06808336533992704, + "learning_rate": 3.205203804927831e-05, + "loss": 0.0004, + "step": 6380 + }, + { + "epoch": 5.250616269515201, + "grad_norm": 0.05873173134989313, + "learning_rate": 3.188482806780805e-05, + "loss": 0.0005, + "step": 6390 + }, + { + "epoch": 5.258833196384552, + "grad_norm": 0.060291263727607985, + "learning_rate": 3.1717850967973436e-05, + "loss": 0.0004, + "step": 6400 + }, + { + "epoch": 5.267050123253903, + "grad_norm": 0.07338281443158083, + "learning_rate": 3.155110889636743e-05, + "loss": 0.0005, + "step": 6410 + }, + { + "epoch": 5.275267050123254, + "grad_norm": 0.015554863760311712, + "learning_rate": 3.1384603996561536e-05, + "loss": 0.0004, + "step": 6420 + }, + { + "epoch": 5.2834839769926045, + "grad_norm": 0.05453872990957685, + "learning_rate": 3.121833840907827e-05, + "loss": 0.0004, + "step": 6430 + }, + { + "epoch": 5.291700903861956, + "grad_norm": 0.10341056353205272, + "learning_rate": 3.105231427136364e-05, + "loss": 0.0004, + "step": 6440 + }, + { + "epoch": 5.299917830731307, + "grad_norm": 0.07217333358079112, + "learning_rate": 3.0886533717759646e-05, + "loss": 0.0004, + "step": 6450 + }, + { + "epoch": 5.308134757600658, + "grad_norm": 0.07514808206178152, + "learning_rate": 3.072099887947694e-05, + "loss": 0.0004, + "step": 6460 + }, + { + "epoch": 5.316351684470008, + "grad_norm": 0.11154664662078861, + "learning_rate": 3.0555711884567274e-05, + "loss": 0.0005, + "step": 6470 + }, + { + "epoch": 5.324568611339359, + "grad_norm": 0.03281456569097846, + "learning_rate": 3.039067485789624e-05, + "loss": 0.0005, + "step": 6480 + }, + { + "epoch": 5.33278553820871, + "grad_norm": 0.0710881728814634, + "learning_rate": 3.0225889921115968e-05, + "loss": 0.0005, + "step": 6490 + }, + { + "epoch": 5.341002465078061, + "grad_norm": 0.14689380660843607, + "learning_rate": 3.0061359192637807e-05, + "loss": 0.0005, + "step": 6500 + }, + { + "epoch": 5.349219391947412, + "grad_norm": 0.105183709054319, + "learning_rate": 2.9897084787605056e-05, + "loss": 0.0005, + "step": 6510 + }, + { + "epoch": 5.357436318816762, + "grad_norm": 0.06230394941058038, + "learning_rate": 2.9733068817865867e-05, + "loss": 0.0005, + "step": 6520 + }, + { + "epoch": 5.365653245686113, + "grad_norm": 0.06797242773932805, + "learning_rate": 2.9569313391946045e-05, + "loss": 0.0004, + "step": 6530 + }, + { + "epoch": 5.373870172555464, + "grad_norm": 0.08738962151430633, + "learning_rate": 2.9405820615021946e-05, + "loss": 0.0004, + "step": 6540 + }, + { + "epoch": 5.382087099424815, + "grad_norm": 0.11197797750902788, + "learning_rate": 2.924259258889336e-05, + "loss": 0.0004, + "step": 6550 + }, + { + "epoch": 5.390304026294166, + "grad_norm": 0.05766340439882582, + "learning_rate": 2.9079631411956622e-05, + "loss": 0.0004, + "step": 6560 + }, + { + "epoch": 5.3985209531635165, + "grad_norm": 0.08026402981652679, + "learning_rate": 2.8916939179177525e-05, + "loss": 0.0004, + "step": 6570 + }, + { + "epoch": 5.406737880032868, + "grad_norm": 0.025327894538300512, + "learning_rate": 2.8754517982064365e-05, + "loss": 0.0003, + "step": 6580 + }, + { + "epoch": 5.414954806902219, + "grad_norm": 0.07737650579928691, + "learning_rate": 2.8592369908641204e-05, + "loss": 0.0005, + "step": 6590 + }, + { + "epoch": 5.42317173377157, + "grad_norm": 0.04100390753406471, + "learning_rate": 2.8430497043420874e-05, + "loss": 0.0004, + "step": 6600 + }, + { + "epoch": 5.43138866064092, + "grad_norm": 0.05039655298171627, + "learning_rate": 2.8268901467378244e-05, + "loss": 0.0005, + "step": 6610 + }, + { + "epoch": 5.439605587510271, + "grad_norm": 0.05143839392371694, + "learning_rate": 2.8107585257923468e-05, + "loss": 0.0004, + "step": 6620 + }, + { + "epoch": 5.447822514379622, + "grad_norm": 0.09671860360954218, + "learning_rate": 2.794655048887527e-05, + "loss": 0.0004, + "step": 6630 + }, + { + "epoch": 5.456039441248973, + "grad_norm": 0.07623773548841196, + "learning_rate": 2.7785799230434285e-05, + "loss": 0.0004, + "step": 6640 + }, + { + "epoch": 5.464256368118324, + "grad_norm": 0.039349058403749575, + "learning_rate": 2.762533354915644e-05, + "loss": 0.0003, + "step": 6650 + }, + { + "epoch": 5.472473294987674, + "grad_norm": 0.03143513466209904, + "learning_rate": 2.74651555079264e-05, + "loss": 0.0003, + "step": 6660 + }, + { + "epoch": 5.480690221857025, + "grad_norm": 0.1016420660696977, + "learning_rate": 2.730526716593103e-05, + "loss": 0.0004, + "step": 6670 + }, + { + "epoch": 5.488907148726376, + "grad_norm": 0.03215006085035493, + "learning_rate": 2.7145670578632954e-05, + "loss": 0.0004, + "step": 6680 + }, + { + "epoch": 5.497124075595727, + "grad_norm": 0.03185067396753877, + "learning_rate": 2.698636779774409e-05, + "loss": 0.0004, + "step": 6690 + }, + { + "epoch": 5.505341002465078, + "grad_norm": 0.07371257590687681, + "learning_rate": 2.6827360871199303e-05, + "loss": 0.0004, + "step": 6700 + }, + { + "epoch": 5.513557929334429, + "grad_norm": 0.03675321228109619, + "learning_rate": 2.666865184313007e-05, + "loss": 0.0004, + "step": 6710 + }, + { + "epoch": 5.52177485620378, + "grad_norm": 0.08372300486264095, + "learning_rate": 2.65102427538382e-05, + "loss": 0.0004, + "step": 6720 + }, + { + "epoch": 5.529991783073131, + "grad_norm": 0.026384399698642574, + "learning_rate": 2.6352135639769603e-05, + "loss": 0.0005, + "step": 6730 + }, + { + "epoch": 5.538208709942482, + "grad_norm": 0.08305188033996418, + "learning_rate": 2.6194332533488118e-05, + "loss": 0.0004, + "step": 6740 + }, + { + "epoch": 5.546425636811833, + "grad_norm": 0.06840448507379106, + "learning_rate": 2.6036835463649373e-05, + "loss": 0.0004, + "step": 6750 + }, + { + "epoch": 5.554642563681183, + "grad_norm": 0.09187541350212843, + "learning_rate": 2.5879646454974703e-05, + "loss": 0.0004, + "step": 6760 + }, + { + "epoch": 5.562859490550534, + "grad_norm": 0.053287878516641324, + "learning_rate": 2.5722767528225145e-05, + "loss": 0.0004, + "step": 6770 + }, + { + "epoch": 5.571076417419885, + "grad_norm": 0.09406204592908068, + "learning_rate": 2.5566200700175424e-05, + "loss": 0.0004, + "step": 6780 + }, + { + "epoch": 5.579293344289236, + "grad_norm": 0.04406812751959359, + "learning_rate": 2.5409947983588066e-05, + "loss": 0.0003, + "step": 6790 + }, + { + "epoch": 5.587510271158587, + "grad_norm": 0.030062788432492375, + "learning_rate": 2.525401138718748e-05, + "loss": 0.0005, + "step": 6800 + }, + { + "epoch": 5.595727198027937, + "grad_norm": 0.04793629527099629, + "learning_rate": 2.509839291563417e-05, + "loss": 0.0003, + "step": 6810 + }, + { + "epoch": 5.603944124897288, + "grad_norm": 0.08190662399628697, + "learning_rate": 2.494309456949895e-05, + "loss": 0.0004, + "step": 6820 + }, + { + "epoch": 5.612161051766639, + "grad_norm": 0.07282716965679993, + "learning_rate": 2.4788118345237216e-05, + "loss": 0.0003, + "step": 6830 + }, + { + "epoch": 5.62037797863599, + "grad_norm": 0.06130522877328794, + "learning_rate": 2.46334662351633e-05, + "loss": 0.0003, + "step": 6840 + }, + { + "epoch": 5.6285949055053415, + "grad_norm": 0.08370123842365505, + "learning_rate": 2.447914022742484e-05, + "loss": 0.0003, + "step": 6850 + }, + { + "epoch": 5.636811832374692, + "grad_norm": 0.1198481107171395, + "learning_rate": 2.432514230597724e-05, + "loss": 0.0003, + "step": 6860 + }, + { + "epoch": 5.645028759244043, + "grad_norm": 0.027595439395732554, + "learning_rate": 2.4171474450558135e-05, + "loss": 0.0003, + "step": 6870 + }, + { + "epoch": 5.653245686113394, + "grad_norm": 0.03839023351802191, + "learning_rate": 2.401813863666198e-05, + "loss": 0.0003, + "step": 6880 + }, + { + "epoch": 5.661462612982745, + "grad_norm": 0.06461758960870899, + "learning_rate": 2.386513683551462e-05, + "loss": 0.0004, + "step": 6890 + }, + { + "epoch": 5.669679539852095, + "grad_norm": 0.023606444211567684, + "learning_rate": 2.3712471014047954e-05, + "loss": 0.0004, + "step": 6900 + }, + { + "epoch": 5.677896466721446, + "grad_norm": 0.05298603587797932, + "learning_rate": 2.356014313487468e-05, + "loss": 0.0004, + "step": 6910 + }, + { + "epoch": 5.686113393590797, + "grad_norm": 0.0395970100991095, + "learning_rate": 2.3408155156263017e-05, + "loss": 0.0004, + "step": 6920 + }, + { + "epoch": 5.694330320460148, + "grad_norm": 0.06108735488244887, + "learning_rate": 2.3256509032111573e-05, + "loss": 0.0003, + "step": 6930 + }, + { + "epoch": 5.702547247329499, + "grad_norm": 0.05720993942255118, + "learning_rate": 2.3105206711924192e-05, + "loss": 0.0004, + "step": 6940 + }, + { + "epoch": 5.710764174198849, + "grad_norm": 0.03319392825608919, + "learning_rate": 2.2954250140784917e-05, + "loss": 0.0003, + "step": 6950 + }, + { + "epoch": 5.7189811010682, + "grad_norm": 0.031406464035255106, + "learning_rate": 2.280364125933297e-05, + "loss": 0.0004, + "step": 6960 + }, + { + "epoch": 5.727198027937551, + "grad_norm": 0.07637068453852462, + "learning_rate": 2.265338200373781e-05, + "loss": 0.0004, + "step": 6970 + }, + { + "epoch": 5.735414954806902, + "grad_norm": 0.09279300817558471, + "learning_rate": 2.250347430567425e-05, + "loss": 0.0004, + "step": 6980 + }, + { + "epoch": 5.743631881676253, + "grad_norm": 0.0410805223696103, + "learning_rate": 2.2353920092297597e-05, + "loss": 0.0003, + "step": 6990 + }, + { + "epoch": 5.751848808545604, + "grad_norm": 0.036955379551948345, + "learning_rate": 2.220472128621891e-05, + "loss": 0.0003, + "step": 7000 + }, + { + "epoch": 5.760065735414955, + "grad_norm": 0.07446562979673903, + "learning_rate": 2.205587980548027e-05, + "loss": 0.0004, + "step": 7010 + }, + { + "epoch": 5.768282662284306, + "grad_norm": 0.08457527682561192, + "learning_rate": 2.1907397563530123e-05, + "loss": 0.0003, + "step": 7020 + }, + { + "epoch": 5.776499589153657, + "grad_norm": 0.049495160014423105, + "learning_rate": 2.1759276469198663e-05, + "loss": 0.0003, + "step": 7030 + }, + { + "epoch": 5.784716516023007, + "grad_norm": 0.05289579163460128, + "learning_rate": 2.161151842667333e-05, + "loss": 0.0003, + "step": 7040 + }, + { + "epoch": 5.792933442892358, + "grad_norm": 0.09570814913112108, + "learning_rate": 2.1464125335474295e-05, + "loss": 0.0004, + "step": 7050 + }, + { + "epoch": 5.801150369761709, + "grad_norm": 0.04169698283933633, + "learning_rate": 2.1317099090430064e-05, + "loss": 0.0003, + "step": 7060 + }, + { + "epoch": 5.80936729663106, + "grad_norm": 0.09601091360074317, + "learning_rate": 2.1170441581653105e-05, + "loss": 0.0004, + "step": 7070 + }, + { + "epoch": 5.817584223500411, + "grad_norm": 0.07839062447329599, + "learning_rate": 2.1024154694515607e-05, + "loss": 0.0003, + "step": 7080 + }, + { + "epoch": 5.825801150369761, + "grad_norm": 0.06065371678537851, + "learning_rate": 2.087824030962509e-05, + "loss": 0.0004, + "step": 7090 + }, + { + "epoch": 5.834018077239112, + "grad_norm": 0.06011971438518955, + "learning_rate": 2.07327003028004e-05, + "loss": 0.0003, + "step": 7100 + }, + { + "epoch": 5.842235004108463, + "grad_norm": 0.023544089324709156, + "learning_rate": 2.0587536545047515e-05, + "loss": 0.0004, + "step": 7110 + }, + { + "epoch": 5.8504519309778145, + "grad_norm": 0.024142141503590467, + "learning_rate": 2.044275090253549e-05, + "loss": 0.0003, + "step": 7120 + }, + { + "epoch": 5.8586688578471655, + "grad_norm": 0.085220089838988, + "learning_rate": 2.0298345236572475e-05, + "loss": 0.0003, + "step": 7130 + }, + { + "epoch": 5.866885784716516, + "grad_norm": 0.06464247146385971, + "learning_rate": 2.015432140358178e-05, + "loss": 0.0003, + "step": 7140 + }, + { + "epoch": 5.875102711585867, + "grad_norm": 0.0361183442372013, + "learning_rate": 2.0010681255078083e-05, + "loss": 0.0003, + "step": 7150 + }, + { + "epoch": 5.883319638455218, + "grad_norm": 0.09666371583074877, + "learning_rate": 1.9867426637643465e-05, + "loss": 0.0003, + "step": 7160 + }, + { + "epoch": 5.891536565324569, + "grad_norm": 0.10791930898412454, + "learning_rate": 1.972455939290383e-05, + "loss": 0.0003, + "step": 7170 + }, + { + "epoch": 5.89975349219392, + "grad_norm": 0.07162077176153998, + "learning_rate": 1.958208135750515e-05, + "loss": 0.0003, + "step": 7180 + }, + { + "epoch": 5.90797041906327, + "grad_norm": 0.09425460191541986, + "learning_rate": 1.9439994363089886e-05, + "loss": 0.0003, + "step": 7190 + }, + { + "epoch": 5.916187345932621, + "grad_norm": 0.06589632396501101, + "learning_rate": 1.9298300236273404e-05, + "loss": 0.0004, + "step": 7200 + }, + { + "epoch": 5.924404272801972, + "grad_norm": 0.010080707203050375, + "learning_rate": 1.9157000798620528e-05, + "loss": 0.0003, + "step": 7210 + }, + { + "epoch": 5.932621199671323, + "grad_norm": 0.01782355670166851, + "learning_rate": 1.9016097866622157e-05, + "loss": 0.0003, + "step": 7220 + }, + { + "epoch": 5.940838126540674, + "grad_norm": 0.08541339041338375, + "learning_rate": 1.8875593251671768e-05, + "loss": 0.0003, + "step": 7230 + }, + { + "epoch": 5.949055053410024, + "grad_norm": 0.0403383903520616, + "learning_rate": 1.873548876004231e-05, + "loss": 0.0003, + "step": 7240 + }, + { + "epoch": 5.957271980279375, + "grad_norm": 0.043433275297285155, + "learning_rate": 1.859578619286285e-05, + "loss": 0.0003, + "step": 7250 + }, + { + "epoch": 5.9654889071487265, + "grad_norm": 0.1289654416690901, + "learning_rate": 1.8456487346095558e-05, + "loss": 0.0003, + "step": 7260 + }, + { + "epoch": 5.9737058340180775, + "grad_norm": 0.042691112085732355, + "learning_rate": 1.83175940105124e-05, + "loss": 0.0003, + "step": 7270 + }, + { + "epoch": 5.981922760887429, + "grad_norm": 0.050366717355632, + "learning_rate": 1.8179107971672334e-05, + "loss": 0.0003, + "step": 7280 + }, + { + "epoch": 5.990139687756779, + "grad_norm": 0.032149964430054835, + "learning_rate": 1.804103100989828e-05, + "loss": 0.0004, + "step": 7290 + }, + { + "epoch": 5.99835661462613, + "grad_norm": 0.03816603372892226, + "learning_rate": 1.7903364900254155e-05, + "loss": 0.0003, + "step": 7300 + }, + { + "epoch": 6.006573541495481, + "grad_norm": 0.012556891088670277, + "learning_rate": 1.7766111412522168e-05, + "loss": 0.0002, + "step": 7310 + }, + { + "epoch": 6.014790468364832, + "grad_norm": 0.024052938509466562, + "learning_rate": 1.762927231117999e-05, + "loss": 0.0002, + "step": 7320 + }, + { + "epoch": 6.023007395234182, + "grad_norm": 0.01613085525687816, + "learning_rate": 1.7492849355378176e-05, + "loss": 0.0001, + "step": 7330 + }, + { + "epoch": 6.031224322103533, + "grad_norm": 0.013358103836048428, + "learning_rate": 1.7356844298917368e-05, + "loss": 0.0001, + "step": 7340 + }, + { + "epoch": 6.039441248972884, + "grad_norm": 0.018304861319451658, + "learning_rate": 1.7221258890225904e-05, + "loss": 0.0001, + "step": 7350 + }, + { + "epoch": 6.047658175842235, + "grad_norm": 0.042929680513244396, + "learning_rate": 1.7086094872337348e-05, + "loss": 0.0001, + "step": 7360 + }, + { + "epoch": 6.055875102711586, + "grad_norm": 0.022654296993854545, + "learning_rate": 1.6951353982867924e-05, + "loss": 0.0001, + "step": 7370 + }, + { + "epoch": 6.064092029580936, + "grad_norm": 0.03052173775190244, + "learning_rate": 1.6817037953994346e-05, + "loss": 0.0001, + "step": 7380 + }, + { + "epoch": 6.0723089564502875, + "grad_norm": 0.012079575561742524, + "learning_rate": 1.6683148512431458e-05, + "loss": 0.0001, + "step": 7390 + }, + { + "epoch": 6.0805258833196385, + "grad_norm": 0.011855108747319118, + "learning_rate": 1.6549687379410116e-05, + "loss": 0.0001, + "step": 7400 + }, + { + "epoch": 6.0887428101889896, + "grad_norm": 0.049849793761952144, + "learning_rate": 1.6416656270654936e-05, + "loss": 0.0002, + "step": 7410 + }, + { + "epoch": 6.096959737058341, + "grad_norm": 0.006559317600966976, + "learning_rate": 1.6284056896362315e-05, + "loss": 0.0001, + "step": 7420 + }, + { + "epoch": 6.105176663927691, + "grad_norm": 0.009016673626101654, + "learning_rate": 1.6151890961178507e-05, + "loss": 0.0001, + "step": 7430 + }, + { + "epoch": 6.113393590797042, + "grad_norm": 0.038050373510506955, + "learning_rate": 1.60201601641776e-05, + "loss": 0.0001, + "step": 7440 + }, + { + "epoch": 6.121610517666393, + "grad_norm": 0.009729819196359553, + "learning_rate": 1.5888866198839673e-05, + "loss": 0.0001, + "step": 7450 + }, + { + "epoch": 6.129827444535744, + "grad_norm": 0.01731626903054496, + "learning_rate": 1.5758010753029107e-05, + "loss": 0.0001, + "step": 7460 + }, + { + "epoch": 6.138044371405094, + "grad_norm": 0.006882906771255923, + "learning_rate": 1.5627595508972888e-05, + "loss": 0.0001, + "step": 7470 + }, + { + "epoch": 6.146261298274445, + "grad_norm": 0.007530518587175814, + "learning_rate": 1.5497622143238837e-05, + "loss": 0.0001, + "step": 7480 + }, + { + "epoch": 6.154478225143796, + "grad_norm": 0.0150356577859745, + "learning_rate": 1.5368092326714232e-05, + "loss": 0.0001, + "step": 7490 + }, + { + "epoch": 6.162695152013147, + "grad_norm": 0.00850621136154103, + "learning_rate": 1.5239007724584259e-05, + "loss": 0.0001, + "step": 7500 + }, + { + "epoch": 6.170912078882498, + "grad_norm": 0.014139621107141917, + "learning_rate": 1.5110369996310575e-05, + "loss": 0.0001, + "step": 7510 + }, + { + "epoch": 6.179129005751848, + "grad_norm": 0.002269181011869001, + "learning_rate": 1.4982180795609979e-05, + "loss": 0.0001, + "step": 7520 + }, + { + "epoch": 6.1873459326211995, + "grad_norm": 0.009622470374378507, + "learning_rate": 1.4854441770433191e-05, + "loss": 0.0001, + "step": 7530 + }, + { + "epoch": 6.1955628594905505, + "grad_norm": 0.0059196089458452335, + "learning_rate": 1.4727154562943702e-05, + "loss": 0.0002, + "step": 7540 + }, + { + "epoch": 6.203779786359902, + "grad_norm": 0.005981743474287708, + "learning_rate": 1.4600320809496516e-05, + "loss": 0.0001, + "step": 7550 + }, + { + "epoch": 6.211996713229253, + "grad_norm": 0.021469388752185552, + "learning_rate": 1.4473942140617253e-05, + "loss": 0.0001, + "step": 7560 + }, + { + "epoch": 6.220213640098603, + "grad_norm": 0.010831060240943258, + "learning_rate": 1.4348020180981181e-05, + "loss": 0.0001, + "step": 7570 + }, + { + "epoch": 6.228430566967954, + "grad_norm": 0.006043329779194204, + "learning_rate": 1.422255654939224e-05, + "loss": 0.0001, + "step": 7580 + }, + { + "epoch": 6.236647493837305, + "grad_norm": 0.019894586606467153, + "learning_rate": 1.4097552858762276e-05, + "loss": 0.0001, + "step": 7590 + }, + { + "epoch": 6.244864420706656, + "grad_norm": 0.004455977975307598, + "learning_rate": 1.3973010716090323e-05, + "loss": 0.0001, + "step": 7600 + }, + { + "epoch": 6.253081347576007, + "grad_norm": 0.011322910304387227, + "learning_rate": 1.3848931722441977e-05, + "loss": 0.0001, + "step": 7610 + }, + { + "epoch": 6.261298274445357, + "grad_norm": 0.016922483519890043, + "learning_rate": 1.3725317472928729e-05, + "loss": 0.0001, + "step": 7620 + }, + { + "epoch": 6.269515201314708, + "grad_norm": 0.03652252021560336, + "learning_rate": 1.3602169556687455e-05, + "loss": 0.0001, + "step": 7630 + }, + { + "epoch": 6.277732128184059, + "grad_norm": 0.011138599633904338, + "learning_rate": 1.3479489556860136e-05, + "loss": 0.0001, + "step": 7640 + }, + { + "epoch": 6.28594905505341, + "grad_norm": 0.0287764852733891, + "learning_rate": 1.335727905057334e-05, + "loss": 0.0001, + "step": 7650 + }, + { + "epoch": 6.294165981922761, + "grad_norm": 0.00885819329873813, + "learning_rate": 1.323553960891798e-05, + "loss": 0.0001, + "step": 7660 + }, + { + "epoch": 6.3023829087921115, + "grad_norm": 0.0028294177943224753, + "learning_rate": 1.3114272796929177e-05, + "loss": 0.0001, + "step": 7670 + }, + { + "epoch": 6.3105998356614625, + "grad_norm": 0.027477569713298226, + "learning_rate": 1.2993480173566142e-05, + "loss": 0.0001, + "step": 7680 + }, + { + "epoch": 6.318816762530814, + "grad_norm": 0.007490571813813741, + "learning_rate": 1.2873163291692065e-05, + "loss": 0.0001, + "step": 7690 + }, + { + "epoch": 6.327033689400165, + "grad_norm": 0.024335485165704607, + "learning_rate": 1.2753323698054138e-05, + "loss": 0.0001, + "step": 7700 + }, + { + "epoch": 6.335250616269515, + "grad_norm": 0.004371079746914589, + "learning_rate": 1.2633962933263826e-05, + "loss": 0.0001, + "step": 7710 + }, + { + "epoch": 6.343467543138866, + "grad_norm": 0.006251537497769062, + "learning_rate": 1.25150825317769e-05, + "loss": 0.0001, + "step": 7720 + }, + { + "epoch": 6.351684470008217, + "grad_norm": 0.020331898183790224, + "learning_rate": 1.2396684021873734e-05, + "loss": 0.0001, + "step": 7730 + }, + { + "epoch": 6.359901396877568, + "grad_norm": 0.004655603593615033, + "learning_rate": 1.2278768925639722e-05, + "loss": 0.0001, + "step": 7740 + }, + { + "epoch": 6.368118323746919, + "grad_norm": 0.0031523431906211427, + "learning_rate": 1.2161338758945717e-05, + "loss": 0.0001, + "step": 7750 + }, + { + "epoch": 6.376335250616269, + "grad_norm": 0.004989701336351937, + "learning_rate": 1.2044395031428469e-05, + "loss": 0.0001, + "step": 7760 + }, + { + "epoch": 6.38455217748562, + "grad_norm": 0.01789297524523158, + "learning_rate": 1.1927939246471209e-05, + "loss": 0.0001, + "step": 7770 + }, + { + "epoch": 6.392769104354971, + "grad_norm": 0.006979472930007758, + "learning_rate": 1.181197290118446e-05, + "loss": 0.0001, + "step": 7780 + }, + { + "epoch": 6.400986031224322, + "grad_norm": 0.01068601094717032, + "learning_rate": 1.1696497486386631e-05, + "loss": 0.0001, + "step": 7790 + }, + { + "epoch": 6.409202958093673, + "grad_norm": 0.005693654595461927, + "learning_rate": 1.158151448658496e-05, + "loss": 0.0001, + "step": 7800 + }, + { + "epoch": 6.4174198849630235, + "grad_norm": 0.0034102916125843753, + "learning_rate": 1.146702537995632e-05, + "loss": 0.0001, + "step": 7810 + }, + { + "epoch": 6.425636811832375, + "grad_norm": 0.004696017573677494, + "learning_rate": 1.1353031638328377e-05, + "loss": 0.0001, + "step": 7820 + }, + { + "epoch": 6.433853738701726, + "grad_norm": 0.005538590125057658, + "learning_rate": 1.1239534727160534e-05, + "loss": 0.0001, + "step": 7830 + }, + { + "epoch": 6.442070665571077, + "grad_norm": 0.0037981161684288198, + "learning_rate": 1.1126536105525092e-05, + "loss": 0.0001, + "step": 7840 + }, + { + "epoch": 6.450287592440427, + "grad_norm": 0.008130546163056942, + "learning_rate": 1.1014037226088609e-05, + "loss": 0.0001, + "step": 7850 + }, + { + "epoch": 6.458504519309778, + "grad_norm": 0.00507003078713952, + "learning_rate": 1.0902039535093107e-05, + "loss": 0.0001, + "step": 7860 + }, + { + "epoch": 6.466721446179129, + "grad_norm": 0.005303142163377792, + "learning_rate": 1.0790544472337543e-05, + "loss": 0.0001, + "step": 7870 + }, + { + "epoch": 6.47493837304848, + "grad_norm": 0.007789976120229994, + "learning_rate": 1.0679553471159238e-05, + "loss": 0.0001, + "step": 7880 + }, + { + "epoch": 6.483155299917831, + "grad_norm": 0.004226289143148964, + "learning_rate": 1.0569067958415552e-05, + "loss": 0.0001, + "step": 7890 + }, + { + "epoch": 6.491372226787181, + "grad_norm": 0.005649609413818498, + "learning_rate": 1.0459089354465463e-05, + "loss": 0.0001, + "step": 7900 + }, + { + "epoch": 6.499589153656532, + "grad_norm": 0.005179102469730814, + "learning_rate": 1.0349619073151273e-05, + "loss": 0.0001, + "step": 7910 + }, + { + "epoch": 6.507806080525883, + "grad_norm": 0.019929860999792926, + "learning_rate": 1.0240658521780571e-05, + "loss": 0.0001, + "step": 7920 + }, + { + "epoch": 6.516023007395234, + "grad_norm": 0.005251184283310909, + "learning_rate": 1.0132209101107998e-05, + "loss": 0.0001, + "step": 7930 + }, + { + "epoch": 6.524239934264585, + "grad_norm": 0.003955322164587, + "learning_rate": 1.0024272205317325e-05, + "loss": 0.0001, + "step": 7940 + }, + { + "epoch": 6.5324568611339355, + "grad_norm": 0.006318495729659461, + "learning_rate": 9.91684922200345e-06, + "loss": 0.0001, + "step": 7950 + }, + { + "epoch": 6.540673788003287, + "grad_norm": 0.0030449070999937953, + "learning_rate": 9.809941532154688e-06, + "loss": 0.0001, + "step": 7960 + }, + { + "epoch": 6.548890714872638, + "grad_norm": 0.032291114822553174, + "learning_rate": 9.703550510134884e-06, + "loss": 0.0001, + "step": 7970 + }, + { + "epoch": 6.557107641741989, + "grad_norm": 0.007393779957982821, + "learning_rate": 9.597677523665828e-06, + "loss": 0.0001, + "step": 7980 + }, + { + "epoch": 6.56532456861134, + "grad_norm": 0.01310408703035522, + "learning_rate": 9.492323933809626e-06, + "loss": 0.0001, + "step": 7990 + }, + { + "epoch": 6.57354149548069, + "grad_norm": 0.00364474267810044, + "learning_rate": 9.387491094951245e-06, + "loss": 0.0001, + "step": 8000 + }, + { + "epoch": 6.581758422350041, + "grad_norm": 0.07013764371458263, + "learning_rate": 9.28318035478108e-06, + "loss": 0.0001, + "step": 8010 + }, + { + "epoch": 6.589975349219392, + "grad_norm": 0.0042762894497602455, + "learning_rate": 9.179393054277563e-06, + "loss": 0.0001, + "step": 8020 + }, + { + "epoch": 6.598192276088743, + "grad_norm": 0.006822177460073707, + "learning_rate": 9.076130527690091e-06, + "loss": 0.0001, + "step": 8030 + }, + { + "epoch": 6.606409202958094, + "grad_norm": 0.04247228512308365, + "learning_rate": 8.973394102521703e-06, + "loss": 0.0001, + "step": 8040 + }, + { + "epoch": 6.614626129827444, + "grad_norm": 0.009417095229872169, + "learning_rate": 8.871185099512102e-06, + "loss": 0.0001, + "step": 8050 + }, + { + "epoch": 6.622843056696795, + "grad_norm": 0.00607486004705313, + "learning_rate": 8.769504832620645e-06, + "loss": 0.0001, + "step": 8060 + }, + { + "epoch": 6.631059983566146, + "grad_norm": 0.008230402113951254, + "learning_rate": 8.668354609009471e-06, + "loss": 0.0001, + "step": 8070 + }, + { + "epoch": 6.639276910435497, + "grad_norm": 0.003767907099494232, + "learning_rate": 8.567735729026704e-06, + "loss": 0.0001, + "step": 8080 + }, + { + "epoch": 6.647493837304848, + "grad_norm": 0.0060625025220406165, + "learning_rate": 8.467649486189665e-06, + "loss": 0.0001, + "step": 8090 + }, + { + "epoch": 6.655710764174199, + "grad_norm": 0.01166742655469823, + "learning_rate": 8.36809716716837e-06, + "loss": 0.0001, + "step": 8100 + }, + { + "epoch": 6.66392769104355, + "grad_norm": 0.009057169894164304, + "learning_rate": 8.269080051768874e-06, + "loss": 0.0001, + "step": 8110 + }, + { + "epoch": 6.672144617912901, + "grad_norm": 0.007341837378722351, + "learning_rate": 8.170599412916863e-06, + "loss": 0.0001, + "step": 8120 + }, + { + "epoch": 6.680361544782252, + "grad_norm": 0.007014560655959058, + "learning_rate": 8.072656516641303e-06, + "loss": 0.0001, + "step": 8130 + }, + { + "epoch": 6.688578471651603, + "grad_norm": 0.0060690230284784295, + "learning_rate": 7.975252622058139e-06, + "loss": 0.0001, + "step": 8140 + }, + { + "epoch": 6.696795398520953, + "grad_norm": 0.0045773532117964, + "learning_rate": 7.878388981354124e-06, + "loss": 0.0001, + "step": 8150 + }, + { + "epoch": 6.705012325390304, + "grad_norm": 0.011011679346444675, + "learning_rate": 7.782066839770714e-06, + "loss": 0.0001, + "step": 8160 + }, + { + "epoch": 6.713229252259655, + "grad_norm": 0.0030627824700418574, + "learning_rate": 7.686287435588063e-06, + "loss": 0.0001, + "step": 8170 + }, + { + "epoch": 6.721446179129006, + "grad_norm": 0.019096815604853528, + "learning_rate": 7.5910520001091e-06, + "loss": 0.0001, + "step": 8180 + }, + { + "epoch": 6.729663105998356, + "grad_norm": 0.0024027337690375786, + "learning_rate": 7.49636175764371e-06, + "loss": 0.0001, + "step": 8190 + }, + { + "epoch": 6.737880032867707, + "grad_norm": 0.011509490405769068, + "learning_rate": 7.402217925492982e-06, + "loss": 0.0001, + "step": 8200 + }, + { + "epoch": 6.746096959737058, + "grad_norm": 0.017331511927844077, + "learning_rate": 7.308621713933567e-06, + "loss": 0.0001, + "step": 8210 + }, + { + "epoch": 6.754313886606409, + "grad_norm": 0.0030539489717396626, + "learning_rate": 7.215574326202118e-06, + "loss": 0.0001, + "step": 8220 + }, + { + "epoch": 6.76253081347576, + "grad_norm": 0.008162144251986641, + "learning_rate": 7.123076958479819e-06, + "loss": 0.0001, + "step": 8230 + }, + { + "epoch": 6.770747740345111, + "grad_norm": 0.01117152027177566, + "learning_rate": 7.031130799877012e-06, + "loss": 0.0001, + "step": 8240 + }, + { + "epoch": 6.778964667214462, + "grad_norm": 0.0022289041007739147, + "learning_rate": 6.939737032417909e-06, + "loss": 0.0001, + "step": 8250 + }, + { + "epoch": 6.787181594083813, + "grad_norm": 0.00250778837799371, + "learning_rate": 6.848896831025397e-06, + "loss": 0.0001, + "step": 8260 + }, + { + "epoch": 6.795398520953164, + "grad_norm": 0.010239617825252774, + "learning_rate": 6.758611363505923e-06, + "loss": 0.0001, + "step": 8270 + }, + { + "epoch": 6.803615447822514, + "grad_norm": 0.004908005944939691, + "learning_rate": 6.668881790534503e-06, + "loss": 0.0001, + "step": 8280 + }, + { + "epoch": 6.811832374691865, + "grad_norm": 0.009545846548302477, + "learning_rate": 6.5797092656397775e-06, + "loss": 0.0001, + "step": 8290 + }, + { + "epoch": 6.820049301561216, + "grad_norm": 0.035864624123359286, + "learning_rate": 6.491094935189201e-06, + "loss": 0.0001, + "step": 8300 + }, + { + "epoch": 6.828266228430567, + "grad_norm": 0.004237835490338293, + "learning_rate": 6.403039938374284e-06, + "loss": 0.0001, + "step": 8310 + }, + { + "epoch": 6.836483155299918, + "grad_norm": 0.004823843556841174, + "learning_rate": 6.315545407195972e-06, + "loss": 0.0001, + "step": 8320 + }, + { + "epoch": 6.844700082169268, + "grad_norm": 0.0057465560303971925, + "learning_rate": 6.228612466450085e-06, + "loss": 0.0001, + "step": 8330 + }, + { + "epoch": 6.852917009038619, + "grad_norm": 0.0021839272209857523, + "learning_rate": 6.14224223371283e-06, + "loss": 0.0001, + "step": 8340 + }, + { + "epoch": 6.86113393590797, + "grad_norm": 0.002802474934316734, + "learning_rate": 6.0564358193264894e-06, + "loss": 0.0001, + "step": 8350 + }, + { + "epoch": 6.869350862777321, + "grad_norm": 0.003229555203702007, + "learning_rate": 5.971194326385093e-06, + "loss": 0.0001, + "step": 8360 + }, + { + "epoch": 6.8775677896466725, + "grad_norm": 0.0070009328714542235, + "learning_rate": 5.886518850720268e-06, + "loss": 0.0001, + "step": 8370 + }, + { + "epoch": 6.885784716516023, + "grad_norm": 0.0160308382727869, + "learning_rate": 5.802410480887138e-06, + "loss": 0.0001, + "step": 8380 + }, + { + "epoch": 6.894001643385374, + "grad_norm": 0.005359781228827345, + "learning_rate": 5.718870298150342e-06, + "loss": 0.0001, + "step": 8390 + }, + { + "epoch": 6.902218570254725, + "grad_norm": 0.005181023653119385, + "learning_rate": 5.635899376470116e-06, + "loss": 0.0001, + "step": 8400 + }, + { + "epoch": 6.910435497124076, + "grad_norm": 0.032974471680409116, + "learning_rate": 5.553498782488498e-06, + "loss": 0.0001, + "step": 8410 + }, + { + "epoch": 6.918652423993427, + "grad_norm": 0.07888453711094053, + "learning_rate": 5.471669575515625e-06, + "loss": 0.0001, + "step": 8420 + }, + { + "epoch": 6.926869350862777, + "grad_norm": 0.006489359356087647, + "learning_rate": 5.3904128075160875e-06, + "loss": 0.0001, + "step": 8430 + }, + { + "epoch": 6.935086277732128, + "grad_norm": 0.0033782347966967757, + "learning_rate": 5.30972952309543e-06, + "loss": 0.0001, + "step": 8440 + }, + { + "epoch": 6.943303204601479, + "grad_norm": 0.013790744321467846, + "learning_rate": 5.22962075948672e-06, + "loss": 0.0001, + "step": 8450 + }, + { + "epoch": 6.95152013147083, + "grad_norm": 0.0103862585210028, + "learning_rate": 5.150087546537202e-06, + "loss": 0.0001, + "step": 8460 + }, + { + "epoch": 6.959737058340181, + "grad_norm": 0.005527170373168249, + "learning_rate": 5.071130906695054e-06, + "loss": 0.0001, + "step": 8470 + }, + { + "epoch": 6.967953985209531, + "grad_norm": 0.008299029914842444, + "learning_rate": 4.992751854996275e-06, + "loss": 0.0001, + "step": 8480 + }, + { + "epoch": 6.976170912078882, + "grad_norm": 0.003940234868927148, + "learning_rate": 4.914951399051604e-06, + "loss": 0.0001, + "step": 8490 + }, + { + "epoch": 6.9843878389482335, + "grad_norm": 0.009562559751631306, + "learning_rate": 4.837730539033564e-06, + "loss": 0.0001, + "step": 8500 + }, + { + "epoch": 6.9926047658175845, + "grad_norm": 0.005061849595818301, + "learning_rate": 4.761090267663626e-06, + "loss": 0.0001, + "step": 8510 + }, + { + "epoch": 7.000821692686935, + "grad_norm": 0.0037394560308653935, + "learning_rate": 4.685031570199477e-06, + "loss": 0.0001, + "step": 8520 + }, + { + "epoch": 7.009038619556286, + "grad_norm": 0.0024890741202245347, + "learning_rate": 4.609555424422241e-06, + "loss": 0.0001, + "step": 8530 + }, + { + "epoch": 7.017255546425637, + "grad_norm": 0.002347883326211967, + "learning_rate": 4.534662800624034e-06, + "loss": 0.0001, + "step": 8540 + }, + { + "epoch": 7.025472473294988, + "grad_norm": 0.0029431823023118807, + "learning_rate": 4.460354661595434e-06, + "loss": 0.0001, + "step": 8550 + }, + { + "epoch": 7.033689400164339, + "grad_norm": 0.0022166565152147576, + "learning_rate": 4.386631962613091e-06, + "loss": 0.0001, + "step": 8560 + }, + { + "epoch": 7.041906327033689, + "grad_norm": 0.0028540031656121905, + "learning_rate": 4.313495651427485e-06, + "loss": 0.0001, + "step": 8570 + }, + { + "epoch": 7.05012325390304, + "grad_norm": 0.0018199061207818505, + "learning_rate": 4.2409466682506944e-06, + "loss": 0.0001, + "step": 8580 + }, + { + "epoch": 7.058340180772391, + "grad_norm": 0.006676679370248783, + "learning_rate": 4.168985945744392e-06, + "loss": 0.0001, + "step": 8590 + }, + { + "epoch": 7.066557107641742, + "grad_norm": 0.0019206661894760565, + "learning_rate": 4.097614409007722e-06, + "loss": 0.0001, + "step": 8600 + }, + { + "epoch": 7.074774034511093, + "grad_norm": 0.0019643797562854657, + "learning_rate": 4.026832975565542e-06, + "loss": 0.0001, + "step": 8610 + }, + { + "epoch": 7.082990961380443, + "grad_norm": 0.002183124484379041, + "learning_rate": 3.956642555356549e-06, + "loss": 0.0001, + "step": 8620 + }, + { + "epoch": 7.091207888249794, + "grad_norm": 0.0022118342028955656, + "learning_rate": 3.887044050721606e-06, + "loss": 0.0001, + "step": 8630 + }, + { + "epoch": 7.0994248151191455, + "grad_norm": 0.00217550665263972, + "learning_rate": 3.818038356392134e-06, + "loss": 0.0001, + "step": 8640 + }, + { + "epoch": 7.1076417419884965, + "grad_norm": 0.0026073824295988037, + "learning_rate": 3.749626359478614e-06, + "loss": 0.0001, + "step": 8650 + }, + { + "epoch": 7.115858668857848, + "grad_norm": 0.0017507130857040154, + "learning_rate": 3.681808939459214e-06, + "loss": 0.0001, + "step": 8660 + }, + { + "epoch": 7.124075595727198, + "grad_norm": 0.0041525426460312055, + "learning_rate": 3.614586968168393e-06, + "loss": 0.0001, + "step": 8670 + }, + { + "epoch": 7.132292522596549, + "grad_norm": 0.002211377850655834, + "learning_rate": 3.547961309785802e-06, + "loss": 0.0001, + "step": 8680 + }, + { + "epoch": 7.1405094494659, + "grad_norm": 0.002389269395151305, + "learning_rate": 3.4819328208250903e-06, + "loss": 0.0001, + "step": 8690 + }, + { + "epoch": 7.148726376335251, + "grad_norm": 0.004046444098601744, + "learning_rate": 3.4165023501229685e-06, + "loss": 0.0001, + "step": 8700 + }, + { + "epoch": 7.156943303204601, + "grad_norm": 0.0025687362955891446, + "learning_rate": 3.351670738828211e-06, + "loss": 0.0001, + "step": 8710 + }, + { + "epoch": 7.165160230073952, + "grad_norm": 0.0018023229176762012, + "learning_rate": 3.2874388203909034e-06, + "loss": 0.0001, + "step": 8720 + }, + { + "epoch": 7.173377156943303, + "grad_norm": 0.0024051048954698436, + "learning_rate": 3.2238074205517443e-06, + "loss": 0.0001, + "step": 8730 + }, + { + "epoch": 7.181594083812654, + "grad_norm": 0.0027352114996948835, + "learning_rate": 3.160777357331346e-06, + "loss": 0.0001, + "step": 8740 + }, + { + "epoch": 7.189811010682005, + "grad_norm": 0.002275545041481053, + "learning_rate": 3.098349441019804e-06, + "loss": 0.0001, + "step": 8750 + }, + { + "epoch": 7.198027937551355, + "grad_norm": 0.0021983931153620206, + "learning_rate": 3.036524474166225e-06, + "loss": 0.0001, + "step": 8760 + }, + { + "epoch": 7.206244864420706, + "grad_norm": 0.003129644806150804, + "learning_rate": 2.975303251568462e-06, + "loss": 0.0001, + "step": 8770 + }, + { + "epoch": 7.2144617912900575, + "grad_norm": 0.002546066057537917, + "learning_rate": 2.9146865602628358e-06, + "loss": 0.0001, + "step": 8780 + }, + { + "epoch": 7.2226787181594085, + "grad_norm": 0.002270070692451648, + "learning_rate": 2.8546751795140438e-06, + "loss": 0.0001, + "step": 8790 + }, + { + "epoch": 7.23089564502876, + "grad_norm": 0.0021887084921120284, + "learning_rate": 2.79526988080519e-06, + "loss": 0.0001, + "step": 8800 + }, + { + "epoch": 7.23911257189811, + "grad_norm": 0.006515329381722771, + "learning_rate": 2.7364714278277636e-06, + "loss": 0.0001, + "step": 8810 + }, + { + "epoch": 7.247329498767461, + "grad_norm": 0.002958401842693592, + "learning_rate": 2.6782805764719254e-06, + "loss": 0.0001, + "step": 8820 + }, + { + "epoch": 7.255546425636812, + "grad_norm": 0.001974263159053724, + "learning_rate": 2.6206980748167076e-06, + "loss": 0.0001, + "step": 8830 + }, + { + "epoch": 7.263763352506163, + "grad_norm": 0.002894554015802063, + "learning_rate": 2.563724663120476e-06, + "loss": 0.0001, + "step": 8840 + }, + { + "epoch": 7.271980279375514, + "grad_norm": 0.0019962141286986794, + "learning_rate": 2.5073610738113373e-06, + "loss": 0.0001, + "step": 8850 + }, + { + "epoch": 7.280197206244864, + "grad_norm": 0.001968009882171414, + "learning_rate": 2.4516080314777647e-06, + "loss": 0.0001, + "step": 8860 + }, + { + "epoch": 7.288414133114215, + "grad_norm": 0.002101796143810661, + "learning_rate": 2.3964662528592974e-06, + "loss": 0.0001, + "step": 8870 + }, + { + "epoch": 7.296631059983566, + "grad_norm": 0.002367055304908569, + "learning_rate": 2.3419364468372894e-06, + "loss": 0.0001, + "step": 8880 + }, + { + "epoch": 7.304847986852917, + "grad_norm": 0.0035310046146143415, + "learning_rate": 2.2880193144257932e-06, + "loss": 0.0001, + "step": 8890 + }, + { + "epoch": 7.313064913722268, + "grad_norm": 0.0029086365245345057, + "learning_rate": 2.2347155487625893e-06, + "loss": 0.0001, + "step": 8900 + }, + { + "epoch": 7.3212818405916185, + "grad_norm": 0.0022606852974026313, + "learning_rate": 2.1820258351002652e-06, + "loss": 0.0001, + "step": 8910 + }, + { + "epoch": 7.3294987674609695, + "grad_norm": 0.003313569613573218, + "learning_rate": 2.1299508507973576e-06, + "loss": 0.0001, + "step": 8920 + }, + { + "epoch": 7.337715694330321, + "grad_norm": 0.0020835003336243446, + "learning_rate": 2.078491265309701e-06, + "loss": 0.0001, + "step": 8930 + }, + { + "epoch": 7.345932621199672, + "grad_norm": 0.003182988577461455, + "learning_rate": 2.027647740181815e-06, + "loss": 0.0001, + "step": 8940 + }, + { + "epoch": 7.354149548069023, + "grad_norm": 0.0018272688094744427, + "learning_rate": 1.977420929038365e-06, + "loss": 0.0001, + "step": 8950 + }, + { + "epoch": 7.362366474938373, + "grad_norm": 0.0029463625824918127, + "learning_rate": 1.9278114775757804e-06, + "loss": 0.0001, + "step": 8960 + }, + { + "epoch": 7.370583401807724, + "grad_norm": 0.00262210222653249, + "learning_rate": 1.878820023553951e-06, + "loss": 0.0001, + "step": 8970 + }, + { + "epoch": 7.378800328677075, + "grad_norm": 0.0027294967523430838, + "learning_rate": 1.8304471967880655e-06, + "loss": 0.0001, + "step": 8980 + }, + { + "epoch": 7.387017255546426, + "grad_norm": 0.0027744334985216556, + "learning_rate": 1.7826936191404297e-06, + "loss": 0.0001, + "step": 8990 + }, + { + "epoch": 7.395234182415776, + "grad_norm": 0.002027432836013355, + "learning_rate": 1.735559904512546e-06, + "loss": 0.0001, + "step": 9000 + }, + { + "epoch": 7.403451109285127, + "grad_norm": 0.0023366777423095937, + "learning_rate": 1.6890466588372178e-06, + "loss": 0.0001, + "step": 9010 + }, + { + "epoch": 7.411668036154478, + "grad_norm": 0.003959171212239567, + "learning_rate": 1.6431544800707078e-06, + "loss": 0.0001, + "step": 9020 + }, + { + "epoch": 7.419884963023829, + "grad_norm": 0.003038517316526668, + "learning_rate": 1.5978839581850924e-06, + "loss": 0.0001, + "step": 9030 + }, + { + "epoch": 7.42810188989318, + "grad_norm": 0.0030635364753688293, + "learning_rate": 1.5532356751606692e-06, + "loss": 0.0001, + "step": 9040 + }, + { + "epoch": 7.4363188167625305, + "grad_norm": 0.004956427570631242, + "learning_rate": 1.509210204978495e-06, + "loss": 0.0001, + "step": 9050 + }, + { + "epoch": 7.4445357436318815, + "grad_norm": 0.0017184708212546133, + "learning_rate": 1.4658081136129432e-06, + "loss": 0.0001, + "step": 9060 + }, + { + "epoch": 7.452752670501233, + "grad_norm": 0.0023069937851967893, + "learning_rate": 1.4230299590245022e-06, + "loss": 0.0001, + "step": 9070 + }, + { + "epoch": 7.460969597370584, + "grad_norm": 0.002102162255454297, + "learning_rate": 1.3808762911525664e-06, + "loss": 0.0001, + "step": 9080 + }, + { + "epoch": 7.469186524239935, + "grad_norm": 0.002215480292371805, + "learning_rate": 1.3393476519083736e-06, + "loss": 0.0001, + "step": 9090 + }, + { + "epoch": 7.477403451109285, + "grad_norm": 0.0036056194163447703, + "learning_rate": 1.298444575168012e-06, + "loss": 0.0001, + "step": 9100 + }, + { + "epoch": 7.485620377978636, + "grad_norm": 0.00285526450732319, + "learning_rate": 1.2581675867655963e-06, + "loss": 0.0001, + "step": 9110 + }, + { + "epoch": 7.493837304847987, + "grad_norm": 0.001955640998053969, + "learning_rate": 1.2185172044865135e-06, + "loss": 0.0001, + "step": 9120 + }, + { + "epoch": 7.502054231717338, + "grad_norm": 0.003042764528202857, + "learning_rate": 1.1794939380607106e-06, + "loss": 0.0001, + "step": 9130 + }, + { + "epoch": 7.510271158586688, + "grad_norm": 0.0030643896288807945, + "learning_rate": 1.1410982891561895e-06, + "loss": 0.0001, + "step": 9140 + }, + { + "epoch": 7.518488085456039, + "grad_norm": 0.003430371455732615, + "learning_rate": 1.1033307513725443e-06, + "loss": 0.0001, + "step": 9150 + }, + { + "epoch": 7.52670501232539, + "grad_norm": 0.002394733913662495, + "learning_rate": 1.0661918102346292e-06, + "loss": 0.0001, + "step": 9160 + }, + { + "epoch": 7.534921939194741, + "grad_norm": 0.002543880393909049, + "learning_rate": 1.0296819431862781e-06, + "loss": 0.0001, + "step": 9170 + }, + { + "epoch": 7.543138866064092, + "grad_norm": 0.0018053156690311556, + "learning_rate": 9.938016195841948e-07, + "loss": 0.0001, + "step": 9180 + }, + { + "epoch": 7.5513557929334425, + "grad_norm": 0.0026998336135635732, + "learning_rate": 9.585513006919456e-07, + "loss": 0.0001, + "step": 9190 + }, + { + "epoch": 7.5595727198027936, + "grad_norm": 0.0024286503863416523, + "learning_rate": 9.239314396739751e-07, + "loss": 0.0001, + "step": 9200 + }, + { + "epoch": 7.567789646672145, + "grad_norm": 0.002072401573430573, + "learning_rate": 8.899424815898006e-07, + "loss": 0.0001, + "step": 9210 + }, + { + "epoch": 7.576006573541496, + "grad_norm": 0.0017131123783006232, + "learning_rate": 8.565848633883156e-07, + "loss": 0.0001, + "step": 9220 + }, + { + "epoch": 7.584223500410847, + "grad_norm": 0.0019265471313105727, + "learning_rate": 8.238590139021396e-07, + "loss": 0.0001, + "step": 9230 + }, + { + "epoch": 7.592440427280197, + "grad_norm": 0.0030467991994696376, + "learning_rate": 7.917653538421221e-07, + "loss": 0.0001, + "step": 9240 + }, + { + "epoch": 7.600657354149548, + "grad_norm": 0.0027561941889771417, + "learning_rate": 7.603042957919248e-07, + "loss": 0.0001, + "step": 9250 + }, + { + "epoch": 7.608874281018899, + "grad_norm": 0.0023882200786924904, + "learning_rate": 7.294762442027314e-07, + "loss": 0.0001, + "step": 9260 + }, + { + "epoch": 7.61709120788825, + "grad_norm": 0.0023159621569301615, + "learning_rate": 6.992815953880405e-07, + "loss": 0.0001, + "step": 9270 + }, + { + "epoch": 7.625308134757601, + "grad_norm": 0.0019927286356167125, + "learning_rate": 6.697207375185532e-07, + "loss": 0.0001, + "step": 9280 + }, + { + "epoch": 7.633525061626951, + "grad_norm": 0.0022629249257758942, + "learning_rate": 6.407940506172217e-07, + "loss": 0.0001, + "step": 9290 + }, + { + "epoch": 7.641741988496302, + "grad_norm": 0.045011492887964105, + "learning_rate": 6.125019065543302e-07, + "loss": 0.0001, + "step": 9300 + }, + { + "epoch": 7.649958915365653, + "grad_norm": 0.002171533068036551, + "learning_rate": 5.848446690427223e-07, + "loss": 0.0001, + "step": 9310 + }, + { + "epoch": 7.658175842235004, + "grad_norm": 0.0031192747437895157, + "learning_rate": 5.578226936331365e-07, + "loss": 0.0001, + "step": 9320 + }, + { + "epoch": 7.666392769104355, + "grad_norm": 0.002314532015307654, + "learning_rate": 5.314363277096224e-07, + "loss": 0.0001, + "step": 9330 + }, + { + "epoch": 7.674609695973706, + "grad_norm": 0.001915994627453927, + "learning_rate": 5.05685910485093e-07, + "loss": 0.0001, + "step": 9340 + }, + { + "epoch": 7.682826622843057, + "grad_norm": 0.0026146749674093666, + "learning_rate": 4.805717729969295e-07, + "loss": 0.0001, + "step": 9350 + }, + { + "epoch": 7.691043549712408, + "grad_norm": 0.0023079795524893498, + "learning_rate": 4.560942381027611e-07, + "loss": 0.0001, + "step": 9360 + }, + { + "epoch": 7.699260476581759, + "grad_norm": 0.0016097946386047883, + "learning_rate": 4.322536204763028e-07, + "loss": 0.0001, + "step": 9370 + }, + { + "epoch": 7.70747740345111, + "grad_norm": 0.002370701734324499, + "learning_rate": 4.0905022660329694e-07, + "loss": 0.0001, + "step": 9380 + }, + { + "epoch": 7.71569433032046, + "grad_norm": 0.0019865542526048587, + "learning_rate": 3.8648435477759427e-07, + "loss": 0.0001, + "step": 9390 + }, + { + "epoch": 7.723911257189811, + "grad_norm": 0.002258789882750112, + "learning_rate": 3.6455629509730136e-07, + "loss": 0.0001, + "step": 9400 + }, + { + "epoch": 7.732128184059162, + "grad_norm": 0.002634160546490344, + "learning_rate": 3.432663294610672e-07, + "loss": 0.0001, + "step": 9410 + }, + { + "epoch": 7.740345110928513, + "grad_norm": 0.0021613881528480656, + "learning_rate": 3.226147315644412e-07, + "loss": 0.0001, + "step": 9420 + }, + { + "epoch": 7.748562037797863, + "grad_norm": 0.0023601149375290773, + "learning_rate": 3.02601766896371e-07, + "loss": 0.0001, + "step": 9430 + }, + { + "epoch": 7.756778964667214, + "grad_norm": 0.0020425588675221165, + "learning_rate": 2.8322769273578244e-07, + "loss": 0.0001, + "step": 9440 + }, + { + "epoch": 7.764995891536565, + "grad_norm": 0.0025293237557369656, + "learning_rate": 2.644927581482659e-07, + "loss": 0.0001, + "step": 9450 + }, + { + "epoch": 7.773212818405916, + "grad_norm": 0.0030372248111876286, + "learning_rate": 2.463972039828899e-07, + "loss": 0.0001, + "step": 9460 + }, + { + "epoch": 7.781429745275267, + "grad_norm": 0.0026193753684198287, + "learning_rate": 2.2894126286909234e-07, + "loss": 0.0001, + "step": 9470 + }, + { + "epoch": 7.789646672144618, + "grad_norm": 0.0022108618026861868, + "learning_rate": 2.121251592136886e-07, + "loss": 0.0001, + "step": 9480 + }, + { + "epoch": 7.797863599013969, + "grad_norm": 0.0021156549924296795, + "learning_rate": 1.9594910919800147e-07, + "loss": 0.0001, + "step": 9490 + }, + { + "epoch": 7.80608052588332, + "grad_norm": 0.002628227826991199, + "learning_rate": 1.8041332077506356e-07, + "loss": 0.0001, + "step": 9500 + }, + { + "epoch": 7.814297452752671, + "grad_norm": 0.0022789543379677394, + "learning_rate": 1.6551799366695265e-07, + "loss": 0.0001, + "step": 9510 + }, + { + "epoch": 7.822514379622021, + "grad_norm": 0.002475135547991725, + "learning_rate": 1.5126331936222704e-07, + "loss": 0.0001, + "step": 9520 + }, + { + "epoch": 7.830731306491372, + "grad_norm": 0.002241320474959172, + "learning_rate": 1.3764948111346098e-07, + "loss": 0.0001, + "step": 9530 + }, + { + "epoch": 7.838948233360723, + "grad_norm": 0.001586339539312268, + "learning_rate": 1.2467665393488536e-07, + "loss": 0.0001, + "step": 9540 + }, + { + "epoch": 7.847165160230074, + "grad_norm": 0.0019988455186237875, + "learning_rate": 1.1234500460014507e-07, + "loss": 0.0001, + "step": 9550 + }, + { + "epoch": 7.855382087099425, + "grad_norm": 0.0026678038133249762, + "learning_rate": 1.0065469164015074e-07, + "loss": 0.0001, + "step": 9560 + }, + { + "epoch": 7.863599013968775, + "grad_norm": 0.0032158810061470836, + "learning_rate": 8.960586534103587e-08, + "loss": 0.0001, + "step": 9570 + }, + { + "epoch": 7.871815940838126, + "grad_norm": 0.0022962158163646286, + "learning_rate": 7.919866774223073e-08, + "loss": 0.0001, + "step": 9580 + }, + { + "epoch": 7.880032867707477, + "grad_norm": 0.0019382682133878015, + "learning_rate": 6.943323263464141e-08, + "loss": 0.0001, + "step": 9590 + }, + { + "epoch": 7.888249794576828, + "grad_norm": 0.0027439318214669307, + "learning_rate": 6.030968555892358e-08, + "loss": 0.0001, + "step": 9600 + }, + { + "epoch": 7.8964667214461794, + "grad_norm": 0.00216562160133885, + "learning_rate": 5.182814380385592e-08, + "loss": 0.0001, + "step": 9610 + }, + { + "epoch": 7.90468364831553, + "grad_norm": 0.002315933879589204, + "learning_rate": 4.398871640486357e-08, + "loss": 0.0001, + "step": 9620 + }, + { + "epoch": 7.912900575184881, + "grad_norm": 0.0019623120220678856, + "learning_rate": 3.679150414258037e-08, + "loss": 0.0001, + "step": 9630 + }, + { + "epoch": 7.921117502054232, + "grad_norm": 0.00213727551582484, + "learning_rate": 3.02365995415943e-08, + "loss": 0.0001, + "step": 9640 + }, + { + "epoch": 7.929334428923583, + "grad_norm": 0.0021035040669768056, + "learning_rate": 2.4324086869204065e-08, + "loss": 0.0001, + "step": 9650 + }, + { + "epoch": 7.937551355792934, + "grad_norm": 0.0027864252055160286, + "learning_rate": 1.9054042134397652e-08, + "loss": 0.0001, + "step": 9660 + }, + { + "epoch": 7.945768282662284, + "grad_norm": 0.0022749610368685666, + "learning_rate": 1.4426533086819849e-08, + "loss": 0.0001, + "step": 9670 + }, + { + "epoch": 7.953985209531635, + "grad_norm": 0.0019970318188962314, + "learning_rate": 1.0441619215939558e-08, + "loss": 0.0001, + "step": 9680 + }, + { + "epoch": 7.962202136400986, + "grad_norm": 0.0018496564953550234, + "learning_rate": 7.0993517502671024e-09, + "loss": 0.0001, + "step": 9690 + }, + { + "epoch": 7.970419063270337, + "grad_norm": 0.003120026264055217, + "learning_rate": 4.39977365669364e-09, + "loss": 0.0001, + "step": 9700 + }, + { + "epoch": 7.978635990139688, + "grad_norm": 0.0027182376181643492, + "learning_rate": 2.342919639958252e-09, + "loss": 0.0001, + "step": 9710 + }, + { + "epoch": 7.986852917009038, + "grad_norm": 0.004404733338971566, + "learning_rate": 9.288161421705521e-10, + "loss": 0.0001, + "step": 9720 + }, + { + "epoch": 7.995069843878389, + "grad_norm": 0.0027499857957761086, + "learning_rate": 1.5748134250537227e-10, + "loss": 0.0001, + "step": 9730 + }, + { + "epoch": 8.0, + "step": 9736, + "total_flos": 575212087148544.0, + "train_loss": 0.11833881080070156, + "train_runtime": 79150.9203, + "train_samples_per_second": 62.964, + "train_steps_per_second": 0.123 + } + ], + "logging_steps": 10, + "max_steps": 9736, + "num_input_tokens_seen": 0, + "num_train_epochs": 8, + "save_steps": 1800, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 575212087148544.0, + "train_batch_size": 128, + "trial_name": null, + "trial_params": null +}