diff --git "a/checkpoint-2010/trainer_state.json" "b/checkpoint-2010/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-2010/trainer_state.json" @@ -0,0 +1,14103 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.30026889752016733, + "eval_steps": 500, + "global_step": 2010, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00014938751120406333, + "grad_norm": 0.5712769031524658, + "learning_rate": 2e-05, + "loss": 1.7864, + "step": 1 + }, + { + "epoch": 0.00029877502240812666, + "grad_norm": 0.8328073620796204, + "learning_rate": 4e-05, + "loss": 1.954, + "step": 2 + }, + { + "epoch": 0.00044816253361219, + "grad_norm": 0.3498156666755676, + "learning_rate": 6e-05, + "loss": 1.0704, + "step": 3 + }, + { + "epoch": 0.0005975500448162533, + "grad_norm": 0.46848297119140625, + "learning_rate": 8e-05, + "loss": 1.3401, + "step": 4 + }, + { + "epoch": 0.0007469375560203167, + "grad_norm": 0.5157446265220642, + "learning_rate": 0.0001, + "loss": 1.2173, + "step": 5 + }, + { + "epoch": 0.00089632506722438, + "grad_norm": 0.34888526797294617, + "learning_rate": 0.00012, + "loss": 1.2156, + "step": 6 + }, + { + "epoch": 0.0010457125784284435, + "grad_norm": 0.25781184434890747, + "learning_rate": 0.00014, + "loss": 1.1595, + "step": 7 + }, + { + "epoch": 0.0011951000896325067, + "grad_norm": 0.39395585656166077, + "learning_rate": 0.00016, + "loss": 1.447, + "step": 8 + }, + { + "epoch": 0.00134448760083657, + "grad_norm": 0.42594340443611145, + "learning_rate": 0.00018, + "loss": 0.8779, + "step": 9 + }, + { + "epoch": 0.0014938751120406335, + "grad_norm": 0.29849162697792053, + "learning_rate": 0.0002, + "loss": 1.214, + "step": 10 + }, + { + "epoch": 0.0016432626232446967, + "grad_norm": 0.206129252910614, + "learning_rate": 0.00019999998895420804, + "loss": 1.222, + "step": 11 + }, + { + "epoch": 0.00179265013444876, + "grad_norm": 0.20143035054206848, + "learning_rate": 0.0001999999558168346, + "loss": 1.2729, + "step": 12 + }, + { + "epoch": 0.0019420376456528235, + "grad_norm": 0.3609657883644104, + "learning_rate": 0.00019999990058788703, + "loss": 1.14, + "step": 13 + }, + { + "epoch": 0.002091425156856887, + "grad_norm": 0.22608326375484467, + "learning_rate": 0.00019999982326737747, + "loss": 1.2138, + "step": 14 + }, + { + "epoch": 0.00224081266806095, + "grad_norm": 0.40498849749565125, + "learning_rate": 0.00019999972385532303, + "loss": 1.4689, + "step": 15 + }, + { + "epoch": 0.0023902001792650133, + "grad_norm": 0.215151846408844, + "learning_rate": 0.0001999996023517457, + "loss": 1.0072, + "step": 16 + }, + { + "epoch": 0.0025395876904690767, + "grad_norm": 0.28476935625076294, + "learning_rate": 0.0001999994587566723, + "loss": 0.9636, + "step": 17 + }, + { + "epoch": 0.00268897520167314, + "grad_norm": 0.2294227033853531, + "learning_rate": 0.0001999992930701345, + "loss": 0.8722, + "step": 18 + }, + { + "epoch": 0.0028383627128772035, + "grad_norm": 0.24489282071590424, + "learning_rate": 0.00019999910529216902, + "loss": 0.913, + "step": 19 + }, + { + "epoch": 0.002987750224081267, + "grad_norm": 0.24592871963977814, + "learning_rate": 0.00019999889542281728, + "loss": 1.2072, + "step": 20 + }, + { + "epoch": 0.00313713773528533, + "grad_norm": 0.19988738000392914, + "learning_rate": 0.0001999986634621256, + "loss": 1.0463, + "step": 21 + }, + { + "epoch": 0.0032865252464893933, + "grad_norm": 0.16961582005023956, + "learning_rate": 0.00019999840941014525, + "loss": 0.8549, + "step": 22 + }, + { + "epoch": 0.0034359127576934568, + "grad_norm": 0.1631389707326889, + "learning_rate": 0.0001999981332669324, + "loss": 0.8112, + "step": 23 + }, + { + "epoch": 0.00358530026889752, + "grad_norm": 0.1402096003293991, + "learning_rate": 0.00019999783503254803, + "loss": 1.0202, + "step": 24 + }, + { + "epoch": 0.0037346877801015836, + "grad_norm": 0.3608367443084717, + "learning_rate": 0.000199997514707058, + "loss": 1.4076, + "step": 25 + }, + { + "epoch": 0.003884075291305647, + "grad_norm": 0.15870416164398193, + "learning_rate": 0.0001999971722905331, + "loss": 0.9372, + "step": 26 + }, + { + "epoch": 0.00403346280250971, + "grad_norm": 0.21479852497577667, + "learning_rate": 0.00019999680778304897, + "loss": 1.0237, + "step": 27 + }, + { + "epoch": 0.004182850313713774, + "grad_norm": 0.6759325861930847, + "learning_rate": 0.00019999642118468614, + "loss": 1.7162, + "step": 28 + }, + { + "epoch": 0.004332237824917837, + "grad_norm": 0.15937522053718567, + "learning_rate": 0.00019999601249552998, + "loss": 0.7851, + "step": 29 + }, + { + "epoch": 0.0044816253361219, + "grad_norm": 0.12428741902112961, + "learning_rate": 0.00019999558171567082, + "loss": 0.9342, + "step": 30 + }, + { + "epoch": 0.004631012847325963, + "grad_norm": 0.2223120778799057, + "learning_rate": 0.0001999951288452038, + "loss": 1.1871, + "step": 31 + }, + { + "epoch": 0.004780400358530027, + "grad_norm": 0.19341352581977844, + "learning_rate": 0.000199994653884229, + "loss": 0.9797, + "step": 32 + }, + { + "epoch": 0.00492978786973409, + "grad_norm": 0.41403812170028687, + "learning_rate": 0.0001999941568328513, + "loss": 0.9629, + "step": 33 + }, + { + "epoch": 0.005079175380938153, + "grad_norm": 0.29959213733673096, + "learning_rate": 0.00019999363769118055, + "loss": 1.231, + "step": 34 + }, + { + "epoch": 0.005228562892142217, + "grad_norm": 0.17178373038768768, + "learning_rate": 0.00019999309645933142, + "loss": 0.7834, + "step": 35 + }, + { + "epoch": 0.00537795040334628, + "grad_norm": 0.2903209626674652, + "learning_rate": 0.00019999253313742344, + "loss": 1.2852, + "step": 36 + }, + { + "epoch": 0.005527337914550344, + "grad_norm": 0.3148847818374634, + "learning_rate": 0.00019999194772558112, + "loss": 1.376, + "step": 37 + }, + { + "epoch": 0.005676725425754407, + "grad_norm": 0.4141716957092285, + "learning_rate": 0.00019999134022393375, + "loss": 0.9684, + "step": 38 + }, + { + "epoch": 0.0058261129369584705, + "grad_norm": 0.2509758770465851, + "learning_rate": 0.00019999071063261554, + "loss": 1.1722, + "step": 39 + }, + { + "epoch": 0.005975500448162534, + "grad_norm": 0.1800830066204071, + "learning_rate": 0.0001999900589517656, + "loss": 1.0264, + "step": 40 + }, + { + "epoch": 0.006124887959366597, + "grad_norm": 0.19982534646987915, + "learning_rate": 0.00019998938518152787, + "loss": 0.7343, + "step": 41 + }, + { + "epoch": 0.00627427547057066, + "grad_norm": 0.1816824972629547, + "learning_rate": 0.0001999886893220512, + "loss": 0.9136, + "step": 42 + }, + { + "epoch": 0.006423662981774723, + "grad_norm": 0.216154083609581, + "learning_rate": 0.0001999879713734893, + "loss": 1.0128, + "step": 43 + }, + { + "epoch": 0.006573050492978787, + "grad_norm": 0.16445758938789368, + "learning_rate": 0.0001999872313360008, + "loss": 0.9593, + "step": 44 + }, + { + "epoch": 0.00672243800418285, + "grad_norm": 0.14644479751586914, + "learning_rate": 0.00019998646920974919, + "loss": 0.9424, + "step": 45 + }, + { + "epoch": 0.0068718255153869135, + "grad_norm": 0.3356267809867859, + "learning_rate": 0.00019998568499490283, + "loss": 0.8878, + "step": 46 + }, + { + "epoch": 0.007021213026590977, + "grad_norm": 0.4974273443222046, + "learning_rate": 0.00019998487869163497, + "loss": 1.338, + "step": 47 + }, + { + "epoch": 0.00717060053779504, + "grad_norm": 0.1881432980298996, + "learning_rate": 0.00019998405030012371, + "loss": 0.882, + "step": 48 + }, + { + "epoch": 0.007319988048999104, + "grad_norm": 0.1419169157743454, + "learning_rate": 0.0001999831998205521, + "loss": 0.7255, + "step": 49 + }, + { + "epoch": 0.007469375560203167, + "grad_norm": 0.16964636743068695, + "learning_rate": 0.00019998232725310796, + "loss": 1.0151, + "step": 50 + }, + { + "epoch": 0.007618763071407231, + "grad_norm": 0.18110951781272888, + "learning_rate": 0.0001999814325979841, + "loss": 0.9705, + "step": 51 + }, + { + "epoch": 0.007768150582611294, + "grad_norm": 0.13390210270881653, + "learning_rate": 0.00019998051585537818, + "loss": 0.6999, + "step": 52 + }, + { + "epoch": 0.007917538093815357, + "grad_norm": 0.13011352717876434, + "learning_rate": 0.00019997957702549269, + "loss": 0.5496, + "step": 53 + }, + { + "epoch": 0.00806692560501942, + "grad_norm": 0.30082619190216064, + "learning_rate": 0.00019997861610853503, + "loss": 1.1625, + "step": 54 + }, + { + "epoch": 0.008216313116223483, + "grad_norm": 0.18591542541980743, + "learning_rate": 0.00019997763310471752, + "loss": 0.8686, + "step": 55 + }, + { + "epoch": 0.008365700627427548, + "grad_norm": 0.20125310122966766, + "learning_rate": 0.00019997662801425725, + "loss": 0.7371, + "step": 56 + }, + { + "epoch": 0.00851508813863161, + "grad_norm": 0.14571277797222137, + "learning_rate": 0.00019997560083737632, + "loss": 1.0109, + "step": 57 + }, + { + "epoch": 0.008664475649835674, + "grad_norm": 0.1365530639886856, + "learning_rate": 0.00019997455157430165, + "loss": 0.7709, + "step": 58 + }, + { + "epoch": 0.008813863161039737, + "grad_norm": 0.14306791126728058, + "learning_rate": 0.000199973480225265, + "loss": 0.7487, + "step": 59 + }, + { + "epoch": 0.0089632506722438, + "grad_norm": 0.2880021929740906, + "learning_rate": 0.00019997238679050308, + "loss": 1.2318, + "step": 60 + }, + { + "epoch": 0.009112638183447864, + "grad_norm": 0.1913180649280548, + "learning_rate": 0.00019997127127025746, + "loss": 0.8861, + "step": 61 + }, + { + "epoch": 0.009262025694651926, + "grad_norm": 0.44290822744369507, + "learning_rate": 0.00019997013366477453, + "loss": 1.3866, + "step": 62 + }, + { + "epoch": 0.00941141320585599, + "grad_norm": 0.12451759725809097, + "learning_rate": 0.00019996897397430563, + "loss": 0.855, + "step": 63 + }, + { + "epoch": 0.009560800717060053, + "grad_norm": 0.2885231375694275, + "learning_rate": 0.00019996779219910696, + "loss": 1.147, + "step": 64 + }, + { + "epoch": 0.009710188228264117, + "grad_norm": 0.14056627452373505, + "learning_rate": 0.00019996658833943957, + "loss": 0.7883, + "step": 65 + }, + { + "epoch": 0.00985957573946818, + "grad_norm": 0.14305633306503296, + "learning_rate": 0.00019996536239556942, + "loss": 0.5779, + "step": 66 + }, + { + "epoch": 0.010008963250672244, + "grad_norm": 0.17166224122047424, + "learning_rate": 0.00019996411436776738, + "loss": 0.9014, + "step": 67 + }, + { + "epoch": 0.010158350761876307, + "grad_norm": 0.15985064208507538, + "learning_rate": 0.0001999628442563091, + "loss": 0.9639, + "step": 68 + }, + { + "epoch": 0.010307738273080371, + "grad_norm": 0.21808995306491852, + "learning_rate": 0.0001999615520614752, + "loss": 0.7525, + "step": 69 + }, + { + "epoch": 0.010457125784284434, + "grad_norm": 0.21660713851451874, + "learning_rate": 0.00019996023778355113, + "loss": 0.9379, + "step": 70 + }, + { + "epoch": 0.010606513295488498, + "grad_norm": 0.20963691174983978, + "learning_rate": 0.00019995890142282728, + "loss": 0.6632, + "step": 71 + }, + { + "epoch": 0.01075590080669256, + "grad_norm": 0.2161312997341156, + "learning_rate": 0.00019995754297959882, + "loss": 0.8932, + "step": 72 + }, + { + "epoch": 0.010905288317896623, + "grad_norm": 0.17720285058021545, + "learning_rate": 0.00019995616245416584, + "loss": 0.5283, + "step": 73 + }, + { + "epoch": 0.011054675829100687, + "grad_norm": 0.18077623844146729, + "learning_rate": 0.0001999547598468334, + "loss": 0.603, + "step": 74 + }, + { + "epoch": 0.01120406334030475, + "grad_norm": 0.19094090163707733, + "learning_rate": 0.00019995333515791125, + "loss": 0.9712, + "step": 75 + }, + { + "epoch": 0.011353450851508814, + "grad_norm": 0.2597976326942444, + "learning_rate": 0.00019995188838771425, + "loss": 0.7913, + "step": 76 + }, + { + "epoch": 0.011502838362712877, + "grad_norm": 0.26235243678092957, + "learning_rate": 0.00019995041953656194, + "loss": 0.7578, + "step": 77 + }, + { + "epoch": 0.011652225873916941, + "grad_norm": 0.1809563785791397, + "learning_rate": 0.0001999489286047788, + "loss": 0.9168, + "step": 78 + }, + { + "epoch": 0.011801613385121004, + "grad_norm": 0.22458118200302124, + "learning_rate": 0.0001999474155926942, + "loss": 1.1456, + "step": 79 + }, + { + "epoch": 0.011951000896325068, + "grad_norm": 0.49804338812828064, + "learning_rate": 0.00019994588050064243, + "loss": 1.1896, + "step": 80 + }, + { + "epoch": 0.01210038840752913, + "grad_norm": 0.2247380167245865, + "learning_rate": 0.00019994432332896258, + "loss": 0.958, + "step": 81 + }, + { + "epoch": 0.012249775918733195, + "grad_norm": 0.17649094760417938, + "learning_rate": 0.00019994274407799872, + "loss": 0.9417, + "step": 82 + }, + { + "epoch": 0.012399163429937257, + "grad_norm": 0.24156750738620758, + "learning_rate": 0.00019994114274809964, + "loss": 0.9959, + "step": 83 + }, + { + "epoch": 0.01254855094114132, + "grad_norm": 0.2841317355632782, + "learning_rate": 0.00019993951933961913, + "loss": 1.296, + "step": 84 + }, + { + "epoch": 0.012697938452345384, + "grad_norm": 0.26792535185813904, + "learning_rate": 0.00019993787385291588, + "loss": 1.0296, + "step": 85 + }, + { + "epoch": 0.012847325963549447, + "grad_norm": 0.11657895892858505, + "learning_rate": 0.00019993620628835332, + "loss": 0.6993, + "step": 86 + }, + { + "epoch": 0.01299671347475351, + "grad_norm": 0.15463922917842865, + "learning_rate": 0.0001999345166462999, + "loss": 0.6925, + "step": 87 + }, + { + "epoch": 0.013146100985957573, + "grad_norm": 0.1776818037033081, + "learning_rate": 0.0001999328049271289, + "loss": 0.7629, + "step": 88 + }, + { + "epoch": 0.013295488497161638, + "grad_norm": 0.16185057163238525, + "learning_rate": 0.00019993107113121844, + "loss": 1.0736, + "step": 89 + }, + { + "epoch": 0.0134448760083657, + "grad_norm": 0.3227018415927887, + "learning_rate": 0.0001999293152589515, + "loss": 0.8586, + "step": 90 + }, + { + "epoch": 0.013594263519569764, + "grad_norm": 0.16486088931560516, + "learning_rate": 0.00019992753731071602, + "loss": 1.0614, + "step": 91 + }, + { + "epoch": 0.013743651030773827, + "grad_norm": 0.1495029777288437, + "learning_rate": 0.0001999257372869048, + "loss": 0.8561, + "step": 92 + }, + { + "epoch": 0.013893038541977891, + "grad_norm": 0.16044558584690094, + "learning_rate": 0.00019992391518791546, + "loss": 1.0294, + "step": 93 + }, + { + "epoch": 0.014042426053181954, + "grad_norm": 0.16251197457313538, + "learning_rate": 0.00019992207101415053, + "loss": 0.838, + "step": 94 + }, + { + "epoch": 0.014191813564386018, + "grad_norm": 0.47265177965164185, + "learning_rate": 0.00019992020476601745, + "loss": 1.4965, + "step": 95 + }, + { + "epoch": 0.01434120107559008, + "grad_norm": 0.16032789647579193, + "learning_rate": 0.00019991831644392848, + "loss": 0.9541, + "step": 96 + }, + { + "epoch": 0.014490588586794143, + "grad_norm": 0.1563224196434021, + "learning_rate": 0.0001999164060483008, + "loss": 0.6754, + "step": 97 + }, + { + "epoch": 0.014639976097998207, + "grad_norm": 0.2469077855348587, + "learning_rate": 0.00019991447357955639, + "loss": 0.9911, + "step": 98 + }, + { + "epoch": 0.01478936360920227, + "grad_norm": 0.13441061973571777, + "learning_rate": 0.00019991251903812225, + "loss": 0.7342, + "step": 99 + }, + { + "epoch": 0.014938751120406334, + "grad_norm": 0.20193825662136078, + "learning_rate": 0.00019991054242443008, + "loss": 0.6834, + "step": 100 + }, + { + "epoch": 0.015088138631610397, + "grad_norm": 0.1552024483680725, + "learning_rate": 0.0001999085437389166, + "loss": 0.8257, + "step": 101 + }, + { + "epoch": 0.015237526142814461, + "grad_norm": 0.25607559084892273, + "learning_rate": 0.00019990652298202335, + "loss": 1.1036, + "step": 102 + }, + { + "epoch": 0.015386913654018524, + "grad_norm": 0.11925558000802994, + "learning_rate": 0.00019990448015419675, + "loss": 0.6345, + "step": 103 + }, + { + "epoch": 0.015536301165222588, + "grad_norm": 0.28443989157676697, + "learning_rate": 0.00019990241525588804, + "loss": 1.1217, + "step": 104 + }, + { + "epoch": 0.01568568867642665, + "grad_norm": 0.15237168967723846, + "learning_rate": 0.00019990032828755345, + "loss": 0.7735, + "step": 105 + }, + { + "epoch": 0.015835076187630713, + "grad_norm": 0.173554927110672, + "learning_rate": 0.000199898219249654, + "loss": 1.1093, + "step": 106 + }, + { + "epoch": 0.01598446369883478, + "grad_norm": 0.14035388827323914, + "learning_rate": 0.0001998960881426556, + "loss": 0.8206, + "step": 107 + }, + { + "epoch": 0.01613385121003884, + "grad_norm": 0.13699080049991608, + "learning_rate": 0.00019989393496702907, + "loss": 1.0426, + "step": 108 + }, + { + "epoch": 0.016283238721242904, + "grad_norm": 0.1313813477754593, + "learning_rate": 0.00019989175972325005, + "loss": 0.7804, + "step": 109 + }, + { + "epoch": 0.016432626232446967, + "grad_norm": 0.15459921956062317, + "learning_rate": 0.00019988956241179912, + "loss": 0.9949, + "step": 110 + }, + { + "epoch": 0.01658201374365103, + "grad_norm": 0.20009522140026093, + "learning_rate": 0.00019988734303316168, + "loss": 1.0159, + "step": 111 + }, + { + "epoch": 0.016731401254855095, + "grad_norm": 0.17909283936023712, + "learning_rate": 0.00019988510158782804, + "loss": 0.5281, + "step": 112 + }, + { + "epoch": 0.016880788766059158, + "grad_norm": 0.6786065697669983, + "learning_rate": 0.00019988283807629334, + "loss": 1.2808, + "step": 113 + }, + { + "epoch": 0.01703017627726322, + "grad_norm": 0.15410082042217255, + "learning_rate": 0.00019988055249905767, + "loss": 1.0755, + "step": 114 + }, + { + "epoch": 0.017179563788467283, + "grad_norm": 0.19600564241409302, + "learning_rate": 0.00019987824485662593, + "loss": 1.2168, + "step": 115 + }, + { + "epoch": 0.01732895129967135, + "grad_norm": 0.1903967261314392, + "learning_rate": 0.00019987591514950787, + "loss": 0.9987, + "step": 116 + }, + { + "epoch": 0.01747833881087541, + "grad_norm": 0.13745927810668945, + "learning_rate": 0.00019987356337821822, + "loss": 0.9328, + "step": 117 + }, + { + "epoch": 0.017627726322079474, + "grad_norm": 0.18400371074676514, + "learning_rate": 0.00019987118954327654, + "loss": 0.6686, + "step": 118 + }, + { + "epoch": 0.017777113833283537, + "grad_norm": 0.29325294494628906, + "learning_rate": 0.00019986879364520716, + "loss": 1.173, + "step": 119 + }, + { + "epoch": 0.0179265013444876, + "grad_norm": 0.15400448441505432, + "learning_rate": 0.00019986637568453945, + "loss": 0.897, + "step": 120 + }, + { + "epoch": 0.018075888855691665, + "grad_norm": 0.2178712785243988, + "learning_rate": 0.00019986393566180755, + "loss": 1.2885, + "step": 121 + }, + { + "epoch": 0.018225276366895728, + "grad_norm": 0.15217332541942596, + "learning_rate": 0.00019986147357755048, + "loss": 0.9847, + "step": 122 + }, + { + "epoch": 0.01837466387809979, + "grad_norm": 0.35806745290756226, + "learning_rate": 0.0001998589894323122, + "loss": 1.0114, + "step": 123 + }, + { + "epoch": 0.018524051389303853, + "grad_norm": 0.46978825330734253, + "learning_rate": 0.00019985648322664145, + "loss": 1.33, + "step": 124 + }, + { + "epoch": 0.01867343890050792, + "grad_norm": 0.3191573917865753, + "learning_rate": 0.00019985395496109192, + "loss": 0.8691, + "step": 125 + }, + { + "epoch": 0.01882282641171198, + "grad_norm": 0.6782920360565186, + "learning_rate": 0.00019985140463622215, + "loss": 1.7387, + "step": 126 + }, + { + "epoch": 0.018972213922916044, + "grad_norm": 0.4556753933429718, + "learning_rate": 0.0001998488322525955, + "loss": 1.9508, + "step": 127 + }, + { + "epoch": 0.019121601434120106, + "grad_norm": 0.1401677280664444, + "learning_rate": 0.0001998462378107803, + "loss": 0.8127, + "step": 128 + }, + { + "epoch": 0.019270988945324172, + "grad_norm": 0.1187644675374031, + "learning_rate": 0.00019984362131134968, + "loss": 0.5421, + "step": 129 + }, + { + "epoch": 0.019420376456528235, + "grad_norm": 0.17097711563110352, + "learning_rate": 0.0001998409827548817, + "loss": 0.7335, + "step": 130 + }, + { + "epoch": 0.019569763967732298, + "grad_norm": 0.12986791133880615, + "learning_rate": 0.00019983832214195917, + "loss": 0.7151, + "step": 131 + }, + { + "epoch": 0.01971915147893636, + "grad_norm": 0.14413940906524658, + "learning_rate": 0.00019983563947316996, + "loss": 0.6833, + "step": 132 + }, + { + "epoch": 0.019868538990140423, + "grad_norm": 0.1424313336610794, + "learning_rate": 0.00019983293474910667, + "loss": 0.8199, + "step": 133 + }, + { + "epoch": 0.02001792650134449, + "grad_norm": 0.11715007573366165, + "learning_rate": 0.00019983020797036683, + "loss": 0.6079, + "step": 134 + }, + { + "epoch": 0.02016731401254855, + "grad_norm": 0.11654297262430191, + "learning_rate": 0.00019982745913755282, + "loss": 0.5302, + "step": 135 + }, + { + "epoch": 0.020316701523752614, + "grad_norm": 0.1768021434545517, + "learning_rate": 0.00019982468825127187, + "loss": 0.5412, + "step": 136 + }, + { + "epoch": 0.020466089034956676, + "grad_norm": 0.2620059549808502, + "learning_rate": 0.00019982189531213618, + "loss": 1.0321, + "step": 137 + }, + { + "epoch": 0.020615476546160742, + "grad_norm": 0.1472969800233841, + "learning_rate": 0.0001998190803207627, + "loss": 1.0097, + "step": 138 + }, + { + "epoch": 0.020764864057364805, + "grad_norm": 0.1889103651046753, + "learning_rate": 0.00019981624327777332, + "loss": 0.7423, + "step": 139 + }, + { + "epoch": 0.020914251568568867, + "grad_norm": 0.20594674348831177, + "learning_rate": 0.0001998133841837948, + "loss": 1.3315, + "step": 140 + }, + { + "epoch": 0.02106363907977293, + "grad_norm": 0.19783402979373932, + "learning_rate": 0.00019981050303945877, + "loss": 1.0435, + "step": 141 + }, + { + "epoch": 0.021213026590976996, + "grad_norm": 0.12695230543613434, + "learning_rate": 0.00019980759984540168, + "loss": 0.8, + "step": 142 + }, + { + "epoch": 0.02136241410218106, + "grad_norm": 0.18909701704978943, + "learning_rate": 0.0001998046746022649, + "loss": 0.7549, + "step": 143 + }, + { + "epoch": 0.02151180161338512, + "grad_norm": 0.2600228488445282, + "learning_rate": 0.0001998017273106947, + "loss": 0.6657, + "step": 144 + }, + { + "epoch": 0.021661189124589184, + "grad_norm": 0.2999661862850189, + "learning_rate": 0.00019979875797134216, + "loss": 0.9626, + "step": 145 + }, + { + "epoch": 0.021810576635793246, + "grad_norm": 0.2698724567890167, + "learning_rate": 0.00019979576658486325, + "loss": 0.8856, + "step": 146 + }, + { + "epoch": 0.021959964146997312, + "grad_norm": 0.14443442225456238, + "learning_rate": 0.0001997927531519188, + "loss": 0.9285, + "step": 147 + }, + { + "epoch": 0.022109351658201375, + "grad_norm": 0.21873806416988373, + "learning_rate": 0.00019978971767317457, + "loss": 0.9565, + "step": 148 + }, + { + "epoch": 0.022258739169405437, + "grad_norm": 0.2041054517030716, + "learning_rate": 0.0001997866601493011, + "loss": 0.9455, + "step": 149 + }, + { + "epoch": 0.0224081266806095, + "grad_norm": 0.19350048899650574, + "learning_rate": 0.00019978358058097388, + "loss": 0.8047, + "step": 150 + }, + { + "epoch": 0.022557514191813566, + "grad_norm": 0.25492972135543823, + "learning_rate": 0.00019978047896887323, + "loss": 1.234, + "step": 151 + }, + { + "epoch": 0.02270690170301763, + "grad_norm": 0.1414925754070282, + "learning_rate": 0.0001997773553136843, + "loss": 0.9703, + "step": 152 + }, + { + "epoch": 0.02285628921422169, + "grad_norm": 0.15996232628822327, + "learning_rate": 0.00019977420961609721, + "loss": 1.1475, + "step": 153 + }, + { + "epoch": 0.023005676725425753, + "grad_norm": 0.1661243587732315, + "learning_rate": 0.00019977104187680688, + "loss": 0.8574, + "step": 154 + }, + { + "epoch": 0.02315506423662982, + "grad_norm": 0.17797410488128662, + "learning_rate": 0.00019976785209651309, + "loss": 0.6742, + "step": 155 + }, + { + "epoch": 0.023304451747833882, + "grad_norm": 0.1401919722557068, + "learning_rate": 0.00019976464027592053, + "loss": 0.7065, + "step": 156 + }, + { + "epoch": 0.023453839259037945, + "grad_norm": 0.3743472695350647, + "learning_rate": 0.00019976140641573875, + "loss": 1.1391, + "step": 157 + }, + { + "epoch": 0.023603226770242007, + "grad_norm": 0.12242946028709412, + "learning_rate": 0.00019975815051668217, + "loss": 0.7517, + "step": 158 + }, + { + "epoch": 0.02375261428144607, + "grad_norm": 0.12851974368095398, + "learning_rate": 0.00019975487257947004, + "loss": 0.5795, + "step": 159 + }, + { + "epoch": 0.023902001792650136, + "grad_norm": 0.15131576359272003, + "learning_rate": 0.0001997515726048265, + "loss": 0.5735, + "step": 160 + }, + { + "epoch": 0.024051389303854198, + "grad_norm": 0.2606663703918457, + "learning_rate": 0.00019974825059348062, + "loss": 0.7677, + "step": 161 + }, + { + "epoch": 0.02420077681505826, + "grad_norm": 0.3116661608219147, + "learning_rate": 0.00019974490654616625, + "loss": 1.1465, + "step": 162 + }, + { + "epoch": 0.024350164326262323, + "grad_norm": 0.15296891331672668, + "learning_rate": 0.00019974154046362212, + "loss": 0.9154, + "step": 163 + }, + { + "epoch": 0.02449955183746639, + "grad_norm": 0.12761889398097992, + "learning_rate": 0.0001997381523465919, + "loss": 0.7902, + "step": 164 + }, + { + "epoch": 0.024648939348670452, + "grad_norm": 0.11566565185785294, + "learning_rate": 0.00019973474219582405, + "loss": 0.5558, + "step": 165 + }, + { + "epoch": 0.024798326859874514, + "grad_norm": 0.12765946984291077, + "learning_rate": 0.00019973131001207195, + "loss": 0.7314, + "step": 166 + }, + { + "epoch": 0.024947714371078577, + "grad_norm": 0.12568241357803345, + "learning_rate": 0.00019972785579609376, + "loss": 0.9476, + "step": 167 + }, + { + "epoch": 0.02509710188228264, + "grad_norm": 0.21633781492710114, + "learning_rate": 0.00019972437954865265, + "loss": 1.3121, + "step": 168 + }, + { + "epoch": 0.025246489393486705, + "grad_norm": 0.14913877844810486, + "learning_rate": 0.00019972088127051657, + "loss": 0.8247, + "step": 169 + }, + { + "epoch": 0.025395876904690768, + "grad_norm": 0.16602329909801483, + "learning_rate": 0.00019971736096245825, + "loss": 1.064, + "step": 170 + }, + { + "epoch": 0.02554526441589483, + "grad_norm": 0.28753912448883057, + "learning_rate": 0.00019971381862525552, + "loss": 1.4888, + "step": 171 + }, + { + "epoch": 0.025694651927098893, + "grad_norm": 0.13800154626369476, + "learning_rate": 0.00019971025425969083, + "loss": 0.5945, + "step": 172 + }, + { + "epoch": 0.02584403943830296, + "grad_norm": 0.22065165638923645, + "learning_rate": 0.00019970666786655166, + "loss": 0.8695, + "step": 173 + }, + { + "epoch": 0.02599342694950702, + "grad_norm": 0.3128964900970459, + "learning_rate": 0.0001997030594466303, + "loss": 1.1673, + "step": 174 + }, + { + "epoch": 0.026142814460711084, + "grad_norm": 0.20117323100566864, + "learning_rate": 0.00019969942900072387, + "loss": 1.0395, + "step": 175 + }, + { + "epoch": 0.026292201971915147, + "grad_norm": 0.14214551448822021, + "learning_rate": 0.00019969577652963444, + "loss": 0.5135, + "step": 176 + }, + { + "epoch": 0.026441589483119213, + "grad_norm": 0.20523157715797424, + "learning_rate": 0.00019969210203416883, + "loss": 1.0156, + "step": 177 + }, + { + "epoch": 0.026590976994323275, + "grad_norm": 0.1385200172662735, + "learning_rate": 0.0001996884055151389, + "loss": 0.5202, + "step": 178 + }, + { + "epoch": 0.026740364505527338, + "grad_norm": 0.14539465308189392, + "learning_rate": 0.00019968468697336117, + "loss": 0.6127, + "step": 179 + }, + { + "epoch": 0.0268897520167314, + "grad_norm": 0.2886626720428467, + "learning_rate": 0.00019968094640965717, + "loss": 1.213, + "step": 180 + }, + { + "epoch": 0.027039139527935463, + "grad_norm": 0.22695106267929077, + "learning_rate": 0.00019967718382485323, + "loss": 1.137, + "step": 181 + }, + { + "epoch": 0.02718852703913953, + "grad_norm": 0.17426621913909912, + "learning_rate": 0.00019967339921978062, + "loss": 0.8978, + "step": 182 + }, + { + "epoch": 0.02733791455034359, + "grad_norm": 0.2875833809375763, + "learning_rate": 0.00019966959259527534, + "loss": 0.8688, + "step": 183 + }, + { + "epoch": 0.027487302061547654, + "grad_norm": 0.14763344824314117, + "learning_rate": 0.00019966576395217837, + "loss": 0.564, + "step": 184 + }, + { + "epoch": 0.027636689572751717, + "grad_norm": 0.24988707900047302, + "learning_rate": 0.00019966191329133555, + "loss": 0.429, + "step": 185 + }, + { + "epoch": 0.027786077083955783, + "grad_norm": 0.16043034195899963, + "learning_rate": 0.0001996580406135975, + "loss": 0.7251, + "step": 186 + }, + { + "epoch": 0.027935464595159845, + "grad_norm": 0.2776719033718109, + "learning_rate": 0.00019965414591981975, + "loss": 1.3513, + "step": 187 + }, + { + "epoch": 0.028084852106363908, + "grad_norm": 0.1749371886253357, + "learning_rate": 0.00019965022921086275, + "loss": 0.7383, + "step": 188 + }, + { + "epoch": 0.02823423961756797, + "grad_norm": 0.2881135940551758, + "learning_rate": 0.00019964629048759176, + "loss": 1.0511, + "step": 189 + }, + { + "epoch": 0.028383627128772036, + "grad_norm": 0.11646547168493271, + "learning_rate": 0.00019964232975087687, + "loss": 0.6575, + "step": 190 + }, + { + "epoch": 0.0285330146399761, + "grad_norm": 0.1457265317440033, + "learning_rate": 0.0001996383470015931, + "loss": 0.7538, + "step": 191 + }, + { + "epoch": 0.02868240215118016, + "grad_norm": 0.13400131464004517, + "learning_rate": 0.00019963434224062025, + "loss": 0.7773, + "step": 192 + }, + { + "epoch": 0.028831789662384224, + "grad_norm": 0.23350481688976288, + "learning_rate": 0.0001996303154688431, + "loss": 1.2769, + "step": 193 + }, + { + "epoch": 0.028981177173588286, + "grad_norm": 0.563207745552063, + "learning_rate": 0.0001996262666871512, + "loss": 1.5822, + "step": 194 + }, + { + "epoch": 0.029130564684792352, + "grad_norm": 0.5377495884895325, + "learning_rate": 0.00019962219589643898, + "loss": 1.4911, + "step": 195 + }, + { + "epoch": 0.029279952195996415, + "grad_norm": 0.15726317465305328, + "learning_rate": 0.00019961810309760577, + "loss": 0.7824, + "step": 196 + }, + { + "epoch": 0.029429339707200478, + "grad_norm": 0.15040062367916107, + "learning_rate": 0.00019961398829155568, + "loss": 0.7541, + "step": 197 + }, + { + "epoch": 0.02957872721840454, + "grad_norm": 0.12620890140533447, + "learning_rate": 0.00019960985147919778, + "loss": 0.932, + "step": 198 + }, + { + "epoch": 0.029728114729608606, + "grad_norm": 0.15029945969581604, + "learning_rate": 0.00019960569266144597, + "loss": 0.9588, + "step": 199 + }, + { + "epoch": 0.02987750224081267, + "grad_norm": 0.16770395636558533, + "learning_rate": 0.00019960151183921897, + "loss": 0.5964, + "step": 200 + }, + { + "epoch": 0.03002688975201673, + "grad_norm": 0.12649193406105042, + "learning_rate": 0.0001995973090134404, + "loss": 0.8221, + "step": 201 + }, + { + "epoch": 0.030176277263220794, + "grad_norm": 0.15091091394424438, + "learning_rate": 0.00019959308418503877, + "loss": 0.8811, + "step": 202 + }, + { + "epoch": 0.03032566477442486, + "grad_norm": 0.12077690660953522, + "learning_rate": 0.00019958883735494732, + "loss": 0.6411, + "step": 203 + }, + { + "epoch": 0.030475052285628922, + "grad_norm": 0.15071533620357513, + "learning_rate": 0.00019958456852410433, + "loss": 0.7742, + "step": 204 + }, + { + "epoch": 0.030624439796832985, + "grad_norm": 0.1368575096130371, + "learning_rate": 0.00019958027769345277, + "loss": 1.1054, + "step": 205 + }, + { + "epoch": 0.030773827308037047, + "grad_norm": 0.22218400239944458, + "learning_rate": 0.0001995759648639406, + "loss": 0.8127, + "step": 206 + }, + { + "epoch": 0.03092321481924111, + "grad_norm": 0.12783220410346985, + "learning_rate": 0.00019957163003652063, + "loss": 0.6916, + "step": 207 + }, + { + "epoch": 0.031072602330445176, + "grad_norm": 0.18626387417316437, + "learning_rate": 0.00019956727321215044, + "loss": 0.8217, + "step": 208 + }, + { + "epoch": 0.03122198984164924, + "grad_norm": 0.12861424684524536, + "learning_rate": 0.0001995628943917925, + "loss": 1.0327, + "step": 209 + }, + { + "epoch": 0.0313713773528533, + "grad_norm": 0.17585206031799316, + "learning_rate": 0.00019955849357641424, + "loss": 0.6836, + "step": 210 + }, + { + "epoch": 0.03152076486405737, + "grad_norm": 0.1541229784488678, + "learning_rate": 0.0001995540707669878, + "loss": 1.1849, + "step": 211 + }, + { + "epoch": 0.031670152375261426, + "grad_norm": 0.13117974996566772, + "learning_rate": 0.00019954962596449024, + "loss": 0.7779, + "step": 212 + }, + { + "epoch": 0.03181953988646549, + "grad_norm": 0.3847130537033081, + "learning_rate": 0.00019954515916990358, + "loss": 1.4158, + "step": 213 + }, + { + "epoch": 0.03196892739766956, + "grad_norm": 0.21019677817821503, + "learning_rate": 0.0001995406703842145, + "loss": 1.0807, + "step": 214 + }, + { + "epoch": 0.03211831490887362, + "grad_norm": 0.10954124480485916, + "learning_rate": 0.0001995361596084147, + "loss": 0.7142, + "step": 215 + }, + { + "epoch": 0.03226770242007768, + "grad_norm": 0.25598999857902527, + "learning_rate": 0.0001995316268435007, + "loss": 0.7528, + "step": 216 + }, + { + "epoch": 0.03241708993128174, + "grad_norm": 0.16298946738243103, + "learning_rate": 0.0001995270720904738, + "loss": 0.6978, + "step": 217 + }, + { + "epoch": 0.03256647744248581, + "grad_norm": 0.37723109126091003, + "learning_rate": 0.00019952249535034025, + "loss": 1.0603, + "step": 218 + }, + { + "epoch": 0.032715864953689874, + "grad_norm": 0.1600012332201004, + "learning_rate": 0.00019951789662411113, + "loss": 0.7634, + "step": 219 + }, + { + "epoch": 0.03286525246489393, + "grad_norm": 0.1618419587612152, + "learning_rate": 0.00019951327591280236, + "loss": 0.8068, + "step": 220 + }, + { + "epoch": 0.033014639976098, + "grad_norm": 0.11850997805595398, + "learning_rate": 0.00019950863321743475, + "loss": 0.7883, + "step": 221 + }, + { + "epoch": 0.03316402748730206, + "grad_norm": 0.18700255453586578, + "learning_rate": 0.0001995039685390339, + "loss": 0.5821, + "step": 222 + }, + { + "epoch": 0.033313414998506125, + "grad_norm": 0.19264543056488037, + "learning_rate": 0.00019949928187863036, + "loss": 0.7578, + "step": 223 + }, + { + "epoch": 0.03346280250971019, + "grad_norm": 0.23828521370887756, + "learning_rate": 0.00019949457323725946, + "loss": 1.1739, + "step": 224 + }, + { + "epoch": 0.03361219002091425, + "grad_norm": 0.2178596705198288, + "learning_rate": 0.0001994898426159614, + "loss": 0.6075, + "step": 225 + }, + { + "epoch": 0.033761577532118316, + "grad_norm": 0.13766655325889587, + "learning_rate": 0.0001994850900157813, + "loss": 0.6941, + "step": 226 + }, + { + "epoch": 0.03391096504332238, + "grad_norm": 0.345782071352005, + "learning_rate": 0.00019948031543776904, + "loss": 1.0944, + "step": 227 + }, + { + "epoch": 0.03406035255452644, + "grad_norm": 0.1483844369649887, + "learning_rate": 0.0001994755188829794, + "loss": 0.9752, + "step": 228 + }, + { + "epoch": 0.03420974006573051, + "grad_norm": 0.16494938731193542, + "learning_rate": 0.00019947070035247205, + "loss": 0.7408, + "step": 229 + }, + { + "epoch": 0.034359127576934566, + "grad_norm": 0.15683765709400177, + "learning_rate": 0.00019946585984731142, + "loss": 0.6498, + "step": 230 + }, + { + "epoch": 0.03450851508813863, + "grad_norm": 0.17898918688297272, + "learning_rate": 0.0001994609973685669, + "loss": 0.7338, + "step": 231 + }, + { + "epoch": 0.0346579025993427, + "grad_norm": 0.15439556539058685, + "learning_rate": 0.00019945611291731274, + "loss": 0.854, + "step": 232 + }, + { + "epoch": 0.03480729011054676, + "grad_norm": 0.16742883622646332, + "learning_rate": 0.0001994512064946279, + "loss": 0.5851, + "step": 233 + }, + { + "epoch": 0.03495667762175082, + "grad_norm": 0.17511384189128876, + "learning_rate": 0.00019944627810159632, + "loss": 0.569, + "step": 234 + }, + { + "epoch": 0.03510606513295488, + "grad_norm": 0.26125359535217285, + "learning_rate": 0.0001994413277393067, + "loss": 0.862, + "step": 235 + }, + { + "epoch": 0.03525545264415895, + "grad_norm": 0.1439584493637085, + "learning_rate": 0.00019943635540885279, + "loss": 1.1311, + "step": 236 + }, + { + "epoch": 0.035404840155363014, + "grad_norm": 0.146185502409935, + "learning_rate": 0.00019943136111133294, + "loss": 0.9574, + "step": 237 + }, + { + "epoch": 0.03555422766656707, + "grad_norm": 5.1866774559021, + "learning_rate": 0.00019942634484785052, + "loss": 3.2188, + "step": 238 + }, + { + "epoch": 0.03570361517777114, + "grad_norm": 0.13537189364433289, + "learning_rate": 0.00019942130661951372, + "loss": 0.6154, + "step": 239 + }, + { + "epoch": 0.0358530026889752, + "grad_norm": 0.13716812431812286, + "learning_rate": 0.00019941624642743548, + "loss": 0.7604, + "step": 240 + }, + { + "epoch": 0.036002390200179264, + "grad_norm": 0.37018468976020813, + "learning_rate": 0.0001994111642727338, + "loss": 0.8598, + "step": 241 + }, + { + "epoch": 0.03615177771138333, + "grad_norm": 0.1747826188802719, + "learning_rate": 0.0001994060601565313, + "loss": 0.6497, + "step": 242 + }, + { + "epoch": 0.03630116522258739, + "grad_norm": 0.12225501239299774, + "learning_rate": 0.0001994009340799556, + "loss": 0.7278, + "step": 243 + }, + { + "epoch": 0.036450552733791455, + "grad_norm": 0.17295411229133606, + "learning_rate": 0.00019939578604413912, + "loss": 0.8747, + "step": 244 + }, + { + "epoch": 0.03659994024499552, + "grad_norm": 0.1815291941165924, + "learning_rate": 0.00019939061605021917, + "loss": 0.7242, + "step": 245 + }, + { + "epoch": 0.03674932775619958, + "grad_norm": 0.25227025151252747, + "learning_rate": 0.00019938542409933787, + "loss": 0.7818, + "step": 246 + }, + { + "epoch": 0.036898715267403646, + "grad_norm": 0.280819296836853, + "learning_rate": 0.00019938021019264221, + "loss": 0.7471, + "step": 247 + }, + { + "epoch": 0.037048102778607706, + "grad_norm": 0.1746496856212616, + "learning_rate": 0.000199374974331284, + "loss": 0.9149, + "step": 248 + }, + { + "epoch": 0.03719749028981177, + "grad_norm": 0.2506274878978729, + "learning_rate": 0.00019936971651641995, + "loss": 0.8666, + "step": 249 + }, + { + "epoch": 0.03734687780101584, + "grad_norm": 0.12307952344417572, + "learning_rate": 0.00019936443674921158, + "loss": 0.7737, + "step": 250 + }, + { + "epoch": 0.0374962653122199, + "grad_norm": 0.13615377247333527, + "learning_rate": 0.0001993591350308253, + "loss": 0.7592, + "step": 251 + }, + { + "epoch": 0.03764565282342396, + "grad_norm": 0.16808447241783142, + "learning_rate": 0.0001993538113624323, + "loss": 0.8599, + "step": 252 + }, + { + "epoch": 0.03779504033462802, + "grad_norm": 0.11544547975063324, + "learning_rate": 0.00019934846574520872, + "loss": 0.7348, + "step": 253 + }, + { + "epoch": 0.03794442784583209, + "grad_norm": 0.20908010005950928, + "learning_rate": 0.00019934309818033544, + "loss": 0.6674, + "step": 254 + }, + { + "epoch": 0.038093815357036154, + "grad_norm": 0.1379510909318924, + "learning_rate": 0.00019933770866899825, + "loss": 0.7295, + "step": 255 + }, + { + "epoch": 0.03824320286824021, + "grad_norm": 0.15058402717113495, + "learning_rate": 0.0001993322972123878, + "loss": 1.1005, + "step": 256 + }, + { + "epoch": 0.03839259037944428, + "grad_norm": 0.1941765695810318, + "learning_rate": 0.00019932686381169955, + "loss": 0.7658, + "step": 257 + }, + { + "epoch": 0.038541977890648345, + "grad_norm": 0.1889600306749344, + "learning_rate": 0.0001993214084681338, + "loss": 0.9765, + "step": 258 + }, + { + "epoch": 0.038691365401852404, + "grad_norm": 0.1466747522354126, + "learning_rate": 0.00019931593118289578, + "loss": 0.5899, + "step": 259 + }, + { + "epoch": 0.03884075291305647, + "grad_norm": 0.2811049520969391, + "learning_rate": 0.00019931043195719548, + "loss": 0.8255, + "step": 260 + }, + { + "epoch": 0.03899014042426053, + "grad_norm": 0.1345093995332718, + "learning_rate": 0.00019930491079224772, + "loss": 0.9015, + "step": 261 + }, + { + "epoch": 0.039139527935464595, + "grad_norm": 0.9843081831932068, + "learning_rate": 0.00019929936768927232, + "loss": 1.5562, + "step": 262 + }, + { + "epoch": 0.03928891544666866, + "grad_norm": 0.18343736231327057, + "learning_rate": 0.00019929380264949376, + "loss": 0.8783, + "step": 263 + }, + { + "epoch": 0.03943830295787272, + "grad_norm": 0.29955071210861206, + "learning_rate": 0.00019928821567414144, + "loss": 1.0868, + "step": 264 + }, + { + "epoch": 0.039587690469076786, + "grad_norm": 0.16872237622737885, + "learning_rate": 0.00019928260676444965, + "loss": 0.7375, + "step": 265 + }, + { + "epoch": 0.039737077980280845, + "grad_norm": 0.1343865543603897, + "learning_rate": 0.00019927697592165747, + "loss": 1.0279, + "step": 266 + }, + { + "epoch": 0.03988646549148491, + "grad_norm": 0.2587420642375946, + "learning_rate": 0.00019927132314700885, + "loss": 0.8529, + "step": 267 + }, + { + "epoch": 0.04003585300268898, + "grad_norm": 0.23096689581871033, + "learning_rate": 0.00019926564844175256, + "loss": 0.8726, + "step": 268 + }, + { + "epoch": 0.040185240513893036, + "grad_norm": 0.13639822602272034, + "learning_rate": 0.00019925995180714224, + "loss": 0.6957, + "step": 269 + }, + { + "epoch": 0.0403346280250971, + "grad_norm": 0.13751177489757538, + "learning_rate": 0.00019925423324443638, + "loss": 0.7239, + "step": 270 + }, + { + "epoch": 0.04048401553630117, + "grad_norm": 0.12963712215423584, + "learning_rate": 0.0001992484927548983, + "loss": 0.686, + "step": 271 + }, + { + "epoch": 0.04063340304750523, + "grad_norm": 0.13808754086494446, + "learning_rate": 0.00019924273033979613, + "loss": 0.973, + "step": 272 + }, + { + "epoch": 0.04078279055870929, + "grad_norm": 0.10808500647544861, + "learning_rate": 0.0001992369460004029, + "loss": 0.6725, + "step": 273 + }, + { + "epoch": 0.04093217806991335, + "grad_norm": 0.14959432184696198, + "learning_rate": 0.0001992311397379965, + "loss": 0.9753, + "step": 274 + }, + { + "epoch": 0.04108156558111742, + "grad_norm": 0.21491862833499908, + "learning_rate": 0.00019922531155385954, + "loss": 0.9506, + "step": 275 + }, + { + "epoch": 0.041230953092321485, + "grad_norm": 0.16956715285778046, + "learning_rate": 0.00019921946144927966, + "loss": 1.0777, + "step": 276 + }, + { + "epoch": 0.041380340603525544, + "grad_norm": 0.17658768594264984, + "learning_rate": 0.00019921358942554917, + "loss": 0.6994, + "step": 277 + }, + { + "epoch": 0.04152972811472961, + "grad_norm": 0.20894332230091095, + "learning_rate": 0.0001992076954839653, + "loss": 1.0217, + "step": 278 + }, + { + "epoch": 0.04167911562593367, + "grad_norm": 0.15689632296562195, + "learning_rate": 0.00019920177962583015, + "loss": 0.4844, + "step": 279 + }, + { + "epoch": 0.041828503137137735, + "grad_norm": 0.24510236084461212, + "learning_rate": 0.00019919584185245062, + "loss": 0.8981, + "step": 280 + }, + { + "epoch": 0.0419778906483418, + "grad_norm": 0.1385307013988495, + "learning_rate": 0.00019918988216513844, + "loss": 0.9278, + "step": 281 + }, + { + "epoch": 0.04212727815954586, + "grad_norm": 0.15266434848308563, + "learning_rate": 0.00019918390056521018, + "loss": 0.7803, + "step": 282 + }, + { + "epoch": 0.042276665670749926, + "grad_norm": 0.920637309551239, + "learning_rate": 0.00019917789705398728, + "loss": 2.0334, + "step": 283 + }, + { + "epoch": 0.04242605318195399, + "grad_norm": 0.16596724092960358, + "learning_rate": 0.00019917187163279605, + "loss": 0.5195, + "step": 284 + }, + { + "epoch": 0.04257544069315805, + "grad_norm": 0.21556362509727478, + "learning_rate": 0.00019916582430296758, + "loss": 1.0858, + "step": 285 + }, + { + "epoch": 0.04272482820436212, + "grad_norm": 0.13357070088386536, + "learning_rate": 0.00019915975506583778, + "loss": 0.7042, + "step": 286 + }, + { + "epoch": 0.042874215715566176, + "grad_norm": 0.13493870198726654, + "learning_rate": 0.00019915366392274752, + "loss": 0.7329, + "step": 287 + }, + { + "epoch": 0.04302360322677024, + "grad_norm": 0.13088954985141754, + "learning_rate": 0.00019914755087504236, + "loss": 0.6911, + "step": 288 + }, + { + "epoch": 0.04317299073797431, + "grad_norm": 0.12659801542758942, + "learning_rate": 0.0001991414159240728, + "loss": 0.9972, + "step": 289 + }, + { + "epoch": 0.04332237824917837, + "grad_norm": 0.14888939261436462, + "learning_rate": 0.00019913525907119418, + "loss": 0.917, + "step": 290 + }, + { + "epoch": 0.04347176576038243, + "grad_norm": 0.15505638718605042, + "learning_rate": 0.00019912908031776655, + "loss": 0.5189, + "step": 291 + }, + { + "epoch": 0.04362115327158649, + "grad_norm": 0.1389274001121521, + "learning_rate": 0.000199122879665155, + "loss": 0.9571, + "step": 292 + }, + { + "epoch": 0.04377054078279056, + "grad_norm": 0.5212099552154541, + "learning_rate": 0.0001991166571147293, + "loss": 1.7933, + "step": 293 + }, + { + "epoch": 0.043919928293994624, + "grad_norm": 0.15124285221099854, + "learning_rate": 0.0001991104126678641, + "loss": 0.8451, + "step": 294 + }, + { + "epoch": 0.04406931580519868, + "grad_norm": 0.12665098905563354, + "learning_rate": 0.0001991041463259389, + "loss": 0.8399, + "step": 295 + }, + { + "epoch": 0.04421870331640275, + "grad_norm": 0.41863369941711426, + "learning_rate": 0.00019909785809033806, + "loss": 0.6476, + "step": 296 + }, + { + "epoch": 0.044368090827606815, + "grad_norm": 0.12471672147512436, + "learning_rate": 0.00019909154796245076, + "loss": 0.8806, + "step": 297 + }, + { + "epoch": 0.044517478338810874, + "grad_norm": 0.21696336567401886, + "learning_rate": 0.00019908521594367098, + "loss": 1.0237, + "step": 298 + }, + { + "epoch": 0.04466686585001494, + "grad_norm": 0.10489077866077423, + "learning_rate": 0.00019907886203539757, + "loss": 0.5955, + "step": 299 + }, + { + "epoch": 0.044816253361219, + "grad_norm": 0.17574696242809296, + "learning_rate": 0.0001990724862390342, + "loss": 0.669, + "step": 300 + }, + { + "epoch": 0.044965640872423066, + "grad_norm": 0.16877828538417816, + "learning_rate": 0.00019906608855598939, + "loss": 0.8991, + "step": 301 + }, + { + "epoch": 0.04511502838362713, + "grad_norm": 0.16386249661445618, + "learning_rate": 0.0001990596689876765, + "loss": 0.9378, + "step": 302 + }, + { + "epoch": 0.04526441589483119, + "grad_norm": 0.1148744598031044, + "learning_rate": 0.00019905322753551368, + "loss": 0.5595, + "step": 303 + }, + { + "epoch": 0.04541380340603526, + "grad_norm": 0.1777895838022232, + "learning_rate": 0.00019904676420092404, + "loss": 0.9161, + "step": 304 + }, + { + "epoch": 0.045563190917239316, + "grad_norm": 0.14002953469753265, + "learning_rate": 0.0001990402789853353, + "loss": 0.6209, + "step": 305 + }, + { + "epoch": 0.04571257842844338, + "grad_norm": 0.3308524489402771, + "learning_rate": 0.00019903377189018024, + "loss": 0.9947, + "step": 306 + }, + { + "epoch": 0.04586196593964745, + "grad_norm": 0.16171899437904358, + "learning_rate": 0.00019902724291689637, + "loss": 0.9893, + "step": 307 + }, + { + "epoch": 0.04601135345085151, + "grad_norm": 0.1441776603460312, + "learning_rate": 0.000199020692066926, + "loss": 0.8799, + "step": 308 + }, + { + "epoch": 0.04616074096205557, + "grad_norm": 0.12858013808727264, + "learning_rate": 0.00019901411934171638, + "loss": 0.6444, + "step": 309 + }, + { + "epoch": 0.04631012847325964, + "grad_norm": 0.14368069171905518, + "learning_rate": 0.00019900752474271945, + "loss": 0.4694, + "step": 310 + }, + { + "epoch": 0.0464595159844637, + "grad_norm": 0.16599714756011963, + "learning_rate": 0.00019900090827139214, + "loss": 0.7362, + "step": 311 + }, + { + "epoch": 0.046608903495667764, + "grad_norm": 0.18922938406467438, + "learning_rate": 0.0001989942699291961, + "loss": 0.9204, + "step": 312 + }, + { + "epoch": 0.04675829100687182, + "grad_norm": 0.17513027787208557, + "learning_rate": 0.00019898760971759783, + "loss": 0.5341, + "step": 313 + }, + { + "epoch": 0.04690767851807589, + "grad_norm": 0.19131267070770264, + "learning_rate": 0.0001989809276380687, + "loss": 1.0719, + "step": 314 + }, + { + "epoch": 0.047057066029279955, + "grad_norm": 0.1646454632282257, + "learning_rate": 0.00019897422369208488, + "loss": 0.9124, + "step": 315 + }, + { + "epoch": 0.047206453540484014, + "grad_norm": 0.17317859828472137, + "learning_rate": 0.00019896749788112737, + "loss": 0.7388, + "step": 316 + }, + { + "epoch": 0.04735584105168808, + "grad_norm": 0.2774486839771271, + "learning_rate": 0.00019896075020668202, + "loss": 0.8183, + "step": 317 + }, + { + "epoch": 0.04750522856289214, + "grad_norm": 0.14470195770263672, + "learning_rate": 0.0001989539806702395, + "loss": 0.8357, + "step": 318 + }, + { + "epoch": 0.047654616074096205, + "grad_norm": 0.1452193409204483, + "learning_rate": 0.00019894718927329524, + "loss": 0.9126, + "step": 319 + }, + { + "epoch": 0.04780400358530027, + "grad_norm": 0.15406997501850128, + "learning_rate": 0.0001989403760173497, + "loss": 0.6407, + "step": 320 + }, + { + "epoch": 0.04795339109650433, + "grad_norm": 0.2753758728504181, + "learning_rate": 0.00019893354090390791, + "loss": 1.1197, + "step": 321 + }, + { + "epoch": 0.048102778607708396, + "grad_norm": 0.14286638796329498, + "learning_rate": 0.00019892668393447997, + "loss": 0.7519, + "step": 322 + }, + { + "epoch": 0.04825216611891246, + "grad_norm": 0.27783840894699097, + "learning_rate": 0.0001989198051105806, + "loss": 0.8173, + "step": 323 + }, + { + "epoch": 0.04840155363011652, + "grad_norm": 0.14072780311107635, + "learning_rate": 0.00019891290443372944, + "loss": 0.7387, + "step": 324 + }, + { + "epoch": 0.04855094114132059, + "grad_norm": 0.17845518887043, + "learning_rate": 0.00019890598190545102, + "loss": 0.9028, + "step": 325 + }, + { + "epoch": 0.048700328652524646, + "grad_norm": 0.16201937198638916, + "learning_rate": 0.0001988990375272746, + "loss": 0.6775, + "step": 326 + }, + { + "epoch": 0.04884971616372871, + "grad_norm": 0.3842596709728241, + "learning_rate": 0.00019889207130073432, + "loss": 1.166, + "step": 327 + }, + { + "epoch": 0.04899910367493278, + "grad_norm": 0.1811336725950241, + "learning_rate": 0.0001988850832273691, + "loss": 0.5225, + "step": 328 + }, + { + "epoch": 0.04914849118613684, + "grad_norm": 0.1348968744277954, + "learning_rate": 0.0001988780733087228, + "loss": 0.6588, + "step": 329 + }, + { + "epoch": 0.049297878697340904, + "grad_norm": 0.15417031943798065, + "learning_rate": 0.0001988710415463439, + "loss": 1.0178, + "step": 330 + }, + { + "epoch": 0.04944726620854496, + "grad_norm": 0.1732081174850464, + "learning_rate": 0.0001988639879417859, + "loss": 0.9023, + "step": 331 + }, + { + "epoch": 0.04959665371974903, + "grad_norm": 0.20192794501781464, + "learning_rate": 0.00019885691249660702, + "loss": 0.9105, + "step": 332 + }, + { + "epoch": 0.049746041230953095, + "grad_norm": 0.17755641043186188, + "learning_rate": 0.0001988498152123704, + "loss": 0.9222, + "step": 333 + }, + { + "epoch": 0.049895428742157154, + "grad_norm": 0.10902781784534454, + "learning_rate": 0.00019884269609064386, + "loss": 0.6073, + "step": 334 + }, + { + "epoch": 0.05004481625336122, + "grad_norm": 0.19006772339344025, + "learning_rate": 0.00019883555513300019, + "loss": 1.1671, + "step": 335 + }, + { + "epoch": 0.05019420376456528, + "grad_norm": 0.1430109441280365, + "learning_rate": 0.0001988283923410169, + "loss": 0.9732, + "step": 336 + }, + { + "epoch": 0.050343591275769345, + "grad_norm": 0.2234259843826294, + "learning_rate": 0.00019882120771627638, + "loss": 1.3184, + "step": 337 + }, + { + "epoch": 0.05049297878697341, + "grad_norm": 0.1166071966290474, + "learning_rate": 0.00019881400126036582, + "loss": 0.6163, + "step": 338 + }, + { + "epoch": 0.05064236629817747, + "grad_norm": 0.12911243736743927, + "learning_rate": 0.0001988067729748773, + "loss": 0.9934, + "step": 339 + }, + { + "epoch": 0.050791753809381536, + "grad_norm": 0.14424997568130493, + "learning_rate": 0.00019879952286140754, + "loss": 0.6436, + "step": 340 + }, + { + "epoch": 0.0509411413205856, + "grad_norm": 0.17660249769687653, + "learning_rate": 0.00019879225092155834, + "loss": 0.8772, + "step": 341 + }, + { + "epoch": 0.05109052883178966, + "grad_norm": 0.126225546002388, + "learning_rate": 0.0001987849571569361, + "loss": 0.5238, + "step": 342 + }, + { + "epoch": 0.05123991634299373, + "grad_norm": 0.1732265204191208, + "learning_rate": 0.00019877764156915213, + "loss": 0.7972, + "step": 343 + }, + { + "epoch": 0.051389303854197786, + "grad_norm": 0.2713650166988373, + "learning_rate": 0.0001987703041598226, + "loss": 0.9757, + "step": 344 + }, + { + "epoch": 0.05153869136540185, + "grad_norm": 0.1454247683286667, + "learning_rate": 0.00019876294493056845, + "loss": 0.9633, + "step": 345 + }, + { + "epoch": 0.05168807887660592, + "grad_norm": 0.12023693323135376, + "learning_rate": 0.00019875556388301543, + "loss": 0.5894, + "step": 346 + }, + { + "epoch": 0.05183746638780998, + "grad_norm": 0.15992896258831024, + "learning_rate": 0.0001987481610187941, + "loss": 0.7281, + "step": 347 + }, + { + "epoch": 0.05198685389901404, + "grad_norm": 0.15995372831821442, + "learning_rate": 0.00019874073633953997, + "loss": 0.6077, + "step": 348 + }, + { + "epoch": 0.0521362414102181, + "grad_norm": 0.12626518309116364, + "learning_rate": 0.0001987332898468932, + "loss": 0.8621, + "step": 349 + }, + { + "epoch": 0.05228562892142217, + "grad_norm": 0.1364617496728897, + "learning_rate": 0.00019872582154249884, + "loss": 0.7432, + "step": 350 + }, + { + "epoch": 0.052435016432626234, + "grad_norm": 0.17311328649520874, + "learning_rate": 0.00019871833142800675, + "loss": 0.7599, + "step": 351 + }, + { + "epoch": 0.052584403943830293, + "grad_norm": 0.18369294703006744, + "learning_rate": 0.00019871081950507163, + "loss": 0.7234, + "step": 352 + }, + { + "epoch": 0.05273379145503436, + "grad_norm": 0.12449700385332108, + "learning_rate": 0.00019870328577535303, + "loss": 0.8287, + "step": 353 + }, + { + "epoch": 0.052883178966238426, + "grad_norm": 0.10272317379713058, + "learning_rate": 0.00019869573024051517, + "loss": 0.5652, + "step": 354 + }, + { + "epoch": 0.053032566477442485, + "grad_norm": 0.16783328354358673, + "learning_rate": 0.00019868815290222726, + "loss": 0.5231, + "step": 355 + }, + { + "epoch": 0.05318195398864655, + "grad_norm": 0.26311105489730835, + "learning_rate": 0.00019868055376216323, + "loss": 0.6964, + "step": 356 + }, + { + "epoch": 0.05333134149985061, + "grad_norm": 0.32151147723197937, + "learning_rate": 0.00019867293282200188, + "loss": 1.0218, + "step": 357 + }, + { + "epoch": 0.053480729011054676, + "grad_norm": 0.23724707961082458, + "learning_rate": 0.00019866529008342673, + "loss": 1.0303, + "step": 358 + }, + { + "epoch": 0.05363011652225874, + "grad_norm": 0.1722903996706009, + "learning_rate": 0.00019865762554812624, + "loss": 0.9047, + "step": 359 + }, + { + "epoch": 0.0537795040334628, + "grad_norm": 0.20011122524738312, + "learning_rate": 0.00019864993921779361, + "loss": 0.8151, + "step": 360 + }, + { + "epoch": 0.05392889154466687, + "grad_norm": 0.13171933591365814, + "learning_rate": 0.0001986422310941269, + "loss": 0.7425, + "step": 361 + }, + { + "epoch": 0.054078279055870926, + "grad_norm": 0.14532333612442017, + "learning_rate": 0.0001986345011788289, + "loss": 0.8868, + "step": 362 + }, + { + "epoch": 0.05422766656707499, + "grad_norm": 0.17356187105178833, + "learning_rate": 0.00019862674947360729, + "loss": 0.4887, + "step": 363 + }, + { + "epoch": 0.05437705407827906, + "grad_norm": 0.24108925461769104, + "learning_rate": 0.00019861897598017457, + "loss": 1.1633, + "step": 364 + }, + { + "epoch": 0.05452644158948312, + "grad_norm": 0.16384924948215485, + "learning_rate": 0.00019861118070024802, + "loss": 0.8454, + "step": 365 + }, + { + "epoch": 0.05467582910068718, + "grad_norm": 0.1604813188314438, + "learning_rate": 0.00019860336363554973, + "loss": 0.6332, + "step": 366 + }, + { + "epoch": 0.05482521661189125, + "grad_norm": 0.19130012392997742, + "learning_rate": 0.00019859552478780659, + "loss": 0.9221, + "step": 367 + }, + { + "epoch": 0.05497460412309531, + "grad_norm": 0.22768980264663696, + "learning_rate": 0.0001985876641587504, + "loss": 0.9983, + "step": 368 + }, + { + "epoch": 0.055123991634299374, + "grad_norm": 0.13694654405117035, + "learning_rate": 0.0001985797817501176, + "loss": 0.633, + "step": 369 + }, + { + "epoch": 0.05527337914550343, + "grad_norm": 0.1287558525800705, + "learning_rate": 0.00019857187756364958, + "loss": 0.6729, + "step": 370 + }, + { + "epoch": 0.0554227666567075, + "grad_norm": 0.12864962220191956, + "learning_rate": 0.00019856395160109256, + "loss": 0.8324, + "step": 371 + }, + { + "epoch": 0.055572154167911565, + "grad_norm": 0.15732339024543762, + "learning_rate": 0.00019855600386419744, + "loss": 0.7524, + "step": 372 + }, + { + "epoch": 0.055721541679115624, + "grad_norm": 0.19376401603221893, + "learning_rate": 0.00019854803435472, + "loss": 0.9524, + "step": 373 + }, + { + "epoch": 0.05587092919031969, + "grad_norm": 0.17447201907634735, + "learning_rate": 0.00019854004307442088, + "loss": 0.9536, + "step": 374 + }, + { + "epoch": 0.05602031670152375, + "grad_norm": 0.1653999537229538, + "learning_rate": 0.00019853203002506543, + "loss": 0.6763, + "step": 375 + }, + { + "epoch": 0.056169704212727815, + "grad_norm": 0.21575714647769928, + "learning_rate": 0.0001985239952084239, + "loss": 1.2311, + "step": 376 + }, + { + "epoch": 0.05631909172393188, + "grad_norm": 0.18763162195682526, + "learning_rate": 0.0001985159386262713, + "loss": 0.7904, + "step": 377 + }, + { + "epoch": 0.05646847923513594, + "grad_norm": 0.14876295626163483, + "learning_rate": 0.0001985078602803874, + "loss": 0.7818, + "step": 378 + }, + { + "epoch": 0.056617866746340006, + "grad_norm": 0.15621663630008698, + "learning_rate": 0.0001984997601725569, + "loss": 0.8913, + "step": 379 + }, + { + "epoch": 0.05676725425754407, + "grad_norm": 0.154473677277565, + "learning_rate": 0.00019849163830456922, + "loss": 0.5824, + "step": 380 + }, + { + "epoch": 0.05691664176874813, + "grad_norm": 0.17189285159111023, + "learning_rate": 0.00019848349467821864, + "loss": 0.7574, + "step": 381 + }, + { + "epoch": 0.0570660292799522, + "grad_norm": 0.17504605650901794, + "learning_rate": 0.00019847532929530415, + "loss": 0.746, + "step": 382 + }, + { + "epoch": 0.05721541679115626, + "grad_norm": 0.14172236621379852, + "learning_rate": 0.00019846714215762966, + "loss": 0.9479, + "step": 383 + }, + { + "epoch": 0.05736480430236032, + "grad_norm": 0.18013326823711395, + "learning_rate": 0.00019845893326700384, + "loss": 0.7438, + "step": 384 + }, + { + "epoch": 0.05751419181356439, + "grad_norm": 0.12909965217113495, + "learning_rate": 0.00019845070262524016, + "loss": 0.6857, + "step": 385 + }, + { + "epoch": 0.05766357932476845, + "grad_norm": 0.15466244518756866, + "learning_rate": 0.00019844245023415685, + "loss": 0.9448, + "step": 386 + }, + { + "epoch": 0.057812966835972514, + "grad_norm": 0.6459704637527466, + "learning_rate": 0.0001984341760955771, + "loss": 1.7624, + "step": 387 + }, + { + "epoch": 0.05796235434717657, + "grad_norm": 0.12828199565410614, + "learning_rate": 0.0001984258802113287, + "loss": 0.6321, + "step": 388 + }, + { + "epoch": 0.05811174185838064, + "grad_norm": 0.19720108807086945, + "learning_rate": 0.0001984175625832444, + "loss": 0.742, + "step": 389 + }, + { + "epoch": 0.058261129369584705, + "grad_norm": 0.1498088240623474, + "learning_rate": 0.0001984092232131616, + "loss": 0.8326, + "step": 390 + }, + { + "epoch": 0.058410516880788764, + "grad_norm": 0.21480423212051392, + "learning_rate": 0.0001984008621029227, + "loss": 0.7597, + "step": 391 + }, + { + "epoch": 0.05855990439199283, + "grad_norm": 0.29573026299476624, + "learning_rate": 0.0001983924792543748, + "loss": 1.3511, + "step": 392 + }, + { + "epoch": 0.058709291903196896, + "grad_norm": 0.1259469836950302, + "learning_rate": 0.0001983840746693698, + "loss": 0.7294, + "step": 393 + }, + { + "epoch": 0.058858679414400955, + "grad_norm": 0.14026756584644318, + "learning_rate": 0.00019837564834976432, + "loss": 0.6073, + "step": 394 + }, + { + "epoch": 0.05900806692560502, + "grad_norm": 0.1396723836660385, + "learning_rate": 0.00019836720029741995, + "loss": 0.532, + "step": 395 + }, + { + "epoch": 0.05915745443680908, + "grad_norm": 0.33731967210769653, + "learning_rate": 0.000198358730514203, + "loss": 1.0597, + "step": 396 + }, + { + "epoch": 0.059306841948013146, + "grad_norm": 0.14965583384037018, + "learning_rate": 0.00019835023900198454, + "loss": 0.9816, + "step": 397 + }, + { + "epoch": 0.05945622945921721, + "grad_norm": 0.18267478048801422, + "learning_rate": 0.0001983417257626405, + "loss": 0.7945, + "step": 398 + }, + { + "epoch": 0.05960561697042127, + "grad_norm": 0.39048805832862854, + "learning_rate": 0.0001983331907980516, + "loss": 1.1221, + "step": 399 + }, + { + "epoch": 0.05975500448162534, + "grad_norm": 0.11438261717557907, + "learning_rate": 0.00019832463411010331, + "loss": 0.6491, + "step": 400 + }, + { + "epoch": 0.059904391992829396, + "grad_norm": 0.1477927714586258, + "learning_rate": 0.00019831605570068596, + "loss": 0.8062, + "step": 401 + }, + { + "epoch": 0.06005377950403346, + "grad_norm": 0.12166056036949158, + "learning_rate": 0.0001983074555716947, + "loss": 0.8609, + "step": 402 + }, + { + "epoch": 0.06020316701523753, + "grad_norm": 0.13299763202667236, + "learning_rate": 0.00019829883372502935, + "loss": 0.9124, + "step": 403 + }, + { + "epoch": 0.06035255452644159, + "grad_norm": 0.5741954445838928, + "learning_rate": 0.00019829019016259468, + "loss": 1.4774, + "step": 404 + }, + { + "epoch": 0.060501942037645653, + "grad_norm": 0.12422844022512436, + "learning_rate": 0.00019828152488630016, + "loss": 0.7628, + "step": 405 + }, + { + "epoch": 0.06065132954884972, + "grad_norm": 0.22834278643131256, + "learning_rate": 0.00019827283789806011, + "loss": 1.0135, + "step": 406 + }, + { + "epoch": 0.06080071706005378, + "grad_norm": 0.2153419405221939, + "learning_rate": 0.00019826412919979358, + "loss": 1.0723, + "step": 407 + }, + { + "epoch": 0.060950104571257845, + "grad_norm": 0.18992449343204498, + "learning_rate": 0.0001982553987934245, + "loss": 0.5903, + "step": 408 + }, + { + "epoch": 0.061099492082461904, + "grad_norm": 0.14883364737033844, + "learning_rate": 0.00019824664668088155, + "loss": 0.7192, + "step": 409 + }, + { + "epoch": 0.06124887959366597, + "grad_norm": 0.12347893416881561, + "learning_rate": 0.0001982378728640982, + "loss": 0.785, + "step": 410 + }, + { + "epoch": 0.061398267104870036, + "grad_norm": 0.14353862404823303, + "learning_rate": 0.0001982290773450127, + "loss": 0.6293, + "step": 411 + }, + { + "epoch": 0.061547654616074095, + "grad_norm": 0.13001669943332672, + "learning_rate": 0.00019822026012556818, + "loss": 0.7083, + "step": 412 + }, + { + "epoch": 0.06169704212727816, + "grad_norm": 0.15990330278873444, + "learning_rate": 0.00019821142120771246, + "loss": 0.6908, + "step": 413 + }, + { + "epoch": 0.06184642963848222, + "grad_norm": 0.27101776003837585, + "learning_rate": 0.0001982025605933982, + "loss": 0.7311, + "step": 414 + }, + { + "epoch": 0.061995817149686286, + "grad_norm": 0.3568134903907776, + "learning_rate": 0.00019819367828458287, + "loss": 1.0127, + "step": 415 + }, + { + "epoch": 0.06214520466089035, + "grad_norm": 0.13109736144542694, + "learning_rate": 0.0001981847742832287, + "loss": 0.8776, + "step": 416 + }, + { + "epoch": 0.06229459217209441, + "grad_norm": 0.1522371917963028, + "learning_rate": 0.0001981758485913027, + "loss": 0.9513, + "step": 417 + }, + { + "epoch": 0.06244397968329848, + "grad_norm": 0.15717531740665436, + "learning_rate": 0.00019816690121077674, + "loss": 1.084, + "step": 418 + }, + { + "epoch": 0.06259336719450254, + "grad_norm": 0.26732301712036133, + "learning_rate": 0.00019815793214362742, + "loss": 1.0219, + "step": 419 + }, + { + "epoch": 0.0627427547057066, + "grad_norm": 0.1595582216978073, + "learning_rate": 0.00019814894139183614, + "loss": 0.7484, + "step": 420 + }, + { + "epoch": 0.06289214221691067, + "grad_norm": 0.24443097412586212, + "learning_rate": 0.00019813992895738908, + "loss": 1.1702, + "step": 421 + }, + { + "epoch": 0.06304152972811473, + "grad_norm": 0.12789739668369293, + "learning_rate": 0.00019813089484227732, + "loss": 0.9188, + "step": 422 + }, + { + "epoch": 0.06319091723931879, + "grad_norm": 0.13589276373386383, + "learning_rate": 0.00019812183904849653, + "loss": 0.5864, + "step": 423 + }, + { + "epoch": 0.06334030475052285, + "grad_norm": 0.20833753049373627, + "learning_rate": 0.00019811276157804733, + "loss": 0.9182, + "step": 424 + }, + { + "epoch": 0.06348969226172692, + "grad_norm": 0.16936369240283966, + "learning_rate": 0.0001981036624329351, + "loss": 0.7968, + "step": 425 + }, + { + "epoch": 0.06363907977293098, + "grad_norm": 0.21851451694965363, + "learning_rate": 0.00019809454161516993, + "loss": 0.6956, + "step": 426 + }, + { + "epoch": 0.06378846728413505, + "grad_norm": 0.15087386965751648, + "learning_rate": 0.0001980853991267668, + "loss": 1.0246, + "step": 427 + }, + { + "epoch": 0.06393785479533912, + "grad_norm": 0.12114626169204712, + "learning_rate": 0.00019807623496974537, + "loss": 0.7678, + "step": 428 + }, + { + "epoch": 0.06408724230654317, + "grad_norm": 0.20948128402233124, + "learning_rate": 0.00019806704914613024, + "loss": 1.1529, + "step": 429 + }, + { + "epoch": 0.06423662981774723, + "grad_norm": 0.1780691146850586, + "learning_rate": 0.0001980578416579506, + "loss": 0.9683, + "step": 430 + }, + { + "epoch": 0.0643860173289513, + "grad_norm": 0.14259085059165955, + "learning_rate": 0.00019804861250724063, + "loss": 0.4694, + "step": 431 + }, + { + "epoch": 0.06453540484015537, + "grad_norm": 0.2003784030675888, + "learning_rate": 0.00019803936169603912, + "loss": 0.6101, + "step": 432 + }, + { + "epoch": 0.06468479235135943, + "grad_norm": 0.19409967958927155, + "learning_rate": 0.00019803008922638976, + "loss": 1.2219, + "step": 433 + }, + { + "epoch": 0.06483417986256348, + "grad_norm": 0.12308470904827118, + "learning_rate": 0.00019802079510034096, + "loss": 0.8568, + "step": 434 + }, + { + "epoch": 0.06498356737376755, + "grad_norm": 0.3875686228275299, + "learning_rate": 0.00019801147931994596, + "loss": 1.027, + "step": 435 + }, + { + "epoch": 0.06513295488497162, + "grad_norm": 0.21796062588691711, + "learning_rate": 0.00019800214188726276, + "loss": 1.2517, + "step": 436 + }, + { + "epoch": 0.06528234239617568, + "grad_norm": 0.14813588559627533, + "learning_rate": 0.00019799278280435413, + "loss": 0.8185, + "step": 437 + }, + { + "epoch": 0.06543172990737975, + "grad_norm": 0.215727299451828, + "learning_rate": 0.00019798340207328766, + "loss": 1.3439, + "step": 438 + }, + { + "epoch": 0.0655811174185838, + "grad_norm": 0.25380611419677734, + "learning_rate": 0.0001979739996961357, + "loss": 1.2215, + "step": 439 + }, + { + "epoch": 0.06573050492978787, + "grad_norm": 0.16782647371292114, + "learning_rate": 0.00019796457567497537, + "loss": 0.9321, + "step": 440 + }, + { + "epoch": 0.06587989244099193, + "grad_norm": 0.13334587216377258, + "learning_rate": 0.0001979551300118886, + "loss": 0.8722, + "step": 441 + }, + { + "epoch": 0.066029279952196, + "grad_norm": 0.27328580617904663, + "learning_rate": 0.0001979456627089621, + "loss": 0.9013, + "step": 442 + }, + { + "epoch": 0.06617866746340006, + "grad_norm": 0.10864201933145523, + "learning_rate": 0.0001979361737682873, + "loss": 0.6201, + "step": 443 + }, + { + "epoch": 0.06632805497460412, + "grad_norm": 0.1800958216190338, + "learning_rate": 0.0001979266631919605, + "loss": 0.727, + "step": 444 + }, + { + "epoch": 0.06647744248580818, + "grad_norm": 0.14912551641464233, + "learning_rate": 0.00019791713098208272, + "loss": 0.6877, + "step": 445 + }, + { + "epoch": 0.06662682999701225, + "grad_norm": 0.3135051727294922, + "learning_rate": 0.00019790757714075979, + "loss": 0.9517, + "step": 446 + }, + { + "epoch": 0.06677621750821632, + "grad_norm": 0.23976337909698486, + "learning_rate": 0.0001978980016701023, + "loss": 0.9952, + "step": 447 + }, + { + "epoch": 0.06692560501942038, + "grad_norm": 0.16540992259979248, + "learning_rate": 0.00019788840457222556, + "loss": 0.796, + "step": 448 + }, + { + "epoch": 0.06707499253062443, + "grad_norm": 0.19656099379062653, + "learning_rate": 0.00019787878584924984, + "loss": 0.7593, + "step": 449 + }, + { + "epoch": 0.0672243800418285, + "grad_norm": 0.12927167117595673, + "learning_rate": 0.0001978691455033, + "loss": 0.819, + "step": 450 + }, + { + "epoch": 0.06737376755303257, + "grad_norm": 0.19154123961925507, + "learning_rate": 0.00019785948353650572, + "loss": 0.9713, + "step": 451 + }, + { + "epoch": 0.06752315506423663, + "grad_norm": 0.12065298110246658, + "learning_rate": 0.0001978497999510015, + "loss": 0.7942, + "step": 452 + }, + { + "epoch": 0.0676725425754407, + "grad_norm": 0.1475251317024231, + "learning_rate": 0.00019784009474892666, + "loss": 0.8598, + "step": 453 + }, + { + "epoch": 0.06782193008664476, + "grad_norm": 0.14072950184345245, + "learning_rate": 0.00019783036793242516, + "loss": 0.8956, + "step": 454 + }, + { + "epoch": 0.06797131759784882, + "grad_norm": 0.20561757683753967, + "learning_rate": 0.00019782061950364584, + "loss": 0.9214, + "step": 455 + }, + { + "epoch": 0.06812070510905288, + "grad_norm": 0.1547221839427948, + "learning_rate": 0.00019781084946474226, + "loss": 0.9042, + "step": 456 + }, + { + "epoch": 0.06827009262025695, + "grad_norm": 0.233870729804039, + "learning_rate": 0.0001978010578178728, + "loss": 1.0727, + "step": 457 + }, + { + "epoch": 0.06841948013146101, + "grad_norm": 0.19565890729427338, + "learning_rate": 0.00019779124456520056, + "loss": 0.8425, + "step": 458 + }, + { + "epoch": 0.06856886764266508, + "grad_norm": 0.15683157742023468, + "learning_rate": 0.00019778140970889348, + "loss": 0.7989, + "step": 459 + }, + { + "epoch": 0.06871825515386913, + "grad_norm": 0.12861517071723938, + "learning_rate": 0.0001977715532511242, + "loss": 0.8849, + "step": 460 + }, + { + "epoch": 0.0688676426650732, + "grad_norm": 0.12733756005764008, + "learning_rate": 0.00019776167519407022, + "loss": 0.7984, + "step": 461 + }, + { + "epoch": 0.06901703017627726, + "grad_norm": 0.14494583010673523, + "learning_rate": 0.0001977517755399137, + "loss": 0.8234, + "step": 462 + }, + { + "epoch": 0.06916641768748133, + "grad_norm": 0.15694382786750793, + "learning_rate": 0.00019774185429084166, + "loss": 0.8476, + "step": 463 + }, + { + "epoch": 0.0693158051986854, + "grad_norm": 0.2537318766117096, + "learning_rate": 0.00019773191144904586, + "loss": 0.9928, + "step": 464 + }, + { + "epoch": 0.06946519270988945, + "grad_norm": 0.45019641518592834, + "learning_rate": 0.0001977219470167228, + "loss": 1.5181, + "step": 465 + }, + { + "epoch": 0.06961458022109351, + "grad_norm": 0.18655377626419067, + "learning_rate": 0.00019771196099607386, + "loss": 0.7463, + "step": 466 + }, + { + "epoch": 0.06976396773229758, + "grad_norm": 0.1978912502527237, + "learning_rate": 0.00019770195338930503, + "loss": 1.0358, + "step": 467 + }, + { + "epoch": 0.06991335524350165, + "grad_norm": 0.15748754143714905, + "learning_rate": 0.00019769192419862716, + "loss": 0.6154, + "step": 468 + }, + { + "epoch": 0.07006274275470571, + "grad_norm": 0.17752663791179657, + "learning_rate": 0.00019768187342625592, + "loss": 0.6289, + "step": 469 + }, + { + "epoch": 0.07021213026590976, + "grad_norm": 0.12329550832509995, + "learning_rate": 0.0001976718010744116, + "loss": 0.9018, + "step": 470 + }, + { + "epoch": 0.07036151777711383, + "grad_norm": 0.17692221701145172, + "learning_rate": 0.00019766170714531937, + "loss": 1.0348, + "step": 471 + }, + { + "epoch": 0.0705109052883179, + "grad_norm": 0.2912501096725464, + "learning_rate": 0.00019765159164120916, + "loss": 0.9097, + "step": 472 + }, + { + "epoch": 0.07066029279952196, + "grad_norm": 0.118867889046669, + "learning_rate": 0.00019764145456431566, + "loss": 0.6933, + "step": 473 + }, + { + "epoch": 0.07080968031072603, + "grad_norm": 0.15778154134750366, + "learning_rate": 0.00019763129591687827, + "loss": 0.5504, + "step": 474 + }, + { + "epoch": 0.07095906782193008, + "grad_norm": 0.14886140823364258, + "learning_rate": 0.00019762111570114122, + "loss": 0.8209, + "step": 475 + }, + { + "epoch": 0.07110845533313415, + "grad_norm": 0.10714101046323776, + "learning_rate": 0.00019761091391935347, + "loss": 0.6473, + "step": 476 + }, + { + "epoch": 0.07125784284433821, + "grad_norm": 0.15396277606487274, + "learning_rate": 0.00019760069057376875, + "loss": 0.8179, + "step": 477 + }, + { + "epoch": 0.07140723035554228, + "grad_norm": 0.1531289666891098, + "learning_rate": 0.00019759044566664558, + "loss": 0.9508, + "step": 478 + }, + { + "epoch": 0.07155661786674634, + "grad_norm": 0.11772772669792175, + "learning_rate": 0.0001975801792002472, + "loss": 0.6606, + "step": 479 + }, + { + "epoch": 0.0717060053779504, + "grad_norm": 0.16291484236717224, + "learning_rate": 0.00019756989117684164, + "loss": 0.6476, + "step": 480 + }, + { + "epoch": 0.07185539288915446, + "grad_norm": 0.15869931876659393, + "learning_rate": 0.00019755958159870172, + "loss": 0.923, + "step": 481 + }, + { + "epoch": 0.07200478040035853, + "grad_norm": 0.1591719686985016, + "learning_rate": 0.00019754925046810493, + "loss": 0.7225, + "step": 482 + }, + { + "epoch": 0.0721541679115626, + "grad_norm": 0.1619904637336731, + "learning_rate": 0.00019753889778733363, + "loss": 0.7185, + "step": 483 + }, + { + "epoch": 0.07230355542276666, + "grad_norm": 0.12438057363033295, + "learning_rate": 0.00019752852355867486, + "loss": 0.7659, + "step": 484 + }, + { + "epoch": 0.07245294293397073, + "grad_norm": 0.11964958906173706, + "learning_rate": 0.00019751812778442046, + "loss": 0.7564, + "step": 485 + }, + { + "epoch": 0.07260233044517478, + "grad_norm": 0.2836957573890686, + "learning_rate": 0.00019750771046686704, + "loss": 1.4225, + "step": 486 + }, + { + "epoch": 0.07275171795637884, + "grad_norm": 0.13266032934188843, + "learning_rate": 0.00019749727160831593, + "loss": 0.7952, + "step": 487 + }, + { + "epoch": 0.07290110546758291, + "grad_norm": 0.12412168830633163, + "learning_rate": 0.00019748681121107325, + "loss": 0.9047, + "step": 488 + }, + { + "epoch": 0.07305049297878698, + "grad_norm": 0.20590822398662567, + "learning_rate": 0.00019747632927744982, + "loss": 0.7496, + "step": 489 + }, + { + "epoch": 0.07319988048999104, + "grad_norm": 0.18052905797958374, + "learning_rate": 0.00019746582580976136, + "loss": 0.6724, + "step": 490 + }, + { + "epoch": 0.0733492680011951, + "grad_norm": 0.24314431846141815, + "learning_rate": 0.0001974553008103282, + "loss": 0.8552, + "step": 491 + }, + { + "epoch": 0.07349865551239916, + "grad_norm": 0.14932356774806976, + "learning_rate": 0.00019744475428147546, + "loss": 0.8878, + "step": 492 + }, + { + "epoch": 0.07364804302360323, + "grad_norm": 0.12878401577472687, + "learning_rate": 0.00019743418622553303, + "loss": 0.7097, + "step": 493 + }, + { + "epoch": 0.07379743053480729, + "grad_norm": 0.1838715523481369, + "learning_rate": 0.00019742359664483563, + "loss": 1.0581, + "step": 494 + }, + { + "epoch": 0.07394681804601136, + "grad_norm": 0.13782045245170593, + "learning_rate": 0.0001974129855417226, + "loss": 1.0439, + "step": 495 + }, + { + "epoch": 0.07409620555721541, + "grad_norm": 0.179475337266922, + "learning_rate": 0.00019740235291853812, + "loss": 0.7708, + "step": 496 + }, + { + "epoch": 0.07424559306841948, + "grad_norm": 0.1903340369462967, + "learning_rate": 0.0001973916987776311, + "loss": 1.3356, + "step": 497 + }, + { + "epoch": 0.07439498057962354, + "grad_norm": 0.13480018079280853, + "learning_rate": 0.00019738102312135523, + "loss": 0.5876, + "step": 498 + }, + { + "epoch": 0.07454436809082761, + "grad_norm": 0.10856112837791443, + "learning_rate": 0.0001973703259520689, + "loss": 0.6991, + "step": 499 + }, + { + "epoch": 0.07469375560203168, + "grad_norm": 0.3036371171474457, + "learning_rate": 0.0001973596072721353, + "loss": 1.0438, + "step": 500 + }, + { + "epoch": 0.07484314311323573, + "grad_norm": 0.12173090875148773, + "learning_rate": 0.0001973488670839224, + "loss": 0.5836, + "step": 501 + }, + { + "epoch": 0.0749925306244398, + "grad_norm": 0.2667033076286316, + "learning_rate": 0.00019733810538980281, + "loss": 0.49, + "step": 502 + }, + { + "epoch": 0.07514191813564386, + "grad_norm": 0.092626191675663, + "learning_rate": 0.00019732732219215397, + "loss": 0.3618, + "step": 503 + }, + { + "epoch": 0.07529130564684793, + "grad_norm": 0.12614378333091736, + "learning_rate": 0.0001973165174933581, + "loss": 0.8307, + "step": 504 + }, + { + "epoch": 0.07544069315805199, + "grad_norm": 0.14259332418441772, + "learning_rate": 0.00019730569129580206, + "loss": 0.8972, + "step": 505 + }, + { + "epoch": 0.07559008066925604, + "grad_norm": 0.12121133506298065, + "learning_rate": 0.0001972948436018776, + "loss": 0.8446, + "step": 506 + }, + { + "epoch": 0.07573946818046011, + "grad_norm": 0.16819995641708374, + "learning_rate": 0.00019728397441398112, + "loss": 0.8122, + "step": 507 + }, + { + "epoch": 0.07588885569166418, + "grad_norm": 0.14065992832183838, + "learning_rate": 0.00019727308373451377, + "loss": 0.457, + "step": 508 + }, + { + "epoch": 0.07603824320286824, + "grad_norm": 0.16299694776535034, + "learning_rate": 0.0001972621715658815, + "loss": 0.878, + "step": 509 + }, + { + "epoch": 0.07618763071407231, + "grad_norm": 0.1941678524017334, + "learning_rate": 0.000197251237910495, + "loss": 0.9534, + "step": 510 + }, + { + "epoch": 0.07633701822527637, + "grad_norm": 0.14302954077720642, + "learning_rate": 0.00019724028277076964, + "loss": 0.8577, + "step": 511 + }, + { + "epoch": 0.07648640573648043, + "grad_norm": 0.19309553503990173, + "learning_rate": 0.00019722930614912563, + "loss": 0.9864, + "step": 512 + }, + { + "epoch": 0.07663579324768449, + "grad_norm": 0.1626858413219452, + "learning_rate": 0.00019721830804798787, + "loss": 0.8104, + "step": 513 + }, + { + "epoch": 0.07678518075888856, + "grad_norm": 0.11680503189563751, + "learning_rate": 0.00019720728846978598, + "loss": 0.7225, + "step": 514 + }, + { + "epoch": 0.07693456827009262, + "grad_norm": 0.11943230032920837, + "learning_rate": 0.0001971962474169544, + "loss": 0.895, + "step": 515 + }, + { + "epoch": 0.07708395578129669, + "grad_norm": 0.2065763920545578, + "learning_rate": 0.00019718518489193225, + "loss": 0.7482, + "step": 516 + }, + { + "epoch": 0.07723334329250074, + "grad_norm": 0.13309244811534882, + "learning_rate": 0.0001971741008971634, + "loss": 0.5227, + "step": 517 + }, + { + "epoch": 0.07738273080370481, + "grad_norm": 0.13642869889736176, + "learning_rate": 0.00019716299543509654, + "loss": 0.6441, + "step": 518 + }, + { + "epoch": 0.07753211831490887, + "grad_norm": 0.18893510103225708, + "learning_rate": 0.00019715186850818498, + "loss": 0.7619, + "step": 519 + }, + { + "epoch": 0.07768150582611294, + "grad_norm": 0.3364134132862091, + "learning_rate": 0.00019714072011888686, + "loss": 1.0618, + "step": 520 + }, + { + "epoch": 0.077830893337317, + "grad_norm": 0.1486874371767044, + "learning_rate": 0.00019712955026966506, + "loss": 0.7846, + "step": 521 + }, + { + "epoch": 0.07798028084852106, + "grad_norm": 0.21529722213745117, + "learning_rate": 0.00019711835896298713, + "loss": 0.4896, + "step": 522 + }, + { + "epoch": 0.07812966835972512, + "grad_norm": 0.1832679957151413, + "learning_rate": 0.00019710714620132546, + "loss": 0.8258, + "step": 523 + }, + { + "epoch": 0.07827905587092919, + "grad_norm": 0.189329594373703, + "learning_rate": 0.00019709591198715707, + "loss": 0.9139, + "step": 524 + }, + { + "epoch": 0.07842844338213326, + "grad_norm": 0.13310536742210388, + "learning_rate": 0.0001970846563229638, + "loss": 0.9531, + "step": 525 + }, + { + "epoch": 0.07857783089333732, + "grad_norm": 0.5338667631149292, + "learning_rate": 0.00019707337921123221, + "loss": 1.0734, + "step": 526 + }, + { + "epoch": 0.07872721840454137, + "grad_norm": 0.6397843360900879, + "learning_rate": 0.0001970620806544536, + "loss": 1.4513, + "step": 527 + }, + { + "epoch": 0.07887660591574544, + "grad_norm": 0.14354926347732544, + "learning_rate": 0.00019705076065512398, + "loss": 0.96, + "step": 528 + }, + { + "epoch": 0.0790259934269495, + "grad_norm": 0.11550958454608917, + "learning_rate": 0.00019703941921574413, + "loss": 0.7829, + "step": 529 + }, + { + "epoch": 0.07917538093815357, + "grad_norm": 0.14045141637325287, + "learning_rate": 0.00019702805633881957, + "loss": 0.6976, + "step": 530 + }, + { + "epoch": 0.07932476844935764, + "grad_norm": 0.18493744730949402, + "learning_rate": 0.00019701667202686048, + "loss": 1.1065, + "step": 531 + }, + { + "epoch": 0.07947415596056169, + "grad_norm": 0.2148275524377823, + "learning_rate": 0.0001970052662823819, + "loss": 1.1823, + "step": 532 + }, + { + "epoch": 0.07962354347176576, + "grad_norm": 0.1770048886537552, + "learning_rate": 0.0001969938391079035, + "loss": 0.8534, + "step": 533 + }, + { + "epoch": 0.07977293098296982, + "grad_norm": 0.11045973002910614, + "learning_rate": 0.00019698239050594977, + "loss": 0.628, + "step": 534 + }, + { + "epoch": 0.07992231849417389, + "grad_norm": 0.113344706594944, + "learning_rate": 0.0001969709204790498, + "loss": 0.7655, + "step": 535 + }, + { + "epoch": 0.08007170600537795, + "grad_norm": 0.13493701815605164, + "learning_rate": 0.0001969594290297376, + "loss": 0.6173, + "step": 536 + }, + { + "epoch": 0.08022109351658202, + "grad_norm": 0.10995621234178543, + "learning_rate": 0.00019694791616055177, + "loss": 0.8091, + "step": 537 + }, + { + "epoch": 0.08037048102778607, + "grad_norm": 0.1691836416721344, + "learning_rate": 0.00019693638187403563, + "loss": 0.9478, + "step": 538 + }, + { + "epoch": 0.08051986853899014, + "grad_norm": 0.1570545732975006, + "learning_rate": 0.0001969248261727374, + "loss": 0.8514, + "step": 539 + }, + { + "epoch": 0.0806692560501942, + "grad_norm": 0.14363858103752136, + "learning_rate": 0.00019691324905920984, + "loss": 0.5037, + "step": 540 + }, + { + "epoch": 0.08081864356139827, + "grad_norm": 0.12334268540143967, + "learning_rate": 0.00019690165053601056, + "loss": 0.9084, + "step": 541 + }, + { + "epoch": 0.08096803107260234, + "grad_norm": 0.3128170073032379, + "learning_rate": 0.0001968900306057018, + "loss": 0.737, + "step": 542 + }, + { + "epoch": 0.08111741858380639, + "grad_norm": 0.16364172101020813, + "learning_rate": 0.00019687838927085066, + "loss": 0.4789, + "step": 543 + }, + { + "epoch": 0.08126680609501045, + "grad_norm": 0.15874239802360535, + "learning_rate": 0.0001968667265340288, + "loss": 1.0205, + "step": 544 + }, + { + "epoch": 0.08141619360621452, + "grad_norm": 0.13250084221363068, + "learning_rate": 0.00019685504239781278, + "loss": 0.8377, + "step": 545 + }, + { + "epoch": 0.08156558111741859, + "grad_norm": 0.13961811363697052, + "learning_rate": 0.00019684333686478383, + "loss": 0.857, + "step": 546 + }, + { + "epoch": 0.08171496862862265, + "grad_norm": 0.1638427972793579, + "learning_rate": 0.0001968316099375278, + "loss": 0.6918, + "step": 547 + }, + { + "epoch": 0.0818643561398267, + "grad_norm": 0.260032057762146, + "learning_rate": 0.00019681986161863542, + "loss": 0.965, + "step": 548 + }, + { + "epoch": 0.08201374365103077, + "grad_norm": 0.12642204761505127, + "learning_rate": 0.00019680809191070203, + "loss": 0.5236, + "step": 549 + }, + { + "epoch": 0.08216313116223484, + "grad_norm": 0.18395595252513885, + "learning_rate": 0.00019679630081632782, + "loss": 1.0722, + "step": 550 + }, + { + "epoch": 0.0823125186734389, + "grad_norm": 0.14553199708461761, + "learning_rate": 0.0001967844883381176, + "loss": 0.7844, + "step": 551 + }, + { + "epoch": 0.08246190618464297, + "grad_norm": 0.13094641268253326, + "learning_rate": 0.00019677265447868086, + "loss": 0.5937, + "step": 552 + }, + { + "epoch": 0.08261129369584702, + "grad_norm": 0.1621766984462738, + "learning_rate": 0.00019676079924063196, + "loss": 0.7064, + "step": 553 + }, + { + "epoch": 0.08276068120705109, + "grad_norm": 0.16490310430526733, + "learning_rate": 0.0001967489226265899, + "loss": 0.654, + "step": 554 + }, + { + "epoch": 0.08291006871825515, + "grad_norm": 0.1863924264907837, + "learning_rate": 0.00019673702463917842, + "loss": 0.8091, + "step": 555 + }, + { + "epoch": 0.08305945622945922, + "grad_norm": 0.1565409004688263, + "learning_rate": 0.00019672510528102597, + "loss": 0.583, + "step": 556 + }, + { + "epoch": 0.08320884374066329, + "grad_norm": 0.2025175541639328, + "learning_rate": 0.0001967131645547657, + "loss": 0.8749, + "step": 557 + }, + { + "epoch": 0.08335823125186734, + "grad_norm": 0.23709741234779358, + "learning_rate": 0.0001967012024630355, + "loss": 1.1301, + "step": 558 + }, + { + "epoch": 0.0835076187630714, + "grad_norm": 0.33181461691856384, + "learning_rate": 0.00019668921900847805, + "loss": 1.0527, + "step": 559 + }, + { + "epoch": 0.08365700627427547, + "grad_norm": 0.11158733814954758, + "learning_rate": 0.00019667721419374065, + "loss": 0.6433, + "step": 560 + }, + { + "epoch": 0.08380639378547954, + "grad_norm": 0.21710413694381714, + "learning_rate": 0.00019666518802147534, + "loss": 0.7117, + "step": 561 + }, + { + "epoch": 0.0839557812966836, + "grad_norm": 0.1480075567960739, + "learning_rate": 0.00019665314049433888, + "loss": 0.8019, + "step": 562 + }, + { + "epoch": 0.08410516880788767, + "grad_norm": 0.15463611483573914, + "learning_rate": 0.00019664107161499277, + "loss": 0.679, + "step": 563 + }, + { + "epoch": 0.08425455631909172, + "grad_norm": 0.13593655824661255, + "learning_rate": 0.00019662898138610323, + "loss": 0.7527, + "step": 564 + }, + { + "epoch": 0.08440394383029579, + "grad_norm": 0.1662757396697998, + "learning_rate": 0.0001966168698103412, + "loss": 0.7486, + "step": 565 + }, + { + "epoch": 0.08455333134149985, + "grad_norm": 0.10743851959705353, + "learning_rate": 0.00019660473689038228, + "loss": 0.6341, + "step": 566 + }, + { + "epoch": 0.08470271885270392, + "grad_norm": 0.15268459916114807, + "learning_rate": 0.00019659258262890683, + "loss": 0.8236, + "step": 567 + }, + { + "epoch": 0.08485210636390798, + "grad_norm": 0.14038819074630737, + "learning_rate": 0.00019658040702859997, + "loss": 0.8065, + "step": 568 + }, + { + "epoch": 0.08500149387511204, + "grad_norm": 0.15066532790660858, + "learning_rate": 0.0001965682100921514, + "loss": 0.9532, + "step": 569 + }, + { + "epoch": 0.0851508813863161, + "grad_norm": 0.1580052673816681, + "learning_rate": 0.00019655599182225565, + "loss": 1.0969, + "step": 570 + }, + { + "epoch": 0.08530026889752017, + "grad_norm": 0.16247299313545227, + "learning_rate": 0.000196543752221612, + "loss": 0.791, + "step": 571 + }, + { + "epoch": 0.08544965640872423, + "grad_norm": 0.1083194687962532, + "learning_rate": 0.00019653149129292426, + "loss": 0.6203, + "step": 572 + }, + { + "epoch": 0.0855990439199283, + "grad_norm": 0.19076959788799286, + "learning_rate": 0.0001965192090389011, + "loss": 0.6709, + "step": 573 + }, + { + "epoch": 0.08574843143113235, + "grad_norm": 0.15673895180225372, + "learning_rate": 0.00019650690546225592, + "loss": 0.7816, + "step": 574 + }, + { + "epoch": 0.08589781894233642, + "grad_norm": 0.13381899893283844, + "learning_rate": 0.00019649458056570672, + "loss": 0.8227, + "step": 575 + }, + { + "epoch": 0.08604720645354048, + "grad_norm": 0.10087165981531143, + "learning_rate": 0.00019648223435197627, + "loss": 0.5839, + "step": 576 + }, + { + "epoch": 0.08619659396474455, + "grad_norm": 0.17345178127288818, + "learning_rate": 0.00019646986682379206, + "loss": 0.8435, + "step": 577 + }, + { + "epoch": 0.08634598147594862, + "grad_norm": 0.16065159440040588, + "learning_rate": 0.00019645747798388628, + "loss": 0.8124, + "step": 578 + }, + { + "epoch": 0.08649536898715267, + "grad_norm": 0.32011693716049194, + "learning_rate": 0.0001964450678349958, + "loss": 0.8119, + "step": 579 + }, + { + "epoch": 0.08664475649835673, + "grad_norm": 0.13616526126861572, + "learning_rate": 0.0001964326363798622, + "loss": 0.8056, + "step": 580 + }, + { + "epoch": 0.0867941440095608, + "grad_norm": 0.28961271047592163, + "learning_rate": 0.00019642018362123182, + "loss": 1.0182, + "step": 581 + }, + { + "epoch": 0.08694353152076487, + "grad_norm": 0.16200341284275055, + "learning_rate": 0.00019640770956185567, + "loss": 0.9604, + "step": 582 + }, + { + "epoch": 0.08709291903196893, + "grad_norm": 0.1426459699869156, + "learning_rate": 0.00019639521420448947, + "loss": 0.6555, + "step": 583 + }, + { + "epoch": 0.08724230654317298, + "grad_norm": 0.11389955133199692, + "learning_rate": 0.0001963826975518936, + "loss": 0.6878, + "step": 584 + }, + { + "epoch": 0.08739169405437705, + "grad_norm": 0.12384461611509323, + "learning_rate": 0.00019637015960683322, + "loss": 0.612, + "step": 585 + }, + { + "epoch": 0.08754108156558112, + "grad_norm": 0.1436391919851303, + "learning_rate": 0.00019635760037207817, + "loss": 0.7069, + "step": 586 + }, + { + "epoch": 0.08769046907678518, + "grad_norm": 0.1199880838394165, + "learning_rate": 0.00019634501985040296, + "loss": 0.7547, + "step": 587 + }, + { + "epoch": 0.08783985658798925, + "grad_norm": 0.16157324612140656, + "learning_rate": 0.00019633241804458687, + "loss": 0.7839, + "step": 588 + }, + { + "epoch": 0.0879892440991933, + "grad_norm": 0.16362355649471283, + "learning_rate": 0.00019631979495741378, + "loss": 0.7197, + "step": 589 + }, + { + "epoch": 0.08813863161039737, + "grad_norm": 0.1794954091310501, + "learning_rate": 0.00019630715059167238, + "loss": 0.6936, + "step": 590 + }, + { + "epoch": 0.08828801912160143, + "grad_norm": 0.3147851228713989, + "learning_rate": 0.00019629448495015597, + "loss": 1.8882, + "step": 591 + }, + { + "epoch": 0.0884374066328055, + "grad_norm": 0.1880149394273758, + "learning_rate": 0.0001962817980356626, + "loss": 0.7354, + "step": 592 + }, + { + "epoch": 0.08858679414400956, + "grad_norm": 0.13019341230392456, + "learning_rate": 0.00019626908985099503, + "loss": 0.9714, + "step": 593 + }, + { + "epoch": 0.08873618165521363, + "grad_norm": 0.18782681226730347, + "learning_rate": 0.0001962563603989607, + "loss": 0.9334, + "step": 594 + }, + { + "epoch": 0.08888556916641768, + "grad_norm": 0.1618777960538864, + "learning_rate": 0.00019624360968237172, + "loss": 0.8143, + "step": 595 + }, + { + "epoch": 0.08903495667762175, + "grad_norm": 0.13201208412647247, + "learning_rate": 0.00019623083770404492, + "loss": 0.9077, + "step": 596 + }, + { + "epoch": 0.08918434418882581, + "grad_norm": 0.11902808398008347, + "learning_rate": 0.0001962180444668019, + "loss": 1.014, + "step": 597 + }, + { + "epoch": 0.08933373170002988, + "grad_norm": 0.17474393546581268, + "learning_rate": 0.0001962052299734688, + "loss": 0.8466, + "step": 598 + }, + { + "epoch": 0.08948311921123395, + "grad_norm": 0.2875848412513733, + "learning_rate": 0.00019619239422687663, + "loss": 0.8023, + "step": 599 + }, + { + "epoch": 0.089632506722438, + "grad_norm": 0.13769172132015228, + "learning_rate": 0.00019617953722986096, + "loss": 0.9026, + "step": 600 + }, + { + "epoch": 0.08978189423364207, + "grad_norm": 0.30760055780410767, + "learning_rate": 0.00019616665898526206, + "loss": 1.0103, + "step": 601 + }, + { + "epoch": 0.08993128174484613, + "grad_norm": 0.5774983763694763, + "learning_rate": 0.00019615375949592504, + "loss": 1.4516, + "step": 602 + }, + { + "epoch": 0.0900806692560502, + "grad_norm": 0.13711205124855042, + "learning_rate": 0.00019614083876469954, + "loss": 0.7543, + "step": 603 + }, + { + "epoch": 0.09023005676725426, + "grad_norm": 0.1261071413755417, + "learning_rate": 0.00019612789679443997, + "loss": 0.6952, + "step": 604 + }, + { + "epoch": 0.09037944427845832, + "grad_norm": 0.13197720050811768, + "learning_rate": 0.00019611493358800538, + "loss": 0.906, + "step": 605 + }, + { + "epoch": 0.09052883178966238, + "grad_norm": 0.12066524475812912, + "learning_rate": 0.00019610194914825962, + "loss": 0.7621, + "step": 606 + }, + { + "epoch": 0.09067821930086645, + "grad_norm": 0.1722782999277115, + "learning_rate": 0.00019608894347807108, + "loss": 0.8798, + "step": 607 + }, + { + "epoch": 0.09082760681207051, + "grad_norm": 0.1058247983455658, + "learning_rate": 0.000196075916580313, + "loss": 0.6246, + "step": 608 + }, + { + "epoch": 0.09097699432327458, + "grad_norm": 0.17862388491630554, + "learning_rate": 0.00019606286845786315, + "loss": 0.6607, + "step": 609 + }, + { + "epoch": 0.09112638183447863, + "grad_norm": 0.1704971194267273, + "learning_rate": 0.0001960497991136041, + "loss": 0.8648, + "step": 610 + }, + { + "epoch": 0.0912757693456827, + "grad_norm": 0.12244229018688202, + "learning_rate": 0.00019603670855042308, + "loss": 0.7546, + "step": 611 + }, + { + "epoch": 0.09142515685688676, + "grad_norm": 0.1924651712179184, + "learning_rate": 0.00019602359677121199, + "loss": 0.7931, + "step": 612 + }, + { + "epoch": 0.09157454436809083, + "grad_norm": 0.1722334921360016, + "learning_rate": 0.00019601046377886746, + "loss": 1.259, + "step": 613 + }, + { + "epoch": 0.0917239318792949, + "grad_norm": 0.11828108131885529, + "learning_rate": 0.0001959973095762907, + "loss": 0.8418, + "step": 614 + }, + { + "epoch": 0.09187331939049895, + "grad_norm": 0.10462518781423569, + "learning_rate": 0.0001959841341663878, + "loss": 0.5707, + "step": 615 + }, + { + "epoch": 0.09202270690170301, + "grad_norm": 0.14686742424964905, + "learning_rate": 0.00019597093755206936, + "loss": 1.0242, + "step": 616 + }, + { + "epoch": 0.09217209441290708, + "grad_norm": 0.1107390895485878, + "learning_rate": 0.00019595771973625068, + "loss": 0.6829, + "step": 617 + }, + { + "epoch": 0.09232148192411115, + "grad_norm": 0.18590034544467926, + "learning_rate": 0.00019594448072185182, + "loss": 0.7426, + "step": 618 + }, + { + "epoch": 0.09247086943531521, + "grad_norm": 0.12690337002277374, + "learning_rate": 0.00019593122051179748, + "loss": 0.978, + "step": 619 + }, + { + "epoch": 0.09262025694651928, + "grad_norm": 0.19712156057357788, + "learning_rate": 0.00019591793910901707, + "loss": 0.8153, + "step": 620 + }, + { + "epoch": 0.09276964445772333, + "grad_norm": 0.16727258265018463, + "learning_rate": 0.00019590463651644464, + "loss": 0.8827, + "step": 621 + }, + { + "epoch": 0.0929190319689274, + "grad_norm": 0.21075357496738434, + "learning_rate": 0.00019589131273701894, + "loss": 0.624, + "step": 622 + }, + { + "epoch": 0.09306841948013146, + "grad_norm": 0.5324035286903381, + "learning_rate": 0.00019587796777368347, + "loss": 1.808, + "step": 623 + }, + { + "epoch": 0.09321780699133553, + "grad_norm": 0.12788911163806915, + "learning_rate": 0.00019586460162938622, + "loss": 0.8374, + "step": 624 + }, + { + "epoch": 0.0933671945025396, + "grad_norm": 0.1606101393699646, + "learning_rate": 0.00019585121430708012, + "loss": 0.6427, + "step": 625 + }, + { + "epoch": 0.09351658201374365, + "grad_norm": 0.1087862104177475, + "learning_rate": 0.00019583780580972253, + "loss": 0.5977, + "step": 626 + }, + { + "epoch": 0.09366596952494771, + "grad_norm": 0.1577569991350174, + "learning_rate": 0.00019582437614027565, + "loss": 0.7819, + "step": 627 + }, + { + "epoch": 0.09381535703615178, + "grad_norm": 0.17475225031375885, + "learning_rate": 0.00019581092530170633, + "loss": 1.0605, + "step": 628 + }, + { + "epoch": 0.09396474454735584, + "grad_norm": 0.12946583330631256, + "learning_rate": 0.000195797453296986, + "loss": 0.832, + "step": 629 + }, + { + "epoch": 0.09411413205855991, + "grad_norm": 0.14866258203983307, + "learning_rate": 0.00019578396012909092, + "loss": 0.7355, + "step": 630 + }, + { + "epoch": 0.09426351956976396, + "grad_norm": 0.13466285169124603, + "learning_rate": 0.00019577044580100189, + "loss": 0.6529, + "step": 631 + }, + { + "epoch": 0.09441290708096803, + "grad_norm": 0.11548906564712524, + "learning_rate": 0.00019575691031570446, + "loss": 0.7818, + "step": 632 + }, + { + "epoch": 0.0945622945921721, + "grad_norm": 0.18528424203395844, + "learning_rate": 0.00019574335367618883, + "loss": 1.0789, + "step": 633 + }, + { + "epoch": 0.09471168210337616, + "grad_norm": 0.1468898057937622, + "learning_rate": 0.00019572977588544986, + "loss": 0.9756, + "step": 634 + }, + { + "epoch": 0.09486106961458023, + "grad_norm": 0.16121335327625275, + "learning_rate": 0.00019571617694648713, + "loss": 0.7288, + "step": 635 + }, + { + "epoch": 0.09501045712578428, + "grad_norm": 0.12135611474514008, + "learning_rate": 0.00019570255686230485, + "loss": 0.5934, + "step": 636 + }, + { + "epoch": 0.09515984463698834, + "grad_norm": 0.12789905071258545, + "learning_rate": 0.0001956889156359119, + "loss": 0.6069, + "step": 637 + }, + { + "epoch": 0.09530923214819241, + "grad_norm": 0.359639436006546, + "learning_rate": 0.00019567525327032187, + "loss": 1.3909, + "step": 638 + }, + { + "epoch": 0.09545861965939648, + "grad_norm": 0.18724483251571655, + "learning_rate": 0.00019566156976855297, + "loss": 0.6935, + "step": 639 + }, + { + "epoch": 0.09560800717060054, + "grad_norm": 0.19026364386081696, + "learning_rate": 0.0001956478651336281, + "loss": 0.7966, + "step": 640 + }, + { + "epoch": 0.0957573946818046, + "grad_norm": 0.26914387941360474, + "learning_rate": 0.00019563413936857484, + "loss": 0.8772, + "step": 641 + }, + { + "epoch": 0.09590678219300866, + "grad_norm": 0.1765991449356079, + "learning_rate": 0.00019562039247642546, + "loss": 0.7672, + "step": 642 + }, + { + "epoch": 0.09605616970421273, + "grad_norm": 0.2534937560558319, + "learning_rate": 0.00019560662446021677, + "loss": 1.2049, + "step": 643 + }, + { + "epoch": 0.09620555721541679, + "grad_norm": 0.14177022874355316, + "learning_rate": 0.00019559283532299043, + "loss": 0.5386, + "step": 644 + }, + { + "epoch": 0.09635494472662086, + "grad_norm": 0.1188812330365181, + "learning_rate": 0.00019557902506779268, + "loss": 0.6466, + "step": 645 + }, + { + "epoch": 0.09650433223782492, + "grad_norm": 0.12482491880655289, + "learning_rate": 0.00019556519369767438, + "loss": 0.6614, + "step": 646 + }, + { + "epoch": 0.09665371974902898, + "grad_norm": 0.1450030505657196, + "learning_rate": 0.00019555134121569112, + "loss": 0.7014, + "step": 647 + }, + { + "epoch": 0.09680310726023304, + "grad_norm": 0.12003795802593231, + "learning_rate": 0.0001955374676249031, + "loss": 0.8168, + "step": 648 + }, + { + "epoch": 0.09695249477143711, + "grad_norm": 0.1340424120426178, + "learning_rate": 0.0001955235729283753, + "loss": 0.9613, + "step": 649 + }, + { + "epoch": 0.09710188228264117, + "grad_norm": 0.19842645525932312, + "learning_rate": 0.0001955096571291772, + "loss": 0.8049, + "step": 650 + }, + { + "epoch": 0.09725126979384524, + "grad_norm": 0.256247341632843, + "learning_rate": 0.00019549572023038305, + "loss": 1.3097, + "step": 651 + }, + { + "epoch": 0.09740065730504929, + "grad_norm": 0.4698760509490967, + "learning_rate": 0.0001954817622350717, + "loss": 1.6247, + "step": 652 + }, + { + "epoch": 0.09755004481625336, + "grad_norm": 0.16017255187034607, + "learning_rate": 0.00019546778314632674, + "loss": 0.9872, + "step": 653 + }, + { + "epoch": 0.09769943232745743, + "grad_norm": 0.13327959179878235, + "learning_rate": 0.00019545378296723635, + "loss": 0.8846, + "step": 654 + }, + { + "epoch": 0.09784881983866149, + "grad_norm": 0.3522357642650604, + "learning_rate": 0.0001954397617008934, + "loss": 1.2639, + "step": 655 + }, + { + "epoch": 0.09799820734986556, + "grad_norm": 0.14365145564079285, + "learning_rate": 0.0001954257193503954, + "loss": 0.6805, + "step": 656 + }, + { + "epoch": 0.09814759486106961, + "grad_norm": 0.09565582871437073, + "learning_rate": 0.00019541165591884454, + "loss": 0.5522, + "step": 657 + }, + { + "epoch": 0.09829698237227368, + "grad_norm": 0.11068779975175858, + "learning_rate": 0.0001953975714093476, + "loss": 0.6697, + "step": 658 + }, + { + "epoch": 0.09844636988347774, + "grad_norm": 0.15267115831375122, + "learning_rate": 0.00019538346582501616, + "loss": 0.743, + "step": 659 + }, + { + "epoch": 0.09859575739468181, + "grad_norm": 0.13380476832389832, + "learning_rate": 0.00019536933916896633, + "loss": 0.7273, + "step": 660 + }, + { + "epoch": 0.09874514490588587, + "grad_norm": 0.37871164083480835, + "learning_rate": 0.0001953551914443189, + "loss": 0.9902, + "step": 661 + }, + { + "epoch": 0.09889453241708993, + "grad_norm": 0.16891950368881226, + "learning_rate": 0.00019534102265419932, + "loss": 1.0287, + "step": 662 + }, + { + "epoch": 0.09904391992829399, + "grad_norm": 0.15239520370960236, + "learning_rate": 0.00019532683280173768, + "loss": 0.9578, + "step": 663 + }, + { + "epoch": 0.09919330743949806, + "grad_norm": 0.1469845324754715, + "learning_rate": 0.00019531262189006882, + "loss": 0.8758, + "step": 664 + }, + { + "epoch": 0.09934269495070212, + "grad_norm": 0.15531109273433685, + "learning_rate": 0.00019529838992233208, + "loss": 1.0518, + "step": 665 + }, + { + "epoch": 0.09949208246190619, + "grad_norm": 0.19832894206047058, + "learning_rate": 0.0001952841369016716, + "loss": 0.6697, + "step": 666 + }, + { + "epoch": 0.09964146997311024, + "grad_norm": 0.12545038759708405, + "learning_rate": 0.00019526986283123601, + "loss": 0.9188, + "step": 667 + }, + { + "epoch": 0.09979085748431431, + "grad_norm": 0.15024740993976593, + "learning_rate": 0.00019525556771417875, + "loss": 0.9019, + "step": 668 + }, + { + "epoch": 0.09994024499551837, + "grad_norm": 0.17119628190994263, + "learning_rate": 0.0001952412515536578, + "loss": 0.7931, + "step": 669 + }, + { + "epoch": 0.10008963250672244, + "grad_norm": 0.2737690806388855, + "learning_rate": 0.00019522691435283585, + "loss": 1.1618, + "step": 670 + }, + { + "epoch": 0.1002390200179265, + "grad_norm": 0.16071948409080505, + "learning_rate": 0.00019521255611488022, + "loss": 0.4868, + "step": 671 + }, + { + "epoch": 0.10038840752913056, + "grad_norm": 0.14076560735702515, + "learning_rate": 0.00019519817684296285, + "loss": 0.6771, + "step": 672 + }, + { + "epoch": 0.10053779504033462, + "grad_norm": 0.1131826639175415, + "learning_rate": 0.00019518377654026032, + "loss": 0.768, + "step": 673 + }, + { + "epoch": 0.10068718255153869, + "grad_norm": 0.15727178752422333, + "learning_rate": 0.00019516935520995393, + "loss": 0.8779, + "step": 674 + }, + { + "epoch": 0.10083657006274276, + "grad_norm": 0.14138628542423248, + "learning_rate": 0.00019515491285522957, + "loss": 0.8227, + "step": 675 + }, + { + "epoch": 0.10098595757394682, + "grad_norm": 0.2561928331851959, + "learning_rate": 0.0001951404494792778, + "loss": 0.9701, + "step": 676 + }, + { + "epoch": 0.10113534508515089, + "grad_norm": 0.22002875804901123, + "learning_rate": 0.00019512596508529378, + "loss": 0.8702, + "step": 677 + }, + { + "epoch": 0.10128473259635494, + "grad_norm": 0.2049333155155182, + "learning_rate": 0.00019511145967647737, + "loss": 0.8744, + "step": 678 + }, + { + "epoch": 0.101434120107559, + "grad_norm": 0.19357016682624817, + "learning_rate": 0.00019509693325603302, + "loss": 0.6567, + "step": 679 + }, + { + "epoch": 0.10158350761876307, + "grad_norm": 0.14756688475608826, + "learning_rate": 0.00019508238582716984, + "loss": 0.7318, + "step": 680 + }, + { + "epoch": 0.10173289512996714, + "grad_norm": 0.16614054143428802, + "learning_rate": 0.00019506781739310163, + "loss": 0.9397, + "step": 681 + }, + { + "epoch": 0.1018822826411712, + "grad_norm": 0.11248838156461716, + "learning_rate": 0.00019505322795704676, + "loss": 0.7042, + "step": 682 + }, + { + "epoch": 0.10203167015237526, + "grad_norm": 0.14995478093624115, + "learning_rate": 0.00019503861752222826, + "loss": 0.7761, + "step": 683 + }, + { + "epoch": 0.10218105766357932, + "grad_norm": 0.15232588350772858, + "learning_rate": 0.0001950239860918738, + "loss": 0.7433, + "step": 684 + }, + { + "epoch": 0.10233044517478339, + "grad_norm": 0.17888033390045166, + "learning_rate": 0.00019500933366921571, + "loss": 1.0783, + "step": 685 + }, + { + "epoch": 0.10247983268598745, + "grad_norm": 0.16491816937923431, + "learning_rate": 0.00019499466025749097, + "loss": 0.6902, + "step": 686 + }, + { + "epoch": 0.10262922019719152, + "grad_norm": 0.29704946279525757, + "learning_rate": 0.00019497996585994112, + "loss": 0.8224, + "step": 687 + }, + { + "epoch": 0.10277860770839557, + "grad_norm": 0.13493022322654724, + "learning_rate": 0.00019496525047981242, + "loss": 0.6601, + "step": 688 + }, + { + "epoch": 0.10292799521959964, + "grad_norm": 0.24183045327663422, + "learning_rate": 0.0001949505141203557, + "loss": 1.0734, + "step": 689 + }, + { + "epoch": 0.1030773827308037, + "grad_norm": 0.41064098477363586, + "learning_rate": 0.00019493575678482649, + "loss": 1.1831, + "step": 690 + }, + { + "epoch": 0.10322677024200777, + "grad_norm": 0.21319490671157837, + "learning_rate": 0.0001949209784764849, + "loss": 0.9954, + "step": 691 + }, + { + "epoch": 0.10337615775321184, + "grad_norm": 0.15545840561389923, + "learning_rate": 0.0001949061791985957, + "loss": 0.831, + "step": 692 + }, + { + "epoch": 0.10352554526441589, + "grad_norm": 0.13938447833061218, + "learning_rate": 0.00019489135895442826, + "loss": 0.8778, + "step": 693 + }, + { + "epoch": 0.10367493277561995, + "grad_norm": 0.27140700817108154, + "learning_rate": 0.00019487651774725663, + "loss": 0.7168, + "step": 694 + }, + { + "epoch": 0.10382432028682402, + "grad_norm": 0.17713186144828796, + "learning_rate": 0.00019486165558035948, + "loss": 0.7476, + "step": 695 + }, + { + "epoch": 0.10397370779802809, + "grad_norm": 0.19945883750915527, + "learning_rate": 0.00019484677245702004, + "loss": 0.5964, + "step": 696 + }, + { + "epoch": 0.10412309530923215, + "grad_norm": 0.27336767315864563, + "learning_rate": 0.0001948318683805263, + "loss": 1.0354, + "step": 697 + }, + { + "epoch": 0.1042724828204362, + "grad_norm": 0.15666867792606354, + "learning_rate": 0.0001948169433541708, + "loss": 0.6364, + "step": 698 + }, + { + "epoch": 0.10442187033164027, + "grad_norm": 0.17508071660995483, + "learning_rate": 0.00019480199738125068, + "loss": 0.7329, + "step": 699 + }, + { + "epoch": 0.10457125784284434, + "grad_norm": 0.16016682982444763, + "learning_rate": 0.00019478703046506773, + "loss": 0.6733, + "step": 700 + }, + { + "epoch": 0.1047206453540484, + "grad_norm": 0.10869868099689484, + "learning_rate": 0.0001947720426089284, + "loss": 0.6218, + "step": 701 + }, + { + "epoch": 0.10487003286525247, + "grad_norm": 0.37434107065200806, + "learning_rate": 0.00019475703381614375, + "loss": 1.1511, + "step": 702 + }, + { + "epoch": 0.10501942037645653, + "grad_norm": 0.2148595005273819, + "learning_rate": 0.00019474200409002945, + "loss": 1.0165, + "step": 703 + }, + { + "epoch": 0.10516880788766059, + "grad_norm": 0.1994418203830719, + "learning_rate": 0.00019472695343390585, + "loss": 0.9581, + "step": 704 + }, + { + "epoch": 0.10531819539886465, + "grad_norm": 0.14765793085098267, + "learning_rate": 0.0001947118818510978, + "loss": 0.9704, + "step": 705 + }, + { + "epoch": 0.10546758291006872, + "grad_norm": 0.3004849851131439, + "learning_rate": 0.00019469678934493488, + "loss": 1.3147, + "step": 706 + }, + { + "epoch": 0.10561697042127279, + "grad_norm": 0.14334464073181152, + "learning_rate": 0.0001946816759187513, + "loss": 0.5802, + "step": 707 + }, + { + "epoch": 0.10576635793247685, + "grad_norm": 0.20262622833251953, + "learning_rate": 0.0001946665415758858, + "loss": 1.0572, + "step": 708 + }, + { + "epoch": 0.1059157454436809, + "grad_norm": 0.1506635844707489, + "learning_rate": 0.00019465138631968184, + "loss": 1.264, + "step": 709 + }, + { + "epoch": 0.10606513295488497, + "grad_norm": 0.1906273066997528, + "learning_rate": 0.00019463621015348748, + "loss": 0.9111, + "step": 710 + }, + { + "epoch": 0.10621452046608904, + "grad_norm": 0.14025966823101044, + "learning_rate": 0.0001946210130806553, + "loss": 0.7685, + "step": 711 + }, + { + "epoch": 0.1063639079772931, + "grad_norm": 0.13425306975841522, + "learning_rate": 0.00019460579510454263, + "loss": 0.8569, + "step": 712 + }, + { + "epoch": 0.10651329548849717, + "grad_norm": 0.2626805305480957, + "learning_rate": 0.0001945905562285113, + "loss": 1.1358, + "step": 713 + }, + { + "epoch": 0.10666268299970122, + "grad_norm": 0.12032942473888397, + "learning_rate": 0.00019457529645592792, + "loss": 0.9112, + "step": 714 + }, + { + "epoch": 0.10681207051090529, + "grad_norm": 0.16017121076583862, + "learning_rate": 0.0001945600157901635, + "loss": 0.4619, + "step": 715 + }, + { + "epoch": 0.10696145802210935, + "grad_norm": 0.1190473660826683, + "learning_rate": 0.00019454471423459389, + "loss": 0.7032, + "step": 716 + }, + { + "epoch": 0.10711084553331342, + "grad_norm": 0.15523146092891693, + "learning_rate": 0.00019452939179259937, + "loss": 0.6421, + "step": 717 + }, + { + "epoch": 0.10726023304451748, + "grad_norm": 0.27652865648269653, + "learning_rate": 0.00019451404846756494, + "loss": 0.8187, + "step": 718 + }, + { + "epoch": 0.10740962055572154, + "grad_norm": 0.2987957000732422, + "learning_rate": 0.00019449868426288017, + "loss": 0.8263, + "step": 719 + }, + { + "epoch": 0.1075590080669256, + "grad_norm": 0.143573597073555, + "learning_rate": 0.00019448329918193927, + "loss": 0.7988, + "step": 720 + }, + { + "epoch": 0.10770839557812967, + "grad_norm": 0.1809718906879425, + "learning_rate": 0.00019446789322814106, + "loss": 0.7172, + "step": 721 + }, + { + "epoch": 0.10785778308933373, + "grad_norm": 0.3178151845932007, + "learning_rate": 0.00019445246640488893, + "loss": 0.8902, + "step": 722 + }, + { + "epoch": 0.1080071706005378, + "grad_norm": 0.17539192736148834, + "learning_rate": 0.00019443701871559092, + "loss": 1.0598, + "step": 723 + }, + { + "epoch": 0.10815655811174185, + "grad_norm": 0.1401323676109314, + "learning_rate": 0.00019442155016365965, + "loss": 0.772, + "step": 724 + }, + { + "epoch": 0.10830594562294592, + "grad_norm": 0.14516496658325195, + "learning_rate": 0.0001944060607525124, + "loss": 1.0332, + "step": 725 + }, + { + "epoch": 0.10845533313414998, + "grad_norm": 0.12078884243965149, + "learning_rate": 0.00019439055048557101, + "loss": 0.4937, + "step": 726 + }, + { + "epoch": 0.10860472064535405, + "grad_norm": 0.12402236461639404, + "learning_rate": 0.00019437501936626198, + "loss": 0.8666, + "step": 727 + }, + { + "epoch": 0.10875410815655812, + "grad_norm": 0.2173183113336563, + "learning_rate": 0.00019435946739801633, + "loss": 0.7866, + "step": 728 + }, + { + "epoch": 0.10890349566776218, + "grad_norm": 0.18823803961277008, + "learning_rate": 0.00019434389458426976, + "loss": 0.6379, + "step": 729 + }, + { + "epoch": 0.10905288317896623, + "grad_norm": 0.1408422738313675, + "learning_rate": 0.00019432830092846253, + "loss": 0.7941, + "step": 730 + }, + { + "epoch": 0.1092022706901703, + "grad_norm": 0.11038701981306076, + "learning_rate": 0.00019431268643403958, + "loss": 0.7818, + "step": 731 + }, + { + "epoch": 0.10935165820137437, + "grad_norm": 0.2575507164001465, + "learning_rate": 0.0001942970511044503, + "loss": 0.8922, + "step": 732 + }, + { + "epoch": 0.10950104571257843, + "grad_norm": 0.17680765688419342, + "learning_rate": 0.00019428139494314888, + "loss": 0.9351, + "step": 733 + }, + { + "epoch": 0.1096504332237825, + "grad_norm": 0.13490507006645203, + "learning_rate": 0.00019426571795359398, + "loss": 0.8378, + "step": 734 + }, + { + "epoch": 0.10979982073498655, + "grad_norm": 0.28408023715019226, + "learning_rate": 0.0001942500201392489, + "loss": 0.7867, + "step": 735 + }, + { + "epoch": 0.10994920824619062, + "grad_norm": 0.16506126523017883, + "learning_rate": 0.0001942343015035815, + "loss": 1.1124, + "step": 736 + }, + { + "epoch": 0.11009859575739468, + "grad_norm": 0.2039964199066162, + "learning_rate": 0.00019421856205006433, + "loss": 0.9931, + "step": 737 + }, + { + "epoch": 0.11024798326859875, + "grad_norm": 0.14856329560279846, + "learning_rate": 0.00019420280178217443, + "loss": 0.7827, + "step": 738 + }, + { + "epoch": 0.11039737077980281, + "grad_norm": 0.1818045973777771, + "learning_rate": 0.00019418702070339355, + "loss": 0.5399, + "step": 739 + }, + { + "epoch": 0.11054675829100687, + "grad_norm": 0.2717982232570648, + "learning_rate": 0.00019417121881720793, + "loss": 1.2213, + "step": 740 + }, + { + "epoch": 0.11069614580221093, + "grad_norm": 0.12363403290510178, + "learning_rate": 0.0001941553961271085, + "loss": 0.7174, + "step": 741 + }, + { + "epoch": 0.110845533313415, + "grad_norm": 0.22691577672958374, + "learning_rate": 0.0001941395526365907, + "loss": 1.1225, + "step": 742 + }, + { + "epoch": 0.11099492082461906, + "grad_norm": 0.16864041984081268, + "learning_rate": 0.00019412368834915462, + "loss": 0.5674, + "step": 743 + }, + { + "epoch": 0.11114430833582313, + "grad_norm": 0.20602700114250183, + "learning_rate": 0.00019410780326830498, + "loss": 0.9028, + "step": 744 + }, + { + "epoch": 0.11129369584702718, + "grad_norm": 0.13901367783546448, + "learning_rate": 0.00019409189739755096, + "loss": 0.6309, + "step": 745 + }, + { + "epoch": 0.11144308335823125, + "grad_norm": 0.20725879073143005, + "learning_rate": 0.0001940759707404065, + "loss": 0.5478, + "step": 746 + }, + { + "epoch": 0.11159247086943531, + "grad_norm": 0.14670120179653168, + "learning_rate": 0.00019406002330039001, + "loss": 0.5592, + "step": 747 + }, + { + "epoch": 0.11174185838063938, + "grad_norm": 0.17827396094799042, + "learning_rate": 0.00019404405508102455, + "loss": 0.8181, + "step": 748 + }, + { + "epoch": 0.11189124589184345, + "grad_norm": 0.1741906702518463, + "learning_rate": 0.00019402806608583774, + "loss": 1.1636, + "step": 749 + }, + { + "epoch": 0.1120406334030475, + "grad_norm": 0.12306499481201172, + "learning_rate": 0.00019401205631836178, + "loss": 0.7564, + "step": 750 + }, + { + "epoch": 0.11219002091425156, + "grad_norm": 0.12863090634346008, + "learning_rate": 0.00019399602578213353, + "loss": 0.739, + "step": 751 + }, + { + "epoch": 0.11233940842545563, + "grad_norm": 0.11944539844989777, + "learning_rate": 0.00019397997448069435, + "loss": 0.5797, + "step": 752 + }, + { + "epoch": 0.1124887959366597, + "grad_norm": 0.11944981664419174, + "learning_rate": 0.00019396390241759024, + "loss": 0.7503, + "step": 753 + }, + { + "epoch": 0.11263818344786376, + "grad_norm": 0.12803085148334503, + "learning_rate": 0.00019394780959637177, + "loss": 0.984, + "step": 754 + }, + { + "epoch": 0.11278757095906783, + "grad_norm": 8.864607810974121, + "learning_rate": 0.0001939316960205941, + "loss": 4.3697, + "step": 755 + }, + { + "epoch": 0.11293695847027188, + "grad_norm": 0.15747593343257904, + "learning_rate": 0.000193915561693817, + "loss": 0.7344, + "step": 756 + }, + { + "epoch": 0.11308634598147595, + "grad_norm": 0.13283927738666534, + "learning_rate": 0.00019389940661960478, + "loss": 0.5302, + "step": 757 + }, + { + "epoch": 0.11323573349268001, + "grad_norm": 0.12874341011047363, + "learning_rate": 0.00019388323080152633, + "loss": 0.828, + "step": 758 + }, + { + "epoch": 0.11338512100388408, + "grad_norm": 0.4698888063430786, + "learning_rate": 0.00019386703424315518, + "loss": 1.5101, + "step": 759 + }, + { + "epoch": 0.11353450851508815, + "grad_norm": 0.12892121076583862, + "learning_rate": 0.00019385081694806936, + "loss": 0.7772, + "step": 760 + }, + { + "epoch": 0.1136838960262922, + "grad_norm": 0.28520870208740234, + "learning_rate": 0.00019383457891985158, + "loss": 0.928, + "step": 761 + }, + { + "epoch": 0.11383328353749626, + "grad_norm": 1.7955225706100464, + "learning_rate": 0.00019381832016208904, + "loss": 0.955, + "step": 762 + }, + { + "epoch": 0.11398267104870033, + "grad_norm": 0.14767761528491974, + "learning_rate": 0.00019380204067837356, + "loss": 0.6337, + "step": 763 + }, + { + "epoch": 0.1141320585599044, + "grad_norm": 0.171101912856102, + "learning_rate": 0.0001937857404723016, + "loss": 0.7905, + "step": 764 + }, + { + "epoch": 0.11428144607110846, + "grad_norm": 0.33381524682044983, + "learning_rate": 0.000193769419547474, + "loss": 0.9788, + "step": 765 + }, + { + "epoch": 0.11443083358231251, + "grad_norm": 0.1214352399110794, + "learning_rate": 0.00019375307790749647, + "loss": 0.9152, + "step": 766 + }, + { + "epoch": 0.11458022109351658, + "grad_norm": 0.3238249123096466, + "learning_rate": 0.00019373671555597902, + "loss": 0.9923, + "step": 767 + }, + { + "epoch": 0.11472960860472065, + "grad_norm": 0.14854952692985535, + "learning_rate": 0.0001937203324965364, + "loss": 0.7194, + "step": 768 + }, + { + "epoch": 0.11487899611592471, + "grad_norm": 0.13817191123962402, + "learning_rate": 0.00019370392873278784, + "loss": 0.7175, + "step": 769 + }, + { + "epoch": 0.11502838362712878, + "grad_norm": 0.24061405658721924, + "learning_rate": 0.0001936875042683573, + "loss": 0.9647, + "step": 770 + }, + { + "epoch": 0.11517777113833283, + "grad_norm": 0.1889171153306961, + "learning_rate": 0.00019367105910687307, + "loss": 0.9805, + "step": 771 + }, + { + "epoch": 0.1153271586495369, + "grad_norm": 0.2141283005475998, + "learning_rate": 0.00019365459325196825, + "loss": 0.7467, + "step": 772 + }, + { + "epoch": 0.11547654616074096, + "grad_norm": 0.20832307636737823, + "learning_rate": 0.00019363810670728032, + "loss": 0.8881, + "step": 773 + }, + { + "epoch": 0.11562593367194503, + "grad_norm": 0.19508862495422363, + "learning_rate": 0.00019362159947645152, + "loss": 0.9594, + "step": 774 + }, + { + "epoch": 0.1157753211831491, + "grad_norm": 0.16393844783306122, + "learning_rate": 0.00019360507156312848, + "loss": 0.8358, + "step": 775 + }, + { + "epoch": 0.11592470869435315, + "grad_norm": 0.17555637657642365, + "learning_rate": 0.00019358852297096253, + "loss": 0.7137, + "step": 776 + }, + { + "epoch": 0.11607409620555721, + "grad_norm": 0.16748878359794617, + "learning_rate": 0.00019357195370360946, + "loss": 0.8945, + "step": 777 + }, + { + "epoch": 0.11622348371676128, + "grad_norm": 0.11165950447320938, + "learning_rate": 0.00019355536376472972, + "loss": 0.5322, + "step": 778 + }, + { + "epoch": 0.11637287122796534, + "grad_norm": 0.12036951631307602, + "learning_rate": 0.00019353875315798828, + "loss": 0.824, + "step": 779 + }, + { + "epoch": 0.11652225873916941, + "grad_norm": 0.2941713333129883, + "learning_rate": 0.0001935221218870547, + "loss": 1.1819, + "step": 780 + }, + { + "epoch": 0.11667164625037346, + "grad_norm": 0.16016024351119995, + "learning_rate": 0.0001935054699556031, + "loss": 0.8635, + "step": 781 + }, + { + "epoch": 0.11682103376157753, + "grad_norm": 0.16375142335891724, + "learning_rate": 0.0001934887973673121, + "loss": 1.0148, + "step": 782 + }, + { + "epoch": 0.1169704212727816, + "grad_norm": 0.3563995659351349, + "learning_rate": 0.000193472104125865, + "loss": 0.9712, + "step": 783 + }, + { + "epoch": 0.11711980878398566, + "grad_norm": 0.19017907977104187, + "learning_rate": 0.0001934553902349496, + "loss": 0.9978, + "step": 784 + }, + { + "epoch": 0.11726919629518973, + "grad_norm": 0.14762148261070251, + "learning_rate": 0.00019343865569825818, + "loss": 0.8258, + "step": 785 + }, + { + "epoch": 0.11741858380639379, + "grad_norm": 0.19044718146324158, + "learning_rate": 0.00019342190051948777, + "loss": 1.0279, + "step": 786 + }, + { + "epoch": 0.11756797131759784, + "grad_norm": 0.15434981882572174, + "learning_rate": 0.0001934051247023398, + "loss": 0.6936, + "step": 787 + }, + { + "epoch": 0.11771735882880191, + "grad_norm": 0.21782897412776947, + "learning_rate": 0.0001933883282505203, + "loss": 0.8074, + "step": 788 + }, + { + "epoch": 0.11786674634000598, + "grad_norm": 0.1382855623960495, + "learning_rate": 0.00019337151116773993, + "loss": 0.76, + "step": 789 + }, + { + "epoch": 0.11801613385121004, + "grad_norm": 0.12988682091236115, + "learning_rate": 0.00019335467345771377, + "loss": 0.7619, + "step": 790 + }, + { + "epoch": 0.11816552136241411, + "grad_norm": 0.14587444067001343, + "learning_rate": 0.0001933378151241616, + "loss": 0.6618, + "step": 791 + }, + { + "epoch": 0.11831490887361816, + "grad_norm": 0.13079139590263367, + "learning_rate": 0.0001933209361708077, + "loss": 0.7573, + "step": 792 + }, + { + "epoch": 0.11846429638482223, + "grad_norm": 0.27695533633232117, + "learning_rate": 0.00019330403660138085, + "loss": 1.0731, + "step": 793 + }, + { + "epoch": 0.11861368389602629, + "grad_norm": 0.1361035704612732, + "learning_rate": 0.00019328711641961445, + "loss": 0.8251, + "step": 794 + }, + { + "epoch": 0.11876307140723036, + "grad_norm": 0.16837474703788757, + "learning_rate": 0.00019327017562924644, + "loss": 0.9535, + "step": 795 + }, + { + "epoch": 0.11891245891843442, + "grad_norm": 0.31326931715011597, + "learning_rate": 0.00019325321423401933, + "loss": 0.9828, + "step": 796 + }, + { + "epoch": 0.11906184642963848, + "grad_norm": 0.16937793791294098, + "learning_rate": 0.00019323623223768015, + "loss": 0.9445, + "step": 797 + }, + { + "epoch": 0.11921123394084254, + "grad_norm": 0.15119615197181702, + "learning_rate": 0.00019321922964398046, + "loss": 0.7483, + "step": 798 + }, + { + "epoch": 0.11936062145204661, + "grad_norm": 0.1750727891921997, + "learning_rate": 0.00019320220645667645, + "loss": 0.9364, + "step": 799 + }, + { + "epoch": 0.11951000896325067, + "grad_norm": 0.22479920089244843, + "learning_rate": 0.00019318516267952874, + "loss": 0.854, + "step": 800 + }, + { + "epoch": 0.11965939647445474, + "grad_norm": 0.12636275589466095, + "learning_rate": 0.00019316809831630265, + "loss": 0.8127, + "step": 801 + }, + { + "epoch": 0.11980878398565879, + "grad_norm": 0.1225171908736229, + "learning_rate": 0.00019315101337076792, + "loss": 0.7809, + "step": 802 + }, + { + "epoch": 0.11995817149686286, + "grad_norm": 0.24531179666519165, + "learning_rate": 0.00019313390784669895, + "loss": 1.0615, + "step": 803 + }, + { + "epoch": 0.12010755900806692, + "grad_norm": 0.15142390131950378, + "learning_rate": 0.0001931167817478745, + "loss": 0.913, + "step": 804 + }, + { + "epoch": 0.12025694651927099, + "grad_norm": 0.13152721524238586, + "learning_rate": 0.00019309963507807811, + "loss": 0.6284, + "step": 805 + }, + { + "epoch": 0.12040633403047506, + "grad_norm": 0.10537417232990265, + "learning_rate": 0.0001930824678410977, + "loss": 0.6071, + "step": 806 + }, + { + "epoch": 0.12055572154167911, + "grad_norm": 0.13866813480854034, + "learning_rate": 0.0001930652800407258, + "loss": 0.7461, + "step": 807 + }, + { + "epoch": 0.12070510905288317, + "grad_norm": 0.16239172220230103, + "learning_rate": 0.00019304807168075944, + "loss": 0.8089, + "step": 808 + }, + { + "epoch": 0.12085449656408724, + "grad_norm": 0.12834633886814117, + "learning_rate": 0.00019303084276500027, + "loss": 0.8459, + "step": 809 + }, + { + "epoch": 0.12100388407529131, + "grad_norm": 0.17036044597625732, + "learning_rate": 0.00019301359329725436, + "loss": 0.8131, + "step": 810 + }, + { + "epoch": 0.12115327158649537, + "grad_norm": 0.14274178445339203, + "learning_rate": 0.00019299632328133247, + "loss": 0.9325, + "step": 811 + }, + { + "epoch": 0.12130265909769944, + "grad_norm": 0.16712737083435059, + "learning_rate": 0.00019297903272104977, + "loss": 0.88, + "step": 812 + }, + { + "epoch": 0.12145204660890349, + "grad_norm": 0.1521551012992859, + "learning_rate": 0.00019296172162022604, + "loss": 0.9838, + "step": 813 + }, + { + "epoch": 0.12160143412010756, + "grad_norm": 0.1645859032869339, + "learning_rate": 0.00019294438998268554, + "loss": 0.6344, + "step": 814 + }, + { + "epoch": 0.12175082163131162, + "grad_norm": 0.15957693755626678, + "learning_rate": 0.00019292703781225717, + "loss": 0.926, + "step": 815 + }, + { + "epoch": 0.12190020914251569, + "grad_norm": 0.13982924818992615, + "learning_rate": 0.00019290966511277422, + "loss": 0.849, + "step": 816 + }, + { + "epoch": 0.12204959665371976, + "grad_norm": 0.16864308714866638, + "learning_rate": 0.00019289227188807467, + "loss": 0.6237, + "step": 817 + }, + { + "epoch": 0.12219898416492381, + "grad_norm": 0.14285649359226227, + "learning_rate": 0.00019287485814200087, + "loss": 0.6809, + "step": 818 + }, + { + "epoch": 0.12234837167612787, + "grad_norm": 0.6627392768859863, + "learning_rate": 0.00019285742387839988, + "loss": 1.9625, + "step": 819 + }, + { + "epoch": 0.12249775918733194, + "grad_norm": 0.13897117972373962, + "learning_rate": 0.00019283996910112318, + "loss": 1.0312, + "step": 820 + }, + { + "epoch": 0.122647146698536, + "grad_norm": 0.15520478785037994, + "learning_rate": 0.00019282249381402677, + "loss": 0.8191, + "step": 821 + }, + { + "epoch": 0.12279653420974007, + "grad_norm": 0.24286359548568726, + "learning_rate": 0.00019280499802097126, + "loss": 1.1401, + "step": 822 + }, + { + "epoch": 0.12294592172094412, + "grad_norm": 0.15701743960380554, + "learning_rate": 0.00019278748172582173, + "loss": 0.7389, + "step": 823 + }, + { + "epoch": 0.12309530923214819, + "grad_norm": 0.16320520639419556, + "learning_rate": 0.0001927699449324478, + "loss": 0.9232, + "step": 824 + }, + { + "epoch": 0.12324469674335226, + "grad_norm": 0.13166747987270355, + "learning_rate": 0.00019275238764472364, + "loss": 0.648, + "step": 825 + }, + { + "epoch": 0.12339408425455632, + "grad_norm": 0.1490585058927536, + "learning_rate": 0.00019273480986652794, + "loss": 0.8485, + "step": 826 + }, + { + "epoch": 0.12354347176576039, + "grad_norm": 0.1607678383588791, + "learning_rate": 0.00019271721160174388, + "loss": 0.7415, + "step": 827 + }, + { + "epoch": 0.12369285927696444, + "grad_norm": 0.13538645207881927, + "learning_rate": 0.0001926995928542592, + "loss": 0.702, + "step": 828 + }, + { + "epoch": 0.1238422467881685, + "grad_norm": 0.17759506404399872, + "learning_rate": 0.00019268195362796622, + "loss": 0.819, + "step": 829 + }, + { + "epoch": 0.12399163429937257, + "grad_norm": 0.14529088139533997, + "learning_rate": 0.00019266429392676164, + "loss": 0.5262, + "step": 830 + }, + { + "epoch": 0.12414102181057664, + "grad_norm": 0.16848066449165344, + "learning_rate": 0.0001926466137545468, + "loss": 0.6761, + "step": 831 + }, + { + "epoch": 0.1242904093217807, + "grad_norm": 0.33541253209114075, + "learning_rate": 0.00019262891311522755, + "loss": 0.9293, + "step": 832 + }, + { + "epoch": 0.12443979683298476, + "grad_norm": 0.2218908667564392, + "learning_rate": 0.00019261119201271422, + "loss": 1.0644, + "step": 833 + }, + { + "epoch": 0.12458918434418882, + "grad_norm": 0.1724868267774582, + "learning_rate": 0.0001925934504509217, + "loss": 0.8411, + "step": 834 + }, + { + "epoch": 0.12473857185539289, + "grad_norm": 0.12873922288417816, + "learning_rate": 0.00019257568843376936, + "loss": 0.7466, + "step": 835 + }, + { + "epoch": 0.12488795936659695, + "grad_norm": 0.13556797802448273, + "learning_rate": 0.00019255790596518112, + "loss": 0.675, + "step": 836 + }, + { + "epoch": 0.12503734687780102, + "grad_norm": 0.12460656464099884, + "learning_rate": 0.00019254010304908543, + "loss": 0.6401, + "step": 837 + }, + { + "epoch": 0.12518673438900507, + "grad_norm": 0.16490042209625244, + "learning_rate": 0.00019252227968941522, + "loss": 0.8934, + "step": 838 + }, + { + "epoch": 0.12533612190020915, + "grad_norm": 0.13448311388492584, + "learning_rate": 0.00019250443589010792, + "loss": 0.7896, + "step": 839 + }, + { + "epoch": 0.1254855094114132, + "grad_norm": 0.16792979836463928, + "learning_rate": 0.00019248657165510556, + "loss": 0.74, + "step": 840 + }, + { + "epoch": 0.12563489692261726, + "grad_norm": 0.1711784452199936, + "learning_rate": 0.00019246868698835458, + "loss": 0.5992, + "step": 841 + }, + { + "epoch": 0.12578428443382134, + "grad_norm": 0.12499507516622543, + "learning_rate": 0.00019245078189380604, + "loss": 0.8517, + "step": 842 + }, + { + "epoch": 0.1259336719450254, + "grad_norm": 0.09736009687185287, + "learning_rate": 0.00019243285637541544, + "loss": 0.6209, + "step": 843 + }, + { + "epoch": 0.12608305945622947, + "grad_norm": 0.21913480758666992, + "learning_rate": 0.0001924149104371428, + "loss": 0.9499, + "step": 844 + }, + { + "epoch": 0.12623244696743352, + "grad_norm": 0.23238122463226318, + "learning_rate": 0.00019239694408295266, + "loss": 1.1666, + "step": 845 + }, + { + "epoch": 0.12638183447863757, + "grad_norm": 0.24529285728931427, + "learning_rate": 0.00019237895731681408, + "loss": 0.734, + "step": 846 + }, + { + "epoch": 0.12653122198984165, + "grad_norm": 0.16577094793319702, + "learning_rate": 0.00019236095014270064, + "loss": 1.0808, + "step": 847 + }, + { + "epoch": 0.1266806095010457, + "grad_norm": 0.08479466289281845, + "learning_rate": 0.0001923429225645904, + "loss": 0.4277, + "step": 848 + }, + { + "epoch": 0.12682999701224978, + "grad_norm": 0.15633448958396912, + "learning_rate": 0.00019232487458646587, + "loss": 0.8547, + "step": 849 + }, + { + "epoch": 0.12697938452345384, + "grad_norm": 0.2966431975364685, + "learning_rate": 0.00019230680621231425, + "loss": 0.7691, + "step": 850 + }, + { + "epoch": 0.12712877203465792, + "grad_norm": 0.21817810833454132, + "learning_rate": 0.00019228871744612704, + "loss": 0.9089, + "step": 851 + }, + { + "epoch": 0.12727815954586197, + "grad_norm": 0.1577853262424469, + "learning_rate": 0.0001922706082919004, + "loss": 0.7345, + "step": 852 + }, + { + "epoch": 0.12742754705706602, + "grad_norm": 0.14371559023857117, + "learning_rate": 0.00019225247875363487, + "loss": 0.9727, + "step": 853 + }, + { + "epoch": 0.1275769345682701, + "grad_norm": 0.10418539494276047, + "learning_rate": 0.0001922343288353356, + "loss": 0.6287, + "step": 854 + }, + { + "epoch": 0.12772632207947415, + "grad_norm": 0.14414125680923462, + "learning_rate": 0.00019221615854101215, + "loss": 1.0211, + "step": 855 + }, + { + "epoch": 0.12787570959067823, + "grad_norm": 0.14058874547481537, + "learning_rate": 0.00019219796787467867, + "loss": 0.9388, + "step": 856 + }, + { + "epoch": 0.12802509710188228, + "grad_norm": 0.24741698801517487, + "learning_rate": 0.00019217975684035374, + "loss": 0.8997, + "step": 857 + }, + { + "epoch": 0.12817448461308634, + "grad_norm": 0.1672530174255371, + "learning_rate": 0.00019216152544206049, + "loss": 0.9332, + "step": 858 + }, + { + "epoch": 0.12832387212429042, + "grad_norm": 0.30283141136169434, + "learning_rate": 0.00019214327368382645, + "loss": 1.1119, + "step": 859 + }, + { + "epoch": 0.12847325963549447, + "grad_norm": 0.41509029269218445, + "learning_rate": 0.00019212500156968383, + "loss": 1.0261, + "step": 860 + }, + { + "epoch": 0.12862264714669855, + "grad_norm": 0.1252591907978058, + "learning_rate": 0.00019210670910366917, + "loss": 0.7836, + "step": 861 + }, + { + "epoch": 0.1287720346579026, + "grad_norm": 0.1359296441078186, + "learning_rate": 0.00019208839628982358, + "loss": 0.9255, + "step": 862 + }, + { + "epoch": 0.12892142216910665, + "grad_norm": 0.22711493074893951, + "learning_rate": 0.0001920700631321926, + "loss": 1.1905, + "step": 863 + }, + { + "epoch": 0.12907080968031073, + "grad_norm": 0.13732872903347015, + "learning_rate": 0.00019205170963482643, + "loss": 0.7808, + "step": 864 + }, + { + "epoch": 0.12922019719151479, + "grad_norm": 0.3582336902618408, + "learning_rate": 0.00019203333580177954, + "loss": 0.7557, + "step": 865 + }, + { + "epoch": 0.12936958470271887, + "grad_norm": 0.14303429424762726, + "learning_rate": 0.00019201494163711104, + "loss": 0.898, + "step": 866 + }, + { + "epoch": 0.12951897221392292, + "grad_norm": 0.11724136024713516, + "learning_rate": 0.00019199652714488446, + "loss": 0.6867, + "step": 867 + }, + { + "epoch": 0.12966835972512697, + "grad_norm": 0.14605864882469177, + "learning_rate": 0.00019197809232916795, + "loss": 0.7361, + "step": 868 + }, + { + "epoch": 0.12981774723633105, + "grad_norm": 0.15446977317333221, + "learning_rate": 0.00019195963719403393, + "loss": 0.7941, + "step": 869 + }, + { + "epoch": 0.1299671347475351, + "grad_norm": 0.17148366570472717, + "learning_rate": 0.00019194116174355954, + "loss": 0.7493, + "step": 870 + }, + { + "epoch": 0.13011652225873918, + "grad_norm": 0.12092535197734833, + "learning_rate": 0.0001919226659818262, + "loss": 0.8044, + "step": 871 + }, + { + "epoch": 0.13026590976994323, + "grad_norm": 0.174177348613739, + "learning_rate": 0.00019190414991291998, + "loss": 0.8565, + "step": 872 + }, + { + "epoch": 0.13041529728114729, + "grad_norm": 0.14217911660671234, + "learning_rate": 0.0001918856135409314, + "loss": 0.7967, + "step": 873 + }, + { + "epoch": 0.13056468479235137, + "grad_norm": 0.2161906510591507, + "learning_rate": 0.00019186705686995533, + "loss": 1.0536, + "step": 874 + }, + { + "epoch": 0.13071407230355542, + "grad_norm": 0.11044277995824814, + "learning_rate": 0.00019184847990409134, + "loss": 0.7443, + "step": 875 + }, + { + "epoch": 0.1308634598147595, + "grad_norm": 0.4287470877170563, + "learning_rate": 0.0001918298826474433, + "loss": 1.1078, + "step": 876 + }, + { + "epoch": 0.13101284732596355, + "grad_norm": 0.313516229391098, + "learning_rate": 0.00019181126510411974, + "loss": 1.1343, + "step": 877 + }, + { + "epoch": 0.1311622348371676, + "grad_norm": 0.15673229098320007, + "learning_rate": 0.0001917926272782334, + "loss": 1.0282, + "step": 878 + }, + { + "epoch": 0.13131162234837168, + "grad_norm": 0.14207200706005096, + "learning_rate": 0.00019177396917390187, + "loss": 0.7487, + "step": 879 + }, + { + "epoch": 0.13146100985957573, + "grad_norm": 0.21555255353450775, + "learning_rate": 0.00019175529079524687, + "loss": 0.8816, + "step": 880 + }, + { + "epoch": 0.1316103973707798, + "grad_norm": 0.10295391082763672, + "learning_rate": 0.00019173659214639482, + "loss": 0.7433, + "step": 881 + }, + { + "epoch": 0.13175978488198387, + "grad_norm": 0.12901347875595093, + "learning_rate": 0.00019171787323147654, + "loss": 0.828, + "step": 882 + }, + { + "epoch": 0.13190917239318792, + "grad_norm": 0.1917589008808136, + "learning_rate": 0.00019169913405462733, + "loss": 0.8845, + "step": 883 + }, + { + "epoch": 0.132058559904392, + "grad_norm": 0.14522455632686615, + "learning_rate": 0.00019168037461998695, + "loss": 0.8876, + "step": 884 + }, + { + "epoch": 0.13220794741559605, + "grad_norm": 0.33967098593711853, + "learning_rate": 0.0001916615949316997, + "loss": 0.93, + "step": 885 + }, + { + "epoch": 0.13235733492680013, + "grad_norm": 0.12859250605106354, + "learning_rate": 0.00019164279499391427, + "loss": 0.8351, + "step": 886 + }, + { + "epoch": 0.13250672243800418, + "grad_norm": 0.11414708197116852, + "learning_rate": 0.00019162397481078386, + "loss": 0.5253, + "step": 887 + }, + { + "epoch": 0.13265610994920823, + "grad_norm": 0.6967369914054871, + "learning_rate": 0.00019160513438646617, + "loss": 1.0658, + "step": 888 + }, + { + "epoch": 0.13280549746041231, + "grad_norm": 0.2920320928096771, + "learning_rate": 0.00019158627372512337, + "loss": 1.0202, + "step": 889 + }, + { + "epoch": 0.13295488497161637, + "grad_norm": 0.1285315304994583, + "learning_rate": 0.00019156739283092205, + "loss": 0.5447, + "step": 890 + }, + { + "epoch": 0.13310427248282045, + "grad_norm": 0.6523221135139465, + "learning_rate": 0.00019154849170803327, + "loss": 1.2813, + "step": 891 + }, + { + "epoch": 0.1332536599940245, + "grad_norm": 0.11954151839017868, + "learning_rate": 0.00019152957036063265, + "loss": 0.7564, + "step": 892 + }, + { + "epoch": 0.13340304750522855, + "grad_norm": 0.10571929812431335, + "learning_rate": 0.0001915106287929002, + "loss": 0.4487, + "step": 893 + }, + { + "epoch": 0.13355243501643263, + "grad_norm": 0.24934682250022888, + "learning_rate": 0.00019149166700902032, + "loss": 0.7512, + "step": 894 + }, + { + "epoch": 0.13370182252763668, + "grad_norm": 0.1705646514892578, + "learning_rate": 0.00019147268501318212, + "loss": 0.6647, + "step": 895 + }, + { + "epoch": 0.13385121003884076, + "grad_norm": 0.1977798193693161, + "learning_rate": 0.0001914536828095789, + "loss": 1.0058, + "step": 896 + }, + { + "epoch": 0.13400059755004481, + "grad_norm": 0.16323387622833252, + "learning_rate": 0.00019143466040240863, + "loss": 1.0301, + "step": 897 + }, + { + "epoch": 0.13414998506124887, + "grad_norm": 0.15375447273254395, + "learning_rate": 0.0001914156177958736, + "loss": 0.5928, + "step": 898 + }, + { + "epoch": 0.13429937257245295, + "grad_norm": 0.2192535251379013, + "learning_rate": 0.00019139655499418067, + "loss": 0.6105, + "step": 899 + }, + { + "epoch": 0.134448760083657, + "grad_norm": 0.35383331775665283, + "learning_rate": 0.0001913774720015411, + "loss": 0.9193, + "step": 900 + }, + { + "epoch": 0.13459814759486108, + "grad_norm": 0.2478116750717163, + "learning_rate": 0.0001913583688221706, + "loss": 1.0428, + "step": 901 + }, + { + "epoch": 0.13474753510606513, + "grad_norm": 0.14307573437690735, + "learning_rate": 0.00019133924546028942, + "loss": 0.8198, + "step": 902 + }, + { + "epoch": 0.13489692261726918, + "grad_norm": 0.27853766083717346, + "learning_rate": 0.00019132010192012214, + "loss": 0.9841, + "step": 903 + }, + { + "epoch": 0.13504631012847326, + "grad_norm": 0.13049887120723724, + "learning_rate": 0.00019130093820589791, + "loss": 0.5867, + "step": 904 + }, + { + "epoch": 0.13519569763967731, + "grad_norm": 0.1877085268497467, + "learning_rate": 0.0001912817543218503, + "loss": 1.0579, + "step": 905 + }, + { + "epoch": 0.1353450851508814, + "grad_norm": 0.18851879239082336, + "learning_rate": 0.00019126255027221735, + "loss": 0.5173, + "step": 906 + }, + { + "epoch": 0.13549447266208545, + "grad_norm": 0.1649446338415146, + "learning_rate": 0.00019124332606124152, + "loss": 0.8469, + "step": 907 + }, + { + "epoch": 0.13564386017328953, + "grad_norm": 0.1373070925474167, + "learning_rate": 0.00019122408169316976, + "loss": 0.7754, + "step": 908 + }, + { + "epoch": 0.13579324768449358, + "grad_norm": 0.13052485883235931, + "learning_rate": 0.00019120481717225342, + "loss": 0.9105, + "step": 909 + }, + { + "epoch": 0.13594263519569763, + "grad_norm": 0.15370683372020721, + "learning_rate": 0.00019118553250274832, + "loss": 0.6646, + "step": 910 + }, + { + "epoch": 0.1360920227069017, + "grad_norm": 0.1601782590150833, + "learning_rate": 0.00019116622768891483, + "loss": 0.7487, + "step": 911 + }, + { + "epoch": 0.13624141021810576, + "grad_norm": 0.19794587790966034, + "learning_rate": 0.00019114690273501765, + "loss": 0.5437, + "step": 912 + }, + { + "epoch": 0.13639079772930984, + "grad_norm": 0.12566497921943665, + "learning_rate": 0.00019112755764532594, + "loss": 0.8602, + "step": 913 + }, + { + "epoch": 0.1365401852405139, + "grad_norm": 0.5696573257446289, + "learning_rate": 0.00019110819242411337, + "loss": 1.1799, + "step": 914 + }, + { + "epoch": 0.13668957275171795, + "grad_norm": 0.11676638573408127, + "learning_rate": 0.00019108880707565802, + "loss": 0.6958, + "step": 915 + }, + { + "epoch": 0.13683896026292203, + "grad_norm": 0.12003248929977417, + "learning_rate": 0.00019106940160424244, + "loss": 0.5425, + "step": 916 + }, + { + "epoch": 0.13698834777412608, + "grad_norm": 0.265358030796051, + "learning_rate": 0.00019104997601415352, + "loss": 0.673, + "step": 917 + }, + { + "epoch": 0.13713773528533016, + "grad_norm": 0.17708387970924377, + "learning_rate": 0.0001910305303096828, + "loss": 0.7885, + "step": 918 + }, + { + "epoch": 0.1372871227965342, + "grad_norm": 0.17588114738464355, + "learning_rate": 0.00019101106449512605, + "loss": 0.8977, + "step": 919 + }, + { + "epoch": 0.13743651030773826, + "grad_norm": 0.20153087377548218, + "learning_rate": 0.0001909915785747836, + "loss": 0.6529, + "step": 920 + }, + { + "epoch": 0.13758589781894234, + "grad_norm": 0.14944609999656677, + "learning_rate": 0.00019097207255296022, + "loss": 0.7525, + "step": 921 + }, + { + "epoch": 0.1377352853301464, + "grad_norm": 0.29275017976760864, + "learning_rate": 0.00019095254643396512, + "loss": 1.0195, + "step": 922 + }, + { + "epoch": 0.13788467284135048, + "grad_norm": 0.26294517517089844, + "learning_rate": 0.00019093300022211186, + "loss": 1.2764, + "step": 923 + }, + { + "epoch": 0.13803406035255453, + "grad_norm": 0.10997014492750168, + "learning_rate": 0.0001909134339217186, + "loss": 0.5097, + "step": 924 + }, + { + "epoch": 0.13818344786375858, + "grad_norm": 0.25661003589630127, + "learning_rate": 0.00019089384753710772, + "loss": 1.1546, + "step": 925 + }, + { + "epoch": 0.13833283537496266, + "grad_norm": 0.14332082867622375, + "learning_rate": 0.00019087424107260627, + "loss": 0.6839, + "step": 926 + }, + { + "epoch": 0.1384822228861667, + "grad_norm": 0.29279524087905884, + "learning_rate": 0.0001908546145325456, + "loss": 0.8088, + "step": 927 + }, + { + "epoch": 0.1386316103973708, + "grad_norm": 0.2230231910943985, + "learning_rate": 0.00019083496792126153, + "loss": 0.7403, + "step": 928 + }, + { + "epoch": 0.13878099790857484, + "grad_norm": 0.20737332105636597, + "learning_rate": 0.00019081530124309427, + "loss": 1.1846, + "step": 929 + }, + { + "epoch": 0.1389303854197789, + "grad_norm": 0.2875371277332306, + "learning_rate": 0.00019079561450238854, + "loss": 0.9049, + "step": 930 + }, + { + "epoch": 0.13907977293098298, + "grad_norm": 0.12367036938667297, + "learning_rate": 0.00019077590770349344, + "loss": 0.8605, + "step": 931 + }, + { + "epoch": 0.13922916044218703, + "grad_norm": 0.34429073333740234, + "learning_rate": 0.00019075618085076247, + "loss": 1.0233, + "step": 932 + }, + { + "epoch": 0.1393785479533911, + "grad_norm": 0.15258093178272247, + "learning_rate": 0.00019073643394855368, + "loss": 0.8381, + "step": 933 + }, + { + "epoch": 0.13952793546459516, + "grad_norm": 0.15117698907852173, + "learning_rate": 0.00019071666700122946, + "loss": 0.6532, + "step": 934 + }, + { + "epoch": 0.1396773229757992, + "grad_norm": 0.11969941109418869, + "learning_rate": 0.0001906968800131566, + "loss": 0.7025, + "step": 935 + }, + { + "epoch": 0.1398267104870033, + "grad_norm": 0.12973757088184357, + "learning_rate": 0.00019067707298870638, + "loss": 0.6276, + "step": 936 + }, + { + "epoch": 0.13997609799820734, + "grad_norm": 0.24265740811824799, + "learning_rate": 0.0001906572459322545, + "loss": 0.8548, + "step": 937 + }, + { + "epoch": 0.14012548550941142, + "grad_norm": 0.18727362155914307, + "learning_rate": 0.00019063739884818103, + "loss": 0.5687, + "step": 938 + }, + { + "epoch": 0.14027487302061548, + "grad_norm": 0.13480857014656067, + "learning_rate": 0.00019061753174087054, + "loss": 0.9462, + "step": 939 + }, + { + "epoch": 0.14042426053181953, + "grad_norm": 0.2221670150756836, + "learning_rate": 0.000190597644614712, + "loss": 0.6429, + "step": 940 + }, + { + "epoch": 0.1405736480430236, + "grad_norm": 0.13278596103191376, + "learning_rate": 0.00019057773747409879, + "loss": 0.839, + "step": 941 + }, + { + "epoch": 0.14072303555422766, + "grad_norm": 0.12343950569629669, + "learning_rate": 0.00019055781032342864, + "loss": 0.8807, + "step": 942 + }, + { + "epoch": 0.14087242306543174, + "grad_norm": 0.1782051920890808, + "learning_rate": 0.00019053786316710386, + "loss": 1.0982, + "step": 943 + }, + { + "epoch": 0.1410218105766358, + "grad_norm": 0.1475050151348114, + "learning_rate": 0.00019051789600953102, + "loss": 1.0571, + "step": 944 + }, + { + "epoch": 0.14117119808783984, + "grad_norm": 0.12406841665506363, + "learning_rate": 0.00019049790885512126, + "loss": 0.641, + "step": 945 + }, + { + "epoch": 0.14132058559904392, + "grad_norm": 0.13010482490062714, + "learning_rate": 0.00019047790170829003, + "loss": 0.7616, + "step": 946 + }, + { + "epoch": 0.14146997311024798, + "grad_norm": 0.14790335297584534, + "learning_rate": 0.00019045787457345722, + "loss": 0.6587, + "step": 947 + }, + { + "epoch": 0.14161936062145206, + "grad_norm": 0.11543125659227371, + "learning_rate": 0.00019043782745504711, + "loss": 0.7014, + "step": 948 + }, + { + "epoch": 0.1417687481326561, + "grad_norm": 0.12451034784317017, + "learning_rate": 0.00019041776035748847, + "loss": 0.5931, + "step": 949 + }, + { + "epoch": 0.14191813564386016, + "grad_norm": 0.1254526674747467, + "learning_rate": 0.00019039767328521442, + "loss": 0.7772, + "step": 950 + }, + { + "epoch": 0.14206752315506424, + "grad_norm": 0.21236388385295868, + "learning_rate": 0.00019037756624266252, + "loss": 0.6836, + "step": 951 + }, + { + "epoch": 0.1422169106662683, + "grad_norm": 0.1704844981431961, + "learning_rate": 0.0001903574392342747, + "loss": 0.5561, + "step": 952 + }, + { + "epoch": 0.14236629817747237, + "grad_norm": 0.1470273733139038, + "learning_rate": 0.0001903372922644974, + "loss": 0.7572, + "step": 953 + }, + { + "epoch": 0.14251568568867642, + "grad_norm": 0.1643514782190323, + "learning_rate": 0.00019031712533778137, + "loss": 0.8037, + "step": 954 + }, + { + "epoch": 0.14266507319988048, + "grad_norm": 0.1775134652853012, + "learning_rate": 0.0001902969384585818, + "loss": 0.6325, + "step": 955 + }, + { + "epoch": 0.14281446071108456, + "grad_norm": 0.16977538168430328, + "learning_rate": 0.00019027673163135827, + "loss": 0.4522, + "step": 956 + }, + { + "epoch": 0.1429638482222886, + "grad_norm": 0.2153913825750351, + "learning_rate": 0.00019025650486057484, + "loss": 0.6189, + "step": 957 + }, + { + "epoch": 0.1431132357334927, + "grad_norm": 0.12824203073978424, + "learning_rate": 0.00019023625815069989, + "loss": 0.5429, + "step": 958 + }, + { + "epoch": 0.14326262324469674, + "grad_norm": 0.13086850941181183, + "learning_rate": 0.00019021599150620625, + "loss": 0.854, + "step": 959 + }, + { + "epoch": 0.1434120107559008, + "grad_norm": 0.13822171092033386, + "learning_rate": 0.00019019570493157114, + "loss": 0.7447, + "step": 960 + }, + { + "epoch": 0.14356139826710487, + "grad_norm": 0.13974325358867645, + "learning_rate": 0.00019017539843127617, + "loss": 0.9451, + "step": 961 + }, + { + "epoch": 0.14371078577830892, + "grad_norm": 0.17371952533721924, + "learning_rate": 0.0001901550720098074, + "loss": 0.9114, + "step": 962 + }, + { + "epoch": 0.143860173289513, + "grad_norm": 0.2538747191429138, + "learning_rate": 0.00019013472567165523, + "loss": 1.1961, + "step": 963 + }, + { + "epoch": 0.14400956080071706, + "grad_norm": 0.110201895236969, + "learning_rate": 0.00019011435942131448, + "loss": 0.4677, + "step": 964 + }, + { + "epoch": 0.14415894831192114, + "grad_norm": 0.11606165021657944, + "learning_rate": 0.00019009397326328443, + "loss": 0.7407, + "step": 965 + }, + { + "epoch": 0.1443083358231252, + "grad_norm": 0.1762661188840866, + "learning_rate": 0.00019007356720206865, + "loss": 0.7177, + "step": 966 + }, + { + "epoch": 0.14445772333432924, + "grad_norm": 0.16712108254432678, + "learning_rate": 0.0001900531412421752, + "loss": 0.858, + "step": 967 + }, + { + "epoch": 0.14460711084553332, + "grad_norm": 0.17348961532115936, + "learning_rate": 0.00019003269538811647, + "loss": 1.1409, + "step": 968 + }, + { + "epoch": 0.14475649835673737, + "grad_norm": 0.13761204481124878, + "learning_rate": 0.0001900122296444093, + "loss": 0.6666, + "step": 969 + }, + { + "epoch": 0.14490588586794145, + "grad_norm": 0.20189805328845978, + "learning_rate": 0.00018999174401557488, + "loss": 0.736, + "step": 970 + }, + { + "epoch": 0.1450552733791455, + "grad_norm": 0.15712390840053558, + "learning_rate": 0.0001899712385061388, + "loss": 0.7429, + "step": 971 + }, + { + "epoch": 0.14520466089034956, + "grad_norm": 0.12312103062868118, + "learning_rate": 0.00018995071312063105, + "loss": 0.8887, + "step": 972 + }, + { + "epoch": 0.14535404840155364, + "grad_norm": 0.16373762488365173, + "learning_rate": 0.00018993016786358603, + "loss": 1.0923, + "step": 973 + }, + { + "epoch": 0.1455034359127577, + "grad_norm": 0.1286284178495407, + "learning_rate": 0.00018990960273954254, + "loss": 0.5527, + "step": 974 + }, + { + "epoch": 0.14565282342396177, + "grad_norm": 0.1538635641336441, + "learning_rate": 0.0001898890177530437, + "loss": 0.8092, + "step": 975 + }, + { + "epoch": 0.14580221093516582, + "grad_norm": 0.14671294391155243, + "learning_rate": 0.00018986841290863704, + "loss": 0.7165, + "step": 976 + }, + { + "epoch": 0.14595159844636987, + "grad_norm": 0.1533249318599701, + "learning_rate": 0.00018984778821087454, + "loss": 0.6822, + "step": 977 + }, + { + "epoch": 0.14610098595757395, + "grad_norm": 0.1579664945602417, + "learning_rate": 0.0001898271436643125, + "loss": 0.8128, + "step": 978 + }, + { + "epoch": 0.146250373468778, + "grad_norm": 0.1557191014289856, + "learning_rate": 0.00018980647927351166, + "loss": 0.8699, + "step": 979 + }, + { + "epoch": 0.14639976097998209, + "grad_norm": 0.1736837476491928, + "learning_rate": 0.00018978579504303706, + "loss": 1.2032, + "step": 980 + }, + { + "epoch": 0.14654914849118614, + "grad_norm": 0.17181052267551422, + "learning_rate": 0.00018976509097745826, + "loss": 0.7804, + "step": 981 + }, + { + "epoch": 0.1466985360023902, + "grad_norm": 0.20371553301811218, + "learning_rate": 0.000189744367081349, + "loss": 0.9971, + "step": 982 + }, + { + "epoch": 0.14684792351359427, + "grad_norm": 0.11608695983886719, + "learning_rate": 0.00018972362335928757, + "loss": 0.7474, + "step": 983 + }, + { + "epoch": 0.14699731102479832, + "grad_norm": 0.23304304480552673, + "learning_rate": 0.00018970285981585662, + "loss": 1.0192, + "step": 984 + }, + { + "epoch": 0.1471466985360024, + "grad_norm": 0.18036411702632904, + "learning_rate": 0.0001896820764556431, + "loss": 1.1699, + "step": 985 + }, + { + "epoch": 0.14729608604720645, + "grad_norm": 0.45688414573669434, + "learning_rate": 0.00018966127328323842, + "loss": 1.5615, + "step": 986 + }, + { + "epoch": 0.1474454735584105, + "grad_norm": 0.11910564452409744, + "learning_rate": 0.00018964045030323828, + "loss": 0.5669, + "step": 987 + }, + { + "epoch": 0.14759486106961459, + "grad_norm": 0.15722838044166565, + "learning_rate": 0.00018961960752024288, + "loss": 0.7162, + "step": 988 + }, + { + "epoch": 0.14774424858081864, + "grad_norm": 0.13349586725234985, + "learning_rate": 0.00018959874493885666, + "loss": 0.8232, + "step": 989 + }, + { + "epoch": 0.14789363609202272, + "grad_norm": 0.2732388973236084, + "learning_rate": 0.0001895778625636885, + "loss": 0.9254, + "step": 990 + }, + { + "epoch": 0.14804302360322677, + "grad_norm": 0.12720026075839996, + "learning_rate": 0.00018955696039935167, + "loss": 0.6276, + "step": 991 + }, + { + "epoch": 0.14819241111443082, + "grad_norm": 0.12968410551548004, + "learning_rate": 0.00018953603845046378, + "loss": 0.7688, + "step": 992 + }, + { + "epoch": 0.1483417986256349, + "grad_norm": 0.2596571445465088, + "learning_rate": 0.0001895150967216468, + "loss": 1.0049, + "step": 993 + }, + { + "epoch": 0.14849118613683895, + "grad_norm": 0.12322080880403519, + "learning_rate": 0.00018949413521752713, + "loss": 0.7845, + "step": 994 + }, + { + "epoch": 0.14864057364804303, + "grad_norm": 0.12585557997226715, + "learning_rate": 0.00018947315394273546, + "loss": 0.5276, + "step": 995 + }, + { + "epoch": 0.14878996115924709, + "grad_norm": 0.16490721702575684, + "learning_rate": 0.00018945215290190693, + "loss": 0.5467, + "step": 996 + }, + { + "epoch": 0.14893934867045114, + "grad_norm": 0.20895147323608398, + "learning_rate": 0.00018943113209968094, + "loss": 0.6746, + "step": 997 + }, + { + "epoch": 0.14908873618165522, + "grad_norm": 0.15641289949417114, + "learning_rate": 0.00018941009154070136, + "loss": 0.5641, + "step": 998 + }, + { + "epoch": 0.14923812369285927, + "grad_norm": 0.12262630462646484, + "learning_rate": 0.0001893890312296164, + "loss": 0.5929, + "step": 999 + }, + { + "epoch": 0.14938751120406335, + "grad_norm": 0.15614943206310272, + "learning_rate": 0.00018936795117107855, + "loss": 0.7035, + "step": 1000 + }, + { + "epoch": 0.1495368987152674, + "grad_norm": 0.13168151676654816, + "learning_rate": 0.00018934685136974482, + "loss": 0.7546, + "step": 1001 + }, + { + "epoch": 0.14968628622647145, + "grad_norm": 0.12817221879959106, + "learning_rate": 0.0001893257318302764, + "loss": 0.6542, + "step": 1002 + }, + { + "epoch": 0.14983567373767553, + "grad_norm": 0.22590519487857819, + "learning_rate": 0.000189304592557339, + "loss": 0.9049, + "step": 1003 + }, + { + "epoch": 0.1499850612488796, + "grad_norm": 0.1316809505224228, + "learning_rate": 0.00018928343355560258, + "loss": 0.8921, + "step": 1004 + }, + { + "epoch": 0.15013444876008367, + "grad_norm": 0.16760273277759552, + "learning_rate": 0.00018926225482974153, + "loss": 0.5918, + "step": 1005 + }, + { + "epoch": 0.15028383627128772, + "grad_norm": 0.1197691559791565, + "learning_rate": 0.00018924105638443452, + "loss": 0.6532, + "step": 1006 + }, + { + "epoch": 0.15043322378249177, + "grad_norm": 0.1392420530319214, + "learning_rate": 0.00018921983822436467, + "loss": 0.9638, + "step": 1007 + }, + { + "epoch": 0.15058261129369585, + "grad_norm": 0.17294703423976898, + "learning_rate": 0.0001891986003542194, + "loss": 0.7818, + "step": 1008 + }, + { + "epoch": 0.1507319988048999, + "grad_norm": 0.2133634239435196, + "learning_rate": 0.00018917734277869042, + "loss": 0.8933, + "step": 1009 + }, + { + "epoch": 0.15088138631610398, + "grad_norm": 0.13073426485061646, + "learning_rate": 0.00018915606550247397, + "loss": 0.7142, + "step": 1010 + }, + { + "epoch": 0.15103077382730803, + "grad_norm": 0.2918628454208374, + "learning_rate": 0.0001891347685302705, + "loss": 0.9469, + "step": 1011 + }, + { + "epoch": 0.1511801613385121, + "grad_norm": 0.12934882938861847, + "learning_rate": 0.0001891134518667848, + "loss": 0.8446, + "step": 1012 + }, + { + "epoch": 0.15132954884971617, + "grad_norm": 0.12790237367153168, + "learning_rate": 0.00018909211551672615, + "loss": 0.5309, + "step": 1013 + }, + { + "epoch": 0.15147893636092022, + "grad_norm": 0.2520146071910858, + "learning_rate": 0.000189070759484808, + "loss": 0.8729, + "step": 1014 + }, + { + "epoch": 0.1516283238721243, + "grad_norm": 0.1819266974925995, + "learning_rate": 0.00018904938377574827, + "loss": 0.8364, + "step": 1015 + }, + { + "epoch": 0.15177771138332835, + "grad_norm": 0.13670526444911957, + "learning_rate": 0.0001890279883942692, + "loss": 1.178, + "step": 1016 + }, + { + "epoch": 0.15192709889453243, + "grad_norm": 0.15499313175678253, + "learning_rate": 0.00018900657334509736, + "loss": 0.6144, + "step": 1017 + }, + { + "epoch": 0.15207648640573648, + "grad_norm": 0.13412916660308838, + "learning_rate": 0.0001889851386329637, + "loss": 0.7415, + "step": 1018 + }, + { + "epoch": 0.15222587391694054, + "grad_norm": 0.17868481576442719, + "learning_rate": 0.00018896368426260345, + "loss": 1.0732, + "step": 1019 + }, + { + "epoch": 0.15237526142814461, + "grad_norm": 0.197604700922966, + "learning_rate": 0.00018894221023875622, + "loss": 0.8012, + "step": 1020 + }, + { + "epoch": 0.15252464893934867, + "grad_norm": 0.13587865233421326, + "learning_rate": 0.000188920716566166, + "loss": 0.5255, + "step": 1021 + }, + { + "epoch": 0.15267403645055275, + "grad_norm": 0.13020993769168854, + "learning_rate": 0.00018889920324958106, + "loss": 0.5617, + "step": 1022 + }, + { + "epoch": 0.1528234239617568, + "grad_norm": 0.19373324513435364, + "learning_rate": 0.00018887767029375401, + "loss": 1.085, + "step": 1023 + }, + { + "epoch": 0.15297281147296085, + "grad_norm": 0.11815980821847916, + "learning_rate": 0.00018885611770344185, + "loss": 0.7142, + "step": 1024 + }, + { + "epoch": 0.15312219898416493, + "grad_norm": 0.21390031278133392, + "learning_rate": 0.00018883454548340587, + "loss": 1.0385, + "step": 1025 + }, + { + "epoch": 0.15327158649536898, + "grad_norm": 0.19770096242427826, + "learning_rate": 0.00018881295363841174, + "loss": 0.7931, + "step": 1026 + }, + { + "epoch": 0.15342097400657306, + "grad_norm": 0.1958816796541214, + "learning_rate": 0.00018879134217322942, + "loss": 0.8291, + "step": 1027 + }, + { + "epoch": 0.15357036151777712, + "grad_norm": 0.11775219440460205, + "learning_rate": 0.00018876971109263324, + "loss": 0.6206, + "step": 1028 + }, + { + "epoch": 0.15371974902898117, + "grad_norm": 0.11675494909286499, + "learning_rate": 0.00018874806040140185, + "loss": 0.5483, + "step": 1029 + }, + { + "epoch": 0.15386913654018525, + "grad_norm": 0.33525991439819336, + "learning_rate": 0.00018872639010431822, + "loss": 1.6754, + "step": 1030 + }, + { + "epoch": 0.1540185240513893, + "grad_norm": 0.15912093222141266, + "learning_rate": 0.00018870470020616967, + "loss": 0.7152, + "step": 1031 + }, + { + "epoch": 0.15416791156259338, + "grad_norm": 0.1797519475221634, + "learning_rate": 0.0001886829907117478, + "loss": 0.5595, + "step": 1032 + }, + { + "epoch": 0.15431729907379743, + "grad_norm": 0.163869708776474, + "learning_rate": 0.0001886612616258486, + "loss": 0.6443, + "step": 1033 + }, + { + "epoch": 0.15446668658500148, + "grad_norm": 0.21159547567367554, + "learning_rate": 0.00018863951295327244, + "loss": 0.8727, + "step": 1034 + }, + { + "epoch": 0.15461607409620556, + "grad_norm": 0.17695502936840057, + "learning_rate": 0.00018861774469882385, + "loss": 1.269, + "step": 1035 + }, + { + "epoch": 0.15476546160740962, + "grad_norm": 0.23800405859947205, + "learning_rate": 0.00018859595686731187, + "loss": 1.2047, + "step": 1036 + }, + { + "epoch": 0.1549148491186137, + "grad_norm": 0.11121930181980133, + "learning_rate": 0.0001885741494635497, + "loss": 0.6537, + "step": 1037 + }, + { + "epoch": 0.15506423662981775, + "grad_norm": 0.10217446833848953, + "learning_rate": 0.00018855232249235498, + "loss": 0.4512, + "step": 1038 + }, + { + "epoch": 0.1552136241410218, + "grad_norm": 0.2477329820394516, + "learning_rate": 0.0001885304759585496, + "loss": 0.8929, + "step": 1039 + }, + { + "epoch": 0.15536301165222588, + "grad_norm": 0.13516084849834442, + "learning_rate": 0.00018850860986695985, + "loss": 0.6883, + "step": 1040 + }, + { + "epoch": 0.15551239916342993, + "grad_norm": 0.14974398910999298, + "learning_rate": 0.0001884867242224163, + "loss": 0.6658, + "step": 1041 + }, + { + "epoch": 0.155661786674634, + "grad_norm": 0.12428376823663712, + "learning_rate": 0.00018846481902975377, + "loss": 0.8312, + "step": 1042 + }, + { + "epoch": 0.15581117418583806, + "grad_norm": 0.10979831218719482, + "learning_rate": 0.0001884428942938115, + "loss": 0.6737, + "step": 1043 + }, + { + "epoch": 0.15596056169704212, + "grad_norm": 0.18722845613956451, + "learning_rate": 0.00018842095001943306, + "loss": 1.0435, + "step": 1044 + }, + { + "epoch": 0.1561099492082462, + "grad_norm": 0.1588689237833023, + "learning_rate": 0.00018839898621146625, + "loss": 0.792, + "step": 1045 + }, + { + "epoch": 0.15625933671945025, + "grad_norm": 0.10412801802158356, + "learning_rate": 0.00018837700287476316, + "loss": 0.4995, + "step": 1046 + }, + { + "epoch": 0.15640872423065433, + "grad_norm": 0.1294378787279129, + "learning_rate": 0.00018835500001418036, + "loss": 0.8798, + "step": 1047 + }, + { + "epoch": 0.15655811174185838, + "grad_norm": 0.16280725598335266, + "learning_rate": 0.00018833297763457858, + "loss": 0.6571, + "step": 1048 + }, + { + "epoch": 0.15670749925306243, + "grad_norm": 0.12636205554008484, + "learning_rate": 0.0001883109357408229, + "loss": 0.5629, + "step": 1049 + }, + { + "epoch": 0.1568568867642665, + "grad_norm": 0.22952638566493988, + "learning_rate": 0.00018828887433778278, + "loss": 0.9414, + "step": 1050 + }, + { + "epoch": 0.15700627427547056, + "grad_norm": 0.18174390494823456, + "learning_rate": 0.00018826679343033186, + "loss": 0.6405, + "step": 1051 + }, + { + "epoch": 0.15715566178667464, + "grad_norm": 0.4626804292201996, + "learning_rate": 0.00018824469302334822, + "loss": 1.1709, + "step": 1052 + }, + { + "epoch": 0.1573050492978787, + "grad_norm": 0.17580588161945343, + "learning_rate": 0.00018822257312171416, + "loss": 0.7962, + "step": 1053 + }, + { + "epoch": 0.15745443680908275, + "grad_norm": 0.1333022564649582, + "learning_rate": 0.0001882004337303163, + "loss": 0.7548, + "step": 1054 + }, + { + "epoch": 0.15760382432028683, + "grad_norm": 1.2796801328659058, + "learning_rate": 0.00018817827485404564, + "loss": 1.9406, + "step": 1055 + }, + { + "epoch": 0.15775321183149088, + "grad_norm": 0.17332683503627777, + "learning_rate": 0.0001881560964977974, + "loss": 0.8372, + "step": 1056 + }, + { + "epoch": 0.15790259934269496, + "grad_norm": 0.15442058444023132, + "learning_rate": 0.0001881338986664711, + "loss": 0.8606, + "step": 1057 + }, + { + "epoch": 0.158051986853899, + "grad_norm": 0.14709503948688507, + "learning_rate": 0.0001881116813649706, + "loss": 0.8004, + "step": 1058 + }, + { + "epoch": 0.15820137436510306, + "grad_norm": 0.18778932094573975, + "learning_rate": 0.00018808944459820407, + "loss": 0.9717, + "step": 1059 + }, + { + "epoch": 0.15835076187630714, + "grad_norm": 0.1773632913827896, + "learning_rate": 0.00018806718837108402, + "loss": 0.981, + "step": 1060 + }, + { + "epoch": 0.1585001493875112, + "grad_norm": 0.13269007205963135, + "learning_rate": 0.00018804491268852708, + "loss": 0.8844, + "step": 1061 + }, + { + "epoch": 0.15864953689871528, + "grad_norm": 0.17859001457691193, + "learning_rate": 0.00018802261755545443, + "loss": 0.5105, + "step": 1062 + }, + { + "epoch": 0.15879892440991933, + "grad_norm": 0.1438104659318924, + "learning_rate": 0.0001880003029767913, + "loss": 0.7643, + "step": 1063 + }, + { + "epoch": 0.15894831192112338, + "grad_norm": 0.13282090425491333, + "learning_rate": 0.0001879779689574674, + "loss": 0.821, + "step": 1064 + }, + { + "epoch": 0.15909769943232746, + "grad_norm": 0.24478429555892944, + "learning_rate": 0.0001879556155024167, + "loss": 0.6762, + "step": 1065 + }, + { + "epoch": 0.1592470869435315, + "grad_norm": 0.11228256672620773, + "learning_rate": 0.00018793324261657737, + "loss": 0.8163, + "step": 1066 + }, + { + "epoch": 0.1593964744547356, + "grad_norm": 0.10839138180017471, + "learning_rate": 0.00018791085030489194, + "loss": 0.6536, + "step": 1067 + }, + { + "epoch": 0.15954586196593964, + "grad_norm": 0.14781688153743744, + "learning_rate": 0.00018788843857230726, + "loss": 0.7572, + "step": 1068 + }, + { + "epoch": 0.1596952494771437, + "grad_norm": 0.18171058595180511, + "learning_rate": 0.00018786600742377437, + "loss": 0.9202, + "step": 1069 + }, + { + "epoch": 0.15984463698834778, + "grad_norm": 0.11682315170764923, + "learning_rate": 0.00018784355686424876, + "loss": 0.7, + "step": 1070 + }, + { + "epoch": 0.15999402449955183, + "grad_norm": 0.09597928076982498, + "learning_rate": 0.00018782108689869006, + "loss": 0.621, + "step": 1071 + }, + { + "epoch": 0.1601434120107559, + "grad_norm": 0.15037846565246582, + "learning_rate": 0.00018779859753206225, + "loss": 0.7951, + "step": 1072 + }, + { + "epoch": 0.16029279952195996, + "grad_norm": 0.11888540536165237, + "learning_rate": 0.00018777608876933358, + "loss": 0.7986, + "step": 1073 + }, + { + "epoch": 0.16044218703316404, + "grad_norm": 0.4484766721725464, + "learning_rate": 0.00018775356061547662, + "loss": 1.231, + "step": 1074 + }, + { + "epoch": 0.1605915745443681, + "grad_norm": 0.13222385942935944, + "learning_rate": 0.00018773101307546815, + "loss": 0.7957, + "step": 1075 + }, + { + "epoch": 0.16074096205557215, + "grad_norm": 0.18500378727912903, + "learning_rate": 0.00018770844615428932, + "loss": 0.695, + "step": 1076 + }, + { + "epoch": 0.16089034956677623, + "grad_norm": 0.10332745313644409, + "learning_rate": 0.0001876858598569255, + "loss": 0.556, + "step": 1077 + }, + { + "epoch": 0.16103973707798028, + "grad_norm": 0.0973210334777832, + "learning_rate": 0.00018766325418836637, + "loss": 0.5246, + "step": 1078 + }, + { + "epoch": 0.16118912458918436, + "grad_norm": 0.15679813921451569, + "learning_rate": 0.00018764062915360588, + "loss": 0.7182, + "step": 1079 + }, + { + "epoch": 0.1613385121003884, + "grad_norm": 0.12232697755098343, + "learning_rate": 0.00018761798475764224, + "loss": 0.7503, + "step": 1080 + }, + { + "epoch": 0.16148789961159246, + "grad_norm": 0.13506333529949188, + "learning_rate": 0.00018759532100547799, + "loss": 0.6708, + "step": 1081 + }, + { + "epoch": 0.16163728712279654, + "grad_norm": 0.1789216846227646, + "learning_rate": 0.00018757263790211988, + "loss": 1.0662, + "step": 1082 + }, + { + "epoch": 0.1617866746340006, + "grad_norm": 0.1154765784740448, + "learning_rate": 0.00018754993545257894, + "loss": 0.6931, + "step": 1083 + }, + { + "epoch": 0.16193606214520467, + "grad_norm": 0.11015768349170685, + "learning_rate": 0.0001875272136618706, + "loss": 0.8244, + "step": 1084 + }, + { + "epoch": 0.16208544965640873, + "grad_norm": 0.14176665246486664, + "learning_rate": 0.00018750447253501436, + "loss": 0.8434, + "step": 1085 + }, + { + "epoch": 0.16223483716761278, + "grad_norm": 0.13398586213588715, + "learning_rate": 0.00018748171207703417, + "loss": 0.8307, + "step": 1086 + }, + { + "epoch": 0.16238422467881686, + "grad_norm": 0.18871843814849854, + "learning_rate": 0.00018745893229295813, + "loss": 0.7193, + "step": 1087 + }, + { + "epoch": 0.1625336121900209, + "grad_norm": 0.1976425051689148, + "learning_rate": 0.00018743613318781868, + "loss": 0.9341, + "step": 1088 + }, + { + "epoch": 0.162682999701225, + "grad_norm": 0.12541085481643677, + "learning_rate": 0.00018741331476665246, + "loss": 0.7364, + "step": 1089 + }, + { + "epoch": 0.16283238721242904, + "grad_norm": 0.3148552179336548, + "learning_rate": 0.00018739047703450048, + "loss": 1.0103, + "step": 1090 + }, + { + "epoch": 0.1629817747236331, + "grad_norm": 0.24794772267341614, + "learning_rate": 0.00018736761999640792, + "loss": 0.7772, + "step": 1091 + }, + { + "epoch": 0.16313116223483717, + "grad_norm": 0.12505854666233063, + "learning_rate": 0.00018734474365742428, + "loss": 0.7458, + "step": 1092 + }, + { + "epoch": 0.16328054974604123, + "grad_norm": 0.13425281643867493, + "learning_rate": 0.0001873218480226033, + "loss": 0.6537, + "step": 1093 + }, + { + "epoch": 0.1634299372572453, + "grad_norm": 0.29929280281066895, + "learning_rate": 0.00018729893309700295, + "loss": 1.2375, + "step": 1094 + }, + { + "epoch": 0.16357932476844936, + "grad_norm": 0.13596223294734955, + "learning_rate": 0.00018727599888568555, + "loss": 0.956, + "step": 1095 + }, + { + "epoch": 0.1637287122796534, + "grad_norm": 0.13480113446712494, + "learning_rate": 0.0001872530453937176, + "loss": 0.8181, + "step": 1096 + }, + { + "epoch": 0.1638780997908575, + "grad_norm": 0.14178095757961273, + "learning_rate": 0.00018723007262616998, + "loss": 0.7811, + "step": 1097 + }, + { + "epoch": 0.16402748730206154, + "grad_norm": 0.11537758260965347, + "learning_rate": 0.0001872070805881176, + "loss": 0.7824, + "step": 1098 + }, + { + "epoch": 0.16417687481326562, + "grad_norm": 0.15772311389446259, + "learning_rate": 0.00018718406928463986, + "loss": 0.679, + "step": 1099 + }, + { + "epoch": 0.16432626232446967, + "grad_norm": 0.22873377799987793, + "learning_rate": 0.00018716103872082026, + "loss": 0.4674, + "step": 1100 + }, + { + "epoch": 0.16447564983567373, + "grad_norm": 0.14255951344966888, + "learning_rate": 0.00018713798890174666, + "loss": 0.7014, + "step": 1101 + }, + { + "epoch": 0.1646250373468778, + "grad_norm": 0.15469695627689362, + "learning_rate": 0.00018711491983251113, + "loss": 0.8959, + "step": 1102 + }, + { + "epoch": 0.16477442485808186, + "grad_norm": 0.10443451255559921, + "learning_rate": 0.00018709183151820996, + "loss": 0.548, + "step": 1103 + }, + { + "epoch": 0.16492381236928594, + "grad_norm": 0.10996398329734802, + "learning_rate": 0.00018706872396394376, + "loss": 0.6857, + "step": 1104 + }, + { + "epoch": 0.16507319988049, + "grad_norm": 0.15236075222492218, + "learning_rate": 0.00018704559717481732, + "loss": 0.796, + "step": 1105 + }, + { + "epoch": 0.16522258739169404, + "grad_norm": 0.19824904203414917, + "learning_rate": 0.00018702245115593974, + "loss": 0.6438, + "step": 1106 + }, + { + "epoch": 0.16537197490289812, + "grad_norm": 0.17726042866706848, + "learning_rate": 0.00018699928591242438, + "loss": 0.9564, + "step": 1107 + }, + { + "epoch": 0.16552136241410217, + "grad_norm": 0.12704910337924957, + "learning_rate": 0.0001869761014493887, + "loss": 0.5796, + "step": 1108 + }, + { + "epoch": 0.16567074992530625, + "grad_norm": 0.22029517590999603, + "learning_rate": 0.00018695289777195462, + "loss": 0.7932, + "step": 1109 + }, + { + "epoch": 0.1658201374365103, + "grad_norm": 0.1251213401556015, + "learning_rate": 0.00018692967488524812, + "loss": 0.8287, + "step": 1110 + }, + { + "epoch": 0.16596952494771436, + "grad_norm": 0.16409267485141754, + "learning_rate": 0.00018690643279439958, + "loss": 0.6792, + "step": 1111 + }, + { + "epoch": 0.16611891245891844, + "grad_norm": 0.11526962369680405, + "learning_rate": 0.0001868831715045435, + "loss": 0.5228, + "step": 1112 + }, + { + "epoch": 0.1662682999701225, + "grad_norm": 0.14684948325157166, + "learning_rate": 0.00018685989102081867, + "loss": 0.6461, + "step": 1113 + }, + { + "epoch": 0.16641768748132657, + "grad_norm": 0.18166372179985046, + "learning_rate": 0.00018683659134836813, + "loss": 0.8579, + "step": 1114 + }, + { + "epoch": 0.16656707499253062, + "grad_norm": 0.12297386676073074, + "learning_rate": 0.00018681327249233913, + "loss": 0.8628, + "step": 1115 + }, + { + "epoch": 0.16671646250373467, + "grad_norm": 0.1687268614768982, + "learning_rate": 0.00018678993445788323, + "loss": 0.9859, + "step": 1116 + }, + { + "epoch": 0.16686585001493875, + "grad_norm": 0.10087472945451736, + "learning_rate": 0.00018676657725015606, + "loss": 0.6073, + "step": 1117 + }, + { + "epoch": 0.1670152375261428, + "grad_norm": 0.12146350741386414, + "learning_rate": 0.00018674320087431768, + "loss": 0.68, + "step": 1118 + }, + { + "epoch": 0.1671646250373469, + "grad_norm": 0.13258694112300873, + "learning_rate": 0.0001867198053355323, + "loss": 0.8528, + "step": 1119 + }, + { + "epoch": 0.16731401254855094, + "grad_norm": 0.15267039835453033, + "learning_rate": 0.00018669639063896836, + "loss": 0.7095, + "step": 1120 + }, + { + "epoch": 0.167463400059755, + "grad_norm": 0.12942974269390106, + "learning_rate": 0.00018667295678979852, + "loss": 0.8282, + "step": 1121 + }, + { + "epoch": 0.16761278757095907, + "grad_norm": 0.1521858274936676, + "learning_rate": 0.0001866495037931997, + "loss": 0.8557, + "step": 1122 + }, + { + "epoch": 0.16776217508216312, + "grad_norm": 0.1155681312084198, + "learning_rate": 0.000186626031654353, + "loss": 0.6572, + "step": 1123 + }, + { + "epoch": 0.1679115625933672, + "grad_norm": 0.13675054907798767, + "learning_rate": 0.00018660254037844388, + "loss": 0.7218, + "step": 1124 + }, + { + "epoch": 0.16806095010457126, + "grad_norm": 0.13993573188781738, + "learning_rate": 0.00018657902997066183, + "loss": 0.6353, + "step": 1125 + }, + { + "epoch": 0.16821033761577533, + "grad_norm": 0.16351041197776794, + "learning_rate": 0.00018655550043620073, + "loss": 1.027, + "step": 1126 + }, + { + "epoch": 0.1683597251269794, + "grad_norm": 0.2799745500087738, + "learning_rate": 0.00018653195178025864, + "loss": 0.7067, + "step": 1127 + }, + { + "epoch": 0.16850911263818344, + "grad_norm": 0.19105161726474762, + "learning_rate": 0.0001865083840080378, + "loss": 0.7683, + "step": 1128 + }, + { + "epoch": 0.16865850014938752, + "grad_norm": 0.22695805132389069, + "learning_rate": 0.00018648479712474468, + "loss": 1.0468, + "step": 1129 + }, + { + "epoch": 0.16880788766059157, + "grad_norm": 0.17338265478610992, + "learning_rate": 0.00018646119113559006, + "loss": 0.8898, + "step": 1130 + }, + { + "epoch": 0.16895727517179565, + "grad_norm": 0.12882818281650543, + "learning_rate": 0.00018643756604578885, + "loss": 0.8652, + "step": 1131 + }, + { + "epoch": 0.1691066626829997, + "grad_norm": 0.2222001552581787, + "learning_rate": 0.00018641392186056016, + "loss": 0.7363, + "step": 1132 + }, + { + "epoch": 0.16925605019420376, + "grad_norm": 0.17244432866573334, + "learning_rate": 0.00018639025858512741, + "loss": 0.5881, + "step": 1133 + }, + { + "epoch": 0.16940543770540784, + "grad_norm": 0.1207844614982605, + "learning_rate": 0.0001863665762247182, + "loss": 0.832, + "step": 1134 + }, + { + "epoch": 0.1695548252166119, + "grad_norm": 0.20044654607772827, + "learning_rate": 0.00018634287478456432, + "loss": 0.9488, + "step": 1135 + }, + { + "epoch": 0.16970421272781597, + "grad_norm": 0.13231255114078522, + "learning_rate": 0.00018631915426990184, + "loss": 0.5556, + "step": 1136 + }, + { + "epoch": 0.16985360023902002, + "grad_norm": 0.444367378950119, + "learning_rate": 0.00018629541468597092, + "loss": 1.1058, + "step": 1137 + }, + { + "epoch": 0.17000298775022407, + "grad_norm": 0.1951618492603302, + "learning_rate": 0.00018627165603801605, + "loss": 0.9483, + "step": 1138 + }, + { + "epoch": 0.17015237526142815, + "grad_norm": 0.2736690044403076, + "learning_rate": 0.00018624787833128588, + "loss": 0.8482, + "step": 1139 + }, + { + "epoch": 0.1703017627726322, + "grad_norm": 0.15237697958946228, + "learning_rate": 0.0001862240815710333, + "loss": 0.6764, + "step": 1140 + }, + { + "epoch": 0.17045115028383628, + "grad_norm": 0.14255677163600922, + "learning_rate": 0.0001862002657625154, + "loss": 0.7176, + "step": 1141 + }, + { + "epoch": 0.17060053779504034, + "grad_norm": 0.20565101504325867, + "learning_rate": 0.0001861764309109934, + "loss": 1.1075, + "step": 1142 + }, + { + "epoch": 0.1707499253062444, + "grad_norm": 0.11338746547698975, + "learning_rate": 0.0001861525770217329, + "loss": 0.8231, + "step": 1143 + }, + { + "epoch": 0.17089931281744847, + "grad_norm": 0.34042519330978394, + "learning_rate": 0.00018612870410000354, + "loss": 1.1565, + "step": 1144 + }, + { + "epoch": 0.17104870032865252, + "grad_norm": 0.12565311789512634, + "learning_rate": 0.00018610481215107925, + "loss": 0.6115, + "step": 1145 + }, + { + "epoch": 0.1711980878398566, + "grad_norm": 0.1262931525707245, + "learning_rate": 0.00018608090118023808, + "loss": 0.6307, + "step": 1146 + }, + { + "epoch": 0.17134747535106065, + "grad_norm": 0.4232707917690277, + "learning_rate": 0.00018605697119276242, + "loss": 0.8655, + "step": 1147 + }, + { + "epoch": 0.1714968628622647, + "grad_norm": 0.16684143245220184, + "learning_rate": 0.00018603302219393874, + "loss": 0.678, + "step": 1148 + }, + { + "epoch": 0.17164625037346878, + "grad_norm": 0.17487527430057526, + "learning_rate": 0.00018600905418905776, + "loss": 0.4725, + "step": 1149 + }, + { + "epoch": 0.17179563788467284, + "grad_norm": 0.1387338787317276, + "learning_rate": 0.0001859850671834144, + "loss": 0.8142, + "step": 1150 + }, + { + "epoch": 0.17194502539587692, + "grad_norm": 0.15392735600471497, + "learning_rate": 0.00018596106118230779, + "loss": 0.7672, + "step": 1151 + }, + { + "epoch": 0.17209441290708097, + "grad_norm": 0.251715749502182, + "learning_rate": 0.0001859370361910412, + "loss": 0.742, + "step": 1152 + }, + { + "epoch": 0.17224380041828502, + "grad_norm": 0.18483877182006836, + "learning_rate": 0.00018591299221492214, + "loss": 0.8683, + "step": 1153 + }, + { + "epoch": 0.1723931879294891, + "grad_norm": 0.14035139977931976, + "learning_rate": 0.00018588892925926228, + "loss": 0.6233, + "step": 1154 + }, + { + "epoch": 0.17254257544069315, + "grad_norm": 0.43573567271232605, + "learning_rate": 0.00018586484732937758, + "loss": 0.9892, + "step": 1155 + }, + { + "epoch": 0.17269196295189723, + "grad_norm": 0.11906327307224274, + "learning_rate": 0.00018584074643058807, + "loss": 0.7038, + "step": 1156 + }, + { + "epoch": 0.17284135046310128, + "grad_norm": 0.19945280253887177, + "learning_rate": 0.00018581662656821802, + "loss": 0.6789, + "step": 1157 + }, + { + "epoch": 0.17299073797430534, + "grad_norm": 0.25069642066955566, + "learning_rate": 0.00018579248774759586, + "loss": 0.9036, + "step": 1158 + }, + { + "epoch": 0.17314012548550942, + "grad_norm": 0.1946730762720108, + "learning_rate": 0.00018576832997405432, + "loss": 0.9371, + "step": 1159 + }, + { + "epoch": 0.17328951299671347, + "grad_norm": 0.16489769518375397, + "learning_rate": 0.00018574415325293018, + "loss": 0.5388, + "step": 1160 + }, + { + "epoch": 0.17343890050791755, + "grad_norm": 0.13044323027133942, + "learning_rate": 0.00018571995758956446, + "loss": 0.6756, + "step": 1161 + }, + { + "epoch": 0.1735882880191216, + "grad_norm": 0.13318844139575958, + "learning_rate": 0.00018569574298930237, + "loss": 0.9302, + "step": 1162 + }, + { + "epoch": 0.17373767553032565, + "grad_norm": 0.1677417755126953, + "learning_rate": 0.0001856715094574933, + "loss": 0.9213, + "step": 1163 + }, + { + "epoch": 0.17388706304152973, + "grad_norm": 0.11865703761577606, + "learning_rate": 0.00018564725699949083, + "loss": 0.8194, + "step": 1164 + }, + { + "epoch": 0.17403645055273378, + "grad_norm": 0.16482964158058167, + "learning_rate": 0.0001856229856206527, + "loss": 0.7889, + "step": 1165 + }, + { + "epoch": 0.17418583806393786, + "grad_norm": 0.13027256727218628, + "learning_rate": 0.0001855986953263409, + "loss": 0.7222, + "step": 1166 + }, + { + "epoch": 0.17433522557514192, + "grad_norm": 0.15511631965637207, + "learning_rate": 0.00018557438612192142, + "loss": 1.2247, + "step": 1167 + }, + { + "epoch": 0.17448461308634597, + "grad_norm": 0.16678637266159058, + "learning_rate": 0.00018555005801276463, + "loss": 0.5846, + "step": 1168 + }, + { + "epoch": 0.17463400059755005, + "grad_norm": 0.15397848188877106, + "learning_rate": 0.00018552571100424503, + "loss": 0.7163, + "step": 1169 + }, + { + "epoch": 0.1747833881087541, + "grad_norm": 0.3140554130077362, + "learning_rate": 0.00018550134510174115, + "loss": 1.1924, + "step": 1170 + }, + { + "epoch": 0.17493277561995818, + "grad_norm": 0.14024198055267334, + "learning_rate": 0.0001854769603106359, + "loss": 0.9647, + "step": 1171 + }, + { + "epoch": 0.17508216313116223, + "grad_norm": 0.2069367915391922, + "learning_rate": 0.0001854525566363162, + "loss": 0.9767, + "step": 1172 + }, + { + "epoch": 0.17523155064236628, + "grad_norm": 0.13816118240356445, + "learning_rate": 0.0001854281340841733, + "loss": 0.542, + "step": 1173 + }, + { + "epoch": 0.17538093815357036, + "grad_norm": 0.14822708070278168, + "learning_rate": 0.00018540369265960242, + "loss": 0.9574, + "step": 1174 + }, + { + "epoch": 0.17553032566477442, + "grad_norm": 0.1753738522529602, + "learning_rate": 0.00018537923236800315, + "loss": 0.8379, + "step": 1175 + }, + { + "epoch": 0.1756797131759785, + "grad_norm": 0.13829240202903748, + "learning_rate": 0.00018535475321477906, + "loss": 0.7342, + "step": 1176 + }, + { + "epoch": 0.17582910068718255, + "grad_norm": 0.15101328492164612, + "learning_rate": 0.00018533025520533805, + "loss": 0.8258, + "step": 1177 + }, + { + "epoch": 0.1759784881983866, + "grad_norm": 0.4592054784297943, + "learning_rate": 0.00018530573834509215, + "loss": 1.2556, + "step": 1178 + }, + { + "epoch": 0.17612787570959068, + "grad_norm": 0.1719399243593216, + "learning_rate": 0.00018528120263945744, + "loss": 0.9381, + "step": 1179 + }, + { + "epoch": 0.17627726322079473, + "grad_norm": 0.11263933032751083, + "learning_rate": 0.0001852566480938543, + "loss": 0.7793, + "step": 1180 + }, + { + "epoch": 0.1764266507319988, + "grad_norm": 0.16535364091396332, + "learning_rate": 0.00018523207471370716, + "loss": 0.6528, + "step": 1181 + }, + { + "epoch": 0.17657603824320287, + "grad_norm": 0.1505178064107895, + "learning_rate": 0.00018520748250444474, + "loss": 0.6552, + "step": 1182 + }, + { + "epoch": 0.17672542575440695, + "grad_norm": 0.12604665756225586, + "learning_rate": 0.00018518287147149983, + "loss": 0.6862, + "step": 1183 + }, + { + "epoch": 0.176874813265611, + "grad_norm": 0.12405974417924881, + "learning_rate": 0.00018515824162030934, + "loss": 0.7987, + "step": 1184 + }, + { + "epoch": 0.17702420077681505, + "grad_norm": 0.15863847732543945, + "learning_rate": 0.00018513359295631448, + "loss": 0.7515, + "step": 1185 + }, + { + "epoch": 0.17717358828801913, + "grad_norm": 0.16431525349617004, + "learning_rate": 0.00018510892548496047, + "loss": 0.6625, + "step": 1186 + }, + { + "epoch": 0.17732297579922318, + "grad_norm": 0.2458639144897461, + "learning_rate": 0.00018508423921169675, + "loss": 1.3889, + "step": 1187 + }, + { + "epoch": 0.17747236331042726, + "grad_norm": 0.1566503345966339, + "learning_rate": 0.00018505953414197696, + "loss": 0.8438, + "step": 1188 + }, + { + "epoch": 0.1776217508216313, + "grad_norm": 0.12780135869979858, + "learning_rate": 0.00018503481028125877, + "loss": 0.9441, + "step": 1189 + }, + { + "epoch": 0.17777113833283537, + "grad_norm": 0.2947339117527008, + "learning_rate": 0.00018501006763500414, + "loss": 1.2509, + "step": 1190 + }, + { + "epoch": 0.17792052584403945, + "grad_norm": 0.23375345766544342, + "learning_rate": 0.00018498530620867908, + "loss": 0.6705, + "step": 1191 + }, + { + "epoch": 0.1780699133552435, + "grad_norm": 0.2791805863380432, + "learning_rate": 0.00018496052600775376, + "loss": 0.9555, + "step": 1192 + }, + { + "epoch": 0.17821930086644758, + "grad_norm": 0.14220461249351501, + "learning_rate": 0.00018493572703770253, + "loss": 0.8023, + "step": 1193 + }, + { + "epoch": 0.17836868837765163, + "grad_norm": 0.10689699649810791, + "learning_rate": 0.0001849109093040039, + "loss": 0.3987, + "step": 1194 + }, + { + "epoch": 0.17851807588885568, + "grad_norm": 0.11420352756977081, + "learning_rate": 0.0001848860728121405, + "loss": 0.6444, + "step": 1195 + }, + { + "epoch": 0.17866746340005976, + "grad_norm": 0.1247362345457077, + "learning_rate": 0.00018486121756759906, + "loss": 0.8246, + "step": 1196 + }, + { + "epoch": 0.1788168509112638, + "grad_norm": 0.15353257954120636, + "learning_rate": 0.00018483634357587057, + "loss": 0.5427, + "step": 1197 + }, + { + "epoch": 0.1789662384224679, + "grad_norm": 0.11908576637506485, + "learning_rate": 0.00018481145084245002, + "loss": 0.671, + "step": 1198 + }, + { + "epoch": 0.17911562593367195, + "grad_norm": 0.1491960883140564, + "learning_rate": 0.0001847865393728366, + "loss": 0.9276, + "step": 1199 + }, + { + "epoch": 0.179265013444876, + "grad_norm": 0.16004832088947296, + "learning_rate": 0.00018476160917253373, + "loss": 0.9802, + "step": 1200 + }, + { + "epoch": 0.17941440095608008, + "grad_norm": 0.20184099674224854, + "learning_rate": 0.00018473666024704883, + "loss": 0.8308, + "step": 1201 + }, + { + "epoch": 0.17956378846728413, + "grad_norm": 0.16696742177009583, + "learning_rate": 0.0001847116926018935, + "loss": 1.1185, + "step": 1202 + }, + { + "epoch": 0.1797131759784882, + "grad_norm": 0.1517256200313568, + "learning_rate": 0.00018468670624258353, + "loss": 0.7536, + "step": 1203 + }, + { + "epoch": 0.17986256348969226, + "grad_norm": 0.16679219901561737, + "learning_rate": 0.0001846617011746388, + "loss": 0.6424, + "step": 1204 + }, + { + "epoch": 0.18001195100089631, + "grad_norm": 0.15376229584217072, + "learning_rate": 0.0001846366774035833, + "loss": 0.9011, + "step": 1205 + }, + { + "epoch": 0.1801613385121004, + "grad_norm": 0.13043151795864105, + "learning_rate": 0.00018461163493494517, + "loss": 0.9194, + "step": 1206 + }, + { + "epoch": 0.18031072602330445, + "grad_norm": 0.15427030622959137, + "learning_rate": 0.0001845865737742567, + "loss": 0.8592, + "step": 1207 + }, + { + "epoch": 0.18046011353450853, + "grad_norm": 0.4822232127189636, + "learning_rate": 0.0001845614939270543, + "loss": 1.6313, + "step": 1208 + }, + { + "epoch": 0.18060950104571258, + "grad_norm": 0.15565815567970276, + "learning_rate": 0.00018453639539887853, + "loss": 0.7739, + "step": 1209 + }, + { + "epoch": 0.18075888855691663, + "grad_norm": 0.11118531227111816, + "learning_rate": 0.00018451127819527402, + "loss": 0.6064, + "step": 1210 + }, + { + "epoch": 0.1809082760681207, + "grad_norm": 0.14477503299713135, + "learning_rate": 0.0001844861423217896, + "loss": 0.9683, + "step": 1211 + }, + { + "epoch": 0.18105766357932476, + "grad_norm": 0.1929170936346054, + "learning_rate": 0.00018446098778397807, + "loss": 0.6135, + "step": 1212 + }, + { + "epoch": 0.18120705109052884, + "grad_norm": 0.13953596353530884, + "learning_rate": 0.0001844358145873966, + "loss": 1.0582, + "step": 1213 + }, + { + "epoch": 0.1813564386017329, + "grad_norm": 0.1348966360092163, + "learning_rate": 0.00018441062273760628, + "loss": 0.4698, + "step": 1214 + }, + { + "epoch": 0.18150582611293695, + "grad_norm": 0.1345159113407135, + "learning_rate": 0.0001843854122401724, + "loss": 0.848, + "step": 1215 + }, + { + "epoch": 0.18165521362414103, + "grad_norm": 0.11462259292602539, + "learning_rate": 0.00018436018310066435, + "loss": 0.6769, + "step": 1216 + }, + { + "epoch": 0.18180460113534508, + "grad_norm": 0.14895157516002655, + "learning_rate": 0.0001843349353246557, + "loss": 0.5784, + "step": 1217 + }, + { + "epoch": 0.18195398864654916, + "grad_norm": 0.13859333097934723, + "learning_rate": 0.000184309668917724, + "loss": 0.903, + "step": 1218 + }, + { + "epoch": 0.1821033761577532, + "grad_norm": 0.11198470741510391, + "learning_rate": 0.00018428438388545107, + "loss": 0.6247, + "step": 1219 + }, + { + "epoch": 0.18225276366895726, + "grad_norm": 0.17095430195331573, + "learning_rate": 0.0001842590802334227, + "loss": 0.5969, + "step": 1220 + }, + { + "epoch": 0.18240215118016134, + "grad_norm": 0.33381956815719604, + "learning_rate": 0.00018423375796722895, + "loss": 1.1353, + "step": 1221 + }, + { + "epoch": 0.1825515386913654, + "grad_norm": 0.11893988400697708, + "learning_rate": 0.00018420841709246383, + "loss": 0.9319, + "step": 1222 + }, + { + "epoch": 0.18270092620256947, + "grad_norm": 0.11547727882862091, + "learning_rate": 0.00018418305761472566, + "loss": 0.5685, + "step": 1223 + }, + { + "epoch": 0.18285031371377353, + "grad_norm": 0.18776455521583557, + "learning_rate": 0.0001841576795396166, + "loss": 0.7411, + "step": 1224 + }, + { + "epoch": 0.18299970122497758, + "grad_norm": 0.28690654039382935, + "learning_rate": 0.00018413228287274315, + "loss": 0.965, + "step": 1225 + }, + { + "epoch": 0.18314908873618166, + "grad_norm": 0.12270329147577286, + "learning_rate": 0.00018410686761971586, + "loss": 0.6724, + "step": 1226 + }, + { + "epoch": 0.1832984762473857, + "grad_norm": 0.19342897832393646, + "learning_rate": 0.0001840814337861493, + "loss": 0.8486, + "step": 1227 + }, + { + "epoch": 0.1834478637585898, + "grad_norm": 0.12105520814657211, + "learning_rate": 0.00018405598137766224, + "loss": 0.9189, + "step": 1228 + }, + { + "epoch": 0.18359725126979384, + "grad_norm": 0.14361943304538727, + "learning_rate": 0.00018403051039987754, + "loss": 0.4516, + "step": 1229 + }, + { + "epoch": 0.1837466387809979, + "grad_norm": 0.23831139504909515, + "learning_rate": 0.00018400502085842208, + "loss": 1.2297, + "step": 1230 + }, + { + "epoch": 0.18389602629220198, + "grad_norm": 0.19276589155197144, + "learning_rate": 0.00018397951275892695, + "loss": 0.6379, + "step": 1231 + }, + { + "epoch": 0.18404541380340603, + "grad_norm": 0.32586970925331116, + "learning_rate": 0.00018395398610702733, + "loss": 1.0752, + "step": 1232 + }, + { + "epoch": 0.1841948013146101, + "grad_norm": 0.1523769348859787, + "learning_rate": 0.00018392844090836237, + "loss": 1.0371, + "step": 1233 + }, + { + "epoch": 0.18434418882581416, + "grad_norm": 0.1339939385652542, + "learning_rate": 0.00018390287716857546, + "loss": 0.5838, + "step": 1234 + }, + { + "epoch": 0.18449357633701824, + "grad_norm": 0.17885640263557434, + "learning_rate": 0.00018387729489331402, + "loss": 1.0965, + "step": 1235 + }, + { + "epoch": 0.1846429638482223, + "grad_norm": 0.2260708212852478, + "learning_rate": 0.00018385169408822964, + "loss": 1.0085, + "step": 1236 + }, + { + "epoch": 0.18479235135942634, + "grad_norm": 0.20628871023654938, + "learning_rate": 0.00018382607475897787, + "loss": 0.4401, + "step": 1237 + }, + { + "epoch": 0.18494173887063042, + "grad_norm": 0.17405936121940613, + "learning_rate": 0.0001838004369112184, + "loss": 0.5253, + "step": 1238 + }, + { + "epoch": 0.18509112638183448, + "grad_norm": 0.11227359622716904, + "learning_rate": 0.00018377478055061516, + "loss": 0.6422, + "step": 1239 + }, + { + "epoch": 0.18524051389303856, + "grad_norm": 0.18113653361797333, + "learning_rate": 0.00018374910568283594, + "loss": 0.5326, + "step": 1240 + }, + { + "epoch": 0.1853899014042426, + "grad_norm": 0.12269013375043869, + "learning_rate": 0.00018372341231355274, + "loss": 0.6607, + "step": 1241 + }, + { + "epoch": 0.18553928891544666, + "grad_norm": 0.31763309240341187, + "learning_rate": 0.00018369770044844168, + "loss": 0.8666, + "step": 1242 + }, + { + "epoch": 0.18568867642665074, + "grad_norm": 0.17516912519931793, + "learning_rate": 0.00018367197009318286, + "loss": 1.0452, + "step": 1243 + }, + { + "epoch": 0.1858380639378548, + "grad_norm": 0.11332941800355911, + "learning_rate": 0.00018364622125346055, + "loss": 0.4925, + "step": 1244 + }, + { + "epoch": 0.18598745144905887, + "grad_norm": 0.1446741372346878, + "learning_rate": 0.0001836204539349631, + "loss": 0.6967, + "step": 1245 + }, + { + "epoch": 0.18613683896026292, + "grad_norm": 0.177836611866951, + "learning_rate": 0.0001835946681433829, + "loss": 0.8671, + "step": 1246 + }, + { + "epoch": 0.18628622647146698, + "grad_norm": 0.1338125616312027, + "learning_rate": 0.00018356886388441645, + "loss": 0.9787, + "step": 1247 + }, + { + "epoch": 0.18643561398267106, + "grad_norm": 0.1504998356103897, + "learning_rate": 0.00018354304116376425, + "loss": 0.7123, + "step": 1248 + }, + { + "epoch": 0.1865850014938751, + "grad_norm": 0.11992106586694717, + "learning_rate": 0.00018351719998713106, + "loss": 0.8016, + "step": 1249 + }, + { + "epoch": 0.1867343890050792, + "grad_norm": 0.13362877070903778, + "learning_rate": 0.0001834913403602255, + "loss": 0.6951, + "step": 1250 + }, + { + "epoch": 0.18688377651628324, + "grad_norm": 0.19289879500865936, + "learning_rate": 0.00018346546228876047, + "loss": 0.9931, + "step": 1251 + }, + { + "epoch": 0.1870331640274873, + "grad_norm": 0.19565518200397491, + "learning_rate": 0.00018343956577845276, + "loss": 1.0326, + "step": 1252 + }, + { + "epoch": 0.18718255153869137, + "grad_norm": 0.2017303854227066, + "learning_rate": 0.00018341365083502335, + "loss": 0.8296, + "step": 1253 + }, + { + "epoch": 0.18733193904989542, + "grad_norm": 0.14418242871761322, + "learning_rate": 0.00018338771746419726, + "loss": 0.9156, + "step": 1254 + }, + { + "epoch": 0.1874813265610995, + "grad_norm": 0.30200129747390747, + "learning_rate": 0.00018336176567170363, + "loss": 0.8995, + "step": 1255 + }, + { + "epoch": 0.18763071407230356, + "grad_norm": 0.2137260138988495, + "learning_rate": 0.00018333579546327556, + "loss": 0.9876, + "step": 1256 + }, + { + "epoch": 0.1877801015835076, + "grad_norm": 0.1436002552509308, + "learning_rate": 0.0001833098068446503, + "loss": 0.7803, + "step": 1257 + }, + { + "epoch": 0.1879294890947117, + "grad_norm": 0.213914155960083, + "learning_rate": 0.00018328379982156915, + "loss": 0.8788, + "step": 1258 + }, + { + "epoch": 0.18807887660591574, + "grad_norm": 0.18235763907432556, + "learning_rate": 0.00018325777439977747, + "loss": 0.6727, + "step": 1259 + }, + { + "epoch": 0.18822826411711982, + "grad_norm": 0.184591144323349, + "learning_rate": 0.00018323173058502472, + "loss": 0.7196, + "step": 1260 + }, + { + "epoch": 0.18837765162832387, + "grad_norm": 0.1435452103614807, + "learning_rate": 0.00018320566838306431, + "loss": 0.9443, + "step": 1261 + }, + { + "epoch": 0.18852703913952792, + "grad_norm": 0.1862800419330597, + "learning_rate": 0.00018317958779965387, + "loss": 0.704, + "step": 1262 + }, + { + "epoch": 0.188676426650732, + "grad_norm": 0.1972978413105011, + "learning_rate": 0.000183153488840555, + "loss": 0.7077, + "step": 1263 + }, + { + "epoch": 0.18882581416193606, + "grad_norm": 0.13390903174877167, + "learning_rate": 0.00018312737151153334, + "loss": 0.658, + "step": 1264 + }, + { + "epoch": 0.18897520167314014, + "grad_norm": 0.3871189057826996, + "learning_rate": 0.00018310123581835868, + "loss": 1.2051, + "step": 1265 + }, + { + "epoch": 0.1891245891843442, + "grad_norm": 0.15519294142723083, + "learning_rate": 0.00018307508176680472, + "loss": 0.6735, + "step": 1266 + }, + { + "epoch": 0.18927397669554824, + "grad_norm": 0.13292644917964935, + "learning_rate": 0.0001830489093626494, + "loss": 0.5655, + "step": 1267 + }, + { + "epoch": 0.18942336420675232, + "grad_norm": 0.12040957808494568, + "learning_rate": 0.00018302271861167456, + "loss": 0.6574, + "step": 1268 + }, + { + "epoch": 0.18957275171795637, + "grad_norm": 0.17542392015457153, + "learning_rate": 0.00018299650951966614, + "loss": 0.7045, + "step": 1269 + }, + { + "epoch": 0.18972213922916045, + "grad_norm": 0.12977173924446106, + "learning_rate": 0.0001829702820924142, + "loss": 0.8453, + "step": 1270 + }, + { + "epoch": 0.1898715267403645, + "grad_norm": 0.3114042282104492, + "learning_rate": 0.00018294403633571275, + "loss": 0.9561, + "step": 1271 + }, + { + "epoch": 0.19002091425156856, + "grad_norm": 0.1601576954126358, + "learning_rate": 0.00018291777225535994, + "loss": 0.8407, + "step": 1272 + }, + { + "epoch": 0.19017030176277264, + "grad_norm": 0.12280754745006561, + "learning_rate": 0.00018289148985715784, + "loss": 0.5095, + "step": 1273 + }, + { + "epoch": 0.1903196892739767, + "grad_norm": 0.12627717852592468, + "learning_rate": 0.00018286518914691272, + "loss": 0.5383, + "step": 1274 + }, + { + "epoch": 0.19046907678518077, + "grad_norm": 0.13007701933383942, + "learning_rate": 0.0001828388701304348, + "loss": 0.8532, + "step": 1275 + }, + { + "epoch": 0.19061846429638482, + "grad_norm": 0.13100433349609375, + "learning_rate": 0.00018281253281353838, + "loss": 0.822, + "step": 1276 + }, + { + "epoch": 0.19076785180758887, + "grad_norm": 0.1324629783630371, + "learning_rate": 0.0001827861772020418, + "loss": 0.7904, + "step": 1277 + }, + { + "epoch": 0.19091723931879295, + "grad_norm": 0.13711746037006378, + "learning_rate": 0.00018275980330176737, + "loss": 0.498, + "step": 1278 + }, + { + "epoch": 0.191066626829997, + "grad_norm": 0.12584207952022552, + "learning_rate": 0.00018273341111854153, + "loss": 1.0128, + "step": 1279 + }, + { + "epoch": 0.19121601434120108, + "grad_norm": 0.20175109803676605, + "learning_rate": 0.00018270700065819477, + "loss": 0.6671, + "step": 1280 + }, + { + "epoch": 0.19136540185240514, + "grad_norm": 0.17598234117031097, + "learning_rate": 0.00018268057192656156, + "loss": 1.0968, + "step": 1281 + }, + { + "epoch": 0.1915147893636092, + "grad_norm": 0.13300736248493195, + "learning_rate": 0.00018265412492948042, + "loss": 0.7158, + "step": 1282 + }, + { + "epoch": 0.19166417687481327, + "grad_norm": 0.1788729578256607, + "learning_rate": 0.00018262765967279386, + "loss": 0.6676, + "step": 1283 + }, + { + "epoch": 0.19181356438601732, + "grad_norm": 0.23207572102546692, + "learning_rate": 0.0001826011761623486, + "loss": 0.9164, + "step": 1284 + }, + { + "epoch": 0.1919629518972214, + "grad_norm": 0.1788550764322281, + "learning_rate": 0.00018257467440399515, + "loss": 1.1415, + "step": 1285 + }, + { + "epoch": 0.19211233940842545, + "grad_norm": 0.10894573479890823, + "learning_rate": 0.0001825481544035882, + "loss": 0.716, + "step": 1286 + }, + { + "epoch": 0.1922617269196295, + "grad_norm": 0.12275734543800354, + "learning_rate": 0.00018252161616698646, + "loss": 0.6821, + "step": 1287 + }, + { + "epoch": 0.19241111443083359, + "grad_norm": 0.12295856326818466, + "learning_rate": 0.00018249505970005262, + "loss": 0.6685, + "step": 1288 + }, + { + "epoch": 0.19256050194203764, + "grad_norm": 0.25377437472343445, + "learning_rate": 0.00018246848500865347, + "loss": 0.9663, + "step": 1289 + }, + { + "epoch": 0.19270988945324172, + "grad_norm": 0.12609486281871796, + "learning_rate": 0.00018244189209865974, + "loss": 0.8103, + "step": 1290 + }, + { + "epoch": 0.19285927696444577, + "grad_norm": 0.18513751029968262, + "learning_rate": 0.0001824152809759462, + "loss": 1.1248, + "step": 1291 + }, + { + "epoch": 0.19300866447564985, + "grad_norm": 0.13006910681724548, + "learning_rate": 0.00018238865164639173, + "loss": 0.6911, + "step": 1292 + }, + { + "epoch": 0.1931580519868539, + "grad_norm": 0.14160241186618805, + "learning_rate": 0.00018236200411587915, + "loss": 0.5689, + "step": 1293 + }, + { + "epoch": 0.19330743949805795, + "grad_norm": 0.17542794346809387, + "learning_rate": 0.0001823353383902953, + "loss": 0.6207, + "step": 1294 + }, + { + "epoch": 0.19345682700926203, + "grad_norm": 0.251878559589386, + "learning_rate": 0.00018230865447553107, + "loss": 0.9893, + "step": 1295 + }, + { + "epoch": 0.19360621452046609, + "grad_norm": 0.14598213136196136, + "learning_rate": 0.0001822819523774814, + "loss": 0.823, + "step": 1296 + }, + { + "epoch": 0.19375560203167017, + "grad_norm": 0.14267736673355103, + "learning_rate": 0.0001822552321020451, + "loss": 0.7624, + "step": 1297 + }, + { + "epoch": 0.19390498954287422, + "grad_norm": 0.2022227942943573, + "learning_rate": 0.00018222849365512523, + "loss": 0.8711, + "step": 1298 + }, + { + "epoch": 0.19405437705407827, + "grad_norm": 0.29249998927116394, + "learning_rate": 0.0001822017370426287, + "loss": 0.9054, + "step": 1299 + }, + { + "epoch": 0.19420376456528235, + "grad_norm": 0.12678225338459015, + "learning_rate": 0.0001821749622704664, + "loss": 0.6689, + "step": 1300 + }, + { + "epoch": 0.1943531520764864, + "grad_norm": 0.1714026778936386, + "learning_rate": 0.00018214816934455333, + "loss": 0.7199, + "step": 1301 + }, + { + "epoch": 0.19450253958769048, + "grad_norm": 0.18611545860767365, + "learning_rate": 0.00018212135827080857, + "loss": 0.9119, + "step": 1302 + }, + { + "epoch": 0.19465192709889453, + "grad_norm": 0.12778407335281372, + "learning_rate": 0.00018209452905515496, + "loss": 0.435, + "step": 1303 + }, + { + "epoch": 0.19480131461009859, + "grad_norm": 0.3280845880508423, + "learning_rate": 0.00018206768170351962, + "loss": 0.9022, + "step": 1304 + }, + { + "epoch": 0.19495070212130267, + "grad_norm": 0.13167747855186462, + "learning_rate": 0.0001820408162218335, + "loss": 0.8323, + "step": 1305 + }, + { + "epoch": 0.19510008963250672, + "grad_norm": 0.14176693558692932, + "learning_rate": 0.0001820139326160316, + "loss": 0.5333, + "step": 1306 + }, + { + "epoch": 0.1952494771437108, + "grad_norm": 0.20718522369861603, + "learning_rate": 0.00018198703089205293, + "loss": 0.8403, + "step": 1307 + }, + { + "epoch": 0.19539886465491485, + "grad_norm": 0.15032266080379486, + "learning_rate": 0.00018196011105584058, + "loss": 1.1825, + "step": 1308 + }, + { + "epoch": 0.1955482521661189, + "grad_norm": 0.13833296298980713, + "learning_rate": 0.0001819331731133415, + "loss": 1.0256, + "step": 1309 + }, + { + "epoch": 0.19569763967732298, + "grad_norm": 0.15338090062141418, + "learning_rate": 0.00018190621707050671, + "loss": 0.7619, + "step": 1310 + }, + { + "epoch": 0.19584702718852703, + "grad_norm": 0.15736734867095947, + "learning_rate": 0.00018187924293329124, + "loss": 0.9195, + "step": 1311 + }, + { + "epoch": 0.19599641469973111, + "grad_norm": 0.1757577806711197, + "learning_rate": 0.0001818522507076541, + "loss": 0.8208, + "step": 1312 + }, + { + "epoch": 0.19614580221093517, + "grad_norm": 0.16938672959804535, + "learning_rate": 0.00018182524039955832, + "loss": 0.6394, + "step": 1313 + }, + { + "epoch": 0.19629518972213922, + "grad_norm": 0.11935129016637802, + "learning_rate": 0.00018179821201497092, + "loss": 0.8837, + "step": 1314 + }, + { + "epoch": 0.1964445772333433, + "grad_norm": 0.1422419250011444, + "learning_rate": 0.00018177116555986283, + "loss": 0.8579, + "step": 1315 + }, + { + "epoch": 0.19659396474454735, + "grad_norm": 0.16648200154304504, + "learning_rate": 0.0001817441010402091, + "loss": 0.4853, + "step": 1316 + }, + { + "epoch": 0.19674335225575143, + "grad_norm": 0.16787053644657135, + "learning_rate": 0.00018171701846198866, + "loss": 0.5068, + "step": 1317 + }, + { + "epoch": 0.19689273976695548, + "grad_norm": 0.13628660142421722, + "learning_rate": 0.00018168991783118452, + "loss": 0.7062, + "step": 1318 + }, + { + "epoch": 0.19704212727815953, + "grad_norm": 0.11356477439403534, + "learning_rate": 0.00018166279915378364, + "loss": 0.5145, + "step": 1319 + }, + { + "epoch": 0.19719151478936361, + "grad_norm": 0.1292952299118042, + "learning_rate": 0.00018163566243577697, + "loss": 0.7452, + "step": 1320 + }, + { + "epoch": 0.19734090230056767, + "grad_norm": 0.17471875250339508, + "learning_rate": 0.00018160850768315941, + "loss": 0.9093, + "step": 1321 + }, + { + "epoch": 0.19749028981177175, + "grad_norm": 0.1547754555940628, + "learning_rate": 0.0001815813349019299, + "loss": 0.6851, + "step": 1322 + }, + { + "epoch": 0.1976396773229758, + "grad_norm": 0.11787121742963791, + "learning_rate": 0.00018155414409809132, + "loss": 0.8243, + "step": 1323 + }, + { + "epoch": 0.19778906483417985, + "grad_norm": 0.13274557888507843, + "learning_rate": 0.00018152693527765057, + "loss": 0.7042, + "step": 1324 + }, + { + "epoch": 0.19793845234538393, + "grad_norm": 0.1521092653274536, + "learning_rate": 0.00018149970844661849, + "loss": 0.5151, + "step": 1325 + }, + { + "epoch": 0.19808783985658798, + "grad_norm": 0.13765467703342438, + "learning_rate": 0.0001814724636110099, + "loss": 0.8489, + "step": 1326 + }, + { + "epoch": 0.19823722736779206, + "grad_norm": 0.26832038164138794, + "learning_rate": 0.00018144520077684369, + "loss": 0.7142, + "step": 1327 + }, + { + "epoch": 0.19838661487899611, + "grad_norm": 0.10923701524734497, + "learning_rate": 0.00018141791995014255, + "loss": 0.6142, + "step": 1328 + }, + { + "epoch": 0.19853600239020017, + "grad_norm": 0.1848011165857315, + "learning_rate": 0.00018139062113693333, + "loss": 0.8898, + "step": 1329 + }, + { + "epoch": 0.19868538990140425, + "grad_norm": 0.12784364819526672, + "learning_rate": 0.00018136330434324674, + "loss": 0.8846, + "step": 1330 + }, + { + "epoch": 0.1988347774126083, + "grad_norm": 0.13698415458202362, + "learning_rate": 0.00018133596957511748, + "loss": 0.7992, + "step": 1331 + }, + { + "epoch": 0.19898416492381238, + "grad_norm": 0.16566555202007294, + "learning_rate": 0.00018130861683858426, + "loss": 0.8587, + "step": 1332 + }, + { + "epoch": 0.19913355243501643, + "grad_norm": 0.12000430375337601, + "learning_rate": 0.0001812812461396897, + "loss": 0.815, + "step": 1333 + }, + { + "epoch": 0.19928293994622048, + "grad_norm": 0.11932047456502914, + "learning_rate": 0.00018125385748448048, + "loss": 0.8236, + "step": 1334 + }, + { + "epoch": 0.19943232745742456, + "grad_norm": 0.14856404066085815, + "learning_rate": 0.00018122645087900708, + "loss": 0.5428, + "step": 1335 + }, + { + "epoch": 0.19958171496862862, + "grad_norm": 0.22421959042549133, + "learning_rate": 0.00018119902632932416, + "loss": 1.4216, + "step": 1336 + }, + { + "epoch": 0.1997311024798327, + "grad_norm": 0.1441725790500641, + "learning_rate": 0.00018117158384149023, + "loss": 1.0382, + "step": 1337 + }, + { + "epoch": 0.19988048999103675, + "grad_norm": 0.14890751242637634, + "learning_rate": 0.0001811441234215677, + "loss": 0.6485, + "step": 1338 + }, + { + "epoch": 0.2000298775022408, + "grad_norm": 0.24321304261684418, + "learning_rate": 0.00018111664507562304, + "loss": 0.8343, + "step": 1339 + }, + { + "epoch": 0.20017926501344488, + "grad_norm": 0.20097461342811584, + "learning_rate": 0.0001810891488097267, + "loss": 0.9692, + "step": 1340 + }, + { + "epoch": 0.20032865252464893, + "grad_norm": 0.15573135018348694, + "learning_rate": 0.00018106163462995297, + "loss": 0.5672, + "step": 1341 + }, + { + "epoch": 0.200478040035853, + "grad_norm": 0.1436656415462494, + "learning_rate": 0.00018103410254238021, + "loss": 0.7355, + "step": 1342 + }, + { + "epoch": 0.20062742754705706, + "grad_norm": 0.28203514218330383, + "learning_rate": 0.00018100655255309068, + "loss": 1.5894, + "step": 1343 + }, + { + "epoch": 0.20077681505826112, + "grad_norm": 0.15477240085601807, + "learning_rate": 0.0001809789846681706, + "loss": 0.9943, + "step": 1344 + }, + { + "epoch": 0.2009262025694652, + "grad_norm": 0.20752249658107758, + "learning_rate": 0.0001809513988937102, + "loss": 0.6022, + "step": 1345 + }, + { + "epoch": 0.20107559008066925, + "grad_norm": 0.12718671560287476, + "learning_rate": 0.00018092379523580357, + "loss": 0.7869, + "step": 1346 + }, + { + "epoch": 0.20122497759187333, + "grad_norm": 0.14520499110221863, + "learning_rate": 0.0001808961737005488, + "loss": 0.8307, + "step": 1347 + }, + { + "epoch": 0.20137436510307738, + "grad_norm": 0.18366779386997223, + "learning_rate": 0.00018086853429404793, + "loss": 1.0939, + "step": 1348 + }, + { + "epoch": 0.20152375261428146, + "grad_norm": 0.15729284286499023, + "learning_rate": 0.00018084087702240694, + "loss": 0.7747, + "step": 1349 + }, + { + "epoch": 0.2016731401254855, + "grad_norm": 0.23096966743469238, + "learning_rate": 0.00018081320189173577, + "loss": 0.8718, + "step": 1350 + }, + { + "epoch": 0.20182252763668956, + "grad_norm": 0.17240802943706512, + "learning_rate": 0.00018078550890814827, + "loss": 0.7155, + "step": 1351 + }, + { + "epoch": 0.20197191514789364, + "grad_norm": 0.1575528234243393, + "learning_rate": 0.0001807577980777623, + "loss": 0.7335, + "step": 1352 + }, + { + "epoch": 0.2021213026590977, + "grad_norm": 0.16896086931228638, + "learning_rate": 0.00018073006940669956, + "loss": 1.0208, + "step": 1353 + }, + { + "epoch": 0.20227069017030178, + "grad_norm": 0.2874300479888916, + "learning_rate": 0.00018070232290108584, + "loss": 1.075, + "step": 1354 + }, + { + "epoch": 0.20242007768150583, + "grad_norm": 0.12777869403362274, + "learning_rate": 0.0001806745585670507, + "loss": 0.4453, + "step": 1355 + }, + { + "epoch": 0.20256946519270988, + "grad_norm": 0.28166279196739197, + "learning_rate": 0.00018064677641072775, + "loss": 0.9604, + "step": 1356 + }, + { + "epoch": 0.20271885270391396, + "grad_norm": 0.1567370742559433, + "learning_rate": 0.0001806189764382545, + "loss": 0.8008, + "step": 1357 + }, + { + "epoch": 0.202868240215118, + "grad_norm": 0.12733782827854156, + "learning_rate": 0.00018059115865577249, + "loss": 0.8105, + "step": 1358 + }, + { + "epoch": 0.2030176277263221, + "grad_norm": 0.1359173059463501, + "learning_rate": 0.00018056332306942696, + "loss": 0.8775, + "step": 1359 + }, + { + "epoch": 0.20316701523752614, + "grad_norm": 0.10855043679475784, + "learning_rate": 0.00018053546968536735, + "loss": 0.6566, + "step": 1360 + }, + { + "epoch": 0.2033164027487302, + "grad_norm": 0.1817154735326767, + "learning_rate": 0.00018050759850974683, + "loss": 0.4578, + "step": 1361 + }, + { + "epoch": 0.20346579025993428, + "grad_norm": 0.13071177899837494, + "learning_rate": 0.00018047970954872264, + "loss": 0.6817, + "step": 1362 + }, + { + "epoch": 0.20361517777113833, + "grad_norm": 0.26592084765434265, + "learning_rate": 0.0001804518028084559, + "loss": 0.9979, + "step": 1363 + }, + { + "epoch": 0.2037645652823424, + "grad_norm": 0.17801976203918457, + "learning_rate": 0.0001804238782951116, + "loss": 0.5385, + "step": 1364 + }, + { + "epoch": 0.20391395279354646, + "grad_norm": 0.13625523447990417, + "learning_rate": 0.00018039593601485874, + "loss": 0.7867, + "step": 1365 + }, + { + "epoch": 0.2040633403047505, + "grad_norm": 0.12862275540828705, + "learning_rate": 0.00018036797597387023, + "loss": 0.8467, + "step": 1366 + }, + { + "epoch": 0.2042127278159546, + "grad_norm": 0.709415078163147, + "learning_rate": 0.00018033999817832286, + "loss": 1.4571, + "step": 1367 + }, + { + "epoch": 0.20436211532715864, + "grad_norm": 0.13892705738544464, + "learning_rate": 0.00018031200263439736, + "loss": 0.7435, + "step": 1368 + }, + { + "epoch": 0.20451150283836272, + "grad_norm": 0.14498738944530487, + "learning_rate": 0.0001802839893482784, + "loss": 1.0691, + "step": 1369 + }, + { + "epoch": 0.20466089034956678, + "grad_norm": 0.34637802839279175, + "learning_rate": 0.00018025595832615459, + "loss": 0.8286, + "step": 1370 + }, + { + "epoch": 0.20481027786077083, + "grad_norm": 0.12984217703342438, + "learning_rate": 0.00018022790957421836, + "loss": 0.6684, + "step": 1371 + }, + { + "epoch": 0.2049596653719749, + "grad_norm": 0.10101130604743958, + "learning_rate": 0.00018019984309866619, + "loss": 0.5913, + "step": 1372 + }, + { + "epoch": 0.20510905288317896, + "grad_norm": 0.20061518251895905, + "learning_rate": 0.00018017175890569834, + "loss": 1.0633, + "step": 1373 + }, + { + "epoch": 0.20525844039438304, + "grad_norm": 0.17312417924404144, + "learning_rate": 0.00018014365700151912, + "loss": 0.8474, + "step": 1374 + }, + { + "epoch": 0.2054078279055871, + "grad_norm": 0.14086449146270752, + "learning_rate": 0.00018011553739233666, + "loss": 0.4827, + "step": 1375 + }, + { + "epoch": 0.20555721541679114, + "grad_norm": 0.21013930439949036, + "learning_rate": 0.000180087400084363, + "loss": 0.8096, + "step": 1376 + }, + { + "epoch": 0.20570660292799522, + "grad_norm": 0.2212519496679306, + "learning_rate": 0.0001800592450838142, + "loss": 1.0053, + "step": 1377 + }, + { + "epoch": 0.20585599043919928, + "grad_norm": 0.15844549238681793, + "learning_rate": 0.00018003107239691004, + "loss": 0.7247, + "step": 1378 + }, + { + "epoch": 0.20600537795040336, + "grad_norm": 0.12048251181840897, + "learning_rate": 0.00018000288202987437, + "loss": 0.6577, + "step": 1379 + }, + { + "epoch": 0.2061547654616074, + "grad_norm": 0.15412500500679016, + "learning_rate": 0.00017997467398893488, + "loss": 0.8612, + "step": 1380 + }, + { + "epoch": 0.20630415297281146, + "grad_norm": 0.2080097794532776, + "learning_rate": 0.00017994644828032316, + "loss": 1.1714, + "step": 1381 + }, + { + "epoch": 0.20645354048401554, + "grad_norm": 0.10042672604322433, + "learning_rate": 0.00017991820491027472, + "loss": 0.4871, + "step": 1382 + }, + { + "epoch": 0.2066029279952196, + "grad_norm": 0.12917503714561462, + "learning_rate": 0.00017988994388502903, + "loss": 0.9816, + "step": 1383 + }, + { + "epoch": 0.20675231550642367, + "grad_norm": 0.1451895833015442, + "learning_rate": 0.0001798616652108293, + "loss": 0.7564, + "step": 1384 + }, + { + "epoch": 0.20690170301762772, + "grad_norm": 0.12688519060611725, + "learning_rate": 0.0001798333688939228, + "loss": 0.9622, + "step": 1385 + }, + { + "epoch": 0.20705109052883178, + "grad_norm": 0.11722414195537567, + "learning_rate": 0.00017980505494056062, + "loss": 0.7, + "step": 1386 + }, + { + "epoch": 0.20720047804003586, + "grad_norm": 0.12333224713802338, + "learning_rate": 0.00017977672335699776, + "loss": 0.6952, + "step": 1387 + }, + { + "epoch": 0.2073498655512399, + "grad_norm": 0.11864780634641647, + "learning_rate": 0.00017974837414949307, + "loss": 0.6831, + "step": 1388 + }, + { + "epoch": 0.207499253062444, + "grad_norm": 0.16104672849178314, + "learning_rate": 0.00017972000732430942, + "loss": 0.8635, + "step": 1389 + }, + { + "epoch": 0.20764864057364804, + "grad_norm": 0.13945555686950684, + "learning_rate": 0.00017969162288771347, + "loss": 0.9024, + "step": 1390 + }, + { + "epoch": 0.2077980280848521, + "grad_norm": 0.11853618174791336, + "learning_rate": 0.00017966322084597572, + "loss": 0.7451, + "step": 1391 + }, + { + "epoch": 0.20794741559605617, + "grad_norm": 0.12260474264621735, + "learning_rate": 0.0001796348012053707, + "loss": 0.5602, + "step": 1392 + }, + { + "epoch": 0.20809680310726023, + "grad_norm": 0.2517350912094116, + "learning_rate": 0.0001796063639721768, + "loss": 0.6186, + "step": 1393 + }, + { + "epoch": 0.2082461906184643, + "grad_norm": 0.2064390927553177, + "learning_rate": 0.00017957790915267615, + "loss": 0.8175, + "step": 1394 + }, + { + "epoch": 0.20839557812966836, + "grad_norm": 0.24569760262966156, + "learning_rate": 0.0001795494367531549, + "loss": 0.8564, + "step": 1395 + }, + { + "epoch": 0.2085449656408724, + "grad_norm": 0.2579876780509949, + "learning_rate": 0.0001795209467799031, + "loss": 1.3103, + "step": 1396 + }, + { + "epoch": 0.2086943531520765, + "grad_norm": 0.24793849885463715, + "learning_rate": 0.00017949243923921462, + "loss": 0.8798, + "step": 1397 + }, + { + "epoch": 0.20884374066328054, + "grad_norm": 0.32119929790496826, + "learning_rate": 0.0001794639141373872, + "loss": 0.9218, + "step": 1398 + }, + { + "epoch": 0.20899312817448462, + "grad_norm": 0.17471039295196533, + "learning_rate": 0.00017943537148072252, + "loss": 0.5358, + "step": 1399 + }, + { + "epoch": 0.20914251568568867, + "grad_norm": 0.1545010507106781, + "learning_rate": 0.00017940681127552604, + "loss": 0.7135, + "step": 1400 + }, + { + "epoch": 0.20929190319689275, + "grad_norm": 0.15277154743671417, + "learning_rate": 0.00017937823352810725, + "loss": 0.8968, + "step": 1401 + }, + { + "epoch": 0.2094412907080968, + "grad_norm": 0.14847959578037262, + "learning_rate": 0.0001793496382447794, + "loss": 0.6325, + "step": 1402 + }, + { + "epoch": 0.20959067821930086, + "grad_norm": 0.14463500678539276, + "learning_rate": 0.00017932102543185963, + "loss": 0.4591, + "step": 1403 + }, + { + "epoch": 0.20974006573050494, + "grad_norm": 0.12656496465206146, + "learning_rate": 0.00017929239509566894, + "loss": 0.8605, + "step": 1404 + }, + { + "epoch": 0.209889453241709, + "grad_norm": 0.3494865298271179, + "learning_rate": 0.00017926374724253226, + "loss": 1.0462, + "step": 1405 + }, + { + "epoch": 0.21003884075291307, + "grad_norm": 0.2256232500076294, + "learning_rate": 0.00017923508187877834, + "loss": 0.6951, + "step": 1406 + }, + { + "epoch": 0.21018822826411712, + "grad_norm": 0.14871495962142944, + "learning_rate": 0.00017920639901073976, + "loss": 0.6265, + "step": 1407 + }, + { + "epoch": 0.21033761577532117, + "grad_norm": 0.12966501712799072, + "learning_rate": 0.00017917769864475314, + "loss": 0.8843, + "step": 1408 + }, + { + "epoch": 0.21048700328652525, + "grad_norm": 0.14218322932720184, + "learning_rate": 0.00017914898078715875, + "loss": 0.5568, + "step": 1409 + }, + { + "epoch": 0.2106363907977293, + "grad_norm": 0.17389464378356934, + "learning_rate": 0.00017912024544430088, + "loss": 1.0362, + "step": 1410 + }, + { + "epoch": 0.21078577830893339, + "grad_norm": 0.18730613589286804, + "learning_rate": 0.00017909149262252755, + "loss": 0.5673, + "step": 1411 + }, + { + "epoch": 0.21093516582013744, + "grad_norm": 0.14220395684242249, + "learning_rate": 0.0001790627223281908, + "loss": 0.9562, + "step": 1412 + }, + { + "epoch": 0.2110845533313415, + "grad_norm": 0.15215925872325897, + "learning_rate": 0.00017903393456764635, + "loss": 0.9981, + "step": 1413 + }, + { + "epoch": 0.21123394084254557, + "grad_norm": 0.22159068286418915, + "learning_rate": 0.00017900512934725397, + "loss": 0.9884, + "step": 1414 + }, + { + "epoch": 0.21138332835374962, + "grad_norm": 0.13190264999866486, + "learning_rate": 0.0001789763066733771, + "loss": 0.7929, + "step": 1415 + }, + { + "epoch": 0.2115327158649537, + "grad_norm": 0.13530825078487396, + "learning_rate": 0.0001789474665523832, + "loss": 0.8792, + "step": 1416 + }, + { + "epoch": 0.21168210337615775, + "grad_norm": 0.14375026524066925, + "learning_rate": 0.00017891860899064342, + "loss": 0.8539, + "step": 1417 + }, + { + "epoch": 0.2118314908873618, + "grad_norm": 0.13115522265434265, + "learning_rate": 0.00017888973399453296, + "loss": 0.8168, + "step": 1418 + }, + { + "epoch": 0.2119808783985659, + "grad_norm": 0.6125191450119019, + "learning_rate": 0.0001788608415704307, + "loss": 1.2228, + "step": 1419 + }, + { + "epoch": 0.21213026590976994, + "grad_norm": 0.11920984089374542, + "learning_rate": 0.00017883193172471944, + "loss": 0.8112, + "step": 1420 + }, + { + "epoch": 0.21227965342097402, + "grad_norm": 0.16258184611797333, + "learning_rate": 0.00017880300446378584, + "loss": 0.8269, + "step": 1421 + }, + { + "epoch": 0.21242904093217807, + "grad_norm": 0.18811704218387604, + "learning_rate": 0.00017877405979402038, + "loss": 0.7807, + "step": 1422 + }, + { + "epoch": 0.21257842844338212, + "grad_norm": 0.18149599432945251, + "learning_rate": 0.00017874509772181738, + "loss": 0.7866, + "step": 1423 + }, + { + "epoch": 0.2127278159545862, + "grad_norm": 0.2281455546617508, + "learning_rate": 0.00017871611825357502, + "loss": 0.9836, + "step": 1424 + }, + { + "epoch": 0.21287720346579025, + "grad_norm": 0.1633453071117401, + "learning_rate": 0.00017868712139569535, + "loss": 0.9219, + "step": 1425 + }, + { + "epoch": 0.21302659097699433, + "grad_norm": 0.6165900230407715, + "learning_rate": 0.00017865810715458427, + "loss": 1.3481, + "step": 1426 + }, + { + "epoch": 0.2131759784881984, + "grad_norm": 0.17836418747901917, + "learning_rate": 0.0001786290755366514, + "loss": 0.9347, + "step": 1427 + }, + { + "epoch": 0.21332536599940244, + "grad_norm": 0.19245529174804688, + "learning_rate": 0.00017860002654831032, + "loss": 0.4793, + "step": 1428 + }, + { + "epoch": 0.21347475351060652, + "grad_norm": 0.13874202966690063, + "learning_rate": 0.00017857096019597844, + "loss": 0.8586, + "step": 1429 + }, + { + "epoch": 0.21362414102181057, + "grad_norm": 0.26488029956817627, + "learning_rate": 0.00017854187648607694, + "loss": 1.1623, + "step": 1430 + }, + { + "epoch": 0.21377352853301465, + "grad_norm": 0.10857315361499786, + "learning_rate": 0.00017851277542503085, + "loss": 0.6721, + "step": 1431 + }, + { + "epoch": 0.2139229160442187, + "grad_norm": 0.1778874397277832, + "learning_rate": 0.00017848365701926913, + "loss": 0.8528, + "step": 1432 + }, + { + "epoch": 0.21407230355542275, + "grad_norm": 0.1809145212173462, + "learning_rate": 0.00017845452127522444, + "loss": 1.001, + "step": 1433 + }, + { + "epoch": 0.21422169106662683, + "grad_norm": 0.17359715700149536, + "learning_rate": 0.00017842536819933337, + "loss": 1.0188, + "step": 1434 + }, + { + "epoch": 0.2143710785778309, + "grad_norm": 0.13792802393436432, + "learning_rate": 0.00017839619779803627, + "loss": 0.9398, + "step": 1435 + }, + { + "epoch": 0.21452046608903497, + "grad_norm": 0.1106923520565033, + "learning_rate": 0.0001783670100777773, + "loss": 0.6102, + "step": 1436 + }, + { + "epoch": 0.21466985360023902, + "grad_norm": 0.1782097965478897, + "learning_rate": 0.0001783378050450046, + "loss": 0.4704, + "step": 1437 + }, + { + "epoch": 0.21481924111144307, + "grad_norm": 0.1738913655281067, + "learning_rate": 0.0001783085827061699, + "loss": 0.7214, + "step": 1438 + }, + { + "epoch": 0.21496862862264715, + "grad_norm": 0.15943187475204468, + "learning_rate": 0.00017827934306772897, + "loss": 0.962, + "step": 1439 + }, + { + "epoch": 0.2151180161338512, + "grad_norm": 0.14920490980148315, + "learning_rate": 0.00017825008613614127, + "loss": 0.6464, + "step": 1440 + }, + { + "epoch": 0.21526740364505528, + "grad_norm": 0.23284317553043365, + "learning_rate": 0.0001782208119178701, + "loss": 0.9425, + "step": 1441 + }, + { + "epoch": 0.21541679115625934, + "grad_norm": 0.12878303229808807, + "learning_rate": 0.00017819152041938265, + "loss": 0.7147, + "step": 1442 + }, + { + "epoch": 0.2155661786674634, + "grad_norm": 0.13469137251377106, + "learning_rate": 0.00017816221164714983, + "loss": 0.8396, + "step": 1443 + }, + { + "epoch": 0.21571556617866747, + "grad_norm": 0.13271968066692352, + "learning_rate": 0.00017813288560764647, + "loss": 0.8708, + "step": 1444 + }, + { + "epoch": 0.21586495368987152, + "grad_norm": 0.13078995048999786, + "learning_rate": 0.00017810354230735108, + "loss": 0.5511, + "step": 1445 + }, + { + "epoch": 0.2160143412010756, + "grad_norm": 0.179561585187912, + "learning_rate": 0.00017807418175274612, + "loss": 0.5318, + "step": 1446 + }, + { + "epoch": 0.21616372871227965, + "grad_norm": 0.16479915380477905, + "learning_rate": 0.00017804480395031778, + "loss": 0.9253, + "step": 1447 + }, + { + "epoch": 0.2163131162234837, + "grad_norm": 0.15908388793468475, + "learning_rate": 0.00017801540890655609, + "loss": 0.7711, + "step": 1448 + }, + { + "epoch": 0.21646250373468778, + "grad_norm": 0.16113927960395813, + "learning_rate": 0.00017798599662795483, + "loss": 0.643, + "step": 1449 + }, + { + "epoch": 0.21661189124589184, + "grad_norm": 0.12595628201961517, + "learning_rate": 0.00017795656712101172, + "loss": 0.6977, + "step": 1450 + }, + { + "epoch": 0.21676127875709592, + "grad_norm": 0.15012586116790771, + "learning_rate": 0.00017792712039222815, + "loss": 0.8391, + "step": 1451 + }, + { + "epoch": 0.21691066626829997, + "grad_norm": 0.13315634429454803, + "learning_rate": 0.00017789765644810935, + "loss": 0.6284, + "step": 1452 + }, + { + "epoch": 0.21706005377950402, + "grad_norm": 0.20887716114521027, + "learning_rate": 0.00017786817529516445, + "loss": 0.6472, + "step": 1453 + }, + { + "epoch": 0.2172094412907081, + "grad_norm": 0.12334492802619934, + "learning_rate": 0.00017783867693990624, + "loss": 0.8016, + "step": 1454 + }, + { + "epoch": 0.21735882880191215, + "grad_norm": 0.14215132594108582, + "learning_rate": 0.0001778091613888514, + "loss": 0.6711, + "step": 1455 + }, + { + "epoch": 0.21750821631311623, + "grad_norm": 0.6180028319358826, + "learning_rate": 0.0001777796286485204, + "loss": 1.2348, + "step": 1456 + }, + { + "epoch": 0.21765760382432028, + "grad_norm": 0.15732866525650024, + "learning_rate": 0.00017775007872543745, + "loss": 0.7951, + "step": 1457 + }, + { + "epoch": 0.21780699133552436, + "grad_norm": 0.07052062451839447, + "learning_rate": 0.0001777205116261306, + "loss": 0.3056, + "step": 1458 + }, + { + "epoch": 0.21795637884672842, + "grad_norm": 0.1404731720685959, + "learning_rate": 0.00017769092735713172, + "loss": 0.8616, + "step": 1459 + }, + { + "epoch": 0.21810576635793247, + "grad_norm": 0.24289721250534058, + "learning_rate": 0.0001776613259249764, + "loss": 1.058, + "step": 1460 + }, + { + "epoch": 0.21825515386913655, + "grad_norm": 0.3455839157104492, + "learning_rate": 0.0001776317073362041, + "loss": 0.9287, + "step": 1461 + }, + { + "epoch": 0.2184045413803406, + "grad_norm": 0.181719109416008, + "learning_rate": 0.00017760207159735805, + "loss": 0.851, + "step": 1462 + }, + { + "epoch": 0.21855392889154468, + "grad_norm": 0.29283466935157776, + "learning_rate": 0.0001775724187149852, + "loss": 0.9268, + "step": 1463 + }, + { + "epoch": 0.21870331640274873, + "grad_norm": 0.12602414190769196, + "learning_rate": 0.00017754274869563637, + "loss": 0.8347, + "step": 1464 + }, + { + "epoch": 0.21885270391395278, + "grad_norm": 0.13600635528564453, + "learning_rate": 0.0001775130615458662, + "loss": 0.9986, + "step": 1465 + }, + { + "epoch": 0.21900209142515686, + "grad_norm": 0.15044277906417847, + "learning_rate": 0.00017748335727223294, + "loss": 0.3543, + "step": 1466 + }, + { + "epoch": 0.21915147893636092, + "grad_norm": 0.1423201858997345, + "learning_rate": 0.00017745363588129878, + "loss": 0.9459, + "step": 1467 + }, + { + "epoch": 0.219300866447565, + "grad_norm": 0.13680531084537506, + "learning_rate": 0.00017742389737962966, + "loss": 0.5642, + "step": 1468 + }, + { + "epoch": 0.21945025395876905, + "grad_norm": 0.13396508991718292, + "learning_rate": 0.00017739414177379528, + "loss": 0.6551, + "step": 1469 + }, + { + "epoch": 0.2195996414699731, + "grad_norm": 0.11900012195110321, + "learning_rate": 0.0001773643690703691, + "loss": 0.6664, + "step": 1470 + }, + { + "epoch": 0.21974902898117718, + "grad_norm": 0.12804585695266724, + "learning_rate": 0.00017733457927592846, + "loss": 0.6117, + "step": 1471 + }, + { + "epoch": 0.21989841649238123, + "grad_norm": 0.19028812646865845, + "learning_rate": 0.00017730477239705428, + "loss": 0.5632, + "step": 1472 + }, + { + "epoch": 0.2200478040035853, + "grad_norm": 0.15631790459156036, + "learning_rate": 0.00017727494844033145, + "loss": 1.0133, + "step": 1473 + }, + { + "epoch": 0.22019719151478936, + "grad_norm": 0.14937928318977356, + "learning_rate": 0.00017724510741234858, + "loss": 0.9413, + "step": 1474 + }, + { + "epoch": 0.22034657902599342, + "grad_norm": 0.12716789543628693, + "learning_rate": 0.00017721524931969796, + "loss": 0.5547, + "step": 1475 + }, + { + "epoch": 0.2204959665371975, + "grad_norm": 0.27234596014022827, + "learning_rate": 0.0001771853741689757, + "loss": 1.0554, + "step": 1476 + }, + { + "epoch": 0.22064535404840155, + "grad_norm": 0.1304127275943756, + "learning_rate": 0.00017715548196678177, + "loss": 0.7407, + "step": 1477 + }, + { + "epoch": 0.22079474155960563, + "grad_norm": 0.14339019358158112, + "learning_rate": 0.0001771255727197198, + "loss": 0.9733, + "step": 1478 + }, + { + "epoch": 0.22094412907080968, + "grad_norm": 0.18441441655158997, + "learning_rate": 0.0001770956464343972, + "loss": 0.9436, + "step": 1479 + }, + { + "epoch": 0.22109351658201373, + "grad_norm": 0.13946813344955444, + "learning_rate": 0.00017706570311742516, + "loss": 0.6699, + "step": 1480 + }, + { + "epoch": 0.2212429040932178, + "grad_norm": 0.20996683835983276, + "learning_rate": 0.00017703574277541865, + "loss": 1.0231, + "step": 1481 + }, + { + "epoch": 0.22139229160442186, + "grad_norm": 0.15931503474712372, + "learning_rate": 0.0001770057654149964, + "loss": 0.7794, + "step": 1482 + }, + { + "epoch": 0.22154167911562594, + "grad_norm": 0.14282017946243286, + "learning_rate": 0.00017697577104278084, + "loss": 0.6515, + "step": 1483 + }, + { + "epoch": 0.22169106662683, + "grad_norm": 0.2071201503276825, + "learning_rate": 0.00017694575966539823, + "loss": 0.8382, + "step": 1484 + }, + { + "epoch": 0.22184045413803405, + "grad_norm": 0.16338187456130981, + "learning_rate": 0.00017691573128947853, + "loss": 0.9846, + "step": 1485 + }, + { + "epoch": 0.22198984164923813, + "grad_norm": 0.13437145948410034, + "learning_rate": 0.00017688568592165552, + "loss": 0.8371, + "step": 1486 + }, + { + "epoch": 0.22213922916044218, + "grad_norm": 0.10946273803710938, + "learning_rate": 0.0001768556235685667, + "loss": 0.6741, + "step": 1487 + }, + { + "epoch": 0.22228861667164626, + "grad_norm": 0.19846905767917633, + "learning_rate": 0.00017682554423685329, + "loss": 0.6215, + "step": 1488 + }, + { + "epoch": 0.2224380041828503, + "grad_norm": 0.12315358966588974, + "learning_rate": 0.0001767954479331603, + "loss": 0.6331, + "step": 1489 + }, + { + "epoch": 0.22258739169405437, + "grad_norm": 0.30643758177757263, + "learning_rate": 0.0001767653346641365, + "loss": 0.8563, + "step": 1490 + }, + { + "epoch": 0.22273677920525844, + "grad_norm": 0.1250000298023224, + "learning_rate": 0.00017673520443643436, + "loss": 0.8405, + "step": 1491 + }, + { + "epoch": 0.2228861667164625, + "grad_norm": 0.12933503091335297, + "learning_rate": 0.00017670505725671013, + "loss": 0.7115, + "step": 1492 + }, + { + "epoch": 0.22303555422766658, + "grad_norm": 0.2853139638900757, + "learning_rate": 0.00017667489313162382, + "loss": 0.8367, + "step": 1493 + }, + { + "epoch": 0.22318494173887063, + "grad_norm": 0.13272625207901, + "learning_rate": 0.00017664471206783915, + "loss": 0.7379, + "step": 1494 + }, + { + "epoch": 0.22333432925007468, + "grad_norm": 0.1108209565281868, + "learning_rate": 0.0001766145140720236, + "loss": 0.5442, + "step": 1495 + }, + { + "epoch": 0.22348371676127876, + "grad_norm": 0.1489863097667694, + "learning_rate": 0.00017658429915084835, + "loss": 0.731, + "step": 1496 + }, + { + "epoch": 0.2236331042724828, + "grad_norm": 0.146969735622406, + "learning_rate": 0.0001765540673109884, + "loss": 0.605, + "step": 1497 + }, + { + "epoch": 0.2237824917836869, + "grad_norm": 0.2565854489803314, + "learning_rate": 0.00017652381855912247, + "loss": 1.1015, + "step": 1498 + }, + { + "epoch": 0.22393187929489095, + "grad_norm": 0.1937399059534073, + "learning_rate": 0.00017649355290193288, + "loss": 1.0397, + "step": 1499 + }, + { + "epoch": 0.224081266806095, + "grad_norm": 0.33174678683280945, + "learning_rate": 0.0001764632703461059, + "loss": 1.0074, + "step": 1500 + }, + { + "epoch": 0.22423065431729908, + "grad_norm": 0.12607234716415405, + "learning_rate": 0.00017643297089833134, + "loss": 0.7882, + "step": 1501 + }, + { + "epoch": 0.22438004182850313, + "grad_norm": 0.14097252488136292, + "learning_rate": 0.00017640265456530293, + "loss": 0.8961, + "step": 1502 + }, + { + "epoch": 0.2245294293397072, + "grad_norm": 0.10994578897953033, + "learning_rate": 0.00017637232135371794, + "loss": 0.4649, + "step": 1503 + }, + { + "epoch": 0.22467881685091126, + "grad_norm": 0.2514827847480774, + "learning_rate": 0.0001763419712702775, + "loss": 0.6585, + "step": 1504 + }, + { + "epoch": 0.2248282043621153, + "grad_norm": 0.13599185645580292, + "learning_rate": 0.0001763116043216864, + "loss": 0.692, + "step": 1505 + }, + { + "epoch": 0.2249775918733194, + "grad_norm": 0.12496299296617508, + "learning_rate": 0.00017628122051465322, + "loss": 0.802, + "step": 1506 + }, + { + "epoch": 0.22512697938452345, + "grad_norm": 0.178312286734581, + "learning_rate": 0.00017625081985589016, + "loss": 0.7811, + "step": 1507 + }, + { + "epoch": 0.22527636689572753, + "grad_norm": 0.12510164082050323, + "learning_rate": 0.00017622040235211326, + "loss": 0.4649, + "step": 1508 + }, + { + "epoch": 0.22542575440693158, + "grad_norm": 0.14987626671791077, + "learning_rate": 0.0001761899680100422, + "loss": 0.4765, + "step": 1509 + }, + { + "epoch": 0.22557514191813566, + "grad_norm": 0.554569661617279, + "learning_rate": 0.00017615951683640045, + "loss": 1.674, + "step": 1510 + }, + { + "epoch": 0.2257245294293397, + "grad_norm": 0.16470187902450562, + "learning_rate": 0.0001761290488379151, + "loss": 0.5423, + "step": 1511 + }, + { + "epoch": 0.22587391694054376, + "grad_norm": 0.13838253915309906, + "learning_rate": 0.00017609856402131703, + "loss": 0.5373, + "step": 1512 + }, + { + "epoch": 0.22602330445174784, + "grad_norm": 0.13024069368839264, + "learning_rate": 0.00017606806239334083, + "loss": 0.8255, + "step": 1513 + }, + { + "epoch": 0.2261726919629519, + "grad_norm": 0.13221733272075653, + "learning_rate": 0.00017603754396072483, + "loss": 0.7189, + "step": 1514 + }, + { + "epoch": 0.22632207947415597, + "grad_norm": 0.15409418940544128, + "learning_rate": 0.00017600700873021097, + "loss": 1.0796, + "step": 1515 + }, + { + "epoch": 0.22647146698536003, + "grad_norm": 0.16743522882461548, + "learning_rate": 0.000175976456708545, + "loss": 1.0363, + "step": 1516 + }, + { + "epoch": 0.22662085449656408, + "grad_norm": 0.19694027304649353, + "learning_rate": 0.0001759458879024763, + "loss": 0.8512, + "step": 1517 + }, + { + "epoch": 0.22677024200776816, + "grad_norm": 0.43795064091682434, + "learning_rate": 0.0001759153023187581, + "loss": 1.4031, + "step": 1518 + }, + { + "epoch": 0.2269196295189722, + "grad_norm": 0.22090782225131989, + "learning_rate": 0.00017588469996414715, + "loss": 0.6461, + "step": 1519 + }, + { + "epoch": 0.2270690170301763, + "grad_norm": 0.13713456690311432, + "learning_rate": 0.00017585408084540405, + "loss": 0.8236, + "step": 1520 + }, + { + "epoch": 0.22721840454138034, + "grad_norm": 0.1438591182231903, + "learning_rate": 0.000175823444969293, + "loss": 1.1679, + "step": 1521 + }, + { + "epoch": 0.2273677920525844, + "grad_norm": 0.12054278701543808, + "learning_rate": 0.00017579279234258198, + "loss": 0.6954, + "step": 1522 + }, + { + "epoch": 0.22751717956378847, + "grad_norm": 0.12184090167284012, + "learning_rate": 0.0001757621229720426, + "loss": 0.7207, + "step": 1523 + }, + { + "epoch": 0.22766656707499253, + "grad_norm": 0.36808204650878906, + "learning_rate": 0.00017573143686445034, + "loss": 0.8402, + "step": 1524 + }, + { + "epoch": 0.2278159545861966, + "grad_norm": 0.20362381637096405, + "learning_rate": 0.00017570073402658408, + "loss": 0.7507, + "step": 1525 + }, + { + "epoch": 0.22796534209740066, + "grad_norm": 0.32647380232810974, + "learning_rate": 0.00017567001446522665, + "loss": 0.6814, + "step": 1526 + }, + { + "epoch": 0.2281147296086047, + "grad_norm": 0.18920038640499115, + "learning_rate": 0.00017563927818716447, + "loss": 0.503, + "step": 1527 + }, + { + "epoch": 0.2282641171198088, + "grad_norm": 0.18404003977775574, + "learning_rate": 0.0001756085251991877, + "loss": 0.5798, + "step": 1528 + }, + { + "epoch": 0.22841350463101284, + "grad_norm": 0.1709233671426773, + "learning_rate": 0.0001755777555080901, + "loss": 1.3392, + "step": 1529 + }, + { + "epoch": 0.22856289214221692, + "grad_norm": 0.13622832298278809, + "learning_rate": 0.00017554696912066924, + "loss": 0.7546, + "step": 1530 + }, + { + "epoch": 0.22871227965342097, + "grad_norm": 0.14805671572685242, + "learning_rate": 0.00017551616604372629, + "loss": 0.8027, + "step": 1531 + }, + { + "epoch": 0.22886166716462503, + "grad_norm": 0.2947498559951782, + "learning_rate": 0.00017548534628406616, + "loss": 0.9071, + "step": 1532 + }, + { + "epoch": 0.2290110546758291, + "grad_norm": 0.1275065690279007, + "learning_rate": 0.0001754545098484974, + "loss": 0.6312, + "step": 1533 + }, + { + "epoch": 0.22916044218703316, + "grad_norm": 0.11439994722604752, + "learning_rate": 0.00017542365674383227, + "loss": 0.4962, + "step": 1534 + }, + { + "epoch": 0.22930982969823724, + "grad_norm": 0.14503420889377594, + "learning_rate": 0.00017539278697688672, + "loss": 0.6967, + "step": 1535 + }, + { + "epoch": 0.2294592172094413, + "grad_norm": 0.13170544803142548, + "learning_rate": 0.00017536190055448037, + "loss": 0.9048, + "step": 1536 + }, + { + "epoch": 0.22960860472064534, + "grad_norm": 0.22682088613510132, + "learning_rate": 0.0001753309974834365, + "loss": 0.7506, + "step": 1537 + }, + { + "epoch": 0.22975799223184942, + "grad_norm": 0.16265404224395752, + "learning_rate": 0.00017530007777058213, + "loss": 0.6615, + "step": 1538 + }, + { + "epoch": 0.22990737974305347, + "grad_norm": 0.1368705928325653, + "learning_rate": 0.0001752691414227479, + "loss": 0.8368, + "step": 1539 + }, + { + "epoch": 0.23005676725425755, + "grad_norm": 0.18166819214820862, + "learning_rate": 0.0001752381884467681, + "loss": 0.6836, + "step": 1540 + }, + { + "epoch": 0.2302061547654616, + "grad_norm": 0.11365760862827301, + "learning_rate": 0.00017520721884948075, + "loss": 0.7073, + "step": 1541 + }, + { + "epoch": 0.23035554227666566, + "grad_norm": 0.23853546380996704, + "learning_rate": 0.00017517623263772758, + "loss": 0.9235, + "step": 1542 + }, + { + "epoch": 0.23050492978786974, + "grad_norm": 0.29412710666656494, + "learning_rate": 0.00017514522981835383, + "loss": 1.1033, + "step": 1543 + }, + { + "epoch": 0.2306543172990738, + "grad_norm": 0.13339991867542267, + "learning_rate": 0.00017511421039820863, + "loss": 0.7038, + "step": 1544 + }, + { + "epoch": 0.23080370481027787, + "grad_norm": 0.12489107251167297, + "learning_rate": 0.00017508317438414458, + "loss": 0.5625, + "step": 1545 + }, + { + "epoch": 0.23095309232148192, + "grad_norm": 0.1896982640028, + "learning_rate": 0.00017505212178301805, + "loss": 0.7489, + "step": 1546 + }, + { + "epoch": 0.23110247983268598, + "grad_norm": 0.13012000918388367, + "learning_rate": 0.00017502105260168907, + "loss": 0.615, + "step": 1547 + }, + { + "epoch": 0.23125186734389006, + "grad_norm": 0.2946460247039795, + "learning_rate": 0.00017498996684702132, + "loss": 0.8367, + "step": 1548 + }, + { + "epoch": 0.2314012548550941, + "grad_norm": 0.15457066893577576, + "learning_rate": 0.00017495886452588205, + "loss": 0.5542, + "step": 1549 + }, + { + "epoch": 0.2315506423662982, + "grad_norm": 0.12016140669584274, + "learning_rate": 0.00017492774564514235, + "loss": 0.908, + "step": 1550 + }, + { + "epoch": 0.23170002987750224, + "grad_norm": 0.17320853471755981, + "learning_rate": 0.00017489661021167686, + "loss": 0.816, + "step": 1551 + }, + { + "epoch": 0.2318494173887063, + "grad_norm": 0.12414354085922241, + "learning_rate": 0.00017486545823236385, + "loss": 0.7893, + "step": 1552 + }, + { + "epoch": 0.23199880489991037, + "grad_norm": 0.1295233517885208, + "learning_rate": 0.00017483428971408534, + "loss": 0.9902, + "step": 1553 + }, + { + "epoch": 0.23214819241111442, + "grad_norm": 0.1532648801803589, + "learning_rate": 0.00017480310466372686, + "loss": 0.6839, + "step": 1554 + }, + { + "epoch": 0.2322975799223185, + "grad_norm": 0.15763483941555023, + "learning_rate": 0.00017477190308817778, + "loss": 0.73, + "step": 1555 + }, + { + "epoch": 0.23244696743352256, + "grad_norm": 0.22746244072914124, + "learning_rate": 0.00017474068499433098, + "loss": 0.9003, + "step": 1556 + }, + { + "epoch": 0.2325963549447266, + "grad_norm": 0.11888179183006287, + "learning_rate": 0.00017470945038908304, + "loss": 0.4965, + "step": 1557 + }, + { + "epoch": 0.2327457424559307, + "grad_norm": 0.11085602641105652, + "learning_rate": 0.00017467819927933416, + "loss": 0.4841, + "step": 1558 + }, + { + "epoch": 0.23289512996713474, + "grad_norm": 0.145162433385849, + "learning_rate": 0.0001746469316719882, + "loss": 0.9839, + "step": 1559 + }, + { + "epoch": 0.23304451747833882, + "grad_norm": 0.12735643982887268, + "learning_rate": 0.00017461564757395272, + "loss": 0.601, + "step": 1560 + }, + { + "epoch": 0.23319390498954287, + "grad_norm": 0.1690744012594223, + "learning_rate": 0.00017458434699213883, + "loss": 0.909, + "step": 1561 + }, + { + "epoch": 0.23334329250074692, + "grad_norm": 0.13597418367862701, + "learning_rate": 0.00017455302993346134, + "loss": 0.8371, + "step": 1562 + }, + { + "epoch": 0.233492680011951, + "grad_norm": 0.1151067465543747, + "learning_rate": 0.00017452169640483865, + "loss": 0.7987, + "step": 1563 + }, + { + "epoch": 0.23364206752315506, + "grad_norm": 0.1729961335659027, + "learning_rate": 0.00017449034641319288, + "loss": 0.434, + "step": 1564 + }, + { + "epoch": 0.23379145503435914, + "grad_norm": 0.1317516565322876, + "learning_rate": 0.00017445897996544972, + "loss": 0.9376, + "step": 1565 + }, + { + "epoch": 0.2339408425455632, + "grad_norm": 0.3169648349285126, + "learning_rate": 0.00017442759706853855, + "loss": 0.7284, + "step": 1566 + }, + { + "epoch": 0.23409023005676727, + "grad_norm": 0.15828759968280792, + "learning_rate": 0.0001743961977293923, + "loss": 0.8469, + "step": 1567 + }, + { + "epoch": 0.23423961756797132, + "grad_norm": 0.15548141300678253, + "learning_rate": 0.00017436478195494756, + "loss": 0.6545, + "step": 1568 + }, + { + "epoch": 0.23438900507917537, + "grad_norm": 0.11150869727134705, + "learning_rate": 0.00017433334975214463, + "loss": 0.6005, + "step": 1569 + }, + { + "epoch": 0.23453839259037945, + "grad_norm": 0.1405535489320755, + "learning_rate": 0.00017430190112792737, + "loss": 0.825, + "step": 1570 + }, + { + "epoch": 0.2346877801015835, + "grad_norm": 0.1619478166103363, + "learning_rate": 0.00017427043608924325, + "loss": 0.5656, + "step": 1571 + }, + { + "epoch": 0.23483716761278758, + "grad_norm": 0.1910037249326706, + "learning_rate": 0.00017423895464304342, + "loss": 0.9477, + "step": 1572 + }, + { + "epoch": 0.23498655512399164, + "grad_norm": 0.13986697793006897, + "learning_rate": 0.00017420745679628264, + "loss": 0.5093, + "step": 1573 + }, + { + "epoch": 0.2351359426351957, + "grad_norm": 0.13165700435638428, + "learning_rate": 0.00017417594255591927, + "loss": 0.7117, + "step": 1574 + }, + { + "epoch": 0.23528533014639977, + "grad_norm": 0.14331944286823273, + "learning_rate": 0.00017414441192891529, + "loss": 0.8661, + "step": 1575 + }, + { + "epoch": 0.23543471765760382, + "grad_norm": 0.12297346442937851, + "learning_rate": 0.00017411286492223632, + "loss": 0.9035, + "step": 1576 + }, + { + "epoch": 0.2355841051688079, + "grad_norm": 0.11895941197872162, + "learning_rate": 0.00017408130154285162, + "loss": 0.6656, + "step": 1577 + }, + { + "epoch": 0.23573349268001195, + "grad_norm": 0.1288233995437622, + "learning_rate": 0.000174049721797734, + "loss": 0.8691, + "step": 1578 + }, + { + "epoch": 0.235882880191216, + "grad_norm": 0.10877169668674469, + "learning_rate": 0.00017401812569385998, + "loss": 0.6643, + "step": 1579 + }, + { + "epoch": 0.23603226770242008, + "grad_norm": 0.11538077890872955, + "learning_rate": 0.00017398651323820958, + "loss": 0.7094, + "step": 1580 + }, + { + "epoch": 0.23618165521362414, + "grad_norm": 0.12808218598365784, + "learning_rate": 0.00017395488443776652, + "loss": 0.5041, + "step": 1581 + }, + { + "epoch": 0.23633104272482822, + "grad_norm": 0.26270732283592224, + "learning_rate": 0.00017392323929951812, + "loss": 0.8308, + "step": 1582 + }, + { + "epoch": 0.23648043023603227, + "grad_norm": 0.14876054227352142, + "learning_rate": 0.0001738915778304553, + "loss": 0.5326, + "step": 1583 + }, + { + "epoch": 0.23662981774723632, + "grad_norm": 0.11852624267339706, + "learning_rate": 0.0001738599000375725, + "loss": 0.6705, + "step": 1584 + }, + { + "epoch": 0.2367792052584404, + "grad_norm": 0.21757078170776367, + "learning_rate": 0.00017382820592786791, + "loss": 0.8034, + "step": 1585 + }, + { + "epoch": 0.23692859276964445, + "grad_norm": 0.20093433558940887, + "learning_rate": 0.00017379649550834327, + "loss": 0.5919, + "step": 1586 + }, + { + "epoch": 0.23707798028084853, + "grad_norm": 0.1466897428035736, + "learning_rate": 0.0001737647687860039, + "loss": 0.8727, + "step": 1587 + }, + { + "epoch": 0.23722736779205258, + "grad_norm": 0.13090848922729492, + "learning_rate": 0.00017373302576785874, + "loss": 0.8407, + "step": 1588 + }, + { + "epoch": 0.23737675530325664, + "grad_norm": 0.265153706073761, + "learning_rate": 0.0001737012664609203, + "loss": 0.6411, + "step": 1589 + }, + { + "epoch": 0.23752614281446072, + "grad_norm": 0.1178729236125946, + "learning_rate": 0.00017366949087220472, + "loss": 0.7404, + "step": 1590 + }, + { + "epoch": 0.23767553032566477, + "grad_norm": 0.1343647837638855, + "learning_rate": 0.00017363769900873173, + "loss": 1.0259, + "step": 1591 + }, + { + "epoch": 0.23782491783686885, + "grad_norm": 0.2027769237756729, + "learning_rate": 0.0001736058908775247, + "loss": 0.5823, + "step": 1592 + }, + { + "epoch": 0.2379743053480729, + "grad_norm": 0.12886838614940643, + "learning_rate": 0.00017357406648561052, + "loss": 0.7036, + "step": 1593 + }, + { + "epoch": 0.23812369285927695, + "grad_norm": 0.11439285427331924, + "learning_rate": 0.0001735422258400197, + "loss": 0.6114, + "step": 1594 + }, + { + "epoch": 0.23827308037048103, + "grad_norm": 0.12173706293106079, + "learning_rate": 0.00017351036894778634, + "loss": 0.5466, + "step": 1595 + }, + { + "epoch": 0.23842246788168509, + "grad_norm": 0.15928342938423157, + "learning_rate": 0.0001734784958159481, + "loss": 1.0324, + "step": 1596 + }, + { + "epoch": 0.23857185539288916, + "grad_norm": 0.24861174821853638, + "learning_rate": 0.00017344660645154635, + "loss": 0.954, + "step": 1597 + }, + { + "epoch": 0.23872124290409322, + "grad_norm": 0.24405913054943085, + "learning_rate": 0.00017341470086162586, + "loss": 0.984, + "step": 1598 + }, + { + "epoch": 0.23887063041529727, + "grad_norm": 0.12469593435525894, + "learning_rate": 0.00017338277905323515, + "loss": 0.8056, + "step": 1599 + }, + { + "epoch": 0.23902001792650135, + "grad_norm": 0.13238497078418732, + "learning_rate": 0.0001733508410334262, + "loss": 0.9383, + "step": 1600 + }, + { + "epoch": 0.2391694054377054, + "grad_norm": 0.1149744838476181, + "learning_rate": 0.00017331888680925466, + "loss": 0.7664, + "step": 1601 + }, + { + "epoch": 0.23931879294890948, + "grad_norm": 0.25224781036376953, + "learning_rate": 0.0001732869163877797, + "loss": 0.8615, + "step": 1602 + }, + { + "epoch": 0.23946818046011353, + "grad_norm": 0.12752684950828552, + "learning_rate": 0.00017325492977606413, + "loss": 0.6047, + "step": 1603 + }, + { + "epoch": 0.23961756797131759, + "grad_norm": 0.12648506462574005, + "learning_rate": 0.00017322292698117425, + "loss": 0.6071, + "step": 1604 + }, + { + "epoch": 0.23976695548252167, + "grad_norm": 0.13702405989170074, + "learning_rate": 0.00017319090801018003, + "loss": 0.878, + "step": 1605 + }, + { + "epoch": 0.23991634299372572, + "grad_norm": 0.15288682281970978, + "learning_rate": 0.00017315887287015492, + "loss": 0.807, + "step": 1606 + }, + { + "epoch": 0.2400657305049298, + "grad_norm": 0.2027759850025177, + "learning_rate": 0.00017312682156817602, + "loss": 0.5606, + "step": 1607 + }, + { + "epoch": 0.24021511801613385, + "grad_norm": 0.21917293965816498, + "learning_rate": 0.000173094754111324, + "loss": 0.9878, + "step": 1608 + }, + { + "epoch": 0.2403645055273379, + "grad_norm": 0.24028894305229187, + "learning_rate": 0.000173062670506683, + "loss": 1.0361, + "step": 1609 + }, + { + "epoch": 0.24051389303854198, + "grad_norm": 0.3546265959739685, + "learning_rate": 0.00017303057076134085, + "loss": 1.2186, + "step": 1610 + }, + { + "epoch": 0.24066328054974603, + "grad_norm": 0.12783119082450867, + "learning_rate": 0.0001729984548823889, + "loss": 0.8723, + "step": 1611 + }, + { + "epoch": 0.2408126680609501, + "grad_norm": 0.11363591253757477, + "learning_rate": 0.000172966322876922, + "loss": 0.572, + "step": 1612 + }, + { + "epoch": 0.24096205557215417, + "grad_norm": 0.15105083584785461, + "learning_rate": 0.00017293417475203863, + "loss": 0.7239, + "step": 1613 + }, + { + "epoch": 0.24111144308335822, + "grad_norm": 0.2824059724807739, + "learning_rate": 0.00017290201051484085, + "loss": 1.0547, + "step": 1614 + }, + { + "epoch": 0.2412608305945623, + "grad_norm": 0.21242327988147736, + "learning_rate": 0.00017286983017243424, + "loss": 0.6498, + "step": 1615 + }, + { + "epoch": 0.24141021810576635, + "grad_norm": 0.1261444389820099, + "learning_rate": 0.00017283763373192798, + "loss": 0.8672, + "step": 1616 + }, + { + "epoch": 0.24155960561697043, + "grad_norm": 0.26087701320648193, + "learning_rate": 0.00017280542120043472, + "loss": 0.6234, + "step": 1617 + }, + { + "epoch": 0.24170899312817448, + "grad_norm": 0.1511380523443222, + "learning_rate": 0.00017277319258507073, + "loss": 0.6773, + "step": 1618 + }, + { + "epoch": 0.24185838063937856, + "grad_norm": 0.14391735196113586, + "learning_rate": 0.00017274094789295586, + "loss": 0.8325, + "step": 1619 + }, + { + "epoch": 0.24200776815058261, + "grad_norm": 0.17719805240631104, + "learning_rate": 0.0001727086871312134, + "loss": 0.7249, + "step": 1620 + }, + { + "epoch": 0.24215715566178667, + "grad_norm": 0.1089552789926529, + "learning_rate": 0.00017267641030697034, + "loss": 0.6459, + "step": 1621 + }, + { + "epoch": 0.24230654317299075, + "grad_norm": 0.15426376461982727, + "learning_rate": 0.00017264411742735707, + "loss": 0.8286, + "step": 1622 + }, + { + "epoch": 0.2424559306841948, + "grad_norm": 0.1446702480316162, + "learning_rate": 0.00017261180849950766, + "loss": 0.5768, + "step": 1623 + }, + { + "epoch": 0.24260531819539888, + "grad_norm": 0.23412354290485382, + "learning_rate": 0.00017257948353055963, + "loss": 0.5719, + "step": 1624 + }, + { + "epoch": 0.24275470570660293, + "grad_norm": 0.1535494327545166, + "learning_rate": 0.00017254714252765412, + "loss": 0.7175, + "step": 1625 + }, + { + "epoch": 0.24290409321780698, + "grad_norm": 0.1816725730895996, + "learning_rate": 0.0001725147854979357, + "loss": 0.8062, + "step": 1626 + }, + { + "epoch": 0.24305348072901106, + "grad_norm": 0.2537088692188263, + "learning_rate": 0.00017248241244855264, + "loss": 0.8809, + "step": 1627 + }, + { + "epoch": 0.24320286824021511, + "grad_norm": 0.12580998241901398, + "learning_rate": 0.00017245002338665656, + "loss": 0.5737, + "step": 1628 + }, + { + "epoch": 0.2433522557514192, + "grad_norm": 0.18232382833957672, + "learning_rate": 0.0001724176183194028, + "loss": 0.5088, + "step": 1629 + }, + { + "epoch": 0.24350164326262325, + "grad_norm": 0.14372123777866364, + "learning_rate": 0.00017238519725395007, + "loss": 0.8043, + "step": 1630 + }, + { + "epoch": 0.2436510307738273, + "grad_norm": 0.1633867621421814, + "learning_rate": 0.00017235276019746077, + "loss": 0.8469, + "step": 1631 + }, + { + "epoch": 0.24380041828503138, + "grad_norm": 0.23584675788879395, + "learning_rate": 0.00017232030715710076, + "loss": 0.9823, + "step": 1632 + }, + { + "epoch": 0.24394980579623543, + "grad_norm": 0.13388711214065552, + "learning_rate": 0.00017228783814003936, + "loss": 0.7298, + "step": 1633 + }, + { + "epoch": 0.2440991933074395, + "grad_norm": 0.14213724434375763, + "learning_rate": 0.00017225535315344955, + "loss": 0.7396, + "step": 1634 + }, + { + "epoch": 0.24424858081864356, + "grad_norm": 0.15166963636875153, + "learning_rate": 0.0001722228522045078, + "loss": 0.735, + "step": 1635 + }, + { + "epoch": 0.24439796832984761, + "grad_norm": 0.1204567551612854, + "learning_rate": 0.00017219033530039397, + "loss": 0.6299, + "step": 1636 + }, + { + "epoch": 0.2445473558410517, + "grad_norm": 0.11346624791622162, + "learning_rate": 0.0001721578024482917, + "loss": 0.6771, + "step": 1637 + }, + { + "epoch": 0.24469674335225575, + "grad_norm": 0.16759368777275085, + "learning_rate": 0.00017212525365538792, + "loss": 0.7302, + "step": 1638 + }, + { + "epoch": 0.24484613086345983, + "grad_norm": 0.12218065559864044, + "learning_rate": 0.0001720926889288732, + "loss": 0.7969, + "step": 1639 + }, + { + "epoch": 0.24499551837466388, + "grad_norm": 0.11079906672239304, + "learning_rate": 0.00017206010827594163, + "loss": 0.5828, + "step": 1640 + }, + { + "epoch": 0.24514490588586793, + "grad_norm": 0.1696266084909439, + "learning_rate": 0.00017202751170379075, + "loss": 1.0499, + "step": 1641 + }, + { + "epoch": 0.245294293397072, + "grad_norm": 0.2520163357257843, + "learning_rate": 0.0001719948992196217, + "loss": 1.0662, + "step": 1642 + }, + { + "epoch": 0.24544368090827606, + "grad_norm": 0.1606278270483017, + "learning_rate": 0.00017196227083063906, + "loss": 0.6013, + "step": 1643 + }, + { + "epoch": 0.24559306841948014, + "grad_norm": 0.11815240234136581, + "learning_rate": 0.00017192962654405096, + "loss": 0.5613, + "step": 1644 + }, + { + "epoch": 0.2457424559306842, + "grad_norm": 0.1555798351764679, + "learning_rate": 0.00017189696636706904, + "loss": 0.9341, + "step": 1645 + }, + { + "epoch": 0.24589184344188825, + "grad_norm": 0.1412414312362671, + "learning_rate": 0.00017186429030690848, + "loss": 0.8947, + "step": 1646 + }, + { + "epoch": 0.24604123095309233, + "grad_norm": 0.1420399695634842, + "learning_rate": 0.00017183159837078792, + "loss": 0.7541, + "step": 1647 + }, + { + "epoch": 0.24619061846429638, + "grad_norm": 0.16363643109798431, + "learning_rate": 0.00017179889056592954, + "loss": 0.7413, + "step": 1648 + }, + { + "epoch": 0.24634000597550046, + "grad_norm": 0.27612796425819397, + "learning_rate": 0.000171766166899559, + "loss": 1.1101, + "step": 1649 + }, + { + "epoch": 0.2464893934867045, + "grad_norm": 0.169744610786438, + "learning_rate": 0.00017173342737890544, + "loss": 0.7638, + "step": 1650 + }, + { + "epoch": 0.24663878099790856, + "grad_norm": 0.2400025874376297, + "learning_rate": 0.0001717006720112016, + "loss": 0.6642, + "step": 1651 + }, + { + "epoch": 0.24678816850911264, + "grad_norm": 0.12115002423524857, + "learning_rate": 0.00017166790080368357, + "loss": 0.5818, + "step": 1652 + }, + { + "epoch": 0.2469375560203167, + "grad_norm": 0.11767467111349106, + "learning_rate": 0.00017163511376359116, + "loss": 0.7801, + "step": 1653 + }, + { + "epoch": 0.24708694353152078, + "grad_norm": 0.18294836580753326, + "learning_rate": 0.00017160231089816748, + "loss": 0.9749, + "step": 1654 + }, + { + "epoch": 0.24723633104272483, + "grad_norm": 0.16762547194957733, + "learning_rate": 0.00017156949221465916, + "loss": 0.5872, + "step": 1655 + }, + { + "epoch": 0.24738571855392888, + "grad_norm": 0.28492650389671326, + "learning_rate": 0.00017153665772031643, + "loss": 0.9859, + "step": 1656 + }, + { + "epoch": 0.24753510606513296, + "grad_norm": 0.516017735004425, + "learning_rate": 0.00017150380742239293, + "loss": 1.2148, + "step": 1657 + }, + { + "epoch": 0.247684493576337, + "grad_norm": 0.1820579469203949, + "learning_rate": 0.0001714709413281458, + "loss": 0.8255, + "step": 1658 + }, + { + "epoch": 0.2478338810875411, + "grad_norm": 0.1763305366039276, + "learning_rate": 0.0001714380594448357, + "loss": 0.7998, + "step": 1659 + }, + { + "epoch": 0.24798326859874514, + "grad_norm": 0.1392301470041275, + "learning_rate": 0.00017140516177972676, + "loss": 0.8738, + "step": 1660 + }, + { + "epoch": 0.2481326561099492, + "grad_norm": 0.2311377078294754, + "learning_rate": 0.00017137224834008657, + "loss": 0.9255, + "step": 1661 + }, + { + "epoch": 0.24828204362115328, + "grad_norm": 0.14632488787174225, + "learning_rate": 0.00017133931913318625, + "loss": 0.8224, + "step": 1662 + }, + { + "epoch": 0.24843143113235733, + "grad_norm": 0.1478954702615738, + "learning_rate": 0.00017130637416630035, + "loss": 0.6986, + "step": 1663 + }, + { + "epoch": 0.2485808186435614, + "grad_norm": 0.161132350564003, + "learning_rate": 0.00017127341344670696, + "loss": 0.7352, + "step": 1664 + }, + { + "epoch": 0.24873020615476546, + "grad_norm": 0.13480094075202942, + "learning_rate": 0.00017124043698168764, + "loss": 0.7767, + "step": 1665 + }, + { + "epoch": 0.2488795936659695, + "grad_norm": 0.25371015071868896, + "learning_rate": 0.00017120744477852745, + "loss": 1.0407, + "step": 1666 + }, + { + "epoch": 0.2490289811771736, + "grad_norm": 0.13566429913043976, + "learning_rate": 0.00017117443684451478, + "loss": 0.7924, + "step": 1667 + }, + { + "epoch": 0.24917836868837764, + "grad_norm": 0.15394295752048492, + "learning_rate": 0.00017114141318694167, + "loss": 0.6343, + "step": 1668 + }, + { + "epoch": 0.24932775619958172, + "grad_norm": 0.15498654544353485, + "learning_rate": 0.0001711083738131036, + "loss": 0.5479, + "step": 1669 + }, + { + "epoch": 0.24947714371078578, + "grad_norm": 0.11515049636363983, + "learning_rate": 0.00017107531873029942, + "loss": 0.7509, + "step": 1670 + }, + { + "epoch": 0.24962653122198983, + "grad_norm": 0.12256692349910736, + "learning_rate": 0.00017104224794583158, + "loss": 0.4964, + "step": 1671 + }, + { + "epoch": 0.2497759187331939, + "grad_norm": 0.5561460852622986, + "learning_rate": 0.0001710091614670059, + "loss": 1.3467, + "step": 1672 + }, + { + "epoch": 0.24992530624439796, + "grad_norm": 0.1814536154270172, + "learning_rate": 0.00017097605930113175, + "loss": 0.7138, + "step": 1673 + }, + { + "epoch": 0.25007469375560204, + "grad_norm": 0.19458015263080597, + "learning_rate": 0.00017094294145552188, + "loss": 0.8228, + "step": 1674 + }, + { + "epoch": 0.2502240812668061, + "grad_norm": 0.24046920239925385, + "learning_rate": 0.00017090980793749258, + "loss": 0.6464, + "step": 1675 + }, + { + "epoch": 0.25037346877801014, + "grad_norm": 0.1473151296377182, + "learning_rate": 0.00017087665875436354, + "loss": 0.4695, + "step": 1676 + }, + { + "epoch": 0.2505228562892142, + "grad_norm": 0.1309395283460617, + "learning_rate": 0.00017084349391345796, + "loss": 0.5628, + "step": 1677 + }, + { + "epoch": 0.2506722438004183, + "grad_norm": 0.1466517597436905, + "learning_rate": 0.00017081031342210245, + "loss": 0.8019, + "step": 1678 + }, + { + "epoch": 0.25082163131162233, + "grad_norm": 0.12185301631689072, + "learning_rate": 0.00017077711728762714, + "loss": 0.7219, + "step": 1679 + }, + { + "epoch": 0.2509710188228264, + "grad_norm": 0.17365945875644684, + "learning_rate": 0.0001707439055173656, + "loss": 0.7096, + "step": 1680 + }, + { + "epoch": 0.2511204063340305, + "grad_norm": 0.1582822948694229, + "learning_rate": 0.00017071067811865476, + "loss": 0.7828, + "step": 1681 + }, + { + "epoch": 0.2512697938452345, + "grad_norm": 0.1119188591837883, + "learning_rate": 0.00017067743509883515, + "loss": 0.612, + "step": 1682 + }, + { + "epoch": 0.2514191813564386, + "grad_norm": 0.11489804089069366, + "learning_rate": 0.00017064417646525066, + "loss": 0.653, + "step": 1683 + }, + { + "epoch": 0.2515685688676427, + "grad_norm": 0.11803659796714783, + "learning_rate": 0.00017061090222524863, + "loss": 0.7372, + "step": 1684 + }, + { + "epoch": 0.25171795637884675, + "grad_norm": 0.1338784098625183, + "learning_rate": 0.00017057761238617984, + "loss": 0.5076, + "step": 1685 + }, + { + "epoch": 0.2518673438900508, + "grad_norm": 0.148620143532753, + "learning_rate": 0.00017054430695539864, + "loss": 0.8958, + "step": 1686 + }, + { + "epoch": 0.25201673140125486, + "grad_norm": 0.13378937542438507, + "learning_rate": 0.00017051098594026267, + "loss": 0.5958, + "step": 1687 + }, + { + "epoch": 0.25216611891245894, + "grad_norm": 0.24176190793514252, + "learning_rate": 0.00017047764934813303, + "loss": 1.1407, + "step": 1688 + }, + { + "epoch": 0.25231550642366296, + "grad_norm": 0.18258847296237946, + "learning_rate": 0.00017044429718637437, + "loss": 0.7464, + "step": 1689 + }, + { + "epoch": 0.25246489393486704, + "grad_norm": 0.1730254888534546, + "learning_rate": 0.00017041092946235467, + "loss": 0.5463, + "step": 1690 + }, + { + "epoch": 0.2526142814460711, + "grad_norm": 0.09865910559892654, + "learning_rate": 0.0001703775461834454, + "loss": 0.6086, + "step": 1691 + }, + { + "epoch": 0.25276366895727514, + "grad_norm": 0.16275614500045776, + "learning_rate": 0.00017034414735702145, + "loss": 0.8216, + "step": 1692 + }, + { + "epoch": 0.2529130564684792, + "grad_norm": 0.22617553174495697, + "learning_rate": 0.00017031073299046117, + "loss": 1.282, + "step": 1693 + }, + { + "epoch": 0.2530624439796833, + "grad_norm": 0.13999703526496887, + "learning_rate": 0.0001702773030911463, + "loss": 0.5571, + "step": 1694 + }, + { + "epoch": 0.2532118314908874, + "grad_norm": 0.22081904113292694, + "learning_rate": 0.00017024385766646204, + "loss": 0.716, + "step": 1695 + }, + { + "epoch": 0.2533612190020914, + "grad_norm": 0.1647699475288391, + "learning_rate": 0.00017021039672379703, + "loss": 0.7478, + "step": 1696 + }, + { + "epoch": 0.2535106065132955, + "grad_norm": 1.2335481643676758, + "learning_rate": 0.0001701769202705433, + "loss": 1.2852, + "step": 1697 + }, + { + "epoch": 0.25365999402449957, + "grad_norm": 0.13648250699043274, + "learning_rate": 0.00017014342831409634, + "loss": 0.7554, + "step": 1698 + }, + { + "epoch": 0.2538093815357036, + "grad_norm": 0.17244575917720795, + "learning_rate": 0.00017010992086185505, + "loss": 0.839, + "step": 1699 + }, + { + "epoch": 0.2539587690469077, + "grad_norm": 0.1780940443277359, + "learning_rate": 0.00017007639792122173, + "loss": 0.3983, + "step": 1700 + }, + { + "epoch": 0.25410815655811175, + "grad_norm": 0.10594518482685089, + "learning_rate": 0.00017004285949960218, + "loss": 0.5268, + "step": 1701 + }, + { + "epoch": 0.25425754406931583, + "grad_norm": 0.24656297266483307, + "learning_rate": 0.00017000930560440554, + "loss": 1.0687, + "step": 1702 + }, + { + "epoch": 0.25440693158051986, + "grad_norm": 0.11236447095870972, + "learning_rate": 0.00016997573624304445, + "loss": 0.7733, + "step": 1703 + }, + { + "epoch": 0.25455631909172394, + "grad_norm": 0.19340038299560547, + "learning_rate": 0.0001699421514229348, + "loss": 0.6995, + "step": 1704 + }, + { + "epoch": 0.254705706602928, + "grad_norm": 0.16833987832069397, + "learning_rate": 0.00016990855115149608, + "loss": 0.4307, + "step": 1705 + }, + { + "epoch": 0.25485509411413204, + "grad_norm": 0.14893771708011627, + "learning_rate": 0.00016987493543615115, + "loss": 0.6488, + "step": 1706 + }, + { + "epoch": 0.2550044816253361, + "grad_norm": 0.2579140067100525, + "learning_rate": 0.0001698413042843262, + "loss": 0.7579, + "step": 1707 + }, + { + "epoch": 0.2551538691365402, + "grad_norm": 0.12917333841323853, + "learning_rate": 0.0001698076577034509, + "loss": 0.4014, + "step": 1708 + }, + { + "epoch": 0.2553032566477442, + "grad_norm": 0.12658651173114777, + "learning_rate": 0.00016977399570095836, + "loss": 0.8899, + "step": 1709 + }, + { + "epoch": 0.2554526441589483, + "grad_norm": 0.18011939525604248, + "learning_rate": 0.00016974031828428495, + "loss": 0.4706, + "step": 1710 + }, + { + "epoch": 0.2556020316701524, + "grad_norm": 0.1294494867324829, + "learning_rate": 0.00016970662546087061, + "loss": 0.7174, + "step": 1711 + }, + { + "epoch": 0.25575141918135647, + "grad_norm": 0.11419305950403214, + "learning_rate": 0.00016967291723815863, + "loss": 0.7966, + "step": 1712 + }, + { + "epoch": 0.2559008066925605, + "grad_norm": 0.13392961025238037, + "learning_rate": 0.0001696391936235957, + "loss": 0.8852, + "step": 1713 + }, + { + "epoch": 0.25605019420376457, + "grad_norm": 0.10392757505178452, + "learning_rate": 0.00016960545462463183, + "loss": 0.5903, + "step": 1714 + }, + { + "epoch": 0.25619958171496865, + "grad_norm": 0.1128273457288742, + "learning_rate": 0.00016957170024872053, + "loss": 0.7823, + "step": 1715 + }, + { + "epoch": 0.2563489692261727, + "grad_norm": 0.16612160205841064, + "learning_rate": 0.0001695379305033187, + "loss": 0.6937, + "step": 1716 + }, + { + "epoch": 0.25649835673737675, + "grad_norm": 0.17383776605129242, + "learning_rate": 0.00016950414539588658, + "loss": 0.7404, + "step": 1717 + }, + { + "epoch": 0.25664774424858083, + "grad_norm": 0.19787214696407318, + "learning_rate": 0.00016947034493388786, + "loss": 0.5815, + "step": 1718 + }, + { + "epoch": 0.25679713175978486, + "grad_norm": 0.25284725427627563, + "learning_rate": 0.00016943652912478963, + "loss": 0.8694, + "step": 1719 + }, + { + "epoch": 0.25694651927098894, + "grad_norm": 0.15624594688415527, + "learning_rate": 0.00016940269797606228, + "loss": 0.6359, + "step": 1720 + }, + { + "epoch": 0.257095906782193, + "grad_norm": 0.13568609952926636, + "learning_rate": 0.00016936885149517966, + "loss": 0.6661, + "step": 1721 + }, + { + "epoch": 0.2572452942933971, + "grad_norm": 0.1571207344532013, + "learning_rate": 0.00016933498968961898, + "loss": 0.5693, + "step": 1722 + }, + { + "epoch": 0.2573946818046011, + "grad_norm": 0.18845240771770477, + "learning_rate": 0.0001693011125668609, + "loss": 0.8907, + "step": 1723 + }, + { + "epoch": 0.2575440693158052, + "grad_norm": 0.16922013461589813, + "learning_rate": 0.00016926722013438936, + "loss": 0.7377, + "step": 1724 + }, + { + "epoch": 0.2576934568270093, + "grad_norm": 0.15387395024299622, + "learning_rate": 0.00016923331239969175, + "loss": 0.9347, + "step": 1725 + }, + { + "epoch": 0.2578428443382133, + "grad_norm": 0.19560673832893372, + "learning_rate": 0.00016919938937025886, + "loss": 0.5263, + "step": 1726 + }, + { + "epoch": 0.2579922318494174, + "grad_norm": 1.0470530986785889, + "learning_rate": 0.00016916545105358477, + "loss": 1.2196, + "step": 1727 + }, + { + "epoch": 0.25814161936062147, + "grad_norm": 0.104112908244133, + "learning_rate": 0.00016913149745716703, + "loss": 0.5982, + "step": 1728 + }, + { + "epoch": 0.2582910068718255, + "grad_norm": 0.22203882038593292, + "learning_rate": 0.00016909752858850649, + "loss": 0.9925, + "step": 1729 + }, + { + "epoch": 0.25844039438302957, + "grad_norm": 0.13334202766418457, + "learning_rate": 0.00016906354445510747, + "loss": 0.8159, + "step": 1730 + }, + { + "epoch": 0.25858978189423365, + "grad_norm": 0.1822672188282013, + "learning_rate": 0.00016902954506447756, + "loss": 0.5895, + "step": 1731 + }, + { + "epoch": 0.25873916940543773, + "grad_norm": 0.13168515264987946, + "learning_rate": 0.0001689955304241278, + "loss": 0.8679, + "step": 1732 + }, + { + "epoch": 0.25888855691664175, + "grad_norm": 0.1478189080953598, + "learning_rate": 0.00016896150054157246, + "loss": 1.0138, + "step": 1733 + }, + { + "epoch": 0.25903794442784583, + "grad_norm": 0.12031956017017365, + "learning_rate": 0.0001689274554243294, + "loss": 0.6412, + "step": 1734 + }, + { + "epoch": 0.2591873319390499, + "grad_norm": 0.13320299983024597, + "learning_rate": 0.00016889339507991972, + "loss": 0.8683, + "step": 1735 + }, + { + "epoch": 0.25933671945025394, + "grad_norm": 0.2738458812236786, + "learning_rate": 0.00016885931951586783, + "loss": 1.3964, + "step": 1736 + }, + { + "epoch": 0.259486106961458, + "grad_norm": 0.18454548716545105, + "learning_rate": 0.00016882522873970157, + "loss": 0.6525, + "step": 1737 + }, + { + "epoch": 0.2596354944726621, + "grad_norm": 0.4490959644317627, + "learning_rate": 0.00016879112275895215, + "loss": 0.8293, + "step": 1738 + }, + { + "epoch": 0.2597848819838661, + "grad_norm": 0.4163506329059601, + "learning_rate": 0.0001687570015811541, + "loss": 0.86, + "step": 1739 + }, + { + "epoch": 0.2599342694950702, + "grad_norm": 0.14413362741470337, + "learning_rate": 0.00016872286521384537, + "loss": 0.5609, + "step": 1740 + }, + { + "epoch": 0.2600836570062743, + "grad_norm": 0.1181647926568985, + "learning_rate": 0.0001686887136645672, + "loss": 0.7817, + "step": 1741 + }, + { + "epoch": 0.26023304451747836, + "grad_norm": 0.13126704096794128, + "learning_rate": 0.0001686545469408642, + "loss": 0.7343, + "step": 1742 + }, + { + "epoch": 0.2603824320286824, + "grad_norm": 0.11598172783851624, + "learning_rate": 0.00016862036505028435, + "loss": 0.774, + "step": 1743 + }, + { + "epoch": 0.26053181953988647, + "grad_norm": 0.11938168853521347, + "learning_rate": 0.000168586168000379, + "loss": 0.6485, + "step": 1744 + }, + { + "epoch": 0.26068120705109055, + "grad_norm": 0.15194714069366455, + "learning_rate": 0.00016855195579870276, + "loss": 0.7801, + "step": 1745 + }, + { + "epoch": 0.26083059456229457, + "grad_norm": 0.11620033532381058, + "learning_rate": 0.00016851772845281367, + "loss": 0.6402, + "step": 1746 + }, + { + "epoch": 0.26097998207349865, + "grad_norm": 0.11883977055549622, + "learning_rate": 0.0001684834859702731, + "loss": 0.8832, + "step": 1747 + }, + { + "epoch": 0.26112936958470273, + "grad_norm": 0.17073407769203186, + "learning_rate": 0.00016844922835864575, + "loss": 0.7806, + "step": 1748 + }, + { + "epoch": 0.26127875709590676, + "grad_norm": 0.15572674572467804, + "learning_rate": 0.0001684149556254997, + "loss": 0.7371, + "step": 1749 + }, + { + "epoch": 0.26142814460711084, + "grad_norm": 0.11433901637792587, + "learning_rate": 0.00016838066777840629, + "loss": 0.5395, + "step": 1750 + }, + { + "epoch": 0.2615775321183149, + "grad_norm": 0.23120561242103577, + "learning_rate": 0.0001683463648249403, + "loss": 0.861, + "step": 1751 + }, + { + "epoch": 0.261726919629519, + "grad_norm": 0.21362490952014923, + "learning_rate": 0.00016831204677267975, + "loss": 0.6806, + "step": 1752 + }, + { + "epoch": 0.261876307140723, + "grad_norm": 0.16032354533672333, + "learning_rate": 0.0001682777136292061, + "loss": 0.5942, + "step": 1753 + }, + { + "epoch": 0.2620256946519271, + "grad_norm": 0.11645185202360153, + "learning_rate": 0.00016824336540210402, + "loss": 0.6547, + "step": 1754 + }, + { + "epoch": 0.2621750821631312, + "grad_norm": 0.15311047434806824, + "learning_rate": 0.0001682090020989616, + "loss": 0.6924, + "step": 1755 + }, + { + "epoch": 0.2623244696743352, + "grad_norm": 0.11865057051181793, + "learning_rate": 0.0001681746237273702, + "loss": 0.5077, + "step": 1756 + }, + { + "epoch": 0.2624738571855393, + "grad_norm": 0.16047270596027374, + "learning_rate": 0.00016814023029492465, + "loss": 0.9401, + "step": 1757 + }, + { + "epoch": 0.26262324469674336, + "grad_norm": 0.30298060178756714, + "learning_rate": 0.00016810582180922293, + "loss": 0.9419, + "step": 1758 + }, + { + "epoch": 0.26277263220794744, + "grad_norm": 0.17041756212711334, + "learning_rate": 0.0001680713982778664, + "loss": 0.6029, + "step": 1759 + }, + { + "epoch": 0.26292201971915147, + "grad_norm": 0.12116200476884842, + "learning_rate": 0.00016803695970845985, + "loss": 0.5324, + "step": 1760 + }, + { + "epoch": 0.26307140723035555, + "grad_norm": 0.2546985149383545, + "learning_rate": 0.00016800250610861123, + "loss": 0.8727, + "step": 1761 + }, + { + "epoch": 0.2632207947415596, + "grad_norm": 0.16526702046394348, + "learning_rate": 0.0001679680374859319, + "loss": 0.8323, + "step": 1762 + }, + { + "epoch": 0.26337018225276365, + "grad_norm": 0.13681448996067047, + "learning_rate": 0.00016793355384803652, + "loss": 0.5819, + "step": 1763 + }, + { + "epoch": 0.26351956976396773, + "grad_norm": 0.1449958086013794, + "learning_rate": 0.0001678990552025431, + "loss": 0.8017, + "step": 1764 + }, + { + "epoch": 0.2636689572751718, + "grad_norm": 3.745661973953247, + "learning_rate": 0.00016786454155707289, + "loss": 2.6157, + "step": 1765 + }, + { + "epoch": 0.26381834478637584, + "grad_norm": 0.1667809933423996, + "learning_rate": 0.00016783001291925055, + "loss": 0.5333, + "step": 1766 + }, + { + "epoch": 0.2639677322975799, + "grad_norm": 0.1692322939634323, + "learning_rate": 0.000167795469296704, + "loss": 0.8491, + "step": 1767 + }, + { + "epoch": 0.264117119808784, + "grad_norm": 0.18635401129722595, + "learning_rate": 0.00016776091069706442, + "loss": 0.7551, + "step": 1768 + }, + { + "epoch": 0.2642665073199881, + "grad_norm": 0.1936548948287964, + "learning_rate": 0.00016772633712796643, + "loss": 0.6007, + "step": 1769 + }, + { + "epoch": 0.2644158948311921, + "grad_norm": 0.22752203047275543, + "learning_rate": 0.00016769174859704783, + "loss": 1.2796, + "step": 1770 + }, + { + "epoch": 0.2645652823423962, + "grad_norm": 0.22011956572532654, + "learning_rate": 0.00016765714511194978, + "loss": 0.7837, + "step": 1771 + }, + { + "epoch": 0.26471466985360026, + "grad_norm": 0.12708061933517456, + "learning_rate": 0.00016762252668031674, + "loss": 0.8151, + "step": 1772 + }, + { + "epoch": 0.2648640573648043, + "grad_norm": 0.16485874354839325, + "learning_rate": 0.00016758789330979646, + "loss": 0.7342, + "step": 1773 + }, + { + "epoch": 0.26501344487600836, + "grad_norm": 0.15620820224285126, + "learning_rate": 0.00016755324500804, + "loss": 0.7541, + "step": 1774 + }, + { + "epoch": 0.26516283238721244, + "grad_norm": 0.13292229175567627, + "learning_rate": 0.00016751858178270176, + "loss": 0.7372, + "step": 1775 + }, + { + "epoch": 0.26531221989841647, + "grad_norm": 0.1720595508813858, + "learning_rate": 0.00016748390364143938, + "loss": 0.6299, + "step": 1776 + }, + { + "epoch": 0.26546160740962055, + "grad_norm": 0.15689946711063385, + "learning_rate": 0.00016744921059191377, + "loss": 0.8878, + "step": 1777 + }, + { + "epoch": 0.26561099492082463, + "grad_norm": 0.3296888470649719, + "learning_rate": 0.00016741450264178917, + "loss": 0.8915, + "step": 1778 + }, + { + "epoch": 0.2657603824320287, + "grad_norm": 0.2351844161748886, + "learning_rate": 0.00016737977979873318, + "loss": 1.1235, + "step": 1779 + }, + { + "epoch": 0.26590976994323273, + "grad_norm": 0.19016700983047485, + "learning_rate": 0.00016734504207041663, + "loss": 0.5777, + "step": 1780 + }, + { + "epoch": 0.2660591574544368, + "grad_norm": 0.15265557169914246, + "learning_rate": 0.00016731028946451353, + "loss": 0.5994, + "step": 1781 + }, + { + "epoch": 0.2662085449656409, + "grad_norm": 0.17204977571964264, + "learning_rate": 0.00016727552198870135, + "loss": 0.6589, + "step": 1782 + }, + { + "epoch": 0.2663579324768449, + "grad_norm": 0.12247525155544281, + "learning_rate": 0.00016724073965066082, + "loss": 0.8026, + "step": 1783 + }, + { + "epoch": 0.266507319988049, + "grad_norm": 0.3057596981525421, + "learning_rate": 0.00016720594245807582, + "loss": 0.8267, + "step": 1784 + }, + { + "epoch": 0.2666567074992531, + "grad_norm": 0.1400333046913147, + "learning_rate": 0.00016717113041863369, + "loss": 0.6098, + "step": 1785 + }, + { + "epoch": 0.2668060950104571, + "grad_norm": 0.13758665323257446, + "learning_rate": 0.00016713630354002484, + "loss": 0.8526, + "step": 1786 + }, + { + "epoch": 0.2669554825216612, + "grad_norm": 0.12495380640029907, + "learning_rate": 0.00016710146182994322, + "loss": 0.6674, + "step": 1787 + }, + { + "epoch": 0.26710487003286526, + "grad_norm": 0.12957152724266052, + "learning_rate": 0.00016706660529608583, + "loss": 0.8127, + "step": 1788 + }, + { + "epoch": 0.26725425754406934, + "grad_norm": 0.18763193488121033, + "learning_rate": 0.00016703173394615303, + "loss": 0.9102, + "step": 1789 + }, + { + "epoch": 0.26740364505527336, + "grad_norm": 0.17572201788425446, + "learning_rate": 0.0001669968477878485, + "loss": 0.8989, + "step": 1790 + }, + { + "epoch": 0.26755303256647744, + "grad_norm": 0.16582119464874268, + "learning_rate": 0.00016696194682887914, + "loss": 1.0639, + "step": 1791 + }, + { + "epoch": 0.2677024200776815, + "grad_norm": 0.42142021656036377, + "learning_rate": 0.00016692703107695507, + "loss": 1.1036, + "step": 1792 + }, + { + "epoch": 0.26785180758888555, + "grad_norm": 0.21679432690143585, + "learning_rate": 0.0001668921005397898, + "loss": 0.5188, + "step": 1793 + }, + { + "epoch": 0.26800119510008963, + "grad_norm": 0.16721928119659424, + "learning_rate": 0.00016685715522509994, + "loss": 0.8261, + "step": 1794 + }, + { + "epoch": 0.2681505826112937, + "grad_norm": 0.1881019026041031, + "learning_rate": 0.00016682219514060556, + "loss": 0.9753, + "step": 1795 + }, + { + "epoch": 0.26829997012249773, + "grad_norm": 0.18447300791740417, + "learning_rate": 0.0001667872202940299, + "loss": 0.8063, + "step": 1796 + }, + { + "epoch": 0.2684493576337018, + "grad_norm": 0.13168549537658691, + "learning_rate": 0.00016675223069309935, + "loss": 0.7722, + "step": 1797 + }, + { + "epoch": 0.2685987451449059, + "grad_norm": 0.3020504117012024, + "learning_rate": 0.0001667172263455438, + "loss": 0.8602, + "step": 1798 + }, + { + "epoch": 0.26874813265611, + "grad_norm": 0.13748160004615784, + "learning_rate": 0.00016668220725909617, + "loss": 0.9058, + "step": 1799 + }, + { + "epoch": 0.268897520167314, + "grad_norm": 0.19238565862178802, + "learning_rate": 0.00016664717344149277, + "loss": 0.6365, + "step": 1800 + }, + { + "epoch": 0.2690469076785181, + "grad_norm": 0.2118740677833557, + "learning_rate": 0.00016661212490047314, + "loss": 0.7105, + "step": 1801 + }, + { + "epoch": 0.26919629518972216, + "grad_norm": 0.17175479233264923, + "learning_rate": 0.00016657706164378, + "loss": 0.6657, + "step": 1802 + }, + { + "epoch": 0.2693456827009262, + "grad_norm": 0.1766873598098755, + "learning_rate": 0.00016654198367915948, + "loss": 0.9525, + "step": 1803 + }, + { + "epoch": 0.26949507021213026, + "grad_norm": 0.2608683705329895, + "learning_rate": 0.00016650689101436073, + "loss": 1.3181, + "step": 1804 + }, + { + "epoch": 0.26964445772333434, + "grad_norm": 0.18048213422298431, + "learning_rate": 0.00016647178365713637, + "loss": 0.6132, + "step": 1805 + }, + { + "epoch": 0.26979384523453837, + "grad_norm": 0.17701251804828644, + "learning_rate": 0.00016643666161524217, + "loss": 0.7133, + "step": 1806 + }, + { + "epoch": 0.26994323274574245, + "grad_norm": 0.25974783301353455, + "learning_rate": 0.00016640152489643705, + "loss": 1.0835, + "step": 1807 + }, + { + "epoch": 0.2700926202569465, + "grad_norm": 0.1478545069694519, + "learning_rate": 0.00016636637350848338, + "loss": 0.8298, + "step": 1808 + }, + { + "epoch": 0.2702420077681506, + "grad_norm": 0.12416455149650574, + "learning_rate": 0.00016633120745914663, + "loss": 0.775, + "step": 1809 + }, + { + "epoch": 0.27039139527935463, + "grad_norm": 0.14755703508853912, + "learning_rate": 0.00016629602675619548, + "loss": 0.7025, + "step": 1810 + }, + { + "epoch": 0.2705407827905587, + "grad_norm": 0.2348576784133911, + "learning_rate": 0.00016626083140740193, + "loss": 0.5877, + "step": 1811 + }, + { + "epoch": 0.2706901703017628, + "grad_norm": 0.19373933970928192, + "learning_rate": 0.0001662256214205413, + "loss": 0.6254, + "step": 1812 + }, + { + "epoch": 0.2708395578129668, + "grad_norm": 0.2443433403968811, + "learning_rate": 0.00016619039680339183, + "loss": 1.0805, + "step": 1813 + }, + { + "epoch": 0.2709889453241709, + "grad_norm": 0.1242481917142868, + "learning_rate": 0.00016615515756373533, + "loss": 0.8073, + "step": 1814 + }, + { + "epoch": 0.271138332835375, + "grad_norm": 0.14438317716121674, + "learning_rate": 0.00016611990370935672, + "loss": 0.8175, + "step": 1815 + }, + { + "epoch": 0.27128772034657905, + "grad_norm": 0.24103055894374847, + "learning_rate": 0.00016608463524804407, + "loss": 0.6098, + "step": 1816 + }, + { + "epoch": 0.2714371078577831, + "grad_norm": 0.11732879281044006, + "learning_rate": 0.00016604935218758878, + "loss": 0.5935, + "step": 1817 + }, + { + "epoch": 0.27158649536898716, + "grad_norm": 0.13892412185668945, + "learning_rate": 0.0001660140545357854, + "loss": 0.6891, + "step": 1818 + }, + { + "epoch": 0.27173588288019124, + "grad_norm": 0.1574752926826477, + "learning_rate": 0.0001659787423004318, + "loss": 0.5805, + "step": 1819 + }, + { + "epoch": 0.27188527039139526, + "grad_norm": 0.1563543677330017, + "learning_rate": 0.00016594341548932894, + "loss": 0.7904, + "step": 1820 + }, + { + "epoch": 0.27203465790259934, + "grad_norm": 0.2136799395084381, + "learning_rate": 0.0001659080741102811, + "loss": 0.7477, + "step": 1821 + }, + { + "epoch": 0.2721840454138034, + "grad_norm": 0.11503917723894119, + "learning_rate": 0.0001658727181710958, + "loss": 0.8353, + "step": 1822 + }, + { + "epoch": 0.27233343292500745, + "grad_norm": 0.10798639804124832, + "learning_rate": 0.00016583734767958367, + "loss": 0.7049, + "step": 1823 + }, + { + "epoch": 0.2724828204362115, + "grad_norm": 0.11826090514659882, + "learning_rate": 0.0001658019626435586, + "loss": 0.8643, + "step": 1824 + }, + { + "epoch": 0.2726322079474156, + "grad_norm": 0.19389161467552185, + "learning_rate": 0.00016576656307083778, + "loss": 0.585, + "step": 1825 + }, + { + "epoch": 0.2727815954586197, + "grad_norm": 0.13553759455680847, + "learning_rate": 0.00016573114896924147, + "loss": 0.7292, + "step": 1826 + }, + { + "epoch": 0.2729309829698237, + "grad_norm": 0.1668972373008728, + "learning_rate": 0.00016569572034659324, + "loss": 0.6368, + "step": 1827 + }, + { + "epoch": 0.2730803704810278, + "grad_norm": 0.22873035073280334, + "learning_rate": 0.0001656602772107198, + "loss": 0.8858, + "step": 1828 + }, + { + "epoch": 0.27322975799223187, + "grad_norm": 0.13842542469501495, + "learning_rate": 0.0001656248195694511, + "loss": 0.6738, + "step": 1829 + }, + { + "epoch": 0.2733791455034359, + "grad_norm": 0.14224855601787567, + "learning_rate": 0.00016558934743062035, + "loss": 0.6905, + "step": 1830 + }, + { + "epoch": 0.27352853301464, + "grad_norm": 0.28141728043556213, + "learning_rate": 0.00016555386080206383, + "loss": 0.6765, + "step": 1831 + }, + { + "epoch": 0.27367792052584405, + "grad_norm": 0.16277846693992615, + "learning_rate": 0.00016551835969162118, + "loss": 0.8857, + "step": 1832 + }, + { + "epoch": 0.2738273080370481, + "grad_norm": 0.1706254631280899, + "learning_rate": 0.0001654828441071351, + "loss": 0.6825, + "step": 1833 + }, + { + "epoch": 0.27397669554825216, + "grad_norm": 0.15771432220935822, + "learning_rate": 0.00016544731405645154, + "loss": 0.7739, + "step": 1834 + }, + { + "epoch": 0.27412608305945624, + "grad_norm": 0.15347278118133545, + "learning_rate": 0.00016541176954741973, + "loss": 0.719, + "step": 1835 + }, + { + "epoch": 0.2742754705706603, + "grad_norm": 0.2687349319458008, + "learning_rate": 0.00016537621058789194, + "loss": 0.7605, + "step": 1836 + }, + { + "epoch": 0.27442485808186434, + "grad_norm": 0.1504988670349121, + "learning_rate": 0.0001653406371857237, + "loss": 0.933, + "step": 1837 + }, + { + "epoch": 0.2745742455930684, + "grad_norm": 0.11636245250701904, + "learning_rate": 0.00016530504934877377, + "loss": 0.4423, + "step": 1838 + }, + { + "epoch": 0.2747236331042725, + "grad_norm": 0.13434423506259918, + "learning_rate": 0.00016526944708490407, + "loss": 1.0305, + "step": 1839 + }, + { + "epoch": 0.2748730206154765, + "grad_norm": 0.16145826876163483, + "learning_rate": 0.0001652338304019797, + "loss": 0.4641, + "step": 1840 + }, + { + "epoch": 0.2750224081266806, + "grad_norm": 0.29437121748924255, + "learning_rate": 0.00016519819930786893, + "loss": 0.5978, + "step": 1841 + }, + { + "epoch": 0.2751717956378847, + "grad_norm": 0.1476840227842331, + "learning_rate": 0.00016516255381044323, + "loss": 0.8374, + "step": 1842 + }, + { + "epoch": 0.2753211831490887, + "grad_norm": 0.1587522178888321, + "learning_rate": 0.0001651268939175773, + "loss": 0.701, + "step": 1843 + }, + { + "epoch": 0.2754705706602928, + "grad_norm": 0.18530099093914032, + "learning_rate": 0.00016509121963714896, + "loss": 0.5378, + "step": 1844 + }, + { + "epoch": 0.27561995817149687, + "grad_norm": 0.17483104765415192, + "learning_rate": 0.00016505553097703916, + "loss": 0.7805, + "step": 1845 + }, + { + "epoch": 0.27576934568270095, + "grad_norm": 0.1965104341506958, + "learning_rate": 0.00016501982794513219, + "loss": 0.9346, + "step": 1846 + }, + { + "epoch": 0.275918733193905, + "grad_norm": 0.18628351390361786, + "learning_rate": 0.00016498411054931533, + "loss": 0.5019, + "step": 1847 + }, + { + "epoch": 0.27606812070510905, + "grad_norm": 0.139807790517807, + "learning_rate": 0.00016494837879747916, + "loss": 0.8478, + "step": 1848 + }, + { + "epoch": 0.27621750821631313, + "grad_norm": 0.14454811811447144, + "learning_rate": 0.0001649126326975174, + "loss": 0.7929, + "step": 1849 + }, + { + "epoch": 0.27636689572751716, + "grad_norm": 0.21324700117111206, + "learning_rate": 0.00016487687225732694, + "loss": 0.8371, + "step": 1850 + }, + { + "epoch": 0.27651628323872124, + "grad_norm": 0.24831005930900574, + "learning_rate": 0.00016484109748480776, + "loss": 0.9214, + "step": 1851 + }, + { + "epoch": 0.2766656707499253, + "grad_norm": 0.11538607627153397, + "learning_rate": 0.00016480530838786312, + "loss": 0.6869, + "step": 1852 + }, + { + "epoch": 0.27681505826112934, + "grad_norm": 0.166727215051651, + "learning_rate": 0.0001647695049743994, + "loss": 1.0586, + "step": 1853 + }, + { + "epoch": 0.2769644457723334, + "grad_norm": 0.15276376903057098, + "learning_rate": 0.00016473368725232614, + "loss": 0.7167, + "step": 1854 + }, + { + "epoch": 0.2771138332835375, + "grad_norm": 0.191617950797081, + "learning_rate": 0.00016469785522955604, + "loss": 0.6044, + "step": 1855 + }, + { + "epoch": 0.2772632207947416, + "grad_norm": 0.20602108538150787, + "learning_rate": 0.0001646620089140049, + "loss": 0.768, + "step": 1856 + }, + { + "epoch": 0.2774126083059456, + "grad_norm": 0.19726546108722687, + "learning_rate": 0.0001646261483135919, + "loss": 0.6565, + "step": 1857 + }, + { + "epoch": 0.2775619958171497, + "grad_norm": 0.16957935690879822, + "learning_rate": 0.00016459027343623906, + "loss": 0.5894, + "step": 1858 + }, + { + "epoch": 0.27771138332835377, + "grad_norm": 0.15182845294475555, + "learning_rate": 0.00016455438428987176, + "loss": 1.0849, + "step": 1859 + }, + { + "epoch": 0.2778607708395578, + "grad_norm": 0.11782501637935638, + "learning_rate": 0.00016451848088241847, + "loss": 0.8136, + "step": 1860 + }, + { + "epoch": 0.27801015835076187, + "grad_norm": 0.17523236572742462, + "learning_rate": 0.00016448256322181085, + "loss": 0.8055, + "step": 1861 + }, + { + "epoch": 0.27815954586196595, + "grad_norm": 0.1287536919116974, + "learning_rate": 0.00016444663131598365, + "loss": 0.687, + "step": 1862 + }, + { + "epoch": 0.27830893337317, + "grad_norm": 0.21639418601989746, + "learning_rate": 0.00016441068517287482, + "loss": 1.1522, + "step": 1863 + }, + { + "epoch": 0.27845832088437406, + "grad_norm": 0.13961181044578552, + "learning_rate": 0.00016437472480042544, + "loss": 0.8324, + "step": 1864 + }, + { + "epoch": 0.27860770839557814, + "grad_norm": 0.2107275277376175, + "learning_rate": 0.00016433875020657968, + "loss": 0.9217, + "step": 1865 + }, + { + "epoch": 0.2787570959067822, + "grad_norm": 0.1293714940547943, + "learning_rate": 0.00016430276139928494, + "loss": 0.7742, + "step": 1866 + }, + { + "epoch": 0.27890648341798624, + "grad_norm": 0.1160321906208992, + "learning_rate": 0.00016426675838649172, + "loss": 0.8035, + "step": 1867 + }, + { + "epoch": 0.2790558709291903, + "grad_norm": 0.12264332175254822, + "learning_rate": 0.00016423074117615362, + "loss": 0.7399, + "step": 1868 + }, + { + "epoch": 0.2792052584403944, + "grad_norm": 0.14644618332386017, + "learning_rate": 0.00016419470977622742, + "loss": 0.7063, + "step": 1869 + }, + { + "epoch": 0.2793546459515984, + "grad_norm": 0.15369118750095367, + "learning_rate": 0.00016415866419467308, + "loss": 0.9317, + "step": 1870 + }, + { + "epoch": 0.2795040334628025, + "grad_norm": 0.176469087600708, + "learning_rate": 0.00016412260443945357, + "loss": 0.9826, + "step": 1871 + }, + { + "epoch": 0.2796534209740066, + "grad_norm": 0.12978896498680115, + "learning_rate": 0.00016408653051853505, + "loss": 0.9042, + "step": 1872 + }, + { + "epoch": 0.27980280848521066, + "grad_norm": 0.1965731978416443, + "learning_rate": 0.00016405044243988695, + "loss": 0.6585, + "step": 1873 + }, + { + "epoch": 0.2799521959964147, + "grad_norm": 0.17228776216506958, + "learning_rate": 0.00016401434021148155, + "loss": 0.8369, + "step": 1874 + }, + { + "epoch": 0.28010158350761877, + "grad_norm": 0.2215079665184021, + "learning_rate": 0.00016397822384129444, + "loss": 1.0549, + "step": 1875 + }, + { + "epoch": 0.28025097101882285, + "grad_norm": 0.15248584747314453, + "learning_rate": 0.00016394209333730437, + "loss": 0.7295, + "step": 1876 + }, + { + "epoch": 0.28040035853002687, + "grad_norm": 0.11419791728258133, + "learning_rate": 0.00016390594870749307, + "loss": 0.5573, + "step": 1877 + }, + { + "epoch": 0.28054974604123095, + "grad_norm": 0.1532645970582962, + "learning_rate": 0.0001638697899598455, + "loss": 0.6374, + "step": 1878 + }, + { + "epoch": 0.28069913355243503, + "grad_norm": 0.27644631266593933, + "learning_rate": 0.00016383361710234966, + "loss": 1.0993, + "step": 1879 + }, + { + "epoch": 0.28084852106363906, + "grad_norm": 0.5468740463256836, + "learning_rate": 0.00016379743014299675, + "loss": 1.2103, + "step": 1880 + }, + { + "epoch": 0.28099790857484314, + "grad_norm": 0.2976016700267792, + "learning_rate": 0.000163761229089781, + "loss": 0.8506, + "step": 1881 + }, + { + "epoch": 0.2811472960860472, + "grad_norm": 0.3081592917442322, + "learning_rate": 0.00016372501395069984, + "loss": 0.8399, + "step": 1882 + }, + { + "epoch": 0.2812966835972513, + "grad_norm": 0.5960981249809265, + "learning_rate": 0.00016368878473375372, + "loss": 1.0627, + "step": 1883 + }, + { + "epoch": 0.2814460711084553, + "grad_norm": 0.16949574649333954, + "learning_rate": 0.0001636525414469463, + "loss": 0.6284, + "step": 1884 + }, + { + "epoch": 0.2815954586196594, + "grad_norm": 0.14496923983097076, + "learning_rate": 0.00016361628409828426, + "loss": 0.5671, + "step": 1885 + }, + { + "epoch": 0.2817448461308635, + "grad_norm": 0.1284521222114563, + "learning_rate": 0.00016358001269577743, + "loss": 0.7364, + "step": 1886 + }, + { + "epoch": 0.2818942336420675, + "grad_norm": 0.10659526288509369, + "learning_rate": 0.00016354372724743877, + "loss": 0.6766, + "step": 1887 + }, + { + "epoch": 0.2820436211532716, + "grad_norm": 0.13739757239818573, + "learning_rate": 0.00016350742776128423, + "loss": 0.8011, + "step": 1888 + }, + { + "epoch": 0.28219300866447566, + "grad_norm": 0.14763128757476807, + "learning_rate": 0.00016347111424533302, + "loss": 0.8036, + "step": 1889 + }, + { + "epoch": 0.2823423961756797, + "grad_norm": 0.14890700578689575, + "learning_rate": 0.00016343478670760732, + "loss": 0.9068, + "step": 1890 + }, + { + "epoch": 0.28249178368688377, + "grad_norm": 0.12007980793714523, + "learning_rate": 0.0001633984451561325, + "loss": 0.4907, + "step": 1891 + }, + { + "epoch": 0.28264117119808785, + "grad_norm": 0.258691668510437, + "learning_rate": 0.00016336208959893698, + "loss": 0.7515, + "step": 1892 + }, + { + "epoch": 0.28279055870929193, + "grad_norm": 0.21225577592849731, + "learning_rate": 0.00016332572004405224, + "loss": 0.6392, + "step": 1893 + }, + { + "epoch": 0.28293994622049595, + "grad_norm": 0.15706035494804382, + "learning_rate": 0.00016328933649951293, + "loss": 0.7764, + "step": 1894 + }, + { + "epoch": 0.28308933373170003, + "grad_norm": 0.8196190595626831, + "learning_rate": 0.0001632529389733567, + "loss": 1.6892, + "step": 1895 + }, + { + "epoch": 0.2832387212429041, + "grad_norm": 0.1771058291196823, + "learning_rate": 0.00016321652747362445, + "loss": 1.2452, + "step": 1896 + }, + { + "epoch": 0.28338810875410814, + "grad_norm": 0.17324721813201904, + "learning_rate": 0.00016318010200835992, + "loss": 0.3363, + "step": 1897 + }, + { + "epoch": 0.2835374962653122, + "grad_norm": 0.14077776670455933, + "learning_rate": 0.00016314366258561016, + "loss": 0.495, + "step": 1898 + }, + { + "epoch": 0.2836868837765163, + "grad_norm": 0.1295582801103592, + "learning_rate": 0.00016310720921342518, + "loss": 0.7699, + "step": 1899 + }, + { + "epoch": 0.2838362712877203, + "grad_norm": 0.18893714249134064, + "learning_rate": 0.00016307074189985814, + "loss": 0.9228, + "step": 1900 + }, + { + "epoch": 0.2839856587989244, + "grad_norm": 0.18363916873931885, + "learning_rate": 0.00016303426065296522, + "loss": 0.557, + "step": 1901 + }, + { + "epoch": 0.2841350463101285, + "grad_norm": 0.156977578997612, + "learning_rate": 0.0001629977654808057, + "loss": 0.7707, + "step": 1902 + }, + { + "epoch": 0.28428443382133256, + "grad_norm": 0.23877717554569244, + "learning_rate": 0.00016296125639144192, + "loss": 0.8239, + "step": 1903 + }, + { + "epoch": 0.2844338213325366, + "grad_norm": 0.13115395605564117, + "learning_rate": 0.0001629247333929394, + "loss": 0.7643, + "step": 1904 + }, + { + "epoch": 0.28458320884374066, + "grad_norm": 0.12260690331459045, + "learning_rate": 0.00016288819649336656, + "loss": 0.6755, + "step": 1905 + }, + { + "epoch": 0.28473259635494474, + "grad_norm": 0.1967897266149521, + "learning_rate": 0.00016285164570079504, + "loss": 0.5642, + "step": 1906 + }, + { + "epoch": 0.28488198386614877, + "grad_norm": 0.263201504945755, + "learning_rate": 0.00016281508102329947, + "loss": 0.9733, + "step": 1907 + }, + { + "epoch": 0.28503137137735285, + "grad_norm": 0.18227405846118927, + "learning_rate": 0.00016277850246895753, + "loss": 0.8968, + "step": 1908 + }, + { + "epoch": 0.28518075888855693, + "grad_norm": 0.3462936282157898, + "learning_rate": 0.00016274191004585009, + "loss": 0.9113, + "step": 1909 + }, + { + "epoch": 0.28533014639976095, + "grad_norm": 0.13814112544059753, + "learning_rate": 0.0001627053037620609, + "loss": 0.8253, + "step": 1910 + }, + { + "epoch": 0.28547953391096503, + "grad_norm": 0.27378007769584656, + "learning_rate": 0.0001626686836256769, + "loss": 0.7296, + "step": 1911 + }, + { + "epoch": 0.2856289214221691, + "grad_norm": 0.16056250035762787, + "learning_rate": 0.00016263204964478807, + "loss": 0.9768, + "step": 1912 + }, + { + "epoch": 0.2857783089333732, + "grad_norm": 0.28953564167022705, + "learning_rate": 0.00016259540182748746, + "loss": 0.5664, + "step": 1913 + }, + { + "epoch": 0.2859276964445772, + "grad_norm": 0.22288711369037628, + "learning_rate": 0.00016255874018187113, + "loss": 0.9053, + "step": 1914 + }, + { + "epoch": 0.2860770839557813, + "grad_norm": 0.153893381357193, + "learning_rate": 0.00016252206471603818, + "loss": 0.8007, + "step": 1915 + }, + { + "epoch": 0.2862264714669854, + "grad_norm": 0.2964473068714142, + "learning_rate": 0.00016248537543809085, + "loss": 0.6401, + "step": 1916 + }, + { + "epoch": 0.2863758589781894, + "grad_norm": 0.16791103780269623, + "learning_rate": 0.00016244867235613438, + "loss": 0.8178, + "step": 1917 + }, + { + "epoch": 0.2865252464893935, + "grad_norm": 0.3182014524936676, + "learning_rate": 0.00016241195547827704, + "loss": 0.96, + "step": 1918 + }, + { + "epoch": 0.28667463400059756, + "grad_norm": 0.15290340781211853, + "learning_rate": 0.00016237522481263018, + "loss": 0.9855, + "step": 1919 + }, + { + "epoch": 0.2868240215118016, + "grad_norm": 0.36779022216796875, + "learning_rate": 0.00016233848036730818, + "loss": 1.0061, + "step": 1920 + }, + { + "epoch": 0.28697340902300567, + "grad_norm": 0.2556033432483673, + "learning_rate": 0.0001623017221504285, + "loss": 0.7636, + "step": 1921 + }, + { + "epoch": 0.28712279653420975, + "grad_norm": 0.19862814247608185, + "learning_rate": 0.00016226495017011155, + "loss": 0.8552, + "step": 1922 + }, + { + "epoch": 0.2872721840454138, + "grad_norm": 0.18107233941555023, + "learning_rate": 0.00016222816443448092, + "loss": 0.907, + "step": 1923 + }, + { + "epoch": 0.28742157155661785, + "grad_norm": 0.33394572138786316, + "learning_rate": 0.0001621913649516631, + "loss": 1.0993, + "step": 1924 + }, + { + "epoch": 0.28757095906782193, + "grad_norm": 0.18873590230941772, + "learning_rate": 0.00016215455172978772, + "loss": 1.1411, + "step": 1925 + }, + { + "epoch": 0.287720346579026, + "grad_norm": 0.17854920029640198, + "learning_rate": 0.00016211772477698737, + "loss": 1.1384, + "step": 1926 + }, + { + "epoch": 0.28786973409023003, + "grad_norm": 0.1506814956665039, + "learning_rate": 0.00016208088410139772, + "loss": 1.0847, + "step": 1927 + }, + { + "epoch": 0.2880191216014341, + "grad_norm": 0.11812624335289001, + "learning_rate": 0.0001620440297111575, + "loss": 0.7856, + "step": 1928 + }, + { + "epoch": 0.2881685091126382, + "grad_norm": 0.36629819869995117, + "learning_rate": 0.00016200716161440833, + "loss": 1.1171, + "step": 1929 + }, + { + "epoch": 0.2883178966238423, + "grad_norm": 0.12690721452236176, + "learning_rate": 0.00016197027981929506, + "loss": 0.8426, + "step": 1930 + }, + { + "epoch": 0.2884672841350463, + "grad_norm": 0.17128652334213257, + "learning_rate": 0.00016193338433396538, + "loss": 0.7219, + "step": 1931 + }, + { + "epoch": 0.2886166716462504, + "grad_norm": 0.19333554804325104, + "learning_rate": 0.00016189647516657018, + "loss": 0.4354, + "step": 1932 + }, + { + "epoch": 0.28876605915745446, + "grad_norm": 0.1483379602432251, + "learning_rate": 0.0001618595523252632, + "loss": 0.5917, + "step": 1933 + }, + { + "epoch": 0.2889154466686585, + "grad_norm": 0.18341517448425293, + "learning_rate": 0.0001618226158182013, + "loss": 0.5832, + "step": 1934 + }, + { + "epoch": 0.28906483417986256, + "grad_norm": 0.13665004074573517, + "learning_rate": 0.00016178566565354437, + "loss": 1.0354, + "step": 1935 + }, + { + "epoch": 0.28921422169106664, + "grad_norm": 0.19725897908210754, + "learning_rate": 0.00016174870183945523, + "loss": 0.8194, + "step": 1936 + }, + { + "epoch": 0.28936360920227067, + "grad_norm": 0.1692163348197937, + "learning_rate": 0.0001617117243840998, + "loss": 0.9514, + "step": 1937 + }, + { + "epoch": 0.28951299671347475, + "grad_norm": 0.22694113850593567, + "learning_rate": 0.00016167473329564705, + "loss": 0.6955, + "step": 1938 + }, + { + "epoch": 0.2896623842246788, + "grad_norm": 0.16074326634407043, + "learning_rate": 0.00016163772858226877, + "loss": 0.8205, + "step": 1939 + }, + { + "epoch": 0.2898117717358829, + "grad_norm": 0.1290196180343628, + "learning_rate": 0.00016160071025213998, + "loss": 0.76, + "step": 1940 + }, + { + "epoch": 0.28996115924708693, + "grad_norm": 0.12540550529956818, + "learning_rate": 0.0001615636783134386, + "loss": 0.6737, + "step": 1941 + }, + { + "epoch": 0.290110546758291, + "grad_norm": 0.1340838223695755, + "learning_rate": 0.00016152663277434556, + "loss": 0.9725, + "step": 1942 + }, + { + "epoch": 0.2902599342694951, + "grad_norm": 0.14680004119873047, + "learning_rate": 0.0001614895736430448, + "loss": 0.7528, + "step": 1943 + }, + { + "epoch": 0.2904093217806991, + "grad_norm": 0.2111632525920868, + "learning_rate": 0.0001614525009277233, + "loss": 0.9577, + "step": 1944 + }, + { + "epoch": 0.2905587092919032, + "grad_norm": 0.16464073956012726, + "learning_rate": 0.00016141541463657094, + "loss": 0.7268, + "step": 1945 + }, + { + "epoch": 0.2907080968031073, + "grad_norm": 0.13918276131153107, + "learning_rate": 0.00016137831477778077, + "loss": 0.7131, + "step": 1946 + }, + { + "epoch": 0.2908574843143113, + "grad_norm": 0.18370576202869415, + "learning_rate": 0.00016134120135954865, + "loss": 0.5412, + "step": 1947 + }, + { + "epoch": 0.2910068718255154, + "grad_norm": 0.17753422260284424, + "learning_rate": 0.00016130407439007355, + "loss": 0.9692, + "step": 1948 + }, + { + "epoch": 0.29115625933671946, + "grad_norm": 0.22409065067768097, + "learning_rate": 0.00016126693387755742, + "loss": 0.7073, + "step": 1949 + }, + { + "epoch": 0.29130564684792354, + "grad_norm": 0.11738231033086777, + "learning_rate": 0.0001612297798302052, + "loss": 0.6564, + "step": 1950 + }, + { + "epoch": 0.29145503435912756, + "grad_norm": 0.13526974618434906, + "learning_rate": 0.0001611926122562248, + "loss": 0.7621, + "step": 1951 + }, + { + "epoch": 0.29160442187033164, + "grad_norm": 0.14819179475307465, + "learning_rate": 0.00016115543116382707, + "loss": 0.7947, + "step": 1952 + }, + { + "epoch": 0.2917538093815357, + "grad_norm": 0.1239335685968399, + "learning_rate": 0.00016111823656122596, + "loss": 0.7881, + "step": 1953 + }, + { + "epoch": 0.29190319689273975, + "grad_norm": 0.1371903270483017, + "learning_rate": 0.00016108102845663832, + "loss": 0.7273, + "step": 1954 + }, + { + "epoch": 0.2920525844039438, + "grad_norm": 0.2668689489364624, + "learning_rate": 0.00016104380685828403, + "loss": 0.8244, + "step": 1955 + }, + { + "epoch": 0.2922019719151479, + "grad_norm": 0.27288052439689636, + "learning_rate": 0.00016100657177438592, + "loss": 1.1713, + "step": 1956 + }, + { + "epoch": 0.29235135942635193, + "grad_norm": 0.12943297624588013, + "learning_rate": 0.0001609693232131698, + "loss": 0.5734, + "step": 1957 + }, + { + "epoch": 0.292500746937556, + "grad_norm": 0.1556374877691269, + "learning_rate": 0.0001609320611828645, + "loss": 0.7606, + "step": 1958 + }, + { + "epoch": 0.2926501344487601, + "grad_norm": 0.1746862679719925, + "learning_rate": 0.00016089478569170174, + "loss": 0.7793, + "step": 1959 + }, + { + "epoch": 0.29279952195996417, + "grad_norm": 0.18083463609218597, + "learning_rate": 0.0001608574967479163, + "loss": 0.5813, + "step": 1960 + }, + { + "epoch": 0.2929489094711682, + "grad_norm": 0.25068581104278564, + "learning_rate": 0.00016082019435974593, + "loss": 1.1797, + "step": 1961 + }, + { + "epoch": 0.2930982969823723, + "grad_norm": 0.1725584864616394, + "learning_rate": 0.00016078287853543125, + "loss": 0.7816, + "step": 1962 + }, + { + "epoch": 0.29324768449357635, + "grad_norm": 0.19390223920345306, + "learning_rate": 0.00016074554928321598, + "loss": 0.5733, + "step": 1963 + }, + { + "epoch": 0.2933970720047804, + "grad_norm": 0.1887221783399582, + "learning_rate": 0.00016070820661134668, + "loss": 0.78, + "step": 1964 + }, + { + "epoch": 0.29354645951598446, + "grad_norm": 0.1666865348815918, + "learning_rate": 0.000160670850528073, + "loss": 0.7833, + "step": 1965 + }, + { + "epoch": 0.29369584702718854, + "grad_norm": 0.13192425668239594, + "learning_rate": 0.00016063348104164744, + "loss": 0.706, + "step": 1966 + }, + { + "epoch": 0.29384523453839256, + "grad_norm": 0.17119993269443512, + "learning_rate": 0.00016059609816032556, + "loss": 0.7467, + "step": 1967 + }, + { + "epoch": 0.29399462204959664, + "grad_norm": 0.12513847649097443, + "learning_rate": 0.00016055870189236578, + "loss": 0.6887, + "step": 1968 + }, + { + "epoch": 0.2941440095608007, + "grad_norm": 0.13904425501823425, + "learning_rate": 0.00016052129224602957, + "loss": 0.661, + "step": 1969 + }, + { + "epoch": 0.2942933970720048, + "grad_norm": 0.12954014539718628, + "learning_rate": 0.00016048386922958127, + "loss": 0.774, + "step": 1970 + }, + { + "epoch": 0.2944427845832088, + "grad_norm": 0.17040160298347473, + "learning_rate": 0.00016044643285128827, + "loss": 0.8955, + "step": 1971 + }, + { + "epoch": 0.2945921720944129, + "grad_norm": 0.3505501449108124, + "learning_rate": 0.00016040898311942082, + "loss": 0.6914, + "step": 1972 + }, + { + "epoch": 0.294741559605617, + "grad_norm": 0.14326271414756775, + "learning_rate": 0.00016037152004225215, + "loss": 0.6905, + "step": 1973 + }, + { + "epoch": 0.294890947116821, + "grad_norm": 0.12812425196170807, + "learning_rate": 0.0001603340436280585, + "loss": 0.7456, + "step": 1974 + }, + { + "epoch": 0.2950403346280251, + "grad_norm": 0.20053675770759583, + "learning_rate": 0.00016029655388511896, + "loss": 0.6051, + "step": 1975 + }, + { + "epoch": 0.29518972213922917, + "grad_norm": 0.1312316656112671, + "learning_rate": 0.00016025905082171562, + "loss": 0.6035, + "step": 1976 + }, + { + "epoch": 0.29533910965043325, + "grad_norm": 0.14165259897708893, + "learning_rate": 0.00016022153444613348, + "loss": 0.8824, + "step": 1977 + }, + { + "epoch": 0.2954884971616373, + "grad_norm": 0.17321546375751495, + "learning_rate": 0.00016018400476666055, + "loss": 0.8771, + "step": 1978 + }, + { + "epoch": 0.29563788467284136, + "grad_norm": 0.11130014806985855, + "learning_rate": 0.00016014646179158767, + "loss": 0.6402, + "step": 1979 + }, + { + "epoch": 0.29578727218404544, + "grad_norm": 0.22360561788082123, + "learning_rate": 0.00016010890552920875, + "loss": 0.4773, + "step": 1980 + }, + { + "epoch": 0.29593665969524946, + "grad_norm": 0.1546655297279358, + "learning_rate": 0.00016007133598782048, + "loss": 1.1104, + "step": 1981 + }, + { + "epoch": 0.29608604720645354, + "grad_norm": 0.3024921417236328, + "learning_rate": 0.00016003375317572263, + "loss": 0.6703, + "step": 1982 + }, + { + "epoch": 0.2962354347176576, + "grad_norm": 0.4379715919494629, + "learning_rate": 0.00015999615710121784, + "loss": 1.0485, + "step": 1983 + }, + { + "epoch": 0.29638482222886164, + "grad_norm": 0.1305762678384781, + "learning_rate": 0.00015995854777261161, + "loss": 0.9361, + "step": 1984 + }, + { + "epoch": 0.2965342097400657, + "grad_norm": 0.12884147465229034, + "learning_rate": 0.0001599209251982125, + "loss": 0.9236, + "step": 1985 + }, + { + "epoch": 0.2966835972512698, + "grad_norm": 0.15657733380794525, + "learning_rate": 0.00015988328938633191, + "loss": 0.9226, + "step": 1986 + }, + { + "epoch": 0.2968329847624739, + "grad_norm": 0.15643414855003357, + "learning_rate": 0.00015984564034528422, + "loss": 0.7931, + "step": 1987 + }, + { + "epoch": 0.2969823722736779, + "grad_norm": 0.15294334292411804, + "learning_rate": 0.00015980797808338664, + "loss": 0.9234, + "step": 1988 + }, + { + "epoch": 0.297131759784882, + "grad_norm": 0.17562635242938995, + "learning_rate": 0.00015977030260895943, + "loss": 1.0773, + "step": 1989 + }, + { + "epoch": 0.29728114729608607, + "grad_norm": 0.15384536981582642, + "learning_rate": 0.00015973261393032563, + "loss": 0.7348, + "step": 1990 + }, + { + "epoch": 0.2974305348072901, + "grad_norm": 0.1446111500263214, + "learning_rate": 0.00015969491205581132, + "loss": 0.6682, + "step": 1991 + }, + { + "epoch": 0.29757992231849417, + "grad_norm": 0.12944035232067108, + "learning_rate": 0.0001596571969937454, + "loss": 0.8242, + "step": 1992 + }, + { + "epoch": 0.29772930982969825, + "grad_norm": 0.15892425179481506, + "learning_rate": 0.00015961946875245973, + "loss": 0.6055, + "step": 1993 + }, + { + "epoch": 0.2978786973409023, + "grad_norm": 0.13054129481315613, + "learning_rate": 0.0001595817273402891, + "loss": 0.5977, + "step": 1994 + }, + { + "epoch": 0.29802808485210636, + "grad_norm": 0.2998169958591461, + "learning_rate": 0.0001595439727655712, + "loss": 1.0095, + "step": 1995 + }, + { + "epoch": 0.29817747236331044, + "grad_norm": 0.11575822532176971, + "learning_rate": 0.00015950620503664658, + "loss": 0.831, + "step": 1996 + }, + { + "epoch": 0.2983268598745145, + "grad_norm": 0.176680326461792, + "learning_rate": 0.0001594684241618587, + "loss": 0.8144, + "step": 1997 + }, + { + "epoch": 0.29847624738571854, + "grad_norm": 0.2667008936405182, + "learning_rate": 0.00015943063014955402, + "loss": 1.1887, + "step": 1998 + }, + { + "epoch": 0.2986256348969226, + "grad_norm": 0.2525520324707031, + "learning_rate": 0.0001593928230080818, + "loss": 0.7398, + "step": 1999 + }, + { + "epoch": 0.2987750224081267, + "grad_norm": 0.20424726605415344, + "learning_rate": 0.00015935500274579426, + "loss": 0.9998, + "step": 2000 + }, + { + "epoch": 0.2989244099193307, + "grad_norm": 0.24848030507564545, + "learning_rate": 0.00015931716937104645, + "loss": 0.9428, + "step": 2001 + }, + { + "epoch": 0.2990737974305348, + "grad_norm": 0.1705709546804428, + "learning_rate": 0.00015927932289219642, + "loss": 0.7016, + "step": 2002 + }, + { + "epoch": 0.2992231849417389, + "grad_norm": 0.20898346602916718, + "learning_rate": 0.000159241463317605, + "loss": 1.2988, + "step": 2003 + }, + { + "epoch": 0.2993725724529429, + "grad_norm": 0.1826796531677246, + "learning_rate": 0.00015920359065563604, + "loss": 0.8184, + "step": 2004 + }, + { + "epoch": 0.299521959964147, + "grad_norm": 0.12231456488370895, + "learning_rate": 0.00015916570491465612, + "loss": 0.8552, + "step": 2005 + }, + { + "epoch": 0.29967134747535107, + "grad_norm": 0.10954627394676208, + "learning_rate": 0.0001591278061030349, + "loss": 0.5853, + "step": 2006 + }, + { + "epoch": 0.29982073498655515, + "grad_norm": 0.17874775826931, + "learning_rate": 0.00015908989422914473, + "loss": 0.7906, + "step": 2007 + }, + { + "epoch": 0.2999701224977592, + "grad_norm": 0.12277006357908249, + "learning_rate": 0.00015905196930136097, + "loss": 0.794, + "step": 2008 + }, + { + "epoch": 0.30011951000896325, + "grad_norm": 0.12225458770990372, + "learning_rate": 0.0001590140313280619, + "loss": 0.6106, + "step": 2009 + }, + { + "epoch": 0.30026889752016733, + "grad_norm": 0.16903723776340485, + "learning_rate": 0.0001589760803176286, + "loss": 0.6282, + "step": 2010 + } + ], + "logging_steps": 1, + "max_steps": 6694, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 670, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 1.587795414881403e+18, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}