| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.1859292848953115, | |
| "eval_steps": 375, | |
| "global_step": 1125, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0001652704754624991, | |
| "grad_norm": 1.5024014711380005, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4778, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0001652704754624991, | |
| "eval_loss": 2.897561550140381, | |
| "eval_runtime": 209.0331, | |
| "eval_samples_per_second": 48.753, | |
| "eval_steps_per_second": 24.379, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0003305409509249982, | |
| "grad_norm": 0.955745279788971, | |
| "learning_rate": 4e-05, | |
| "loss": 1.388, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0004958114263874973, | |
| "grad_norm": 0.9341157674789429, | |
| "learning_rate": 6e-05, | |
| "loss": 1.5937, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0006610819018499964, | |
| "grad_norm": 0.7856804132461548, | |
| "learning_rate": 8e-05, | |
| "loss": 2.0326, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0008263523773124955, | |
| "grad_norm": 0.7618529200553894, | |
| "learning_rate": 0.0001, | |
| "loss": 2.5266, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0009916228527749947, | |
| "grad_norm": 0.9779232740402222, | |
| "learning_rate": 0.00012, | |
| "loss": 2.6918, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0011568933282374936, | |
| "grad_norm": 1.0010966062545776, | |
| "learning_rate": 0.00014, | |
| "loss": 2.2979, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0013221638036999927, | |
| "grad_norm": 0.978973925113678, | |
| "learning_rate": 0.00016, | |
| "loss": 2.3689, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0014874342791624919, | |
| "grad_norm": 1.0895302295684814, | |
| "learning_rate": 0.00018, | |
| "loss": 2.7272, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.001652704754624991, | |
| "grad_norm": 1.006224274635315, | |
| "learning_rate": 0.0002, | |
| "loss": 2.3388, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.00181797523008749, | |
| "grad_norm": 0.9786027669906616, | |
| "learning_rate": 0.00019999977772170748, | |
| "loss": 2.3947, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0019832457055499893, | |
| "grad_norm": 1.134806752204895, | |
| "learning_rate": 0.00019999911088781805, | |
| "loss": 2.2153, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.002148516181012488, | |
| "grad_norm": 1.2217847108840942, | |
| "learning_rate": 0.0001999979995012962, | |
| "loss": 2.0339, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.002313786656474987, | |
| "grad_norm": 1.0734370946884155, | |
| "learning_rate": 0.00019999644356708261, | |
| "loss": 2.3898, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0024790571319374863, | |
| "grad_norm": 1.1617718935012817, | |
| "learning_rate": 0.00019999444309209432, | |
| "loss": 2.0164, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0026443276073999855, | |
| "grad_norm": 1.2858864068984985, | |
| "learning_rate": 0.0001999919980852246, | |
| "loss": 2.3209, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0028095980828624846, | |
| "grad_norm": 1.2736631631851196, | |
| "learning_rate": 0.00019998910855734288, | |
| "loss": 2.2074, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0029748685583249838, | |
| "grad_norm": 1.5206106901168823, | |
| "learning_rate": 0.0001999857745212947, | |
| "loss": 2.245, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.003140139033787483, | |
| "grad_norm": 1.4323664903640747, | |
| "learning_rate": 0.00019998199599190178, | |
| "loss": 2.5118, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.003305409509249982, | |
| "grad_norm": 1.5932579040527344, | |
| "learning_rate": 0.0001999777729859618, | |
| "loss": 2.3128, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.003470679984712481, | |
| "grad_norm": 2.1362950801849365, | |
| "learning_rate": 0.00019997310552224846, | |
| "loss": 2.66, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.00363595046017498, | |
| "grad_norm": 1.672585129737854, | |
| "learning_rate": 0.00019996799362151122, | |
| "loss": 2.1904, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.003801220935637479, | |
| "grad_norm": 1.7134736776351929, | |
| "learning_rate": 0.00019996243730647538, | |
| "loss": 2.7138, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.003966491411099979, | |
| "grad_norm": 1.8876935243606567, | |
| "learning_rate": 0.00019995643660184191, | |
| "loss": 1.9226, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.004131761886562477, | |
| "grad_norm": 1.9208993911743164, | |
| "learning_rate": 0.00019994999153428737, | |
| "loss": 2.5529, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.004297032362024976, | |
| "grad_norm": 2.132223129272461, | |
| "learning_rate": 0.00019994310213246368, | |
| "loss": 2.7591, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.004462302837487476, | |
| "grad_norm": 2.1397337913513184, | |
| "learning_rate": 0.00019993576842699816, | |
| "loss": 2.9653, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.004627573312949974, | |
| "grad_norm": 2.2892026901245117, | |
| "learning_rate": 0.0001999279904504933, | |
| "loss": 2.633, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.004792843788412474, | |
| "grad_norm": 2.5149452686309814, | |
| "learning_rate": 0.00019991976823752653, | |
| "loss": 3.079, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.004958114263874973, | |
| "grad_norm": 2.5469651222229004, | |
| "learning_rate": 0.00019991110182465032, | |
| "loss": 3.0389, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.005123384739337472, | |
| "grad_norm": 2.6004953384399414, | |
| "learning_rate": 0.00019990199125039174, | |
| "loss": 2.7452, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.005288655214799971, | |
| "grad_norm": 2.519162893295288, | |
| "learning_rate": 0.00019989243655525247, | |
| "loss": 2.9512, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.0054539256902624705, | |
| "grad_norm": 2.892303943634033, | |
| "learning_rate": 0.00019988243778170853, | |
| "loss": 3.2904, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.005619196165724969, | |
| "grad_norm": 2.8910715579986572, | |
| "learning_rate": 0.0001998719949742101, | |
| "loss": 3.0635, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.005784466641187468, | |
| "grad_norm": 3.138007402420044, | |
| "learning_rate": 0.0001998611081791814, | |
| "loss": 2.8122, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0059497371166499675, | |
| "grad_norm": 3.376237630844116, | |
| "learning_rate": 0.00019984977744502038, | |
| "loss": 2.7496, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.006115007592112466, | |
| "grad_norm": 2.9936282634735107, | |
| "learning_rate": 0.00019983800282209857, | |
| "loss": 2.6468, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.006280278067574966, | |
| "grad_norm": 3.166961908340454, | |
| "learning_rate": 0.00019982578436276082, | |
| "loss": 2.5579, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0064455485430374646, | |
| "grad_norm": 3.5384085178375244, | |
| "learning_rate": 0.00019981312212132512, | |
| "loss": 2.8784, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.006610819018499964, | |
| "grad_norm": 3.787200689315796, | |
| "learning_rate": 0.00019980001615408228, | |
| "loss": 2.4643, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.006776089493962463, | |
| "grad_norm": 3.600571393966675, | |
| "learning_rate": 0.00019978646651929572, | |
| "loss": 2.995, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.006941359969424962, | |
| "grad_norm": 3.7038369178771973, | |
| "learning_rate": 0.00019977247327720128, | |
| "loss": 2.718, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.007106630444887461, | |
| "grad_norm": 4.449253559112549, | |
| "learning_rate": 0.0001997580364900068, | |
| "loss": 2.7885, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.00727190092034996, | |
| "grad_norm": 5.158216953277588, | |
| "learning_rate": 0.000199743156221892, | |
| "loss": 3.0556, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.007437171395812459, | |
| "grad_norm": 5.8202595710754395, | |
| "learning_rate": 0.00019972783253900808, | |
| "loss": 3.0437, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.007602441871274958, | |
| "grad_norm": 5.959027290344238, | |
| "learning_rate": 0.00019971206550947748, | |
| "loss": 2.6739, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.007767712346737458, | |
| "grad_norm": 5.996998310089111, | |
| "learning_rate": 0.00019969585520339354, | |
| "loss": 3.0094, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.007932982822199957, | |
| "grad_norm": 7.538012981414795, | |
| "learning_rate": 0.0001996792016928203, | |
| "loss": 2.6221, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.008098253297662455, | |
| "grad_norm": 8.595839500427246, | |
| "learning_rate": 0.00019966210505179197, | |
| "loss": 3.498, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.008263523773124955, | |
| "grad_norm": 10.21332836151123, | |
| "learning_rate": 0.00019964456535631286, | |
| "loss": 3.8247, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.008428794248587454, | |
| "grad_norm": 22.48410415649414, | |
| "learning_rate": 0.0001996265826843568, | |
| "loss": 2.2788, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.008594064724049952, | |
| "grad_norm": 14.210624694824219, | |
| "learning_rate": 0.00019960815711586696, | |
| "loss": 2.0172, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.008759335199512452, | |
| "grad_norm": 7.03934383392334, | |
| "learning_rate": 0.00019958928873275539, | |
| "loss": 1.9204, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.008924605674974951, | |
| "grad_norm": 2.600945472717285, | |
| "learning_rate": 0.00019956997761890277, | |
| "loss": 1.9673, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.009089876150437451, | |
| "grad_norm": 1.5479152202606201, | |
| "learning_rate": 0.00019955022386015792, | |
| "loss": 2.5609, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.009255146625899949, | |
| "grad_norm": 1.14052414894104, | |
| "learning_rate": 0.00019953002754433743, | |
| "loss": 2.8202, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.009420417101362448, | |
| "grad_norm": 0.8741720914840698, | |
| "learning_rate": 0.00019950938876122542, | |
| "loss": 2.0929, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.009585687576824948, | |
| "grad_norm": 0.9342317581176758, | |
| "learning_rate": 0.00019948830760257291, | |
| "loss": 2.2272, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.009750958052287447, | |
| "grad_norm": 0.8068450689315796, | |
| "learning_rate": 0.0001994667841620976, | |
| "loss": 2.2732, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.009916228527749945, | |
| "grad_norm": 0.9862707257270813, | |
| "learning_rate": 0.00019944481853548335, | |
| "loss": 2.2772, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.010081499003212445, | |
| "grad_norm": 0.8683833479881287, | |
| "learning_rate": 0.00019942241082037982, | |
| "loss": 2.0926, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.010246769478674944, | |
| "grad_norm": 1.0305839776992798, | |
| "learning_rate": 0.00019939956111640197, | |
| "loss": 2.3179, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.010412039954137442, | |
| "grad_norm": 0.9628840684890747, | |
| "learning_rate": 0.00019937626952512964, | |
| "loss": 2.6018, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.010577310429599942, | |
| "grad_norm": 1.0319561958312988, | |
| "learning_rate": 0.0001993525361501072, | |
| "loss": 2.4756, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.010742580905062441, | |
| "grad_norm": 1.0028680562973022, | |
| "learning_rate": 0.00019932836109684286, | |
| "loss": 2.1883, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.010907851380524941, | |
| "grad_norm": 1.1696614027023315, | |
| "learning_rate": 0.00019930374447280845, | |
| "loss": 2.5459, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.011073121855987439, | |
| "grad_norm": 1.0888032913208008, | |
| "learning_rate": 0.00019927868638743875, | |
| "loss": 2.1314, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.011238392331449939, | |
| "grad_norm": 1.256794810295105, | |
| "learning_rate": 0.0001992531869521312, | |
| "loss": 2.2889, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.011403662806912438, | |
| "grad_norm": 1.3103197813034058, | |
| "learning_rate": 0.00019922724628024515, | |
| "loss": 1.9291, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.011568933282374936, | |
| "grad_norm": 1.2497581243515015, | |
| "learning_rate": 0.0001992008644871016, | |
| "loss": 2.12, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.011734203757837436, | |
| "grad_norm": 1.4419199228286743, | |
| "learning_rate": 0.00019917404168998256, | |
| "loss": 2.4166, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.011899474233299935, | |
| "grad_norm": 1.706443428993225, | |
| "learning_rate": 0.0001991467780081305, | |
| "loss": 2.5804, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.012064744708762435, | |
| "grad_norm": 1.7904235124588013, | |
| "learning_rate": 0.00019911907356274795, | |
| "loss": 2.3902, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.012230015184224933, | |
| "grad_norm": 1.8568871021270752, | |
| "learning_rate": 0.00019909092847699683, | |
| "loss": 2.5723, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.012395285659687432, | |
| "grad_norm": 2.180750608444214, | |
| "learning_rate": 0.00019906234287599798, | |
| "loss": 3.2458, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.012560556135149932, | |
| "grad_norm": 1.7413932085037231, | |
| "learning_rate": 0.00019903331688683057, | |
| "loss": 2.3947, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.012725826610612431, | |
| "grad_norm": 2.1169497966766357, | |
| "learning_rate": 0.00019900385063853154, | |
| "loss": 2.7057, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.012891097086074929, | |
| "grad_norm": 2.0730772018432617, | |
| "learning_rate": 0.00019897394426209505, | |
| "loss": 2.9583, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.013056367561537429, | |
| "grad_norm": 2.10861873626709, | |
| "learning_rate": 0.00019894359789047187, | |
| "loss": 3.047, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.013221638036999928, | |
| "grad_norm": 1.964033842086792, | |
| "learning_rate": 0.00019891281165856873, | |
| "loss": 2.6404, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.013386908512462426, | |
| "grad_norm": 2.1377522945404053, | |
| "learning_rate": 0.00019888158570324795, | |
| "loss": 2.7513, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.013552178987924926, | |
| "grad_norm": 2.5152194499969482, | |
| "learning_rate": 0.0001988499201633265, | |
| "loss": 2.9624, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.013717449463387425, | |
| "grad_norm": 2.659669876098633, | |
| "learning_rate": 0.00019881781517957562, | |
| "loss": 2.9105, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.013882719938849925, | |
| "grad_norm": 2.4301917552948, | |
| "learning_rate": 0.0001987852708947202, | |
| "loss": 2.8207, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.014047990414312423, | |
| "grad_norm": 2.968515396118164, | |
| "learning_rate": 0.00019875228745343794, | |
| "loss": 2.6456, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.014213260889774922, | |
| "grad_norm": 3.1119697093963623, | |
| "learning_rate": 0.0001987188650023589, | |
| "loss": 2.8118, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.014378531365237422, | |
| "grad_norm": 3.010894775390625, | |
| "learning_rate": 0.0001986850036900648, | |
| "loss": 2.3691, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.01454380184069992, | |
| "grad_norm": 2.985806941986084, | |
| "learning_rate": 0.00019865070366708836, | |
| "loss": 2.5471, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.01470907231616242, | |
| "grad_norm": 3.533379077911377, | |
| "learning_rate": 0.00019861596508591255, | |
| "loss": 2.7374, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.014874342791624919, | |
| "grad_norm": 3.4825985431671143, | |
| "learning_rate": 0.00019858078810097002, | |
| "loss": 2.6087, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.015039613267087418, | |
| "grad_norm": 3.1555826663970947, | |
| "learning_rate": 0.00019854517286864245, | |
| "loss": 2.4697, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.015204883742549916, | |
| "grad_norm": 4.647697448730469, | |
| "learning_rate": 0.0001985091195472596, | |
| "loss": 2.8638, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.015370154218012416, | |
| "grad_norm": 3.671441078186035, | |
| "learning_rate": 0.0001984726282970989, | |
| "loss": 2.7693, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.015535424693474915, | |
| "grad_norm": 4.510019302368164, | |
| "learning_rate": 0.0001984356992803847, | |
| "loss": 2.415, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.015700695168937413, | |
| "grad_norm": 4.487939357757568, | |
| "learning_rate": 0.00019839833266128724, | |
| "loss": 2.7945, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.015865965644399915, | |
| "grad_norm": 5.875516891479492, | |
| "learning_rate": 0.00019836052860592237, | |
| "loss": 2.836, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.016031236119862412, | |
| "grad_norm": 5.926601886749268, | |
| "learning_rate": 0.0001983222872823505, | |
| "loss": 2.5728, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.01619650659532491, | |
| "grad_norm": 6.333475589752197, | |
| "learning_rate": 0.00019828360886057594, | |
| "loss": 2.876, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.01636177707078741, | |
| "grad_norm": 6.237234115600586, | |
| "learning_rate": 0.00019824449351254616, | |
| "loss": 2.3797, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.01652704754624991, | |
| "grad_norm": 13.625974655151367, | |
| "learning_rate": 0.00019820494141215104, | |
| "loss": 3.2541, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.016692318021712407, | |
| "grad_norm": 23.068504333496094, | |
| "learning_rate": 0.000198164952735222, | |
| "loss": 2.113, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.01685758849717491, | |
| "grad_norm": 8.552973747253418, | |
| "learning_rate": 0.00019812452765953135, | |
| "loss": 1.8912, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.017022858972637406, | |
| "grad_norm": 3.735076427459717, | |
| "learning_rate": 0.00019808366636479147, | |
| "loss": 1.6648, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.017188129448099904, | |
| "grad_norm": 1.945611596107483, | |
| "learning_rate": 0.00019804236903265388, | |
| "loss": 2.1432, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.017353399923562406, | |
| "grad_norm": 1.5284167528152466, | |
| "learning_rate": 0.00019800063584670863, | |
| "loss": 2.7168, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.017518670399024903, | |
| "grad_norm": 0.8499382138252258, | |
| "learning_rate": 0.00019795846699248332, | |
| "loss": 2.2935, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.017683940874487405, | |
| "grad_norm": 1.260697603225708, | |
| "learning_rate": 0.00019791586265744237, | |
| "loss": 2.453, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.017849211349949903, | |
| "grad_norm": 1.3915923833847046, | |
| "learning_rate": 0.00019787282303098617, | |
| "loss": 2.1036, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.0180144818254124, | |
| "grad_norm": 1.0668586492538452, | |
| "learning_rate": 0.0001978293483044502, | |
| "loss": 2.2599, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.018179752300874902, | |
| "grad_norm": 1.0319124460220337, | |
| "learning_rate": 0.00019778543867110426, | |
| "loss": 2.5129, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0183450227763374, | |
| "grad_norm": 1.0817118883132935, | |
| "learning_rate": 0.00019774109432615147, | |
| "loss": 2.5452, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.018510293251799897, | |
| "grad_norm": 0.9889940619468689, | |
| "learning_rate": 0.00019769631546672756, | |
| "loss": 2.2689, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.0186755637272624, | |
| "grad_norm": 1.056303858757019, | |
| "learning_rate": 0.00019765110229189988, | |
| "loss": 2.2817, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.018840834202724897, | |
| "grad_norm": 1.0146815776824951, | |
| "learning_rate": 0.00019760545500266657, | |
| "loss": 2.4426, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.019006104678187394, | |
| "grad_norm": 0.9750092029571533, | |
| "learning_rate": 0.00019755937380195568, | |
| "loss": 2.3229, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.019171375153649896, | |
| "grad_norm": 1.1180509328842163, | |
| "learning_rate": 0.00019751285889462423, | |
| "loss": 2.3687, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.019336645629112394, | |
| "grad_norm": 1.4042870998382568, | |
| "learning_rate": 0.0001974659104874573, | |
| "loss": 2.824, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.019501916104574895, | |
| "grad_norm": 1.190585970878601, | |
| "learning_rate": 0.0001974185287891671, | |
| "loss": 2.2264, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.019667186580037393, | |
| "grad_norm": 1.5875192880630493, | |
| "learning_rate": 0.0001973707140103921, | |
| "loss": 2.6505, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.01983245705549989, | |
| "grad_norm": 1.5820868015289307, | |
| "learning_rate": 0.00019732246636369605, | |
| "loss": 2.2316, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.019997727530962392, | |
| "grad_norm": 1.3869949579238892, | |
| "learning_rate": 0.00019727378606356703, | |
| "loss": 2.0768, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.02016299800642489, | |
| "grad_norm": 1.5926547050476074, | |
| "learning_rate": 0.00019722467332641656, | |
| "loss": 2.6443, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.020328268481887388, | |
| "grad_norm": 1.7064094543457031, | |
| "learning_rate": 0.00019717512837057855, | |
| "loss": 2.4537, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.02049353895734989, | |
| "grad_norm": 1.7296555042266846, | |
| "learning_rate": 0.0001971251514163083, | |
| "loss": 2.287, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.020658809432812387, | |
| "grad_norm": 1.715631127357483, | |
| "learning_rate": 0.0001970747426857817, | |
| "loss": 2.3228, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.020824079908274885, | |
| "grad_norm": 1.858058214187622, | |
| "learning_rate": 0.00019702390240309404, | |
| "loss": 2.6594, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.020989350383737386, | |
| "grad_norm": 1.8299081325531006, | |
| "learning_rate": 0.0001969726307942592, | |
| "loss": 2.9619, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.021154620859199884, | |
| "grad_norm": 1.933591604232788, | |
| "learning_rate": 0.00019692092808720846, | |
| "loss": 2.6431, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.02131989133466238, | |
| "grad_norm": 2.227861166000366, | |
| "learning_rate": 0.0001968687945117896, | |
| "loss": 2.7041, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.021485161810124883, | |
| "grad_norm": 2.2244441509246826, | |
| "learning_rate": 0.00019681623029976588, | |
| "loss": 2.5885, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.02165043228558738, | |
| "grad_norm": 2.0957236289978027, | |
| "learning_rate": 0.00019676323568481498, | |
| "loss": 2.4978, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.021815702761049882, | |
| "grad_norm": 2.260390043258667, | |
| "learning_rate": 0.00019670981090252792, | |
| "loss": 2.3771, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.02198097323651238, | |
| "grad_norm": 2.3658909797668457, | |
| "learning_rate": 0.00019665595619040808, | |
| "loss": 3.0458, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.022146243711974878, | |
| "grad_norm": 2.7521181106567383, | |
| "learning_rate": 0.0001966016717878702, | |
| "loss": 2.9837, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.02231151418743738, | |
| "grad_norm": 2.6171624660491943, | |
| "learning_rate": 0.00019654695793623907, | |
| "loss": 2.5488, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.022476784662899877, | |
| "grad_norm": 2.8344151973724365, | |
| "learning_rate": 0.0001964918148787488, | |
| "loss": 2.3022, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.022642055138362375, | |
| "grad_norm": 3.373476982116699, | |
| "learning_rate": 0.00019643624286054144, | |
| "loss": 2.9102, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.022807325613824876, | |
| "grad_norm": 3.4098997116088867, | |
| "learning_rate": 0.00019638024212866606, | |
| "loss": 2.9409, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.022972596089287374, | |
| "grad_norm": 3.242896795272827, | |
| "learning_rate": 0.0001963238129320776, | |
| "loss": 2.3314, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.023137866564749872, | |
| "grad_norm": 3.7681121826171875, | |
| "learning_rate": 0.00019626695552163578, | |
| "loss": 2.7552, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.023303137040212373, | |
| "grad_norm": 3.4483370780944824, | |
| "learning_rate": 0.00019620967015010395, | |
| "loss": 2.6775, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.02346840751567487, | |
| "grad_norm": 3.8740861415863037, | |
| "learning_rate": 0.00019615195707214803, | |
| "loss": 2.4138, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.023633677991137372, | |
| "grad_norm": 3.6632370948791504, | |
| "learning_rate": 0.0001960938165443353, | |
| "loss": 2.4943, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.02379894846659987, | |
| "grad_norm": 4.024032115936279, | |
| "learning_rate": 0.00019603524882513327, | |
| "loss": 2.6587, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.023964218942062368, | |
| "grad_norm": 4.615535259246826, | |
| "learning_rate": 0.0001959762541749086, | |
| "loss": 2.8448, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.02412948941752487, | |
| "grad_norm": 4.237120151519775, | |
| "learning_rate": 0.00019591683285592593, | |
| "loss": 2.3161, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.024294759892987367, | |
| "grad_norm": 4.228479862213135, | |
| "learning_rate": 0.00019585698513234663, | |
| "loss": 2.3943, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.024460030368449865, | |
| "grad_norm": 5.225378513336182, | |
| "learning_rate": 0.0001957967112702277, | |
| "loss": 2.1276, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.024625300843912366, | |
| "grad_norm": 5.5622735023498535, | |
| "learning_rate": 0.00019573601153752052, | |
| "loss": 2.3912, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.024790571319374864, | |
| "grad_norm": 9.477717399597168, | |
| "learning_rate": 0.00019567488620406983, | |
| "loss": 3.0933, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.024955841794837362, | |
| "grad_norm": 5.711953639984131, | |
| "learning_rate": 0.00019561333554161224, | |
| "loss": 1.6375, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.025121112270299863, | |
| "grad_norm": 5.625733375549316, | |
| "learning_rate": 0.0001955513598237753, | |
| "loss": 1.6595, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.02528638274576236, | |
| "grad_norm": 2.32511568069458, | |
| "learning_rate": 0.00019548895932607621, | |
| "loss": 1.6096, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.025451653221224863, | |
| "grad_norm": 1.23055100440979, | |
| "learning_rate": 0.00019542613432592038, | |
| "loss": 2.1636, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.02561692369668736, | |
| "grad_norm": 0.993299663066864, | |
| "learning_rate": 0.00019536288510260056, | |
| "loss": 2.6037, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.025782194172149858, | |
| "grad_norm": 0.7763637900352478, | |
| "learning_rate": 0.00019529921193729534, | |
| "loss": 2.0913, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.02594746464761236, | |
| "grad_norm": 0.9637672305107117, | |
| "learning_rate": 0.00019523511511306793, | |
| "loss": 2.4378, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.026112735123074857, | |
| "grad_norm": 0.7835621237754822, | |
| "learning_rate": 0.000195170594914865, | |
| "loss": 2.5359, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.026278005598537355, | |
| "grad_norm": 0.9140456318855286, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 2.1616, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.026443276073999857, | |
| "grad_norm": 0.9035102128982544, | |
| "learning_rate": 0.00019504028554572864, | |
| "loss": 2.357, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.026608546549462354, | |
| "grad_norm": 1.1492279767990112, | |
| "learning_rate": 0.00019497449695409408, | |
| "loss": 2.6761, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.026773817024924852, | |
| "grad_norm": 0.9569836854934692, | |
| "learning_rate": 0.00019490828614707916, | |
| "loss": 2.0154, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.026939087500387354, | |
| "grad_norm": 1.024749994277954, | |
| "learning_rate": 0.00019484165341902845, | |
| "loss": 2.1919, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.02710435797584985, | |
| "grad_norm": 1.0347239971160889, | |
| "learning_rate": 0.00019477459906616206, | |
| "loss": 2.3991, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.02726962845131235, | |
| "grad_norm": 1.0091862678527832, | |
| "learning_rate": 0.00019470712338657458, | |
| "loss": 2.3106, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.02743489892677485, | |
| "grad_norm": 1.146797776222229, | |
| "learning_rate": 0.0001946392266802336, | |
| "loss": 2.8065, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.02760016940223735, | |
| "grad_norm": 1.1744192838668823, | |
| "learning_rate": 0.0001945709092489783, | |
| "loss": 2.3922, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.02776543987769985, | |
| "grad_norm": 1.3500045537948608, | |
| "learning_rate": 0.00019450217139651844, | |
| "loss": 2.2811, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.027930710353162348, | |
| "grad_norm": 1.4020427465438843, | |
| "learning_rate": 0.0001944330134284326, | |
| "loss": 2.0844, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.028095980828624845, | |
| "grad_norm": 1.463188886642456, | |
| "learning_rate": 0.00019436343565216711, | |
| "loss": 2.4052, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.028261251304087347, | |
| "grad_norm": 1.5158060789108276, | |
| "learning_rate": 0.00019429343837703455, | |
| "loss": 2.6677, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.028426521779549845, | |
| "grad_norm": 1.5841165781021118, | |
| "learning_rate": 0.0001942230219142124, | |
| "loss": 2.4633, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.028591792255012342, | |
| "grad_norm": 1.6128944158554077, | |
| "learning_rate": 0.0001941521865767417, | |
| "loss": 2.1815, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.028757062730474844, | |
| "grad_norm": 1.8062238693237305, | |
| "learning_rate": 0.0001940809326795256, | |
| "loss": 2.3996, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.02892233320593734, | |
| "grad_norm": 1.8842582702636719, | |
| "learning_rate": 0.000194009260539328, | |
| "loss": 2.9509, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.02908760368139984, | |
| "grad_norm": 2.1278233528137207, | |
| "learning_rate": 0.0001939371704747721, | |
| "loss": 2.4552, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.02925287415686234, | |
| "grad_norm": 2.0727250576019287, | |
| "learning_rate": 0.00019386466280633906, | |
| "loss": 2.5597, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.02941814463232484, | |
| "grad_norm": 2.2283802032470703, | |
| "learning_rate": 0.00019379173785636646, | |
| "loss": 2.7298, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.02958341510778734, | |
| "grad_norm": 2.0877585411071777, | |
| "learning_rate": 0.000193718395949047, | |
| "loss": 2.9711, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.029748685583249838, | |
| "grad_norm": 2.1163971424102783, | |
| "learning_rate": 0.00019364463741042694, | |
| "loss": 2.614, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.029913956058712336, | |
| "grad_norm": 2.237908124923706, | |
| "learning_rate": 0.00019357046256840473, | |
| "loss": 2.4931, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.030079226534174837, | |
| "grad_norm": 2.3826680183410645, | |
| "learning_rate": 0.00019349587175272948, | |
| "loss": 2.6467, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.030244497009637335, | |
| "grad_norm": 2.203145742416382, | |
| "learning_rate": 0.0001934208652949996, | |
| "loss": 2.4464, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.030409767485099833, | |
| "grad_norm": 2.6857528686523438, | |
| "learning_rate": 0.00019334544352866127, | |
| "loss": 2.7842, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.030575037960562334, | |
| "grad_norm": 2.8190665245056152, | |
| "learning_rate": 0.00019326960678900688, | |
| "loss": 2.445, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.03074030843602483, | |
| "grad_norm": 2.9053189754486084, | |
| "learning_rate": 0.00019319335541317361, | |
| "loss": 2.6002, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.03090557891148733, | |
| "grad_norm": 3.018693447113037, | |
| "learning_rate": 0.00019311668974014208, | |
| "loss": 2.7639, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.03107084938694983, | |
| "grad_norm": 3.129119873046875, | |
| "learning_rate": 0.00019303961011073447, | |
| "loss": 2.8864, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.03123611986241233, | |
| "grad_norm": 3.7832894325256348, | |
| "learning_rate": 0.00019296211686761346, | |
| "loss": 2.8485, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.03140139033787483, | |
| "grad_norm": 3.1579928398132324, | |
| "learning_rate": 0.00019288421035528028, | |
| "loss": 2.5657, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.031566660813337324, | |
| "grad_norm": 3.414823055267334, | |
| "learning_rate": 0.00019280589092007352, | |
| "loss": 2.6183, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.03173193128879983, | |
| "grad_norm": 3.225539445877075, | |
| "learning_rate": 0.00019272715891016735, | |
| "loss": 2.4056, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.03189720176426233, | |
| "grad_norm": 4.247017860412598, | |
| "learning_rate": 0.00019264801467557007, | |
| "loss": 2.4858, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.032062472239724825, | |
| "grad_norm": 4.451138019561768, | |
| "learning_rate": 0.00019256845856812266, | |
| "loss": 2.6277, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.03222774271518732, | |
| "grad_norm": 4.106354236602783, | |
| "learning_rate": 0.000192488490941497, | |
| "loss": 2.5092, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.03239301319064982, | |
| "grad_norm": 4.525591850280762, | |
| "learning_rate": 0.00019240811215119448, | |
| "loss": 2.6473, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.032558283666112325, | |
| "grad_norm": 5.480126857757568, | |
| "learning_rate": 0.00019232732255454422, | |
| "loss": 2.7263, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.03272355414157482, | |
| "grad_norm": 4.905240535736084, | |
| "learning_rate": 0.00019224612251070175, | |
| "loss": 2.346, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.03288882461703732, | |
| "grad_norm": 6.290627956390381, | |
| "learning_rate": 0.0001921645123806472, | |
| "loss": 3.0139, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.03305409509249982, | |
| "grad_norm": 9.973276138305664, | |
| "learning_rate": 0.0001920824925271838, | |
| "loss": 2.9487, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03321936556796232, | |
| "grad_norm": 4.321014404296875, | |
| "learning_rate": 0.0001920000633149362, | |
| "loss": 1.5159, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.033384636043424815, | |
| "grad_norm": 3.545685291290283, | |
| "learning_rate": 0.00019191722511034884, | |
| "loss": 1.7116, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.03354990651888732, | |
| "grad_norm": 1.8792541027069092, | |
| "learning_rate": 0.00019183397828168448, | |
| "loss": 1.4233, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.03371517699434982, | |
| "grad_norm": 1.3337411880493164, | |
| "learning_rate": 0.00019175032319902234, | |
| "loss": 2.1826, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.033880447469812315, | |
| "grad_norm": 1.039467453956604, | |
| "learning_rate": 0.00019166626023425662, | |
| "loss": 2.3848, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.03404571794527481, | |
| "grad_norm": 1.1255942583084106, | |
| "learning_rate": 0.00019158178976109476, | |
| "loss": 2.751, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.03421098842073731, | |
| "grad_norm": 0.8885089159011841, | |
| "learning_rate": 0.0001914969121550558, | |
| "loss": 2.1767, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.03437625889619981, | |
| "grad_norm": 0.851932942867279, | |
| "learning_rate": 0.00019141162779346874, | |
| "loss": 2.1687, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.03454152937166231, | |
| "grad_norm": 1.1003276109695435, | |
| "learning_rate": 0.00019132593705547082, | |
| "loss": 2.2528, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.03470679984712481, | |
| "grad_norm": 0.8771634697914124, | |
| "learning_rate": 0.00019123984032200586, | |
| "loss": 2.2655, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.03487207032258731, | |
| "grad_norm": 0.9093711376190186, | |
| "learning_rate": 0.00019115333797582254, | |
| "loss": 2.2472, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.03503734079804981, | |
| "grad_norm": 0.9847548604011536, | |
| "learning_rate": 0.00019106643040147278, | |
| "loss": 2.3438, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.035202611273512305, | |
| "grad_norm": 0.9569659233093262, | |
| "learning_rate": 0.00019097911798530987, | |
| "loss": 2.3969, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.03536788174897481, | |
| "grad_norm": 1.0445773601531982, | |
| "learning_rate": 0.00019089140111548696, | |
| "loss": 2.6232, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.03553315222443731, | |
| "grad_norm": 1.0466011762619019, | |
| "learning_rate": 0.00019080328018195513, | |
| "loss": 2.2336, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.035698422699899805, | |
| "grad_norm": 1.174451470375061, | |
| "learning_rate": 0.0001907147555764618, | |
| "loss": 1.9445, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.0358636931753623, | |
| "grad_norm": 1.181753158569336, | |
| "learning_rate": 0.00019062582769254895, | |
| "loss": 2.1892, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.0360289636508248, | |
| "grad_norm": 1.2002571821212769, | |
| "learning_rate": 0.00019053649692555135, | |
| "loss": 2.0725, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.0361942341262873, | |
| "grad_norm": 1.295405626296997, | |
| "learning_rate": 0.00019044676367259476, | |
| "loss": 2.6805, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.036359504601749804, | |
| "grad_norm": 1.330724835395813, | |
| "learning_rate": 0.00019035662833259432, | |
| "loss": 2.2376, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.0365247750772123, | |
| "grad_norm": 1.3845914602279663, | |
| "learning_rate": 0.00019026609130625257, | |
| "loss": 2.4574, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.0366900455526748, | |
| "grad_norm": 1.5534614324569702, | |
| "learning_rate": 0.00019017515299605788, | |
| "loss": 2.5799, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.0368553160281373, | |
| "grad_norm": 1.6567062139511108, | |
| "learning_rate": 0.00019008381380628247, | |
| "loss": 2.742, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.037020586503599795, | |
| "grad_norm": 1.8053407669067383, | |
| "learning_rate": 0.00018999207414298067, | |
| "loss": 3.0706, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.0371858569790623, | |
| "grad_norm": 1.8480101823806763, | |
| "learning_rate": 0.00018989993441398726, | |
| "loss": 2.3829, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.0373511274545248, | |
| "grad_norm": 1.8867958784103394, | |
| "learning_rate": 0.00018980739502891546, | |
| "loss": 3.0338, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.037516397929987295, | |
| "grad_norm": 1.800199031829834, | |
| "learning_rate": 0.0001897144563991552, | |
| "loss": 2.1971, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.03768166840544979, | |
| "grad_norm": 1.9533302783966064, | |
| "learning_rate": 0.00018962111893787128, | |
| "loss": 2.1371, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.03784693888091229, | |
| "grad_norm": 2.0056862831115723, | |
| "learning_rate": 0.00018952738306000151, | |
| "loss": 2.7914, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.03801220935637479, | |
| "grad_norm": 2.2028658390045166, | |
| "learning_rate": 0.00018943324918225494, | |
| "loss": 2.664, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.038177479831837294, | |
| "grad_norm": 2.3667781352996826, | |
| "learning_rate": 0.0001893387177231099, | |
| "loss": 2.4894, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.03834275030729979, | |
| "grad_norm": 2.0973241329193115, | |
| "learning_rate": 0.0001892437891028122, | |
| "loss": 2.5924, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.03850802078276229, | |
| "grad_norm": 2.24971079826355, | |
| "learning_rate": 0.0001891484637433733, | |
| "loss": 2.2934, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.03867329125822479, | |
| "grad_norm": 2.577932834625244, | |
| "learning_rate": 0.00018905274206856837, | |
| "loss": 2.4019, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.038838561733687285, | |
| "grad_norm": 2.669767141342163, | |
| "learning_rate": 0.00018895662450393438, | |
| "loss": 2.4896, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.03900383220914979, | |
| "grad_norm": 2.795431613922119, | |
| "learning_rate": 0.00018886011147676833, | |
| "loss": 2.4702, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.03916910268461229, | |
| "grad_norm": 2.6597859859466553, | |
| "learning_rate": 0.00018876320341612522, | |
| "loss": 2.5595, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.039334373160074786, | |
| "grad_norm": 2.9854073524475098, | |
| "learning_rate": 0.00018866590075281624, | |
| "loss": 2.3452, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.039499643635537283, | |
| "grad_norm": 3.1764540672302246, | |
| "learning_rate": 0.00018856820391940674, | |
| "loss": 2.3393, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.03966491411099978, | |
| "grad_norm": 2.9728143215179443, | |
| "learning_rate": 0.00018847011335021449, | |
| "loss": 2.5481, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.03983018458646228, | |
| "grad_norm": 4.049989700317383, | |
| "learning_rate": 0.00018837162948130752, | |
| "loss": 3.0019, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.039995455061924784, | |
| "grad_norm": 3.720461845397949, | |
| "learning_rate": 0.00018827275275050233, | |
| "loss": 2.6645, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.04016072553738728, | |
| "grad_norm": 3.652611494064331, | |
| "learning_rate": 0.00018817348359736203, | |
| "loss": 2.4387, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.04032599601284978, | |
| "grad_norm": 4.24875545501709, | |
| "learning_rate": 0.00018807382246319412, | |
| "loss": 2.3321, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.04049126648831228, | |
| "grad_norm": 4.664936065673828, | |
| "learning_rate": 0.00018797376979104872, | |
| "loss": 2.4599, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.040656536963774775, | |
| "grad_norm": 4.748204708099365, | |
| "learning_rate": 0.00018787332602571662, | |
| "loss": 2.2912, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.04082180743923728, | |
| "grad_norm": 5.072621822357178, | |
| "learning_rate": 0.00018777249161372713, | |
| "loss": 2.0861, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.04098707791469978, | |
| "grad_norm": 6.534742832183838, | |
| "learning_rate": 0.00018767126700334634, | |
| "loss": 2.5439, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.041152348390162276, | |
| "grad_norm": 6.339376926422119, | |
| "learning_rate": 0.0001875696526445749, | |
| "loss": 2.5152, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.041317618865624774, | |
| "grad_norm": 10.001213073730469, | |
| "learning_rate": 0.0001874676489891461, | |
| "loss": 3.092, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.04148288934108727, | |
| "grad_norm": 3.0091938972473145, | |
| "learning_rate": 0.00018736525649052394, | |
| "loss": 1.3808, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.04164815981654977, | |
| "grad_norm": 2.9950389862060547, | |
| "learning_rate": 0.00018726247560390099, | |
| "loss": 1.4688, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.041813430292012274, | |
| "grad_norm": 1.7389607429504395, | |
| "learning_rate": 0.00018715930678619644, | |
| "loss": 1.5949, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.04197870076747477, | |
| "grad_norm": 1.4947503805160522, | |
| "learning_rate": 0.00018705575049605413, | |
| "loss": 2.5111, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.04214397124293727, | |
| "grad_norm": 1.1639373302459717, | |
| "learning_rate": 0.00018695180719384029, | |
| "loss": 2.5251, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.04230924171839977, | |
| "grad_norm": 0.8776928782463074, | |
| "learning_rate": 0.00018684747734164177, | |
| "loss": 2.3522, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.042474512193862266, | |
| "grad_norm": 0.7244071364402771, | |
| "learning_rate": 0.00018674276140326376, | |
| "loss": 2.0348, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.04263978266932476, | |
| "grad_norm": 0.9879050254821777, | |
| "learning_rate": 0.00018663765984422786, | |
| "loss": 2.5215, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.04280505314478727, | |
| "grad_norm": 0.8424708843231201, | |
| "learning_rate": 0.00018653217313177004, | |
| "loss": 2.4893, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.042970323620249766, | |
| "grad_norm": 0.8051637411117554, | |
| "learning_rate": 0.00018642630173483832, | |
| "loss": 2.1197, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.043135594095712264, | |
| "grad_norm": 0.9109228849411011, | |
| "learning_rate": 0.00018632004612409103, | |
| "loss": 2.2626, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.04330086457117476, | |
| "grad_norm": 1.0443938970565796, | |
| "learning_rate": 0.00018621340677189453, | |
| "loss": 2.2493, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.04346613504663726, | |
| "grad_norm": 0.9891976714134216, | |
| "learning_rate": 0.00018610638415232097, | |
| "loss": 2.088, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.043631405522099764, | |
| "grad_norm": 0.9806310534477234, | |
| "learning_rate": 0.00018599897874114652, | |
| "loss": 2.3922, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.04379667599756226, | |
| "grad_norm": 1.1193654537200928, | |
| "learning_rate": 0.00018589119101584898, | |
| "loss": 2.53, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.04396194647302476, | |
| "grad_norm": 1.160233974456787, | |
| "learning_rate": 0.00018578302145560584, | |
| "loss": 2.262, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.04412721694848726, | |
| "grad_norm": 1.338555097579956, | |
| "learning_rate": 0.00018567447054129195, | |
| "loss": 2.4142, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.044292487423949756, | |
| "grad_norm": 1.269381046295166, | |
| "learning_rate": 0.00018556553875547754, | |
| "loss": 2.4504, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.044457757899412254, | |
| "grad_norm": 1.391980767250061, | |
| "learning_rate": 0.00018545622658242607, | |
| "loss": 2.326, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.04462302837487476, | |
| "grad_norm": 1.3978862762451172, | |
| "learning_rate": 0.00018534653450809197, | |
| "loss": 2.2657, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.044788298850337256, | |
| "grad_norm": 1.4855581521987915, | |
| "learning_rate": 0.00018523646302011867, | |
| "loss": 2.6734, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.044953569325799754, | |
| "grad_norm": 1.6173784732818604, | |
| "learning_rate": 0.00018512601260783606, | |
| "loss": 2.5432, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.04511883980126225, | |
| "grad_norm": 1.684188961982727, | |
| "learning_rate": 0.00018501518376225887, | |
| "loss": 2.1659, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.04528411027672475, | |
| "grad_norm": 1.8205846548080444, | |
| "learning_rate": 0.00018490397697608395, | |
| "loss": 2.371, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.045449380752187254, | |
| "grad_norm": 1.88100266456604, | |
| "learning_rate": 0.0001847923927436884, | |
| "loss": 2.6993, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.04561465122764975, | |
| "grad_norm": 1.8194174766540527, | |
| "learning_rate": 0.00018468043156112728, | |
| "loss": 2.4791, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.04577992170311225, | |
| "grad_norm": 1.9578603506088257, | |
| "learning_rate": 0.0001845680939261314, | |
| "loss": 2.5443, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.04594519217857475, | |
| "grad_norm": 2.062070846557617, | |
| "learning_rate": 0.00018445538033810515, | |
| "loss": 2.5447, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.046110462654037246, | |
| "grad_norm": 2.3316287994384766, | |
| "learning_rate": 0.00018434229129812418, | |
| "loss": 2.7685, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.046275733129499744, | |
| "grad_norm": 2.1271250247955322, | |
| "learning_rate": 0.0001842288273089332, | |
| "loss": 2.8287, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.04644100360496225, | |
| "grad_norm": 2.2673919200897217, | |
| "learning_rate": 0.00018411498887494396, | |
| "loss": 2.7384, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.046606274080424746, | |
| "grad_norm": 2.4204909801483154, | |
| "learning_rate": 0.00018400077650223263, | |
| "loss": 2.7451, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.046771544555887244, | |
| "grad_norm": 2.4176559448242188, | |
| "learning_rate": 0.0001838861906985379, | |
| "loss": 2.5283, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.04693681503134974, | |
| "grad_norm": 2.111084461212158, | |
| "learning_rate": 0.00018377123197325842, | |
| "loss": 2.6704, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.04710208550681224, | |
| "grad_norm": 2.1875665187835693, | |
| "learning_rate": 0.00018365590083745085, | |
| "loss": 2.5255, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.047267355982274745, | |
| "grad_norm": 2.8831610679626465, | |
| "learning_rate": 0.00018354019780382735, | |
| "loss": 2.5586, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.04743262645773724, | |
| "grad_norm": 2.8357796669006348, | |
| "learning_rate": 0.0001834241233867533, | |
| "loss": 2.6485, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.04759789693319974, | |
| "grad_norm": 3.0059330463409424, | |
| "learning_rate": 0.00018330767810224524, | |
| "loss": 2.5055, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.04776316740866224, | |
| "grad_norm": 3.1243112087249756, | |
| "learning_rate": 0.0001831908624679683, | |
| "loss": 2.1974, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.047928437884124736, | |
| "grad_norm": 3.085310459136963, | |
| "learning_rate": 0.0001830736770032341, | |
| "loss": 2.2185, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.048093708359587234, | |
| "grad_norm": 3.7968411445617676, | |
| "learning_rate": 0.0001829561222289984, | |
| "loss": 2.5619, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.04825897883504974, | |
| "grad_norm": 3.428130865097046, | |
| "learning_rate": 0.00018283819866785853, | |
| "loss": 2.4643, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.048424249310512237, | |
| "grad_norm": 3.883866786956787, | |
| "learning_rate": 0.0001827199068440516, | |
| "loss": 2.7354, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.048589519785974734, | |
| "grad_norm": 4.249161243438721, | |
| "learning_rate": 0.00018260124728345162, | |
| "loss": 2.4175, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.04875479026143723, | |
| "grad_norm": 4.839107036590576, | |
| "learning_rate": 0.00018248222051356754, | |
| "loss": 2.4766, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.04892006073689973, | |
| "grad_norm": 4.305207252502441, | |
| "learning_rate": 0.00018236282706354063, | |
| "loss": 2.6408, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.049085331212362235, | |
| "grad_norm": 4.896646976470947, | |
| "learning_rate": 0.00018224306746414238, | |
| "loss": 2.3499, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.04925060168782473, | |
| "grad_norm": 5.219709396362305, | |
| "learning_rate": 0.00018212294224777197, | |
| "loss": 2.6275, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.04941587216328723, | |
| "grad_norm": 5.9750590324401855, | |
| "learning_rate": 0.00018200245194845399, | |
| "loss": 2.5248, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.04958114263874973, | |
| "grad_norm": 7.813031196594238, | |
| "learning_rate": 0.00018188159710183594, | |
| "loss": 3.1461, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.049746413114212226, | |
| "grad_norm": 2.398712396621704, | |
| "learning_rate": 0.000181760378245186, | |
| "loss": 1.3916, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.049911683589674724, | |
| "grad_norm": 2.724813222885132, | |
| "learning_rate": 0.00018163879591739067, | |
| "loss": 1.4479, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.05007695406513723, | |
| "grad_norm": 2.2264227867126465, | |
| "learning_rate": 0.0001815168506589521, | |
| "loss": 1.7367, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.05024222454059973, | |
| "grad_norm": 1.3800264596939087, | |
| "learning_rate": 0.000181394543011986, | |
| "loss": 2.0049, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.050407495016062225, | |
| "grad_norm": 1.0985920429229736, | |
| "learning_rate": 0.00018127187352021907, | |
| "loss": 2.4378, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.05057276549152472, | |
| "grad_norm": 1.1383318901062012, | |
| "learning_rate": 0.0001811488427289866, | |
| "loss": 2.9098, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.05073803596698722, | |
| "grad_norm": 0.8942216634750366, | |
| "learning_rate": 0.00018102545118523007, | |
| "loss": 1.961, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.050903306442449725, | |
| "grad_norm": 0.8574941754341125, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 2.5375, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.05106857691791222, | |
| "grad_norm": 0.8979453444480896, | |
| "learning_rate": 0.00018077758803592718, | |
| "loss": 2.5273, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.05123384739337472, | |
| "grad_norm": 0.8864741921424866, | |
| "learning_rate": 0.00018065311753227273, | |
| "loss": 2.5011, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.05139911786883722, | |
| "grad_norm": 1.0070165395736694, | |
| "learning_rate": 0.0001805282884798732, | |
| "loss": 2.2337, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.051564388344299716, | |
| "grad_norm": 0.9006021618843079, | |
| "learning_rate": 0.00018040310143366446, | |
| "loss": 1.973, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.051729658819762214, | |
| "grad_norm": 0.8520970344543457, | |
| "learning_rate": 0.00018027755695017368, | |
| "loss": 2.1139, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.05189492929522472, | |
| "grad_norm": 1.1375415325164795, | |
| "learning_rate": 0.00018015165558751717, | |
| "loss": 2.1686, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.05206019977068722, | |
| "grad_norm": 1.0089519023895264, | |
| "learning_rate": 0.00018002539790539773, | |
| "loss": 2.106, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.052225470246149715, | |
| "grad_norm": 1.0937998294830322, | |
| "learning_rate": 0.00017989878446510215, | |
| "loss": 2.3221, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.05239074072161221, | |
| "grad_norm": 1.173466682434082, | |
| "learning_rate": 0.00017977181582949888, | |
| "loss": 2.1659, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.05255601119707471, | |
| "grad_norm": 1.1528087854385376, | |
| "learning_rate": 0.0001796444925630353, | |
| "loss": 2.2379, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.05272128167253721, | |
| "grad_norm": 1.1747756004333496, | |
| "learning_rate": 0.00017951681523173542, | |
| "loss": 2.0929, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.05288655214799971, | |
| "grad_norm": 1.2405529022216797, | |
| "learning_rate": 0.0001793887844031972, | |
| "loss": 2.2142, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.05305182262346221, | |
| "grad_norm": 1.4098936319351196, | |
| "learning_rate": 0.00017926040064659014, | |
| "loss": 2.645, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.05321709309892471, | |
| "grad_norm": 1.395555853843689, | |
| "learning_rate": 0.0001791316645326526, | |
| "loss": 2.0263, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.05338236357438721, | |
| "grad_norm": 1.5886664390563965, | |
| "learning_rate": 0.00017900257663368963, | |
| "loss": 2.0161, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.053547634049849704, | |
| "grad_norm": 1.722530484199524, | |
| "learning_rate": 0.0001788731375235698, | |
| "loss": 2.5239, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.05371290452531221, | |
| "grad_norm": 1.788002371788025, | |
| "learning_rate": 0.00017874334777772327, | |
| "loss": 2.3831, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.05387817500077471, | |
| "grad_norm": 1.8565741777420044, | |
| "learning_rate": 0.00017861320797313892, | |
| "loss": 2.4004, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.054043445476237205, | |
| "grad_norm": 1.7931753396987915, | |
| "learning_rate": 0.0001784827186883618, | |
| "loss": 2.0539, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.0542087159516997, | |
| "grad_norm": 1.898111343383789, | |
| "learning_rate": 0.00017835188050349064, | |
| "loss": 2.3671, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.0543739864271622, | |
| "grad_norm": 2.0476067066192627, | |
| "learning_rate": 0.00017822069400017516, | |
| "loss": 2.1538, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.0545392569026247, | |
| "grad_norm": 2.189760684967041, | |
| "learning_rate": 0.00017808915976161362, | |
| "loss": 2.7559, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.0547045273780872, | |
| "grad_norm": 2.1433463096618652, | |
| "learning_rate": 0.00017795727837255015, | |
| "loss": 2.3307, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.0548697978535497, | |
| "grad_norm": 2.2783188819885254, | |
| "learning_rate": 0.00017782505041927216, | |
| "loss": 2.5076, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.0550350683290122, | |
| "grad_norm": 2.6381657123565674, | |
| "learning_rate": 0.00017769247648960774, | |
| "loss": 2.793, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.0552003388044747, | |
| "grad_norm": 2.3586020469665527, | |
| "learning_rate": 0.00017755955717292296, | |
| "loss": 2.6804, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.055365609279937195, | |
| "grad_norm": 2.7866647243499756, | |
| "learning_rate": 0.00017742629306011944, | |
| "loss": 2.7621, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.0555308797553997, | |
| "grad_norm": 2.6400251388549805, | |
| "learning_rate": 0.00017729268474363154, | |
| "loss": 2.5589, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.0556961502308622, | |
| "grad_norm": 2.8022119998931885, | |
| "learning_rate": 0.0001771587328174239, | |
| "loss": 2.0524, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.055861420706324695, | |
| "grad_norm": 2.8182830810546875, | |
| "learning_rate": 0.0001770244378769885, | |
| "loss": 2.2297, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.05602669118178719, | |
| "grad_norm": 3.4760756492614746, | |
| "learning_rate": 0.0001768898005193425, | |
| "loss": 2.8004, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.05619196165724969, | |
| "grad_norm": 3.309964179992676, | |
| "learning_rate": 0.000176754821343025, | |
| "loss": 2.5758, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.05635723213271219, | |
| "grad_norm": 3.500337839126587, | |
| "learning_rate": 0.0001766195009480949, | |
| "loss": 2.6575, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.05652250260817469, | |
| "grad_norm": 3.5581259727478027, | |
| "learning_rate": 0.0001764838399361279, | |
| "loss": 2.4085, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.05668777308363719, | |
| "grad_norm": 4.1034464836120605, | |
| "learning_rate": 0.00017634783891021393, | |
| "loss": 2.5597, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.05685304355909969, | |
| "grad_norm": 3.7894608974456787, | |
| "learning_rate": 0.00017621149847495458, | |
| "loss": 2.2587, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.05701831403456219, | |
| "grad_norm": 4.066458225250244, | |
| "learning_rate": 0.00017607481923646016, | |
| "loss": 2.2147, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.057183584510024685, | |
| "grad_norm": 4.655360698699951, | |
| "learning_rate": 0.0001759378018023473, | |
| "loss": 2.5648, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.05734885498548719, | |
| "grad_norm": 4.92647123336792, | |
| "learning_rate": 0.00017580044678173592, | |
| "loss": 2.5519, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.05751412546094969, | |
| "grad_norm": 4.746926784515381, | |
| "learning_rate": 0.00017566275478524693, | |
| "loss": 2.2554, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.057679395936412185, | |
| "grad_norm": 5.611936569213867, | |
| "learning_rate": 0.0001755247264249991, | |
| "loss": 2.334, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.05784466641187468, | |
| "grad_norm": 13.957818984985352, | |
| "learning_rate": 0.0001753863623146066, | |
| "loss": 3.2839, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.05800993688733718, | |
| "grad_norm": 1.5909780263900757, | |
| "learning_rate": 0.00017524766306917618, | |
| "loss": 1.3852, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.05817520736279968, | |
| "grad_norm": 1.4798673391342163, | |
| "learning_rate": 0.0001751086293053045, | |
| "loss": 1.3216, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.058340477838262184, | |
| "grad_norm": 1.1548388004302979, | |
| "learning_rate": 0.0001749692616410753, | |
| "loss": 1.5986, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.05850574831372468, | |
| "grad_norm": 1.0224725008010864, | |
| "learning_rate": 0.00017482956069605668, | |
| "loss": 1.9637, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.05867101878918718, | |
| "grad_norm": 1.3334860801696777, | |
| "learning_rate": 0.00017468952709129846, | |
| "loss": 2.8391, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.05883628926464968, | |
| "grad_norm": 0.9578279256820679, | |
| "learning_rate": 0.00017454916144932922, | |
| "loss": 2.3261, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.059001559740112175, | |
| "grad_norm": 1.1700596809387207, | |
| "learning_rate": 0.0001744084643941536, | |
| "loss": 2.2642, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.05916683021557468, | |
| "grad_norm": 0.9093031287193298, | |
| "learning_rate": 0.00017426743655124974, | |
| "loss": 2.2101, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.05933210069103718, | |
| "grad_norm": 1.018043875694275, | |
| "learning_rate": 0.0001741260785475661, | |
| "loss": 2.0616, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.059497371166499675, | |
| "grad_norm": 1.1330430507659912, | |
| "learning_rate": 0.00017398439101151905, | |
| "loss": 2.2073, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.05966264164196217, | |
| "grad_norm": 0.9593504667282104, | |
| "learning_rate": 0.00017384237457298987, | |
| "loss": 1.8434, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.05982791211742467, | |
| "grad_norm": 1.095883846282959, | |
| "learning_rate": 0.00017370002986332193, | |
| "loss": 2.122, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.05999318259288717, | |
| "grad_norm": 1.01414155960083, | |
| "learning_rate": 0.00017355735751531807, | |
| "loss": 2.3286, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.060158453068349674, | |
| "grad_norm": 0.9529690742492676, | |
| "learning_rate": 0.00017341435816323756, | |
| "loss": 2.2769, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.06032372354381217, | |
| "grad_norm": 1.064019799232483, | |
| "learning_rate": 0.00017327103244279348, | |
| "loss": 2.265, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.06048899401927467, | |
| "grad_norm": 1.2005796432495117, | |
| "learning_rate": 0.00017312738099114973, | |
| "loss": 2.8063, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.06065426449473717, | |
| "grad_norm": 1.3219201564788818, | |
| "learning_rate": 0.00017298340444691835, | |
| "loss": 2.1139, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.060819534970199665, | |
| "grad_norm": 1.2771499156951904, | |
| "learning_rate": 0.00017283910345015647, | |
| "loss": 2.5412, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.06098480544566217, | |
| "grad_norm": 1.546258568763733, | |
| "learning_rate": 0.0001726944786423637, | |
| "loss": 2.3801, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.06115007592112467, | |
| "grad_norm": 1.4880259037017822, | |
| "learning_rate": 0.00017254953066647913, | |
| "loss": 2.4388, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.061315346396587166, | |
| "grad_norm": 1.5611305236816406, | |
| "learning_rate": 0.00017240426016687863, | |
| "loss": 2.2293, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.06148061687204966, | |
| "grad_norm": 1.683081865310669, | |
| "learning_rate": 0.00017225866778937165, | |
| "loss": 2.4346, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.06164588734751216, | |
| "grad_norm": 1.4754785299301147, | |
| "learning_rate": 0.00017211275418119876, | |
| "loss": 2.3159, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.06181115782297466, | |
| "grad_norm": 1.701194405555725, | |
| "learning_rate": 0.0001719665199910285, | |
| "loss": 2.6458, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.061976428298437164, | |
| "grad_norm": 1.8477551937103271, | |
| "learning_rate": 0.00017181996586895454, | |
| "loss": 3.1028, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.061976428298437164, | |
| "eval_loss": 2.339935064315796, | |
| "eval_runtime": 208.4998, | |
| "eval_samples_per_second": 48.878, | |
| "eval_steps_per_second": 24.441, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.06214169877389966, | |
| "grad_norm": 1.8668705224990845, | |
| "learning_rate": 0.00017167309246649297, | |
| "loss": 2.7361, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.06230696924936216, | |
| "grad_norm": 1.8983453512191772, | |
| "learning_rate": 0.0001715259004365791, | |
| "loss": 2.4626, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.06247223972482466, | |
| "grad_norm": 2.078368663787842, | |
| "learning_rate": 0.00017137839043356484, | |
| "loss": 2.9497, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.06263751020028716, | |
| "grad_norm": 1.9831960201263428, | |
| "learning_rate": 0.00017123056311321562, | |
| "loss": 2.084, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.06280278067574965, | |
| "grad_norm": 2.193898916244507, | |
| "learning_rate": 0.0001710824191327075, | |
| "loss": 2.5677, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.06296805115121215, | |
| "grad_norm": 2.473846197128296, | |
| "learning_rate": 0.00017093395915062428, | |
| "loss": 2.5835, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.06313332162667465, | |
| "grad_norm": 2.645812511444092, | |
| "learning_rate": 0.00017078518382695465, | |
| "loss": 2.5068, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.06329859210213716, | |
| "grad_norm": 2.563343048095703, | |
| "learning_rate": 0.00017063609382308908, | |
| "loss": 2.4763, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.06346386257759966, | |
| "grad_norm": 2.604940891265869, | |
| "learning_rate": 0.00017048668980181698, | |
| "loss": 2.4875, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.06362913305306216, | |
| "grad_norm": 3.5896048545837402, | |
| "learning_rate": 0.00017033697242732377, | |
| "loss": 2.4614, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.06379440352852465, | |
| "grad_norm": 2.652202844619751, | |
| "learning_rate": 0.0001701869423651879, | |
| "loss": 2.587, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.06395967400398715, | |
| "grad_norm": 3.466193675994873, | |
| "learning_rate": 0.00017003660028237793, | |
| "loss": 3.0918, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.06412494447944965, | |
| "grad_norm": 2.833293914794922, | |
| "learning_rate": 0.00016988594684724947, | |
| "loss": 2.3021, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.06429021495491215, | |
| "grad_norm": 3.474046468734741, | |
| "learning_rate": 0.00016973498272954222, | |
| "loss": 2.9898, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.06445548543037465, | |
| "grad_norm": 3.809767246246338, | |
| "learning_rate": 0.00016958370860037717, | |
| "loss": 2.7766, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.06462075590583714, | |
| "grad_norm": 3.208801507949829, | |
| "learning_rate": 0.00016943212513225345, | |
| "loss": 2.4729, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.06478602638129964, | |
| "grad_norm": 3.7613186836242676, | |
| "learning_rate": 0.00016928023299904533, | |
| "loss": 2.7564, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.06495129685676214, | |
| "grad_norm": 3.4937353134155273, | |
| "learning_rate": 0.0001691280328759992, | |
| "loss": 2.4674, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.06511656733222465, | |
| "grad_norm": 3.5094993114471436, | |
| "learning_rate": 0.00016897552543973084, | |
| "loss": 2.2195, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.06528183780768715, | |
| "grad_norm": 4.127921104431152, | |
| "learning_rate": 0.00016882271136822206, | |
| "loss": 2.3509, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.06544710828314965, | |
| "grad_norm": 3.795407295227051, | |
| "learning_rate": 0.0001686695913408179, | |
| "loss": 2.3811, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.06561237875861214, | |
| "grad_norm": 4.278930187225342, | |
| "learning_rate": 0.0001685161660382235, | |
| "loss": 2.3552, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.06577764923407464, | |
| "grad_norm": 4.8861589431762695, | |
| "learning_rate": 0.00016836243614250113, | |
| "loss": 2.2508, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.06594291970953714, | |
| "grad_norm": 5.97781229019165, | |
| "learning_rate": 0.00016820840233706719, | |
| "loss": 2.4173, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.06610819018499964, | |
| "grad_norm": 11.736589431762695, | |
| "learning_rate": 0.0001680540653066891, | |
| "loss": 3.0601, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.06627346066046214, | |
| "grad_norm": 1.0298711061477661, | |
| "learning_rate": 0.00016789942573748232, | |
| "loss": 1.2516, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.06643873113592463, | |
| "grad_norm": 1.0405280590057373, | |
| "learning_rate": 0.0001677444843169072, | |
| "loss": 1.3653, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.06660400161138713, | |
| "grad_norm": 1.1241669654846191, | |
| "learning_rate": 0.00016758924173376603, | |
| "loss": 1.7322, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.06676927208684963, | |
| "grad_norm": 1.003084421157837, | |
| "learning_rate": 0.0001674336986781999, | |
| "loss": 2.481, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.06693454256231213, | |
| "grad_norm": 0.8039822578430176, | |
| "learning_rate": 0.00016727785584168581, | |
| "loss": 2.4494, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.06709981303777464, | |
| "grad_norm": 0.9330201745033264, | |
| "learning_rate": 0.0001671217139170333, | |
| "loss": 2.2713, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.06726508351323714, | |
| "grad_norm": 0.9125016331672668, | |
| "learning_rate": 0.00016696527359838154, | |
| "loss": 2.2348, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.06743035398869963, | |
| "grad_norm": 0.9843574166297913, | |
| "learning_rate": 0.00016680853558119632, | |
| "loss": 2.231, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.06759562446416213, | |
| "grad_norm": 0.9823955297470093, | |
| "learning_rate": 0.0001666515005622668, | |
| "loss": 2.0948, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.06776089493962463, | |
| "grad_norm": 1.0274838209152222, | |
| "learning_rate": 0.0001664941692397025, | |
| "loss": 2.0928, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.06792616541508713, | |
| "grad_norm": 0.8992162346839905, | |
| "learning_rate": 0.00016633654231293013, | |
| "loss": 1.9845, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.06809143589054963, | |
| "grad_norm": 1.0962395668029785, | |
| "learning_rate": 0.00016617862048269065, | |
| "loss": 2.3938, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.06825670636601212, | |
| "grad_norm": 1.1849541664123535, | |
| "learning_rate": 0.00016602040445103588, | |
| "loss": 2.238, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.06842197684147462, | |
| "grad_norm": 1.0153218507766724, | |
| "learning_rate": 0.00016586189492132566, | |
| "loss": 2.103, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.06858724731693712, | |
| "grad_norm": 1.0640662908554077, | |
| "learning_rate": 0.00016570309259822453, | |
| "loss": 2.2117, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.06875251779239962, | |
| "grad_norm": 1.1224597692489624, | |
| "learning_rate": 0.0001655439981876987, | |
| "loss": 2.119, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.06891778826786213, | |
| "grad_norm": 1.2191041707992554, | |
| "learning_rate": 0.00016538461239701277, | |
| "loss": 2.4141, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.06908305874332463, | |
| "grad_norm": 1.1894046068191528, | |
| "learning_rate": 0.00016522493593472683, | |
| "loss": 1.8937, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.06924832921878712, | |
| "grad_norm": 1.3787667751312256, | |
| "learning_rate": 0.0001650649695106931, | |
| "loss": 2.3488, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.06941359969424962, | |
| "grad_norm": 1.3866989612579346, | |
| "learning_rate": 0.00016490471383605288, | |
| "loss": 2.3129, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.06957887016971212, | |
| "grad_norm": 1.3885571956634521, | |
| "learning_rate": 0.00016474416962323325, | |
| "loss": 2.0808, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.06974414064517462, | |
| "grad_norm": 1.5631028413772583, | |
| "learning_rate": 0.00016458333758594414, | |
| "loss": 2.2406, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.06990941112063712, | |
| "grad_norm": 1.5918340682983398, | |
| "learning_rate": 0.00016442221843917496, | |
| "loss": 2.6892, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.07007468159609961, | |
| "grad_norm": 1.5709396600723267, | |
| "learning_rate": 0.00016426081289919143, | |
| "loss": 2.2324, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.07023995207156211, | |
| "grad_norm": 1.850425124168396, | |
| "learning_rate": 0.0001640991216835326, | |
| "loss": 2.5997, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.07040522254702461, | |
| "grad_norm": 1.835813045501709, | |
| "learning_rate": 0.00016393714551100734, | |
| "loss": 2.984, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.07057049302248711, | |
| "grad_norm": 1.857850432395935, | |
| "learning_rate": 0.0001637748851016914, | |
| "loss": 2.5182, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.07073576349794962, | |
| "grad_norm": 1.8345370292663574, | |
| "learning_rate": 0.00016361234117692413, | |
| "loss": 2.1213, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.07090103397341212, | |
| "grad_norm": 2.125157594680786, | |
| "learning_rate": 0.00016344951445930526, | |
| "loss": 2.7516, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.07106630444887461, | |
| "grad_norm": 2.0694260597229004, | |
| "learning_rate": 0.0001632864056726917, | |
| "loss": 2.7459, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.07123157492433711, | |
| "grad_norm": 2.2040746212005615, | |
| "learning_rate": 0.00016312301554219426, | |
| "loss": 2.6295, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.07139684539979961, | |
| "grad_norm": 2.372509002685547, | |
| "learning_rate": 0.00016295934479417453, | |
| "loss": 2.4616, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.07156211587526211, | |
| "grad_norm": 2.6924386024475098, | |
| "learning_rate": 0.00016279539415624164, | |
| "loss": 2.7477, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.0717273863507246, | |
| "grad_norm": 2.440096855163574, | |
| "learning_rate": 0.0001626311643572489, | |
| "loss": 2.5412, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.0718926568261871, | |
| "grad_norm": 2.819272756576538, | |
| "learning_rate": 0.00016246665612729074, | |
| "loss": 2.6312, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.0720579273016496, | |
| "grad_norm": 2.85014009475708, | |
| "learning_rate": 0.00016230187019769928, | |
| "loss": 2.4503, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.0722231977771121, | |
| "grad_norm": 3.466008186340332, | |
| "learning_rate": 0.00016213680730104124, | |
| "loss": 2.7176, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.0723884682525746, | |
| "grad_norm": 3.0699470043182373, | |
| "learning_rate": 0.0001619714681711146, | |
| "loss": 2.5846, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.07255373872803711, | |
| "grad_norm": 3.049323320388794, | |
| "learning_rate": 0.00016180585354294536, | |
| "loss": 2.267, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.07271900920349961, | |
| "grad_norm": 3.0568227767944336, | |
| "learning_rate": 0.00016163996415278424, | |
| "loss": 2.3126, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.0728842796789621, | |
| "grad_norm": 3.6018261909484863, | |
| "learning_rate": 0.00016147380073810346, | |
| "loss": 2.771, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.0730495501544246, | |
| "grad_norm": 3.663275718688965, | |
| "learning_rate": 0.0001613073640375934, | |
| "loss": 2.744, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.0732148206298871, | |
| "grad_norm": 3.635608434677124, | |
| "learning_rate": 0.00016114065479115946, | |
| "loss": 2.3742, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.0733800911053496, | |
| "grad_norm": 4.326128005981445, | |
| "learning_rate": 0.00016097367373991842, | |
| "loss": 2.2298, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.0735453615808121, | |
| "grad_norm": 4.141120433807373, | |
| "learning_rate": 0.00016080642162619565, | |
| "loss": 1.8221, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.0737106320562746, | |
| "grad_norm": 4.418088436126709, | |
| "learning_rate": 0.0001606388991935214, | |
| "loss": 2.3042, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.07387590253173709, | |
| "grad_norm": 3.9933526515960693, | |
| "learning_rate": 0.0001604711071866277, | |
| "loss": 1.7728, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.07404117300719959, | |
| "grad_norm": 5.002580165863037, | |
| "learning_rate": 0.00016030304635144494, | |
| "loss": 2.19, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.07420644348266209, | |
| "grad_norm": 7.466914176940918, | |
| "learning_rate": 0.00016013471743509862, | |
| "loss": 2.7615, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.0743717139581246, | |
| "grad_norm": 10.264341354370117, | |
| "learning_rate": 0.00015996612118590603, | |
| "loss": 3.5717, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.0745369844335871, | |
| "grad_norm": 1.3243449926376343, | |
| "learning_rate": 0.00015979725835337294, | |
| "loss": 1.2435, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.0747022549090496, | |
| "grad_norm": 0.7879775762557983, | |
| "learning_rate": 0.00015962812968819016, | |
| "loss": 1.2245, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.07486752538451209, | |
| "grad_norm": 0.7144224047660828, | |
| "learning_rate": 0.0001594587359422303, | |
| "loss": 1.3302, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.07503279585997459, | |
| "grad_norm": 1.023307204246521, | |
| "learning_rate": 0.0001592890778685444, | |
| "loss": 2.2432, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.07519806633543709, | |
| "grad_norm": 1.2127289772033691, | |
| "learning_rate": 0.00015911915622135862, | |
| "loss": 2.7845, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.07536333681089959, | |
| "grad_norm": 0.8183092474937439, | |
| "learning_rate": 0.00015894897175607086, | |
| "loss": 2.3426, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.07552860728636208, | |
| "grad_norm": 0.8497331142425537, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 2.4256, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.07569387776182458, | |
| "grad_norm": 0.7920719385147095, | |
| "learning_rate": 0.00015860781739861928, | |
| "loss": 2.0701, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.07585914823728708, | |
| "grad_norm": 0.8244463801383972, | |
| "learning_rate": 0.00015843684902307962, | |
| "loss": 2.1564, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.07602441871274958, | |
| "grad_norm": 0.8627625703811646, | |
| "learning_rate": 0.00015826562086267956, | |
| "loss": 2.4051, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.07618968918821209, | |
| "grad_norm": 0.8703142404556274, | |
| "learning_rate": 0.00015809413367862512, | |
| "loss": 1.9225, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.07635495966367459, | |
| "grad_norm": 0.9556866884231567, | |
| "learning_rate": 0.00015792238823327388, | |
| "loss": 2.1445, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.07652023013913709, | |
| "grad_norm": 0.9048780202865601, | |
| "learning_rate": 0.00015775038529013152, | |
| "loss": 2.1131, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.07668550061459958, | |
| "grad_norm": 1.0929898023605347, | |
| "learning_rate": 0.0001575781256138485, | |
| "loss": 2.3259, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.07685077109006208, | |
| "grad_norm": 1.1149414777755737, | |
| "learning_rate": 0.00015740560997021648, | |
| "loss": 2.3243, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.07701604156552458, | |
| "grad_norm": 1.0898072719573975, | |
| "learning_rate": 0.00015723283912616513, | |
| "loss": 2.3052, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.07718131204098708, | |
| "grad_norm": 1.1040949821472168, | |
| "learning_rate": 0.00015705981384975866, | |
| "loss": 2.3322, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.07734658251644957, | |
| "grad_norm": 1.4544291496276855, | |
| "learning_rate": 0.0001568865349101923, | |
| "loss": 2.1372, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.07751185299191207, | |
| "grad_norm": 1.4453994035720825, | |
| "learning_rate": 0.00015671300307778898, | |
| "loss": 2.3296, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.07767712346737457, | |
| "grad_norm": 1.388993740081787, | |
| "learning_rate": 0.00015653921912399589, | |
| "loss": 2.2287, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.07784239394283707, | |
| "grad_norm": 1.52480947971344, | |
| "learning_rate": 0.00015636518382138107, | |
| "loss": 2.5048, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.07800766441829958, | |
| "grad_norm": 1.651181697845459, | |
| "learning_rate": 0.0001561908979436299, | |
| "loss": 2.1369, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.07817293489376208, | |
| "grad_norm": 1.742471694946289, | |
| "learning_rate": 0.00015601636226554168, | |
| "loss": 2.2678, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.07833820536922458, | |
| "grad_norm": 1.6204532384872437, | |
| "learning_rate": 0.00015584157756302634, | |
| "loss": 1.7443, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.07850347584468707, | |
| "grad_norm": 1.8165796995162964, | |
| "learning_rate": 0.0001556665446131007, | |
| "loss": 2.312, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.07866874632014957, | |
| "grad_norm": 2.0851690769195557, | |
| "learning_rate": 0.00015549126419388536, | |
| "loss": 2.7382, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.07883401679561207, | |
| "grad_norm": 2.0037810802459717, | |
| "learning_rate": 0.0001553157370846009, | |
| "loss": 2.3406, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.07899928727107457, | |
| "grad_norm": 1.9187804460525513, | |
| "learning_rate": 0.00015513996406556465, | |
| "loss": 2.765, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.07916455774653706, | |
| "grad_norm": 2.230612277984619, | |
| "learning_rate": 0.00015496394591818716, | |
| "loss": 2.7399, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.07932982822199956, | |
| "grad_norm": 2.404221773147583, | |
| "learning_rate": 0.0001547876834249687, | |
| "loss": 2.8219, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.07949509869746206, | |
| "grad_norm": 2.461216926574707, | |
| "learning_rate": 0.00015461117736949577, | |
| "loss": 2.6265, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.07966036917292456, | |
| "grad_norm": 2.6747286319732666, | |
| "learning_rate": 0.00015443442853643762, | |
| "loss": 2.5295, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.07982563964838707, | |
| "grad_norm": 2.639806032180786, | |
| "learning_rate": 0.00015425743771154294, | |
| "loss": 2.2147, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.07999091012384957, | |
| "grad_norm": 2.8985249996185303, | |
| "learning_rate": 0.00015408020568163602, | |
| "loss": 2.6202, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.08015618059931207, | |
| "grad_norm": 2.7747952938079834, | |
| "learning_rate": 0.00015390273323461352, | |
| "loss": 2.0239, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.08032145107477456, | |
| "grad_norm": 2.8715693950653076, | |
| "learning_rate": 0.0001537250211594409, | |
| "loss": 2.5015, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.08048672155023706, | |
| "grad_norm": 3.0101988315582275, | |
| "learning_rate": 0.0001535470702461489, | |
| "loss": 2.4754, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.08065199202569956, | |
| "grad_norm": 3.6238367557525635, | |
| "learning_rate": 0.00015336888128583, | |
| "loss": 2.5276, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.08081726250116206, | |
| "grad_norm": 2.935698986053467, | |
| "learning_rate": 0.000153190455070635, | |
| "loss": 2.3233, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.08098253297662455, | |
| "grad_norm": 3.3008511066436768, | |
| "learning_rate": 0.00015301179239376938, | |
| "loss": 2.5714, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.08114780345208705, | |
| "grad_norm": 3.821829080581665, | |
| "learning_rate": 0.00015283289404948976, | |
| "loss": 2.3643, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.08131307392754955, | |
| "grad_norm": 4.2104692459106445, | |
| "learning_rate": 0.0001526537608331006, | |
| "loss": 2.4445, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.08147834440301205, | |
| "grad_norm": 3.4621777534484863, | |
| "learning_rate": 0.00015247439354095041, | |
| "loss": 1.6488, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.08164361487847456, | |
| "grad_norm": 3.889923095703125, | |
| "learning_rate": 0.00015229479297042823, | |
| "loss": 2.0852, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.08180888535393706, | |
| "grad_norm": 4.294791221618652, | |
| "learning_rate": 0.00015211495991996027, | |
| "loss": 2.2366, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.08197415582939956, | |
| "grad_norm": 5.829231262207031, | |
| "learning_rate": 0.0001519348951890062, | |
| "loss": 2.1231, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.08213942630486205, | |
| "grad_norm": 5.648350715637207, | |
| "learning_rate": 0.0001517545995780556, | |
| "loss": 2.2986, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.08230469678032455, | |
| "grad_norm": 7.411054611206055, | |
| "learning_rate": 0.00015157407388862452, | |
| "loss": 3.1934, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.08246996725578705, | |
| "grad_norm": 6.661128520965576, | |
| "learning_rate": 0.00015139331892325179, | |
| "loss": 2.3472, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.08263523773124955, | |
| "grad_norm": 9.719291687011719, | |
| "learning_rate": 0.0001512123354854955, | |
| "loss": 2.4928, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.08280050820671205, | |
| "grad_norm": 1.3126555681228638, | |
| "learning_rate": 0.0001510311243799295, | |
| "loss": 1.2044, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.08296577868217454, | |
| "grad_norm": 0.6947880387306213, | |
| "learning_rate": 0.00015084968641213958, | |
| "loss": 1.3913, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 0.08313104915763704, | |
| "grad_norm": 0.7388302683830261, | |
| "learning_rate": 0.00015066802238872023, | |
| "loss": 1.7062, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 0.08329631963309954, | |
| "grad_norm": 0.7357963919639587, | |
| "learning_rate": 0.0001504861331172709, | |
| "loss": 1.6154, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 0.08346159010856205, | |
| "grad_norm": 0.958837628364563, | |
| "learning_rate": 0.0001503040194063922, | |
| "loss": 2.5589, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.08362686058402455, | |
| "grad_norm": 0.8261339068412781, | |
| "learning_rate": 0.00015012168206568268, | |
| "loss": 2.239, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 0.08379213105948705, | |
| "grad_norm": 0.675197184085846, | |
| "learning_rate": 0.00014993912190573505, | |
| "loss": 1.8415, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 0.08395740153494954, | |
| "grad_norm": 0.773586094379425, | |
| "learning_rate": 0.00014975633973813242, | |
| "loss": 2.225, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 0.08412267201041204, | |
| "grad_norm": 0.8245728015899658, | |
| "learning_rate": 0.00014957333637544503, | |
| "loss": 2.1888, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 0.08428794248587454, | |
| "grad_norm": 0.9109384417533875, | |
| "learning_rate": 0.00014939011263122634, | |
| "loss": 2.5024, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.08445321296133704, | |
| "grad_norm": 0.8955804109573364, | |
| "learning_rate": 0.0001492066693200096, | |
| "loss": 2.2977, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 0.08461848343679954, | |
| "grad_norm": 0.9247772693634033, | |
| "learning_rate": 0.00014902300725730413, | |
| "loss": 2.0606, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 0.08478375391226203, | |
| "grad_norm": 0.8497363924980164, | |
| "learning_rate": 0.00014883912725959167, | |
| "loss": 1.7316, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 0.08494902438772453, | |
| "grad_norm": 1.025578260421753, | |
| "learning_rate": 0.00014865503014432292, | |
| "loss": 1.9945, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 0.08511429486318703, | |
| "grad_norm": 0.9977094531059265, | |
| "learning_rate": 0.00014847071672991367, | |
| "loss": 2.0645, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.08527956533864953, | |
| "grad_norm": 1.045988917350769, | |
| "learning_rate": 0.0001482861878357414, | |
| "loss": 2.0204, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 0.08544483581411204, | |
| "grad_norm": 1.1506321430206299, | |
| "learning_rate": 0.00014810144428214144, | |
| "loss": 2.297, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 0.08561010628957454, | |
| "grad_norm": 1.3619967699050903, | |
| "learning_rate": 0.0001479164868904034, | |
| "loss": 2.4022, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 0.08577537676503703, | |
| "grad_norm": 1.3217016458511353, | |
| "learning_rate": 0.00014773131648276758, | |
| "loss": 2.0652, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 0.08594064724049953, | |
| "grad_norm": 1.31471586227417, | |
| "learning_rate": 0.00014754593388242117, | |
| "loss": 2.5536, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.08610591771596203, | |
| "grad_norm": 1.401208758354187, | |
| "learning_rate": 0.0001473603399134948, | |
| "loss": 2.4343, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 0.08627118819142453, | |
| "grad_norm": 1.359070897102356, | |
| "learning_rate": 0.0001471745354010586, | |
| "loss": 2.3605, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 0.08643645866688703, | |
| "grad_norm": 1.6402932405471802, | |
| "learning_rate": 0.00014698852117111884, | |
| "loss": 2.1437, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 0.08660172914234952, | |
| "grad_norm": 1.6058882474899292, | |
| "learning_rate": 0.000146802298050614, | |
| "loss": 2.5985, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 0.08676699961781202, | |
| "grad_norm": 1.6456177234649658, | |
| "learning_rate": 0.0001466158668674112, | |
| "loss": 2.5085, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.08693227009327452, | |
| "grad_norm": 1.8821338415145874, | |
| "learning_rate": 0.00014642922845030257, | |
| "loss": 2.3757, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 0.08709754056873702, | |
| "grad_norm": 1.9960278272628784, | |
| "learning_rate": 0.0001462423836290015, | |
| "loss": 2.4533, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 0.08726281104419953, | |
| "grad_norm": 1.989198923110962, | |
| "learning_rate": 0.00014605533323413887, | |
| "loss": 2.7163, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 0.08742808151966203, | |
| "grad_norm": 1.9564366340637207, | |
| "learning_rate": 0.00014586807809725962, | |
| "loss": 2.4972, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 0.08759335199512452, | |
| "grad_norm": 2.6321208477020264, | |
| "learning_rate": 0.00014568061905081875, | |
| "loss": 2.5209, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.08775862247058702, | |
| "grad_norm": 2.20314621925354, | |
| "learning_rate": 0.00014549295692817778, | |
| "loss": 2.3181, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 0.08792389294604952, | |
| "grad_norm": 2.3064544200897217, | |
| "learning_rate": 0.00014530509256360102, | |
| "loss": 2.0767, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 0.08808916342151202, | |
| "grad_norm": 2.5148324966430664, | |
| "learning_rate": 0.00014511702679225193, | |
| "loss": 2.7805, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 0.08825443389697452, | |
| "grad_norm": 2.238633632659912, | |
| "learning_rate": 0.0001449287604501893, | |
| "loss": 2.2732, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 0.08841970437243701, | |
| "grad_norm": 2.92490291595459, | |
| "learning_rate": 0.00014474029437436348, | |
| "loss": 2.2284, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.08858497484789951, | |
| "grad_norm": 2.831331729888916, | |
| "learning_rate": 0.00014455162940261285, | |
| "loss": 2.432, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 0.08875024532336201, | |
| "grad_norm": 2.8920512199401855, | |
| "learning_rate": 0.0001443627663736599, | |
| "loss": 2.3652, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 0.08891551579882451, | |
| "grad_norm": 3.0407826900482178, | |
| "learning_rate": 0.00014417370612710778, | |
| "loss": 2.2523, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 0.08908078627428702, | |
| "grad_norm": 3.4025917053222656, | |
| "learning_rate": 0.00014398444950343623, | |
| "loss": 2.463, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 0.08924605674974952, | |
| "grad_norm": 3.815530776977539, | |
| "learning_rate": 0.00014379499734399798, | |
| "loss": 2.5442, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.08941132722521201, | |
| "grad_norm": 3.569391965866089, | |
| "learning_rate": 0.0001436053504910151, | |
| "loss": 2.5601, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 0.08957659770067451, | |
| "grad_norm": 3.856222152709961, | |
| "learning_rate": 0.0001434155097875752, | |
| "loss": 2.659, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 0.08974186817613701, | |
| "grad_norm": 4.079010486602783, | |
| "learning_rate": 0.00014322547607762762, | |
| "loss": 2.0967, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 0.08990713865159951, | |
| "grad_norm": 3.778975248336792, | |
| "learning_rate": 0.0001430352502059797, | |
| "loss": 2.0831, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 0.090072409127062, | |
| "grad_norm": 4.952725887298584, | |
| "learning_rate": 0.0001428448330182931, | |
| "loss": 2.2062, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.0902376796025245, | |
| "grad_norm": 4.95838737487793, | |
| "learning_rate": 0.00014265422536107993, | |
| "loss": 2.3836, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 0.090402950077987, | |
| "grad_norm": 4.887285232543945, | |
| "learning_rate": 0.00014246342808169914, | |
| "loss": 1.9324, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 0.0905682205534495, | |
| "grad_norm": 5.759344577789307, | |
| "learning_rate": 0.00014227244202835257, | |
| "loss": 2.4722, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 0.090733491028912, | |
| "grad_norm": 7.185070514678955, | |
| "learning_rate": 0.0001420812680500813, | |
| "loss": 2.9153, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 0.09089876150437451, | |
| "grad_norm": 9.65413761138916, | |
| "learning_rate": 0.00014188990699676184, | |
| "loss": 3.0405, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.091064031979837, | |
| "grad_norm": 1.4314805269241333, | |
| "learning_rate": 0.00014169835971910238, | |
| "loss": 1.1863, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 0.0912293024552995, | |
| "grad_norm": 0.7541770935058594, | |
| "learning_rate": 0.0001415066270686389, | |
| "loss": 1.4154, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 0.091394572930762, | |
| "grad_norm": 0.6706185936927795, | |
| "learning_rate": 0.00014131470989773158, | |
| "loss": 1.4615, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 0.0915598434062245, | |
| "grad_norm": 0.6045027375221252, | |
| "learning_rate": 0.0001411226090595608, | |
| "loss": 1.7779, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 0.091725113881687, | |
| "grad_norm": 0.8165815472602844, | |
| "learning_rate": 0.00014093032540812348, | |
| "loss": 2.3638, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.0918903843571495, | |
| "grad_norm": 0.7140501141548157, | |
| "learning_rate": 0.0001407378597982293, | |
| "loss": 2.1641, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 0.092055654832612, | |
| "grad_norm": 0.8131698369979858, | |
| "learning_rate": 0.00014054521308549673, | |
| "loss": 2.2842, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 0.09222092530807449, | |
| "grad_norm": 0.7364957332611084, | |
| "learning_rate": 0.0001403523861263495, | |
| "loss": 1.84, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 0.09238619578353699, | |
| "grad_norm": 0.9326314926147461, | |
| "learning_rate": 0.00014015937977801256, | |
| "loss": 2.1904, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 0.09255146625899949, | |
| "grad_norm": 0.8098871111869812, | |
| "learning_rate": 0.00013996619489850822, | |
| "loss": 2.1801, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.092716736734462, | |
| "grad_norm": 0.9290388822555542, | |
| "learning_rate": 0.00013977283234665273, | |
| "loss": 2.1772, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 0.0928820072099245, | |
| "grad_norm": 0.9413136839866638, | |
| "learning_rate": 0.00013957929298205195, | |
| "loss": 2.0678, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 0.093047277685387, | |
| "grad_norm": 1.0153107643127441, | |
| "learning_rate": 0.00013938557766509792, | |
| "loss": 1.9317, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 0.09321254816084949, | |
| "grad_norm": 1.027348279953003, | |
| "learning_rate": 0.0001391916872569648, | |
| "loss": 1.9527, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 0.09337781863631199, | |
| "grad_norm": 1.1003931760787964, | |
| "learning_rate": 0.00013899762261960518, | |
| "loss": 2.2167, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.09354308911177449, | |
| "grad_norm": 1.2371317148208618, | |
| "learning_rate": 0.0001388033846157462, | |
| "loss": 2.3379, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 0.09370835958723699, | |
| "grad_norm": 1.2197039127349854, | |
| "learning_rate": 0.0001386089741088857, | |
| "loss": 2.322, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 0.09387363006269948, | |
| "grad_norm": 1.2260133028030396, | |
| "learning_rate": 0.00013841439196328836, | |
| "loss": 2.0632, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 0.09403890053816198, | |
| "grad_norm": 1.435096025466919, | |
| "learning_rate": 0.00013821963904398193, | |
| "loss": 2.2846, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 0.09420417101362448, | |
| "grad_norm": 1.4923068284988403, | |
| "learning_rate": 0.00013802471621675338, | |
| "loss": 2.2572, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.09436944148908698, | |
| "grad_norm": 1.4718787670135498, | |
| "learning_rate": 0.00013782962434814492, | |
| "loss": 1.9984, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 0.09453471196454949, | |
| "grad_norm": 1.7288150787353516, | |
| "learning_rate": 0.00013763436430545034, | |
| "loss": 2.3539, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 0.09469998244001199, | |
| "grad_norm": 1.6094728708267212, | |
| "learning_rate": 0.00013743893695671096, | |
| "loss": 2.326, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 0.09486525291547449, | |
| "grad_norm": 1.6949067115783691, | |
| "learning_rate": 0.00013724334317071198, | |
| "loss": 2.3055, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 0.09503052339093698, | |
| "grad_norm": 1.8237234354019165, | |
| "learning_rate": 0.00013704758381697844, | |
| "loss": 2.3343, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.09519579386639948, | |
| "grad_norm": 1.8844832181930542, | |
| "learning_rate": 0.00013685165976577146, | |
| "loss": 2.7004, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 0.09536106434186198, | |
| "grad_norm": 2.0341548919677734, | |
| "learning_rate": 0.0001366555718880843, | |
| "loss": 2.2445, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 0.09552633481732448, | |
| "grad_norm": 2.281798839569092, | |
| "learning_rate": 0.00013645932105563844, | |
| "loss": 2.2766, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 0.09569160529278697, | |
| "grad_norm": 1.9859449863433838, | |
| "learning_rate": 0.00013626290814088005, | |
| "loss": 2.4107, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 0.09585687576824947, | |
| "grad_norm": 2.285785436630249, | |
| "learning_rate": 0.00013606633401697557, | |
| "loss": 2.5148, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.09602214624371197, | |
| "grad_norm": 2.3296151161193848, | |
| "learning_rate": 0.00013586959955780824, | |
| "loss": 2.8249, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 0.09618741671917447, | |
| "grad_norm": 2.6428415775299072, | |
| "learning_rate": 0.00013567270563797398, | |
| "loss": 2.1186, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 0.09635268719463698, | |
| "grad_norm": 2.6988117694854736, | |
| "learning_rate": 0.00013547565313277776, | |
| "loss": 2.3376, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 0.09651795767009948, | |
| "grad_norm": 2.885573148727417, | |
| "learning_rate": 0.00013527844291822948, | |
| "loss": 2.9902, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 0.09668322814556198, | |
| "grad_norm": 3.032674551010132, | |
| "learning_rate": 0.0001350810758710401, | |
| "loss": 2.6686, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.09684849862102447, | |
| "grad_norm": 2.723219394683838, | |
| "learning_rate": 0.00013488355286861783, | |
| "loss": 2.3772, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 0.09701376909648697, | |
| "grad_norm": 3.236877202987671, | |
| "learning_rate": 0.0001346858747890642, | |
| "loss": 2.5389, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 0.09717903957194947, | |
| "grad_norm": 3.7379932403564453, | |
| "learning_rate": 0.00013448804251117003, | |
| "loss": 2.3541, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 0.09734431004741197, | |
| "grad_norm": 3.417888641357422, | |
| "learning_rate": 0.0001342900569144119, | |
| "loss": 2.8407, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 0.09750958052287446, | |
| "grad_norm": 3.191343307495117, | |
| "learning_rate": 0.0001340919188789477, | |
| "loss": 2.4034, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.09767485099833696, | |
| "grad_norm": 3.7207911014556885, | |
| "learning_rate": 0.00013389362928561317, | |
| "loss": 2.4484, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 0.09784012147379946, | |
| "grad_norm": 3.920457601547241, | |
| "learning_rate": 0.00013369518901591772, | |
| "loss": 2.5659, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 0.09800539194926196, | |
| "grad_norm": 4.423088550567627, | |
| "learning_rate": 0.00013349659895204067, | |
| "loss": 2.0712, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 0.09817066242472447, | |
| "grad_norm": 4.407251834869385, | |
| "learning_rate": 0.0001332978599768272, | |
| "loss": 2.3152, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 0.09833593290018697, | |
| "grad_norm": 4.617842197418213, | |
| "learning_rate": 0.00013309897297378455, | |
| "loss": 2.1745, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.09850120337564947, | |
| "grad_norm": 4.762388706207275, | |
| "learning_rate": 0.00013289993882707797, | |
| "loss": 2.3794, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 0.09866647385111196, | |
| "grad_norm": 5.113027095794678, | |
| "learning_rate": 0.00013270075842152678, | |
| "loss": 2.6819, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 0.09883174432657446, | |
| "grad_norm": 5.677262306213379, | |
| "learning_rate": 0.00013250143264260074, | |
| "loss": 2.497, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 0.09899701480203696, | |
| "grad_norm": 6.53700065612793, | |
| "learning_rate": 0.0001323019623764156, | |
| "loss": 2.8031, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 0.09916228527749946, | |
| "grad_norm": 7.789911270141602, | |
| "learning_rate": 0.00013210234850972964, | |
| "loss": 2.0391, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.09932755575296195, | |
| "grad_norm": 1.0732189416885376, | |
| "learning_rate": 0.0001319025919299394, | |
| "loss": 1.3476, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 0.09949282622842445, | |
| "grad_norm": 0.7092403173446655, | |
| "learning_rate": 0.00013170269352507597, | |
| "loss": 1.3001, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 0.09965809670388695, | |
| "grad_norm": 0.6339560151100159, | |
| "learning_rate": 0.0001315026541838008, | |
| "loss": 1.2912, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 0.09982336717934945, | |
| "grad_norm": 0.8303507566452026, | |
| "learning_rate": 0.00013130247479540202, | |
| "loss": 1.7372, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 0.09998863765481196, | |
| "grad_norm": 0.8783502578735352, | |
| "learning_rate": 0.00013110215624979025, | |
| "loss": 2.5962, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.10015390813027446, | |
| "grad_norm": 0.9190163612365723, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 2.2247, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 0.10031917860573696, | |
| "grad_norm": 0.8587096929550171, | |
| "learning_rate": 0.00013070110524965954, | |
| "loss": 2.5056, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 0.10048444908119945, | |
| "grad_norm": 0.898530125617981, | |
| "learning_rate": 0.00013050037457803924, | |
| "loss": 2.1565, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 0.10064971955666195, | |
| "grad_norm": 1.0313509702682495, | |
| "learning_rate": 0.0001302995083149953, | |
| "loss": 2.1435, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 0.10081499003212445, | |
| "grad_norm": 0.984456479549408, | |
| "learning_rate": 0.0001300985073534919, | |
| "loss": 2.0734, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.10098026050758695, | |
| "grad_norm": 0.9210591912269592, | |
| "learning_rate": 0.00012989737258709203, | |
| "loss": 2.0853, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 0.10114553098304944, | |
| "grad_norm": 0.8905264735221863, | |
| "learning_rate": 0.00012969610490995358, | |
| "loss": 2.0526, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 0.10131080145851194, | |
| "grad_norm": 0.9354328513145447, | |
| "learning_rate": 0.00012949470521682528, | |
| "loss": 2.0569, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 0.10147607193397444, | |
| "grad_norm": 1.0619162321090698, | |
| "learning_rate": 0.0001292931744030427, | |
| "loss": 2.4487, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 0.10164134240943694, | |
| "grad_norm": 1.0751473903656006, | |
| "learning_rate": 0.0001290915133645243, | |
| "loss": 1.8786, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.10180661288489945, | |
| "grad_norm": 1.222031593322754, | |
| "learning_rate": 0.00012888972299776754, | |
| "loss": 1.9611, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 0.10197188336036195, | |
| "grad_norm": 1.0904767513275146, | |
| "learning_rate": 0.00012868780419984482, | |
| "loss": 2.1155, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 0.10213715383582445, | |
| "grad_norm": 1.478400468826294, | |
| "learning_rate": 0.00012848575786839943, | |
| "loss": 2.4205, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 0.10230242431128694, | |
| "grad_norm": 1.3960856199264526, | |
| "learning_rate": 0.0001282835849016416, | |
| "loss": 2.3638, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 0.10246769478674944, | |
| "grad_norm": 1.3682273626327515, | |
| "learning_rate": 0.00012808128619834461, | |
| "loss": 2.3077, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.10263296526221194, | |
| "grad_norm": 1.4534392356872559, | |
| "learning_rate": 0.0001278788626578407, | |
| "loss": 2.0747, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 0.10279823573767444, | |
| "grad_norm": 1.5964767932891846, | |
| "learning_rate": 0.00012767631518001698, | |
| "loss": 2.7049, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 0.10296350621313693, | |
| "grad_norm": 1.697447657585144, | |
| "learning_rate": 0.00012747364466531163, | |
| "loss": 2.2763, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 0.10312877668859943, | |
| "grad_norm": 1.8660848140716553, | |
| "learning_rate": 0.00012727085201470973, | |
| "loss": 2.3084, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 0.10329404716406193, | |
| "grad_norm": 2.07235050201416, | |
| "learning_rate": 0.00012706793812973941, | |
| "loss": 2.4683, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.10345931763952443, | |
| "grad_norm": 1.8157964944839478, | |
| "learning_rate": 0.0001268649039124677, | |
| "loss": 2.9693, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 0.10362458811498694, | |
| "grad_norm": 2.076352834701538, | |
| "learning_rate": 0.00012666175026549662, | |
| "loss": 2.8005, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 0.10378985859044944, | |
| "grad_norm": 2.222311496734619, | |
| "learning_rate": 0.000126458478091959, | |
| "loss": 2.9366, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 0.10395512906591194, | |
| "grad_norm": 2.3791983127593994, | |
| "learning_rate": 0.00012625508829551473, | |
| "loss": 2.5959, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 0.10412039954137443, | |
| "grad_norm": 2.378978967666626, | |
| "learning_rate": 0.00012605158178034654, | |
| "loss": 2.6831, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.10428567001683693, | |
| "grad_norm": 2.3791182041168213, | |
| "learning_rate": 0.00012584795945115603, | |
| "loss": 2.1217, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 0.10445094049229943, | |
| "grad_norm": 2.5949718952178955, | |
| "learning_rate": 0.0001256442222131597, | |
| "loss": 2.1992, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 0.10461621096776193, | |
| "grad_norm": 2.604978561401367, | |
| "learning_rate": 0.0001254403709720848, | |
| "loss": 2.7245, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 0.10478148144322443, | |
| "grad_norm": 2.956751585006714, | |
| "learning_rate": 0.0001252364066341655, | |
| "loss": 2.7992, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 0.10494675191868692, | |
| "grad_norm": 3.0405449867248535, | |
| "learning_rate": 0.00012503233010613865, | |
| "loss": 2.3531, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.10511202239414942, | |
| "grad_norm": 3.1233065128326416, | |
| "learning_rate": 0.00012482814229523997, | |
| "loss": 2.6056, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 0.10527729286961192, | |
| "grad_norm": 3.64660906791687, | |
| "learning_rate": 0.00012462384410919975, | |
| "loss": 2.8294, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 0.10544256334507442, | |
| "grad_norm": 2.998234272003174, | |
| "learning_rate": 0.00012441943645623903, | |
| "loss": 2.0387, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 0.10560783382053693, | |
| "grad_norm": 3.6462655067443848, | |
| "learning_rate": 0.00012421492024506555, | |
| "loss": 2.5703, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 0.10577310429599943, | |
| "grad_norm": 3.6429123878479004, | |
| "learning_rate": 0.00012401029638486953, | |
| "loss": 2.4506, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.10593837477146192, | |
| "grad_norm": 4.139918804168701, | |
| "learning_rate": 0.0001238055657853198, | |
| "loss": 2.5996, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 0.10610364524692442, | |
| "grad_norm": 3.9799206256866455, | |
| "learning_rate": 0.00012360072935655982, | |
| "loss": 2.5859, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 0.10626891572238692, | |
| "grad_norm": 4.634383678436279, | |
| "learning_rate": 0.00012339578800920332, | |
| "loss": 1.7455, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 0.10643418619784942, | |
| "grad_norm": 4.632512092590332, | |
| "learning_rate": 0.00012319074265433063, | |
| "loss": 2.4171, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 0.10659945667331192, | |
| "grad_norm": 5.244453430175781, | |
| "learning_rate": 0.00012298559420348437, | |
| "loss": 2.2382, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.10676472714877441, | |
| "grad_norm": 4.456907272338867, | |
| "learning_rate": 0.00012278034356866545, | |
| "loss": 2.1996, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 0.10692999762423691, | |
| "grad_norm": 6.22163724899292, | |
| "learning_rate": 0.00012257499166232907, | |
| "loss": 2.0646, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 0.10709526809969941, | |
| "grad_norm": 5.403663158416748, | |
| "learning_rate": 0.0001223695393973807, | |
| "loss": 2.323, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 0.1072605385751619, | |
| "grad_norm": 7.638148307800293, | |
| "learning_rate": 0.0001221639876871719, | |
| "loss": 2.393, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 0.10742580905062442, | |
| "grad_norm": 8.628881454467773, | |
| "learning_rate": 0.0001219583374454963, | |
| "loss": 2.9563, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.10759107952608692, | |
| "grad_norm": 1.3764619827270508, | |
| "learning_rate": 0.00012175258958658564, | |
| "loss": 1.1341, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 0.10775635000154941, | |
| "grad_norm": 0.6377153396606445, | |
| "learning_rate": 0.00012154674502510555, | |
| "loss": 1.2823, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 0.10792162047701191, | |
| "grad_norm": 0.61078941822052, | |
| "learning_rate": 0.00012134080467615159, | |
| "loss": 1.2439, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 0.10808689095247441, | |
| "grad_norm": 0.6451861262321472, | |
| "learning_rate": 0.00012113476945524513, | |
| "loss": 1.6167, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 0.10825216142793691, | |
| "grad_norm": 0.7710828185081482, | |
| "learning_rate": 0.00012092864027832933, | |
| "loss": 2.015, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.1084174319033994, | |
| "grad_norm": 0.961591362953186, | |
| "learning_rate": 0.000120722418061765, | |
| "loss": 2.4769, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 0.1085827023788619, | |
| "grad_norm": 0.8059797286987305, | |
| "learning_rate": 0.0001205161037223266, | |
| "loss": 2.2861, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 0.1087479728543244, | |
| "grad_norm": 0.7990265488624573, | |
| "learning_rate": 0.00012030969817719808, | |
| "loss": 2.2446, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 0.1089132433297869, | |
| "grad_norm": 0.8020466566085815, | |
| "learning_rate": 0.00012010320234396894, | |
| "loss": 2.1129, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 0.1090785138052494, | |
| "grad_norm": 0.7914049625396729, | |
| "learning_rate": 0.00011989661714062999, | |
| "loss": 2.0312, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.10924378428071191, | |
| "grad_norm": 0.99219810962677, | |
| "learning_rate": 0.0001196899434855693, | |
| "loss": 2.0558, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 0.1094090547561744, | |
| "grad_norm": 0.9570561051368713, | |
| "learning_rate": 0.00011948318229756827, | |
| "loss": 2.2061, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 0.1095743252316369, | |
| "grad_norm": 1.194684386253357, | |
| "learning_rate": 0.00011927633449579735, | |
| "loss": 2.4531, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 0.1097395957070994, | |
| "grad_norm": 0.9222526550292969, | |
| "learning_rate": 0.0001190694009998121, | |
| "loss": 2.3778, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 0.1099048661825619, | |
| "grad_norm": 1.067865014076233, | |
| "learning_rate": 0.00011886238272954897, | |
| "loss": 2.2211, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.1100701366580244, | |
| "grad_norm": 1.1547572612762451, | |
| "learning_rate": 0.00011865528060532127, | |
| "loss": 2.3185, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 0.1102354071334869, | |
| "grad_norm": 1.175791621208191, | |
| "learning_rate": 0.0001184480955478152, | |
| "loss": 2.3825, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 0.1104006776089494, | |
| "grad_norm": 1.1731864213943481, | |
| "learning_rate": 0.00011824082847808558, | |
| "loss": 1.7605, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 0.11056594808441189, | |
| "grad_norm": 1.2130746841430664, | |
| "learning_rate": 0.00011803348031755179, | |
| "loss": 2.0597, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 0.11073121855987439, | |
| "grad_norm": 1.3803483247756958, | |
| "learning_rate": 0.0001178260519879937, | |
| "loss": 2.1179, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.11089648903533689, | |
| "grad_norm": 1.342759370803833, | |
| "learning_rate": 0.00011761854441154767, | |
| "loss": 1.7941, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 0.1110617595107994, | |
| "grad_norm": 1.518161416053772, | |
| "learning_rate": 0.00011741095851070228, | |
| "loss": 2.1746, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 0.1112270299862619, | |
| "grad_norm": 1.688681721687317, | |
| "learning_rate": 0.00011720329520829429, | |
| "loss": 2.8293, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 0.1113923004617244, | |
| "grad_norm": 1.6524933576583862, | |
| "learning_rate": 0.0001169955554275046, | |
| "loss": 2.4412, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 0.11155757093718689, | |
| "grad_norm": 1.8030879497528076, | |
| "learning_rate": 0.0001167877400918541, | |
| "loss": 2.5758, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.11172284141264939, | |
| "grad_norm": 1.8228859901428223, | |
| "learning_rate": 0.00011657985012519952, | |
| "loss": 2.3193, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 0.11188811188811189, | |
| "grad_norm": 1.8945361375808716, | |
| "learning_rate": 0.00011637188645172944, | |
| "loss": 2.538, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 0.11205338236357439, | |
| "grad_norm": 1.8033952713012695, | |
| "learning_rate": 0.00011616384999596006, | |
| "loss": 2.5063, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 0.11221865283903688, | |
| "grad_norm": 2.0463850498199463, | |
| "learning_rate": 0.00011595574168273111, | |
| "loss": 2.6117, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 0.11238392331449938, | |
| "grad_norm": 2.145220994949341, | |
| "learning_rate": 0.0001157475624372018, | |
| "loss": 2.8788, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.11254919378996188, | |
| "grad_norm": 2.3135502338409424, | |
| "learning_rate": 0.0001155393131848467, | |
| "loss": 2.9463, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 0.11271446426542438, | |
| "grad_norm": 2.378373622894287, | |
| "learning_rate": 0.00011533099485145155, | |
| "loss": 2.6389, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 0.11287973474088689, | |
| "grad_norm": 2.6474478244781494, | |
| "learning_rate": 0.00011512260836310924, | |
| "loss": 2.498, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 0.11304500521634939, | |
| "grad_norm": 2.6414291858673096, | |
| "learning_rate": 0.00011491415464621562, | |
| "loss": 2.5016, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 0.11321027569181188, | |
| "grad_norm": 3.052006959915161, | |
| "learning_rate": 0.00011470563462746541, | |
| "loss": 2.5619, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.11337554616727438, | |
| "grad_norm": 2.497357130050659, | |
| "learning_rate": 0.00011449704923384812, | |
| "loss": 2.5177, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 0.11354081664273688, | |
| "grad_norm": 3.317051649093628, | |
| "learning_rate": 0.00011428839939264382, | |
| "loss": 2.5258, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 0.11370608711819938, | |
| "grad_norm": 2.784662961959839, | |
| "learning_rate": 0.0001140796860314191, | |
| "loss": 2.3684, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 0.11387135759366188, | |
| "grad_norm": 3.097684860229492, | |
| "learning_rate": 0.00011387091007802297, | |
| "loss": 2.3278, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 0.11403662806912437, | |
| "grad_norm": 3.4619364738464355, | |
| "learning_rate": 0.0001136620724605827, | |
| "loss": 2.7647, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.11420189854458687, | |
| "grad_norm": 3.7742695808410645, | |
| "learning_rate": 0.00011345317410749964, | |
| "loss": 2.1908, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 0.11436716902004937, | |
| "grad_norm": 3.8340349197387695, | |
| "learning_rate": 0.00011324421594744516, | |
| "loss": 2.4625, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 0.11453243949551187, | |
| "grad_norm": 3.795687198638916, | |
| "learning_rate": 0.00011303519890935656, | |
| "loss": 2.5857, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 0.11469770997097438, | |
| "grad_norm": 3.566619873046875, | |
| "learning_rate": 0.00011282612392243286, | |
| "loss": 1.7235, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 0.11486298044643688, | |
| "grad_norm": 4.450596332550049, | |
| "learning_rate": 0.00011261699191613066, | |
| "loss": 2.2944, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.11502825092189937, | |
| "grad_norm": 4.696109294891357, | |
| "learning_rate": 0.00011240780382016005, | |
| "loss": 2.1651, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 0.11519352139736187, | |
| "grad_norm": 4.380517959594727, | |
| "learning_rate": 0.00011219856056448051, | |
| "loss": 1.6746, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 0.11535879187282437, | |
| "grad_norm": 5.125734806060791, | |
| "learning_rate": 0.00011198926307929664, | |
| "loss": 2.0477, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 0.11552406234828687, | |
| "grad_norm": 6.3372392654418945, | |
| "learning_rate": 0.00011177991229505431, | |
| "loss": 2.4232, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 0.11568933282374937, | |
| "grad_norm": 8.878656387329102, | |
| "learning_rate": 0.00011157050914243614, | |
| "loss": 3.2974, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.11585460329921186, | |
| "grad_norm": 0.8543605804443359, | |
| "learning_rate": 0.00011136105455235766, | |
| "loss": 1.0566, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 0.11601987377467436, | |
| "grad_norm": 0.5880201458930969, | |
| "learning_rate": 0.00011115154945596305, | |
| "loss": 1.2489, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 0.11618514425013686, | |
| "grad_norm": 0.623322069644928, | |
| "learning_rate": 0.00011094199478462095, | |
| "loss": 1.5908, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 0.11635041472559936, | |
| "grad_norm": 1.172674536705017, | |
| "learning_rate": 0.00011073239146992054, | |
| "loss": 2.3208, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 0.11651568520106187, | |
| "grad_norm": 0.7992529273033142, | |
| "learning_rate": 0.00011052274044366711, | |
| "loss": 2.0372, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.11668095567652437, | |
| "grad_norm": 0.7547040581703186, | |
| "learning_rate": 0.00011031304263787812, | |
| "loss": 2.1528, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 0.11684622615198687, | |
| "grad_norm": 0.782480776309967, | |
| "learning_rate": 0.00011010329898477891, | |
| "loss": 2.0863, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 0.11701149662744936, | |
| "grad_norm": 0.7878833413124084, | |
| "learning_rate": 0.0001098935104167988, | |
| "loss": 2.2428, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 0.11717676710291186, | |
| "grad_norm": 0.841954231262207, | |
| "learning_rate": 0.00010968367786656663, | |
| "loss": 2.2871, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 0.11734203757837436, | |
| "grad_norm": 0.8345673084259033, | |
| "learning_rate": 0.00010947380226690684, | |
| "loss": 2.0559, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.11750730805383686, | |
| "grad_norm": 0.9443355798721313, | |
| "learning_rate": 0.00010926388455083522, | |
| "loss": 2.0117, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 0.11767257852929935, | |
| "grad_norm": 1.0044097900390625, | |
| "learning_rate": 0.00010905392565155477, | |
| "loss": 2.1702, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 0.11783784900476185, | |
| "grad_norm": 1.0212572813034058, | |
| "learning_rate": 0.00010884392650245165, | |
| "loss": 1.8402, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 0.11800311948022435, | |
| "grad_norm": 0.9890906810760498, | |
| "learning_rate": 0.00010863388803709089, | |
| "loss": 2.0654, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 0.11816838995568685, | |
| "grad_norm": 1.0666841268539429, | |
| "learning_rate": 0.00010842381118921232, | |
| "loss": 2.1339, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.11833366043114936, | |
| "grad_norm": 1.0679786205291748, | |
| "learning_rate": 0.00010821369689272638, | |
| "loss": 2.1859, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 0.11849893090661186, | |
| "grad_norm": 1.2677335739135742, | |
| "learning_rate": 0.00010800354608171003, | |
| "loss": 2.01, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 0.11866420138207436, | |
| "grad_norm": 1.263466715812683, | |
| "learning_rate": 0.00010779335969040252, | |
| "loss": 1.9806, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 0.11882947185753685, | |
| "grad_norm": 1.5951952934265137, | |
| "learning_rate": 0.00010758313865320134, | |
| "loss": 1.8264, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 0.11899474233299935, | |
| "grad_norm": 1.296347975730896, | |
| "learning_rate": 0.00010737288390465792, | |
| "loss": 1.8939, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.11916001280846185, | |
| "grad_norm": 1.5633519887924194, | |
| "learning_rate": 0.00010716259637947357, | |
| "loss": 2.1699, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 0.11932528328392435, | |
| "grad_norm": 1.7896921634674072, | |
| "learning_rate": 0.00010695227701249537, | |
| "loss": 2.6321, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 0.11949055375938684, | |
| "grad_norm": 1.8836790323257446, | |
| "learning_rate": 0.00010674192673871191, | |
| "loss": 2.4108, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 0.11965582423484934, | |
| "grad_norm": 1.7736847400665283, | |
| "learning_rate": 0.00010653154649324917, | |
| "loss": 2.4307, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 0.11982109471031184, | |
| "grad_norm": 1.652762770652771, | |
| "learning_rate": 0.00010632113721136636, | |
| "loss": 2.309, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.11998636518577434, | |
| "grad_norm": 1.7753866910934448, | |
| "learning_rate": 0.00010611069982845183, | |
| "loss": 2.1368, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 0.12015163566123685, | |
| "grad_norm": 2.14203143119812, | |
| "learning_rate": 0.00010590023528001884, | |
| "loss": 2.7191, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 0.12031690613669935, | |
| "grad_norm": 2.011603832244873, | |
| "learning_rate": 0.00010568974450170139, | |
| "loss": 2.4192, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 0.12048217661216185, | |
| "grad_norm": 2.1191020011901855, | |
| "learning_rate": 0.00010547922842925008, | |
| "loss": 2.7372, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 0.12064744708762434, | |
| "grad_norm": 2.30214786529541, | |
| "learning_rate": 0.00010526868799852796, | |
| "loss": 2.8499, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.12081271756308684, | |
| "grad_norm": 2.388906955718994, | |
| "learning_rate": 0.0001050581241455064, | |
| "loss": 2.4864, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 0.12097798803854934, | |
| "grad_norm": 2.5357463359832764, | |
| "learning_rate": 0.00010484753780626089, | |
| "loss": 2.7267, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 0.12114325851401184, | |
| "grad_norm": 2.7825512886047363, | |
| "learning_rate": 0.00010463692991696685, | |
| "loss": 2.7716, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 0.12130852898947433, | |
| "grad_norm": 2.7057747840881348, | |
| "learning_rate": 0.00010442630141389549, | |
| "loss": 2.5268, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 0.12147379946493683, | |
| "grad_norm": 2.3839614391326904, | |
| "learning_rate": 0.00010421565323340971, | |
| "loss": 2.5892, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.12163906994039933, | |
| "grad_norm": 2.9489643573760986, | |
| "learning_rate": 0.00010400498631195992, | |
| "loss": 2.2354, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 0.12180434041586183, | |
| "grad_norm": 2.982599973678589, | |
| "learning_rate": 0.00010379430158607975, | |
| "loss": 2.6725, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 0.12196961089132434, | |
| "grad_norm": 3.166109323501587, | |
| "learning_rate": 0.000103583599992382, | |
| "loss": 2.2997, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 0.12213488136678684, | |
| "grad_norm": 3.4669601917266846, | |
| "learning_rate": 0.0001033728824675545, | |
| "loss": 2.8882, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 0.12230015184224934, | |
| "grad_norm": 3.7491748332977295, | |
| "learning_rate": 0.0001031621499483559, | |
| "loss": 2.2155, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.12246542231771183, | |
| "grad_norm": 3.875016927719116, | |
| "learning_rate": 0.00010295140337161146, | |
| "loss": 2.5984, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 0.12263069279317433, | |
| "grad_norm": 3.715601921081543, | |
| "learning_rate": 0.00010274064367420897, | |
| "loss": 2.2819, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 0.12279596326863683, | |
| "grad_norm": 4.2248759269714355, | |
| "learning_rate": 0.00010252987179309459, | |
| "loss": 2.9126, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 0.12296123374409933, | |
| "grad_norm": 4.149989128112793, | |
| "learning_rate": 0.00010231908866526851, | |
| "loss": 2.0437, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 0.12312650421956182, | |
| "grad_norm": 3.8425707817077637, | |
| "learning_rate": 0.00010210829522778111, | |
| "loss": 2.0575, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.12329177469502432, | |
| "grad_norm": 4.6139726638793945, | |
| "learning_rate": 0.00010189749241772844, | |
| "loss": 2.1791, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 0.12345704517048682, | |
| "grad_norm": 4.846596717834473, | |
| "learning_rate": 0.00010168668117224825, | |
| "loss": 2.3917, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 0.12362231564594932, | |
| "grad_norm": 5.106334209442139, | |
| "learning_rate": 0.00010147586242851585, | |
| "loss": 1.9967, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 0.12378758612141182, | |
| "grad_norm": 5.454449653625488, | |
| "learning_rate": 0.00010126503712373982, | |
| "loss": 2.3207, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 0.12395285659687433, | |
| "grad_norm": 8.876123428344727, | |
| "learning_rate": 0.00010105420619515798, | |
| "loss": 2.8928, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.12395285659687433, | |
| "eval_loss": 2.2081432342529297, | |
| "eval_runtime": 208.9615, | |
| "eval_samples_per_second": 48.77, | |
| "eval_steps_per_second": 24.387, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.12411812707233683, | |
| "grad_norm": 0.7086247205734253, | |
| "learning_rate": 0.00010084337058003303, | |
| "loss": 1.1074, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 0.12428339754779932, | |
| "grad_norm": 0.6974254250526428, | |
| "learning_rate": 0.00010063253121564868, | |
| "loss": 1.3559, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 0.12444866802326182, | |
| "grad_norm": 0.5847852826118469, | |
| "learning_rate": 0.00010042168903930514, | |
| "loss": 1.2863, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 0.12461393849872432, | |
| "grad_norm": 0.6428030133247375, | |
| "learning_rate": 0.00010021084498831522, | |
| "loss": 1.7291, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 0.12477920897418682, | |
| "grad_norm": 0.9877603054046631, | |
| "learning_rate": 0.0001, | |
| "loss": 2.3706, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.12494447944964931, | |
| "grad_norm": 0.7717822194099426, | |
| "learning_rate": 9.97891550116848e-05, | |
| "loss": 2.8035, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 0.1251097499251118, | |
| "grad_norm": 0.6428593993186951, | |
| "learning_rate": 9.957831096069488e-05, | |
| "loss": 1.9591, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 0.1252750204005743, | |
| "grad_norm": 0.6849289536476135, | |
| "learning_rate": 9.936746878435136e-05, | |
| "loss": 1.9275, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 0.1254402908760368, | |
| "grad_norm": 0.8406524062156677, | |
| "learning_rate": 9.915662941996699e-05, | |
| "loss": 2.2426, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 0.1256055613514993, | |
| "grad_norm": 0.8995076417922974, | |
| "learning_rate": 9.894579380484204e-05, | |
| "loss": 2.3196, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.1257708318269618, | |
| "grad_norm": 0.923305094242096, | |
| "learning_rate": 9.873496287626019e-05, | |
| "loss": 2.3969, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 0.1259361023024243, | |
| "grad_norm": 0.8841802477836609, | |
| "learning_rate": 9.852413757148417e-05, | |
| "loss": 2.1089, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 0.1261013727778868, | |
| "grad_norm": 1.0622060298919678, | |
| "learning_rate": 9.831331882775178e-05, | |
| "loss": 2.2627, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 0.1262666432533493, | |
| "grad_norm": 1.0138001441955566, | |
| "learning_rate": 9.81025075822716e-05, | |
| "loss": 1.8757, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 0.1264319137288118, | |
| "grad_norm": 1.0839015245437622, | |
| "learning_rate": 9.789170477221891e-05, | |
| "loss": 2.1583, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.12659718420427432, | |
| "grad_norm": 1.143316388130188, | |
| "learning_rate": 9.76809113347315e-05, | |
| "loss": 1.9662, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 0.12676245467973682, | |
| "grad_norm": 1.298763394355774, | |
| "learning_rate": 9.747012820690543e-05, | |
| "loss": 2.0712, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 0.12692772515519932, | |
| "grad_norm": 1.321671485900879, | |
| "learning_rate": 9.725935632579104e-05, | |
| "loss": 2.1975, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 0.12709299563066181, | |
| "grad_norm": 1.420569658279419, | |
| "learning_rate": 9.704859662838855e-05, | |
| "loss": 2.1219, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 0.1272582661061243, | |
| "grad_norm": 1.458512544631958, | |
| "learning_rate": 9.683785005164411e-05, | |
| "loss": 2.2603, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.1274235365815868, | |
| "grad_norm": 1.5219634771347046, | |
| "learning_rate": 9.662711753244551e-05, | |
| "loss": 2.2249, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 0.1275888070570493, | |
| "grad_norm": 1.691079020500183, | |
| "learning_rate": 9.641640000761802e-05, | |
| "loss": 2.1337, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 0.1277540775325118, | |
| "grad_norm": 1.8330628871917725, | |
| "learning_rate": 9.620569841392029e-05, | |
| "loss": 2.5652, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 0.1279193480079743, | |
| "grad_norm": 1.7486244440078735, | |
| "learning_rate": 9.59950136880401e-05, | |
| "loss": 2.2389, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 0.1280846184834368, | |
| "grad_norm": 1.6367963552474976, | |
| "learning_rate": 9.57843467665903e-05, | |
| "loss": 2.3473, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.1282498889588993, | |
| "grad_norm": 2.0890188217163086, | |
| "learning_rate": 9.557369858610453e-05, | |
| "loss": 2.8439, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 0.1284151594343618, | |
| "grad_norm": 2.399138927459717, | |
| "learning_rate": 9.53630700830332e-05, | |
| "loss": 2.3451, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 0.1285804299098243, | |
| "grad_norm": 2.086455821990967, | |
| "learning_rate": 9.51524621937391e-05, | |
| "loss": 2.4895, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 0.1287457003852868, | |
| "grad_norm": 2.0463106632232666, | |
| "learning_rate": 9.494187585449358e-05, | |
| "loss": 2.7085, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 0.1289109708607493, | |
| "grad_norm": 2.2464828491210938, | |
| "learning_rate": 9.473131200147205e-05, | |
| "loss": 2.5326, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.1290762413362118, | |
| "grad_norm": 2.1522295475006104, | |
| "learning_rate": 9.452077157074994e-05, | |
| "loss": 2.7267, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 0.1292415118116743, | |
| "grad_norm": 2.1915390491485596, | |
| "learning_rate": 9.431025549829862e-05, | |
| "loss": 2.5596, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 0.12940678228713678, | |
| "grad_norm": 2.3839638233184814, | |
| "learning_rate": 9.409976471998118e-05, | |
| "loss": 2.4065, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 0.12957205276259928, | |
| "grad_norm": 2.4610612392425537, | |
| "learning_rate": 9.388930017154819e-05, | |
| "loss": 2.2588, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 0.12973732323806178, | |
| "grad_norm": 2.763357639312744, | |
| "learning_rate": 9.367886278863366e-05, | |
| "loss": 2.1213, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.12990259371352428, | |
| "grad_norm": 2.982503890991211, | |
| "learning_rate": 9.346845350675088e-05, | |
| "loss": 2.3916, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 0.13006786418898678, | |
| "grad_norm": 3.4206717014312744, | |
| "learning_rate": 9.325807326128814e-05, | |
| "loss": 2.3497, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 0.1302331346644493, | |
| "grad_norm": 2.861182928085327, | |
| "learning_rate": 9.304772298750463e-05, | |
| "loss": 2.111, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 0.1303984051399118, | |
| "grad_norm": 3.2037413120269775, | |
| "learning_rate": 9.283740362052642e-05, | |
| "loss": 2.1358, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 0.1305636756153743, | |
| "grad_norm": 3.2721924781799316, | |
| "learning_rate": 9.26271160953421e-05, | |
| "loss": 2.262, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.1307289460908368, | |
| "grad_norm": 3.294147253036499, | |
| "learning_rate": 9.241686134679867e-05, | |
| "loss": 1.9879, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 0.1308942165662993, | |
| "grad_norm": 3.8471827507019043, | |
| "learning_rate": 9.220664030959749e-05, | |
| "loss": 2.0336, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 0.1310594870417618, | |
| "grad_norm": 3.8515522480010986, | |
| "learning_rate": 9.199645391828999e-05, | |
| "loss": 2.1412, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 0.1312247575172243, | |
| "grad_norm": 4.013675212860107, | |
| "learning_rate": 9.178630310727365e-05, | |
| "loss": 2.0929, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 0.1313900279926868, | |
| "grad_norm": 3.8516957759857178, | |
| "learning_rate": 9.157618881078772e-05, | |
| "loss": 2.3487, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.13155529846814928, | |
| "grad_norm": 4.891301155090332, | |
| "learning_rate": 9.136611196290915e-05, | |
| "loss": 2.2354, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 0.13172056894361178, | |
| "grad_norm": 4.761098861694336, | |
| "learning_rate": 9.115607349754834e-05, | |
| "loss": 2.259, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 0.13188583941907428, | |
| "grad_norm": 5.0645599365234375, | |
| "learning_rate": 9.094607434844523e-05, | |
| "loss": 2.0045, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 0.13205110989453678, | |
| "grad_norm": 6.802490711212158, | |
| "learning_rate": 9.07361154491648e-05, | |
| "loss": 2.481, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 0.13221638036999928, | |
| "grad_norm": 9.168866157531738, | |
| "learning_rate": 9.052619773309317e-05, | |
| "loss": 3.1667, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.13238165084546177, | |
| "grad_norm": 0.7689570784568787, | |
| "learning_rate": 9.031632213343339e-05, | |
| "loss": 1.195, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 0.13254692132092427, | |
| "grad_norm": 0.5660879015922546, | |
| "learning_rate": 9.01064895832012e-05, | |
| "loss": 1.3269, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 0.13271219179638677, | |
| "grad_norm": 0.7060874700546265, | |
| "learning_rate": 8.98967010152211e-05, | |
| "loss": 1.4369, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 0.13287746227184927, | |
| "grad_norm": 1.2695504426956177, | |
| "learning_rate": 8.968695736212193e-05, | |
| "loss": 2.2576, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 0.13304273274731176, | |
| "grad_norm": 0.7867845296859741, | |
| "learning_rate": 8.947725955633294e-05, | |
| "loss": 2.6122, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.13320800322277426, | |
| "grad_norm": 0.7278014421463013, | |
| "learning_rate": 8.926760853007946e-05, | |
| "loss": 1.8505, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 0.13337327369823676, | |
| "grad_norm": 0.8305306434631348, | |
| "learning_rate": 8.905800521537905e-05, | |
| "loss": 2.2389, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 0.13353854417369926, | |
| "grad_norm": 0.8625719547271729, | |
| "learning_rate": 8.884845054403699e-05, | |
| "loss": 2.0574, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 0.13370381464916176, | |
| "grad_norm": 0.8179556131362915, | |
| "learning_rate": 8.863894544764236e-05, | |
| "loss": 2.1356, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 0.13386908512462425, | |
| "grad_norm": 0.8811230063438416, | |
| "learning_rate": 8.84294908575639e-05, | |
| "loss": 2.1357, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.13403435560008678, | |
| "grad_norm": 0.8514245748519897, | |
| "learning_rate": 8.822008770494572e-05, | |
| "loss": 1.9962, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 0.13419962607554928, | |
| "grad_norm": 1.0908472537994385, | |
| "learning_rate": 8.801073692070337e-05, | |
| "loss": 2.2666, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 0.13436489655101178, | |
| "grad_norm": 1.0929603576660156, | |
| "learning_rate": 8.780143943551954e-05, | |
| "loss": 2.0224, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 0.13453016702647427, | |
| "grad_norm": 1.0276299715042114, | |
| "learning_rate": 8.759219617983999e-05, | |
| "loss": 2.1321, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 0.13469543750193677, | |
| "grad_norm": 1.0192618370056152, | |
| "learning_rate": 8.738300808386935e-05, | |
| "loss": 2.2379, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.13486070797739927, | |
| "grad_norm": 1.3195905685424805, | |
| "learning_rate": 8.717387607756713e-05, | |
| "loss": 1.7885, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 0.13502597845286177, | |
| "grad_norm": 1.2186145782470703, | |
| "learning_rate": 8.696480109064342e-05, | |
| "loss": 2.0619, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 0.13519124892832426, | |
| "grad_norm": 1.3322213888168335, | |
| "learning_rate": 8.675578405255485e-05, | |
| "loss": 2.0923, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 0.13535651940378676, | |
| "grad_norm": 1.3305845260620117, | |
| "learning_rate": 8.654682589250038e-05, | |
| "loss": 2.1558, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 0.13552178987924926, | |
| "grad_norm": 1.5486564636230469, | |
| "learning_rate": 8.633792753941733e-05, | |
| "loss": 2.0763, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.13568706035471176, | |
| "grad_norm": 1.509981393814087, | |
| "learning_rate": 8.612908992197705e-05, | |
| "loss": 2.1749, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 0.13585233083017426, | |
| "grad_norm": 1.5474811792373657, | |
| "learning_rate": 8.592031396858093e-05, | |
| "loss": 2.3094, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 0.13601760130563675, | |
| "grad_norm": 1.757800579071045, | |
| "learning_rate": 8.571160060735624e-05, | |
| "loss": 2.5309, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 0.13618287178109925, | |
| "grad_norm": 1.9582196474075317, | |
| "learning_rate": 8.550295076615188e-05, | |
| "loss": 2.7508, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 0.13634814225656175, | |
| "grad_norm": 2.000728130340576, | |
| "learning_rate": 8.529436537253458e-05, | |
| "loss": 2.2387, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.13651341273202425, | |
| "grad_norm": 2.1247360706329346, | |
| "learning_rate": 8.508584535378439e-05, | |
| "loss": 2.4557, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 0.13667868320748675, | |
| "grad_norm": 2.0145645141601562, | |
| "learning_rate": 8.487739163689079e-05, | |
| "loss": 2.6309, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 0.13684395368294924, | |
| "grad_norm": 2.161586046218872, | |
| "learning_rate": 8.466900514854847e-05, | |
| "loss": 2.201, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 0.13700922415841174, | |
| "grad_norm": 2.11183762550354, | |
| "learning_rate": 8.446068681515334e-05, | |
| "loss": 2.3269, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 0.13717449463387424, | |
| "grad_norm": 2.305616855621338, | |
| "learning_rate": 8.425243756279824e-05, | |
| "loss": 2.1798, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.13733976510933674, | |
| "grad_norm": 2.301452159881592, | |
| "learning_rate": 8.404425831726894e-05, | |
| "loss": 3.0408, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 0.13750503558479923, | |
| "grad_norm": 2.6158676147460938, | |
| "learning_rate": 8.383615000404e-05, | |
| "loss": 2.9498, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 0.13767030606026176, | |
| "grad_norm": 2.5953450202941895, | |
| "learning_rate": 8.362811354827059e-05, | |
| "loss": 2.6831, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 0.13783557653572426, | |
| "grad_norm": 2.5904581546783447, | |
| "learning_rate": 8.342014987480047e-05, | |
| "loss": 2.4604, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 0.13800084701118676, | |
| "grad_norm": 2.9298088550567627, | |
| "learning_rate": 8.321225990814591e-05, | |
| "loss": 2.7181, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.13816611748664925, | |
| "grad_norm": 3.073090076446533, | |
| "learning_rate": 8.300444457249543e-05, | |
| "loss": 2.3408, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 0.13833138796211175, | |
| "grad_norm": 3.1069414615631104, | |
| "learning_rate": 8.279670479170573e-05, | |
| "loss": 2.3978, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 0.13849665843757425, | |
| "grad_norm": 2.655959367752075, | |
| "learning_rate": 8.258904148929775e-05, | |
| "loss": 2.3696, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 0.13866192891303675, | |
| "grad_norm": 3.284933567047119, | |
| "learning_rate": 8.238145558845235e-05, | |
| "loss": 2.4918, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 0.13882719938849925, | |
| "grad_norm": 3.335923433303833, | |
| "learning_rate": 8.217394801200631e-05, | |
| "loss": 2.579, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.13899246986396174, | |
| "grad_norm": 3.2936391830444336, | |
| "learning_rate": 8.196651968244826e-05, | |
| "loss": 2.2769, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 0.13915774033942424, | |
| "grad_norm": 3.5451266765594482, | |
| "learning_rate": 8.175917152191447e-05, | |
| "loss": 1.9767, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 0.13932301081488674, | |
| "grad_norm": 4.063502788543701, | |
| "learning_rate": 8.15519044521848e-05, | |
| "loss": 2.1579, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 0.13948828129034924, | |
| "grad_norm": 3.840585947036743, | |
| "learning_rate": 8.134471939467874e-05, | |
| "loss": 1.883, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 0.13965355176581173, | |
| "grad_norm": 5.08105993270874, | |
| "learning_rate": 8.113761727045105e-05, | |
| "loss": 2.3574, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.13981882224127423, | |
| "grad_norm": 3.766429901123047, | |
| "learning_rate": 8.093059900018792e-05, | |
| "loss": 1.694, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 0.13998409271673673, | |
| "grad_norm": 5.276097774505615, | |
| "learning_rate": 8.072366550420266e-05, | |
| "loss": 2.1232, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 0.14014936319219923, | |
| "grad_norm": 5.221464157104492, | |
| "learning_rate": 8.051681770243175e-05, | |
| "loss": 2.3516, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 0.14031463366766173, | |
| "grad_norm": 5.557736396789551, | |
| "learning_rate": 8.031005651443073e-05, | |
| "loss": 2.283, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 0.14047990414312422, | |
| "grad_norm": 11.079894065856934, | |
| "learning_rate": 8.010338285937006e-05, | |
| "loss": 2.9246, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.14064517461858672, | |
| "grad_norm": 0.64576655626297, | |
| "learning_rate": 7.989679765603108e-05, | |
| "loss": 1.2148, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 0.14081044509404922, | |
| "grad_norm": 0.571146547794342, | |
| "learning_rate": 7.969030182280192e-05, | |
| "loss": 1.2923, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 0.14097571556951172, | |
| "grad_norm": 0.5226064324378967, | |
| "learning_rate": 7.948389627767343e-05, | |
| "loss": 1.2965, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 0.14114098604497421, | |
| "grad_norm": 0.5962955951690674, | |
| "learning_rate": 7.927758193823501e-05, | |
| "loss": 1.493, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 0.14130625652043674, | |
| "grad_norm": 1.0950610637664795, | |
| "learning_rate": 7.907135972167069e-05, | |
| "loss": 2.145, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.14147152699589924, | |
| "grad_norm": 1.05116605758667, | |
| "learning_rate": 7.88652305447549e-05, | |
| "loss": 2.496, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 0.14163679747136174, | |
| "grad_norm": 0.8175346851348877, | |
| "learning_rate": 7.865919532384844e-05, | |
| "loss": 2.1028, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 0.14180206794682423, | |
| "grad_norm": 0.8164800405502319, | |
| "learning_rate": 7.845325497489449e-05, | |
| "loss": 2.1895, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 0.14196733842228673, | |
| "grad_norm": 0.7507344484329224, | |
| "learning_rate": 7.82474104134144e-05, | |
| "loss": 1.8808, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 0.14213260889774923, | |
| "grad_norm": 0.8547980189323425, | |
| "learning_rate": 7.804166255450373e-05, | |
| "loss": 2.2765, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.14229787937321173, | |
| "grad_norm": 1.1604375839233398, | |
| "learning_rate": 7.783601231282812e-05, | |
| "loss": 2.3694, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 0.14246314984867423, | |
| "grad_norm": 0.9143975377082825, | |
| "learning_rate": 7.763046060261932e-05, | |
| "loss": 1.9546, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 0.14262842032413672, | |
| "grad_norm": 0.8295143246650696, | |
| "learning_rate": 7.742500833767094e-05, | |
| "loss": 1.9922, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 0.14279369079959922, | |
| "grad_norm": 0.9933025240898132, | |
| "learning_rate": 7.721965643133458e-05, | |
| "loss": 1.9879, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 0.14295896127506172, | |
| "grad_norm": 1.0303356647491455, | |
| "learning_rate": 7.701440579651566e-05, | |
| "loss": 2.2688, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.14312423175052422, | |
| "grad_norm": 1.0926827192306519, | |
| "learning_rate": 7.680925734566937e-05, | |
| "loss": 2.0978, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 0.14328950222598671, | |
| "grad_norm": 1.111046552658081, | |
| "learning_rate": 7.660421199079669e-05, | |
| "loss": 2.0009, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 0.1434547727014492, | |
| "grad_norm": 1.3105140924453735, | |
| "learning_rate": 7.639927064344022e-05, | |
| "loss": 2.5226, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 0.1436200431769117, | |
| "grad_norm": 1.3718643188476562, | |
| "learning_rate": 7.619443421468021e-05, | |
| "loss": 2.3989, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 0.1437853136523742, | |
| "grad_norm": 1.386592149734497, | |
| "learning_rate": 7.598970361513051e-05, | |
| "loss": 1.9662, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.1439505841278367, | |
| "grad_norm": 1.6088908910751343, | |
| "learning_rate": 7.578507975493448e-05, | |
| "loss": 1.8162, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 0.1441158546032992, | |
| "grad_norm": 1.6048753261566162, | |
| "learning_rate": 7.558056354376098e-05, | |
| "loss": 2.4007, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 0.1442811250787617, | |
| "grad_norm": 1.7685431241989136, | |
| "learning_rate": 7.537615589080027e-05, | |
| "loss": 2.3788, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 0.1444463955542242, | |
| "grad_norm": 1.5546327829360962, | |
| "learning_rate": 7.517185770476006e-05, | |
| "loss": 2.1635, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 0.1446116660296867, | |
| "grad_norm": 1.7146538496017456, | |
| "learning_rate": 7.496766989386136e-05, | |
| "loss": 2.3133, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.1447769365051492, | |
| "grad_norm": 1.8806188106536865, | |
| "learning_rate": 7.476359336583454e-05, | |
| "loss": 2.395, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 0.14494220698061172, | |
| "grad_norm": 1.8896050453186035, | |
| "learning_rate": 7.455962902791522e-05, | |
| "loss": 2.3525, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 0.14510747745607422, | |
| "grad_norm": 2.016399383544922, | |
| "learning_rate": 7.435577778684033e-05, | |
| "loss": 2.7357, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 0.14527274793153672, | |
| "grad_norm": 1.8287063837051392, | |
| "learning_rate": 7.415204054884399e-05, | |
| "loss": 2.4014, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 0.14543801840699921, | |
| "grad_norm": 2.3520097732543945, | |
| "learning_rate": 7.394841821965345e-05, | |
| "loss": 2.4157, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.1456032888824617, | |
| "grad_norm": 2.3081214427948, | |
| "learning_rate": 7.374491170448525e-05, | |
| "loss": 2.4761, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 0.1457685593579242, | |
| "grad_norm": 2.556182622909546, | |
| "learning_rate": 7.3541521908041e-05, | |
| "loss": 2.0215, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 0.1459338298333867, | |
| "grad_norm": 2.7098548412323, | |
| "learning_rate": 7.33382497345034e-05, | |
| "loss": 2.4724, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 0.1460991003088492, | |
| "grad_norm": 2.6992313861846924, | |
| "learning_rate": 7.313509608753231e-05, | |
| "loss": 2.2446, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 0.1462643707843117, | |
| "grad_norm": 2.4774627685546875, | |
| "learning_rate": 7.293206187026061e-05, | |
| "loss": 2.3697, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.1464296412597742, | |
| "grad_norm": 2.8421108722686768, | |
| "learning_rate": 7.27291479852903e-05, | |
| "loss": 2.5001, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 0.1465949117352367, | |
| "grad_norm": 3.111126184463501, | |
| "learning_rate": 7.252635533468843e-05, | |
| "loss": 2.3817, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 0.1467601822106992, | |
| "grad_norm": 3.4716055393218994, | |
| "learning_rate": 7.232368481998309e-05, | |
| "loss": 2.419, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 0.1469254526861617, | |
| "grad_norm": 3.5414786338806152, | |
| "learning_rate": 7.212113734215932e-05, | |
| "loss": 2.6442, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 0.1470907231616242, | |
| "grad_norm": 3.838484764099121, | |
| "learning_rate": 7.191871380165538e-05, | |
| "loss": 2.2247, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.1472559936370867, | |
| "grad_norm": 3.4011685848236084, | |
| "learning_rate": 7.17164150983584e-05, | |
| "loss": 2.313, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 0.1474212641125492, | |
| "grad_norm": 3.8526804447174072, | |
| "learning_rate": 7.151424213160061e-05, | |
| "loss": 2.1966, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 0.1475865345880117, | |
| "grad_norm": 5.23972225189209, | |
| "learning_rate": 7.131219580015521e-05, | |
| "loss": 1.9577, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 0.14775180506347418, | |
| "grad_norm": 4.628785610198975, | |
| "learning_rate": 7.11102770022325e-05, | |
| "loss": 2.6681, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 0.14791707553893668, | |
| "grad_norm": 4.514820098876953, | |
| "learning_rate": 7.090848663547574e-05, | |
| "loss": 2.4181, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.14808234601439918, | |
| "grad_norm": 4.820570468902588, | |
| "learning_rate": 7.070682559695736e-05, | |
| "loss": 2.2424, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 0.14824761648986168, | |
| "grad_norm": 4.191586017608643, | |
| "learning_rate": 7.050529478317476e-05, | |
| "loss": 1.8049, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 0.14841288696532418, | |
| "grad_norm": 4.894969463348389, | |
| "learning_rate": 7.03038950900464e-05, | |
| "loss": 2.4348, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 0.1485781574407867, | |
| "grad_norm": 5.656913757324219, | |
| "learning_rate": 7.010262741290798e-05, | |
| "loss": 2.0972, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 0.1487434279162492, | |
| "grad_norm": 8.331171989440918, | |
| "learning_rate": 6.990149264650814e-05, | |
| "loss": 2.6102, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.1489086983917117, | |
| "grad_norm": 0.5950672030448914, | |
| "learning_rate": 6.970049168500474e-05, | |
| "loss": 1.1136, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 0.1490739688671742, | |
| "grad_norm": 0.5455874800682068, | |
| "learning_rate": 6.94996254219608e-05, | |
| "loss": 1.345, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 0.1492392393426367, | |
| "grad_norm": 0.5597286820411682, | |
| "learning_rate": 6.929889475034048e-05, | |
| "loss": 1.243, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 0.1494045098180992, | |
| "grad_norm": 0.8704578876495361, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 1.7736, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 0.1495697802935617, | |
| "grad_norm": 0.8955904841423035, | |
| "learning_rate": 6.889784375020978e-05, | |
| "loss": 2.0367, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.14973505076902419, | |
| "grad_norm": 0.8076390624046326, | |
| "learning_rate": 6.869752520459803e-05, | |
| "loss": 2.3345, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 0.14990032124448668, | |
| "grad_norm": 0.7306450009346008, | |
| "learning_rate": 6.849734581619918e-05, | |
| "loss": 1.8771, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 0.15006559171994918, | |
| "grad_norm": 0.7677145004272461, | |
| "learning_rate": 6.829730647492404e-05, | |
| "loss": 2.2029, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 0.15023086219541168, | |
| "grad_norm": 0.7807202935218811, | |
| "learning_rate": 6.80974080700606e-05, | |
| "loss": 1.8954, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 0.15039613267087418, | |
| "grad_norm": 0.8537066578865051, | |
| "learning_rate": 6.789765149027039e-05, | |
| "loss": 2.283, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.15056140314633668, | |
| "grad_norm": 0.8551520109176636, | |
| "learning_rate": 6.769803762358443e-05, | |
| "loss": 1.994, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 0.15072667362179917, | |
| "grad_norm": 1.020125389099121, | |
| "learning_rate": 6.749856735739928e-05, | |
| "loss": 2.3547, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 0.15089194409726167, | |
| "grad_norm": 0.8300055861473083, | |
| "learning_rate": 6.729924157847323e-05, | |
| "loss": 1.9244, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 0.15105721457272417, | |
| "grad_norm": 0.9789553880691528, | |
| "learning_rate": 6.710006117292209e-05, | |
| "loss": 2.1382, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 0.15122248504818667, | |
| "grad_norm": 1.166289210319519, | |
| "learning_rate": 6.690102702621548e-05, | |
| "loss": 2.1411, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.15138775552364916, | |
| "grad_norm": 1.088026523590088, | |
| "learning_rate": 6.670214002317278e-05, | |
| "loss": 1.9739, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 0.15155302599911166, | |
| "grad_norm": 1.1498019695281982, | |
| "learning_rate": 6.650340104795932e-05, | |
| "loss": 2.2583, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 0.15171829647457416, | |
| "grad_norm": 1.260051965713501, | |
| "learning_rate": 6.630481098408228e-05, | |
| "loss": 2.2788, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 0.15188356695003666, | |
| "grad_norm": 1.3153319358825684, | |
| "learning_rate": 6.610637071438686e-05, | |
| "loss": 2.0744, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 0.15204883742549916, | |
| "grad_norm": 1.3484034538269043, | |
| "learning_rate": 6.590808112105232e-05, | |
| "loss": 2.0484, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.15221410790096165, | |
| "grad_norm": 1.6158196926116943, | |
| "learning_rate": 6.570994308558812e-05, | |
| "loss": 2.4273, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 0.15237937837642418, | |
| "grad_norm": 1.6348590850830078, | |
| "learning_rate": 6.551195748882997e-05, | |
| "loss": 2.1496, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 0.15254464885188668, | |
| "grad_norm": 1.6826823949813843, | |
| "learning_rate": 6.531412521093586e-05, | |
| "loss": 2.3494, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 0.15270991932734918, | |
| "grad_norm": 1.6990618705749512, | |
| "learning_rate": 6.51164471313822e-05, | |
| "loss": 2.107, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 0.15287518980281167, | |
| "grad_norm": 1.940071702003479, | |
| "learning_rate": 6.491892412895995e-05, | |
| "loss": 2.1787, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.15304046027827417, | |
| "grad_norm": 1.9719414710998535, | |
| "learning_rate": 6.472155708177052e-05, | |
| "loss": 2.3206, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 0.15320573075373667, | |
| "grad_norm": 2.0704565048217773, | |
| "learning_rate": 6.452434686722224e-05, | |
| "loss": 2.5605, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 0.15337100122919917, | |
| "grad_norm": 2.0246057510375977, | |
| "learning_rate": 6.432729436202604e-05, | |
| "loss": 2.6231, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 0.15353627170466166, | |
| "grad_norm": 2.181257724761963, | |
| "learning_rate": 6.41304004421918e-05, | |
| "loss": 2.124, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 0.15370154218012416, | |
| "grad_norm": 2.3393361568450928, | |
| "learning_rate": 6.393366598302446e-05, | |
| "loss": 2.6182, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.15386681265558666, | |
| "grad_norm": 2.4433555603027344, | |
| "learning_rate": 6.373709185911998e-05, | |
| "loss": 2.4835, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.15403208313104916, | |
| "grad_norm": 2.769249677658081, | |
| "learning_rate": 6.354067894436155e-05, | |
| "loss": 2.5903, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 0.15419735360651166, | |
| "grad_norm": 2.9044911861419678, | |
| "learning_rate": 6.334442811191576e-05, | |
| "loss": 2.7941, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 0.15436262408197415, | |
| "grad_norm": 2.4655826091766357, | |
| "learning_rate": 6.314834023422858e-05, | |
| "loss": 2.2729, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 0.15452789455743665, | |
| "grad_norm": 2.922982931137085, | |
| "learning_rate": 6.295241618302156e-05, | |
| "loss": 2.5217, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.15469316503289915, | |
| "grad_norm": 3.113304853439331, | |
| "learning_rate": 6.275665682928803e-05, | |
| "loss": 2.8372, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 0.15485843550836165, | |
| "grad_norm": 2.803987979888916, | |
| "learning_rate": 6.256106304328905e-05, | |
| "loss": 2.2549, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 0.15502370598382414, | |
| "grad_norm": 2.8876821994781494, | |
| "learning_rate": 6.23656356945497e-05, | |
| "loss": 1.9143, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 0.15518897645928664, | |
| "grad_norm": 3.334991216659546, | |
| "learning_rate": 6.21703756518551e-05, | |
| "loss": 2.1132, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 0.15535424693474914, | |
| "grad_norm": 3.5040950775146484, | |
| "learning_rate": 6.197528378324665e-05, | |
| "loss": 2.6805, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.15551951741021164, | |
| "grad_norm": 3.608353853225708, | |
| "learning_rate": 6.17803609560181e-05, | |
| "loss": 2.5346, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 0.15568478788567414, | |
| "grad_norm": 4.339207649230957, | |
| "learning_rate": 6.158560803671168e-05, | |
| "loss": 2.4727, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 0.15585005836113663, | |
| "grad_norm": 4.191640377044678, | |
| "learning_rate": 6.139102589111435e-05, | |
| "loss": 2.269, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 0.15601532883659916, | |
| "grad_norm": 4.738291263580322, | |
| "learning_rate": 6.119661538425381e-05, | |
| "loss": 2.4109, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 0.15618059931206166, | |
| "grad_norm": 4.653080463409424, | |
| "learning_rate": 6.100237738039484e-05, | |
| "loss": 2.1349, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.15634586978752416, | |
| "grad_norm": 4.324831485748291, | |
| "learning_rate": 6.0808312743035236e-05, | |
| "loss": 1.7562, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 0.15651114026298665, | |
| "grad_norm": 4.334379196166992, | |
| "learning_rate": 6.061442233490211e-05, | |
| "loss": 1.8517, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 0.15667641073844915, | |
| "grad_norm": 5.483864784240723, | |
| "learning_rate": 6.042070701794806e-05, | |
| "loss": 2.3008, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 0.15684168121391165, | |
| "grad_norm": 7.2431535720825195, | |
| "learning_rate": 6.0227167653347305e-05, | |
| "loss": 2.7862, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 0.15700695168937415, | |
| "grad_norm": 10.70998477935791, | |
| "learning_rate": 6.0033805101491794e-05, | |
| "loss": 3.4004, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.15717222216483664, | |
| "grad_norm": 0.5935502052307129, | |
| "learning_rate": 5.98406202219875e-05, | |
| "loss": 1.0626, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 0.15733749264029914, | |
| "grad_norm": 0.5730698704719543, | |
| "learning_rate": 5.964761387365052e-05, | |
| "loss": 1.295, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 0.15750276311576164, | |
| "grad_norm": 0.5663546919822693, | |
| "learning_rate": 5.9454786914503255e-05, | |
| "loss": 1.629, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 0.15766803359122414, | |
| "grad_norm": 0.6374579668045044, | |
| "learning_rate": 5.926214020177074e-05, | |
| "loss": 1.6356, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 0.15783330406668664, | |
| "grad_norm": 0.9674170613288879, | |
| "learning_rate": 5.9069674591876534e-05, | |
| "loss": 2.3043, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.15799857454214913, | |
| "grad_norm": 0.7713983058929443, | |
| "learning_rate": 5.887739094043923e-05, | |
| "loss": 2.0488, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 0.15816384501761163, | |
| "grad_norm": 0.7409908771514893, | |
| "learning_rate": 5.868529010226845e-05, | |
| "loss": 2.0903, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 0.15832911549307413, | |
| "grad_norm": 0.7113223075866699, | |
| "learning_rate": 5.849337293136112e-05, | |
| "loss": 2.0287, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 0.15849438596853663, | |
| "grad_norm": 0.7567322254180908, | |
| "learning_rate": 5.830164028089766e-05, | |
| "loss": 1.8374, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 0.15865965644399913, | |
| "grad_norm": 0.8666066527366638, | |
| "learning_rate": 5.811009300323818e-05, | |
| "loss": 2.1452, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.15882492691946162, | |
| "grad_norm": 0.8505182266235352, | |
| "learning_rate": 5.791873194991872e-05, | |
| "loss": 1.9542, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 0.15899019739492412, | |
| "grad_norm": 1.0273040533065796, | |
| "learning_rate": 5.7727557971647427e-05, | |
| "loss": 2.4823, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 0.15915546787038662, | |
| "grad_norm": 0.9564617872238159, | |
| "learning_rate": 5.7536571918300864e-05, | |
| "loss": 1.7382, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 0.15932073834584912, | |
| "grad_norm": 1.0788862705230713, | |
| "learning_rate": 5.734577463892008e-05, | |
| "loss": 2.2809, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 0.15948600882131161, | |
| "grad_norm": 1.185129165649414, | |
| "learning_rate": 5.7155166981706956e-05, | |
| "loss": 2.2785, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.15965127929677414, | |
| "grad_norm": 1.1670318841934204, | |
| "learning_rate": 5.6964749794020354e-05, | |
| "loss": 1.8238, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 0.15981654977223664, | |
| "grad_norm": 1.2945955991744995, | |
| "learning_rate": 5.6774523922372394e-05, | |
| "loss": 2.3842, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 0.15998182024769914, | |
| "grad_norm": 1.2796839475631714, | |
| "learning_rate": 5.6584490212424804e-05, | |
| "loss": 2.1622, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 0.16014709072316163, | |
| "grad_norm": 1.329115867614746, | |
| "learning_rate": 5.639464950898491e-05, | |
| "loss": 1.9625, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 0.16031236119862413, | |
| "grad_norm": 1.6704708337783813, | |
| "learning_rate": 5.620500265600206e-05, | |
| "loss": 2.6795, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.16047763167408663, | |
| "grad_norm": 1.59796941280365, | |
| "learning_rate": 5.601555049656382e-05, | |
| "loss": 2.4847, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 0.16064290214954913, | |
| "grad_norm": 1.7455967664718628, | |
| "learning_rate": 5.58262938728922e-05, | |
| "loss": 2.2844, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 0.16080817262501163, | |
| "grad_norm": 1.7863965034484863, | |
| "learning_rate": 5.563723362634008e-05, | |
| "loss": 2.3612, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 0.16097344310047412, | |
| "grad_norm": 1.8732479810714722, | |
| "learning_rate": 5.544837059738719e-05, | |
| "loss": 2.3577, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 0.16113871357593662, | |
| "grad_norm": 1.9350277185440063, | |
| "learning_rate": 5.525970562563656e-05, | |
| "loss": 2.2167, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.16130398405139912, | |
| "grad_norm": 1.7780574560165405, | |
| "learning_rate": 5.507123954981073e-05, | |
| "loss": 2.1837, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 0.16146925452686162, | |
| "grad_norm": 2.0393459796905518, | |
| "learning_rate": 5.488297320774807e-05, | |
| "loss": 2.19, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 0.16163452500232411, | |
| "grad_norm": 2.17767596244812, | |
| "learning_rate": 5.4694907436399e-05, | |
| "loss": 2.5205, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 0.1617997954777866, | |
| "grad_norm": 2.2335870265960693, | |
| "learning_rate": 5.4507043071822284e-05, | |
| "loss": 2.465, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 0.1619650659532491, | |
| "grad_norm": 2.2167108058929443, | |
| "learning_rate": 5.431938094918132e-05, | |
| "loss": 2.7937, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.1621303364287116, | |
| "grad_norm": 2.5321850776672363, | |
| "learning_rate": 5.41319219027404e-05, | |
| "loss": 2.8456, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 0.1622956069041741, | |
| "grad_norm": 2.4527344703674316, | |
| "learning_rate": 5.394466676586114e-05, | |
| "loss": 2.688, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 0.1624608773796366, | |
| "grad_norm": 2.6821136474609375, | |
| "learning_rate": 5.375761637099854e-05, | |
| "loss": 2.3072, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 0.1626261478550991, | |
| "grad_norm": 2.780740976333618, | |
| "learning_rate": 5.357077154969742e-05, | |
| "loss": 2.2209, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 0.1627914183305616, | |
| "grad_norm": 2.737525224685669, | |
| "learning_rate": 5.3384133132588784e-05, | |
| "loss": 2.4298, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.1629566888060241, | |
| "grad_norm": 3.062124013900757, | |
| "learning_rate": 5.3197701949386e-05, | |
| "loss": 2.631, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 0.1631219592814866, | |
| "grad_norm": 3.2128493785858154, | |
| "learning_rate": 5.301147882888116e-05, | |
| "loss": 2.1863, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 0.16328722975694912, | |
| "grad_norm": 3.551666498184204, | |
| "learning_rate": 5.28254645989414e-05, | |
| "loss": 2.5422, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 0.16345250023241162, | |
| "grad_norm": 3.4580109119415283, | |
| "learning_rate": 5.2639660086505226e-05, | |
| "loss": 2.0963, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 0.16361777070787412, | |
| "grad_norm": 4.194453716278076, | |
| "learning_rate": 5.2454066117578815e-05, | |
| "loss": 2.5536, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.16378304118333661, | |
| "grad_norm": 3.324228286743164, | |
| "learning_rate": 5.226868351723244e-05, | |
| "loss": 2.4779, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 0.1639483116587991, | |
| "grad_norm": 4.007864475250244, | |
| "learning_rate": 5.2083513109596616e-05, | |
| "loss": 2.3074, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 0.1641135821342616, | |
| "grad_norm": 4.593420028686523, | |
| "learning_rate": 5.189855571785859e-05, | |
| "loss": 2.2186, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 0.1642788526097241, | |
| "grad_norm": 4.065645217895508, | |
| "learning_rate": 5.171381216425863e-05, | |
| "loss": 2.0566, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 0.1644441230851866, | |
| "grad_norm": 4.7412004470825195, | |
| "learning_rate": 5.152928327008635e-05, | |
| "loss": 2.1046, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.1646093935606491, | |
| "grad_norm": 4.305594444274902, | |
| "learning_rate": 5.134496985567714e-05, | |
| "loss": 1.9874, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 0.1647746640361116, | |
| "grad_norm": 5.443896770477295, | |
| "learning_rate": 5.116087274040837e-05, | |
| "loss": 1.8882, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 0.1649399345115741, | |
| "grad_norm": 5.245382785797119, | |
| "learning_rate": 5.0976992742695925e-05, | |
| "loss": 2.0924, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 0.1651052049870366, | |
| "grad_norm": 6.249019622802734, | |
| "learning_rate": 5.07933306799904e-05, | |
| "loss": 2.3488, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 0.1652704754624991, | |
| "grad_norm": 10.814421653747559, | |
| "learning_rate": 5.060988736877366e-05, | |
| "loss": 3.5877, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.1654357459379616, | |
| "grad_norm": 0.7541307806968689, | |
| "learning_rate": 5.042666362455498e-05, | |
| "loss": 1.0407, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 0.1656010164134241, | |
| "grad_norm": 0.6568590998649597, | |
| "learning_rate": 5.024366026186755e-05, | |
| "loss": 1.3005, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 0.1657662868888866, | |
| "grad_norm": 0.5757076144218445, | |
| "learning_rate": 5.006087809426496e-05, | |
| "loss": 1.3772, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 0.16593155736434909, | |
| "grad_norm": 0.6687923073768616, | |
| "learning_rate": 4.987831793431731e-05, | |
| "loss": 1.7875, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 0.16609682783981158, | |
| "grad_norm": 0.653105616569519, | |
| "learning_rate": 4.9695980593607817e-05, | |
| "loss": 2.2733, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.16626209831527408, | |
| "grad_norm": 0.6917572617530823, | |
| "learning_rate": 4.9513866882729146e-05, | |
| "loss": 1.9693, | |
| "step": 1006 | |
| }, | |
| { | |
| "epoch": 0.16642736879073658, | |
| "grad_norm": 0.8558650612831116, | |
| "learning_rate": 4.9331977611279777e-05, | |
| "loss": 2.2021, | |
| "step": 1007 | |
| }, | |
| { | |
| "epoch": 0.16659263926619908, | |
| "grad_norm": 0.789960503578186, | |
| "learning_rate": 4.9150313587860433e-05, | |
| "loss": 1.884, | |
| "step": 1008 | |
| }, | |
| { | |
| "epoch": 0.16675790974166158, | |
| "grad_norm": 0.8318465948104858, | |
| "learning_rate": 4.896887562007054e-05, | |
| "loss": 2.152, | |
| "step": 1009 | |
| }, | |
| { | |
| "epoch": 0.1669231802171241, | |
| "grad_norm": 0.8718820214271545, | |
| "learning_rate": 4.8787664514504504e-05, | |
| "loss": 2.3621, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.1670884506925866, | |
| "grad_norm": 0.8635246157646179, | |
| "learning_rate": 4.860668107674823e-05, | |
| "loss": 1.8975, | |
| "step": 1011 | |
| }, | |
| { | |
| "epoch": 0.1672537211680491, | |
| "grad_norm": 1.0594284534454346, | |
| "learning_rate": 4.8425926111375506e-05, | |
| "loss": 2.5349, | |
| "step": 1012 | |
| }, | |
| { | |
| "epoch": 0.1674189916435116, | |
| "grad_norm": 1.0077821016311646, | |
| "learning_rate": 4.824540042194443e-05, | |
| "loss": 2.3844, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 0.1675842621189741, | |
| "grad_norm": 1.0180336236953735, | |
| "learning_rate": 4.8065104810993856e-05, | |
| "loss": 2.1807, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 0.1677495325944366, | |
| "grad_norm": 1.1051278114318848, | |
| "learning_rate": 4.788504008003978e-05, | |
| "loss": 2.2008, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.1679148030698991, | |
| "grad_norm": 1.114028811454773, | |
| "learning_rate": 4.770520702957182e-05, | |
| "loss": 1.9685, | |
| "step": 1016 | |
| }, | |
| { | |
| "epoch": 0.16808007354536159, | |
| "grad_norm": 1.3594282865524292, | |
| "learning_rate": 4.752560645904962e-05, | |
| "loss": 1.9471, | |
| "step": 1017 | |
| }, | |
| { | |
| "epoch": 0.16824534402082408, | |
| "grad_norm": 1.3758127689361572, | |
| "learning_rate": 4.734623916689941e-05, | |
| "loss": 2.0084, | |
| "step": 1018 | |
| }, | |
| { | |
| "epoch": 0.16841061449628658, | |
| "grad_norm": 1.4578577280044556, | |
| "learning_rate": 4.716710595051022e-05, | |
| "loss": 2.137, | |
| "step": 1019 | |
| }, | |
| { | |
| "epoch": 0.16857588497174908, | |
| "grad_norm": 1.4129718542099, | |
| "learning_rate": 4.698820760623064e-05, | |
| "loss": 2.3047, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.16874115544721158, | |
| "grad_norm": 1.7225353717803955, | |
| "learning_rate": 4.6809544929365004e-05, | |
| "loss": 2.474, | |
| "step": 1021 | |
| }, | |
| { | |
| "epoch": 0.16890642592267407, | |
| "grad_norm": 1.5706661939620972, | |
| "learning_rate": 4.663111871417e-05, | |
| "loss": 2.362, | |
| "step": 1022 | |
| }, | |
| { | |
| "epoch": 0.16907169639813657, | |
| "grad_norm": 1.6746793985366821, | |
| "learning_rate": 4.645292975385111e-05, | |
| "loss": 2.1768, | |
| "step": 1023 | |
| }, | |
| { | |
| "epoch": 0.16923696687359907, | |
| "grad_norm": 2.165961503982544, | |
| "learning_rate": 4.627497884055912e-05, | |
| "loss": 2.8457, | |
| "step": 1024 | |
| }, | |
| { | |
| "epoch": 0.16940223734906157, | |
| "grad_norm": 2.038978099822998, | |
| "learning_rate": 4.609726676538652e-05, | |
| "loss": 2.4012, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.16956750782452407, | |
| "grad_norm": 1.8751431703567505, | |
| "learning_rate": 4.591979431836402e-05, | |
| "loss": 2.5528, | |
| "step": 1026 | |
| }, | |
| { | |
| "epoch": 0.16973277829998656, | |
| "grad_norm": 2.311516761779785, | |
| "learning_rate": 4.574256228845706e-05, | |
| "loss": 2.3069, | |
| "step": 1027 | |
| }, | |
| { | |
| "epoch": 0.16989804877544906, | |
| "grad_norm": 2.6974129676818848, | |
| "learning_rate": 4.5565571463562365e-05, | |
| "loss": 2.6734, | |
| "step": 1028 | |
| }, | |
| { | |
| "epoch": 0.17006331925091156, | |
| "grad_norm": 2.457087993621826, | |
| "learning_rate": 4.5388822630504256e-05, | |
| "loss": 2.4316, | |
| "step": 1029 | |
| }, | |
| { | |
| "epoch": 0.17022858972637406, | |
| "grad_norm": 2.273343086242676, | |
| "learning_rate": 4.521231657503132e-05, | |
| "loss": 2.4713, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.17039386020183656, | |
| "grad_norm": 2.4276673793792725, | |
| "learning_rate": 4.503605408181286e-05, | |
| "loss": 2.2425, | |
| "step": 1031 | |
| }, | |
| { | |
| "epoch": 0.17055913067729905, | |
| "grad_norm": 2.332921266555786, | |
| "learning_rate": 4.486003593443537e-05, | |
| "loss": 2.6586, | |
| "step": 1032 | |
| }, | |
| { | |
| "epoch": 0.17072440115276158, | |
| "grad_norm": 2.7408506870269775, | |
| "learning_rate": 4.468426291539914e-05, | |
| "loss": 2.457, | |
| "step": 1033 | |
| }, | |
| { | |
| "epoch": 0.17088967162822408, | |
| "grad_norm": 3.3169167041778564, | |
| "learning_rate": 4.4508735806114654e-05, | |
| "loss": 2.5411, | |
| "step": 1034 | |
| }, | |
| { | |
| "epoch": 0.17105494210368657, | |
| "grad_norm": 2.913501501083374, | |
| "learning_rate": 4.433345538689929e-05, | |
| "loss": 2.6742, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.17122021257914907, | |
| "grad_norm": 2.984902858734131, | |
| "learning_rate": 4.415842243697369e-05, | |
| "loss": 2.5198, | |
| "step": 1036 | |
| }, | |
| { | |
| "epoch": 0.17138548305461157, | |
| "grad_norm": 3.4609689712524414, | |
| "learning_rate": 4.39836377344583e-05, | |
| "loss": 2.3073, | |
| "step": 1037 | |
| }, | |
| { | |
| "epoch": 0.17155075353007407, | |
| "grad_norm": 3.1210947036743164, | |
| "learning_rate": 4.380910205637012e-05, | |
| "loss": 1.9781, | |
| "step": 1038 | |
| }, | |
| { | |
| "epoch": 0.17171602400553657, | |
| "grad_norm": 4.232224464416504, | |
| "learning_rate": 4.363481617861893e-05, | |
| "loss": 2.7109, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 0.17188129448099906, | |
| "grad_norm": 3.564352035522461, | |
| "learning_rate": 4.346078087600412e-05, | |
| "loss": 2.2455, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.17204656495646156, | |
| "grad_norm": 3.8418354988098145, | |
| "learning_rate": 4.3286996922211034e-05, | |
| "loss": 2.6493, | |
| "step": 1041 | |
| }, | |
| { | |
| "epoch": 0.17221183543192406, | |
| "grad_norm": 4.232987880706787, | |
| "learning_rate": 4.311346508980772e-05, | |
| "loss": 2.1566, | |
| "step": 1042 | |
| }, | |
| { | |
| "epoch": 0.17237710590738656, | |
| "grad_norm": 4.377345085144043, | |
| "learning_rate": 4.2940186150241365e-05, | |
| "loss": 2.5609, | |
| "step": 1043 | |
| }, | |
| { | |
| "epoch": 0.17254237638284906, | |
| "grad_norm": 4.043179988861084, | |
| "learning_rate": 4.27671608738349e-05, | |
| "loss": 1.9728, | |
| "step": 1044 | |
| }, | |
| { | |
| "epoch": 0.17270764685831155, | |
| "grad_norm": 4.7065629959106445, | |
| "learning_rate": 4.2594390029783534e-05, | |
| "loss": 2.2058, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.17287291733377405, | |
| "grad_norm": 5.319474220275879, | |
| "learning_rate": 4.242187438615153e-05, | |
| "loss": 1.6211, | |
| "step": 1046 | |
| }, | |
| { | |
| "epoch": 0.17303818780923655, | |
| "grad_norm": 6.218306541442871, | |
| "learning_rate": 4.224961470986849e-05, | |
| "loss": 2.5782, | |
| "step": 1047 | |
| }, | |
| { | |
| "epoch": 0.17320345828469905, | |
| "grad_norm": 5.0891289710998535, | |
| "learning_rate": 4.207761176672614e-05, | |
| "loss": 2.1981, | |
| "step": 1048 | |
| }, | |
| { | |
| "epoch": 0.17336872876016154, | |
| "grad_norm": 7.03289794921875, | |
| "learning_rate": 4.190586632137491e-05, | |
| "loss": 2.4313, | |
| "step": 1049 | |
| }, | |
| { | |
| "epoch": 0.17353399923562404, | |
| "grad_norm": 8.721726417541504, | |
| "learning_rate": 4.173437913732048e-05, | |
| "loss": 3.1351, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.17369926971108654, | |
| "grad_norm": 0.7477954626083374, | |
| "learning_rate": 4.156315097692037e-05, | |
| "loss": 1.1234, | |
| "step": 1051 | |
| }, | |
| { | |
| "epoch": 0.17386454018654904, | |
| "grad_norm": 0.5343180298805237, | |
| "learning_rate": 4.139218260138074e-05, | |
| "loss": 1.2159, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 0.17402981066201154, | |
| "grad_norm": 0.5169820785522461, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 1.2448, | |
| "step": 1053 | |
| }, | |
| { | |
| "epoch": 0.17419508113747403, | |
| "grad_norm": 0.667062520980835, | |
| "learning_rate": 4.1051028243929125e-05, | |
| "loss": 1.711, | |
| "step": 1054 | |
| }, | |
| { | |
| "epoch": 0.17436035161293656, | |
| "grad_norm": 0.6438874006271362, | |
| "learning_rate": 4.088084377864135e-05, | |
| "loss": 2.1356, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.17452562208839906, | |
| "grad_norm": 0.807325541973114, | |
| "learning_rate": 4.07109221314556e-05, | |
| "loss": 2.1382, | |
| "step": 1056 | |
| }, | |
| { | |
| "epoch": 0.17469089256386156, | |
| "grad_norm": 0.8036754131317139, | |
| "learning_rate": 4.054126405776971e-05, | |
| "loss": 2.1485, | |
| "step": 1057 | |
| }, | |
| { | |
| "epoch": 0.17485616303932405, | |
| "grad_norm": 0.8354204893112183, | |
| "learning_rate": 4.037187031180985e-05, | |
| "loss": 2.103, | |
| "step": 1058 | |
| }, | |
| { | |
| "epoch": 0.17502143351478655, | |
| "grad_norm": 0.8434606194496155, | |
| "learning_rate": 4.020274164662707e-05, | |
| "loss": 2.2437, | |
| "step": 1059 | |
| }, | |
| { | |
| "epoch": 0.17518670399024905, | |
| "grad_norm": 0.847907304763794, | |
| "learning_rate": 4.003387881409397e-05, | |
| "loss": 2.0132, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.17535197446571155, | |
| "grad_norm": 0.8847858309745789, | |
| "learning_rate": 3.986528256490141e-05, | |
| "loss": 2.2499, | |
| "step": 1061 | |
| }, | |
| { | |
| "epoch": 0.17551724494117404, | |
| "grad_norm": 0.9525457620620728, | |
| "learning_rate": 3.969695364855511e-05, | |
| "loss": 2.4571, | |
| "step": 1062 | |
| }, | |
| { | |
| "epoch": 0.17568251541663654, | |
| "grad_norm": 0.9390000700950623, | |
| "learning_rate": 3.952889281337235e-05, | |
| "loss": 1.9478, | |
| "step": 1063 | |
| }, | |
| { | |
| "epoch": 0.17584778589209904, | |
| "grad_norm": 1.0419167280197144, | |
| "learning_rate": 3.93611008064786e-05, | |
| "loss": 2.1195, | |
| "step": 1064 | |
| }, | |
| { | |
| "epoch": 0.17601305636756154, | |
| "grad_norm": 1.073873519897461, | |
| "learning_rate": 3.9193578373804364e-05, | |
| "loss": 1.8843, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.17617832684302404, | |
| "grad_norm": 1.1356242895126343, | |
| "learning_rate": 3.90263262600816e-05, | |
| "loss": 2.0141, | |
| "step": 1066 | |
| }, | |
| { | |
| "epoch": 0.17634359731848653, | |
| "grad_norm": 1.1506868600845337, | |
| "learning_rate": 3.88593452088406e-05, | |
| "loss": 2.3498, | |
| "step": 1067 | |
| }, | |
| { | |
| "epoch": 0.17650886779394903, | |
| "grad_norm": 1.318650722503662, | |
| "learning_rate": 3.869263596240661e-05, | |
| "loss": 1.7925, | |
| "step": 1068 | |
| }, | |
| { | |
| "epoch": 0.17667413826941153, | |
| "grad_norm": 1.4005818367004395, | |
| "learning_rate": 3.8526199261896544e-05, | |
| "loss": 2.2308, | |
| "step": 1069 | |
| }, | |
| { | |
| "epoch": 0.17683940874487403, | |
| "grad_norm": 1.3406119346618652, | |
| "learning_rate": 3.836003584721577e-05, | |
| "loss": 2.2432, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.17700467922033652, | |
| "grad_norm": 1.3795819282531738, | |
| "learning_rate": 3.8194146457054655e-05, | |
| "loss": 2.3123, | |
| "step": 1071 | |
| }, | |
| { | |
| "epoch": 0.17716994969579902, | |
| "grad_norm": 1.3522512912750244, | |
| "learning_rate": 3.802853182888543e-05, | |
| "loss": 2.1964, | |
| "step": 1072 | |
| }, | |
| { | |
| "epoch": 0.17733522017126152, | |
| "grad_norm": 1.5563820600509644, | |
| "learning_rate": 3.786319269895877e-05, | |
| "loss": 2.3375, | |
| "step": 1073 | |
| }, | |
| { | |
| "epoch": 0.17750049064672402, | |
| "grad_norm": 1.721183180809021, | |
| "learning_rate": 3.769812980230074e-05, | |
| "loss": 2.401, | |
| "step": 1074 | |
| }, | |
| { | |
| "epoch": 0.17766576112218652, | |
| "grad_norm": 1.7993472814559937, | |
| "learning_rate": 3.7533343872709294e-05, | |
| "loss": 2.2281, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.17783103159764901, | |
| "grad_norm": 1.783200740814209, | |
| "learning_rate": 3.736883564275112e-05, | |
| "loss": 1.8046, | |
| "step": 1076 | |
| }, | |
| { | |
| "epoch": 0.17799630207311154, | |
| "grad_norm": 1.9802608489990234, | |
| "learning_rate": 3.7204605843758386e-05, | |
| "loss": 2.85, | |
| "step": 1077 | |
| }, | |
| { | |
| "epoch": 0.17816157254857404, | |
| "grad_norm": 2.001185894012451, | |
| "learning_rate": 3.704065520582549e-05, | |
| "loss": 2.2455, | |
| "step": 1078 | |
| }, | |
| { | |
| "epoch": 0.17832684302403654, | |
| "grad_norm": 2.132444143295288, | |
| "learning_rate": 3.6876984457805786e-05, | |
| "loss": 2.5723, | |
| "step": 1079 | |
| }, | |
| { | |
| "epoch": 0.17849211349949903, | |
| "grad_norm": 2.3601012229919434, | |
| "learning_rate": 3.671359432730834e-05, | |
| "loss": 2.2365, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.17865738397496153, | |
| "grad_norm": 2.415203332901001, | |
| "learning_rate": 3.655048554069478e-05, | |
| "loss": 2.4875, | |
| "step": 1081 | |
| }, | |
| { | |
| "epoch": 0.17882265445042403, | |
| "grad_norm": 2.15446400642395, | |
| "learning_rate": 3.638765882307589e-05, | |
| "loss": 2.2508, | |
| "step": 1082 | |
| }, | |
| { | |
| "epoch": 0.17898792492588653, | |
| "grad_norm": 2.504662036895752, | |
| "learning_rate": 3.6225114898308634e-05, | |
| "loss": 2.4326, | |
| "step": 1083 | |
| }, | |
| { | |
| "epoch": 0.17915319540134902, | |
| "grad_norm": 2.3438305854797363, | |
| "learning_rate": 3.6062854488992714e-05, | |
| "loss": 1.9443, | |
| "step": 1084 | |
| }, | |
| { | |
| "epoch": 0.17931846587681152, | |
| "grad_norm": 2.519359827041626, | |
| "learning_rate": 3.5900878316467454e-05, | |
| "loss": 2.6052, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.17948373635227402, | |
| "grad_norm": 2.739835739135742, | |
| "learning_rate": 3.573918710080857e-05, | |
| "loss": 2.4006, | |
| "step": 1086 | |
| }, | |
| { | |
| "epoch": 0.17964900682773652, | |
| "grad_norm": 3.0523221492767334, | |
| "learning_rate": 3.5577781560825066e-05, | |
| "loss": 2.2859, | |
| "step": 1087 | |
| }, | |
| { | |
| "epoch": 0.17981427730319902, | |
| "grad_norm": 3.0159943103790283, | |
| "learning_rate": 3.541666241405588e-05, | |
| "loss": 2.3516, | |
| "step": 1088 | |
| }, | |
| { | |
| "epoch": 0.17997954777866151, | |
| "grad_norm": 3.3250668048858643, | |
| "learning_rate": 3.5255830376766764e-05, | |
| "loss": 2.5885, | |
| "step": 1089 | |
| }, | |
| { | |
| "epoch": 0.180144818254124, | |
| "grad_norm": 3.051696300506592, | |
| "learning_rate": 3.509528616394716e-05, | |
| "loss": 2.226, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.1803100887295865, | |
| "grad_norm": 3.435652732849121, | |
| "learning_rate": 3.4935030489306883e-05, | |
| "loss": 1.9904, | |
| "step": 1091 | |
| }, | |
| { | |
| "epoch": 0.180475359205049, | |
| "grad_norm": 3.8968424797058105, | |
| "learning_rate": 3.4775064065273165e-05, | |
| "loss": 2.3891, | |
| "step": 1092 | |
| }, | |
| { | |
| "epoch": 0.1806406296805115, | |
| "grad_norm": 3.81770396232605, | |
| "learning_rate": 3.4615387602987236e-05, | |
| "loss": 2.3543, | |
| "step": 1093 | |
| }, | |
| { | |
| "epoch": 0.180805900155974, | |
| "grad_norm": 3.6240451335906982, | |
| "learning_rate": 3.445600181230134e-05, | |
| "loss": 2.132, | |
| "step": 1094 | |
| }, | |
| { | |
| "epoch": 0.1809711706314365, | |
| "grad_norm": 4.504293441772461, | |
| "learning_rate": 3.429690740177549e-05, | |
| "loss": 2.3018, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.181136441106899, | |
| "grad_norm": 4.285729885101318, | |
| "learning_rate": 3.413810507867436e-05, | |
| "loss": 1.7408, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 0.1813017115823615, | |
| "grad_norm": 5.172756195068359, | |
| "learning_rate": 3.397959554896415e-05, | |
| "loss": 2.2746, | |
| "step": 1097 | |
| }, | |
| { | |
| "epoch": 0.181466982057824, | |
| "grad_norm": 5.05802583694458, | |
| "learning_rate": 3.3821379517309405e-05, | |
| "loss": 1.7984, | |
| "step": 1098 | |
| }, | |
| { | |
| "epoch": 0.18163225253328652, | |
| "grad_norm": 5.701052665710449, | |
| "learning_rate": 3.3663457687069924e-05, | |
| "loss": 2.2654, | |
| "step": 1099 | |
| }, | |
| { | |
| "epoch": 0.18179752300874902, | |
| "grad_norm": 8.087822914123535, | |
| "learning_rate": 3.350583076029754e-05, | |
| "loss": 2.9708, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.18196279348421152, | |
| "grad_norm": 0.5673682689666748, | |
| "learning_rate": 3.334849943773323e-05, | |
| "loss": 1.1877, | |
| "step": 1101 | |
| }, | |
| { | |
| "epoch": 0.182128063959674, | |
| "grad_norm": 0.6245754361152649, | |
| "learning_rate": 3.319146441880371e-05, | |
| "loss": 1.3206, | |
| "step": 1102 | |
| }, | |
| { | |
| "epoch": 0.1822933344351365, | |
| "grad_norm": 0.5597710013389587, | |
| "learning_rate": 3.3034726401618444e-05, | |
| "loss": 1.3792, | |
| "step": 1103 | |
| }, | |
| { | |
| "epoch": 0.182458604910599, | |
| "grad_norm": 0.7274876832962036, | |
| "learning_rate": 3.28782860829667e-05, | |
| "loss": 1.826, | |
| "step": 1104 | |
| }, | |
| { | |
| "epoch": 0.1826238753860615, | |
| "grad_norm": 0.6294060349464417, | |
| "learning_rate": 3.272214415831418e-05, | |
| "loss": 1.992, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.182789145861524, | |
| "grad_norm": 0.679905891418457, | |
| "learning_rate": 3.2566301321800085e-05, | |
| "loss": 2.0683, | |
| "step": 1106 | |
| }, | |
| { | |
| "epoch": 0.1829544163369865, | |
| "grad_norm": 0.7083233594894409, | |
| "learning_rate": 3.241075826623401e-05, | |
| "loss": 1.7579, | |
| "step": 1107 | |
| }, | |
| { | |
| "epoch": 0.183119686812449, | |
| "grad_norm": 0.7589206099510193, | |
| "learning_rate": 3.225551568309284e-05, | |
| "loss": 1.8437, | |
| "step": 1108 | |
| }, | |
| { | |
| "epoch": 0.1832849572879115, | |
| "grad_norm": 0.8503727316856384, | |
| "learning_rate": 3.210057426251773e-05, | |
| "loss": 1.9717, | |
| "step": 1109 | |
| }, | |
| { | |
| "epoch": 0.183450227763374, | |
| "grad_norm": 0.865169107913971, | |
| "learning_rate": 3.1945934693310896e-05, | |
| "loss": 1.9272, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.1836154982388365, | |
| "grad_norm": 0.7725500464439392, | |
| "learning_rate": 3.179159766293282e-05, | |
| "loss": 2.0634, | |
| "step": 1111 | |
| }, | |
| { | |
| "epoch": 0.183780768714299, | |
| "grad_norm": 0.8942416906356812, | |
| "learning_rate": 3.163756385749889e-05, | |
| "loss": 2.2611, | |
| "step": 1112 | |
| }, | |
| { | |
| "epoch": 0.1839460391897615, | |
| "grad_norm": 1.0037657022476196, | |
| "learning_rate": 3.148383396177653e-05, | |
| "loss": 2.3662, | |
| "step": 1113 | |
| }, | |
| { | |
| "epoch": 0.184111309665224, | |
| "grad_norm": 0.9730199575424194, | |
| "learning_rate": 3.133040865918213e-05, | |
| "loss": 2.3259, | |
| "step": 1114 | |
| }, | |
| { | |
| "epoch": 0.18427658014068649, | |
| "grad_norm": 1.0028016567230225, | |
| "learning_rate": 3.117728863177796e-05, | |
| "loss": 2.0115, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.18444185061614898, | |
| "grad_norm": 1.214603066444397, | |
| "learning_rate": 3.102447456026919e-05, | |
| "loss": 2.2606, | |
| "step": 1116 | |
| }, | |
| { | |
| "epoch": 0.18460712109161148, | |
| "grad_norm": 1.192114233970642, | |
| "learning_rate": 3.0871967124000834e-05, | |
| "loss": 2.033, | |
| "step": 1117 | |
| }, | |
| { | |
| "epoch": 0.18477239156707398, | |
| "grad_norm": 1.2866182327270508, | |
| "learning_rate": 3.0719767000954714e-05, | |
| "loss": 2.2258, | |
| "step": 1118 | |
| }, | |
| { | |
| "epoch": 0.18493766204253648, | |
| "grad_norm": 1.335279107093811, | |
| "learning_rate": 3.056787486774656e-05, | |
| "loss": 1.9696, | |
| "step": 1119 | |
| }, | |
| { | |
| "epoch": 0.18510293251799897, | |
| "grad_norm": 1.5218226909637451, | |
| "learning_rate": 3.041629139962283e-05, | |
| "loss": 2.3885, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.1852682029934615, | |
| "grad_norm": 1.4813919067382812, | |
| "learning_rate": 3.0265017270457775e-05, | |
| "loss": 2.3, | |
| "step": 1121 | |
| }, | |
| { | |
| "epoch": 0.185433473468924, | |
| "grad_norm": 1.670596718788147, | |
| "learning_rate": 3.0114053152750556e-05, | |
| "loss": 2.2173, | |
| "step": 1122 | |
| }, | |
| { | |
| "epoch": 0.1855987439443865, | |
| "grad_norm": 1.7553831338882446, | |
| "learning_rate": 2.9963399717622077e-05, | |
| "loss": 2.3743, | |
| "step": 1123 | |
| }, | |
| { | |
| "epoch": 0.185764014419849, | |
| "grad_norm": 1.7241071462631226, | |
| "learning_rate": 2.98130576348121e-05, | |
| "loss": 2.4326, | |
| "step": 1124 | |
| }, | |
| { | |
| "epoch": 0.1859292848953115, | |
| "grad_norm": 1.8137153387069702, | |
| "learning_rate": 2.966302757267625e-05, | |
| "loss": 2.1402, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.1859292848953115, | |
| "eval_loss": 2.143037796020508, | |
| "eval_runtime": 208.6121, | |
| "eval_samples_per_second": 48.851, | |
| "eval_steps_per_second": 24.428, | |
| "step": 1125 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 375, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 8.861078061790003e+16, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |