{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.524300441826215, "eval_steps": 500, "global_step": 3000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0008415737428992215, "grad_norm": 1220866.75, "learning_rate": 0.0, "loss": 6.7045, "step": 1 }, { "epoch": 0.001683147485798443, "grad_norm": 1159900.75, "learning_rate": 1.0000000000000001e-07, "loss": 5.5719, "step": 2 }, { "epoch": 0.0025247212286976647, "grad_norm": 179539.0, "learning_rate": 2.0000000000000002e-07, "loss": 2.9643, "step": 3 }, { "epoch": 0.003366294971596886, "grad_norm": 60915.5234375, "learning_rate": 3.0000000000000004e-07, "loss": 2.3161, "step": 4 }, { "epoch": 0.004207868714496108, "grad_norm": 41937.33984375, "learning_rate": 4.0000000000000003e-07, "loss": 2.3402, "step": 5 }, { "epoch": 0.005049442457395329, "grad_norm": 25547.8046875, "learning_rate": 5.000000000000001e-07, "loss": 2.0138, "step": 6 }, { "epoch": 0.005891016200294551, "grad_norm": 22664.328125, "learning_rate": 6.000000000000001e-07, "loss": 2.0857, "step": 7 }, { "epoch": 0.006732589943193772, "grad_norm": 14687.037109375, "learning_rate": 7.000000000000001e-07, "loss": 2.0964, "step": 8 }, { "epoch": 0.007574163686092994, "grad_norm": 13980.6904296875, "learning_rate": 8.000000000000001e-07, "loss": 2.4171, "step": 9 }, { "epoch": 0.008415737428992216, "grad_norm": 9177.841796875, "learning_rate": 9.000000000000001e-07, "loss": 2.0959, "step": 10 }, { "epoch": 0.009257311171891438, "grad_norm": 7599.8466796875, "learning_rate": 1.0000000000000002e-06, "loss": 1.7852, "step": 11 }, { "epoch": 0.010098884914790659, "grad_norm": 6499.0732421875, "learning_rate": 1.1e-06, "loss": 1.6924, "step": 12 }, { "epoch": 0.01094045865768988, "grad_norm": 6869.11865234375, "learning_rate": 1.2000000000000002e-06, "loss": 1.6957, "step": 13 }, { "epoch": 0.011782032400589101, "grad_norm": 6527.53662109375, "learning_rate": 1.3e-06, "loss": 1.6633, "step": 14 }, { "epoch": 0.012623606143488323, "grad_norm": 6013.7412109375, "learning_rate": 1.4000000000000001e-06, "loss": 1.8745, "step": 15 }, { "epoch": 0.013465179886387544, "grad_norm": 7231.5712890625, "learning_rate": 1.5e-06, "loss": 1.5122, "step": 16 }, { "epoch": 0.014306753629286767, "grad_norm": 5269.26318359375, "learning_rate": 1.6000000000000001e-06, "loss": 1.8746, "step": 17 }, { "epoch": 0.015148327372185988, "grad_norm": 6138.6845703125, "learning_rate": 1.7000000000000002e-06, "loss": 1.9535, "step": 18 }, { "epoch": 0.01598990111508521, "grad_norm": 5054.857421875, "learning_rate": 1.8000000000000001e-06, "loss": 1.8551, "step": 19 }, { "epoch": 0.016831474857984433, "grad_norm": 5286.1708984375, "learning_rate": 1.9000000000000002e-06, "loss": 1.8266, "step": 20 }, { "epoch": 0.017673048600883652, "grad_norm": 3078.6181640625, "learning_rate": 2.0000000000000003e-06, "loss": 1.5621, "step": 21 }, { "epoch": 0.018514622343782875, "grad_norm": 4725.95654296875, "learning_rate": 2.1000000000000002e-06, "loss": 1.4557, "step": 22 }, { "epoch": 0.019356196086682095, "grad_norm": 3049.34375, "learning_rate": 2.2e-06, "loss": 1.6599, "step": 23 }, { "epoch": 0.020197769829581318, "grad_norm": 2984.544921875, "learning_rate": 2.3000000000000004e-06, "loss": 1.7747, "step": 24 }, { "epoch": 0.021039343572480537, "grad_norm": 6122.8681640625, "learning_rate": 2.4000000000000003e-06, "loss": 2.2994, "step": 25 }, { "epoch": 0.02188091731537976, "grad_norm": 2451.735107421875, "learning_rate": 2.5e-06, "loss": 1.6198, "step": 26 }, { "epoch": 0.022722491058278983, "grad_norm": 2788.764404296875, "learning_rate": 2.6e-06, "loss": 1.6456, "step": 27 }, { "epoch": 0.023564064801178203, "grad_norm": 2508.93408203125, "learning_rate": 2.7000000000000004e-06, "loss": 1.6448, "step": 28 }, { "epoch": 0.024405638544077426, "grad_norm": 2768.5078125, "learning_rate": 2.8000000000000003e-06, "loss": 1.7862, "step": 29 }, { "epoch": 0.025247212286976645, "grad_norm": 1970.8114013671875, "learning_rate": 2.9e-06, "loss": 1.6794, "step": 30 }, { "epoch": 0.02608878602987587, "grad_norm": 2229.092041015625, "learning_rate": 3e-06, "loss": 1.5213, "step": 31 }, { "epoch": 0.026930359772775088, "grad_norm": 2219.89990234375, "learning_rate": 3.1000000000000004e-06, "loss": 1.5728, "step": 32 }, { "epoch": 0.02777193351567431, "grad_norm": 2461.462158203125, "learning_rate": 3.2000000000000003e-06, "loss": 1.8106, "step": 33 }, { "epoch": 0.028613507258573534, "grad_norm": 1786.85693359375, "learning_rate": 3.3000000000000006e-06, "loss": 1.5036, "step": 34 }, { "epoch": 0.029455081001472753, "grad_norm": 1430.862060546875, "learning_rate": 3.4000000000000005e-06, "loss": 1.2056, "step": 35 }, { "epoch": 0.030296654744371977, "grad_norm": 1801.9127197265625, "learning_rate": 3.5e-06, "loss": 1.3607, "step": 36 }, { "epoch": 0.031138228487271196, "grad_norm": 1663.1851806640625, "learning_rate": 3.6000000000000003e-06, "loss": 1.4841, "step": 37 }, { "epoch": 0.03197980223017042, "grad_norm": 2743.8642578125, "learning_rate": 3.7e-06, "loss": 1.629, "step": 38 }, { "epoch": 0.03282137597306964, "grad_norm": 1889.03662109375, "learning_rate": 3.8000000000000005e-06, "loss": 1.5254, "step": 39 }, { "epoch": 0.033662949715968865, "grad_norm": 1672.36279296875, "learning_rate": 3.900000000000001e-06, "loss": 1.3192, "step": 40 }, { "epoch": 0.034504523458868085, "grad_norm": 1701.8798828125, "learning_rate": 4.000000000000001e-06, "loss": 1.5442, "step": 41 }, { "epoch": 0.035346097201767304, "grad_norm": 1488.308349609375, "learning_rate": 4.1e-06, "loss": 1.7156, "step": 42 }, { "epoch": 0.036187670944666524, "grad_norm": 1526.427001953125, "learning_rate": 4.2000000000000004e-06, "loss": 1.8971, "step": 43 }, { "epoch": 0.03702924468756575, "grad_norm": 1431.5985107421875, "learning_rate": 4.3e-06, "loss": 1.7197, "step": 44 }, { "epoch": 0.03787081843046497, "grad_norm": 1157.3121337890625, "learning_rate": 4.4e-06, "loss": 1.1669, "step": 45 }, { "epoch": 0.03871239217336419, "grad_norm": 1217.6405029296875, "learning_rate": 4.5e-06, "loss": 1.7516, "step": 46 }, { "epoch": 0.039553965916263416, "grad_norm": 1782.4837646484375, "learning_rate": 4.600000000000001e-06, "loss": 1.5455, "step": 47 }, { "epoch": 0.040395539659162635, "grad_norm": 1759.3140869140625, "learning_rate": 4.7e-06, "loss": 1.559, "step": 48 }, { "epoch": 0.041237113402061855, "grad_norm": 1439.0577392578125, "learning_rate": 4.800000000000001e-06, "loss": 1.6599, "step": 49 }, { "epoch": 0.042078687144961074, "grad_norm": 1564.3431396484375, "learning_rate": 4.9000000000000005e-06, "loss": 1.666, "step": 50 }, { "epoch": 0.0429202608878603, "grad_norm": 1380.912109375, "learning_rate": 5e-06, "loss": 1.6822, "step": 51 }, { "epoch": 0.04376183463075952, "grad_norm": 1470.3021240234375, "learning_rate": 4.999999002610039e-06, "loss": 1.5958, "step": 52 }, { "epoch": 0.04460340837365874, "grad_norm": 2070.1259765625, "learning_rate": 4.999996010440949e-06, "loss": 1.599, "step": 53 }, { "epoch": 0.045444982116557966, "grad_norm": 1041.359375, "learning_rate": 4.999991023495121e-06, "loss": 1.5113, "step": 54 }, { "epoch": 0.046286555859457186, "grad_norm": 917.322998046875, "learning_rate": 4.9999840417765314e-06, "loss": 1.7166, "step": 55 }, { "epoch": 0.047128129602356406, "grad_norm": 1014.3453979492188, "learning_rate": 4.999975065290751e-06, "loss": 1.4541, "step": 56 }, { "epoch": 0.047969703345255625, "grad_norm": 1376.1556396484375, "learning_rate": 4.999964094044944e-06, "loss": 1.5334, "step": 57 }, { "epoch": 0.04881127708815485, "grad_norm": 763.3536376953125, "learning_rate": 4.999951128047863e-06, "loss": 1.4616, "step": 58 }, { "epoch": 0.04965285083105407, "grad_norm": 782.1888427734375, "learning_rate": 4.999936167309855e-06, "loss": 1.2268, "step": 59 }, { "epoch": 0.05049442457395329, "grad_norm": 1264.4327392578125, "learning_rate": 4.999919211842856e-06, "loss": 1.8287, "step": 60 }, { "epoch": 0.05133599831685252, "grad_norm": 792.3248901367188, "learning_rate": 4.999900261660395e-06, "loss": 1.4994, "step": 61 }, { "epoch": 0.05217757205975174, "grad_norm": 707.390380859375, "learning_rate": 4.999879316777594e-06, "loss": 1.4387, "step": 62 }, { "epoch": 0.053019145802650956, "grad_norm": 753.5997924804688, "learning_rate": 4.999856377211164e-06, "loss": 1.3524, "step": 63 }, { "epoch": 0.053860719545550176, "grad_norm": 1315.026611328125, "learning_rate": 4.99983144297941e-06, "loss": 1.438, "step": 64 }, { "epoch": 0.0547022932884494, "grad_norm": 999.6796875, "learning_rate": 4.999804514102226e-06, "loss": 1.6885, "step": 65 }, { "epoch": 0.05554386703134862, "grad_norm": 1015.9467163085938, "learning_rate": 4.999775590601098e-06, "loss": 1.5048, "step": 66 }, { "epoch": 0.05638544077424784, "grad_norm": 904.549560546875, "learning_rate": 4.999744672499107e-06, "loss": 1.3633, "step": 67 }, { "epoch": 0.05722701451714707, "grad_norm": 802.7755737304688, "learning_rate": 4.999711759820921e-06, "loss": 1.3909, "step": 68 }, { "epoch": 0.05806858826004629, "grad_norm": 988.9061279296875, "learning_rate": 4.999676852592803e-06, "loss": 1.7298, "step": 69 }, { "epoch": 0.05891016200294551, "grad_norm": 804.6920776367188, "learning_rate": 4.9996399508426045e-06, "loss": 1.5774, "step": 70 }, { "epoch": 0.059751735745844727, "grad_norm": 983.9691772460938, "learning_rate": 4.99960105459977e-06, "loss": 1.4138, "step": 71 }, { "epoch": 0.06059330948874395, "grad_norm": 1331.3961181640625, "learning_rate": 4.999560163895335e-06, "loss": 1.5044, "step": 72 }, { "epoch": 0.06143488323164317, "grad_norm": 623.4003295898438, "learning_rate": 4.9995172787619285e-06, "loss": 1.3104, "step": 73 }, { "epoch": 0.06227645697454239, "grad_norm": 602.7695922851562, "learning_rate": 4.999472399233767e-06, "loss": 1.6953, "step": 74 }, { "epoch": 0.06311803071744161, "grad_norm": 963.92724609375, "learning_rate": 4.999425525346661e-06, "loss": 1.4376, "step": 75 }, { "epoch": 0.06395960446034084, "grad_norm": 625.63037109375, "learning_rate": 4.9993766571380124e-06, "loss": 1.6049, "step": 76 }, { "epoch": 0.06480117820324006, "grad_norm": 686.1634521484375, "learning_rate": 4.999325794646813e-06, "loss": 1.5269, "step": 77 }, { "epoch": 0.06564275194613928, "grad_norm": 644.1351928710938, "learning_rate": 4.999272937913647e-06, "loss": 1.3191, "step": 78 }, { "epoch": 0.0664843256890385, "grad_norm": 672.7321166992188, "learning_rate": 4.99921808698069e-06, "loss": 1.1279, "step": 79 }, { "epoch": 0.06732589943193773, "grad_norm": 993.0891723632812, "learning_rate": 4.999161241891707e-06, "loss": 1.6974, "step": 80 }, { "epoch": 0.06816747317483694, "grad_norm": 778.0184936523438, "learning_rate": 4.999102402692056e-06, "loss": 1.6307, "step": 81 }, { "epoch": 0.06900904691773617, "grad_norm": 601.3452758789062, "learning_rate": 4.9990415694286845e-06, "loss": 1.7501, "step": 82 }, { "epoch": 0.06985062066063538, "grad_norm": 640.2106323242188, "learning_rate": 4.998978742150133e-06, "loss": 1.3542, "step": 83 }, { "epoch": 0.07069219440353461, "grad_norm": 906.6204833984375, "learning_rate": 4.998913920906534e-06, "loss": 1.2639, "step": 84 }, { "epoch": 0.07153376814643383, "grad_norm": 454.73944091796875, "learning_rate": 4.998847105749606e-06, "loss": 1.1869, "step": 85 }, { "epoch": 0.07237534188933305, "grad_norm": 697.6354370117188, "learning_rate": 4.998778296732663e-06, "loss": 1.3469, "step": 86 }, { "epoch": 0.07321691563223227, "grad_norm": 518.2359619140625, "learning_rate": 4.998707493910607e-06, "loss": 1.4724, "step": 87 }, { "epoch": 0.0740584893751315, "grad_norm": 793.8117065429688, "learning_rate": 4.998634697339936e-06, "loss": 1.559, "step": 88 }, { "epoch": 0.07490006311803071, "grad_norm": 452.31536865234375, "learning_rate": 4.998559907078732e-06, "loss": 1.1228, "step": 89 }, { "epoch": 0.07574163686092994, "grad_norm": 671.4263305664062, "learning_rate": 4.998483123186673e-06, "loss": 1.3658, "step": 90 }, { "epoch": 0.07658321060382917, "grad_norm": 494.4920959472656, "learning_rate": 4.998404345725023e-06, "loss": 1.4983, "step": 91 }, { "epoch": 0.07742478434672838, "grad_norm": 564.9657592773438, "learning_rate": 4.998323574756643e-06, "loss": 1.4565, "step": 92 }, { "epoch": 0.0782663580896276, "grad_norm": 588.97216796875, "learning_rate": 4.998240810345979e-06, "loss": 1.4131, "step": 93 }, { "epoch": 0.07910793183252683, "grad_norm": 912.6625366210938, "learning_rate": 4.998156052559071e-06, "loss": 1.3753, "step": 94 }, { "epoch": 0.07994950557542604, "grad_norm": 623.6527709960938, "learning_rate": 4.998069301463546e-06, "loss": 1.2489, "step": 95 }, { "epoch": 0.08079107931832527, "grad_norm": 479.7945251464844, "learning_rate": 4.997980557128627e-06, "loss": 1.127, "step": 96 }, { "epoch": 0.08163265306122448, "grad_norm": 445.030517578125, "learning_rate": 4.997889819625121e-06, "loss": 1.2625, "step": 97 }, { "epoch": 0.08247422680412371, "grad_norm": 576.0341796875, "learning_rate": 4.99779708902543e-06, "loss": 1.5363, "step": 98 }, { "epoch": 0.08331580054702294, "grad_norm": 646.4935302734375, "learning_rate": 4.997702365403546e-06, "loss": 1.9151, "step": 99 }, { "epoch": 0.08415737428992215, "grad_norm": 688.9935302734375, "learning_rate": 4.997605648835048e-06, "loss": 1.3353, "step": 100 }, { "epoch": 0.08499894803282138, "grad_norm": 547.1505737304688, "learning_rate": 4.9975069393971076e-06, "loss": 1.2941, "step": 101 }, { "epoch": 0.0858405217757206, "grad_norm": 585.2498779296875, "learning_rate": 4.997406237168488e-06, "loss": 1.4254, "step": 102 }, { "epoch": 0.08668209551861981, "grad_norm": 503.3772277832031, "learning_rate": 4.997303542229539e-06, "loss": 1.2362, "step": 103 }, { "epoch": 0.08752366926151904, "grad_norm": 455.1918640136719, "learning_rate": 4.997198854662203e-06, "loss": 1.182, "step": 104 }, { "epoch": 0.08836524300441827, "grad_norm": 571.8682861328125, "learning_rate": 4.9970921745500104e-06, "loss": 1.5275, "step": 105 }, { "epoch": 0.08920681674731748, "grad_norm": 494.3360900878906, "learning_rate": 4.996983501978084e-06, "loss": 1.3781, "step": 106 }, { "epoch": 0.0900483904902167, "grad_norm": 550.50732421875, "learning_rate": 4.996872837033134e-06, "loss": 1.4291, "step": 107 }, { "epoch": 0.09088996423311593, "grad_norm": 601.7146606445312, "learning_rate": 4.996760179803461e-06, "loss": 1.4825, "step": 108 }, { "epoch": 0.09173153797601515, "grad_norm": 441.38037109375, "learning_rate": 4.996645530378957e-06, "loss": 1.4476, "step": 109 }, { "epoch": 0.09257311171891437, "grad_norm": 413.4273986816406, "learning_rate": 4.996528888851101e-06, "loss": 1.2163, "step": 110 }, { "epoch": 0.09341468546181358, "grad_norm": 580.2811279296875, "learning_rate": 4.996410255312962e-06, "loss": 1.1997, "step": 111 }, { "epoch": 0.09425625920471281, "grad_norm": 347.80322265625, "learning_rate": 4.996289629859201e-06, "loss": 1.3247, "step": 112 }, { "epoch": 0.09509783294761204, "grad_norm": 521.5128173828125, "learning_rate": 4.996167012586065e-06, "loss": 1.3204, "step": 113 }, { "epoch": 0.09593940669051125, "grad_norm": 384.0645446777344, "learning_rate": 4.996042403591393e-06, "loss": 1.3969, "step": 114 }, { "epoch": 0.09678098043341048, "grad_norm": 413.4881591796875, "learning_rate": 4.99591580297461e-06, "loss": 1.4374, "step": 115 }, { "epoch": 0.0976225541763097, "grad_norm": 389.8695983886719, "learning_rate": 4.995787210836734e-06, "loss": 1.1474, "step": 116 }, { "epoch": 0.09846412791920892, "grad_norm": 382.87823486328125, "learning_rate": 4.995656627280371e-06, "loss": 1.3191, "step": 117 }, { "epoch": 0.09930570166210814, "grad_norm": 471.7465515136719, "learning_rate": 4.995524052409713e-06, "loss": 1.5302, "step": 118 }, { "epoch": 0.10014727540500737, "grad_norm": 347.9407958984375, "learning_rate": 4.995389486330543e-06, "loss": 1.1651, "step": 119 }, { "epoch": 0.10098884914790658, "grad_norm": 490.6868896484375, "learning_rate": 4.995252929150234e-06, "loss": 1.2531, "step": 120 }, { "epoch": 0.10183042289080581, "grad_norm": 452.0323486328125, "learning_rate": 4.995114380977747e-06, "loss": 1.3209, "step": 121 }, { "epoch": 0.10267199663370503, "grad_norm": 406.12701416015625, "learning_rate": 4.99497384192363e-06, "loss": 1.5792, "step": 122 }, { "epoch": 0.10351357037660425, "grad_norm": 394.3304138183594, "learning_rate": 4.994831312100023e-06, "loss": 1.2446, "step": 123 }, { "epoch": 0.10435514411950347, "grad_norm": 377.8374938964844, "learning_rate": 4.994686791620649e-06, "loss": 0.9982, "step": 124 }, { "epoch": 0.1051967178624027, "grad_norm": 365.93560791015625, "learning_rate": 4.994540280600825e-06, "loss": 1.1529, "step": 125 }, { "epoch": 0.10603829160530191, "grad_norm": 368.8994140625, "learning_rate": 4.994391779157452e-06, "loss": 1.2085, "step": 126 }, { "epoch": 0.10687986534820114, "grad_norm": 381.2408752441406, "learning_rate": 4.994241287409024e-06, "loss": 1.2553, "step": 127 }, { "epoch": 0.10772143909110035, "grad_norm": 387.8236083984375, "learning_rate": 4.994088805475618e-06, "loss": 1.4945, "step": 128 }, { "epoch": 0.10856301283399958, "grad_norm": 305.5185546875, "learning_rate": 4.993934333478901e-06, "loss": 1.0922, "step": 129 }, { "epoch": 0.1094045865768988, "grad_norm": 408.90057373046875, "learning_rate": 4.993777871542129e-06, "loss": 1.2649, "step": 130 }, { "epoch": 0.11024616031979802, "grad_norm": 420.7751159667969, "learning_rate": 4.993619419790145e-06, "loss": 1.1042, "step": 131 }, { "epoch": 0.11108773406269724, "grad_norm": 331.85504150390625, "learning_rate": 4.993458978349379e-06, "loss": 1.1527, "step": 132 }, { "epoch": 0.11192930780559647, "grad_norm": 558.6054077148438, "learning_rate": 4.993296547347848e-06, "loss": 1.4046, "step": 133 }, { "epoch": 0.11277088154849568, "grad_norm": 365.9854736328125, "learning_rate": 4.99313212691516e-06, "loss": 1.3445, "step": 134 }, { "epoch": 0.11361245529139491, "grad_norm": 312.7580261230469, "learning_rate": 4.992965717182507e-06, "loss": 1.3768, "step": 135 }, { "epoch": 0.11445402903429414, "grad_norm": 460.6139831542969, "learning_rate": 4.992797318282668e-06, "loss": 1.4361, "step": 136 }, { "epoch": 0.11529560277719335, "grad_norm": 378.0125732421875, "learning_rate": 4.992626930350013e-06, "loss": 1.0356, "step": 137 }, { "epoch": 0.11613717652009257, "grad_norm": 435.4710388183594, "learning_rate": 4.992454553520495e-06, "loss": 1.3194, "step": 138 }, { "epoch": 0.1169787502629918, "grad_norm": 347.2532043457031, "learning_rate": 4.992280187931656e-06, "loss": 1.5423, "step": 139 }, { "epoch": 0.11782032400589101, "grad_norm": 314.25067138671875, "learning_rate": 4.992103833722623e-06, "loss": 0.9756, "step": 140 }, { "epoch": 0.11866189774879024, "grad_norm": 304.3933410644531, "learning_rate": 4.991925491034113e-06, "loss": 1.3259, "step": 141 }, { "epoch": 0.11950347149168945, "grad_norm": 301.6213073730469, "learning_rate": 4.9917451600084275e-06, "loss": 1.1353, "step": 142 }, { "epoch": 0.12034504523458868, "grad_norm": 408.5636901855469, "learning_rate": 4.991562840789455e-06, "loss": 1.402, "step": 143 }, { "epoch": 0.1211866189774879, "grad_norm": 340.44317626953125, "learning_rate": 4.991378533522669e-06, "loss": 1.353, "step": 144 }, { "epoch": 0.12202819272038712, "grad_norm": 268.15936279296875, "learning_rate": 4.99119223835513e-06, "loss": 1.2545, "step": 145 }, { "epoch": 0.12286976646328635, "grad_norm": 372.2171325683594, "learning_rate": 4.991003955435487e-06, "loss": 1.2183, "step": 146 }, { "epoch": 0.12371134020618557, "grad_norm": 278.7896423339844, "learning_rate": 4.990813684913972e-06, "loss": 1.1768, "step": 147 }, { "epoch": 0.12455291394908478, "grad_norm": 266.3719787597656, "learning_rate": 4.990621426942404e-06, "loss": 1.2064, "step": 148 }, { "epoch": 0.125394487691984, "grad_norm": 308.0953369140625, "learning_rate": 4.99042718167419e-06, "loss": 1.3979, "step": 149 }, { "epoch": 0.12623606143488322, "grad_norm": 414.5055236816406, "learning_rate": 4.990230949264319e-06, "loss": 1.1933, "step": 150 }, { "epoch": 0.12707763517778245, "grad_norm": 328.179931640625, "learning_rate": 4.990032729869366e-06, "loss": 1.157, "step": 151 }, { "epoch": 0.12791920892068168, "grad_norm": 468.81768798828125, "learning_rate": 4.989832523647494e-06, "loss": 1.15, "step": 152 }, { "epoch": 0.1287607826635809, "grad_norm": 352.3136291503906, "learning_rate": 4.989630330758451e-06, "loss": 1.1188, "step": 153 }, { "epoch": 0.12960235640648013, "grad_norm": 388.4230041503906, "learning_rate": 4.9894261513635674e-06, "loss": 0.9838, "step": 154 }, { "epoch": 0.13044393014937933, "grad_norm": 251.35621643066406, "learning_rate": 4.98921998562576e-06, "loss": 1.0673, "step": 155 }, { "epoch": 0.13128550389227855, "grad_norm": 438.343505859375, "learning_rate": 4.989011833709534e-06, "loss": 1.2557, "step": 156 }, { "epoch": 0.13212707763517778, "grad_norm": 316.020751953125, "learning_rate": 4.988801695780972e-06, "loss": 1.0859, "step": 157 }, { "epoch": 0.132968651378077, "grad_norm": 342.6540222167969, "learning_rate": 4.988589572007749e-06, "loss": 1.122, "step": 158 }, { "epoch": 0.13381022512097623, "grad_norm": 396.69403076171875, "learning_rate": 4.988375462559119e-06, "loss": 1.2415, "step": 159 }, { "epoch": 0.13465179886387546, "grad_norm": 480.0825500488281, "learning_rate": 4.988159367605924e-06, "loss": 1.5157, "step": 160 }, { "epoch": 0.13549337260677466, "grad_norm": 414.4588928222656, "learning_rate": 4.987941287320588e-06, "loss": 1.0775, "step": 161 }, { "epoch": 0.13633494634967389, "grad_norm": 345.3420715332031, "learning_rate": 4.987721221877119e-06, "loss": 1.3017, "step": 162 }, { "epoch": 0.1371765200925731, "grad_norm": 395.992431640625, "learning_rate": 4.987499171451112e-06, "loss": 1.176, "step": 163 }, { "epoch": 0.13801809383547234, "grad_norm": 241.21575927734375, "learning_rate": 4.987275136219742e-06, "loss": 1.1583, "step": 164 }, { "epoch": 0.13885966757837157, "grad_norm": 459.3298034667969, "learning_rate": 4.987049116361769e-06, "loss": 1.2788, "step": 165 }, { "epoch": 0.13970124132127076, "grad_norm": 408.43121337890625, "learning_rate": 4.9868211120575396e-06, "loss": 1.4012, "step": 166 }, { "epoch": 0.14054281506417, "grad_norm": 457.90081787109375, "learning_rate": 4.986591123488978e-06, "loss": 1.445, "step": 167 }, { "epoch": 0.14138438880706922, "grad_norm": 281.84515380859375, "learning_rate": 4.9863591508395965e-06, "loss": 1.3696, "step": 168 }, { "epoch": 0.14222596254996844, "grad_norm": 332.0083923339844, "learning_rate": 4.986125194294489e-06, "loss": 1.0008, "step": 169 }, { "epoch": 0.14306753629286767, "grad_norm": 339.5845031738281, "learning_rate": 4.985889254040332e-06, "loss": 1.3454, "step": 170 }, { "epoch": 0.1439091100357669, "grad_norm": 841.4637451171875, "learning_rate": 4.985651330265383e-06, "loss": 1.2247, "step": 171 }, { "epoch": 0.1447506837786661, "grad_norm": 336.28045654296875, "learning_rate": 4.985411423159487e-06, "loss": 1.3671, "step": 172 }, { "epoch": 0.14559225752156532, "grad_norm": 392.8457946777344, "learning_rate": 4.985169532914069e-06, "loss": 1.1262, "step": 173 }, { "epoch": 0.14643383126446455, "grad_norm": 301.209228515625, "learning_rate": 4.984925659722134e-06, "loss": 1.1485, "step": 174 }, { "epoch": 0.14727540500736377, "grad_norm": 277.6703796386719, "learning_rate": 4.984679803778273e-06, "loss": 1.1578, "step": 175 }, { "epoch": 0.148116978750263, "grad_norm": 302.1859436035156, "learning_rate": 4.984431965278655e-06, "loss": 1.2856, "step": 176 }, { "epoch": 0.1489585524931622, "grad_norm": 300.76025390625, "learning_rate": 4.984182144421037e-06, "loss": 1.0614, "step": 177 }, { "epoch": 0.14980012623606143, "grad_norm": 332.2203674316406, "learning_rate": 4.983930341404751e-06, "loss": 1.1719, "step": 178 }, { "epoch": 0.15064169997896065, "grad_norm": 204.2970733642578, "learning_rate": 4.983676556430715e-06, "loss": 0.9967, "step": 179 }, { "epoch": 0.15148327372185988, "grad_norm": 320.5247802734375, "learning_rate": 4.983420789701427e-06, "loss": 1.255, "step": 180 }, { "epoch": 0.1523248474647591, "grad_norm": 249.64166259765625, "learning_rate": 4.983163041420965e-06, "loss": 0.8233, "step": 181 }, { "epoch": 0.15316642120765833, "grad_norm": 246.04092407226562, "learning_rate": 4.982903311794991e-06, "loss": 1.3713, "step": 182 }, { "epoch": 0.15400799495055753, "grad_norm": 302.2898864746094, "learning_rate": 4.9826416010307464e-06, "loss": 1.159, "step": 183 }, { "epoch": 0.15484956869345676, "grad_norm": 252.8140411376953, "learning_rate": 4.982377909337053e-06, "loss": 1.0992, "step": 184 }, { "epoch": 0.15569114243635598, "grad_norm": 262.8348693847656, "learning_rate": 4.982112236924312e-06, "loss": 0.9386, "step": 185 }, { "epoch": 0.1565327161792552, "grad_norm": 262.5164489746094, "learning_rate": 4.98184458400451e-06, "loss": 1.2239, "step": 186 }, { "epoch": 0.15737428992215444, "grad_norm": 231.84715270996094, "learning_rate": 4.981574950791207e-06, "loss": 0.9828, "step": 187 }, { "epoch": 0.15821586366505366, "grad_norm": 240.302001953125, "learning_rate": 4.981303337499549e-06, "loss": 0.991, "step": 188 }, { "epoch": 0.15905743740795286, "grad_norm": 246.06973266601562, "learning_rate": 4.981029744346258e-06, "loss": 1.1315, "step": 189 }, { "epoch": 0.1598990111508521, "grad_norm": 302.11181640625, "learning_rate": 4.980754171549638e-06, "loss": 1.1767, "step": 190 }, { "epoch": 0.16074058489375131, "grad_norm": 253.83128356933594, "learning_rate": 4.980476619329571e-06, "loss": 1.1659, "step": 191 }, { "epoch": 0.16158215863665054, "grad_norm": 261.5288391113281, "learning_rate": 4.980197087907521e-06, "loss": 1.296, "step": 192 }, { "epoch": 0.16242373237954977, "grad_norm": 240.8744659423828, "learning_rate": 4.979915577506527e-06, "loss": 1.2207, "step": 193 }, { "epoch": 0.16326530612244897, "grad_norm": 247.27427673339844, "learning_rate": 4.979632088351211e-06, "loss": 1.1433, "step": 194 }, { "epoch": 0.1641068798653482, "grad_norm": 225.68511962890625, "learning_rate": 4.979346620667774e-06, "loss": 1.1316, "step": 195 }, { "epoch": 0.16494845360824742, "grad_norm": 235.21524047851562, "learning_rate": 4.979059174683991e-06, "loss": 1.0509, "step": 196 }, { "epoch": 0.16579002735114665, "grad_norm": 211.19992065429688, "learning_rate": 4.9787697506292195e-06, "loss": 1.1184, "step": 197 }, { "epoch": 0.16663160109404587, "grad_norm": 299.4044494628906, "learning_rate": 4.978478348734396e-06, "loss": 1.2405, "step": 198 }, { "epoch": 0.1674731748369451, "grad_norm": 226.4676055908203, "learning_rate": 4.978184969232032e-06, "loss": 1.2609, "step": 199 }, { "epoch": 0.1683147485798443, "grad_norm": 302.3034973144531, "learning_rate": 4.977889612356218e-06, "loss": 1.0461, "step": 200 }, { "epoch": 0.16915632232274352, "grad_norm": 336.5556335449219, "learning_rate": 4.9775922783426246e-06, "loss": 1.1755, "step": 201 }, { "epoch": 0.16999789606564275, "grad_norm": 159.33302307128906, "learning_rate": 4.977292967428497e-06, "loss": 0.8049, "step": 202 }, { "epoch": 0.17083946980854198, "grad_norm": 273.533935546875, "learning_rate": 4.97699167985266e-06, "loss": 0.9883, "step": 203 }, { "epoch": 0.1716810435514412, "grad_norm": 333.1833801269531, "learning_rate": 4.976688415855513e-06, "loss": 1.172, "step": 204 }, { "epoch": 0.17252261729434043, "grad_norm": 317.8067321777344, "learning_rate": 4.976383175679034e-06, "loss": 1.1309, "step": 205 }, { "epoch": 0.17336419103723963, "grad_norm": 275.27545166015625, "learning_rate": 4.97607595956678e-06, "loss": 1.0488, "step": 206 }, { "epoch": 0.17420576478013886, "grad_norm": 225.5316162109375, "learning_rate": 4.97576676776388e-06, "loss": 1.0911, "step": 207 }, { "epoch": 0.17504733852303808, "grad_norm": 236.5315399169922, "learning_rate": 4.975455600517043e-06, "loss": 0.9873, "step": 208 }, { "epoch": 0.1758889122659373, "grad_norm": 200.12295532226562, "learning_rate": 4.975142458074553e-06, "loss": 1.3244, "step": 209 }, { "epoch": 0.17673048600883653, "grad_norm": 836.041259765625, "learning_rate": 4.97482734068627e-06, "loss": 1.3085, "step": 210 }, { "epoch": 0.17757205975173573, "grad_norm": 208.06675720214844, "learning_rate": 4.97451024860363e-06, "loss": 1.2389, "step": 211 }, { "epoch": 0.17841363349463496, "grad_norm": 250.1989288330078, "learning_rate": 4.974191182079644e-06, "loss": 1.0525, "step": 212 }, { "epoch": 0.1792552072375342, "grad_norm": 273.5354919433594, "learning_rate": 4.9738701413689006e-06, "loss": 1.2381, "step": 213 }, { "epoch": 0.1800967809804334, "grad_norm": 646.2859497070312, "learning_rate": 4.97354712672756e-06, "loss": 0.9527, "step": 214 }, { "epoch": 0.18093835472333264, "grad_norm": 351.3960266113281, "learning_rate": 4.97322213841336e-06, "loss": 1.1817, "step": 215 }, { "epoch": 0.18177992846623187, "grad_norm": 270.67529296875, "learning_rate": 4.972895176685614e-06, "loss": 1.2518, "step": 216 }, { "epoch": 0.18262150220913106, "grad_norm": 191.62271118164062, "learning_rate": 4.972566241805208e-06, "loss": 0.991, "step": 217 }, { "epoch": 0.1834630759520303, "grad_norm": 233.8841094970703, "learning_rate": 4.972235334034601e-06, "loss": 1.171, "step": 218 }, { "epoch": 0.18430464969492952, "grad_norm": 260.0747375488281, "learning_rate": 4.971902453637831e-06, "loss": 1.0585, "step": 219 }, { "epoch": 0.18514622343782874, "grad_norm": 339.6416015625, "learning_rate": 4.971567600880507e-06, "loss": 0.9673, "step": 220 }, { "epoch": 0.18598779718072797, "grad_norm": 227.68484497070312, "learning_rate": 4.97123077602981e-06, "loss": 1.1898, "step": 221 }, { "epoch": 0.18682937092362717, "grad_norm": 221.54664611816406, "learning_rate": 4.970891979354499e-06, "loss": 1.0242, "step": 222 }, { "epoch": 0.1876709446665264, "grad_norm": 354.9067077636719, "learning_rate": 4.970551211124902e-06, "loss": 1.3887, "step": 223 }, { "epoch": 0.18851251840942562, "grad_norm": 351.784912109375, "learning_rate": 4.970208471612924e-06, "loss": 1.2994, "step": 224 }, { "epoch": 0.18935409215232485, "grad_norm": 449.1482849121094, "learning_rate": 4.969863761092038e-06, "loss": 1.196, "step": 225 }, { "epoch": 0.19019566589522408, "grad_norm": 306.9056396484375, "learning_rate": 4.9695170798372966e-06, "loss": 0.8384, "step": 226 }, { "epoch": 0.1910372396381233, "grad_norm": 276.563720703125, "learning_rate": 4.969168428125318e-06, "loss": 1.1098, "step": 227 }, { "epoch": 0.1918788133810225, "grad_norm": 269.2872009277344, "learning_rate": 4.968817806234296e-06, "loss": 1.1772, "step": 228 }, { "epoch": 0.19272038712392173, "grad_norm": 231.6291046142578, "learning_rate": 4.968465214443996e-06, "loss": 0.9755, "step": 229 }, { "epoch": 0.19356196086682095, "grad_norm": 256.9276123046875, "learning_rate": 4.968110653035757e-06, "loss": 1.1614, "step": 230 }, { "epoch": 0.19440353460972018, "grad_norm": 243.2183074951172, "learning_rate": 4.967754122292485e-06, "loss": 1.1839, "step": 231 }, { "epoch": 0.1952451083526194, "grad_norm": 250.42715454101562, "learning_rate": 4.967395622498662e-06, "loss": 1.1289, "step": 232 }, { "epoch": 0.19608668209551863, "grad_norm": 220.76181030273438, "learning_rate": 4.9670351539403395e-06, "loss": 1.1068, "step": 233 }, { "epoch": 0.19692825583841783, "grad_norm": 199.235107421875, "learning_rate": 4.9666727169051385e-06, "loss": 0.9976, "step": 234 }, { "epoch": 0.19776982958131706, "grad_norm": 225.22341918945312, "learning_rate": 4.9663083116822515e-06, "loss": 0.9086, "step": 235 }, { "epoch": 0.19861140332421628, "grad_norm": 203.07470703125, "learning_rate": 4.965941938562444e-06, "loss": 0.977, "step": 236 }, { "epoch": 0.1994529770671155, "grad_norm": 204.74114990234375, "learning_rate": 4.965573597838047e-06, "loss": 0.8848, "step": 237 }, { "epoch": 0.20029455081001474, "grad_norm": 196.8950653076172, "learning_rate": 4.965203289802966e-06, "loss": 1.0225, "step": 238 }, { "epoch": 0.20113612455291394, "grad_norm": 200.91964721679688, "learning_rate": 4.964831014752673e-06, "loss": 0.9533, "step": 239 }, { "epoch": 0.20197769829581316, "grad_norm": 214.7240753173828, "learning_rate": 4.964456772984211e-06, "loss": 1.1088, "step": 240 }, { "epoch": 0.2028192720387124, "grad_norm": 324.22711181640625, "learning_rate": 4.964080564796192e-06, "loss": 1.0528, "step": 241 }, { "epoch": 0.20366084578161162, "grad_norm": 363.13189697265625, "learning_rate": 4.963702390488797e-06, "loss": 1.1069, "step": 242 }, { "epoch": 0.20450241952451084, "grad_norm": 176.63153076171875, "learning_rate": 4.963322250363776e-06, "loss": 0.9716, "step": 243 }, { "epoch": 0.20534399326741007, "grad_norm": 209.2146759033203, "learning_rate": 4.962940144724446e-06, "loss": 1.2695, "step": 244 }, { "epoch": 0.20618556701030927, "grad_norm": 230.91806030273438, "learning_rate": 4.9625560738756965e-06, "loss": 1.1252, "step": 245 }, { "epoch": 0.2070271407532085, "grad_norm": 219.78872680664062, "learning_rate": 4.96217003812398e-06, "loss": 0.9209, "step": 246 }, { "epoch": 0.20786871449610772, "grad_norm": 331.7774353027344, "learning_rate": 4.961782037777319e-06, "loss": 1.0271, "step": 247 }, { "epoch": 0.20871028823900695, "grad_norm": 238.35067749023438, "learning_rate": 4.961392073145305e-06, "loss": 1.0553, "step": 248 }, { "epoch": 0.20955186198190617, "grad_norm": 219.87989807128906, "learning_rate": 4.961000144539094e-06, "loss": 1.1054, "step": 249 }, { "epoch": 0.2103934357248054, "grad_norm": 213.93609619140625, "learning_rate": 4.960606252271413e-06, "loss": 1.0722, "step": 250 }, { "epoch": 0.2112350094677046, "grad_norm": 173.34837341308594, "learning_rate": 4.960210396656549e-06, "loss": 0.7469, "step": 251 }, { "epoch": 0.21207658321060383, "grad_norm": 176.90809631347656, "learning_rate": 4.959812578010364e-06, "loss": 0.8142, "step": 252 }, { "epoch": 0.21291815695350305, "grad_norm": 145.44351196289062, "learning_rate": 4.959412796650281e-06, "loss": 0.6995, "step": 253 }, { "epoch": 0.21375973069640228, "grad_norm": 164.6792755126953, "learning_rate": 4.959011052895289e-06, "loss": 0.8751, "step": 254 }, { "epoch": 0.2146013044393015, "grad_norm": 179.0094451904297, "learning_rate": 4.9586073470659465e-06, "loss": 0.8738, "step": 255 }, { "epoch": 0.2154428781822007, "grad_norm": 275.41729736328125, "learning_rate": 4.958201679484373e-06, "loss": 1.1572, "step": 256 }, { "epoch": 0.21628445192509993, "grad_norm": 211.57325744628906, "learning_rate": 4.9577940504742576e-06, "loss": 1.1856, "step": 257 }, { "epoch": 0.21712602566799916, "grad_norm": 164.41294860839844, "learning_rate": 4.95738446036085e-06, "loss": 0.7646, "step": 258 }, { "epoch": 0.21796759941089838, "grad_norm": 204.46368408203125, "learning_rate": 4.95697290947097e-06, "loss": 0.9968, "step": 259 }, { "epoch": 0.2188091731537976, "grad_norm": 166.3022003173828, "learning_rate": 4.956559398132995e-06, "loss": 0.9949, "step": 260 }, { "epoch": 0.21965074689669684, "grad_norm": 183.39434814453125, "learning_rate": 4.956143926676874e-06, "loss": 0.8959, "step": 261 }, { "epoch": 0.22049232063959603, "grad_norm": 184.88546752929688, "learning_rate": 4.955726495434116e-06, "loss": 0.9911, "step": 262 }, { "epoch": 0.22133389438249526, "grad_norm": 261.6977844238281, "learning_rate": 4.955307104737794e-06, "loss": 1.0551, "step": 263 }, { "epoch": 0.2221754681253945, "grad_norm": 206.3694610595703, "learning_rate": 4.954885754922544e-06, "loss": 1.0895, "step": 264 }, { "epoch": 0.22301704186829371, "grad_norm": 192.2148895263672, "learning_rate": 4.954462446324567e-06, "loss": 0.9871, "step": 265 }, { "epoch": 0.22385861561119294, "grad_norm": 181.06787109375, "learning_rate": 4.954037179281626e-06, "loss": 1.0505, "step": 266 }, { "epoch": 0.22470018935409214, "grad_norm": 211.8572540283203, "learning_rate": 4.9536099541330475e-06, "loss": 0.9762, "step": 267 }, { "epoch": 0.22554176309699137, "grad_norm": 229.56036376953125, "learning_rate": 4.953180771219718e-06, "loss": 0.9214, "step": 268 }, { "epoch": 0.2263833368398906, "grad_norm": 175.8040771484375, "learning_rate": 4.952749630884088e-06, "loss": 0.8592, "step": 269 }, { "epoch": 0.22722491058278982, "grad_norm": 205.4430389404297, "learning_rate": 4.952316533470169e-06, "loss": 1.093, "step": 270 }, { "epoch": 0.22806648432568905, "grad_norm": 208.4731903076172, "learning_rate": 4.951881479323537e-06, "loss": 1.0825, "step": 271 }, { "epoch": 0.22890805806858827, "grad_norm": 257.52972412109375, "learning_rate": 4.951444468791325e-06, "loss": 1.0441, "step": 272 }, { "epoch": 0.22974963181148747, "grad_norm": 155.39646911621094, "learning_rate": 4.951005502222229e-06, "loss": 0.7643, "step": 273 }, { "epoch": 0.2305912055543867, "grad_norm": 187.8433380126953, "learning_rate": 4.950564579966506e-06, "loss": 0.8654, "step": 274 }, { "epoch": 0.23143277929728592, "grad_norm": 185.39447021484375, "learning_rate": 4.950121702375973e-06, "loss": 1.096, "step": 275 }, { "epoch": 0.23227435304018515, "grad_norm": 164.85166931152344, "learning_rate": 4.949676869804008e-06, "loss": 0.9997, "step": 276 }, { "epoch": 0.23311592678308438, "grad_norm": 231.10238647460938, "learning_rate": 4.949230082605546e-06, "loss": 0.8651, "step": 277 }, { "epoch": 0.2339575005259836, "grad_norm": 220.4220428466797, "learning_rate": 4.948781341137087e-06, "loss": 1.245, "step": 278 }, { "epoch": 0.2347990742688828, "grad_norm": 193.20501708984375, "learning_rate": 4.948330645756686e-06, "loss": 1.1347, "step": 279 }, { "epoch": 0.23564064801178203, "grad_norm": 138.90548706054688, "learning_rate": 4.947877996823957e-06, "loss": 0.9499, "step": 280 }, { "epoch": 0.23648222175468125, "grad_norm": 217.0552520751953, "learning_rate": 4.947423394700075e-06, "loss": 1.3172, "step": 281 }, { "epoch": 0.23732379549758048, "grad_norm": 214.4515380859375, "learning_rate": 4.946966839747773e-06, "loss": 1.0737, "step": 282 }, { "epoch": 0.2381653692404797, "grad_norm": 222.87826538085938, "learning_rate": 4.946508332331341e-06, "loss": 1.0215, "step": 283 }, { "epoch": 0.2390069429833789, "grad_norm": 242.82290649414062, "learning_rate": 4.946047872816628e-06, "loss": 1.0079, "step": 284 }, { "epoch": 0.23984851672627813, "grad_norm": 190.513427734375, "learning_rate": 4.9455854615710385e-06, "loss": 0.9054, "step": 285 }, { "epoch": 0.24069009046917736, "grad_norm": 180.5825958251953, "learning_rate": 4.945121098963538e-06, "loss": 0.8978, "step": 286 }, { "epoch": 0.24153166421207659, "grad_norm": 222.07452392578125, "learning_rate": 4.944654785364647e-06, "loss": 0.7585, "step": 287 }, { "epoch": 0.2423732379549758, "grad_norm": 323.4588928222656, "learning_rate": 4.94418652114644e-06, "loss": 1.0618, "step": 288 }, { "epoch": 0.24321481169787504, "grad_norm": 178.20001220703125, "learning_rate": 4.943716306682555e-06, "loss": 0.9913, "step": 289 }, { "epoch": 0.24405638544077424, "grad_norm": 211.6590576171875, "learning_rate": 4.943244142348177e-06, "loss": 0.872, "step": 290 }, { "epoch": 0.24489795918367346, "grad_norm": 155.99195861816406, "learning_rate": 4.942770028520055e-06, "loss": 0.82, "step": 291 }, { "epoch": 0.2457395329265727, "grad_norm": 186.73419189453125, "learning_rate": 4.942293965576489e-06, "loss": 1.0884, "step": 292 }, { "epoch": 0.24658110666947192, "grad_norm": 246.95408630371094, "learning_rate": 4.941815953897335e-06, "loss": 1.0209, "step": 293 }, { "epoch": 0.24742268041237114, "grad_norm": 255.74241638183594, "learning_rate": 4.941335993864005e-06, "loss": 1.0589, "step": 294 }, { "epoch": 0.24826425415527034, "grad_norm": 240.6260528564453, "learning_rate": 4.940854085859465e-06, "loss": 1.2218, "step": 295 }, { "epoch": 0.24910582789816957, "grad_norm": 248.5775146484375, "learning_rate": 4.940370230268233e-06, "loss": 0.9993, "step": 296 }, { "epoch": 0.2499474016410688, "grad_norm": 217.07452392578125, "learning_rate": 4.939884427476386e-06, "loss": 1.1372, "step": 297 }, { "epoch": 0.250788975383968, "grad_norm": 196.2444610595703, "learning_rate": 4.9393966778715505e-06, "loss": 0.7739, "step": 298 }, { "epoch": 0.2516305491268672, "grad_norm": 138.59149169921875, "learning_rate": 4.938906981842907e-06, "loss": 0.8014, "step": 299 }, { "epoch": 0.25247212286976645, "grad_norm": 152.51995849609375, "learning_rate": 4.938415339781192e-06, "loss": 1.0103, "step": 300 }, { "epoch": 0.2533136966126657, "grad_norm": 136.28709411621094, "learning_rate": 4.93792175207869e-06, "loss": 0.7962, "step": 301 }, { "epoch": 0.2541552703555649, "grad_norm": 171.7638397216797, "learning_rate": 4.937426219129242e-06, "loss": 1.0729, "step": 302 }, { "epoch": 0.2549968440984641, "grad_norm": 307.9116516113281, "learning_rate": 4.93692874132824e-06, "loss": 1.0687, "step": 303 }, { "epoch": 0.25583841784136335, "grad_norm": 135.71917724609375, "learning_rate": 4.936429319072626e-06, "loss": 0.7077, "step": 304 }, { "epoch": 0.2566799915842626, "grad_norm": 220.42298889160156, "learning_rate": 4.9359279527608975e-06, "loss": 1.295, "step": 305 }, { "epoch": 0.2575215653271618, "grad_norm": 143.9759063720703, "learning_rate": 4.935424642793099e-06, "loss": 0.9554, "step": 306 }, { "epoch": 0.25836313907006103, "grad_norm": 155.3295440673828, "learning_rate": 4.934919389570826e-06, "loss": 0.8669, "step": 307 }, { "epoch": 0.25920471281296026, "grad_norm": 173.58909606933594, "learning_rate": 4.934412193497229e-06, "loss": 0.9683, "step": 308 }, { "epoch": 0.2600462865558595, "grad_norm": 164.60137939453125, "learning_rate": 4.933903054977005e-06, "loss": 1.0304, "step": 309 }, { "epoch": 0.26088786029875866, "grad_norm": 148.02781677246094, "learning_rate": 4.9333919744164005e-06, "loss": 0.9517, "step": 310 }, { "epoch": 0.2617294340416579, "grad_norm": 167.60630798339844, "learning_rate": 4.932878952223213e-06, "loss": 0.9157, "step": 311 }, { "epoch": 0.2625710077845571, "grad_norm": 165.19789123535156, "learning_rate": 4.932363988806791e-06, "loss": 0.7926, "step": 312 }, { "epoch": 0.26341258152745634, "grad_norm": 179.5117645263672, "learning_rate": 4.931847084578028e-06, "loss": 0.9573, "step": 313 }, { "epoch": 0.26425415527035556, "grad_norm": 199.1255645751953, "learning_rate": 4.931328239949369e-06, "loss": 1.1345, "step": 314 }, { "epoch": 0.2650957290132548, "grad_norm": 192.1922607421875, "learning_rate": 4.930807455334807e-06, "loss": 1.0578, "step": 315 }, { "epoch": 0.265937302756154, "grad_norm": 252.30825805664062, "learning_rate": 4.930284731149879e-06, "loss": 1.1559, "step": 316 }, { "epoch": 0.26677887649905324, "grad_norm": 189.8568572998047, "learning_rate": 4.929760067811676e-06, "loss": 1.211, "step": 317 }, { "epoch": 0.26762045024195247, "grad_norm": 184.5948486328125, "learning_rate": 4.929233465738833e-06, "loss": 0.9276, "step": 318 }, { "epoch": 0.2684620239848517, "grad_norm": 270.45855712890625, "learning_rate": 4.928704925351532e-06, "loss": 1.0734, "step": 319 }, { "epoch": 0.2693035977277509, "grad_norm": 224.6494598388672, "learning_rate": 4.9281744470715e-06, "loss": 0.9032, "step": 320 }, { "epoch": 0.2701451714706501, "grad_norm": 141.63299560546875, "learning_rate": 4.927642031322013e-06, "loss": 0.7141, "step": 321 }, { "epoch": 0.2709867452135493, "grad_norm": 178.1015625, "learning_rate": 4.927107678527893e-06, "loss": 0.7973, "step": 322 }, { "epoch": 0.27182831895644854, "grad_norm": 155.43727111816406, "learning_rate": 4.9265713891155055e-06, "loss": 0.9052, "step": 323 }, { "epoch": 0.27266989269934777, "grad_norm": 140.11785888671875, "learning_rate": 4.926033163512761e-06, "loss": 0.7327, "step": 324 }, { "epoch": 0.273511466442247, "grad_norm": 153.34718322753906, "learning_rate": 4.925493002149119e-06, "loss": 0.8394, "step": 325 }, { "epoch": 0.2743530401851462, "grad_norm": 191.73004150390625, "learning_rate": 4.924950905455578e-06, "loss": 0.9868, "step": 326 }, { "epoch": 0.27519461392804545, "grad_norm": 150.3389129638672, "learning_rate": 4.924406873864685e-06, "loss": 0.6314, "step": 327 }, { "epoch": 0.2760361876709447, "grad_norm": 200.07156372070312, "learning_rate": 4.9238609078105285e-06, "loss": 0.889, "step": 328 }, { "epoch": 0.2768777614138439, "grad_norm": 133.3046417236328, "learning_rate": 4.9233130077287425e-06, "loss": 0.7299, "step": 329 }, { "epoch": 0.27771933515674313, "grad_norm": 159.61390686035156, "learning_rate": 4.9227631740565015e-06, "loss": 0.9479, "step": 330 }, { "epoch": 0.27856090889964236, "grad_norm": 148.41209411621094, "learning_rate": 4.922211407232525e-06, "loss": 1.0882, "step": 331 }, { "epoch": 0.2794024826425415, "grad_norm": 167.6956329345703, "learning_rate": 4.921657707697077e-06, "loss": 1.0162, "step": 332 }, { "epoch": 0.28024405638544075, "grad_norm": 200.65501403808594, "learning_rate": 4.9211020758919555e-06, "loss": 1.0155, "step": 333 }, { "epoch": 0.28108563012834, "grad_norm": 168.57797241210938, "learning_rate": 4.9205445122605095e-06, "loss": 0.9225, "step": 334 }, { "epoch": 0.2819272038712392, "grad_norm": 148.2366943359375, "learning_rate": 4.919985017247625e-06, "loss": 0.8756, "step": 335 }, { "epoch": 0.28276877761413843, "grad_norm": 215.92623901367188, "learning_rate": 4.91942359129973e-06, "loss": 1.1371, "step": 336 }, { "epoch": 0.28361035135703766, "grad_norm": 183.8272247314453, "learning_rate": 4.918860234864793e-06, "loss": 0.9762, "step": 337 }, { "epoch": 0.2844519250999369, "grad_norm": 213.24313354492188, "learning_rate": 4.918294948392322e-06, "loss": 1.0352, "step": 338 }, { "epoch": 0.2852934988428361, "grad_norm": 162.19821166992188, "learning_rate": 4.9177277323333665e-06, "loss": 1.1227, "step": 339 }, { "epoch": 0.28613507258573534, "grad_norm": 174.2341766357422, "learning_rate": 4.917158587140515e-06, "loss": 0.9655, "step": 340 }, { "epoch": 0.28697664632863457, "grad_norm": 147.44154357910156, "learning_rate": 4.916587513267895e-06, "loss": 0.8849, "step": 341 }, { "epoch": 0.2878182200715338, "grad_norm": 162.8009033203125, "learning_rate": 4.916014511171173e-06, "loss": 0.9787, "step": 342 }, { "epoch": 0.28865979381443296, "grad_norm": 192.03756713867188, "learning_rate": 4.915439581307555e-06, "loss": 0.8299, "step": 343 }, { "epoch": 0.2895013675573322, "grad_norm": 239.55003356933594, "learning_rate": 4.914862724135784e-06, "loss": 0.8902, "step": 344 }, { "epoch": 0.2903429413002314, "grad_norm": 143.41920471191406, "learning_rate": 4.91428394011614e-06, "loss": 0.8735, "step": 345 }, { "epoch": 0.29118451504313064, "grad_norm": 151.96649169921875, "learning_rate": 4.9137032297104445e-06, "loss": 1.0496, "step": 346 }, { "epoch": 0.29202608878602987, "grad_norm": 143.7827606201172, "learning_rate": 4.913120593382051e-06, "loss": 0.9842, "step": 347 }, { "epoch": 0.2928676625289291, "grad_norm": 214.41659545898438, "learning_rate": 4.912536031595853e-06, "loss": 1.0993, "step": 348 }, { "epoch": 0.2937092362718283, "grad_norm": 192.0199737548828, "learning_rate": 4.911949544818278e-06, "loss": 0.9345, "step": 349 }, { "epoch": 0.29455081001472755, "grad_norm": 140.22434997558594, "learning_rate": 4.911361133517293e-06, "loss": 0.7077, "step": 350 }, { "epoch": 0.2953923837576268, "grad_norm": 149.80943298339844, "learning_rate": 4.910770798162396e-06, "loss": 1.0209, "step": 351 }, { "epoch": 0.296233957500526, "grad_norm": 188.54193115234375, "learning_rate": 4.910178539224626e-06, "loss": 0.7282, "step": 352 }, { "epoch": 0.29707553124342523, "grad_norm": 182.12388610839844, "learning_rate": 4.90958435717655e-06, "loss": 1.1027, "step": 353 }, { "epoch": 0.2979171049863244, "grad_norm": 155.85667419433594, "learning_rate": 4.908988252492274e-06, "loss": 1.1697, "step": 354 }, { "epoch": 0.2987586787292236, "grad_norm": 174.2142791748047, "learning_rate": 4.908390225647438e-06, "loss": 1.1215, "step": 355 }, { "epoch": 0.29960025247212285, "grad_norm": 191.0850830078125, "learning_rate": 4.907790277119214e-06, "loss": 1.1317, "step": 356 }, { "epoch": 0.3004418262150221, "grad_norm": 166.85194396972656, "learning_rate": 4.907188407386308e-06, "loss": 0.9083, "step": 357 }, { "epoch": 0.3012833999579213, "grad_norm": 115.8208236694336, "learning_rate": 4.90658461692896e-06, "loss": 0.8165, "step": 358 }, { "epoch": 0.30212497370082053, "grad_norm": 136.56874084472656, "learning_rate": 4.905978906228942e-06, "loss": 0.7891, "step": 359 }, { "epoch": 0.30296654744371976, "grad_norm": 153.9842071533203, "learning_rate": 4.905371275769555e-06, "loss": 1.1678, "step": 360 }, { "epoch": 0.303808121186619, "grad_norm": 183.4450225830078, "learning_rate": 4.9047617260356375e-06, "loss": 0.9229, "step": 361 }, { "epoch": 0.3046496949295182, "grad_norm": 132.87869262695312, "learning_rate": 4.904150257513555e-06, "loss": 0.8403, "step": 362 }, { "epoch": 0.30549126867241744, "grad_norm": 143.56069946289062, "learning_rate": 4.903536870691206e-06, "loss": 0.9998, "step": 363 }, { "epoch": 0.30633284241531666, "grad_norm": 151.4114532470703, "learning_rate": 4.90292156605802e-06, "loss": 1.0212, "step": 364 }, { "epoch": 0.3071744161582159, "grad_norm": 139.41395568847656, "learning_rate": 4.902304344104954e-06, "loss": 1.0311, "step": 365 }, { "epoch": 0.30801598990111506, "grad_norm": 155.7967071533203, "learning_rate": 4.901685205324499e-06, "loss": 0.9561, "step": 366 }, { "epoch": 0.3088575636440143, "grad_norm": 153.0625457763672, "learning_rate": 4.901064150210672e-06, "loss": 0.984, "step": 367 }, { "epoch": 0.3096991373869135, "grad_norm": 204.0398406982422, "learning_rate": 4.900441179259019e-06, "loss": 1.152, "step": 368 }, { "epoch": 0.31054071112981274, "grad_norm": 178.45762634277344, "learning_rate": 4.899816292966619e-06, "loss": 0.9165, "step": 369 }, { "epoch": 0.31138228487271197, "grad_norm": 154.21519470214844, "learning_rate": 4.899189491832075e-06, "loss": 1.1589, "step": 370 }, { "epoch": 0.3122238586156112, "grad_norm": 152.87892150878906, "learning_rate": 4.8985607763555175e-06, "loss": 0.9606, "step": 371 }, { "epoch": 0.3130654323585104, "grad_norm": 125.60699462890625, "learning_rate": 4.897930147038609e-06, "loss": 0.9289, "step": 372 }, { "epoch": 0.31390700610140965, "grad_norm": 141.3811492919922, "learning_rate": 4.897297604384533e-06, "loss": 0.7742, "step": 373 }, { "epoch": 0.3147485798443089, "grad_norm": 123.92646789550781, "learning_rate": 4.8966631488980055e-06, "loss": 0.8978, "step": 374 }, { "epoch": 0.3155901535872081, "grad_norm": 147.79969787597656, "learning_rate": 4.896026781085265e-06, "loss": 1.042, "step": 375 }, { "epoch": 0.3164317273301073, "grad_norm": 274.53399658203125, "learning_rate": 4.895388501454077e-06, "loss": 1.0025, "step": 376 }, { "epoch": 0.3172733010730065, "grad_norm": 154.58265686035156, "learning_rate": 4.894748310513733e-06, "loss": 0.99, "step": 377 }, { "epoch": 0.3181148748159057, "grad_norm": 110.33751678466797, "learning_rate": 4.894106208775048e-06, "loss": 0.8535, "step": 378 }, { "epoch": 0.31895644855880495, "grad_norm": 118.87710571289062, "learning_rate": 4.893462196750362e-06, "loss": 0.9556, "step": 379 }, { "epoch": 0.3197980223017042, "grad_norm": 119.17797088623047, "learning_rate": 4.8928162749535435e-06, "loss": 0.8088, "step": 380 }, { "epoch": 0.3206395960446034, "grad_norm": 96.09217834472656, "learning_rate": 4.892168443899978e-06, "loss": 0.7595, "step": 381 }, { "epoch": 0.32148116978750263, "grad_norm": 127.31585693359375, "learning_rate": 4.8915187041065785e-06, "loss": 0.8242, "step": 382 }, { "epoch": 0.32232274353040186, "grad_norm": 125.22929382324219, "learning_rate": 4.89086705609178e-06, "loss": 0.8291, "step": 383 }, { "epoch": 0.3231643172733011, "grad_norm": 144.8851776123047, "learning_rate": 4.89021350037554e-06, "loss": 1.1254, "step": 384 }, { "epoch": 0.3240058910162003, "grad_norm": 121.80813598632812, "learning_rate": 4.88955803747934e-06, "loss": 0.7422, "step": 385 }, { "epoch": 0.32484746475909954, "grad_norm": 159.96646118164062, "learning_rate": 4.88890066792618e-06, "loss": 1.0759, "step": 386 }, { "epoch": 0.32568903850199876, "grad_norm": 114.78094482421875, "learning_rate": 4.888241392240584e-06, "loss": 0.563, "step": 387 }, { "epoch": 0.32653061224489793, "grad_norm": 140.37147521972656, "learning_rate": 4.887580210948595e-06, "loss": 0.7917, "step": 388 }, { "epoch": 0.32737218598779716, "grad_norm": 143.88070678710938, "learning_rate": 4.886917124577778e-06, "loss": 1.0749, "step": 389 }, { "epoch": 0.3282137597306964, "grad_norm": 142.34536743164062, "learning_rate": 4.886252133657218e-06, "loss": 0.9067, "step": 390 }, { "epoch": 0.3290553334735956, "grad_norm": 157.0981903076172, "learning_rate": 4.8855852387175185e-06, "loss": 1.1051, "step": 391 }, { "epoch": 0.32989690721649484, "grad_norm": 272.75946044921875, "learning_rate": 4.8849164402908024e-06, "loss": 0.8886, "step": 392 }, { "epoch": 0.33073848095939407, "grad_norm": 132.60775756835938, "learning_rate": 4.884245738910714e-06, "loss": 0.8409, "step": 393 }, { "epoch": 0.3315800547022933, "grad_norm": 127.93871307373047, "learning_rate": 4.883573135112413e-06, "loss": 0.7745, "step": 394 }, { "epoch": 0.3324216284451925, "grad_norm": 119.92588806152344, "learning_rate": 4.882898629432577e-06, "loss": 0.9234, "step": 395 }, { "epoch": 0.33326320218809175, "grad_norm": 149.8199005126953, "learning_rate": 4.882222222409403e-06, "loss": 1.017, "step": 396 }, { "epoch": 0.33410477593099097, "grad_norm": 141.49534606933594, "learning_rate": 4.881543914582604e-06, "loss": 0.7122, "step": 397 }, { "epoch": 0.3349463496738902, "grad_norm": 118.47855377197266, "learning_rate": 4.880863706493411e-06, "loss": 0.928, "step": 398 }, { "epoch": 0.33578792341678937, "grad_norm": 123.57112884521484, "learning_rate": 4.8801815986845685e-06, "loss": 0.8434, "step": 399 }, { "epoch": 0.3366294971596886, "grad_norm": 139.73097229003906, "learning_rate": 4.879497591700339e-06, "loss": 0.7714, "step": 400 }, { "epoch": 0.3374710709025878, "grad_norm": 138.10569763183594, "learning_rate": 4.878811686086502e-06, "loss": 0.8897, "step": 401 }, { "epoch": 0.33831264464548705, "grad_norm": 157.8883514404297, "learning_rate": 4.878123882390347e-06, "loss": 1.1479, "step": 402 }, { "epoch": 0.3391542183883863, "grad_norm": 138.03404235839844, "learning_rate": 4.877434181160681e-06, "loss": 0.9142, "step": 403 }, { "epoch": 0.3399957921312855, "grad_norm": 142.91648864746094, "learning_rate": 4.876742582947826e-06, "loss": 0.8765, "step": 404 }, { "epoch": 0.34083736587418473, "grad_norm": 140.9148406982422, "learning_rate": 4.876049088303616e-06, "loss": 1.0419, "step": 405 }, { "epoch": 0.34167893961708395, "grad_norm": 127.75152587890625, "learning_rate": 4.8753536977814e-06, "loss": 0.839, "step": 406 }, { "epoch": 0.3425205133599832, "grad_norm": 137.12606811523438, "learning_rate": 4.874656411936035e-06, "loss": 0.8453, "step": 407 }, { "epoch": 0.3433620871028824, "grad_norm": 158.32156372070312, "learning_rate": 4.8739572313238984e-06, "loss": 1.0524, "step": 408 }, { "epoch": 0.34420366084578163, "grad_norm": 142.4107208251953, "learning_rate": 4.873256156502871e-06, "loss": 0.9642, "step": 409 }, { "epoch": 0.34504523458868086, "grad_norm": 165.7876739501953, "learning_rate": 4.87255318803235e-06, "loss": 0.8204, "step": 410 }, { "epoch": 0.34588680833158003, "grad_norm": 134.32568359375, "learning_rate": 4.871848326473243e-06, "loss": 0.8822, "step": 411 }, { "epoch": 0.34672838207447926, "grad_norm": 148.0491180419922, "learning_rate": 4.871141572387966e-06, "loss": 0.8069, "step": 412 }, { "epoch": 0.3475699558173785, "grad_norm": 159.5659637451172, "learning_rate": 4.870432926340448e-06, "loss": 1.1664, "step": 413 }, { "epoch": 0.3484115295602777, "grad_norm": 170.69534301757812, "learning_rate": 4.869722388896127e-06, "loss": 1.2899, "step": 414 }, { "epoch": 0.34925310330317694, "grad_norm": 156.73345947265625, "learning_rate": 4.869009960621947e-06, "loss": 0.9722, "step": 415 }, { "epoch": 0.35009467704607616, "grad_norm": 113.5978775024414, "learning_rate": 4.868295642086363e-06, "loss": 0.9496, "step": 416 }, { "epoch": 0.3509362507889754, "grad_norm": 97.61703491210938, "learning_rate": 4.86757943385934e-06, "loss": 0.6348, "step": 417 }, { "epoch": 0.3517778245318746, "grad_norm": 170.4888916015625, "learning_rate": 4.8668613365123485e-06, "loss": 0.8473, "step": 418 }, { "epoch": 0.35261939827477384, "grad_norm": 122.68419647216797, "learning_rate": 4.866141350618367e-06, "loss": 0.9248, "step": 419 }, { "epoch": 0.35346097201767307, "grad_norm": 145.2256622314453, "learning_rate": 4.8654194767518815e-06, "loss": 0.9411, "step": 420 }, { "epoch": 0.3543025457605723, "grad_norm": 123.21392059326172, "learning_rate": 4.864695715488882e-06, "loss": 0.9264, "step": 421 }, { "epoch": 0.35514411950347147, "grad_norm": 127.80828857421875, "learning_rate": 4.863970067406868e-06, "loss": 0.9816, "step": 422 }, { "epoch": 0.3559856932463707, "grad_norm": 108.34115600585938, "learning_rate": 4.863242533084842e-06, "loss": 0.9395, "step": 423 }, { "epoch": 0.3568272669892699, "grad_norm": 106.89410400390625, "learning_rate": 4.862513113103312e-06, "loss": 0.8584, "step": 424 }, { "epoch": 0.35766884073216915, "grad_norm": 110.82831573486328, "learning_rate": 4.861781808044292e-06, "loss": 0.7783, "step": 425 }, { "epoch": 0.3585104144750684, "grad_norm": 170.09359741210938, "learning_rate": 4.861048618491298e-06, "loss": 0.9577, "step": 426 }, { "epoch": 0.3593519882179676, "grad_norm": 126.96100616455078, "learning_rate": 4.8603135450293516e-06, "loss": 0.9036, "step": 427 }, { "epoch": 0.3601935619608668, "grad_norm": 165.4973907470703, "learning_rate": 4.8595765882449766e-06, "loss": 1.0496, "step": 428 }, { "epoch": 0.36103513570376605, "grad_norm": 126.7000503540039, "learning_rate": 4.858837748726199e-06, "loss": 0.9302, "step": 429 }, { "epoch": 0.3618767094466653, "grad_norm": 154.56597900390625, "learning_rate": 4.858097027062548e-06, "loss": 0.9942, "step": 430 }, { "epoch": 0.3627182831895645, "grad_norm": 106.47966766357422, "learning_rate": 4.857354423845054e-06, "loss": 0.7642, "step": 431 }, { "epoch": 0.36355985693246373, "grad_norm": 111.3463363647461, "learning_rate": 4.856609939666251e-06, "loss": 0.8334, "step": 432 }, { "epoch": 0.3644014306753629, "grad_norm": 113.79244995117188, "learning_rate": 4.855863575120168e-06, "loss": 0.9112, "step": 433 }, { "epoch": 0.36524300441826213, "grad_norm": 123.11966705322266, "learning_rate": 4.855115330802342e-06, "loss": 1.0529, "step": 434 }, { "epoch": 0.36608457816116136, "grad_norm": 108.89775085449219, "learning_rate": 4.854365207309803e-06, "loss": 0.7978, "step": 435 }, { "epoch": 0.3669261519040606, "grad_norm": 164.04800415039062, "learning_rate": 4.853613205241086e-06, "loss": 0.9432, "step": 436 }, { "epoch": 0.3677677256469598, "grad_norm": 136.918701171875, "learning_rate": 4.8528593251962206e-06, "loss": 0.982, "step": 437 }, { "epoch": 0.36860929938985904, "grad_norm": 93.6369400024414, "learning_rate": 4.852103567776738e-06, "loss": 0.6342, "step": 438 }, { "epoch": 0.36945087313275826, "grad_norm": 161.514892578125, "learning_rate": 4.851345933585666e-06, "loss": 1.0082, "step": 439 }, { "epoch": 0.3702924468756575, "grad_norm": 120.8597183227539, "learning_rate": 4.850586423227529e-06, "loss": 1.0075, "step": 440 }, { "epoch": 0.3711340206185567, "grad_norm": 146.28897094726562, "learning_rate": 4.849825037308351e-06, "loss": 0.8819, "step": 441 }, { "epoch": 0.37197559436145594, "grad_norm": 133.81219482421875, "learning_rate": 4.849061776435649e-06, "loss": 0.7981, "step": 442 }, { "epoch": 0.37281716810435517, "grad_norm": 126.19327545166016, "learning_rate": 4.8482966412184394e-06, "loss": 0.9471, "step": 443 }, { "epoch": 0.37365874184725434, "grad_norm": 128.17840576171875, "learning_rate": 4.847529632267233e-06, "loss": 0.7527, "step": 444 }, { "epoch": 0.37450031559015357, "grad_norm": 122.65428161621094, "learning_rate": 4.846760750194033e-06, "loss": 0.93, "step": 445 }, { "epoch": 0.3753418893330528, "grad_norm": 127.0604019165039, "learning_rate": 4.845989995612342e-06, "loss": 0.8162, "step": 446 }, { "epoch": 0.376183463075952, "grad_norm": 141.19271850585938, "learning_rate": 4.8452173691371545e-06, "loss": 0.9245, "step": 447 }, { "epoch": 0.37702503681885124, "grad_norm": 145.30938720703125, "learning_rate": 4.844442871384957e-06, "loss": 1.1939, "step": 448 }, { "epoch": 0.37786661056175047, "grad_norm": 95.39657592773438, "learning_rate": 4.843666502973731e-06, "loss": 0.8952, "step": 449 }, { "epoch": 0.3787081843046497, "grad_norm": 143.34825134277344, "learning_rate": 4.84288826452295e-06, "loss": 1.1639, "step": 450 }, { "epoch": 0.3795497580475489, "grad_norm": 110.71271514892578, "learning_rate": 4.842108156653581e-06, "loss": 0.8824, "step": 451 }, { "epoch": 0.38039133179044815, "grad_norm": 100.15702819824219, "learning_rate": 4.84132617998808e-06, "loss": 0.7559, "step": 452 }, { "epoch": 0.3812329055333474, "grad_norm": 108.70226287841797, "learning_rate": 4.840542335150395e-06, "loss": 0.9015, "step": 453 }, { "epoch": 0.3820744792762466, "grad_norm": 127.7223892211914, "learning_rate": 4.839756622765966e-06, "loss": 0.8032, "step": 454 }, { "epoch": 0.38291605301914583, "grad_norm": 183.36643981933594, "learning_rate": 4.838969043461725e-06, "loss": 1.3108, "step": 455 }, { "epoch": 0.383757626762045, "grad_norm": 93.0749740600586, "learning_rate": 4.838179597866085e-06, "loss": 0.6992, "step": 456 }, { "epoch": 0.3845992005049442, "grad_norm": 218.0874481201172, "learning_rate": 4.83738828660896e-06, "loss": 1.1059, "step": 457 }, { "epoch": 0.38544077424784345, "grad_norm": 129.31700134277344, "learning_rate": 4.836595110321744e-06, "loss": 0.686, "step": 458 }, { "epoch": 0.3862823479907427, "grad_norm": 113.3348388671875, "learning_rate": 4.835800069637321e-06, "loss": 0.7151, "step": 459 }, { "epoch": 0.3871239217336419, "grad_norm": 242.91207885742188, "learning_rate": 4.835003165190064e-06, "loss": 1.1047, "step": 460 }, { "epoch": 0.38796549547654113, "grad_norm": 107.70856475830078, "learning_rate": 4.8342043976158344e-06, "loss": 0.8914, "step": 461 }, { "epoch": 0.38880706921944036, "grad_norm": 102.79827117919922, "learning_rate": 4.8334037675519775e-06, "loss": 0.8817, "step": 462 }, { "epoch": 0.3896486429623396, "grad_norm": 87.40475463867188, "learning_rate": 4.832601275637324e-06, "loss": 0.606, "step": 463 }, { "epoch": 0.3904902167052388, "grad_norm": 118.46971893310547, "learning_rate": 4.831796922512194e-06, "loss": 1.0383, "step": 464 }, { "epoch": 0.39133179044813804, "grad_norm": 122.6368179321289, "learning_rate": 4.8309907088183894e-06, "loss": 0.9457, "step": 465 }, { "epoch": 0.39217336419103727, "grad_norm": 123.12105560302734, "learning_rate": 4.830182635199197e-06, "loss": 0.9501, "step": 466 }, { "epoch": 0.39301493793393644, "grad_norm": 168.8681640625, "learning_rate": 4.829372702299391e-06, "loss": 0.7714, "step": 467 }, { "epoch": 0.39385651167683566, "grad_norm": 128.96510314941406, "learning_rate": 4.828560910765223e-06, "loss": 1.0639, "step": 468 }, { "epoch": 0.3946980854197349, "grad_norm": 130.435302734375, "learning_rate": 4.827747261244434e-06, "loss": 0.9294, "step": 469 }, { "epoch": 0.3955396591626341, "grad_norm": 116.36153411865234, "learning_rate": 4.826931754386244e-06, "loss": 0.9145, "step": 470 }, { "epoch": 0.39638123290553334, "grad_norm": 170.28231811523438, "learning_rate": 4.8261143908413554e-06, "loss": 1.1118, "step": 471 }, { "epoch": 0.39722280664843257, "grad_norm": 140.88491821289062, "learning_rate": 4.825295171261952e-06, "loss": 0.8239, "step": 472 }, { "epoch": 0.3980643803913318, "grad_norm": 150.85214233398438, "learning_rate": 4.8244740963017e-06, "loss": 0.8503, "step": 473 }, { "epoch": 0.398905954134231, "grad_norm": 108.1480712890625, "learning_rate": 4.823651166615742e-06, "loss": 1.0041, "step": 474 }, { "epoch": 0.39974752787713025, "grad_norm": 319.3303527832031, "learning_rate": 4.822826382860708e-06, "loss": 1.003, "step": 475 }, { "epoch": 0.4005891016200295, "grad_norm": 123.97578430175781, "learning_rate": 4.8219997456947e-06, "loss": 0.8949, "step": 476 }, { "epoch": 0.4014306753629287, "grad_norm": 209.10208129882812, "learning_rate": 4.821171255777302e-06, "loss": 1.0194, "step": 477 }, { "epoch": 0.4022722491058279, "grad_norm": 141.95541381835938, "learning_rate": 4.820340913769577e-06, "loss": 0.9419, "step": 478 }, { "epoch": 0.4031138228487271, "grad_norm": 140.51358032226562, "learning_rate": 4.819508720334063e-06, "loss": 1.1065, "step": 479 }, { "epoch": 0.4039553965916263, "grad_norm": 97.90631103515625, "learning_rate": 4.818674676134778e-06, "loss": 0.919, "step": 480 }, { "epoch": 0.40479697033452555, "grad_norm": 151.83956909179688, "learning_rate": 4.8178387818372165e-06, "loss": 1.0386, "step": 481 }, { "epoch": 0.4056385440774248, "grad_norm": 112.4679946899414, "learning_rate": 4.817001038108349e-06, "loss": 0.7155, "step": 482 }, { "epoch": 0.406480117820324, "grad_norm": 110.14672088623047, "learning_rate": 4.816161445616618e-06, "loss": 0.8235, "step": 483 }, { "epoch": 0.40732169156322323, "grad_norm": 166.98773193359375, "learning_rate": 4.815320005031949e-06, "loss": 1.3208, "step": 484 }, { "epoch": 0.40816326530612246, "grad_norm": 116.70498657226562, "learning_rate": 4.814476717025734e-06, "loss": 0.9825, "step": 485 }, { "epoch": 0.4090048390490217, "grad_norm": 114.12043762207031, "learning_rate": 4.813631582270844e-06, "loss": 0.9469, "step": 486 }, { "epoch": 0.4098464127919209, "grad_norm": 118.60401916503906, "learning_rate": 4.812784601441622e-06, "loss": 0.8652, "step": 487 }, { "epoch": 0.41068798653482014, "grad_norm": 142.69345092773438, "learning_rate": 4.811935775213884e-06, "loss": 0.7706, "step": 488 }, { "epoch": 0.4115295602777193, "grad_norm": 129.11827087402344, "learning_rate": 4.811085104264919e-06, "loss": 0.986, "step": 489 }, { "epoch": 0.41237113402061853, "grad_norm": 126.18577575683594, "learning_rate": 4.810232589273487e-06, "loss": 0.9475, "step": 490 }, { "epoch": 0.41321270776351776, "grad_norm": 120.70259857177734, "learning_rate": 4.80937823091982e-06, "loss": 0.9647, "step": 491 }, { "epoch": 0.414054281506417, "grad_norm": 113.89292907714844, "learning_rate": 4.808522029885622e-06, "loss": 0.889, "step": 492 }, { "epoch": 0.4148958552493162, "grad_norm": 97.43743896484375, "learning_rate": 4.8076639868540635e-06, "loss": 0.7151, "step": 493 }, { "epoch": 0.41573742899221544, "grad_norm": 105.58618927001953, "learning_rate": 4.8068041025097895e-06, "loss": 0.9715, "step": 494 }, { "epoch": 0.41657900273511467, "grad_norm": 127.87680053710938, "learning_rate": 4.805942377538911e-06, "loss": 0.8187, "step": 495 }, { "epoch": 0.4174205764780139, "grad_norm": 134.050048828125, "learning_rate": 4.805078812629009e-06, "loss": 0.8815, "step": 496 }, { "epoch": 0.4182621502209131, "grad_norm": 108.08966064453125, "learning_rate": 4.804213408469133e-06, "loss": 0.8102, "step": 497 }, { "epoch": 0.41910372396381235, "grad_norm": 129.20497131347656, "learning_rate": 4.803346165749798e-06, "loss": 0.6703, "step": 498 }, { "epoch": 0.4199452977067116, "grad_norm": 146.5807647705078, "learning_rate": 4.802477085162988e-06, "loss": 0.857, "step": 499 }, { "epoch": 0.4207868714496108, "grad_norm": 121.63455200195312, "learning_rate": 4.801606167402152e-06, "loss": 0.9737, "step": 500 }, { "epoch": 0.42162844519250997, "grad_norm": 112.13333892822266, "learning_rate": 4.800733413162206e-06, "loss": 0.8266, "step": 501 }, { "epoch": 0.4224700189354092, "grad_norm": 105.02662658691406, "learning_rate": 4.799858823139532e-06, "loss": 0.7877, "step": 502 }, { "epoch": 0.4233115926783084, "grad_norm": 100.8410415649414, "learning_rate": 4.798982398031976e-06, "loss": 0.8558, "step": 503 }, { "epoch": 0.42415316642120765, "grad_norm": 141.06207275390625, "learning_rate": 4.798104138538848e-06, "loss": 1.1098, "step": 504 }, { "epoch": 0.4249947401641069, "grad_norm": 631.0702514648438, "learning_rate": 4.79722404536092e-06, "loss": 1.371, "step": 505 }, { "epoch": 0.4258363139070061, "grad_norm": 133.77012634277344, "learning_rate": 4.79634211920043e-06, "loss": 1.0773, "step": 506 }, { "epoch": 0.42667788764990533, "grad_norm": 104.05528259277344, "learning_rate": 4.795458360761078e-06, "loss": 0.8034, "step": 507 }, { "epoch": 0.42751946139280456, "grad_norm": 135.33978271484375, "learning_rate": 4.794572770748025e-06, "loss": 0.9243, "step": 508 }, { "epoch": 0.4283610351357038, "grad_norm": 143.45217895507812, "learning_rate": 4.793685349867893e-06, "loss": 1.0566, "step": 509 }, { "epoch": 0.429202608878603, "grad_norm": 136.93051147460938, "learning_rate": 4.792796098828768e-06, "loss": 0.8205, "step": 510 }, { "epoch": 0.43004418262150224, "grad_norm": 124.12754821777344, "learning_rate": 4.791905018340191e-06, "loss": 0.7642, "step": 511 }, { "epoch": 0.4308857563644014, "grad_norm": 119.36278533935547, "learning_rate": 4.791012109113168e-06, "loss": 0.8458, "step": 512 }, { "epoch": 0.43172733010730063, "grad_norm": 123.41871643066406, "learning_rate": 4.790117371860162e-06, "loss": 0.9155, "step": 513 }, { "epoch": 0.43256890385019986, "grad_norm": 99.3111572265625, "learning_rate": 4.789220807295092e-06, "loss": 0.8699, "step": 514 }, { "epoch": 0.4334104775930991, "grad_norm": 115.54022979736328, "learning_rate": 4.788322416133341e-06, "loss": 0.8808, "step": 515 }, { "epoch": 0.4342520513359983, "grad_norm": 108.34188079833984, "learning_rate": 4.7874221990917445e-06, "loss": 0.8901, "step": 516 }, { "epoch": 0.43509362507889754, "grad_norm": 105.6187973022461, "learning_rate": 4.786520156888596e-06, "loss": 0.8886, "step": 517 }, { "epoch": 0.43593519882179677, "grad_norm": 108.33097076416016, "learning_rate": 4.7856162902436465e-06, "loss": 0.9418, "step": 518 }, { "epoch": 0.436776772564696, "grad_norm": 121.62516784667969, "learning_rate": 4.784710599878102e-06, "loss": 0.8255, "step": 519 }, { "epoch": 0.4376183463075952, "grad_norm": 151.26263427734375, "learning_rate": 4.7838030865146236e-06, "loss": 1.1101, "step": 520 }, { "epoch": 0.43845992005049444, "grad_norm": 125.10527801513672, "learning_rate": 4.782893750877326e-06, "loss": 1.101, "step": 521 }, { "epoch": 0.43930149379339367, "grad_norm": 122.728271484375, "learning_rate": 4.781982593691781e-06, "loss": 0.933, "step": 522 }, { "epoch": 0.44014306753629284, "grad_norm": 133.12278747558594, "learning_rate": 4.78106961568501e-06, "loss": 0.9782, "step": 523 }, { "epoch": 0.44098464127919207, "grad_norm": 138.73318481445312, "learning_rate": 4.78015481758549e-06, "loss": 0.834, "step": 524 }, { "epoch": 0.4418262150220913, "grad_norm": 128.3526153564453, "learning_rate": 4.77923820012315e-06, "loss": 0.9034, "step": 525 }, { "epoch": 0.4426677887649905, "grad_norm": 108.43571472167969, "learning_rate": 4.778319764029368e-06, "loss": 0.8541, "step": 526 }, { "epoch": 0.44350936250788975, "grad_norm": 116.94001007080078, "learning_rate": 4.777399510036977e-06, "loss": 0.958, "step": 527 }, { "epoch": 0.444350936250789, "grad_norm": 120.42687225341797, "learning_rate": 4.776477438880258e-06, "loss": 1.0345, "step": 528 }, { "epoch": 0.4451925099936882, "grad_norm": 111.29932403564453, "learning_rate": 4.775553551294942e-06, "loss": 0.8609, "step": 529 }, { "epoch": 0.44603408373658743, "grad_norm": 131.2679901123047, "learning_rate": 4.774627848018211e-06, "loss": 1.1318, "step": 530 }, { "epoch": 0.44687565747948665, "grad_norm": 127.53345489501953, "learning_rate": 4.773700329788696e-06, "loss": 0.9082, "step": 531 }, { "epoch": 0.4477172312223859, "grad_norm": 124.14596557617188, "learning_rate": 4.772770997346471e-06, "loss": 1.0633, "step": 532 }, { "epoch": 0.4485588049652851, "grad_norm": 97.10338592529297, "learning_rate": 4.7718398514330655e-06, "loss": 0.7285, "step": 533 }, { "epoch": 0.4494003787081843, "grad_norm": 86.35016632080078, "learning_rate": 4.770906892791449e-06, "loss": 0.5934, "step": 534 }, { "epoch": 0.4502419524510835, "grad_norm": 112.21371459960938, "learning_rate": 4.7699721221660425e-06, "loss": 1.0579, "step": 535 }, { "epoch": 0.45108352619398273, "grad_norm": 106.98262023925781, "learning_rate": 4.7690355403027086e-06, "loss": 0.8364, "step": 536 }, { "epoch": 0.45192509993688196, "grad_norm": 122.70458984375, "learning_rate": 4.768097147948759e-06, "loss": 0.8282, "step": 537 }, { "epoch": 0.4527666736797812, "grad_norm": 115.28579711914062, "learning_rate": 4.767156945852948e-06, "loss": 0.7753, "step": 538 }, { "epoch": 0.4536082474226804, "grad_norm": 131.470458984375, "learning_rate": 4.7662149347654725e-06, "loss": 1.0653, "step": 539 }, { "epoch": 0.45444982116557964, "grad_norm": 119.49797821044922, "learning_rate": 4.7652711154379765e-06, "loss": 0.9991, "step": 540 }, { "epoch": 0.45529139490847886, "grad_norm": 126.73249053955078, "learning_rate": 4.764325488623544e-06, "loss": 1.1871, "step": 541 }, { "epoch": 0.4561329686513781, "grad_norm": 95.57186889648438, "learning_rate": 4.7633780550767e-06, "loss": 0.9152, "step": 542 }, { "epoch": 0.4569745423942773, "grad_norm": 119.88944244384766, "learning_rate": 4.762428815553416e-06, "loss": 0.818, "step": 543 }, { "epoch": 0.45781611613717654, "grad_norm": 105.79959106445312, "learning_rate": 4.761477770811101e-06, "loss": 0.9361, "step": 544 }, { "epoch": 0.4586576898800757, "grad_norm": 119.57502746582031, "learning_rate": 4.760524921608604e-06, "loss": 0.9895, "step": 545 }, { "epoch": 0.45949926362297494, "grad_norm": 114.09158325195312, "learning_rate": 4.759570268706214e-06, "loss": 1.1529, "step": 546 }, { "epoch": 0.46034083736587417, "grad_norm": 100.0584487915039, "learning_rate": 4.758613812865662e-06, "loss": 0.7656, "step": 547 }, { "epoch": 0.4611824111087734, "grad_norm": 100.03468322753906, "learning_rate": 4.757655554850115e-06, "loss": 0.7772, "step": 548 }, { "epoch": 0.4620239848516726, "grad_norm": 114.32467651367188, "learning_rate": 4.756695495424177e-06, "loss": 0.8262, "step": 549 }, { "epoch": 0.46286555859457185, "grad_norm": 93.30988311767578, "learning_rate": 4.755733635353894e-06, "loss": 0.8206, "step": 550 }, { "epoch": 0.4637071323374711, "grad_norm": 90.66636657714844, "learning_rate": 4.7547699754067425e-06, "loss": 0.7546, "step": 551 }, { "epoch": 0.4645487060803703, "grad_norm": 84.4310302734375, "learning_rate": 4.75380451635164e-06, "loss": 0.7591, "step": 552 }, { "epoch": 0.4653902798232695, "grad_norm": 155.35433959960938, "learning_rate": 4.752837258958937e-06, "loss": 1.2789, "step": 553 }, { "epoch": 0.46623185356616875, "grad_norm": 82.25304412841797, "learning_rate": 4.751868204000421e-06, "loss": 0.686, "step": 554 }, { "epoch": 0.467073427309068, "grad_norm": 110.43043518066406, "learning_rate": 4.750897352249311e-06, "loss": 0.8744, "step": 555 }, { "epoch": 0.4679150010519672, "grad_norm": 104.57023620605469, "learning_rate": 4.7499247044802625e-06, "loss": 0.784, "step": 556 }, { "epoch": 0.4687565747948664, "grad_norm": 82.92748260498047, "learning_rate": 4.748950261469362e-06, "loss": 0.7746, "step": 557 }, { "epoch": 0.4695981485377656, "grad_norm": 118.4659194946289, "learning_rate": 4.747974023994129e-06, "loss": 0.7537, "step": 558 }, { "epoch": 0.47043972228066483, "grad_norm": 108.45428466796875, "learning_rate": 4.746995992833516e-06, "loss": 0.9114, "step": 559 }, { "epoch": 0.47128129602356406, "grad_norm": 92.51221466064453, "learning_rate": 4.746016168767906e-06, "loss": 0.819, "step": 560 }, { "epoch": 0.4721228697664633, "grad_norm": 89.94612884521484, "learning_rate": 4.745034552579111e-06, "loss": 0.8209, "step": 561 }, { "epoch": 0.4729644435093625, "grad_norm": 96.07715606689453, "learning_rate": 4.744051145050375e-06, "loss": 0.7277, "step": 562 }, { "epoch": 0.47380601725226174, "grad_norm": 115.94178771972656, "learning_rate": 4.743065946966371e-06, "loss": 0.8407, "step": 563 }, { "epoch": 0.47464759099516096, "grad_norm": 90.5213394165039, "learning_rate": 4.7420789591132e-06, "loss": 0.7891, "step": 564 }, { "epoch": 0.4754891647380602, "grad_norm": 113.68289947509766, "learning_rate": 4.741090182278391e-06, "loss": 1.0005, "step": 565 }, { "epoch": 0.4763307384809594, "grad_norm": 136.67337036132812, "learning_rate": 4.740099617250902e-06, "loss": 1.1607, "step": 566 }, { "epoch": 0.47717231222385864, "grad_norm": 91.25704193115234, "learning_rate": 4.739107264821115e-06, "loss": 0.6981, "step": 567 }, { "epoch": 0.4780138859667578, "grad_norm": 106.67892456054688, "learning_rate": 4.738113125780841e-06, "loss": 0.7727, "step": 568 }, { "epoch": 0.47885545970965704, "grad_norm": 121.09901428222656, "learning_rate": 4.7371172009233155e-06, "loss": 0.8397, "step": 569 }, { "epoch": 0.47969703345255627, "grad_norm": 80.89665222167969, "learning_rate": 4.736119491043199e-06, "loss": 0.7499, "step": 570 }, { "epoch": 0.4805386071954555, "grad_norm": 130.6654815673828, "learning_rate": 4.735119996936576e-06, "loss": 0.992, "step": 571 }, { "epoch": 0.4813801809383547, "grad_norm": 101.22091674804688, "learning_rate": 4.734118719400954e-06, "loss": 0.9348, "step": 572 }, { "epoch": 0.48222175468125394, "grad_norm": 93.3582534790039, "learning_rate": 4.733115659235265e-06, "loss": 0.8983, "step": 573 }, { "epoch": 0.48306332842415317, "grad_norm": 100.23347473144531, "learning_rate": 4.7321108172398625e-06, "loss": 0.8489, "step": 574 }, { "epoch": 0.4839049021670524, "grad_norm": 83.09318542480469, "learning_rate": 4.731104194216523e-06, "loss": 0.709, "step": 575 }, { "epoch": 0.4847464759099516, "grad_norm": 76.45307159423828, "learning_rate": 4.730095790968442e-06, "loss": 0.7334, "step": 576 }, { "epoch": 0.48558804965285085, "grad_norm": 117.21244812011719, "learning_rate": 4.7290856083002355e-06, "loss": 1.0757, "step": 577 }, { "epoch": 0.4864296233957501, "grad_norm": 90.2735595703125, "learning_rate": 4.728073647017942e-06, "loss": 0.8145, "step": 578 }, { "epoch": 0.48727119713864925, "grad_norm": 104.04305267333984, "learning_rate": 4.727059907929018e-06, "loss": 1.0265, "step": 579 }, { "epoch": 0.4881127708815485, "grad_norm": 103.80632019042969, "learning_rate": 4.726044391842337e-06, "loss": 0.9707, "step": 580 }, { "epoch": 0.4889543446244477, "grad_norm": 91.39692687988281, "learning_rate": 4.72502709956819e-06, "loss": 0.8324, "step": 581 }, { "epoch": 0.4897959183673469, "grad_norm": 103.67784881591797, "learning_rate": 4.7240080319182905e-06, "loss": 0.9852, "step": 582 }, { "epoch": 0.49063749211024615, "grad_norm": 92.56071472167969, "learning_rate": 4.722987189705761e-06, "loss": 0.7628, "step": 583 }, { "epoch": 0.4914790658531454, "grad_norm": 92.0321273803711, "learning_rate": 4.721964573745145e-06, "loss": 0.9345, "step": 584 }, { "epoch": 0.4923206395960446, "grad_norm": 109.95680236816406, "learning_rate": 4.7209401848524015e-06, "loss": 0.9331, "step": 585 }, { "epoch": 0.49316221333894383, "grad_norm": 109.92249298095703, "learning_rate": 4.7199140238449e-06, "loss": 1.0453, "step": 586 }, { "epoch": 0.49400378708184306, "grad_norm": 109.58256530761719, "learning_rate": 4.718886091541429e-06, "loss": 1.1754, "step": 587 }, { "epoch": 0.4948453608247423, "grad_norm": 109.40169525146484, "learning_rate": 4.717856388762187e-06, "loss": 1.0512, "step": 588 }, { "epoch": 0.4956869345676415, "grad_norm": 113.2249755859375, "learning_rate": 4.716824916328786e-06, "loss": 0.8632, "step": 589 }, { "epoch": 0.4965285083105407, "grad_norm": 110.68663787841797, "learning_rate": 4.715791675064251e-06, "loss": 1.1118, "step": 590 }, { "epoch": 0.4973700820534399, "grad_norm": 101.6324691772461, "learning_rate": 4.714756665793015e-06, "loss": 0.9311, "step": 591 }, { "epoch": 0.49821165579633914, "grad_norm": 156.46829223632812, "learning_rate": 4.713719889340929e-06, "loss": 0.943, "step": 592 }, { "epoch": 0.49905322953923836, "grad_norm": 106.6951904296875, "learning_rate": 4.712681346535245e-06, "loss": 0.5679, "step": 593 }, { "epoch": 0.4998948032821376, "grad_norm": 114.27716827392578, "learning_rate": 4.71164103820463e-06, "loss": 0.8942, "step": 594 }, { "epoch": 0.5007363770250368, "grad_norm": 88.36841583251953, "learning_rate": 4.71059896517916e-06, "loss": 0.8059, "step": 595 }, { "epoch": 0.501577950767936, "grad_norm": 77.92396545410156, "learning_rate": 4.709555128290316e-06, "loss": 0.789, "step": 596 }, { "epoch": 0.5024195245108353, "grad_norm": 92.30797576904297, "learning_rate": 4.708509528370988e-06, "loss": 0.794, "step": 597 }, { "epoch": 0.5032610982537344, "grad_norm": 114.52457427978516, "learning_rate": 4.707462166255473e-06, "loss": 0.9309, "step": 598 }, { "epoch": 0.5041026719966337, "grad_norm": 206.55442810058594, "learning_rate": 4.7064130427794745e-06, "loss": 1.8481, "step": 599 }, { "epoch": 0.5049442457395329, "grad_norm": 82.25666809082031, "learning_rate": 4.705362158780099e-06, "loss": 0.7768, "step": 600 }, { "epoch": 0.5057858194824322, "grad_norm": 93.18348693847656, "learning_rate": 4.70430951509586e-06, "loss": 0.8917, "step": 601 }, { "epoch": 0.5066273932253313, "grad_norm": 116.30068969726562, "learning_rate": 4.703255112566676e-06, "loss": 1.0472, "step": 602 }, { "epoch": 0.5074689669682306, "grad_norm": 99.05022430419922, "learning_rate": 4.702198952033866e-06, "loss": 0.9099, "step": 603 }, { "epoch": 0.5083105407111298, "grad_norm": 99.7860336303711, "learning_rate": 4.701141034340153e-06, "loss": 0.9021, "step": 604 }, { "epoch": 0.5091521144540291, "grad_norm": 122.43360137939453, "learning_rate": 4.700081360329664e-06, "loss": 0.8791, "step": 605 }, { "epoch": 0.5099936881969283, "grad_norm": 110.67709350585938, "learning_rate": 4.699019930847923e-06, "loss": 0.9684, "step": 606 }, { "epoch": 0.5108352619398274, "grad_norm": 88.45003509521484, "learning_rate": 4.697956746741858e-06, "loss": 0.5716, "step": 607 }, { "epoch": 0.5116768356827267, "grad_norm": 91.00718688964844, "learning_rate": 4.696891808859797e-06, "loss": 0.8279, "step": 608 }, { "epoch": 0.5125184094256259, "grad_norm": 103.737548828125, "learning_rate": 4.695825118051467e-06, "loss": 0.8202, "step": 609 }, { "epoch": 0.5133599831685252, "grad_norm": 94.69662475585938, "learning_rate": 4.694756675167992e-06, "loss": 0.9734, "step": 610 }, { "epoch": 0.5142015569114243, "grad_norm": 114.9798355102539, "learning_rate": 4.693686481061898e-06, "loss": 1.0855, "step": 611 }, { "epoch": 0.5150431306543236, "grad_norm": 95.23029327392578, "learning_rate": 4.692614536587101e-06, "loss": 0.7172, "step": 612 }, { "epoch": 0.5158847043972228, "grad_norm": 99.6467514038086, "learning_rate": 4.691540842598923e-06, "loss": 0.7589, "step": 613 }, { "epoch": 0.5167262781401221, "grad_norm": 103.03988647460938, "learning_rate": 4.690465399954076e-06, "loss": 0.8081, "step": 614 }, { "epoch": 0.5175678518830212, "grad_norm": 90.93890380859375, "learning_rate": 4.689388209510667e-06, "loss": 0.7581, "step": 615 }, { "epoch": 0.5184094256259205, "grad_norm": 105.35801696777344, "learning_rate": 4.688309272128201e-06, "loss": 0.8253, "step": 616 }, { "epoch": 0.5192509993688197, "grad_norm": 133.68971252441406, "learning_rate": 4.687228588667573e-06, "loss": 0.8134, "step": 617 }, { "epoch": 0.520092573111719, "grad_norm": 107.72319793701172, "learning_rate": 4.686146159991075e-06, "loss": 0.9942, "step": 618 }, { "epoch": 0.5209341468546181, "grad_norm": 110.94938659667969, "learning_rate": 4.685061986962391e-06, "loss": 0.8772, "step": 619 }, { "epoch": 0.5217757205975173, "grad_norm": 102.12600708007812, "learning_rate": 4.6839760704465915e-06, "loss": 0.7658, "step": 620 }, { "epoch": 0.5226172943404166, "grad_norm": 96.44245910644531, "learning_rate": 4.6828884113101454e-06, "loss": 0.9078, "step": 621 }, { "epoch": 0.5234588680833158, "grad_norm": 118.14047241210938, "learning_rate": 4.681799010420909e-06, "loss": 1.0003, "step": 622 }, { "epoch": 0.524300441826215, "grad_norm": 85.10506439208984, "learning_rate": 4.680707868648127e-06, "loss": 0.8619, "step": 623 }, { "epoch": 0.5251420155691142, "grad_norm": 86.90888214111328, "learning_rate": 4.679614986862435e-06, "loss": 0.9455, "step": 624 }, { "epoch": 0.5259835893120135, "grad_norm": 86.18508911132812, "learning_rate": 4.678520365935857e-06, "loss": 0.8459, "step": 625 }, { "epoch": 0.5268251630549127, "grad_norm": 82.22334289550781, "learning_rate": 4.677424006741803e-06, "loss": 0.8193, "step": 626 }, { "epoch": 0.527666736797812, "grad_norm": 80.63818359375, "learning_rate": 4.676325910155072e-06, "loss": 0.9017, "step": 627 }, { "epoch": 0.5285083105407111, "grad_norm": 95.19749450683594, "learning_rate": 4.675226077051849e-06, "loss": 0.9112, "step": 628 }, { "epoch": 0.5293498842836104, "grad_norm": 103.03102111816406, "learning_rate": 4.674124508309702e-06, "loss": 0.7195, "step": 629 }, { "epoch": 0.5301914580265096, "grad_norm": 100.74110412597656, "learning_rate": 4.673021204807588e-06, "loss": 0.8151, "step": 630 }, { "epoch": 0.5310330317694087, "grad_norm": 148.33775329589844, "learning_rate": 4.671916167425844e-06, "loss": 1.073, "step": 631 }, { "epoch": 0.531874605512308, "grad_norm": 131.19943237304688, "learning_rate": 4.670809397046195e-06, "loss": 1.0741, "step": 632 }, { "epoch": 0.5327161792552072, "grad_norm": 109.5487289428711, "learning_rate": 4.669700894551744e-06, "loss": 0.715, "step": 633 }, { "epoch": 0.5335577529981065, "grad_norm": 90.64634704589844, "learning_rate": 4.6685906608269795e-06, "loss": 0.8414, "step": 634 }, { "epoch": 0.5343993267410057, "grad_norm": 144.5821075439453, "learning_rate": 4.667478696757771e-06, "loss": 1.2275, "step": 635 }, { "epoch": 0.5352409004839049, "grad_norm": 94.56639099121094, "learning_rate": 4.666365003231367e-06, "loss": 0.8112, "step": 636 }, { "epoch": 0.5360824742268041, "grad_norm": 88.34557342529297, "learning_rate": 4.665249581136396e-06, "loss": 0.7514, "step": 637 }, { "epoch": 0.5369240479697034, "grad_norm": 112.50604248046875, "learning_rate": 4.6641324313628685e-06, "loss": 0.8959, "step": 638 }, { "epoch": 0.5377656217126026, "grad_norm": 93.03011322021484, "learning_rate": 4.6630135548021714e-06, "loss": 0.7076, "step": 639 }, { "epoch": 0.5386071954555018, "grad_norm": 97.32340240478516, "learning_rate": 4.661892952347069e-06, "loss": 0.9593, "step": 640 }, { "epoch": 0.539448769198401, "grad_norm": 98.1068115234375, "learning_rate": 4.660770624891703e-06, "loss": 0.9363, "step": 641 }, { "epoch": 0.5402903429413002, "grad_norm": 84.54730987548828, "learning_rate": 4.659646573331593e-06, "loss": 0.7832, "step": 642 }, { "epoch": 0.5411319166841995, "grad_norm": 98.7477035522461, "learning_rate": 4.658520798563632e-06, "loss": 0.9713, "step": 643 }, { "epoch": 0.5419734904270986, "grad_norm": 86.43669128417969, "learning_rate": 4.65739330148609e-06, "loss": 0.9527, "step": 644 }, { "epoch": 0.5428150641699979, "grad_norm": 86.94834899902344, "learning_rate": 4.656264082998611e-06, "loss": 0.7682, "step": 645 }, { "epoch": 0.5436566379128971, "grad_norm": 82.33644104003906, "learning_rate": 4.65513314400221e-06, "loss": 0.7847, "step": 646 }, { "epoch": 0.5444982116557964, "grad_norm": 77.75247192382812, "learning_rate": 4.654000485399278e-06, "loss": 0.8332, "step": 647 }, { "epoch": 0.5453397853986955, "grad_norm": 87.43701171875, "learning_rate": 4.652866108093578e-06, "loss": 0.7374, "step": 648 }, { "epoch": 0.5461813591415948, "grad_norm": 104.72925567626953, "learning_rate": 4.6517300129902415e-06, "loss": 1.07, "step": 649 }, { "epoch": 0.547022932884494, "grad_norm": 83.67887878417969, "learning_rate": 4.650592200995773e-06, "loss": 0.893, "step": 650 }, { "epoch": 0.5478645066273933, "grad_norm": 74.65118408203125, "learning_rate": 4.649452673018047e-06, "loss": 0.7125, "step": 651 }, { "epoch": 0.5487060803702924, "grad_norm": 97.34910583496094, "learning_rate": 4.648311429966304e-06, "loss": 0.8686, "step": 652 }, { "epoch": 0.5495476541131916, "grad_norm": 91.98160552978516, "learning_rate": 4.647168472751158e-06, "loss": 0.8258, "step": 653 }, { "epoch": 0.5503892278560909, "grad_norm": 86.38584899902344, "learning_rate": 4.6460238022845885e-06, "loss": 0.6989, "step": 654 }, { "epoch": 0.5512308015989901, "grad_norm": 98.5851058959961, "learning_rate": 4.64487741947994e-06, "loss": 1.05, "step": 655 }, { "epoch": 0.5520723753418894, "grad_norm": 74.66211700439453, "learning_rate": 4.643729325251926e-06, "loss": 0.8472, "step": 656 }, { "epoch": 0.5529139490847885, "grad_norm": 100.06787872314453, "learning_rate": 4.642579520516624e-06, "loss": 0.8999, "step": 657 }, { "epoch": 0.5537555228276878, "grad_norm": 111.21283721923828, "learning_rate": 4.641428006191477e-06, "loss": 0.8044, "step": 658 }, { "epoch": 0.554597096570587, "grad_norm": 102.51060485839844, "learning_rate": 4.6402747831952936e-06, "loss": 0.9319, "step": 659 }, { "epoch": 0.5554386703134863, "grad_norm": 86.2106704711914, "learning_rate": 4.639119852448242e-06, "loss": 0.9393, "step": 660 }, { "epoch": 0.5562802440563854, "grad_norm": 100.43547821044922, "learning_rate": 4.637963214871857e-06, "loss": 0.9186, "step": 661 }, { "epoch": 0.5571218177992847, "grad_norm": 87.44974517822266, "learning_rate": 4.636804871389033e-06, "loss": 0.775, "step": 662 }, { "epoch": 0.5579633915421839, "grad_norm": 86.83062744140625, "learning_rate": 4.635644822924026e-06, "loss": 0.8735, "step": 663 }, { "epoch": 0.558804965285083, "grad_norm": 93.20692443847656, "learning_rate": 4.6344830704024525e-06, "loss": 0.9683, "step": 664 }, { "epoch": 0.5596465390279823, "grad_norm": 103.13855743408203, "learning_rate": 4.633319614751288e-06, "loss": 0.789, "step": 665 }, { "epoch": 0.5604881127708815, "grad_norm": 85.04438781738281, "learning_rate": 4.63215445689887e-06, "loss": 0.8745, "step": 666 }, { "epoch": 0.5613296865137808, "grad_norm": 92.2657699584961, "learning_rate": 4.6309875977748895e-06, "loss": 0.6973, "step": 667 }, { "epoch": 0.56217126025668, "grad_norm": 107.98641967773438, "learning_rate": 4.629819038310399e-06, "loss": 0.9694, "step": 668 }, { "epoch": 0.5630128339995792, "grad_norm": 105.74818420410156, "learning_rate": 4.628648779437805e-06, "loss": 1.0651, "step": 669 }, { "epoch": 0.5638544077424784, "grad_norm": 93.05474853515625, "learning_rate": 4.6274768220908715e-06, "loss": 0.8287, "step": 670 }, { "epoch": 0.5646959814853777, "grad_norm": 101.81683349609375, "learning_rate": 4.626303167204717e-06, "loss": 1.0467, "step": 671 }, { "epoch": 0.5655375552282769, "grad_norm": 108.88080596923828, "learning_rate": 4.6251278157158155e-06, "loss": 1.0961, "step": 672 }, { "epoch": 0.5663791289711761, "grad_norm": 99.9627456665039, "learning_rate": 4.623950768561993e-06, "loss": 0.9827, "step": 673 }, { "epoch": 0.5672207027140753, "grad_norm": 74.0265884399414, "learning_rate": 4.622772026682431e-06, "loss": 0.7854, "step": 674 }, { "epoch": 0.5680622764569745, "grad_norm": 97.81053161621094, "learning_rate": 4.621591591017659e-06, "loss": 0.8921, "step": 675 }, { "epoch": 0.5689038501998738, "grad_norm": 98.89409637451172, "learning_rate": 4.6204094625095645e-06, "loss": 1.0033, "step": 676 }, { "epoch": 0.5697454239427729, "grad_norm": 260.55377197265625, "learning_rate": 4.619225642101379e-06, "loss": 0.6616, "step": 677 }, { "epoch": 0.5705869976856722, "grad_norm": 88.79230499267578, "learning_rate": 4.618040130737689e-06, "loss": 0.8913, "step": 678 }, { "epoch": 0.5714285714285714, "grad_norm": 88.83299255371094, "learning_rate": 4.616852929364426e-06, "loss": 0.9479, "step": 679 }, { "epoch": 0.5722701451714707, "grad_norm": 96.11365509033203, "learning_rate": 4.615664038928874e-06, "loss": 0.9281, "step": 680 }, { "epoch": 0.5731117189143698, "grad_norm": 85.7844009399414, "learning_rate": 4.6144734603796625e-06, "loss": 0.6493, "step": 681 }, { "epoch": 0.5739532926572691, "grad_norm": 92.83222961425781, "learning_rate": 4.613281194666768e-06, "loss": 0.8968, "step": 682 }, { "epoch": 0.5747948664001683, "grad_norm": 99.30359649658203, "learning_rate": 4.612087242741513e-06, "loss": 0.841, "step": 683 }, { "epoch": 0.5756364401430676, "grad_norm": 82.88372802734375, "learning_rate": 4.610891605556568e-06, "loss": 0.7645, "step": 684 }, { "epoch": 0.5764780138859668, "grad_norm": 101.14331817626953, "learning_rate": 4.609694284065945e-06, "loss": 0.9215, "step": 685 }, { "epoch": 0.5773195876288659, "grad_norm": 94.408203125, "learning_rate": 4.608495279225e-06, "loss": 0.7959, "step": 686 }, { "epoch": 0.5781611613717652, "grad_norm": 105.04963684082031, "learning_rate": 4.607294591990435e-06, "loss": 1.0119, "step": 687 }, { "epoch": 0.5790027351146644, "grad_norm": 110.18367767333984, "learning_rate": 4.606092223320293e-06, "loss": 1.0655, "step": 688 }, { "epoch": 0.5798443088575637, "grad_norm": 97.40960693359375, "learning_rate": 4.604888174173957e-06, "loss": 0.9047, "step": 689 }, { "epoch": 0.5806858826004628, "grad_norm": 82.741943359375, "learning_rate": 4.603682445512153e-06, "loss": 0.6259, "step": 690 }, { "epoch": 0.5815274563433621, "grad_norm": 75.7352523803711, "learning_rate": 4.602475038296946e-06, "loss": 0.6802, "step": 691 }, { "epoch": 0.5823690300862613, "grad_norm": 92.63795471191406, "learning_rate": 4.601265953491741e-06, "loss": 0.7462, "step": 692 }, { "epoch": 0.5832106038291606, "grad_norm": 107.34371948242188, "learning_rate": 4.60005519206128e-06, "loss": 0.9377, "step": 693 }, { "epoch": 0.5840521775720597, "grad_norm": 89.72711181640625, "learning_rate": 4.5988427549716465e-06, "loss": 0.8217, "step": 694 }, { "epoch": 0.584893751314959, "grad_norm": 98.20314025878906, "learning_rate": 4.597628643190256e-06, "loss": 0.8939, "step": 695 }, { "epoch": 0.5857353250578582, "grad_norm": 113.12525177001953, "learning_rate": 4.596412857685865e-06, "loss": 0.9188, "step": 696 }, { "epoch": 0.5865768988007574, "grad_norm": 103.85059356689453, "learning_rate": 4.595195399428562e-06, "loss": 1.0561, "step": 697 }, { "epoch": 0.5874184725436566, "grad_norm": 108.72787475585938, "learning_rate": 4.5939762693897704e-06, "loss": 0.8937, "step": 698 }, { "epoch": 0.5882600462865558, "grad_norm": 99.8670425415039, "learning_rate": 4.592755468542251e-06, "loss": 0.8651, "step": 699 }, { "epoch": 0.5891016200294551, "grad_norm": 91.36952209472656, "learning_rate": 4.591532997860094e-06, "loss": 0.8302, "step": 700 }, { "epoch": 0.5899431937723543, "grad_norm": 99.10494995117188, "learning_rate": 4.590308858318725e-06, "loss": 0.8918, "step": 701 }, { "epoch": 0.5907847675152536, "grad_norm": 88.3750228881836, "learning_rate": 4.589083050894898e-06, "loss": 0.8718, "step": 702 }, { "epoch": 0.5916263412581527, "grad_norm": 82.0086669921875, "learning_rate": 4.587855576566699e-06, "loss": 0.7089, "step": 703 }, { "epoch": 0.592467915001052, "grad_norm": 75.7967529296875, "learning_rate": 4.586626436313546e-06, "loss": 0.7745, "step": 704 }, { "epoch": 0.5933094887439512, "grad_norm": 95.32615661621094, "learning_rate": 4.585395631116184e-06, "loss": 0.925, "step": 705 }, { "epoch": 0.5941510624868505, "grad_norm": 91.64199829101562, "learning_rate": 4.584163161956687e-06, "loss": 0.7124, "step": 706 }, { "epoch": 0.5949926362297496, "grad_norm": 96.98888397216797, "learning_rate": 4.582929029818457e-06, "loss": 0.8887, "step": 707 }, { "epoch": 0.5958342099726488, "grad_norm": 130.055908203125, "learning_rate": 4.5816932356862236e-06, "loss": 1.0215, "step": 708 }, { "epoch": 0.5966757837155481, "grad_norm": 91.83544158935547, "learning_rate": 4.58045578054604e-06, "loss": 0.9852, "step": 709 }, { "epoch": 0.5975173574584473, "grad_norm": 121.96644592285156, "learning_rate": 4.579216665385288e-06, "loss": 0.7534, "step": 710 }, { "epoch": 0.5983589312013465, "grad_norm": 95.6441650390625, "learning_rate": 4.577975891192672e-06, "loss": 0.848, "step": 711 }, { "epoch": 0.5992005049442457, "grad_norm": 90.83454895019531, "learning_rate": 4.57673345895822e-06, "loss": 0.8351, "step": 712 }, { "epoch": 0.600042078687145, "grad_norm": 83.06048583984375, "learning_rate": 4.575489369673285e-06, "loss": 0.9874, "step": 713 }, { "epoch": 0.6008836524300442, "grad_norm": 81.80094146728516, "learning_rate": 4.5742436243305385e-06, "loss": 0.9504, "step": 714 }, { "epoch": 0.6017252261729434, "grad_norm": 79.86914825439453, "learning_rate": 4.572996223923979e-06, "loss": 0.8618, "step": 715 }, { "epoch": 0.6025667999158426, "grad_norm": 90.9134750366211, "learning_rate": 4.571747169448918e-06, "loss": 0.6886, "step": 716 }, { "epoch": 0.6034083736587419, "grad_norm": 76.10773468017578, "learning_rate": 4.570496461901994e-06, "loss": 0.7606, "step": 717 }, { "epoch": 0.6042499474016411, "grad_norm": 84.14070129394531, "learning_rate": 4.569244102281161e-06, "loss": 0.8042, "step": 718 }, { "epoch": 0.6050915211445403, "grad_norm": 90.08390045166016, "learning_rate": 4.567990091585691e-06, "loss": 0.8287, "step": 719 }, { "epoch": 0.6059330948874395, "grad_norm": 81.404296875, "learning_rate": 4.566734430816174e-06, "loss": 0.7245, "step": 720 }, { "epoch": 0.6067746686303387, "grad_norm": 91.73657989501953, "learning_rate": 4.565477120974517e-06, "loss": 0.7274, "step": 721 }, { "epoch": 0.607616242373238, "grad_norm": 99.26171112060547, "learning_rate": 4.564218163063944e-06, "loss": 0.8198, "step": 722 }, { "epoch": 0.6084578161161371, "grad_norm": 85.09310913085938, "learning_rate": 4.562957558088992e-06, "loss": 0.6707, "step": 723 }, { "epoch": 0.6092993898590364, "grad_norm": 63.053466796875, "learning_rate": 4.56169530705551e-06, "loss": 0.5859, "step": 724 }, { "epoch": 0.6101409636019356, "grad_norm": 93.56147003173828, "learning_rate": 4.560431410970667e-06, "loss": 0.815, "step": 725 }, { "epoch": 0.6109825373448349, "grad_norm": 83.997314453125, "learning_rate": 4.55916587084294e-06, "loss": 0.7257, "step": 726 }, { "epoch": 0.611824111087734, "grad_norm": 81.07512664794922, "learning_rate": 4.5578986876821165e-06, "loss": 0.7207, "step": 727 }, { "epoch": 0.6126656848306333, "grad_norm": 89.04486846923828, "learning_rate": 4.556629862499299e-06, "loss": 0.7028, "step": 728 }, { "epoch": 0.6135072585735325, "grad_norm": 115.55461120605469, "learning_rate": 4.555359396306898e-06, "loss": 0.9597, "step": 729 }, { "epoch": 0.6143488323164318, "grad_norm": 78.45919799804688, "learning_rate": 4.5540872901186325e-06, "loss": 0.737, "step": 730 }, { "epoch": 0.615190406059331, "grad_norm": 105.65302276611328, "learning_rate": 4.552813544949533e-06, "loss": 0.8943, "step": 731 }, { "epoch": 0.6160319798022301, "grad_norm": 86.27177429199219, "learning_rate": 4.551538161815935e-06, "loss": 0.6789, "step": 732 }, { "epoch": 0.6168735535451294, "grad_norm": 111.21141052246094, "learning_rate": 4.550261141735483e-06, "loss": 1.0777, "step": 733 }, { "epoch": 0.6177151272880286, "grad_norm": 75.09874725341797, "learning_rate": 4.548982485727125e-06, "loss": 0.6761, "step": 734 }, { "epoch": 0.6185567010309279, "grad_norm": 108.20618438720703, "learning_rate": 4.547702194811117e-06, "loss": 1.0618, "step": 735 }, { "epoch": 0.619398274773827, "grad_norm": 84.72933959960938, "learning_rate": 4.546420270009018e-06, "loss": 0.7433, "step": 736 }, { "epoch": 0.6202398485167263, "grad_norm": 89.56011962890625, "learning_rate": 4.5451367123436905e-06, "loss": 0.8188, "step": 737 }, { "epoch": 0.6210814222596255, "grad_norm": 91.85026550292969, "learning_rate": 4.543851522839303e-06, "loss": 0.9414, "step": 738 }, { "epoch": 0.6219229960025248, "grad_norm": 84.19593048095703, "learning_rate": 4.542564702521321e-06, "loss": 0.8061, "step": 739 }, { "epoch": 0.6227645697454239, "grad_norm": 71.855224609375, "learning_rate": 4.541276252416516e-06, "loss": 0.7419, "step": 740 }, { "epoch": 0.6236061434883232, "grad_norm": 120.99076080322266, "learning_rate": 4.539986173552955e-06, "loss": 0.8866, "step": 741 }, { "epoch": 0.6244477172312224, "grad_norm": 98.39362335205078, "learning_rate": 4.53869446696001e-06, "loss": 0.8933, "step": 742 }, { "epoch": 0.6252892909741216, "grad_norm": 89.82457733154297, "learning_rate": 4.537401133668347e-06, "loss": 0.8136, "step": 743 }, { "epoch": 0.6261308647170208, "grad_norm": 93.52442932128906, "learning_rate": 4.536106174709933e-06, "loss": 0.8235, "step": 744 }, { "epoch": 0.62697243845992, "grad_norm": 118.60028076171875, "learning_rate": 4.534809591118033e-06, "loss": 1.1256, "step": 745 }, { "epoch": 0.6278140122028193, "grad_norm": 72.14741516113281, "learning_rate": 4.533511383927204e-06, "loss": 0.7877, "step": 746 }, { "epoch": 0.6286555859457185, "grad_norm": 73.36578369140625, "learning_rate": 4.532211554173303e-06, "loss": 0.7578, "step": 747 }, { "epoch": 0.6294971596886177, "grad_norm": 120.38871002197266, "learning_rate": 4.530910102893478e-06, "loss": 1.3195, "step": 748 }, { "epoch": 0.6303387334315169, "grad_norm": 97.33341979980469, "learning_rate": 4.529607031126173e-06, "loss": 1.1318, "step": 749 }, { "epoch": 0.6311803071744162, "grad_norm": 81.80436706542969, "learning_rate": 4.528302339911125e-06, "loss": 0.7518, "step": 750 }, { "epoch": 0.6320218809173154, "grad_norm": 86.9666519165039, "learning_rate": 4.526996030289363e-06, "loss": 0.8304, "step": 751 }, { "epoch": 0.6328634546602147, "grad_norm": 77.8279800415039, "learning_rate": 4.525688103303206e-06, "loss": 0.7867, "step": 752 }, { "epoch": 0.6337050284031138, "grad_norm": 67.7917251586914, "learning_rate": 4.524378559996267e-06, "loss": 0.7116, "step": 753 }, { "epoch": 0.634546602146013, "grad_norm": 82.60803985595703, "learning_rate": 4.5230674014134425e-06, "loss": 0.7545, "step": 754 }, { "epoch": 0.6353881758889123, "grad_norm": 84.61490631103516, "learning_rate": 4.521754628600925e-06, "loss": 0.7934, "step": 755 }, { "epoch": 0.6362297496318114, "grad_norm": 87.46534729003906, "learning_rate": 4.52044024260619e-06, "loss": 0.8427, "step": 756 }, { "epoch": 0.6370713233747107, "grad_norm": 165.8030548095703, "learning_rate": 4.519124244478002e-06, "loss": 0.7902, "step": 757 }, { "epoch": 0.6379128971176099, "grad_norm": 72.46358489990234, "learning_rate": 4.517806635266412e-06, "loss": 0.6943, "step": 758 }, { "epoch": 0.6387544708605092, "grad_norm": 85.27467346191406, "learning_rate": 4.5164874160227555e-06, "loss": 1.0195, "step": 759 }, { "epoch": 0.6395960446034084, "grad_norm": 166.4009552001953, "learning_rate": 4.5151665877996554e-06, "loss": 0.8665, "step": 760 }, { "epoch": 0.6404376183463076, "grad_norm": 120.591552734375, "learning_rate": 4.513844151651013e-06, "loss": 0.9978, "step": 761 }, { "epoch": 0.6412791920892068, "grad_norm": 84.2991714477539, "learning_rate": 4.512520108632019e-06, "loss": 1.0485, "step": 762 }, { "epoch": 0.6421207658321061, "grad_norm": 81.13017272949219, "learning_rate": 4.5111944597991405e-06, "loss": 0.6065, "step": 763 }, { "epoch": 0.6429623395750053, "grad_norm": 83.88569641113281, "learning_rate": 4.5098672062101316e-06, "loss": 0.9585, "step": 764 }, { "epoch": 0.6438039133179044, "grad_norm": 87.67027282714844, "learning_rate": 4.5085383489240205e-06, "loss": 0.5934, "step": 765 }, { "epoch": 0.6446454870608037, "grad_norm": 144.13389587402344, "learning_rate": 4.5072078890011206e-06, "loss": 1.012, "step": 766 }, { "epoch": 0.6454870608037029, "grad_norm": 110.44563293457031, "learning_rate": 4.5058758275030215e-06, "loss": 0.96, "step": 767 }, { "epoch": 0.6463286345466022, "grad_norm": 94.65059661865234, "learning_rate": 4.5045421654925905e-06, "loss": 0.7704, "step": 768 }, { "epoch": 0.6471702082895013, "grad_norm": 80.62268829345703, "learning_rate": 4.503206904033973e-06, "loss": 0.88, "step": 769 }, { "epoch": 0.6480117820324006, "grad_norm": 80.2649154663086, "learning_rate": 4.501870044192589e-06, "loss": 0.7423, "step": 770 }, { "epoch": 0.6488533557752998, "grad_norm": 182.56639099121094, "learning_rate": 4.5005315870351365e-06, "loss": 0.7048, "step": 771 }, { "epoch": 0.6496949295181991, "grad_norm": 143.60955810546875, "learning_rate": 4.499191533629586e-06, "loss": 0.7246, "step": 772 }, { "epoch": 0.6505365032610982, "grad_norm": 70.18280792236328, "learning_rate": 4.4978498850451815e-06, "loss": 0.7484, "step": 773 }, { "epoch": 0.6513780770039975, "grad_norm": 71.79647064208984, "learning_rate": 4.4965066423524405e-06, "loss": 0.8044, "step": 774 }, { "epoch": 0.6522196507468967, "grad_norm": 88.14927673339844, "learning_rate": 4.495161806623153e-06, "loss": 0.9869, "step": 775 }, { "epoch": 0.6530612244897959, "grad_norm": 96.6579818725586, "learning_rate": 4.493815378930379e-06, "loss": 1.0004, "step": 776 }, { "epoch": 0.6539027982326951, "grad_norm": 75.86454772949219, "learning_rate": 4.492467360348451e-06, "loss": 0.8189, "step": 777 }, { "epoch": 0.6547443719755943, "grad_norm": 167.09136962890625, "learning_rate": 4.491117751952966e-06, "loss": 0.8028, "step": 778 }, { "epoch": 0.6555859457184936, "grad_norm": 82.6733627319336, "learning_rate": 4.489766554820795e-06, "loss": 0.8775, "step": 779 }, { "epoch": 0.6564275194613928, "grad_norm": 84.00127410888672, "learning_rate": 4.488413770030074e-06, "loss": 0.866, "step": 780 }, { "epoch": 0.657269093204292, "grad_norm": 95.56571197509766, "learning_rate": 4.4870593986602055e-06, "loss": 0.7944, "step": 781 }, { "epoch": 0.6581106669471912, "grad_norm": 76.46224975585938, "learning_rate": 4.48570344179186e-06, "loss": 0.7765, "step": 782 }, { "epoch": 0.6589522406900905, "grad_norm": 95.7029800415039, "learning_rate": 4.48434590050697e-06, "loss": 0.9271, "step": 783 }, { "epoch": 0.6597938144329897, "grad_norm": 74.58334350585938, "learning_rate": 4.482986775888735e-06, "loss": 0.8118, "step": 784 }, { "epoch": 0.660635388175889, "grad_norm": 87.86770629882812, "learning_rate": 4.481626069021616e-06, "loss": 0.9888, "step": 785 }, { "epoch": 0.6614769619187881, "grad_norm": 86.0281753540039, "learning_rate": 4.480263780991338e-06, "loss": 0.7254, "step": 786 }, { "epoch": 0.6623185356616873, "grad_norm": 105.17807006835938, "learning_rate": 4.478899912884888e-06, "loss": 0.8356, "step": 787 }, { "epoch": 0.6631601094045866, "grad_norm": 86.4527816772461, "learning_rate": 4.47753446579051e-06, "loss": 0.7711, "step": 788 }, { "epoch": 0.6640016831474858, "grad_norm": 87.57376861572266, "learning_rate": 4.476167440797713e-06, "loss": 0.9198, "step": 789 }, { "epoch": 0.664843256890385, "grad_norm": 75.603759765625, "learning_rate": 4.474798838997262e-06, "loss": 0.7014, "step": 790 }, { "epoch": 0.6656848306332842, "grad_norm": 74.6491470336914, "learning_rate": 4.47342866148118e-06, "loss": 0.7734, "step": 791 }, { "epoch": 0.6665264043761835, "grad_norm": 78.29629516601562, "learning_rate": 4.472056909342748e-06, "loss": 0.7462, "step": 792 }, { "epoch": 0.6673679781190827, "grad_norm": 84.12577819824219, "learning_rate": 4.470683583676504e-06, "loss": 0.8609, "step": 793 }, { "epoch": 0.6682095518619819, "grad_norm": 84.50099182128906, "learning_rate": 4.469308685578241e-06, "loss": 0.9761, "step": 794 }, { "epoch": 0.6690511256048811, "grad_norm": 78.68242645263672, "learning_rate": 4.467932216145006e-06, "loss": 0.7648, "step": 795 }, { "epoch": 0.6698926993477804, "grad_norm": 78.8578872680664, "learning_rate": 4.466554176475102e-06, "loss": 0.8425, "step": 796 }, { "epoch": 0.6707342730906796, "grad_norm": 99.00289154052734, "learning_rate": 4.465174567668081e-06, "loss": 0.9641, "step": 797 }, { "epoch": 0.6715758468335787, "grad_norm": 70.02457427978516, "learning_rate": 4.463793390824752e-06, "loss": 0.7527, "step": 798 }, { "epoch": 0.672417420576478, "grad_norm": 95.30015563964844, "learning_rate": 4.4624106470471704e-06, "loss": 0.9996, "step": 799 }, { "epoch": 0.6732589943193772, "grad_norm": 72.76326751708984, "learning_rate": 4.461026337438646e-06, "loss": 0.7481, "step": 800 }, { "epoch": 0.6741005680622765, "grad_norm": 112.60525512695312, "learning_rate": 4.459640463103733e-06, "loss": 1.0518, "step": 801 }, { "epoch": 0.6749421418051756, "grad_norm": 74.76424407958984, "learning_rate": 4.45825302514824e-06, "loss": 0.7562, "step": 802 }, { "epoch": 0.6757837155480749, "grad_norm": 76.3163070678711, "learning_rate": 4.456864024679219e-06, "loss": 0.8997, "step": 803 }, { "epoch": 0.6766252892909741, "grad_norm": 81.21278381347656, "learning_rate": 4.4554734628049714e-06, "loss": 0.7025, "step": 804 }, { "epoch": 0.6774668630338734, "grad_norm": 97.0674819946289, "learning_rate": 4.454081340635042e-06, "loss": 0.8344, "step": 805 }, { "epoch": 0.6783084367767725, "grad_norm": 74.96123504638672, "learning_rate": 4.4526876592802205e-06, "loss": 0.7259, "step": 806 }, { "epoch": 0.6791500105196718, "grad_norm": 85.4431381225586, "learning_rate": 4.451292419852544e-06, "loss": 0.8788, "step": 807 }, { "epoch": 0.679991584262571, "grad_norm": 90.90916442871094, "learning_rate": 4.449895623465291e-06, "loss": 1.0065, "step": 808 }, { "epoch": 0.6808331580054703, "grad_norm": 79.66510009765625, "learning_rate": 4.44849727123298e-06, "loss": 0.7831, "step": 809 }, { "epoch": 0.6816747317483695, "grad_norm": 87.62796783447266, "learning_rate": 4.4470973642713745e-06, "loss": 0.8283, "step": 810 }, { "epoch": 0.6825163054912686, "grad_norm": 106.29439544677734, "learning_rate": 4.445695903697477e-06, "loss": 0.8776, "step": 811 }, { "epoch": 0.6833578792341679, "grad_norm": 86.14681243896484, "learning_rate": 4.444292890629529e-06, "loss": 0.7907, "step": 812 }, { "epoch": 0.6841994529770671, "grad_norm": 111.23360443115234, "learning_rate": 4.442888326187012e-06, "loss": 0.9141, "step": 813 }, { "epoch": 0.6850410267199664, "grad_norm": 80.4526138305664, "learning_rate": 4.441482211490644e-06, "loss": 0.7239, "step": 814 }, { "epoch": 0.6858826004628655, "grad_norm": 78.80682373046875, "learning_rate": 4.440074547662382e-06, "loss": 0.7941, "step": 815 }, { "epoch": 0.6867241742057648, "grad_norm": 79.59489440917969, "learning_rate": 4.438665335825416e-06, "loss": 0.8866, "step": 816 }, { "epoch": 0.687565747948664, "grad_norm": 72.70225524902344, "learning_rate": 4.4372545771041755e-06, "loss": 0.6917, "step": 817 }, { "epoch": 0.6884073216915633, "grad_norm": 89.61310577392578, "learning_rate": 4.43584227262432e-06, "loss": 1.0724, "step": 818 }, { "epoch": 0.6892488954344624, "grad_norm": 108.31492614746094, "learning_rate": 4.4344284235127435e-06, "loss": 1.0526, "step": 819 }, { "epoch": 0.6900904691773617, "grad_norm": 64.2813491821289, "learning_rate": 4.433013030897575e-06, "loss": 0.5364, "step": 820 }, { "epoch": 0.6909320429202609, "grad_norm": 83.37664794921875, "learning_rate": 4.431596095908173e-06, "loss": 0.7599, "step": 821 }, { "epoch": 0.6917736166631601, "grad_norm": 82.26396942138672, "learning_rate": 4.430177619675125e-06, "loss": 0.8282, "step": 822 }, { "epoch": 0.6926151904060593, "grad_norm": 90.45413970947266, "learning_rate": 4.428757603330251e-06, "loss": 0.853, "step": 823 }, { "epoch": 0.6934567641489585, "grad_norm": 106.56082916259766, "learning_rate": 4.4273360480066e-06, "loss": 1.0193, "step": 824 }, { "epoch": 0.6942983378918578, "grad_norm": 94.9800033569336, "learning_rate": 4.425912954838448e-06, "loss": 0.8258, "step": 825 }, { "epoch": 0.695139911634757, "grad_norm": 90.53028106689453, "learning_rate": 4.424488324961297e-06, "loss": 0.8751, "step": 826 }, { "epoch": 0.6959814853776563, "grad_norm": 101.14625549316406, "learning_rate": 4.423062159511877e-06, "loss": 0.9244, "step": 827 }, { "epoch": 0.6968230591205554, "grad_norm": 79.58170318603516, "learning_rate": 4.421634459628141e-06, "loss": 0.7385, "step": 828 }, { "epoch": 0.6976646328634547, "grad_norm": 86.39154052734375, "learning_rate": 4.420205226449271e-06, "loss": 0.8993, "step": 829 }, { "epoch": 0.6985062066063539, "grad_norm": 129.21112060546875, "learning_rate": 4.418774461115666e-06, "loss": 0.8467, "step": 830 }, { "epoch": 0.6993477803492532, "grad_norm": 74.53318786621094, "learning_rate": 4.417342164768953e-06, "loss": 0.7344, "step": 831 }, { "epoch": 0.7001893540921523, "grad_norm": 86.27154541015625, "learning_rate": 4.4159083385519765e-06, "loss": 0.7898, "step": 832 }, { "epoch": 0.7010309278350515, "grad_norm": 70.78800964355469, "learning_rate": 4.4144729836088055e-06, "loss": 0.7605, "step": 833 }, { "epoch": 0.7018725015779508, "grad_norm": 89.29914855957031, "learning_rate": 4.413036101084725e-06, "loss": 0.9406, "step": 834 }, { "epoch": 0.70271407532085, "grad_norm": 92.68386840820312, "learning_rate": 4.411597692126242e-06, "loss": 0.9553, "step": 835 }, { "epoch": 0.7035556490637492, "grad_norm": 71.86311340332031, "learning_rate": 4.410157757881081e-06, "loss": 0.7552, "step": 836 }, { "epoch": 0.7043972228066484, "grad_norm": 71.1396713256836, "learning_rate": 4.4087162994981805e-06, "loss": 0.718, "step": 837 }, { "epoch": 0.7052387965495477, "grad_norm": 82.484619140625, "learning_rate": 4.407273318127699e-06, "loss": 0.9462, "step": 838 }, { "epoch": 0.7060803702924469, "grad_norm": 84.81983184814453, "learning_rate": 4.4058288149210085e-06, "loss": 0.9518, "step": 839 }, { "epoch": 0.7069219440353461, "grad_norm": 81.63459014892578, "learning_rate": 4.4043827910306944e-06, "loss": 0.9702, "step": 840 }, { "epoch": 0.7077635177782453, "grad_norm": 91.86296081542969, "learning_rate": 4.402935247610558e-06, "loss": 0.9159, "step": 841 }, { "epoch": 0.7086050915211446, "grad_norm": 70.82006072998047, "learning_rate": 4.401486185815611e-06, "loss": 0.6971, "step": 842 }, { "epoch": 0.7094466652640438, "grad_norm": 93.87294006347656, "learning_rate": 4.400035606802076e-06, "loss": 0.859, "step": 843 }, { "epoch": 0.7102882390069429, "grad_norm": 109.06575775146484, "learning_rate": 4.398583511727388e-06, "loss": 0.9812, "step": 844 }, { "epoch": 0.7111298127498422, "grad_norm": 116.77679443359375, "learning_rate": 4.397129901750192e-06, "loss": 0.9749, "step": 845 }, { "epoch": 0.7119713864927414, "grad_norm": 69.3492431640625, "learning_rate": 4.39567477803034e-06, "loss": 0.7911, "step": 846 }, { "epoch": 0.7128129602356407, "grad_norm": 78.79708099365234, "learning_rate": 4.394218141728893e-06, "loss": 0.9871, "step": 847 }, { "epoch": 0.7136545339785398, "grad_norm": 58.36460876464844, "learning_rate": 4.392759994008118e-06, "loss": 0.7013, "step": 848 }, { "epoch": 0.7144961077214391, "grad_norm": 77.56861114501953, "learning_rate": 4.391300336031489e-06, "loss": 0.8301, "step": 849 }, { "epoch": 0.7153376814643383, "grad_norm": 64.96390533447266, "learning_rate": 4.3898391689636834e-06, "loss": 0.7755, "step": 850 }, { "epoch": 0.7161792552072376, "grad_norm": 75.965087890625, "learning_rate": 4.388376493970585e-06, "loss": 0.7274, "step": 851 }, { "epoch": 0.7170208289501367, "grad_norm": 76.60365295410156, "learning_rate": 4.38691231221928e-06, "loss": 1.0479, "step": 852 }, { "epoch": 0.717862402693036, "grad_norm": 97.6857681274414, "learning_rate": 4.385446624878057e-06, "loss": 1.1624, "step": 853 }, { "epoch": 0.7187039764359352, "grad_norm": 83.55355072021484, "learning_rate": 4.383979433116403e-06, "loss": 0.8381, "step": 854 }, { "epoch": 0.7195455501788344, "grad_norm": 99.44181823730469, "learning_rate": 4.382510738105011e-06, "loss": 1.0486, "step": 855 }, { "epoch": 0.7203871239217337, "grad_norm": 71.99688720703125, "learning_rate": 4.381040541015766e-06, "loss": 0.8731, "step": 856 }, { "epoch": 0.7212286976646328, "grad_norm": 86.90930938720703, "learning_rate": 4.379568843021761e-06, "loss": 0.9857, "step": 857 }, { "epoch": 0.7220702714075321, "grad_norm": 81.49408721923828, "learning_rate": 4.378095645297279e-06, "loss": 0.8581, "step": 858 }, { "epoch": 0.7229118451504313, "grad_norm": 74.4553451538086, "learning_rate": 4.376620949017801e-06, "loss": 0.851, "step": 859 }, { "epoch": 0.7237534188933306, "grad_norm": 92.5252685546875, "learning_rate": 4.3751447553600065e-06, "loss": 0.7895, "step": 860 }, { "epoch": 0.7245949926362297, "grad_norm": 70.13682556152344, "learning_rate": 4.3736670655017675e-06, "loss": 0.8247, "step": 861 }, { "epoch": 0.725436566379129, "grad_norm": 87.14791870117188, "learning_rate": 4.372187880622149e-06, "loss": 0.8072, "step": 862 }, { "epoch": 0.7262781401220282, "grad_norm": 214.95753479003906, "learning_rate": 4.370707201901414e-06, "loss": 0.8754, "step": 863 }, { "epoch": 0.7271197138649275, "grad_norm": 85.0881118774414, "learning_rate": 4.369225030521011e-06, "loss": 0.9258, "step": 864 }, { "epoch": 0.7279612876078266, "grad_norm": 79.23942565917969, "learning_rate": 4.367741367663582e-06, "loss": 0.9405, "step": 865 }, { "epoch": 0.7288028613507258, "grad_norm": 77.9674301147461, "learning_rate": 4.36625621451296e-06, "loss": 0.8144, "step": 866 }, { "epoch": 0.7296444350936251, "grad_norm": 58.98210906982422, "learning_rate": 4.364769572254169e-06, "loss": 0.5694, "step": 867 }, { "epoch": 0.7304860088365243, "grad_norm": 103.85771179199219, "learning_rate": 4.363281442073414e-06, "loss": 0.812, "step": 868 }, { "epoch": 0.7313275825794235, "grad_norm": 96.6537094116211, "learning_rate": 4.361791825158095e-06, "loss": 0.8771, "step": 869 }, { "epoch": 0.7321691563223227, "grad_norm": 71.61028289794922, "learning_rate": 4.360300722696794e-06, "loss": 0.7049, "step": 870 }, { "epoch": 0.733010730065222, "grad_norm": 67.1308364868164, "learning_rate": 4.3588081358792796e-06, "loss": 0.7327, "step": 871 }, { "epoch": 0.7338523038081212, "grad_norm": 87.99250030517578, "learning_rate": 4.357314065896506e-06, "loss": 0.8441, "step": 872 }, { "epoch": 0.7346938775510204, "grad_norm": 91.286376953125, "learning_rate": 4.355818513940608e-06, "loss": 0.8906, "step": 873 }, { "epoch": 0.7355354512939196, "grad_norm": 77.45460510253906, "learning_rate": 4.354321481204904e-06, "loss": 0.8006, "step": 874 }, { "epoch": 0.7363770250368189, "grad_norm": 83.27759552001953, "learning_rate": 4.352822968883896e-06, "loss": 0.6706, "step": 875 }, { "epoch": 0.7372185987797181, "grad_norm": 131.8611297607422, "learning_rate": 4.351322978173264e-06, "loss": 0.7503, "step": 876 }, { "epoch": 0.7380601725226172, "grad_norm": 66.10797119140625, "learning_rate": 4.3498215102698695e-06, "loss": 0.6593, "step": 877 }, { "epoch": 0.7389017462655165, "grad_norm": 81.42675018310547, "learning_rate": 4.348318566371749e-06, "loss": 0.8397, "step": 878 }, { "epoch": 0.7397433200084157, "grad_norm": 76.12470245361328, "learning_rate": 4.346814147678124e-06, "loss": 0.7452, "step": 879 }, { "epoch": 0.740584893751315, "grad_norm": 88.55060577392578, "learning_rate": 4.345308255389383e-06, "loss": 1.0045, "step": 880 }, { "epoch": 0.7414264674942141, "grad_norm": 69.48717498779297, "learning_rate": 4.343800890707099e-06, "loss": 0.6494, "step": 881 }, { "epoch": 0.7422680412371134, "grad_norm": 73.43086242675781, "learning_rate": 4.342292054834016e-06, "loss": 0.7483, "step": 882 }, { "epoch": 0.7431096149800126, "grad_norm": 102.50550079345703, "learning_rate": 4.340781748974051e-06, "loss": 0.7571, "step": 883 }, { "epoch": 0.7439511887229119, "grad_norm": 74.73184204101562, "learning_rate": 4.339269974332295e-06, "loss": 0.7497, "step": 884 }, { "epoch": 0.744792762465811, "grad_norm": 105.30278778076172, "learning_rate": 4.337756732115011e-06, "loss": 0.8387, "step": 885 }, { "epoch": 0.7456343362087103, "grad_norm": 95.76217651367188, "learning_rate": 4.336242023529635e-06, "loss": 0.913, "step": 886 }, { "epoch": 0.7464759099516095, "grad_norm": 81.93257904052734, "learning_rate": 4.334725849784771e-06, "loss": 0.8567, "step": 887 }, { "epoch": 0.7473174836945087, "grad_norm": 86.54633331298828, "learning_rate": 4.3332082120901895e-06, "loss": 0.9386, "step": 888 }, { "epoch": 0.748159057437408, "grad_norm": 77.6705093383789, "learning_rate": 4.331689111656834e-06, "loss": 0.8797, "step": 889 }, { "epoch": 0.7490006311803071, "grad_norm": 63.00545120239258, "learning_rate": 4.3301685496968114e-06, "loss": 0.7039, "step": 890 }, { "epoch": 0.7498422049232064, "grad_norm": 92.33924102783203, "learning_rate": 4.328646527423398e-06, "loss": 0.8826, "step": 891 }, { "epoch": 0.7506837786661056, "grad_norm": 74.36640167236328, "learning_rate": 4.3271230460510325e-06, "loss": 0.7637, "step": 892 }, { "epoch": 0.7515253524090049, "grad_norm": 428.94378662109375, "learning_rate": 4.325598106795319e-06, "loss": 0.8922, "step": 893 }, { "epoch": 0.752366926151904, "grad_norm": 82.86460876464844, "learning_rate": 4.324071710873025e-06, "loss": 0.8002, "step": 894 }, { "epoch": 0.7532084998948033, "grad_norm": 115.68354034423828, "learning_rate": 4.322543859502081e-06, "loss": 1.0063, "step": 895 }, { "epoch": 0.7540500736377025, "grad_norm": 83.29459381103516, "learning_rate": 4.3210145539015755e-06, "loss": 0.8483, "step": 896 }, { "epoch": 0.7548916473806018, "grad_norm": 73.96419525146484, "learning_rate": 4.319483795291762e-06, "loss": 0.8241, "step": 897 }, { "epoch": 0.7557332211235009, "grad_norm": 77.9550552368164, "learning_rate": 4.31795158489405e-06, "loss": 0.7826, "step": 898 }, { "epoch": 0.7565747948664001, "grad_norm": 103.94831085205078, "learning_rate": 4.316417923931008e-06, "loss": 0.8422, "step": 899 }, { "epoch": 0.7574163686092994, "grad_norm": 107.88838195800781, "learning_rate": 4.314882813626364e-06, "loss": 1.0209, "step": 900 }, { "epoch": 0.7582579423521986, "grad_norm": 76.37074279785156, "learning_rate": 4.313346255205001e-06, "loss": 0.8941, "step": 901 }, { "epoch": 0.7590995160950978, "grad_norm": 61.950862884521484, "learning_rate": 4.311808249892955e-06, "loss": 0.624, "step": 902 }, { "epoch": 0.759941089837997, "grad_norm": 89.68816375732422, "learning_rate": 4.310268798917422e-06, "loss": 1.0453, "step": 903 }, { "epoch": 0.7607826635808963, "grad_norm": 74.37384033203125, "learning_rate": 4.308727903506745e-06, "loss": 0.933, "step": 904 }, { "epoch": 0.7616242373237955, "grad_norm": 72.46370697021484, "learning_rate": 4.307185564890426e-06, "loss": 0.7905, "step": 905 }, { "epoch": 0.7624658110666948, "grad_norm": 66.4675064086914, "learning_rate": 4.305641784299113e-06, "loss": 0.7179, "step": 906 }, { "epoch": 0.7633073848095939, "grad_norm": 75.27861022949219, "learning_rate": 4.304096562964608e-06, "loss": 0.8811, "step": 907 }, { "epoch": 0.7641489585524932, "grad_norm": 61.52951431274414, "learning_rate": 4.302549902119863e-06, "loss": 0.6958, "step": 908 }, { "epoch": 0.7649905322953924, "grad_norm": 67.17115020751953, "learning_rate": 4.301001802998974e-06, "loss": 0.7002, "step": 909 }, { "epoch": 0.7658321060382917, "grad_norm": 98.0196304321289, "learning_rate": 4.299452266837191e-06, "loss": 0.9624, "step": 910 }, { "epoch": 0.7666736797811908, "grad_norm": 83.28306579589844, "learning_rate": 4.2979012948709055e-06, "loss": 0.9621, "step": 911 }, { "epoch": 0.76751525352409, "grad_norm": 76.3790283203125, "learning_rate": 4.296348888337657e-06, "loss": 0.6902, "step": 912 }, { "epoch": 0.7683568272669893, "grad_norm": 87.94258880615234, "learning_rate": 4.29479504847613e-06, "loss": 0.8355, "step": 913 }, { "epoch": 0.7691984010098885, "grad_norm": 92.80834197998047, "learning_rate": 4.2932397765261516e-06, "loss": 1.0073, "step": 914 }, { "epoch": 0.7700399747527877, "grad_norm": 71.12057495117188, "learning_rate": 4.29168307372869e-06, "loss": 0.7336, "step": 915 }, { "epoch": 0.7708815484956869, "grad_norm": 67.44596862792969, "learning_rate": 4.290124941325861e-06, "loss": 0.7769, "step": 916 }, { "epoch": 0.7717231222385862, "grad_norm": 71.54752349853516, "learning_rate": 4.288565380560913e-06, "loss": 0.726, "step": 917 }, { "epoch": 0.7725646959814854, "grad_norm": 148.19482421875, "learning_rate": 4.287004392678241e-06, "loss": 0.9742, "step": 918 }, { "epoch": 0.7734062697243846, "grad_norm": 61.75517272949219, "learning_rate": 4.2854419789233744e-06, "loss": 0.6872, "step": 919 }, { "epoch": 0.7742478434672838, "grad_norm": 74.51992797851562, "learning_rate": 4.2838781405429825e-06, "loss": 0.6292, "step": 920 }, { "epoch": 0.7750894172101831, "grad_norm": 87.5741958618164, "learning_rate": 4.28231287878487e-06, "loss": 0.8581, "step": 921 }, { "epoch": 0.7759309909530823, "grad_norm": 105.05841064453125, "learning_rate": 4.280746194897979e-06, "loss": 0.8532, "step": 922 }, { "epoch": 0.7767725646959814, "grad_norm": 80.18090057373047, "learning_rate": 4.279178090132386e-06, "loss": 0.7673, "step": 923 }, { "epoch": 0.7776141384388807, "grad_norm": 117.09395599365234, "learning_rate": 4.277608565739297e-06, "loss": 0.9147, "step": 924 }, { "epoch": 0.7784557121817799, "grad_norm": 98.15302276611328, "learning_rate": 4.276037622971059e-06, "loss": 0.8127, "step": 925 }, { "epoch": 0.7792972859246792, "grad_norm": 81.57997131347656, "learning_rate": 4.274465263081142e-06, "loss": 0.7557, "step": 926 }, { "epoch": 0.7801388596675783, "grad_norm": 110.0119400024414, "learning_rate": 4.272891487324152e-06, "loss": 0.7321, "step": 927 }, { "epoch": 0.7809804334104776, "grad_norm": 126.8502197265625, "learning_rate": 4.271316296955825e-06, "loss": 0.8107, "step": 928 }, { "epoch": 0.7818220071533768, "grad_norm": 72.49397277832031, "learning_rate": 4.269739693233024e-06, "loss": 0.6845, "step": 929 }, { "epoch": 0.7826635808962761, "grad_norm": 74.62886810302734, "learning_rate": 4.268161677413738e-06, "loss": 0.714, "step": 930 }, { "epoch": 0.7835051546391752, "grad_norm": 78.56126403808594, "learning_rate": 4.266582250757086e-06, "loss": 0.7087, "step": 931 }, { "epoch": 0.7843467283820745, "grad_norm": 107.03934478759766, "learning_rate": 4.265001414523312e-06, "loss": 0.8672, "step": 932 }, { "epoch": 0.7851883021249737, "grad_norm": 88.3293685913086, "learning_rate": 4.263419169973784e-06, "loss": 0.9362, "step": 933 }, { "epoch": 0.7860298758678729, "grad_norm": 87.77210235595703, "learning_rate": 4.261835518370992e-06, "loss": 0.7801, "step": 934 }, { "epoch": 0.7868714496107722, "grad_norm": 86.31080627441406, "learning_rate": 4.2602504609785526e-06, "loss": 0.9737, "step": 935 }, { "epoch": 0.7877130233536713, "grad_norm": 84.96116638183594, "learning_rate": 4.2586639990612015e-06, "loss": 0.7804, "step": 936 }, { "epoch": 0.7885545970965706, "grad_norm": 74.81805419921875, "learning_rate": 4.257076133884795e-06, "loss": 0.8313, "step": 937 }, { "epoch": 0.7893961708394698, "grad_norm": 98.99002838134766, "learning_rate": 4.255486866716311e-06, "loss": 0.8527, "step": 938 }, { "epoch": 0.7902377445823691, "grad_norm": 92.40936279296875, "learning_rate": 4.253896198823844e-06, "loss": 0.9288, "step": 939 }, { "epoch": 0.7910793183252682, "grad_norm": 99.4084243774414, "learning_rate": 4.252304131476607e-06, "loss": 0.9823, "step": 940 }, { "epoch": 0.7919208920681675, "grad_norm": 72.1796646118164, "learning_rate": 4.2507106659449295e-06, "loss": 0.7174, "step": 941 }, { "epoch": 0.7927624658110667, "grad_norm": 71.70301055908203, "learning_rate": 4.249115803500257e-06, "loss": 0.8033, "step": 942 }, { "epoch": 0.793604039553966, "grad_norm": 68.06544494628906, "learning_rate": 4.247519545415149e-06, "loss": 0.7506, "step": 943 }, { "epoch": 0.7944456132968651, "grad_norm": 62.31113052368164, "learning_rate": 4.245921892963279e-06, "loss": 0.6848, "step": 944 }, { "epoch": 0.7952871870397643, "grad_norm": 84.2646255493164, "learning_rate": 4.2443228474194335e-06, "loss": 0.9001, "step": 945 }, { "epoch": 0.7961287607826636, "grad_norm": 87.14163208007812, "learning_rate": 4.24272241005951e-06, "loss": 1.0412, "step": 946 }, { "epoch": 0.7969703345255628, "grad_norm": 89.12403869628906, "learning_rate": 4.241120582160516e-06, "loss": 0.8883, "step": 947 }, { "epoch": 0.797811908268462, "grad_norm": 79.68348693847656, "learning_rate": 4.239517365000569e-06, "loss": 0.9308, "step": 948 }, { "epoch": 0.7986534820113612, "grad_norm": 79.94398498535156, "learning_rate": 4.237912759858896e-06, "loss": 0.8429, "step": 949 }, { "epoch": 0.7994950557542605, "grad_norm": 73.92453002929688, "learning_rate": 4.236306768015831e-06, "loss": 0.8694, "step": 950 }, { "epoch": 0.8003366294971597, "grad_norm": 76.87979888916016, "learning_rate": 4.234699390752812e-06, "loss": 0.8229, "step": 951 }, { "epoch": 0.801178203240059, "grad_norm": 66.13666534423828, "learning_rate": 4.233090629352388e-06, "loss": 0.7658, "step": 952 }, { "epoch": 0.8020197769829581, "grad_norm": 77.92546844482422, "learning_rate": 4.231480485098205e-06, "loss": 0.7735, "step": 953 }, { "epoch": 0.8028613507258574, "grad_norm": 91.31938934326172, "learning_rate": 4.229868959275019e-06, "loss": 1.0128, "step": 954 }, { "epoch": 0.8037029244687566, "grad_norm": 103.90299224853516, "learning_rate": 4.228256053168685e-06, "loss": 1.1743, "step": 955 }, { "epoch": 0.8045444982116557, "grad_norm": 80.80589294433594, "learning_rate": 4.226641768066161e-06, "loss": 0.8783, "step": 956 }, { "epoch": 0.805386071954555, "grad_norm": 93.89417266845703, "learning_rate": 4.225026105255503e-06, "loss": 1.0368, "step": 957 }, { "epoch": 0.8062276456974542, "grad_norm": 72.91015625, "learning_rate": 4.223409066025867e-06, "loss": 0.8157, "step": 958 }, { "epoch": 0.8070692194403535, "grad_norm": 67.31309509277344, "learning_rate": 4.221790651667511e-06, "loss": 0.7382, "step": 959 }, { "epoch": 0.8079107931832527, "grad_norm": 80.1012191772461, "learning_rate": 4.220170863471785e-06, "loss": 0.9073, "step": 960 }, { "epoch": 0.8087523669261519, "grad_norm": 82.77555847167969, "learning_rate": 4.218549702731138e-06, "loss": 0.9337, "step": 961 }, { "epoch": 0.8095939406690511, "grad_norm": 84.62960815429688, "learning_rate": 4.216927170739113e-06, "loss": 0.8536, "step": 962 }, { "epoch": 0.8104355144119504, "grad_norm": 72.1750259399414, "learning_rate": 4.2153032687903475e-06, "loss": 0.8958, "step": 963 }, { "epoch": 0.8112770881548496, "grad_norm": 73.40558624267578, "learning_rate": 4.213677998180574e-06, "loss": 0.8339, "step": 964 }, { "epoch": 0.8121186618977488, "grad_norm": 99.02385711669922, "learning_rate": 4.212051360206614e-06, "loss": 0.7137, "step": 965 }, { "epoch": 0.812960235640648, "grad_norm": 116.62301635742188, "learning_rate": 4.21042335616638e-06, "loss": 1.3858, "step": 966 }, { "epoch": 0.8138018093835472, "grad_norm": 72.70298767089844, "learning_rate": 4.208793987358879e-06, "loss": 0.6905, "step": 967 }, { "epoch": 0.8146433831264465, "grad_norm": 78.67459106445312, "learning_rate": 4.207163255084202e-06, "loss": 0.9802, "step": 968 }, { "epoch": 0.8154849568693456, "grad_norm": 110.66291809082031, "learning_rate": 4.205531160643531e-06, "loss": 0.6545, "step": 969 }, { "epoch": 0.8163265306122449, "grad_norm": 71.75804138183594, "learning_rate": 4.203897705339132e-06, "loss": 0.9127, "step": 970 }, { "epoch": 0.8171681043551441, "grad_norm": 80.5187759399414, "learning_rate": 4.20226289047436e-06, "loss": 0.7607, "step": 971 }, { "epoch": 0.8180096780980434, "grad_norm": 75.0357666015625, "learning_rate": 4.200626717353653e-06, "loss": 0.7147, "step": 972 }, { "epoch": 0.8188512518409425, "grad_norm": 73.03073120117188, "learning_rate": 4.198989187282532e-06, "loss": 0.5297, "step": 973 }, { "epoch": 0.8196928255838418, "grad_norm": 84.6255111694336, "learning_rate": 4.197350301567603e-06, "loss": 0.8926, "step": 974 }, { "epoch": 0.820534399326741, "grad_norm": 66.8243637084961, "learning_rate": 4.195710061516552e-06, "loss": 0.8253, "step": 975 }, { "epoch": 0.8213759730696403, "grad_norm": 57.61470031738281, "learning_rate": 4.194068468438147e-06, "loss": 0.6581, "step": 976 }, { "epoch": 0.8222175468125394, "grad_norm": 66.89012908935547, "learning_rate": 4.192425523642234e-06, "loss": 0.81, "step": 977 }, { "epoch": 0.8230591205554386, "grad_norm": 76.0096435546875, "learning_rate": 4.1907812284397395e-06, "loss": 0.8723, "step": 978 }, { "epoch": 0.8239006942983379, "grad_norm": 80.26139831542969, "learning_rate": 4.189135584142664e-06, "loss": 0.8389, "step": 979 }, { "epoch": 0.8247422680412371, "grad_norm": 71.46955871582031, "learning_rate": 4.187488592064089e-06, "loss": 0.7342, "step": 980 }, { "epoch": 0.8255838417841364, "grad_norm": 68.54132843017578, "learning_rate": 4.185840253518168e-06, "loss": 0.9466, "step": 981 }, { "epoch": 0.8264254155270355, "grad_norm": 74.63121032714844, "learning_rate": 4.18419056982013e-06, "loss": 0.9002, "step": 982 }, { "epoch": 0.8272669892699348, "grad_norm": 68.22480010986328, "learning_rate": 4.182539542286279e-06, "loss": 0.7849, "step": 983 }, { "epoch": 0.828108563012834, "grad_norm": 67.76121520996094, "learning_rate": 4.1808871722339875e-06, "loss": 0.4417, "step": 984 }, { "epoch": 0.8289501367557333, "grad_norm": 80.49524688720703, "learning_rate": 4.179233460981702e-06, "loss": 0.9594, "step": 985 }, { "epoch": 0.8297917104986324, "grad_norm": 82.17792510986328, "learning_rate": 4.177578409848939e-06, "loss": 0.842, "step": 986 }, { "epoch": 0.8306332842415317, "grad_norm": 81.26715850830078, "learning_rate": 4.1759220201562835e-06, "loss": 0.9353, "step": 987 }, { "epoch": 0.8314748579844309, "grad_norm": 77.98991394042969, "learning_rate": 4.174264293225389e-06, "loss": 0.753, "step": 988 }, { "epoch": 0.83231643172733, "grad_norm": 83.93816375732422, "learning_rate": 4.172605230378975e-06, "loss": 0.9018, "step": 989 }, { "epoch": 0.8331580054702293, "grad_norm": 74.68817138671875, "learning_rate": 4.170944832940827e-06, "loss": 0.6308, "step": 990 }, { "epoch": 0.8339995792131285, "grad_norm": 78.59561157226562, "learning_rate": 4.169283102235797e-06, "loss": 0.867, "step": 991 }, { "epoch": 0.8348411529560278, "grad_norm": 65.0259780883789, "learning_rate": 4.167620039589799e-06, "loss": 0.8615, "step": 992 }, { "epoch": 0.835682726698927, "grad_norm": 79.45830535888672, "learning_rate": 4.1659556463298124e-06, "loss": 0.9266, "step": 993 }, { "epoch": 0.8365243004418262, "grad_norm": 81.86888122558594, "learning_rate": 4.164289923783874e-06, "loss": 0.8939, "step": 994 }, { "epoch": 0.8373658741847254, "grad_norm": 81.57584381103516, "learning_rate": 4.162622873281086e-06, "loss": 0.8272, "step": 995 }, { "epoch": 0.8382074479276247, "grad_norm": 68.12931823730469, "learning_rate": 4.160954496151607e-06, "loss": 0.778, "step": 996 }, { "epoch": 0.8390490216705239, "grad_norm": 77.3565673828125, "learning_rate": 4.159284793726655e-06, "loss": 0.695, "step": 997 }, { "epoch": 0.8398905954134231, "grad_norm": 72.69764709472656, "learning_rate": 4.157613767338505e-06, "loss": 0.8958, "step": 998 }, { "epoch": 0.8407321691563223, "grad_norm": 72.09540557861328, "learning_rate": 4.15594141832049e-06, "loss": 0.6611, "step": 999 }, { "epoch": 0.8415737428992216, "grad_norm": 66.83956909179688, "learning_rate": 4.154267748006996e-06, "loss": 0.7814, "step": 1000 }, { "epoch": 0.8424153166421208, "grad_norm": 69.9083251953125, "learning_rate": 4.152592757733467e-06, "loss": 0.7695, "step": 1001 }, { "epoch": 0.8432568903850199, "grad_norm": 81.96198272705078, "learning_rate": 4.150916448836395e-06, "loss": 0.8362, "step": 1002 }, { "epoch": 0.8440984641279192, "grad_norm": 60.79277420043945, "learning_rate": 4.149238822653329e-06, "loss": 0.6429, "step": 1003 }, { "epoch": 0.8449400378708184, "grad_norm": 73.13953399658203, "learning_rate": 4.147559880522866e-06, "loss": 0.8873, "step": 1004 }, { "epoch": 0.8457816116137177, "grad_norm": 75.53364562988281, "learning_rate": 4.145879623784653e-06, "loss": 0.7998, "step": 1005 }, { "epoch": 0.8466231853566168, "grad_norm": 65.25211334228516, "learning_rate": 4.144198053779389e-06, "loss": 0.839, "step": 1006 }, { "epoch": 0.8474647590995161, "grad_norm": 74.82688903808594, "learning_rate": 4.142515171848819e-06, "loss": 1.0005, "step": 1007 }, { "epoch": 0.8483063328424153, "grad_norm": 65.3867416381836, "learning_rate": 4.140830979335733e-06, "loss": 0.5801, "step": 1008 }, { "epoch": 0.8491479065853146, "grad_norm": 67.73045349121094, "learning_rate": 4.139145477583968e-06, "loss": 0.8043, "step": 1009 }, { "epoch": 0.8499894803282138, "grad_norm": 92.07035064697266, "learning_rate": 4.137458667938407e-06, "loss": 0.888, "step": 1010 }, { "epoch": 0.850831054071113, "grad_norm": 73.86541748046875, "learning_rate": 4.135770551744977e-06, "loss": 0.8436, "step": 1011 }, { "epoch": 0.8516726278140122, "grad_norm": 83.31813049316406, "learning_rate": 4.134081130350644e-06, "loss": 1.0081, "step": 1012 }, { "epoch": 0.8525142015569114, "grad_norm": 76.76337432861328, "learning_rate": 4.132390405103419e-06, "loss": 0.7978, "step": 1013 }, { "epoch": 0.8533557752998107, "grad_norm": 96.06390380859375, "learning_rate": 4.13069837735235e-06, "loss": 0.9046, "step": 1014 }, { "epoch": 0.8541973490427098, "grad_norm": 74.8203353881836, "learning_rate": 4.1290050484475285e-06, "loss": 0.9066, "step": 1015 }, { "epoch": 0.8550389227856091, "grad_norm": 76.6991958618164, "learning_rate": 4.127310419740081e-06, "loss": 0.8842, "step": 1016 }, { "epoch": 0.8558804965285083, "grad_norm": 59.34732437133789, "learning_rate": 4.12561449258217e-06, "loss": 0.7031, "step": 1017 }, { "epoch": 0.8567220702714076, "grad_norm": 71.81965637207031, "learning_rate": 4.123917268326999e-06, "loss": 0.8189, "step": 1018 }, { "epoch": 0.8575636440143067, "grad_norm": 58.81732940673828, "learning_rate": 4.122218748328802e-06, "loss": 0.8027, "step": 1019 }, { "epoch": 0.858405217757206, "grad_norm": 75.25838470458984, "learning_rate": 4.12051893394285e-06, "loss": 0.8217, "step": 1020 }, { "epoch": 0.8592467915001052, "grad_norm": 71.95877075195312, "learning_rate": 4.118817826525443e-06, "loss": 0.6275, "step": 1021 }, { "epoch": 0.8600883652430045, "grad_norm": 82.2073974609375, "learning_rate": 4.1171154274339164e-06, "loss": 0.9913, "step": 1022 }, { "epoch": 0.8609299389859036, "grad_norm": 80.4325942993164, "learning_rate": 4.1154117380266346e-06, "loss": 0.6675, "step": 1023 }, { "epoch": 0.8617715127288028, "grad_norm": 79.63078308105469, "learning_rate": 4.113706759662992e-06, "loss": 0.7476, "step": 1024 }, { "epoch": 0.8626130864717021, "grad_norm": 124.51466369628906, "learning_rate": 4.112000493703411e-06, "loss": 0.7556, "step": 1025 }, { "epoch": 0.8634546602146013, "grad_norm": 69.8863525390625, "learning_rate": 4.11029294150934e-06, "loss": 0.8307, "step": 1026 }, { "epoch": 0.8642962339575005, "grad_norm": 151.2084197998047, "learning_rate": 4.108584104443259e-06, "loss": 0.7547, "step": 1027 }, { "epoch": 0.8651378077003997, "grad_norm": 58.940773010253906, "learning_rate": 4.106873983868667e-06, "loss": 0.8035, "step": 1028 }, { "epoch": 0.865979381443299, "grad_norm": 68.62104034423828, "learning_rate": 4.10516258115009e-06, "loss": 0.7035, "step": 1029 }, { "epoch": 0.8668209551861982, "grad_norm": 99.0032730102539, "learning_rate": 4.1034498976530755e-06, "loss": 1.0287, "step": 1030 }, { "epoch": 0.8676625289290975, "grad_norm": 89.32034301757812, "learning_rate": 4.101735934744197e-06, "loss": 0.9512, "step": 1031 }, { "epoch": 0.8685041026719966, "grad_norm": 59.35560989379883, "learning_rate": 4.100020693791043e-06, "loss": 0.7646, "step": 1032 }, { "epoch": 0.8693456764148959, "grad_norm": 59.45350646972656, "learning_rate": 4.098304176162228e-06, "loss": 0.697, "step": 1033 }, { "epoch": 0.8701872501577951, "grad_norm": 86.22278594970703, "learning_rate": 4.09658638322738e-06, "loss": 1.0046, "step": 1034 }, { "epoch": 0.8710288239006942, "grad_norm": 69.21746063232422, "learning_rate": 4.094867316357146e-06, "loss": 0.6598, "step": 1035 }, { "epoch": 0.8718703976435935, "grad_norm": 71.82403564453125, "learning_rate": 4.093146976923191e-06, "loss": 0.7445, "step": 1036 }, { "epoch": 0.8727119713864927, "grad_norm": 72.35151672363281, "learning_rate": 4.091425366298195e-06, "loss": 0.82, "step": 1037 }, { "epoch": 0.873553545129392, "grad_norm": 79.09820556640625, "learning_rate": 4.08970248585585e-06, "loss": 0.7896, "step": 1038 }, { "epoch": 0.8743951188722912, "grad_norm": 70.99044036865234, "learning_rate": 4.087978336970864e-06, "loss": 0.7507, "step": 1039 }, { "epoch": 0.8752366926151904, "grad_norm": 64.83211517333984, "learning_rate": 4.0862529210189566e-06, "loss": 0.8278, "step": 1040 }, { "epoch": 0.8760782663580896, "grad_norm": 88.5916519165039, "learning_rate": 4.084526239376857e-06, "loss": 0.8102, "step": 1041 }, { "epoch": 0.8769198401009889, "grad_norm": 100.17524719238281, "learning_rate": 4.082798293422305e-06, "loss": 0.9703, "step": 1042 }, { "epoch": 0.8777614138438881, "grad_norm": 80.35630798339844, "learning_rate": 4.08106908453405e-06, "loss": 0.8869, "step": 1043 }, { "epoch": 0.8786029875867873, "grad_norm": 72.74226379394531, "learning_rate": 4.079338614091848e-06, "loss": 0.73, "step": 1044 }, { "epoch": 0.8794445613296865, "grad_norm": 87.65748596191406, "learning_rate": 4.077606883476462e-06, "loss": 0.9816, "step": 1045 }, { "epoch": 0.8802861350725857, "grad_norm": 64.14987182617188, "learning_rate": 4.075873894069661e-06, "loss": 0.6598, "step": 1046 }, { "epoch": 0.881127708815485, "grad_norm": 62.39348220825195, "learning_rate": 4.074139647254218e-06, "loss": 0.6845, "step": 1047 }, { "epoch": 0.8819692825583841, "grad_norm": 57.6712532043457, "learning_rate": 4.072404144413909e-06, "loss": 0.6839, "step": 1048 }, { "epoch": 0.8828108563012834, "grad_norm": 65.15045928955078, "learning_rate": 4.070667386933512e-06, "loss": 0.874, "step": 1049 }, { "epoch": 0.8836524300441826, "grad_norm": 84.79566192626953, "learning_rate": 4.068929376198808e-06, "loss": 0.8565, "step": 1050 }, { "epoch": 0.8844940037870819, "grad_norm": 85.54796600341797, "learning_rate": 4.067190113596576e-06, "loss": 1.0899, "step": 1051 }, { "epoch": 0.885335577529981, "grad_norm": 66.54473876953125, "learning_rate": 4.065449600514594e-06, "loss": 0.7514, "step": 1052 }, { "epoch": 0.8861771512728803, "grad_norm": 70.7523422241211, "learning_rate": 4.0637078383416385e-06, "loss": 0.7151, "step": 1053 }, { "epoch": 0.8870187250157795, "grad_norm": 78.33436584472656, "learning_rate": 4.061964828467482e-06, "loss": 0.8094, "step": 1054 }, { "epoch": 0.8878602987586788, "grad_norm": 63.319393157958984, "learning_rate": 4.060220572282892e-06, "loss": 0.7593, "step": 1055 }, { "epoch": 0.888701872501578, "grad_norm": 77.5903549194336, "learning_rate": 4.058475071179635e-06, "loss": 0.7143, "step": 1056 }, { "epoch": 0.8895434462444771, "grad_norm": 77.97570037841797, "learning_rate": 4.056728326550463e-06, "loss": 0.811, "step": 1057 }, { "epoch": 0.8903850199873764, "grad_norm": 80.57714080810547, "learning_rate": 4.0549803397891265e-06, "loss": 0.8381, "step": 1058 }, { "epoch": 0.8912265937302756, "grad_norm": 62.823917388916016, "learning_rate": 4.053231112290365e-06, "loss": 0.7561, "step": 1059 }, { "epoch": 0.8920681674731749, "grad_norm": 68.00716400146484, "learning_rate": 4.051480645449909e-06, "loss": 0.7462, "step": 1060 }, { "epoch": 0.892909741216074, "grad_norm": 60.848243713378906, "learning_rate": 4.049728940664474e-06, "loss": 0.6313, "step": 1061 }, { "epoch": 0.8937513149589733, "grad_norm": 69.16812896728516, "learning_rate": 4.04797599933177e-06, "loss": 0.8213, "step": 1062 }, { "epoch": 0.8945928887018725, "grad_norm": 88.9859848022461, "learning_rate": 4.046221822850487e-06, "loss": 0.9066, "step": 1063 }, { "epoch": 0.8954344624447718, "grad_norm": 135.78160095214844, "learning_rate": 4.044466412620305e-06, "loss": 0.7938, "step": 1064 }, { "epoch": 0.8962760361876709, "grad_norm": 89.73754119873047, "learning_rate": 4.042709770041885e-06, "loss": 0.926, "step": 1065 }, { "epoch": 0.8971176099305702, "grad_norm": 68.3105239868164, "learning_rate": 4.040951896516875e-06, "loss": 0.671, "step": 1066 }, { "epoch": 0.8979591836734694, "grad_norm": 57.36161422729492, "learning_rate": 4.039192793447902e-06, "loss": 0.6333, "step": 1067 }, { "epoch": 0.8988007574163686, "grad_norm": 91.2534408569336, "learning_rate": 4.037432462238578e-06, "loss": 0.9629, "step": 1068 }, { "epoch": 0.8996423311592678, "grad_norm": 97.27592468261719, "learning_rate": 4.035670904293488e-06, "loss": 1.0401, "step": 1069 }, { "epoch": 0.900483904902167, "grad_norm": 96.43426513671875, "learning_rate": 4.033908121018204e-06, "loss": 0.9545, "step": 1070 }, { "epoch": 0.9013254786450663, "grad_norm": 66.45618438720703, "learning_rate": 4.03214411381927e-06, "loss": 0.7543, "step": 1071 }, { "epoch": 0.9021670523879655, "grad_norm": 107.90736389160156, "learning_rate": 4.030378884104209e-06, "loss": 0.7016, "step": 1072 }, { "epoch": 0.9030086261308647, "grad_norm": 77.58516693115234, "learning_rate": 4.0286124332815186e-06, "loss": 0.8714, "step": 1073 }, { "epoch": 0.9038501998737639, "grad_norm": 65.78289794921875, "learning_rate": 4.0268447627606725e-06, "loss": 0.5679, "step": 1074 }, { "epoch": 0.9046917736166632, "grad_norm": 71.64932250976562, "learning_rate": 4.025075873952113e-06, "loss": 0.8684, "step": 1075 }, { "epoch": 0.9055333473595624, "grad_norm": 69.41444396972656, "learning_rate": 4.023305768267261e-06, "loss": 0.7746, "step": 1076 }, { "epoch": 0.9063749211024617, "grad_norm": 80.02897644042969, "learning_rate": 4.021534447118502e-06, "loss": 0.8513, "step": 1077 }, { "epoch": 0.9072164948453608, "grad_norm": 69.74525451660156, "learning_rate": 4.019761911919198e-06, "loss": 0.823, "step": 1078 }, { "epoch": 0.90805806858826, "grad_norm": 57.38936996459961, "learning_rate": 4.017988164083673e-06, "loss": 0.6093, "step": 1079 }, { "epoch": 0.9088996423311593, "grad_norm": 68.37493133544922, "learning_rate": 4.016213205027222e-06, "loss": 0.7365, "step": 1080 }, { "epoch": 0.9097412160740584, "grad_norm": 61.99602127075195, "learning_rate": 4.0144370361661075e-06, "loss": 0.7312, "step": 1081 }, { "epoch": 0.9105827898169577, "grad_norm": 73.24932861328125, "learning_rate": 4.012659658917555e-06, "loss": 0.7137, "step": 1082 }, { "epoch": 0.9114243635598569, "grad_norm": 73.64488983154297, "learning_rate": 4.010881074699755e-06, "loss": 0.8696, "step": 1083 }, { "epoch": 0.9122659373027562, "grad_norm": 82.7219467163086, "learning_rate": 4.009101284931861e-06, "loss": 0.8684, "step": 1084 }, { "epoch": 0.9131075110456554, "grad_norm": 86.55023193359375, "learning_rate": 4.0073202910339895e-06, "loss": 0.9902, "step": 1085 }, { "epoch": 0.9139490847885546, "grad_norm": 61.76776123046875, "learning_rate": 4.005538094427216e-06, "loss": 0.767, "step": 1086 }, { "epoch": 0.9147906585314538, "grad_norm": 70.59973907470703, "learning_rate": 4.003754696533576e-06, "loss": 0.8766, "step": 1087 }, { "epoch": 0.9156322322743531, "grad_norm": 69.00643920898438, "learning_rate": 4.001970098776066e-06, "loss": 0.9194, "step": 1088 }, { "epoch": 0.9164738060172523, "grad_norm": 88.84870147705078, "learning_rate": 4.000184302578636e-06, "loss": 1.0529, "step": 1089 }, { "epoch": 0.9173153797601514, "grad_norm": 88.0756607055664, "learning_rate": 3.998397309366194e-06, "loss": 0.891, "step": 1090 }, { "epoch": 0.9181569535030507, "grad_norm": 79.1492691040039, "learning_rate": 3.996609120564605e-06, "loss": 0.7358, "step": 1091 }, { "epoch": 0.9189985272459499, "grad_norm": 76.13194274902344, "learning_rate": 3.9948197376006855e-06, "loss": 0.8981, "step": 1092 }, { "epoch": 0.9198401009888492, "grad_norm": 69.00740051269531, "learning_rate": 3.993029161902205e-06, "loss": 0.8097, "step": 1093 }, { "epoch": 0.9206816747317483, "grad_norm": 75.37550354003906, "learning_rate": 3.991237394897886e-06, "loss": 0.9167, "step": 1094 }, { "epoch": 0.9215232484746476, "grad_norm": 86.60401916503906, "learning_rate": 3.9894444380174005e-06, "loss": 1.0947, "step": 1095 }, { "epoch": 0.9223648222175468, "grad_norm": 68.166015625, "learning_rate": 3.987650292691371e-06, "loss": 0.7399, "step": 1096 }, { "epoch": 0.9232063959604461, "grad_norm": 88.67655181884766, "learning_rate": 3.985854960351365e-06, "loss": 0.9703, "step": 1097 }, { "epoch": 0.9240479697033452, "grad_norm": 67.01685333251953, "learning_rate": 3.9840584424299034e-06, "loss": 0.7215, "step": 1098 }, { "epoch": 0.9248895434462445, "grad_norm": 79.09455871582031, "learning_rate": 3.982260740360447e-06, "loss": 0.8123, "step": 1099 }, { "epoch": 0.9257311171891437, "grad_norm": 74.57783508300781, "learning_rate": 3.980461855577403e-06, "loss": 0.9284, "step": 1100 }, { "epoch": 0.926572690932043, "grad_norm": 78.28739929199219, "learning_rate": 3.978661789516126e-06, "loss": 0.9738, "step": 1101 }, { "epoch": 0.9274142646749421, "grad_norm": 69.92387390136719, "learning_rate": 3.976860543612907e-06, "loss": 0.7838, "step": 1102 }, { "epoch": 0.9282558384178413, "grad_norm": 69.45232391357422, "learning_rate": 3.975058119304984e-06, "loss": 0.8191, "step": 1103 }, { "epoch": 0.9290974121607406, "grad_norm": 85.00932312011719, "learning_rate": 3.973254518030532e-06, "loss": 0.8866, "step": 1104 }, { "epoch": 0.9299389859036398, "grad_norm": 68.20960235595703, "learning_rate": 3.971449741228666e-06, "loss": 0.8838, "step": 1105 }, { "epoch": 0.930780559646539, "grad_norm": 68.52924346923828, "learning_rate": 3.969643790339441e-06, "loss": 0.7283, "step": 1106 }, { "epoch": 0.9316221333894382, "grad_norm": 85.05404663085938, "learning_rate": 3.967836666803844e-06, "loss": 0.8443, "step": 1107 }, { "epoch": 0.9324637071323375, "grad_norm": 70.1175765991211, "learning_rate": 3.966028372063801e-06, "loss": 0.7288, "step": 1108 }, { "epoch": 0.9333052808752367, "grad_norm": 72.65882873535156, "learning_rate": 3.9642189075621726e-06, "loss": 0.8062, "step": 1109 }, { "epoch": 0.934146854618136, "grad_norm": 71.01957702636719, "learning_rate": 3.962408274742753e-06, "loss": 0.9103, "step": 1110 }, { "epoch": 0.9349884283610351, "grad_norm": 86.82347106933594, "learning_rate": 3.960596475050266e-06, "loss": 0.6355, "step": 1111 }, { "epoch": 0.9358300021039344, "grad_norm": 93.42919158935547, "learning_rate": 3.95878350993037e-06, "loss": 0.7063, "step": 1112 }, { "epoch": 0.9366715758468336, "grad_norm": 92.69883728027344, "learning_rate": 3.956969380829651e-06, "loss": 0.8995, "step": 1113 }, { "epoch": 0.9375131495897328, "grad_norm": 63.439788818359375, "learning_rate": 3.955154089195624e-06, "loss": 0.7768, "step": 1114 }, { "epoch": 0.938354723332632, "grad_norm": 79.63054656982422, "learning_rate": 3.953337636476731e-06, "loss": 0.8086, "step": 1115 }, { "epoch": 0.9391962970755312, "grad_norm": 79.53690338134766, "learning_rate": 3.951520024122344e-06, "loss": 0.8836, "step": 1116 }, { "epoch": 0.9400378708184305, "grad_norm": 63.9859619140625, "learning_rate": 3.949701253582755e-06, "loss": 0.7912, "step": 1117 }, { "epoch": 0.9408794445613297, "grad_norm": 66.5926742553711, "learning_rate": 3.947881326309184e-06, "loss": 0.8147, "step": 1118 }, { "epoch": 0.9417210183042289, "grad_norm": 79.54521179199219, "learning_rate": 3.946060243753773e-06, "loss": 0.9673, "step": 1119 }, { "epoch": 0.9425625920471281, "grad_norm": 86.67349243164062, "learning_rate": 3.944238007369585e-06, "loss": 0.8635, "step": 1120 }, { "epoch": 0.9434041657900274, "grad_norm": 57.8299674987793, "learning_rate": 3.942414618610604e-06, "loss": 0.5855, "step": 1121 }, { "epoch": 0.9442457395329266, "grad_norm": 81.97209167480469, "learning_rate": 3.940590078931734e-06, "loss": 0.7143, "step": 1122 }, { "epoch": 0.9450873132758258, "grad_norm": 53.280616760253906, "learning_rate": 3.938764389788798e-06, "loss": 0.6815, "step": 1123 }, { "epoch": 0.945928887018725, "grad_norm": 90.65757751464844, "learning_rate": 3.936937552638534e-06, "loss": 1.0972, "step": 1124 }, { "epoch": 0.9467704607616242, "grad_norm": 97.28402709960938, "learning_rate": 3.935109568938598e-06, "loss": 0.7802, "step": 1125 }, { "epoch": 0.9476120345045235, "grad_norm": 62.99158477783203, "learning_rate": 3.9332804401475585e-06, "loss": 0.6851, "step": 1126 }, { "epoch": 0.9484536082474226, "grad_norm": 75.19110870361328, "learning_rate": 3.9314501677249015e-06, "loss": 0.7416, "step": 1127 }, { "epoch": 0.9492951819903219, "grad_norm": 86.4506607055664, "learning_rate": 3.929618753131023e-06, "loss": 0.9711, "step": 1128 }, { "epoch": 0.9501367557332211, "grad_norm": 68.20625305175781, "learning_rate": 3.927786197827228e-06, "loss": 0.7296, "step": 1129 }, { "epoch": 0.9509783294761204, "grad_norm": 79.4578857421875, "learning_rate": 3.9259525032757375e-06, "loss": 0.9495, "step": 1130 }, { "epoch": 0.9518199032190195, "grad_norm": 59.97657775878906, "learning_rate": 3.924117670939676e-06, "loss": 0.5803, "step": 1131 }, { "epoch": 0.9526614769619188, "grad_norm": 69.94535064697266, "learning_rate": 3.92228170228308e-06, "loss": 0.8632, "step": 1132 }, { "epoch": 0.953503050704818, "grad_norm": 70.57887268066406, "learning_rate": 3.920444598770889e-06, "loss": 0.8572, "step": 1133 }, { "epoch": 0.9543446244477173, "grad_norm": 72.97533416748047, "learning_rate": 3.918606361868952e-06, "loss": 0.7592, "step": 1134 }, { "epoch": 0.9551861981906165, "grad_norm": 76.23296356201172, "learning_rate": 3.916766993044018e-06, "loss": 0.774, "step": 1135 }, { "epoch": 0.9560277719335156, "grad_norm": 69.50141143798828, "learning_rate": 3.914926493763743e-06, "loss": 0.8404, "step": 1136 }, { "epoch": 0.9568693456764149, "grad_norm": 61.13656997680664, "learning_rate": 3.9130848654966835e-06, "loss": 0.8523, "step": 1137 }, { "epoch": 0.9577109194193141, "grad_norm": 76.27356719970703, "learning_rate": 3.911242109712296e-06, "loss": 0.6067, "step": 1138 }, { "epoch": 0.9585524931622134, "grad_norm": 70.06895446777344, "learning_rate": 3.909398227880937e-06, "loss": 0.5791, "step": 1139 }, { "epoch": 0.9593940669051125, "grad_norm": 58.804359436035156, "learning_rate": 3.907553221473863e-06, "loss": 0.7047, "step": 1140 }, { "epoch": 0.9602356406480118, "grad_norm": 91.75293731689453, "learning_rate": 3.9057070919632265e-06, "loss": 1.0091, "step": 1141 }, { "epoch": 0.961077214390911, "grad_norm": 69.82617950439453, "learning_rate": 3.903859840822076e-06, "loss": 0.8492, "step": 1142 }, { "epoch": 0.9619187881338103, "grad_norm": 50.22442626953125, "learning_rate": 3.902011469524355e-06, "loss": 0.6491, "step": 1143 }, { "epoch": 0.9627603618767094, "grad_norm": 74.7395248413086, "learning_rate": 3.900161979544901e-06, "loss": 1.0323, "step": 1144 }, { "epoch": 0.9636019356196087, "grad_norm": 70.21725463867188, "learning_rate": 3.898311372359445e-06, "loss": 0.6371, "step": 1145 }, { "epoch": 0.9644435093625079, "grad_norm": 79.02413177490234, "learning_rate": 3.896459649444607e-06, "loss": 0.8239, "step": 1146 }, { "epoch": 0.9652850831054071, "grad_norm": 83.71379852294922, "learning_rate": 3.894606812277901e-06, "loss": 0.9468, "step": 1147 }, { "epoch": 0.9661266568483063, "grad_norm": 64.89826202392578, "learning_rate": 3.892752862337728e-06, "loss": 0.7002, "step": 1148 }, { "epoch": 0.9669682305912055, "grad_norm": 66.54814147949219, "learning_rate": 3.890897801103375e-06, "loss": 0.649, "step": 1149 }, { "epoch": 0.9678098043341048, "grad_norm": 101.90099334716797, "learning_rate": 3.889041630055017e-06, "loss": 0.7801, "step": 1150 }, { "epoch": 0.968651378077004, "grad_norm": 63.44615936279297, "learning_rate": 3.887184350673718e-06, "loss": 0.6731, "step": 1151 }, { "epoch": 0.9694929518199032, "grad_norm": 58.71804428100586, "learning_rate": 3.885325964441422e-06, "loss": 0.7604, "step": 1152 }, { "epoch": 0.9703345255628024, "grad_norm": 61.444766998291016, "learning_rate": 3.883466472840958e-06, "loss": 0.7737, "step": 1153 }, { "epoch": 0.9711760993057017, "grad_norm": 74.5656967163086, "learning_rate": 3.881605877356035e-06, "loss": 0.8615, "step": 1154 }, { "epoch": 0.9720176730486009, "grad_norm": 102.34452056884766, "learning_rate": 3.8797441794712465e-06, "loss": 1.0315, "step": 1155 }, { "epoch": 0.9728592467915002, "grad_norm": 80.07135772705078, "learning_rate": 3.877881380672063e-06, "loss": 0.8032, "step": 1156 }, { "epoch": 0.9737008205343993, "grad_norm": 71.28023529052734, "learning_rate": 3.876017482444832e-06, "loss": 0.8414, "step": 1157 }, { "epoch": 0.9745423942772985, "grad_norm": 69.85945129394531, "learning_rate": 3.874152486276783e-06, "loss": 0.8079, "step": 1158 }, { "epoch": 0.9753839680201978, "grad_norm": 66.5711441040039, "learning_rate": 3.872286393656017e-06, "loss": 0.722, "step": 1159 }, { "epoch": 0.976225541763097, "grad_norm": 177.06283569335938, "learning_rate": 3.870419206071513e-06, "loss": 0.766, "step": 1160 }, { "epoch": 0.9770671155059962, "grad_norm": 92.01974487304688, "learning_rate": 3.86855092501312e-06, "loss": 1.0476, "step": 1161 }, { "epoch": 0.9779086892488954, "grad_norm": 74.33403015136719, "learning_rate": 3.866681551971565e-06, "loss": 0.9104, "step": 1162 }, { "epoch": 0.9787502629917947, "grad_norm": 66.5916748046875, "learning_rate": 3.864811088438441e-06, "loss": 0.7797, "step": 1163 }, { "epoch": 0.9795918367346939, "grad_norm": 54.91474533081055, "learning_rate": 3.862939535906213e-06, "loss": 0.7155, "step": 1164 }, { "epoch": 0.9804334104775931, "grad_norm": 61.410614013671875, "learning_rate": 3.8610668958682165e-06, "loss": 0.7638, "step": 1165 }, { "epoch": 0.9812749842204923, "grad_norm": 71.29328155517578, "learning_rate": 3.859193169818653e-06, "loss": 0.7566, "step": 1166 }, { "epoch": 0.9821165579633916, "grad_norm": 68.58023834228516, "learning_rate": 3.85731835925259e-06, "loss": 0.9146, "step": 1167 }, { "epoch": 0.9829581317062908, "grad_norm": 63.331214904785156, "learning_rate": 3.8554424656659625e-06, "loss": 0.8314, "step": 1168 }, { "epoch": 0.9837997054491899, "grad_norm": 78.91651916503906, "learning_rate": 3.853565490555568e-06, "loss": 0.8704, "step": 1169 }, { "epoch": 0.9846412791920892, "grad_norm": 67.55413818359375, "learning_rate": 3.851687435419067e-06, "loss": 0.7655, "step": 1170 }, { "epoch": 0.9854828529349884, "grad_norm": 69.61270904541016, "learning_rate": 3.849808301754983e-06, "loss": 0.8157, "step": 1171 }, { "epoch": 0.9863244266778877, "grad_norm": 62.39061737060547, "learning_rate": 3.847928091062698e-06, "loss": 0.8609, "step": 1172 }, { "epoch": 0.9871660004207868, "grad_norm": 73.21965026855469, "learning_rate": 3.846046804842456e-06, "loss": 0.6772, "step": 1173 }, { "epoch": 0.9880075741636861, "grad_norm": 52.76114273071289, "learning_rate": 3.844164444595356e-06, "loss": 0.6729, "step": 1174 }, { "epoch": 0.9888491479065853, "grad_norm": 65.8831558227539, "learning_rate": 3.842281011823359e-06, "loss": 0.6839, "step": 1175 }, { "epoch": 0.9896907216494846, "grad_norm": 63.77772903442383, "learning_rate": 3.840396508029275e-06, "loss": 0.8853, "step": 1176 }, { "epoch": 0.9905322953923837, "grad_norm": 70.06664276123047, "learning_rate": 3.8385109347167735e-06, "loss": 0.6873, "step": 1177 }, { "epoch": 0.991373869135283, "grad_norm": 81.78180694580078, "learning_rate": 3.836624293390378e-06, "loss": 0.9209, "step": 1178 }, { "epoch": 0.9922154428781822, "grad_norm": 66.87544250488281, "learning_rate": 3.834736585555458e-06, "loss": 0.6199, "step": 1179 }, { "epoch": 0.9930570166210814, "grad_norm": 64.75494384765625, "learning_rate": 3.832847812718242e-06, "loss": 0.9293, "step": 1180 }, { "epoch": 0.9938985903639806, "grad_norm": 67.62046813964844, "learning_rate": 3.830957976385803e-06, "loss": 0.8471, "step": 1181 }, { "epoch": 0.9947401641068798, "grad_norm": 72.29881286621094, "learning_rate": 3.829067078066063e-06, "loss": 0.8786, "step": 1182 }, { "epoch": 0.9955817378497791, "grad_norm": 57.97904968261719, "learning_rate": 3.827175119267794e-06, "loss": 0.716, "step": 1183 }, { "epoch": 0.9964233115926783, "grad_norm": 57.827091217041016, "learning_rate": 3.825282101500612e-06, "loss": 0.7305, "step": 1184 }, { "epoch": 0.9972648853355776, "grad_norm": 70.0528335571289, "learning_rate": 3.823388026274979e-06, "loss": 0.7974, "step": 1185 }, { "epoch": 0.9981064590784767, "grad_norm": 73.72193908691406, "learning_rate": 3.821492895102199e-06, "loss": 0.8964, "step": 1186 }, { "epoch": 0.998948032821376, "grad_norm": 67.32428741455078, "learning_rate": 3.819596709494421e-06, "loss": 0.7493, "step": 1187 }, { "epoch": 0.9997896065642752, "grad_norm": 64.63594818115234, "learning_rate": 3.817699470964634e-06, "loss": 0.8268, "step": 1188 }, { "epoch": 1.0008415737428993, "grad_norm": 118.80487823486328, "learning_rate": 3.815801181026666e-06, "loss": 1.8068, "step": 1189 }, { "epoch": 1.0016831474857983, "grad_norm": 62.071449279785156, "learning_rate": 3.813901841195188e-06, "loss": 0.7766, "step": 1190 }, { "epoch": 1.0025247212286976, "grad_norm": 71.35749053955078, "learning_rate": 3.812001452985704e-06, "loss": 0.7491, "step": 1191 }, { "epoch": 1.003366294971597, "grad_norm": 65.00326538085938, "learning_rate": 3.8101000179145564e-06, "loss": 0.7828, "step": 1192 }, { "epoch": 1.0042078687144962, "grad_norm": 65.48424530029297, "learning_rate": 3.8081975374989237e-06, "loss": 0.7553, "step": 1193 }, { "epoch": 1.0050494424573952, "grad_norm": 71.35240936279297, "learning_rate": 3.8062940132568183e-06, "loss": 0.8126, "step": 1194 }, { "epoch": 1.0058910162002945, "grad_norm": 75.91024780273438, "learning_rate": 3.804389446707083e-06, "loss": 0.9366, "step": 1195 }, { "epoch": 1.0067325899431938, "grad_norm": 56.03441619873047, "learning_rate": 3.8024838393693965e-06, "loss": 0.6938, "step": 1196 }, { "epoch": 1.007574163686093, "grad_norm": 75.76158142089844, "learning_rate": 3.8005771927642655e-06, "loss": 0.7218, "step": 1197 }, { "epoch": 1.0084157374289922, "grad_norm": 72.80091857910156, "learning_rate": 3.7986695084130244e-06, "loss": 0.7922, "step": 1198 }, { "epoch": 1.0092573111718914, "grad_norm": 67.57899475097656, "learning_rate": 3.7967607878378375e-06, "loss": 0.8067, "step": 1199 }, { "epoch": 1.0100988849147907, "grad_norm": 64.86554718017578, "learning_rate": 3.794851032561698e-06, "loss": 0.7498, "step": 1200 }, { "epoch": 1.0109404586576898, "grad_norm": 63.07143783569336, "learning_rate": 3.7929402441084207e-06, "loss": 0.8218, "step": 1201 }, { "epoch": 1.011782032400589, "grad_norm": 70.4422378540039, "learning_rate": 3.7910284240026474e-06, "loss": 0.8098, "step": 1202 }, { "epoch": 1.0126236061434883, "grad_norm": 73.65013885498047, "learning_rate": 3.7891155737698415e-06, "loss": 0.8102, "step": 1203 }, { "epoch": 1.0134651798863876, "grad_norm": 61.945465087890625, "learning_rate": 3.787201694936289e-06, "loss": 0.8075, "step": 1204 }, { "epoch": 1.0143067536292867, "grad_norm": 67.951904296875, "learning_rate": 3.785286789029098e-06, "loss": 0.6416, "step": 1205 }, { "epoch": 1.015148327372186, "grad_norm": 68.7704086303711, "learning_rate": 3.783370857576194e-06, "loss": 0.6408, "step": 1206 }, { "epoch": 1.0159899011150852, "grad_norm": 69.95020294189453, "learning_rate": 3.7814539021063213e-06, "loss": 0.731, "step": 1207 }, { "epoch": 1.0168314748579845, "grad_norm": 75.72080993652344, "learning_rate": 3.7795359241490425e-06, "loss": 0.6204, "step": 1208 }, { "epoch": 1.0176730486008836, "grad_norm": 69.6406478881836, "learning_rate": 3.7776169252347344e-06, "loss": 0.7158, "step": 1209 }, { "epoch": 1.0185146223437829, "grad_norm": 74.77608489990234, "learning_rate": 3.7756969068945893e-06, "loss": 0.7458, "step": 1210 }, { "epoch": 1.0193561960866822, "grad_norm": 97.5343017578125, "learning_rate": 3.773775870660613e-06, "loss": 0.9934, "step": 1211 }, { "epoch": 1.0201977698295812, "grad_norm": 62.32524490356445, "learning_rate": 3.7718538180656237e-06, "loss": 0.6901, "step": 1212 }, { "epoch": 1.0210393435724805, "grad_norm": 90.14856719970703, "learning_rate": 3.7699307506432496e-06, "loss": 0.9022, "step": 1213 }, { "epoch": 1.0218809173153798, "grad_norm": 86.78469848632812, "learning_rate": 3.7680066699279294e-06, "loss": 0.8197, "step": 1214 }, { "epoch": 1.022722491058279, "grad_norm": 67.51494598388672, "learning_rate": 3.76608157745491e-06, "loss": 0.7087, "step": 1215 }, { "epoch": 1.0235640648011781, "grad_norm": 69.369873046875, "learning_rate": 3.764155474760246e-06, "loss": 0.774, "step": 1216 }, { "epoch": 1.0244056385440774, "grad_norm": 65.6505355834961, "learning_rate": 3.762228363380797e-06, "loss": 0.6347, "step": 1217 }, { "epoch": 1.0252472122869767, "grad_norm": 55.137664794921875, "learning_rate": 3.760300244854229e-06, "loss": 0.4952, "step": 1218 }, { "epoch": 1.026088786029876, "grad_norm": 67.7928466796875, "learning_rate": 3.758371120719011e-06, "loss": 0.6969, "step": 1219 }, { "epoch": 1.026930359772775, "grad_norm": 76.3191909790039, "learning_rate": 3.756440992514414e-06, "loss": 0.7473, "step": 1220 }, { "epoch": 1.0277719335156743, "grad_norm": 81.80817413330078, "learning_rate": 3.754509861780511e-06, "loss": 0.7316, "step": 1221 }, { "epoch": 1.0286135072585736, "grad_norm": 65.12232208251953, "learning_rate": 3.7525777300581723e-06, "loss": 0.7829, "step": 1222 }, { "epoch": 1.0294550810014726, "grad_norm": 136.61734008789062, "learning_rate": 3.7506445988890715e-06, "loss": 0.9183, "step": 1223 }, { "epoch": 1.030296654744372, "grad_norm": 76.92049407958984, "learning_rate": 3.748710469815675e-06, "loss": 0.815, "step": 1224 }, { "epoch": 1.0311382284872712, "grad_norm": 76.00663757324219, "learning_rate": 3.7467753443812493e-06, "loss": 0.6683, "step": 1225 }, { "epoch": 1.0319798022301705, "grad_norm": 82.0116195678711, "learning_rate": 3.7448392241298532e-06, "loss": 0.8531, "step": 1226 }, { "epoch": 1.0328213759730696, "grad_norm": 68.03218841552734, "learning_rate": 3.74290211060634e-06, "loss": 0.7932, "step": 1227 }, { "epoch": 1.0336629497159688, "grad_norm": 85.45331573486328, "learning_rate": 3.7409640053563565e-06, "loss": 0.8355, "step": 1228 }, { "epoch": 1.0345045234588681, "grad_norm": 111.25006103515625, "learning_rate": 3.739024909926339e-06, "loss": 0.94, "step": 1229 }, { "epoch": 1.0353460972017674, "grad_norm": 80.6521224975586, "learning_rate": 3.7370848258635163e-06, "loss": 0.6197, "step": 1230 }, { "epoch": 1.0361876709446665, "grad_norm": 80.95207977294922, "learning_rate": 3.735143754715903e-06, "loss": 0.8114, "step": 1231 }, { "epoch": 1.0370292446875657, "grad_norm": 73.94544219970703, "learning_rate": 3.733201698032305e-06, "loss": 0.692, "step": 1232 }, { "epoch": 1.037870818430465, "grad_norm": 67.47168731689453, "learning_rate": 3.7312586573623103e-06, "loss": 0.7783, "step": 1233 }, { "epoch": 1.038712392173364, "grad_norm": 64.5782241821289, "learning_rate": 3.729314634256297e-06, "loss": 0.8575, "step": 1234 }, { "epoch": 1.0395539659162634, "grad_norm": 80.02053833007812, "learning_rate": 3.7273696302654223e-06, "loss": 0.6189, "step": 1235 }, { "epoch": 1.0403955396591626, "grad_norm": 69.40652465820312, "learning_rate": 3.7254236469416287e-06, "loss": 0.8405, "step": 1236 }, { "epoch": 1.041237113402062, "grad_norm": 77.50457763671875, "learning_rate": 3.72347668583764e-06, "loss": 0.7768, "step": 1237 }, { "epoch": 1.042078687144961, "grad_norm": 58.28728485107422, "learning_rate": 3.7215287485069598e-06, "loss": 0.6804, "step": 1238 }, { "epoch": 1.0429202608878603, "grad_norm": 81.27555084228516, "learning_rate": 3.7195798365038695e-06, "loss": 0.887, "step": 1239 }, { "epoch": 1.0437618346307596, "grad_norm": 107.18325805664062, "learning_rate": 3.7176299513834308e-06, "loss": 1.1082, "step": 1240 }, { "epoch": 1.0446034083736588, "grad_norm": 71.24678802490234, "learning_rate": 3.7156790947014797e-06, "loss": 0.8752, "step": 1241 }, { "epoch": 1.045444982116558, "grad_norm": 64.26017761230469, "learning_rate": 3.713727268014628e-06, "loss": 0.5882, "step": 1242 }, { "epoch": 1.0462865558594572, "grad_norm": 82.4547119140625, "learning_rate": 3.711774472880262e-06, "loss": 0.7684, "step": 1243 }, { "epoch": 1.0471281296023565, "grad_norm": 73.11531066894531, "learning_rate": 3.709820710856539e-06, "loss": 0.7802, "step": 1244 }, { "epoch": 1.0479697033452555, "grad_norm": 83.8969497680664, "learning_rate": 3.707865983502391e-06, "loss": 0.8114, "step": 1245 }, { "epoch": 1.0488112770881548, "grad_norm": 69.56085205078125, "learning_rate": 3.7059102923775175e-06, "loss": 0.6799, "step": 1246 }, { "epoch": 1.049652850831054, "grad_norm": 63.41591262817383, "learning_rate": 3.7039536390423873e-06, "loss": 0.7341, "step": 1247 }, { "epoch": 1.0504944245739534, "grad_norm": 56.824581146240234, "learning_rate": 3.7019960250582383e-06, "loss": 0.6494, "step": 1248 }, { "epoch": 1.0513359983168524, "grad_norm": 74.19931030273438, "learning_rate": 3.700037451987074e-06, "loss": 0.6798, "step": 1249 }, { "epoch": 1.0521775720597517, "grad_norm": 65.28782653808594, "learning_rate": 3.698077921391663e-06, "loss": 0.6992, "step": 1250 }, { "epoch": 1.053019145802651, "grad_norm": 119.53057861328125, "learning_rate": 3.6961174348355383e-06, "loss": 1.2016, "step": 1251 }, { "epoch": 1.0538607195455503, "grad_norm": 59.280277252197266, "learning_rate": 3.694155993882996e-06, "loss": 0.5885, "step": 1252 }, { "epoch": 1.0547022932884493, "grad_norm": 84.87373352050781, "learning_rate": 3.692193600099092e-06, "loss": 0.7715, "step": 1253 }, { "epoch": 1.0555438670313486, "grad_norm": 71.60598754882812, "learning_rate": 3.690230255049646e-06, "loss": 0.8461, "step": 1254 }, { "epoch": 1.056385440774248, "grad_norm": 81.19401550292969, "learning_rate": 3.6882659603012325e-06, "loss": 0.7127, "step": 1255 }, { "epoch": 1.0572270145171472, "grad_norm": 90.88418579101562, "learning_rate": 3.6863007174211873e-06, "loss": 0.8569, "step": 1256 }, { "epoch": 1.0580685882600462, "grad_norm": 60.42268753051758, "learning_rate": 3.6843345279776e-06, "loss": 0.6123, "step": 1257 }, { "epoch": 1.0589101620029455, "grad_norm": 68.3895263671875, "learning_rate": 3.6823673935393168e-06, "loss": 0.6432, "step": 1258 }, { "epoch": 1.0597517357458448, "grad_norm": 69.6331558227539, "learning_rate": 3.6803993156759388e-06, "loss": 0.7601, "step": 1259 }, { "epoch": 1.0605933094887439, "grad_norm": 67.74185180664062, "learning_rate": 3.6784302959578178e-06, "loss": 0.9226, "step": 1260 }, { "epoch": 1.0614348832316431, "grad_norm": 77.12291717529297, "learning_rate": 3.6764603359560585e-06, "loss": 0.862, "step": 1261 }, { "epoch": 1.0622764569745424, "grad_norm": 71.94026947021484, "learning_rate": 3.674489437242516e-06, "loss": 0.7163, "step": 1262 }, { "epoch": 1.0631180307174417, "grad_norm": 84.82723999023438, "learning_rate": 3.6725176013897938e-06, "loss": 0.654, "step": 1263 }, { "epoch": 1.0639596044603408, "grad_norm": 55.34287643432617, "learning_rate": 3.6705448299712433e-06, "loss": 0.4786, "step": 1264 }, { "epoch": 1.06480117820324, "grad_norm": 78.00069427490234, "learning_rate": 3.668571124560962e-06, "loss": 0.8155, "step": 1265 }, { "epoch": 1.0656427519461393, "grad_norm": 68.24934387207031, "learning_rate": 3.666596486733794e-06, "loss": 0.8296, "step": 1266 }, { "epoch": 1.0664843256890384, "grad_norm": 62.619388580322266, "learning_rate": 3.664620918065325e-06, "loss": 0.8412, "step": 1267 }, { "epoch": 1.0673258994319377, "grad_norm": 62.24843978881836, "learning_rate": 3.6626444201318865e-06, "loss": 0.6693, "step": 1268 }, { "epoch": 1.068167473174837, "grad_norm": 72.28678894042969, "learning_rate": 3.6606669945105494e-06, "loss": 0.8794, "step": 1269 }, { "epoch": 1.0690090469177362, "grad_norm": 63.4146728515625, "learning_rate": 3.6586886427791245e-06, "loss": 0.7898, "step": 1270 }, { "epoch": 1.0698506206606353, "grad_norm": 82.39759826660156, "learning_rate": 3.6567093665161636e-06, "loss": 0.811, "step": 1271 }, { "epoch": 1.0706921944035346, "grad_norm": 76.58573150634766, "learning_rate": 3.6547291673009533e-06, "loss": 0.889, "step": 1272 }, { "epoch": 1.0715337681464339, "grad_norm": 68.4346694946289, "learning_rate": 3.6527480467135196e-06, "loss": 0.7371, "step": 1273 }, { "epoch": 1.0723753418893331, "grad_norm": 69.64159393310547, "learning_rate": 3.6507660063346216e-06, "loss": 0.7767, "step": 1274 }, { "epoch": 1.0732169156322322, "grad_norm": 72.70745086669922, "learning_rate": 3.648783047745753e-06, "loss": 0.7388, "step": 1275 }, { "epoch": 1.0740584893751315, "grad_norm": 61.45391845703125, "learning_rate": 3.6467991725291406e-06, "loss": 0.7133, "step": 1276 }, { "epoch": 1.0749000631180308, "grad_norm": 59.521297454833984, "learning_rate": 3.644814382267742e-06, "loss": 0.6792, "step": 1277 }, { "epoch": 1.07574163686093, "grad_norm": 72.21150207519531, "learning_rate": 3.6428286785452454e-06, "loss": 0.7792, "step": 1278 }, { "epoch": 1.076583210603829, "grad_norm": 95.27154541015625, "learning_rate": 3.640842062946067e-06, "loss": 0.9082, "step": 1279 }, { "epoch": 1.0774247843467284, "grad_norm": 62.20063781738281, "learning_rate": 3.638854537055352e-06, "loss": 0.645, "step": 1280 }, { "epoch": 1.0782663580896277, "grad_norm": 61.36935043334961, "learning_rate": 3.6368661024589702e-06, "loss": 0.7807, "step": 1281 }, { "epoch": 1.0791079318325267, "grad_norm": 68.19465637207031, "learning_rate": 3.634876760743518e-06, "loss": 0.8405, "step": 1282 }, { "epoch": 1.079949505575426, "grad_norm": 99.6678466796875, "learning_rate": 3.6328865134963144e-06, "loss": 1.0025, "step": 1283 }, { "epoch": 1.0807910793183253, "grad_norm": 67.33273315429688, "learning_rate": 3.6308953623054025e-06, "loss": 0.693, "step": 1284 }, { "epoch": 1.0816326530612246, "grad_norm": 78.76142120361328, "learning_rate": 3.628903308759545e-06, "loss": 0.855, "step": 1285 }, { "epoch": 1.0824742268041236, "grad_norm": 68.64391326904297, "learning_rate": 3.6269103544482245e-06, "loss": 0.8117, "step": 1286 }, { "epoch": 1.083315800547023, "grad_norm": 79.25052642822266, "learning_rate": 3.6249165009616436e-06, "loss": 0.6921, "step": 1287 }, { "epoch": 1.0841573742899222, "grad_norm": 82.95649719238281, "learning_rate": 3.622921749890723e-06, "loss": 0.7743, "step": 1288 }, { "epoch": 1.0849989480328213, "grad_norm": 73.58319854736328, "learning_rate": 3.620926102827097e-06, "loss": 0.8296, "step": 1289 }, { "epoch": 1.0858405217757205, "grad_norm": 71.20716857910156, "learning_rate": 3.618929561363117e-06, "loss": 0.7582, "step": 1290 }, { "epoch": 1.0866820955186198, "grad_norm": 79.03595733642578, "learning_rate": 3.616932127091848e-06, "loss": 0.8906, "step": 1291 }, { "epoch": 1.087523669261519, "grad_norm": 94.94636535644531, "learning_rate": 3.6149338016070647e-06, "loss": 0.9492, "step": 1292 }, { "epoch": 1.0883652430044182, "grad_norm": 72.05924987792969, "learning_rate": 3.612934586503257e-06, "loss": 0.5987, "step": 1293 }, { "epoch": 1.0892068167473175, "grad_norm": 80.31555938720703, "learning_rate": 3.610934483375621e-06, "loss": 0.9135, "step": 1294 }, { "epoch": 1.0900483904902167, "grad_norm": 80.71312713623047, "learning_rate": 3.6089334938200644e-06, "loss": 0.8973, "step": 1295 }, { "epoch": 1.090889964233116, "grad_norm": 69.00127410888672, "learning_rate": 3.6069316194331995e-06, "loss": 0.7587, "step": 1296 }, { "epoch": 1.091731537976015, "grad_norm": 63.67856979370117, "learning_rate": 3.6049288618123456e-06, "loss": 0.7781, "step": 1297 }, { "epoch": 1.0925731117189144, "grad_norm": 78.16109466552734, "learning_rate": 3.6029252225555283e-06, "loss": 0.8975, "step": 1298 }, { "epoch": 1.0934146854618136, "grad_norm": 82.5450668334961, "learning_rate": 3.6009207032614748e-06, "loss": 0.7018, "step": 1299 }, { "epoch": 1.094256259204713, "grad_norm": 67.94577026367188, "learning_rate": 3.598915305529615e-06, "loss": 0.7293, "step": 1300 }, { "epoch": 1.095097832947612, "grad_norm": 76.59442138671875, "learning_rate": 3.596909030960079e-06, "loss": 0.8041, "step": 1301 }, { "epoch": 1.0959394066905113, "grad_norm": 59.258644104003906, "learning_rate": 3.5949018811536984e-06, "loss": 0.8281, "step": 1302 }, { "epoch": 1.0967809804334105, "grad_norm": 81.13563537597656, "learning_rate": 3.5928938577120015e-06, "loss": 0.7705, "step": 1303 }, { "epoch": 1.0976225541763096, "grad_norm": 84.97859954833984, "learning_rate": 3.590884962237214e-06, "loss": 0.7768, "step": 1304 }, { "epoch": 1.0984641279192089, "grad_norm": 75.2193832397461, "learning_rate": 3.5888751963322587e-06, "loss": 0.7057, "step": 1305 }, { "epoch": 1.0993057016621082, "grad_norm": 62.515525817871094, "learning_rate": 3.586864561600751e-06, "loss": 0.607, "step": 1306 }, { "epoch": 1.1001472754050075, "grad_norm": 85.55615234375, "learning_rate": 3.5848530596470005e-06, "loss": 0.9196, "step": 1307 }, { "epoch": 1.1009888491479065, "grad_norm": 87.32113647460938, "learning_rate": 3.5828406920760085e-06, "loss": 1.0084, "step": 1308 }, { "epoch": 1.1018304228908058, "grad_norm": 72.80986785888672, "learning_rate": 3.580827460493468e-06, "loss": 0.6663, "step": 1309 }, { "epoch": 1.102671996633705, "grad_norm": 82.09026336669922, "learning_rate": 3.5788133665057595e-06, "loss": 0.8127, "step": 1310 }, { "epoch": 1.1035135703766041, "grad_norm": 89.75031280517578, "learning_rate": 3.5767984117199535e-06, "loss": 0.8213, "step": 1311 }, { "epoch": 1.1043551441195034, "grad_norm": 81.2878646850586, "learning_rate": 3.5747825977438062e-06, "loss": 0.8556, "step": 1312 }, { "epoch": 1.1051967178624027, "grad_norm": 73.67637634277344, "learning_rate": 3.57276592618576e-06, "loss": 0.7873, "step": 1313 }, { "epoch": 1.106038291605302, "grad_norm": 93.03740692138672, "learning_rate": 3.5707483986549407e-06, "loss": 0.7661, "step": 1314 }, { "epoch": 1.106879865348201, "grad_norm": 73.1768569946289, "learning_rate": 3.568730016761159e-06, "loss": 0.8781, "step": 1315 }, { "epoch": 1.1077214390911003, "grad_norm": 52.64601516723633, "learning_rate": 3.566710782114904e-06, "loss": 0.5982, "step": 1316 }, { "epoch": 1.1085630128339996, "grad_norm": 67.40423583984375, "learning_rate": 3.5646906963273482e-06, "loss": 0.6887, "step": 1317 }, { "epoch": 1.1094045865768989, "grad_norm": 76.27008819580078, "learning_rate": 3.5626697610103424e-06, "loss": 0.8059, "step": 1318 }, { "epoch": 1.110246160319798, "grad_norm": 68.23004150390625, "learning_rate": 3.5606479777764142e-06, "loss": 0.7669, "step": 1319 }, { "epoch": 1.1110877340626972, "grad_norm": 68.53443145751953, "learning_rate": 3.55862534823877e-06, "loss": 0.669, "step": 1320 }, { "epoch": 1.1119293078055965, "grad_norm": 60.3872184753418, "learning_rate": 3.556601874011289e-06, "loss": 0.6156, "step": 1321 }, { "epoch": 1.1127708815484958, "grad_norm": 57.09178161621094, "learning_rate": 3.5545775567085263e-06, "loss": 0.8307, "step": 1322 }, { "epoch": 1.1136124552913949, "grad_norm": 77.64226531982422, "learning_rate": 3.552552397945708e-06, "loss": 0.963, "step": 1323 }, { "epoch": 1.1144540290342941, "grad_norm": 67.55712127685547, "learning_rate": 3.550526399338734e-06, "loss": 0.7474, "step": 1324 }, { "epoch": 1.1152956027771934, "grad_norm": 79.86772155761719, "learning_rate": 3.5484995625041713e-06, "loss": 0.7521, "step": 1325 }, { "epoch": 1.1161371765200925, "grad_norm": 132.9801483154297, "learning_rate": 3.5464718890592583e-06, "loss": 0.6854, "step": 1326 }, { "epoch": 1.1169787502629918, "grad_norm": 79.50981903076172, "learning_rate": 3.5444433806218993e-06, "loss": 0.9638, "step": 1327 }, { "epoch": 1.117820324005891, "grad_norm": 84.78490447998047, "learning_rate": 3.5424140388106652e-06, "loss": 0.7145, "step": 1328 }, { "epoch": 1.1186618977487903, "grad_norm": 68.53125, "learning_rate": 3.540383865244793e-06, "loss": 0.5723, "step": 1329 }, { "epoch": 1.1195034714916894, "grad_norm": 65.09300231933594, "learning_rate": 3.538352861544182e-06, "loss": 0.6238, "step": 1330 }, { "epoch": 1.1203450452345887, "grad_norm": 148.02352905273438, "learning_rate": 3.536321029329395e-06, "loss": 0.6512, "step": 1331 }, { "epoch": 1.121186618977488, "grad_norm": 67.34825897216797, "learning_rate": 3.5342883702216534e-06, "loss": 0.7211, "step": 1332 }, { "epoch": 1.122028192720387, "grad_norm": 83.49595642089844, "learning_rate": 3.5322548858428424e-06, "loss": 0.8661, "step": 1333 }, { "epoch": 1.1228697664632863, "grad_norm": 66.2895736694336, "learning_rate": 3.530220577815502e-06, "loss": 0.7105, "step": 1334 }, { "epoch": 1.1237113402061856, "grad_norm": 114.4986801147461, "learning_rate": 3.5281854477628317e-06, "loss": 0.7341, "step": 1335 }, { "epoch": 1.1245529139490849, "grad_norm": 75.82929992675781, "learning_rate": 3.526149497308686e-06, "loss": 0.7586, "step": 1336 }, { "epoch": 1.125394487691984, "grad_norm": 84.63530731201172, "learning_rate": 3.524112728077575e-06, "loss": 0.9024, "step": 1337 }, { "epoch": 1.1262360614348832, "grad_norm": 81.52574920654297, "learning_rate": 3.52207514169466e-06, "loss": 0.8914, "step": 1338 }, { "epoch": 1.1270776351777825, "grad_norm": 97.83477020263672, "learning_rate": 3.5200367397857553e-06, "loss": 1.0349, "step": 1339 }, { "epoch": 1.1279192089206818, "grad_norm": 90.05326080322266, "learning_rate": 3.5179975239773266e-06, "loss": 0.952, "step": 1340 }, { "epoch": 1.1287607826635808, "grad_norm": 73.44329071044922, "learning_rate": 3.5159574958964895e-06, "loss": 0.7403, "step": 1341 }, { "epoch": 1.12960235640648, "grad_norm": 56.738311767578125, "learning_rate": 3.513916657171006e-06, "loss": 0.6308, "step": 1342 }, { "epoch": 1.1304439301493794, "grad_norm": 60.14042282104492, "learning_rate": 3.5118750094292853e-06, "loss": 0.6327, "step": 1343 }, { "epoch": 1.1312855038922787, "grad_norm": 73.36322021484375, "learning_rate": 3.509832554300384e-06, "loss": 0.7424, "step": 1344 }, { "epoch": 1.1321270776351777, "grad_norm": 96.14129638671875, "learning_rate": 3.5077892934140006e-06, "loss": 0.9552, "step": 1345 }, { "epoch": 1.132968651378077, "grad_norm": 85.27511596679688, "learning_rate": 3.5057452284004767e-06, "loss": 0.9572, "step": 1346 }, { "epoch": 1.1338102251209763, "grad_norm": 65.14854431152344, "learning_rate": 3.5037003608907975e-06, "loss": 0.7357, "step": 1347 }, { "epoch": 1.1346517988638753, "grad_norm": 85.38391876220703, "learning_rate": 3.5016546925165866e-06, "loss": 0.8456, "step": 1348 }, { "epoch": 1.1354933726067746, "grad_norm": 83.07268524169922, "learning_rate": 3.499608224910107e-06, "loss": 1.0145, "step": 1349 }, { "epoch": 1.136334946349674, "grad_norm": 82.41991424560547, "learning_rate": 3.497560959704261e-06, "loss": 0.9077, "step": 1350 }, { "epoch": 1.1371765200925732, "grad_norm": 87.86802673339844, "learning_rate": 3.495512898532585e-06, "loss": 1.0205, "step": 1351 }, { "epoch": 1.1380180938354723, "grad_norm": 67.33047485351562, "learning_rate": 3.4934640430292512e-06, "loss": 0.6918, "step": 1352 }, { "epoch": 1.1388596675783715, "grad_norm": 58.26289367675781, "learning_rate": 3.491414394829066e-06, "loss": 0.6877, "step": 1353 }, { "epoch": 1.1397012413212708, "grad_norm": 73.31893157958984, "learning_rate": 3.4893639555674695e-06, "loss": 0.8029, "step": 1354 }, { "epoch": 1.1405428150641699, "grad_norm": 68.03288269042969, "learning_rate": 3.487312726880531e-06, "loss": 0.7889, "step": 1355 }, { "epoch": 1.1413843888070692, "grad_norm": 66.80435180664062, "learning_rate": 3.48526071040495e-06, "loss": 0.7886, "step": 1356 }, { "epoch": 1.1422259625499684, "grad_norm": 70.782958984375, "learning_rate": 3.483207907778055e-06, "loss": 0.8967, "step": 1357 }, { "epoch": 1.1430675362928677, "grad_norm": 70.4962158203125, "learning_rate": 3.481154320637802e-06, "loss": 0.6492, "step": 1358 }, { "epoch": 1.143909110035767, "grad_norm": 63.329246520996094, "learning_rate": 3.4790999506227737e-06, "loss": 0.7749, "step": 1359 }, { "epoch": 1.144750683778666, "grad_norm": 75.54255676269531, "learning_rate": 3.4770447993721747e-06, "loss": 0.8227, "step": 1360 }, { "epoch": 1.1455922575215653, "grad_norm": 62.19999694824219, "learning_rate": 3.474988868525836e-06, "loss": 0.7421, "step": 1361 }, { "epoch": 1.1464338312644646, "grad_norm": 84.12921142578125, "learning_rate": 3.4729321597242095e-06, "loss": 0.8211, "step": 1362 }, { "epoch": 1.1472754050073637, "grad_norm": 66.9234390258789, "learning_rate": 3.470874674608367e-06, "loss": 0.6031, "step": 1363 }, { "epoch": 1.148116978750263, "grad_norm": 58.342872619628906, "learning_rate": 3.468816414820001e-06, "loss": 0.6996, "step": 1364 }, { "epoch": 1.1489585524931623, "grad_norm": 69.41954040527344, "learning_rate": 3.4667573820014215e-06, "loss": 0.7793, "step": 1365 }, { "epoch": 1.1498001262360615, "grad_norm": 93.1528549194336, "learning_rate": 3.464697577795556e-06, "loss": 0.8255, "step": 1366 }, { "epoch": 1.1506416999789606, "grad_norm": 60.05856704711914, "learning_rate": 3.4626370038459455e-06, "loss": 0.5099, "step": 1367 }, { "epoch": 1.1514832737218599, "grad_norm": 82.09307861328125, "learning_rate": 3.4605756617967485e-06, "loss": 0.7799, "step": 1368 }, { "epoch": 1.1523248474647592, "grad_norm": 132.58616638183594, "learning_rate": 3.458513553292733e-06, "loss": 0.7991, "step": 1369 }, { "epoch": 1.1531664212076582, "grad_norm": 62.63915252685547, "learning_rate": 3.456450679979281e-06, "loss": 0.6916, "step": 1370 }, { "epoch": 1.1540079949505575, "grad_norm": 74.68429565429688, "learning_rate": 3.454387043502383e-06, "loss": 0.8426, "step": 1371 }, { "epoch": 1.1548495686934568, "grad_norm": 64.6526107788086, "learning_rate": 3.45232264550864e-06, "loss": 0.7868, "step": 1372 }, { "epoch": 1.155691142436356, "grad_norm": 71.72866821289062, "learning_rate": 3.4502574876452594e-06, "loss": 0.9175, "step": 1373 }, { "epoch": 1.1565327161792551, "grad_norm": 65.70030975341797, "learning_rate": 3.448191571560055e-06, "loss": 0.6926, "step": 1374 }, { "epoch": 1.1573742899221544, "grad_norm": 90.4015121459961, "learning_rate": 3.4461248989014466e-06, "loss": 0.9557, "step": 1375 }, { "epoch": 1.1582158636650537, "grad_norm": 110.88024139404297, "learning_rate": 3.444057471318457e-06, "loss": 0.748, "step": 1376 }, { "epoch": 1.1590574374079528, "grad_norm": 66.52738952636719, "learning_rate": 3.4419892904607124e-06, "loss": 0.7728, "step": 1377 }, { "epoch": 1.159899011150852, "grad_norm": 81.06018829345703, "learning_rate": 3.439920357978437e-06, "loss": 0.8179, "step": 1378 }, { "epoch": 1.1607405848937513, "grad_norm": 67.53968048095703, "learning_rate": 3.437850675522457e-06, "loss": 0.779, "step": 1379 }, { "epoch": 1.1615821586366506, "grad_norm": 64.2708511352539, "learning_rate": 3.435780244744199e-06, "loss": 0.8699, "step": 1380 }, { "epoch": 1.1624237323795499, "grad_norm": 65.01960754394531, "learning_rate": 3.433709067295683e-06, "loss": 0.6326, "step": 1381 }, { "epoch": 1.163265306122449, "grad_norm": 61.47157669067383, "learning_rate": 3.4316371448295256e-06, "loss": 0.7747, "step": 1382 }, { "epoch": 1.1641068798653482, "grad_norm": 74.15666198730469, "learning_rate": 3.4295644789989397e-06, "loss": 0.7089, "step": 1383 }, { "epoch": 1.1649484536082475, "grad_norm": 59.576751708984375, "learning_rate": 3.427491071457729e-06, "loss": 0.7315, "step": 1384 }, { "epoch": 1.1657900273511466, "grad_norm": 57.37242126464844, "learning_rate": 3.425416923860292e-06, "loss": 0.7324, "step": 1385 }, { "epoch": 1.1666316010940458, "grad_norm": 67.24441528320312, "learning_rate": 3.4233420378616144e-06, "loss": 0.8165, "step": 1386 }, { "epoch": 1.1674731748369451, "grad_norm": 72.99162292480469, "learning_rate": 3.421266415117274e-06, "loss": 0.8523, "step": 1387 }, { "epoch": 1.1683147485798444, "grad_norm": 83.09717559814453, "learning_rate": 3.4191900572834325e-06, "loss": 0.6511, "step": 1388 }, { "epoch": 1.1691563223227435, "grad_norm": 73.0621337890625, "learning_rate": 3.4171129660168433e-06, "loss": 0.7932, "step": 1389 }, { "epoch": 1.1699978960656428, "grad_norm": 63.88787841796875, "learning_rate": 3.4150351429748406e-06, "loss": 0.7402, "step": 1390 }, { "epoch": 1.170839469808542, "grad_norm": 68.88009643554688, "learning_rate": 3.4129565898153448e-06, "loss": 0.6272, "step": 1391 }, { "epoch": 1.171681043551441, "grad_norm": 75.77027893066406, "learning_rate": 3.4108773081968594e-06, "loss": 0.8729, "step": 1392 }, { "epoch": 1.1725226172943404, "grad_norm": 63.99619674682617, "learning_rate": 3.408797299778467e-06, "loss": 0.7848, "step": 1393 }, { "epoch": 1.1733641910372397, "grad_norm": 64.0927505493164, "learning_rate": 3.406716566219831e-06, "loss": 0.8276, "step": 1394 }, { "epoch": 1.174205764780139, "grad_norm": 224.97900390625, "learning_rate": 3.4046351091811945e-06, "loss": 0.8918, "step": 1395 }, { "epoch": 1.175047338523038, "grad_norm": 79.71916198730469, "learning_rate": 3.4025529303233764e-06, "loss": 0.8166, "step": 1396 }, { "epoch": 1.1758889122659373, "grad_norm": 72.94778442382812, "learning_rate": 3.400470031307772e-06, "loss": 0.6815, "step": 1397 }, { "epoch": 1.1767304860088366, "grad_norm": 71.34203338623047, "learning_rate": 3.3983864137963517e-06, "loss": 0.6691, "step": 1398 }, { "epoch": 1.1775720597517356, "grad_norm": 75.5274658203125, "learning_rate": 3.396302079451659e-06, "loss": 0.8337, "step": 1399 }, { "epoch": 1.178413633494635, "grad_norm": 81.76123046875, "learning_rate": 3.3942170299368094e-06, "loss": 0.9503, "step": 1400 }, { "epoch": 1.1792552072375342, "grad_norm": 198.34930419921875, "learning_rate": 3.3921312669154883e-06, "loss": 0.6294, "step": 1401 }, { "epoch": 1.1800967809804335, "grad_norm": 73.18561553955078, "learning_rate": 3.3900447920519507e-06, "loss": 0.8083, "step": 1402 }, { "epoch": 1.1809383547233328, "grad_norm": 77.73960876464844, "learning_rate": 3.387957607011021e-06, "loss": 0.7997, "step": 1403 }, { "epoch": 1.1817799284662318, "grad_norm": 69.83214569091797, "learning_rate": 3.3858697134580875e-06, "loss": 0.5873, "step": 1404 }, { "epoch": 1.182621502209131, "grad_norm": 51.2693977355957, "learning_rate": 3.383781113059107e-06, "loss": 0.7206, "step": 1405 }, { "epoch": 1.1834630759520304, "grad_norm": 70.56368255615234, "learning_rate": 3.3816918074805982e-06, "loss": 0.7665, "step": 1406 }, { "epoch": 1.1843046496949294, "grad_norm": 59.05124282836914, "learning_rate": 3.379601798389642e-06, "loss": 0.6055, "step": 1407 }, { "epoch": 1.1851462234378287, "grad_norm": 61.26966094970703, "learning_rate": 3.3775110874538834e-06, "loss": 0.6029, "step": 1408 }, { "epoch": 1.185987797180728, "grad_norm": 65.18826293945312, "learning_rate": 3.3754196763415255e-06, "loss": 0.8414, "step": 1409 }, { "epoch": 1.1868293709236273, "grad_norm": 74.95602416992188, "learning_rate": 3.373327566721329e-06, "loss": 0.9045, "step": 1410 }, { "epoch": 1.1876709446665263, "grad_norm": 54.38325500488281, "learning_rate": 3.371234760262614e-06, "loss": 0.6686, "step": 1411 }, { "epoch": 1.1885125184094256, "grad_norm": 78.43930053710938, "learning_rate": 3.3691412586352552e-06, "loss": 0.7651, "step": 1412 }, { "epoch": 1.189354092152325, "grad_norm": 72.94366455078125, "learning_rate": 3.3670470635096837e-06, "loss": 0.9488, "step": 1413 }, { "epoch": 1.190195665895224, "grad_norm": 64.21903991699219, "learning_rate": 3.364952176556882e-06, "loss": 0.5698, "step": 1414 }, { "epoch": 1.1910372396381232, "grad_norm": 79.49015045166016, "learning_rate": 3.362856599448386e-06, "loss": 0.9068, "step": 1415 }, { "epoch": 1.1918788133810225, "grad_norm": 70.82846069335938, "learning_rate": 3.360760333856282e-06, "loss": 0.7983, "step": 1416 }, { "epoch": 1.1927203871239218, "grad_norm": 85.83604431152344, "learning_rate": 3.3586633814532043e-06, "loss": 0.76, "step": 1417 }, { "epoch": 1.1935619608668209, "grad_norm": 70.05342102050781, "learning_rate": 3.3565657439123368e-06, "loss": 0.7129, "step": 1418 }, { "epoch": 1.1944035346097202, "grad_norm": 85.30274963378906, "learning_rate": 3.3544674229074103e-06, "loss": 0.9235, "step": 1419 }, { "epoch": 1.1952451083526194, "grad_norm": 126.79804229736328, "learning_rate": 3.3523684201126987e-06, "loss": 0.5791, "step": 1420 }, { "epoch": 1.1960866820955187, "grad_norm": 67.60092163085938, "learning_rate": 3.3502687372030235e-06, "loss": 0.758, "step": 1421 }, { "epoch": 1.1969282558384178, "grad_norm": 98.4145278930664, "learning_rate": 3.3481683758537454e-06, "loss": 0.9113, "step": 1422 }, { "epoch": 1.197769829581317, "grad_norm": 63.639102935791016, "learning_rate": 3.3460673377407676e-06, "loss": 0.7124, "step": 1423 }, { "epoch": 1.1986114033242163, "grad_norm": 60.17888259887695, "learning_rate": 3.343965624540535e-06, "loss": 0.6899, "step": 1424 }, { "epoch": 1.1994529770671156, "grad_norm": 86.66293334960938, "learning_rate": 3.341863237930029e-06, "loss": 1.0092, "step": 1425 }, { "epoch": 1.2002945508100147, "grad_norm": 76.53636169433594, "learning_rate": 3.3397601795867684e-06, "loss": 0.7176, "step": 1426 }, { "epoch": 1.201136124552914, "grad_norm": 72.173583984375, "learning_rate": 3.3376564511888095e-06, "loss": 0.8557, "step": 1427 }, { "epoch": 1.2019776982958132, "grad_norm": 58.85466003417969, "learning_rate": 3.335552054414742e-06, "loss": 0.6287, "step": 1428 }, { "epoch": 1.2028192720387123, "grad_norm": 58.9666748046875, "learning_rate": 3.333446990943689e-06, "loss": 0.6785, "step": 1429 }, { "epoch": 1.2036608457816116, "grad_norm": 57.108009338378906, "learning_rate": 3.3313412624553066e-06, "loss": 0.7176, "step": 1430 }, { "epoch": 1.2045024195245109, "grad_norm": 68.65019226074219, "learning_rate": 3.3292348706297807e-06, "loss": 0.8955, "step": 1431 }, { "epoch": 1.2053439932674102, "grad_norm": 82.8580322265625, "learning_rate": 3.327127817147825e-06, "loss": 0.933, "step": 1432 }, { "epoch": 1.2061855670103092, "grad_norm": 74.54534149169922, "learning_rate": 3.3250201036906843e-06, "loss": 0.6931, "step": 1433 }, { "epoch": 1.2070271407532085, "grad_norm": 56.394012451171875, "learning_rate": 3.322911731940128e-06, "loss": 0.5127, "step": 1434 }, { "epoch": 1.2078687144961078, "grad_norm": 71.54861450195312, "learning_rate": 3.3208027035784513e-06, "loss": 0.84, "step": 1435 }, { "epoch": 1.2087102882390068, "grad_norm": 71.24790954589844, "learning_rate": 3.3186930202884725e-06, "loss": 0.7949, "step": 1436 }, { "epoch": 1.2095518619819061, "grad_norm": 62.81068801879883, "learning_rate": 3.316582683753534e-06, "loss": 0.7365, "step": 1437 }, { "epoch": 1.2103934357248054, "grad_norm": 68.5630111694336, "learning_rate": 3.314471695657498e-06, "loss": 0.5798, "step": 1438 }, { "epoch": 1.2112350094677047, "grad_norm": 66.36724853515625, "learning_rate": 3.3123600576847475e-06, "loss": 0.8426, "step": 1439 }, { "epoch": 1.2120765832106037, "grad_norm": 77.36634063720703, "learning_rate": 3.3102477715201835e-06, "loss": 0.9764, "step": 1440 }, { "epoch": 1.212918156953503, "grad_norm": 77.31050872802734, "learning_rate": 3.3081348388492247e-06, "loss": 0.9308, "step": 1441 }, { "epoch": 1.2137597306964023, "grad_norm": 74.0624771118164, "learning_rate": 3.306021261357805e-06, "loss": 0.8019, "step": 1442 }, { "epoch": 1.2146013044393016, "grad_norm": 63.09843063354492, "learning_rate": 3.303907040732374e-06, "loss": 0.7373, "step": 1443 }, { "epoch": 1.2154428781822006, "grad_norm": 69.69409942626953, "learning_rate": 3.301792178659893e-06, "loss": 0.8323, "step": 1444 }, { "epoch": 1.2162844519251, "grad_norm": 48.24660110473633, "learning_rate": 3.299676676827834e-06, "loss": 0.6096, "step": 1445 }, { "epoch": 1.2171260256679992, "grad_norm": 71.8796615600586, "learning_rate": 3.297560536924185e-06, "loss": 0.757, "step": 1446 }, { "epoch": 1.2179675994108985, "grad_norm": 74.03809356689453, "learning_rate": 3.295443760637437e-06, "loss": 0.7347, "step": 1447 }, { "epoch": 1.2188091731537976, "grad_norm": 75.21710968017578, "learning_rate": 3.2933263496565908e-06, "loss": 0.8716, "step": 1448 }, { "epoch": 1.2196507468966968, "grad_norm": 77.39705657958984, "learning_rate": 3.2912083056711545e-06, "loss": 0.7371, "step": 1449 }, { "epoch": 1.2204923206395961, "grad_norm": 82.25308990478516, "learning_rate": 3.2890896303711417e-06, "loss": 0.6782, "step": 1450 }, { "epoch": 1.2213338943824952, "grad_norm": 57.6148681640625, "learning_rate": 3.2869703254470677e-06, "loss": 0.8079, "step": 1451 }, { "epoch": 1.2221754681253945, "grad_norm": 62.759891510009766, "learning_rate": 3.284850392589951e-06, "loss": 0.7079, "step": 1452 }, { "epoch": 1.2230170418682937, "grad_norm": 61.7097053527832, "learning_rate": 3.282729833491312e-06, "loss": 0.8856, "step": 1453 }, { "epoch": 1.223858615611193, "grad_norm": 63.70922088623047, "learning_rate": 3.2806086498431704e-06, "loss": 0.8202, "step": 1454 }, { "epoch": 1.224700189354092, "grad_norm": 59.12576675415039, "learning_rate": 3.278486843338044e-06, "loss": 0.6588, "step": 1455 }, { "epoch": 1.2255417630969914, "grad_norm": 69.03202056884766, "learning_rate": 3.276364415668947e-06, "loss": 0.9184, "step": 1456 }, { "epoch": 1.2263833368398906, "grad_norm": 63.162418365478516, "learning_rate": 3.2742413685293906e-06, "loss": 0.6063, "step": 1457 }, { "epoch": 1.2272249105827897, "grad_norm": 47.80107879638672, "learning_rate": 3.272117703613379e-06, "loss": 0.6087, "step": 1458 }, { "epoch": 1.228066484325689, "grad_norm": 69.43952178955078, "learning_rate": 3.2699934226154093e-06, "loss": 0.7593, "step": 1459 }, { "epoch": 1.2289080580685883, "grad_norm": 80.62454986572266, "learning_rate": 3.267868527230473e-06, "loss": 0.7964, "step": 1460 }, { "epoch": 1.2297496318114876, "grad_norm": 85.75826263427734, "learning_rate": 3.2657430191540467e-06, "loss": 0.7348, "step": 1461 }, { "epoch": 1.2305912055543866, "grad_norm": 67.71492004394531, "learning_rate": 3.2636169000821007e-06, "loss": 0.8163, "step": 1462 }, { "epoch": 1.231432779297286, "grad_norm": 59.86122131347656, "learning_rate": 3.2614901717110893e-06, "loss": 0.6843, "step": 1463 }, { "epoch": 1.2322743530401852, "grad_norm": 61.91895294189453, "learning_rate": 3.259362835737956e-06, "loss": 0.6392, "step": 1464 }, { "epoch": 1.2331159267830845, "grad_norm": 62.762943267822266, "learning_rate": 3.257234893860126e-06, "loss": 0.641, "step": 1465 }, { "epoch": 1.2339575005259835, "grad_norm": 56.892093658447266, "learning_rate": 3.255106347775512e-06, "loss": 0.6627, "step": 1466 }, { "epoch": 1.2347990742688828, "grad_norm": 64.63336944580078, "learning_rate": 3.2529771991825043e-06, "loss": 0.7472, "step": 1467 }, { "epoch": 1.235640648011782, "grad_norm": 81.38465881347656, "learning_rate": 3.250847449779977e-06, "loss": 0.7303, "step": 1468 }, { "epoch": 1.2364822217546814, "grad_norm": 66.22145080566406, "learning_rate": 3.2487171012672825e-06, "loss": 0.9322, "step": 1469 }, { "epoch": 1.2373237954975804, "grad_norm": 77.17510986328125, "learning_rate": 3.2465861553442514e-06, "loss": 0.9179, "step": 1470 }, { "epoch": 1.2381653692404797, "grad_norm": 84.92408752441406, "learning_rate": 3.2444546137111905e-06, "loss": 0.8133, "step": 1471 }, { "epoch": 1.239006942983379, "grad_norm": 82.94153594970703, "learning_rate": 3.2423224780688827e-06, "loss": 0.7668, "step": 1472 }, { "epoch": 1.239848516726278, "grad_norm": 89.27391815185547, "learning_rate": 3.240189750118584e-06, "loss": 0.8618, "step": 1473 }, { "epoch": 1.2406900904691773, "grad_norm": 82.90675354003906, "learning_rate": 3.2380564315620254e-06, "loss": 0.9421, "step": 1474 }, { "epoch": 1.2415316642120766, "grad_norm": 80.6243667602539, "learning_rate": 3.2359225241014057e-06, "loss": 0.6881, "step": 1475 }, { "epoch": 1.242373237954976, "grad_norm": 79.01913452148438, "learning_rate": 3.2337880294393953e-06, "loss": 0.7986, "step": 1476 }, { "epoch": 1.243214811697875, "grad_norm": 63.30573272705078, "learning_rate": 3.231652949279133e-06, "loss": 0.7715, "step": 1477 }, { "epoch": 1.2440563854407742, "grad_norm": 67.03791809082031, "learning_rate": 3.2295172853242252e-06, "loss": 0.7908, "step": 1478 }, { "epoch": 1.2448979591836735, "grad_norm": 68.64743041992188, "learning_rate": 3.2273810392787443e-06, "loss": 0.8484, "step": 1479 }, { "epoch": 1.2457395329265726, "grad_norm": 67.99822998046875, "learning_rate": 3.225244212847225e-06, "loss": 0.7503, "step": 1480 }, { "epoch": 1.2465811066694719, "grad_norm": 58.34498596191406, "learning_rate": 3.2231068077346676e-06, "loss": 0.6159, "step": 1481 }, { "epoch": 1.2474226804123711, "grad_norm": 53.544925689697266, "learning_rate": 3.220968825646534e-06, "loss": 0.6952, "step": 1482 }, { "epoch": 1.2482642541552704, "grad_norm": 66.74227905273438, "learning_rate": 3.218830268288744e-06, "loss": 0.6318, "step": 1483 }, { "epoch": 1.2491058278981695, "grad_norm": 69.69791412353516, "learning_rate": 3.21669113736768e-06, "loss": 0.606, "step": 1484 }, { "epoch": 1.2499474016410688, "grad_norm": 84.73765563964844, "learning_rate": 3.2145514345901786e-06, "loss": 0.791, "step": 1485 }, { "epoch": 1.250788975383968, "grad_norm": 75.47090148925781, "learning_rate": 3.212411161663535e-06, "loss": 0.831, "step": 1486 }, { "epoch": 1.251630549126867, "grad_norm": 88.83540344238281, "learning_rate": 3.2102703202954984e-06, "loss": 0.8213, "step": 1487 }, { "epoch": 1.2524721228697664, "grad_norm": 64.17558288574219, "learning_rate": 3.208128912194272e-06, "loss": 0.763, "step": 1488 }, { "epoch": 1.2533136966126657, "grad_norm": 66.27608489990234, "learning_rate": 3.2059869390685106e-06, "loss": 0.7587, "step": 1489 }, { "epoch": 1.254155270355565, "grad_norm": 69.03297424316406, "learning_rate": 3.2038444026273207e-06, "loss": 0.7366, "step": 1490 }, { "epoch": 1.2549968440984642, "grad_norm": 60.40120315551758, "learning_rate": 3.2017013045802576e-06, "loss": 0.6389, "step": 1491 }, { "epoch": 1.2558384178413633, "grad_norm": 66.19833374023438, "learning_rate": 3.1995576466373242e-06, "loss": 0.7644, "step": 1492 }, { "epoch": 1.2566799915842626, "grad_norm": 68.6891098022461, "learning_rate": 3.197413430508971e-06, "loss": 0.5621, "step": 1493 }, { "epoch": 1.2575215653271619, "grad_norm": 88.04609680175781, "learning_rate": 3.1952686579060943e-06, "loss": 0.9415, "step": 1494 }, { "epoch": 1.258363139070061, "grad_norm": 77.23043823242188, "learning_rate": 3.193123330540034e-06, "loss": 0.6554, "step": 1495 }, { "epoch": 1.2592047128129602, "grad_norm": 62.28450012207031, "learning_rate": 3.190977450122571e-06, "loss": 0.6336, "step": 1496 }, { "epoch": 1.2600462865558595, "grad_norm": 84.38449096679688, "learning_rate": 3.188831018365931e-06, "loss": 0.7267, "step": 1497 }, { "epoch": 1.2608878602987588, "grad_norm": 55.069732666015625, "learning_rate": 3.1866840369827764e-06, "loss": 0.668, "step": 1498 }, { "epoch": 1.2617294340416578, "grad_norm": 51.937782287597656, "learning_rate": 3.1845365076862095e-06, "loss": 0.6898, "step": 1499 }, { "epoch": 1.262571007784557, "grad_norm": 73.25384521484375, "learning_rate": 3.18238843218977e-06, "loss": 0.6591, "step": 1500 }, { "epoch": 1.2634125815274564, "grad_norm": 72.16972351074219, "learning_rate": 3.1802398122074317e-06, "loss": 0.5857, "step": 1501 }, { "epoch": 1.2642541552703555, "grad_norm": 71.87747955322266, "learning_rate": 3.1780906494536064e-06, "loss": 0.7514, "step": 1502 }, { "epoch": 1.2650957290132547, "grad_norm": 82.81014251708984, "learning_rate": 3.1759409456431357e-06, "loss": 0.9101, "step": 1503 }, { "epoch": 1.265937302756154, "grad_norm": 71.5657730102539, "learning_rate": 3.1737907024912935e-06, "loss": 0.9653, "step": 1504 }, { "epoch": 1.2667788764990533, "grad_norm": 82.05487823486328, "learning_rate": 3.171639921713785e-06, "loss": 0.7225, "step": 1505 }, { "epoch": 1.2676204502419526, "grad_norm": 68.70574951171875, "learning_rate": 3.1694886050267438e-06, "loss": 0.7879, "step": 1506 }, { "epoch": 1.2684620239848516, "grad_norm": 141.91429138183594, "learning_rate": 3.167336754146731e-06, "loss": 0.8457, "step": 1507 }, { "epoch": 1.269303597727751, "grad_norm": 59.09758758544922, "learning_rate": 3.165184370790735e-06, "loss": 0.6574, "step": 1508 }, { "epoch": 1.27014517147065, "grad_norm": 64.79000854492188, "learning_rate": 3.1630314566761676e-06, "loss": 0.5976, "step": 1509 }, { "epoch": 1.2709867452135493, "grad_norm": 76.0698013305664, "learning_rate": 3.160878013520865e-06, "loss": 0.7067, "step": 1510 }, { "epoch": 1.2718283189564485, "grad_norm": 53.58982467651367, "learning_rate": 3.1587240430430848e-06, "loss": 0.5869, "step": 1511 }, { "epoch": 1.2726698926993478, "grad_norm": 64.46701049804688, "learning_rate": 3.1565695469615063e-06, "loss": 0.6679, "step": 1512 }, { "epoch": 1.273511466442247, "grad_norm": 84.78414916992188, "learning_rate": 3.1544145269952288e-06, "loss": 0.7975, "step": 1513 }, { "epoch": 1.2743530401851462, "grad_norm": 54.48835754394531, "learning_rate": 3.1522589848637662e-06, "loss": 0.7745, "step": 1514 }, { "epoch": 1.2751946139280455, "grad_norm": 78.3157730102539, "learning_rate": 3.1501029222870526e-06, "loss": 0.7623, "step": 1515 }, { "epoch": 1.2760361876709447, "grad_norm": 65.5987548828125, "learning_rate": 3.147946340985436e-06, "loss": 0.5745, "step": 1516 }, { "epoch": 1.2768777614138438, "grad_norm": 56.74808120727539, "learning_rate": 3.145789242679678e-06, "loss": 0.577, "step": 1517 }, { "epoch": 1.277719335156743, "grad_norm": 60.50911331176758, "learning_rate": 3.143631629090954e-06, "loss": 0.6998, "step": 1518 }, { "epoch": 1.2785609088996424, "grad_norm": 70.05846405029297, "learning_rate": 3.1414735019408506e-06, "loss": 0.6781, "step": 1519 }, { "epoch": 1.2794024826425416, "grad_norm": 52.80854797363281, "learning_rate": 3.1393148629513604e-06, "loss": 0.5671, "step": 1520 }, { "epoch": 1.2802440563854407, "grad_norm": 59.13666915893555, "learning_rate": 3.1371557138448895e-06, "loss": 0.6924, "step": 1521 }, { "epoch": 1.28108563012834, "grad_norm": 71.01053619384766, "learning_rate": 3.1349960563442483e-06, "loss": 0.9118, "step": 1522 }, { "epoch": 1.2819272038712393, "grad_norm": 98.56394958496094, "learning_rate": 3.1328358921726525e-06, "loss": 0.6676, "step": 1523 }, { "epoch": 1.2827687776141383, "grad_norm": 70.43609619140625, "learning_rate": 3.1306752230537247e-06, "loss": 0.6809, "step": 1524 }, { "epoch": 1.2836103513570376, "grad_norm": 75.69036865234375, "learning_rate": 3.128514050711487e-06, "loss": 0.8077, "step": 1525 }, { "epoch": 1.2844519250999369, "grad_norm": 73.75038146972656, "learning_rate": 3.1263523768703656e-06, "loss": 0.7001, "step": 1526 }, { "epoch": 1.2852934988428362, "grad_norm": 60.24272537231445, "learning_rate": 3.124190203255186e-06, "loss": 0.65, "step": 1527 }, { "epoch": 1.2861350725857355, "grad_norm": 94.17372131347656, "learning_rate": 3.1220275315911718e-06, "loss": 0.6881, "step": 1528 }, { "epoch": 1.2869766463286345, "grad_norm": 70.98792266845703, "learning_rate": 3.119864363603945e-06, "loss": 0.877, "step": 1529 }, { "epoch": 1.2878182200715338, "grad_norm": 79.87841033935547, "learning_rate": 3.1177007010195225e-06, "loss": 0.7708, "step": 1530 }, { "epoch": 1.2886597938144329, "grad_norm": 64.53260040283203, "learning_rate": 3.1155365455643183e-06, "loss": 0.642, "step": 1531 }, { "epoch": 1.2895013675573321, "grad_norm": 69.79100799560547, "learning_rate": 3.113371898965137e-06, "loss": 0.7633, "step": 1532 }, { "epoch": 1.2903429413002314, "grad_norm": 80.67230987548828, "learning_rate": 3.111206762949175e-06, "loss": 0.8915, "step": 1533 }, { "epoch": 1.2911845150431307, "grad_norm": 77.89718627929688, "learning_rate": 3.1090411392440214e-06, "loss": 0.798, "step": 1534 }, { "epoch": 1.29202608878603, "grad_norm": 90.8302001953125, "learning_rate": 3.1068750295776534e-06, "loss": 0.7016, "step": 1535 }, { "epoch": 1.292867662528929, "grad_norm": 82.23298645019531, "learning_rate": 3.104708435678435e-06, "loss": 0.8851, "step": 1536 }, { "epoch": 1.2937092362718283, "grad_norm": 71.48120880126953, "learning_rate": 3.1025413592751173e-06, "loss": 0.784, "step": 1537 }, { "epoch": 1.2945508100147276, "grad_norm": 82.49555206298828, "learning_rate": 3.1003738020968375e-06, "loss": 0.7891, "step": 1538 }, { "epoch": 1.2953923837576267, "grad_norm": 84.50588989257812, "learning_rate": 3.098205765873115e-06, "loss": 0.7263, "step": 1539 }, { "epoch": 1.296233957500526, "grad_norm": 67.03771209716797, "learning_rate": 3.0960372523338522e-06, "loss": 0.7631, "step": 1540 }, { "epoch": 1.2970755312434252, "grad_norm": 57.258365631103516, "learning_rate": 3.0938682632093314e-06, "loss": 0.6434, "step": 1541 }, { "epoch": 1.2979171049863245, "grad_norm": 97.4935302734375, "learning_rate": 3.091698800230215e-06, "loss": 0.6032, "step": 1542 }, { "epoch": 1.2987586787292236, "grad_norm": 68.4141845703125, "learning_rate": 3.0895288651275435e-06, "loss": 0.7371, "step": 1543 }, { "epoch": 1.2996002524721229, "grad_norm": 54.85114669799805, "learning_rate": 3.0873584596327345e-06, "loss": 0.6731, "step": 1544 }, { "epoch": 1.3004418262150221, "grad_norm": 86.47489929199219, "learning_rate": 3.0851875854775803e-06, "loss": 1.0307, "step": 1545 }, { "epoch": 1.3012833999579212, "grad_norm": 67.76560974121094, "learning_rate": 3.0830162443942473e-06, "loss": 0.8864, "step": 1546 }, { "epoch": 1.3021249737008205, "grad_norm": 78.3753433227539, "learning_rate": 3.080844438115275e-06, "loss": 0.8383, "step": 1547 }, { "epoch": 1.3029665474437198, "grad_norm": 55.671783447265625, "learning_rate": 3.0786721683735726e-06, "loss": 0.5977, "step": 1548 }, { "epoch": 1.303808121186619, "grad_norm": 78.6380615234375, "learning_rate": 3.0764994369024215e-06, "loss": 0.7576, "step": 1549 }, { "epoch": 1.3046496949295183, "grad_norm": 59.780235290527344, "learning_rate": 3.0743262454354683e-06, "loss": 0.6514, "step": 1550 }, { "epoch": 1.3054912686724174, "grad_norm": 58.70412826538086, "learning_rate": 3.0721525957067305e-06, "loss": 0.6283, "step": 1551 }, { "epoch": 1.3063328424153167, "grad_norm": 68.63858795166016, "learning_rate": 3.069978489450588e-06, "loss": 0.7468, "step": 1552 }, { "epoch": 1.307174416158216, "grad_norm": 66.73802947998047, "learning_rate": 3.067803928401786e-06, "loss": 0.7989, "step": 1553 }, { "epoch": 1.308015989901115, "grad_norm": 67.49393463134766, "learning_rate": 3.0656289142954345e-06, "loss": 0.7638, "step": 1554 }, { "epoch": 1.3088575636440143, "grad_norm": 68.90960693359375, "learning_rate": 3.063453448867001e-06, "loss": 0.6757, "step": 1555 }, { "epoch": 1.3096991373869136, "grad_norm": 79.73001861572266, "learning_rate": 3.061277533852317e-06, "loss": 0.8807, "step": 1556 }, { "epoch": 1.3105407111298129, "grad_norm": 76.04195404052734, "learning_rate": 3.0591011709875712e-06, "loss": 0.7572, "step": 1557 }, { "epoch": 1.311382284872712, "grad_norm": 75.8448715209961, "learning_rate": 3.0569243620093084e-06, "loss": 0.62, "step": 1558 }, { "epoch": 1.3122238586156112, "grad_norm": 62.169151306152344, "learning_rate": 3.054747108654431e-06, "loss": 0.7032, "step": 1559 }, { "epoch": 1.3130654323585105, "grad_norm": 59.739234924316406, "learning_rate": 3.0525694126601963e-06, "loss": 0.6344, "step": 1560 }, { "epoch": 1.3139070061014095, "grad_norm": 63.75718307495117, "learning_rate": 3.050391275764213e-06, "loss": 0.5988, "step": 1561 }, { "epoch": 1.3147485798443088, "grad_norm": 94.91593170166016, "learning_rate": 3.0482126997044425e-06, "loss": 0.7454, "step": 1562 }, { "epoch": 1.315590153587208, "grad_norm": 81.72116088867188, "learning_rate": 3.046033686219198e-06, "loss": 0.7939, "step": 1563 }, { "epoch": 1.3164317273301074, "grad_norm": 72.60472869873047, "learning_rate": 3.043854237047139e-06, "loss": 0.9237, "step": 1564 }, { "epoch": 1.3172733010730064, "grad_norm": 64.98355102539062, "learning_rate": 3.0416743539272754e-06, "loss": 0.7229, "step": 1565 }, { "epoch": 1.3181148748159057, "grad_norm": 69.29059600830078, "learning_rate": 3.0394940385989603e-06, "loss": 0.6549, "step": 1566 }, { "epoch": 1.318956448558805, "grad_norm": 77.21969604492188, "learning_rate": 3.037313292801895e-06, "loss": 0.8292, "step": 1567 }, { "epoch": 1.319798022301704, "grad_norm": 89.2873306274414, "learning_rate": 3.0351321182761224e-06, "loss": 0.8574, "step": 1568 }, { "epoch": 1.3206395960446033, "grad_norm": 89.90650177001953, "learning_rate": 3.032950516762027e-06, "loss": 0.9938, "step": 1569 }, { "epoch": 1.3214811697875026, "grad_norm": 74.41616821289062, "learning_rate": 3.0307684900003355e-06, "loss": 0.7855, "step": 1570 }, { "epoch": 1.322322743530402, "grad_norm": 80.36529541015625, "learning_rate": 3.0285860397321126e-06, "loss": 0.9729, "step": 1571 }, { "epoch": 1.3231643172733012, "grad_norm": 63.00858688354492, "learning_rate": 3.0264031676987616e-06, "loss": 0.6507, "step": 1572 }, { "epoch": 1.3240058910162003, "grad_norm": 73.80546569824219, "learning_rate": 3.0242198756420234e-06, "loss": 0.7323, "step": 1573 }, { "epoch": 1.3248474647590995, "grad_norm": 71.82361602783203, "learning_rate": 3.0220361653039707e-06, "loss": 0.7676, "step": 1574 }, { "epoch": 1.3256890385019988, "grad_norm": 67.37881469726562, "learning_rate": 3.0198520384270147e-06, "loss": 0.6754, "step": 1575 }, { "epoch": 1.3265306122448979, "grad_norm": 73.1670913696289, "learning_rate": 3.017667496753894e-06, "loss": 0.8173, "step": 1576 }, { "epoch": 1.3273721859877972, "grad_norm": 78.4403076171875, "learning_rate": 3.0154825420276816e-06, "loss": 0.8628, "step": 1577 }, { "epoch": 1.3282137597306964, "grad_norm": 70.75596618652344, "learning_rate": 3.013297175991779e-06, "loss": 0.8315, "step": 1578 }, { "epoch": 1.3290553334735957, "grad_norm": 58.30497741699219, "learning_rate": 3.011111400389916e-06, "loss": 0.6339, "step": 1579 }, { "epoch": 1.3298969072164948, "grad_norm": 81.0142593383789, "learning_rate": 3.008925216966149e-06, "loss": 0.678, "step": 1580 }, { "epoch": 1.330738480959394, "grad_norm": 88.39791107177734, "learning_rate": 3.00673862746486e-06, "loss": 0.8672, "step": 1581 }, { "epoch": 1.3315800547022933, "grad_norm": 87.34832763671875, "learning_rate": 3.0045516336307547e-06, "loss": 0.8872, "step": 1582 }, { "epoch": 1.3324216284451924, "grad_norm": 75.19876098632812, "learning_rate": 3.0023642372088614e-06, "loss": 0.8358, "step": 1583 }, { "epoch": 1.3332632021880917, "grad_norm": 69.73424530029297, "learning_rate": 3.0001764399445305e-06, "loss": 0.77, "step": 1584 }, { "epoch": 1.334104775930991, "grad_norm": 59.559471130371094, "learning_rate": 2.997988243583432e-06, "loss": 0.5481, "step": 1585 }, { "epoch": 1.3349463496738903, "grad_norm": 72.986328125, "learning_rate": 2.9957996498715525e-06, "loss": 0.7308, "step": 1586 }, { "epoch": 1.3357879234167893, "grad_norm": 70.01398468017578, "learning_rate": 2.9936106605551985e-06, "loss": 0.7372, "step": 1587 }, { "epoch": 1.3366294971596886, "grad_norm": 66.65174865722656, "learning_rate": 2.9914212773809898e-06, "loss": 0.7694, "step": 1588 }, { "epoch": 1.3374710709025879, "grad_norm": 85.71355438232422, "learning_rate": 2.9892315020958624e-06, "loss": 0.874, "step": 1589 }, { "epoch": 1.338312644645487, "grad_norm": 68.76534271240234, "learning_rate": 2.9870413364470635e-06, "loss": 0.736, "step": 1590 }, { "epoch": 1.3391542183883862, "grad_norm": 84.54867553710938, "learning_rate": 2.984850782182153e-06, "loss": 0.7407, "step": 1591 }, { "epoch": 1.3399957921312855, "grad_norm": 63.563621520996094, "learning_rate": 2.9826598410489998e-06, "loss": 0.6142, "step": 1592 }, { "epoch": 1.3408373658741848, "grad_norm": 68.36559295654297, "learning_rate": 2.9804685147957824e-06, "loss": 0.6722, "step": 1593 }, { "epoch": 1.341678939617084, "grad_norm": 56.75374984741211, "learning_rate": 2.9782768051709867e-06, "loss": 0.6059, "step": 1594 }, { "epoch": 1.3425205133599831, "grad_norm": 66.35613250732422, "learning_rate": 2.9760847139234024e-06, "loss": 0.9102, "step": 1595 }, { "epoch": 1.3433620871028824, "grad_norm": 74.88129425048828, "learning_rate": 2.973892242802127e-06, "loss": 0.5021, "step": 1596 }, { "epoch": 1.3442036608457817, "grad_norm": 54.18392562866211, "learning_rate": 2.971699393556559e-06, "loss": 0.4953, "step": 1597 }, { "epoch": 1.3450452345886807, "grad_norm": 165.30419921875, "learning_rate": 2.969506167936399e-06, "loss": 0.4721, "step": 1598 }, { "epoch": 1.34588680833158, "grad_norm": 84.15676879882812, "learning_rate": 2.9673125676916476e-06, "loss": 0.9234, "step": 1599 }, { "epoch": 1.3467283820744793, "grad_norm": 60.92980194091797, "learning_rate": 2.9651185945726054e-06, "loss": 0.6403, "step": 1600 }, { "epoch": 1.3475699558173786, "grad_norm": 52.536869049072266, "learning_rate": 2.9629242503298695e-06, "loss": 0.6488, "step": 1601 }, { "epoch": 1.3484115295602777, "grad_norm": 52.28572463989258, "learning_rate": 2.9607295367143337e-06, "loss": 0.6984, "step": 1602 }, { "epoch": 1.349253103303177, "grad_norm": 73.1734848022461, "learning_rate": 2.958534455477185e-06, "loss": 0.8367, "step": 1603 }, { "epoch": 1.3500946770460762, "grad_norm": 73.1240234375, "learning_rate": 2.956339008369906e-06, "loss": 0.7781, "step": 1604 }, { "epoch": 1.3509362507889753, "grad_norm": 63.2278938293457, "learning_rate": 2.9541431971442706e-06, "loss": 0.7579, "step": 1605 }, { "epoch": 1.3517778245318746, "grad_norm": 71.2304458618164, "learning_rate": 2.951947023552343e-06, "loss": 0.8171, "step": 1606 }, { "epoch": 1.3526193982747738, "grad_norm": 73.63424682617188, "learning_rate": 2.949750489346475e-06, "loss": 0.7502, "step": 1607 }, { "epoch": 1.3534609720176731, "grad_norm": 66.96131896972656, "learning_rate": 2.947553596279309e-06, "loss": 0.8436, "step": 1608 }, { "epoch": 1.3543025457605724, "grad_norm": 61.75212097167969, "learning_rate": 2.945356346103771e-06, "loss": 0.7046, "step": 1609 }, { "epoch": 1.3551441195034715, "grad_norm": 70.67098999023438, "learning_rate": 2.943158740573074e-06, "loss": 0.7237, "step": 1610 }, { "epoch": 1.3559856932463707, "grad_norm": 75.321533203125, "learning_rate": 2.9409607814407136e-06, "loss": 0.8022, "step": 1611 }, { "epoch": 1.3568272669892698, "grad_norm": 66.69032287597656, "learning_rate": 2.938762470460468e-06, "loss": 0.7304, "step": 1612 }, { "epoch": 1.357668840732169, "grad_norm": 70.29718780517578, "learning_rate": 2.9365638093863947e-06, "loss": 0.8823, "step": 1613 }, { "epoch": 1.3585104144750684, "grad_norm": 61.76929473876953, "learning_rate": 2.934364799972833e-06, "loss": 0.8264, "step": 1614 }, { "epoch": 1.3593519882179677, "grad_norm": 102.89973449707031, "learning_rate": 2.9321654439743985e-06, "loss": 0.6024, "step": 1615 }, { "epoch": 1.360193561960867, "grad_norm": 81.0641860961914, "learning_rate": 2.9299657431459834e-06, "loss": 0.9365, "step": 1616 }, { "epoch": 1.361035135703766, "grad_norm": 75.15452575683594, "learning_rate": 2.9277656992427558e-06, "loss": 0.6629, "step": 1617 }, { "epoch": 1.3618767094466653, "grad_norm": 103.4858169555664, "learning_rate": 2.925565314020156e-06, "loss": 0.7863, "step": 1618 }, { "epoch": 1.3627182831895646, "grad_norm": 72.90863800048828, "learning_rate": 2.9233645892338997e-06, "loss": 0.6465, "step": 1619 }, { "epoch": 1.3635598569324636, "grad_norm": 63.42647933959961, "learning_rate": 2.921163526639969e-06, "loss": 0.7184, "step": 1620 }, { "epoch": 1.364401430675363, "grad_norm": 71.44855499267578, "learning_rate": 2.9189621279946203e-06, "loss": 0.8488, "step": 1621 }, { "epoch": 1.3652430044182622, "grad_norm": 68.80278015136719, "learning_rate": 2.916760395054374e-06, "loss": 0.7289, "step": 1622 }, { "epoch": 1.3660845781611615, "grad_norm": 77.66200256347656, "learning_rate": 2.914558329576021e-06, "loss": 0.932, "step": 1623 }, { "epoch": 1.3669261519040605, "grad_norm": 97.39385223388672, "learning_rate": 2.9123559333166145e-06, "loss": 0.76, "step": 1624 }, { "epoch": 1.3677677256469598, "grad_norm": 87.82623291015625, "learning_rate": 2.9101532080334732e-06, "loss": 0.9349, "step": 1625 }, { "epoch": 1.368609299389859, "grad_norm": 71.79322052001953, "learning_rate": 2.907950155484178e-06, "loss": 0.75, "step": 1626 }, { "epoch": 1.3694508731327582, "grad_norm": 67.104736328125, "learning_rate": 2.9057467774265706e-06, "loss": 0.6786, "step": 1627 }, { "epoch": 1.3702924468756574, "grad_norm": 64.74378967285156, "learning_rate": 2.9035430756187532e-06, "loss": 0.7288, "step": 1628 }, { "epoch": 1.3711340206185567, "grad_norm": 71.50525665283203, "learning_rate": 2.901339051819086e-06, "loss": 0.7327, "step": 1629 }, { "epoch": 1.371975594361456, "grad_norm": 63.350547790527344, "learning_rate": 2.899134707786185e-06, "loss": 0.7037, "step": 1630 }, { "epoch": 1.3728171681043553, "grad_norm": 70.74079895019531, "learning_rate": 2.8969300452789235e-06, "loss": 0.7773, "step": 1631 }, { "epoch": 1.3736587418472543, "grad_norm": 68.98648071289062, "learning_rate": 2.894725066056428e-06, "loss": 0.5523, "step": 1632 }, { "epoch": 1.3745003155901536, "grad_norm": 89.67284393310547, "learning_rate": 2.8925197718780772e-06, "loss": 0.9865, "step": 1633 }, { "epoch": 1.3753418893330527, "grad_norm": 64.90214538574219, "learning_rate": 2.890314164503503e-06, "loss": 0.7149, "step": 1634 }, { "epoch": 1.376183463075952, "grad_norm": 65.8893051147461, "learning_rate": 2.888108245692585e-06, "loss": 0.852, "step": 1635 }, { "epoch": 1.3770250368188512, "grad_norm": 73.79243469238281, "learning_rate": 2.8859020172054527e-06, "loss": 0.7735, "step": 1636 }, { "epoch": 1.3778666105617505, "grad_norm": 64.22339630126953, "learning_rate": 2.883695480802481e-06, "loss": 0.7819, "step": 1637 }, { "epoch": 1.3787081843046498, "grad_norm": 64.70500946044922, "learning_rate": 2.881488638244293e-06, "loss": 0.6551, "step": 1638 }, { "epoch": 1.3795497580475489, "grad_norm": 91.891845703125, "learning_rate": 2.8792814912917546e-06, "loss": 0.9065, "step": 1639 }, { "epoch": 1.3803913317904482, "grad_norm": 85.52303314208984, "learning_rate": 2.877074041705974e-06, "loss": 0.8086, "step": 1640 }, { "epoch": 1.3812329055333474, "grad_norm": 45.623069763183594, "learning_rate": 2.8748662912483026e-06, "loss": 0.5107, "step": 1641 }, { "epoch": 1.3820744792762465, "grad_norm": 79.50008392333984, "learning_rate": 2.8726582416803306e-06, "loss": 0.8424, "step": 1642 }, { "epoch": 1.3829160530191458, "grad_norm": 57.23161697387695, "learning_rate": 2.870449894763887e-06, "loss": 0.732, "step": 1643 }, { "epoch": 1.383757626762045, "grad_norm": 138.73167419433594, "learning_rate": 2.8682412522610382e-06, "loss": 0.6207, "step": 1644 }, { "epoch": 1.3845992005049443, "grad_norm": 64.35865783691406, "learning_rate": 2.8660323159340874e-06, "loss": 0.6844, "step": 1645 }, { "epoch": 1.3854407742478434, "grad_norm": 80.82450103759766, "learning_rate": 2.8638230875455697e-06, "loss": 0.7879, "step": 1646 }, { "epoch": 1.3862823479907427, "grad_norm": 59.76946258544922, "learning_rate": 2.8616135688582565e-06, "loss": 0.5935, "step": 1647 }, { "epoch": 1.387123921733642, "grad_norm": 79.74017333984375, "learning_rate": 2.8594037616351477e-06, "loss": 0.9341, "step": 1648 }, { "epoch": 1.387965495476541, "grad_norm": 65.20237731933594, "learning_rate": 2.8571936676394755e-06, "loss": 0.7487, "step": 1649 }, { "epoch": 1.3888070692194403, "grad_norm": 67.59424591064453, "learning_rate": 2.854983288634702e-06, "loss": 0.8008, "step": 1650 }, { "epoch": 1.3896486429623396, "grad_norm": 57.513370513916016, "learning_rate": 2.8527726263845123e-06, "loss": 0.7612, "step": 1651 }, { "epoch": 1.3904902167052389, "grad_norm": 60.59592819213867, "learning_rate": 2.8505616826528214e-06, "loss": 0.6841, "step": 1652 }, { "epoch": 1.3913317904481382, "grad_norm": 75.38861083984375, "learning_rate": 2.8483504592037687e-06, "loss": 0.8172, "step": 1653 }, { "epoch": 1.3921733641910372, "grad_norm": 68.17393493652344, "learning_rate": 2.8461389578017145e-06, "loss": 0.734, "step": 1654 }, { "epoch": 1.3930149379339365, "grad_norm": 67.03765106201172, "learning_rate": 2.8439271802112427e-06, "loss": 0.5979, "step": 1655 }, { "epoch": 1.3938565116768356, "grad_norm": 82.4088363647461, "learning_rate": 2.8417151281971576e-06, "loss": 0.6884, "step": 1656 }, { "epoch": 1.3946980854197348, "grad_norm": 63.67273712158203, "learning_rate": 2.839502803524482e-06, "loss": 0.7441, "step": 1657 }, { "epoch": 1.3955396591626341, "grad_norm": 77.49272918701172, "learning_rate": 2.837290207958455e-06, "loss": 0.7249, "step": 1658 }, { "epoch": 1.3963812329055334, "grad_norm": 59.528076171875, "learning_rate": 2.8350773432645346e-06, "loss": 0.6164, "step": 1659 }, { "epoch": 1.3972228066484327, "grad_norm": 60.476280212402344, "learning_rate": 2.8328642112083916e-06, "loss": 0.6363, "step": 1660 }, { "epoch": 1.3980643803913317, "grad_norm": 68.28182220458984, "learning_rate": 2.8306508135559095e-06, "loss": 0.6591, "step": 1661 }, { "epoch": 1.398905954134231, "grad_norm": 74.33509826660156, "learning_rate": 2.8284371520731857e-06, "loss": 0.8051, "step": 1662 }, { "epoch": 1.3997475278771303, "grad_norm": 193.60691833496094, "learning_rate": 2.8262232285265285e-06, "loss": 0.7696, "step": 1663 }, { "epoch": 1.4005891016200294, "grad_norm": 53.84804916381836, "learning_rate": 2.8240090446824512e-06, "loss": 0.5555, "step": 1664 }, { "epoch": 1.4014306753629286, "grad_norm": 62.28761291503906, "learning_rate": 2.821794602307679e-06, "loss": 0.5864, "step": 1665 }, { "epoch": 1.402272249105828, "grad_norm": 78.90672302246094, "learning_rate": 2.8195799031691424e-06, "loss": 0.8775, "step": 1666 }, { "epoch": 1.4031138228487272, "grad_norm": 54.90179443359375, "learning_rate": 2.817364949033976e-06, "loss": 0.5934, "step": 1667 }, { "epoch": 1.4039553965916263, "grad_norm": 78.57133483886719, "learning_rate": 2.8151497416695184e-06, "loss": 0.6741, "step": 1668 }, { "epoch": 1.4047969703345256, "grad_norm": 72.7953872680664, "learning_rate": 2.8129342828433092e-06, "loss": 0.8454, "step": 1669 }, { "epoch": 1.4056385440774248, "grad_norm": 63.27824401855469, "learning_rate": 2.8107185743230904e-06, "loss": 0.7299, "step": 1670 }, { "epoch": 1.406480117820324, "grad_norm": 65.07614135742188, "learning_rate": 2.8085026178768014e-06, "loss": 0.5279, "step": 1671 }, { "epoch": 1.4073216915632232, "grad_norm": 71.12516784667969, "learning_rate": 2.8062864152725816e-06, "loss": 0.8251, "step": 1672 }, { "epoch": 1.4081632653061225, "grad_norm": 83.48336029052734, "learning_rate": 2.8040699682787648e-06, "loss": 0.9559, "step": 1673 }, { "epoch": 1.4090048390490217, "grad_norm": 72.82669830322266, "learning_rate": 2.8018532786638806e-06, "loss": 0.7378, "step": 1674 }, { "epoch": 1.409846412791921, "grad_norm": 65.47310638427734, "learning_rate": 2.799636348196653e-06, "loss": 0.715, "step": 1675 }, { "epoch": 1.41068798653482, "grad_norm": 60.22334671020508, "learning_rate": 2.7974191786459965e-06, "loss": 0.6188, "step": 1676 }, { "epoch": 1.4115295602777194, "grad_norm": 71.9939193725586, "learning_rate": 2.7952017717810174e-06, "loss": 0.5973, "step": 1677 }, { "epoch": 1.4123711340206184, "grad_norm": 94.24855041503906, "learning_rate": 2.7929841293710116e-06, "loss": 0.9587, "step": 1678 }, { "epoch": 1.4132127077635177, "grad_norm": 64.97252655029297, "learning_rate": 2.790766253185462e-06, "loss": 0.7037, "step": 1679 }, { "epoch": 1.414054281506417, "grad_norm": 95.64558410644531, "learning_rate": 2.7885481449940383e-06, "loss": 1.0062, "step": 1680 }, { "epoch": 1.4148958552493163, "grad_norm": 61.63941192626953, "learning_rate": 2.7863298065665962e-06, "loss": 0.7186, "step": 1681 }, { "epoch": 1.4157374289922156, "grad_norm": 71.3286361694336, "learning_rate": 2.7841112396731747e-06, "loss": 0.8337, "step": 1682 }, { "epoch": 1.4165790027351146, "grad_norm": 67.52711486816406, "learning_rate": 2.7818924460839934e-06, "loss": 0.7435, "step": 1683 }, { "epoch": 1.417420576478014, "grad_norm": 65.43138122558594, "learning_rate": 2.7796734275694557e-06, "loss": 0.7566, "step": 1684 }, { "epoch": 1.4182621502209132, "grad_norm": 74.34310913085938, "learning_rate": 2.7774541859001434e-06, "loss": 0.7623, "step": 1685 }, { "epoch": 1.4191037239638122, "grad_norm": 70.51815795898438, "learning_rate": 2.7752347228468145e-06, "loss": 0.713, "step": 1686 }, { "epoch": 1.4199452977067115, "grad_norm": 86.21179962158203, "learning_rate": 2.773015040180406e-06, "loss": 0.8835, "step": 1687 }, { "epoch": 1.4207868714496108, "grad_norm": 79.187255859375, "learning_rate": 2.7707951396720296e-06, "loss": 0.7705, "step": 1688 }, { "epoch": 1.42162844519251, "grad_norm": 99.74498748779297, "learning_rate": 2.7685750230929697e-06, "loss": 0.9959, "step": 1689 }, { "epoch": 1.4224700189354091, "grad_norm": 72.84029388427734, "learning_rate": 2.7663546922146846e-06, "loss": 0.7988, "step": 1690 }, { "epoch": 1.4233115926783084, "grad_norm": 64.78691864013672, "learning_rate": 2.764134148808802e-06, "loss": 0.6735, "step": 1691 }, { "epoch": 1.4241531664212077, "grad_norm": 66.76640319824219, "learning_rate": 2.7619133946471203e-06, "loss": 0.7184, "step": 1692 }, { "epoch": 1.4249947401641068, "grad_norm": 67.89680480957031, "learning_rate": 2.7596924315016068e-06, "loss": 0.6836, "step": 1693 }, { "epoch": 1.425836313907006, "grad_norm": 53.331573486328125, "learning_rate": 2.757471261144394e-06, "loss": 0.6315, "step": 1694 }, { "epoch": 1.4266778876499053, "grad_norm": 78.04473876953125, "learning_rate": 2.7552498853477798e-06, "loss": 0.8534, "step": 1695 }, { "epoch": 1.4275194613928046, "grad_norm": 66.13089752197266, "learning_rate": 2.753028305884227e-06, "loss": 0.8333, "step": 1696 }, { "epoch": 1.428361035135704, "grad_norm": 66.52342987060547, "learning_rate": 2.7508065245263598e-06, "loss": 0.7317, "step": 1697 }, { "epoch": 1.429202608878603, "grad_norm": 68.26254272460938, "learning_rate": 2.7485845430469648e-06, "loss": 0.7913, "step": 1698 }, { "epoch": 1.4300441826215022, "grad_norm": 91.65166473388672, "learning_rate": 2.7463623632189877e-06, "loss": 0.7921, "step": 1699 }, { "epoch": 1.4308857563644013, "grad_norm": 56.97581481933594, "learning_rate": 2.7441399868155328e-06, "loss": 0.722, "step": 1700 }, { "epoch": 1.4317273301073006, "grad_norm": 56.74054718017578, "learning_rate": 2.741917415609859e-06, "loss": 0.521, "step": 1701 }, { "epoch": 1.4325689038501999, "grad_norm": 59.926578521728516, "learning_rate": 2.739694651375385e-06, "loss": 0.6443, "step": 1702 }, { "epoch": 1.4334104775930991, "grad_norm": 60.563411712646484, "learning_rate": 2.737471695885679e-06, "loss": 0.6518, "step": 1703 }, { "epoch": 1.4342520513359984, "grad_norm": 81.89253997802734, "learning_rate": 2.735248550914464e-06, "loss": 0.8534, "step": 1704 }, { "epoch": 1.4350936250788975, "grad_norm": 57.36185836791992, "learning_rate": 2.733025218235616e-06, "loss": 0.6834, "step": 1705 }, { "epoch": 1.4359351988217968, "grad_norm": 66.09939575195312, "learning_rate": 2.7308016996231563e-06, "loss": 0.9084, "step": 1706 }, { "epoch": 1.436776772564696, "grad_norm": 75.37176513671875, "learning_rate": 2.728577996851258e-06, "loss": 0.85, "step": 1707 }, { "epoch": 1.437618346307595, "grad_norm": 68.36734008789062, "learning_rate": 2.7263541116942404e-06, "loss": 0.7047, "step": 1708 }, { "epoch": 1.4384599200504944, "grad_norm": 78.23722076416016, "learning_rate": 2.7241300459265676e-06, "loss": 0.8309, "step": 1709 }, { "epoch": 1.4393014937933937, "grad_norm": 75.7884521484375, "learning_rate": 2.7219058013228478e-06, "loss": 0.8572, "step": 1710 }, { "epoch": 1.440143067536293, "grad_norm": 56.12018966674805, "learning_rate": 2.719681379657834e-06, "loss": 0.6193, "step": 1711 }, { "epoch": 1.440984641279192, "grad_norm": 65.65968322753906, "learning_rate": 2.7174567827064174e-06, "loss": 0.5725, "step": 1712 }, { "epoch": 1.4418262150220913, "grad_norm": 71.11070251464844, "learning_rate": 2.7152320122436316e-06, "loss": 0.8121, "step": 1713 }, { "epoch": 1.4426677887649906, "grad_norm": 66.6751937866211, "learning_rate": 2.7130070700446464e-06, "loss": 0.6744, "step": 1714 }, { "epoch": 1.4435093625078896, "grad_norm": 54.40220260620117, "learning_rate": 2.7107819578847704e-06, "loss": 0.6712, "step": 1715 }, { "epoch": 1.444350936250789, "grad_norm": 77.96082305908203, "learning_rate": 2.708556677539448e-06, "loss": 0.7327, "step": 1716 }, { "epoch": 1.4451925099936882, "grad_norm": 82.51429748535156, "learning_rate": 2.7063312307842553e-06, "loss": 0.8126, "step": 1717 }, { "epoch": 1.4460340837365875, "grad_norm": 73.80522155761719, "learning_rate": 2.704105619394905e-06, "loss": 0.8241, "step": 1718 }, { "epoch": 1.4468756574794868, "grad_norm": 72.76254272460938, "learning_rate": 2.701879845147237e-06, "loss": 0.8046, "step": 1719 }, { "epoch": 1.4477172312223858, "grad_norm": 97.87123107910156, "learning_rate": 2.6996539098172247e-06, "loss": 1.0595, "step": 1720 }, { "epoch": 1.448558804965285, "grad_norm": 66.60394287109375, "learning_rate": 2.6974278151809686e-06, "loss": 0.7177, "step": 1721 }, { "epoch": 1.4494003787081842, "grad_norm": 66.91735076904297, "learning_rate": 2.6952015630146955e-06, "loss": 0.8048, "step": 1722 }, { "epoch": 1.4502419524510834, "grad_norm": 64.95494079589844, "learning_rate": 2.692975155094758e-06, "loss": 0.8128, "step": 1723 }, { "epoch": 1.4510835261939827, "grad_norm": 60.2796630859375, "learning_rate": 2.690748593197635e-06, "loss": 0.6849, "step": 1724 }, { "epoch": 1.451925099936882, "grad_norm": 63.44791793823242, "learning_rate": 2.688521879099926e-06, "loss": 0.6255, "step": 1725 }, { "epoch": 1.4527666736797813, "grad_norm": 69.22535705566406, "learning_rate": 2.686295014578354e-06, "loss": 0.7813, "step": 1726 }, { "epoch": 1.4536082474226804, "grad_norm": 50.263423919677734, "learning_rate": 2.684068001409759e-06, "loss": 0.7111, "step": 1727 }, { "epoch": 1.4544498211655796, "grad_norm": 63.202091217041016, "learning_rate": 2.6818408413711035e-06, "loss": 0.6883, "step": 1728 }, { "epoch": 1.455291394908479, "grad_norm": 58.402469635009766, "learning_rate": 2.6796135362394635e-06, "loss": 0.6139, "step": 1729 }, { "epoch": 1.456132968651378, "grad_norm": 71.20877075195312, "learning_rate": 2.6773860877920334e-06, "loss": 0.8273, "step": 1730 }, { "epoch": 1.4569745423942773, "grad_norm": 51.11784744262695, "learning_rate": 2.6751584978061208e-06, "loss": 0.5177, "step": 1731 }, { "epoch": 1.4578161161371765, "grad_norm": 71.94078063964844, "learning_rate": 2.6729307680591454e-06, "loss": 0.8767, "step": 1732 }, { "epoch": 1.4586576898800758, "grad_norm": 82.49336242675781, "learning_rate": 2.6707029003286416e-06, "loss": 1.0348, "step": 1733 }, { "epoch": 1.4594992636229749, "grad_norm": 71.91825103759766, "learning_rate": 2.6684748963922503e-06, "loss": 0.8245, "step": 1734 }, { "epoch": 1.4603408373658742, "grad_norm": 67.69684600830078, "learning_rate": 2.6662467580277217e-06, "loss": 0.8122, "step": 1735 }, { "epoch": 1.4611824111087734, "grad_norm": 60.152103424072266, "learning_rate": 2.6640184870129156e-06, "loss": 0.7082, "step": 1736 }, { "epoch": 1.4620239848516725, "grad_norm": 83.35700988769531, "learning_rate": 2.661790085125795e-06, "loss": 0.8148, "step": 1737 }, { "epoch": 1.4628655585945718, "grad_norm": 83.12397003173828, "learning_rate": 2.6595615541444297e-06, "loss": 0.8294, "step": 1738 }, { "epoch": 1.463707132337471, "grad_norm": 48.19567108154297, "learning_rate": 2.65733289584699e-06, "loss": 0.5771, "step": 1739 }, { "epoch": 1.4645487060803704, "grad_norm": 68.2626724243164, "learning_rate": 2.65510411201175e-06, "loss": 0.7642, "step": 1740 }, { "epoch": 1.4653902798232696, "grad_norm": 71.0654067993164, "learning_rate": 2.6528752044170823e-06, "loss": 0.6847, "step": 1741 }, { "epoch": 1.4662318535661687, "grad_norm": 59.572898864746094, "learning_rate": 2.650646174841459e-06, "loss": 0.8298, "step": 1742 }, { "epoch": 1.467073427309068, "grad_norm": 85.9233627319336, "learning_rate": 2.64841702506345e-06, "loss": 0.9816, "step": 1743 }, { "epoch": 1.4679150010519673, "grad_norm": 62.2779426574707, "learning_rate": 2.6461877568617195e-06, "loss": 0.6507, "step": 1744 }, { "epoch": 1.4687565747948663, "grad_norm": 62.837310791015625, "learning_rate": 2.6439583720150287e-06, "loss": 0.6898, "step": 1745 }, { "epoch": 1.4695981485377656, "grad_norm": 73.21057891845703, "learning_rate": 2.6417288723022296e-06, "loss": 0.7929, "step": 1746 }, { "epoch": 1.4704397222806649, "grad_norm": 78.05528259277344, "learning_rate": 2.6394992595022667e-06, "loss": 0.8539, "step": 1747 }, { "epoch": 1.4712812960235642, "grad_norm": 71.5974349975586, "learning_rate": 2.637269535394175e-06, "loss": 0.8032, "step": 1748 }, { "epoch": 1.4721228697664632, "grad_norm": 88.62098693847656, "learning_rate": 2.635039701757078e-06, "loss": 1.0008, "step": 1749 }, { "epoch": 1.4729644435093625, "grad_norm": 75.0531997680664, "learning_rate": 2.6328097603701875e-06, "loss": 0.8675, "step": 1750 }, { "epoch": 1.4738060172522618, "grad_norm": 53.15607452392578, "learning_rate": 2.630579713012799e-06, "loss": 0.5601, "step": 1751 }, { "epoch": 1.4746475909951609, "grad_norm": 60.73322296142578, "learning_rate": 2.6283495614642943e-06, "loss": 0.7335, "step": 1752 }, { "epoch": 1.4754891647380601, "grad_norm": 70.41018676757812, "learning_rate": 2.6261193075041388e-06, "loss": 0.6636, "step": 1753 }, { "epoch": 1.4763307384809594, "grad_norm": 58.46347427368164, "learning_rate": 2.6238889529118777e-06, "loss": 0.6997, "step": 1754 }, { "epoch": 1.4771723122238587, "grad_norm": 73.29708099365234, "learning_rate": 2.6216584994671384e-06, "loss": 0.8565, "step": 1755 }, { "epoch": 1.4780138859667578, "grad_norm": 64.69960021972656, "learning_rate": 2.6194279489496264e-06, "loss": 0.5535, "step": 1756 }, { "epoch": 1.478855459709657, "grad_norm": 58.43681716918945, "learning_rate": 2.617197303139124e-06, "loss": 0.7176, "step": 1757 }, { "epoch": 1.4796970334525563, "grad_norm": 73.35099029541016, "learning_rate": 2.614966563815491e-06, "loss": 0.7436, "step": 1758 }, { "epoch": 1.4805386071954554, "grad_norm": 86.46379852294922, "learning_rate": 2.6127357327586607e-06, "loss": 0.9225, "step": 1759 }, { "epoch": 1.4813801809383547, "grad_norm": 66.31956481933594, "learning_rate": 2.610504811748639e-06, "loss": 0.7184, "step": 1760 }, { "epoch": 1.482221754681254, "grad_norm": 62.675777435302734, "learning_rate": 2.6082738025655064e-06, "loss": 0.6338, "step": 1761 }, { "epoch": 1.4830633284241532, "grad_norm": 60.13311767578125, "learning_rate": 2.6060427069894105e-06, "loss": 0.8495, "step": 1762 }, { "epoch": 1.4839049021670525, "grad_norm": 63.45058822631836, "learning_rate": 2.6038115268005694e-06, "loss": 0.739, "step": 1763 }, { "epoch": 1.4847464759099516, "grad_norm": 61.45596694946289, "learning_rate": 2.601580263779268e-06, "loss": 0.8252, "step": 1764 }, { "epoch": 1.4855880496528509, "grad_norm": 84.68161010742188, "learning_rate": 2.599348919705859e-06, "loss": 0.7826, "step": 1765 }, { "epoch": 1.4864296233957501, "grad_norm": 59.868900299072266, "learning_rate": 2.597117496360758e-06, "loss": 0.5912, "step": 1766 }, { "epoch": 1.4872711971386492, "grad_norm": 70.5306167602539, "learning_rate": 2.594885995524444e-06, "loss": 0.7172, "step": 1767 }, { "epoch": 1.4881127708815485, "grad_norm": 71.05335998535156, "learning_rate": 2.5926544189774584e-06, "loss": 0.6855, "step": 1768 }, { "epoch": 1.4889543446244478, "grad_norm": 52.01316833496094, "learning_rate": 2.5904227685004023e-06, "loss": 0.5923, "step": 1769 }, { "epoch": 1.489795918367347, "grad_norm": 65.62788391113281, "learning_rate": 2.5881910458739373e-06, "loss": 0.6316, "step": 1770 }, { "epoch": 1.490637492110246, "grad_norm": 84.27230834960938, "learning_rate": 2.5859592528787813e-06, "loss": 0.8697, "step": 1771 }, { "epoch": 1.4914790658531454, "grad_norm": 70.32630157470703, "learning_rate": 2.5837273912957084e-06, "loss": 0.7927, "step": 1772 }, { "epoch": 1.4923206395960447, "grad_norm": 78.77819061279297, "learning_rate": 2.5814954629055477e-06, "loss": 0.7227, "step": 1773 }, { "epoch": 1.4931622133389437, "grad_norm": 56.90717315673828, "learning_rate": 2.5792634694891822e-06, "loss": 0.5666, "step": 1774 }, { "epoch": 1.494003787081843, "grad_norm": 82.15579986572266, "learning_rate": 2.5770314128275453e-06, "loss": 0.728, "step": 1775 }, { "epoch": 1.4948453608247423, "grad_norm": 57.27631759643555, "learning_rate": 2.574799294701622e-06, "loss": 0.7312, "step": 1776 }, { "epoch": 1.4956869345676416, "grad_norm": 60.32022476196289, "learning_rate": 2.572567116892447e-06, "loss": 0.5993, "step": 1777 }, { "epoch": 1.4965285083105406, "grad_norm": 74.37557983398438, "learning_rate": 2.5703348811811003e-06, "loss": 0.7395, "step": 1778 }, { "epoch": 1.49737008205344, "grad_norm": 73.13529205322266, "learning_rate": 2.56810258934871e-06, "loss": 0.8346, "step": 1779 }, { "epoch": 1.4982116557963392, "grad_norm": 81.80269622802734, "learning_rate": 2.565870243176448e-06, "loss": 0.8701, "step": 1780 }, { "epoch": 1.4990532295392383, "grad_norm": 52.8936767578125, "learning_rate": 2.5636378444455312e-06, "loss": 0.5246, "step": 1781 }, { "epoch": 1.4998948032821375, "grad_norm": 69.68518829345703, "learning_rate": 2.5614053949372174e-06, "loss": 0.8555, "step": 1782 }, { "epoch": 1.5007363770250368, "grad_norm": 71.54790496826172, "learning_rate": 2.559172896432803e-06, "loss": 0.6966, "step": 1783 }, { "epoch": 1.501577950767936, "grad_norm": 73.0295181274414, "learning_rate": 2.5569403507136258e-06, "loss": 0.72, "step": 1784 }, { "epoch": 1.5024195245108354, "grad_norm": 63.13631820678711, "learning_rate": 2.554707759561062e-06, "loss": 0.681, "step": 1785 }, { "epoch": 1.5032610982537344, "grad_norm": 64.40337371826172, "learning_rate": 2.5524751247565216e-06, "loss": 0.6048, "step": 1786 }, { "epoch": 1.5041026719966337, "grad_norm": 66.20853424072266, "learning_rate": 2.5502424480814515e-06, "loss": 0.7552, "step": 1787 }, { "epoch": 1.5049442457395328, "grad_norm": 65.15352630615234, "learning_rate": 2.54800973131733e-06, "loss": 0.698, "step": 1788 }, { "epoch": 1.505785819482432, "grad_norm": 68.215087890625, "learning_rate": 2.5457769762456702e-06, "loss": 0.6937, "step": 1789 }, { "epoch": 1.5066273932253313, "grad_norm": 64.3287582397461, "learning_rate": 2.5435441846480125e-06, "loss": 0.7626, "step": 1790 }, { "epoch": 1.5074689669682306, "grad_norm": 84.3650131225586, "learning_rate": 2.541311358305929e-06, "loss": 0.735, "step": 1791 }, { "epoch": 1.50831054071113, "grad_norm": 59.81379318237305, "learning_rate": 2.539078499001018e-06, "loss": 0.659, "step": 1792 }, { "epoch": 1.5091521144540292, "grad_norm": 70.40740203857422, "learning_rate": 2.5368456085149044e-06, "loss": 0.8409, "step": 1793 }, { "epoch": 1.5099936881969283, "grad_norm": 62.056270599365234, "learning_rate": 2.5346126886292406e-06, "loss": 0.4953, "step": 1794 }, { "epoch": 1.5108352619398273, "grad_norm": 72.21955108642578, "learning_rate": 2.532379741125697e-06, "loss": 0.8509, "step": 1795 }, { "epoch": 1.5116768356827266, "grad_norm": 61.44864273071289, "learning_rate": 2.53014676778597e-06, "loss": 0.7606, "step": 1796 }, { "epoch": 1.5125184094256259, "grad_norm": 80.21410369873047, "learning_rate": 2.527913770391776e-06, "loss": 0.8352, "step": 1797 }, { "epoch": 1.5133599831685252, "grad_norm": 68.3894271850586, "learning_rate": 2.5256807507248515e-06, "loss": 0.8391, "step": 1798 }, { "epoch": 1.5142015569114244, "grad_norm": 64.38005065917969, "learning_rate": 2.5234477105669487e-06, "loss": 0.669, "step": 1799 }, { "epoch": 1.5150431306543237, "grad_norm": 66.75924682617188, "learning_rate": 2.521214651699836e-06, "loss": 0.6427, "step": 1800 }, { "epoch": 1.5158847043972228, "grad_norm": 74.76290893554688, "learning_rate": 2.518981575905299e-06, "loss": 0.7038, "step": 1801 }, { "epoch": 1.516726278140122, "grad_norm": 72.84453582763672, "learning_rate": 2.5167484849651357e-06, "loss": 0.619, "step": 1802 }, { "epoch": 1.5175678518830211, "grad_norm": 50.17288589477539, "learning_rate": 2.5145153806611555e-06, "loss": 0.6522, "step": 1803 }, { "epoch": 1.5184094256259204, "grad_norm": 55.1636962890625, "learning_rate": 2.51228226477518e-06, "loss": 0.5761, "step": 1804 }, { "epoch": 1.5192509993688197, "grad_norm": 82.204345703125, "learning_rate": 2.5100491390890373e-06, "loss": 0.8975, "step": 1805 }, { "epoch": 1.520092573111719, "grad_norm": 66.0625991821289, "learning_rate": 2.5078160053845672e-06, "loss": 0.6909, "step": 1806 }, { "epoch": 1.5209341468546183, "grad_norm": 60.695919036865234, "learning_rate": 2.5055828654436125e-06, "loss": 0.5098, "step": 1807 }, { "epoch": 1.5217757205975173, "grad_norm": 73.6211929321289, "learning_rate": 2.503349721048022e-06, "loss": 0.8039, "step": 1808 }, { "epoch": 1.5226172943404166, "grad_norm": 88.235595703125, "learning_rate": 2.50111657397965e-06, "loss": 0.8847, "step": 1809 }, { "epoch": 1.5234588680833157, "grad_norm": 72.74187469482422, "learning_rate": 2.4988834260203505e-06, "loss": 0.6765, "step": 1810 }, { "epoch": 1.524300441826215, "grad_norm": 62.39958953857422, "learning_rate": 2.4966502789519782e-06, "loss": 0.6344, "step": 1811 }, { "epoch": 1.5251420155691142, "grad_norm": 67.40599060058594, "learning_rate": 2.4944171345563883e-06, "loss": 0.6808, "step": 1812 }, { "epoch": 1.5259835893120135, "grad_norm": 77.7221908569336, "learning_rate": 2.4921839946154336e-06, "loss": 0.9695, "step": 1813 }, { "epoch": 1.5268251630549128, "grad_norm": 68.94850158691406, "learning_rate": 2.489950860910963e-06, "loss": 0.7738, "step": 1814 }, { "epoch": 1.527666736797812, "grad_norm": 69.1143569946289, "learning_rate": 2.487717735224821e-06, "loss": 0.7607, "step": 1815 }, { "epoch": 1.5285083105407111, "grad_norm": 71.65644073486328, "learning_rate": 2.485484619338845e-06, "loss": 0.6638, "step": 1816 }, { "epoch": 1.5293498842836104, "grad_norm": 56.82776641845703, "learning_rate": 2.4832515150348647e-06, "loss": 0.6694, "step": 1817 }, { "epoch": 1.5301914580265095, "grad_norm": 73.23497009277344, "learning_rate": 2.4810184240947012e-06, "loss": 0.7134, "step": 1818 }, { "epoch": 1.5310330317694087, "grad_norm": 67.1120834350586, "learning_rate": 2.4787853483001644e-06, "loss": 0.6523, "step": 1819 }, { "epoch": 1.531874605512308, "grad_norm": 80.06694030761719, "learning_rate": 2.4765522894330526e-06, "loss": 0.8407, "step": 1820 }, { "epoch": 1.5327161792552073, "grad_norm": 61.792110443115234, "learning_rate": 2.474319249275149e-06, "loss": 0.6341, "step": 1821 }, { "epoch": 1.5335577529981066, "grad_norm": 66.96337127685547, "learning_rate": 2.472086229608224e-06, "loss": 0.806, "step": 1822 }, { "epoch": 1.5343993267410057, "grad_norm": 76.55535125732422, "learning_rate": 2.469853232214031e-06, "loss": 0.8714, "step": 1823 }, { "epoch": 1.535240900483905, "grad_norm": 85.6926040649414, "learning_rate": 2.4676202588743043e-06, "loss": 0.7707, "step": 1824 }, { "epoch": 1.536082474226804, "grad_norm": 80.15267944335938, "learning_rate": 2.4653873113707607e-06, "loss": 0.8354, "step": 1825 }, { "epoch": 1.5369240479697033, "grad_norm": 53.12759017944336, "learning_rate": 2.463154391485095e-06, "loss": 0.6233, "step": 1826 }, { "epoch": 1.5377656217126026, "grad_norm": 61.406944274902344, "learning_rate": 2.4609215009989823e-06, "loss": 0.7844, "step": 1827 }, { "epoch": 1.5386071954555018, "grad_norm": 59.22563171386719, "learning_rate": 2.4586886416940716e-06, "loss": 0.7773, "step": 1828 }, { "epoch": 1.5394487691984011, "grad_norm": 74.83763122558594, "learning_rate": 2.4564558153519875e-06, "loss": 0.8684, "step": 1829 }, { "epoch": 1.5402903429413002, "grad_norm": 92.87191772460938, "learning_rate": 2.45422302375433e-06, "loss": 0.8001, "step": 1830 }, { "epoch": 1.5411319166841995, "grad_norm": 61.79736328125, "learning_rate": 2.45199026868267e-06, "loss": 0.8162, "step": 1831 }, { "epoch": 1.5419734904270985, "grad_norm": 142.7338104248047, "learning_rate": 2.4497575519185497e-06, "loss": 1.0086, "step": 1832 }, { "epoch": 1.5428150641699978, "grad_norm": 81.91464233398438, "learning_rate": 2.447524875243479e-06, "loss": 0.841, "step": 1833 }, { "epoch": 1.543656637912897, "grad_norm": 54.11069869995117, "learning_rate": 2.445292240438939e-06, "loss": 0.6511, "step": 1834 }, { "epoch": 1.5444982116557964, "grad_norm": 56.75102233886719, "learning_rate": 2.443059649286375e-06, "loss": 0.7371, "step": 1835 }, { "epoch": 1.5453397853986957, "grad_norm": 58.729736328125, "learning_rate": 2.4408271035671984e-06, "loss": 0.7164, "step": 1836 }, { "epoch": 1.546181359141595, "grad_norm": 75.8735122680664, "learning_rate": 2.438594605062784e-06, "loss": 0.8903, "step": 1837 }, { "epoch": 1.547022932884494, "grad_norm": 59.3784294128418, "learning_rate": 2.436362155554469e-06, "loss": 0.7606, "step": 1838 }, { "epoch": 1.5478645066273933, "grad_norm": 71.88601684570312, "learning_rate": 2.4341297568235524e-06, "loss": 0.9239, "step": 1839 }, { "epoch": 1.5487060803702923, "grad_norm": 66.5792465209961, "learning_rate": 2.431897410651291e-06, "loss": 0.8731, "step": 1840 }, { "epoch": 1.5495476541131916, "grad_norm": 51.85613250732422, "learning_rate": 2.429665118818901e-06, "loss": 0.7003, "step": 1841 }, { "epoch": 1.550389227856091, "grad_norm": 51.30876159667969, "learning_rate": 2.4274328831075543e-06, "loss": 0.5885, "step": 1842 }, { "epoch": 1.5512308015989902, "grad_norm": 94.05731201171875, "learning_rate": 2.4252007052983784e-06, "loss": 1.5032, "step": 1843 }, { "epoch": 1.5520723753418895, "grad_norm": 69.79046630859375, "learning_rate": 2.4229685871724555e-06, "loss": 0.7215, "step": 1844 }, { "epoch": 1.5529139490847885, "grad_norm": 66.31012725830078, "learning_rate": 2.4207365305108186e-06, "loss": 0.8458, "step": 1845 }, { "epoch": 1.5537555228276878, "grad_norm": 68.99163055419922, "learning_rate": 2.4185045370944527e-06, "loss": 0.7926, "step": 1846 }, { "epoch": 1.5545970965705869, "grad_norm": 91.59426879882812, "learning_rate": 2.4162726087042924e-06, "loss": 0.9716, "step": 1847 }, { "epoch": 1.5554386703134861, "grad_norm": 97.00667572021484, "learning_rate": 2.4140407471212195e-06, "loss": 0.9322, "step": 1848 }, { "epoch": 1.5562802440563854, "grad_norm": 110.47148132324219, "learning_rate": 2.4118089541260635e-06, "loss": 0.8022, "step": 1849 }, { "epoch": 1.5571218177992847, "grad_norm": 66.46359252929688, "learning_rate": 2.409577231499598e-06, "loss": 0.8061, "step": 1850 }, { "epoch": 1.557963391542184, "grad_norm": 58.62633514404297, "learning_rate": 2.407345581022543e-06, "loss": 0.7221, "step": 1851 }, { "epoch": 1.558804965285083, "grad_norm": 78.39395904541016, "learning_rate": 2.405114004475557e-06, "loss": 1.0168, "step": 1852 }, { "epoch": 1.5596465390279823, "grad_norm": 61.914886474609375, "learning_rate": 2.4028825036392427e-06, "loss": 0.7306, "step": 1853 }, { "epoch": 1.5604881127708814, "grad_norm": 75.09632110595703, "learning_rate": 2.4006510802941417e-06, "loss": 0.887, "step": 1854 }, { "epoch": 1.5613296865137807, "grad_norm": 54.39837646484375, "learning_rate": 2.3984197362207328e-06, "loss": 0.6556, "step": 1855 }, { "epoch": 1.56217126025668, "grad_norm": 68.95436096191406, "learning_rate": 2.3961884731994314e-06, "loss": 0.8248, "step": 1856 }, { "epoch": 1.5630128339995792, "grad_norm": 64.99290466308594, "learning_rate": 2.3939572930105903e-06, "loss": 0.8434, "step": 1857 }, { "epoch": 1.5638544077424785, "grad_norm": 73.03610229492188, "learning_rate": 2.3917261974344944e-06, "loss": 0.84, "step": 1858 }, { "epoch": 1.5646959814853778, "grad_norm": 59.284454345703125, "learning_rate": 2.3894951882513613e-06, "loss": 0.7346, "step": 1859 }, { "epoch": 1.5655375552282769, "grad_norm": 83.41828918457031, "learning_rate": 2.38726426724134e-06, "loss": 0.7097, "step": 1860 }, { "epoch": 1.5663791289711761, "grad_norm": 74.30712127685547, "learning_rate": 2.38503343618451e-06, "loss": 0.8443, "step": 1861 }, { "epoch": 1.5672207027140752, "grad_norm": 68.92391967773438, "learning_rate": 2.3828026968608768e-06, "loss": 0.7339, "step": 1862 }, { "epoch": 1.5680622764569745, "grad_norm": 60.97203063964844, "learning_rate": 2.3805720510503744e-06, "loss": 0.7265, "step": 1863 }, { "epoch": 1.5689038501998738, "grad_norm": 70.96116638183594, "learning_rate": 2.378341500532862e-06, "loss": 0.8255, "step": 1864 }, { "epoch": 1.569745423942773, "grad_norm": 60.73342514038086, "learning_rate": 2.3761110470881227e-06, "loss": 0.6874, "step": 1865 }, { "epoch": 1.5705869976856723, "grad_norm": 83.00879669189453, "learning_rate": 2.3738806924958617e-06, "loss": 0.7904, "step": 1866 }, { "epoch": 1.5714285714285714, "grad_norm": 63.39756393432617, "learning_rate": 2.371650438535706e-06, "loss": 0.8388, "step": 1867 }, { "epoch": 1.5722701451714707, "grad_norm": 64.87178039550781, "learning_rate": 2.3694202869872015e-06, "loss": 0.7687, "step": 1868 }, { "epoch": 1.5731117189143697, "grad_norm": 51.909698486328125, "learning_rate": 2.367190239629813e-06, "loss": 0.6358, "step": 1869 }, { "epoch": 1.573953292657269, "grad_norm": 60.68330001831055, "learning_rate": 2.364960298242922e-06, "loss": 0.7397, "step": 1870 }, { "epoch": 1.5747948664001683, "grad_norm": 66.23773956298828, "learning_rate": 2.362730464605825e-06, "loss": 0.6968, "step": 1871 }, { "epoch": 1.5756364401430676, "grad_norm": 74.61249542236328, "learning_rate": 2.3605007404977333e-06, "loss": 0.7697, "step": 1872 }, { "epoch": 1.5764780138859669, "grad_norm": 61.38346862792969, "learning_rate": 2.3582711276977704e-06, "loss": 0.8497, "step": 1873 }, { "epoch": 1.577319587628866, "grad_norm": 60.430137634277344, "learning_rate": 2.3560416279849717e-06, "loss": 0.7473, "step": 1874 }, { "epoch": 1.5781611613717652, "grad_norm": 57.78585433959961, "learning_rate": 2.3538122431382805e-06, "loss": 0.5718, "step": 1875 }, { "epoch": 1.5790027351146643, "grad_norm": 69.59358215332031, "learning_rate": 2.3515829749365516e-06, "loss": 0.7033, "step": 1876 }, { "epoch": 1.5798443088575636, "grad_norm": 72.73912811279297, "learning_rate": 2.3493538251585426e-06, "loss": 0.6765, "step": 1877 }, { "epoch": 1.5806858826004628, "grad_norm": 73.55653381347656, "learning_rate": 2.347124795582919e-06, "loss": 0.7669, "step": 1878 }, { "epoch": 1.5815274563433621, "grad_norm": 78.63309478759766, "learning_rate": 2.3448958879882513e-06, "loss": 0.8789, "step": 1879 }, { "epoch": 1.5823690300862614, "grad_norm": 82.12600708007812, "learning_rate": 2.3426671041530107e-06, "loss": 0.7361, "step": 1880 }, { "epoch": 1.5832106038291607, "grad_norm": 91.17112731933594, "learning_rate": 2.340438445855571e-06, "loss": 0.8799, "step": 1881 }, { "epoch": 1.5840521775720597, "grad_norm": 67.19610595703125, "learning_rate": 2.3382099148742054e-06, "loss": 0.7402, "step": 1882 }, { "epoch": 1.584893751314959, "grad_norm": 72.3564682006836, "learning_rate": 2.335981512987085e-06, "loss": 0.6634, "step": 1883 }, { "epoch": 1.585735325057858, "grad_norm": 91.96434783935547, "learning_rate": 2.333753241972279e-06, "loss": 0.8571, "step": 1884 }, { "epoch": 1.5865768988007574, "grad_norm": 66.343017578125, "learning_rate": 2.3315251036077514e-06, "loss": 0.8014, "step": 1885 }, { "epoch": 1.5874184725436566, "grad_norm": 73.31749725341797, "learning_rate": 2.3292970996713596e-06, "loss": 0.8267, "step": 1886 }, { "epoch": 1.588260046286556, "grad_norm": 57.65916061401367, "learning_rate": 2.327069231940855e-06, "loss": 0.7508, "step": 1887 }, { "epoch": 1.5891016200294552, "grad_norm": 66.38873291015625, "learning_rate": 2.32484150219388e-06, "loss": 0.703, "step": 1888 }, { "epoch": 1.5899431937723543, "grad_norm": 77.54417419433594, "learning_rate": 2.3226139122079674e-06, "loss": 0.7594, "step": 1889 }, { "epoch": 1.5907847675152536, "grad_norm": 65.63406372070312, "learning_rate": 2.3203864637605373e-06, "loss": 0.8108, "step": 1890 }, { "epoch": 1.5916263412581526, "grad_norm": 69.9978256225586, "learning_rate": 2.3181591586288974e-06, "loss": 0.7563, "step": 1891 }, { "epoch": 1.592467915001052, "grad_norm": 66.27239990234375, "learning_rate": 2.3159319985902413e-06, "loss": 0.5842, "step": 1892 }, { "epoch": 1.5933094887439512, "grad_norm": 70.13909912109375, "learning_rate": 2.313704985421647e-06, "loss": 1.0032, "step": 1893 }, { "epoch": 1.5941510624868505, "grad_norm": 70.11542510986328, "learning_rate": 2.311478120900074e-06, "loss": 0.676, "step": 1894 }, { "epoch": 1.5949926362297497, "grad_norm": 63.457454681396484, "learning_rate": 2.3092514068023657e-06, "loss": 0.7231, "step": 1895 }, { "epoch": 1.5958342099726488, "grad_norm": 73.18553924560547, "learning_rate": 2.3070248449052423e-06, "loss": 0.7353, "step": 1896 }, { "epoch": 1.596675783715548, "grad_norm": 63.250892639160156, "learning_rate": 2.3047984369853053e-06, "loss": 0.7342, "step": 1897 }, { "epoch": 1.5975173574584471, "grad_norm": 71.97840118408203, "learning_rate": 2.3025721848190322e-06, "loss": 0.7405, "step": 1898 }, { "epoch": 1.5983589312013464, "grad_norm": 80.704345703125, "learning_rate": 2.3003460901827757e-06, "loss": 0.7872, "step": 1899 }, { "epoch": 1.5992005049442457, "grad_norm": 60.682701110839844, "learning_rate": 2.2981201548527637e-06, "loss": 0.5739, "step": 1900 }, { "epoch": 1.600042078687145, "grad_norm": 66.49835968017578, "learning_rate": 2.295894380605096e-06, "loss": 0.6998, "step": 1901 }, { "epoch": 1.6008836524300443, "grad_norm": 76.23893737792969, "learning_rate": 2.293668769215745e-06, "loss": 0.6962, "step": 1902 }, { "epoch": 1.6017252261729436, "grad_norm": 70.72975158691406, "learning_rate": 2.291443322460553e-06, "loss": 0.8068, "step": 1903 }, { "epoch": 1.6025667999158426, "grad_norm": 60.4953727722168, "learning_rate": 2.28921804211523e-06, "loss": 0.5729, "step": 1904 }, { "epoch": 1.603408373658742, "grad_norm": 59.01346206665039, "learning_rate": 2.2869929299553544e-06, "loss": 0.7066, "step": 1905 }, { "epoch": 1.604249947401641, "grad_norm": 63.7917594909668, "learning_rate": 2.2847679877563692e-06, "loss": 0.6744, "step": 1906 }, { "epoch": 1.6050915211445402, "grad_norm": 62.64045715332031, "learning_rate": 2.282543217293583e-06, "loss": 0.6869, "step": 1907 }, { "epoch": 1.6059330948874395, "grad_norm": 58.16526412963867, "learning_rate": 2.2803186203421664e-06, "loss": 0.6462, "step": 1908 }, { "epoch": 1.6067746686303388, "grad_norm": 71.87863159179688, "learning_rate": 2.278094198677152e-06, "loss": 0.5625, "step": 1909 }, { "epoch": 1.607616242373238, "grad_norm": 89.81664276123047, "learning_rate": 2.2758699540734337e-06, "loss": 0.8705, "step": 1910 }, { "epoch": 1.6084578161161371, "grad_norm": 75.28933715820312, "learning_rate": 2.2736458883057604e-06, "loss": 0.6407, "step": 1911 }, { "epoch": 1.6092993898590364, "grad_norm": 61.30549240112305, "learning_rate": 2.2714220031487426e-06, "loss": 0.6456, "step": 1912 }, { "epoch": 1.6101409636019355, "grad_norm": 76.59605407714844, "learning_rate": 2.269198300376844e-06, "loss": 0.6283, "step": 1913 }, { "epoch": 1.6109825373448348, "grad_norm": 69.38159942626953, "learning_rate": 2.2669747817643845e-06, "loss": 0.6952, "step": 1914 }, { "epoch": 1.611824111087734, "grad_norm": 79.64710998535156, "learning_rate": 2.2647514490855354e-06, "loss": 0.7826, "step": 1915 }, { "epoch": 1.6126656848306333, "grad_norm": 56.66648483276367, "learning_rate": 2.262528304114321e-06, "loss": 0.4689, "step": 1916 }, { "epoch": 1.6135072585735326, "grad_norm": 51.78862762451172, "learning_rate": 2.2603053486246153e-06, "loss": 0.5611, "step": 1917 }, { "epoch": 1.614348832316432, "grad_norm": 62.394996643066406, "learning_rate": 2.2580825843901404e-06, "loss": 0.7488, "step": 1918 }, { "epoch": 1.615190406059331, "grad_norm": 78.04889678955078, "learning_rate": 2.2558600131844672e-06, "loss": 0.7673, "step": 1919 }, { "epoch": 1.61603197980223, "grad_norm": 86.36821746826172, "learning_rate": 2.253637636781013e-06, "loss": 0.9458, "step": 1920 }, { "epoch": 1.6168735535451293, "grad_norm": 66.14805603027344, "learning_rate": 2.251415456953036e-06, "loss": 1.2449, "step": 1921 }, { "epoch": 1.6177151272880286, "grad_norm": 73.35967254638672, "learning_rate": 2.2491934754736415e-06, "loss": 0.8074, "step": 1922 }, { "epoch": 1.6185567010309279, "grad_norm": 73.34542846679688, "learning_rate": 2.2469716941157744e-06, "loss": 0.8084, "step": 1923 }, { "epoch": 1.6193982747738271, "grad_norm": 58.43281936645508, "learning_rate": 2.2447501146522215e-06, "loss": 0.6528, "step": 1924 }, { "epoch": 1.6202398485167264, "grad_norm": 68.62159729003906, "learning_rate": 2.242528738855607e-06, "loss": 0.7318, "step": 1925 }, { "epoch": 1.6210814222596255, "grad_norm": 50.025611877441406, "learning_rate": 2.2403075684983937e-06, "loss": 0.5296, "step": 1926 }, { "epoch": 1.6219229960025248, "grad_norm": 64.96892547607422, "learning_rate": 2.23808660535288e-06, "loss": 0.8676, "step": 1927 }, { "epoch": 1.6227645697454238, "grad_norm": 61.32883834838867, "learning_rate": 2.2358658511911986e-06, "loss": 0.651, "step": 1928 }, { "epoch": 1.623606143488323, "grad_norm": 79.6020278930664, "learning_rate": 2.2336453077853167e-06, "loss": 1.0572, "step": 1929 }, { "epoch": 1.6244477172312224, "grad_norm": 53.11130905151367, "learning_rate": 2.231424976907031e-06, "loss": 0.6287, "step": 1930 }, { "epoch": 1.6252892909741217, "grad_norm": 70.36692810058594, "learning_rate": 2.229204860327971e-06, "loss": 0.6388, "step": 1931 }, { "epoch": 1.626130864717021, "grad_norm": 53.396766662597656, "learning_rate": 2.2269849598195944e-06, "loss": 0.4792, "step": 1932 }, { "epoch": 1.62697243845992, "grad_norm": 66.66010284423828, "learning_rate": 2.224765277153186e-06, "loss": 0.8598, "step": 1933 }, { "epoch": 1.6278140122028193, "grad_norm": 66.52577209472656, "learning_rate": 2.2225458140998574e-06, "loss": 0.8062, "step": 1934 }, { "epoch": 1.6286555859457184, "grad_norm": 81.40502166748047, "learning_rate": 2.2203265724305447e-06, "loss": 0.8584, "step": 1935 }, { "epoch": 1.6294971596886176, "grad_norm": 63.646629333496094, "learning_rate": 2.218107553916007e-06, "loss": 0.7144, "step": 1936 }, { "epoch": 1.630338733431517, "grad_norm": 65.940673828125, "learning_rate": 2.2158887603268265e-06, "loss": 0.6844, "step": 1937 }, { "epoch": 1.6311803071744162, "grad_norm": 73.90335845947266, "learning_rate": 2.213670193433404e-06, "loss": 0.7667, "step": 1938 }, { "epoch": 1.6320218809173155, "grad_norm": 60.94840621948242, "learning_rate": 2.211451855005962e-06, "loss": 0.8619, "step": 1939 }, { "epoch": 1.6328634546602148, "grad_norm": 48.2357177734375, "learning_rate": 2.209233746814539e-06, "loss": 0.5923, "step": 1940 }, { "epoch": 1.6337050284031138, "grad_norm": 80.9052963256836, "learning_rate": 2.2070158706289897e-06, "loss": 0.651, "step": 1941 }, { "epoch": 1.6345466021460129, "grad_norm": 61.8893928527832, "learning_rate": 2.2047982282189834e-06, "loss": 0.6907, "step": 1942 }, { "epoch": 1.6353881758889122, "grad_norm": 57.87552261352539, "learning_rate": 2.2025808213540044e-06, "loss": 0.7559, "step": 1943 }, { "epoch": 1.6362297496318114, "grad_norm": 60.96064376831055, "learning_rate": 2.2003636518033478e-06, "loss": 0.6719, "step": 1944 }, { "epoch": 1.6370713233747107, "grad_norm": 69.94816589355469, "learning_rate": 2.1981467213361198e-06, "loss": 0.8268, "step": 1945 }, { "epoch": 1.63791289711761, "grad_norm": 74.3624038696289, "learning_rate": 2.195930031721236e-06, "loss": 0.822, "step": 1946 }, { "epoch": 1.6387544708605093, "grad_norm": 58.52619552612305, "learning_rate": 2.193713584727419e-06, "loss": 0.6349, "step": 1947 }, { "epoch": 1.6395960446034084, "grad_norm": 83.47274780273438, "learning_rate": 2.191497382123199e-06, "loss": 0.7253, "step": 1948 }, { "epoch": 1.6404376183463076, "grad_norm": 80.07342529296875, "learning_rate": 2.189281425676911e-06, "loss": 0.921, "step": 1949 }, { "epoch": 1.6412791920892067, "grad_norm": 72.1954574584961, "learning_rate": 2.1870657171566916e-06, "loss": 0.8344, "step": 1950 }, { "epoch": 1.642120765832106, "grad_norm": 53.89166259765625, "learning_rate": 2.1848502583304824e-06, "loss": 0.5467, "step": 1951 }, { "epoch": 1.6429623395750053, "grad_norm": 66.69710540771484, "learning_rate": 2.1826350509660245e-06, "loss": 0.7852, "step": 1952 }, { "epoch": 1.6438039133179045, "grad_norm": 58.29098129272461, "learning_rate": 2.1804200968308575e-06, "loss": 0.6756, "step": 1953 }, { "epoch": 1.6446454870608038, "grad_norm": 67.6562271118164, "learning_rate": 2.1782053976923213e-06, "loss": 0.6601, "step": 1954 }, { "epoch": 1.6454870608037029, "grad_norm": 65.10391998291016, "learning_rate": 2.1759909553175496e-06, "loss": 0.6831, "step": 1955 }, { "epoch": 1.6463286345466022, "grad_norm": 70.75611877441406, "learning_rate": 2.1737767714734727e-06, "loss": 0.7738, "step": 1956 }, { "epoch": 1.6471702082895012, "grad_norm": 62.865966796875, "learning_rate": 2.1715628479268142e-06, "loss": 0.7, "step": 1957 }, { "epoch": 1.6480117820324005, "grad_norm": 62.8587532043457, "learning_rate": 2.169349186444091e-06, "loss": 0.729, "step": 1958 }, { "epoch": 1.6488533557752998, "grad_norm": 69.52644348144531, "learning_rate": 2.167135788791609e-06, "loss": 0.7733, "step": 1959 }, { "epoch": 1.649694929518199, "grad_norm": 81.00640106201172, "learning_rate": 2.1649226567354654e-06, "loss": 0.8087, "step": 1960 }, { "epoch": 1.6505365032610984, "grad_norm": 72.51599884033203, "learning_rate": 2.162709792041545e-06, "loss": 0.7809, "step": 1961 }, { "epoch": 1.6513780770039976, "grad_norm": 79.64278411865234, "learning_rate": 2.1604971964755183e-06, "loss": 0.9589, "step": 1962 }, { "epoch": 1.6522196507468967, "grad_norm": 68.44630432128906, "learning_rate": 2.1582848718028424e-06, "loss": 0.9324, "step": 1963 }, { "epoch": 1.6530612244897958, "grad_norm": 65.26687622070312, "learning_rate": 2.156072819788758e-06, "loss": 0.7008, "step": 1964 }, { "epoch": 1.653902798232695, "grad_norm": 112.8241195678711, "learning_rate": 2.1538610421982868e-06, "loss": 0.9982, "step": 1965 }, { "epoch": 1.6547443719755943, "grad_norm": 66.75296783447266, "learning_rate": 2.151649540796233e-06, "loss": 0.7606, "step": 1966 }, { "epoch": 1.6555859457184936, "grad_norm": 53.238224029541016, "learning_rate": 2.14943831734718e-06, "loss": 0.7262, "step": 1967 }, { "epoch": 1.6564275194613929, "grad_norm": 58.80339813232422, "learning_rate": 2.1472273736154893e-06, "loss": 0.742, "step": 1968 }, { "epoch": 1.6572690932042922, "grad_norm": 63.22513198852539, "learning_rate": 2.1450167113653e-06, "loss": 0.5513, "step": 1969 }, { "epoch": 1.6581106669471912, "grad_norm": 92.08692169189453, "learning_rate": 2.142806332360525e-06, "loss": 0.6216, "step": 1970 }, { "epoch": 1.6589522406900905, "grad_norm": 58.243263244628906, "learning_rate": 2.140596238364853e-06, "loss": 0.6815, "step": 1971 }, { "epoch": 1.6597938144329896, "grad_norm": 65.3012466430664, "learning_rate": 2.138386431141745e-06, "loss": 0.772, "step": 1972 }, { "epoch": 1.6606353881758888, "grad_norm": 65.20149230957031, "learning_rate": 2.136176912454431e-06, "loss": 0.7605, "step": 1973 }, { "epoch": 1.6614769619187881, "grad_norm": 71.35391235351562, "learning_rate": 2.133967684065914e-06, "loss": 0.8941, "step": 1974 }, { "epoch": 1.6623185356616874, "grad_norm": 75.7702407836914, "learning_rate": 2.131758747738962e-06, "loss": 0.8523, "step": 1975 }, { "epoch": 1.6631601094045867, "grad_norm": 83.33497619628906, "learning_rate": 2.1295501052361135e-06, "loss": 0.8715, "step": 1976 }, { "epoch": 1.6640016831474858, "grad_norm": 72.53399658203125, "learning_rate": 2.1273417583196702e-06, "loss": 0.7308, "step": 1977 }, { "epoch": 1.664843256890385, "grad_norm": 57.52545928955078, "learning_rate": 2.1251337087516978e-06, "loss": 0.6671, "step": 1978 }, { "epoch": 1.665684830633284, "grad_norm": 67.9999008178711, "learning_rate": 2.1229259582940263e-06, "loss": 0.6903, "step": 1979 }, { "epoch": 1.6665264043761834, "grad_norm": 70.0722427368164, "learning_rate": 2.1207185087082463e-06, "loss": 0.8178, "step": 1980 }, { "epoch": 1.6673679781190827, "grad_norm": 59.60759353637695, "learning_rate": 2.1185113617557077e-06, "loss": 0.6901, "step": 1981 }, { "epoch": 1.668209551861982, "grad_norm": 56.739009857177734, "learning_rate": 2.1163045191975196e-06, "loss": 0.5001, "step": 1982 }, { "epoch": 1.6690511256048812, "grad_norm": 63.38383865356445, "learning_rate": 2.114097982794548e-06, "loss": 0.823, "step": 1983 }, { "epoch": 1.6698926993477805, "grad_norm": 63.312259674072266, "learning_rate": 2.1118917543074154e-06, "loss": 0.77, "step": 1984 }, { "epoch": 1.6707342730906796, "grad_norm": 70.6395034790039, "learning_rate": 2.1096858354964975e-06, "loss": 0.6799, "step": 1985 }, { "epoch": 1.6715758468335786, "grad_norm": 67.7114028930664, "learning_rate": 2.107480228121923e-06, "loss": 0.7599, "step": 1986 }, { "epoch": 1.672417420576478, "grad_norm": 71.95157623291016, "learning_rate": 2.105274933943573e-06, "loss": 0.5956, "step": 1987 }, { "epoch": 1.6732589943193772, "grad_norm": 59.94184112548828, "learning_rate": 2.1030699547210773e-06, "loss": 0.6317, "step": 1988 }, { "epoch": 1.6741005680622765, "grad_norm": 80.0818099975586, "learning_rate": 2.1008652922138156e-06, "loss": 0.5644, "step": 1989 }, { "epoch": 1.6749421418051758, "grad_norm": 62.6698112487793, "learning_rate": 2.098660948180915e-06, "loss": 0.6761, "step": 1990 }, { "epoch": 1.675783715548075, "grad_norm": 73.26113891601562, "learning_rate": 2.096456924381247e-06, "loss": 0.7827, "step": 1991 }, { "epoch": 1.676625289290974, "grad_norm": 62.22536849975586, "learning_rate": 2.09425322257343e-06, "loss": 0.7173, "step": 1992 }, { "epoch": 1.6774668630338734, "grad_norm": 63.22867965698242, "learning_rate": 2.0920498445158227e-06, "loss": 0.7695, "step": 1993 }, { "epoch": 1.6783084367767724, "grad_norm": 71.86830139160156, "learning_rate": 2.0898467919665276e-06, "loss": 0.8375, "step": 1994 }, { "epoch": 1.6791500105196717, "grad_norm": 82.2249526977539, "learning_rate": 2.0876440666833863e-06, "loss": 0.7509, "step": 1995 }, { "epoch": 1.679991584262571, "grad_norm": 63.76021194458008, "learning_rate": 2.0854416704239795e-06, "loss": 0.6561, "step": 1996 }, { "epoch": 1.6808331580054703, "grad_norm": 72.8883056640625, "learning_rate": 2.083239604945626e-06, "loss": 0.9157, "step": 1997 }, { "epoch": 1.6816747317483696, "grad_norm": 57.80598831176758, "learning_rate": 2.081037872005381e-06, "loss": 0.6497, "step": 1998 }, { "epoch": 1.6825163054912686, "grad_norm": 82.14009857177734, "learning_rate": 2.0788364733600313e-06, "loss": 0.7494, "step": 1999 }, { "epoch": 1.683357879234168, "grad_norm": 71.7037124633789, "learning_rate": 2.076635410766101e-06, "loss": 0.6929, "step": 2000 }, { "epoch": 1.684199452977067, "grad_norm": 70.10237884521484, "learning_rate": 2.0744346859798442e-06, "loss": 0.749, "step": 2001 }, { "epoch": 1.6850410267199663, "grad_norm": 56.99276351928711, "learning_rate": 2.0722343007572447e-06, "loss": 0.484, "step": 2002 }, { "epoch": 1.6858826004628655, "grad_norm": 66.94182586669922, "learning_rate": 2.0700342568540166e-06, "loss": 0.783, "step": 2003 }, { "epoch": 1.6867241742057648, "grad_norm": 77.5578384399414, "learning_rate": 2.0678345560256015e-06, "loss": 0.7812, "step": 2004 }, { "epoch": 1.687565747948664, "grad_norm": 83.48815155029297, "learning_rate": 2.065635200027167e-06, "loss": 0.8372, "step": 2005 }, { "epoch": 1.6884073216915634, "grad_norm": 67.20012664794922, "learning_rate": 2.0634361906136053e-06, "loss": 0.7154, "step": 2006 }, { "epoch": 1.6892488954344624, "grad_norm": 74.85514068603516, "learning_rate": 2.0612375295395324e-06, "loss": 1.0268, "step": 2007 }, { "epoch": 1.6900904691773617, "grad_norm": 67.56674194335938, "learning_rate": 2.0590392185592877e-06, "loss": 0.7718, "step": 2008 }, { "epoch": 1.6909320429202608, "grad_norm": 60.73134231567383, "learning_rate": 2.0568412594269273e-06, "loss": 0.7355, "step": 2009 }, { "epoch": 1.69177361666316, "grad_norm": 79.8938217163086, "learning_rate": 2.0546436538962304e-06, "loss": 0.8255, "step": 2010 }, { "epoch": 1.6926151904060593, "grad_norm": 70.16535949707031, "learning_rate": 2.0524464037206924e-06, "loss": 0.6692, "step": 2011 }, { "epoch": 1.6934567641489586, "grad_norm": 74.3699722290039, "learning_rate": 2.050249510653526e-06, "loss": 0.7874, "step": 2012 }, { "epoch": 1.694298337891858, "grad_norm": 57.0115966796875, "learning_rate": 2.048052976447658e-06, "loss": 0.6623, "step": 2013 }, { "epoch": 1.695139911634757, "grad_norm": 62.181549072265625, "learning_rate": 2.04585680285573e-06, "loss": 0.8497, "step": 2014 }, { "epoch": 1.6959814853776563, "grad_norm": 63.5701789855957, "learning_rate": 2.0436609916300943e-06, "loss": 0.7479, "step": 2015 }, { "epoch": 1.6968230591205553, "grad_norm": 64.78250885009766, "learning_rate": 2.041465544522816e-06, "loss": 0.6866, "step": 2016 }, { "epoch": 1.6976646328634546, "grad_norm": 63.110042572021484, "learning_rate": 2.039270463285668e-06, "loss": 0.621, "step": 2017 }, { "epoch": 1.6985062066063539, "grad_norm": 61.109371185302734, "learning_rate": 2.0370757496701313e-06, "loss": 0.6547, "step": 2018 }, { "epoch": 1.6993477803492532, "grad_norm": 64.15480041503906, "learning_rate": 2.0348814054273954e-06, "loss": 0.5722, "step": 2019 }, { "epoch": 1.7001893540921524, "grad_norm": 73.68436431884766, "learning_rate": 2.0326874323083532e-06, "loss": 0.6723, "step": 2020 }, { "epoch": 1.7010309278350515, "grad_norm": 80.33219146728516, "learning_rate": 2.030493832063602e-06, "loss": 0.9425, "step": 2021 }, { "epoch": 1.7018725015779508, "grad_norm": 67.62129211425781, "learning_rate": 2.0283006064434417e-06, "loss": 0.6225, "step": 2022 }, { "epoch": 1.7027140753208498, "grad_norm": 67.28459167480469, "learning_rate": 2.0261077571978737e-06, "loss": 0.5826, "step": 2023 }, { "epoch": 1.7035556490637491, "grad_norm": 76.38748931884766, "learning_rate": 2.023915286076598e-06, "loss": 0.7516, "step": 2024 }, { "epoch": 1.7043972228066484, "grad_norm": 76.17875671386719, "learning_rate": 2.0217231948290146e-06, "loss": 0.9187, "step": 2025 }, { "epoch": 1.7052387965495477, "grad_norm": 80.77780151367188, "learning_rate": 2.019531485204218e-06, "loss": 0.7201, "step": 2026 }, { "epoch": 1.706080370292447, "grad_norm": 82.1442642211914, "learning_rate": 2.0173401589510007e-06, "loss": 0.7575, "step": 2027 }, { "epoch": 1.7069219440353463, "grad_norm": 58.396751403808594, "learning_rate": 2.0151492178178477e-06, "loss": 0.6444, "step": 2028 }, { "epoch": 1.7077635177782453, "grad_norm": 66.89722442626953, "learning_rate": 2.0129586635529373e-06, "loss": 0.7763, "step": 2029 }, { "epoch": 1.7086050915211446, "grad_norm": 79.16987609863281, "learning_rate": 2.0107684979041385e-06, "loss": 0.8056, "step": 2030 }, { "epoch": 1.7094466652640437, "grad_norm": 56.09518814086914, "learning_rate": 2.0085787226190106e-06, "loss": 0.6896, "step": 2031 }, { "epoch": 1.710288239006943, "grad_norm": 74.3250732421875, "learning_rate": 2.0063893394448024e-06, "loss": 0.7337, "step": 2032 }, { "epoch": 1.7111298127498422, "grad_norm": 76.9399185180664, "learning_rate": 2.0042003501284483e-06, "loss": 0.9612, "step": 2033 }, { "epoch": 1.7119713864927415, "grad_norm": 67.16383361816406, "learning_rate": 2.002011756416569e-06, "loss": 0.6726, "step": 2034 }, { "epoch": 1.7128129602356408, "grad_norm": 66.2113037109375, "learning_rate": 1.99982356005547e-06, "loss": 0.8357, "step": 2035 }, { "epoch": 1.7136545339785398, "grad_norm": 71.09184265136719, "learning_rate": 1.997635762791139e-06, "loss": 0.6289, "step": 2036 }, { "epoch": 1.7144961077214391, "grad_norm": 84.5699691772461, "learning_rate": 1.995448366369246e-06, "loss": 0.9711, "step": 2037 }, { "epoch": 1.7153376814643382, "grad_norm": 65.88434600830078, "learning_rate": 1.9932613725351407e-06, "loss": 0.7449, "step": 2038 }, { "epoch": 1.7161792552072375, "grad_norm": 81.14341735839844, "learning_rate": 1.9910747830338512e-06, "loss": 0.6928, "step": 2039 }, { "epoch": 1.7170208289501367, "grad_norm": 60.15339279174805, "learning_rate": 1.9888885996100843e-06, "loss": 0.7836, "step": 2040 }, { "epoch": 1.717862402693036, "grad_norm": 72.8441162109375, "learning_rate": 1.9867028240082214e-06, "loss": 0.9141, "step": 2041 }, { "epoch": 1.7187039764359353, "grad_norm": 73.18795013427734, "learning_rate": 1.984517457972319e-06, "loss": 0.7361, "step": 2042 }, { "epoch": 1.7195455501788344, "grad_norm": 77.13481903076172, "learning_rate": 1.982332503246107e-06, "loss": 0.7836, "step": 2043 }, { "epoch": 1.7203871239217337, "grad_norm": 78.78730773925781, "learning_rate": 1.980147961572986e-06, "loss": 0.7982, "step": 2044 }, { "epoch": 1.7212286976646327, "grad_norm": 83.6219482421875, "learning_rate": 1.9779638346960297e-06, "loss": 0.894, "step": 2045 }, { "epoch": 1.722070271407532, "grad_norm": 65.20450592041016, "learning_rate": 1.975780124357977e-06, "loss": 0.6408, "step": 2046 }, { "epoch": 1.7229118451504313, "grad_norm": 67.91110229492188, "learning_rate": 1.9735968323012384e-06, "loss": 0.8016, "step": 2047 }, { "epoch": 1.7237534188933306, "grad_norm": 65.83489990234375, "learning_rate": 1.971413960267888e-06, "loss": 0.8645, "step": 2048 }, { "epoch": 1.7245949926362298, "grad_norm": 53.356876373291016, "learning_rate": 1.969231509999665e-06, "loss": 0.6035, "step": 2049 }, { "epoch": 1.7254365663791291, "grad_norm": 81.52629089355469, "learning_rate": 1.9670494832379735e-06, "loss": 0.8186, "step": 2050 }, { "epoch": 1.7262781401220282, "grad_norm": 63.8055534362793, "learning_rate": 1.964867881723878e-06, "loss": 0.7779, "step": 2051 }, { "epoch": 1.7271197138649275, "grad_norm": 71.25120544433594, "learning_rate": 1.962686707198106e-06, "loss": 0.7822, "step": 2052 }, { "epoch": 1.7279612876078265, "grad_norm": 68.70378875732422, "learning_rate": 1.960505961401041e-06, "loss": 0.7607, "step": 2053 }, { "epoch": 1.7288028613507258, "grad_norm": 61.30226516723633, "learning_rate": 1.9583256460727263e-06, "loss": 0.6749, "step": 2054 }, { "epoch": 1.729644435093625, "grad_norm": 56.17864990234375, "learning_rate": 1.956145762952862e-06, "loss": 0.6608, "step": 2055 }, { "epoch": 1.7304860088365244, "grad_norm": 59.89890670776367, "learning_rate": 1.9539663137808034e-06, "loss": 0.7246, "step": 2056 }, { "epoch": 1.7313275825794237, "grad_norm": 84.15071105957031, "learning_rate": 1.9517873002955584e-06, "loss": 0.6847, "step": 2057 }, { "epoch": 1.7321691563223227, "grad_norm": 64.42945098876953, "learning_rate": 1.949608724235788e-06, "loss": 0.7775, "step": 2058 }, { "epoch": 1.733010730065222, "grad_norm": 72.126708984375, "learning_rate": 1.947430587339805e-06, "loss": 0.8892, "step": 2059 }, { "epoch": 1.733852303808121, "grad_norm": 67.61111450195312, "learning_rate": 1.94525289134557e-06, "loss": 0.662, "step": 2060 }, { "epoch": 1.7346938775510203, "grad_norm": 64.27236938476562, "learning_rate": 1.943075637990693e-06, "loss": 0.6804, "step": 2061 }, { "epoch": 1.7355354512939196, "grad_norm": 56.20612335205078, "learning_rate": 1.94089882901243e-06, "loss": 0.6295, "step": 2062 }, { "epoch": 1.736377025036819, "grad_norm": 58.25652313232422, "learning_rate": 1.9387224661476837e-06, "loss": 0.7555, "step": 2063 }, { "epoch": 1.7372185987797182, "grad_norm": 50.20823669433594, "learning_rate": 1.9365465511329994e-06, "loss": 0.5574, "step": 2064 }, { "epoch": 1.7380601725226172, "grad_norm": 71.83638000488281, "learning_rate": 1.9343710857045663e-06, "loss": 0.7911, "step": 2065 }, { "epoch": 1.7389017462655165, "grad_norm": 62.68744659423828, "learning_rate": 1.932196071598214e-06, "loss": 0.7785, "step": 2066 }, { "epoch": 1.7397433200084156, "grad_norm": 65.7825927734375, "learning_rate": 1.9300215105494127e-06, "loss": 0.6123, "step": 2067 }, { "epoch": 1.7405848937513149, "grad_norm": 107.66278839111328, "learning_rate": 1.9278474042932703e-06, "loss": 0.7594, "step": 2068 }, { "epoch": 1.7414264674942141, "grad_norm": 69.9992446899414, "learning_rate": 1.925673754564532e-06, "loss": 0.8265, "step": 2069 }, { "epoch": 1.7422680412371134, "grad_norm": 68.45075225830078, "learning_rate": 1.9235005630975794e-06, "loss": 0.7149, "step": 2070 }, { "epoch": 1.7431096149800127, "grad_norm": 70.57120513916016, "learning_rate": 1.921327831626428e-06, "loss": 0.814, "step": 2071 }, { "epoch": 1.743951188722912, "grad_norm": 52.48041915893555, "learning_rate": 1.9191555618847257e-06, "loss": 0.6502, "step": 2072 }, { "epoch": 1.744792762465811, "grad_norm": 58.065547943115234, "learning_rate": 1.9169837556057536e-06, "loss": 0.7074, "step": 2073 }, { "epoch": 1.7456343362087103, "grad_norm": 67.77338409423828, "learning_rate": 1.9148124145224205e-06, "loss": 0.7019, "step": 2074 }, { "epoch": 1.7464759099516094, "grad_norm": 63.35089874267578, "learning_rate": 1.9126415403672663e-06, "loss": 0.6285, "step": 2075 }, { "epoch": 1.7473174836945087, "grad_norm": 82.49103546142578, "learning_rate": 1.9104711348724573e-06, "loss": 0.8969, "step": 2076 }, { "epoch": 1.748159057437408, "grad_norm": 59.18016052246094, "learning_rate": 1.908301199769786e-06, "loss": 0.5908, "step": 2077 }, { "epoch": 1.7490006311803072, "grad_norm": 76.43251037597656, "learning_rate": 1.9061317367906694e-06, "loss": 0.97, "step": 2078 }, { "epoch": 1.7498422049232065, "grad_norm": 66.44189453125, "learning_rate": 1.9039627476661484e-06, "loss": 0.6332, "step": 2079 }, { "epoch": 1.7506837786661056, "grad_norm": 57.63457107543945, "learning_rate": 1.9017942341268853e-06, "loss": 0.7882, "step": 2080 }, { "epoch": 1.7515253524090049, "grad_norm": 68.29576873779297, "learning_rate": 1.8996261979031627e-06, "loss": 0.8223, "step": 2081 }, { "epoch": 1.752366926151904, "grad_norm": 64.6242446899414, "learning_rate": 1.897458640724883e-06, "loss": 0.7203, "step": 2082 }, { "epoch": 1.7532084998948032, "grad_norm": 57.31889724731445, "learning_rate": 1.8952915643215659e-06, "loss": 0.5814, "step": 2083 }, { "epoch": 1.7540500736377025, "grad_norm": 65.94902801513672, "learning_rate": 1.8931249704223477e-06, "loss": 0.7261, "step": 2084 }, { "epoch": 1.7548916473806018, "grad_norm": 55.26048278808594, "learning_rate": 1.890958860755979e-06, "loss": 0.7176, "step": 2085 }, { "epoch": 1.755733221123501, "grad_norm": 74.53553009033203, "learning_rate": 1.8887932370508257e-06, "loss": 0.8328, "step": 2086 }, { "epoch": 1.7565747948664001, "grad_norm": 55.07794952392578, "learning_rate": 1.8866281010348641e-06, "loss": 0.7367, "step": 2087 }, { "epoch": 1.7574163686092994, "grad_norm": 68.0139389038086, "learning_rate": 1.884463454435682e-06, "loss": 0.8022, "step": 2088 }, { "epoch": 1.7582579423521985, "grad_norm": 59.173465728759766, "learning_rate": 1.8822992989804772e-06, "loss": 0.6236, "step": 2089 }, { "epoch": 1.7590995160950977, "grad_norm": 60.35981369018555, "learning_rate": 1.8801356363960556e-06, "loss": 0.694, "step": 2090 }, { "epoch": 1.759941089837997, "grad_norm": 62.7681770324707, "learning_rate": 1.8779724684088284e-06, "loss": 0.6438, "step": 2091 }, { "epoch": 1.7607826635808963, "grad_norm": 54.19706726074219, "learning_rate": 1.8758097967448143e-06, "loss": 0.4935, "step": 2092 }, { "epoch": 1.7616242373237956, "grad_norm": 75.36858367919922, "learning_rate": 1.8736476231296344e-06, "loss": 0.8261, "step": 2093 }, { "epoch": 1.7624658110666949, "grad_norm": 58.33365249633789, "learning_rate": 1.871485949288513e-06, "loss": 0.7537, "step": 2094 }, { "epoch": 1.763307384809594, "grad_norm": 66.77893829345703, "learning_rate": 1.8693247769462757e-06, "loss": 0.6265, "step": 2095 }, { "epoch": 1.7641489585524932, "grad_norm": 85.07715606689453, "learning_rate": 1.8671641078273483e-06, "loss": 0.6661, "step": 2096 }, { "epoch": 1.7649905322953923, "grad_norm": 67.86998748779297, "learning_rate": 1.8650039436557532e-06, "loss": 0.7555, "step": 2097 }, { "epoch": 1.7658321060382915, "grad_norm": 75.93537902832031, "learning_rate": 1.8628442861551115e-06, "loss": 0.6683, "step": 2098 }, { "epoch": 1.7666736797811908, "grad_norm": 64.42959594726562, "learning_rate": 1.8606851370486407e-06, "loss": 0.8069, "step": 2099 }, { "epoch": 1.7675152535240901, "grad_norm": 122.44876098632812, "learning_rate": 1.8585264980591511e-06, "loss": 0.6346, "step": 2100 }, { "epoch": 1.7683568272669894, "grad_norm": 54.100425720214844, "learning_rate": 1.8563683709090463e-06, "loss": 0.5711, "step": 2101 }, { "epoch": 1.7691984010098885, "grad_norm": 57.87714385986328, "learning_rate": 1.8542107573203224e-06, "loss": 0.575, "step": 2102 }, { "epoch": 1.7700399747527877, "grad_norm": 58.95901107788086, "learning_rate": 1.8520536590145649e-06, "loss": 0.6917, "step": 2103 }, { "epoch": 1.7708815484956868, "grad_norm": 76.74815368652344, "learning_rate": 1.8498970777129488e-06, "loss": 0.784, "step": 2104 }, { "epoch": 1.771723122238586, "grad_norm": 84.59148406982422, "learning_rate": 1.8477410151362352e-06, "loss": 0.8739, "step": 2105 }, { "epoch": 1.7725646959814854, "grad_norm": 55.115177154541016, "learning_rate": 1.845585473004773e-06, "loss": 0.5427, "step": 2106 }, { "epoch": 1.7734062697243846, "grad_norm": 54.50093460083008, "learning_rate": 1.843430453038494e-06, "loss": 0.6424, "step": 2107 }, { "epoch": 1.774247843467284, "grad_norm": 75.58036041259766, "learning_rate": 1.8412759569569159e-06, "loss": 0.7836, "step": 2108 }, { "epoch": 1.7750894172101832, "grad_norm": 83.39302062988281, "learning_rate": 1.839121986479136e-06, "loss": 0.6811, "step": 2109 }, { "epoch": 1.7759309909530823, "grad_norm": 66.5321044921875, "learning_rate": 1.8369685433238333e-06, "loss": 0.4496, "step": 2110 }, { "epoch": 1.7767725646959813, "grad_norm": 69.394775390625, "learning_rate": 1.8348156292092659e-06, "loss": 0.8011, "step": 2111 }, { "epoch": 1.7776141384388806, "grad_norm": 73.21894073486328, "learning_rate": 1.8326632458532695e-06, "loss": 0.8602, "step": 2112 }, { "epoch": 1.77845571218178, "grad_norm": 57.68058776855469, "learning_rate": 1.8305113949732573e-06, "loss": 0.6457, "step": 2113 }, { "epoch": 1.7792972859246792, "grad_norm": 62.928504943847656, "learning_rate": 1.828360078286216e-06, "loss": 0.7843, "step": 2114 }, { "epoch": 1.7801388596675785, "grad_norm": 56.39037322998047, "learning_rate": 1.8262092975087076e-06, "loss": 0.5209, "step": 2115 }, { "epoch": 1.7809804334104777, "grad_norm": 59.846981048583984, "learning_rate": 1.824059054356865e-06, "loss": 0.7138, "step": 2116 }, { "epoch": 1.7818220071533768, "grad_norm": 58.56599807739258, "learning_rate": 1.8219093505463944e-06, "loss": 0.8367, "step": 2117 }, { "epoch": 1.782663580896276, "grad_norm": 64.58570098876953, "learning_rate": 1.8197601877925685e-06, "loss": 0.7276, "step": 2118 }, { "epoch": 1.7835051546391751, "grad_norm": 59.75026321411133, "learning_rate": 1.8176115678102314e-06, "loss": 0.781, "step": 2119 }, { "epoch": 1.7843467283820744, "grad_norm": 89.35760498046875, "learning_rate": 1.8154634923137916e-06, "loss": 0.9094, "step": 2120 }, { "epoch": 1.7851883021249737, "grad_norm": 56.73500442504883, "learning_rate": 1.8133159630172243e-06, "loss": 0.657, "step": 2121 }, { "epoch": 1.786029875867873, "grad_norm": 73.52031707763672, "learning_rate": 1.8111689816340697e-06, "loss": 0.7437, "step": 2122 }, { "epoch": 1.7868714496107723, "grad_norm": 84.91339111328125, "learning_rate": 1.8090225498774295e-06, "loss": 0.8218, "step": 2123 }, { "epoch": 1.7877130233536713, "grad_norm": 84.25125885009766, "learning_rate": 1.806876669459967e-06, "loss": 0.7746, "step": 2124 }, { "epoch": 1.7885545970965706, "grad_norm": 67.50647735595703, "learning_rate": 1.8047313420939063e-06, "loss": 0.707, "step": 2125 }, { "epoch": 1.7893961708394697, "grad_norm": 69.52603149414062, "learning_rate": 1.8025865694910296e-06, "loss": 1.0066, "step": 2126 }, { "epoch": 1.790237744582369, "grad_norm": 50.65574645996094, "learning_rate": 1.800442353362677e-06, "loss": 0.5244, "step": 2127 }, { "epoch": 1.7910793183252682, "grad_norm": 66.66259765625, "learning_rate": 1.7982986954197434e-06, "loss": 0.8929, "step": 2128 }, { "epoch": 1.7919208920681675, "grad_norm": 74.70015716552734, "learning_rate": 1.7961555973726801e-06, "loss": 0.6895, "step": 2129 }, { "epoch": 1.7927624658110668, "grad_norm": 67.10930633544922, "learning_rate": 1.7940130609314898e-06, "loss": 0.76, "step": 2130 }, { "epoch": 1.793604039553966, "grad_norm": 66.6857681274414, "learning_rate": 1.7918710878057282e-06, "loss": 0.626, "step": 2131 }, { "epoch": 1.7944456132968651, "grad_norm": 63.96589660644531, "learning_rate": 1.7897296797045018e-06, "loss": 0.649, "step": 2132 }, { "epoch": 1.7952871870397642, "grad_norm": 57.037078857421875, "learning_rate": 1.7875888383364654e-06, "loss": 0.7395, "step": 2133 }, { "epoch": 1.7961287607826635, "grad_norm": 68.79027557373047, "learning_rate": 1.7854485654098216e-06, "loss": 0.9107, "step": 2134 }, { "epoch": 1.7969703345255628, "grad_norm": 72.25999450683594, "learning_rate": 1.7833088626323204e-06, "loss": 0.5938, "step": 2135 }, { "epoch": 1.797811908268462, "grad_norm": 66.58133697509766, "learning_rate": 1.7811697317112559e-06, "loss": 0.7369, "step": 2136 }, { "epoch": 1.7986534820113613, "grad_norm": 58.60413360595703, "learning_rate": 1.7790311743534663e-06, "loss": 0.6582, "step": 2137 }, { "epoch": 1.7994950557542606, "grad_norm": 76.59229278564453, "learning_rate": 1.7768931922653324e-06, "loss": 0.803, "step": 2138 }, { "epoch": 1.8003366294971597, "grad_norm": 54.43056106567383, "learning_rate": 1.7747557871527755e-06, "loss": 0.6851, "step": 2139 }, { "epoch": 1.801178203240059, "grad_norm": 68.74797058105469, "learning_rate": 1.7726189607212574e-06, "loss": 0.624, "step": 2140 }, { "epoch": 1.802019776982958, "grad_norm": 67.81819152832031, "learning_rate": 1.7704827146757758e-06, "loss": 0.8438, "step": 2141 }, { "epoch": 1.8028613507258573, "grad_norm": 58.922576904296875, "learning_rate": 1.7683470507208683e-06, "loss": 0.7595, "step": 2142 }, { "epoch": 1.8037029244687566, "grad_norm": 59.12992477416992, "learning_rate": 1.7662119705606061e-06, "loss": 0.8185, "step": 2143 }, { "epoch": 1.8045444982116559, "grad_norm": 57.263153076171875, "learning_rate": 1.7640774758985958e-06, "loss": 0.7675, "step": 2144 }, { "epoch": 1.8053860719545551, "grad_norm": 54.074764251708984, "learning_rate": 1.7619435684379754e-06, "loss": 0.5313, "step": 2145 }, { "epoch": 1.8062276456974542, "grad_norm": 64.90428924560547, "learning_rate": 1.7598102498814162e-06, "loss": 0.6689, "step": 2146 }, { "epoch": 1.8070692194403535, "grad_norm": 73.16099548339844, "learning_rate": 1.7576775219311181e-06, "loss": 0.7455, "step": 2147 }, { "epoch": 1.8079107931832525, "grad_norm": 64.098876953125, "learning_rate": 1.7555453862888108e-06, "loss": 0.524, "step": 2148 }, { "epoch": 1.8087523669261518, "grad_norm": 56.784202575683594, "learning_rate": 1.75341384465575e-06, "loss": 0.6232, "step": 2149 }, { "epoch": 1.809593940669051, "grad_norm": 86.95955657958984, "learning_rate": 1.7512828987327185e-06, "loss": 0.8403, "step": 2150 }, { "epoch": 1.8104355144119504, "grad_norm": 97.39279174804688, "learning_rate": 1.7491525502200237e-06, "loss": 0.9458, "step": 2151 }, { "epoch": 1.8112770881548497, "grad_norm": 60.38591766357422, "learning_rate": 1.7470228008174964e-06, "loss": 0.7653, "step": 2152 }, { "epoch": 1.812118661897749, "grad_norm": 62.22574996948242, "learning_rate": 1.7448936522244888e-06, "loss": 0.7138, "step": 2153 }, { "epoch": 1.812960235640648, "grad_norm": 66.15081024169922, "learning_rate": 1.7427651061398743e-06, "loss": 0.6096, "step": 2154 }, { "epoch": 1.813801809383547, "grad_norm": 76.48223114013672, "learning_rate": 1.7406371642620451e-06, "loss": 0.9563, "step": 2155 }, { "epoch": 1.8146433831264464, "grad_norm": 59.00917053222656, "learning_rate": 1.7385098282889115e-06, "loss": 0.7433, "step": 2156 }, { "epoch": 1.8154849568693456, "grad_norm": 72.06829833984375, "learning_rate": 1.7363830999179004e-06, "loss": 0.8908, "step": 2157 }, { "epoch": 1.816326530612245, "grad_norm": 71.82278442382812, "learning_rate": 1.7342569808459541e-06, "loss": 0.8161, "step": 2158 }, { "epoch": 1.8171681043551442, "grad_norm": 72.5805435180664, "learning_rate": 1.7321314727695279e-06, "loss": 0.7711, "step": 2159 }, { "epoch": 1.8180096780980435, "grad_norm": 61.565006256103516, "learning_rate": 1.730006577384591e-06, "loss": 0.7901, "step": 2160 }, { "epoch": 1.8188512518409425, "grad_norm": 76.91460418701172, "learning_rate": 1.727882296386622e-06, "loss": 0.7171, "step": 2161 }, { "epoch": 1.8196928255838418, "grad_norm": 62.8109130859375, "learning_rate": 1.72575863147061e-06, "loss": 0.8574, "step": 2162 }, { "epoch": 1.8205343993267409, "grad_norm": 64.93528747558594, "learning_rate": 1.7236355843310535e-06, "loss": 0.8641, "step": 2163 }, { "epoch": 1.8213759730696402, "grad_norm": 61.25568389892578, "learning_rate": 1.7215131566619567e-06, "loss": 0.6553, "step": 2164 }, { "epoch": 1.8222175468125394, "grad_norm": 61.76858901977539, "learning_rate": 1.7193913501568298e-06, "loss": 0.7845, "step": 2165 }, { "epoch": 1.8230591205554387, "grad_norm": 65.39705657958984, "learning_rate": 1.7172701665086882e-06, "loss": 0.6767, "step": 2166 }, { "epoch": 1.823900694298338, "grad_norm": 68.75941467285156, "learning_rate": 1.7151496074100499e-06, "loss": 0.8299, "step": 2167 }, { "epoch": 1.824742268041237, "grad_norm": 67.01436614990234, "learning_rate": 1.7130296745529334e-06, "loss": 0.6991, "step": 2168 }, { "epoch": 1.8255838417841364, "grad_norm": 59.835479736328125, "learning_rate": 1.7109103696288593e-06, "loss": 0.6792, "step": 2169 }, { "epoch": 1.8264254155270354, "grad_norm": 73.47028350830078, "learning_rate": 1.7087916943288457e-06, "loss": 0.9334, "step": 2170 }, { "epoch": 1.8272669892699347, "grad_norm": 67.07899475097656, "learning_rate": 1.70667365034341e-06, "loss": 0.7283, "step": 2171 }, { "epoch": 1.828108563012834, "grad_norm": 53.611454010009766, "learning_rate": 1.704556239362564e-06, "loss": 0.611, "step": 2172 }, { "epoch": 1.8289501367557333, "grad_norm": 66.5965347290039, "learning_rate": 1.702439463075816e-06, "loss": 0.7968, "step": 2173 }, { "epoch": 1.8297917104986325, "grad_norm": 60.807987213134766, "learning_rate": 1.700323323172166e-06, "loss": 0.5905, "step": 2174 }, { "epoch": 1.8306332842415318, "grad_norm": 73.02649688720703, "learning_rate": 1.6982078213401083e-06, "loss": 0.5884, "step": 2175 }, { "epoch": 1.8314748579844309, "grad_norm": 63.57632064819336, "learning_rate": 1.6960929592676268e-06, "loss": 0.6883, "step": 2176 }, { "epoch": 1.83231643172733, "grad_norm": 63.84977340698242, "learning_rate": 1.6939787386421951e-06, "loss": 0.6668, "step": 2177 }, { "epoch": 1.8331580054702292, "grad_norm": 76.59451293945312, "learning_rate": 1.6918651611507752e-06, "loss": 0.7618, "step": 2178 }, { "epoch": 1.8339995792131285, "grad_norm": 60.18157958984375, "learning_rate": 1.6897522284798165e-06, "loss": 0.6913, "step": 2179 }, { "epoch": 1.8348411529560278, "grad_norm": 68.55657958984375, "learning_rate": 1.6876399423152525e-06, "loss": 0.7262, "step": 2180 }, { "epoch": 1.835682726698927, "grad_norm": 68.82770538330078, "learning_rate": 1.685528304342502e-06, "loss": 0.8593, "step": 2181 }, { "epoch": 1.8365243004418264, "grad_norm": 62.159400939941406, "learning_rate": 1.6834173162464662e-06, "loss": 0.6638, "step": 2182 }, { "epoch": 1.8373658741847254, "grad_norm": 85.95907592773438, "learning_rate": 1.6813069797115277e-06, "loss": 0.8754, "step": 2183 }, { "epoch": 1.8382074479276247, "grad_norm": 66.26023864746094, "learning_rate": 1.6791972964215502e-06, "loss": 0.6105, "step": 2184 }, { "epoch": 1.8390490216705238, "grad_norm": 62.79727554321289, "learning_rate": 1.6770882680598732e-06, "loss": 0.6177, "step": 2185 }, { "epoch": 1.839890595413423, "grad_norm": 54.53998947143555, "learning_rate": 1.6749798963093167e-06, "loss": 0.7558, "step": 2186 }, { "epoch": 1.8407321691563223, "grad_norm": 82.19766235351562, "learning_rate": 1.672872182852176e-06, "loss": 0.8732, "step": 2187 }, { "epoch": 1.8415737428992216, "grad_norm": 85.38581085205078, "learning_rate": 1.670765129370221e-06, "loss": 0.6903, "step": 2188 }, { "epoch": 1.8424153166421209, "grad_norm": 110.23831939697266, "learning_rate": 1.6686587375446944e-06, "loss": 1.0256, "step": 2189 }, { "epoch": 1.84325689038502, "grad_norm": 79.85043334960938, "learning_rate": 1.6665530090563118e-06, "loss": 1.0438, "step": 2190 }, { "epoch": 1.8440984641279192, "grad_norm": 62.538761138916016, "learning_rate": 1.6644479455852594e-06, "loss": 0.7486, "step": 2191 }, { "epoch": 1.8449400378708183, "grad_norm": 69.23018646240234, "learning_rate": 1.6623435488111915e-06, "loss": 0.7848, "step": 2192 }, { "epoch": 1.8457816116137176, "grad_norm": 60.08961868286133, "learning_rate": 1.6602398204132325e-06, "loss": 0.5754, "step": 2193 }, { "epoch": 1.8466231853566168, "grad_norm": 64.63136291503906, "learning_rate": 1.6581367620699723e-06, "loss": 0.7078, "step": 2194 }, { "epoch": 1.8474647590995161, "grad_norm": 61.370113372802734, "learning_rate": 1.6560343754594658e-06, "loss": 0.6519, "step": 2195 }, { "epoch": 1.8483063328424154, "grad_norm": 57.12778854370117, "learning_rate": 1.6539326622592328e-06, "loss": 0.5928, "step": 2196 }, { "epoch": 1.8491479065853147, "grad_norm": 58.63680648803711, "learning_rate": 1.6518316241462556e-06, "loss": 0.7515, "step": 2197 }, { "epoch": 1.8499894803282138, "grad_norm": 57.86534118652344, "learning_rate": 1.6497312627969775e-06, "loss": 0.6928, "step": 2198 }, { "epoch": 1.850831054071113, "grad_norm": 77.90965270996094, "learning_rate": 1.6476315798873015e-06, "loss": 0.9641, "step": 2199 }, { "epoch": 1.851672627814012, "grad_norm": 53.760345458984375, "learning_rate": 1.6455325770925905e-06, "loss": 0.6735, "step": 2200 }, { "epoch": 1.8525142015569114, "grad_norm": 99.69976043701172, "learning_rate": 1.6434342560876636e-06, "loss": 1.1589, "step": 2201 }, { "epoch": 1.8533557752998107, "grad_norm": 64.50054168701172, "learning_rate": 1.6413366185467965e-06, "loss": 0.7704, "step": 2202 }, { "epoch": 1.85419734904271, "grad_norm": 54.418067932128906, "learning_rate": 1.6392396661437188e-06, "loss": 0.6597, "step": 2203 }, { "epoch": 1.8550389227856092, "grad_norm": 59.051536560058594, "learning_rate": 1.6371434005516146e-06, "loss": 0.6146, "step": 2204 }, { "epoch": 1.8558804965285083, "grad_norm": 89.67035675048828, "learning_rate": 1.6350478234431187e-06, "loss": 0.88, "step": 2205 }, { "epoch": 1.8567220702714076, "grad_norm": 54.25040817260742, "learning_rate": 1.6329529364903171e-06, "loss": 0.5999, "step": 2206 }, { "epoch": 1.8575636440143066, "grad_norm": 83.056884765625, "learning_rate": 1.6308587413647454e-06, "loss": 0.8977, "step": 2207 }, { "epoch": 1.858405217757206, "grad_norm": 77.22631072998047, "learning_rate": 1.628765239737387e-06, "loss": 0.7787, "step": 2208 }, { "epoch": 1.8592467915001052, "grad_norm": 68.0757827758789, "learning_rate": 1.626672433278672e-06, "loss": 0.8219, "step": 2209 }, { "epoch": 1.8600883652430045, "grad_norm": 77.16809844970703, "learning_rate": 1.6245803236584755e-06, "loss": 0.7615, "step": 2210 }, { "epoch": 1.8609299389859038, "grad_norm": 65.29447937011719, "learning_rate": 1.6224889125461168e-06, "loss": 0.6396, "step": 2211 }, { "epoch": 1.8617715127288028, "grad_norm": 85.58439636230469, "learning_rate": 1.620398201610358e-06, "loss": 0.8211, "step": 2212 }, { "epoch": 1.862613086471702, "grad_norm": 74.19503784179688, "learning_rate": 1.6183081925194028e-06, "loss": 0.7028, "step": 2213 }, { "epoch": 1.8634546602146012, "grad_norm": 74.5559310913086, "learning_rate": 1.6162188869408935e-06, "loss": 0.7414, "step": 2214 }, { "epoch": 1.8642962339575004, "grad_norm": 74.82536315917969, "learning_rate": 1.614130286541913e-06, "loss": 0.7786, "step": 2215 }, { "epoch": 1.8651378077003997, "grad_norm": 73.92768096923828, "learning_rate": 1.6120423929889798e-06, "loss": 0.828, "step": 2216 }, { "epoch": 1.865979381443299, "grad_norm": 65.88578033447266, "learning_rate": 1.6099552079480502e-06, "loss": 0.6288, "step": 2217 }, { "epoch": 1.8668209551861983, "grad_norm": 73.93595886230469, "learning_rate": 1.6078687330845124e-06, "loss": 0.7046, "step": 2218 }, { "epoch": 1.8676625289290976, "grad_norm": 68.39221954345703, "learning_rate": 1.605782970063191e-06, "loss": 0.7618, "step": 2219 }, { "epoch": 1.8685041026719966, "grad_norm": 61.75254440307617, "learning_rate": 1.603697920548341e-06, "loss": 0.7601, "step": 2220 }, { "epoch": 1.869345676414896, "grad_norm": 80.1451187133789, "learning_rate": 1.601613586203648e-06, "loss": 0.8277, "step": 2221 }, { "epoch": 1.870187250157795, "grad_norm": 71.23120880126953, "learning_rate": 1.5995299686922281e-06, "loss": 0.7912, "step": 2222 }, { "epoch": 1.8710288239006942, "grad_norm": 77.2643814086914, "learning_rate": 1.5974470696766236e-06, "loss": 0.7487, "step": 2223 }, { "epoch": 1.8718703976435935, "grad_norm": 70.44129943847656, "learning_rate": 1.5953648908188057e-06, "loss": 0.7631, "step": 2224 }, { "epoch": 1.8727119713864928, "grad_norm": 66.0828857421875, "learning_rate": 1.593283433780169e-06, "loss": 0.5765, "step": 2225 }, { "epoch": 1.873553545129392, "grad_norm": 56.88987731933594, "learning_rate": 1.5912027002215334e-06, "loss": 0.6828, "step": 2226 }, { "epoch": 1.8743951188722912, "grad_norm": 78.15242004394531, "learning_rate": 1.5891226918031405e-06, "loss": 0.9266, "step": 2227 }, { "epoch": 1.8752366926151904, "grad_norm": 71.46236419677734, "learning_rate": 1.587043410184656e-06, "loss": 0.782, "step": 2228 }, { "epoch": 1.8760782663580895, "grad_norm": 58.207584381103516, "learning_rate": 1.5849648570251609e-06, "loss": 0.6328, "step": 2229 }, { "epoch": 1.8769198401009888, "grad_norm": 70.5195541381836, "learning_rate": 1.5828870339831581e-06, "loss": 0.6815, "step": 2230 }, { "epoch": 1.877761413843888, "grad_norm": 86.5136489868164, "learning_rate": 1.5808099427165685e-06, "loss": 0.6899, "step": 2231 }, { "epoch": 1.8786029875867873, "grad_norm": 69.32208251953125, "learning_rate": 1.5787335848827276e-06, "loss": 0.8282, "step": 2232 }, { "epoch": 1.8794445613296866, "grad_norm": 118.74019622802734, "learning_rate": 1.5766579621383862e-06, "loss": 0.745, "step": 2233 }, { "epoch": 1.8802861350725857, "grad_norm": 63.61555862426758, "learning_rate": 1.5745830761397087e-06, "loss": 0.6513, "step": 2234 }, { "epoch": 1.881127708815485, "grad_norm": 67.32327270507812, "learning_rate": 1.5725089285422717e-06, "loss": 0.7587, "step": 2235 }, { "epoch": 1.881969282558384, "grad_norm": 44.54482650756836, "learning_rate": 1.5704355210010618e-06, "loss": 0.6055, "step": 2236 }, { "epoch": 1.8828108563012833, "grad_norm": 72.0876235961914, "learning_rate": 1.5683628551704755e-06, "loss": 0.8302, "step": 2237 }, { "epoch": 1.8836524300441826, "grad_norm": 69.63773345947266, "learning_rate": 1.5662909327043186e-06, "loss": 0.7435, "step": 2238 }, { "epoch": 1.8844940037870819, "grad_norm": 74.05252075195312, "learning_rate": 1.564219755255802e-06, "loss": 0.7485, "step": 2239 }, { "epoch": 1.8853355775299812, "grad_norm": 67.08032989501953, "learning_rate": 1.5621493244775432e-06, "loss": 0.7111, "step": 2240 }, { "epoch": 1.8861771512728804, "grad_norm": 61.99530029296875, "learning_rate": 1.5600796420215641e-06, "loss": 0.6949, "step": 2241 }, { "epoch": 1.8870187250157795, "grad_norm": 86.49744415283203, "learning_rate": 1.5580107095392889e-06, "loss": 0.9101, "step": 2242 }, { "epoch": 1.8878602987586788, "grad_norm": 88.76123046875, "learning_rate": 1.5559425286815434e-06, "loss": 0.6711, "step": 2243 }, { "epoch": 1.8887018725015778, "grad_norm": 65.21556854248047, "learning_rate": 1.5538751010985536e-06, "loss": 0.6391, "step": 2244 }, { "epoch": 1.8895434462444771, "grad_norm": 71.65559387207031, "learning_rate": 1.5518084284399453e-06, "loss": 0.6321, "step": 2245 }, { "epoch": 1.8903850199873764, "grad_norm": 75.29273986816406, "learning_rate": 1.5497425123547416e-06, "loss": 0.7162, "step": 2246 }, { "epoch": 1.8912265937302757, "grad_norm": 71.78056335449219, "learning_rate": 1.5476773544913609e-06, "loss": 0.7598, "step": 2247 }, { "epoch": 1.892068167473175, "grad_norm": 63.79790115356445, "learning_rate": 1.545612956497618e-06, "loss": 0.6993, "step": 2248 }, { "epoch": 1.892909741216074, "grad_norm": 57.5540885925293, "learning_rate": 1.5435493200207202e-06, "loss": 0.6928, "step": 2249 }, { "epoch": 1.8937513149589733, "grad_norm": 63.54572296142578, "learning_rate": 1.5414864467072676e-06, "loss": 0.7804, "step": 2250 }, { "epoch": 1.8945928887018724, "grad_norm": 55.57268524169922, "learning_rate": 1.5394243382032526e-06, "loss": 0.5618, "step": 2251 }, { "epoch": 1.8954344624447717, "grad_norm": 55.75808334350586, "learning_rate": 1.5373629961540549e-06, "loss": 0.656, "step": 2252 }, { "epoch": 1.896276036187671, "grad_norm": 80.97319793701172, "learning_rate": 1.535302422204445e-06, "loss": 0.8692, "step": 2253 }, { "epoch": 1.8971176099305702, "grad_norm": 74.42911529541016, "learning_rate": 1.5332426179985791e-06, "loss": 0.7752, "step": 2254 }, { "epoch": 1.8979591836734695, "grad_norm": 78.87651062011719, "learning_rate": 1.5311835851799997e-06, "loss": 0.8745, "step": 2255 }, { "epoch": 1.8988007574163686, "grad_norm": 66.46542358398438, "learning_rate": 1.529125325391634e-06, "loss": 0.8628, "step": 2256 }, { "epoch": 1.8996423311592678, "grad_norm": 59.02506637573242, "learning_rate": 1.5270678402757914e-06, "loss": 0.704, "step": 2257 }, { "epoch": 1.900483904902167, "grad_norm": 78.31338500976562, "learning_rate": 1.5250111314741644e-06, "loss": 0.7776, "step": 2258 }, { "epoch": 1.9013254786450662, "grad_norm": 58.565773010253906, "learning_rate": 1.5229552006278259e-06, "loss": 0.5033, "step": 2259 }, { "epoch": 1.9021670523879655, "grad_norm": 72.6630630493164, "learning_rate": 1.520900049377227e-06, "loss": 0.6292, "step": 2260 }, { "epoch": 1.9030086261308647, "grad_norm": 85.1754150390625, "learning_rate": 1.5188456793621987e-06, "loss": 0.8217, "step": 2261 }, { "epoch": 1.903850199873764, "grad_norm": 56.88100814819336, "learning_rate": 1.5167920922219459e-06, "loss": 0.5536, "step": 2262 }, { "epoch": 1.9046917736166633, "grad_norm": 60.75733184814453, "learning_rate": 1.5147392895950508e-06, "loss": 0.7346, "step": 2263 }, { "epoch": 1.9055333473595624, "grad_norm": 59.35121536254883, "learning_rate": 1.5126872731194698e-06, "loss": 0.7426, "step": 2264 }, { "epoch": 1.9063749211024617, "grad_norm": 85.046142578125, "learning_rate": 1.5106360444325307e-06, "loss": 0.7998, "step": 2265 }, { "epoch": 1.9072164948453607, "grad_norm": 56.76308059692383, "learning_rate": 1.5085856051709338e-06, "loss": 0.6138, "step": 2266 }, { "epoch": 1.90805806858826, "grad_norm": 62.7496223449707, "learning_rate": 1.5065359569707494e-06, "loss": 0.6749, "step": 2267 }, { "epoch": 1.9088996423311593, "grad_norm": 63.72941970825195, "learning_rate": 1.5044871014674157e-06, "loss": 0.6104, "step": 2268 }, { "epoch": 1.9097412160740586, "grad_norm": 67.2701187133789, "learning_rate": 1.5024390402957392e-06, "loss": 0.7457, "step": 2269 }, { "epoch": 1.9105827898169578, "grad_norm": 62.71966552734375, "learning_rate": 1.5003917750898928e-06, "loss": 0.7163, "step": 2270 }, { "epoch": 1.911424363559857, "grad_norm": 51.9465217590332, "learning_rate": 1.4983453074834136e-06, "loss": 0.5534, "step": 2271 }, { "epoch": 1.9122659373027562, "grad_norm": 65.47900390625, "learning_rate": 1.4962996391092038e-06, "loss": 0.8001, "step": 2272 }, { "epoch": 1.9131075110456552, "grad_norm": 75.9030532836914, "learning_rate": 1.4942547715995242e-06, "loss": 0.8953, "step": 2273 }, { "epoch": 1.9139490847885545, "grad_norm": 74.58604431152344, "learning_rate": 1.4922107065860009e-06, "loss": 0.7393, "step": 2274 }, { "epoch": 1.9147906585314538, "grad_norm": 72.80326080322266, "learning_rate": 1.4901674456996167e-06, "loss": 0.8472, "step": 2275 }, { "epoch": 1.915632232274353, "grad_norm": 69.97894287109375, "learning_rate": 1.4881249905707151e-06, "loss": 0.7002, "step": 2276 }, { "epoch": 1.9164738060172524, "grad_norm": 63.33930206298828, "learning_rate": 1.486083342828995e-06, "loss": 0.8482, "step": 2277 }, { "epoch": 1.9173153797601514, "grad_norm": 78.06482696533203, "learning_rate": 1.4840425041035112e-06, "loss": 0.824, "step": 2278 }, { "epoch": 1.9181569535030507, "grad_norm": 65.84388732910156, "learning_rate": 1.4820024760226743e-06, "loss": 0.6919, "step": 2279 }, { "epoch": 1.9189985272459498, "grad_norm": 62.79630661010742, "learning_rate": 1.4799632602142461e-06, "loss": 0.6189, "step": 2280 }, { "epoch": 1.919840100988849, "grad_norm": 58.355709075927734, "learning_rate": 1.4779248583053412e-06, "loss": 0.6215, "step": 2281 }, { "epoch": 1.9206816747317483, "grad_norm": 63.83758544921875, "learning_rate": 1.4758872719224259e-06, "loss": 0.8167, "step": 2282 }, { "epoch": 1.9215232484746476, "grad_norm": 60.83406066894531, "learning_rate": 1.4738505026913142e-06, "loss": 0.7229, "step": 2283 }, { "epoch": 1.922364822217547, "grad_norm": 66.23934173583984, "learning_rate": 1.4718145522371685e-06, "loss": 0.7516, "step": 2284 }, { "epoch": 1.9232063959604462, "grad_norm": 71.29521942138672, "learning_rate": 1.4697794221844986e-06, "loss": 0.6389, "step": 2285 }, { "epoch": 1.9240479697033452, "grad_norm": 57.2097053527832, "learning_rate": 1.4677451141571586e-06, "loss": 0.6305, "step": 2286 }, { "epoch": 1.9248895434462445, "grad_norm": 83.174072265625, "learning_rate": 1.4657116297783472e-06, "loss": 0.7423, "step": 2287 }, { "epoch": 1.9257311171891436, "grad_norm": 60.69619369506836, "learning_rate": 1.4636789706706062e-06, "loss": 0.6056, "step": 2288 }, { "epoch": 1.9265726909320429, "grad_norm": 73.55635070800781, "learning_rate": 1.4616471384558185e-06, "loss": 0.9381, "step": 2289 }, { "epoch": 1.9274142646749421, "grad_norm": 70.80742645263672, "learning_rate": 1.4596161347552074e-06, "loss": 0.4943, "step": 2290 }, { "epoch": 1.9282558384178414, "grad_norm": 61.46763229370117, "learning_rate": 1.4575859611893356e-06, "loss": 0.7255, "step": 2291 }, { "epoch": 1.9290974121607407, "grad_norm": 73.93780517578125, "learning_rate": 1.4555566193781013e-06, "loss": 0.8606, "step": 2292 }, { "epoch": 1.9299389859036398, "grad_norm": 79.21611785888672, "learning_rate": 1.4535281109407426e-06, "loss": 0.9096, "step": 2293 }, { "epoch": 1.930780559646539, "grad_norm": 76.24579620361328, "learning_rate": 1.451500437495829e-06, "loss": 0.7817, "step": 2294 }, { "epoch": 1.9316221333894381, "grad_norm": 72.4844741821289, "learning_rate": 1.4494736006612668e-06, "loss": 0.7279, "step": 2295 }, { "epoch": 1.9324637071323374, "grad_norm": 70.94538879394531, "learning_rate": 1.4474476020542916e-06, "loss": 0.7566, "step": 2296 }, { "epoch": 1.9333052808752367, "grad_norm": 63.378597259521484, "learning_rate": 1.4454224432914744e-06, "loss": 0.7534, "step": 2297 }, { "epoch": 1.934146854618136, "grad_norm": 71.35863494873047, "learning_rate": 1.443398125988712e-06, "loss": 0.8834, "step": 2298 }, { "epoch": 1.9349884283610352, "grad_norm": 70.34723663330078, "learning_rate": 1.441374651761231e-06, "loss": 0.8989, "step": 2299 }, { "epoch": 1.9358300021039345, "grad_norm": 57.3638801574707, "learning_rate": 1.4393520222235868e-06, "loss": 0.5678, "step": 2300 }, { "epoch": 1.9366715758468336, "grad_norm": 57.16642761230469, "learning_rate": 1.4373302389896587e-06, "loss": 0.755, "step": 2301 }, { "epoch": 1.9375131495897326, "grad_norm": 63.867515563964844, "learning_rate": 1.4353093036726534e-06, "loss": 0.5697, "step": 2302 }, { "epoch": 1.938354723332632, "grad_norm": 72.28782653808594, "learning_rate": 1.4332892178850972e-06, "loss": 0.8688, "step": 2303 }, { "epoch": 1.9391962970755312, "grad_norm": 60.91026306152344, "learning_rate": 1.4312699832388428e-06, "loss": 0.6112, "step": 2304 }, { "epoch": 1.9400378708184305, "grad_norm": 51.18401336669922, "learning_rate": 1.4292516013450595e-06, "loss": 0.5192, "step": 2305 }, { "epoch": 1.9408794445613298, "grad_norm": 60.899539947509766, "learning_rate": 1.4272340738142406e-06, "loss": 0.6408, "step": 2306 }, { "epoch": 1.941721018304229, "grad_norm": 60.82663345336914, "learning_rate": 1.4252174022561937e-06, "loss": 0.6614, "step": 2307 }, { "epoch": 1.9425625920471281, "grad_norm": 55.083961486816406, "learning_rate": 1.4232015882800467e-06, "loss": 0.7775, "step": 2308 }, { "epoch": 1.9434041657900274, "grad_norm": 67.09272003173828, "learning_rate": 1.4211866334942403e-06, "loss": 0.6193, "step": 2309 }, { "epoch": 1.9442457395329265, "grad_norm": 78.66014862060547, "learning_rate": 1.4191725395065326e-06, "loss": 0.876, "step": 2310 }, { "epoch": 1.9450873132758257, "grad_norm": 69.67564392089844, "learning_rate": 1.417159307923992e-06, "loss": 0.6553, "step": 2311 }, { "epoch": 1.945928887018725, "grad_norm": 76.12279510498047, "learning_rate": 1.415146940353e-06, "loss": 0.8337, "step": 2312 }, { "epoch": 1.9467704607616243, "grad_norm": 72.69770050048828, "learning_rate": 1.4131354383992497e-06, "loss": 0.7773, "step": 2313 }, { "epoch": 1.9476120345045236, "grad_norm": 90.46495056152344, "learning_rate": 1.4111248036677413e-06, "loss": 0.744, "step": 2314 }, { "epoch": 1.9484536082474226, "grad_norm": 64.65869140625, "learning_rate": 1.409115037762786e-06, "loss": 0.6586, "step": 2315 }, { "epoch": 1.949295181990322, "grad_norm": 270.3387756347656, "learning_rate": 1.4071061422879996e-06, "loss": 1.2586, "step": 2316 }, { "epoch": 1.950136755733221, "grad_norm": 75.58063507080078, "learning_rate": 1.405098118846303e-06, "loss": 0.6379, "step": 2317 }, { "epoch": 1.9509783294761203, "grad_norm": 71.5633544921875, "learning_rate": 1.4030909690399218e-06, "loss": 0.6761, "step": 2318 }, { "epoch": 1.9518199032190195, "grad_norm": 72.47621154785156, "learning_rate": 1.4010846944703868e-06, "loss": 0.8109, "step": 2319 }, { "epoch": 1.9526614769619188, "grad_norm": 77.21080017089844, "learning_rate": 1.3990792967385265e-06, "loss": 0.8222, "step": 2320 }, { "epoch": 1.9535030507048181, "grad_norm": 66.69099426269531, "learning_rate": 1.397074777444473e-06, "loss": 0.7484, "step": 2321 }, { "epoch": 1.9543446244477174, "grad_norm": 72.10345458984375, "learning_rate": 1.395071138187655e-06, "loss": 0.78, "step": 2322 }, { "epoch": 1.9551861981906165, "grad_norm": 75.02869415283203, "learning_rate": 1.3930683805668022e-06, "loss": 0.8524, "step": 2323 }, { "epoch": 1.9560277719335155, "grad_norm": 65.50741577148438, "learning_rate": 1.3910665061799366e-06, "loss": 0.5282, "step": 2324 }, { "epoch": 1.9568693456764148, "grad_norm": 68.63650512695312, "learning_rate": 1.38906551662438e-06, "loss": 0.6473, "step": 2325 }, { "epoch": 1.957710919419314, "grad_norm": 86.39122772216797, "learning_rate": 1.3870654134967438e-06, "loss": 0.9033, "step": 2326 }, { "epoch": 1.9585524931622134, "grad_norm": 60.809146881103516, "learning_rate": 1.385066198392936e-06, "loss": 0.484, "step": 2327 }, { "epoch": 1.9593940669051126, "grad_norm": 64.33039855957031, "learning_rate": 1.3830678729081526e-06, "loss": 0.8284, "step": 2328 }, { "epoch": 1.960235640648012, "grad_norm": 61.75801086425781, "learning_rate": 1.3810704386368834e-06, "loss": 0.7528, "step": 2329 }, { "epoch": 1.961077214390911, "grad_norm": 93.34917449951172, "learning_rate": 1.379073897172904e-06, "loss": 0.9207, "step": 2330 }, { "epoch": 1.9619187881338103, "grad_norm": 58.484214782714844, "learning_rate": 1.3770782501092778e-06, "loss": 0.5444, "step": 2331 }, { "epoch": 1.9627603618767093, "grad_norm": 68.56871032714844, "learning_rate": 1.3750834990383572e-06, "loss": 0.8807, "step": 2332 }, { "epoch": 1.9636019356196086, "grad_norm": 53.71663284301758, "learning_rate": 1.3730896455517762e-06, "loss": 0.5559, "step": 2333 }, { "epoch": 1.964443509362508, "grad_norm": 56.146427154541016, "learning_rate": 1.3710966912404566e-06, "loss": 0.6799, "step": 2334 }, { "epoch": 1.9652850831054072, "grad_norm": 92.54448699951172, "learning_rate": 1.3691046376945983e-06, "loss": 0.7772, "step": 2335 }, { "epoch": 1.9661266568483065, "grad_norm": 73.73626708984375, "learning_rate": 1.367113486503686e-06, "loss": 0.6257, "step": 2336 }, { "epoch": 1.9669682305912055, "grad_norm": 74.22389221191406, "learning_rate": 1.3651232392564822e-06, "loss": 0.7922, "step": 2337 }, { "epoch": 1.9678098043341048, "grad_norm": 74.11428833007812, "learning_rate": 1.3631338975410302e-06, "loss": 0.8146, "step": 2338 }, { "epoch": 1.9686513780770039, "grad_norm": 84.78289031982422, "learning_rate": 1.361145462944648e-06, "loss": 0.8851, "step": 2339 }, { "epoch": 1.9694929518199031, "grad_norm": 85.74117279052734, "learning_rate": 1.3591579370539334e-06, "loss": 0.8934, "step": 2340 }, { "epoch": 1.9703345255628024, "grad_norm": 59.489906311035156, "learning_rate": 1.3571713214547546e-06, "loss": 0.7244, "step": 2341 }, { "epoch": 1.9711760993057017, "grad_norm": 72.61528015136719, "learning_rate": 1.3551856177322587e-06, "loss": 0.8778, "step": 2342 }, { "epoch": 1.972017673048601, "grad_norm": 66.0748519897461, "learning_rate": 1.3532008274708605e-06, "loss": 0.7046, "step": 2343 }, { "epoch": 1.9728592467915003, "grad_norm": 52.10646057128906, "learning_rate": 1.351216952254248e-06, "loss": 0.5669, "step": 2344 }, { "epoch": 1.9737008205343993, "grad_norm": 72.16187286376953, "learning_rate": 1.3492339936653798e-06, "loss": 0.9286, "step": 2345 }, { "epoch": 1.9745423942772984, "grad_norm": 63.71145248413086, "learning_rate": 1.3472519532864815e-06, "loss": 0.7192, "step": 2346 }, { "epoch": 1.9753839680201977, "grad_norm": 65.3427505493164, "learning_rate": 1.3452708326990477e-06, "loss": 0.6995, "step": 2347 }, { "epoch": 1.976225541763097, "grad_norm": 74.12447357177734, "learning_rate": 1.343290633483837e-06, "loss": 0.7666, "step": 2348 }, { "epoch": 1.9770671155059962, "grad_norm": 70.56058502197266, "learning_rate": 1.341311357220876e-06, "loss": 0.8012, "step": 2349 }, { "epoch": 1.9779086892488955, "grad_norm": 63.311458587646484, "learning_rate": 1.3393330054894504e-06, "loss": 0.7388, "step": 2350 }, { "epoch": 1.9787502629917948, "grad_norm": 90.02980041503906, "learning_rate": 1.3373555798681137e-06, "loss": 1.0757, "step": 2351 }, { "epoch": 1.9795918367346939, "grad_norm": 65.29212188720703, "learning_rate": 1.3353790819346746e-06, "loss": 0.7583, "step": 2352 }, { "epoch": 1.9804334104775931, "grad_norm": 74.32219696044922, "learning_rate": 1.3334035132662065e-06, "loss": 0.7544, "step": 2353 }, { "epoch": 1.9812749842204922, "grad_norm": 60.60328674316406, "learning_rate": 1.3314288754390376e-06, "loss": 0.6951, "step": 2354 }, { "epoch": 1.9821165579633915, "grad_norm": 63.507564544677734, "learning_rate": 1.3294551700287571e-06, "loss": 0.6197, "step": 2355 }, { "epoch": 1.9829581317062908, "grad_norm": 51.96760559082031, "learning_rate": 1.327482398610207e-06, "loss": 0.6158, "step": 2356 }, { "epoch": 1.98379970544919, "grad_norm": 74.8888168334961, "learning_rate": 1.325510562757484e-06, "loss": 0.7534, "step": 2357 }, { "epoch": 1.9846412791920893, "grad_norm": 66.71295928955078, "learning_rate": 1.3235396640439422e-06, "loss": 0.8724, "step": 2358 }, { "epoch": 1.9854828529349884, "grad_norm": 66.0381851196289, "learning_rate": 1.3215697040421826e-06, "loss": 0.6828, "step": 2359 }, { "epoch": 1.9863244266778877, "grad_norm": 70.55795288085938, "learning_rate": 1.3196006843240633e-06, "loss": 0.8511, "step": 2360 }, { "epoch": 1.9871660004207867, "grad_norm": 63.4446907043457, "learning_rate": 1.3176326064606843e-06, "loss": 0.7485, "step": 2361 }, { "epoch": 1.988007574163686, "grad_norm": 77.5648193359375, "learning_rate": 1.3156654720224018e-06, "loss": 0.7181, "step": 2362 }, { "epoch": 1.9888491479065853, "grad_norm": 57.53008270263672, "learning_rate": 1.3136992825788142e-06, "loss": 0.6573, "step": 2363 }, { "epoch": 1.9896907216494846, "grad_norm": 57.448753356933594, "learning_rate": 1.3117340396987687e-06, "loss": 0.7521, "step": 2364 }, { "epoch": 1.9905322953923839, "grad_norm": 61.140846252441406, "learning_rate": 1.309769744950355e-06, "loss": 0.7161, "step": 2365 }, { "epoch": 1.9913738691352831, "grad_norm": 51.21188735961914, "learning_rate": 1.307806399900909e-06, "loss": 0.5591, "step": 2366 }, { "epoch": 1.9922154428781822, "grad_norm": 67.9342269897461, "learning_rate": 1.3058440061170047e-06, "loss": 0.7145, "step": 2367 }, { "epoch": 1.9930570166210813, "grad_norm": 80.97782897949219, "learning_rate": 1.3038825651644626e-06, "loss": 0.9007, "step": 2368 }, { "epoch": 1.9938985903639805, "grad_norm": 59.37981414794922, "learning_rate": 1.3019220786083376e-06, "loss": 0.7504, "step": 2369 }, { "epoch": 1.9947401641068798, "grad_norm": 74.92973327636719, "learning_rate": 1.299962548012927e-06, "loss": 0.8484, "step": 2370 }, { "epoch": 1.995581737849779, "grad_norm": 70.0312728881836, "learning_rate": 1.2980039749417618e-06, "loss": 0.751, "step": 2371 }, { "epoch": 1.9964233115926784, "grad_norm": 61.27229690551758, "learning_rate": 1.2960463609576135e-06, "loss": 0.8234, "step": 2372 }, { "epoch": 1.9972648853355777, "grad_norm": 93.65629577636719, "learning_rate": 1.2940897076224837e-06, "loss": 0.7821, "step": 2373 }, { "epoch": 1.9981064590784767, "grad_norm": 93.00342559814453, "learning_rate": 1.2921340164976092e-06, "loss": 0.9813, "step": 2374 }, { "epoch": 1.998948032821376, "grad_norm": 66.82344055175781, "learning_rate": 1.2901792891434616e-06, "loss": 0.6761, "step": 2375 }, { "epoch": 1.999789606564275, "grad_norm": 63.64382553100586, "learning_rate": 1.2882255271197389e-06, "loss": 0.5141, "step": 2376 }, { "epoch": 2.0, "grad_norm": 197.7196807861328, "learning_rate": 1.2862727319853729e-06, "loss": 0.9352, "step": 2377 }, { "epoch": 2.0008415737428993, "grad_norm": 68.72442626953125, "learning_rate": 1.2843209052985205e-06, "loss": 0.7375, "step": 2378 }, { "epoch": 2.0016831474857986, "grad_norm": 73.65171813964844, "learning_rate": 1.2823700486165699e-06, "loss": 0.6918, "step": 2379 }, { "epoch": 2.002524721228698, "grad_norm": 71.76443481445312, "learning_rate": 1.2804201634961305e-06, "loss": 0.8057, "step": 2380 }, { "epoch": 2.0033662949715967, "grad_norm": 75.31938171386719, "learning_rate": 1.2784712514930415e-06, "loss": 0.7052, "step": 2381 }, { "epoch": 2.004207868714496, "grad_norm": 66.46198272705078, "learning_rate": 1.2765233141623602e-06, "loss": 0.7448, "step": 2382 }, { "epoch": 2.0050494424573952, "grad_norm": 46.86164855957031, "learning_rate": 1.2745763530583721e-06, "loss": 0.5279, "step": 2383 }, { "epoch": 2.0058910162002945, "grad_norm": 59.78047561645508, "learning_rate": 1.2726303697345781e-06, "loss": 0.6646, "step": 2384 }, { "epoch": 2.006732589943194, "grad_norm": 44.50751876831055, "learning_rate": 1.270685365743704e-06, "loss": 0.5851, "step": 2385 }, { "epoch": 2.007574163686093, "grad_norm": 64.33128356933594, "learning_rate": 1.2687413426376905e-06, "loss": 0.7577, "step": 2386 }, { "epoch": 2.0084157374289924, "grad_norm": 79.38579559326172, "learning_rate": 1.2667983019676961e-06, "loss": 0.8151, "step": 2387 }, { "epoch": 2.0092573111718917, "grad_norm": 56.066436767578125, "learning_rate": 1.264856245284098e-06, "loss": 0.5714, "step": 2388 }, { "epoch": 2.0100988849147905, "grad_norm": 78.3211441040039, "learning_rate": 1.2629151741364847e-06, "loss": 0.7935, "step": 2389 }, { "epoch": 2.0109404586576898, "grad_norm": 83.97476959228516, "learning_rate": 1.2609750900736622e-06, "loss": 0.7445, "step": 2390 }, { "epoch": 2.011782032400589, "grad_norm": 71.06140899658203, "learning_rate": 1.2590359946436443e-06, "loss": 0.8628, "step": 2391 }, { "epoch": 2.0126236061434883, "grad_norm": 57.746360778808594, "learning_rate": 1.2570978893936608e-06, "loss": 0.6732, "step": 2392 }, { "epoch": 2.0134651798863876, "grad_norm": 87.6710205078125, "learning_rate": 1.255160775870147e-06, "loss": 0.898, "step": 2393 }, { "epoch": 2.014306753629287, "grad_norm": 73.00880432128906, "learning_rate": 1.253224655618751e-06, "loss": 0.7046, "step": 2394 }, { "epoch": 2.015148327372186, "grad_norm": 56.755889892578125, "learning_rate": 1.2512895301843244e-06, "loss": 0.5928, "step": 2395 }, { "epoch": 2.015989901115085, "grad_norm": 70.450439453125, "learning_rate": 1.2493554011109289e-06, "loss": 0.5954, "step": 2396 }, { "epoch": 2.0168314748579843, "grad_norm": 63.83796310424805, "learning_rate": 1.247422269941827e-06, "loss": 0.65, "step": 2397 }, { "epoch": 2.0176730486008836, "grad_norm": 51.25503158569336, "learning_rate": 1.2454901382194895e-06, "loss": 0.5288, "step": 2398 }, { "epoch": 2.018514622343783, "grad_norm": 62.279022216796875, "learning_rate": 1.2435590074855863e-06, "loss": 0.6754, "step": 2399 }, { "epoch": 2.019356196086682, "grad_norm": 65.7779769897461, "learning_rate": 1.241628879280989e-06, "loss": 0.5908, "step": 2400 }, { "epoch": 2.0201977698295814, "grad_norm": 55.950439453125, "learning_rate": 1.2396997551457712e-06, "loss": 0.7067, "step": 2401 }, { "epoch": 2.0210393435724807, "grad_norm": 67.66270446777344, "learning_rate": 1.237771636619203e-06, "loss": 0.7854, "step": 2402 }, { "epoch": 2.0218809173153796, "grad_norm": 81.34107971191406, "learning_rate": 1.235844525239755e-06, "loss": 0.8033, "step": 2403 }, { "epoch": 2.022722491058279, "grad_norm": 81.92566680908203, "learning_rate": 1.2339184225450912e-06, "loss": 0.8637, "step": 2404 }, { "epoch": 2.023564064801178, "grad_norm": 69.52362060546875, "learning_rate": 1.2319933300720723e-06, "loss": 0.6985, "step": 2405 }, { "epoch": 2.0244056385440774, "grad_norm": 69.33657836914062, "learning_rate": 1.2300692493567515e-06, "loss": 0.6924, "step": 2406 }, { "epoch": 2.0252472122869767, "grad_norm": 58.44219207763672, "learning_rate": 1.2281461819343776e-06, "loss": 0.5913, "step": 2407 }, { "epoch": 2.026088786029876, "grad_norm": 62.818328857421875, "learning_rate": 1.2262241293393877e-06, "loss": 0.6074, "step": 2408 }, { "epoch": 2.0269303597727752, "grad_norm": 74.00282287597656, "learning_rate": 1.2243030931054117e-06, "loss": 0.7489, "step": 2409 }, { "epoch": 2.0277719335156745, "grad_norm": 73.24765014648438, "learning_rate": 1.2223830747652663e-06, "loss": 0.6461, "step": 2410 }, { "epoch": 2.0286135072585734, "grad_norm": 73.88299560546875, "learning_rate": 1.2204640758509588e-06, "loss": 0.8073, "step": 2411 }, { "epoch": 2.0294550810014726, "grad_norm": 83.63514709472656, "learning_rate": 1.218546097893679e-06, "loss": 0.8902, "step": 2412 }, { "epoch": 2.030296654744372, "grad_norm": 91.70342254638672, "learning_rate": 1.216629142423807e-06, "loss": 0.9449, "step": 2413 }, { "epoch": 2.031138228487271, "grad_norm": 75.2408218383789, "learning_rate": 1.2147132109709023e-06, "loss": 0.7626, "step": 2414 }, { "epoch": 2.0319798022301705, "grad_norm": 62.704891204833984, "learning_rate": 1.2127983050637116e-06, "loss": 0.5581, "step": 2415 }, { "epoch": 2.0328213759730698, "grad_norm": 61.1127815246582, "learning_rate": 1.210884426230159e-06, "loss": 0.5484, "step": 2416 }, { "epoch": 2.033662949715969, "grad_norm": 75.96637725830078, "learning_rate": 1.2089715759973537e-06, "loss": 0.8051, "step": 2417 }, { "epoch": 2.034504523458868, "grad_norm": 66.8486099243164, "learning_rate": 1.2070597558915804e-06, "loss": 0.6753, "step": 2418 }, { "epoch": 2.035346097201767, "grad_norm": 75.64108276367188, "learning_rate": 1.2051489674383027e-06, "loss": 0.9247, "step": 2419 }, { "epoch": 2.0361876709446665, "grad_norm": 67.49555969238281, "learning_rate": 1.203239212162163e-06, "loss": 0.7821, "step": 2420 }, { "epoch": 2.0370292446875657, "grad_norm": 52.557071685791016, "learning_rate": 1.2013304915869767e-06, "loss": 0.5832, "step": 2421 }, { "epoch": 2.037870818430465, "grad_norm": 61.864376068115234, "learning_rate": 1.1994228072357362e-06, "loss": 0.729, "step": 2422 }, { "epoch": 2.0387123921733643, "grad_norm": 55.379207611083984, "learning_rate": 1.1975161606306035e-06, "loss": 0.552, "step": 2423 }, { "epoch": 2.0395539659162636, "grad_norm": 83.2946548461914, "learning_rate": 1.1956105532929174e-06, "loss": 0.9164, "step": 2424 }, { "epoch": 2.0403955396591624, "grad_norm": 69.54570007324219, "learning_rate": 1.1937059867431825e-06, "loss": 0.8997, "step": 2425 }, { "epoch": 2.0412371134020617, "grad_norm": 68.32620239257812, "learning_rate": 1.1918024625010767e-06, "loss": 0.7443, "step": 2426 }, { "epoch": 2.042078687144961, "grad_norm": 61.021785736083984, "learning_rate": 1.1898999820854438e-06, "loss": 0.6541, "step": 2427 }, { "epoch": 2.0429202608878603, "grad_norm": 73.8634033203125, "learning_rate": 1.1879985470142969e-06, "loss": 0.6883, "step": 2428 }, { "epoch": 2.0437618346307596, "grad_norm": 73.61117553710938, "learning_rate": 1.1860981588048124e-06, "loss": 0.6408, "step": 2429 }, { "epoch": 2.044603408373659, "grad_norm": 69.99700164794922, "learning_rate": 1.1841988189733342e-06, "loss": 0.694, "step": 2430 }, { "epoch": 2.045444982116558, "grad_norm": 68.9922866821289, "learning_rate": 1.1823005290353672e-06, "loss": 0.6156, "step": 2431 }, { "epoch": 2.0462865558594574, "grad_norm": 71.89803314208984, "learning_rate": 1.1804032905055796e-06, "loss": 0.9109, "step": 2432 }, { "epoch": 2.0471281296023562, "grad_norm": 67.3031005859375, "learning_rate": 1.178507104897802e-06, "loss": 0.6873, "step": 2433 }, { "epoch": 2.0479697033452555, "grad_norm": 64.5659408569336, "learning_rate": 1.1766119737250217e-06, "loss": 0.7768, "step": 2434 }, { "epoch": 2.048811277088155, "grad_norm": 67.28711700439453, "learning_rate": 1.1747178984993887e-06, "loss": 0.609, "step": 2435 }, { "epoch": 2.049652850831054, "grad_norm": 72.9075927734375, "learning_rate": 1.1728248807322063e-06, "loss": 0.7171, "step": 2436 }, { "epoch": 2.0504944245739534, "grad_norm": 62.915245056152344, "learning_rate": 1.1709329219339374e-06, "loss": 0.6878, "step": 2437 }, { "epoch": 2.0513359983168526, "grad_norm": 66.7945785522461, "learning_rate": 1.1690420236141976e-06, "loss": 0.7398, "step": 2438 }, { "epoch": 2.052177572059752, "grad_norm": 81.240966796875, "learning_rate": 1.1671521872817584e-06, "loss": 0.7949, "step": 2439 }, { "epoch": 2.0530191458026508, "grad_norm": 69.78356170654297, "learning_rate": 1.1652634144445415e-06, "loss": 0.6583, "step": 2440 }, { "epoch": 2.05386071954555, "grad_norm": 68.70997619628906, "learning_rate": 1.1633757066096229e-06, "loss": 0.6203, "step": 2441 }, { "epoch": 2.0547022932884493, "grad_norm": 55.60271453857422, "learning_rate": 1.1614890652832267e-06, "loss": 0.57, "step": 2442 }, { "epoch": 2.0555438670313486, "grad_norm": 71.62289428710938, "learning_rate": 1.1596034919707253e-06, "loss": 0.6958, "step": 2443 }, { "epoch": 2.056385440774248, "grad_norm": 56.20205307006836, "learning_rate": 1.1577189881766422e-06, "loss": 0.6184, "step": 2444 }, { "epoch": 2.057227014517147, "grad_norm": 66.36256408691406, "learning_rate": 1.1558355554046434e-06, "loss": 0.7245, "step": 2445 }, { "epoch": 2.0580685882600465, "grad_norm": 88.98446655273438, "learning_rate": 1.153953195157545e-06, "loss": 0.8945, "step": 2446 }, { "epoch": 2.0589101620029453, "grad_norm": 63.11686325073242, "learning_rate": 1.152071908937303e-06, "loss": 0.6214, "step": 2447 }, { "epoch": 2.0597517357458446, "grad_norm": 65.78763580322266, "learning_rate": 1.150191698245019e-06, "loss": 0.7413, "step": 2448 }, { "epoch": 2.060593309488744, "grad_norm": 68.73076629638672, "learning_rate": 1.1483125645809343e-06, "loss": 0.6814, "step": 2449 }, { "epoch": 2.061434883231643, "grad_norm": 69.8797607421875, "learning_rate": 1.1464345094444338e-06, "loss": 0.6081, "step": 2450 }, { "epoch": 2.0622764569745424, "grad_norm": 73.53306579589844, "learning_rate": 1.1445575343340384e-06, "loss": 0.5956, "step": 2451 }, { "epoch": 2.0631180307174417, "grad_norm": 74.26734924316406, "learning_rate": 1.1426816407474113e-06, "loss": 0.7987, "step": 2452 }, { "epoch": 2.063959604460341, "grad_norm": 85.37826538085938, "learning_rate": 1.140806830181348e-06, "loss": 0.6925, "step": 2453 }, { "epoch": 2.0648011782032403, "grad_norm": 60.54194259643555, "learning_rate": 1.1389331041317844e-06, "loss": 0.6173, "step": 2454 }, { "epoch": 2.065642751946139, "grad_norm": 61.3946647644043, "learning_rate": 1.1370604640937873e-06, "loss": 0.6793, "step": 2455 }, { "epoch": 2.0664843256890384, "grad_norm": 73.27828979492188, "learning_rate": 1.1351889115615602e-06, "loss": 0.7443, "step": 2456 }, { "epoch": 2.0673258994319377, "grad_norm": 90.79389190673828, "learning_rate": 1.1333184480284353e-06, "loss": 0.7576, "step": 2457 }, { "epoch": 2.068167473174837, "grad_norm": 72.02031707763672, "learning_rate": 1.13144907498688e-06, "loss": 0.648, "step": 2458 }, { "epoch": 2.0690090469177362, "grad_norm": 72.92021179199219, "learning_rate": 1.1295807939284875e-06, "loss": 0.72, "step": 2459 }, { "epoch": 2.0698506206606355, "grad_norm": 67.07632446289062, "learning_rate": 1.1277136063439837e-06, "loss": 0.6992, "step": 2460 }, { "epoch": 2.070692194403535, "grad_norm": 63.59321594238281, "learning_rate": 1.1258475137232183e-06, "loss": 0.6769, "step": 2461 }, { "epoch": 2.0715337681464336, "grad_norm": 84.59838104248047, "learning_rate": 1.1239825175551682e-06, "loss": 0.8736, "step": 2462 }, { "epoch": 2.072375341889333, "grad_norm": 74.90754699707031, "learning_rate": 1.1221186193279387e-06, "loss": 0.7427, "step": 2463 }, { "epoch": 2.073216915632232, "grad_norm": 78.17488098144531, "learning_rate": 1.1202558205287539e-06, "loss": 0.6659, "step": 2464 }, { "epoch": 2.0740584893751315, "grad_norm": 104.38905334472656, "learning_rate": 1.1183941226439657e-06, "loss": 0.8333, "step": 2465 }, { "epoch": 2.0749000631180308, "grad_norm": 71.7644271850586, "learning_rate": 1.1165335271590427e-06, "loss": 1.2413, "step": 2466 }, { "epoch": 2.07574163686093, "grad_norm": 68.65503692626953, "learning_rate": 1.1146740355585787e-06, "loss": 0.6689, "step": 2467 }, { "epoch": 2.0765832106038293, "grad_norm": 80.6414566040039, "learning_rate": 1.1128156493262818e-06, "loss": 0.802, "step": 2468 }, { "epoch": 2.077424784346728, "grad_norm": 72.85330963134766, "learning_rate": 1.110958369944983e-06, "loss": 0.6855, "step": 2469 }, { "epoch": 2.0782663580896275, "grad_norm": 84.65975189208984, "learning_rate": 1.1091021988966261e-06, "loss": 0.8109, "step": 2470 }, { "epoch": 2.0791079318325267, "grad_norm": 85.22624969482422, "learning_rate": 1.1072471376622732e-06, "loss": 0.7163, "step": 2471 }, { "epoch": 2.079949505575426, "grad_norm": 66.9729232788086, "learning_rate": 1.1053931877220988e-06, "loss": 0.6104, "step": 2472 }, { "epoch": 2.0807910793183253, "grad_norm": 80.03248596191406, "learning_rate": 1.1035403505553933e-06, "loss": 0.6977, "step": 2473 }, { "epoch": 2.0816326530612246, "grad_norm": 70.52706146240234, "learning_rate": 1.1016886276405567e-06, "loss": 0.6296, "step": 2474 }, { "epoch": 2.082474226804124, "grad_norm": 74.24388885498047, "learning_rate": 1.0998380204551e-06, "loss": 0.8894, "step": 2475 }, { "epoch": 2.083315800547023, "grad_norm": 65.65377044677734, "learning_rate": 1.0979885304756465e-06, "loss": 0.5534, "step": 2476 }, { "epoch": 2.084157374289922, "grad_norm": 61.11000061035156, "learning_rate": 1.0961401591779246e-06, "loss": 0.7157, "step": 2477 }, { "epoch": 2.0849989480328213, "grad_norm": 60.91055679321289, "learning_rate": 1.0942929080367743e-06, "loss": 0.5756, "step": 2478 }, { "epoch": 2.0858405217757205, "grad_norm": 61.728511810302734, "learning_rate": 1.0924467785261366e-06, "loss": 0.6726, "step": 2479 }, { "epoch": 2.08668209551862, "grad_norm": 91.92497253417969, "learning_rate": 1.0906017721190632e-06, "loss": 0.9503, "step": 2480 }, { "epoch": 2.087523669261519, "grad_norm": 76.24710845947266, "learning_rate": 1.088757890287704e-06, "loss": 0.869, "step": 2481 }, { "epoch": 2.0883652430044184, "grad_norm": 64.9084701538086, "learning_rate": 1.086915134503317e-06, "loss": 0.6312, "step": 2482 }, { "epoch": 2.0892068167473177, "grad_norm": 102.50822448730469, "learning_rate": 1.0850735062362564e-06, "loss": 1.0367, "step": 2483 }, { "epoch": 2.0900483904902165, "grad_norm": 80.54660034179688, "learning_rate": 1.083233006955982e-06, "loss": 0.7209, "step": 2484 }, { "epoch": 2.090889964233116, "grad_norm": 68.0789566040039, "learning_rate": 1.0813936381310482e-06, "loss": 0.7194, "step": 2485 }, { "epoch": 2.091731537976015, "grad_norm": 70.04940032958984, "learning_rate": 1.0795554012291113e-06, "loss": 0.6234, "step": 2486 }, { "epoch": 2.0925731117189144, "grad_norm": 89.34388732910156, "learning_rate": 1.0777182977169211e-06, "loss": 0.8364, "step": 2487 }, { "epoch": 2.0934146854618136, "grad_norm": 74.53938293457031, "learning_rate": 1.0758823290603243e-06, "loss": 0.7133, "step": 2488 }, { "epoch": 2.094256259204713, "grad_norm": 71.77678680419922, "learning_rate": 1.0740474967242637e-06, "loss": 0.7494, "step": 2489 }, { "epoch": 2.095097832947612, "grad_norm": 65.68650817871094, "learning_rate": 1.072213802172772e-06, "loss": 0.5725, "step": 2490 }, { "epoch": 2.095939406690511, "grad_norm": 75.179443359375, "learning_rate": 1.0703812468689784e-06, "loss": 0.7371, "step": 2491 }, { "epoch": 2.0967809804334103, "grad_norm": 56.12419891357422, "learning_rate": 1.0685498322750993e-06, "loss": 0.7039, "step": 2492 }, { "epoch": 2.0976225541763096, "grad_norm": 82.42550659179688, "learning_rate": 1.0667195598524423e-06, "loss": 0.795, "step": 2493 }, { "epoch": 2.098464127919209, "grad_norm": 75.12179565429688, "learning_rate": 1.064890431061403e-06, "loss": 0.8092, "step": 2494 }, { "epoch": 2.099305701662108, "grad_norm": 76.6459732055664, "learning_rate": 1.063062447361467e-06, "loss": 0.7452, "step": 2495 }, { "epoch": 2.1001472754050075, "grad_norm": 87.87409210205078, "learning_rate": 1.0612356102112025e-06, "loss": 1.1205, "step": 2496 }, { "epoch": 2.1009888491479067, "grad_norm": 65.37862396240234, "learning_rate": 1.0594099210682667e-06, "loss": 0.7343, "step": 2497 }, { "epoch": 2.101830422890806, "grad_norm": 73.89498901367188, "learning_rate": 1.0575853813893966e-06, "loss": 0.6213, "step": 2498 }, { "epoch": 2.102671996633705, "grad_norm": 62.228633880615234, "learning_rate": 1.0557619926304164e-06, "loss": 0.6226, "step": 2499 }, { "epoch": 2.103513570376604, "grad_norm": 71.1729736328125, "learning_rate": 1.0539397562462278e-06, "loss": 0.7481, "step": 2500 }, { "epoch": 2.1043551441195034, "grad_norm": 108.12519836425781, "learning_rate": 1.0521186736908171e-06, "loss": 0.8752, "step": 2501 }, { "epoch": 2.1051967178624027, "grad_norm": 69.51334381103516, "learning_rate": 1.0502987464172457e-06, "loss": 0.7707, "step": 2502 }, { "epoch": 2.106038291605302, "grad_norm": 74.9326171875, "learning_rate": 1.0484799758776573e-06, "loss": 0.7145, "step": 2503 }, { "epoch": 2.1068798653482013, "grad_norm": 81.75310516357422, "learning_rate": 1.0466623635232699e-06, "loss": 0.7077, "step": 2504 }, { "epoch": 2.1077214390911005, "grad_norm": 63.8576545715332, "learning_rate": 1.044845910804377e-06, "loss": 0.4491, "step": 2505 }, { "epoch": 2.1085630128339994, "grad_norm": 68.5901870727539, "learning_rate": 1.0430306191703504e-06, "loss": 0.6742, "step": 2506 }, { "epoch": 2.1094045865768987, "grad_norm": 65.15589141845703, "learning_rate": 1.0412164900696303e-06, "loss": 0.6712, "step": 2507 }, { "epoch": 2.110246160319798, "grad_norm": 66.7618637084961, "learning_rate": 1.0394035249497347e-06, "loss": 0.6602, "step": 2508 }, { "epoch": 2.1110877340626972, "grad_norm": 73.76443481445312, "learning_rate": 1.0375917252572477e-06, "loss": 0.7051, "step": 2509 }, { "epoch": 2.1119293078055965, "grad_norm": 62.10883331298828, "learning_rate": 1.035781092437828e-06, "loss": 0.6902, "step": 2510 }, { "epoch": 2.112770881548496, "grad_norm": 72.45101165771484, "learning_rate": 1.0339716279361993e-06, "loss": 0.8273, "step": 2511 }, { "epoch": 2.113612455291395, "grad_norm": 77.44570922851562, "learning_rate": 1.0321633331961573e-06, "loss": 0.7852, "step": 2512 }, { "epoch": 2.1144540290342944, "grad_norm": 73.0498046875, "learning_rate": 1.0303562096605597e-06, "loss": 0.6374, "step": 2513 }, { "epoch": 2.115295602777193, "grad_norm": 62.08723449707031, "learning_rate": 1.028550258771334e-06, "loss": 0.621, "step": 2514 }, { "epoch": 2.1161371765200925, "grad_norm": 66.72229766845703, "learning_rate": 1.0267454819694677e-06, "loss": 0.6706, "step": 2515 }, { "epoch": 2.1169787502629918, "grad_norm": 77.8282241821289, "learning_rate": 1.0249418806950166e-06, "loss": 0.6352, "step": 2516 }, { "epoch": 2.117820324005891, "grad_norm": 61.727569580078125, "learning_rate": 1.0231394563870938e-06, "loss": 0.65, "step": 2517 }, { "epoch": 2.1186618977487903, "grad_norm": 59.97420883178711, "learning_rate": 1.0213382104838752e-06, "loss": 0.585, "step": 2518 }, { "epoch": 2.1195034714916896, "grad_norm": 86.899169921875, "learning_rate": 1.0195381444225977e-06, "loss": 0.8229, "step": 2519 }, { "epoch": 2.120345045234589, "grad_norm": 90.7455062866211, "learning_rate": 1.0177392596395542e-06, "loss": 0.8266, "step": 2520 }, { "epoch": 2.1211866189774877, "grad_norm": 70.87030792236328, "learning_rate": 1.0159415575700976e-06, "loss": 0.7682, "step": 2521 }, { "epoch": 2.122028192720387, "grad_norm": 103.03855895996094, "learning_rate": 1.014145039648635e-06, "loss": 0.8716, "step": 2522 }, { "epoch": 2.1228697664632863, "grad_norm": 71.49374389648438, "learning_rate": 1.0123497073086303e-06, "loss": 0.9124, "step": 2523 }, { "epoch": 2.1237113402061856, "grad_norm": 63.0020637512207, "learning_rate": 1.0105555619825997e-06, "loss": 0.5423, "step": 2524 }, { "epoch": 2.124552913949085, "grad_norm": 73.98238372802734, "learning_rate": 1.0087626051021142e-06, "loss": 0.8097, "step": 2525 }, { "epoch": 2.125394487691984, "grad_norm": 75.4309310913086, "learning_rate": 1.0069708380977948e-06, "loss": 0.869, "step": 2526 }, { "epoch": 2.1262360614348834, "grad_norm": 70.3682632446289, "learning_rate": 1.0051802623993147e-06, "loss": 0.7256, "step": 2527 }, { "epoch": 2.1270776351777823, "grad_norm": 68.74978637695312, "learning_rate": 1.0033908794353947e-06, "loss": 0.5867, "step": 2528 }, { "epoch": 2.1279192089206815, "grad_norm": 70.8469467163086, "learning_rate": 1.001602690633806e-06, "loss": 0.7549, "step": 2529 }, { "epoch": 2.128760782663581, "grad_norm": 80.92948150634766, "learning_rate": 9.998156974213652e-07, "loss": 0.7448, "step": 2530 }, { "epoch": 2.12960235640648, "grad_norm": 60.65598678588867, "learning_rate": 9.980299012239347e-07, "loss": 0.5948, "step": 2531 }, { "epoch": 2.1304439301493794, "grad_norm": 48.92744064331055, "learning_rate": 9.962453034664246e-07, "loss": 0.5996, "step": 2532 }, { "epoch": 2.1312855038922787, "grad_norm": 65.03863525390625, "learning_rate": 9.944619055727848e-07, "loss": 0.673, "step": 2533 }, { "epoch": 2.132127077635178, "grad_norm": 65.4809799194336, "learning_rate": 9.926797089660116e-07, "loss": 0.5711, "step": 2534 }, { "epoch": 2.132968651378077, "grad_norm": 71.00261688232422, "learning_rate": 9.908987150681402e-07, "loss": 0.8347, "step": 2535 }, { "epoch": 2.133810225120976, "grad_norm": 67.29861450195312, "learning_rate": 9.891189253002466e-07, "loss": 0.6725, "step": 2536 }, { "epoch": 2.1346517988638753, "grad_norm": 65.30101013183594, "learning_rate": 9.873403410824463e-07, "loss": 0.8104, "step": 2537 }, { "epoch": 2.1354933726067746, "grad_norm": 84.99264526367188, "learning_rate": 9.85562963833894e-07, "loss": 0.648, "step": 2538 }, { "epoch": 2.136334946349674, "grad_norm": 68.99024200439453, "learning_rate": 9.83786794972779e-07, "loss": 0.6247, "step": 2539 }, { "epoch": 2.137176520092573, "grad_norm": 69.75712585449219, "learning_rate": 9.820118359163288e-07, "loss": 0.7422, "step": 2540 }, { "epoch": 2.1380180938354725, "grad_norm": 59.413421630859375, "learning_rate": 9.802380880808032e-07, "loss": 0.5832, "step": 2541 }, { "epoch": 2.1388596675783718, "grad_norm": 60.82857131958008, "learning_rate": 9.784655528814987e-07, "loss": 0.5496, "step": 2542 }, { "epoch": 2.1397012413212706, "grad_norm": 83.707763671875, "learning_rate": 9.766942317327403e-07, "loss": 0.7174, "step": 2543 }, { "epoch": 2.14054281506417, "grad_norm": 48.741363525390625, "learning_rate": 9.749241260478879e-07, "loss": 0.3006, "step": 2544 }, { "epoch": 2.141384388807069, "grad_norm": 69.54522705078125, "learning_rate": 9.731552372393287e-07, "loss": 0.7039, "step": 2545 }, { "epoch": 2.1422259625499684, "grad_norm": 71.36038970947266, "learning_rate": 9.713875667184819e-07, "loss": 0.7666, "step": 2546 }, { "epoch": 2.1430675362928677, "grad_norm": 97.82860565185547, "learning_rate": 9.696211158957914e-07, "loss": 0.8601, "step": 2547 }, { "epoch": 2.143909110035767, "grad_norm": 77.60707092285156, "learning_rate": 9.678558861807305e-07, "loss": 0.7382, "step": 2548 }, { "epoch": 2.1447506837786663, "grad_norm": 75.60054779052734, "learning_rate": 9.660918789817969e-07, "loss": 0.6905, "step": 2549 }, { "epoch": 2.145592257521565, "grad_norm": 84.70235443115234, "learning_rate": 9.64329095706512e-07, "loss": 0.5114, "step": 2550 }, { "epoch": 2.1464338312644644, "grad_norm": 75.57403564453125, "learning_rate": 9.625675377614235e-07, "loss": 0.6717, "step": 2551 }, { "epoch": 2.1472754050073637, "grad_norm": 64.758056640625, "learning_rate": 9.608072065520976e-07, "loss": 0.7525, "step": 2552 }, { "epoch": 2.148116978750263, "grad_norm": 62.87696838378906, "learning_rate": 9.590481034831255e-07, "loss": 0.6162, "step": 2553 }, { "epoch": 2.1489585524931623, "grad_norm": 75.20830535888672, "learning_rate": 9.572902299581151e-07, "loss": 0.6972, "step": 2554 }, { "epoch": 2.1498001262360615, "grad_norm": 71.37078857421875, "learning_rate": 9.555335873796963e-07, "loss": 0.7619, "step": 2555 }, { "epoch": 2.150641699978961, "grad_norm": 73.39781188964844, "learning_rate": 9.53778177149513e-07, "loss": 0.8032, "step": 2556 }, { "epoch": 2.15148327372186, "grad_norm": 91.35111236572266, "learning_rate": 9.520240006682307e-07, "loss": 0.6181, "step": 2557 }, { "epoch": 2.152324847464759, "grad_norm": 64.5626449584961, "learning_rate": 9.502710593355252e-07, "loss": 0.7019, "step": 2558 }, { "epoch": 2.153166421207658, "grad_norm": 74.30403137207031, "learning_rate": 9.485193545500918e-07, "loss": 0.6625, "step": 2559 }, { "epoch": 2.1540079949505575, "grad_norm": 70.13634490966797, "learning_rate": 9.467688877096346e-07, "loss": 0.6766, "step": 2560 }, { "epoch": 2.154849568693457, "grad_norm": 63.890769958496094, "learning_rate": 9.450196602108735e-07, "loss": 0.6317, "step": 2561 }, { "epoch": 2.155691142436356, "grad_norm": 66.91095733642578, "learning_rate": 9.432716734495378e-07, "loss": 0.6279, "step": 2562 }, { "epoch": 2.1565327161792553, "grad_norm": 59.37250518798828, "learning_rate": 9.415249288203662e-07, "loss": 0.7143, "step": 2563 }, { "epoch": 2.1573742899221546, "grad_norm": 96.72491455078125, "learning_rate": 9.397794277171083e-07, "loss": 0.8763, "step": 2564 }, { "epoch": 2.1582158636650535, "grad_norm": 76.71916961669922, "learning_rate": 9.38035171532519e-07, "loss": 0.8947, "step": 2565 }, { "epoch": 2.1590574374079528, "grad_norm": 88.00267791748047, "learning_rate": 9.362921616583628e-07, "loss": 0.7661, "step": 2566 }, { "epoch": 2.159899011150852, "grad_norm": 82.27729034423828, "learning_rate": 9.345503994854066e-07, "loss": 0.8139, "step": 2567 }, { "epoch": 2.1607405848937513, "grad_norm": 53.226661682128906, "learning_rate": 9.328098864034246e-07, "loss": 0.5347, "step": 2568 }, { "epoch": 2.1615821586366506, "grad_norm": 76.812744140625, "learning_rate": 9.310706238011915e-07, "loss": 0.7518, "step": 2569 }, { "epoch": 2.16242373237955, "grad_norm": 82.11021423339844, "learning_rate": 9.293326130664878e-07, "loss": 0.7259, "step": 2570 }, { "epoch": 2.163265306122449, "grad_norm": 60.77939987182617, "learning_rate": 9.275958555860909e-07, "loss": 0.5585, "step": 2571 }, { "epoch": 2.164106879865348, "grad_norm": 58.36905288696289, "learning_rate": 9.258603527457824e-07, "loss": 0.5984, "step": 2572 }, { "epoch": 2.1649484536082473, "grad_norm": 88.13652038574219, "learning_rate": 9.241261059303385e-07, "loss": 0.703, "step": 2573 }, { "epoch": 2.1657900273511466, "grad_norm": 74.98705291748047, "learning_rate": 9.223931165235381e-07, "loss": 0.6954, "step": 2574 }, { "epoch": 2.166631601094046, "grad_norm": 59.873783111572266, "learning_rate": 9.206613859081529e-07, "loss": 0.689, "step": 2575 }, { "epoch": 2.167473174836945, "grad_norm": 63.8530158996582, "learning_rate": 9.189309154659504e-07, "loss": 0.6467, "step": 2576 }, { "epoch": 2.1683147485798444, "grad_norm": 64.32845306396484, "learning_rate": 9.172017065776956e-07, "loss": 0.6098, "step": 2577 }, { "epoch": 2.1691563223227437, "grad_norm": 61.077857971191406, "learning_rate": 9.154737606231435e-07, "loss": 0.6059, "step": 2578 }, { "epoch": 2.1699978960656425, "grad_norm": 83.24121856689453, "learning_rate": 9.137470789810449e-07, "loss": 0.6332, "step": 2579 }, { "epoch": 2.170839469808542, "grad_norm": 69.9027328491211, "learning_rate": 9.120216630291365e-07, "loss": 0.7594, "step": 2580 }, { "epoch": 2.171681043551441, "grad_norm": 60.3941650390625, "learning_rate": 9.102975141441514e-07, "loss": 0.5929, "step": 2581 }, { "epoch": 2.1725226172943404, "grad_norm": 61.47883224487305, "learning_rate": 9.085746337018061e-07, "loss": 0.732, "step": 2582 }, { "epoch": 2.1733641910372397, "grad_norm": 98.2105941772461, "learning_rate": 9.068530230768099e-07, "loss": 0.7443, "step": 2583 }, { "epoch": 2.174205764780139, "grad_norm": 65.96863555908203, "learning_rate": 9.051326836428548e-07, "loss": 0.6803, "step": 2584 }, { "epoch": 2.175047338523038, "grad_norm": 91.28777313232422, "learning_rate": 9.034136167726215e-07, "loss": 0.7161, "step": 2585 }, { "epoch": 2.1758889122659375, "grad_norm": 86.08289337158203, "learning_rate": 9.016958238377726e-07, "loss": 0.7628, "step": 2586 }, { "epoch": 2.1767304860088363, "grad_norm": 65.36969757080078, "learning_rate": 8.999793062089571e-07, "loss": 0.689, "step": 2587 }, { "epoch": 2.1775720597517356, "grad_norm": 90.16263580322266, "learning_rate": 8.982640652558037e-07, "loss": 0.8336, "step": 2588 }, { "epoch": 2.178413633494635, "grad_norm": 74.20494842529297, "learning_rate": 8.965501023469251e-07, "loss": 0.5702, "step": 2589 }, { "epoch": 2.179255207237534, "grad_norm": 65.81613159179688, "learning_rate": 8.948374188499109e-07, "loss": 0.6538, "step": 2590 }, { "epoch": 2.1800967809804335, "grad_norm": 86.37291717529297, "learning_rate": 8.93126016131334e-07, "loss": 0.7678, "step": 2591 }, { "epoch": 2.1809383547233328, "grad_norm": 84.37739562988281, "learning_rate": 8.914158955567421e-07, "loss": 0.8994, "step": 2592 }, { "epoch": 2.181779928466232, "grad_norm": 102.40328979492188, "learning_rate": 8.897070584906597e-07, "loss": 0.7576, "step": 2593 }, { "epoch": 2.182621502209131, "grad_norm": 55.37407684326172, "learning_rate": 8.879995062965902e-07, "loss": 0.5465, "step": 2594 }, { "epoch": 2.18346307595203, "grad_norm": 68.9563980102539, "learning_rate": 8.862932403370084e-07, "loss": 0.7728, "step": 2595 }, { "epoch": 2.1843046496949294, "grad_norm": 69.24717712402344, "learning_rate": 8.845882619733658e-07, "loss": 0.6903, "step": 2596 }, { "epoch": 2.1851462234378287, "grad_norm": 94.22429656982422, "learning_rate": 8.828845725660836e-07, "loss": 0.9523, "step": 2597 }, { "epoch": 2.185987797180728, "grad_norm": 69.71115112304688, "learning_rate": 8.811821734745577e-07, "loss": 0.5651, "step": 2598 }, { "epoch": 2.1868293709236273, "grad_norm": 78.55773162841797, "learning_rate": 8.794810660571506e-07, "loss": 0.7817, "step": 2599 }, { "epoch": 2.1876709446665266, "grad_norm": 73.78901672363281, "learning_rate": 8.777812516711984e-07, "loss": 0.7284, "step": 2600 }, { "epoch": 2.188512518409426, "grad_norm": 66.3755111694336, "learning_rate": 8.760827316730011e-07, "loss": 0.5779, "step": 2601 }, { "epoch": 2.1893540921523247, "grad_norm": 61.488487243652344, "learning_rate": 8.743855074178306e-07, "loss": 0.6865, "step": 2602 }, { "epoch": 2.190195665895224, "grad_norm": 80.92797088623047, "learning_rate": 8.726895802599202e-07, "loss": 0.787, "step": 2603 }, { "epoch": 2.1910372396381232, "grad_norm": 79.75867462158203, "learning_rate": 8.709949515524724e-07, "loss": 0.8862, "step": 2604 }, { "epoch": 2.1918788133810225, "grad_norm": 79.74058532714844, "learning_rate": 8.693016226476505e-07, "loss": 0.5738, "step": 2605 }, { "epoch": 2.192720387123922, "grad_norm": 94.07272338867188, "learning_rate": 8.676095948965818e-07, "loss": 0.808, "step": 2606 }, { "epoch": 2.193561960866821, "grad_norm": 75.76996612548828, "learning_rate": 8.659188696493564e-07, "loss": 0.8395, "step": 2607 }, { "epoch": 2.1944035346097204, "grad_norm": 75.81172943115234, "learning_rate": 8.642294482550231e-07, "loss": 0.8318, "step": 2608 }, { "epoch": 2.195245108352619, "grad_norm": 101.07645416259766, "learning_rate": 8.625413320615931e-07, "loss": 0.9219, "step": 2609 }, { "epoch": 2.1960866820955185, "grad_norm": 83.057861328125, "learning_rate": 8.608545224160323e-07, "loss": 0.7046, "step": 2610 }, { "epoch": 2.1969282558384178, "grad_norm": 55.087825775146484, "learning_rate": 8.591690206642684e-07, "loss": 0.6379, "step": 2611 }, { "epoch": 2.197769829581317, "grad_norm": 86.99334716796875, "learning_rate": 8.574848281511816e-07, "loss": 0.8289, "step": 2612 }, { "epoch": 2.1986114033242163, "grad_norm": 76.90739440917969, "learning_rate": 8.55801946220611e-07, "loss": 0.7161, "step": 2613 }, { "epoch": 2.1994529770671156, "grad_norm": 63.99169921875, "learning_rate": 8.541203762153466e-07, "loss": 0.5903, "step": 2614 }, { "epoch": 2.200294550810015, "grad_norm": 80.21016693115234, "learning_rate": 8.524401194771351e-07, "loss": 0.7175, "step": 2615 }, { "epoch": 2.2011361245529137, "grad_norm": 69.41879272460938, "learning_rate": 8.507611773466712e-07, "loss": 0.788, "step": 2616 }, { "epoch": 2.201977698295813, "grad_norm": 93.61493682861328, "learning_rate": 8.490835511636053e-07, "loss": 0.879, "step": 2617 }, { "epoch": 2.2028192720387123, "grad_norm": 93.2294692993164, "learning_rate": 8.474072422665341e-07, "loss": 0.9065, "step": 2618 }, { "epoch": 2.2036608457816116, "grad_norm": 78.117919921875, "learning_rate": 8.457322519930039e-07, "loss": 0.8399, "step": 2619 }, { "epoch": 2.204502419524511, "grad_norm": 71.18528747558594, "learning_rate": 8.44058581679511e-07, "loss": 0.6947, "step": 2620 }, { "epoch": 2.20534399326741, "grad_norm": 68.76024627685547, "learning_rate": 8.423862326614954e-07, "loss": 0.7072, "step": 2621 }, { "epoch": 2.2061855670103094, "grad_norm": 65.00563049316406, "learning_rate": 8.407152062733462e-07, "loss": 0.6758, "step": 2622 }, { "epoch": 2.2070271407532083, "grad_norm": 76.2518539428711, "learning_rate": 8.390455038483944e-07, "loss": 0.7294, "step": 2623 }, { "epoch": 2.2078687144961076, "grad_norm": 92.82779693603516, "learning_rate": 8.373771267189155e-07, "loss": 0.8257, "step": 2624 }, { "epoch": 2.208710288239007, "grad_norm": 71.85044860839844, "learning_rate": 8.357100762161266e-07, "loss": 0.724, "step": 2625 }, { "epoch": 2.209551861981906, "grad_norm": 75.96949768066406, "learning_rate": 8.340443536701892e-07, "loss": 0.8478, "step": 2626 }, { "epoch": 2.2103934357248054, "grad_norm": 82.55245971679688, "learning_rate": 8.323799604102018e-07, "loss": 0.7485, "step": 2627 }, { "epoch": 2.2112350094677047, "grad_norm": 86.44733428955078, "learning_rate": 8.307168977642047e-07, "loss": 0.936, "step": 2628 }, { "epoch": 2.212076583210604, "grad_norm": 76.61569213867188, "learning_rate": 8.290551670591743e-07, "loss": 0.7845, "step": 2629 }, { "epoch": 2.2129181569535032, "grad_norm": 69.5757827758789, "learning_rate": 8.273947696210269e-07, "loss": 0.7181, "step": 2630 }, { "epoch": 2.213759730696402, "grad_norm": 88.36131286621094, "learning_rate": 8.257357067746122e-07, "loss": 0.8341, "step": 2631 }, { "epoch": 2.2146013044393014, "grad_norm": 85.6727523803711, "learning_rate": 8.240779798437174e-07, "loss": 0.8354, "step": 2632 }, { "epoch": 2.2154428781822006, "grad_norm": 76.74469757080078, "learning_rate": 8.224215901510613e-07, "loss": 0.7763, "step": 2633 }, { "epoch": 2.2162844519251, "grad_norm": 81.24677276611328, "learning_rate": 8.207665390182987e-07, "loss": 0.8769, "step": 2634 }, { "epoch": 2.217126025667999, "grad_norm": 59.80624771118164, "learning_rate": 8.191128277660129e-07, "loss": 0.6884, "step": 2635 }, { "epoch": 2.2179675994108985, "grad_norm": 93.44509887695312, "learning_rate": 8.174604577137219e-07, "loss": 1.0181, "step": 2636 }, { "epoch": 2.2188091731537978, "grad_norm": 62.65293502807617, "learning_rate": 8.158094301798705e-07, "loss": 0.6754, "step": 2637 }, { "epoch": 2.219650746896697, "grad_norm": 63.386810302734375, "learning_rate": 8.141597464818324e-07, "loss": 0.5182, "step": 2638 }, { "epoch": 2.220492320639596, "grad_norm": 83.7011947631836, "learning_rate": 8.125114079359119e-07, "loss": 0.7442, "step": 2639 }, { "epoch": 2.221333894382495, "grad_norm": 74.85529327392578, "learning_rate": 8.108644158573362e-07, "loss": 0.7306, "step": 2640 }, { "epoch": 2.2221754681253945, "grad_norm": 62.198394775390625, "learning_rate": 8.092187715602615e-07, "loss": 0.6761, "step": 2641 }, { "epoch": 2.2230170418682937, "grad_norm": 94.11766815185547, "learning_rate": 8.075744763577659e-07, "loss": 0.6081, "step": 2642 }, { "epoch": 2.223858615611193, "grad_norm": 64.26312255859375, "learning_rate": 8.059315315618533e-07, "loss": 0.6243, "step": 2643 }, { "epoch": 2.2247001893540923, "grad_norm": 59.57717514038086, "learning_rate": 8.04289938483448e-07, "loss": 0.6755, "step": 2644 }, { "epoch": 2.2255417630969916, "grad_norm": 73.62777709960938, "learning_rate": 8.026496984323975e-07, "loss": 0.6898, "step": 2645 }, { "epoch": 2.2263833368398904, "grad_norm": 68.1532211303711, "learning_rate": 8.010108127174684e-07, "loss": 0.6488, "step": 2646 }, { "epoch": 2.2272249105827897, "grad_norm": 72.72462463378906, "learning_rate": 7.993732826463483e-07, "loss": 0.5844, "step": 2647 }, { "epoch": 2.228066484325689, "grad_norm": 69.8485107421875, "learning_rate": 7.977371095256403e-07, "loss": 0.7049, "step": 2648 }, { "epoch": 2.2289080580685883, "grad_norm": 64.51768493652344, "learning_rate": 7.961022946608685e-07, "loss": 0.7371, "step": 2649 }, { "epoch": 2.2297496318114876, "grad_norm": 67.44911193847656, "learning_rate": 7.9446883935647e-07, "loss": 0.6575, "step": 2650 }, { "epoch": 2.230591205554387, "grad_norm": 98.60176849365234, "learning_rate": 7.928367449157981e-07, "loss": 0.799, "step": 2651 }, { "epoch": 2.231432779297286, "grad_norm": 62.87820816040039, "learning_rate": 7.912060126411217e-07, "loss": 0.7146, "step": 2652 }, { "epoch": 2.232274353040185, "grad_norm": 74.87606811523438, "learning_rate": 7.8957664383362e-07, "loss": 0.7964, "step": 2653 }, { "epoch": 2.2331159267830842, "grad_norm": 72.17474365234375, "learning_rate": 7.879486397933874e-07, "loss": 0.6262, "step": 2654 }, { "epoch": 2.2339575005259835, "grad_norm": 74.49185943603516, "learning_rate": 7.863220018194265e-07, "loss": 0.6359, "step": 2655 }, { "epoch": 2.234799074268883, "grad_norm": 60.609596252441406, "learning_rate": 7.846967312096529e-07, "loss": 0.4705, "step": 2656 }, { "epoch": 2.235640648011782, "grad_norm": 76.58203887939453, "learning_rate": 7.830728292608875e-07, "loss": 0.6328, "step": 2657 }, { "epoch": 2.2364822217546814, "grad_norm": 78.09967041015625, "learning_rate": 7.814502972688625e-07, "loss": 0.8127, "step": 2658 }, { "epoch": 2.2373237954975806, "grad_norm": 71.87503051757812, "learning_rate": 7.798291365282149e-07, "loss": 0.781, "step": 2659 }, { "epoch": 2.2381653692404795, "grad_norm": 69.74879455566406, "learning_rate": 7.78209348332489e-07, "loss": 0.6539, "step": 2660 }, { "epoch": 2.2390069429833788, "grad_norm": 75.3714599609375, "learning_rate": 7.765909339741328e-07, "loss": 0.819, "step": 2661 }, { "epoch": 2.239848516726278, "grad_norm": 78.81387329101562, "learning_rate": 7.749738947444973e-07, "loss": 0.8472, "step": 2662 }, { "epoch": 2.2406900904691773, "grad_norm": 64.38542938232422, "learning_rate": 7.733582319338398e-07, "loss": 0.7762, "step": 2663 }, { "epoch": 2.2415316642120766, "grad_norm": 57.374385833740234, "learning_rate": 7.717439468313146e-07, "loss": 0.5568, "step": 2664 }, { "epoch": 2.242373237954976, "grad_norm": 71.208740234375, "learning_rate": 7.701310407249812e-07, "loss": 0.7077, "step": 2665 }, { "epoch": 2.243214811697875, "grad_norm": 71.51923370361328, "learning_rate": 7.685195149017949e-07, "loss": 0.7042, "step": 2666 }, { "epoch": 2.244056385440774, "grad_norm": 95.73368072509766, "learning_rate": 7.669093706476139e-07, "loss": 0.7494, "step": 2667 }, { "epoch": 2.2448979591836733, "grad_norm": 79.88150024414062, "learning_rate": 7.653006092471882e-07, "loss": 0.7396, "step": 2668 }, { "epoch": 2.2457395329265726, "grad_norm": 72.77428436279297, "learning_rate": 7.636932319841702e-07, "loss": 0.6887, "step": 2669 }, { "epoch": 2.246581106669472, "grad_norm": 69.95852661132812, "learning_rate": 7.620872401411044e-07, "loss": 0.8481, "step": 2670 }, { "epoch": 2.247422680412371, "grad_norm": 76.32493591308594, "learning_rate": 7.604826349994318e-07, "loss": 0.743, "step": 2671 }, { "epoch": 2.2482642541552704, "grad_norm": 77.95675659179688, "learning_rate": 7.588794178394849e-07, "loss": 0.7578, "step": 2672 }, { "epoch": 2.2491058278981697, "grad_norm": 68.78948974609375, "learning_rate": 7.57277589940491e-07, "loss": 0.661, "step": 2673 }, { "epoch": 2.249947401641069, "grad_norm": 55.85639953613281, "learning_rate": 7.556771525805667e-07, "loss": 0.5971, "step": 2674 }, { "epoch": 2.250788975383968, "grad_norm": 63.55598831176758, "learning_rate": 7.540781070367214e-07, "loss": 0.6268, "step": 2675 }, { "epoch": 2.251630549126867, "grad_norm": 59.03841781616211, "learning_rate": 7.524804545848513e-07, "loss": 0.615, "step": 2676 }, { "epoch": 2.2524721228697664, "grad_norm": 76.84010314941406, "learning_rate": 7.508841964997438e-07, "loss": 0.4644, "step": 2677 }, { "epoch": 2.2533136966126657, "grad_norm": 51.794551849365234, "learning_rate": 7.492893340550708e-07, "loss": 0.5472, "step": 2678 }, { "epoch": 2.254155270355565, "grad_norm": 53.754249572753906, "learning_rate": 7.476958685233937e-07, "loss": 0.5048, "step": 2679 }, { "epoch": 2.2549968440984642, "grad_norm": 73.86393737792969, "learning_rate": 7.461038011761571e-07, "loss": 0.7377, "step": 2680 }, { "epoch": 2.2558384178413635, "grad_norm": 71.97015380859375, "learning_rate": 7.445131332836894e-07, "loss": 0.7591, "step": 2681 }, { "epoch": 2.256679991584263, "grad_norm": 78.24359130859375, "learning_rate": 7.429238661152055e-07, "loss": 0.6848, "step": 2682 }, { "epoch": 2.2575215653271616, "grad_norm": 65.17981719970703, "learning_rate": 7.413360009387991e-07, "loss": 0.5589, "step": 2683 }, { "epoch": 2.258363139070061, "grad_norm": 83.3587417602539, "learning_rate": 7.397495390214482e-07, "loss": 0.9073, "step": 2684 }, { "epoch": 2.25920471281296, "grad_norm": 58.31077575683594, "learning_rate": 7.381644816290085e-07, "loss": 0.5773, "step": 2685 }, { "epoch": 2.2600462865558595, "grad_norm": 75.99929809570312, "learning_rate": 7.365808300262176e-07, "loss": 0.8195, "step": 2686 }, { "epoch": 2.2608878602987588, "grad_norm": 88.0613784790039, "learning_rate": 7.349985854766883e-07, "loss": 1.4363, "step": 2687 }, { "epoch": 2.261729434041658, "grad_norm": 63.5888557434082, "learning_rate": 7.334177492429143e-07, "loss": 0.5412, "step": 2688 }, { "epoch": 2.2625710077845573, "grad_norm": 66.94778442382812, "learning_rate": 7.318383225862621e-07, "loss": 0.7216, "step": 2689 }, { "epoch": 2.263412581527456, "grad_norm": 74.33531951904297, "learning_rate": 7.302603067669769e-07, "loss": 0.9001, "step": 2690 }, { "epoch": 2.2642541552703555, "grad_norm": 73.48979187011719, "learning_rate": 7.286837030441749e-07, "loss": 0.7569, "step": 2691 }, { "epoch": 2.2650957290132547, "grad_norm": 75.95056915283203, "learning_rate": 7.271085126758482e-07, "loss": 0.7879, "step": 2692 }, { "epoch": 2.265937302756154, "grad_norm": 63.37393569946289, "learning_rate": 7.255347369188595e-07, "loss": 0.6456, "step": 2693 }, { "epoch": 2.2667788764990533, "grad_norm": 81.6666488647461, "learning_rate": 7.239623770289426e-07, "loss": 0.9573, "step": 2694 }, { "epoch": 2.2676204502419526, "grad_norm": 63.326683044433594, "learning_rate": 7.223914342607036e-07, "loss": 0.8561, "step": 2695 }, { "epoch": 2.268462023984852, "grad_norm": 73.0787582397461, "learning_rate": 7.208219098676153e-07, "loss": 0.6298, "step": 2696 }, { "epoch": 2.2693035977277507, "grad_norm": 72.39900207519531, "learning_rate": 7.192538051020215e-07, "loss": 0.8104, "step": 2697 }, { "epoch": 2.27014517147065, "grad_norm": 76.27084350585938, "learning_rate": 7.176871212151299e-07, "loss": 0.763, "step": 2698 }, { "epoch": 2.2709867452135493, "grad_norm": 66.07930755615234, "learning_rate": 7.161218594570182e-07, "loss": 0.6167, "step": 2699 }, { "epoch": 2.2718283189564485, "grad_norm": 68.93179321289062, "learning_rate": 7.14558021076626e-07, "loss": 0.867, "step": 2700 }, { "epoch": 2.272669892699348, "grad_norm": 90.92766571044922, "learning_rate": 7.129956073217598e-07, "loss": 0.8113, "step": 2701 }, { "epoch": 2.273511466442247, "grad_norm": 68.71321105957031, "learning_rate": 7.114346194390869e-07, "loss": 0.6634, "step": 2702 }, { "epoch": 2.2743530401851464, "grad_norm": 63.138648986816406, "learning_rate": 7.098750586741399e-07, "loss": 0.6428, "step": 2703 }, { "epoch": 2.2751946139280452, "grad_norm": 77.28284454345703, "learning_rate": 7.083169262713094e-07, "loss": 0.6732, "step": 2704 }, { "epoch": 2.2760361876709445, "grad_norm": 82.3548583984375, "learning_rate": 7.067602234738494e-07, "loss": 0.654, "step": 2705 }, { "epoch": 2.276877761413844, "grad_norm": 60.5097770690918, "learning_rate": 7.052049515238709e-07, "loss": 0.5642, "step": 2706 }, { "epoch": 2.277719335156743, "grad_norm": 65.79631805419922, "learning_rate": 7.036511116623432e-07, "loss": 0.5801, "step": 2707 }, { "epoch": 2.2785609088996424, "grad_norm": 68.15387725830078, "learning_rate": 7.020987051290953e-07, "loss": 0.6225, "step": 2708 }, { "epoch": 2.2794024826425416, "grad_norm": 68.56729125976562, "learning_rate": 7.005477331628096e-07, "loss": 0.6634, "step": 2709 }, { "epoch": 2.280244056385441, "grad_norm": 69.78826904296875, "learning_rate": 6.989981970010265e-07, "loss": 0.6271, "step": 2710 }, { "epoch": 2.2810856301283398, "grad_norm": 64.544189453125, "learning_rate": 6.974500978801385e-07, "loss": 0.5273, "step": 2711 }, { "epoch": 2.281927203871239, "grad_norm": 69.36445617675781, "learning_rate": 6.959034370353929e-07, "loss": 0.5533, "step": 2712 }, { "epoch": 2.2827687776141383, "grad_norm": 62.86524200439453, "learning_rate": 6.943582157008879e-07, "loss": 0.6382, "step": 2713 }, { "epoch": 2.2836103513570376, "grad_norm": 69.18538665771484, "learning_rate": 6.928144351095756e-07, "loss": 0.7346, "step": 2714 }, { "epoch": 2.284451925099937, "grad_norm": 88.08509063720703, "learning_rate": 6.912720964932554e-07, "loss": 0.8908, "step": 2715 }, { "epoch": 2.285293498842836, "grad_norm": 76.39439392089844, "learning_rate": 6.897312010825796e-07, "loss": 0.6876, "step": 2716 }, { "epoch": 2.2861350725857355, "grad_norm": 71.68861389160156, "learning_rate": 6.881917501070451e-07, "loss": 0.8289, "step": 2717 }, { "epoch": 2.2869766463286347, "grad_norm": 82.28411102294922, "learning_rate": 6.866537447950003e-07, "loss": 0.6893, "step": 2718 }, { "epoch": 2.287818220071534, "grad_norm": 77.50567626953125, "learning_rate": 6.851171863736361e-07, "loss": 0.6614, "step": 2719 }, { "epoch": 2.288659793814433, "grad_norm": 84.4049072265625, "learning_rate": 6.835820760689923e-07, "loss": 0.9143, "step": 2720 }, { "epoch": 2.289501367557332, "grad_norm": 63.697662353515625, "learning_rate": 6.820484151059509e-07, "loss": 0.7862, "step": 2721 }, { "epoch": 2.2903429413002314, "grad_norm": 69.99846649169922, "learning_rate": 6.805162047082392e-07, "loss": 0.6691, "step": 2722 }, { "epoch": 2.2911845150431307, "grad_norm": 67.50333404541016, "learning_rate": 6.789854460984258e-07, "loss": 0.6738, "step": 2723 }, { "epoch": 2.29202608878603, "grad_norm": 68.89779663085938, "learning_rate": 6.774561404979205e-07, "loss": 0.6846, "step": 2724 }, { "epoch": 2.2928676625289293, "grad_norm": 61.03339385986328, "learning_rate": 6.759282891269758e-07, "loss": 0.6553, "step": 2725 }, { "epoch": 2.2937092362718285, "grad_norm": 85.70636749267578, "learning_rate": 6.744018932046817e-07, "loss": 0.9082, "step": 2726 }, { "epoch": 2.2945508100147274, "grad_norm": 65.80377960205078, "learning_rate": 6.728769539489685e-07, "loss": 0.6463, "step": 2727 }, { "epoch": 2.2953923837576267, "grad_norm": 69.08267974853516, "learning_rate": 6.713534725766027e-07, "loss": 0.612, "step": 2728 }, { "epoch": 2.296233957500526, "grad_norm": 71.7514877319336, "learning_rate": 6.698314503031892e-07, "loss": 0.7709, "step": 2729 }, { "epoch": 2.2970755312434252, "grad_norm": 75.49186706542969, "learning_rate": 6.683108883431668e-07, "loss": 0.7778, "step": 2730 }, { "epoch": 2.2979171049863245, "grad_norm": 98.48780822753906, "learning_rate": 6.667917879098113e-07, "loss": 1.031, "step": 2731 }, { "epoch": 2.298758678729224, "grad_norm": 67.31439971923828, "learning_rate": 6.652741502152298e-07, "loss": 0.7394, "step": 2732 }, { "epoch": 2.299600252472123, "grad_norm": 62.94004440307617, "learning_rate": 6.63757976470365e-07, "loss": 0.6961, "step": 2733 }, { "epoch": 2.300441826215022, "grad_norm": 76.52629089355469, "learning_rate": 6.622432678849883e-07, "loss": 0.8654, "step": 2734 }, { "epoch": 2.301283399957921, "grad_norm": 95.00666809082031, "learning_rate": 6.607300256677057e-07, "loss": 0.6629, "step": 2735 }, { "epoch": 2.3021249737008205, "grad_norm": 69.1645278930664, "learning_rate": 6.592182510259503e-07, "loss": 0.7036, "step": 2736 }, { "epoch": 2.3029665474437198, "grad_norm": 82.000244140625, "learning_rate": 6.577079451659849e-07, "loss": 0.8454, "step": 2737 }, { "epoch": 2.303808121186619, "grad_norm": 69.08513641357422, "learning_rate": 6.561991092929016e-07, "loss": 0.6672, "step": 2738 }, { "epoch": 2.3046496949295183, "grad_norm": 71.79818725585938, "learning_rate": 6.546917446106168e-07, "loss": 0.7542, "step": 2739 }, { "epoch": 2.3054912686724176, "grad_norm": 68.78752899169922, "learning_rate": 6.531858523218773e-07, "loss": 0.5922, "step": 2740 }, { "epoch": 2.3063328424153164, "grad_norm": 83.79581451416016, "learning_rate": 6.516814336282504e-07, "loss": 0.7229, "step": 2741 }, { "epoch": 2.3071744161582157, "grad_norm": 66.08184814453125, "learning_rate": 6.501784897301313e-07, "loss": 0.3769, "step": 2742 }, { "epoch": 2.308015989901115, "grad_norm": 76.68889617919922, "learning_rate": 6.486770218267357e-07, "loss": 0.6826, "step": 2743 }, { "epoch": 2.3088575636440143, "grad_norm": 72.55879211425781, "learning_rate": 6.471770311161041e-07, "loss": 0.7003, "step": 2744 }, { "epoch": 2.3096991373869136, "grad_norm": 72.34886932373047, "learning_rate": 6.456785187950959e-07, "loss": 0.5606, "step": 2745 }, { "epoch": 2.310540711129813, "grad_norm": 83.10501098632812, "learning_rate": 6.441814860593929e-07, "loss": 0.7568, "step": 2746 }, { "epoch": 2.311382284872712, "grad_norm": 59.49289321899414, "learning_rate": 6.426859341034944e-07, "loss": 0.6194, "step": 2747 }, { "epoch": 2.312223858615611, "grad_norm": 68.54898071289062, "learning_rate": 6.411918641207204e-07, "loss": 0.6046, "step": 2748 }, { "epoch": 2.3130654323585103, "grad_norm": 78.61849212646484, "learning_rate": 6.396992773032068e-07, "loss": 0.7696, "step": 2749 }, { "epoch": 2.3139070061014095, "grad_norm": 69.59107208251953, "learning_rate": 6.382081748419058e-07, "loss": 0.5495, "step": 2750 }, { "epoch": 2.314748579844309, "grad_norm": 72.25382232666016, "learning_rate": 6.36718557926587e-07, "loss": 0.7249, "step": 2751 }, { "epoch": 2.315590153587208, "grad_norm": 66.62847900390625, "learning_rate": 6.352304277458321e-07, "loss": 0.7342, "step": 2752 }, { "epoch": 2.3164317273301074, "grad_norm": 67.72795104980469, "learning_rate": 6.337437854870399e-07, "loss": 0.7752, "step": 2753 }, { "epoch": 2.3172733010730067, "grad_norm": 60.34444046020508, "learning_rate": 6.32258632336418e-07, "loss": 0.525, "step": 2754 }, { "epoch": 2.3181148748159055, "grad_norm": 64.30903625488281, "learning_rate": 6.307749694789906e-07, "loss": 0.7063, "step": 2755 }, { "epoch": 2.318956448558805, "grad_norm": 76.66483306884766, "learning_rate": 6.292927980985869e-07, "loss": 0.7567, "step": 2756 }, { "epoch": 2.319798022301704, "grad_norm": 98.90345764160156, "learning_rate": 6.278121193778514e-07, "loss": 0.7554, "step": 2757 }, { "epoch": 2.3206395960446033, "grad_norm": 69.31681823730469, "learning_rate": 6.263329344982339e-07, "loss": 0.6618, "step": 2758 }, { "epoch": 2.3214811697875026, "grad_norm": 70.44660186767578, "learning_rate": 6.24855244639995e-07, "loss": 0.7237, "step": 2759 }, { "epoch": 2.322322743530402, "grad_norm": 80.642578125, "learning_rate": 6.233790509821999e-07, "loss": 0.7881, "step": 2760 }, { "epoch": 2.323164317273301, "grad_norm": 67.77523803710938, "learning_rate": 6.219043547027226e-07, "loss": 0.6395, "step": 2761 }, { "epoch": 2.3240058910162005, "grad_norm": 69.06492614746094, "learning_rate": 6.204311569782395e-07, "loss": 0.6886, "step": 2762 }, { "epoch": 2.3248474647590998, "grad_norm": 72.3539810180664, "learning_rate": 6.189594589842343e-07, "loss": 0.7279, "step": 2763 }, { "epoch": 2.3256890385019986, "grad_norm": 79.99612426757812, "learning_rate": 6.174892618949904e-07, "loss": 0.7797, "step": 2764 }, { "epoch": 2.326530612244898, "grad_norm": 72.51342010498047, "learning_rate": 6.160205668835977e-07, "loss": 0.8849, "step": 2765 }, { "epoch": 2.327372185987797, "grad_norm": 98.62345886230469, "learning_rate": 6.14553375121944e-07, "loss": 0.931, "step": 2766 }, { "epoch": 2.3282137597306964, "grad_norm": 63.61135482788086, "learning_rate": 6.130876877807201e-07, "loss": 0.6797, "step": 2767 }, { "epoch": 2.3290553334735957, "grad_norm": 76.73944854736328, "learning_rate": 6.116235060294154e-07, "loss": 0.6953, "step": 2768 }, { "epoch": 2.329896907216495, "grad_norm": 70.87306213378906, "learning_rate": 6.101608310363172e-07, "loss": 0.7429, "step": 2769 }, { "epoch": 2.3307384809593943, "grad_norm": 78.2562484741211, "learning_rate": 6.086996639685124e-07, "loss": 0.7101, "step": 2770 }, { "epoch": 2.331580054702293, "grad_norm": 77.42034912109375, "learning_rate": 6.072400059918828e-07, "loss": 0.7568, "step": 2771 }, { "epoch": 2.3324216284451924, "grad_norm": 75.82662963867188, "learning_rate": 6.057818582711081e-07, "loss": 0.7037, "step": 2772 }, { "epoch": 2.3332632021880917, "grad_norm": 51.16324996948242, "learning_rate": 6.043252219696602e-07, "loss": 0.5751, "step": 2773 }, { "epoch": 2.334104775930991, "grad_norm": 79.32745361328125, "learning_rate": 6.028700982498084e-07, "loss": 0.7968, "step": 2774 }, { "epoch": 2.3349463496738903, "grad_norm": 80.3613052368164, "learning_rate": 6.01416488272612e-07, "loss": 0.7536, "step": 2775 }, { "epoch": 2.3357879234167895, "grad_norm": 78.13218688964844, "learning_rate": 5.999643931979249e-07, "loss": 0.6145, "step": 2776 }, { "epoch": 2.336629497159689, "grad_norm": 88.33989715576172, "learning_rate": 5.9851381418439e-07, "loss": 0.8331, "step": 2777 }, { "epoch": 2.3374710709025877, "grad_norm": 74.43746185302734, "learning_rate": 5.970647523894427e-07, "loss": 0.8918, "step": 2778 }, { "epoch": 2.338312644645487, "grad_norm": 66.07038879394531, "learning_rate": 5.956172089693058e-07, "loss": 0.6996, "step": 2779 }, { "epoch": 2.339154218388386, "grad_norm": 85.45646667480469, "learning_rate": 5.941711850789924e-07, "loss": 0.8507, "step": 2780 }, { "epoch": 2.3399957921312855, "grad_norm": 80.66790008544922, "learning_rate": 5.927266818723021e-07, "loss": 0.8123, "step": 2781 }, { "epoch": 2.340837365874185, "grad_norm": 57.894527435302734, "learning_rate": 5.912837005018201e-07, "loss": 0.5644, "step": 2782 }, { "epoch": 2.341678939617084, "grad_norm": 72.54692077636719, "learning_rate": 5.898422421189204e-07, "loss": 0.6876, "step": 2783 }, { "epoch": 2.3425205133599833, "grad_norm": 75.5073013305664, "learning_rate": 5.884023078737583e-07, "loss": 0.5964, "step": 2784 }, { "epoch": 2.343362087102882, "grad_norm": 80.3395767211914, "learning_rate": 5.869638989152757e-07, "loss": 0.6719, "step": 2785 }, { "epoch": 2.3442036608457815, "grad_norm": 78.1523666381836, "learning_rate": 5.855270163911953e-07, "loss": 0.8764, "step": 2786 }, { "epoch": 2.3450452345886807, "grad_norm": 94.764404296875, "learning_rate": 5.840916614480241e-07, "loss": 0.7906, "step": 2787 }, { "epoch": 2.34588680833158, "grad_norm": 89.36188507080078, "learning_rate": 5.826578352310477e-07, "loss": 0.8468, "step": 2788 }, { "epoch": 2.3467283820744793, "grad_norm": 70.78152465820312, "learning_rate": 5.812255388843346e-07, "loss": 0.5724, "step": 2789 }, { "epoch": 2.3475699558173786, "grad_norm": 80.20957946777344, "learning_rate": 5.797947735507295e-07, "loss": 0.7106, "step": 2790 }, { "epoch": 2.348411529560278, "grad_norm": 81.46155548095703, "learning_rate": 5.783655403718589e-07, "loss": 0.8191, "step": 2791 }, { "epoch": 2.3492531033031767, "grad_norm": 65.57109069824219, "learning_rate": 5.769378404881235e-07, "loss": 0.733, "step": 2792 }, { "epoch": 2.350094677046076, "grad_norm": 82.96961975097656, "learning_rate": 5.755116750387035e-07, "loss": 0.8243, "step": 2793 }, { "epoch": 2.3509362507889753, "grad_norm": 70.33444213867188, "learning_rate": 5.74087045161553e-07, "loss": 0.66, "step": 2794 }, { "epoch": 2.3517778245318746, "grad_norm": 62.43494415283203, "learning_rate": 5.726639519934002e-07, "loss": 0.5087, "step": 2795 }, { "epoch": 2.352619398274774, "grad_norm": 87.62031555175781, "learning_rate": 5.712423966697495e-07, "loss": 0.6205, "step": 2796 }, { "epoch": 2.353460972017673, "grad_norm": 95.37403106689453, "learning_rate": 5.698223803248759e-07, "loss": 0.8199, "step": 2797 }, { "epoch": 2.3543025457605724, "grad_norm": 86.87307739257812, "learning_rate": 5.684039040918285e-07, "loss": 0.7423, "step": 2798 }, { "epoch": 2.3551441195034712, "grad_norm": 75.36544799804688, "learning_rate": 5.669869691024258e-07, "loss": 0.7015, "step": 2799 }, { "epoch": 2.3559856932463705, "grad_norm": 89.22855377197266, "learning_rate": 5.655715764872575e-07, "loss": 0.7582, "step": 2800 }, { "epoch": 2.35682726698927, "grad_norm": 83.80813598632812, "learning_rate": 5.641577273756812e-07, "loss": 0.6119, "step": 2801 }, { "epoch": 2.357668840732169, "grad_norm": 57.1900520324707, "learning_rate": 5.627454228958257e-07, "loss": 0.4919, "step": 2802 }, { "epoch": 2.3585104144750684, "grad_norm": 81.27507781982422, "learning_rate": 5.61334664174584e-07, "loss": 0.7169, "step": 2803 }, { "epoch": 2.3593519882179677, "grad_norm": 104.19719696044922, "learning_rate": 5.599254523376191e-07, "loss": 0.926, "step": 2804 }, { "epoch": 2.360193561960867, "grad_norm": 74.32052612304688, "learning_rate": 5.585177885093562e-07, "loss": 0.7827, "step": 2805 }, { "epoch": 2.361035135703766, "grad_norm": 87.31022644042969, "learning_rate": 5.571116738129889e-07, "loss": 0.9509, "step": 2806 }, { "epoch": 2.3618767094466655, "grad_norm": 74.00389862060547, "learning_rate": 5.557071093704713e-07, "loss": 0.8326, "step": 2807 }, { "epoch": 2.3627182831895643, "grad_norm": 85.74870300292969, "learning_rate": 5.543040963025237e-07, "loss": 0.9886, "step": 2808 }, { "epoch": 2.3635598569324636, "grad_norm": 84.3199462890625, "learning_rate": 5.529026357286257e-07, "loss": 0.7827, "step": 2809 }, { "epoch": 2.364401430675363, "grad_norm": 60.698997497558594, "learning_rate": 5.515027287670205e-07, "loss": 0.4828, "step": 2810 }, { "epoch": 2.365243004418262, "grad_norm": 71.0528793334961, "learning_rate": 5.501043765347102e-07, "loss": 0.7467, "step": 2811 }, { "epoch": 2.3660845781611615, "grad_norm": 65.58141326904297, "learning_rate": 5.487075801474562e-07, "loss": 0.6683, "step": 2812 }, { "epoch": 2.3669261519040607, "grad_norm": 79.87923431396484, "learning_rate": 5.473123407197802e-07, "loss": 0.7423, "step": 2813 }, { "epoch": 2.36776772564696, "grad_norm": 87.9638442993164, "learning_rate": 5.459186593649594e-07, "loss": 0.8497, "step": 2814 }, { "epoch": 2.368609299389859, "grad_norm": 65.41155242919922, "learning_rate": 5.445265371950298e-07, "loss": 0.6166, "step": 2815 }, { "epoch": 2.369450873132758, "grad_norm": 83.76545715332031, "learning_rate": 5.43135975320781e-07, "loss": 0.7609, "step": 2816 }, { "epoch": 2.3702924468756574, "grad_norm": 68.06028747558594, "learning_rate": 5.417469748517606e-07, "loss": 0.7898, "step": 2817 }, { "epoch": 2.3711340206185567, "grad_norm": 71.13483428955078, "learning_rate": 5.403595368962672e-07, "loss": 0.7531, "step": 2818 }, { "epoch": 2.371975594361456, "grad_norm": 94.18714904785156, "learning_rate": 5.389736625613553e-07, "loss": 0.8381, "step": 2819 }, { "epoch": 2.3728171681043553, "grad_norm": 57.60980224609375, "learning_rate": 5.375893529528298e-07, "loss": 0.6088, "step": 2820 }, { "epoch": 2.3736587418472546, "grad_norm": 66.8813705444336, "learning_rate": 5.362066091752488e-07, "loss": 0.6743, "step": 2821 }, { "epoch": 2.3745003155901534, "grad_norm": 72.43280792236328, "learning_rate": 5.348254323319188e-07, "loss": 0.7803, "step": 2822 }, { "epoch": 2.3753418893330527, "grad_norm": 68.30451202392578, "learning_rate": 5.334458235248991e-07, "loss": 0.7013, "step": 2823 }, { "epoch": 2.376183463075952, "grad_norm": 78.18911743164062, "learning_rate": 5.320677838549948e-07, "loss": 0.7499, "step": 2824 }, { "epoch": 2.3770250368188512, "grad_norm": 97.57591247558594, "learning_rate": 5.306913144217598e-07, "loss": 0.8554, "step": 2825 }, { "epoch": 2.3778666105617505, "grad_norm": 63.472843170166016, "learning_rate": 5.29316416323497e-07, "loss": 0.63, "step": 2826 }, { "epoch": 2.37870818430465, "grad_norm": 78.7544174194336, "learning_rate": 5.279430906572528e-07, "loss": 0.6776, "step": 2827 }, { "epoch": 2.379549758047549, "grad_norm": 78.45001220703125, "learning_rate": 5.265713385188215e-07, "loss": 0.7812, "step": 2828 }, { "epoch": 2.380391331790448, "grad_norm": 58.25782012939453, "learning_rate": 5.252011610027388e-07, "loss": 0.5565, "step": 2829 }, { "epoch": 2.381232905533347, "grad_norm": 76.55278778076172, "learning_rate": 5.238325592022875e-07, "loss": 0.7931, "step": 2830 }, { "epoch": 2.3820744792762465, "grad_norm": 71.27601623535156, "learning_rate": 5.224655342094898e-07, "loss": 0.7598, "step": 2831 }, { "epoch": 2.3829160530191458, "grad_norm": 93.5136947631836, "learning_rate": 5.211000871151126e-07, "loss": 0.8727, "step": 2832 }, { "epoch": 2.383757626762045, "grad_norm": 67.13955688476562, "learning_rate": 5.197362190086613e-07, "loss": 0.5739, "step": 2833 }, { "epoch": 2.3845992005049443, "grad_norm": 92.276123046875, "learning_rate": 5.183739309783842e-07, "loss": 1.0032, "step": 2834 }, { "epoch": 2.3854407742478436, "grad_norm": 73.4191665649414, "learning_rate": 5.170132241112655e-07, "loss": 0.7186, "step": 2835 }, { "epoch": 2.3862823479907425, "grad_norm": 70.24175262451172, "learning_rate": 5.156540994930306e-07, "loss": 0.764, "step": 2836 }, { "epoch": 2.3871239217336417, "grad_norm": 76.55288696289062, "learning_rate": 5.142965582081413e-07, "loss": 0.7587, "step": 2837 }, { "epoch": 2.387965495476541, "grad_norm": 73.71908569335938, "learning_rate": 5.129406013397948e-07, "loss": 0.7861, "step": 2838 }, { "epoch": 2.3888070692194403, "grad_norm": 72.42681884765625, "learning_rate": 5.11586229969927e-07, "loss": 0.7079, "step": 2839 }, { "epoch": 2.3896486429623396, "grad_norm": 88.2432861328125, "learning_rate": 5.102334451792054e-07, "loss": 0.6677, "step": 2840 }, { "epoch": 2.390490216705239, "grad_norm": 86.31133270263672, "learning_rate": 5.088822480470349e-07, "loss": 0.6334, "step": 2841 }, { "epoch": 2.391331790448138, "grad_norm": 64.02345275878906, "learning_rate": 5.075326396515501e-07, "loss": 0.5847, "step": 2842 }, { "epoch": 2.3921733641910374, "grad_norm": 79.86390686035156, "learning_rate": 5.06184621069622e-07, "loss": 0.7749, "step": 2843 }, { "epoch": 2.3930149379339363, "grad_norm": 58.0899772644043, "learning_rate": 5.048381933768479e-07, "loss": 0.7013, "step": 2844 }, { "epoch": 2.3938565116768356, "grad_norm": 69.51956176757812, "learning_rate": 5.034933576475609e-07, "loss": 0.6276, "step": 2845 }, { "epoch": 2.394698085419735, "grad_norm": 70.69096374511719, "learning_rate": 5.021501149548197e-07, "loss": 0.7173, "step": 2846 }, { "epoch": 2.395539659162634, "grad_norm": 65.6499252319336, "learning_rate": 5.008084663704155e-07, "loss": 0.6346, "step": 2847 }, { "epoch": 2.3963812329055334, "grad_norm": 63.65243911743164, "learning_rate": 4.994684129648641e-07, "loss": 0.5175, "step": 2848 }, { "epoch": 2.3972228066484327, "grad_norm": 84.8652572631836, "learning_rate": 4.981299558074116e-07, "loss": 0.731, "step": 2849 }, { "epoch": 2.398064380391332, "grad_norm": 75.05459594726562, "learning_rate": 4.967930959660277e-07, "loss": 0.7957, "step": 2850 }, { "epoch": 2.3989059541342312, "grad_norm": 69.55000305175781, "learning_rate": 4.954578345074101e-07, "loss": 0.6918, "step": 2851 }, { "epoch": 2.39974752787713, "grad_norm": 87.55575561523438, "learning_rate": 4.941241724969789e-07, "loss": 0.8302, "step": 2852 }, { "epoch": 2.4005891016200294, "grad_norm": 83.13668823242188, "learning_rate": 4.927921109988798e-07, "loss": 0.8951, "step": 2853 }, { "epoch": 2.4014306753629286, "grad_norm": 63.46538162231445, "learning_rate": 4.914616510759798e-07, "loss": 0.6953, "step": 2854 }, { "epoch": 2.402272249105828, "grad_norm": 93.41539001464844, "learning_rate": 4.901327937898698e-07, "loss": 0.8709, "step": 2855 }, { "epoch": 2.403113822848727, "grad_norm": 72.72631072998047, "learning_rate": 4.888055402008602e-07, "loss": 0.6018, "step": 2856 }, { "epoch": 2.4039553965916265, "grad_norm": 77.36554718017578, "learning_rate": 4.87479891367982e-07, "loss": 0.8762, "step": 2857 }, { "epoch": 2.4047969703345258, "grad_norm": 81.09053039550781, "learning_rate": 4.861558483489878e-07, "loss": 0.77, "step": 2858 }, { "epoch": 2.4056385440774246, "grad_norm": 67.20418548583984, "learning_rate": 4.848334122003456e-07, "loss": 0.7321, "step": 2859 }, { "epoch": 2.406480117820324, "grad_norm": 66.99989318847656, "learning_rate": 4.835125839772448e-07, "loss": 0.6287, "step": 2860 }, { "epoch": 2.407321691563223, "grad_norm": 95.92698669433594, "learning_rate": 4.821933647335883e-07, "loss": 0.7052, "step": 2861 }, { "epoch": 2.4081632653061225, "grad_norm": 60.03333282470703, "learning_rate": 4.808757555219986e-07, "loss": 0.685, "step": 2862 }, { "epoch": 2.4090048390490217, "grad_norm": 83.08409118652344, "learning_rate": 4.795597573938105e-07, "loss": 0.826, "step": 2863 }, { "epoch": 2.409846412791921, "grad_norm": 90.52806091308594, "learning_rate": 4.78245371399076e-07, "loss": 0.9159, "step": 2864 }, { "epoch": 2.4106879865348203, "grad_norm": 70.5450668334961, "learning_rate": 4.769325985865577e-07, "loss": 0.7109, "step": 2865 }, { "epoch": 2.411529560277719, "grad_norm": 70.94446563720703, "learning_rate": 4.756214400037343e-07, "loss": 0.7586, "step": 2866 }, { "epoch": 2.4123711340206184, "grad_norm": 64.8140869140625, "learning_rate": 4.743118966967938e-07, "loss": 0.6505, "step": 2867 }, { "epoch": 2.4132127077635177, "grad_norm": 55.723838806152344, "learning_rate": 4.7300396971063784e-07, "loss": 0.5858, "step": 2868 }, { "epoch": 2.414054281506417, "grad_norm": 62.51237869262695, "learning_rate": 4.716976600888759e-07, "loss": 0.6471, "step": 2869 }, { "epoch": 2.4148958552493163, "grad_norm": 57.743011474609375, "learning_rate": 4.703929688738279e-07, "loss": 0.5322, "step": 2870 }, { "epoch": 2.4157374289922156, "grad_norm": 73.57013702392578, "learning_rate": 4.690898971065236e-07, "loss": 0.78, "step": 2871 }, { "epoch": 2.416579002735115, "grad_norm": 67.57293701171875, "learning_rate": 4.677884458266982e-07, "loss": 0.569, "step": 2872 }, { "epoch": 2.4174205764780137, "grad_norm": 89.22135925292969, "learning_rate": 4.664886160727966e-07, "loss": 0.8086, "step": 2873 }, { "epoch": 2.418262150220913, "grad_norm": 76.50499725341797, "learning_rate": 4.6519040888196755e-07, "loss": 0.7092, "step": 2874 }, { "epoch": 2.4191037239638122, "grad_norm": 70.07254791259766, "learning_rate": 4.638938252900671e-07, "loss": 0.5811, "step": 2875 }, { "epoch": 2.4199452977067115, "grad_norm": 69.00104522705078, "learning_rate": 4.625988663316533e-07, "loss": 0.6891, "step": 2876 }, { "epoch": 2.420786871449611, "grad_norm": 79.72003936767578, "learning_rate": 4.6130553303999127e-07, "loss": 0.5434, "step": 2877 }, { "epoch": 2.42162844519251, "grad_norm": 64.02023315429688, "learning_rate": 4.600138264470455e-07, "loss": 0.583, "step": 2878 }, { "epoch": 2.4224700189354094, "grad_norm": 85.77101135253906, "learning_rate": 4.587237475834852e-07, "loss": 0.8188, "step": 2879 }, { "epoch": 2.423311592678308, "grad_norm": 75.75797271728516, "learning_rate": 4.574352974786797e-07, "loss": 0.6148, "step": 2880 }, { "epoch": 2.4241531664212075, "grad_norm": 70.56800079345703, "learning_rate": 4.561484771606975e-07, "loss": 0.6727, "step": 2881 }, { "epoch": 2.4249947401641068, "grad_norm": 74.41127014160156, "learning_rate": 4.548632876563097e-07, "loss": 0.7229, "step": 2882 }, { "epoch": 2.425836313907006, "grad_norm": 60.94454574584961, "learning_rate": 4.5357972999098274e-07, "loss": 0.7146, "step": 2883 }, { "epoch": 2.4266778876499053, "grad_norm": 69.30078887939453, "learning_rate": 4.5229780518888403e-07, "loss": 0.7364, "step": 2884 }, { "epoch": 2.4275194613928046, "grad_norm": 66.25672912597656, "learning_rate": 4.5101751427287564e-07, "loss": 0.5268, "step": 2885 }, { "epoch": 2.428361035135704, "grad_norm": 84.91715240478516, "learning_rate": 4.4973885826451816e-07, "loss": 0.5426, "step": 2886 }, { "epoch": 2.429202608878603, "grad_norm": 71.32837677001953, "learning_rate": 4.484618381840658e-07, "loss": 0.5602, "step": 2887 }, { "epoch": 2.4300441826215025, "grad_norm": 85.11842346191406, "learning_rate": 4.4718645505046816e-07, "loss": 0.6415, "step": 2888 }, { "epoch": 2.4308857563644013, "grad_norm": 71.9579849243164, "learning_rate": 4.4591270988136815e-07, "loss": 0.7654, "step": 2889 }, { "epoch": 2.4317273301073006, "grad_norm": 65.82129669189453, "learning_rate": 4.446406036931036e-07, "loss": 0.7394, "step": 2890 }, { "epoch": 2.4325689038502, "grad_norm": 55.85297393798828, "learning_rate": 4.433701375007018e-07, "loss": 0.4927, "step": 2891 }, { "epoch": 2.433410477593099, "grad_norm": 64.38798522949219, "learning_rate": 4.4210131231788465e-07, "loss": 0.5416, "step": 2892 }, { "epoch": 2.4342520513359984, "grad_norm": 68.10429382324219, "learning_rate": 4.4083412915706114e-07, "loss": 0.825, "step": 2893 }, { "epoch": 2.4350936250788977, "grad_norm": 67.20697021484375, "learning_rate": 4.395685890293333e-07, "loss": 0.6683, "step": 2894 }, { "epoch": 2.435935198821797, "grad_norm": 68.6749038696289, "learning_rate": 4.383046929444898e-07, "loss": 0.6102, "step": 2895 }, { "epoch": 2.436776772564696, "grad_norm": 74.43025207519531, "learning_rate": 4.3704244191100947e-07, "loss": 0.7007, "step": 2896 }, { "epoch": 2.437618346307595, "grad_norm": 65.49515533447266, "learning_rate": 4.3578183693605615e-07, "loss": 0.5542, "step": 2897 }, { "epoch": 2.4384599200504944, "grad_norm": 77.37419128417969, "learning_rate": 4.3452287902548295e-07, "loss": 0.7009, "step": 2898 }, { "epoch": 2.4393014937933937, "grad_norm": 53.074153900146484, "learning_rate": 4.3326556918382694e-07, "loss": 0.5877, "step": 2899 }, { "epoch": 2.440143067536293, "grad_norm": 83.15350341796875, "learning_rate": 4.3200990841431007e-07, "loss": 0.7741, "step": 2900 }, { "epoch": 2.4409846412791922, "grad_norm": 89.71104431152344, "learning_rate": 4.307558977188403e-07, "loss": 0.7063, "step": 2901 }, { "epoch": 2.4418262150220915, "grad_norm": 76.31108856201172, "learning_rate": 4.295035380980064e-07, "loss": 0.7377, "step": 2902 }, { "epoch": 2.4426677887649904, "grad_norm": 101.99163055419922, "learning_rate": 4.2825283055108274e-07, "loss": 0.8148, "step": 2903 }, { "epoch": 2.4435093625078896, "grad_norm": 70.75888061523438, "learning_rate": 4.2700377607602204e-07, "loss": 0.7201, "step": 2904 }, { "epoch": 2.444350936250789, "grad_norm": 91.8196029663086, "learning_rate": 4.257563756694616e-07, "loss": 0.8204, "step": 2905 }, { "epoch": 2.445192509993688, "grad_norm": 85.7092056274414, "learning_rate": 4.245106303267155e-07, "loss": 0.8377, "step": 2906 }, { "epoch": 2.4460340837365875, "grad_norm": 74.576416015625, "learning_rate": 4.232665410417802e-07, "loss": 0.5651, "step": 2907 }, { "epoch": 2.4468756574794868, "grad_norm": 77.78511810302734, "learning_rate": 4.220241088073282e-07, "loss": 0.9072, "step": 2908 }, { "epoch": 2.447717231222386, "grad_norm": 66.56268310546875, "learning_rate": 4.2078333461471247e-07, "loss": 0.5572, "step": 2909 }, { "epoch": 2.448558804965285, "grad_norm": 65.0099868774414, "learning_rate": 4.1954421945396007e-07, "loss": 0.6724, "step": 2910 }, { "epoch": 2.449400378708184, "grad_norm": 74.9384994506836, "learning_rate": 4.1830676431377727e-07, "loss": 0.8451, "step": 2911 }, { "epoch": 2.4502419524510834, "grad_norm": 68.10995483398438, "learning_rate": 4.170709701815437e-07, "loss": 0.6404, "step": 2912 }, { "epoch": 2.4510835261939827, "grad_norm": 81.80453491210938, "learning_rate": 4.1583683804331336e-07, "loss": 0.8797, "step": 2913 }, { "epoch": 2.451925099936882, "grad_norm": 95.01277160644531, "learning_rate": 4.1460436888381666e-07, "loss": 0.7832, "step": 2914 }, { "epoch": 2.4527666736797813, "grad_norm": 82.091796875, "learning_rate": 4.1337356368645414e-07, "loss": 0.843, "step": 2915 }, { "epoch": 2.4536082474226806, "grad_norm": 74.73335266113281, "learning_rate": 4.1214442343330133e-07, "loss": 0.6345, "step": 2916 }, { "epoch": 2.4544498211655794, "grad_norm": 66.23681640625, "learning_rate": 4.1091694910510236e-07, "loss": 0.6309, "step": 2917 }, { "epoch": 2.4552913949084787, "grad_norm": 67.25305938720703, "learning_rate": 4.0969114168127545e-07, "loss": 0.647, "step": 2918 }, { "epoch": 2.456132968651378, "grad_norm": 85.36168670654297, "learning_rate": 4.0846700213990537e-07, "loss": 0.7821, "step": 2919 }, { "epoch": 2.4569745423942773, "grad_norm": 77.59706115722656, "learning_rate": 4.0724453145774916e-07, "loss": 0.7125, "step": 2920 }, { "epoch": 2.4578161161371765, "grad_norm": 73.59081268310547, "learning_rate": 4.0602373061022965e-07, "loss": 0.6388, "step": 2921 }, { "epoch": 2.458657689880076, "grad_norm": 59.99535369873047, "learning_rate": 4.0480460057143927e-07, "loss": 0.657, "step": 2922 }, { "epoch": 2.459499263622975, "grad_norm": 84.95344543457031, "learning_rate": 4.035871423141355e-07, "loss": 0.7777, "step": 2923 }, { "epoch": 2.460340837365874, "grad_norm": 65.10357666015625, "learning_rate": 4.0237135680974425e-07, "loss": 0.7684, "step": 2924 }, { "epoch": 2.4611824111087732, "grad_norm": 85.78872680664062, "learning_rate": 4.0115724502835436e-07, "loss": 0.7728, "step": 2925 }, { "epoch": 2.4620239848516725, "grad_norm": 55.186553955078125, "learning_rate": 3.9994480793871996e-07, "loss": 0.5056, "step": 2926 }, { "epoch": 2.462865558594572, "grad_norm": 64.6177749633789, "learning_rate": 3.9873404650825983e-07, "loss": 0.7035, "step": 2927 }, { "epoch": 2.463707132337471, "grad_norm": 81.97435760498047, "learning_rate": 3.9752496170305436e-07, "loss": 0.7332, "step": 2928 }, { "epoch": 2.4645487060803704, "grad_norm": 72.62235260009766, "learning_rate": 3.9631755448784753e-07, "loss": 0.8292, "step": 2929 }, { "epoch": 2.4653902798232696, "grad_norm": 81.98554992675781, "learning_rate": 3.9511182582604305e-07, "loss": 0.8335, "step": 2930 }, { "epoch": 2.466231853566169, "grad_norm": 77.68927764892578, "learning_rate": 3.939077766797078e-07, "loss": 0.6359, "step": 2931 }, { "epoch": 2.467073427309068, "grad_norm": 65.85250854492188, "learning_rate": 3.927054080095649e-07, "loss": 0.5673, "step": 2932 }, { "epoch": 2.467915001051967, "grad_norm": 83.57113647460938, "learning_rate": 3.915047207750006e-07, "loss": 0.5669, "step": 2933 }, { "epoch": 2.4687565747948663, "grad_norm": 87.88668823242188, "learning_rate": 3.903057159340559e-07, "loss": 0.8293, "step": 2934 }, { "epoch": 2.4695981485377656, "grad_norm": 84.85686492919922, "learning_rate": 3.891083944434329e-07, "loss": 0.8363, "step": 2935 }, { "epoch": 2.470439722280665, "grad_norm": 77.37671661376953, "learning_rate": 3.87912757258487e-07, "loss": 0.7659, "step": 2936 }, { "epoch": 2.471281296023564, "grad_norm": 68.12232208251953, "learning_rate": 3.86718805333233e-07, "loss": 0.6278, "step": 2937 }, { "epoch": 2.4721228697664634, "grad_norm": 71.66863250732422, "learning_rate": 3.855265396203381e-07, "loss": 0.4967, "step": 2938 }, { "epoch": 2.4729644435093627, "grad_norm": 75.73211669921875, "learning_rate": 3.843359610711267e-07, "loss": 0.6937, "step": 2939 }, { "epoch": 2.4738060172522616, "grad_norm": 80.80725860595703, "learning_rate": 3.8314707063557445e-07, "loss": 0.7193, "step": 2940 }, { "epoch": 2.474647590995161, "grad_norm": 123.07603454589844, "learning_rate": 3.819598692623122e-07, "loss": 0.8792, "step": 2941 }, { "epoch": 2.47548916473806, "grad_norm": 64.66447448730469, "learning_rate": 3.807743578986217e-07, "loss": 0.6556, "step": 2942 }, { "epoch": 2.4763307384809594, "grad_norm": 80.24279022216797, "learning_rate": 3.795905374904363e-07, "loss": 0.5938, "step": 2943 }, { "epoch": 2.4771723122238587, "grad_norm": 84.3942642211914, "learning_rate": 3.784084089823412e-07, "loss": 0.7238, "step": 2944 }, { "epoch": 2.478013885966758, "grad_norm": 78.06907653808594, "learning_rate": 3.772279733175702e-07, "loss": 0.7026, "step": 2945 }, { "epoch": 2.4788554597096573, "grad_norm": 73.81548309326172, "learning_rate": 3.7604923143800767e-07, "loss": 0.6441, "step": 2946 }, { "epoch": 2.479697033452556, "grad_norm": 65.31498718261719, "learning_rate": 3.7487218428418517e-07, "loss": 0.5947, "step": 2947 }, { "epoch": 2.4805386071954554, "grad_norm": 74.80921936035156, "learning_rate": 3.7369683279528364e-07, "loss": 0.6549, "step": 2948 }, { "epoch": 2.4813801809383547, "grad_norm": 66.89586639404297, "learning_rate": 3.72523177909129e-07, "loss": 0.595, "step": 2949 }, { "epoch": 2.482221754681254, "grad_norm": 80.47528839111328, "learning_rate": 3.713512205621958e-07, "loss": 0.7593, "step": 2950 }, { "epoch": 2.4830633284241532, "grad_norm": 83.587890625, "learning_rate": 3.7018096168960154e-07, "loss": 0.8495, "step": 2951 }, { "epoch": 2.4839049021670525, "grad_norm": 66.08248138427734, "learning_rate": 3.69012402225111e-07, "loss": 0.5155, "step": 2952 }, { "epoch": 2.484746475909952, "grad_norm": 89.72796630859375, "learning_rate": 3.6784554310113047e-07, "loss": 0.9008, "step": 2953 }, { "epoch": 2.4855880496528506, "grad_norm": 71.32404327392578, "learning_rate": 3.666803852487122e-07, "loss": 0.5415, "step": 2954 }, { "epoch": 2.48642962339575, "grad_norm": 65.65059661865234, "learning_rate": 3.6551692959754885e-07, "loss": 0.58, "step": 2955 }, { "epoch": 2.487271197138649, "grad_norm": 105.35753631591797, "learning_rate": 3.6435517707597514e-07, "loss": 1.0298, "step": 2956 }, { "epoch": 2.4881127708815485, "grad_norm": 74.5104751586914, "learning_rate": 3.6319512861096807e-07, "loss": 0.8663, "step": 2957 }, { "epoch": 2.4889543446244478, "grad_norm": 66.81471252441406, "learning_rate": 3.6203678512814366e-07, "loss": 0.6707, "step": 2958 }, { "epoch": 2.489795918367347, "grad_norm": 73.4651107788086, "learning_rate": 3.6088014755175873e-07, "loss": 0.5985, "step": 2959 }, { "epoch": 2.4906374921102463, "grad_norm": 91.64862823486328, "learning_rate": 3.5972521680470747e-07, "loss": 0.7358, "step": 2960 }, { "epoch": 2.491479065853145, "grad_norm": 60.11470031738281, "learning_rate": 3.585719938085236e-07, "loss": 0.6411, "step": 2961 }, { "epoch": 2.4923206395960444, "grad_norm": 67.0931167602539, "learning_rate": 3.574204794833769e-07, "loss": 0.7048, "step": 2962 }, { "epoch": 2.4931622133389437, "grad_norm": 93.97506713867188, "learning_rate": 3.562706747480754e-07, "loss": 0.7478, "step": 2963 }, { "epoch": 2.494003787081843, "grad_norm": 66.70235443115234, "learning_rate": 3.5512258052006087e-07, "loss": 0.7373, "step": 2964 }, { "epoch": 2.4948453608247423, "grad_norm": 62.17636489868164, "learning_rate": 3.539761977154124e-07, "loss": 0.5253, "step": 2965 }, { "epoch": 2.4956869345676416, "grad_norm": 64.12118530273438, "learning_rate": 3.5283152724884165e-07, "loss": 0.6718, "step": 2966 }, { "epoch": 2.496528508310541, "grad_norm": 75.38851165771484, "learning_rate": 3.5168857003369613e-07, "loss": 0.8652, "step": 2967 }, { "epoch": 2.4973700820534397, "grad_norm": 64.30504608154297, "learning_rate": 3.5054732698195424e-07, "loss": 0.7016, "step": 2968 }, { "epoch": 2.498211655796339, "grad_norm": 66.50183868408203, "learning_rate": 3.494077990042272e-07, "loss": 0.7028, "step": 2969 }, { "epoch": 2.4990532295392383, "grad_norm": 52.156211853027344, "learning_rate": 3.482699870097589e-07, "loss": 0.4556, "step": 2970 }, { "epoch": 2.4998948032821375, "grad_norm": 85.39350891113281, "learning_rate": 3.471338919064221e-07, "loss": 0.7236, "step": 2971 }, { "epoch": 2.500736377025037, "grad_norm": 58.85251998901367, "learning_rate": 3.459995146007217e-07, "loss": 0.5998, "step": 2972 }, { "epoch": 2.501577950767936, "grad_norm": 74.82879638671875, "learning_rate": 3.4486685599778995e-07, "loss": 0.5815, "step": 2973 }, { "epoch": 2.5024195245108354, "grad_norm": 121.61479187011719, "learning_rate": 3.4373591700139013e-07, "loss": 0.9555, "step": 2974 }, { "epoch": 2.503261098253734, "grad_norm": 57.69901657104492, "learning_rate": 3.4260669851391067e-07, "loss": 0.6584, "step": 2975 }, { "epoch": 2.504102671996634, "grad_norm": 71.45643615722656, "learning_rate": 3.414792014363691e-07, "loss": 0.597, "step": 2976 }, { "epoch": 2.504944245739533, "grad_norm": 74.47058868408203, "learning_rate": 3.403534266684083e-07, "loss": 0.634, "step": 2977 }, { "epoch": 2.505785819482432, "grad_norm": 75.10823059082031, "learning_rate": 3.392293751082981e-07, "loss": 0.8324, "step": 2978 }, { "epoch": 2.5066273932253313, "grad_norm": 67.17010498046875, "learning_rate": 3.381070476529319e-07, "loss": 0.6373, "step": 2979 }, { "epoch": 2.5074689669682306, "grad_norm": 71.45992279052734, "learning_rate": 3.3698644519782944e-07, "loss": 0.7788, "step": 2980 }, { "epoch": 2.50831054071113, "grad_norm": 74.49483489990234, "learning_rate": 3.3586756863713153e-07, "loss": 0.6999, "step": 2981 }, { "epoch": 2.509152114454029, "grad_norm": 69.73905944824219, "learning_rate": 3.347504188636042e-07, "loss": 0.7447, "step": 2982 }, { "epoch": 2.5099936881969285, "grad_norm": 78.26557159423828, "learning_rate": 3.3363499676863366e-07, "loss": 0.7851, "step": 2983 }, { "epoch": 2.5108352619398273, "grad_norm": 71.55675506591797, "learning_rate": 3.3252130324222975e-07, "loss": 0.6733, "step": 2984 }, { "epoch": 2.5116768356827266, "grad_norm": 74.66873168945312, "learning_rate": 3.314093391730205e-07, "loss": 0.4865, "step": 2985 }, { "epoch": 2.512518409425626, "grad_norm": 56.66135787963867, "learning_rate": 3.3029910544825653e-07, "loss": 0.5577, "step": 2986 }, { "epoch": 2.513359983168525, "grad_norm": 68.96792602539062, "learning_rate": 3.291906029538064e-07, "loss": 0.6793, "step": 2987 }, { "epoch": 2.5142015569114244, "grad_norm": 87.43818664550781, "learning_rate": 3.2808383257415645e-07, "loss": 0.6944, "step": 2988 }, { "epoch": 2.5150431306543237, "grad_norm": 98.43070220947266, "learning_rate": 3.2697879519241333e-07, "loss": 1.0489, "step": 2989 }, { "epoch": 2.515884704397223, "grad_norm": 92.46650695800781, "learning_rate": 3.258754916902987e-07, "loss": 0.6346, "step": 2990 }, { "epoch": 2.516726278140122, "grad_norm": 70.23343658447266, "learning_rate": 3.247739229481522e-07, "loss": 0.7804, "step": 2991 }, { "epoch": 2.517567851883021, "grad_norm": 73.9881591796875, "learning_rate": 3.236740898449284e-07, "loss": 0.6821, "step": 2992 }, { "epoch": 2.5184094256259204, "grad_norm": 72.21820831298828, "learning_rate": 3.225759932581976e-07, "loss": 0.674, "step": 2993 }, { "epoch": 2.5192509993688197, "grad_norm": 86.99113464355469, "learning_rate": 3.214796340641438e-07, "loss": 0.8204, "step": 2994 }, { "epoch": 2.520092573111719, "grad_norm": 68.23062133789062, "learning_rate": 3.2038501313756576e-07, "loss": 0.5949, "step": 2995 }, { "epoch": 2.5209341468546183, "grad_norm": 62.9173698425293, "learning_rate": 3.1929213135187347e-07, "loss": 0.621, "step": 2996 }, { "epoch": 2.5217757205975175, "grad_norm": 70.73606872558594, "learning_rate": 3.182009895790919e-07, "loss": 0.67, "step": 2997 }, { "epoch": 2.5226172943404164, "grad_norm": 76.61640930175781, "learning_rate": 3.1711158868985487e-07, "loss": 0.6759, "step": 2998 }, { "epoch": 2.5234588680833157, "grad_norm": 82.52962493896484, "learning_rate": 3.1602392955340925e-07, "loss": 0.6979, "step": 2999 }, { "epoch": 2.524300441826215, "grad_norm": 71.40148162841797, "learning_rate": 3.149380130376109e-07, "loss": 0.6493, "step": 3000 } ], "logging_steps": 1, "max_steps": 3567, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.645349309207378e+17, "train_batch_size": 2, "trial_name": null, "trial_params": null }