{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.0989328740449605, "eval_steps": 500, "global_step": 11000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 8.993897640450955e-06, "grad_norm": 624.5777587890625, "learning_rate": 0.0, "loss": 17.9534, "step": 1 }, { "epoch": 1.798779528090191e-05, "grad_norm": 1787.1439208984375, "learning_rate": 9.090909090909091e-08, "loss": 17.9833, "step": 2 }, { "epoch": 2.6981692921352862e-05, "grad_norm": 543.7571411132812, "learning_rate": 1.8181818181818183e-07, "loss": 17.4185, "step": 3 }, { "epoch": 3.597559056180382e-05, "grad_norm": 1464.9036865234375, "learning_rate": 2.7272727272727274e-07, "loss": 17.0435, "step": 4 }, { "epoch": 4.496948820225477e-05, "grad_norm": 433.85675048828125, "learning_rate": 3.6363636363636366e-07, "loss": 17.6478, "step": 5 }, { "epoch": 5.3963385842705725e-05, "grad_norm": 618.152587890625, "learning_rate": 4.545454545454545e-07, "loss": 17.6259, "step": 6 }, { "epoch": 6.295728348315668e-05, "grad_norm": 723.3334350585938, "learning_rate": 5.454545454545455e-07, "loss": 18.6397, "step": 7 }, { "epoch": 7.195118112360764e-05, "grad_norm": 842.21875, "learning_rate": 6.363636363636364e-07, "loss": 17.0937, "step": 8 }, { "epoch": 8.094507876405858e-05, "grad_norm": 575.5238647460938, "learning_rate": 7.272727272727273e-07, "loss": 17.2475, "step": 9 }, { "epoch": 8.993897640450954e-05, "grad_norm": 1246.06787109375, "learning_rate": 8.181818181818183e-07, "loss": 17.6788, "step": 10 }, { "epoch": 9.89328740449605e-05, "grad_norm": 703.3920288085938, "learning_rate": 9.09090909090909e-07, "loss": 17.6445, "step": 11 }, { "epoch": 0.00010792677168541145, "grad_norm": 789.861328125, "learning_rate": 1.0000000000000002e-06, "loss": 18.0761, "step": 12 }, { "epoch": 0.0001169206693258624, "grad_norm": 1688.398193359375, "learning_rate": 1.090909090909091e-06, "loss": 17.1804, "step": 13 }, { "epoch": 0.00012591456696631336, "grad_norm": 970.5971069335938, "learning_rate": 1.1818181818181818e-06, "loss": 17.8007, "step": 14 }, { "epoch": 0.00013490846460676432, "grad_norm": 484.73492431640625, "learning_rate": 1.2727272727272728e-06, "loss": 17.3304, "step": 15 }, { "epoch": 0.00014390236224721528, "grad_norm": 1148.683349609375, "learning_rate": 1.3636363636363636e-06, "loss": 18.1804, "step": 16 }, { "epoch": 0.0001528962598876662, "grad_norm": 1276.140380859375, "learning_rate": 1.4545454545454546e-06, "loss": 17.8103, "step": 17 }, { "epoch": 0.00016189015752811716, "grad_norm": 1091.2808837890625, "learning_rate": 1.5454545454545457e-06, "loss": 17.7373, "step": 18 }, { "epoch": 0.00017088405516856812, "grad_norm": 1217.447021484375, "learning_rate": 1.6363636363636367e-06, "loss": 17.1021, "step": 19 }, { "epoch": 0.00017987795280901907, "grad_norm": 1050.6259765625, "learning_rate": 1.7272727272727273e-06, "loss": 17.8152, "step": 20 }, { "epoch": 0.00018887185044947003, "grad_norm": 1252.3167724609375, "learning_rate": 1.818181818181818e-06, "loss": 17.0886, "step": 21 }, { "epoch": 0.000197865748089921, "grad_norm": 1881.7572021484375, "learning_rate": 1.9090909090909095e-06, "loss": 16.6945, "step": 22 }, { "epoch": 0.00020685964573037194, "grad_norm": 969.9746704101562, "learning_rate": 2.0000000000000003e-06, "loss": 17.8223, "step": 23 }, { "epoch": 0.0002158535433708229, "grad_norm": 1115.3941650390625, "learning_rate": 2.090909090909091e-06, "loss": 16.8948, "step": 24 }, { "epoch": 0.00022484744101127386, "grad_norm": 1183.36328125, "learning_rate": 2.181818181818182e-06, "loss": 16.6381, "step": 25 }, { "epoch": 0.0002338413386517248, "grad_norm": 530.5995483398438, "learning_rate": 2.2727272727272728e-06, "loss": 17.2105, "step": 26 }, { "epoch": 0.00024283523629217577, "grad_norm": 1587.966552734375, "learning_rate": 2.3636363636363636e-06, "loss": 17.1436, "step": 27 }, { "epoch": 0.0002518291339326267, "grad_norm": 523.8456420898438, "learning_rate": 2.4545454545454544e-06, "loss": 17.1746, "step": 28 }, { "epoch": 0.0002608230315730777, "grad_norm": 452.8838806152344, "learning_rate": 2.5454545454545456e-06, "loss": 16.8128, "step": 29 }, { "epoch": 0.00026981692921352864, "grad_norm": 1319.495849609375, "learning_rate": 2.6363636363636364e-06, "loss": 17.1558, "step": 30 }, { "epoch": 0.0002788108268539796, "grad_norm": 343.67791748046875, "learning_rate": 2.7272727272727272e-06, "loss": 16.7303, "step": 31 }, { "epoch": 0.00028780472449443055, "grad_norm": 670.576904296875, "learning_rate": 2.8181818181818185e-06, "loss": 16.2864, "step": 32 }, { "epoch": 0.0002967986221348815, "grad_norm": 586.6473388671875, "learning_rate": 2.9090909090909093e-06, "loss": 16.3865, "step": 33 }, { "epoch": 0.0003057925197753324, "grad_norm": 401.2021789550781, "learning_rate": 3e-06, "loss": 16.2615, "step": 34 }, { "epoch": 0.00031478641741578337, "grad_norm": 550.7803955078125, "learning_rate": 3.0909090909090913e-06, "loss": 16.0867, "step": 35 }, { "epoch": 0.0003237803150562343, "grad_norm": 449.146728515625, "learning_rate": 3.1818181818181817e-06, "loss": 16.7781, "step": 36 }, { "epoch": 0.0003327742126966853, "grad_norm": 236.89505004882812, "learning_rate": 3.2727272727272733e-06, "loss": 16.7936, "step": 37 }, { "epoch": 0.00034176811033713624, "grad_norm": 260.0072021484375, "learning_rate": 3.363636363636364e-06, "loss": 16.7107, "step": 38 }, { "epoch": 0.0003507620079775872, "grad_norm": 362.1142578125, "learning_rate": 3.4545454545454545e-06, "loss": 15.8317, "step": 39 }, { "epoch": 0.00035975590561803815, "grad_norm": 176.87826538085938, "learning_rate": 3.5454545454545454e-06, "loss": 16.213, "step": 40 }, { "epoch": 0.0003687498032584891, "grad_norm": 311.5877685546875, "learning_rate": 3.636363636363636e-06, "loss": 17.5375, "step": 41 }, { "epoch": 0.00037774370089894006, "grad_norm": 157.58517456054688, "learning_rate": 3.727272727272727e-06, "loss": 15.7588, "step": 42 }, { "epoch": 0.000386737598539391, "grad_norm": 180.6936492919922, "learning_rate": 3.818181818181819e-06, "loss": 16.0854, "step": 43 }, { "epoch": 0.000395731496179842, "grad_norm": 256.3274841308594, "learning_rate": 3.90909090909091e-06, "loss": 15.3464, "step": 44 }, { "epoch": 0.00040472539382029293, "grad_norm": 831.1060791015625, "learning_rate": 4.000000000000001e-06, "loss": 16.4678, "step": 45 }, { "epoch": 0.0004137192914607439, "grad_norm": 295.3250732421875, "learning_rate": 4.0909090909090915e-06, "loss": 16.0141, "step": 46 }, { "epoch": 0.00042271318910119484, "grad_norm": 260.3460388183594, "learning_rate": 4.181818181818182e-06, "loss": 16.0095, "step": 47 }, { "epoch": 0.0004317070867416458, "grad_norm": 753.9771118164062, "learning_rate": 4.272727272727273e-06, "loss": 15.0963, "step": 48 }, { "epoch": 0.00044070098438209676, "grad_norm": 602.2676391601562, "learning_rate": 4.363636363636364e-06, "loss": 15.3355, "step": 49 }, { "epoch": 0.0004496948820225477, "grad_norm": 550.3294677734375, "learning_rate": 4.454545454545455e-06, "loss": 15.6943, "step": 50 }, { "epoch": 0.00045868877966299867, "grad_norm": 525.8038330078125, "learning_rate": 4.5454545454545455e-06, "loss": 16.3719, "step": 51 }, { "epoch": 0.0004676826773034496, "grad_norm": 331.14923095703125, "learning_rate": 4.636363636363636e-06, "loss": 15.1211, "step": 52 }, { "epoch": 0.0004766765749439006, "grad_norm": 355.77093505859375, "learning_rate": 4.727272727272727e-06, "loss": 15.1063, "step": 53 }, { "epoch": 0.00048567047258435154, "grad_norm": 3031.103515625, "learning_rate": 4.818181818181818e-06, "loss": 15.4299, "step": 54 }, { "epoch": 0.0004946643702248024, "grad_norm": 418.1381530761719, "learning_rate": 4.909090909090909e-06, "loss": 15.4191, "step": 55 }, { "epoch": 0.0005036582678652535, "grad_norm": 911.3734130859375, "learning_rate": 5e-06, "loss": 15.5616, "step": 56 }, { "epoch": 0.0005126521655057044, "grad_norm": 185.1163330078125, "learning_rate": 5.090909090909091e-06, "loss": 14.8612, "step": 57 }, { "epoch": 0.0005216460631461554, "grad_norm": 627.5923461914062, "learning_rate": 5.181818181818182e-06, "loss": 14.6524, "step": 58 }, { "epoch": 0.0005306399607866063, "grad_norm": 139.53121948242188, "learning_rate": 5.272727272727273e-06, "loss": 15.3519, "step": 59 }, { "epoch": 0.0005396338584270573, "grad_norm": 830.2334594726562, "learning_rate": 5.363636363636364e-06, "loss": 14.6705, "step": 60 }, { "epoch": 0.0005486277560675082, "grad_norm": 1547.2772216796875, "learning_rate": 5.4545454545454545e-06, "loss": 14.7645, "step": 61 }, { "epoch": 0.0005576216537079592, "grad_norm": 1675.0274658203125, "learning_rate": 5.545454545454546e-06, "loss": 14.1208, "step": 62 }, { "epoch": 0.0005666155513484101, "grad_norm": 405.97027587890625, "learning_rate": 5.636363636363637e-06, "loss": 15.0928, "step": 63 }, { "epoch": 0.0005756094489888611, "grad_norm": 611.3431396484375, "learning_rate": 5.727272727272728e-06, "loss": 14.0582, "step": 64 }, { "epoch": 0.000584603346629312, "grad_norm": 3743.0625, "learning_rate": 5.8181818181818185e-06, "loss": 14.564, "step": 65 }, { "epoch": 0.000593597244269763, "grad_norm": 2080.346923828125, "learning_rate": 5.909090909090909e-06, "loss": 15.3672, "step": 66 }, { "epoch": 0.0006025911419102139, "grad_norm": 847.2843627929688, "learning_rate": 6e-06, "loss": 13.6193, "step": 67 }, { "epoch": 0.0006115850395506648, "grad_norm": 710.4701538085938, "learning_rate": 6.090909090909091e-06, "loss": 13.9402, "step": 68 }, { "epoch": 0.0006205789371911158, "grad_norm": 565.0715942382812, "learning_rate": 6.181818181818183e-06, "loss": 15.2938, "step": 69 }, { "epoch": 0.0006295728348315667, "grad_norm": 715.6685180664062, "learning_rate": 6.2727272727272734e-06, "loss": 14.6494, "step": 70 }, { "epoch": 0.0006385667324720177, "grad_norm": 686.0399780273438, "learning_rate": 6.363636363636363e-06, "loss": 13.6058, "step": 71 }, { "epoch": 0.0006475606301124686, "grad_norm": 903.3383178710938, "learning_rate": 6.454545454545455e-06, "loss": 14.07, "step": 72 }, { "epoch": 0.0006565545277529197, "grad_norm": 281.6870422363281, "learning_rate": 6.545454545454547e-06, "loss": 14.7523, "step": 73 }, { "epoch": 0.0006655484253933706, "grad_norm": 806.2594604492188, "learning_rate": 6.636363636363637e-06, "loss": 12.7969, "step": 74 }, { "epoch": 0.0006745423230338216, "grad_norm": 399.9375, "learning_rate": 6.727272727272728e-06, "loss": 14.0856, "step": 75 }, { "epoch": 0.0006835362206742725, "grad_norm": 641.3349609375, "learning_rate": 6.818181818181818e-06, "loss": 12.93, "step": 76 }, { "epoch": 0.0006925301183147235, "grad_norm": 395.2854309082031, "learning_rate": 6.909090909090909e-06, "loss": 13.2835, "step": 77 }, { "epoch": 0.0007015240159551744, "grad_norm": 838.3772583007812, "learning_rate": 7.000000000000001e-06, "loss": 12.6786, "step": 78 }, { "epoch": 0.0007105179135956254, "grad_norm": 196.3134002685547, "learning_rate": 7.090909090909091e-06, "loss": 12.5385, "step": 79 }, { "epoch": 0.0007195118112360763, "grad_norm": 1469.98046875, "learning_rate": 7.181818181818182e-06, "loss": 12.5102, "step": 80 }, { "epoch": 0.0007285057088765273, "grad_norm": 538.8250122070312, "learning_rate": 7.272727272727272e-06, "loss": 12.4218, "step": 81 }, { "epoch": 0.0007374996065169782, "grad_norm": 384.11102294921875, "learning_rate": 7.363636363636364e-06, "loss": 12.2247, "step": 82 }, { "epoch": 0.0007464935041574292, "grad_norm": 225.18310546875, "learning_rate": 7.454545454545454e-06, "loss": 11.9204, "step": 83 }, { "epoch": 0.0007554874017978801, "grad_norm": 209.81674194335938, "learning_rate": 7.545454545454546e-06, "loss": 12.9126, "step": 84 }, { "epoch": 0.0007644812994383311, "grad_norm": 237.86912536621094, "learning_rate": 7.636363636363638e-06, "loss": 11.8694, "step": 85 }, { "epoch": 0.000773475197078782, "grad_norm": 231.4341278076172, "learning_rate": 7.727272727272727e-06, "loss": 11.5688, "step": 86 }, { "epoch": 0.000782469094719233, "grad_norm": 245.27500915527344, "learning_rate": 7.81818181818182e-06, "loss": 11.2395, "step": 87 }, { "epoch": 0.000791462992359684, "grad_norm": 375.0210876464844, "learning_rate": 7.909090909090909e-06, "loss": 11.0175, "step": 88 }, { "epoch": 0.000800456890000135, "grad_norm": 345.97308349609375, "learning_rate": 8.000000000000001e-06, "loss": 10.774, "step": 89 }, { "epoch": 0.0008094507876405859, "grad_norm": 206.9285125732422, "learning_rate": 8.09090909090909e-06, "loss": 10.9838, "step": 90 }, { "epoch": 0.0008184446852810368, "grad_norm": 290.99908447265625, "learning_rate": 8.181818181818183e-06, "loss": 10.8066, "step": 91 }, { "epoch": 0.0008274385829214878, "grad_norm": 455.3750915527344, "learning_rate": 8.272727272727274e-06, "loss": 10.1416, "step": 92 }, { "epoch": 0.0008364324805619387, "grad_norm": 345.7048645019531, "learning_rate": 8.363636363636365e-06, "loss": 10.3245, "step": 93 }, { "epoch": 0.0008454263782023897, "grad_norm": 171.3020782470703, "learning_rate": 8.454545454545455e-06, "loss": 9.6277, "step": 94 }, { "epoch": 0.0008544202758428406, "grad_norm": 267.947265625, "learning_rate": 8.545454545454546e-06, "loss": 10.3132, "step": 95 }, { "epoch": 0.0008634141734832916, "grad_norm": 387.4131164550781, "learning_rate": 8.636363636363637e-06, "loss": 10.5936, "step": 96 }, { "epoch": 0.0008724080711237425, "grad_norm": 417.14166259765625, "learning_rate": 8.727272727272728e-06, "loss": 9.4039, "step": 97 }, { "epoch": 0.0008814019687641935, "grad_norm": 212.96841430664062, "learning_rate": 8.818181818181819e-06, "loss": 8.9306, "step": 98 }, { "epoch": 0.0008903958664046444, "grad_norm": 179.31338500976562, "learning_rate": 8.90909090909091e-06, "loss": 8.1835, "step": 99 }, { "epoch": 0.0008993897640450954, "grad_norm": 259.7271423339844, "learning_rate": 9e-06, "loss": 9.3655, "step": 100 }, { "epoch": 0.0009083836616855463, "grad_norm": 553.3068237304688, "learning_rate": 9.090909090909091e-06, "loss": 7.9431, "step": 101 }, { "epoch": 0.0009173775593259973, "grad_norm": 1172.50341796875, "learning_rate": 9.181818181818182e-06, "loss": 7.8488, "step": 102 }, { "epoch": 0.0009263714569664482, "grad_norm": 241.0426025390625, "learning_rate": 9.272727272727273e-06, "loss": 6.5786, "step": 103 }, { "epoch": 0.0009353653546068993, "grad_norm": 540.9775390625, "learning_rate": 9.363636363636365e-06, "loss": 8.4175, "step": 104 }, { "epoch": 0.0009443592522473502, "grad_norm": 241.18971252441406, "learning_rate": 9.454545454545454e-06, "loss": 6.1545, "step": 105 }, { "epoch": 0.0009533531498878012, "grad_norm": 370.6540222167969, "learning_rate": 9.545454545454547e-06, "loss": 5.4816, "step": 106 }, { "epoch": 0.0009623470475282521, "grad_norm": 316.3355712890625, "learning_rate": 9.636363636363636e-06, "loss": 5.4708, "step": 107 }, { "epoch": 0.0009713409451687031, "grad_norm": 335.52239990234375, "learning_rate": 9.727272727272728e-06, "loss": 5.9066, "step": 108 }, { "epoch": 0.000980334842809154, "grad_norm": 212.37066650390625, "learning_rate": 9.818181818181818e-06, "loss": 6.5996, "step": 109 }, { "epoch": 0.0009893287404496049, "grad_norm": 158.48117065429688, "learning_rate": 9.90909090909091e-06, "loss": 5.0258, "step": 110 }, { "epoch": 0.000998322638090056, "grad_norm": 406.5316162109375, "learning_rate": 1e-05, "loss": 5.2208, "step": 111 }, { "epoch": 0.001007316535730507, "grad_norm": 220.3340301513672, "learning_rate": 1.0090909090909092e-05, "loss": 5.7037, "step": 112 }, { "epoch": 0.001016310433370958, "grad_norm": 117.0799789428711, "learning_rate": 1.0181818181818182e-05, "loss": 4.5769, "step": 113 }, { "epoch": 0.0010253043310114087, "grad_norm": 86.51113891601562, "learning_rate": 1.0272727272727273e-05, "loss": 4.5308, "step": 114 }, { "epoch": 0.0010342982286518597, "grad_norm": 119.74539184570312, "learning_rate": 1.0363636363636364e-05, "loss": 5.0915, "step": 115 }, { "epoch": 0.0010432921262923107, "grad_norm": 61.35383224487305, "learning_rate": 1.0454545454545455e-05, "loss": 4.4312, "step": 116 }, { "epoch": 0.0010522860239327615, "grad_norm": 58.55775451660156, "learning_rate": 1.0545454545454546e-05, "loss": 4.2491, "step": 117 }, { "epoch": 0.0010612799215732125, "grad_norm": 120.29472351074219, "learning_rate": 1.0636363636363638e-05, "loss": 4.385, "step": 118 }, { "epoch": 0.0010702738192136635, "grad_norm": 73.53257751464844, "learning_rate": 1.0727272727272727e-05, "loss": 4.9063, "step": 119 }, { "epoch": 0.0010792677168541146, "grad_norm": 110.20689392089844, "learning_rate": 1.081818181818182e-05, "loss": 4.8728, "step": 120 }, { "epoch": 0.0010882616144945653, "grad_norm": 85.38801574707031, "learning_rate": 1.0909090909090909e-05, "loss": 5.0187, "step": 121 }, { "epoch": 0.0010972555121350164, "grad_norm": 153.42276000976562, "learning_rate": 1.1000000000000001e-05, "loss": 4.186, "step": 122 }, { "epoch": 0.0011062494097754674, "grad_norm": 64.13314819335938, "learning_rate": 1.1090909090909092e-05, "loss": 4.0873, "step": 123 }, { "epoch": 0.0011152433074159184, "grad_norm": 72.71946716308594, "learning_rate": 1.1181818181818183e-05, "loss": 4.5826, "step": 124 }, { "epoch": 0.0011242372050563692, "grad_norm": 157.0845947265625, "learning_rate": 1.1272727272727274e-05, "loss": 4.8166, "step": 125 }, { "epoch": 0.0011332311026968202, "grad_norm": 87.81234741210938, "learning_rate": 1.1363636363636365e-05, "loss": 4.6643, "step": 126 }, { "epoch": 0.0011422250003372712, "grad_norm": 77.61016845703125, "learning_rate": 1.1454545454545455e-05, "loss": 4.7834, "step": 127 }, { "epoch": 0.0011512188979777222, "grad_norm": 92.09603118896484, "learning_rate": 1.1545454545454545e-05, "loss": 5.0563, "step": 128 }, { "epoch": 0.001160212795618173, "grad_norm": 72.69209289550781, "learning_rate": 1.1636363636363637e-05, "loss": 3.8202, "step": 129 }, { "epoch": 0.001169206693258624, "grad_norm": 62.451534271240234, "learning_rate": 1.1727272727272728e-05, "loss": 3.8198, "step": 130 }, { "epoch": 0.001178200590899075, "grad_norm": 136.73240661621094, "learning_rate": 1.1818181818181819e-05, "loss": 4.7253, "step": 131 }, { "epoch": 0.001187194488539526, "grad_norm": 186.7635040283203, "learning_rate": 1.190909090909091e-05, "loss": 4.0171, "step": 132 }, { "epoch": 0.0011961883861799768, "grad_norm": 150.6204833984375, "learning_rate": 1.2e-05, "loss": 3.9884, "step": 133 }, { "epoch": 0.0012051822838204278, "grad_norm": 144.12637329101562, "learning_rate": 1.2090909090909091e-05, "loss": 3.6953, "step": 134 }, { "epoch": 0.0012141761814608788, "grad_norm": 66.29656982421875, "learning_rate": 1.2181818181818182e-05, "loss": 4.5341, "step": 135 }, { "epoch": 0.0012231700791013296, "grad_norm": 122.22705078125, "learning_rate": 1.2272727272727273e-05, "loss": 5.4335, "step": 136 }, { "epoch": 0.0012321639767417807, "grad_norm": 120.67552185058594, "learning_rate": 1.2363636363636365e-05, "loss": 3.7019, "step": 137 }, { "epoch": 0.0012411578743822317, "grad_norm": 60.150630950927734, "learning_rate": 1.2454545454545454e-05, "loss": 3.3658, "step": 138 }, { "epoch": 0.0012501517720226827, "grad_norm": 148.12303161621094, "learning_rate": 1.2545454545454547e-05, "loss": 3.3753, "step": 139 }, { "epoch": 0.0012591456696631335, "grad_norm": 113.60433959960938, "learning_rate": 1.2636363636363638e-05, "loss": 3.999, "step": 140 }, { "epoch": 0.0012681395673035845, "grad_norm": 149.37197875976562, "learning_rate": 1.2727272727272727e-05, "loss": 4.0167, "step": 141 }, { "epoch": 0.0012771334649440355, "grad_norm": 93.42738342285156, "learning_rate": 1.2818181818181818e-05, "loss": 3.5332, "step": 142 }, { "epoch": 0.0012861273625844865, "grad_norm": 125.68160247802734, "learning_rate": 1.290909090909091e-05, "loss": 4.0157, "step": 143 }, { "epoch": 0.0012951212602249373, "grad_norm": 153.78225708007812, "learning_rate": 1.3000000000000001e-05, "loss": 3.9976, "step": 144 }, { "epoch": 0.0013041151578653883, "grad_norm": 180.440673828125, "learning_rate": 1.3090909090909093e-05, "loss": 3.8925, "step": 145 }, { "epoch": 0.0013131090555058393, "grad_norm": 95.41625213623047, "learning_rate": 1.318181818181818e-05, "loss": 3.7465, "step": 146 }, { "epoch": 0.0013221029531462903, "grad_norm": 122.04198455810547, "learning_rate": 1.3272727272727273e-05, "loss": 3.8933, "step": 147 }, { "epoch": 0.0013310968507867411, "grad_norm": 107.50264739990234, "learning_rate": 1.3363636363636364e-05, "loss": 4.019, "step": 148 }, { "epoch": 0.0013400907484271921, "grad_norm": 98.88459777832031, "learning_rate": 1.3454545454545457e-05, "loss": 4.0481, "step": 149 }, { "epoch": 0.0013490846460676431, "grad_norm": 160.04849243164062, "learning_rate": 1.3545454545454547e-05, "loss": 3.3143, "step": 150 }, { "epoch": 0.0013580785437080941, "grad_norm": 92.48059844970703, "learning_rate": 1.3636363636363637e-05, "loss": 3.3451, "step": 151 }, { "epoch": 0.001367072441348545, "grad_norm": 160.9433135986328, "learning_rate": 1.3727272727272727e-05, "loss": 3.67, "step": 152 }, { "epoch": 0.001376066338988996, "grad_norm": 155.68333435058594, "learning_rate": 1.3818181818181818e-05, "loss": 3.6389, "step": 153 }, { "epoch": 0.001385060236629447, "grad_norm": 243.73287963867188, "learning_rate": 1.390909090909091e-05, "loss": 3.9784, "step": 154 }, { "epoch": 0.001394054134269898, "grad_norm": 247.50204467773438, "learning_rate": 1.4000000000000001e-05, "loss": 4.2037, "step": 155 }, { "epoch": 0.0014030480319103488, "grad_norm": 137.28292846679688, "learning_rate": 1.409090909090909e-05, "loss": 3.4788, "step": 156 }, { "epoch": 0.0014120419295507998, "grad_norm": 143.1851806640625, "learning_rate": 1.4181818181818181e-05, "loss": 3.0819, "step": 157 }, { "epoch": 0.0014210358271912508, "grad_norm": 239.31680297851562, "learning_rate": 1.4272727272727274e-05, "loss": 3.7093, "step": 158 }, { "epoch": 0.0014300297248317016, "grad_norm": 223.9289093017578, "learning_rate": 1.4363636363636365e-05, "loss": 4.1168, "step": 159 }, { "epoch": 0.0014390236224721526, "grad_norm": 140.57806396484375, "learning_rate": 1.4454545454545454e-05, "loss": 4.0792, "step": 160 }, { "epoch": 0.0014480175201126036, "grad_norm": 187.22354125976562, "learning_rate": 1.4545454545454545e-05, "loss": 3.3268, "step": 161 }, { "epoch": 0.0014570114177530546, "grad_norm": 206.4626922607422, "learning_rate": 1.4636363636363637e-05, "loss": 3.9791, "step": 162 }, { "epoch": 0.0014660053153935054, "grad_norm": 608.0723266601562, "learning_rate": 1.4727272727272728e-05, "loss": 3.7737, "step": 163 }, { "epoch": 0.0014749992130339564, "grad_norm": 162.69952392578125, "learning_rate": 1.481818181818182e-05, "loss": 3.4041, "step": 164 }, { "epoch": 0.0014839931106744074, "grad_norm": 98.44976043701172, "learning_rate": 1.4909090909090908e-05, "loss": 3.2322, "step": 165 }, { "epoch": 0.0014929870083148584, "grad_norm": 110.34259796142578, "learning_rate": 1.5e-05, "loss": 3.4905, "step": 166 }, { "epoch": 0.0015019809059553092, "grad_norm": 157.68157958984375, "learning_rate": 1.5090909090909091e-05, "loss": 3.7214, "step": 167 }, { "epoch": 0.0015109748035957602, "grad_norm": 149.73728942871094, "learning_rate": 1.5181818181818184e-05, "loss": 3.937, "step": 168 }, { "epoch": 0.0015199687012362113, "grad_norm": 156.5905303955078, "learning_rate": 1.5272727272727276e-05, "loss": 2.9693, "step": 169 }, { "epoch": 0.0015289625988766623, "grad_norm": 192.3663330078125, "learning_rate": 1.5363636363636362e-05, "loss": 3.0357, "step": 170 }, { "epoch": 0.001537956496517113, "grad_norm": 106.77263641357422, "learning_rate": 1.5454545454545454e-05, "loss": 4.5128, "step": 171 }, { "epoch": 0.001546950394157564, "grad_norm": 225.660888671875, "learning_rate": 1.5545454545454547e-05, "loss": 3.5855, "step": 172 }, { "epoch": 0.001555944291798015, "grad_norm": 179.28977966308594, "learning_rate": 1.563636363636364e-05, "loss": 3.9906, "step": 173 }, { "epoch": 0.001564938189438466, "grad_norm": 120.28265380859375, "learning_rate": 1.572727272727273e-05, "loss": 3.1352, "step": 174 }, { "epoch": 0.0015739320870789169, "grad_norm": 107.6299819946289, "learning_rate": 1.5818181818181818e-05, "loss": 3.7572, "step": 175 }, { "epoch": 0.001582925984719368, "grad_norm": 257.49505615234375, "learning_rate": 1.590909090909091e-05, "loss": 3.3607, "step": 176 }, { "epoch": 0.001591919882359819, "grad_norm": 246.9903564453125, "learning_rate": 1.6000000000000003e-05, "loss": 3.3219, "step": 177 }, { "epoch": 0.00160091378000027, "grad_norm": 137.95632934570312, "learning_rate": 1.6090909090909092e-05, "loss": 3.1963, "step": 178 }, { "epoch": 0.0016099076776407207, "grad_norm": 79.78209686279297, "learning_rate": 1.618181818181818e-05, "loss": 3.9131, "step": 179 }, { "epoch": 0.0016189015752811717, "grad_norm": 92.68106079101562, "learning_rate": 1.6272727272727273e-05, "loss": 3.792, "step": 180 }, { "epoch": 0.0016278954729216227, "grad_norm": 286.2545166015625, "learning_rate": 1.6363636363636366e-05, "loss": 3.1824, "step": 181 }, { "epoch": 0.0016368893705620735, "grad_norm": 374.82830810546875, "learning_rate": 1.6454545454545455e-05, "loss": 3.6738, "step": 182 }, { "epoch": 0.0016458832682025245, "grad_norm": 263.861083984375, "learning_rate": 1.6545454545454548e-05, "loss": 3.6791, "step": 183 }, { "epoch": 0.0016548771658429755, "grad_norm": 279.99603271484375, "learning_rate": 1.6636363636363637e-05, "loss": 3.0861, "step": 184 }, { "epoch": 0.0016638710634834266, "grad_norm": 330.89990234375, "learning_rate": 1.672727272727273e-05, "loss": 3.091, "step": 185 }, { "epoch": 0.0016728649611238774, "grad_norm": 168.16578674316406, "learning_rate": 1.6818181818181818e-05, "loss": 3.1189, "step": 186 }, { "epoch": 0.0016818588587643284, "grad_norm": 136.7516632080078, "learning_rate": 1.690909090909091e-05, "loss": 3.5708, "step": 187 }, { "epoch": 0.0016908527564047794, "grad_norm": 371.912109375, "learning_rate": 1.7000000000000003e-05, "loss": 3.5552, "step": 188 }, { "epoch": 0.0016998466540452304, "grad_norm": 133.0180206298828, "learning_rate": 1.7090909090909092e-05, "loss": 3.1068, "step": 189 }, { "epoch": 0.0017088405516856812, "grad_norm": 122.30583953857422, "learning_rate": 1.718181818181818e-05, "loss": 3.2482, "step": 190 }, { "epoch": 0.0017178344493261322, "grad_norm": 139.85595703125, "learning_rate": 1.7272727272727274e-05, "loss": 3.1804, "step": 191 }, { "epoch": 0.0017268283469665832, "grad_norm": 85.83197021484375, "learning_rate": 1.7363636363636366e-05, "loss": 3.5153, "step": 192 }, { "epoch": 0.0017358222446070342, "grad_norm": 82.42160034179688, "learning_rate": 1.7454545454545456e-05, "loss": 3.5141, "step": 193 }, { "epoch": 0.001744816142247485, "grad_norm": 77.44760131835938, "learning_rate": 1.7545454545454545e-05, "loss": 3.0701, "step": 194 }, { "epoch": 0.001753810039887936, "grad_norm": 111.4917984008789, "learning_rate": 1.7636363636363637e-05, "loss": 3.075, "step": 195 }, { "epoch": 0.001762803937528387, "grad_norm": 85.65130615234375, "learning_rate": 1.772727272727273e-05, "loss": 3.1711, "step": 196 }, { "epoch": 0.001771797835168838, "grad_norm": 64.50166320800781, "learning_rate": 1.781818181818182e-05, "loss": 3.0664, "step": 197 }, { "epoch": 0.0017807917328092888, "grad_norm": 251.89498901367188, "learning_rate": 1.7909090909090908e-05, "loss": 3.7399, "step": 198 }, { "epoch": 0.0017897856304497398, "grad_norm": 157.12791442871094, "learning_rate": 1.8e-05, "loss": 3.4729, "step": 199 }, { "epoch": 0.0017987795280901909, "grad_norm": 152.71597290039062, "learning_rate": 1.8090909090909093e-05, "loss": 3.0034, "step": 200 }, { "epoch": 0.0018077734257306419, "grad_norm": 263.5816650390625, "learning_rate": 1.8181818181818182e-05, "loss": 3.4371, "step": 201 }, { "epoch": 0.0018167673233710927, "grad_norm": 119.90428924560547, "learning_rate": 1.8272727272727275e-05, "loss": 3.3518, "step": 202 }, { "epoch": 0.0018257612210115437, "grad_norm": 190.151123046875, "learning_rate": 1.8363636363636364e-05, "loss": 3.7226, "step": 203 }, { "epoch": 0.0018347551186519947, "grad_norm": 180.32394409179688, "learning_rate": 1.8454545454545456e-05, "loss": 3.1088, "step": 204 }, { "epoch": 0.0018437490162924455, "grad_norm": 70.09622955322266, "learning_rate": 1.8545454545454545e-05, "loss": 3.1764, "step": 205 }, { "epoch": 0.0018527429139328965, "grad_norm": 135.13621520996094, "learning_rate": 1.8636363636363638e-05, "loss": 3.3676, "step": 206 }, { "epoch": 0.0018617368115733475, "grad_norm": 225.05950927734375, "learning_rate": 1.872727272727273e-05, "loss": 3.66, "step": 207 }, { "epoch": 0.0018707307092137985, "grad_norm": 122.18893432617188, "learning_rate": 1.881818181818182e-05, "loss": 2.9396, "step": 208 }, { "epoch": 0.0018797246068542493, "grad_norm": 85.67851257324219, "learning_rate": 1.890909090909091e-05, "loss": 2.81, "step": 209 }, { "epoch": 0.0018887185044947003, "grad_norm": 112.75260162353516, "learning_rate": 1.9e-05, "loss": 3.1241, "step": 210 }, { "epoch": 0.0018977124021351513, "grad_norm": 76.12918853759766, "learning_rate": 1.9090909090909094e-05, "loss": 3.0039, "step": 211 }, { "epoch": 0.0019067062997756023, "grad_norm": 103.438232421875, "learning_rate": 1.9181818181818183e-05, "loss": 3.3159, "step": 212 }, { "epoch": 0.0019157001974160531, "grad_norm": 64.7052993774414, "learning_rate": 1.9272727272727272e-05, "loss": 2.8556, "step": 213 }, { "epoch": 0.0019246940950565041, "grad_norm": 67.88910675048828, "learning_rate": 1.9363636363636364e-05, "loss": 3.1821, "step": 214 }, { "epoch": 0.0019336879926969551, "grad_norm": 138.4066925048828, "learning_rate": 1.9454545454545457e-05, "loss": 3.1407, "step": 215 }, { "epoch": 0.0019426818903374062, "grad_norm": 48.77667999267578, "learning_rate": 1.9545454545454546e-05, "loss": 3.2218, "step": 216 }, { "epoch": 0.001951675787977857, "grad_norm": 129.01211547851562, "learning_rate": 1.9636363636363635e-05, "loss": 3.1226, "step": 217 }, { "epoch": 0.001960669685618308, "grad_norm": 137.63018798828125, "learning_rate": 1.9727272727272728e-05, "loss": 3.1247, "step": 218 }, { "epoch": 0.001969663583258759, "grad_norm": 69.55642700195312, "learning_rate": 1.981818181818182e-05, "loss": 3.2127, "step": 219 }, { "epoch": 0.0019786574808992098, "grad_norm": 139.25526428222656, "learning_rate": 1.990909090909091e-05, "loss": 3.0208, "step": 220 }, { "epoch": 0.001987651378539661, "grad_norm": 95.1934814453125, "learning_rate": 2e-05, "loss": 3.1067, "step": 221 }, { "epoch": 0.001996645276180112, "grad_norm": 92.59369659423828, "learning_rate": 2.009090909090909e-05, "loss": 3.3906, "step": 222 }, { "epoch": 0.0020056391738205626, "grad_norm": 111.0574951171875, "learning_rate": 2.0181818181818183e-05, "loss": 3.7034, "step": 223 }, { "epoch": 0.002014633071461014, "grad_norm": 189.96214294433594, "learning_rate": 2.0272727272727272e-05, "loss": 3.1057, "step": 224 }, { "epoch": 0.0020236269691014646, "grad_norm": 61.11247634887695, "learning_rate": 2.0363636363636365e-05, "loss": 3.2343, "step": 225 }, { "epoch": 0.002032620866741916, "grad_norm": 35.40169906616211, "learning_rate": 2.0454545454545457e-05, "loss": 2.7804, "step": 226 }, { "epoch": 0.0020416147643823666, "grad_norm": 52.55931091308594, "learning_rate": 2.0545454545454546e-05, "loss": 3.3604, "step": 227 }, { "epoch": 0.0020506086620228174, "grad_norm": 72.27637481689453, "learning_rate": 2.0636363636363636e-05, "loss": 3.1553, "step": 228 }, { "epoch": 0.0020596025596632686, "grad_norm": 40.75813674926758, "learning_rate": 2.0727272727272728e-05, "loss": 3.083, "step": 229 }, { "epoch": 0.0020685964573037194, "grad_norm": 91.76944732666016, "learning_rate": 2.081818181818182e-05, "loss": 2.9126, "step": 230 }, { "epoch": 0.0020775903549441702, "grad_norm": 91.73649597167969, "learning_rate": 2.090909090909091e-05, "loss": 2.8581, "step": 231 }, { "epoch": 0.0020865842525846215, "grad_norm": 57.67620086669922, "learning_rate": 2.1e-05, "loss": 2.9455, "step": 232 }, { "epoch": 0.0020955781502250722, "grad_norm": 31.565837860107422, "learning_rate": 2.109090909090909e-05, "loss": 2.7875, "step": 233 }, { "epoch": 0.002104572047865523, "grad_norm": 61.9347038269043, "learning_rate": 2.1181818181818184e-05, "loss": 2.9887, "step": 234 }, { "epoch": 0.0021135659455059743, "grad_norm": 68.58342742919922, "learning_rate": 2.1272727272727276e-05, "loss": 2.8304, "step": 235 }, { "epoch": 0.002122559843146425, "grad_norm": 69.1456069946289, "learning_rate": 2.1363636363636362e-05, "loss": 3.029, "step": 236 }, { "epoch": 0.0021315537407868763, "grad_norm": 96.83798217773438, "learning_rate": 2.1454545454545455e-05, "loss": 3.1787, "step": 237 }, { "epoch": 0.002140547638427327, "grad_norm": 69.93182373046875, "learning_rate": 2.1545454545454547e-05, "loss": 2.9907, "step": 238 }, { "epoch": 0.002149541536067778, "grad_norm": 98.68118286132812, "learning_rate": 2.163636363636364e-05, "loss": 2.9288, "step": 239 }, { "epoch": 0.002158535433708229, "grad_norm": 59.91753387451172, "learning_rate": 2.172727272727273e-05, "loss": 3.9726, "step": 240 }, { "epoch": 0.00216752933134868, "grad_norm": 195.9169921875, "learning_rate": 2.1818181818181818e-05, "loss": 3.191, "step": 241 }, { "epoch": 0.0021765232289891307, "grad_norm": 115.80036163330078, "learning_rate": 2.190909090909091e-05, "loss": 3.262, "step": 242 }, { "epoch": 0.002185517126629582, "grad_norm": 192.85691833496094, "learning_rate": 2.2000000000000003e-05, "loss": 3.0146, "step": 243 }, { "epoch": 0.0021945110242700327, "grad_norm": 239.3960723876953, "learning_rate": 2.2090909090909092e-05, "loss": 3.0813, "step": 244 }, { "epoch": 0.002203504921910484, "grad_norm": 80.44845581054688, "learning_rate": 2.2181818181818184e-05, "loss": 2.8697, "step": 245 }, { "epoch": 0.0022124988195509347, "grad_norm": 215.9846649169922, "learning_rate": 2.2272727272727274e-05, "loss": 3.1669, "step": 246 }, { "epoch": 0.0022214927171913855, "grad_norm": 343.25738525390625, "learning_rate": 2.2363636363636366e-05, "loss": 3.1174, "step": 247 }, { "epoch": 0.0022304866148318368, "grad_norm": 384.26153564453125, "learning_rate": 2.2454545454545455e-05, "loss": 3.008, "step": 248 }, { "epoch": 0.0022394805124722876, "grad_norm": 124.88306427001953, "learning_rate": 2.2545454545454548e-05, "loss": 2.9283, "step": 249 }, { "epoch": 0.0022484744101127383, "grad_norm": 103.40723419189453, "learning_rate": 2.2636363636363637e-05, "loss": 2.8574, "step": 250 }, { "epoch": 0.0022574683077531896, "grad_norm": 160.85443115234375, "learning_rate": 2.272727272727273e-05, "loss": 3.5287, "step": 251 }, { "epoch": 0.0022664622053936404, "grad_norm": 45.96371078491211, "learning_rate": 2.281818181818182e-05, "loss": 2.8872, "step": 252 }, { "epoch": 0.002275456103034091, "grad_norm": 90.8829116821289, "learning_rate": 2.290909090909091e-05, "loss": 3.2123, "step": 253 }, { "epoch": 0.0022844500006745424, "grad_norm": 91.640869140625, "learning_rate": 2.3000000000000003e-05, "loss": 2.7109, "step": 254 }, { "epoch": 0.002293443898314993, "grad_norm": 47.6387939453125, "learning_rate": 2.309090909090909e-05, "loss": 3.4242, "step": 255 }, { "epoch": 0.0023024377959554444, "grad_norm": 108.52681732177734, "learning_rate": 2.318181818181818e-05, "loss": 3.0449, "step": 256 }, { "epoch": 0.002311431693595895, "grad_norm": 62.707340240478516, "learning_rate": 2.3272727272727274e-05, "loss": 3.1075, "step": 257 }, { "epoch": 0.002320425591236346, "grad_norm": 115.00871276855469, "learning_rate": 2.3363636363636367e-05, "loss": 2.7261, "step": 258 }, { "epoch": 0.0023294194888767972, "grad_norm": 136.74339294433594, "learning_rate": 2.3454545454545456e-05, "loss": 3.1472, "step": 259 }, { "epoch": 0.002338413386517248, "grad_norm": 100.91593933105469, "learning_rate": 2.3545454545454545e-05, "loss": 2.7715, "step": 260 }, { "epoch": 0.002347407284157699, "grad_norm": 35.69953536987305, "learning_rate": 2.3636363636363637e-05, "loss": 2.691, "step": 261 }, { "epoch": 0.00235640118179815, "grad_norm": 35.81983184814453, "learning_rate": 2.372727272727273e-05, "loss": 2.8004, "step": 262 }, { "epoch": 0.002365395079438601, "grad_norm": 72.92234802246094, "learning_rate": 2.381818181818182e-05, "loss": 2.7828, "step": 263 }, { "epoch": 0.002374388977079052, "grad_norm": 62.81526565551758, "learning_rate": 2.390909090909091e-05, "loss": 2.9331, "step": 264 }, { "epoch": 0.002383382874719503, "grad_norm": 265.2322692871094, "learning_rate": 2.4e-05, "loss": 2.9601, "step": 265 }, { "epoch": 0.0023923767723599536, "grad_norm": 134.36927795410156, "learning_rate": 2.4090909090909093e-05, "loss": 2.8622, "step": 266 }, { "epoch": 0.002401370670000405, "grad_norm": 72.66773223876953, "learning_rate": 2.4181818181818182e-05, "loss": 2.9547, "step": 267 }, { "epoch": 0.0024103645676408557, "grad_norm": 79.63045501708984, "learning_rate": 2.4272727272727275e-05, "loss": 2.8842, "step": 268 }, { "epoch": 0.0024193584652813065, "grad_norm": 34.807342529296875, "learning_rate": 2.4363636363636364e-05, "loss": 2.9106, "step": 269 }, { "epoch": 0.0024283523629217577, "grad_norm": 64.01009368896484, "learning_rate": 2.4454545454545456e-05, "loss": 3.0516, "step": 270 }, { "epoch": 0.0024373462605622085, "grad_norm": 123.88774108886719, "learning_rate": 2.4545454545454545e-05, "loss": 3.1205, "step": 271 }, { "epoch": 0.0024463401582026593, "grad_norm": 32.183597564697266, "learning_rate": 2.4636363636363638e-05, "loss": 2.7969, "step": 272 }, { "epoch": 0.0024553340558431105, "grad_norm": 34.47355270385742, "learning_rate": 2.472727272727273e-05, "loss": 2.6981, "step": 273 }, { "epoch": 0.0024643279534835613, "grad_norm": 35.006256103515625, "learning_rate": 2.481818181818182e-05, "loss": 2.6709, "step": 274 }, { "epoch": 0.0024733218511240125, "grad_norm": 42.49452590942383, "learning_rate": 2.490909090909091e-05, "loss": 2.6843, "step": 275 }, { "epoch": 0.0024823157487644633, "grad_norm": 29.628705978393555, "learning_rate": 2.5e-05, "loss": 2.7749, "step": 276 }, { "epoch": 0.002491309646404914, "grad_norm": 33.55371856689453, "learning_rate": 2.5090909090909094e-05, "loss": 2.7034, "step": 277 }, { "epoch": 0.0025003035440453653, "grad_norm": 33.486297607421875, "learning_rate": 2.5181818181818183e-05, "loss": 2.7372, "step": 278 }, { "epoch": 0.002509297441685816, "grad_norm": 26.911794662475586, "learning_rate": 2.5272727272727275e-05, "loss": 2.8555, "step": 279 }, { "epoch": 0.002518291339326267, "grad_norm": 28.314470291137695, "learning_rate": 2.5363636363636368e-05, "loss": 2.8832, "step": 280 }, { "epoch": 0.002527285236966718, "grad_norm": 86.84230041503906, "learning_rate": 2.5454545454545454e-05, "loss": 2.7332, "step": 281 }, { "epoch": 0.002536279134607169, "grad_norm": 73.28656005859375, "learning_rate": 2.5545454545454546e-05, "loss": 2.7645, "step": 282 }, { "epoch": 0.00254527303224762, "grad_norm": 50.48367691040039, "learning_rate": 2.5636363636363635e-05, "loss": 2.7545, "step": 283 }, { "epoch": 0.002554266929888071, "grad_norm": 79.51824188232422, "learning_rate": 2.5727272727272728e-05, "loss": 2.7741, "step": 284 }, { "epoch": 0.0025632608275285218, "grad_norm": 96.66320037841797, "learning_rate": 2.581818181818182e-05, "loss": 2.8547, "step": 285 }, { "epoch": 0.002572254725168973, "grad_norm": 41.353294372558594, "learning_rate": 2.590909090909091e-05, "loss": 2.7092, "step": 286 }, { "epoch": 0.002581248622809424, "grad_norm": 68.37921905517578, "learning_rate": 2.6000000000000002e-05, "loss": 2.6085, "step": 287 }, { "epoch": 0.0025902425204498746, "grad_norm": 51.16245651245117, "learning_rate": 2.6090909090909094e-05, "loss": 2.7847, "step": 288 }, { "epoch": 0.002599236418090326, "grad_norm": 34.79716491699219, "learning_rate": 2.6181818181818187e-05, "loss": 2.7926, "step": 289 }, { "epoch": 0.0026082303157307766, "grad_norm": 349.29339599609375, "learning_rate": 2.6272727272727276e-05, "loss": 2.6578, "step": 290 }, { "epoch": 0.002617224213371228, "grad_norm": 299.22015380859375, "learning_rate": 2.636363636363636e-05, "loss": 3.0245, "step": 291 }, { "epoch": 0.0026262181110116786, "grad_norm": 99.52041625976562, "learning_rate": 2.6454545454545454e-05, "loss": 2.8194, "step": 292 }, { "epoch": 0.0026352120086521294, "grad_norm": 52.22737503051758, "learning_rate": 2.6545454545454547e-05, "loss": 2.8177, "step": 293 }, { "epoch": 0.0026442059062925806, "grad_norm": 96.27494049072266, "learning_rate": 2.6636363636363636e-05, "loss": 3.0188, "step": 294 }, { "epoch": 0.0026531998039330314, "grad_norm": 197.97125244140625, "learning_rate": 2.6727272727272728e-05, "loss": 2.8892, "step": 295 }, { "epoch": 0.0026621937015734822, "grad_norm": 473.30499267578125, "learning_rate": 2.681818181818182e-05, "loss": 2.6927, "step": 296 }, { "epoch": 0.0026711875992139335, "grad_norm": 295.10626220703125, "learning_rate": 2.6909090909090913e-05, "loss": 2.7727, "step": 297 }, { "epoch": 0.0026801814968543843, "grad_norm": 419.4664001464844, "learning_rate": 2.7000000000000002e-05, "loss": 2.9912, "step": 298 }, { "epoch": 0.002689175394494835, "grad_norm": 899.4752197265625, "learning_rate": 2.7090909090909095e-05, "loss": 2.9282, "step": 299 }, { "epoch": 0.0026981692921352863, "grad_norm": 511.102783203125, "learning_rate": 2.718181818181818e-05, "loss": 3.1278, "step": 300 }, { "epoch": 0.002707163189775737, "grad_norm": 355.4637145996094, "learning_rate": 2.7272727272727273e-05, "loss": 2.7197, "step": 301 }, { "epoch": 0.0027161570874161883, "grad_norm": 87.07572937011719, "learning_rate": 2.7363636363636362e-05, "loss": 2.9893, "step": 302 }, { "epoch": 0.002725150985056639, "grad_norm": 351.97991943359375, "learning_rate": 2.7454545454545455e-05, "loss": 3.0283, "step": 303 }, { "epoch": 0.00273414488269709, "grad_norm": 347.8600158691406, "learning_rate": 2.7545454545454547e-05, "loss": 2.8555, "step": 304 }, { "epoch": 0.002743138780337541, "grad_norm": 146.10769653320312, "learning_rate": 2.7636363636363636e-05, "loss": 2.9316, "step": 305 }, { "epoch": 0.002752132677977992, "grad_norm": 42.5224609375, "learning_rate": 2.772727272727273e-05, "loss": 2.7596, "step": 306 }, { "epoch": 0.0027611265756184427, "grad_norm": 39.49795150756836, "learning_rate": 2.781818181818182e-05, "loss": 2.7516, "step": 307 }, { "epoch": 0.002770120473258894, "grad_norm": 22.18710708618164, "learning_rate": 2.7909090909090914e-05, "loss": 2.6042, "step": 308 }, { "epoch": 0.0027791143708993447, "grad_norm": 26.737184524536133, "learning_rate": 2.8000000000000003e-05, "loss": 2.7415, "step": 309 }, { "epoch": 0.002788108268539796, "grad_norm": 14.636752128601074, "learning_rate": 2.809090909090909e-05, "loss": 2.8259, "step": 310 }, { "epoch": 0.0027971021661802467, "grad_norm": 43.52849578857422, "learning_rate": 2.818181818181818e-05, "loss": 2.722, "step": 311 }, { "epoch": 0.0028060960638206975, "grad_norm": 70.25341033935547, "learning_rate": 2.8272727272727274e-05, "loss": 2.6344, "step": 312 }, { "epoch": 0.0028150899614611488, "grad_norm": 56.68631362915039, "learning_rate": 2.8363636363636363e-05, "loss": 2.7435, "step": 313 }, { "epoch": 0.0028240838591015996, "grad_norm": 31.749832153320312, "learning_rate": 2.8454545454545455e-05, "loss": 2.5849, "step": 314 }, { "epoch": 0.0028330777567420504, "grad_norm": 50.61017990112305, "learning_rate": 2.8545454545454548e-05, "loss": 2.6884, "step": 315 }, { "epoch": 0.0028420716543825016, "grad_norm": 17.99905014038086, "learning_rate": 2.863636363636364e-05, "loss": 2.6863, "step": 316 }, { "epoch": 0.0028510655520229524, "grad_norm": 52.78300476074219, "learning_rate": 2.872727272727273e-05, "loss": 2.5679, "step": 317 }, { "epoch": 0.002860059449663403, "grad_norm": 21.387653350830078, "learning_rate": 2.8818181818181822e-05, "loss": 2.6698, "step": 318 }, { "epoch": 0.0028690533473038544, "grad_norm": 21.18263053894043, "learning_rate": 2.8909090909090908e-05, "loss": 2.7708, "step": 319 }, { "epoch": 0.002878047244944305, "grad_norm": 33.6380615234375, "learning_rate": 2.9e-05, "loss": 2.6107, "step": 320 }, { "epoch": 0.0028870411425847564, "grad_norm": 26.449613571166992, "learning_rate": 2.909090909090909e-05, "loss": 2.6022, "step": 321 }, { "epoch": 0.002896035040225207, "grad_norm": 67.54065704345703, "learning_rate": 2.9181818181818182e-05, "loss": 2.8179, "step": 322 }, { "epoch": 0.002905028937865658, "grad_norm": 94.57304382324219, "learning_rate": 2.9272727272727274e-05, "loss": 2.7015, "step": 323 }, { "epoch": 0.0029140228355061092, "grad_norm": 39.76059341430664, "learning_rate": 2.9363636363636367e-05, "loss": 2.6519, "step": 324 }, { "epoch": 0.00292301673314656, "grad_norm": 14.038307189941406, "learning_rate": 2.9454545454545456e-05, "loss": 2.6765, "step": 325 }, { "epoch": 0.002932010630787011, "grad_norm": 150.26202392578125, "learning_rate": 2.954545454545455e-05, "loss": 2.8133, "step": 326 }, { "epoch": 0.002941004528427462, "grad_norm": 87.42321014404297, "learning_rate": 2.963636363636364e-05, "loss": 2.7432, "step": 327 }, { "epoch": 0.002949998426067913, "grad_norm": 80.30967712402344, "learning_rate": 2.972727272727273e-05, "loss": 2.8271, "step": 328 }, { "epoch": 0.002958992323708364, "grad_norm": 30.537057876586914, "learning_rate": 2.9818181818181816e-05, "loss": 2.6839, "step": 329 }, { "epoch": 0.002967986221348815, "grad_norm": 35.343265533447266, "learning_rate": 2.9909090909090908e-05, "loss": 2.7018, "step": 330 }, { "epoch": 0.0029769801189892657, "grad_norm": 70.98170471191406, "learning_rate": 3e-05, "loss": 2.6828, "step": 331 }, { "epoch": 0.002985974016629717, "grad_norm": 28.612892150878906, "learning_rate": 3.0090909090909093e-05, "loss": 2.5517, "step": 332 }, { "epoch": 0.0029949679142701677, "grad_norm": 60.58148193359375, "learning_rate": 3.0181818181818182e-05, "loss": 2.6384, "step": 333 }, { "epoch": 0.0030039618119106185, "grad_norm": 100.5984115600586, "learning_rate": 3.0272727272727275e-05, "loss": 2.7623, "step": 334 }, { "epoch": 0.0030129557095510697, "grad_norm": 136.9729766845703, "learning_rate": 3.0363636363636367e-05, "loss": 2.6914, "step": 335 }, { "epoch": 0.0030219496071915205, "grad_norm": 34.682979583740234, "learning_rate": 3.0454545454545456e-05, "loss": 2.7192, "step": 336 }, { "epoch": 0.0030309435048319717, "grad_norm": 77.67842864990234, "learning_rate": 3.054545454545455e-05, "loss": 2.4685, "step": 337 }, { "epoch": 0.0030399374024724225, "grad_norm": 96.45396423339844, "learning_rate": 3.063636363636364e-05, "loss": 2.5484, "step": 338 }, { "epoch": 0.0030489313001128733, "grad_norm": 52.39377975463867, "learning_rate": 3.0727272727272724e-05, "loss": 2.5122, "step": 339 }, { "epoch": 0.0030579251977533245, "grad_norm": 48.01491928100586, "learning_rate": 3.0818181818181816e-05, "loss": 2.6634, "step": 340 }, { "epoch": 0.0030669190953937753, "grad_norm": 63.227745056152344, "learning_rate": 3.090909090909091e-05, "loss": 2.5317, "step": 341 }, { "epoch": 0.003075912993034226, "grad_norm": 46.99110794067383, "learning_rate": 3.1e-05, "loss": 2.6247, "step": 342 }, { "epoch": 0.0030849068906746773, "grad_norm": 39.3392448425293, "learning_rate": 3.1090909090909094e-05, "loss": 2.6409, "step": 343 }, { "epoch": 0.003093900788315128, "grad_norm": 123.8673324584961, "learning_rate": 3.1181818181818186e-05, "loss": 2.5517, "step": 344 }, { "epoch": 0.003102894685955579, "grad_norm": 40.15580368041992, "learning_rate": 3.127272727272728e-05, "loss": 2.6248, "step": 345 }, { "epoch": 0.00311188858359603, "grad_norm": 22.070646286010742, "learning_rate": 3.1363636363636365e-05, "loss": 2.6096, "step": 346 }, { "epoch": 0.003120882481236481, "grad_norm": 17.01295280456543, "learning_rate": 3.145454545454546e-05, "loss": 2.5886, "step": 347 }, { "epoch": 0.003129876378876932, "grad_norm": 502.77716064453125, "learning_rate": 3.154545454545454e-05, "loss": 2.7185, "step": 348 }, { "epoch": 0.003138870276517383, "grad_norm": 220.7124786376953, "learning_rate": 3.1636363636363635e-05, "loss": 2.652, "step": 349 }, { "epoch": 0.0031478641741578338, "grad_norm": 40.6119384765625, "learning_rate": 3.172727272727273e-05, "loss": 2.5447, "step": 350 }, { "epoch": 0.003156858071798285, "grad_norm": 28.14447021484375, "learning_rate": 3.181818181818182e-05, "loss": 2.6028, "step": 351 }, { "epoch": 0.003165851969438736, "grad_norm": 100.99872589111328, "learning_rate": 3.190909090909091e-05, "loss": 2.4951, "step": 352 }, { "epoch": 0.0031748458670791866, "grad_norm": 184.5791778564453, "learning_rate": 3.2000000000000005e-05, "loss": 2.6375, "step": 353 }, { "epoch": 0.003183839764719638, "grad_norm": 66.14623260498047, "learning_rate": 3.209090909090909e-05, "loss": 2.6241, "step": 354 }, { "epoch": 0.0031928336623600886, "grad_norm": 89.26773071289062, "learning_rate": 3.2181818181818184e-05, "loss": 2.6124, "step": 355 }, { "epoch": 0.00320182756000054, "grad_norm": 33.29132843017578, "learning_rate": 3.2272727272727276e-05, "loss": 2.6274, "step": 356 }, { "epoch": 0.0032108214576409906, "grad_norm": 40.92216873168945, "learning_rate": 3.236363636363636e-05, "loss": 2.6266, "step": 357 }, { "epoch": 0.0032198153552814414, "grad_norm": 77.89151000976562, "learning_rate": 3.2454545454545454e-05, "loss": 2.5986, "step": 358 }, { "epoch": 0.0032288092529218927, "grad_norm": 23.291912078857422, "learning_rate": 3.254545454545455e-05, "loss": 2.5735, "step": 359 }, { "epoch": 0.0032378031505623434, "grad_norm": 41.3476448059082, "learning_rate": 3.263636363636364e-05, "loss": 2.5837, "step": 360 }, { "epoch": 0.0032467970482027942, "grad_norm": 54.72057342529297, "learning_rate": 3.272727272727273e-05, "loss": 2.6716, "step": 361 }, { "epoch": 0.0032557909458432455, "grad_norm": 41.65103530883789, "learning_rate": 3.281818181818182e-05, "loss": 2.7165, "step": 362 }, { "epoch": 0.0032647848434836963, "grad_norm": 28.426193237304688, "learning_rate": 3.290909090909091e-05, "loss": 2.5034, "step": 363 }, { "epoch": 0.003273778741124147, "grad_norm": 26.71957015991211, "learning_rate": 3.3e-05, "loss": 2.6379, "step": 364 }, { "epoch": 0.0032827726387645983, "grad_norm": 79.43277740478516, "learning_rate": 3.3090909090909095e-05, "loss": 2.6091, "step": 365 }, { "epoch": 0.003291766536405049, "grad_norm": 23.48211669921875, "learning_rate": 3.318181818181819e-05, "loss": 2.6034, "step": 366 }, { "epoch": 0.0033007604340455003, "grad_norm": 18.23671531677246, "learning_rate": 3.327272727272727e-05, "loss": 2.5584, "step": 367 }, { "epoch": 0.003309754331685951, "grad_norm": 11.874900817871094, "learning_rate": 3.3363636363636366e-05, "loss": 2.5881, "step": 368 }, { "epoch": 0.003318748229326402, "grad_norm": 18.68777847290039, "learning_rate": 3.345454545454546e-05, "loss": 2.627, "step": 369 }, { "epoch": 0.003327742126966853, "grad_norm": 32.40439987182617, "learning_rate": 3.3545454545454544e-05, "loss": 2.5897, "step": 370 }, { "epoch": 0.003336736024607304, "grad_norm": 36.62127685546875, "learning_rate": 3.3636363636363636e-05, "loss": 2.5898, "step": 371 }, { "epoch": 0.0033457299222477547, "grad_norm": 12.973684310913086, "learning_rate": 3.372727272727273e-05, "loss": 2.6294, "step": 372 }, { "epoch": 0.003354723819888206, "grad_norm": 8.09320068359375, "learning_rate": 3.381818181818182e-05, "loss": 2.5505, "step": 373 }, { "epoch": 0.0033637177175286567, "grad_norm": 11.420305252075195, "learning_rate": 3.3909090909090914e-05, "loss": 2.5784, "step": 374 }, { "epoch": 0.003372711615169108, "grad_norm": 12.049633026123047, "learning_rate": 3.4000000000000007e-05, "loss": 2.6101, "step": 375 }, { "epoch": 0.0033817055128095587, "grad_norm": 14.55691909790039, "learning_rate": 3.409090909090909e-05, "loss": 2.4904, "step": 376 }, { "epoch": 0.0033906994104500095, "grad_norm": 14.35408878326416, "learning_rate": 3.4181818181818185e-05, "loss": 2.5721, "step": 377 }, { "epoch": 0.0033996933080904608, "grad_norm": 16.358585357666016, "learning_rate": 3.427272727272727e-05, "loss": 2.5846, "step": 378 }, { "epoch": 0.0034086872057309116, "grad_norm": 13.180251121520996, "learning_rate": 3.436363636363636e-05, "loss": 2.5942, "step": 379 }, { "epoch": 0.0034176811033713624, "grad_norm": 11.884993553161621, "learning_rate": 3.4454545454545455e-05, "loss": 2.5825, "step": 380 }, { "epoch": 0.0034266750010118136, "grad_norm": 21.903844833374023, "learning_rate": 3.454545454545455e-05, "loss": 2.6786, "step": 381 }, { "epoch": 0.0034356688986522644, "grad_norm": 10.655231475830078, "learning_rate": 3.463636363636364e-05, "loss": 2.5477, "step": 382 }, { "epoch": 0.0034446627962927156, "grad_norm": 8.134358406066895, "learning_rate": 3.472727272727273e-05, "loss": 2.6264, "step": 383 }, { "epoch": 0.0034536566939331664, "grad_norm": 10.995036125183105, "learning_rate": 3.481818181818182e-05, "loss": 2.6314, "step": 384 }, { "epoch": 0.003462650591573617, "grad_norm": 11.038241386413574, "learning_rate": 3.490909090909091e-05, "loss": 2.499, "step": 385 }, { "epoch": 0.0034716444892140684, "grad_norm": 4.939131259918213, "learning_rate": 3.5e-05, "loss": 2.5725, "step": 386 }, { "epoch": 0.003480638386854519, "grad_norm": 8.1621732711792, "learning_rate": 3.509090909090909e-05, "loss": 2.5686, "step": 387 }, { "epoch": 0.00348963228449497, "grad_norm": 8.263984680175781, "learning_rate": 3.518181818181818e-05, "loss": 2.5254, "step": 388 }, { "epoch": 0.0034986261821354212, "grad_norm": 23.33096694946289, "learning_rate": 3.5272727272727274e-05, "loss": 2.5912, "step": 389 }, { "epoch": 0.003507620079775872, "grad_norm": 18.071269989013672, "learning_rate": 3.536363636363637e-05, "loss": 2.5056, "step": 390 }, { "epoch": 0.003516613977416323, "grad_norm": 18.0259952545166, "learning_rate": 3.545454545454546e-05, "loss": 2.4462, "step": 391 }, { "epoch": 0.003525607875056774, "grad_norm": 10.46397590637207, "learning_rate": 3.5545454545454545e-05, "loss": 2.527, "step": 392 }, { "epoch": 0.003534601772697225, "grad_norm": 6.038736343383789, "learning_rate": 3.563636363636364e-05, "loss": 2.5446, "step": 393 }, { "epoch": 0.003543595670337676, "grad_norm": 25.153114318847656, "learning_rate": 3.572727272727273e-05, "loss": 2.7408, "step": 394 }, { "epoch": 0.003552589567978127, "grad_norm": 19.05910873413086, "learning_rate": 3.5818181818181816e-05, "loss": 2.5632, "step": 395 }, { "epoch": 0.0035615834656185777, "grad_norm": 20.161680221557617, "learning_rate": 3.590909090909091e-05, "loss": 2.4902, "step": 396 }, { "epoch": 0.003570577363259029, "grad_norm": 12.120384216308594, "learning_rate": 3.6e-05, "loss": 2.4759, "step": 397 }, { "epoch": 0.0035795712608994797, "grad_norm": 9.16979694366455, "learning_rate": 3.6090909090909093e-05, "loss": 2.6742, "step": 398 }, { "epoch": 0.0035885651585399305, "grad_norm": 5.7415571212768555, "learning_rate": 3.6181818181818186e-05, "loss": 2.5058, "step": 399 }, { "epoch": 0.0035975590561803817, "grad_norm": 8.898907661437988, "learning_rate": 3.627272727272727e-05, "loss": 2.5235, "step": 400 }, { "epoch": 0.0036065529538208325, "grad_norm": 10.796667098999023, "learning_rate": 3.6363636363636364e-05, "loss": 2.4847, "step": 401 }, { "epoch": 0.0036155468514612837, "grad_norm": 13.13328742980957, "learning_rate": 3.645454545454546e-05, "loss": 2.6087, "step": 402 }, { "epoch": 0.0036245407491017345, "grad_norm": 5.896069526672363, "learning_rate": 3.654545454545455e-05, "loss": 2.5116, "step": 403 }, { "epoch": 0.0036335346467421853, "grad_norm": 12.354032516479492, "learning_rate": 3.663636363636364e-05, "loss": 2.4865, "step": 404 }, { "epoch": 0.0036425285443826365, "grad_norm": 9.572775840759277, "learning_rate": 3.672727272727273e-05, "loss": 2.5348, "step": 405 }, { "epoch": 0.0036515224420230873, "grad_norm": 18.182926177978516, "learning_rate": 3.681818181818182e-05, "loss": 2.641, "step": 406 }, { "epoch": 0.003660516339663538, "grad_norm": 16.372573852539062, "learning_rate": 3.690909090909091e-05, "loss": 2.4032, "step": 407 }, { "epoch": 0.0036695102373039894, "grad_norm": 6.369616985321045, "learning_rate": 3.7e-05, "loss": 2.5007, "step": 408 }, { "epoch": 0.00367850413494444, "grad_norm": 5.87434720993042, "learning_rate": 3.709090909090909e-05, "loss": 2.5089, "step": 409 }, { "epoch": 0.003687498032584891, "grad_norm": 11.947864532470703, "learning_rate": 3.718181818181818e-05, "loss": 2.5573, "step": 410 }, { "epoch": 0.003696491930225342, "grad_norm": 13.243257522583008, "learning_rate": 3.7272727272727276e-05, "loss": 2.4119, "step": 411 }, { "epoch": 0.003705485827865793, "grad_norm": 96.9281005859375, "learning_rate": 3.736363636363637e-05, "loss": 2.5608, "step": 412 }, { "epoch": 0.003714479725506244, "grad_norm": 15.962468147277832, "learning_rate": 3.745454545454546e-05, "loss": 2.5225, "step": 413 }, { "epoch": 0.003723473623146695, "grad_norm": 5.415879726409912, "learning_rate": 3.7545454545454546e-05, "loss": 2.4449, "step": 414 }, { "epoch": 0.0037324675207871458, "grad_norm": 24.821794509887695, "learning_rate": 3.763636363636364e-05, "loss": 2.5894, "step": 415 }, { "epoch": 0.003741461418427597, "grad_norm": 12.479686737060547, "learning_rate": 3.7727272727272725e-05, "loss": 2.6217, "step": 416 }, { "epoch": 0.003750455316068048, "grad_norm": 6.780008792877197, "learning_rate": 3.781818181818182e-05, "loss": 2.4162, "step": 417 }, { "epoch": 0.0037594492137084986, "grad_norm": 317.8321228027344, "learning_rate": 3.790909090909091e-05, "loss": 2.5253, "step": 418 }, { "epoch": 0.00376844311134895, "grad_norm": 30.4754581451416, "learning_rate": 3.8e-05, "loss": 2.6148, "step": 419 }, { "epoch": 0.0037774370089894006, "grad_norm": 16.09713363647461, "learning_rate": 3.8090909090909095e-05, "loss": 2.5931, "step": 420 }, { "epoch": 0.003786430906629852, "grad_norm": 11.552228927612305, "learning_rate": 3.818181818181819e-05, "loss": 2.5226, "step": 421 }, { "epoch": 0.0037954248042703026, "grad_norm": 20.038066864013672, "learning_rate": 3.827272727272728e-05, "loss": 2.4769, "step": 422 }, { "epoch": 0.0038044187019107534, "grad_norm": 10.651211738586426, "learning_rate": 3.8363636363636365e-05, "loss": 2.5161, "step": 423 }, { "epoch": 0.0038134125995512047, "grad_norm": 227.53170776367188, "learning_rate": 3.845454545454545e-05, "loss": 2.5614, "step": 424 }, { "epoch": 0.0038224064971916554, "grad_norm": 24.951213836669922, "learning_rate": 3.8545454545454544e-05, "loss": 2.5896, "step": 425 }, { "epoch": 0.0038314003948321062, "grad_norm": 13.660984992980957, "learning_rate": 3.8636363636363636e-05, "loss": 2.4454, "step": 426 }, { "epoch": 0.0038403942924725575, "grad_norm": 6.037498474121094, "learning_rate": 3.872727272727273e-05, "loss": 2.5347, "step": 427 }, { "epoch": 0.0038493881901130083, "grad_norm": 5.03461217880249, "learning_rate": 3.881818181818182e-05, "loss": 2.5223, "step": 428 }, { "epoch": 0.0038583820877534595, "grad_norm": 6.264018535614014, "learning_rate": 3.8909090909090914e-05, "loss": 2.518, "step": 429 }, { "epoch": 0.0038673759853939103, "grad_norm": 4.939131736755371, "learning_rate": 3.9000000000000006e-05, "loss": 2.3866, "step": 430 }, { "epoch": 0.003876369883034361, "grad_norm": 6.09281587600708, "learning_rate": 3.909090909090909e-05, "loss": 2.5, "step": 431 }, { "epoch": 0.0038853637806748123, "grad_norm": 8.432394027709961, "learning_rate": 3.9181818181818184e-05, "loss": 2.5482, "step": 432 }, { "epoch": 0.003894357678315263, "grad_norm": 7.908455848693848, "learning_rate": 3.927272727272727e-05, "loss": 2.39, "step": 433 }, { "epoch": 0.003903351575955714, "grad_norm": 5.349055767059326, "learning_rate": 3.936363636363636e-05, "loss": 2.5227, "step": 434 }, { "epoch": 0.003912345473596165, "grad_norm": 8.727982521057129, "learning_rate": 3.9454545454545455e-05, "loss": 2.5689, "step": 435 }, { "epoch": 0.003921339371236616, "grad_norm": 12.721184730529785, "learning_rate": 3.954545454545455e-05, "loss": 2.5216, "step": 436 }, { "epoch": 0.003930333268877067, "grad_norm": 10.756434440612793, "learning_rate": 3.963636363636364e-05, "loss": 2.5229, "step": 437 }, { "epoch": 0.003939327166517518, "grad_norm": 84.30569458007812, "learning_rate": 3.972727272727273e-05, "loss": 2.5331, "step": 438 }, { "epoch": 0.003948321064157969, "grad_norm": 15.8399658203125, "learning_rate": 3.981818181818182e-05, "loss": 2.4312, "step": 439 }, { "epoch": 0.0039573149617984195, "grad_norm": 8.966629028320312, "learning_rate": 3.990909090909091e-05, "loss": 2.5436, "step": 440 }, { "epoch": 0.003966308859438871, "grad_norm": 7.498310089111328, "learning_rate": 4e-05, "loss": 2.555, "step": 441 }, { "epoch": 0.003975302757079322, "grad_norm": 11.625467300415039, "learning_rate": 4.009090909090909e-05, "loss": 2.4952, "step": 442 }, { "epoch": 0.003984296654719772, "grad_norm": 7.1807661056518555, "learning_rate": 4.018181818181818e-05, "loss": 2.4346, "step": 443 }, { "epoch": 0.003993290552360224, "grad_norm": 185.90045166015625, "learning_rate": 4.0272727272727274e-05, "loss": 2.554, "step": 444 }, { "epoch": 0.004002284450000675, "grad_norm": 22.226383209228516, "learning_rate": 4.0363636363636367e-05, "loss": 2.4563, "step": 445 }, { "epoch": 0.004011278347641125, "grad_norm": 5.050864219665527, "learning_rate": 4.045454545454546e-05, "loss": 2.4988, "step": 446 }, { "epoch": 0.004020272245281576, "grad_norm": 5.204012870788574, "learning_rate": 4.0545454545454545e-05, "loss": 2.393, "step": 447 }, { "epoch": 0.004029266142922028, "grad_norm": 4.5905866622924805, "learning_rate": 4.063636363636364e-05, "loss": 2.4568, "step": 448 }, { "epoch": 0.004038260040562478, "grad_norm": 6.743342399597168, "learning_rate": 4.072727272727273e-05, "loss": 2.386, "step": 449 }, { "epoch": 0.004047253938202929, "grad_norm": 7.234485149383545, "learning_rate": 4.081818181818182e-05, "loss": 2.4349, "step": 450 }, { "epoch": 0.00405624783584338, "grad_norm": 11.250282287597656, "learning_rate": 4.0909090909090915e-05, "loss": 2.4458, "step": 451 }, { "epoch": 0.004065241733483832, "grad_norm": 6.4098286628723145, "learning_rate": 4.1e-05, "loss": 2.4528, "step": 452 }, { "epoch": 0.004074235631124282, "grad_norm": 5.250341415405273, "learning_rate": 4.109090909090909e-05, "loss": 2.5672, "step": 453 }, { "epoch": 0.004083229528764733, "grad_norm": 5.866156578063965, "learning_rate": 4.1181818181818185e-05, "loss": 2.5054, "step": 454 }, { "epoch": 0.0040922234264051845, "grad_norm": 6.161355495452881, "learning_rate": 4.127272727272727e-05, "loss": 2.442, "step": 455 }, { "epoch": 0.004101217324045635, "grad_norm": 4.930241107940674, "learning_rate": 4.1363636363636364e-05, "loss": 2.3989, "step": 456 }, { "epoch": 0.004110211221686086, "grad_norm": 6.114069938659668, "learning_rate": 4.1454545454545456e-05, "loss": 2.5388, "step": 457 }, { "epoch": 0.004119205119326537, "grad_norm": 5.344629764556885, "learning_rate": 4.154545454545455e-05, "loss": 2.4695, "step": 458 }, { "epoch": 0.004128199016966988, "grad_norm": 6.4735002517700195, "learning_rate": 4.163636363636364e-05, "loss": 2.4764, "step": 459 }, { "epoch": 0.004137192914607439, "grad_norm": 4.274179458618164, "learning_rate": 4.1727272727272734e-05, "loss": 2.4462, "step": 460 }, { "epoch": 0.00414618681224789, "grad_norm": 5.283795356750488, "learning_rate": 4.181818181818182e-05, "loss": 2.5101, "step": 461 }, { "epoch": 0.0041551807098883405, "grad_norm": 3.78226375579834, "learning_rate": 4.190909090909091e-05, "loss": 2.3994, "step": 462 }, { "epoch": 0.004164174607528792, "grad_norm": 3.4593284130096436, "learning_rate": 4.2e-05, "loss": 2.4868, "step": 463 }, { "epoch": 0.004173168505169243, "grad_norm": 3.445216178894043, "learning_rate": 4.209090909090909e-05, "loss": 2.435, "step": 464 }, { "epoch": 0.004182162402809693, "grad_norm": 11.3976411819458, "learning_rate": 4.218181818181818e-05, "loss": 2.361, "step": 465 }, { "epoch": 0.0041911563004501445, "grad_norm": 7.932605266571045, "learning_rate": 4.2272727272727275e-05, "loss": 2.3981, "step": 466 }, { "epoch": 0.004200150198090596, "grad_norm": 5.212550163269043, "learning_rate": 4.236363636363637e-05, "loss": 2.3614, "step": 467 }, { "epoch": 0.004209144095731046, "grad_norm": 4.533825397491455, "learning_rate": 4.245454545454546e-05, "loss": 2.3672, "step": 468 }, { "epoch": 0.004218137993371497, "grad_norm": 4.939641952514648, "learning_rate": 4.254545454545455e-05, "loss": 2.4213, "step": 469 }, { "epoch": 0.0042271318910119485, "grad_norm": 5.2856364250183105, "learning_rate": 4.263636363636364e-05, "loss": 2.486, "step": 470 }, { "epoch": 0.0042361257886524, "grad_norm": 2.2887861728668213, "learning_rate": 4.2727272727272724e-05, "loss": 2.4456, "step": 471 }, { "epoch": 0.00424511968629285, "grad_norm": 8.504525184631348, "learning_rate": 4.281818181818182e-05, "loss": 2.5319, "step": 472 }, { "epoch": 0.004254113583933301, "grad_norm": 5.444023132324219, "learning_rate": 4.290909090909091e-05, "loss": 2.4518, "step": 473 }, { "epoch": 0.004263107481573753, "grad_norm": 3.858370780944824, "learning_rate": 4.3e-05, "loss": 2.4155, "step": 474 }, { "epoch": 0.004272101379214203, "grad_norm": 6.466838836669922, "learning_rate": 4.3090909090909094e-05, "loss": 2.5241, "step": 475 }, { "epoch": 0.004281095276854654, "grad_norm": 5.948734760284424, "learning_rate": 4.318181818181819e-05, "loss": 2.5933, "step": 476 }, { "epoch": 0.004290089174495105, "grad_norm": 4.375283718109131, "learning_rate": 4.327272727272728e-05, "loss": 2.4722, "step": 477 }, { "epoch": 0.004299083072135556, "grad_norm": 5.097462177276611, "learning_rate": 4.3363636363636365e-05, "loss": 2.4599, "step": 478 }, { "epoch": 0.004308076969776007, "grad_norm": 6.355868339538574, "learning_rate": 4.345454545454546e-05, "loss": 2.4053, "step": 479 }, { "epoch": 0.004317070867416458, "grad_norm": 6.523740291595459, "learning_rate": 4.354545454545454e-05, "loss": 2.4362, "step": 480 }, { "epoch": 0.004326064765056909, "grad_norm": 11.158427238464355, "learning_rate": 4.3636363636363636e-05, "loss": 2.5399, "step": 481 }, { "epoch": 0.00433505866269736, "grad_norm": 6.568932056427002, "learning_rate": 4.372727272727273e-05, "loss": 2.5359, "step": 482 }, { "epoch": 0.004344052560337811, "grad_norm": 5.511691570281982, "learning_rate": 4.381818181818182e-05, "loss": 2.4101, "step": 483 }, { "epoch": 0.004353046457978261, "grad_norm": 4.5280890464782715, "learning_rate": 4.390909090909091e-05, "loss": 2.5112, "step": 484 }, { "epoch": 0.004362040355618713, "grad_norm": 2.9648256301879883, "learning_rate": 4.4000000000000006e-05, "loss": 2.3865, "step": 485 }, { "epoch": 0.004371034253259164, "grad_norm": 2.3086631298065186, "learning_rate": 4.409090909090909e-05, "loss": 2.3406, "step": 486 }, { "epoch": 0.004380028150899614, "grad_norm": 4.144700527191162, "learning_rate": 4.4181818181818184e-05, "loss": 2.3723, "step": 487 }, { "epoch": 0.004389022048540065, "grad_norm": 6.331112384796143, "learning_rate": 4.4272727272727276e-05, "loss": 2.5191, "step": 488 }, { "epoch": 0.004398015946180517, "grad_norm": 2.395456552505493, "learning_rate": 4.436363636363637e-05, "loss": 2.4398, "step": 489 }, { "epoch": 0.004407009843820968, "grad_norm": 11.222804069519043, "learning_rate": 4.4454545454545455e-05, "loss": 2.5287, "step": 490 }, { "epoch": 0.004416003741461418, "grad_norm": 7.065359115600586, "learning_rate": 4.454545454545455e-05, "loss": 2.3943, "step": 491 }, { "epoch": 0.0044249976391018695, "grad_norm": 2.2675955295562744, "learning_rate": 4.463636363636364e-05, "loss": 2.4372, "step": 492 }, { "epoch": 0.004433991536742321, "grad_norm": 4.432316303253174, "learning_rate": 4.472727272727273e-05, "loss": 2.5259, "step": 493 }, { "epoch": 0.004442985434382771, "grad_norm": 3.251943826675415, "learning_rate": 4.481818181818182e-05, "loss": 2.4694, "step": 494 }, { "epoch": 0.004451979332023222, "grad_norm": 5.685604095458984, "learning_rate": 4.490909090909091e-05, "loss": 2.5528, "step": 495 }, { "epoch": 0.0044609732296636735, "grad_norm": 5.714761734008789, "learning_rate": 4.5e-05, "loss": 2.5311, "step": 496 }, { "epoch": 0.004469967127304124, "grad_norm": 6.29110860824585, "learning_rate": 4.5090909090909095e-05, "loss": 2.4541, "step": 497 }, { "epoch": 0.004478961024944575, "grad_norm": 26.393896102905273, "learning_rate": 4.518181818181819e-05, "loss": 2.3905, "step": 498 }, { "epoch": 0.004487954922585026, "grad_norm": 29.4892635345459, "learning_rate": 4.5272727272727274e-05, "loss": 2.5434, "step": 499 }, { "epoch": 0.004496948820225477, "grad_norm": 9.63700008392334, "learning_rate": 4.5363636363636366e-05, "loss": 2.4066, "step": 500 }, { "epoch": 0.004505942717865928, "grad_norm": 9.156270027160645, "learning_rate": 4.545454545454546e-05, "loss": 2.4767, "step": 501 }, { "epoch": 0.004514936615506379, "grad_norm": 137.2318878173828, "learning_rate": 4.5545454545454544e-05, "loss": 2.4479, "step": 502 }, { "epoch": 0.0045239305131468295, "grad_norm": 195.78118896484375, "learning_rate": 4.563636363636364e-05, "loss": 2.6008, "step": 503 }, { "epoch": 0.004532924410787281, "grad_norm": 35.147743225097656, "learning_rate": 4.572727272727273e-05, "loss": 2.4295, "step": 504 }, { "epoch": 0.004541918308427732, "grad_norm": 5.983067989349365, "learning_rate": 4.581818181818182e-05, "loss": 2.4059, "step": 505 }, { "epoch": 0.004550912206068182, "grad_norm": 20.23373794555664, "learning_rate": 4.5909090909090914e-05, "loss": 2.44, "step": 506 }, { "epoch": 0.0045599061037086335, "grad_norm": 4.173520088195801, "learning_rate": 4.600000000000001e-05, "loss": 2.3349, "step": 507 }, { "epoch": 0.004568900001349085, "grad_norm": 9.3209228515625, "learning_rate": 4.609090909090909e-05, "loss": 2.4971, "step": 508 }, { "epoch": 0.004577893898989536, "grad_norm": 7.05019998550415, "learning_rate": 4.618181818181818e-05, "loss": 2.502, "step": 509 }, { "epoch": 0.004586887796629986, "grad_norm": 4.440962314605713, "learning_rate": 4.627272727272727e-05, "loss": 2.3926, "step": 510 }, { "epoch": 0.004595881694270438, "grad_norm": 4.18477201461792, "learning_rate": 4.636363636363636e-05, "loss": 2.5137, "step": 511 }, { "epoch": 0.004604875591910889, "grad_norm": 7.083902835845947, "learning_rate": 4.6454545454545456e-05, "loss": 2.4064, "step": 512 }, { "epoch": 0.004613869489551339, "grad_norm": 48.055904388427734, "learning_rate": 4.654545454545455e-05, "loss": 2.3294, "step": 513 }, { "epoch": 0.00462286338719179, "grad_norm": 5.847009181976318, "learning_rate": 4.663636363636364e-05, "loss": 2.3623, "step": 514 }, { "epoch": 0.004631857284832242, "grad_norm": 6.544424533843994, "learning_rate": 4.672727272727273e-05, "loss": 2.4563, "step": 515 }, { "epoch": 0.004640851182472692, "grad_norm": 3.1533761024475098, "learning_rate": 4.681818181818182e-05, "loss": 2.421, "step": 516 }, { "epoch": 0.004649845080113143, "grad_norm": 4.543294429779053, "learning_rate": 4.690909090909091e-05, "loss": 2.3646, "step": 517 }, { "epoch": 0.0046588389777535944, "grad_norm": 4.146971702575684, "learning_rate": 4.7e-05, "loss": 2.3599, "step": 518 }, { "epoch": 0.004667832875394045, "grad_norm": 2.5263845920562744, "learning_rate": 4.709090909090909e-05, "loss": 2.447, "step": 519 }, { "epoch": 0.004676826773034496, "grad_norm": 3.316958427429199, "learning_rate": 4.718181818181818e-05, "loss": 2.5234, "step": 520 }, { "epoch": 0.004685820670674947, "grad_norm": 3.836557626724243, "learning_rate": 4.7272727272727275e-05, "loss": 2.3482, "step": 521 }, { "epoch": 0.004694814568315398, "grad_norm": 4.416954517364502, "learning_rate": 4.736363636363637e-05, "loss": 2.4686, "step": 522 }, { "epoch": 0.004703808465955849, "grad_norm": 132.07891845703125, "learning_rate": 4.745454545454546e-05, "loss": 2.4657, "step": 523 }, { "epoch": 0.0047128023635963, "grad_norm": 759.2149047851562, "learning_rate": 4.7545454545454545e-05, "loss": 2.7605, "step": 524 }, { "epoch": 0.0047217962612367504, "grad_norm": 186.16273498535156, "learning_rate": 4.763636363636364e-05, "loss": 2.5167, "step": 525 }, { "epoch": 0.004730790158877202, "grad_norm": 21.49215316772461, "learning_rate": 4.772727272727273e-05, "loss": 2.4627, "step": 526 }, { "epoch": 0.004739784056517653, "grad_norm": 5.0005083084106445, "learning_rate": 4.781818181818182e-05, "loss": 2.4019, "step": 527 }, { "epoch": 0.004748777954158104, "grad_norm": 2.7786004543304443, "learning_rate": 4.790909090909091e-05, "loss": 2.4485, "step": 528 }, { "epoch": 0.0047577718517985545, "grad_norm": 2.6912715435028076, "learning_rate": 4.8e-05, "loss": 2.4206, "step": 529 }, { "epoch": 0.004766765749439006, "grad_norm": 3.564931631088257, "learning_rate": 4.8090909090909094e-05, "loss": 2.4458, "step": 530 }, { "epoch": 0.004775759647079457, "grad_norm": 4.527584075927734, "learning_rate": 4.8181818181818186e-05, "loss": 2.3245, "step": 531 }, { "epoch": 0.004784753544719907, "grad_norm": 10.07675838470459, "learning_rate": 4.827272727272727e-05, "loss": 2.4126, "step": 532 }, { "epoch": 0.0047937474423603585, "grad_norm": 6.41603422164917, "learning_rate": 4.8363636363636364e-05, "loss": 2.3938, "step": 533 }, { "epoch": 0.00480274134000081, "grad_norm": 3.609283685684204, "learning_rate": 4.845454545454546e-05, "loss": 2.3916, "step": 534 }, { "epoch": 0.00481173523764126, "grad_norm": 1.7395579814910889, "learning_rate": 4.854545454545455e-05, "loss": 2.5133, "step": 535 }, { "epoch": 0.004820729135281711, "grad_norm": 2.8066775798797607, "learning_rate": 4.863636363636364e-05, "loss": 2.4421, "step": 536 }, { "epoch": 0.004829723032922163, "grad_norm": 8.41027545928955, "learning_rate": 4.872727272727273e-05, "loss": 2.3096, "step": 537 }, { "epoch": 0.004838716930562613, "grad_norm": 5.382160663604736, "learning_rate": 4.881818181818182e-05, "loss": 2.4843, "step": 538 }, { "epoch": 0.004847710828203064, "grad_norm": 3.2162673473358154, "learning_rate": 4.890909090909091e-05, "loss": 2.4602, "step": 539 }, { "epoch": 0.004856704725843515, "grad_norm": 2.2050821781158447, "learning_rate": 4.9e-05, "loss": 2.4471, "step": 540 }, { "epoch": 0.004865698623483966, "grad_norm": 3.3169596195220947, "learning_rate": 4.909090909090909e-05, "loss": 2.5143, "step": 541 }, { "epoch": 0.004874692521124417, "grad_norm": 5.980579376220703, "learning_rate": 4.9181818181818183e-05, "loss": 2.5125, "step": 542 }, { "epoch": 0.004883686418764868, "grad_norm": 7.0032196044921875, "learning_rate": 4.9272727272727276e-05, "loss": 2.3135, "step": 543 }, { "epoch": 0.0048926803164053186, "grad_norm": 2.2724649906158447, "learning_rate": 4.936363636363637e-05, "loss": 2.4153, "step": 544 }, { "epoch": 0.00490167421404577, "grad_norm": 53.67974853515625, "learning_rate": 4.945454545454546e-05, "loss": 2.5726, "step": 545 }, { "epoch": 0.004910668111686221, "grad_norm": 4.723302364349365, "learning_rate": 4.9545454545454553e-05, "loss": 2.3774, "step": 546 }, { "epoch": 0.004919662009326672, "grad_norm": 2.6171138286590576, "learning_rate": 4.963636363636364e-05, "loss": 2.4464, "step": 547 }, { "epoch": 0.004928655906967123, "grad_norm": 2.6987946033477783, "learning_rate": 4.9727272727272725e-05, "loss": 2.466, "step": 548 }, { "epoch": 0.004937649804607574, "grad_norm": 10.407817840576172, "learning_rate": 4.981818181818182e-05, "loss": 2.4244, "step": 549 }, { "epoch": 0.004946643702248025, "grad_norm": 4.350247859954834, "learning_rate": 4.990909090909091e-05, "loss": 2.4616, "step": 550 }, { "epoch": 0.004955637599888475, "grad_norm": 6.2453107833862305, "learning_rate": 5e-05, "loss": 2.4426, "step": 551 }, { "epoch": 0.004964631497528927, "grad_norm": 37.29060745239258, "learning_rate": 4.9999998870263466e-05, "loss": 2.3924, "step": 552 }, { "epoch": 0.004973625395169378, "grad_norm": 74.37647247314453, "learning_rate": 4.999999548105396e-05, "loss": 2.4737, "step": 553 }, { "epoch": 0.004982619292809828, "grad_norm": 14.813892364501953, "learning_rate": 4.999998983237179e-05, "loss": 2.4317, "step": 554 }, { "epoch": 0.0049916131904502795, "grad_norm": 8.05070972442627, "learning_rate": 4.9999981924217466e-05, "loss": 2.5713, "step": 555 }, { "epoch": 0.005000607088090731, "grad_norm": 6.338493824005127, "learning_rate": 4.99999717565917e-05, "loss": 2.396, "step": 556 }, { "epoch": 0.005009600985731181, "grad_norm": 18.641860961914062, "learning_rate": 4.999995932949543e-05, "loss": 2.4908, "step": 557 }, { "epoch": 0.005018594883371632, "grad_norm": 82.1209945678711, "learning_rate": 4.999994464292975e-05, "loss": 2.516, "step": 558 }, { "epoch": 0.0050275887810120835, "grad_norm": 8.401792526245117, "learning_rate": 4.9999927696896e-05, "loss": 2.4075, "step": 559 }, { "epoch": 0.005036582678652534, "grad_norm": 3.3167531490325928, "learning_rate": 4.999990849139571e-05, "loss": 2.5289, "step": 560 }, { "epoch": 0.005045576576292985, "grad_norm": 3.5516357421875, "learning_rate": 4.9999887026430626e-05, "loss": 2.5607, "step": 561 }, { "epoch": 0.005054570473933436, "grad_norm": 5.763249397277832, "learning_rate": 4.999986330200268e-05, "loss": 2.3415, "step": 562 }, { "epoch": 0.0050635643715738875, "grad_norm": 5.255375385284424, "learning_rate": 4.999983731811402e-05, "loss": 2.4213, "step": 563 }, { "epoch": 0.005072558269214338, "grad_norm": 4.464290618896484, "learning_rate": 4.999980907476698e-05, "loss": 2.4283, "step": 564 }, { "epoch": 0.005081552166854789, "grad_norm": 4.298441410064697, "learning_rate": 4.999977857196413e-05, "loss": 2.3682, "step": 565 }, { "epoch": 0.00509054606449524, "grad_norm": 6.778446197509766, "learning_rate": 4.9999745809708215e-05, "loss": 2.3273, "step": 566 }, { "epoch": 0.005099539962135691, "grad_norm": 868.2053833007812, "learning_rate": 4.999971078800221e-05, "loss": 2.9579, "step": 567 }, { "epoch": 0.005108533859776142, "grad_norm": 1063.8876953125, "learning_rate": 4.9999673506849263e-05, "loss": 2.9642, "step": 568 }, { "epoch": 0.005117527757416593, "grad_norm": 12.19197940826416, "learning_rate": 4.999963396625276e-05, "loss": 2.4405, "step": 569 }, { "epoch": 0.0051265216550570435, "grad_norm": 24.950546264648438, "learning_rate": 4.9999592166216256e-05, "loss": 2.4331, "step": 570 }, { "epoch": 0.005135515552697495, "grad_norm": 943.664794921875, "learning_rate": 4.9999548106743545e-05, "loss": 3.0833, "step": 571 }, { "epoch": 0.005144509450337946, "grad_norm": 1022.8340454101562, "learning_rate": 4.99995017878386e-05, "loss": 3.0055, "step": 572 }, { "epoch": 0.005153503347978396, "grad_norm": 29.700937271118164, "learning_rate": 4.999945320950562e-05, "loss": 2.42, "step": 573 }, { "epoch": 0.005162497245618848, "grad_norm": 3.144054889678955, "learning_rate": 4.9999402371748984e-05, "loss": 2.3299, "step": 574 }, { "epoch": 0.005171491143259299, "grad_norm": 1.2397363185882568, "learning_rate": 4.999934927457328e-05, "loss": 2.4631, "step": 575 }, { "epoch": 0.005180485040899749, "grad_norm": 5.255294322967529, "learning_rate": 4.999929391798332e-05, "loss": 2.3657, "step": 576 }, { "epoch": 0.0051894789385402, "grad_norm": 5.393404006958008, "learning_rate": 4.9999236301984104e-05, "loss": 2.5346, "step": 577 }, { "epoch": 0.005198472836180652, "grad_norm": 3.3912456035614014, "learning_rate": 4.999917642658084e-05, "loss": 2.4842, "step": 578 }, { "epoch": 0.005207466733821102, "grad_norm": 1.8318792581558228, "learning_rate": 4.999911429177893e-05, "loss": 2.4364, "step": 579 }, { "epoch": 0.005216460631461553, "grad_norm": 26.505447387695312, "learning_rate": 4.9999049897584e-05, "loss": 2.313, "step": 580 }, { "epoch": 0.005225454529102004, "grad_norm": 3.3377368450164795, "learning_rate": 4.9998983244001874e-05, "loss": 2.3303, "step": 581 }, { "epoch": 0.005234448426742456, "grad_norm": 6.476102352142334, "learning_rate": 4.999891433103856e-05, "loss": 2.4122, "step": 582 }, { "epoch": 0.005243442324382906, "grad_norm": 5.718252182006836, "learning_rate": 4.99988431587003e-05, "loss": 2.4131, "step": 583 }, { "epoch": 0.005252436222023357, "grad_norm": 20.037609100341797, "learning_rate": 4.999876972699352e-05, "loss": 2.5064, "step": 584 }, { "epoch": 0.0052614301196638085, "grad_norm": 10.646677017211914, "learning_rate": 4.9998694035924866e-05, "loss": 2.4485, "step": 585 }, { "epoch": 0.005270424017304259, "grad_norm": 3.244764804840088, "learning_rate": 4.999861608550116e-05, "loss": 2.4147, "step": 586 }, { "epoch": 0.00527941791494471, "grad_norm": 2.5757675170898438, "learning_rate": 4.999853587572947e-05, "loss": 2.3996, "step": 587 }, { "epoch": 0.005288411812585161, "grad_norm": 2.946089744567871, "learning_rate": 4.9998453406617025e-05, "loss": 2.4282, "step": 588 }, { "epoch": 0.005297405710225612, "grad_norm": 4.174030303955078, "learning_rate": 4.999836867817129e-05, "loss": 2.5452, "step": 589 }, { "epoch": 0.005306399607866063, "grad_norm": 2.7353034019470215, "learning_rate": 4.999828169039992e-05, "loss": 2.4936, "step": 590 }, { "epoch": 0.005315393505506514, "grad_norm": 2.9000155925750732, "learning_rate": 4.9998192443310776e-05, "loss": 2.4056, "step": 591 }, { "epoch": 0.0053243874031469645, "grad_norm": 2.387249231338501, "learning_rate": 4.999810093691192e-05, "loss": 2.5121, "step": 592 }, { "epoch": 0.005333381300787416, "grad_norm": 1.5600703954696655, "learning_rate": 4.9998007171211634e-05, "loss": 2.2865, "step": 593 }, { "epoch": 0.005342375198427867, "grad_norm": 3.0678749084472656, "learning_rate": 4.999791114621839e-05, "loss": 2.3634, "step": 594 }, { "epoch": 0.005351369096068317, "grad_norm": 1.6179578304290771, "learning_rate": 4.999781286194085e-05, "loss": 2.3452, "step": 595 }, { "epoch": 0.0053603629937087685, "grad_norm": 2.9978394508361816, "learning_rate": 4.999771231838792e-05, "loss": 2.4297, "step": 596 }, { "epoch": 0.00536935689134922, "grad_norm": 2.1894114017486572, "learning_rate": 4.999760951556867e-05, "loss": 2.4439, "step": 597 }, { "epoch": 0.00537835078898967, "grad_norm": 1.9343936443328857, "learning_rate": 4.9997504453492395e-05, "loss": 2.3994, "step": 598 }, { "epoch": 0.005387344686630121, "grad_norm": 1.511374592781067, "learning_rate": 4.99973971321686e-05, "loss": 2.3782, "step": 599 }, { "epoch": 0.0053963385842705726, "grad_norm": 2.0215413570404053, "learning_rate": 4.999728755160697e-05, "loss": 2.3963, "step": 600 }, { "epoch": 0.005405332481911024, "grad_norm": 2.0300323963165283, "learning_rate": 4.999717571181742e-05, "loss": 2.381, "step": 601 }, { "epoch": 0.005414326379551474, "grad_norm": 1.4765416383743286, "learning_rate": 4.999706161281005e-05, "loss": 2.5344, "step": 602 }, { "epoch": 0.005423320277191925, "grad_norm": 1.7686407566070557, "learning_rate": 4.999694525459518e-05, "loss": 2.3815, "step": 603 }, { "epoch": 0.005432314174832377, "grad_norm": 2.907012939453125, "learning_rate": 4.9996826637183326e-05, "loss": 2.4307, "step": 604 }, { "epoch": 0.005441308072472827, "grad_norm": 1.7695300579071045, "learning_rate": 4.9996705760585194e-05, "loss": 2.4209, "step": 605 }, { "epoch": 0.005450301970113278, "grad_norm": 2.1694586277008057, "learning_rate": 4.9996582624811725e-05, "loss": 2.3793, "step": 606 }, { "epoch": 0.005459295867753729, "grad_norm": 2.7593190670013428, "learning_rate": 4.9996457229874046e-05, "loss": 2.4173, "step": 607 }, { "epoch": 0.00546828976539418, "grad_norm": 3.392613649368286, "learning_rate": 4.999632957578349e-05, "loss": 2.4865, "step": 608 }, { "epoch": 0.005477283663034631, "grad_norm": 4.626733303070068, "learning_rate": 4.999619966255158e-05, "loss": 2.4389, "step": 609 }, { "epoch": 0.005486277560675082, "grad_norm": 1.0661565065383911, "learning_rate": 4.9996067490190066e-05, "loss": 2.4327, "step": 610 }, { "epoch": 0.005495271458315533, "grad_norm": 1.619776725769043, "learning_rate": 4.999593305871091e-05, "loss": 2.4746, "step": 611 }, { "epoch": 0.005504265355955984, "grad_norm": 1.2350534200668335, "learning_rate": 4.9995796368126234e-05, "loss": 2.3905, "step": 612 }, { "epoch": 0.005513259253596435, "grad_norm": 2.012613296508789, "learning_rate": 4.999565741844841e-05, "loss": 2.4483, "step": 613 }, { "epoch": 0.005522253151236885, "grad_norm": 1.6660722494125366, "learning_rate": 4.9995516209689996e-05, "loss": 2.2968, "step": 614 }, { "epoch": 0.005531247048877337, "grad_norm": 1.593167781829834, "learning_rate": 4.999537274186375e-05, "loss": 2.4306, "step": 615 }, { "epoch": 0.005540240946517788, "grad_norm": 1.9116402864456177, "learning_rate": 4.999522701498263e-05, "loss": 2.371, "step": 616 }, { "epoch": 0.005549234844158238, "grad_norm": 1.4229307174682617, "learning_rate": 4.9995079029059824e-05, "loss": 2.3387, "step": 617 }, { "epoch": 0.0055582287417986894, "grad_norm": 2.0541815757751465, "learning_rate": 4.9994928784108694e-05, "loss": 2.3974, "step": 618 }, { "epoch": 0.005567222639439141, "grad_norm": 1.4975587129592896, "learning_rate": 4.999477628014282e-05, "loss": 2.3433, "step": 619 }, { "epoch": 0.005576216537079592, "grad_norm": 2.6742827892303467, "learning_rate": 4.999462151717599e-05, "loss": 2.3205, "step": 620 }, { "epoch": 0.005585210434720042, "grad_norm": 2.0126726627349854, "learning_rate": 4.9994464495222196e-05, "loss": 2.2967, "step": 621 }, { "epoch": 0.0055942043323604935, "grad_norm": 1.6105343103408813, "learning_rate": 4.999430521429561e-05, "loss": 2.4415, "step": 622 }, { "epoch": 0.005603198230000945, "grad_norm": 2.3501596450805664, "learning_rate": 4.9994143674410655e-05, "loss": 2.3561, "step": 623 }, { "epoch": 0.005612192127641395, "grad_norm": 2.1336519718170166, "learning_rate": 4.99939798755819e-05, "loss": 2.3675, "step": 624 }, { "epoch": 0.005621186025281846, "grad_norm": 2.012542247772217, "learning_rate": 4.999381381782417e-05, "loss": 2.4565, "step": 625 }, { "epoch": 0.0056301799229222975, "grad_norm": 1.5894619226455688, "learning_rate": 4.999364550115248e-05, "loss": 2.3818, "step": 626 }, { "epoch": 0.005639173820562748, "grad_norm": 1.5222396850585938, "learning_rate": 4.999347492558202e-05, "loss": 2.308, "step": 627 }, { "epoch": 0.005648167718203199, "grad_norm": 2.7180862426757812, "learning_rate": 4.999330209112822e-05, "loss": 2.5078, "step": 628 }, { "epoch": 0.00565716161584365, "grad_norm": 1.638213038444519, "learning_rate": 4.99931269978067e-05, "loss": 2.4567, "step": 629 }, { "epoch": 0.005666155513484101, "grad_norm": 2.0418717861175537, "learning_rate": 4.999294964563328e-05, "loss": 2.4698, "step": 630 }, { "epoch": 0.005675149411124552, "grad_norm": 2.8325579166412354, "learning_rate": 4.9992770034623994e-05, "loss": 2.273, "step": 631 }, { "epoch": 0.005684143308765003, "grad_norm": 2.4228262901306152, "learning_rate": 4.9992588164795064e-05, "loss": 2.3302, "step": 632 }, { "epoch": 0.0056931372064054535, "grad_norm": 2.3301501274108887, "learning_rate": 4.9992404036162944e-05, "loss": 2.3703, "step": 633 }, { "epoch": 0.005702131104045905, "grad_norm": 2.912381172180176, "learning_rate": 4.9992217648744254e-05, "loss": 2.4629, "step": 634 }, { "epoch": 0.005711125001686356, "grad_norm": 2.8964321613311768, "learning_rate": 4.999202900255586e-05, "loss": 2.5102, "step": 635 }, { "epoch": 0.005720118899326806, "grad_norm": 2.1238880157470703, "learning_rate": 4.999183809761481e-05, "loss": 2.3765, "step": 636 }, { "epoch": 0.0057291127969672576, "grad_norm": 1.820768117904663, "learning_rate": 4.999164493393835e-05, "loss": 2.4077, "step": 637 }, { "epoch": 0.005738106694607709, "grad_norm": 1.8464839458465576, "learning_rate": 4.999144951154394e-05, "loss": 2.4221, "step": 638 }, { "epoch": 0.00574710059224816, "grad_norm": 1.9260097742080688, "learning_rate": 4.999125183044924e-05, "loss": 2.2966, "step": 639 }, { "epoch": 0.00575609448988861, "grad_norm": 2.3923349380493164, "learning_rate": 4.9991051890672114e-05, "loss": 2.3924, "step": 640 }, { "epoch": 0.005765088387529062, "grad_norm": 1.808197021484375, "learning_rate": 4.999084969223063e-05, "loss": 2.4124, "step": 641 }, { "epoch": 0.005774082285169513, "grad_norm": 2.1454949378967285, "learning_rate": 4.999064523514309e-05, "loss": 2.4176, "step": 642 }, { "epoch": 0.005783076182809963, "grad_norm": 1.7350833415985107, "learning_rate": 4.999043851942793e-05, "loss": 2.3737, "step": 643 }, { "epoch": 0.005792070080450414, "grad_norm": 2.89910888671875, "learning_rate": 4.999022954510386e-05, "loss": 2.5145, "step": 644 }, { "epoch": 0.005801063978090866, "grad_norm": 2.385995388031006, "learning_rate": 4.999001831218977e-05, "loss": 2.3942, "step": 645 }, { "epoch": 0.005810057875731316, "grad_norm": 38.989044189453125, "learning_rate": 4.9989804820704735e-05, "loss": 2.3923, "step": 646 }, { "epoch": 0.005819051773371767, "grad_norm": 105.62820434570312, "learning_rate": 4.998958907066805e-05, "loss": 2.4966, "step": 647 }, { "epoch": 0.0058280456710122185, "grad_norm": 10.19742202758789, "learning_rate": 4.998937106209923e-05, "loss": 2.3541, "step": 648 }, { "epoch": 0.005837039568652669, "grad_norm": 2.646786689758301, "learning_rate": 4.998915079501798e-05, "loss": 2.3183, "step": 649 }, { "epoch": 0.00584603346629312, "grad_norm": 2.7256410121917725, "learning_rate": 4.998892826944418e-05, "loss": 2.5195, "step": 650 }, { "epoch": 0.005855027363933571, "grad_norm": 1.814576268196106, "learning_rate": 4.998870348539797e-05, "loss": 2.4928, "step": 651 }, { "epoch": 0.005864021261574022, "grad_norm": 65.62152862548828, "learning_rate": 4.998847644289965e-05, "loss": 2.346, "step": 652 }, { "epoch": 0.005873015159214473, "grad_norm": 156.02301025390625, "learning_rate": 4.998824714196974e-05, "loss": 2.5314, "step": 653 }, { "epoch": 0.005882009056854924, "grad_norm": 18.26453971862793, "learning_rate": 4.9988015582628974e-05, "loss": 2.4169, "step": 654 }, { "epoch": 0.005891002954495375, "grad_norm": 1.6023188829421997, "learning_rate": 4.998778176489827e-05, "loss": 2.3979, "step": 655 }, { "epoch": 0.005899996852135826, "grad_norm": 3.3964672088623047, "learning_rate": 4.998754568879876e-05, "loss": 2.3552, "step": 656 }, { "epoch": 0.005908990749776277, "grad_norm": 318.99041748046875, "learning_rate": 4.9987307354351796e-05, "loss": 2.4107, "step": 657 }, { "epoch": 0.005917984647416728, "grad_norm": 1.386103868484497, "learning_rate": 4.998706676157891e-05, "loss": 2.3398, "step": 658 }, { "epoch": 0.0059269785450571785, "grad_norm": 1.273573875427246, "learning_rate": 4.9986823910501834e-05, "loss": 2.3497, "step": 659 }, { "epoch": 0.00593597244269763, "grad_norm": 1.5252665281295776, "learning_rate": 4.998657880114254e-05, "loss": 2.3604, "step": 660 }, { "epoch": 0.005944966340338081, "grad_norm": 7.399944305419922, "learning_rate": 4.9986331433523156e-05, "loss": 2.3463, "step": 661 }, { "epoch": 0.005953960237978531, "grad_norm": 348.9754943847656, "learning_rate": 4.9986081807666055e-05, "loss": 2.5269, "step": 662 }, { "epoch": 0.0059629541356189825, "grad_norm": 1212.12109375, "learning_rate": 4.9985829923593786e-05, "loss": 2.9826, "step": 663 }, { "epoch": 0.005971948033259434, "grad_norm": 250.246826171875, "learning_rate": 4.998557578132912e-05, "loss": 2.4467, "step": 664 }, { "epoch": 0.005980941930899884, "grad_norm": 2.0885791778564453, "learning_rate": 4.998531938089503e-05, "loss": 2.4626, "step": 665 }, { "epoch": 0.005989935828540335, "grad_norm": 1.5140365362167358, "learning_rate": 4.998506072231469e-05, "loss": 2.3628, "step": 666 }, { "epoch": 0.005998929726180787, "grad_norm": 1.3232437372207642, "learning_rate": 4.998479980561148e-05, "loss": 2.3055, "step": 667 }, { "epoch": 0.006007923623821237, "grad_norm": 1.0808279514312744, "learning_rate": 4.9984536630808964e-05, "loss": 2.3661, "step": 668 }, { "epoch": 0.006016917521461688, "grad_norm": 1.9327250719070435, "learning_rate": 4.998427119793094e-05, "loss": 2.3704, "step": 669 }, { "epoch": 0.006025911419102139, "grad_norm": 4.848217964172363, "learning_rate": 4.99840035070014e-05, "loss": 2.4662, "step": 670 }, { "epoch": 0.00603490531674259, "grad_norm": 4.094484806060791, "learning_rate": 4.9983733558044534e-05, "loss": 2.4309, "step": 671 }, { "epoch": 0.006043899214383041, "grad_norm": 2.386913537979126, "learning_rate": 4.9983461351084735e-05, "loss": 2.3768, "step": 672 }, { "epoch": 0.006052893112023492, "grad_norm": 2.4711503982543945, "learning_rate": 4.9983186886146606e-05, "loss": 2.4355, "step": 673 }, { "epoch": 0.0060618870096639434, "grad_norm": 1.4953632354736328, "learning_rate": 4.998291016325496e-05, "loss": 2.2473, "step": 674 }, { "epoch": 0.006070880907304394, "grad_norm": 3.127026081085205, "learning_rate": 4.99826311824348e-05, "loss": 2.3246, "step": 675 }, { "epoch": 0.006079874804944845, "grad_norm": 3.4192206859588623, "learning_rate": 4.998234994371135e-05, "loss": 2.3357, "step": 676 }, { "epoch": 0.006088868702585296, "grad_norm": 1.615310788154602, "learning_rate": 4.998206644711002e-05, "loss": 2.3901, "step": 677 }, { "epoch": 0.006097862600225747, "grad_norm": 1.7600175142288208, "learning_rate": 4.998178069265643e-05, "loss": 2.3917, "step": 678 }, { "epoch": 0.006106856497866198, "grad_norm": 1.5513185262680054, "learning_rate": 4.998149268037641e-05, "loss": 2.3612, "step": 679 }, { "epoch": 0.006115850395506649, "grad_norm": 1.5127142667770386, "learning_rate": 4.9981202410295984e-05, "loss": 2.3874, "step": 680 }, { "epoch": 0.006124844293147099, "grad_norm": 4.144172668457031, "learning_rate": 4.99809098824414e-05, "loss": 2.4896, "step": 681 }, { "epoch": 0.006133838190787551, "grad_norm": 2.2086148262023926, "learning_rate": 4.9980615096839076e-05, "loss": 2.418, "step": 682 }, { "epoch": 0.006142832088428002, "grad_norm": 19.644895553588867, "learning_rate": 4.9980318053515686e-05, "loss": 2.5861, "step": 683 }, { "epoch": 0.006151825986068452, "grad_norm": 5.292164325714111, "learning_rate": 4.9980018752498035e-05, "loss": 2.3599, "step": 684 }, { "epoch": 0.0061608198837089035, "grad_norm": 59.39234161376953, "learning_rate": 4.997971719381321e-05, "loss": 2.4926, "step": 685 }, { "epoch": 0.006169813781349355, "grad_norm": 164.944580078125, "learning_rate": 4.9979413377488446e-05, "loss": 2.426, "step": 686 }, { "epoch": 0.006178807678989805, "grad_norm": 2.1656153202056885, "learning_rate": 4.997910730355121e-05, "loss": 2.3332, "step": 687 }, { "epoch": 0.006187801576630256, "grad_norm": 361.9581604003906, "learning_rate": 4.997879897202916e-05, "loss": 2.4859, "step": 688 }, { "epoch": 0.0061967954742707075, "grad_norm": 81.85458374023438, "learning_rate": 4.9978488382950165e-05, "loss": 2.4685, "step": 689 }, { "epoch": 0.006205789371911158, "grad_norm": 5.230372428894043, "learning_rate": 4.9978175536342286e-05, "loss": 2.419, "step": 690 }, { "epoch": 0.006214783269551609, "grad_norm": 2.3871922492980957, "learning_rate": 4.997786043223382e-05, "loss": 2.3261, "step": 691 }, { "epoch": 0.00622377716719206, "grad_norm": 7.315066814422607, "learning_rate": 4.997754307065322e-05, "loss": 2.4856, "step": 692 }, { "epoch": 0.0062327710648325116, "grad_norm": 4.792977333068848, "learning_rate": 4.997722345162918e-05, "loss": 2.373, "step": 693 }, { "epoch": 0.006241764962472962, "grad_norm": 4.047577381134033, "learning_rate": 4.997690157519059e-05, "loss": 2.3665, "step": 694 }, { "epoch": 0.006250758860113413, "grad_norm": 1.6489315032958984, "learning_rate": 4.997657744136654e-05, "loss": 2.2563, "step": 695 }, { "epoch": 0.006259752757753864, "grad_norm": 1.2394626140594482, "learning_rate": 4.997625105018633e-05, "loss": 2.2744, "step": 696 }, { "epoch": 0.006268746655394315, "grad_norm": 1.6913032531738281, "learning_rate": 4.997592240167945e-05, "loss": 2.3004, "step": 697 }, { "epoch": 0.006277740553034766, "grad_norm": 1.1033881902694702, "learning_rate": 4.99755914958756e-05, "loss": 2.3581, "step": 698 }, { "epoch": 0.006286734450675217, "grad_norm": 1.4236403703689575, "learning_rate": 4.99752583328047e-05, "loss": 2.3757, "step": 699 }, { "epoch": 0.0062957283483156675, "grad_norm": 1.4072418212890625, "learning_rate": 4.9974922912496846e-05, "loss": 2.367, "step": 700 }, { "epoch": 0.006304722245956119, "grad_norm": 1.4323331117630005, "learning_rate": 4.997458523498236e-05, "loss": 2.368, "step": 701 }, { "epoch": 0.00631371614359657, "grad_norm": 3.138033390045166, "learning_rate": 4.9974245300291763e-05, "loss": 2.4071, "step": 702 }, { "epoch": 0.00632271004123702, "grad_norm": 1.720033049583435, "learning_rate": 4.997390310845578e-05, "loss": 2.4112, "step": 703 }, { "epoch": 0.006331703938877472, "grad_norm": 1.968177318572998, "learning_rate": 4.9973558659505325e-05, "loss": 2.3203, "step": 704 }, { "epoch": 0.006340697836517923, "grad_norm": 2.108110189437866, "learning_rate": 4.997321195347154e-05, "loss": 2.4735, "step": 705 }, { "epoch": 0.006349691734158373, "grad_norm": 1.26445734500885, "learning_rate": 4.997286299038576e-05, "loss": 2.527, "step": 706 }, { "epoch": 0.006358685631798824, "grad_norm": 2.095703601837158, "learning_rate": 4.997251177027952e-05, "loss": 2.4369, "step": 707 }, { "epoch": 0.006367679529439276, "grad_norm": 1.7823727130889893, "learning_rate": 4.9972158293184565e-05, "loss": 2.3546, "step": 708 }, { "epoch": 0.006376673427079726, "grad_norm": 1.430592656135559, "learning_rate": 4.997180255913284e-05, "loss": 2.4506, "step": 709 }, { "epoch": 0.006385667324720177, "grad_norm": 2.089142084121704, "learning_rate": 4.99714445681565e-05, "loss": 2.3887, "step": 710 }, { "epoch": 0.0063946612223606284, "grad_norm": 1.1167904138565063, "learning_rate": 4.99710843202879e-05, "loss": 2.3572, "step": 711 }, { "epoch": 0.00640365512000108, "grad_norm": 1.7042675018310547, "learning_rate": 4.9970721815559585e-05, "loss": 2.3336, "step": 712 }, { "epoch": 0.00641264901764153, "grad_norm": 1.5516256093978882, "learning_rate": 4.9970357054004335e-05, "loss": 2.3428, "step": 713 }, { "epoch": 0.006421642915281981, "grad_norm": 0.7995926737785339, "learning_rate": 4.996999003565511e-05, "loss": 2.3922, "step": 714 }, { "epoch": 0.0064306368129224325, "grad_norm": 1.3005210161209106, "learning_rate": 4.9969620760545074e-05, "loss": 2.3788, "step": 715 }, { "epoch": 0.006439630710562883, "grad_norm": 1.9167686700820923, "learning_rate": 4.996924922870762e-05, "loss": 2.4571, "step": 716 }, { "epoch": 0.006448624608203334, "grad_norm": 1.8720837831497192, "learning_rate": 4.996887544017631e-05, "loss": 2.3795, "step": 717 }, { "epoch": 0.006457618505843785, "grad_norm": 1.3605936765670776, "learning_rate": 4.996849939498493e-05, "loss": 2.3193, "step": 718 }, { "epoch": 0.006466612403484236, "grad_norm": 1.6093920469284058, "learning_rate": 4.9968121093167476e-05, "loss": 2.4547, "step": 719 }, { "epoch": 0.006475606301124687, "grad_norm": 1.8580682277679443, "learning_rate": 4.9967740534758126e-05, "loss": 2.3832, "step": 720 }, { "epoch": 0.006484600198765138, "grad_norm": 1.1695489883422852, "learning_rate": 4.9967357719791286e-05, "loss": 2.3307, "step": 721 }, { "epoch": 0.0064935940964055885, "grad_norm": 1.8113250732421875, "learning_rate": 4.9966972648301536e-05, "loss": 2.4475, "step": 722 }, { "epoch": 0.00650258799404604, "grad_norm": 1.3906854391098022, "learning_rate": 4.99665853203237e-05, "loss": 2.3736, "step": 723 }, { "epoch": 0.006511581891686491, "grad_norm": 1.8299096822738647, "learning_rate": 4.996619573589277e-05, "loss": 2.3624, "step": 724 }, { "epoch": 0.006520575789326941, "grad_norm": 2.2322638034820557, "learning_rate": 4.9965803895043964e-05, "loss": 2.4239, "step": 725 }, { "epoch": 0.0065295696869673925, "grad_norm": 1.2125953435897827, "learning_rate": 4.996540979781269e-05, "loss": 2.3379, "step": 726 }, { "epoch": 0.006538563584607844, "grad_norm": 2.416262149810791, "learning_rate": 4.996501344423456e-05, "loss": 2.4828, "step": 727 }, { "epoch": 0.006547557482248294, "grad_norm": 1.60369873046875, "learning_rate": 4.996461483434542e-05, "loss": 2.4416, "step": 728 }, { "epoch": 0.006556551379888745, "grad_norm": 1.5307985544204712, "learning_rate": 4.996421396818127e-05, "loss": 2.4244, "step": 729 }, { "epoch": 0.0065655452775291966, "grad_norm": 0.987632691860199, "learning_rate": 4.996381084577836e-05, "loss": 2.3704, "step": 730 }, { "epoch": 0.006574539175169648, "grad_norm": 0.9207761883735657, "learning_rate": 4.996340546717311e-05, "loss": 2.3548, "step": 731 }, { "epoch": 0.006583533072810098, "grad_norm": 1.4703799486160278, "learning_rate": 4.996299783240217e-05, "loss": 2.5025, "step": 732 }, { "epoch": 0.006592526970450549, "grad_norm": 1.1804099082946777, "learning_rate": 4.996258794150236e-05, "loss": 2.4147, "step": 733 }, { "epoch": 0.006601520868091001, "grad_norm": 2.1551332473754883, "learning_rate": 4.9962175794510744e-05, "loss": 2.4238, "step": 734 }, { "epoch": 0.006610514765731451, "grad_norm": 1.1678895950317383, "learning_rate": 4.996176139146457e-05, "loss": 2.4213, "step": 735 }, { "epoch": 0.006619508663371902, "grad_norm": 2.0252041816711426, "learning_rate": 4.9961344732401294e-05, "loss": 2.4356, "step": 736 }, { "epoch": 0.006628502561012353, "grad_norm": 0.9819325804710388, "learning_rate": 4.9960925817358564e-05, "loss": 2.3012, "step": 737 }, { "epoch": 0.006637496458652804, "grad_norm": 1.136311650276184, "learning_rate": 4.996050464637423e-05, "loss": 2.3593, "step": 738 }, { "epoch": 0.006646490356293255, "grad_norm": 0.9213350415229797, "learning_rate": 4.996008121948639e-05, "loss": 2.3494, "step": 739 }, { "epoch": 0.006655484253933706, "grad_norm": 1.3651710748672485, "learning_rate": 4.995965553673329e-05, "loss": 2.5314, "step": 740 }, { "epoch": 0.006664478151574157, "grad_norm": 1.3020514249801636, "learning_rate": 4.995922759815339e-05, "loss": 2.2532, "step": 741 }, { "epoch": 0.006673472049214608, "grad_norm": 0.775888204574585, "learning_rate": 4.99587974037854e-05, "loss": 2.3791, "step": 742 }, { "epoch": 0.006682465946855059, "grad_norm": 1.7162033319473267, "learning_rate": 4.995836495366818e-05, "loss": 2.3848, "step": 743 }, { "epoch": 0.006691459844495509, "grad_norm": 1.0520635843276978, "learning_rate": 4.9957930247840824e-05, "loss": 2.3923, "step": 744 }, { "epoch": 0.006700453742135961, "grad_norm": 1.6035592555999756, "learning_rate": 4.99574932863426e-05, "loss": 2.5291, "step": 745 }, { "epoch": 0.006709447639776412, "grad_norm": 1.0390980243682861, "learning_rate": 4.9957054069213025e-05, "loss": 2.3656, "step": 746 }, { "epoch": 0.006718441537416863, "grad_norm": 1.1373931169509888, "learning_rate": 4.995661259649179e-05, "loss": 2.3262, "step": 747 }, { "epoch": 0.0067274354350573135, "grad_norm": 1.5664631128311157, "learning_rate": 4.9956168868218787e-05, "loss": 2.3749, "step": 748 }, { "epoch": 0.006736429332697765, "grad_norm": 1.1574504375457764, "learning_rate": 4.9955722884434114e-05, "loss": 2.3114, "step": 749 }, { "epoch": 0.006745423230338216, "grad_norm": 0.8074120283126831, "learning_rate": 4.99552746451781e-05, "loss": 2.4724, "step": 750 }, { "epoch": 0.006754417127978666, "grad_norm": 0.8828337788581848, "learning_rate": 4.995482415049123e-05, "loss": 2.3239, "step": 751 }, { "epoch": 0.0067634110256191175, "grad_norm": 1.0525522232055664, "learning_rate": 4.995437140041424e-05, "loss": 2.3863, "step": 752 }, { "epoch": 0.006772404923259569, "grad_norm": 2.016676902770996, "learning_rate": 4.995391639498803e-05, "loss": 2.402, "step": 753 }, { "epoch": 0.006781398820900019, "grad_norm": 1.6840852499008179, "learning_rate": 4.995345913425374e-05, "loss": 2.3384, "step": 754 }, { "epoch": 0.00679039271854047, "grad_norm": 1.2580803632736206, "learning_rate": 4.99529996182527e-05, "loss": 2.3912, "step": 755 }, { "epoch": 0.0067993866161809215, "grad_norm": 2.412668466567993, "learning_rate": 4.995253784702642e-05, "loss": 2.3987, "step": 756 }, { "epoch": 0.006808380513821372, "grad_norm": 2.4408586025238037, "learning_rate": 4.995207382061666e-05, "loss": 2.3739, "step": 757 }, { "epoch": 0.006817374411461823, "grad_norm": 2.665259599685669, "learning_rate": 4.995160753906534e-05, "loss": 2.5046, "step": 758 }, { "epoch": 0.006826368309102274, "grad_norm": 2.536120653152466, "learning_rate": 4.9951139002414596e-05, "loss": 2.4292, "step": 759 }, { "epoch": 0.006835362206742725, "grad_norm": 1.1645842790603638, "learning_rate": 4.995066821070679e-05, "loss": 2.373, "step": 760 }, { "epoch": 0.006844356104383176, "grad_norm": 2.160353899002075, "learning_rate": 4.995019516398447e-05, "loss": 2.5006, "step": 761 }, { "epoch": 0.006853350002023627, "grad_norm": 2.830785036087036, "learning_rate": 4.994971986229038e-05, "loss": 2.471, "step": 762 }, { "epoch": 0.0068623438996640775, "grad_norm": 2.361668825149536, "learning_rate": 4.9949242305667486e-05, "loss": 2.3206, "step": 763 }, { "epoch": 0.006871337797304529, "grad_norm": 1.6988106966018677, "learning_rate": 4.994876249415894e-05, "loss": 2.4014, "step": 764 }, { "epoch": 0.00688033169494498, "grad_norm": 0.9352366924285889, "learning_rate": 4.9948280427808115e-05, "loss": 2.402, "step": 765 }, { "epoch": 0.006889325592585431, "grad_norm": 1.5733873844146729, "learning_rate": 4.994779610665858e-05, "loss": 2.421, "step": 766 }, { "epoch": 0.006898319490225882, "grad_norm": 1.1752417087554932, "learning_rate": 4.9947309530754104e-05, "loss": 2.2774, "step": 767 }, { "epoch": 0.006907313387866333, "grad_norm": 1.1936308145523071, "learning_rate": 4.9946820700138666e-05, "loss": 2.2928, "step": 768 }, { "epoch": 0.006916307285506784, "grad_norm": 1.9820133447647095, "learning_rate": 4.994632961485644e-05, "loss": 2.3744, "step": 769 }, { "epoch": 0.006925301183147234, "grad_norm": 2.30741810798645, "learning_rate": 4.9945836274951805e-05, "loss": 2.3732, "step": 770 }, { "epoch": 0.006934295080787686, "grad_norm": 2.1270976066589355, "learning_rate": 4.994534068046937e-05, "loss": 2.3706, "step": 771 }, { "epoch": 0.006943288978428137, "grad_norm": 1.8759336471557617, "learning_rate": 4.994484283145391e-05, "loss": 2.4928, "step": 772 }, { "epoch": 0.006952282876068587, "grad_norm": 1.6399668455123901, "learning_rate": 4.994434272795042e-05, "loss": 2.4569, "step": 773 }, { "epoch": 0.006961276773709038, "grad_norm": 1.5422440767288208, "learning_rate": 4.99438403700041e-05, "loss": 2.4189, "step": 774 }, { "epoch": 0.00697027067134949, "grad_norm": 2.132363796234131, "learning_rate": 4.994333575766036e-05, "loss": 2.4798, "step": 775 }, { "epoch": 0.00697926456898994, "grad_norm": 1.4157288074493408, "learning_rate": 4.99428288909648e-05, "loss": 2.4164, "step": 776 }, { "epoch": 0.006988258466630391, "grad_norm": 1.1873482465744019, "learning_rate": 4.994231976996323e-05, "loss": 2.413, "step": 777 }, { "epoch": 0.0069972523642708425, "grad_norm": 0.9601787328720093, "learning_rate": 4.994180839470166e-05, "loss": 2.3641, "step": 778 }, { "epoch": 0.007006246261911293, "grad_norm": 1.6519640684127808, "learning_rate": 4.994129476522632e-05, "loss": 2.3609, "step": 779 }, { "epoch": 0.007015240159551744, "grad_norm": 1.522507667541504, "learning_rate": 4.994077888158362e-05, "loss": 2.2939, "step": 780 }, { "epoch": 0.007024234057192195, "grad_norm": 1.2533833980560303, "learning_rate": 4.994026074382019e-05, "loss": 2.3539, "step": 781 }, { "epoch": 0.007033227954832646, "grad_norm": 1.0521180629730225, "learning_rate": 4.9939740351982856e-05, "loss": 2.3217, "step": 782 }, { "epoch": 0.007042221852473097, "grad_norm": 1.5616939067840576, "learning_rate": 4.993921770611866e-05, "loss": 2.3804, "step": 783 }, { "epoch": 0.007051215750113548, "grad_norm": 0.9750461578369141, "learning_rate": 4.9938692806274814e-05, "loss": 2.328, "step": 784 }, { "epoch": 0.007060209647753999, "grad_norm": 1.0921258926391602, "learning_rate": 4.9938165652498787e-05, "loss": 2.2191, "step": 785 }, { "epoch": 0.00706920354539445, "grad_norm": 1.053160548210144, "learning_rate": 4.993763624483821e-05, "loss": 2.3964, "step": 786 }, { "epoch": 0.007078197443034901, "grad_norm": 2.779226303100586, "learning_rate": 4.993710458334092e-05, "loss": 2.4296, "step": 787 }, { "epoch": 0.007087191340675352, "grad_norm": 1.360906720161438, "learning_rate": 4.9936570668054984e-05, "loss": 2.2808, "step": 788 }, { "epoch": 0.0070961852383158025, "grad_norm": 1.4774829149246216, "learning_rate": 4.993603449902865e-05, "loss": 2.2954, "step": 789 }, { "epoch": 0.007105179135956254, "grad_norm": 1.9357240200042725, "learning_rate": 4.9935496076310376e-05, "loss": 2.296, "step": 790 }, { "epoch": 0.007114173033596705, "grad_norm": 2.325702667236328, "learning_rate": 4.9934955399948824e-05, "loss": 2.339, "step": 791 }, { "epoch": 0.007123166931237155, "grad_norm": 1.0266544818878174, "learning_rate": 4.993441246999287e-05, "loss": 2.3156, "step": 792 }, { "epoch": 0.0071321608288776065, "grad_norm": 1.1733542680740356, "learning_rate": 4.993386728649156e-05, "loss": 2.3455, "step": 793 }, { "epoch": 0.007141154726518058, "grad_norm": 1.7009702920913696, "learning_rate": 4.9933319849494185e-05, "loss": 2.3087, "step": 794 }, { "epoch": 0.007150148624158508, "grad_norm": 1.306169867515564, "learning_rate": 4.9932770159050225e-05, "loss": 2.3605, "step": 795 }, { "epoch": 0.007159142521798959, "grad_norm": 2.564422845840454, "learning_rate": 4.9932218215209346e-05, "loss": 2.4546, "step": 796 }, { "epoch": 0.007168136419439411, "grad_norm": 1.0364539623260498, "learning_rate": 4.993166401802144e-05, "loss": 2.4305, "step": 797 }, { "epoch": 0.007177130317079861, "grad_norm": 1.7015596628189087, "learning_rate": 4.99311075675366e-05, "loss": 2.3859, "step": 798 }, { "epoch": 0.007186124214720312, "grad_norm": 1.8677973747253418, "learning_rate": 4.9930548863805105e-05, "loss": 2.466, "step": 799 }, { "epoch": 0.007195118112360763, "grad_norm": 1.9033832550048828, "learning_rate": 4.992998790687746e-05, "loss": 2.3308, "step": 800 }, { "epoch": 0.007204112010001214, "grad_norm": 0.9066283702850342, "learning_rate": 4.992942469680436e-05, "loss": 2.282, "step": 801 }, { "epoch": 0.007213105907641665, "grad_norm": 1.59651517868042, "learning_rate": 4.99288592336367e-05, "loss": 2.4003, "step": 802 }, { "epoch": 0.007222099805282116, "grad_norm": 1.3170840740203857, "learning_rate": 4.992829151742561e-05, "loss": 2.3541, "step": 803 }, { "epoch": 0.0072310937029225674, "grad_norm": 1.7328461408615112, "learning_rate": 4.9927721548222374e-05, "loss": 2.3629, "step": 804 }, { "epoch": 0.007240087600563018, "grad_norm": 1.3761106729507446, "learning_rate": 4.992714932607851e-05, "loss": 2.3344, "step": 805 }, { "epoch": 0.007249081498203469, "grad_norm": 1.4148802757263184, "learning_rate": 4.992657485104575e-05, "loss": 2.2926, "step": 806 }, { "epoch": 0.00725807539584392, "grad_norm": 1.0844882726669312, "learning_rate": 4.9925998123175996e-05, "loss": 2.4877, "step": 807 }, { "epoch": 0.007267069293484371, "grad_norm": 2.088745355606079, "learning_rate": 4.992541914252138e-05, "loss": 2.3863, "step": 808 }, { "epoch": 0.007276063191124822, "grad_norm": 1.5522487163543701, "learning_rate": 4.992483790913423e-05, "loss": 2.315, "step": 809 }, { "epoch": 0.007285057088765273, "grad_norm": 1.7547898292541504, "learning_rate": 4.992425442306708e-05, "loss": 2.3906, "step": 810 }, { "epoch": 0.0072940509864057234, "grad_norm": 1.3132327795028687, "learning_rate": 4.992366868437266e-05, "loss": 2.3774, "step": 811 }, { "epoch": 0.007303044884046175, "grad_norm": 1.0966075658798218, "learning_rate": 4.992308069310391e-05, "loss": 2.394, "step": 812 }, { "epoch": 0.007312038781686626, "grad_norm": 1.5437568426132202, "learning_rate": 4.992249044931396e-05, "loss": 2.3382, "step": 813 }, { "epoch": 0.007321032679327076, "grad_norm": 0.9832726120948792, "learning_rate": 4.992189795305618e-05, "loss": 2.3258, "step": 814 }, { "epoch": 0.0073300265769675275, "grad_norm": 0.9568870067596436, "learning_rate": 4.9921303204384104e-05, "loss": 2.463, "step": 815 }, { "epoch": 0.007339020474607979, "grad_norm": 0.9940560460090637, "learning_rate": 4.99207062033515e-05, "loss": 2.4384, "step": 816 }, { "epoch": 0.007348014372248429, "grad_norm": 0.773494303226471, "learning_rate": 4.992010695001229e-05, "loss": 2.4385, "step": 817 }, { "epoch": 0.00735700826988888, "grad_norm": 0.9313816428184509, "learning_rate": 4.9919505444420666e-05, "loss": 2.3089, "step": 818 }, { "epoch": 0.0073660021675293315, "grad_norm": 1.4051645994186401, "learning_rate": 4.991890168663098e-05, "loss": 2.4094, "step": 819 }, { "epoch": 0.007374996065169782, "grad_norm": 0.708573043346405, "learning_rate": 4.99182956766978e-05, "loss": 2.3811, "step": 820 }, { "epoch": 0.007383989962810233, "grad_norm": 1.4287480115890503, "learning_rate": 4.99176874146759e-05, "loss": 2.3966, "step": 821 }, { "epoch": 0.007392983860450684, "grad_norm": 0.9948827028274536, "learning_rate": 4.991707690062025e-05, "loss": 2.2247, "step": 822 }, { "epoch": 0.0074019777580911356, "grad_norm": 0.7052820324897766, "learning_rate": 4.9916464134586024e-05, "loss": 2.4127, "step": 823 }, { "epoch": 0.007410971655731586, "grad_norm": 0.8648499250411987, "learning_rate": 4.991584911662861e-05, "loss": 2.3283, "step": 824 }, { "epoch": 0.007419965553372037, "grad_norm": 1.2007049322128296, "learning_rate": 4.991523184680359e-05, "loss": 2.3553, "step": 825 }, { "epoch": 0.007428959451012488, "grad_norm": 1.5342097282409668, "learning_rate": 4.991461232516675e-05, "loss": 2.4664, "step": 826 }, { "epoch": 0.007437953348652939, "grad_norm": 1.252152681350708, "learning_rate": 4.991399055177408e-05, "loss": 2.3259, "step": 827 }, { "epoch": 0.00744694724629339, "grad_norm": 1.6909875869750977, "learning_rate": 4.991336652668179e-05, "loss": 2.5274, "step": 828 }, { "epoch": 0.007455941143933841, "grad_norm": 1.041088342666626, "learning_rate": 4.991274024994626e-05, "loss": 2.5211, "step": 829 }, { "epoch": 0.0074649350415742916, "grad_norm": 1.0081087350845337, "learning_rate": 4.99121117216241e-05, "loss": 2.2854, "step": 830 }, { "epoch": 0.007473928939214743, "grad_norm": 0.7612619996070862, "learning_rate": 4.9911480941772115e-05, "loss": 2.4526, "step": 831 }, { "epoch": 0.007482922836855194, "grad_norm": 1.263737440109253, "learning_rate": 4.991084791044732e-05, "loss": 2.3774, "step": 832 }, { "epoch": 0.007491916734495644, "grad_norm": 1.2397956848144531, "learning_rate": 4.991021262770692e-05, "loss": 2.4083, "step": 833 }, { "epoch": 0.007500910632136096, "grad_norm": 1.0493193864822388, "learning_rate": 4.9909575093608334e-05, "loss": 2.3255, "step": 834 }, { "epoch": 0.007509904529776547, "grad_norm": 1.3190057277679443, "learning_rate": 4.990893530820918e-05, "loss": 2.3486, "step": 835 }, { "epoch": 0.007518898427416997, "grad_norm": 1.0106217861175537, "learning_rate": 4.9908293271567286e-05, "loss": 2.2965, "step": 836 }, { "epoch": 0.007527892325057448, "grad_norm": 2.0340452194213867, "learning_rate": 4.990764898374067e-05, "loss": 2.344, "step": 837 }, { "epoch": 0.0075368862226979, "grad_norm": 0.9287651777267456, "learning_rate": 4.990700244478758e-05, "loss": 2.2849, "step": 838 }, { "epoch": 0.00754588012033835, "grad_norm": 1.7290759086608887, "learning_rate": 4.990635365476643e-05, "loss": 2.3609, "step": 839 }, { "epoch": 0.007554874017978801, "grad_norm": 1.1983683109283447, "learning_rate": 4.990570261373585e-05, "loss": 2.4809, "step": 840 }, { "epoch": 0.0075638679156192525, "grad_norm": 1.5039162635803223, "learning_rate": 4.9905049321754706e-05, "loss": 2.3046, "step": 841 }, { "epoch": 0.007572861813259704, "grad_norm": 1.4246561527252197, "learning_rate": 4.990439377888203e-05, "loss": 2.3694, "step": 842 }, { "epoch": 0.007581855710900154, "grad_norm": 1.9078946113586426, "learning_rate": 4.990373598517707e-05, "loss": 2.3287, "step": 843 }, { "epoch": 0.007590849608540605, "grad_norm": 2.58236026763916, "learning_rate": 4.9903075940699265e-05, "loss": 2.4055, "step": 844 }, { "epoch": 0.0075998435061810565, "grad_norm": 1.0287704467773438, "learning_rate": 4.990241364550829e-05, "loss": 2.2513, "step": 845 }, { "epoch": 0.007608837403821507, "grad_norm": 1.243863821029663, "learning_rate": 4.990174909966399e-05, "loss": 2.3509, "step": 846 }, { "epoch": 0.007617831301461958, "grad_norm": 2.5628280639648438, "learning_rate": 4.990108230322643e-05, "loss": 2.278, "step": 847 }, { "epoch": 0.007626825199102409, "grad_norm": 1.5202381610870361, "learning_rate": 4.9900413256255876e-05, "loss": 2.3038, "step": 848 }, { "epoch": 0.00763581909674286, "grad_norm": 2.9225621223449707, "learning_rate": 4.989974195881279e-05, "loss": 2.427, "step": 849 }, { "epoch": 0.007644812994383311, "grad_norm": 1.8009926080703735, "learning_rate": 4.989906841095784e-05, "loss": 2.4446, "step": 850 }, { "epoch": 0.007653806892023762, "grad_norm": 1.1386442184448242, "learning_rate": 4.989839261275191e-05, "loss": 2.3672, "step": 851 }, { "epoch": 0.0076628007896642125, "grad_norm": 2.892897605895996, "learning_rate": 4.989771456425607e-05, "loss": 2.3714, "step": 852 }, { "epoch": 0.007671794687304664, "grad_norm": 2.243138074874878, "learning_rate": 4.989703426553161e-05, "loss": 2.2661, "step": 853 }, { "epoch": 0.007680788584945115, "grad_norm": 4.217054843902588, "learning_rate": 4.989635171664e-05, "loss": 2.3475, "step": 854 }, { "epoch": 0.007689782482585565, "grad_norm": 1.5125004053115845, "learning_rate": 4.989566691764295e-05, "loss": 2.316, "step": 855 }, { "epoch": 0.0076987763802260165, "grad_norm": 1.1293413639068604, "learning_rate": 4.989497986860234e-05, "loss": 2.2831, "step": 856 }, { "epoch": 0.007707770277866468, "grad_norm": 1.9106346368789673, "learning_rate": 4.989429056958026e-05, "loss": 2.426, "step": 857 }, { "epoch": 0.007716764175506919, "grad_norm": 0.665691077709198, "learning_rate": 4.989359902063901e-05, "loss": 2.374, "step": 858 }, { "epoch": 0.007725758073147369, "grad_norm": 0.8827734589576721, "learning_rate": 4.98929052218411e-05, "loss": 2.2281, "step": 859 }, { "epoch": 0.007734751970787821, "grad_norm": 1.4260919094085693, "learning_rate": 4.989220917324922e-05, "loss": 2.3352, "step": 860 }, { "epoch": 0.007743745868428272, "grad_norm": 2.3282699584960938, "learning_rate": 4.9891510874926293e-05, "loss": 2.3059, "step": 861 }, { "epoch": 0.007752739766068722, "grad_norm": 2.224483013153076, "learning_rate": 4.989081032693542e-05, "loss": 2.3128, "step": 862 }, { "epoch": 0.007761733663709173, "grad_norm": 0.88776034116745, "learning_rate": 4.989010752933992e-05, "loss": 2.3525, "step": 863 }, { "epoch": 0.007770727561349625, "grad_norm": 2.1006646156311035, "learning_rate": 4.988940248220332e-05, "loss": 2.3066, "step": 864 }, { "epoch": 0.007779721458990075, "grad_norm": 1.7027292251586914, "learning_rate": 4.9888695185589325e-05, "loss": 2.3422, "step": 865 }, { "epoch": 0.007788715356630526, "grad_norm": 0.925018846988678, "learning_rate": 4.988798563956186e-05, "loss": 2.327, "step": 866 }, { "epoch": 0.007797709254270977, "grad_norm": 1.78342604637146, "learning_rate": 4.988727384418507e-05, "loss": 2.3251, "step": 867 }, { "epoch": 0.007806703151911428, "grad_norm": 1.8967210054397583, "learning_rate": 4.9886559799523275e-05, "loss": 2.4215, "step": 868 }, { "epoch": 0.00781569704955188, "grad_norm": 1.4762319326400757, "learning_rate": 4.9885843505641014e-05, "loss": 2.4299, "step": 869 }, { "epoch": 0.00782469094719233, "grad_norm": 1.4177857637405396, "learning_rate": 4.988512496260301e-05, "loss": 2.2442, "step": 870 }, { "epoch": 0.00783368484483278, "grad_norm": 3.988291025161743, "learning_rate": 4.988440417047423e-05, "loss": 2.3613, "step": 871 }, { "epoch": 0.007842678742473233, "grad_norm": 2.524430751800537, "learning_rate": 4.988368112931979e-05, "loss": 2.3907, "step": 872 }, { "epoch": 0.007851672640113683, "grad_norm": 2.3439295291900635, "learning_rate": 4.988295583920506e-05, "loss": 2.4879, "step": 873 }, { "epoch": 0.007860666537754133, "grad_norm": 1.3418093919754028, "learning_rate": 4.988222830019559e-05, "loss": 2.3817, "step": 874 }, { "epoch": 0.007869660435394586, "grad_norm": 2.416830539703369, "learning_rate": 4.988149851235712e-05, "loss": 2.3644, "step": 875 }, { "epoch": 0.007878654333035036, "grad_norm": 5.460973739624023, "learning_rate": 4.988076647575562e-05, "loss": 2.4638, "step": 876 }, { "epoch": 0.007887648230675486, "grad_norm": 7.149078369140625, "learning_rate": 4.9880032190457226e-05, "loss": 2.3455, "step": 877 }, { "epoch": 0.007896642128315938, "grad_norm": 7.11635160446167, "learning_rate": 4.9879295656528334e-05, "loss": 2.55, "step": 878 }, { "epoch": 0.007905636025956389, "grad_norm": 6.192065238952637, "learning_rate": 4.9878556874035504e-05, "loss": 2.3529, "step": 879 }, { "epoch": 0.007914629923596839, "grad_norm": 3.8169753551483154, "learning_rate": 4.9877815843045495e-05, "loss": 2.3865, "step": 880 }, { "epoch": 0.007923623821237291, "grad_norm": 1.9059431552886963, "learning_rate": 4.9877072563625285e-05, "loss": 2.2723, "step": 881 }, { "epoch": 0.007932617718877742, "grad_norm": 1.1022893190383911, "learning_rate": 4.987632703584206e-05, "loss": 2.452, "step": 882 }, { "epoch": 0.007941611616518192, "grad_norm": 1.7090600728988647, "learning_rate": 4.987557925976318e-05, "loss": 2.3762, "step": 883 }, { "epoch": 0.007950605514158644, "grad_norm": 1.3606576919555664, "learning_rate": 4.987482923545625e-05, "loss": 2.3087, "step": 884 }, { "epoch": 0.007959599411799094, "grad_norm": 1.358298897743225, "learning_rate": 4.987407696298904e-05, "loss": 2.3496, "step": 885 }, { "epoch": 0.007968593309439545, "grad_norm": 1.6232304573059082, "learning_rate": 4.987332244242955e-05, "loss": 2.3479, "step": 886 }, { "epoch": 0.007977587207079997, "grad_norm": 1.5596123933792114, "learning_rate": 4.9872565673845974e-05, "loss": 2.4425, "step": 887 }, { "epoch": 0.007986581104720447, "grad_norm": 1.5620065927505493, "learning_rate": 4.98718066573067e-05, "loss": 2.3169, "step": 888 }, { "epoch": 0.007995575002360897, "grad_norm": 2.1944029331207275, "learning_rate": 4.987104539288032e-05, "loss": 2.42, "step": 889 }, { "epoch": 0.00800456890000135, "grad_norm": 1.7498599290847778, "learning_rate": 4.9870281880635656e-05, "loss": 2.3805, "step": 890 }, { "epoch": 0.0080135627976418, "grad_norm": 1.612972378730774, "learning_rate": 4.98695161206417e-05, "loss": 2.3094, "step": 891 }, { "epoch": 0.00802255669528225, "grad_norm": 2.5815799236297607, "learning_rate": 4.986874811296767e-05, "loss": 2.4157, "step": 892 }, { "epoch": 0.008031550592922702, "grad_norm": 17.66390037536621, "learning_rate": 4.9867977857682965e-05, "loss": 2.4277, "step": 893 }, { "epoch": 0.008040544490563153, "grad_norm": 46.72271728515625, "learning_rate": 4.9867205354857206e-05, "loss": 2.3533, "step": 894 }, { "epoch": 0.008049538388203603, "grad_norm": 2.8205180168151855, "learning_rate": 4.9866430604560214e-05, "loss": 2.372, "step": 895 }, { "epoch": 0.008058532285844055, "grad_norm": 4.408201217651367, "learning_rate": 4.986565360686201e-05, "loss": 2.4591, "step": 896 }, { "epoch": 0.008067526183484506, "grad_norm": 1.0639312267303467, "learning_rate": 4.986487436183282e-05, "loss": 2.3197, "step": 897 }, { "epoch": 0.008076520081124956, "grad_norm": 1.9649598598480225, "learning_rate": 4.986409286954306e-05, "loss": 2.2815, "step": 898 }, { "epoch": 0.008085513978765408, "grad_norm": 2.717100143432617, "learning_rate": 4.986330913006337e-05, "loss": 2.4578, "step": 899 }, { "epoch": 0.008094507876405858, "grad_norm": 2.211270332336426, "learning_rate": 4.986252314346458e-05, "loss": 2.2714, "step": 900 }, { "epoch": 0.008103501774046309, "grad_norm": 2.953294515609741, "learning_rate": 4.986173490981773e-05, "loss": 2.3534, "step": 901 }, { "epoch": 0.00811249567168676, "grad_norm": 20.0116024017334, "learning_rate": 4.986094442919406e-05, "loss": 2.246, "step": 902 }, { "epoch": 0.008121489569327211, "grad_norm": 56.963191986083984, "learning_rate": 4.9860151701665004e-05, "loss": 2.3506, "step": 903 }, { "epoch": 0.008130483466967663, "grad_norm": 2.4784321784973145, "learning_rate": 4.985935672730222e-05, "loss": 2.3257, "step": 904 }, { "epoch": 0.008139477364608114, "grad_norm": 1.7748329639434814, "learning_rate": 4.985855950617755e-05, "loss": 2.3586, "step": 905 }, { "epoch": 0.008148471262248564, "grad_norm": 2.1759116649627686, "learning_rate": 4.985776003836304e-05, "loss": 2.3651, "step": 906 }, { "epoch": 0.008157465159889016, "grad_norm": 1.564903974533081, "learning_rate": 4.985695832393097e-05, "loss": 2.408, "step": 907 }, { "epoch": 0.008166459057529466, "grad_norm": 0.7693034410476685, "learning_rate": 4.985615436295377e-05, "loss": 2.2915, "step": 908 }, { "epoch": 0.008175452955169917, "grad_norm": 3.041149854660034, "learning_rate": 4.9855348155504104e-05, "loss": 2.3744, "step": 909 }, { "epoch": 0.008184446852810369, "grad_norm": 2.2738850116729736, "learning_rate": 4.9854539701654855e-05, "loss": 2.3491, "step": 910 }, { "epoch": 0.00819344075045082, "grad_norm": 2.5227854251861572, "learning_rate": 4.9853729001479075e-05, "loss": 2.373, "step": 911 }, { "epoch": 0.00820243464809127, "grad_norm": 1.8525545597076416, "learning_rate": 4.9852916055050035e-05, "loss": 2.3316, "step": 912 }, { "epoch": 0.008211428545731722, "grad_norm": 1.6783733367919922, "learning_rate": 4.985210086244121e-05, "loss": 2.3053, "step": 913 }, { "epoch": 0.008220422443372172, "grad_norm": 3.593388319015503, "learning_rate": 4.985128342372629e-05, "loss": 2.4132, "step": 914 }, { "epoch": 0.008229416341012622, "grad_norm": 1.6809263229370117, "learning_rate": 4.985046373897913e-05, "loss": 2.388, "step": 915 }, { "epoch": 0.008238410238653075, "grad_norm": 1.7394527196884155, "learning_rate": 4.984964180827383e-05, "loss": 2.2643, "step": 916 }, { "epoch": 0.008247404136293525, "grad_norm": 1.6720887422561646, "learning_rate": 4.984881763168466e-05, "loss": 2.3783, "step": 917 }, { "epoch": 0.008256398033933975, "grad_norm": 114.69852447509766, "learning_rate": 4.9847991209286124e-05, "loss": 2.3095, "step": 918 }, { "epoch": 0.008265391931574427, "grad_norm": 2.773820400238037, "learning_rate": 4.9847162541152906e-05, "loss": 2.3849, "step": 919 }, { "epoch": 0.008274385829214878, "grad_norm": 2.6709611415863037, "learning_rate": 4.98463316273599e-05, "loss": 2.2671, "step": 920 }, { "epoch": 0.008283379726855328, "grad_norm": 2.1583499908447266, "learning_rate": 4.984549846798221e-05, "loss": 2.3633, "step": 921 }, { "epoch": 0.00829237362449578, "grad_norm": 2.721205472946167, "learning_rate": 4.984466306309512e-05, "loss": 2.3699, "step": 922 }, { "epoch": 0.00830136752213623, "grad_norm": 3.0230939388275146, "learning_rate": 4.9843825412774145e-05, "loss": 2.3702, "step": 923 }, { "epoch": 0.008310361419776681, "grad_norm": 2.07593035697937, "learning_rate": 4.9842985517094996e-05, "loss": 2.2955, "step": 924 }, { "epoch": 0.008319355317417133, "grad_norm": 8.249615669250488, "learning_rate": 4.984214337613357e-05, "loss": 2.3203, "step": 925 }, { "epoch": 0.008328349215057583, "grad_norm": 3.146702766418457, "learning_rate": 4.9841298989965984e-05, "loss": 2.4488, "step": 926 }, { "epoch": 0.008337343112698034, "grad_norm": 3.03059458732605, "learning_rate": 4.984045235866855e-05, "loss": 2.3512, "step": 927 }, { "epoch": 0.008346337010338486, "grad_norm": 2.2177159786224365, "learning_rate": 4.983960348231779e-05, "loss": 2.3528, "step": 928 }, { "epoch": 0.008355330907978936, "grad_norm": 6.031072616577148, "learning_rate": 4.983875236099042e-05, "loss": 2.4123, "step": 929 }, { "epoch": 0.008364324805619387, "grad_norm": 4.340119361877441, "learning_rate": 4.983789899476337e-05, "loss": 2.4107, "step": 930 }, { "epoch": 0.008373318703259839, "grad_norm": 14.878398895263672, "learning_rate": 4.9837043383713753e-05, "loss": 2.4057, "step": 931 }, { "epoch": 0.008382312600900289, "grad_norm": 8.246759414672852, "learning_rate": 4.9836185527918924e-05, "loss": 2.3712, "step": 932 }, { "epoch": 0.00839130649854074, "grad_norm": 7.280088901519775, "learning_rate": 4.983532542745638e-05, "loss": 2.4236, "step": 933 }, { "epoch": 0.008400300396181191, "grad_norm": 4.351801872253418, "learning_rate": 4.983446308240388e-05, "loss": 2.3233, "step": 934 }, { "epoch": 0.008409294293821642, "grad_norm": 2.8788869380950928, "learning_rate": 4.983359849283936e-05, "loss": 2.3607, "step": 935 }, { "epoch": 0.008418288191462092, "grad_norm": 49.51881408691406, "learning_rate": 4.9832731658840956e-05, "loss": 2.2114, "step": 936 }, { "epoch": 0.008427282089102544, "grad_norm": 16.22137451171875, "learning_rate": 4.983186258048701e-05, "loss": 2.3123, "step": 937 }, { "epoch": 0.008436275986742995, "grad_norm": 1.5916885137557983, "learning_rate": 4.9830991257856074e-05, "loss": 2.332, "step": 938 }, { "epoch": 0.008445269884383445, "grad_norm": 2.1934919357299805, "learning_rate": 4.983011769102689e-05, "loss": 2.3972, "step": 939 }, { "epoch": 0.008454263782023897, "grad_norm": 1.4318190813064575, "learning_rate": 4.982924188007841e-05, "loss": 2.3014, "step": 940 }, { "epoch": 0.008463257679664347, "grad_norm": 3.1829514503479004, "learning_rate": 4.982836382508981e-05, "loss": 2.538, "step": 941 }, { "epoch": 0.0084722515773048, "grad_norm": 1.1255837678909302, "learning_rate": 4.9827483526140414e-05, "loss": 2.3257, "step": 942 }, { "epoch": 0.00848124547494525, "grad_norm": 2.0071754455566406, "learning_rate": 4.98266009833098e-05, "loss": 2.4001, "step": 943 }, { "epoch": 0.0084902393725857, "grad_norm": 1.5771818161010742, "learning_rate": 4.982571619667774e-05, "loss": 2.3978, "step": 944 }, { "epoch": 0.008499233270226152, "grad_norm": 2.3359062671661377, "learning_rate": 4.9824829166324186e-05, "loss": 2.3576, "step": 945 }, { "epoch": 0.008508227167866603, "grad_norm": 85.4216537475586, "learning_rate": 4.982393989232931e-05, "loss": 2.3332, "step": 946 }, { "epoch": 0.008517221065507053, "grad_norm": 3.0369982719421387, "learning_rate": 4.982304837477348e-05, "loss": 2.3304, "step": 947 }, { "epoch": 0.008526214963147505, "grad_norm": 2.0758142471313477, "learning_rate": 4.982215461373728e-05, "loss": 2.3491, "step": 948 }, { "epoch": 0.008535208860787956, "grad_norm": 2.0540716648101807, "learning_rate": 4.982125860930148e-05, "loss": 2.4209, "step": 949 }, { "epoch": 0.008544202758428406, "grad_norm": 2.7230679988861084, "learning_rate": 4.982036036154706e-05, "loss": 2.4095, "step": 950 }, { "epoch": 0.008553196656068858, "grad_norm": 2.2635655403137207, "learning_rate": 4.981945987055521e-05, "loss": 2.3302, "step": 951 }, { "epoch": 0.008562190553709308, "grad_norm": 2.8918027877807617, "learning_rate": 4.9818557136407304e-05, "loss": 2.2959, "step": 952 }, { "epoch": 0.008571184451349759, "grad_norm": 0.9320359230041504, "learning_rate": 4.9817652159184946e-05, "loss": 2.297, "step": 953 }, { "epoch": 0.00858017834899021, "grad_norm": 1.637445092201233, "learning_rate": 4.9816744938969904e-05, "loss": 2.3192, "step": 954 }, { "epoch": 0.008589172246630661, "grad_norm": 0.8960732221603394, "learning_rate": 4.98158354758442e-05, "loss": 2.351, "step": 955 }, { "epoch": 0.008598166144271112, "grad_norm": 0.9512754082679749, "learning_rate": 4.9814923769890004e-05, "loss": 2.3218, "step": 956 }, { "epoch": 0.008607160041911564, "grad_norm": 1.2916690111160278, "learning_rate": 4.9814009821189734e-05, "loss": 2.4091, "step": 957 }, { "epoch": 0.008616153939552014, "grad_norm": 0.800326406955719, "learning_rate": 4.981309362982598e-05, "loss": 2.3718, "step": 958 }, { "epoch": 0.008625147837192464, "grad_norm": 1.2276074886322021, "learning_rate": 4.981217519588155e-05, "loss": 2.4268, "step": 959 }, { "epoch": 0.008634141734832916, "grad_norm": 1.0263639688491821, "learning_rate": 4.981125451943944e-05, "loss": 2.3521, "step": 960 }, { "epoch": 0.008643135632473367, "grad_norm": 2.415966749191284, "learning_rate": 4.981033160058289e-05, "loss": 2.383, "step": 961 }, { "epoch": 0.008652129530113817, "grad_norm": 1.3266791105270386, "learning_rate": 4.9809406439395286e-05, "loss": 2.3436, "step": 962 }, { "epoch": 0.00866112342775427, "grad_norm": 2.955108404159546, "learning_rate": 4.980847903596025e-05, "loss": 2.3794, "step": 963 }, { "epoch": 0.00867011732539472, "grad_norm": 1.9091037511825562, "learning_rate": 4.98075493903616e-05, "loss": 2.375, "step": 964 }, { "epoch": 0.00867911122303517, "grad_norm": 6.714507579803467, "learning_rate": 4.9806617502683356e-05, "loss": 2.3218, "step": 965 }, { "epoch": 0.008688105120675622, "grad_norm": 1.7211779356002808, "learning_rate": 4.9805683373009746e-05, "loss": 2.3044, "step": 966 }, { "epoch": 0.008697099018316072, "grad_norm": 1.9710533618927002, "learning_rate": 4.980474700142519e-05, "loss": 2.3285, "step": 967 }, { "epoch": 0.008706092915956523, "grad_norm": 1.4440377950668335, "learning_rate": 4.980380838801431e-05, "loss": 2.326, "step": 968 }, { "epoch": 0.008715086813596975, "grad_norm": 2.5308713912963867, "learning_rate": 4.980286753286195e-05, "loss": 2.3517, "step": 969 }, { "epoch": 0.008724080711237425, "grad_norm": 2.4353814125061035, "learning_rate": 4.980192443605314e-05, "loss": 2.3247, "step": 970 }, { "epoch": 0.008733074608877876, "grad_norm": 1.2030874490737915, "learning_rate": 4.9800979097673106e-05, "loss": 2.2785, "step": 971 }, { "epoch": 0.008742068506518328, "grad_norm": 4.3923540115356445, "learning_rate": 4.98000315178073e-05, "loss": 2.4156, "step": 972 }, { "epoch": 0.008751062404158778, "grad_norm": 2.5101020336151123, "learning_rate": 4.979908169654136e-05, "loss": 2.3458, "step": 973 }, { "epoch": 0.008760056301799228, "grad_norm": 2.055736780166626, "learning_rate": 4.9798129633961124e-05, "loss": 2.366, "step": 974 }, { "epoch": 0.00876905019943968, "grad_norm": 1.0032880306243896, "learning_rate": 4.979717533015264e-05, "loss": 2.2664, "step": 975 }, { "epoch": 0.00877804409708013, "grad_norm": 2.2960662841796875, "learning_rate": 4.979621878520216e-05, "loss": 2.4005, "step": 976 }, { "epoch": 0.008787037994720581, "grad_norm": 1.7479488849639893, "learning_rate": 4.979525999919614e-05, "loss": 2.2448, "step": 977 }, { "epoch": 0.008796031892361033, "grad_norm": 1.5676523447036743, "learning_rate": 4.9794298972221225e-05, "loss": 2.3536, "step": 978 }, { "epoch": 0.008805025790001484, "grad_norm": 78.5261459350586, "learning_rate": 4.979333570436427e-05, "loss": 2.2746, "step": 979 }, { "epoch": 0.008814019687641936, "grad_norm": 9.62118911743164, "learning_rate": 4.979237019571235e-05, "loss": 2.2938, "step": 980 }, { "epoch": 0.008823013585282386, "grad_norm": 2.8171589374542236, "learning_rate": 4.979140244635271e-05, "loss": 2.3666, "step": 981 }, { "epoch": 0.008832007482922836, "grad_norm": 1.6373629570007324, "learning_rate": 4.979043245637282e-05, "loss": 2.244, "step": 982 }, { "epoch": 0.008841001380563289, "grad_norm": 2.18770694732666, "learning_rate": 4.978946022586035e-05, "loss": 2.376, "step": 983 }, { "epoch": 0.008849995278203739, "grad_norm": 2.425063371658325, "learning_rate": 4.978848575490316e-05, "loss": 2.272, "step": 984 }, { "epoch": 0.00885898917584419, "grad_norm": 1.0710437297821045, "learning_rate": 4.9787509043589334e-05, "loss": 2.3817, "step": 985 }, { "epoch": 0.008867983073484641, "grad_norm": 1.5350664854049683, "learning_rate": 4.9786530092007135e-05, "loss": 2.3173, "step": 986 }, { "epoch": 0.008876976971125092, "grad_norm": 2.1459872722625732, "learning_rate": 4.9785548900245045e-05, "loss": 2.3475, "step": 987 }, { "epoch": 0.008885970868765542, "grad_norm": 2.152179718017578, "learning_rate": 4.978456546839174e-05, "loss": 2.3343, "step": 988 }, { "epoch": 0.008894964766405994, "grad_norm": 1.4224385023117065, "learning_rate": 4.978357979653611e-05, "loss": 2.3542, "step": 989 }, { "epoch": 0.008903958664046445, "grad_norm": 1.3504154682159424, "learning_rate": 4.9782591884767224e-05, "loss": 2.3621, "step": 990 }, { "epoch": 0.008912952561686895, "grad_norm": 7.234045505523682, "learning_rate": 4.978160173317438e-05, "loss": 2.377, "step": 991 }, { "epoch": 0.008921946459327347, "grad_norm": 2.711791753768921, "learning_rate": 4.978060934184707e-05, "loss": 2.406, "step": 992 }, { "epoch": 0.008930940356967797, "grad_norm": 1.7958492040634155, "learning_rate": 4.9779614710874974e-05, "loss": 2.3133, "step": 993 }, { "epoch": 0.008939934254608248, "grad_norm": 4.097232818603516, "learning_rate": 4.977861784034799e-05, "loss": 2.4167, "step": 994 }, { "epoch": 0.0089489281522487, "grad_norm": 1.4877700805664062, "learning_rate": 4.977761873035622e-05, "loss": 2.4374, "step": 995 }, { "epoch": 0.00895792204988915, "grad_norm": 19.301212310791016, "learning_rate": 4.977661738098995e-05, "loss": 2.4801, "step": 996 }, { "epoch": 0.0089669159475296, "grad_norm": 1.8256185054779053, "learning_rate": 4.977561379233969e-05, "loss": 2.4585, "step": 997 }, { "epoch": 0.008975909845170053, "grad_norm": 3.464038848876953, "learning_rate": 4.9774607964496144e-05, "loss": 2.3755, "step": 998 }, { "epoch": 0.008984903742810503, "grad_norm": 2.2096097469329834, "learning_rate": 4.9773599897550214e-05, "loss": 2.4156, "step": 999 }, { "epoch": 0.008993897640450953, "grad_norm": 4.295968532562256, "learning_rate": 4.977258959159301e-05, "loss": 2.3988, "step": 1000 }, { "epoch": 0.009002891538091405, "grad_norm": 1.1036388874053955, "learning_rate": 4.9771577046715846e-05, "loss": 2.3178, "step": 1001 }, { "epoch": 0.009011885435731856, "grad_norm": 1.7228236198425293, "learning_rate": 4.9770562263010225e-05, "loss": 2.2841, "step": 1002 }, { "epoch": 0.009020879333372306, "grad_norm": 2.547356128692627, "learning_rate": 4.976954524056787e-05, "loss": 2.4064, "step": 1003 }, { "epoch": 0.009029873231012758, "grad_norm": 2.5266928672790527, "learning_rate": 4.976852597948069e-05, "loss": 2.3287, "step": 1004 }, { "epoch": 0.009038867128653209, "grad_norm": 2.128464460372925, "learning_rate": 4.9767504479840816e-05, "loss": 2.3066, "step": 1005 }, { "epoch": 0.009047861026293659, "grad_norm": 1.5903282165527344, "learning_rate": 4.976648074174056e-05, "loss": 2.3114, "step": 1006 }, { "epoch": 0.009056854923934111, "grad_norm": 284.1065368652344, "learning_rate": 4.976545476527246e-05, "loss": 2.4002, "step": 1007 }, { "epoch": 0.009065848821574561, "grad_norm": 446.20068359375, "learning_rate": 4.976442655052922e-05, "loss": 2.5204, "step": 1008 }, { "epoch": 0.009074842719215012, "grad_norm": 4.401348114013672, "learning_rate": 4.976339609760379e-05, "loss": 2.3228, "step": 1009 }, { "epoch": 0.009083836616855464, "grad_norm": 1.9144341945648193, "learning_rate": 4.9762363406589295e-05, "loss": 2.3604, "step": 1010 }, { "epoch": 0.009092830514495914, "grad_norm": 18.24340057373047, "learning_rate": 4.976132847757906e-05, "loss": 2.3804, "step": 1011 }, { "epoch": 0.009101824412136365, "grad_norm": 1.3318119049072266, "learning_rate": 4.976029131066664e-05, "loss": 2.2884, "step": 1012 }, { "epoch": 0.009110818309776817, "grad_norm": 1.1708160638809204, "learning_rate": 4.975925190594575e-05, "loss": 2.3304, "step": 1013 }, { "epoch": 0.009119812207417267, "grad_norm": 1.9248275756835938, "learning_rate": 4.9758210263510345e-05, "loss": 2.4356, "step": 1014 }, { "epoch": 0.00912880610505772, "grad_norm": 1.1926578283309937, "learning_rate": 4.9757166383454565e-05, "loss": 2.3249, "step": 1015 }, { "epoch": 0.00913780000269817, "grad_norm": 1.5307071208953857, "learning_rate": 4.975612026587275e-05, "loss": 2.3416, "step": 1016 }, { "epoch": 0.00914679390033862, "grad_norm": 1.9043503999710083, "learning_rate": 4.9755071910859456e-05, "loss": 2.4471, "step": 1017 }, { "epoch": 0.009155787797979072, "grad_norm": 1.5771394968032837, "learning_rate": 4.975402131850942e-05, "loss": 2.3845, "step": 1018 }, { "epoch": 0.009164781695619522, "grad_norm": 1.0784991979599, "learning_rate": 4.9752968488917604e-05, "loss": 2.2247, "step": 1019 }, { "epoch": 0.009173775593259973, "grad_norm": 11.075843811035156, "learning_rate": 4.975191342217917e-05, "loss": 2.2718, "step": 1020 }, { "epoch": 0.009182769490900425, "grad_norm": 11.83606243133545, "learning_rate": 4.975085611838944e-05, "loss": 2.3569, "step": 1021 }, { "epoch": 0.009191763388540875, "grad_norm": 6.07887601852417, "learning_rate": 4.9749796577644e-05, "loss": 2.4011, "step": 1022 }, { "epoch": 0.009200757286181326, "grad_norm": 1.940321445465088, "learning_rate": 4.974873480003861e-05, "loss": 2.2614, "step": 1023 }, { "epoch": 0.009209751183821778, "grad_norm": 3.0341317653656006, "learning_rate": 4.974767078566922e-05, "loss": 2.4608, "step": 1024 }, { "epoch": 0.009218745081462228, "grad_norm": 2.68550181388855, "learning_rate": 4.9746604534632004e-05, "loss": 2.2984, "step": 1025 }, { "epoch": 0.009227738979102678, "grad_norm": 4.981675624847412, "learning_rate": 4.9745536047023324e-05, "loss": 2.3595, "step": 1026 }, { "epoch": 0.00923673287674313, "grad_norm": 3.624321222305298, "learning_rate": 4.974446532293975e-05, "loss": 2.3652, "step": 1027 }, { "epoch": 0.00924572677438358, "grad_norm": 2.262476682662964, "learning_rate": 4.9743392362478046e-05, "loss": 2.3697, "step": 1028 }, { "epoch": 0.009254720672024031, "grad_norm": 2.0064961910247803, "learning_rate": 4.9742317165735196e-05, "loss": 2.3023, "step": 1029 }, { "epoch": 0.009263714569664483, "grad_norm": 1.8211779594421387, "learning_rate": 4.9741239732808377e-05, "loss": 2.3624, "step": 1030 }, { "epoch": 0.009272708467304934, "grad_norm": 1.7825077772140503, "learning_rate": 4.974016006379495e-05, "loss": 2.3577, "step": 1031 }, { "epoch": 0.009281702364945384, "grad_norm": 2.8979403972625732, "learning_rate": 4.9739078158792514e-05, "loss": 2.4009, "step": 1032 }, { "epoch": 0.009290696262585836, "grad_norm": 3.5438995361328125, "learning_rate": 4.973799401789884e-05, "loss": 2.393, "step": 1033 }, { "epoch": 0.009299690160226286, "grad_norm": 2.3994574546813965, "learning_rate": 4.9736907641211904e-05, "loss": 2.3648, "step": 1034 }, { "epoch": 0.009308684057866737, "grad_norm": 1.2672652006149292, "learning_rate": 4.97358190288299e-05, "loss": 2.2841, "step": 1035 }, { "epoch": 0.009317677955507189, "grad_norm": 1.8701341152191162, "learning_rate": 4.973472818085122e-05, "loss": 2.341, "step": 1036 }, { "epoch": 0.00932667185314764, "grad_norm": 2.4897868633270264, "learning_rate": 4.973363509737445e-05, "loss": 2.3821, "step": 1037 }, { "epoch": 0.00933566575078809, "grad_norm": 2.0629661083221436, "learning_rate": 4.973253977849838e-05, "loss": 2.3637, "step": 1038 }, { "epoch": 0.009344659648428542, "grad_norm": 1.559417963027954, "learning_rate": 4.9731442224322e-05, "loss": 2.2815, "step": 1039 }, { "epoch": 0.009353653546068992, "grad_norm": 1.5687872171401978, "learning_rate": 4.973034243494452e-05, "loss": 2.3517, "step": 1040 }, { "epoch": 0.009362647443709442, "grad_norm": 1.0377593040466309, "learning_rate": 4.9729240410465316e-05, "loss": 2.2756, "step": 1041 }, { "epoch": 0.009371641341349895, "grad_norm": 1.5999455451965332, "learning_rate": 4.9728136150984e-05, "loss": 2.3488, "step": 1042 }, { "epoch": 0.009380635238990345, "grad_norm": 1.3381636142730713, "learning_rate": 4.972702965660039e-05, "loss": 2.3803, "step": 1043 }, { "epoch": 0.009389629136630795, "grad_norm": 1.8739471435546875, "learning_rate": 4.972592092741446e-05, "loss": 2.3593, "step": 1044 }, { "epoch": 0.009398623034271247, "grad_norm": 1.6740717887878418, "learning_rate": 4.972480996352644e-05, "loss": 2.3002, "step": 1045 }, { "epoch": 0.009407616931911698, "grad_norm": 1.1257169246673584, "learning_rate": 4.972369676503672e-05, "loss": 2.349, "step": 1046 }, { "epoch": 0.009416610829552148, "grad_norm": 37.121726989746094, "learning_rate": 4.9722581332045906e-05, "loss": 2.3289, "step": 1047 }, { "epoch": 0.0094256047271926, "grad_norm": 1.7633475065231323, "learning_rate": 4.9721463664654835e-05, "loss": 2.375, "step": 1048 }, { "epoch": 0.00943459862483305, "grad_norm": 2.242678642272949, "learning_rate": 4.97203437629645e-05, "loss": 2.2794, "step": 1049 }, { "epoch": 0.009443592522473501, "grad_norm": 2.0801889896392822, "learning_rate": 4.971922162707612e-05, "loss": 2.2789, "step": 1050 }, { "epoch": 0.009452586420113953, "grad_norm": 1.684820294380188, "learning_rate": 4.971809725709112e-05, "loss": 2.4247, "step": 1051 }, { "epoch": 0.009461580317754403, "grad_norm": 2.8138086795806885, "learning_rate": 4.971697065311111e-05, "loss": 2.3125, "step": 1052 }, { "epoch": 0.009470574215394855, "grad_norm": 1.4332460165023804, "learning_rate": 4.971584181523792e-05, "loss": 2.3261, "step": 1053 }, { "epoch": 0.009479568113035306, "grad_norm": 1.1635377407073975, "learning_rate": 4.9714710743573566e-05, "loss": 2.3102, "step": 1054 }, { "epoch": 0.009488562010675756, "grad_norm": 12.210381507873535, "learning_rate": 4.971357743822027e-05, "loss": 2.3955, "step": 1055 }, { "epoch": 0.009497555908316208, "grad_norm": 2.8512823581695557, "learning_rate": 4.971244189928047e-05, "loss": 2.2482, "step": 1056 }, { "epoch": 0.009506549805956659, "grad_norm": 2.8980586528778076, "learning_rate": 4.971130412685679e-05, "loss": 2.2319, "step": 1057 }, { "epoch": 0.009515543703597109, "grad_norm": 2.3892319202423096, "learning_rate": 4.971016412105206e-05, "loss": 2.4365, "step": 1058 }, { "epoch": 0.009524537601237561, "grad_norm": 1.2553093433380127, "learning_rate": 4.970902188196931e-05, "loss": 2.2569, "step": 1059 }, { "epoch": 0.009533531498878011, "grad_norm": 1.172286868095398, "learning_rate": 4.970787740971178e-05, "loss": 2.2426, "step": 1060 }, { "epoch": 0.009542525396518462, "grad_norm": 1.9281389713287354, "learning_rate": 4.97067307043829e-05, "loss": 2.3148, "step": 1061 }, { "epoch": 0.009551519294158914, "grad_norm": 5.392895221710205, "learning_rate": 4.970558176608631e-05, "loss": 2.3564, "step": 1062 }, { "epoch": 0.009560513191799364, "grad_norm": 111.55917358398438, "learning_rate": 4.970443059492585e-05, "loss": 2.6691, "step": 1063 }, { "epoch": 0.009569507089439815, "grad_norm": 79.27272033691406, "learning_rate": 4.9703277191005556e-05, "loss": 2.7401, "step": 1064 }, { "epoch": 0.009578500987080267, "grad_norm": 64.21418762207031, "learning_rate": 4.9702121554429695e-05, "loss": 2.7139, "step": 1065 }, { "epoch": 0.009587494884720717, "grad_norm": 238.8323211669922, "learning_rate": 4.970096368530268e-05, "loss": 2.5298, "step": 1066 }, { "epoch": 0.009596488782361167, "grad_norm": 5.2633185386657715, "learning_rate": 4.969980358372918e-05, "loss": 2.4147, "step": 1067 }, { "epoch": 0.00960548268000162, "grad_norm": 3.5609946250915527, "learning_rate": 4.9698641249814036e-05, "loss": 2.2809, "step": 1068 }, { "epoch": 0.00961447657764207, "grad_norm": 186.07591247558594, "learning_rate": 4.96974766836623e-05, "loss": 2.438, "step": 1069 }, { "epoch": 0.00962347047528252, "grad_norm": 37.068077087402344, "learning_rate": 4.9696309885379225e-05, "loss": 2.3861, "step": 1070 }, { "epoch": 0.009632464372922972, "grad_norm": 4.245080947875977, "learning_rate": 4.9695140855070254e-05, "loss": 2.3331, "step": 1071 }, { "epoch": 0.009641458270563423, "grad_norm": 1.162161111831665, "learning_rate": 4.9693969592841054e-05, "loss": 2.2964, "step": 1072 }, { "epoch": 0.009650452168203873, "grad_norm": 1.679726481437683, "learning_rate": 4.969279609879749e-05, "loss": 2.3182, "step": 1073 }, { "epoch": 0.009659446065844325, "grad_norm": 322.6358642578125, "learning_rate": 4.96916203730456e-05, "loss": 2.2659, "step": 1074 }, { "epoch": 0.009668439963484775, "grad_norm": 1.6927112340927124, "learning_rate": 4.969044241569166e-05, "loss": 2.3038, "step": 1075 }, { "epoch": 0.009677433861125226, "grad_norm": 106.5965805053711, "learning_rate": 4.968926222684213e-05, "loss": 2.3294, "step": 1076 }, { "epoch": 0.009686427758765678, "grad_norm": 5.916990756988525, "learning_rate": 4.968807980660367e-05, "loss": 2.3532, "step": 1077 }, { "epoch": 0.009695421656406128, "grad_norm": 1.4046456813812256, "learning_rate": 4.968689515508316e-05, "loss": 2.3392, "step": 1078 }, { "epoch": 0.009704415554046579, "grad_norm": 2.997627019882202, "learning_rate": 4.968570827238764e-05, "loss": 2.3581, "step": 1079 }, { "epoch": 0.00971340945168703, "grad_norm": 2.466237783432007, "learning_rate": 4.968451915862441e-05, "loss": 2.3702, "step": 1080 }, { "epoch": 0.009722403349327481, "grad_norm": 6.010172367095947, "learning_rate": 4.968332781390092e-05, "loss": 2.339, "step": 1081 }, { "epoch": 0.009731397246967931, "grad_norm": 1.6601020097732544, "learning_rate": 4.968213423832485e-05, "loss": 2.306, "step": 1082 }, { "epoch": 0.009740391144608384, "grad_norm": 1.492484450340271, "learning_rate": 4.9680938432004074e-05, "loss": 2.3155, "step": 1083 }, { "epoch": 0.009749385042248834, "grad_norm": 1.6578973531723022, "learning_rate": 4.967974039504666e-05, "loss": 2.2692, "step": 1084 }, { "epoch": 0.009758378939889284, "grad_norm": 1.943566918373108, "learning_rate": 4.9678540127560904e-05, "loss": 2.4199, "step": 1085 }, { "epoch": 0.009767372837529736, "grad_norm": 1.309987187385559, "learning_rate": 4.9677337629655264e-05, "loss": 2.2808, "step": 1086 }, { "epoch": 0.009776366735170187, "grad_norm": 1.8059741258621216, "learning_rate": 4.967613290143843e-05, "loss": 2.3359, "step": 1087 }, { "epoch": 0.009785360632810637, "grad_norm": 1.6679922342300415, "learning_rate": 4.967492594301928e-05, "loss": 2.3139, "step": 1088 }, { "epoch": 0.00979435453045109, "grad_norm": 1.5452219247817993, "learning_rate": 4.9673716754506906e-05, "loss": 2.3277, "step": 1089 }, { "epoch": 0.00980334842809154, "grad_norm": 0.9614936709403992, "learning_rate": 4.967250533601059e-05, "loss": 2.2946, "step": 1090 }, { "epoch": 0.009812342325731992, "grad_norm": 1.8186993598937988, "learning_rate": 4.967129168763981e-05, "loss": 2.3634, "step": 1091 }, { "epoch": 0.009821336223372442, "grad_norm": 3.445974826812744, "learning_rate": 4.967007580950427e-05, "loss": 2.3921, "step": 1092 }, { "epoch": 0.009830330121012892, "grad_norm": 214.7694549560547, "learning_rate": 4.9668857701713844e-05, "loss": 2.3131, "step": 1093 }, { "epoch": 0.009839324018653344, "grad_norm": 129.35501098632812, "learning_rate": 4.9667637364378625e-05, "loss": 2.3979, "step": 1094 }, { "epoch": 0.009848317916293795, "grad_norm": 22.181787490844727, "learning_rate": 4.966641479760892e-05, "loss": 2.3343, "step": 1095 }, { "epoch": 0.009857311813934245, "grad_norm": 2.1598572731018066, "learning_rate": 4.966519000151522e-05, "loss": 2.3796, "step": 1096 }, { "epoch": 0.009866305711574697, "grad_norm": 2.760488748550415, "learning_rate": 4.96639629762082e-05, "loss": 2.3354, "step": 1097 }, { "epoch": 0.009875299609215148, "grad_norm": 23.494564056396484, "learning_rate": 4.966273372179878e-05, "loss": 2.301, "step": 1098 }, { "epoch": 0.009884293506855598, "grad_norm": 958.7348022460938, "learning_rate": 4.966150223839805e-05, "loss": 2.6469, "step": 1099 }, { "epoch": 0.00989328740449605, "grad_norm": 3.113955497741699, "learning_rate": 4.966026852611731e-05, "loss": 2.4272, "step": 1100 }, { "epoch": 0.0099022813021365, "grad_norm": 52.463993072509766, "learning_rate": 4.965903258506806e-05, "loss": 2.3236, "step": 1101 }, { "epoch": 0.00991127519977695, "grad_norm": 1.7886731624603271, "learning_rate": 4.9657794415362014e-05, "loss": 2.2811, "step": 1102 }, { "epoch": 0.009920269097417403, "grad_norm": 2.48513126373291, "learning_rate": 4.965655401711106e-05, "loss": 2.2932, "step": 1103 }, { "epoch": 0.009929262995057853, "grad_norm": 1.5509709119796753, "learning_rate": 4.965531139042731e-05, "loss": 2.3416, "step": 1104 }, { "epoch": 0.009938256892698304, "grad_norm": 4.904428482055664, "learning_rate": 4.9654066535423075e-05, "loss": 2.3302, "step": 1105 }, { "epoch": 0.009947250790338756, "grad_norm": 2.888819932937622, "learning_rate": 4.9652819452210865e-05, "loss": 2.3515, "step": 1106 }, { "epoch": 0.009956244687979206, "grad_norm": 1.615666389465332, "learning_rate": 4.9651570140903384e-05, "loss": 2.2589, "step": 1107 }, { "epoch": 0.009965238585619656, "grad_norm": 1.6766654253005981, "learning_rate": 4.965031860161355e-05, "loss": 2.3282, "step": 1108 }, { "epoch": 0.009974232483260109, "grad_norm": 2.3949713706970215, "learning_rate": 4.9649064834454474e-05, "loss": 2.2685, "step": 1109 }, { "epoch": 0.009983226380900559, "grad_norm": 1.4269980192184448, "learning_rate": 4.9647808839539455e-05, "loss": 2.3944, "step": 1110 }, { "epoch": 0.00999222027854101, "grad_norm": 1.2341117858886719, "learning_rate": 4.9646550616982034e-05, "loss": 2.2753, "step": 1111 }, { "epoch": 0.010001214176181461, "grad_norm": 1.9939247369766235, "learning_rate": 4.9645290166895906e-05, "loss": 2.3253, "step": 1112 }, { "epoch": 0.010010208073821912, "grad_norm": 1.176496982574463, "learning_rate": 4.9644027489395005e-05, "loss": 2.3194, "step": 1113 }, { "epoch": 0.010019201971462362, "grad_norm": 2.368574380874634, "learning_rate": 4.964276258459345e-05, "loss": 2.3987, "step": 1114 }, { "epoch": 0.010028195869102814, "grad_norm": 1.1555325984954834, "learning_rate": 4.9641495452605545e-05, "loss": 2.2783, "step": 1115 }, { "epoch": 0.010037189766743265, "grad_norm": 11.08785343170166, "learning_rate": 4.964022609354583e-05, "loss": 2.3416, "step": 1116 }, { "epoch": 0.010046183664383715, "grad_norm": 3.6222121715545654, "learning_rate": 4.9638954507529014e-05, "loss": 2.3072, "step": 1117 }, { "epoch": 0.010055177562024167, "grad_norm": 5.869036674499512, "learning_rate": 4.963768069467003e-05, "loss": 2.3855, "step": 1118 }, { "epoch": 0.010064171459664617, "grad_norm": 1.9966332912445068, "learning_rate": 4.9636404655084e-05, "loss": 2.3254, "step": 1119 }, { "epoch": 0.010073165357305068, "grad_norm": 4.625076770782471, "learning_rate": 4.9635126388886256e-05, "loss": 2.3371, "step": 1120 }, { "epoch": 0.01008215925494552, "grad_norm": 3.0359678268432617, "learning_rate": 4.963384589619233e-05, "loss": 2.2586, "step": 1121 }, { "epoch": 0.01009115315258597, "grad_norm": 1.8090620040893555, "learning_rate": 4.963256317711794e-05, "loss": 2.2716, "step": 1122 }, { "epoch": 0.01010014705022642, "grad_norm": 1.4668354988098145, "learning_rate": 4.963127823177902e-05, "loss": 2.3089, "step": 1123 }, { "epoch": 0.010109140947866873, "grad_norm": 2.646360158920288, "learning_rate": 4.96299910602917e-05, "loss": 2.3216, "step": 1124 }, { "epoch": 0.010118134845507323, "grad_norm": 1.7903180122375488, "learning_rate": 4.9628701662772325e-05, "loss": 2.2168, "step": 1125 }, { "epoch": 0.010127128743147775, "grad_norm": 1.1596676111221313, "learning_rate": 4.962741003933742e-05, "loss": 2.2366, "step": 1126 }, { "epoch": 0.010136122640788225, "grad_norm": 2.070632219314575, "learning_rate": 4.9626116190103724e-05, "loss": 2.3454, "step": 1127 }, { "epoch": 0.010145116538428676, "grad_norm": 0.9789115786552429, "learning_rate": 4.9624820115188174e-05, "loss": 2.2597, "step": 1128 }, { "epoch": 0.010154110436069128, "grad_norm": 1.8018876314163208, "learning_rate": 4.9623521814707894e-05, "loss": 2.305, "step": 1129 }, { "epoch": 0.010163104333709578, "grad_norm": 0.8252420425415039, "learning_rate": 4.962222128878025e-05, "loss": 2.2355, "step": 1130 }, { "epoch": 0.010172098231350029, "grad_norm": 1.1904879808425903, "learning_rate": 4.962091853752275e-05, "loss": 2.3075, "step": 1131 }, { "epoch": 0.01018109212899048, "grad_norm": 1.8671382665634155, "learning_rate": 4.961961356105317e-05, "loss": 2.2624, "step": 1132 }, { "epoch": 0.010190086026630931, "grad_norm": 1.6544936895370483, "learning_rate": 4.961830635948942e-05, "loss": 2.3444, "step": 1133 }, { "epoch": 0.010199079924271381, "grad_norm": 2.3454082012176514, "learning_rate": 4.9616996932949666e-05, "loss": 2.4088, "step": 1134 }, { "epoch": 0.010208073821911834, "grad_norm": 1.764322280883789, "learning_rate": 4.961568528155224e-05, "loss": 2.2591, "step": 1135 }, { "epoch": 0.010217067719552284, "grad_norm": 1.2431902885437012, "learning_rate": 4.9614371405415696e-05, "loss": 2.3133, "step": 1136 }, { "epoch": 0.010226061617192734, "grad_norm": 1.7182575464248657, "learning_rate": 4.9613055304658776e-05, "loss": 2.2835, "step": 1137 }, { "epoch": 0.010235055514833186, "grad_norm": 16.691743850708008, "learning_rate": 4.961173697940042e-05, "loss": 2.2866, "step": 1138 }, { "epoch": 0.010244049412473637, "grad_norm": 1.0919997692108154, "learning_rate": 4.9610416429759797e-05, "loss": 2.3138, "step": 1139 }, { "epoch": 0.010253043310114087, "grad_norm": 2.4553284645080566, "learning_rate": 4.960909365585624e-05, "loss": 2.4208, "step": 1140 }, { "epoch": 0.01026203720775454, "grad_norm": 2.8286502361297607, "learning_rate": 4.9607768657809306e-05, "loss": 2.3583, "step": 1141 }, { "epoch": 0.01027103110539499, "grad_norm": 1.5299936532974243, "learning_rate": 4.960644143573875e-05, "loss": 2.3979, "step": 1142 }, { "epoch": 0.01028002500303544, "grad_norm": 1.3973337411880493, "learning_rate": 4.960511198976451e-05, "loss": 2.2907, "step": 1143 }, { "epoch": 0.010289018900675892, "grad_norm": 1.4424344301223755, "learning_rate": 4.960378032000676e-05, "loss": 2.2768, "step": 1144 }, { "epoch": 0.010298012798316342, "grad_norm": 1.1546728610992432, "learning_rate": 4.960244642658585e-05, "loss": 2.2212, "step": 1145 }, { "epoch": 0.010307006695956793, "grad_norm": 3.8030622005462646, "learning_rate": 4.960111030962232e-05, "loss": 2.3394, "step": 1146 }, { "epoch": 0.010316000593597245, "grad_norm": 2.399183750152588, "learning_rate": 4.959977196923694e-05, "loss": 2.3322, "step": 1147 }, { "epoch": 0.010324994491237695, "grad_norm": 3.7612507343292236, "learning_rate": 4.959843140555067e-05, "loss": 2.3258, "step": 1148 }, { "epoch": 0.010333988388878146, "grad_norm": 2.4283108711242676, "learning_rate": 4.959708861868466e-05, "loss": 2.3865, "step": 1149 }, { "epoch": 0.010342982286518598, "grad_norm": 2.4122281074523926, "learning_rate": 4.959574360876028e-05, "loss": 2.3813, "step": 1150 }, { "epoch": 0.010351976184159048, "grad_norm": 2.301945447921753, "learning_rate": 4.959439637589909e-05, "loss": 2.3788, "step": 1151 }, { "epoch": 0.010360970081799498, "grad_norm": 2.392803907394409, "learning_rate": 4.959304692022284e-05, "loss": 2.3019, "step": 1152 }, { "epoch": 0.01036996397943995, "grad_norm": 1.652524709701538, "learning_rate": 4.9591695241853495e-05, "loss": 2.2857, "step": 1153 }, { "epoch": 0.0103789578770804, "grad_norm": 1.127707600593567, "learning_rate": 4.959034134091323e-05, "loss": 2.2377, "step": 1154 }, { "epoch": 0.010387951774720851, "grad_norm": 2.509152412414551, "learning_rate": 4.95889852175244e-05, "loss": 2.3167, "step": 1155 }, { "epoch": 0.010396945672361303, "grad_norm": 2.5287764072418213, "learning_rate": 4.958762687180956e-05, "loss": 2.3174, "step": 1156 }, { "epoch": 0.010405939570001754, "grad_norm": 2.6051721572875977, "learning_rate": 4.958626630389149e-05, "loss": 2.3245, "step": 1157 }, { "epoch": 0.010414933467642204, "grad_norm": 3.279592275619507, "learning_rate": 4.958490351389316e-05, "loss": 2.3171, "step": 1158 }, { "epoch": 0.010423927365282656, "grad_norm": 3.551753520965576, "learning_rate": 4.958353850193773e-05, "loss": 2.2767, "step": 1159 }, { "epoch": 0.010432921262923106, "grad_norm": 19.09039306640625, "learning_rate": 4.958217126814857e-05, "loss": 2.3301, "step": 1160 }, { "epoch": 0.010441915160563557, "grad_norm": 1.6319879293441772, "learning_rate": 4.958080181264925e-05, "loss": 2.3, "step": 1161 }, { "epoch": 0.010450909058204009, "grad_norm": 1.3658466339111328, "learning_rate": 4.9579430135563534e-05, "loss": 2.3458, "step": 1162 }, { "epoch": 0.01045990295584446, "grad_norm": 1.20742666721344, "learning_rate": 4.95780562370154e-05, "loss": 2.2795, "step": 1163 }, { "epoch": 0.010468896853484911, "grad_norm": 1.6854474544525146, "learning_rate": 4.957668011712902e-05, "loss": 2.3564, "step": 1164 }, { "epoch": 0.010477890751125362, "grad_norm": 2.292989492416382, "learning_rate": 4.957530177602876e-05, "loss": 2.3595, "step": 1165 }, { "epoch": 0.010486884648765812, "grad_norm": 1.4159866571426392, "learning_rate": 4.9573921213839194e-05, "loss": 2.3254, "step": 1166 }, { "epoch": 0.010495878546406264, "grad_norm": 1.691332221031189, "learning_rate": 4.9572538430685094e-05, "loss": 2.3564, "step": 1167 }, { "epoch": 0.010504872444046714, "grad_norm": 2.640218734741211, "learning_rate": 4.957115342669144e-05, "loss": 2.3182, "step": 1168 }, { "epoch": 0.010513866341687165, "grad_norm": 1.481858253479004, "learning_rate": 4.956976620198342e-05, "loss": 2.2696, "step": 1169 }, { "epoch": 0.010522860239327617, "grad_norm": 1.0896230936050415, "learning_rate": 4.956837675668637e-05, "loss": 2.2795, "step": 1170 }, { "epoch": 0.010531854136968067, "grad_norm": 1.335352897644043, "learning_rate": 4.9566985090925904e-05, "loss": 2.3225, "step": 1171 }, { "epoch": 0.010540848034608518, "grad_norm": 8.702324867248535, "learning_rate": 4.956559120482779e-05, "loss": 2.3123, "step": 1172 }, { "epoch": 0.01054984193224897, "grad_norm": 2.045257329940796, "learning_rate": 4.9564195098518e-05, "loss": 2.3944, "step": 1173 }, { "epoch": 0.01055883582988942, "grad_norm": 1.743530511856079, "learning_rate": 4.956279677212271e-05, "loss": 2.3436, "step": 1174 }, { "epoch": 0.01056782972752987, "grad_norm": 2.2260138988494873, "learning_rate": 4.956139622576831e-05, "loss": 2.4495, "step": 1175 }, { "epoch": 0.010576823625170323, "grad_norm": 1.0863250494003296, "learning_rate": 4.9559993459581375e-05, "loss": 2.3514, "step": 1176 }, { "epoch": 0.010585817522810773, "grad_norm": 1.050481915473938, "learning_rate": 4.9558588473688685e-05, "loss": 2.3664, "step": 1177 }, { "epoch": 0.010594811420451223, "grad_norm": 2.1504368782043457, "learning_rate": 4.9557181268217227e-05, "loss": 2.3724, "step": 1178 }, { "epoch": 0.010603805318091675, "grad_norm": 3.4603984355926514, "learning_rate": 4.955577184329416e-05, "loss": 2.3355, "step": 1179 }, { "epoch": 0.010612799215732126, "grad_norm": 1.2105674743652344, "learning_rate": 4.955436019904689e-05, "loss": 2.3539, "step": 1180 }, { "epoch": 0.010621793113372576, "grad_norm": 1.112475037574768, "learning_rate": 4.9552946335603e-05, "loss": 2.2702, "step": 1181 }, { "epoch": 0.010630787011013028, "grad_norm": 1.2488195896148682, "learning_rate": 4.9551530253090263e-05, "loss": 2.2923, "step": 1182 }, { "epoch": 0.010639780908653479, "grad_norm": 1.3281227350234985, "learning_rate": 4.9550111951636666e-05, "loss": 2.2355, "step": 1183 }, { "epoch": 0.010648774806293929, "grad_norm": 2.6969902515411377, "learning_rate": 4.954869143137039e-05, "loss": 2.288, "step": 1184 }, { "epoch": 0.010657768703934381, "grad_norm": 1.9387229681015015, "learning_rate": 4.9547268692419834e-05, "loss": 2.3412, "step": 1185 }, { "epoch": 0.010666762601574831, "grad_norm": 0.9136784672737122, "learning_rate": 4.954584373491357e-05, "loss": 2.2832, "step": 1186 }, { "epoch": 0.010675756499215282, "grad_norm": 1.5051130056381226, "learning_rate": 4.9544416558980376e-05, "loss": 2.3421, "step": 1187 }, { "epoch": 0.010684750396855734, "grad_norm": 0.7083832621574402, "learning_rate": 4.9542987164749264e-05, "loss": 2.2245, "step": 1188 }, { "epoch": 0.010693744294496184, "grad_norm": 0.9482129812240601, "learning_rate": 4.95415555523494e-05, "loss": 2.2884, "step": 1189 }, { "epoch": 0.010702738192136635, "grad_norm": 1.8723399639129639, "learning_rate": 4.954012172191018e-05, "loss": 2.3801, "step": 1190 }, { "epoch": 0.010711732089777087, "grad_norm": 1.4367190599441528, "learning_rate": 4.95386856735612e-05, "loss": 2.2496, "step": 1191 }, { "epoch": 0.010720725987417537, "grad_norm": 1.2865971326828003, "learning_rate": 4.953724740743223e-05, "loss": 2.3046, "step": 1192 }, { "epoch": 0.010729719885057987, "grad_norm": 1.185848593711853, "learning_rate": 4.9535806923653275e-05, "loss": 2.2834, "step": 1193 }, { "epoch": 0.01073871378269844, "grad_norm": 1.4188623428344727, "learning_rate": 4.953436422235451e-05, "loss": 2.2954, "step": 1194 }, { "epoch": 0.01074770768033889, "grad_norm": 1.5299527645111084, "learning_rate": 4.953291930366634e-05, "loss": 2.2884, "step": 1195 }, { "epoch": 0.01075670157797934, "grad_norm": 1.187511682510376, "learning_rate": 4.953147216771935e-05, "loss": 2.359, "step": 1196 }, { "epoch": 0.010765695475619792, "grad_norm": 1.9731473922729492, "learning_rate": 4.953002281464432e-05, "loss": 2.2737, "step": 1197 }, { "epoch": 0.010774689373260243, "grad_norm": 0.9455071687698364, "learning_rate": 4.952857124457226e-05, "loss": 2.2927, "step": 1198 }, { "epoch": 0.010783683270900695, "grad_norm": 1.5308231115341187, "learning_rate": 4.952711745763434e-05, "loss": 2.3734, "step": 1199 }, { "epoch": 0.010792677168541145, "grad_norm": 1.7899839878082275, "learning_rate": 4.952566145396197e-05, "loss": 2.1991, "step": 1200 }, { "epoch": 0.010801671066181595, "grad_norm": 1.5891908407211304, "learning_rate": 4.952420323368673e-05, "loss": 2.4113, "step": 1201 }, { "epoch": 0.010810664963822048, "grad_norm": 1.4638893604278564, "learning_rate": 4.952274279694042e-05, "loss": 2.3024, "step": 1202 }, { "epoch": 0.010819658861462498, "grad_norm": 2.119044542312622, "learning_rate": 4.952128014385503e-05, "loss": 2.308, "step": 1203 }, { "epoch": 0.010828652759102948, "grad_norm": 1.6738792657852173, "learning_rate": 4.9519815274562756e-05, "loss": 2.2302, "step": 1204 }, { "epoch": 0.0108376466567434, "grad_norm": 1.505308747291565, "learning_rate": 4.9518348189195984e-05, "loss": 2.3468, "step": 1205 }, { "epoch": 0.01084664055438385, "grad_norm": 1.9311424493789673, "learning_rate": 4.9516878887887305e-05, "loss": 2.224, "step": 1206 }, { "epoch": 0.010855634452024301, "grad_norm": 2.1454856395721436, "learning_rate": 4.9515407370769526e-05, "loss": 2.338, "step": 1207 }, { "epoch": 0.010864628349664753, "grad_norm": 1.1918364763259888, "learning_rate": 4.951393363797564e-05, "loss": 2.3401, "step": 1208 }, { "epoch": 0.010873622247305204, "grad_norm": 1.2202930450439453, "learning_rate": 4.951245768963883e-05, "loss": 2.3299, "step": 1209 }, { "epoch": 0.010882616144945654, "grad_norm": 1.1868610382080078, "learning_rate": 4.95109795258925e-05, "loss": 2.3391, "step": 1210 }, { "epoch": 0.010891610042586106, "grad_norm": 1.0417895317077637, "learning_rate": 4.9509499146870236e-05, "loss": 2.2816, "step": 1211 }, { "epoch": 0.010900603940226556, "grad_norm": 2.5350186824798584, "learning_rate": 4.9508016552705835e-05, "loss": 2.3242, "step": 1212 }, { "epoch": 0.010909597837867007, "grad_norm": 1.461430311203003, "learning_rate": 4.9506531743533316e-05, "loss": 2.34, "step": 1213 }, { "epoch": 0.010918591735507459, "grad_norm": 3.288555145263672, "learning_rate": 4.9505044719486836e-05, "loss": 2.4029, "step": 1214 }, { "epoch": 0.01092758563314791, "grad_norm": 2.890031337738037, "learning_rate": 4.9503555480700815e-05, "loss": 2.3105, "step": 1215 }, { "epoch": 0.01093657953078836, "grad_norm": 5.223935127258301, "learning_rate": 4.9502064027309836e-05, "loss": 2.2961, "step": 1216 }, { "epoch": 0.010945573428428812, "grad_norm": 96.04720306396484, "learning_rate": 4.950057035944871e-05, "loss": 2.4028, "step": 1217 }, { "epoch": 0.010954567326069262, "grad_norm": 2.35675311088562, "learning_rate": 4.949907447725242e-05, "loss": 2.3561, "step": 1218 }, { "epoch": 0.010963561223709712, "grad_norm": 2.025588274002075, "learning_rate": 4.949757638085617e-05, "loss": 2.3425, "step": 1219 }, { "epoch": 0.010972555121350164, "grad_norm": 1.438686490058899, "learning_rate": 4.9496076070395346e-05, "loss": 2.3929, "step": 1220 }, { "epoch": 0.010981549018990615, "grad_norm": 2.8650848865509033, "learning_rate": 4.949457354600556e-05, "loss": 2.2802, "step": 1221 }, { "epoch": 0.010990542916631065, "grad_norm": 1.5126683712005615, "learning_rate": 4.94930688078226e-05, "loss": 2.2844, "step": 1222 }, { "epoch": 0.010999536814271517, "grad_norm": 1.913718819618225, "learning_rate": 4.949156185598246e-05, "loss": 2.3641, "step": 1223 }, { "epoch": 0.011008530711911968, "grad_norm": 1.2559486627578735, "learning_rate": 4.949005269062134e-05, "loss": 2.233, "step": 1224 }, { "epoch": 0.011017524609552418, "grad_norm": 1.4811545610427856, "learning_rate": 4.948854131187564e-05, "loss": 2.3706, "step": 1225 }, { "epoch": 0.01102651850719287, "grad_norm": 1.481111764907837, "learning_rate": 4.948702771988195e-05, "loss": 2.3131, "step": 1226 }, { "epoch": 0.01103551240483332, "grad_norm": 1.9949358701705933, "learning_rate": 4.948551191477707e-05, "loss": 2.3826, "step": 1227 }, { "epoch": 0.01104450630247377, "grad_norm": 2.204502582550049, "learning_rate": 4.9483993896697996e-05, "loss": 2.342, "step": 1228 }, { "epoch": 0.011053500200114223, "grad_norm": 2.171661376953125, "learning_rate": 4.948247366578192e-05, "loss": 2.3162, "step": 1229 }, { "epoch": 0.011062494097754673, "grad_norm": 2.7246861457824707, "learning_rate": 4.9480951222166254e-05, "loss": 2.3667, "step": 1230 }, { "epoch": 0.011071487995395124, "grad_norm": 1.9337729215621948, "learning_rate": 4.9479426565988585e-05, "loss": 2.3223, "step": 1231 }, { "epoch": 0.011080481893035576, "grad_norm": 2.4414193630218506, "learning_rate": 4.947789969738671e-05, "loss": 2.3164, "step": 1232 }, { "epoch": 0.011089475790676026, "grad_norm": 2.173196315765381, "learning_rate": 4.9476370616498617e-05, "loss": 2.2918, "step": 1233 }, { "epoch": 0.011098469688316476, "grad_norm": 2.7107467651367188, "learning_rate": 4.947483932346252e-05, "loss": 2.3572, "step": 1234 }, { "epoch": 0.011107463585956929, "grad_norm": 1.258103370666504, "learning_rate": 4.9473305818416806e-05, "loss": 2.2838, "step": 1235 }, { "epoch": 0.011116457483597379, "grad_norm": 2.1360626220703125, "learning_rate": 4.947177010150007e-05, "loss": 2.3383, "step": 1236 }, { "epoch": 0.011125451381237831, "grad_norm": 1.0039118528366089, "learning_rate": 4.9470232172851114e-05, "loss": 2.3596, "step": 1237 }, { "epoch": 0.011134445278878281, "grad_norm": 2.0521559715270996, "learning_rate": 4.9468692032608926e-05, "loss": 2.3033, "step": 1238 }, { "epoch": 0.011143439176518732, "grad_norm": 1.234839677810669, "learning_rate": 4.946714968091271e-05, "loss": 2.246, "step": 1239 }, { "epoch": 0.011152433074159184, "grad_norm": 1.0917353630065918, "learning_rate": 4.946560511790186e-05, "loss": 2.2894, "step": 1240 }, { "epoch": 0.011161426971799634, "grad_norm": 1.0826351642608643, "learning_rate": 4.946405834371597e-05, "loss": 2.3392, "step": 1241 }, { "epoch": 0.011170420869440085, "grad_norm": 1.1655303239822388, "learning_rate": 4.9462509358494844e-05, "loss": 2.3199, "step": 1242 }, { "epoch": 0.011179414767080537, "grad_norm": 1.2636141777038574, "learning_rate": 4.946095816237847e-05, "loss": 2.3306, "step": 1243 }, { "epoch": 0.011188408664720987, "grad_norm": 2.016878604888916, "learning_rate": 4.945940475550703e-05, "loss": 2.3402, "step": 1244 }, { "epoch": 0.011197402562361437, "grad_norm": 1.3355700969696045, "learning_rate": 4.945784913802095e-05, "loss": 2.243, "step": 1245 }, { "epoch": 0.01120639646000189, "grad_norm": 2.1973514556884766, "learning_rate": 4.9456291310060806e-05, "loss": 2.3478, "step": 1246 }, { "epoch": 0.01121539035764234, "grad_norm": 1.1568211317062378, "learning_rate": 4.9454731271767395e-05, "loss": 2.3208, "step": 1247 }, { "epoch": 0.01122438425528279, "grad_norm": 1.5626015663146973, "learning_rate": 4.945316902328171e-05, "loss": 2.2441, "step": 1248 }, { "epoch": 0.011233378152923242, "grad_norm": 2.322676420211792, "learning_rate": 4.9451604564744944e-05, "loss": 2.3715, "step": 1249 }, { "epoch": 0.011242372050563693, "grad_norm": 2.1260242462158203, "learning_rate": 4.9450037896298495e-05, "loss": 2.341, "step": 1250 }, { "epoch": 0.011251365948204143, "grad_norm": 1.714827537536621, "learning_rate": 4.9448469018083965e-05, "loss": 2.2521, "step": 1251 }, { "epoch": 0.011260359845844595, "grad_norm": 6.027016639709473, "learning_rate": 4.944689793024313e-05, "loss": 2.2911, "step": 1252 }, { "epoch": 0.011269353743485045, "grad_norm": 2.1543612480163574, "learning_rate": 4.9445324632918e-05, "loss": 2.3697, "step": 1253 }, { "epoch": 0.011278347641125496, "grad_norm": 1.6868714094161987, "learning_rate": 4.944374912625075e-05, "loss": 2.3858, "step": 1254 }, { "epoch": 0.011287341538765948, "grad_norm": 1.0732194185256958, "learning_rate": 4.944217141038379e-05, "loss": 2.3034, "step": 1255 }, { "epoch": 0.011296335436406398, "grad_norm": 1.4649441242218018, "learning_rate": 4.9440591485459706e-05, "loss": 2.3401, "step": 1256 }, { "epoch": 0.011305329334046849, "grad_norm": 1.239201545715332, "learning_rate": 4.943900935162128e-05, "loss": 2.3145, "step": 1257 }, { "epoch": 0.0113143232316873, "grad_norm": 1.889357566833496, "learning_rate": 4.9437425009011526e-05, "loss": 2.2848, "step": 1258 }, { "epoch": 0.011323317129327751, "grad_norm": 1.7509350776672363, "learning_rate": 4.9435838457773606e-05, "loss": 2.2887, "step": 1259 }, { "epoch": 0.011332311026968201, "grad_norm": 2.5500247478485107, "learning_rate": 4.943424969805094e-05, "loss": 2.302, "step": 1260 }, { "epoch": 0.011341304924608653, "grad_norm": 2.0146055221557617, "learning_rate": 4.94326587299871e-05, "loss": 2.3183, "step": 1261 }, { "epoch": 0.011350298822249104, "grad_norm": 2.1095030307769775, "learning_rate": 4.9431065553725874e-05, "loss": 2.293, "step": 1262 }, { "epoch": 0.011359292719889554, "grad_norm": 0.8967338800430298, "learning_rate": 4.942947016941126e-05, "loss": 2.2293, "step": 1263 }, { "epoch": 0.011368286617530006, "grad_norm": 1.6413196325302124, "learning_rate": 4.942787257718745e-05, "loss": 2.2868, "step": 1264 }, { "epoch": 0.011377280515170457, "grad_norm": 0.9218994379043579, "learning_rate": 4.9426272777198836e-05, "loss": 2.3297, "step": 1265 }, { "epoch": 0.011386274412810907, "grad_norm": 2.4999451637268066, "learning_rate": 4.9424670769589984e-05, "loss": 2.2538, "step": 1266 }, { "epoch": 0.011395268310451359, "grad_norm": 1.861364483833313, "learning_rate": 4.9423066554505706e-05, "loss": 2.3073, "step": 1267 }, { "epoch": 0.01140426220809181, "grad_norm": 2.0318455696105957, "learning_rate": 4.942146013209097e-05, "loss": 2.3113, "step": 1268 }, { "epoch": 0.01141325610573226, "grad_norm": 1.3827846050262451, "learning_rate": 4.9419851502490976e-05, "loss": 2.2541, "step": 1269 }, { "epoch": 0.011422250003372712, "grad_norm": 1.509055733680725, "learning_rate": 4.9418240665851106e-05, "loss": 2.3673, "step": 1270 }, { "epoch": 0.011431243901013162, "grad_norm": 1.5659303665161133, "learning_rate": 4.941662762231695e-05, "loss": 2.3634, "step": 1271 }, { "epoch": 0.011440237798653613, "grad_norm": 1.633475661277771, "learning_rate": 4.941501237203428e-05, "loss": 2.3204, "step": 1272 }, { "epoch": 0.011449231696294065, "grad_norm": 1.589223027229309, "learning_rate": 4.9413394915149094e-05, "loss": 2.3345, "step": 1273 }, { "epoch": 0.011458225593934515, "grad_norm": 1.170649528503418, "learning_rate": 4.941177525180757e-05, "loss": 2.2869, "step": 1274 }, { "epoch": 0.011467219491574967, "grad_norm": 0.9948899745941162, "learning_rate": 4.94101533821561e-05, "loss": 2.2896, "step": 1275 }, { "epoch": 0.011476213389215418, "grad_norm": 0.8960604667663574, "learning_rate": 4.9408529306341255e-05, "loss": 2.2698, "step": 1276 }, { "epoch": 0.011485207286855868, "grad_norm": 2.304053783416748, "learning_rate": 4.940690302450982e-05, "loss": 2.3175, "step": 1277 }, { "epoch": 0.01149420118449632, "grad_norm": 1.649854063987732, "learning_rate": 4.940527453680878e-05, "loss": 2.3434, "step": 1278 }, { "epoch": 0.01150319508213677, "grad_norm": 1.9188902378082275, "learning_rate": 4.940364384338532e-05, "loss": 2.3023, "step": 1279 }, { "epoch": 0.01151218897977722, "grad_norm": 1.5064107179641724, "learning_rate": 4.94020109443868e-05, "loss": 2.3428, "step": 1280 }, { "epoch": 0.011521182877417673, "grad_norm": 1.1684613227844238, "learning_rate": 4.9400375839960826e-05, "loss": 2.274, "step": 1281 }, { "epoch": 0.011530176775058123, "grad_norm": 1.879447340965271, "learning_rate": 4.9398738530255164e-05, "loss": 2.311, "step": 1282 }, { "epoch": 0.011539170672698574, "grad_norm": 5.295740127563477, "learning_rate": 4.9397099015417794e-05, "loss": 2.3564, "step": 1283 }, { "epoch": 0.011548164570339026, "grad_norm": 1.2326210737228394, "learning_rate": 4.9395457295596895e-05, "loss": 2.2755, "step": 1284 }, { "epoch": 0.011557158467979476, "grad_norm": 1.1373960971832275, "learning_rate": 4.939381337094084e-05, "loss": 2.2385, "step": 1285 }, { "epoch": 0.011566152365619926, "grad_norm": 1.6129990816116333, "learning_rate": 4.9392167241598206e-05, "loss": 2.3477, "step": 1286 }, { "epoch": 0.011575146263260378, "grad_norm": 1.6414514780044556, "learning_rate": 4.939051890771777e-05, "loss": 2.3078, "step": 1287 }, { "epoch": 0.011584140160900829, "grad_norm": 1.6887080669403076, "learning_rate": 4.938886836944851e-05, "loss": 2.2672, "step": 1288 }, { "epoch": 0.01159313405854128, "grad_norm": 1.3660883903503418, "learning_rate": 4.938721562693959e-05, "loss": 2.3015, "step": 1289 }, { "epoch": 0.011602127956181731, "grad_norm": 0.9072462916374207, "learning_rate": 4.93855606803404e-05, "loss": 2.2724, "step": 1290 }, { "epoch": 0.011611121853822182, "grad_norm": 1.0197970867156982, "learning_rate": 4.938390352980049e-05, "loss": 2.322, "step": 1291 }, { "epoch": 0.011620115751462632, "grad_norm": 1.6257226467132568, "learning_rate": 4.938224417546965e-05, "loss": 2.292, "step": 1292 }, { "epoch": 0.011629109649103084, "grad_norm": 0.8640071749687195, "learning_rate": 4.938058261749784e-05, "loss": 2.3107, "step": 1293 }, { "epoch": 0.011638103546743534, "grad_norm": 2.1623289585113525, "learning_rate": 4.937891885603524e-05, "loss": 2.3685, "step": 1294 }, { "epoch": 0.011647097444383985, "grad_norm": 2.623000144958496, "learning_rate": 4.9377252891232205e-05, "loss": 2.4302, "step": 1295 }, { "epoch": 0.011656091342024437, "grad_norm": 4.7325873374938965, "learning_rate": 4.937558472323932e-05, "loss": 2.3178, "step": 1296 }, { "epoch": 0.011665085239664887, "grad_norm": 2.6786139011383057, "learning_rate": 4.937391435220733e-05, "loss": 2.3095, "step": 1297 }, { "epoch": 0.011674079137305338, "grad_norm": 1.3070074319839478, "learning_rate": 4.937224177828722e-05, "loss": 2.3799, "step": 1298 }, { "epoch": 0.01168307303494579, "grad_norm": 4.2565693855285645, "learning_rate": 4.937056700163015e-05, "loss": 2.3069, "step": 1299 }, { "epoch": 0.01169206693258624, "grad_norm": 2.372837543487549, "learning_rate": 4.936889002238749e-05, "loss": 2.2814, "step": 1300 }, { "epoch": 0.01170106083022669, "grad_norm": 4.036974906921387, "learning_rate": 4.936721084071079e-05, "loss": 2.3156, "step": 1301 }, { "epoch": 0.011710054727867143, "grad_norm": 1.7612476348876953, "learning_rate": 4.9365529456751823e-05, "loss": 2.3362, "step": 1302 }, { "epoch": 0.011719048625507593, "grad_norm": 1.62795889377594, "learning_rate": 4.936384587066255e-05, "loss": 2.3077, "step": 1303 }, { "epoch": 0.011728042523148043, "grad_norm": 1.8024780750274658, "learning_rate": 4.936216008259513e-05, "loss": 2.2378, "step": 1304 }, { "epoch": 0.011737036420788495, "grad_norm": 1.9589920043945312, "learning_rate": 4.9360472092701914e-05, "loss": 2.3331, "step": 1305 }, { "epoch": 0.011746030318428946, "grad_norm": 5.74002742767334, "learning_rate": 4.935878190113548e-05, "loss": 2.272, "step": 1306 }, { "epoch": 0.011755024216069396, "grad_norm": 1.2023650407791138, "learning_rate": 4.935708950804857e-05, "loss": 2.3192, "step": 1307 }, { "epoch": 0.011764018113709848, "grad_norm": 2.29329252243042, "learning_rate": 4.935539491359415e-05, "loss": 2.4965, "step": 1308 }, { "epoch": 0.011773012011350299, "grad_norm": 5.815248489379883, "learning_rate": 4.935369811792536e-05, "loss": 2.2611, "step": 1309 }, { "epoch": 0.01178200590899075, "grad_norm": 1.0783790349960327, "learning_rate": 4.935199912119558e-05, "loss": 2.315, "step": 1310 }, { "epoch": 0.011790999806631201, "grad_norm": 1.660359263420105, "learning_rate": 4.935029792355834e-05, "loss": 2.2553, "step": 1311 }, { "epoch": 0.011799993704271651, "grad_norm": 1.8036712408065796, "learning_rate": 4.93485945251674e-05, "loss": 2.333, "step": 1312 }, { "epoch": 0.011808987601912103, "grad_norm": 1.599664330482483, "learning_rate": 4.934688892617672e-05, "loss": 2.3116, "step": 1313 }, { "epoch": 0.011817981499552554, "grad_norm": 4.0947957038879395, "learning_rate": 4.934518112674044e-05, "loss": 2.3851, "step": 1314 }, { "epoch": 0.011826975397193004, "grad_norm": 1.2490711212158203, "learning_rate": 4.9343471127012916e-05, "loss": 2.3138, "step": 1315 }, { "epoch": 0.011835969294833456, "grad_norm": 1.5987191200256348, "learning_rate": 4.9341758927148685e-05, "loss": 2.2805, "step": 1316 }, { "epoch": 0.011844963192473907, "grad_norm": 1.4530214071273804, "learning_rate": 4.934004452730251e-05, "loss": 2.2581, "step": 1317 }, { "epoch": 0.011853957090114357, "grad_norm": 1.4896512031555176, "learning_rate": 4.933832792762932e-05, "loss": 2.2328, "step": 1318 }, { "epoch": 0.011862950987754809, "grad_norm": 1.3908450603485107, "learning_rate": 4.933660912828427e-05, "loss": 2.2771, "step": 1319 }, { "epoch": 0.01187194488539526, "grad_norm": 1.1553617715835571, "learning_rate": 4.93348881294227e-05, "loss": 2.2609, "step": 1320 }, { "epoch": 0.01188093878303571, "grad_norm": 3.907634735107422, "learning_rate": 4.933316493120015e-05, "loss": 2.3196, "step": 1321 }, { "epoch": 0.011889932680676162, "grad_norm": 1.322037935256958, "learning_rate": 4.933143953377236e-05, "loss": 2.2876, "step": 1322 }, { "epoch": 0.011898926578316612, "grad_norm": 2.2371411323547363, "learning_rate": 4.932971193729528e-05, "loss": 2.2772, "step": 1323 }, { "epoch": 0.011907920475957063, "grad_norm": 1.2199677228927612, "learning_rate": 4.9327982141925036e-05, "loss": 2.3, "step": 1324 }, { "epoch": 0.011916914373597515, "grad_norm": 1.5293182134628296, "learning_rate": 4.9326250147817975e-05, "loss": 2.3158, "step": 1325 }, { "epoch": 0.011925908271237965, "grad_norm": 0.74586421251297, "learning_rate": 4.932451595513062e-05, "loss": 2.3088, "step": 1326 }, { "epoch": 0.011934902168878415, "grad_norm": 1.2650678157806396, "learning_rate": 4.932277956401972e-05, "loss": 2.329, "step": 1327 }, { "epoch": 0.011943896066518868, "grad_norm": 1.5978035926818848, "learning_rate": 4.93210409746422e-05, "loss": 2.2952, "step": 1328 }, { "epoch": 0.011952889964159318, "grad_norm": 1.2295321226119995, "learning_rate": 4.931930018715519e-05, "loss": 2.2475, "step": 1329 }, { "epoch": 0.011961883861799768, "grad_norm": 1.0471268892288208, "learning_rate": 4.931755720171603e-05, "loss": 2.2794, "step": 1330 }, { "epoch": 0.01197087775944022, "grad_norm": 2.31353497505188, "learning_rate": 4.931581201848224e-05, "loss": 2.356, "step": 1331 }, { "epoch": 0.01197987165708067, "grad_norm": 1.8516645431518555, "learning_rate": 4.931406463761154e-05, "loss": 2.3354, "step": 1332 }, { "epoch": 0.011988865554721121, "grad_norm": 1.6288126707077026, "learning_rate": 4.931231505926188e-05, "loss": 2.2762, "step": 1333 }, { "epoch": 0.011997859452361573, "grad_norm": 1.9435045719146729, "learning_rate": 4.931056328359137e-05, "loss": 2.3321, "step": 1334 }, { "epoch": 0.012006853350002024, "grad_norm": 1.2276501655578613, "learning_rate": 4.930880931075834e-05, "loss": 2.3042, "step": 1335 }, { "epoch": 0.012015847247642474, "grad_norm": 1.3298304080963135, "learning_rate": 4.93070531409213e-05, "loss": 2.3203, "step": 1336 }, { "epoch": 0.012024841145282926, "grad_norm": 0.9763184785842896, "learning_rate": 4.930529477423898e-05, "loss": 2.243, "step": 1337 }, { "epoch": 0.012033835042923376, "grad_norm": 1.9828107357025146, "learning_rate": 4.9303534210870295e-05, "loss": 2.3426, "step": 1338 }, { "epoch": 0.012042828940563827, "grad_norm": 2.406665802001953, "learning_rate": 4.930177145097438e-05, "loss": 2.3129, "step": 1339 }, { "epoch": 0.012051822838204279, "grad_norm": 1.736283302307129, "learning_rate": 4.9300006494710516e-05, "loss": 2.266, "step": 1340 }, { "epoch": 0.01206081673584473, "grad_norm": 1.6486119031906128, "learning_rate": 4.929823934223825e-05, "loss": 2.2789, "step": 1341 }, { "epoch": 0.01206981063348518, "grad_norm": 2.0691566467285156, "learning_rate": 4.929646999371729e-05, "loss": 2.3114, "step": 1342 }, { "epoch": 0.012078804531125632, "grad_norm": 1.2073266506195068, "learning_rate": 4.929469844930753e-05, "loss": 2.344, "step": 1343 }, { "epoch": 0.012087798428766082, "grad_norm": 1.0628546476364136, "learning_rate": 4.9292924709169095e-05, "loss": 2.305, "step": 1344 }, { "epoch": 0.012096792326406532, "grad_norm": 0.9728950262069702, "learning_rate": 4.929114877346229e-05, "loss": 2.265, "step": 1345 }, { "epoch": 0.012105786224046984, "grad_norm": 1.0010732412338257, "learning_rate": 4.928937064234763e-05, "loss": 2.2663, "step": 1346 }, { "epoch": 0.012114780121687435, "grad_norm": 1.1296560764312744, "learning_rate": 4.928759031598581e-05, "loss": 2.371, "step": 1347 }, { "epoch": 0.012123774019327887, "grad_norm": 1.5296412706375122, "learning_rate": 4.928580779453774e-05, "loss": 2.2256, "step": 1348 }, { "epoch": 0.012132767916968337, "grad_norm": 1.653023362159729, "learning_rate": 4.928402307816451e-05, "loss": 2.3913, "step": 1349 }, { "epoch": 0.012141761814608788, "grad_norm": 3.8353521823883057, "learning_rate": 4.928223616702744e-05, "loss": 2.4941, "step": 1350 }, { "epoch": 0.01215075571224924, "grad_norm": 2.857924699783325, "learning_rate": 4.928044706128803e-05, "loss": 2.421, "step": 1351 }, { "epoch": 0.01215974960988969, "grad_norm": 2.749598979949951, "learning_rate": 4.927865576110795e-05, "loss": 2.348, "step": 1352 }, { "epoch": 0.01216874350753014, "grad_norm": 1.8370927572250366, "learning_rate": 4.9276862266649114e-05, "loss": 2.3351, "step": 1353 }, { "epoch": 0.012177737405170592, "grad_norm": 0.851827085018158, "learning_rate": 4.9275066578073626e-05, "loss": 2.3425, "step": 1354 }, { "epoch": 0.012186731302811043, "grad_norm": 0.9431597590446472, "learning_rate": 4.927326869554376e-05, "loss": 2.3447, "step": 1355 }, { "epoch": 0.012195725200451493, "grad_norm": 1.717712640762329, "learning_rate": 4.9271468619222016e-05, "loss": 2.2972, "step": 1356 }, { "epoch": 0.012204719098091945, "grad_norm": 2.3689746856689453, "learning_rate": 4.926966634927108e-05, "loss": 2.2116, "step": 1357 }, { "epoch": 0.012213712995732396, "grad_norm": 1.7761175632476807, "learning_rate": 4.926786188585384e-05, "loss": 2.3788, "step": 1358 }, { "epoch": 0.012222706893372846, "grad_norm": 1.5584611892700195, "learning_rate": 4.9266055229133386e-05, "loss": 2.3209, "step": 1359 }, { "epoch": 0.012231700791013298, "grad_norm": 1.228891134262085, "learning_rate": 4.9264246379272996e-05, "loss": 2.3825, "step": 1360 }, { "epoch": 0.012240694688653748, "grad_norm": 4.562474727630615, "learning_rate": 4.9262435336436154e-05, "loss": 2.2612, "step": 1361 }, { "epoch": 0.012249688586294199, "grad_norm": 1.772830605506897, "learning_rate": 4.9260622100786534e-05, "loss": 2.2911, "step": 1362 }, { "epoch": 0.012258682483934651, "grad_norm": 2.08131742477417, "learning_rate": 4.9258806672488025e-05, "loss": 2.3708, "step": 1363 }, { "epoch": 0.012267676381575101, "grad_norm": 1.5537928342819214, "learning_rate": 4.9256989051704696e-05, "loss": 2.2242, "step": 1364 }, { "epoch": 0.012276670279215552, "grad_norm": 3.594076633453369, "learning_rate": 4.925516923860083e-05, "loss": 2.3497, "step": 1365 }, { "epoch": 0.012285664176856004, "grad_norm": 1.2255253791809082, "learning_rate": 4.925334723334089e-05, "loss": 2.3622, "step": 1366 }, { "epoch": 0.012294658074496454, "grad_norm": 1.7830429077148438, "learning_rate": 4.925152303608955e-05, "loss": 2.321, "step": 1367 }, { "epoch": 0.012303651972136904, "grad_norm": 0.9183161854743958, "learning_rate": 4.924969664701168e-05, "loss": 2.288, "step": 1368 }, { "epoch": 0.012312645869777357, "grad_norm": 0.824897289276123, "learning_rate": 4.9247868066272345e-05, "loss": 2.3118, "step": 1369 }, { "epoch": 0.012321639767417807, "grad_norm": 0.9935973286628723, "learning_rate": 4.924603729403681e-05, "loss": 2.2862, "step": 1370 }, { "epoch": 0.012330633665058257, "grad_norm": 1.8484669923782349, "learning_rate": 4.9244204330470544e-05, "loss": 2.3656, "step": 1371 }, { "epoch": 0.01233962756269871, "grad_norm": 1.6523222923278809, "learning_rate": 4.924236917573921e-05, "loss": 2.3523, "step": 1372 }, { "epoch": 0.01234862146033916, "grad_norm": 1.3153334856033325, "learning_rate": 4.924053183000865e-05, "loss": 2.3487, "step": 1373 }, { "epoch": 0.01235761535797961, "grad_norm": 2.817600727081299, "learning_rate": 4.923869229344494e-05, "loss": 2.3055, "step": 1374 }, { "epoch": 0.012366609255620062, "grad_norm": 1.8938976526260376, "learning_rate": 4.923685056621432e-05, "loss": 2.3513, "step": 1375 }, { "epoch": 0.012375603153260513, "grad_norm": 1.8110777139663696, "learning_rate": 4.923500664848326e-05, "loss": 2.3197, "step": 1376 }, { "epoch": 0.012384597050900963, "grad_norm": 1.3235276937484741, "learning_rate": 4.92331605404184e-05, "loss": 2.3093, "step": 1377 }, { "epoch": 0.012393590948541415, "grad_norm": 1.1013685464859009, "learning_rate": 4.92313122421866e-05, "loss": 2.3095, "step": 1378 }, { "epoch": 0.012402584846181865, "grad_norm": 3.155381679534912, "learning_rate": 4.922946175395489e-05, "loss": 2.2927, "step": 1379 }, { "epoch": 0.012411578743822316, "grad_norm": 1.7338447570800781, "learning_rate": 4.922760907589052e-05, "loss": 2.3123, "step": 1380 }, { "epoch": 0.012420572641462768, "grad_norm": 1.9885655641555786, "learning_rate": 4.922575420816095e-05, "loss": 2.374, "step": 1381 }, { "epoch": 0.012429566539103218, "grad_norm": 1.0699632167816162, "learning_rate": 4.9223897150933804e-05, "loss": 2.3657, "step": 1382 }, { "epoch": 0.012438560436743669, "grad_norm": 2.571831703186035, "learning_rate": 4.922203790437693e-05, "loss": 2.3649, "step": 1383 }, { "epoch": 0.01244755433438412, "grad_norm": 2.8510687351226807, "learning_rate": 4.922017646865835e-05, "loss": 2.3133, "step": 1384 }, { "epoch": 0.012456548232024571, "grad_norm": 2.8858659267425537, "learning_rate": 4.9218312843946314e-05, "loss": 2.3696, "step": 1385 }, { "epoch": 0.012465542129665023, "grad_norm": 2.153928518295288, "learning_rate": 4.921644703040925e-05, "loss": 2.2591, "step": 1386 }, { "epoch": 0.012474536027305473, "grad_norm": 1.840311884880066, "learning_rate": 4.9214579028215776e-05, "loss": 2.3336, "step": 1387 }, { "epoch": 0.012483529924945924, "grad_norm": 1.258995532989502, "learning_rate": 4.9212708837534746e-05, "loss": 2.2697, "step": 1388 }, { "epoch": 0.012492523822586376, "grad_norm": 1.5013561248779297, "learning_rate": 4.9210836458535165e-05, "loss": 2.3315, "step": 1389 }, { "epoch": 0.012501517720226826, "grad_norm": 2.9934911727905273, "learning_rate": 4.9208961891386265e-05, "loss": 2.3513, "step": 1390 }, { "epoch": 0.012510511617867277, "grad_norm": 1.2505568265914917, "learning_rate": 4.920708513625746e-05, "loss": 2.304, "step": 1391 }, { "epoch": 0.012519505515507729, "grad_norm": 3.9782907962799072, "learning_rate": 4.9205206193318374e-05, "loss": 2.3204, "step": 1392 }, { "epoch": 0.012528499413148179, "grad_norm": 4.457254409790039, "learning_rate": 4.920332506273883e-05, "loss": 2.2693, "step": 1393 }, { "epoch": 0.01253749331078863, "grad_norm": 1.9486520290374756, "learning_rate": 4.9201441744688835e-05, "loss": 2.298, "step": 1394 }, { "epoch": 0.012546487208429082, "grad_norm": 1.8635858297348022, "learning_rate": 4.91995562393386e-05, "loss": 2.296, "step": 1395 }, { "epoch": 0.012555481106069532, "grad_norm": 1.4830387830734253, "learning_rate": 4.9197668546858536e-05, "loss": 2.3269, "step": 1396 }, { "epoch": 0.012564475003709982, "grad_norm": 0.9352790117263794, "learning_rate": 4.919577866741926e-05, "loss": 2.288, "step": 1397 }, { "epoch": 0.012573468901350434, "grad_norm": 1.602815866470337, "learning_rate": 4.919388660119156e-05, "loss": 2.2405, "step": 1398 }, { "epoch": 0.012582462798990885, "grad_norm": 1.2547131776809692, "learning_rate": 4.919199234834646e-05, "loss": 2.2738, "step": 1399 }, { "epoch": 0.012591456696631335, "grad_norm": 1.011612057685852, "learning_rate": 4.919009590905514e-05, "loss": 2.3585, "step": 1400 }, { "epoch": 0.012600450594271787, "grad_norm": 0.7901346683502197, "learning_rate": 4.9188197283489015e-05, "loss": 2.2434, "step": 1401 }, { "epoch": 0.012609444491912238, "grad_norm": 1.0523431301116943, "learning_rate": 4.918629647181966e-05, "loss": 2.2917, "step": 1402 }, { "epoch": 0.012618438389552688, "grad_norm": 1.2329866886138916, "learning_rate": 4.9184393474218894e-05, "loss": 2.3502, "step": 1403 }, { "epoch": 0.01262743228719314, "grad_norm": 1.7527052164077759, "learning_rate": 4.9182488290858686e-05, "loss": 2.3325, "step": 1404 }, { "epoch": 0.01263642618483359, "grad_norm": 0.9112148284912109, "learning_rate": 4.9180580921911236e-05, "loss": 2.3672, "step": 1405 }, { "epoch": 0.01264542008247404, "grad_norm": 1.489206314086914, "learning_rate": 4.917867136754893e-05, "loss": 2.2949, "step": 1406 }, { "epoch": 0.012654413980114493, "grad_norm": 2.510047674179077, "learning_rate": 4.9176759627944346e-05, "loss": 2.2474, "step": 1407 }, { "epoch": 0.012663407877754943, "grad_norm": 1.6785236597061157, "learning_rate": 4.917484570327027e-05, "loss": 2.3413, "step": 1408 }, { "epoch": 0.012672401775395394, "grad_norm": 0.9878116846084595, "learning_rate": 4.917292959369968e-05, "loss": 2.2201, "step": 1409 }, { "epoch": 0.012681395673035846, "grad_norm": 1.2093920707702637, "learning_rate": 4.917101129940575e-05, "loss": 2.3082, "step": 1410 }, { "epoch": 0.012690389570676296, "grad_norm": 1.625248670578003, "learning_rate": 4.916909082056185e-05, "loss": 2.3316, "step": 1411 }, { "epoch": 0.012699383468316746, "grad_norm": 1.675307035446167, "learning_rate": 4.916716815734156e-05, "loss": 2.355, "step": 1412 }, { "epoch": 0.012708377365957198, "grad_norm": 1.44845449924469, "learning_rate": 4.9165243309918634e-05, "loss": 2.3005, "step": 1413 }, { "epoch": 0.012717371263597649, "grad_norm": 1.645935297012329, "learning_rate": 4.9163316278467056e-05, "loss": 2.3009, "step": 1414 }, { "epoch": 0.0127263651612381, "grad_norm": 1.3288215398788452, "learning_rate": 4.916138706316097e-05, "loss": 2.3164, "step": 1415 }, { "epoch": 0.012735359058878551, "grad_norm": 2.2171990871429443, "learning_rate": 4.9159455664174756e-05, "loss": 2.3462, "step": 1416 }, { "epoch": 0.012744352956519002, "grad_norm": 1.350297451019287, "learning_rate": 4.915752208168295e-05, "loss": 2.363, "step": 1417 }, { "epoch": 0.012753346854159452, "grad_norm": 0.9702437520027161, "learning_rate": 4.915558631586032e-05, "loss": 2.2286, "step": 1418 }, { "epoch": 0.012762340751799904, "grad_norm": 1.020673155784607, "learning_rate": 4.9153648366881826e-05, "loss": 2.2715, "step": 1419 }, { "epoch": 0.012771334649440354, "grad_norm": 0.8472902178764343, "learning_rate": 4.9151708234922603e-05, "loss": 2.2092, "step": 1420 }, { "epoch": 0.012780328547080807, "grad_norm": 1.295495867729187, "learning_rate": 4.9149765920158005e-05, "loss": 2.3591, "step": 1421 }, { "epoch": 0.012789322444721257, "grad_norm": 1.7508115768432617, "learning_rate": 4.914782142276357e-05, "loss": 2.3541, "step": 1422 }, { "epoch": 0.012798316342361707, "grad_norm": 1.0777674913406372, "learning_rate": 4.9145874742915056e-05, "loss": 2.3678, "step": 1423 }, { "epoch": 0.01280731024000216, "grad_norm": 1.720782995223999, "learning_rate": 4.9143925880788386e-05, "loss": 2.3476, "step": 1424 }, { "epoch": 0.01281630413764261, "grad_norm": 2.026348829269409, "learning_rate": 4.9141974836559704e-05, "loss": 2.343, "step": 1425 }, { "epoch": 0.01282529803528306, "grad_norm": 1.1215814352035522, "learning_rate": 4.9140021610405326e-05, "loss": 2.2349, "step": 1426 }, { "epoch": 0.012834291932923512, "grad_norm": 0.973241925239563, "learning_rate": 4.913806620250181e-05, "loss": 2.2499, "step": 1427 }, { "epoch": 0.012843285830563963, "grad_norm": 1.0042587518692017, "learning_rate": 4.9136108613025866e-05, "loss": 2.2721, "step": 1428 }, { "epoch": 0.012852279728204413, "grad_norm": 1.5908631086349487, "learning_rate": 4.913414884215442e-05, "loss": 2.2816, "step": 1429 }, { "epoch": 0.012861273625844865, "grad_norm": 1.480427622795105, "learning_rate": 4.9132186890064605e-05, "loss": 2.3384, "step": 1430 }, { "epoch": 0.012870267523485315, "grad_norm": 1.840626835823059, "learning_rate": 4.913022275693372e-05, "loss": 2.3776, "step": 1431 }, { "epoch": 0.012879261421125766, "grad_norm": 2.2271974086761475, "learning_rate": 4.91282564429393e-05, "loss": 2.2949, "step": 1432 }, { "epoch": 0.012888255318766218, "grad_norm": 1.788423776626587, "learning_rate": 4.9126287948259045e-05, "loss": 2.3391, "step": 1433 }, { "epoch": 0.012897249216406668, "grad_norm": 2.0390164852142334, "learning_rate": 4.912431727307088e-05, "loss": 2.3828, "step": 1434 }, { "epoch": 0.012906243114047119, "grad_norm": 1.7438021898269653, "learning_rate": 4.9122344417552904e-05, "loss": 2.3585, "step": 1435 }, { "epoch": 0.01291523701168757, "grad_norm": 1.5729365348815918, "learning_rate": 4.912036938188342e-05, "loss": 2.2889, "step": 1436 }, { "epoch": 0.012924230909328021, "grad_norm": 1.4151298999786377, "learning_rate": 4.911839216624093e-05, "loss": 2.2024, "step": 1437 }, { "epoch": 0.012933224806968471, "grad_norm": 1.3459993600845337, "learning_rate": 4.911641277080413e-05, "loss": 2.257, "step": 1438 }, { "epoch": 0.012942218704608923, "grad_norm": 1.3020594120025635, "learning_rate": 4.911443119575193e-05, "loss": 2.3676, "step": 1439 }, { "epoch": 0.012951212602249374, "grad_norm": 0.9443081021308899, "learning_rate": 4.911244744126341e-05, "loss": 2.2889, "step": 1440 }, { "epoch": 0.012960206499889824, "grad_norm": 1.1419777870178223, "learning_rate": 4.911046150751786e-05, "loss": 2.3973, "step": 1441 }, { "epoch": 0.012969200397530276, "grad_norm": 0.7631124258041382, "learning_rate": 4.910847339469477e-05, "loss": 2.1734, "step": 1442 }, { "epoch": 0.012978194295170727, "grad_norm": 1.3308863639831543, "learning_rate": 4.9106483102973824e-05, "loss": 2.3016, "step": 1443 }, { "epoch": 0.012987188192811177, "grad_norm": 0.6941406726837158, "learning_rate": 4.9104490632534894e-05, "loss": 2.2565, "step": 1444 }, { "epoch": 0.012996182090451629, "grad_norm": 1.0932955741882324, "learning_rate": 4.910249598355806e-05, "loss": 2.2987, "step": 1445 }, { "epoch": 0.01300517598809208, "grad_norm": 0.8946712613105774, "learning_rate": 4.910049915622361e-05, "loss": 2.2626, "step": 1446 }, { "epoch": 0.01301416988573253, "grad_norm": 1.3275916576385498, "learning_rate": 4.9098500150712e-05, "loss": 2.3549, "step": 1447 }, { "epoch": 0.013023163783372982, "grad_norm": 0.7752215266227722, "learning_rate": 4.9096498967203905e-05, "loss": 2.241, "step": 1448 }, { "epoch": 0.013032157681013432, "grad_norm": 1.0660349130630493, "learning_rate": 4.909449560588019e-05, "loss": 2.291, "step": 1449 }, { "epoch": 0.013041151578653883, "grad_norm": 1.3548780679702759, "learning_rate": 4.9092490066921916e-05, "loss": 2.3213, "step": 1450 }, { "epoch": 0.013050145476294335, "grad_norm": 1.3699947595596313, "learning_rate": 4.9090482350510336e-05, "loss": 2.2725, "step": 1451 }, { "epoch": 0.013059139373934785, "grad_norm": 1.3277685642242432, "learning_rate": 4.9088472456826904e-05, "loss": 2.2799, "step": 1452 }, { "epoch": 0.013068133271575235, "grad_norm": 1.296507716178894, "learning_rate": 4.908646038605329e-05, "loss": 2.3449, "step": 1453 }, { "epoch": 0.013077127169215687, "grad_norm": 1.3192994594573975, "learning_rate": 4.908444613837132e-05, "loss": 2.2762, "step": 1454 }, { "epoch": 0.013086121066856138, "grad_norm": 1.0131722688674927, "learning_rate": 4.908242971396305e-05, "loss": 2.2564, "step": 1455 }, { "epoch": 0.013095114964496588, "grad_norm": 1.6245476007461548, "learning_rate": 4.908041111301073e-05, "loss": 2.3634, "step": 1456 }, { "epoch": 0.01310410886213704, "grad_norm": 1.0569870471954346, "learning_rate": 4.907839033569679e-05, "loss": 2.3689, "step": 1457 }, { "epoch": 0.01311310275977749, "grad_norm": 0.6416109800338745, "learning_rate": 4.907636738220386e-05, "loss": 2.2693, "step": 1458 }, { "epoch": 0.013122096657417943, "grad_norm": 0.8165781497955322, "learning_rate": 4.907434225271478e-05, "loss": 2.2533, "step": 1459 }, { "epoch": 0.013131090555058393, "grad_norm": 1.4047127962112427, "learning_rate": 4.907231494741258e-05, "loss": 2.3631, "step": 1460 }, { "epoch": 0.013140084452698843, "grad_norm": 1.3052713871002197, "learning_rate": 4.907028546648048e-05, "loss": 2.2588, "step": 1461 }, { "epoch": 0.013149078350339296, "grad_norm": 1.8216019868850708, "learning_rate": 4.9068253810101915e-05, "loss": 2.3131, "step": 1462 }, { "epoch": 0.013158072247979746, "grad_norm": 1.5135239362716675, "learning_rate": 4.906621997846049e-05, "loss": 2.3277, "step": 1463 }, { "epoch": 0.013167066145620196, "grad_norm": 0.896512508392334, "learning_rate": 4.906418397174002e-05, "loss": 2.3493, "step": 1464 }, { "epoch": 0.013176060043260648, "grad_norm": 7.013914585113525, "learning_rate": 4.9062145790124526e-05, "loss": 2.2922, "step": 1465 }, { "epoch": 0.013185053940901099, "grad_norm": 1.9243741035461426, "learning_rate": 4.9060105433798215e-05, "loss": 2.3571, "step": 1466 }, { "epoch": 0.013194047838541549, "grad_norm": 1.5016390085220337, "learning_rate": 4.905806290294549e-05, "loss": 2.3823, "step": 1467 }, { "epoch": 0.013203041736182001, "grad_norm": 1.5999473333358765, "learning_rate": 4.905601819775095e-05, "loss": 2.3691, "step": 1468 }, { "epoch": 0.013212035633822452, "grad_norm": 1.295243263244629, "learning_rate": 4.90539713183994e-05, "loss": 2.2992, "step": 1469 }, { "epoch": 0.013221029531462902, "grad_norm": 2.173689842224121, "learning_rate": 4.9051922265075834e-05, "loss": 2.2334, "step": 1470 }, { "epoch": 0.013230023429103354, "grad_norm": 1.455386757850647, "learning_rate": 4.904987103796543e-05, "loss": 2.2295, "step": 1471 }, { "epoch": 0.013239017326743804, "grad_norm": 1.8496041297912598, "learning_rate": 4.904781763725359e-05, "loss": 2.3184, "step": 1472 }, { "epoch": 0.013248011224384255, "grad_norm": 0.9949283003807068, "learning_rate": 4.90457620631259e-05, "loss": 2.2462, "step": 1473 }, { "epoch": 0.013257005122024707, "grad_norm": 5.564322471618652, "learning_rate": 4.904370431576813e-05, "loss": 2.395, "step": 1474 }, { "epoch": 0.013265999019665157, "grad_norm": 1.4406638145446777, "learning_rate": 4.904164439536626e-05, "loss": 2.3453, "step": 1475 }, { "epoch": 0.013274992917305608, "grad_norm": 1.9769619703292847, "learning_rate": 4.9039582302106465e-05, "loss": 2.2512, "step": 1476 }, { "epoch": 0.01328398681494606, "grad_norm": 2.204035520553589, "learning_rate": 4.9037518036175126e-05, "loss": 2.3678, "step": 1477 }, { "epoch": 0.01329298071258651, "grad_norm": 0.815660834312439, "learning_rate": 4.903545159775879e-05, "loss": 2.2468, "step": 1478 }, { "epoch": 0.01330197461022696, "grad_norm": 5.241702556610107, "learning_rate": 4.903338298704422e-05, "loss": 2.3255, "step": 1479 }, { "epoch": 0.013310968507867412, "grad_norm": 1.594947099685669, "learning_rate": 4.9031312204218396e-05, "loss": 2.3213, "step": 1480 }, { "epoch": 0.013319962405507863, "grad_norm": 2.852376937866211, "learning_rate": 4.902923924946844e-05, "loss": 2.3545, "step": 1481 }, { "epoch": 0.013328956303148313, "grad_norm": 1.9794753789901733, "learning_rate": 4.902716412298174e-05, "loss": 2.296, "step": 1482 }, { "epoch": 0.013337950200788765, "grad_norm": 1.202060341835022, "learning_rate": 4.902508682494582e-05, "loss": 2.3212, "step": 1483 }, { "epoch": 0.013346944098429216, "grad_norm": 1.2876771688461304, "learning_rate": 4.9023007355548436e-05, "loss": 2.2472, "step": 1484 }, { "epoch": 0.013355937996069666, "grad_norm": 8.125530242919922, "learning_rate": 4.9020925714977514e-05, "loss": 2.4545, "step": 1485 }, { "epoch": 0.013364931893710118, "grad_norm": 2.8215906620025635, "learning_rate": 4.901884190342121e-05, "loss": 2.3704, "step": 1486 }, { "epoch": 0.013373925791350568, "grad_norm": 1.6175618171691895, "learning_rate": 4.901675592106783e-05, "loss": 2.3454, "step": 1487 }, { "epoch": 0.013382919688991019, "grad_norm": 1.2598589658737183, "learning_rate": 4.9014667768105936e-05, "loss": 2.3632, "step": 1488 }, { "epoch": 0.013391913586631471, "grad_norm": 1.0440917015075684, "learning_rate": 4.901257744472423e-05, "loss": 2.3011, "step": 1489 }, { "epoch": 0.013400907484271921, "grad_norm": 0.7207644581794739, "learning_rate": 4.901048495111164e-05, "loss": 2.2114, "step": 1490 }, { "epoch": 0.013409901381912372, "grad_norm": 1.503522515296936, "learning_rate": 4.9008390287457276e-05, "loss": 2.2791, "step": 1491 }, { "epoch": 0.013418895279552824, "grad_norm": 1.039432406425476, "learning_rate": 4.9006293453950455e-05, "loss": 2.2587, "step": 1492 }, { "epoch": 0.013427889177193274, "grad_norm": 1.23433518409729, "learning_rate": 4.900419445078069e-05, "loss": 2.322, "step": 1493 }, { "epoch": 0.013436883074833726, "grad_norm": 1.8497687578201294, "learning_rate": 4.900209327813769e-05, "loss": 2.2566, "step": 1494 }, { "epoch": 0.013445876972474177, "grad_norm": 1.0818395614624023, "learning_rate": 4.8999989936211356e-05, "loss": 2.2234, "step": 1495 }, { "epoch": 0.013454870870114627, "grad_norm": 1.1293376684188843, "learning_rate": 4.899788442519177e-05, "loss": 2.2535, "step": 1496 }, { "epoch": 0.013463864767755079, "grad_norm": 1.1049357652664185, "learning_rate": 4.8995776745269254e-05, "loss": 2.235, "step": 1497 }, { "epoch": 0.01347285866539553, "grad_norm": 1.1415754556655884, "learning_rate": 4.8993666896634274e-05, "loss": 2.3691, "step": 1498 }, { "epoch": 0.01348185256303598, "grad_norm": 1.2333590984344482, "learning_rate": 4.899155487947752e-05, "loss": 2.2423, "step": 1499 }, { "epoch": 0.013490846460676432, "grad_norm": 1.201111078262329, "learning_rate": 4.898944069398989e-05, "loss": 2.3714, "step": 1500 }, { "epoch": 0.013499840358316882, "grad_norm": 1.2897664308547974, "learning_rate": 4.898732434036244e-05, "loss": 2.322, "step": 1501 }, { "epoch": 0.013508834255957333, "grad_norm": 1.4544669389724731, "learning_rate": 4.898520581878645e-05, "loss": 2.3104, "step": 1502 }, { "epoch": 0.013517828153597785, "grad_norm": 1.377790093421936, "learning_rate": 4.89830851294534e-05, "loss": 2.268, "step": 1503 }, { "epoch": 0.013526822051238235, "grad_norm": 2.3503384590148926, "learning_rate": 4.898096227255494e-05, "loss": 2.3085, "step": 1504 }, { "epoch": 0.013535815948878685, "grad_norm": 1.3760664463043213, "learning_rate": 4.897883724828295e-05, "loss": 2.2564, "step": 1505 }, { "epoch": 0.013544809846519137, "grad_norm": 1.2119131088256836, "learning_rate": 4.8976710056829476e-05, "loss": 2.3232, "step": 1506 }, { "epoch": 0.013553803744159588, "grad_norm": 0.9780412912368774, "learning_rate": 4.8974580698386765e-05, "loss": 2.2853, "step": 1507 }, { "epoch": 0.013562797641800038, "grad_norm": 1.0036619901657104, "learning_rate": 4.8972449173147276e-05, "loss": 2.2166, "step": 1508 }, { "epoch": 0.01357179153944049, "grad_norm": 0.927261471748352, "learning_rate": 4.897031548130365e-05, "loss": 2.3161, "step": 1509 }, { "epoch": 0.01358078543708094, "grad_norm": 1.3687413930892944, "learning_rate": 4.896817962304874e-05, "loss": 2.3037, "step": 1510 }, { "epoch": 0.013589779334721391, "grad_norm": 0.7189801931381226, "learning_rate": 4.896604159857556e-05, "loss": 2.3202, "step": 1511 }, { "epoch": 0.013598773232361843, "grad_norm": 0.8859209418296814, "learning_rate": 4.8963901408077364e-05, "loss": 2.2335, "step": 1512 }, { "epoch": 0.013607767130002293, "grad_norm": 1.1698659658432007, "learning_rate": 4.896175905174757e-05, "loss": 2.3372, "step": 1513 }, { "epoch": 0.013616761027642744, "grad_norm": 1.1946779489517212, "learning_rate": 4.8959614529779796e-05, "loss": 2.2648, "step": 1514 }, { "epoch": 0.013625754925283196, "grad_norm": 1.2864960432052612, "learning_rate": 4.895746784236787e-05, "loss": 2.34, "step": 1515 }, { "epoch": 0.013634748822923646, "grad_norm": 1.3298262357711792, "learning_rate": 4.895531898970581e-05, "loss": 2.2857, "step": 1516 }, { "epoch": 0.013643742720564097, "grad_norm": 1.0850532054901123, "learning_rate": 4.895316797198781e-05, "loss": 2.3234, "step": 1517 }, { "epoch": 0.013652736618204549, "grad_norm": 1.1582186222076416, "learning_rate": 4.89510147894083e-05, "loss": 2.2708, "step": 1518 }, { "epoch": 0.013661730515844999, "grad_norm": 0.9520682692527771, "learning_rate": 4.8948859442161874e-05, "loss": 2.3413, "step": 1519 }, { "epoch": 0.01367072441348545, "grad_norm": 1.1450742483139038, "learning_rate": 4.894670193044332e-05, "loss": 2.3047, "step": 1520 }, { "epoch": 0.013679718311125902, "grad_norm": 1.4632195234298706, "learning_rate": 4.894454225444764e-05, "loss": 2.325, "step": 1521 }, { "epoch": 0.013688712208766352, "grad_norm": 1.5279765129089355, "learning_rate": 4.894238041437003e-05, "loss": 2.3494, "step": 1522 }, { "epoch": 0.013697706106406802, "grad_norm": 1.4543125629425049, "learning_rate": 4.894021641040585e-05, "loss": 2.3153, "step": 1523 }, { "epoch": 0.013706700004047254, "grad_norm": 1.0561600923538208, "learning_rate": 4.893805024275071e-05, "loss": 2.3385, "step": 1524 }, { "epoch": 0.013715693901687705, "grad_norm": 1.064774751663208, "learning_rate": 4.8935881911600366e-05, "loss": 2.2873, "step": 1525 }, { "epoch": 0.013724687799328155, "grad_norm": 1.2532635927200317, "learning_rate": 4.89337114171508e-05, "loss": 2.3004, "step": 1526 }, { "epoch": 0.013733681696968607, "grad_norm": 2.683222770690918, "learning_rate": 4.893153875959817e-05, "loss": 2.3142, "step": 1527 }, { "epoch": 0.013742675594609058, "grad_norm": 1.316571831703186, "learning_rate": 4.8929363939138854e-05, "loss": 2.3365, "step": 1528 }, { "epoch": 0.013751669492249508, "grad_norm": 1.5955365896224976, "learning_rate": 4.89271869559694e-05, "loss": 2.2927, "step": 1529 }, { "epoch": 0.01376066338988996, "grad_norm": 1.0240486860275269, "learning_rate": 4.8925007810286546e-05, "loss": 2.2553, "step": 1530 }, { "epoch": 0.01376965728753041, "grad_norm": 0.8262383937835693, "learning_rate": 4.8922826502287275e-05, "loss": 2.2031, "step": 1531 }, { "epoch": 0.013778651185170862, "grad_norm": 0.7110381722450256, "learning_rate": 4.8920643032168703e-05, "loss": 2.3212, "step": 1532 }, { "epoch": 0.013787645082811313, "grad_norm": 1.495509386062622, "learning_rate": 4.891845740012818e-05, "loss": 2.3231, "step": 1533 }, { "epoch": 0.013796638980451763, "grad_norm": 1.0226750373840332, "learning_rate": 4.891626960636323e-05, "loss": 2.2767, "step": 1534 }, { "epoch": 0.013805632878092215, "grad_norm": 0.8253571391105652, "learning_rate": 4.8914079651071606e-05, "loss": 2.2549, "step": 1535 }, { "epoch": 0.013814626775732666, "grad_norm": 1.2516210079193115, "learning_rate": 4.891188753445122e-05, "loss": 2.294, "step": 1536 }, { "epoch": 0.013823620673373116, "grad_norm": 0.818970263004303, "learning_rate": 4.89096932567002e-05, "loss": 2.2824, "step": 1537 }, { "epoch": 0.013832614571013568, "grad_norm": 0.9809755086898804, "learning_rate": 4.890749681801684e-05, "loss": 2.2402, "step": 1538 }, { "epoch": 0.013841608468654018, "grad_norm": 1.5677928924560547, "learning_rate": 4.890529821859968e-05, "loss": 2.3347, "step": 1539 }, { "epoch": 0.013850602366294469, "grad_norm": 0.8905024528503418, "learning_rate": 4.890309745864741e-05, "loss": 2.2615, "step": 1540 }, { "epoch": 0.01385959626393492, "grad_norm": 0.9016208052635193, "learning_rate": 4.8900894538358944e-05, "loss": 2.2615, "step": 1541 }, { "epoch": 0.013868590161575371, "grad_norm": 1.0534595251083374, "learning_rate": 4.8898689457933363e-05, "loss": 2.2869, "step": 1542 }, { "epoch": 0.013877584059215822, "grad_norm": 1.4061229228973389, "learning_rate": 4.889648221756999e-05, "loss": 2.3209, "step": 1543 }, { "epoch": 0.013886577956856274, "grad_norm": 1.2702196836471558, "learning_rate": 4.889427281746828e-05, "loss": 2.2897, "step": 1544 }, { "epoch": 0.013895571854496724, "grad_norm": 1.3807512521743774, "learning_rate": 4.8892061257827924e-05, "loss": 2.2773, "step": 1545 }, { "epoch": 0.013904565752137174, "grad_norm": 0.9990711808204651, "learning_rate": 4.888984753884882e-05, "loss": 2.2724, "step": 1546 }, { "epoch": 0.013913559649777626, "grad_norm": 1.0630393028259277, "learning_rate": 4.888763166073101e-05, "loss": 2.2663, "step": 1547 }, { "epoch": 0.013922553547418077, "grad_norm": 0.831275224685669, "learning_rate": 4.8885413623674804e-05, "loss": 2.3437, "step": 1548 }, { "epoch": 0.013931547445058527, "grad_norm": 1.5377510786056519, "learning_rate": 4.8883193427880625e-05, "loss": 2.2938, "step": 1549 }, { "epoch": 0.01394054134269898, "grad_norm": 0.9480425715446472, "learning_rate": 4.8880971073549154e-05, "loss": 2.3141, "step": 1550 }, { "epoch": 0.01394953524033943, "grad_norm": 1.0191166400909424, "learning_rate": 4.887874656088124e-05, "loss": 2.2693, "step": 1551 }, { "epoch": 0.01395852913797988, "grad_norm": 1.0550453662872314, "learning_rate": 4.8876519890077926e-05, "loss": 2.2977, "step": 1552 }, { "epoch": 0.013967523035620332, "grad_norm": 0.9050216674804688, "learning_rate": 4.887429106134047e-05, "loss": 2.279, "step": 1553 }, { "epoch": 0.013976516933260782, "grad_norm": 1.4496594667434692, "learning_rate": 4.88720600748703e-05, "loss": 2.3405, "step": 1554 }, { "epoch": 0.013985510830901233, "grad_norm": 1.1739283800125122, "learning_rate": 4.8869826930869045e-05, "loss": 2.2464, "step": 1555 }, { "epoch": 0.013994504728541685, "grad_norm": 0.8753737807273865, "learning_rate": 4.886759162953856e-05, "loss": 2.2698, "step": 1556 }, { "epoch": 0.014003498626182135, "grad_norm": 1.2392778396606445, "learning_rate": 4.886535417108084e-05, "loss": 2.2212, "step": 1557 }, { "epoch": 0.014012492523822586, "grad_norm": 0.8766874670982361, "learning_rate": 4.886311455569811e-05, "loss": 2.3099, "step": 1558 }, { "epoch": 0.014021486421463038, "grad_norm": 1.513805627822876, "learning_rate": 4.88608727835928e-05, "loss": 2.3002, "step": 1559 }, { "epoch": 0.014030480319103488, "grad_norm": 0.7318100929260254, "learning_rate": 4.88586288549675e-05, "loss": 2.2333, "step": 1560 }, { "epoch": 0.014039474216743938, "grad_norm": 1.0444992780685425, "learning_rate": 4.8856382770025034e-05, "loss": 2.3115, "step": 1561 }, { "epoch": 0.01404846811438439, "grad_norm": 1.5187288522720337, "learning_rate": 4.885413452896838e-05, "loss": 2.2705, "step": 1562 }, { "epoch": 0.014057462012024841, "grad_norm": 1.7314192056655884, "learning_rate": 4.885188413200075e-05, "loss": 2.2405, "step": 1563 }, { "epoch": 0.014066455909665291, "grad_norm": 1.0305639505386353, "learning_rate": 4.884963157932552e-05, "loss": 2.235, "step": 1564 }, { "epoch": 0.014075449807305743, "grad_norm": 1.9407317638397217, "learning_rate": 4.8847376871146265e-05, "loss": 2.331, "step": 1565 }, { "epoch": 0.014084443704946194, "grad_norm": 1.0195425748825073, "learning_rate": 4.884512000766679e-05, "loss": 2.3955, "step": 1566 }, { "epoch": 0.014093437602586644, "grad_norm": 0.9120759963989258, "learning_rate": 4.884286098909104e-05, "loss": 2.2323, "step": 1567 }, { "epoch": 0.014102431500227096, "grad_norm": 1.0566129684448242, "learning_rate": 4.884059981562321e-05, "loss": 2.2677, "step": 1568 }, { "epoch": 0.014111425397867547, "grad_norm": 1.2720448970794678, "learning_rate": 4.883833648746763e-05, "loss": 2.2322, "step": 1569 }, { "epoch": 0.014120419295507999, "grad_norm": 1.0273852348327637, "learning_rate": 4.8836071004828886e-05, "loss": 2.172, "step": 1570 }, { "epoch": 0.014129413193148449, "grad_norm": 1.0984374284744263, "learning_rate": 4.8833803367911717e-05, "loss": 2.2965, "step": 1571 }, { "epoch": 0.0141384070907889, "grad_norm": 1.053094506263733, "learning_rate": 4.883153357692107e-05, "loss": 2.3454, "step": 1572 }, { "epoch": 0.014147400988429351, "grad_norm": 1.8766710758209229, "learning_rate": 4.882926163206209e-05, "loss": 2.2171, "step": 1573 }, { "epoch": 0.014156394886069802, "grad_norm": 1.362017035484314, "learning_rate": 4.8826987533540114e-05, "loss": 2.3389, "step": 1574 }, { "epoch": 0.014165388783710252, "grad_norm": 1.2835819721221924, "learning_rate": 4.882471128156066e-05, "loss": 2.3169, "step": 1575 }, { "epoch": 0.014174382681350704, "grad_norm": 1.1079405546188354, "learning_rate": 4.882243287632947e-05, "loss": 2.3226, "step": 1576 }, { "epoch": 0.014183376578991155, "grad_norm": 1.3148959875106812, "learning_rate": 4.882015231805245e-05, "loss": 2.3007, "step": 1577 }, { "epoch": 0.014192370476631605, "grad_norm": 1.1894735097885132, "learning_rate": 4.881786960693572e-05, "loss": 2.3892, "step": 1578 }, { "epoch": 0.014201364374272057, "grad_norm": 1.3177655935287476, "learning_rate": 4.8815584743185593e-05, "loss": 2.2949, "step": 1579 }, { "epoch": 0.014210358271912507, "grad_norm": 1.26853346824646, "learning_rate": 4.881329772700857e-05, "loss": 2.3299, "step": 1580 }, { "epoch": 0.014219352169552958, "grad_norm": 1.016809344291687, "learning_rate": 4.881100855861134e-05, "loss": 2.3164, "step": 1581 }, { "epoch": 0.01422834606719341, "grad_norm": 0.8750474452972412, "learning_rate": 4.8808717238200815e-05, "loss": 2.3421, "step": 1582 }, { "epoch": 0.01423733996483386, "grad_norm": 1.335290789604187, "learning_rate": 4.880642376598407e-05, "loss": 2.2986, "step": 1583 }, { "epoch": 0.01424633386247431, "grad_norm": 1.4852817058563232, "learning_rate": 4.8804128142168374e-05, "loss": 2.2835, "step": 1584 }, { "epoch": 0.014255327760114763, "grad_norm": 0.7631269097328186, "learning_rate": 4.880183036696123e-05, "loss": 2.2489, "step": 1585 }, { "epoch": 0.014264321657755213, "grad_norm": 1.0256742238998413, "learning_rate": 4.8799530440570284e-05, "loss": 2.2861, "step": 1586 }, { "epoch": 0.014273315555395663, "grad_norm": 1.2452783584594727, "learning_rate": 4.8797228363203416e-05, "loss": 2.2486, "step": 1587 }, { "epoch": 0.014282309453036116, "grad_norm": 0.9505096077919006, "learning_rate": 4.879492413506869e-05, "loss": 2.2661, "step": 1588 }, { "epoch": 0.014291303350676566, "grad_norm": 1.1314697265625, "learning_rate": 4.879261775637434e-05, "loss": 2.402, "step": 1589 }, { "epoch": 0.014300297248317016, "grad_norm": 0.9363741278648376, "learning_rate": 4.879030922732882e-05, "loss": 2.2796, "step": 1590 }, { "epoch": 0.014309291145957468, "grad_norm": 0.8884981274604797, "learning_rate": 4.878799854814079e-05, "loss": 2.2422, "step": 1591 }, { "epoch": 0.014318285043597919, "grad_norm": 0.917630136013031, "learning_rate": 4.8785685719019064e-05, "loss": 2.2508, "step": 1592 }, { "epoch": 0.014327278941238369, "grad_norm": 1.1392261981964111, "learning_rate": 4.878337074017268e-05, "loss": 2.2778, "step": 1593 }, { "epoch": 0.014336272838878821, "grad_norm": 0.961719810962677, "learning_rate": 4.878105361181088e-05, "loss": 2.25, "step": 1594 }, { "epoch": 0.014345266736519272, "grad_norm": 1.5704357624053955, "learning_rate": 4.877873433414306e-05, "loss": 2.2166, "step": 1595 }, { "epoch": 0.014354260634159722, "grad_norm": 1.2725164890289307, "learning_rate": 4.877641290737884e-05, "loss": 2.2487, "step": 1596 }, { "epoch": 0.014363254531800174, "grad_norm": 1.314294457435608, "learning_rate": 4.8774089331728034e-05, "loss": 2.3055, "step": 1597 }, { "epoch": 0.014372248429440624, "grad_norm": 1.0970916748046875, "learning_rate": 4.877176360740064e-05, "loss": 2.2969, "step": 1598 }, { "epoch": 0.014381242327081075, "grad_norm": 0.7976924777030945, "learning_rate": 4.8769435734606864e-05, "loss": 2.3035, "step": 1599 }, { "epoch": 0.014390236224721527, "grad_norm": 1.0993620157241821, "learning_rate": 4.8767105713557085e-05, "loss": 2.302, "step": 1600 }, { "epoch": 0.014399230122361977, "grad_norm": 1.098041296005249, "learning_rate": 4.8764773544461886e-05, "loss": 2.3381, "step": 1601 }, { "epoch": 0.014408224020002428, "grad_norm": 0.9061229825019836, "learning_rate": 4.8762439227532056e-05, "loss": 2.2633, "step": 1602 }, { "epoch": 0.01441721791764288, "grad_norm": 0.6966328620910645, "learning_rate": 4.876010276297857e-05, "loss": 2.2389, "step": 1603 }, { "epoch": 0.01442621181528333, "grad_norm": 1.0640064477920532, "learning_rate": 4.8757764151012575e-05, "loss": 2.2952, "step": 1604 }, { "epoch": 0.014435205712923782, "grad_norm": 0.7718282341957092, "learning_rate": 4.875542339184546e-05, "loss": 2.2782, "step": 1605 }, { "epoch": 0.014444199610564232, "grad_norm": 1.0525380373001099, "learning_rate": 4.875308048568875e-05, "loss": 2.2899, "step": 1606 }, { "epoch": 0.014453193508204683, "grad_norm": 0.8518139123916626, "learning_rate": 4.875073543275422e-05, "loss": 2.1997, "step": 1607 }, { "epoch": 0.014462187405845135, "grad_norm": 0.6957027316093445, "learning_rate": 4.874838823325381e-05, "loss": 2.2464, "step": 1608 }, { "epoch": 0.014471181303485585, "grad_norm": 0.7280006408691406, "learning_rate": 4.874603888739964e-05, "loss": 2.1706, "step": 1609 }, { "epoch": 0.014480175201126036, "grad_norm": 1.2138913869857788, "learning_rate": 4.874368739540405e-05, "loss": 2.4041, "step": 1610 }, { "epoch": 0.014489169098766488, "grad_norm": 1.1543684005737305, "learning_rate": 4.874133375747957e-05, "loss": 2.2338, "step": 1611 }, { "epoch": 0.014498162996406938, "grad_norm": 1.113680124282837, "learning_rate": 4.873897797383893e-05, "loss": 2.2884, "step": 1612 }, { "epoch": 0.014507156894047388, "grad_norm": 1.1381075382232666, "learning_rate": 4.873662004469501e-05, "loss": 2.2889, "step": 1613 }, { "epoch": 0.01451615079168784, "grad_norm": 0.8906194567680359, "learning_rate": 4.8734259970260954e-05, "loss": 2.2815, "step": 1614 }, { "epoch": 0.014525144689328291, "grad_norm": 1.0186251401901245, "learning_rate": 4.8731897750750047e-05, "loss": 2.2484, "step": 1615 }, { "epoch": 0.014534138586968741, "grad_norm": 2.067753314971924, "learning_rate": 4.872953338637577e-05, "loss": 2.2734, "step": 1616 }, { "epoch": 0.014543132484609193, "grad_norm": 1.6547949314117432, "learning_rate": 4.872716687735183e-05, "loss": 2.3562, "step": 1617 }, { "epoch": 0.014552126382249644, "grad_norm": 1.2500650882720947, "learning_rate": 4.872479822389211e-05, "loss": 2.3027, "step": 1618 }, { "epoch": 0.014561120279890094, "grad_norm": 2.5716488361358643, "learning_rate": 4.8722427426210675e-05, "loss": 2.2621, "step": 1619 }, { "epoch": 0.014570114177530546, "grad_norm": 1.515515923500061, "learning_rate": 4.87200544845218e-05, "loss": 2.3118, "step": 1620 }, { "epoch": 0.014579108075170997, "grad_norm": 1.1795251369476318, "learning_rate": 4.871767939903996e-05, "loss": 2.3822, "step": 1621 }, { "epoch": 0.014588101972811447, "grad_norm": 1.2546466588974, "learning_rate": 4.871530216997979e-05, "loss": 2.3991, "step": 1622 }, { "epoch": 0.014597095870451899, "grad_norm": 1.5921902656555176, "learning_rate": 4.871292279755616e-05, "loss": 2.3393, "step": 1623 }, { "epoch": 0.01460608976809235, "grad_norm": 1.4096040725708008, "learning_rate": 4.8710541281984105e-05, "loss": 2.2775, "step": 1624 }, { "epoch": 0.0146150836657328, "grad_norm": 0.7142196297645569, "learning_rate": 4.870815762347887e-05, "loss": 2.2274, "step": 1625 }, { "epoch": 0.014624077563373252, "grad_norm": 1.1841425895690918, "learning_rate": 4.870577182225589e-05, "loss": 2.1523, "step": 1626 }, { "epoch": 0.014633071461013702, "grad_norm": 1.2558950185775757, "learning_rate": 4.870338387853078e-05, "loss": 2.3211, "step": 1627 }, { "epoch": 0.014642065358654152, "grad_norm": 0.8260988593101501, "learning_rate": 4.870099379251937e-05, "loss": 2.2266, "step": 1628 }, { "epoch": 0.014651059256294605, "grad_norm": 1.4167429208755493, "learning_rate": 4.8698601564437675e-05, "loss": 2.2658, "step": 1629 }, { "epoch": 0.014660053153935055, "grad_norm": 1.6624324321746826, "learning_rate": 4.869620719450189e-05, "loss": 2.2527, "step": 1630 }, { "epoch": 0.014669047051575505, "grad_norm": 1.6745294332504272, "learning_rate": 4.8693810682928423e-05, "loss": 2.2864, "step": 1631 }, { "epoch": 0.014678040949215957, "grad_norm": 1.4230668544769287, "learning_rate": 4.869141202993386e-05, "loss": 2.3429, "step": 1632 }, { "epoch": 0.014687034846856408, "grad_norm": 123.09016418457031, "learning_rate": 4.8689011235735006e-05, "loss": 2.3163, "step": 1633 }, { "epoch": 0.014696028744496858, "grad_norm": 66.58489990234375, "learning_rate": 4.868660830054883e-05, "loss": 2.3473, "step": 1634 }, { "epoch": 0.01470502264213731, "grad_norm": 1.6765671968460083, "learning_rate": 4.8684203224592506e-05, "loss": 2.312, "step": 1635 }, { "epoch": 0.01471401653977776, "grad_norm": 44.038726806640625, "learning_rate": 4.86817960080834e-05, "loss": 2.2916, "step": 1636 }, { "epoch": 0.014723010437418211, "grad_norm": 1.037361979484558, "learning_rate": 4.867938665123909e-05, "loss": 2.3141, "step": 1637 }, { "epoch": 0.014732004335058663, "grad_norm": 1.3186761140823364, "learning_rate": 4.8676975154277314e-05, "loss": 2.322, "step": 1638 }, { "epoch": 0.014740998232699113, "grad_norm": 1.5475441217422485, "learning_rate": 4.867456151741603e-05, "loss": 2.3099, "step": 1639 }, { "epoch": 0.014749992130339564, "grad_norm": 1.6584702730178833, "learning_rate": 4.8672145740873374e-05, "loss": 2.3032, "step": 1640 }, { "epoch": 0.014758986027980016, "grad_norm": 1.1423327922821045, "learning_rate": 4.866972782486768e-05, "loss": 2.3606, "step": 1641 }, { "epoch": 0.014767979925620466, "grad_norm": 1.4102569818496704, "learning_rate": 4.866730776961749e-05, "loss": 2.3171, "step": 1642 }, { "epoch": 0.014776973823260918, "grad_norm": 4.442851543426514, "learning_rate": 4.8664885575341506e-05, "loss": 2.3512, "step": 1643 }, { "epoch": 0.014785967720901369, "grad_norm": 1.1988860368728638, "learning_rate": 4.8662461242258664e-05, "loss": 2.3322, "step": 1644 }, { "epoch": 0.014794961618541819, "grad_norm": 1.2642006874084473, "learning_rate": 4.8660034770588046e-05, "loss": 2.2811, "step": 1645 }, { "epoch": 0.014803955516182271, "grad_norm": 0.9310919046401978, "learning_rate": 4.8657606160548986e-05, "loss": 2.3692, "step": 1646 }, { "epoch": 0.014812949413822721, "grad_norm": 1.4296855926513672, "learning_rate": 4.8655175412360954e-05, "loss": 2.3736, "step": 1647 }, { "epoch": 0.014821943311463172, "grad_norm": 1.0482155084609985, "learning_rate": 4.8652742526243644e-05, "loss": 2.3192, "step": 1648 }, { "epoch": 0.014830937209103624, "grad_norm": 0.7324535846710205, "learning_rate": 4.8650307502416945e-05, "loss": 2.2758, "step": 1649 }, { "epoch": 0.014839931106744074, "grad_norm": 1.776188850402832, "learning_rate": 4.864787034110093e-05, "loss": 2.2996, "step": 1650 }, { "epoch": 0.014848925004384525, "grad_norm": 1.8748433589935303, "learning_rate": 4.864543104251587e-05, "loss": 2.311, "step": 1651 }, { "epoch": 0.014857918902024977, "grad_norm": 0.8143200874328613, "learning_rate": 4.864298960688222e-05, "loss": 2.254, "step": 1652 }, { "epoch": 0.014866912799665427, "grad_norm": 1.3252731561660767, "learning_rate": 4.864054603442062e-05, "loss": 2.243, "step": 1653 }, { "epoch": 0.014875906697305877, "grad_norm": 1.9002904891967773, "learning_rate": 4.863810032535195e-05, "loss": 2.3136, "step": 1654 }, { "epoch": 0.01488490059494633, "grad_norm": 1.0166923999786377, "learning_rate": 4.863565247989723e-05, "loss": 2.3262, "step": 1655 }, { "epoch": 0.01489389449258678, "grad_norm": 1.0350232124328613, "learning_rate": 4.863320249827769e-05, "loss": 2.338, "step": 1656 }, { "epoch": 0.01490288839022723, "grad_norm": 1.132045865058899, "learning_rate": 4.863075038071477e-05, "loss": 2.2896, "step": 1657 }, { "epoch": 0.014911882287867682, "grad_norm": 1.8662331104278564, "learning_rate": 4.862829612743008e-05, "loss": 2.4228, "step": 1658 }, { "epoch": 0.014920876185508133, "grad_norm": 2.886127233505249, "learning_rate": 4.862583973864544e-05, "loss": 2.3359, "step": 1659 }, { "epoch": 0.014929870083148583, "grad_norm": 1.339038610458374, "learning_rate": 4.8623381214582855e-05, "loss": 2.3592, "step": 1660 }, { "epoch": 0.014938863980789035, "grad_norm": 1.4620894193649292, "learning_rate": 4.862092055546451e-05, "loss": 2.3012, "step": 1661 }, { "epoch": 0.014947857878429486, "grad_norm": 1.0699290037155151, "learning_rate": 4.861845776151281e-05, "loss": 2.3676, "step": 1662 }, { "epoch": 0.014956851776069936, "grad_norm": 1.0566192865371704, "learning_rate": 4.861599283295035e-05, "loss": 2.276, "step": 1663 }, { "epoch": 0.014965845673710388, "grad_norm": 1.0292271375656128, "learning_rate": 4.861352576999989e-05, "loss": 2.2916, "step": 1664 }, { "epoch": 0.014974839571350838, "grad_norm": 1.2903679609298706, "learning_rate": 4.86110565728844e-05, "loss": 2.3391, "step": 1665 }, { "epoch": 0.014983833468991289, "grad_norm": 1.051493763923645, "learning_rate": 4.8608585241827046e-05, "loss": 2.3237, "step": 1666 }, { "epoch": 0.01499282736663174, "grad_norm": 1.1401419639587402, "learning_rate": 4.8606111777051186e-05, "loss": 2.3291, "step": 1667 }, { "epoch": 0.015001821264272191, "grad_norm": 1.4332388639450073, "learning_rate": 4.860363617878037e-05, "loss": 2.262, "step": 1668 }, { "epoch": 0.015010815161912642, "grad_norm": 0.9544563889503479, "learning_rate": 4.8601158447238346e-05, "loss": 2.2888, "step": 1669 }, { "epoch": 0.015019809059553094, "grad_norm": 0.9660168290138245, "learning_rate": 4.859867858264905e-05, "loss": 2.2314, "step": 1670 }, { "epoch": 0.015028802957193544, "grad_norm": 0.9694017171859741, "learning_rate": 4.8596196585236595e-05, "loss": 2.2047, "step": 1671 }, { "epoch": 0.015037796854833994, "grad_norm": 1.5287903547286987, "learning_rate": 4.85937124552253e-05, "loss": 2.3677, "step": 1672 }, { "epoch": 0.015046790752474446, "grad_norm": 0.8434200286865234, "learning_rate": 4.8591226192839696e-05, "loss": 2.2631, "step": 1673 }, { "epoch": 0.015055784650114897, "grad_norm": 0.9062039852142334, "learning_rate": 4.8588737798304476e-05, "loss": 2.258, "step": 1674 }, { "epoch": 0.015064778547755347, "grad_norm": 1.3914508819580078, "learning_rate": 4.8586247271844546e-05, "loss": 2.306, "step": 1675 }, { "epoch": 0.0150737724453958, "grad_norm": 1.364774465560913, "learning_rate": 4.858375461368499e-05, "loss": 2.2833, "step": 1676 }, { "epoch": 0.01508276634303625, "grad_norm": 0.8096304535865784, "learning_rate": 4.858125982405108e-05, "loss": 2.2704, "step": 1677 }, { "epoch": 0.0150917602406767, "grad_norm": 0.9544671773910522, "learning_rate": 4.857876290316833e-05, "loss": 2.2346, "step": 1678 }, { "epoch": 0.015100754138317152, "grad_norm": 1.6853290796279907, "learning_rate": 4.857626385126237e-05, "loss": 2.3401, "step": 1679 }, { "epoch": 0.015109748035957602, "grad_norm": 1.2222957611083984, "learning_rate": 4.857376266855909e-05, "loss": 2.2968, "step": 1680 }, { "epoch": 0.015118741933598055, "grad_norm": 1.121512770652771, "learning_rate": 4.8571259355284515e-05, "loss": 2.3299, "step": 1681 }, { "epoch": 0.015127735831238505, "grad_norm": 1.0393519401550293, "learning_rate": 4.856875391166492e-05, "loss": 2.2729, "step": 1682 }, { "epoch": 0.015136729728878955, "grad_norm": 0.924287736415863, "learning_rate": 4.856624633792673e-05, "loss": 2.2261, "step": 1683 }, { "epoch": 0.015145723626519407, "grad_norm": 1.5494250059127808, "learning_rate": 4.856373663429657e-05, "loss": 2.2968, "step": 1684 }, { "epoch": 0.015154717524159858, "grad_norm": 1.256866455078125, "learning_rate": 4.856122480100129e-05, "loss": 2.3281, "step": 1685 }, { "epoch": 0.015163711421800308, "grad_norm": 2.0648908615112305, "learning_rate": 4.8558710838267885e-05, "loss": 2.2923, "step": 1686 }, { "epoch": 0.01517270531944076, "grad_norm": 5.36793851852417, "learning_rate": 4.855619474632357e-05, "loss": 2.3124, "step": 1687 }, { "epoch": 0.01518169921708121, "grad_norm": 0.9082876443862915, "learning_rate": 4.855367652539574e-05, "loss": 2.1946, "step": 1688 }, { "epoch": 0.015190693114721661, "grad_norm": 2.38844895362854, "learning_rate": 4.8551156175712e-05, "loss": 2.29, "step": 1689 }, { "epoch": 0.015199687012362113, "grad_norm": 3.0658109188079834, "learning_rate": 4.854863369750014e-05, "loss": 2.2714, "step": 1690 }, { "epoch": 0.015208680910002563, "grad_norm": 1.5102766752243042, "learning_rate": 4.854610909098812e-05, "loss": 2.2177, "step": 1691 }, { "epoch": 0.015217674807643014, "grad_norm": 0.9167364239692688, "learning_rate": 4.854358235640413e-05, "loss": 2.243, "step": 1692 }, { "epoch": 0.015226668705283466, "grad_norm": 1.907095193862915, "learning_rate": 4.8541053493976516e-05, "loss": 2.34, "step": 1693 }, { "epoch": 0.015235662602923916, "grad_norm": 0.6987553238868713, "learning_rate": 4.8538522503933846e-05, "loss": 2.2872, "step": 1694 }, { "epoch": 0.015244656500564367, "grad_norm": 0.8857932090759277, "learning_rate": 4.853598938650487e-05, "loss": 2.2709, "step": 1695 }, { "epoch": 0.015253650398204819, "grad_norm": 1.0220707654953003, "learning_rate": 4.853345414191852e-05, "loss": 2.3388, "step": 1696 }, { "epoch": 0.015262644295845269, "grad_norm": 0.9603571891784668, "learning_rate": 4.853091677040395e-05, "loss": 2.2508, "step": 1697 }, { "epoch": 0.01527163819348572, "grad_norm": 1.0661171674728394, "learning_rate": 4.852837727219044e-05, "loss": 2.3104, "step": 1698 }, { "epoch": 0.015280632091126171, "grad_norm": 1.0047944784164429, "learning_rate": 4.852583564750755e-05, "loss": 2.346, "step": 1699 }, { "epoch": 0.015289625988766622, "grad_norm": 1.9199333190917969, "learning_rate": 4.8523291896584977e-05, "loss": 2.3277, "step": 1700 }, { "epoch": 0.015298619886407072, "grad_norm": 1.2526764869689941, "learning_rate": 4.8520746019652605e-05, "loss": 2.2917, "step": 1701 }, { "epoch": 0.015307613784047524, "grad_norm": 1.4083847999572754, "learning_rate": 4.851819801694056e-05, "loss": 2.2945, "step": 1702 }, { "epoch": 0.015316607681687975, "grad_norm": 1.1006942987442017, "learning_rate": 4.85156478886791e-05, "loss": 2.2471, "step": 1703 }, { "epoch": 0.015325601579328425, "grad_norm": 0.9158415794372559, "learning_rate": 4.851309563509872e-05, "loss": 2.2687, "step": 1704 }, { "epoch": 0.015334595476968877, "grad_norm": 1.537262201309204, "learning_rate": 4.851054125643008e-05, "loss": 2.2725, "step": 1705 }, { "epoch": 0.015343589374609327, "grad_norm": 0.9294925332069397, "learning_rate": 4.850798475290403e-05, "loss": 2.3882, "step": 1706 }, { "epoch": 0.015352583272249778, "grad_norm": 0.7939320802688599, "learning_rate": 4.850542612475165e-05, "loss": 2.2375, "step": 1707 }, { "epoch": 0.01536157716989023, "grad_norm": 0.7283051013946533, "learning_rate": 4.8502865372204184e-05, "loss": 2.247, "step": 1708 }, { "epoch": 0.01537057106753068, "grad_norm": 1.1219340562820435, "learning_rate": 4.850030249549305e-05, "loss": 2.2645, "step": 1709 }, { "epoch": 0.01537956496517113, "grad_norm": 0.6406038403511047, "learning_rate": 4.849773749484989e-05, "loss": 2.264, "step": 1710 }, { "epoch": 0.015388558862811583, "grad_norm": 1.3027560710906982, "learning_rate": 4.849517037050652e-05, "loss": 2.2326, "step": 1711 }, { "epoch": 0.015397552760452033, "grad_norm": 1.0084408521652222, "learning_rate": 4.849260112269497e-05, "loss": 2.3088, "step": 1712 }, { "epoch": 0.015406546658092483, "grad_norm": 0.5874270796775818, "learning_rate": 4.849002975164743e-05, "loss": 2.2642, "step": 1713 }, { "epoch": 0.015415540555732936, "grad_norm": 1.212501883506775, "learning_rate": 4.8487456257596305e-05, "loss": 2.2844, "step": 1714 }, { "epoch": 0.015424534453373386, "grad_norm": 0.9356610774993896, "learning_rate": 4.8484880640774176e-05, "loss": 2.2396, "step": 1715 }, { "epoch": 0.015433528351013838, "grad_norm": 1.258927583694458, "learning_rate": 4.848230290141383e-05, "loss": 2.2819, "step": 1716 }, { "epoch": 0.015442522248654288, "grad_norm": 0.7848899960517883, "learning_rate": 4.8479723039748246e-05, "loss": 2.2881, "step": 1717 }, { "epoch": 0.015451516146294739, "grad_norm": 1.0405381917953491, "learning_rate": 4.847714105601059e-05, "loss": 2.2882, "step": 1718 }, { "epoch": 0.01546051004393519, "grad_norm": 0.8394889831542969, "learning_rate": 4.84745569504342e-05, "loss": 2.2942, "step": 1719 }, { "epoch": 0.015469503941575641, "grad_norm": 0.628038227558136, "learning_rate": 4.8471970723252645e-05, "loss": 2.3152, "step": 1720 }, { "epoch": 0.015478497839216092, "grad_norm": 1.528188943862915, "learning_rate": 4.846938237469966e-05, "loss": 2.345, "step": 1721 }, { "epoch": 0.015487491736856544, "grad_norm": 1.0436937808990479, "learning_rate": 4.8466791905009165e-05, "loss": 2.2968, "step": 1722 }, { "epoch": 0.015496485634496994, "grad_norm": 1.2217814922332764, "learning_rate": 4.84641993144153e-05, "loss": 2.3549, "step": 1723 }, { "epoch": 0.015505479532137444, "grad_norm": 1.5608245134353638, "learning_rate": 4.8461604603152376e-05, "loss": 2.3443, "step": 1724 }, { "epoch": 0.015514473429777896, "grad_norm": 0.9040743112564087, "learning_rate": 4.84590077714549e-05, "loss": 2.2734, "step": 1725 }, { "epoch": 0.015523467327418347, "grad_norm": 1.0167921781539917, "learning_rate": 4.8456408819557564e-05, "loss": 2.2381, "step": 1726 }, { "epoch": 0.015532461225058797, "grad_norm": 0.7634522318840027, "learning_rate": 4.845380774769527e-05, "loss": 2.2552, "step": 1727 }, { "epoch": 0.01554145512269925, "grad_norm": 1.1253379583358765, "learning_rate": 4.845120455610309e-05, "loss": 2.3381, "step": 1728 }, { "epoch": 0.0155504490203397, "grad_norm": 1.342286229133606, "learning_rate": 4.84485992450163e-05, "loss": 2.2358, "step": 1729 }, { "epoch": 0.01555944291798015, "grad_norm": 1.160684585571289, "learning_rate": 4.844599181467037e-05, "loss": 2.3146, "step": 1730 }, { "epoch": 0.015568436815620602, "grad_norm": 0.8372664451599121, "learning_rate": 4.844338226530095e-05, "loss": 2.2624, "step": 1731 }, { "epoch": 0.015577430713261052, "grad_norm": 0.6305686235427856, "learning_rate": 4.8440770597143906e-05, "loss": 2.2075, "step": 1732 }, { "epoch": 0.015586424610901503, "grad_norm": 1.278459072113037, "learning_rate": 4.843815681043525e-05, "loss": 2.2647, "step": 1733 }, { "epoch": 0.015595418508541955, "grad_norm": 1.4778318405151367, "learning_rate": 4.843554090541123e-05, "loss": 2.2812, "step": 1734 }, { "epoch": 0.015604412406182405, "grad_norm": 0.843654215335846, "learning_rate": 4.843292288230826e-05, "loss": 2.2355, "step": 1735 }, { "epoch": 0.015613406303822856, "grad_norm": 1.2035799026489258, "learning_rate": 4.843030274136297e-05, "loss": 2.2931, "step": 1736 }, { "epoch": 0.015622400201463308, "grad_norm": 0.9506595134735107, "learning_rate": 4.842768048281216e-05, "loss": 2.2913, "step": 1737 }, { "epoch": 0.01563139409910376, "grad_norm": 1.4588580131530762, "learning_rate": 4.8425056106892805e-05, "loss": 2.3381, "step": 1738 }, { "epoch": 0.01564038799674421, "grad_norm": 1.281014084815979, "learning_rate": 4.842242961384211e-05, "loss": 2.2844, "step": 1739 }, { "epoch": 0.01564938189438466, "grad_norm": 1.2635339498519897, "learning_rate": 4.841980100389747e-05, "loss": 2.3184, "step": 1740 }, { "epoch": 0.01565837579202511, "grad_norm": 0.9929319024085999, "learning_rate": 4.841717027729643e-05, "loss": 2.3244, "step": 1741 }, { "epoch": 0.01566736968966556, "grad_norm": 0.9049252867698669, "learning_rate": 4.841453743427678e-05, "loss": 2.3104, "step": 1742 }, { "epoch": 0.01567636358730601, "grad_norm": 1.3217833042144775, "learning_rate": 4.8411902475076435e-05, "loss": 2.3355, "step": 1743 }, { "epoch": 0.015685357484946465, "grad_norm": 1.3570677042007446, "learning_rate": 4.8409265399933574e-05, "loss": 2.2917, "step": 1744 }, { "epoch": 0.015694351382586916, "grad_norm": 1.055814266204834, "learning_rate": 4.8406626209086515e-05, "loss": 2.3087, "step": 1745 }, { "epoch": 0.015703345280227366, "grad_norm": 0.8690663576126099, "learning_rate": 4.8403984902773796e-05, "loss": 2.3487, "step": 1746 }, { "epoch": 0.015712339177867816, "grad_norm": 1.463718056678772, "learning_rate": 4.840134148123413e-05, "loss": 2.3123, "step": 1747 }, { "epoch": 0.015721333075508267, "grad_norm": 1.0361788272857666, "learning_rate": 4.839869594470642e-05, "loss": 2.2572, "step": 1748 }, { "epoch": 0.015730326973148717, "grad_norm": 1.4331319332122803, "learning_rate": 4.839604829342978e-05, "loss": 2.25, "step": 1749 }, { "epoch": 0.01573932087078917, "grad_norm": 1.7820483446121216, "learning_rate": 4.83933985276435e-05, "loss": 2.3103, "step": 1750 }, { "epoch": 0.01574831476842962, "grad_norm": 1.5386706590652466, "learning_rate": 4.839074664758704e-05, "loss": 2.2559, "step": 1751 }, { "epoch": 0.015757308666070072, "grad_norm": 1.9421716928482056, "learning_rate": 4.8388092653500105e-05, "loss": 2.3133, "step": 1752 }, { "epoch": 0.015766302563710522, "grad_norm": 0.7473291754722595, "learning_rate": 4.838543654562255e-05, "loss": 2.2426, "step": 1753 }, { "epoch": 0.015775296461350972, "grad_norm": 2.3026628494262695, "learning_rate": 4.838277832419442e-05, "loss": 2.4159, "step": 1754 }, { "epoch": 0.015784290358991423, "grad_norm": 1.0959279537200928, "learning_rate": 4.838011798945598e-05, "loss": 2.2783, "step": 1755 }, { "epoch": 0.015793284256631877, "grad_norm": 1.1502763032913208, "learning_rate": 4.837745554164765e-05, "loss": 2.2898, "step": 1756 }, { "epoch": 0.015802278154272327, "grad_norm": 0.7148979902267456, "learning_rate": 4.8374790981010075e-05, "loss": 2.1665, "step": 1757 }, { "epoch": 0.015811272051912777, "grad_norm": 1.093146562576294, "learning_rate": 4.8372124307784075e-05, "loss": 2.2296, "step": 1758 }, { "epoch": 0.015820265949553228, "grad_norm": 0.8140142560005188, "learning_rate": 4.836945552221065e-05, "loss": 2.2643, "step": 1759 }, { "epoch": 0.015829259847193678, "grad_norm": 0.9954135417938232, "learning_rate": 4.8366784624531e-05, "loss": 2.2318, "step": 1760 }, { "epoch": 0.01583825374483413, "grad_norm": 1.4276254177093506, "learning_rate": 4.8364111614986527e-05, "loss": 2.2922, "step": 1761 }, { "epoch": 0.015847247642474582, "grad_norm": 0.871088445186615, "learning_rate": 4.836143649381881e-05, "loss": 2.3666, "step": 1762 }, { "epoch": 0.015856241540115033, "grad_norm": 1.3774094581604004, "learning_rate": 4.8358759261269626e-05, "loss": 2.3687, "step": 1763 }, { "epoch": 0.015865235437755483, "grad_norm": 0.8023905754089355, "learning_rate": 4.8356079917580944e-05, "loss": 2.2467, "step": 1764 }, { "epoch": 0.015874229335395933, "grad_norm": 1.524170994758606, "learning_rate": 4.8353398462994916e-05, "loss": 2.3557, "step": 1765 }, { "epoch": 0.015883223233036384, "grad_norm": 1.3545712232589722, "learning_rate": 4.835071489775388e-05, "loss": 2.2675, "step": 1766 }, { "epoch": 0.015892217130676834, "grad_norm": 0.9542168974876404, "learning_rate": 4.834802922210039e-05, "loss": 2.2941, "step": 1767 }, { "epoch": 0.015901211028317288, "grad_norm": 1.1913856267929077, "learning_rate": 4.834534143627717e-05, "loss": 2.2461, "step": 1768 }, { "epoch": 0.01591020492595774, "grad_norm": 0.9957913160324097, "learning_rate": 4.8342651540527136e-05, "loss": 2.2682, "step": 1769 }, { "epoch": 0.01591919882359819, "grad_norm": 1.2226651906967163, "learning_rate": 4.8339959535093395e-05, "loss": 2.3396, "step": 1770 }, { "epoch": 0.01592819272123864, "grad_norm": 2.445849657058716, "learning_rate": 4.8337265420219245e-05, "loss": 2.2851, "step": 1771 }, { "epoch": 0.01593718661887909, "grad_norm": 2.062366247177124, "learning_rate": 4.83345691961482e-05, "loss": 2.2925, "step": 1772 }, { "epoch": 0.015946180516519543, "grad_norm": 1.8406349420547485, "learning_rate": 4.8331870863123905e-05, "loss": 2.2293, "step": 1773 }, { "epoch": 0.015955174414159994, "grad_norm": 0.8395392894744873, "learning_rate": 4.832917042139027e-05, "loss": 2.3894, "step": 1774 }, { "epoch": 0.015964168311800444, "grad_norm": 1.0644644498825073, "learning_rate": 4.8326467871191325e-05, "loss": 2.3345, "step": 1775 }, { "epoch": 0.015973162209440894, "grad_norm": 0.9206362962722778, "learning_rate": 4.8323763212771354e-05, "loss": 2.2374, "step": 1776 }, { "epoch": 0.015982156107081345, "grad_norm": 0.7812778353691101, "learning_rate": 4.832105644637478e-05, "loss": 2.254, "step": 1777 }, { "epoch": 0.015991150004721795, "grad_norm": 1.4544132947921753, "learning_rate": 4.831834757224624e-05, "loss": 2.2713, "step": 1778 }, { "epoch": 0.01600014390236225, "grad_norm": 1.4938353300094604, "learning_rate": 4.8315636590630566e-05, "loss": 2.3449, "step": 1779 }, { "epoch": 0.0160091378000027, "grad_norm": 0.861350953578949, "learning_rate": 4.831292350177277e-05, "loss": 2.2377, "step": 1780 }, { "epoch": 0.01601813169764315, "grad_norm": 0.9754862785339355, "learning_rate": 4.831020830591806e-05, "loss": 2.3128, "step": 1781 }, { "epoch": 0.0160271255952836, "grad_norm": 0.8671904802322388, "learning_rate": 4.830749100331183e-05, "loss": 2.2829, "step": 1782 }, { "epoch": 0.01603611949292405, "grad_norm": 1.046902060508728, "learning_rate": 4.830477159419966e-05, "loss": 2.2218, "step": 1783 }, { "epoch": 0.0160451133905645, "grad_norm": 1.9188002347946167, "learning_rate": 4.830205007882735e-05, "loss": 2.2688, "step": 1784 }, { "epoch": 0.016054107288204954, "grad_norm": 1.4355274438858032, "learning_rate": 4.829932645744084e-05, "loss": 2.3428, "step": 1785 }, { "epoch": 0.016063101185845405, "grad_norm": 1.168654441833496, "learning_rate": 4.829660073028631e-05, "loss": 2.2224, "step": 1786 }, { "epoch": 0.016072095083485855, "grad_norm": 1.3810064792633057, "learning_rate": 4.8293872897610096e-05, "loss": 2.2886, "step": 1787 }, { "epoch": 0.016081088981126306, "grad_norm": 1.2109360694885254, "learning_rate": 4.829114295965874e-05, "loss": 2.2377, "step": 1788 }, { "epoch": 0.016090082878766756, "grad_norm": 1.4794343709945679, "learning_rate": 4.828841091667897e-05, "loss": 2.2686, "step": 1789 }, { "epoch": 0.016099076776407206, "grad_norm": 1.6576998233795166, "learning_rate": 4.8285676768917704e-05, "loss": 2.2041, "step": 1790 }, { "epoch": 0.01610807067404766, "grad_norm": 1.1506435871124268, "learning_rate": 4.828294051662206e-05, "loss": 2.2795, "step": 1791 }, { "epoch": 0.01611706457168811, "grad_norm": 1.8207429647445679, "learning_rate": 4.828020216003932e-05, "loss": 2.2831, "step": 1792 }, { "epoch": 0.01612605846932856, "grad_norm": 1.3562047481536865, "learning_rate": 4.827746169941698e-05, "loss": 2.2682, "step": 1793 }, { "epoch": 0.01613505236696901, "grad_norm": 1.6682422161102295, "learning_rate": 4.827471913500274e-05, "loss": 2.3876, "step": 1794 }, { "epoch": 0.01614404626460946, "grad_norm": 1.1676632165908813, "learning_rate": 4.827197446704444e-05, "loss": 2.2827, "step": 1795 }, { "epoch": 0.016153040162249912, "grad_norm": 0.7973687052726746, "learning_rate": 4.826922769579016e-05, "loss": 2.2901, "step": 1796 }, { "epoch": 0.016162034059890366, "grad_norm": 1.4986774921417236, "learning_rate": 4.826647882148815e-05, "loss": 2.2922, "step": 1797 }, { "epoch": 0.016171027957530816, "grad_norm": 1.4185477495193481, "learning_rate": 4.826372784438684e-05, "loss": 2.3027, "step": 1798 }, { "epoch": 0.016180021855171266, "grad_norm": 1.560782790184021, "learning_rate": 4.826097476473486e-05, "loss": 2.3667, "step": 1799 }, { "epoch": 0.016189015752811717, "grad_norm": 1.1071298122406006, "learning_rate": 4.8258219582781046e-05, "loss": 2.2883, "step": 1800 }, { "epoch": 0.016198009650452167, "grad_norm": 1.3657180070877075, "learning_rate": 4.825546229877439e-05, "loss": 2.3031, "step": 1801 }, { "epoch": 0.016207003548092618, "grad_norm": 1.2129443883895874, "learning_rate": 4.82527029129641e-05, "loss": 2.2703, "step": 1802 }, { "epoch": 0.01621599744573307, "grad_norm": 0.9238573312759399, "learning_rate": 4.824994142559957e-05, "loss": 2.2827, "step": 1803 }, { "epoch": 0.01622499134337352, "grad_norm": 1.260115146636963, "learning_rate": 4.824717783693038e-05, "loss": 2.3157, "step": 1804 }, { "epoch": 0.016233985241013972, "grad_norm": 1.4081841707229614, "learning_rate": 4.8244412147206284e-05, "loss": 2.2942, "step": 1805 }, { "epoch": 0.016242979138654422, "grad_norm": 0.9430930614471436, "learning_rate": 4.824164435667727e-05, "loss": 2.2714, "step": 1806 }, { "epoch": 0.016251973036294873, "grad_norm": 1.180860996246338, "learning_rate": 4.8238874465593464e-05, "loss": 2.319, "step": 1807 }, { "epoch": 0.016260966933935327, "grad_norm": 1.352108359336853, "learning_rate": 4.823610247420521e-05, "loss": 2.3218, "step": 1808 }, { "epoch": 0.016269960831575777, "grad_norm": 2.0896854400634766, "learning_rate": 4.8233328382763057e-05, "loss": 2.3172, "step": 1809 }, { "epoch": 0.016278954729216227, "grad_norm": 1.1523449420928955, "learning_rate": 4.8230552191517694e-05, "loss": 2.2646, "step": 1810 }, { "epoch": 0.016287948626856678, "grad_norm": 1.7392079830169678, "learning_rate": 4.8227773900720056e-05, "loss": 2.2986, "step": 1811 }, { "epoch": 0.016296942524497128, "grad_norm": 0.9787827134132385, "learning_rate": 4.822499351062123e-05, "loss": 2.2402, "step": 1812 }, { "epoch": 0.01630593642213758, "grad_norm": 1.443446397781372, "learning_rate": 4.8222211021472506e-05, "loss": 2.3217, "step": 1813 }, { "epoch": 0.016314930319778032, "grad_norm": 0.7974103689193726, "learning_rate": 4.821942643352536e-05, "loss": 2.2411, "step": 1814 }, { "epoch": 0.016323924217418483, "grad_norm": 1.7783901691436768, "learning_rate": 4.821663974703147e-05, "loss": 2.4581, "step": 1815 }, { "epoch": 0.016332918115058933, "grad_norm": 0.8028185963630676, "learning_rate": 4.821385096224268e-05, "loss": 2.2644, "step": 1816 }, { "epoch": 0.016341912012699383, "grad_norm": 0.8400540351867676, "learning_rate": 4.821106007941105e-05, "loss": 2.2547, "step": 1817 }, { "epoch": 0.016350905910339834, "grad_norm": 1.2375069856643677, "learning_rate": 4.820826709878881e-05, "loss": 2.2746, "step": 1818 }, { "epoch": 0.016359899807980284, "grad_norm": 0.670217752456665, "learning_rate": 4.820547202062837e-05, "loss": 2.2338, "step": 1819 }, { "epoch": 0.016368893705620738, "grad_norm": 1.0312367677688599, "learning_rate": 4.8202674845182384e-05, "loss": 2.2935, "step": 1820 }, { "epoch": 0.016377887603261188, "grad_norm": 0.7886951565742493, "learning_rate": 4.819987557270363e-05, "loss": 2.1898, "step": 1821 }, { "epoch": 0.01638688150090164, "grad_norm": 0.848126232624054, "learning_rate": 4.819707420344511e-05, "loss": 2.2792, "step": 1822 }, { "epoch": 0.01639587539854209, "grad_norm": 1.2181836366653442, "learning_rate": 4.819427073766001e-05, "loss": 2.298, "step": 1823 }, { "epoch": 0.01640486929618254, "grad_norm": 1.0177326202392578, "learning_rate": 4.8191465175601706e-05, "loss": 2.3741, "step": 1824 }, { "epoch": 0.01641386319382299, "grad_norm": 1.1135579347610474, "learning_rate": 4.818865751752376e-05, "loss": 2.3023, "step": 1825 }, { "epoch": 0.016422857091463443, "grad_norm": 0.6524264216423035, "learning_rate": 4.818584776367993e-05, "loss": 2.2265, "step": 1826 }, { "epoch": 0.016431850989103894, "grad_norm": 1.0768200159072876, "learning_rate": 4.8183035914324136e-05, "loss": 2.2416, "step": 1827 }, { "epoch": 0.016440844886744344, "grad_norm": 1.1255426406860352, "learning_rate": 4.818022196971054e-05, "loss": 2.3167, "step": 1828 }, { "epoch": 0.016449838784384795, "grad_norm": 0.659188985824585, "learning_rate": 4.817740593009344e-05, "loss": 2.2551, "step": 1829 }, { "epoch": 0.016458832682025245, "grad_norm": 1.033416748046875, "learning_rate": 4.8174587795727364e-05, "loss": 2.2822, "step": 1830 }, { "epoch": 0.016467826579665695, "grad_norm": 1.2327380180358887, "learning_rate": 4.8171767566867e-05, "loss": 2.2626, "step": 1831 }, { "epoch": 0.01647682047730615, "grad_norm": 1.0146552324295044, "learning_rate": 4.816894524376725e-05, "loss": 2.3406, "step": 1832 }, { "epoch": 0.0164858143749466, "grad_norm": 0.994263231754303, "learning_rate": 4.816612082668318e-05, "loss": 2.2253, "step": 1833 }, { "epoch": 0.01649480827258705, "grad_norm": 0.9802228808403015, "learning_rate": 4.816329431587006e-05, "loss": 2.2665, "step": 1834 }, { "epoch": 0.0165038021702275, "grad_norm": 0.693946897983551, "learning_rate": 4.816046571158335e-05, "loss": 2.2454, "step": 1835 }, { "epoch": 0.01651279606786795, "grad_norm": 1.3491873741149902, "learning_rate": 4.8157635014078694e-05, "loss": 2.2501, "step": 1836 }, { "epoch": 0.0165217899655084, "grad_norm": 1.929073691368103, "learning_rate": 4.8154802223611935e-05, "loss": 2.369, "step": 1837 }, { "epoch": 0.016530783863148855, "grad_norm": 1.6979625225067139, "learning_rate": 4.815196734043909e-05, "loss": 2.3639, "step": 1838 }, { "epoch": 0.016539777760789305, "grad_norm": 1.1496517658233643, "learning_rate": 4.8149130364816364e-05, "loss": 2.2529, "step": 1839 }, { "epoch": 0.016548771658429755, "grad_norm": 1.0107629299163818, "learning_rate": 4.8146291297000176e-05, "loss": 2.3402, "step": 1840 }, { "epoch": 0.016557765556070206, "grad_norm": 1.1265355348587036, "learning_rate": 4.8143450137247114e-05, "loss": 2.3118, "step": 1841 }, { "epoch": 0.016566759453710656, "grad_norm": 1.4963544607162476, "learning_rate": 4.814060688581396e-05, "loss": 2.2731, "step": 1842 }, { "epoch": 0.016575753351351107, "grad_norm": 1.1520501375198364, "learning_rate": 4.813776154295767e-05, "loss": 2.2518, "step": 1843 }, { "epoch": 0.01658474724899156, "grad_norm": 1.1177328824996948, "learning_rate": 4.813491410893543e-05, "loss": 2.1875, "step": 1844 }, { "epoch": 0.01659374114663201, "grad_norm": 0.7695892453193665, "learning_rate": 4.813206458400455e-05, "loss": 2.2137, "step": 1845 }, { "epoch": 0.01660273504427246, "grad_norm": 1.1445571184158325, "learning_rate": 4.812921296842261e-05, "loss": 2.2929, "step": 1846 }, { "epoch": 0.01661172894191291, "grad_norm": 0.9211652874946594, "learning_rate": 4.81263592624473e-05, "loss": 2.2681, "step": 1847 }, { "epoch": 0.016620722839553362, "grad_norm": 1.8441261053085327, "learning_rate": 4.812350346633655e-05, "loss": 2.3464, "step": 1848 }, { "epoch": 0.016629716737193816, "grad_norm": 0.8590016961097717, "learning_rate": 4.812064558034847e-05, "loss": 2.2257, "step": 1849 }, { "epoch": 0.016638710634834266, "grad_norm": 0.8752726912498474, "learning_rate": 4.811778560474135e-05, "loss": 2.231, "step": 1850 }, { "epoch": 0.016647704532474716, "grad_norm": 1.0980899333953857, "learning_rate": 4.811492353977366e-05, "loss": 2.3283, "step": 1851 }, { "epoch": 0.016656698430115167, "grad_norm": 1.6323179006576538, "learning_rate": 4.811205938570408e-05, "loss": 2.1941, "step": 1852 }, { "epoch": 0.016665692327755617, "grad_norm": 1.0961986780166626, "learning_rate": 4.810919314279146e-05, "loss": 2.2665, "step": 1853 }, { "epoch": 0.016674686225396067, "grad_norm": 0.9938058853149414, "learning_rate": 4.810632481129487e-05, "loss": 2.281, "step": 1854 }, { "epoch": 0.01668368012303652, "grad_norm": 0.6960379481315613, "learning_rate": 4.8103454391473514e-05, "loss": 2.2964, "step": 1855 }, { "epoch": 0.01669267402067697, "grad_norm": 0.9237589240074158, "learning_rate": 4.810058188358685e-05, "loss": 2.3102, "step": 1856 }, { "epoch": 0.016701667918317422, "grad_norm": 0.6621735692024231, "learning_rate": 4.809770728789447e-05, "loss": 2.2571, "step": 1857 }, { "epoch": 0.016710661815957872, "grad_norm": 0.8308791518211365, "learning_rate": 4.8094830604656185e-05, "loss": 2.3512, "step": 1858 }, { "epoch": 0.016719655713598323, "grad_norm": 1.0479060411453247, "learning_rate": 4.8091951834131984e-05, "loss": 2.3564, "step": 1859 }, { "epoch": 0.016728649611238773, "grad_norm": 1.0854172706604004, "learning_rate": 4.8089070976582054e-05, "loss": 2.2815, "step": 1860 }, { "epoch": 0.016737643508879227, "grad_norm": 1.0308998823165894, "learning_rate": 4.808618803226675e-05, "loss": 2.2354, "step": 1861 }, { "epoch": 0.016746637406519677, "grad_norm": 1.7791664600372314, "learning_rate": 4.808330300144665e-05, "loss": 2.2741, "step": 1862 }, { "epoch": 0.016755631304160128, "grad_norm": 1.4389320611953735, "learning_rate": 4.8080415884382474e-05, "loss": 2.2824, "step": 1863 }, { "epoch": 0.016764625201800578, "grad_norm": 1.377921223640442, "learning_rate": 4.8077526681335176e-05, "loss": 2.2444, "step": 1864 }, { "epoch": 0.01677361909944103, "grad_norm": 0.8582924008369446, "learning_rate": 4.807463539256588e-05, "loss": 2.2474, "step": 1865 }, { "epoch": 0.01678261299708148, "grad_norm": 0.9905087947845459, "learning_rate": 4.807174201833589e-05, "loss": 2.2381, "step": 1866 }, { "epoch": 0.016791606894721933, "grad_norm": 1.1994690895080566, "learning_rate": 4.8068846558906696e-05, "loss": 2.2416, "step": 1867 }, { "epoch": 0.016800600792362383, "grad_norm": 0.7634406089782715, "learning_rate": 4.806594901454001e-05, "loss": 2.2251, "step": 1868 }, { "epoch": 0.016809594690002833, "grad_norm": 0.8624756336212158, "learning_rate": 4.806304938549769e-05, "loss": 2.3237, "step": 1869 }, { "epoch": 0.016818588587643284, "grad_norm": 0.7684710025787354, "learning_rate": 4.806014767204181e-05, "loss": 2.1954, "step": 1870 }, { "epoch": 0.016827582485283734, "grad_norm": 1.9442144632339478, "learning_rate": 4.805724387443462e-05, "loss": 2.2917, "step": 1871 }, { "epoch": 0.016836576382924184, "grad_norm": 1.1653467416763306, "learning_rate": 4.8054337992938566e-05, "loss": 2.243, "step": 1872 }, { "epoch": 0.016845570280564638, "grad_norm": 1.2989853620529175, "learning_rate": 4.805143002781628e-05, "loss": 2.2672, "step": 1873 }, { "epoch": 0.01685456417820509, "grad_norm": 0.5724433064460754, "learning_rate": 4.804851997933057e-05, "loss": 2.2178, "step": 1874 }, { "epoch": 0.01686355807584554, "grad_norm": 0.9845910668373108, "learning_rate": 4.8045607847744454e-05, "loss": 2.303, "step": 1875 }, { "epoch": 0.01687255197348599, "grad_norm": 0.8734267950057983, "learning_rate": 4.804269363332112e-05, "loss": 2.2648, "step": 1876 }, { "epoch": 0.01688154587112644, "grad_norm": 0.7489177584648132, "learning_rate": 4.8039777336323957e-05, "loss": 2.2902, "step": 1877 }, { "epoch": 0.01689053976876689, "grad_norm": 1.152745246887207, "learning_rate": 4.803685895701654e-05, "loss": 2.3479, "step": 1878 }, { "epoch": 0.016899533666407344, "grad_norm": 0.8649763464927673, "learning_rate": 4.803393849566261e-05, "loss": 2.2638, "step": 1879 }, { "epoch": 0.016908527564047794, "grad_norm": 0.957438588142395, "learning_rate": 4.803101595252614e-05, "loss": 2.3078, "step": 1880 }, { "epoch": 0.016917521461688245, "grad_norm": 0.749431848526001, "learning_rate": 4.802809132787125e-05, "loss": 2.1633, "step": 1881 }, { "epoch": 0.016926515359328695, "grad_norm": 0.8619750738143921, "learning_rate": 4.8025164621962284e-05, "loss": 2.2075, "step": 1882 }, { "epoch": 0.016935509256969145, "grad_norm": 1.1175601482391357, "learning_rate": 4.802223583506372e-05, "loss": 2.2293, "step": 1883 }, { "epoch": 0.0169445031546096, "grad_norm": 1.3970891237258911, "learning_rate": 4.8019304967440294e-05, "loss": 2.2848, "step": 1884 }, { "epoch": 0.01695349705225005, "grad_norm": 0.780992329120636, "learning_rate": 4.801637201935687e-05, "loss": 2.1801, "step": 1885 }, { "epoch": 0.0169624909498905, "grad_norm": 0.994679868221283, "learning_rate": 4.801343699107854e-05, "loss": 2.309, "step": 1886 }, { "epoch": 0.01697148484753095, "grad_norm": 0.9754831194877625, "learning_rate": 4.801049988287055e-05, "loss": 2.2761, "step": 1887 }, { "epoch": 0.0169804787451714, "grad_norm": 1.1257673501968384, "learning_rate": 4.8007560694998375e-05, "loss": 2.2436, "step": 1888 }, { "epoch": 0.01698947264281185, "grad_norm": 0.8542078733444214, "learning_rate": 4.800461942772765e-05, "loss": 2.2935, "step": 1889 }, { "epoch": 0.016998466540452305, "grad_norm": 0.7948276400566101, "learning_rate": 4.80016760813242e-05, "loss": 2.2553, "step": 1890 }, { "epoch": 0.017007460438092755, "grad_norm": 1.023480772972107, "learning_rate": 4.7998730656054035e-05, "loss": 2.2589, "step": 1891 }, { "epoch": 0.017016454335733205, "grad_norm": 0.5797092914581299, "learning_rate": 4.7995783152183366e-05, "loss": 2.272, "step": 1892 }, { "epoch": 0.017025448233373656, "grad_norm": 0.8182694315910339, "learning_rate": 4.799283356997859e-05, "loss": 2.2563, "step": 1893 }, { "epoch": 0.017034442131014106, "grad_norm": 1.257016658782959, "learning_rate": 4.798988190970628e-05, "loss": 2.2578, "step": 1894 }, { "epoch": 0.017043436028654557, "grad_norm": 1.1566047668457031, "learning_rate": 4.79869281716332e-05, "loss": 2.2644, "step": 1895 }, { "epoch": 0.01705242992629501, "grad_norm": 1.223054051399231, "learning_rate": 4.7983972356026315e-05, "loss": 2.3659, "step": 1896 }, { "epoch": 0.01706142382393546, "grad_norm": 0.6454828977584839, "learning_rate": 4.798101446315276e-05, "loss": 2.2161, "step": 1897 }, { "epoch": 0.01707041772157591, "grad_norm": 0.9303452968597412, "learning_rate": 4.797805449327988e-05, "loss": 2.2114, "step": 1898 }, { "epoch": 0.01707941161921636, "grad_norm": 1.0017653703689575, "learning_rate": 4.797509244667518e-05, "loss": 2.2505, "step": 1899 }, { "epoch": 0.017088405516856812, "grad_norm": 0.8761176466941833, "learning_rate": 4.797212832360637e-05, "loss": 2.3295, "step": 1900 }, { "epoch": 0.017097399414497262, "grad_norm": 1.3909958600997925, "learning_rate": 4.796916212434135e-05, "loss": 2.2344, "step": 1901 }, { "epoch": 0.017106393312137716, "grad_norm": 1.188353419303894, "learning_rate": 4.7966193849148194e-05, "loss": 2.3737, "step": 1902 }, { "epoch": 0.017115387209778166, "grad_norm": 0.9582924842834473, "learning_rate": 4.7963223498295176e-05, "loss": 2.2416, "step": 1903 }, { "epoch": 0.017124381107418617, "grad_norm": 0.6712927222251892, "learning_rate": 4.796025107205075e-05, "loss": 2.2281, "step": 1904 }, { "epoch": 0.017133375005059067, "grad_norm": 1.0080678462982178, "learning_rate": 4.7957276570683566e-05, "loss": 2.2302, "step": 1905 }, { "epoch": 0.017142368902699517, "grad_norm": 0.8767439126968384, "learning_rate": 4.7954299994462454e-05, "loss": 2.3459, "step": 1906 }, { "epoch": 0.017151362800339968, "grad_norm": 1.0786638259887695, "learning_rate": 4.795132134365643e-05, "loss": 2.2993, "step": 1907 }, { "epoch": 0.01716035669798042, "grad_norm": 1.2587511539459229, "learning_rate": 4.79483406185347e-05, "loss": 2.2827, "step": 1908 }, { "epoch": 0.017169350595620872, "grad_norm": 1.4085206985473633, "learning_rate": 4.7945357819366675e-05, "loss": 2.2252, "step": 1909 }, { "epoch": 0.017178344493261322, "grad_norm": 1.057995080947876, "learning_rate": 4.794237294642192e-05, "loss": 2.3504, "step": 1910 }, { "epoch": 0.017187338390901773, "grad_norm": 1.0978472232818604, "learning_rate": 4.793938599997021e-05, "loss": 2.3538, "step": 1911 }, { "epoch": 0.017196332288542223, "grad_norm": 0.9442919492721558, "learning_rate": 4.7936396980281496e-05, "loss": 2.3188, "step": 1912 }, { "epoch": 0.017205326186182673, "grad_norm": 0.8239895105361938, "learning_rate": 4.793340588762593e-05, "loss": 2.2172, "step": 1913 }, { "epoch": 0.017214320083823127, "grad_norm": 1.4184783697128296, "learning_rate": 4.793041272227384e-05, "loss": 2.283, "step": 1914 }, { "epoch": 0.017223313981463578, "grad_norm": 0.6340658068656921, "learning_rate": 4.792741748449575e-05, "loss": 2.1854, "step": 1915 }, { "epoch": 0.017232307879104028, "grad_norm": 1.1299843788146973, "learning_rate": 4.792442017456237e-05, "loss": 2.3031, "step": 1916 }, { "epoch": 0.01724130177674448, "grad_norm": 1.4297564029693604, "learning_rate": 4.792142079274457e-05, "loss": 2.2117, "step": 1917 }, { "epoch": 0.01725029567438493, "grad_norm": 0.6740981340408325, "learning_rate": 4.791841933931346e-05, "loss": 2.1921, "step": 1918 }, { "epoch": 0.017259289572025382, "grad_norm": 1.3782836198806763, "learning_rate": 4.7915415814540296e-05, "loss": 2.2988, "step": 1919 }, { "epoch": 0.017268283469665833, "grad_norm": 2.1514110565185547, "learning_rate": 4.791241021869653e-05, "loss": 2.292, "step": 1920 }, { "epoch": 0.017277277367306283, "grad_norm": 0.9755523204803467, "learning_rate": 4.79094025520538e-05, "loss": 2.1852, "step": 1921 }, { "epoch": 0.017286271264946734, "grad_norm": 0.5889203548431396, "learning_rate": 4.790639281488396e-05, "loss": 2.1865, "step": 1922 }, { "epoch": 0.017295265162587184, "grad_norm": 0.8238359093666077, "learning_rate": 4.7903381007459e-05, "loss": 2.2827, "step": 1923 }, { "epoch": 0.017304259060227634, "grad_norm": 1.3062047958374023, "learning_rate": 4.790036713005114e-05, "loss": 2.3012, "step": 1924 }, { "epoch": 0.017313252957868088, "grad_norm": 0.6288627982139587, "learning_rate": 4.7897351182932756e-05, "loss": 2.1774, "step": 1925 }, { "epoch": 0.01732224685550854, "grad_norm": 1.2543439865112305, "learning_rate": 4.789433316637644e-05, "loss": 2.3061, "step": 1926 }, { "epoch": 0.01733124075314899, "grad_norm": 1.6748862266540527, "learning_rate": 4.7891313080654954e-05, "loss": 2.2744, "step": 1927 }, { "epoch": 0.01734023465078944, "grad_norm": 1.1074962615966797, "learning_rate": 4.788829092604124e-05, "loss": 2.2932, "step": 1928 }, { "epoch": 0.01734922854842989, "grad_norm": 1.0261276960372925, "learning_rate": 4.788526670280845e-05, "loss": 2.254, "step": 1929 }, { "epoch": 0.01735822244607034, "grad_norm": 1.3492799997329712, "learning_rate": 4.7882240411229905e-05, "loss": 2.3212, "step": 1930 }, { "epoch": 0.017367216343710794, "grad_norm": 0.8169590830802917, "learning_rate": 4.787921205157912e-05, "loss": 2.1939, "step": 1931 }, { "epoch": 0.017376210241351244, "grad_norm": 2.293365240097046, "learning_rate": 4.78761816241298e-05, "loss": 2.3158, "step": 1932 }, { "epoch": 0.017385204138991694, "grad_norm": 0.652710497379303, "learning_rate": 4.7873149129155816e-05, "loss": 2.2172, "step": 1933 }, { "epoch": 0.017394198036632145, "grad_norm": 0.9734316468238831, "learning_rate": 4.787011456693125e-05, "loss": 2.2232, "step": 1934 }, { "epoch": 0.017403191934272595, "grad_norm": 1.9363571405410767, "learning_rate": 4.786707793773037e-05, "loss": 2.3442, "step": 1935 }, { "epoch": 0.017412185831913046, "grad_norm": 0.9665070176124573, "learning_rate": 4.786403924182761e-05, "loss": 2.2754, "step": 1936 }, { "epoch": 0.0174211797295535, "grad_norm": 1.055876612663269, "learning_rate": 4.786099847949761e-05, "loss": 2.2133, "step": 1937 }, { "epoch": 0.01743017362719395, "grad_norm": 1.0245312452316284, "learning_rate": 4.78579556510152e-05, "loss": 2.2364, "step": 1938 }, { "epoch": 0.0174391675248344, "grad_norm": 0.9658739566802979, "learning_rate": 4.785491075665537e-05, "loss": 2.124, "step": 1939 }, { "epoch": 0.01744816142247485, "grad_norm": 1.2974826097488403, "learning_rate": 4.7851863796693326e-05, "loss": 2.2864, "step": 1940 }, { "epoch": 0.0174571553201153, "grad_norm": 1.042313814163208, "learning_rate": 4.784881477140445e-05, "loss": 2.2755, "step": 1941 }, { "epoch": 0.01746614921775575, "grad_norm": 1.1423113346099854, "learning_rate": 4.7845763681064296e-05, "loss": 2.2599, "step": 1942 }, { "epoch": 0.017475143115396205, "grad_norm": 0.994827151298523, "learning_rate": 4.784271052594863e-05, "loss": 2.2505, "step": 1943 }, { "epoch": 0.017484137013036655, "grad_norm": 1.4856171607971191, "learning_rate": 4.7839655306333395e-05, "loss": 2.2876, "step": 1944 }, { "epoch": 0.017493130910677106, "grad_norm": 1.0082606077194214, "learning_rate": 4.783659802249472e-05, "loss": 2.315, "step": 1945 }, { "epoch": 0.017502124808317556, "grad_norm": 1.24881911277771, "learning_rate": 4.7833538674708905e-05, "loss": 2.3216, "step": 1946 }, { "epoch": 0.017511118705958006, "grad_norm": 0.852787435054779, "learning_rate": 4.783047726325246e-05, "loss": 2.1809, "step": 1947 }, { "epoch": 0.017520112603598457, "grad_norm": 0.8327879905700684, "learning_rate": 4.7827413788402077e-05, "loss": 2.2645, "step": 1948 }, { "epoch": 0.01752910650123891, "grad_norm": 1.182180643081665, "learning_rate": 4.7824348250434615e-05, "loss": 2.3767, "step": 1949 }, { "epoch": 0.01753810039887936, "grad_norm": 1.3073945045471191, "learning_rate": 4.782128064962714e-05, "loss": 2.284, "step": 1950 }, { "epoch": 0.01754709429651981, "grad_norm": 0.8969705700874329, "learning_rate": 4.78182109862569e-05, "loss": 2.2808, "step": 1951 }, { "epoch": 0.01755608819416026, "grad_norm": 1.233926773071289, "learning_rate": 4.781513926060135e-05, "loss": 2.2481, "step": 1952 }, { "epoch": 0.017565082091800712, "grad_norm": 0.8716261982917786, "learning_rate": 4.7812065472938075e-05, "loss": 2.1913, "step": 1953 }, { "epoch": 0.017574075989441162, "grad_norm": 1.4345555305480957, "learning_rate": 4.780898962354489e-05, "loss": 2.1927, "step": 1954 }, { "epoch": 0.017583069887081616, "grad_norm": 0.8084832429885864, "learning_rate": 4.780591171269979e-05, "loss": 2.3321, "step": 1955 }, { "epoch": 0.017592063784722067, "grad_norm": 0.8963560461997986, "learning_rate": 4.7802831740680955e-05, "loss": 2.3104, "step": 1956 }, { "epoch": 0.017601057682362517, "grad_norm": 0.7294359803199768, "learning_rate": 4.779974970776675e-05, "loss": 2.2292, "step": 1957 }, { "epoch": 0.017610051580002967, "grad_norm": 0.9283105134963989, "learning_rate": 4.779666561423573e-05, "loss": 2.3325, "step": 1958 }, { "epoch": 0.017619045477643418, "grad_norm": 1.5552716255187988, "learning_rate": 4.779357946036661e-05, "loss": 2.307, "step": 1959 }, { "epoch": 0.01762803937528387, "grad_norm": 1.5735383033752441, "learning_rate": 4.7790491246438345e-05, "loss": 2.2355, "step": 1960 }, { "epoch": 0.017637033272924322, "grad_norm": 1.307526707649231, "learning_rate": 4.7787400972730024e-05, "loss": 2.2477, "step": 1961 }, { "epoch": 0.017646027170564772, "grad_norm": 1.1407091617584229, "learning_rate": 4.7784308639520945e-05, "loss": 2.2384, "step": 1962 }, { "epoch": 0.017655021068205223, "grad_norm": 1.6965885162353516, "learning_rate": 4.77812142470906e-05, "loss": 2.1858, "step": 1963 }, { "epoch": 0.017664014965845673, "grad_norm": 1.4773070812225342, "learning_rate": 4.777811779571864e-05, "loss": 2.2493, "step": 1964 }, { "epoch": 0.017673008863486123, "grad_norm": 1.2642821073532104, "learning_rate": 4.777501928568494e-05, "loss": 2.2737, "step": 1965 }, { "epoch": 0.017682002761126577, "grad_norm": 1.8974300622940063, "learning_rate": 4.777191871726952e-05, "loss": 2.1704, "step": 1966 }, { "epoch": 0.017690996658767028, "grad_norm": 4.7509870529174805, "learning_rate": 4.7768816090752614e-05, "loss": 2.3157, "step": 1967 }, { "epoch": 0.017699990556407478, "grad_norm": 1.0007964372634888, "learning_rate": 4.776571140641464e-05, "loss": 2.2045, "step": 1968 }, { "epoch": 0.017708984454047928, "grad_norm": 1.1411465406417847, "learning_rate": 4.776260466453618e-05, "loss": 2.2924, "step": 1969 }, { "epoch": 0.01771797835168838, "grad_norm": 0.6316106915473938, "learning_rate": 4.775949586539803e-05, "loss": 2.1816, "step": 1970 }, { "epoch": 0.01772697224932883, "grad_norm": 0.6891862750053406, "learning_rate": 4.7756385009281165e-05, "loss": 2.2516, "step": 1971 }, { "epoch": 0.017735966146969283, "grad_norm": 1.0875741243362427, "learning_rate": 4.775327209646673e-05, "loss": 2.1976, "step": 1972 }, { "epoch": 0.017744960044609733, "grad_norm": 1.7854807376861572, "learning_rate": 4.775015712723607e-05, "loss": 2.3174, "step": 1973 }, { "epoch": 0.017753953942250184, "grad_norm": 1.4643340110778809, "learning_rate": 4.774704010187072e-05, "loss": 2.2586, "step": 1974 }, { "epoch": 0.017762947839890634, "grad_norm": 0.8734381794929504, "learning_rate": 4.7743921020652384e-05, "loss": 2.2768, "step": 1975 }, { "epoch": 0.017771941737531084, "grad_norm": 0.7068848013877869, "learning_rate": 4.774079988386296e-05, "loss": 2.2247, "step": 1976 }, { "epoch": 0.017780935635171535, "grad_norm": 1.885035753250122, "learning_rate": 4.773767669178454e-05, "loss": 2.3304, "step": 1977 }, { "epoch": 0.01778992953281199, "grad_norm": 1.7003952264785767, "learning_rate": 4.773455144469939e-05, "loss": 2.2169, "step": 1978 }, { "epoch": 0.01779892343045244, "grad_norm": 0.9710990786552429, "learning_rate": 4.7731424142889966e-05, "loss": 2.2928, "step": 1979 }, { "epoch": 0.01780791732809289, "grad_norm": 1.119415044784546, "learning_rate": 4.7728294786638914e-05, "loss": 2.2104, "step": 1980 }, { "epoch": 0.01781691122573334, "grad_norm": 1.5172535181045532, "learning_rate": 4.7725163376229064e-05, "loss": 2.2277, "step": 1981 }, { "epoch": 0.01782590512337379, "grad_norm": 1.0494482517242432, "learning_rate": 4.772202991194342e-05, "loss": 2.2825, "step": 1982 }, { "epoch": 0.01783489902101424, "grad_norm": 3.9221458435058594, "learning_rate": 4.77188943940652e-05, "loss": 2.1774, "step": 1983 }, { "epoch": 0.017843892918654694, "grad_norm": 1.619274377822876, "learning_rate": 4.771575682287776e-05, "loss": 2.2868, "step": 1984 }, { "epoch": 0.017852886816295144, "grad_norm": 2.6348698139190674, "learning_rate": 4.771261719866469e-05, "loss": 2.3313, "step": 1985 }, { "epoch": 0.017861880713935595, "grad_norm": 1.3733265399932861, "learning_rate": 4.770947552170974e-05, "loss": 2.1807, "step": 1986 }, { "epoch": 0.017870874611576045, "grad_norm": 1.2614948749542236, "learning_rate": 4.770633179229686e-05, "loss": 2.279, "step": 1987 }, { "epoch": 0.017879868509216496, "grad_norm": 1.1199651956558228, "learning_rate": 4.770318601071016e-05, "loss": 2.3157, "step": 1988 }, { "epoch": 0.017888862406856946, "grad_norm": 0.6340201497077942, "learning_rate": 4.7700038177233965e-05, "loss": 2.2394, "step": 1989 }, { "epoch": 0.0178978563044974, "grad_norm": 1.096011996269226, "learning_rate": 4.769688829215278e-05, "loss": 2.2629, "step": 1990 }, { "epoch": 0.01790685020213785, "grad_norm": 0.7549166679382324, "learning_rate": 4.769373635575126e-05, "loss": 2.2123, "step": 1991 }, { "epoch": 0.0179158440997783, "grad_norm": 1.0398750305175781, "learning_rate": 4.7690582368314304e-05, "loss": 2.2757, "step": 1992 }, { "epoch": 0.01792483799741875, "grad_norm": 0.9786906242370605, "learning_rate": 4.768742633012695e-05, "loss": 2.295, "step": 1993 }, { "epoch": 0.0179338318950592, "grad_norm": 1.1991000175476074, "learning_rate": 4.768426824147445e-05, "loss": 2.2628, "step": 1994 }, { "epoch": 0.017942825792699655, "grad_norm": 1.0576684474945068, "learning_rate": 4.768110810264221e-05, "loss": 2.2412, "step": 1995 }, { "epoch": 0.017951819690340105, "grad_norm": 0.6487658619880676, "learning_rate": 4.767794591391585e-05, "loss": 2.2332, "step": 1996 }, { "epoch": 0.017960813587980556, "grad_norm": 0.7374446988105774, "learning_rate": 4.767478167558117e-05, "loss": 2.3014, "step": 1997 }, { "epoch": 0.017969807485621006, "grad_norm": 1.2109500169754028, "learning_rate": 4.767161538792414e-05, "loss": 2.249, "step": 1998 }, { "epoch": 0.017978801383261456, "grad_norm": 0.9672105312347412, "learning_rate": 4.766844705123093e-05, "loss": 2.3134, "step": 1999 }, { "epoch": 0.017987795280901907, "grad_norm": 0.717119038105011, "learning_rate": 4.766527666578791e-05, "loss": 2.2366, "step": 2000 }, { "epoch": 0.01799678917854236, "grad_norm": 0.787848174571991, "learning_rate": 4.7662104231881574e-05, "loss": 2.2259, "step": 2001 }, { "epoch": 0.01800578307618281, "grad_norm": 0.8986544609069824, "learning_rate": 4.765892974979867e-05, "loss": 2.3011, "step": 2002 }, { "epoch": 0.01801477697382326, "grad_norm": 0.8208872079849243, "learning_rate": 4.7655753219826114e-05, "loss": 2.222, "step": 2003 }, { "epoch": 0.01802377087146371, "grad_norm": 0.9681172966957092, "learning_rate": 4.765257464225098e-05, "loss": 2.1809, "step": 2004 }, { "epoch": 0.018032764769104162, "grad_norm": 1.122541069984436, "learning_rate": 4.764939401736055e-05, "loss": 2.263, "step": 2005 }, { "epoch": 0.018041758666744612, "grad_norm": 1.1489046812057495, "learning_rate": 4.7646211345442284e-05, "loss": 2.2422, "step": 2006 }, { "epoch": 0.018050752564385066, "grad_norm": 1.0339548587799072, "learning_rate": 4.7643026626783824e-05, "loss": 2.2929, "step": 2007 }, { "epoch": 0.018059746462025517, "grad_norm": 1.1917225122451782, "learning_rate": 4.7639839861673015e-05, "loss": 2.2667, "step": 2008 }, { "epoch": 0.018068740359665967, "grad_norm": 1.2372184991836548, "learning_rate": 4.763665105039786e-05, "loss": 2.3402, "step": 2009 }, { "epoch": 0.018077734257306417, "grad_norm": 1.6296262741088867, "learning_rate": 4.763346019324657e-05, "loss": 2.2316, "step": 2010 }, { "epoch": 0.018086728154946868, "grad_norm": 1.5122408866882324, "learning_rate": 4.763026729050752e-05, "loss": 2.2416, "step": 2011 }, { "epoch": 0.018095722052587318, "grad_norm": 1.1967326402664185, "learning_rate": 4.762707234246929e-05, "loss": 2.3104, "step": 2012 }, { "epoch": 0.018104715950227772, "grad_norm": 0.8677659630775452, "learning_rate": 4.7623875349420635e-05, "loss": 2.198, "step": 2013 }, { "epoch": 0.018113709847868222, "grad_norm": 0.6742106676101685, "learning_rate": 4.762067631165049e-05, "loss": 2.2668, "step": 2014 }, { "epoch": 0.018122703745508673, "grad_norm": 1.3347848653793335, "learning_rate": 4.7617475229447986e-05, "loss": 2.297, "step": 2015 }, { "epoch": 0.018131697643149123, "grad_norm": 1.2756644487380981, "learning_rate": 4.7614272103102435e-05, "loss": 2.3188, "step": 2016 }, { "epoch": 0.018140691540789573, "grad_norm": 0.792405903339386, "learning_rate": 4.7611066932903334e-05, "loss": 2.291, "step": 2017 }, { "epoch": 0.018149685438430024, "grad_norm": 1.003030776977539, "learning_rate": 4.760785971914035e-05, "loss": 2.2849, "step": 2018 }, { "epoch": 0.018158679336070477, "grad_norm": 1.0209025144577026, "learning_rate": 4.7604650462103356e-05, "loss": 2.2053, "step": 2019 }, { "epoch": 0.018167673233710928, "grad_norm": 1.173720121383667, "learning_rate": 4.7601439162082405e-05, "loss": 2.2555, "step": 2020 }, { "epoch": 0.018176667131351378, "grad_norm": 0.7822023034095764, "learning_rate": 4.759822581936772e-05, "loss": 2.2815, "step": 2021 }, { "epoch": 0.01818566102899183, "grad_norm": 1.2155734300613403, "learning_rate": 4.759501043424974e-05, "loss": 2.268, "step": 2022 }, { "epoch": 0.01819465492663228, "grad_norm": 1.2870901823043823, "learning_rate": 4.7591793007019045e-05, "loss": 2.2225, "step": 2023 }, { "epoch": 0.01820364882427273, "grad_norm": 0.9671716690063477, "learning_rate": 4.758857353796643e-05, "loss": 2.2535, "step": 2024 }, { "epoch": 0.018212642721913183, "grad_norm": 1.100695252418518, "learning_rate": 4.7585352027382877e-05, "loss": 2.1722, "step": 2025 }, { "epoch": 0.018221636619553633, "grad_norm": 1.2669881582260132, "learning_rate": 4.758212847555953e-05, "loss": 2.2141, "step": 2026 }, { "epoch": 0.018230630517194084, "grad_norm": 1.1758842468261719, "learning_rate": 4.757890288278773e-05, "loss": 2.3188, "step": 2027 }, { "epoch": 0.018239624414834534, "grad_norm": 0.7398145198822021, "learning_rate": 4.757567524935901e-05, "loss": 2.2565, "step": 2028 }, { "epoch": 0.018248618312474985, "grad_norm": 1.2391008138656616, "learning_rate": 4.757244557556508e-05, "loss": 2.2822, "step": 2029 }, { "epoch": 0.01825761221011544, "grad_norm": 0.7369070053100586, "learning_rate": 4.756921386169783e-05, "loss": 2.2857, "step": 2030 }, { "epoch": 0.01826660610775589, "grad_norm": 0.6738448739051819, "learning_rate": 4.756598010804935e-05, "loss": 2.2603, "step": 2031 }, { "epoch": 0.01827560000539634, "grad_norm": 0.7800471782684326, "learning_rate": 4.756274431491188e-05, "loss": 2.2641, "step": 2032 }, { "epoch": 0.01828459390303679, "grad_norm": 1.374998688697815, "learning_rate": 4.7559506482577886e-05, "loss": 2.3034, "step": 2033 }, { "epoch": 0.01829358780067724, "grad_norm": 0.8791331052780151, "learning_rate": 4.7556266611339996e-05, "loss": 2.1869, "step": 2034 }, { "epoch": 0.01830258169831769, "grad_norm": 0.8267035484313965, "learning_rate": 4.7553024701491026e-05, "loss": 2.1641, "step": 2035 }, { "epoch": 0.018311575595958144, "grad_norm": 0.6286002397537231, "learning_rate": 4.754978075332398e-05, "loss": 2.2398, "step": 2036 }, { "epoch": 0.018320569493598594, "grad_norm": 1.0337588787078857, "learning_rate": 4.754653476713203e-05, "loss": 2.2865, "step": 2037 }, { "epoch": 0.018329563391239045, "grad_norm": 1.0305794477462769, "learning_rate": 4.754328674320856e-05, "loss": 2.213, "step": 2038 }, { "epoch": 0.018338557288879495, "grad_norm": 1.2267616987228394, "learning_rate": 4.754003668184711e-05, "loss": 2.2579, "step": 2039 }, { "epoch": 0.018347551186519945, "grad_norm": 1.3421757221221924, "learning_rate": 4.753678458334142e-05, "loss": 2.2516, "step": 2040 }, { "epoch": 0.018356545084160396, "grad_norm": 1.190474033355713, "learning_rate": 4.7533530447985416e-05, "loss": 2.3461, "step": 2041 }, { "epoch": 0.01836553898180085, "grad_norm": 1.4007905721664429, "learning_rate": 4.753027427607321e-05, "loss": 2.2789, "step": 2042 }, { "epoch": 0.0183745328794413, "grad_norm": 0.7736991047859192, "learning_rate": 4.752701606789908e-05, "loss": 2.2531, "step": 2043 }, { "epoch": 0.01838352677708175, "grad_norm": 1.0103511810302734, "learning_rate": 4.7523755823757496e-05, "loss": 2.2942, "step": 2044 }, { "epoch": 0.0183925206747222, "grad_norm": 1.0329315662384033, "learning_rate": 4.7520493543943126e-05, "loss": 2.2865, "step": 2045 }, { "epoch": 0.01840151457236265, "grad_norm": 0.8820033073425293, "learning_rate": 4.75172292287508e-05, "loss": 2.2318, "step": 2046 }, { "epoch": 0.0184105084700031, "grad_norm": 1.3961974382400513, "learning_rate": 4.751396287847556e-05, "loss": 2.2446, "step": 2047 }, { "epoch": 0.018419502367643555, "grad_norm": 0.6718161702156067, "learning_rate": 4.751069449341259e-05, "loss": 2.2673, "step": 2048 }, { "epoch": 0.018428496265284006, "grad_norm": 0.6252725124359131, "learning_rate": 4.7507424073857306e-05, "loss": 2.2101, "step": 2049 }, { "epoch": 0.018437490162924456, "grad_norm": 0.6773392558097839, "learning_rate": 4.750415162010528e-05, "loss": 2.1566, "step": 2050 }, { "epoch": 0.018446484060564906, "grad_norm": 1.1595820188522339, "learning_rate": 4.750087713245227e-05, "loss": 2.3324, "step": 2051 }, { "epoch": 0.018455477958205357, "grad_norm": 1.264174222946167, "learning_rate": 4.7497600611194225e-05, "loss": 2.22, "step": 2052 }, { "epoch": 0.018464471855845807, "grad_norm": 9.973942756652832, "learning_rate": 4.749432205662726e-05, "loss": 2.2713, "step": 2053 }, { "epoch": 0.01847346575348626, "grad_norm": 0.7811609506607056, "learning_rate": 4.74910414690477e-05, "loss": 2.2938, "step": 2054 }, { "epoch": 0.01848245965112671, "grad_norm": 0.7520526647567749, "learning_rate": 4.748775884875204e-05, "loss": 2.3017, "step": 2055 }, { "epoch": 0.01849145354876716, "grad_norm": 0.9412525296211243, "learning_rate": 4.748447419603696e-05, "loss": 2.2834, "step": 2056 }, { "epoch": 0.018500447446407612, "grad_norm": 1.0499240159988403, "learning_rate": 4.748118751119931e-05, "loss": 2.2876, "step": 2057 }, { "epoch": 0.018509441344048062, "grad_norm": 1.3839200735092163, "learning_rate": 4.747789879453615e-05, "loss": 2.2737, "step": 2058 }, { "epoch": 0.018518435241688513, "grad_norm": 0.8923033475875854, "learning_rate": 4.747460804634471e-05, "loss": 2.2669, "step": 2059 }, { "epoch": 0.018527429139328967, "grad_norm": 1.0484960079193115, "learning_rate": 4.7471315266922404e-05, "loss": 2.306, "step": 2060 }, { "epoch": 0.018536423036969417, "grad_norm": 1.178171157836914, "learning_rate": 4.7468020456566833e-05, "loss": 2.261, "step": 2061 }, { "epoch": 0.018545416934609867, "grad_norm": 0.9641180038452148, "learning_rate": 4.746472361557577e-05, "loss": 2.2529, "step": 2062 }, { "epoch": 0.018554410832250318, "grad_norm": 0.9128633141517639, "learning_rate": 4.746142474424717e-05, "loss": 2.2334, "step": 2063 }, { "epoch": 0.018563404729890768, "grad_norm": 1.6732821464538574, "learning_rate": 4.745812384287921e-05, "loss": 2.3148, "step": 2064 }, { "epoch": 0.01857239862753122, "grad_norm": 1.3879365921020508, "learning_rate": 4.7454820911770205e-05, "loss": 2.196, "step": 2065 }, { "epoch": 0.018581392525171672, "grad_norm": 0.5516320466995239, "learning_rate": 4.745151595121867e-05, "loss": 2.2615, "step": 2066 }, { "epoch": 0.018590386422812123, "grad_norm": 0.701082706451416, "learning_rate": 4.7448208961523314e-05, "loss": 2.2474, "step": 2067 }, { "epoch": 0.018599380320452573, "grad_norm": 1.2852147817611694, "learning_rate": 4.7444899942983004e-05, "loss": 2.2463, "step": 2068 }, { "epoch": 0.018608374218093023, "grad_norm": 0.7896435856819153, "learning_rate": 4.7441588895896805e-05, "loss": 2.2526, "step": 2069 }, { "epoch": 0.018617368115733474, "grad_norm": 0.7853023409843445, "learning_rate": 4.743827582056398e-05, "loss": 2.2339, "step": 2070 }, { "epoch": 0.018626362013373927, "grad_norm": 0.8163451552391052, "learning_rate": 4.743496071728396e-05, "loss": 2.2701, "step": 2071 }, { "epoch": 0.018635355911014378, "grad_norm": 1.1163349151611328, "learning_rate": 4.7431643586356355e-05, "loss": 2.2695, "step": 2072 }, { "epoch": 0.018644349808654828, "grad_norm": 1.006850004196167, "learning_rate": 4.742832442808096e-05, "loss": 2.2588, "step": 2073 }, { "epoch": 0.01865334370629528, "grad_norm": 1.424700140953064, "learning_rate": 4.742500324275776e-05, "loss": 2.2457, "step": 2074 }, { "epoch": 0.01866233760393573, "grad_norm": 0.9460026621818542, "learning_rate": 4.7421680030686925e-05, "loss": 2.2516, "step": 2075 }, { "epoch": 0.01867133150157618, "grad_norm": 0.9021950960159302, "learning_rate": 4.7418354792168794e-05, "loss": 2.2694, "step": 2076 }, { "epoch": 0.018680325399216633, "grad_norm": 0.9865133762359619, "learning_rate": 4.7415027527503916e-05, "loss": 2.1929, "step": 2077 }, { "epoch": 0.018689319296857083, "grad_norm": 0.7852327227592468, "learning_rate": 4.741169823699299e-05, "loss": 2.222, "step": 2078 }, { "epoch": 0.018698313194497534, "grad_norm": 1.5320199728012085, "learning_rate": 4.74083669209369e-05, "loss": 2.319, "step": 2079 }, { "epoch": 0.018707307092137984, "grad_norm": 1.0572597980499268, "learning_rate": 4.7405033579636756e-05, "loss": 2.2543, "step": 2080 }, { "epoch": 0.018716300989778435, "grad_norm": 0.7703912258148193, "learning_rate": 4.7401698213393815e-05, "loss": 2.2224, "step": 2081 }, { "epoch": 0.018725294887418885, "grad_norm": 0.9162212610244751, "learning_rate": 4.7398360822509514e-05, "loss": 2.2896, "step": 2082 }, { "epoch": 0.01873428878505934, "grad_norm": 1.6657401323318481, "learning_rate": 4.7395021407285486e-05, "loss": 2.2894, "step": 2083 }, { "epoch": 0.01874328268269979, "grad_norm": 0.785179078578949, "learning_rate": 4.739167996802355e-05, "loss": 2.267, "step": 2084 }, { "epoch": 0.01875227658034024, "grad_norm": 0.9193000793457031, "learning_rate": 4.738833650502569e-05, "loss": 2.3276, "step": 2085 }, { "epoch": 0.01876127047798069, "grad_norm": 0.6749075651168823, "learning_rate": 4.738499101859409e-05, "loss": 2.2389, "step": 2086 }, { "epoch": 0.01877026437562114, "grad_norm": 0.9294242858886719, "learning_rate": 4.738164350903113e-05, "loss": 2.2387, "step": 2087 }, { "epoch": 0.01877925827326159, "grad_norm": 0.6857464909553528, "learning_rate": 4.737829397663932e-05, "loss": 2.2784, "step": 2088 }, { "epoch": 0.018788252170902044, "grad_norm": 1.080670714378357, "learning_rate": 4.73749424217214e-05, "loss": 2.2793, "step": 2089 }, { "epoch": 0.018797246068542495, "grad_norm": 0.7008550763130188, "learning_rate": 4.73715888445803e-05, "loss": 2.2591, "step": 2090 }, { "epoch": 0.018806239966182945, "grad_norm": 1.3462494611740112, "learning_rate": 4.736823324551909e-05, "loss": 2.1636, "step": 2091 }, { "epoch": 0.018815233863823395, "grad_norm": 0.996548056602478, "learning_rate": 4.7364875624841056e-05, "loss": 2.3015, "step": 2092 }, { "epoch": 0.018824227761463846, "grad_norm": 1.0654503107070923, "learning_rate": 4.7361515982849644e-05, "loss": 2.266, "step": 2093 }, { "epoch": 0.018833221659104296, "grad_norm": 2.0192370414733887, "learning_rate": 4.7358154319848515e-05, "loss": 2.3698, "step": 2094 }, { "epoch": 0.01884221555674475, "grad_norm": 0.8046656847000122, "learning_rate": 4.7354790636141464e-05, "loss": 2.2319, "step": 2095 }, { "epoch": 0.0188512094543852, "grad_norm": 1.2429351806640625, "learning_rate": 4.735142493203253e-05, "loss": 2.2777, "step": 2096 }, { "epoch": 0.01886020335202565, "grad_norm": 2.4257755279541016, "learning_rate": 4.734805720782588e-05, "loss": 2.2688, "step": 2097 }, { "epoch": 0.0188691972496661, "grad_norm": 0.7009837031364441, "learning_rate": 4.734468746382589e-05, "loss": 2.2507, "step": 2098 }, { "epoch": 0.01887819114730655, "grad_norm": 0.9130544662475586, "learning_rate": 4.7341315700337115e-05, "loss": 2.3064, "step": 2099 }, { "epoch": 0.018887185044947002, "grad_norm": 1.0194252729415894, "learning_rate": 4.7337941917664285e-05, "loss": 2.2676, "step": 2100 }, { "epoch": 0.018896178942587456, "grad_norm": 1.5217162370681763, "learning_rate": 4.733456611611233e-05, "loss": 2.2814, "step": 2101 }, { "epoch": 0.018905172840227906, "grad_norm": 0.8659536242485046, "learning_rate": 4.733118829598635e-05, "loss": 2.2768, "step": 2102 }, { "epoch": 0.018914166737868356, "grad_norm": 0.7141968011856079, "learning_rate": 4.7327808457591626e-05, "loss": 2.2292, "step": 2103 }, { "epoch": 0.018923160635508807, "grad_norm": 0.6548545956611633, "learning_rate": 4.7324426601233617e-05, "loss": 2.2473, "step": 2104 }, { "epoch": 0.018932154533149257, "grad_norm": 0.6614185571670532, "learning_rate": 4.732104272721798e-05, "loss": 2.2625, "step": 2105 }, { "epoch": 0.01894114843078971, "grad_norm": 0.8646777868270874, "learning_rate": 4.731765683585054e-05, "loss": 2.1958, "step": 2106 }, { "epoch": 0.01895014232843016, "grad_norm": 0.7505583167076111, "learning_rate": 4.731426892743732e-05, "loss": 2.2519, "step": 2107 }, { "epoch": 0.01895913622607061, "grad_norm": 1.1736984252929688, "learning_rate": 4.7310879002284506e-05, "loss": 2.2927, "step": 2108 }, { "epoch": 0.018968130123711062, "grad_norm": 0.9012004137039185, "learning_rate": 4.730748706069849e-05, "loss": 2.3158, "step": 2109 }, { "epoch": 0.018977124021351512, "grad_norm": 1.2254165410995483, "learning_rate": 4.7304093102985804e-05, "loss": 2.3304, "step": 2110 }, { "epoch": 0.018986117918991963, "grad_norm": 0.8770359754562378, "learning_rate": 4.7300697129453215e-05, "loss": 2.2612, "step": 2111 }, { "epoch": 0.018995111816632416, "grad_norm": 0.7073201537132263, "learning_rate": 4.729729914040763e-05, "loss": 2.2204, "step": 2112 }, { "epoch": 0.019004105714272867, "grad_norm": 1.1034274101257324, "learning_rate": 4.7293899136156184e-05, "loss": 2.276, "step": 2113 }, { "epoch": 0.019013099611913317, "grad_norm": 0.6854431629180908, "learning_rate": 4.729049711700614e-05, "loss": 2.3044, "step": 2114 }, { "epoch": 0.019022093509553768, "grad_norm": 1.634405255317688, "learning_rate": 4.728709308326497e-05, "loss": 2.2313, "step": 2115 }, { "epoch": 0.019031087407194218, "grad_norm": 0.7387341856956482, "learning_rate": 4.728368703524034e-05, "loss": 2.2248, "step": 2116 }, { "epoch": 0.01904008130483467, "grad_norm": 1.1643847227096558, "learning_rate": 4.728027897324007e-05, "loss": 2.2813, "step": 2117 }, { "epoch": 0.019049075202475122, "grad_norm": 1.0862387418746948, "learning_rate": 4.727686889757219e-05, "loss": 2.3492, "step": 2118 }, { "epoch": 0.019058069100115572, "grad_norm": 0.9885034561157227, "learning_rate": 4.727345680854489e-05, "loss": 2.2607, "step": 2119 }, { "epoch": 0.019067062997756023, "grad_norm": 0.6856438517570496, "learning_rate": 4.7270042706466564e-05, "loss": 2.2502, "step": 2120 }, { "epoch": 0.019076056895396473, "grad_norm": 0.8556082844734192, "learning_rate": 4.7266626591645755e-05, "loss": 2.2719, "step": 2121 }, { "epoch": 0.019085050793036924, "grad_norm": 1.0042799711227417, "learning_rate": 4.7263208464391214e-05, "loss": 2.2606, "step": 2122 }, { "epoch": 0.019094044690677374, "grad_norm": 0.7154176235198975, "learning_rate": 4.7259788325011887e-05, "loss": 2.2773, "step": 2123 }, { "epoch": 0.019103038588317828, "grad_norm": 0.9029390811920166, "learning_rate": 4.725636617381686e-05, "loss": 2.2285, "step": 2124 }, { "epoch": 0.019112032485958278, "grad_norm": 1.173692226409912, "learning_rate": 4.725294201111543e-05, "loss": 2.2295, "step": 2125 }, { "epoch": 0.01912102638359873, "grad_norm": 0.7895246744155884, "learning_rate": 4.724951583721707e-05, "loss": 2.2207, "step": 2126 }, { "epoch": 0.01913002028123918, "grad_norm": 1.2068692445755005, "learning_rate": 4.7246087652431437e-05, "loss": 2.2925, "step": 2127 }, { "epoch": 0.01913901417887963, "grad_norm": 1.2170716524124146, "learning_rate": 4.724265745706836e-05, "loss": 2.2699, "step": 2128 }, { "epoch": 0.01914800807652008, "grad_norm": 1.0181915760040283, "learning_rate": 4.723922525143786e-05, "loss": 2.2817, "step": 2129 }, { "epoch": 0.019157001974160533, "grad_norm": 1.227689266204834, "learning_rate": 4.723579103585014e-05, "loss": 2.301, "step": 2130 }, { "epoch": 0.019165995871800984, "grad_norm": 1.2610079050064087, "learning_rate": 4.723235481061558e-05, "loss": 2.2527, "step": 2131 }, { "epoch": 0.019174989769441434, "grad_norm": 0.9254682660102844, "learning_rate": 4.722891657604473e-05, "loss": 2.302, "step": 2132 }, { "epoch": 0.019183983667081884, "grad_norm": 0.8742602467536926, "learning_rate": 4.722547633244834e-05, "loss": 2.2721, "step": 2133 }, { "epoch": 0.019192977564722335, "grad_norm": 1.3151696920394897, "learning_rate": 4.722203408013734e-05, "loss": 2.3188, "step": 2134 }, { "epoch": 0.019201971462362785, "grad_norm": 1.0550148487091064, "learning_rate": 4.721858981942284e-05, "loss": 2.3254, "step": 2135 }, { "epoch": 0.01921096536000324, "grad_norm": 1.1391528844833374, "learning_rate": 4.7215143550616124e-05, "loss": 2.2822, "step": 2136 }, { "epoch": 0.01921995925764369, "grad_norm": 1.5544606447219849, "learning_rate": 4.7211695274028664e-05, "loss": 2.2393, "step": 2137 }, { "epoch": 0.01922895315528414, "grad_norm": 1.161718487739563, "learning_rate": 4.720824498997211e-05, "loss": 2.2774, "step": 2138 }, { "epoch": 0.01923794705292459, "grad_norm": 1.1359113454818726, "learning_rate": 4.7204792698758296e-05, "loss": 2.2255, "step": 2139 }, { "epoch": 0.01924694095056504, "grad_norm": 1.1177033185958862, "learning_rate": 4.7201338400699224e-05, "loss": 2.3056, "step": 2140 }, { "epoch": 0.019255934848205494, "grad_norm": 1.4053698778152466, "learning_rate": 4.719788209610711e-05, "loss": 2.3237, "step": 2141 }, { "epoch": 0.019264928745845945, "grad_norm": 0.9074521064758301, "learning_rate": 4.7194423785294314e-05, "loss": 2.2571, "step": 2142 }, { "epoch": 0.019273922643486395, "grad_norm": 1.11603581905365, "learning_rate": 4.719096346857341e-05, "loss": 2.295, "step": 2143 }, { "epoch": 0.019282916541126845, "grad_norm": 1.2183903455734253, "learning_rate": 4.718750114625713e-05, "loss": 2.2918, "step": 2144 }, { "epoch": 0.019291910438767296, "grad_norm": 1.084152102470398, "learning_rate": 4.7184036818658386e-05, "loss": 2.2967, "step": 2145 }, { "epoch": 0.019300904336407746, "grad_norm": 1.1626291275024414, "learning_rate": 4.71805704860903e-05, "loss": 2.2791, "step": 2146 }, { "epoch": 0.0193098982340482, "grad_norm": 0.745373010635376, "learning_rate": 4.717710214886614e-05, "loss": 2.2097, "step": 2147 }, { "epoch": 0.01931889213168865, "grad_norm": 1.3231112957000732, "learning_rate": 4.717363180729937e-05, "loss": 2.3148, "step": 2148 }, { "epoch": 0.0193278860293291, "grad_norm": 1.5443940162658691, "learning_rate": 4.7170159461703645e-05, "loss": 2.2622, "step": 2149 }, { "epoch": 0.01933687992696955, "grad_norm": 1.0226582288742065, "learning_rate": 4.716668511239279e-05, "loss": 2.2974, "step": 2150 }, { "epoch": 0.01934587382461, "grad_norm": 0.9293818473815918, "learning_rate": 4.716320875968081e-05, "loss": 2.2171, "step": 2151 }, { "epoch": 0.01935486772225045, "grad_norm": 0.6801403760910034, "learning_rate": 4.715973040388189e-05, "loss": 2.2629, "step": 2152 }, { "epoch": 0.019363861619890906, "grad_norm": 1.138388752937317, "learning_rate": 4.71562500453104e-05, "loss": 2.2569, "step": 2153 }, { "epoch": 0.019372855517531356, "grad_norm": 1.8464343547821045, "learning_rate": 4.715276768428091e-05, "loss": 2.293, "step": 2154 }, { "epoch": 0.019381849415171806, "grad_norm": 1.1198322772979736, "learning_rate": 4.714928332110813e-05, "loss": 2.2619, "step": 2155 }, { "epoch": 0.019390843312812257, "grad_norm": 0.9041375517845154, "learning_rate": 4.714579695610698e-05, "loss": 2.3172, "step": 2156 }, { "epoch": 0.019399837210452707, "grad_norm": 1.5511722564697266, "learning_rate": 4.7142308589592556e-05, "loss": 2.1921, "step": 2157 }, { "epoch": 0.019408831108093157, "grad_norm": 0.9330231547355652, "learning_rate": 4.713881822188013e-05, "loss": 2.3008, "step": 2158 }, { "epoch": 0.01941782500573361, "grad_norm": 0.8872137069702148, "learning_rate": 4.713532585328516e-05, "loss": 2.2675, "step": 2159 }, { "epoch": 0.01942681890337406, "grad_norm": 0.7595241069793701, "learning_rate": 4.713183148412329e-05, "loss": 2.2634, "step": 2160 }, { "epoch": 0.019435812801014512, "grad_norm": 1.0190390348434448, "learning_rate": 4.7128335114710314e-05, "loss": 2.282, "step": 2161 }, { "epoch": 0.019444806698654962, "grad_norm": 0.9127533435821533, "learning_rate": 4.712483674536225e-05, "loss": 2.3731, "step": 2162 }, { "epoch": 0.019453800596295413, "grad_norm": 1.6705224514007568, "learning_rate": 4.7121336376395276e-05, "loss": 2.2921, "step": 2163 }, { "epoch": 0.019462794493935863, "grad_norm": 0.8918855786323547, "learning_rate": 4.711783400812575e-05, "loss": 2.2729, "step": 2164 }, { "epoch": 0.019471788391576317, "grad_norm": 1.0690165758132935, "learning_rate": 4.71143296408702e-05, "loss": 2.2783, "step": 2165 }, { "epoch": 0.019480782289216767, "grad_norm": 0.9630061984062195, "learning_rate": 4.711082327494536e-05, "loss": 2.265, "step": 2166 }, { "epoch": 0.019489776186857218, "grad_norm": 1.0091564655303955, "learning_rate": 4.7107314910668134e-05, "loss": 2.301, "step": 2167 }, { "epoch": 0.019498770084497668, "grad_norm": 0.8768916726112366, "learning_rate": 4.710380454835559e-05, "loss": 2.2611, "step": 2168 }, { "epoch": 0.019507763982138118, "grad_norm": 0.75152587890625, "learning_rate": 4.7100292188325e-05, "loss": 2.2601, "step": 2169 }, { "epoch": 0.01951675787977857, "grad_norm": 1.0343595743179321, "learning_rate": 4.7096777830893815e-05, "loss": 2.2083, "step": 2170 }, { "epoch": 0.019525751777419022, "grad_norm": 0.8038623332977295, "learning_rate": 4.709326147637965e-05, "loss": 2.2523, "step": 2171 }, { "epoch": 0.019534745675059473, "grad_norm": 0.9332152605056763, "learning_rate": 4.70897431251003e-05, "loss": 2.2489, "step": 2172 }, { "epoch": 0.019543739572699923, "grad_norm": 0.8469945192337036, "learning_rate": 4.708622277737376e-05, "loss": 2.301, "step": 2173 }, { "epoch": 0.019552733470340374, "grad_norm": 1.0699492692947388, "learning_rate": 4.70827004335182e-05, "loss": 2.1979, "step": 2174 }, { "epoch": 0.019561727367980824, "grad_norm": 1.2029427289962769, "learning_rate": 4.7079176093851965e-05, "loss": 2.322, "step": 2175 }, { "epoch": 0.019570721265621274, "grad_norm": 1.2178703546524048, "learning_rate": 4.7075649758693565e-05, "loss": 2.1813, "step": 2176 }, { "epoch": 0.019579715163261728, "grad_norm": 0.9799526929855347, "learning_rate": 4.707212142836173e-05, "loss": 2.2817, "step": 2177 }, { "epoch": 0.01958870906090218, "grad_norm": 0.7159790992736816, "learning_rate": 4.706859110317532e-05, "loss": 2.2446, "step": 2178 }, { "epoch": 0.01959770295854263, "grad_norm": 1.7315138578414917, "learning_rate": 4.7065058783453424e-05, "loss": 2.3949, "step": 2179 }, { "epoch": 0.01960669685618308, "grad_norm": 1.375238299369812, "learning_rate": 4.7061524469515285e-05, "loss": 2.2587, "step": 2180 }, { "epoch": 0.01961569075382353, "grad_norm": 1.0229532718658447, "learning_rate": 4.7057988161680324e-05, "loss": 2.2102, "step": 2181 }, { "epoch": 0.019624684651463983, "grad_norm": 0.9224338531494141, "learning_rate": 4.7054449860268144e-05, "loss": 2.1597, "step": 2182 }, { "epoch": 0.019633678549104434, "grad_norm": 0.561843991279602, "learning_rate": 4.705090956559856e-05, "loss": 2.2338, "step": 2183 }, { "epoch": 0.019642672446744884, "grad_norm": 1.0539823770523071, "learning_rate": 4.7047367277991506e-05, "loss": 2.19, "step": 2184 }, { "epoch": 0.019651666344385334, "grad_norm": 1.7389276027679443, "learning_rate": 4.704382299776715e-05, "loss": 2.2407, "step": 2185 }, { "epoch": 0.019660660242025785, "grad_norm": 2.4400722980499268, "learning_rate": 4.704027672524581e-05, "loss": 2.2884, "step": 2186 }, { "epoch": 0.019669654139666235, "grad_norm": 2.1247634887695312, "learning_rate": 4.7036728460748e-05, "loss": 2.1692, "step": 2187 }, { "epoch": 0.01967864803730669, "grad_norm": 0.9995978474617004, "learning_rate": 4.7033178204594414e-05, "loss": 2.2591, "step": 2188 }, { "epoch": 0.01968764193494714, "grad_norm": 1.5498327016830444, "learning_rate": 4.702962595710591e-05, "loss": 2.1985, "step": 2189 }, { "epoch": 0.01969663583258759, "grad_norm": 0.9117842316627502, "learning_rate": 4.7026071718603536e-05, "loss": 2.2747, "step": 2190 }, { "epoch": 0.01970562973022804, "grad_norm": 1.9296698570251465, "learning_rate": 4.7022515489408536e-05, "loss": 2.2605, "step": 2191 }, { "epoch": 0.01971462362786849, "grad_norm": 0.8542248606681824, "learning_rate": 4.70189572698423e-05, "loss": 2.2153, "step": 2192 }, { "epoch": 0.01972361752550894, "grad_norm": 1.0186160802841187, "learning_rate": 4.7015397060226417e-05, "loss": 2.2874, "step": 2193 }, { "epoch": 0.019732611423149395, "grad_norm": 1.1415942907333374, "learning_rate": 4.701183486088267e-05, "loss": 2.2807, "step": 2194 }, { "epoch": 0.019741605320789845, "grad_norm": 0.7824199795722961, "learning_rate": 4.700827067213299e-05, "loss": 2.1992, "step": 2195 }, { "epoch": 0.019750599218430295, "grad_norm": 1.9839529991149902, "learning_rate": 4.7004704494299526e-05, "loss": 2.3262, "step": 2196 }, { "epoch": 0.019759593116070746, "grad_norm": 1.6303540468215942, "learning_rate": 4.700113632770456e-05, "loss": 2.2494, "step": 2197 }, { "epoch": 0.019768587013711196, "grad_norm": 1.9447354078292847, "learning_rate": 4.699756617267059e-05, "loss": 2.2498, "step": 2198 }, { "epoch": 0.019777580911351646, "grad_norm": 0.8938925862312317, "learning_rate": 4.6993994029520285e-05, "loss": 2.28, "step": 2199 }, { "epoch": 0.0197865748089921, "grad_norm": 1.4122340679168701, "learning_rate": 4.699041989857649e-05, "loss": 2.255, "step": 2200 }, { "epoch": 0.01979556870663255, "grad_norm": 1.2862732410430908, "learning_rate": 4.698684378016222e-05, "loss": 2.1884, "step": 2201 }, { "epoch": 0.019804562604273, "grad_norm": 2.5184741020202637, "learning_rate": 4.698326567460071e-05, "loss": 2.3101, "step": 2202 }, { "epoch": 0.01981355650191345, "grad_norm": 1.8277627229690552, "learning_rate": 4.697968558221532e-05, "loss": 2.2916, "step": 2203 }, { "epoch": 0.0198225503995539, "grad_norm": 0.8037735819816589, "learning_rate": 4.697610350332962e-05, "loss": 2.2114, "step": 2204 }, { "epoch": 0.019831544297194352, "grad_norm": 2.05893611907959, "learning_rate": 4.697251943826736e-05, "loss": 2.312, "step": 2205 }, { "epoch": 0.019840538194834806, "grad_norm": 1.143593192100525, "learning_rate": 4.696893338735246e-05, "loss": 2.2213, "step": 2206 }, { "epoch": 0.019849532092475256, "grad_norm": 1.3206249475479126, "learning_rate": 4.696534535090901e-05, "loss": 2.2009, "step": 2207 }, { "epoch": 0.019858525990115707, "grad_norm": 0.8536596894264221, "learning_rate": 4.6961755329261316e-05, "loss": 2.3212, "step": 2208 }, { "epoch": 0.019867519887756157, "grad_norm": 1.07578706741333, "learning_rate": 4.6958163322733836e-05, "loss": 2.2964, "step": 2209 }, { "epoch": 0.019876513785396607, "grad_norm": 1.3569079637527466, "learning_rate": 4.69545693316512e-05, "loss": 2.2969, "step": 2210 }, { "epoch": 0.019885507683037058, "grad_norm": 1.2989603281021118, "learning_rate": 4.6950973356338234e-05, "loss": 2.2707, "step": 2211 }, { "epoch": 0.01989450158067751, "grad_norm": 1.117103099822998, "learning_rate": 4.694737539711994e-05, "loss": 2.2768, "step": 2212 }, { "epoch": 0.019903495478317962, "grad_norm": 1.3671441078186035, "learning_rate": 4.6943775454321504e-05, "loss": 2.2557, "step": 2213 }, { "epoch": 0.019912489375958412, "grad_norm": 1.77822744846344, "learning_rate": 4.6940173528268274e-05, "loss": 2.2608, "step": 2214 }, { "epoch": 0.019921483273598863, "grad_norm": 1.5484094619750977, "learning_rate": 4.693656961928579e-05, "loss": 2.3006, "step": 2215 }, { "epoch": 0.019930477171239313, "grad_norm": 0.9517895579338074, "learning_rate": 4.693296372769977e-05, "loss": 2.2517, "step": 2216 }, { "epoch": 0.019939471068879767, "grad_norm": 1.3722822666168213, "learning_rate": 4.692935585383612e-05, "loss": 2.3061, "step": 2217 }, { "epoch": 0.019948464966520217, "grad_norm": 1.2710357904434204, "learning_rate": 4.69257459980209e-05, "loss": 2.2946, "step": 2218 }, { "epoch": 0.019957458864160667, "grad_norm": 0.76896071434021, "learning_rate": 4.6922134160580386e-05, "loss": 2.2581, "step": 2219 }, { "epoch": 0.019966452761801118, "grad_norm": 0.7425410747528076, "learning_rate": 4.6918520341840986e-05, "loss": 2.3409, "step": 2220 }, { "epoch": 0.019975446659441568, "grad_norm": 1.2122324705123901, "learning_rate": 4.6914904542129334e-05, "loss": 2.2828, "step": 2221 }, { "epoch": 0.01998444055708202, "grad_norm": 1.227325439453125, "learning_rate": 4.691128676177221e-05, "loss": 2.3047, "step": 2222 }, { "epoch": 0.019993434454722472, "grad_norm": 0.9643160104751587, "learning_rate": 4.690766700109659e-05, "loss": 2.2962, "step": 2223 }, { "epoch": 0.020002428352362923, "grad_norm": 1.1272917985916138, "learning_rate": 4.690404526042963e-05, "loss": 2.2196, "step": 2224 }, { "epoch": 0.020011422250003373, "grad_norm": 0.9037061929702759, "learning_rate": 4.6900421540098646e-05, "loss": 2.2359, "step": 2225 }, { "epoch": 0.020020416147643823, "grad_norm": 2.1908326148986816, "learning_rate": 4.689679584043115e-05, "loss": 2.2762, "step": 2226 }, { "epoch": 0.020029410045284274, "grad_norm": 0.933113694190979, "learning_rate": 4.689316816175484e-05, "loss": 2.2607, "step": 2227 }, { "epoch": 0.020038403942924724, "grad_norm": 1.0624946355819702, "learning_rate": 4.6889538504397566e-05, "loss": 2.2503, "step": 2228 }, { "epoch": 0.020047397840565178, "grad_norm": 0.898786723613739, "learning_rate": 4.688590686868738e-05, "loss": 2.2297, "step": 2229 }, { "epoch": 0.02005639173820563, "grad_norm": 0.9204374551773071, "learning_rate": 4.68822732549525e-05, "loss": 2.2615, "step": 2230 }, { "epoch": 0.02006538563584608, "grad_norm": 1.1698166131973267, "learning_rate": 4.687863766352134e-05, "loss": 2.3257, "step": 2231 }, { "epoch": 0.02007437953348653, "grad_norm": 0.8170420527458191, "learning_rate": 4.6875000094722476e-05, "loss": 2.2767, "step": 2232 }, { "epoch": 0.02008337343112698, "grad_norm": 1.3359299898147583, "learning_rate": 4.687136054888466e-05, "loss": 2.2749, "step": 2233 }, { "epoch": 0.02009236732876743, "grad_norm": 0.5163419842720032, "learning_rate": 4.686771902633684e-05, "loss": 2.2225, "step": 2234 }, { "epoch": 0.020101361226407884, "grad_norm": 0.7911099791526794, "learning_rate": 4.686407552740812e-05, "loss": 2.2601, "step": 2235 }, { "epoch": 0.020110355124048334, "grad_norm": 1.295609951019287, "learning_rate": 4.6860430052427814e-05, "loss": 2.2134, "step": 2236 }, { "epoch": 0.020119349021688784, "grad_norm": 1.028989315032959, "learning_rate": 4.685678260172538e-05, "loss": 2.2523, "step": 2237 }, { "epoch": 0.020128342919329235, "grad_norm": 0.9491502046585083, "learning_rate": 4.6853133175630476e-05, "loss": 2.2543, "step": 2238 }, { "epoch": 0.020137336816969685, "grad_norm": 0.9951165914535522, "learning_rate": 4.6849481774472934e-05, "loss": 2.3439, "step": 2239 }, { "epoch": 0.020146330714610135, "grad_norm": 1.1918082237243652, "learning_rate": 4.684582839858277e-05, "loss": 2.2269, "step": 2240 }, { "epoch": 0.02015532461225059, "grad_norm": 1.0417859554290771, "learning_rate": 4.684217304829016e-05, "loss": 2.1471, "step": 2241 }, { "epoch": 0.02016431850989104, "grad_norm": 1.2586559057235718, "learning_rate": 4.683851572392548e-05, "loss": 2.3167, "step": 2242 }, { "epoch": 0.02017331240753149, "grad_norm": 1.1549022197723389, "learning_rate": 4.6834856425819266e-05, "loss": 2.3416, "step": 2243 }, { "epoch": 0.02018230630517194, "grad_norm": 1.1947603225708008, "learning_rate": 4.683119515430224e-05, "loss": 2.2699, "step": 2244 }, { "epoch": 0.02019130020281239, "grad_norm": 0.9119445085525513, "learning_rate": 4.682753190970533e-05, "loss": 2.2507, "step": 2245 }, { "epoch": 0.02020029410045284, "grad_norm": 1.2544097900390625, "learning_rate": 4.6823866692359586e-05, "loss": 2.2758, "step": 2246 }, { "epoch": 0.020209287998093295, "grad_norm": 1.0736241340637207, "learning_rate": 4.6820199502596276e-05, "loss": 2.2722, "step": 2247 }, { "epoch": 0.020218281895733745, "grad_norm": 1.415103554725647, "learning_rate": 4.681653034074684e-05, "loss": 2.3335, "step": 2248 }, { "epoch": 0.020227275793374196, "grad_norm": 0.6900818347930908, "learning_rate": 4.681285920714289e-05, "loss": 2.1638, "step": 2249 }, { "epoch": 0.020236269691014646, "grad_norm": 0.8814685344696045, "learning_rate": 4.680918610211622e-05, "loss": 2.1492, "step": 2250 }, { "epoch": 0.020245263588655096, "grad_norm": 0.8556309938430786, "learning_rate": 4.68055110259988e-05, "loss": 2.1946, "step": 2251 }, { "epoch": 0.02025425748629555, "grad_norm": 1.007909893989563, "learning_rate": 4.680183397912278e-05, "loss": 2.2751, "step": 2252 }, { "epoch": 0.020263251383936, "grad_norm": 1.0881547927856445, "learning_rate": 4.67981549618205e-05, "loss": 2.3669, "step": 2253 }, { "epoch": 0.02027224528157645, "grad_norm": 0.7005136609077454, "learning_rate": 4.679447397442443e-05, "loss": 2.2185, "step": 2254 }, { "epoch": 0.0202812391792169, "grad_norm": 0.9973048567771912, "learning_rate": 4.6790791017267297e-05, "loss": 2.1632, "step": 2255 }, { "epoch": 0.02029023307685735, "grad_norm": 1.1708922386169434, "learning_rate": 4.678710609068193e-05, "loss": 2.2071, "step": 2256 }, { "epoch": 0.020299226974497802, "grad_norm": 0.8599388599395752, "learning_rate": 4.678341919500139e-05, "loss": 2.2452, "step": 2257 }, { "epoch": 0.020308220872138256, "grad_norm": 0.7157385349273682, "learning_rate": 4.677973033055889e-05, "loss": 2.2244, "step": 2258 }, { "epoch": 0.020317214769778706, "grad_norm": 0.7259404063224792, "learning_rate": 4.6776039497687816e-05, "loss": 2.232, "step": 2259 }, { "epoch": 0.020326208667419157, "grad_norm": 1.5459734201431274, "learning_rate": 4.6772346696721745e-05, "loss": 2.2148, "step": 2260 }, { "epoch": 0.020335202565059607, "grad_norm": 1.325190782546997, "learning_rate": 4.6768651927994434e-05, "loss": 2.3667, "step": 2261 }, { "epoch": 0.020344196462700057, "grad_norm": 0.7426716685295105, "learning_rate": 4.676495519183981e-05, "loss": 2.2936, "step": 2262 }, { "epoch": 0.020353190360340508, "grad_norm": 0.9877602458000183, "learning_rate": 4.676125648859198e-05, "loss": 2.2045, "step": 2263 }, { "epoch": 0.02036218425798096, "grad_norm": 0.903903603553772, "learning_rate": 4.675755581858522e-05, "loss": 2.223, "step": 2264 }, { "epoch": 0.020371178155621412, "grad_norm": 0.6788469552993774, "learning_rate": 4.6753853182154006e-05, "loss": 2.2215, "step": 2265 }, { "epoch": 0.020380172053261862, "grad_norm": 1.1426076889038086, "learning_rate": 4.675014857963297e-05, "loss": 2.2313, "step": 2266 }, { "epoch": 0.020389165950902313, "grad_norm": 1.1569961309432983, "learning_rate": 4.674644201135694e-05, "loss": 2.2706, "step": 2267 }, { "epoch": 0.020398159848542763, "grad_norm": 1.0449674129486084, "learning_rate": 4.674273347766089e-05, "loss": 2.2692, "step": 2268 }, { "epoch": 0.020407153746183213, "grad_norm": 1.0808453559875488, "learning_rate": 4.6739022978880023e-05, "loss": 2.2219, "step": 2269 }, { "epoch": 0.020416147643823667, "grad_norm": 1.3297439813613892, "learning_rate": 4.6735310515349665e-05, "loss": 2.1784, "step": 2270 }, { "epoch": 0.020425141541464117, "grad_norm": 1.626508355140686, "learning_rate": 4.673159608740536e-05, "loss": 2.2606, "step": 2271 }, { "epoch": 0.020434135439104568, "grad_norm": 1.2485791444778442, "learning_rate": 4.6727879695382804e-05, "loss": 2.2292, "step": 2272 }, { "epoch": 0.020443129336745018, "grad_norm": 1.4958404302597046, "learning_rate": 4.672416133961789e-05, "loss": 2.3006, "step": 2273 }, { "epoch": 0.02045212323438547, "grad_norm": 0.5167723894119263, "learning_rate": 4.672044102044667e-05, "loss": 2.1836, "step": 2274 }, { "epoch": 0.02046111713202592, "grad_norm": 1.8018296957015991, "learning_rate": 4.6716718738205384e-05, "loss": 2.2126, "step": 2275 }, { "epoch": 0.020470111029666373, "grad_norm": 0.9538870453834534, "learning_rate": 4.671299449323045e-05, "loss": 2.2102, "step": 2276 }, { "epoch": 0.020479104927306823, "grad_norm": 1.1056751012802124, "learning_rate": 4.6709268285858464e-05, "loss": 2.3291, "step": 2277 }, { "epoch": 0.020488098824947273, "grad_norm": 0.8626651167869568, "learning_rate": 4.670554011642619e-05, "loss": 2.1962, "step": 2278 }, { "epoch": 0.020497092722587724, "grad_norm": 1.1622825860977173, "learning_rate": 4.6701809985270586e-05, "loss": 2.2507, "step": 2279 }, { "epoch": 0.020506086620228174, "grad_norm": 1.0902806520462036, "learning_rate": 4.6698077892728765e-05, "loss": 2.2624, "step": 2280 }, { "epoch": 0.020515080517868624, "grad_norm": 1.4502575397491455, "learning_rate": 4.669434383913803e-05, "loss": 2.2853, "step": 2281 }, { "epoch": 0.02052407441550908, "grad_norm": 1.2830830812454224, "learning_rate": 4.6690607824835875e-05, "loss": 2.2477, "step": 2282 }, { "epoch": 0.02053306831314953, "grad_norm": 1.2313460111618042, "learning_rate": 4.668686985015994e-05, "loss": 2.2874, "step": 2283 }, { "epoch": 0.02054206221078998, "grad_norm": 0.7200677990913391, "learning_rate": 4.6683129915448076e-05, "loss": 2.2656, "step": 2284 }, { "epoch": 0.02055105610843043, "grad_norm": 1.3321913480758667, "learning_rate": 4.6679388021038285e-05, "loss": 2.2744, "step": 2285 }, { "epoch": 0.02056005000607088, "grad_norm": 1.0952403545379639, "learning_rate": 4.667564416726875e-05, "loss": 2.2452, "step": 2286 }, { "epoch": 0.020569043903711334, "grad_norm": 0.9069653153419495, "learning_rate": 4.667189835447785e-05, "loss": 2.2574, "step": 2287 }, { "epoch": 0.020578037801351784, "grad_norm": 0.7422159910202026, "learning_rate": 4.666815058300411e-05, "loss": 2.2295, "step": 2288 }, { "epoch": 0.020587031698992234, "grad_norm": 1.071475625038147, "learning_rate": 4.666440085318626e-05, "loss": 2.3362, "step": 2289 }, { "epoch": 0.020596025596632685, "grad_norm": 1.1728519201278687, "learning_rate": 4.66606491653632e-05, "loss": 2.225, "step": 2290 }, { "epoch": 0.020605019494273135, "grad_norm": 1.3963611125946045, "learning_rate": 4.6656895519873997e-05, "loss": 2.2349, "step": 2291 }, { "epoch": 0.020614013391913585, "grad_norm": 1.096144676208496, "learning_rate": 4.665313991705791e-05, "loss": 2.2711, "step": 2292 }, { "epoch": 0.02062300728955404, "grad_norm": 1.1083519458770752, "learning_rate": 4.664938235725436e-05, "loss": 2.2612, "step": 2293 }, { "epoch": 0.02063200118719449, "grad_norm": 1.0622618198394775, "learning_rate": 4.664562284080295e-05, "loss": 2.2165, "step": 2294 }, { "epoch": 0.02064099508483494, "grad_norm": 1.2845836877822876, "learning_rate": 4.664186136804346e-05, "loss": 2.2862, "step": 2295 }, { "epoch": 0.02064998898247539, "grad_norm": 1.5346875190734863, "learning_rate": 4.663809793931585e-05, "loss": 2.1872, "step": 2296 }, { "epoch": 0.02065898288011584, "grad_norm": 1.0051254034042358, "learning_rate": 4.6634332554960255e-05, "loss": 2.2732, "step": 2297 }, { "epoch": 0.02066797677775629, "grad_norm": 1.5149986743927002, "learning_rate": 4.663056521531698e-05, "loss": 2.3743, "step": 2298 }, { "epoch": 0.020676970675396745, "grad_norm": 0.8976969718933105, "learning_rate": 4.6626795920726527e-05, "loss": 2.2714, "step": 2299 }, { "epoch": 0.020685964573037195, "grad_norm": 0.8198716044425964, "learning_rate": 4.662302467152955e-05, "loss": 2.2284, "step": 2300 }, { "epoch": 0.020694958470677646, "grad_norm": 0.9765826463699341, "learning_rate": 4.66192514680669e-05, "loss": 2.2667, "step": 2301 }, { "epoch": 0.020703952368318096, "grad_norm": 1.591596007347107, "learning_rate": 4.6615476310679585e-05, "loss": 2.2219, "step": 2302 }, { "epoch": 0.020712946265958546, "grad_norm": 1.2211425304412842, "learning_rate": 4.6611699199708804e-05, "loss": 2.2806, "step": 2303 }, { "epoch": 0.020721940163598997, "grad_norm": 1.56206214427948, "learning_rate": 4.6607920135495916e-05, "loss": 2.2481, "step": 2304 }, { "epoch": 0.02073093406123945, "grad_norm": 0.8573588132858276, "learning_rate": 4.6604139118382494e-05, "loss": 2.2527, "step": 2305 }, { "epoch": 0.0207399279588799, "grad_norm": 1.1655347347259521, "learning_rate": 4.660035614871024e-05, "loss": 2.1892, "step": 2306 }, { "epoch": 0.02074892185652035, "grad_norm": 1.4244165420532227, "learning_rate": 4.659657122682107e-05, "loss": 2.307, "step": 2307 }, { "epoch": 0.0207579157541608, "grad_norm": 1.1542502641677856, "learning_rate": 4.659278435305706e-05, "loss": 2.3164, "step": 2308 }, { "epoch": 0.020766909651801252, "grad_norm": 1.123436450958252, "learning_rate": 4.658899552776045e-05, "loss": 2.2913, "step": 2309 }, { "epoch": 0.020775903549441702, "grad_norm": 1.5499385595321655, "learning_rate": 4.658520475127368e-05, "loss": 2.3659, "step": 2310 }, { "epoch": 0.020784897447082156, "grad_norm": 1.5019572973251343, "learning_rate": 4.6581412023939354e-05, "loss": 2.1872, "step": 2311 }, { "epoch": 0.020793891344722606, "grad_norm": 0.77071213722229, "learning_rate": 4.657761734610024e-05, "loss": 2.2373, "step": 2312 }, { "epoch": 0.020802885242363057, "grad_norm": 0.8370568156242371, "learning_rate": 4.657382071809934e-05, "loss": 2.1683, "step": 2313 }, { "epoch": 0.020811879140003507, "grad_norm": 1.0323280096054077, "learning_rate": 4.657002214027974e-05, "loss": 2.2018, "step": 2314 }, { "epoch": 0.020820873037643958, "grad_norm": 0.9621389508247375, "learning_rate": 4.656622161298478e-05, "loss": 2.1993, "step": 2315 }, { "epoch": 0.020829866935284408, "grad_norm": 1.353743553161621, "learning_rate": 4.656241913655793e-05, "loss": 2.3285, "step": 2316 }, { "epoch": 0.020838860832924862, "grad_norm": 0.8924314975738525, "learning_rate": 4.6558614711342874e-05, "loss": 2.2443, "step": 2317 }, { "epoch": 0.020847854730565312, "grad_norm": 0.8163611888885498, "learning_rate": 4.655480833768344e-05, "loss": 2.3378, "step": 2318 }, { "epoch": 0.020856848628205762, "grad_norm": 0.983971118927002, "learning_rate": 4.655100001592365e-05, "loss": 2.2452, "step": 2319 }, { "epoch": 0.020865842525846213, "grad_norm": 1.0396231412887573, "learning_rate": 4.654718974640768e-05, "loss": 2.2673, "step": 2320 }, { "epoch": 0.020874836423486663, "grad_norm": 1.0256354808807373, "learning_rate": 4.654337752947992e-05, "loss": 2.2545, "step": 2321 }, { "epoch": 0.020883830321127114, "grad_norm": 0.881330132484436, "learning_rate": 4.65395633654849e-05, "loss": 2.2316, "step": 2322 }, { "epoch": 0.020892824218767567, "grad_norm": 0.9010043144226074, "learning_rate": 4.653574725476734e-05, "loss": 2.239, "step": 2323 }, { "epoch": 0.020901818116408018, "grad_norm": 1.4055900573730469, "learning_rate": 4.653192919767214e-05, "loss": 2.2991, "step": 2324 }, { "epoch": 0.020910812014048468, "grad_norm": 0.8476502895355225, "learning_rate": 4.6528109194544376e-05, "loss": 2.2553, "step": 2325 }, { "epoch": 0.02091980591168892, "grad_norm": 0.5734707713127136, "learning_rate": 4.6524287245729295e-05, "loss": 2.2857, "step": 2326 }, { "epoch": 0.02092879980932937, "grad_norm": 0.6379580497741699, "learning_rate": 4.652046335157231e-05, "loss": 2.2246, "step": 2327 }, { "epoch": 0.020937793706969823, "grad_norm": 0.6498323082923889, "learning_rate": 4.651663751241903e-05, "loss": 2.2383, "step": 2328 }, { "epoch": 0.020946787604610273, "grad_norm": 0.964203417301178, "learning_rate": 4.651280972861521e-05, "loss": 2.237, "step": 2329 }, { "epoch": 0.020955781502250723, "grad_norm": 1.2168457508087158, "learning_rate": 4.650898000050684e-05, "loss": 2.252, "step": 2330 }, { "epoch": 0.020964775399891174, "grad_norm": 1.6756229400634766, "learning_rate": 4.650514832844002e-05, "loss": 2.2529, "step": 2331 }, { "epoch": 0.020973769297531624, "grad_norm": 0.7146030068397522, "learning_rate": 4.6501314712761054e-05, "loss": 2.2332, "step": 2332 }, { "epoch": 0.020982763195172074, "grad_norm": 1.8755948543548584, "learning_rate": 4.649747915381643e-05, "loss": 2.2888, "step": 2333 }, { "epoch": 0.02099175709281253, "grad_norm": 1.340332269668579, "learning_rate": 4.649364165195279e-05, "loss": 2.3014, "step": 2334 }, { "epoch": 0.02100075099045298, "grad_norm": 1.3762927055358887, "learning_rate": 4.648980220751696e-05, "loss": 2.2052, "step": 2335 }, { "epoch": 0.02100974488809343, "grad_norm": 0.8822158575057983, "learning_rate": 4.648596082085597e-05, "loss": 2.2069, "step": 2336 }, { "epoch": 0.02101873878573388, "grad_norm": 0.7100244164466858, "learning_rate": 4.648211749231698e-05, "loss": 2.2107, "step": 2337 }, { "epoch": 0.02102773268337433, "grad_norm": 0.7201721668243408, "learning_rate": 4.6478272222247346e-05, "loss": 2.2748, "step": 2338 }, { "epoch": 0.02103672658101478, "grad_norm": 0.8511122465133667, "learning_rate": 4.64744250109946e-05, "loss": 2.179, "step": 2339 }, { "epoch": 0.021045720478655234, "grad_norm": 0.5459071397781372, "learning_rate": 4.647057585890646e-05, "loss": 2.2068, "step": 2340 }, { "epoch": 0.021054714376295684, "grad_norm": 0.6787444353103638, "learning_rate": 4.646672476633079e-05, "loss": 2.3003, "step": 2341 }, { "epoch": 0.021063708273936135, "grad_norm": 0.9851678013801575, "learning_rate": 4.646287173361566e-05, "loss": 2.2709, "step": 2342 }, { "epoch": 0.021072702171576585, "grad_norm": 1.2711679935455322, "learning_rate": 4.645901676110931e-05, "loss": 2.2155, "step": 2343 }, { "epoch": 0.021081696069217035, "grad_norm": 0.9089391827583313, "learning_rate": 4.645515984916013e-05, "loss": 2.1684, "step": 2344 }, { "epoch": 0.021090689966857486, "grad_norm": 0.8341915607452393, "learning_rate": 4.645130099811671e-05, "loss": 2.2106, "step": 2345 }, { "epoch": 0.02109968386449794, "grad_norm": 2.038611650466919, "learning_rate": 4.644744020832782e-05, "loss": 2.2873, "step": 2346 }, { "epoch": 0.02110867776213839, "grad_norm": 0.7331545352935791, "learning_rate": 4.644357748014238e-05, "loss": 2.2409, "step": 2347 }, { "epoch": 0.02111767165977884, "grad_norm": 0.7537279725074768, "learning_rate": 4.643971281390951e-05, "loss": 2.2718, "step": 2348 }, { "epoch": 0.02112666555741929, "grad_norm": 0.9624699354171753, "learning_rate": 4.6435846209978484e-05, "loss": 2.2073, "step": 2349 }, { "epoch": 0.02113565945505974, "grad_norm": 0.9245498180389404, "learning_rate": 4.643197766869877e-05, "loss": 2.3417, "step": 2350 }, { "epoch": 0.02114465335270019, "grad_norm": 0.8140974640846252, "learning_rate": 4.642810719041999e-05, "loss": 2.2553, "step": 2351 }, { "epoch": 0.021153647250340645, "grad_norm": 0.8038978576660156, "learning_rate": 4.642423477549197e-05, "loss": 2.2359, "step": 2352 }, { "epoch": 0.021162641147981096, "grad_norm": 0.6892197728157043, "learning_rate": 4.6420360424264686e-05, "loss": 2.2389, "step": 2353 }, { "epoch": 0.021171635045621546, "grad_norm": 0.6683658361434937, "learning_rate": 4.641648413708829e-05, "loss": 2.2103, "step": 2354 }, { "epoch": 0.021180628943261996, "grad_norm": 0.9551239013671875, "learning_rate": 4.6412605914313144e-05, "loss": 2.383, "step": 2355 }, { "epoch": 0.021189622840902447, "grad_norm": 0.6828892230987549, "learning_rate": 4.640872575628973e-05, "loss": 2.2525, "step": 2356 }, { "epoch": 0.021198616738542897, "grad_norm": 1.6641943454742432, "learning_rate": 4.640484366336874e-05, "loss": 2.2238, "step": 2357 }, { "epoch": 0.02120761063618335, "grad_norm": 2.0551323890686035, "learning_rate": 4.640095963590103e-05, "loss": 2.2678, "step": 2358 }, { "epoch": 0.0212166045338238, "grad_norm": 1.018084168434143, "learning_rate": 4.6397073674237646e-05, "loss": 2.2821, "step": 2359 }, { "epoch": 0.02122559843146425, "grad_norm": 0.989090621471405, "learning_rate": 4.6393185778729797e-05, "loss": 2.2107, "step": 2360 }, { "epoch": 0.021234592329104702, "grad_norm": 0.6363390684127808, "learning_rate": 4.638929594972885e-05, "loss": 2.182, "step": 2361 }, { "epoch": 0.021243586226745152, "grad_norm": 1.4775336980819702, "learning_rate": 4.638540418758638e-05, "loss": 2.2547, "step": 2362 }, { "epoch": 0.021252580124385606, "grad_norm": 0.9519386291503906, "learning_rate": 4.638151049265411e-05, "loss": 2.2659, "step": 2363 }, { "epoch": 0.021261574022026056, "grad_norm": 1.448249340057373, "learning_rate": 4.637761486528395e-05, "loss": 2.2301, "step": 2364 }, { "epoch": 0.021270567919666507, "grad_norm": 1.1718530654907227, "learning_rate": 4.637371730582799e-05, "loss": 2.2586, "step": 2365 }, { "epoch": 0.021279561817306957, "grad_norm": 0.6762630343437195, "learning_rate": 4.6369817814638475e-05, "loss": 2.2485, "step": 2366 }, { "epoch": 0.021288555714947408, "grad_norm": 1.4585444927215576, "learning_rate": 4.6365916392067854e-05, "loss": 2.2709, "step": 2367 }, { "epoch": 0.021297549612587858, "grad_norm": 0.9264916181564331, "learning_rate": 4.636201303846871e-05, "loss": 2.2145, "step": 2368 }, { "epoch": 0.02130654351022831, "grad_norm": 0.9253841042518616, "learning_rate": 4.6358107754193845e-05, "loss": 2.2309, "step": 2369 }, { "epoch": 0.021315537407868762, "grad_norm": 0.975245475769043, "learning_rate": 4.6354200539596204e-05, "loss": 2.2407, "step": 2370 }, { "epoch": 0.021324531305509212, "grad_norm": 1.1457244157791138, "learning_rate": 4.6350291395028924e-05, "loss": 2.2391, "step": 2371 }, { "epoch": 0.021333525203149663, "grad_norm": 1.1978095769882202, "learning_rate": 4.634638032084529e-05, "loss": 2.266, "step": 2372 }, { "epoch": 0.021342519100790113, "grad_norm": 0.6161963939666748, "learning_rate": 4.634246731739881e-05, "loss": 2.2444, "step": 2373 }, { "epoch": 0.021351512998430563, "grad_norm": 1.6567800045013428, "learning_rate": 4.633855238504312e-05, "loss": 2.3361, "step": 2374 }, { "epoch": 0.021360506896071017, "grad_norm": 0.7269092798233032, "learning_rate": 4.6334635524132046e-05, "loss": 2.2811, "step": 2375 }, { "epoch": 0.021369500793711468, "grad_norm": 0.8117492198944092, "learning_rate": 4.63307167350196e-05, "loss": 2.2469, "step": 2376 }, { "epoch": 0.021378494691351918, "grad_norm": 0.8311796188354492, "learning_rate": 4.6326796018059936e-05, "loss": 2.2317, "step": 2377 }, { "epoch": 0.02138748858899237, "grad_norm": 1.207515001296997, "learning_rate": 4.632287337360743e-05, "loss": 2.3174, "step": 2378 }, { "epoch": 0.02139648248663282, "grad_norm": 1.5077379941940308, "learning_rate": 4.6318948802016594e-05, "loss": 2.2187, "step": 2379 }, { "epoch": 0.02140547638427327, "grad_norm": 0.7120034694671631, "learning_rate": 4.631502230364213e-05, "loss": 2.1883, "step": 2380 }, { "epoch": 0.021414470281913723, "grad_norm": 1.0836539268493652, "learning_rate": 4.6311093878838904e-05, "loss": 2.2348, "step": 2381 }, { "epoch": 0.021423464179554173, "grad_norm": 0.8691428899765015, "learning_rate": 4.630716352796197e-05, "loss": 2.2094, "step": 2382 }, { "epoch": 0.021432458077194624, "grad_norm": 1.982783317565918, "learning_rate": 4.6303231251366553e-05, "loss": 2.3301, "step": 2383 }, { "epoch": 0.021441451974835074, "grad_norm": 1.250273585319519, "learning_rate": 4.629929704940803e-05, "loss": 2.3091, "step": 2384 }, { "epoch": 0.021450445872475524, "grad_norm": 1.0423836708068848, "learning_rate": 4.629536092244199e-05, "loss": 2.2455, "step": 2385 }, { "epoch": 0.021459439770115975, "grad_norm": 1.1859381198883057, "learning_rate": 4.629142287082417e-05, "loss": 2.3021, "step": 2386 }, { "epoch": 0.02146843366775643, "grad_norm": 0.71415776014328, "learning_rate": 4.628748289491047e-05, "loss": 2.2061, "step": 2387 }, { "epoch": 0.02147742756539688, "grad_norm": 0.8777947425842285, "learning_rate": 4.6283540995057004e-05, "loss": 2.3139, "step": 2388 }, { "epoch": 0.02148642146303733, "grad_norm": 1.4611034393310547, "learning_rate": 4.6279597171620024e-05, "loss": 2.2528, "step": 2389 }, { "epoch": 0.02149541536067778, "grad_norm": 0.9472832083702087, "learning_rate": 4.6275651424955976e-05, "loss": 2.2364, "step": 2390 }, { "epoch": 0.02150440925831823, "grad_norm": 1.0596520900726318, "learning_rate": 4.627170375542147e-05, "loss": 2.3269, "step": 2391 }, { "epoch": 0.02151340315595868, "grad_norm": 1.184664249420166, "learning_rate": 4.626775416337329e-05, "loss": 2.225, "step": 2392 }, { "epoch": 0.021522397053599134, "grad_norm": 1.1677136421203613, "learning_rate": 4.626380264916839e-05, "loss": 2.2367, "step": 2393 }, { "epoch": 0.021531390951239585, "grad_norm": 0.8704460859298706, "learning_rate": 4.625984921316392e-05, "loss": 2.2238, "step": 2394 }, { "epoch": 0.021540384848880035, "grad_norm": 1.036882758140564, "learning_rate": 4.625589385571717e-05, "loss": 2.2676, "step": 2395 }, { "epoch": 0.021549378746520485, "grad_norm": 1.7169899940490723, "learning_rate": 4.6251936577185627e-05, "loss": 2.3373, "step": 2396 }, { "epoch": 0.021558372644160936, "grad_norm": 2.0544495582580566, "learning_rate": 4.624797737792696e-05, "loss": 2.1965, "step": 2397 }, { "epoch": 0.02156736654180139, "grad_norm": 1.1191158294677734, "learning_rate": 4.624401625829897e-05, "loss": 2.279, "step": 2398 }, { "epoch": 0.02157636043944184, "grad_norm": 1.2953102588653564, "learning_rate": 4.6240053218659674e-05, "loss": 2.2681, "step": 2399 }, { "epoch": 0.02158535433708229, "grad_norm": 0.9371036291122437, "learning_rate": 4.623608825936725e-05, "loss": 2.2541, "step": 2400 }, { "epoch": 0.02159434823472274, "grad_norm": 1.141536831855774, "learning_rate": 4.6232121380780034e-05, "loss": 2.2278, "step": 2401 }, { "epoch": 0.02160334213236319, "grad_norm": 0.9859936237335205, "learning_rate": 4.622815258325657e-05, "loss": 2.2253, "step": 2402 }, { "epoch": 0.02161233603000364, "grad_norm": 1.1577844619750977, "learning_rate": 4.6224181867155535e-05, "loss": 2.2451, "step": 2403 }, { "epoch": 0.021621329927644095, "grad_norm": 0.9990515112876892, "learning_rate": 4.62202092328358e-05, "loss": 2.1978, "step": 2404 }, { "epoch": 0.021630323825284545, "grad_norm": 0.9159051775932312, "learning_rate": 4.621623468065642e-05, "loss": 2.2474, "step": 2405 }, { "epoch": 0.021639317722924996, "grad_norm": 0.7199147343635559, "learning_rate": 4.621225821097659e-05, "loss": 2.2941, "step": 2406 }, { "epoch": 0.021648311620565446, "grad_norm": 1.3560916185379028, "learning_rate": 4.620827982415572e-05, "loss": 2.3239, "step": 2407 }, { "epoch": 0.021657305518205897, "grad_norm": 0.6104757189750671, "learning_rate": 4.6204299520553366e-05, "loss": 2.2537, "step": 2408 }, { "epoch": 0.021666299415846347, "grad_norm": 0.7127137780189514, "learning_rate": 4.620031730052925e-05, "loss": 2.2331, "step": 2409 }, { "epoch": 0.0216752933134868, "grad_norm": 0.6301649808883667, "learning_rate": 4.6196333164443295e-05, "loss": 2.2383, "step": 2410 }, { "epoch": 0.02168428721112725, "grad_norm": 0.9867540001869202, "learning_rate": 4.619234711265558e-05, "loss": 2.3066, "step": 2411 }, { "epoch": 0.0216932811087677, "grad_norm": 1.131047248840332, "learning_rate": 4.618835914552636e-05, "loss": 2.2745, "step": 2412 }, { "epoch": 0.021702275006408152, "grad_norm": 0.7229920029640198, "learning_rate": 4.6184369263416065e-05, "loss": 2.2262, "step": 2413 }, { "epoch": 0.021711268904048602, "grad_norm": 1.1453962326049805, "learning_rate": 4.61803774666853e-05, "loss": 2.2821, "step": 2414 }, { "epoch": 0.021720262801689053, "grad_norm": 0.8127686381340027, "learning_rate": 4.617638375569482e-05, "loss": 2.2728, "step": 2415 }, { "epoch": 0.021729256699329506, "grad_norm": 1.263199806213379, "learning_rate": 4.617238813080559e-05, "loss": 2.2057, "step": 2416 }, { "epoch": 0.021738250596969957, "grad_norm": 0.8185906410217285, "learning_rate": 4.616839059237873e-05, "loss": 2.1911, "step": 2417 }, { "epoch": 0.021747244494610407, "grad_norm": 0.9190717935562134, "learning_rate": 4.616439114077552e-05, "loss": 2.2555, "step": 2418 }, { "epoch": 0.021756238392250857, "grad_norm": 1.1180697679519653, "learning_rate": 4.616038977635744e-05, "loss": 2.2448, "step": 2419 }, { "epoch": 0.021765232289891308, "grad_norm": 1.5997012853622437, "learning_rate": 4.6156386499486126e-05, "loss": 2.2896, "step": 2420 }, { "epoch": 0.021774226187531758, "grad_norm": 1.8077616691589355, "learning_rate": 4.6152381310523387e-05, "loss": 2.2255, "step": 2421 }, { "epoch": 0.021783220085172212, "grad_norm": 0.7581536173820496, "learning_rate": 4.614837420983121e-05, "loss": 2.2224, "step": 2422 }, { "epoch": 0.021792213982812662, "grad_norm": 1.212648868560791, "learning_rate": 4.614436519777175e-05, "loss": 2.2551, "step": 2423 }, { "epoch": 0.021801207880453113, "grad_norm": 1.5475356578826904, "learning_rate": 4.614035427470734e-05, "loss": 2.3298, "step": 2424 }, { "epoch": 0.021810201778093563, "grad_norm": 1.1770330667495728, "learning_rate": 4.613634144100047e-05, "loss": 2.2851, "step": 2425 }, { "epoch": 0.021819195675734013, "grad_norm": 1.1748696565628052, "learning_rate": 4.613232669701384e-05, "loss": 2.3658, "step": 2426 }, { "epoch": 0.021828189573374464, "grad_norm": 1.0885498523712158, "learning_rate": 4.612831004311028e-05, "loss": 2.2426, "step": 2427 }, { "epoch": 0.021837183471014918, "grad_norm": 0.9773613810539246, "learning_rate": 4.612429147965281e-05, "loss": 2.2741, "step": 2428 }, { "epoch": 0.021846177368655368, "grad_norm": 0.8444312214851379, "learning_rate": 4.6120271007004644e-05, "loss": 2.2963, "step": 2429 }, { "epoch": 0.02185517126629582, "grad_norm": 1.1682432889938354, "learning_rate": 4.6116248625529124e-05, "loss": 2.3036, "step": 2430 }, { "epoch": 0.02186416516393627, "grad_norm": 0.8337119817733765, "learning_rate": 4.61122243355898e-05, "loss": 2.2595, "step": 2431 }, { "epoch": 0.02187315906157672, "grad_norm": 1.8767850399017334, "learning_rate": 4.610819813755038e-05, "loss": 2.2302, "step": 2432 }, { "epoch": 0.02188215295921717, "grad_norm": 0.6706061959266663, "learning_rate": 4.610417003177475e-05, "loss": 2.2549, "step": 2433 }, { "epoch": 0.021891146856857623, "grad_norm": 1.3031163215637207, "learning_rate": 4.610014001862696e-05, "loss": 2.2468, "step": 2434 }, { "epoch": 0.021900140754498074, "grad_norm": 1.4461766481399536, "learning_rate": 4.609610809847125e-05, "loss": 2.2294, "step": 2435 }, { "epoch": 0.021909134652138524, "grad_norm": 0.7525919675827026, "learning_rate": 4.6092074271672e-05, "loss": 2.2836, "step": 2436 }, { "epoch": 0.021918128549778974, "grad_norm": 0.6679876446723938, "learning_rate": 4.60880385385938e-05, "loss": 2.2092, "step": 2437 }, { "epoch": 0.021927122447419425, "grad_norm": 0.949408769607544, "learning_rate": 4.60840008996014e-05, "loss": 2.2405, "step": 2438 }, { "epoch": 0.02193611634505988, "grad_norm": 1.3952058553695679, "learning_rate": 4.60799613550597e-05, "loss": 2.253, "step": 2439 }, { "epoch": 0.02194511024270033, "grad_norm": 1.5070608854293823, "learning_rate": 4.60759199053338e-05, "loss": 2.2592, "step": 2440 }, { "epoch": 0.02195410414034078, "grad_norm": 1.6125280857086182, "learning_rate": 4.6071876550788964e-05, "loss": 2.298, "step": 2441 }, { "epoch": 0.02196309803798123, "grad_norm": 1.1626113653182983, "learning_rate": 4.606783129179062e-05, "loss": 2.2481, "step": 2442 }, { "epoch": 0.02197209193562168, "grad_norm": 2.1960456371307373, "learning_rate": 4.6063784128704367e-05, "loss": 2.2754, "step": 2443 }, { "epoch": 0.02198108583326213, "grad_norm": 0.976567804813385, "learning_rate": 4.6059735061896e-05, "loss": 2.3605, "step": 2444 }, { "epoch": 0.021990079730902584, "grad_norm": 1.1528140306472778, "learning_rate": 4.605568409173147e-05, "loss": 2.1756, "step": 2445 }, { "epoch": 0.021999073628543035, "grad_norm": 0.9914456009864807, "learning_rate": 4.605163121857688e-05, "loss": 2.2079, "step": 2446 }, { "epoch": 0.022008067526183485, "grad_norm": 1.0502045154571533, "learning_rate": 4.604757644279854e-05, "loss": 2.269, "step": 2447 }, { "epoch": 0.022017061423823935, "grad_norm": 0.8106114268302917, "learning_rate": 4.604351976476291e-05, "loss": 2.2711, "step": 2448 }, { "epoch": 0.022026055321464386, "grad_norm": 1.1567299365997314, "learning_rate": 4.603946118483663e-05, "loss": 2.2162, "step": 2449 }, { "epoch": 0.022035049219104836, "grad_norm": 0.9873222708702087, "learning_rate": 4.603540070338651e-05, "loss": 2.227, "step": 2450 }, { "epoch": 0.02204404311674529, "grad_norm": 1.212885856628418, "learning_rate": 4.6031338320779534e-05, "loss": 2.4176, "step": 2451 }, { "epoch": 0.02205303701438574, "grad_norm": 0.5234179496765137, "learning_rate": 4.602727403738285e-05, "loss": 2.2308, "step": 2452 }, { "epoch": 0.02206203091202619, "grad_norm": 1.3271576166152954, "learning_rate": 4.602320785356379e-05, "loss": 2.2608, "step": 2453 }, { "epoch": 0.02207102480966664, "grad_norm": 1.5926169157028198, "learning_rate": 4.601913976968985e-05, "loss": 2.3979, "step": 2454 }, { "epoch": 0.02208001870730709, "grad_norm": 1.3101963996887207, "learning_rate": 4.60150697861287e-05, "loss": 2.273, "step": 2455 }, { "epoch": 0.02208901260494754, "grad_norm": 0.7954499125480652, "learning_rate": 4.601099790324817e-05, "loss": 2.2524, "step": 2456 }, { "epoch": 0.022098006502587995, "grad_norm": 1.0100698471069336, "learning_rate": 4.600692412141628e-05, "loss": 2.22, "step": 2457 }, { "epoch": 0.022107000400228446, "grad_norm": 1.2023860216140747, "learning_rate": 4.600284844100122e-05, "loss": 2.2569, "step": 2458 }, { "epoch": 0.022115994297868896, "grad_norm": 0.7265771627426147, "learning_rate": 4.599877086237134e-05, "loss": 2.2551, "step": 2459 }, { "epoch": 0.022124988195509347, "grad_norm": 0.9623656272888184, "learning_rate": 4.5994691385895164e-05, "loss": 2.2875, "step": 2460 }, { "epoch": 0.022133982093149797, "grad_norm": 1.0737924575805664, "learning_rate": 4.5990610011941396e-05, "loss": 2.2687, "step": 2461 }, { "epoch": 0.022142975990790247, "grad_norm": 1.1508426666259766, "learning_rate": 4.5986526740878904e-05, "loss": 2.2963, "step": 2462 }, { "epoch": 0.0221519698884307, "grad_norm": 1.0979071855545044, "learning_rate": 4.598244157307673e-05, "loss": 2.2345, "step": 2463 }, { "epoch": 0.02216096378607115, "grad_norm": 0.7382902503013611, "learning_rate": 4.597835450890409e-05, "loss": 2.2307, "step": 2464 }, { "epoch": 0.022169957683711602, "grad_norm": 0.6234745383262634, "learning_rate": 4.597426554873037e-05, "loss": 2.2033, "step": 2465 }, { "epoch": 0.022178951581352052, "grad_norm": 1.0507071018218994, "learning_rate": 4.5970174692925106e-05, "loss": 2.3109, "step": 2466 }, { "epoch": 0.022187945478992502, "grad_norm": 1.4410667419433594, "learning_rate": 4.596608194185805e-05, "loss": 2.2408, "step": 2467 }, { "epoch": 0.022196939376632953, "grad_norm": 1.038637399673462, "learning_rate": 4.59619872958991e-05, "loss": 2.182, "step": 2468 }, { "epoch": 0.022205933274273407, "grad_norm": 0.918613612651825, "learning_rate": 4.59578907554183e-05, "loss": 2.304, "step": 2469 }, { "epoch": 0.022214927171913857, "grad_norm": 0.8182939291000366, "learning_rate": 4.595379232078592e-05, "loss": 2.2524, "step": 2470 }, { "epoch": 0.022223921069554307, "grad_norm": 1.1375149488449097, "learning_rate": 4.594969199237235e-05, "loss": 2.2709, "step": 2471 }, { "epoch": 0.022232914967194758, "grad_norm": 0.8590173125267029, "learning_rate": 4.594558977054819e-05, "loss": 2.255, "step": 2472 }, { "epoch": 0.022241908864835208, "grad_norm": 0.9720834493637085, "learning_rate": 4.594148565568418e-05, "loss": 2.2804, "step": 2473 }, { "epoch": 0.022250902762475662, "grad_norm": 1.0119317770004272, "learning_rate": 4.593737964815126e-05, "loss": 2.1655, "step": 2474 }, { "epoch": 0.022259896660116112, "grad_norm": 2.301095485687256, "learning_rate": 4.593327174832052e-05, "loss": 2.2773, "step": 2475 }, { "epoch": 0.022268890557756563, "grad_norm": 0.8834993839263916, "learning_rate": 4.592916195656322e-05, "loss": 2.2404, "step": 2476 }, { "epoch": 0.022277884455397013, "grad_norm": 1.3549675941467285, "learning_rate": 4.5925050273250804e-05, "loss": 2.2716, "step": 2477 }, { "epoch": 0.022286878353037463, "grad_norm": 0.7502055764198303, "learning_rate": 4.592093669875489e-05, "loss": 2.1591, "step": 2478 }, { "epoch": 0.022295872250677914, "grad_norm": 0.9009088277816772, "learning_rate": 4.591682123344725e-05, "loss": 2.246, "step": 2479 }, { "epoch": 0.022304866148318368, "grad_norm": 0.8956072330474854, "learning_rate": 4.5912703877699836e-05, "loss": 2.2368, "step": 2480 }, { "epoch": 0.022313860045958818, "grad_norm": 1.073777675628662, "learning_rate": 4.590858463188477e-05, "loss": 2.2327, "step": 2481 }, { "epoch": 0.02232285394359927, "grad_norm": 1.652058720588684, "learning_rate": 4.5904463496374344e-05, "loss": 2.163, "step": 2482 }, { "epoch": 0.02233184784123972, "grad_norm": 0.9745768308639526, "learning_rate": 4.590034047154103e-05, "loss": 2.2075, "step": 2483 }, { "epoch": 0.02234084173888017, "grad_norm": 1.6262750625610352, "learning_rate": 4.589621555775745e-05, "loss": 2.264, "step": 2484 }, { "epoch": 0.02234983563652062, "grad_norm": 1.239117980003357, "learning_rate": 4.589208875539641e-05, "loss": 2.1744, "step": 2485 }, { "epoch": 0.022358829534161073, "grad_norm": 0.5564536452293396, "learning_rate": 4.58879600648309e-05, "loss": 2.24, "step": 2486 }, { "epoch": 0.022367823431801524, "grad_norm": 1.1510002613067627, "learning_rate": 4.588382948643406e-05, "loss": 2.2732, "step": 2487 }, { "epoch": 0.022376817329441974, "grad_norm": 1.0595316886901855, "learning_rate": 4.5879697020579207e-05, "loss": 2.299, "step": 2488 }, { "epoch": 0.022385811227082424, "grad_norm": 0.8138825297355652, "learning_rate": 4.587556266763981e-05, "loss": 2.1529, "step": 2489 }, { "epoch": 0.022394805124722875, "grad_norm": 1.7369400262832642, "learning_rate": 4.587142642798956e-05, "loss": 2.2892, "step": 2490 }, { "epoch": 0.022403799022363325, "grad_norm": 0.9728208184242249, "learning_rate": 4.586728830200227e-05, "loss": 2.2401, "step": 2491 }, { "epoch": 0.02241279292000378, "grad_norm": 0.8631722927093506, "learning_rate": 4.586314829005194e-05, "loss": 2.2637, "step": 2492 }, { "epoch": 0.02242178681764423, "grad_norm": 1.6548008918762207, "learning_rate": 4.585900639251273e-05, "loss": 2.2009, "step": 2493 }, { "epoch": 0.02243078071528468, "grad_norm": 1.578078031539917, "learning_rate": 4.5854862609759e-05, "loss": 2.1976, "step": 2494 }, { "epoch": 0.02243977461292513, "grad_norm": 1.4733911752700806, "learning_rate": 4.5850716942165245e-05, "loss": 2.19, "step": 2495 }, { "epoch": 0.02244876851056558, "grad_norm": 1.2819656133651733, "learning_rate": 4.584656939010615e-05, "loss": 2.2529, "step": 2496 }, { "epoch": 0.02245776240820603, "grad_norm": 1.1497126817703247, "learning_rate": 4.584241995395657e-05, "loss": 2.3063, "step": 2497 }, { "epoch": 0.022466756305846484, "grad_norm": 1.979767918586731, "learning_rate": 4.5838268634091524e-05, "loss": 2.3157, "step": 2498 }, { "epoch": 0.022475750203486935, "grad_norm": 1.2814713716506958, "learning_rate": 4.5834115430886206e-05, "loss": 2.2581, "step": 2499 }, { "epoch": 0.022484744101127385, "grad_norm": 1.1708576679229736, "learning_rate": 4.582996034471597e-05, "loss": 2.2758, "step": 2500 }, { "epoch": 0.022493737998767836, "grad_norm": 1.1838024854660034, "learning_rate": 4.582580337595636e-05, "loss": 2.2262, "step": 2501 }, { "epoch": 0.022502731896408286, "grad_norm": 0.8205706477165222, "learning_rate": 4.5821644524983074e-05, "loss": 2.1926, "step": 2502 }, { "epoch": 0.022511725794048736, "grad_norm": 1.034786343574524, "learning_rate": 4.581748379217198e-05, "loss": 2.2258, "step": 2503 }, { "epoch": 0.02252071969168919, "grad_norm": 2.0054707527160645, "learning_rate": 4.581332117789912e-05, "loss": 2.304, "step": 2504 }, { "epoch": 0.02252971358932964, "grad_norm": 1.4698878526687622, "learning_rate": 4.5809156682540716e-05, "loss": 2.2415, "step": 2505 }, { "epoch": 0.02253870748697009, "grad_norm": 1.8093228340148926, "learning_rate": 4.580499030647314e-05, "loss": 2.2856, "step": 2506 }, { "epoch": 0.02254770138461054, "grad_norm": 0.9136622548103333, "learning_rate": 4.580082205007295e-05, "loss": 2.2393, "step": 2507 }, { "epoch": 0.02255669528225099, "grad_norm": 1.3996533155441284, "learning_rate": 4.579665191371687e-05, "loss": 2.259, "step": 2508 }, { "epoch": 0.022565689179891445, "grad_norm": 1.1649353504180908, "learning_rate": 4.579247989778179e-05, "loss": 2.238, "step": 2509 }, { "epoch": 0.022574683077531896, "grad_norm": 1.3719139099121094, "learning_rate": 4.5788306002644775e-05, "loss": 2.2097, "step": 2510 }, { "epoch": 0.022583676975172346, "grad_norm": 0.6472741365432739, "learning_rate": 4.5784130228683056e-05, "loss": 2.2299, "step": 2511 }, { "epoch": 0.022592670872812796, "grad_norm": 1.0743552446365356, "learning_rate": 4.577995257627403e-05, "loss": 2.225, "step": 2512 }, { "epoch": 0.022601664770453247, "grad_norm": 1.2717546224594116, "learning_rate": 4.5775773045795266e-05, "loss": 2.2432, "step": 2513 }, { "epoch": 0.022610658668093697, "grad_norm": 1.3228983879089355, "learning_rate": 4.5771591637624525e-05, "loss": 2.328, "step": 2514 }, { "epoch": 0.02261965256573415, "grad_norm": 2.2070436477661133, "learning_rate": 4.57674083521397e-05, "loss": 2.3074, "step": 2515 }, { "epoch": 0.0226286464633746, "grad_norm": 1.2498551607131958, "learning_rate": 4.576322318971887e-05, "loss": 2.2014, "step": 2516 }, { "epoch": 0.02263764036101505, "grad_norm": 1.386706829071045, "learning_rate": 4.57590361507403e-05, "loss": 2.1787, "step": 2517 }, { "epoch": 0.022646634258655502, "grad_norm": 0.6084449887275696, "learning_rate": 4.57548472355824e-05, "loss": 2.2226, "step": 2518 }, { "epoch": 0.022655628156295952, "grad_norm": 0.8788933157920837, "learning_rate": 4.5750656444623764e-05, "loss": 2.2693, "step": 2519 }, { "epoch": 0.022664622053936403, "grad_norm": 1.1350479125976562, "learning_rate": 4.574646377824315e-05, "loss": 2.2905, "step": 2520 }, { "epoch": 0.022673615951576857, "grad_norm": 1.170608639717102, "learning_rate": 4.574226923681948e-05, "loss": 2.2435, "step": 2521 }, { "epoch": 0.022682609849217307, "grad_norm": 1.4048583507537842, "learning_rate": 4.573807282073186e-05, "loss": 2.2109, "step": 2522 }, { "epoch": 0.022691603746857757, "grad_norm": 0.9129926562309265, "learning_rate": 4.5733874530359565e-05, "loss": 2.2214, "step": 2523 }, { "epoch": 0.022700597644498208, "grad_norm": 0.673080563545227, "learning_rate": 4.5729674366082015e-05, "loss": 2.2324, "step": 2524 }, { "epoch": 0.022709591542138658, "grad_norm": 1.585649847984314, "learning_rate": 4.5725472328278825e-05, "loss": 2.2117, "step": 2525 }, { "epoch": 0.02271858543977911, "grad_norm": 0.8523921370506287, "learning_rate": 4.572126841732976e-05, "loss": 2.2719, "step": 2526 }, { "epoch": 0.022727579337419562, "grad_norm": 1.2965589761734009, "learning_rate": 4.5717062633614796e-05, "loss": 2.3182, "step": 2527 }, { "epoch": 0.022736573235060013, "grad_norm": 0.7020949721336365, "learning_rate": 4.571285497751401e-05, "loss": 2.2515, "step": 2528 }, { "epoch": 0.022745567132700463, "grad_norm": 0.9996892213821411, "learning_rate": 4.5708645449407704e-05, "loss": 2.2471, "step": 2529 }, { "epoch": 0.022754561030340913, "grad_norm": 0.6850045323371887, "learning_rate": 4.570443404967632e-05, "loss": 2.2311, "step": 2530 }, { "epoch": 0.022763554927981364, "grad_norm": 1.3833266496658325, "learning_rate": 4.5700220778700504e-05, "loss": 2.3024, "step": 2531 }, { "epoch": 0.022772548825621814, "grad_norm": 1.2717101573944092, "learning_rate": 4.5696005636861025e-05, "loss": 2.2217, "step": 2532 }, { "epoch": 0.022781542723262268, "grad_norm": 1.463228702545166, "learning_rate": 4.569178862453885e-05, "loss": 2.2839, "step": 2533 }, { "epoch": 0.022790536620902718, "grad_norm": 0.9617488980293274, "learning_rate": 4.568756974211509e-05, "loss": 2.1448, "step": 2534 }, { "epoch": 0.02279953051854317, "grad_norm": 1.5313831567764282, "learning_rate": 4.568334898997107e-05, "loss": 2.3163, "step": 2535 }, { "epoch": 0.02280852441618362, "grad_norm": 0.9104557633399963, "learning_rate": 4.567912636848826e-05, "loss": 2.2596, "step": 2536 }, { "epoch": 0.02281751831382407, "grad_norm": 0.9424986243247986, "learning_rate": 4.567490187804827e-05, "loss": 2.2443, "step": 2537 }, { "epoch": 0.02282651221146452, "grad_norm": 0.5520268082618713, "learning_rate": 4.567067551903291e-05, "loss": 2.2307, "step": 2538 }, { "epoch": 0.022835506109104974, "grad_norm": 1.104384422302246, "learning_rate": 4.566644729182418e-05, "loss": 2.1736, "step": 2539 }, { "epoch": 0.022844500006745424, "grad_norm": 0.965882420539856, "learning_rate": 4.566221719680419e-05, "loss": 2.2103, "step": 2540 }, { "epoch": 0.022853493904385874, "grad_norm": 1.3724310398101807, "learning_rate": 4.5657985234355275e-05, "loss": 2.2025, "step": 2541 }, { "epoch": 0.022862487802026325, "grad_norm": 0.7048248648643494, "learning_rate": 4.565375140485989e-05, "loss": 2.1743, "step": 2542 }, { "epoch": 0.022871481699666775, "grad_norm": 1.675199270248413, "learning_rate": 4.564951570870071e-05, "loss": 2.2242, "step": 2543 }, { "epoch": 0.022880475597307225, "grad_norm": 1.0177119970321655, "learning_rate": 4.564527814626054e-05, "loss": 2.249, "step": 2544 }, { "epoch": 0.02288946949494768, "grad_norm": 1.3488800525665283, "learning_rate": 4.564103871792237e-05, "loss": 2.2911, "step": 2545 }, { "epoch": 0.02289846339258813, "grad_norm": 1.0901708602905273, "learning_rate": 4.563679742406935e-05, "loss": 2.2741, "step": 2546 }, { "epoch": 0.02290745729022858, "grad_norm": 0.9563313126564026, "learning_rate": 4.563255426508481e-05, "loss": 2.233, "step": 2547 }, { "epoch": 0.02291645118786903, "grad_norm": 0.8600868582725525, "learning_rate": 4.562830924135224e-05, "loss": 2.249, "step": 2548 }, { "epoch": 0.02292544508550948, "grad_norm": 0.7287613153457642, "learning_rate": 4.56240623532553e-05, "loss": 2.2215, "step": 2549 }, { "epoch": 0.022934438983149934, "grad_norm": 1.1409497261047363, "learning_rate": 4.5619813601177815e-05, "loss": 2.3441, "step": 2550 }, { "epoch": 0.022943432880790385, "grad_norm": 0.8541669249534607, "learning_rate": 4.561556298550379e-05, "loss": 2.2254, "step": 2551 }, { "epoch": 0.022952426778430835, "grad_norm": 0.5671873688697815, "learning_rate": 4.561131050661738e-05, "loss": 2.2263, "step": 2552 }, { "epoch": 0.022961420676071286, "grad_norm": 0.9204468131065369, "learning_rate": 4.560705616490294e-05, "loss": 2.2138, "step": 2553 }, { "epoch": 0.022970414573711736, "grad_norm": 1.1867889165878296, "learning_rate": 4.560279996074495e-05, "loss": 2.2312, "step": 2554 }, { "epoch": 0.022979408471352186, "grad_norm": 1.6452332735061646, "learning_rate": 4.5598541894528097e-05, "loss": 2.2443, "step": 2555 }, { "epoch": 0.02298840236899264, "grad_norm": 1.0195550918579102, "learning_rate": 4.559428196663721e-05, "loss": 2.2259, "step": 2556 }, { "epoch": 0.02299739626663309, "grad_norm": 1.008714199066162, "learning_rate": 4.55900201774573e-05, "loss": 2.3348, "step": 2557 }, { "epoch": 0.02300639016427354, "grad_norm": 1.3930459022521973, "learning_rate": 4.558575652737355e-05, "loss": 2.254, "step": 2558 }, { "epoch": 0.02301538406191399, "grad_norm": 1.5778610706329346, "learning_rate": 4.558149101677129e-05, "loss": 2.2367, "step": 2559 }, { "epoch": 0.02302437795955444, "grad_norm": 0.76688152551651, "learning_rate": 4.557722364603605e-05, "loss": 2.2101, "step": 2560 }, { "epoch": 0.023033371857194892, "grad_norm": 1.290571928024292, "learning_rate": 4.5572954415553496e-05, "loss": 2.3356, "step": 2561 }, { "epoch": 0.023042365754835346, "grad_norm": 0.89118891954422, "learning_rate": 4.556868332570948e-05, "loss": 2.1998, "step": 2562 }, { "epoch": 0.023051359652475796, "grad_norm": 1.0452582836151123, "learning_rate": 4.556441037689003e-05, "loss": 2.2682, "step": 2563 }, { "epoch": 0.023060353550116246, "grad_norm": 0.8912217020988464, "learning_rate": 4.556013556948131e-05, "loss": 2.2066, "step": 2564 }, { "epoch": 0.023069347447756697, "grad_norm": 1.1093764305114746, "learning_rate": 4.5555858903869685e-05, "loss": 2.2301, "step": 2565 }, { "epoch": 0.023078341345397147, "grad_norm": 1.3917089700698853, "learning_rate": 4.5551580380441675e-05, "loss": 2.2358, "step": 2566 }, { "epoch": 0.023087335243037597, "grad_norm": 1.1660871505737305, "learning_rate": 4.554729999958396e-05, "loss": 2.268, "step": 2567 }, { "epoch": 0.02309632914067805, "grad_norm": 1.3867371082305908, "learning_rate": 4.554301776168341e-05, "loss": 2.2502, "step": 2568 }, { "epoch": 0.0231053230383185, "grad_norm": 1.0169161558151245, "learning_rate": 4.5538733667127034e-05, "loss": 2.229, "step": 2569 }, { "epoch": 0.023114316935958952, "grad_norm": 1.5149816274642944, "learning_rate": 4.553444771630204e-05, "loss": 2.2522, "step": 2570 }, { "epoch": 0.023123310833599402, "grad_norm": 2.189971685409546, "learning_rate": 4.553015990959578e-05, "loss": 2.2245, "step": 2571 }, { "epoch": 0.023132304731239853, "grad_norm": 1.3959707021713257, "learning_rate": 4.5525870247395774e-05, "loss": 2.2102, "step": 2572 }, { "epoch": 0.023141298628880303, "grad_norm": 1.179862141609192, "learning_rate": 4.552157873008972e-05, "loss": 2.204, "step": 2573 }, { "epoch": 0.023150292526520757, "grad_norm": 0.883988618850708, "learning_rate": 4.55172853580655e-05, "loss": 2.2155, "step": 2574 }, { "epoch": 0.023159286424161207, "grad_norm": 1.3593411445617676, "learning_rate": 4.551299013171111e-05, "loss": 2.1952, "step": 2575 }, { "epoch": 0.023168280321801658, "grad_norm": 1.560001015663147, "learning_rate": 4.550869305141478e-05, "loss": 2.2533, "step": 2576 }, { "epoch": 0.023177274219442108, "grad_norm": 1.2084921598434448, "learning_rate": 4.550439411756486e-05, "loss": 2.2545, "step": 2577 }, { "epoch": 0.02318626811708256, "grad_norm": 0.7104119062423706, "learning_rate": 4.550009333054988e-05, "loss": 2.2108, "step": 2578 }, { "epoch": 0.02319526201472301, "grad_norm": 1.3518625497817993, "learning_rate": 4.549579069075856e-05, "loss": 2.2617, "step": 2579 }, { "epoch": 0.023204255912363463, "grad_norm": 1.0677927732467651, "learning_rate": 4.549148619857974e-05, "loss": 2.3475, "step": 2580 }, { "epoch": 0.023213249810003913, "grad_norm": 0.9304141402244568, "learning_rate": 4.548717985440247e-05, "loss": 2.2621, "step": 2581 }, { "epoch": 0.023222243707644363, "grad_norm": 1.611052393913269, "learning_rate": 4.548287165861595e-05, "loss": 2.3008, "step": 2582 }, { "epoch": 0.023231237605284814, "grad_norm": 1.7965089082717896, "learning_rate": 4.547856161160956e-05, "loss": 2.1588, "step": 2583 }, { "epoch": 0.023240231502925264, "grad_norm": 1.0640138387680054, "learning_rate": 4.547424971377282e-05, "loss": 2.3459, "step": 2584 }, { "epoch": 0.023249225400565718, "grad_norm": 1.0721553564071655, "learning_rate": 4.546993596549545e-05, "loss": 2.2124, "step": 2585 }, { "epoch": 0.023258219298206168, "grad_norm": 1.303740382194519, "learning_rate": 4.546562036716732e-05, "loss": 2.2978, "step": 2586 }, { "epoch": 0.02326721319584662, "grad_norm": 1.569815993309021, "learning_rate": 4.5461302919178453e-05, "loss": 2.2364, "step": 2587 }, { "epoch": 0.02327620709348707, "grad_norm": 2.0101001262664795, "learning_rate": 4.545698362191907e-05, "loss": 2.2604, "step": 2588 }, { "epoch": 0.02328520099112752, "grad_norm": 1.2652915716171265, "learning_rate": 4.5452662475779554e-05, "loss": 2.2162, "step": 2589 }, { "epoch": 0.02329419488876797, "grad_norm": 2.0101308822631836, "learning_rate": 4.5448339481150415e-05, "loss": 2.1842, "step": 2590 }, { "epoch": 0.023303188786408423, "grad_norm": 0.8240359425544739, "learning_rate": 4.544401463842239e-05, "loss": 2.2204, "step": 2591 }, { "epoch": 0.023312182684048874, "grad_norm": 0.8013392686843872, "learning_rate": 4.543968794798634e-05, "loss": 2.2083, "step": 2592 }, { "epoch": 0.023321176581689324, "grad_norm": 1.8335515260696411, "learning_rate": 4.543535941023331e-05, "loss": 2.2449, "step": 2593 }, { "epoch": 0.023330170479329775, "grad_norm": 1.0454282760620117, "learning_rate": 4.543102902555451e-05, "loss": 2.2402, "step": 2594 }, { "epoch": 0.023339164376970225, "grad_norm": 1.1600691080093384, "learning_rate": 4.542669679434131e-05, "loss": 2.284, "step": 2595 }, { "epoch": 0.023348158274610675, "grad_norm": 1.5985760688781738, "learning_rate": 4.542236271698526e-05, "loss": 2.2584, "step": 2596 }, { "epoch": 0.02335715217225113, "grad_norm": 1.0974044799804688, "learning_rate": 4.541802679387806e-05, "loss": 2.3022, "step": 2597 }, { "epoch": 0.02336614606989158, "grad_norm": 1.4724164009094238, "learning_rate": 4.541368902541159e-05, "loss": 2.2888, "step": 2598 }, { "epoch": 0.02337513996753203, "grad_norm": 1.6678922176361084, "learning_rate": 4.54093494119779e-05, "loss": 2.3076, "step": 2599 }, { "epoch": 0.02338413386517248, "grad_norm": 2.133077621459961, "learning_rate": 4.5405007953969185e-05, "loss": 2.2835, "step": 2600 }, { "epoch": 0.02339312776281293, "grad_norm": 0.9044418931007385, "learning_rate": 4.5400664651777835e-05, "loss": 2.2824, "step": 2601 }, { "epoch": 0.02340212166045338, "grad_norm": 1.033979892730713, "learning_rate": 4.539631950579638e-05, "loss": 2.2543, "step": 2602 }, { "epoch": 0.023411115558093835, "grad_norm": 0.8132662177085876, "learning_rate": 4.539197251641755e-05, "loss": 2.2507, "step": 2603 }, { "epoch": 0.023420109455734285, "grad_norm": 0.8884284496307373, "learning_rate": 4.53876236840342e-05, "loss": 2.3025, "step": 2604 }, { "epoch": 0.023429103353374735, "grad_norm": 1.0126274824142456, "learning_rate": 4.538327300903938e-05, "loss": 2.1972, "step": 2605 }, { "epoch": 0.023438097251015186, "grad_norm": 1.0416536331176758, "learning_rate": 4.537892049182631e-05, "loss": 2.2261, "step": 2606 }, { "epoch": 0.023447091148655636, "grad_norm": 1.072108268737793, "learning_rate": 4.537456613278834e-05, "loss": 2.2172, "step": 2607 }, { "epoch": 0.023456085046296087, "grad_norm": 0.7475091218948364, "learning_rate": 4.537020993231904e-05, "loss": 2.271, "step": 2608 }, { "epoch": 0.02346507894393654, "grad_norm": 1.058565378189087, "learning_rate": 4.536585189081211e-05, "loss": 2.2329, "step": 2609 }, { "epoch": 0.02347407284157699, "grad_norm": 1.4133237600326538, "learning_rate": 4.5361492008661414e-05, "loss": 2.3256, "step": 2610 }, { "epoch": 0.02348306673921744, "grad_norm": 1.0400463342666626, "learning_rate": 4.535713028626101e-05, "loss": 2.2413, "step": 2611 }, { "epoch": 0.02349206063685789, "grad_norm": 1.5575670003890991, "learning_rate": 4.535276672400509e-05, "loss": 2.2939, "step": 2612 }, { "epoch": 0.023501054534498342, "grad_norm": 0.7127779126167297, "learning_rate": 4.5348401322288044e-05, "loss": 2.3061, "step": 2613 }, { "epoch": 0.023510048432138792, "grad_norm": 1.7266398668289185, "learning_rate": 4.5344034081504395e-05, "loss": 2.2164, "step": 2614 }, { "epoch": 0.023519042329779246, "grad_norm": 1.448909878730774, "learning_rate": 4.5339665002048864e-05, "loss": 2.3292, "step": 2615 }, { "epoch": 0.023528036227419696, "grad_norm": 2.2207140922546387, "learning_rate": 4.533529408431631e-05, "loss": 2.2511, "step": 2616 }, { "epoch": 0.023537030125060147, "grad_norm": 1.0009530782699585, "learning_rate": 4.5330921328701795e-05, "loss": 2.2472, "step": 2617 }, { "epoch": 0.023546024022700597, "grad_norm": 1.0049372911453247, "learning_rate": 4.53265467356005e-05, "loss": 2.2408, "step": 2618 }, { "epoch": 0.023555017920341047, "grad_norm": 1.2947500944137573, "learning_rate": 4.532217030540781e-05, "loss": 2.341, "step": 2619 }, { "epoch": 0.0235640118179815, "grad_norm": 0.9671518206596375, "learning_rate": 4.531779203851925e-05, "loss": 2.2664, "step": 2620 }, { "epoch": 0.02357300571562195, "grad_norm": 0.8629701733589172, "learning_rate": 4.531341193533053e-05, "loss": 2.2472, "step": 2621 }, { "epoch": 0.023581999613262402, "grad_norm": 1.0224510431289673, "learning_rate": 4.530902999623752e-05, "loss": 2.2775, "step": 2622 }, { "epoch": 0.023590993510902852, "grad_norm": 0.912785530090332, "learning_rate": 4.530464622163625e-05, "loss": 2.1826, "step": 2623 }, { "epoch": 0.023599987408543303, "grad_norm": 1.2613143920898438, "learning_rate": 4.530026061192294e-05, "loss": 2.2931, "step": 2624 }, { "epoch": 0.023608981306183753, "grad_norm": 1.243145227432251, "learning_rate": 4.529587316749393e-05, "loss": 2.2655, "step": 2625 }, { "epoch": 0.023617975203824207, "grad_norm": 1.0698224306106567, "learning_rate": 4.529148388874577e-05, "loss": 2.2815, "step": 2626 }, { "epoch": 0.023626969101464657, "grad_norm": 1.0359392166137695, "learning_rate": 4.528709277607515e-05, "loss": 2.3021, "step": 2627 }, { "epoch": 0.023635962999105108, "grad_norm": 0.9310892820358276, "learning_rate": 4.528269982987893e-05, "loss": 2.2528, "step": 2628 }, { "epoch": 0.023644956896745558, "grad_norm": 1.1447091102600098, "learning_rate": 4.527830505055415e-05, "loss": 2.1569, "step": 2629 }, { "epoch": 0.02365395079438601, "grad_norm": 1.0823932886123657, "learning_rate": 4.5273908438498e-05, "loss": 2.3175, "step": 2630 }, { "epoch": 0.02366294469202646, "grad_norm": 1.3747798204421997, "learning_rate": 4.526950999410785e-05, "loss": 2.2827, "step": 2631 }, { "epoch": 0.023671938589666913, "grad_norm": 1.4402918815612793, "learning_rate": 4.5265109717781214e-05, "loss": 2.2342, "step": 2632 }, { "epoch": 0.023680932487307363, "grad_norm": 1.0666195154190063, "learning_rate": 4.526070760991579e-05, "loss": 2.2629, "step": 2633 }, { "epoch": 0.023689926384947813, "grad_norm": 1.4574238061904907, "learning_rate": 4.525630367090944e-05, "loss": 2.2216, "step": 2634 }, { "epoch": 0.023698920282588264, "grad_norm": 1.0301612615585327, "learning_rate": 4.525189790116017e-05, "loss": 2.2133, "step": 2635 }, { "epoch": 0.023707914180228714, "grad_norm": 1.3526188135147095, "learning_rate": 4.5247490301066195e-05, "loss": 2.2232, "step": 2636 }, { "epoch": 0.023716908077869164, "grad_norm": 1.041427493095398, "learning_rate": 4.5243080871025846e-05, "loss": 2.2126, "step": 2637 }, { "epoch": 0.023725901975509618, "grad_norm": 1.0333713293075562, "learning_rate": 4.523866961143766e-05, "loss": 2.2311, "step": 2638 }, { "epoch": 0.02373489587315007, "grad_norm": 1.6763808727264404, "learning_rate": 4.52342565227003e-05, "loss": 2.2903, "step": 2639 }, { "epoch": 0.02374388977079052, "grad_norm": 0.8471630811691284, "learning_rate": 4.522984160521264e-05, "loss": 2.2011, "step": 2640 }, { "epoch": 0.02375288366843097, "grad_norm": 2.2285099029541016, "learning_rate": 4.522542485937369e-05, "loss": 2.246, "step": 2641 }, { "epoch": 0.02376187756607142, "grad_norm": 1.2699719667434692, "learning_rate": 4.522100628558262e-05, "loss": 2.277, "step": 2642 }, { "epoch": 0.02377087146371187, "grad_norm": 2.37869930267334, "learning_rate": 4.5216585884238784e-05, "loss": 2.3462, "step": 2643 }, { "epoch": 0.023779865361352324, "grad_norm": 1.8265730142593384, "learning_rate": 4.521216365574169e-05, "loss": 2.2279, "step": 2644 }, { "epoch": 0.023788859258992774, "grad_norm": 1.8037865161895752, "learning_rate": 4.5207739600491015e-05, "loss": 2.2298, "step": 2645 }, { "epoch": 0.023797853156633225, "grad_norm": 1.1285464763641357, "learning_rate": 4.52033137188866e-05, "loss": 2.195, "step": 2646 }, { "epoch": 0.023806847054273675, "grad_norm": 2.237424612045288, "learning_rate": 4.5198886011328456e-05, "loss": 2.2995, "step": 2647 }, { "epoch": 0.023815840951914125, "grad_norm": 1.0455268621444702, "learning_rate": 4.519445647821675e-05, "loss": 2.2023, "step": 2648 }, { "epoch": 0.023824834849554576, "grad_norm": 1.7602901458740234, "learning_rate": 4.519002511995182e-05, "loss": 2.2459, "step": 2649 }, { "epoch": 0.02383382874719503, "grad_norm": 1.3636770248413086, "learning_rate": 4.5185591936934156e-05, "loss": 2.2457, "step": 2650 }, { "epoch": 0.02384282264483548, "grad_norm": 0.9138522744178772, "learning_rate": 4.518115692956445e-05, "loss": 2.3012, "step": 2651 }, { "epoch": 0.02385181654247593, "grad_norm": 1.1165839433670044, "learning_rate": 4.517672009824351e-05, "loss": 2.318, "step": 2652 }, { "epoch": 0.02386081044011638, "grad_norm": 0.9331222176551819, "learning_rate": 4.5172281443372345e-05, "loss": 2.2191, "step": 2653 }, { "epoch": 0.02386980433775683, "grad_norm": 0.9418733716011047, "learning_rate": 4.51678409653521e-05, "loss": 2.2663, "step": 2654 }, { "epoch": 0.02387879823539728, "grad_norm": 1.0345649719238281, "learning_rate": 4.516339866458412e-05, "loss": 2.1965, "step": 2655 }, { "epoch": 0.023887792133037735, "grad_norm": 1.151200532913208, "learning_rate": 4.515895454146989e-05, "loss": 2.2283, "step": 2656 }, { "epoch": 0.023896786030678185, "grad_norm": 0.7773505449295044, "learning_rate": 4.515450859641106e-05, "loss": 2.3259, "step": 2657 }, { "epoch": 0.023905779928318636, "grad_norm": 0.8433427810668945, "learning_rate": 4.515006082980945e-05, "loss": 2.2225, "step": 2658 }, { "epoch": 0.023914773825959086, "grad_norm": 1.3869307041168213, "learning_rate": 4.514561124206704e-05, "loss": 2.2896, "step": 2659 }, { "epoch": 0.023923767723599536, "grad_norm": 0.9080814123153687, "learning_rate": 4.5141159833585996e-05, "loss": 2.2798, "step": 2660 }, { "epoch": 0.02393276162123999, "grad_norm": 1.081661343574524, "learning_rate": 4.513670660476861e-05, "loss": 2.3324, "step": 2661 }, { "epoch": 0.02394175551888044, "grad_norm": 1.0186505317687988, "learning_rate": 4.513225155601738e-05, "loss": 2.2748, "step": 2662 }, { "epoch": 0.02395074941652089, "grad_norm": 1.4184377193450928, "learning_rate": 4.512779468773494e-05, "loss": 2.2791, "step": 2663 }, { "epoch": 0.02395974331416134, "grad_norm": 0.862434446811676, "learning_rate": 4.512333600032408e-05, "loss": 2.2809, "step": 2664 }, { "epoch": 0.023968737211801792, "grad_norm": 1.0760747194290161, "learning_rate": 4.51188754941878e-05, "loss": 2.2285, "step": 2665 }, { "epoch": 0.023977731109442242, "grad_norm": 0.5988388061523438, "learning_rate": 4.5114413169729226e-05, "loss": 2.2193, "step": 2666 }, { "epoch": 0.023986725007082696, "grad_norm": 1.162620186805725, "learning_rate": 4.510994902735165e-05, "loss": 2.1986, "step": 2667 }, { "epoch": 0.023995718904723146, "grad_norm": 0.9914802312850952, "learning_rate": 4.510548306745854e-05, "loss": 2.2387, "step": 2668 }, { "epoch": 0.024004712802363597, "grad_norm": 1.204278826713562, "learning_rate": 4.510101529045353e-05, "loss": 2.2638, "step": 2669 }, { "epoch": 0.024013706700004047, "grad_norm": 1.4169708490371704, "learning_rate": 4.50965456967404e-05, "loss": 2.2825, "step": 2670 }, { "epoch": 0.024022700597644497, "grad_norm": 0.5760794281959534, "learning_rate": 4.509207428672313e-05, "loss": 2.2878, "step": 2671 }, { "epoch": 0.024031694495284948, "grad_norm": 0.8500551581382751, "learning_rate": 4.508760106080582e-05, "loss": 2.2804, "step": 2672 }, { "epoch": 0.0240406883929254, "grad_norm": 1.1358448266983032, "learning_rate": 4.5083126019392765e-05, "loss": 2.149, "step": 2673 }, { "epoch": 0.024049682290565852, "grad_norm": 0.7983713150024414, "learning_rate": 4.5078649162888406e-05, "loss": 2.1833, "step": 2674 }, { "epoch": 0.024058676188206302, "grad_norm": 0.9522508978843689, "learning_rate": 4.507417049169737e-05, "loss": 2.1989, "step": 2675 }, { "epoch": 0.024067670085846753, "grad_norm": 0.6413620710372925, "learning_rate": 4.5069690006224424e-05, "loss": 2.2132, "step": 2676 }, { "epoch": 0.024076663983487203, "grad_norm": 0.9654516577720642, "learning_rate": 4.506520770687451e-05, "loss": 2.2585, "step": 2677 }, { "epoch": 0.024085657881127653, "grad_norm": 1.1428955793380737, "learning_rate": 4.5060723594052744e-05, "loss": 2.2761, "step": 2678 }, { "epoch": 0.024094651778768107, "grad_norm": 0.983109712600708, "learning_rate": 4.505623766816438e-05, "loss": 2.2192, "step": 2679 }, { "epoch": 0.024103645676408558, "grad_norm": 1.0218700170516968, "learning_rate": 4.505174992961486e-05, "loss": 2.1696, "step": 2680 }, { "epoch": 0.024112639574049008, "grad_norm": 1.75558602809906, "learning_rate": 4.504726037880978e-05, "loss": 2.3086, "step": 2681 }, { "epoch": 0.02412163347168946, "grad_norm": 0.6948557496070862, "learning_rate": 4.50427690161549e-05, "loss": 2.2761, "step": 2682 }, { "epoch": 0.02413062736932991, "grad_norm": 1.0534065961837769, "learning_rate": 4.5038275842056145e-05, "loss": 2.2063, "step": 2683 }, { "epoch": 0.02413962126697036, "grad_norm": 1.3086248636245728, "learning_rate": 4.5033780856919605e-05, "loss": 2.3193, "step": 2684 }, { "epoch": 0.024148615164610813, "grad_norm": 0.7874822616577148, "learning_rate": 4.502928406115152e-05, "loss": 2.345, "step": 2685 }, { "epoch": 0.024157609062251263, "grad_norm": 0.6520964503288269, "learning_rate": 4.5024785455158326e-05, "loss": 2.2663, "step": 2686 }, { "epoch": 0.024166602959891714, "grad_norm": 1.0472631454467773, "learning_rate": 4.502028503934659e-05, "loss": 2.2338, "step": 2687 }, { "epoch": 0.024175596857532164, "grad_norm": 0.6436008214950562, "learning_rate": 4.5015782814123055e-05, "loss": 2.2205, "step": 2688 }, { "epoch": 0.024184590755172614, "grad_norm": 0.9097863435745239, "learning_rate": 4.501127877989463e-05, "loss": 2.2112, "step": 2689 }, { "epoch": 0.024193584652813065, "grad_norm": 1.2385507822036743, "learning_rate": 4.5006772937068376e-05, "loss": 2.3492, "step": 2690 }, { "epoch": 0.02420257855045352, "grad_norm": 0.5195785164833069, "learning_rate": 4.5002265286051536e-05, "loss": 2.2721, "step": 2691 }, { "epoch": 0.02421157244809397, "grad_norm": 0.7700551152229309, "learning_rate": 4.499775582725151e-05, "loss": 2.2677, "step": 2692 }, { "epoch": 0.02422056634573442, "grad_norm": 1.2729090452194214, "learning_rate": 4.4993244561075846e-05, "loss": 2.2143, "step": 2693 }, { "epoch": 0.02422956024337487, "grad_norm": 0.9438384175300598, "learning_rate": 4.4988731487932275e-05, "loss": 2.2401, "step": 2694 }, { "epoch": 0.02423855414101532, "grad_norm": 1.58765709400177, "learning_rate": 4.498421660822868e-05, "loss": 2.2622, "step": 2695 }, { "epoch": 0.024247548038655774, "grad_norm": 3.647616147994995, "learning_rate": 4.497969992237312e-05, "loss": 2.3004, "step": 2696 }, { "epoch": 0.024256541936296224, "grad_norm": 1.0209457874298096, "learning_rate": 4.49751814307738e-05, "loss": 2.2865, "step": 2697 }, { "epoch": 0.024265535833936674, "grad_norm": 1.604297161102295, "learning_rate": 4.49706611338391e-05, "loss": 2.2621, "step": 2698 }, { "epoch": 0.024274529731577125, "grad_norm": 0.5525092482566833, "learning_rate": 4.496613903197755e-05, "loss": 2.2917, "step": 2699 }, { "epoch": 0.024283523629217575, "grad_norm": 0.8003180027008057, "learning_rate": 4.4961615125597863e-05, "loss": 2.2241, "step": 2700 }, { "epoch": 0.024292517526858026, "grad_norm": 1.307331919670105, "learning_rate": 4.49570894151089e-05, "loss": 2.2478, "step": 2701 }, { "epoch": 0.02430151142449848, "grad_norm": 0.9818620085716248, "learning_rate": 4.495256190091969e-05, "loss": 2.2222, "step": 2702 }, { "epoch": 0.02431050532213893, "grad_norm": 1.9315427541732788, "learning_rate": 4.4948032583439436e-05, "loss": 2.2932, "step": 2703 }, { "epoch": 0.02431949921977938, "grad_norm": 0.8354949951171875, "learning_rate": 4.4943501463077476e-05, "loss": 2.2349, "step": 2704 }, { "epoch": 0.02432849311741983, "grad_norm": 0.9704464673995972, "learning_rate": 4.493896854024333e-05, "loss": 2.2843, "step": 2705 }, { "epoch": 0.02433748701506028, "grad_norm": 0.9038994908332825, "learning_rate": 4.4934433815346696e-05, "loss": 2.2172, "step": 2706 }, { "epoch": 0.02434648091270073, "grad_norm": 0.8859515190124512, "learning_rate": 4.49298972887974e-05, "loss": 2.2454, "step": 2707 }, { "epoch": 0.024355474810341185, "grad_norm": 0.8267603516578674, "learning_rate": 4.492535896100545e-05, "loss": 2.2076, "step": 2708 }, { "epoch": 0.024364468707981635, "grad_norm": 0.8111536502838135, "learning_rate": 4.492081883238102e-05, "loss": 2.2373, "step": 2709 }, { "epoch": 0.024373462605622086, "grad_norm": 1.0119287967681885, "learning_rate": 4.4916276903334445e-05, "loss": 2.2612, "step": 2710 }, { "epoch": 0.024382456503262536, "grad_norm": 1.3758295774459839, "learning_rate": 4.491173317427622e-05, "loss": 2.1848, "step": 2711 }, { "epoch": 0.024391450400902986, "grad_norm": 1.051419734954834, "learning_rate": 4.490718764561699e-05, "loss": 2.2696, "step": 2712 }, { "epoch": 0.024400444298543437, "grad_norm": 1.2014377117156982, "learning_rate": 4.4902640317767584e-05, "loss": 2.2307, "step": 2713 }, { "epoch": 0.02440943819618389, "grad_norm": 1.5124366283416748, "learning_rate": 4.489809119113898e-05, "loss": 2.229, "step": 2714 }, { "epoch": 0.02441843209382434, "grad_norm": 0.9914209246635437, "learning_rate": 4.489354026614233e-05, "loss": 2.2433, "step": 2715 }, { "epoch": 0.02442742599146479, "grad_norm": 1.1648746728897095, "learning_rate": 4.488898754318894e-05, "loss": 2.1839, "step": 2716 }, { "epoch": 0.02443641988910524, "grad_norm": 1.3846435546875, "learning_rate": 4.4884433022690275e-05, "loss": 2.2765, "step": 2717 }, { "epoch": 0.024445413786745692, "grad_norm": 0.8690109252929688, "learning_rate": 4.487987670505798e-05, "loss": 2.239, "step": 2718 }, { "epoch": 0.024454407684386142, "grad_norm": 2.078702211380005, "learning_rate": 4.487531859070383e-05, "loss": 2.2065, "step": 2719 }, { "epoch": 0.024463401582026596, "grad_norm": 1.689876675605774, "learning_rate": 4.48707586800398e-05, "loss": 2.2433, "step": 2720 }, { "epoch": 0.024472395479667047, "grad_norm": 0.8730366826057434, "learning_rate": 4.4866196973478e-05, "loss": 2.264, "step": 2721 }, { "epoch": 0.024481389377307497, "grad_norm": 0.8175548911094666, "learning_rate": 4.4861633471430716e-05, "loss": 2.2558, "step": 2722 }, { "epoch": 0.024490383274947947, "grad_norm": 1.802256464958191, "learning_rate": 4.4857068174310396e-05, "loss": 2.2379, "step": 2723 }, { "epoch": 0.024499377172588398, "grad_norm": 1.1021754741668701, "learning_rate": 4.485250108252964e-05, "loss": 2.2384, "step": 2724 }, { "epoch": 0.024508371070228848, "grad_norm": 0.8987403512001038, "learning_rate": 4.484793219650122e-05, "loss": 2.3051, "step": 2725 }, { "epoch": 0.024517364967869302, "grad_norm": 1.4956508874893188, "learning_rate": 4.4843361516638075e-05, "loss": 2.2197, "step": 2726 }, { "epoch": 0.024526358865509752, "grad_norm": 1.1700496673583984, "learning_rate": 4.4838789043353265e-05, "loss": 2.3221, "step": 2727 }, { "epoch": 0.024535352763150203, "grad_norm": 0.8588183522224426, "learning_rate": 4.4834214777060094e-05, "loss": 2.2779, "step": 2728 }, { "epoch": 0.024544346660790653, "grad_norm": 0.6465746760368347, "learning_rate": 4.4829638718171954e-05, "loss": 2.2634, "step": 2729 }, { "epoch": 0.024553340558431103, "grad_norm": 0.818744957447052, "learning_rate": 4.4825060867102406e-05, "loss": 2.1845, "step": 2730 }, { "epoch": 0.024562334456071557, "grad_norm": 0.9117552042007446, "learning_rate": 4.482048122426523e-05, "loss": 2.2117, "step": 2731 }, { "epoch": 0.024571328353712008, "grad_norm": 1.2231776714324951, "learning_rate": 4.48158997900743e-05, "loss": 2.2658, "step": 2732 }, { "epoch": 0.024580322251352458, "grad_norm": 1.3326233625411987, "learning_rate": 4.4811316564943694e-05, "loss": 2.303, "step": 2733 }, { "epoch": 0.024589316148992908, "grad_norm": 0.6749585270881653, "learning_rate": 4.4806731549287636e-05, "loss": 2.2185, "step": 2734 }, { "epoch": 0.02459831004663336, "grad_norm": 1.6628086566925049, "learning_rate": 4.4802144743520516e-05, "loss": 2.2672, "step": 2735 }, { "epoch": 0.02460730394427381, "grad_norm": 0.8773925304412842, "learning_rate": 4.479755614805688e-05, "loss": 2.2649, "step": 2736 }, { "epoch": 0.024616297841914263, "grad_norm": 0.8830981850624084, "learning_rate": 4.4792965763311446e-05, "loss": 2.2492, "step": 2737 }, { "epoch": 0.024625291739554713, "grad_norm": 1.0465861558914185, "learning_rate": 4.4788373589699085e-05, "loss": 2.2969, "step": 2738 }, { "epoch": 0.024634285637195164, "grad_norm": 1.027933120727539, "learning_rate": 4.478377962763483e-05, "loss": 2.2103, "step": 2739 }, { "epoch": 0.024643279534835614, "grad_norm": 0.9336531758308411, "learning_rate": 4.477918387753388e-05, "loss": 2.2252, "step": 2740 }, { "epoch": 0.024652273432476064, "grad_norm": 0.6650957465171814, "learning_rate": 4.4774586339811605e-05, "loss": 2.1969, "step": 2741 }, { "epoch": 0.024661267330116515, "grad_norm": 1.0099353790283203, "learning_rate": 4.4769987014883505e-05, "loss": 2.312, "step": 2742 }, { "epoch": 0.02467026122775697, "grad_norm": 0.895894467830658, "learning_rate": 4.476538590316529e-05, "loss": 2.1998, "step": 2743 }, { "epoch": 0.02467925512539742, "grad_norm": 0.855859637260437, "learning_rate": 4.476078300507277e-05, "loss": 2.2237, "step": 2744 }, { "epoch": 0.02468824902303787, "grad_norm": 1.0555497407913208, "learning_rate": 4.475617832102197e-05, "loss": 2.1843, "step": 2745 }, { "epoch": 0.02469724292067832, "grad_norm": 1.2171953916549683, "learning_rate": 4.4751571851429056e-05, "loss": 2.2858, "step": 2746 }, { "epoch": 0.02470623681831877, "grad_norm": 1.3389594554901123, "learning_rate": 4.474696359671035e-05, "loss": 2.1957, "step": 2747 }, { "epoch": 0.02471523071595922, "grad_norm": 0.8535686731338501, "learning_rate": 4.474235355728235e-05, "loss": 2.2345, "step": 2748 }, { "epoch": 0.024724224613599674, "grad_norm": 0.9718269109725952, "learning_rate": 4.473774173356169e-05, "loss": 2.2301, "step": 2749 }, { "epoch": 0.024733218511240124, "grad_norm": 1.7145953178405762, "learning_rate": 4.47331281259652e-05, "loss": 2.2588, "step": 2750 }, { "epoch": 0.024742212408880575, "grad_norm": 0.5039273500442505, "learning_rate": 4.4728512734909844e-05, "loss": 2.1909, "step": 2751 }, { "epoch": 0.024751206306521025, "grad_norm": 0.6773383617401123, "learning_rate": 4.4723895560812756e-05, "loss": 2.2925, "step": 2752 }, { "epoch": 0.024760200204161475, "grad_norm": 0.9763181805610657, "learning_rate": 4.471927660409123e-05, "loss": 2.2946, "step": 2753 }, { "epoch": 0.024769194101801926, "grad_norm": 0.7764590382575989, "learning_rate": 4.471465586516272e-05, "loss": 2.2581, "step": 2754 }, { "epoch": 0.02477818799944238, "grad_norm": 1.3218764066696167, "learning_rate": 4.471003334444486e-05, "loss": 2.226, "step": 2755 }, { "epoch": 0.02478718189708283, "grad_norm": 1.3302581310272217, "learning_rate": 4.470540904235541e-05, "loss": 2.2984, "step": 2756 }, { "epoch": 0.02479617579472328, "grad_norm": 0.4093705415725708, "learning_rate": 4.4700782959312316e-05, "loss": 2.1936, "step": 2757 }, { "epoch": 0.02480516969236373, "grad_norm": 0.9448533654212952, "learning_rate": 4.4696155095733674e-05, "loss": 2.2325, "step": 2758 }, { "epoch": 0.02481416359000418, "grad_norm": 1.1990740299224854, "learning_rate": 4.469152545203775e-05, "loss": 2.3203, "step": 2759 }, { "epoch": 0.02482315748764463, "grad_norm": 0.6376228928565979, "learning_rate": 4.4686894028642966e-05, "loss": 2.252, "step": 2760 }, { "epoch": 0.024832151385285085, "grad_norm": 0.7667058110237122, "learning_rate": 4.468226082596792e-05, "loss": 2.262, "step": 2761 }, { "epoch": 0.024841145282925536, "grad_norm": 1.0270947217941284, "learning_rate": 4.467762584443131e-05, "loss": 2.2588, "step": 2762 }, { "epoch": 0.024850139180565986, "grad_norm": 0.7357791662216187, "learning_rate": 4.46729890844521e-05, "loss": 2.2979, "step": 2763 }, { "epoch": 0.024859133078206436, "grad_norm": 1.2101958990097046, "learning_rate": 4.466835054644931e-05, "loss": 2.2958, "step": 2764 }, { "epoch": 0.024868126975846887, "grad_norm": 0.7955397367477417, "learning_rate": 4.466371023084219e-05, "loss": 2.2479, "step": 2765 }, { "epoch": 0.024877120873487337, "grad_norm": 0.8184387683868408, "learning_rate": 4.465906813805012e-05, "loss": 2.2356, "step": 2766 }, { "epoch": 0.02488611477112779, "grad_norm": 1.4352951049804688, "learning_rate": 4.465442426849264e-05, "loss": 2.228, "step": 2767 }, { "epoch": 0.02489510866876824, "grad_norm": 0.5096254944801331, "learning_rate": 4.4649778622589464e-05, "loss": 2.2394, "step": 2768 }, { "epoch": 0.02490410256640869, "grad_norm": 1.160958170890808, "learning_rate": 4.464513120076047e-05, "loss": 2.3005, "step": 2769 }, { "epoch": 0.024913096464049142, "grad_norm": 0.6801103353500366, "learning_rate": 4.464048200342568e-05, "loss": 2.2404, "step": 2770 }, { "epoch": 0.024922090361689592, "grad_norm": 2.984139919281006, "learning_rate": 4.463583103100527e-05, "loss": 2.2855, "step": 2771 }, { "epoch": 0.024931084259330046, "grad_norm": 1.7741076946258545, "learning_rate": 4.46311782839196e-05, "loss": 2.2514, "step": 2772 }, { "epoch": 0.024940078156970497, "grad_norm": 1.638950228691101, "learning_rate": 4.46265237625892e-05, "loss": 2.2419, "step": 2773 }, { "epoch": 0.024949072054610947, "grad_norm": 1.0804576873779297, "learning_rate": 4.462186746743471e-05, "loss": 2.2606, "step": 2774 }, { "epoch": 0.024958065952251397, "grad_norm": 1.2346988916397095, "learning_rate": 4.461720939887697e-05, "loss": 2.2235, "step": 2775 }, { "epoch": 0.024967059849891848, "grad_norm": 1.6586073637008667, "learning_rate": 4.4612549557336974e-05, "loss": 2.2571, "step": 2776 }, { "epoch": 0.024976053747532298, "grad_norm": 1.156713604927063, "learning_rate": 4.460788794323588e-05, "loss": 2.2623, "step": 2777 }, { "epoch": 0.024985047645172752, "grad_norm": 0.7579923868179321, "learning_rate": 4.4603224556994994e-05, "loss": 2.2789, "step": 2778 }, { "epoch": 0.024994041542813202, "grad_norm": 1.0331007242202759, "learning_rate": 4.459855939903578e-05, "loss": 2.1965, "step": 2779 }, { "epoch": 0.025003035440453653, "grad_norm": 1.8126881122589111, "learning_rate": 4.459389246977989e-05, "loss": 2.3436, "step": 2780 }, { "epoch": 0.025012029338094103, "grad_norm": 2.52854323387146, "learning_rate": 4.4589223769649094e-05, "loss": 2.3618, "step": 2781 }, { "epoch": 0.025021023235734553, "grad_norm": 1.2825706005096436, "learning_rate": 4.458455329906535e-05, "loss": 2.212, "step": 2782 }, { "epoch": 0.025030017133375004, "grad_norm": 1.0408897399902344, "learning_rate": 4.457988105845078e-05, "loss": 2.1501, "step": 2783 }, { "epoch": 0.025039011031015457, "grad_norm": 0.8186978697776794, "learning_rate": 4.457520704822765e-05, "loss": 2.1665, "step": 2784 }, { "epoch": 0.025048004928655908, "grad_norm": 0.6502106785774231, "learning_rate": 4.457053126881839e-05, "loss": 2.1628, "step": 2785 }, { "epoch": 0.025056998826296358, "grad_norm": 2.2323830127716064, "learning_rate": 4.456585372064559e-05, "loss": 2.2641, "step": 2786 }, { "epoch": 0.02506599272393681, "grad_norm": 1.3796271085739136, "learning_rate": 4.456117440413201e-05, "loss": 2.2436, "step": 2787 }, { "epoch": 0.02507498662157726, "grad_norm": 1.2113150358200073, "learning_rate": 4.455649331970055e-05, "loss": 2.2499, "step": 2788 }, { "epoch": 0.02508398051921771, "grad_norm": 1.1902210712432861, "learning_rate": 4.45518104677743e-05, "loss": 2.2159, "step": 2789 }, { "epoch": 0.025092974416858163, "grad_norm": 1.389224648475647, "learning_rate": 4.4547125848776474e-05, "loss": 2.2832, "step": 2790 }, { "epoch": 0.025101968314498613, "grad_norm": 1.2478233575820923, "learning_rate": 4.454243946313047e-05, "loss": 2.228, "step": 2791 }, { "epoch": 0.025110962212139064, "grad_norm": 1.4874475002288818, "learning_rate": 4.453775131125983e-05, "loss": 2.2515, "step": 2792 }, { "epoch": 0.025119956109779514, "grad_norm": 0.9949284195899963, "learning_rate": 4.453306139358828e-05, "loss": 2.2674, "step": 2793 }, { "epoch": 0.025128950007419965, "grad_norm": 2.011068105697632, "learning_rate": 4.4528369710539674e-05, "loss": 2.2445, "step": 2794 }, { "epoch": 0.025137943905060415, "grad_norm": 0.9138484597206116, "learning_rate": 4.452367626253805e-05, "loss": 2.2049, "step": 2795 }, { "epoch": 0.02514693780270087, "grad_norm": 0.9498387575149536, "learning_rate": 4.451898105000759e-05, "loss": 2.205, "step": 2796 }, { "epoch": 0.02515593170034132, "grad_norm": 0.7584311962127686, "learning_rate": 4.451428407337266e-05, "loss": 2.2711, "step": 2797 }, { "epoch": 0.02516492559798177, "grad_norm": 1.5933650732040405, "learning_rate": 4.450958533305774e-05, "loss": 2.2274, "step": 2798 }, { "epoch": 0.02517391949562222, "grad_norm": 1.7942460775375366, "learning_rate": 4.450488482948751e-05, "loss": 2.2583, "step": 2799 }, { "epoch": 0.02518291339326267, "grad_norm": 1.8657792806625366, "learning_rate": 4.450018256308681e-05, "loss": 2.2747, "step": 2800 }, { "epoch": 0.02519190729090312, "grad_norm": 1.5537186861038208, "learning_rate": 4.449547853428061e-05, "loss": 2.2326, "step": 2801 }, { "epoch": 0.025200901188543574, "grad_norm": 1.0531367063522339, "learning_rate": 4.449077274349406e-05, "loss": 2.2645, "step": 2802 }, { "epoch": 0.025209895086184025, "grad_norm": 0.7962214350700378, "learning_rate": 4.448606519115246e-05, "loss": 2.2634, "step": 2803 }, { "epoch": 0.025218888983824475, "grad_norm": 0.9789595007896423, "learning_rate": 4.448135587768128e-05, "loss": 2.2206, "step": 2804 }, { "epoch": 0.025227882881464925, "grad_norm": 1.5801173448562622, "learning_rate": 4.447664480350614e-05, "loss": 2.2545, "step": 2805 }, { "epoch": 0.025236876779105376, "grad_norm": 0.5547583103179932, "learning_rate": 4.4471931969052816e-05, "loss": 2.1988, "step": 2806 }, { "epoch": 0.02524587067674583, "grad_norm": 0.5494580864906311, "learning_rate": 4.446721737474726e-05, "loss": 2.21, "step": 2807 }, { "epoch": 0.02525486457438628, "grad_norm": 0.9488596320152283, "learning_rate": 4.446250102101556e-05, "loss": 2.2458, "step": 2808 }, { "epoch": 0.02526385847202673, "grad_norm": 1.2917429208755493, "learning_rate": 4.445778290828398e-05, "loss": 2.2376, "step": 2809 }, { "epoch": 0.02527285236966718, "grad_norm": 2.000288248062134, "learning_rate": 4.445306303697894e-05, "loss": 2.3441, "step": 2810 }, { "epoch": 0.02528184626730763, "grad_norm": 1.634665846824646, "learning_rate": 4.4448341407527024e-05, "loss": 2.2565, "step": 2811 }, { "epoch": 0.02529084016494808, "grad_norm": 1.1367703676223755, "learning_rate": 4.444361802035495e-05, "loss": 2.2192, "step": 2812 }, { "epoch": 0.025299834062588535, "grad_norm": 1.0770435333251953, "learning_rate": 4.443889287588963e-05, "loss": 2.2067, "step": 2813 }, { "epoch": 0.025308827960228986, "grad_norm": 1.151777982711792, "learning_rate": 4.44341659745581e-05, "loss": 2.2712, "step": 2814 }, { "epoch": 0.025317821857869436, "grad_norm": 1.3772189617156982, "learning_rate": 4.442943731678759e-05, "loss": 2.265, "step": 2815 }, { "epoch": 0.025326815755509886, "grad_norm": 1.0883219242095947, "learning_rate": 4.4424706903005465e-05, "loss": 2.2351, "step": 2816 }, { "epoch": 0.025335809653150337, "grad_norm": 1.065712571144104, "learning_rate": 4.4419974733639244e-05, "loss": 2.2502, "step": 2817 }, { "epoch": 0.025344803550790787, "grad_norm": 1.7275992631912231, "learning_rate": 4.441524080911663e-05, "loss": 2.2584, "step": 2818 }, { "epoch": 0.02535379744843124, "grad_norm": 0.8800598978996277, "learning_rate": 4.4410505129865456e-05, "loss": 2.2497, "step": 2819 }, { "epoch": 0.02536279134607169, "grad_norm": 1.9890000820159912, "learning_rate": 4.440576769631375e-05, "loss": 2.2969, "step": 2820 }, { "epoch": 0.02537178524371214, "grad_norm": 1.5857577323913574, "learning_rate": 4.440102850888965e-05, "loss": 2.2363, "step": 2821 }, { "epoch": 0.025380779141352592, "grad_norm": 0.8814457058906555, "learning_rate": 4.439628756802149e-05, "loss": 2.2162, "step": 2822 }, { "epoch": 0.025389773038993042, "grad_norm": 1.8704408407211304, "learning_rate": 4.439154487413776e-05, "loss": 2.288, "step": 2823 }, { "epoch": 0.025398766936633493, "grad_norm": 1.4949895143508911, "learning_rate": 4.4386800427667084e-05, "loss": 2.2798, "step": 2824 }, { "epoch": 0.025407760834273947, "grad_norm": 0.8873275518417358, "learning_rate": 4.4382054229038275e-05, "loss": 2.2341, "step": 2825 }, { "epoch": 0.025416754731914397, "grad_norm": 0.5728654861450195, "learning_rate": 4.437730627868027e-05, "loss": 2.2266, "step": 2826 }, { "epoch": 0.025425748629554847, "grad_norm": 0.7610718011856079, "learning_rate": 4.4372556577022204e-05, "loss": 2.2451, "step": 2827 }, { "epoch": 0.025434742527195298, "grad_norm": 0.9848352074623108, "learning_rate": 4.436780512449334e-05, "loss": 2.2699, "step": 2828 }, { "epoch": 0.025443736424835748, "grad_norm": 1.4243015050888062, "learning_rate": 4.436305192152311e-05, "loss": 2.2506, "step": 2829 }, { "epoch": 0.0254527303224762, "grad_norm": 1.2924723625183105, "learning_rate": 4.4358296968541105e-05, "loss": 2.2069, "step": 2830 }, { "epoch": 0.025461724220116652, "grad_norm": 0.8533881902694702, "learning_rate": 4.4353540265977064e-05, "loss": 2.1837, "step": 2831 }, { "epoch": 0.025470718117757103, "grad_norm": 0.6762183308601379, "learning_rate": 4.43487818142609e-05, "loss": 2.1718, "step": 2832 }, { "epoch": 0.025479712015397553, "grad_norm": 2.873892307281494, "learning_rate": 4.4344021613822685e-05, "loss": 2.3103, "step": 2833 }, { "epoch": 0.025488705913038003, "grad_norm": 1.3380135297775269, "learning_rate": 4.433925966509263e-05, "loss": 2.257, "step": 2834 }, { "epoch": 0.025497699810678454, "grad_norm": 1.6459331512451172, "learning_rate": 4.433449596850111e-05, "loss": 2.2573, "step": 2835 }, { "epoch": 0.025506693708318904, "grad_norm": 0.8595320582389832, "learning_rate": 4.432973052447867e-05, "loss": 2.2418, "step": 2836 }, { "epoch": 0.025515687605959358, "grad_norm": 0.7581088542938232, "learning_rate": 4.432496333345601e-05, "loss": 2.2028, "step": 2837 }, { "epoch": 0.025524681503599808, "grad_norm": 1.7214889526367188, "learning_rate": 4.432019439586398e-05, "loss": 2.2875, "step": 2838 }, { "epoch": 0.02553367540124026, "grad_norm": 0.7399866580963135, "learning_rate": 4.4315423712133595e-05, "loss": 2.228, "step": 2839 }, { "epoch": 0.02554266929888071, "grad_norm": 0.9629696607589722, "learning_rate": 4.4310651282696015e-05, "loss": 2.2756, "step": 2840 }, { "epoch": 0.02555166319652116, "grad_norm": 0.8558395504951477, "learning_rate": 4.430587710798256e-05, "loss": 2.221, "step": 2841 }, { "epoch": 0.025560657094161613, "grad_norm": 0.6496704816818237, "learning_rate": 4.430110118842474e-05, "loss": 2.2092, "step": 2842 }, { "epoch": 0.025569650991802063, "grad_norm": 1.168129563331604, "learning_rate": 4.4296323524454186e-05, "loss": 2.2289, "step": 2843 }, { "epoch": 0.025578644889442514, "grad_norm": 0.959107518196106, "learning_rate": 4.4291544116502695e-05, "loss": 2.2875, "step": 2844 }, { "epoch": 0.025587638787082964, "grad_norm": 1.3436812162399292, "learning_rate": 4.4286762965002214e-05, "loss": 2.2784, "step": 2845 }, { "epoch": 0.025596632684723414, "grad_norm": 1.1457875967025757, "learning_rate": 4.4281980070384886e-05, "loss": 2.2304, "step": 2846 }, { "epoch": 0.025605626582363865, "grad_norm": 0.9224822521209717, "learning_rate": 4.4277195433082955e-05, "loss": 2.2738, "step": 2847 }, { "epoch": 0.02561462048000432, "grad_norm": 1.034532904624939, "learning_rate": 4.4272409053528875e-05, "loss": 2.303, "step": 2848 }, { "epoch": 0.02562361437764477, "grad_norm": 0.7546110153198242, "learning_rate": 4.4267620932155216e-05, "loss": 2.1952, "step": 2849 }, { "epoch": 0.02563260827528522, "grad_norm": 0.9413643479347229, "learning_rate": 4.426283106939474e-05, "loss": 2.1748, "step": 2850 }, { "epoch": 0.02564160217292567, "grad_norm": 1.059566617012024, "learning_rate": 4.4258039465680326e-05, "loss": 2.2622, "step": 2851 }, { "epoch": 0.02565059607056612, "grad_norm": 5.238901138305664, "learning_rate": 4.425324612144506e-05, "loss": 2.3093, "step": 2852 }, { "epoch": 0.02565958996820657, "grad_norm": 0.7189644575119019, "learning_rate": 4.424845103712214e-05, "loss": 2.2175, "step": 2853 }, { "epoch": 0.025668583865847024, "grad_norm": 0.9174880385398865, "learning_rate": 4.424365421314495e-05, "loss": 2.2465, "step": 2854 }, { "epoch": 0.025677577763487475, "grad_norm": 0.7002683281898499, "learning_rate": 4.423885564994702e-05, "loss": 2.285, "step": 2855 }, { "epoch": 0.025686571661127925, "grad_norm": 1.6672979593276978, "learning_rate": 4.423405534796204e-05, "loss": 2.2238, "step": 2856 }, { "epoch": 0.025695565558768375, "grad_norm": 0.574082612991333, "learning_rate": 4.422925330762385e-05, "loss": 2.2126, "step": 2857 }, { "epoch": 0.025704559456408826, "grad_norm": 1.393553376197815, "learning_rate": 4.4224449529366465e-05, "loss": 2.3212, "step": 2858 }, { "epoch": 0.025713553354049276, "grad_norm": 0.9704312086105347, "learning_rate": 4.421964401362404e-05, "loss": 2.1611, "step": 2859 }, { "epoch": 0.02572254725168973, "grad_norm": 1.0220420360565186, "learning_rate": 4.421483676083088e-05, "loss": 2.2211, "step": 2860 }, { "epoch": 0.02573154114933018, "grad_norm": 1.4195069074630737, "learning_rate": 4.421002777142148e-05, "loss": 2.3202, "step": 2861 }, { "epoch": 0.02574053504697063, "grad_norm": 0.6300917267799377, "learning_rate": 4.4205217045830454e-05, "loss": 2.1453, "step": 2862 }, { "epoch": 0.02574952894461108, "grad_norm": 1.351525902748108, "learning_rate": 4.420040458449261e-05, "loss": 2.2807, "step": 2863 }, { "epoch": 0.02575852284225153, "grad_norm": 1.2535600662231445, "learning_rate": 4.419559038784287e-05, "loss": 2.2565, "step": 2864 }, { "epoch": 0.025767516739891982, "grad_norm": 1.135172963142395, "learning_rate": 4.419077445631634e-05, "loss": 2.2775, "step": 2865 }, { "epoch": 0.025776510637532436, "grad_norm": 1.8445550203323364, "learning_rate": 4.4185956790348296e-05, "loss": 2.2868, "step": 2866 }, { "epoch": 0.025785504535172886, "grad_norm": 1.069567084312439, "learning_rate": 4.418113739037414e-05, "loss": 2.2454, "step": 2867 }, { "epoch": 0.025794498432813336, "grad_norm": 1.228572964668274, "learning_rate": 4.4176316256829446e-05, "loss": 2.3004, "step": 2868 }, { "epoch": 0.025803492330453787, "grad_norm": 1.7094563245773315, "learning_rate": 4.4171493390149945e-05, "loss": 2.2916, "step": 2869 }, { "epoch": 0.025812486228094237, "grad_norm": 0.6854211688041687, "learning_rate": 4.4166668790771526e-05, "loss": 2.2955, "step": 2870 }, { "epoch": 0.025821480125734687, "grad_norm": 0.7049437165260315, "learning_rate": 4.416184245913022e-05, "loss": 2.2233, "step": 2871 }, { "epoch": 0.02583047402337514, "grad_norm": 1.7591519355773926, "learning_rate": 4.415701439566223e-05, "loss": 2.2874, "step": 2872 }, { "epoch": 0.02583946792101559, "grad_norm": 1.0526173114776611, "learning_rate": 4.415218460080392e-05, "loss": 2.2578, "step": 2873 }, { "epoch": 0.025848461818656042, "grad_norm": 1.1280505657196045, "learning_rate": 4.414735307499179e-05, "loss": 2.2621, "step": 2874 }, { "epoch": 0.025857455716296492, "grad_norm": 0.7453281879425049, "learning_rate": 4.414251981866251e-05, "loss": 2.2134, "step": 2875 }, { "epoch": 0.025866449613936943, "grad_norm": 2.4870922565460205, "learning_rate": 4.413768483225292e-05, "loss": 2.2538, "step": 2876 }, { "epoch": 0.025875443511577396, "grad_norm": 1.2357620000839233, "learning_rate": 4.4132848116199974e-05, "loss": 2.1583, "step": 2877 }, { "epoch": 0.025884437409217847, "grad_norm": 1.4477052688598633, "learning_rate": 4.4128009670940835e-05, "loss": 2.23, "step": 2878 }, { "epoch": 0.025893431306858297, "grad_norm": 1.3295570611953735, "learning_rate": 4.412316949691277e-05, "loss": 2.2579, "step": 2879 }, { "epoch": 0.025902425204498748, "grad_norm": 1.4048742055892944, "learning_rate": 4.411832759455326e-05, "loss": 2.2436, "step": 2880 }, { "epoch": 0.025911419102139198, "grad_norm": 0.774686872959137, "learning_rate": 4.411348396429989e-05, "loss": 2.2148, "step": 2881 }, { "epoch": 0.02592041299977965, "grad_norm": 0.7920633554458618, "learning_rate": 4.4108638606590425e-05, "loss": 2.2795, "step": 2882 }, { "epoch": 0.025929406897420102, "grad_norm": 0.5762454867362976, "learning_rate": 4.4103791521862784e-05, "loss": 2.2135, "step": 2883 }, { "epoch": 0.025938400795060552, "grad_norm": 0.8527852296829224, "learning_rate": 4.4098942710555035e-05, "loss": 2.2749, "step": 2884 }, { "epoch": 0.025947394692701003, "grad_norm": 1.0628376007080078, "learning_rate": 4.4094092173105426e-05, "loss": 2.2216, "step": 2885 }, { "epoch": 0.025956388590341453, "grad_norm": 1.1779074668884277, "learning_rate": 4.408923990995233e-05, "loss": 2.2547, "step": 2886 }, { "epoch": 0.025965382487981904, "grad_norm": 0.8033862113952637, "learning_rate": 4.408438592153429e-05, "loss": 2.2975, "step": 2887 }, { "epoch": 0.025974376385622354, "grad_norm": 0.5644456744194031, "learning_rate": 4.407953020829001e-05, "loss": 2.2465, "step": 2888 }, { "epoch": 0.025983370283262808, "grad_norm": 0.9251390099525452, "learning_rate": 4.407467277065834e-05, "loss": 2.1307, "step": 2889 }, { "epoch": 0.025992364180903258, "grad_norm": 0.8781344890594482, "learning_rate": 4.406981360907829e-05, "loss": 2.2561, "step": 2890 }, { "epoch": 0.02600135807854371, "grad_norm": 1.6725481748580933, "learning_rate": 4.4064952723989026e-05, "loss": 2.2712, "step": 2891 }, { "epoch": 0.02601035197618416, "grad_norm": 0.6144312620162964, "learning_rate": 4.4060090115829866e-05, "loss": 2.2125, "step": 2892 }, { "epoch": 0.02601934587382461, "grad_norm": 1.0812171697616577, "learning_rate": 4.40552257850403e-05, "loss": 2.2661, "step": 2893 }, { "epoch": 0.02602833977146506, "grad_norm": 1.6886677742004395, "learning_rate": 4.405035973205994e-05, "loss": 2.245, "step": 2894 }, { "epoch": 0.026037333669105513, "grad_norm": 1.0997364521026611, "learning_rate": 4.40454919573286e-05, "loss": 2.2034, "step": 2895 }, { "epoch": 0.026046327566745964, "grad_norm": 0.7956640124320984, "learning_rate": 4.404062246128621e-05, "loss": 2.2259, "step": 2896 }, { "epoch": 0.026055321464386414, "grad_norm": 1.0289183855056763, "learning_rate": 4.4035751244372856e-05, "loss": 2.2497, "step": 2897 }, { "epoch": 0.026064315362026864, "grad_norm": 1.0860953330993652, "learning_rate": 4.403087830702882e-05, "loss": 2.3148, "step": 2898 }, { "epoch": 0.026073309259667315, "grad_norm": 0.7038049697875977, "learning_rate": 4.40260036496945e-05, "loss": 2.2235, "step": 2899 }, { "epoch": 0.026082303157307765, "grad_norm": 0.7975780367851257, "learning_rate": 4.402112727281046e-05, "loss": 2.2342, "step": 2900 }, { "epoch": 0.02609129705494822, "grad_norm": 0.8370307087898254, "learning_rate": 4.4016249176817424e-05, "loss": 2.2343, "step": 2901 }, { "epoch": 0.02610029095258867, "grad_norm": 2.1911425590515137, "learning_rate": 4.401136936215628e-05, "loss": 2.3213, "step": 2902 }, { "epoch": 0.02610928485022912, "grad_norm": 1.0455611944198608, "learning_rate": 4.400648782926804e-05, "loss": 2.2441, "step": 2903 }, { "epoch": 0.02611827874786957, "grad_norm": 1.0040595531463623, "learning_rate": 4.40016045785939e-05, "loss": 2.2149, "step": 2904 }, { "epoch": 0.02612727264551002, "grad_norm": 0.8932300209999084, "learning_rate": 4.399671961057522e-05, "loss": 2.1851, "step": 2905 }, { "epoch": 0.02613626654315047, "grad_norm": 1.8096182346343994, "learning_rate": 4.399183292565347e-05, "loss": 2.2585, "step": 2906 }, { "epoch": 0.026145260440790925, "grad_norm": 0.7082765102386475, "learning_rate": 4.3986944524270315e-05, "loss": 2.2329, "step": 2907 }, { "epoch": 0.026154254338431375, "grad_norm": 1.4207112789154053, "learning_rate": 4.398205440686758e-05, "loss": 2.291, "step": 2908 }, { "epoch": 0.026163248236071825, "grad_norm": 0.7311521768569946, "learning_rate": 4.3977162573887196e-05, "loss": 2.2545, "step": 2909 }, { "epoch": 0.026172242133712276, "grad_norm": 0.9081767201423645, "learning_rate": 4.3972269025771304e-05, "loss": 2.2053, "step": 2910 }, { "epoch": 0.026181236031352726, "grad_norm": 0.6936066150665283, "learning_rate": 4.3967373762962184e-05, "loss": 2.1878, "step": 2911 }, { "epoch": 0.026190229928993176, "grad_norm": 0.7016533017158508, "learning_rate": 4.396247678590224e-05, "loss": 2.2393, "step": 2912 }, { "epoch": 0.02619922382663363, "grad_norm": 0.6537289619445801, "learning_rate": 4.3957578095034075e-05, "loss": 2.2512, "step": 2913 }, { "epoch": 0.02620821772427408, "grad_norm": 0.949700653553009, "learning_rate": 4.395267769080042e-05, "loss": 2.2411, "step": 2914 }, { "epoch": 0.02621721162191453, "grad_norm": 0.7158918976783752, "learning_rate": 4.394777557364417e-05, "loss": 2.2558, "step": 2915 }, { "epoch": 0.02622620551955498, "grad_norm": 1.9634621143341064, "learning_rate": 4.3942871744008374e-05, "loss": 2.2528, "step": 2916 }, { "epoch": 0.02623519941719543, "grad_norm": 1.4459389448165894, "learning_rate": 4.393796620233623e-05, "loss": 2.2326, "step": 2917 }, { "epoch": 0.026244193314835886, "grad_norm": 0.9597479701042175, "learning_rate": 4.3933058949071096e-05, "loss": 2.2511, "step": 2918 }, { "epoch": 0.026253187212476336, "grad_norm": 0.7239498496055603, "learning_rate": 4.392814998465649e-05, "loss": 2.2523, "step": 2919 }, { "epoch": 0.026262181110116786, "grad_norm": 0.9697819352149963, "learning_rate": 4.392323930953608e-05, "loss": 2.2566, "step": 2920 }, { "epoch": 0.026271175007757237, "grad_norm": 1.3568336963653564, "learning_rate": 4.391832692415368e-05, "loss": 2.1759, "step": 2921 }, { "epoch": 0.026280168905397687, "grad_norm": 0.9375148415565491, "learning_rate": 4.391341282895327e-05, "loss": 2.1796, "step": 2922 }, { "epoch": 0.026289162803038137, "grad_norm": 0.8493642210960388, "learning_rate": 4.3908497024378984e-05, "loss": 2.2326, "step": 2923 }, { "epoch": 0.02629815670067859, "grad_norm": 1.27517569065094, "learning_rate": 4.3903579510875095e-05, "loss": 2.2219, "step": 2924 }, { "epoch": 0.02630715059831904, "grad_norm": 2.5239834785461426, "learning_rate": 4.389866028888606e-05, "loss": 2.2375, "step": 2925 }, { "epoch": 0.026316144495959492, "grad_norm": 1.8744539022445679, "learning_rate": 4.389373935885646e-05, "loss": 2.2298, "step": 2926 }, { "epoch": 0.026325138393599942, "grad_norm": 1.1497787237167358, "learning_rate": 4.388881672123105e-05, "loss": 2.2389, "step": 2927 }, { "epoch": 0.026334132291240393, "grad_norm": 0.5491324067115784, "learning_rate": 4.388389237645473e-05, "loss": 2.2046, "step": 2928 }, { "epoch": 0.026343126188880843, "grad_norm": 1.1366963386535645, "learning_rate": 4.387896632497255e-05, "loss": 2.2149, "step": 2929 }, { "epoch": 0.026352120086521297, "grad_norm": 0.9640880227088928, "learning_rate": 4.387403856722974e-05, "loss": 2.261, "step": 2930 }, { "epoch": 0.026361113984161747, "grad_norm": 1.2636257410049438, "learning_rate": 4.386910910367164e-05, "loss": 2.238, "step": 2931 }, { "epoch": 0.026370107881802198, "grad_norm": 0.9522882699966431, "learning_rate": 4.386417793474379e-05, "loss": 2.2545, "step": 2932 }, { "epoch": 0.026379101779442648, "grad_norm": 0.9146173596382141, "learning_rate": 4.385924506089186e-05, "loss": 2.1864, "step": 2933 }, { "epoch": 0.026388095677083098, "grad_norm": 1.5206266641616821, "learning_rate": 4.385431048256167e-05, "loss": 2.2772, "step": 2934 }, { "epoch": 0.02639708957472355, "grad_norm": 0.8813039064407349, "learning_rate": 4.384937420019921e-05, "loss": 2.1912, "step": 2935 }, { "epoch": 0.026406083472364002, "grad_norm": 0.6774020791053772, "learning_rate": 4.3844436214250616e-05, "loss": 2.1672, "step": 2936 }, { "epoch": 0.026415077370004453, "grad_norm": 0.7874820232391357, "learning_rate": 4.3839496525162164e-05, "loss": 2.2341, "step": 2937 }, { "epoch": 0.026424071267644903, "grad_norm": 0.9825344681739807, "learning_rate": 4.383455513338032e-05, "loss": 2.2333, "step": 2938 }, { "epoch": 0.026433065165285353, "grad_norm": 1.0263444185256958, "learning_rate": 4.3829612039351664e-05, "loss": 2.2298, "step": 2939 }, { "epoch": 0.026442059062925804, "grad_norm": 1.5711126327514648, "learning_rate": 4.382466724352295e-05, "loss": 2.2472, "step": 2940 }, { "epoch": 0.026451052960566254, "grad_norm": 0.9089137315750122, "learning_rate": 4.38197207463411e-05, "loss": 2.2358, "step": 2941 }, { "epoch": 0.026460046858206708, "grad_norm": 1.5508273839950562, "learning_rate": 4.3814772548253144e-05, "loss": 2.1809, "step": 2942 }, { "epoch": 0.02646904075584716, "grad_norm": 1.0541675090789795, "learning_rate": 4.380982264970632e-05, "loss": 2.237, "step": 2943 }, { "epoch": 0.02647803465348761, "grad_norm": 1.1791822910308838, "learning_rate": 4.3804871051147975e-05, "loss": 2.2605, "step": 2944 }, { "epoch": 0.02648702855112806, "grad_norm": 0.5384827256202698, "learning_rate": 4.379991775302565e-05, "loss": 2.2355, "step": 2945 }, { "epoch": 0.02649602244876851, "grad_norm": 1.2648590803146362, "learning_rate": 4.379496275578701e-05, "loss": 2.1982, "step": 2946 }, { "epoch": 0.02650501634640896, "grad_norm": 0.8019316792488098, "learning_rate": 4.379000605987987e-05, "loss": 2.2496, "step": 2947 }, { "epoch": 0.026514010244049414, "grad_norm": 0.9262810349464417, "learning_rate": 4.3785047665752234e-05, "loss": 2.2342, "step": 2948 }, { "epoch": 0.026523004141689864, "grad_norm": 0.5679707527160645, "learning_rate": 4.3780087573852213e-05, "loss": 2.1917, "step": 2949 }, { "epoch": 0.026531998039330314, "grad_norm": 0.5395865440368652, "learning_rate": 4.377512578462811e-05, "loss": 2.2309, "step": 2950 }, { "epoch": 0.026540991936970765, "grad_norm": 1.0233451128005981, "learning_rate": 4.377016229852836e-05, "loss": 2.227, "step": 2951 }, { "epoch": 0.026549985834611215, "grad_norm": 1.0830150842666626, "learning_rate": 4.376519711600155e-05, "loss": 2.2622, "step": 2952 }, { "epoch": 0.02655897973225167, "grad_norm": 0.7942443490028381, "learning_rate": 4.3760230237496455e-05, "loss": 2.2707, "step": 2953 }, { "epoch": 0.02656797362989212, "grad_norm": 1.4294872283935547, "learning_rate": 4.3755261663461945e-05, "loss": 2.1308, "step": 2954 }, { "epoch": 0.02657696752753257, "grad_norm": 1.8997098207473755, "learning_rate": 4.3750291394347094e-05, "loss": 2.217, "step": 2955 }, { "epoch": 0.02658596142517302, "grad_norm": 1.725861668586731, "learning_rate": 4.374531943060109e-05, "loss": 2.2537, "step": 2956 }, { "epoch": 0.02659495532281347, "grad_norm": 0.5923041701316833, "learning_rate": 4.374034577267333e-05, "loss": 2.2584, "step": 2957 }, { "epoch": 0.02660394922045392, "grad_norm": 0.8764998316764832, "learning_rate": 4.373537042101329e-05, "loss": 2.2816, "step": 2958 }, { "epoch": 0.026612943118094375, "grad_norm": 0.7707982659339905, "learning_rate": 4.373039337607065e-05, "loss": 2.2481, "step": 2959 }, { "epoch": 0.026621937015734825, "grad_norm": 0.8721486926078796, "learning_rate": 4.3725414638295235e-05, "loss": 2.2286, "step": 2960 }, { "epoch": 0.026630930913375275, "grad_norm": 0.612036406993866, "learning_rate": 4.372043420813702e-05, "loss": 2.2143, "step": 2961 }, { "epoch": 0.026639924811015726, "grad_norm": 0.7116363644599915, "learning_rate": 4.3715452086046114e-05, "loss": 2.2654, "step": 2962 }, { "epoch": 0.026648918708656176, "grad_norm": 1.5265060663223267, "learning_rate": 4.3710468272472806e-05, "loss": 2.2672, "step": 2963 }, { "epoch": 0.026657912606296626, "grad_norm": 1.0516798496246338, "learning_rate": 4.3705482767867535e-05, "loss": 2.2461, "step": 2964 }, { "epoch": 0.02666690650393708, "grad_norm": 0.6807788014411926, "learning_rate": 4.3700495572680886e-05, "loss": 2.188, "step": 2965 }, { "epoch": 0.02667590040157753, "grad_norm": 0.8392161726951599, "learning_rate": 4.3695506687363576e-05, "loss": 2.2128, "step": 2966 }, { "epoch": 0.02668489429921798, "grad_norm": 0.8234874606132507, "learning_rate": 4.369051611236652e-05, "loss": 2.2992, "step": 2967 }, { "epoch": 0.02669388819685843, "grad_norm": 0.9318340420722961, "learning_rate": 4.368552384814074e-05, "loss": 2.2517, "step": 2968 }, { "epoch": 0.02670288209449888, "grad_norm": 1.3431357145309448, "learning_rate": 4.368052989513745e-05, "loss": 2.1426, "step": 2969 }, { "epoch": 0.026711875992139332, "grad_norm": 1.1976364850997925, "learning_rate": 4.367553425380798e-05, "loss": 2.1732, "step": 2970 }, { "epoch": 0.026720869889779786, "grad_norm": 1.1573487520217896, "learning_rate": 4.367053692460385e-05, "loss": 2.2425, "step": 2971 }, { "epoch": 0.026729863787420236, "grad_norm": 1.3955656290054321, "learning_rate": 4.3665537907976695e-05, "loss": 2.2644, "step": 2972 }, { "epoch": 0.026738857685060687, "grad_norm": 1.4529247283935547, "learning_rate": 4.366053720437833e-05, "loss": 2.3632, "step": 2973 }, { "epoch": 0.026747851582701137, "grad_norm": 0.9167463779449463, "learning_rate": 4.365553481426071e-05, "loss": 2.1637, "step": 2974 }, { "epoch": 0.026756845480341587, "grad_norm": 1.2519160509109497, "learning_rate": 4.3650530738075955e-05, "loss": 2.2536, "step": 2975 }, { "epoch": 0.026765839377982038, "grad_norm": 1.2280174493789673, "learning_rate": 4.364552497627632e-05, "loss": 2.2381, "step": 2976 }, { "epoch": 0.02677483327562249, "grad_norm": 0.6842616200447083, "learning_rate": 4.364051752931422e-05, "loss": 2.2133, "step": 2977 }, { "epoch": 0.026783827173262942, "grad_norm": 1.0951298475265503, "learning_rate": 4.363550839764222e-05, "loss": 2.3072, "step": 2978 }, { "epoch": 0.026792821070903392, "grad_norm": 1.1005184650421143, "learning_rate": 4.363049758171304e-05, "loss": 2.2413, "step": 2979 }, { "epoch": 0.026801814968543843, "grad_norm": 1.0165858268737793, "learning_rate": 4.3625485081979565e-05, "loss": 2.2096, "step": 2980 }, { "epoch": 0.026810808866184293, "grad_norm": 0.7957262396812439, "learning_rate": 4.362047089889481e-05, "loss": 2.2388, "step": 2981 }, { "epoch": 0.026819802763824743, "grad_norm": 1.1797778606414795, "learning_rate": 4.3615455032911946e-05, "loss": 2.1272, "step": 2982 }, { "epoch": 0.026828796661465197, "grad_norm": 0.8988558053970337, "learning_rate": 4.36104374844843e-05, "loss": 2.2288, "step": 2983 }, { "epoch": 0.026837790559105647, "grad_norm": 0.6336374282836914, "learning_rate": 4.360541825406537e-05, "loss": 2.1597, "step": 2984 }, { "epoch": 0.026846784456746098, "grad_norm": 2.3540353775024414, "learning_rate": 4.3600397342108776e-05, "loss": 2.2661, "step": 2985 }, { "epoch": 0.026855778354386548, "grad_norm": 1.1871908903121948, "learning_rate": 4.359537474906831e-05, "loss": 2.2179, "step": 2986 }, { "epoch": 0.026864772252027, "grad_norm": 0.867724597454071, "learning_rate": 4.359035047539789e-05, "loss": 2.2182, "step": 2987 }, { "epoch": 0.026873766149667452, "grad_norm": 1.1920708417892456, "learning_rate": 4.358532452155163e-05, "loss": 2.2325, "step": 2988 }, { "epoch": 0.026882760047307903, "grad_norm": 0.940662682056427, "learning_rate": 4.3580296887983754e-05, "loss": 2.2367, "step": 2989 }, { "epoch": 0.026891753944948353, "grad_norm": 0.7341658473014832, "learning_rate": 4.357526757514866e-05, "loss": 2.2548, "step": 2990 }, { "epoch": 0.026900747842588803, "grad_norm": 1.3730934858322144, "learning_rate": 4.357023658350089e-05, "loss": 2.187, "step": 2991 }, { "epoch": 0.026909741740229254, "grad_norm": 1.3443313837051392, "learning_rate": 4.356520391349513e-05, "loss": 2.2327, "step": 2992 }, { "epoch": 0.026918735637869704, "grad_norm": 0.6057613492012024, "learning_rate": 4.356016956558625e-05, "loss": 2.1788, "step": 2993 }, { "epoch": 0.026927729535510158, "grad_norm": 0.9672314524650574, "learning_rate": 4.3555133540229234e-05, "loss": 2.257, "step": 2994 }, { "epoch": 0.02693672343315061, "grad_norm": 1.084049940109253, "learning_rate": 4.355009583787923e-05, "loss": 2.2628, "step": 2995 }, { "epoch": 0.02694571733079106, "grad_norm": 0.9928491711616516, "learning_rate": 4.354505645899155e-05, "loss": 2.3137, "step": 2996 }, { "epoch": 0.02695471122843151, "grad_norm": 1.1049350500106812, "learning_rate": 4.354001540402164e-05, "loss": 2.3299, "step": 2997 }, { "epoch": 0.02696370512607196, "grad_norm": 1.051131010055542, "learning_rate": 4.35349726734251e-05, "loss": 2.2063, "step": 2998 }, { "epoch": 0.02697269902371241, "grad_norm": 1.0858001708984375, "learning_rate": 4.35299282676577e-05, "loss": 2.2553, "step": 2999 }, { "epoch": 0.026981692921352864, "grad_norm": 0.9315809607505798, "learning_rate": 4.352488218717535e-05, "loss": 2.1781, "step": 3000 }, { "epoch": 0.026990686818993314, "grad_norm": 164.34661865234375, "learning_rate": 4.351983443243409e-05, "loss": 10.5018, "step": 3001 }, { "epoch": 0.026999680716633764, "grad_norm": 233.47007751464844, "learning_rate": 4.351478500389015e-05, "loss": 9.416, "step": 3002 }, { "epoch": 0.027008674614274215, "grad_norm": 230.17076110839844, "learning_rate": 4.350973390199988e-05, "loss": 6.9619, "step": 3003 }, { "epoch": 0.027017668511914665, "grad_norm": 92.68528747558594, "learning_rate": 4.3504681127219793e-05, "loss": 7.6505, "step": 3004 }, { "epoch": 0.027026662409555115, "grad_norm": 131.80685424804688, "learning_rate": 4.3499626680006565e-05, "loss": 5.9782, "step": 3005 }, { "epoch": 0.02703565630719557, "grad_norm": 118.5185546875, "learning_rate": 4.3494570560817e-05, "loss": 5.2924, "step": 3006 }, { "epoch": 0.02704465020483602, "grad_norm": 71.25559997558594, "learning_rate": 4.348951277010807e-05, "loss": 4.8907, "step": 3007 }, { "epoch": 0.02705364410247647, "grad_norm": 83.69850158691406, "learning_rate": 4.3484453308336884e-05, "loss": 4.4758, "step": 3008 }, { "epoch": 0.02706263800011692, "grad_norm": 65.12527465820312, "learning_rate": 4.347939217596073e-05, "loss": 4.2166, "step": 3009 }, { "epoch": 0.02707163189775737, "grad_norm": 23.537372589111328, "learning_rate": 4.3474329373437e-05, "loss": 3.8609, "step": 3010 }, { "epoch": 0.02708062579539782, "grad_norm": 31.924802780151367, "learning_rate": 4.346926490122329e-05, "loss": 3.5514, "step": 3011 }, { "epoch": 0.027089619693038275, "grad_norm": 50.665122985839844, "learning_rate": 4.3464198759777316e-05, "loss": 3.484, "step": 3012 }, { "epoch": 0.027098613590678725, "grad_norm": 26.982336044311523, "learning_rate": 4.3459130949556935e-05, "loss": 3.4117, "step": 3013 }, { "epoch": 0.027107607488319176, "grad_norm": 18.523277282714844, "learning_rate": 4.345406147102019e-05, "loss": 3.0359, "step": 3014 }, { "epoch": 0.027116601385959626, "grad_norm": 15.057336807250977, "learning_rate": 4.3448990324625244e-05, "loss": 3.0179, "step": 3015 }, { "epoch": 0.027125595283600076, "grad_norm": 24.163381576538086, "learning_rate": 4.3443917510830425e-05, "loss": 3.6586, "step": 3016 }, { "epoch": 0.027134589181240527, "grad_norm": 20.82476806640625, "learning_rate": 4.34388430300942e-05, "loss": 2.8927, "step": 3017 }, { "epoch": 0.02714358307888098, "grad_norm": 9.124222755432129, "learning_rate": 4.343376688287522e-05, "loss": 3.1373, "step": 3018 }, { "epoch": 0.02715257697652143, "grad_norm": 8.222960472106934, "learning_rate": 4.3428689069632224e-05, "loss": 2.9814, "step": 3019 }, { "epoch": 0.02716157087416188, "grad_norm": 15.590635299682617, "learning_rate": 4.342360959082417e-05, "loss": 3.0653, "step": 3020 }, { "epoch": 0.02717056477180233, "grad_norm": 6.372993469238281, "learning_rate": 4.341852844691012e-05, "loss": 2.4734, "step": 3021 }, { "epoch": 0.027179558669442782, "grad_norm": 5.414846897125244, "learning_rate": 4.3413445638349315e-05, "loss": 2.9569, "step": 3022 }, { "epoch": 0.027188552567083232, "grad_norm": 5.210756778717041, "learning_rate": 4.3408361165601124e-05, "loss": 2.96, "step": 3023 }, { "epoch": 0.027197546464723686, "grad_norm": 7.2568488121032715, "learning_rate": 4.340327502912508e-05, "loss": 2.7437, "step": 3024 }, { "epoch": 0.027206540362364137, "grad_norm": 7.10967493057251, "learning_rate": 4.339818722938086e-05, "loss": 3.0871, "step": 3025 }, { "epoch": 0.027215534260004587, "grad_norm": 7.578812599182129, "learning_rate": 4.3393097766828293e-05, "loss": 2.4839, "step": 3026 }, { "epoch": 0.027224528157645037, "grad_norm": 4.991209030151367, "learning_rate": 4.338800664192736e-05, "loss": 2.8594, "step": 3027 }, { "epoch": 0.027233522055285488, "grad_norm": 5.278810024261475, "learning_rate": 4.3382913855138195e-05, "loss": 2.9281, "step": 3028 }, { "epoch": 0.02724251595292594, "grad_norm": 4.117546558380127, "learning_rate": 4.337781940692108e-05, "loss": 2.7895, "step": 3029 }, { "epoch": 0.027251509850566392, "grad_norm": 5.537631988525391, "learning_rate": 4.337272329773644e-05, "loss": 2.5872, "step": 3030 }, { "epoch": 0.027260503748206842, "grad_norm": 23.576351165771484, "learning_rate": 4.3367625528044855e-05, "loss": 2.9932, "step": 3031 }, { "epoch": 0.027269497645847292, "grad_norm": 6.555099964141846, "learning_rate": 4.336252609830706e-05, "loss": 2.6899, "step": 3032 }, { "epoch": 0.027278491543487743, "grad_norm": 5.0880446434021, "learning_rate": 4.335742500898393e-05, "loss": 2.5752, "step": 3033 }, { "epoch": 0.027287485441128193, "grad_norm": 5.363535404205322, "learning_rate": 4.335232226053651e-05, "loss": 3.0538, "step": 3034 }, { "epoch": 0.027296479338768647, "grad_norm": 8.71766185760498, "learning_rate": 4.334721785342597e-05, "loss": 3.6094, "step": 3035 }, { "epoch": 0.027305473236409097, "grad_norm": 5.477747917175293, "learning_rate": 4.334211178811364e-05, "loss": 3.0263, "step": 3036 }, { "epoch": 0.027314467134049548, "grad_norm": 5.022616863250732, "learning_rate": 4.3337004065061e-05, "loss": 2.9335, "step": 3037 }, { "epoch": 0.027323461031689998, "grad_norm": 5.908479690551758, "learning_rate": 4.3331894684729686e-05, "loss": 3.1842, "step": 3038 }, { "epoch": 0.02733245492933045, "grad_norm": 3.5384485721588135, "learning_rate": 4.332678364758147e-05, "loss": 2.6095, "step": 3039 }, { "epoch": 0.0273414488269709, "grad_norm": 4.0285820960998535, "learning_rate": 4.3321670954078295e-05, "loss": 2.6206, "step": 3040 }, { "epoch": 0.027350442724611353, "grad_norm": 5.005476951599121, "learning_rate": 4.3316556604682236e-05, "loss": 2.759, "step": 3041 }, { "epoch": 0.027359436622251803, "grad_norm": 4.230103015899658, "learning_rate": 4.3311440599855516e-05, "loss": 2.4586, "step": 3042 }, { "epoch": 0.027368430519892253, "grad_norm": 3.368016004562378, "learning_rate": 4.330632294006052e-05, "loss": 2.4077, "step": 3043 }, { "epoch": 0.027377424417532704, "grad_norm": 5.533243656158447, "learning_rate": 4.330120362575977e-05, "loss": 3.0158, "step": 3044 }, { "epoch": 0.027386418315173154, "grad_norm": 6.317595958709717, "learning_rate": 4.329608265741595e-05, "loss": 3.0374, "step": 3045 }, { "epoch": 0.027395412212813604, "grad_norm": 6.311054706573486, "learning_rate": 4.329096003549189e-05, "loss": 2.7498, "step": 3046 }, { "epoch": 0.02740440611045406, "grad_norm": 3.863466739654541, "learning_rate": 4.328583576045057e-05, "loss": 2.7746, "step": 3047 }, { "epoch": 0.02741340000809451, "grad_norm": 4.124538421630859, "learning_rate": 4.3280709832755094e-05, "loss": 2.8438, "step": 3048 }, { "epoch": 0.02742239390573496, "grad_norm": 1.9168682098388672, "learning_rate": 4.3275582252868764e-05, "loss": 2.4297, "step": 3049 }, { "epoch": 0.02743138780337541, "grad_norm": 2.903076648712158, "learning_rate": 4.3270453021254994e-05, "loss": 2.4087, "step": 3050 }, { "epoch": 0.02744038170101586, "grad_norm": 6.916593551635742, "learning_rate": 4.3265322138377354e-05, "loss": 2.8712, "step": 3051 }, { "epoch": 0.02744937559865631, "grad_norm": 4.9376983642578125, "learning_rate": 4.326018960469957e-05, "loss": 2.3593, "step": 3052 }, { "epoch": 0.027458369496296764, "grad_norm": 3.7438056468963623, "learning_rate": 4.325505542068553e-05, "loss": 2.6833, "step": 3053 }, { "epoch": 0.027467363393937214, "grad_norm": 3.5138394832611084, "learning_rate": 4.324991958679924e-05, "loss": 2.634, "step": 3054 }, { "epoch": 0.027476357291577665, "grad_norm": 3.7332141399383545, "learning_rate": 4.324478210350487e-05, "loss": 2.8016, "step": 3055 }, { "epoch": 0.027485351189218115, "grad_norm": 3.705418109893799, "learning_rate": 4.323964297126675e-05, "loss": 2.8465, "step": 3056 }, { "epoch": 0.027494345086858565, "grad_norm": 2.137070894241333, "learning_rate": 4.3234502190549345e-05, "loss": 2.3547, "step": 3057 }, { "epoch": 0.027503338984499016, "grad_norm": 6.680044651031494, "learning_rate": 4.322935976181727e-05, "loss": 2.6226, "step": 3058 }, { "epoch": 0.02751233288213947, "grad_norm": 4.464818000793457, "learning_rate": 4.3224215685535294e-05, "loss": 3.0795, "step": 3059 }, { "epoch": 0.02752132677977992, "grad_norm": 4.232580661773682, "learning_rate": 4.3219069962168336e-05, "loss": 2.6682, "step": 3060 }, { "epoch": 0.02753032067742037, "grad_norm": 4.284449577331543, "learning_rate": 4.321392259218145e-05, "loss": 2.7556, "step": 3061 }, { "epoch": 0.02753931457506082, "grad_norm": 2.653223752975464, "learning_rate": 4.3208773576039865e-05, "loss": 2.5548, "step": 3062 }, { "epoch": 0.02754830847270127, "grad_norm": 2.9533090591430664, "learning_rate": 4.320362291420894e-05, "loss": 2.7221, "step": 3063 }, { "epoch": 0.027557302370341725, "grad_norm": 2.6692256927490234, "learning_rate": 4.319847060715419e-05, "loss": 2.7942, "step": 3064 }, { "epoch": 0.027566296267982175, "grad_norm": 4.573818683624268, "learning_rate": 4.3193316655341255e-05, "loss": 3.1693, "step": 3065 }, { "epoch": 0.027575290165622626, "grad_norm": 3.019530773162842, "learning_rate": 4.318816105923596e-05, "loss": 2.7297, "step": 3066 }, { "epoch": 0.027584284063263076, "grad_norm": 2.9747629165649414, "learning_rate": 4.318300381930426e-05, "loss": 2.8603, "step": 3067 }, { "epoch": 0.027593277960903526, "grad_norm": 3.9518773555755615, "learning_rate": 4.317784493601226e-05, "loss": 2.8617, "step": 3068 }, { "epoch": 0.027602271858543977, "grad_norm": 4.637152194976807, "learning_rate": 4.3172684409826216e-05, "loss": 2.9972, "step": 3069 }, { "epoch": 0.02761126575618443, "grad_norm": 2.022292137145996, "learning_rate": 4.316752224121252e-05, "loss": 2.2848, "step": 3070 }, { "epoch": 0.02762025965382488, "grad_norm": 9.13784122467041, "learning_rate": 4.316235843063775e-05, "loss": 2.7052, "step": 3071 }, { "epoch": 0.02762925355146533, "grad_norm": 4.085910797119141, "learning_rate": 4.315719297856857e-05, "loss": 2.8563, "step": 3072 }, { "epoch": 0.02763824744910578, "grad_norm": 4.014984607696533, "learning_rate": 4.3152025885471856e-05, "loss": 2.9059, "step": 3073 }, { "epoch": 0.027647241346746232, "grad_norm": 3.9625484943389893, "learning_rate": 4.314685715181459e-05, "loss": 2.7529, "step": 3074 }, { "epoch": 0.027656235244386682, "grad_norm": 1.9315695762634277, "learning_rate": 4.314168677806393e-05, "loss": 2.5964, "step": 3075 }, { "epoch": 0.027665229142027136, "grad_norm": 1.4846476316452026, "learning_rate": 4.313651476468715e-05, "loss": 2.3712, "step": 3076 }, { "epoch": 0.027674223039667586, "grad_norm": 2.028611421585083, "learning_rate": 4.31313411121517e-05, "loss": 2.7029, "step": 3077 }, { "epoch": 0.027683216937308037, "grad_norm": 4.492123603820801, "learning_rate": 4.3126165820925174e-05, "loss": 2.432, "step": 3078 }, { "epoch": 0.027692210834948487, "grad_norm": 1.8397845029830933, "learning_rate": 4.3120988891475306e-05, "loss": 2.5761, "step": 3079 }, { "epoch": 0.027701204732588938, "grad_norm": 2.2878644466400146, "learning_rate": 4.311581032426998e-05, "loss": 2.7872, "step": 3080 }, { "epoch": 0.027710198630229388, "grad_norm": 1.6362402439117432, "learning_rate": 4.311063011977723e-05, "loss": 2.3746, "step": 3081 }, { "epoch": 0.02771919252786984, "grad_norm": 2.5059030055999756, "learning_rate": 4.310544827846523e-05, "loss": 2.6604, "step": 3082 }, { "epoch": 0.027728186425510292, "grad_norm": 2.533919334411621, "learning_rate": 4.310026480080233e-05, "loss": 2.972, "step": 3083 }, { "epoch": 0.027737180323150742, "grad_norm": 4.72614049911499, "learning_rate": 4.3095079687256976e-05, "loss": 2.6965, "step": 3084 }, { "epoch": 0.027746174220791193, "grad_norm": 2.4846439361572266, "learning_rate": 4.3089892938297814e-05, "loss": 3.2602, "step": 3085 }, { "epoch": 0.027755168118431643, "grad_norm": 1.5414323806762695, "learning_rate": 4.308470455439362e-05, "loss": 2.5499, "step": 3086 }, { "epoch": 0.027764162016072094, "grad_norm": 36.09817123413086, "learning_rate": 4.307951453601331e-05, "loss": 2.7794, "step": 3087 }, { "epoch": 0.027773155913712547, "grad_norm": 29.671833038330078, "learning_rate": 4.307432288362594e-05, "loss": 2.5502, "step": 3088 }, { "epoch": 0.027782149811352998, "grad_norm": 6.110573768615723, "learning_rate": 4.306912959770074e-05, "loss": 2.8371, "step": 3089 }, { "epoch": 0.027791143708993448, "grad_norm": 2.1736464500427246, "learning_rate": 4.3063934678707076e-05, "loss": 2.645, "step": 3090 }, { "epoch": 0.0278001376066339, "grad_norm": 1.6040300130844116, "learning_rate": 4.305873812711444e-05, "loss": 2.4166, "step": 3091 }, { "epoch": 0.02780913150427435, "grad_norm": 2.115355968475342, "learning_rate": 4.305353994339252e-05, "loss": 2.6523, "step": 3092 }, { "epoch": 0.0278181254019148, "grad_norm": 2.474229335784912, "learning_rate": 4.304834012801109e-05, "loss": 2.7929, "step": 3093 }, { "epoch": 0.027827119299555253, "grad_norm": 2.3080015182495117, "learning_rate": 4.304313868144013e-05, "loss": 2.8269, "step": 3094 }, { "epoch": 0.027836113197195703, "grad_norm": 1.780664086341858, "learning_rate": 4.3037935604149725e-05, "loss": 2.7244, "step": 3095 }, { "epoch": 0.027845107094836154, "grad_norm": 2.1991262435913086, "learning_rate": 4.303273089661013e-05, "loss": 2.8051, "step": 3096 }, { "epoch": 0.027854100992476604, "grad_norm": 2.9452619552612305, "learning_rate": 4.3027524559291735e-05, "loss": 2.9417, "step": 3097 }, { "epoch": 0.027863094890117054, "grad_norm": 1.3555408716201782, "learning_rate": 4.302231659266509e-05, "loss": 2.3402, "step": 3098 }, { "epoch": 0.027872088787757508, "grad_norm": 2.3037586212158203, "learning_rate": 4.301710699720089e-05, "loss": 3.0345, "step": 3099 }, { "epoch": 0.02788108268539796, "grad_norm": 2.7883963584899902, "learning_rate": 4.301189577336996e-05, "loss": 2.7237, "step": 3100 }, { "epoch": 0.02789007658303841, "grad_norm": 1.7890801429748535, "learning_rate": 4.3006682921643296e-05, "loss": 2.8227, "step": 3101 }, { "epoch": 0.02789907048067886, "grad_norm": 1.8575735092163086, "learning_rate": 4.300146844249202e-05, "loss": 2.4907, "step": 3102 }, { "epoch": 0.02790806437831931, "grad_norm": 2.660573720932007, "learning_rate": 4.2996252336387414e-05, "loss": 2.633, "step": 3103 }, { "epoch": 0.02791705827595976, "grad_norm": 1.3236286640167236, "learning_rate": 4.299103460380091e-05, "loss": 2.4318, "step": 3104 }, { "epoch": 0.027926052173600214, "grad_norm": 2.138324499130249, "learning_rate": 4.298581524520408e-05, "loss": 2.5578, "step": 3105 }, { "epoch": 0.027935046071240664, "grad_norm": 2.475240707397461, "learning_rate": 4.2980594261068636e-05, "loss": 3.1449, "step": 3106 }, { "epoch": 0.027944039968881115, "grad_norm": 2.030167579650879, "learning_rate": 4.297537165186645e-05, "loss": 2.6711, "step": 3107 }, { "epoch": 0.027953033866521565, "grad_norm": 2.4538168907165527, "learning_rate": 4.2970147418069536e-05, "loss": 2.4143, "step": 3108 }, { "epoch": 0.027962027764162015, "grad_norm": 1.8144887685775757, "learning_rate": 4.2964921560150054e-05, "loss": 2.716, "step": 3109 }, { "epoch": 0.027971021661802466, "grad_norm": 2.146554946899414, "learning_rate": 4.2959694078580315e-05, "loss": 2.9401, "step": 3110 }, { "epoch": 0.02798001555944292, "grad_norm": 1.9110617637634277, "learning_rate": 4.295446497383276e-05, "loss": 2.8912, "step": 3111 }, { "epoch": 0.02798900945708337, "grad_norm": 1.7476952075958252, "learning_rate": 4.294923424638001e-05, "loss": 2.6514, "step": 3112 }, { "epoch": 0.02799800335472382, "grad_norm": 1.7322272062301636, "learning_rate": 4.29440018966948e-05, "loss": 2.932, "step": 3113 }, { "epoch": 0.02800699725236427, "grad_norm": 1.5488052368164062, "learning_rate": 4.293876792525002e-05, "loss": 2.5939, "step": 3114 }, { "epoch": 0.02801599115000472, "grad_norm": 1.2806622982025146, "learning_rate": 4.293353233251872e-05, "loss": 2.2836, "step": 3115 }, { "epoch": 0.02802498504764517, "grad_norm": 2.36108136177063, "learning_rate": 4.292829511897409e-05, "loss": 2.8845, "step": 3116 }, { "epoch": 0.028033978945285625, "grad_norm": 1.91974675655365, "learning_rate": 4.292305628508945e-05, "loss": 2.5302, "step": 3117 }, { "epoch": 0.028042972842926076, "grad_norm": 2.0816078186035156, "learning_rate": 4.291781583133829e-05, "loss": 2.5868, "step": 3118 }, { "epoch": 0.028051966740566526, "grad_norm": 1.8561017513275146, "learning_rate": 4.291257375819424e-05, "loss": 2.7714, "step": 3119 }, { "epoch": 0.028060960638206976, "grad_norm": 2.3980798721313477, "learning_rate": 4.290733006613107e-05, "loss": 2.7932, "step": 3120 }, { "epoch": 0.028069954535847427, "grad_norm": 1.2275316715240479, "learning_rate": 4.290208475562268e-05, "loss": 2.3113, "step": 3121 }, { "epoch": 0.028078948433487877, "grad_norm": 2.2904083728790283, "learning_rate": 4.2896837827143164e-05, "loss": 2.7037, "step": 3122 }, { "epoch": 0.02808794233112833, "grad_norm": 1.7793763875961304, "learning_rate": 4.289158928116672e-05, "loss": 2.6535, "step": 3123 }, { "epoch": 0.02809693622876878, "grad_norm": 1.7803759574890137, "learning_rate": 4.288633911816772e-05, "loss": 2.6141, "step": 3124 }, { "epoch": 0.02810593012640923, "grad_norm": 1.8528448343276978, "learning_rate": 4.288108733862064e-05, "loss": 2.7916, "step": 3125 }, { "epoch": 0.028114924024049682, "grad_norm": 2.605508804321289, "learning_rate": 4.287583394300015e-05, "loss": 3.0234, "step": 3126 }, { "epoch": 0.028123917921690132, "grad_norm": 2.0806243419647217, "learning_rate": 4.287057893178105e-05, "loss": 2.2836, "step": 3127 }, { "epoch": 0.028132911819330583, "grad_norm": 1.8188472986221313, "learning_rate": 4.286532230543826e-05, "loss": 2.7764, "step": 3128 }, { "epoch": 0.028141905716971036, "grad_norm": 2.1568543910980225, "learning_rate": 4.2860064064446896e-05, "loss": 2.5176, "step": 3129 }, { "epoch": 0.028150899614611487, "grad_norm": 2.104617118835449, "learning_rate": 4.2854804209282176e-05, "loss": 3.0131, "step": 3130 }, { "epoch": 0.028159893512251937, "grad_norm": 2.2620439529418945, "learning_rate": 4.2849542740419484e-05, "loss": 2.7694, "step": 3131 }, { "epoch": 0.028168887409892387, "grad_norm": 1.6867849826812744, "learning_rate": 4.2844279658334344e-05, "loss": 2.7187, "step": 3132 }, { "epoch": 0.028177881307532838, "grad_norm": 1.5512785911560059, "learning_rate": 4.283901496350243e-05, "loss": 2.2364, "step": 3133 }, { "epoch": 0.028186875205173288, "grad_norm": 1.7382086515426636, "learning_rate": 4.283374865639955e-05, "loss": 2.5764, "step": 3134 }, { "epoch": 0.028195869102813742, "grad_norm": 2.1844022274017334, "learning_rate": 4.2828480737501684e-05, "loss": 2.7991, "step": 3135 }, { "epoch": 0.028204863000454192, "grad_norm": 2.8010752201080322, "learning_rate": 4.282321120728493e-05, "loss": 2.9881, "step": 3136 }, { "epoch": 0.028213856898094643, "grad_norm": 2.2111332416534424, "learning_rate": 4.281794006622554e-05, "loss": 2.6648, "step": 3137 }, { "epoch": 0.028222850795735093, "grad_norm": 2.12583065032959, "learning_rate": 4.2812667314799924e-05, "loss": 2.795, "step": 3138 }, { "epoch": 0.028231844693375543, "grad_norm": 1.688863754272461, "learning_rate": 4.2807392953484624e-05, "loss": 2.8978, "step": 3139 }, { "epoch": 0.028240838591015997, "grad_norm": 2.12074875831604, "learning_rate": 4.280211698275632e-05, "loss": 2.8603, "step": 3140 }, { "epoch": 0.028249832488656448, "grad_norm": 2.3826851844787598, "learning_rate": 4.279683940309187e-05, "loss": 2.8939, "step": 3141 }, { "epoch": 0.028258826386296898, "grad_norm": 2.2437212467193604, "learning_rate": 4.2791560214968244e-05, "loss": 2.5449, "step": 3142 }, { "epoch": 0.02826782028393735, "grad_norm": 2.227304220199585, "learning_rate": 4.278627941886256e-05, "loss": 2.9616, "step": 3143 }, { "epoch": 0.0282768141815778, "grad_norm": 0.8882721066474915, "learning_rate": 4.27809970152521e-05, "loss": 2.3797, "step": 3144 }, { "epoch": 0.02828580807921825, "grad_norm": 1.1870465278625488, "learning_rate": 4.277571300461429e-05, "loss": 2.3518, "step": 3145 }, { "epoch": 0.028294801976858703, "grad_norm": 2.0124857425689697, "learning_rate": 4.2770427387426684e-05, "loss": 2.8119, "step": 3146 }, { "epoch": 0.028303795874499153, "grad_norm": 2.1447248458862305, "learning_rate": 4.2765140164166984e-05, "loss": 2.8031, "step": 3147 }, { "epoch": 0.028312789772139604, "grad_norm": 1.1922261714935303, "learning_rate": 4.275985133531306e-05, "loss": 2.6453, "step": 3148 }, { "epoch": 0.028321783669780054, "grad_norm": 1.924161672592163, "learning_rate": 4.2754560901342893e-05, "loss": 2.9501, "step": 3149 }, { "epoch": 0.028330777567420504, "grad_norm": 1.5944032669067383, "learning_rate": 4.2749268862734634e-05, "loss": 2.4608, "step": 3150 }, { "epoch": 0.028339771465060955, "grad_norm": 1.1359895467758179, "learning_rate": 4.274397521996658e-05, "loss": 2.5973, "step": 3151 }, { "epoch": 0.02834876536270141, "grad_norm": 1.8088732957839966, "learning_rate": 4.273867997351716e-05, "loss": 2.6777, "step": 3152 }, { "epoch": 0.02835775926034186, "grad_norm": 1.335385799407959, "learning_rate": 4.273338312386495e-05, "loss": 2.4901, "step": 3153 }, { "epoch": 0.02836675315798231, "grad_norm": 1.6990774869918823, "learning_rate": 4.272808467148867e-05, "loss": 2.5551, "step": 3154 }, { "epoch": 0.02837574705562276, "grad_norm": 1.7167222499847412, "learning_rate": 4.2722784616867194e-05, "loss": 2.6493, "step": 3155 }, { "epoch": 0.02838474095326321, "grad_norm": 2.01309871673584, "learning_rate": 4.271748296047953e-05, "loss": 2.6202, "step": 3156 }, { "epoch": 0.02839373485090366, "grad_norm": 1.5969191789627075, "learning_rate": 4.271217970280484e-05, "loss": 2.8733, "step": 3157 }, { "epoch": 0.028402728748544114, "grad_norm": 2.0551114082336426, "learning_rate": 4.270687484432243e-05, "loss": 2.9857, "step": 3158 }, { "epoch": 0.028411722646184565, "grad_norm": 1.429969072341919, "learning_rate": 4.270156838551175e-05, "loss": 2.5453, "step": 3159 }, { "epoch": 0.028420716543825015, "grad_norm": 1.4842311143875122, "learning_rate": 4.269626032685238e-05, "loss": 2.3403, "step": 3160 }, { "epoch": 0.028429710441465465, "grad_norm": 2.7709639072418213, "learning_rate": 4.269095066882406e-05, "loss": 2.593, "step": 3161 }, { "epoch": 0.028438704339105916, "grad_norm": 1.6040809154510498, "learning_rate": 4.268563941190668e-05, "loss": 2.5595, "step": 3162 }, { "epoch": 0.028447698236746366, "grad_norm": 2.7668447494506836, "learning_rate": 4.2680326556580243e-05, "loss": 2.758, "step": 3163 }, { "epoch": 0.02845669213438682, "grad_norm": 2.8213260173797607, "learning_rate": 4.2675012103324955e-05, "loss": 2.9854, "step": 3164 }, { "epoch": 0.02846568603202727, "grad_norm": 1.925601840019226, "learning_rate": 4.26696960526211e-05, "loss": 2.7617, "step": 3165 }, { "epoch": 0.02847467992966772, "grad_norm": 1.4193756580352783, "learning_rate": 4.266437840494915e-05, "loss": 2.4942, "step": 3166 }, { "epoch": 0.02848367382730817, "grad_norm": 1.5050954818725586, "learning_rate": 4.2659059160789704e-05, "loss": 2.7118, "step": 3167 }, { "epoch": 0.02849266772494862, "grad_norm": 1.589365839958191, "learning_rate": 4.265373832062352e-05, "loss": 2.4921, "step": 3168 }, { "epoch": 0.02850166162258907, "grad_norm": 1.522916555404663, "learning_rate": 4.2648415884931476e-05, "loss": 2.6795, "step": 3169 }, { "epoch": 0.028510655520229525, "grad_norm": 1.8381823301315308, "learning_rate": 4.264309185419462e-05, "loss": 2.52, "step": 3170 }, { "epoch": 0.028519649417869976, "grad_norm": 1.6782310009002686, "learning_rate": 4.2637766228894114e-05, "loss": 2.5847, "step": 3171 }, { "epoch": 0.028528643315510426, "grad_norm": 1.6060789823532104, "learning_rate": 4.26324390095113e-05, "loss": 2.7593, "step": 3172 }, { "epoch": 0.028537637213150877, "grad_norm": 1.9160151481628418, "learning_rate": 4.262711019652764e-05, "loss": 2.78, "step": 3173 }, { "epoch": 0.028546631110791327, "grad_norm": 1.9057458639144897, "learning_rate": 4.262177979042474e-05, "loss": 2.6166, "step": 3174 }, { "epoch": 0.02855562500843178, "grad_norm": 1.3022676706314087, "learning_rate": 4.2616447791684375e-05, "loss": 2.7484, "step": 3175 }, { "epoch": 0.02856461890607223, "grad_norm": 1.7138577699661255, "learning_rate": 4.261111420078843e-05, "loss": 2.5383, "step": 3176 }, { "epoch": 0.02857361280371268, "grad_norm": 1.5676833391189575, "learning_rate": 4.2605779018218954e-05, "loss": 2.4569, "step": 3177 }, { "epoch": 0.028582606701353132, "grad_norm": 2.05310320854187, "learning_rate": 4.2600442244458135e-05, "loss": 2.9552, "step": 3178 }, { "epoch": 0.028591600598993582, "grad_norm": 2.0233919620513916, "learning_rate": 4.25951038799883e-05, "loss": 3.0334, "step": 3179 }, { "epoch": 0.028600594496634033, "grad_norm": 3.085414171218872, "learning_rate": 4.258976392529192e-05, "loss": 2.7832, "step": 3180 }, { "epoch": 0.028609588394274486, "grad_norm": 1.612722396850586, "learning_rate": 4.258442238085163e-05, "loss": 2.594, "step": 3181 }, { "epoch": 0.028618582291914937, "grad_norm": 2.3569014072418213, "learning_rate": 4.25790792471502e-05, "loss": 2.5299, "step": 3182 }, { "epoch": 0.028627576189555387, "grad_norm": 1.8709267377853394, "learning_rate": 4.2573734524670505e-05, "loss": 2.5311, "step": 3183 }, { "epoch": 0.028636570087195837, "grad_norm": 1.603458046913147, "learning_rate": 4.256838821389562e-05, "loss": 2.4503, "step": 3184 }, { "epoch": 0.028645563984836288, "grad_norm": 2.5641987323760986, "learning_rate": 4.256304031530872e-05, "loss": 2.845, "step": 3185 }, { "epoch": 0.028654557882476738, "grad_norm": 1.6905276775360107, "learning_rate": 4.255769082939316e-05, "loss": 2.5367, "step": 3186 }, { "epoch": 0.028663551780117192, "grad_norm": 1.830762505531311, "learning_rate": 4.255233975663242e-05, "loss": 2.837, "step": 3187 }, { "epoch": 0.028672545677757642, "grad_norm": 2.191316843032837, "learning_rate": 4.2546987097510117e-05, "loss": 2.9806, "step": 3188 }, { "epoch": 0.028681539575398093, "grad_norm": 1.8977962732315063, "learning_rate": 4.254163285251002e-05, "loss": 3.025, "step": 3189 }, { "epoch": 0.028690533473038543, "grad_norm": 1.6529710292816162, "learning_rate": 4.2536277022116035e-05, "loss": 2.552, "step": 3190 }, { "epoch": 0.028699527370678993, "grad_norm": 1.0790417194366455, "learning_rate": 4.2530919606812216e-05, "loss": 2.2753, "step": 3191 }, { "epoch": 0.028708521268319444, "grad_norm": 2.420034170150757, "learning_rate": 4.252556060708277e-05, "loss": 2.9323, "step": 3192 }, { "epoch": 0.028717515165959898, "grad_norm": 2.2509121894836426, "learning_rate": 4.252020002341204e-05, "loss": 2.9628, "step": 3193 }, { "epoch": 0.028726509063600348, "grad_norm": 2.715315341949463, "learning_rate": 4.251483785628451e-05, "loss": 2.8185, "step": 3194 }, { "epoch": 0.0287355029612408, "grad_norm": 1.6430678367614746, "learning_rate": 4.250947410618478e-05, "loss": 2.5061, "step": 3195 }, { "epoch": 0.02874449685888125, "grad_norm": 2.390767812728882, "learning_rate": 4.250410877359765e-05, "loss": 2.7288, "step": 3196 }, { "epoch": 0.0287534907565217, "grad_norm": 1.9465256929397583, "learning_rate": 4.249874185900803e-05, "loss": 2.6338, "step": 3197 }, { "epoch": 0.02876248465416215, "grad_norm": 1.866338849067688, "learning_rate": 4.2493373362900954e-05, "loss": 2.6736, "step": 3198 }, { "epoch": 0.028771478551802603, "grad_norm": 1.1112959384918213, "learning_rate": 4.248800328576165e-05, "loss": 2.4035, "step": 3199 }, { "epoch": 0.028780472449443054, "grad_norm": 1.5329484939575195, "learning_rate": 4.2482631628075435e-05, "loss": 2.6123, "step": 3200 }, { "epoch": 0.028789466347083504, "grad_norm": 1.3881475925445557, "learning_rate": 4.247725839032781e-05, "loss": 2.431, "step": 3201 }, { "epoch": 0.028798460244723954, "grad_norm": 2.378202199935913, "learning_rate": 4.247188357300439e-05, "loss": 3.0914, "step": 3202 }, { "epoch": 0.028807454142364405, "grad_norm": 1.3993823528289795, "learning_rate": 4.246650717659096e-05, "loss": 2.312, "step": 3203 }, { "epoch": 0.028816448040004855, "grad_norm": 1.8119425773620605, "learning_rate": 4.246112920157342e-05, "loss": 2.5053, "step": 3204 }, { "epoch": 0.02882544193764531, "grad_norm": 1.6248605251312256, "learning_rate": 4.2455749648437834e-05, "loss": 2.4053, "step": 3205 }, { "epoch": 0.02883443583528576, "grad_norm": 1.926376461982727, "learning_rate": 4.2450368517670395e-05, "loss": 2.5846, "step": 3206 }, { "epoch": 0.02884342973292621, "grad_norm": 2.732304573059082, "learning_rate": 4.244498580975744e-05, "loss": 2.6683, "step": 3207 }, { "epoch": 0.02885242363056666, "grad_norm": 2.5231528282165527, "learning_rate": 4.243960152518547e-05, "loss": 2.8991, "step": 3208 }, { "epoch": 0.02886141752820711, "grad_norm": 2.0483527183532715, "learning_rate": 4.2434215664441086e-05, "loss": 2.5622, "step": 3209 }, { "epoch": 0.028870411425847564, "grad_norm": 2.0705831050872803, "learning_rate": 4.242882822801108e-05, "loss": 2.5068, "step": 3210 }, { "epoch": 0.028879405323488015, "grad_norm": 1.5434489250183105, "learning_rate": 4.242343921638234e-05, "loss": 2.5495, "step": 3211 }, { "epoch": 0.028888399221128465, "grad_norm": 1.5506601333618164, "learning_rate": 4.241804863004194e-05, "loss": 2.6403, "step": 3212 }, { "epoch": 0.028897393118768915, "grad_norm": 1.8174827098846436, "learning_rate": 4.241265646947706e-05, "loss": 2.4508, "step": 3213 }, { "epoch": 0.028906387016409366, "grad_norm": 2.242428779602051, "learning_rate": 4.240726273517505e-05, "loss": 2.7667, "step": 3214 }, { "epoch": 0.028915380914049816, "grad_norm": 1.613316297531128, "learning_rate": 4.240186742762337e-05, "loss": 2.7172, "step": 3215 }, { "epoch": 0.02892437481169027, "grad_norm": 1.6559053659439087, "learning_rate": 4.2396470547309666e-05, "loss": 2.6727, "step": 3216 }, { "epoch": 0.02893336870933072, "grad_norm": 2.246370792388916, "learning_rate": 4.2391072094721694e-05, "loss": 3.399, "step": 3217 }, { "epoch": 0.02894236260697117, "grad_norm": 1.4741957187652588, "learning_rate": 4.238567207034735e-05, "loss": 2.4987, "step": 3218 }, { "epoch": 0.02895135650461162, "grad_norm": 1.8105964660644531, "learning_rate": 4.23802704746747e-05, "loss": 2.715, "step": 3219 }, { "epoch": 0.02896035040225207, "grad_norm": 1.6966323852539062, "learning_rate": 4.237486730819191e-05, "loss": 2.6311, "step": 3220 }, { "epoch": 0.02896934429989252, "grad_norm": 2.000488758087158, "learning_rate": 4.236946257138734e-05, "loss": 2.7612, "step": 3221 }, { "epoch": 0.028978338197532975, "grad_norm": 1.4468376636505127, "learning_rate": 4.2364056264749454e-05, "loss": 2.6344, "step": 3222 }, { "epoch": 0.028987332095173426, "grad_norm": 1.6795893907546997, "learning_rate": 4.235864838876686e-05, "loss": 2.5052, "step": 3223 }, { "epoch": 0.028996325992813876, "grad_norm": 1.6664032936096191, "learning_rate": 4.235323894392832e-05, "loss": 2.4581, "step": 3224 }, { "epoch": 0.029005319890454326, "grad_norm": 1.7170703411102295, "learning_rate": 4.2347827930722745e-05, "loss": 2.6503, "step": 3225 }, { "epoch": 0.029014313788094777, "grad_norm": 1.5472594499588013, "learning_rate": 4.234241534963916e-05, "loss": 2.5205, "step": 3226 }, { "epoch": 0.029023307685735227, "grad_norm": 1.2882448434829712, "learning_rate": 4.2337001201166755e-05, "loss": 2.3607, "step": 3227 }, { "epoch": 0.02903230158337568, "grad_norm": 1.7825978994369507, "learning_rate": 4.2331585485794854e-05, "loss": 2.9767, "step": 3228 }, { "epoch": 0.02904129548101613, "grad_norm": 1.5966273546218872, "learning_rate": 4.232616820401293e-05, "loss": 2.5371, "step": 3229 }, { "epoch": 0.029050289378656582, "grad_norm": 0.8911131620407104, "learning_rate": 4.2320749356310586e-05, "loss": 2.2281, "step": 3230 }, { "epoch": 0.029059283276297032, "grad_norm": 2.448517084121704, "learning_rate": 4.231532894317757e-05, "loss": 3.1325, "step": 3231 }, { "epoch": 0.029068277173937482, "grad_norm": 1.8377784490585327, "learning_rate": 4.2309906965103766e-05, "loss": 2.7684, "step": 3232 }, { "epoch": 0.029077271071577933, "grad_norm": 1.2117584943771362, "learning_rate": 4.230448342257922e-05, "loss": 2.417, "step": 3233 }, { "epoch": 0.029086264969218387, "grad_norm": 1.8754435777664185, "learning_rate": 4.229905831609411e-05, "loss": 2.7622, "step": 3234 }, { "epoch": 0.029095258866858837, "grad_norm": 1.3211241960525513, "learning_rate": 4.229363164613874e-05, "loss": 2.6762, "step": 3235 }, { "epoch": 0.029104252764499287, "grad_norm": 1.9874247312545776, "learning_rate": 4.2288203413203564e-05, "loss": 2.893, "step": 3236 }, { "epoch": 0.029113246662139738, "grad_norm": 1.6713495254516602, "learning_rate": 4.228277361777918e-05, "loss": 2.7332, "step": 3237 }, { "epoch": 0.029122240559780188, "grad_norm": 1.6191916465759277, "learning_rate": 4.2277342260356336e-05, "loss": 2.9543, "step": 3238 }, { "epoch": 0.02913123445742064, "grad_norm": 2.6305012702941895, "learning_rate": 4.227190934142591e-05, "loss": 2.3787, "step": 3239 }, { "epoch": 0.029140228355061092, "grad_norm": 0.9331474304199219, "learning_rate": 4.226647486147892e-05, "loss": 2.5366, "step": 3240 }, { "epoch": 0.029149222252701543, "grad_norm": 1.760728120803833, "learning_rate": 4.2261038821006537e-05, "loss": 2.6324, "step": 3241 }, { "epoch": 0.029158216150341993, "grad_norm": 1.4972813129425049, "learning_rate": 4.225560122050005e-05, "loss": 2.4743, "step": 3242 }, { "epoch": 0.029167210047982443, "grad_norm": 1.3395090103149414, "learning_rate": 4.225016206045091e-05, "loss": 2.385, "step": 3243 }, { "epoch": 0.029176203945622894, "grad_norm": 1.7381865978240967, "learning_rate": 4.224472134135071e-05, "loss": 2.3299, "step": 3244 }, { "epoch": 0.029185197843263344, "grad_norm": 2.068230628967285, "learning_rate": 4.223927906369117e-05, "loss": 2.7988, "step": 3245 }, { "epoch": 0.029194191740903798, "grad_norm": 2.257577657699585, "learning_rate": 4.223383522796415e-05, "loss": 2.66, "step": 3246 }, { "epoch": 0.02920318563854425, "grad_norm": 1.5361354351043701, "learning_rate": 4.222838983466167e-05, "loss": 2.5582, "step": 3247 }, { "epoch": 0.0292121795361847, "grad_norm": 2.198725700378418, "learning_rate": 4.222294288427587e-05, "loss": 2.8861, "step": 3248 }, { "epoch": 0.02922117343382515, "grad_norm": 1.4787107706069946, "learning_rate": 4.2217494377299046e-05, "loss": 2.3817, "step": 3249 }, { "epoch": 0.0292301673314656, "grad_norm": 1.5107005834579468, "learning_rate": 4.221204431422362e-05, "loss": 2.5192, "step": 3250 }, { "epoch": 0.029239161229106053, "grad_norm": 2.053328514099121, "learning_rate": 4.220659269554217e-05, "loss": 2.765, "step": 3251 }, { "epoch": 0.029248155126746504, "grad_norm": 2.016929864883423, "learning_rate": 4.2201139521747404e-05, "loss": 2.9479, "step": 3252 }, { "epoch": 0.029257149024386954, "grad_norm": 1.3982702493667603, "learning_rate": 4.2195684793332186e-05, "loss": 2.4806, "step": 3253 }, { "epoch": 0.029266142922027404, "grad_norm": 1.6902679204940796, "learning_rate": 4.219022851078949e-05, "loss": 2.6731, "step": 3254 }, { "epoch": 0.029275136819667855, "grad_norm": 1.8485438823699951, "learning_rate": 4.218477067461245e-05, "loss": 2.6969, "step": 3255 }, { "epoch": 0.029284130717308305, "grad_norm": 1.9012726545333862, "learning_rate": 4.217931128529436e-05, "loss": 2.6516, "step": 3256 }, { "epoch": 0.02929312461494876, "grad_norm": 1.122360110282898, "learning_rate": 4.217385034332861e-05, "loss": 2.5806, "step": 3257 }, { "epoch": 0.02930211851258921, "grad_norm": 1.8455142974853516, "learning_rate": 4.216838784920877e-05, "loss": 2.8081, "step": 3258 }, { "epoch": 0.02931111241022966, "grad_norm": 1.8813096284866333, "learning_rate": 4.216292380342853e-05, "loss": 2.5427, "step": 3259 }, { "epoch": 0.02932010630787011, "grad_norm": 1.201625108718872, "learning_rate": 4.215745820648172e-05, "loss": 2.5825, "step": 3260 }, { "epoch": 0.02932910020551056, "grad_norm": 1.4453039169311523, "learning_rate": 4.215199105886232e-05, "loss": 2.4669, "step": 3261 }, { "epoch": 0.02933809410315101, "grad_norm": 1.3902549743652344, "learning_rate": 4.214652236106443e-05, "loss": 2.8198, "step": 3262 }, { "epoch": 0.029347088000791464, "grad_norm": 1.605765461921692, "learning_rate": 4.214105211358234e-05, "loss": 2.4493, "step": 3263 }, { "epoch": 0.029356081898431915, "grad_norm": 1.795940637588501, "learning_rate": 4.213558031691041e-05, "loss": 2.6149, "step": 3264 }, { "epoch": 0.029365075796072365, "grad_norm": 2.3953378200531006, "learning_rate": 4.213010697154319e-05, "loss": 2.9536, "step": 3265 }, { "epoch": 0.029374069693712816, "grad_norm": 1.6743098497390747, "learning_rate": 4.212463207797535e-05, "loss": 2.7996, "step": 3266 }, { "epoch": 0.029383063591353266, "grad_norm": 2.18847393989563, "learning_rate": 4.211915563670171e-05, "loss": 2.6463, "step": 3267 }, { "epoch": 0.029392057488993716, "grad_norm": 1.7241605520248413, "learning_rate": 4.211367764821722e-05, "loss": 2.8344, "step": 3268 }, { "epoch": 0.02940105138663417, "grad_norm": 1.193372130393982, "learning_rate": 4.210819811301698e-05, "loss": 2.4941, "step": 3269 }, { "epoch": 0.02941004528427462, "grad_norm": 1.720690131187439, "learning_rate": 4.210271703159622e-05, "loss": 2.7105, "step": 3270 }, { "epoch": 0.02941903918191507, "grad_norm": 1.5158485174179077, "learning_rate": 4.209723440445032e-05, "loss": 2.4802, "step": 3271 }, { "epoch": 0.02942803307955552, "grad_norm": 1.6816502809524536, "learning_rate": 4.209175023207479e-05, "loss": 2.8874, "step": 3272 }, { "epoch": 0.02943702697719597, "grad_norm": 2.0808212757110596, "learning_rate": 4.208626451496528e-05, "loss": 2.8933, "step": 3273 }, { "epoch": 0.029446020874836422, "grad_norm": 2.326096534729004, "learning_rate": 4.2080777253617584e-05, "loss": 2.5204, "step": 3274 }, { "epoch": 0.029455014772476876, "grad_norm": 1.2439253330230713, "learning_rate": 4.207528844852764e-05, "loss": 2.3031, "step": 3275 }, { "epoch": 0.029464008670117326, "grad_norm": 1.4350587129592896, "learning_rate": 4.2069798100191525e-05, "loss": 2.643, "step": 3276 }, { "epoch": 0.029473002567757776, "grad_norm": 1.4610927104949951, "learning_rate": 4.2064306209105444e-05, "loss": 2.7064, "step": 3277 }, { "epoch": 0.029481996465398227, "grad_norm": 2.155223846435547, "learning_rate": 4.205881277576574e-05, "loss": 2.726, "step": 3278 }, { "epoch": 0.029490990363038677, "grad_norm": 1.868330717086792, "learning_rate": 4.205331780066892e-05, "loss": 2.6359, "step": 3279 }, { "epoch": 0.029499984260679128, "grad_norm": 1.6553901433944702, "learning_rate": 4.2047821284311594e-05, "loss": 2.736, "step": 3280 }, { "epoch": 0.02950897815831958, "grad_norm": 1.3665984869003296, "learning_rate": 4.204232322719055e-05, "loss": 2.3049, "step": 3281 }, { "epoch": 0.02951797205596003, "grad_norm": 1.627935767173767, "learning_rate": 4.203682362980269e-05, "loss": 2.7761, "step": 3282 }, { "epoch": 0.029526965953600482, "grad_norm": 1.703913927078247, "learning_rate": 4.203132249264507e-05, "loss": 2.6342, "step": 3283 }, { "epoch": 0.029535959851240932, "grad_norm": 1.2964797019958496, "learning_rate": 4.202581981621485e-05, "loss": 2.5291, "step": 3284 }, { "epoch": 0.029544953748881383, "grad_norm": 1.6619776487350464, "learning_rate": 4.202031560100938e-05, "loss": 2.6128, "step": 3285 }, { "epoch": 0.029553947646521837, "grad_norm": 2.1519932746887207, "learning_rate": 4.201480984752612e-05, "loss": 3.1258, "step": 3286 }, { "epoch": 0.029562941544162287, "grad_norm": 1.3609532117843628, "learning_rate": 4.200930255626267e-05, "loss": 2.4268, "step": 3287 }, { "epoch": 0.029571935441802737, "grad_norm": 2.0980749130249023, "learning_rate": 4.2003793727716776e-05, "loss": 2.9522, "step": 3288 }, { "epoch": 0.029580929339443188, "grad_norm": 1.1343724727630615, "learning_rate": 4.1998283362386324e-05, "loss": 2.4151, "step": 3289 }, { "epoch": 0.029589923237083638, "grad_norm": 1.675140619277954, "learning_rate": 4.199277146076933e-05, "loss": 2.8902, "step": 3290 }, { "epoch": 0.02959891713472409, "grad_norm": 1.4978352785110474, "learning_rate": 4.198725802336395e-05, "loss": 2.6583, "step": 3291 }, { "epoch": 0.029607911032364542, "grad_norm": 1.5554269552230835, "learning_rate": 4.198174305066849e-05, "loss": 2.674, "step": 3292 }, { "epoch": 0.029616904930004993, "grad_norm": 1.489122748374939, "learning_rate": 4.197622654318138e-05, "loss": 2.6959, "step": 3293 }, { "epoch": 0.029625898827645443, "grad_norm": 1.2885017395019531, "learning_rate": 4.197070850140121e-05, "loss": 2.4992, "step": 3294 }, { "epoch": 0.029634892725285893, "grad_norm": 2.560407876968384, "learning_rate": 4.196518892582668e-05, "loss": 2.9846, "step": 3295 }, { "epoch": 0.029643886622926344, "grad_norm": 2.4655470848083496, "learning_rate": 4.195966781695665e-05, "loss": 3.1773, "step": 3296 }, { "epoch": 0.029652880520566794, "grad_norm": 1.3560019731521606, "learning_rate": 4.19541451752901e-05, "loss": 2.2974, "step": 3297 }, { "epoch": 0.029661874418207248, "grad_norm": 1.349364995956421, "learning_rate": 4.194862100132619e-05, "loss": 2.4187, "step": 3298 }, { "epoch": 0.029670868315847698, "grad_norm": 0.9441218972206116, "learning_rate": 4.194309529556415e-05, "loss": 2.2283, "step": 3299 }, { "epoch": 0.02967986221348815, "grad_norm": 1.5616647005081177, "learning_rate": 4.193756805850342e-05, "loss": 2.5831, "step": 3300 }, { "epoch": 0.0296888561111286, "grad_norm": 2.0731680393218994, "learning_rate": 4.193203929064353e-05, "loss": 2.555, "step": 3301 }, { "epoch": 0.02969785000876905, "grad_norm": 1.5503902435302734, "learning_rate": 4.1926508992484165e-05, "loss": 2.5863, "step": 3302 }, { "epoch": 0.0297068439064095, "grad_norm": 1.8494412899017334, "learning_rate": 4.192097716452514e-05, "loss": 2.9367, "step": 3303 }, { "epoch": 0.029715837804049954, "grad_norm": 1.490099310874939, "learning_rate": 4.1915443807266444e-05, "loss": 2.5875, "step": 3304 }, { "epoch": 0.029724831701690404, "grad_norm": 1.962157964706421, "learning_rate": 4.190990892120814e-05, "loss": 2.8404, "step": 3305 }, { "epoch": 0.029733825599330854, "grad_norm": 2.141503095626831, "learning_rate": 4.1904372506850484e-05, "loss": 3.0512, "step": 3306 }, { "epoch": 0.029742819496971305, "grad_norm": 1.869417428970337, "learning_rate": 4.189883456469386e-05, "loss": 2.7167, "step": 3307 }, { "epoch": 0.029751813394611755, "grad_norm": 2.0307211875915527, "learning_rate": 4.189329509523876e-05, "loss": 2.531, "step": 3308 }, { "epoch": 0.029760807292252205, "grad_norm": 2.1704916954040527, "learning_rate": 4.1887754098985846e-05, "loss": 2.8415, "step": 3309 }, { "epoch": 0.02976980118989266, "grad_norm": 1.736915111541748, "learning_rate": 4.1882211576435906e-05, "loss": 2.6702, "step": 3310 }, { "epoch": 0.02977879508753311, "grad_norm": 1.813555359840393, "learning_rate": 4.187666752808987e-05, "loss": 2.5396, "step": 3311 }, { "epoch": 0.02978778898517356, "grad_norm": 1.2786608934402466, "learning_rate": 4.18711219544488e-05, "loss": 2.5725, "step": 3312 }, { "epoch": 0.02979678288281401, "grad_norm": 2.590571165084839, "learning_rate": 4.18655748560139e-05, "loss": 2.7994, "step": 3313 }, { "epoch": 0.02980577678045446, "grad_norm": 1.6217520236968994, "learning_rate": 4.186002623328651e-05, "loss": 2.3895, "step": 3314 }, { "epoch": 0.02981477067809491, "grad_norm": 1.7878055572509766, "learning_rate": 4.1854476086768115e-05, "loss": 2.7066, "step": 3315 }, { "epoch": 0.029823764575735365, "grad_norm": 1.9915719032287598, "learning_rate": 4.1848924416960314e-05, "loss": 2.5151, "step": 3316 }, { "epoch": 0.029832758473375815, "grad_norm": 1.3386199474334717, "learning_rate": 4.184337122436488e-05, "loss": 2.3277, "step": 3317 }, { "epoch": 0.029841752371016265, "grad_norm": 2.1768219470977783, "learning_rate": 4.1837816509483696e-05, "loss": 3.2877, "step": 3318 }, { "epoch": 0.029850746268656716, "grad_norm": 1.6496798992156982, "learning_rate": 4.183226027281879e-05, "loss": 2.7192, "step": 3319 }, { "epoch": 0.029859740166297166, "grad_norm": 2.1622986793518066, "learning_rate": 4.182670251487233e-05, "loss": 3.092, "step": 3320 }, { "epoch": 0.02986873406393762, "grad_norm": 2.205345869064331, "learning_rate": 4.182114323614662e-05, "loss": 2.8845, "step": 3321 }, { "epoch": 0.02987772796157807, "grad_norm": 2.4858412742614746, "learning_rate": 4.1815582437144106e-05, "loss": 2.7079, "step": 3322 }, { "epoch": 0.02988672185921852, "grad_norm": 1.6224535703659058, "learning_rate": 4.181002011836737e-05, "loss": 2.6017, "step": 3323 }, { "epoch": 0.02989571575685897, "grad_norm": 1.8362705707550049, "learning_rate": 4.180445628031911e-05, "loss": 3.0393, "step": 3324 }, { "epoch": 0.02990470965449942, "grad_norm": 2.492368221282959, "learning_rate": 4.17988909235022e-05, "loss": 3.1368, "step": 3325 }, { "epoch": 0.029913703552139872, "grad_norm": 2.4074878692626953, "learning_rate": 4.1793324048419626e-05, "loss": 2.3552, "step": 3326 }, { "epoch": 0.029922697449780326, "grad_norm": 2.3631813526153564, "learning_rate": 4.17877556555745e-05, "loss": 2.8693, "step": 3327 }, { "epoch": 0.029931691347420776, "grad_norm": 1.311640739440918, "learning_rate": 4.1782185745470106e-05, "loss": 2.2689, "step": 3328 }, { "epoch": 0.029940685245061226, "grad_norm": 1.8539131879806519, "learning_rate": 4.1776614318609845e-05, "loss": 2.7497, "step": 3329 }, { "epoch": 0.029949679142701677, "grad_norm": 1.2482659816741943, "learning_rate": 4.177104137549724e-05, "loss": 2.3946, "step": 3330 }, { "epoch": 0.029958673040342127, "grad_norm": 2.1096694469451904, "learning_rate": 4.1765466916635995e-05, "loss": 2.6731, "step": 3331 }, { "epoch": 0.029967666937982577, "grad_norm": 2.0338621139526367, "learning_rate": 4.17598909425299e-05, "loss": 2.8158, "step": 3332 }, { "epoch": 0.02997666083562303, "grad_norm": 1.609251856803894, "learning_rate": 4.175431345368291e-05, "loss": 2.4903, "step": 3333 }, { "epoch": 0.02998565473326348, "grad_norm": 1.5098083019256592, "learning_rate": 4.174873445059913e-05, "loss": 2.5547, "step": 3334 }, { "epoch": 0.029994648630903932, "grad_norm": 1.0906245708465576, "learning_rate": 4.174315393378276e-05, "loss": 2.345, "step": 3335 }, { "epoch": 0.030003642528544382, "grad_norm": 1.4871244430541992, "learning_rate": 4.173757190373817e-05, "loss": 2.8292, "step": 3336 }, { "epoch": 0.030012636426184833, "grad_norm": 1.6301637887954712, "learning_rate": 4.173198836096987e-05, "loss": 2.5211, "step": 3337 }, { "epoch": 0.030021630323825283, "grad_norm": 1.815865159034729, "learning_rate": 4.1726403305982476e-05, "loss": 2.6535, "step": 3338 }, { "epoch": 0.030030624221465737, "grad_norm": 1.7673953771591187, "learning_rate": 4.172081673928077e-05, "loss": 2.6829, "step": 3339 }, { "epoch": 0.030039618119106187, "grad_norm": 2.2503175735473633, "learning_rate": 4.1715228661369654e-05, "loss": 2.6606, "step": 3340 }, { "epoch": 0.030048612016746638, "grad_norm": 1.9588162899017334, "learning_rate": 4.170963907275418e-05, "loss": 2.6899, "step": 3341 }, { "epoch": 0.030057605914387088, "grad_norm": 2.0876410007476807, "learning_rate": 4.170404797393952e-05, "loss": 3.0206, "step": 3342 }, { "epoch": 0.03006659981202754, "grad_norm": 1.7342253923416138, "learning_rate": 4.1698455365431e-05, "loss": 2.7104, "step": 3343 }, { "epoch": 0.03007559370966799, "grad_norm": 2.1743319034576416, "learning_rate": 4.1692861247734066e-05, "loss": 2.8804, "step": 3344 }, { "epoch": 0.030084587607308443, "grad_norm": 1.967452049255371, "learning_rate": 4.1687265621354314e-05, "loss": 2.5481, "step": 3345 }, { "epoch": 0.030093581504948893, "grad_norm": 1.9557037353515625, "learning_rate": 4.1681668486797475e-05, "loss": 2.9304, "step": 3346 }, { "epoch": 0.030102575402589343, "grad_norm": 1.7088018655776978, "learning_rate": 4.1676069844569404e-05, "loss": 2.4527, "step": 3347 }, { "epoch": 0.030111569300229794, "grad_norm": 1.9367979764938354, "learning_rate": 4.1670469695176097e-05, "loss": 2.73, "step": 3348 }, { "epoch": 0.030120563197870244, "grad_norm": 2.0835793018341064, "learning_rate": 4.1664868039123694e-05, "loss": 2.7046, "step": 3349 }, { "epoch": 0.030129557095510694, "grad_norm": 1.492823839187622, "learning_rate": 4.165926487691848e-05, "loss": 2.7355, "step": 3350 }, { "epoch": 0.030138550993151148, "grad_norm": 1.7221592664718628, "learning_rate": 4.1653660209066835e-05, "loss": 2.6852, "step": 3351 }, { "epoch": 0.0301475448907916, "grad_norm": 1.7473608255386353, "learning_rate": 4.164805403607532e-05, "loss": 2.8036, "step": 3352 }, { "epoch": 0.03015653878843205, "grad_norm": 1.8439804315567017, "learning_rate": 4.164244635845063e-05, "loss": 2.7067, "step": 3353 }, { "epoch": 0.0301655326860725, "grad_norm": 1.849801778793335, "learning_rate": 4.163683717669955e-05, "loss": 2.4718, "step": 3354 }, { "epoch": 0.03017452658371295, "grad_norm": 1.4369633197784424, "learning_rate": 4.1631226491329044e-05, "loss": 2.7084, "step": 3355 }, { "epoch": 0.0301835204813534, "grad_norm": 2.3263518810272217, "learning_rate": 4.16256143028462e-05, "loss": 2.8929, "step": 3356 }, { "epoch": 0.030192514378993854, "grad_norm": 1.0721911191940308, "learning_rate": 4.162000061175826e-05, "loss": 2.2188, "step": 3357 }, { "epoch": 0.030201508276634304, "grad_norm": 1.4041659832000732, "learning_rate": 4.161438541857255e-05, "loss": 2.444, "step": 3358 }, { "epoch": 0.030210502174274755, "grad_norm": 0.9677774906158447, "learning_rate": 4.1608768723796585e-05, "loss": 2.3541, "step": 3359 }, { "epoch": 0.030219496071915205, "grad_norm": 1.8552231788635254, "learning_rate": 4.1603150527937996e-05, "loss": 2.6203, "step": 3360 }, { "epoch": 0.030228489969555655, "grad_norm": 1.82222580909729, "learning_rate": 4.159753083150454e-05, "loss": 2.6765, "step": 3361 }, { "epoch": 0.03023748386719611, "grad_norm": 1.7040854692459106, "learning_rate": 4.1591909635004136e-05, "loss": 2.8338, "step": 3362 }, { "epoch": 0.03024647776483656, "grad_norm": 2.1298177242279053, "learning_rate": 4.15862869389448e-05, "loss": 2.9086, "step": 3363 }, { "epoch": 0.03025547166247701, "grad_norm": 1.8585193157196045, "learning_rate": 4.158066274383471e-05, "loss": 2.6013, "step": 3364 }, { "epoch": 0.03026446556011746, "grad_norm": 2.540275812149048, "learning_rate": 4.15750370501822e-05, "loss": 2.8693, "step": 3365 }, { "epoch": 0.03027345945775791, "grad_norm": 1.7587987184524536, "learning_rate": 4.1569409858495675e-05, "loss": 2.9522, "step": 3366 }, { "epoch": 0.03028245335539836, "grad_norm": 1.1225632429122925, "learning_rate": 4.156378116928375e-05, "loss": 2.3912, "step": 3367 }, { "epoch": 0.030291447253038815, "grad_norm": 1.9423259496688843, "learning_rate": 4.155815098305511e-05, "loss": 2.5091, "step": 3368 }, { "epoch": 0.030300441150679265, "grad_norm": 1.577430009841919, "learning_rate": 4.1552519300318626e-05, "loss": 2.5518, "step": 3369 }, { "epoch": 0.030309435048319715, "grad_norm": 1.4187660217285156, "learning_rate": 4.1546886121583274e-05, "loss": 2.7032, "step": 3370 }, { "epoch": 0.030318428945960166, "grad_norm": 1.5907630920410156, "learning_rate": 4.154125144735818e-05, "loss": 2.53, "step": 3371 }, { "epoch": 0.030327422843600616, "grad_norm": 1.9868648052215576, "learning_rate": 4.153561527815259e-05, "loss": 2.8392, "step": 3372 }, { "epoch": 0.030336416741241067, "grad_norm": 1.945101022720337, "learning_rate": 4.152997761447591e-05, "loss": 2.9582, "step": 3373 }, { "epoch": 0.03034541063888152, "grad_norm": 2.755678176879883, "learning_rate": 4.152433845683766e-05, "loss": 2.9925, "step": 3374 }, { "epoch": 0.03035440453652197, "grad_norm": 1.6906394958496094, "learning_rate": 4.151869780574749e-05, "loss": 2.599, "step": 3375 }, { "epoch": 0.03036339843416242, "grad_norm": 1.9711891412734985, "learning_rate": 4.1513055661715214e-05, "loss": 2.7005, "step": 3376 }, { "epoch": 0.03037239233180287, "grad_norm": 1.442321538925171, "learning_rate": 4.1507412025250745e-05, "loss": 2.451, "step": 3377 }, { "epoch": 0.030381386229443322, "grad_norm": 1.8602179288864136, "learning_rate": 4.150176689686417e-05, "loss": 2.5068, "step": 3378 }, { "epoch": 0.030390380127083772, "grad_norm": 1.7945653200149536, "learning_rate": 4.149612027706566e-05, "loss": 2.4345, "step": 3379 }, { "epoch": 0.030399374024724226, "grad_norm": 1.3439702987670898, "learning_rate": 4.149047216636558e-05, "loss": 2.512, "step": 3380 }, { "epoch": 0.030408367922364676, "grad_norm": 1.5020204782485962, "learning_rate": 4.148482256527438e-05, "loss": 2.6476, "step": 3381 }, { "epoch": 0.030417361820005127, "grad_norm": 1.37667715549469, "learning_rate": 4.147917147430267e-05, "loss": 2.3348, "step": 3382 }, { "epoch": 0.030426355717645577, "grad_norm": 1.3744052648544312, "learning_rate": 4.1473518893961206e-05, "loss": 2.7139, "step": 3383 }, { "epoch": 0.030435349615286027, "grad_norm": 2.120690107345581, "learning_rate": 4.146786482476084e-05, "loss": 3.0738, "step": 3384 }, { "epoch": 0.030444343512926478, "grad_norm": 1.953344702720642, "learning_rate": 4.146220926721259e-05, "loss": 2.8722, "step": 3385 }, { "epoch": 0.03045333741056693, "grad_norm": 1.4912606477737427, "learning_rate": 4.1456552221827596e-05, "loss": 2.4628, "step": 3386 }, { "epoch": 0.030462331308207382, "grad_norm": 1.322853684425354, "learning_rate": 4.145089368911714e-05, "loss": 2.3417, "step": 3387 }, { "epoch": 0.030471325205847832, "grad_norm": 1.8122005462646484, "learning_rate": 4.1445233669592634e-05, "loss": 2.8805, "step": 3388 }, { "epoch": 0.030480319103488283, "grad_norm": 1.1497490406036377, "learning_rate": 4.143957216376561e-05, "loss": 2.3248, "step": 3389 }, { "epoch": 0.030489313001128733, "grad_norm": 2.028402328491211, "learning_rate": 4.143390917214778e-05, "loss": 2.8319, "step": 3390 }, { "epoch": 0.030498306898769183, "grad_norm": 1.292473316192627, "learning_rate": 4.142824469525093e-05, "loss": 2.4919, "step": 3391 }, { "epoch": 0.030507300796409637, "grad_norm": 1.769448161125183, "learning_rate": 4.142257873358701e-05, "loss": 2.4771, "step": 3392 }, { "epoch": 0.030516294694050088, "grad_norm": 1.2355117797851562, "learning_rate": 4.141691128766813e-05, "loss": 2.392, "step": 3393 }, { "epoch": 0.030525288591690538, "grad_norm": 1.804715633392334, "learning_rate": 4.141124235800649e-05, "loss": 3.0569, "step": 3394 }, { "epoch": 0.03053428248933099, "grad_norm": 1.7024048566818237, "learning_rate": 4.140557194511443e-05, "loss": 2.9215, "step": 3395 }, { "epoch": 0.03054327638697144, "grad_norm": 1.8269582986831665, "learning_rate": 4.139990004950446e-05, "loss": 2.6109, "step": 3396 }, { "epoch": 0.030552270284611893, "grad_norm": 1.944586157798767, "learning_rate": 4.1394226671689174e-05, "loss": 2.8459, "step": 3397 }, { "epoch": 0.030561264182252343, "grad_norm": 2.3425145149230957, "learning_rate": 4.1388551812181355e-05, "loss": 2.353, "step": 3398 }, { "epoch": 0.030570258079892793, "grad_norm": 2.1298961639404297, "learning_rate": 4.138287547149387e-05, "loss": 2.5982, "step": 3399 }, { "epoch": 0.030579251977533244, "grad_norm": 2.354739189147949, "learning_rate": 4.137719765013974e-05, "loss": 2.7469, "step": 3400 }, { "epoch": 0.030588245875173694, "grad_norm": 1.9223195314407349, "learning_rate": 4.137151834863213e-05, "loss": 2.7908, "step": 3401 }, { "epoch": 0.030597239772814144, "grad_norm": 1.799376368522644, "learning_rate": 4.136583756748433e-05, "loss": 2.6189, "step": 3402 }, { "epoch": 0.030606233670454598, "grad_norm": 1.208121657371521, "learning_rate": 4.136015530720975e-05, "loss": 2.3416, "step": 3403 }, { "epoch": 0.03061522756809505, "grad_norm": 2.05855655670166, "learning_rate": 4.135447156832196e-05, "loss": 2.7653, "step": 3404 }, { "epoch": 0.0306242214657355, "grad_norm": 1.6012519598007202, "learning_rate": 4.1348786351334645e-05, "loss": 2.3059, "step": 3405 }, { "epoch": 0.03063321536337595, "grad_norm": 1.7506383657455444, "learning_rate": 4.134309965676163e-05, "loss": 2.6803, "step": 3406 }, { "epoch": 0.0306422092610164, "grad_norm": 1.551652431488037, "learning_rate": 4.1337411485116875e-05, "loss": 2.7992, "step": 3407 }, { "epoch": 0.03065120315865685, "grad_norm": 1.6702601909637451, "learning_rate": 4.133172183691446e-05, "loss": 2.7935, "step": 3408 }, { "epoch": 0.030660197056297304, "grad_norm": 1.5679682493209839, "learning_rate": 4.132603071266862e-05, "loss": 2.8631, "step": 3409 }, { "epoch": 0.030669190953937754, "grad_norm": 1.9372650384902954, "learning_rate": 4.1320338112893705e-05, "loss": 2.7535, "step": 3410 }, { "epoch": 0.030678184851578204, "grad_norm": 2.1970160007476807, "learning_rate": 4.131464403810422e-05, "loss": 2.7999, "step": 3411 }, { "epoch": 0.030687178749218655, "grad_norm": 1.669127345085144, "learning_rate": 4.130894848881477e-05, "loss": 2.5203, "step": 3412 }, { "epoch": 0.030696172646859105, "grad_norm": 1.1519346237182617, "learning_rate": 4.130325146554013e-05, "loss": 2.4899, "step": 3413 }, { "epoch": 0.030705166544499556, "grad_norm": 1.8278120756149292, "learning_rate": 4.129755296879517e-05, "loss": 2.5686, "step": 3414 }, { "epoch": 0.03071416044214001, "grad_norm": 1.328052282333374, "learning_rate": 4.129185299909494e-05, "loss": 2.5573, "step": 3415 }, { "epoch": 0.03072315433978046, "grad_norm": 1.7503916025161743, "learning_rate": 4.128615155695458e-05, "loss": 2.5814, "step": 3416 }, { "epoch": 0.03073214823742091, "grad_norm": 1.403752326965332, "learning_rate": 4.128044864288938e-05, "loss": 2.6349, "step": 3417 }, { "epoch": 0.03074114213506136, "grad_norm": 1.3199284076690674, "learning_rate": 4.1274744257414776e-05, "loss": 2.5955, "step": 3418 }, { "epoch": 0.03075013603270181, "grad_norm": 2.0915558338165283, "learning_rate": 4.1269038401046314e-05, "loss": 2.8574, "step": 3419 }, { "epoch": 0.03075912993034226, "grad_norm": 1.8908058404922485, "learning_rate": 4.126333107429968e-05, "loss": 2.5446, "step": 3420 }, { "epoch": 0.030768123827982715, "grad_norm": 1.575299620628357, "learning_rate": 4.12576222776907e-05, "loss": 2.6538, "step": 3421 }, { "epoch": 0.030777117725623165, "grad_norm": 1.8053762912750244, "learning_rate": 4.1251912011735326e-05, "loss": 2.6544, "step": 3422 }, { "epoch": 0.030786111623263616, "grad_norm": 1.7617080211639404, "learning_rate": 4.1246200276949665e-05, "loss": 2.5586, "step": 3423 }, { "epoch": 0.030795105520904066, "grad_norm": 30.55015754699707, "learning_rate": 4.1240487073849906e-05, "loss": 2.4594, "step": 3424 }, { "epoch": 0.030804099418544516, "grad_norm": 2.149658441543579, "learning_rate": 4.1234772402952435e-05, "loss": 2.8973, "step": 3425 }, { "epoch": 0.030813093316184967, "grad_norm": 1.165056824684143, "learning_rate": 4.1229056264773705e-05, "loss": 2.2917, "step": 3426 }, { "epoch": 0.03082208721382542, "grad_norm": 0.9712805151939392, "learning_rate": 4.122333865983036e-05, "loss": 2.3733, "step": 3427 }, { "epoch": 0.03083108111146587, "grad_norm": 2.1642415523529053, "learning_rate": 4.121761958863913e-05, "loss": 2.4644, "step": 3428 }, { "epoch": 0.03084007500910632, "grad_norm": 1.9858040809631348, "learning_rate": 4.121189905171692e-05, "loss": 2.9691, "step": 3429 }, { "epoch": 0.030849068906746772, "grad_norm": 1.6129240989685059, "learning_rate": 4.1206177049580744e-05, "loss": 2.5952, "step": 3430 }, { "epoch": 0.030858062804387222, "grad_norm": 1.7416294813156128, "learning_rate": 4.120045358274772e-05, "loss": 2.733, "step": 3431 }, { "epoch": 0.030867056702027676, "grad_norm": 2.312134027481079, "learning_rate": 4.1194728651735166e-05, "loss": 2.7164, "step": 3432 }, { "epoch": 0.030876050599668126, "grad_norm": 2.586787462234497, "learning_rate": 4.118900225706047e-05, "loss": 2.7187, "step": 3433 }, { "epoch": 0.030885044497308577, "grad_norm": 1.7719311714172363, "learning_rate": 4.11832743992412e-05, "loss": 2.6564, "step": 3434 }, { "epoch": 0.030894038394949027, "grad_norm": 2.077239990234375, "learning_rate": 4.117754507879501e-05, "loss": 2.56, "step": 3435 }, { "epoch": 0.030903032292589477, "grad_norm": 2.227294683456421, "learning_rate": 4.117181429623973e-05, "loss": 2.623, "step": 3436 }, { "epoch": 0.030912026190229928, "grad_norm": 2.6154046058654785, "learning_rate": 4.116608205209329e-05, "loss": 2.9939, "step": 3437 }, { "epoch": 0.03092102008787038, "grad_norm": 2.1306374073028564, "learning_rate": 4.116034834687376e-05, "loss": 2.4143, "step": 3438 }, { "epoch": 0.030930013985510832, "grad_norm": 1.6579190492630005, "learning_rate": 4.1154613181099354e-05, "loss": 2.9209, "step": 3439 }, { "epoch": 0.030939007883151282, "grad_norm": 1.8854484558105469, "learning_rate": 4.1148876555288416e-05, "loss": 2.7184, "step": 3440 }, { "epoch": 0.030948001780791733, "grad_norm": 1.557192087173462, "learning_rate": 4.1143138469959405e-05, "loss": 2.509, "step": 3441 }, { "epoch": 0.030956995678432183, "grad_norm": 0.9559489488601685, "learning_rate": 4.1137398925630935e-05, "loss": 2.377, "step": 3442 }, { "epoch": 0.030965989576072633, "grad_norm": 2.699507236480713, "learning_rate": 4.113165792282172e-05, "loss": 2.519, "step": 3443 }, { "epoch": 0.030974983473713087, "grad_norm": 1.8598675727844238, "learning_rate": 4.112591546205064e-05, "loss": 2.7318, "step": 3444 }, { "epoch": 0.030983977371353538, "grad_norm": 1.840263843536377, "learning_rate": 4.112017154383669e-05, "loss": 2.4547, "step": 3445 }, { "epoch": 0.030992971268993988, "grad_norm": 1.5272835493087769, "learning_rate": 4.1114426168699004e-05, "loss": 2.3579, "step": 3446 }, { "epoch": 0.03100196516663444, "grad_norm": 1.304485559463501, "learning_rate": 4.110867933715683e-05, "loss": 2.4517, "step": 3447 }, { "epoch": 0.03101095906427489, "grad_norm": 1.5052766799926758, "learning_rate": 4.1102931049729576e-05, "loss": 2.1765, "step": 3448 }, { "epoch": 0.03101995296191534, "grad_norm": 2.0382137298583984, "learning_rate": 4.109718130693676e-05, "loss": 2.6833, "step": 3449 }, { "epoch": 0.031028946859555793, "grad_norm": 2.3191866874694824, "learning_rate": 4.109143010929803e-05, "loss": 2.9831, "step": 3450 }, { "epoch": 0.031037940757196243, "grad_norm": 2.4846222400665283, "learning_rate": 4.108567745733318e-05, "loss": 2.4307, "step": 3451 }, { "epoch": 0.031046934654836694, "grad_norm": 1.7377508878707886, "learning_rate": 4.1079923351562125e-05, "loss": 2.7333, "step": 3452 }, { "epoch": 0.031055928552477144, "grad_norm": 1.6193459033966064, "learning_rate": 4.107416779250493e-05, "loss": 2.6744, "step": 3453 }, { "epoch": 0.031064922450117594, "grad_norm": 1.7543203830718994, "learning_rate": 4.106841078068175e-05, "loss": 2.7356, "step": 3454 }, { "epoch": 0.031073916347758045, "grad_norm": 1.059053659439087, "learning_rate": 4.106265231661292e-05, "loss": 2.5098, "step": 3455 }, { "epoch": 0.0310829102453985, "grad_norm": 2.606616258621216, "learning_rate": 4.1056892400818865e-05, "loss": 2.7764, "step": 3456 }, { "epoch": 0.03109190414303895, "grad_norm": 1.665492296218872, "learning_rate": 4.105113103382017e-05, "loss": 2.7263, "step": 3457 }, { "epoch": 0.0311008980406794, "grad_norm": 1.4017540216445923, "learning_rate": 4.104536821613755e-05, "loss": 2.401, "step": 3458 }, { "epoch": 0.03110989193831985, "grad_norm": 1.5780314207077026, "learning_rate": 4.103960394829183e-05, "loss": 2.6173, "step": 3459 }, { "epoch": 0.0311188858359603, "grad_norm": 1.835100531578064, "learning_rate": 4.103383823080398e-05, "loss": 2.5836, "step": 3460 }, { "epoch": 0.03112787973360075, "grad_norm": 1.877487301826477, "learning_rate": 4.1028071064195104e-05, "loss": 2.6385, "step": 3461 }, { "epoch": 0.031136873631241204, "grad_norm": 1.6154340505599976, "learning_rate": 4.102230244898643e-05, "loss": 2.604, "step": 3462 }, { "epoch": 0.031145867528881654, "grad_norm": 1.6481633186340332, "learning_rate": 4.101653238569931e-05, "loss": 2.4717, "step": 3463 }, { "epoch": 0.031154861426522105, "grad_norm": 1.5445773601531982, "learning_rate": 4.101076087485525e-05, "loss": 2.6888, "step": 3464 }, { "epoch": 0.031163855324162555, "grad_norm": 3.517937183380127, "learning_rate": 4.100498791697587e-05, "loss": 2.8667, "step": 3465 }, { "epoch": 0.031172849221803006, "grad_norm": 2.025331974029541, "learning_rate": 4.099921351258292e-05, "loss": 2.2594, "step": 3466 }, { "epoch": 0.031181843119443456, "grad_norm": 1.456588864326477, "learning_rate": 4.0993437662198284e-05, "loss": 2.4566, "step": 3467 }, { "epoch": 0.03119083701708391, "grad_norm": 1.8370065689086914, "learning_rate": 4.098766036634398e-05, "loss": 2.5281, "step": 3468 }, { "epoch": 0.03119983091472436, "grad_norm": 1.6744706630706787, "learning_rate": 4.098188162554215e-05, "loss": 2.5053, "step": 3469 }, { "epoch": 0.03120882481236481, "grad_norm": 1.7700225114822388, "learning_rate": 4.097610144031507e-05, "loss": 2.5348, "step": 3470 }, { "epoch": 0.03121781871000526, "grad_norm": 2.3166873455047607, "learning_rate": 4.097031981118515e-05, "loss": 2.6887, "step": 3471 }, { "epoch": 0.03122681260764571, "grad_norm": 1.9202619791030884, "learning_rate": 4.096453673867494e-05, "loss": 2.9097, "step": 3472 }, { "epoch": 0.031235806505286165, "grad_norm": 4.900930404663086, "learning_rate": 4.0958752223307084e-05, "loss": 2.6857, "step": 3473 }, { "epoch": 0.031244800402926615, "grad_norm": 1.7613332271575928, "learning_rate": 4.09529662656044e-05, "loss": 2.4869, "step": 3474 }, { "epoch": 0.031253794300567066, "grad_norm": 1.9844918251037598, "learning_rate": 4.09471788660898e-05, "loss": 2.742, "step": 3475 }, { "epoch": 0.03126278819820752, "grad_norm": 1.06586754322052, "learning_rate": 4.094139002528635e-05, "loss": 2.3659, "step": 3476 }, { "epoch": 0.031271782095847966, "grad_norm": 2.041654109954834, "learning_rate": 4.093559974371725e-05, "loss": 2.2376, "step": 3477 }, { "epoch": 0.03128077599348842, "grad_norm": 2.0764882564544678, "learning_rate": 4.09298080219058e-05, "loss": 2.6743, "step": 3478 }, { "epoch": 0.03128976989112887, "grad_norm": 1.8052029609680176, "learning_rate": 4.092401486037546e-05, "loss": 2.6677, "step": 3479 }, { "epoch": 0.03129876378876932, "grad_norm": 1.7542539834976196, "learning_rate": 4.091822025964981e-05, "loss": 2.5687, "step": 3480 }, { "epoch": 0.03130775768640977, "grad_norm": 0.9073329567909241, "learning_rate": 4.091242422025256e-05, "loss": 2.2343, "step": 3481 }, { "epoch": 0.03131675158405022, "grad_norm": 1.8987321853637695, "learning_rate": 4.090662674270755e-05, "loss": 2.975, "step": 3482 }, { "epoch": 0.031325745481690676, "grad_norm": 3.2611277103424072, "learning_rate": 4.090082782753875e-05, "loss": 2.9234, "step": 3483 }, { "epoch": 0.03133473937933112, "grad_norm": 2.131120204925537, "learning_rate": 4.0895027475270256e-05, "loss": 2.6537, "step": 3484 }, { "epoch": 0.031343733276971576, "grad_norm": 0.641789972782135, "learning_rate": 4.088922568642629e-05, "loss": 2.2268, "step": 3485 }, { "epoch": 0.03135272717461202, "grad_norm": 2.522510528564453, "learning_rate": 4.088342246153123e-05, "loss": 2.8768, "step": 3486 }, { "epoch": 0.03136172107225248, "grad_norm": 1.68440580368042, "learning_rate": 4.0877617801109555e-05, "loss": 2.7239, "step": 3487 }, { "epoch": 0.03137071496989293, "grad_norm": 1.6203149557113647, "learning_rate": 4.0871811705685883e-05, "loss": 2.6152, "step": 3488 }, { "epoch": 0.03137970886753338, "grad_norm": 1.756134271621704, "learning_rate": 4.0866004175784974e-05, "loss": 2.5835, "step": 3489 }, { "epoch": 0.03138870276517383, "grad_norm": 1.8467273712158203, "learning_rate": 4.086019521193169e-05, "loss": 2.9523, "step": 3490 }, { "epoch": 0.03139769666281428, "grad_norm": 2.914285182952881, "learning_rate": 4.085438481465104e-05, "loss": 2.7041, "step": 3491 }, { "epoch": 0.03140669056045473, "grad_norm": 1.7277684211730957, "learning_rate": 4.0848572984468174e-05, "loss": 2.5323, "step": 3492 }, { "epoch": 0.03141568445809518, "grad_norm": 1.2283086776733398, "learning_rate": 4.0842759721908355e-05, "loss": 2.271, "step": 3493 }, { "epoch": 0.03142467835573563, "grad_norm": 1.4307879209518433, "learning_rate": 4.083694502749697e-05, "loss": 2.6751, "step": 3494 }, { "epoch": 0.03143367225337609, "grad_norm": 1.7581567764282227, "learning_rate": 4.0831128901759556e-05, "loss": 2.5879, "step": 3495 }, { "epoch": 0.031442666151016534, "grad_norm": 1.9575295448303223, "learning_rate": 4.082531134522176e-05, "loss": 2.8242, "step": 3496 }, { "epoch": 0.03145166004865699, "grad_norm": 1.4894213676452637, "learning_rate": 4.081949235840937e-05, "loss": 2.427, "step": 3497 }, { "epoch": 0.031460653946297434, "grad_norm": 1.4826219081878662, "learning_rate": 4.081367194184831e-05, "loss": 2.5871, "step": 3498 }, { "epoch": 0.03146964784393789, "grad_norm": 1.6851222515106201, "learning_rate": 4.0807850096064605e-05, "loss": 2.3567, "step": 3499 }, { "epoch": 0.03147864174157834, "grad_norm": 1.5882151126861572, "learning_rate": 4.080202682158443e-05, "loss": 2.6965, "step": 3500 }, { "epoch": 0.03148763563921879, "grad_norm": 1.7023595571517944, "learning_rate": 4.0796202118934105e-05, "loss": 2.5957, "step": 3501 }, { "epoch": 0.03149662953685924, "grad_norm": 1.581197738647461, "learning_rate": 4.079037598864004e-05, "loss": 2.6929, "step": 3502 }, { "epoch": 0.03150562343449969, "grad_norm": 1.6653720140457153, "learning_rate": 4.07845484312288e-05, "loss": 2.405, "step": 3503 }, { "epoch": 0.031514617332140143, "grad_norm": 1.8919742107391357, "learning_rate": 4.077871944722707e-05, "loss": 2.7877, "step": 3504 }, { "epoch": 0.0315236112297806, "grad_norm": 1.8624435663223267, "learning_rate": 4.0772889037161674e-05, "loss": 2.8142, "step": 3505 }, { "epoch": 0.031532605127421044, "grad_norm": 0.5594093203544617, "learning_rate": 4.076705720155956e-05, "loss": 2.2481, "step": 3506 }, { "epoch": 0.0315415990250615, "grad_norm": 1.7148337364196777, "learning_rate": 4.07612239409478e-05, "loss": 2.7492, "step": 3507 }, { "epoch": 0.031550592922701945, "grad_norm": 1.8705363273620605, "learning_rate": 4.0755389255853596e-05, "loss": 2.6029, "step": 3508 }, { "epoch": 0.0315595868203424, "grad_norm": 1.7896169424057007, "learning_rate": 4.074955314680427e-05, "loss": 2.5484, "step": 3509 }, { "epoch": 0.031568580717982846, "grad_norm": 1.7371689081192017, "learning_rate": 4.0743715614327317e-05, "loss": 2.828, "step": 3510 }, { "epoch": 0.0315775746156233, "grad_norm": 1.962634801864624, "learning_rate": 4.073787665895029e-05, "loss": 2.7962, "step": 3511 }, { "epoch": 0.03158656851326375, "grad_norm": 1.2074649333953857, "learning_rate": 4.073203628120091e-05, "loss": 2.5396, "step": 3512 }, { "epoch": 0.0315955624109042, "grad_norm": 1.8012992143630981, "learning_rate": 4.0726194481607064e-05, "loss": 2.6088, "step": 3513 }, { "epoch": 0.031604556308544654, "grad_norm": 2.21783709526062, "learning_rate": 4.072035126069668e-05, "loss": 2.6591, "step": 3514 }, { "epoch": 0.0316135502061851, "grad_norm": 1.6806777715682983, "learning_rate": 4.071450661899789e-05, "loss": 2.5983, "step": 3515 }, { "epoch": 0.031622544103825555, "grad_norm": 1.1537336111068726, "learning_rate": 4.0708660557038915e-05, "loss": 2.2663, "step": 3516 }, { "epoch": 0.03163153800146601, "grad_norm": 1.3641184568405151, "learning_rate": 4.0702813075348116e-05, "loss": 2.3932, "step": 3517 }, { "epoch": 0.031640531899106455, "grad_norm": 1.874319076538086, "learning_rate": 4.0696964174453986e-05, "loss": 2.3891, "step": 3518 }, { "epoch": 0.03164952579674691, "grad_norm": 1.4884498119354248, "learning_rate": 4.069111385488515e-05, "loss": 2.5408, "step": 3519 }, { "epoch": 0.031658519694387356, "grad_norm": 1.3867486715316772, "learning_rate": 4.0685262117170336e-05, "loss": 2.5402, "step": 3520 }, { "epoch": 0.03166751359202781, "grad_norm": 1.367681860923767, "learning_rate": 4.067940896183843e-05, "loss": 2.3799, "step": 3521 }, { "epoch": 0.03167650748966826, "grad_norm": 1.206986904144287, "learning_rate": 4.067355438941842e-05, "loss": 2.3757, "step": 3522 }, { "epoch": 0.03168550138730871, "grad_norm": 1.5842413902282715, "learning_rate": 4.0667698400439466e-05, "loss": 2.621, "step": 3523 }, { "epoch": 0.031694495284949165, "grad_norm": 1.6283831596374512, "learning_rate": 4.066184099543079e-05, "loss": 2.5809, "step": 3524 }, { "epoch": 0.03170348918258961, "grad_norm": 2.4499545097351074, "learning_rate": 4.06559821749218e-05, "loss": 3.0529, "step": 3525 }, { "epoch": 0.031712483080230065, "grad_norm": 1.7996556758880615, "learning_rate": 4.065012193944201e-05, "loss": 2.3898, "step": 3526 }, { "epoch": 0.03172147697787051, "grad_norm": 2.1088201999664307, "learning_rate": 4.064426028952105e-05, "loss": 2.7646, "step": 3527 }, { "epoch": 0.031730470875510966, "grad_norm": 2.679507255554199, "learning_rate": 4.0638397225688695e-05, "loss": 2.5167, "step": 3528 }, { "epoch": 0.03173946477315142, "grad_norm": 2.2626891136169434, "learning_rate": 4.063253274847484e-05, "loss": 2.8624, "step": 3529 }, { "epoch": 0.03174845867079187, "grad_norm": 1.4691919088363647, "learning_rate": 4.0626666858409524e-05, "loss": 2.5003, "step": 3530 }, { "epoch": 0.03175745256843232, "grad_norm": 3.65517520904541, "learning_rate": 4.062079955602288e-05, "loss": 3.0182, "step": 3531 }, { "epoch": 0.03176644646607277, "grad_norm": 2.3816535472869873, "learning_rate": 4.06149308418452e-05, "loss": 2.8654, "step": 3532 }, { "epoch": 0.03177544036371322, "grad_norm": 1.3753752708435059, "learning_rate": 4.060906071640689e-05, "loss": 2.4594, "step": 3533 }, { "epoch": 0.03178443426135367, "grad_norm": 1.4689233303070068, "learning_rate": 4.060318918023849e-05, "loss": 2.5227, "step": 3534 }, { "epoch": 0.03179342815899412, "grad_norm": 1.3339725732803345, "learning_rate": 4.0597316233870644e-05, "loss": 2.3698, "step": 3535 }, { "epoch": 0.031802422056634576, "grad_norm": 1.263561725616455, "learning_rate": 4.0591441877834167e-05, "loss": 2.5347, "step": 3536 }, { "epoch": 0.03181141595427502, "grad_norm": 2.0930984020233154, "learning_rate": 4.058556611265997e-05, "loss": 2.6513, "step": 3537 }, { "epoch": 0.03182040985191548, "grad_norm": 1.5537431240081787, "learning_rate": 4.057968893887909e-05, "loss": 2.6657, "step": 3538 }, { "epoch": 0.03182940374955592, "grad_norm": 1.763088583946228, "learning_rate": 4.05738103570227e-05, "loss": 2.7179, "step": 3539 }, { "epoch": 0.03183839764719638, "grad_norm": 1.123969316482544, "learning_rate": 4.056793036762211e-05, "loss": 2.6029, "step": 3540 }, { "epoch": 0.03184739154483683, "grad_norm": 2.392209053039551, "learning_rate": 4.056204897120874e-05, "loss": 2.9682, "step": 3541 }, { "epoch": 0.03185638544247728, "grad_norm": 1.0905828475952148, "learning_rate": 4.055616616831415e-05, "loss": 2.4293, "step": 3542 }, { "epoch": 0.03186537934011773, "grad_norm": 1.6227232217788696, "learning_rate": 4.0550281959470023e-05, "loss": 2.5869, "step": 3543 }, { "epoch": 0.03187437323775818, "grad_norm": 1.4821815490722656, "learning_rate": 4.054439634520815e-05, "loss": 2.6369, "step": 3544 }, { "epoch": 0.03188336713539863, "grad_norm": 1.594957709312439, "learning_rate": 4.0538509326060484e-05, "loss": 2.8543, "step": 3545 }, { "epoch": 0.031892361033039086, "grad_norm": 1.8716062307357788, "learning_rate": 4.053262090255908e-05, "loss": 2.7421, "step": 3546 }, { "epoch": 0.03190135493067953, "grad_norm": 1.734070897102356, "learning_rate": 4.052673107523613e-05, "loss": 2.4905, "step": 3547 }, { "epoch": 0.03191034882831999, "grad_norm": 1.724765419960022, "learning_rate": 4.0520839844623945e-05, "loss": 2.9236, "step": 3548 }, { "epoch": 0.031919342725960434, "grad_norm": 2.0353503227233887, "learning_rate": 4.051494721125499e-05, "loss": 2.3993, "step": 3549 }, { "epoch": 0.03192833662360089, "grad_norm": 0.7322614192962646, "learning_rate": 4.050905317566179e-05, "loss": 2.2122, "step": 3550 }, { "epoch": 0.031937330521241335, "grad_norm": 1.6296218633651733, "learning_rate": 4.050315773837708e-05, "loss": 2.6714, "step": 3551 }, { "epoch": 0.03194632441888179, "grad_norm": 1.5368154048919678, "learning_rate": 4.049726089993368e-05, "loss": 2.7107, "step": 3552 }, { "epoch": 0.03195531831652224, "grad_norm": 1.8119735717773438, "learning_rate": 4.049136266086453e-05, "loss": 2.5896, "step": 3553 }, { "epoch": 0.03196431221416269, "grad_norm": 2.5196995735168457, "learning_rate": 4.048546302170271e-05, "loss": 2.3777, "step": 3554 }, { "epoch": 0.03197330611180314, "grad_norm": 1.0583696365356445, "learning_rate": 4.0479561982981415e-05, "loss": 2.3624, "step": 3555 }, { "epoch": 0.03198230000944359, "grad_norm": 0.8278627395629883, "learning_rate": 4.047365954523398e-05, "loss": 2.1862, "step": 3556 }, { "epoch": 0.031991293907084044, "grad_norm": 1.4801901578903198, "learning_rate": 4.046775570899387e-05, "loss": 2.4405, "step": 3557 }, { "epoch": 0.0320002878047245, "grad_norm": 2.147960662841797, "learning_rate": 4.046185047479465e-05, "loss": 2.9149, "step": 3558 }, { "epoch": 0.032009281702364945, "grad_norm": 1.89235520362854, "learning_rate": 4.045594384317005e-05, "loss": 2.5563, "step": 3559 }, { "epoch": 0.0320182756000054, "grad_norm": 1.9691011905670166, "learning_rate": 4.045003581465388e-05, "loss": 2.6087, "step": 3560 }, { "epoch": 0.032027269497645845, "grad_norm": 1.5244743824005127, "learning_rate": 4.0444126389780126e-05, "loss": 2.588, "step": 3561 }, { "epoch": 0.0320362633952863, "grad_norm": 1.8930597305297852, "learning_rate": 4.043821556908286e-05, "loss": 2.6866, "step": 3562 }, { "epoch": 0.032045257292926746, "grad_norm": 0.9449501633644104, "learning_rate": 4.04323033530963e-05, "loss": 2.3257, "step": 3563 }, { "epoch": 0.0320542511905672, "grad_norm": 3.0702691078186035, "learning_rate": 4.042638974235478e-05, "loss": 2.3377, "step": 3564 }, { "epoch": 0.032063245088207654, "grad_norm": 1.8839943408966064, "learning_rate": 4.042047473739278e-05, "loss": 2.6441, "step": 3565 }, { "epoch": 0.0320722389858481, "grad_norm": 1.928968906402588, "learning_rate": 4.041455833874488e-05, "loss": 2.4406, "step": 3566 }, { "epoch": 0.032081232883488554, "grad_norm": 1.6300245523452759, "learning_rate": 4.04086405469458e-05, "loss": 2.7544, "step": 3567 }, { "epoch": 0.032090226781129, "grad_norm": 2.8023571968078613, "learning_rate": 4.040272136253037e-05, "loss": 2.8009, "step": 3568 }, { "epoch": 0.032099220678769455, "grad_norm": 2.413008689880371, "learning_rate": 4.03968007860336e-05, "loss": 2.6393, "step": 3569 }, { "epoch": 0.03210821457640991, "grad_norm": 1.7155275344848633, "learning_rate": 4.039087881799054e-05, "loss": 2.6861, "step": 3570 }, { "epoch": 0.032117208474050356, "grad_norm": 1.8222769498825073, "learning_rate": 4.038495545893643e-05, "loss": 2.8935, "step": 3571 }, { "epoch": 0.03212620237169081, "grad_norm": 2.3011767864227295, "learning_rate": 4.037903070940663e-05, "loss": 2.6748, "step": 3572 }, { "epoch": 0.032135196269331257, "grad_norm": 1.7426681518554688, "learning_rate": 4.0373104569936594e-05, "loss": 2.759, "step": 3573 }, { "epoch": 0.03214419016697171, "grad_norm": 3.6192615032196045, "learning_rate": 4.0367177041061913e-05, "loss": 2.6811, "step": 3574 }, { "epoch": 0.03215318406461216, "grad_norm": 1.873392939567566, "learning_rate": 4.0361248123318335e-05, "loss": 3.0157, "step": 3575 }, { "epoch": 0.03216217796225261, "grad_norm": 1.473142147064209, "learning_rate": 4.03553178172417e-05, "loss": 2.2212, "step": 3576 }, { "epoch": 0.032171171859893065, "grad_norm": 2.0809760093688965, "learning_rate": 4.034938612336797e-05, "loss": 2.4585, "step": 3577 }, { "epoch": 0.03218016575753351, "grad_norm": 1.4646273851394653, "learning_rate": 4.034345304223327e-05, "loss": 2.5326, "step": 3578 }, { "epoch": 0.032189159655173966, "grad_norm": 1.238966941833496, "learning_rate": 4.03375185743738e-05, "loss": 2.5825, "step": 3579 }, { "epoch": 0.03219815355281441, "grad_norm": 1.5712103843688965, "learning_rate": 4.033158272032592e-05, "loss": 2.625, "step": 3580 }, { "epoch": 0.032207147450454866, "grad_norm": 1.7034969329833984, "learning_rate": 4.0325645480626114e-05, "loss": 2.8374, "step": 3581 }, { "epoch": 0.03221614134809532, "grad_norm": 1.5879398584365845, "learning_rate": 4.031970685581097e-05, "loss": 2.4241, "step": 3582 }, { "epoch": 0.03222513524573577, "grad_norm": 1.8574512004852295, "learning_rate": 4.031376684641722e-05, "loss": 2.6999, "step": 3583 }, { "epoch": 0.03223412914337622, "grad_norm": 1.1938265562057495, "learning_rate": 4.030782545298173e-05, "loss": 2.4209, "step": 3584 }, { "epoch": 0.03224312304101667, "grad_norm": 1.8512554168701172, "learning_rate": 4.0301882676041456e-05, "loss": 2.6119, "step": 3585 }, { "epoch": 0.03225211693865712, "grad_norm": 2.2730696201324463, "learning_rate": 4.029593851613351e-05, "loss": 2.5533, "step": 3586 }, { "epoch": 0.032261110836297575, "grad_norm": 0.7881747484207153, "learning_rate": 4.028999297379511e-05, "loss": 2.2206, "step": 3587 }, { "epoch": 0.03227010473393802, "grad_norm": 1.7286643981933594, "learning_rate": 4.028404604956362e-05, "loss": 2.7824, "step": 3588 }, { "epoch": 0.032279098631578476, "grad_norm": 2.317812919616699, "learning_rate": 4.0278097743976503e-05, "loss": 2.8735, "step": 3589 }, { "epoch": 0.03228809252921892, "grad_norm": 1.2177726030349731, "learning_rate": 4.027214805757138e-05, "loss": 2.2985, "step": 3590 }, { "epoch": 0.03229708642685938, "grad_norm": 2.1486735343933105, "learning_rate": 4.0266196990885955e-05, "loss": 2.9185, "step": 3591 }, { "epoch": 0.032306080324499824, "grad_norm": 2.2497661113739014, "learning_rate": 4.026024454445809e-05, "loss": 2.8424, "step": 3592 }, { "epoch": 0.03231507422214028, "grad_norm": 1.717777967453003, "learning_rate": 4.0254290718825774e-05, "loss": 2.6471, "step": 3593 }, { "epoch": 0.03232406811978073, "grad_norm": 2.029930591583252, "learning_rate": 4.0248335514527075e-05, "loss": 2.6074, "step": 3594 }, { "epoch": 0.03233306201742118, "grad_norm": 1.8137168884277344, "learning_rate": 4.024237893210024e-05, "loss": 2.5146, "step": 3595 }, { "epoch": 0.03234205591506163, "grad_norm": 1.8733121156692505, "learning_rate": 4.023642097208362e-05, "loss": 2.4684, "step": 3596 }, { "epoch": 0.03235104981270208, "grad_norm": 1.5022145509719849, "learning_rate": 4.0230461635015695e-05, "loss": 2.616, "step": 3597 }, { "epoch": 0.03236004371034253, "grad_norm": 1.6463229656219482, "learning_rate": 4.0224500921435035e-05, "loss": 2.6135, "step": 3598 }, { "epoch": 0.03236903760798299, "grad_norm": 2.0474395751953125, "learning_rate": 4.021853883188038e-05, "loss": 2.4848, "step": 3599 }, { "epoch": 0.032378031505623434, "grad_norm": 1.5436025857925415, "learning_rate": 4.0212575366890587e-05, "loss": 2.6402, "step": 3600 }, { "epoch": 0.03238702540326389, "grad_norm": 1.5756020545959473, "learning_rate": 4.020661052700461e-05, "loss": 2.6303, "step": 3601 }, { "epoch": 0.032396019300904334, "grad_norm": 2.007960081100464, "learning_rate": 4.020064431276156e-05, "loss": 2.822, "step": 3602 }, { "epoch": 0.03240501319854479, "grad_norm": 1.9185757637023926, "learning_rate": 4.0194676724700645e-05, "loss": 2.6486, "step": 3603 }, { "epoch": 0.032414007096185235, "grad_norm": 2.931666135787964, "learning_rate": 4.018870776336122e-05, "loss": 2.4643, "step": 3604 }, { "epoch": 0.03242300099382569, "grad_norm": 1.0200412273406982, "learning_rate": 4.0182737429282734e-05, "loss": 2.3513, "step": 3605 }, { "epoch": 0.03243199489146614, "grad_norm": 1.7788134813308716, "learning_rate": 4.017676572300481e-05, "loss": 2.8979, "step": 3606 }, { "epoch": 0.03244098878910659, "grad_norm": 1.6414347887039185, "learning_rate": 4.017079264506713e-05, "loss": 2.8639, "step": 3607 }, { "epoch": 0.03244998268674704, "grad_norm": 1.728614091873169, "learning_rate": 4.016481819600957e-05, "loss": 2.7283, "step": 3608 }, { "epoch": 0.03245897658438749, "grad_norm": 2.4541995525360107, "learning_rate": 4.0158842376372064e-05, "loss": 2.9056, "step": 3609 }, { "epoch": 0.032467970482027944, "grad_norm": 1.4306966066360474, "learning_rate": 4.015286518669471e-05, "loss": 2.4496, "step": 3610 }, { "epoch": 0.0324769643796684, "grad_norm": 1.9879299402236938, "learning_rate": 4.014688662751772e-05, "loss": 2.9079, "step": 3611 }, { "epoch": 0.032485958277308845, "grad_norm": 2.0274124145507812, "learning_rate": 4.014090669938144e-05, "loss": 2.7617, "step": 3612 }, { "epoch": 0.0324949521749493, "grad_norm": 1.451941728591919, "learning_rate": 4.013492540282632e-05, "loss": 2.636, "step": 3613 }, { "epoch": 0.032503946072589746, "grad_norm": 1.4582048654556274, "learning_rate": 4.012894273839294e-05, "loss": 2.4876, "step": 3614 }, { "epoch": 0.0325129399702302, "grad_norm": 1.113447904586792, "learning_rate": 4.012295870662202e-05, "loss": 2.3139, "step": 3615 }, { "epoch": 0.03252193386787065, "grad_norm": 1.3956459760665894, "learning_rate": 4.0116973308054365e-05, "loss": 2.4382, "step": 3616 }, { "epoch": 0.0325309277655111, "grad_norm": 1.6250743865966797, "learning_rate": 4.011098654323096e-05, "loss": 2.8399, "step": 3617 }, { "epoch": 0.032539921663151554, "grad_norm": 4.397181510925293, "learning_rate": 4.0104998412692854e-05, "loss": 2.5378, "step": 3618 }, { "epoch": 0.032548915560792, "grad_norm": 1.4838683605194092, "learning_rate": 4.0099008916981266e-05, "loss": 2.6187, "step": 3619 }, { "epoch": 0.032557909458432455, "grad_norm": 2.029542922973633, "learning_rate": 4.009301805663752e-05, "loss": 2.4478, "step": 3620 }, { "epoch": 0.0325669033560729, "grad_norm": 1.5884758234024048, "learning_rate": 4.008702583220306e-05, "loss": 2.4364, "step": 3621 }, { "epoch": 0.032575897253713355, "grad_norm": 2.0209741592407227, "learning_rate": 4.008103224421945e-05, "loss": 2.6443, "step": 3622 }, { "epoch": 0.03258489115135381, "grad_norm": 2.0704917907714844, "learning_rate": 4.007503729322839e-05, "loss": 2.5671, "step": 3623 }, { "epoch": 0.032593885048994256, "grad_norm": 2.178070306777954, "learning_rate": 4.00690409797717e-05, "loss": 2.9684, "step": 3624 }, { "epoch": 0.03260287894663471, "grad_norm": 1.854955792427063, "learning_rate": 4.0063043304391315e-05, "loss": 2.4085, "step": 3625 }, { "epoch": 0.03261187284427516, "grad_norm": 2.9342503547668457, "learning_rate": 4.005704426762931e-05, "loss": 2.6673, "step": 3626 }, { "epoch": 0.03262086674191561, "grad_norm": 1.8421697616577148, "learning_rate": 4.005104387002785e-05, "loss": 2.6374, "step": 3627 }, { "epoch": 0.032629860639556064, "grad_norm": 2.1421334743499756, "learning_rate": 4.004504211212926e-05, "loss": 2.5565, "step": 3628 }, { "epoch": 0.03263885453719651, "grad_norm": 1.7831922769546509, "learning_rate": 4.003903899447597e-05, "loss": 2.7665, "step": 3629 }, { "epoch": 0.032647848434836965, "grad_norm": 1.7842991352081299, "learning_rate": 4.003303451761053e-05, "loss": 2.7217, "step": 3630 }, { "epoch": 0.03265684233247741, "grad_norm": 1.8717799186706543, "learning_rate": 4.002702868207563e-05, "loss": 2.7771, "step": 3631 }, { "epoch": 0.032665836230117866, "grad_norm": 1.3628075122833252, "learning_rate": 4.0021021488414064e-05, "loss": 2.309, "step": 3632 }, { "epoch": 0.03267483012775831, "grad_norm": 2.338263511657715, "learning_rate": 4.001501293716875e-05, "loss": 2.5995, "step": 3633 }, { "epoch": 0.03268382402539877, "grad_norm": 1.7249672412872314, "learning_rate": 4.0009003028882744e-05, "loss": 2.484, "step": 3634 }, { "epoch": 0.03269281792303922, "grad_norm": 1.7328068017959595, "learning_rate": 4.0002991764099204e-05, "loss": 2.5352, "step": 3635 }, { "epoch": 0.03270181182067967, "grad_norm": 1.8944132328033447, "learning_rate": 3.999697914336143e-05, "loss": 2.7617, "step": 3636 }, { "epoch": 0.03271080571832012, "grad_norm": 2.2706499099731445, "learning_rate": 3.9990965167212837e-05, "loss": 2.9333, "step": 3637 }, { "epoch": 0.03271979961596057, "grad_norm": 2.28812575340271, "learning_rate": 3.998494983619696e-05, "loss": 2.5904, "step": 3638 }, { "epoch": 0.03272879351360102, "grad_norm": 1.4702509641647339, "learning_rate": 3.997893315085746e-05, "loss": 2.4018, "step": 3639 }, { "epoch": 0.032737787411241476, "grad_norm": 1.989013433456421, "learning_rate": 3.9972915111738105e-05, "loss": 2.909, "step": 3640 }, { "epoch": 0.03274678130888192, "grad_norm": 2.069377899169922, "learning_rate": 3.996689571938282e-05, "loss": 3.0038, "step": 3641 }, { "epoch": 0.032755775206522376, "grad_norm": 1.9793119430541992, "learning_rate": 3.996087497433562e-05, "loss": 2.5693, "step": 3642 }, { "epoch": 0.03276476910416282, "grad_norm": 1.7479356527328491, "learning_rate": 3.9954852877140647e-05, "loss": 2.4704, "step": 3643 }, { "epoch": 0.03277376300180328, "grad_norm": 2.429288148880005, "learning_rate": 3.994882942834218e-05, "loss": 2.8486, "step": 3644 }, { "epoch": 0.032782756899443724, "grad_norm": 1.435154914855957, "learning_rate": 3.994280462848462e-05, "loss": 2.6056, "step": 3645 }, { "epoch": 0.03279175079708418, "grad_norm": 2.0152359008789062, "learning_rate": 3.9936778478112463e-05, "loss": 2.8415, "step": 3646 }, { "epoch": 0.03280074469472463, "grad_norm": 1.3133597373962402, "learning_rate": 3.993075097777036e-05, "loss": 2.3228, "step": 3647 }, { "epoch": 0.03280973859236508, "grad_norm": 1.7511314153671265, "learning_rate": 3.9924722128003064e-05, "loss": 2.4705, "step": 3648 }, { "epoch": 0.03281873249000553, "grad_norm": 2.1378157138824463, "learning_rate": 3.9918691929355466e-05, "loss": 2.8542, "step": 3649 }, { "epoch": 0.03282772638764598, "grad_norm": 1.2630749940872192, "learning_rate": 3.9912660382372545e-05, "loss": 2.3809, "step": 3650 }, { "epoch": 0.03283672028528643, "grad_norm": 2.39906907081604, "learning_rate": 3.990662748759946e-05, "loss": 2.7286, "step": 3651 }, { "epoch": 0.03284571418292689, "grad_norm": 2.020158052444458, "learning_rate": 3.990059324558142e-05, "loss": 2.86, "step": 3652 }, { "epoch": 0.032854708080567334, "grad_norm": 1.7078821659088135, "learning_rate": 3.9894557656863823e-05, "loss": 2.6189, "step": 3653 }, { "epoch": 0.03286370197820779, "grad_norm": 1.0153225660324097, "learning_rate": 3.988852072199214e-05, "loss": 2.3834, "step": 3654 }, { "epoch": 0.032872695875848235, "grad_norm": 1.4995334148406982, "learning_rate": 3.9882482441512e-05, "loss": 2.4677, "step": 3655 }, { "epoch": 0.03288168977348869, "grad_norm": 1.2416150569915771, "learning_rate": 3.987644281596913e-05, "loss": 2.3653, "step": 3656 }, { "epoch": 0.03289068367112914, "grad_norm": 1.6740343570709229, "learning_rate": 3.9870401845909375e-05, "loss": 2.5451, "step": 3657 }, { "epoch": 0.03289967756876959, "grad_norm": 2.1310760974884033, "learning_rate": 3.986435953187871e-05, "loss": 2.7053, "step": 3658 }, { "epoch": 0.03290867146641004, "grad_norm": 1.6410943269729614, "learning_rate": 3.985831587442326e-05, "loss": 2.5896, "step": 3659 }, { "epoch": 0.03291766536405049, "grad_norm": 1.9079653024673462, "learning_rate": 3.985227087408921e-05, "loss": 2.6016, "step": 3660 }, { "epoch": 0.032926659261690944, "grad_norm": 1.3547496795654297, "learning_rate": 3.9846224531422925e-05, "loss": 2.5951, "step": 3661 }, { "epoch": 0.03293565315933139, "grad_norm": 1.515633463859558, "learning_rate": 3.984017684697085e-05, "loss": 2.5972, "step": 3662 }, { "epoch": 0.032944647056971844, "grad_norm": 1.860808253288269, "learning_rate": 3.983412782127958e-05, "loss": 2.8165, "step": 3663 }, { "epoch": 0.0329536409546123, "grad_norm": 2.0141289234161377, "learning_rate": 3.9828077454895815e-05, "loss": 2.731, "step": 3664 }, { "epoch": 0.032962634852252745, "grad_norm": 1.6866053342819214, "learning_rate": 3.982202574836638e-05, "loss": 2.567, "step": 3665 }, { "epoch": 0.0329716287498932, "grad_norm": 2.0215282440185547, "learning_rate": 3.9815972702238225e-05, "loss": 2.7155, "step": 3666 }, { "epoch": 0.032980622647533646, "grad_norm": 1.6232118606567383, "learning_rate": 3.980991831705842e-05, "loss": 2.7144, "step": 3667 }, { "epoch": 0.0329896165451741, "grad_norm": 1.8770339488983154, "learning_rate": 3.980386259337414e-05, "loss": 2.6233, "step": 3668 }, { "epoch": 0.032998610442814554, "grad_norm": 1.5729364156723022, "learning_rate": 3.979780553173271e-05, "loss": 2.6431, "step": 3669 }, { "epoch": 0.033007604340455, "grad_norm": 1.8756521940231323, "learning_rate": 3.9791747132681554e-05, "loss": 2.7806, "step": 3670 }, { "epoch": 0.033016598238095454, "grad_norm": 2.144597053527832, "learning_rate": 3.978568739676822e-05, "loss": 2.9947, "step": 3671 }, { "epoch": 0.0330255921357359, "grad_norm": 1.7208480834960938, "learning_rate": 3.9779626324540395e-05, "loss": 2.2922, "step": 3672 }, { "epoch": 0.033034586033376355, "grad_norm": 1.8949164152145386, "learning_rate": 3.977356391654585e-05, "loss": 2.2378, "step": 3673 }, { "epoch": 0.0330435799310168, "grad_norm": 2.2295749187469482, "learning_rate": 3.976750017333252e-05, "loss": 3.082, "step": 3674 }, { "epoch": 0.033052573828657256, "grad_norm": 1.5571682453155518, "learning_rate": 3.976143509544843e-05, "loss": 2.716, "step": 3675 }, { "epoch": 0.03306156772629771, "grad_norm": 1.548628568649292, "learning_rate": 3.9755368683441735e-05, "loss": 2.2846, "step": 3676 }, { "epoch": 0.033070561623938156, "grad_norm": 1.3922665119171143, "learning_rate": 3.974930093786071e-05, "loss": 2.5234, "step": 3677 }, { "epoch": 0.03307955552157861, "grad_norm": 1.7753299474716187, "learning_rate": 3.974323185925375e-05, "loss": 2.5931, "step": 3678 }, { "epoch": 0.03308854941921906, "grad_norm": 1.8115192651748657, "learning_rate": 3.973716144816938e-05, "loss": 2.8823, "step": 3679 }, { "epoch": 0.03309754331685951, "grad_norm": 1.9802744388580322, "learning_rate": 3.973108970515623e-05, "loss": 2.8719, "step": 3680 }, { "epoch": 0.033106537214499965, "grad_norm": 1.7217671871185303, "learning_rate": 3.9725016630763054e-05, "loss": 2.8143, "step": 3681 }, { "epoch": 0.03311553111214041, "grad_norm": 3.014029026031494, "learning_rate": 3.971894222553875e-05, "loss": 3.129, "step": 3682 }, { "epoch": 0.033124525009780866, "grad_norm": 1.3727129697799683, "learning_rate": 3.971286649003229e-05, "loss": 2.4753, "step": 3683 }, { "epoch": 0.03313351890742131, "grad_norm": 2.260828971862793, "learning_rate": 3.9706789424792804e-05, "loss": 2.9526, "step": 3684 }, { "epoch": 0.033142512805061766, "grad_norm": 1.5782235860824585, "learning_rate": 3.9700711030369534e-05, "loss": 2.776, "step": 3685 }, { "epoch": 0.03315150670270221, "grad_norm": 2.6127612590789795, "learning_rate": 3.969463130731183e-05, "loss": 2.7875, "step": 3686 }, { "epoch": 0.03316050060034267, "grad_norm": 2.2365381717681885, "learning_rate": 3.9688550256169185e-05, "loss": 2.3133, "step": 3687 }, { "epoch": 0.03316949449798312, "grad_norm": 1.371523141860962, "learning_rate": 3.968246787749118e-05, "loss": 2.6505, "step": 3688 }, { "epoch": 0.03317848839562357, "grad_norm": 1.7309523820877075, "learning_rate": 3.9676384171827547e-05, "loss": 2.7576, "step": 3689 }, { "epoch": 0.03318748229326402, "grad_norm": 1.478163242340088, "learning_rate": 3.9670299139728116e-05, "loss": 2.5234, "step": 3690 }, { "epoch": 0.03319647619090447, "grad_norm": 1.4423015117645264, "learning_rate": 3.966421278174286e-05, "loss": 2.4761, "step": 3691 }, { "epoch": 0.03320547008854492, "grad_norm": 1.5634043216705322, "learning_rate": 3.965812509842184e-05, "loss": 2.6364, "step": 3692 }, { "epoch": 0.033214463986185376, "grad_norm": 1.393546223640442, "learning_rate": 3.965203609031527e-05, "loss": 2.4415, "step": 3693 }, { "epoch": 0.03322345788382582, "grad_norm": 1.9648776054382324, "learning_rate": 3.964594575797346e-05, "loss": 2.9337, "step": 3694 }, { "epoch": 0.03323245178146628, "grad_norm": 1.545791506767273, "learning_rate": 3.963985410194684e-05, "loss": 2.6881, "step": 3695 }, { "epoch": 0.033241445679106724, "grad_norm": 1.937265157699585, "learning_rate": 3.963376112278597e-05, "loss": 2.8216, "step": 3696 }, { "epoch": 0.03325043957674718, "grad_norm": 2.0932743549346924, "learning_rate": 3.9627666821041545e-05, "loss": 2.6351, "step": 3697 }, { "epoch": 0.03325943347438763, "grad_norm": 1.917465329170227, "learning_rate": 3.962157119726434e-05, "loss": 2.5523, "step": 3698 }, { "epoch": 0.03326842737202808, "grad_norm": 2.0996181964874268, "learning_rate": 3.961547425200528e-05, "loss": 2.6505, "step": 3699 }, { "epoch": 0.03327742126966853, "grad_norm": 2.450352430343628, "learning_rate": 3.9609375985815405e-05, "loss": 2.8337, "step": 3700 }, { "epoch": 0.03328641516730898, "grad_norm": 2.124682664871216, "learning_rate": 3.960327639924586e-05, "loss": 2.3837, "step": 3701 }, { "epoch": 0.03329540906494943, "grad_norm": 1.2955535650253296, "learning_rate": 3.959717549284793e-05, "loss": 2.5468, "step": 3702 }, { "epoch": 0.03330440296258988, "grad_norm": 2.2891628742218018, "learning_rate": 3.9591073267172996e-05, "loss": 2.7056, "step": 3703 }, { "epoch": 0.033313396860230333, "grad_norm": 1.7695691585540771, "learning_rate": 3.958496972277257e-05, "loss": 2.6396, "step": 3704 }, { "epoch": 0.03332239075787079, "grad_norm": 2.0943799018859863, "learning_rate": 3.95788648601983e-05, "loss": 2.893, "step": 3705 }, { "epoch": 0.033331384655511234, "grad_norm": 2.072481155395508, "learning_rate": 3.957275868000192e-05, "loss": 2.716, "step": 3706 }, { "epoch": 0.03334037855315169, "grad_norm": 1.734165906906128, "learning_rate": 3.956665118273532e-05, "loss": 2.8471, "step": 3707 }, { "epoch": 0.033349372450792135, "grad_norm": 1.7622182369232178, "learning_rate": 3.956054236895046e-05, "loss": 2.7351, "step": 3708 }, { "epoch": 0.03335836634843259, "grad_norm": 1.5400911569595337, "learning_rate": 3.955443223919947e-05, "loss": 2.5137, "step": 3709 }, { "epoch": 0.03336736024607304, "grad_norm": 1.7568118572235107, "learning_rate": 3.9548320794034574e-05, "loss": 2.5336, "step": 3710 }, { "epoch": 0.03337635414371349, "grad_norm": 1.3058832883834839, "learning_rate": 3.954220803400811e-05, "loss": 2.3056, "step": 3711 }, { "epoch": 0.03338534804135394, "grad_norm": 2.0841336250305176, "learning_rate": 3.953609395967255e-05, "loss": 2.3692, "step": 3712 }, { "epoch": 0.03339434193899439, "grad_norm": 1.345820665359497, "learning_rate": 3.952997857158047e-05, "loss": 2.3321, "step": 3713 }, { "epoch": 0.033403335836634844, "grad_norm": 1.3058717250823975, "learning_rate": 3.952386187028459e-05, "loss": 2.5238, "step": 3714 }, { "epoch": 0.03341232973427529, "grad_norm": 1.8011940717697144, "learning_rate": 3.9517743856337706e-05, "loss": 2.7084, "step": 3715 }, { "epoch": 0.033421323631915745, "grad_norm": 1.44242525100708, "learning_rate": 3.951162453029278e-05, "loss": 2.5532, "step": 3716 }, { "epoch": 0.0334303175295562, "grad_norm": 1.6459295749664307, "learning_rate": 3.950550389270285e-05, "loss": 2.5109, "step": 3717 }, { "epoch": 0.033439311427196645, "grad_norm": 2.143077850341797, "learning_rate": 3.949938194412111e-05, "loss": 2.5907, "step": 3718 }, { "epoch": 0.0334483053248371, "grad_norm": 2.743124485015869, "learning_rate": 3.949325868510083e-05, "loss": 2.5837, "step": 3719 }, { "epoch": 0.033457299222477546, "grad_norm": 1.444530963897705, "learning_rate": 3.9487134116195454e-05, "loss": 2.4761, "step": 3720 }, { "epoch": 0.033466293120118, "grad_norm": 1.695156216621399, "learning_rate": 3.94810082379585e-05, "loss": 2.4988, "step": 3721 }, { "epoch": 0.033475287017758454, "grad_norm": 1.5944336652755737, "learning_rate": 3.947488105094362e-05, "loss": 2.3387, "step": 3722 }, { "epoch": 0.0334842809153989, "grad_norm": 2.474018096923828, "learning_rate": 3.9468752555704584e-05, "loss": 2.467, "step": 3723 }, { "epoch": 0.033493274813039355, "grad_norm": 2.3061063289642334, "learning_rate": 3.946262275279528e-05, "loss": 2.763, "step": 3724 }, { "epoch": 0.0335022687106798, "grad_norm": 1.7110357284545898, "learning_rate": 3.94564916427697e-05, "loss": 2.5504, "step": 3725 }, { "epoch": 0.033511262608320255, "grad_norm": 2.3513896465301514, "learning_rate": 3.945035922618197e-05, "loss": 2.7294, "step": 3726 }, { "epoch": 0.03352025650596071, "grad_norm": 1.7288864850997925, "learning_rate": 3.9444225503586355e-05, "loss": 2.7605, "step": 3727 }, { "epoch": 0.033529250403601156, "grad_norm": 1.5982245206832886, "learning_rate": 3.943809047553719e-05, "loss": 2.3483, "step": 3728 }, { "epoch": 0.03353824430124161, "grad_norm": 1.836276650428772, "learning_rate": 3.9431954142588956e-05, "loss": 2.3282, "step": 3729 }, { "epoch": 0.03354723819888206, "grad_norm": 1.5374042987823486, "learning_rate": 3.942581650529625e-05, "loss": 2.747, "step": 3730 }, { "epoch": 0.03355623209652251, "grad_norm": 1.4159756898880005, "learning_rate": 3.94196775642138e-05, "loss": 2.4583, "step": 3731 }, { "epoch": 0.03356522599416296, "grad_norm": 1.6580169200897217, "learning_rate": 3.941353731989641e-05, "loss": 2.6339, "step": 3732 }, { "epoch": 0.03357421989180341, "grad_norm": 1.7374281883239746, "learning_rate": 3.940739577289905e-05, "loss": 2.5716, "step": 3733 }, { "epoch": 0.033583213789443865, "grad_norm": 1.6529935598373413, "learning_rate": 3.940125292377677e-05, "loss": 2.588, "step": 3734 }, { "epoch": 0.03359220768708431, "grad_norm": 2.4541523456573486, "learning_rate": 3.939510877308477e-05, "loss": 2.9344, "step": 3735 }, { "epoch": 0.033601201584724766, "grad_norm": 1.5681049823760986, "learning_rate": 3.938896332137834e-05, "loss": 2.6118, "step": 3736 }, { "epoch": 0.03361019548236521, "grad_norm": 1.6124038696289062, "learning_rate": 3.93828165692129e-05, "loss": 2.6259, "step": 3737 }, { "epoch": 0.03361918938000567, "grad_norm": 1.1825910806655884, "learning_rate": 3.9376668517144e-05, "loss": 2.3936, "step": 3738 }, { "epoch": 0.03362818327764612, "grad_norm": 2.166827440261841, "learning_rate": 3.937051916572728e-05, "loss": 2.5845, "step": 3739 }, { "epoch": 0.03363717717528657, "grad_norm": 1.6936359405517578, "learning_rate": 3.9364368515518514e-05, "loss": 2.7419, "step": 3740 }, { "epoch": 0.03364617107292702, "grad_norm": 1.8919154405593872, "learning_rate": 3.935821656707359e-05, "loss": 2.5458, "step": 3741 }, { "epoch": 0.03365516497056747, "grad_norm": 1.908681035041809, "learning_rate": 3.9352063320948527e-05, "loss": 2.5486, "step": 3742 }, { "epoch": 0.03366415886820792, "grad_norm": 2.0029842853546143, "learning_rate": 3.934590877769944e-05, "loss": 3.1182, "step": 3743 }, { "epoch": 0.03367315276584837, "grad_norm": 1.7686482667922974, "learning_rate": 3.9339752937882565e-05, "loss": 2.5542, "step": 3744 }, { "epoch": 0.03368214666348882, "grad_norm": 2.1062309741973877, "learning_rate": 3.933359580205427e-05, "loss": 2.8004, "step": 3745 }, { "epoch": 0.033691140561129276, "grad_norm": 1.2170573472976685, "learning_rate": 3.9327437370771014e-05, "loss": 2.4618, "step": 3746 }, { "epoch": 0.03370013445876972, "grad_norm": 1.4031569957733154, "learning_rate": 3.932127764458941e-05, "loss": 2.5223, "step": 3747 }, { "epoch": 0.03370912835641018, "grad_norm": 1.8125271797180176, "learning_rate": 3.931511662406616e-05, "loss": 2.7107, "step": 3748 }, { "epoch": 0.033718122254050624, "grad_norm": 1.8426556587219238, "learning_rate": 3.9308954309758085e-05, "loss": 2.5903, "step": 3749 }, { "epoch": 0.03372711615169108, "grad_norm": 2.4040496349334717, "learning_rate": 3.930279070222213e-05, "loss": 2.823, "step": 3750 }, { "epoch": 0.03373611004933153, "grad_norm": 1.5169763565063477, "learning_rate": 3.9296625802015356e-05, "loss": 2.4748, "step": 3751 }, { "epoch": 0.03374510394697198, "grad_norm": 2.2871909141540527, "learning_rate": 3.929045960969494e-05, "loss": 2.433, "step": 3752 }, { "epoch": 0.03375409784461243, "grad_norm": 1.4511253833770752, "learning_rate": 3.928429212581819e-05, "loss": 2.4272, "step": 3753 }, { "epoch": 0.03376309174225288, "grad_norm": 1.5718731880187988, "learning_rate": 3.927812335094248e-05, "loss": 2.6088, "step": 3754 }, { "epoch": 0.03377208563989333, "grad_norm": 1.250670313835144, "learning_rate": 3.927195328562538e-05, "loss": 2.2208, "step": 3755 }, { "epoch": 0.03378107953753378, "grad_norm": 1.8173960447311401, "learning_rate": 3.9265781930424514e-05, "loss": 2.5212, "step": 3756 }, { "epoch": 0.033790073435174234, "grad_norm": 1.8730297088623047, "learning_rate": 3.925960928589764e-05, "loss": 2.92, "step": 3757 }, { "epoch": 0.03379906733281469, "grad_norm": 2.004077196121216, "learning_rate": 3.925343535260263e-05, "loss": 2.5782, "step": 3758 }, { "epoch": 0.033808061230455135, "grad_norm": 2.0933520793914795, "learning_rate": 3.9247260131097495e-05, "loss": 2.6451, "step": 3759 }, { "epoch": 0.03381705512809559, "grad_norm": 1.1789861917495728, "learning_rate": 3.924108362194033e-05, "loss": 2.309, "step": 3760 }, { "epoch": 0.033826049025736035, "grad_norm": 2.004755735397339, "learning_rate": 3.923490582568937e-05, "loss": 2.8363, "step": 3761 }, { "epoch": 0.03383504292337649, "grad_norm": 2.082210063934326, "learning_rate": 3.9228726742902956e-05, "loss": 2.8728, "step": 3762 }, { "epoch": 0.03384403682101694, "grad_norm": 1.9876586198806763, "learning_rate": 3.9222546374139533e-05, "loss": 2.6362, "step": 3763 }, { "epoch": 0.03385303071865739, "grad_norm": 1.1386542320251465, "learning_rate": 3.92163647199577e-05, "loss": 2.2778, "step": 3764 }, { "epoch": 0.033862024616297844, "grad_norm": 1.4220439195632935, "learning_rate": 3.921018178091613e-05, "loss": 2.4866, "step": 3765 }, { "epoch": 0.03387101851393829, "grad_norm": 1.7021104097366333, "learning_rate": 3.9203997557573645e-05, "loss": 2.4146, "step": 3766 }, { "epoch": 0.033880012411578744, "grad_norm": 1.6175121068954468, "learning_rate": 3.9197812050489146e-05, "loss": 2.7797, "step": 3767 }, { "epoch": 0.0338890063092192, "grad_norm": 1.7629095315933228, "learning_rate": 3.9191625260221695e-05, "loss": 2.4803, "step": 3768 }, { "epoch": 0.033898000206859645, "grad_norm": 1.41362726688385, "learning_rate": 3.9185437187330444e-05, "loss": 2.4763, "step": 3769 }, { "epoch": 0.0339069941045001, "grad_norm": 1.6962143182754517, "learning_rate": 3.917924783237466e-05, "loss": 2.5634, "step": 3770 }, { "epoch": 0.033915988002140546, "grad_norm": 1.8387856483459473, "learning_rate": 3.9173057195913715e-05, "loss": 2.838, "step": 3771 }, { "epoch": 0.033924981899781, "grad_norm": 2.63030743598938, "learning_rate": 3.9166865278507134e-05, "loss": 2.5913, "step": 3772 }, { "epoch": 0.033933975797421446, "grad_norm": 1.8950709104537964, "learning_rate": 3.916067208071453e-05, "loss": 2.5956, "step": 3773 }, { "epoch": 0.0339429696950619, "grad_norm": 1.8872044086456299, "learning_rate": 3.915447760309564e-05, "loss": 2.5757, "step": 3774 }, { "epoch": 0.033951963592702354, "grad_norm": 1.7051669359207153, "learning_rate": 3.91482818462103e-05, "loss": 2.7333, "step": 3775 }, { "epoch": 0.0339609574903428, "grad_norm": 1.7347087860107422, "learning_rate": 3.9142084810618495e-05, "loss": 2.6077, "step": 3776 }, { "epoch": 0.033969951387983255, "grad_norm": 2.178447723388672, "learning_rate": 3.9135886496880295e-05, "loss": 2.8282, "step": 3777 }, { "epoch": 0.0339789452856237, "grad_norm": 1.6145431995391846, "learning_rate": 3.91296869055559e-05, "loss": 2.488, "step": 3778 }, { "epoch": 0.033987939183264156, "grad_norm": 1.5070369243621826, "learning_rate": 3.912348603720561e-05, "loss": 2.648, "step": 3779 }, { "epoch": 0.03399693308090461, "grad_norm": 1.606884479522705, "learning_rate": 3.911728389238988e-05, "loss": 2.5296, "step": 3780 }, { "epoch": 0.034005926978545056, "grad_norm": 1.9123921394348145, "learning_rate": 3.911108047166924e-05, "loss": 2.5827, "step": 3781 }, { "epoch": 0.03401492087618551, "grad_norm": 1.2321248054504395, "learning_rate": 3.910487577560433e-05, "loss": 2.2468, "step": 3782 }, { "epoch": 0.03402391477382596, "grad_norm": 2.0566344261169434, "learning_rate": 3.909866980475595e-05, "loss": 2.8495, "step": 3783 }, { "epoch": 0.03403290867146641, "grad_norm": 1.7181134223937988, "learning_rate": 3.909246255968497e-05, "loss": 2.5143, "step": 3784 }, { "epoch": 0.03404190256910686, "grad_norm": 1.181848168373108, "learning_rate": 3.9086254040952416e-05, "loss": 2.5611, "step": 3785 }, { "epoch": 0.03405089646674731, "grad_norm": 1.5317614078521729, "learning_rate": 3.908004424911939e-05, "loss": 2.5974, "step": 3786 }, { "epoch": 0.034059890364387765, "grad_norm": 1.4889416694641113, "learning_rate": 3.9073833184747124e-05, "loss": 2.2715, "step": 3787 }, { "epoch": 0.03406888426202821, "grad_norm": 1.863641619682312, "learning_rate": 3.906762084839698e-05, "loss": 2.6339, "step": 3788 }, { "epoch": 0.034077878159668666, "grad_norm": 2.096662998199463, "learning_rate": 3.9061407240630416e-05, "loss": 2.9869, "step": 3789 }, { "epoch": 0.03408687205730911, "grad_norm": 1.7585359811782837, "learning_rate": 3.905519236200901e-05, "loss": 2.758, "step": 3790 }, { "epoch": 0.03409586595494957, "grad_norm": 1.658840537071228, "learning_rate": 3.9048976213094454e-05, "loss": 2.6998, "step": 3791 }, { "epoch": 0.03410485985259002, "grad_norm": 2.226508855819702, "learning_rate": 3.904275879444857e-05, "loss": 3.0113, "step": 3792 }, { "epoch": 0.03411385375023047, "grad_norm": 1.5728014707565308, "learning_rate": 3.903654010663326e-05, "loss": 2.5176, "step": 3793 }, { "epoch": 0.03412284764787092, "grad_norm": 1.1497588157653809, "learning_rate": 3.9030320150210586e-05, "loss": 2.3361, "step": 3794 }, { "epoch": 0.03413184154551137, "grad_norm": 1.9912992715835571, "learning_rate": 3.9024098925742683e-05, "loss": 2.6025, "step": 3795 }, { "epoch": 0.03414083544315182, "grad_norm": 1.8596857786178589, "learning_rate": 3.901787643379182e-05, "loss": 2.7259, "step": 3796 }, { "epoch": 0.03414982934079227, "grad_norm": 1.710956335067749, "learning_rate": 3.901165267492039e-05, "loss": 2.5212, "step": 3797 }, { "epoch": 0.03415882323843272, "grad_norm": 1.7185235023498535, "learning_rate": 3.9005427649690884e-05, "loss": 2.5378, "step": 3798 }, { "epoch": 0.03416781713607318, "grad_norm": 1.6389755010604858, "learning_rate": 3.8999201358665904e-05, "loss": 2.8447, "step": 3799 }, { "epoch": 0.034176811033713624, "grad_norm": 2.3170218467712402, "learning_rate": 3.899297380240819e-05, "loss": 3.1485, "step": 3800 }, { "epoch": 0.03418580493135408, "grad_norm": 2.130007743835449, "learning_rate": 3.898674498148058e-05, "loss": 3.0142, "step": 3801 }, { "epoch": 0.034194798828994524, "grad_norm": 1.632299780845642, "learning_rate": 3.898051489644601e-05, "loss": 2.6872, "step": 3802 }, { "epoch": 0.03420379272663498, "grad_norm": 1.9686185121536255, "learning_rate": 3.897428354786758e-05, "loss": 2.5402, "step": 3803 }, { "epoch": 0.03421278662427543, "grad_norm": 1.5067311525344849, "learning_rate": 3.8968050936308445e-05, "loss": 2.9395, "step": 3804 }, { "epoch": 0.03422178052191588, "grad_norm": 2.0449464321136475, "learning_rate": 3.896181706233191e-05, "loss": 3.007, "step": 3805 }, { "epoch": 0.03423077441955633, "grad_norm": 1.9509621858596802, "learning_rate": 3.8955581926501394e-05, "loss": 2.6705, "step": 3806 }, { "epoch": 0.03423976831719678, "grad_norm": 1.4788622856140137, "learning_rate": 3.894934552938041e-05, "loss": 2.653, "step": 3807 }, { "epoch": 0.03424876221483723, "grad_norm": 1.43935227394104, "learning_rate": 3.8943107871532605e-05, "loss": 2.4352, "step": 3808 }, { "epoch": 0.03425775611247769, "grad_norm": 1.1588003635406494, "learning_rate": 3.893686895352173e-05, "loss": 2.3809, "step": 3809 }, { "epoch": 0.034266750010118134, "grad_norm": 1.7148849964141846, "learning_rate": 3.893062877591164e-05, "loss": 2.696, "step": 3810 }, { "epoch": 0.03427574390775859, "grad_norm": 0.9685558676719666, "learning_rate": 3.8924387339266335e-05, "loss": 2.3507, "step": 3811 }, { "epoch": 0.034284737805399035, "grad_norm": 2.233031749725342, "learning_rate": 3.89181446441499e-05, "loss": 2.7828, "step": 3812 }, { "epoch": 0.03429373170303949, "grad_norm": 1.3864998817443848, "learning_rate": 3.891190069112654e-05, "loss": 2.4639, "step": 3813 }, { "epoch": 0.034302725600679936, "grad_norm": 1.9515326023101807, "learning_rate": 3.8905655480760586e-05, "loss": 2.725, "step": 3814 }, { "epoch": 0.03431171949832039, "grad_norm": 2.051046848297119, "learning_rate": 3.889940901361646e-05, "loss": 2.4164, "step": 3815 }, { "epoch": 0.03432071339596084, "grad_norm": 1.664448857307434, "learning_rate": 3.8893161290258725e-05, "loss": 2.6775, "step": 3816 }, { "epoch": 0.03432970729360129, "grad_norm": 1.2279576063156128, "learning_rate": 3.8886912311252036e-05, "loss": 2.3782, "step": 3817 }, { "epoch": 0.034338701191241744, "grad_norm": 1.379246711730957, "learning_rate": 3.888066207716117e-05, "loss": 2.7122, "step": 3818 }, { "epoch": 0.03434769508888219, "grad_norm": 1.3827881813049316, "learning_rate": 3.8874410588551013e-05, "loss": 2.3183, "step": 3819 }, { "epoch": 0.034356688986522645, "grad_norm": 2.289158344268799, "learning_rate": 3.8868157845986574e-05, "loss": 2.6657, "step": 3820 }, { "epoch": 0.0343656828841631, "grad_norm": 1.6705955266952515, "learning_rate": 3.8861903850032965e-05, "loss": 2.7126, "step": 3821 }, { "epoch": 0.034374676781803545, "grad_norm": 1.5154824256896973, "learning_rate": 3.885564860125542e-05, "loss": 2.6962, "step": 3822 }, { "epoch": 0.034383670679444, "grad_norm": 1.7197299003601074, "learning_rate": 3.884939210021928e-05, "loss": 2.521, "step": 3823 }, { "epoch": 0.034392664577084446, "grad_norm": 1.7329707145690918, "learning_rate": 3.8843134347489995e-05, "loss": 2.9022, "step": 3824 }, { "epoch": 0.0344016584747249, "grad_norm": 2.734928846359253, "learning_rate": 3.883687534363314e-05, "loss": 2.6113, "step": 3825 }, { "epoch": 0.03441065237236535, "grad_norm": 2.2386152744293213, "learning_rate": 3.883061508921439e-05, "loss": 2.5311, "step": 3826 }, { "epoch": 0.0344196462700058, "grad_norm": 1.2120344638824463, "learning_rate": 3.8824353584799555e-05, "loss": 2.6159, "step": 3827 }, { "epoch": 0.034428640167646254, "grad_norm": 1.0491162538528442, "learning_rate": 3.881809083095452e-05, "loss": 2.4251, "step": 3828 }, { "epoch": 0.0344376340652867, "grad_norm": 1.0927135944366455, "learning_rate": 3.8811826828245334e-05, "loss": 2.4579, "step": 3829 }, { "epoch": 0.034446627962927155, "grad_norm": 1.7338138818740845, "learning_rate": 3.88055615772381e-05, "loss": 2.4096, "step": 3830 }, { "epoch": 0.0344556218605676, "grad_norm": 1.7562298774719238, "learning_rate": 3.8799295078499096e-05, "loss": 2.559, "step": 3831 }, { "epoch": 0.034464615758208056, "grad_norm": 1.166852355003357, "learning_rate": 3.879302733259466e-05, "loss": 2.4343, "step": 3832 }, { "epoch": 0.03447360965584851, "grad_norm": 1.2102848291397095, "learning_rate": 3.8786758340091275e-05, "loss": 2.3425, "step": 3833 }, { "epoch": 0.03448260355348896, "grad_norm": 2.028939962387085, "learning_rate": 3.8780488101555516e-05, "loss": 2.7564, "step": 3834 }, { "epoch": 0.03449159745112941, "grad_norm": 1.6344228982925415, "learning_rate": 3.877421661755408e-05, "loss": 2.5322, "step": 3835 }, { "epoch": 0.03450059134876986, "grad_norm": 1.682499647140503, "learning_rate": 3.87679438886538e-05, "loss": 2.6216, "step": 3836 }, { "epoch": 0.03450958524641031, "grad_norm": 2.1760239601135254, "learning_rate": 3.876166991542156e-05, "loss": 2.639, "step": 3837 }, { "epoch": 0.034518579144050765, "grad_norm": 1.99851655960083, "learning_rate": 3.875539469842443e-05, "loss": 2.3263, "step": 3838 }, { "epoch": 0.03452757304169121, "grad_norm": 1.7676385641098022, "learning_rate": 3.874911823822954e-05, "loss": 2.5761, "step": 3839 }, { "epoch": 0.034536566939331666, "grad_norm": 1.6867426633834839, "learning_rate": 3.874284053540416e-05, "loss": 2.4851, "step": 3840 }, { "epoch": 0.03454556083697211, "grad_norm": 1.9702297449111938, "learning_rate": 3.8736561590515644e-05, "loss": 2.7933, "step": 3841 }, { "epoch": 0.034554554734612566, "grad_norm": 1.429770827293396, "learning_rate": 3.8730281404131486e-05, "loss": 2.485, "step": 3842 }, { "epoch": 0.03456354863225301, "grad_norm": 1.7759859561920166, "learning_rate": 3.872399997681929e-05, "loss": 2.7344, "step": 3843 }, { "epoch": 0.03457254252989347, "grad_norm": 1.8051917552947998, "learning_rate": 3.8717717309146755e-05, "loss": 2.525, "step": 3844 }, { "epoch": 0.03458153642753392, "grad_norm": 4.420403003692627, "learning_rate": 3.8711433401681706e-05, "loss": 2.365, "step": 3845 }, { "epoch": 0.03459053032517437, "grad_norm": 2.1113529205322266, "learning_rate": 3.8705148254992074e-05, "loss": 2.7846, "step": 3846 }, { "epoch": 0.03459952422281482, "grad_norm": 1.6200953722000122, "learning_rate": 3.869886186964591e-05, "loss": 2.4128, "step": 3847 }, { "epoch": 0.03460851812045527, "grad_norm": 2.0413262844085693, "learning_rate": 3.8692574246211355e-05, "loss": 2.5387, "step": 3848 }, { "epoch": 0.03461751201809572, "grad_norm": 1.9400897026062012, "learning_rate": 3.8686285385256695e-05, "loss": 2.7377, "step": 3849 }, { "epoch": 0.034626505915736176, "grad_norm": 2.07602596282959, "learning_rate": 3.8679995287350305e-05, "loss": 2.5417, "step": 3850 }, { "epoch": 0.03463549981337662, "grad_norm": 1.4244205951690674, "learning_rate": 3.867370395306068e-05, "loss": 2.4852, "step": 3851 }, { "epoch": 0.03464449371101708, "grad_norm": 2.268200635910034, "learning_rate": 3.866741138295641e-05, "loss": 2.6286, "step": 3852 }, { "epoch": 0.034653487608657524, "grad_norm": 1.4995604753494263, "learning_rate": 3.866111757760622e-05, "loss": 2.6617, "step": 3853 }, { "epoch": 0.03466248150629798, "grad_norm": 2.1164870262145996, "learning_rate": 3.865482253757894e-05, "loss": 2.6375, "step": 3854 }, { "epoch": 0.034671475403938425, "grad_norm": 1.523382544517517, "learning_rate": 3.8648526263443515e-05, "loss": 2.5738, "step": 3855 }, { "epoch": 0.03468046930157888, "grad_norm": 1.5741539001464844, "learning_rate": 3.864222875576899e-05, "loss": 2.5846, "step": 3856 }, { "epoch": 0.03468946319921933, "grad_norm": 1.9759364128112793, "learning_rate": 3.863593001512451e-05, "loss": 2.6433, "step": 3857 }, { "epoch": 0.03469845709685978, "grad_norm": 1.631495714187622, "learning_rate": 3.8629630042079366e-05, "loss": 2.6589, "step": 3858 }, { "epoch": 0.03470745099450023, "grad_norm": 1.8326332569122314, "learning_rate": 3.862332883720294e-05, "loss": 2.5591, "step": 3859 }, { "epoch": 0.03471644489214068, "grad_norm": 1.615007758140564, "learning_rate": 3.8617026401064724e-05, "loss": 2.5935, "step": 3860 }, { "epoch": 0.034725438789781134, "grad_norm": 1.8782062530517578, "learning_rate": 3.861072273423434e-05, "loss": 2.7231, "step": 3861 }, { "epoch": 0.03473443268742159, "grad_norm": 1.1915667057037354, "learning_rate": 3.860441783728149e-05, "loss": 2.4193, "step": 3862 }, { "epoch": 0.034743426585062034, "grad_norm": 1.6690047979354858, "learning_rate": 3.859811171077601e-05, "loss": 2.6014, "step": 3863 }, { "epoch": 0.03475242048270249, "grad_norm": 1.5591877698898315, "learning_rate": 3.859180435528784e-05, "loss": 2.5917, "step": 3864 }, { "epoch": 0.034761414380342935, "grad_norm": 2.1981186866760254, "learning_rate": 3.858549577138703e-05, "loss": 2.729, "step": 3865 }, { "epoch": 0.03477040827798339, "grad_norm": 1.3632287979125977, "learning_rate": 3.8579185959643745e-05, "loss": 2.3231, "step": 3866 }, { "epoch": 0.034779402175623836, "grad_norm": 1.5948312282562256, "learning_rate": 3.857287492062826e-05, "loss": 2.8963, "step": 3867 }, { "epoch": 0.03478839607326429, "grad_norm": 2.1141817569732666, "learning_rate": 3.856656265491097e-05, "loss": 2.5269, "step": 3868 }, { "epoch": 0.034797389970904744, "grad_norm": 2.057330369949341, "learning_rate": 3.856024916306234e-05, "loss": 2.8568, "step": 3869 }, { "epoch": 0.03480638386854519, "grad_norm": 1.6950689554214478, "learning_rate": 3.8553934445653005e-05, "loss": 2.4368, "step": 3870 }, { "epoch": 0.034815377766185644, "grad_norm": 1.4940918684005737, "learning_rate": 3.854761850325368e-05, "loss": 2.435, "step": 3871 }, { "epoch": 0.03482437166382609, "grad_norm": 1.6880420446395874, "learning_rate": 3.854130133643518e-05, "loss": 2.4816, "step": 3872 }, { "epoch": 0.034833365561466545, "grad_norm": 1.461518406867981, "learning_rate": 3.853498294576845e-05, "loss": 2.5196, "step": 3873 }, { "epoch": 0.034842359459107, "grad_norm": 2.253464698791504, "learning_rate": 3.852866333182454e-05, "loss": 2.8601, "step": 3874 }, { "epoch": 0.034851353356747446, "grad_norm": 1.5065401792526245, "learning_rate": 3.8522342495174624e-05, "loss": 2.71, "step": 3875 }, { "epoch": 0.0348603472543879, "grad_norm": 2.0962822437286377, "learning_rate": 3.851602043638994e-05, "loss": 2.7501, "step": 3876 }, { "epoch": 0.034869341152028346, "grad_norm": 1.858170509338379, "learning_rate": 3.8509697156041894e-05, "loss": 2.662, "step": 3877 }, { "epoch": 0.0348783350496688, "grad_norm": 1.485366702079773, "learning_rate": 3.850337265470197e-05, "loss": 2.6135, "step": 3878 }, { "epoch": 0.034887328947309254, "grad_norm": 1.6054610013961792, "learning_rate": 3.849704693294177e-05, "loss": 2.541, "step": 3879 }, { "epoch": 0.0348963228449497, "grad_norm": 1.9356967210769653, "learning_rate": 3.849071999133301e-05, "loss": 2.8238, "step": 3880 }, { "epoch": 0.034905316742590155, "grad_norm": 2.195230007171631, "learning_rate": 3.848439183044751e-05, "loss": 2.661, "step": 3881 }, { "epoch": 0.0349143106402306, "grad_norm": 1.8316466808319092, "learning_rate": 3.8478062450857196e-05, "loss": 2.5901, "step": 3882 }, { "epoch": 0.034923304537871055, "grad_norm": 1.6934648752212524, "learning_rate": 3.8471731853134117e-05, "loss": 2.4817, "step": 3883 }, { "epoch": 0.0349322984355115, "grad_norm": 1.5313518047332764, "learning_rate": 3.846540003785042e-05, "loss": 2.5051, "step": 3884 }, { "epoch": 0.034941292333151956, "grad_norm": 1.9450091123580933, "learning_rate": 3.845906700557837e-05, "loss": 2.7325, "step": 3885 }, { "epoch": 0.03495028623079241, "grad_norm": 1.3464125394821167, "learning_rate": 3.8452732756890345e-05, "loss": 2.3746, "step": 3886 }, { "epoch": 0.03495928012843286, "grad_norm": 1.5809268951416016, "learning_rate": 3.8446397292358826e-05, "loss": 2.8419, "step": 3887 }, { "epoch": 0.03496827402607331, "grad_norm": 1.642280101776123, "learning_rate": 3.8440060612556394e-05, "loss": 2.8574, "step": 3888 }, { "epoch": 0.03497726792371376, "grad_norm": 1.8861252069473267, "learning_rate": 3.843372271805577e-05, "loss": 2.5058, "step": 3889 }, { "epoch": 0.03498626182135421, "grad_norm": 1.7825548648834229, "learning_rate": 3.8427383609429754e-05, "loss": 2.8508, "step": 3890 }, { "epoch": 0.034995255718994665, "grad_norm": 1.6319572925567627, "learning_rate": 3.842104328725127e-05, "loss": 2.4633, "step": 3891 }, { "epoch": 0.03500424961663511, "grad_norm": 1.6886553764343262, "learning_rate": 3.841470175209335e-05, "loss": 2.7815, "step": 3892 }, { "epoch": 0.035013243514275566, "grad_norm": 1.6851814985275269, "learning_rate": 3.840835900452912e-05, "loss": 2.6989, "step": 3893 }, { "epoch": 0.03502223741191601, "grad_norm": 1.7669893503189087, "learning_rate": 3.8402015045131864e-05, "loss": 2.6625, "step": 3894 }, { "epoch": 0.03503123130955647, "grad_norm": 1.5724925994873047, "learning_rate": 3.8395669874474915e-05, "loss": 2.7612, "step": 3895 }, { "epoch": 0.035040225207196914, "grad_norm": 1.4397239685058594, "learning_rate": 3.838932349313176e-05, "loss": 2.4731, "step": 3896 }, { "epoch": 0.03504921910483737, "grad_norm": 1.5780889987945557, "learning_rate": 3.838297590167596e-05, "loss": 2.5597, "step": 3897 }, { "epoch": 0.03505821300247782, "grad_norm": 1.708639144897461, "learning_rate": 3.837662710068122e-05, "loss": 2.9167, "step": 3898 }, { "epoch": 0.03506720690011827, "grad_norm": 0.8521286249160767, "learning_rate": 3.8370277090721326e-05, "loss": 2.2348, "step": 3899 }, { "epoch": 0.03507620079775872, "grad_norm": 1.723006010055542, "learning_rate": 3.8363925872370185e-05, "loss": 2.5022, "step": 3900 }, { "epoch": 0.03508519469539917, "grad_norm": 1.7737760543823242, "learning_rate": 3.8357573446201825e-05, "loss": 2.5942, "step": 3901 }, { "epoch": 0.03509418859303962, "grad_norm": 2.2909297943115234, "learning_rate": 3.8351219812790365e-05, "loss": 2.6121, "step": 3902 }, { "epoch": 0.03510318249068008, "grad_norm": 1.132799744606018, "learning_rate": 3.8344864972710046e-05, "loss": 2.4527, "step": 3903 }, { "epoch": 0.03511217638832052, "grad_norm": 2.065223217010498, "learning_rate": 3.833850892653519e-05, "loss": 2.6102, "step": 3904 }, { "epoch": 0.03512117028596098, "grad_norm": 1.8666112422943115, "learning_rate": 3.8332151674840275e-05, "loss": 2.6474, "step": 3905 }, { "epoch": 0.035130164183601424, "grad_norm": 1.326658010482788, "learning_rate": 3.832579321819985e-05, "loss": 2.5063, "step": 3906 }, { "epoch": 0.03513915808124188, "grad_norm": 1.9494609832763672, "learning_rate": 3.831943355718858e-05, "loss": 2.7836, "step": 3907 }, { "epoch": 0.035148151978882325, "grad_norm": 2.1442699432373047, "learning_rate": 3.831307269238126e-05, "loss": 2.4317, "step": 3908 }, { "epoch": 0.03515714587652278, "grad_norm": 1.8793662786483765, "learning_rate": 3.830671062435276e-05, "loss": 2.3412, "step": 3909 }, { "epoch": 0.03516613977416323, "grad_norm": 1.7804378271102905, "learning_rate": 3.8300347353678104e-05, "loss": 2.5329, "step": 3910 }, { "epoch": 0.03517513367180368, "grad_norm": 2.306265354156494, "learning_rate": 3.8293982880932365e-05, "loss": 2.951, "step": 3911 }, { "epoch": 0.03518412756944413, "grad_norm": 1.8697285652160645, "learning_rate": 3.828761720669077e-05, "loss": 2.5736, "step": 3912 }, { "epoch": 0.03519312146708458, "grad_norm": 1.7361235618591309, "learning_rate": 3.8281250331528654e-05, "loss": 2.7868, "step": 3913 }, { "epoch": 0.035202115364725034, "grad_norm": 2.20723032951355, "learning_rate": 3.827488225602144e-05, "loss": 2.6723, "step": 3914 }, { "epoch": 0.03521110926236549, "grad_norm": 1.6256860494613647, "learning_rate": 3.8268512980744654e-05, "loss": 2.6061, "step": 3915 }, { "epoch": 0.035220103160005935, "grad_norm": 1.695840835571289, "learning_rate": 3.826214250627397e-05, "loss": 2.7202, "step": 3916 }, { "epoch": 0.03522909705764639, "grad_norm": 1.5654244422912598, "learning_rate": 3.825577083318512e-05, "loss": 2.673, "step": 3917 }, { "epoch": 0.035238090955286835, "grad_norm": 2.016876697540283, "learning_rate": 3.824939796205399e-05, "loss": 2.6118, "step": 3918 }, { "epoch": 0.03524708485292729, "grad_norm": 1.3711811304092407, "learning_rate": 3.8243023893456535e-05, "loss": 2.5986, "step": 3919 }, { "epoch": 0.03525607875056774, "grad_norm": 2.200509548187256, "learning_rate": 3.8236648627968845e-05, "loss": 2.794, "step": 3920 }, { "epoch": 0.03526507264820819, "grad_norm": 1.6370943784713745, "learning_rate": 3.823027216616711e-05, "loss": 2.5932, "step": 3921 }, { "epoch": 0.035274066545848644, "grad_norm": 1.639208197593689, "learning_rate": 3.8223894508627634e-05, "loss": 2.7175, "step": 3922 }, { "epoch": 0.03528306044348909, "grad_norm": 1.8653144836425781, "learning_rate": 3.821751565592681e-05, "loss": 2.7628, "step": 3923 }, { "epoch": 0.035292054341129545, "grad_norm": 1.5254727602005005, "learning_rate": 3.8211135608641155e-05, "loss": 2.6358, "step": 3924 }, { "epoch": 0.03530104823876999, "grad_norm": 1.2655279636383057, "learning_rate": 3.8204754367347305e-05, "loss": 2.5887, "step": 3925 }, { "epoch": 0.035310042136410445, "grad_norm": 1.2939910888671875, "learning_rate": 3.8198371932621965e-05, "loss": 2.4298, "step": 3926 }, { "epoch": 0.0353190360340509, "grad_norm": 2.161081552505493, "learning_rate": 3.8191988305042e-05, "loss": 2.5974, "step": 3927 }, { "epoch": 0.035328029931691346, "grad_norm": 1.5839754343032837, "learning_rate": 3.818560348518433e-05, "loss": 2.6794, "step": 3928 }, { "epoch": 0.0353370238293318, "grad_norm": 1.494784951210022, "learning_rate": 3.8179217473626016e-05, "loss": 2.4829, "step": 3929 }, { "epoch": 0.03534601772697225, "grad_norm": 1.567044734954834, "learning_rate": 3.817283027094423e-05, "loss": 2.506, "step": 3930 }, { "epoch": 0.0353550116246127, "grad_norm": 1.396832823753357, "learning_rate": 3.8166441877716234e-05, "loss": 2.469, "step": 3931 }, { "epoch": 0.035364005522253154, "grad_norm": 1.8299264907836914, "learning_rate": 3.8160052294519396e-05, "loss": 2.6279, "step": 3932 }, { "epoch": 0.0353729994198936, "grad_norm": 1.4904509782791138, "learning_rate": 3.815366152193122e-05, "loss": 2.5309, "step": 3933 }, { "epoch": 0.035381993317534055, "grad_norm": 1.536381721496582, "learning_rate": 3.814726956052927e-05, "loss": 2.3388, "step": 3934 }, { "epoch": 0.0353909872151745, "grad_norm": 1.244486689567566, "learning_rate": 3.814087641089126e-05, "loss": 2.4803, "step": 3935 }, { "epoch": 0.035399981112814956, "grad_norm": 2.0747857093811035, "learning_rate": 3.8134482073594996e-05, "loss": 2.8542, "step": 3936 }, { "epoch": 0.0354089750104554, "grad_norm": 1.7315131425857544, "learning_rate": 3.812808654921839e-05, "loss": 2.7047, "step": 3937 }, { "epoch": 0.035417968908095857, "grad_norm": 1.5149033069610596, "learning_rate": 3.812168983833946e-05, "loss": 2.7371, "step": 3938 }, { "epoch": 0.03542696280573631, "grad_norm": 0.9867162704467773, "learning_rate": 3.8115291941536345e-05, "loss": 2.2698, "step": 3939 }, { "epoch": 0.03543595670337676, "grad_norm": 1.6082206964492798, "learning_rate": 3.810889285938726e-05, "loss": 2.8659, "step": 3940 }, { "epoch": 0.03544495060101721, "grad_norm": 1.6938294172286987, "learning_rate": 3.810249259247056e-05, "loss": 2.8551, "step": 3941 }, { "epoch": 0.03545394449865766, "grad_norm": 1.7532322406768799, "learning_rate": 3.80960911413647e-05, "loss": 2.4738, "step": 3942 }, { "epoch": 0.03546293839629811, "grad_norm": 1.9644179344177246, "learning_rate": 3.808968850664821e-05, "loss": 2.595, "step": 3943 }, { "epoch": 0.035471932293938566, "grad_norm": 1.488019347190857, "learning_rate": 3.808328468889979e-05, "loss": 2.5625, "step": 3944 }, { "epoch": 0.03548092619157901, "grad_norm": 1.5021404027938843, "learning_rate": 3.807687968869819e-05, "loss": 2.6566, "step": 3945 }, { "epoch": 0.035489920089219466, "grad_norm": 1.7135361433029175, "learning_rate": 3.8070473506622285e-05, "loss": 2.4953, "step": 3946 }, { "epoch": 0.03549891398685991, "grad_norm": 1.4986759424209595, "learning_rate": 3.8064066143251067e-05, "loss": 2.43, "step": 3947 }, { "epoch": 0.03550790788450037, "grad_norm": 1.4706056118011475, "learning_rate": 3.805765759916362e-05, "loss": 2.6096, "step": 3948 }, { "epoch": 0.03551690178214082, "grad_norm": 1.6926414966583252, "learning_rate": 3.805124787493913e-05, "loss": 2.5765, "step": 3949 }, { "epoch": 0.03552589567978127, "grad_norm": 1.020189642906189, "learning_rate": 3.8044836971156936e-05, "loss": 2.3719, "step": 3950 }, { "epoch": 0.03553488957742172, "grad_norm": 1.2501990795135498, "learning_rate": 3.8038424888396416e-05, "loss": 2.2974, "step": 3951 }, { "epoch": 0.03554388347506217, "grad_norm": 1.3079867362976074, "learning_rate": 3.80320116272371e-05, "loss": 2.3562, "step": 3952 }, { "epoch": 0.03555287737270262, "grad_norm": 1.9358408451080322, "learning_rate": 3.802559718825862e-05, "loss": 2.8102, "step": 3953 }, { "epoch": 0.03556187127034307, "grad_norm": 1.8057698011398315, "learning_rate": 3.8019181572040684e-05, "loss": 2.5991, "step": 3954 }, { "epoch": 0.03557086516798352, "grad_norm": 1.7238205671310425, "learning_rate": 3.801276477916314e-05, "loss": 2.8221, "step": 3955 }, { "epoch": 0.03557985906562398, "grad_norm": 1.0883442163467407, "learning_rate": 3.8006346810205935e-05, "loss": 2.3182, "step": 3956 }, { "epoch": 0.035588852963264424, "grad_norm": 2.09382700920105, "learning_rate": 3.7999927665749115e-05, "loss": 2.7966, "step": 3957 }, { "epoch": 0.03559784686090488, "grad_norm": 1.6835544109344482, "learning_rate": 3.799350734637283e-05, "loss": 2.8195, "step": 3958 }, { "epoch": 0.035606840758545324, "grad_norm": 2.779787540435791, "learning_rate": 3.7987085852657354e-05, "loss": 2.7463, "step": 3959 }, { "epoch": 0.03561583465618578, "grad_norm": 1.6887489557266235, "learning_rate": 3.798066318518305e-05, "loss": 2.4041, "step": 3960 }, { "epoch": 0.03562482855382623, "grad_norm": 0.8220937252044678, "learning_rate": 3.797423934453038e-05, "loss": 2.2052, "step": 3961 }, { "epoch": 0.03563382245146668, "grad_norm": 1.6060198545455933, "learning_rate": 3.7967814331279936e-05, "loss": 2.4687, "step": 3962 }, { "epoch": 0.03564281634910713, "grad_norm": 1.652868628501892, "learning_rate": 3.79613881460124e-05, "loss": 2.5785, "step": 3963 }, { "epoch": 0.03565181024674758, "grad_norm": 1.6599810123443604, "learning_rate": 3.795496078930857e-05, "loss": 2.5716, "step": 3964 }, { "epoch": 0.035660804144388034, "grad_norm": 1.1881047487258911, "learning_rate": 3.794853226174933e-05, "loss": 2.2826, "step": 3965 }, { "epoch": 0.03566979804202848, "grad_norm": 2.086540937423706, "learning_rate": 3.794210256391569e-05, "loss": 2.5935, "step": 3966 }, { "epoch": 0.035678791939668934, "grad_norm": 1.8542962074279785, "learning_rate": 3.793567169638877e-05, "loss": 2.5869, "step": 3967 }, { "epoch": 0.03568778583730939, "grad_norm": 1.8763463497161865, "learning_rate": 3.7929239659749764e-05, "loss": 2.6939, "step": 3968 }, { "epoch": 0.035696779734949835, "grad_norm": 1.8982267379760742, "learning_rate": 3.7922806454580016e-05, "loss": 2.7375, "step": 3969 }, { "epoch": 0.03570577363259029, "grad_norm": 1.299750566482544, "learning_rate": 3.7916372081460926e-05, "loss": 2.458, "step": 3970 }, { "epoch": 0.035714767530230736, "grad_norm": 2.212277889251709, "learning_rate": 3.790993654097405e-05, "loss": 2.6269, "step": 3971 }, { "epoch": 0.03572376142787119, "grad_norm": 1.6694337129592896, "learning_rate": 3.7903499833701006e-05, "loss": 2.728, "step": 3972 }, { "epoch": 0.03573275532551164, "grad_norm": 1.4249786138534546, "learning_rate": 3.789706196022356e-05, "loss": 2.5879, "step": 3973 }, { "epoch": 0.03574174922315209, "grad_norm": 1.7112280130386353, "learning_rate": 3.789062292112353e-05, "loss": 2.8498, "step": 3974 }, { "epoch": 0.035750743120792544, "grad_norm": 1.7559258937835693, "learning_rate": 3.78841827169829e-05, "loss": 2.5612, "step": 3975 }, { "epoch": 0.03575973701843299, "grad_norm": 1.4105560779571533, "learning_rate": 3.78777413483837e-05, "loss": 2.4915, "step": 3976 }, { "epoch": 0.035768730916073445, "grad_norm": 1.6962757110595703, "learning_rate": 3.787129881590812e-05, "loss": 2.4906, "step": 3977 }, { "epoch": 0.03577772481371389, "grad_norm": 1.51368248462677, "learning_rate": 3.7864855120138404e-05, "loss": 2.4636, "step": 3978 }, { "epoch": 0.035786718711354346, "grad_norm": 1.522635817527771, "learning_rate": 3.785841026165694e-05, "loss": 2.5469, "step": 3979 }, { "epoch": 0.0357957126089948, "grad_norm": 1.8758580684661865, "learning_rate": 3.785196424104622e-05, "loss": 2.8156, "step": 3980 }, { "epoch": 0.035804706506635246, "grad_norm": 2.1665091514587402, "learning_rate": 3.784551705888881e-05, "loss": 2.6329, "step": 3981 }, { "epoch": 0.0358137004042757, "grad_norm": 1.4234635829925537, "learning_rate": 3.78390687157674e-05, "loss": 2.6155, "step": 3982 }, { "epoch": 0.03582269430191615, "grad_norm": 1.3771456480026245, "learning_rate": 3.783261921226479e-05, "loss": 2.3836, "step": 3983 }, { "epoch": 0.0358316881995566, "grad_norm": 1.4243730306625366, "learning_rate": 3.782616854896387e-05, "loss": 2.5016, "step": 3984 }, { "epoch": 0.035840682097197055, "grad_norm": 1.8237055540084839, "learning_rate": 3.781971672644766e-05, "loss": 2.5527, "step": 3985 }, { "epoch": 0.0358496759948375, "grad_norm": 1.8247952461242676, "learning_rate": 3.7813263745299254e-05, "loss": 2.6595, "step": 3986 }, { "epoch": 0.035858669892477955, "grad_norm": 1.623121738433838, "learning_rate": 3.780680960610188e-05, "loss": 2.5388, "step": 3987 }, { "epoch": 0.0358676637901184, "grad_norm": 1.4101901054382324, "learning_rate": 3.7800354309438855e-05, "loss": 2.5394, "step": 3988 }, { "epoch": 0.035876657687758856, "grad_norm": 1.4663554430007935, "learning_rate": 3.779389785589358e-05, "loss": 2.5181, "step": 3989 }, { "epoch": 0.03588565158539931, "grad_norm": 1.6531736850738525, "learning_rate": 3.7787440246049606e-05, "loss": 2.7278, "step": 3990 }, { "epoch": 0.03589464548303976, "grad_norm": 2.1047871112823486, "learning_rate": 3.778098148049055e-05, "loss": 2.567, "step": 3991 }, { "epoch": 0.03590363938068021, "grad_norm": 1.8902673721313477, "learning_rate": 3.777452155980016e-05, "loss": 2.7586, "step": 3992 }, { "epoch": 0.03591263327832066, "grad_norm": 1.7767066955566406, "learning_rate": 3.7768060484562266e-05, "loss": 2.6277, "step": 3993 }, { "epoch": 0.03592162717596111, "grad_norm": 1.826487421989441, "learning_rate": 3.776159825536082e-05, "loss": 2.7717, "step": 3994 }, { "epoch": 0.03593062107360156, "grad_norm": 1.7087175846099854, "learning_rate": 3.7755134872779876e-05, "loss": 2.5912, "step": 3995 }, { "epoch": 0.03593961497124201, "grad_norm": 1.1982694864273071, "learning_rate": 3.774867033740357e-05, "loss": 2.4847, "step": 3996 }, { "epoch": 0.035948608868882466, "grad_norm": 1.9269697666168213, "learning_rate": 3.774220464981617e-05, "loss": 2.7287, "step": 3997 }, { "epoch": 0.03595760276652291, "grad_norm": 2.528010606765747, "learning_rate": 3.773573781060205e-05, "loss": 2.8895, "step": 3998 }, { "epoch": 0.03596659666416337, "grad_norm": 1.885959506034851, "learning_rate": 3.7729269820345656e-05, "loss": 2.5339, "step": 3999 }, { "epoch": 0.035975590561803814, "grad_norm": 1.432624101638794, "learning_rate": 3.772280067963157e-05, "loss": 2.6634, "step": 4000 }, { "epoch": 0.03598458445944427, "grad_norm": 1.3743621110916138, "learning_rate": 3.771633038904446e-05, "loss": 2.4497, "step": 4001 }, { "epoch": 0.03599357835708472, "grad_norm": 1.1649539470672607, "learning_rate": 3.770985894916911e-05, "loss": 2.1873, "step": 4002 }, { "epoch": 0.03600257225472517, "grad_norm": 1.6138865947723389, "learning_rate": 3.770338636059039e-05, "loss": 2.4367, "step": 4003 }, { "epoch": 0.03601156615236562, "grad_norm": 1.4940006732940674, "learning_rate": 3.7696912623893306e-05, "loss": 2.6396, "step": 4004 }, { "epoch": 0.03602056005000607, "grad_norm": 2.0938210487365723, "learning_rate": 3.7690437739662924e-05, "loss": 2.9243, "step": 4005 }, { "epoch": 0.03602955394764652, "grad_norm": 1.769730567932129, "learning_rate": 3.768396170848445e-05, "loss": 2.389, "step": 4006 }, { "epoch": 0.03603854784528697, "grad_norm": 1.319413423538208, "learning_rate": 3.7677484530943185e-05, "loss": 2.3946, "step": 4007 }, { "epoch": 0.03604754174292742, "grad_norm": 2.0449068546295166, "learning_rate": 3.767100620762452e-05, "loss": 2.8691, "step": 4008 }, { "epoch": 0.03605653564056788, "grad_norm": 1.9554694890975952, "learning_rate": 3.766452673911396e-05, "loss": 2.617, "step": 4009 }, { "epoch": 0.036065529538208324, "grad_norm": 2.0682313442230225, "learning_rate": 3.765804612599712e-05, "loss": 2.5595, "step": 4010 }, { "epoch": 0.03607452343584878, "grad_norm": 1.355016827583313, "learning_rate": 3.76515643688597e-05, "loss": 2.3904, "step": 4011 }, { "epoch": 0.036083517333489225, "grad_norm": 2.199749231338501, "learning_rate": 3.7645081468287516e-05, "loss": 2.8973, "step": 4012 }, { "epoch": 0.03609251123112968, "grad_norm": 2.007657527923584, "learning_rate": 3.7638597424866496e-05, "loss": 2.8144, "step": 4013 }, { "epoch": 0.03610150512877013, "grad_norm": 3.0730481147766113, "learning_rate": 3.763211223918265e-05, "loss": 2.6557, "step": 4014 }, { "epoch": 0.03611049902641058, "grad_norm": 2.644803524017334, "learning_rate": 3.762562591182211e-05, "loss": 2.3854, "step": 4015 }, { "epoch": 0.03611949292405103, "grad_norm": 1.7670005559921265, "learning_rate": 3.76191384433711e-05, "loss": 2.5255, "step": 4016 }, { "epoch": 0.03612848682169148, "grad_norm": 1.9304391145706177, "learning_rate": 3.7612649834415945e-05, "loss": 2.3179, "step": 4017 }, { "epoch": 0.036137480719331934, "grad_norm": 1.6926542520523071, "learning_rate": 3.760616008554309e-05, "loss": 2.5747, "step": 4018 }, { "epoch": 0.03614647461697238, "grad_norm": 1.543218731880188, "learning_rate": 3.7599669197339046e-05, "loss": 2.5255, "step": 4019 }, { "epoch": 0.036155468514612835, "grad_norm": 2.283803939819336, "learning_rate": 3.7593177170390484e-05, "loss": 2.8969, "step": 4020 }, { "epoch": 0.03616446241225329, "grad_norm": 1.8534523248672485, "learning_rate": 3.7586684005284144e-05, "loss": 2.4367, "step": 4021 }, { "epoch": 0.036173456309893735, "grad_norm": 2.2874274253845215, "learning_rate": 3.758018970260685e-05, "loss": 2.7699, "step": 4022 }, { "epoch": 0.03618245020753419, "grad_norm": 1.625935435295105, "learning_rate": 3.7573694262945555e-05, "loss": 2.5963, "step": 4023 }, { "epoch": 0.036191444105174636, "grad_norm": 1.4244784116744995, "learning_rate": 3.756719768688733e-05, "loss": 2.4079, "step": 4024 }, { "epoch": 0.03620043800281509, "grad_norm": 2.1171045303344727, "learning_rate": 3.7560699975019305e-05, "loss": 2.7405, "step": 4025 }, { "epoch": 0.036209431900455544, "grad_norm": 2.8666746616363525, "learning_rate": 3.7554201127928744e-05, "loss": 3.0037, "step": 4026 }, { "epoch": 0.03621842579809599, "grad_norm": 1.9545055627822876, "learning_rate": 3.7547701146203005e-05, "loss": 2.7076, "step": 4027 }, { "epoch": 0.036227419695736444, "grad_norm": 1.8545596599578857, "learning_rate": 3.754120003042957e-05, "loss": 2.6929, "step": 4028 }, { "epoch": 0.03623641359337689, "grad_norm": 2.066159248352051, "learning_rate": 3.753469778119597e-05, "loss": 2.8501, "step": 4029 }, { "epoch": 0.036245407491017345, "grad_norm": 1.427412748336792, "learning_rate": 3.752819439908989e-05, "loss": 2.3595, "step": 4030 }, { "epoch": 0.0362544013886578, "grad_norm": 1.6074880361557007, "learning_rate": 3.752168988469909e-05, "loss": 2.6297, "step": 4031 }, { "epoch": 0.036263395286298246, "grad_norm": 1.9218696355819702, "learning_rate": 3.7515184238611444e-05, "loss": 2.7563, "step": 4032 }, { "epoch": 0.0362723891839387, "grad_norm": 1.2809534072875977, "learning_rate": 3.750867746141493e-05, "loss": 2.3596, "step": 4033 }, { "epoch": 0.03628138308157915, "grad_norm": 1.4224462509155273, "learning_rate": 3.750216955369763e-05, "loss": 2.6259, "step": 4034 }, { "epoch": 0.0362903769792196, "grad_norm": 1.8553327322006226, "learning_rate": 3.74956605160477e-05, "loss": 2.8105, "step": 4035 }, { "epoch": 0.03629937087686005, "grad_norm": 2.743291139602661, "learning_rate": 3.748915034905344e-05, "loss": 2.8409, "step": 4036 }, { "epoch": 0.0363083647745005, "grad_norm": 1.8983391523361206, "learning_rate": 3.7482639053303213e-05, "loss": 2.6472, "step": 4037 }, { "epoch": 0.036317358672140955, "grad_norm": 1.2686783075332642, "learning_rate": 3.747612662938552e-05, "loss": 2.4268, "step": 4038 }, { "epoch": 0.0363263525697814, "grad_norm": 1.7441028356552124, "learning_rate": 3.7469613077888935e-05, "loss": 2.8538, "step": 4039 }, { "epoch": 0.036335346467421856, "grad_norm": 1.9671118259429932, "learning_rate": 3.7463098399402155e-05, "loss": 2.761, "step": 4040 }, { "epoch": 0.0363443403650623, "grad_norm": 3.5343072414398193, "learning_rate": 3.745658259451397e-05, "loss": 2.6099, "step": 4041 }, { "epoch": 0.036353334262702756, "grad_norm": 1.5958629846572876, "learning_rate": 3.745006566381326e-05, "loss": 2.552, "step": 4042 }, { "epoch": 0.03636232816034321, "grad_norm": 1.7036460638046265, "learning_rate": 3.744354760788903e-05, "loss": 2.5941, "step": 4043 }, { "epoch": 0.03637132205798366, "grad_norm": 1.691109299659729, "learning_rate": 3.743702842733037e-05, "loss": 2.4759, "step": 4044 }, { "epoch": 0.03638031595562411, "grad_norm": 1.2308858633041382, "learning_rate": 3.743050812272648e-05, "loss": 2.3478, "step": 4045 }, { "epoch": 0.03638930985326456, "grad_norm": 1.94650399684906, "learning_rate": 3.742398669466665e-05, "loss": 2.7348, "step": 4046 }, { "epoch": 0.03639830375090501, "grad_norm": 1.5214192867279053, "learning_rate": 3.7417464143740285e-05, "loss": 2.9086, "step": 4047 }, { "epoch": 0.03640729764854546, "grad_norm": 1.6351613998413086, "learning_rate": 3.741094047053688e-05, "loss": 2.8157, "step": 4048 }, { "epoch": 0.03641629154618591, "grad_norm": 1.7503023147583008, "learning_rate": 3.7404415675646054e-05, "loss": 2.7555, "step": 4049 }, { "epoch": 0.036425285443826366, "grad_norm": 1.8643602132797241, "learning_rate": 3.7397889759657494e-05, "loss": 2.5389, "step": 4050 }, { "epoch": 0.03643427934146681, "grad_norm": 2.19431209564209, "learning_rate": 3.739136272316102e-05, "loss": 2.8508, "step": 4051 }, { "epoch": 0.03644327323910727, "grad_norm": 1.7215909957885742, "learning_rate": 3.738483456674652e-05, "loss": 2.515, "step": 4052 }, { "epoch": 0.036452267136747714, "grad_norm": 1.6797105073928833, "learning_rate": 3.737830529100401e-05, "loss": 2.6595, "step": 4053 }, { "epoch": 0.03646126103438817, "grad_norm": 1.6473201513290405, "learning_rate": 3.73717748965236e-05, "loss": 2.5623, "step": 4054 }, { "epoch": 0.03647025493202862, "grad_norm": 1.972875952720642, "learning_rate": 3.736524338389551e-05, "loss": 2.956, "step": 4055 }, { "epoch": 0.03647924882966907, "grad_norm": 2.373939037322998, "learning_rate": 3.735871075371004e-05, "loss": 3.0774, "step": 4056 }, { "epoch": 0.03648824272730952, "grad_norm": 1.0875765085220337, "learning_rate": 3.73521770065576e-05, "loss": 2.6266, "step": 4057 }, { "epoch": 0.03649723662494997, "grad_norm": 1.4054456949234009, "learning_rate": 3.7345642143028715e-05, "loss": 2.5411, "step": 4058 }, { "epoch": 0.03650623052259042, "grad_norm": 1.2090612649917603, "learning_rate": 3.7339106163713985e-05, "loss": 2.3178, "step": 4059 }, { "epoch": 0.03651522442023088, "grad_norm": 1.2831110954284668, "learning_rate": 3.733256906920413e-05, "loss": 2.4904, "step": 4060 }, { "epoch": 0.036524218317871324, "grad_norm": 1.9785423278808594, "learning_rate": 3.732603086008996e-05, "loss": 2.6778, "step": 4061 }, { "epoch": 0.03653321221551178, "grad_norm": 1.5432411432266235, "learning_rate": 3.73194915369624e-05, "loss": 2.4493, "step": 4062 }, { "epoch": 0.036542206113152224, "grad_norm": 1.8665194511413574, "learning_rate": 3.731295110041247e-05, "loss": 2.6372, "step": 4063 }, { "epoch": 0.03655120001079268, "grad_norm": 2.1811342239379883, "learning_rate": 3.7306409551031284e-05, "loss": 2.9368, "step": 4064 }, { "epoch": 0.036560193908433125, "grad_norm": 1.6699353456497192, "learning_rate": 3.729986688941005e-05, "loss": 2.5691, "step": 4065 }, { "epoch": 0.03656918780607358, "grad_norm": 1.9747369289398193, "learning_rate": 3.72933231161401e-05, "loss": 2.6233, "step": 4066 }, { "epoch": 0.03657818170371403, "grad_norm": 2.062697649002075, "learning_rate": 3.728677823181284e-05, "loss": 2.9062, "step": 4067 }, { "epoch": 0.03658717560135448, "grad_norm": 1.6749913692474365, "learning_rate": 3.7280232237019805e-05, "loss": 2.4827, "step": 4068 }, { "epoch": 0.036596169498994933, "grad_norm": 3.639761447906494, "learning_rate": 3.727368513235261e-05, "loss": 2.5164, "step": 4069 }, { "epoch": 0.03660516339663538, "grad_norm": 2.0140538215637207, "learning_rate": 3.726713691840297e-05, "loss": 2.8498, "step": 4070 }, { "epoch": 0.036614157294275834, "grad_norm": 1.7058147192001343, "learning_rate": 3.726058759576271e-05, "loss": 2.991, "step": 4071 }, { "epoch": 0.03662315119191629, "grad_norm": 1.702327847480774, "learning_rate": 3.725403716502375e-05, "loss": 2.5956, "step": 4072 }, { "epoch": 0.036632145089556735, "grad_norm": 1.4156453609466553, "learning_rate": 3.72474856267781e-05, "loss": 2.5563, "step": 4073 }, { "epoch": 0.03664113898719719, "grad_norm": 1.4307552576065063, "learning_rate": 3.7240932981617896e-05, "loss": 2.4745, "step": 4074 }, { "epoch": 0.036650132884837636, "grad_norm": 1.1816962957382202, "learning_rate": 3.7234379230135355e-05, "loss": 2.3692, "step": 4075 }, { "epoch": 0.03665912678247809, "grad_norm": 2.2284088134765625, "learning_rate": 3.72278243729228e-05, "loss": 2.4147, "step": 4076 }, { "epoch": 0.036668120680118536, "grad_norm": 1.9954692125320435, "learning_rate": 3.7221268410572646e-05, "loss": 2.5132, "step": 4077 }, { "epoch": 0.03667711457775899, "grad_norm": 2.214566946029663, "learning_rate": 3.721471134367742e-05, "loss": 3.0952, "step": 4078 }, { "epoch": 0.036686108475399444, "grad_norm": 1.5208942890167236, "learning_rate": 3.720815317282973e-05, "loss": 2.6538, "step": 4079 }, { "epoch": 0.03669510237303989, "grad_norm": 2.009974479675293, "learning_rate": 3.7201593898622316e-05, "loss": 2.9122, "step": 4080 }, { "epoch": 0.036704096270680345, "grad_norm": 1.0230293273925781, "learning_rate": 3.719503352164799e-05, "loss": 2.3577, "step": 4081 }, { "epoch": 0.03671309016832079, "grad_norm": 1.600180745124817, "learning_rate": 3.718847204249966e-05, "loss": 2.4774, "step": 4082 }, { "epoch": 0.036722084065961245, "grad_norm": 1.630920171737671, "learning_rate": 3.718190946177036e-05, "loss": 2.6603, "step": 4083 }, { "epoch": 0.0367310779636017, "grad_norm": 2.1396260261535645, "learning_rate": 3.71753457800532e-05, "loss": 2.8348, "step": 4084 }, { "epoch": 0.036740071861242146, "grad_norm": 2.4375789165496826, "learning_rate": 3.716878099794141e-05, "loss": 2.3906, "step": 4085 }, { "epoch": 0.0367490657588826, "grad_norm": 1.6788620948791504, "learning_rate": 3.71622151160283e-05, "loss": 2.8422, "step": 4086 }, { "epoch": 0.03675805965652305, "grad_norm": 1.9501675367355347, "learning_rate": 3.715564813490728e-05, "loss": 2.6057, "step": 4087 }, { "epoch": 0.0367670535541635, "grad_norm": 1.597767949104309, "learning_rate": 3.714908005517188e-05, "loss": 2.6799, "step": 4088 }, { "epoch": 0.03677604745180395, "grad_norm": 1.3483024835586548, "learning_rate": 3.714251087741571e-05, "loss": 2.5193, "step": 4089 }, { "epoch": 0.0367850413494444, "grad_norm": 1.8438688516616821, "learning_rate": 3.7135940602232484e-05, "loss": 2.6838, "step": 4090 }, { "epoch": 0.036794035247084855, "grad_norm": 1.6526838541030884, "learning_rate": 3.712936923021602e-05, "loss": 2.7763, "step": 4091 }, { "epoch": 0.0368030291447253, "grad_norm": 1.9978703260421753, "learning_rate": 3.712279676196023e-05, "loss": 2.5823, "step": 4092 }, { "epoch": 0.036812023042365756, "grad_norm": 2.5050573348999023, "learning_rate": 3.711622319805913e-05, "loss": 3.0019, "step": 4093 }, { "epoch": 0.0368210169400062, "grad_norm": 1.944429636001587, "learning_rate": 3.710964853910682e-05, "loss": 2.7912, "step": 4094 }, { "epoch": 0.03683001083764666, "grad_norm": 1.8593462705612183, "learning_rate": 3.710307278569752e-05, "loss": 2.7686, "step": 4095 }, { "epoch": 0.03683900473528711, "grad_norm": 1.949704885482788, "learning_rate": 3.709649593842554e-05, "loss": 2.8324, "step": 4096 }, { "epoch": 0.03684799863292756, "grad_norm": 1.7294214963912964, "learning_rate": 3.708991799788528e-05, "loss": 2.5913, "step": 4097 }, { "epoch": 0.03685699253056801, "grad_norm": 1.6455267667770386, "learning_rate": 3.708333896467126e-05, "loss": 2.5988, "step": 4098 }, { "epoch": 0.03686598642820846, "grad_norm": 2.0253350734710693, "learning_rate": 3.707675883937809e-05, "loss": 2.65, "step": 4099 }, { "epoch": 0.03687498032584891, "grad_norm": 1.2221102714538574, "learning_rate": 3.707017762260045e-05, "loss": 2.3617, "step": 4100 }, { "epoch": 0.036883974223489366, "grad_norm": 2.509685754776001, "learning_rate": 3.706359531493316e-05, "loss": 2.9686, "step": 4101 }, { "epoch": 0.03689296812112981, "grad_norm": 1.6992993354797363, "learning_rate": 3.705701191697112e-05, "loss": 2.3959, "step": 4102 }, { "epoch": 0.03690196201877027, "grad_norm": 1.2788963317871094, "learning_rate": 3.7050427429309326e-05, "loss": 2.5163, "step": 4103 }, { "epoch": 0.03691095591641071, "grad_norm": 1.4854936599731445, "learning_rate": 3.704384185254288e-05, "loss": 2.6378, "step": 4104 }, { "epoch": 0.03691994981405117, "grad_norm": 2.0132758617401123, "learning_rate": 3.7037255187266986e-05, "loss": 2.9208, "step": 4105 }, { "epoch": 0.036928943711691614, "grad_norm": 2.370893716812134, "learning_rate": 3.7030667434076936e-05, "loss": 2.7282, "step": 4106 }, { "epoch": 0.03693793760933207, "grad_norm": 1.1999024152755737, "learning_rate": 3.702407859356812e-05, "loss": 2.478, "step": 4107 }, { "epoch": 0.03694693150697252, "grad_norm": 1.4941229820251465, "learning_rate": 3.701748866633603e-05, "loss": 2.5054, "step": 4108 }, { "epoch": 0.03695592540461297, "grad_norm": 1.4290263652801514, "learning_rate": 3.701089765297625e-05, "loss": 2.5293, "step": 4109 }, { "epoch": 0.03696491930225342, "grad_norm": 1.7559183835983276, "learning_rate": 3.700430555408448e-05, "loss": 2.5883, "step": 4110 }, { "epoch": 0.03697391319989387, "grad_norm": 1.4555747509002686, "learning_rate": 3.699771237025652e-05, "loss": 2.3577, "step": 4111 }, { "epoch": 0.03698290709753432, "grad_norm": 1.51092529296875, "learning_rate": 3.699111810208822e-05, "loss": 2.7549, "step": 4112 }, { "epoch": 0.03699190099517478, "grad_norm": 1.9780025482177734, "learning_rate": 3.6984522750175595e-05, "loss": 2.9842, "step": 4113 }, { "epoch": 0.037000894892815224, "grad_norm": 1.8395339250564575, "learning_rate": 3.697792631511471e-05, "loss": 2.4663, "step": 4114 }, { "epoch": 0.03700988879045568, "grad_norm": 1.9106343984603882, "learning_rate": 3.697132879750174e-05, "loss": 2.6615, "step": 4115 }, { "epoch": 0.037018882688096125, "grad_norm": 1.4448469877243042, "learning_rate": 3.696473019793297e-05, "loss": 2.7119, "step": 4116 }, { "epoch": 0.03702787658573658, "grad_norm": 1.8414264917373657, "learning_rate": 3.695813051700478e-05, "loss": 2.6977, "step": 4117 }, { "epoch": 0.037036870483377025, "grad_norm": 2.0560827255249023, "learning_rate": 3.695152975531362e-05, "loss": 2.4433, "step": 4118 }, { "epoch": 0.03704586438101748, "grad_norm": 1.8280144929885864, "learning_rate": 3.6944927913456085e-05, "loss": 2.6598, "step": 4119 }, { "epoch": 0.03705485827865793, "grad_norm": 1.814423680305481, "learning_rate": 3.693832499202883e-05, "loss": 2.7352, "step": 4120 }, { "epoch": 0.03706385217629838, "grad_norm": 1.5943603515625, "learning_rate": 3.693172099162861e-05, "loss": 2.8392, "step": 4121 }, { "epoch": 0.037072846073938834, "grad_norm": 1.2433197498321533, "learning_rate": 3.6925115912852304e-05, "loss": 2.4366, "step": 4122 }, { "epoch": 0.03708183997157928, "grad_norm": 1.3588217496871948, "learning_rate": 3.6918509756296874e-05, "loss": 2.4545, "step": 4123 }, { "epoch": 0.037090833869219735, "grad_norm": 2.015580892562866, "learning_rate": 3.6911902522559366e-05, "loss": 2.8942, "step": 4124 }, { "epoch": 0.03709982776686019, "grad_norm": 1.076526403427124, "learning_rate": 3.6905294212236937e-05, "loss": 2.1767, "step": 4125 }, { "epoch": 0.037108821664500635, "grad_norm": 2.231714963912964, "learning_rate": 3.689868482592684e-05, "loss": 2.9095, "step": 4126 }, { "epoch": 0.03711781556214109, "grad_norm": 1.8697946071624756, "learning_rate": 3.6892074364226425e-05, "loss": 2.666, "step": 4127 }, { "epoch": 0.037126809459781536, "grad_norm": 2.266927480697632, "learning_rate": 3.688546282773314e-05, "loss": 2.9588, "step": 4128 }, { "epoch": 0.03713580335742199, "grad_norm": 1.3627322912216187, "learning_rate": 3.687885021704452e-05, "loss": 2.5105, "step": 4129 }, { "epoch": 0.03714479725506244, "grad_norm": 1.5865817070007324, "learning_rate": 3.6872236532758226e-05, "loss": 2.64, "step": 4130 }, { "epoch": 0.03715379115270289, "grad_norm": 1.5910104513168335, "learning_rate": 3.686562177547197e-05, "loss": 2.5723, "step": 4131 }, { "epoch": 0.037162785050343344, "grad_norm": 1.4348665475845337, "learning_rate": 3.68590059457836e-05, "loss": 2.6792, "step": 4132 }, { "epoch": 0.03717177894798379, "grad_norm": 2.272031784057617, "learning_rate": 3.685238904429105e-05, "loss": 3.1204, "step": 4133 }, { "epoch": 0.037180772845624245, "grad_norm": 2.6473848819732666, "learning_rate": 3.684577107159235e-05, "loss": 2.835, "step": 4134 }, { "epoch": 0.03718976674326469, "grad_norm": 1.9858535528182983, "learning_rate": 3.683915202828561e-05, "loss": 2.8486, "step": 4135 }, { "epoch": 0.037198760640905146, "grad_norm": 1.8657938241958618, "learning_rate": 3.683253191496906e-05, "loss": 2.9089, "step": 4136 }, { "epoch": 0.0372077545385456, "grad_norm": 1.1115305423736572, "learning_rate": 3.6825910732241026e-05, "loss": 2.5345, "step": 4137 }, { "epoch": 0.037216748436186047, "grad_norm": 1.6933058500289917, "learning_rate": 3.6819288480699914e-05, "loss": 2.6203, "step": 4138 }, { "epoch": 0.0372257423338265, "grad_norm": 1.1565539836883545, "learning_rate": 3.681266516094424e-05, "loss": 2.2939, "step": 4139 }, { "epoch": 0.03723473623146695, "grad_norm": 1.7512598037719727, "learning_rate": 3.6806040773572615e-05, "loss": 2.6723, "step": 4140 }, { "epoch": 0.0372437301291074, "grad_norm": 1.5645256042480469, "learning_rate": 3.679941531918375e-05, "loss": 2.4118, "step": 4141 }, { "epoch": 0.037252724026747855, "grad_norm": 1.7808257341384888, "learning_rate": 3.679278879837642e-05, "loss": 2.7171, "step": 4142 }, { "epoch": 0.0372617179243883, "grad_norm": 1.5328327417373657, "learning_rate": 3.6786161211749547e-05, "loss": 2.4808, "step": 4143 }, { "epoch": 0.037270711822028756, "grad_norm": 1.0457390546798706, "learning_rate": 3.6779532559902124e-05, "loss": 2.3495, "step": 4144 }, { "epoch": 0.0372797057196692, "grad_norm": 2.1519205570220947, "learning_rate": 3.6772902843433234e-05, "loss": 2.6863, "step": 4145 }, { "epoch": 0.037288699617309656, "grad_norm": 1.3976410627365112, "learning_rate": 3.676627206294207e-05, "loss": 2.431, "step": 4146 }, { "epoch": 0.0372976935149501, "grad_norm": 2.4560089111328125, "learning_rate": 3.67596402190279e-05, "loss": 3.1191, "step": 4147 }, { "epoch": 0.03730668741259056, "grad_norm": 1.8305248022079468, "learning_rate": 3.675300731229012e-05, "loss": 2.7167, "step": 4148 }, { "epoch": 0.03731568131023101, "grad_norm": 1.561865210533142, "learning_rate": 3.6746373343328194e-05, "loss": 2.2532, "step": 4149 }, { "epoch": 0.03732467520787146, "grad_norm": 1.7163256406784058, "learning_rate": 3.67397383127417e-05, "loss": 2.6165, "step": 4150 }, { "epoch": 0.03733366910551191, "grad_norm": 1.4743196964263916, "learning_rate": 3.67331022211303e-05, "loss": 2.3291, "step": 4151 }, { "epoch": 0.03734266300315236, "grad_norm": 1.7658414840698242, "learning_rate": 3.672646506909377e-05, "loss": 2.63, "step": 4152 }, { "epoch": 0.03735165690079281, "grad_norm": 1.6462267637252808, "learning_rate": 3.671982685723194e-05, "loss": 2.5831, "step": 4153 }, { "epoch": 0.037360650798433266, "grad_norm": 1.920540452003479, "learning_rate": 3.6713187586144804e-05, "loss": 2.6606, "step": 4154 }, { "epoch": 0.03736964469607371, "grad_norm": 1.2835266590118408, "learning_rate": 3.6706547256432374e-05, "loss": 2.2566, "step": 4155 }, { "epoch": 0.03737863859371417, "grad_norm": 1.3071614503860474, "learning_rate": 3.6699905868694816e-05, "loss": 2.5571, "step": 4156 }, { "epoch": 0.037387632491354614, "grad_norm": 1.5810612440109253, "learning_rate": 3.6693263423532366e-05, "loss": 2.6131, "step": 4157 }, { "epoch": 0.03739662638899507, "grad_norm": 1.6603367328643799, "learning_rate": 3.6686619921545366e-05, "loss": 2.8064, "step": 4158 }, { "epoch": 0.037405620286635514, "grad_norm": 1.4719425439834595, "learning_rate": 3.667997536333424e-05, "loss": 2.419, "step": 4159 }, { "epoch": 0.03741461418427597, "grad_norm": 2.1609861850738525, "learning_rate": 3.667332974949953e-05, "loss": 2.9898, "step": 4160 }, { "epoch": 0.03742360808191642, "grad_norm": 1.9102791547775269, "learning_rate": 3.6666683080641846e-05, "loss": 2.71, "step": 4161 }, { "epoch": 0.03743260197955687, "grad_norm": 1.3296762704849243, "learning_rate": 3.666003535736191e-05, "loss": 2.5407, "step": 4162 }, { "epoch": 0.03744159587719732, "grad_norm": 1.9834681749343872, "learning_rate": 3.665338658026054e-05, "loss": 3.0626, "step": 4163 }, { "epoch": 0.03745058977483777, "grad_norm": 1.6137361526489258, "learning_rate": 3.6646736749938646e-05, "loss": 2.5439, "step": 4164 }, { "epoch": 0.037459583672478224, "grad_norm": 1.8425134420394897, "learning_rate": 3.664008586699722e-05, "loss": 2.8228, "step": 4165 }, { "epoch": 0.03746857757011868, "grad_norm": 1.759131669998169, "learning_rate": 3.663343393203738e-05, "loss": 2.7877, "step": 4166 }, { "epoch": 0.037477571467759124, "grad_norm": 1.8265564441680908, "learning_rate": 3.66267809456603e-05, "loss": 2.5506, "step": 4167 }, { "epoch": 0.03748656536539958, "grad_norm": 1.987113118171692, "learning_rate": 3.6620126908467295e-05, "loss": 2.7417, "step": 4168 }, { "epoch": 0.037495559263040025, "grad_norm": 1.5526129007339478, "learning_rate": 3.661347182105973e-05, "loss": 2.6822, "step": 4169 }, { "epoch": 0.03750455316068048, "grad_norm": 1.6587305068969727, "learning_rate": 3.66068156840391e-05, "loss": 2.8926, "step": 4170 }, { "epoch": 0.03751354705832093, "grad_norm": 1.580249547958374, "learning_rate": 3.660015849800695e-05, "loss": 2.6012, "step": 4171 }, { "epoch": 0.03752254095596138, "grad_norm": 1.640329360961914, "learning_rate": 3.659350026356498e-05, "loss": 2.4536, "step": 4172 }, { "epoch": 0.03753153485360183, "grad_norm": 0.9011757373809814, "learning_rate": 3.6586840981314954e-05, "loss": 2.2821, "step": 4173 }, { "epoch": 0.03754052875124228, "grad_norm": 2.93022084236145, "learning_rate": 3.6580180651858705e-05, "loss": 2.8748, "step": 4174 }, { "epoch": 0.037549522648882734, "grad_norm": 1.9294586181640625, "learning_rate": 3.6573519275798214e-05, "loss": 2.9824, "step": 4175 }, { "epoch": 0.03755851654652318, "grad_norm": 1.8179420232772827, "learning_rate": 3.6566856853735516e-05, "loss": 2.5961, "step": 4176 }, { "epoch": 0.037567510444163635, "grad_norm": 1.606546401977539, "learning_rate": 3.6560193386272743e-05, "loss": 2.7079, "step": 4177 }, { "epoch": 0.03757650434180409, "grad_norm": 1.6996022462844849, "learning_rate": 3.655352887401215e-05, "loss": 2.6165, "step": 4178 }, { "epoch": 0.037585498239444536, "grad_norm": 2.0482177734375, "learning_rate": 3.654686331755606e-05, "loss": 2.7317, "step": 4179 }, { "epoch": 0.03759449213708499, "grad_norm": 1.6851061582565308, "learning_rate": 3.65401967175069e-05, "loss": 2.5154, "step": 4180 }, { "epoch": 0.037603486034725436, "grad_norm": 1.6685789823532104, "learning_rate": 3.65335290744672e-05, "loss": 2.6621, "step": 4181 }, { "epoch": 0.03761247993236589, "grad_norm": 1.5178762674331665, "learning_rate": 3.652686038903956e-05, "loss": 2.4476, "step": 4182 }, { "epoch": 0.037621473830006344, "grad_norm": 1.5181981325149536, "learning_rate": 3.6520190661826695e-05, "loss": 2.479, "step": 4183 }, { "epoch": 0.03763046772764679, "grad_norm": 2.1660878658294678, "learning_rate": 3.65135198934314e-05, "loss": 2.8787, "step": 4184 }, { "epoch": 0.037639461625287245, "grad_norm": 0.9992207884788513, "learning_rate": 3.6506848084456577e-05, "loss": 2.404, "step": 4185 }, { "epoch": 0.03764845552292769, "grad_norm": 1.7691996097564697, "learning_rate": 3.6500175235505224e-05, "loss": 2.7348, "step": 4186 }, { "epoch": 0.037657449420568145, "grad_norm": 1.1741465330123901, "learning_rate": 3.6493501347180416e-05, "loss": 2.4317, "step": 4187 }, { "epoch": 0.03766644331820859, "grad_norm": 1.5431565046310425, "learning_rate": 3.648682642008534e-05, "loss": 2.526, "step": 4188 }, { "epoch": 0.037675437215849046, "grad_norm": 1.803174376487732, "learning_rate": 3.648015045482327e-05, "loss": 2.6102, "step": 4189 }, { "epoch": 0.0376844311134895, "grad_norm": 1.9306353330612183, "learning_rate": 3.647347345199756e-05, "loss": 2.8937, "step": 4190 }, { "epoch": 0.03769342501112995, "grad_norm": 1.8507392406463623, "learning_rate": 3.646679541221168e-05, "loss": 2.7811, "step": 4191 }, { "epoch": 0.0377024189087704, "grad_norm": 1.8445013761520386, "learning_rate": 3.6460116336069176e-05, "loss": 2.648, "step": 4192 }, { "epoch": 0.03771141280641085, "grad_norm": 1.3630894422531128, "learning_rate": 3.6453436224173704e-05, "loss": 2.7791, "step": 4193 }, { "epoch": 0.0377204067040513, "grad_norm": 1.9898568391799927, "learning_rate": 3.6446755077129006e-05, "loss": 2.5679, "step": 4194 }, { "epoch": 0.037729400601691755, "grad_norm": 1.6992641687393188, "learning_rate": 3.644007289553891e-05, "loss": 2.8017, "step": 4195 }, { "epoch": 0.0377383944993322, "grad_norm": 2.0244321823120117, "learning_rate": 3.643338968000736e-05, "loss": 2.7977, "step": 4196 }, { "epoch": 0.037747388396972656, "grad_norm": 1.9462965726852417, "learning_rate": 3.6426705431138356e-05, "loss": 2.7506, "step": 4197 }, { "epoch": 0.0377563822946131, "grad_norm": 1.6517012119293213, "learning_rate": 3.642002014953602e-05, "loss": 2.9809, "step": 4198 }, { "epoch": 0.03776537619225356, "grad_norm": 1.1507667303085327, "learning_rate": 3.6413333835804566e-05, "loss": 2.5947, "step": 4199 }, { "epoch": 0.037774370089894004, "grad_norm": 2.1364071369171143, "learning_rate": 3.640664649054829e-05, "loss": 2.7627, "step": 4200 }, { "epoch": 0.03778336398753446, "grad_norm": 1.7480672597885132, "learning_rate": 3.6399958114371595e-05, "loss": 2.9183, "step": 4201 }, { "epoch": 0.03779235788517491, "grad_norm": 1.5112261772155762, "learning_rate": 3.639326870787896e-05, "loss": 2.6186, "step": 4202 }, { "epoch": 0.03780135178281536, "grad_norm": 2.006838083267212, "learning_rate": 3.6386578271674984e-05, "loss": 2.645, "step": 4203 }, { "epoch": 0.03781034568045581, "grad_norm": 1.7307850122451782, "learning_rate": 3.6379886806364316e-05, "loss": 2.6628, "step": 4204 }, { "epoch": 0.03781933957809626, "grad_norm": 1.6781830787658691, "learning_rate": 3.637319431255174e-05, "loss": 2.6826, "step": 4205 }, { "epoch": 0.03782833347573671, "grad_norm": 2.135406255722046, "learning_rate": 3.6366500790842115e-05, "loss": 2.9326, "step": 4206 }, { "epoch": 0.037837327373377166, "grad_norm": 1.6108239889144897, "learning_rate": 3.635980624184038e-05, "loss": 2.5736, "step": 4207 }, { "epoch": 0.03784632127101761, "grad_norm": 1.1693636178970337, "learning_rate": 3.63531106661516e-05, "loss": 2.3475, "step": 4208 }, { "epoch": 0.03785531516865807, "grad_norm": 1.004227638244629, "learning_rate": 3.6346414064380914e-05, "loss": 2.3797, "step": 4209 }, { "epoch": 0.037864309066298514, "grad_norm": 2.4279301166534424, "learning_rate": 3.6339716437133535e-05, "loss": 2.7203, "step": 4210 }, { "epoch": 0.03787330296393897, "grad_norm": 1.6359918117523193, "learning_rate": 3.6333017785014813e-05, "loss": 2.6078, "step": 4211 }, { "epoch": 0.03788229686157942, "grad_norm": 2.117284059524536, "learning_rate": 3.632631810863013e-05, "loss": 2.5356, "step": 4212 }, { "epoch": 0.03789129075921987, "grad_norm": 1.6277445554733276, "learning_rate": 3.631961740858503e-05, "loss": 2.6332, "step": 4213 }, { "epoch": 0.03790028465686032, "grad_norm": 2.192307949066162, "learning_rate": 3.631291568548509e-05, "loss": 3.0889, "step": 4214 }, { "epoch": 0.03790927855450077, "grad_norm": 1.536473035812378, "learning_rate": 3.630621293993602e-05, "loss": 2.445, "step": 4215 }, { "epoch": 0.03791827245214122, "grad_norm": 1.896742820739746, "learning_rate": 3.629950917254362e-05, "loss": 2.8473, "step": 4216 }, { "epoch": 0.03792726634978167, "grad_norm": 1.7203538417816162, "learning_rate": 3.6292804383913736e-05, "loss": 2.6004, "step": 4217 }, { "epoch": 0.037936260247422124, "grad_norm": 1.602813720703125, "learning_rate": 3.628609857465236e-05, "loss": 2.5068, "step": 4218 }, { "epoch": 0.03794525414506258, "grad_norm": 1.5803674459457397, "learning_rate": 3.627939174536554e-05, "loss": 2.6202, "step": 4219 }, { "epoch": 0.037954248042703025, "grad_norm": 1.743503212928772, "learning_rate": 3.627268389665945e-05, "loss": 2.66, "step": 4220 }, { "epoch": 0.03796324194034348, "grad_norm": 1.522035837173462, "learning_rate": 3.626597502914033e-05, "loss": 2.5258, "step": 4221 }, { "epoch": 0.037972235837983925, "grad_norm": 1.5961893796920776, "learning_rate": 3.625926514341452e-05, "loss": 2.7007, "step": 4222 }, { "epoch": 0.03798122973562438, "grad_norm": 1.9801156520843506, "learning_rate": 3.625255424008847e-05, "loss": 2.5701, "step": 4223 }, { "epoch": 0.03799022363326483, "grad_norm": 1.6422834396362305, "learning_rate": 3.624584231976867e-05, "loss": 2.8375, "step": 4224 }, { "epoch": 0.03799921753090528, "grad_norm": 1.5878595113754272, "learning_rate": 3.623912938306176e-05, "loss": 2.6944, "step": 4225 }, { "epoch": 0.038008211428545734, "grad_norm": 1.545140027999878, "learning_rate": 3.623241543057444e-05, "loss": 2.4875, "step": 4226 }, { "epoch": 0.03801720532618618, "grad_norm": 1.3952250480651855, "learning_rate": 3.622570046291352e-05, "loss": 2.7565, "step": 4227 }, { "epoch": 0.038026199223826634, "grad_norm": 1.0368702411651611, "learning_rate": 3.6218984480685874e-05, "loss": 2.464, "step": 4228 }, { "epoch": 0.03803519312146708, "grad_norm": 1.9220932722091675, "learning_rate": 3.6212267484498504e-05, "loss": 2.7844, "step": 4229 }, { "epoch": 0.038044187019107535, "grad_norm": 1.066354751586914, "learning_rate": 3.620554947495847e-05, "loss": 2.2811, "step": 4230 }, { "epoch": 0.03805318091674799, "grad_norm": 2.2018909454345703, "learning_rate": 3.6198830452672945e-05, "loss": 2.6836, "step": 4231 }, { "epoch": 0.038062174814388436, "grad_norm": 1.4963877201080322, "learning_rate": 3.619211041824918e-05, "loss": 2.4242, "step": 4232 }, { "epoch": 0.03807116871202889, "grad_norm": 2.618581771850586, "learning_rate": 3.618538937229454e-05, "loss": 2.4333, "step": 4233 }, { "epoch": 0.03808016260966934, "grad_norm": 1.335446834564209, "learning_rate": 3.617866731541645e-05, "loss": 2.4428, "step": 4234 }, { "epoch": 0.03808915650730979, "grad_norm": 2.2028064727783203, "learning_rate": 3.6171944248222455e-05, "loss": 2.9351, "step": 4235 }, { "epoch": 0.038098150404950244, "grad_norm": 1.4644043445587158, "learning_rate": 3.616522017132017e-05, "loss": 2.8336, "step": 4236 }, { "epoch": 0.03810714430259069, "grad_norm": 2.2431800365448, "learning_rate": 3.61584950853173e-05, "loss": 2.9413, "step": 4237 }, { "epoch": 0.038116138200231145, "grad_norm": 1.5709748268127441, "learning_rate": 3.615176899082168e-05, "loss": 2.9361, "step": 4238 }, { "epoch": 0.03812513209787159, "grad_norm": 1.5906323194503784, "learning_rate": 3.614504188844118e-05, "loss": 2.6648, "step": 4239 }, { "epoch": 0.038134125995512046, "grad_norm": 1.6523988246917725, "learning_rate": 3.6138313778783804e-05, "loss": 2.4733, "step": 4240 }, { "epoch": 0.03814311989315249, "grad_norm": 1.8378541469573975, "learning_rate": 3.613158466245763e-05, "loss": 2.5551, "step": 4241 }, { "epoch": 0.038152113790792946, "grad_norm": 1.1005371809005737, "learning_rate": 3.612485454007082e-05, "loss": 2.4303, "step": 4242 }, { "epoch": 0.0381611076884334, "grad_norm": 1.7779488563537598, "learning_rate": 3.611812341223163e-05, "loss": 2.5161, "step": 4243 }, { "epoch": 0.03817010158607385, "grad_norm": 1.5981870889663696, "learning_rate": 3.6111391279548435e-05, "loss": 2.4626, "step": 4244 }, { "epoch": 0.0381790954837143, "grad_norm": 1.8398759365081787, "learning_rate": 3.6104658142629654e-05, "loss": 2.8202, "step": 4245 }, { "epoch": 0.03818808938135475, "grad_norm": 2.253457546234131, "learning_rate": 3.609792400208384e-05, "loss": 2.959, "step": 4246 }, { "epoch": 0.0381970832789952, "grad_norm": 2.1177985668182373, "learning_rate": 3.6091188858519607e-05, "loss": 3.1084, "step": 4247 }, { "epoch": 0.038206077176635656, "grad_norm": 2.1399285793304443, "learning_rate": 3.608445271254566e-05, "loss": 3.0283, "step": 4248 }, { "epoch": 0.0382150710742761, "grad_norm": 1.414995789527893, "learning_rate": 3.607771556477083e-05, "loss": 2.4179, "step": 4249 }, { "epoch": 0.038224064971916556, "grad_norm": 2.0489118099212646, "learning_rate": 3.6070977415803986e-05, "loss": 2.6192, "step": 4250 }, { "epoch": 0.038233058869557, "grad_norm": 1.5025792121887207, "learning_rate": 3.6064238266254145e-05, "loss": 2.3609, "step": 4251 }, { "epoch": 0.03824205276719746, "grad_norm": 1.8622610569000244, "learning_rate": 3.6057498116730355e-05, "loss": 2.5868, "step": 4252 }, { "epoch": 0.03825104666483791, "grad_norm": 1.668278455734253, "learning_rate": 3.60507569678418e-05, "loss": 2.5965, "step": 4253 }, { "epoch": 0.03826004056247836, "grad_norm": 1.7714275121688843, "learning_rate": 3.6044014820197725e-05, "loss": 2.4935, "step": 4254 }, { "epoch": 0.03826903446011881, "grad_norm": 1.8485870361328125, "learning_rate": 3.603727167440749e-05, "loss": 3.1364, "step": 4255 }, { "epoch": 0.03827802835775926, "grad_norm": 3.196810722351074, "learning_rate": 3.603052753108053e-05, "loss": 2.4766, "step": 4256 }, { "epoch": 0.03828702225539971, "grad_norm": 1.8554177284240723, "learning_rate": 3.602378239082638e-05, "loss": 2.6847, "step": 4257 }, { "epoch": 0.03829601615304016, "grad_norm": 1.901282787322998, "learning_rate": 3.601703625425464e-05, "loss": 2.3526, "step": 4258 }, { "epoch": 0.03830501005068061, "grad_norm": 1.6393566131591797, "learning_rate": 3.601028912197504e-05, "loss": 2.4394, "step": 4259 }, { "epoch": 0.03831400394832107, "grad_norm": 1.8540844917297363, "learning_rate": 3.600354099459737e-05, "loss": 2.6421, "step": 4260 }, { "epoch": 0.038322997845961514, "grad_norm": 1.7406470775604248, "learning_rate": 3.59967918727315e-05, "loss": 2.7568, "step": 4261 }, { "epoch": 0.03833199174360197, "grad_norm": 1.7599694728851318, "learning_rate": 3.599004175698744e-05, "loss": 2.7779, "step": 4262 }, { "epoch": 0.038340985641242414, "grad_norm": 1.754137635231018, "learning_rate": 3.5983290647975235e-05, "loss": 2.6339, "step": 4263 }, { "epoch": 0.03834997953888287, "grad_norm": 1.3696986436843872, "learning_rate": 3.597653854630506e-05, "loss": 2.407, "step": 4264 }, { "epoch": 0.03835897343652332, "grad_norm": 2.1887638568878174, "learning_rate": 3.5969785452587154e-05, "loss": 2.8942, "step": 4265 }, { "epoch": 0.03836796733416377, "grad_norm": 1.8246631622314453, "learning_rate": 3.596303136743185e-05, "loss": 2.7172, "step": 4266 }, { "epoch": 0.03837696123180422, "grad_norm": 1.7691447734832764, "learning_rate": 3.5956276291449584e-05, "loss": 2.6538, "step": 4267 }, { "epoch": 0.03838595512944467, "grad_norm": 1.7444006204605103, "learning_rate": 3.594952022525086e-05, "loss": 2.7788, "step": 4268 }, { "epoch": 0.038394949027085123, "grad_norm": 1.821008324623108, "learning_rate": 3.5942763169446295e-05, "loss": 2.5657, "step": 4269 }, { "epoch": 0.03840394292472557, "grad_norm": 2.0695290565490723, "learning_rate": 3.593600512464658e-05, "loss": 2.8352, "step": 4270 }, { "epoch": 0.038412936822366024, "grad_norm": 2.0044031143188477, "learning_rate": 3.592924609146251e-05, "loss": 2.67, "step": 4271 }, { "epoch": 0.03842193072000648, "grad_norm": 2.1275365352630615, "learning_rate": 3.5922486070504944e-05, "loss": 2.7943, "step": 4272 }, { "epoch": 0.038430924617646925, "grad_norm": 1.8445630073547363, "learning_rate": 3.591572506238485e-05, "loss": 2.866, "step": 4273 }, { "epoch": 0.03843991851528738, "grad_norm": 1.286908745765686, "learning_rate": 3.590896306771329e-05, "loss": 2.2643, "step": 4274 }, { "epoch": 0.038448912412927826, "grad_norm": 1.0647387504577637, "learning_rate": 3.5902200087101384e-05, "loss": 2.3385, "step": 4275 }, { "epoch": 0.03845790631056828, "grad_norm": 1.543147325515747, "learning_rate": 3.5895436121160386e-05, "loss": 2.6101, "step": 4276 }, { "epoch": 0.03846690020820873, "grad_norm": 1.9546363353729248, "learning_rate": 3.588867117050161e-05, "loss": 2.8192, "step": 4277 }, { "epoch": 0.03847589410584918, "grad_norm": 1.3303875923156738, "learning_rate": 3.588190523573646e-05, "loss": 2.4411, "step": 4278 }, { "epoch": 0.038484888003489634, "grad_norm": 1.1169207096099854, "learning_rate": 3.587513831747643e-05, "loss": 2.3689, "step": 4279 }, { "epoch": 0.03849388190113008, "grad_norm": 2.1327593326568604, "learning_rate": 3.586837041633312e-05, "loss": 2.3658, "step": 4280 }, { "epoch": 0.038502875798770535, "grad_norm": 1.462516188621521, "learning_rate": 3.586160153291819e-05, "loss": 2.6075, "step": 4281 }, { "epoch": 0.03851186969641099, "grad_norm": 2.189162015914917, "learning_rate": 3.585483166784342e-05, "loss": 2.7289, "step": 4282 }, { "epoch": 0.038520863594051435, "grad_norm": 1.484312653541565, "learning_rate": 3.584806082172065e-05, "loss": 2.4246, "step": 4283 }, { "epoch": 0.03852985749169189, "grad_norm": 1.7973440885543823, "learning_rate": 3.5841288995161824e-05, "loss": 2.6112, "step": 4284 }, { "epoch": 0.038538851389332336, "grad_norm": 2.1871590614318848, "learning_rate": 3.5834516188778986e-05, "loss": 2.6399, "step": 4285 }, { "epoch": 0.03854784528697279, "grad_norm": 2.1736574172973633, "learning_rate": 3.5827742403184245e-05, "loss": 3.2479, "step": 4286 }, { "epoch": 0.03855683918461324, "grad_norm": 1.611697793006897, "learning_rate": 3.58209676389898e-05, "loss": 2.6555, "step": 4287 }, { "epoch": 0.03856583308225369, "grad_norm": 1.4008960723876953, "learning_rate": 3.5814191896807966e-05, "loss": 2.4069, "step": 4288 }, { "epoch": 0.038574826979894145, "grad_norm": 2.3634822368621826, "learning_rate": 3.580741517725111e-05, "loss": 2.8854, "step": 4289 }, { "epoch": 0.03858382087753459, "grad_norm": 1.5215293169021606, "learning_rate": 3.5800637480931706e-05, "loss": 2.4744, "step": 4290 }, { "epoch": 0.038592814775175045, "grad_norm": 2.121868371963501, "learning_rate": 3.579385880846232e-05, "loss": 3.2247, "step": 4291 }, { "epoch": 0.03860180867281549, "grad_norm": 1.7381349802017212, "learning_rate": 3.578707916045561e-05, "loss": 2.6726, "step": 4292 }, { "epoch": 0.038610802570455946, "grad_norm": 2.1851325035095215, "learning_rate": 3.5780298537524304e-05, "loss": 2.6884, "step": 4293 }, { "epoch": 0.0386197964680964, "grad_norm": 1.636803150177002, "learning_rate": 3.577351694028123e-05, "loss": 2.6447, "step": 4294 }, { "epoch": 0.03862879036573685, "grad_norm": 2.382847309112549, "learning_rate": 3.576673436933929e-05, "loss": 3.2471, "step": 4295 }, { "epoch": 0.0386377842633773, "grad_norm": 1.954850196838379, "learning_rate": 3.575995082531149e-05, "loss": 2.3439, "step": 4296 }, { "epoch": 0.03864677816101775, "grad_norm": 1.2926579713821411, "learning_rate": 3.5753166308810935e-05, "loss": 2.4343, "step": 4297 }, { "epoch": 0.0386557720586582, "grad_norm": 1.4956152439117432, "learning_rate": 3.574638082045079e-05, "loss": 2.513, "step": 4298 }, { "epoch": 0.03866476595629865, "grad_norm": 1.0894114971160889, "learning_rate": 3.573959436084432e-05, "loss": 2.3049, "step": 4299 }, { "epoch": 0.0386737598539391, "grad_norm": 1.350954294204712, "learning_rate": 3.573280693060488e-05, "loss": 2.5152, "step": 4300 }, { "epoch": 0.038682753751579556, "grad_norm": 1.9989923238754272, "learning_rate": 3.5726018530345915e-05, "loss": 2.5578, "step": 4301 }, { "epoch": 0.03869174764922, "grad_norm": 1.2624728679656982, "learning_rate": 3.571922916068094e-05, "loss": 2.6948, "step": 4302 }, { "epoch": 0.03870074154686046, "grad_norm": 1.4550580978393555, "learning_rate": 3.571243882222358e-05, "loss": 2.3884, "step": 4303 }, { "epoch": 0.0387097354445009, "grad_norm": 1.4997247457504272, "learning_rate": 3.570564751558754e-05, "loss": 2.4649, "step": 4304 }, { "epoch": 0.03871872934214136, "grad_norm": 1.1561869382858276, "learning_rate": 3.5698855241386604e-05, "loss": 2.4944, "step": 4305 }, { "epoch": 0.03872772323978181, "grad_norm": 1.2742180824279785, "learning_rate": 3.569206200023466e-05, "loss": 2.3859, "step": 4306 }, { "epoch": 0.03873671713742226, "grad_norm": 1.5813676118850708, "learning_rate": 3.5685267792745676e-05, "loss": 2.3774, "step": 4307 }, { "epoch": 0.03874571103506271, "grad_norm": 2.2800655364990234, "learning_rate": 3.5678472619533685e-05, "loss": 2.6511, "step": 4308 }, { "epoch": 0.03875470493270316, "grad_norm": 1.6807173490524292, "learning_rate": 3.567167648121285e-05, "loss": 2.6631, "step": 4309 }, { "epoch": 0.03876369883034361, "grad_norm": 1.101976990699768, "learning_rate": 3.566487937839738e-05, "loss": 2.3719, "step": 4310 }, { "epoch": 0.03877269272798406, "grad_norm": 2.251037120819092, "learning_rate": 3.565808131170161e-05, "loss": 2.7368, "step": 4311 }, { "epoch": 0.03878168662562451, "grad_norm": 1.722577452659607, "learning_rate": 3.565128228173992e-05, "loss": 2.6672, "step": 4312 }, { "epoch": 0.03879068052326497, "grad_norm": 1.5068058967590332, "learning_rate": 3.564448228912682e-05, "loss": 2.8053, "step": 4313 }, { "epoch": 0.038799674420905414, "grad_norm": 1.4974627494812012, "learning_rate": 3.563768133447687e-05, "loss": 2.6453, "step": 4314 }, { "epoch": 0.03880866831854587, "grad_norm": 1.9071520566940308, "learning_rate": 3.563087941840474e-05, "loss": 2.8295, "step": 4315 }, { "epoch": 0.038817662216186315, "grad_norm": 1.6086000204086304, "learning_rate": 3.562407654152518e-05, "loss": 2.6657, "step": 4316 }, { "epoch": 0.03882665611382677, "grad_norm": 1.879254698753357, "learning_rate": 3.561727270445302e-05, "loss": 2.9091, "step": 4317 }, { "epoch": 0.03883565001146722, "grad_norm": 1.3812000751495361, "learning_rate": 3.56104679078032e-05, "loss": 2.409, "step": 4318 }, { "epoch": 0.03884464390910767, "grad_norm": 1.790460228919983, "learning_rate": 3.560366215219071e-05, "loss": 2.7385, "step": 4319 }, { "epoch": 0.03885363780674812, "grad_norm": 1.94392991065979, "learning_rate": 3.559685543823066e-05, "loss": 2.9594, "step": 4320 }, { "epoch": 0.03886263170438857, "grad_norm": 1.9204518795013428, "learning_rate": 3.559004776653823e-05, "loss": 2.916, "step": 4321 }, { "epoch": 0.038871625602029024, "grad_norm": 1.5620052814483643, "learning_rate": 3.5583239137728684e-05, "loss": 2.4004, "step": 4322 }, { "epoch": 0.03888061949966948, "grad_norm": 1.964674711227417, "learning_rate": 3.557642955241739e-05, "loss": 2.4973, "step": 4323 }, { "epoch": 0.038889613397309925, "grad_norm": 1.6786314249038696, "learning_rate": 3.556961901121978e-05, "loss": 2.5619, "step": 4324 }, { "epoch": 0.03889860729495038, "grad_norm": 1.907672643661499, "learning_rate": 3.556280751475139e-05, "loss": 2.8849, "step": 4325 }, { "epoch": 0.038907601192590825, "grad_norm": 1.3249690532684326, "learning_rate": 3.555599506362784e-05, "loss": 2.5072, "step": 4326 }, { "epoch": 0.03891659509023128, "grad_norm": 1.4135900735855103, "learning_rate": 3.554918165846482e-05, "loss": 2.5984, "step": 4327 }, { "epoch": 0.038925588987871726, "grad_norm": 1.8034052848815918, "learning_rate": 3.5542367299878134e-05, "loss": 2.454, "step": 4328 }, { "epoch": 0.03893458288551218, "grad_norm": 1.6253139972686768, "learning_rate": 3.553555198848363e-05, "loss": 2.6505, "step": 4329 }, { "epoch": 0.038943576783152634, "grad_norm": 0.7888286709785461, "learning_rate": 3.55287357248973e-05, "loss": 2.301, "step": 4330 }, { "epoch": 0.03895257068079308, "grad_norm": 1.5398356914520264, "learning_rate": 3.552191850973517e-05, "loss": 2.6061, "step": 4331 }, { "epoch": 0.038961564578433534, "grad_norm": 1.7515863180160522, "learning_rate": 3.551510034361337e-05, "loss": 2.5629, "step": 4332 }, { "epoch": 0.03897055847607398, "grad_norm": 1.551504135131836, "learning_rate": 3.5508281227148136e-05, "loss": 2.6178, "step": 4333 }, { "epoch": 0.038979552373714435, "grad_norm": 1.4227750301361084, "learning_rate": 3.550146116095577e-05, "loss": 2.3598, "step": 4334 }, { "epoch": 0.03898854627135489, "grad_norm": 2.0063674449920654, "learning_rate": 3.549464014565265e-05, "loss": 2.6713, "step": 4335 }, { "epoch": 0.038997540168995336, "grad_norm": 1.8867048025131226, "learning_rate": 3.5487818181855255e-05, "loss": 2.7738, "step": 4336 }, { "epoch": 0.03900653406663579, "grad_norm": 2.125711679458618, "learning_rate": 3.5480995270180147e-05, "loss": 2.5441, "step": 4337 }, { "epoch": 0.039015527964276236, "grad_norm": 2.9075558185577393, "learning_rate": 3.547417141124397e-05, "loss": 2.9193, "step": 4338 }, { "epoch": 0.03902452186191669, "grad_norm": 0.7535821795463562, "learning_rate": 3.546734660566347e-05, "loss": 2.1875, "step": 4339 }, { "epoch": 0.03903351575955714, "grad_norm": 2.155046224594116, "learning_rate": 3.5460520854055465e-05, "loss": 2.6175, "step": 4340 }, { "epoch": 0.03904250965719759, "grad_norm": 2.0640103816986084, "learning_rate": 3.545369415703685e-05, "loss": 2.8799, "step": 4341 }, { "epoch": 0.039051503554838045, "grad_norm": 1.6441601514816284, "learning_rate": 3.544686651522462e-05, "loss": 2.4371, "step": 4342 }, { "epoch": 0.03906049745247849, "grad_norm": 1.3233158588409424, "learning_rate": 3.5440037929235835e-05, "loss": 2.4543, "step": 4343 }, { "epoch": 0.039069491350118946, "grad_norm": 1.5683834552764893, "learning_rate": 3.543320839968767e-05, "loss": 2.3575, "step": 4344 }, { "epoch": 0.03907848524775939, "grad_norm": 1.7488805055618286, "learning_rate": 3.5426377927197366e-05, "loss": 2.6761, "step": 4345 }, { "epoch": 0.039087479145399846, "grad_norm": 1.2281190156936646, "learning_rate": 3.5419546512382266e-05, "loss": 2.2523, "step": 4346 }, { "epoch": 0.0390964730430403, "grad_norm": 1.394361138343811, "learning_rate": 3.541271415585977e-05, "loss": 2.4065, "step": 4347 }, { "epoch": 0.03910546694068075, "grad_norm": 2.032426118850708, "learning_rate": 3.540588085824738e-05, "loss": 2.6984, "step": 4348 }, { "epoch": 0.0391144608383212, "grad_norm": 2.1634323596954346, "learning_rate": 3.539904662016269e-05, "loss": 2.7117, "step": 4349 }, { "epoch": 0.03912345473596165, "grad_norm": 1.4262696504592896, "learning_rate": 3.5392211442223365e-05, "loss": 2.6451, "step": 4350 }, { "epoch": 0.0391324486336021, "grad_norm": 1.7022312879562378, "learning_rate": 3.5385375325047166e-05, "loss": 2.6229, "step": 4351 }, { "epoch": 0.03914144253124255, "grad_norm": 1.6547703742980957, "learning_rate": 3.5378538269251924e-05, "loss": 2.5219, "step": 4352 }, { "epoch": 0.039150436428883, "grad_norm": 1.2035777568817139, "learning_rate": 3.537170027545558e-05, "loss": 2.2913, "step": 4353 }, { "epoch": 0.039159430326523456, "grad_norm": 1.5100306272506714, "learning_rate": 3.536486134427613e-05, "loss": 2.6053, "step": 4354 }, { "epoch": 0.0391684242241639, "grad_norm": 1.6072312593460083, "learning_rate": 3.535802147633169e-05, "loss": 2.3762, "step": 4355 }, { "epoch": 0.03917741812180436, "grad_norm": 1.7713837623596191, "learning_rate": 3.5351180672240414e-05, "loss": 2.8351, "step": 4356 }, { "epoch": 0.039186412019444804, "grad_norm": 1.6928741931915283, "learning_rate": 3.534433893262058e-05, "loss": 2.7943, "step": 4357 }, { "epoch": 0.03919540591708526, "grad_norm": 2.0104000568389893, "learning_rate": 3.533749625809053e-05, "loss": 2.6927, "step": 4358 }, { "epoch": 0.03920439981472571, "grad_norm": 1.8786591291427612, "learning_rate": 3.533065264926872e-05, "loss": 2.6771, "step": 4359 }, { "epoch": 0.03921339371236616, "grad_norm": 1.4345358610153198, "learning_rate": 3.532380810677364e-05, "loss": 2.7038, "step": 4360 }, { "epoch": 0.03922238761000661, "grad_norm": 1.8107023239135742, "learning_rate": 3.531696263122389e-05, "loss": 2.7433, "step": 4361 }, { "epoch": 0.03923138150764706, "grad_norm": 1.9260114431381226, "learning_rate": 3.531011622323819e-05, "loss": 2.736, "step": 4362 }, { "epoch": 0.03924037540528751, "grad_norm": 1.774664282798767, "learning_rate": 3.530326888343528e-05, "loss": 2.5294, "step": 4363 }, { "epoch": 0.03924936930292797, "grad_norm": 1.245728850364685, "learning_rate": 3.529642061243403e-05, "loss": 2.3588, "step": 4364 }, { "epoch": 0.039258363200568414, "grad_norm": 1.693515419960022, "learning_rate": 3.528957141085338e-05, "loss": 2.955, "step": 4365 }, { "epoch": 0.03926735709820887, "grad_norm": 1.9320275783538818, "learning_rate": 3.5282721279312344e-05, "loss": 2.7687, "step": 4366 }, { "epoch": 0.039276350995849314, "grad_norm": 1.8973227739334106, "learning_rate": 3.527587021843003e-05, "loss": 2.8137, "step": 4367 }, { "epoch": 0.03928534489348977, "grad_norm": 1.685325026512146, "learning_rate": 3.526901822882564e-05, "loss": 2.3037, "step": 4368 }, { "epoch": 0.039294338791130215, "grad_norm": 1.1917459964752197, "learning_rate": 3.526216531111845e-05, "loss": 2.496, "step": 4369 }, { "epoch": 0.03930333268877067, "grad_norm": 2.1434285640716553, "learning_rate": 3.52553114659278e-05, "loss": 2.6407, "step": 4370 }, { "epoch": 0.03931232658641112, "grad_norm": 1.7201480865478516, "learning_rate": 3.524845669387315e-05, "loss": 2.7087, "step": 4371 }, { "epoch": 0.03932132048405157, "grad_norm": 2.0255320072174072, "learning_rate": 3.524160099557402e-05, "loss": 2.8905, "step": 4372 }, { "epoch": 0.03933031438169202, "grad_norm": 1.9365555047988892, "learning_rate": 3.523474437165003e-05, "loss": 2.9264, "step": 4373 }, { "epoch": 0.03933930827933247, "grad_norm": 1.744293212890625, "learning_rate": 3.5227886822720856e-05, "loss": 2.5874, "step": 4374 }, { "epoch": 0.039348302176972924, "grad_norm": 1.5093646049499512, "learning_rate": 3.522102834940629e-05, "loss": 2.4974, "step": 4375 }, { "epoch": 0.03935729607461338, "grad_norm": 1.504341721534729, "learning_rate": 3.52141689523262e-05, "loss": 2.5562, "step": 4376 }, { "epoch": 0.039366289972253825, "grad_norm": 1.7945085763931274, "learning_rate": 3.5207308632100525e-05, "loss": 2.5964, "step": 4377 }, { "epoch": 0.03937528386989428, "grad_norm": 2.223681926727295, "learning_rate": 3.520044738934928e-05, "loss": 3.0254, "step": 4378 }, { "epoch": 0.039384277767534726, "grad_norm": 1.3787826299667358, "learning_rate": 3.519358522469259e-05, "loss": 2.5034, "step": 4379 }, { "epoch": 0.03939327166517518, "grad_norm": 1.3108664751052856, "learning_rate": 3.518672213875065e-05, "loss": 2.5666, "step": 4380 }, { "epoch": 0.039402265562815626, "grad_norm": 2.226986885070801, "learning_rate": 3.517985813214373e-05, "loss": 2.414, "step": 4381 }, { "epoch": 0.03941125946045608, "grad_norm": 1.986231803894043, "learning_rate": 3.5172993205492196e-05, "loss": 2.8109, "step": 4382 }, { "epoch": 0.039420253358096534, "grad_norm": 1.989922046661377, "learning_rate": 3.51661273594165e-05, "loss": 2.8148, "step": 4383 }, { "epoch": 0.03942924725573698, "grad_norm": 1.7651724815368652, "learning_rate": 3.515926059453717e-05, "loss": 2.7446, "step": 4384 }, { "epoch": 0.039438241153377435, "grad_norm": 1.3000984191894531, "learning_rate": 3.51523929114748e-05, "loss": 2.3835, "step": 4385 }, { "epoch": 0.03944723505101788, "grad_norm": 1.4386498928070068, "learning_rate": 3.5145524310850083e-05, "loss": 2.4604, "step": 4386 }, { "epoch": 0.039456228948658335, "grad_norm": 2.1265454292297363, "learning_rate": 3.513865479328382e-05, "loss": 3.0179, "step": 4387 }, { "epoch": 0.03946522284629879, "grad_norm": 1.4675655364990234, "learning_rate": 3.5131784359396856e-05, "loss": 2.4757, "step": 4388 }, { "epoch": 0.039474216743939236, "grad_norm": 1.4228352308273315, "learning_rate": 3.5124913009810134e-05, "loss": 2.4315, "step": 4389 }, { "epoch": 0.03948321064157969, "grad_norm": 1.6823338270187378, "learning_rate": 3.511804074514468e-05, "loss": 2.7603, "step": 4390 }, { "epoch": 0.03949220453922014, "grad_norm": 1.7290035486221313, "learning_rate": 3.511116756602161e-05, "loss": 2.5845, "step": 4391 }, { "epoch": 0.03950119843686059, "grad_norm": 1.5887150764465332, "learning_rate": 3.51042934730621e-05, "loss": 2.5248, "step": 4392 }, { "epoch": 0.039510192334501044, "grad_norm": 1.3977969884872437, "learning_rate": 3.509741846688742e-05, "loss": 2.6048, "step": 4393 }, { "epoch": 0.03951918623214149, "grad_norm": 1.2486575841903687, "learning_rate": 3.509054254811894e-05, "loss": 2.4291, "step": 4394 }, { "epoch": 0.039528180129781945, "grad_norm": 1.6769219636917114, "learning_rate": 3.50836657173781e-05, "loss": 2.666, "step": 4395 }, { "epoch": 0.03953717402742239, "grad_norm": 1.6732099056243896, "learning_rate": 3.507678797528641e-05, "loss": 2.7069, "step": 4396 }, { "epoch": 0.039546167925062846, "grad_norm": 2.027407169342041, "learning_rate": 3.5069909322465476e-05, "loss": 2.454, "step": 4397 }, { "epoch": 0.03955516182270329, "grad_norm": 1.7403335571289062, "learning_rate": 3.5063029759536984e-05, "loss": 2.6695, "step": 4398 }, { "epoch": 0.03956415572034375, "grad_norm": 1.4808553457260132, "learning_rate": 3.5056149287122694e-05, "loss": 2.6774, "step": 4399 }, { "epoch": 0.0395731496179842, "grad_norm": 1.306771159172058, "learning_rate": 3.504926790584447e-05, "loss": 2.5001, "step": 4400 }, { "epoch": 0.03958214351562465, "grad_norm": 1.8227893114089966, "learning_rate": 3.504238561632424e-05, "loss": 2.6397, "step": 4401 }, { "epoch": 0.0395911374132651, "grad_norm": 1.2531253099441528, "learning_rate": 3.503550241918401e-05, "loss": 2.3492, "step": 4402 }, { "epoch": 0.03960013131090555, "grad_norm": 1.1466059684753418, "learning_rate": 3.502861831504588e-05, "loss": 2.1999, "step": 4403 }, { "epoch": 0.039609125208546, "grad_norm": 1.1088780164718628, "learning_rate": 3.5021733304532035e-05, "loss": 2.437, "step": 4404 }, { "epoch": 0.039618119106186456, "grad_norm": 1.262181282043457, "learning_rate": 3.501484738826472e-05, "loss": 2.4443, "step": 4405 }, { "epoch": 0.0396271130038269, "grad_norm": 1.7331327199935913, "learning_rate": 3.5007960566866295e-05, "loss": 2.5644, "step": 4406 }, { "epoch": 0.039636106901467356, "grad_norm": 1.5468355417251587, "learning_rate": 3.500107284095917e-05, "loss": 2.6608, "step": 4407 }, { "epoch": 0.0396451007991078, "grad_norm": 1.5333197116851807, "learning_rate": 3.499418421116585e-05, "loss": 2.8033, "step": 4408 }, { "epoch": 0.03965409469674826, "grad_norm": 0.7521847486495972, "learning_rate": 3.498729467810892e-05, "loss": 2.1783, "step": 4409 }, { "epoch": 0.039663088594388704, "grad_norm": 1.8610211610794067, "learning_rate": 3.4980404242411064e-05, "loss": 2.6953, "step": 4410 }, { "epoch": 0.03967208249202916, "grad_norm": 1.763418197631836, "learning_rate": 3.4973512904695026e-05, "loss": 2.5912, "step": 4411 }, { "epoch": 0.03968107638966961, "grad_norm": 1.4699841737747192, "learning_rate": 3.496662066558363e-05, "loss": 2.4483, "step": 4412 }, { "epoch": 0.03969007028731006, "grad_norm": 1.2002185583114624, "learning_rate": 3.4959727525699795e-05, "loss": 2.199, "step": 4413 }, { "epoch": 0.03969906418495051, "grad_norm": 1.607157588005066, "learning_rate": 3.4952833485666505e-05, "loss": 2.6154, "step": 4414 }, { "epoch": 0.03970805808259096, "grad_norm": 1.4585423469543457, "learning_rate": 3.494593854610685e-05, "loss": 2.4836, "step": 4415 }, { "epoch": 0.03971705198023141, "grad_norm": 1.4144543409347534, "learning_rate": 3.493904270764398e-05, "loss": 2.3806, "step": 4416 }, { "epoch": 0.03972604587787187, "grad_norm": 1.6575531959533691, "learning_rate": 3.493214597090114e-05, "loss": 2.3582, "step": 4417 }, { "epoch": 0.039735039775512314, "grad_norm": 1.6461251974105835, "learning_rate": 3.4925248336501647e-05, "loss": 2.4329, "step": 4418 }, { "epoch": 0.03974403367315277, "grad_norm": 1.5483429431915283, "learning_rate": 3.4918349805068904e-05, "loss": 2.5624, "step": 4419 }, { "epoch": 0.039753027570793215, "grad_norm": 1.8989605903625488, "learning_rate": 3.4911450377226376e-05, "loss": 2.8697, "step": 4420 }, { "epoch": 0.03976202146843367, "grad_norm": 1.650023102760315, "learning_rate": 3.4904550053597644e-05, "loss": 2.7448, "step": 4421 }, { "epoch": 0.039771015366074115, "grad_norm": 2.0357933044433594, "learning_rate": 3.4897648834806346e-05, "loss": 2.7076, "step": 4422 }, { "epoch": 0.03978000926371457, "grad_norm": 1.4775516986846924, "learning_rate": 3.489074672147621e-05, "loss": 2.5032, "step": 4423 }, { "epoch": 0.03978900316135502, "grad_norm": 0.9905308485031128, "learning_rate": 3.488384371423103e-05, "loss": 2.2327, "step": 4424 }, { "epoch": 0.03979799705899547, "grad_norm": 2.209672212600708, "learning_rate": 3.487693981369472e-05, "loss": 2.5062, "step": 4425 }, { "epoch": 0.039806990956635924, "grad_norm": 1.4259769916534424, "learning_rate": 3.487003502049122e-05, "loss": 2.4929, "step": 4426 }, { "epoch": 0.03981598485427637, "grad_norm": 1.531160593032837, "learning_rate": 3.486312933524457e-05, "loss": 2.701, "step": 4427 }, { "epoch": 0.039824978751916824, "grad_norm": 2.003316640853882, "learning_rate": 3.485622275857893e-05, "loss": 2.6786, "step": 4428 }, { "epoch": 0.03983397264955728, "grad_norm": 2.143021821975708, "learning_rate": 3.484931529111849e-05, "loss": 2.8903, "step": 4429 }, { "epoch": 0.039842966547197725, "grad_norm": 1.134189248085022, "learning_rate": 3.484240693348754e-05, "loss": 2.4817, "step": 4430 }, { "epoch": 0.03985196044483818, "grad_norm": 1.1328734159469604, "learning_rate": 3.483549768631046e-05, "loss": 2.4301, "step": 4431 }, { "epoch": 0.039860954342478626, "grad_norm": 1.0830706357955933, "learning_rate": 3.482858755021169e-05, "loss": 2.3771, "step": 4432 }, { "epoch": 0.03986994824011908, "grad_norm": 1.5341674089431763, "learning_rate": 3.482167652581576e-05, "loss": 2.595, "step": 4433 }, { "epoch": 0.039878942137759534, "grad_norm": 1.6236790418624878, "learning_rate": 3.4814764613747286e-05, "loss": 2.6369, "step": 4434 }, { "epoch": 0.03988793603539998, "grad_norm": 2.0757153034210205, "learning_rate": 3.480785181463096e-05, "loss": 2.9628, "step": 4435 }, { "epoch": 0.039896929933040434, "grad_norm": 2.1510813236236572, "learning_rate": 3.4800938129091546e-05, "loss": 2.8639, "step": 4436 }, { "epoch": 0.03990592383068088, "grad_norm": 2.207315683364868, "learning_rate": 3.4794023557753905e-05, "loss": 2.8288, "step": 4437 }, { "epoch": 0.039914917728321335, "grad_norm": 1.5657546520233154, "learning_rate": 3.478710810124296e-05, "loss": 2.6311, "step": 4438 }, { "epoch": 0.03992391162596178, "grad_norm": 2.174151659011841, "learning_rate": 3.478019176018374e-05, "loss": 2.5777, "step": 4439 }, { "epoch": 0.039932905523602236, "grad_norm": 2.2875301837921143, "learning_rate": 3.477327453520131e-05, "loss": 2.7791, "step": 4440 }, { "epoch": 0.03994189942124269, "grad_norm": 1.6117360591888428, "learning_rate": 3.4766356426920854e-05, "loss": 2.5424, "step": 4441 }, { "epoch": 0.039950893318883136, "grad_norm": 1.8857017755508423, "learning_rate": 3.475943743596763e-05, "loss": 2.7944, "step": 4442 }, { "epoch": 0.03995988721652359, "grad_norm": 1.6579383611679077, "learning_rate": 3.475251756296696e-05, "loss": 2.8687, "step": 4443 }, { "epoch": 0.03996888111416404, "grad_norm": 1.9988772869110107, "learning_rate": 3.4745596808544255e-05, "loss": 2.7117, "step": 4444 }, { "epoch": 0.03997787501180449, "grad_norm": 1.342341423034668, "learning_rate": 3.473867517332501e-05, "loss": 2.5034, "step": 4445 }, { "epoch": 0.039986868909444945, "grad_norm": 0.8930309414863586, "learning_rate": 3.4731752657934794e-05, "loss": 2.2383, "step": 4446 }, { "epoch": 0.03999586280708539, "grad_norm": 1.6584042310714722, "learning_rate": 3.472482926299925e-05, "loss": 2.6804, "step": 4447 }, { "epoch": 0.040004856704725845, "grad_norm": 2.0250391960144043, "learning_rate": 3.4717904989144116e-05, "loss": 2.542, "step": 4448 }, { "epoch": 0.04001385060236629, "grad_norm": 1.6680017709732056, "learning_rate": 3.4710979836995196e-05, "loss": 2.4203, "step": 4449 }, { "epoch": 0.040022844500006746, "grad_norm": 1.6525627374649048, "learning_rate": 3.470405380717837e-05, "loss": 2.7034, "step": 4450 }, { "epoch": 0.04003183839764719, "grad_norm": 1.646492838859558, "learning_rate": 3.469712690031962e-05, "loss": 2.631, "step": 4451 }, { "epoch": 0.04004083229528765, "grad_norm": 1.8490898609161377, "learning_rate": 3.4690199117044984e-05, "loss": 2.7012, "step": 4452 }, { "epoch": 0.0400498261929281, "grad_norm": 1.8973115682601929, "learning_rate": 3.4683270457980596e-05, "loss": 2.5221, "step": 4453 }, { "epoch": 0.04005882009056855, "grad_norm": 1.6556031703948975, "learning_rate": 3.467634092375265e-05, "loss": 2.8215, "step": 4454 }, { "epoch": 0.040067813988209, "grad_norm": 1.9410135746002197, "learning_rate": 3.466941051498743e-05, "loss": 2.7658, "step": 4455 }, { "epoch": 0.04007680788584945, "grad_norm": 1.7445154190063477, "learning_rate": 3.4662479232311306e-05, "loss": 2.566, "step": 4456 }, { "epoch": 0.0400858017834899, "grad_norm": 2.4811577796936035, "learning_rate": 3.465554707635072e-05, "loss": 2.7698, "step": 4457 }, { "epoch": 0.040094795681130356, "grad_norm": 1.096279263496399, "learning_rate": 3.464861404773218e-05, "loss": 2.2764, "step": 4458 }, { "epoch": 0.0401037895787708, "grad_norm": 1.4868017435073853, "learning_rate": 3.464168014708231e-05, "loss": 2.6138, "step": 4459 }, { "epoch": 0.04011278347641126, "grad_norm": 1.756840705871582, "learning_rate": 3.4634745375027766e-05, "loss": 2.6191, "step": 4460 }, { "epoch": 0.040121777374051704, "grad_norm": 1.2942440509796143, "learning_rate": 3.46278097321953e-05, "loss": 2.5027, "step": 4461 }, { "epoch": 0.04013077127169216, "grad_norm": 1.5372575521469116, "learning_rate": 3.462087321921178e-05, "loss": 2.4667, "step": 4462 }, { "epoch": 0.04013976516933261, "grad_norm": 1.5835827589035034, "learning_rate": 3.4613935836704085e-05, "loss": 2.4364, "step": 4463 }, { "epoch": 0.04014875906697306, "grad_norm": 2.036045551300049, "learning_rate": 3.460699758529923e-05, "loss": 2.7655, "step": 4464 }, { "epoch": 0.04015775296461351, "grad_norm": 1.1974716186523438, "learning_rate": 3.460005846562428e-05, "loss": 2.6474, "step": 4465 }, { "epoch": 0.04016674686225396, "grad_norm": 2.375235080718994, "learning_rate": 3.4593118478306395e-05, "loss": 2.6763, "step": 4466 }, { "epoch": 0.04017574075989441, "grad_norm": 1.5288751125335693, "learning_rate": 3.458617762397279e-05, "loss": 2.4722, "step": 4467 }, { "epoch": 0.04018473465753486, "grad_norm": 1.7602866888046265, "learning_rate": 3.457923590325078e-05, "loss": 2.6297, "step": 4468 }, { "epoch": 0.04019372855517531, "grad_norm": 1.6864972114562988, "learning_rate": 3.457229331676774e-05, "loss": 2.6688, "step": 4469 }, { "epoch": 0.04020272245281577, "grad_norm": 1.3883715867996216, "learning_rate": 3.4565349865151145e-05, "loss": 2.4815, "step": 4470 }, { "epoch": 0.040211716350456214, "grad_norm": 1.700974702835083, "learning_rate": 3.4558405549028526e-05, "loss": 2.6083, "step": 4471 }, { "epoch": 0.04022071024809667, "grad_norm": 1.5239630937576294, "learning_rate": 3.4551460369027515e-05, "loss": 2.4352, "step": 4472 }, { "epoch": 0.040229704145737115, "grad_norm": 1.8302757740020752, "learning_rate": 3.45445143257758e-05, "loss": 2.8833, "step": 4473 }, { "epoch": 0.04023869804337757, "grad_norm": 1.8634439706802368, "learning_rate": 3.453756741990117e-05, "loss": 2.8475, "step": 4474 }, { "epoch": 0.04024769194101802, "grad_norm": 1.4408118724822998, "learning_rate": 3.453061965203146e-05, "loss": 2.4627, "step": 4475 }, { "epoch": 0.04025668583865847, "grad_norm": 1.31882905960083, "learning_rate": 3.4523671022794616e-05, "loss": 2.5511, "step": 4476 }, { "epoch": 0.04026567973629892, "grad_norm": 1.2851413488388062, "learning_rate": 3.4516721532818634e-05, "loss": 2.4882, "step": 4477 }, { "epoch": 0.04027467363393937, "grad_norm": 1.8844201564788818, "learning_rate": 3.450977118273162e-05, "loss": 2.4979, "step": 4478 }, { "epoch": 0.040283667531579824, "grad_norm": 1.7700430154800415, "learning_rate": 3.4502819973161726e-05, "loss": 2.9012, "step": 4479 }, { "epoch": 0.04029266142922027, "grad_norm": 1.6792809963226318, "learning_rate": 3.44958679047372e-05, "loss": 2.4297, "step": 4480 }, { "epoch": 0.040301655326860725, "grad_norm": 1.6150203943252563, "learning_rate": 3.448891497808636e-05, "loss": 2.6351, "step": 4481 }, { "epoch": 0.04031064922450118, "grad_norm": 1.0985921621322632, "learning_rate": 3.448196119383761e-05, "loss": 2.3564, "step": 4482 }, { "epoch": 0.040319643122141625, "grad_norm": 1.9644205570220947, "learning_rate": 3.447500655261941e-05, "loss": 2.8218, "step": 4483 }, { "epoch": 0.04032863701978208, "grad_norm": 1.7132774591445923, "learning_rate": 3.4468051055060335e-05, "loss": 2.5403, "step": 4484 }, { "epoch": 0.040337630917422526, "grad_norm": 1.4813334941864014, "learning_rate": 3.4461094701788994e-05, "loss": 2.812, "step": 4485 }, { "epoch": 0.04034662481506298, "grad_norm": 2.1658928394317627, "learning_rate": 3.445413749343411e-05, "loss": 2.8571, "step": 4486 }, { "epoch": 0.040355618712703434, "grad_norm": 2.029980182647705, "learning_rate": 3.444717943062446e-05, "loss": 2.5587, "step": 4487 }, { "epoch": 0.04036461261034388, "grad_norm": 1.1957958936691284, "learning_rate": 3.444022051398891e-05, "loss": 2.3306, "step": 4488 }, { "epoch": 0.040373606507984335, "grad_norm": 2.160168409347534, "learning_rate": 3.4433260744156396e-05, "loss": 2.57, "step": 4489 }, { "epoch": 0.04038260040562478, "grad_norm": 1.827613353729248, "learning_rate": 3.442630012175594e-05, "loss": 2.7772, "step": 4490 }, { "epoch": 0.040391594303265235, "grad_norm": 1.4935654401779175, "learning_rate": 3.4419338647416624e-05, "loss": 2.575, "step": 4491 }, { "epoch": 0.04040058820090568, "grad_norm": 0.6489777565002441, "learning_rate": 3.4412376321767634e-05, "loss": 2.2031, "step": 4492 }, { "epoch": 0.040409582098546136, "grad_norm": 1.3631253242492676, "learning_rate": 3.440541314543821e-05, "loss": 2.3248, "step": 4493 }, { "epoch": 0.04041857599618659, "grad_norm": 1.762289047241211, "learning_rate": 3.439844911905768e-05, "loss": 2.7871, "step": 4494 }, { "epoch": 0.04042756989382704, "grad_norm": 1.7072608470916748, "learning_rate": 3.439148424325544e-05, "loss": 2.9286, "step": 4495 }, { "epoch": 0.04043656379146749, "grad_norm": 1.9634130001068115, "learning_rate": 3.4384518518660966e-05, "loss": 2.937, "step": 4496 }, { "epoch": 0.04044555768910794, "grad_norm": 1.2109774351119995, "learning_rate": 3.437755194590382e-05, "loss": 2.4719, "step": 4497 }, { "epoch": 0.04045455158674839, "grad_norm": 2.03463077545166, "learning_rate": 3.4370584525613626e-05, "loss": 2.9721, "step": 4498 }, { "epoch": 0.040463545484388845, "grad_norm": 2.147233486175537, "learning_rate": 3.436361625842009e-05, "loss": 2.4693, "step": 4499 }, { "epoch": 0.04047253938202929, "grad_norm": 1.6777721643447876, "learning_rate": 3.435664714495301e-05, "loss": 2.6647, "step": 4500 }, { "epoch": 0.040481533279669746, "grad_norm": 1.3682037591934204, "learning_rate": 3.4349677185842245e-05, "loss": 2.419, "step": 4501 }, { "epoch": 0.04049052717731019, "grad_norm": 1.2877957820892334, "learning_rate": 3.4342706381717724e-05, "loss": 2.3347, "step": 4502 }, { "epoch": 0.040499521074950647, "grad_norm": 1.927704930305481, "learning_rate": 3.4335734733209456e-05, "loss": 2.6737, "step": 4503 }, { "epoch": 0.0405085149725911, "grad_norm": 1.489983320236206, "learning_rate": 3.432876224094754e-05, "loss": 2.3987, "step": 4504 }, { "epoch": 0.04051750887023155, "grad_norm": 1.9137423038482666, "learning_rate": 3.432178890556214e-05, "loss": 2.7115, "step": 4505 }, { "epoch": 0.040526502767872, "grad_norm": 1.9031157493591309, "learning_rate": 3.43148147276835e-05, "loss": 2.7025, "step": 4506 }, { "epoch": 0.04053549666551245, "grad_norm": 2.0486972332000732, "learning_rate": 3.430783970794195e-05, "loss": 2.6409, "step": 4507 }, { "epoch": 0.0405444905631529, "grad_norm": 1.305566668510437, "learning_rate": 3.4300863846967855e-05, "loss": 2.4271, "step": 4508 }, { "epoch": 0.04055348446079335, "grad_norm": 1.3594260215759277, "learning_rate": 3.429388714539172e-05, "loss": 2.4814, "step": 4509 }, { "epoch": 0.0405624783584338, "grad_norm": 2.0376737117767334, "learning_rate": 3.4286909603844065e-05, "loss": 2.7429, "step": 4510 }, { "epoch": 0.040571472256074256, "grad_norm": 1.8565876483917236, "learning_rate": 3.427993122295552e-05, "loss": 2.7838, "step": 4511 }, { "epoch": 0.0405804661537147, "grad_norm": 1.471709966659546, "learning_rate": 3.427295200335679e-05, "loss": 2.4019, "step": 4512 }, { "epoch": 0.04058946005135516, "grad_norm": 1.5323163270950317, "learning_rate": 3.4265971945678656e-05, "loss": 2.7501, "step": 4513 }, { "epoch": 0.040598453948995604, "grad_norm": 1.3368377685546875, "learning_rate": 3.425899105055194e-05, "loss": 2.3621, "step": 4514 }, { "epoch": 0.04060744784663606, "grad_norm": 1.8829818964004517, "learning_rate": 3.4252009318607596e-05, "loss": 2.6378, "step": 4515 }, { "epoch": 0.04061644174427651, "grad_norm": 1.703145980834961, "learning_rate": 3.424502675047662e-05, "loss": 2.7413, "step": 4516 }, { "epoch": 0.04062543564191696, "grad_norm": 1.9699541330337524, "learning_rate": 3.423804334679008e-05, "loss": 2.6654, "step": 4517 }, { "epoch": 0.04063442953955741, "grad_norm": 1.1089390516281128, "learning_rate": 3.4231059108179124e-05, "loss": 2.3331, "step": 4518 }, { "epoch": 0.04064342343719786, "grad_norm": 1.6895328760147095, "learning_rate": 3.4224074035275e-05, "loss": 2.5687, "step": 4519 }, { "epoch": 0.04065241733483831, "grad_norm": 1.1253955364227295, "learning_rate": 3.4217088128708996e-05, "loss": 2.4489, "step": 4520 }, { "epoch": 0.04066141123247876, "grad_norm": 1.767910361289978, "learning_rate": 3.421010138911249e-05, "loss": 2.62, "step": 4521 }, { "epoch": 0.040670405130119214, "grad_norm": 1.2594718933105469, "learning_rate": 3.4203113817116957e-05, "loss": 2.3721, "step": 4522 }, { "epoch": 0.04067939902775967, "grad_norm": 1.66157865524292, "learning_rate": 3.41961254133539e-05, "loss": 2.8821, "step": 4523 }, { "epoch": 0.040688392925400114, "grad_norm": 1.711624026298523, "learning_rate": 3.418913617845492e-05, "loss": 2.6468, "step": 4524 }, { "epoch": 0.04069738682304057, "grad_norm": 1.8693872690200806, "learning_rate": 3.4182146113051724e-05, "loss": 2.8611, "step": 4525 }, { "epoch": 0.040706380720681015, "grad_norm": 1.7919337749481201, "learning_rate": 3.4175155217776055e-05, "loss": 2.5152, "step": 4526 }, { "epoch": 0.04071537461832147, "grad_norm": 2.169097661972046, "learning_rate": 3.416816349325974e-05, "loss": 2.7601, "step": 4527 }, { "epoch": 0.04072436851596192, "grad_norm": 2.1731159687042236, "learning_rate": 3.4161170940134674e-05, "loss": 2.6308, "step": 4528 }, { "epoch": 0.04073336241360237, "grad_norm": 1.6365631818771362, "learning_rate": 3.415417755903285e-05, "loss": 2.4543, "step": 4529 }, { "epoch": 0.040742356311242824, "grad_norm": 1.2217626571655273, "learning_rate": 3.414718335058632e-05, "loss": 2.5604, "step": 4530 }, { "epoch": 0.04075135020888327, "grad_norm": 1.8155542612075806, "learning_rate": 3.414018831542721e-05, "loss": 2.7137, "step": 4531 }, { "epoch": 0.040760344106523724, "grad_norm": 1.8236017227172852, "learning_rate": 3.413319245418773e-05, "loss": 2.7263, "step": 4532 }, { "epoch": 0.04076933800416417, "grad_norm": 2.212207317352295, "learning_rate": 3.412619576750014e-05, "loss": 2.6846, "step": 4533 }, { "epoch": 0.040778331901804625, "grad_norm": 1.4749689102172852, "learning_rate": 3.411919825599682e-05, "loss": 2.5617, "step": 4534 }, { "epoch": 0.04078732579944508, "grad_norm": 1.2227028608322144, "learning_rate": 3.411219992031018e-05, "loss": 2.3344, "step": 4535 }, { "epoch": 0.040796319697085526, "grad_norm": 1.3611161708831787, "learning_rate": 3.410520076107273e-05, "loss": 2.4367, "step": 4536 }, { "epoch": 0.04080531359472598, "grad_norm": 1.6731725931167603, "learning_rate": 3.409820077891704e-05, "loss": 3.1346, "step": 4537 }, { "epoch": 0.040814307492366426, "grad_norm": 1.790431261062622, "learning_rate": 3.409119997447576e-05, "loss": 2.7567, "step": 4538 }, { "epoch": 0.04082330139000688, "grad_norm": 1.8608852624893188, "learning_rate": 3.408419834838162e-05, "loss": 2.7446, "step": 4539 }, { "epoch": 0.040832295287647334, "grad_norm": 2.216593027114868, "learning_rate": 3.4077195901267416e-05, "loss": 2.371, "step": 4540 }, { "epoch": 0.04084128918528778, "grad_norm": 1.81519615650177, "learning_rate": 3.4070192633766025e-05, "loss": 2.484, "step": 4541 }, { "epoch": 0.040850283082928235, "grad_norm": 1.705349326133728, "learning_rate": 3.406318854651039e-05, "loss": 2.2296, "step": 4542 }, { "epoch": 0.04085927698056868, "grad_norm": 1.6653869152069092, "learning_rate": 3.405618364013355e-05, "loss": 2.6356, "step": 4543 }, { "epoch": 0.040868270878209136, "grad_norm": 2.0056703090667725, "learning_rate": 3.4049177915268576e-05, "loss": 2.5356, "step": 4544 }, { "epoch": 0.04087726477584959, "grad_norm": 2.012026309967041, "learning_rate": 3.404217137254865e-05, "loss": 2.7394, "step": 4545 }, { "epoch": 0.040886258673490036, "grad_norm": 2.070591449737549, "learning_rate": 3.403516401260701e-05, "loss": 2.4474, "step": 4546 }, { "epoch": 0.04089525257113049, "grad_norm": 1.8404104709625244, "learning_rate": 3.402815583607699e-05, "loss": 2.4538, "step": 4547 }, { "epoch": 0.04090424646877094, "grad_norm": 1.6194474697113037, "learning_rate": 3.402114684359196e-05, "loss": 2.6018, "step": 4548 }, { "epoch": 0.04091324036641139, "grad_norm": 1.331263780593872, "learning_rate": 3.40141370357854e-05, "loss": 2.5448, "step": 4549 }, { "epoch": 0.04092223426405184, "grad_norm": 1.5331249237060547, "learning_rate": 3.400712641329085e-05, "loss": 2.3865, "step": 4550 }, { "epoch": 0.04093122816169229, "grad_norm": 2.546968698501587, "learning_rate": 3.4000114976741906e-05, "loss": 2.6718, "step": 4551 }, { "epoch": 0.040940222059332745, "grad_norm": 1.736614465713501, "learning_rate": 3.3993102726772263e-05, "loss": 2.2962, "step": 4552 }, { "epoch": 0.04094921595697319, "grad_norm": 2.0874712467193604, "learning_rate": 3.398608966401568e-05, "loss": 2.3755, "step": 4553 }, { "epoch": 0.040958209854613646, "grad_norm": 1.640486717224121, "learning_rate": 3.397907578910601e-05, "loss": 2.7328, "step": 4554 }, { "epoch": 0.04096720375225409, "grad_norm": 1.5212615728378296, "learning_rate": 3.397206110267713e-05, "loss": 2.5586, "step": 4555 }, { "epoch": 0.04097619764989455, "grad_norm": 1.4394758939743042, "learning_rate": 3.396504560536303e-05, "loss": 2.6906, "step": 4556 }, { "epoch": 0.040985191547535, "grad_norm": 1.7915948629379272, "learning_rate": 3.3958029297797765e-05, "loss": 2.4545, "step": 4557 }, { "epoch": 0.04099418544517545, "grad_norm": 1.9176812171936035, "learning_rate": 3.3951012180615465e-05, "loss": 2.4578, "step": 4558 }, { "epoch": 0.0410031793428159, "grad_norm": 1.6473138332366943, "learning_rate": 3.3943994254450326e-05, "loss": 2.4772, "step": 4559 }, { "epoch": 0.04101217324045635, "grad_norm": 1.5448583364486694, "learning_rate": 3.3936975519936615e-05, "loss": 2.3364, "step": 4560 }, { "epoch": 0.0410211671380968, "grad_norm": 2.9649338722229004, "learning_rate": 3.392995597770868e-05, "loss": 2.8273, "step": 4561 }, { "epoch": 0.04103016103573725, "grad_norm": 1.5005501508712769, "learning_rate": 3.3922935628400946e-05, "loss": 2.802, "step": 4562 }, { "epoch": 0.0410391549333777, "grad_norm": 2.2600884437561035, "learning_rate": 3.3915914472647905e-05, "loss": 3.0044, "step": 4563 }, { "epoch": 0.04104814883101816, "grad_norm": 1.529415249824524, "learning_rate": 3.390889251108412e-05, "loss": 2.5899, "step": 4564 }, { "epoch": 0.041057142728658604, "grad_norm": 1.412465214729309, "learning_rate": 3.390186974434422e-05, "loss": 2.5551, "step": 4565 }, { "epoch": 0.04106613662629906, "grad_norm": 1.8531643152236938, "learning_rate": 3.389484617306292e-05, "loss": 2.7161, "step": 4566 }, { "epoch": 0.041075130523939504, "grad_norm": 1.5859479904174805, "learning_rate": 3.3887821797875e-05, "loss": 2.5086, "step": 4567 }, { "epoch": 0.04108412442157996, "grad_norm": 1.07433021068573, "learning_rate": 3.3880796619415325e-05, "loss": 2.3211, "step": 4568 }, { "epoch": 0.04109311831922041, "grad_norm": 1.9476953744888306, "learning_rate": 3.387377063831882e-05, "loss": 2.6793, "step": 4569 }, { "epoch": 0.04110211221686086, "grad_norm": 1.6459232568740845, "learning_rate": 3.3866743855220466e-05, "loss": 2.4452, "step": 4570 }, { "epoch": 0.04111110611450131, "grad_norm": 2.4576823711395264, "learning_rate": 3.385971627075537e-05, "loss": 2.987, "step": 4571 }, { "epoch": 0.04112010001214176, "grad_norm": 1.5391840934753418, "learning_rate": 3.385268788555866e-05, "loss": 2.5501, "step": 4572 }, { "epoch": 0.04112909390978221, "grad_norm": 1.842389464378357, "learning_rate": 3.384565870026555e-05, "loss": 2.8069, "step": 4573 }, { "epoch": 0.04113808780742267, "grad_norm": 1.5965651273727417, "learning_rate": 3.383862871551133e-05, "loss": 2.6844, "step": 4574 }, { "epoch": 0.041147081705063114, "grad_norm": 1.4844931364059448, "learning_rate": 3.3831597931931366e-05, "loss": 2.3938, "step": 4575 }, { "epoch": 0.04115607560270357, "grad_norm": 2.0001444816589355, "learning_rate": 3.38245663501611e-05, "loss": 2.6596, "step": 4576 }, { "epoch": 0.041165069500344015, "grad_norm": 1.378292441368103, "learning_rate": 3.381753397083604e-05, "loss": 2.6313, "step": 4577 }, { "epoch": 0.04117406339798447, "grad_norm": 1.7484112977981567, "learning_rate": 3.381050079459174e-05, "loss": 2.7941, "step": 4578 }, { "epoch": 0.041183057295624916, "grad_norm": 2.3297412395477295, "learning_rate": 3.3803466822063876e-05, "loss": 3.0041, "step": 4579 }, { "epoch": 0.04119205119326537, "grad_norm": 1.6530609130859375, "learning_rate": 3.379643205388816e-05, "loss": 2.6387, "step": 4580 }, { "epoch": 0.04120104509090582, "grad_norm": 1.8463517427444458, "learning_rate": 3.378939649070039e-05, "loss": 2.7388, "step": 4581 }, { "epoch": 0.04121003898854627, "grad_norm": 1.535904884338379, "learning_rate": 3.3782360133136436e-05, "loss": 2.9111, "step": 4582 }, { "epoch": 0.041219032886186724, "grad_norm": 1.6279395818710327, "learning_rate": 3.377532298183223e-05, "loss": 2.4783, "step": 4583 }, { "epoch": 0.04122802678382717, "grad_norm": 1.798724889755249, "learning_rate": 3.376828503742379e-05, "loss": 2.6048, "step": 4584 }, { "epoch": 0.041237020681467625, "grad_norm": 1.5005005598068237, "learning_rate": 3.3761246300547195e-05, "loss": 2.5365, "step": 4585 }, { "epoch": 0.04124601457910808, "grad_norm": 1.480992317199707, "learning_rate": 3.3754206771838586e-05, "loss": 2.3769, "step": 4586 }, { "epoch": 0.041255008476748525, "grad_norm": 1.798842191696167, "learning_rate": 3.374716645193421e-05, "loss": 2.5815, "step": 4587 }, { "epoch": 0.04126400237438898, "grad_norm": 1.1243653297424316, "learning_rate": 3.3740125341470336e-05, "loss": 2.4424, "step": 4588 }, { "epoch": 0.041272996272029426, "grad_norm": 1.381732702255249, "learning_rate": 3.373308344108336e-05, "loss": 2.4909, "step": 4589 }, { "epoch": 0.04128199016966988, "grad_norm": 1.3383620977401733, "learning_rate": 3.37260407514097e-05, "loss": 2.6066, "step": 4590 }, { "epoch": 0.04129098406731033, "grad_norm": 1.5011320114135742, "learning_rate": 3.371899727308588e-05, "loss": 2.4809, "step": 4591 }, { "epoch": 0.04129997796495078, "grad_norm": 2.196830987930298, "learning_rate": 3.3711953006748484e-05, "loss": 2.8057, "step": 4592 }, { "epoch": 0.041308971862591234, "grad_norm": 1.4781379699707031, "learning_rate": 3.3704907953034156e-05, "loss": 2.5931, "step": 4593 }, { "epoch": 0.04131796576023168, "grad_norm": 2.405683755874634, "learning_rate": 3.3697862112579614e-05, "loss": 2.8353, "step": 4594 }, { "epoch": 0.041326959657872135, "grad_norm": 2.1085116863250732, "learning_rate": 3.369081548602167e-05, "loss": 2.5366, "step": 4595 }, { "epoch": 0.04133595355551258, "grad_norm": 1.5938177108764648, "learning_rate": 3.368376807399719e-05, "loss": 2.3557, "step": 4596 }, { "epoch": 0.041344947453153036, "grad_norm": 1.4676162004470825, "learning_rate": 3.367671987714309e-05, "loss": 2.5809, "step": 4597 }, { "epoch": 0.04135394135079349, "grad_norm": 1.7108780145645142, "learning_rate": 3.366967089609641e-05, "loss": 2.6274, "step": 4598 }, { "epoch": 0.04136293524843394, "grad_norm": 1.1815482378005981, "learning_rate": 3.3662621131494204e-05, "loss": 2.3358, "step": 4599 }, { "epoch": 0.04137192914607439, "grad_norm": 2.3759050369262695, "learning_rate": 3.3655570583973626e-05, "loss": 3.0781, "step": 4600 }, { "epoch": 0.04138092304371484, "grad_norm": 1.5695337057113647, "learning_rate": 3.364851925417191e-05, "loss": 2.4684, "step": 4601 }, { "epoch": 0.04138991694135529, "grad_norm": 1.6752275228500366, "learning_rate": 3.3641467142726335e-05, "loss": 2.527, "step": 4602 }, { "epoch": 0.04139891083899574, "grad_norm": 1.283116340637207, "learning_rate": 3.363441425027427e-05, "loss": 2.5367, "step": 4603 }, { "epoch": 0.04140790473663619, "grad_norm": 1.3377468585968018, "learning_rate": 3.362736057745314e-05, "loss": 2.3675, "step": 4604 }, { "epoch": 0.041416898634276646, "grad_norm": 1.0460261106491089, "learning_rate": 3.362030612490046e-05, "loss": 2.3839, "step": 4605 }, { "epoch": 0.04142589253191709, "grad_norm": 1.961141586303711, "learning_rate": 3.3613250893253796e-05, "loss": 2.6839, "step": 4606 }, { "epoch": 0.041434886429557546, "grad_norm": 1.8977138996124268, "learning_rate": 3.3606194883150785e-05, "loss": 2.5074, "step": 4607 }, { "epoch": 0.04144388032719799, "grad_norm": 1.5131011009216309, "learning_rate": 3.359913809522917e-05, "loss": 2.6798, "step": 4608 }, { "epoch": 0.04145287422483845, "grad_norm": 1.6956440210342407, "learning_rate": 3.3592080530126695e-05, "loss": 2.905, "step": 4609 }, { "epoch": 0.0414618681224789, "grad_norm": 1.4057440757751465, "learning_rate": 3.358502218848125e-05, "loss": 2.4178, "step": 4610 }, { "epoch": 0.04147086202011935, "grad_norm": 1.9208256006240845, "learning_rate": 3.357796307093074e-05, "loss": 2.5447, "step": 4611 }, { "epoch": 0.0414798559177598, "grad_norm": 1.6720367670059204, "learning_rate": 3.357090317811317e-05, "loss": 2.648, "step": 4612 }, { "epoch": 0.04148884981540025, "grad_norm": 0.7702749967575073, "learning_rate": 3.3563842510666596e-05, "loss": 2.2464, "step": 4613 }, { "epoch": 0.0414978437130407, "grad_norm": 2.3850934505462646, "learning_rate": 3.355678106922917e-05, "loss": 2.9942, "step": 4614 }, { "epoch": 0.041506837610681156, "grad_norm": 2.2529892921447754, "learning_rate": 3.354971885443908e-05, "loss": 2.7996, "step": 4615 }, { "epoch": 0.0415158315083216, "grad_norm": 1.7430754899978638, "learning_rate": 3.354265586693461e-05, "loss": 2.5012, "step": 4616 }, { "epoch": 0.04152482540596206, "grad_norm": 1.8782989978790283, "learning_rate": 3.353559210735411e-05, "loss": 2.8207, "step": 4617 }, { "epoch": 0.041533819303602504, "grad_norm": 1.5955222845077515, "learning_rate": 3.352852757633598e-05, "loss": 2.6702, "step": 4618 }, { "epoch": 0.04154281320124296, "grad_norm": 1.5827083587646484, "learning_rate": 3.352146227451872e-05, "loss": 2.6058, "step": 4619 }, { "epoch": 0.041551807098883405, "grad_norm": 1.2010825872421265, "learning_rate": 3.351439620254088e-05, "loss": 2.3895, "step": 4620 }, { "epoch": 0.04156080099652386, "grad_norm": 2.451399564743042, "learning_rate": 3.350732936104108e-05, "loss": 2.6379, "step": 4621 }, { "epoch": 0.04156979489416431, "grad_norm": 1.447615385055542, "learning_rate": 3.3500261750658015e-05, "loss": 2.5274, "step": 4622 }, { "epoch": 0.04157878879180476, "grad_norm": 2.0332369804382324, "learning_rate": 3.3493193372030444e-05, "loss": 3.1782, "step": 4623 }, { "epoch": 0.04158778268944521, "grad_norm": 1.5351821184158325, "learning_rate": 3.348612422579721e-05, "loss": 2.4618, "step": 4624 }, { "epoch": 0.04159677658708566, "grad_norm": 1.2563327550888062, "learning_rate": 3.347905431259721e-05, "loss": 2.4104, "step": 4625 }, { "epoch": 0.041605770484726114, "grad_norm": 1.7778736352920532, "learning_rate": 3.347198363306942e-05, "loss": 2.4391, "step": 4626 }, { "epoch": 0.04161476438236657, "grad_norm": 1.5894896984100342, "learning_rate": 3.346491218785286e-05, "loss": 2.6834, "step": 4627 }, { "epoch": 0.041623758280007014, "grad_norm": 1.5762192010879517, "learning_rate": 3.345783997758666e-05, "loss": 2.4112, "step": 4628 }, { "epoch": 0.04163275217764747, "grad_norm": 1.5952914953231812, "learning_rate": 3.345076700291e-05, "loss": 2.5754, "step": 4629 }, { "epoch": 0.041641746075287915, "grad_norm": 1.690817952156067, "learning_rate": 3.3443693264462105e-05, "loss": 2.6092, "step": 4630 }, { "epoch": 0.04165073997292837, "grad_norm": 1.4484082460403442, "learning_rate": 3.343661876288232e-05, "loss": 2.3399, "step": 4631 }, { "epoch": 0.041659733870568816, "grad_norm": 1.2962037324905396, "learning_rate": 3.342954349881001e-05, "loss": 2.4756, "step": 4632 }, { "epoch": 0.04166872776820927, "grad_norm": 1.4534395933151245, "learning_rate": 3.342246747288465e-05, "loss": 2.5564, "step": 4633 }, { "epoch": 0.041677721665849723, "grad_norm": 1.6429251432418823, "learning_rate": 3.341539068574574e-05, "loss": 2.5459, "step": 4634 }, { "epoch": 0.04168671556349017, "grad_norm": 1.7204504013061523, "learning_rate": 3.3408313138032894e-05, "loss": 2.7779, "step": 4635 }, { "epoch": 0.041695709461130624, "grad_norm": 1.541718602180481, "learning_rate": 3.3401234830385756e-05, "loss": 2.4597, "step": 4636 }, { "epoch": 0.04170470335877107, "grad_norm": 2.1124956607818604, "learning_rate": 3.3394155763444066e-05, "loss": 2.7906, "step": 4637 }, { "epoch": 0.041713697256411525, "grad_norm": 1.579913854598999, "learning_rate": 3.3387075937847614e-05, "loss": 2.6698, "step": 4638 }, { "epoch": 0.04172269115405198, "grad_norm": 2.334123373031616, "learning_rate": 3.337999535423628e-05, "loss": 2.9127, "step": 4639 }, { "epoch": 0.041731685051692426, "grad_norm": 1.5778982639312744, "learning_rate": 3.337291401324999e-05, "loss": 2.5995, "step": 4640 }, { "epoch": 0.04174067894933288, "grad_norm": 1.4730827808380127, "learning_rate": 3.336583191552876e-05, "loss": 2.5253, "step": 4641 }, { "epoch": 0.041749672846973326, "grad_norm": 1.7080423831939697, "learning_rate": 3.3358749061712636e-05, "loss": 2.6087, "step": 4642 }, { "epoch": 0.04175866674461378, "grad_norm": 1.9072245359420776, "learning_rate": 3.335166545244178e-05, "loss": 2.6923, "step": 4643 }, { "epoch": 0.04176766064225423, "grad_norm": 1.6694685220718384, "learning_rate": 3.33445810883564e-05, "loss": 2.5307, "step": 4644 }, { "epoch": 0.04177665453989468, "grad_norm": 1.576827883720398, "learning_rate": 3.333749597009676e-05, "loss": 2.5448, "step": 4645 }, { "epoch": 0.041785648437535135, "grad_norm": 1.2934821844100952, "learning_rate": 3.3330410098303224e-05, "loss": 2.386, "step": 4646 }, { "epoch": 0.04179464233517558, "grad_norm": 1.8246935606002808, "learning_rate": 3.33233234736162e-05, "loss": 2.4604, "step": 4647 }, { "epoch": 0.041803636232816035, "grad_norm": 1.4678345918655396, "learning_rate": 3.331623609667615e-05, "loss": 2.559, "step": 4648 }, { "epoch": 0.04181263013045648, "grad_norm": 0.997654139995575, "learning_rate": 3.330914796812366e-05, "loss": 2.167, "step": 4649 }, { "epoch": 0.041821624028096936, "grad_norm": 1.4292304515838623, "learning_rate": 3.330205908859931e-05, "loss": 2.4796, "step": 4650 }, { "epoch": 0.04183061792573739, "grad_norm": 2.186976194381714, "learning_rate": 3.32949694587438e-05, "loss": 2.7528, "step": 4651 }, { "epoch": 0.04183961182337784, "grad_norm": 2.329317808151245, "learning_rate": 3.328787907919789e-05, "loss": 2.9671, "step": 4652 }, { "epoch": 0.04184860572101829, "grad_norm": 1.978234887123108, "learning_rate": 3.3280787950602396e-05, "loss": 2.5239, "step": 4653 }, { "epoch": 0.04185759961865874, "grad_norm": 2.285611629486084, "learning_rate": 3.327369607359821e-05, "loss": 2.4422, "step": 4654 }, { "epoch": 0.04186659351629919, "grad_norm": 0.4941359758377075, "learning_rate": 3.3266603448826285e-05, "loss": 2.2771, "step": 4655 }, { "epoch": 0.041875587413939645, "grad_norm": 1.0876973867416382, "learning_rate": 3.3259510076927644e-05, "loss": 2.312, "step": 4656 }, { "epoch": 0.04188458131158009, "grad_norm": 1.9674811363220215, "learning_rate": 3.325241595854337e-05, "loss": 2.9532, "step": 4657 }, { "epoch": 0.041893575209220546, "grad_norm": 1.954727053642273, "learning_rate": 3.3245321094314634e-05, "loss": 2.8345, "step": 4658 }, { "epoch": 0.04190256910686099, "grad_norm": 1.766262412071228, "learning_rate": 3.323822548488266e-05, "loss": 2.6944, "step": 4659 }, { "epoch": 0.04191156300450145, "grad_norm": 1.6745797395706177, "learning_rate": 3.3231129130888736e-05, "loss": 2.7347, "step": 4660 }, { "epoch": 0.041920556902141894, "grad_norm": 1.5070199966430664, "learning_rate": 3.322403203297424e-05, "loss": 2.6745, "step": 4661 }, { "epoch": 0.04192955079978235, "grad_norm": 1.539087176322937, "learning_rate": 3.321693419178059e-05, "loss": 2.6386, "step": 4662 }, { "epoch": 0.0419385446974228, "grad_norm": 1.2758136987686157, "learning_rate": 3.3209835607949256e-05, "loss": 2.6667, "step": 4663 }, { "epoch": 0.04194753859506325, "grad_norm": 1.7248706817626953, "learning_rate": 3.320273628212184e-05, "loss": 2.6953, "step": 4664 }, { "epoch": 0.0419565324927037, "grad_norm": 1.42571222782135, "learning_rate": 3.319563621493994e-05, "loss": 2.4099, "step": 4665 }, { "epoch": 0.04196552639034415, "grad_norm": 1.375084400177002, "learning_rate": 3.318853540704527e-05, "loss": 2.384, "step": 4666 }, { "epoch": 0.0419745202879846, "grad_norm": 1.7001525163650513, "learning_rate": 3.31814338590796e-05, "loss": 2.452, "step": 4667 }, { "epoch": 0.04198351418562506, "grad_norm": 1.8392138481140137, "learning_rate": 3.3174331571684745e-05, "loss": 2.901, "step": 4668 }, { "epoch": 0.0419925080832655, "grad_norm": 2.1236114501953125, "learning_rate": 3.3167228545502614e-05, "loss": 3.1402, "step": 4669 }, { "epoch": 0.04200150198090596, "grad_norm": 1.993749737739563, "learning_rate": 3.316012478117515e-05, "loss": 2.9297, "step": 4670 }, { "epoch": 0.042010495878546404, "grad_norm": 2.1093852519989014, "learning_rate": 3.315302027934441e-05, "loss": 3.2164, "step": 4671 }, { "epoch": 0.04201948977618686, "grad_norm": 2.006256580352783, "learning_rate": 3.314591504065247e-05, "loss": 2.6904, "step": 4672 }, { "epoch": 0.042028483673827305, "grad_norm": 1.2842860221862793, "learning_rate": 3.3138809065741513e-05, "loss": 2.5436, "step": 4673 }, { "epoch": 0.04203747757146776, "grad_norm": 1.2264803647994995, "learning_rate": 3.313170235525375e-05, "loss": 2.5199, "step": 4674 }, { "epoch": 0.04204647146910821, "grad_norm": 1.6028876304626465, "learning_rate": 3.312459490983149e-05, "loss": 2.6273, "step": 4675 }, { "epoch": 0.04205546536674866, "grad_norm": 1.693145990371704, "learning_rate": 3.311748673011709e-05, "loss": 2.6408, "step": 4676 }, { "epoch": 0.04206445926438911, "grad_norm": 1.558530330657959, "learning_rate": 3.311037781675299e-05, "loss": 2.3839, "step": 4677 }, { "epoch": 0.04207345316202956, "grad_norm": 1.3199396133422852, "learning_rate": 3.310326817038167e-05, "loss": 2.2556, "step": 4678 }, { "epoch": 0.042082447059670014, "grad_norm": 1.6429229974746704, "learning_rate": 3.3096157791645706e-05, "loss": 2.6981, "step": 4679 }, { "epoch": 0.04209144095731047, "grad_norm": 1.6013883352279663, "learning_rate": 3.308904668118772e-05, "loss": 2.6018, "step": 4680 }, { "epoch": 0.042100434854950915, "grad_norm": 1.9112818241119385, "learning_rate": 3.3081934839650403e-05, "loss": 2.5279, "step": 4681 }, { "epoch": 0.04210942875259137, "grad_norm": 2.030717611312866, "learning_rate": 3.307482226767653e-05, "loss": 2.6129, "step": 4682 }, { "epoch": 0.042118422650231815, "grad_norm": 1.7557600736618042, "learning_rate": 3.306770896590891e-05, "loss": 2.9215, "step": 4683 }, { "epoch": 0.04212741654787227, "grad_norm": 1.7730679512023926, "learning_rate": 3.3060594934990444e-05, "loss": 2.6647, "step": 4684 }, { "epoch": 0.04213641044551272, "grad_norm": 1.7879236936569214, "learning_rate": 3.30534801755641e-05, "loss": 2.6655, "step": 4685 }, { "epoch": 0.04214540434315317, "grad_norm": 1.4589954614639282, "learning_rate": 3.304636468827288e-05, "loss": 2.6035, "step": 4686 }, { "epoch": 0.042154398240793624, "grad_norm": 2.1267483234405518, "learning_rate": 3.3039248473759885e-05, "loss": 2.3906, "step": 4687 }, { "epoch": 0.04216339213843407, "grad_norm": 1.8902310132980347, "learning_rate": 3.303213153266828e-05, "loss": 2.664, "step": 4688 }, { "epoch": 0.042172386036074525, "grad_norm": 2.0066797733306885, "learning_rate": 3.302501386564128e-05, "loss": 2.7388, "step": 4689 }, { "epoch": 0.04218137993371497, "grad_norm": 1.6837626695632935, "learning_rate": 3.301789547332217e-05, "loss": 2.2127, "step": 4690 }, { "epoch": 0.042190373831355425, "grad_norm": 2.1508405208587646, "learning_rate": 3.30107763563543e-05, "loss": 2.5425, "step": 4691 }, { "epoch": 0.04219936772899588, "grad_norm": 1.310161828994751, "learning_rate": 3.300365651538109e-05, "loss": 2.5479, "step": 4692 }, { "epoch": 0.042208361626636326, "grad_norm": 1.9451030492782593, "learning_rate": 3.299653595104603e-05, "loss": 2.9108, "step": 4693 }, { "epoch": 0.04221735552427678, "grad_norm": 0.9771146178245544, "learning_rate": 3.2989414663992656e-05, "loss": 2.2515, "step": 4694 }, { "epoch": 0.04222634942191723, "grad_norm": 1.9318994283676147, "learning_rate": 3.298229265486459e-05, "loss": 2.4956, "step": 4695 }, { "epoch": 0.04223534331955768, "grad_norm": 1.4127867221832275, "learning_rate": 3.2975169924305524e-05, "loss": 2.4571, "step": 4696 }, { "epoch": 0.042244337217198134, "grad_norm": 2.520545482635498, "learning_rate": 3.296804647295919e-05, "loss": 2.6667, "step": 4697 }, { "epoch": 0.04225333111483858, "grad_norm": 2.2998175621032715, "learning_rate": 3.2960922301469385e-05, "loss": 2.6951, "step": 4698 }, { "epoch": 0.042262325012479035, "grad_norm": 1.916703224182129, "learning_rate": 3.295379741048e-05, "loss": 2.622, "step": 4699 }, { "epoch": 0.04227131891011948, "grad_norm": 1.6789857149124146, "learning_rate": 3.294667180063498e-05, "loss": 2.8209, "step": 4700 }, { "epoch": 0.042280312807759936, "grad_norm": 2.387068510055542, "learning_rate": 3.293954547257832e-05, "loss": 2.7355, "step": 4701 }, { "epoch": 0.04228930670540038, "grad_norm": 0.847571849822998, "learning_rate": 3.2932418426954085e-05, "loss": 2.2538, "step": 4702 }, { "epoch": 0.042298300603040836, "grad_norm": 1.199466347694397, "learning_rate": 3.2925290664406424e-05, "loss": 2.2523, "step": 4703 }, { "epoch": 0.04230729450068129, "grad_norm": 1.8785889148712158, "learning_rate": 3.291816218557953e-05, "loss": 2.7116, "step": 4704 }, { "epoch": 0.04231628839832174, "grad_norm": 1.7401007413864136, "learning_rate": 3.2911032991117654e-05, "loss": 2.5238, "step": 4705 }, { "epoch": 0.04232528229596219, "grad_norm": 1.211975336074829, "learning_rate": 3.290390308166515e-05, "loss": 2.3417, "step": 4706 }, { "epoch": 0.04233427619360264, "grad_norm": 1.8889391422271729, "learning_rate": 3.2896772457866394e-05, "loss": 2.5592, "step": 4707 }, { "epoch": 0.04234327009124309, "grad_norm": 1.976694107055664, "learning_rate": 3.288964112036586e-05, "loss": 2.3654, "step": 4708 }, { "epoch": 0.042352263988883546, "grad_norm": 2.0012381076812744, "learning_rate": 3.2882509069808044e-05, "loss": 2.6834, "step": 4709 }, { "epoch": 0.04236125788652399, "grad_norm": 1.4077636003494263, "learning_rate": 3.287537630683755e-05, "loss": 2.4315, "step": 4710 }, { "epoch": 0.042370251784164446, "grad_norm": 1.9626319408416748, "learning_rate": 3.286824283209903e-05, "loss": 2.5738, "step": 4711 }, { "epoch": 0.04237924568180489, "grad_norm": 2.2071990966796875, "learning_rate": 3.2861108646237205e-05, "loss": 2.7786, "step": 4712 }, { "epoch": 0.04238823957944535, "grad_norm": 1.4909621477127075, "learning_rate": 3.2853973749896836e-05, "loss": 2.4514, "step": 4713 }, { "epoch": 0.042397233477085794, "grad_norm": 1.4495676755905151, "learning_rate": 3.284683814372279e-05, "loss": 2.582, "step": 4714 }, { "epoch": 0.04240622737472625, "grad_norm": 1.6287481784820557, "learning_rate": 3.2839701828359956e-05, "loss": 2.6513, "step": 4715 }, { "epoch": 0.0424152212723667, "grad_norm": 1.1344316005706787, "learning_rate": 3.283256480445332e-05, "loss": 2.4028, "step": 4716 }, { "epoch": 0.04242421517000715, "grad_norm": 1.6105525493621826, "learning_rate": 3.282542707264792e-05, "loss": 2.6245, "step": 4717 }, { "epoch": 0.0424332090676476, "grad_norm": 1.444323182106018, "learning_rate": 3.281828863358884e-05, "loss": 2.4384, "step": 4718 }, { "epoch": 0.04244220296528805, "grad_norm": 1.8188976049423218, "learning_rate": 3.2811149487921256e-05, "loss": 2.7255, "step": 4719 }, { "epoch": 0.0424511968629285, "grad_norm": 1.241472840309143, "learning_rate": 3.28040096362904e-05, "loss": 2.3893, "step": 4720 }, { "epoch": 0.04246019076056896, "grad_norm": 1.8419268131256104, "learning_rate": 3.2796869079341555e-05, "loss": 2.5899, "step": 4721 }, { "epoch": 0.042469184658209404, "grad_norm": 1.281195878982544, "learning_rate": 3.278972781772008e-05, "loss": 2.4792, "step": 4722 }, { "epoch": 0.04247817855584986, "grad_norm": 1.9545069932937622, "learning_rate": 3.278258585207139e-05, "loss": 2.6511, "step": 4723 }, { "epoch": 0.042487172453490304, "grad_norm": 0.9819701910018921, "learning_rate": 3.277544318304099e-05, "loss": 2.4255, "step": 4724 }, { "epoch": 0.04249616635113076, "grad_norm": 1.7710702419281006, "learning_rate": 3.2768299811274394e-05, "loss": 2.8044, "step": 4725 }, { "epoch": 0.04250516024877121, "grad_norm": 1.517311930656433, "learning_rate": 3.276115573741724e-05, "loss": 2.6968, "step": 4726 }, { "epoch": 0.04251415414641166, "grad_norm": 1.551377773284912, "learning_rate": 3.275401096211518e-05, "loss": 2.5472, "step": 4727 }, { "epoch": 0.04252314804405211, "grad_norm": 2.2503421306610107, "learning_rate": 3.274686548601397e-05, "loss": 2.8231, "step": 4728 }, { "epoch": 0.04253214194169256, "grad_norm": 1.433037519454956, "learning_rate": 3.2739719309759386e-05, "loss": 2.527, "step": 4729 }, { "epoch": 0.042541135839333014, "grad_norm": 1.8937129974365234, "learning_rate": 3.273257243399733e-05, "loss": 2.5775, "step": 4730 }, { "epoch": 0.04255012973697346, "grad_norm": 1.5321577787399292, "learning_rate": 3.272542485937369e-05, "loss": 2.503, "step": 4731 }, { "epoch": 0.042559123634613914, "grad_norm": 2.0804898738861084, "learning_rate": 3.271827658653448e-05, "loss": 2.827, "step": 4732 }, { "epoch": 0.04256811753225437, "grad_norm": 1.4726321697235107, "learning_rate": 3.2711127616125735e-05, "loss": 2.7232, "step": 4733 }, { "epoch": 0.042577111429894815, "grad_norm": 1.1932293176651, "learning_rate": 3.270397794879359e-05, "loss": 2.3128, "step": 4734 }, { "epoch": 0.04258610532753527, "grad_norm": 1.8715955018997192, "learning_rate": 3.269682758518421e-05, "loss": 2.503, "step": 4735 }, { "epoch": 0.042595099225175716, "grad_norm": 0.7359278798103333, "learning_rate": 3.268967652594385e-05, "loss": 2.2271, "step": 4736 }, { "epoch": 0.04260409312281617, "grad_norm": 1.7299964427947998, "learning_rate": 3.268252477171881e-05, "loss": 2.7503, "step": 4737 }, { "epoch": 0.04261308702045662, "grad_norm": 1.0410046577453613, "learning_rate": 3.267537232315546e-05, "loss": 2.243, "step": 4738 }, { "epoch": 0.04262208091809707, "grad_norm": 1.8028756380081177, "learning_rate": 3.2668219180900215e-05, "loss": 2.6745, "step": 4739 }, { "epoch": 0.042631074815737524, "grad_norm": 3.4239003658294678, "learning_rate": 3.266106534559958e-05, "loss": 2.4212, "step": 4740 }, { "epoch": 0.04264006871337797, "grad_norm": 1.6264952421188354, "learning_rate": 3.265391081790012e-05, "loss": 2.857, "step": 4741 }, { "epoch": 0.042649062611018425, "grad_norm": 0.9997032284736633, "learning_rate": 3.264675559844844e-05, "loss": 2.2855, "step": 4742 }, { "epoch": 0.04265805650865887, "grad_norm": 1.844890832901001, "learning_rate": 3.263959968789122e-05, "loss": 2.4691, "step": 4743 }, { "epoch": 0.042667050406299326, "grad_norm": 2.1101036071777344, "learning_rate": 3.2632443086875216e-05, "loss": 2.5873, "step": 4744 }, { "epoch": 0.04267604430393978, "grad_norm": 1.5795233249664307, "learning_rate": 3.262528579604723e-05, "loss": 2.5396, "step": 4745 }, { "epoch": 0.042685038201580226, "grad_norm": 1.624883770942688, "learning_rate": 3.261812781605412e-05, "loss": 2.5855, "step": 4746 }, { "epoch": 0.04269403209922068, "grad_norm": 1.6244808435440063, "learning_rate": 3.2610969147542824e-05, "loss": 2.7624, "step": 4747 }, { "epoch": 0.04270302599686113, "grad_norm": 1.637644648551941, "learning_rate": 3.2603809791160336e-05, "loss": 2.4442, "step": 4748 }, { "epoch": 0.04271201989450158, "grad_norm": 1.2056527137756348, "learning_rate": 3.2596649747553704e-05, "loss": 2.4709, "step": 4749 }, { "epoch": 0.042721013792142035, "grad_norm": 1.1548292636871338, "learning_rate": 3.258948901737006e-05, "loss": 2.4743, "step": 4750 }, { "epoch": 0.04273000768978248, "grad_norm": 1.9205800294876099, "learning_rate": 3.258232760125657e-05, "loss": 2.7958, "step": 4751 }, { "epoch": 0.042739001587422935, "grad_norm": 1.3681516647338867, "learning_rate": 3.257516549986048e-05, "loss": 2.5589, "step": 4752 }, { "epoch": 0.04274799548506338, "grad_norm": 2.5354042053222656, "learning_rate": 3.2568002713829084e-05, "loss": 2.4137, "step": 4753 }, { "epoch": 0.042756989382703836, "grad_norm": 1.7041724920272827, "learning_rate": 3.2560839243809755e-05, "loss": 2.4469, "step": 4754 }, { "epoch": 0.04276598328034428, "grad_norm": 2.2354888916015625, "learning_rate": 3.2553675090449917e-05, "loss": 2.8855, "step": 4755 }, { "epoch": 0.04277497717798474, "grad_norm": 1.4591480493545532, "learning_rate": 3.2546510254397074e-05, "loss": 2.5972, "step": 4756 }, { "epoch": 0.04278397107562519, "grad_norm": 1.905314564704895, "learning_rate": 3.2539344736298754e-05, "loss": 2.644, "step": 4757 }, { "epoch": 0.04279296497326564, "grad_norm": 1.3529906272888184, "learning_rate": 3.2532178536802576e-05, "loss": 2.6714, "step": 4758 }, { "epoch": 0.04280195887090609, "grad_norm": 1.5484410524368286, "learning_rate": 3.252501165655622e-05, "loss": 2.5592, "step": 4759 }, { "epoch": 0.04281095276854654, "grad_norm": 1.2399762868881226, "learning_rate": 3.2517844096207414e-05, "loss": 2.2032, "step": 4760 }, { "epoch": 0.04281994666618699, "grad_norm": 1.4927465915679932, "learning_rate": 3.251067585640395e-05, "loss": 2.5744, "step": 4761 }, { "epoch": 0.042828940563827446, "grad_norm": 1.6844521760940552, "learning_rate": 3.25035069377937e-05, "loss": 2.627, "step": 4762 }, { "epoch": 0.04283793446146789, "grad_norm": 1.8556263446807861, "learning_rate": 3.249633734102457e-05, "loss": 2.449, "step": 4763 }, { "epoch": 0.04284692835910835, "grad_norm": 1.9691869020462036, "learning_rate": 3.2489167066744547e-05, "loss": 2.5966, "step": 4764 }, { "epoch": 0.042855922256748794, "grad_norm": 1.5686497688293457, "learning_rate": 3.2481996115601675e-05, "loss": 2.3756, "step": 4765 }, { "epoch": 0.04286491615438925, "grad_norm": 1.1151002645492554, "learning_rate": 3.247482448824405e-05, "loss": 2.3256, "step": 4766 }, { "epoch": 0.0428739100520297, "grad_norm": 2.0757579803466797, "learning_rate": 3.2467652185319834e-05, "loss": 2.8262, "step": 4767 }, { "epoch": 0.04288290394967015, "grad_norm": 1.9868398904800415, "learning_rate": 3.2460479207477266e-05, "loss": 2.7895, "step": 4768 }, { "epoch": 0.0428918978473106, "grad_norm": 2.0347063541412354, "learning_rate": 3.2453305555364616e-05, "loss": 2.6905, "step": 4769 }, { "epoch": 0.04290089174495105, "grad_norm": 1.4504588842391968, "learning_rate": 3.244613122963024e-05, "loss": 2.8229, "step": 4770 }, { "epoch": 0.0429098856425915, "grad_norm": 1.375630497932434, "learning_rate": 3.243895623092254e-05, "loss": 2.5179, "step": 4771 }, { "epoch": 0.04291887954023195, "grad_norm": 1.163741111755371, "learning_rate": 3.243178055988999e-05, "loss": 2.4339, "step": 4772 }, { "epoch": 0.0429278734378724, "grad_norm": 1.758355736732483, "learning_rate": 3.2424604217181123e-05, "loss": 2.6423, "step": 4773 }, { "epoch": 0.04293686733551286, "grad_norm": 1.8166369199752808, "learning_rate": 3.241742720344453e-05, "loss": 2.5095, "step": 4774 }, { "epoch": 0.042945861233153304, "grad_norm": 1.3265454769134521, "learning_rate": 3.241024951932885e-05, "loss": 2.3047, "step": 4775 }, { "epoch": 0.04295485513079376, "grad_norm": 2.067720413208008, "learning_rate": 3.240307116548279e-05, "loss": 2.7693, "step": 4776 }, { "epoch": 0.042963849028434205, "grad_norm": 1.985103726387024, "learning_rate": 3.239589214255514e-05, "loss": 3.0657, "step": 4777 }, { "epoch": 0.04297284292607466, "grad_norm": 2.2700774669647217, "learning_rate": 3.238871245119472e-05, "loss": 2.8032, "step": 4778 }, { "epoch": 0.04298183682371511, "grad_norm": 1.6183042526245117, "learning_rate": 3.2381532092050434e-05, "loss": 2.3773, "step": 4779 }, { "epoch": 0.04299083072135556, "grad_norm": 1.5326024293899536, "learning_rate": 3.237435106577123e-05, "loss": 2.642, "step": 4780 }, { "epoch": 0.04299982461899601, "grad_norm": 1.8681467771530151, "learning_rate": 3.236716937300611e-05, "loss": 2.853, "step": 4781 }, { "epoch": 0.04300881851663646, "grad_norm": 1.55057692527771, "learning_rate": 3.235998701440416e-05, "loss": 2.6307, "step": 4782 }, { "epoch": 0.043017812414276914, "grad_norm": 1.833404302597046, "learning_rate": 3.235280399061451e-05, "loss": 2.4815, "step": 4783 }, { "epoch": 0.04302680631191736, "grad_norm": 0.9509752988815308, "learning_rate": 3.234562030228636e-05, "loss": 2.244, "step": 4784 }, { "epoch": 0.043035800209557815, "grad_norm": 1.8320599794387817, "learning_rate": 3.233843595006896e-05, "loss": 2.8493, "step": 4785 }, { "epoch": 0.04304479410719827, "grad_norm": 2.1329898834228516, "learning_rate": 3.2331250934611624e-05, "loss": 2.5785, "step": 4786 }, { "epoch": 0.043053788004838715, "grad_norm": 1.1131519079208374, "learning_rate": 3.232406525656372e-05, "loss": 2.3292, "step": 4787 }, { "epoch": 0.04306278190247917, "grad_norm": 1.6067297458648682, "learning_rate": 3.2316878916574686e-05, "loss": 2.6608, "step": 4788 }, { "epoch": 0.043071775800119616, "grad_norm": 1.4645466804504395, "learning_rate": 3.230969191529402e-05, "loss": 2.3569, "step": 4789 }, { "epoch": 0.04308076969776007, "grad_norm": 1.7485642433166504, "learning_rate": 3.230250425337127e-05, "loss": 2.5871, "step": 4790 }, { "epoch": 0.043089763595400524, "grad_norm": 1.8147869110107422, "learning_rate": 3.2295315931456054e-05, "loss": 2.5279, "step": 4791 }, { "epoch": 0.04309875749304097, "grad_norm": 0.8546193242073059, "learning_rate": 3.228812695019804e-05, "loss": 2.2256, "step": 4792 }, { "epoch": 0.043107751390681424, "grad_norm": 1.471081256866455, "learning_rate": 3.2280937310246965e-05, "loss": 2.6835, "step": 4793 }, { "epoch": 0.04311674528832187, "grad_norm": 1.5582671165466309, "learning_rate": 3.227374701225262e-05, "loss": 2.7125, "step": 4794 }, { "epoch": 0.043125739185962325, "grad_norm": 1.769566535949707, "learning_rate": 3.226655605686485e-05, "loss": 2.6045, "step": 4795 }, { "epoch": 0.04313473308360278, "grad_norm": 1.883786678314209, "learning_rate": 3.2259364444733564e-05, "loss": 2.6662, "step": 4796 }, { "epoch": 0.043143726981243226, "grad_norm": 1.635026454925537, "learning_rate": 3.225217217650876e-05, "loss": 2.5102, "step": 4797 }, { "epoch": 0.04315272087888368, "grad_norm": 1.7826025485992432, "learning_rate": 3.224497925284042e-05, "loss": 2.7494, "step": 4798 }, { "epoch": 0.04316171477652413, "grad_norm": 1.4440743923187256, "learning_rate": 3.2237785674378675e-05, "loss": 2.6553, "step": 4799 }, { "epoch": 0.04317070867416458, "grad_norm": 1.399863600730896, "learning_rate": 3.223059144177365e-05, "loss": 2.4829, "step": 4800 }, { "epoch": 0.04317970257180503, "grad_norm": 1.5225317478179932, "learning_rate": 3.222339655567556e-05, "loss": 2.451, "step": 4801 }, { "epoch": 0.04318869646944548, "grad_norm": 1.4932067394256592, "learning_rate": 3.221620101673467e-05, "loss": 2.5977, "step": 4802 }, { "epoch": 0.043197690367085935, "grad_norm": 1.3739020824432373, "learning_rate": 3.2209004825601306e-05, "loss": 2.3873, "step": 4803 }, { "epoch": 0.04320668426472638, "grad_norm": 2.041126012802124, "learning_rate": 3.220180798292585e-05, "loss": 2.8397, "step": 4804 }, { "epoch": 0.043215678162366836, "grad_norm": 1.7425001859664917, "learning_rate": 3.219461048935874e-05, "loss": 2.6198, "step": 4805 }, { "epoch": 0.04322467206000728, "grad_norm": 1.8059793710708618, "learning_rate": 3.2187412345550495e-05, "loss": 2.5656, "step": 4806 }, { "epoch": 0.043233665957647736, "grad_norm": 1.2928029298782349, "learning_rate": 3.2180213552151664e-05, "loss": 2.3714, "step": 4807 }, { "epoch": 0.04324265985528819, "grad_norm": 2.1226985454559326, "learning_rate": 3.217301410981285e-05, "loss": 2.7373, "step": 4808 }, { "epoch": 0.04325165375292864, "grad_norm": 2.3012092113494873, "learning_rate": 3.216581401918476e-05, "loss": 2.5767, "step": 4809 }, { "epoch": 0.04326064765056909, "grad_norm": 1.9242141246795654, "learning_rate": 3.215861328091812e-05, "loss": 2.764, "step": 4810 }, { "epoch": 0.04326964154820954, "grad_norm": 1.371201753616333, "learning_rate": 3.215141189566371e-05, "loss": 2.3177, "step": 4811 }, { "epoch": 0.04327863544584999, "grad_norm": 2.829190731048584, "learning_rate": 3.2144209864072405e-05, "loss": 2.7522, "step": 4812 }, { "epoch": 0.04328762934349044, "grad_norm": 1.586220383644104, "learning_rate": 3.2137007186795105e-05, "loss": 2.8399, "step": 4813 }, { "epoch": 0.04329662324113089, "grad_norm": 2.1337711811065674, "learning_rate": 3.212980386448279e-05, "loss": 2.7788, "step": 4814 }, { "epoch": 0.043305617138771346, "grad_norm": 1.832259178161621, "learning_rate": 3.212259989778647e-05, "loss": 2.8099, "step": 4815 }, { "epoch": 0.04331461103641179, "grad_norm": 2.1353793144226074, "learning_rate": 3.211539528735725e-05, "loss": 2.9799, "step": 4816 }, { "epoch": 0.04332360493405225, "grad_norm": 1.7874630689620972, "learning_rate": 3.210819003384626e-05, "loss": 2.6738, "step": 4817 }, { "epoch": 0.043332598831692694, "grad_norm": 1.4372024536132812, "learning_rate": 3.2100984137904715e-05, "loss": 2.6028, "step": 4818 }, { "epoch": 0.04334159272933315, "grad_norm": 1.8646783828735352, "learning_rate": 3.2093777600183875e-05, "loss": 2.3945, "step": 4819 }, { "epoch": 0.0433505866269736, "grad_norm": 1.7390061616897583, "learning_rate": 3.2086570421335055e-05, "loss": 2.2633, "step": 4820 }, { "epoch": 0.04335958052461405, "grad_norm": 1.2693020105361938, "learning_rate": 3.207936260200963e-05, "loss": 2.5478, "step": 4821 }, { "epoch": 0.0433685744222545, "grad_norm": 1.791790246963501, "learning_rate": 3.207215414285905e-05, "loss": 2.6881, "step": 4822 }, { "epoch": 0.04337756831989495, "grad_norm": 1.591543197631836, "learning_rate": 3.2064945044534785e-05, "loss": 2.4626, "step": 4823 }, { "epoch": 0.0433865622175354, "grad_norm": 0.8383256196975708, "learning_rate": 3.2057735307688396e-05, "loss": 2.2143, "step": 4824 }, { "epoch": 0.04339555611517585, "grad_norm": 1.9999315738677979, "learning_rate": 3.2050524932971494e-05, "loss": 2.4005, "step": 4825 }, { "epoch": 0.043404550012816304, "grad_norm": 1.468301773071289, "learning_rate": 3.2043313921035743e-05, "loss": 2.5357, "step": 4826 }, { "epoch": 0.04341354391045676, "grad_norm": 1.8559321165084839, "learning_rate": 3.2036102272532866e-05, "loss": 2.8342, "step": 4827 }, { "epoch": 0.043422537808097204, "grad_norm": 1.519051194190979, "learning_rate": 3.202888998811465e-05, "loss": 2.4929, "step": 4828 }, { "epoch": 0.04343153170573766, "grad_norm": 1.36306893825531, "learning_rate": 3.202167706843292e-05, "loss": 2.3363, "step": 4829 }, { "epoch": 0.043440525603378105, "grad_norm": 1.4653385877609253, "learning_rate": 3.201446351413958e-05, "loss": 2.4584, "step": 4830 }, { "epoch": 0.04344951950101856, "grad_norm": 1.0268492698669434, "learning_rate": 3.200724932588659e-05, "loss": 2.2659, "step": 4831 }, { "epoch": 0.04345851339865901, "grad_norm": 1.7247837781906128, "learning_rate": 3.200003450432595e-05, "loss": 2.5481, "step": 4832 }, { "epoch": 0.04346750729629946, "grad_norm": 1.1157300472259521, "learning_rate": 3.199281905010973e-05, "loss": 2.3945, "step": 4833 }, { "epoch": 0.043476501193939913, "grad_norm": 1.6224695444107056, "learning_rate": 3.198560296389005e-05, "loss": 2.6797, "step": 4834 }, { "epoch": 0.04348549509158036, "grad_norm": 1.2878233194351196, "learning_rate": 3.197838624631911e-05, "loss": 2.4884, "step": 4835 }, { "epoch": 0.043494488989220814, "grad_norm": 1.4827512502670288, "learning_rate": 3.1971168898049134e-05, "loss": 2.4421, "step": 4836 }, { "epoch": 0.04350348288686127, "grad_norm": 1.3665564060211182, "learning_rate": 3.196395091973242e-05, "loss": 2.5913, "step": 4837 }, { "epoch": 0.043512476784501715, "grad_norm": 1.407931923866272, "learning_rate": 3.195673231202132e-05, "loss": 2.4668, "step": 4838 }, { "epoch": 0.04352147068214217, "grad_norm": 1.6518893241882324, "learning_rate": 3.194951307556825e-05, "loss": 2.5407, "step": 4839 }, { "epoch": 0.043530464579782616, "grad_norm": 1.2894777059555054, "learning_rate": 3.194229321102568e-05, "loss": 2.5186, "step": 4840 }, { "epoch": 0.04353945847742307, "grad_norm": 1.7532262802124023, "learning_rate": 3.1935072719046115e-05, "loss": 2.6551, "step": 4841 }, { "epoch": 0.043548452375063516, "grad_norm": 1.4723098278045654, "learning_rate": 3.1927851600282155e-05, "loss": 2.8411, "step": 4842 }, { "epoch": 0.04355744627270397, "grad_norm": 1.4466543197631836, "learning_rate": 3.1920629855386434e-05, "loss": 2.6678, "step": 4843 }, { "epoch": 0.043566440170344424, "grad_norm": 1.5341876745224, "learning_rate": 3.1913407485011626e-05, "loss": 2.6694, "step": 4844 }, { "epoch": 0.04357543406798487, "grad_norm": 1.7431820631027222, "learning_rate": 3.190618448981051e-05, "loss": 2.8419, "step": 4845 }, { "epoch": 0.043584427965625325, "grad_norm": 1.5638906955718994, "learning_rate": 3.1898960870435876e-05, "loss": 2.6839, "step": 4846 }, { "epoch": 0.04359342186326577, "grad_norm": 1.9973000288009644, "learning_rate": 3.189173662754058e-05, "loss": 2.4296, "step": 4847 }, { "epoch": 0.043602415760906225, "grad_norm": 2.183122396469116, "learning_rate": 3.1884511761777556e-05, "loss": 2.8508, "step": 4848 }, { "epoch": 0.04361140965854668, "grad_norm": 1.253006100654602, "learning_rate": 3.187728627379978e-05, "loss": 2.3467, "step": 4849 }, { "epoch": 0.043620403556187126, "grad_norm": 0.8560361862182617, "learning_rate": 3.187006016426028e-05, "loss": 2.1467, "step": 4850 }, { "epoch": 0.04362939745382758, "grad_norm": 1.7876077890396118, "learning_rate": 3.186283343381213e-05, "loss": 2.7284, "step": 4851 }, { "epoch": 0.04363839135146803, "grad_norm": 1.8128454685211182, "learning_rate": 3.185560608310849e-05, "loss": 2.8153, "step": 4852 }, { "epoch": 0.04364738524910848, "grad_norm": 1.9536702632904053, "learning_rate": 3.184837811280256e-05, "loss": 2.5837, "step": 4853 }, { "epoch": 0.04365637914674893, "grad_norm": 1.5009000301361084, "learning_rate": 3.184114952354759e-05, "loss": 2.5575, "step": 4854 }, { "epoch": 0.04366537304438938, "grad_norm": 1.5603307485580444, "learning_rate": 3.18339203159969e-05, "loss": 2.5166, "step": 4855 }, { "epoch": 0.043674366942029835, "grad_norm": 1.4802589416503906, "learning_rate": 3.182669049080385e-05, "loss": 2.5004, "step": 4856 }, { "epoch": 0.04368336083967028, "grad_norm": 1.6827392578125, "learning_rate": 3.1819460048621866e-05, "loss": 2.6966, "step": 4857 }, { "epoch": 0.043692354737310736, "grad_norm": 1.465554118156433, "learning_rate": 3.1812228990104425e-05, "loss": 2.5753, "step": 4858 }, { "epoch": 0.04370134863495118, "grad_norm": 1.1929665803909302, "learning_rate": 3.1804997315905074e-05, "loss": 2.4548, "step": 4859 }, { "epoch": 0.04371034253259164, "grad_norm": 1.111159086227417, "learning_rate": 3.1797765026677384e-05, "loss": 2.3787, "step": 4860 }, { "epoch": 0.04371933643023209, "grad_norm": 2.035193681716919, "learning_rate": 3.179053212307502e-05, "loss": 2.7413, "step": 4861 }, { "epoch": 0.04372833032787254, "grad_norm": 1.0538839101791382, "learning_rate": 3.1783298605751674e-05, "loss": 2.458, "step": 4862 }, { "epoch": 0.04373732422551299, "grad_norm": 1.6932134628295898, "learning_rate": 3.1776064475361114e-05, "loss": 2.5425, "step": 4863 }, { "epoch": 0.04374631812315344, "grad_norm": 2.171297311782837, "learning_rate": 3.1768829732557135e-05, "loss": 2.4131, "step": 4864 }, { "epoch": 0.04375531202079389, "grad_norm": 1.6380075216293335, "learning_rate": 3.176159437799362e-05, "loss": 2.5193, "step": 4865 }, { "epoch": 0.04376430591843434, "grad_norm": 1.3870933055877686, "learning_rate": 3.175435841232448e-05, "loss": 2.4888, "step": 4866 }, { "epoch": 0.04377329981607479, "grad_norm": 1.820525884628296, "learning_rate": 3.174712183620371e-05, "loss": 2.6247, "step": 4867 }, { "epoch": 0.043782293713715247, "grad_norm": 1.718639612197876, "learning_rate": 3.173988465028534e-05, "loss": 2.4682, "step": 4868 }, { "epoch": 0.04379128761135569, "grad_norm": 1.581529974937439, "learning_rate": 3.1732646855223445e-05, "loss": 2.2831, "step": 4869 }, { "epoch": 0.04380028150899615, "grad_norm": 1.7485156059265137, "learning_rate": 3.172540845167219e-05, "loss": 2.7521, "step": 4870 }, { "epoch": 0.043809275406636594, "grad_norm": 1.8699551820755005, "learning_rate": 3.171816944028576e-05, "loss": 2.4633, "step": 4871 }, { "epoch": 0.04381826930427705, "grad_norm": 1.8832323551177979, "learning_rate": 3.1710929821718406e-05, "loss": 2.6558, "step": 4872 }, { "epoch": 0.0438272632019175, "grad_norm": 2.081618070602417, "learning_rate": 3.1703689596624444e-05, "loss": 2.6785, "step": 4873 }, { "epoch": 0.04383625709955795, "grad_norm": 2.257828950881958, "learning_rate": 3.169644876565824e-05, "loss": 2.4952, "step": 4874 }, { "epoch": 0.0438452509971984, "grad_norm": 1.8698188066482544, "learning_rate": 3.1689207329474214e-05, "loss": 2.9311, "step": 4875 }, { "epoch": 0.04385424489483885, "grad_norm": 1.618011474609375, "learning_rate": 3.168196528872682e-05, "loss": 2.586, "step": 4876 }, { "epoch": 0.0438632387924793, "grad_norm": 1.476027250289917, "learning_rate": 3.167472264407062e-05, "loss": 2.6718, "step": 4877 }, { "epoch": 0.04387223269011976, "grad_norm": 1.798018455505371, "learning_rate": 3.1667479396160166e-05, "loss": 2.6903, "step": 4878 }, { "epoch": 0.043881226587760204, "grad_norm": 1.8508659601211548, "learning_rate": 3.16602355456501e-05, "loss": 2.9198, "step": 4879 }, { "epoch": 0.04389022048540066, "grad_norm": 2.1838157176971436, "learning_rate": 3.165299109319513e-05, "loss": 3.0519, "step": 4880 }, { "epoch": 0.043899214383041105, "grad_norm": 2.091712474822998, "learning_rate": 3.1645746039449984e-05, "loss": 2.892, "step": 4881 }, { "epoch": 0.04390820828068156, "grad_norm": 1.7486668825149536, "learning_rate": 3.163850038506948e-05, "loss": 2.8, "step": 4882 }, { "epoch": 0.043917202178322005, "grad_norm": 1.5357357263565063, "learning_rate": 3.163125413070844e-05, "loss": 2.6482, "step": 4883 }, { "epoch": 0.04392619607596246, "grad_norm": 1.6757526397705078, "learning_rate": 3.1624007277021814e-05, "loss": 2.9482, "step": 4884 }, { "epoch": 0.04393518997360291, "grad_norm": 1.966107964515686, "learning_rate": 3.161675982466454e-05, "loss": 3.0476, "step": 4885 }, { "epoch": 0.04394418387124336, "grad_norm": 1.823981761932373, "learning_rate": 3.160951177429165e-05, "loss": 2.7886, "step": 4886 }, { "epoch": 0.043953177768883814, "grad_norm": 2.0122733116149902, "learning_rate": 3.160226312655819e-05, "loss": 2.8613, "step": 4887 }, { "epoch": 0.04396217166652426, "grad_norm": 1.7700142860412598, "learning_rate": 3.1595013882119295e-05, "loss": 2.6428, "step": 4888 }, { "epoch": 0.043971165564164715, "grad_norm": 1.823222041130066, "learning_rate": 3.1587764041630156e-05, "loss": 2.6494, "step": 4889 }, { "epoch": 0.04398015946180517, "grad_norm": 1.9896769523620605, "learning_rate": 3.1580513605746006e-05, "loss": 2.3552, "step": 4890 }, { "epoch": 0.043989153359445615, "grad_norm": 1.7289910316467285, "learning_rate": 3.157326257512212e-05, "loss": 2.6182, "step": 4891 }, { "epoch": 0.04399814725708607, "grad_norm": 1.6736879348754883, "learning_rate": 3.156601095041384e-05, "loss": 2.7022, "step": 4892 }, { "epoch": 0.044007141154726516, "grad_norm": 2.227177143096924, "learning_rate": 3.155875873227657e-05, "loss": 2.7856, "step": 4893 }, { "epoch": 0.04401613505236697, "grad_norm": 1.2402243614196777, "learning_rate": 3.155150592136574e-05, "loss": 2.3386, "step": 4894 }, { "epoch": 0.04402512895000742, "grad_norm": 1.5600290298461914, "learning_rate": 3.154425251833686e-05, "loss": 2.5507, "step": 4895 }, { "epoch": 0.04403412284764787, "grad_norm": 1.7119842767715454, "learning_rate": 3.1536998523845494e-05, "loss": 2.5763, "step": 4896 }, { "epoch": 0.044043116745288324, "grad_norm": 1.676302194595337, "learning_rate": 3.152974393854725e-05, "loss": 2.7746, "step": 4897 }, { "epoch": 0.04405211064292877, "grad_norm": 1.600170612335205, "learning_rate": 3.152248876309777e-05, "loss": 2.5513, "step": 4898 }, { "epoch": 0.044061104540569225, "grad_norm": 1.942423939704895, "learning_rate": 3.151523299815278e-05, "loss": 2.3487, "step": 4899 }, { "epoch": 0.04407009843820967, "grad_norm": 1.9994333982467651, "learning_rate": 3.1507976644368055e-05, "loss": 2.8068, "step": 4900 }, { "epoch": 0.044079092335850126, "grad_norm": 2.1874430179595947, "learning_rate": 3.150071970239941e-05, "loss": 2.572, "step": 4901 }, { "epoch": 0.04408808623349058, "grad_norm": 1.8684486150741577, "learning_rate": 3.149346217290271e-05, "loss": 2.5316, "step": 4902 }, { "epoch": 0.044097080131131026, "grad_norm": 2.0040788650512695, "learning_rate": 3.14862040565339e-05, "loss": 2.5933, "step": 4903 }, { "epoch": 0.04410607402877148, "grad_norm": 1.1934393644332886, "learning_rate": 3.1478945353948964e-05, "loss": 2.5192, "step": 4904 }, { "epoch": 0.04411506792641193, "grad_norm": 1.9914473295211792, "learning_rate": 3.147168606580392e-05, "loss": 2.8696, "step": 4905 }, { "epoch": 0.04412406182405238, "grad_norm": 0.8495921492576599, "learning_rate": 3.146442619275486e-05, "loss": 2.4164, "step": 4906 }, { "epoch": 0.044133055721692835, "grad_norm": 1.964126706123352, "learning_rate": 3.145716573545792e-05, "loss": 2.7937, "step": 4907 }, { "epoch": 0.04414204961933328, "grad_norm": 1.5502870082855225, "learning_rate": 3.144990469456929e-05, "loss": 2.6583, "step": 4908 }, { "epoch": 0.044151043516973736, "grad_norm": 2.274113655090332, "learning_rate": 3.144264307074523e-05, "loss": 2.9175, "step": 4909 }, { "epoch": 0.04416003741461418, "grad_norm": 2.1979973316192627, "learning_rate": 3.143538086464203e-05, "loss": 2.8784, "step": 4910 }, { "epoch": 0.044169031312254636, "grad_norm": 1.661430835723877, "learning_rate": 3.142811807691603e-05, "loss": 2.6107, "step": 4911 }, { "epoch": 0.04417802520989508, "grad_norm": 1.777230978012085, "learning_rate": 3.142085470822365e-05, "loss": 2.6591, "step": 4912 }, { "epoch": 0.04418701910753554, "grad_norm": 1.7770907878875732, "learning_rate": 3.141359075922133e-05, "loss": 2.5501, "step": 4913 }, { "epoch": 0.04419601300517599, "grad_norm": 1.1145879030227661, "learning_rate": 3.1406326230565596e-05, "loss": 2.2786, "step": 4914 }, { "epoch": 0.04420500690281644, "grad_norm": 1.2006455659866333, "learning_rate": 3.139906112291299e-05, "loss": 2.4666, "step": 4915 }, { "epoch": 0.04421400080045689, "grad_norm": 1.917510747909546, "learning_rate": 3.139179543692013e-05, "loss": 2.9385, "step": 4916 }, { "epoch": 0.04422299469809734, "grad_norm": 1.833746075630188, "learning_rate": 3.138452917324369e-05, "loss": 2.4921, "step": 4917 }, { "epoch": 0.04423198859573779, "grad_norm": 1.3472036123275757, "learning_rate": 3.137726233254038e-05, "loss": 2.5719, "step": 4918 }, { "epoch": 0.044240982493378246, "grad_norm": 1.3911651372909546, "learning_rate": 3.1369994915466965e-05, "loss": 2.4117, "step": 4919 }, { "epoch": 0.04424997639101869, "grad_norm": 1.5583547353744507, "learning_rate": 3.1362726922680276e-05, "loss": 2.6047, "step": 4920 }, { "epoch": 0.04425897028865915, "grad_norm": 1.6851825714111328, "learning_rate": 3.135545835483718e-05, "loss": 2.6919, "step": 4921 }, { "epoch": 0.044267964186299594, "grad_norm": 1.3815250396728516, "learning_rate": 3.13481892125946e-05, "loss": 2.4199, "step": 4922 }, { "epoch": 0.04427695808394005, "grad_norm": 2.088809013366699, "learning_rate": 3.134091949660951e-05, "loss": 2.4972, "step": 4923 }, { "epoch": 0.044285951981580494, "grad_norm": 1.8053690195083618, "learning_rate": 3.133364920753896e-05, "loss": 2.5207, "step": 4924 }, { "epoch": 0.04429494587922095, "grad_norm": 1.7516417503356934, "learning_rate": 3.132637834604001e-05, "loss": 2.6177, "step": 4925 }, { "epoch": 0.0443039397768614, "grad_norm": 1.690057396888733, "learning_rate": 3.13191069127698e-05, "loss": 2.6195, "step": 4926 }, { "epoch": 0.04431293367450185, "grad_norm": 1.394157886505127, "learning_rate": 3.1311834908385516e-05, "loss": 2.4477, "step": 4927 }, { "epoch": 0.0443219275721423, "grad_norm": 1.5291836261749268, "learning_rate": 3.1304562333544384e-05, "loss": 2.5993, "step": 4928 }, { "epoch": 0.04433092146978275, "grad_norm": 1.5407929420471191, "learning_rate": 3.129728918890371e-05, "loss": 2.4676, "step": 4929 }, { "epoch": 0.044339915367423204, "grad_norm": 1.9808015823364258, "learning_rate": 3.129001547512081e-05, "loss": 2.4995, "step": 4930 }, { "epoch": 0.04434890926506366, "grad_norm": 1.2944581508636475, "learning_rate": 3.128274119285309e-05, "loss": 2.529, "step": 4931 }, { "epoch": 0.044357903162704104, "grad_norm": 1.7623978853225708, "learning_rate": 3.127546634275799e-05, "loss": 2.4678, "step": 4932 }, { "epoch": 0.04436689706034456, "grad_norm": 1.463850498199463, "learning_rate": 3.1268190925493005e-05, "loss": 2.5086, "step": 4933 }, { "epoch": 0.044375890957985005, "grad_norm": 1.7895580530166626, "learning_rate": 3.1260914941715675e-05, "loss": 2.4027, "step": 4934 }, { "epoch": 0.04438488485562546, "grad_norm": 1.2661067247390747, "learning_rate": 3.125363839208359e-05, "loss": 2.4954, "step": 4935 }, { "epoch": 0.044393878753265906, "grad_norm": 1.5510870218276978, "learning_rate": 3.12463612772544e-05, "loss": 2.3692, "step": 4936 }, { "epoch": 0.04440287265090636, "grad_norm": 1.5418957471847534, "learning_rate": 3.123908359788581e-05, "loss": 2.3933, "step": 4937 }, { "epoch": 0.04441186654854681, "grad_norm": 1.7235668897628784, "learning_rate": 3.1231805354635566e-05, "loss": 2.7539, "step": 4938 }, { "epoch": 0.04442086044618726, "grad_norm": 1.8912891149520874, "learning_rate": 3.1224526548161475e-05, "loss": 2.6492, "step": 4939 }, { "epoch": 0.044429854343827714, "grad_norm": 1.8366245031356812, "learning_rate": 3.121724717912138e-05, "loss": 2.5169, "step": 4940 }, { "epoch": 0.04443884824146816, "grad_norm": 1.5158265829086304, "learning_rate": 3.120996724817316e-05, "loss": 2.7184, "step": 4941 }, { "epoch": 0.044447842139108615, "grad_norm": 1.7331664562225342, "learning_rate": 3.1202686755974805e-05, "loss": 2.5453, "step": 4942 }, { "epoch": 0.04445683603674907, "grad_norm": 1.9798636436462402, "learning_rate": 3.11954057031843e-05, "loss": 2.8501, "step": 4943 }, { "epoch": 0.044465829934389516, "grad_norm": 1.8991148471832275, "learning_rate": 3.1188124090459696e-05, "loss": 2.7807, "step": 4944 }, { "epoch": 0.04447482383202997, "grad_norm": 1.309145450592041, "learning_rate": 3.118084191845911e-05, "loss": 2.5511, "step": 4945 }, { "epoch": 0.044483817729670416, "grad_norm": 1.4936025142669678, "learning_rate": 3.1173559187840686e-05, "loss": 2.4483, "step": 4946 }, { "epoch": 0.04449281162731087, "grad_norm": 1.7720167636871338, "learning_rate": 3.116627589926263e-05, "loss": 2.9374, "step": 4947 }, { "epoch": 0.044501805524951324, "grad_norm": 1.6636334657669067, "learning_rate": 3.1158992053383204e-05, "loss": 2.6438, "step": 4948 }, { "epoch": 0.04451079942259177, "grad_norm": 1.4325810670852661, "learning_rate": 3.11517076508607e-05, "loss": 2.3605, "step": 4949 }, { "epoch": 0.044519793320232225, "grad_norm": 1.920971155166626, "learning_rate": 3.114442269235349e-05, "loss": 2.8788, "step": 4950 }, { "epoch": 0.04452878721787267, "grad_norm": 1.2860749959945679, "learning_rate": 3.1137137178519985e-05, "loss": 2.6103, "step": 4951 }, { "epoch": 0.044537781115513125, "grad_norm": 1.5505845546722412, "learning_rate": 3.1129851110018623e-05, "loss": 2.7366, "step": 4952 }, { "epoch": 0.04454677501315357, "grad_norm": 2.139430522918701, "learning_rate": 3.112256448750792e-05, "loss": 2.6376, "step": 4953 }, { "epoch": 0.044555768910794026, "grad_norm": 1.49161696434021, "learning_rate": 3.1115277311646434e-05, "loss": 2.6069, "step": 4954 }, { "epoch": 0.04456476280843448, "grad_norm": 2.70503568649292, "learning_rate": 3.1107989583092776e-05, "loss": 2.6098, "step": 4955 }, { "epoch": 0.04457375670607493, "grad_norm": 1.9890803098678589, "learning_rate": 3.110070130250559e-05, "loss": 2.7244, "step": 4956 }, { "epoch": 0.04458275060371538, "grad_norm": 1.4074339866638184, "learning_rate": 3.1093412470543596e-05, "loss": 2.5568, "step": 4957 }, { "epoch": 0.04459174450135583, "grad_norm": 2.1812543869018555, "learning_rate": 3.108612308786554e-05, "loss": 2.7692, "step": 4958 }, { "epoch": 0.04460073839899628, "grad_norm": 1.502809762954712, "learning_rate": 3.1078833155130244e-05, "loss": 2.6233, "step": 4959 }, { "epoch": 0.044609732296636735, "grad_norm": 2.028017520904541, "learning_rate": 3.1071542672996554e-05, "loss": 2.4982, "step": 4960 }, { "epoch": 0.04461872619427718, "grad_norm": 2.257486581802368, "learning_rate": 3.106425164212338e-05, "loss": 2.6607, "step": 4961 }, { "epoch": 0.044627720091917636, "grad_norm": 1.631014347076416, "learning_rate": 3.105696006316966e-05, "loss": 2.6445, "step": 4962 }, { "epoch": 0.04463671398955808, "grad_norm": 1.5814671516418457, "learning_rate": 3.104966793679442e-05, "loss": 2.6182, "step": 4963 }, { "epoch": 0.04464570788719854, "grad_norm": 1.6949599981307983, "learning_rate": 3.1042375263656716e-05, "loss": 2.663, "step": 4964 }, { "epoch": 0.044654701784838984, "grad_norm": 1.674662470817566, "learning_rate": 3.103508204441564e-05, "loss": 2.5388, "step": 4965 }, { "epoch": 0.04466369568247944, "grad_norm": 1.3896543979644775, "learning_rate": 3.102778827973034e-05, "loss": 2.3906, "step": 4966 }, { "epoch": 0.04467268958011989, "grad_norm": 1.819649338722229, "learning_rate": 3.1020493970260046e-05, "loss": 2.3766, "step": 4967 }, { "epoch": 0.04468168347776034, "grad_norm": 1.8337081670761108, "learning_rate": 3.1013199116663976e-05, "loss": 2.7209, "step": 4968 }, { "epoch": 0.04469067737540079, "grad_norm": 1.7790607213974, "learning_rate": 3.100590371960146e-05, "loss": 2.5894, "step": 4969 }, { "epoch": 0.04469967127304124, "grad_norm": 2.2056140899658203, "learning_rate": 3.099860777973182e-05, "loss": 2.9981, "step": 4970 }, { "epoch": 0.04470866517068169, "grad_norm": 1.2698993682861328, "learning_rate": 3.0991311297714476e-05, "loss": 2.3938, "step": 4971 }, { "epoch": 0.044717659068322146, "grad_norm": 1.0504125356674194, "learning_rate": 3.098401427420887e-05, "loss": 2.3329, "step": 4972 }, { "epoch": 0.04472665296596259, "grad_norm": 1.6110376119613647, "learning_rate": 3.0976716709874496e-05, "loss": 2.6333, "step": 4973 }, { "epoch": 0.04473564686360305, "grad_norm": 1.8951574563980103, "learning_rate": 3.0969418605370915e-05, "loss": 2.7821, "step": 4974 }, { "epoch": 0.044744640761243494, "grad_norm": 1.8190077543258667, "learning_rate": 3.096211996135771e-05, "loss": 2.586, "step": 4975 }, { "epoch": 0.04475363465888395, "grad_norm": 1.5987716913223267, "learning_rate": 3.0954820778494516e-05, "loss": 2.8599, "step": 4976 }, { "epoch": 0.044762628556524395, "grad_norm": 1.6270591020584106, "learning_rate": 3.094752105744103e-05, "loss": 2.5339, "step": 4977 }, { "epoch": 0.04477162245416485, "grad_norm": 1.4538217782974243, "learning_rate": 3.0940220798857e-05, "loss": 2.588, "step": 4978 }, { "epoch": 0.0447806163518053, "grad_norm": 1.2604185342788696, "learning_rate": 3.093292000340222e-05, "loss": 2.5627, "step": 4979 }, { "epoch": 0.04478961024944575, "grad_norm": 1.1532069444656372, "learning_rate": 3.0925618671736514e-05, "loss": 2.372, "step": 4980 }, { "epoch": 0.0447986041470862, "grad_norm": 2.0813069343566895, "learning_rate": 3.091831680451978e-05, "loss": 2.6314, "step": 4981 }, { "epoch": 0.04480759804472665, "grad_norm": 1.4380773305892944, "learning_rate": 3.091101440241195e-05, "loss": 2.5787, "step": 4982 }, { "epoch": 0.044816591942367104, "grad_norm": 2.592958450317383, "learning_rate": 3.090371146607299e-05, "loss": 3.0337, "step": 4983 }, { "epoch": 0.04482558584000756, "grad_norm": 1.5731925964355469, "learning_rate": 3.0896407996162954e-05, "loss": 2.5844, "step": 4984 }, { "epoch": 0.044834579737648005, "grad_norm": 1.8328701257705688, "learning_rate": 3.088910399334192e-05, "loss": 2.616, "step": 4985 }, { "epoch": 0.04484357363528846, "grad_norm": 1.8047834634780884, "learning_rate": 3.088179945827001e-05, "loss": 2.749, "step": 4986 }, { "epoch": 0.044852567532928905, "grad_norm": 1.407589316368103, "learning_rate": 3.08744943916074e-05, "loss": 2.7702, "step": 4987 }, { "epoch": 0.04486156143056936, "grad_norm": 1.7275018692016602, "learning_rate": 3.086718879401431e-05, "loss": 2.79, "step": 4988 }, { "epoch": 0.04487055532820981, "grad_norm": 1.672323226928711, "learning_rate": 3.0859882666151016e-05, "loss": 2.6404, "step": 4989 }, { "epoch": 0.04487954922585026, "grad_norm": 1.808928370475769, "learning_rate": 3.0852576008677843e-05, "loss": 2.4486, "step": 4990 }, { "epoch": 0.044888543123490714, "grad_norm": 1.9773120880126953, "learning_rate": 3.084526882225515e-05, "loss": 2.857, "step": 4991 }, { "epoch": 0.04489753702113116, "grad_norm": 1.33017098903656, "learning_rate": 3.083796110754336e-05, "loss": 2.5942, "step": 4992 }, { "epoch": 0.044906530918771614, "grad_norm": 0.9888942241668701, "learning_rate": 3.0830652865202934e-05, "loss": 2.342, "step": 4993 }, { "epoch": 0.04491552481641206, "grad_norm": 1.722355842590332, "learning_rate": 3.082334409589438e-05, "loss": 2.8489, "step": 4994 }, { "epoch": 0.044924518714052515, "grad_norm": 2.310755491256714, "learning_rate": 3.081603480027826e-05, "loss": 2.4433, "step": 4995 }, { "epoch": 0.04493351261169297, "grad_norm": 1.9311875104904175, "learning_rate": 3.080872497901518e-05, "loss": 2.691, "step": 4996 }, { "epoch": 0.044942506509333416, "grad_norm": 1.7248361110687256, "learning_rate": 3.080141463276579e-05, "loss": 2.8531, "step": 4997 }, { "epoch": 0.04495150040697387, "grad_norm": 1.54302978515625, "learning_rate": 3.07941037621908e-05, "loss": 2.4229, "step": 4998 }, { "epoch": 0.04496049430461432, "grad_norm": 1.9160557985305786, "learning_rate": 3.078679236795094e-05, "loss": 2.5384, "step": 4999 }, { "epoch": 0.04496948820225477, "grad_norm": 1.584702730178833, "learning_rate": 3.0779480450707035e-05, "loss": 2.4491, "step": 5000 }, { "epoch": 0.044978482099895224, "grad_norm": 1.8709205389022827, "learning_rate": 3.07721680111199e-05, "loss": 2.9629, "step": 5001 }, { "epoch": 0.04498747599753567, "grad_norm": 1.5241249799728394, "learning_rate": 3.076485504985044e-05, "loss": 2.4922, "step": 5002 }, { "epoch": 0.044996469895176125, "grad_norm": 3.0251266956329346, "learning_rate": 3.0757541567559594e-05, "loss": 2.6104, "step": 5003 }, { "epoch": 0.04500546379281657, "grad_norm": 2.3069543838500977, "learning_rate": 3.0750227564908346e-05, "loss": 2.6396, "step": 5004 }, { "epoch": 0.045014457690457026, "grad_norm": 1.8334057331085205, "learning_rate": 3.0742913042557716e-05, "loss": 2.6131, "step": 5005 }, { "epoch": 0.04502345158809747, "grad_norm": 1.7716655731201172, "learning_rate": 3.073559800116879e-05, "loss": 2.928, "step": 5006 }, { "epoch": 0.045032445485737926, "grad_norm": 1.2755329608917236, "learning_rate": 3.07282824414027e-05, "loss": 2.4254, "step": 5007 }, { "epoch": 0.04504143938337838, "grad_norm": 1.7272790670394897, "learning_rate": 3.072096636392061e-05, "loss": 2.764, "step": 5008 }, { "epoch": 0.04505043328101883, "grad_norm": 1.2483549118041992, "learning_rate": 3.071364976938375e-05, "loss": 2.3351, "step": 5009 }, { "epoch": 0.04505942717865928, "grad_norm": 1.8516470193862915, "learning_rate": 3.0706332658453376e-05, "loss": 2.7186, "step": 5010 }, { "epoch": 0.04506842107629973, "grad_norm": 1.9911584854125977, "learning_rate": 3.069901503179079e-05, "loss": 2.8056, "step": 5011 }, { "epoch": 0.04507741497394018, "grad_norm": 1.767412543296814, "learning_rate": 3.0691696890057375e-05, "loss": 2.4289, "step": 5012 }, { "epoch": 0.045086408871580635, "grad_norm": 2.2160532474517822, "learning_rate": 3.068437823391452e-05, "loss": 2.7941, "step": 5013 }, { "epoch": 0.04509540276922108, "grad_norm": 1.6904268264770508, "learning_rate": 3.067705906402368e-05, "loss": 2.4902, "step": 5014 }, { "epoch": 0.045104396666861536, "grad_norm": 1.5813502073287964, "learning_rate": 3.066973938104637e-05, "loss": 2.5499, "step": 5015 }, { "epoch": 0.04511339056450198, "grad_norm": 1.948965311050415, "learning_rate": 3.0662419185644115e-05, "loss": 3.0078, "step": 5016 }, { "epoch": 0.04512238446214244, "grad_norm": 1.7763299942016602, "learning_rate": 3.065509847847851e-05, "loss": 2.8469, "step": 5017 }, { "epoch": 0.04513137835978289, "grad_norm": 1.749259352684021, "learning_rate": 3.06477772602112e-05, "loss": 2.4018, "step": 5018 }, { "epoch": 0.04514037225742334, "grad_norm": 1.723496675491333, "learning_rate": 3.064045553150386e-05, "loss": 2.654, "step": 5019 }, { "epoch": 0.04514936615506379, "grad_norm": 1.814733862876892, "learning_rate": 3.063313329301822e-05, "loss": 2.5809, "step": 5020 }, { "epoch": 0.04515836005270424, "grad_norm": 1.6291708946228027, "learning_rate": 3.062581054541607e-05, "loss": 2.4184, "step": 5021 }, { "epoch": 0.04516735395034469, "grad_norm": 1.272611379623413, "learning_rate": 3.0618487289359224e-05, "loss": 2.3584, "step": 5022 }, { "epoch": 0.04517634784798514, "grad_norm": 1.7761573791503906, "learning_rate": 3.061116352550955e-05, "loss": 2.606, "step": 5023 }, { "epoch": 0.04518534174562559, "grad_norm": 1.4271390438079834, "learning_rate": 3.060383925452895e-05, "loss": 2.8723, "step": 5024 }, { "epoch": 0.04519433564326605, "grad_norm": 1.8552244901657104, "learning_rate": 3.05965144770794e-05, "loss": 2.7903, "step": 5025 }, { "epoch": 0.045203329540906494, "grad_norm": 1.3911141157150269, "learning_rate": 3.0589189193822895e-05, "loss": 2.4318, "step": 5026 }, { "epoch": 0.04521232343854695, "grad_norm": 1.696758508682251, "learning_rate": 3.058186340542149e-05, "loss": 2.607, "step": 5027 }, { "epoch": 0.045221317336187394, "grad_norm": 1.5247018337249756, "learning_rate": 3.05745371125373e-05, "loss": 2.468, "step": 5028 }, { "epoch": 0.04523031123382785, "grad_norm": 1.7392948865890503, "learning_rate": 3.056721031583243e-05, "loss": 2.4478, "step": 5029 }, { "epoch": 0.0452393051314683, "grad_norm": 1.7458994388580322, "learning_rate": 3.05598830159691e-05, "loss": 2.9361, "step": 5030 }, { "epoch": 0.04524829902910875, "grad_norm": 1.3605570793151855, "learning_rate": 3.0552555213609525e-05, "loss": 2.607, "step": 5031 }, { "epoch": 0.0452572929267492, "grad_norm": 1.8071479797363281, "learning_rate": 3.0545226909415993e-05, "loss": 2.4425, "step": 5032 }, { "epoch": 0.04526628682438965, "grad_norm": 1.8920683860778809, "learning_rate": 3.053789810405082e-05, "loss": 2.3496, "step": 5033 }, { "epoch": 0.0452752807220301, "grad_norm": 1.9249176979064941, "learning_rate": 3.053056879817639e-05, "loss": 2.7834, "step": 5034 }, { "epoch": 0.04528427461967055, "grad_norm": 1.2770425081253052, "learning_rate": 3.0523238992455104e-05, "loss": 2.4139, "step": 5035 }, { "epoch": 0.045293268517311004, "grad_norm": 1.4331575632095337, "learning_rate": 3.0515908687549426e-05, "loss": 2.3642, "step": 5036 }, { "epoch": 0.04530226241495146, "grad_norm": 2.9937305450439453, "learning_rate": 3.050857788412187e-05, "loss": 2.7241, "step": 5037 }, { "epoch": 0.045311256312591905, "grad_norm": 1.5616581439971924, "learning_rate": 3.050124658283497e-05, "loss": 2.5705, "step": 5038 }, { "epoch": 0.04532025021023236, "grad_norm": 1.5489085912704468, "learning_rate": 3.0493914784351328e-05, "loss": 2.3469, "step": 5039 }, { "epoch": 0.045329244107872806, "grad_norm": 1.3081082105636597, "learning_rate": 3.0486582489333598e-05, "loss": 2.5882, "step": 5040 }, { "epoch": 0.04533823800551326, "grad_norm": 1.4067621231079102, "learning_rate": 3.047924969844444e-05, "loss": 2.5637, "step": 5041 }, { "epoch": 0.04534723190315371, "grad_norm": 1.4100406169891357, "learning_rate": 3.0471916412346602e-05, "loss": 2.5387, "step": 5042 }, { "epoch": 0.04535622580079416, "grad_norm": 2.41214919090271, "learning_rate": 3.0464582631702854e-05, "loss": 3.0559, "step": 5043 }, { "epoch": 0.045365219698434614, "grad_norm": 1.674769401550293, "learning_rate": 3.0457248357176004e-05, "loss": 2.6326, "step": 5044 }, { "epoch": 0.04537421359607506, "grad_norm": 0.9204654097557068, "learning_rate": 3.044991358942893e-05, "loss": 2.321, "step": 5045 }, { "epoch": 0.045383207493715515, "grad_norm": 1.47079336643219, "learning_rate": 3.0442578329124544e-05, "loss": 2.4254, "step": 5046 }, { "epoch": 0.04539220139135596, "grad_norm": 1.364933967590332, "learning_rate": 3.0435242576925783e-05, "loss": 2.4812, "step": 5047 }, { "epoch": 0.045401195288996415, "grad_norm": 2.0005533695220947, "learning_rate": 3.042790633349565e-05, "loss": 2.7981, "step": 5048 }, { "epoch": 0.04541018918663687, "grad_norm": 1.9175353050231934, "learning_rate": 3.042056959949719e-05, "loss": 2.5403, "step": 5049 }, { "epoch": 0.045419183084277316, "grad_norm": 1.4885615110397339, "learning_rate": 3.0413232375593497e-05, "loss": 2.6525, "step": 5050 }, { "epoch": 0.04542817698191777, "grad_norm": 1.6191736459732056, "learning_rate": 3.040589466244768e-05, "loss": 2.7562, "step": 5051 }, { "epoch": 0.04543717087955822, "grad_norm": 1.667685627937317, "learning_rate": 3.0398556460722938e-05, "loss": 2.7545, "step": 5052 }, { "epoch": 0.04544616477719867, "grad_norm": 1.3980356454849243, "learning_rate": 3.039121777108247e-05, "loss": 2.3061, "step": 5053 }, { "epoch": 0.045455158674839125, "grad_norm": 1.1342909336090088, "learning_rate": 3.0383878594189552e-05, "loss": 2.315, "step": 5054 }, { "epoch": 0.04546415257247957, "grad_norm": 1.570957064628601, "learning_rate": 3.0376538930707483e-05, "loss": 2.7309, "step": 5055 }, { "epoch": 0.045473146470120025, "grad_norm": 1.6409451961517334, "learning_rate": 3.036919878129961e-05, "loss": 2.6471, "step": 5056 }, { "epoch": 0.04548214036776047, "grad_norm": 2.060184955596924, "learning_rate": 3.0361858146629345e-05, "loss": 2.6167, "step": 5057 }, { "epoch": 0.045491134265400926, "grad_norm": 1.931464672088623, "learning_rate": 3.0354517027360118e-05, "loss": 2.6093, "step": 5058 }, { "epoch": 0.04550012816304138, "grad_norm": 1.5562903881072998, "learning_rate": 3.03471754241554e-05, "loss": 2.4283, "step": 5059 }, { "epoch": 0.04550912206068183, "grad_norm": 1.6758593320846558, "learning_rate": 3.033983333767873e-05, "loss": 2.4065, "step": 5060 }, { "epoch": 0.04551811595832228, "grad_norm": 1.096084713935852, "learning_rate": 3.0332490768593675e-05, "loss": 2.3151, "step": 5061 }, { "epoch": 0.04552710985596273, "grad_norm": 1.9293029308319092, "learning_rate": 3.0325147717563847e-05, "loss": 2.7121, "step": 5062 }, { "epoch": 0.04553610375360318, "grad_norm": 2.1146297454833984, "learning_rate": 3.0317804185252906e-05, "loss": 2.894, "step": 5063 }, { "epoch": 0.04554509765124363, "grad_norm": 1.6669520139694214, "learning_rate": 3.0310460172324545e-05, "loss": 2.7343, "step": 5064 }, { "epoch": 0.04555409154888408, "grad_norm": 1.8885160684585571, "learning_rate": 3.0303115679442524e-05, "loss": 2.7764, "step": 5065 }, { "epoch": 0.045563085446524536, "grad_norm": 1.2782649993896484, "learning_rate": 3.029577070727061e-05, "loss": 2.3805, "step": 5066 }, { "epoch": 0.04557207934416498, "grad_norm": 1.0914913415908813, "learning_rate": 3.0288425256472647e-05, "loss": 2.2815, "step": 5067 }, { "epoch": 0.045581073241805437, "grad_norm": 1.7914258241653442, "learning_rate": 3.0281079327712504e-05, "loss": 3.0127, "step": 5068 }, { "epoch": 0.04559006713944588, "grad_norm": 1.4227867126464844, "learning_rate": 3.027373292165411e-05, "loss": 2.4947, "step": 5069 }, { "epoch": 0.04559906103708634, "grad_norm": 1.2338289022445679, "learning_rate": 3.0266386038961402e-05, "loss": 2.3272, "step": 5070 }, { "epoch": 0.04560805493472679, "grad_norm": 1.7619990110397339, "learning_rate": 3.02590386802984e-05, "loss": 2.5791, "step": 5071 }, { "epoch": 0.04561704883236724, "grad_norm": 1.4732246398925781, "learning_rate": 3.025169084632915e-05, "loss": 2.5587, "step": 5072 }, { "epoch": 0.04562604273000769, "grad_norm": 1.7394261360168457, "learning_rate": 3.0244342537717734e-05, "loss": 2.4976, "step": 5073 }, { "epoch": 0.04563503662764814, "grad_norm": 2.0734071731567383, "learning_rate": 3.0236993755128283e-05, "loss": 2.7964, "step": 5074 }, { "epoch": 0.04564403052528859, "grad_norm": 2.045464515686035, "learning_rate": 3.022964449922499e-05, "loss": 2.5525, "step": 5075 }, { "epoch": 0.04565302442292904, "grad_norm": 1.946346640586853, "learning_rate": 3.0222294770672053e-05, "loss": 2.6466, "step": 5076 }, { "epoch": 0.04566201832056949, "grad_norm": 1.2477601766586304, "learning_rate": 3.021494457013374e-05, "loss": 2.4648, "step": 5077 }, { "epoch": 0.04567101221820995, "grad_norm": 1.615607500076294, "learning_rate": 3.0207593898274354e-05, "loss": 2.6151, "step": 5078 }, { "epoch": 0.045680006115850394, "grad_norm": 2.0936262607574463, "learning_rate": 3.0200242755758246e-05, "loss": 2.8528, "step": 5079 }, { "epoch": 0.04568900001349085, "grad_norm": 1.4663069248199463, "learning_rate": 3.019289114324979e-05, "loss": 2.6939, "step": 5080 }, { "epoch": 0.045697993911131295, "grad_norm": 1.257479190826416, "learning_rate": 3.018553906141343e-05, "loss": 2.4352, "step": 5081 }, { "epoch": 0.04570698780877175, "grad_norm": 1.5102845430374146, "learning_rate": 3.0178186510913635e-05, "loss": 2.4013, "step": 5082 }, { "epoch": 0.0457159817064122, "grad_norm": 1.96958327293396, "learning_rate": 3.017083349241492e-05, "loss": 2.5701, "step": 5083 }, { "epoch": 0.04572497560405265, "grad_norm": 2.1231656074523926, "learning_rate": 3.0163480006581845e-05, "loss": 3.2086, "step": 5084 }, { "epoch": 0.0457339695016931, "grad_norm": 1.332980751991272, "learning_rate": 3.015612605407901e-05, "loss": 2.4072, "step": 5085 }, { "epoch": 0.04574296339933355, "grad_norm": 1.1533913612365723, "learning_rate": 3.014877163557105e-05, "loss": 2.2873, "step": 5086 }, { "epoch": 0.045751957296974004, "grad_norm": 2.037344217300415, "learning_rate": 3.014141675172266e-05, "loss": 2.6076, "step": 5087 }, { "epoch": 0.04576095119461445, "grad_norm": 2.292468547821045, "learning_rate": 3.0134061403198566e-05, "loss": 2.5844, "step": 5088 }, { "epoch": 0.045769945092254904, "grad_norm": 2.7759413719177246, "learning_rate": 3.0126705590663522e-05, "loss": 2.6147, "step": 5089 }, { "epoch": 0.04577893898989536, "grad_norm": 2.1645495891571045, "learning_rate": 3.011934931478235e-05, "loss": 2.5493, "step": 5090 }, { "epoch": 0.045787932887535805, "grad_norm": 1.8371108770370483, "learning_rate": 3.0111992576219906e-05, "loss": 2.623, "step": 5091 }, { "epoch": 0.04579692678517626, "grad_norm": 1.5752075910568237, "learning_rate": 3.0104635375641083e-05, "loss": 2.5742, "step": 5092 }, { "epoch": 0.045805920682816706, "grad_norm": 2.317467212677002, "learning_rate": 3.009727771371081e-05, "loss": 2.9054, "step": 5093 }, { "epoch": 0.04581491458045716, "grad_norm": 1.1759788990020752, "learning_rate": 3.008991959109406e-05, "loss": 2.3143, "step": 5094 }, { "epoch": 0.045823908478097614, "grad_norm": 1.7885408401489258, "learning_rate": 3.0082561008455868e-05, "loss": 2.7512, "step": 5095 }, { "epoch": 0.04583290237573806, "grad_norm": 1.7976435422897339, "learning_rate": 3.0075201966461285e-05, "loss": 2.6566, "step": 5096 }, { "epoch": 0.045841896273378514, "grad_norm": 2.1845715045928955, "learning_rate": 3.006784246577541e-05, "loss": 2.8685, "step": 5097 }, { "epoch": 0.04585089017101896, "grad_norm": 2.3951046466827393, "learning_rate": 3.0060482507063398e-05, "loss": 2.4439, "step": 5098 }, { "epoch": 0.045859884068659415, "grad_norm": 1.68977689743042, "learning_rate": 3.0053122090990426e-05, "loss": 2.5536, "step": 5099 }, { "epoch": 0.04586887796629987, "grad_norm": 1.9814519882202148, "learning_rate": 3.0045761218221725e-05, "loss": 2.3756, "step": 5100 }, { "epoch": 0.045877871863940316, "grad_norm": 1.5072954893112183, "learning_rate": 3.0038399889422553e-05, "loss": 2.6168, "step": 5101 }, { "epoch": 0.04588686576158077, "grad_norm": 1.6794332265853882, "learning_rate": 3.003103810525823e-05, "loss": 2.4933, "step": 5102 }, { "epoch": 0.045895859659221216, "grad_norm": 1.3541661500930786, "learning_rate": 3.0023675866394098e-05, "loss": 2.3843, "step": 5103 }, { "epoch": 0.04590485355686167, "grad_norm": 1.829084873199463, "learning_rate": 3.0016313173495554e-05, "loss": 2.5127, "step": 5104 }, { "epoch": 0.04591384745450212, "grad_norm": 1.355436086654663, "learning_rate": 3.0008950027228033e-05, "loss": 2.4456, "step": 5105 }, { "epoch": 0.04592284135214257, "grad_norm": 1.928631067276001, "learning_rate": 3.0001586428257005e-05, "loss": 2.6834, "step": 5106 }, { "epoch": 0.045931835249783025, "grad_norm": 1.778913140296936, "learning_rate": 2.9994222377247978e-05, "loss": 2.6592, "step": 5107 }, { "epoch": 0.04594082914742347, "grad_norm": 1.3901762962341309, "learning_rate": 2.998685787486651e-05, "loss": 2.4827, "step": 5108 }, { "epoch": 0.045949823045063926, "grad_norm": 1.3923003673553467, "learning_rate": 2.9979492921778195e-05, "loss": 2.4478, "step": 5109 }, { "epoch": 0.04595881694270437, "grad_norm": 1.6988650560379028, "learning_rate": 2.997212751864868e-05, "loss": 2.6855, "step": 5110 }, { "epoch": 0.045967810840344826, "grad_norm": 2.099416971206665, "learning_rate": 2.996476166614364e-05, "loss": 2.6984, "step": 5111 }, { "epoch": 0.04597680473798528, "grad_norm": 2.5289113521575928, "learning_rate": 2.9957395364928782e-05, "loss": 2.8414, "step": 5112 }, { "epoch": 0.04598579863562573, "grad_norm": 1.9750722646713257, "learning_rate": 2.995002861566987e-05, "loss": 2.7798, "step": 5113 }, { "epoch": 0.04599479253326618, "grad_norm": 1.7288841009140015, "learning_rate": 2.994266141903271e-05, "loss": 2.7035, "step": 5114 }, { "epoch": 0.04600378643090663, "grad_norm": 1.4493359327316284, "learning_rate": 2.9935293775683126e-05, "loss": 2.4098, "step": 5115 }, { "epoch": 0.04601278032854708, "grad_norm": 1.7378653287887573, "learning_rate": 2.9927925686287006e-05, "loss": 2.6495, "step": 5116 }, { "epoch": 0.04602177422618753, "grad_norm": 1.769308090209961, "learning_rate": 2.9920557151510283e-05, "loss": 2.6873, "step": 5117 }, { "epoch": 0.04603076812382798, "grad_norm": 2.1859922409057617, "learning_rate": 2.991318817201889e-05, "loss": 2.9808, "step": 5118 }, { "epoch": 0.046039762021468436, "grad_norm": 2.0419669151306152, "learning_rate": 2.9905818748478848e-05, "loss": 2.8831, "step": 5119 }, { "epoch": 0.04604875591910888, "grad_norm": 1.2818669080734253, "learning_rate": 2.9898448881556197e-05, "loss": 2.5953, "step": 5120 }, { "epoch": 0.04605774981674934, "grad_norm": 1.695204496383667, "learning_rate": 2.9891078571917004e-05, "loss": 2.5061, "step": 5121 }, { "epoch": 0.046066743714389784, "grad_norm": 1.3887547254562378, "learning_rate": 2.9883707820227403e-05, "loss": 2.4702, "step": 5122 }, { "epoch": 0.04607573761203024, "grad_norm": 1.948201060295105, "learning_rate": 2.987633662715355e-05, "loss": 2.6319, "step": 5123 }, { "epoch": 0.04608473150967069, "grad_norm": 1.607787013053894, "learning_rate": 2.9868964993361636e-05, "loss": 2.5709, "step": 5124 }, { "epoch": 0.04609372540731114, "grad_norm": 1.4717470407485962, "learning_rate": 2.9861592919517917e-05, "loss": 2.7237, "step": 5125 }, { "epoch": 0.04610271930495159, "grad_norm": 2.0451717376708984, "learning_rate": 2.9854220406288668e-05, "loss": 2.9634, "step": 5126 }, { "epoch": 0.04611171320259204, "grad_norm": 2.2258031368255615, "learning_rate": 2.984684745434021e-05, "loss": 2.7244, "step": 5127 }, { "epoch": 0.04612070710023249, "grad_norm": 1.4981200695037842, "learning_rate": 2.983947406433889e-05, "loss": 2.4508, "step": 5128 }, { "epoch": 0.04612970099787295, "grad_norm": 1.9707835912704468, "learning_rate": 2.9832100236951126e-05, "loss": 2.9065, "step": 5129 }, { "epoch": 0.046138694895513394, "grad_norm": 1.2059922218322754, "learning_rate": 2.982472597284334e-05, "loss": 2.3168, "step": 5130 }, { "epoch": 0.04614768879315385, "grad_norm": 1.391526460647583, "learning_rate": 2.981735127268202e-05, "loss": 2.3505, "step": 5131 }, { "epoch": 0.046156682690794294, "grad_norm": 1.423432469367981, "learning_rate": 2.9809976137133677e-05, "loss": 2.4984, "step": 5132 }, { "epoch": 0.04616567658843475, "grad_norm": 1.7532024383544922, "learning_rate": 2.980260056686488e-05, "loss": 3.1256, "step": 5133 }, { "epoch": 0.046174670486075195, "grad_norm": 2.0758934020996094, "learning_rate": 2.9795224562542212e-05, "loss": 2.4406, "step": 5134 }, { "epoch": 0.04618366438371565, "grad_norm": 1.6501590013504028, "learning_rate": 2.978784812483232e-05, "loss": 2.6105, "step": 5135 }, { "epoch": 0.0461926582813561, "grad_norm": 1.8880908489227295, "learning_rate": 2.9780471254401864e-05, "loss": 2.5347, "step": 5136 }, { "epoch": 0.04620165217899655, "grad_norm": 1.5346909761428833, "learning_rate": 2.9773093951917563e-05, "loss": 2.5613, "step": 5137 }, { "epoch": 0.046210646076637, "grad_norm": 2.4164583683013916, "learning_rate": 2.9765716218046175e-05, "loss": 2.7078, "step": 5138 }, { "epoch": 0.04621963997427745, "grad_norm": 1.0697046518325806, "learning_rate": 2.975833805345448e-05, "loss": 2.4798, "step": 5139 }, { "epoch": 0.046228633871917904, "grad_norm": 1.6173056364059448, "learning_rate": 2.9750959458809336e-05, "loss": 2.7935, "step": 5140 }, { "epoch": 0.04623762776955836, "grad_norm": 1.935653567314148, "learning_rate": 2.9743580434777587e-05, "loss": 2.7484, "step": 5141 }, { "epoch": 0.046246621667198805, "grad_norm": 1.8236017227172852, "learning_rate": 2.973620098202614e-05, "loss": 2.7775, "step": 5142 }, { "epoch": 0.04625561556483926, "grad_norm": 2.207111358642578, "learning_rate": 2.9728821101221948e-05, "loss": 2.5935, "step": 5143 }, { "epoch": 0.046264609462479706, "grad_norm": 1.8763893842697144, "learning_rate": 2.9721440793032007e-05, "loss": 2.5396, "step": 5144 }, { "epoch": 0.04627360336012016, "grad_norm": 1.4599844217300415, "learning_rate": 2.9714060058123328e-05, "loss": 2.4401, "step": 5145 }, { "epoch": 0.046282597257760606, "grad_norm": 1.6280019283294678, "learning_rate": 2.970667889716298e-05, "loss": 2.6788, "step": 5146 }, { "epoch": 0.04629159115540106, "grad_norm": 1.6209971904754639, "learning_rate": 2.9699297310818063e-05, "loss": 2.6018, "step": 5147 }, { "epoch": 0.046300585053041514, "grad_norm": 1.2544103860855103, "learning_rate": 2.9691915299755718e-05, "loss": 2.3951, "step": 5148 }, { "epoch": 0.04630957895068196, "grad_norm": 1.9226372241973877, "learning_rate": 2.9684532864643122e-05, "loss": 2.6469, "step": 5149 }, { "epoch": 0.046318572848322415, "grad_norm": 1.7031368017196655, "learning_rate": 2.9677150006147485e-05, "loss": 2.7254, "step": 5150 }, { "epoch": 0.04632756674596286, "grad_norm": 1.9292023181915283, "learning_rate": 2.966976672493607e-05, "loss": 2.7639, "step": 5151 }, { "epoch": 0.046336560643603315, "grad_norm": 1.0474390983581543, "learning_rate": 2.9662383021676175e-05, "loss": 2.3964, "step": 5152 }, { "epoch": 0.04634555454124377, "grad_norm": 1.5797382593154907, "learning_rate": 2.965499889703512e-05, "loss": 2.6032, "step": 5153 }, { "epoch": 0.046354548438884216, "grad_norm": 1.6905065774917603, "learning_rate": 2.964761435168028e-05, "loss": 2.4237, "step": 5154 }, { "epoch": 0.04636354233652467, "grad_norm": 1.4787241220474243, "learning_rate": 2.9640229386279063e-05, "loss": 2.665, "step": 5155 }, { "epoch": 0.04637253623416512, "grad_norm": 2.0938563346862793, "learning_rate": 2.9632844001498905e-05, "loss": 2.917, "step": 5156 }, { "epoch": 0.04638153013180557, "grad_norm": 1.7919472455978394, "learning_rate": 2.96254581980073e-05, "loss": 2.6249, "step": 5157 }, { "epoch": 0.04639052402944602, "grad_norm": 1.541499376296997, "learning_rate": 2.9618071976471773e-05, "loss": 2.4, "step": 5158 }, { "epoch": 0.04639951792708647, "grad_norm": 1.5933542251586914, "learning_rate": 2.961068533755986e-05, "loss": 2.4098, "step": 5159 }, { "epoch": 0.046408511824726925, "grad_norm": 1.2546675205230713, "learning_rate": 2.9603298281939178e-05, "loss": 2.2463, "step": 5160 }, { "epoch": 0.04641750572236737, "grad_norm": 1.6619253158569336, "learning_rate": 2.9595910810277365e-05, "loss": 2.6708, "step": 5161 }, { "epoch": 0.046426499620007826, "grad_norm": 1.2502926588058472, "learning_rate": 2.9588522923242073e-05, "loss": 2.409, "step": 5162 }, { "epoch": 0.04643549351764827, "grad_norm": 1.6257622241973877, "learning_rate": 2.9581134621501023e-05, "loss": 2.651, "step": 5163 }, { "epoch": 0.04644448741528873, "grad_norm": 1.651869773864746, "learning_rate": 2.9573745905721967e-05, "loss": 2.6822, "step": 5164 }, { "epoch": 0.04645348131292918, "grad_norm": 1.7246384620666504, "learning_rate": 2.9566356776572672e-05, "loss": 2.6701, "step": 5165 }, { "epoch": 0.04646247521056963, "grad_norm": 0.8797203302383423, "learning_rate": 2.9558967234720974e-05, "loss": 2.2718, "step": 5166 }, { "epoch": 0.04647146910821008, "grad_norm": 1.9863002300262451, "learning_rate": 2.955157728083473e-05, "loss": 2.9756, "step": 5167 }, { "epoch": 0.04648046300585053, "grad_norm": 1.8400846719741821, "learning_rate": 2.9544186915581835e-05, "loss": 2.6955, "step": 5168 }, { "epoch": 0.04648945690349098, "grad_norm": 1.346163272857666, "learning_rate": 2.9536796139630217e-05, "loss": 2.4444, "step": 5169 }, { "epoch": 0.046498450801131436, "grad_norm": 1.7258570194244385, "learning_rate": 2.952940495364786e-05, "loss": 2.4908, "step": 5170 }, { "epoch": 0.04650744469877188, "grad_norm": 1.6468545198440552, "learning_rate": 2.952201335830275e-05, "loss": 2.7478, "step": 5171 }, { "epoch": 0.046516438596412336, "grad_norm": 1.8136582374572754, "learning_rate": 2.951462135426295e-05, "loss": 2.7809, "step": 5172 }, { "epoch": 0.04652543249405278, "grad_norm": 1.2907079458236694, "learning_rate": 2.950722894219653e-05, "loss": 2.4895, "step": 5173 }, { "epoch": 0.04653442639169324, "grad_norm": 1.8509353399276733, "learning_rate": 2.9499836122771612e-05, "loss": 2.616, "step": 5174 }, { "epoch": 0.046543420289333684, "grad_norm": 1.0441138744354248, "learning_rate": 2.949244289665636e-05, "loss": 2.3142, "step": 5175 }, { "epoch": 0.04655241418697414, "grad_norm": 2.1249680519104004, "learning_rate": 2.948504926451896e-05, "loss": 2.7469, "step": 5176 }, { "epoch": 0.04656140808461459, "grad_norm": 1.3761355876922607, "learning_rate": 2.947765522702763e-05, "loss": 2.6273, "step": 5177 }, { "epoch": 0.04657040198225504, "grad_norm": 1.4108328819274902, "learning_rate": 2.947026078485064e-05, "loss": 2.3476, "step": 5178 }, { "epoch": 0.04657939587989549, "grad_norm": 1.9041452407836914, "learning_rate": 2.9462865938656304e-05, "loss": 2.7305, "step": 5179 }, { "epoch": 0.04658838977753594, "grad_norm": 1.9314167499542236, "learning_rate": 2.9455470689112942e-05, "loss": 2.6778, "step": 5180 }, { "epoch": 0.04659738367517639, "grad_norm": 1.2239360809326172, "learning_rate": 2.9448075036888944e-05, "loss": 2.4874, "step": 5181 }, { "epoch": 0.04660637757281685, "grad_norm": 1.396801233291626, "learning_rate": 2.944067898265272e-05, "loss": 2.3815, "step": 5182 }, { "epoch": 0.046615371470457294, "grad_norm": 2.151740312576294, "learning_rate": 2.9433282527072715e-05, "loss": 2.8012, "step": 5183 }, { "epoch": 0.04662436536809775, "grad_norm": 1.368455171585083, "learning_rate": 2.94258856708174e-05, "loss": 2.4233, "step": 5184 }, { "epoch": 0.046633359265738195, "grad_norm": 1.4328324794769287, "learning_rate": 2.9418488414555306e-05, "loss": 2.2163, "step": 5185 }, { "epoch": 0.04664235316337865, "grad_norm": 1.6792142391204834, "learning_rate": 2.941109075895499e-05, "loss": 2.6598, "step": 5186 }, { "epoch": 0.046651347061019095, "grad_norm": 1.2814385890960693, "learning_rate": 2.9403692704685038e-05, "loss": 2.4127, "step": 5187 }, { "epoch": 0.04666034095865955, "grad_norm": 1.639710783958435, "learning_rate": 2.9396294252414092e-05, "loss": 2.5498, "step": 5188 }, { "epoch": 0.0466693348563, "grad_norm": 1.6983884572982788, "learning_rate": 2.93888954028108e-05, "loss": 2.531, "step": 5189 }, { "epoch": 0.04667832875394045, "grad_norm": 1.7214444875717163, "learning_rate": 2.938149615654388e-05, "loss": 2.4761, "step": 5190 }, { "epoch": 0.046687322651580904, "grad_norm": 1.5819867849349976, "learning_rate": 2.9374096514282047e-05, "loss": 2.6747, "step": 5191 }, { "epoch": 0.04669631654922135, "grad_norm": 1.6921467781066895, "learning_rate": 2.9366696476694083e-05, "loss": 2.7604, "step": 5192 }, { "epoch": 0.046705310446861804, "grad_norm": 1.6340128183364868, "learning_rate": 2.9359296044448794e-05, "loss": 2.5087, "step": 5193 }, { "epoch": 0.04671430434450226, "grad_norm": 1.5635424852371216, "learning_rate": 2.9351895218215032e-05, "loss": 2.5091, "step": 5194 }, { "epoch": 0.046723298242142705, "grad_norm": 1.4454176425933838, "learning_rate": 2.9344493998661654e-05, "loss": 2.4917, "step": 5195 }, { "epoch": 0.04673229213978316, "grad_norm": 1.6475478410720825, "learning_rate": 2.93370923864576e-05, "loss": 2.7814, "step": 5196 }, { "epoch": 0.046741286037423606, "grad_norm": 1.303037166595459, "learning_rate": 2.93296903822718e-05, "loss": 2.2953, "step": 5197 }, { "epoch": 0.04675027993506406, "grad_norm": 1.5207717418670654, "learning_rate": 2.932228798677325e-05, "loss": 2.4378, "step": 5198 }, { "epoch": 0.04675927383270451, "grad_norm": 1.457772970199585, "learning_rate": 2.9314885200630963e-05, "loss": 2.7072, "step": 5199 }, { "epoch": 0.04676826773034496, "grad_norm": 1.7443163394927979, "learning_rate": 2.930748202451401e-05, "loss": 2.6237, "step": 5200 }, { "epoch": 0.046777261627985414, "grad_norm": 1.805065631866455, "learning_rate": 2.9300078459091462e-05, "loss": 2.8297, "step": 5201 }, { "epoch": 0.04678625552562586, "grad_norm": 1.2514055967330933, "learning_rate": 2.929267450503246e-05, "loss": 2.2584, "step": 5202 }, { "epoch": 0.046795249423266315, "grad_norm": 2.0525450706481934, "learning_rate": 2.9285270163006158e-05, "loss": 2.7271, "step": 5203 }, { "epoch": 0.04680424332090676, "grad_norm": 0.7115890979766846, "learning_rate": 2.927786543368175e-05, "loss": 2.2094, "step": 5204 }, { "epoch": 0.046813237218547216, "grad_norm": 1.720037817955017, "learning_rate": 2.9270460317728483e-05, "loss": 2.8293, "step": 5205 }, { "epoch": 0.04682223111618767, "grad_norm": 1.8226114511489868, "learning_rate": 2.92630548158156e-05, "loss": 2.6216, "step": 5206 }, { "epoch": 0.046831225013828116, "grad_norm": 1.6012665033340454, "learning_rate": 2.9255648928612418e-05, "loss": 2.5447, "step": 5207 }, { "epoch": 0.04684021891146857, "grad_norm": 2.202984571456909, "learning_rate": 2.9248242656788267e-05, "loss": 2.6775, "step": 5208 }, { "epoch": 0.04684921280910902, "grad_norm": 1.2732160091400146, "learning_rate": 2.924083600101252e-05, "loss": 2.4321, "step": 5209 }, { "epoch": 0.04685820670674947, "grad_norm": 1.099295973777771, "learning_rate": 2.9233428961954597e-05, "loss": 2.343, "step": 5210 }, { "epoch": 0.046867200604389925, "grad_norm": 1.663979172706604, "learning_rate": 2.9226021540283914e-05, "loss": 2.6534, "step": 5211 }, { "epoch": 0.04687619450203037, "grad_norm": 1.4272494316101074, "learning_rate": 2.921861373666996e-05, "loss": 2.2824, "step": 5212 }, { "epoch": 0.046885188399670825, "grad_norm": 1.613844394683838, "learning_rate": 2.9211205551782235e-05, "loss": 2.4242, "step": 5213 }, { "epoch": 0.04689418229731127, "grad_norm": 1.438167929649353, "learning_rate": 2.9203796986290282e-05, "loss": 2.5453, "step": 5214 }, { "epoch": 0.046903176194951726, "grad_norm": 1.644103765487671, "learning_rate": 2.9196388040863693e-05, "loss": 2.8228, "step": 5215 }, { "epoch": 0.04691217009259217, "grad_norm": 1.6755105257034302, "learning_rate": 2.918897871617207e-05, "loss": 2.5258, "step": 5216 }, { "epoch": 0.04692116399023263, "grad_norm": 2.0151047706604004, "learning_rate": 2.9181569012885072e-05, "loss": 2.8851, "step": 5217 }, { "epoch": 0.04693015788787308, "grad_norm": 1.8524338006973267, "learning_rate": 2.917415893167237e-05, "loss": 2.5268, "step": 5218 }, { "epoch": 0.04693915178551353, "grad_norm": 1.767845869064331, "learning_rate": 2.9166748473203675e-05, "loss": 2.4927, "step": 5219 }, { "epoch": 0.04694814568315398, "grad_norm": 0.9855478405952454, "learning_rate": 2.9159337638148736e-05, "loss": 2.3079, "step": 5220 }, { "epoch": 0.04695713958079443, "grad_norm": 1.7930481433868408, "learning_rate": 2.9151926427177346e-05, "loss": 2.5362, "step": 5221 }, { "epoch": 0.04696613347843488, "grad_norm": 1.5947368144989014, "learning_rate": 2.9144514840959315e-05, "loss": 2.7291, "step": 5222 }, { "epoch": 0.046975127376075336, "grad_norm": 1.7026904821395874, "learning_rate": 2.9137102880164503e-05, "loss": 2.5202, "step": 5223 }, { "epoch": 0.04698412127371578, "grad_norm": 1.2936599254608154, "learning_rate": 2.9129690545462783e-05, "loss": 2.3914, "step": 5224 }, { "epoch": 0.04699311517135624, "grad_norm": 2.0711419582366943, "learning_rate": 2.9122277837524085e-05, "loss": 2.8107, "step": 5225 }, { "epoch": 0.047002109068996684, "grad_norm": 1.3051625490188599, "learning_rate": 2.9114864757018352e-05, "loss": 2.489, "step": 5226 }, { "epoch": 0.04701110296663714, "grad_norm": 2.285130500793457, "learning_rate": 2.9107451304615574e-05, "loss": 2.5149, "step": 5227 }, { "epoch": 0.047020096864277584, "grad_norm": 1.6021844148635864, "learning_rate": 2.910003748098577e-05, "loss": 2.7929, "step": 5228 }, { "epoch": 0.04702909076191804, "grad_norm": 1.6722134351730347, "learning_rate": 2.9092623286799e-05, "loss": 2.7119, "step": 5229 }, { "epoch": 0.04703808465955849, "grad_norm": 1.3409628868103027, "learning_rate": 2.9085208722725342e-05, "loss": 2.4896, "step": 5230 }, { "epoch": 0.04704707855719894, "grad_norm": 2.099393606185913, "learning_rate": 2.9077793789434926e-05, "loss": 2.6594, "step": 5231 }, { "epoch": 0.04705607245483939, "grad_norm": 1.5398826599121094, "learning_rate": 2.907037848759789e-05, "loss": 2.6034, "step": 5232 }, { "epoch": 0.04706506635247984, "grad_norm": 1.584682822227478, "learning_rate": 2.906296281788443e-05, "loss": 2.5939, "step": 5233 }, { "epoch": 0.04707406025012029, "grad_norm": 2.1052792072296143, "learning_rate": 2.905554678096477e-05, "loss": 2.8001, "step": 5234 }, { "epoch": 0.04708305414776075, "grad_norm": 1.983749508857727, "learning_rate": 2.904813037750917e-05, "loss": 2.7751, "step": 5235 }, { "epoch": 0.047092048045401194, "grad_norm": 2.158062219619751, "learning_rate": 2.9040713608187898e-05, "loss": 2.5441, "step": 5236 }, { "epoch": 0.04710104194304165, "grad_norm": 1.6893008947372437, "learning_rate": 2.9033296473671278e-05, "loss": 2.6167, "step": 5237 }, { "epoch": 0.047110035840682095, "grad_norm": 2.031414747238159, "learning_rate": 2.9025878974629678e-05, "loss": 2.8122, "step": 5238 }, { "epoch": 0.04711902973832255, "grad_norm": 2.14980411529541, "learning_rate": 2.9018461111733465e-05, "loss": 2.3937, "step": 5239 }, { "epoch": 0.047128023635963, "grad_norm": 1.8360207080841064, "learning_rate": 2.901104288565306e-05, "loss": 2.3453, "step": 5240 }, { "epoch": 0.04713701753360345, "grad_norm": 1.106438159942627, "learning_rate": 2.900362429705893e-05, "loss": 2.2909, "step": 5241 }, { "epoch": 0.0471460114312439, "grad_norm": 2.0300140380859375, "learning_rate": 2.899620534662154e-05, "loss": 2.498, "step": 5242 }, { "epoch": 0.04715500532888435, "grad_norm": 1.639299988746643, "learning_rate": 2.8988786035011424e-05, "loss": 2.3622, "step": 5243 }, { "epoch": 0.047163999226524804, "grad_norm": 1.7668744325637817, "learning_rate": 2.8981366362899113e-05, "loss": 2.4163, "step": 5244 }, { "epoch": 0.04717299312416525, "grad_norm": 1.7007426023483276, "learning_rate": 2.897394633095521e-05, "loss": 2.4837, "step": 5245 }, { "epoch": 0.047181987021805705, "grad_norm": 1.2346627712249756, "learning_rate": 2.896652593985031e-05, "loss": 2.396, "step": 5246 }, { "epoch": 0.04719098091944616, "grad_norm": 1.3548049926757812, "learning_rate": 2.895910519025507e-05, "loss": 2.3987, "step": 5247 }, { "epoch": 0.047199974817086605, "grad_norm": 1.4447277784347534, "learning_rate": 2.8951684082840163e-05, "loss": 2.6099, "step": 5248 }, { "epoch": 0.04720896871472706, "grad_norm": 2.3226258754730225, "learning_rate": 2.8944262618276307e-05, "loss": 2.8904, "step": 5249 }, { "epoch": 0.047217962612367506, "grad_norm": 1.620275616645813, "learning_rate": 2.8936840797234244e-05, "loss": 2.7269, "step": 5250 }, { "epoch": 0.04722695651000796, "grad_norm": 2.4259755611419678, "learning_rate": 2.8929418620384753e-05, "loss": 2.7952, "step": 5251 }, { "epoch": 0.047235950407648414, "grad_norm": 2.2827463150024414, "learning_rate": 2.892199608839864e-05, "loss": 2.6698, "step": 5252 }, { "epoch": 0.04724494430528886, "grad_norm": 2.194251298904419, "learning_rate": 2.8914573201946748e-05, "loss": 2.7341, "step": 5253 }, { "epoch": 0.047253938202929315, "grad_norm": 2.0551271438598633, "learning_rate": 2.890714996169994e-05, "loss": 2.7242, "step": 5254 }, { "epoch": 0.04726293210056976, "grad_norm": 2.0763614177703857, "learning_rate": 2.8899726368329128e-05, "loss": 2.5467, "step": 5255 }, { "epoch": 0.047271925998210215, "grad_norm": 1.40879487991333, "learning_rate": 2.8892302422505245e-05, "loss": 2.5517, "step": 5256 }, { "epoch": 0.04728091989585066, "grad_norm": 1.4810956716537476, "learning_rate": 2.8884878124899262e-05, "loss": 2.4504, "step": 5257 }, { "epoch": 0.047289913793491116, "grad_norm": 1.8421162366867065, "learning_rate": 2.8877453476182188e-05, "loss": 2.6144, "step": 5258 }, { "epoch": 0.04729890769113157, "grad_norm": 1.4061473608016968, "learning_rate": 2.8870028477025042e-05, "loss": 2.3344, "step": 5259 }, { "epoch": 0.04730790158877202, "grad_norm": 1.9830044507980347, "learning_rate": 2.8862603128098882e-05, "loss": 2.6408, "step": 5260 }, { "epoch": 0.04731689548641247, "grad_norm": 1.7336560487747192, "learning_rate": 2.8855177430074814e-05, "loss": 2.5493, "step": 5261 }, { "epoch": 0.04732588938405292, "grad_norm": 1.4854834079742432, "learning_rate": 2.8847751383623967e-05, "loss": 2.2552, "step": 5262 }, { "epoch": 0.04733488328169337, "grad_norm": 1.7303838729858398, "learning_rate": 2.884032498941749e-05, "loss": 2.3878, "step": 5263 }, { "epoch": 0.047343877179333825, "grad_norm": 1.8792762756347656, "learning_rate": 2.8832898248126576e-05, "loss": 2.7041, "step": 5264 }, { "epoch": 0.04735287107697427, "grad_norm": 1.7298250198364258, "learning_rate": 2.882547116042245e-05, "loss": 2.6376, "step": 5265 }, { "epoch": 0.047361864974614726, "grad_norm": 2.1919264793395996, "learning_rate": 2.881804372697637e-05, "loss": 2.5059, "step": 5266 }, { "epoch": 0.04737085887225517, "grad_norm": 2.0822348594665527, "learning_rate": 2.8810615948459596e-05, "loss": 2.4894, "step": 5267 }, { "epoch": 0.047379852769895626, "grad_norm": 2.8173961639404297, "learning_rate": 2.8803187825543464e-05, "loss": 2.4796, "step": 5268 }, { "epoch": 0.04738884666753607, "grad_norm": 2.1377177238464355, "learning_rate": 2.879575935889931e-05, "loss": 2.6413, "step": 5269 }, { "epoch": 0.04739784056517653, "grad_norm": 1.363299012184143, "learning_rate": 2.878833054919851e-05, "loss": 2.5599, "step": 5270 }, { "epoch": 0.04740683446281698, "grad_norm": 1.9942094087600708, "learning_rate": 2.8780901397112485e-05, "loss": 2.7767, "step": 5271 }, { "epoch": 0.04741582836045743, "grad_norm": 0.7499995827674866, "learning_rate": 2.8773471903312654e-05, "loss": 2.1711, "step": 5272 }, { "epoch": 0.04742482225809788, "grad_norm": 1.7738511562347412, "learning_rate": 2.876604206847051e-05, "loss": 2.5403, "step": 5273 }, { "epoch": 0.04743381615573833, "grad_norm": 1.567488670349121, "learning_rate": 2.875861189325753e-05, "loss": 2.5406, "step": 5274 }, { "epoch": 0.04744281005337878, "grad_norm": 1.9065743684768677, "learning_rate": 2.8751181378345255e-05, "loss": 2.7363, "step": 5275 }, { "epoch": 0.047451803951019236, "grad_norm": 1.6800588369369507, "learning_rate": 2.8743750524405254e-05, "loss": 2.4789, "step": 5276 }, { "epoch": 0.04746079784865968, "grad_norm": 2.2995874881744385, "learning_rate": 2.8736319332109106e-05, "loss": 2.6033, "step": 5277 }, { "epoch": 0.04746979174630014, "grad_norm": 1.4265490770339966, "learning_rate": 2.8728887802128447e-05, "loss": 2.4776, "step": 5278 }, { "epoch": 0.047478785643940584, "grad_norm": 2.0243916511535645, "learning_rate": 2.872145593513492e-05, "loss": 2.3664, "step": 5279 }, { "epoch": 0.04748777954158104, "grad_norm": 2.419698715209961, "learning_rate": 2.8714023731800223e-05, "loss": 2.7118, "step": 5280 }, { "epoch": 0.04749677343922149, "grad_norm": 1.7180263996124268, "learning_rate": 2.870659119279605e-05, "loss": 2.7285, "step": 5281 }, { "epoch": 0.04750576733686194, "grad_norm": 1.818234920501709, "learning_rate": 2.8699158318794168e-05, "loss": 2.675, "step": 5282 }, { "epoch": 0.04751476123450239, "grad_norm": 1.863985300064087, "learning_rate": 2.8691725110466334e-05, "loss": 2.6524, "step": 5283 }, { "epoch": 0.04752375513214284, "grad_norm": 1.6576730012893677, "learning_rate": 2.868429156848436e-05, "loss": 2.5863, "step": 5284 }, { "epoch": 0.04753274902978329, "grad_norm": 1.546093463897705, "learning_rate": 2.8676857693520086e-05, "loss": 2.3584, "step": 5285 }, { "epoch": 0.04754174292742374, "grad_norm": 1.7879523038864136, "learning_rate": 2.866942348624538e-05, "loss": 2.3671, "step": 5286 }, { "epoch": 0.047550736825064194, "grad_norm": 1.4459052085876465, "learning_rate": 2.866198894733212e-05, "loss": 2.5162, "step": 5287 }, { "epoch": 0.04755973072270465, "grad_norm": 2.052431344985962, "learning_rate": 2.865455407745225e-05, "loss": 2.8307, "step": 5288 }, { "epoch": 0.047568724620345094, "grad_norm": 1.518136739730835, "learning_rate": 2.8647118877277717e-05, "loss": 2.5537, "step": 5289 }, { "epoch": 0.04757771851798555, "grad_norm": 2.058666706085205, "learning_rate": 2.8639683347480508e-05, "loss": 2.8607, "step": 5290 }, { "epoch": 0.047586712415625995, "grad_norm": 1.613295316696167, "learning_rate": 2.863224748873264e-05, "loss": 2.7717, "step": 5291 }, { "epoch": 0.04759570631326645, "grad_norm": 1.958187222480774, "learning_rate": 2.862481130170615e-05, "loss": 2.6947, "step": 5292 }, { "epoch": 0.0476047002109069, "grad_norm": 1.8828284740447998, "learning_rate": 2.8617374787073125e-05, "loss": 2.4513, "step": 5293 }, { "epoch": 0.04761369410854735, "grad_norm": 1.576209545135498, "learning_rate": 2.860993794550567e-05, "loss": 2.5765, "step": 5294 }, { "epoch": 0.047622688006187804, "grad_norm": 2.156716823577881, "learning_rate": 2.86025007776759e-05, "loss": 2.7356, "step": 5295 }, { "epoch": 0.04763168190382825, "grad_norm": 2.0356082916259766, "learning_rate": 2.8595063284255996e-05, "loss": 2.5891, "step": 5296 }, { "epoch": 0.047640675801468704, "grad_norm": 1.4186902046203613, "learning_rate": 2.858762546591814e-05, "loss": 2.3932, "step": 5297 }, { "epoch": 0.04764966969910915, "grad_norm": 1.5686438083648682, "learning_rate": 2.8580187323334555e-05, "loss": 2.5809, "step": 5298 }, { "epoch": 0.047658663596749605, "grad_norm": 1.2073408365249634, "learning_rate": 2.85727488571775e-05, "loss": 2.3264, "step": 5299 }, { "epoch": 0.04766765749439006, "grad_norm": 1.7875230312347412, "learning_rate": 2.856531006811926e-05, "loss": 2.6475, "step": 5300 }, { "epoch": 0.047676651392030506, "grad_norm": 1.226456642150879, "learning_rate": 2.8557870956832132e-05, "loss": 2.4492, "step": 5301 }, { "epoch": 0.04768564528967096, "grad_norm": 2.373202323913574, "learning_rate": 2.8550431523988457e-05, "loss": 2.4635, "step": 5302 }, { "epoch": 0.047694639187311406, "grad_norm": 1.819726824760437, "learning_rate": 2.8542991770260608e-05, "loss": 2.5686, "step": 5303 }, { "epoch": 0.04770363308495186, "grad_norm": 0.92728191614151, "learning_rate": 2.8535551696320972e-05, "loss": 2.2686, "step": 5304 }, { "epoch": 0.047712626982592314, "grad_norm": 1.2622363567352295, "learning_rate": 2.8528111302841986e-05, "loss": 2.5616, "step": 5305 }, { "epoch": 0.04772162088023276, "grad_norm": 1.5455939769744873, "learning_rate": 2.85206705904961e-05, "loss": 2.5349, "step": 5306 }, { "epoch": 0.047730614777873215, "grad_norm": 1.6113271713256836, "learning_rate": 2.85132295599558e-05, "loss": 2.7793, "step": 5307 }, { "epoch": 0.04773960867551366, "grad_norm": 1.5609160661697388, "learning_rate": 2.8505788211893597e-05, "loss": 2.5168, "step": 5308 }, { "epoch": 0.047748602573154116, "grad_norm": 1.989532470703125, "learning_rate": 2.849834654698203e-05, "loss": 2.7698, "step": 5309 }, { "epoch": 0.04775759647079456, "grad_norm": 1.7944594621658325, "learning_rate": 2.8490904565893665e-05, "loss": 2.5268, "step": 5310 }, { "epoch": 0.047766590368435016, "grad_norm": 1.1479564905166626, "learning_rate": 2.8483462269301114e-05, "loss": 2.5025, "step": 5311 }, { "epoch": 0.04777558426607547, "grad_norm": 1.4925159215927124, "learning_rate": 2.8476019657876997e-05, "loss": 2.687, "step": 5312 }, { "epoch": 0.04778457816371592, "grad_norm": 1.8640931844711304, "learning_rate": 2.8468576732293957e-05, "loss": 2.495, "step": 5313 }, { "epoch": 0.04779357206135637, "grad_norm": 1.891969919204712, "learning_rate": 2.846113349322469e-05, "loss": 2.5606, "step": 5314 }, { "epoch": 0.04780256595899682, "grad_norm": 1.2041245698928833, "learning_rate": 2.8453689941341916e-05, "loss": 2.4673, "step": 5315 }, { "epoch": 0.04781155985663727, "grad_norm": 1.781201720237732, "learning_rate": 2.8446246077318355e-05, "loss": 2.9016, "step": 5316 }, { "epoch": 0.047820553754277725, "grad_norm": 1.6694881916046143, "learning_rate": 2.843880190182679e-05, "loss": 2.4861, "step": 5317 }, { "epoch": 0.04782954765191817, "grad_norm": 1.1396961212158203, "learning_rate": 2.843135741554001e-05, "loss": 2.1838, "step": 5318 }, { "epoch": 0.047838541549558626, "grad_norm": 1.6247366666793823, "learning_rate": 2.8423912619130842e-05, "loss": 2.6573, "step": 5319 }, { "epoch": 0.04784753544719907, "grad_norm": 1.180901288986206, "learning_rate": 2.8416467513272145e-05, "loss": 2.3446, "step": 5320 }, { "epoch": 0.04785652934483953, "grad_norm": 1.625337839126587, "learning_rate": 2.8409022098636796e-05, "loss": 2.6282, "step": 5321 }, { "epoch": 0.04786552324247998, "grad_norm": 1.8294172286987305, "learning_rate": 2.840157637589769e-05, "loss": 2.5745, "step": 5322 }, { "epoch": 0.04787451714012043, "grad_norm": 1.7276185750961304, "learning_rate": 2.8394130345727782e-05, "loss": 2.5047, "step": 5323 }, { "epoch": 0.04788351103776088, "grad_norm": 1.9368029832839966, "learning_rate": 2.8386684008800035e-05, "loss": 2.3889, "step": 5324 }, { "epoch": 0.04789250493540133, "grad_norm": 1.251679539680481, "learning_rate": 2.8379237365787426e-05, "loss": 2.4265, "step": 5325 }, { "epoch": 0.04790149883304178, "grad_norm": 1.7430205345153809, "learning_rate": 2.8371790417362987e-05, "loss": 2.6542, "step": 5326 }, { "epoch": 0.04791049273068223, "grad_norm": 1.8688266277313232, "learning_rate": 2.836434316419976e-05, "loss": 2.7402, "step": 5327 }, { "epoch": 0.04791948662832268, "grad_norm": 1.4358022212982178, "learning_rate": 2.8356895606970822e-05, "loss": 2.4244, "step": 5328 }, { "epoch": 0.04792848052596314, "grad_norm": 1.9426873922348022, "learning_rate": 2.8349447746349277e-05, "loss": 2.7376, "step": 5329 }, { "epoch": 0.047937474423603584, "grad_norm": 1.4784505367279053, "learning_rate": 2.8341999583008256e-05, "loss": 2.654, "step": 5330 }, { "epoch": 0.04794646832124404, "grad_norm": 2.09908390045166, "learning_rate": 2.8334551117620907e-05, "loss": 2.8642, "step": 5331 }, { "epoch": 0.047955462218884484, "grad_norm": 1.655156135559082, "learning_rate": 2.8327102350860417e-05, "loss": 2.4376, "step": 5332 }, { "epoch": 0.04796445611652494, "grad_norm": 1.8526558876037598, "learning_rate": 2.83196532834e-05, "loss": 2.6347, "step": 5333 }, { "epoch": 0.04797345001416539, "grad_norm": 1.7711069583892822, "learning_rate": 2.83122039159129e-05, "loss": 2.5217, "step": 5334 }, { "epoch": 0.04798244391180584, "grad_norm": 1.5121440887451172, "learning_rate": 2.830475424907238e-05, "loss": 2.681, "step": 5335 }, { "epoch": 0.04799143780944629, "grad_norm": 1.3594595193862915, "learning_rate": 2.8297304283551728e-05, "loss": 2.7003, "step": 5336 }, { "epoch": 0.04800043170708674, "grad_norm": 1.9257395267486572, "learning_rate": 2.8289854020024265e-05, "loss": 2.6975, "step": 5337 }, { "epoch": 0.04800942560472719, "grad_norm": 1.927554965019226, "learning_rate": 2.8282403459163337e-05, "loss": 2.6486, "step": 5338 }, { "epoch": 0.04801841950236764, "grad_norm": 1.3678126335144043, "learning_rate": 2.8274952601642324e-05, "loss": 2.53, "step": 5339 }, { "epoch": 0.048027413400008094, "grad_norm": 1.7239545583724976, "learning_rate": 2.8267501448134615e-05, "loss": 2.6874, "step": 5340 }, { "epoch": 0.04803640729764855, "grad_norm": 2.243119478225708, "learning_rate": 2.8260049999313646e-05, "loss": 2.8175, "step": 5341 }, { "epoch": 0.048045401195288995, "grad_norm": 1.489894986152649, "learning_rate": 2.825259825585288e-05, "loss": 2.4272, "step": 5342 }, { "epoch": 0.04805439509292945, "grad_norm": 2.6559815406799316, "learning_rate": 2.8245146218425783e-05, "loss": 2.8273, "step": 5343 }, { "epoch": 0.048063388990569895, "grad_norm": 1.5977489948272705, "learning_rate": 2.8237693887705863e-05, "loss": 2.5254, "step": 5344 }, { "epoch": 0.04807238288821035, "grad_norm": 1.868375539779663, "learning_rate": 2.823024126436666e-05, "loss": 2.3552, "step": 5345 }, { "epoch": 0.0480813767858508, "grad_norm": 1.6592895984649658, "learning_rate": 2.8222788349081726e-05, "loss": 2.5227, "step": 5346 }, { "epoch": 0.04809037068349125, "grad_norm": 2.590772867202759, "learning_rate": 2.8215335142524657e-05, "loss": 2.5193, "step": 5347 }, { "epoch": 0.048099364581131704, "grad_norm": 2.1459147930145264, "learning_rate": 2.8207881645369056e-05, "loss": 2.965, "step": 5348 }, { "epoch": 0.04810835847877215, "grad_norm": 1.895775556564331, "learning_rate": 2.8200427858288574e-05, "loss": 2.4973, "step": 5349 }, { "epoch": 0.048117352376412605, "grad_norm": 2.0931107997894287, "learning_rate": 2.8192973781956865e-05, "loss": 2.6843, "step": 5350 }, { "epoch": 0.04812634627405306, "grad_norm": 1.8844516277313232, "learning_rate": 2.8185519417047624e-05, "loss": 2.6063, "step": 5351 }, { "epoch": 0.048135340171693505, "grad_norm": 1.7844175100326538, "learning_rate": 2.8178064764234567e-05, "loss": 2.5581, "step": 5352 }, { "epoch": 0.04814433406933396, "grad_norm": 1.9636223316192627, "learning_rate": 2.8170609824191446e-05, "loss": 2.4498, "step": 5353 }, { "epoch": 0.048153327966974406, "grad_norm": 1.2392829656600952, "learning_rate": 2.8163154597592023e-05, "loss": 2.3, "step": 5354 }, { "epoch": 0.04816232186461486, "grad_norm": 1.7848246097564697, "learning_rate": 2.8155699085110087e-05, "loss": 2.5693, "step": 5355 }, { "epoch": 0.04817131576225531, "grad_norm": 1.8972123861312866, "learning_rate": 2.814824328741948e-05, "loss": 2.8433, "step": 5356 }, { "epoch": 0.04818030965989576, "grad_norm": 1.809592604637146, "learning_rate": 2.8140787205194024e-05, "loss": 2.7892, "step": 5357 }, { "epoch": 0.048189303557536214, "grad_norm": 1.1615781784057617, "learning_rate": 2.8133330839107608e-05, "loss": 2.335, "step": 5358 }, { "epoch": 0.04819829745517666, "grad_norm": 1.5520434379577637, "learning_rate": 2.8125874189834127e-05, "loss": 2.559, "step": 5359 }, { "epoch": 0.048207291352817115, "grad_norm": 1.1923514604568481, "learning_rate": 2.81184172580475e-05, "loss": 2.4476, "step": 5360 }, { "epoch": 0.04821628525045756, "grad_norm": 1.428335189819336, "learning_rate": 2.811096004442168e-05, "loss": 2.3997, "step": 5361 }, { "epoch": 0.048225279148098016, "grad_norm": 1.6450732946395874, "learning_rate": 2.8103502549630645e-05, "loss": 2.693, "step": 5362 }, { "epoch": 0.04823427304573847, "grad_norm": 1.846496820449829, "learning_rate": 2.8096044774348397e-05, "loss": 2.7573, "step": 5363 }, { "epoch": 0.04824326694337892, "grad_norm": 1.4310005903244019, "learning_rate": 2.808858671924895e-05, "loss": 2.5487, "step": 5364 }, { "epoch": 0.04825226084101937, "grad_norm": 0.8058369755744934, "learning_rate": 2.8081128385006366e-05, "loss": 2.3366, "step": 5365 }, { "epoch": 0.04826125473865982, "grad_norm": 1.7896026372909546, "learning_rate": 2.8073669772294714e-05, "loss": 2.358, "step": 5366 }, { "epoch": 0.04827024863630027, "grad_norm": 1.3942644596099854, "learning_rate": 2.8066210881788098e-05, "loss": 2.4211, "step": 5367 }, { "epoch": 0.04827924253394072, "grad_norm": 1.5411781072616577, "learning_rate": 2.8058751714160646e-05, "loss": 2.7435, "step": 5368 }, { "epoch": 0.04828823643158117, "grad_norm": 1.893233060836792, "learning_rate": 2.8051292270086503e-05, "loss": 2.9442, "step": 5369 }, { "epoch": 0.048297230329221626, "grad_norm": 1.8343504667282104, "learning_rate": 2.804383255023986e-05, "loss": 2.872, "step": 5370 }, { "epoch": 0.04830622422686207, "grad_norm": 1.975637435913086, "learning_rate": 2.8036372555294917e-05, "loss": 2.8423, "step": 5371 }, { "epoch": 0.048315218124502526, "grad_norm": 1.5206884145736694, "learning_rate": 2.802891228592588e-05, "loss": 2.2836, "step": 5372 }, { "epoch": 0.04832421202214297, "grad_norm": 1.817371129989624, "learning_rate": 2.8021451742807016e-05, "loss": 2.7486, "step": 5373 }, { "epoch": 0.04833320591978343, "grad_norm": 1.1582752466201782, "learning_rate": 2.80139909266126e-05, "loss": 2.5512, "step": 5374 }, { "epoch": 0.04834219981742388, "grad_norm": 0.9388465285301208, "learning_rate": 2.8006529838016925e-05, "loss": 2.2188, "step": 5375 }, { "epoch": 0.04835119371506433, "grad_norm": 1.8211524486541748, "learning_rate": 2.799906847769433e-05, "loss": 2.487, "step": 5376 }, { "epoch": 0.04836018761270478, "grad_norm": 1.8007928133010864, "learning_rate": 2.7991606846319147e-05, "loss": 2.8064, "step": 5377 }, { "epoch": 0.04836918151034523, "grad_norm": 1.602394461631775, "learning_rate": 2.798414494456577e-05, "loss": 2.5672, "step": 5378 }, { "epoch": 0.04837817540798568, "grad_norm": 1.844016194343567, "learning_rate": 2.7976682773108577e-05, "loss": 2.8387, "step": 5379 }, { "epoch": 0.04838716930562613, "grad_norm": 1.5409318208694458, "learning_rate": 2.7969220332622003e-05, "loss": 2.4544, "step": 5380 }, { "epoch": 0.04839616320326658, "grad_norm": 1.8291146755218506, "learning_rate": 2.79617576237805e-05, "loss": 2.6878, "step": 5381 }, { "epoch": 0.04840515710090704, "grad_norm": 1.2410657405853271, "learning_rate": 2.7954294647258524e-05, "loss": 2.2298, "step": 5382 }, { "epoch": 0.048414150998547484, "grad_norm": 2.0232298374176025, "learning_rate": 2.794683140373059e-05, "loss": 2.7659, "step": 5383 }, { "epoch": 0.04842314489618794, "grad_norm": 0.986171305179596, "learning_rate": 2.7939367893871205e-05, "loss": 2.3288, "step": 5384 }, { "epoch": 0.048432138793828385, "grad_norm": 2.2204384803771973, "learning_rate": 2.793190411835491e-05, "loss": 2.5131, "step": 5385 }, { "epoch": 0.04844113269146884, "grad_norm": 2.1671559810638428, "learning_rate": 2.7924440077856283e-05, "loss": 2.7023, "step": 5386 }, { "epoch": 0.04845012658910929, "grad_norm": 2.247406482696533, "learning_rate": 2.7916975773049907e-05, "loss": 2.8568, "step": 5387 }, { "epoch": 0.04845912048674974, "grad_norm": 1.615600824356079, "learning_rate": 2.7909511204610405e-05, "loss": 2.6602, "step": 5388 }, { "epoch": 0.04846811438439019, "grad_norm": 1.3901774883270264, "learning_rate": 2.790204637321242e-05, "loss": 2.494, "step": 5389 }, { "epoch": 0.04847710828203064, "grad_norm": 1.191828727722168, "learning_rate": 2.7894581279530607e-05, "loss": 2.4156, "step": 5390 }, { "epoch": 0.048486102179671094, "grad_norm": 1.283498764038086, "learning_rate": 2.788711592423966e-05, "loss": 2.3748, "step": 5391 }, { "epoch": 0.04849509607731155, "grad_norm": 2.2340354919433594, "learning_rate": 2.7879650308014278e-05, "loss": 2.5551, "step": 5392 }, { "epoch": 0.048504089974951994, "grad_norm": 1.740108847618103, "learning_rate": 2.7872184431529202e-05, "loss": 2.4473, "step": 5393 }, { "epoch": 0.04851308387259245, "grad_norm": 2.474130392074585, "learning_rate": 2.786471829545919e-05, "loss": 2.6165, "step": 5394 }, { "epoch": 0.048522077770232895, "grad_norm": 1.444397211074829, "learning_rate": 2.785725190047903e-05, "loss": 2.3544, "step": 5395 }, { "epoch": 0.04853107166787335, "grad_norm": 1.1834009885787964, "learning_rate": 2.7849785247263515e-05, "loss": 2.4937, "step": 5396 }, { "epoch": 0.048540065565513796, "grad_norm": 1.6160849332809448, "learning_rate": 2.7842318336487483e-05, "loss": 2.4148, "step": 5397 }, { "epoch": 0.04854905946315425, "grad_norm": 1.2364749908447266, "learning_rate": 2.7834851168825782e-05, "loss": 2.3211, "step": 5398 }, { "epoch": 0.0485580533607947, "grad_norm": 1.4156618118286133, "learning_rate": 2.7827383744953284e-05, "loss": 2.417, "step": 5399 }, { "epoch": 0.04856704725843515, "grad_norm": 1.1593453884124756, "learning_rate": 2.7819916065544888e-05, "loss": 2.445, "step": 5400 }, { "epoch": 0.048576041156075604, "grad_norm": 1.9699820280075073, "learning_rate": 2.781244813127552e-05, "loss": 2.7259, "step": 5401 }, { "epoch": 0.04858503505371605, "grad_norm": 1.5834218263626099, "learning_rate": 2.7804979942820113e-05, "loss": 2.3876, "step": 5402 }, { "epoch": 0.048594028951356505, "grad_norm": 1.6256732940673828, "learning_rate": 2.779751150085364e-05, "loss": 2.7628, "step": 5403 }, { "epoch": 0.04860302284899696, "grad_norm": 1.8367480039596558, "learning_rate": 2.7790042806051093e-05, "loss": 2.7208, "step": 5404 }, { "epoch": 0.048612016746637406, "grad_norm": 1.2901114225387573, "learning_rate": 2.7782573859087487e-05, "loss": 2.3461, "step": 5405 }, { "epoch": 0.04862101064427786, "grad_norm": 1.1269105672836304, "learning_rate": 2.7775104660637847e-05, "loss": 2.4267, "step": 5406 }, { "epoch": 0.048630004541918306, "grad_norm": 1.7177660465240479, "learning_rate": 2.7767635211377242e-05, "loss": 2.652, "step": 5407 }, { "epoch": 0.04863899843955876, "grad_norm": 1.5640833377838135, "learning_rate": 2.776016551198074e-05, "loss": 2.4455, "step": 5408 }, { "epoch": 0.04864799233719921, "grad_norm": 2.192736864089966, "learning_rate": 2.7752695563123454e-05, "loss": 2.5947, "step": 5409 }, { "epoch": 0.04865698623483966, "grad_norm": 1.988682746887207, "learning_rate": 2.774522536548051e-05, "loss": 2.7446, "step": 5410 }, { "epoch": 0.048665980132480115, "grad_norm": 1.4174946546554565, "learning_rate": 2.7737754919727054e-05, "loss": 2.5696, "step": 5411 }, { "epoch": 0.04867497403012056, "grad_norm": 1.6843223571777344, "learning_rate": 2.773028422653826e-05, "loss": 2.6734, "step": 5412 }, { "epoch": 0.048683967927761015, "grad_norm": 1.529298186302185, "learning_rate": 2.7722813286589316e-05, "loss": 2.5997, "step": 5413 }, { "epoch": 0.04869296182540146, "grad_norm": 1.4150757789611816, "learning_rate": 2.7715342100555442e-05, "loss": 2.5178, "step": 5414 }, { "epoch": 0.048701955723041916, "grad_norm": 1.0565379858016968, "learning_rate": 2.7707870669111868e-05, "loss": 2.2932, "step": 5415 }, { "epoch": 0.04871094962068237, "grad_norm": 1.9287523031234741, "learning_rate": 2.7700398992933862e-05, "loss": 2.5552, "step": 5416 }, { "epoch": 0.04871994351832282, "grad_norm": 1.7044265270233154, "learning_rate": 2.7692927072696702e-05, "loss": 2.5362, "step": 5417 }, { "epoch": 0.04872893741596327, "grad_norm": 1.8625701665878296, "learning_rate": 2.7685454909075704e-05, "loss": 2.7191, "step": 5418 }, { "epoch": 0.04873793131360372, "grad_norm": 1.9767954349517822, "learning_rate": 2.7677982502746176e-05, "loss": 2.7019, "step": 5419 }, { "epoch": 0.04874692521124417, "grad_norm": 1.0995333194732666, "learning_rate": 2.7670509854383476e-05, "loss": 2.3309, "step": 5420 }, { "epoch": 0.04875591910888462, "grad_norm": 1.460827350616455, "learning_rate": 2.7663036964662965e-05, "loss": 2.6219, "step": 5421 }, { "epoch": 0.04876491300652507, "grad_norm": 1.6824133396148682, "learning_rate": 2.7655563834260045e-05, "loss": 2.7069, "step": 5422 }, { "epoch": 0.048773906904165526, "grad_norm": 1.6983197927474976, "learning_rate": 2.7648090463850122e-05, "loss": 2.5189, "step": 5423 }, { "epoch": 0.04878290080180597, "grad_norm": 1.7577269077301025, "learning_rate": 2.764061685410865e-05, "loss": 2.6506, "step": 5424 }, { "epoch": 0.04879189469944643, "grad_norm": 1.7396293878555298, "learning_rate": 2.763314300571106e-05, "loss": 2.543, "step": 5425 }, { "epoch": 0.048800888597086874, "grad_norm": 1.7171064615249634, "learning_rate": 2.762566891933285e-05, "loss": 2.7442, "step": 5426 }, { "epoch": 0.04880988249472733, "grad_norm": 1.6552278995513916, "learning_rate": 2.76181945956495e-05, "loss": 2.5864, "step": 5427 }, { "epoch": 0.04881887639236778, "grad_norm": 1.2829135656356812, "learning_rate": 2.761072003533654e-05, "loss": 2.2965, "step": 5428 }, { "epoch": 0.04882787029000823, "grad_norm": 1.727217435836792, "learning_rate": 2.7603245239069527e-05, "loss": 2.5937, "step": 5429 }, { "epoch": 0.04883686418764868, "grad_norm": 1.3382437229156494, "learning_rate": 2.7595770207524006e-05, "loss": 2.4274, "step": 5430 }, { "epoch": 0.04884585808528913, "grad_norm": 1.5735969543457031, "learning_rate": 2.758829494137557e-05, "loss": 2.9328, "step": 5431 }, { "epoch": 0.04885485198292958, "grad_norm": 2.004376173019409, "learning_rate": 2.7580819441299826e-05, "loss": 2.421, "step": 5432 }, { "epoch": 0.048863845880570037, "grad_norm": 1.5710150003433228, "learning_rate": 2.7573343707972404e-05, "loss": 2.4159, "step": 5433 }, { "epoch": 0.04887283977821048, "grad_norm": 2.141671895980835, "learning_rate": 2.7565867742068946e-05, "loss": 2.7711, "step": 5434 }, { "epoch": 0.04888183367585094, "grad_norm": 1.3417831659317017, "learning_rate": 2.755839154426513e-05, "loss": 2.4856, "step": 5435 }, { "epoch": 0.048890827573491384, "grad_norm": 1.7312544584274292, "learning_rate": 2.7550915115236635e-05, "loss": 2.8582, "step": 5436 }, { "epoch": 0.04889982147113184, "grad_norm": 1.481497883796692, "learning_rate": 2.754343845565918e-05, "loss": 2.3582, "step": 5437 }, { "epoch": 0.048908815368772285, "grad_norm": 2.2603156566619873, "learning_rate": 2.75359615662085e-05, "loss": 2.9566, "step": 5438 }, { "epoch": 0.04891780926641274, "grad_norm": 1.5680174827575684, "learning_rate": 2.752848444756035e-05, "loss": 2.4699, "step": 5439 }, { "epoch": 0.04892680316405319, "grad_norm": 1.8056877851486206, "learning_rate": 2.7521007100390496e-05, "loss": 2.6511, "step": 5440 }, { "epoch": 0.04893579706169364, "grad_norm": 1.3601665496826172, "learning_rate": 2.7513529525374726e-05, "loss": 2.4907, "step": 5441 }, { "epoch": 0.04894479095933409, "grad_norm": 2.1910386085510254, "learning_rate": 2.750605172318888e-05, "loss": 2.6564, "step": 5442 }, { "epoch": 0.04895378485697454, "grad_norm": 1.2967439889907837, "learning_rate": 2.7498573694508765e-05, "loss": 2.4424, "step": 5443 }, { "epoch": 0.048962778754614994, "grad_norm": 1.3125334978103638, "learning_rate": 2.7491095440010256e-05, "loss": 2.536, "step": 5444 }, { "epoch": 0.04897177265225545, "grad_norm": 1.8147680759429932, "learning_rate": 2.7483616960369224e-05, "loss": 2.6212, "step": 5445 }, { "epoch": 0.048980766549895895, "grad_norm": 1.3005270957946777, "learning_rate": 2.7476138256261575e-05, "loss": 2.4248, "step": 5446 }, { "epoch": 0.04898976044753635, "grad_norm": 1.530658483505249, "learning_rate": 2.7468659328363206e-05, "loss": 2.5309, "step": 5447 }, { "epoch": 0.048998754345176795, "grad_norm": 2.2262630462646484, "learning_rate": 2.746118017735007e-05, "loss": 2.3441, "step": 5448 }, { "epoch": 0.04900774824281725, "grad_norm": 1.8445262908935547, "learning_rate": 2.745370080389812e-05, "loss": 2.6602, "step": 5449 }, { "epoch": 0.049016742140457696, "grad_norm": 1.9671952724456787, "learning_rate": 2.744622120868333e-05, "loss": 2.6317, "step": 5450 }, { "epoch": 0.04902573603809815, "grad_norm": 1.6081132888793945, "learning_rate": 2.7438741392381705e-05, "loss": 2.5662, "step": 5451 }, { "epoch": 0.049034729935738604, "grad_norm": 1.9018937349319458, "learning_rate": 2.7431261355669263e-05, "loss": 2.4032, "step": 5452 }, { "epoch": 0.04904372383337905, "grad_norm": 3.2856714725494385, "learning_rate": 2.742378109922204e-05, "loss": 2.5393, "step": 5453 }, { "epoch": 0.049052717731019504, "grad_norm": 1.9488612413406372, "learning_rate": 2.741630062371609e-05, "loss": 2.9758, "step": 5454 }, { "epoch": 0.04906171162865995, "grad_norm": 1.7597286701202393, "learning_rate": 2.740881992982749e-05, "loss": 2.5591, "step": 5455 }, { "epoch": 0.049070705526300405, "grad_norm": 1.5036661624908447, "learning_rate": 2.740133901823234e-05, "loss": 2.5878, "step": 5456 }, { "epoch": 0.04907969942394086, "grad_norm": 1.724717140197754, "learning_rate": 2.7393857889606756e-05, "loss": 2.3302, "step": 5457 }, { "epoch": 0.049088693321581306, "grad_norm": 2.0018980503082275, "learning_rate": 2.7386376544626868e-05, "loss": 2.6995, "step": 5458 }, { "epoch": 0.04909768721922176, "grad_norm": 1.4452589750289917, "learning_rate": 2.7378894983968844e-05, "loss": 2.464, "step": 5459 }, { "epoch": 0.04910668111686221, "grad_norm": 2.1344518661499023, "learning_rate": 2.737141320830886e-05, "loss": 2.7724, "step": 5460 }, { "epoch": 0.04911567501450266, "grad_norm": 1.5835623741149902, "learning_rate": 2.7363931218323103e-05, "loss": 2.5239, "step": 5461 }, { "epoch": 0.049124668912143114, "grad_norm": 1.4448487758636475, "learning_rate": 2.7356449014687784e-05, "loss": 2.601, "step": 5462 }, { "epoch": 0.04913366280978356, "grad_norm": 1.6544418334960938, "learning_rate": 2.734896659807914e-05, "loss": 2.5649, "step": 5463 }, { "epoch": 0.049142656707424015, "grad_norm": 1.9056940078735352, "learning_rate": 2.7341483969173427e-05, "loss": 2.5872, "step": 5464 }, { "epoch": 0.04915165060506446, "grad_norm": 1.722733974456787, "learning_rate": 2.7334001128646912e-05, "loss": 2.4816, "step": 5465 }, { "epoch": 0.049160644502704916, "grad_norm": 1.6705702543258667, "learning_rate": 2.7326518077175895e-05, "loss": 2.7282, "step": 5466 }, { "epoch": 0.04916963840034536, "grad_norm": 1.6898210048675537, "learning_rate": 2.7319034815436674e-05, "loss": 2.55, "step": 5467 }, { "epoch": 0.049178632297985816, "grad_norm": 1.6964843273162842, "learning_rate": 2.731155134410559e-05, "loss": 2.7832, "step": 5468 }, { "epoch": 0.04918762619562627, "grad_norm": 2.2571399211883545, "learning_rate": 2.7304067663858984e-05, "loss": 2.7541, "step": 5469 }, { "epoch": 0.04919662009326672, "grad_norm": 1.6744041442871094, "learning_rate": 2.729658377537322e-05, "loss": 2.5035, "step": 5470 }, { "epoch": 0.04920561399090717, "grad_norm": 1.7214312553405762, "learning_rate": 2.7289099679324686e-05, "loss": 2.7987, "step": 5471 }, { "epoch": 0.04921460788854762, "grad_norm": 1.6673256158828735, "learning_rate": 2.7281615376389797e-05, "loss": 2.6719, "step": 5472 }, { "epoch": 0.04922360178618807, "grad_norm": 1.786631464958191, "learning_rate": 2.727413086724496e-05, "loss": 2.8373, "step": 5473 }, { "epoch": 0.049232595683828526, "grad_norm": 2.3701305389404297, "learning_rate": 2.726664615256664e-05, "loss": 2.6063, "step": 5474 }, { "epoch": 0.04924158958146897, "grad_norm": 1.386600136756897, "learning_rate": 2.7259161233031268e-05, "loss": 2.4389, "step": 5475 }, { "epoch": 0.049250583479109426, "grad_norm": 1.3467158079147339, "learning_rate": 2.7251676109315338e-05, "loss": 2.5282, "step": 5476 }, { "epoch": 0.04925957737674987, "grad_norm": 1.498895525932312, "learning_rate": 2.7244190782095348e-05, "loss": 2.283, "step": 5477 }, { "epoch": 0.04926857127439033, "grad_norm": 1.5738468170166016, "learning_rate": 2.7236705252047816e-05, "loss": 2.4669, "step": 5478 }, { "epoch": 0.049277565172030774, "grad_norm": 1.369022011756897, "learning_rate": 2.722921951984927e-05, "loss": 2.6136, "step": 5479 }, { "epoch": 0.04928655906967123, "grad_norm": 1.758817434310913, "learning_rate": 2.7221733586176257e-05, "loss": 2.9457, "step": 5480 }, { "epoch": 0.04929555296731168, "grad_norm": 1.1232898235321045, "learning_rate": 2.721424745170537e-05, "loss": 2.4236, "step": 5481 }, { "epoch": 0.04930454686495213, "grad_norm": 1.5534754991531372, "learning_rate": 2.7206761117113172e-05, "loss": 2.5592, "step": 5482 }, { "epoch": 0.04931354076259258, "grad_norm": 0.5999318361282349, "learning_rate": 2.7199274583076283e-05, "loss": 2.3165, "step": 5483 }, { "epoch": 0.04932253466023303, "grad_norm": 1.4238353967666626, "learning_rate": 2.7191787850271334e-05, "loss": 2.4947, "step": 5484 }, { "epoch": 0.04933152855787348, "grad_norm": 1.2538528442382812, "learning_rate": 2.7184300919374955e-05, "loss": 2.4655, "step": 5485 }, { "epoch": 0.04934052245551394, "grad_norm": 1.7042330503463745, "learning_rate": 2.717681379106381e-05, "loss": 2.5662, "step": 5486 }, { "epoch": 0.049349516353154384, "grad_norm": 1.0492637157440186, "learning_rate": 2.716932646601458e-05, "loss": 2.3645, "step": 5487 }, { "epoch": 0.04935851025079484, "grad_norm": 1.2124112844467163, "learning_rate": 2.716183894490396e-05, "loss": 2.4397, "step": 5488 }, { "epoch": 0.049367504148435284, "grad_norm": 1.3977283239364624, "learning_rate": 2.7154351228408663e-05, "loss": 2.4686, "step": 5489 }, { "epoch": 0.04937649804607574, "grad_norm": 1.900085687637329, "learning_rate": 2.7146863317205427e-05, "loss": 2.5603, "step": 5490 }, { "epoch": 0.049385491943716185, "grad_norm": 1.7887026071548462, "learning_rate": 2.7139375211970996e-05, "loss": 2.7185, "step": 5491 }, { "epoch": 0.04939448584135664, "grad_norm": 1.4319369792938232, "learning_rate": 2.7131886913382138e-05, "loss": 2.4296, "step": 5492 }, { "epoch": 0.04940347973899709, "grad_norm": 1.7374756336212158, "learning_rate": 2.7124398422115632e-05, "loss": 2.8236, "step": 5493 }, { "epoch": 0.04941247363663754, "grad_norm": 1.9848905801773071, "learning_rate": 2.7116909738848286e-05, "loss": 2.971, "step": 5494 }, { "epoch": 0.049421467534277994, "grad_norm": 1.770845890045166, "learning_rate": 2.7109420864256925e-05, "loss": 2.5095, "step": 5495 }, { "epoch": 0.04943046143191844, "grad_norm": 1.9845768213272095, "learning_rate": 2.7101931799018377e-05, "loss": 2.3344, "step": 5496 }, { "epoch": 0.049439455329558894, "grad_norm": 1.4472713470458984, "learning_rate": 2.7094442543809495e-05, "loss": 2.604, "step": 5497 }, { "epoch": 0.04944844922719935, "grad_norm": 1.4648559093475342, "learning_rate": 2.7086953099307144e-05, "loss": 2.5511, "step": 5498 }, { "epoch": 0.049457443124839795, "grad_norm": 1.415381669998169, "learning_rate": 2.7079463466188228e-05, "loss": 2.4568, "step": 5499 }, { "epoch": 0.04946643702248025, "grad_norm": 1.386671543121338, "learning_rate": 2.7071973645129638e-05, "loss": 2.3099, "step": 5500 }, { "epoch": 0.049475430920120696, "grad_norm": 1.5002654790878296, "learning_rate": 2.7064483636808313e-05, "loss": 2.5876, "step": 5501 }, { "epoch": 0.04948442481776115, "grad_norm": 1.941899299621582, "learning_rate": 2.7056993441901167e-05, "loss": 2.6011, "step": 5502 }, { "epoch": 0.0494934187154016, "grad_norm": 1.3325577974319458, "learning_rate": 2.704950306108518e-05, "loss": 2.3554, "step": 5503 }, { "epoch": 0.04950241261304205, "grad_norm": 2.0461130142211914, "learning_rate": 2.7042012495037312e-05, "loss": 2.8666, "step": 5504 }, { "epoch": 0.049511406510682504, "grad_norm": 1.5261331796646118, "learning_rate": 2.7034521744434548e-05, "loss": 2.525, "step": 5505 }, { "epoch": 0.04952040040832295, "grad_norm": 1.6937530040740967, "learning_rate": 2.7027030809953903e-05, "loss": 2.7936, "step": 5506 }, { "epoch": 0.049529394305963405, "grad_norm": 1.6830159425735474, "learning_rate": 2.70195396922724e-05, "loss": 2.6023, "step": 5507 }, { "epoch": 0.04953838820360385, "grad_norm": 1.4869492053985596, "learning_rate": 2.7012048392067075e-05, "loss": 2.361, "step": 5508 }, { "epoch": 0.049547382101244306, "grad_norm": 2.1239917278289795, "learning_rate": 2.7004556910014982e-05, "loss": 2.8761, "step": 5509 }, { "epoch": 0.04955637599888476, "grad_norm": 0.827623188495636, "learning_rate": 2.699706524679319e-05, "loss": 2.2492, "step": 5510 }, { "epoch": 0.049565369896525206, "grad_norm": 1.667885184288025, "learning_rate": 2.6989573403078794e-05, "loss": 2.5366, "step": 5511 }, { "epoch": 0.04957436379416566, "grad_norm": 1.158420443534851, "learning_rate": 2.6982081379548896e-05, "loss": 2.3124, "step": 5512 }, { "epoch": 0.04958335769180611, "grad_norm": 1.717388391494751, "learning_rate": 2.6974589176880626e-05, "loss": 2.5363, "step": 5513 }, { "epoch": 0.04959235158944656, "grad_norm": 2.4027514457702637, "learning_rate": 2.6967096795751106e-05, "loss": 2.6003, "step": 5514 }, { "epoch": 0.049601345487087015, "grad_norm": 2.2710819244384766, "learning_rate": 2.69596042368375e-05, "loss": 2.5823, "step": 5515 }, { "epoch": 0.04961033938472746, "grad_norm": 1.6417595148086548, "learning_rate": 2.6952111500816974e-05, "loss": 2.9502, "step": 5516 }, { "epoch": 0.049619333282367915, "grad_norm": 1.9703943729400635, "learning_rate": 2.6944618588366706e-05, "loss": 2.4131, "step": 5517 }, { "epoch": 0.04962832718000836, "grad_norm": 1.9474563598632812, "learning_rate": 2.6937125500163907e-05, "loss": 2.8109, "step": 5518 }, { "epoch": 0.049637321077648816, "grad_norm": 2.083252191543579, "learning_rate": 2.69296322368858e-05, "loss": 2.8479, "step": 5519 }, { "epoch": 0.04964631497528926, "grad_norm": 2.196695327758789, "learning_rate": 2.69221387992096e-05, "loss": 2.7333, "step": 5520 }, { "epoch": 0.04965530887292972, "grad_norm": 1.5955878496170044, "learning_rate": 2.691464518781257e-05, "loss": 2.556, "step": 5521 }, { "epoch": 0.04966430277057017, "grad_norm": 1.3178226947784424, "learning_rate": 2.6907151403371965e-05, "loss": 2.5497, "step": 5522 }, { "epoch": 0.04967329666821062, "grad_norm": 1.513507604598999, "learning_rate": 2.689965744656508e-05, "loss": 2.4288, "step": 5523 }, { "epoch": 0.04968229056585107, "grad_norm": 1.7322967052459717, "learning_rate": 2.6892163318069198e-05, "loss": 2.7344, "step": 5524 }, { "epoch": 0.04969128446349152, "grad_norm": 1.8035452365875244, "learning_rate": 2.6884669018561638e-05, "loss": 2.697, "step": 5525 }, { "epoch": 0.04970027836113197, "grad_norm": 2.411703586578369, "learning_rate": 2.6877174548719706e-05, "loss": 2.7914, "step": 5526 }, { "epoch": 0.049709272258772426, "grad_norm": 1.5920251607894897, "learning_rate": 2.6869679909220775e-05, "loss": 2.6969, "step": 5527 }, { "epoch": 0.04971826615641287, "grad_norm": 1.9231245517730713, "learning_rate": 2.6862185100742176e-05, "loss": 2.7938, "step": 5528 }, { "epoch": 0.04972726005405333, "grad_norm": 2.063218593597412, "learning_rate": 2.6854690123961308e-05, "loss": 2.5635, "step": 5529 }, { "epoch": 0.049736253951693774, "grad_norm": 2.000607490539551, "learning_rate": 2.6847194979555534e-05, "loss": 2.5253, "step": 5530 }, { "epoch": 0.04974524784933423, "grad_norm": 2.0608558654785156, "learning_rate": 2.6839699668202273e-05, "loss": 2.7952, "step": 5531 }, { "epoch": 0.049754241746974674, "grad_norm": 1.5483404397964478, "learning_rate": 2.6832204190578935e-05, "loss": 2.5938, "step": 5532 }, { "epoch": 0.04976323564461513, "grad_norm": 0.5339071750640869, "learning_rate": 2.6824708547362953e-05, "loss": 2.2079, "step": 5533 }, { "epoch": 0.04977222954225558, "grad_norm": 1.517595648765564, "learning_rate": 2.681721273923178e-05, "loss": 2.4943, "step": 5534 }, { "epoch": 0.04978122343989603, "grad_norm": 1.227372646331787, "learning_rate": 2.680971676686287e-05, "loss": 2.4352, "step": 5535 }, { "epoch": 0.04979021733753648, "grad_norm": 1.3495638370513916, "learning_rate": 2.680222063093372e-05, "loss": 2.5986, "step": 5536 }, { "epoch": 0.04979921123517693, "grad_norm": 1.9719418287277222, "learning_rate": 2.6794724332121814e-05, "loss": 2.5303, "step": 5537 }, { "epoch": 0.04980820513281738, "grad_norm": 1.7108186483383179, "learning_rate": 2.678722787110464e-05, "loss": 2.7629, "step": 5538 }, { "epoch": 0.04981719903045784, "grad_norm": 1.7618340253829956, "learning_rate": 2.6779731248559743e-05, "loss": 2.7019, "step": 5539 }, { "epoch": 0.049826192928098284, "grad_norm": 1.9476468563079834, "learning_rate": 2.6772234465164654e-05, "loss": 2.8312, "step": 5540 }, { "epoch": 0.04983518682573874, "grad_norm": 1.7647939920425415, "learning_rate": 2.6764737521596915e-05, "loss": 2.419, "step": 5541 }, { "epoch": 0.049844180723379185, "grad_norm": 2.5476913452148438, "learning_rate": 2.675724041853411e-05, "loss": 2.5658, "step": 5542 }, { "epoch": 0.04985317462101964, "grad_norm": 1.1266337633132935, "learning_rate": 2.6749743156653807e-05, "loss": 2.3675, "step": 5543 }, { "epoch": 0.04986216851866009, "grad_norm": 1.9909474849700928, "learning_rate": 2.6742245736633603e-05, "loss": 2.4744, "step": 5544 }, { "epoch": 0.04987116241630054, "grad_norm": 1.865558385848999, "learning_rate": 2.6734748159151102e-05, "loss": 2.6592, "step": 5545 }, { "epoch": 0.04988015631394099, "grad_norm": 1.667876124382019, "learning_rate": 2.672725042488393e-05, "loss": 2.6016, "step": 5546 }, { "epoch": 0.04988915021158144, "grad_norm": 1.6822179555892944, "learning_rate": 2.671975253450973e-05, "loss": 2.6796, "step": 5547 }, { "epoch": 0.049898144109221894, "grad_norm": 1.636737585067749, "learning_rate": 2.671225448870614e-05, "loss": 2.5075, "step": 5548 }, { "epoch": 0.04990713800686234, "grad_norm": 1.618526816368103, "learning_rate": 2.6704756288150846e-05, "loss": 2.4613, "step": 5549 }, { "epoch": 0.049916131904502795, "grad_norm": 1.2868871688842773, "learning_rate": 2.6697257933521508e-05, "loss": 2.3316, "step": 5550 }, { "epoch": 0.04992512580214325, "grad_norm": 2.1501612663269043, "learning_rate": 2.668975942549583e-05, "loss": 2.7672, "step": 5551 }, { "epoch": 0.049934119699783695, "grad_norm": 1.1656924486160278, "learning_rate": 2.668226076475151e-05, "loss": 2.4829, "step": 5552 }, { "epoch": 0.04994311359742415, "grad_norm": 1.70994234085083, "learning_rate": 2.667476195196627e-05, "loss": 2.5211, "step": 5553 }, { "epoch": 0.049952107495064596, "grad_norm": 2.1176867485046387, "learning_rate": 2.6667262987817854e-05, "loss": 2.9154, "step": 5554 }, { "epoch": 0.04996110139270505, "grad_norm": 1.5952043533325195, "learning_rate": 2.665976387298401e-05, "loss": 2.5818, "step": 5555 }, { "epoch": 0.049970095290345504, "grad_norm": 1.7997909784317017, "learning_rate": 2.6652264608142484e-05, "loss": 2.5513, "step": 5556 }, { "epoch": 0.04997908918798595, "grad_norm": 1.9216657876968384, "learning_rate": 2.664476519397107e-05, "loss": 2.6349, "step": 5557 }, { "epoch": 0.049988083085626404, "grad_norm": 1.7526179552078247, "learning_rate": 2.663726563114755e-05, "loss": 2.614, "step": 5558 }, { "epoch": 0.04999707698326685, "grad_norm": 1.4640496969223022, "learning_rate": 2.6629765920349725e-05, "loss": 2.4994, "step": 5559 }, { "epoch": 0.050006070880907305, "grad_norm": 1.2515791654586792, "learning_rate": 2.662226606225541e-05, "loss": 2.4463, "step": 5560 }, { "epoch": 0.05001506477854775, "grad_norm": 2.2378218173980713, "learning_rate": 2.661476605754244e-05, "loss": 3.0487, "step": 5561 }, { "epoch": 0.050024058676188206, "grad_norm": 1.3359589576721191, "learning_rate": 2.6607265906888645e-05, "loss": 2.4199, "step": 5562 }, { "epoch": 0.05003305257382866, "grad_norm": 1.6423211097717285, "learning_rate": 2.659976561097189e-05, "loss": 2.4637, "step": 5563 }, { "epoch": 0.05004204647146911, "grad_norm": 2.9247517585754395, "learning_rate": 2.659226517047005e-05, "loss": 2.4904, "step": 5564 }, { "epoch": 0.05005104036910956, "grad_norm": 1.7688404321670532, "learning_rate": 2.658476458606099e-05, "loss": 2.8532, "step": 5565 }, { "epoch": 0.05006003426675001, "grad_norm": 1.3896958827972412, "learning_rate": 2.6577263858422622e-05, "loss": 2.4644, "step": 5566 }, { "epoch": 0.05006902816439046, "grad_norm": 1.8986960649490356, "learning_rate": 2.656976298823284e-05, "loss": 2.3834, "step": 5567 }, { "epoch": 0.050078022062030915, "grad_norm": 1.1192917823791504, "learning_rate": 2.6562261976169566e-05, "loss": 2.2831, "step": 5568 }, { "epoch": 0.05008701595967136, "grad_norm": 2.0229685306549072, "learning_rate": 2.655476082291074e-05, "loss": 2.4688, "step": 5569 }, { "epoch": 0.050096009857311816, "grad_norm": 1.5438594818115234, "learning_rate": 2.654725952913431e-05, "loss": 2.5538, "step": 5570 }, { "epoch": 0.05010500375495226, "grad_norm": 1.5116268396377563, "learning_rate": 2.653975809551823e-05, "loss": 2.6395, "step": 5571 }, { "epoch": 0.050113997652592716, "grad_norm": 1.5090124607086182, "learning_rate": 2.6532256522740474e-05, "loss": 2.3747, "step": 5572 }, { "epoch": 0.05012299155023317, "grad_norm": 1.4434394836425781, "learning_rate": 2.652475481147902e-05, "loss": 2.6127, "step": 5573 }, { "epoch": 0.05013198544787362, "grad_norm": 1.3658517599105835, "learning_rate": 2.6517252962411865e-05, "loss": 2.4179, "step": 5574 }, { "epoch": 0.05014097934551407, "grad_norm": 1.499442219734192, "learning_rate": 2.6509750976217017e-05, "loss": 2.5566, "step": 5575 }, { "epoch": 0.05014997324315452, "grad_norm": 2.001338481903076, "learning_rate": 2.650224885357251e-05, "loss": 2.8129, "step": 5576 }, { "epoch": 0.05015896714079497, "grad_norm": 1.4080146551132202, "learning_rate": 2.6494746595156365e-05, "loss": 2.4425, "step": 5577 }, { "epoch": 0.05016796103843542, "grad_norm": 1.6663414239883423, "learning_rate": 2.6487244201646645e-05, "loss": 2.499, "step": 5578 }, { "epoch": 0.05017695493607587, "grad_norm": 2.074810028076172, "learning_rate": 2.6479741673721392e-05, "loss": 2.9187, "step": 5579 }, { "epoch": 0.050185948833716326, "grad_norm": 2.0353024005889893, "learning_rate": 2.6472239012058677e-05, "loss": 2.5578, "step": 5580 }, { "epoch": 0.05019494273135677, "grad_norm": 2.1320505142211914, "learning_rate": 2.646473621733658e-05, "loss": 2.7362, "step": 5581 }, { "epoch": 0.05020393662899723, "grad_norm": 2.412168502807617, "learning_rate": 2.6457233290233202e-05, "loss": 2.7401, "step": 5582 }, { "epoch": 0.050212930526637674, "grad_norm": 1.713963270187378, "learning_rate": 2.6449730231426658e-05, "loss": 2.6387, "step": 5583 }, { "epoch": 0.05022192442427813, "grad_norm": 1.4034247398376465, "learning_rate": 2.6442227041595053e-05, "loss": 2.4867, "step": 5584 }, { "epoch": 0.05023091832191858, "grad_norm": 1.1011943817138672, "learning_rate": 2.6434723721416517e-05, "loss": 2.2568, "step": 5585 }, { "epoch": 0.05023991221955903, "grad_norm": 2.7355363368988037, "learning_rate": 2.6427220271569203e-05, "loss": 2.5886, "step": 5586 }, { "epoch": 0.05024890611719948, "grad_norm": 1.5769543647766113, "learning_rate": 2.6419716692731257e-05, "loss": 2.4495, "step": 5587 }, { "epoch": 0.05025790001483993, "grad_norm": 1.7275733947753906, "learning_rate": 2.6412212985580843e-05, "loss": 2.6843, "step": 5588 }, { "epoch": 0.05026689391248038, "grad_norm": 1.1602823734283447, "learning_rate": 2.6404709150796137e-05, "loss": 2.4122, "step": 5589 }, { "epoch": 0.05027588781012083, "grad_norm": 1.7689666748046875, "learning_rate": 2.639720518905534e-05, "loss": 2.8547, "step": 5590 }, { "epoch": 0.050284881707761284, "grad_norm": 2.287754774093628, "learning_rate": 2.6389701101036636e-05, "loss": 3.0272, "step": 5591 }, { "epoch": 0.05029387560540174, "grad_norm": 1.755926489830017, "learning_rate": 2.6382196887418244e-05, "loss": 2.7299, "step": 5592 }, { "epoch": 0.050302869503042184, "grad_norm": 1.311213731765747, "learning_rate": 2.6374692548878395e-05, "loss": 2.3967, "step": 5593 }, { "epoch": 0.05031186340068264, "grad_norm": 1.6412413120269775, "learning_rate": 2.6367188086095303e-05, "loss": 2.4286, "step": 5594 }, { "epoch": 0.050320857298323085, "grad_norm": 1.0997518301010132, "learning_rate": 2.6359683499747224e-05, "loss": 2.2816, "step": 5595 }, { "epoch": 0.05032985119596354, "grad_norm": 1.763487696647644, "learning_rate": 2.6352178790512422e-05, "loss": 2.7185, "step": 5596 }, { "epoch": 0.05033884509360399, "grad_norm": 1.149881362915039, "learning_rate": 2.6344673959069156e-05, "loss": 2.5189, "step": 5597 }, { "epoch": 0.05034783899124444, "grad_norm": 2.1016149520874023, "learning_rate": 2.63371690060957e-05, "loss": 2.6213, "step": 5598 }, { "epoch": 0.05035683288888489, "grad_norm": 1.616387128829956, "learning_rate": 2.632966393227036e-05, "loss": 2.6302, "step": 5599 }, { "epoch": 0.05036582678652534, "grad_norm": 2.0580921173095703, "learning_rate": 2.632215873827142e-05, "loss": 2.6708, "step": 5600 }, { "epoch": 0.050374820684165794, "grad_norm": 3.094907760620117, "learning_rate": 2.6314653424777193e-05, "loss": 2.5417, "step": 5601 }, { "epoch": 0.05038381458180624, "grad_norm": 1.293916940689087, "learning_rate": 2.6307147992466013e-05, "loss": 2.5197, "step": 5602 }, { "epoch": 0.050392808479446695, "grad_norm": 1.298364520072937, "learning_rate": 2.6299642442016198e-05, "loss": 2.7644, "step": 5603 }, { "epoch": 0.05040180237708715, "grad_norm": 1.7221360206604004, "learning_rate": 2.62921367741061e-05, "loss": 2.6758, "step": 5604 }, { "epoch": 0.050410796274727596, "grad_norm": 1.7285192012786865, "learning_rate": 2.6284630989414078e-05, "loss": 2.6601, "step": 5605 }, { "epoch": 0.05041979017236805, "grad_norm": 2.055485486984253, "learning_rate": 2.6277125088618492e-05, "loss": 2.8213, "step": 5606 }, { "epoch": 0.050428784070008496, "grad_norm": 1.1942862272262573, "learning_rate": 2.626961907239771e-05, "loss": 2.3221, "step": 5607 }, { "epoch": 0.05043777796764895, "grad_norm": 1.899765133857727, "learning_rate": 2.6262112941430127e-05, "loss": 2.7297, "step": 5608 }, { "epoch": 0.050446771865289404, "grad_norm": 2.2265801429748535, "learning_rate": 2.6254606696394135e-05, "loss": 2.8307, "step": 5609 }, { "epoch": 0.05045576576292985, "grad_norm": 1.8032729625701904, "learning_rate": 2.624710033796814e-05, "loss": 2.7153, "step": 5610 }, { "epoch": 0.050464759660570305, "grad_norm": 1.2635847330093384, "learning_rate": 2.623959386683056e-05, "loss": 2.2997, "step": 5611 }, { "epoch": 0.05047375355821075, "grad_norm": 1.6811776161193848, "learning_rate": 2.6232087283659818e-05, "loss": 2.7006, "step": 5612 }, { "epoch": 0.050482747455851205, "grad_norm": 1.9073692560195923, "learning_rate": 2.622458058913436e-05, "loss": 2.6722, "step": 5613 }, { "epoch": 0.05049174135349166, "grad_norm": 1.4261687994003296, "learning_rate": 2.6217073783932634e-05, "loss": 2.5353, "step": 5614 }, { "epoch": 0.050500735251132106, "grad_norm": 1.1898109912872314, "learning_rate": 2.6209566868733075e-05, "loss": 2.2101, "step": 5615 }, { "epoch": 0.05050972914877256, "grad_norm": 1.4951753616333008, "learning_rate": 2.6202059844214176e-05, "loss": 2.5737, "step": 5616 }, { "epoch": 0.05051872304641301, "grad_norm": 0.8305614590644836, "learning_rate": 2.6194552711054398e-05, "loss": 2.2384, "step": 5617 }, { "epoch": 0.05052771694405346, "grad_norm": 1.9361591339111328, "learning_rate": 2.618704546993223e-05, "loss": 2.9717, "step": 5618 }, { "epoch": 0.05053671084169391, "grad_norm": 1.4848889112472534, "learning_rate": 2.617953812152618e-05, "loss": 2.7647, "step": 5619 }, { "epoch": 0.05054570473933436, "grad_norm": 1.5303785800933838, "learning_rate": 2.6172030666514734e-05, "loss": 2.7746, "step": 5620 }, { "epoch": 0.050554698636974815, "grad_norm": 1.34975004196167, "learning_rate": 2.6164523105576432e-05, "loss": 2.3615, "step": 5621 }, { "epoch": 0.05056369253461526, "grad_norm": 1.69707453250885, "learning_rate": 2.6157015439389774e-05, "loss": 2.6066, "step": 5622 }, { "epoch": 0.050572686432255716, "grad_norm": 1.2968112230300903, "learning_rate": 2.6149507668633318e-05, "loss": 2.3105, "step": 5623 }, { "epoch": 0.05058168032989616, "grad_norm": 1.9787888526916504, "learning_rate": 2.6141999793985583e-05, "loss": 2.6917, "step": 5624 }, { "epoch": 0.05059067422753662, "grad_norm": 1.6411546468734741, "learning_rate": 2.6134491816125146e-05, "loss": 2.5663, "step": 5625 }, { "epoch": 0.05059966812517707, "grad_norm": 1.2326549291610718, "learning_rate": 2.612698373573056e-05, "loss": 2.3716, "step": 5626 }, { "epoch": 0.05060866202281752, "grad_norm": 2.6172499656677246, "learning_rate": 2.6119475553480406e-05, "loss": 2.5903, "step": 5627 }, { "epoch": 0.05061765592045797, "grad_norm": 2.9417128562927246, "learning_rate": 2.6111967270053246e-05, "loss": 2.3849, "step": 5628 }, { "epoch": 0.05062664981809842, "grad_norm": 2.041621446609497, "learning_rate": 2.610445888612768e-05, "loss": 2.5323, "step": 5629 }, { "epoch": 0.05063564371573887, "grad_norm": 1.565558671951294, "learning_rate": 2.6096950402382308e-05, "loss": 2.4827, "step": 5630 }, { "epoch": 0.05064463761337932, "grad_norm": 2.2812423706054688, "learning_rate": 2.6089441819495747e-05, "loss": 2.5929, "step": 5631 }, { "epoch": 0.05065363151101977, "grad_norm": 0.7207363247871399, "learning_rate": 2.6081933138146607e-05, "loss": 2.2173, "step": 5632 }, { "epoch": 0.050662625408660227, "grad_norm": 1.7431845664978027, "learning_rate": 2.6074424359013517e-05, "loss": 2.6117, "step": 5633 }, { "epoch": 0.05067161930630067, "grad_norm": 1.6367127895355225, "learning_rate": 2.6066915482775113e-05, "loss": 2.5117, "step": 5634 }, { "epoch": 0.05068061320394113, "grad_norm": 1.7251893281936646, "learning_rate": 2.605940651011003e-05, "loss": 2.5594, "step": 5635 }, { "epoch": 0.050689607101581574, "grad_norm": 1.3751147985458374, "learning_rate": 2.6051897441696927e-05, "loss": 2.5108, "step": 5636 }, { "epoch": 0.05069860099922203, "grad_norm": 2.0286812782287598, "learning_rate": 2.6044388278214478e-05, "loss": 2.8021, "step": 5637 }, { "epoch": 0.05070759489686248, "grad_norm": 1.4338791370391846, "learning_rate": 2.6036879020341327e-05, "loss": 2.6964, "step": 5638 }, { "epoch": 0.05071658879450293, "grad_norm": 2.530937671661377, "learning_rate": 2.602936966875617e-05, "loss": 2.7466, "step": 5639 }, { "epoch": 0.05072558269214338, "grad_norm": 1.8172882795333862, "learning_rate": 2.6021860224137694e-05, "loss": 2.6524, "step": 5640 }, { "epoch": 0.05073457658978383, "grad_norm": 1.7562512159347534, "learning_rate": 2.6014350687164595e-05, "loss": 2.3209, "step": 5641 }, { "epoch": 0.05074357048742428, "grad_norm": 1.5149487257003784, "learning_rate": 2.600684105851557e-05, "loss": 2.4523, "step": 5642 }, { "epoch": 0.05075256438506473, "grad_norm": 1.7390376329421997, "learning_rate": 2.5999331338869336e-05, "loss": 2.721, "step": 5643 }, { "epoch": 0.050761558282705184, "grad_norm": 2.058654546737671, "learning_rate": 2.599182152890461e-05, "loss": 2.6694, "step": 5644 }, { "epoch": 0.05077055218034564, "grad_norm": 1.1723181009292603, "learning_rate": 2.5984311629300124e-05, "loss": 2.3174, "step": 5645 }, { "epoch": 0.050779546077986085, "grad_norm": 1.8678486347198486, "learning_rate": 2.5976801640734605e-05, "loss": 2.4659, "step": 5646 }, { "epoch": 0.05078853997562654, "grad_norm": 1.6455680131912231, "learning_rate": 2.5969291563886812e-05, "loss": 2.5511, "step": 5647 }, { "epoch": 0.050797533873266985, "grad_norm": 1.584984540939331, "learning_rate": 2.5961781399435497e-05, "loss": 2.567, "step": 5648 }, { "epoch": 0.05080652777090744, "grad_norm": 1.6565089225769043, "learning_rate": 2.5954271148059418e-05, "loss": 2.5342, "step": 5649 }, { "epoch": 0.05081552166854789, "grad_norm": 1.1301003694534302, "learning_rate": 2.5946760810437327e-05, "loss": 2.3093, "step": 5650 }, { "epoch": 0.05082451556618834, "grad_norm": 1.3659570217132568, "learning_rate": 2.593925038724802e-05, "loss": 2.417, "step": 5651 }, { "epoch": 0.050833509463828794, "grad_norm": 2.6098177433013916, "learning_rate": 2.593173987917027e-05, "loss": 2.5289, "step": 5652 }, { "epoch": 0.05084250336146924, "grad_norm": 1.7693769931793213, "learning_rate": 2.5924229286882877e-05, "loss": 2.478, "step": 5653 }, { "epoch": 0.050851497259109694, "grad_norm": 1.7723768949508667, "learning_rate": 2.5916718611064643e-05, "loss": 2.7528, "step": 5654 }, { "epoch": 0.05086049115675015, "grad_norm": 1.3806558847427368, "learning_rate": 2.5909207852394363e-05, "loss": 2.857, "step": 5655 }, { "epoch": 0.050869485054390595, "grad_norm": 1.4030601978302002, "learning_rate": 2.5901697011550858e-05, "loss": 2.4908, "step": 5656 }, { "epoch": 0.05087847895203105, "grad_norm": 1.7481403350830078, "learning_rate": 2.5894186089212947e-05, "loss": 2.7237, "step": 5657 }, { "epoch": 0.050887472849671496, "grad_norm": 1.2980701923370361, "learning_rate": 2.5886675086059454e-05, "loss": 2.5545, "step": 5658 }, { "epoch": 0.05089646674731195, "grad_norm": 2.013845443725586, "learning_rate": 2.5879164002769228e-05, "loss": 2.6363, "step": 5659 }, { "epoch": 0.0509054606449524, "grad_norm": 1.3968217372894287, "learning_rate": 2.5871652840021104e-05, "loss": 2.69, "step": 5660 }, { "epoch": 0.05091445454259285, "grad_norm": 1.5900264978408813, "learning_rate": 2.5864141598493942e-05, "loss": 2.6622, "step": 5661 }, { "epoch": 0.050923448440233304, "grad_norm": 1.9684886932373047, "learning_rate": 2.5856630278866588e-05, "loss": 2.7172, "step": 5662 }, { "epoch": 0.05093244233787375, "grad_norm": 1.0023187398910522, "learning_rate": 2.5849118881817913e-05, "loss": 2.4435, "step": 5663 }, { "epoch": 0.050941436235514205, "grad_norm": 1.667944073677063, "learning_rate": 2.5841607408026785e-05, "loss": 2.7098, "step": 5664 }, { "epoch": 0.05095043013315465, "grad_norm": 1.5394574403762817, "learning_rate": 2.5834095858172085e-05, "loss": 2.5243, "step": 5665 }, { "epoch": 0.050959424030795106, "grad_norm": 1.437364935874939, "learning_rate": 2.5826584232932706e-05, "loss": 2.5359, "step": 5666 }, { "epoch": 0.05096841792843556, "grad_norm": 2.2309987545013428, "learning_rate": 2.5819072532987532e-05, "loss": 2.6592, "step": 5667 }, { "epoch": 0.050977411826076006, "grad_norm": 2.100717306137085, "learning_rate": 2.5811560759015463e-05, "loss": 2.7186, "step": 5668 }, { "epoch": 0.05098640572371646, "grad_norm": 1.7888046503067017, "learning_rate": 2.5804048911695412e-05, "loss": 2.5924, "step": 5669 }, { "epoch": 0.05099539962135691, "grad_norm": 1.4380017518997192, "learning_rate": 2.5796536991706284e-05, "loss": 2.3794, "step": 5670 }, { "epoch": 0.05100439351899736, "grad_norm": 0.9121497273445129, "learning_rate": 2.5789024999726996e-05, "loss": 2.4321, "step": 5671 }, { "epoch": 0.05101338741663781, "grad_norm": 1.4581583738327026, "learning_rate": 2.5781512936436482e-05, "loss": 2.6013, "step": 5672 }, { "epoch": 0.05102238131427826, "grad_norm": 1.4425793886184692, "learning_rate": 2.577400080251367e-05, "loss": 2.5358, "step": 5673 }, { "epoch": 0.051031375211918716, "grad_norm": 1.394791841506958, "learning_rate": 2.57664885986375e-05, "loss": 2.4486, "step": 5674 }, { "epoch": 0.05104036910955916, "grad_norm": 1.524783968925476, "learning_rate": 2.575897632548691e-05, "loss": 2.5267, "step": 5675 }, { "epoch": 0.051049363007199616, "grad_norm": 1.5146634578704834, "learning_rate": 2.575146398374087e-05, "loss": 2.6092, "step": 5676 }, { "epoch": 0.05105835690484006, "grad_norm": 1.6495107412338257, "learning_rate": 2.5743951574078314e-05, "loss": 2.5872, "step": 5677 }, { "epoch": 0.05106735080248052, "grad_norm": 1.1332262754440308, "learning_rate": 2.5736439097178216e-05, "loss": 2.4682, "step": 5678 }, { "epoch": 0.05107634470012097, "grad_norm": 1.8231858015060425, "learning_rate": 2.5728926553719557e-05, "loss": 2.6832, "step": 5679 }, { "epoch": 0.05108533859776142, "grad_norm": 1.9676088094711304, "learning_rate": 2.5721413944381288e-05, "loss": 2.8934, "step": 5680 }, { "epoch": 0.05109433249540187, "grad_norm": 2.6704723834991455, "learning_rate": 2.5713901269842404e-05, "loss": 2.5715, "step": 5681 }, { "epoch": 0.05110332639304232, "grad_norm": 1.63522469997406, "learning_rate": 2.57063885307819e-05, "loss": 2.6661, "step": 5682 }, { "epoch": 0.05111232029068277, "grad_norm": 1.5169837474822998, "learning_rate": 2.5698875727878763e-05, "loss": 2.527, "step": 5683 }, { "epoch": 0.051121314188323226, "grad_norm": 1.0579332113265991, "learning_rate": 2.569136286181198e-05, "loss": 2.4083, "step": 5684 }, { "epoch": 0.05113030808596367, "grad_norm": 1.660075306892395, "learning_rate": 2.568384993326058e-05, "loss": 2.5331, "step": 5685 }, { "epoch": 0.05113930198360413, "grad_norm": 1.8644059896469116, "learning_rate": 2.5676336942903547e-05, "loss": 2.709, "step": 5686 }, { "epoch": 0.051148295881244574, "grad_norm": 1.5392996072769165, "learning_rate": 2.5668823891419913e-05, "loss": 2.586, "step": 5687 }, { "epoch": 0.05115728977888503, "grad_norm": 2.6424272060394287, "learning_rate": 2.5661310779488695e-05, "loss": 2.82, "step": 5688 }, { "epoch": 0.051166283676525474, "grad_norm": 1.7052751779556274, "learning_rate": 2.5653797607788925e-05, "loss": 2.8215, "step": 5689 }, { "epoch": 0.05117527757416593, "grad_norm": 1.737575650215149, "learning_rate": 2.5646284376999628e-05, "loss": 2.5822, "step": 5690 }, { "epoch": 0.05118427147180638, "grad_norm": 2.478792667388916, "learning_rate": 2.5638771087799857e-05, "loss": 2.953, "step": 5691 }, { "epoch": 0.05119326536944683, "grad_norm": 1.025389552116394, "learning_rate": 2.563125774086863e-05, "loss": 2.365, "step": 5692 }, { "epoch": 0.05120225926708728, "grad_norm": 1.9161714315414429, "learning_rate": 2.5623744336885007e-05, "loss": 2.3603, "step": 5693 }, { "epoch": 0.05121125316472773, "grad_norm": 1.9706228971481323, "learning_rate": 2.5616230876528047e-05, "loss": 2.4984, "step": 5694 }, { "epoch": 0.051220247062368184, "grad_norm": 1.475502610206604, "learning_rate": 2.5608717360476803e-05, "loss": 2.4936, "step": 5695 }, { "epoch": 0.05122924096000864, "grad_norm": 2.1352591514587402, "learning_rate": 2.5601203789410342e-05, "loss": 2.4263, "step": 5696 }, { "epoch": 0.051238234857649084, "grad_norm": 1.0987627506256104, "learning_rate": 2.559369016400773e-05, "loss": 2.3574, "step": 5697 }, { "epoch": 0.05124722875528954, "grad_norm": 2.082176685333252, "learning_rate": 2.5586176484948043e-05, "loss": 2.7801, "step": 5698 }, { "epoch": 0.051256222652929985, "grad_norm": 1.6106228828430176, "learning_rate": 2.5578662752910347e-05, "loss": 2.7036, "step": 5699 }, { "epoch": 0.05126521655057044, "grad_norm": 1.4153437614440918, "learning_rate": 2.5571148968573743e-05, "loss": 2.6554, "step": 5700 }, { "epoch": 0.051274210448210886, "grad_norm": 1.7262197732925415, "learning_rate": 2.5563635132617302e-05, "loss": 2.3652, "step": 5701 }, { "epoch": 0.05128320434585134, "grad_norm": 1.683018445968628, "learning_rate": 2.5556121245720133e-05, "loss": 2.4396, "step": 5702 }, { "epoch": 0.05129219824349179, "grad_norm": 1.8770898580551147, "learning_rate": 2.5548607308561323e-05, "loss": 2.5842, "step": 5703 }, { "epoch": 0.05130119214113224, "grad_norm": 2.093803882598877, "learning_rate": 2.554109332181998e-05, "loss": 2.8134, "step": 5704 }, { "epoch": 0.051310186038772694, "grad_norm": 1.7443187236785889, "learning_rate": 2.5533579286175202e-05, "loss": 2.7325, "step": 5705 }, { "epoch": 0.05131917993641314, "grad_norm": 1.7994638681411743, "learning_rate": 2.5526065202306098e-05, "loss": 2.7357, "step": 5706 }, { "epoch": 0.051328173834053595, "grad_norm": 1.1174148321151733, "learning_rate": 2.5518551070891788e-05, "loss": 2.3026, "step": 5707 }, { "epoch": 0.05133716773169405, "grad_norm": 1.4081101417541504, "learning_rate": 2.55110368926114e-05, "loss": 2.3842, "step": 5708 }, { "epoch": 0.051346161629334496, "grad_norm": 1.3615401983261108, "learning_rate": 2.5503522668144035e-05, "loss": 2.4586, "step": 5709 }, { "epoch": 0.05135515552697495, "grad_norm": 3.271449565887451, "learning_rate": 2.5496008398168843e-05, "loss": 2.7474, "step": 5710 }, { "epoch": 0.051364149424615396, "grad_norm": 1.4275376796722412, "learning_rate": 2.5488494083364944e-05, "loss": 2.5052, "step": 5711 }, { "epoch": 0.05137314332225585, "grad_norm": 1.3856492042541504, "learning_rate": 2.5480979724411475e-05, "loss": 2.5459, "step": 5712 }, { "epoch": 0.0513821372198963, "grad_norm": 1.8587093353271484, "learning_rate": 2.5473465321987578e-05, "loss": 2.6716, "step": 5713 }, { "epoch": 0.05139113111753675, "grad_norm": 1.8865816593170166, "learning_rate": 2.54659508767724e-05, "loss": 2.4446, "step": 5714 }, { "epoch": 0.051400125015177205, "grad_norm": 1.9292010068893433, "learning_rate": 2.5458436389445073e-05, "loss": 2.9046, "step": 5715 }, { "epoch": 0.05140911891281765, "grad_norm": 1.5918726921081543, "learning_rate": 2.545092186068476e-05, "loss": 2.7347, "step": 5716 }, { "epoch": 0.051418112810458105, "grad_norm": 2.113602876663208, "learning_rate": 2.5443407291170622e-05, "loss": 2.5389, "step": 5717 }, { "epoch": 0.05142710670809855, "grad_norm": 1.7031230926513672, "learning_rate": 2.5435892681581808e-05, "loss": 2.4595, "step": 5718 }, { "epoch": 0.051436100605739006, "grad_norm": 1.7899807691574097, "learning_rate": 2.5428378032597482e-05, "loss": 2.7105, "step": 5719 }, { "epoch": 0.05144509450337946, "grad_norm": 2.7382519245147705, "learning_rate": 2.542086334489681e-05, "loss": 2.6626, "step": 5720 }, { "epoch": 0.05145408840101991, "grad_norm": 1.6151330471038818, "learning_rate": 2.5413348619158967e-05, "loss": 2.6739, "step": 5721 }, { "epoch": 0.05146308229866036, "grad_norm": 2.175398111343384, "learning_rate": 2.5405833856063115e-05, "loss": 2.7716, "step": 5722 }, { "epoch": 0.05147207619630081, "grad_norm": 1.6193041801452637, "learning_rate": 2.5398319056288432e-05, "loss": 2.6495, "step": 5723 }, { "epoch": 0.05148107009394126, "grad_norm": 2.3205552101135254, "learning_rate": 2.5390804220514113e-05, "loss": 2.8373, "step": 5724 }, { "epoch": 0.051490063991581715, "grad_norm": 1.6913609504699707, "learning_rate": 2.5383289349419327e-05, "loss": 2.6643, "step": 5725 }, { "epoch": 0.05149905788922216, "grad_norm": 1.6914935111999512, "learning_rate": 2.5375774443683265e-05, "loss": 2.3733, "step": 5726 }, { "epoch": 0.051508051786862616, "grad_norm": 1.337083101272583, "learning_rate": 2.5368259503985102e-05, "loss": 2.5532, "step": 5727 }, { "epoch": 0.05151704568450306, "grad_norm": 1.5660969018936157, "learning_rate": 2.536074453100405e-05, "loss": 2.7679, "step": 5728 }, { "epoch": 0.05152603958214352, "grad_norm": 1.5381436347961426, "learning_rate": 2.5353229525419293e-05, "loss": 2.7056, "step": 5729 }, { "epoch": 0.051535033479783963, "grad_norm": 1.0120609998703003, "learning_rate": 2.5345714487910034e-05, "loss": 2.2939, "step": 5730 }, { "epoch": 0.05154402737742442, "grad_norm": 1.4623103141784668, "learning_rate": 2.533819941915547e-05, "loss": 2.4368, "step": 5731 }, { "epoch": 0.05155302127506487, "grad_norm": 1.4275404214859009, "learning_rate": 2.5330684319834814e-05, "loss": 2.4952, "step": 5732 }, { "epoch": 0.05156201517270532, "grad_norm": 1.6544805765151978, "learning_rate": 2.5323169190627255e-05, "loss": 2.5623, "step": 5733 }, { "epoch": 0.05157100907034577, "grad_norm": 1.5572311878204346, "learning_rate": 2.5315654032212016e-05, "loss": 2.5175, "step": 5734 }, { "epoch": 0.05158000296798622, "grad_norm": 1.3152873516082764, "learning_rate": 2.530813884526831e-05, "loss": 2.4723, "step": 5735 }, { "epoch": 0.05158899686562667, "grad_norm": 1.9200719594955444, "learning_rate": 2.530062363047534e-05, "loss": 2.4964, "step": 5736 }, { "epoch": 0.051597990763267126, "grad_norm": 1.397160530090332, "learning_rate": 2.529310838851233e-05, "loss": 2.4458, "step": 5737 }, { "epoch": 0.05160698466090757, "grad_norm": 0.9311887621879578, "learning_rate": 2.528559312005851e-05, "loss": 2.2869, "step": 5738 }, { "epoch": 0.05161597855854803, "grad_norm": 1.2999449968338013, "learning_rate": 2.5278077825793083e-05, "loss": 2.4668, "step": 5739 }, { "epoch": 0.051624972456188474, "grad_norm": 2.2117865085601807, "learning_rate": 2.5270562506395284e-05, "loss": 2.5125, "step": 5740 }, { "epoch": 0.05163396635382893, "grad_norm": 1.7617484331130981, "learning_rate": 2.5263047162544336e-05, "loss": 2.3805, "step": 5741 }, { "epoch": 0.051642960251469375, "grad_norm": 1.66494619846344, "learning_rate": 2.525553179491947e-05, "loss": 2.4003, "step": 5742 }, { "epoch": 0.05165195414910983, "grad_norm": 1.1821058988571167, "learning_rate": 2.5248016404199908e-05, "loss": 2.5074, "step": 5743 }, { "epoch": 0.05166094804675028, "grad_norm": 1.4768482446670532, "learning_rate": 2.5240500991064903e-05, "loss": 2.4944, "step": 5744 }, { "epoch": 0.05166994194439073, "grad_norm": 1.306566596031189, "learning_rate": 2.523298555619367e-05, "loss": 2.4354, "step": 5745 }, { "epoch": 0.05167893584203118, "grad_norm": 1.626410722732544, "learning_rate": 2.5225470100265457e-05, "loss": 2.7334, "step": 5746 }, { "epoch": 0.05168792973967163, "grad_norm": 1.7012288570404053, "learning_rate": 2.52179546239595e-05, "loss": 2.7556, "step": 5747 }, { "epoch": 0.051696923637312084, "grad_norm": 1.7441771030426025, "learning_rate": 2.521043912795503e-05, "loss": 2.5484, "step": 5748 }, { "epoch": 0.05170591753495254, "grad_norm": 1.7051887512207031, "learning_rate": 2.5202923612931308e-05, "loss": 2.5202, "step": 5749 }, { "epoch": 0.051714911432592985, "grad_norm": 1.3768846988677979, "learning_rate": 2.5195408079567566e-05, "loss": 2.4962, "step": 5750 }, { "epoch": 0.05172390533023344, "grad_norm": 1.3151389360427856, "learning_rate": 2.518789252854305e-05, "loss": 2.6759, "step": 5751 }, { "epoch": 0.051732899227873885, "grad_norm": 1.199027180671692, "learning_rate": 2.518037696053701e-05, "loss": 2.2783, "step": 5752 }, { "epoch": 0.05174189312551434, "grad_norm": 2.040562391281128, "learning_rate": 2.5172861376228695e-05, "loss": 2.5961, "step": 5753 }, { "epoch": 0.05175088702315479, "grad_norm": 2.2873504161834717, "learning_rate": 2.5165345776297355e-05, "loss": 2.5352, "step": 5754 }, { "epoch": 0.05175988092079524, "grad_norm": 1.7497658729553223, "learning_rate": 2.515783016142224e-05, "loss": 2.3151, "step": 5755 }, { "epoch": 0.051768874818435694, "grad_norm": 1.6206729412078857, "learning_rate": 2.5150314532282616e-05, "loss": 2.5364, "step": 5756 }, { "epoch": 0.05177786871607614, "grad_norm": 2.8635201454162598, "learning_rate": 2.514279888955771e-05, "loss": 3.0189, "step": 5757 }, { "epoch": 0.051786862613716594, "grad_norm": 2.476752281188965, "learning_rate": 2.5135283233926803e-05, "loss": 2.7318, "step": 5758 }, { "epoch": 0.05179585651135704, "grad_norm": 1.3079414367675781, "learning_rate": 2.5127767566069144e-05, "loss": 2.4175, "step": 5759 }, { "epoch": 0.051804850408997495, "grad_norm": 1.3087109327316284, "learning_rate": 2.5120251886663986e-05, "loss": 2.5092, "step": 5760 }, { "epoch": 0.05181384430663795, "grad_norm": 1.4002362489700317, "learning_rate": 2.5112736196390595e-05, "loss": 2.3874, "step": 5761 }, { "epoch": 0.051822838204278396, "grad_norm": 2.0239415168762207, "learning_rate": 2.510522049592823e-05, "loss": 2.5189, "step": 5762 }, { "epoch": 0.05183183210191885, "grad_norm": 2.1252546310424805, "learning_rate": 2.5097704785956145e-05, "loss": 2.6577, "step": 5763 }, { "epoch": 0.0518408259995593, "grad_norm": 2.0474720001220703, "learning_rate": 2.5090189067153608e-05, "loss": 3.1034, "step": 5764 }, { "epoch": 0.05184981989719975, "grad_norm": 1.351408839225769, "learning_rate": 2.508267334019988e-05, "loss": 2.4237, "step": 5765 }, { "epoch": 0.051858813794840204, "grad_norm": 1.5209397077560425, "learning_rate": 2.507515760577423e-05, "loss": 2.518, "step": 5766 }, { "epoch": 0.05186780769248065, "grad_norm": 1.6421632766723633, "learning_rate": 2.5067641864555907e-05, "loss": 2.6036, "step": 5767 }, { "epoch": 0.051876801590121105, "grad_norm": 2.024538993835449, "learning_rate": 2.506012611722419e-05, "loss": 2.6293, "step": 5768 }, { "epoch": 0.05188579548776155, "grad_norm": 1.8205511569976807, "learning_rate": 2.505261036445834e-05, "loss": 2.6587, "step": 5769 }, { "epoch": 0.051894789385402006, "grad_norm": 1.937888741493225, "learning_rate": 2.5045094606937618e-05, "loss": 2.7376, "step": 5770 }, { "epoch": 0.05190378328304245, "grad_norm": 1.5759918689727783, "learning_rate": 2.5037578845341296e-05, "loss": 2.5323, "step": 5771 }, { "epoch": 0.051912777180682906, "grad_norm": 1.462791085243225, "learning_rate": 2.5030063080348642e-05, "loss": 2.6217, "step": 5772 }, { "epoch": 0.05192177107832336, "grad_norm": 1.519185185432434, "learning_rate": 2.502254731263892e-05, "loss": 2.4677, "step": 5773 }, { "epoch": 0.05193076497596381, "grad_norm": 2.2266361713409424, "learning_rate": 2.5015031542891397e-05, "loss": 2.6881, "step": 5774 }, { "epoch": 0.05193975887360426, "grad_norm": 1.8688948154449463, "learning_rate": 2.5007515771785334e-05, "loss": 2.5729, "step": 5775 }, { "epoch": 0.05194875277124471, "grad_norm": 2.2908875942230225, "learning_rate": 2.5e-05, "loss": 2.772, "step": 5776 }, { "epoch": 0.05195774666888516, "grad_norm": 1.4406849145889282, "learning_rate": 2.4992484228214675e-05, "loss": 2.5551, "step": 5777 }, { "epoch": 0.051966740566525615, "grad_norm": 1.7537091970443726, "learning_rate": 2.498496845710861e-05, "loss": 2.5157, "step": 5778 }, { "epoch": 0.05197573446416606, "grad_norm": 1.3089593648910522, "learning_rate": 2.4977452687361085e-05, "loss": 2.4803, "step": 5779 }, { "epoch": 0.051984728361806516, "grad_norm": 1.4574275016784668, "learning_rate": 2.496993691965136e-05, "loss": 2.4197, "step": 5780 }, { "epoch": 0.05199372225944696, "grad_norm": 1.202369213104248, "learning_rate": 2.4962421154658707e-05, "loss": 2.3345, "step": 5781 }, { "epoch": 0.05200271615708742, "grad_norm": 1.5907267332077026, "learning_rate": 2.4954905393062387e-05, "loss": 2.4484, "step": 5782 }, { "epoch": 0.052011710054727864, "grad_norm": 2.0568301677703857, "learning_rate": 2.4947389635541662e-05, "loss": 2.8117, "step": 5783 }, { "epoch": 0.05202070395236832, "grad_norm": 2.1567723751068115, "learning_rate": 2.4939873882775818e-05, "loss": 2.6806, "step": 5784 }, { "epoch": 0.05202969785000877, "grad_norm": 1.2421561479568481, "learning_rate": 2.4932358135444092e-05, "loss": 2.5101, "step": 5785 }, { "epoch": 0.05203869174764922, "grad_norm": 2.155468463897705, "learning_rate": 2.4924842394225777e-05, "loss": 2.8207, "step": 5786 }, { "epoch": 0.05204768564528967, "grad_norm": 1.3350147008895874, "learning_rate": 2.4917326659800123e-05, "loss": 2.399, "step": 5787 }, { "epoch": 0.05205667954293012, "grad_norm": 1.2629586458206177, "learning_rate": 2.4909810932846398e-05, "loss": 2.4653, "step": 5788 }, { "epoch": 0.05206567344057057, "grad_norm": 2.0579874515533447, "learning_rate": 2.4902295214043867e-05, "loss": 2.9479, "step": 5789 }, { "epoch": 0.05207466733821103, "grad_norm": 2.6926751136779785, "learning_rate": 2.4894779504071774e-05, "loss": 2.6168, "step": 5790 }, { "epoch": 0.052083661235851474, "grad_norm": 2.5529487133026123, "learning_rate": 2.488726380360941e-05, "loss": 2.702, "step": 5791 }, { "epoch": 0.05209265513349193, "grad_norm": 1.48592209815979, "learning_rate": 2.4879748113336013e-05, "loss": 2.4954, "step": 5792 }, { "epoch": 0.052101649031132374, "grad_norm": 1.5285180807113647, "learning_rate": 2.487223243393086e-05, "loss": 2.4055, "step": 5793 }, { "epoch": 0.05211064292877283, "grad_norm": 1.1473162174224854, "learning_rate": 2.4864716766073202e-05, "loss": 2.3112, "step": 5794 }, { "epoch": 0.05211963682641328, "grad_norm": 1.338918685913086, "learning_rate": 2.4857201110442294e-05, "loss": 2.559, "step": 5795 }, { "epoch": 0.05212863072405373, "grad_norm": 0.9491031169891357, "learning_rate": 2.48496854677174e-05, "loss": 2.3647, "step": 5796 }, { "epoch": 0.05213762462169418, "grad_norm": 1.2293965816497803, "learning_rate": 2.484216983857776e-05, "loss": 2.529, "step": 5797 }, { "epoch": 0.05214661851933463, "grad_norm": 1.489202618598938, "learning_rate": 2.483465422370265e-05, "loss": 2.5194, "step": 5798 }, { "epoch": 0.05215561241697508, "grad_norm": 2.1044607162475586, "learning_rate": 2.4827138623771307e-05, "loss": 3.0618, "step": 5799 }, { "epoch": 0.05216460631461553, "grad_norm": 2.1535236835479736, "learning_rate": 2.4819623039462992e-05, "loss": 2.4571, "step": 5800 }, { "epoch": 0.052173600212255984, "grad_norm": 1.313037633895874, "learning_rate": 2.4812107471456954e-05, "loss": 2.4489, "step": 5801 }, { "epoch": 0.05218259410989644, "grad_norm": 2.0069146156311035, "learning_rate": 2.480459192043244e-05, "loss": 2.8404, "step": 5802 }, { "epoch": 0.052191588007536885, "grad_norm": 1.5892188549041748, "learning_rate": 2.47970763870687e-05, "loss": 2.7515, "step": 5803 }, { "epoch": 0.05220058190517734, "grad_norm": 1.9770779609680176, "learning_rate": 2.4789560872044967e-05, "loss": 2.7888, "step": 5804 }, { "epoch": 0.052209575802817786, "grad_norm": 0.9381443858146667, "learning_rate": 2.4782045376040508e-05, "loss": 2.2517, "step": 5805 }, { "epoch": 0.05221856970045824, "grad_norm": 2.002169132232666, "learning_rate": 2.4774529899734552e-05, "loss": 2.7943, "step": 5806 }, { "epoch": 0.05222756359809869, "grad_norm": 1.40768301486969, "learning_rate": 2.4767014443806335e-05, "loss": 2.3723, "step": 5807 }, { "epoch": 0.05223655749573914, "grad_norm": 1.903940200805664, "learning_rate": 2.4759499008935103e-05, "loss": 2.2625, "step": 5808 }, { "epoch": 0.052245551393379594, "grad_norm": 2.1659562587738037, "learning_rate": 2.475198359580009e-05, "loss": 2.6755, "step": 5809 }, { "epoch": 0.05225454529102004, "grad_norm": 1.6180616617202759, "learning_rate": 2.474446820508054e-05, "loss": 2.4473, "step": 5810 }, { "epoch": 0.052263539188660495, "grad_norm": 2.186347007751465, "learning_rate": 2.4736952837455667e-05, "loss": 2.7268, "step": 5811 }, { "epoch": 0.05227253308630094, "grad_norm": 1.6272828578948975, "learning_rate": 2.472943749360472e-05, "loss": 2.3569, "step": 5812 }, { "epoch": 0.052281526983941395, "grad_norm": 2.0502071380615234, "learning_rate": 2.4721922174206923e-05, "loss": 2.7706, "step": 5813 }, { "epoch": 0.05229052088158185, "grad_norm": 2.2824580669403076, "learning_rate": 2.47144068799415e-05, "loss": 2.4716, "step": 5814 }, { "epoch": 0.052299514779222296, "grad_norm": 1.8339191675186157, "learning_rate": 2.4706891611487675e-05, "loss": 2.4472, "step": 5815 }, { "epoch": 0.05230850867686275, "grad_norm": 1.390195608139038, "learning_rate": 2.4699376369524663e-05, "loss": 2.5938, "step": 5816 }, { "epoch": 0.0523175025745032, "grad_norm": 1.2134333848953247, "learning_rate": 2.46918611547317e-05, "loss": 2.3706, "step": 5817 }, { "epoch": 0.05232649647214365, "grad_norm": 1.086940050125122, "learning_rate": 2.4684345967787994e-05, "loss": 2.4391, "step": 5818 }, { "epoch": 0.052335490369784105, "grad_norm": 1.5594123601913452, "learning_rate": 2.4676830809372747e-05, "loss": 2.5782, "step": 5819 }, { "epoch": 0.05234448426742455, "grad_norm": 2.591390609741211, "learning_rate": 2.46693156801652e-05, "loss": 2.7295, "step": 5820 }, { "epoch": 0.052353478165065005, "grad_norm": 2.4333863258361816, "learning_rate": 2.466180058084453e-05, "loss": 2.4679, "step": 5821 }, { "epoch": 0.05236247206270545, "grad_norm": 2.063784122467041, "learning_rate": 2.4654285512089975e-05, "loss": 2.7187, "step": 5822 }, { "epoch": 0.052371465960345906, "grad_norm": 1.744113564491272, "learning_rate": 2.464677047458071e-05, "loss": 2.5543, "step": 5823 }, { "epoch": 0.05238045985798635, "grad_norm": 1.6951110363006592, "learning_rate": 2.4639255468995956e-05, "loss": 2.7804, "step": 5824 }, { "epoch": 0.05238945375562681, "grad_norm": 1.3151508569717407, "learning_rate": 2.4631740496014904e-05, "loss": 2.4742, "step": 5825 }, { "epoch": 0.05239844765326726, "grad_norm": 1.6186232566833496, "learning_rate": 2.4624225556316744e-05, "loss": 2.4379, "step": 5826 }, { "epoch": 0.05240744155090771, "grad_norm": 1.4871190786361694, "learning_rate": 2.461671065058068e-05, "loss": 2.4902, "step": 5827 }, { "epoch": 0.05241643544854816, "grad_norm": 1.7747842073440552, "learning_rate": 2.460919577948589e-05, "loss": 2.2593, "step": 5828 }, { "epoch": 0.05242542934618861, "grad_norm": 2.033120632171631, "learning_rate": 2.460168094371157e-05, "loss": 2.8444, "step": 5829 }, { "epoch": 0.05243442324382906, "grad_norm": 1.4245985746383667, "learning_rate": 2.4594166143936898e-05, "loss": 2.448, "step": 5830 }, { "epoch": 0.052443417141469516, "grad_norm": 2.1502089500427246, "learning_rate": 2.458665138084104e-05, "loss": 2.6447, "step": 5831 }, { "epoch": 0.05245241103910996, "grad_norm": 2.0094094276428223, "learning_rate": 2.4579136655103198e-05, "loss": 2.8827, "step": 5832 }, { "epoch": 0.052461404936750416, "grad_norm": 0.9886398315429688, "learning_rate": 2.457162196740252e-05, "loss": 2.3037, "step": 5833 }, { "epoch": 0.05247039883439086, "grad_norm": 1.1315010786056519, "learning_rate": 2.4564107318418198e-05, "loss": 2.2581, "step": 5834 }, { "epoch": 0.05247939273203132, "grad_norm": 2.104717969894409, "learning_rate": 2.455659270882938e-05, "loss": 2.5948, "step": 5835 }, { "epoch": 0.05248838662967177, "grad_norm": 1.8888541460037231, "learning_rate": 2.4549078139315244e-05, "loss": 2.6961, "step": 5836 }, { "epoch": 0.05249738052731222, "grad_norm": 1.6489293575286865, "learning_rate": 2.454156361055494e-05, "loss": 2.6055, "step": 5837 }, { "epoch": 0.05250637442495267, "grad_norm": 1.7375037670135498, "learning_rate": 2.453404912322761e-05, "loss": 2.5755, "step": 5838 }, { "epoch": 0.05251536832259312, "grad_norm": 1.621427059173584, "learning_rate": 2.4526534678012428e-05, "loss": 2.4475, "step": 5839 }, { "epoch": 0.05252436222023357, "grad_norm": 1.2269055843353271, "learning_rate": 2.4519020275588524e-05, "loss": 2.3666, "step": 5840 }, { "epoch": 0.05253335611787402, "grad_norm": 2.151425361633301, "learning_rate": 2.451150591663506e-05, "loss": 2.6665, "step": 5841 }, { "epoch": 0.05254235001551447, "grad_norm": 2.134730339050293, "learning_rate": 2.4503991601831163e-05, "loss": 2.5787, "step": 5842 }, { "epoch": 0.05255134391315493, "grad_norm": 1.956053614616394, "learning_rate": 2.4496477331855967e-05, "loss": 2.9552, "step": 5843 }, { "epoch": 0.052560337810795374, "grad_norm": 1.5520232915878296, "learning_rate": 2.4488963107388613e-05, "loss": 2.7876, "step": 5844 }, { "epoch": 0.05256933170843583, "grad_norm": 1.6283907890319824, "learning_rate": 2.448144892910821e-05, "loss": 2.8355, "step": 5845 }, { "epoch": 0.052578325606076275, "grad_norm": 0.6182486414909363, "learning_rate": 2.4473934797693908e-05, "loss": 2.2052, "step": 5846 }, { "epoch": 0.05258731950371673, "grad_norm": 1.3155750036239624, "learning_rate": 2.44664207138248e-05, "loss": 2.3943, "step": 5847 }, { "epoch": 0.05259631340135718, "grad_norm": 1.6774827241897583, "learning_rate": 2.4458906678180023e-05, "loss": 2.6037, "step": 5848 }, { "epoch": 0.05260530729899763, "grad_norm": 1.488465428352356, "learning_rate": 2.445139269143868e-05, "loss": 2.6045, "step": 5849 }, { "epoch": 0.05261430119663808, "grad_norm": 1.7021033763885498, "learning_rate": 2.4443878754279866e-05, "loss": 2.4099, "step": 5850 }, { "epoch": 0.05262329509427853, "grad_norm": 1.801295518875122, "learning_rate": 2.44363648673827e-05, "loss": 2.7676, "step": 5851 }, { "epoch": 0.052632288991918984, "grad_norm": 1.4828482866287231, "learning_rate": 2.442885103142626e-05, "loss": 2.488, "step": 5852 }, { "epoch": 0.05264128288955943, "grad_norm": 1.5936075448989868, "learning_rate": 2.4421337247089655e-05, "loss": 2.5824, "step": 5853 }, { "epoch": 0.052650276787199884, "grad_norm": 1.9260646104812622, "learning_rate": 2.441382351505197e-05, "loss": 2.6653, "step": 5854 }, { "epoch": 0.05265927068484034, "grad_norm": 1.4796940088272095, "learning_rate": 2.440630983599227e-05, "loss": 2.4685, "step": 5855 }, { "epoch": 0.052668264582480785, "grad_norm": 1.2301009893417358, "learning_rate": 2.439879621058966e-05, "loss": 2.527, "step": 5856 }, { "epoch": 0.05267725848012124, "grad_norm": 1.7434604167938232, "learning_rate": 2.43912826395232e-05, "loss": 2.7151, "step": 5857 }, { "epoch": 0.052686252377761686, "grad_norm": 1.520922064781189, "learning_rate": 2.438376912347196e-05, "loss": 2.7587, "step": 5858 }, { "epoch": 0.05269524627540214, "grad_norm": 2.4592230319976807, "learning_rate": 2.4376255663115002e-05, "loss": 2.6749, "step": 5859 }, { "epoch": 0.052704240173042594, "grad_norm": 1.344713807106018, "learning_rate": 2.4368742259131378e-05, "loss": 2.4975, "step": 5860 }, { "epoch": 0.05271323407068304, "grad_norm": 1.603704571723938, "learning_rate": 2.4361228912200155e-05, "loss": 2.4625, "step": 5861 }, { "epoch": 0.052722227968323494, "grad_norm": 1.7258961200714111, "learning_rate": 2.435371562300037e-05, "loss": 2.4891, "step": 5862 }, { "epoch": 0.05273122186596394, "grad_norm": 1.6150670051574707, "learning_rate": 2.4346202392211077e-05, "loss": 2.679, "step": 5863 }, { "epoch": 0.052740215763604395, "grad_norm": 1.5912383794784546, "learning_rate": 2.4338689220511304e-05, "loss": 2.5494, "step": 5864 }, { "epoch": 0.05274920966124485, "grad_norm": 1.9099267721176147, "learning_rate": 2.4331176108580092e-05, "loss": 2.7518, "step": 5865 }, { "epoch": 0.052758203558885296, "grad_norm": 1.8701752424240112, "learning_rate": 2.4323663057096463e-05, "loss": 2.6189, "step": 5866 }, { "epoch": 0.05276719745652575, "grad_norm": 1.9831511974334717, "learning_rate": 2.4316150066739426e-05, "loss": 2.851, "step": 5867 }, { "epoch": 0.052776191354166196, "grad_norm": 2.082138776779175, "learning_rate": 2.4308637138188022e-05, "loss": 2.7568, "step": 5868 }, { "epoch": 0.05278518525180665, "grad_norm": 1.3364025354385376, "learning_rate": 2.4301124272121246e-05, "loss": 2.4621, "step": 5869 }, { "epoch": 0.0527941791494471, "grad_norm": 2.555736541748047, "learning_rate": 2.4293611469218107e-05, "loss": 2.7208, "step": 5870 }, { "epoch": 0.05280317304708755, "grad_norm": 1.7353935241699219, "learning_rate": 2.42860987301576e-05, "loss": 2.5661, "step": 5871 }, { "epoch": 0.052812166944728005, "grad_norm": 1.6793981790542603, "learning_rate": 2.4278586055618714e-05, "loss": 2.4781, "step": 5872 }, { "epoch": 0.05282116084236845, "grad_norm": 1.3408640623092651, "learning_rate": 2.4271073446280455e-05, "loss": 2.3716, "step": 5873 }, { "epoch": 0.052830154740008906, "grad_norm": 1.790439248085022, "learning_rate": 2.4263560902821783e-05, "loss": 2.8595, "step": 5874 }, { "epoch": 0.05283914863764935, "grad_norm": 1.4342747926712036, "learning_rate": 2.425604842592169e-05, "loss": 2.672, "step": 5875 }, { "epoch": 0.052848142535289806, "grad_norm": 1.777701497077942, "learning_rate": 2.4248536016259135e-05, "loss": 2.5992, "step": 5876 }, { "epoch": 0.05285713643293026, "grad_norm": 2.306591272354126, "learning_rate": 2.424102367451309e-05, "loss": 2.7793, "step": 5877 }, { "epoch": 0.05286613033057071, "grad_norm": 2.292839288711548, "learning_rate": 2.4233511401362514e-05, "loss": 2.6517, "step": 5878 }, { "epoch": 0.05287512422821116, "grad_norm": 1.8292174339294434, "learning_rate": 2.4225999197486333e-05, "loss": 2.48, "step": 5879 }, { "epoch": 0.05288411812585161, "grad_norm": 1.1913433074951172, "learning_rate": 2.4218487063563524e-05, "loss": 2.5051, "step": 5880 }, { "epoch": 0.05289311202349206, "grad_norm": 1.3367847204208374, "learning_rate": 2.4210975000273003e-05, "loss": 2.5479, "step": 5881 }, { "epoch": 0.05290210592113251, "grad_norm": 1.8221735954284668, "learning_rate": 2.420346300829372e-05, "loss": 2.6028, "step": 5882 }, { "epoch": 0.05291109981877296, "grad_norm": 1.7667906284332275, "learning_rate": 2.4195951088304597e-05, "loss": 2.5152, "step": 5883 }, { "epoch": 0.052920093716413416, "grad_norm": 1.725054383277893, "learning_rate": 2.418843924098454e-05, "loss": 2.516, "step": 5884 }, { "epoch": 0.05292908761405386, "grad_norm": 1.4462120532989502, "learning_rate": 2.4180927467012477e-05, "loss": 2.4635, "step": 5885 }, { "epoch": 0.05293808151169432, "grad_norm": 1.3528188467025757, "learning_rate": 2.4173415767067297e-05, "loss": 2.4384, "step": 5886 }, { "epoch": 0.052947075409334764, "grad_norm": 2.1253302097320557, "learning_rate": 2.4165904141827918e-05, "loss": 2.6494, "step": 5887 }, { "epoch": 0.05295606930697522, "grad_norm": 1.7043321132659912, "learning_rate": 2.4158392591973214e-05, "loss": 2.7826, "step": 5888 }, { "epoch": 0.05296506320461567, "grad_norm": 1.8390326499938965, "learning_rate": 2.4150881118182093e-05, "loss": 2.6041, "step": 5889 }, { "epoch": 0.05297405710225612, "grad_norm": 2.6684787273406982, "learning_rate": 2.4143369721133418e-05, "loss": 2.7261, "step": 5890 }, { "epoch": 0.05298305099989657, "grad_norm": 1.7630671262741089, "learning_rate": 2.4135858401506064e-05, "loss": 2.5052, "step": 5891 }, { "epoch": 0.05299204489753702, "grad_norm": 1.0537275075912476, "learning_rate": 2.4128347159978902e-05, "loss": 2.4016, "step": 5892 }, { "epoch": 0.05300103879517747, "grad_norm": 1.7829983234405518, "learning_rate": 2.4120835997230775e-05, "loss": 2.5298, "step": 5893 }, { "epoch": 0.05301003269281792, "grad_norm": 1.4010602235794067, "learning_rate": 2.4113324913940548e-05, "loss": 2.5024, "step": 5894 }, { "epoch": 0.053019026590458374, "grad_norm": 2.780203104019165, "learning_rate": 2.410581391078707e-05, "loss": 2.9008, "step": 5895 }, { "epoch": 0.05302802048809883, "grad_norm": 1.4883394241333008, "learning_rate": 2.4098302988449148e-05, "loss": 2.3594, "step": 5896 }, { "epoch": 0.053037014385739274, "grad_norm": 1.904987096786499, "learning_rate": 2.4090792147605647e-05, "loss": 2.5817, "step": 5897 }, { "epoch": 0.05304600828337973, "grad_norm": 1.8060717582702637, "learning_rate": 2.4083281388935363e-05, "loss": 2.5141, "step": 5898 }, { "epoch": 0.053055002181020175, "grad_norm": 1.814956545829773, "learning_rate": 2.4075770713117125e-05, "loss": 2.6086, "step": 5899 }, { "epoch": 0.05306399607866063, "grad_norm": 1.6023788452148438, "learning_rate": 2.4068260120829726e-05, "loss": 2.5393, "step": 5900 }, { "epoch": 0.05307298997630108, "grad_norm": 2.0525007247924805, "learning_rate": 2.4060749612751988e-05, "loss": 2.357, "step": 5901 }, { "epoch": 0.05308198387394153, "grad_norm": 1.9559475183486938, "learning_rate": 2.4053239189562682e-05, "loss": 2.5799, "step": 5902 }, { "epoch": 0.05309097777158198, "grad_norm": 2.085028648376465, "learning_rate": 2.404572885194059e-05, "loss": 2.8248, "step": 5903 }, { "epoch": 0.05309997166922243, "grad_norm": 1.8274368047714233, "learning_rate": 2.403821860056451e-05, "loss": 2.5292, "step": 5904 }, { "epoch": 0.053108965566862884, "grad_norm": 1.3587042093276978, "learning_rate": 2.4030708436113187e-05, "loss": 2.7148, "step": 5905 }, { "epoch": 0.05311795946450334, "grad_norm": 2.056291103363037, "learning_rate": 2.4023198359265398e-05, "loss": 2.5514, "step": 5906 }, { "epoch": 0.053126953362143785, "grad_norm": 1.5181262493133545, "learning_rate": 2.4015688370699892e-05, "loss": 2.5248, "step": 5907 }, { "epoch": 0.05313594725978424, "grad_norm": 1.4825575351715088, "learning_rate": 2.4008178471095397e-05, "loss": 2.425, "step": 5908 }, { "epoch": 0.053144941157424685, "grad_norm": 2.04168701171875, "learning_rate": 2.4000668661130674e-05, "loss": 2.8616, "step": 5909 }, { "epoch": 0.05315393505506514, "grad_norm": 1.1936098337173462, "learning_rate": 2.399315894148443e-05, "loss": 2.4645, "step": 5910 }, { "epoch": 0.053162928952705586, "grad_norm": 2.1933062076568604, "learning_rate": 2.398564931283541e-05, "loss": 2.5667, "step": 5911 }, { "epoch": 0.05317192285034604, "grad_norm": 1.8604774475097656, "learning_rate": 2.397813977586231e-05, "loss": 2.834, "step": 5912 }, { "epoch": 0.053180916747986494, "grad_norm": 2.0463852882385254, "learning_rate": 2.3970630331243834e-05, "loss": 2.8839, "step": 5913 }, { "epoch": 0.05318991064562694, "grad_norm": 1.821907877922058, "learning_rate": 2.3963120979658682e-05, "loss": 2.6721, "step": 5914 }, { "epoch": 0.053198904543267395, "grad_norm": 1.5905489921569824, "learning_rate": 2.395561172178553e-05, "loss": 2.6021, "step": 5915 }, { "epoch": 0.05320789844090784, "grad_norm": 2.4909250736236572, "learning_rate": 2.3948102558303076e-05, "loss": 2.8578, "step": 5916 }, { "epoch": 0.053216892338548295, "grad_norm": 1.4876762628555298, "learning_rate": 2.3940593489889972e-05, "loss": 2.4682, "step": 5917 }, { "epoch": 0.05322588623618875, "grad_norm": 1.8617318868637085, "learning_rate": 2.3933084517224893e-05, "loss": 2.7955, "step": 5918 }, { "epoch": 0.053234880133829196, "grad_norm": 1.9057890176773071, "learning_rate": 2.392557564098649e-05, "loss": 2.9166, "step": 5919 }, { "epoch": 0.05324387403146965, "grad_norm": 1.8451768159866333, "learning_rate": 2.3918066861853396e-05, "loss": 2.5929, "step": 5920 }, { "epoch": 0.0532528679291101, "grad_norm": 0.9136981964111328, "learning_rate": 2.391055818050426e-05, "loss": 2.2133, "step": 5921 }, { "epoch": 0.05326186182675055, "grad_norm": 1.1256299018859863, "learning_rate": 2.3903049597617688e-05, "loss": 2.429, "step": 5922 }, { "epoch": 0.053270855724391, "grad_norm": 0.9792494177818298, "learning_rate": 2.389554111387233e-05, "loss": 2.0884, "step": 5923 }, { "epoch": 0.05327984962203145, "grad_norm": 1.219143033027649, "learning_rate": 2.388803272994677e-05, "loss": 2.2354, "step": 5924 }, { "epoch": 0.053288843519671905, "grad_norm": 2.56982159614563, "learning_rate": 2.3880524446519603e-05, "loss": 2.9676, "step": 5925 }, { "epoch": 0.05329783741731235, "grad_norm": 1.4375818967819214, "learning_rate": 2.387301626426944e-05, "loss": 2.6236, "step": 5926 }, { "epoch": 0.053306831314952806, "grad_norm": 1.695178747177124, "learning_rate": 2.3865508183874853e-05, "loss": 2.4137, "step": 5927 }, { "epoch": 0.05331582521259325, "grad_norm": 1.115891933441162, "learning_rate": 2.385800020601442e-05, "loss": 2.4572, "step": 5928 }, { "epoch": 0.05332481911023371, "grad_norm": 1.4353976249694824, "learning_rate": 2.3850492331366685e-05, "loss": 2.4834, "step": 5929 }, { "epoch": 0.05333381300787416, "grad_norm": 2.0355355739593506, "learning_rate": 2.3842984560610228e-05, "loss": 2.7798, "step": 5930 }, { "epoch": 0.05334280690551461, "grad_norm": 1.5113332271575928, "learning_rate": 2.3835476894423577e-05, "loss": 2.6061, "step": 5931 }, { "epoch": 0.05335180080315506, "grad_norm": 1.3468754291534424, "learning_rate": 2.382796933348527e-05, "loss": 2.3744, "step": 5932 }, { "epoch": 0.05336079470079551, "grad_norm": 1.533708930015564, "learning_rate": 2.3820461878473828e-05, "loss": 2.5239, "step": 5933 }, { "epoch": 0.05336978859843596, "grad_norm": 1.9419997930526733, "learning_rate": 2.3812954530067768e-05, "loss": 2.4396, "step": 5934 }, { "epoch": 0.05337878249607641, "grad_norm": 2.2364251613616943, "learning_rate": 2.3805447288945608e-05, "loss": 2.8952, "step": 5935 }, { "epoch": 0.05338777639371686, "grad_norm": 1.7243225574493408, "learning_rate": 2.3797940155785836e-05, "loss": 2.5727, "step": 5936 }, { "epoch": 0.053396770291357316, "grad_norm": 0.9541964530944824, "learning_rate": 2.3790433131266927e-05, "loss": 2.3676, "step": 5937 }, { "epoch": 0.05340576418899776, "grad_norm": 1.5545345544815063, "learning_rate": 2.3782926216067378e-05, "loss": 2.631, "step": 5938 }, { "epoch": 0.05341475808663822, "grad_norm": 1.425553560256958, "learning_rate": 2.377541941086564e-05, "loss": 2.5422, "step": 5939 }, { "epoch": 0.053423751984278664, "grad_norm": 1.5131886005401611, "learning_rate": 2.3767912716340188e-05, "loss": 2.5225, "step": 5940 }, { "epoch": 0.05343274588191912, "grad_norm": 2.108018636703491, "learning_rate": 2.3760406133169443e-05, "loss": 2.5807, "step": 5941 }, { "epoch": 0.05344173977955957, "grad_norm": 1.6878293752670288, "learning_rate": 2.3752899662031866e-05, "loss": 2.6418, "step": 5942 }, { "epoch": 0.05345073367720002, "grad_norm": 1.5397592782974243, "learning_rate": 2.3745393303605877e-05, "loss": 2.461, "step": 5943 }, { "epoch": 0.05345972757484047, "grad_norm": 1.3599765300750732, "learning_rate": 2.3737887058569878e-05, "loss": 2.4714, "step": 5944 }, { "epoch": 0.05346872147248092, "grad_norm": 1.943172574043274, "learning_rate": 2.37303809276023e-05, "loss": 2.6791, "step": 5945 }, { "epoch": 0.05347771537012137, "grad_norm": 1.4677302837371826, "learning_rate": 2.3722874911381517e-05, "loss": 2.6196, "step": 5946 }, { "epoch": 0.05348670926776183, "grad_norm": 1.4859697818756104, "learning_rate": 2.3715369010585928e-05, "loss": 2.4117, "step": 5947 }, { "epoch": 0.053495703165402274, "grad_norm": 1.688707947731018, "learning_rate": 2.3707863225893905e-05, "loss": 2.7533, "step": 5948 }, { "epoch": 0.05350469706304273, "grad_norm": 2.4043705463409424, "learning_rate": 2.3700357557983804e-05, "loss": 2.8938, "step": 5949 }, { "epoch": 0.053513690960683175, "grad_norm": 1.2930631637573242, "learning_rate": 2.3692852007533996e-05, "loss": 2.2851, "step": 5950 }, { "epoch": 0.05352268485832363, "grad_norm": 1.6568694114685059, "learning_rate": 2.368534657522281e-05, "loss": 2.5707, "step": 5951 }, { "epoch": 0.053531678755964075, "grad_norm": 1.541953444480896, "learning_rate": 2.367784126172859e-05, "loss": 2.3715, "step": 5952 }, { "epoch": 0.05354067265360453, "grad_norm": 1.9886503219604492, "learning_rate": 2.3670336067729647e-05, "loss": 2.8707, "step": 5953 }, { "epoch": 0.05354966655124498, "grad_norm": 1.7902733087539673, "learning_rate": 2.36628309939043e-05, "loss": 2.5852, "step": 5954 }, { "epoch": 0.05355866044888543, "grad_norm": 1.9307847023010254, "learning_rate": 2.3655326040930856e-05, "loss": 2.6769, "step": 5955 }, { "epoch": 0.053567654346525884, "grad_norm": 2.917771816253662, "learning_rate": 2.364782120948758e-05, "loss": 2.9109, "step": 5956 }, { "epoch": 0.05357664824416633, "grad_norm": 1.4667404890060425, "learning_rate": 2.3640316500252778e-05, "loss": 2.4514, "step": 5957 }, { "epoch": 0.053585642141806784, "grad_norm": 1.4749462604522705, "learning_rate": 2.36328119139047e-05, "loss": 2.5519, "step": 5958 }, { "epoch": 0.05359463603944724, "grad_norm": 1.4232072830200195, "learning_rate": 2.3625307451121614e-05, "loss": 2.4205, "step": 5959 }, { "epoch": 0.053603629937087685, "grad_norm": 2.134577751159668, "learning_rate": 2.3617803112581762e-05, "loss": 2.5879, "step": 5960 }, { "epoch": 0.05361262383472814, "grad_norm": 1.5492606163024902, "learning_rate": 2.361029889896337e-05, "loss": 2.6368, "step": 5961 }, { "epoch": 0.053621617732368586, "grad_norm": 1.6439663171768188, "learning_rate": 2.3602794810944672e-05, "loss": 2.6189, "step": 5962 }, { "epoch": 0.05363061163000904, "grad_norm": 1.1377358436584473, "learning_rate": 2.3595290849203862e-05, "loss": 2.3046, "step": 5963 }, { "epoch": 0.053639605527649487, "grad_norm": 1.9758307933807373, "learning_rate": 2.3587787014419166e-05, "loss": 2.8006, "step": 5964 }, { "epoch": 0.05364859942528994, "grad_norm": 1.5293102264404297, "learning_rate": 2.3580283307268746e-05, "loss": 2.5253, "step": 5965 }, { "epoch": 0.053657593322930394, "grad_norm": 1.6428358554840088, "learning_rate": 2.35727797284308e-05, "loss": 2.5198, "step": 5966 }, { "epoch": 0.05366658722057084, "grad_norm": 1.772044062614441, "learning_rate": 2.3565276278583485e-05, "loss": 2.3332, "step": 5967 }, { "epoch": 0.053675581118211295, "grad_norm": 1.6230669021606445, "learning_rate": 2.3557772958404952e-05, "loss": 2.5023, "step": 5968 }, { "epoch": 0.05368457501585174, "grad_norm": 1.7991867065429688, "learning_rate": 2.355026976857335e-05, "loss": 2.5798, "step": 5969 }, { "epoch": 0.053693568913492196, "grad_norm": 2.6399035453796387, "learning_rate": 2.3542766709766794e-05, "loss": 2.7353, "step": 5970 }, { "epoch": 0.05370256281113265, "grad_norm": 1.8663451671600342, "learning_rate": 2.3535263782663424e-05, "loss": 2.9252, "step": 5971 }, { "epoch": 0.053711556708773096, "grad_norm": 2.2301018238067627, "learning_rate": 2.352776098794134e-05, "loss": 2.4252, "step": 5972 }, { "epoch": 0.05372055060641355, "grad_norm": 1.646695852279663, "learning_rate": 2.3520258326278617e-05, "loss": 2.5929, "step": 5973 }, { "epoch": 0.053729544504054, "grad_norm": 1.9013340473175049, "learning_rate": 2.351275579835336e-05, "loss": 2.8534, "step": 5974 }, { "epoch": 0.05373853840169445, "grad_norm": 1.9206992387771606, "learning_rate": 2.350525340484363e-05, "loss": 2.6478, "step": 5975 }, { "epoch": 0.053747532299334905, "grad_norm": 1.996207594871521, "learning_rate": 2.3497751146427493e-05, "loss": 2.4125, "step": 5976 }, { "epoch": 0.05375652619697535, "grad_norm": 1.551431655883789, "learning_rate": 2.349024902378298e-05, "loss": 2.5076, "step": 5977 }, { "epoch": 0.053765520094615805, "grad_norm": 1.4903490543365479, "learning_rate": 2.3482747037588144e-05, "loss": 2.5088, "step": 5978 }, { "epoch": 0.05377451399225625, "grad_norm": 1.543755292892456, "learning_rate": 2.3475245188520994e-05, "loss": 2.6823, "step": 5979 }, { "epoch": 0.053783507889896706, "grad_norm": 1.6668524742126465, "learning_rate": 2.346774347725953e-05, "loss": 2.4824, "step": 5980 }, { "epoch": 0.05379250178753715, "grad_norm": 1.8498460054397583, "learning_rate": 2.3460241904481776e-05, "loss": 2.7478, "step": 5981 }, { "epoch": 0.05380149568517761, "grad_norm": 1.647156000137329, "learning_rate": 2.3452740470865693e-05, "loss": 2.4817, "step": 5982 }, { "epoch": 0.05381048958281806, "grad_norm": 1.6297043561935425, "learning_rate": 2.3445239177089263e-05, "loss": 2.4954, "step": 5983 }, { "epoch": 0.05381948348045851, "grad_norm": 1.443075180053711, "learning_rate": 2.3437738023830443e-05, "loss": 2.5264, "step": 5984 }, { "epoch": 0.05382847737809896, "grad_norm": 1.8748619556427002, "learning_rate": 2.3430237011767167e-05, "loss": 2.85, "step": 5985 }, { "epoch": 0.05383747127573941, "grad_norm": 1.3310132026672363, "learning_rate": 2.342273614157739e-05, "loss": 2.4855, "step": 5986 }, { "epoch": 0.05384646517337986, "grad_norm": 1.545436143875122, "learning_rate": 2.3415235413939013e-05, "loss": 2.5644, "step": 5987 }, { "epoch": 0.053855459071020316, "grad_norm": 1.3878560066223145, "learning_rate": 2.3407734829529956e-05, "loss": 2.7063, "step": 5988 }, { "epoch": 0.05386445296866076, "grad_norm": 1.8372458219528198, "learning_rate": 2.3400234389028107e-05, "loss": 2.7, "step": 5989 }, { "epoch": 0.05387344686630122, "grad_norm": 2.150509834289551, "learning_rate": 2.339273409311136e-05, "loss": 2.9464, "step": 5990 }, { "epoch": 0.053882440763941664, "grad_norm": 1.7004873752593994, "learning_rate": 2.3385233942457573e-05, "loss": 2.6769, "step": 5991 }, { "epoch": 0.05389143466158212, "grad_norm": 1.6265954971313477, "learning_rate": 2.337773393774459e-05, "loss": 2.5753, "step": 5992 }, { "epoch": 0.053900428559222564, "grad_norm": 1.368878722190857, "learning_rate": 2.337023407965028e-05, "loss": 2.5475, "step": 5993 }, { "epoch": 0.05390942245686302, "grad_norm": 1.2708847522735596, "learning_rate": 2.3362734368852452e-05, "loss": 2.4368, "step": 5994 }, { "epoch": 0.05391841635450347, "grad_norm": 0.6865407228469849, "learning_rate": 2.3355234806028935e-05, "loss": 2.1745, "step": 5995 }, { "epoch": 0.05392741025214392, "grad_norm": 2.3195066452026367, "learning_rate": 2.334773539185752e-05, "loss": 2.8863, "step": 5996 }, { "epoch": 0.05393640414978437, "grad_norm": 1.5505002737045288, "learning_rate": 2.3340236127015998e-05, "loss": 2.4555, "step": 5997 }, { "epoch": 0.05394539804742482, "grad_norm": 1.5066300630569458, "learning_rate": 2.3332737012182152e-05, "loss": 2.6699, "step": 5998 }, { "epoch": 0.05395439194506527, "grad_norm": 1.2803144454956055, "learning_rate": 2.332523804803373e-05, "loss": 2.3748, "step": 5999 }, { "epoch": 0.05396338584270573, "grad_norm": 1.3652963638305664, "learning_rate": 2.33177392352485e-05, "loss": 2.4638, "step": 6000 }, { "epoch": 0.053972379740346174, "grad_norm": 8.561768531799316, "learning_rate": 2.3310240574504185e-05, "loss": 2.5795, "step": 6001 }, { "epoch": 0.05398137363798663, "grad_norm": 30.444522857666016, "learning_rate": 2.33027420664785e-05, "loss": 2.3517, "step": 6002 }, { "epoch": 0.053990367535627075, "grad_norm": 46.808650970458984, "learning_rate": 2.3295243711849167e-05, "loss": 2.7977, "step": 6003 }, { "epoch": 0.05399936143326753, "grad_norm": 26.58209228515625, "learning_rate": 2.3287745511293858e-05, "loss": 2.7344, "step": 6004 }, { "epoch": 0.054008355330907976, "grad_norm": 1.3081345558166504, "learning_rate": 2.3280247465490276e-05, "loss": 2.4311, "step": 6005 }, { "epoch": 0.05401734922854843, "grad_norm": 66.0512924194336, "learning_rate": 2.3272749575116068e-05, "loss": 2.3109, "step": 6006 }, { "epoch": 0.05402634312618888, "grad_norm": 50.3670654296875, "learning_rate": 2.32652518408489e-05, "loss": 2.0634, "step": 6007 }, { "epoch": 0.05403533702382933, "grad_norm": 21.430620193481445, "learning_rate": 2.3257754263366406e-05, "loss": 2.0925, "step": 6008 }, { "epoch": 0.054044330921469784, "grad_norm": 17.556602478027344, "learning_rate": 2.3250256843346195e-05, "loss": 1.2633, "step": 6009 }, { "epoch": 0.05405332481911023, "grad_norm": 12.050970077514648, "learning_rate": 2.3242759581465897e-05, "loss": 1.9548, "step": 6010 }, { "epoch": 0.054062318716750685, "grad_norm": 26.461490631103516, "learning_rate": 2.323526247840308e-05, "loss": 1.4213, "step": 6011 }, { "epoch": 0.05407131261439114, "grad_norm": 31.688602447509766, "learning_rate": 2.3227765534835355e-05, "loss": 1.1326, "step": 6012 }, { "epoch": 0.054080306512031585, "grad_norm": 26.472997665405273, "learning_rate": 2.3220268751440266e-05, "loss": 1.5705, "step": 6013 }, { "epoch": 0.05408930040967204, "grad_norm": 31.002883911132812, "learning_rate": 2.3212772128895364e-05, "loss": 2.6079, "step": 6014 }, { "epoch": 0.054098294307312486, "grad_norm": 13.641185760498047, "learning_rate": 2.3205275667878202e-05, "loss": 2.6369, "step": 6015 }, { "epoch": 0.05410728820495294, "grad_norm": 11.556583404541016, "learning_rate": 2.3197779369066285e-05, "loss": 1.8253, "step": 6016 }, { "epoch": 0.054116282102593394, "grad_norm": 12.298347473144531, "learning_rate": 2.3190283233137132e-05, "loss": 2.5473, "step": 6017 }, { "epoch": 0.05412527600023384, "grad_norm": 10.502002716064453, "learning_rate": 2.3182787260768223e-05, "loss": 2.1875, "step": 6018 }, { "epoch": 0.054134269897874294, "grad_norm": 10.063284873962402, "learning_rate": 2.3175291452637053e-05, "loss": 2.0704, "step": 6019 }, { "epoch": 0.05414326379551474, "grad_norm": 9.092700958251953, "learning_rate": 2.3167795809421078e-05, "loss": 1.5131, "step": 6020 }, { "epoch": 0.054152257693155195, "grad_norm": 8.348368644714355, "learning_rate": 2.3160300331797733e-05, "loss": 2.0864, "step": 6021 }, { "epoch": 0.05416125159079564, "grad_norm": 5.228555202484131, "learning_rate": 2.3152805020444472e-05, "loss": 2.7645, "step": 6022 }, { "epoch": 0.054170245488436096, "grad_norm": 7.453481674194336, "learning_rate": 2.3145309876038698e-05, "loss": 1.433, "step": 6023 }, { "epoch": 0.05417923938607655, "grad_norm": 5.262151718139648, "learning_rate": 2.3137814899257826e-05, "loss": 1.3424, "step": 6024 }, { "epoch": 0.054188233283717, "grad_norm": 8.084968566894531, "learning_rate": 2.313032009077924e-05, "loss": 2.8621, "step": 6025 }, { "epoch": 0.05419722718135745, "grad_norm": 4.899521827697754, "learning_rate": 2.3122825451280296e-05, "loss": 1.1736, "step": 6026 }, { "epoch": 0.0542062210789979, "grad_norm": 5.419795989990234, "learning_rate": 2.3115330981438378e-05, "loss": 2.4588, "step": 6027 }, { "epoch": 0.05421521497663835, "grad_norm": 3.5538246631622314, "learning_rate": 2.3107836681930804e-05, "loss": 1.6232, "step": 6028 }, { "epoch": 0.054224208874278805, "grad_norm": 4.523129940032959, "learning_rate": 2.3100342553434924e-05, "loss": 2.572, "step": 6029 }, { "epoch": 0.05423320277191925, "grad_norm": 6.215458393096924, "learning_rate": 2.309284859662803e-05, "loss": 2.7346, "step": 6030 }, { "epoch": 0.054242196669559706, "grad_norm": 7.305747032165527, "learning_rate": 2.3085354812187433e-05, "loss": 1.7266, "step": 6031 }, { "epoch": 0.05425119056720015, "grad_norm": 7.495891571044922, "learning_rate": 2.307786120079041e-05, "loss": 1.4121, "step": 6032 }, { "epoch": 0.054260184464840606, "grad_norm": 2.5185563564300537, "learning_rate": 2.3070367763114207e-05, "loss": 1.8736, "step": 6033 }, { "epoch": 0.05426917836248105, "grad_norm": 52.50704574584961, "learning_rate": 2.3062874499836095e-05, "loss": 1.2806, "step": 6034 }, { "epoch": 0.05427817226012151, "grad_norm": 8.260412216186523, "learning_rate": 2.3055381411633293e-05, "loss": 2.2876, "step": 6035 }, { "epoch": 0.05428716615776196, "grad_norm": 5.010391712188721, "learning_rate": 2.3047888499183035e-05, "loss": 2.052, "step": 6036 }, { "epoch": 0.05429616005540241, "grad_norm": 7.096980094909668, "learning_rate": 2.304039576316251e-05, "loss": 2.0483, "step": 6037 }, { "epoch": 0.05430515395304286, "grad_norm": 3.580015182495117, "learning_rate": 2.30329032042489e-05, "loss": 2.4286, "step": 6038 }, { "epoch": 0.05431414785068331, "grad_norm": 10.168733596801758, "learning_rate": 2.3025410823119383e-05, "loss": 1.6694, "step": 6039 }, { "epoch": 0.05432314174832376, "grad_norm": 10.835968971252441, "learning_rate": 2.30179186204511e-05, "loss": 1.3768, "step": 6040 }, { "epoch": 0.054332135645964216, "grad_norm": 10.224164009094238, "learning_rate": 2.3010426596921208e-05, "loss": 1.6802, "step": 6041 }, { "epoch": 0.05434112954360466, "grad_norm": 6.913486957550049, "learning_rate": 2.300293475320681e-05, "loss": 1.3605, "step": 6042 }, { "epoch": 0.05435012344124512, "grad_norm": 11.18375301361084, "learning_rate": 2.2995443089985024e-05, "loss": 1.2124, "step": 6043 }, { "epoch": 0.054359117338885564, "grad_norm": 6.755508899688721, "learning_rate": 2.2987951607932934e-05, "loss": 1.9404, "step": 6044 }, { "epoch": 0.05436811123652602, "grad_norm": 4.8031182289123535, "learning_rate": 2.2980460307727607e-05, "loss": 2.5084, "step": 6045 }, { "epoch": 0.054377105134166465, "grad_norm": 14.957075119018555, "learning_rate": 2.2972969190046103e-05, "loss": 0.5642, "step": 6046 }, { "epoch": 0.05438609903180692, "grad_norm": 7.064863681793213, "learning_rate": 2.296547825556545e-05, "loss": 1.9892, "step": 6047 }, { "epoch": 0.05439509292944737, "grad_norm": 8.535187721252441, "learning_rate": 2.2957987504962693e-05, "loss": 1.228, "step": 6048 }, { "epoch": 0.05440408682708782, "grad_norm": 4.162393093109131, "learning_rate": 2.2950496938914828e-05, "loss": 1.9587, "step": 6049 }, { "epoch": 0.05441308072472827, "grad_norm": 3.723822832107544, "learning_rate": 2.294300655809884e-05, "loss": 2.6003, "step": 6050 }, { "epoch": 0.05442207462236872, "grad_norm": 6.49167537689209, "learning_rate": 2.2935516363191693e-05, "loss": 2.2158, "step": 6051 }, { "epoch": 0.054431068520009174, "grad_norm": 2.5646724700927734, "learning_rate": 2.292802635487036e-05, "loss": 2.4706, "step": 6052 }, { "epoch": 0.05444006241764963, "grad_norm": 5.627834796905518, "learning_rate": 2.2920536533811778e-05, "loss": 1.3096, "step": 6053 }, { "epoch": 0.054449056315290074, "grad_norm": 8.793120384216309, "learning_rate": 2.2913046900692852e-05, "loss": 1.1548, "step": 6054 }, { "epoch": 0.05445805021293053, "grad_norm": 2.382612705230713, "learning_rate": 2.2905557456190514e-05, "loss": 2.5824, "step": 6055 }, { "epoch": 0.054467044110570975, "grad_norm": 4.64486837387085, "learning_rate": 2.2898068200981636e-05, "loss": 2.261, "step": 6056 }, { "epoch": 0.05447603800821143, "grad_norm": 8.325465202331543, "learning_rate": 2.289057913574308e-05, "loss": 1.4336, "step": 6057 }, { "epoch": 0.05448503190585188, "grad_norm": 3.936396837234497, "learning_rate": 2.2883090261151717e-05, "loss": 1.5706, "step": 6058 }, { "epoch": 0.05449402580349233, "grad_norm": 12.387532234191895, "learning_rate": 2.2875601577884367e-05, "loss": 1.3432, "step": 6059 }, { "epoch": 0.054503019701132784, "grad_norm": 6.321553707122803, "learning_rate": 2.286811308661787e-05, "loss": 1.6034, "step": 6060 }, { "epoch": 0.05451201359877323, "grad_norm": 3.134871244430542, "learning_rate": 2.2860624788029013e-05, "loss": 1.8539, "step": 6061 }, { "epoch": 0.054521007496413684, "grad_norm": 6.100884914398193, "learning_rate": 2.2853136682794575e-05, "loss": 1.2591, "step": 6062 }, { "epoch": 0.05453000139405413, "grad_norm": 3.9176132678985596, "learning_rate": 2.284564877159134e-05, "loss": 1.7938, "step": 6063 }, { "epoch": 0.054538995291694585, "grad_norm": 7.112562656402588, "learning_rate": 2.2838161055096045e-05, "loss": 0.3987, "step": 6064 }, { "epoch": 0.05454798918933504, "grad_norm": 3.6403708457946777, "learning_rate": 2.2830673533985427e-05, "loss": 2.6434, "step": 6065 }, { "epoch": 0.054556983086975486, "grad_norm": 5.859287738800049, "learning_rate": 2.2823186208936205e-05, "loss": 1.78, "step": 6066 }, { "epoch": 0.05456597698461594, "grad_norm": 3.2749979496002197, "learning_rate": 2.281569908062505e-05, "loss": 1.4955, "step": 6067 }, { "epoch": 0.054574970882256386, "grad_norm": 9.30294418334961, "learning_rate": 2.2808212149728675e-05, "loss": 0.6988, "step": 6068 }, { "epoch": 0.05458396477989684, "grad_norm": 3.409794807434082, "learning_rate": 2.2800725416923713e-05, "loss": 1.696, "step": 6069 }, { "epoch": 0.054592958677537294, "grad_norm": 12.415543556213379, "learning_rate": 2.279323888288683e-05, "loss": 0.9052, "step": 6070 }, { "epoch": 0.05460195257517774, "grad_norm": 4.216876029968262, "learning_rate": 2.2785752548294635e-05, "loss": 2.068, "step": 6071 }, { "epoch": 0.054610946472818195, "grad_norm": 25.05821990966797, "learning_rate": 2.2778266413823746e-05, "loss": 0.9441, "step": 6072 }, { "epoch": 0.05461994037045864, "grad_norm": 6.035775661468506, "learning_rate": 2.2770780480150744e-05, "loss": 1.4896, "step": 6073 }, { "epoch": 0.054628934268099096, "grad_norm": 3.6467649936676025, "learning_rate": 2.276329474795219e-05, "loss": 1.9379, "step": 6074 }, { "epoch": 0.05463792816573954, "grad_norm": 9.076715469360352, "learning_rate": 2.275580921790466e-05, "loss": 0.6178, "step": 6075 }, { "epoch": 0.054646922063379996, "grad_norm": 5.287787437438965, "learning_rate": 2.2748323890684665e-05, "loss": 2.1115, "step": 6076 }, { "epoch": 0.05465591596102045, "grad_norm": 3.4206957817077637, "learning_rate": 2.2740838766968738e-05, "loss": 2.0636, "step": 6077 }, { "epoch": 0.0546649098586609, "grad_norm": 2.805154323577881, "learning_rate": 2.2733353847433377e-05, "loss": 1.8488, "step": 6078 }, { "epoch": 0.05467390375630135, "grad_norm": 3.129277467727661, "learning_rate": 2.272586913275504e-05, "loss": 1.9764, "step": 6079 }, { "epoch": 0.0546828976539418, "grad_norm": 5.055434226989746, "learning_rate": 2.2718384623610212e-05, "loss": 0.7067, "step": 6080 }, { "epoch": 0.05469189155158225, "grad_norm": 2.2621185779571533, "learning_rate": 2.2710900320675313e-05, "loss": 2.5273, "step": 6081 }, { "epoch": 0.054700885449222705, "grad_norm": 3.5199007987976074, "learning_rate": 2.2703416224626785e-05, "loss": 1.1777, "step": 6082 }, { "epoch": 0.05470987934686315, "grad_norm": 3.0664474964141846, "learning_rate": 2.269593233614102e-05, "loss": 2.633, "step": 6083 }, { "epoch": 0.054718873244503606, "grad_norm": 4.919866561889648, "learning_rate": 2.2688448655894415e-05, "loss": 2.0408, "step": 6084 }, { "epoch": 0.05472786714214405, "grad_norm": 2.315762758255005, "learning_rate": 2.268096518456333e-05, "loss": 2.3987, "step": 6085 }, { "epoch": 0.05473686103978451, "grad_norm": 3.88278865814209, "learning_rate": 2.267348192282411e-05, "loss": 1.8866, "step": 6086 }, { "epoch": 0.05474585493742496, "grad_norm": 3.168590545654297, "learning_rate": 2.2665998871353093e-05, "loss": 2.1341, "step": 6087 }, { "epoch": 0.05475484883506541, "grad_norm": 14.265596389770508, "learning_rate": 2.2658516030826575e-05, "loss": 0.9002, "step": 6088 }, { "epoch": 0.05476384273270586, "grad_norm": 2.022181749343872, "learning_rate": 2.2651033401920865e-05, "loss": 2.6933, "step": 6089 }, { "epoch": 0.05477283663034631, "grad_norm": 1.8882665634155273, "learning_rate": 2.264355098531223e-05, "loss": 2.7517, "step": 6090 }, { "epoch": 0.05478183052798676, "grad_norm": 3.0396976470947266, "learning_rate": 2.2636068781676902e-05, "loss": 1.8161, "step": 6091 }, { "epoch": 0.05479082442562721, "grad_norm": 3.5915870666503906, "learning_rate": 2.2628586791691146e-05, "loss": 1.3672, "step": 6092 }, { "epoch": 0.05479981832326766, "grad_norm": 5.994241237640381, "learning_rate": 2.2621105016031155e-05, "loss": 0.8143, "step": 6093 }, { "epoch": 0.05480881222090812, "grad_norm": 2.4487392902374268, "learning_rate": 2.2613623455373135e-05, "loss": 1.314, "step": 6094 }, { "epoch": 0.054817806118548563, "grad_norm": 3.0818159580230713, "learning_rate": 2.2606142110393247e-05, "loss": 2.3395, "step": 6095 }, { "epoch": 0.05482680001618902, "grad_norm": 2.4132611751556396, "learning_rate": 2.2598660981767668e-05, "loss": 2.2215, "step": 6096 }, { "epoch": 0.054835793913829464, "grad_norm": 16.991966247558594, "learning_rate": 2.259118007017252e-05, "loss": 1.0896, "step": 6097 }, { "epoch": 0.05484478781146992, "grad_norm": 2.3984532356262207, "learning_rate": 2.2583699376283913e-05, "loss": 1.2661, "step": 6098 }, { "epoch": 0.05485378170911037, "grad_norm": 3.3444926738739014, "learning_rate": 2.2576218900777965e-05, "loss": 2.7561, "step": 6099 }, { "epoch": 0.05486277560675082, "grad_norm": 2.729473829269409, "learning_rate": 2.2568738644330736e-05, "loss": 1.4185, "step": 6100 }, { "epoch": 0.05487176950439127, "grad_norm": 3.140753984451294, "learning_rate": 2.2561258607618297e-05, "loss": 1.4368, "step": 6101 }, { "epoch": 0.05488076340203172, "grad_norm": 24.866329193115234, "learning_rate": 2.2553778791316676e-05, "loss": 1.4914, "step": 6102 }, { "epoch": 0.05488975729967217, "grad_norm": 3.92926025390625, "learning_rate": 2.254629919610188e-05, "loss": 1.9254, "step": 6103 }, { "epoch": 0.05489875119731262, "grad_norm": 40.42650604248047, "learning_rate": 2.253881982264994e-05, "loss": 1.1453, "step": 6104 }, { "epoch": 0.054907745094953074, "grad_norm": 2.8445723056793213, "learning_rate": 2.2531340671636793e-05, "loss": 2.1614, "step": 6105 }, { "epoch": 0.05491673899259353, "grad_norm": 3.1292738914489746, "learning_rate": 2.2523861743738434e-05, "loss": 1.6877, "step": 6106 }, { "epoch": 0.054925732890233975, "grad_norm": 2.2905261516571045, "learning_rate": 2.2516383039630775e-05, "loss": 2.1912, "step": 6107 }, { "epoch": 0.05493472678787443, "grad_norm": 16.2624568939209, "learning_rate": 2.2508904559989746e-05, "loss": 1.2539, "step": 6108 }, { "epoch": 0.054943720685514875, "grad_norm": 2.674515724182129, "learning_rate": 2.2501426305491244e-05, "loss": 2.4076, "step": 6109 }, { "epoch": 0.05495271458315533, "grad_norm": 2.3374290466308594, "learning_rate": 2.249394827681113e-05, "loss": 2.4671, "step": 6110 }, { "epoch": 0.05496170848079578, "grad_norm": 4.148759365081787, "learning_rate": 2.2486470474625276e-05, "loss": 0.7887, "step": 6111 }, { "epoch": 0.05497070237843623, "grad_norm": 5.5961198806762695, "learning_rate": 2.2478992899609513e-05, "loss": 1.7441, "step": 6112 }, { "epoch": 0.054979696276076684, "grad_norm": 3.1160314083099365, "learning_rate": 2.2471515552439653e-05, "loss": 1.7323, "step": 6113 }, { "epoch": 0.05498869017371713, "grad_norm": 2.557417154312134, "learning_rate": 2.2464038433791503e-05, "loss": 1.7734, "step": 6114 }, { "epoch": 0.054997684071357585, "grad_norm": 2.7809932231903076, "learning_rate": 2.2456561544340822e-05, "loss": 1.6029, "step": 6115 }, { "epoch": 0.05500667796899803, "grad_norm": 3.519197463989258, "learning_rate": 2.244908488476337e-05, "loss": 2.0611, "step": 6116 }, { "epoch": 0.055015671866638485, "grad_norm": 2.8692901134490967, "learning_rate": 2.2441608455734873e-05, "loss": 1.4219, "step": 6117 }, { "epoch": 0.05502466576427894, "grad_norm": 2.538878917694092, "learning_rate": 2.2434132257931057e-05, "loss": 2.2823, "step": 6118 }, { "epoch": 0.055033659661919386, "grad_norm": 2.155076026916504, "learning_rate": 2.2426656292027598e-05, "loss": 2.1488, "step": 6119 }, { "epoch": 0.05504265355955984, "grad_norm": 3.9612300395965576, "learning_rate": 2.241918055870018e-05, "loss": 1.1275, "step": 6120 }, { "epoch": 0.05505164745720029, "grad_norm": 2.9449501037597656, "learning_rate": 2.241170505862444e-05, "loss": 1.7123, "step": 6121 }, { "epoch": 0.05506064135484074, "grad_norm": 2.2580056190490723, "learning_rate": 2.2404229792475996e-05, "loss": 2.1236, "step": 6122 }, { "epoch": 0.055069635252481194, "grad_norm": 3.2904276847839355, "learning_rate": 2.2396754760930482e-05, "loss": 1.0412, "step": 6123 }, { "epoch": 0.05507862915012164, "grad_norm": 14.750919342041016, "learning_rate": 2.2389279964663457e-05, "loss": 1.5286, "step": 6124 }, { "epoch": 0.055087623047762095, "grad_norm": 23.832317352294922, "learning_rate": 2.238180540435051e-05, "loss": 1.5796, "step": 6125 }, { "epoch": 0.05509661694540254, "grad_norm": 3.8665926456451416, "learning_rate": 2.2374331080667164e-05, "loss": 2.24, "step": 6126 }, { "epoch": 0.055105610843042996, "grad_norm": 3.47674298286438, "learning_rate": 2.2366856994288948e-05, "loss": 2.3177, "step": 6127 }, { "epoch": 0.05511460474068345, "grad_norm": 3.8887908458709717, "learning_rate": 2.2359383145891364e-05, "loss": 1.9745, "step": 6128 }, { "epoch": 0.0551235986383239, "grad_norm": 4.059725761413574, "learning_rate": 2.2351909536149873e-05, "loss": 2.742, "step": 6129 }, { "epoch": 0.05513259253596435, "grad_norm": 4.455172061920166, "learning_rate": 2.234443616573996e-05, "loss": 1.8417, "step": 6130 }, { "epoch": 0.0551415864336048, "grad_norm": 13.044368743896484, "learning_rate": 2.2336963035337037e-05, "loss": 1.1441, "step": 6131 }, { "epoch": 0.05515058033124525, "grad_norm": 2.5078694820404053, "learning_rate": 2.2329490145616534e-05, "loss": 1.9061, "step": 6132 }, { "epoch": 0.0551595742288857, "grad_norm": 2.4225621223449707, "learning_rate": 2.2322017497253833e-05, "loss": 2.7189, "step": 6133 }, { "epoch": 0.05516856812652615, "grad_norm": 2.188112258911133, "learning_rate": 2.2314545090924305e-05, "loss": 2.7843, "step": 6134 }, { "epoch": 0.055177562024166606, "grad_norm": 2.38330078125, "learning_rate": 2.23070729273033e-05, "loss": 1.5193, "step": 6135 }, { "epoch": 0.05518655592180705, "grad_norm": 3.5744378566741943, "learning_rate": 2.229960100706614e-05, "loss": 1.589, "step": 6136 }, { "epoch": 0.055195549819447506, "grad_norm": 15.688020706176758, "learning_rate": 2.2292129330888138e-05, "loss": 1.6368, "step": 6137 }, { "epoch": 0.05520454371708795, "grad_norm": 2.9993014335632324, "learning_rate": 2.228465789944457e-05, "loss": 1.8199, "step": 6138 }, { "epoch": 0.05521353761472841, "grad_norm": 18.3117618560791, "learning_rate": 2.2277186713410687e-05, "loss": 1.5702, "step": 6139 }, { "epoch": 0.05522253151236886, "grad_norm": 3.393406391143799, "learning_rate": 2.2269715773461748e-05, "loss": 1.293, "step": 6140 }, { "epoch": 0.05523152541000931, "grad_norm": 20.536762237548828, "learning_rate": 2.2262245080272948e-05, "loss": 1.7487, "step": 6141 }, { "epoch": 0.05524051930764976, "grad_norm": 3.2395663261413574, "learning_rate": 2.2254774634519497e-05, "loss": 1.1567, "step": 6142 }, { "epoch": 0.05524951320529021, "grad_norm": 30.984800338745117, "learning_rate": 2.2247304436876555e-05, "loss": 0.648, "step": 6143 }, { "epoch": 0.05525850710293066, "grad_norm": 3.473153829574585, "learning_rate": 2.2239834488019266e-05, "loss": 1.9446, "step": 6144 }, { "epoch": 0.05526750100057111, "grad_norm": 2.356274127960205, "learning_rate": 2.223236478862277e-05, "loss": 2.8513, "step": 6145 }, { "epoch": 0.05527649489821156, "grad_norm": 2.1269514560699463, "learning_rate": 2.2224895339362152e-05, "loss": 1.9339, "step": 6146 }, { "epoch": 0.05528548879585202, "grad_norm": 3.6906542778015137, "learning_rate": 2.221742614091252e-05, "loss": 1.635, "step": 6147 }, { "epoch": 0.055294482693492464, "grad_norm": 12.786334991455078, "learning_rate": 2.2209957193948906e-05, "loss": 1.2077, "step": 6148 }, { "epoch": 0.05530347659113292, "grad_norm": 6.445119380950928, "learning_rate": 2.2202488499146365e-05, "loss": 1.6834, "step": 6149 }, { "epoch": 0.055312470488773365, "grad_norm": 9.692169189453125, "learning_rate": 2.2195020057179896e-05, "loss": 1.8805, "step": 6150 }, { "epoch": 0.05532146438641382, "grad_norm": 3.338568687438965, "learning_rate": 2.2187551868724485e-05, "loss": 1.9701, "step": 6151 }, { "epoch": 0.05533045828405427, "grad_norm": 20.414819717407227, "learning_rate": 2.2180083934455118e-05, "loss": 0.574, "step": 6152 }, { "epoch": 0.05533945218169472, "grad_norm": 4.726652145385742, "learning_rate": 2.2172616255046715e-05, "loss": 1.9826, "step": 6153 }, { "epoch": 0.05534844607933517, "grad_norm": 7.753073215484619, "learning_rate": 2.216514883117422e-05, "loss": 1.2156, "step": 6154 }, { "epoch": 0.05535743997697562, "grad_norm": 4.65565299987793, "learning_rate": 2.2157681663512523e-05, "loss": 1.2287, "step": 6155 }, { "epoch": 0.055366433874616074, "grad_norm": 2.5945425033569336, "learning_rate": 2.2150214752736488e-05, "loss": 1.8859, "step": 6156 }, { "epoch": 0.05537542777225652, "grad_norm": 3.9225683212280273, "learning_rate": 2.214274809952098e-05, "loss": 1.6138, "step": 6157 }, { "epoch": 0.055384421669896974, "grad_norm": 7.451877117156982, "learning_rate": 2.2135281704540812e-05, "loss": 1.2464, "step": 6158 }, { "epoch": 0.05539341556753743, "grad_norm": 2.7763326168060303, "learning_rate": 2.2127815568470807e-05, "loss": 2.4019, "step": 6159 }, { "epoch": 0.055402409465177875, "grad_norm": 2.7130672931671143, "learning_rate": 2.2120349691985724e-05, "loss": 2.3917, "step": 6160 }, { "epoch": 0.05541140336281833, "grad_norm": 4.72904634475708, "learning_rate": 2.2112884075760347e-05, "loss": 0.7046, "step": 6161 }, { "epoch": 0.055420397260458776, "grad_norm": 4.815637588500977, "learning_rate": 2.21054187204694e-05, "loss": 1.8139, "step": 6162 }, { "epoch": 0.05542939115809923, "grad_norm": 3.030322790145874, "learning_rate": 2.209795362678758e-05, "loss": 2.3998, "step": 6163 }, { "epoch": 0.05543838505573968, "grad_norm": 3.562424898147583, "learning_rate": 2.2090488795389598e-05, "loss": 2.1496, "step": 6164 }, { "epoch": 0.05544737895338013, "grad_norm": 11.08246898651123, "learning_rate": 2.208302422695009e-05, "loss": 0.9906, "step": 6165 }, { "epoch": 0.055456372851020584, "grad_norm": 2.023056745529175, "learning_rate": 2.207555992214372e-05, "loss": 1.4478, "step": 6166 }, { "epoch": 0.05546536674866103, "grad_norm": 19.961692810058594, "learning_rate": 2.2068095881645098e-05, "loss": 1.1133, "step": 6167 }, { "epoch": 0.055474360646301485, "grad_norm": 2.932074785232544, "learning_rate": 2.20606321061288e-05, "loss": 1.315, "step": 6168 }, { "epoch": 0.05548335454394194, "grad_norm": 10.371999740600586, "learning_rate": 2.2053168596269416e-05, "loss": 1.679, "step": 6169 }, { "epoch": 0.055492348441582386, "grad_norm": 3.3435633182525635, "learning_rate": 2.2045705352741475e-05, "loss": 1.6286, "step": 6170 }, { "epoch": 0.05550134233922284, "grad_norm": 2.9308342933654785, "learning_rate": 2.203824237621951e-05, "loss": 2.181, "step": 6171 }, { "epoch": 0.055510336236863286, "grad_norm": 2.9388580322265625, "learning_rate": 2.2030779667377996e-05, "loss": 2.0083, "step": 6172 }, { "epoch": 0.05551933013450374, "grad_norm": 2.917715549468994, "learning_rate": 2.2023317226891425e-05, "loss": 2.1427, "step": 6173 }, { "epoch": 0.05552832403214419, "grad_norm": 2.784933090209961, "learning_rate": 2.201585505543424e-05, "loss": 1.4825, "step": 6174 }, { "epoch": 0.05553731792978464, "grad_norm": 3.748136281967163, "learning_rate": 2.200839315368086e-05, "loss": 1.2718, "step": 6175 }, { "epoch": 0.055546311827425095, "grad_norm": 4.2398481369018555, "learning_rate": 2.200093152230568e-05, "loss": 1.6586, "step": 6176 }, { "epoch": 0.05555530572506554, "grad_norm": 17.677833557128906, "learning_rate": 2.1993470161983078e-05, "loss": 1.6874, "step": 6177 }, { "epoch": 0.055564299622705995, "grad_norm": 2.41558575630188, "learning_rate": 2.198600907338741e-05, "loss": 1.9149, "step": 6178 }, { "epoch": 0.05557329352034644, "grad_norm": 3.2188799381256104, "learning_rate": 2.1978548257192997e-05, "loss": 1.6912, "step": 6179 }, { "epoch": 0.055582287417986896, "grad_norm": 1.9941376447677612, "learning_rate": 2.1971087714074123e-05, "loss": 2.702, "step": 6180 }, { "epoch": 0.05559128131562735, "grad_norm": 19.626028060913086, "learning_rate": 2.1963627444705096e-05, "loss": 1.5656, "step": 6181 }, { "epoch": 0.0556002752132678, "grad_norm": 20.244123458862305, "learning_rate": 2.195616744976014e-05, "loss": 1.3758, "step": 6182 }, { "epoch": 0.05560926911090825, "grad_norm": 2.689199686050415, "learning_rate": 2.19487077299135e-05, "loss": 1.8238, "step": 6183 }, { "epoch": 0.0556182630085487, "grad_norm": 2.916536331176758, "learning_rate": 2.1941248285839356e-05, "loss": 1.6023, "step": 6184 }, { "epoch": 0.05562725690618915, "grad_norm": 6.748956203460693, "learning_rate": 2.1933789118211908e-05, "loss": 1.2312, "step": 6185 }, { "epoch": 0.0556362508038296, "grad_norm": 2.4179911613464355, "learning_rate": 2.1926330227705298e-05, "loss": 1.5009, "step": 6186 }, { "epoch": 0.05564524470147005, "grad_norm": 2.3461201190948486, "learning_rate": 2.191887161499364e-05, "loss": 1.3317, "step": 6187 }, { "epoch": 0.055654238599110506, "grad_norm": 2.5428967475891113, "learning_rate": 2.1911413280751056e-05, "loss": 2.1279, "step": 6188 }, { "epoch": 0.05566323249675095, "grad_norm": 4.420546531677246, "learning_rate": 2.190395522565161e-05, "loss": 1.8772, "step": 6189 }, { "epoch": 0.05567222639439141, "grad_norm": 2.879868984222412, "learning_rate": 2.189649745036936e-05, "loss": 2.2818, "step": 6190 }, { "epoch": 0.055681220292031854, "grad_norm": 2.5673410892486572, "learning_rate": 2.188903995557833e-05, "loss": 2.4746, "step": 6191 }, { "epoch": 0.05569021418967231, "grad_norm": 4.337410926818848, "learning_rate": 2.18815827419525e-05, "loss": 2.479, "step": 6192 }, { "epoch": 0.05569920808731276, "grad_norm": 3.0120127201080322, "learning_rate": 2.1874125810165882e-05, "loss": 1.0206, "step": 6193 }, { "epoch": 0.05570820198495321, "grad_norm": 12.669408798217773, "learning_rate": 2.186666916089239e-05, "loss": 0.5895, "step": 6194 }, { "epoch": 0.05571719588259366, "grad_norm": 12.113533020019531, "learning_rate": 2.1859212794805982e-05, "loss": 1.1895, "step": 6195 }, { "epoch": 0.05572618978023411, "grad_norm": 3.013603925704956, "learning_rate": 2.1851756712580527e-05, "loss": 1.7896, "step": 6196 }, { "epoch": 0.05573518367787456, "grad_norm": 3.3793959617614746, "learning_rate": 2.1844300914889916e-05, "loss": 1.9024, "step": 6197 }, { "epoch": 0.055744177575515017, "grad_norm": 2.9360849857330322, "learning_rate": 2.183684540240799e-05, "loss": 1.7563, "step": 6198 }, { "epoch": 0.05575317147315546, "grad_norm": 2.961888551712036, "learning_rate": 2.1829390175808557e-05, "loss": 1.2812, "step": 6199 }, { "epoch": 0.05576216537079592, "grad_norm": 3.5812339782714844, "learning_rate": 2.182193523576544e-05, "loss": 2.4044, "step": 6200 }, { "epoch": 0.055771159268436364, "grad_norm": 9.988757133483887, "learning_rate": 2.1814480582952375e-05, "loss": 0.7419, "step": 6201 }, { "epoch": 0.05578015316607682, "grad_norm": 4.482209205627441, "learning_rate": 2.180702621804314e-05, "loss": 1.4934, "step": 6202 }, { "epoch": 0.055789147063717265, "grad_norm": 2.4447500705718994, "learning_rate": 2.1799572141711435e-05, "loss": 1.3551, "step": 6203 }, { "epoch": 0.05579814096135772, "grad_norm": 2.571885585784912, "learning_rate": 2.1792118354630947e-05, "loss": 2.4378, "step": 6204 }, { "epoch": 0.05580713485899817, "grad_norm": 2.751645565032959, "learning_rate": 2.1784664857475352e-05, "loss": 2.5431, "step": 6205 }, { "epoch": 0.05581612875663862, "grad_norm": 2.5077931880950928, "learning_rate": 2.1777211650918276e-05, "loss": 2.4171, "step": 6206 }, { "epoch": 0.05582512265427907, "grad_norm": 2.199172258377075, "learning_rate": 2.176975873563335e-05, "loss": 2.4049, "step": 6207 }, { "epoch": 0.05583411655191952, "grad_norm": 8.00373649597168, "learning_rate": 2.1762306112294146e-05, "loss": 1.4579, "step": 6208 }, { "epoch": 0.055843110449559974, "grad_norm": 3.142775058746338, "learning_rate": 2.175485378157422e-05, "loss": 1.5826, "step": 6209 }, { "epoch": 0.05585210434720043, "grad_norm": 3.8785059452056885, "learning_rate": 2.1747401744147125e-05, "loss": 1.0432, "step": 6210 }, { "epoch": 0.055861098244840875, "grad_norm": 3.81628680229187, "learning_rate": 2.1739950000686353e-05, "loss": 1.5247, "step": 6211 }, { "epoch": 0.05587009214248133, "grad_norm": 12.882800102233887, "learning_rate": 2.173249855186539e-05, "loss": 1.2719, "step": 6212 }, { "epoch": 0.055879086040121775, "grad_norm": 19.181909561157227, "learning_rate": 2.172504739835768e-05, "loss": 1.177, "step": 6213 }, { "epoch": 0.05588807993776223, "grad_norm": 2.1506097316741943, "learning_rate": 2.171759654083667e-05, "loss": 2.3512, "step": 6214 }, { "epoch": 0.055897073835402676, "grad_norm": 3.5262041091918945, "learning_rate": 2.1710145979975745e-05, "loss": 0.9197, "step": 6215 }, { "epoch": 0.05590606773304313, "grad_norm": 2.293360948562622, "learning_rate": 2.1702695716448278e-05, "loss": 2.3101, "step": 6216 }, { "epoch": 0.055915061630683584, "grad_norm": 3.358121871948242, "learning_rate": 2.169524575092763e-05, "loss": 1.121, "step": 6217 }, { "epoch": 0.05592405552832403, "grad_norm": 2.4571564197540283, "learning_rate": 2.16877960840871e-05, "loss": 1.8223, "step": 6218 }, { "epoch": 0.055933049425964484, "grad_norm": 12.92473316192627, "learning_rate": 2.1680346716600003e-05, "loss": 1.1379, "step": 6219 }, { "epoch": 0.05594204332360493, "grad_norm": 2.330613136291504, "learning_rate": 2.1672897649139595e-05, "loss": 2.5645, "step": 6220 }, { "epoch": 0.055951037221245385, "grad_norm": 2.533477783203125, "learning_rate": 2.16654488823791e-05, "loss": 2.4728, "step": 6221 }, { "epoch": 0.05596003111888584, "grad_norm": 2.2142488956451416, "learning_rate": 2.1658000416991757e-05, "loss": 2.2052, "step": 6222 }, { "epoch": 0.055969025016526286, "grad_norm": 18.351781845092773, "learning_rate": 2.1650552253650726e-05, "loss": 1.006, "step": 6223 }, { "epoch": 0.05597801891416674, "grad_norm": 2.1849327087402344, "learning_rate": 2.164310439302918e-05, "loss": 1.6033, "step": 6224 }, { "epoch": 0.05598701281180719, "grad_norm": 2.844468832015991, "learning_rate": 2.163565683580024e-05, "loss": 2.2513, "step": 6225 }, { "epoch": 0.05599600670944764, "grad_norm": 11.849068641662598, "learning_rate": 2.1628209582637022e-05, "loss": 1.8391, "step": 6226 }, { "epoch": 0.05600500060708809, "grad_norm": 7.11011266708374, "learning_rate": 2.1620762634212586e-05, "loss": 1.6206, "step": 6227 }, { "epoch": 0.05601399450472854, "grad_norm": 2.392850160598755, "learning_rate": 2.1613315991199974e-05, "loss": 2.3945, "step": 6228 }, { "epoch": 0.056022988402368995, "grad_norm": 3.4287517070770264, "learning_rate": 2.1605869654272224e-05, "loss": 1.5801, "step": 6229 }, { "epoch": 0.05603198230000944, "grad_norm": 5.507307529449463, "learning_rate": 2.1598423624102308e-05, "loss": 1.7657, "step": 6230 }, { "epoch": 0.056040976197649896, "grad_norm": 2.9046053886413574, "learning_rate": 2.1590977901363213e-05, "loss": 1.3603, "step": 6231 }, { "epoch": 0.05604997009529034, "grad_norm": 2.508179187774658, "learning_rate": 2.158353248672786e-05, "loss": 1.7931, "step": 6232 }, { "epoch": 0.056058963992930796, "grad_norm": 2.3827052116394043, "learning_rate": 2.157608738086916e-05, "loss": 1.4991, "step": 6233 }, { "epoch": 0.05606795789057125, "grad_norm": 2.9062235355377197, "learning_rate": 2.1568642584459998e-05, "loss": 2.5555, "step": 6234 }, { "epoch": 0.0560769517882117, "grad_norm": 3.4296040534973145, "learning_rate": 2.1561198098173213e-05, "loss": 1.8331, "step": 6235 }, { "epoch": 0.05608594568585215, "grad_norm": 3.8812642097473145, "learning_rate": 2.1553753922681648e-05, "loss": 0.9856, "step": 6236 }, { "epoch": 0.0560949395834926, "grad_norm": 8.40546703338623, "learning_rate": 2.154631005865809e-05, "loss": 1.7791, "step": 6237 }, { "epoch": 0.05610393348113305, "grad_norm": 3.9450039863586426, "learning_rate": 2.153886650677531e-05, "loss": 1.7934, "step": 6238 }, { "epoch": 0.056112927378773506, "grad_norm": 7.022414684295654, "learning_rate": 2.1531423267706045e-05, "loss": 2.0756, "step": 6239 }, { "epoch": 0.05612192127641395, "grad_norm": 3.8032031059265137, "learning_rate": 2.1523980342123012e-05, "loss": 2.2705, "step": 6240 }, { "epoch": 0.056130915174054406, "grad_norm": 2.433954954147339, "learning_rate": 2.1516537730698892e-05, "loss": 2.5716, "step": 6241 }, { "epoch": 0.05613990907169485, "grad_norm": 2.9761388301849365, "learning_rate": 2.150909543410633e-05, "loss": 2.2701, "step": 6242 }, { "epoch": 0.05614890296933531, "grad_norm": 2.16066312789917, "learning_rate": 2.1501653453017977e-05, "loss": 2.7589, "step": 6243 }, { "epoch": 0.056157896866975754, "grad_norm": 3.28078293800354, "learning_rate": 2.1494211788106412e-05, "loss": 2.2034, "step": 6244 }, { "epoch": 0.05616689076461621, "grad_norm": 4.770376205444336, "learning_rate": 2.1486770440044208e-05, "loss": 0.6532, "step": 6245 }, { "epoch": 0.05617588466225666, "grad_norm": 2.5346269607543945, "learning_rate": 2.1479329409503908e-05, "loss": 2.3016, "step": 6246 }, { "epoch": 0.05618487855989711, "grad_norm": 2.8442142009735107, "learning_rate": 2.1471888697158017e-05, "loss": 1.9297, "step": 6247 }, { "epoch": 0.05619387245753756, "grad_norm": 16.182777404785156, "learning_rate": 2.1464448303679037e-05, "loss": 1.5083, "step": 6248 }, { "epoch": 0.05620286635517801, "grad_norm": 1.8999619483947754, "learning_rate": 2.1457008229739394e-05, "loss": 2.6008, "step": 6249 }, { "epoch": 0.05621186025281846, "grad_norm": 13.008435249328613, "learning_rate": 2.144956847601155e-05, "loss": 1.7397, "step": 6250 }, { "epoch": 0.05622085415045892, "grad_norm": 2.2940781116485596, "learning_rate": 2.1442129043167874e-05, "loss": 2.6527, "step": 6251 }, { "epoch": 0.056229848048099364, "grad_norm": 2.343794345855713, "learning_rate": 2.1434689931880746e-05, "loss": 2.5595, "step": 6252 }, { "epoch": 0.05623884194573982, "grad_norm": 2.733752965927124, "learning_rate": 2.14272511428225e-05, "loss": 2.4935, "step": 6253 }, { "epoch": 0.056247835843380264, "grad_norm": 2.8838021755218506, "learning_rate": 2.141981267666544e-05, "loss": 1.5649, "step": 6254 }, { "epoch": 0.05625682974102072, "grad_norm": 2.6545658111572266, "learning_rate": 2.1412374534081867e-05, "loss": 1.9968, "step": 6255 }, { "epoch": 0.056265823638661165, "grad_norm": 2.8792500495910645, "learning_rate": 2.1404936715744017e-05, "loss": 2.076, "step": 6256 }, { "epoch": 0.05627481753630162, "grad_norm": 16.38611602783203, "learning_rate": 2.1397499222324106e-05, "loss": 1.4053, "step": 6257 }, { "epoch": 0.05628381143394207, "grad_norm": 2.3547239303588867, "learning_rate": 2.1390062054494342e-05, "loss": 1.4585, "step": 6258 }, { "epoch": 0.05629280533158252, "grad_norm": 2.9338173866271973, "learning_rate": 2.1382625212926877e-05, "loss": 1.2407, "step": 6259 }, { "epoch": 0.056301799229222974, "grad_norm": 1.7624162435531616, "learning_rate": 2.1375188698293855e-05, "loss": 2.7783, "step": 6260 }, { "epoch": 0.05631079312686342, "grad_norm": 3.0959529876708984, "learning_rate": 2.1367752511267363e-05, "loss": 1.9486, "step": 6261 }, { "epoch": 0.056319787024503874, "grad_norm": 3.2445828914642334, "learning_rate": 2.1360316652519498e-05, "loss": 1.642, "step": 6262 }, { "epoch": 0.05632878092214433, "grad_norm": 20.136554718017578, "learning_rate": 2.1352881122722292e-05, "loss": 1.4497, "step": 6263 }, { "epoch": 0.056337774819784775, "grad_norm": 17.72723388671875, "learning_rate": 2.134544592254775e-05, "loss": 2.1751, "step": 6264 }, { "epoch": 0.05634676871742523, "grad_norm": 2.7108194828033447, "learning_rate": 2.1338011052667887e-05, "loss": 1.4844, "step": 6265 }, { "epoch": 0.056355762615065676, "grad_norm": 7.621509552001953, "learning_rate": 2.133057651375463e-05, "loss": 1.8174, "step": 6266 }, { "epoch": 0.05636475651270613, "grad_norm": 2.9656200408935547, "learning_rate": 2.132314230647992e-05, "loss": 1.9062, "step": 6267 }, { "epoch": 0.056373750410346576, "grad_norm": 27.929521560668945, "learning_rate": 2.1315708431515648e-05, "loss": 1.23, "step": 6268 }, { "epoch": 0.05638274430798703, "grad_norm": 2.4254283905029297, "learning_rate": 2.1308274889533672e-05, "loss": 2.3319, "step": 6269 }, { "epoch": 0.056391738205627484, "grad_norm": 19.066814422607422, "learning_rate": 2.1300841681205844e-05, "loss": 0.6703, "step": 6270 }, { "epoch": 0.05640073210326793, "grad_norm": 7.909316062927246, "learning_rate": 2.1293408807203947e-05, "loss": 1.1783, "step": 6271 }, { "epoch": 0.056409726000908385, "grad_norm": 35.94742965698242, "learning_rate": 2.1285976268199786e-05, "loss": 0.8573, "step": 6272 }, { "epoch": 0.05641871989854883, "grad_norm": 2.381544828414917, "learning_rate": 2.1278544064865082e-05, "loss": 1.8439, "step": 6273 }, { "epoch": 0.056427713796189286, "grad_norm": 4.207617282867432, "learning_rate": 2.127111219787156e-05, "loss": 1.3694, "step": 6274 }, { "epoch": 0.05643670769382974, "grad_norm": 4.684695720672607, "learning_rate": 2.12636806678909e-05, "loss": 1.3301, "step": 6275 }, { "epoch": 0.056445701591470186, "grad_norm": 2.93285870552063, "learning_rate": 2.125624947559475e-05, "loss": 2.0967, "step": 6276 }, { "epoch": 0.05645469548911064, "grad_norm": 3.4943158626556396, "learning_rate": 2.1248818621654747e-05, "loss": 1.4437, "step": 6277 }, { "epoch": 0.05646368938675109, "grad_norm": 3.573370933532715, "learning_rate": 2.1241388106742472e-05, "loss": 1.2171, "step": 6278 }, { "epoch": 0.05647268328439154, "grad_norm": 2.237787961959839, "learning_rate": 2.1233957931529498e-05, "loss": 1.5954, "step": 6279 }, { "epoch": 0.056481677182031995, "grad_norm": 2.586348056793213, "learning_rate": 2.1226528096687348e-05, "loss": 1.7099, "step": 6280 }, { "epoch": 0.05649067107967244, "grad_norm": 14.099623680114746, "learning_rate": 2.121909860288752e-05, "loss": 1.0898, "step": 6281 }, { "epoch": 0.056499664977312895, "grad_norm": 16.291643142700195, "learning_rate": 2.1211669450801493e-05, "loss": 1.3471, "step": 6282 }, { "epoch": 0.05650865887495334, "grad_norm": 2.4896080493927, "learning_rate": 2.1204240641100693e-05, "loss": 1.9505, "step": 6283 }, { "epoch": 0.056517652772593796, "grad_norm": 2.65415358543396, "learning_rate": 2.1196812174456542e-05, "loss": 1.0508, "step": 6284 }, { "epoch": 0.05652664667023424, "grad_norm": 2.9677140712738037, "learning_rate": 2.1189384051540414e-05, "loss": 1.6557, "step": 6285 }, { "epoch": 0.0565356405678747, "grad_norm": 2.855534315109253, "learning_rate": 2.1181956273023636e-05, "loss": 1.8546, "step": 6286 }, { "epoch": 0.05654463446551515, "grad_norm": 9.758406639099121, "learning_rate": 2.117452883957755e-05, "loss": 1.3311, "step": 6287 }, { "epoch": 0.0565536283631556, "grad_norm": 3.132784128189087, "learning_rate": 2.1167101751873426e-05, "loss": 1.7921, "step": 6288 }, { "epoch": 0.05656262226079605, "grad_norm": 2.633016586303711, "learning_rate": 2.115967501058252e-05, "loss": 2.6431, "step": 6289 }, { "epoch": 0.0565716161584365, "grad_norm": 2.6179652214050293, "learning_rate": 2.1152248616376035e-05, "loss": 1.8622, "step": 6290 }, { "epoch": 0.05658061005607695, "grad_norm": 15.081232070922852, "learning_rate": 2.114482256992519e-05, "loss": 2.1729, "step": 6291 }, { "epoch": 0.056589603953717406, "grad_norm": 12.115928649902344, "learning_rate": 2.1137396871901127e-05, "loss": 1.4622, "step": 6292 }, { "epoch": 0.05659859785135785, "grad_norm": 2.528446674346924, "learning_rate": 2.1129971522974967e-05, "loss": 1.5292, "step": 6293 }, { "epoch": 0.05660759174899831, "grad_norm": 2.659414529800415, "learning_rate": 2.112254652381782e-05, "loss": 2.2914, "step": 6294 }, { "epoch": 0.056616585646638753, "grad_norm": 2.6402690410614014, "learning_rate": 2.1115121875100737e-05, "loss": 2.2049, "step": 6295 }, { "epoch": 0.05662557954427921, "grad_norm": 7.7310333251953125, "learning_rate": 2.110769757749476e-05, "loss": 1.5032, "step": 6296 }, { "epoch": 0.056634573441919654, "grad_norm": 2.951051950454712, "learning_rate": 2.1100273631670885e-05, "loss": 1.6391, "step": 6297 }, { "epoch": 0.05664356733956011, "grad_norm": 2.186422348022461, "learning_rate": 2.1092850038300068e-05, "loss": 1.6816, "step": 6298 }, { "epoch": 0.05665256123720056, "grad_norm": 9.809473037719727, "learning_rate": 2.108542679805326e-05, "loss": 0.7024, "step": 6299 }, { "epoch": 0.05666155513484101, "grad_norm": 2.435192584991455, "learning_rate": 2.1078003911601363e-05, "loss": 2.6926, "step": 6300 }, { "epoch": 0.05667054903248146, "grad_norm": 2.0794177055358887, "learning_rate": 2.1070581379615253e-05, "loss": 2.6454, "step": 6301 }, { "epoch": 0.05667954293012191, "grad_norm": 4.821994781494141, "learning_rate": 2.1063159202765752e-05, "loss": 1.8856, "step": 6302 }, { "epoch": 0.05668853682776236, "grad_norm": 2.8244547843933105, "learning_rate": 2.1055737381723695e-05, "loss": 1.4694, "step": 6303 }, { "epoch": 0.05669753072540282, "grad_norm": 2.803273916244507, "learning_rate": 2.1048315917159846e-05, "loss": 1.5697, "step": 6304 }, { "epoch": 0.056706524623043264, "grad_norm": 2.9177563190460205, "learning_rate": 2.1040894809744935e-05, "loss": 1.8473, "step": 6305 }, { "epoch": 0.05671551852068372, "grad_norm": 2.3076798915863037, "learning_rate": 2.10334740601497e-05, "loss": 2.0082, "step": 6306 }, { "epoch": 0.056724512418324165, "grad_norm": 2.6534836292266846, "learning_rate": 2.10260536690448e-05, "loss": 2.1316, "step": 6307 }, { "epoch": 0.05673350631596462, "grad_norm": 12.253898620605469, "learning_rate": 2.1018633637100892e-05, "loss": 1.389, "step": 6308 }, { "epoch": 0.05674250021360507, "grad_norm": 3.0577569007873535, "learning_rate": 2.1011213964988592e-05, "loss": 1.4222, "step": 6309 }, { "epoch": 0.05675149411124552, "grad_norm": 1.8533822298049927, "learning_rate": 2.100379465337846e-05, "loss": 2.3228, "step": 6310 }, { "epoch": 0.05676048800888597, "grad_norm": 16.16748046875, "learning_rate": 2.099637570294108e-05, "loss": 1.1632, "step": 6311 }, { "epoch": 0.05676948190652642, "grad_norm": 15.719714164733887, "learning_rate": 2.0988957114346938e-05, "loss": 1.4396, "step": 6312 }, { "epoch": 0.056778475804166874, "grad_norm": 2.2765607833862305, "learning_rate": 2.0981538888266545e-05, "loss": 2.0616, "step": 6313 }, { "epoch": 0.05678746970180732, "grad_norm": 18.196887969970703, "learning_rate": 2.097412102537033e-05, "loss": 1.2325, "step": 6314 }, { "epoch": 0.056796463599447775, "grad_norm": 2.157334804534912, "learning_rate": 2.0966703526328728e-05, "loss": 1.6111, "step": 6315 }, { "epoch": 0.05680545749708823, "grad_norm": 2.390056848526001, "learning_rate": 2.0959286391812118e-05, "loss": 2.6253, "step": 6316 }, { "epoch": 0.056814451394728675, "grad_norm": 2.915530204772949, "learning_rate": 2.095186962249084e-05, "loss": 1.46, "step": 6317 }, { "epoch": 0.05682344529236913, "grad_norm": 13.05138111114502, "learning_rate": 2.0944453219035233e-05, "loss": 1.7993, "step": 6318 }, { "epoch": 0.056832439190009576, "grad_norm": 28.52082061767578, "learning_rate": 2.0937037182115568e-05, "loss": 0.4926, "step": 6319 }, { "epoch": 0.05684143308765003, "grad_norm": 3.4435784816741943, "learning_rate": 2.0929621512402113e-05, "loss": 1.0587, "step": 6320 }, { "epoch": 0.056850426985290484, "grad_norm": 1.9852454662322998, "learning_rate": 2.0922206210565087e-05, "loss": 2.4059, "step": 6321 }, { "epoch": 0.05685942088293093, "grad_norm": 2.6948111057281494, "learning_rate": 2.0914791277274664e-05, "loss": 2.0176, "step": 6322 }, { "epoch": 0.056868414780571384, "grad_norm": 3.093379020690918, "learning_rate": 2.090737671320101e-05, "loss": 1.8733, "step": 6323 }, { "epoch": 0.05687740867821183, "grad_norm": 2.532243251800537, "learning_rate": 2.0899962519014233e-05, "loss": 2.0908, "step": 6324 }, { "epoch": 0.056886402575852285, "grad_norm": 7.895179748535156, "learning_rate": 2.0892548695384436e-05, "loss": 1.4084, "step": 6325 }, { "epoch": 0.05689539647349273, "grad_norm": 4.437901496887207, "learning_rate": 2.088513524298165e-05, "loss": 1.5124, "step": 6326 }, { "epoch": 0.056904390371133186, "grad_norm": 2.9714579582214355, "learning_rate": 2.087772216247592e-05, "loss": 1.6195, "step": 6327 }, { "epoch": 0.05691338426877364, "grad_norm": 2.888932704925537, "learning_rate": 2.0870309454537223e-05, "loss": 1.6631, "step": 6328 }, { "epoch": 0.05692237816641409, "grad_norm": 11.384979248046875, "learning_rate": 2.0862897119835502e-05, "loss": 1.5179, "step": 6329 }, { "epoch": 0.05693137206405454, "grad_norm": 7.945837020874023, "learning_rate": 2.085548515904069e-05, "loss": 1.3152, "step": 6330 }, { "epoch": 0.05694036596169499, "grad_norm": 3.1505637168884277, "learning_rate": 2.0848073572822657e-05, "loss": 1.9868, "step": 6331 }, { "epoch": 0.05694935985933544, "grad_norm": 2.708022356033325, "learning_rate": 2.084066236185127e-05, "loss": 2.3743, "step": 6332 }, { "epoch": 0.056958353756975895, "grad_norm": 2.373565435409546, "learning_rate": 2.0833251526796338e-05, "loss": 2.9485, "step": 6333 }, { "epoch": 0.05696734765461634, "grad_norm": 2.676525831222534, "learning_rate": 2.0825841068327636e-05, "loss": 2.0032, "step": 6334 }, { "epoch": 0.056976341552256796, "grad_norm": 3.7996482849121094, "learning_rate": 2.081843098711493e-05, "loss": 1.624, "step": 6335 }, { "epoch": 0.05698533544989724, "grad_norm": 2.537379026412964, "learning_rate": 2.0811021283827925e-05, "loss": 1.3926, "step": 6336 }, { "epoch": 0.056994329347537696, "grad_norm": 3.1293327808380127, "learning_rate": 2.080361195913631e-05, "loss": 1.8455, "step": 6337 }, { "epoch": 0.05700332324517814, "grad_norm": 2.2671213150024414, "learning_rate": 2.0796203013709714e-05, "loss": 1.6819, "step": 6338 }, { "epoch": 0.0570123171428186, "grad_norm": 2.497199773788452, "learning_rate": 2.0788794448217774e-05, "loss": 2.4116, "step": 6339 }, { "epoch": 0.05702131104045905, "grad_norm": 2.207385778427124, "learning_rate": 2.0781386263330054e-05, "loss": 2.7244, "step": 6340 }, { "epoch": 0.0570303049380995, "grad_norm": 4.197400093078613, "learning_rate": 2.077397845971609e-05, "loss": 1.1807, "step": 6341 }, { "epoch": 0.05703929883573995, "grad_norm": 2.6844072341918945, "learning_rate": 2.0766571038045413e-05, "loss": 2.0707, "step": 6342 }, { "epoch": 0.0570482927333804, "grad_norm": 2.3496956825256348, "learning_rate": 2.0759163998987478e-05, "loss": 1.5911, "step": 6343 }, { "epoch": 0.05705728663102085, "grad_norm": 16.175556182861328, "learning_rate": 2.0751757343211735e-05, "loss": 1.6211, "step": 6344 }, { "epoch": 0.057066280528661306, "grad_norm": 16.20012664794922, "learning_rate": 2.0744351071387595e-05, "loss": 1.5251, "step": 6345 }, { "epoch": 0.05707527442630175, "grad_norm": 13.04784870147705, "learning_rate": 2.0736945184184405e-05, "loss": 1.7072, "step": 6346 }, { "epoch": 0.05708426832394221, "grad_norm": 13.542201042175293, "learning_rate": 2.0729539682271533e-05, "loss": 0.7641, "step": 6347 }, { "epoch": 0.057093262221582654, "grad_norm": 4.1186370849609375, "learning_rate": 2.072213456631825e-05, "loss": 1.123, "step": 6348 }, { "epoch": 0.05710225611922311, "grad_norm": 2.6331264972686768, "learning_rate": 2.0714729836993848e-05, "loss": 1.9331, "step": 6349 }, { "epoch": 0.05711125001686356, "grad_norm": 2.0374081134796143, "learning_rate": 2.070732549496755e-05, "loss": 1.8223, "step": 6350 }, { "epoch": 0.05712024391450401, "grad_norm": 3.224442720413208, "learning_rate": 2.0699921540908544e-05, "loss": 1.6153, "step": 6351 }, { "epoch": 0.05712923781214446, "grad_norm": 18.276050567626953, "learning_rate": 2.0692517975486e-05, "loss": 0.8466, "step": 6352 }, { "epoch": 0.05713823170978491, "grad_norm": 2.3955886363983154, "learning_rate": 2.0685114799369036e-05, "loss": 1.633, "step": 6353 }, { "epoch": 0.05714722560742536, "grad_norm": 15.73517894744873, "learning_rate": 2.0677712013226755e-05, "loss": 1.8711, "step": 6354 }, { "epoch": 0.05715621950506581, "grad_norm": 2.989649534225464, "learning_rate": 2.0670309617728205e-05, "loss": 1.7886, "step": 6355 }, { "epoch": 0.057165213402706264, "grad_norm": 1.8378089666366577, "learning_rate": 2.0662907613542404e-05, "loss": 2.0345, "step": 6356 }, { "epoch": 0.05717420730034672, "grad_norm": 2.6122114658355713, "learning_rate": 2.0655506001338348e-05, "loss": 1.8918, "step": 6357 }, { "epoch": 0.057183201197987164, "grad_norm": 13.182963371276855, "learning_rate": 2.0648104781784977e-05, "loss": 1.3149, "step": 6358 }, { "epoch": 0.05719219509562762, "grad_norm": 2.218418836593628, "learning_rate": 2.0640703955551212e-05, "loss": 1.9247, "step": 6359 }, { "epoch": 0.057201188993268065, "grad_norm": 2.7176547050476074, "learning_rate": 2.063330352330592e-05, "loss": 1.4015, "step": 6360 }, { "epoch": 0.05721018289090852, "grad_norm": 3.333738327026367, "learning_rate": 2.062590348571796e-05, "loss": 1.4865, "step": 6361 }, { "epoch": 0.05721917678854897, "grad_norm": 4.254212379455566, "learning_rate": 2.061850384345613e-05, "loss": 0.8509, "step": 6362 }, { "epoch": 0.05722817068618942, "grad_norm": 2.120281934738159, "learning_rate": 2.0611104597189203e-05, "loss": 2.2425, "step": 6363 }, { "epoch": 0.05723716458382987, "grad_norm": 2.1150994300842285, "learning_rate": 2.060370574758591e-05, "loss": 2.4427, "step": 6364 }, { "epoch": 0.05724615848147032, "grad_norm": 7.26578950881958, "learning_rate": 2.0596307295314958e-05, "loss": 1.6847, "step": 6365 }, { "epoch": 0.057255152379110774, "grad_norm": 2.660614490509033, "learning_rate": 2.0588909241045018e-05, "loss": 2.6144, "step": 6366 }, { "epoch": 0.05726414627675122, "grad_norm": 17.35543441772461, "learning_rate": 2.0581511585444697e-05, "loss": 1.1321, "step": 6367 }, { "epoch": 0.057273140174391675, "grad_norm": 2.2359516620635986, "learning_rate": 2.057411432918261e-05, "loss": 2.1672, "step": 6368 }, { "epoch": 0.05728213407203213, "grad_norm": 3.178501605987549, "learning_rate": 2.05667174729273e-05, "loss": 1.039, "step": 6369 }, { "epoch": 0.057291127969672576, "grad_norm": 2.4403162002563477, "learning_rate": 2.0559321017347285e-05, "loss": 2.6722, "step": 6370 }, { "epoch": 0.05730012186731303, "grad_norm": 2.660313367843628, "learning_rate": 2.0551924963111062e-05, "loss": 1.3854, "step": 6371 }, { "epoch": 0.057309115764953476, "grad_norm": 1.9335484504699707, "learning_rate": 2.0544529310887057e-05, "loss": 2.2218, "step": 6372 }, { "epoch": 0.05731810966259393, "grad_norm": 14.634410858154297, "learning_rate": 2.0537134061343705e-05, "loss": 1.6169, "step": 6373 }, { "epoch": 0.057327103560234384, "grad_norm": 25.48896026611328, "learning_rate": 2.0529739215149368e-05, "loss": 1.727, "step": 6374 }, { "epoch": 0.05733609745787483, "grad_norm": 25.207948684692383, "learning_rate": 2.052234477297238e-05, "loss": 0.8325, "step": 6375 }, { "epoch": 0.057345091355515285, "grad_norm": 2.4240188598632812, "learning_rate": 2.0514950735481052e-05, "loss": 2.0687, "step": 6376 }, { "epoch": 0.05735408525315573, "grad_norm": 2.369910955429077, "learning_rate": 2.0507557103343643e-05, "loss": 2.3271, "step": 6377 }, { "epoch": 0.057363079150796185, "grad_norm": 4.019639015197754, "learning_rate": 2.0500163877228394e-05, "loss": 1.8369, "step": 6378 }, { "epoch": 0.05737207304843663, "grad_norm": 3.287919521331787, "learning_rate": 2.0492771057803472e-05, "loss": 1.2406, "step": 6379 }, { "epoch": 0.057381066946077086, "grad_norm": 3.3355560302734375, "learning_rate": 2.0485378645737057e-05, "loss": 1.2114, "step": 6380 }, { "epoch": 0.05739006084371754, "grad_norm": 2.579101324081421, "learning_rate": 2.047798664169726e-05, "loss": 1.7743, "step": 6381 }, { "epoch": 0.05739905474135799, "grad_norm": 2.6730761528015137, "learning_rate": 2.0470595046352146e-05, "loss": 1.9408, "step": 6382 }, { "epoch": 0.05740804863899844, "grad_norm": 2.4464967250823975, "learning_rate": 2.046320386036979e-05, "loss": 1.4872, "step": 6383 }, { "epoch": 0.05741704253663889, "grad_norm": 2.2445757389068604, "learning_rate": 2.0455813084418167e-05, "loss": 2.9604, "step": 6384 }, { "epoch": 0.05742603643427934, "grad_norm": 2.403839111328125, "learning_rate": 2.0448422719165276e-05, "loss": 1.8134, "step": 6385 }, { "epoch": 0.057435030331919795, "grad_norm": 4.374133586883545, "learning_rate": 2.0441032765279032e-05, "loss": 0.4029, "step": 6386 }, { "epoch": 0.05744402422956024, "grad_norm": 19.10008430480957, "learning_rate": 2.0433643223427333e-05, "loss": 1.3162, "step": 6387 }, { "epoch": 0.057453018127200696, "grad_norm": 2.658632278442383, "learning_rate": 2.0426254094278045e-05, "loss": 1.489, "step": 6388 }, { "epoch": 0.05746201202484114, "grad_norm": 14.87643814086914, "learning_rate": 2.0418865378498976e-05, "loss": 1.7346, "step": 6389 }, { "epoch": 0.0574710059224816, "grad_norm": 26.448732376098633, "learning_rate": 2.0411477076757932e-05, "loss": 1.4744, "step": 6390 }, { "epoch": 0.05747999982012205, "grad_norm": 2.6052193641662598, "learning_rate": 2.040408918972264e-05, "loss": 1.8978, "step": 6391 }, { "epoch": 0.0574889937177625, "grad_norm": 7.886629581451416, "learning_rate": 2.0396701718060824e-05, "loss": 1.4347, "step": 6392 }, { "epoch": 0.05749798761540295, "grad_norm": 3.7090625762939453, "learning_rate": 2.0389314662440148e-05, "loss": 1.7855, "step": 6393 }, { "epoch": 0.0575069815130434, "grad_norm": 2.470893144607544, "learning_rate": 2.0381928023528236e-05, "loss": 1.8614, "step": 6394 }, { "epoch": 0.05751597541068385, "grad_norm": 2.256927013397217, "learning_rate": 2.0374541801992704e-05, "loss": 2.0686, "step": 6395 }, { "epoch": 0.0575249693083243, "grad_norm": 2.7326135635375977, "learning_rate": 2.036715599850109e-05, "loss": 1.2534, "step": 6396 }, { "epoch": 0.05753396320596475, "grad_norm": 2.2289021015167236, "learning_rate": 2.0359770613720943e-05, "loss": 1.6671, "step": 6397 }, { "epoch": 0.057542957103605206, "grad_norm": 2.526602029800415, "learning_rate": 2.0352385648319727e-05, "loss": 1.1653, "step": 6398 }, { "epoch": 0.05755195100124565, "grad_norm": 2.6694443225860596, "learning_rate": 2.0345001102964883e-05, "loss": 2.546, "step": 6399 }, { "epoch": 0.05756094489888611, "grad_norm": 2.2658984661102295, "learning_rate": 2.0337616978323834e-05, "loss": 2.4662, "step": 6400 }, { "epoch": 0.057569938796526554, "grad_norm": 3.6554791927337646, "learning_rate": 2.033023327506393e-05, "loss": 1.8237, "step": 6401 }, { "epoch": 0.05757893269416701, "grad_norm": 4.296962261199951, "learning_rate": 2.0322849993852518e-05, "loss": 1.2387, "step": 6402 }, { "epoch": 0.05758792659180746, "grad_norm": 2.9105324745178223, "learning_rate": 2.031546713535688e-05, "loss": 1.6094, "step": 6403 }, { "epoch": 0.05759692048944791, "grad_norm": 15.876191139221191, "learning_rate": 2.0308084700244288e-05, "loss": 2.0286, "step": 6404 }, { "epoch": 0.05760591438708836, "grad_norm": 2.269028902053833, "learning_rate": 2.0300702689181943e-05, "loss": 1.7316, "step": 6405 }, { "epoch": 0.05761490828472881, "grad_norm": 2.6354310512542725, "learning_rate": 2.0293321102837025e-05, "loss": 2.5103, "step": 6406 }, { "epoch": 0.05762390218236926, "grad_norm": 28.39348793029785, "learning_rate": 2.028593994187668e-05, "loss": 0.9943, "step": 6407 }, { "epoch": 0.05763289608000971, "grad_norm": 2.4896888732910156, "learning_rate": 2.0278559206967995e-05, "loss": 2.1608, "step": 6408 }, { "epoch": 0.057641889977650164, "grad_norm": 2.3891003131866455, "learning_rate": 2.0271178898778054e-05, "loss": 2.4538, "step": 6409 }, { "epoch": 0.05765088387529062, "grad_norm": 3.8266241550445557, "learning_rate": 2.026379901797387e-05, "loss": 1.0031, "step": 6410 }, { "epoch": 0.057659877772931065, "grad_norm": 3.0318777561187744, "learning_rate": 2.025641956522242e-05, "loss": 1.9482, "step": 6411 }, { "epoch": 0.05766887167057152, "grad_norm": 19.674015045166016, "learning_rate": 2.0249040541190673e-05, "loss": 0.9858, "step": 6412 }, { "epoch": 0.057677865568211965, "grad_norm": 2.831857204437256, "learning_rate": 2.0241661946545515e-05, "loss": 1.6323, "step": 6413 }, { "epoch": 0.05768685946585242, "grad_norm": 2.419520616531372, "learning_rate": 2.0234283781953834e-05, "loss": 1.451, "step": 6414 }, { "epoch": 0.05769585336349287, "grad_norm": 8.786890029907227, "learning_rate": 2.0226906048082436e-05, "loss": 1.9439, "step": 6415 }, { "epoch": 0.05770484726113332, "grad_norm": 3.155348062515259, "learning_rate": 2.0219528745598145e-05, "loss": 1.831, "step": 6416 }, { "epoch": 0.057713841158773774, "grad_norm": 2.0709402561187744, "learning_rate": 2.0212151875167694e-05, "loss": 1.6512, "step": 6417 }, { "epoch": 0.05772283505641422, "grad_norm": 7.520233154296875, "learning_rate": 2.0204775437457794e-05, "loss": 1.8473, "step": 6418 }, { "epoch": 0.057731828954054674, "grad_norm": 18.292129516601562, "learning_rate": 2.0197399433135124e-05, "loss": 1.2022, "step": 6419 }, { "epoch": 0.05774082285169513, "grad_norm": 23.11542320251465, "learning_rate": 2.019002386286632e-05, "loss": 1.9401, "step": 6420 }, { "epoch": 0.057749816749335575, "grad_norm": 3.3946585655212402, "learning_rate": 2.0182648727317986e-05, "loss": 1.3887, "step": 6421 }, { "epoch": 0.05775881064697603, "grad_norm": 2.5101568698883057, "learning_rate": 2.017527402715667e-05, "loss": 1.7251, "step": 6422 }, { "epoch": 0.057767804544616476, "grad_norm": 2.554558277130127, "learning_rate": 2.016789976304888e-05, "loss": 2.4999, "step": 6423 }, { "epoch": 0.05777679844225693, "grad_norm": 9.816383361816406, "learning_rate": 2.0160525935661114e-05, "loss": 0.8838, "step": 6424 }, { "epoch": 0.05778579233989738, "grad_norm": 23.778934478759766, "learning_rate": 2.0153152545659798e-05, "loss": 0.8116, "step": 6425 }, { "epoch": 0.05779478623753783, "grad_norm": 10.523138046264648, "learning_rate": 2.0145779593711338e-05, "loss": 1.5302, "step": 6426 }, { "epoch": 0.057803780135178284, "grad_norm": 2.7287659645080566, "learning_rate": 2.0138407080482085e-05, "loss": 1.761, "step": 6427 }, { "epoch": 0.05781277403281873, "grad_norm": 9.90892219543457, "learning_rate": 2.0131035006638367e-05, "loss": 1.0585, "step": 6428 }, { "epoch": 0.057821767930459185, "grad_norm": 5.50095796585083, "learning_rate": 2.0123663372846462e-05, "loss": 1.5476, "step": 6429 }, { "epoch": 0.05783076182809963, "grad_norm": 13.56203556060791, "learning_rate": 2.01162921797726e-05, "loss": 1.2747, "step": 6430 }, { "epoch": 0.057839755725740086, "grad_norm": 3.4567155838012695, "learning_rate": 2.0108921428083e-05, "loss": 1.4041, "step": 6431 }, { "epoch": 0.05784874962338054, "grad_norm": 7.2275261878967285, "learning_rate": 2.010155111844381e-05, "loss": 1.09, "step": 6432 }, { "epoch": 0.057857743521020986, "grad_norm": 3.660776138305664, "learning_rate": 2.0094181251521158e-05, "loss": 1.3669, "step": 6433 }, { "epoch": 0.05786673741866144, "grad_norm": 3.8161041736602783, "learning_rate": 2.008681182798112e-05, "loss": 1.9409, "step": 6434 }, { "epoch": 0.05787573131630189, "grad_norm": 3.999359607696533, "learning_rate": 2.0079442848489723e-05, "loss": 1.3106, "step": 6435 }, { "epoch": 0.05788472521394234, "grad_norm": 2.669426918029785, "learning_rate": 2.0072074313712997e-05, "loss": 2.7587, "step": 6436 }, { "epoch": 0.05789371911158279, "grad_norm": 2.585167169570923, "learning_rate": 2.0064706224316873e-05, "loss": 2.5476, "step": 6437 }, { "epoch": 0.05790271300922324, "grad_norm": 4.069382667541504, "learning_rate": 2.0057338580967297e-05, "loss": 1.1355, "step": 6438 }, { "epoch": 0.057911706906863696, "grad_norm": 4.714339256286621, "learning_rate": 2.0049971384330134e-05, "loss": 0.9022, "step": 6439 }, { "epoch": 0.05792070080450414, "grad_norm": 5.08834171295166, "learning_rate": 2.0042604635071224e-05, "loss": 1.2651, "step": 6440 }, { "epoch": 0.057929694702144596, "grad_norm": 4.037235736846924, "learning_rate": 2.003523833385637e-05, "loss": 1.6344, "step": 6441 }, { "epoch": 0.05793868859978504, "grad_norm": 3.8975400924682617, "learning_rate": 2.0027872481351316e-05, "loss": 1.1604, "step": 6442 }, { "epoch": 0.0579476824974255, "grad_norm": 2.945565938949585, "learning_rate": 2.0020507078221807e-05, "loss": 2.3234, "step": 6443 }, { "epoch": 0.05795667639506595, "grad_norm": 3.208897590637207, "learning_rate": 2.001314212513349e-05, "loss": 1.6261, "step": 6444 }, { "epoch": 0.0579656702927064, "grad_norm": 2.4258456230163574, "learning_rate": 2.0005777622752027e-05, "loss": 2.4796, "step": 6445 }, { "epoch": 0.05797466419034685, "grad_norm": 3.2588186264038086, "learning_rate": 1.9998413571743008e-05, "loss": 2.3286, "step": 6446 }, { "epoch": 0.0579836580879873, "grad_norm": 19.9918212890625, "learning_rate": 1.9991049972771972e-05, "loss": 0.928, "step": 6447 }, { "epoch": 0.05799265198562775, "grad_norm": 2.2868740558624268, "learning_rate": 1.998368682650445e-05, "loss": 2.2247, "step": 6448 }, { "epoch": 0.0580016458832682, "grad_norm": 12.528562545776367, "learning_rate": 1.9976324133605905e-05, "loss": 0.579, "step": 6449 }, { "epoch": 0.05801063978090865, "grad_norm": 2.789999485015869, "learning_rate": 1.9968961894741778e-05, "loss": 1.1375, "step": 6450 }, { "epoch": 0.05801963367854911, "grad_norm": 2.4791452884674072, "learning_rate": 1.9961600110577456e-05, "loss": 1.7186, "step": 6451 }, { "epoch": 0.058028627576189554, "grad_norm": 17.784744262695312, "learning_rate": 1.995423878177828e-05, "loss": 1.6898, "step": 6452 }, { "epoch": 0.05803762147383001, "grad_norm": 15.664779663085938, "learning_rate": 1.9946877909009583e-05, "loss": 1.1526, "step": 6453 }, { "epoch": 0.058046615371470454, "grad_norm": 32.6697998046875, "learning_rate": 1.9939517492936605e-05, "loss": 0.9223, "step": 6454 }, { "epoch": 0.05805560926911091, "grad_norm": 2.2503957748413086, "learning_rate": 1.9932157534224592e-05, "loss": 1.8973, "step": 6455 }, { "epoch": 0.05806460316675136, "grad_norm": 3.5948755741119385, "learning_rate": 1.9924798033538717e-05, "loss": 1.1321, "step": 6456 }, { "epoch": 0.05807359706439181, "grad_norm": 3.452266216278076, "learning_rate": 1.9917438991544135e-05, "loss": 2.3568, "step": 6457 }, { "epoch": 0.05808259096203226, "grad_norm": 2.9770164489746094, "learning_rate": 1.9910080408905944e-05, "loss": 2.1743, "step": 6458 }, { "epoch": 0.05809158485967271, "grad_norm": 2.674774169921875, "learning_rate": 1.9902722286289193e-05, "loss": 1.1607, "step": 6459 }, { "epoch": 0.058100578757313164, "grad_norm": 7.900119304656982, "learning_rate": 1.9895364624358923e-05, "loss": 0.7391, "step": 6460 }, { "epoch": 0.05810957265495362, "grad_norm": 9.098516464233398, "learning_rate": 1.9888007423780093e-05, "loss": 0.5409, "step": 6461 }, { "epoch": 0.058118566552594064, "grad_norm": 2.5825071334838867, "learning_rate": 1.9880650685217654e-05, "loss": 2.6572, "step": 6462 }, { "epoch": 0.05812756045023452, "grad_norm": 2.904989719390869, "learning_rate": 1.9873294409336487e-05, "loss": 2.2745, "step": 6463 }, { "epoch": 0.058136554347874965, "grad_norm": 3.7500507831573486, "learning_rate": 1.9865938596801444e-05, "loss": 1.9351, "step": 6464 }, { "epoch": 0.05814554824551542, "grad_norm": 2.8578133583068848, "learning_rate": 1.9858583248277348e-05, "loss": 2.1145, "step": 6465 }, { "epoch": 0.058154542143155866, "grad_norm": 2.0568318367004395, "learning_rate": 1.9851228364428947e-05, "loss": 2.4296, "step": 6466 }, { "epoch": 0.05816353604079632, "grad_norm": 2.9942333698272705, "learning_rate": 1.9843873945920996e-05, "loss": 2.619, "step": 6467 }, { "epoch": 0.05817252993843677, "grad_norm": 4.164976119995117, "learning_rate": 1.9836519993418158e-05, "loss": 0.5806, "step": 6468 }, { "epoch": 0.05818152383607722, "grad_norm": 2.2337660789489746, "learning_rate": 1.9829166507585083e-05, "loss": 2.0185, "step": 6469 }, { "epoch": 0.058190517733717674, "grad_norm": 3.8816988468170166, "learning_rate": 1.9821813489086374e-05, "loss": 1.7069, "step": 6470 }, { "epoch": 0.05819951163135812, "grad_norm": 2.6530208587646484, "learning_rate": 1.9814460938586572e-05, "loss": 1.9393, "step": 6471 }, { "epoch": 0.058208505528998575, "grad_norm": 2.31097674369812, "learning_rate": 1.9807108856750215e-05, "loss": 2.6563, "step": 6472 }, { "epoch": 0.05821749942663903, "grad_norm": 23.212249755859375, "learning_rate": 1.9799757244241756e-05, "loss": 0.3321, "step": 6473 }, { "epoch": 0.058226493324279475, "grad_norm": 2.426488161087036, "learning_rate": 1.979240610172565e-05, "loss": 2.2524, "step": 6474 }, { "epoch": 0.05823548722191993, "grad_norm": 2.605823040008545, "learning_rate": 1.9785055429866268e-05, "loss": 1.9705, "step": 6475 }, { "epoch": 0.058244481119560376, "grad_norm": 7.130190849304199, "learning_rate": 1.9777705229327952e-05, "loss": 1.9408, "step": 6476 }, { "epoch": 0.05825347501720083, "grad_norm": 2.809830904006958, "learning_rate": 1.9770355500775016e-05, "loss": 1.7742, "step": 6477 }, { "epoch": 0.05826246891484128, "grad_norm": 2.0550358295440674, "learning_rate": 1.9763006244871713e-05, "loss": 1.7648, "step": 6478 }, { "epoch": 0.05827146281248173, "grad_norm": 2.8451969623565674, "learning_rate": 1.975565746228227e-05, "loss": 1.3877, "step": 6479 }, { "epoch": 0.058280456710122185, "grad_norm": 2.7054555416107178, "learning_rate": 1.974830915367086e-05, "loss": 1.1743, "step": 6480 }, { "epoch": 0.05828945060776263, "grad_norm": 2.2886221408843994, "learning_rate": 1.97409613197016e-05, "loss": 2.3088, "step": 6481 }, { "epoch": 0.058298444505403085, "grad_norm": 2.226559638977051, "learning_rate": 1.9733613961038604e-05, "loss": 2.2417, "step": 6482 }, { "epoch": 0.05830743840304353, "grad_norm": 2.8491687774658203, "learning_rate": 1.97262670783459e-05, "loss": 1.9199, "step": 6483 }, { "epoch": 0.058316432300683986, "grad_norm": 3.0353951454162598, "learning_rate": 1.97189206722875e-05, "loss": 1.4488, "step": 6484 }, { "epoch": 0.05832542619832444, "grad_norm": 2.865309476852417, "learning_rate": 1.9711574743527352e-05, "loss": 2.075, "step": 6485 }, { "epoch": 0.05833442009596489, "grad_norm": 28.847633361816406, "learning_rate": 1.9704229292729392e-05, "loss": 1.0479, "step": 6486 }, { "epoch": 0.05834341399360534, "grad_norm": 2.540621519088745, "learning_rate": 1.969688432055749e-05, "loss": 1.8954, "step": 6487 }, { "epoch": 0.05835240789124579, "grad_norm": 2.4780142307281494, "learning_rate": 1.968953982767546e-05, "loss": 2.2197, "step": 6488 }, { "epoch": 0.05836140178888624, "grad_norm": 2.6780593395233154, "learning_rate": 1.9682195814747106e-05, "loss": 1.7278, "step": 6489 }, { "epoch": 0.05837039568652669, "grad_norm": 2.037623167037964, "learning_rate": 1.9674852282436156e-05, "loss": 2.3033, "step": 6490 }, { "epoch": 0.05837938958416714, "grad_norm": 14.164401054382324, "learning_rate": 1.9667509231406334e-05, "loss": 1.3261, "step": 6491 }, { "epoch": 0.058388383481807596, "grad_norm": 19.788419723510742, "learning_rate": 1.9660166662321284e-05, "loss": 2.0836, "step": 6492 }, { "epoch": 0.05839737737944804, "grad_norm": 3.308771848678589, "learning_rate": 1.9652824575844607e-05, "loss": 1.6992, "step": 6493 }, { "epoch": 0.0584063712770885, "grad_norm": 2.6091716289520264, "learning_rate": 1.9645482972639894e-05, "loss": 1.363, "step": 6494 }, { "epoch": 0.05841536517472894, "grad_norm": 22.233936309814453, "learning_rate": 1.9638141853370658e-05, "loss": 1.5755, "step": 6495 }, { "epoch": 0.0584243590723694, "grad_norm": 12.296062469482422, "learning_rate": 1.9630801218700395e-05, "loss": 1.1479, "step": 6496 }, { "epoch": 0.05843335297000985, "grad_norm": 3.4417147636413574, "learning_rate": 1.962346106929252e-05, "loss": 1.3138, "step": 6497 }, { "epoch": 0.0584423468676503, "grad_norm": 2.148881435394287, "learning_rate": 1.9616121405810457e-05, "loss": 2.5596, "step": 6498 }, { "epoch": 0.05845134076529075, "grad_norm": 17.20873260498047, "learning_rate": 1.9608782228917537e-05, "loss": 2.1863, "step": 6499 }, { "epoch": 0.0584603346629312, "grad_norm": 2.8482587337493896, "learning_rate": 1.9601443539277064e-05, "loss": 1.3674, "step": 6500 }, { "epoch": 0.05846932856057165, "grad_norm": 2.939865827560425, "learning_rate": 1.9594105337552323e-05, "loss": 2.3925, "step": 6501 }, { "epoch": 0.058478322458212106, "grad_norm": 4.005784511566162, "learning_rate": 1.958676762440651e-05, "loss": 1.3237, "step": 6502 }, { "epoch": 0.05848731635585255, "grad_norm": 3.520458221435547, "learning_rate": 1.9579430400502812e-05, "loss": 2.1682, "step": 6503 }, { "epoch": 0.05849631025349301, "grad_norm": 2.2008867263793945, "learning_rate": 1.957209366650436e-05, "loss": 2.5212, "step": 6504 }, { "epoch": 0.058505304151133454, "grad_norm": 3.3753859996795654, "learning_rate": 1.9564757423074226e-05, "loss": 1.7041, "step": 6505 }, { "epoch": 0.05851429804877391, "grad_norm": 3.3430190086364746, "learning_rate": 1.955742167087547e-05, "loss": 1.8636, "step": 6506 }, { "epoch": 0.058523291946414355, "grad_norm": 8.093205451965332, "learning_rate": 1.955008641057107e-05, "loss": 1.3781, "step": 6507 }, { "epoch": 0.05853228584405481, "grad_norm": 2.18335223197937, "learning_rate": 1.9542751642824e-05, "loss": 2.5798, "step": 6508 }, { "epoch": 0.05854127974169526, "grad_norm": 19.54896354675293, "learning_rate": 1.9535417368297155e-05, "loss": 0.5501, "step": 6509 }, { "epoch": 0.05855027363933571, "grad_norm": 2.49882173538208, "learning_rate": 1.9528083587653407e-05, "loss": 1.7408, "step": 6510 }, { "epoch": 0.05855926753697616, "grad_norm": 2.727187395095825, "learning_rate": 1.9520750301555572e-05, "loss": 1.5692, "step": 6511 }, { "epoch": 0.05856826143461661, "grad_norm": 12.678126335144043, "learning_rate": 1.951341751066641e-05, "loss": 1.6307, "step": 6512 }, { "epoch": 0.058577255332257064, "grad_norm": 10.207828521728516, "learning_rate": 1.9506085215648675e-05, "loss": 2.2474, "step": 6513 }, { "epoch": 0.05858624922989752, "grad_norm": 14.422455787658691, "learning_rate": 1.9498753417165026e-05, "loss": 1.1954, "step": 6514 }, { "epoch": 0.058595243127537965, "grad_norm": 2.7489914894104004, "learning_rate": 1.9491422115878137e-05, "loss": 2.1331, "step": 6515 }, { "epoch": 0.05860423702517842, "grad_norm": 2.319197416305542, "learning_rate": 1.9484091312450577e-05, "loss": 1.7631, "step": 6516 }, { "epoch": 0.058613230922818865, "grad_norm": 2.238410472869873, "learning_rate": 1.94767610075449e-05, "loss": 2.6209, "step": 6517 }, { "epoch": 0.05862222482045932, "grad_norm": 3.0509519577026367, "learning_rate": 1.946943120182362e-05, "loss": 1.6145, "step": 6518 }, { "epoch": 0.058631218718099766, "grad_norm": 6.5008769035339355, "learning_rate": 1.9462101895949182e-05, "loss": 1.8693, "step": 6519 }, { "epoch": 0.05864021261574022, "grad_norm": 2.7315433025360107, "learning_rate": 1.9454773090584016e-05, "loss": 1.7082, "step": 6520 }, { "epoch": 0.058649206513380674, "grad_norm": 10.120878219604492, "learning_rate": 1.9447444786390477e-05, "loss": 1.3871, "step": 6521 }, { "epoch": 0.05865820041102112, "grad_norm": 3.0052132606506348, "learning_rate": 1.9440116984030905e-05, "loss": 1.1575, "step": 6522 }, { "epoch": 0.058667194308661574, "grad_norm": 3.030221462249756, "learning_rate": 1.9432789684167578e-05, "loss": 2.1833, "step": 6523 }, { "epoch": 0.05867618820630202, "grad_norm": 10.06636905670166, "learning_rate": 1.9425462887462712e-05, "loss": 0.3842, "step": 6524 }, { "epoch": 0.058685182103942475, "grad_norm": 2.8629565238952637, "learning_rate": 1.9418136594578513e-05, "loss": 1.6357, "step": 6525 }, { "epoch": 0.05869417600158293, "grad_norm": 2.7838056087493896, "learning_rate": 1.9410810806177104e-05, "loss": 1.9363, "step": 6526 }, { "epoch": 0.058703169899223376, "grad_norm": 2.4999585151672363, "learning_rate": 1.940348552292061e-05, "loss": 1.5261, "step": 6527 }, { "epoch": 0.05871216379686383, "grad_norm": 7.145956039428711, "learning_rate": 1.939616074547106e-05, "loss": 1.3359, "step": 6528 }, { "epoch": 0.058721157694504277, "grad_norm": 7.789540767669678, "learning_rate": 1.9388836474490457e-05, "loss": 1.5065, "step": 6529 }, { "epoch": 0.05873015159214473, "grad_norm": 9.590216636657715, "learning_rate": 1.9381512710640782e-05, "loss": 1.7396, "step": 6530 }, { "epoch": 0.058739145489785184, "grad_norm": 10.438943862915039, "learning_rate": 1.937418945458393e-05, "loss": 1.428, "step": 6531 }, { "epoch": 0.05874813938742563, "grad_norm": 3.1550650596618652, "learning_rate": 1.936686670698178e-05, "loss": 1.8673, "step": 6532 }, { "epoch": 0.058757133285066085, "grad_norm": 10.121783256530762, "learning_rate": 1.935954446849614e-05, "loss": 0.8122, "step": 6533 }, { "epoch": 0.05876612718270653, "grad_norm": 4.011422157287598, "learning_rate": 1.9352222739788806e-05, "loss": 0.7091, "step": 6534 }, { "epoch": 0.058775121080346986, "grad_norm": 3.330423355102539, "learning_rate": 1.93449015215215e-05, "loss": 1.4226, "step": 6535 }, { "epoch": 0.05878411497798743, "grad_norm": 4.029369354248047, "learning_rate": 1.9337580814355888e-05, "loss": 1.9829, "step": 6536 }, { "epoch": 0.058793108875627886, "grad_norm": 4.382944107055664, "learning_rate": 1.9330260618953637e-05, "loss": 0.9663, "step": 6537 }, { "epoch": 0.05880210277326834, "grad_norm": 3.01891827583313, "learning_rate": 1.9322940935976316e-05, "loss": 1.332, "step": 6538 }, { "epoch": 0.05881109667090879, "grad_norm": 2.61749267578125, "learning_rate": 1.9315621766085487e-05, "loss": 1.6071, "step": 6539 }, { "epoch": 0.05882009056854924, "grad_norm": 6.944577693939209, "learning_rate": 1.9308303109942638e-05, "loss": 1.1431, "step": 6540 }, { "epoch": 0.05882908446618969, "grad_norm": 4.014144420623779, "learning_rate": 1.9300984968209213e-05, "loss": 1.9958, "step": 6541 }, { "epoch": 0.05883807836383014, "grad_norm": 2.7646446228027344, "learning_rate": 1.929366734154664e-05, "loss": 1.958, "step": 6542 }, { "epoch": 0.058847072261470595, "grad_norm": 2.529576063156128, "learning_rate": 1.928635023061626e-05, "loss": 1.6532, "step": 6543 }, { "epoch": 0.05885606615911104, "grad_norm": 7.87863302230835, "learning_rate": 1.927903363607939e-05, "loss": 0.8548, "step": 6544 }, { "epoch": 0.058865060056751496, "grad_norm": 8.309576034545898, "learning_rate": 1.92717175585973e-05, "loss": 0.4826, "step": 6545 }, { "epoch": 0.05887405395439194, "grad_norm": 2.670325994491577, "learning_rate": 1.9264401998831213e-05, "loss": 2.3087, "step": 6546 }, { "epoch": 0.0588830478520324, "grad_norm": 10.164125442504883, "learning_rate": 1.9257086957442293e-05, "loss": 1.5805, "step": 6547 }, { "epoch": 0.058892041749672844, "grad_norm": 3.2059366703033447, "learning_rate": 1.9249772435091663e-05, "loss": 2.562, "step": 6548 }, { "epoch": 0.0589010356473133, "grad_norm": 2.547980546951294, "learning_rate": 1.9242458432440412e-05, "loss": 1.529, "step": 6549 }, { "epoch": 0.05891002954495375, "grad_norm": 2.774374008178711, "learning_rate": 1.923514495014956e-05, "loss": 1.3568, "step": 6550 }, { "epoch": 0.0589190234425942, "grad_norm": 2.351557493209839, "learning_rate": 1.9227831988880107e-05, "loss": 1.6628, "step": 6551 }, { "epoch": 0.05892801734023465, "grad_norm": 2.7569615840911865, "learning_rate": 1.922051954929298e-05, "loss": 1.9106, "step": 6552 }, { "epoch": 0.0589370112378751, "grad_norm": 30.082990646362305, "learning_rate": 1.9213207632049063e-05, "loss": 0.9545, "step": 6553 }, { "epoch": 0.05894600513551555, "grad_norm": 3.184044122695923, "learning_rate": 1.9205896237809214e-05, "loss": 2.2979, "step": 6554 }, { "epoch": 0.05895499903315601, "grad_norm": 2.6807069778442383, "learning_rate": 1.9198585367234212e-05, "loss": 2.4533, "step": 6555 }, { "epoch": 0.058963992930796454, "grad_norm": 2.121166944503784, "learning_rate": 1.9191275020984827e-05, "loss": 2.4727, "step": 6556 }, { "epoch": 0.05897298682843691, "grad_norm": 4.65942907333374, "learning_rate": 1.9183965199721745e-05, "loss": 1.5788, "step": 6557 }, { "epoch": 0.058981980726077354, "grad_norm": 19.009233474731445, "learning_rate": 1.9176655904105627e-05, "loss": 1.6755, "step": 6558 }, { "epoch": 0.05899097462371781, "grad_norm": 3.573309898376465, "learning_rate": 1.916934713479708e-05, "loss": 1.7594, "step": 6559 }, { "epoch": 0.058999968521358255, "grad_norm": 2.227522611618042, "learning_rate": 1.9162038892456642e-05, "loss": 2.6417, "step": 6560 }, { "epoch": 0.05900896241899871, "grad_norm": 9.792318344116211, "learning_rate": 1.9154731177744856e-05, "loss": 1.5937, "step": 6561 }, { "epoch": 0.05901795631663916, "grad_norm": 17.490169525146484, "learning_rate": 1.914742399132216e-05, "loss": 0.9149, "step": 6562 }, { "epoch": 0.05902695021427961, "grad_norm": 3.377448320388794, "learning_rate": 1.9140117333848986e-05, "loss": 1.7479, "step": 6563 }, { "epoch": 0.05903594411192006, "grad_norm": 3.3244500160217285, "learning_rate": 1.9132811205985698e-05, "loss": 1.4022, "step": 6564 }, { "epoch": 0.05904493800956051, "grad_norm": 25.91261100769043, "learning_rate": 1.9125505608392608e-05, "loss": 0.6384, "step": 6565 }, { "epoch": 0.059053931907200964, "grad_norm": 2.5823590755462646, "learning_rate": 1.911820054173e-05, "loss": 2.5028, "step": 6566 }, { "epoch": 0.05906292580484142, "grad_norm": 3.0214791297912598, "learning_rate": 1.911089600665808e-05, "loss": 2.7204, "step": 6567 }, { "epoch": 0.059071919702481865, "grad_norm": 2.701676607131958, "learning_rate": 1.9103592003837045e-05, "loss": 1.6236, "step": 6568 }, { "epoch": 0.05908091360012232, "grad_norm": 1.9165301322937012, "learning_rate": 1.9096288533927018e-05, "loss": 2.4745, "step": 6569 }, { "epoch": 0.059089907497762766, "grad_norm": 2.6567108631134033, "learning_rate": 1.9088985597588062e-05, "loss": 2.9176, "step": 6570 }, { "epoch": 0.05909890139540322, "grad_norm": 3.2588613033294678, "learning_rate": 1.9081683195480226e-05, "loss": 1.1742, "step": 6571 }, { "epoch": 0.05910789529304367, "grad_norm": 2.1932291984558105, "learning_rate": 1.907438132826349e-05, "loss": 2.7343, "step": 6572 }, { "epoch": 0.05911688919068412, "grad_norm": 11.294676780700684, "learning_rate": 1.9067079996597788e-05, "loss": 1.2251, "step": 6573 }, { "epoch": 0.059125883088324574, "grad_norm": 5.469202995300293, "learning_rate": 1.9059779201142996e-05, "loss": 0.8091, "step": 6574 }, { "epoch": 0.05913487698596502, "grad_norm": 26.694005966186523, "learning_rate": 1.9052478942558972e-05, "loss": 0.6369, "step": 6575 }, { "epoch": 0.059143870883605475, "grad_norm": 3.520059823989868, "learning_rate": 1.9045179221505497e-05, "loss": 1.1988, "step": 6576 }, { "epoch": 0.05915286478124592, "grad_norm": 2.875689744949341, "learning_rate": 1.90378800386423e-05, "loss": 2.5534, "step": 6577 }, { "epoch": 0.059161858678886375, "grad_norm": 4.101829528808594, "learning_rate": 1.903058139462909e-05, "loss": 1.1893, "step": 6578 }, { "epoch": 0.05917085257652683, "grad_norm": 4.052967071533203, "learning_rate": 1.90232832901255e-05, "loss": 1.2524, "step": 6579 }, { "epoch": 0.059179846474167276, "grad_norm": 1.8344266414642334, "learning_rate": 1.9015985725791136e-05, "loss": 2.5684, "step": 6580 }, { "epoch": 0.05918884037180773, "grad_norm": 2.945596933364868, "learning_rate": 1.9008688702285534e-05, "loss": 1.7941, "step": 6581 }, { "epoch": 0.05919783426944818, "grad_norm": 7.495866775512695, "learning_rate": 1.9001392220268183e-05, "loss": 1.6657, "step": 6582 }, { "epoch": 0.05920682816708863, "grad_norm": 2.6404523849487305, "learning_rate": 1.8994096280398555e-05, "loss": 1.8153, "step": 6583 }, { "epoch": 0.059215822064729084, "grad_norm": 20.710723876953125, "learning_rate": 1.8986800883336023e-05, "loss": 1.769, "step": 6584 }, { "epoch": 0.05922481596236953, "grad_norm": 12.697158813476562, "learning_rate": 1.8979506029739964e-05, "loss": 0.9506, "step": 6585 }, { "epoch": 0.059233809860009985, "grad_norm": 6.881481647491455, "learning_rate": 1.8972211720269657e-05, "loss": 1.6443, "step": 6586 }, { "epoch": 0.05924280375765043, "grad_norm": 21.107011795043945, "learning_rate": 1.896491795558437e-05, "loss": 0.9566, "step": 6587 }, { "epoch": 0.059251797655290886, "grad_norm": 3.888643264770508, "learning_rate": 1.8957624736343293e-05, "loss": 0.819, "step": 6588 }, { "epoch": 0.05926079155293133, "grad_norm": 10.521493911743164, "learning_rate": 1.895033206320558e-05, "loss": 0.9851, "step": 6589 }, { "epoch": 0.05926978545057179, "grad_norm": 2.9224371910095215, "learning_rate": 1.8943039936830344e-05, "loss": 1.6165, "step": 6590 }, { "epoch": 0.05927877934821224, "grad_norm": 3.1406807899475098, "learning_rate": 1.8935748357876625e-05, "loss": 1.6925, "step": 6591 }, { "epoch": 0.05928777324585269, "grad_norm": 13.552597999572754, "learning_rate": 1.892845732700345e-05, "loss": 1.6666, "step": 6592 }, { "epoch": 0.05929676714349314, "grad_norm": 2.9640791416168213, "learning_rate": 1.8921166844869762e-05, "loss": 1.7649, "step": 6593 }, { "epoch": 0.05930576104113359, "grad_norm": 9.565706253051758, "learning_rate": 1.891387691213446e-05, "loss": 1.564, "step": 6594 }, { "epoch": 0.05931475493877404, "grad_norm": 2.756864070892334, "learning_rate": 1.8906587529456416e-05, "loss": 1.6998, "step": 6595 }, { "epoch": 0.059323748836414496, "grad_norm": 3.4438209533691406, "learning_rate": 1.8899298697494415e-05, "loss": 1.7939, "step": 6596 }, { "epoch": 0.05933274273405494, "grad_norm": 14.067632675170898, "learning_rate": 1.8892010416907237e-05, "loss": 1.0992, "step": 6597 }, { "epoch": 0.059341736631695396, "grad_norm": 5.0262298583984375, "learning_rate": 1.8884722688353572e-05, "loss": 1.2649, "step": 6598 }, { "epoch": 0.05935073052933584, "grad_norm": 6.599509239196777, "learning_rate": 1.8877435512492083e-05, "loss": 1.3599, "step": 6599 }, { "epoch": 0.0593597244269763, "grad_norm": 2.725496292114258, "learning_rate": 1.8870148889981386e-05, "loss": 1.9404, "step": 6600 }, { "epoch": 0.059368718324616744, "grad_norm": 11.138749122619629, "learning_rate": 1.8862862821480025e-05, "loss": 1.0209, "step": 6601 }, { "epoch": 0.0593777122222572, "grad_norm": 2.460085153579712, "learning_rate": 1.8855577307646513e-05, "loss": 1.9791, "step": 6602 }, { "epoch": 0.05938670611989765, "grad_norm": 2.6262340545654297, "learning_rate": 1.88482923491393e-05, "loss": 1.9492, "step": 6603 }, { "epoch": 0.0593957000175381, "grad_norm": 2.9305648803710938, "learning_rate": 1.8841007946616802e-05, "loss": 2.0758, "step": 6604 }, { "epoch": 0.05940469391517855, "grad_norm": 8.712248802185059, "learning_rate": 1.883372410073738e-05, "loss": 1.2558, "step": 6605 }, { "epoch": 0.059413687812819, "grad_norm": 13.92618179321289, "learning_rate": 1.8826440812159323e-05, "loss": 1.041, "step": 6606 }, { "epoch": 0.05942268171045945, "grad_norm": 2.963809013366699, "learning_rate": 1.8819158081540893e-05, "loss": 1.9977, "step": 6607 }, { "epoch": 0.05943167560809991, "grad_norm": 4.597587585449219, "learning_rate": 1.8811875909540303e-05, "loss": 1.1267, "step": 6608 }, { "epoch": 0.059440669505740354, "grad_norm": 2.227841377258301, "learning_rate": 1.8804594296815706e-05, "loss": 1.9155, "step": 6609 }, { "epoch": 0.05944966340338081, "grad_norm": 4.358119010925293, "learning_rate": 1.8797313244025194e-05, "loss": 1.927, "step": 6610 }, { "epoch": 0.059458657301021255, "grad_norm": 2.3430521488189697, "learning_rate": 1.879003275182684e-05, "loss": 2.7294, "step": 6611 }, { "epoch": 0.05946765119866171, "grad_norm": 10.078500747680664, "learning_rate": 1.8782752820878634e-05, "loss": 0.9683, "step": 6612 }, { "epoch": 0.05947664509630216, "grad_norm": 6.4128851890563965, "learning_rate": 1.877547345183853e-05, "loss": 0.2749, "step": 6613 }, { "epoch": 0.05948563899394261, "grad_norm": 2.7094130516052246, "learning_rate": 1.8768194645364433e-05, "loss": 1.8492, "step": 6614 }, { "epoch": 0.05949463289158306, "grad_norm": 5.571656703948975, "learning_rate": 1.8760916402114183e-05, "loss": 1.2818, "step": 6615 }, { "epoch": 0.05950362678922351, "grad_norm": 4.103518962860107, "learning_rate": 1.87536387227456e-05, "loss": 0.9018, "step": 6616 }, { "epoch": 0.059512620686863964, "grad_norm": 4.148079872131348, "learning_rate": 1.874636160791642e-05, "loss": 1.9676, "step": 6617 }, { "epoch": 0.05952161458450441, "grad_norm": 3.983520984649658, "learning_rate": 1.8739085058284334e-05, "loss": 1.5297, "step": 6618 }, { "epoch": 0.059530608482144864, "grad_norm": 4.034151554107666, "learning_rate": 1.8731809074507004e-05, "loss": 1.2463, "step": 6619 }, { "epoch": 0.05953960237978532, "grad_norm": 5.187223434448242, "learning_rate": 1.8724533657242012e-05, "loss": 1.226, "step": 6620 }, { "epoch": 0.059548596277425765, "grad_norm": 2.279810905456543, "learning_rate": 1.8717258807146918e-05, "loss": 1.7829, "step": 6621 }, { "epoch": 0.05955759017506622, "grad_norm": 3.725925922393799, "learning_rate": 1.8709984524879194e-05, "loss": 0.6196, "step": 6622 }, { "epoch": 0.059566584072706666, "grad_norm": 14.505255699157715, "learning_rate": 1.87027108110963e-05, "loss": 1.6731, "step": 6623 }, { "epoch": 0.05957557797034712, "grad_norm": 2.220571994781494, "learning_rate": 1.869543766645562e-05, "loss": 1.5901, "step": 6624 }, { "epoch": 0.059584571867987574, "grad_norm": 7.195810794830322, "learning_rate": 1.868816509161449e-05, "loss": 1.0478, "step": 6625 }, { "epoch": 0.05959356576562802, "grad_norm": 2.5324652194976807, "learning_rate": 1.8680893087230204e-05, "loss": 1.7961, "step": 6626 }, { "epoch": 0.059602559663268474, "grad_norm": 2.784029245376587, "learning_rate": 1.8673621653959994e-05, "loss": 2.3362, "step": 6627 }, { "epoch": 0.05961155356090892, "grad_norm": 2.305018424987793, "learning_rate": 1.866635079246105e-05, "loss": 1.4162, "step": 6628 }, { "epoch": 0.059620547458549375, "grad_norm": 2.1424453258514404, "learning_rate": 1.8659080503390496e-05, "loss": 2.7725, "step": 6629 }, { "epoch": 0.05962954135618982, "grad_norm": 2.4894542694091797, "learning_rate": 1.8651810787405406e-05, "loss": 1.7674, "step": 6630 }, { "epoch": 0.059638535253830276, "grad_norm": 3.8705365657806396, "learning_rate": 1.8644541645162834e-05, "loss": 1.8896, "step": 6631 }, { "epoch": 0.05964752915147073, "grad_norm": 2.491145610809326, "learning_rate": 1.8637273077319727e-05, "loss": 1.4985, "step": 6632 }, { "epoch": 0.059656523049111176, "grad_norm": 2.301893949508667, "learning_rate": 1.8630005084533038e-05, "loss": 1.8351, "step": 6633 }, { "epoch": 0.05966551694675163, "grad_norm": 2.112844705581665, "learning_rate": 1.8622737667459623e-05, "loss": 2.7326, "step": 6634 }, { "epoch": 0.05967451084439208, "grad_norm": 3.7631590366363525, "learning_rate": 1.8615470826756316e-05, "loss": 0.8922, "step": 6635 }, { "epoch": 0.05968350474203253, "grad_norm": 2.6156089305877686, "learning_rate": 1.8608204563079873e-05, "loss": 2.2224, "step": 6636 }, { "epoch": 0.059692498639672985, "grad_norm": 3.3677923679351807, "learning_rate": 1.8600938877087012e-05, "loss": 2.7818, "step": 6637 }, { "epoch": 0.05970149253731343, "grad_norm": 2.7572484016418457, "learning_rate": 1.859367376943441e-05, "loss": 1.2799, "step": 6638 }, { "epoch": 0.059710486434953886, "grad_norm": 4.677101135253906, "learning_rate": 1.8586409240778666e-05, "loss": 1.7743, "step": 6639 }, { "epoch": 0.05971948033259433, "grad_norm": 18.695388793945312, "learning_rate": 1.8579145291776355e-05, "loss": 1.3434, "step": 6640 }, { "epoch": 0.059728474230234786, "grad_norm": 3.123318672180176, "learning_rate": 1.8571881923083975e-05, "loss": 0.8079, "step": 6641 }, { "epoch": 0.05973746812787524, "grad_norm": 2.6110875606536865, "learning_rate": 1.856461913535798e-05, "loss": 1.3865, "step": 6642 }, { "epoch": 0.05974646202551569, "grad_norm": 2.838191270828247, "learning_rate": 1.855735692925478e-05, "loss": 2.1471, "step": 6643 }, { "epoch": 0.05975545592315614, "grad_norm": 18.637258529663086, "learning_rate": 1.855009530543071e-05, "loss": 1.526, "step": 6644 }, { "epoch": 0.05976444982079659, "grad_norm": 25.188926696777344, "learning_rate": 1.8542834264542092e-05, "loss": 0.8833, "step": 6645 }, { "epoch": 0.05977344371843704, "grad_norm": 2.409428119659424, "learning_rate": 1.8535573807245155e-05, "loss": 2.0114, "step": 6646 }, { "epoch": 0.05978243761607749, "grad_norm": 2.8818531036376953, "learning_rate": 1.8528313934196086e-05, "loss": 1.5804, "step": 6647 }, { "epoch": 0.05979143151371794, "grad_norm": 2.7890708446502686, "learning_rate": 1.8521054646051042e-05, "loss": 2.9301, "step": 6648 }, { "epoch": 0.059800425411358396, "grad_norm": 1.8796849250793457, "learning_rate": 1.8513795943466095e-05, "loss": 2.7363, "step": 6649 }, { "epoch": 0.05980941930899884, "grad_norm": 2.8280186653137207, "learning_rate": 1.8506537827097293e-05, "loss": 0.8704, "step": 6650 }, { "epoch": 0.0598184132066393, "grad_norm": 2.248157262802124, "learning_rate": 1.8499280297600594e-05, "loss": 2.2208, "step": 6651 }, { "epoch": 0.059827407104279744, "grad_norm": 2.069932460784912, "learning_rate": 1.849202335563195e-05, "loss": 2.4344, "step": 6652 }, { "epoch": 0.0598364010019202, "grad_norm": 11.113990783691406, "learning_rate": 1.848476700184723e-05, "loss": 1.9356, "step": 6653 }, { "epoch": 0.05984539489956065, "grad_norm": 18.680438995361328, "learning_rate": 1.8477511236902234e-05, "loss": 0.8618, "step": 6654 }, { "epoch": 0.0598543887972011, "grad_norm": 2.578897476196289, "learning_rate": 1.8470256061452763e-05, "loss": 2.1667, "step": 6655 }, { "epoch": 0.05986338269484155, "grad_norm": 22.744592666625977, "learning_rate": 1.8463001476154508e-05, "loss": 1.7075, "step": 6656 }, { "epoch": 0.059872376592482, "grad_norm": 2.1052870750427246, "learning_rate": 1.8455747481663143e-05, "loss": 1.6702, "step": 6657 }, { "epoch": 0.05988137049012245, "grad_norm": 13.573596954345703, "learning_rate": 1.844849407863427e-05, "loss": 1.6556, "step": 6658 }, { "epoch": 0.0598903643877629, "grad_norm": 2.8984689712524414, "learning_rate": 1.844124126772344e-05, "loss": 1.9723, "step": 6659 }, { "epoch": 0.059899358285403353, "grad_norm": 5.0649261474609375, "learning_rate": 1.8433989049586166e-05, "loss": 0.4806, "step": 6660 }, { "epoch": 0.05990835218304381, "grad_norm": 18.968994140625, "learning_rate": 1.842673742487788e-05, "loss": 1.358, "step": 6661 }, { "epoch": 0.059917346080684254, "grad_norm": 3.1620590686798096, "learning_rate": 1.8419486394254e-05, "loss": 1.227, "step": 6662 }, { "epoch": 0.05992633997832471, "grad_norm": 23.896516799926758, "learning_rate": 1.841223595836984e-05, "loss": 0.58, "step": 6663 }, { "epoch": 0.059935333875965155, "grad_norm": 9.880253791809082, "learning_rate": 1.8404986117880707e-05, "loss": 1.4702, "step": 6664 }, { "epoch": 0.05994432777360561, "grad_norm": 3.683051824569702, "learning_rate": 1.8397736873441822e-05, "loss": 1.0047, "step": 6665 }, { "epoch": 0.05995332167124606, "grad_norm": 2.4225258827209473, "learning_rate": 1.8390488225708365e-05, "loss": 1.5816, "step": 6666 }, { "epoch": 0.05996231556888651, "grad_norm": 13.632396697998047, "learning_rate": 1.8383240175335464e-05, "loss": 1.5952, "step": 6667 }, { "epoch": 0.05997130946652696, "grad_norm": 15.089790344238281, "learning_rate": 1.8375992722978185e-05, "loss": 1.6623, "step": 6668 }, { "epoch": 0.05998030336416741, "grad_norm": 2.536569833755493, "learning_rate": 1.836874586929156e-05, "loss": 2.4644, "step": 6669 }, { "epoch": 0.059989297261807864, "grad_norm": 7.968502998352051, "learning_rate": 1.8361499614930532e-05, "loss": 1.618, "step": 6670 }, { "epoch": 0.05999829115944831, "grad_norm": 2.473602533340454, "learning_rate": 1.8354253960550015e-05, "loss": 2.6311, "step": 6671 }, { "epoch": 0.060007285057088765, "grad_norm": 2.2671923637390137, "learning_rate": 1.834700890680488e-05, "loss": 2.8425, "step": 6672 }, { "epoch": 0.06001627895472922, "grad_norm": 2.451450824737549, "learning_rate": 1.8339764454349898e-05, "loss": 2.6393, "step": 6673 }, { "epoch": 0.060025272852369665, "grad_norm": 15.039514541625977, "learning_rate": 1.8332520603839843e-05, "loss": 1.1695, "step": 6674 }, { "epoch": 0.06003426675001012, "grad_norm": 2.9280993938446045, "learning_rate": 1.8325277355929386e-05, "loss": 1.7498, "step": 6675 }, { "epoch": 0.060043260647650566, "grad_norm": 2.968531370162964, "learning_rate": 1.831803471127318e-05, "loss": 1.3954, "step": 6676 }, { "epoch": 0.06005225454529102, "grad_norm": 2.3663103580474854, "learning_rate": 1.83107926705258e-05, "loss": 2.0899, "step": 6677 }, { "epoch": 0.060061248442931474, "grad_norm": 3.229858160018921, "learning_rate": 1.8303551234341763e-05, "loss": 1.749, "step": 6678 }, { "epoch": 0.06007024234057192, "grad_norm": 2.767094135284424, "learning_rate": 1.8296310403375562e-05, "loss": 2.2116, "step": 6679 }, { "epoch": 0.060079236238212375, "grad_norm": 11.261690139770508, "learning_rate": 1.8289070178281596e-05, "loss": 1.5123, "step": 6680 }, { "epoch": 0.06008823013585282, "grad_norm": 2.298550605773926, "learning_rate": 1.8281830559714246e-05, "loss": 2.1626, "step": 6681 }, { "epoch": 0.060097224033493275, "grad_norm": 4.448683261871338, "learning_rate": 1.827459154832782e-05, "loss": 1.5473, "step": 6682 }, { "epoch": 0.06010621793113373, "grad_norm": 2.1440136432647705, "learning_rate": 1.8267353144776554e-05, "loss": 2.1819, "step": 6683 }, { "epoch": 0.060115211828774176, "grad_norm": 7.903903007507324, "learning_rate": 1.826011534971467e-05, "loss": 1.0847, "step": 6684 }, { "epoch": 0.06012420572641463, "grad_norm": 3.2730488777160645, "learning_rate": 1.825287816379629e-05, "loss": 1.6751, "step": 6685 }, { "epoch": 0.06013319962405508, "grad_norm": 12.839073181152344, "learning_rate": 1.824564158767552e-05, "loss": 1.9412, "step": 6686 }, { "epoch": 0.06014219352169553, "grad_norm": 2.049246072769165, "learning_rate": 1.823840562200638e-05, "loss": 2.6686, "step": 6687 }, { "epoch": 0.06015118741933598, "grad_norm": 10.420867919921875, "learning_rate": 1.823117026744287e-05, "loss": 1.5663, "step": 6688 }, { "epoch": 0.06016018131697643, "grad_norm": 4.97452974319458, "learning_rate": 1.8223935524638898e-05, "loss": 1.3612, "step": 6689 }, { "epoch": 0.060169175214616885, "grad_norm": 2.7087695598602295, "learning_rate": 1.821670139424833e-05, "loss": 1.6199, "step": 6690 }, { "epoch": 0.06017816911225733, "grad_norm": 2.5365874767303467, "learning_rate": 1.820946787692499e-05, "loss": 1.891, "step": 6691 }, { "epoch": 0.060187163009897786, "grad_norm": 3.5897364616394043, "learning_rate": 1.820223497332262e-05, "loss": 1.4501, "step": 6692 }, { "epoch": 0.06019615690753823, "grad_norm": 2.2628254890441895, "learning_rate": 1.8195002684094935e-05, "loss": 2.8015, "step": 6693 }, { "epoch": 0.06020515080517869, "grad_norm": 4.192201137542725, "learning_rate": 1.818777100989558e-05, "loss": 1.3012, "step": 6694 }, { "epoch": 0.06021414470281914, "grad_norm": 2.122249126434326, "learning_rate": 1.818053995137814e-05, "loss": 2.6023, "step": 6695 }, { "epoch": 0.06022313860045959, "grad_norm": 9.424440383911133, "learning_rate": 1.8173309509196156e-05, "loss": 0.9633, "step": 6696 }, { "epoch": 0.06023213249810004, "grad_norm": 13.783613204956055, "learning_rate": 1.8166079684003104e-05, "loss": 1.1006, "step": 6697 }, { "epoch": 0.06024112639574049, "grad_norm": 3.6755220890045166, "learning_rate": 1.8158850476452417e-05, "loss": 1.0725, "step": 6698 }, { "epoch": 0.06025012029338094, "grad_norm": 3.1129958629608154, "learning_rate": 1.815162188719744e-05, "loss": 2.7167, "step": 6699 }, { "epoch": 0.06025911419102139, "grad_norm": 3.005376100540161, "learning_rate": 1.814439391689151e-05, "loss": 2.3989, "step": 6700 }, { "epoch": 0.06026810808866184, "grad_norm": 1.9845877885818481, "learning_rate": 1.813716656618788e-05, "loss": 1.9122, "step": 6701 }, { "epoch": 0.060277101986302296, "grad_norm": 3.8678629398345947, "learning_rate": 1.812993983573973e-05, "loss": 1.4906, "step": 6702 }, { "epoch": 0.06028609588394274, "grad_norm": 2.5896217823028564, "learning_rate": 1.8122713726200225e-05, "loss": 1.2833, "step": 6703 }, { "epoch": 0.0602950897815832, "grad_norm": 14.450841903686523, "learning_rate": 1.8115488238222443e-05, "loss": 1.1478, "step": 6704 }, { "epoch": 0.060304083679223644, "grad_norm": 3.198265790939331, "learning_rate": 1.8108263372459424e-05, "loss": 0.7789, "step": 6705 }, { "epoch": 0.0603130775768641, "grad_norm": 8.263344764709473, "learning_rate": 1.810103912956414e-05, "loss": 1.2371, "step": 6706 }, { "epoch": 0.06032207147450455, "grad_norm": 3.021094799041748, "learning_rate": 1.8093815510189493e-05, "loss": 0.9298, "step": 6707 }, { "epoch": 0.060331065372145, "grad_norm": 2.198812246322632, "learning_rate": 1.8086592514988377e-05, "loss": 2.4895, "step": 6708 }, { "epoch": 0.06034005926978545, "grad_norm": 2.786489248275757, "learning_rate": 1.8079370144613572e-05, "loss": 2.0779, "step": 6709 }, { "epoch": 0.0603490531674259, "grad_norm": 2.6370580196380615, "learning_rate": 1.8072148399717847e-05, "loss": 1.3903, "step": 6710 }, { "epoch": 0.06035804706506635, "grad_norm": 14.251638412475586, "learning_rate": 1.806492728095389e-05, "loss": 1.4835, "step": 6711 }, { "epoch": 0.0603670409627068, "grad_norm": 3.429210662841797, "learning_rate": 1.805770678897433e-05, "loss": 1.1617, "step": 6712 }, { "epoch": 0.060376034860347254, "grad_norm": 6.721078395843506, "learning_rate": 1.8050486924431756e-05, "loss": 1.8457, "step": 6713 }, { "epoch": 0.06038502875798771, "grad_norm": 2.8929355144500732, "learning_rate": 1.804326768797868e-05, "loss": 2.1566, "step": 6714 }, { "epoch": 0.060394022655628155, "grad_norm": 3.1088545322418213, "learning_rate": 1.8036049080267585e-05, "loss": 1.4011, "step": 6715 }, { "epoch": 0.06040301655326861, "grad_norm": 9.007380485534668, "learning_rate": 1.8028831101950865e-05, "loss": 1.2959, "step": 6716 }, { "epoch": 0.060412010450909055, "grad_norm": 3.260287284851074, "learning_rate": 1.8021613753680893e-05, "loss": 0.7997, "step": 6717 }, { "epoch": 0.06042100434854951, "grad_norm": 2.9218850135803223, "learning_rate": 1.801439703610995e-05, "loss": 2.2289, "step": 6718 }, { "epoch": 0.06042999824618996, "grad_norm": 18.04792594909668, "learning_rate": 1.8007180949890277e-05, "loss": 1.8518, "step": 6719 }, { "epoch": 0.06043899214383041, "grad_norm": 2.5040647983551025, "learning_rate": 1.799996549567406e-05, "loss": 2.4736, "step": 6720 }, { "epoch": 0.060447986041470864, "grad_norm": 10.411816596984863, "learning_rate": 1.7992750674113412e-05, "loss": 1.8267, "step": 6721 }, { "epoch": 0.06045697993911131, "grad_norm": 24.81153106689453, "learning_rate": 1.798553648586042e-05, "loss": 1.6643, "step": 6722 }, { "epoch": 0.060465973836751764, "grad_norm": 2.613628625869751, "learning_rate": 1.7978322931567087e-05, "loss": 0.9744, "step": 6723 }, { "epoch": 0.06047496773439222, "grad_norm": 2.60172963142395, "learning_rate": 1.797111001188535e-05, "loss": 1.3779, "step": 6724 }, { "epoch": 0.060483961632032665, "grad_norm": 16.25913429260254, "learning_rate": 1.796389772746714e-05, "loss": 1.6323, "step": 6725 }, { "epoch": 0.06049295552967312, "grad_norm": 3.517164707183838, "learning_rate": 1.795668607896426e-05, "loss": 0.5317, "step": 6726 }, { "epoch": 0.060501949427313566, "grad_norm": 2.567338705062866, "learning_rate": 1.7949475067028512e-05, "loss": 2.063, "step": 6727 }, { "epoch": 0.06051094332495402, "grad_norm": 2.78887939453125, "learning_rate": 1.7942264692311606e-05, "loss": 1.6849, "step": 6728 }, { "epoch": 0.060519937222594467, "grad_norm": 2.525953531265259, "learning_rate": 1.793505495546522e-05, "loss": 2.3471, "step": 6729 }, { "epoch": 0.06052893112023492, "grad_norm": 22.19233512878418, "learning_rate": 1.7927845857140962e-05, "loss": 1.5044, "step": 6730 }, { "epoch": 0.060537925017875374, "grad_norm": 2.1336121559143066, "learning_rate": 1.7920637397990374e-05, "loss": 2.7355, "step": 6731 }, { "epoch": 0.06054691891551582, "grad_norm": 2.7160325050354004, "learning_rate": 1.7913429578664947e-05, "loss": 1.8143, "step": 6732 }, { "epoch": 0.060555912813156275, "grad_norm": 3.39473819732666, "learning_rate": 1.7906222399816124e-05, "loss": 2.0922, "step": 6733 }, { "epoch": 0.06056490671079672, "grad_norm": 2.4111459255218506, "learning_rate": 1.789901586209529e-05, "loss": 2.5591, "step": 6734 }, { "epoch": 0.060573900608437176, "grad_norm": 2.172318935394287, "learning_rate": 1.789180996615375e-05, "loss": 2.466, "step": 6735 }, { "epoch": 0.06058289450607763, "grad_norm": 2.645334005355835, "learning_rate": 1.7884604712642757e-05, "loss": 2.3679, "step": 6736 }, { "epoch": 0.060591888403718076, "grad_norm": 2.459177017211914, "learning_rate": 1.787740010221354e-05, "loss": 2.2076, "step": 6737 }, { "epoch": 0.06060088230135853, "grad_norm": 2.6307263374328613, "learning_rate": 1.787019613551722e-05, "loss": 2.0179, "step": 6738 }, { "epoch": 0.06060987619899898, "grad_norm": 2.956991195678711, "learning_rate": 1.7862992813204904e-05, "loss": 1.3039, "step": 6739 }, { "epoch": 0.06061887009663943, "grad_norm": 2.584132194519043, "learning_rate": 1.7855790135927597e-05, "loss": 1.8614, "step": 6740 }, { "epoch": 0.06062786399427988, "grad_norm": 2.7039072513580322, "learning_rate": 1.7848588104336293e-05, "loss": 1.5933, "step": 6741 }, { "epoch": 0.06063685789192033, "grad_norm": 1.6194078922271729, "learning_rate": 1.7841386719081892e-05, "loss": 2.5811, "step": 6742 }, { "epoch": 0.060645851789560785, "grad_norm": 8.392783164978027, "learning_rate": 1.7834185980815237e-05, "loss": 1.0917, "step": 6743 }, { "epoch": 0.06065484568720123, "grad_norm": 4.236525058746338, "learning_rate": 1.7826985890187148e-05, "loss": 0.6726, "step": 6744 }, { "epoch": 0.060663839584841686, "grad_norm": 2.341742992401123, "learning_rate": 1.7819786447848346e-05, "loss": 1.9852, "step": 6745 }, { "epoch": 0.06067283348248213, "grad_norm": 2.317648410797119, "learning_rate": 1.781258765444951e-05, "loss": 2.1637, "step": 6746 }, { "epoch": 0.06068182738012259, "grad_norm": 1.8525720834732056, "learning_rate": 1.7805389510641263e-05, "loss": 2.1124, "step": 6747 }, { "epoch": 0.06069082127776304, "grad_norm": 2.714545249938965, "learning_rate": 1.779819201707415e-05, "loss": 2.0176, "step": 6748 }, { "epoch": 0.06069981517540349, "grad_norm": 2.904561996459961, "learning_rate": 1.77909951743987e-05, "loss": 2.0452, "step": 6749 }, { "epoch": 0.06070880907304394, "grad_norm": 22.829641342163086, "learning_rate": 1.778379898326533e-05, "loss": 1.3866, "step": 6750 }, { "epoch": 0.06071780297068439, "grad_norm": 16.22712516784668, "learning_rate": 1.7776603444324445e-05, "loss": 1.3549, "step": 6751 }, { "epoch": 0.06072679686832484, "grad_norm": 2.7552332878112793, "learning_rate": 1.7769408558226352e-05, "loss": 2.6288, "step": 6752 }, { "epoch": 0.060735790765965296, "grad_norm": 2.981102466583252, "learning_rate": 1.7762214325621334e-05, "loss": 1.7417, "step": 6753 }, { "epoch": 0.06074478466360574, "grad_norm": 3.1007378101348877, "learning_rate": 1.775502074715959e-05, "loss": 1.5732, "step": 6754 }, { "epoch": 0.0607537785612462, "grad_norm": 2.366981267929077, "learning_rate": 1.7747827823491252e-05, "loss": 2.3658, "step": 6755 }, { "epoch": 0.060762772458886644, "grad_norm": 2.6934525966644287, "learning_rate": 1.7740635555266438e-05, "loss": 1.4737, "step": 6756 }, { "epoch": 0.0607717663565271, "grad_norm": 2.12931489944458, "learning_rate": 1.7733443943135154e-05, "loss": 2.1889, "step": 6757 }, { "epoch": 0.060780760254167544, "grad_norm": 18.11400032043457, "learning_rate": 1.772625298774739e-05, "loss": 1.5962, "step": 6758 }, { "epoch": 0.060789754151808, "grad_norm": 2.2134110927581787, "learning_rate": 1.771906268975304e-05, "loss": 2.586, "step": 6759 }, { "epoch": 0.06079874804944845, "grad_norm": 2.3148183822631836, "learning_rate": 1.7711873049801965e-05, "loss": 1.9667, "step": 6760 }, { "epoch": 0.0608077419470889, "grad_norm": 22.3873291015625, "learning_rate": 1.7704684068543952e-05, "loss": 1.2807, "step": 6761 }, { "epoch": 0.06081673584472935, "grad_norm": 2.5328056812286377, "learning_rate": 1.769749574662873e-05, "loss": 2.3556, "step": 6762 }, { "epoch": 0.0608257297423698, "grad_norm": 2.2443337440490723, "learning_rate": 1.7690308084705986e-05, "loss": 2.5113, "step": 6763 }, { "epoch": 0.06083472364001025, "grad_norm": 16.840190887451172, "learning_rate": 1.768312108342531e-05, "loss": 1.1574, "step": 6764 }, { "epoch": 0.06084371753765071, "grad_norm": 2.0883712768554688, "learning_rate": 1.7675934743436285e-05, "loss": 2.4917, "step": 6765 }, { "epoch": 0.060852711435291154, "grad_norm": 8.18382740020752, "learning_rate": 1.7668749065388385e-05, "loss": 1.9275, "step": 6766 }, { "epoch": 0.06086170533293161, "grad_norm": 2.1323227882385254, "learning_rate": 1.7661564049931046e-05, "loss": 1.8158, "step": 6767 }, { "epoch": 0.060870699230572055, "grad_norm": 1.999347448348999, "learning_rate": 1.7654379697713647e-05, "loss": 2.5692, "step": 6768 }, { "epoch": 0.06087969312821251, "grad_norm": 3.06050181388855, "learning_rate": 1.7647196009385486e-05, "loss": 1.5442, "step": 6769 }, { "epoch": 0.060888687025852956, "grad_norm": 2.49541974067688, "learning_rate": 1.7640012985595845e-05, "loss": 1.9228, "step": 6770 }, { "epoch": 0.06089768092349341, "grad_norm": 3.116194725036621, "learning_rate": 1.76328306269939e-05, "loss": 1.5985, "step": 6771 }, { "epoch": 0.06090667482113386, "grad_norm": 2.8555798530578613, "learning_rate": 1.762564893422878e-05, "loss": 2.1069, "step": 6772 }, { "epoch": 0.06091566871877431, "grad_norm": 2.399186611175537, "learning_rate": 1.7618467907949572e-05, "loss": 2.2421, "step": 6773 }, { "epoch": 0.060924662616414764, "grad_norm": 3.337900400161743, "learning_rate": 1.761128754880528e-05, "loss": 1.5396, "step": 6774 }, { "epoch": 0.06093365651405521, "grad_norm": 3.310943126678467, "learning_rate": 1.7604107857444868e-05, "loss": 1.5639, "step": 6775 }, { "epoch": 0.060942650411695665, "grad_norm": 2.225010633468628, "learning_rate": 1.759692883451721e-05, "loss": 2.1811, "step": 6776 }, { "epoch": 0.06095164430933612, "grad_norm": 3.1634066104888916, "learning_rate": 1.758975048067116e-05, "loss": 1.7798, "step": 6777 }, { "epoch": 0.060960638206976565, "grad_norm": 2.7137365341186523, "learning_rate": 1.7582572796555483e-05, "loss": 1.7363, "step": 6778 }, { "epoch": 0.06096963210461702, "grad_norm": 19.51657485961914, "learning_rate": 1.7575395782818872e-05, "loss": 1.6376, "step": 6779 }, { "epoch": 0.060978626002257466, "grad_norm": 12.683406829833984, "learning_rate": 1.7568219440110006e-05, "loss": 1.4155, "step": 6780 }, { "epoch": 0.06098761989989792, "grad_norm": 6.9439239501953125, "learning_rate": 1.756104376907746e-05, "loss": 0.685, "step": 6781 }, { "epoch": 0.06099661379753837, "grad_norm": 7.166369438171387, "learning_rate": 1.7553868770369768e-05, "loss": 0.9881, "step": 6782 }, { "epoch": 0.06100560769517882, "grad_norm": 2.188406229019165, "learning_rate": 1.7546694444635393e-05, "loss": 1.6099, "step": 6783 }, { "epoch": 0.061014601592819274, "grad_norm": 3.0362045764923096, "learning_rate": 1.753952079252274e-05, "loss": 2.3532, "step": 6784 }, { "epoch": 0.06102359549045972, "grad_norm": 3.201456069946289, "learning_rate": 1.753234781468017e-05, "loss": 1.6104, "step": 6785 }, { "epoch": 0.061032589388100175, "grad_norm": 2.152585744857788, "learning_rate": 1.752517551175596e-05, "loss": 2.1203, "step": 6786 }, { "epoch": 0.06104158328574062, "grad_norm": 11.25692367553711, "learning_rate": 1.751800388439833e-05, "loss": 2.0128, "step": 6787 }, { "epoch": 0.061050577183381076, "grad_norm": 1.9888319969177246, "learning_rate": 1.7510832933255463e-05, "loss": 2.0968, "step": 6788 }, { "epoch": 0.06105957108102153, "grad_norm": 2.5560390949249268, "learning_rate": 1.7503662658975433e-05, "loss": 1.8171, "step": 6789 }, { "epoch": 0.06106856497866198, "grad_norm": 5.711238384246826, "learning_rate": 1.7496493062206308e-05, "loss": 1.0822, "step": 6790 }, { "epoch": 0.06107755887630243, "grad_norm": 2.4225924015045166, "learning_rate": 1.748932414359605e-05, "loss": 1.5847, "step": 6791 }, { "epoch": 0.06108655277394288, "grad_norm": 2.802313804626465, "learning_rate": 1.7482155903792595e-05, "loss": 1.3381, "step": 6792 }, { "epoch": 0.06109554667158333, "grad_norm": 14.206372261047363, "learning_rate": 1.7474988343443784e-05, "loss": 0.6573, "step": 6793 }, { "epoch": 0.061104540569223785, "grad_norm": 2.9164159297943115, "learning_rate": 1.7467821463197433e-05, "loss": 1.0913, "step": 6794 }, { "epoch": 0.06111353446686423, "grad_norm": 3.021409749984741, "learning_rate": 1.7460655263701252e-05, "loss": 1.8305, "step": 6795 }, { "epoch": 0.061122528364504686, "grad_norm": 9.052611351013184, "learning_rate": 1.745348974560293e-05, "loss": 1.4754, "step": 6796 }, { "epoch": 0.06113152226214513, "grad_norm": 2.8910129070281982, "learning_rate": 1.7446324909550082e-05, "loss": 1.5785, "step": 6797 }, { "epoch": 0.061140516159785586, "grad_norm": 2.7377922534942627, "learning_rate": 1.7439160756190244e-05, "loss": 2.2957, "step": 6798 }, { "epoch": 0.06114951005742603, "grad_norm": 2.813603162765503, "learning_rate": 1.7431997286170922e-05, "loss": 1.4215, "step": 6799 }, { "epoch": 0.06115850395506649, "grad_norm": 3.370530128479004, "learning_rate": 1.7424834500139537e-05, "loss": 2.1644, "step": 6800 }, { "epoch": 0.06116749785270694, "grad_norm": 23.401470184326172, "learning_rate": 1.741767239874344e-05, "loss": 0.5365, "step": 6801 }, { "epoch": 0.06117649175034739, "grad_norm": 1.9897246360778809, "learning_rate": 1.7410510982629952e-05, "loss": 1.8538, "step": 6802 }, { "epoch": 0.06118548564798784, "grad_norm": 2.7945919036865234, "learning_rate": 1.7403350252446295e-05, "loss": 2.367, "step": 6803 }, { "epoch": 0.06119447954562829, "grad_norm": 3.2923667430877686, "learning_rate": 1.7396190208839674e-05, "loss": 1.7868, "step": 6804 }, { "epoch": 0.06120347344326874, "grad_norm": 2.5750818252563477, "learning_rate": 1.7389030852457178e-05, "loss": 1.9048, "step": 6805 }, { "epoch": 0.061212467340909196, "grad_norm": 9.561416625976562, "learning_rate": 1.7381872183945887e-05, "loss": 1.2901, "step": 6806 }, { "epoch": 0.06122146123854964, "grad_norm": 4.497103691101074, "learning_rate": 1.737471420395278e-05, "loss": 0.3039, "step": 6807 }, { "epoch": 0.0612304551361901, "grad_norm": 2.284252643585205, "learning_rate": 1.736755691312479e-05, "loss": 2.1131, "step": 6808 }, { "epoch": 0.061239449033830544, "grad_norm": 3.305314064025879, "learning_rate": 1.736040031210879e-05, "loss": 0.4313, "step": 6809 }, { "epoch": 0.061248442931471, "grad_norm": 2.6212713718414307, "learning_rate": 1.7353244401551566e-05, "loss": 1.95, "step": 6810 }, { "epoch": 0.061257436829111445, "grad_norm": 2.58530855178833, "learning_rate": 1.734608918209989e-05, "loss": 2.606, "step": 6811 }, { "epoch": 0.0612664307267519, "grad_norm": 2.2478580474853516, "learning_rate": 1.7338934654400427e-05, "loss": 2.142, "step": 6812 }, { "epoch": 0.06127542462439235, "grad_norm": 2.3349175453186035, "learning_rate": 1.7331780819099787e-05, "loss": 2.7895, "step": 6813 }, { "epoch": 0.0612844185220328, "grad_norm": 3.1901426315307617, "learning_rate": 1.732462767684455e-05, "loss": 1.5995, "step": 6814 }, { "epoch": 0.06129341241967325, "grad_norm": 2.0909032821655273, "learning_rate": 1.7317475228281193e-05, "loss": 2.5984, "step": 6815 }, { "epoch": 0.0613024063173137, "grad_norm": 8.650952339172363, "learning_rate": 1.7310323474056154e-05, "loss": 2.0465, "step": 6816 }, { "epoch": 0.061311400214954154, "grad_norm": 2.4058337211608887, "learning_rate": 1.7303172414815782e-05, "loss": 1.6312, "step": 6817 }, { "epoch": 0.06132039411259461, "grad_norm": 2.3439877033233643, "learning_rate": 1.7296022051206413e-05, "loss": 1.9711, "step": 6818 }, { "epoch": 0.061329388010235054, "grad_norm": 2.0898194313049316, "learning_rate": 1.728887238387427e-05, "loss": 2.1921, "step": 6819 }, { "epoch": 0.06133838190787551, "grad_norm": 38.804656982421875, "learning_rate": 1.7281723413465527e-05, "loss": 0.4588, "step": 6820 }, { "epoch": 0.061347375805515955, "grad_norm": 1.9776620864868164, "learning_rate": 1.7274575140626318e-05, "loss": 2.2225, "step": 6821 }, { "epoch": 0.06135636970315641, "grad_norm": 2.0212206840515137, "learning_rate": 1.726742756600268e-05, "loss": 1.4299, "step": 6822 }, { "epoch": 0.061365363600796856, "grad_norm": 20.499862670898438, "learning_rate": 1.7260280690240616e-05, "loss": 1.84, "step": 6823 }, { "epoch": 0.06137435749843731, "grad_norm": 11.62765121459961, "learning_rate": 1.7253134513986046e-05, "loss": 1.025, "step": 6824 }, { "epoch": 0.061383351396077764, "grad_norm": 31.66212272644043, "learning_rate": 1.7245989037884823e-05, "loss": 0.7619, "step": 6825 }, { "epoch": 0.06139234529371821, "grad_norm": 2.5690653324127197, "learning_rate": 1.723884426258277e-05, "loss": 2.5922, "step": 6826 }, { "epoch": 0.061401339191358664, "grad_norm": 2.359375476837158, "learning_rate": 1.7231700188725605e-05, "loss": 2.1406, "step": 6827 }, { "epoch": 0.06141033308899911, "grad_norm": 1.8326659202575684, "learning_rate": 1.722455681695902e-05, "loss": 1.954, "step": 6828 }, { "epoch": 0.061419326986639565, "grad_norm": 2.4582297801971436, "learning_rate": 1.7217414147928606e-05, "loss": 1.8618, "step": 6829 }, { "epoch": 0.06142832088428002, "grad_norm": 1.7646799087524414, "learning_rate": 1.7210272182279925e-05, "loss": 2.7161, "step": 6830 }, { "epoch": 0.061437314781920466, "grad_norm": 2.17374324798584, "learning_rate": 1.7203130920658457e-05, "loss": 2.0781, "step": 6831 }, { "epoch": 0.06144630867956092, "grad_norm": 2.6762869358062744, "learning_rate": 1.7195990363709604e-05, "loss": 1.5976, "step": 6832 }, { "epoch": 0.061455302577201366, "grad_norm": 2.5894246101379395, "learning_rate": 1.7188850512078746e-05, "loss": 1.8962, "step": 6833 }, { "epoch": 0.06146429647484182, "grad_norm": 16.531248092651367, "learning_rate": 1.7181711366411158e-05, "loss": 1.6598, "step": 6834 }, { "epoch": 0.061473290372482274, "grad_norm": 18.904142379760742, "learning_rate": 1.7174572927352084e-05, "loss": 1.0883, "step": 6835 }, { "epoch": 0.06148228427012272, "grad_norm": 1.647371768951416, "learning_rate": 1.7167435195546684e-05, "loss": 2.0303, "step": 6836 }, { "epoch": 0.061491278167763175, "grad_norm": 3.1938998699188232, "learning_rate": 1.7160298171640043e-05, "loss": 1.5316, "step": 6837 }, { "epoch": 0.06150027206540362, "grad_norm": 1.9349373579025269, "learning_rate": 1.7153161856277218e-05, "loss": 2.6101, "step": 6838 }, { "epoch": 0.061509265963044076, "grad_norm": 2.3370859622955322, "learning_rate": 1.714602625010316e-05, "loss": 1.5503, "step": 6839 }, { "epoch": 0.06151825986068452, "grad_norm": 2.1845452785491943, "learning_rate": 1.7138891353762804e-05, "loss": 1.7122, "step": 6840 }, { "epoch": 0.061527253758324976, "grad_norm": 12.487761497497559, "learning_rate": 1.7131757167900968e-05, "loss": 1.4148, "step": 6841 }, { "epoch": 0.06153624765596543, "grad_norm": 10.17249870300293, "learning_rate": 1.712462369316245e-05, "loss": 1.1871, "step": 6842 }, { "epoch": 0.06154524155360588, "grad_norm": 6.851061820983887, "learning_rate": 1.7117490930191965e-05, "loss": 1.3632, "step": 6843 }, { "epoch": 0.06155423545124633, "grad_norm": 3.142359495162964, "learning_rate": 1.7110358879634154e-05, "loss": 1.6025, "step": 6844 }, { "epoch": 0.06156322934888678, "grad_norm": 2.310985803604126, "learning_rate": 1.7103227542133615e-05, "loss": 1.8644, "step": 6845 }, { "epoch": 0.06157222324652723, "grad_norm": 16.061140060424805, "learning_rate": 1.7096096918334852e-05, "loss": 1.3155, "step": 6846 }, { "epoch": 0.061581217144167685, "grad_norm": 3.5950095653533936, "learning_rate": 1.7088967008882348e-05, "loss": 1.2674, "step": 6847 }, { "epoch": 0.06159021104180813, "grad_norm": 4.173348426818848, "learning_rate": 1.7081837814420486e-05, "loss": 1.0349, "step": 6848 }, { "epoch": 0.061599204939448586, "grad_norm": 12.319186210632324, "learning_rate": 1.7074709335593585e-05, "loss": 1.3588, "step": 6849 }, { "epoch": 0.06160819883708903, "grad_norm": 2.69256329536438, "learning_rate": 1.706758157304592e-05, "loss": 2.5489, "step": 6850 }, { "epoch": 0.06161719273472949, "grad_norm": 2.295644998550415, "learning_rate": 1.7060454527421688e-05, "loss": 2.5007, "step": 6851 }, { "epoch": 0.061626186632369934, "grad_norm": 8.913114547729492, "learning_rate": 1.7053328199365028e-05, "loss": 1.0398, "step": 6852 }, { "epoch": 0.06163518053001039, "grad_norm": 2.516648292541504, "learning_rate": 1.7046202589520008e-05, "loss": 2.4054, "step": 6853 }, { "epoch": 0.06164417442765084, "grad_norm": 11.649989128112793, "learning_rate": 1.703907769853062e-05, "loss": 1.0175, "step": 6854 }, { "epoch": 0.06165316832529129, "grad_norm": 9.739009857177734, "learning_rate": 1.7031953527040824e-05, "loss": 1.7451, "step": 6855 }, { "epoch": 0.06166216222293174, "grad_norm": 7.5058274269104, "learning_rate": 1.702483007569448e-05, "loss": 0.4094, "step": 6856 }, { "epoch": 0.06167115612057219, "grad_norm": 2.50956654548645, "learning_rate": 1.7017707345135413e-05, "loss": 2.3281, "step": 6857 }, { "epoch": 0.06168015001821264, "grad_norm": 3.73134708404541, "learning_rate": 1.7010585336007347e-05, "loss": 1.3506, "step": 6858 }, { "epoch": 0.0616891439158531, "grad_norm": 6.492552280426025, "learning_rate": 1.700346404895398e-05, "loss": 1.282, "step": 6859 }, { "epoch": 0.061698137813493543, "grad_norm": 4.816607475280762, "learning_rate": 1.699634348461892e-05, "loss": 1.0221, "step": 6860 }, { "epoch": 0.061707131711134, "grad_norm": 2.3561134338378906, "learning_rate": 1.6989223643645706e-05, "loss": 1.9755, "step": 6861 }, { "epoch": 0.061716125608774444, "grad_norm": 2.5398473739624023, "learning_rate": 1.6982104526677838e-05, "loss": 1.566, "step": 6862 }, { "epoch": 0.0617251195064149, "grad_norm": 3.807440996170044, "learning_rate": 1.6974986134358722e-05, "loss": 1.2217, "step": 6863 }, { "epoch": 0.06173411340405535, "grad_norm": 10.134356498718262, "learning_rate": 1.6967868467331723e-05, "loss": 1.5913, "step": 6864 }, { "epoch": 0.0617431073016958, "grad_norm": 2.7101023197174072, "learning_rate": 1.696075152624012e-05, "loss": 2.5092, "step": 6865 }, { "epoch": 0.06175210119933625, "grad_norm": 2.5130116939544678, "learning_rate": 1.6953635311727124e-05, "loss": 2.2097, "step": 6866 }, { "epoch": 0.0617610950969767, "grad_norm": 18.617935180664062, "learning_rate": 1.6946519824435914e-05, "loss": 0.7069, "step": 6867 }, { "epoch": 0.06177008899461715, "grad_norm": 2.9305264949798584, "learning_rate": 1.6939405065009555e-05, "loss": 2.5729, "step": 6868 }, { "epoch": 0.0617790828922576, "grad_norm": 3.264005661010742, "learning_rate": 1.6932291034091097e-05, "loss": 1.2568, "step": 6869 }, { "epoch": 0.061788076789898054, "grad_norm": 4.835871696472168, "learning_rate": 1.692517773232348e-05, "loss": 1.7413, "step": 6870 }, { "epoch": 0.06179707068753851, "grad_norm": 6.369422912597656, "learning_rate": 1.6918065160349602e-05, "loss": 0.7859, "step": 6871 }, { "epoch": 0.061806064585178955, "grad_norm": 3.4242584705352783, "learning_rate": 1.6910953318812294e-05, "loss": 2.1857, "step": 6872 }, { "epoch": 0.06181505848281941, "grad_norm": 3.782236337661743, "learning_rate": 1.69038422083543e-05, "loss": 1.4743, "step": 6873 }, { "epoch": 0.061824052380459855, "grad_norm": 4.521255016326904, "learning_rate": 1.689673182961834e-05, "loss": 1.4491, "step": 6874 }, { "epoch": 0.06183304627810031, "grad_norm": 11.229693412780762, "learning_rate": 1.6889622183247018e-05, "loss": 1.6393, "step": 6875 }, { "epoch": 0.06184204017574076, "grad_norm": 2.50127911567688, "learning_rate": 1.6882513269882917e-05, "loss": 2.5266, "step": 6876 }, { "epoch": 0.06185103407338121, "grad_norm": 3.596621036529541, "learning_rate": 1.687540509016852e-05, "loss": 1.3276, "step": 6877 }, { "epoch": 0.061860027971021664, "grad_norm": 2.4558517932891846, "learning_rate": 1.6868297644746257e-05, "loss": 2.3191, "step": 6878 }, { "epoch": 0.06186902186866211, "grad_norm": 2.831799030303955, "learning_rate": 1.68611909342585e-05, "loss": 1.6474, "step": 6879 }, { "epoch": 0.061878015766302565, "grad_norm": 3.5365660190582275, "learning_rate": 1.685408495934753e-05, "loss": 1.5039, "step": 6880 }, { "epoch": 0.06188700966394301, "grad_norm": 2.2994115352630615, "learning_rate": 1.6846979720655597e-05, "loss": 1.8285, "step": 6881 }, { "epoch": 0.061896003561583465, "grad_norm": 3.129082202911377, "learning_rate": 1.6839875218824846e-05, "loss": 2.2093, "step": 6882 }, { "epoch": 0.06190499745922392, "grad_norm": 8.503429412841797, "learning_rate": 1.683277145449739e-05, "loss": 1.2447, "step": 6883 }, { "epoch": 0.061913991356864366, "grad_norm": 3.0854365825653076, "learning_rate": 1.6825668428315257e-05, "loss": 1.5491, "step": 6884 }, { "epoch": 0.06192298525450482, "grad_norm": 7.196030616760254, "learning_rate": 1.68185661409204e-05, "loss": 1.5097, "step": 6885 }, { "epoch": 0.06193197915214527, "grad_norm": 3.8629772663116455, "learning_rate": 1.681146459295473e-05, "loss": 2.0093, "step": 6886 }, { "epoch": 0.06194097304978572, "grad_norm": 2.8229196071624756, "learning_rate": 1.6804363785060056e-05, "loss": 1.9936, "step": 6887 }, { "epoch": 0.061949966947426174, "grad_norm": 2.5826327800750732, "learning_rate": 1.679726371787817e-05, "loss": 2.1828, "step": 6888 }, { "epoch": 0.06195896084506662, "grad_norm": 14.148650169372559, "learning_rate": 1.679016439205075e-05, "loss": 1.6665, "step": 6889 }, { "epoch": 0.061967954742707075, "grad_norm": 10.776053428649902, "learning_rate": 1.6783065808219424e-05, "loss": 1.1, "step": 6890 }, { "epoch": 0.06197694864034752, "grad_norm": 13.958891868591309, "learning_rate": 1.6775967967025762e-05, "loss": 1.6885, "step": 6891 }, { "epoch": 0.061985942537987976, "grad_norm": 2.0987589359283447, "learning_rate": 1.6768870869111257e-05, "loss": 2.647, "step": 6892 }, { "epoch": 0.06199493643562842, "grad_norm": 10.9698486328125, "learning_rate": 1.6761774515117342e-05, "loss": 1.665, "step": 6893 }, { "epoch": 0.06200393033326888, "grad_norm": 3.5586113929748535, "learning_rate": 1.675467890568536e-05, "loss": 0.9208, "step": 6894 }, { "epoch": 0.06201292423090933, "grad_norm": 6.7031941413879395, "learning_rate": 1.6747584041456633e-05, "loss": 1.6161, "step": 6895 }, { "epoch": 0.06202191812854978, "grad_norm": 2.885362148284912, "learning_rate": 1.674048992307237e-05, "loss": 1.6508, "step": 6896 }, { "epoch": 0.06203091202619023, "grad_norm": 2.2258493900299072, "learning_rate": 1.6733396551173718e-05, "loss": 3.0755, "step": 6897 }, { "epoch": 0.06203990592383068, "grad_norm": 8.171675682067871, "learning_rate": 1.6726303926401792e-05, "loss": 1.9662, "step": 6898 }, { "epoch": 0.06204889982147113, "grad_norm": 9.282742500305176, "learning_rate": 1.6719212049397603e-05, "loss": 1.3081, "step": 6899 }, { "epoch": 0.062057893719111586, "grad_norm": 3.3027126789093018, "learning_rate": 1.6712120920802114e-05, "loss": 1.0294, "step": 6900 }, { "epoch": 0.06206688761675203, "grad_norm": 2.732379674911499, "learning_rate": 1.670503054125621e-05, "loss": 1.8069, "step": 6901 }, { "epoch": 0.062075881514392486, "grad_norm": 7.260043144226074, "learning_rate": 1.66979409114007e-05, "loss": 0.745, "step": 6902 }, { "epoch": 0.06208487541203293, "grad_norm": 4.662838459014893, "learning_rate": 1.669085203187636e-05, "loss": 1.3712, "step": 6903 }, { "epoch": 0.06209386930967339, "grad_norm": 19.084169387817383, "learning_rate": 1.668376390332385e-05, "loss": 0.8187, "step": 6904 }, { "epoch": 0.06210286320731384, "grad_norm": 3.058396577835083, "learning_rate": 1.667667652638381e-05, "loss": 1.3162, "step": 6905 }, { "epoch": 0.06211185710495429, "grad_norm": 13.545879364013672, "learning_rate": 1.6669589901696778e-05, "loss": 1.4712, "step": 6906 }, { "epoch": 0.06212085100259474, "grad_norm": 2.548748016357422, "learning_rate": 1.666250402990324e-05, "loss": 1.6652, "step": 6907 }, { "epoch": 0.06212984490023519, "grad_norm": 2.409233808517456, "learning_rate": 1.6655418911643613e-05, "loss": 2.9, "step": 6908 }, { "epoch": 0.06213883879787564, "grad_norm": 2.4233267307281494, "learning_rate": 1.6648334547558226e-05, "loss": 1.6523, "step": 6909 }, { "epoch": 0.06214783269551609, "grad_norm": 2.3080103397369385, "learning_rate": 1.6641250938287374e-05, "loss": 2.7037, "step": 6910 }, { "epoch": 0.06215682659315654, "grad_norm": 6.413151741027832, "learning_rate": 1.663416808447125e-05, "loss": 1.4817, "step": 6911 }, { "epoch": 0.062165820490797, "grad_norm": 7.9092912673950195, "learning_rate": 1.6627085986750013e-05, "loss": 1.6366, "step": 6912 }, { "epoch": 0.062174814388437444, "grad_norm": 2.2800369262695312, "learning_rate": 1.6620004645763727e-05, "loss": 2.4657, "step": 6913 }, { "epoch": 0.0621838082860779, "grad_norm": 4.0057759284973145, "learning_rate": 1.6612924062152385e-05, "loss": 2.0102, "step": 6914 }, { "epoch": 0.062192802183718345, "grad_norm": 2.281846523284912, "learning_rate": 1.6605844236555943e-05, "loss": 2.4688, "step": 6915 }, { "epoch": 0.0622017960813588, "grad_norm": 2.443542957305908, "learning_rate": 1.6598765169614243e-05, "loss": 2.2603, "step": 6916 }, { "epoch": 0.06221078997899925, "grad_norm": 3.5124151706695557, "learning_rate": 1.6591686861967115e-05, "loss": 1.6756, "step": 6917 }, { "epoch": 0.0622197838766397, "grad_norm": 8.56857681274414, "learning_rate": 1.6584609314254263e-05, "loss": 1.6083, "step": 6918 }, { "epoch": 0.06222877777428015, "grad_norm": 4.0522565841674805, "learning_rate": 1.6577532527115358e-05, "loss": 1.7448, "step": 6919 }, { "epoch": 0.0622377716719206, "grad_norm": 2.273977041244507, "learning_rate": 1.6570456501189998e-05, "loss": 2.1611, "step": 6920 }, { "epoch": 0.062246765569561054, "grad_norm": 2.0989067554473877, "learning_rate": 1.6563381237117688e-05, "loss": 2.528, "step": 6921 }, { "epoch": 0.0622557594672015, "grad_norm": 11.600334167480469, "learning_rate": 1.65563067355379e-05, "loss": 1.4075, "step": 6922 }, { "epoch": 0.062264753364841954, "grad_norm": 3.3630218505859375, "learning_rate": 1.654923299709001e-05, "loss": 2.0566, "step": 6923 }, { "epoch": 0.06227374726248241, "grad_norm": 7.2170000076293945, "learning_rate": 1.6542160022413345e-05, "loss": 0.9841, "step": 6924 }, { "epoch": 0.062282741160122855, "grad_norm": 2.8736133575439453, "learning_rate": 1.6535087812147148e-05, "loss": 2.0878, "step": 6925 }, { "epoch": 0.06229173505776331, "grad_norm": 2.4254932403564453, "learning_rate": 1.6528016366930592e-05, "loss": 2.2769, "step": 6926 }, { "epoch": 0.062300728955403756, "grad_norm": 4.19509744644165, "learning_rate": 1.65209456874028e-05, "loss": 0.9149, "step": 6927 }, { "epoch": 0.06230972285304421, "grad_norm": 11.718706130981445, "learning_rate": 1.651387577420279e-05, "loss": 1.1238, "step": 6928 }, { "epoch": 0.06231871675068466, "grad_norm": 2.249149799346924, "learning_rate": 1.650680662796956e-05, "loss": 2.5989, "step": 6929 }, { "epoch": 0.06232771064832511, "grad_norm": 3.312978982925415, "learning_rate": 1.6499738249341995e-05, "loss": 1.1683, "step": 6930 }, { "epoch": 0.062336704545965564, "grad_norm": 3.109739303588867, "learning_rate": 1.6492670638958924e-05, "loss": 1.4572, "step": 6931 }, { "epoch": 0.06234569844360601, "grad_norm": 9.480429649353027, "learning_rate": 1.6485603797459126e-05, "loss": 1.3982, "step": 6932 }, { "epoch": 0.062354692341246465, "grad_norm": 3.654844284057617, "learning_rate": 1.6478537725481283e-05, "loss": 1.79, "step": 6933 }, { "epoch": 0.06236368623888691, "grad_norm": 3.074113130569458, "learning_rate": 1.6471472423664027e-05, "loss": 1.2132, "step": 6934 }, { "epoch": 0.062372680136527366, "grad_norm": 2.2239301204681396, "learning_rate": 1.6464407892645895e-05, "loss": 2.5193, "step": 6935 }, { "epoch": 0.06238167403416782, "grad_norm": 2.2510602474212646, "learning_rate": 1.6457344133065396e-05, "loss": 2.0225, "step": 6936 }, { "epoch": 0.062390667931808266, "grad_norm": 1.921875238418579, "learning_rate": 1.645028114556093e-05, "loss": 1.5815, "step": 6937 }, { "epoch": 0.06239966182944872, "grad_norm": 2.142667055130005, "learning_rate": 1.6443218930770836e-05, "loss": 2.4696, "step": 6938 }, { "epoch": 0.06240865572708917, "grad_norm": 2.6350958347320557, "learning_rate": 1.643615748933341e-05, "loss": 2.1264, "step": 6939 }, { "epoch": 0.06241764962472962, "grad_norm": 2.5245814323425293, "learning_rate": 1.6429096821886837e-05, "loss": 1.5394, "step": 6940 }, { "epoch": 0.062426643522370075, "grad_norm": 17.933218002319336, "learning_rate": 1.6422036929069272e-05, "loss": 0.8294, "step": 6941 }, { "epoch": 0.06243563742001052, "grad_norm": 12.732117652893066, "learning_rate": 1.641497781151877e-05, "loss": 1.302, "step": 6942 }, { "epoch": 0.062444631317650975, "grad_norm": 2.1731724739074707, "learning_rate": 1.640791946987331e-05, "loss": 2.4854, "step": 6943 }, { "epoch": 0.06245362521529142, "grad_norm": 2.5787248611450195, "learning_rate": 1.6400861904770847e-05, "loss": 1.7367, "step": 6944 }, { "epoch": 0.062462619112931876, "grad_norm": 25.53697395324707, "learning_rate": 1.639380511684921e-05, "loss": 1.2294, "step": 6945 }, { "epoch": 0.06247161301057233, "grad_norm": 2.835294723510742, "learning_rate": 1.6386749106746213e-05, "loss": 1.9399, "step": 6946 }, { "epoch": 0.06248060690821278, "grad_norm": 2.5896270275115967, "learning_rate": 1.6379693875099546e-05, "loss": 1.459, "step": 6947 }, { "epoch": 0.06248960080585323, "grad_norm": 2.1901252269744873, "learning_rate": 1.6372639422546865e-05, "loss": 2.8702, "step": 6948 }, { "epoch": 0.06249859470349368, "grad_norm": 3.443615436553955, "learning_rate": 1.636558574972574e-05, "loss": 1.2856, "step": 6949 }, { "epoch": 0.06250758860113413, "grad_norm": 15.168377876281738, "learning_rate": 1.6358532857273667e-05, "loss": 1.8008, "step": 6950 }, { "epoch": 0.06251658249877458, "grad_norm": 3.8941876888275146, "learning_rate": 1.63514807458281e-05, "loss": 1.9078, "step": 6951 }, { "epoch": 0.06252557639641504, "grad_norm": 11.058781623840332, "learning_rate": 1.6344429416026373e-05, "loss": 1.7387, "step": 6952 }, { "epoch": 0.06253457029405549, "grad_norm": 32.83675003051758, "learning_rate": 1.6337378868505805e-05, "loss": 0.8622, "step": 6953 }, { "epoch": 0.06254356419169593, "grad_norm": 15.836210250854492, "learning_rate": 1.63303291039036e-05, "loss": 0.9694, "step": 6954 }, { "epoch": 0.06255255808933638, "grad_norm": 1.691310167312622, "learning_rate": 1.632328012285691e-05, "loss": 2.6087, "step": 6955 }, { "epoch": 0.06256155198697684, "grad_norm": 13.643056869506836, "learning_rate": 1.6316231926002825e-05, "loss": 0.9059, "step": 6956 }, { "epoch": 0.06257054588461729, "grad_norm": 11.355583190917969, "learning_rate": 1.6309184513978328e-05, "loss": 1.4397, "step": 6957 }, { "epoch": 0.06257953978225773, "grad_norm": 3.4354569911956787, "learning_rate": 1.630213788742039e-05, "loss": 1.2883, "step": 6958 }, { "epoch": 0.0625885336798982, "grad_norm": 2.282205581665039, "learning_rate": 1.6295092046965846e-05, "loss": 2.0484, "step": 6959 }, { "epoch": 0.06259752757753864, "grad_norm": 2.334864616394043, "learning_rate": 1.628804699325152e-05, "loss": 1.7991, "step": 6960 }, { "epoch": 0.06260652147517909, "grad_norm": 2.5012574195861816, "learning_rate": 1.6281002726914125e-05, "loss": 2.0123, "step": 6961 }, { "epoch": 0.06261551537281954, "grad_norm": 13.616608619689941, "learning_rate": 1.6273959248590304e-05, "loss": 1.8276, "step": 6962 }, { "epoch": 0.06262450927046, "grad_norm": 17.859516143798828, "learning_rate": 1.626691655891665e-05, "loss": 1.4233, "step": 6963 }, { "epoch": 0.06263350316810044, "grad_norm": 3.044219732284546, "learning_rate": 1.6259874658529663e-05, "loss": 1.3474, "step": 6964 }, { "epoch": 0.06264249706574089, "grad_norm": 2.365917682647705, "learning_rate": 1.6252833548065802e-05, "loss": 1.693, "step": 6965 }, { "epoch": 0.06265149096338135, "grad_norm": 3.9540035724639893, "learning_rate": 1.624579322816142e-05, "loss": 1.872, "step": 6966 }, { "epoch": 0.0626604848610218, "grad_norm": 2.7007029056549072, "learning_rate": 1.623875369945281e-05, "loss": 1.6039, "step": 6967 }, { "epoch": 0.06266947875866224, "grad_norm": 3.306238889694214, "learning_rate": 1.6231714962576214e-05, "loss": 1.1963, "step": 6968 }, { "epoch": 0.06267847265630269, "grad_norm": 3.1140811443328857, "learning_rate": 1.6224677018167767e-05, "loss": 1.4147, "step": 6969 }, { "epoch": 0.06268746655394315, "grad_norm": 2.313880205154419, "learning_rate": 1.621763986686357e-05, "loss": 1.8739, "step": 6970 }, { "epoch": 0.0626964604515836, "grad_norm": 14.347667694091797, "learning_rate": 1.6210603509299605e-05, "loss": 1.6237, "step": 6971 }, { "epoch": 0.06270545434922405, "grad_norm": 3.190976619720459, "learning_rate": 1.6203567946111843e-05, "loss": 2.0587, "step": 6972 }, { "epoch": 0.06271444824686451, "grad_norm": 2.610801935195923, "learning_rate": 1.6196533177936133e-05, "loss": 1.51, "step": 6973 }, { "epoch": 0.06272344214450495, "grad_norm": 3.3358728885650635, "learning_rate": 1.6189499205408267e-05, "loss": 1.1224, "step": 6974 }, { "epoch": 0.0627324360421454, "grad_norm": 3.3101789951324463, "learning_rate": 1.6182466029163975e-05, "loss": 2.194, "step": 6975 }, { "epoch": 0.06274142993978586, "grad_norm": 9.242555618286133, "learning_rate": 1.61754336498389e-05, "loss": 1.4831, "step": 6976 }, { "epoch": 0.06275042383742631, "grad_norm": 2.8136472702026367, "learning_rate": 1.616840206806864e-05, "loss": 1.4805, "step": 6977 }, { "epoch": 0.06275941773506676, "grad_norm": 2.353768825531006, "learning_rate": 1.616137128448868e-05, "loss": 2.6849, "step": 6978 }, { "epoch": 0.0627684116327072, "grad_norm": 13.492554664611816, "learning_rate": 1.615434129973446e-05, "loss": 0.8985, "step": 6979 }, { "epoch": 0.06277740553034766, "grad_norm": 4.195044994354248, "learning_rate": 1.6147312114441354e-05, "loss": 0.9833, "step": 6980 }, { "epoch": 0.06278639942798811, "grad_norm": 2.620311975479126, "learning_rate": 1.6140283729244636e-05, "loss": 3.0332, "step": 6981 }, { "epoch": 0.06279539332562856, "grad_norm": 2.8967549800872803, "learning_rate": 1.613325614477954e-05, "loss": 1.7035, "step": 6982 }, { "epoch": 0.06280438722326902, "grad_norm": 2.278616189956665, "learning_rate": 1.612622936168119e-05, "loss": 2.6436, "step": 6983 }, { "epoch": 0.06281338112090946, "grad_norm": 1.9630157947540283, "learning_rate": 1.6119203380584685e-05, "loss": 2.1144, "step": 6984 }, { "epoch": 0.06282237501854991, "grad_norm": 2.7598955631256104, "learning_rate": 1.611217820212501e-05, "loss": 2.0961, "step": 6985 }, { "epoch": 0.06283136891619036, "grad_norm": 2.344496011734009, "learning_rate": 1.6105153826937085e-05, "loss": 1.8833, "step": 6986 }, { "epoch": 0.06284036281383082, "grad_norm": 3.3308751583099365, "learning_rate": 1.609813025565579e-05, "loss": 1.7836, "step": 6987 }, { "epoch": 0.06284935671147127, "grad_norm": 2.553237199783325, "learning_rate": 1.609110748891589e-05, "loss": 1.5547, "step": 6988 }, { "epoch": 0.06285835060911171, "grad_norm": 2.120405673980713, "learning_rate": 1.60840855273521e-05, "loss": 2.0288, "step": 6989 }, { "epoch": 0.06286734450675217, "grad_norm": 2.1654486656188965, "learning_rate": 1.607706437159906e-05, "loss": 2.1934, "step": 6990 }, { "epoch": 0.06287633840439262, "grad_norm": 1.9527755975723267, "learning_rate": 1.607004402229132e-05, "loss": 1.6209, "step": 6991 }, { "epoch": 0.06288533230203307, "grad_norm": 2.329308271408081, "learning_rate": 1.6063024480063395e-05, "loss": 2.5897, "step": 6992 }, { "epoch": 0.06289432619967353, "grad_norm": 2.9307596683502197, "learning_rate": 1.605600574554968e-05, "loss": 1.0092, "step": 6993 }, { "epoch": 0.06290332009731398, "grad_norm": 1.8784570693969727, "learning_rate": 1.604898781938454e-05, "loss": 2.5981, "step": 6994 }, { "epoch": 0.06291231399495442, "grad_norm": 1.795873761177063, "learning_rate": 1.604197070220224e-05, "loss": 2.4717, "step": 6995 }, { "epoch": 0.06292130789259487, "grad_norm": 3.028172016143799, "learning_rate": 1.6034954394636976e-05, "loss": 1.4883, "step": 6996 }, { "epoch": 0.06293030179023533, "grad_norm": 2.5840628147125244, "learning_rate": 1.602793889732288e-05, "loss": 2.6851, "step": 6997 }, { "epoch": 0.06293929568787578, "grad_norm": 23.659656524658203, "learning_rate": 1.6020924210893996e-05, "loss": 1.5296, "step": 6998 }, { "epoch": 0.06294828958551622, "grad_norm": 3.4459853172302246, "learning_rate": 1.6013910335984316e-05, "loss": 1.2248, "step": 6999 }, { "epoch": 0.06295728348315668, "grad_norm": 14.121413230895996, "learning_rate": 1.6006897273227732e-05, "loss": 1.8431, "step": 7000 }, { "epoch": 0.06296627738079713, "grad_norm": 1.897733211517334, "learning_rate": 1.59998850232581e-05, "loss": 2.8433, "step": 7001 }, { "epoch": 0.06297527127843758, "grad_norm": 3.568979263305664, "learning_rate": 1.5992873586709162e-05, "loss": 1.1135, "step": 7002 }, { "epoch": 0.06298426517607802, "grad_norm": 3.1072142124176025, "learning_rate": 1.5985862964214604e-05, "loss": 1.7139, "step": 7003 }, { "epoch": 0.06299325907371849, "grad_norm": 1.690329670906067, "learning_rate": 1.5978853156408047e-05, "loss": 2.5994, "step": 7004 }, { "epoch": 0.06300225297135893, "grad_norm": 13.85278034210205, "learning_rate": 1.5971844163923012e-05, "loss": 1.571, "step": 7005 }, { "epoch": 0.06301124686899938, "grad_norm": 18.419038772583008, "learning_rate": 1.596483598739299e-05, "loss": 0.9945, "step": 7006 }, { "epoch": 0.06302024076663984, "grad_norm": 2.393202304840088, "learning_rate": 1.595782862745136e-05, "loss": 1.9181, "step": 7007 }, { "epoch": 0.06302923466428029, "grad_norm": 2.4174959659576416, "learning_rate": 1.5950822084731426e-05, "loss": 2.0252, "step": 7008 }, { "epoch": 0.06303822856192073, "grad_norm": 2.3917958736419678, "learning_rate": 1.594381635986646e-05, "loss": 2.2941, "step": 7009 }, { "epoch": 0.0630472224595612, "grad_norm": 24.227783203125, "learning_rate": 1.5936811453489608e-05, "loss": 1.1477, "step": 7010 }, { "epoch": 0.06305621635720164, "grad_norm": 2.6175034046173096, "learning_rate": 1.5929807366233977e-05, "loss": 1.9182, "step": 7011 }, { "epoch": 0.06306521025484209, "grad_norm": 3.5782413482666016, "learning_rate": 1.5922804098732583e-05, "loss": 0.7209, "step": 7012 }, { "epoch": 0.06307420415248254, "grad_norm": 1.8882362842559814, "learning_rate": 1.5915801651618384e-05, "loss": 1.9886, "step": 7013 }, { "epoch": 0.063083198050123, "grad_norm": 2.6205508708953857, "learning_rate": 1.590880002552425e-05, "loss": 2.7644, "step": 7014 }, { "epoch": 0.06309219194776344, "grad_norm": 19.572023391723633, "learning_rate": 1.5901799221082963e-05, "loss": 1.0436, "step": 7015 }, { "epoch": 0.06310118584540389, "grad_norm": 12.836889266967773, "learning_rate": 1.5894799238927276e-05, "loss": 1.0734, "step": 7016 }, { "epoch": 0.06311017974304435, "grad_norm": 2.1536571979522705, "learning_rate": 1.588780007968982e-05, "loss": 2.7206, "step": 7017 }, { "epoch": 0.0631191736406848, "grad_norm": 3.0678131580352783, "learning_rate": 1.588080174400318e-05, "loss": 1.5703, "step": 7018 }, { "epoch": 0.06312816753832524, "grad_norm": 2.0831589698791504, "learning_rate": 1.5873804232499863e-05, "loss": 1.5996, "step": 7019 }, { "epoch": 0.06313716143596569, "grad_norm": 2.547739267349243, "learning_rate": 1.5866807545812278e-05, "loss": 1.5259, "step": 7020 }, { "epoch": 0.06314615533360615, "grad_norm": 18.157777786254883, "learning_rate": 1.5859811684572794e-05, "loss": 1.1562, "step": 7021 }, { "epoch": 0.0631551492312466, "grad_norm": 14.985655784606934, "learning_rate": 1.5852816649413678e-05, "loss": 1.5469, "step": 7022 }, { "epoch": 0.06316414312888705, "grad_norm": 2.613154649734497, "learning_rate": 1.584582244096715e-05, "loss": 2.2631, "step": 7023 }, { "epoch": 0.0631731370265275, "grad_norm": 2.6303482055664062, "learning_rate": 1.583882905986533e-05, "loss": 2.2035, "step": 7024 }, { "epoch": 0.06318213092416795, "grad_norm": 3.135220766067505, "learning_rate": 1.583183650674027e-05, "loss": 1.0139, "step": 7025 }, { "epoch": 0.0631911248218084, "grad_norm": 2.646552801132202, "learning_rate": 1.5824844782223954e-05, "loss": 1.643, "step": 7026 }, { "epoch": 0.06320011871944885, "grad_norm": 2.409923791885376, "learning_rate": 1.5817853886948275e-05, "loss": 1.2554, "step": 7027 }, { "epoch": 0.06320911261708931, "grad_norm": 3.282770872116089, "learning_rate": 1.581086382154508e-05, "loss": 1.6905, "step": 7028 }, { "epoch": 0.06321810651472975, "grad_norm": 2.5910356044769287, "learning_rate": 1.5803874586646107e-05, "loss": 1.8669, "step": 7029 }, { "epoch": 0.0632271004123702, "grad_norm": 1.952449917793274, "learning_rate": 1.5796886182883053e-05, "loss": 2.4282, "step": 7030 }, { "epoch": 0.06323609431001066, "grad_norm": 36.665611267089844, "learning_rate": 1.578989861088751e-05, "loss": 0.582, "step": 7031 }, { "epoch": 0.06324508820765111, "grad_norm": 2.221360206604004, "learning_rate": 1.5782911871291006e-05, "loss": 2.0805, "step": 7032 }, { "epoch": 0.06325408210529156, "grad_norm": 2.362574815750122, "learning_rate": 1.5775925964725008e-05, "loss": 2.1185, "step": 7033 }, { "epoch": 0.06326307600293202, "grad_norm": 2.937331199645996, "learning_rate": 1.5768940891820875e-05, "loss": 0.856, "step": 7034 }, { "epoch": 0.06327206990057246, "grad_norm": 2.6021344661712646, "learning_rate": 1.576195665320993e-05, "loss": 2.6216, "step": 7035 }, { "epoch": 0.06328106379821291, "grad_norm": 2.1090142726898193, "learning_rate": 1.5754973249523387e-05, "loss": 2.2102, "step": 7036 }, { "epoch": 0.06329005769585336, "grad_norm": 2.204972743988037, "learning_rate": 1.5747990681392407e-05, "loss": 2.8079, "step": 7037 }, { "epoch": 0.06329905159349382, "grad_norm": 14.845032691955566, "learning_rate": 1.574100894944806e-05, "loss": 1.1345, "step": 7038 }, { "epoch": 0.06330804549113427, "grad_norm": 22.365507125854492, "learning_rate": 1.5734028054321353e-05, "loss": 0.9431, "step": 7039 }, { "epoch": 0.06331703938877471, "grad_norm": 3.059663772583008, "learning_rate": 1.5727047996643217e-05, "loss": 2.0107, "step": 7040 }, { "epoch": 0.06332603328641517, "grad_norm": 34.626956939697266, "learning_rate": 1.5720068777044476e-05, "loss": 0.6824, "step": 7041 }, { "epoch": 0.06333502718405562, "grad_norm": 2.4826507568359375, "learning_rate": 1.5713090396155944e-05, "loss": 2.0849, "step": 7042 }, { "epoch": 0.06334402108169607, "grad_norm": 16.825258255004883, "learning_rate": 1.5706112854608293e-05, "loss": 0.4423, "step": 7043 }, { "epoch": 0.06335301497933651, "grad_norm": 2.0915133953094482, "learning_rate": 1.5699136153032147e-05, "loss": 1.4399, "step": 7044 }, { "epoch": 0.06336200887697697, "grad_norm": 2.513396978378296, "learning_rate": 1.5692160292058067e-05, "loss": 1.8772, "step": 7045 }, { "epoch": 0.06337100277461742, "grad_norm": 13.570737838745117, "learning_rate": 1.56851852723165e-05, "loss": 1.5708, "step": 7046 }, { "epoch": 0.06337999667225787, "grad_norm": 3.0628161430358887, "learning_rate": 1.5678211094437863e-05, "loss": 0.8833, "step": 7047 }, { "epoch": 0.06338899056989833, "grad_norm": 12.630725860595703, "learning_rate": 1.567123775905246e-05, "loss": 0.6404, "step": 7048 }, { "epoch": 0.06339798446753878, "grad_norm": 3.7964212894439697, "learning_rate": 1.566426526679055e-05, "loss": 1.5936, "step": 7049 }, { "epoch": 0.06340697836517922, "grad_norm": 2.449104070663452, "learning_rate": 1.565729361828229e-05, "loss": 2.423, "step": 7050 }, { "epoch": 0.06341597226281968, "grad_norm": 6.916489601135254, "learning_rate": 1.5650322814157764e-05, "loss": 1.8151, "step": 7051 }, { "epoch": 0.06342496616046013, "grad_norm": 20.17441749572754, "learning_rate": 1.5643352855046996e-05, "loss": 1.7908, "step": 7052 }, { "epoch": 0.06343396005810058, "grad_norm": 2.6306347846984863, "learning_rate": 1.563638374157991e-05, "loss": 1.5927, "step": 7053 }, { "epoch": 0.06344295395574102, "grad_norm": 2.4185361862182617, "learning_rate": 1.5629415474386383e-05, "loss": 1.8998, "step": 7054 }, { "epoch": 0.06345194785338149, "grad_norm": 2.192922353744507, "learning_rate": 1.5622448054096196e-05, "loss": 2.4948, "step": 7055 }, { "epoch": 0.06346094175102193, "grad_norm": 2.273730754852295, "learning_rate": 1.561548148133904e-05, "loss": 2.5358, "step": 7056 }, { "epoch": 0.06346993564866238, "grad_norm": 3.636176347732544, "learning_rate": 1.560851575674457e-05, "loss": 1.7968, "step": 7057 }, { "epoch": 0.06347892954630284, "grad_norm": 2.131446599960327, "learning_rate": 1.560155088094233e-05, "loss": 2.6808, "step": 7058 }, { "epoch": 0.06348792344394329, "grad_norm": 2.571272134780884, "learning_rate": 1.5594586854561796e-05, "loss": 2.2709, "step": 7059 }, { "epoch": 0.06349691734158373, "grad_norm": 2.4138011932373047, "learning_rate": 1.5587623678232365e-05, "loss": 2.0152, "step": 7060 }, { "epoch": 0.06350591123922418, "grad_norm": 2.9981343746185303, "learning_rate": 1.5580661352583375e-05, "loss": 1.248, "step": 7061 }, { "epoch": 0.06351490513686464, "grad_norm": 11.865092277526855, "learning_rate": 1.5573699878244072e-05, "loss": 0.9153, "step": 7062 }, { "epoch": 0.06352389903450509, "grad_norm": 12.198990821838379, "learning_rate": 1.5566739255843606e-05, "loss": 0.5096, "step": 7063 }, { "epoch": 0.06353289293214553, "grad_norm": 5.061520099639893, "learning_rate": 1.55597794860111e-05, "loss": 0.6745, "step": 7064 }, { "epoch": 0.063541886829786, "grad_norm": 2.464644193649292, "learning_rate": 1.5552820569375546e-05, "loss": 2.6734, "step": 7065 }, { "epoch": 0.06355088072742644, "grad_norm": 5.192516803741455, "learning_rate": 1.55458625065659e-05, "loss": 0.2916, "step": 7066 }, { "epoch": 0.06355987462506689, "grad_norm": 2.6343419551849365, "learning_rate": 1.553890529821102e-05, "loss": 2.0541, "step": 7067 }, { "epoch": 0.06356886852270734, "grad_norm": 4.703368186950684, "learning_rate": 1.5531948944939674e-05, "loss": 0.6752, "step": 7068 }, { "epoch": 0.0635778624203478, "grad_norm": 13.252143859863281, "learning_rate": 1.5524993447380596e-05, "loss": 1.5838, "step": 7069 }, { "epoch": 0.06358685631798824, "grad_norm": 2.459758758544922, "learning_rate": 1.5518038806162395e-05, "loss": 2.1588, "step": 7070 }, { "epoch": 0.06359585021562869, "grad_norm": 2.9583377838134766, "learning_rate": 1.5511085021913644e-05, "loss": 1.6555, "step": 7071 }, { "epoch": 0.06360484411326915, "grad_norm": 7.3621506690979, "learning_rate": 1.5504132095262808e-05, "loss": 0.4906, "step": 7072 }, { "epoch": 0.0636138380109096, "grad_norm": 12.136138916015625, "learning_rate": 1.549718002683828e-05, "loss": 0.7946, "step": 7073 }, { "epoch": 0.06362283190855005, "grad_norm": 3.6464860439300537, "learning_rate": 1.549022881726839e-05, "loss": 1.3873, "step": 7074 }, { "epoch": 0.0636318258061905, "grad_norm": 2.625944137573242, "learning_rate": 1.5483278467181365e-05, "loss": 2.6057, "step": 7075 }, { "epoch": 0.06364081970383095, "grad_norm": 10.509140968322754, "learning_rate": 1.5476328977205397e-05, "loss": 0.9563, "step": 7076 }, { "epoch": 0.0636498136014714, "grad_norm": 3.368196725845337, "learning_rate": 1.546938034796854e-05, "loss": 2.022, "step": 7077 }, { "epoch": 0.06365880749911185, "grad_norm": 3.087388277053833, "learning_rate": 1.546243258009884e-05, "loss": 2.1144, "step": 7078 }, { "epoch": 0.06366780139675231, "grad_norm": 3.9050137996673584, "learning_rate": 1.5455485674224203e-05, "loss": 1.343, "step": 7079 }, { "epoch": 0.06367679529439275, "grad_norm": 2.859652042388916, "learning_rate": 1.5448539630972487e-05, "loss": 2.6436, "step": 7080 }, { "epoch": 0.0636857891920332, "grad_norm": 4.0156474113464355, "learning_rate": 1.544159445097148e-05, "loss": 1.2849, "step": 7081 }, { "epoch": 0.06369478308967366, "grad_norm": 2.285734176635742, "learning_rate": 1.5434650134848858e-05, "loss": 2.627, "step": 7082 }, { "epoch": 0.06370377698731411, "grad_norm": 8.565053939819336, "learning_rate": 1.5427706683232264e-05, "loss": 1.1015, "step": 7083 }, { "epoch": 0.06371277088495456, "grad_norm": 2.0902609825134277, "learning_rate": 1.542076409674923e-05, "loss": 2.2742, "step": 7084 }, { "epoch": 0.063721764782595, "grad_norm": 3.622171640396118, "learning_rate": 1.541382237602721e-05, "loss": 1.0044, "step": 7085 }, { "epoch": 0.06373075868023546, "grad_norm": 25.81324577331543, "learning_rate": 1.5406881521693607e-05, "loss": 0.5354, "step": 7086 }, { "epoch": 0.06373975257787591, "grad_norm": 3.3398964405059814, "learning_rate": 1.5399941534375716e-05, "loss": 1.5168, "step": 7087 }, { "epoch": 0.06374874647551636, "grad_norm": 2.4238946437835693, "learning_rate": 1.539300241470077e-05, "loss": 2.0374, "step": 7088 }, { "epoch": 0.06375774037315682, "grad_norm": 3.4558799266815186, "learning_rate": 1.538606416329591e-05, "loss": 1.8003, "step": 7089 }, { "epoch": 0.06376673427079727, "grad_norm": 2.9527530670166016, "learning_rate": 1.537912678078823e-05, "loss": 2.1582, "step": 7090 }, { "epoch": 0.06377572816843771, "grad_norm": 2.976674795150757, "learning_rate": 1.5372190267804704e-05, "loss": 1.1339, "step": 7091 }, { "epoch": 0.06378472206607817, "grad_norm": 3.276317596435547, "learning_rate": 1.5365254624972243e-05, "loss": 0.969, "step": 7092 }, { "epoch": 0.06379371596371862, "grad_norm": 3.3662211894989014, "learning_rate": 1.53583198529177e-05, "loss": 0.9891, "step": 7093 }, { "epoch": 0.06380270986135907, "grad_norm": 2.1138246059417725, "learning_rate": 1.5351385952267815e-05, "loss": 1.4343, "step": 7094 }, { "epoch": 0.06381170375899951, "grad_norm": 3.527120351791382, "learning_rate": 1.5344452923649287e-05, "loss": 0.6161, "step": 7095 }, { "epoch": 0.06382069765663997, "grad_norm": 2.678518772125244, "learning_rate": 1.5337520767688703e-05, "loss": 2.2009, "step": 7096 }, { "epoch": 0.06382969155428042, "grad_norm": 2.083892822265625, "learning_rate": 1.533058948501257e-05, "loss": 2.4093, "step": 7097 }, { "epoch": 0.06383868545192087, "grad_norm": 3.13010311126709, "learning_rate": 1.5323659076247353e-05, "loss": 1.8899, "step": 7098 }, { "epoch": 0.06384767934956133, "grad_norm": 2.605388879776001, "learning_rate": 1.531672954201941e-05, "loss": 1.9151, "step": 7099 }, { "epoch": 0.06385667324720178, "grad_norm": 16.400718688964844, "learning_rate": 1.5309800882955015e-05, "loss": 1.0773, "step": 7100 }, { "epoch": 0.06386566714484222, "grad_norm": 2.528052806854248, "learning_rate": 1.5302873099680377e-05, "loss": 2.1183, "step": 7101 }, { "epoch": 0.06387466104248267, "grad_norm": 2.5693488121032715, "learning_rate": 1.5295946192821632e-05, "loss": 1.8415, "step": 7102 }, { "epoch": 0.06388365494012313, "grad_norm": 2.520010471343994, "learning_rate": 1.5289020163004816e-05, "loss": 2.6603, "step": 7103 }, { "epoch": 0.06389264883776358, "grad_norm": 24.048973083496094, "learning_rate": 1.528209501085589e-05, "loss": 1.6348, "step": 7104 }, { "epoch": 0.06390164273540402, "grad_norm": 2.188539505004883, "learning_rate": 1.527517073700076e-05, "loss": 2.2003, "step": 7105 }, { "epoch": 0.06391063663304448, "grad_norm": 2.091101884841919, "learning_rate": 1.5268247342065215e-05, "loss": 1.3441, "step": 7106 }, { "epoch": 0.06391963053068493, "grad_norm": 2.1699206829071045, "learning_rate": 1.5261324826675e-05, "loss": 1.7453, "step": 7107 }, { "epoch": 0.06392862442832538, "grad_norm": 1.7131186723709106, "learning_rate": 1.5254403191455757e-05, "loss": 1.6408, "step": 7108 }, { "epoch": 0.06393761832596583, "grad_norm": 15.139542579650879, "learning_rate": 1.5247482437033048e-05, "loss": 1.4945, "step": 7109 }, { "epoch": 0.06394661222360629, "grad_norm": 2.267333984375, "learning_rate": 1.5240562564032379e-05, "loss": 1.9629, "step": 7110 }, { "epoch": 0.06395560612124673, "grad_norm": 2.079526662826538, "learning_rate": 1.5233643573079148e-05, "loss": 2.3054, "step": 7111 }, { "epoch": 0.06396460001888718, "grad_norm": 3.2960269451141357, "learning_rate": 1.5226725464798697e-05, "loss": 1.0148, "step": 7112 }, { "epoch": 0.06397359391652764, "grad_norm": 2.3764917850494385, "learning_rate": 1.5219808239816269e-05, "loss": 1.7843, "step": 7113 }, { "epoch": 0.06398258781416809, "grad_norm": 2.322683572769165, "learning_rate": 1.5212891898757043e-05, "loss": 2.6984, "step": 7114 }, { "epoch": 0.06399158171180853, "grad_norm": 3.4145307540893555, "learning_rate": 1.5205976442246106e-05, "loss": 0.829, "step": 7115 }, { "epoch": 0.064000575609449, "grad_norm": 33.660987854003906, "learning_rate": 1.5199061870908456e-05, "loss": 1.0694, "step": 7116 }, { "epoch": 0.06400956950708944, "grad_norm": 2.534173011779785, "learning_rate": 1.5192148185369048e-05, "loss": 1.8747, "step": 7117 }, { "epoch": 0.06401856340472989, "grad_norm": 2.1169633865356445, "learning_rate": 1.5185235386252717e-05, "loss": 2.4317, "step": 7118 }, { "epoch": 0.06402755730237034, "grad_norm": 2.3483314514160156, "learning_rate": 1.5178323474184248e-05, "loss": 2.2683, "step": 7119 }, { "epoch": 0.0640365512000108, "grad_norm": 25.805248260498047, "learning_rate": 1.517141244978832e-05, "loss": 1.6385, "step": 7120 }, { "epoch": 0.06404554509765124, "grad_norm": 2.0849366188049316, "learning_rate": 1.5164502313689549e-05, "loss": 1.9244, "step": 7121 }, { "epoch": 0.06405453899529169, "grad_norm": 2.8958888053894043, "learning_rate": 1.5157593066512466e-05, "loss": 1.6728, "step": 7122 }, { "epoch": 0.06406353289293215, "grad_norm": 2.2577335834503174, "learning_rate": 1.5150684708881513e-05, "loss": 2.1597, "step": 7123 }, { "epoch": 0.0640725267905726, "grad_norm": 2.7618706226348877, "learning_rate": 1.5143777241421076e-05, "loss": 1.9094, "step": 7124 }, { "epoch": 0.06408152068821305, "grad_norm": 2.4996960163116455, "learning_rate": 1.5136870664755427e-05, "loss": 1.9817, "step": 7125 }, { "epoch": 0.06409051458585349, "grad_norm": 2.074495315551758, "learning_rate": 1.5129964979508792e-05, "loss": 2.2283, "step": 7126 }, { "epoch": 0.06409950848349395, "grad_norm": 1.8286796808242798, "learning_rate": 1.5123060186305291e-05, "loss": 2.3151, "step": 7127 }, { "epoch": 0.0641085023811344, "grad_norm": 1.9239012002944946, "learning_rate": 1.5116156285768968e-05, "loss": 2.1768, "step": 7128 }, { "epoch": 0.06411749627877485, "grad_norm": 1.9778456687927246, "learning_rate": 1.5109253278523799e-05, "loss": 2.0991, "step": 7129 }, { "epoch": 0.06412649017641531, "grad_norm": 1.5320045948028564, "learning_rate": 1.5102351165193653e-05, "loss": 2.2942, "step": 7130 }, { "epoch": 0.06413548407405575, "grad_norm": 1.7614909410476685, "learning_rate": 1.509544994640236e-05, "loss": 2.4898, "step": 7131 }, { "epoch": 0.0641444779716962, "grad_norm": 27.51972770690918, "learning_rate": 1.5088549622773631e-05, "loss": 1.8838, "step": 7132 }, { "epoch": 0.06415347186933666, "grad_norm": 4.370514869689941, "learning_rate": 1.5081650194931102e-05, "loss": 1.3846, "step": 7133 }, { "epoch": 0.06416246576697711, "grad_norm": 2.9464340209960938, "learning_rate": 1.5074751663498358e-05, "loss": 1.1694, "step": 7134 }, { "epoch": 0.06417145966461756, "grad_norm": 2.168503761291504, "learning_rate": 1.506785402909886e-05, "loss": 2.2359, "step": 7135 }, { "epoch": 0.064180453562258, "grad_norm": 2.7048540115356445, "learning_rate": 1.506095729235602e-05, "loss": 2.1701, "step": 7136 }, { "epoch": 0.06418944745989846, "grad_norm": 22.32532501220703, "learning_rate": 1.505406145389316e-05, "loss": 1.1626, "step": 7137 }, { "epoch": 0.06419844135753891, "grad_norm": 2.7903199195861816, "learning_rate": 1.5047166514333499e-05, "loss": 2.0908, "step": 7138 }, { "epoch": 0.06420743525517936, "grad_norm": 2.317195415496826, "learning_rate": 1.5040272474300218e-05, "loss": 2.1243, "step": 7139 }, { "epoch": 0.06421642915281982, "grad_norm": 3.2866880893707275, "learning_rate": 1.5033379334416375e-05, "loss": 1.8792, "step": 7140 }, { "epoch": 0.06422542305046026, "grad_norm": 13.174358367919922, "learning_rate": 1.5026487095304981e-05, "loss": 1.4644, "step": 7141 }, { "epoch": 0.06423441694810071, "grad_norm": 1.9743348360061646, "learning_rate": 1.5019595757588937e-05, "loss": 2.4883, "step": 7142 }, { "epoch": 0.06424341084574116, "grad_norm": 2.735666513442993, "learning_rate": 1.501270532189108e-05, "loss": 1.3278, "step": 7143 }, { "epoch": 0.06425240474338162, "grad_norm": 2.003028392791748, "learning_rate": 1.5005815788834163e-05, "loss": 1.3177, "step": 7144 }, { "epoch": 0.06426139864102207, "grad_norm": 15.080216407775879, "learning_rate": 1.4998927159040838e-05, "loss": 1.0641, "step": 7145 }, { "epoch": 0.06427039253866251, "grad_norm": 1.8097902536392212, "learning_rate": 1.4992039433133714e-05, "loss": 2.7528, "step": 7146 }, { "epoch": 0.06427938643630297, "grad_norm": 13.611855506896973, "learning_rate": 1.4985152611735276e-05, "loss": 1.7198, "step": 7147 }, { "epoch": 0.06428838033394342, "grad_norm": 2.272010564804077, "learning_rate": 1.4978266695467969e-05, "loss": 2.0352, "step": 7148 }, { "epoch": 0.06429737423158387, "grad_norm": 1.4817227125167847, "learning_rate": 1.4971381684954123e-05, "loss": 2.5599, "step": 7149 }, { "epoch": 0.06430636812922431, "grad_norm": 17.364986419677734, "learning_rate": 1.4964497580815995e-05, "loss": 1.6977, "step": 7150 }, { "epoch": 0.06431536202686478, "grad_norm": 6.324827194213867, "learning_rate": 1.495761438367577e-05, "loss": 1.4227, "step": 7151 }, { "epoch": 0.06432435592450522, "grad_norm": 3.3217289447784424, "learning_rate": 1.4950732094155528e-05, "loss": 1.9296, "step": 7152 }, { "epoch": 0.06433334982214567, "grad_norm": 2.3464229106903076, "learning_rate": 1.494385071287731e-05, "loss": 1.9389, "step": 7153 }, { "epoch": 0.06434234371978613, "grad_norm": 1.5508692264556885, "learning_rate": 1.4936970240463025e-05, "loss": 2.4903, "step": 7154 }, { "epoch": 0.06435133761742658, "grad_norm": 1.9412373304367065, "learning_rate": 1.493009067753453e-05, "loss": 2.8383, "step": 7155 }, { "epoch": 0.06436033151506702, "grad_norm": 4.206516742706299, "learning_rate": 1.4923212024713602e-05, "loss": 1.3069, "step": 7156 }, { "epoch": 0.06436932541270748, "grad_norm": 2.564056396484375, "learning_rate": 1.4916334282621907e-05, "loss": 1.9505, "step": 7157 }, { "epoch": 0.06437831931034793, "grad_norm": 3.4418184757232666, "learning_rate": 1.4909457451881065e-05, "loss": 0.9065, "step": 7158 }, { "epoch": 0.06438731320798838, "grad_norm": 1.6705466508865356, "learning_rate": 1.4902581533112581e-05, "loss": 1.8731, "step": 7159 }, { "epoch": 0.06439630710562882, "grad_norm": 2.5022683143615723, "learning_rate": 1.4895706526937914e-05, "loss": 2.3859, "step": 7160 }, { "epoch": 0.06440530100326929, "grad_norm": 2.7944841384887695, "learning_rate": 1.4888832433978403e-05, "loss": 2.1291, "step": 7161 }, { "epoch": 0.06441429490090973, "grad_norm": 18.60266876220703, "learning_rate": 1.4881959254855324e-05, "loss": 1.7291, "step": 7162 }, { "epoch": 0.06442328879855018, "grad_norm": 4.488522052764893, "learning_rate": 1.4875086990189868e-05, "loss": 2.2223, "step": 7163 }, { "epoch": 0.06443228269619064, "grad_norm": 11.14511489868164, "learning_rate": 1.4868215640603144e-05, "loss": 1.1642, "step": 7164 }, { "epoch": 0.06444127659383109, "grad_norm": 2.628340005874634, "learning_rate": 1.4861345206716181e-05, "loss": 2.0914, "step": 7165 }, { "epoch": 0.06445027049147153, "grad_norm": 2.804176092147827, "learning_rate": 1.485447568914991e-05, "loss": 1.8152, "step": 7166 }, { "epoch": 0.06445926438911198, "grad_norm": 1.9006277322769165, "learning_rate": 1.4847607088525208e-05, "loss": 2.5645, "step": 7167 }, { "epoch": 0.06446825828675244, "grad_norm": 19.007688522338867, "learning_rate": 1.4840739405462844e-05, "loss": 0.8382, "step": 7168 }, { "epoch": 0.06447725218439289, "grad_norm": 2.2427115440368652, "learning_rate": 1.4833872640583502e-05, "loss": 1.8906, "step": 7169 }, { "epoch": 0.06448624608203334, "grad_norm": 6.574840545654297, "learning_rate": 1.4827006794507808e-05, "loss": 1.746, "step": 7170 }, { "epoch": 0.0644952399796738, "grad_norm": 2.1611521244049072, "learning_rate": 1.4820141867856269e-05, "loss": 1.9165, "step": 7171 }, { "epoch": 0.06450423387731424, "grad_norm": 2.6925387382507324, "learning_rate": 1.4813277861249358e-05, "loss": 1.4233, "step": 7172 }, { "epoch": 0.06451322777495469, "grad_norm": 3.532763719558716, "learning_rate": 1.4806414775307418e-05, "loss": 0.8674, "step": 7173 }, { "epoch": 0.06452222167259515, "grad_norm": 3.064758777618408, "learning_rate": 1.479955261065072e-05, "loss": 2.1589, "step": 7174 }, { "epoch": 0.0645312155702356, "grad_norm": 3.265244245529175, "learning_rate": 1.4792691367899482e-05, "loss": 1.3263, "step": 7175 }, { "epoch": 0.06454020946787604, "grad_norm": 2.3557612895965576, "learning_rate": 1.4785831047673799e-05, "loss": 2.4774, "step": 7176 }, { "epoch": 0.06454920336551649, "grad_norm": 2.9515275955200195, "learning_rate": 1.4778971650593709e-05, "loss": 2.6876, "step": 7177 }, { "epoch": 0.06455819726315695, "grad_norm": 8.767289161682129, "learning_rate": 1.4772113177279143e-05, "loss": 1.3401, "step": 7178 }, { "epoch": 0.0645671911607974, "grad_norm": 3.1873881816864014, "learning_rate": 1.4765255628349978e-05, "loss": 1.1251, "step": 7179 }, { "epoch": 0.06457618505843785, "grad_norm": 2.5379910469055176, "learning_rate": 1.4758399004425987e-05, "loss": 2.6269, "step": 7180 }, { "epoch": 0.0645851789560783, "grad_norm": 2.0282537937164307, "learning_rate": 1.4751543306126855e-05, "loss": 1.8367, "step": 7181 }, { "epoch": 0.06459417285371875, "grad_norm": 2.7894678115844727, "learning_rate": 1.474468853407221e-05, "loss": 1.8669, "step": 7182 }, { "epoch": 0.0646031667513592, "grad_norm": 11.707415580749512, "learning_rate": 1.4737834688881563e-05, "loss": 1.1473, "step": 7183 }, { "epoch": 0.06461216064899965, "grad_norm": 3.1099376678466797, "learning_rate": 1.4730981771174369e-05, "loss": 2.1995, "step": 7184 }, { "epoch": 0.06462115454664011, "grad_norm": 2.022697925567627, "learning_rate": 1.4724129781569978e-05, "loss": 2.7271, "step": 7185 }, { "epoch": 0.06463014844428056, "grad_norm": 4.246668338775635, "learning_rate": 1.471727872068766e-05, "loss": 0.8403, "step": 7186 }, { "epoch": 0.064639142341921, "grad_norm": 16.263151168823242, "learning_rate": 1.471042858914663e-05, "loss": 1.6794, "step": 7187 }, { "epoch": 0.06464813623956146, "grad_norm": 2.6921253204345703, "learning_rate": 1.470357938756597e-05, "loss": 1.2142, "step": 7188 }, { "epoch": 0.06465713013720191, "grad_norm": 2.7652039527893066, "learning_rate": 1.4696731116564721e-05, "loss": 2.158, "step": 7189 }, { "epoch": 0.06466612403484236, "grad_norm": 2.5495898723602295, "learning_rate": 1.4689883776761815e-05, "loss": 1.7543, "step": 7190 }, { "epoch": 0.0646751179324828, "grad_norm": 2.487795352935791, "learning_rate": 1.468303736877611e-05, "loss": 1.913, "step": 7191 }, { "epoch": 0.06468411183012326, "grad_norm": 8.063586235046387, "learning_rate": 1.4676191893226376e-05, "loss": 1.0703, "step": 7192 }, { "epoch": 0.06469310572776371, "grad_norm": 6.015879154205322, "learning_rate": 1.4669347350731289e-05, "loss": 1.7765, "step": 7193 }, { "epoch": 0.06470209962540416, "grad_norm": 7.342743396759033, "learning_rate": 1.4662503741909473e-05, "loss": 1.4932, "step": 7194 }, { "epoch": 0.06471109352304462, "grad_norm": 2.617896556854248, "learning_rate": 1.465566106737942e-05, "loss": 2.0175, "step": 7195 }, { "epoch": 0.06472008742068507, "grad_norm": 3.546590805053711, "learning_rate": 1.4648819327759588e-05, "loss": 1.1889, "step": 7196 }, { "epoch": 0.06472908131832551, "grad_norm": 3.10819935798645, "learning_rate": 1.464197852366832e-05, "loss": 1.9394, "step": 7197 }, { "epoch": 0.06473807521596597, "grad_norm": 12.691519737243652, "learning_rate": 1.4635138655723868e-05, "loss": 0.4585, "step": 7198 }, { "epoch": 0.06474706911360642, "grad_norm": 2.2071359157562256, "learning_rate": 1.4628299724544425e-05, "loss": 1.5732, "step": 7199 }, { "epoch": 0.06475606301124687, "grad_norm": 3.1042661666870117, "learning_rate": 1.4621461730748075e-05, "loss": 1.5461, "step": 7200 }, { "epoch": 0.06476505690888731, "grad_norm": 1.906923532485962, "learning_rate": 1.4614624674952842e-05, "loss": 1.8117, "step": 7201 }, { "epoch": 0.06477405080652777, "grad_norm": 3.746511459350586, "learning_rate": 1.4607788557776635e-05, "loss": 1.0045, "step": 7202 }, { "epoch": 0.06478304470416822, "grad_norm": 1.9512404203414917, "learning_rate": 1.4600953379837315e-05, "loss": 1.8145, "step": 7203 }, { "epoch": 0.06479203860180867, "grad_norm": 2.3406012058258057, "learning_rate": 1.4594119141752632e-05, "loss": 1.3782, "step": 7204 }, { "epoch": 0.06480103249944913, "grad_norm": 2.680323362350464, "learning_rate": 1.4587285844140239e-05, "loss": 1.6048, "step": 7205 }, { "epoch": 0.06481002639708958, "grad_norm": 25.773448944091797, "learning_rate": 1.4580453487617745e-05, "loss": 1.546, "step": 7206 }, { "epoch": 0.06481902029473002, "grad_norm": 2.2152693271636963, "learning_rate": 1.4573622072802632e-05, "loss": 1.8548, "step": 7207 }, { "epoch": 0.06482801419237047, "grad_norm": 2.5267579555511475, "learning_rate": 1.4566791600312336e-05, "loss": 1.9277, "step": 7208 }, { "epoch": 0.06483700809001093, "grad_norm": 2.048952341079712, "learning_rate": 1.4559962070764177e-05, "loss": 2.5557, "step": 7209 }, { "epoch": 0.06484600198765138, "grad_norm": 2.142747640609741, "learning_rate": 1.4553133484775389e-05, "loss": 2.4224, "step": 7210 }, { "epoch": 0.06485499588529182, "grad_norm": 2.6934332847595215, "learning_rate": 1.4546305842963156e-05, "loss": 1.0828, "step": 7211 }, { "epoch": 0.06486398978293229, "grad_norm": 2.2142257690429688, "learning_rate": 1.4539479145944532e-05, "loss": 2.63, "step": 7212 }, { "epoch": 0.06487298368057273, "grad_norm": 2.3251419067382812, "learning_rate": 1.4532653394336526e-05, "loss": 2.3465, "step": 7213 }, { "epoch": 0.06488197757821318, "grad_norm": 3.1577024459838867, "learning_rate": 1.452582858875603e-05, "loss": 1.6797, "step": 7214 }, { "epoch": 0.06489097147585364, "grad_norm": 1.9250415563583374, "learning_rate": 1.4519004729819863e-05, "loss": 1.9057, "step": 7215 }, { "epoch": 0.06489996537349409, "grad_norm": 2.6029698848724365, "learning_rate": 1.4512181818144763e-05, "loss": 2.219, "step": 7216 }, { "epoch": 0.06490895927113453, "grad_norm": 2.9124090671539307, "learning_rate": 1.4505359854347361e-05, "loss": 1.3931, "step": 7217 }, { "epoch": 0.06491795316877498, "grad_norm": 25.28011703491211, "learning_rate": 1.4498538839044243e-05, "loss": 1.6308, "step": 7218 }, { "epoch": 0.06492694706641544, "grad_norm": 2.9577879905700684, "learning_rate": 1.4491718772851861e-05, "loss": 2.0967, "step": 7219 }, { "epoch": 0.06493594096405589, "grad_norm": 15.822932243347168, "learning_rate": 1.448489965638663e-05, "loss": 1.769, "step": 7220 }, { "epoch": 0.06494493486169634, "grad_norm": 2.699000120162964, "learning_rate": 1.4478081490264841e-05, "loss": 1.3492, "step": 7221 }, { "epoch": 0.0649539287593368, "grad_norm": 1.8944684267044067, "learning_rate": 1.4471264275102706e-05, "loss": 2.1242, "step": 7222 }, { "epoch": 0.06496292265697724, "grad_norm": 9.56735897064209, "learning_rate": 1.4464448011516373e-05, "loss": 0.7703, "step": 7223 }, { "epoch": 0.06497191655461769, "grad_norm": 2.2507500648498535, "learning_rate": 1.445763270012187e-05, "loss": 1.9054, "step": 7224 }, { "epoch": 0.06498091045225814, "grad_norm": 2.4835989475250244, "learning_rate": 1.445081834153518e-05, "loss": 1.3678, "step": 7225 }, { "epoch": 0.0649899043498986, "grad_norm": 1.8652349710464478, "learning_rate": 1.4444004936372165e-05, "loss": 2.4097, "step": 7226 }, { "epoch": 0.06499889824753904, "grad_norm": 2.3474395275115967, "learning_rate": 1.4437192485248605e-05, "loss": 2.0317, "step": 7227 }, { "epoch": 0.06500789214517949, "grad_norm": 2.879873275756836, "learning_rate": 1.4430380988780218e-05, "loss": 1.7465, "step": 7228 }, { "epoch": 0.06501688604281995, "grad_norm": 1.8781880140304565, "learning_rate": 1.4423570447582616e-05, "loss": 2.5886, "step": 7229 }, { "epoch": 0.0650258799404604, "grad_norm": 2.8615472316741943, "learning_rate": 1.4416760862271322e-05, "loss": 1.4603, "step": 7230 }, { "epoch": 0.06503487383810085, "grad_norm": 26.44091033935547, "learning_rate": 1.4409952233461777e-05, "loss": 1.5769, "step": 7231 }, { "epoch": 0.0650438677357413, "grad_norm": 1.749017596244812, "learning_rate": 1.4403144561769347e-05, "loss": 1.8413, "step": 7232 }, { "epoch": 0.06505286163338175, "grad_norm": 8.014484405517578, "learning_rate": 1.43963378478093e-05, "loss": 1.8727, "step": 7233 }, { "epoch": 0.0650618555310222, "grad_norm": 2.9849650859832764, "learning_rate": 1.4389532092196806e-05, "loss": 2.2715, "step": 7234 }, { "epoch": 0.06507084942866265, "grad_norm": 12.942831039428711, "learning_rate": 1.4382727295546983e-05, "loss": 1.0112, "step": 7235 }, { "epoch": 0.06507984332630311, "grad_norm": 1.9260118007659912, "learning_rate": 1.4375923458474821e-05, "loss": 2.423, "step": 7236 }, { "epoch": 0.06508883722394355, "grad_norm": 2.2176833152770996, "learning_rate": 1.4369120581595264e-05, "loss": 1.8215, "step": 7237 }, { "epoch": 0.065097831121584, "grad_norm": 8.55763053894043, "learning_rate": 1.4362318665523139e-05, "loss": 1.2586, "step": 7238 }, { "epoch": 0.06510682501922446, "grad_norm": 1.8093565702438354, "learning_rate": 1.4355517710873184e-05, "loss": 2.617, "step": 7239 }, { "epoch": 0.06511581891686491, "grad_norm": 3.650974988937378, "learning_rate": 1.434871771826008e-05, "loss": 0.9183, "step": 7240 }, { "epoch": 0.06512481281450536, "grad_norm": 15.136865615844727, "learning_rate": 1.4341918688298388e-05, "loss": 1.8072, "step": 7241 }, { "epoch": 0.0651338067121458, "grad_norm": 16.135677337646484, "learning_rate": 1.4335120621602619e-05, "loss": 1.0, "step": 7242 }, { "epoch": 0.06514280060978626, "grad_norm": 15.449339866638184, "learning_rate": 1.4328323518787158e-05, "loss": 1.6645, "step": 7243 }, { "epoch": 0.06515179450742671, "grad_norm": 6.9825849533081055, "learning_rate": 1.4321527380466321e-05, "loss": 2.0534, "step": 7244 }, { "epoch": 0.06516078840506716, "grad_norm": 3.103581190109253, "learning_rate": 1.4314732207254339e-05, "loss": 1.922, "step": 7245 }, { "epoch": 0.06516978230270762, "grad_norm": 2.820382595062256, "learning_rate": 1.4307937999765341e-05, "loss": 1.2967, "step": 7246 }, { "epoch": 0.06517877620034807, "grad_norm": 4.233402729034424, "learning_rate": 1.43011447586134e-05, "loss": 0.8907, "step": 7247 }, { "epoch": 0.06518777009798851, "grad_norm": 3.0921506881713867, "learning_rate": 1.4294352484412463e-05, "loss": 1.6451, "step": 7248 }, { "epoch": 0.06519676399562896, "grad_norm": 13.504074096679688, "learning_rate": 1.4287561177776425e-05, "loss": 0.5496, "step": 7249 }, { "epoch": 0.06520575789326942, "grad_norm": 2.5957345962524414, "learning_rate": 1.4280770839319071e-05, "loss": 2.1721, "step": 7250 }, { "epoch": 0.06521475179090987, "grad_norm": 2.680569887161255, "learning_rate": 1.4273981469654093e-05, "loss": 1.7759, "step": 7251 }, { "epoch": 0.06522374568855031, "grad_norm": 2.547994613647461, "learning_rate": 1.4267193069395126e-05, "loss": 1.5927, "step": 7252 }, { "epoch": 0.06523273958619077, "grad_norm": 2.0627989768981934, "learning_rate": 1.4260405639155677e-05, "loss": 2.6422, "step": 7253 }, { "epoch": 0.06524173348383122, "grad_norm": 2.327659845352173, "learning_rate": 1.4253619179549213e-05, "loss": 1.7148, "step": 7254 }, { "epoch": 0.06525072738147167, "grad_norm": 2.020167112350464, "learning_rate": 1.424683369118907e-05, "loss": 1.5771, "step": 7255 }, { "epoch": 0.06525972127911213, "grad_norm": 6.751526355743408, "learning_rate": 1.4240049174688514e-05, "loss": 1.5653, "step": 7256 }, { "epoch": 0.06526871517675258, "grad_norm": 3.1241369247436523, "learning_rate": 1.4233265630660716e-05, "loss": 1.2019, "step": 7257 }, { "epoch": 0.06527770907439302, "grad_norm": 2.4903204441070557, "learning_rate": 1.4226483059718781e-05, "loss": 2.0264, "step": 7258 }, { "epoch": 0.06528670297203347, "grad_norm": 18.840492248535156, "learning_rate": 1.4219701462475706e-05, "loss": 1.0171, "step": 7259 }, { "epoch": 0.06529569686967393, "grad_norm": 2.86942458152771, "learning_rate": 1.4212920839544392e-05, "loss": 1.9678, "step": 7260 }, { "epoch": 0.06530469076731438, "grad_norm": 12.284415245056152, "learning_rate": 1.4206141191537682e-05, "loss": 1.1265, "step": 7261 }, { "epoch": 0.06531368466495482, "grad_norm": 2.3154706954956055, "learning_rate": 1.4199362519068305e-05, "loss": 1.8294, "step": 7262 }, { "epoch": 0.06532267856259528, "grad_norm": 16.916719436645508, "learning_rate": 1.4192584822748897e-05, "loss": 1.3151, "step": 7263 }, { "epoch": 0.06533167246023573, "grad_norm": 3.1445419788360596, "learning_rate": 1.4185808103192044e-05, "loss": 2.5375, "step": 7264 }, { "epoch": 0.06534066635787618, "grad_norm": 2.118372917175293, "learning_rate": 1.4179032361010198e-05, "loss": 2.8715, "step": 7265 }, { "epoch": 0.06534966025551663, "grad_norm": 1.961380124092102, "learning_rate": 1.4172257596815761e-05, "loss": 1.4381, "step": 7266 }, { "epoch": 0.06535865415315709, "grad_norm": 2.762683629989624, "learning_rate": 1.4165483811221008e-05, "loss": 1.2984, "step": 7267 }, { "epoch": 0.06536764805079753, "grad_norm": 2.617133378982544, "learning_rate": 1.4158711004838171e-05, "loss": 2.5019, "step": 7268 }, { "epoch": 0.06537664194843798, "grad_norm": 2.5514116287231445, "learning_rate": 1.4151939178279355e-05, "loss": 1.4752, "step": 7269 }, { "epoch": 0.06538563584607844, "grad_norm": 2.983217716217041, "learning_rate": 1.4145168332156591e-05, "loss": 2.0006, "step": 7270 }, { "epoch": 0.06539462974371889, "grad_norm": 2.3705365657806396, "learning_rate": 1.413839846708182e-05, "loss": 2.347, "step": 7271 }, { "epoch": 0.06540362364135933, "grad_norm": 3.0273239612579346, "learning_rate": 1.4131629583666888e-05, "loss": 1.197, "step": 7272 }, { "epoch": 0.0654126175389998, "grad_norm": 9.69559097290039, "learning_rate": 1.412486168252358e-05, "loss": 1.4427, "step": 7273 }, { "epoch": 0.06542161143664024, "grad_norm": 2.2493896484375, "learning_rate": 1.4118094764263556e-05, "loss": 2.3326, "step": 7274 }, { "epoch": 0.06543060533428069, "grad_norm": 2.07794451713562, "learning_rate": 1.41113288294984e-05, "loss": 1.8342, "step": 7275 }, { "epoch": 0.06543959923192114, "grad_norm": 18.926271438598633, "learning_rate": 1.4104563878839621e-05, "loss": 1.6756, "step": 7276 }, { "epoch": 0.0654485931295616, "grad_norm": 2.157378911972046, "learning_rate": 1.4097799912898615e-05, "loss": 2.278, "step": 7277 }, { "epoch": 0.06545758702720204, "grad_norm": 1.6683062314987183, "learning_rate": 1.4091036932286721e-05, "loss": 2.4241, "step": 7278 }, { "epoch": 0.06546658092484249, "grad_norm": 2.4996492862701416, "learning_rate": 1.408427493761515e-05, "loss": 1.6831, "step": 7279 }, { "epoch": 0.06547557482248295, "grad_norm": 2.0474095344543457, "learning_rate": 1.407751392949506e-05, "loss": 2.0491, "step": 7280 }, { "epoch": 0.0654845687201234, "grad_norm": 2.117410182952881, "learning_rate": 1.4070753908537498e-05, "loss": 1.7877, "step": 7281 }, { "epoch": 0.06549356261776385, "grad_norm": 1.850964069366455, "learning_rate": 1.4063994875353414e-05, "loss": 2.6038, "step": 7282 }, { "epoch": 0.06550255651540429, "grad_norm": 2.4589288234710693, "learning_rate": 1.4057236830553704e-05, "loss": 2.228, "step": 7283 }, { "epoch": 0.06551155041304475, "grad_norm": 2.1605210304260254, "learning_rate": 1.4050479774749142e-05, "loss": 2.2794, "step": 7284 }, { "epoch": 0.0655205443106852, "grad_norm": 2.9277408123016357, "learning_rate": 1.4043723708550426e-05, "loss": 1.7623, "step": 7285 }, { "epoch": 0.06552953820832565, "grad_norm": 1.7544682025909424, "learning_rate": 1.4036968632568163e-05, "loss": 2.5372, "step": 7286 }, { "epoch": 0.06553853210596611, "grad_norm": 16.888607025146484, "learning_rate": 1.4030214547412852e-05, "loss": 1.2142, "step": 7287 }, { "epoch": 0.06554752600360655, "grad_norm": 1.791275978088379, "learning_rate": 1.4023461453694947e-05, "loss": 2.579, "step": 7288 }, { "epoch": 0.065556519901247, "grad_norm": 22.057205200195312, "learning_rate": 1.4016709352024759e-05, "loss": 1.5794, "step": 7289 }, { "epoch": 0.06556551379888745, "grad_norm": 3.697232961654663, "learning_rate": 1.4009958243012563e-05, "loss": 1.5292, "step": 7290 }, { "epoch": 0.06557450769652791, "grad_norm": 2.2854814529418945, "learning_rate": 1.4003208127268503e-05, "loss": 2.1003, "step": 7291 }, { "epoch": 0.06558350159416836, "grad_norm": 3.414905548095703, "learning_rate": 1.399645900540264e-05, "loss": 1.475, "step": 7292 }, { "epoch": 0.0655924954918088, "grad_norm": 3.748809814453125, "learning_rate": 1.3989710878024965e-05, "loss": 0.9849, "step": 7293 }, { "epoch": 0.06560148938944926, "grad_norm": 2.875699520111084, "learning_rate": 1.3982963745745352e-05, "loss": 1.4496, "step": 7294 }, { "epoch": 0.06561048328708971, "grad_norm": 8.257377624511719, "learning_rate": 1.3976217609173626e-05, "loss": 2.0648, "step": 7295 }, { "epoch": 0.06561947718473016, "grad_norm": 15.016207695007324, "learning_rate": 1.3969472468919461e-05, "loss": 1.2349, "step": 7296 }, { "epoch": 0.06562847108237062, "grad_norm": 2.6840832233428955, "learning_rate": 1.3962728325592505e-05, "loss": 1.9109, "step": 7297 }, { "epoch": 0.06563746498001106, "grad_norm": 2.2667829990386963, "learning_rate": 1.3955985179802278e-05, "loss": 1.9407, "step": 7298 }, { "epoch": 0.06564645887765151, "grad_norm": 2.201249599456787, "learning_rate": 1.394924303215821e-05, "loss": 2.887, "step": 7299 }, { "epoch": 0.06565545277529196, "grad_norm": 1.972356915473938, "learning_rate": 1.3942501883269659e-05, "loss": 2.6602, "step": 7300 }, { "epoch": 0.06566444667293242, "grad_norm": 1.8010921478271484, "learning_rate": 1.3935761733745865e-05, "loss": 2.52, "step": 7301 }, { "epoch": 0.06567344057057287, "grad_norm": 2.2631280422210693, "learning_rate": 1.3929022584196017e-05, "loss": 2.3423, "step": 7302 }, { "epoch": 0.06568243446821331, "grad_norm": 2.416001796722412, "learning_rate": 1.3922284435229183e-05, "loss": 1.8096, "step": 7303 }, { "epoch": 0.06569142836585377, "grad_norm": 2.1106414794921875, "learning_rate": 1.3915547287454342e-05, "loss": 2.3659, "step": 7304 }, { "epoch": 0.06570042226349422, "grad_norm": 3.4338419437408447, "learning_rate": 1.3908811141480408e-05, "loss": 1.0661, "step": 7305 }, { "epoch": 0.06570941616113467, "grad_norm": 2.5574376583099365, "learning_rate": 1.3902075997916165e-05, "loss": 2.0544, "step": 7306 }, { "epoch": 0.06571841005877511, "grad_norm": 2.4188098907470703, "learning_rate": 1.3895341857370347e-05, "loss": 2.6483, "step": 7307 }, { "epoch": 0.06572740395641558, "grad_norm": 4.950438022613525, "learning_rate": 1.3888608720451567e-05, "loss": 1.3266, "step": 7308 }, { "epoch": 0.06573639785405602, "grad_norm": 2.5879526138305664, "learning_rate": 1.3881876587768369e-05, "loss": 1.6013, "step": 7309 }, { "epoch": 0.06574539175169647, "grad_norm": 2.901867151260376, "learning_rate": 1.387514545992919e-05, "loss": 1.9605, "step": 7310 }, { "epoch": 0.06575438564933693, "grad_norm": 2.0366432666778564, "learning_rate": 1.386841533754238e-05, "loss": 2.5904, "step": 7311 }, { "epoch": 0.06576337954697738, "grad_norm": 1.7611061334609985, "learning_rate": 1.3861686221216196e-05, "loss": 2.8593, "step": 7312 }, { "epoch": 0.06577237344461782, "grad_norm": 2.6881654262542725, "learning_rate": 1.3854958111558822e-05, "loss": 1.9501, "step": 7313 }, { "epoch": 0.06578136734225828, "grad_norm": 20.363847732543945, "learning_rate": 1.3848231009178331e-05, "loss": 1.3362, "step": 7314 }, { "epoch": 0.06579036123989873, "grad_norm": 2.3472909927368164, "learning_rate": 1.3841504914682706e-05, "loss": 1.6769, "step": 7315 }, { "epoch": 0.06579935513753918, "grad_norm": 1.8644871711730957, "learning_rate": 1.383477982867984e-05, "loss": 2.0218, "step": 7316 }, { "epoch": 0.06580834903517963, "grad_norm": 2.1275253295898438, "learning_rate": 1.3828055751777558e-05, "loss": 2.4044, "step": 7317 }, { "epoch": 0.06581734293282009, "grad_norm": 2.357118606567383, "learning_rate": 1.382133268458355e-05, "loss": 1.8066, "step": 7318 }, { "epoch": 0.06582633683046053, "grad_norm": 2.7896718978881836, "learning_rate": 1.3814610627705465e-05, "loss": 1.5646, "step": 7319 }, { "epoch": 0.06583533072810098, "grad_norm": 18.563817977905273, "learning_rate": 1.3807889581750816e-05, "loss": 0.8885, "step": 7320 }, { "epoch": 0.06584432462574144, "grad_norm": 1.7681366205215454, "learning_rate": 1.3801169547327059e-05, "loss": 1.7067, "step": 7321 }, { "epoch": 0.06585331852338189, "grad_norm": 1.8607304096221924, "learning_rate": 1.3794450525041536e-05, "loss": 2.5152, "step": 7322 }, { "epoch": 0.06586231242102233, "grad_norm": 2.6187779903411865, "learning_rate": 1.3787732515501498e-05, "loss": 2.4943, "step": 7323 }, { "epoch": 0.06587130631866278, "grad_norm": 20.09831428527832, "learning_rate": 1.3781015519314125e-05, "loss": 0.7989, "step": 7324 }, { "epoch": 0.06588030021630324, "grad_norm": 2.1458120346069336, "learning_rate": 1.3774299537086489e-05, "loss": 2.001, "step": 7325 }, { "epoch": 0.06588929411394369, "grad_norm": 16.635454177856445, "learning_rate": 1.3767584569425562e-05, "loss": 1.4103, "step": 7326 }, { "epoch": 0.06589828801158414, "grad_norm": 13.330873489379883, "learning_rate": 1.3760870616938248e-05, "loss": 1.7646, "step": 7327 }, { "epoch": 0.0659072819092246, "grad_norm": 2.4919869899749756, "learning_rate": 1.3754157680231333e-05, "loss": 1.2861, "step": 7328 }, { "epoch": 0.06591627580686504, "grad_norm": 20.062164306640625, "learning_rate": 1.3747445759911543e-05, "loss": 1.5257, "step": 7329 }, { "epoch": 0.06592526970450549, "grad_norm": 2.275524854660034, "learning_rate": 1.3740734856585475e-05, "loss": 1.3624, "step": 7330 }, { "epoch": 0.06593426360214594, "grad_norm": 2.251749038696289, "learning_rate": 1.373402497085967e-05, "loss": 1.8825, "step": 7331 }, { "epoch": 0.0659432574997864, "grad_norm": 2.4817285537719727, "learning_rate": 1.3727316103340549e-05, "loss": 3.0386, "step": 7332 }, { "epoch": 0.06595225139742684, "grad_norm": 11.646502494812012, "learning_rate": 1.3720608254634462e-05, "loss": 1.6438, "step": 7333 }, { "epoch": 0.06596124529506729, "grad_norm": 2.6431822776794434, "learning_rate": 1.3713901425347652e-05, "loss": 1.6207, "step": 7334 }, { "epoch": 0.06597023919270775, "grad_norm": 29.859729766845703, "learning_rate": 1.3707195616086268e-05, "loss": 0.8555, "step": 7335 }, { "epoch": 0.0659792330903482, "grad_norm": 2.5260865688323975, "learning_rate": 1.3700490827456391e-05, "loss": 1.521, "step": 7336 }, { "epoch": 0.06598822698798865, "grad_norm": 1.8957929611206055, "learning_rate": 1.3693787060063967e-05, "loss": 1.9806, "step": 7337 }, { "epoch": 0.06599722088562911, "grad_norm": 2.1439948081970215, "learning_rate": 1.3687084314514907e-05, "loss": 2.5588, "step": 7338 }, { "epoch": 0.06600621478326955, "grad_norm": 2.180783271789551, "learning_rate": 1.3680382591414975e-05, "loss": 2.8639, "step": 7339 }, { "epoch": 0.06601520868091, "grad_norm": 16.442365646362305, "learning_rate": 1.3673681891369878e-05, "loss": 1.1083, "step": 7340 }, { "epoch": 0.06602420257855045, "grad_norm": 21.699871063232422, "learning_rate": 1.3666982214985208e-05, "loss": 1.6259, "step": 7341 }, { "epoch": 0.06603319647619091, "grad_norm": 14.942243576049805, "learning_rate": 1.3660283562866465e-05, "loss": 0.8932, "step": 7342 }, { "epoch": 0.06604219037383136, "grad_norm": 1.8684356212615967, "learning_rate": 1.3653585935619095e-05, "loss": 2.7074, "step": 7343 }, { "epoch": 0.0660511842714718, "grad_norm": 2.714939594268799, "learning_rate": 1.3646889333848398e-05, "loss": 2.3177, "step": 7344 }, { "epoch": 0.06606017816911226, "grad_norm": 2.37528657913208, "learning_rate": 1.3640193758159619e-05, "loss": 1.7034, "step": 7345 }, { "epoch": 0.06606917206675271, "grad_norm": 2.6503753662109375, "learning_rate": 1.3633499209157896e-05, "loss": 2.0569, "step": 7346 }, { "epoch": 0.06607816596439316, "grad_norm": 2.1705965995788574, "learning_rate": 1.3626805687448262e-05, "loss": 2.4408, "step": 7347 }, { "epoch": 0.0660871598620336, "grad_norm": 3.061659574508667, "learning_rate": 1.3620113193635688e-05, "loss": 1.5963, "step": 7348 }, { "epoch": 0.06609615375967406, "grad_norm": 3.854473829269409, "learning_rate": 1.3613421728325018e-05, "loss": 1.4947, "step": 7349 }, { "epoch": 0.06610514765731451, "grad_norm": 2.600071907043457, "learning_rate": 1.3606731292121036e-05, "loss": 2.5819, "step": 7350 }, { "epoch": 0.06611414155495496, "grad_norm": 20.439876556396484, "learning_rate": 1.3600041885628409e-05, "loss": 1.7366, "step": 7351 }, { "epoch": 0.06612313545259542, "grad_norm": 2.429896354675293, "learning_rate": 1.3593353509451706e-05, "loss": 2.0034, "step": 7352 }, { "epoch": 0.06613212935023587, "grad_norm": 2.0705246925354004, "learning_rate": 1.3586666164195438e-05, "loss": 2.1422, "step": 7353 }, { "epoch": 0.06614112324787631, "grad_norm": 2.7265846729278564, "learning_rate": 1.3579979850463986e-05, "loss": 1.3898, "step": 7354 }, { "epoch": 0.06615011714551677, "grad_norm": 25.28412628173828, "learning_rate": 1.3573294568861659e-05, "loss": 0.8527, "step": 7355 }, { "epoch": 0.06615911104315722, "grad_norm": 18.61055564880371, "learning_rate": 1.3566610319992656e-05, "loss": 0.4805, "step": 7356 }, { "epoch": 0.06616810494079767, "grad_norm": 11.594995498657227, "learning_rate": 1.355992710446109e-05, "loss": 0.5669, "step": 7357 }, { "epoch": 0.06617709883843811, "grad_norm": 13.945464134216309, "learning_rate": 1.3553244922871e-05, "loss": 1.1149, "step": 7358 }, { "epoch": 0.06618609273607858, "grad_norm": 2.4399380683898926, "learning_rate": 1.3546563775826293e-05, "loss": 1.4872, "step": 7359 }, { "epoch": 0.06619508663371902, "grad_norm": 5.053860664367676, "learning_rate": 1.353988366393083e-05, "loss": 0.6283, "step": 7360 }, { "epoch": 0.06620408053135947, "grad_norm": 3.352267026901245, "learning_rate": 1.3533204587788323e-05, "loss": 1.5243, "step": 7361 }, { "epoch": 0.06621307442899993, "grad_norm": 2.993485927581787, "learning_rate": 1.3526526548002444e-05, "loss": 1.5107, "step": 7362 }, { "epoch": 0.06622206832664038, "grad_norm": 10.216487884521484, "learning_rate": 1.3519849545176739e-05, "loss": 0.852, "step": 7363 }, { "epoch": 0.06623106222428082, "grad_norm": 16.0078067779541, "learning_rate": 1.3513173579914657e-05, "loss": 0.3537, "step": 7364 }, { "epoch": 0.06624005612192127, "grad_norm": 2.4158687591552734, "learning_rate": 1.3506498652819583e-05, "loss": 2.3401, "step": 7365 }, { "epoch": 0.06624905001956173, "grad_norm": 3.4030745029449463, "learning_rate": 1.3499824764494773e-05, "loss": 1.8174, "step": 7366 }, { "epoch": 0.06625804391720218, "grad_norm": 2.940709352493286, "learning_rate": 1.3493151915543422e-05, "loss": 2.0537, "step": 7367 }, { "epoch": 0.06626703781484262, "grad_norm": 2.336372137069702, "learning_rate": 1.3486480106568607e-05, "loss": 2.6951, "step": 7368 }, { "epoch": 0.06627603171248309, "grad_norm": 17.95156478881836, "learning_rate": 1.3479809338173322e-05, "loss": 1.0976, "step": 7369 }, { "epoch": 0.06628502561012353, "grad_norm": 2.3232758045196533, "learning_rate": 1.3473139610960455e-05, "loss": 2.6817, "step": 7370 }, { "epoch": 0.06629401950776398, "grad_norm": 14.135737419128418, "learning_rate": 1.346647092553281e-05, "loss": 1.6228, "step": 7371 }, { "epoch": 0.06630301340540443, "grad_norm": 3.2749624252319336, "learning_rate": 1.3459803282493105e-05, "loss": 1.5858, "step": 7372 }, { "epoch": 0.06631200730304489, "grad_norm": 3.2507405281066895, "learning_rate": 1.3453136682443939e-05, "loss": 1.5147, "step": 7373 }, { "epoch": 0.06632100120068533, "grad_norm": 3.297355890274048, "learning_rate": 1.3446471125987853e-05, "loss": 1.3905, "step": 7374 }, { "epoch": 0.06632999509832578, "grad_norm": 2.5735931396484375, "learning_rate": 1.3439806613727266e-05, "loss": 2.095, "step": 7375 }, { "epoch": 0.06633898899596624, "grad_norm": 3.7717723846435547, "learning_rate": 1.3433143146264493e-05, "loss": 0.862, "step": 7376 }, { "epoch": 0.06634798289360669, "grad_norm": 2.6106808185577393, "learning_rate": 1.3426480724201792e-05, "loss": 1.8628, "step": 7377 }, { "epoch": 0.06635697679124714, "grad_norm": 10.38595199584961, "learning_rate": 1.3419819348141289e-05, "loss": 0.8228, "step": 7378 }, { "epoch": 0.0663659706888876, "grad_norm": 2.322053909301758, "learning_rate": 1.341315901868505e-05, "loss": 1.9468, "step": 7379 }, { "epoch": 0.06637496458652804, "grad_norm": 3.2721023559570312, "learning_rate": 1.3406499736435019e-05, "loss": 1.494, "step": 7380 }, { "epoch": 0.06638395848416849, "grad_norm": 12.453540802001953, "learning_rate": 1.3399841501993055e-05, "loss": 1.0565, "step": 7381 }, { "epoch": 0.06639295238180894, "grad_norm": 10.103490829467773, "learning_rate": 1.3393184315960918e-05, "loss": 1.6723, "step": 7382 }, { "epoch": 0.0664019462794494, "grad_norm": 2.640012741088867, "learning_rate": 1.3386528178940275e-05, "loss": 1.7938, "step": 7383 }, { "epoch": 0.06641094017708984, "grad_norm": 3.008842945098877, "learning_rate": 1.3379873091532713e-05, "loss": 2.1077, "step": 7384 }, { "epoch": 0.06641993407473029, "grad_norm": 10.735822677612305, "learning_rate": 1.3373219054339697e-05, "loss": 1.7838, "step": 7385 }, { "epoch": 0.06642892797237075, "grad_norm": 2.0581045150756836, "learning_rate": 1.3366566067962627e-05, "loss": 2.6464, "step": 7386 }, { "epoch": 0.0664379218700112, "grad_norm": 9.040849685668945, "learning_rate": 1.3359914133002788e-05, "loss": 1.1607, "step": 7387 }, { "epoch": 0.06644691576765165, "grad_norm": 2.467780351638794, "learning_rate": 1.335326325006136e-05, "loss": 2.2017, "step": 7388 }, { "epoch": 0.06645590966529209, "grad_norm": 16.44783592224121, "learning_rate": 1.3346613419739468e-05, "loss": 1.4227, "step": 7389 }, { "epoch": 0.06646490356293255, "grad_norm": 2.927867889404297, "learning_rate": 1.3339964642638089e-05, "loss": 1.9413, "step": 7390 }, { "epoch": 0.066473897460573, "grad_norm": 2.898233413696289, "learning_rate": 1.3333316919358157e-05, "loss": 2.0513, "step": 7391 }, { "epoch": 0.06648289135821345, "grad_norm": 2.209581136703491, "learning_rate": 1.3326670250500479e-05, "loss": 1.559, "step": 7392 }, { "epoch": 0.06649188525585391, "grad_norm": 7.386668682098389, "learning_rate": 1.3320024636665757e-05, "loss": 0.9522, "step": 7393 }, { "epoch": 0.06650087915349435, "grad_norm": 2.235161542892456, "learning_rate": 1.3313380078454637e-05, "loss": 2.0808, "step": 7394 }, { "epoch": 0.0665098730511348, "grad_norm": 4.26155948638916, "learning_rate": 1.330673657646764e-05, "loss": 0.7718, "step": 7395 }, { "epoch": 0.06651886694877526, "grad_norm": 3.4046242237091064, "learning_rate": 1.3300094131305196e-05, "loss": 1.5636, "step": 7396 }, { "epoch": 0.06652786084641571, "grad_norm": 4.448303699493408, "learning_rate": 1.329345274356763e-05, "loss": 0.969, "step": 7397 }, { "epoch": 0.06653685474405616, "grad_norm": 10.55585765838623, "learning_rate": 1.328681241385521e-05, "loss": 1.3862, "step": 7398 }, { "epoch": 0.0665458486416966, "grad_norm": 13.543828964233398, "learning_rate": 1.3280173142768065e-05, "loss": 1.5263, "step": 7399 }, { "epoch": 0.06655484253933706, "grad_norm": 29.993446350097656, "learning_rate": 1.327353493090624e-05, "loss": 0.8242, "step": 7400 }, { "epoch": 0.06656383643697751, "grad_norm": 14.158581733703613, "learning_rate": 1.3266897778869702e-05, "loss": 1.6452, "step": 7401 }, { "epoch": 0.06657283033461796, "grad_norm": 6.051938056945801, "learning_rate": 1.3260261687258299e-05, "loss": 0.4921, "step": 7402 }, { "epoch": 0.06658182423225842, "grad_norm": 11.187843322753906, "learning_rate": 1.3253626656671809e-05, "loss": 1.7483, "step": 7403 }, { "epoch": 0.06659081812989887, "grad_norm": 8.11154842376709, "learning_rate": 1.3246992687709889e-05, "loss": 1.1311, "step": 7404 }, { "epoch": 0.06659981202753931, "grad_norm": 2.1741347312927246, "learning_rate": 1.32403597809721e-05, "loss": 2.4914, "step": 7405 }, { "epoch": 0.06660880592517976, "grad_norm": 2.8336918354034424, "learning_rate": 1.3233727937057939e-05, "loss": 1.9984, "step": 7406 }, { "epoch": 0.06661779982282022, "grad_norm": 12.559601783752441, "learning_rate": 1.3227097156566765e-05, "loss": 1.6011, "step": 7407 }, { "epoch": 0.06662679372046067, "grad_norm": 12.900568962097168, "learning_rate": 1.3220467440097878e-05, "loss": 1.7171, "step": 7408 }, { "epoch": 0.06663578761810111, "grad_norm": 2.679935932159424, "learning_rate": 1.3213838788250454e-05, "loss": 2.944, "step": 7409 }, { "epoch": 0.06664478151574157, "grad_norm": 7.543893337249756, "learning_rate": 1.3207211201623587e-05, "loss": 0.8365, "step": 7410 }, { "epoch": 0.06665377541338202, "grad_norm": 2.667296886444092, "learning_rate": 1.3200584680816268e-05, "loss": 2.0089, "step": 7411 }, { "epoch": 0.06666276931102247, "grad_norm": 2.3266844749450684, "learning_rate": 1.3193959226427388e-05, "loss": 1.944, "step": 7412 }, { "epoch": 0.06667176320866292, "grad_norm": 4.174129962921143, "learning_rate": 1.3187334839055765e-05, "loss": 1.9954, "step": 7413 }, { "epoch": 0.06668075710630338, "grad_norm": 4.220705032348633, "learning_rate": 1.3180711519300087e-05, "loss": 1.6304, "step": 7414 }, { "epoch": 0.06668975100394382, "grad_norm": 2.809621572494507, "learning_rate": 1.3174089267758983e-05, "loss": 1.9968, "step": 7415 }, { "epoch": 0.06669874490158427, "grad_norm": 2.8889994621276855, "learning_rate": 1.3167468085030949e-05, "loss": 2.119, "step": 7416 }, { "epoch": 0.06670773879922473, "grad_norm": 4.552759647369385, "learning_rate": 1.3160847971714396e-05, "loss": 0.9192, "step": 7417 }, { "epoch": 0.06671673269686518, "grad_norm": 2.463593006134033, "learning_rate": 1.3154228928407664e-05, "loss": 2.0798, "step": 7418 }, { "epoch": 0.06672572659450562, "grad_norm": 3.701843023300171, "learning_rate": 1.314761095570895e-05, "loss": 1.0529, "step": 7419 }, { "epoch": 0.06673472049214609, "grad_norm": 7.371886253356934, "learning_rate": 1.3140994054216399e-05, "loss": 0.657, "step": 7420 }, { "epoch": 0.06674371438978653, "grad_norm": 7.950949192047119, "learning_rate": 1.3134378224528027e-05, "loss": 1.0141, "step": 7421 }, { "epoch": 0.06675270828742698, "grad_norm": 2.2003672122955322, "learning_rate": 1.3127763467241777e-05, "loss": 2.2862, "step": 7422 }, { "epoch": 0.06676170218506743, "grad_norm": 4.835981845855713, "learning_rate": 1.3121149782955477e-05, "loss": 0.6769, "step": 7423 }, { "epoch": 0.06677069608270789, "grad_norm": 3.959409236907959, "learning_rate": 1.3114537172266866e-05, "loss": 0.9888, "step": 7424 }, { "epoch": 0.06677968998034833, "grad_norm": 4.713810443878174, "learning_rate": 1.3107925635773583e-05, "loss": 1.7223, "step": 7425 }, { "epoch": 0.06678868387798878, "grad_norm": 4.613863945007324, "learning_rate": 1.3101315174073162e-05, "loss": 0.3576, "step": 7426 }, { "epoch": 0.06679767777562924, "grad_norm": 2.927076578140259, "learning_rate": 1.3094705787763071e-05, "loss": 2.1737, "step": 7427 }, { "epoch": 0.06680667167326969, "grad_norm": 2.565784454345703, "learning_rate": 1.3088097477440645e-05, "loss": 2.3838, "step": 7428 }, { "epoch": 0.06681566557091013, "grad_norm": 3.0369668006896973, "learning_rate": 1.3081490243703128e-05, "loss": 2.4407, "step": 7429 }, { "epoch": 0.06682465946855058, "grad_norm": 4.1036834716796875, "learning_rate": 1.3074884087147698e-05, "loss": 1.6495, "step": 7430 }, { "epoch": 0.06683365336619104, "grad_norm": 3.470780611038208, "learning_rate": 1.3068279008371387e-05, "loss": 1.5135, "step": 7431 }, { "epoch": 0.06684264726383149, "grad_norm": 2.5547945499420166, "learning_rate": 1.306167500797118e-05, "loss": 2.5048, "step": 7432 }, { "epoch": 0.06685164116147194, "grad_norm": 2.5781147480010986, "learning_rate": 1.3055072086543926e-05, "loss": 1.4287, "step": 7433 }, { "epoch": 0.0668606350591124, "grad_norm": 2.761427879333496, "learning_rate": 1.3048470244686382e-05, "loss": 2.0811, "step": 7434 }, { "epoch": 0.06686962895675284, "grad_norm": 2.5016403198242188, "learning_rate": 1.304186948299523e-05, "loss": 1.1308, "step": 7435 }, { "epoch": 0.06687862285439329, "grad_norm": 2.8876519203186035, "learning_rate": 1.3035269802067038e-05, "loss": 1.2519, "step": 7436 }, { "epoch": 0.06688761675203375, "grad_norm": 2.153805732727051, "learning_rate": 1.3028671202498261e-05, "loss": 2.0882, "step": 7437 }, { "epoch": 0.0668966106496742, "grad_norm": 2.519996166229248, "learning_rate": 1.30220736848853e-05, "loss": 1.9441, "step": 7438 }, { "epoch": 0.06690560454731465, "grad_norm": 2.79070782661438, "learning_rate": 1.3015477249824416e-05, "loss": 1.8557, "step": 7439 }, { "epoch": 0.06691459844495509, "grad_norm": 4.141666412353516, "learning_rate": 1.3008881897911787e-05, "loss": 0.3678, "step": 7440 }, { "epoch": 0.06692359234259555, "grad_norm": 21.016775131225586, "learning_rate": 1.3002287629743487e-05, "loss": 1.2961, "step": 7441 }, { "epoch": 0.066932586240236, "grad_norm": 7.851491451263428, "learning_rate": 1.2995694445915521e-05, "loss": 1.2501, "step": 7442 }, { "epoch": 0.06694158013787645, "grad_norm": 2.1285510063171387, "learning_rate": 1.2989102347023748e-05, "loss": 1.752, "step": 7443 }, { "epoch": 0.06695057403551691, "grad_norm": 1.8368957042694092, "learning_rate": 1.2982511333663982e-05, "loss": 1.7862, "step": 7444 }, { "epoch": 0.06695956793315735, "grad_norm": 12.646016120910645, "learning_rate": 1.2975921406431893e-05, "loss": 0.9315, "step": 7445 }, { "epoch": 0.0669685618307978, "grad_norm": 1.9761593341827393, "learning_rate": 1.2969332565923067e-05, "loss": 2.4826, "step": 7446 }, { "epoch": 0.06697755572843825, "grad_norm": 2.796893358230591, "learning_rate": 1.2962744812733018e-05, "loss": 2.5383, "step": 7447 }, { "epoch": 0.06698654962607871, "grad_norm": 2.8077685832977295, "learning_rate": 1.2956158147457115e-05, "loss": 1.0646, "step": 7448 }, { "epoch": 0.06699554352371916, "grad_norm": 2.4287304878234863, "learning_rate": 1.2949572570690676e-05, "loss": 1.7419, "step": 7449 }, { "epoch": 0.0670045374213596, "grad_norm": 2.9902524948120117, "learning_rate": 1.294298808302889e-05, "loss": 2.7763, "step": 7450 }, { "epoch": 0.06701353131900006, "grad_norm": 2.2515323162078857, "learning_rate": 1.2936404685066852e-05, "loss": 2.0459, "step": 7451 }, { "epoch": 0.06702252521664051, "grad_norm": 1.8336336612701416, "learning_rate": 1.2929822377399565e-05, "loss": 1.5046, "step": 7452 }, { "epoch": 0.06703151911428096, "grad_norm": 17.372774124145508, "learning_rate": 1.2923241160621923e-05, "loss": 0.6153, "step": 7453 }, { "epoch": 0.06704051301192142, "grad_norm": 18.788496017456055, "learning_rate": 1.2916661035328744e-05, "loss": 1.4778, "step": 7454 }, { "epoch": 0.06704950690956187, "grad_norm": 2.936103582382202, "learning_rate": 1.2910082002114717e-05, "loss": 2.005, "step": 7455 }, { "epoch": 0.06705850080720231, "grad_norm": 3.2086775302886963, "learning_rate": 1.2903504061574467e-05, "loss": 2.1569, "step": 7456 }, { "epoch": 0.06706749470484276, "grad_norm": 3.380913019180298, "learning_rate": 1.289692721430249e-05, "loss": 1.1624, "step": 7457 }, { "epoch": 0.06707648860248322, "grad_norm": 7.244720458984375, "learning_rate": 1.2890351460893185e-05, "loss": 1.8318, "step": 7458 }, { "epoch": 0.06708548250012367, "grad_norm": 2.377488851547241, "learning_rate": 1.2883776801940884e-05, "loss": 2.6702, "step": 7459 }, { "epoch": 0.06709447639776411, "grad_norm": 3.942568302154541, "learning_rate": 1.287720323803977e-05, "loss": 2.3684, "step": 7460 }, { "epoch": 0.06710347029540457, "grad_norm": 1.8597475290298462, "learning_rate": 1.2870630769783983e-05, "loss": 2.0557, "step": 7461 }, { "epoch": 0.06711246419304502, "grad_norm": 2.3877320289611816, "learning_rate": 1.2864059397767515e-05, "loss": 2.1942, "step": 7462 }, { "epoch": 0.06712145809068547, "grad_norm": 3.59509015083313, "learning_rate": 1.2857489122584294e-05, "loss": 0.953, "step": 7463 }, { "epoch": 0.06713045198832591, "grad_norm": 2.1839165687561035, "learning_rate": 1.2850919944828125e-05, "loss": 2.2508, "step": 7464 }, { "epoch": 0.06713944588596638, "grad_norm": 18.595548629760742, "learning_rate": 1.2844351865092728e-05, "loss": 1.704, "step": 7465 }, { "epoch": 0.06714843978360682, "grad_norm": 1.7759292125701904, "learning_rate": 1.2837784883971716e-05, "loss": 2.4452, "step": 7466 }, { "epoch": 0.06715743368124727, "grad_norm": 2.530778408050537, "learning_rate": 1.2831219002058595e-05, "loss": 2.1823, "step": 7467 }, { "epoch": 0.06716642757888773, "grad_norm": 3.223365306854248, "learning_rate": 1.2824654219946803e-05, "loss": 1.5703, "step": 7468 }, { "epoch": 0.06717542147652818, "grad_norm": 2.988006353378296, "learning_rate": 1.2818090538229649e-05, "loss": 1.4217, "step": 7469 }, { "epoch": 0.06718441537416862, "grad_norm": 17.802593231201172, "learning_rate": 1.2811527957500345e-05, "loss": 0.7995, "step": 7470 }, { "epoch": 0.06719340927180907, "grad_norm": 2.217027187347412, "learning_rate": 1.280496647835202e-05, "loss": 2.0061, "step": 7471 }, { "epoch": 0.06720240316944953, "grad_norm": 2.1776299476623535, "learning_rate": 1.2798406101377686e-05, "loss": 2.0718, "step": 7472 }, { "epoch": 0.06721139706708998, "grad_norm": 17.286046981811523, "learning_rate": 1.279184682717027e-05, "loss": 1.4338, "step": 7473 }, { "epoch": 0.06722039096473043, "grad_norm": 12.366009712219238, "learning_rate": 1.2785288656322581e-05, "loss": 1.2911, "step": 7474 }, { "epoch": 0.06722938486237089, "grad_norm": 2.4144086837768555, "learning_rate": 1.2778731589427356e-05, "loss": 1.9999, "step": 7475 }, { "epoch": 0.06723837876001133, "grad_norm": 2.4588143825531006, "learning_rate": 1.2772175627077205e-05, "loss": 2.4259, "step": 7476 }, { "epoch": 0.06724737265765178, "grad_norm": 13.27937126159668, "learning_rate": 1.2765620769864642e-05, "loss": 1.2095, "step": 7477 }, { "epoch": 0.06725636655529224, "grad_norm": 2.910149574279785, "learning_rate": 1.2759067018382103e-05, "loss": 1.91, "step": 7478 }, { "epoch": 0.06726536045293269, "grad_norm": 3.961604118347168, "learning_rate": 1.2752514373221906e-05, "loss": 0.7495, "step": 7479 }, { "epoch": 0.06727435435057313, "grad_norm": 3.092334508895874, "learning_rate": 1.2745962834976266e-05, "loss": 1.2564, "step": 7480 }, { "epoch": 0.06728334824821358, "grad_norm": 3.7224583625793457, "learning_rate": 1.2739412404237306e-05, "loss": 0.9991, "step": 7481 }, { "epoch": 0.06729234214585404, "grad_norm": 2.7171199321746826, "learning_rate": 1.2732863081597039e-05, "loss": 1.9863, "step": 7482 }, { "epoch": 0.06730133604349449, "grad_norm": 3.136533498764038, "learning_rate": 1.2726314867647399e-05, "loss": 1.7632, "step": 7483 }, { "epoch": 0.06731032994113494, "grad_norm": 16.064512252807617, "learning_rate": 1.2719767762980195e-05, "loss": 1.2183, "step": 7484 }, { "epoch": 0.0673193238387754, "grad_norm": 15.365964889526367, "learning_rate": 1.2713221768187161e-05, "loss": 1.2707, "step": 7485 }, { "epoch": 0.06732831773641584, "grad_norm": 2.6311559677124023, "learning_rate": 1.2706676883859903e-05, "loss": 1.4371, "step": 7486 }, { "epoch": 0.06733731163405629, "grad_norm": 10.762723922729492, "learning_rate": 1.2700133110589954e-05, "loss": 1.1134, "step": 7487 }, { "epoch": 0.06734630553169674, "grad_norm": 3.7226362228393555, "learning_rate": 1.2693590448968729e-05, "loss": 0.4661, "step": 7488 }, { "epoch": 0.0673552994293372, "grad_norm": 2.6902101039886475, "learning_rate": 1.2687048899587529e-05, "loss": 2.2092, "step": 7489 }, { "epoch": 0.06736429332697765, "grad_norm": 2.691549062728882, "learning_rate": 1.26805084630376e-05, "loss": 1.8804, "step": 7490 }, { "epoch": 0.06737328722461809, "grad_norm": 2.270472288131714, "learning_rate": 1.2673969139910047e-05, "loss": 2.8176, "step": 7491 }, { "epoch": 0.06738228112225855, "grad_norm": 16.40492820739746, "learning_rate": 1.2667430930795877e-05, "loss": 1.1685, "step": 7492 }, { "epoch": 0.067391275019899, "grad_norm": 11.521434783935547, "learning_rate": 1.2660893836286026e-05, "loss": 0.9737, "step": 7493 }, { "epoch": 0.06740026891753945, "grad_norm": 5.7537522315979, "learning_rate": 1.2654357856971296e-05, "loss": 1.8264, "step": 7494 }, { "epoch": 0.06740926281517991, "grad_norm": 10.655234336853027, "learning_rate": 1.264782299344241e-05, "loss": 1.2141, "step": 7495 }, { "epoch": 0.06741825671282035, "grad_norm": 2.4427032470703125, "learning_rate": 1.2641289246289961e-05, "loss": 2.4043, "step": 7496 }, { "epoch": 0.0674272506104608, "grad_norm": 2.550358533859253, "learning_rate": 1.2634756616104492e-05, "loss": 1.5552, "step": 7497 }, { "epoch": 0.06743624450810125, "grad_norm": 9.364204406738281, "learning_rate": 1.2628225103476405e-05, "loss": 1.6675, "step": 7498 }, { "epoch": 0.06744523840574171, "grad_norm": 3.1350038051605225, "learning_rate": 1.2621694708995993e-05, "loss": 1.4215, "step": 7499 }, { "epoch": 0.06745423230338216, "grad_norm": 5.245044231414795, "learning_rate": 1.261516543325349e-05, "loss": 1.2126, "step": 7500 }, { "epoch": 0.0674632262010226, "grad_norm": 2.6766717433929443, "learning_rate": 1.2608637276838986e-05, "loss": 1.4179, "step": 7501 }, { "epoch": 0.06747222009866306, "grad_norm": 2.5846779346466064, "learning_rate": 1.260211024034251e-05, "loss": 2.0548, "step": 7502 }, { "epoch": 0.06748121399630351, "grad_norm": 14.444372177124023, "learning_rate": 1.2595584324353943e-05, "loss": 1.6133, "step": 7503 }, { "epoch": 0.06749020789394396, "grad_norm": 2.1766254901885986, "learning_rate": 1.2589059529463118e-05, "loss": 2.0941, "step": 7504 }, { "epoch": 0.0674992017915844, "grad_norm": 9.318672180175781, "learning_rate": 1.2582535856259722e-05, "loss": 1.4803, "step": 7505 }, { "epoch": 0.06750819568922486, "grad_norm": 3.5538227558135986, "learning_rate": 1.2576013305333362e-05, "loss": 1.4928, "step": 7506 }, { "epoch": 0.06751718958686531, "grad_norm": 3.7145187854766846, "learning_rate": 1.2569491877273537e-05, "loss": 1.712, "step": 7507 }, { "epoch": 0.06752618348450576, "grad_norm": 2.1243221759796143, "learning_rate": 1.2562971572669635e-05, "loss": 2.727, "step": 7508 }, { "epoch": 0.06753517738214622, "grad_norm": 3.0571625232696533, "learning_rate": 1.2556452392110978e-05, "loss": 1.6155, "step": 7509 }, { "epoch": 0.06754417127978667, "grad_norm": 3.7958009243011475, "learning_rate": 1.254993433618675e-05, "loss": 1.1103, "step": 7510 }, { "epoch": 0.06755316517742711, "grad_norm": 2.760913610458374, "learning_rate": 1.2543417405486035e-05, "loss": 1.0114, "step": 7511 }, { "epoch": 0.06756215907506756, "grad_norm": 3.1639721393585205, "learning_rate": 1.253690160059785e-05, "loss": 1.5123, "step": 7512 }, { "epoch": 0.06757115297270802, "grad_norm": 2.452826499938965, "learning_rate": 1.2530386922111066e-05, "loss": 1.8362, "step": 7513 }, { "epoch": 0.06758014687034847, "grad_norm": 3.2646031379699707, "learning_rate": 1.2523873370614489e-05, "loss": 1.5288, "step": 7514 }, { "epoch": 0.06758914076798891, "grad_norm": 15.003978729248047, "learning_rate": 1.2517360946696787e-05, "loss": 1.2965, "step": 7515 }, { "epoch": 0.06759813466562938, "grad_norm": 3.883488416671753, "learning_rate": 1.2510849650946571e-05, "loss": 1.4359, "step": 7516 }, { "epoch": 0.06760712856326982, "grad_norm": 2.5038037300109863, "learning_rate": 1.250433948395231e-05, "loss": 1.9168, "step": 7517 }, { "epoch": 0.06761612246091027, "grad_norm": 2.643780469894409, "learning_rate": 1.2497830446302374e-05, "loss": 1.8227, "step": 7518 }, { "epoch": 0.06762511635855073, "grad_norm": 2.516181468963623, "learning_rate": 1.2491322538585068e-05, "loss": 2.3471, "step": 7519 }, { "epoch": 0.06763411025619118, "grad_norm": 3.208531618118286, "learning_rate": 1.248481576138856e-05, "loss": 1.9262, "step": 7520 }, { "epoch": 0.06764310415383162, "grad_norm": 2.4678173065185547, "learning_rate": 1.247831011530092e-05, "loss": 1.9299, "step": 7521 }, { "epoch": 0.06765209805147207, "grad_norm": 2.4060721397399902, "learning_rate": 1.2471805600910125e-05, "loss": 2.9329, "step": 7522 }, { "epoch": 0.06766109194911253, "grad_norm": 16.902263641357422, "learning_rate": 1.2465302218804038e-05, "loss": 0.8899, "step": 7523 }, { "epoch": 0.06767008584675298, "grad_norm": 2.599264144897461, "learning_rate": 1.2458799969570441e-05, "loss": 1.4752, "step": 7524 }, { "epoch": 0.06767907974439342, "grad_norm": 2.4081010818481445, "learning_rate": 1.245229885379699e-05, "loss": 1.9173, "step": 7525 }, { "epoch": 0.06768807364203389, "grad_norm": 2.924421787261963, "learning_rate": 1.244579887207126e-05, "loss": 1.7654, "step": 7526 }, { "epoch": 0.06769706753967433, "grad_norm": 2.6418874263763428, "learning_rate": 1.2439300024980697e-05, "loss": 1.6384, "step": 7527 }, { "epoch": 0.06770606143731478, "grad_norm": 2.1183888912200928, "learning_rate": 1.2432802313112676e-05, "loss": 1.9022, "step": 7528 }, { "epoch": 0.06771505533495523, "grad_norm": 8.622857093811035, "learning_rate": 1.2426305737054447e-05, "loss": 1.32, "step": 7529 }, { "epoch": 0.06772404923259569, "grad_norm": 3.260547637939453, "learning_rate": 1.2419810297393153e-05, "loss": 0.8183, "step": 7530 }, { "epoch": 0.06773304313023613, "grad_norm": 2.1091504096984863, "learning_rate": 1.2413315994715864e-05, "loss": 2.3499, "step": 7531 }, { "epoch": 0.06774203702787658, "grad_norm": 3.3165125846862793, "learning_rate": 1.2406822829609508e-05, "loss": 1.6463, "step": 7532 }, { "epoch": 0.06775103092551704, "grad_norm": 2.495814561843872, "learning_rate": 1.2400330802660951e-05, "loss": 1.8018, "step": 7533 }, { "epoch": 0.06776002482315749, "grad_norm": 22.09170913696289, "learning_rate": 1.2393839914456922e-05, "loss": 0.5321, "step": 7534 }, { "epoch": 0.06776901872079794, "grad_norm": 13.0279541015625, "learning_rate": 1.2387350165584066e-05, "loss": 0.5153, "step": 7535 }, { "epoch": 0.0677780126184384, "grad_norm": 2.0033605098724365, "learning_rate": 1.2380861556628915e-05, "loss": 1.5801, "step": 7536 }, { "epoch": 0.06778700651607884, "grad_norm": 3.276632308959961, "learning_rate": 1.2374374088177895e-05, "loss": 1.6038, "step": 7537 }, { "epoch": 0.06779600041371929, "grad_norm": 2.2077853679656982, "learning_rate": 1.2367887760817357e-05, "loss": 2.0209, "step": 7538 }, { "epoch": 0.06780499431135974, "grad_norm": 3.502872943878174, "learning_rate": 1.2361402575133506e-05, "loss": 0.8688, "step": 7539 }, { "epoch": 0.0678139882090002, "grad_norm": 2.5795295238494873, "learning_rate": 1.2354918531712486e-05, "loss": 2.1882, "step": 7540 }, { "epoch": 0.06782298210664064, "grad_norm": 2.2094879150390625, "learning_rate": 1.2348435631140311e-05, "loss": 2.0718, "step": 7541 }, { "epoch": 0.06783197600428109, "grad_norm": 1.7709827423095703, "learning_rate": 1.2341953874002887e-05, "loss": 2.6178, "step": 7542 }, { "epoch": 0.06784096990192155, "grad_norm": 2.0635743141174316, "learning_rate": 1.2335473260886046e-05, "loss": 1.5329, "step": 7543 }, { "epoch": 0.067849963799562, "grad_norm": 19.324451446533203, "learning_rate": 1.2328993792375479e-05, "loss": 1.3049, "step": 7544 }, { "epoch": 0.06785895769720245, "grad_norm": 2.1759650707244873, "learning_rate": 1.2322515469056818e-05, "loss": 2.2418, "step": 7545 }, { "epoch": 0.06786795159484289, "grad_norm": 2.4787328243255615, "learning_rate": 1.2316038291515553e-05, "loss": 1.4992, "step": 7546 }, { "epoch": 0.06787694549248335, "grad_norm": 2.5223536491394043, "learning_rate": 1.2309562260337073e-05, "loss": 1.3203, "step": 7547 }, { "epoch": 0.0678859393901238, "grad_norm": 19.034639358520508, "learning_rate": 1.23030873761067e-05, "loss": 1.6608, "step": 7548 }, { "epoch": 0.06789493328776425, "grad_norm": 2.4854562282562256, "learning_rate": 1.2296613639409613e-05, "loss": 1.9821, "step": 7549 }, { "epoch": 0.06790392718540471, "grad_norm": 2.244494915008545, "learning_rate": 1.22901410508309e-05, "loss": 2.0781, "step": 7550 }, { "epoch": 0.06791292108304516, "grad_norm": 2.9512171745300293, "learning_rate": 1.2283669610955542e-05, "loss": 1.8116, "step": 7551 }, { "epoch": 0.0679219149806856, "grad_norm": 3.013715982437134, "learning_rate": 1.2277199320368437e-05, "loss": 2.0227, "step": 7552 }, { "epoch": 0.06793090887832605, "grad_norm": 2.092479705810547, "learning_rate": 1.2270730179654353e-05, "loss": 2.8606, "step": 7553 }, { "epoch": 0.06793990277596651, "grad_norm": 18.71356773376465, "learning_rate": 1.2264262189397954e-05, "loss": 0.439, "step": 7554 }, { "epoch": 0.06794889667360696, "grad_norm": 16.735002517700195, "learning_rate": 1.2257795350183832e-05, "loss": 1.5543, "step": 7555 }, { "epoch": 0.0679578905712474, "grad_norm": 1.7709567546844482, "learning_rate": 1.225132966259643e-05, "loss": 2.1007, "step": 7556 }, { "epoch": 0.06796688446888786, "grad_norm": 6.511504650115967, "learning_rate": 1.2244865127220132e-05, "loss": 1.7596, "step": 7557 }, { "epoch": 0.06797587836652831, "grad_norm": 1.9947764873504639, "learning_rate": 1.2238401744639185e-05, "loss": 2.4832, "step": 7558 }, { "epoch": 0.06798487226416876, "grad_norm": 3.0881078243255615, "learning_rate": 1.2231939515437735e-05, "loss": 1.9398, "step": 7559 }, { "epoch": 0.06799386616180922, "grad_norm": 2.799192428588867, "learning_rate": 1.2225478440199845e-05, "loss": 1.3119, "step": 7560 }, { "epoch": 0.06800286005944967, "grad_norm": 2.1572911739349365, "learning_rate": 1.2219018519509456e-05, "loss": 1.9339, "step": 7561 }, { "epoch": 0.06801185395709011, "grad_norm": 2.4737086296081543, "learning_rate": 1.2212559753950406e-05, "loss": 1.795, "step": 7562 }, { "epoch": 0.06802084785473056, "grad_norm": 15.629800796508789, "learning_rate": 1.2206102144106424e-05, "loss": 0.9663, "step": 7563 }, { "epoch": 0.06802984175237102, "grad_norm": 1.5537933111190796, "learning_rate": 1.2199645690561156e-05, "loss": 2.3135, "step": 7564 }, { "epoch": 0.06803883565001147, "grad_norm": 23.293912887573242, "learning_rate": 1.2193190393898123e-05, "loss": 1.146, "step": 7565 }, { "epoch": 0.06804782954765191, "grad_norm": 3.071608304977417, "learning_rate": 1.2186736254700742e-05, "loss": 1.4797, "step": 7566 }, { "epoch": 0.06805682344529237, "grad_norm": 2.054856061935425, "learning_rate": 1.2180283273552346e-05, "loss": 1.9191, "step": 7567 }, { "epoch": 0.06806581734293282, "grad_norm": 3.018178939819336, "learning_rate": 1.2173831451036127e-05, "loss": 1.7602, "step": 7568 }, { "epoch": 0.06807481124057327, "grad_norm": 2.5414774417877197, "learning_rate": 1.216738078773522e-05, "loss": 1.68, "step": 7569 }, { "epoch": 0.06808380513821372, "grad_norm": 3.4251198768615723, "learning_rate": 1.2160931284232612e-05, "loss": 2.5539, "step": 7570 }, { "epoch": 0.06809279903585418, "grad_norm": 2.3364083766937256, "learning_rate": 1.2154482941111197e-05, "loss": 2.4577, "step": 7571 }, { "epoch": 0.06810179293349462, "grad_norm": 21.757678985595703, "learning_rate": 1.2148035758953784e-05, "loss": 1.5507, "step": 7572 }, { "epoch": 0.06811078683113507, "grad_norm": 8.77946949005127, "learning_rate": 1.214158973834305e-05, "loss": 1.3064, "step": 7573 }, { "epoch": 0.06811978072877553, "grad_norm": 1.7822763919830322, "learning_rate": 1.2135144879861595e-05, "loss": 2.1118, "step": 7574 }, { "epoch": 0.06812877462641598, "grad_norm": 2.3318839073181152, "learning_rate": 1.212870118409189e-05, "loss": 1.9178, "step": 7575 }, { "epoch": 0.06813776852405642, "grad_norm": 28.162935256958008, "learning_rate": 1.2122258651616306e-05, "loss": 0.9264, "step": 7576 }, { "epoch": 0.06814676242169689, "grad_norm": 20.089954376220703, "learning_rate": 1.2115817283017115e-05, "loss": 1.408, "step": 7577 }, { "epoch": 0.06815575631933733, "grad_norm": 1.6402385234832764, "learning_rate": 1.2109377078876472e-05, "loss": 2.3444, "step": 7578 }, { "epoch": 0.06816475021697778, "grad_norm": 2.3454720973968506, "learning_rate": 1.2102938039776452e-05, "loss": 2.0264, "step": 7579 }, { "epoch": 0.06817374411461823, "grad_norm": 4.296626567840576, "learning_rate": 1.209650016629899e-05, "loss": 0.8519, "step": 7580 }, { "epoch": 0.06818273801225869, "grad_norm": 2.9259378910064697, "learning_rate": 1.2090063459025955e-05, "loss": 1.7252, "step": 7581 }, { "epoch": 0.06819173190989913, "grad_norm": 3.180196523666382, "learning_rate": 1.2083627918539079e-05, "loss": 2.4904, "step": 7582 }, { "epoch": 0.06820072580753958, "grad_norm": 1.9969412088394165, "learning_rate": 1.2077193545419991e-05, "loss": 2.1776, "step": 7583 }, { "epoch": 0.06820971970518004, "grad_norm": 1.9868875741958618, "learning_rate": 1.207076034025024e-05, "loss": 2.6213, "step": 7584 }, { "epoch": 0.06821871360282049, "grad_norm": 9.008501052856445, "learning_rate": 1.2064328303611234e-05, "loss": 1.2097, "step": 7585 }, { "epoch": 0.06822770750046094, "grad_norm": 14.09090518951416, "learning_rate": 1.2057897436084308e-05, "loss": 1.6489, "step": 7586 }, { "epoch": 0.06823670139810138, "grad_norm": 1.9417130947113037, "learning_rate": 1.2051467738250677e-05, "loss": 2.0565, "step": 7587 }, { "epoch": 0.06824569529574184, "grad_norm": 2.0543813705444336, "learning_rate": 1.2045039210691433e-05, "loss": 2.0712, "step": 7588 }, { "epoch": 0.06825468919338229, "grad_norm": 8.63691520690918, "learning_rate": 1.2038611853987603e-05, "loss": 0.7394, "step": 7589 }, { "epoch": 0.06826368309102274, "grad_norm": 5.067418575286865, "learning_rate": 1.203218566872007e-05, "loss": 1.3792, "step": 7590 }, { "epoch": 0.0682726769886632, "grad_norm": 25.563112258911133, "learning_rate": 1.202576065546963e-05, "loss": 1.4189, "step": 7591 }, { "epoch": 0.06828167088630364, "grad_norm": 4.284435749053955, "learning_rate": 1.2019336814816956e-05, "loss": 0.807, "step": 7592 }, { "epoch": 0.06829066478394409, "grad_norm": 2.513610601425171, "learning_rate": 1.201291414734265e-05, "loss": 1.7373, "step": 7593 }, { "epoch": 0.06829965868158454, "grad_norm": 3.085834503173828, "learning_rate": 1.2006492653627174e-05, "loss": 2.0071, "step": 7594 }, { "epoch": 0.068308652579225, "grad_norm": 2.1197617053985596, "learning_rate": 1.2000072334250889e-05, "loss": 2.0815, "step": 7595 }, { "epoch": 0.06831764647686545, "grad_norm": 17.493480682373047, "learning_rate": 1.1993653189794071e-05, "loss": 1.1983, "step": 7596 }, { "epoch": 0.06832664037450589, "grad_norm": 10.679052352905273, "learning_rate": 1.198723522083686e-05, "loss": 0.961, "step": 7597 }, { "epoch": 0.06833563427214635, "grad_norm": 2.5732438564300537, "learning_rate": 1.1980818427959322e-05, "loss": 1.4322, "step": 7598 }, { "epoch": 0.0683446281697868, "grad_norm": 2.4029345512390137, "learning_rate": 1.1974402811741395e-05, "loss": 2.2012, "step": 7599 }, { "epoch": 0.06835362206742725, "grad_norm": 11.754157066345215, "learning_rate": 1.1967988372762898e-05, "loss": 0.7777, "step": 7600 }, { "epoch": 0.06836261596506771, "grad_norm": 14.665637969970703, "learning_rate": 1.1961575111603588e-05, "loss": 1.629, "step": 7601 }, { "epoch": 0.06837160986270815, "grad_norm": 15.430252075195312, "learning_rate": 1.1955163028843063e-05, "loss": 0.9142, "step": 7602 }, { "epoch": 0.0683806037603486, "grad_norm": 16.41813850402832, "learning_rate": 1.1948752125060864e-05, "loss": 1.7615, "step": 7603 }, { "epoch": 0.06838959765798905, "grad_norm": 2.6624674797058105, "learning_rate": 1.194234240083639e-05, "loss": 1.8085, "step": 7604 }, { "epoch": 0.06839859155562951, "grad_norm": 8.591865539550781, "learning_rate": 1.1935933856748946e-05, "loss": 0.8873, "step": 7605 }, { "epoch": 0.06840758545326996, "grad_norm": 2.560727596282959, "learning_rate": 1.1929526493377726e-05, "loss": 1.9823, "step": 7606 }, { "epoch": 0.0684165793509104, "grad_norm": 7.718970775604248, "learning_rate": 1.1923120311301817e-05, "loss": 1.5188, "step": 7607 }, { "epoch": 0.06842557324855086, "grad_norm": 3.4351813793182373, "learning_rate": 1.1916715311100216e-05, "loss": 1.416, "step": 7608 }, { "epoch": 0.06843456714619131, "grad_norm": 3.0983052253723145, "learning_rate": 1.1910311493351784e-05, "loss": 2.2757, "step": 7609 }, { "epoch": 0.06844356104383176, "grad_norm": 4.733487606048584, "learning_rate": 1.190390885863531e-05, "loss": 0.9591, "step": 7610 }, { "epoch": 0.0684525549414722, "grad_norm": 7.8430633544921875, "learning_rate": 1.1897507407529448e-05, "loss": 1.1266, "step": 7611 }, { "epoch": 0.06846154883911267, "grad_norm": 2.7414586544036865, "learning_rate": 1.1891107140612746e-05, "loss": 2.2265, "step": 7612 }, { "epoch": 0.06847054273675311, "grad_norm": 2.2961926460266113, "learning_rate": 1.1884708058463668e-05, "loss": 2.6881, "step": 7613 }, { "epoch": 0.06847953663439356, "grad_norm": 5.876256942749023, "learning_rate": 1.187831016166054e-05, "loss": 1.1474, "step": 7614 }, { "epoch": 0.06848853053203402, "grad_norm": 7.716744899749756, "learning_rate": 1.1871913450781613e-05, "loss": 1.1562, "step": 7615 }, { "epoch": 0.06849752442967447, "grad_norm": 2.630424976348877, "learning_rate": 1.186551792640501e-05, "loss": 2.6316, "step": 7616 }, { "epoch": 0.06850651832731491, "grad_norm": 8.983920097351074, "learning_rate": 1.1859123589108739e-05, "loss": 0.9544, "step": 7617 }, { "epoch": 0.06851551222495537, "grad_norm": 3.1006956100463867, "learning_rate": 1.1852730439470735e-05, "loss": 1.1145, "step": 7618 }, { "epoch": 0.06852450612259582, "grad_norm": 9.6243896484375, "learning_rate": 1.1846338478068791e-05, "loss": 1.4505, "step": 7619 }, { "epoch": 0.06853350002023627, "grad_norm": 3.2238898277282715, "learning_rate": 1.1839947705480611e-05, "loss": 1.6206, "step": 7620 }, { "epoch": 0.06854249391787672, "grad_norm": 2.5554795265197754, "learning_rate": 1.183355812228377e-05, "loss": 2.5341, "step": 7621 }, { "epoch": 0.06855148781551718, "grad_norm": 6.619315147399902, "learning_rate": 1.1827169729055774e-05, "loss": 1.1971, "step": 7622 }, { "epoch": 0.06856048171315762, "grad_norm": 9.483216285705566, "learning_rate": 1.1820782526373991e-05, "loss": 0.9231, "step": 7623 }, { "epoch": 0.06856947561079807, "grad_norm": 6.560564994812012, "learning_rate": 1.1814396514815676e-05, "loss": 1.45, "step": 7624 }, { "epoch": 0.06857846950843853, "grad_norm": 17.151676177978516, "learning_rate": 1.1808011694958013e-05, "loss": 0.9018, "step": 7625 }, { "epoch": 0.06858746340607898, "grad_norm": 2.63995099067688, "learning_rate": 1.1801628067378031e-05, "loss": 2.5003, "step": 7626 }, { "epoch": 0.06859645730371942, "grad_norm": 3.262554407119751, "learning_rate": 1.1795245632652701e-05, "loss": 1.7275, "step": 7627 }, { "epoch": 0.06860545120135987, "grad_norm": 3.6214044094085693, "learning_rate": 1.1788864391358837e-05, "loss": 1.3706, "step": 7628 }, { "epoch": 0.06861444509900033, "grad_norm": 4.048209190368652, "learning_rate": 1.1782484344073188e-05, "loss": 1.4199, "step": 7629 }, { "epoch": 0.06862343899664078, "grad_norm": 2.972478151321411, "learning_rate": 1.1776105491372369e-05, "loss": 2.1404, "step": 7630 }, { "epoch": 0.06863243289428123, "grad_norm": 3.224045991897583, "learning_rate": 1.1769727833832892e-05, "loss": 2.2711, "step": 7631 }, { "epoch": 0.06864142679192169, "grad_norm": 7.9035844802856445, "learning_rate": 1.1763351372031164e-05, "loss": 1.7558, "step": 7632 }, { "epoch": 0.06865042068956213, "grad_norm": 3.863961696624756, "learning_rate": 1.1756976106543469e-05, "loss": 1.3137, "step": 7633 }, { "epoch": 0.06865941458720258, "grad_norm": 2.326716423034668, "learning_rate": 1.175060203794602e-05, "loss": 2.8212, "step": 7634 }, { "epoch": 0.06866840848484304, "grad_norm": 14.39902114868164, "learning_rate": 1.1744229166814888e-05, "loss": 0.6032, "step": 7635 }, { "epoch": 0.06867740238248349, "grad_norm": 9.924027442932129, "learning_rate": 1.1737857493726037e-05, "loss": 0.2209, "step": 7636 }, { "epoch": 0.06868639628012393, "grad_norm": 3.7069575786590576, "learning_rate": 1.173148701925535e-05, "loss": 1.5562, "step": 7637 }, { "epoch": 0.06869539017776438, "grad_norm": 3.0234386920928955, "learning_rate": 1.1725117743978567e-05, "loss": 1.8706, "step": 7638 }, { "epoch": 0.06870438407540484, "grad_norm": 4.064068794250488, "learning_rate": 1.1718749668471349e-05, "loss": 1.5169, "step": 7639 }, { "epoch": 0.06871337797304529, "grad_norm": 3.829908609390259, "learning_rate": 1.1712382793309234e-05, "loss": 1.1521, "step": 7640 }, { "epoch": 0.06872237187068574, "grad_norm": 4.622248649597168, "learning_rate": 1.170601711906764e-05, "loss": 0.9082, "step": 7641 }, { "epoch": 0.0687313657683262, "grad_norm": 2.313192844390869, "learning_rate": 1.1699652646321907e-05, "loss": 2.584, "step": 7642 }, { "epoch": 0.06874035966596664, "grad_norm": 5.707856178283691, "learning_rate": 1.1693289375647235e-05, "loss": 1.2214, "step": 7643 }, { "epoch": 0.06874935356360709, "grad_norm": 13.47648811340332, "learning_rate": 1.1686927307618745e-05, "loss": 0.5777, "step": 7644 }, { "epoch": 0.06875834746124754, "grad_norm": 11.374863624572754, "learning_rate": 1.1680566442811425e-05, "loss": 1.1305, "step": 7645 }, { "epoch": 0.068767341358888, "grad_norm": 2.3179712295532227, "learning_rate": 1.1674206781800162e-05, "loss": 2.0875, "step": 7646 }, { "epoch": 0.06877633525652845, "grad_norm": 3.710357904434204, "learning_rate": 1.1667848325159738e-05, "loss": 1.0958, "step": 7647 }, { "epoch": 0.06878532915416889, "grad_norm": 10.855904579162598, "learning_rate": 1.1661491073464815e-05, "loss": 1.1042, "step": 7648 }, { "epoch": 0.06879432305180935, "grad_norm": 8.016056060791016, "learning_rate": 1.1655135027289967e-05, "loss": 1.0538, "step": 7649 }, { "epoch": 0.0688033169494498, "grad_norm": 12.71065902709961, "learning_rate": 1.1648780187209634e-05, "loss": 1.1618, "step": 7650 }, { "epoch": 0.06881231084709025, "grad_norm": 3.500521421432495, "learning_rate": 1.1642426553798174e-05, "loss": 2.3348, "step": 7651 }, { "epoch": 0.0688213047447307, "grad_norm": 3.4516489505767822, "learning_rate": 1.1636074127629817e-05, "loss": 0.8808, "step": 7652 }, { "epoch": 0.06883029864237115, "grad_norm": 9.896587371826172, "learning_rate": 1.1629722909278678e-05, "loss": 0.926, "step": 7653 }, { "epoch": 0.0688392925400116, "grad_norm": 4.2149834632873535, "learning_rate": 1.1623372899318788e-05, "loss": 0.7751, "step": 7654 }, { "epoch": 0.06884828643765205, "grad_norm": 2.8125457763671875, "learning_rate": 1.161702409832404e-05, "loss": 2.1936, "step": 7655 }, { "epoch": 0.06885728033529251, "grad_norm": 11.007015228271484, "learning_rate": 1.1610676506868246e-05, "loss": 0.9731, "step": 7656 }, { "epoch": 0.06886627423293296, "grad_norm": 2.5170414447784424, "learning_rate": 1.1604330125525079e-05, "loss": 2.6655, "step": 7657 }, { "epoch": 0.0688752681305734, "grad_norm": 12.899922370910645, "learning_rate": 1.1597984954868135e-05, "loss": 0.9097, "step": 7658 }, { "epoch": 0.06888426202821386, "grad_norm": 2.3711836338043213, "learning_rate": 1.1591640995470879e-05, "loss": 2.5241, "step": 7659 }, { "epoch": 0.06889325592585431, "grad_norm": 3.0300772190093994, "learning_rate": 1.1585298247906664e-05, "loss": 1.6869, "step": 7660 }, { "epoch": 0.06890224982349476, "grad_norm": 8.79556655883789, "learning_rate": 1.1578956712748745e-05, "loss": 1.3991, "step": 7661 }, { "epoch": 0.0689112437211352, "grad_norm": 2.4712953567504883, "learning_rate": 1.1572616390570253e-05, "loss": 2.9632, "step": 7662 }, { "epoch": 0.06892023761877566, "grad_norm": 9.133211135864258, "learning_rate": 1.1566277281944237e-05, "loss": 0.7174, "step": 7663 }, { "epoch": 0.06892923151641611, "grad_norm": 2.7888669967651367, "learning_rate": 1.1559939387443613e-05, "loss": 2.3295, "step": 7664 }, { "epoch": 0.06893822541405656, "grad_norm": 3.19560170173645, "learning_rate": 1.1553602707641181e-05, "loss": 1.682, "step": 7665 }, { "epoch": 0.06894721931169702, "grad_norm": 2.473867416381836, "learning_rate": 1.1547267243109664e-05, "loss": 2.6448, "step": 7666 }, { "epoch": 0.06895621320933747, "grad_norm": 2.8383748531341553, "learning_rate": 1.154093299442163e-05, "loss": 1.759, "step": 7667 }, { "epoch": 0.06896520710697791, "grad_norm": 3.452755928039551, "learning_rate": 1.1534599962149587e-05, "loss": 2.1836, "step": 7668 }, { "epoch": 0.06897420100461836, "grad_norm": 7.701223850250244, "learning_rate": 1.1528268146865886e-05, "loss": 0.4537, "step": 7669 }, { "epoch": 0.06898319490225882, "grad_norm": 4.252696990966797, "learning_rate": 1.1521937549142808e-05, "loss": 1.6389, "step": 7670 }, { "epoch": 0.06899218879989927, "grad_norm": 3.6394996643066406, "learning_rate": 1.1515608169552497e-05, "loss": 1.2392, "step": 7671 }, { "epoch": 0.06900118269753971, "grad_norm": 2.6226277351379395, "learning_rate": 1.1509280008666987e-05, "loss": 2.4443, "step": 7672 }, { "epoch": 0.06901017659518018, "grad_norm": 3.9947621822357178, "learning_rate": 1.1502953067058228e-05, "loss": 1.4203, "step": 7673 }, { "epoch": 0.06901917049282062, "grad_norm": 4.908615589141846, "learning_rate": 1.149662734529803e-05, "loss": 1.6439, "step": 7674 }, { "epoch": 0.06902816439046107, "grad_norm": 3.0820846557617188, "learning_rate": 1.1490302843958112e-05, "loss": 1.183, "step": 7675 }, { "epoch": 0.06903715828810153, "grad_norm": 3.378774404525757, "learning_rate": 1.148397956361007e-05, "loss": 1.3856, "step": 7676 }, { "epoch": 0.06904615218574198, "grad_norm": 2.577611207962036, "learning_rate": 1.1477657504825386e-05, "loss": 2.0433, "step": 7677 }, { "epoch": 0.06905514608338242, "grad_norm": 2.321681261062622, "learning_rate": 1.1471336668175464e-05, "loss": 2.5816, "step": 7678 }, { "epoch": 0.06906413998102287, "grad_norm": 14.833800315856934, "learning_rate": 1.146501705423155e-05, "loss": 1.1387, "step": 7679 }, { "epoch": 0.06907313387866333, "grad_norm": 3.071552038192749, "learning_rate": 1.1458698663564827e-05, "loss": 1.5042, "step": 7680 }, { "epoch": 0.06908212777630378, "grad_norm": 9.448452949523926, "learning_rate": 1.1452381496746323e-05, "loss": 1.142, "step": 7681 }, { "epoch": 0.06909112167394423, "grad_norm": 2.2068090438842773, "learning_rate": 1.1446065554346997e-05, "loss": 2.5139, "step": 7682 }, { "epoch": 0.06910011557158469, "grad_norm": 14.019685745239258, "learning_rate": 1.1439750836937665e-05, "loss": 1.1629, "step": 7683 }, { "epoch": 0.06910910946922513, "grad_norm": 3.4466209411621094, "learning_rate": 1.143343734508904e-05, "loss": 1.6593, "step": 7684 }, { "epoch": 0.06911810336686558, "grad_norm": 11.112135887145996, "learning_rate": 1.1427125079371745e-05, "loss": 1.8362, "step": 7685 }, { "epoch": 0.06912709726450603, "grad_norm": 3.028397560119629, "learning_rate": 1.1420814040356262e-05, "loss": 1.3699, "step": 7686 }, { "epoch": 0.06913609116214649, "grad_norm": 1.9145199060440063, "learning_rate": 1.1414504228612983e-05, "loss": 2.2582, "step": 7687 }, { "epoch": 0.06914508505978693, "grad_norm": 19.079757690429688, "learning_rate": 1.1408195644712169e-05, "loss": 0.8359, "step": 7688 }, { "epoch": 0.06915407895742738, "grad_norm": 3.070014238357544, "learning_rate": 1.1401888289224e-05, "loss": 1.4054, "step": 7689 }, { "epoch": 0.06916307285506784, "grad_norm": 14.800333023071289, "learning_rate": 1.1395582162718525e-05, "loss": 1.5405, "step": 7690 }, { "epoch": 0.06917206675270829, "grad_norm": 1.9088122844696045, "learning_rate": 1.1389277265765666e-05, "loss": 2.5431, "step": 7691 }, { "epoch": 0.06918106065034874, "grad_norm": 8.660789489746094, "learning_rate": 1.1382973598935278e-05, "loss": 0.7197, "step": 7692 }, { "epoch": 0.06919005454798918, "grad_norm": 3.4990761280059814, "learning_rate": 1.1376671162797062e-05, "loss": 0.9389, "step": 7693 }, { "epoch": 0.06919904844562964, "grad_norm": 20.414005279541016, "learning_rate": 1.137036995792064e-05, "loss": 0.7309, "step": 7694 }, { "epoch": 0.06920804234327009, "grad_norm": 2.7831664085388184, "learning_rate": 1.1364069984875503e-05, "loss": 2.204, "step": 7695 }, { "epoch": 0.06921703624091054, "grad_norm": 3.3964288234710693, "learning_rate": 1.1357771244231023e-05, "loss": 1.3027, "step": 7696 }, { "epoch": 0.069226030138551, "grad_norm": 8.857315063476562, "learning_rate": 1.1351473736556492e-05, "loss": 2.0963, "step": 7697 }, { "epoch": 0.06923502403619144, "grad_norm": 2.5584216117858887, "learning_rate": 1.1345177462421052e-05, "loss": 2.0038, "step": 7698 }, { "epoch": 0.06924401793383189, "grad_norm": 17.442121505737305, "learning_rate": 1.1338882422393779e-05, "loss": 0.6653, "step": 7699 }, { "epoch": 0.06925301183147235, "grad_norm": 3.3891165256500244, "learning_rate": 1.1332588617043596e-05, "loss": 1.0135, "step": 7700 }, { "epoch": 0.0692620057291128, "grad_norm": 3.676039457321167, "learning_rate": 1.1326296046939333e-05, "loss": 0.8594, "step": 7701 }, { "epoch": 0.06927099962675325, "grad_norm": 2.573209285736084, "learning_rate": 1.1320004712649706e-05, "loss": 1.6109, "step": 7702 }, { "epoch": 0.0692799935243937, "grad_norm": 2.9935739040374756, "learning_rate": 1.1313714614743306e-05, "loss": 1.8175, "step": 7703 }, { "epoch": 0.06928898742203415, "grad_norm": 3.0742979049682617, "learning_rate": 1.130742575378865e-05, "loss": 1.9026, "step": 7704 }, { "epoch": 0.0692979813196746, "grad_norm": 16.420867919921875, "learning_rate": 1.1301138130354095e-05, "loss": 1.2764, "step": 7705 }, { "epoch": 0.06930697521731505, "grad_norm": 11.115218162536621, "learning_rate": 1.129485174500793e-05, "loss": 1.6603, "step": 7706 }, { "epoch": 0.06931596911495551, "grad_norm": 2.8926546573638916, "learning_rate": 1.1288566598318303e-05, "loss": 1.4724, "step": 7707 }, { "epoch": 0.06932496301259596, "grad_norm": 3.4349045753479004, "learning_rate": 1.128228269085325e-05, "loss": 1.2772, "step": 7708 }, { "epoch": 0.0693339569102364, "grad_norm": 2.79854154586792, "learning_rate": 1.1276000023180719e-05, "loss": 1.021, "step": 7709 }, { "epoch": 0.06934295080787685, "grad_norm": 2.961493730545044, "learning_rate": 1.1269718595868514e-05, "loss": 2.2991, "step": 7710 }, { "epoch": 0.06935194470551731, "grad_norm": 17.567848205566406, "learning_rate": 1.1263438409484363e-05, "loss": 1.5683, "step": 7711 }, { "epoch": 0.06936093860315776, "grad_norm": 2.908236026763916, "learning_rate": 1.1257159464595855e-05, "loss": 2.011, "step": 7712 }, { "epoch": 0.0693699325007982, "grad_norm": 2.7864015102386475, "learning_rate": 1.1250881761770459e-05, "loss": 2.3774, "step": 7713 }, { "epoch": 0.06937892639843866, "grad_norm": 2.5261855125427246, "learning_rate": 1.1244605301575572e-05, "loss": 2.0369, "step": 7714 }, { "epoch": 0.06938792029607911, "grad_norm": 9.40592098236084, "learning_rate": 1.1238330084578441e-05, "loss": 1.0389, "step": 7715 }, { "epoch": 0.06939691419371956, "grad_norm": 2.111018180847168, "learning_rate": 1.1232056111346215e-05, "loss": 2.8103, "step": 7716 }, { "epoch": 0.06940590809136002, "grad_norm": 2.70007061958313, "learning_rate": 1.1225783382445925e-05, "loss": 1.1066, "step": 7717 }, { "epoch": 0.06941490198900047, "grad_norm": 2.814936876296997, "learning_rate": 1.121951189844449e-05, "loss": 2.996, "step": 7718 }, { "epoch": 0.06942389588664091, "grad_norm": 2.7893102169036865, "learning_rate": 1.1213241659908736e-05, "loss": 1.8691, "step": 7719 }, { "epoch": 0.06943288978428136, "grad_norm": 14.396574974060059, "learning_rate": 1.1206972667405341e-05, "loss": 1.6921, "step": 7720 }, { "epoch": 0.06944188368192182, "grad_norm": 25.153654098510742, "learning_rate": 1.120070492150091e-05, "loss": 0.6785, "step": 7721 }, { "epoch": 0.06945087757956227, "grad_norm": 18.681631088256836, "learning_rate": 1.1194438422761894e-05, "loss": 1.6293, "step": 7722 }, { "epoch": 0.06945987147720271, "grad_norm": 7.984598636627197, "learning_rate": 1.118817317175467e-05, "loss": 1.3012, "step": 7723 }, { "epoch": 0.06946886537484318, "grad_norm": 10.673260688781738, "learning_rate": 1.1181909169045482e-05, "loss": 1.6312, "step": 7724 }, { "epoch": 0.06947785927248362, "grad_norm": 13.621559143066406, "learning_rate": 1.117564641520045e-05, "loss": 1.0058, "step": 7725 }, { "epoch": 0.06948685317012407, "grad_norm": 2.781395435333252, "learning_rate": 1.1169384910785614e-05, "loss": 2.3703, "step": 7726 }, { "epoch": 0.06949584706776452, "grad_norm": 8.250746726989746, "learning_rate": 1.1163124656366861e-05, "loss": 0.4772, "step": 7727 }, { "epoch": 0.06950484096540498, "grad_norm": 2.6771883964538574, "learning_rate": 1.1156865652510009e-05, "loss": 1.6469, "step": 7728 }, { "epoch": 0.06951383486304542, "grad_norm": 2.435119390487671, "learning_rate": 1.1150607899780727e-05, "loss": 2.2362, "step": 7729 }, { "epoch": 0.06952282876068587, "grad_norm": 11.62525463104248, "learning_rate": 1.1144351398744587e-05, "loss": 1.4584, "step": 7730 }, { "epoch": 0.06953182265832633, "grad_norm": 3.305568218231201, "learning_rate": 1.1138096149967044e-05, "loss": 2.0974, "step": 7731 }, { "epoch": 0.06954081655596678, "grad_norm": 2.767754316329956, "learning_rate": 1.113184215401343e-05, "loss": 2.1868, "step": 7732 }, { "epoch": 0.06954981045360722, "grad_norm": 4.386049747467041, "learning_rate": 1.1125589411448994e-05, "loss": 0.7566, "step": 7733 }, { "epoch": 0.06955880435124767, "grad_norm": 7.314235687255859, "learning_rate": 1.1119337922838832e-05, "loss": 1.6401, "step": 7734 }, { "epoch": 0.06956779824888813, "grad_norm": 17.86168670654297, "learning_rate": 1.111308768874797e-05, "loss": 1.5583, "step": 7735 }, { "epoch": 0.06957679214652858, "grad_norm": 4.037002086639404, "learning_rate": 1.1106838709741282e-05, "loss": 1.6031, "step": 7736 }, { "epoch": 0.06958578604416903, "grad_norm": 5.390331268310547, "learning_rate": 1.1100590986383538e-05, "loss": 0.4796, "step": 7737 }, { "epoch": 0.06959477994180949, "grad_norm": 4.459733009338379, "learning_rate": 1.1094344519239422e-05, "loss": 0.8678, "step": 7738 }, { "epoch": 0.06960377383944993, "grad_norm": 10.114385604858398, "learning_rate": 1.108809930887346e-05, "loss": 1.6709, "step": 7739 }, { "epoch": 0.06961276773709038, "grad_norm": 14.246051788330078, "learning_rate": 1.1081855355850105e-05, "loss": 1.365, "step": 7740 }, { "epoch": 0.06962176163473084, "grad_norm": 2.7820980548858643, "learning_rate": 1.107561266073367e-05, "loss": 1.814, "step": 7741 }, { "epoch": 0.06963075553237129, "grad_norm": 4.280333042144775, "learning_rate": 1.1069371224088368e-05, "loss": 1.2815, "step": 7742 }, { "epoch": 0.06963974943001174, "grad_norm": 2.201251268386841, "learning_rate": 1.1063131046478279e-05, "loss": 2.8429, "step": 7743 }, { "epoch": 0.06964874332765218, "grad_norm": 3.757612466812134, "learning_rate": 1.1056892128467406e-05, "loss": 0.9855, "step": 7744 }, { "epoch": 0.06965773722529264, "grad_norm": 3.973146915435791, "learning_rate": 1.1050654470619601e-05, "loss": 1.9514, "step": 7745 }, { "epoch": 0.06966673112293309, "grad_norm": 7.101302146911621, "learning_rate": 1.1044418073498614e-05, "loss": 1.0624, "step": 7746 }, { "epoch": 0.06967572502057354, "grad_norm": 12.136322975158691, "learning_rate": 1.1038182937668096e-05, "loss": 1.6048, "step": 7747 }, { "epoch": 0.069684718918214, "grad_norm": 3.2175707817077637, "learning_rate": 1.1031949063691566e-05, "loss": 1.4075, "step": 7748 }, { "epoch": 0.06969371281585444, "grad_norm": 2.3161559104919434, "learning_rate": 1.1025716452132425e-05, "loss": 2.8535, "step": 7749 }, { "epoch": 0.06970270671349489, "grad_norm": 3.782304525375366, "learning_rate": 1.1019485103553992e-05, "loss": 2.0448, "step": 7750 }, { "epoch": 0.06971170061113534, "grad_norm": 2.756338119506836, "learning_rate": 1.1013255018519425e-05, "loss": 2.1568, "step": 7751 }, { "epoch": 0.0697206945087758, "grad_norm": 3.097546339035034, "learning_rate": 1.1007026197591813e-05, "loss": 1.5015, "step": 7752 }, { "epoch": 0.06972968840641625, "grad_norm": 11.921220779418945, "learning_rate": 1.10007986413341e-05, "loss": 0.3298, "step": 7753 }, { "epoch": 0.06973868230405669, "grad_norm": 3.088573455810547, "learning_rate": 1.099457235030912e-05, "loss": 1.4383, "step": 7754 }, { "epoch": 0.06974767620169715, "grad_norm": 2.5230464935302734, "learning_rate": 1.0988347325079615e-05, "loss": 1.9781, "step": 7755 }, { "epoch": 0.0697566700993376, "grad_norm": 2.998831272125244, "learning_rate": 1.0982123566208185e-05, "loss": 1.3263, "step": 7756 }, { "epoch": 0.06976566399697805, "grad_norm": 2.663666248321533, "learning_rate": 1.097590107425733e-05, "loss": 2.4061, "step": 7757 }, { "epoch": 0.06977465789461851, "grad_norm": 2.4916985034942627, "learning_rate": 1.0969679849789421e-05, "loss": 2.3311, "step": 7758 }, { "epoch": 0.06978365179225895, "grad_norm": 2.5721542835235596, "learning_rate": 1.0963459893366743e-05, "loss": 1.5811, "step": 7759 }, { "epoch": 0.0697926456898994, "grad_norm": 2.482348918914795, "learning_rate": 1.0957241205551441e-05, "loss": 2.0934, "step": 7760 }, { "epoch": 0.06980163958753985, "grad_norm": 2.481661319732666, "learning_rate": 1.0951023786905545e-05, "loss": 2.6912, "step": 7761 }, { "epoch": 0.06981063348518031, "grad_norm": 2.659541130065918, "learning_rate": 1.0944807637990998e-05, "loss": 1.5286, "step": 7762 }, { "epoch": 0.06981962738282076, "grad_norm": 4.3550190925598145, "learning_rate": 1.0938592759369584e-05, "loss": 0.4812, "step": 7763 }, { "epoch": 0.0698286212804612, "grad_norm": 2.184509754180908, "learning_rate": 1.0932379151603022e-05, "loss": 2.1275, "step": 7764 }, { "epoch": 0.06983761517810166, "grad_norm": 13.078556060791016, "learning_rate": 1.092616681525288e-05, "loss": 0.8459, "step": 7765 }, { "epoch": 0.06984660907574211, "grad_norm": 2.0539724826812744, "learning_rate": 1.0919955750880614e-05, "loss": 1.9769, "step": 7766 }, { "epoch": 0.06985560297338256, "grad_norm": 2.431086301803589, "learning_rate": 1.091374595904759e-05, "loss": 2.1532, "step": 7767 }, { "epoch": 0.069864596871023, "grad_norm": 12.550600051879883, "learning_rate": 1.0907537440315024e-05, "loss": 1.3976, "step": 7768 }, { "epoch": 0.06987359076866347, "grad_norm": 16.17007064819336, "learning_rate": 1.0901330195244053e-05, "loss": 1.0344, "step": 7769 }, { "epoch": 0.06988258466630391, "grad_norm": 2.1913671493530273, "learning_rate": 1.0895124224395676e-05, "loss": 1.9166, "step": 7770 }, { "epoch": 0.06989157856394436, "grad_norm": 2.9290831089019775, "learning_rate": 1.0888919528330777e-05, "loss": 1.5895, "step": 7771 }, { "epoch": 0.06990057246158482, "grad_norm": 2.298097848892212, "learning_rate": 1.088271610761013e-05, "loss": 1.8824, "step": 7772 }, { "epoch": 0.06990956635922527, "grad_norm": 7.890809535980225, "learning_rate": 1.0876513962794391e-05, "loss": 1.0838, "step": 7773 }, { "epoch": 0.06991856025686571, "grad_norm": 2.6974871158599854, "learning_rate": 1.0870313094444111e-05, "loss": 1.4502, "step": 7774 }, { "epoch": 0.06992755415450616, "grad_norm": 19.417449951171875, "learning_rate": 1.086411350311971e-05, "loss": 1.7303, "step": 7775 }, { "epoch": 0.06993654805214662, "grad_norm": 13.14277458190918, "learning_rate": 1.0857915189381513e-05, "loss": 1.2879, "step": 7776 }, { "epoch": 0.06994554194978707, "grad_norm": 3.2016210556030273, "learning_rate": 1.0851718153789705e-05, "loss": 1.3502, "step": 7777 }, { "epoch": 0.06995453584742752, "grad_norm": 3.2819430828094482, "learning_rate": 1.0845522396904367e-05, "loss": 1.4175, "step": 7778 }, { "epoch": 0.06996352974506798, "grad_norm": 2.626351833343506, "learning_rate": 1.0839327919285475e-05, "loss": 2.4042, "step": 7779 }, { "epoch": 0.06997252364270842, "grad_norm": 13.379611015319824, "learning_rate": 1.0833134721492864e-05, "loss": 1.4826, "step": 7780 }, { "epoch": 0.06998151754034887, "grad_norm": 10.39612102508545, "learning_rate": 1.0826942804086288e-05, "loss": 0.9642, "step": 7781 }, { "epoch": 0.06999051143798933, "grad_norm": 2.4484856128692627, "learning_rate": 1.0820752167625354e-05, "loss": 2.3873, "step": 7782 }, { "epoch": 0.06999950533562978, "grad_norm": 31.240718841552734, "learning_rate": 1.0814562812669557e-05, "loss": 0.5429, "step": 7783 }, { "epoch": 0.07000849923327022, "grad_norm": 3.0164635181427, "learning_rate": 1.0808374739778304e-05, "loss": 1.5568, "step": 7784 }, { "epoch": 0.07001749313091067, "grad_norm": 5.801798343658447, "learning_rate": 1.0802187949510856e-05, "loss": 1.9248, "step": 7785 }, { "epoch": 0.07002648702855113, "grad_norm": 2.5459415912628174, "learning_rate": 1.0796002442426367e-05, "loss": 2.3828, "step": 7786 }, { "epoch": 0.07003548092619158, "grad_norm": 2.7284786701202393, "learning_rate": 1.078981821908387e-05, "loss": 2.3877, "step": 7787 }, { "epoch": 0.07004447482383203, "grad_norm": 3.3579297065734863, "learning_rate": 1.0783635280042306e-05, "loss": 1.6592, "step": 7788 }, { "epoch": 0.07005346872147249, "grad_norm": 3.2986838817596436, "learning_rate": 1.0777453625860472e-05, "loss": 1.938, "step": 7789 }, { "epoch": 0.07006246261911293, "grad_norm": 8.538931846618652, "learning_rate": 1.077127325709705e-05, "loss": 1.7558, "step": 7790 }, { "epoch": 0.07007145651675338, "grad_norm": 13.491140365600586, "learning_rate": 1.0765094174310636e-05, "loss": 0.8695, "step": 7791 }, { "epoch": 0.07008045041439383, "grad_norm": 15.009162902832031, "learning_rate": 1.075891637805967e-05, "loss": 0.6768, "step": 7792 }, { "epoch": 0.07008944431203429, "grad_norm": 2.3264636993408203, "learning_rate": 1.0752739868902509e-05, "loss": 2.0711, "step": 7793 }, { "epoch": 0.07009843820967473, "grad_norm": 7.555400371551514, "learning_rate": 1.0746564647397373e-05, "loss": 0.7385, "step": 7794 }, { "epoch": 0.07010743210731518, "grad_norm": 2.480832099914551, "learning_rate": 1.0740390714102364e-05, "loss": 1.9573, "step": 7795 }, { "epoch": 0.07011642600495564, "grad_norm": 2.242246627807617, "learning_rate": 1.0734218069575492e-05, "loss": 2.6146, "step": 7796 }, { "epoch": 0.07012541990259609, "grad_norm": 13.4462251663208, "learning_rate": 1.0728046714374615e-05, "loss": 1.4627, "step": 7797 }, { "epoch": 0.07013441380023654, "grad_norm": 9.146303176879883, "learning_rate": 1.0721876649057513e-05, "loss": 1.0368, "step": 7798 }, { "epoch": 0.070143407697877, "grad_norm": 3.104645013809204, "learning_rate": 1.0715707874181819e-05, "loss": 2.1988, "step": 7799 }, { "epoch": 0.07015240159551744, "grad_norm": 2.4473438262939453, "learning_rate": 1.0709540390305061e-05, "loss": 1.916, "step": 7800 }, { "epoch": 0.07016139549315789, "grad_norm": 2.2877867221832275, "learning_rate": 1.0703374197984653e-05, "loss": 2.4803, "step": 7801 }, { "epoch": 0.07017038939079834, "grad_norm": 2.0289695262908936, "learning_rate": 1.0697209297777878e-05, "loss": 2.109, "step": 7802 }, { "epoch": 0.0701793832884388, "grad_norm": 3.27011775970459, "learning_rate": 1.0691045690241927e-05, "loss": 1.7005, "step": 7803 }, { "epoch": 0.07018837718607925, "grad_norm": 2.7007625102996826, "learning_rate": 1.0684883375933846e-05, "loss": 2.2689, "step": 7804 }, { "epoch": 0.07019737108371969, "grad_norm": 11.025697708129883, "learning_rate": 1.0678722355410595e-05, "loss": 0.89, "step": 7805 }, { "epoch": 0.07020636498136015, "grad_norm": 9.441170692443848, "learning_rate": 1.067256262922899e-05, "loss": 1.4142, "step": 7806 }, { "epoch": 0.0702153588790006, "grad_norm": 14.113212585449219, "learning_rate": 1.0666404197945738e-05, "loss": 1.1888, "step": 7807 }, { "epoch": 0.07022435277664105, "grad_norm": 3.2713208198547363, "learning_rate": 1.0660247062117443e-05, "loss": 2.0139, "step": 7808 }, { "epoch": 0.0702333466742815, "grad_norm": 3.121835708618164, "learning_rate": 1.0654091222300564e-05, "loss": 2.3007, "step": 7809 }, { "epoch": 0.07024234057192195, "grad_norm": 2.1815884113311768, "learning_rate": 1.0647936679051476e-05, "loss": 2.6876, "step": 7810 }, { "epoch": 0.0702513344695624, "grad_norm": 2.9724295139312744, "learning_rate": 1.064178343292641e-05, "loss": 2.3302, "step": 7811 }, { "epoch": 0.07026032836720285, "grad_norm": 2.9969406127929688, "learning_rate": 1.0635631484481496e-05, "loss": 1.9293, "step": 7812 }, { "epoch": 0.07026932226484331, "grad_norm": 8.466670036315918, "learning_rate": 1.0629480834272734e-05, "loss": 1.108, "step": 7813 }, { "epoch": 0.07027831616248376, "grad_norm": 3.9270787239074707, "learning_rate": 1.0623331482856009e-05, "loss": 1.3083, "step": 7814 }, { "epoch": 0.0702873100601242, "grad_norm": 3.096050977706909, "learning_rate": 1.0617183430787105e-05, "loss": 1.9534, "step": 7815 }, { "epoch": 0.07029630395776465, "grad_norm": 4.875589370727539, "learning_rate": 1.0611036678621666e-05, "loss": 0.4673, "step": 7816 }, { "epoch": 0.07030529785540511, "grad_norm": 5.171274662017822, "learning_rate": 1.060489122691524e-05, "loss": 1.9571, "step": 7817 }, { "epoch": 0.07031429175304556, "grad_norm": 9.276239395141602, "learning_rate": 1.0598747076223239e-05, "loss": 0.564, "step": 7818 }, { "epoch": 0.070323285650686, "grad_norm": 2.360788106918335, "learning_rate": 1.0592604227100958e-05, "loss": 1.8669, "step": 7819 }, { "epoch": 0.07033227954832647, "grad_norm": 2.452322483062744, "learning_rate": 1.0586462680103596e-05, "loss": 2.4349, "step": 7820 }, { "epoch": 0.07034127344596691, "grad_norm": 17.850767135620117, "learning_rate": 1.0580322435786204e-05, "loss": 1.6466, "step": 7821 }, { "epoch": 0.07035026734360736, "grad_norm": 12.194485664367676, "learning_rate": 1.0574183494703748e-05, "loss": 0.481, "step": 7822 }, { "epoch": 0.07035926124124782, "grad_norm": 9.175263404846191, "learning_rate": 1.0568045857411041e-05, "loss": 1.7374, "step": 7823 }, { "epoch": 0.07036825513888827, "grad_norm": 2.7750988006591797, "learning_rate": 1.0561909524462812e-05, "loss": 2.0851, "step": 7824 }, { "epoch": 0.07037724903652871, "grad_norm": 9.63393783569336, "learning_rate": 1.055577449641365e-05, "loss": 0.7293, "step": 7825 }, { "epoch": 0.07038624293416916, "grad_norm": 2.6138696670532227, "learning_rate": 1.0549640773818029e-05, "loss": 1.9145, "step": 7826 }, { "epoch": 0.07039523683180962, "grad_norm": 3.3945772647857666, "learning_rate": 1.0543508357230311e-05, "loss": 2.0101, "step": 7827 }, { "epoch": 0.07040423072945007, "grad_norm": 2.4644217491149902, "learning_rate": 1.053737724720473e-05, "loss": 2.1267, "step": 7828 }, { "epoch": 0.07041322462709051, "grad_norm": 3.0670833587646484, "learning_rate": 1.0531247444295422e-05, "loss": 1.0146, "step": 7829 }, { "epoch": 0.07042221852473098, "grad_norm": 2.253021478652954, "learning_rate": 1.0525118949056386e-05, "loss": 2.35, "step": 7830 }, { "epoch": 0.07043121242237142, "grad_norm": 9.398479461669922, "learning_rate": 1.05189917620415e-05, "loss": 1.0357, "step": 7831 }, { "epoch": 0.07044020632001187, "grad_norm": 11.610546112060547, "learning_rate": 1.0512865883804548e-05, "loss": 0.3528, "step": 7832 }, { "epoch": 0.07044920021765232, "grad_norm": 6.845404148101807, "learning_rate": 1.0506741314899166e-05, "loss": 0.4384, "step": 7833 }, { "epoch": 0.07045819411529278, "grad_norm": 3.206779956817627, "learning_rate": 1.0500618055878902e-05, "loss": 1.7805, "step": 7834 }, { "epoch": 0.07046718801293322, "grad_norm": 15.861336708068848, "learning_rate": 1.0494496107297153e-05, "loss": 0.8285, "step": 7835 }, { "epoch": 0.07047618191057367, "grad_norm": 3.9714221954345703, "learning_rate": 1.0488375469707226e-05, "loss": 1.4877, "step": 7836 }, { "epoch": 0.07048517580821413, "grad_norm": 2.617220878601074, "learning_rate": 1.0482256143662298e-05, "loss": 1.8445, "step": 7837 }, { "epoch": 0.07049416970585458, "grad_norm": 13.074334144592285, "learning_rate": 1.0476138129715413e-05, "loss": 1.5326, "step": 7838 }, { "epoch": 0.07050316360349503, "grad_norm": 15.692573547363281, "learning_rate": 1.0470021428419528e-05, "loss": 1.0334, "step": 7839 }, { "epoch": 0.07051215750113549, "grad_norm": 9.369380950927734, "learning_rate": 1.0463906040327453e-05, "loss": 0.6317, "step": 7840 }, { "epoch": 0.07052115139877593, "grad_norm": 3.4399800300598145, "learning_rate": 1.0457791965991898e-05, "loss": 1.832, "step": 7841 }, { "epoch": 0.07053014529641638, "grad_norm": 11.185629844665527, "learning_rate": 1.0451679205965439e-05, "loss": 1.5186, "step": 7842 }, { "epoch": 0.07053913919405683, "grad_norm": 2.8009049892425537, "learning_rate": 1.0445567760800535e-05, "loss": 2.4622, "step": 7843 }, { "epoch": 0.07054813309169729, "grad_norm": 8.275794982910156, "learning_rate": 1.0439457631049549e-05, "loss": 0.9633, "step": 7844 }, { "epoch": 0.07055712698933773, "grad_norm": 2.3691978454589844, "learning_rate": 1.0433348817264691e-05, "loss": 2.2629, "step": 7845 }, { "epoch": 0.07056612088697818, "grad_norm": 3.392915725708008, "learning_rate": 1.0427241319998083e-05, "loss": 1.4131, "step": 7846 }, { "epoch": 0.07057511478461864, "grad_norm": 10.512736320495605, "learning_rate": 1.0421135139801702e-05, "loss": 1.1983, "step": 7847 }, { "epoch": 0.07058410868225909, "grad_norm": 14.063602447509766, "learning_rate": 1.041503027722743e-05, "loss": 0.6536, "step": 7848 }, { "epoch": 0.07059310257989954, "grad_norm": 2.575723171234131, "learning_rate": 1.0408926732827015e-05, "loss": 1.6012, "step": 7849 }, { "epoch": 0.07060209647753998, "grad_norm": 3.5188283920288086, "learning_rate": 1.0402824507152075e-05, "loss": 2.0674, "step": 7850 }, { "epoch": 0.07061109037518044, "grad_norm": 2.8089754581451416, "learning_rate": 1.0396723600754143e-05, "loss": 1.6365, "step": 7851 }, { "epoch": 0.07062008427282089, "grad_norm": 3.6967525482177734, "learning_rate": 1.039062401418459e-05, "loss": 1.5658, "step": 7852 }, { "epoch": 0.07062907817046134, "grad_norm": 2.8592636585235596, "learning_rate": 1.0384525747994716e-05, "loss": 2.3021, "step": 7853 }, { "epoch": 0.0706380720681018, "grad_norm": 2.3881325721740723, "learning_rate": 1.0378428802735662e-05, "loss": 2.6529, "step": 7854 }, { "epoch": 0.07064706596574225, "grad_norm": 10.175576210021973, "learning_rate": 1.0372333178958462e-05, "loss": 1.0855, "step": 7855 }, { "epoch": 0.07065605986338269, "grad_norm": 2.949366807937622, "learning_rate": 1.0366238877214035e-05, "loss": 2.2549, "step": 7856 }, { "epoch": 0.07066505376102315, "grad_norm": 3.594358205795288, "learning_rate": 1.0360145898053166e-05, "loss": 1.3721, "step": 7857 }, { "epoch": 0.0706740476586636, "grad_norm": 18.41791343688965, "learning_rate": 1.0354054242026553e-05, "loss": 0.5466, "step": 7858 }, { "epoch": 0.07068304155630405, "grad_norm": 11.709961891174316, "learning_rate": 1.034796390968474e-05, "loss": 1.4331, "step": 7859 }, { "epoch": 0.0706920354539445, "grad_norm": 2.870100259780884, "learning_rate": 1.0341874901578161e-05, "loss": 2.0638, "step": 7860 }, { "epoch": 0.07070102935158495, "grad_norm": 4.05856990814209, "learning_rate": 1.0335787218257148e-05, "loss": 1.1753, "step": 7861 }, { "epoch": 0.0707100232492254, "grad_norm": 3.3138887882232666, "learning_rate": 1.032970086027188e-05, "loss": 1.4623, "step": 7862 }, { "epoch": 0.07071901714686585, "grad_norm": 3.5967283248901367, "learning_rate": 1.0323615828172456e-05, "loss": 1.7466, "step": 7863 }, { "epoch": 0.07072801104450631, "grad_norm": 2.7744758129119873, "learning_rate": 1.031753212250882e-05, "loss": 2.0838, "step": 7864 }, { "epoch": 0.07073700494214676, "grad_norm": 3.3881754875183105, "learning_rate": 1.031144974383082e-05, "loss": 1.5549, "step": 7865 }, { "epoch": 0.0707459988397872, "grad_norm": 2.5720326900482178, "learning_rate": 1.0305368692688174e-05, "loss": 2.5558, "step": 7866 }, { "epoch": 0.07075499273742765, "grad_norm": 3.72341251373291, "learning_rate": 1.0299288969630477e-05, "loss": 1.3737, "step": 7867 }, { "epoch": 0.07076398663506811, "grad_norm": 2.103339672088623, "learning_rate": 1.0293210575207199e-05, "loss": 2.5798, "step": 7868 }, { "epoch": 0.07077298053270856, "grad_norm": 5.038074016571045, "learning_rate": 1.028713350996772e-05, "loss": 1.8841, "step": 7869 }, { "epoch": 0.070781974430349, "grad_norm": 2.9219369888305664, "learning_rate": 1.0281057774461265e-05, "loss": 2.4089, "step": 7870 }, { "epoch": 0.07079096832798946, "grad_norm": 11.693636894226074, "learning_rate": 1.0274983369236953e-05, "loss": 0.5465, "step": 7871 }, { "epoch": 0.07079996222562991, "grad_norm": 10.161890983581543, "learning_rate": 1.0268910294843775e-05, "loss": 1.8765, "step": 7872 }, { "epoch": 0.07080895612327036, "grad_norm": 2.551340341567993, "learning_rate": 1.0262838551830628e-05, "loss": 1.9849, "step": 7873 }, { "epoch": 0.0708179500209108, "grad_norm": 8.391427040100098, "learning_rate": 1.0256768140746251e-05, "loss": 1.3328, "step": 7874 }, { "epoch": 0.07082694391855127, "grad_norm": 2.4468679428100586, "learning_rate": 1.0250699062139296e-05, "loss": 2.6034, "step": 7875 }, { "epoch": 0.07083593781619171, "grad_norm": 10.71274471282959, "learning_rate": 1.0244631316558267e-05, "loss": 1.3429, "step": 7876 }, { "epoch": 0.07084493171383216, "grad_norm": 2.34578800201416, "learning_rate": 1.0238564904551574e-05, "loss": 1.9167, "step": 7877 }, { "epoch": 0.07085392561147262, "grad_norm": 2.710768222808838, "learning_rate": 1.0232499826667483e-05, "loss": 1.6779, "step": 7878 }, { "epoch": 0.07086291950911307, "grad_norm": 2.6167924404144287, "learning_rate": 1.0226436083454144e-05, "loss": 2.2585, "step": 7879 }, { "epoch": 0.07087191340675351, "grad_norm": 2.370802402496338, "learning_rate": 1.0220373675459611e-05, "loss": 2.4304, "step": 7880 }, { "epoch": 0.07088090730439398, "grad_norm": 2.537224054336548, "learning_rate": 1.0214312603231782e-05, "loss": 2.0706, "step": 7881 }, { "epoch": 0.07088990120203442, "grad_norm": 2.155724287033081, "learning_rate": 1.0208252867318455e-05, "loss": 1.9569, "step": 7882 }, { "epoch": 0.07089889509967487, "grad_norm": 3.1207456588745117, "learning_rate": 1.0202194468267302e-05, "loss": 1.5396, "step": 7883 }, { "epoch": 0.07090788899731532, "grad_norm": 2.0763354301452637, "learning_rate": 1.0196137406625864e-05, "loss": 1.5141, "step": 7884 }, { "epoch": 0.07091688289495578, "grad_norm": 2.776745557785034, "learning_rate": 1.0190081682941593e-05, "loss": 1.6629, "step": 7885 }, { "epoch": 0.07092587679259622, "grad_norm": 14.40388298034668, "learning_rate": 1.0184027297761778e-05, "loss": 1.6215, "step": 7886 }, { "epoch": 0.07093487069023667, "grad_norm": 20.654918670654297, "learning_rate": 1.0177974251633623e-05, "loss": 1.6613, "step": 7887 }, { "epoch": 0.07094386458787713, "grad_norm": 2.840684175491333, "learning_rate": 1.0171922545104184e-05, "loss": 1.909, "step": 7888 }, { "epoch": 0.07095285848551758, "grad_norm": 2.13383412361145, "learning_rate": 1.0165872178720424e-05, "loss": 2.4141, "step": 7889 }, { "epoch": 0.07096185238315803, "grad_norm": 2.6795191764831543, "learning_rate": 1.0159823153029158e-05, "loss": 1.3069, "step": 7890 }, { "epoch": 0.07097084628079847, "grad_norm": 23.427581787109375, "learning_rate": 1.0153775468577079e-05, "loss": 1.6, "step": 7891 }, { "epoch": 0.07097984017843893, "grad_norm": 9.635080337524414, "learning_rate": 1.0147729125910792e-05, "loss": 0.6278, "step": 7892 }, { "epoch": 0.07098883407607938, "grad_norm": 9.102328300476074, "learning_rate": 1.0141684125576742e-05, "loss": 1.2905, "step": 7893 }, { "epoch": 0.07099782797371983, "grad_norm": 2.456193208694458, "learning_rate": 1.0135640468121286e-05, "loss": 2.3873, "step": 7894 }, { "epoch": 0.07100682187136029, "grad_norm": 8.373101234436035, "learning_rate": 1.0129598154090633e-05, "loss": 1.1067, "step": 7895 }, { "epoch": 0.07101581576900073, "grad_norm": 3.1816797256469727, "learning_rate": 1.0123557184030882e-05, "loss": 1.4649, "step": 7896 }, { "epoch": 0.07102480966664118, "grad_norm": 2.2343616485595703, "learning_rate": 1.0117517558488007e-05, "loss": 2.257, "step": 7897 }, { "epoch": 0.07103380356428164, "grad_norm": 2.409592866897583, "learning_rate": 1.011147927800786e-05, "loss": 2.0724, "step": 7898 }, { "epoch": 0.07104279746192209, "grad_norm": 20.449905395507812, "learning_rate": 1.0105442343136184e-05, "loss": 1.1891, "step": 7899 }, { "epoch": 0.07105179135956254, "grad_norm": 13.7026948928833, "learning_rate": 1.009940675441858e-05, "loss": 1.1311, "step": 7900 }, { "epoch": 0.07106078525720298, "grad_norm": 20.361679077148438, "learning_rate": 1.0093372512400551e-05, "loss": 1.0926, "step": 7901 }, { "epoch": 0.07106977915484344, "grad_norm": 2.034632444381714, "learning_rate": 1.0087339617627459e-05, "loss": 1.5523, "step": 7902 }, { "epoch": 0.07107877305248389, "grad_norm": 2.7574620246887207, "learning_rate": 1.0081308070644541e-05, "loss": 2.7382, "step": 7903 }, { "epoch": 0.07108776695012434, "grad_norm": 2.5789480209350586, "learning_rate": 1.0075277871996938e-05, "loss": 2.3767, "step": 7904 }, { "epoch": 0.0710967608477648, "grad_norm": 2.984908103942871, "learning_rate": 1.006924902222964e-05, "loss": 1.6014, "step": 7905 }, { "epoch": 0.07110575474540524, "grad_norm": 2.4387362003326416, "learning_rate": 1.0063221521887539e-05, "loss": 2.2503, "step": 7906 }, { "epoch": 0.07111474864304569, "grad_norm": 18.051734924316406, "learning_rate": 1.0057195371515388e-05, "loss": 1.7612, "step": 7907 }, { "epoch": 0.07112374254068614, "grad_norm": 12.83596134185791, "learning_rate": 1.0051170571657817e-05, "loss": 1.2237, "step": 7908 }, { "epoch": 0.0711327364383266, "grad_norm": 3.3270487785339355, "learning_rate": 1.0045147122859358e-05, "loss": 0.9449, "step": 7909 }, { "epoch": 0.07114173033596705, "grad_norm": 3.082066535949707, "learning_rate": 1.0039125025664392e-05, "loss": 1.9937, "step": 7910 }, { "epoch": 0.07115072423360749, "grad_norm": 13.52994441986084, "learning_rate": 1.0033104280617192e-05, "loss": 1.6045, "step": 7911 }, { "epoch": 0.07115971813124795, "grad_norm": 2.7167365550994873, "learning_rate": 1.0027084888261899e-05, "loss": 1.3554, "step": 7912 }, { "epoch": 0.0711687120288884, "grad_norm": 3.500373125076294, "learning_rate": 1.0021066849142552e-05, "loss": 2.1009, "step": 7913 }, { "epoch": 0.07117770592652885, "grad_norm": 2.3876397609710693, "learning_rate": 1.001505016380305e-05, "loss": 1.7407, "step": 7914 }, { "epoch": 0.0711866998241693, "grad_norm": 4.446011066436768, "learning_rate": 1.0009034832787168e-05, "loss": 0.8565, "step": 7915 }, { "epoch": 0.07119569372180976, "grad_norm": 3.1796340942382812, "learning_rate": 1.0003020856638576e-05, "loss": 1.4597, "step": 7916 }, { "epoch": 0.0712046876194502, "grad_norm": 20.496091842651367, "learning_rate": 9.997008235900798e-06, "loss": 0.4588, "step": 7917 }, { "epoch": 0.07121368151709065, "grad_norm": 2.174514055252075, "learning_rate": 9.990996971117264e-06, "loss": 2.5471, "step": 7918 }, { "epoch": 0.07122267541473111, "grad_norm": 2.2124266624450684, "learning_rate": 9.984987062831259e-06, "loss": 2.6372, "step": 7919 }, { "epoch": 0.07123166931237156, "grad_norm": 2.485452651977539, "learning_rate": 9.97897851158594e-06, "loss": 2.8249, "step": 7920 }, { "epoch": 0.071240663210012, "grad_norm": 2.23134708404541, "learning_rate": 9.972971317924374e-06, "loss": 1.9676, "step": 7921 }, { "epoch": 0.07124965710765246, "grad_norm": 14.58145809173584, "learning_rate": 9.966965482389474e-06, "loss": 1.1222, "step": 7922 }, { "epoch": 0.07125865100529291, "grad_norm": 2.502971887588501, "learning_rate": 9.960961005524034e-06, "loss": 2.4611, "step": 7923 }, { "epoch": 0.07126764490293336, "grad_norm": 5.517295837402344, "learning_rate": 9.954957887870745e-06, "loss": 1.1285, "step": 7924 }, { "epoch": 0.0712766388005738, "grad_norm": 18.651851654052734, "learning_rate": 9.94895612997216e-06, "loss": 1.754, "step": 7925 }, { "epoch": 0.07128563269821427, "grad_norm": 3.193704605102539, "learning_rate": 9.942955732370707e-06, "loss": 1.475, "step": 7926 }, { "epoch": 0.07129462659585471, "grad_norm": 2.1933178901672363, "learning_rate": 9.936956695608687e-06, "loss": 2.2234, "step": 7927 }, { "epoch": 0.07130362049349516, "grad_norm": 2.305403470993042, "learning_rate": 9.930959020228305e-06, "loss": 1.7993, "step": 7928 }, { "epoch": 0.07131261439113562, "grad_norm": 8.406499862670898, "learning_rate": 9.92496270677161e-06, "loss": 1.6506, "step": 7929 }, { "epoch": 0.07132160828877607, "grad_norm": 12.955229759216309, "learning_rate": 9.918967755780554e-06, "loss": 1.9779, "step": 7930 }, { "epoch": 0.07133060218641651, "grad_norm": 9.62811279296875, "learning_rate": 9.912974167796951e-06, "loss": 1.3762, "step": 7931 }, { "epoch": 0.07133959608405696, "grad_norm": 2.6418187618255615, "learning_rate": 9.90698194336248e-06, "loss": 1.8492, "step": 7932 }, { "epoch": 0.07134858998169742, "grad_norm": 2.3361380100250244, "learning_rate": 9.900991083018735e-06, "loss": 2.0652, "step": 7933 }, { "epoch": 0.07135758387933787, "grad_norm": 3.000971555709839, "learning_rate": 9.895001587307144e-06, "loss": 1.7139, "step": 7934 }, { "epoch": 0.07136657777697832, "grad_norm": 3.0679099559783936, "learning_rate": 9.889013456769048e-06, "loss": 1.548, "step": 7935 }, { "epoch": 0.07137557167461878, "grad_norm": 12.926019668579102, "learning_rate": 9.88302669194564e-06, "loss": 1.1626, "step": 7936 }, { "epoch": 0.07138456557225922, "grad_norm": 9.879298210144043, "learning_rate": 9.877041293377995e-06, "loss": 1.5752, "step": 7937 }, { "epoch": 0.07139355946989967, "grad_norm": 13.228421211242676, "learning_rate": 9.871057261607073e-06, "loss": 1.2836, "step": 7938 }, { "epoch": 0.07140255336754013, "grad_norm": 6.566876411437988, "learning_rate": 9.865074597173688e-06, "loss": 1.1693, "step": 7939 }, { "epoch": 0.07141154726518058, "grad_norm": 12.614394187927246, "learning_rate": 9.859093300618569e-06, "loss": 1.4641, "step": 7940 }, { "epoch": 0.07142054116282102, "grad_norm": 2.7377986907958984, "learning_rate": 9.853113372482279e-06, "loss": 1.8983, "step": 7941 }, { "epoch": 0.07142953506046147, "grad_norm": 2.669889450073242, "learning_rate": 9.847134813305295e-06, "loss": 2.1905, "step": 7942 }, { "epoch": 0.07143852895810193, "grad_norm": 2.7398736476898193, "learning_rate": 9.841157623627947e-06, "loss": 2.0236, "step": 7943 }, { "epoch": 0.07144752285574238, "grad_norm": 2.161689281463623, "learning_rate": 9.835181803990438e-06, "loss": 1.9723, "step": 7944 }, { "epoch": 0.07145651675338283, "grad_norm": 3.151482105255127, "learning_rate": 9.829207354932871e-06, "loss": 2.1679, "step": 7945 }, { "epoch": 0.07146551065102329, "grad_norm": 10.823966026306152, "learning_rate": 9.823234276995194e-06, "loss": 1.5303, "step": 7946 }, { "epoch": 0.07147450454866373, "grad_norm": 10.064932823181152, "learning_rate": 9.817262570717264e-06, "loss": 1.0721, "step": 7947 }, { "epoch": 0.07148349844630418, "grad_norm": 10.214683532714844, "learning_rate": 9.81129223663879e-06, "loss": 0.9079, "step": 7948 }, { "epoch": 0.07149249234394463, "grad_norm": 21.03414535522461, "learning_rate": 9.805323275299356e-06, "loss": 1.4947, "step": 7949 }, { "epoch": 0.07150148624158509, "grad_norm": 2.132843017578125, "learning_rate": 9.799355687238444e-06, "loss": 2.0734, "step": 7950 }, { "epoch": 0.07151048013922554, "grad_norm": 2.9797720909118652, "learning_rate": 9.793389472995393e-06, "loss": 1.6623, "step": 7951 }, { "epoch": 0.07151947403686598, "grad_norm": 6.121119499206543, "learning_rate": 9.787424633109424e-06, "loss": 1.0636, "step": 7952 }, { "epoch": 0.07152846793450644, "grad_norm": 2.177581787109375, "learning_rate": 9.78146116811962e-06, "loss": 2.5217, "step": 7953 }, { "epoch": 0.07153746183214689, "grad_norm": 14.867919921875, "learning_rate": 9.775499078564973e-06, "loss": 0.8312, "step": 7954 }, { "epoch": 0.07154645572978734, "grad_norm": 3.101828098297119, "learning_rate": 9.769538364984321e-06, "loss": 1.675, "step": 7955 }, { "epoch": 0.07155544962742778, "grad_norm": 3.1572229862213135, "learning_rate": 9.763579027916378e-06, "loss": 1.9474, "step": 7956 }, { "epoch": 0.07156444352506824, "grad_norm": 14.743952751159668, "learning_rate": 9.75762106789976e-06, "loss": 0.888, "step": 7957 }, { "epoch": 0.07157343742270869, "grad_norm": 3.1423094272613525, "learning_rate": 9.751664485472926e-06, "loss": 1.9415, "step": 7958 }, { "epoch": 0.07158243132034914, "grad_norm": 12.959692001342773, "learning_rate": 9.745709281174237e-06, "loss": 1.5161, "step": 7959 }, { "epoch": 0.0715914252179896, "grad_norm": 3.690518856048584, "learning_rate": 9.739755455541913e-06, "loss": 0.948, "step": 7960 }, { "epoch": 0.07160041911563005, "grad_norm": 11.736944198608398, "learning_rate": 9.733803009114045e-06, "loss": 1.2984, "step": 7961 }, { "epoch": 0.07160941301327049, "grad_norm": 8.099568367004395, "learning_rate": 9.727851942428628e-06, "loss": 0.9947, "step": 7962 }, { "epoch": 0.07161840691091095, "grad_norm": 3.0657191276550293, "learning_rate": 9.721902256023494e-06, "loss": 1.4847, "step": 7963 }, { "epoch": 0.0716274008085514, "grad_norm": 2.3718807697296143, "learning_rate": 9.715953950436384e-06, "loss": 2.0841, "step": 7964 }, { "epoch": 0.07163639470619185, "grad_norm": 12.403515815734863, "learning_rate": 9.710007026204895e-06, "loss": 1.3463, "step": 7965 }, { "epoch": 0.0716453886038323, "grad_norm": 4.307412147521973, "learning_rate": 9.704061483866503e-06, "loss": 1.4332, "step": 7966 }, { "epoch": 0.07165438250147275, "grad_norm": 3.314279556274414, "learning_rate": 9.698117323958555e-06, "loss": 1.9251, "step": 7967 }, { "epoch": 0.0716633763991132, "grad_norm": 3.286432981491089, "learning_rate": 9.692174547018276e-06, "loss": 1.9844, "step": 7968 }, { "epoch": 0.07167237029675365, "grad_norm": 3.5284271240234375, "learning_rate": 9.686233153582783e-06, "loss": 1.4009, "step": 7969 }, { "epoch": 0.07168136419439411, "grad_norm": 5.2899274826049805, "learning_rate": 9.680293144189031e-06, "loss": 1.2516, "step": 7970 }, { "epoch": 0.07169035809203456, "grad_norm": 6.068678379058838, "learning_rate": 9.674354519373894e-06, "loss": 1.0234, "step": 7971 }, { "epoch": 0.071699351989675, "grad_norm": 2.981302261352539, "learning_rate": 9.66841727967409e-06, "loss": 1.8216, "step": 7972 }, { "epoch": 0.07170834588731545, "grad_norm": 3.4765255451202393, "learning_rate": 9.662481425626205e-06, "loss": 1.0812, "step": 7973 }, { "epoch": 0.07171733978495591, "grad_norm": 2.442885398864746, "learning_rate": 9.65654695776674e-06, "loss": 2.7242, "step": 7974 }, { "epoch": 0.07172633368259636, "grad_norm": 3.0737783908843994, "learning_rate": 9.650613876632025e-06, "loss": 2.2615, "step": 7975 }, { "epoch": 0.0717353275802368, "grad_norm": 2.968156099319458, "learning_rate": 9.644682182758306e-06, "loss": 1.7107, "step": 7976 }, { "epoch": 0.07174432147787727, "grad_norm": 2.986185073852539, "learning_rate": 9.63875187668166e-06, "loss": 1.9781, "step": 7977 }, { "epoch": 0.07175331537551771, "grad_norm": 2.6641554832458496, "learning_rate": 9.632822958938084e-06, "loss": 1.2106, "step": 7978 }, { "epoch": 0.07176230927315816, "grad_norm": 2.7148959636688232, "learning_rate": 9.626895430063415e-06, "loss": 2.0079, "step": 7979 }, { "epoch": 0.07177130317079862, "grad_norm": 2.845229148864746, "learning_rate": 9.620969290593382e-06, "loss": 2.5636, "step": 7980 }, { "epoch": 0.07178029706843907, "grad_norm": 2.9812958240509033, "learning_rate": 9.615044541063575e-06, "loss": 1.5542, "step": 7981 }, { "epoch": 0.07178929096607951, "grad_norm": 2.369962215423584, "learning_rate": 9.609121182009467e-06, "loss": 2.2307, "step": 7982 }, { "epoch": 0.07179828486371996, "grad_norm": 6.120273113250732, "learning_rate": 9.603199213966415e-06, "loss": 1.4362, "step": 7983 }, { "epoch": 0.07180727876136042, "grad_norm": 3.041490316390991, "learning_rate": 9.597278637469634e-06, "loss": 2.6699, "step": 7984 }, { "epoch": 0.07181627265900087, "grad_norm": 11.987273216247559, "learning_rate": 9.59135945305421e-06, "loss": 1.1782, "step": 7985 }, { "epoch": 0.07182526655664132, "grad_norm": 13.781732559204102, "learning_rate": 9.585441661255132e-06, "loss": 1.035, "step": 7986 }, { "epoch": 0.07183426045428178, "grad_norm": 3.345792293548584, "learning_rate": 9.579525262607226e-06, "loss": 1.6372, "step": 7987 }, { "epoch": 0.07184325435192222, "grad_norm": 2.5252113342285156, "learning_rate": 9.573610257645222e-06, "loss": 2.6369, "step": 7988 }, { "epoch": 0.07185224824956267, "grad_norm": 2.547802448272705, "learning_rate": 9.567696646903701e-06, "loss": 2.6298, "step": 7989 }, { "epoch": 0.07186124214720312, "grad_norm": 17.655176162719727, "learning_rate": 9.561784430917143e-06, "loss": 1.0203, "step": 7990 }, { "epoch": 0.07187023604484358, "grad_norm": 2.6433029174804688, "learning_rate": 9.555873610219878e-06, "loss": 1.9431, "step": 7991 }, { "epoch": 0.07187922994248402, "grad_norm": 3.3626229763031006, "learning_rate": 9.549964185346125e-06, "loss": 1.9873, "step": 7992 }, { "epoch": 0.07188822384012447, "grad_norm": 2.226472854614258, "learning_rate": 9.544056156829965e-06, "loss": 2.1867, "step": 7993 }, { "epoch": 0.07189721773776493, "grad_norm": 7.985034465789795, "learning_rate": 9.53814952520535e-06, "loss": 1.0558, "step": 7994 }, { "epoch": 0.07190621163540538, "grad_norm": 14.843088150024414, "learning_rate": 9.53224429100614e-06, "loss": 1.533, "step": 7995 }, { "epoch": 0.07191520553304583, "grad_norm": 2.476980209350586, "learning_rate": 9.526340454766028e-06, "loss": 1.6, "step": 7996 }, { "epoch": 0.07192419943068627, "grad_norm": 4.3616132736206055, "learning_rate": 9.520438017018593e-06, "loss": 0.7875, "step": 7997 }, { "epoch": 0.07193319332832673, "grad_norm": 13.74548053741455, "learning_rate": 9.514536978297303e-06, "loss": 1.5939, "step": 7998 }, { "epoch": 0.07194218722596718, "grad_norm": 4.670053958892822, "learning_rate": 9.508637339135474e-06, "loss": 1.0497, "step": 7999 }, { "epoch": 0.07195118112360763, "grad_norm": 3.0543830394744873, "learning_rate": 9.502739100066324e-06, "loss": 2.0275, "step": 8000 }, { "epoch": 0.07196017502124809, "grad_norm": 2.1187851428985596, "learning_rate": 9.496842261622921e-06, "loss": 0.454, "step": 8001 }, { "epoch": 0.07196916891888853, "grad_norm": 13.957679748535156, "learning_rate": 9.490946824338209e-06, "loss": 0.6223, "step": 8002 }, { "epoch": 0.07197816281652898, "grad_norm": 3.4832582473754883, "learning_rate": 9.485052788745024e-06, "loss": 0.5375, "step": 8003 }, { "epoch": 0.07198715671416944, "grad_norm": 3.2531754970550537, "learning_rate": 9.479160155376052e-06, "loss": 0.6874, "step": 8004 }, { "epoch": 0.07199615061180989, "grad_norm": 2.774319887161255, "learning_rate": 9.473268924763872e-06, "loss": 0.3947, "step": 8005 }, { "epoch": 0.07200514450945034, "grad_norm": 2.585522174835205, "learning_rate": 9.467379097440926e-06, "loss": 0.5275, "step": 8006 }, { "epoch": 0.07201413840709078, "grad_norm": 2.0792791843414307, "learning_rate": 9.461490673939525e-06, "loss": 0.335, "step": 8007 }, { "epoch": 0.07202313230473124, "grad_norm": 13.13854694366455, "learning_rate": 9.455603654791862e-06, "loss": 0.4072, "step": 8008 }, { "epoch": 0.07203212620237169, "grad_norm": 2.2537503242492676, "learning_rate": 9.449718040529987e-06, "loss": 0.2508, "step": 8009 }, { "epoch": 0.07204112010001214, "grad_norm": 2.707529067993164, "learning_rate": 9.443833831685855e-06, "loss": 0.5591, "step": 8010 }, { "epoch": 0.0720501139976526, "grad_norm": 2.0195443630218506, "learning_rate": 9.437951028791256e-06, "loss": 0.332, "step": 8011 }, { "epoch": 0.07205910789529305, "grad_norm": 2.7832727432250977, "learning_rate": 9.432069632377891e-06, "loss": 0.4131, "step": 8012 }, { "epoch": 0.07206810179293349, "grad_norm": 2.055440902709961, "learning_rate": 9.426189642977304e-06, "loss": 0.2304, "step": 8013 }, { "epoch": 0.07207709569057394, "grad_norm": 4.295065402984619, "learning_rate": 9.420311061120915e-06, "loss": 0.0763, "step": 8014 }, { "epoch": 0.0720860895882144, "grad_norm": 2.4270107746124268, "learning_rate": 9.414433887340038e-06, "loss": 0.4094, "step": 8015 }, { "epoch": 0.07209508348585485, "grad_norm": 11.051129341125488, "learning_rate": 9.40855812216583e-06, "loss": 0.6662, "step": 8016 }, { "epoch": 0.0721040773834953, "grad_norm": 2.3169748783111572, "learning_rate": 9.402683766129355e-06, "loss": 0.1517, "step": 8017 }, { "epoch": 0.07211307128113575, "grad_norm": 1.9346528053283691, "learning_rate": 9.396810819761514e-06, "loss": 0.2844, "step": 8018 }, { "epoch": 0.0721220651787762, "grad_norm": 2.5651659965515137, "learning_rate": 9.39093928359311e-06, "loss": 0.5689, "step": 8019 }, { "epoch": 0.07213105907641665, "grad_norm": 2.9450926780700684, "learning_rate": 9.385069158154805e-06, "loss": 0.5228, "step": 8020 }, { "epoch": 0.07214005297405711, "grad_norm": 2.1852712631225586, "learning_rate": 9.379200443977127e-06, "loss": 0.3326, "step": 8021 }, { "epoch": 0.07214904687169756, "grad_norm": 2.960617780685425, "learning_rate": 9.373333141590487e-06, "loss": 0.5254, "step": 8022 }, { "epoch": 0.072158040769338, "grad_norm": 1.8727906942367554, "learning_rate": 9.36746725152516e-06, "loss": 0.5903, "step": 8023 }, { "epoch": 0.07216703466697845, "grad_norm": 10.550286293029785, "learning_rate": 9.361602774311312e-06, "loss": 0.749, "step": 8024 }, { "epoch": 0.07217602856461891, "grad_norm": 2.7190825939178467, "learning_rate": 9.355739710478962e-06, "loss": 0.8513, "step": 8025 }, { "epoch": 0.07218502246225936, "grad_norm": 2.014705181121826, "learning_rate": 9.349878060557999e-06, "loss": 0.1967, "step": 8026 }, { "epoch": 0.0721940163598998, "grad_norm": 7.790019989013672, "learning_rate": 9.344017825078204e-06, "loss": 0.559, "step": 8027 }, { "epoch": 0.07220301025754026, "grad_norm": 2.722329616546631, "learning_rate": 9.33815900456921e-06, "loss": 0.8572, "step": 8028 }, { "epoch": 0.07221200415518071, "grad_norm": 2.5518741607666016, "learning_rate": 9.332301599560542e-06, "loss": 0.4648, "step": 8029 }, { "epoch": 0.07222099805282116, "grad_norm": 2.4830873012542725, "learning_rate": 9.326445610581571e-06, "loss": 0.5978, "step": 8030 }, { "epoch": 0.0722299919504616, "grad_norm": 5.647598743438721, "learning_rate": 9.320591038161574e-06, "loss": 0.5063, "step": 8031 }, { "epoch": 0.07223898584810207, "grad_norm": 19.449995040893555, "learning_rate": 9.31473788282967e-06, "loss": 0.9213, "step": 8032 }, { "epoch": 0.07224797974574251, "grad_norm": 14.857604026794434, "learning_rate": 9.308886145114853e-06, "loss": 0.4738, "step": 8033 }, { "epoch": 0.07225697364338296, "grad_norm": 1.7895861864089966, "learning_rate": 9.303035825546013e-06, "loss": 0.2259, "step": 8034 }, { "epoch": 0.07226596754102342, "grad_norm": 2.3839521408081055, "learning_rate": 9.297186924651889e-06, "loss": 0.2758, "step": 8035 }, { "epoch": 0.07227496143866387, "grad_norm": 2.55733323097229, "learning_rate": 9.291339442961094e-06, "loss": 1.023, "step": 8036 }, { "epoch": 0.07228395533630431, "grad_norm": 1.909386396408081, "learning_rate": 9.285493381002122e-06, "loss": 0.3676, "step": 8037 }, { "epoch": 0.07229294923394476, "grad_norm": 3.216871738433838, "learning_rate": 9.279648739303322e-06, "loss": 0.5668, "step": 8038 }, { "epoch": 0.07230194313158522, "grad_norm": 3.226766347885132, "learning_rate": 9.273805518392947e-06, "loss": 0.5118, "step": 8039 }, { "epoch": 0.07231093702922567, "grad_norm": 24.29452133178711, "learning_rate": 9.267963718799083e-06, "loss": 0.6776, "step": 8040 }, { "epoch": 0.07231993092686612, "grad_norm": 2.252685308456421, "learning_rate": 9.262123341049719e-06, "loss": 0.2178, "step": 8041 }, { "epoch": 0.07232892482450658, "grad_norm": 3.832818031311035, "learning_rate": 9.25628438567269e-06, "loss": 0.5979, "step": 8042 }, { "epoch": 0.07233791872214702, "grad_norm": 1.6380866765975952, "learning_rate": 9.250446853195726e-06, "loss": 0.3072, "step": 8043 }, { "epoch": 0.07234691261978747, "grad_norm": 2.3324036598205566, "learning_rate": 9.244610744146412e-06, "loss": 0.7017, "step": 8044 }, { "epoch": 0.07235590651742793, "grad_norm": 1.7288490533828735, "learning_rate": 9.238776059052201e-06, "loss": 0.41, "step": 8045 }, { "epoch": 0.07236490041506838, "grad_norm": 1.4399373531341553, "learning_rate": 9.232942798440442e-06, "loss": 0.5992, "step": 8046 }, { "epoch": 0.07237389431270883, "grad_norm": 4.845678329467773, "learning_rate": 9.227110962838329e-06, "loss": 0.4615, "step": 8047 }, { "epoch": 0.07238288821034927, "grad_norm": 15.50008487701416, "learning_rate": 9.221280552772928e-06, "loss": 0.434, "step": 8048 }, { "epoch": 0.07239188210798973, "grad_norm": 2.1923787593841553, "learning_rate": 9.215451568771207e-06, "loss": 0.1333, "step": 8049 }, { "epoch": 0.07240087600563018, "grad_norm": 8.612996101379395, "learning_rate": 9.209624011359972e-06, "loss": 0.7673, "step": 8050 }, { "epoch": 0.07240986990327063, "grad_norm": 3.687276840209961, "learning_rate": 9.203797881065906e-06, "loss": 0.1762, "step": 8051 }, { "epoch": 0.07241886380091109, "grad_norm": 2.19718599319458, "learning_rate": 9.19797317841557e-06, "loss": 0.2962, "step": 8052 }, { "epoch": 0.07242785769855153, "grad_norm": 3.593055486679077, "learning_rate": 9.192149903935405e-06, "loss": 0.1626, "step": 8053 }, { "epoch": 0.07243685159619198, "grad_norm": 16.1307430267334, "learning_rate": 9.186328058151694e-06, "loss": 0.4344, "step": 8054 }, { "epoch": 0.07244584549383243, "grad_norm": 2.944232702255249, "learning_rate": 9.18050764159063e-06, "loss": 0.1076, "step": 8055 }, { "epoch": 0.07245483939147289, "grad_norm": 1.7776223421096802, "learning_rate": 9.174688654778243e-06, "loss": 0.3008, "step": 8056 }, { "epoch": 0.07246383328911334, "grad_norm": 25.719097137451172, "learning_rate": 9.168871098240445e-06, "loss": 0.6065, "step": 8057 }, { "epoch": 0.07247282718675378, "grad_norm": 15.278233528137207, "learning_rate": 9.163054972503033e-06, "loss": 0.4196, "step": 8058 }, { "epoch": 0.07248182108439424, "grad_norm": 2.2293994426727295, "learning_rate": 9.157240278091647e-06, "loss": 0.4416, "step": 8059 }, { "epoch": 0.07249081498203469, "grad_norm": 10.67837142944336, "learning_rate": 9.151427015531827e-06, "loss": 0.4574, "step": 8060 }, { "epoch": 0.07249980887967514, "grad_norm": 4.156530380249023, "learning_rate": 9.145615185348963e-06, "loss": 0.3634, "step": 8061 }, { "epoch": 0.0725088027773156, "grad_norm": 1.9457863569259644, "learning_rate": 9.13980478806832e-06, "loss": 0.5745, "step": 8062 }, { "epoch": 0.07251779667495604, "grad_norm": 2.6250786781311035, "learning_rate": 9.13399582421504e-06, "loss": 0.5149, "step": 8063 }, { "epoch": 0.07252679057259649, "grad_norm": 3.796468496322632, "learning_rate": 9.128188294314119e-06, "loss": 0.304, "step": 8064 }, { "epoch": 0.07253578447023694, "grad_norm": 6.524633884429932, "learning_rate": 9.12238219889045e-06, "loss": 0.5726, "step": 8065 }, { "epoch": 0.0725447783678774, "grad_norm": 15.931842803955078, "learning_rate": 9.11657753846877e-06, "loss": 0.67, "step": 8066 }, { "epoch": 0.07255377226551785, "grad_norm": 1.8539962768554688, "learning_rate": 9.110774313573711e-06, "loss": 0.3787, "step": 8067 }, { "epoch": 0.0725627661631583, "grad_norm": 2.1245322227478027, "learning_rate": 9.104972524729757e-06, "loss": 0.338, "step": 8068 }, { "epoch": 0.07257176006079875, "grad_norm": 2.618295669555664, "learning_rate": 9.099172172461257e-06, "loss": 0.501, "step": 8069 }, { "epoch": 0.0725807539584392, "grad_norm": 2.2690818309783936, "learning_rate": 9.093373257292456e-06, "loss": 0.6485, "step": 8070 }, { "epoch": 0.07258974785607965, "grad_norm": 2.862367630004883, "learning_rate": 9.087575779747437e-06, "loss": 0.2202, "step": 8071 }, { "epoch": 0.0725987417537201, "grad_norm": 1.9833537340164185, "learning_rate": 9.08177974035019e-06, "loss": 0.4646, "step": 8072 }, { "epoch": 0.07260773565136056, "grad_norm": 12.641139030456543, "learning_rate": 9.075985139624547e-06, "loss": 0.4859, "step": 8073 }, { "epoch": 0.072616729549001, "grad_norm": 2.418081045150757, "learning_rate": 9.0701919780942e-06, "loss": 0.2273, "step": 8074 }, { "epoch": 0.07262572344664145, "grad_norm": 3.1736831665039062, "learning_rate": 9.064400256282757e-06, "loss": 0.3873, "step": 8075 }, { "epoch": 0.07263471734428191, "grad_norm": 2.8360769748687744, "learning_rate": 9.058609974713655e-06, "loss": 0.2435, "step": 8076 }, { "epoch": 0.07264371124192236, "grad_norm": 2.452237606048584, "learning_rate": 9.052821133910212e-06, "loss": 0.8012, "step": 8077 }, { "epoch": 0.0726527051395628, "grad_norm": 11.075590133666992, "learning_rate": 9.047033734395616e-06, "loss": 0.7165, "step": 8078 }, { "epoch": 0.07266169903720326, "grad_norm": 2.9808032512664795, "learning_rate": 9.041247776692922e-06, "loss": 0.4822, "step": 8079 }, { "epoch": 0.07267069293484371, "grad_norm": 9.114850044250488, "learning_rate": 9.035463261325069e-06, "loss": 0.8862, "step": 8080 }, { "epoch": 0.07267968683248416, "grad_norm": 14.733414649963379, "learning_rate": 9.029680188814846e-06, "loss": 0.9033, "step": 8081 }, { "epoch": 0.0726886807301246, "grad_norm": 1.7730376720428467, "learning_rate": 9.023898559684934e-06, "loss": 0.5352, "step": 8082 }, { "epoch": 0.07269767462776507, "grad_norm": 2.7182602882385254, "learning_rate": 9.018118374457854e-06, "loss": 0.4223, "step": 8083 }, { "epoch": 0.07270666852540551, "grad_norm": 9.048175811767578, "learning_rate": 9.012339633656027e-06, "loss": 0.9464, "step": 8084 }, { "epoch": 0.07271566242304596, "grad_norm": 2.6334028244018555, "learning_rate": 9.006562337801727e-06, "loss": 0.3834, "step": 8085 }, { "epoch": 0.07272465632068642, "grad_norm": 2.50465989112854, "learning_rate": 9.000786487417085e-06, "loss": 0.1604, "step": 8086 }, { "epoch": 0.07273365021832687, "grad_norm": 7.97613525390625, "learning_rate": 8.995012083024137e-06, "loss": 0.3846, "step": 8087 }, { "epoch": 0.07274264411596731, "grad_norm": 16.955997467041016, "learning_rate": 8.98923912514475e-06, "loss": 0.5748, "step": 8088 }, { "epoch": 0.07275163801360776, "grad_norm": 2.0455105304718018, "learning_rate": 8.983467614300692e-06, "loss": 0.4961, "step": 8089 }, { "epoch": 0.07276063191124822, "grad_norm": 3.1607725620269775, "learning_rate": 8.977697551013581e-06, "loss": 0.521, "step": 8090 }, { "epoch": 0.07276962580888867, "grad_norm": 10.927391052246094, "learning_rate": 8.971928935804907e-06, "loss": 0.6121, "step": 8091 }, { "epoch": 0.07277861970652912, "grad_norm": 2.031308650970459, "learning_rate": 8.966161769196033e-06, "loss": 0.2144, "step": 8092 }, { "epoch": 0.07278761360416958, "grad_norm": 2.3426878452301025, "learning_rate": 8.960396051708175e-06, "loss": 0.2216, "step": 8093 }, { "epoch": 0.07279660750181002, "grad_norm": 3.224581480026245, "learning_rate": 8.954631783862457e-06, "loss": 0.0765, "step": 8094 }, { "epoch": 0.07280560139945047, "grad_norm": 2.1579346656799316, "learning_rate": 8.94886896617983e-06, "loss": 0.1952, "step": 8095 }, { "epoch": 0.07281459529709092, "grad_norm": 2.102949380874634, "learning_rate": 8.94310759918114e-06, "loss": 0.2873, "step": 8096 }, { "epoch": 0.07282358919473138, "grad_norm": 3.2937419414520264, "learning_rate": 8.937347683387095e-06, "loss": 0.3459, "step": 8097 }, { "epoch": 0.07283258309237182, "grad_norm": 13.320281028747559, "learning_rate": 8.931589219318254e-06, "loss": 0.5377, "step": 8098 }, { "epoch": 0.07284157699001227, "grad_norm": 1.8870320320129395, "learning_rate": 8.925832207495082e-06, "loss": 0.2679, "step": 8099 }, { "epoch": 0.07285057088765273, "grad_norm": 2.731064796447754, "learning_rate": 8.920076648437872e-06, "loss": 0.7507, "step": 8100 }, { "epoch": 0.07285956478529318, "grad_norm": 2.2182765007019043, "learning_rate": 8.914322542666822e-06, "loss": 0.4478, "step": 8101 }, { "epoch": 0.07286855868293363, "grad_norm": 2.892962694168091, "learning_rate": 8.908569890701976e-06, "loss": 0.293, "step": 8102 }, { "epoch": 0.07287755258057409, "grad_norm": 2.413796901702881, "learning_rate": 8.902818693063242e-06, "loss": 0.3885, "step": 8103 }, { "epoch": 0.07288654647821453, "grad_norm": 1.9481632709503174, "learning_rate": 8.897068950270426e-06, "loss": 0.2711, "step": 8104 }, { "epoch": 0.07289554037585498, "grad_norm": 1.9487875699996948, "learning_rate": 8.891320662843172e-06, "loss": 0.337, "step": 8105 }, { "epoch": 0.07290453427349543, "grad_norm": 2.6892054080963135, "learning_rate": 8.885573831301006e-06, "loss": 0.5046, "step": 8106 }, { "epoch": 0.07291352817113589, "grad_norm": 2.1422605514526367, "learning_rate": 8.879828456163311e-06, "loss": 0.1742, "step": 8107 }, { "epoch": 0.07292252206877634, "grad_norm": 13.568390846252441, "learning_rate": 8.874084537949364e-06, "loss": 0.6681, "step": 8108 }, { "epoch": 0.07293151596641678, "grad_norm": 3.1841228008270264, "learning_rate": 8.86834207717829e-06, "loss": 0.5104, "step": 8109 }, { "epoch": 0.07294050986405724, "grad_norm": 2.9196670055389404, "learning_rate": 8.862601074369076e-06, "loss": 0.6328, "step": 8110 }, { "epoch": 0.07294950376169769, "grad_norm": 2.317739486694336, "learning_rate": 8.8568615300406e-06, "loss": 0.3178, "step": 8111 }, { "epoch": 0.07295849765933814, "grad_norm": 2.0369739532470703, "learning_rate": 8.851123444711585e-06, "loss": 0.3089, "step": 8112 }, { "epoch": 0.07296749155697858, "grad_norm": 19.293964385986328, "learning_rate": 8.845386818900647e-06, "loss": 0.9018, "step": 8113 }, { "epoch": 0.07297648545461904, "grad_norm": 2.9696309566497803, "learning_rate": 8.839651653126247e-06, "loss": 0.1993, "step": 8114 }, { "epoch": 0.07298547935225949, "grad_norm": 2.4831736087799072, "learning_rate": 8.833917947906718e-06, "loss": 0.0682, "step": 8115 }, { "epoch": 0.07299447324989994, "grad_norm": 2.5887749195098877, "learning_rate": 8.828185703760277e-06, "loss": 0.1962, "step": 8116 }, { "epoch": 0.0730034671475404, "grad_norm": 2.4396655559539795, "learning_rate": 8.822454921204995e-06, "loss": 0.5605, "step": 8117 }, { "epoch": 0.07301246104518085, "grad_norm": 2.5267703533172607, "learning_rate": 8.816725600758812e-06, "loss": 0.2594, "step": 8118 }, { "epoch": 0.07302145494282129, "grad_norm": 2.5800461769104004, "learning_rate": 8.810997742939531e-06, "loss": 0.5748, "step": 8119 }, { "epoch": 0.07303044884046175, "grad_norm": 13.075410842895508, "learning_rate": 8.805271348264843e-06, "loss": 0.6296, "step": 8120 }, { "epoch": 0.0730394427381022, "grad_norm": 17.428457260131836, "learning_rate": 8.799546417252289e-06, "loss": 0.3502, "step": 8121 }, { "epoch": 0.07304843663574265, "grad_norm": 12.9984769821167, "learning_rate": 8.79382295041927e-06, "loss": 0.9058, "step": 8122 }, { "epoch": 0.0730574305333831, "grad_norm": 2.358675956726074, "learning_rate": 8.788100948283085e-06, "loss": 0.5076, "step": 8123 }, { "epoch": 0.07306642443102355, "grad_norm": 1.788003921508789, "learning_rate": 8.782380411360866e-06, "loss": 0.2669, "step": 8124 }, { "epoch": 0.073075418328664, "grad_norm": 1.932532548904419, "learning_rate": 8.776661340169648e-06, "loss": 0.3612, "step": 8125 }, { "epoch": 0.07308441222630445, "grad_norm": 1.6108349561691284, "learning_rate": 8.770943735226303e-06, "loss": 0.0711, "step": 8126 }, { "epoch": 0.07309340612394491, "grad_norm": 21.182085037231445, "learning_rate": 8.765227597047574e-06, "loss": 0.4959, "step": 8127 }, { "epoch": 0.07310240002158536, "grad_norm": 25.964664459228516, "learning_rate": 8.759512926150093e-06, "loss": 0.4352, "step": 8128 }, { "epoch": 0.0731113939192258, "grad_norm": 2.1910901069641113, "learning_rate": 8.753799723050337e-06, "loss": 0.4479, "step": 8129 }, { "epoch": 0.07312038781686625, "grad_norm": 1.7647202014923096, "learning_rate": 8.74808798826467e-06, "loss": 0.566, "step": 8130 }, { "epoch": 0.07312938171450671, "grad_norm": 6.087286949157715, "learning_rate": 8.742377722309306e-06, "loss": 0.7981, "step": 8131 }, { "epoch": 0.07313837561214716, "grad_norm": 3.0404930114746094, "learning_rate": 8.73666892570033e-06, "loss": 0.5499, "step": 8132 }, { "epoch": 0.0731473695097876, "grad_norm": 3.086827516555786, "learning_rate": 8.730961598953702e-06, "loss": 0.4677, "step": 8133 }, { "epoch": 0.07315636340742807, "grad_norm": 8.759130477905273, "learning_rate": 8.725255742585231e-06, "loss": 0.3992, "step": 8134 }, { "epoch": 0.07316535730506851, "grad_norm": 2.961829900741577, "learning_rate": 8.719551357110625e-06, "loss": 0.3273, "step": 8135 }, { "epoch": 0.07317435120270896, "grad_norm": 22.77712631225586, "learning_rate": 8.713848443045426e-06, "loss": 0.4166, "step": 8136 }, { "epoch": 0.0731833451003494, "grad_norm": 13.581802368164062, "learning_rate": 8.708147000905064e-06, "loss": 0.3732, "step": 8137 }, { "epoch": 0.07319233899798987, "grad_norm": 1.8044394254684448, "learning_rate": 8.702447031204833e-06, "loss": 0.197, "step": 8138 }, { "epoch": 0.07320133289563031, "grad_norm": 2.89473295211792, "learning_rate": 8.696748534459879e-06, "loss": 0.2864, "step": 8139 }, { "epoch": 0.07321032679327076, "grad_norm": 2.626605272293091, "learning_rate": 8.691051511185236e-06, "loss": 0.1703, "step": 8140 }, { "epoch": 0.07321932069091122, "grad_norm": 2.979314088821411, "learning_rate": 8.685355961895784e-06, "loss": 0.3027, "step": 8141 }, { "epoch": 0.07322831458855167, "grad_norm": 2.140768527984619, "learning_rate": 8.679661887106294e-06, "loss": 0.2605, "step": 8142 }, { "epoch": 0.07323730848619212, "grad_norm": 2.3162736892700195, "learning_rate": 8.673969287331388e-06, "loss": 0.5828, "step": 8143 }, { "epoch": 0.07324630238383258, "grad_norm": 1.833219289779663, "learning_rate": 8.668278163085541e-06, "loss": 0.0341, "step": 8144 }, { "epoch": 0.07325529628147302, "grad_norm": 3.101125717163086, "learning_rate": 8.662588514883132e-06, "loss": 0.0848, "step": 8145 }, { "epoch": 0.07326429017911347, "grad_norm": 2.584658145904541, "learning_rate": 8.656900343238376e-06, "loss": 0.5329, "step": 8146 }, { "epoch": 0.07327328407675392, "grad_norm": 2.4890832901000977, "learning_rate": 8.651213648665363e-06, "loss": 0.3863, "step": 8147 }, { "epoch": 0.07328227797439438, "grad_norm": 2.2825021743774414, "learning_rate": 8.645528431678043e-06, "loss": 0.4234, "step": 8148 }, { "epoch": 0.07329127187203482, "grad_norm": 8.422805786132812, "learning_rate": 8.639844692790255e-06, "loss": 0.7653, "step": 8149 }, { "epoch": 0.07330026576967527, "grad_norm": 12.653806686401367, "learning_rate": 8.634162432515682e-06, "loss": 0.6682, "step": 8150 }, { "epoch": 0.07330925966731573, "grad_norm": 1.9887844324111938, "learning_rate": 8.628481651367876e-06, "loss": 0.7028, "step": 8151 }, { "epoch": 0.07331825356495618, "grad_norm": 11.435428619384766, "learning_rate": 8.622802349860268e-06, "loss": 0.6332, "step": 8152 }, { "epoch": 0.07332724746259663, "grad_norm": 1.7186335325241089, "learning_rate": 8.617124528506138e-06, "loss": 0.2213, "step": 8153 }, { "epoch": 0.07333624136023707, "grad_norm": 11.095010757446289, "learning_rate": 8.611448187818655e-06, "loss": 0.3842, "step": 8154 }, { "epoch": 0.07334523525787753, "grad_norm": 7.284315586090088, "learning_rate": 8.605773328310832e-06, "loss": 0.6077, "step": 8155 }, { "epoch": 0.07335422915551798, "grad_norm": 2.049041271209717, "learning_rate": 8.60009995049555e-06, "loss": 0.3692, "step": 8156 }, { "epoch": 0.07336322305315843, "grad_norm": 3.6013524532318115, "learning_rate": 8.594428054885575e-06, "loss": 0.4696, "step": 8157 }, { "epoch": 0.07337221695079889, "grad_norm": 3.329341411590576, "learning_rate": 8.588757641993517e-06, "loss": 0.4764, "step": 8158 }, { "epoch": 0.07338121084843933, "grad_norm": 2.105095624923706, "learning_rate": 8.583088712331872e-06, "loss": 0.567, "step": 8159 }, { "epoch": 0.07339020474607978, "grad_norm": 1.866402268409729, "learning_rate": 8.577421266412988e-06, "loss": 0.051, "step": 8160 }, { "epoch": 0.07339919864372024, "grad_norm": 1.7433239221572876, "learning_rate": 8.57175530474908e-06, "loss": 0.3884, "step": 8161 }, { "epoch": 0.07340819254136069, "grad_norm": 11.74752426147461, "learning_rate": 8.566090827852234e-06, "loss": 0.6209, "step": 8162 }, { "epoch": 0.07341718643900114, "grad_norm": 2.7486960887908936, "learning_rate": 8.56042783623439e-06, "loss": 0.4342, "step": 8163 }, { "epoch": 0.07342618033664158, "grad_norm": 17.22389793395996, "learning_rate": 8.554766330407377e-06, "loss": 0.5053, "step": 8164 }, { "epoch": 0.07343517423428204, "grad_norm": 2.356005907058716, "learning_rate": 8.549106310882863e-06, "loss": 0.498, "step": 8165 }, { "epoch": 0.07344416813192249, "grad_norm": 13.029666900634766, "learning_rate": 8.543447778172408e-06, "loss": 0.3976, "step": 8166 }, { "epoch": 0.07345316202956294, "grad_norm": 11.154169082641602, "learning_rate": 8.537790732787418e-06, "loss": 0.379, "step": 8167 }, { "epoch": 0.0734621559272034, "grad_norm": 2.0181055068969727, "learning_rate": 8.532135175239162e-06, "loss": 0.5208, "step": 8168 }, { "epoch": 0.07347114982484385, "grad_norm": 4.095487594604492, "learning_rate": 8.5264811060388e-06, "loss": 0.4816, "step": 8169 }, { "epoch": 0.07348014372248429, "grad_norm": 3.6038312911987305, "learning_rate": 8.520828525697322e-06, "loss": 0.2325, "step": 8170 }, { "epoch": 0.07348913762012474, "grad_norm": 2.7021899223327637, "learning_rate": 8.515177434725621e-06, "loss": 0.3883, "step": 8171 }, { "epoch": 0.0734981315177652, "grad_norm": 2.9641387462615967, "learning_rate": 8.509527833634429e-06, "loss": 0.3311, "step": 8172 }, { "epoch": 0.07350712541540565, "grad_norm": 2.340041399002075, "learning_rate": 8.503879722934346e-06, "loss": 0.4311, "step": 8173 }, { "epoch": 0.0735161193130461, "grad_norm": 2.863030195236206, "learning_rate": 8.49823310313584e-06, "loss": 0.3401, "step": 8174 }, { "epoch": 0.07352511321068655, "grad_norm": 2.339021682739258, "learning_rate": 8.492587974749258e-06, "loss": 0.2525, "step": 8175 }, { "epoch": 0.073534107108327, "grad_norm": 4.319813251495361, "learning_rate": 8.486944338284797e-06, "loss": 0.5524, "step": 8176 }, { "epoch": 0.07354310100596745, "grad_norm": 2.1442811489105225, "learning_rate": 8.48130219425251e-06, "loss": 0.4619, "step": 8177 }, { "epoch": 0.0735520949036079, "grad_norm": 2.1957919597625732, "learning_rate": 8.475661543162347e-06, "loss": 0.2042, "step": 8178 }, { "epoch": 0.07356108880124836, "grad_norm": 2.9848668575286865, "learning_rate": 8.470022385524095e-06, "loss": 0.6675, "step": 8179 }, { "epoch": 0.0735700826988888, "grad_norm": 2.573397159576416, "learning_rate": 8.464384721847406e-06, "loss": 0.4833, "step": 8180 }, { "epoch": 0.07357907659652925, "grad_norm": 2.296194076538086, "learning_rate": 8.458748552641826e-06, "loss": 0.2712, "step": 8181 }, { "epoch": 0.07358807049416971, "grad_norm": 2.693282127380371, "learning_rate": 8.453113878416727e-06, "loss": 0.1881, "step": 8182 }, { "epoch": 0.07359706439181016, "grad_norm": 1.7272191047668457, "learning_rate": 8.447480699681378e-06, "loss": 0.32, "step": 8183 }, { "epoch": 0.0736060582894506, "grad_norm": 1.9834359884262085, "learning_rate": 8.441849016944889e-06, "loss": 0.4492, "step": 8184 }, { "epoch": 0.07361505218709107, "grad_norm": 2.2247140407562256, "learning_rate": 8.436218830716258e-06, "loss": 0.2838, "step": 8185 }, { "epoch": 0.07362404608473151, "grad_norm": 3.182877540588379, "learning_rate": 8.430590141504328e-06, "loss": 0.3025, "step": 8186 }, { "epoch": 0.07363303998237196, "grad_norm": 1.7847747802734375, "learning_rate": 8.424962949817813e-06, "loss": 0.703, "step": 8187 }, { "epoch": 0.0736420338800124, "grad_norm": 1.8570622205734253, "learning_rate": 8.419337256165295e-06, "loss": 0.6767, "step": 8188 }, { "epoch": 0.07365102777765287, "grad_norm": 3.5622217655181885, "learning_rate": 8.413713061055206e-06, "loss": 0.5766, "step": 8189 }, { "epoch": 0.07366002167529331, "grad_norm": 1.8237643241882324, "learning_rate": 8.408090364995876e-06, "loss": 0.0377, "step": 8190 }, { "epoch": 0.07366901557293376, "grad_norm": 2.0029048919677734, "learning_rate": 8.402469168495466e-06, "loss": 0.1667, "step": 8191 }, { "epoch": 0.07367800947057422, "grad_norm": 12.418427467346191, "learning_rate": 8.396849472062007e-06, "loss": 0.3854, "step": 8192 }, { "epoch": 0.07368700336821467, "grad_norm": 25.209667205810547, "learning_rate": 8.39123127620342e-06, "loss": 0.3705, "step": 8193 }, { "epoch": 0.07369599726585511, "grad_norm": 2.574199676513672, "learning_rate": 8.38561458142745e-06, "loss": 0.3998, "step": 8194 }, { "epoch": 0.07370499116349556, "grad_norm": 2.9913346767425537, "learning_rate": 8.379999388241746e-06, "loss": 0.3865, "step": 8195 }, { "epoch": 0.07371398506113602, "grad_norm": 3.187852621078491, "learning_rate": 8.374385697153792e-06, "loss": 0.4333, "step": 8196 }, { "epoch": 0.07372297895877647, "grad_norm": 24.58479881286621, "learning_rate": 8.368773508670954e-06, "loss": 0.7223, "step": 8197 }, { "epoch": 0.07373197285641692, "grad_norm": 2.3335509300231934, "learning_rate": 8.363162823300455e-06, "loss": 0.4224, "step": 8198 }, { "epoch": 0.07374096675405738, "grad_norm": 2.6524505615234375, "learning_rate": 8.357553641549374e-06, "loss": 0.7456, "step": 8199 }, { "epoch": 0.07374996065169782, "grad_norm": 2.0766255855560303, "learning_rate": 8.351945963924673e-06, "loss": 0.1617, "step": 8200 }, { "epoch": 0.07375895454933827, "grad_norm": 2.2182605266571045, "learning_rate": 8.346339790933166e-06, "loss": 0.4109, "step": 8201 }, { "epoch": 0.07376794844697873, "grad_norm": 17.99481201171875, "learning_rate": 8.340735123081531e-06, "loss": 0.6135, "step": 8202 }, { "epoch": 0.07377694234461918, "grad_norm": 2.68619966506958, "learning_rate": 8.335131960876314e-06, "loss": 0.6139, "step": 8203 }, { "epoch": 0.07378593624225963, "grad_norm": 2.500063896179199, "learning_rate": 8.329530304823908e-06, "loss": 0.3654, "step": 8204 }, { "epoch": 0.07379493013990007, "grad_norm": 12.428422927856445, "learning_rate": 8.323930155430607e-06, "loss": 0.3897, "step": 8205 }, { "epoch": 0.07380392403754053, "grad_norm": 3.5907578468322754, "learning_rate": 8.318331513202529e-06, "loss": 0.6048, "step": 8206 }, { "epoch": 0.07381291793518098, "grad_norm": 7.951860427856445, "learning_rate": 8.31273437864569e-06, "loss": 0.42, "step": 8207 }, { "epoch": 0.07382191183282143, "grad_norm": 2.374204397201538, "learning_rate": 8.307138752265933e-06, "loss": 0.3764, "step": 8208 }, { "epoch": 0.07383090573046189, "grad_norm": 1.8350147008895874, "learning_rate": 8.301544634569005e-06, "loss": 0.2126, "step": 8209 }, { "epoch": 0.07383989962810233, "grad_norm": 2.500581979751587, "learning_rate": 8.295952026060485e-06, "loss": 0.535, "step": 8210 }, { "epoch": 0.07384889352574278, "grad_norm": 1.7098686695098877, "learning_rate": 8.290360927245824e-06, "loss": 0.2691, "step": 8211 }, { "epoch": 0.07385788742338323, "grad_norm": 3.109178066253662, "learning_rate": 8.284771338630352e-06, "loss": 0.566, "step": 8212 }, { "epoch": 0.07386688132102369, "grad_norm": 11.9787015914917, "learning_rate": 8.27918326071923e-06, "loss": 0.7556, "step": 8213 }, { "epoch": 0.07387587521866414, "grad_norm": 11.597487449645996, "learning_rate": 8.273596694017527e-06, "loss": 0.5539, "step": 8214 }, { "epoch": 0.07388486911630458, "grad_norm": 2.208312749862671, "learning_rate": 8.268011639030137e-06, "loss": 0.7125, "step": 8215 }, { "epoch": 0.07389386301394504, "grad_norm": 20.2325439453125, "learning_rate": 8.262428096261834e-06, "loss": 0.3769, "step": 8216 }, { "epoch": 0.07390285691158549, "grad_norm": 2.7790095806121826, "learning_rate": 8.256846066217249e-06, "loss": 0.6586, "step": 8217 }, { "epoch": 0.07391185080922594, "grad_norm": 18.069026947021484, "learning_rate": 8.251265549400877e-06, "loss": 0.4304, "step": 8218 }, { "epoch": 0.07392084470686638, "grad_norm": 26.501508712768555, "learning_rate": 8.24568654631709e-06, "loss": 0.3142, "step": 8219 }, { "epoch": 0.07392983860450685, "grad_norm": 31.926240921020508, "learning_rate": 8.24010905747011e-06, "loss": 0.8899, "step": 8220 }, { "epoch": 0.07393883250214729, "grad_norm": 2.0915513038635254, "learning_rate": 8.23453308336401e-06, "loss": 0.2811, "step": 8221 }, { "epoch": 0.07394782639978774, "grad_norm": 3.056089401245117, "learning_rate": 8.22895862450276e-06, "loss": 0.0776, "step": 8222 }, { "epoch": 0.0739568202974282, "grad_norm": 1.7837276458740234, "learning_rate": 8.223385681390158e-06, "loss": 0.5893, "step": 8223 }, { "epoch": 0.07396581419506865, "grad_norm": 14.064417839050293, "learning_rate": 8.217814254529898e-06, "loss": 0.4237, "step": 8224 }, { "epoch": 0.0739748080927091, "grad_norm": 2.103680372238159, "learning_rate": 8.212244344425498e-06, "loss": 0.3834, "step": 8225 }, { "epoch": 0.07398380199034955, "grad_norm": 2.6024959087371826, "learning_rate": 8.206675951580381e-06, "loss": 0.3768, "step": 8226 }, { "epoch": 0.07399279588799, "grad_norm": 10.737861633300781, "learning_rate": 8.201109076497803e-06, "loss": 0.5209, "step": 8227 }, { "epoch": 0.07400178978563045, "grad_norm": 2.3815054893493652, "learning_rate": 8.195543719680884e-06, "loss": 0.3885, "step": 8228 }, { "epoch": 0.0740107836832709, "grad_norm": 16.65320587158203, "learning_rate": 8.189979881632634e-06, "loss": 0.3482, "step": 8229 }, { "epoch": 0.07401977758091136, "grad_norm": 8.860481262207031, "learning_rate": 8.184417562855895e-06, "loss": 0.5502, "step": 8230 }, { "epoch": 0.0740287714785518, "grad_norm": 2.71755051612854, "learning_rate": 8.178856763853382e-06, "loss": 0.3787, "step": 8231 }, { "epoch": 0.07403776537619225, "grad_norm": 8.761086463928223, "learning_rate": 8.173297485127679e-06, "loss": 0.6483, "step": 8232 }, { "epoch": 0.07404675927383271, "grad_norm": 13.971957206726074, "learning_rate": 8.167739727181214e-06, "loss": 0.4524, "step": 8233 }, { "epoch": 0.07405575317147316, "grad_norm": 2.302755832672119, "learning_rate": 8.162183490516315e-06, "loss": 0.799, "step": 8234 }, { "epoch": 0.0740647470691136, "grad_norm": 2.110431671142578, "learning_rate": 8.156628775635123e-06, "loss": 0.3438, "step": 8235 }, { "epoch": 0.07407374096675405, "grad_norm": 2.00355863571167, "learning_rate": 8.151075583039692e-06, "loss": 0.3751, "step": 8236 }, { "epoch": 0.07408273486439451, "grad_norm": 15.68254280090332, "learning_rate": 8.145523913231889e-06, "loss": 0.7476, "step": 8237 }, { "epoch": 0.07409172876203496, "grad_norm": 3.4384729862213135, "learning_rate": 8.139973766713493e-06, "loss": 0.3167, "step": 8238 }, { "epoch": 0.0741007226596754, "grad_norm": 1.8447102308273315, "learning_rate": 8.134425143986107e-06, "loss": 0.2031, "step": 8239 }, { "epoch": 0.07410971655731587, "grad_norm": 3.924147367477417, "learning_rate": 8.1288780455512e-06, "loss": 0.732, "step": 8240 }, { "epoch": 0.07411871045495631, "grad_norm": 9.159504890441895, "learning_rate": 8.123332471910134e-06, "loss": 0.422, "step": 8241 }, { "epoch": 0.07412770435259676, "grad_norm": 11.69211196899414, "learning_rate": 8.117788423564098e-06, "loss": 0.6301, "step": 8242 }, { "epoch": 0.07413669825023722, "grad_norm": 2.443594455718994, "learning_rate": 8.112245901014165e-06, "loss": 0.8092, "step": 8243 }, { "epoch": 0.07414569214787767, "grad_norm": 4.146954536437988, "learning_rate": 8.106704904761254e-06, "loss": 0.6703, "step": 8244 }, { "epoch": 0.07415468604551811, "grad_norm": 15.589235305786133, "learning_rate": 8.10116543530615e-06, "loss": 0.2075, "step": 8245 }, { "epoch": 0.07416367994315856, "grad_norm": 5.756708145141602, "learning_rate": 8.09562749314952e-06, "loss": 0.6494, "step": 8246 }, { "epoch": 0.07417267384079902, "grad_norm": 7.108509540557861, "learning_rate": 8.090091078791861e-06, "loss": 0.5569, "step": 8247 }, { "epoch": 0.07418166773843947, "grad_norm": 3.0335135459899902, "learning_rate": 8.084556192733567e-06, "loss": 0.2506, "step": 8248 }, { "epoch": 0.07419066163607992, "grad_norm": 2.8295106887817383, "learning_rate": 8.079022835474853e-06, "loss": 0.3764, "step": 8249 }, { "epoch": 0.07419965553372038, "grad_norm": 3.9015421867370605, "learning_rate": 8.07349100751584e-06, "loss": 0.3707, "step": 8250 }, { "epoch": 0.07420864943136082, "grad_norm": 7.591548442840576, "learning_rate": 8.067960709356478e-06, "loss": 0.8536, "step": 8251 }, { "epoch": 0.07421764332900127, "grad_norm": 3.1867504119873047, "learning_rate": 8.062431941496578e-06, "loss": 0.3394, "step": 8252 }, { "epoch": 0.07422663722664172, "grad_norm": 2.1505815982818604, "learning_rate": 8.056904704435847e-06, "loss": 0.3264, "step": 8253 }, { "epoch": 0.07423563112428218, "grad_norm": 2.9525704383850098, "learning_rate": 8.051378998673812e-06, "loss": 0.7361, "step": 8254 }, { "epoch": 0.07424462502192263, "grad_norm": 3.466583251953125, "learning_rate": 8.045854824709891e-06, "loss": 0.6303, "step": 8255 }, { "epoch": 0.07425361891956307, "grad_norm": 3.7350988388061523, "learning_rate": 8.040332183043353e-06, "loss": 0.206, "step": 8256 }, { "epoch": 0.07426261281720353, "grad_norm": 2.5881829261779785, "learning_rate": 8.034811074173323e-06, "loss": 0.3759, "step": 8257 }, { "epoch": 0.07427160671484398, "grad_norm": 4.885267734527588, "learning_rate": 8.029291498598799e-06, "loss": 0.4233, "step": 8258 }, { "epoch": 0.07428060061248443, "grad_norm": 19.469518661499023, "learning_rate": 8.023773456818618e-06, "loss": 0.8724, "step": 8259 }, { "epoch": 0.07428959451012487, "grad_norm": 2.2432098388671875, "learning_rate": 8.018256949331516e-06, "loss": 0.2485, "step": 8260 }, { "epoch": 0.07429858840776533, "grad_norm": 6.379098892211914, "learning_rate": 8.012741976636048e-06, "loss": 0.5346, "step": 8261 }, { "epoch": 0.07430758230540578, "grad_norm": 14.416951179504395, "learning_rate": 8.007228539230676e-06, "loss": 0.5892, "step": 8262 }, { "epoch": 0.07431657620304623, "grad_norm": 3.4716198444366455, "learning_rate": 8.001716637613682e-06, "loss": 0.2289, "step": 8263 }, { "epoch": 0.07432557010068669, "grad_norm": 4.194447040557861, "learning_rate": 7.996206272283225e-06, "loss": 0.3885, "step": 8264 }, { "epoch": 0.07433456399832714, "grad_norm": 2.053662061691284, "learning_rate": 7.990697443737335e-06, "loss": 0.4795, "step": 8265 }, { "epoch": 0.07434355789596758, "grad_norm": 2.1799447536468506, "learning_rate": 7.985190152473881e-06, "loss": 0.5576, "step": 8266 }, { "epoch": 0.07435255179360804, "grad_norm": 3.8870489597320557, "learning_rate": 7.979684398990625e-06, "loss": 0.3184, "step": 8267 }, { "epoch": 0.07436154569124849, "grad_norm": 3.754389524459839, "learning_rate": 7.974180183785158e-06, "loss": 0.2678, "step": 8268 }, { "epoch": 0.07437053958888894, "grad_norm": 3.826350212097168, "learning_rate": 7.96867750735494e-06, "loss": 0.7356, "step": 8269 }, { "epoch": 0.07437953348652938, "grad_norm": 2.232435703277588, "learning_rate": 7.963176370197311e-06, "loss": 0.4086, "step": 8270 }, { "epoch": 0.07438852738416984, "grad_norm": 3.052854061126709, "learning_rate": 7.957676772809452e-06, "loss": 0.3376, "step": 8271 }, { "epoch": 0.07439752128181029, "grad_norm": 4.09420108795166, "learning_rate": 7.95217871568841e-06, "loss": 0.5246, "step": 8272 }, { "epoch": 0.07440651517945074, "grad_norm": 11.645196914672852, "learning_rate": 7.946682199331088e-06, "loss": 0.4386, "step": 8273 }, { "epoch": 0.0744155090770912, "grad_norm": 1.8426505327224731, "learning_rate": 7.941187224234264e-06, "loss": 0.5012, "step": 8274 }, { "epoch": 0.07442450297473165, "grad_norm": 2.397636651992798, "learning_rate": 7.935693790894567e-06, "loss": 0.6042, "step": 8275 }, { "epoch": 0.07443349687237209, "grad_norm": 17.004579544067383, "learning_rate": 7.930201899808475e-06, "loss": 0.6737, "step": 8276 }, { "epoch": 0.07444249077001254, "grad_norm": 3.213139295578003, "learning_rate": 7.924711551472358e-06, "loss": 0.5191, "step": 8277 }, { "epoch": 0.074451484667653, "grad_norm": 3.069812774658203, "learning_rate": 7.919222746382413e-06, "loss": 0.3108, "step": 8278 }, { "epoch": 0.07446047856529345, "grad_norm": 2.011059284210205, "learning_rate": 7.913735485034724e-06, "loss": 0.3283, "step": 8279 }, { "epoch": 0.0744694724629339, "grad_norm": 2.287670373916626, "learning_rate": 7.908249767925219e-06, "loss": 0.4516, "step": 8280 }, { "epoch": 0.07447846636057436, "grad_norm": 3.22733736038208, "learning_rate": 7.902765595549683e-06, "loss": 0.345, "step": 8281 }, { "epoch": 0.0744874602582148, "grad_norm": 3.048682928085327, "learning_rate": 7.897282968403782e-06, "loss": 0.6308, "step": 8282 }, { "epoch": 0.07449645415585525, "grad_norm": 2.0129129886627197, "learning_rate": 7.89180188698302e-06, "loss": 0.1306, "step": 8283 }, { "epoch": 0.07450544805349571, "grad_norm": 3.5746707916259766, "learning_rate": 7.886322351782783e-06, "loss": 0.1772, "step": 8284 }, { "epoch": 0.07451444195113616, "grad_norm": 2.2430124282836914, "learning_rate": 7.880844363298299e-06, "loss": 0.8113, "step": 8285 }, { "epoch": 0.0745234358487766, "grad_norm": 7.8214616775512695, "learning_rate": 7.87536792202466e-06, "loss": 0.7182, "step": 8286 }, { "epoch": 0.07453242974641705, "grad_norm": 2.2660605907440186, "learning_rate": 7.869893028456824e-06, "loss": 0.5276, "step": 8287 }, { "epoch": 0.07454142364405751, "grad_norm": 2.325359582901001, "learning_rate": 7.864419683089597e-06, "loss": 0.4556, "step": 8288 }, { "epoch": 0.07455041754169796, "grad_norm": 1.9171491861343384, "learning_rate": 7.858947886417672e-06, "loss": 0.512, "step": 8289 }, { "epoch": 0.0745594114393384, "grad_norm": 2.6914420127868652, "learning_rate": 7.853477638935564e-06, "loss": 0.4563, "step": 8290 }, { "epoch": 0.07456840533697887, "grad_norm": 3.312748670578003, "learning_rate": 7.848008941137688e-06, "loss": 0.184, "step": 8291 }, { "epoch": 0.07457739923461931, "grad_norm": 10.215374946594238, "learning_rate": 7.842541793518287e-06, "loss": 0.4848, "step": 8292 }, { "epoch": 0.07458639313225976, "grad_norm": 2.245692491531372, "learning_rate": 7.837076196571472e-06, "loss": 0.4994, "step": 8293 }, { "epoch": 0.0745953870299002, "grad_norm": 2.8157076835632324, "learning_rate": 7.831612150791232e-06, "loss": 0.4557, "step": 8294 }, { "epoch": 0.07460438092754067, "grad_norm": 2.1680712699890137, "learning_rate": 7.826149656671386e-06, "loss": 0.3044, "step": 8295 }, { "epoch": 0.07461337482518111, "grad_norm": 2.2796974182128906, "learning_rate": 7.820688714705644e-06, "loss": 0.4857, "step": 8296 }, { "epoch": 0.07462236872282156, "grad_norm": 2.0493624210357666, "learning_rate": 7.81522932538755e-06, "loss": 0.2531, "step": 8297 }, { "epoch": 0.07463136262046202, "grad_norm": 1.6253422498703003, "learning_rate": 7.809771489210518e-06, "loss": 0.2897, "step": 8298 }, { "epoch": 0.07464035651810247, "grad_norm": 16.003507614135742, "learning_rate": 7.80431520666782e-06, "loss": 0.7299, "step": 8299 }, { "epoch": 0.07464935041574292, "grad_norm": 1.7532169818878174, "learning_rate": 7.798860478252598e-06, "loss": 0.3064, "step": 8300 }, { "epoch": 0.07465834431338338, "grad_norm": 29.62613296508789, "learning_rate": 7.793407304457836e-06, "loss": 0.8517, "step": 8301 }, { "epoch": 0.07466733821102382, "grad_norm": 1.9708420038223267, "learning_rate": 7.787955685776383e-06, "loss": 0.5276, "step": 8302 }, { "epoch": 0.07467633210866427, "grad_norm": 2.0618531703948975, "learning_rate": 7.782505622700965e-06, "loss": 0.3378, "step": 8303 }, { "epoch": 0.07468532600630472, "grad_norm": 4.270646572113037, "learning_rate": 7.77705711572414e-06, "loss": 0.4902, "step": 8304 }, { "epoch": 0.07469431990394518, "grad_norm": 43.73309326171875, "learning_rate": 7.771610165338337e-06, "loss": 0.9638, "step": 8305 }, { "epoch": 0.07470331380158562, "grad_norm": 1.9851726293563843, "learning_rate": 7.766164772035856e-06, "loss": 0.4748, "step": 8306 }, { "epoch": 0.07471230769922607, "grad_norm": 1.8635668754577637, "learning_rate": 7.760720936308836e-06, "loss": 0.4307, "step": 8307 }, { "epoch": 0.07472130159686653, "grad_norm": 21.381223678588867, "learning_rate": 7.755278658649293e-06, "loss": 0.8957, "step": 8308 }, { "epoch": 0.07473029549450698, "grad_norm": 2.810300350189209, "learning_rate": 7.74983793954909e-06, "loss": 0.1538, "step": 8309 }, { "epoch": 0.07473928939214743, "grad_norm": 5.268619537353516, "learning_rate": 7.744398779499948e-06, "loss": 0.324, "step": 8310 }, { "epoch": 0.07474828328978787, "grad_norm": 1.5638600587844849, "learning_rate": 7.738961178993468e-06, "loss": 0.1496, "step": 8311 }, { "epoch": 0.07475727718742833, "grad_norm": 23.081026077270508, "learning_rate": 7.733525138521081e-06, "loss": 0.5433, "step": 8312 }, { "epoch": 0.07476627108506878, "grad_norm": 2.6441478729248047, "learning_rate": 7.728090658574094e-06, "loss": 0.5382, "step": 8313 }, { "epoch": 0.07477526498270923, "grad_norm": 4.179241180419922, "learning_rate": 7.722657739643663e-06, "loss": 0.2968, "step": 8314 }, { "epoch": 0.07478425888034969, "grad_norm": 2.899247884750366, "learning_rate": 7.717226382220825e-06, "loss": 0.3831, "step": 8315 }, { "epoch": 0.07479325277799014, "grad_norm": 11.158818244934082, "learning_rate": 7.711796586796449e-06, "loss": 0.6001, "step": 8316 }, { "epoch": 0.07480224667563058, "grad_norm": 9.037158012390137, "learning_rate": 7.706368353861269e-06, "loss": 0.6078, "step": 8317 }, { "epoch": 0.07481124057327103, "grad_norm": 26.79269027709961, "learning_rate": 7.700941683905899e-06, "loss": 0.497, "step": 8318 }, { "epoch": 0.07482023447091149, "grad_norm": 11.275419235229492, "learning_rate": 7.695516577420774e-06, "loss": 0.6846, "step": 8319 }, { "epoch": 0.07482922836855194, "grad_norm": 1.4525642395019531, "learning_rate": 7.690093034896236e-06, "loss": 0.2637, "step": 8320 }, { "epoch": 0.07483822226619238, "grad_norm": 2.5811216831207275, "learning_rate": 7.684671056822443e-06, "loss": 0.2432, "step": 8321 }, { "epoch": 0.07484721616383284, "grad_norm": 3.8531107902526855, "learning_rate": 7.67925064368942e-06, "loss": 0.5441, "step": 8322 }, { "epoch": 0.07485621006147329, "grad_norm": 1.8020200729370117, "learning_rate": 7.673831795987075e-06, "loss": 0.3704, "step": 8323 }, { "epoch": 0.07486520395911374, "grad_norm": 2.189771890640259, "learning_rate": 7.668414514205145e-06, "loss": 0.415, "step": 8324 }, { "epoch": 0.0748741978567542, "grad_norm": 1.4273436069488525, "learning_rate": 7.662998798833248e-06, "loss": 0.301, "step": 8325 }, { "epoch": 0.07488319175439465, "grad_norm": 3.126162528991699, "learning_rate": 7.657584650360847e-06, "loss": 0.244, "step": 8326 }, { "epoch": 0.07489218565203509, "grad_norm": 21.114208221435547, "learning_rate": 7.652172069277264e-06, "loss": 0.4786, "step": 8327 }, { "epoch": 0.07490117954967554, "grad_norm": 3.6259968280792236, "learning_rate": 7.646761056071686e-06, "loss": 0.5222, "step": 8328 }, { "epoch": 0.074910173447316, "grad_norm": 2.0097270011901855, "learning_rate": 7.641351611233144e-06, "loss": 0.8035, "step": 8329 }, { "epoch": 0.07491916734495645, "grad_norm": 2.768550157546997, "learning_rate": 7.635943735250553e-06, "loss": 0.3537, "step": 8330 }, { "epoch": 0.0749281612425969, "grad_norm": 1.670568823814392, "learning_rate": 7.630537428612658e-06, "loss": 0.1265, "step": 8331 }, { "epoch": 0.07493715514023735, "grad_norm": 8.332469940185547, "learning_rate": 7.625132691808087e-06, "loss": 0.6113, "step": 8332 }, { "epoch": 0.0749461490378778, "grad_norm": 11.012784004211426, "learning_rate": 7.619729525325312e-06, "loss": 0.9463, "step": 8333 }, { "epoch": 0.07495514293551825, "grad_norm": 2.2509493827819824, "learning_rate": 7.614327929652651e-06, "loss": 0.3951, "step": 8334 }, { "epoch": 0.0749641368331587, "grad_norm": 2.1660218238830566, "learning_rate": 7.6089279052783156e-06, "loss": 0.4341, "step": 8335 }, { "epoch": 0.07497313073079916, "grad_norm": 2.561673402786255, "learning_rate": 7.603529452690333e-06, "loss": 0.6151, "step": 8336 }, { "epoch": 0.0749821246284396, "grad_norm": 2.5340304374694824, "learning_rate": 7.598132572376632e-06, "loss": 0.7803, "step": 8337 }, { "epoch": 0.07499111852608005, "grad_norm": 2.543431520462036, "learning_rate": 7.5927372648249565e-06, "loss": 0.1647, "step": 8338 }, { "epoch": 0.07500011242372051, "grad_norm": 2.653242588043213, "learning_rate": 7.587343530522945e-06, "loss": 0.5755, "step": 8339 }, { "epoch": 0.07500910632136096, "grad_norm": 11.948578834533691, "learning_rate": 7.581951369958068e-06, "loss": 0.3736, "step": 8340 }, { "epoch": 0.0750181002190014, "grad_norm": 3.0051419734954834, "learning_rate": 7.576560783617668e-06, "loss": 0.2426, "step": 8341 }, { "epoch": 0.07502709411664187, "grad_norm": 1.9533677101135254, "learning_rate": 7.571171771988936e-06, "loss": 0.5172, "step": 8342 }, { "epoch": 0.07503608801428231, "grad_norm": 3.365807294845581, "learning_rate": 7.565784335558917e-06, "loss": 0.6914, "step": 8343 }, { "epoch": 0.07504508191192276, "grad_norm": 2.7553908824920654, "learning_rate": 7.560398474814542e-06, "loss": 0.3015, "step": 8344 }, { "epoch": 0.0750540758095632, "grad_norm": 1.9740166664123535, "learning_rate": 7.555014190242565e-06, "loss": 0.4936, "step": 8345 }, { "epoch": 0.07506306970720367, "grad_norm": 19.346588134765625, "learning_rate": 7.549631482329611e-06, "loss": 1.0177, "step": 8346 }, { "epoch": 0.07507206360484411, "grad_norm": 22.328462600708008, "learning_rate": 7.544250351562174e-06, "loss": 0.7552, "step": 8347 }, { "epoch": 0.07508105750248456, "grad_norm": 23.09140968322754, "learning_rate": 7.538870798426581e-06, "loss": 0.6119, "step": 8348 }, { "epoch": 0.07509005140012502, "grad_norm": 1.5785826444625854, "learning_rate": 7.533492823409044e-06, "loss": 0.6379, "step": 8349 }, { "epoch": 0.07509904529776547, "grad_norm": 2.5550878047943115, "learning_rate": 7.528116426995604e-06, "loss": 0.4156, "step": 8350 }, { "epoch": 0.07510803919540592, "grad_norm": 12.63647747039795, "learning_rate": 7.522741609672193e-06, "loss": 0.7346, "step": 8351 }, { "epoch": 0.07511703309304636, "grad_norm": 2.8971354961395264, "learning_rate": 7.517368371924569e-06, "loss": 0.3954, "step": 8352 }, { "epoch": 0.07512602699068682, "grad_norm": 17.67403793334961, "learning_rate": 7.511996714238353e-06, "loss": 0.3786, "step": 8353 }, { "epoch": 0.07513502088832727, "grad_norm": 8.775568962097168, "learning_rate": 7.5066266370990455e-06, "loss": 0.9853, "step": 8354 }, { "epoch": 0.07514401478596772, "grad_norm": 2.9338245391845703, "learning_rate": 7.501258140991979e-06, "loss": 0.4549, "step": 8355 }, { "epoch": 0.07515300868360818, "grad_norm": 2.4050064086914062, "learning_rate": 7.495891226402355e-06, "loss": 0.2072, "step": 8356 }, { "epoch": 0.07516200258124862, "grad_norm": 3.098163366317749, "learning_rate": 7.490525893815228e-06, "loss": 0.6315, "step": 8357 }, { "epoch": 0.07517099647888907, "grad_norm": 2.5195558071136475, "learning_rate": 7.485162143715502e-06, "loss": 0.1166, "step": 8358 }, { "epoch": 0.07517999037652952, "grad_norm": 15.306504249572754, "learning_rate": 7.479799976587964e-06, "loss": 0.8335, "step": 8359 }, { "epoch": 0.07518898427416998, "grad_norm": 1.8734607696533203, "learning_rate": 7.474439392917226e-06, "loss": 0.1504, "step": 8360 }, { "epoch": 0.07519797817181043, "grad_norm": 2.376333475112915, "learning_rate": 7.469080393187786e-06, "loss": 0.3349, "step": 8361 }, { "epoch": 0.07520697206945087, "grad_norm": 3.4623210430145264, "learning_rate": 7.463722977883977e-06, "loss": 0.7466, "step": 8362 }, { "epoch": 0.07521596596709133, "grad_norm": 16.70173454284668, "learning_rate": 7.458367147489989e-06, "loss": 0.4693, "step": 8363 }, { "epoch": 0.07522495986473178, "grad_norm": 2.4429683685302734, "learning_rate": 7.453012902489895e-06, "loss": 0.6598, "step": 8364 }, { "epoch": 0.07523395376237223, "grad_norm": 15.451314926147461, "learning_rate": 7.447660243367582e-06, "loss": 0.8257, "step": 8365 }, { "epoch": 0.07524294766001269, "grad_norm": 15.858338356018066, "learning_rate": 7.44230917060684e-06, "loss": 0.6204, "step": 8366 }, { "epoch": 0.07525194155765313, "grad_norm": 9.798660278320312, "learning_rate": 7.436959684691286e-06, "loss": 0.6457, "step": 8367 }, { "epoch": 0.07526093545529358, "grad_norm": 2.102044105529785, "learning_rate": 7.431611786104395e-06, "loss": 0.4808, "step": 8368 }, { "epoch": 0.07526992935293403, "grad_norm": 2.358524799346924, "learning_rate": 7.426265475329508e-06, "loss": 0.6013, "step": 8369 }, { "epoch": 0.07527892325057449, "grad_norm": 2.8481228351593018, "learning_rate": 7.420920752849814e-06, "loss": 0.8547, "step": 8370 }, { "epoch": 0.07528791714821494, "grad_norm": 2.1256022453308105, "learning_rate": 7.415577619148373e-06, "loss": 0.3744, "step": 8371 }, { "epoch": 0.07529691104585538, "grad_norm": 2.2499139308929443, "learning_rate": 7.410236074708077e-06, "loss": 0.5805, "step": 8372 }, { "epoch": 0.07530590494349584, "grad_norm": 2.8053696155548096, "learning_rate": 7.404896120011709e-06, "loss": 0.3635, "step": 8373 }, { "epoch": 0.07531489884113629, "grad_norm": 2.166879415512085, "learning_rate": 7.399557755541878e-06, "loss": 0.5664, "step": 8374 }, { "epoch": 0.07532389273877674, "grad_norm": 2.523719310760498, "learning_rate": 7.39422098178105e-06, "loss": 0.288, "step": 8375 }, { "epoch": 0.07533288663641718, "grad_norm": 1.8111780881881714, "learning_rate": 7.388885799211573e-06, "loss": 0.3934, "step": 8376 }, { "epoch": 0.07534188053405765, "grad_norm": 2.3016088008880615, "learning_rate": 7.3835522083156225e-06, "loss": 0.7713, "step": 8377 }, { "epoch": 0.07535087443169809, "grad_norm": 2.2202272415161133, "learning_rate": 7.378220209575256e-06, "loss": 0.6349, "step": 8378 }, { "epoch": 0.07535986832933854, "grad_norm": 20.465255737304688, "learning_rate": 7.372889803472358e-06, "loss": 0.5667, "step": 8379 }, { "epoch": 0.075368862226979, "grad_norm": 15.492762565612793, "learning_rate": 7.3675609904887005e-06, "loss": 0.5155, "step": 8380 }, { "epoch": 0.07537785612461945, "grad_norm": 2.2641491889953613, "learning_rate": 7.3622337711058895e-06, "loss": 0.384, "step": 8381 }, { "epoch": 0.0753868500222599, "grad_norm": 3.8160417079925537, "learning_rate": 7.356908145805394e-06, "loss": 0.0996, "step": 8382 }, { "epoch": 0.07539584391990035, "grad_norm": 17.645496368408203, "learning_rate": 7.351584115068535e-06, "loss": 0.3148, "step": 8383 }, { "epoch": 0.0754048378175408, "grad_norm": 2.148362636566162, "learning_rate": 7.346261679376487e-06, "loss": 0.455, "step": 8384 }, { "epoch": 0.07541383171518125, "grad_norm": 10.706427574157715, "learning_rate": 7.3409408392103e-06, "loss": 0.9372, "step": 8385 }, { "epoch": 0.0754228256128217, "grad_norm": 2.426600694656372, "learning_rate": 7.3356215950508595e-06, "loss": 0.5345, "step": 8386 }, { "epoch": 0.07543181951046216, "grad_norm": 13.185954093933105, "learning_rate": 7.330303947378906e-06, "loss": 0.9535, "step": 8387 }, { "epoch": 0.0754408134081026, "grad_norm": 2.788595199584961, "learning_rate": 7.324987896675056e-06, "loss": 0.7337, "step": 8388 }, { "epoch": 0.07544980730574305, "grad_norm": 15.210824012756348, "learning_rate": 7.3196734434197555e-06, "loss": 0.5564, "step": 8389 }, { "epoch": 0.07545880120338351, "grad_norm": 3.196861743927002, "learning_rate": 7.314360588093333e-06, "loss": 0.2462, "step": 8390 }, { "epoch": 0.07546779510102396, "grad_norm": 2.7113866806030273, "learning_rate": 7.309049331175943e-06, "loss": 0.3866, "step": 8391 }, { "epoch": 0.0754767889986644, "grad_norm": 2.1298115253448486, "learning_rate": 7.303739673147628e-06, "loss": 0.3813, "step": 8392 }, { "epoch": 0.07548578289630485, "grad_norm": 14.145027160644531, "learning_rate": 7.29843161448826e-06, "loss": 0.674, "step": 8393 }, { "epoch": 0.07549477679394531, "grad_norm": 18.484895706176758, "learning_rate": 7.293125155677566e-06, "loss": 0.6286, "step": 8394 }, { "epoch": 0.07550377069158576, "grad_norm": 1.6861330270767212, "learning_rate": 7.287820297195158e-06, "loss": 0.2428, "step": 8395 }, { "epoch": 0.0755127645892262, "grad_norm": 1.7396950721740723, "learning_rate": 7.282517039520475e-06, "loss": 0.5846, "step": 8396 }, { "epoch": 0.07552175848686667, "grad_norm": 1.909887433052063, "learning_rate": 7.277215383132816e-06, "loss": 0.1632, "step": 8397 }, { "epoch": 0.07553075238450711, "grad_norm": 2.5469295978546143, "learning_rate": 7.271915328511341e-06, "loss": 0.5811, "step": 8398 }, { "epoch": 0.07553974628214756, "grad_norm": 3.538804531097412, "learning_rate": 7.2666168761350575e-06, "loss": 0.1847, "step": 8399 }, { "epoch": 0.07554874017978801, "grad_norm": 3.0350868701934814, "learning_rate": 7.2613200264828484e-06, "loss": 0.4144, "step": 8400 }, { "epoch": 0.07555773407742847, "grad_norm": 4.19800329208374, "learning_rate": 7.256024780033418e-06, "loss": 0.2756, "step": 8401 }, { "epoch": 0.07556672797506891, "grad_norm": 3.8818087577819824, "learning_rate": 7.250731137265365e-06, "loss": 0.3715, "step": 8402 }, { "epoch": 0.07557572187270936, "grad_norm": 1.8385734558105469, "learning_rate": 7.2454390986571056e-06, "loss": 0.3613, "step": 8403 }, { "epoch": 0.07558471577034982, "grad_norm": 2.2281687259674072, "learning_rate": 7.240148664686946e-06, "loss": 0.6033, "step": 8404 }, { "epoch": 0.07559370966799027, "grad_norm": 14.098738670349121, "learning_rate": 7.234859835833021e-06, "loss": 0.8411, "step": 8405 }, { "epoch": 0.07560270356563072, "grad_norm": 20.930604934692383, "learning_rate": 7.2295726125733206e-06, "loss": 0.5034, "step": 8406 }, { "epoch": 0.07561169746327118, "grad_norm": 3.1239442825317383, "learning_rate": 7.2242869953857125e-06, "loss": 0.2921, "step": 8407 }, { "epoch": 0.07562069136091162, "grad_norm": 17.029890060424805, "learning_rate": 7.219002984747894e-06, "loss": 0.6407, "step": 8408 }, { "epoch": 0.07562968525855207, "grad_norm": 2.2584304809570312, "learning_rate": 7.21372058113744e-06, "loss": 0.9368, "step": 8409 }, { "epoch": 0.07563867915619252, "grad_norm": 2.1793556213378906, "learning_rate": 7.208439785031762e-06, "loss": 0.5491, "step": 8410 }, { "epoch": 0.07564767305383298, "grad_norm": 19.853107452392578, "learning_rate": 7.203160596908134e-06, "loss": 0.5917, "step": 8411 }, { "epoch": 0.07565666695147343, "grad_norm": 37.039546966552734, "learning_rate": 7.197883017243681e-06, "loss": 0.5518, "step": 8412 }, { "epoch": 0.07566566084911387, "grad_norm": 30.96685218811035, "learning_rate": 7.192607046515379e-06, "loss": 0.7759, "step": 8413 }, { "epoch": 0.07567465474675433, "grad_norm": 4.251175880432129, "learning_rate": 7.187332685200079e-06, "loss": 0.2228, "step": 8414 }, { "epoch": 0.07568364864439478, "grad_norm": 14.476906776428223, "learning_rate": 7.182059933774457e-06, "loss": 0.4357, "step": 8415 }, { "epoch": 0.07569264254203523, "grad_norm": 1.9290579557418823, "learning_rate": 7.176788792715075e-06, "loss": 0.1292, "step": 8416 }, { "epoch": 0.07570163643967567, "grad_norm": 10.355091094970703, "learning_rate": 7.171519262498322e-06, "loss": 0.6999, "step": 8417 }, { "epoch": 0.07571063033731613, "grad_norm": 1.9296400547027588, "learning_rate": 7.1662513436004495e-06, "loss": 0.4699, "step": 8418 }, { "epoch": 0.07571962423495658, "grad_norm": 3.229037046432495, "learning_rate": 7.160985036497578e-06, "loss": 0.5457, "step": 8419 }, { "epoch": 0.07572861813259703, "grad_norm": 19.452308654785156, "learning_rate": 7.155720341665656e-06, "loss": 0.9458, "step": 8420 }, { "epoch": 0.07573761203023749, "grad_norm": 2.3186912536621094, "learning_rate": 7.1504572595805204e-06, "loss": 0.2876, "step": 8421 }, { "epoch": 0.07574660592787794, "grad_norm": 2.0589683055877686, "learning_rate": 7.145195790717829e-06, "loss": 0.5964, "step": 8422 }, { "epoch": 0.07575559982551838, "grad_norm": 2.236511707305908, "learning_rate": 7.1399359355531104e-06, "loss": 0.3668, "step": 8423 }, { "epoch": 0.07576459372315884, "grad_norm": 12.503987312316895, "learning_rate": 7.134677694561745e-06, "loss": 0.539, "step": 8424 }, { "epoch": 0.07577358762079929, "grad_norm": 3.1838631629943848, "learning_rate": 7.1294210682189585e-06, "loss": 0.2813, "step": 8425 }, { "epoch": 0.07578258151843974, "grad_norm": 2.469102144241333, "learning_rate": 7.124166056999854e-06, "loss": 0.3508, "step": 8426 }, { "epoch": 0.07579157541608018, "grad_norm": 2.9100167751312256, "learning_rate": 7.118912661379368e-06, "loss": 0.6283, "step": 8427 }, { "epoch": 0.07580056931372064, "grad_norm": 3.3599023818969727, "learning_rate": 7.113660881832288e-06, "loss": 0.3093, "step": 8428 }, { "epoch": 0.07580956321136109, "grad_norm": 2.322831392288208, "learning_rate": 7.10841071883328e-06, "loss": 0.4922, "step": 8429 }, { "epoch": 0.07581855710900154, "grad_norm": 3.3498294353485107, "learning_rate": 7.1031621728568336e-06, "loss": 0.3848, "step": 8430 }, { "epoch": 0.075827551006642, "grad_norm": 3.723520278930664, "learning_rate": 7.097915244377318e-06, "loss": 0.6097, "step": 8431 }, { "epoch": 0.07583654490428245, "grad_norm": 2.836073875427246, "learning_rate": 7.092669933868934e-06, "loss": 0.5735, "step": 8432 }, { "epoch": 0.0758455388019229, "grad_norm": 3.0568454265594482, "learning_rate": 7.087426241805761e-06, "loss": 0.574, "step": 8433 }, { "epoch": 0.07585453269956334, "grad_norm": 2.8390696048736572, "learning_rate": 7.08218416866171e-06, "loss": 0.4232, "step": 8434 }, { "epoch": 0.0758635265972038, "grad_norm": 2.5054826736450195, "learning_rate": 7.076943714910547e-06, "loss": 0.743, "step": 8435 }, { "epoch": 0.07587252049484425, "grad_norm": 6.082218170166016, "learning_rate": 7.071704881025915e-06, "loss": 0.6795, "step": 8436 }, { "epoch": 0.0758815143924847, "grad_norm": 3.5660533905029297, "learning_rate": 7.066467667481283e-06, "loss": 0.5196, "step": 8437 }, { "epoch": 0.07589050829012516, "grad_norm": 22.15796661376953, "learning_rate": 7.061232074749985e-06, "loss": 0.8035, "step": 8438 }, { "epoch": 0.0758995021877656, "grad_norm": 3.030656099319458, "learning_rate": 7.055998103305214e-06, "loss": 0.6298, "step": 8439 }, { "epoch": 0.07590849608540605, "grad_norm": 3.146868944168091, "learning_rate": 7.050765753619998e-06, "loss": 0.3555, "step": 8440 }, { "epoch": 0.0759174899830465, "grad_norm": 3.4990429878234863, "learning_rate": 7.045535026167246e-06, "loss": 0.4967, "step": 8441 }, { "epoch": 0.07592648388068696, "grad_norm": 19.15091323852539, "learning_rate": 7.040305921419693e-06, "loss": 0.3703, "step": 8442 }, { "epoch": 0.0759354777783274, "grad_norm": 2.0827150344848633, "learning_rate": 7.035078439849954e-06, "loss": 0.2474, "step": 8443 }, { "epoch": 0.07594447167596785, "grad_norm": 1.8849478960037231, "learning_rate": 7.029852581930466e-06, "loss": 0.2017, "step": 8444 }, { "epoch": 0.07595346557360831, "grad_norm": 1.8229984045028687, "learning_rate": 7.0246283481335555e-06, "loss": 0.2405, "step": 8445 }, { "epoch": 0.07596245947124876, "grad_norm": 3.5028562545776367, "learning_rate": 7.019405738931372e-06, "loss": 0.1866, "step": 8446 }, { "epoch": 0.0759714533688892, "grad_norm": 8.966373443603516, "learning_rate": 7.014184754795925e-06, "loss": 0.4855, "step": 8447 }, { "epoch": 0.07598044726652967, "grad_norm": 2.479067325592041, "learning_rate": 7.008965396199091e-06, "loss": 0.5424, "step": 8448 }, { "epoch": 0.07598944116417011, "grad_norm": 2.4170613288879395, "learning_rate": 7.003747663612581e-06, "loss": 0.7225, "step": 8449 }, { "epoch": 0.07599843506181056, "grad_norm": 12.3326997756958, "learning_rate": 6.998531557507981e-06, "loss": 0.5877, "step": 8450 }, { "epoch": 0.076007428959451, "grad_norm": 20.80951499938965, "learning_rate": 6.993317078356709e-06, "loss": 0.9344, "step": 8451 }, { "epoch": 0.07601642285709147, "grad_norm": 13.292749404907227, "learning_rate": 6.988104226630046e-06, "loss": 0.7085, "step": 8452 }, { "epoch": 0.07602541675473191, "grad_norm": 10.45980453491211, "learning_rate": 6.982893002799121e-06, "loss": 0.789, "step": 8453 }, { "epoch": 0.07603441065237236, "grad_norm": 2.0548675060272217, "learning_rate": 6.9776834073349115e-06, "loss": 0.1354, "step": 8454 }, { "epoch": 0.07604340455001282, "grad_norm": 2.8869025707244873, "learning_rate": 6.972475440708271e-06, "loss": 0.578, "step": 8455 }, { "epoch": 0.07605239844765327, "grad_norm": 2.1925857067108154, "learning_rate": 6.967269103389875e-06, "loss": 0.3088, "step": 8456 }, { "epoch": 0.07606139234529372, "grad_norm": 3.395970344543457, "learning_rate": 6.962064395850282e-06, "loss": 0.2862, "step": 8457 }, { "epoch": 0.07607038624293416, "grad_norm": 3.243396759033203, "learning_rate": 6.956861318559882e-06, "loss": 0.3018, "step": 8458 }, { "epoch": 0.07607938014057462, "grad_norm": 2.501438617706299, "learning_rate": 6.951659871988911e-06, "loss": 0.5661, "step": 8459 }, { "epoch": 0.07608837403821507, "grad_norm": 3.975895643234253, "learning_rate": 6.94646005660749e-06, "loss": 0.2903, "step": 8460 }, { "epoch": 0.07609736793585552, "grad_norm": 26.230566024780273, "learning_rate": 6.941261872885554e-06, "loss": 0.6285, "step": 8461 }, { "epoch": 0.07610636183349598, "grad_norm": 2.6870737075805664, "learning_rate": 6.936065321292928e-06, "loss": 0.3028, "step": 8462 }, { "epoch": 0.07611535573113642, "grad_norm": 2.307006597518921, "learning_rate": 6.930870402299261e-06, "loss": 0.4089, "step": 8463 }, { "epoch": 0.07612434962877687, "grad_norm": 2.772844076156616, "learning_rate": 6.925677116374057e-06, "loss": 0.4299, "step": 8464 }, { "epoch": 0.07613334352641733, "grad_norm": 5.921109199523926, "learning_rate": 6.920485463986695e-06, "loss": 0.5838, "step": 8465 }, { "epoch": 0.07614233742405778, "grad_norm": 11.866799354553223, "learning_rate": 6.9152954456063816e-06, "loss": 0.6381, "step": 8466 }, { "epoch": 0.07615133132169823, "grad_norm": 2.2103734016418457, "learning_rate": 6.91010706170219e-06, "loss": 0.2061, "step": 8467 }, { "epoch": 0.07616032521933867, "grad_norm": 3.628046751022339, "learning_rate": 6.904920312743026e-06, "loss": 0.3221, "step": 8468 }, { "epoch": 0.07616931911697913, "grad_norm": 3.034322500228882, "learning_rate": 6.899735199197685e-06, "loss": 0.6115, "step": 8469 }, { "epoch": 0.07617831301461958, "grad_norm": 2.6680655479431152, "learning_rate": 6.894551721534778e-06, "loss": 0.117, "step": 8470 }, { "epoch": 0.07618730691226003, "grad_norm": 2.7111713886260986, "learning_rate": 6.889369880222776e-06, "loss": 0.1392, "step": 8471 }, { "epoch": 0.07619630080990049, "grad_norm": 2.3511595726013184, "learning_rate": 6.884189675730026e-06, "loss": 0.2585, "step": 8472 }, { "epoch": 0.07620529470754094, "grad_norm": 2.1063947677612305, "learning_rate": 6.8790111085246935e-06, "loss": 0.6226, "step": 8473 }, { "epoch": 0.07621428860518138, "grad_norm": 2.5690245628356934, "learning_rate": 6.873834179074828e-06, "loss": 0.7566, "step": 8474 }, { "epoch": 0.07622328250282183, "grad_norm": 2.0814449787139893, "learning_rate": 6.868658887848303e-06, "loss": 0.4354, "step": 8475 }, { "epoch": 0.07623227640046229, "grad_norm": 2.554464817047119, "learning_rate": 6.863485235312853e-06, "loss": 0.6975, "step": 8476 }, { "epoch": 0.07624127029810274, "grad_norm": 11.453028678894043, "learning_rate": 6.858313221936078e-06, "loss": 0.6198, "step": 8477 }, { "epoch": 0.07625026419574318, "grad_norm": 2.1385841369628906, "learning_rate": 6.853142848185412e-06, "loss": 0.3331, "step": 8478 }, { "epoch": 0.07625925809338364, "grad_norm": 2.4501469135284424, "learning_rate": 6.847974114528144e-06, "loss": 0.4384, "step": 8479 }, { "epoch": 0.07626825199102409, "grad_norm": 3.0832719802856445, "learning_rate": 6.84280702143143e-06, "loss": 0.5297, "step": 8480 }, { "epoch": 0.07627724588866454, "grad_norm": 2.1655638217926025, "learning_rate": 6.837641569362261e-06, "loss": 0.5665, "step": 8481 }, { "epoch": 0.07628623978630499, "grad_norm": 3.0070834159851074, "learning_rate": 6.832477758787484e-06, "loss": 0.1943, "step": 8482 }, { "epoch": 0.07629523368394545, "grad_norm": 2.1729929447174072, "learning_rate": 6.8273155901737915e-06, "loss": 0.1536, "step": 8483 }, { "epoch": 0.07630422758158589, "grad_norm": 2.268489122390747, "learning_rate": 6.822155063987748e-06, "loss": 0.4201, "step": 8484 }, { "epoch": 0.07631322147922634, "grad_norm": 1.7633932828903198, "learning_rate": 6.816996180695742e-06, "loss": 0.3316, "step": 8485 }, { "epoch": 0.0763222153768668, "grad_norm": 7.150792121887207, "learning_rate": 6.811838940764046e-06, "loss": 1.0349, "step": 8486 }, { "epoch": 0.07633120927450725, "grad_norm": 35.661231994628906, "learning_rate": 6.806683344658754e-06, "loss": 0.608, "step": 8487 }, { "epoch": 0.0763402031721477, "grad_norm": 2.02586030960083, "learning_rate": 6.801529392845821e-06, "loss": 0.422, "step": 8488 }, { "epoch": 0.07634919706978816, "grad_norm": 1.84721839427948, "learning_rate": 6.796377085791064e-06, "loss": 0.0523, "step": 8489 }, { "epoch": 0.0763581909674286, "grad_norm": 2.320284128189087, "learning_rate": 6.7912264239601314e-06, "loss": 0.1864, "step": 8490 }, { "epoch": 0.07636718486506905, "grad_norm": 2.5138509273529053, "learning_rate": 6.78607740781855e-06, "loss": 0.7275, "step": 8491 }, { "epoch": 0.0763761787627095, "grad_norm": 2.4561431407928467, "learning_rate": 6.780930037831673e-06, "loss": 0.4594, "step": 8492 }, { "epoch": 0.07638517266034996, "grad_norm": 2.9062304496765137, "learning_rate": 6.775784314464717e-06, "loss": 0.3187, "step": 8493 }, { "epoch": 0.0763941665579904, "grad_norm": 12.140205383300781, "learning_rate": 6.770640238182744e-06, "loss": 0.75, "step": 8494 }, { "epoch": 0.07640316045563085, "grad_norm": 2.593737840652466, "learning_rate": 6.765497809450663e-06, "loss": 0.4555, "step": 8495 }, { "epoch": 0.07641215435327131, "grad_norm": 7.649170398712158, "learning_rate": 6.760357028733258e-06, "loss": 0.5267, "step": 8496 }, { "epoch": 0.07642114825091176, "grad_norm": 16.310869216918945, "learning_rate": 6.75521789649513e-06, "loss": 0.5138, "step": 8497 }, { "epoch": 0.0764301421485522, "grad_norm": 1.8025157451629639, "learning_rate": 6.750080413200768e-06, "loss": 0.3216, "step": 8498 }, { "epoch": 0.07643913604619265, "grad_norm": 2.9307503700256348, "learning_rate": 6.7449445793144775e-06, "loss": 0.5892, "step": 8499 }, { "epoch": 0.07644812994383311, "grad_norm": 1.81307053565979, "learning_rate": 6.739810395300427e-06, "loss": 0.045, "step": 8500 }, { "epoch": 0.07645712384147356, "grad_norm": 2.9829680919647217, "learning_rate": 6.7346778616226515e-06, "loss": 0.4872, "step": 8501 }, { "epoch": 0.076466117739114, "grad_norm": 1.8427367210388184, "learning_rate": 6.729546978745011e-06, "loss": 0.2699, "step": 8502 }, { "epoch": 0.07647511163675447, "grad_norm": 3.6779720783233643, "learning_rate": 6.72441774713124e-06, "loss": 0.2623, "step": 8503 }, { "epoch": 0.07648410553439491, "grad_norm": 2.353724956512451, "learning_rate": 6.71929016724491e-06, "loss": 0.4185, "step": 8504 }, { "epoch": 0.07649309943203536, "grad_norm": 2.3168256282806396, "learning_rate": 6.714164239549437e-06, "loss": 0.4416, "step": 8505 }, { "epoch": 0.07650209332967582, "grad_norm": 2.4434094429016113, "learning_rate": 6.70903996450811e-06, "loss": 0.706, "step": 8506 }, { "epoch": 0.07651108722731627, "grad_norm": 3.555872917175293, "learning_rate": 6.703917342584051e-06, "loss": 0.5872, "step": 8507 }, { "epoch": 0.07652008112495672, "grad_norm": 2.5877866744995117, "learning_rate": 6.698796374240238e-06, "loss": 0.6416, "step": 8508 }, { "epoch": 0.07652907502259716, "grad_norm": 2.5680055618286133, "learning_rate": 6.693677059939485e-06, "loss": 0.0709, "step": 8509 }, { "epoch": 0.07653806892023762, "grad_norm": 2.2810006141662598, "learning_rate": 6.688559400144489e-06, "loss": 0.4594, "step": 8510 }, { "epoch": 0.07654706281787807, "grad_norm": 2.20872163772583, "learning_rate": 6.683443395317773e-06, "loss": 0.5215, "step": 8511 }, { "epoch": 0.07655605671551852, "grad_norm": 12.120711326599121, "learning_rate": 6.678329045921705e-06, "loss": 0.5893, "step": 8512 }, { "epoch": 0.07656505061315898, "grad_norm": 2.881378173828125, "learning_rate": 6.6732163524185325e-06, "loss": 0.4396, "step": 8513 }, { "epoch": 0.07657404451079942, "grad_norm": 14.304339408874512, "learning_rate": 6.6681053152703175e-06, "loss": 1.1793, "step": 8514 }, { "epoch": 0.07658303840843987, "grad_norm": 17.181011199951172, "learning_rate": 6.662995934939007e-06, "loss": 0.5493, "step": 8515 }, { "epoch": 0.07659203230608032, "grad_norm": 17.379901885986328, "learning_rate": 6.657888211886371e-06, "loss": 0.6013, "step": 8516 }, { "epoch": 0.07660102620372078, "grad_norm": 2.7400476932525635, "learning_rate": 6.652782146574035e-06, "loss": 0.405, "step": 8517 }, { "epoch": 0.07661002010136123, "grad_norm": 3.0534861087799072, "learning_rate": 6.647677739463495e-06, "loss": 0.4034, "step": 8518 }, { "epoch": 0.07661901399900167, "grad_norm": 2.2172017097473145, "learning_rate": 6.642574991016065e-06, "loss": 0.3171, "step": 8519 }, { "epoch": 0.07662800789664213, "grad_norm": 20.93010902404785, "learning_rate": 6.637473901692942e-06, "loss": 0.3488, "step": 8520 }, { "epoch": 0.07663700179428258, "grad_norm": 11.227700233459473, "learning_rate": 6.6323744719551496e-06, "loss": 0.5692, "step": 8521 }, { "epoch": 0.07664599569192303, "grad_norm": 2.3191726207733154, "learning_rate": 6.627276702263568e-06, "loss": 0.7944, "step": 8522 }, { "epoch": 0.07665498958956349, "grad_norm": 2.031020164489746, "learning_rate": 6.622180593078931e-06, "loss": 0.3414, "step": 8523 }, { "epoch": 0.07666398348720393, "grad_norm": 7.019061088562012, "learning_rate": 6.617086144861806e-06, "loss": 0.7917, "step": 8524 }, { "epoch": 0.07667297738484438, "grad_norm": 3.6143500804901123, "learning_rate": 6.6119933580726435e-06, "loss": 0.4352, "step": 8525 }, { "epoch": 0.07668197128248483, "grad_norm": 2.5386765003204346, "learning_rate": 6.606902233171711e-06, "loss": 0.0717, "step": 8526 }, { "epoch": 0.07669096518012529, "grad_norm": 13.323572158813477, "learning_rate": 6.601812770619145e-06, "loss": 0.6679, "step": 8527 }, { "epoch": 0.07669995907776574, "grad_norm": 2.2283942699432373, "learning_rate": 6.596724970874929e-06, "loss": 0.4071, "step": 8528 }, { "epoch": 0.07670895297540618, "grad_norm": 2.547369956970215, "learning_rate": 6.591638834398878e-06, "loss": 0.5761, "step": 8529 }, { "epoch": 0.07671794687304664, "grad_norm": 2.13053822517395, "learning_rate": 6.586554361650688e-06, "loss": 0.2572, "step": 8530 }, { "epoch": 0.07672694077068709, "grad_norm": 2.968857526779175, "learning_rate": 6.5814715530898745e-06, "loss": 0.373, "step": 8531 }, { "epoch": 0.07673593466832754, "grad_norm": 18.817556381225586, "learning_rate": 6.576390409175831e-06, "loss": 0.7543, "step": 8532 }, { "epoch": 0.07674492856596798, "grad_norm": 2.081512689590454, "learning_rate": 6.571310930367772e-06, "loss": 0.3945, "step": 8533 }, { "epoch": 0.07675392246360845, "grad_norm": 15.282315254211426, "learning_rate": 6.566233117124787e-06, "loss": 0.5225, "step": 8534 }, { "epoch": 0.07676291636124889, "grad_norm": 2.042193651199341, "learning_rate": 6.561156969905799e-06, "loss": 0.1219, "step": 8535 }, { "epoch": 0.07677191025888934, "grad_norm": 13.365447998046875, "learning_rate": 6.556082489169582e-06, "loss": 0.4146, "step": 8536 }, { "epoch": 0.0767809041565298, "grad_norm": 2.057011842727661, "learning_rate": 6.551009675374764e-06, "loss": 0.6575, "step": 8537 }, { "epoch": 0.07678989805417025, "grad_norm": 1.9905726909637451, "learning_rate": 6.545938528979812e-06, "loss": 0.3071, "step": 8538 }, { "epoch": 0.0767988919518107, "grad_norm": 2.593597650527954, "learning_rate": 6.540869050443066e-06, "loss": 0.4172, "step": 8539 }, { "epoch": 0.07680788584945114, "grad_norm": 12.828767776489258, "learning_rate": 6.535801240222694e-06, "loss": 0.6158, "step": 8540 }, { "epoch": 0.0768168797470916, "grad_norm": 3.228611707687378, "learning_rate": 6.530735098776711e-06, "loss": 0.4955, "step": 8541 }, { "epoch": 0.07682587364473205, "grad_norm": 3.3525731563568115, "learning_rate": 6.525670626563002e-06, "loss": 0.2427, "step": 8542 }, { "epoch": 0.0768348675423725, "grad_norm": 24.730186462402344, "learning_rate": 6.520607824039274e-06, "loss": 0.5823, "step": 8543 }, { "epoch": 0.07684386144001296, "grad_norm": 18.143980026245117, "learning_rate": 6.515546691663116e-06, "loss": 0.4558, "step": 8544 }, { "epoch": 0.0768528553376534, "grad_norm": 2.5841526985168457, "learning_rate": 6.5104872298919315e-06, "loss": 0.46, "step": 8545 }, { "epoch": 0.07686184923529385, "grad_norm": 1.7799733877182007, "learning_rate": 6.5054294391830025e-06, "loss": 0.3946, "step": 8546 }, { "epoch": 0.07687084313293431, "grad_norm": 2.5202860832214355, "learning_rate": 6.500373319993438e-06, "loss": 0.4141, "step": 8547 }, { "epoch": 0.07687983703057476, "grad_norm": 18.925262451171875, "learning_rate": 6.49531887278021e-06, "loss": 0.6408, "step": 8548 }, { "epoch": 0.0768888309282152, "grad_norm": 2.0021419525146484, "learning_rate": 6.4902660980001304e-06, "loss": 0.384, "step": 8549 }, { "epoch": 0.07689782482585565, "grad_norm": 11.515953063964844, "learning_rate": 6.485214996109856e-06, "loss": 0.5408, "step": 8550 }, { "epoch": 0.07690681872349611, "grad_norm": 9.568404197692871, "learning_rate": 6.480165567565913e-06, "loss": 0.4919, "step": 8551 }, { "epoch": 0.07691581262113656, "grad_norm": 16.55735969543457, "learning_rate": 6.47511781282466e-06, "loss": 0.6, "step": 8552 }, { "epoch": 0.076924806518777, "grad_norm": 10.302879333496094, "learning_rate": 6.4700717323422975e-06, "loss": 0.5955, "step": 8553 }, { "epoch": 0.07693380041641747, "grad_norm": 2.116886615753174, "learning_rate": 6.465027326574902e-06, "loss": 0.2939, "step": 8554 }, { "epoch": 0.07694279431405791, "grad_norm": 5.08701229095459, "learning_rate": 6.459984595978363e-06, "loss": 0.2666, "step": 8555 }, { "epoch": 0.07695178821169836, "grad_norm": 3.17059326171875, "learning_rate": 6.454943541008457e-06, "loss": 0.4647, "step": 8556 }, { "epoch": 0.07696078210933881, "grad_norm": 1.8975650072097778, "learning_rate": 6.449904162120768e-06, "loss": 0.3189, "step": 8557 }, { "epoch": 0.07696977600697927, "grad_norm": 3.4936256408691406, "learning_rate": 6.444866459770768e-06, "loss": 0.1357, "step": 8558 }, { "epoch": 0.07697876990461971, "grad_norm": 2.590592384338379, "learning_rate": 6.439830434413754e-06, "loss": 0.6961, "step": 8559 }, { "epoch": 0.07698776380226016, "grad_norm": 20.780460357666016, "learning_rate": 6.434796086504863e-06, "loss": 0.9501, "step": 8560 }, { "epoch": 0.07699675769990062, "grad_norm": 12.355781555175781, "learning_rate": 6.429763416499115e-06, "loss": 0.4107, "step": 8561 }, { "epoch": 0.07700575159754107, "grad_norm": 2.900789499282837, "learning_rate": 6.424732424851346e-06, "loss": 0.3088, "step": 8562 }, { "epoch": 0.07701474549518152, "grad_norm": 13.895474433898926, "learning_rate": 6.419703112016254e-06, "loss": 0.5729, "step": 8563 }, { "epoch": 0.07702373939282198, "grad_norm": 2.789640426635742, "learning_rate": 6.414675478448381e-06, "loss": 0.2237, "step": 8564 }, { "epoch": 0.07703273329046242, "grad_norm": 16.162189483642578, "learning_rate": 6.409649524602113e-06, "loss": 0.5151, "step": 8565 }, { "epoch": 0.07704172718810287, "grad_norm": 2.0734598636627197, "learning_rate": 6.404625250931704e-06, "loss": 0.3819, "step": 8566 }, { "epoch": 0.07705072108574332, "grad_norm": 14.013766288757324, "learning_rate": 6.399602657891226e-06, "loss": 0.4566, "step": 8567 }, { "epoch": 0.07705971498338378, "grad_norm": 9.412793159484863, "learning_rate": 6.3945817459346355e-06, "loss": 0.4433, "step": 8568 }, { "epoch": 0.07706870888102423, "grad_norm": 7.404946804046631, "learning_rate": 6.389562515515707e-06, "loss": 0.638, "step": 8569 }, { "epoch": 0.07707770277866467, "grad_norm": 11.771021842956543, "learning_rate": 6.384544967088063e-06, "loss": 0.4879, "step": 8570 }, { "epoch": 0.07708669667630513, "grad_norm": 2.331397533416748, "learning_rate": 6.379529101105202e-06, "loss": 0.5675, "step": 8571 }, { "epoch": 0.07709569057394558, "grad_norm": 5.562939643859863, "learning_rate": 6.374514918020439e-06, "loss": 0.2529, "step": 8572 }, { "epoch": 0.07710468447158603, "grad_norm": 9.183422088623047, "learning_rate": 6.36950241828696e-06, "loss": 0.6433, "step": 8573 }, { "epoch": 0.07711367836922647, "grad_norm": 2.576367139816284, "learning_rate": 6.36449160235778e-06, "loss": 0.605, "step": 8574 }, { "epoch": 0.07712267226686693, "grad_norm": 11.078079223632812, "learning_rate": 6.359482470685785e-06, "loss": 0.3676, "step": 8575 }, { "epoch": 0.07713166616450738, "grad_norm": 2.2848994731903076, "learning_rate": 6.354475023723686e-06, "loss": 0.5985, "step": 8576 }, { "epoch": 0.07714066006214783, "grad_norm": 3.6298444271087646, "learning_rate": 6.349469261924049e-06, "loss": 0.1995, "step": 8577 }, { "epoch": 0.07714965395978829, "grad_norm": 4.289686679840088, "learning_rate": 6.344465185739293e-06, "loss": 0.5248, "step": 8578 }, { "epoch": 0.07715864785742874, "grad_norm": 1.9761842489242554, "learning_rate": 6.33946279562167e-06, "loss": 0.5571, "step": 8579 }, { "epoch": 0.07716764175506918, "grad_norm": 2.334505319595337, "learning_rate": 6.3344620920233106e-06, "loss": 0.7583, "step": 8580 }, { "epoch": 0.07717663565270963, "grad_norm": 3.386277675628662, "learning_rate": 6.329463075396161e-06, "loss": 0.7654, "step": 8581 }, { "epoch": 0.07718562955035009, "grad_norm": 3.2110586166381836, "learning_rate": 6.324465746192021e-06, "loss": 0.4239, "step": 8582 }, { "epoch": 0.07719462344799054, "grad_norm": 2.9215939044952393, "learning_rate": 6.319470104862557e-06, "loss": 0.4204, "step": 8583 }, { "epoch": 0.07720361734563098, "grad_norm": 5.440718173980713, "learning_rate": 6.314476151859258e-06, "loss": 0.4594, "step": 8584 }, { "epoch": 0.07721261124327145, "grad_norm": 10.232893943786621, "learning_rate": 6.309483887633486e-06, "loss": 0.2702, "step": 8585 }, { "epoch": 0.07722160514091189, "grad_norm": 2.065317153930664, "learning_rate": 6.30449331263642e-06, "loss": 0.6504, "step": 8586 }, { "epoch": 0.07723059903855234, "grad_norm": 2.323225975036621, "learning_rate": 6.299504427319119e-06, "loss": 0.2729, "step": 8587 }, { "epoch": 0.0772395929361928, "grad_norm": 2.1499862670898438, "learning_rate": 6.294517232132466e-06, "loss": 0.2222, "step": 8588 }, { "epoch": 0.07724858683383325, "grad_norm": 2.3435842990875244, "learning_rate": 6.28953172752719e-06, "loss": 0.6167, "step": 8589 }, { "epoch": 0.0772575807314737, "grad_norm": 2.541618824005127, "learning_rate": 6.284547913953892e-06, "loss": 0.3149, "step": 8590 }, { "epoch": 0.07726657462911414, "grad_norm": 2.2765510082244873, "learning_rate": 6.279565791862993e-06, "loss": 0.4035, "step": 8591 }, { "epoch": 0.0772755685267546, "grad_norm": 2.4645254611968994, "learning_rate": 6.274585361704774e-06, "loss": 0.3259, "step": 8592 }, { "epoch": 0.07728456242439505, "grad_norm": 10.092869758605957, "learning_rate": 6.269606623929361e-06, "loss": 0.3813, "step": 8593 }, { "epoch": 0.0772935563220355, "grad_norm": 2.8616526126861572, "learning_rate": 6.264629578986719e-06, "loss": 0.596, "step": 8594 }, { "epoch": 0.07730255021967596, "grad_norm": 2.1883254051208496, "learning_rate": 6.259654227326681e-06, "loss": 0.3358, "step": 8595 }, { "epoch": 0.0773115441173164, "grad_norm": 2.933424949645996, "learning_rate": 6.254680569398905e-06, "loss": 0.4856, "step": 8596 }, { "epoch": 0.07732053801495685, "grad_norm": 2.6414012908935547, "learning_rate": 6.249708605652913e-06, "loss": 0.2146, "step": 8597 }, { "epoch": 0.0773295319125973, "grad_norm": 2.4077508449554443, "learning_rate": 6.244738336538053e-06, "loss": 0.6184, "step": 8598 }, { "epoch": 0.07733852581023776, "grad_norm": 8.816022872924805, "learning_rate": 6.23976976250355e-06, "loss": 0.8545, "step": 8599 }, { "epoch": 0.0773475197078782, "grad_norm": 4.384307861328125, "learning_rate": 6.23480288399845e-06, "loss": 0.3125, "step": 8600 }, { "epoch": 0.07735651360551865, "grad_norm": 13.387408256530762, "learning_rate": 6.229837701471644e-06, "loss": 0.6308, "step": 8601 }, { "epoch": 0.07736550750315911, "grad_norm": 2.0515894889831543, "learning_rate": 6.224874215371895e-06, "loss": 0.479, "step": 8602 }, { "epoch": 0.07737450140079956, "grad_norm": 1.6801950931549072, "learning_rate": 6.219912426147795e-06, "loss": 0.3316, "step": 8603 }, { "epoch": 0.07738349529844, "grad_norm": 2.489037036895752, "learning_rate": 6.214952334247773e-06, "loss": 0.3407, "step": 8604 }, { "epoch": 0.07739248919608047, "grad_norm": 2.3672592639923096, "learning_rate": 6.209993940120132e-06, "loss": 0.3305, "step": 8605 }, { "epoch": 0.07740148309372091, "grad_norm": 3.227330207824707, "learning_rate": 6.205037244213e-06, "loss": 0.3452, "step": 8606 }, { "epoch": 0.07741047699136136, "grad_norm": 10.03866195678711, "learning_rate": 6.200082246974357e-06, "loss": 0.5024, "step": 8607 }, { "epoch": 0.0774194708890018, "grad_norm": 31.65299415588379, "learning_rate": 6.195128948852025e-06, "loss": 0.6102, "step": 8608 }, { "epoch": 0.07742846478664227, "grad_norm": 2.1720807552337646, "learning_rate": 6.190177350293688e-06, "loss": 0.4137, "step": 8609 }, { "epoch": 0.07743745868428271, "grad_norm": 2.1393256187438965, "learning_rate": 6.185227451746858e-06, "loss": 0.6931, "step": 8610 }, { "epoch": 0.07744645258192316, "grad_norm": 18.740888595581055, "learning_rate": 6.18027925365891e-06, "loss": 0.3858, "step": 8611 }, { "epoch": 0.07745544647956362, "grad_norm": 15.605084419250488, "learning_rate": 6.175332756477054e-06, "loss": 0.6679, "step": 8612 }, { "epoch": 0.07746444037720407, "grad_norm": 10.622133255004883, "learning_rate": 6.170387960648338e-06, "loss": 0.7531, "step": 8613 }, { "epoch": 0.07747343427484452, "grad_norm": 1.8697258234024048, "learning_rate": 6.165444866619685e-06, "loss": 0.4301, "step": 8614 }, { "epoch": 0.07748242817248496, "grad_norm": 2.1339550018310547, "learning_rate": 6.160503474837831e-06, "loss": 0.2667, "step": 8615 }, { "epoch": 0.07749142207012542, "grad_norm": 2.234271287918091, "learning_rate": 6.155563785749388e-06, "loss": 0.2999, "step": 8616 }, { "epoch": 0.07750041596776587, "grad_norm": 2.066335439682007, "learning_rate": 6.1506257998007935e-06, "loss": 0.4743, "step": 8617 }, { "epoch": 0.07750940986540632, "grad_norm": 1.8551105260849, "learning_rate": 6.145689517438336e-06, "loss": 0.3472, "step": 8618 }, { "epoch": 0.07751840376304678, "grad_norm": 17.549970626831055, "learning_rate": 6.140754939108151e-06, "loss": 0.3132, "step": 8619 }, { "epoch": 0.07752739766068723, "grad_norm": 3.7020175457000732, "learning_rate": 6.1358220652562136e-06, "loss": 0.3846, "step": 8620 }, { "epoch": 0.07753639155832767, "grad_norm": 2.381591796875, "learning_rate": 6.130890896328368e-06, "loss": 0.4587, "step": 8621 }, { "epoch": 0.07754538545596812, "grad_norm": 2.712545871734619, "learning_rate": 6.125961432770269e-06, "loss": 0.3358, "step": 8622 }, { "epoch": 0.07755437935360858, "grad_norm": 6.771638870239258, "learning_rate": 6.121033675027455e-06, "loss": 0.7219, "step": 8623 }, { "epoch": 0.07756337325124903, "grad_norm": 2.505872964859009, "learning_rate": 6.116107623545281e-06, "loss": 0.6295, "step": 8624 }, { "epoch": 0.07757236714888947, "grad_norm": 17.856887817382812, "learning_rate": 6.111183278768956e-06, "loss": 0.2407, "step": 8625 }, { "epoch": 0.07758136104652993, "grad_norm": 3.9547901153564453, "learning_rate": 6.106260641143546e-06, "loss": 0.3271, "step": 8626 }, { "epoch": 0.07759035494417038, "grad_norm": 2.929793119430542, "learning_rate": 6.10133971111394e-06, "loss": 0.5331, "step": 8627 }, { "epoch": 0.07759934884181083, "grad_norm": 2.5462756156921387, "learning_rate": 6.096420489124904e-06, "loss": 0.8249, "step": 8628 }, { "epoch": 0.07760834273945129, "grad_norm": 2.305673837661743, "learning_rate": 6.0915029756210246e-06, "loss": 0.5524, "step": 8629 }, { "epoch": 0.07761733663709174, "grad_norm": 2.3520689010620117, "learning_rate": 6.086587171046731e-06, "loss": 0.3108, "step": 8630 }, { "epoch": 0.07762633053473218, "grad_norm": 19.842628479003906, "learning_rate": 6.081673075846323e-06, "loss": 0.5777, "step": 8631 }, { "epoch": 0.07763532443237263, "grad_norm": 16.0567626953125, "learning_rate": 6.076760690463925e-06, "loss": 0.9555, "step": 8632 }, { "epoch": 0.07764431833001309, "grad_norm": 3.253995895385742, "learning_rate": 6.071850015343517e-06, "loss": 0.5766, "step": 8633 }, { "epoch": 0.07765331222765354, "grad_norm": 3.199002742767334, "learning_rate": 6.066941050928906e-06, "loss": 0.4113, "step": 8634 }, { "epoch": 0.07766230612529398, "grad_norm": 2.0351850986480713, "learning_rate": 6.06203379766378e-06, "loss": 0.3805, "step": 8635 }, { "epoch": 0.07767130002293444, "grad_norm": 2.626953125, "learning_rate": 6.057128255991637e-06, "loss": 0.5248, "step": 8636 }, { "epoch": 0.07768029392057489, "grad_norm": 2.771376371383667, "learning_rate": 6.052224426355835e-06, "loss": 0.3308, "step": 8637 }, { "epoch": 0.07768928781821534, "grad_norm": 2.424006223678589, "learning_rate": 6.047322309199588e-06, "loss": 0.0658, "step": 8638 }, { "epoch": 0.07769828171585579, "grad_norm": 1.8645613193511963, "learning_rate": 6.042421904965926e-06, "loss": 0.2561, "step": 8639 }, { "epoch": 0.07770727561349625, "grad_norm": 9.079527854919434, "learning_rate": 6.037523214097762e-06, "loss": 0.5336, "step": 8640 }, { "epoch": 0.07771626951113669, "grad_norm": 2.4184341430664062, "learning_rate": 6.032626237037825e-06, "loss": 0.6186, "step": 8641 }, { "epoch": 0.07772526340877714, "grad_norm": 2.3993263244628906, "learning_rate": 6.027730974228693e-06, "loss": 0.4027, "step": 8642 }, { "epoch": 0.0777342573064176, "grad_norm": 2.2896649837493896, "learning_rate": 6.022837426112804e-06, "loss": 0.3563, "step": 8643 }, { "epoch": 0.07774325120405805, "grad_norm": 19.735862731933594, "learning_rate": 6.0179455931324236e-06, "loss": 0.4667, "step": 8644 }, { "epoch": 0.0777522451016985, "grad_norm": 4.01049280166626, "learning_rate": 6.013055475729681e-06, "loss": 0.2297, "step": 8645 }, { "epoch": 0.07776123899933896, "grad_norm": 2.139413833618164, "learning_rate": 6.008167074346535e-06, "loss": 0.467, "step": 8646 }, { "epoch": 0.0777702328969794, "grad_norm": 6.244221210479736, "learning_rate": 6.003280389424789e-06, "loss": 0.5951, "step": 8647 }, { "epoch": 0.07777922679461985, "grad_norm": 12.899266242980957, "learning_rate": 5.998395421406103e-06, "loss": 0.4533, "step": 8648 }, { "epoch": 0.0777882206922603, "grad_norm": 13.609940528869629, "learning_rate": 5.993512170731963e-06, "loss": 0.559, "step": 8649 }, { "epoch": 0.07779721458990076, "grad_norm": 4.68431282043457, "learning_rate": 5.9886306378437285e-06, "loss": 0.374, "step": 8650 }, { "epoch": 0.0778062084875412, "grad_norm": 14.071749687194824, "learning_rate": 5.983750823182574e-06, "loss": 0.5796, "step": 8651 }, { "epoch": 0.07781520238518165, "grad_norm": 15.9473876953125, "learning_rate": 5.978872727189544e-06, "loss": 0.7054, "step": 8652 }, { "epoch": 0.07782419628282211, "grad_norm": 2.4189295768737793, "learning_rate": 5.973996350305508e-06, "loss": 0.4207, "step": 8653 }, { "epoch": 0.07783319018046256, "grad_norm": 2.073854446411133, "learning_rate": 5.969121692971183e-06, "loss": 0.7163, "step": 8654 }, { "epoch": 0.077842184078103, "grad_norm": 19.888473510742188, "learning_rate": 5.964248755627147e-06, "loss": 0.4928, "step": 8655 }, { "epoch": 0.07785117797574345, "grad_norm": 2.198866844177246, "learning_rate": 5.959377538713798e-06, "loss": 0.5635, "step": 8656 }, { "epoch": 0.07786017187338391, "grad_norm": 2.8893628120422363, "learning_rate": 5.954508042671403e-06, "loss": 0.2424, "step": 8657 }, { "epoch": 0.07786916577102436, "grad_norm": 2.419403553009033, "learning_rate": 5.9496402679400594e-06, "loss": 0.1721, "step": 8658 }, { "epoch": 0.0778781596686648, "grad_norm": 2.8401055335998535, "learning_rate": 5.944774214959703e-06, "loss": 0.3381, "step": 8659 }, { "epoch": 0.07788715356630527, "grad_norm": 2.514561653137207, "learning_rate": 5.939909884170134e-06, "loss": 0.1414, "step": 8660 }, { "epoch": 0.07789614746394571, "grad_norm": 9.896312713623047, "learning_rate": 5.93504727601098e-06, "loss": 0.7237, "step": 8661 }, { "epoch": 0.07790514136158616, "grad_norm": 2.9159083366394043, "learning_rate": 5.930186390921719e-06, "loss": 0.3811, "step": 8662 }, { "epoch": 0.07791413525922661, "grad_norm": 2.139326333999634, "learning_rate": 5.925327229341662e-06, "loss": 0.0837, "step": 8663 }, { "epoch": 0.07792312915686707, "grad_norm": 1.8465874195098877, "learning_rate": 5.920469791709993e-06, "loss": 0.1769, "step": 8664 }, { "epoch": 0.07793212305450752, "grad_norm": 14.94510555267334, "learning_rate": 5.915614078465714e-06, "loss": 0.4828, "step": 8665 }, { "epoch": 0.07794111695214796, "grad_norm": 3.372276544570923, "learning_rate": 5.91076009004767e-06, "loss": 0.2976, "step": 8666 }, { "epoch": 0.07795011084978842, "grad_norm": 6.905248165130615, "learning_rate": 5.905907826894578e-06, "loss": 0.3263, "step": 8667 }, { "epoch": 0.07795910474742887, "grad_norm": 2.68544602394104, "learning_rate": 5.901057289444961e-06, "loss": 0.4003, "step": 8668 }, { "epoch": 0.07796809864506932, "grad_norm": 1.8799420595169067, "learning_rate": 5.896208478137222e-06, "loss": 0.6286, "step": 8669 }, { "epoch": 0.07797709254270978, "grad_norm": 2.285896062850952, "learning_rate": 5.891361393409584e-06, "loss": 0.4892, "step": 8670 }, { "epoch": 0.07798608644035022, "grad_norm": 10.26109504699707, "learning_rate": 5.886516035700113e-06, "loss": 0.8287, "step": 8671 }, { "epoch": 0.07799508033799067, "grad_norm": 3.21868634223938, "learning_rate": 5.881672405446745e-06, "loss": 0.5885, "step": 8672 }, { "epoch": 0.07800407423563112, "grad_norm": 3.813856601715088, "learning_rate": 5.8768305030872304e-06, "loss": 0.2132, "step": 8673 }, { "epoch": 0.07801306813327158, "grad_norm": 2.1322598457336426, "learning_rate": 5.871990329059177e-06, "loss": 0.5809, "step": 8674 }, { "epoch": 0.07802206203091203, "grad_norm": 4.375406265258789, "learning_rate": 5.867151883800026e-06, "loss": 0.6264, "step": 8675 }, { "epoch": 0.07803105592855247, "grad_norm": 27.577295303344727, "learning_rate": 5.86231516774709e-06, "loss": 0.8534, "step": 8676 }, { "epoch": 0.07804004982619293, "grad_norm": 21.102691650390625, "learning_rate": 5.857480181337494e-06, "loss": 0.3185, "step": 8677 }, { "epoch": 0.07804904372383338, "grad_norm": 2.195747137069702, "learning_rate": 5.852646925008212e-06, "loss": 0.4524, "step": 8678 }, { "epoch": 0.07805803762147383, "grad_norm": 2.1234207153320312, "learning_rate": 5.847815399196088e-06, "loss": 0.2563, "step": 8679 }, { "epoch": 0.07806703151911427, "grad_norm": 2.788010835647583, "learning_rate": 5.842985604337769e-06, "loss": 0.6863, "step": 8680 }, { "epoch": 0.07807602541675474, "grad_norm": 3.339674472808838, "learning_rate": 5.838157540869785e-06, "loss": 0.6034, "step": 8681 }, { "epoch": 0.07808501931439518, "grad_norm": 2.257326364517212, "learning_rate": 5.833331209228482e-06, "loss": 0.1219, "step": 8682 }, { "epoch": 0.07809401321203563, "grad_norm": 2.2681376934051514, "learning_rate": 5.828506609850054e-06, "loss": 0.4623, "step": 8683 }, { "epoch": 0.07810300710967609, "grad_norm": 2.856335401535034, "learning_rate": 5.823683743170555e-06, "loss": 0.085, "step": 8684 }, { "epoch": 0.07811200100731654, "grad_norm": 2.8257346153259277, "learning_rate": 5.818862609625858e-06, "loss": 0.4564, "step": 8685 }, { "epoch": 0.07812099490495698, "grad_norm": 1.948840618133545, "learning_rate": 5.814043209651706e-06, "loss": 0.6917, "step": 8686 }, { "epoch": 0.07812998880259744, "grad_norm": 3.300077438354492, "learning_rate": 5.809225543683661e-06, "loss": 0.27, "step": 8687 }, { "epoch": 0.07813898270023789, "grad_norm": 14.279667854309082, "learning_rate": 5.804409612157141e-06, "loss": 0.6671, "step": 8688 }, { "epoch": 0.07814797659787834, "grad_norm": 13.372077941894531, "learning_rate": 5.799595415507405e-06, "loss": 0.621, "step": 8689 }, { "epoch": 0.07815697049551878, "grad_norm": 2.022092580795288, "learning_rate": 5.79478295416955e-06, "loss": 0.7759, "step": 8690 }, { "epoch": 0.07816596439315925, "grad_norm": 2.5065970420837402, "learning_rate": 5.78997222857853e-06, "loss": 0.6323, "step": 8691 }, { "epoch": 0.07817495829079969, "grad_norm": 2.261521339416504, "learning_rate": 5.785163239169122e-06, "loss": 0.487, "step": 8692 }, { "epoch": 0.07818395218844014, "grad_norm": 24.173044204711914, "learning_rate": 5.780355986375971e-06, "loss": 0.7281, "step": 8693 }, { "epoch": 0.0781929460860806, "grad_norm": 7.8742876052856445, "learning_rate": 5.775550470633542e-06, "loss": 0.7542, "step": 8694 }, { "epoch": 0.07820193998372105, "grad_norm": 2.1035964488983154, "learning_rate": 5.77074669237615e-06, "loss": 0.4585, "step": 8695 }, { "epoch": 0.0782109338813615, "grad_norm": 17.473493576049805, "learning_rate": 5.7659446520379665e-06, "loss": 0.5451, "step": 8696 }, { "epoch": 0.07821992777900194, "grad_norm": 2.5300252437591553, "learning_rate": 5.761144350052983e-06, "loss": 0.6237, "step": 8697 }, { "epoch": 0.0782289216766424, "grad_norm": 2.361579656600952, "learning_rate": 5.756345786855055e-06, "loss": 0.4821, "step": 8698 }, { "epoch": 0.07823791557428285, "grad_norm": 10.816925048828125, "learning_rate": 5.751548962877859e-06, "loss": 0.722, "step": 8699 }, { "epoch": 0.0782469094719233, "grad_norm": 11.43142318725586, "learning_rate": 5.746753878554945e-06, "loss": 0.6723, "step": 8700 }, { "epoch": 0.07825590336956376, "grad_norm": 6.0857696533203125, "learning_rate": 5.741960534319677e-06, "loss": 0.5345, "step": 8701 }, { "epoch": 0.0782648972672042, "grad_norm": 2.1867709159851074, "learning_rate": 5.737168930605272e-06, "loss": 0.2509, "step": 8702 }, { "epoch": 0.07827389116484465, "grad_norm": 2.169151782989502, "learning_rate": 5.7323790678447925e-06, "loss": 0.5381, "step": 8703 }, { "epoch": 0.0782828850624851, "grad_norm": 16.014318466186523, "learning_rate": 5.727590946471132e-06, "loss": 0.422, "step": 8704 }, { "epoch": 0.07829187896012556, "grad_norm": 2.0033771991729736, "learning_rate": 5.722804566917047e-06, "loss": 0.3833, "step": 8705 }, { "epoch": 0.078300872857766, "grad_norm": 2.5681850910186768, "learning_rate": 5.718019929615126e-06, "loss": 0.4864, "step": 8706 }, { "epoch": 0.07830986675540645, "grad_norm": 2.0554540157318115, "learning_rate": 5.7132370349977846e-06, "loss": 0.4513, "step": 8707 }, { "epoch": 0.07831886065304691, "grad_norm": 3.072221279144287, "learning_rate": 5.708455883497316e-06, "loss": 0.6463, "step": 8708 }, { "epoch": 0.07832785455068736, "grad_norm": 21.566171646118164, "learning_rate": 5.703676475545816e-06, "loss": 0.3012, "step": 8709 }, { "epoch": 0.0783368484483278, "grad_norm": 2.5937037467956543, "learning_rate": 5.69889881157526e-06, "loss": 0.4457, "step": 8710 }, { "epoch": 0.07834584234596827, "grad_norm": 2.0509679317474365, "learning_rate": 5.694122892017439e-06, "loss": 0.2122, "step": 8711 }, { "epoch": 0.07835483624360871, "grad_norm": 2.1414542198181152, "learning_rate": 5.689348717303991e-06, "loss": 0.2122, "step": 8712 }, { "epoch": 0.07836383014124916, "grad_norm": 3.0531036853790283, "learning_rate": 5.684576287866411e-06, "loss": 0.1629, "step": 8713 }, { "epoch": 0.07837282403888961, "grad_norm": 27.87185287475586, "learning_rate": 5.6798056041360155e-06, "loss": 0.5549, "step": 8714 }, { "epoch": 0.07838181793653007, "grad_norm": 2.3610658645629883, "learning_rate": 5.675036666543987e-06, "loss": 0.1695, "step": 8715 }, { "epoch": 0.07839081183417052, "grad_norm": 2.2163338661193848, "learning_rate": 5.670269475521331e-06, "loss": 0.5164, "step": 8716 }, { "epoch": 0.07839980573181096, "grad_norm": 8.188508987426758, "learning_rate": 5.6655040314988974e-06, "loss": 0.7792, "step": 8717 }, { "epoch": 0.07840879962945142, "grad_norm": 2.938602924346924, "learning_rate": 5.660740334907383e-06, "loss": 0.4555, "step": 8718 }, { "epoch": 0.07841779352709187, "grad_norm": 3.1793713569641113, "learning_rate": 5.655978386177321e-06, "loss": 0.6037, "step": 8719 }, { "epoch": 0.07842678742473232, "grad_norm": 2.292339324951172, "learning_rate": 5.6512181857391044e-06, "loss": 0.1424, "step": 8720 }, { "epoch": 0.07843578132237276, "grad_norm": 2.6644344329833984, "learning_rate": 5.646459734022938e-06, "loss": 0.5594, "step": 8721 }, { "epoch": 0.07844477522001322, "grad_norm": 17.576213836669922, "learning_rate": 5.641703031458903e-06, "loss": 0.4784, "step": 8722 }, { "epoch": 0.07845376911765367, "grad_norm": 15.48034954071045, "learning_rate": 5.636948078476898e-06, "loss": 0.4362, "step": 8723 }, { "epoch": 0.07846276301529412, "grad_norm": 2.2496631145477295, "learning_rate": 5.632194875506663e-06, "loss": 0.476, "step": 8724 }, { "epoch": 0.07847175691293458, "grad_norm": 3.1441240310668945, "learning_rate": 5.627443422977799e-06, "loss": 0.3387, "step": 8725 }, { "epoch": 0.07848075081057503, "grad_norm": 2.0450809001922607, "learning_rate": 5.622693721319727e-06, "loss": 0.6701, "step": 8726 }, { "epoch": 0.07848974470821547, "grad_norm": 2.4113621711730957, "learning_rate": 5.61794577096173e-06, "loss": 0.5643, "step": 8727 }, { "epoch": 0.07849873860585593, "grad_norm": 2.790735960006714, "learning_rate": 5.613199572332919e-06, "loss": 0.439, "step": 8728 }, { "epoch": 0.07850773250349638, "grad_norm": 2.6012871265411377, "learning_rate": 5.608455125862247e-06, "loss": 0.2534, "step": 8729 }, { "epoch": 0.07851672640113683, "grad_norm": 8.018048286437988, "learning_rate": 5.603712431978509e-06, "loss": 0.6092, "step": 8730 }, { "epoch": 0.07852572029877727, "grad_norm": 16.22110366821289, "learning_rate": 5.598971491110355e-06, "loss": 0.4902, "step": 8731 }, { "epoch": 0.07853471419641773, "grad_norm": 3.6707422733306885, "learning_rate": 5.5942323036862616e-06, "loss": 0.305, "step": 8732 }, { "epoch": 0.07854370809405818, "grad_norm": 12.197420120239258, "learning_rate": 5.5894948701345425e-06, "loss": 0.5955, "step": 8733 }, { "epoch": 0.07855270199169863, "grad_norm": 11.177788734436035, "learning_rate": 5.584759190883379e-06, "loss": 0.615, "step": 8734 }, { "epoch": 0.07856169588933909, "grad_norm": 15.95755386352539, "learning_rate": 5.580025266360764e-06, "loss": 0.7773, "step": 8735 }, { "epoch": 0.07857068978697954, "grad_norm": 2.0720229148864746, "learning_rate": 5.575293096994544e-06, "loss": 0.3922, "step": 8736 }, { "epoch": 0.07857968368461998, "grad_norm": 13.94814395904541, "learning_rate": 5.5705626832124155e-06, "loss": 0.5041, "step": 8737 }, { "epoch": 0.07858867758226043, "grad_norm": 6.381041526794434, "learning_rate": 5.565834025441899e-06, "loss": 0.591, "step": 8738 }, { "epoch": 0.07859767147990089, "grad_norm": 7.2903032302856445, "learning_rate": 5.561107124110379e-06, "loss": 0.8528, "step": 8739 }, { "epoch": 0.07860666537754134, "grad_norm": 17.489423751831055, "learning_rate": 5.556381979645051e-06, "loss": 0.6099, "step": 8740 }, { "epoch": 0.07861565927518178, "grad_norm": 2.387732982635498, "learning_rate": 5.55165859247298e-06, "loss": 0.3745, "step": 8741 }, { "epoch": 0.07862465317282225, "grad_norm": 8.49859619140625, "learning_rate": 5.546936963021063e-06, "loss": 1.0434, "step": 8742 }, { "epoch": 0.07863364707046269, "grad_norm": 3.398775815963745, "learning_rate": 5.542217091716026e-06, "loss": 0.3735, "step": 8743 }, { "epoch": 0.07864264096810314, "grad_norm": 3.6540868282318115, "learning_rate": 5.537498978984451e-06, "loss": 0.5491, "step": 8744 }, { "epoch": 0.0786516348657436, "grad_norm": 2.695204973220825, "learning_rate": 5.53278262525275e-06, "loss": 0.763, "step": 8745 }, { "epoch": 0.07866062876338405, "grad_norm": 3.3248682022094727, "learning_rate": 5.528068030947192e-06, "loss": 0.3004, "step": 8746 }, { "epoch": 0.0786696226610245, "grad_norm": 2.2041213512420654, "learning_rate": 5.523355196493874e-06, "loss": 0.6334, "step": 8747 }, { "epoch": 0.07867861655866494, "grad_norm": 3.9659769535064697, "learning_rate": 5.5186441223187255e-06, "loss": 0.7769, "step": 8748 }, { "epoch": 0.0786876104563054, "grad_norm": 14.180432319641113, "learning_rate": 5.513934808847545e-06, "loss": 0.5934, "step": 8749 }, { "epoch": 0.07869660435394585, "grad_norm": 2.5563855171203613, "learning_rate": 5.5092272565059416e-06, "loss": 0.3985, "step": 8750 }, { "epoch": 0.0787055982515863, "grad_norm": 2.048389434814453, "learning_rate": 5.5045214657193925e-06, "loss": 0.8371, "step": 8751 }, { "epoch": 0.07871459214922676, "grad_norm": 3.1459527015686035, "learning_rate": 5.4998174369131875e-06, "loss": 0.4907, "step": 8752 }, { "epoch": 0.0787235860468672, "grad_norm": 5.419997215270996, "learning_rate": 5.495115170512485e-06, "loss": 0.7983, "step": 8753 }, { "epoch": 0.07873257994450765, "grad_norm": 7.855379104614258, "learning_rate": 5.4904146669422645e-06, "loss": 0.8562, "step": 8754 }, { "epoch": 0.0787415738421481, "grad_norm": 7.787425994873047, "learning_rate": 5.485715926627344e-06, "loss": 0.6191, "step": 8755 }, { "epoch": 0.07875056773978856, "grad_norm": 4.416382312774658, "learning_rate": 5.481018949992409e-06, "loss": 0.4477, "step": 8756 }, { "epoch": 0.078759561637429, "grad_norm": 10.479920387268066, "learning_rate": 5.476323737461955e-06, "loss": 0.8811, "step": 8757 }, { "epoch": 0.07876855553506945, "grad_norm": 9.014201164245605, "learning_rate": 5.471630289460333e-06, "loss": 0.6998, "step": 8758 }, { "epoch": 0.07877754943270991, "grad_norm": 3.824124336242676, "learning_rate": 5.4669386064117315e-06, "loss": 0.2506, "step": 8759 }, { "epoch": 0.07878654333035036, "grad_norm": 9.214349746704102, "learning_rate": 5.462248688740171e-06, "loss": 0.5271, "step": 8760 }, { "epoch": 0.0787955372279908, "grad_norm": 7.941821098327637, "learning_rate": 5.457560536869538e-06, "loss": 0.8641, "step": 8761 }, { "epoch": 0.07880453112563125, "grad_norm": 5.569979190826416, "learning_rate": 5.452874151223528e-06, "loss": 0.6767, "step": 8762 }, { "epoch": 0.07881352502327171, "grad_norm": 5.156221389770508, "learning_rate": 5.448189532225703e-06, "loss": 0.3515, "step": 8763 }, { "epoch": 0.07882251892091216, "grad_norm": 3.2014248371124268, "learning_rate": 5.443506680299442e-06, "loss": 0.5584, "step": 8764 }, { "epoch": 0.07883151281855261, "grad_norm": 1.9857287406921387, "learning_rate": 5.438825595867991e-06, "loss": 0.5156, "step": 8765 }, { "epoch": 0.07884050671619307, "grad_norm": 8.511829376220703, "learning_rate": 5.434146279354413e-06, "loss": 0.9141, "step": 8766 }, { "epoch": 0.07884950061383351, "grad_norm": 2.7598185539245605, "learning_rate": 5.429468731181611e-06, "loss": 0.4808, "step": 8767 }, { "epoch": 0.07885849451147396, "grad_norm": 12.431516647338867, "learning_rate": 5.424792951772353e-06, "loss": 0.4244, "step": 8768 }, { "epoch": 0.07886748840911442, "grad_norm": 3.1539955139160156, "learning_rate": 5.420118941549218e-06, "loss": 0.4031, "step": 8769 }, { "epoch": 0.07887648230675487, "grad_norm": 3.340597629547119, "learning_rate": 5.415446700934649e-06, "loss": 0.5528, "step": 8770 }, { "epoch": 0.07888547620439532, "grad_norm": 3.126983880996704, "learning_rate": 5.410776230350914e-06, "loss": 0.4183, "step": 8771 }, { "epoch": 0.07889447010203576, "grad_norm": 2.597900629043579, "learning_rate": 5.40610753022012e-06, "loss": 0.4065, "step": 8772 }, { "epoch": 0.07890346399967622, "grad_norm": 3.351914167404175, "learning_rate": 5.401440600964225e-06, "loss": 0.3759, "step": 8773 }, { "epoch": 0.07891245789731667, "grad_norm": 2.604290723800659, "learning_rate": 5.396775443005012e-06, "loss": 0.1688, "step": 8774 }, { "epoch": 0.07892145179495712, "grad_norm": 14.023604393005371, "learning_rate": 5.3921120567641255e-06, "loss": 0.5451, "step": 8775 }, { "epoch": 0.07893044569259758, "grad_norm": 5.411497116088867, "learning_rate": 5.387450442663025e-06, "loss": 0.5671, "step": 8776 }, { "epoch": 0.07893943959023803, "grad_norm": 2.8056252002716064, "learning_rate": 5.382790601123036e-06, "loss": 0.5201, "step": 8777 }, { "epoch": 0.07894843348787847, "grad_norm": 3.057980537414551, "learning_rate": 5.378132532565302e-06, "loss": 0.2302, "step": 8778 }, { "epoch": 0.07895742738551892, "grad_norm": 9.765243530273438, "learning_rate": 5.373476237410807e-06, "loss": 0.5759, "step": 8779 }, { "epoch": 0.07896642128315938, "grad_norm": 2.587297201156616, "learning_rate": 5.368821716080399e-06, "loss": 0.3213, "step": 8780 }, { "epoch": 0.07897541518079983, "grad_norm": 2.3547441959381104, "learning_rate": 5.364168968994731e-06, "loss": 0.2786, "step": 8781 }, { "epoch": 0.07898440907844027, "grad_norm": 3.2472124099731445, "learning_rate": 5.359517996574329e-06, "loss": 0.1607, "step": 8782 }, { "epoch": 0.07899340297608073, "grad_norm": 3.4559268951416016, "learning_rate": 5.354868799239534e-06, "loss": 0.4236, "step": 8783 }, { "epoch": 0.07900239687372118, "grad_norm": 22.450702667236328, "learning_rate": 5.350221377410531e-06, "loss": 0.7611, "step": 8784 }, { "epoch": 0.07901139077136163, "grad_norm": 4.387184143066406, "learning_rate": 5.34557573150736e-06, "loss": 0.1059, "step": 8785 }, { "epoch": 0.07902038466900209, "grad_norm": 10.625408172607422, "learning_rate": 5.340931861949888e-06, "loss": 0.6475, "step": 8786 }, { "epoch": 0.07902937856664254, "grad_norm": 2.509444236755371, "learning_rate": 5.336289769157818e-06, "loss": 0.6607, "step": 8787 }, { "epoch": 0.07903837246428298, "grad_norm": 2.7124786376953125, "learning_rate": 5.331649453550696e-06, "loss": 0.5583, "step": 8788 }, { "epoch": 0.07904736636192343, "grad_norm": 2.490349292755127, "learning_rate": 5.327010915547909e-06, "loss": 0.3093, "step": 8789 }, { "epoch": 0.07905636025956389, "grad_norm": 5.968339920043945, "learning_rate": 5.322374155568688e-06, "loss": 0.4502, "step": 8790 }, { "epoch": 0.07906535415720434, "grad_norm": 12.033771514892578, "learning_rate": 5.317739174032091e-06, "loss": 0.4196, "step": 8791 }, { "epoch": 0.07907434805484478, "grad_norm": 2.5643792152404785, "learning_rate": 5.313105971357035e-06, "loss": 0.4905, "step": 8792 }, { "epoch": 0.07908334195248524, "grad_norm": 2.588120222091675, "learning_rate": 5.30847454796225e-06, "loss": 0.3612, "step": 8793 }, { "epoch": 0.07909233585012569, "grad_norm": 2.2810487747192383, "learning_rate": 5.303844904266331e-06, "loss": 0.1832, "step": 8794 }, { "epoch": 0.07910132974776614, "grad_norm": 2.2935726642608643, "learning_rate": 5.2992170406876925e-06, "loss": 0.1702, "step": 8795 }, { "epoch": 0.07911032364540659, "grad_norm": 2.577547073364258, "learning_rate": 5.294590957644593e-06, "loss": 0.6459, "step": 8796 }, { "epoch": 0.07911931754304705, "grad_norm": 3.460430383682251, "learning_rate": 5.289966655555145e-06, "loss": 0.4955, "step": 8797 }, { "epoch": 0.0791283114406875, "grad_norm": 2.590026617050171, "learning_rate": 5.2853441348372805e-06, "loss": 0.4125, "step": 8798 }, { "epoch": 0.07913730533832794, "grad_norm": 2.189363956451416, "learning_rate": 5.2807233959087795e-06, "loss": 0.2375, "step": 8799 }, { "epoch": 0.0791462992359684, "grad_norm": 2.09395694732666, "learning_rate": 5.276104439187257e-06, "loss": 0.5799, "step": 8800 }, { "epoch": 0.07915529313360885, "grad_norm": 2.7425131797790527, "learning_rate": 5.271487265090163e-06, "loss": 0.4146, "step": 8801 }, { "epoch": 0.0791642870312493, "grad_norm": 1.9323197603225708, "learning_rate": 5.266871874034807e-06, "loss": 0.6143, "step": 8802 }, { "epoch": 0.07917328092888974, "grad_norm": 2.734656572341919, "learning_rate": 5.262258266438311e-06, "loss": 0.241, "step": 8803 }, { "epoch": 0.0791822748265302, "grad_norm": 2.6195759773254395, "learning_rate": 5.257646442717659e-06, "loss": 0.5521, "step": 8804 }, { "epoch": 0.07919126872417065, "grad_norm": 3.2786998748779297, "learning_rate": 5.253036403289649e-06, "loss": 0.614, "step": 8805 }, { "epoch": 0.0792002626218111, "grad_norm": 2.4782216548919678, "learning_rate": 5.248428148570947e-06, "loss": 0.4935, "step": 8806 }, { "epoch": 0.07920925651945156, "grad_norm": 2.181596279144287, "learning_rate": 5.243821678978034e-06, "loss": 0.1514, "step": 8807 }, { "epoch": 0.079218250417092, "grad_norm": 2.9279613494873047, "learning_rate": 5.239216994927232e-06, "loss": 0.2462, "step": 8808 }, { "epoch": 0.07922724431473245, "grad_norm": 13.366905212402344, "learning_rate": 5.23461409683472e-06, "loss": 0.8401, "step": 8809 }, { "epoch": 0.07923623821237291, "grad_norm": 2.088473320007324, "learning_rate": 5.230012985116489e-06, "loss": 0.361, "step": 8810 }, { "epoch": 0.07924523211001336, "grad_norm": 2.8704099655151367, "learning_rate": 5.225413660188397e-06, "loss": 0.4477, "step": 8811 }, { "epoch": 0.0792542260076538, "grad_norm": 2.5839004516601562, "learning_rate": 5.220816122466119e-06, "loss": 0.6563, "step": 8812 }, { "epoch": 0.07926321990529425, "grad_norm": 2.2406938076019287, "learning_rate": 5.216220372365177e-06, "loss": 0.2099, "step": 8813 }, { "epoch": 0.07927221380293471, "grad_norm": 2.7712998390197754, "learning_rate": 5.211626410300926e-06, "loss": 0.414, "step": 8814 }, { "epoch": 0.07928120770057516, "grad_norm": 28.289419174194336, "learning_rate": 5.20703423668856e-06, "loss": 0.9036, "step": 8815 }, { "epoch": 0.0792902015982156, "grad_norm": 2.383646249771118, "learning_rate": 5.202443851943126e-06, "loss": 0.2987, "step": 8816 }, { "epoch": 0.07929919549585607, "grad_norm": 2.3161234855651855, "learning_rate": 5.197855256479486e-06, "loss": 0.2957, "step": 8817 }, { "epoch": 0.07930818939349651, "grad_norm": 2.5021378993988037, "learning_rate": 5.193268450712369e-06, "loss": 0.2195, "step": 8818 }, { "epoch": 0.07931718329113696, "grad_norm": 2.044412136077881, "learning_rate": 5.1886834350563126e-06, "loss": 0.3489, "step": 8819 }, { "epoch": 0.07932617718877741, "grad_norm": 1.8845330476760864, "learning_rate": 5.184100209925702e-06, "loss": 0.3543, "step": 8820 }, { "epoch": 0.07933517108641787, "grad_norm": 1.9840916395187378, "learning_rate": 5.179518775734777e-06, "loss": 0.4258, "step": 8821 }, { "epoch": 0.07934416498405832, "grad_norm": 3.2726032733917236, "learning_rate": 5.174939132897588e-06, "loss": 0.4565, "step": 8822 }, { "epoch": 0.07935315888169876, "grad_norm": 2.5899510383605957, "learning_rate": 5.170361281828054e-06, "loss": 0.3798, "step": 8823 }, { "epoch": 0.07936215277933922, "grad_norm": 1.6353586912155151, "learning_rate": 5.165785222939909e-06, "loss": 0.2599, "step": 8824 }, { "epoch": 0.07937114667697967, "grad_norm": 1.9741524457931519, "learning_rate": 5.161210956646725e-06, "loss": 0.2397, "step": 8825 }, { "epoch": 0.07938014057462012, "grad_norm": 2.533841133117676, "learning_rate": 5.156638483361934e-06, "loss": 0.3699, "step": 8826 }, { "epoch": 0.07938913447226058, "grad_norm": 3.8074417114257812, "learning_rate": 5.152067803498783e-06, "loss": 0.6321, "step": 8827 }, { "epoch": 0.07939812836990102, "grad_norm": 5.009308338165283, "learning_rate": 5.147498917470364e-06, "loss": 0.2038, "step": 8828 }, { "epoch": 0.07940712226754147, "grad_norm": 17.50306510925293, "learning_rate": 5.142931825689606e-06, "loss": 0.5525, "step": 8829 }, { "epoch": 0.07941611616518192, "grad_norm": 17.107677459716797, "learning_rate": 5.138366528569285e-06, "loss": 0.5965, "step": 8830 }, { "epoch": 0.07942511006282238, "grad_norm": 19.23002052307129, "learning_rate": 5.133803026522005e-06, "loss": 0.5332, "step": 8831 }, { "epoch": 0.07943410396046283, "grad_norm": 1.9657243490219116, "learning_rate": 5.129241319960201e-06, "loss": 0.6281, "step": 8832 }, { "epoch": 0.07944309785810327, "grad_norm": 2.331686496734619, "learning_rate": 5.124681409296173e-06, "loss": 0.1404, "step": 8833 }, { "epoch": 0.07945209175574373, "grad_norm": 1.627040982246399, "learning_rate": 5.120123294942022e-06, "loss": 0.3068, "step": 8834 }, { "epoch": 0.07946108565338418, "grad_norm": 14.607524871826172, "learning_rate": 5.115566977309724e-06, "loss": 0.961, "step": 8835 }, { "epoch": 0.07947007955102463, "grad_norm": 1.8412764072418213, "learning_rate": 5.111012456811065e-06, "loss": 0.2828, "step": 8836 }, { "epoch": 0.07947907344866507, "grad_norm": 2.5792322158813477, "learning_rate": 5.106459733857669e-06, "loss": 0.5624, "step": 8837 }, { "epoch": 0.07948806734630554, "grad_norm": 8.076607704162598, "learning_rate": 5.1019088088610225e-06, "loss": 0.6205, "step": 8838 }, { "epoch": 0.07949706124394598, "grad_norm": 20.171409606933594, "learning_rate": 5.097359682232416e-06, "loss": 0.5452, "step": 8839 }, { "epoch": 0.07950605514158643, "grad_norm": 3.843249559402466, "learning_rate": 5.092812354383014e-06, "loss": 0.3177, "step": 8840 }, { "epoch": 0.07951504903922689, "grad_norm": 24.530900955200195, "learning_rate": 5.0882668257237866e-06, "loss": 0.5981, "step": 8841 }, { "epoch": 0.07952404293686734, "grad_norm": 1.8272533416748047, "learning_rate": 5.083723096665557e-06, "loss": 0.3638, "step": 8842 }, { "epoch": 0.07953303683450778, "grad_norm": 2.3236474990844727, "learning_rate": 5.079181167618985e-06, "loss": 0.5656, "step": 8843 }, { "epoch": 0.07954203073214823, "grad_norm": 2.5037972927093506, "learning_rate": 5.074641038994552e-06, "loss": 0.5018, "step": 8844 }, { "epoch": 0.07955102462978869, "grad_norm": 1.7665491104125977, "learning_rate": 5.070102711202607e-06, "loss": 0.4271, "step": 8845 }, { "epoch": 0.07956001852742914, "grad_norm": 2.1413958072662354, "learning_rate": 5.065566184653306e-06, "loss": 0.3636, "step": 8846 }, { "epoch": 0.07956901242506959, "grad_norm": 2.020465850830078, "learning_rate": 5.061031459756668e-06, "loss": 0.4394, "step": 8847 }, { "epoch": 0.07957800632271005, "grad_norm": 13.31071662902832, "learning_rate": 5.056498536922533e-06, "loss": 0.4889, "step": 8848 }, { "epoch": 0.07958700022035049, "grad_norm": 36.89935302734375, "learning_rate": 5.0519674165605686e-06, "loss": 0.6335, "step": 8849 }, { "epoch": 0.07959599411799094, "grad_norm": 2.2389895915985107, "learning_rate": 5.047438099080312e-06, "loss": 0.2562, "step": 8850 }, { "epoch": 0.0796049880156314, "grad_norm": 18.970510482788086, "learning_rate": 5.0429105848911e-06, "loss": 0.819, "step": 8851 }, { "epoch": 0.07961398191327185, "grad_norm": 2.1258649826049805, "learning_rate": 5.0383848744021415e-06, "loss": 0.2361, "step": 8852 }, { "epoch": 0.0796229758109123, "grad_norm": 18.134977340698242, "learning_rate": 5.033860968022455e-06, "loss": 0.509, "step": 8853 }, { "epoch": 0.07963196970855274, "grad_norm": 3.0490493774414062, "learning_rate": 5.029338866160912e-06, "loss": 0.5743, "step": 8854 }, { "epoch": 0.0796409636061932, "grad_norm": 12.72742748260498, "learning_rate": 5.024818569226203e-06, "loss": 0.5703, "step": 8855 }, { "epoch": 0.07964995750383365, "grad_norm": 20.08206558227539, "learning_rate": 5.0203000776268825e-06, "loss": 0.7938, "step": 8856 }, { "epoch": 0.0796589514014741, "grad_norm": 4.140419960021973, "learning_rate": 5.0157833917713225e-06, "loss": 0.363, "step": 8857 }, { "epoch": 0.07966794529911456, "grad_norm": 2.5401253700256348, "learning_rate": 5.011268512067727e-06, "loss": 0.4041, "step": 8858 }, { "epoch": 0.079676939196755, "grad_norm": 2.385608434677124, "learning_rate": 5.006755438924157e-06, "loss": 0.5872, "step": 8859 }, { "epoch": 0.07968593309439545, "grad_norm": 3.919597864151001, "learning_rate": 5.002244172748499e-06, "loss": 0.3602, "step": 8860 }, { "epoch": 0.0796949269920359, "grad_norm": 20.672170639038086, "learning_rate": 4.997734713948466e-06, "loss": 1.0562, "step": 8861 }, { "epoch": 0.07970392088967636, "grad_norm": 16.389211654663086, "learning_rate": 4.9932270629316295e-06, "loss": 0.5061, "step": 8862 }, { "epoch": 0.0797129147873168, "grad_norm": 2.2002735137939453, "learning_rate": 4.988721220105375e-06, "loss": 0.3539, "step": 8863 }, { "epoch": 0.07972190868495725, "grad_norm": 3.687103271484375, "learning_rate": 4.984217185876949e-06, "loss": 0.4017, "step": 8864 }, { "epoch": 0.07973090258259771, "grad_norm": 2.5829663276672363, "learning_rate": 4.979714960653417e-06, "loss": 0.5027, "step": 8865 }, { "epoch": 0.07973989648023816, "grad_norm": 2.0373120307922363, "learning_rate": 4.975214544841675e-06, "loss": 0.3071, "step": 8866 }, { "epoch": 0.0797488903778786, "grad_norm": 3.176018714904785, "learning_rate": 4.970715938848478e-06, "loss": 0.3157, "step": 8867 }, { "epoch": 0.07975788427551907, "grad_norm": 3.8163211345672607, "learning_rate": 4.9662191430804026e-06, "loss": 0.5661, "step": 8868 }, { "epoch": 0.07976687817315951, "grad_norm": 2.0761983394622803, "learning_rate": 4.961724157943862e-06, "loss": 0.2474, "step": 8869 }, { "epoch": 0.07977587207079996, "grad_norm": 10.680991172790527, "learning_rate": 4.957230983845102e-06, "loss": 0.6779, "step": 8870 }, { "epoch": 0.07978486596844041, "grad_norm": 12.067887306213379, "learning_rate": 4.952739621190225e-06, "loss": 0.827, "step": 8871 }, { "epoch": 0.07979385986608087, "grad_norm": 2.191086769104004, "learning_rate": 4.948250070385147e-06, "loss": 0.6218, "step": 8872 }, { "epoch": 0.07980285376372132, "grad_norm": 3.4661500453948975, "learning_rate": 4.9437623318356216e-06, "loss": 0.4347, "step": 8873 }, { "epoch": 0.07981184766136176, "grad_norm": 3.439639091491699, "learning_rate": 4.939276405947263e-06, "loss": 0.206, "step": 8874 }, { "epoch": 0.07982084155900222, "grad_norm": 2.4973063468933105, "learning_rate": 4.934792293125487e-06, "loss": 0.2742, "step": 8875 }, { "epoch": 0.07982983545664267, "grad_norm": 2.023829460144043, "learning_rate": 4.930309993775578e-06, "loss": 0.5747, "step": 8876 }, { "epoch": 0.07983882935428312, "grad_norm": 12.470130920410156, "learning_rate": 4.925829508302635e-06, "loss": 0.8627, "step": 8877 }, { "epoch": 0.07984782325192356, "grad_norm": 2.3985342979431152, "learning_rate": 4.9213508371115935e-06, "loss": 0.5394, "step": 8878 }, { "epoch": 0.07985681714956402, "grad_norm": 2.4940054416656494, "learning_rate": 4.916873980607239e-06, "loss": 0.4059, "step": 8879 }, { "epoch": 0.07986581104720447, "grad_norm": 11.600824356079102, "learning_rate": 4.912398939194179e-06, "loss": 0.5378, "step": 8880 }, { "epoch": 0.07987480494484492, "grad_norm": 16.816486358642578, "learning_rate": 4.907925713276873e-06, "loss": 0.6525, "step": 8881 }, { "epoch": 0.07988379884248538, "grad_norm": 2.1641411781311035, "learning_rate": 4.9034543032596e-06, "loss": 0.3648, "step": 8882 }, { "epoch": 0.07989279274012583, "grad_norm": 2.226412057876587, "learning_rate": 4.898984709546478e-06, "loss": 0.6184, "step": 8883 }, { "epoch": 0.07990178663776627, "grad_norm": 1.930140733718872, "learning_rate": 4.894516932541468e-06, "loss": 0.209, "step": 8884 }, { "epoch": 0.07991078053540672, "grad_norm": 11.77936840057373, "learning_rate": 4.890050972648355e-06, "loss": 0.7073, "step": 8885 }, { "epoch": 0.07991977443304718, "grad_norm": 3.313514471054077, "learning_rate": 4.885586830270783e-06, "loss": 0.1857, "step": 8886 }, { "epoch": 0.07992876833068763, "grad_norm": 2.1629884243011475, "learning_rate": 4.881124505812201e-06, "loss": 0.3524, "step": 8887 }, { "epoch": 0.07993776222832807, "grad_norm": 3.123626947402954, "learning_rate": 4.87666399967592e-06, "loss": 0.4209, "step": 8888 }, { "epoch": 0.07994675612596853, "grad_norm": 2.0363497734069824, "learning_rate": 4.872205312265074e-06, "loss": 0.2716, "step": 8889 }, { "epoch": 0.07995575002360898, "grad_norm": 16.72089195251465, "learning_rate": 4.867748443982626e-06, "loss": 0.5558, "step": 8890 }, { "epoch": 0.07996474392124943, "grad_norm": 2.12092924118042, "learning_rate": 4.863293395231391e-06, "loss": 0.521, "step": 8891 }, { "epoch": 0.07997373781888989, "grad_norm": 2.584256649017334, "learning_rate": 4.858840166414008e-06, "loss": 0.2588, "step": 8892 }, { "epoch": 0.07998273171653034, "grad_norm": 20.586660385131836, "learning_rate": 4.85438875793296e-06, "loss": 0.4603, "step": 8893 }, { "epoch": 0.07999172561417078, "grad_norm": 2.1543471813201904, "learning_rate": 4.8499391701905514e-06, "loss": 0.4122, "step": 8894 }, { "epoch": 0.08000071951181123, "grad_norm": 10.75645923614502, "learning_rate": 4.845491403588944e-06, "loss": 0.413, "step": 8895 }, { "epoch": 0.08000971340945169, "grad_norm": 1.8475173711776733, "learning_rate": 4.841045458530113e-06, "loss": 0.2924, "step": 8896 }, { "epoch": 0.08001870730709214, "grad_norm": 2.467892646789551, "learning_rate": 4.836601335415883e-06, "loss": 0.4969, "step": 8897 }, { "epoch": 0.08002770120473258, "grad_norm": 6.826835632324219, "learning_rate": 4.832159034647904e-06, "loss": 0.7285, "step": 8898 }, { "epoch": 0.08003669510237305, "grad_norm": 2.397552967071533, "learning_rate": 4.827718556627661e-06, "loss": 0.494, "step": 8899 }, { "epoch": 0.08004568900001349, "grad_norm": 2.204367160797119, "learning_rate": 4.823279901756497e-06, "loss": 0.4207, "step": 8900 }, { "epoch": 0.08005468289765394, "grad_norm": 5.1364426612854, "learning_rate": 4.8188430704355605e-06, "loss": 0.6713, "step": 8901 }, { "epoch": 0.08006367679529439, "grad_norm": 17.49321746826172, "learning_rate": 4.8144080630658426e-06, "loss": 0.5764, "step": 8902 }, { "epoch": 0.08007267069293485, "grad_norm": 18.444746017456055, "learning_rate": 4.80997488004819e-06, "loss": 0.9354, "step": 8903 }, { "epoch": 0.0800816645905753, "grad_norm": 20.839223861694336, "learning_rate": 4.805543521783254e-06, "loss": 0.2956, "step": 8904 }, { "epoch": 0.08009065848821574, "grad_norm": 3.1189846992492676, "learning_rate": 4.801113988671549e-06, "loss": 0.7829, "step": 8905 }, { "epoch": 0.0800996523858562, "grad_norm": 9.420787811279297, "learning_rate": 4.7966862811134e-06, "loss": 0.5272, "step": 8906 }, { "epoch": 0.08010864628349665, "grad_norm": 6.454324722290039, "learning_rate": 4.792260399508988e-06, "loss": 0.7963, "step": 8907 }, { "epoch": 0.0801176401811371, "grad_norm": 2.279379367828369, "learning_rate": 4.787836344258314e-06, "loss": 0.1689, "step": 8908 }, { "epoch": 0.08012663407877756, "grad_norm": 2.2387571334838867, "learning_rate": 4.783414115761223e-06, "loss": 0.5097, "step": 8909 }, { "epoch": 0.080135627976418, "grad_norm": 3.8532278537750244, "learning_rate": 4.778993714417382e-06, "loss": 0.3432, "step": 8910 }, { "epoch": 0.08014462187405845, "grad_norm": 2.60074520111084, "learning_rate": 4.7745751406263165e-06, "loss": 0.6059, "step": 8911 }, { "epoch": 0.0801536157716989, "grad_norm": 16.335256576538086, "learning_rate": 4.770158394787364e-06, "loss": 0.4389, "step": 8912 }, { "epoch": 0.08016260966933936, "grad_norm": 2.697990894317627, "learning_rate": 4.765743477299706e-06, "loss": 0.2662, "step": 8913 }, { "epoch": 0.0801716035669798, "grad_norm": 10.952118873596191, "learning_rate": 4.761330388562349e-06, "loss": 0.4339, "step": 8914 }, { "epoch": 0.08018059746462025, "grad_norm": 10.12090015411377, "learning_rate": 4.756919128974161e-06, "loss": 0.5435, "step": 8915 }, { "epoch": 0.08018959136226071, "grad_norm": 3.0190677642822266, "learning_rate": 4.752509698933811e-06, "loss": 0.6518, "step": 8916 }, { "epoch": 0.08019858525990116, "grad_norm": 2.509678363800049, "learning_rate": 4.7481020988398324e-06, "loss": 0.3763, "step": 8917 }, { "epoch": 0.0802075791575416, "grad_norm": 2.403158187866211, "learning_rate": 4.743696329090566e-06, "loss": 0.528, "step": 8918 }, { "epoch": 0.08021657305518205, "grad_norm": 14.297457695007324, "learning_rate": 4.739292390084213e-06, "loss": 0.9017, "step": 8919 }, { "epoch": 0.08022556695282251, "grad_norm": 3.13204288482666, "learning_rate": 4.73489028221879e-06, "loss": 0.4084, "step": 8920 }, { "epoch": 0.08023456085046296, "grad_norm": 2.122828245162964, "learning_rate": 4.730490005892152e-06, "loss": 0.2208, "step": 8921 }, { "epoch": 0.08024355474810341, "grad_norm": 5.310396671295166, "learning_rate": 4.726091561502e-06, "loss": 0.8207, "step": 8922 }, { "epoch": 0.08025254864574387, "grad_norm": 16.45570945739746, "learning_rate": 4.721694949445854e-06, "loss": 0.558, "step": 8923 }, { "epoch": 0.08026154254338431, "grad_norm": 3.0207180976867676, "learning_rate": 4.7173001701210775e-06, "loss": 0.2568, "step": 8924 }, { "epoch": 0.08027053644102476, "grad_norm": 2.0149803161621094, "learning_rate": 4.712907223924864e-06, "loss": 0.2918, "step": 8925 }, { "epoch": 0.08027953033866522, "grad_norm": 11.348580360412598, "learning_rate": 4.708516111254238e-06, "loss": 0.4909, "step": 8926 }, { "epoch": 0.08028852423630567, "grad_norm": 6.8168721199035645, "learning_rate": 4.704126832506078e-06, "loss": 0.5794, "step": 8927 }, { "epoch": 0.08029751813394612, "grad_norm": 2.334338426589966, "learning_rate": 4.699739388077065e-06, "loss": 0.6725, "step": 8928 }, { "epoch": 0.08030651203158656, "grad_norm": 8.8694429397583, "learning_rate": 4.695353778363748e-06, "loss": 0.5872, "step": 8929 }, { "epoch": 0.08031550592922702, "grad_norm": 15.095367431640625, "learning_rate": 4.690970003762487e-06, "loss": 0.3098, "step": 8930 }, { "epoch": 0.08032449982686747, "grad_norm": 4.005987167358398, "learning_rate": 4.686588064669475e-06, "loss": 0.4075, "step": 8931 }, { "epoch": 0.08033349372450792, "grad_norm": 2.3128013610839844, "learning_rate": 4.682207961480758e-06, "loss": 0.345, "step": 8932 }, { "epoch": 0.08034248762214838, "grad_norm": 15.118335723876953, "learning_rate": 4.677829694592198e-06, "loss": 0.4611, "step": 8933 }, { "epoch": 0.08035148151978883, "grad_norm": 2.7541911602020264, "learning_rate": 4.673453264399505e-06, "loss": 0.3255, "step": 8934 }, { "epoch": 0.08036047541742927, "grad_norm": 1.8807973861694336, "learning_rate": 4.669078671298207e-06, "loss": 0.5143, "step": 8935 }, { "epoch": 0.08036946931506972, "grad_norm": 5.419625282287598, "learning_rate": 4.664705915683687e-06, "loss": 0.4055, "step": 8936 }, { "epoch": 0.08037846321271018, "grad_norm": 3.764873743057251, "learning_rate": 4.660334997951141e-06, "loss": 0.3474, "step": 8937 }, { "epoch": 0.08038745711035063, "grad_norm": 3.5351150035858154, "learning_rate": 4.655965918495611e-06, "loss": 0.4547, "step": 8938 }, { "epoch": 0.08039645100799107, "grad_norm": 9.871148109436035, "learning_rate": 4.651598677711968e-06, "loss": 0.3272, "step": 8939 }, { "epoch": 0.08040544490563153, "grad_norm": 2.1882691383361816, "learning_rate": 4.647233275994914e-06, "loss": 0.5453, "step": 8940 }, { "epoch": 0.08041443880327198, "grad_norm": 15.966837882995605, "learning_rate": 4.642869713738998e-06, "loss": 0.4602, "step": 8941 }, { "epoch": 0.08042343270091243, "grad_norm": 3.1917788982391357, "learning_rate": 4.638507991338592e-06, "loss": 0.2255, "step": 8942 }, { "epoch": 0.08043242659855288, "grad_norm": 2.2850184440612793, "learning_rate": 4.634148109187897e-06, "loss": 0.4913, "step": 8943 }, { "epoch": 0.08044142049619334, "grad_norm": 3.560523748397827, "learning_rate": 4.629790067680964e-06, "loss": 0.3354, "step": 8944 }, { "epoch": 0.08045041439383378, "grad_norm": 1.8460215330123901, "learning_rate": 4.6254338672116565e-06, "loss": 0.5228, "step": 8945 }, { "epoch": 0.08045940829147423, "grad_norm": 2.1840078830718994, "learning_rate": 4.621079508173698e-06, "loss": 0.6208, "step": 8946 }, { "epoch": 0.08046840218911469, "grad_norm": 6.9107208251953125, "learning_rate": 4.6167269909606184e-06, "loss": 0.391, "step": 8947 }, { "epoch": 0.08047739608675514, "grad_norm": 3.5338592529296875, "learning_rate": 4.612376315965802e-06, "loss": 0.266, "step": 8948 }, { "epoch": 0.08048638998439558, "grad_norm": 2.172254800796509, "learning_rate": 4.6080274835824584e-06, "loss": 0.4573, "step": 8949 }, { "epoch": 0.08049538388203605, "grad_norm": 10.115431785583496, "learning_rate": 4.603680494203616e-06, "loss": 0.7453, "step": 8950 }, { "epoch": 0.08050437777967649, "grad_norm": 1.9423998594284058, "learning_rate": 4.59933534822217e-06, "loss": 0.538, "step": 8951 }, { "epoch": 0.08051337167731694, "grad_norm": 2.639410972595215, "learning_rate": 4.594992046030819e-06, "loss": 0.3979, "step": 8952 }, { "epoch": 0.08052236557495739, "grad_norm": 3.1448888778686523, "learning_rate": 4.590650588022111e-06, "loss": 0.4368, "step": 8953 }, { "epoch": 0.08053135947259785, "grad_norm": 14.889731407165527, "learning_rate": 4.5863109745884175e-06, "loss": 0.6304, "step": 8954 }, { "epoch": 0.0805403533702383, "grad_norm": 2.908590316772461, "learning_rate": 4.581973206121948e-06, "loss": 0.1008, "step": 8955 }, { "epoch": 0.08054934726787874, "grad_norm": 13.374216079711914, "learning_rate": 4.577637283014749e-06, "loss": 0.5187, "step": 8956 }, { "epoch": 0.0805583411655192, "grad_norm": 2.772568464279175, "learning_rate": 4.573303205658691e-06, "loss": 0.4039, "step": 8957 }, { "epoch": 0.08056733506315965, "grad_norm": 2.4788992404937744, "learning_rate": 4.568970974445494e-06, "loss": 0.2469, "step": 8958 }, { "epoch": 0.0805763289608001, "grad_norm": 2.7696681022644043, "learning_rate": 4.5646405897666885e-06, "loss": 0.3401, "step": 8959 }, { "epoch": 0.08058532285844054, "grad_norm": 2.395367383956909, "learning_rate": 4.5603120520136595e-06, "loss": 0.4378, "step": 8960 }, { "epoch": 0.080594316756081, "grad_norm": 5.301259517669678, "learning_rate": 4.555985361577614e-06, "loss": 0.3192, "step": 8961 }, { "epoch": 0.08060331065372145, "grad_norm": 13.287841796875, "learning_rate": 4.551660518849582e-06, "loss": 0.3959, "step": 8962 }, { "epoch": 0.0806123045513619, "grad_norm": 2.4400134086608887, "learning_rate": 4.5473375242204545e-06, "loss": 0.5593, "step": 8963 }, { "epoch": 0.08062129844900236, "grad_norm": 2.6977083683013916, "learning_rate": 4.543016378080925e-06, "loss": 0.378, "step": 8964 }, { "epoch": 0.0806302923466428, "grad_norm": 2.1807363033294678, "learning_rate": 4.53869708082155e-06, "loss": 0.4773, "step": 8965 }, { "epoch": 0.08063928624428325, "grad_norm": 2.3806545734405518, "learning_rate": 4.534379632832692e-06, "loss": 0.5761, "step": 8966 }, { "epoch": 0.08064828014192371, "grad_norm": 16.53265380859375, "learning_rate": 4.530064034504558e-06, "loss": 0.24, "step": 8967 }, { "epoch": 0.08065727403956416, "grad_norm": 3.052042007446289, "learning_rate": 4.525750286227187e-06, "loss": 0.3052, "step": 8968 }, { "epoch": 0.0806662679372046, "grad_norm": 2.041405200958252, "learning_rate": 4.521438388390448e-06, "loss": 0.4812, "step": 8969 }, { "epoch": 0.08067526183484505, "grad_norm": 2.87654972076416, "learning_rate": 4.5171283413840555e-06, "loss": 0.3006, "step": 8970 }, { "epoch": 0.08068425573248551, "grad_norm": 3.0747203826904297, "learning_rate": 4.512820145597535e-06, "loss": 0.7083, "step": 8971 }, { "epoch": 0.08069324963012596, "grad_norm": 2.7605905532836914, "learning_rate": 4.5085138014202695e-06, "loss": 0.6404, "step": 8972 }, { "epoch": 0.0807022435277664, "grad_norm": 2.0310094356536865, "learning_rate": 4.5042093092414525e-06, "loss": 0.1686, "step": 8973 }, { "epoch": 0.08071123742540687, "grad_norm": 10.061881065368652, "learning_rate": 4.499906669450119e-06, "loss": 0.9341, "step": 8974 }, { "epoch": 0.08072023132304731, "grad_norm": 16.75160789489746, "learning_rate": 4.495605882435144e-06, "loss": 0.548, "step": 8975 }, { "epoch": 0.08072922522068776, "grad_norm": 3.645458936691284, "learning_rate": 4.491306948585219e-06, "loss": 0.4157, "step": 8976 }, { "epoch": 0.08073821911832821, "grad_norm": 2.3686017990112305, "learning_rate": 4.487009868288888e-06, "loss": 0.5635, "step": 8977 }, { "epoch": 0.08074721301596867, "grad_norm": 8.279006004333496, "learning_rate": 4.48271464193451e-06, "loss": 0.5064, "step": 8978 }, { "epoch": 0.08075620691360912, "grad_norm": 2.2112388610839844, "learning_rate": 4.478421269910283e-06, "loss": 0.3482, "step": 8979 }, { "epoch": 0.08076520081124956, "grad_norm": 3.061876058578491, "learning_rate": 4.4741297526042365e-06, "loss": 0.3184, "step": 8980 }, { "epoch": 0.08077419470889002, "grad_norm": 2.2360568046569824, "learning_rate": 4.469840090404229e-06, "loss": 0.6595, "step": 8981 }, { "epoch": 0.08078318860653047, "grad_norm": 4.248483657836914, "learning_rate": 4.465552283697969e-06, "loss": 0.3563, "step": 8982 }, { "epoch": 0.08079218250417092, "grad_norm": 2.526104211807251, "learning_rate": 4.461266332872965e-06, "loss": 0.2634, "step": 8983 }, { "epoch": 0.08080117640181136, "grad_norm": 12.839669227600098, "learning_rate": 4.456982238316598e-06, "loss": 0.4138, "step": 8984 }, { "epoch": 0.08081017029945183, "grad_norm": 2.7711856365203857, "learning_rate": 4.452700000416046e-06, "loss": 0.2025, "step": 8985 }, { "epoch": 0.08081916419709227, "grad_norm": 14.222490310668945, "learning_rate": 4.448419619558333e-06, "loss": 0.3298, "step": 8986 }, { "epoch": 0.08082815809473272, "grad_norm": 1.7590553760528564, "learning_rate": 4.444141096130322e-06, "loss": 0.1934, "step": 8987 }, { "epoch": 0.08083715199237318, "grad_norm": 17.47429847717285, "learning_rate": 4.439864430518692e-06, "loss": 0.6286, "step": 8988 }, { "epoch": 0.08084614589001363, "grad_norm": 2.4971866607666016, "learning_rate": 4.435589623109976e-06, "loss": 0.2922, "step": 8989 }, { "epoch": 0.08085513978765407, "grad_norm": 1.9885319471359253, "learning_rate": 4.43131667429052e-06, "loss": 0.3254, "step": 8990 }, { "epoch": 0.08086413368529453, "grad_norm": 2.167391061782837, "learning_rate": 4.427045584446501e-06, "loss": 0.3401, "step": 8991 }, { "epoch": 0.08087312758293498, "grad_norm": 2.4769771099090576, "learning_rate": 4.422776353963951e-06, "loss": 0.4433, "step": 8992 }, { "epoch": 0.08088212148057543, "grad_norm": 2.0252585411071777, "learning_rate": 4.4185089832287084e-06, "loss": 0.8117, "step": 8993 }, { "epoch": 0.08089111537821587, "grad_norm": 16.92536735534668, "learning_rate": 4.4142434726264556e-06, "loss": 0.7901, "step": 8994 }, { "epoch": 0.08090010927585634, "grad_norm": 2.8722572326660156, "learning_rate": 4.409979822542701e-06, "loss": 0.173, "step": 8995 }, { "epoch": 0.08090910317349678, "grad_norm": 20.80576515197754, "learning_rate": 4.405718033362797e-06, "loss": 0.9303, "step": 8996 }, { "epoch": 0.08091809707113723, "grad_norm": 1.9689501523971558, "learning_rate": 4.401458105471912e-06, "loss": 0.588, "step": 8997 }, { "epoch": 0.08092709096877769, "grad_norm": 8.156305313110352, "learning_rate": 4.397200039255053e-06, "loss": 0.5461, "step": 8998 }, { "epoch": 0.08093608486641814, "grad_norm": 2.213296890258789, "learning_rate": 4.392943835097069e-06, "loss": 0.4966, "step": 8999 }, { "epoch": 0.08094507876405858, "grad_norm": 3.0512382984161377, "learning_rate": 4.3886894933826165e-06, "loss": 0.3496, "step": 9000 }, { "epoch": 0.08095407266169903, "grad_norm": 2.4487929344177246, "learning_rate": 4.384437014496215e-06, "loss": 0.6442, "step": 9001 }, { "epoch": 0.08096306655933949, "grad_norm": 2.60485577583313, "learning_rate": 4.380186398822192e-06, "loss": 0.2979, "step": 9002 }, { "epoch": 0.08097206045697994, "grad_norm": 17.486248016357422, "learning_rate": 4.375937646744704e-06, "loss": 0.4357, "step": 9003 }, { "epoch": 0.08098105435462039, "grad_norm": 4.298687934875488, "learning_rate": 4.371690758647765e-06, "loss": 0.6975, "step": 9004 }, { "epoch": 0.08099004825226085, "grad_norm": 2.491285562515259, "learning_rate": 4.367445734915188e-06, "loss": 0.5802, "step": 9005 }, { "epoch": 0.08099904214990129, "grad_norm": 2.42629075050354, "learning_rate": 4.36320257593065e-06, "loss": 0.306, "step": 9006 }, { "epoch": 0.08100803604754174, "grad_norm": 2.205735921859741, "learning_rate": 4.358961282077634e-06, "loss": 0.2582, "step": 9007 }, { "epoch": 0.0810170299451822, "grad_norm": 3.2575342655181885, "learning_rate": 4.354721853739466e-06, "loss": 0.1979, "step": 9008 }, { "epoch": 0.08102602384282265, "grad_norm": 2.096877336502075, "learning_rate": 4.350484291299297e-06, "loss": 0.3808, "step": 9009 }, { "epoch": 0.0810350177404631, "grad_norm": 2.0723259449005127, "learning_rate": 4.346248595140112e-06, "loss": 0.3663, "step": 9010 }, { "epoch": 0.08104401163810354, "grad_norm": 1.881786823272705, "learning_rate": 4.342014765644739e-06, "loss": 0.3651, "step": 9011 }, { "epoch": 0.081053005535744, "grad_norm": 13.191213607788086, "learning_rate": 4.337782803195814e-06, "loss": 0.7965, "step": 9012 }, { "epoch": 0.08106199943338445, "grad_norm": 2.1202831268310547, "learning_rate": 4.333552708175831e-06, "loss": 0.4036, "step": 9013 }, { "epoch": 0.0810709933310249, "grad_norm": 15.038684844970703, "learning_rate": 4.329324480967092e-06, "loss": 0.5304, "step": 9014 }, { "epoch": 0.08107998722866536, "grad_norm": 2.372816801071167, "learning_rate": 4.325098121951737e-06, "loss": 0.5688, "step": 9015 }, { "epoch": 0.0810889811263058, "grad_norm": 2.281562089920044, "learning_rate": 4.320873631511749e-06, "loss": 0.6123, "step": 9016 }, { "epoch": 0.08109797502394625, "grad_norm": 3.754397392272949, "learning_rate": 4.316651010028924e-06, "loss": 0.5304, "step": 9017 }, { "epoch": 0.0811069689215867, "grad_norm": 2.710911273956299, "learning_rate": 4.312430257884908e-06, "loss": 0.4529, "step": 9018 }, { "epoch": 0.08111596281922716, "grad_norm": 27.56716537475586, "learning_rate": 4.308211375461163e-06, "loss": 0.6311, "step": 9019 }, { "epoch": 0.0811249567168676, "grad_norm": 1.7390607595443726, "learning_rate": 4.3039943631389776e-06, "loss": 0.3094, "step": 9020 }, { "epoch": 0.08113395061450805, "grad_norm": 2.763099431991577, "learning_rate": 4.299779221299499e-06, "loss": 0.5362, "step": 9021 }, { "epoch": 0.08114294451214851, "grad_norm": 3.889509439468384, "learning_rate": 4.295565950323677e-06, "loss": 0.4894, "step": 9022 }, { "epoch": 0.08115193840978896, "grad_norm": 2.316570997238159, "learning_rate": 4.291354550592303e-06, "loss": 0.5031, "step": 9023 }, { "epoch": 0.0811609323074294, "grad_norm": 11.130158424377441, "learning_rate": 4.2871450224859945e-06, "loss": 0.8064, "step": 9024 }, { "epoch": 0.08116992620506985, "grad_norm": 1.7977807521820068, "learning_rate": 4.282937366385215e-06, "loss": 0.2435, "step": 9025 }, { "epoch": 0.08117892010271031, "grad_norm": 17.052837371826172, "learning_rate": 4.278731582670239e-06, "loss": 0.3436, "step": 9026 }, { "epoch": 0.08118791400035076, "grad_norm": 2.4594645500183105, "learning_rate": 4.274527671721179e-06, "loss": 0.5077, "step": 9027 }, { "epoch": 0.08119690789799121, "grad_norm": 2.3307888507843018, "learning_rate": 4.27032563391799e-06, "loss": 0.4078, "step": 9028 }, { "epoch": 0.08120590179563167, "grad_norm": 3.6115989685058594, "learning_rate": 4.266125469640439e-06, "loss": 0.4327, "step": 9029 }, { "epoch": 0.08121489569327212, "grad_norm": 2.131692886352539, "learning_rate": 4.261927179268138e-06, "loss": 0.2781, "step": 9030 }, { "epoch": 0.08122388959091256, "grad_norm": 2.219752073287964, "learning_rate": 4.257730763180523e-06, "loss": 0.2047, "step": 9031 }, { "epoch": 0.08123288348855302, "grad_norm": 19.357563018798828, "learning_rate": 4.253536221756851e-06, "loss": 0.4769, "step": 9032 }, { "epoch": 0.08124187738619347, "grad_norm": 11.460504531860352, "learning_rate": 4.249343555376239e-06, "loss": 0.5135, "step": 9033 }, { "epoch": 0.08125087128383392, "grad_norm": 2.3526668548583984, "learning_rate": 4.245152764417604e-06, "loss": 0.4383, "step": 9034 }, { "epoch": 0.08125986518147436, "grad_norm": 3.387362241744995, "learning_rate": 4.2409638492597e-06, "loss": 0.3517, "step": 9035 }, { "epoch": 0.08126885907911482, "grad_norm": 2.6866440773010254, "learning_rate": 4.236776810281132e-06, "loss": 0.2232, "step": 9036 }, { "epoch": 0.08127785297675527, "grad_norm": 2.387037754058838, "learning_rate": 4.232591647860309e-06, "loss": 0.4146, "step": 9037 }, { "epoch": 0.08128684687439572, "grad_norm": 2.6158275604248047, "learning_rate": 4.228408362375485e-06, "loss": 0.4718, "step": 9038 }, { "epoch": 0.08129584077203618, "grad_norm": 13.73037338256836, "learning_rate": 4.224226954204732e-06, "loss": 0.3309, "step": 9039 }, { "epoch": 0.08130483466967663, "grad_norm": 3.1018266677856445, "learning_rate": 4.220047423725978e-06, "loss": 0.422, "step": 9040 }, { "epoch": 0.08131382856731707, "grad_norm": 2.1801998615264893, "learning_rate": 4.215869771316949e-06, "loss": 0.5481, "step": 9041 }, { "epoch": 0.08132282246495752, "grad_norm": 15.600550651550293, "learning_rate": 4.211693997355229e-06, "loss": 0.7026, "step": 9042 }, { "epoch": 0.08133181636259798, "grad_norm": 9.73424243927002, "learning_rate": 4.207520102218213e-06, "loss": 0.6503, "step": 9043 }, { "epoch": 0.08134081026023843, "grad_norm": 2.1408605575561523, "learning_rate": 4.2033480862831294e-06, "loss": 0.2707, "step": 9044 }, { "epoch": 0.08134980415787887, "grad_norm": 2.40228271484375, "learning_rate": 4.199177949927052e-06, "loss": 0.3617, "step": 9045 }, { "epoch": 0.08135879805551934, "grad_norm": 2.4796602725982666, "learning_rate": 4.195009693526858e-06, "loss": 0.3061, "step": 9046 }, { "epoch": 0.08136779195315978, "grad_norm": 18.645360946655273, "learning_rate": 4.190843317459287e-06, "loss": 0.5961, "step": 9047 }, { "epoch": 0.08137678585080023, "grad_norm": 3.5053205490112305, "learning_rate": 4.186678822100884e-06, "loss": 0.3253, "step": 9048 }, { "epoch": 0.08138577974844069, "grad_norm": 14.237509727478027, "learning_rate": 4.18251620782803e-06, "loss": 0.7066, "step": 9049 }, { "epoch": 0.08139477364608114, "grad_norm": 2.2247695922851562, "learning_rate": 4.178355475016937e-06, "loss": 0.4948, "step": 9050 }, { "epoch": 0.08140376754372158, "grad_norm": 2.210862636566162, "learning_rate": 4.1741966240436445e-06, "loss": 0.8722, "step": 9051 }, { "epoch": 0.08141276144136203, "grad_norm": 2.788623332977295, "learning_rate": 4.170039655284033e-06, "loss": 0.7408, "step": 9052 }, { "epoch": 0.08142175533900249, "grad_norm": 1.6668742895126343, "learning_rate": 4.1658845691137985e-06, "loss": 0.6355, "step": 9053 }, { "epoch": 0.08143074923664294, "grad_norm": 2.7795987129211426, "learning_rate": 4.161731365908481e-06, "loss": 0.3572, "step": 9054 }, { "epoch": 0.08143974313428338, "grad_norm": 12.985000610351562, "learning_rate": 4.1575800460434355e-06, "loss": 1.0041, "step": 9055 }, { "epoch": 0.08144873703192385, "grad_norm": 2.049252986907959, "learning_rate": 4.153430609893852e-06, "loss": 0.3906, "step": 9056 }, { "epoch": 0.08145773092956429, "grad_norm": 2.153320550918579, "learning_rate": 4.14928305783476e-06, "loss": 0.4429, "step": 9057 }, { "epoch": 0.08146672482720474, "grad_norm": 2.559793710708618, "learning_rate": 4.145137390241003e-06, "loss": 0.1767, "step": 9058 }, { "epoch": 0.08147571872484519, "grad_norm": 25.920860290527344, "learning_rate": 4.14099360748727e-06, "loss": 0.8662, "step": 9059 }, { "epoch": 0.08148471262248565, "grad_norm": 2.972421646118164, "learning_rate": 4.136851709948061e-06, "loss": 0.5889, "step": 9060 }, { "epoch": 0.0814937065201261, "grad_norm": 2.2121500968933105, "learning_rate": 4.132711697997732e-06, "loss": 0.5245, "step": 9061 }, { "epoch": 0.08150270041776654, "grad_norm": 2.09810471534729, "learning_rate": 4.128573572010439e-06, "loss": 0.6812, "step": 9062 }, { "epoch": 0.081511694315407, "grad_norm": 14.111913681030273, "learning_rate": 4.124437332360187e-06, "loss": 0.6652, "step": 9063 }, { "epoch": 0.08152068821304745, "grad_norm": 1.6867791414260864, "learning_rate": 4.120302979420806e-06, "loss": 0.2492, "step": 9064 }, { "epoch": 0.0815296821106879, "grad_norm": 20.642337799072266, "learning_rate": 4.116170513565942e-06, "loss": 0.4665, "step": 9065 }, { "epoch": 0.08153867600832834, "grad_norm": 24.799428939819336, "learning_rate": 4.112039935169101e-06, "loss": 0.3826, "step": 9066 }, { "epoch": 0.0815476699059688, "grad_norm": 5.830845832824707, "learning_rate": 4.107911244603593e-06, "loss": 0.2299, "step": 9067 }, { "epoch": 0.08155666380360925, "grad_norm": 37.57333755493164, "learning_rate": 4.103784442242556e-06, "loss": 0.4714, "step": 9068 }, { "epoch": 0.0815656577012497, "grad_norm": 2.464226007461548, "learning_rate": 4.09965952845898e-06, "loss": 0.4063, "step": 9069 }, { "epoch": 0.08157465159889016, "grad_norm": 3.17732310295105, "learning_rate": 4.0955365036256575e-06, "loss": 0.375, "step": 9070 }, { "epoch": 0.0815836454965306, "grad_norm": 2.188575267791748, "learning_rate": 4.0914153681152365e-06, "loss": 0.2968, "step": 9071 }, { "epoch": 0.08159263939417105, "grad_norm": 18.24871826171875, "learning_rate": 4.0872961223001715e-06, "loss": 0.4042, "step": 9072 }, { "epoch": 0.08160163329181151, "grad_norm": 13.566888809204102, "learning_rate": 4.083178766552753e-06, "loss": 0.7581, "step": 9073 }, { "epoch": 0.08161062718945196, "grad_norm": 3.124279499053955, "learning_rate": 4.079063301245115e-06, "loss": 0.162, "step": 9074 }, { "epoch": 0.0816196210870924, "grad_norm": 3.2211811542510986, "learning_rate": 4.074949726749195e-06, "loss": 0.4436, "step": 9075 }, { "epoch": 0.08162861498473285, "grad_norm": 3.612926721572876, "learning_rate": 4.070838043436786e-06, "loss": 0.5338, "step": 9076 }, { "epoch": 0.08163760888237331, "grad_norm": 11.028691291809082, "learning_rate": 4.066728251679491e-06, "loss": 0.917, "step": 9077 }, { "epoch": 0.08164660278001376, "grad_norm": 16.629621505737305, "learning_rate": 4.062620351848748e-06, "loss": 0.5055, "step": 9078 }, { "epoch": 0.08165559667765421, "grad_norm": 6.275465488433838, "learning_rate": 4.058514344315828e-06, "loss": 0.5055, "step": 9079 }, { "epoch": 0.08166459057529467, "grad_norm": 33.06439971923828, "learning_rate": 4.054410229451819e-06, "loss": 0.652, "step": 9080 }, { "epoch": 0.08167358447293512, "grad_norm": 2.364936351776123, "learning_rate": 4.050308007627656e-06, "loss": 0.7626, "step": 9081 }, { "epoch": 0.08168257837057556, "grad_norm": 1.8892992734909058, "learning_rate": 4.046207679214087e-06, "loss": 0.2308, "step": 9082 }, { "epoch": 0.08169157226821601, "grad_norm": 3.918940544128418, "learning_rate": 4.042109244581704e-06, "loss": 0.3562, "step": 9083 }, { "epoch": 0.08170056616585647, "grad_norm": 3.0579004287719727, "learning_rate": 4.0380127041009134e-06, "loss": 0.608, "step": 9084 }, { "epoch": 0.08170956006349692, "grad_norm": 2.3308796882629395, "learning_rate": 4.03391805814195e-06, "loss": 0.7745, "step": 9085 }, { "epoch": 0.08171855396113736, "grad_norm": 16.30845069885254, "learning_rate": 4.029825307074894e-06, "loss": 0.3971, "step": 9086 }, { "epoch": 0.08172754785877782, "grad_norm": 2.3344545364379883, "learning_rate": 4.025734451269636e-06, "loss": 0.3128, "step": 9087 }, { "epoch": 0.08173654175641827, "grad_norm": 2.2971129417419434, "learning_rate": 4.021645491095913e-06, "loss": 0.2447, "step": 9088 }, { "epoch": 0.08174553565405872, "grad_norm": 2.1029787063598633, "learning_rate": 4.017558426923271e-06, "loss": 0.5447, "step": 9089 }, { "epoch": 0.08175452955169918, "grad_norm": 2.1476402282714844, "learning_rate": 4.013473259121095e-06, "loss": 0.35, "step": 9090 }, { "epoch": 0.08176352344933963, "grad_norm": 2.465406656265259, "learning_rate": 4.009389988058607e-06, "loss": 0.3316, "step": 9091 }, { "epoch": 0.08177251734698007, "grad_norm": 2.5240068435668945, "learning_rate": 4.005308614104841e-06, "loss": 0.3905, "step": 9092 }, { "epoch": 0.08178151124462052, "grad_norm": 3.2704391479492188, "learning_rate": 4.0012291376286704e-06, "loss": 0.9243, "step": 9093 }, { "epoch": 0.08179050514226098, "grad_norm": 2.196274757385254, "learning_rate": 3.997151558998785e-06, "loss": 0.3297, "step": 9094 }, { "epoch": 0.08179949903990143, "grad_norm": 9.726493835449219, "learning_rate": 3.993075878583727e-06, "loss": 0.4638, "step": 9095 }, { "epoch": 0.08180849293754187, "grad_norm": 11.358208656311035, "learning_rate": 3.989002096751843e-06, "loss": 0.3662, "step": 9096 }, { "epoch": 0.08181748683518233, "grad_norm": 12.008241653442383, "learning_rate": 3.984930213871313e-06, "loss": 0.5579, "step": 9097 }, { "epoch": 0.08182648073282278, "grad_norm": 12.932305335998535, "learning_rate": 3.98086023031016e-06, "loss": 0.8013, "step": 9098 }, { "epoch": 0.08183547463046323, "grad_norm": 2.154299020767212, "learning_rate": 3.976792146436212e-06, "loss": 0.3845, "step": 9099 }, { "epoch": 0.08184446852810368, "grad_norm": 14.196988105773926, "learning_rate": 3.972725962617155e-06, "loss": 0.5661, "step": 9100 }, { "epoch": 0.08185346242574414, "grad_norm": 3.9034554958343506, "learning_rate": 3.968661679220468e-06, "loss": 0.2995, "step": 9101 }, { "epoch": 0.08186245632338458, "grad_norm": 2.1760594844818115, "learning_rate": 3.964599296613492e-06, "loss": 0.6357, "step": 9102 }, { "epoch": 0.08187145022102503, "grad_norm": 1.9595752954483032, "learning_rate": 3.960538815163375e-06, "loss": 0.6917, "step": 9103 }, { "epoch": 0.08188044411866549, "grad_norm": 2.6058528423309326, "learning_rate": 3.956480235237098e-06, "loss": 0.456, "step": 9104 }, { "epoch": 0.08188943801630594, "grad_norm": 1.711787223815918, "learning_rate": 3.9524235572014704e-06, "loss": 0.2908, "step": 9105 }, { "epoch": 0.08189843191394638, "grad_norm": 2.1313161849975586, "learning_rate": 3.9483687814231255e-06, "loss": 0.4527, "step": 9106 }, { "epoch": 0.08190742581158683, "grad_norm": 2.9876997470855713, "learning_rate": 3.944315908268539e-06, "loss": 0.2887, "step": 9107 }, { "epoch": 0.08191641970922729, "grad_norm": 5.0111870765686035, "learning_rate": 3.9402649381040045e-06, "loss": 0.3728, "step": 9108 }, { "epoch": 0.08192541360686774, "grad_norm": 1.9264228343963623, "learning_rate": 3.936215871295634e-06, "loss": 0.3439, "step": 9109 }, { "epoch": 0.08193440750450819, "grad_norm": 2.0958425998687744, "learning_rate": 3.9321687082093914e-06, "loss": 0.0833, "step": 9110 }, { "epoch": 0.08194340140214865, "grad_norm": 1.6851773262023926, "learning_rate": 3.928123449211041e-06, "loss": 0.3367, "step": 9111 }, { "epoch": 0.0819523952997891, "grad_norm": 7.489118576049805, "learning_rate": 3.924080094666205e-06, "loss": 0.7625, "step": 9112 }, { "epoch": 0.08196138919742954, "grad_norm": 3.3973162174224854, "learning_rate": 3.920038644940302e-06, "loss": 0.4645, "step": 9113 }, { "epoch": 0.08197038309507, "grad_norm": 12.321308135986328, "learning_rate": 3.915999100398607e-06, "loss": 0.6818, "step": 9114 }, { "epoch": 0.08197937699271045, "grad_norm": 4.214868545532227, "learning_rate": 3.911961461406202e-06, "loss": 0.2022, "step": 9115 }, { "epoch": 0.0819883708903509, "grad_norm": 8.194269180297852, "learning_rate": 3.907925728328002e-06, "loss": 0.93, "step": 9116 }, { "epoch": 0.08199736478799134, "grad_norm": 2.01670503616333, "learning_rate": 3.903891901528759e-06, "loss": 0.2679, "step": 9117 }, { "epoch": 0.0820063586856318, "grad_norm": 2.5628528594970703, "learning_rate": 3.899859981373047e-06, "loss": 0.361, "step": 9118 }, { "epoch": 0.08201535258327225, "grad_norm": 1.8492794036865234, "learning_rate": 3.895829968225259e-06, "loss": 0.3081, "step": 9119 }, { "epoch": 0.0820243464809127, "grad_norm": 9.62203311920166, "learning_rate": 3.891801862449629e-06, "loss": 0.5071, "step": 9120 }, { "epoch": 0.08203334037855316, "grad_norm": 1.9114071130752563, "learning_rate": 3.887775664410204e-06, "loss": 0.2496, "step": 9121 }, { "epoch": 0.0820423342761936, "grad_norm": 10.906203269958496, "learning_rate": 3.883751374470882e-06, "loss": 0.8647, "step": 9122 }, { "epoch": 0.08205132817383405, "grad_norm": 14.47404670715332, "learning_rate": 3.87972899299536e-06, "loss": 0.3212, "step": 9123 }, { "epoch": 0.0820603220714745, "grad_norm": 2.957961320877075, "learning_rate": 3.875708520347187e-06, "loss": 0.44, "step": 9124 }, { "epoch": 0.08206931596911496, "grad_norm": 3.1887166500091553, "learning_rate": 3.8716899568897184e-06, "loss": 0.3479, "step": 9125 }, { "epoch": 0.0820783098667554, "grad_norm": 14.07762336730957, "learning_rate": 3.867673302986161e-06, "loss": 0.6591, "step": 9126 }, { "epoch": 0.08208730376439585, "grad_norm": 2.569822311401367, "learning_rate": 3.863658558999531e-06, "loss": 0.3935, "step": 9127 }, { "epoch": 0.08209629766203631, "grad_norm": 2.339246988296509, "learning_rate": 3.859645725292665e-06, "loss": 0.6178, "step": 9128 }, { "epoch": 0.08210529155967676, "grad_norm": 2.829576253890991, "learning_rate": 3.855634802228253e-06, "loss": 0.3136, "step": 9129 }, { "epoch": 0.08211428545731721, "grad_norm": 7.7395124435424805, "learning_rate": 3.851625790168789e-06, "loss": 0.7282, "step": 9130 }, { "epoch": 0.08212327935495767, "grad_norm": 2.8332879543304443, "learning_rate": 3.847618689476612e-06, "loss": 0.4414, "step": 9131 }, { "epoch": 0.08213227325259811, "grad_norm": 2.2751011848449707, "learning_rate": 3.8436135005138764e-06, "loss": 0.2104, "step": 9132 }, { "epoch": 0.08214126715023856, "grad_norm": 2.022834300994873, "learning_rate": 3.839610223642562e-06, "loss": 0.3914, "step": 9133 }, { "epoch": 0.08215026104787901, "grad_norm": 3.7132797241210938, "learning_rate": 3.835608859224485e-06, "loss": 0.2961, "step": 9134 }, { "epoch": 0.08215925494551947, "grad_norm": 20.188867568969727, "learning_rate": 3.831609407621276e-06, "loss": 0.282, "step": 9135 }, { "epoch": 0.08216824884315992, "grad_norm": 11.575118064880371, "learning_rate": 3.827611869194414e-06, "loss": 1.0765, "step": 9136 }, { "epoch": 0.08217724274080036, "grad_norm": 2.546353340148926, "learning_rate": 3.823616244305184e-06, "loss": 0.1838, "step": 9137 }, { "epoch": 0.08218623663844082, "grad_norm": 16.982341766357422, "learning_rate": 3.819622533314709e-06, "loss": 0.5527, "step": 9138 }, { "epoch": 0.08219523053608127, "grad_norm": 2.8152575492858887, "learning_rate": 3.815630736583939e-06, "loss": 0.3559, "step": 9139 }, { "epoch": 0.08220422443372172, "grad_norm": 2.042654037475586, "learning_rate": 3.8116408544736386e-06, "loss": 0.1827, "step": 9140 }, { "epoch": 0.08221321833136216, "grad_norm": 3.3286428451538086, "learning_rate": 3.8076528873444204e-06, "loss": 0.2268, "step": 9141 }, { "epoch": 0.08222221222900263, "grad_norm": 7.108248233795166, "learning_rate": 3.8036668355567045e-06, "loss": 0.5761, "step": 9142 }, { "epoch": 0.08223120612664307, "grad_norm": 1.8409966230392456, "learning_rate": 3.7996826994707524e-06, "loss": 0.1095, "step": 9143 }, { "epoch": 0.08224020002428352, "grad_norm": 12.403912544250488, "learning_rate": 3.7957004794466427e-06, "loss": 0.7503, "step": 9144 }, { "epoch": 0.08224919392192398, "grad_norm": 1.9142820835113525, "learning_rate": 3.7917201758442793e-06, "loss": 0.2805, "step": 9145 }, { "epoch": 0.08225818781956443, "grad_norm": 2.262678384780884, "learning_rate": 3.787741789023408e-06, "loss": 0.5894, "step": 9146 }, { "epoch": 0.08226718171720487, "grad_norm": 2.825627088546753, "learning_rate": 3.783765319343588e-06, "loss": 0.1817, "step": 9147 }, { "epoch": 0.08227617561484533, "grad_norm": 3.66426944732666, "learning_rate": 3.779790767164204e-06, "loss": 0.4749, "step": 9148 }, { "epoch": 0.08228516951248578, "grad_norm": 2.6804239749908447, "learning_rate": 3.7758181328444745e-06, "loss": 0.4153, "step": 9149 }, { "epoch": 0.08229416341012623, "grad_norm": 2.0987348556518555, "learning_rate": 3.7718474167434375e-06, "loss": 0.1619, "step": 9150 }, { "epoch": 0.08230315730776668, "grad_norm": 4.360361099243164, "learning_rate": 3.7678786192199694e-06, "loss": 0.7762, "step": 9151 }, { "epoch": 0.08231215120540714, "grad_norm": 16.1990966796875, "learning_rate": 3.7639117406327555e-06, "loss": 0.8367, "step": 9152 }, { "epoch": 0.08232114510304758, "grad_norm": 16.703054428100586, "learning_rate": 3.7599467813403315e-06, "loss": 0.4989, "step": 9153 }, { "epoch": 0.08233013900068803, "grad_norm": 2.0212745666503906, "learning_rate": 3.7559837417010357e-06, "loss": 0.2016, "step": 9154 }, { "epoch": 0.08233913289832849, "grad_norm": 2.754594326019287, "learning_rate": 3.752022622073051e-06, "loss": 0.4263, "step": 9155 }, { "epoch": 0.08234812679596894, "grad_norm": 11.7836332321167, "learning_rate": 3.7480634228143767e-06, "loss": 0.4851, "step": 9156 }, { "epoch": 0.08235712069360938, "grad_norm": 16.74056053161621, "learning_rate": 3.744106144282833e-06, "loss": 0.2773, "step": 9157 }, { "epoch": 0.08236611459124983, "grad_norm": 9.33006477355957, "learning_rate": 3.7401507868360857e-06, "loss": 0.4493, "step": 9158 }, { "epoch": 0.08237510848889029, "grad_norm": 1.9322936534881592, "learning_rate": 3.7361973508316134e-06, "loss": 0.4505, "step": 9159 }, { "epoch": 0.08238410238653074, "grad_norm": 3.0434370040893555, "learning_rate": 3.7322458366267187e-06, "loss": 0.3916, "step": 9160 }, { "epoch": 0.08239309628417119, "grad_norm": 2.579259157180786, "learning_rate": 3.728296244578533e-06, "loss": 0.2675, "step": 9161 }, { "epoch": 0.08240209018181165, "grad_norm": 3.8165040016174316, "learning_rate": 3.724348575044026e-06, "loss": 0.4139, "step": 9162 }, { "epoch": 0.0824110840794521, "grad_norm": 8.297204971313477, "learning_rate": 3.720402828379979e-06, "loss": 0.5813, "step": 9163 }, { "epoch": 0.08242007797709254, "grad_norm": 2.4020628929138184, "learning_rate": 3.7164590049429987e-06, "loss": 0.4768, "step": 9164 }, { "epoch": 0.08242907187473299, "grad_norm": 2.3291280269622803, "learning_rate": 3.712517105089533e-06, "loss": 0.5856, "step": 9165 }, { "epoch": 0.08243806577237345, "grad_norm": 18.748844146728516, "learning_rate": 3.708577129175839e-06, "loss": 0.3552, "step": 9166 }, { "epoch": 0.0824470596700139, "grad_norm": 11.367530822753906, "learning_rate": 3.704639077558014e-06, "loss": 0.689, "step": 9167 }, { "epoch": 0.08245605356765434, "grad_norm": 3.1343164443969727, "learning_rate": 3.700702950591972e-06, "loss": 0.4535, "step": 9168 }, { "epoch": 0.0824650474652948, "grad_norm": 11.51198959350586, "learning_rate": 3.696768748633453e-06, "loss": 0.6309, "step": 9169 }, { "epoch": 0.08247404136293525, "grad_norm": 3.2477829456329346, "learning_rate": 3.6928364720380303e-06, "loss": 0.8763, "step": 9170 }, { "epoch": 0.0824830352605757, "grad_norm": 2.267542839050293, "learning_rate": 3.6889061211610926e-06, "loss": 0.265, "step": 9171 }, { "epoch": 0.08249202915821616, "grad_norm": 2.7692511081695557, "learning_rate": 3.684977696357872e-06, "loss": 0.6175, "step": 9172 }, { "epoch": 0.0825010230558566, "grad_norm": 21.229583740234375, "learning_rate": 3.6810511979834066e-06, "loss": 0.327, "step": 9173 }, { "epoch": 0.08251001695349705, "grad_norm": 2.051239013671875, "learning_rate": 3.677126626392574e-06, "loss": 0.6079, "step": 9174 }, { "epoch": 0.0825190108511375, "grad_norm": 2.668128728866577, "learning_rate": 3.6732039819400683e-06, "loss": 0.5726, "step": 9175 }, { "epoch": 0.08252800474877796, "grad_norm": 1.7586511373519897, "learning_rate": 3.669283264980408e-06, "loss": 0.225, "step": 9176 }, { "epoch": 0.0825369986464184, "grad_norm": 2.6297824382781982, "learning_rate": 3.6653644758679577e-06, "loss": 0.3623, "step": 9177 }, { "epoch": 0.08254599254405885, "grad_norm": 2.4282329082489014, "learning_rate": 3.661447614956881e-06, "loss": 0.4676, "step": 9178 }, { "epoch": 0.08255498644169931, "grad_norm": 2.3072924613952637, "learning_rate": 3.6575326826011917e-06, "loss": 0.7781, "step": 9179 }, { "epoch": 0.08256398033933976, "grad_norm": 2.6451523303985596, "learning_rate": 3.653619679154707e-06, "loss": 0.3288, "step": 9180 }, { "epoch": 0.0825729742369802, "grad_norm": 2.0466341972351074, "learning_rate": 3.6497086049710803e-06, "loss": 0.1809, "step": 9181 }, { "epoch": 0.08258196813462065, "grad_norm": 1.9705641269683838, "learning_rate": 3.645799460403798e-06, "loss": 0.3866, "step": 9182 }, { "epoch": 0.08259096203226111, "grad_norm": 2.0328099727630615, "learning_rate": 3.6418922458061525e-06, "loss": 0.3492, "step": 9183 }, { "epoch": 0.08259995592990156, "grad_norm": 1.9617674350738525, "learning_rate": 3.6379869615312893e-06, "loss": 0.3542, "step": 9184 }, { "epoch": 0.08260894982754201, "grad_norm": 19.678730010986328, "learning_rate": 3.634083607932154e-06, "loss": 0.6313, "step": 9185 }, { "epoch": 0.08261794372518247, "grad_norm": 16.674808502197266, "learning_rate": 3.630182185361522e-06, "loss": 0.2723, "step": 9186 }, { "epoch": 0.08262693762282292, "grad_norm": 2.338763475418091, "learning_rate": 3.6262826941720146e-06, "loss": 0.6734, "step": 9187 }, { "epoch": 0.08263593152046336, "grad_norm": 2.75555682182312, "learning_rate": 3.6223851347160528e-06, "loss": 0.3917, "step": 9188 }, { "epoch": 0.08264492541810382, "grad_norm": 15.53157901763916, "learning_rate": 3.6184895073458985e-06, "loss": 0.6574, "step": 9189 }, { "epoch": 0.08265391931574427, "grad_norm": 3.3062403202056885, "learning_rate": 3.6145958124136262e-06, "loss": 0.4242, "step": 9190 }, { "epoch": 0.08266291321338472, "grad_norm": 4.561279296875, "learning_rate": 3.610704050271155e-06, "loss": 0.4064, "step": 9191 }, { "epoch": 0.08267190711102516, "grad_norm": 10.893577575683594, "learning_rate": 3.6068142212702134e-06, "loss": 0.6503, "step": 9192 }, { "epoch": 0.08268090100866562, "grad_norm": 16.741561889648438, "learning_rate": 3.602926325762354e-06, "loss": 0.8556, "step": 9193 }, { "epoch": 0.08268989490630607, "grad_norm": 17.145788192749023, "learning_rate": 3.599040364098971e-06, "loss": 0.5532, "step": 9194 }, { "epoch": 0.08269888880394652, "grad_norm": 25.46247673034668, "learning_rate": 3.595156336631264e-06, "loss": 0.5354, "step": 9195 }, { "epoch": 0.08270788270158698, "grad_norm": 18.02614974975586, "learning_rate": 3.591274243710277e-06, "loss": 0.2033, "step": 9196 }, { "epoch": 0.08271687659922743, "grad_norm": 19.908035278320312, "learning_rate": 3.5873940856868656e-06, "loss": 0.2287, "step": 9197 }, { "epoch": 0.08272587049686787, "grad_norm": 11.63160228729248, "learning_rate": 3.5835158629117046e-06, "loss": 0.3759, "step": 9198 }, { "epoch": 0.08273486439450832, "grad_norm": 17.7515811920166, "learning_rate": 3.579639575735319e-06, "loss": 0.4105, "step": 9199 }, { "epoch": 0.08274385829214878, "grad_norm": 2.880474090576172, "learning_rate": 3.575765224508032e-06, "loss": 0.2809, "step": 9200 }, { "epoch": 0.08275285218978923, "grad_norm": 3.366959810256958, "learning_rate": 3.571892809580013e-06, "loss": 0.1297, "step": 9201 }, { "epoch": 0.08276184608742967, "grad_norm": 2.5648040771484375, "learning_rate": 3.5680223313012407e-06, "loss": 0.4452, "step": 9202 }, { "epoch": 0.08277083998507014, "grad_norm": 2.515718460083008, "learning_rate": 3.564153790021524e-06, "loss": 0.2588, "step": 9203 }, { "epoch": 0.08277983388271058, "grad_norm": 24.981529235839844, "learning_rate": 3.5602871860904995e-06, "loss": 0.3405, "step": 9204 }, { "epoch": 0.08278882778035103, "grad_norm": 3.020270824432373, "learning_rate": 3.5564225198576217e-06, "loss": 0.3515, "step": 9205 }, { "epoch": 0.08279782167799148, "grad_norm": 3.940751314163208, "learning_rate": 3.552559791672186e-06, "loss": 0.3908, "step": 9206 }, { "epoch": 0.08280681557563194, "grad_norm": 12.586211204528809, "learning_rate": 3.5486990018832877e-06, "loss": 0.6327, "step": 9207 }, { "epoch": 0.08281580947327238, "grad_norm": 8.222445487976074, "learning_rate": 3.544840150839876e-06, "loss": 0.4761, "step": 9208 }, { "epoch": 0.08282480337091283, "grad_norm": 20.368221282958984, "learning_rate": 3.5409832388906992e-06, "loss": 0.6775, "step": 9209 }, { "epoch": 0.08283379726855329, "grad_norm": 10.908215522766113, "learning_rate": 3.53712826638434e-06, "loss": 0.6692, "step": 9210 }, { "epoch": 0.08284279116619374, "grad_norm": 3.2175064086914062, "learning_rate": 3.5332752336692144e-06, "loss": 0.0994, "step": 9211 }, { "epoch": 0.08285178506383419, "grad_norm": 2.279534339904785, "learning_rate": 3.5294241410935457e-06, "loss": 0.2642, "step": 9212 }, { "epoch": 0.08286077896147465, "grad_norm": 2.331230401992798, "learning_rate": 3.5255749890054036e-06, "loss": 0.4212, "step": 9213 }, { "epoch": 0.08286977285911509, "grad_norm": 2.5104658603668213, "learning_rate": 3.521727777752662e-06, "loss": 0.2629, "step": 9214 }, { "epoch": 0.08287876675675554, "grad_norm": 8.215341567993164, "learning_rate": 3.5178825076830236e-06, "loss": 0.6608, "step": 9215 }, { "epoch": 0.08288776065439599, "grad_norm": 1.929746389389038, "learning_rate": 3.5140391791440325e-06, "loss": 0.3264, "step": 9216 }, { "epoch": 0.08289675455203645, "grad_norm": 2.4106602668762207, "learning_rate": 3.5101977924830387e-06, "loss": 0.6828, "step": 9217 }, { "epoch": 0.0829057484496769, "grad_norm": 3.5676252841949463, "learning_rate": 3.50635834804722e-06, "loss": 0.2755, "step": 9218 }, { "epoch": 0.08291474234731734, "grad_norm": 2.21635103225708, "learning_rate": 3.502520846183577e-06, "loss": 0.8888, "step": 9219 }, { "epoch": 0.0829237362449578, "grad_norm": 11.82616138458252, "learning_rate": 3.4986852872389507e-06, "loss": 0.28, "step": 9220 }, { "epoch": 0.08293273014259825, "grad_norm": 6.135422706604004, "learning_rate": 3.4948516715599874e-06, "loss": 0.611, "step": 9221 }, { "epoch": 0.0829417240402387, "grad_norm": 2.125746726989746, "learning_rate": 3.491019999493164e-06, "loss": 0.5626, "step": 9222 }, { "epoch": 0.08295071793787914, "grad_norm": 9.708209991455078, "learning_rate": 3.4871902713847882e-06, "loss": 0.6503, "step": 9223 }, { "epoch": 0.0829597118355196, "grad_norm": 4.0734782218933105, "learning_rate": 3.4833624875809794e-06, "loss": 0.4329, "step": 9224 }, { "epoch": 0.08296870573316005, "grad_norm": 12.711557388305664, "learning_rate": 3.4795366484277002e-06, "loss": 0.1313, "step": 9225 }, { "epoch": 0.0829776996308005, "grad_norm": 9.173335075378418, "learning_rate": 3.475712754270716e-06, "loss": 0.5021, "step": 9226 }, { "epoch": 0.08298669352844096, "grad_norm": 10.044154167175293, "learning_rate": 3.4718908054556243e-06, "loss": 0.6161, "step": 9227 }, { "epoch": 0.0829956874260814, "grad_norm": 2.3693830966949463, "learning_rate": 3.4680708023278607e-06, "loss": 0.2634, "step": 9228 }, { "epoch": 0.08300468132372185, "grad_norm": 2.851691484451294, "learning_rate": 3.4642527452326656e-06, "loss": 0.2918, "step": 9229 }, { "epoch": 0.08301367522136231, "grad_norm": 2.9579646587371826, "learning_rate": 3.46043663451511e-06, "loss": 0.3118, "step": 9230 }, { "epoch": 0.08302266911900276, "grad_norm": 17.27808380126953, "learning_rate": 3.456622470520085e-06, "loss": 0.4401, "step": 9231 }, { "epoch": 0.0830316630166432, "grad_norm": 12.731069564819336, "learning_rate": 3.45281025359232e-06, "loss": 0.5284, "step": 9232 }, { "epoch": 0.08304065691428365, "grad_norm": 2.4219558238983154, "learning_rate": 3.4489999840763595e-06, "loss": 0.503, "step": 9233 }, { "epoch": 0.08304965081192411, "grad_norm": 3.8079111576080322, "learning_rate": 3.4451916623165605e-06, "loss": 0.0703, "step": 9234 }, { "epoch": 0.08305864470956456, "grad_norm": 2.4116618633270264, "learning_rate": 3.4413852886571284e-06, "loss": 0.5894, "step": 9235 }, { "epoch": 0.08306763860720501, "grad_norm": 17.64954376220703, "learning_rate": 3.4375808634420666e-06, "loss": 0.5128, "step": 9236 }, { "epoch": 0.08307663250484547, "grad_norm": 11.363709449768066, "learning_rate": 3.433778387015227e-06, "loss": 0.4914, "step": 9237 }, { "epoch": 0.08308562640248592, "grad_norm": 3.26381778717041, "learning_rate": 3.429977859720265e-06, "loss": 0.2351, "step": 9238 }, { "epoch": 0.08309462030012636, "grad_norm": 3.518810510635376, "learning_rate": 3.4261792819006684e-06, "loss": 0.1637, "step": 9239 }, { "epoch": 0.08310361419776681, "grad_norm": 15.165038108825684, "learning_rate": 3.4223826538997554e-06, "loss": 0.6441, "step": 9240 }, { "epoch": 0.08311260809540727, "grad_norm": 4.960395812988281, "learning_rate": 3.418587976060653e-06, "loss": 0.6923, "step": 9241 }, { "epoch": 0.08312160199304772, "grad_norm": 2.6836676597595215, "learning_rate": 3.414795248726327e-06, "loss": 0.497, "step": 9242 }, { "epoch": 0.08313059589068816, "grad_norm": 10.067849159240723, "learning_rate": 3.4110044722395568e-06, "loss": 0.8885, "step": 9243 }, { "epoch": 0.08313958978832862, "grad_norm": 9.978849411010742, "learning_rate": 3.407215646942949e-06, "loss": 0.2612, "step": 9244 }, { "epoch": 0.08314858368596907, "grad_norm": 13.13457202911377, "learning_rate": 3.4034287731789323e-06, "loss": 0.3927, "step": 9245 }, { "epoch": 0.08315757758360952, "grad_norm": 5.33831262588501, "learning_rate": 3.3996438512897577e-06, "loss": 0.4674, "step": 9246 }, { "epoch": 0.08316657148124997, "grad_norm": 3.876798391342163, "learning_rate": 3.39586088161751e-06, "loss": 0.5323, "step": 9247 }, { "epoch": 0.08317556537889043, "grad_norm": 7.671564102172852, "learning_rate": 3.3920798645040802e-06, "loss": 0.9011, "step": 9248 }, { "epoch": 0.08318455927653087, "grad_norm": 2.2106986045837402, "learning_rate": 3.3883008002912054e-06, "loss": 0.3696, "step": 9249 }, { "epoch": 0.08319355317417132, "grad_norm": 2.1730875968933105, "learning_rate": 3.3845236893204235e-06, "loss": 0.1702, "step": 9250 }, { "epoch": 0.08320254707181178, "grad_norm": 3.2394394874572754, "learning_rate": 3.3807485319331034e-06, "loss": 0.1611, "step": 9251 }, { "epoch": 0.08321154096945223, "grad_norm": 2.126110792160034, "learning_rate": 3.3769753284704526e-06, "loss": 0.3574, "step": 9252 }, { "epoch": 0.08322053486709267, "grad_norm": 2.692476272583008, "learning_rate": 3.3732040792734733e-06, "loss": 0.3246, "step": 9253 }, { "epoch": 0.08322952876473313, "grad_norm": 2.8567025661468506, "learning_rate": 3.3694347846830205e-06, "loss": 0.4008, "step": 9254 }, { "epoch": 0.08323852266237358, "grad_norm": 3.072559356689453, "learning_rate": 3.365667445039747e-06, "loss": 0.1675, "step": 9255 }, { "epoch": 0.08324751656001403, "grad_norm": 4.044977188110352, "learning_rate": 3.361902060684155e-06, "loss": 0.6626, "step": 9256 }, { "epoch": 0.08325651045765448, "grad_norm": 3.1464929580688477, "learning_rate": 3.3581386319565477e-06, "loss": 0.4792, "step": 9257 }, { "epoch": 0.08326550435529494, "grad_norm": 3.443794012069702, "learning_rate": 3.3543771591970587e-06, "loss": 0.3352, "step": 9258 }, { "epoch": 0.08327449825293538, "grad_norm": 2.6905598640441895, "learning_rate": 3.3506176427456486e-06, "loss": 0.357, "step": 9259 }, { "epoch": 0.08328349215057583, "grad_norm": 11.054201126098633, "learning_rate": 3.346860082942091e-06, "loss": 0.5714, "step": 9260 }, { "epoch": 0.08329248604821629, "grad_norm": 3.0288822650909424, "learning_rate": 3.3431044801260025e-06, "loss": 0.4728, "step": 9261 }, { "epoch": 0.08330147994585674, "grad_norm": 4.276911735534668, "learning_rate": 3.3393508346368035e-06, "loss": 0.377, "step": 9262 }, { "epoch": 0.08331047384349718, "grad_norm": 3.2177586555480957, "learning_rate": 3.3355991468137394e-06, "loss": 0.4786, "step": 9263 }, { "epoch": 0.08331946774113763, "grad_norm": 13.510106086730957, "learning_rate": 3.331849416995894e-06, "loss": 0.474, "step": 9264 }, { "epoch": 0.08332846163877809, "grad_norm": 2.5032739639282227, "learning_rate": 3.3281016455221555e-06, "loss": 0.0511, "step": 9265 }, { "epoch": 0.08333745553641854, "grad_norm": 6.9032368659973145, "learning_rate": 3.3243558327312518e-06, "loss": 0.6473, "step": 9266 }, { "epoch": 0.08334644943405899, "grad_norm": 4.140089988708496, "learning_rate": 3.3206119789617158e-06, "loss": 0.7142, "step": 9267 }, { "epoch": 0.08335544333169945, "grad_norm": 2.159254789352417, "learning_rate": 3.3168700845519236e-06, "loss": 0.4806, "step": 9268 }, { "epoch": 0.0833644372293399, "grad_norm": 1.817166805267334, "learning_rate": 3.3131301498400603e-06, "loss": 0.5697, "step": 9269 }, { "epoch": 0.08337343112698034, "grad_norm": 12.205202102661133, "learning_rate": 3.309392175164125e-06, "loss": 0.4631, "step": 9270 }, { "epoch": 0.0833824250246208, "grad_norm": 2.6663858890533447, "learning_rate": 3.3056561608619697e-06, "loss": 0.1727, "step": 9271 }, { "epoch": 0.08339141892226125, "grad_norm": 21.011825561523438, "learning_rate": 3.3019221072712436e-06, "loss": 0.6291, "step": 9272 }, { "epoch": 0.0834004128199017, "grad_norm": 14.809337615966797, "learning_rate": 3.298190014729424e-06, "loss": 0.248, "step": 9273 }, { "epoch": 0.08340940671754214, "grad_norm": 2.9277091026306152, "learning_rate": 3.2944598835738193e-06, "loss": 0.355, "step": 9274 }, { "epoch": 0.0834184006151826, "grad_norm": 1.9730212688446045, "learning_rate": 3.2907317141415428e-06, "loss": 0.3538, "step": 9275 }, { "epoch": 0.08342739451282305, "grad_norm": 2.683565378189087, "learning_rate": 3.2870055067695556e-06, "loss": 0.7342, "step": 9276 }, { "epoch": 0.0834363884104635, "grad_norm": 2.804011821746826, "learning_rate": 3.283281261794621e-06, "loss": 0.6541, "step": 9277 }, { "epoch": 0.08344538230810396, "grad_norm": 12.462835311889648, "learning_rate": 3.2795589795533377e-06, "loss": 0.3891, "step": 9278 }, { "epoch": 0.0834543762057444, "grad_norm": 2.0254738330841064, "learning_rate": 3.2758386603821136e-06, "loss": 0.3629, "step": 9279 }, { "epoch": 0.08346337010338485, "grad_norm": 6.154382228851318, "learning_rate": 3.272120304617196e-06, "loss": 0.7493, "step": 9280 }, { "epoch": 0.0834723640010253, "grad_norm": 4.673314571380615, "learning_rate": 3.2684039125946448e-06, "loss": 0.7052, "step": 9281 }, { "epoch": 0.08348135789866576, "grad_norm": 14.066244125366211, "learning_rate": 3.2646894846503322e-06, "loss": 0.6278, "step": 9282 }, { "epoch": 0.0834903517963062, "grad_norm": 2.5550284385681152, "learning_rate": 3.260977021119979e-06, "loss": 0.72, "step": 9283 }, { "epoch": 0.08349934569394665, "grad_norm": 2.1449270248413086, "learning_rate": 3.2572665223391084e-06, "loss": 0.3178, "step": 9284 }, { "epoch": 0.08350833959158711, "grad_norm": 11.895355224609375, "learning_rate": 3.2535579886430718e-06, "loss": 0.6566, "step": 9285 }, { "epoch": 0.08351733348922756, "grad_norm": 2.619170665740967, "learning_rate": 3.249851420367031e-06, "loss": 0.3653, "step": 9286 }, { "epoch": 0.08352632738686801, "grad_norm": 14.472835540771484, "learning_rate": 3.246146817845999e-06, "loss": 0.718, "step": 9287 }, { "epoch": 0.08353532128450845, "grad_norm": 2.520326614379883, "learning_rate": 3.242444181414789e-06, "loss": 0.47, "step": 9288 }, { "epoch": 0.08354431518214891, "grad_norm": 3.775430917739868, "learning_rate": 3.2387435114080293e-06, "loss": 0.334, "step": 9289 }, { "epoch": 0.08355330907978936, "grad_norm": 2.7488667964935303, "learning_rate": 3.235044808160198e-06, "loss": 0.4567, "step": 9290 }, { "epoch": 0.08356230297742981, "grad_norm": 3.2547285556793213, "learning_rate": 3.2313480720055745e-06, "loss": 0.3831, "step": 9291 }, { "epoch": 0.08357129687507027, "grad_norm": 4.195824146270752, "learning_rate": 3.2276533032782606e-06, "loss": 0.2523, "step": 9292 }, { "epoch": 0.08358029077271072, "grad_norm": 2.809607982635498, "learning_rate": 3.2239605023121927e-06, "loss": 0.7611, "step": 9293 }, { "epoch": 0.08358928467035116, "grad_norm": 3.129742383956909, "learning_rate": 3.2202696694411143e-06, "loss": 0.4042, "step": 9294 }, { "epoch": 0.08359827856799162, "grad_norm": 2.309900999069214, "learning_rate": 3.2165808049986117e-06, "loss": 0.3437, "step": 9295 }, { "epoch": 0.08360727246563207, "grad_norm": 12.10960578918457, "learning_rate": 3.2128939093180655e-06, "loss": 0.753, "step": 9296 }, { "epoch": 0.08361626636327252, "grad_norm": 7.126574516296387, "learning_rate": 3.209208982732706e-06, "loss": 0.6057, "step": 9297 }, { "epoch": 0.08362526026091296, "grad_norm": 2.260127544403076, "learning_rate": 3.20552602557557e-06, "loss": 0.3071, "step": 9298 }, { "epoch": 0.08363425415855343, "grad_norm": 18.082124710083008, "learning_rate": 3.201845038179513e-06, "loss": 0.2529, "step": 9299 }, { "epoch": 0.08364324805619387, "grad_norm": 17.74921226501465, "learning_rate": 3.1981660208772247e-06, "loss": 0.5736, "step": 9300 }, { "epoch": 0.08365224195383432, "grad_norm": 2.415400505065918, "learning_rate": 3.194488974001203e-06, "loss": 0.7392, "step": 9301 }, { "epoch": 0.08366123585147478, "grad_norm": 1.9702637195587158, "learning_rate": 3.1908138978837848e-06, "loss": 0.3695, "step": 9302 }, { "epoch": 0.08367022974911523, "grad_norm": 2.3711888790130615, "learning_rate": 3.1871407928571183e-06, "loss": 0.8507, "step": 9303 }, { "epoch": 0.08367922364675567, "grad_norm": 14.154026985168457, "learning_rate": 3.183469659253163e-06, "loss": 0.3508, "step": 9304 }, { "epoch": 0.08368821754439612, "grad_norm": 8.344785690307617, "learning_rate": 3.179800497403729e-06, "loss": 0.7509, "step": 9305 }, { "epoch": 0.08369721144203658, "grad_norm": 2.2819080352783203, "learning_rate": 3.1761333076404164e-06, "loss": 0.4206, "step": 9306 }, { "epoch": 0.08370620533967703, "grad_norm": 19.32234764099121, "learning_rate": 3.1724680902946753e-06, "loss": 0.5642, "step": 9307 }, { "epoch": 0.08371519923731748, "grad_norm": 2.1817240715026855, "learning_rate": 3.1688048456977537e-06, "loss": 0.4263, "step": 9308 }, { "epoch": 0.08372419313495794, "grad_norm": 11.79926586151123, "learning_rate": 3.1651435741807372e-06, "loss": 0.4442, "step": 9309 }, { "epoch": 0.08373318703259838, "grad_norm": 2.004343271255493, "learning_rate": 3.1614842760745277e-06, "loss": 0.401, "step": 9310 }, { "epoch": 0.08374218093023883, "grad_norm": 2.4981749057769775, "learning_rate": 3.1578269517098412e-06, "loss": 0.3697, "step": 9311 }, { "epoch": 0.08375117482787929, "grad_norm": 2.174621820449829, "learning_rate": 3.1541716014172357e-06, "loss": 0.1466, "step": 9312 }, { "epoch": 0.08376016872551974, "grad_norm": 3.131197214126587, "learning_rate": 3.150518225527069e-06, "loss": 0.451, "step": 9313 }, { "epoch": 0.08376916262316018, "grad_norm": 7.191627502441406, "learning_rate": 3.1468668243695294e-06, "loss": 0.5454, "step": 9314 }, { "epoch": 0.08377815652080063, "grad_norm": 2.3303890228271484, "learning_rate": 3.1432173982746285e-06, "loss": 0.8424, "step": 9315 }, { "epoch": 0.08378715041844109, "grad_norm": 2.3219566345214844, "learning_rate": 3.1395699475721935e-06, "loss": 0.5266, "step": 9316 }, { "epoch": 0.08379614431608154, "grad_norm": 4.828557968139648, "learning_rate": 3.1359244725918835e-06, "loss": 0.6718, "step": 9317 }, { "epoch": 0.08380513821372199, "grad_norm": 2.8991153240203857, "learning_rate": 3.1322809736631654e-06, "loss": 0.1933, "step": 9318 }, { "epoch": 0.08381413211136245, "grad_norm": 3.8561556339263916, "learning_rate": 3.1286394511153448e-06, "loss": 0.3812, "step": 9319 }, { "epoch": 0.0838231260090029, "grad_norm": 2.681469202041626, "learning_rate": 3.124999905277526e-06, "loss": 0.4135, "step": 9320 }, { "epoch": 0.08383211990664334, "grad_norm": 2.1060051918029785, "learning_rate": 3.1213623364786613e-06, "loss": 0.3465, "step": 9321 }, { "epoch": 0.08384111380428379, "grad_norm": 3.415558099746704, "learning_rate": 3.1177267450475e-06, "loss": 0.2675, "step": 9322 }, { "epoch": 0.08385010770192425, "grad_norm": 2.6998443603515625, "learning_rate": 3.114093131312623e-06, "loss": 0.4039, "step": 9323 }, { "epoch": 0.0838591015995647, "grad_norm": 2.9553821086883545, "learning_rate": 3.11046149560244e-06, "loss": 0.3413, "step": 9324 }, { "epoch": 0.08386809549720514, "grad_norm": 3.214954376220703, "learning_rate": 3.106831838245164e-06, "loss": 0.3162, "step": 9325 }, { "epoch": 0.0838770893948456, "grad_norm": 5.212258338928223, "learning_rate": 3.103204159568851e-06, "loss": 0.4573, "step": 9326 }, { "epoch": 0.08388608329248605, "grad_norm": 2.841947078704834, "learning_rate": 3.099578459901359e-06, "loss": 0.5349, "step": 9327 }, { "epoch": 0.0838950771901265, "grad_norm": 2.1919305324554443, "learning_rate": 3.095954739570378e-06, "loss": 0.4373, "step": 9328 }, { "epoch": 0.08390407108776694, "grad_norm": 2.8264899253845215, "learning_rate": 3.092332998903416e-06, "loss": 0.2696, "step": 9329 }, { "epoch": 0.0839130649854074, "grad_norm": 1.7750661373138428, "learning_rate": 3.088713238227792e-06, "loss": 0.1372, "step": 9330 }, { "epoch": 0.08392205888304785, "grad_norm": 14.3159761428833, "learning_rate": 3.085095457870671e-06, "loss": 0.1905, "step": 9331 }, { "epoch": 0.0839310527806883, "grad_norm": 2.761253833770752, "learning_rate": 3.0814796581590145e-06, "loss": 0.5165, "step": 9332 }, { "epoch": 0.08394004667832876, "grad_norm": 3.6564149856567383, "learning_rate": 3.077865839419622e-06, "loss": 0.7288, "step": 9333 }, { "epoch": 0.0839490405759692, "grad_norm": 13.530410766601562, "learning_rate": 3.0742540019791017e-06, "loss": 0.577, "step": 9334 }, { "epoch": 0.08395803447360965, "grad_norm": 3.4010725021362305, "learning_rate": 3.070644146163881e-06, "loss": 0.2834, "step": 9335 }, { "epoch": 0.08396702837125011, "grad_norm": 3.551132917404175, "learning_rate": 3.0670362723002303e-06, "loss": 0.3268, "step": 9336 }, { "epoch": 0.08397602226889056, "grad_norm": 17.973268508911133, "learning_rate": 3.063430380714211e-06, "loss": 0.228, "step": 9337 }, { "epoch": 0.083985016166531, "grad_norm": 1.9814374446868896, "learning_rate": 3.0598264717317305e-06, "loss": 0.6922, "step": 9338 }, { "epoch": 0.08399401006417145, "grad_norm": 2.4144837856292725, "learning_rate": 3.0562245456785014e-06, "loss": 0.404, "step": 9339 }, { "epoch": 0.08400300396181191, "grad_norm": 1.9291023015975952, "learning_rate": 3.0526246028800637e-06, "loss": 0.4432, "step": 9340 }, { "epoch": 0.08401199785945236, "grad_norm": 2.247400999069214, "learning_rate": 3.0490266436617675e-06, "loss": 0.5488, "step": 9341 }, { "epoch": 0.08402099175709281, "grad_norm": 24.72880744934082, "learning_rate": 3.045430668348806e-06, "loss": 0.654, "step": 9342 }, { "epoch": 0.08402998565473327, "grad_norm": 2.9814374446868896, "learning_rate": 3.0418366772661744e-06, "loss": 0.4683, "step": 9343 }, { "epoch": 0.08403897955237372, "grad_norm": 2.074248790740967, "learning_rate": 3.038244670738685e-06, "loss": 0.7584, "step": 9344 }, { "epoch": 0.08404797345001416, "grad_norm": 1.8493696451187134, "learning_rate": 3.0346546490909926e-06, "loss": 0.3173, "step": 9345 }, { "epoch": 0.08405696734765461, "grad_norm": 2.0235595703125, "learning_rate": 3.031066612647554e-06, "loss": 0.3813, "step": 9346 }, { "epoch": 0.08406596124529507, "grad_norm": 2.417133331298828, "learning_rate": 3.0274805617326457e-06, "loss": 0.478, "step": 9347 }, { "epoch": 0.08407495514293552, "grad_norm": 2.9586265087127686, "learning_rate": 3.0238964966703838e-06, "loss": 0.5128, "step": 9348 }, { "epoch": 0.08408394904057596, "grad_norm": 3.059847354888916, "learning_rate": 3.020314417784681e-06, "loss": 0.6088, "step": 9349 }, { "epoch": 0.08409294293821643, "grad_norm": 3.402155637741089, "learning_rate": 3.0167343253992923e-06, "loss": 0.5319, "step": 9350 }, { "epoch": 0.08410193683585687, "grad_norm": 3.4936559200286865, "learning_rate": 3.013156219837776e-06, "loss": 0.3839, "step": 9351 }, { "epoch": 0.08411093073349732, "grad_norm": 2.7222068309783936, "learning_rate": 3.009580101423512e-06, "loss": 0.4742, "step": 9352 }, { "epoch": 0.08411992463113778, "grad_norm": 2.118896722793579, "learning_rate": 3.0060059704797196e-06, "loss": 0.2351, "step": 9353 }, { "epoch": 0.08412891852877823, "grad_norm": 2.8689827919006348, "learning_rate": 3.002433827329415e-06, "loss": 0.3657, "step": 9354 }, { "epoch": 0.08413791242641867, "grad_norm": 2.712507486343384, "learning_rate": 2.998863672295449e-06, "loss": 0.6947, "step": 9355 }, { "epoch": 0.08414690632405912, "grad_norm": 2.3650596141815186, "learning_rate": 2.995295505700488e-06, "loss": 0.4735, "step": 9356 }, { "epoch": 0.08415590022169958, "grad_norm": 2.310042381286621, "learning_rate": 2.99172932786701e-06, "loss": 0.3104, "step": 9357 }, { "epoch": 0.08416489411934003, "grad_norm": 9.799020767211914, "learning_rate": 2.9881651391173356e-06, "loss": 0.4685, "step": 9358 }, { "epoch": 0.08417388801698047, "grad_norm": 3.462599754333496, "learning_rate": 2.9846029397735816e-06, "loss": 0.1702, "step": 9359 }, { "epoch": 0.08418288191462094, "grad_norm": 1.8467154502868652, "learning_rate": 2.981042730157707e-06, "loss": 0.3938, "step": 9360 }, { "epoch": 0.08419187581226138, "grad_norm": 7.577988147735596, "learning_rate": 2.977484510591469e-06, "loss": 0.7582, "step": 9361 }, { "epoch": 0.08420086970990183, "grad_norm": 12.434563636779785, "learning_rate": 2.9739282813964627e-06, "loss": 0.7952, "step": 9362 }, { "epoch": 0.08420986360754228, "grad_norm": 3.0625429153442383, "learning_rate": 2.9703740428940946e-06, "loss": 0.1632, "step": 9363 }, { "epoch": 0.08421885750518274, "grad_norm": 2.457343101501465, "learning_rate": 2.9668217954055887e-06, "loss": 0.4234, "step": 9364 }, { "epoch": 0.08422785140282318, "grad_norm": 27.72540855407715, "learning_rate": 2.9632715392520023e-06, "loss": 1.279, "step": 9365 }, { "epoch": 0.08423684530046363, "grad_norm": 1.943159818649292, "learning_rate": 2.9597232747541894e-06, "loss": 0.3252, "step": 9366 }, { "epoch": 0.08424583919810409, "grad_norm": 26.840885162353516, "learning_rate": 2.9561770022328544e-06, "loss": 0.7302, "step": 9367 }, { "epoch": 0.08425483309574454, "grad_norm": 1.7450023889541626, "learning_rate": 2.9526327220084998e-06, "loss": 0.4238, "step": 9368 }, { "epoch": 0.08426382699338499, "grad_norm": 3.25118088722229, "learning_rate": 2.9490904344014497e-06, "loss": 0.6111, "step": 9369 }, { "epoch": 0.08427282089102545, "grad_norm": 2.4347734451293945, "learning_rate": 2.9455501397318562e-06, "loss": 0.2216, "step": 9370 }, { "epoch": 0.08428181478866589, "grad_norm": 19.409339904785156, "learning_rate": 2.9420118383196803e-06, "loss": 0.2613, "step": 9371 }, { "epoch": 0.08429080868630634, "grad_norm": 1.9349833726882935, "learning_rate": 2.9384755304847217e-06, "loss": 0.3905, "step": 9372 }, { "epoch": 0.08429980258394679, "grad_norm": 2.1129205226898193, "learning_rate": 2.9349412165465773e-06, "loss": 0.4183, "step": 9373 }, { "epoch": 0.08430879648158725, "grad_norm": 3.157784938812256, "learning_rate": 2.9314088968246832e-06, "loss": 0.3913, "step": 9374 }, { "epoch": 0.0843177903792277, "grad_norm": 10.159321784973145, "learning_rate": 2.927878571638282e-06, "loss": 0.4853, "step": 9375 }, { "epoch": 0.08432678427686814, "grad_norm": 3.28212308883667, "learning_rate": 2.9243502413064368e-06, "loss": 0.5876, "step": 9376 }, { "epoch": 0.0843357781745086, "grad_norm": 27.838329315185547, "learning_rate": 2.920823906148043e-06, "loss": 0.9143, "step": 9377 }, { "epoch": 0.08434477207214905, "grad_norm": 2.942870855331421, "learning_rate": 2.9172995664818016e-06, "loss": 0.3231, "step": 9378 }, { "epoch": 0.0843537659697895, "grad_norm": 2.27071475982666, "learning_rate": 2.9137772226262404e-06, "loss": 0.3046, "step": 9379 }, { "epoch": 0.08436275986742994, "grad_norm": 8.190023422241211, "learning_rate": 2.9102568748997056e-06, "loss": 0.8004, "step": 9380 }, { "epoch": 0.0843717537650704, "grad_norm": 2.8314530849456787, "learning_rate": 2.9067385236203586e-06, "loss": 0.4381, "step": 9381 }, { "epoch": 0.08438074766271085, "grad_norm": 2.578510046005249, "learning_rate": 2.9032221691061903e-06, "loss": 0.522, "step": 9382 }, { "epoch": 0.0843897415603513, "grad_norm": 2.277554750442505, "learning_rate": 2.899707811675001e-06, "loss": 0.5941, "step": 9383 }, { "epoch": 0.08439873545799176, "grad_norm": 2.6325647830963135, "learning_rate": 2.896195451644415e-06, "loss": 0.5793, "step": 9384 }, { "epoch": 0.0844077293556322, "grad_norm": 2.050834894180298, "learning_rate": 2.892685089331873e-06, "loss": 0.4244, "step": 9385 }, { "epoch": 0.08441672325327265, "grad_norm": 1.9019603729248047, "learning_rate": 2.889176725054643e-06, "loss": 0.3069, "step": 9386 }, { "epoch": 0.0844257171509131, "grad_norm": 2.3769941329956055, "learning_rate": 2.8856703591298077e-06, "loss": 0.3585, "step": 9387 }, { "epoch": 0.08443471104855356, "grad_norm": 2.485809326171875, "learning_rate": 2.8821659918742584e-06, "loss": 0.4001, "step": 9388 }, { "epoch": 0.084443704946194, "grad_norm": 2.2413151264190674, "learning_rate": 2.8786636236047298e-06, "loss": 0.5901, "step": 9389 }, { "epoch": 0.08445269884383445, "grad_norm": 16.721126556396484, "learning_rate": 2.87516325463775e-06, "loss": 0.28, "step": 9390 }, { "epoch": 0.08446169274147491, "grad_norm": 2.4173879623413086, "learning_rate": 2.8716648852896903e-06, "loss": 0.8788, "step": 9391 }, { "epoch": 0.08447068663911536, "grad_norm": 2.0514750480651855, "learning_rate": 2.8681685158767208e-06, "loss": 0.3381, "step": 9392 }, { "epoch": 0.08447968053675581, "grad_norm": 2.6332969665527344, "learning_rate": 2.8646741467148406e-06, "loss": 0.6094, "step": 9393 }, { "epoch": 0.08448867443439627, "grad_norm": 2.4646201133728027, "learning_rate": 2.861181778119873e-06, "loss": 0.7335, "step": 9394 }, { "epoch": 0.08449766833203672, "grad_norm": 2.7836811542510986, "learning_rate": 2.8576914104074425e-06, "loss": 0.4887, "step": 9395 }, { "epoch": 0.08450666222967716, "grad_norm": 2.130242109298706, "learning_rate": 2.8542030438930224e-06, "loss": 0.3224, "step": 9396 }, { "epoch": 0.08451565612731761, "grad_norm": 3.7454633712768555, "learning_rate": 2.850716678891874e-06, "loss": 0.3961, "step": 9397 }, { "epoch": 0.08452465002495807, "grad_norm": 2.284501075744629, "learning_rate": 2.847232315719095e-06, "loss": 0.4758, "step": 9398 }, { "epoch": 0.08453364392259852, "grad_norm": 7.187203884124756, "learning_rate": 2.8437499546896006e-06, "loss": 0.6611, "step": 9399 }, { "epoch": 0.08454263782023896, "grad_norm": 18.341724395751953, "learning_rate": 2.840269596118114e-06, "loss": 0.4499, "step": 9400 }, { "epoch": 0.08455163171787942, "grad_norm": 20.619915008544922, "learning_rate": 2.8367912403191977e-06, "loss": 0.5758, "step": 9401 }, { "epoch": 0.08456062561551987, "grad_norm": 16.37345314025879, "learning_rate": 2.8333148876072137e-06, "loss": 0.5638, "step": 9402 }, { "epoch": 0.08456961951316032, "grad_norm": 1.8689115047454834, "learning_rate": 2.8298405382963576e-06, "loss": 0.4974, "step": 9403 }, { "epoch": 0.08457861341080077, "grad_norm": 4.023101806640625, "learning_rate": 2.8263681927006347e-06, "loss": 0.3462, "step": 9404 }, { "epoch": 0.08458760730844123, "grad_norm": 1.999869704246521, "learning_rate": 2.8228978511338656e-06, "loss": 0.7084, "step": 9405 }, { "epoch": 0.08459660120608167, "grad_norm": 1.8517991304397583, "learning_rate": 2.8194295139097048e-06, "loss": 0.391, "step": 9406 }, { "epoch": 0.08460559510372212, "grad_norm": 21.838876724243164, "learning_rate": 2.8159631813416105e-06, "loss": 0.4484, "step": 9407 }, { "epoch": 0.08461458900136258, "grad_norm": 1.9997739791870117, "learning_rate": 2.812498853742876e-06, "loss": 0.2677, "step": 9408 }, { "epoch": 0.08462358289900303, "grad_norm": 4.261723518371582, "learning_rate": 2.8090365314265927e-06, "loss": 0.467, "step": 9409 }, { "epoch": 0.08463257679664347, "grad_norm": 2.419429302215576, "learning_rate": 2.805576214705688e-06, "loss": 0.2735, "step": 9410 }, { "epoch": 0.08464157069428394, "grad_norm": 3.171677827835083, "learning_rate": 2.802117903892898e-06, "loss": 0.3528, "step": 9411 }, { "epoch": 0.08465056459192438, "grad_norm": 16.777341842651367, "learning_rate": 2.798661599300778e-06, "loss": 0.7086, "step": 9412 }, { "epoch": 0.08465955848956483, "grad_norm": 15.355764389038086, "learning_rate": 2.7952073012417145e-06, "loss": 0.4702, "step": 9413 }, { "epoch": 0.08466855238720528, "grad_norm": 1.8823792934417725, "learning_rate": 2.791755010027894e-06, "loss": 0.375, "step": 9414 }, { "epoch": 0.08467754628484574, "grad_norm": 2.64400053024292, "learning_rate": 2.7883047259713386e-06, "loss": 0.7503, "step": 9415 }, { "epoch": 0.08468654018248618, "grad_norm": 3.3309271335601807, "learning_rate": 2.7848564493838767e-06, "loss": 0.184, "step": 9416 }, { "epoch": 0.08469553408012663, "grad_norm": 3.923962116241455, "learning_rate": 2.781410180577157e-06, "loss": 0.6477, "step": 9417 }, { "epoch": 0.08470452797776709, "grad_norm": 2.3149287700653076, "learning_rate": 2.7779659198626574e-06, "loss": 0.3684, "step": 9418 }, { "epoch": 0.08471352187540754, "grad_norm": 2.028630495071411, "learning_rate": 2.7745236675516597e-06, "loss": 0.712, "step": 9419 }, { "epoch": 0.08472251577304798, "grad_norm": 2.3241405487060547, "learning_rate": 2.771083423955276e-06, "loss": 0.4238, "step": 9420 }, { "epoch": 0.08473150967068843, "grad_norm": 12.497243881225586, "learning_rate": 2.7676451893844266e-06, "loss": 0.7994, "step": 9421 }, { "epoch": 0.08474050356832889, "grad_norm": 18.84054946899414, "learning_rate": 2.7642089641498614e-06, "loss": 0.5236, "step": 9422 }, { "epoch": 0.08474949746596934, "grad_norm": 2.084601402282715, "learning_rate": 2.7607747485621395e-06, "loss": 0.3897, "step": 9423 }, { "epoch": 0.08475849136360979, "grad_norm": 10.839725494384766, "learning_rate": 2.757342542931643e-06, "loss": 0.7574, "step": 9424 }, { "epoch": 0.08476748526125025, "grad_norm": 21.919221878051758, "learning_rate": 2.7539123475685683e-06, "loss": 0.5563, "step": 9425 }, { "epoch": 0.0847764791588907, "grad_norm": 2.37664794921875, "learning_rate": 2.750484162782929e-06, "loss": 0.1642, "step": 9426 }, { "epoch": 0.08478547305653114, "grad_norm": 11.044760704040527, "learning_rate": 2.7470579888845743e-06, "loss": 0.6388, "step": 9427 }, { "epoch": 0.08479446695417159, "grad_norm": 3.108893394470215, "learning_rate": 2.743633826183145e-06, "loss": 0.5794, "step": 9428 }, { "epoch": 0.08480346085181205, "grad_norm": 2.417186975479126, "learning_rate": 2.7402116749881163e-06, "loss": 0.4407, "step": 9429 }, { "epoch": 0.0848124547494525, "grad_norm": 2.7383344173431396, "learning_rate": 2.7367915356087847e-06, "loss": 0.1945, "step": 9430 }, { "epoch": 0.08482144864709294, "grad_norm": 2.123339891433716, "learning_rate": 2.733373408354248e-06, "loss": 0.272, "step": 9431 }, { "epoch": 0.0848304425447334, "grad_norm": 14.984065055847168, "learning_rate": 2.7299572935334445e-06, "loss": 0.5131, "step": 9432 }, { "epoch": 0.08483943644237385, "grad_norm": 3.2400100231170654, "learning_rate": 2.726543191455114e-06, "loss": 0.3495, "step": 9433 }, { "epoch": 0.0848484303400143, "grad_norm": 2.5283684730529785, "learning_rate": 2.7231311024278115e-06, "loss": 0.4674, "step": 9434 }, { "epoch": 0.08485742423765476, "grad_norm": 2.4240636825561523, "learning_rate": 2.719721026759933e-06, "loss": 0.2383, "step": 9435 }, { "epoch": 0.0848664181352952, "grad_norm": 2.438467264175415, "learning_rate": 2.7163129647596625e-06, "loss": 0.6041, "step": 9436 }, { "epoch": 0.08487541203293565, "grad_norm": 2.546339511871338, "learning_rate": 2.7129069167350306e-06, "loss": 0.0685, "step": 9437 }, { "epoch": 0.0848844059305761, "grad_norm": 3.56160306930542, "learning_rate": 2.7095028829938667e-06, "loss": 0.448, "step": 9438 }, { "epoch": 0.08489339982821656, "grad_norm": 2.4544968605041504, "learning_rate": 2.706100863843822e-06, "loss": 0.8398, "step": 9439 }, { "epoch": 0.084902393725857, "grad_norm": 2.048142433166504, "learning_rate": 2.7027008595923693e-06, "loss": 0.3631, "step": 9440 }, { "epoch": 0.08491138762349745, "grad_norm": 2.2665297985076904, "learning_rate": 2.6993028705467883e-06, "loss": 0.375, "step": 9441 }, { "epoch": 0.08492038152113791, "grad_norm": 3.355891466140747, "learning_rate": 2.6959068970142e-06, "loss": 0.2531, "step": 9442 }, { "epoch": 0.08492937541877836, "grad_norm": 2.3667941093444824, "learning_rate": 2.6925129393015197e-06, "loss": 0.5107, "step": 9443 }, { "epoch": 0.08493836931641881, "grad_norm": 2.61555552482605, "learning_rate": 2.6891209977154943e-06, "loss": 0.4891, "step": 9444 }, { "epoch": 0.08494736321405925, "grad_norm": 2.2762389183044434, "learning_rate": 2.685731072562683e-06, "loss": 0.5037, "step": 9445 }, { "epoch": 0.08495635711169972, "grad_norm": 12.113385200500488, "learning_rate": 2.682343164149456e-06, "loss": 0.5659, "step": 9446 }, { "epoch": 0.08496535100934016, "grad_norm": 2.4915685653686523, "learning_rate": 2.6789572727820233e-06, "loss": 0.471, "step": 9447 }, { "epoch": 0.08497434490698061, "grad_norm": 3.1150970458984375, "learning_rate": 2.6755733987663816e-06, "loss": 0.5219, "step": 9448 }, { "epoch": 0.08498333880462107, "grad_norm": 2.2192747592926025, "learning_rate": 2.6721915424083783e-06, "loss": 0.399, "step": 9449 }, { "epoch": 0.08499233270226152, "grad_norm": 14.257675170898438, "learning_rate": 2.6688117040136464e-06, "loss": 0.3145, "step": 9450 }, { "epoch": 0.08500132659990196, "grad_norm": 16.568647384643555, "learning_rate": 2.6654338838876665e-06, "loss": 0.7843, "step": 9451 }, { "epoch": 0.08501032049754242, "grad_norm": 2.062504291534424, "learning_rate": 2.662058082335714e-06, "loss": 0.6724, "step": 9452 }, { "epoch": 0.08501931439518287, "grad_norm": 16.548728942871094, "learning_rate": 2.6586842996628918e-06, "loss": 0.5644, "step": 9453 }, { "epoch": 0.08502830829282332, "grad_norm": 2.0558764934539795, "learning_rate": 2.6553125361741177e-06, "loss": 0.1603, "step": 9454 }, { "epoch": 0.08503730219046376, "grad_norm": 2.165374517440796, "learning_rate": 2.651942792174125e-06, "loss": 0.2538, "step": 9455 }, { "epoch": 0.08504629608810423, "grad_norm": 15.185065269470215, "learning_rate": 2.6485750679674766e-06, "loss": 0.59, "step": 9456 }, { "epoch": 0.08505528998574467, "grad_norm": 2.481633424758911, "learning_rate": 2.645209363858539e-06, "loss": 0.4606, "step": 9457 }, { "epoch": 0.08506428388338512, "grad_norm": 2.085165500640869, "learning_rate": 2.6418456801514948e-06, "loss": 0.3153, "step": 9458 }, { "epoch": 0.08507327778102558, "grad_norm": 2.7872402667999268, "learning_rate": 2.6384840171503612e-06, "loss": 0.5077, "step": 9459 }, { "epoch": 0.08508227167866603, "grad_norm": 2.9239084720611572, "learning_rate": 2.6351243751589512e-06, "loss": 0.5625, "step": 9460 }, { "epoch": 0.08509126557630647, "grad_norm": 15.20152473449707, "learning_rate": 2.6317667544809134e-06, "loss": 0.4556, "step": 9461 }, { "epoch": 0.08510025947394692, "grad_norm": 13.656667709350586, "learning_rate": 2.6284111554197023e-06, "loss": 0.6239, "step": 9462 }, { "epoch": 0.08510925337158738, "grad_norm": 2.33200740814209, "learning_rate": 2.6250575782785948e-06, "loss": 0.3749, "step": 9463 }, { "epoch": 0.08511824726922783, "grad_norm": 2.0074329376220703, "learning_rate": 2.621706023360687e-06, "loss": 0.4716, "step": 9464 }, { "epoch": 0.08512724116686828, "grad_norm": 13.378423690795898, "learning_rate": 2.6183564909688817e-06, "loss": 0.5522, "step": 9465 }, { "epoch": 0.08513623506450874, "grad_norm": 2.3718879222869873, "learning_rate": 2.615008981405906e-06, "loss": 0.5046, "step": 9466 }, { "epoch": 0.08514522896214918, "grad_norm": 12.710310935974121, "learning_rate": 2.6116634949743147e-06, "loss": 0.4882, "step": 9467 }, { "epoch": 0.08515422285978963, "grad_norm": 2.385256767272949, "learning_rate": 2.608320031976458e-06, "loss": 0.7818, "step": 9468 }, { "epoch": 0.08516321675743008, "grad_norm": 1.9105899333953857, "learning_rate": 2.6049785927145213e-06, "loss": 0.2313, "step": 9469 }, { "epoch": 0.08517221065507054, "grad_norm": 2.262575626373291, "learning_rate": 2.601639177490489e-06, "loss": 0.4738, "step": 9470 }, { "epoch": 0.08518120455271098, "grad_norm": 2.4758667945861816, "learning_rate": 2.5983017866061915e-06, "loss": 0.3884, "step": 9471 }, { "epoch": 0.08519019845035143, "grad_norm": 2.1431171894073486, "learning_rate": 2.5949664203632428e-06, "loss": 0.3235, "step": 9472 }, { "epoch": 0.08519919234799189, "grad_norm": 5.519144535064697, "learning_rate": 2.5916330790630994e-06, "loss": 0.0871, "step": 9473 }, { "epoch": 0.08520818624563234, "grad_norm": 3.051807403564453, "learning_rate": 2.5883017630070198e-06, "loss": 0.2871, "step": 9474 }, { "epoch": 0.08521718014327279, "grad_norm": 2.646359443664551, "learning_rate": 2.584972472496089e-06, "loss": 0.5303, "step": 9475 }, { "epoch": 0.08522617404091325, "grad_norm": 15.510170936584473, "learning_rate": 2.581645207831204e-06, "loss": 0.3289, "step": 9476 }, { "epoch": 0.0852351679385537, "grad_norm": 2.47595477104187, "learning_rate": 2.5783199693130755e-06, "loss": 0.2479, "step": 9477 }, { "epoch": 0.08524416183619414, "grad_norm": 2.2484023571014404, "learning_rate": 2.5749967572422403e-06, "loss": 0.3827, "step": 9478 }, { "epoch": 0.08525315573383459, "grad_norm": 2.1156747341156006, "learning_rate": 2.5716755719190438e-06, "loss": 0.15, "step": 9479 }, { "epoch": 0.08526214963147505, "grad_norm": 3.149153709411621, "learning_rate": 2.5683564136436523e-06, "loss": 0.5833, "step": 9480 }, { "epoch": 0.0852711435291155, "grad_norm": 19.189315795898438, "learning_rate": 2.565039282716045e-06, "loss": 0.4846, "step": 9481 }, { "epoch": 0.08528013742675594, "grad_norm": 4.080819606781006, "learning_rate": 2.5617241794360175e-06, "loss": 0.3284, "step": 9482 }, { "epoch": 0.0852891313243964, "grad_norm": 1.9336843490600586, "learning_rate": 2.558411104103198e-06, "loss": 0.1533, "step": 9483 }, { "epoch": 0.08529812522203685, "grad_norm": 2.617645502090454, "learning_rate": 2.555100057017004e-06, "loss": 0.3786, "step": 9484 }, { "epoch": 0.0853071191196773, "grad_norm": 2.388834238052368, "learning_rate": 2.5517910384766946e-06, "loss": 0.3596, "step": 9485 }, { "epoch": 0.08531611301731774, "grad_norm": 2.111809730529785, "learning_rate": 2.54848404878133e-06, "loss": 0.2475, "step": 9486 }, { "epoch": 0.0853251069149582, "grad_norm": 2.3988468647003174, "learning_rate": 2.545179088229796e-06, "loss": 0.42, "step": 9487 }, { "epoch": 0.08533410081259865, "grad_norm": 2.7583999633789062, "learning_rate": 2.541876157120793e-06, "loss": 0.3301, "step": 9488 }, { "epoch": 0.0853430947102391, "grad_norm": 15.69168758392334, "learning_rate": 2.538575255752826e-06, "loss": 0.4664, "step": 9489 }, { "epoch": 0.08535208860787956, "grad_norm": 29.93404197692871, "learning_rate": 2.53527638442424e-06, "loss": 0.7295, "step": 9490 }, { "epoch": 0.08536108250552, "grad_norm": 17.005455017089844, "learning_rate": 2.531979543433172e-06, "loss": 0.4115, "step": 9491 }, { "epoch": 0.08537007640316045, "grad_norm": 2.5028131008148193, "learning_rate": 2.5286847330775963e-06, "loss": 0.4106, "step": 9492 }, { "epoch": 0.08537907030080091, "grad_norm": 22.111228942871094, "learning_rate": 2.52539195365529e-06, "loss": 0.5511, "step": 9493 }, { "epoch": 0.08538806419844136, "grad_norm": 2.396345615386963, "learning_rate": 2.522101205463853e-06, "loss": 0.6961, "step": 9494 }, { "epoch": 0.08539705809608181, "grad_norm": 3.1891682147979736, "learning_rate": 2.518812488800698e-06, "loss": 0.3326, "step": 9495 }, { "epoch": 0.08540605199372225, "grad_norm": 10.696627616882324, "learning_rate": 2.515525803963048e-06, "loss": 0.7616, "step": 9496 }, { "epoch": 0.08541504589136271, "grad_norm": 3.3830254077911377, "learning_rate": 2.5122411512479652e-06, "loss": 0.1632, "step": 9497 }, { "epoch": 0.08542403978900316, "grad_norm": 18.55963134765625, "learning_rate": 2.5089585309523045e-06, "loss": 0.7807, "step": 9498 }, { "epoch": 0.08543303368664361, "grad_norm": 6.377307415008545, "learning_rate": 2.5056779433727416e-06, "loss": 0.6902, "step": 9499 }, { "epoch": 0.08544202758428407, "grad_norm": 2.826587438583374, "learning_rate": 2.502399388805782e-06, "loss": 0.7826, "step": 9500 }, { "epoch": 0.08545102148192452, "grad_norm": 1.7489840984344482, "learning_rate": 2.499122867547729e-06, "loss": 0.3381, "step": 9501 }, { "epoch": 0.08546001537956496, "grad_norm": 14.833584785461426, "learning_rate": 2.495848379894719e-06, "loss": 1.0805, "step": 9502 }, { "epoch": 0.08546900927720541, "grad_norm": 20.304306030273438, "learning_rate": 2.492575926142687e-06, "loss": 0.488, "step": 9503 }, { "epoch": 0.08547800317484587, "grad_norm": 19.525833129882812, "learning_rate": 2.4893055065874083e-06, "loss": 0.8297, "step": 9504 }, { "epoch": 0.08548699707248632, "grad_norm": 7.045970439910889, "learning_rate": 2.4860371215244484e-06, "loss": 0.5869, "step": 9505 }, { "epoch": 0.08549599097012676, "grad_norm": 2.5358314514160156, "learning_rate": 2.482770771249199e-06, "loss": 0.6409, "step": 9506 }, { "epoch": 0.08550498486776723, "grad_norm": 1.8922733068466187, "learning_rate": 2.4795064560568794e-06, "loss": 0.3031, "step": 9507 }, { "epoch": 0.08551397876540767, "grad_norm": 2.813944101333618, "learning_rate": 2.476244176242509e-06, "loss": 0.422, "step": 9508 }, { "epoch": 0.08552297266304812, "grad_norm": 23.700780868530273, "learning_rate": 2.472983932100928e-06, "loss": 0.541, "step": 9509 }, { "epoch": 0.08553196656068857, "grad_norm": 2.637640953063965, "learning_rate": 2.469725723926797e-06, "loss": 0.7247, "step": 9510 }, { "epoch": 0.08554096045832903, "grad_norm": 14.024820327758789, "learning_rate": 2.466469552014583e-06, "loss": 0.2864, "step": 9511 }, { "epoch": 0.08554995435596947, "grad_norm": 2.1995534896850586, "learning_rate": 2.463215416658582e-06, "loss": 0.19, "step": 9512 }, { "epoch": 0.08555894825360992, "grad_norm": 2.196166515350342, "learning_rate": 2.459963318152894e-06, "loss": 0.6558, "step": 9513 }, { "epoch": 0.08556794215125038, "grad_norm": 1.8787803649902344, "learning_rate": 2.4567132567914485e-06, "loss": 0.2458, "step": 9514 }, { "epoch": 0.08557693604889083, "grad_norm": 2.5143184661865234, "learning_rate": 2.453465232867971e-06, "loss": 0.4561, "step": 9515 }, { "epoch": 0.08558592994653128, "grad_norm": 2.649121046066284, "learning_rate": 2.4502192466760276e-06, "loss": 0.3037, "step": 9516 }, { "epoch": 0.08559492384417174, "grad_norm": 13.883164405822754, "learning_rate": 2.446975298508977e-06, "loss": 0.2377, "step": 9517 }, { "epoch": 0.08560391774181218, "grad_norm": 8.927351951599121, "learning_rate": 2.443733388660005e-06, "loss": 0.676, "step": 9518 }, { "epoch": 0.08561291163945263, "grad_norm": 3.9072227478027344, "learning_rate": 2.4404935174221153e-06, "loss": 0.3157, "step": 9519 }, { "epoch": 0.08562190553709308, "grad_norm": 16.768709182739258, "learning_rate": 2.4372556850881245e-06, "loss": 0.5497, "step": 9520 }, { "epoch": 0.08563089943473354, "grad_norm": 4.237083911895752, "learning_rate": 2.434019891950659e-06, "loss": 0.2306, "step": 9521 }, { "epoch": 0.08563989333237398, "grad_norm": 2.4882092475891113, "learning_rate": 2.4307861383021714e-06, "loss": 0.5831, "step": 9522 }, { "epoch": 0.08564888723001443, "grad_norm": 2.331524133682251, "learning_rate": 2.427554424434925e-06, "loss": 0.2764, "step": 9523 }, { "epoch": 0.08565788112765489, "grad_norm": 2.9415273666381836, "learning_rate": 2.4243247506409964e-06, "loss": 0.3254, "step": 9524 }, { "epoch": 0.08566687502529534, "grad_norm": 3.2992615699768066, "learning_rate": 2.4210971172122733e-06, "loss": 0.4979, "step": 9525 }, { "epoch": 0.08567586892293579, "grad_norm": 3.521392583847046, "learning_rate": 2.4178715244404794e-06, "loss": 0.3542, "step": 9526 }, { "epoch": 0.08568486282057623, "grad_norm": 3.0344655513763428, "learning_rate": 2.414647972617129e-06, "loss": 0.0715, "step": 9527 }, { "epoch": 0.0856938567182167, "grad_norm": 3.340543270111084, "learning_rate": 2.4114264620335725e-06, "loss": 0.495, "step": 9528 }, { "epoch": 0.08570285061585714, "grad_norm": 13.211137771606445, "learning_rate": 2.4082069929809633e-06, "loss": 0.8536, "step": 9529 }, { "epoch": 0.08571184451349759, "grad_norm": 1.750065565109253, "learning_rate": 2.4049895657502654e-06, "loss": 0.6327, "step": 9530 }, { "epoch": 0.08572083841113805, "grad_norm": 2.143393039703369, "learning_rate": 2.401774180632277e-06, "loss": 0.2651, "step": 9531 }, { "epoch": 0.0857298323087785, "grad_norm": 2.509995937347412, "learning_rate": 2.3985608379175962e-06, "loss": 0.5038, "step": 9532 }, { "epoch": 0.08573882620641894, "grad_norm": 2.765270233154297, "learning_rate": 2.3953495378966434e-06, "loss": 0.6275, "step": 9533 }, { "epoch": 0.0857478201040594, "grad_norm": 2.150019884109497, "learning_rate": 2.3921402808596562e-06, "loss": 0.446, "step": 9534 }, { "epoch": 0.08575681400169985, "grad_norm": 8.322054862976074, "learning_rate": 2.3889330670966743e-06, "loss": 0.7445, "step": 9535 }, { "epoch": 0.0857658078993403, "grad_norm": 21.087223052978516, "learning_rate": 2.3857278968975714e-06, "loss": 0.4439, "step": 9536 }, { "epoch": 0.08577480179698074, "grad_norm": 2.4440252780914307, "learning_rate": 2.3825247705520164e-06, "loss": 0.5465, "step": 9537 }, { "epoch": 0.0857837956946212, "grad_norm": 2.4335973262786865, "learning_rate": 2.379323688349516e-06, "loss": 0.3932, "step": 9538 }, { "epoch": 0.08579278959226165, "grad_norm": 2.0799670219421387, "learning_rate": 2.37612465057937e-06, "loss": 0.3952, "step": 9539 }, { "epoch": 0.0858017834899021, "grad_norm": 2.245213508605957, "learning_rate": 2.3729276575307157e-06, "loss": 0.3512, "step": 9540 }, { "epoch": 0.08581077738754256, "grad_norm": 2.2693030834198, "learning_rate": 2.3697327094924865e-06, "loss": 0.4195, "step": 9541 }, { "epoch": 0.085819771285183, "grad_norm": 1.8791792392730713, "learning_rate": 2.3665398067534374e-06, "loss": 0.7123, "step": 9542 }, { "epoch": 0.08582876518282345, "grad_norm": 2.0120136737823486, "learning_rate": 2.363348949602143e-06, "loss": 0.2366, "step": 9543 }, { "epoch": 0.0858377590804639, "grad_norm": 2.651966094970703, "learning_rate": 2.360160138326986e-06, "loss": 0.3119, "step": 9544 }, { "epoch": 0.08584675297810436, "grad_norm": 14.346939086914062, "learning_rate": 2.356973373216173e-06, "loss": 0.3485, "step": 9545 }, { "epoch": 0.0858557468757448, "grad_norm": 1.789201259613037, "learning_rate": 2.3537886545577203e-06, "loss": 0.2489, "step": 9546 }, { "epoch": 0.08586474077338525, "grad_norm": 2.0454533100128174, "learning_rate": 2.3506059826394505e-06, "loss": 0.1909, "step": 9547 }, { "epoch": 0.08587373467102571, "grad_norm": 2.550328493118286, "learning_rate": 2.3474253577490195e-06, "loss": 0.4231, "step": 9548 }, { "epoch": 0.08588272856866616, "grad_norm": 12.014039993286133, "learning_rate": 2.3442467801738863e-06, "loss": 0.6573, "step": 9549 }, { "epoch": 0.08589172246630661, "grad_norm": 1.6459977626800537, "learning_rate": 2.3410702502013272e-06, "loss": 0.5892, "step": 9550 }, { "epoch": 0.08590071636394706, "grad_norm": 3.000657796859741, "learning_rate": 2.3378957681184283e-06, "loss": 0.5781, "step": 9551 }, { "epoch": 0.08590971026158752, "grad_norm": 3.106677770614624, "learning_rate": 2.334723334212102e-06, "loss": 0.2026, "step": 9552 }, { "epoch": 0.08591870415922796, "grad_norm": 1.9744471311569214, "learning_rate": 2.3315529487690697e-06, "loss": 0.3287, "step": 9553 }, { "epoch": 0.08592769805686841, "grad_norm": 1.8726552724838257, "learning_rate": 2.3283846120758625e-06, "loss": 0.2598, "step": 9554 }, { "epoch": 0.08593669195450887, "grad_norm": 2.1141297817230225, "learning_rate": 2.3252183244188376e-06, "loss": 0.4005, "step": 9555 }, { "epoch": 0.08594568585214932, "grad_norm": 2.3847877979278564, "learning_rate": 2.3220540860841496e-06, "loss": 0.4652, "step": 9556 }, { "epoch": 0.08595467974978976, "grad_norm": 2.1709065437316895, "learning_rate": 2.3188918973577944e-06, "loss": 0.4155, "step": 9557 }, { "epoch": 0.08596367364743022, "grad_norm": 2.138018846511841, "learning_rate": 2.3157317585255606e-06, "loss": 0.4872, "step": 9558 }, { "epoch": 0.08597266754507067, "grad_norm": 12.653657913208008, "learning_rate": 2.3125736698730498e-06, "loss": 0.5632, "step": 9559 }, { "epoch": 0.08598166144271112, "grad_norm": 15.975092887878418, "learning_rate": 2.309417631685698e-06, "loss": 0.3881, "step": 9560 }, { "epoch": 0.08599065534035157, "grad_norm": 1.9277199506759644, "learning_rate": 2.3062636442487377e-06, "loss": 0.7074, "step": 9561 }, { "epoch": 0.08599964923799203, "grad_norm": 2.6024467945098877, "learning_rate": 2.30311170784723e-06, "loss": 0.2914, "step": 9562 }, { "epoch": 0.08600864313563247, "grad_norm": 1.9357316493988037, "learning_rate": 2.299961822766039e-06, "loss": 0.3964, "step": 9563 }, { "epoch": 0.08601763703327292, "grad_norm": 1.6334236860275269, "learning_rate": 2.296813989289848e-06, "loss": 0.2056, "step": 9564 }, { "epoch": 0.08602663093091338, "grad_norm": 2.031198501586914, "learning_rate": 2.2936682077031535e-06, "loss": 0.363, "step": 9565 }, { "epoch": 0.08603562482855383, "grad_norm": 18.700061798095703, "learning_rate": 2.290524478290265e-06, "loss": 0.3926, "step": 9566 }, { "epoch": 0.08604461872619427, "grad_norm": 5.6434502601623535, "learning_rate": 2.2873828013353193e-06, "loss": 0.916, "step": 9567 }, { "epoch": 0.08605361262383472, "grad_norm": 9.695127487182617, "learning_rate": 2.2842431771222445e-06, "loss": 0.8563, "step": 9568 }, { "epoch": 0.08606260652147518, "grad_norm": 3.056016206741333, "learning_rate": 2.2811056059348105e-06, "loss": 0.6017, "step": 9569 }, { "epoch": 0.08607160041911563, "grad_norm": 2.4778199195861816, "learning_rate": 2.277970088056583e-06, "loss": 0.6229, "step": 9570 }, { "epoch": 0.08608059431675608, "grad_norm": 1.7745670080184937, "learning_rate": 2.2748366237709374e-06, "loss": 0.6682, "step": 9571 }, { "epoch": 0.08608958821439654, "grad_norm": 3.2467076778411865, "learning_rate": 2.2717052133610863e-06, "loss": 0.2119, "step": 9572 }, { "epoch": 0.08609858211203698, "grad_norm": 15.355948448181152, "learning_rate": 2.2685758571100306e-06, "loss": 0.3372, "step": 9573 }, { "epoch": 0.08610757600967743, "grad_norm": 6.696120262145996, "learning_rate": 2.2654485553006116e-06, "loss": 0.781, "step": 9574 }, { "epoch": 0.08611656990731789, "grad_norm": 1.81423020362854, "learning_rate": 2.2623233082154638e-06, "loss": 0.3742, "step": 9575 }, { "epoch": 0.08612556380495834, "grad_norm": 2.3714756965637207, "learning_rate": 2.2592001161370392e-06, "loss": 0.2659, "step": 9576 }, { "epoch": 0.08613455770259879, "grad_norm": 1.6056122779846191, "learning_rate": 2.25607897934762e-06, "loss": 0.0269, "step": 9577 }, { "epoch": 0.08614355160023923, "grad_norm": 4.079221248626709, "learning_rate": 2.2529598981292838e-06, "loss": 0.3882, "step": 9578 }, { "epoch": 0.08615254549787969, "grad_norm": 10.946501731872559, "learning_rate": 2.24984287276393e-06, "loss": 0.5037, "step": 9579 }, { "epoch": 0.08616153939552014, "grad_norm": 2.0312554836273193, "learning_rate": 2.2467279035332688e-06, "loss": 0.4054, "step": 9580 }, { "epoch": 0.08617053329316059, "grad_norm": 2.33662486076355, "learning_rate": 2.243614990718837e-06, "loss": 0.2278, "step": 9581 }, { "epoch": 0.08617952719080105, "grad_norm": 2.4808366298675537, "learning_rate": 2.24050413460197e-06, "loss": 0.4001, "step": 9582 }, { "epoch": 0.0861885210884415, "grad_norm": 3.686713695526123, "learning_rate": 2.237395335463821e-06, "loss": 0.2937, "step": 9583 }, { "epoch": 0.08619751498608194, "grad_norm": 4.303175449371338, "learning_rate": 2.234288593585368e-06, "loss": 0.3209, "step": 9584 }, { "epoch": 0.08620650888372239, "grad_norm": 3.144434928894043, "learning_rate": 2.2311839092473864e-06, "loss": 0.349, "step": 9585 }, { "epoch": 0.08621550278136285, "grad_norm": 16.885595321655273, "learning_rate": 2.2280812827304846e-06, "loss": 0.72, "step": 9586 }, { "epoch": 0.0862244966790033, "grad_norm": 18.20671272277832, "learning_rate": 2.224980714315067e-06, "loss": 0.7718, "step": 9587 }, { "epoch": 0.08623349057664374, "grad_norm": 3.462977170944214, "learning_rate": 2.2218822042813556e-06, "loss": 0.3151, "step": 9588 }, { "epoch": 0.0862424844742842, "grad_norm": 2.7201766967773438, "learning_rate": 2.2187857529094023e-06, "loss": 0.2439, "step": 9589 }, { "epoch": 0.08625147837192465, "grad_norm": 20.229528427124023, "learning_rate": 2.2156913604790554e-06, "loss": 0.4532, "step": 9590 }, { "epoch": 0.0862604722695651, "grad_norm": 2.361391544342041, "learning_rate": 2.2125990272699792e-06, "loss": 0.0777, "step": 9591 }, { "epoch": 0.08626946616720556, "grad_norm": 2.021571159362793, "learning_rate": 2.2095087535616566e-06, "loss": 0.3728, "step": 9592 }, { "epoch": 0.086278460064846, "grad_norm": 3.2938437461853027, "learning_rate": 2.2064205396333886e-06, "loss": 0.4298, "step": 9593 }, { "epoch": 0.08628745396248645, "grad_norm": 2.653848171234131, "learning_rate": 2.20333438576428e-06, "loss": 0.3403, "step": 9594 }, { "epoch": 0.0862964478601269, "grad_norm": 1.8881056308746338, "learning_rate": 2.2002502922332526e-06, "loss": 0.5233, "step": 9595 }, { "epoch": 0.08630544175776736, "grad_norm": 1.7546637058258057, "learning_rate": 2.197168259319049e-06, "loss": 0.4973, "step": 9596 }, { "epoch": 0.0863144356554078, "grad_norm": 18.125537872314453, "learning_rate": 2.1940882873002137e-06, "loss": 0.6185, "step": 9597 }, { "epoch": 0.08632342955304825, "grad_norm": 2.449212074279785, "learning_rate": 2.1910103764551187e-06, "loss": 0.5185, "step": 9598 }, { "epoch": 0.08633242345068871, "grad_norm": 2.945300817489624, "learning_rate": 2.1879345270619357e-06, "loss": 0.3868, "step": 9599 }, { "epoch": 0.08634141734832916, "grad_norm": 2.2764101028442383, "learning_rate": 2.184860739398656e-06, "loss": 0.435, "step": 9600 }, { "epoch": 0.08635041124596961, "grad_norm": 2.534269094467163, "learning_rate": 2.1817890137430934e-06, "loss": 0.6432, "step": 9601 }, { "epoch": 0.08635940514361005, "grad_norm": 2.873824119567871, "learning_rate": 2.1787193503728572e-06, "loss": 0.4463, "step": 9602 }, { "epoch": 0.08636839904125052, "grad_norm": 3.3243532180786133, "learning_rate": 2.175651749565388e-06, "loss": 0.5243, "step": 9603 }, { "epoch": 0.08637739293889096, "grad_norm": 6.872348308563232, "learning_rate": 2.1725862115979322e-06, "loss": 0.6466, "step": 9604 }, { "epoch": 0.08638638683653141, "grad_norm": 2.065627336502075, "learning_rate": 2.169522736747545e-06, "loss": 0.3021, "step": 9605 }, { "epoch": 0.08639538073417187, "grad_norm": 27.75898551940918, "learning_rate": 2.1664613252911025e-06, "loss": 0.7871, "step": 9606 }, { "epoch": 0.08640437463181232, "grad_norm": 22.297895431518555, "learning_rate": 2.1634019775052863e-06, "loss": 0.8125, "step": 9607 }, { "epoch": 0.08641336852945276, "grad_norm": 2.334813117980957, "learning_rate": 2.160344693666605e-06, "loss": 0.3016, "step": 9608 }, { "epoch": 0.08642236242709321, "grad_norm": 15.89322280883789, "learning_rate": 2.157289474051369e-06, "loss": 0.5061, "step": 9609 }, { "epoch": 0.08643135632473367, "grad_norm": 3.140714645385742, "learning_rate": 2.1542363189357066e-06, "loss": 0.2459, "step": 9610 }, { "epoch": 0.08644035022237412, "grad_norm": 2.4492952823638916, "learning_rate": 2.151185228595559e-06, "loss": 0.5434, "step": 9611 }, { "epoch": 0.08644934412001457, "grad_norm": 4.304100513458252, "learning_rate": 2.1481362033066776e-06, "loss": 0.382, "step": 9612 }, { "epoch": 0.08645833801765503, "grad_norm": 3.6109261512756348, "learning_rate": 2.1450892433446336e-06, "loss": 0.3197, "step": 9613 }, { "epoch": 0.08646733191529547, "grad_norm": 3.0183043479919434, "learning_rate": 2.1420443489848036e-06, "loss": 0.414, "step": 9614 }, { "epoch": 0.08647632581293592, "grad_norm": 2.0153942108154297, "learning_rate": 2.13900152050239e-06, "loss": 0.2753, "step": 9615 }, { "epoch": 0.08648531971057638, "grad_norm": 2.2939672470092773, "learning_rate": 2.1359607581723895e-06, "loss": 0.2234, "step": 9616 }, { "epoch": 0.08649431360821683, "grad_norm": 2.6051807403564453, "learning_rate": 2.1329220622696323e-06, "loss": 0.6819, "step": 9617 }, { "epoch": 0.08650330750585727, "grad_norm": 1.9278583526611328, "learning_rate": 2.1298854330687513e-06, "loss": 0.5551, "step": 9618 }, { "epoch": 0.08651230140349772, "grad_norm": 14.928765296936035, "learning_rate": 2.126850870844188e-06, "loss": 0.4725, "step": 9619 }, { "epoch": 0.08652129530113818, "grad_norm": 39.12944793701172, "learning_rate": 2.1238183758702096e-06, "loss": 0.7872, "step": 9620 }, { "epoch": 0.08653028919877863, "grad_norm": 1.832507610321045, "learning_rate": 2.1207879484208797e-06, "loss": 0.215, "step": 9621 }, { "epoch": 0.08653928309641908, "grad_norm": 15.512027740478516, "learning_rate": 2.1177595887700963e-06, "loss": 0.6225, "step": 9622 }, { "epoch": 0.08654827699405954, "grad_norm": 1.6923744678497314, "learning_rate": 2.114733297191554e-06, "loss": 0.3424, "step": 9623 }, { "epoch": 0.08655727089169998, "grad_norm": 1.962090015411377, "learning_rate": 2.111709073958762e-06, "loss": 0.2916, "step": 9624 }, { "epoch": 0.08656626478934043, "grad_norm": 2.0746610164642334, "learning_rate": 2.1086869193450544e-06, "loss": 0.4747, "step": 9625 }, { "epoch": 0.08657525868698088, "grad_norm": 2.480990171432495, "learning_rate": 2.1056668336235622e-06, "loss": 0.1804, "step": 9626 }, { "epoch": 0.08658425258462134, "grad_norm": 1.8842133283615112, "learning_rate": 2.102648817067249e-06, "loss": 0.2633, "step": 9627 }, { "epoch": 0.08659324648226178, "grad_norm": 16.97816276550293, "learning_rate": 2.099632869948867e-06, "loss": 0.5556, "step": 9628 }, { "epoch": 0.08660224037990223, "grad_norm": 14.873313903808594, "learning_rate": 2.0966189925410023e-06, "loss": 0.7083, "step": 9629 }, { "epoch": 0.08661123427754269, "grad_norm": 3.007197856903076, "learning_rate": 2.093607185116045e-06, "loss": 0.4262, "step": 9630 }, { "epoch": 0.08662022817518314, "grad_norm": 2.0665805339813232, "learning_rate": 2.090597447946194e-06, "loss": 0.4594, "step": 9631 }, { "epoch": 0.08662922207282359, "grad_norm": 20.189451217651367, "learning_rate": 2.0875897813034727e-06, "loss": 0.736, "step": 9632 }, { "epoch": 0.08663821597046405, "grad_norm": 3.0001986026763916, "learning_rate": 2.0845841854597094e-06, "loss": 0.6557, "step": 9633 }, { "epoch": 0.0866472098681045, "grad_norm": 2.3889217376708984, "learning_rate": 2.081580660686544e-06, "loss": 0.4331, "step": 9634 }, { "epoch": 0.08665620376574494, "grad_norm": 2.420957088470459, "learning_rate": 2.078579207255432e-06, "loss": 0.1605, "step": 9635 }, { "epoch": 0.08666519766338539, "grad_norm": 1.7880176305770874, "learning_rate": 2.0755798254376374e-06, "loss": 0.2707, "step": 9636 }, { "epoch": 0.08667419156102585, "grad_norm": 20.28321647644043, "learning_rate": 2.072582515504254e-06, "loss": 0.5098, "step": 9637 }, { "epoch": 0.0866831854586663, "grad_norm": 2.927788496017456, "learning_rate": 2.069587277726159e-06, "loss": 0.5714, "step": 9638 }, { "epoch": 0.08669217935630674, "grad_norm": 11.538216590881348, "learning_rate": 2.0665941123740735e-06, "loss": 0.8062, "step": 9639 }, { "epoch": 0.0867011732539472, "grad_norm": 13.863561630249023, "learning_rate": 2.0636030197185076e-06, "loss": 0.8051, "step": 9640 }, { "epoch": 0.08671016715158765, "grad_norm": 16.10287857055664, "learning_rate": 2.060614000029798e-06, "loss": 0.408, "step": 9641 }, { "epoch": 0.0867191610492281, "grad_norm": 1.772739052772522, "learning_rate": 2.0576270535780875e-06, "loss": 0.1491, "step": 9642 }, { "epoch": 0.08672815494686854, "grad_norm": 7.253314971923828, "learning_rate": 2.054642180633326e-06, "loss": 0.6426, "step": 9643 }, { "epoch": 0.086737148844509, "grad_norm": 3.805497646331787, "learning_rate": 2.0516593814652957e-06, "loss": 0.3502, "step": 9644 }, { "epoch": 0.08674614274214945, "grad_norm": 14.363755226135254, "learning_rate": 2.0486786563435723e-06, "loss": 0.624, "step": 9645 }, { "epoch": 0.0867551366397899, "grad_norm": 2.0775015354156494, "learning_rate": 2.045700005537546e-06, "loss": 0.3836, "step": 9646 }, { "epoch": 0.08676413053743036, "grad_norm": 9.603816986083984, "learning_rate": 2.0427234293164343e-06, "loss": 0.5047, "step": 9647 }, { "epoch": 0.0867731244350708, "grad_norm": 3.159724473953247, "learning_rate": 2.0397489279492506e-06, "loss": 0.615, "step": 9648 }, { "epoch": 0.08678211833271125, "grad_norm": 1.8404182195663452, "learning_rate": 2.036776501704829e-06, "loss": 0.4637, "step": 9649 }, { "epoch": 0.0867911122303517, "grad_norm": 19.50966453552246, "learning_rate": 2.033806150851808e-06, "loss": 0.5153, "step": 9650 }, { "epoch": 0.08680010612799216, "grad_norm": 2.6862316131591797, "learning_rate": 2.030837875658656e-06, "loss": 0.0605, "step": 9651 }, { "epoch": 0.08680910002563261, "grad_norm": 1.83784019947052, "learning_rate": 2.0278716763936334e-06, "loss": 0.3483, "step": 9652 }, { "epoch": 0.08681809392327305, "grad_norm": 8.964920997619629, "learning_rate": 2.024907553324823e-06, "loss": 0.4978, "step": 9653 }, { "epoch": 0.08682708782091351, "grad_norm": 2.4592087268829346, "learning_rate": 2.0219455067201243e-06, "loss": 0.882, "step": 9654 }, { "epoch": 0.08683608171855396, "grad_norm": 2.8874287605285645, "learning_rate": 2.018985536847237e-06, "loss": 0.1886, "step": 9655 }, { "epoch": 0.08684507561619441, "grad_norm": 10.78700065612793, "learning_rate": 2.016027643973689e-06, "loss": 0.6232, "step": 9656 }, { "epoch": 0.08685406951383487, "grad_norm": 2.0402326583862305, "learning_rate": 2.0130718283668027e-06, "loss": 0.1941, "step": 9657 }, { "epoch": 0.08686306341147532, "grad_norm": 19.481279373168945, "learning_rate": 2.010118090293728e-06, "loss": 0.4484, "step": 9658 }, { "epoch": 0.08687205730911576, "grad_norm": 9.368060111999512, "learning_rate": 2.007166430021415e-06, "loss": 0.4377, "step": 9659 }, { "epoch": 0.08688105120675621, "grad_norm": 3.339308977127075, "learning_rate": 2.0042168478166368e-06, "loss": 0.2712, "step": 9660 }, { "epoch": 0.08689004510439667, "grad_norm": 2.505600929260254, "learning_rate": 2.0012693439459716e-06, "loss": 0.3328, "step": 9661 }, { "epoch": 0.08689903900203712, "grad_norm": 2.2200069427490234, "learning_rate": 1.9983239186758067e-06, "loss": 0.3136, "step": 9662 }, { "epoch": 0.08690803289967756, "grad_norm": 14.606632232666016, "learning_rate": 1.995380572272354e-06, "loss": 0.8209, "step": 9663 }, { "epoch": 0.08691702679731803, "grad_norm": 2.0567100048065186, "learning_rate": 1.9924393050016254e-06, "loss": 0.3294, "step": 9664 }, { "epoch": 0.08692602069495847, "grad_norm": 2.3378868103027344, "learning_rate": 1.9895001171294475e-06, "loss": 0.5322, "step": 9665 }, { "epoch": 0.08693501459259892, "grad_norm": 2.6362812519073486, "learning_rate": 1.9865630089214687e-06, "loss": 0.4114, "step": 9666 }, { "epoch": 0.08694400849023937, "grad_norm": 15.783991813659668, "learning_rate": 1.9836279806431324e-06, "loss": 0.7741, "step": 9667 }, { "epoch": 0.08695300238787983, "grad_norm": 10.642529487609863, "learning_rate": 1.9806950325597153e-06, "loss": 0.618, "step": 9668 }, { "epoch": 0.08696199628552027, "grad_norm": 3.3231678009033203, "learning_rate": 1.977764164936277e-06, "loss": 0.2418, "step": 9669 }, { "epoch": 0.08697099018316072, "grad_norm": 1.970430850982666, "learning_rate": 1.9748353780377234e-06, "loss": 0.5421, "step": 9670 }, { "epoch": 0.08697998408080118, "grad_norm": 2.43890643119812, "learning_rate": 1.97190867212875e-06, "loss": 0.5218, "step": 9671 }, { "epoch": 0.08698897797844163, "grad_norm": 19.597335815429688, "learning_rate": 1.96898404747386e-06, "loss": 0.3846, "step": 9672 }, { "epoch": 0.08699797187608208, "grad_norm": 19.45095443725586, "learning_rate": 1.9660615043373883e-06, "loss": 0.2195, "step": 9673 }, { "epoch": 0.08700696577372254, "grad_norm": 2.604999542236328, "learning_rate": 1.9631410429834683e-06, "loss": 0.3309, "step": 9674 }, { "epoch": 0.08701595967136298, "grad_norm": 2.5117998123168945, "learning_rate": 1.960222663676048e-06, "loss": 0.1805, "step": 9675 }, { "epoch": 0.08702495356900343, "grad_norm": 3.897082567214966, "learning_rate": 1.9573063666788875e-06, "loss": 0.2533, "step": 9676 }, { "epoch": 0.08703394746664388, "grad_norm": 3.0702950954437256, "learning_rate": 1.9543921522555517e-06, "loss": 0.5771, "step": 9677 }, { "epoch": 0.08704294136428434, "grad_norm": 3.1275010108947754, "learning_rate": 1.9514800206694383e-06, "loss": 0.4382, "step": 9678 }, { "epoch": 0.08705193526192478, "grad_norm": 15.408703804016113, "learning_rate": 1.948569972183728e-06, "loss": 0.56, "step": 9679 }, { "epoch": 0.08706092915956523, "grad_norm": 2.7715413570404053, "learning_rate": 1.9456620070614385e-06, "loss": 0.4885, "step": 9680 }, { "epoch": 0.08706992305720569, "grad_norm": 15.680237770080566, "learning_rate": 1.9427561255653816e-06, "loss": 0.3663, "step": 9681 }, { "epoch": 0.08707891695484614, "grad_norm": 3.454932451248169, "learning_rate": 1.9398523279581944e-06, "loss": 0.7685, "step": 9682 }, { "epoch": 0.08708791085248659, "grad_norm": 1.9935405254364014, "learning_rate": 1.936950614502317e-06, "loss": 0.3695, "step": 9683 }, { "epoch": 0.08709690475012703, "grad_norm": 2.999997138977051, "learning_rate": 1.9340509854599954e-06, "loss": 0.6755, "step": 9684 }, { "epoch": 0.0871058986477675, "grad_norm": 2.7419486045837402, "learning_rate": 1.9311534410933056e-06, "loss": 0.6174, "step": 9685 }, { "epoch": 0.08711489254540794, "grad_norm": 2.4722437858581543, "learning_rate": 1.9282579816641168e-06, "loss": 0.2603, "step": 9686 }, { "epoch": 0.08712388644304839, "grad_norm": 2.622192621231079, "learning_rate": 1.925364607434124e-06, "loss": 0.4951, "step": 9687 }, { "epoch": 0.08713288034068885, "grad_norm": 17.64230728149414, "learning_rate": 1.9224733186648245e-06, "loss": 0.5104, "step": 9688 }, { "epoch": 0.0871418742383293, "grad_norm": 2.567950963973999, "learning_rate": 1.9195841156175314e-06, "loss": 0.309, "step": 9689 }, { "epoch": 0.08715086813596974, "grad_norm": 8.9426908493042, "learning_rate": 1.9166969985533634e-06, "loss": 0.4978, "step": 9690 }, { "epoch": 0.08715986203361019, "grad_norm": 2.402336597442627, "learning_rate": 1.9138119677332532e-06, "loss": 0.3312, "step": 9691 }, { "epoch": 0.08716885593125065, "grad_norm": 17.854761123657227, "learning_rate": 1.9109290234179567e-06, "loss": 0.3095, "step": 9692 }, { "epoch": 0.0871778498288911, "grad_norm": 12.812525749206543, "learning_rate": 1.9080481658680206e-06, "loss": 0.5213, "step": 9693 }, { "epoch": 0.08718684372653154, "grad_norm": 21.247356414794922, "learning_rate": 1.9051693953438226e-06, "loss": 0.7606, "step": 9694 }, { "epoch": 0.087195837624172, "grad_norm": 2.7413055896759033, "learning_rate": 1.9022927121055384e-06, "loss": 0.2692, "step": 9695 }, { "epoch": 0.08720483152181245, "grad_norm": 8.566044807434082, "learning_rate": 1.8994181164131569e-06, "loss": 0.61, "step": 9696 }, { "epoch": 0.0872138254194529, "grad_norm": 11.142141342163086, "learning_rate": 1.8965456085264871e-06, "loss": 0.6522, "step": 9697 }, { "epoch": 0.08722281931709336, "grad_norm": 9.227420806884766, "learning_rate": 1.893675188705138e-06, "loss": 0.643, "step": 9698 }, { "epoch": 0.0872318132147338, "grad_norm": 13.214836120605469, "learning_rate": 1.890806857208538e-06, "loss": 0.533, "step": 9699 }, { "epoch": 0.08724080711237425, "grad_norm": 2.5840389728546143, "learning_rate": 1.8879406142959272e-06, "loss": 0.2134, "step": 9700 }, { "epoch": 0.0872498010100147, "grad_norm": 9.2205171585083, "learning_rate": 1.8850764602263426e-06, "loss": 0.6794, "step": 9701 }, { "epoch": 0.08725879490765516, "grad_norm": 13.95486068725586, "learning_rate": 1.8822143952586552e-06, "loss": 0.5526, "step": 9702 }, { "epoch": 0.0872677888052956, "grad_norm": 3.069917678833008, "learning_rate": 1.87935441965153e-06, "loss": 0.5472, "step": 9703 }, { "epoch": 0.08727678270293605, "grad_norm": 17.66819190979004, "learning_rate": 1.8764965336634493e-06, "loss": 0.3634, "step": 9704 }, { "epoch": 0.08728577660057651, "grad_norm": 2.670758008956909, "learning_rate": 1.8736407375527009e-06, "loss": 0.8305, "step": 9705 }, { "epoch": 0.08729477049821696, "grad_norm": 2.932675838470459, "learning_rate": 1.8707870315773978e-06, "loss": 0.2664, "step": 9706 }, { "epoch": 0.08730376439585741, "grad_norm": 2.3460233211517334, "learning_rate": 1.8679354159954505e-06, "loss": 0.4355, "step": 9707 }, { "epoch": 0.08731275829349786, "grad_norm": 1.6942137479782104, "learning_rate": 1.8650858910645775e-06, "loss": 0.2124, "step": 9708 }, { "epoch": 0.08732175219113832, "grad_norm": 9.032496452331543, "learning_rate": 1.8622384570423286e-06, "loss": 0.7884, "step": 9709 }, { "epoch": 0.08733074608877876, "grad_norm": 3.821301221847534, "learning_rate": 1.859393114186042e-06, "loss": 0.326, "step": 9710 }, { "epoch": 0.08733973998641921, "grad_norm": 2.435805082321167, "learning_rate": 1.856549862752885e-06, "loss": 0.2942, "step": 9711 }, { "epoch": 0.08734873388405967, "grad_norm": 3.1363141536712646, "learning_rate": 1.8537087029998234e-06, "loss": 0.9695, "step": 9712 }, { "epoch": 0.08735772778170012, "grad_norm": 2.8801462650299072, "learning_rate": 1.850869635183633e-06, "loss": 0.1447, "step": 9713 }, { "epoch": 0.08736672167934056, "grad_norm": 3.410707473754883, "learning_rate": 1.8480326595609166e-06, "loss": 0.3352, "step": 9714 }, { "epoch": 0.08737571557698103, "grad_norm": 2.4565818309783936, "learning_rate": 1.8451977763880691e-06, "loss": 0.5424, "step": 9715 }, { "epoch": 0.08738470947462147, "grad_norm": 2.033688545227051, "learning_rate": 1.8423649859213076e-06, "loss": 0.2531, "step": 9716 }, { "epoch": 0.08739370337226192, "grad_norm": 3.220705032348633, "learning_rate": 1.8395342884166528e-06, "loss": 0.5433, "step": 9717 }, { "epoch": 0.08740269726990237, "grad_norm": 2.213092803955078, "learning_rate": 1.836705684129944e-06, "loss": 0.4884, "step": 9718 }, { "epoch": 0.08741169116754283, "grad_norm": 11.246356010437012, "learning_rate": 1.833879173316827e-06, "loss": 0.8315, "step": 9719 }, { "epoch": 0.08742068506518327, "grad_norm": 2.1133575439453125, "learning_rate": 1.831054756232753e-06, "loss": 0.4946, "step": 9720 }, { "epoch": 0.08742967896282372, "grad_norm": 4.090102195739746, "learning_rate": 1.8282324331330008e-06, "loss": 0.2409, "step": 9721 }, { "epoch": 0.08743867286046418, "grad_norm": 15.296724319458008, "learning_rate": 1.8254122042726363e-06, "loss": 0.7135, "step": 9722 }, { "epoch": 0.08744766675810463, "grad_norm": 2.32477068901062, "learning_rate": 1.8225940699065608e-06, "loss": 0.2968, "step": 9723 }, { "epoch": 0.08745666065574507, "grad_norm": 15.905804634094238, "learning_rate": 1.8197780302894679e-06, "loss": 0.5266, "step": 9724 }, { "epoch": 0.08746565455338552, "grad_norm": 3.8122336864471436, "learning_rate": 1.8169640856758651e-06, "loss": 0.1769, "step": 9725 }, { "epoch": 0.08747464845102598, "grad_norm": 11.75333309173584, "learning_rate": 1.8141522363200797e-06, "loss": 0.6149, "step": 9726 }, { "epoch": 0.08748364234866643, "grad_norm": 12.670889854431152, "learning_rate": 1.8113424824762416e-06, "loss": 0.5188, "step": 9727 }, { "epoch": 0.08749263624630688, "grad_norm": 18.57587242126465, "learning_rate": 1.8085348243982946e-06, "loss": 0.3251, "step": 9728 }, { "epoch": 0.08750163014394734, "grad_norm": 2.6837239265441895, "learning_rate": 1.805729262339989e-06, "loss": 0.3438, "step": 9729 }, { "epoch": 0.08751062404158778, "grad_norm": 21.413005828857422, "learning_rate": 1.8029257965548912e-06, "loss": 0.6827, "step": 9730 }, { "epoch": 0.08751961793922823, "grad_norm": 1.8608872890472412, "learning_rate": 1.8001244272963763e-06, "loss": 0.5537, "step": 9731 }, { "epoch": 0.08752861183686868, "grad_norm": 28.32935905456543, "learning_rate": 1.7973251548176196e-06, "loss": 0.6578, "step": 9732 }, { "epoch": 0.08753760573450914, "grad_norm": 2.6420114040374756, "learning_rate": 1.7945279793716297e-06, "loss": 0.8363, "step": 9733 }, { "epoch": 0.08754659963214959, "grad_norm": 2.3167309761047363, "learning_rate": 1.791732901211199e-06, "loss": 0.8966, "step": 9734 }, { "epoch": 0.08755559352979003, "grad_norm": 2.933284044265747, "learning_rate": 1.7889399205889584e-06, "loss": 0.499, "step": 9735 }, { "epoch": 0.08756458742743049, "grad_norm": 2.247015953063965, "learning_rate": 1.7861490377573258e-06, "loss": 0.3394, "step": 9736 }, { "epoch": 0.08757358132507094, "grad_norm": 15.688582420349121, "learning_rate": 1.7833602529685328e-06, "loss": 0.9689, "step": 9737 }, { "epoch": 0.08758257522271139, "grad_norm": 2.2688663005828857, "learning_rate": 1.7805735664746415e-06, "loss": 0.4178, "step": 9738 }, { "epoch": 0.08759156912035185, "grad_norm": 5.685274124145508, "learning_rate": 1.7777889785274948e-06, "loss": 0.5514, "step": 9739 }, { "epoch": 0.0876005630179923, "grad_norm": 8.82175064086914, "learning_rate": 1.775006489378772e-06, "loss": 0.5898, "step": 9740 }, { "epoch": 0.08760955691563274, "grad_norm": 16.77676010131836, "learning_rate": 1.7722260992799472e-06, "loss": 0.5326, "step": 9741 }, { "epoch": 0.08761855081327319, "grad_norm": 9.058639526367188, "learning_rate": 1.7694478084823024e-06, "loss": 0.5002, "step": 9742 }, { "epoch": 0.08762754471091365, "grad_norm": 3.2491040229797363, "learning_rate": 1.766671617236948e-06, "loss": 0.1093, "step": 9743 }, { "epoch": 0.0876365386085541, "grad_norm": 17.599815368652344, "learning_rate": 1.7638975257947888e-06, "loss": 0.6839, "step": 9744 }, { "epoch": 0.08764553250619454, "grad_norm": 2.15906023979187, "learning_rate": 1.7611255344065437e-06, "loss": 0.2592, "step": 9745 }, { "epoch": 0.087654526403835, "grad_norm": 12.49172592163086, "learning_rate": 1.7583556433227372e-06, "loss": 0.6864, "step": 9746 }, { "epoch": 0.08766352030147545, "grad_norm": 15.471363067626953, "learning_rate": 1.7555878527937164e-06, "loss": 0.5743, "step": 9747 }, { "epoch": 0.0876725141991159, "grad_norm": 2.110888957977295, "learning_rate": 1.7528221630696312e-06, "loss": 0.5284, "step": 9748 }, { "epoch": 0.08768150809675634, "grad_norm": 7.656508445739746, "learning_rate": 1.7500585744004317e-06, "loss": 0.4937, "step": 9749 }, { "epoch": 0.0876905019943968, "grad_norm": 2.7950546741485596, "learning_rate": 1.747297087035904e-06, "loss": 0.5325, "step": 9750 }, { "epoch": 0.08769949589203725, "grad_norm": 2.8739724159240723, "learning_rate": 1.7445377012256126e-06, "loss": 0.3183, "step": 9751 }, { "epoch": 0.0877084897896777, "grad_norm": 3.8870766162872314, "learning_rate": 1.7417804172189606e-06, "loss": 0.4243, "step": 9752 }, { "epoch": 0.08771748368731816, "grad_norm": 3.0464255809783936, "learning_rate": 1.7390252352651403e-06, "loss": 0.5638, "step": 9753 }, { "epoch": 0.0877264775849586, "grad_norm": 2.6601507663726807, "learning_rate": 1.7362721556131639e-06, "loss": 0.3885, "step": 9754 }, { "epoch": 0.08773547148259905, "grad_norm": 4.419112682342529, "learning_rate": 1.7335211785118543e-06, "loss": 0.4746, "step": 9755 }, { "epoch": 0.08774446538023951, "grad_norm": 2.0523440837860107, "learning_rate": 1.7307723042098377e-06, "loss": 0.3522, "step": 9756 }, { "epoch": 0.08775345927787996, "grad_norm": 5.574007987976074, "learning_rate": 1.7280255329555572e-06, "loss": 0.4553, "step": 9757 }, { "epoch": 0.08776245317552041, "grad_norm": 8.291093826293945, "learning_rate": 1.7252808649972668e-06, "loss": 0.6541, "step": 9758 }, { "epoch": 0.08777144707316085, "grad_norm": 6.543187141418457, "learning_rate": 1.7225383005830209e-06, "loss": 0.586, "step": 9759 }, { "epoch": 0.08778044097080132, "grad_norm": 16.438766479492188, "learning_rate": 1.7197978399606907e-06, "loss": 0.7173, "step": 9760 }, { "epoch": 0.08778943486844176, "grad_norm": 10.391491889953613, "learning_rate": 1.7170594833779503e-06, "loss": 0.7698, "step": 9761 }, { "epoch": 0.08779842876608221, "grad_norm": 2.5981178283691406, "learning_rate": 1.7143232310823015e-06, "loss": 0.2095, "step": 9762 }, { "epoch": 0.08780742266372267, "grad_norm": 10.582021713256836, "learning_rate": 1.7115890833210358e-06, "loss": 0.2614, "step": 9763 }, { "epoch": 0.08781641656136312, "grad_norm": 13.761950492858887, "learning_rate": 1.7088570403412663e-06, "loss": 0.505, "step": 9764 }, { "epoch": 0.08782541045900356, "grad_norm": 2.8013458251953125, "learning_rate": 1.7061271023899095e-06, "loss": 0.4154, "step": 9765 }, { "epoch": 0.08783440435664401, "grad_norm": 2.7434513568878174, "learning_rate": 1.703399269713693e-06, "loss": 0.3207, "step": 9766 }, { "epoch": 0.08784339825428447, "grad_norm": 3.9271509647369385, "learning_rate": 1.7006735425591585e-06, "loss": 0.2689, "step": 9767 }, { "epoch": 0.08785239215192492, "grad_norm": 2.5061428546905518, "learning_rate": 1.6979499211726535e-06, "loss": 0.2812, "step": 9768 }, { "epoch": 0.08786138604956537, "grad_norm": 3.318490982055664, "learning_rate": 1.6952284058003366e-06, "loss": 0.517, "step": 9769 }, { "epoch": 0.08787037994720583, "grad_norm": 11.466612815856934, "learning_rate": 1.6925089966881746e-06, "loss": 0.5191, "step": 9770 }, { "epoch": 0.08787937384484627, "grad_norm": 2.27654767036438, "learning_rate": 1.6897916940819465e-06, "loss": 0.2896, "step": 9771 }, { "epoch": 0.08788836774248672, "grad_norm": 10.248982429504395, "learning_rate": 1.687076498227233e-06, "loss": 0.6132, "step": 9772 }, { "epoch": 0.08789736164012717, "grad_norm": 2.3980846405029297, "learning_rate": 1.6843634093694384e-06, "loss": 0.8303, "step": 9773 }, { "epoch": 0.08790635553776763, "grad_norm": 3.8808581829071045, "learning_rate": 1.681652427753766e-06, "loss": 0.2687, "step": 9774 }, { "epoch": 0.08791534943540807, "grad_norm": 3.3370285034179688, "learning_rate": 1.6789435536252258e-06, "loss": 0.4512, "step": 9775 }, { "epoch": 0.08792434333304852, "grad_norm": 3.5827982425689697, "learning_rate": 1.6762367872286522e-06, "loss": 0.3602, "step": 9776 }, { "epoch": 0.08793333723068898, "grad_norm": 2.1934738159179688, "learning_rate": 1.6735321288086747e-06, "loss": 0.8031, "step": 9777 }, { "epoch": 0.08794233112832943, "grad_norm": 17.079708099365234, "learning_rate": 1.6708295786097338e-06, "loss": 0.5799, "step": 9778 }, { "epoch": 0.08795132502596988, "grad_norm": 10.130239486694336, "learning_rate": 1.6681291368760927e-06, "loss": 0.8073, "step": 9779 }, { "epoch": 0.08796031892361034, "grad_norm": 3.165437936782837, "learning_rate": 1.6654308038518058e-06, "loss": 0.6359, "step": 9780 }, { "epoch": 0.08796931282125078, "grad_norm": 13.447351455688477, "learning_rate": 1.662734579780753e-06, "loss": 0.5611, "step": 9781 }, { "epoch": 0.08797830671889123, "grad_norm": 2.5436949729919434, "learning_rate": 1.6600404649066065e-06, "loss": 0.4685, "step": 9782 }, { "epoch": 0.08798730061653168, "grad_norm": 17.04020118713379, "learning_rate": 1.6573484594728683e-06, "loss": 0.1659, "step": 9783 }, { "epoch": 0.08799629451417214, "grad_norm": 2.719273328781128, "learning_rate": 1.6546585637228329e-06, "loss": 0.5397, "step": 9784 }, { "epoch": 0.08800528841181258, "grad_norm": 4.2893195152282715, "learning_rate": 1.6519707778996112e-06, "loss": 0.3224, "step": 9785 }, { "epoch": 0.08801428230945303, "grad_norm": 2.4379141330718994, "learning_rate": 1.6492851022461232e-06, "loss": 0.0761, "step": 9786 }, { "epoch": 0.08802327620709349, "grad_norm": 3.063967227935791, "learning_rate": 1.6466015370050909e-06, "loss": 0.2452, "step": 9787 }, { "epoch": 0.08803227010473394, "grad_norm": 9.164079666137695, "learning_rate": 1.643920082419062e-06, "loss": 0.6935, "step": 9788 }, { "epoch": 0.08804126400237439, "grad_norm": 10.397014617919922, "learning_rate": 1.6412407387303792e-06, "loss": 0.5167, "step": 9789 }, { "epoch": 0.08805025790001483, "grad_norm": 14.228179931640625, "learning_rate": 1.6385635061811954e-06, "loss": 0.3416, "step": 9790 }, { "epoch": 0.0880592517976553, "grad_norm": 8.423290252685547, "learning_rate": 1.6358883850134816e-06, "loss": 0.5167, "step": 9791 }, { "epoch": 0.08806824569529574, "grad_norm": 2.2792229652404785, "learning_rate": 1.633215375469005e-06, "loss": 0.3657, "step": 9792 }, { "epoch": 0.08807723959293619, "grad_norm": 2.8076658248901367, "learning_rate": 1.6305444777893587e-06, "loss": 0.6786, "step": 9793 }, { "epoch": 0.08808623349057665, "grad_norm": 3.032611846923828, "learning_rate": 1.6278756922159327e-06, "loss": 0.72, "step": 9794 }, { "epoch": 0.0880952273882171, "grad_norm": 6.016335964202881, "learning_rate": 1.6252090189899204e-06, "loss": 0.8046, "step": 9795 }, { "epoch": 0.08810422128585754, "grad_norm": 9.613625526428223, "learning_rate": 1.6225444583523458e-06, "loss": 0.5825, "step": 9796 }, { "epoch": 0.088113215183498, "grad_norm": 3.294189929962158, "learning_rate": 1.6198820105440189e-06, "loss": 0.0796, "step": 9797 }, { "epoch": 0.08812220908113845, "grad_norm": 8.399353981018066, "learning_rate": 1.6172216758055776e-06, "loss": 0.511, "step": 9798 }, { "epoch": 0.0881312029787789, "grad_norm": 4.364357948303223, "learning_rate": 1.6145634543774546e-06, "loss": 0.3147, "step": 9799 }, { "epoch": 0.08814019687641934, "grad_norm": 8.792668342590332, "learning_rate": 1.6119073464998968e-06, "loss": 0.4021, "step": 9800 }, { "epoch": 0.0881491907740598, "grad_norm": 2.413689374923706, "learning_rate": 1.6092533524129622e-06, "loss": 0.4461, "step": 9801 }, { "epoch": 0.08815818467170025, "grad_norm": 12.344671249389648, "learning_rate": 1.6066014723565116e-06, "loss": 0.7345, "step": 9802 }, { "epoch": 0.0881671785693407, "grad_norm": 2.124892473220825, "learning_rate": 1.6039517065702253e-06, "loss": 0.2847, "step": 9803 }, { "epoch": 0.08817617246698116, "grad_norm": 3.1611173152923584, "learning_rate": 1.6013040552935814e-06, "loss": 0.3646, "step": 9804 }, { "epoch": 0.0881851663646216, "grad_norm": 2.7340142726898193, "learning_rate": 1.5986585187658776e-06, "loss": 0.6828, "step": 9805 }, { "epoch": 0.08819416026226205, "grad_norm": 9.615241050720215, "learning_rate": 1.5960150972262083e-06, "loss": 0.8614, "step": 9806 }, { "epoch": 0.0882031541599025, "grad_norm": 6.50739049911499, "learning_rate": 1.5933737909134855e-06, "loss": 0.5613, "step": 9807 }, { "epoch": 0.08821214805754296, "grad_norm": 2.8332552909851074, "learning_rate": 1.5907346000664292e-06, "loss": 0.3902, "step": 9808 }, { "epoch": 0.08822114195518341, "grad_norm": 3.1210014820098877, "learning_rate": 1.5880975249235651e-06, "loss": 0.2657, "step": 9809 }, { "epoch": 0.08823013585282385, "grad_norm": 9.909236907958984, "learning_rate": 1.5854625657232308e-06, "loss": 0.6307, "step": 9810 }, { "epoch": 0.08823912975046432, "grad_norm": 2.2233476638793945, "learning_rate": 1.582829722703566e-06, "loss": 0.5866, "step": 9811 }, { "epoch": 0.08824812364810476, "grad_norm": 2.296057939529419, "learning_rate": 1.5801989961025305e-06, "loss": 0.449, "step": 9812 }, { "epoch": 0.08825711754574521, "grad_norm": 3.4890871047973633, "learning_rate": 1.5775703861578866e-06, "loss": 0.5431, "step": 9813 }, { "epoch": 0.08826611144338567, "grad_norm": 5.272711277008057, "learning_rate": 1.5749438931072003e-06, "loss": 0.5907, "step": 9814 }, { "epoch": 0.08827510534102612, "grad_norm": 2.38638973236084, "learning_rate": 1.5723195171878564e-06, "loss": 0.339, "step": 9815 }, { "epoch": 0.08828409923866656, "grad_norm": 2.187223434448242, "learning_rate": 1.5696972586370346e-06, "loss": 0.3435, "step": 9816 }, { "epoch": 0.08829309313630701, "grad_norm": 14.268525123596191, "learning_rate": 1.5670771176917397e-06, "loss": 0.2896, "step": 9817 }, { "epoch": 0.08830208703394747, "grad_norm": 8.670781135559082, "learning_rate": 1.5644590945887771e-06, "loss": 0.541, "step": 9818 }, { "epoch": 0.08831108093158792, "grad_norm": 12.827265739440918, "learning_rate": 1.5618431895647544e-06, "loss": 0.4007, "step": 9819 }, { "epoch": 0.08832007482922836, "grad_norm": 12.282525062561035, "learning_rate": 1.559229402856105e-06, "loss": 0.3768, "step": 9820 }, { "epoch": 0.08832906872686883, "grad_norm": 2.7534303665161133, "learning_rate": 1.5566177346990478e-06, "loss": 0.2644, "step": 9821 }, { "epoch": 0.08833806262450927, "grad_norm": 2.432169198989868, "learning_rate": 1.5540081853296307e-06, "loss": 0.194, "step": 9822 }, { "epoch": 0.08834705652214972, "grad_norm": 3.0955545902252197, "learning_rate": 1.5514007549836978e-06, "loss": 0.692, "step": 9823 }, { "epoch": 0.08835605041979017, "grad_norm": 2.865490198135376, "learning_rate": 1.5487954438969139e-06, "loss": 0.3458, "step": 9824 }, { "epoch": 0.08836504431743063, "grad_norm": 8.29062557220459, "learning_rate": 1.5461922523047345e-06, "loss": 0.4655, "step": 9825 }, { "epoch": 0.08837403821507107, "grad_norm": 1.9534810781478882, "learning_rate": 1.5435911804424357e-06, "loss": 0.3856, "step": 9826 }, { "epoch": 0.08838303211271152, "grad_norm": 14.521223068237305, "learning_rate": 1.5409922285451046e-06, "loss": 0.5145, "step": 9827 }, { "epoch": 0.08839202601035198, "grad_norm": 9.103110313415527, "learning_rate": 1.538395396847625e-06, "loss": 0.4671, "step": 9828 }, { "epoch": 0.08840101990799243, "grad_norm": 2.981839418411255, "learning_rate": 1.5358006855847012e-06, "loss": 0.5996, "step": 9829 }, { "epoch": 0.08841001380563288, "grad_norm": 9.724048614501953, "learning_rate": 1.5332080949908372e-06, "loss": 0.2589, "step": 9830 }, { "epoch": 0.08841900770327332, "grad_norm": 2.2390856742858887, "learning_rate": 1.5306176253003457e-06, "loss": 0.2745, "step": 9831 }, { "epoch": 0.08842800160091378, "grad_norm": 2.359731674194336, "learning_rate": 1.5280292767473587e-06, "loss": 0.5552, "step": 9832 }, { "epoch": 0.08843699549855423, "grad_norm": 2.429841995239258, "learning_rate": 1.5254430495658e-06, "loss": 0.4193, "step": 9833 }, { "epoch": 0.08844598939619468, "grad_norm": 3.269611120223999, "learning_rate": 1.5228589439894163e-06, "loss": 0.2608, "step": 9834 }, { "epoch": 0.08845498329383514, "grad_norm": 3.5449206829071045, "learning_rate": 1.5202769602517515e-06, "loss": 0.4039, "step": 9835 }, { "epoch": 0.08846397719147558, "grad_norm": 7.630363464355469, "learning_rate": 1.5176970985861656e-06, "loss": 0.8569, "step": 9836 }, { "epoch": 0.08847297108911603, "grad_norm": 4.195070266723633, "learning_rate": 1.515119359225825e-06, "loss": 0.3075, "step": 9837 }, { "epoch": 0.08848196498675649, "grad_norm": 3.321758508682251, "learning_rate": 1.5125437424036965e-06, "loss": 0.1975, "step": 9838 }, { "epoch": 0.08849095888439694, "grad_norm": 2.2496745586395264, "learning_rate": 1.5099702483525713e-06, "loss": 0.6772, "step": 9839 }, { "epoch": 0.08849995278203739, "grad_norm": 2.3820455074310303, "learning_rate": 1.5073988773050329e-06, "loss": 0.3947, "step": 9840 }, { "epoch": 0.08850894667967783, "grad_norm": 2.6054558753967285, "learning_rate": 1.5048296294934788e-06, "loss": 0.442, "step": 9841 }, { "epoch": 0.0885179405773183, "grad_norm": 4.209651470184326, "learning_rate": 1.5022625051501148e-06, "loss": 0.3465, "step": 9842 }, { "epoch": 0.08852693447495874, "grad_norm": 14.101722717285156, "learning_rate": 1.4996975045069556e-06, "loss": 0.3605, "step": 9843 }, { "epoch": 0.08853592837259919, "grad_norm": 3.5759451389312744, "learning_rate": 1.4971346277958236e-06, "loss": 0.4725, "step": 9844 }, { "epoch": 0.08854492227023965, "grad_norm": 2.1126461029052734, "learning_rate": 1.494573875248348e-06, "loss": 0.6444, "step": 9845 }, { "epoch": 0.0885539161678801, "grad_norm": 3.4386279582977295, "learning_rate": 1.4920152470959707e-06, "loss": 0.4237, "step": 9846 }, { "epoch": 0.08856291006552054, "grad_norm": 2.8550894260406494, "learning_rate": 1.4894587435699298e-06, "loss": 0.6163, "step": 9847 }, { "epoch": 0.08857190396316099, "grad_norm": 3.440537214279175, "learning_rate": 1.486904364901287e-06, "loss": 0.562, "step": 9848 }, { "epoch": 0.08858089786080145, "grad_norm": 2.0158095359802246, "learning_rate": 1.4843521113209052e-06, "loss": 0.0782, "step": 9849 }, { "epoch": 0.0885898917584419, "grad_norm": 3.1422581672668457, "learning_rate": 1.4818019830594444e-06, "loss": 0.4835, "step": 9850 }, { "epoch": 0.08859888565608234, "grad_norm": 4.744100093841553, "learning_rate": 1.479253980347392e-06, "loss": 0.5669, "step": 9851 }, { "epoch": 0.0886078795537228, "grad_norm": 2.2168402671813965, "learning_rate": 1.4767081034150282e-06, "loss": 0.4707, "step": 9852 }, { "epoch": 0.08861687345136325, "grad_norm": 13.55176830291748, "learning_rate": 1.4741643524924492e-06, "loss": 0.4341, "step": 9853 }, { "epoch": 0.0886258673490037, "grad_norm": 13.602510452270508, "learning_rate": 1.4716227278095602e-06, "loss": 0.382, "step": 9854 }, { "epoch": 0.08863486124664416, "grad_norm": 2.8089756965637207, "learning_rate": 1.4690832295960633e-06, "loss": 0.2567, "step": 9855 }, { "epoch": 0.0886438551442846, "grad_norm": 2.3977129459381104, "learning_rate": 1.4665458580814805e-06, "loss": 0.4531, "step": 9856 }, { "epoch": 0.08865284904192505, "grad_norm": 5.187553882598877, "learning_rate": 1.4640106134951316e-06, "loss": 0.3389, "step": 9857 }, { "epoch": 0.0886618429395655, "grad_norm": 10.735997200012207, "learning_rate": 1.4614774960661548e-06, "loss": 0.2642, "step": 9858 }, { "epoch": 0.08867083683720596, "grad_norm": 3.5421736240386963, "learning_rate": 1.4589465060234898e-06, "loss": 0.3527, "step": 9859 }, { "epoch": 0.08867983073484641, "grad_norm": 24.593372344970703, "learning_rate": 1.4564176435958782e-06, "loss": 0.4466, "step": 9860 }, { "epoch": 0.08868882463248685, "grad_norm": 8.70399284362793, "learning_rate": 1.4538909090118846e-06, "loss": 0.5182, "step": 9861 }, { "epoch": 0.08869781853012731, "grad_norm": 2.694197177886963, "learning_rate": 1.4513663024998652e-06, "loss": 0.5858, "step": 9862 }, { "epoch": 0.08870681242776776, "grad_norm": 2.434936046600342, "learning_rate": 1.4488438242879988e-06, "loss": 0.4805, "step": 9863 }, { "epoch": 0.08871580632540821, "grad_norm": 2.9406652450561523, "learning_rate": 1.4463234746042581e-06, "loss": 0.3497, "step": 9864 }, { "epoch": 0.08872480022304866, "grad_norm": 2.575796365737915, "learning_rate": 1.4438052536764335e-06, "loss": 0.6972, "step": 9865 }, { "epoch": 0.08873379412068912, "grad_norm": 4.343408107757568, "learning_rate": 1.4412891617321177e-06, "loss": 0.596, "step": 9866 }, { "epoch": 0.08874278801832956, "grad_norm": 2.73545241355896, "learning_rate": 1.4387751989987098e-06, "loss": 0.3567, "step": 9867 }, { "epoch": 0.08875178191597001, "grad_norm": 15.733132362365723, "learning_rate": 1.4362633657034248e-06, "loss": 0.5944, "step": 9868 }, { "epoch": 0.08876077581361047, "grad_norm": 8.434510231018066, "learning_rate": 1.4337536620732755e-06, "loss": 0.7374, "step": 9869 }, { "epoch": 0.08876976971125092, "grad_norm": 5.012548923492432, "learning_rate": 1.4312460883350865e-06, "loss": 0.4922, "step": 9870 }, { "epoch": 0.08877876360889136, "grad_norm": 3.075345993041992, "learning_rate": 1.42874064471549e-06, "loss": 0.29, "step": 9871 }, { "epoch": 0.08878775750653181, "grad_norm": 2.887016534805298, "learning_rate": 1.4262373314409217e-06, "loss": 0.344, "step": 9872 }, { "epoch": 0.08879675140417227, "grad_norm": 14.622725486755371, "learning_rate": 1.4237361487376366e-06, "loss": 0.7418, "step": 9873 }, { "epoch": 0.08880574530181272, "grad_norm": 2.7165002822875977, "learning_rate": 1.4212370968316791e-06, "loss": 0.2107, "step": 9874 }, { "epoch": 0.08881473919945317, "grad_norm": 2.6514344215393066, "learning_rate": 1.4187401759489183e-06, "loss": 0.5487, "step": 9875 }, { "epoch": 0.08882373309709363, "grad_norm": 2.1376216411590576, "learning_rate": 1.4162453863150183e-06, "loss": 0.385, "step": 9876 }, { "epoch": 0.08883272699473407, "grad_norm": 2.9520671367645264, "learning_rate": 1.4137527281554625e-06, "loss": 0.3838, "step": 9877 }, { "epoch": 0.08884172089237452, "grad_norm": 2.3759846687316895, "learning_rate": 1.411262201695529e-06, "loss": 0.6262, "step": 9878 }, { "epoch": 0.08885071479001498, "grad_norm": 1.8475316762924194, "learning_rate": 1.4087738071603075e-06, "loss": 0.5096, "step": 9879 }, { "epoch": 0.08885970868765543, "grad_norm": 2.1953511238098145, "learning_rate": 1.406287544774701e-06, "loss": 0.5713, "step": 9880 }, { "epoch": 0.08886870258529588, "grad_norm": 9.518521308898926, "learning_rate": 1.4038034147634106e-06, "loss": 0.4055, "step": 9881 }, { "epoch": 0.08887769648293632, "grad_norm": 17.102031707763672, "learning_rate": 1.4013214173509536e-06, "loss": 0.1642, "step": 9882 }, { "epoch": 0.08888669038057678, "grad_norm": 7.864806175231934, "learning_rate": 1.398841552761651e-06, "loss": 0.3876, "step": 9883 }, { "epoch": 0.08889568427821723, "grad_norm": 21.254619598388672, "learning_rate": 1.3963638212196261e-06, "loss": 0.5087, "step": 9884 }, { "epoch": 0.08890467817585768, "grad_norm": 2.4115965366363525, "learning_rate": 1.3938882229488164e-06, "loss": 0.4805, "step": 9885 }, { "epoch": 0.08891367207349814, "grad_norm": 2.870159864425659, "learning_rate": 1.3914147581729598e-06, "loss": 0.5319, "step": 9886 }, { "epoch": 0.08892266597113858, "grad_norm": 2.7346487045288086, "learning_rate": 1.3889434271156105e-06, "loss": 0.4804, "step": 9887 }, { "epoch": 0.08893165986877903, "grad_norm": 3.1735358238220215, "learning_rate": 1.3864742300001176e-06, "loss": 0.4167, "step": 9888 }, { "epoch": 0.08894065376641948, "grad_norm": 22.24134635925293, "learning_rate": 1.3840071670496557e-06, "loss": 0.8109, "step": 9889 }, { "epoch": 0.08894964766405994, "grad_norm": 3.104448080062866, "learning_rate": 1.3815422384871879e-06, "loss": 0.08, "step": 9890 }, { "epoch": 0.08895864156170039, "grad_norm": 1.8880174160003662, "learning_rate": 1.3790794445354888e-06, "loss": 0.4496, "step": 9891 }, { "epoch": 0.08896763545934083, "grad_norm": 2.2775676250457764, "learning_rate": 1.3766187854171524e-06, "loss": 0.3205, "step": 9892 }, { "epoch": 0.0889766293569813, "grad_norm": 2.518336057662964, "learning_rate": 1.3741602613545596e-06, "loss": 0.5005, "step": 9893 }, { "epoch": 0.08898562325462174, "grad_norm": 3.516066789627075, "learning_rate": 1.3717038725699183e-06, "loss": 0.7422, "step": 9894 }, { "epoch": 0.08899461715226219, "grad_norm": 2.1812708377838135, "learning_rate": 1.3692496192852316e-06, "loss": 0.5695, "step": 9895 }, { "epoch": 0.08900361104990265, "grad_norm": 2.285642623901367, "learning_rate": 1.366797501722311e-06, "loss": 0.2604, "step": 9896 }, { "epoch": 0.0890126049475431, "grad_norm": 28.236122131347656, "learning_rate": 1.364347520102774e-06, "loss": 0.3233, "step": 9897 }, { "epoch": 0.08902159884518354, "grad_norm": 2.9958646297454834, "learning_rate": 1.3618996746480538e-06, "loss": 0.5584, "step": 9898 }, { "epoch": 0.08903059274282399, "grad_norm": 2.2649452686309814, "learning_rate": 1.3594539655793797e-06, "loss": 0.4323, "step": 9899 }, { "epoch": 0.08903958664046445, "grad_norm": 5.526645660400391, "learning_rate": 1.3570103931177886e-06, "loss": 0.7206, "step": 9900 }, { "epoch": 0.0890485805381049, "grad_norm": 1.7384631633758545, "learning_rate": 1.3545689574841342e-06, "loss": 0.5908, "step": 9901 }, { "epoch": 0.08905757443574534, "grad_norm": 13.133833885192871, "learning_rate": 1.3521296588990706e-06, "loss": 0.5085, "step": 9902 }, { "epoch": 0.0890665683333858, "grad_norm": 3.2282843589782715, "learning_rate": 1.3496924975830521e-06, "loss": 0.0807, "step": 9903 }, { "epoch": 0.08907556223102625, "grad_norm": 3.0995407104492188, "learning_rate": 1.3472574737563553e-06, "loss": 0.5522, "step": 9904 }, { "epoch": 0.0890845561286667, "grad_norm": 5.472433090209961, "learning_rate": 1.3448245876390486e-06, "loss": 1.0334, "step": 9905 }, { "epoch": 0.08909355002630714, "grad_norm": 11.140021324157715, "learning_rate": 1.3423938394510199e-06, "loss": 0.3848, "step": 9906 }, { "epoch": 0.0891025439239476, "grad_norm": 2.4692625999450684, "learning_rate": 1.3399652294119547e-06, "loss": 0.0942, "step": 9907 }, { "epoch": 0.08911153782158805, "grad_norm": 3.031285047531128, "learning_rate": 1.337538757741344e-06, "loss": 0.4213, "step": 9908 }, { "epoch": 0.0891205317192285, "grad_norm": 14.746382713317871, "learning_rate": 1.3351144246584956e-06, "loss": 0.6856, "step": 9909 }, { "epoch": 0.08912952561686896, "grad_norm": 6.528025150299072, "learning_rate": 1.332692230382518e-06, "loss": 0.1257, "step": 9910 }, { "epoch": 0.0891385195145094, "grad_norm": 2.762235641479492, "learning_rate": 1.3302721751323215e-06, "loss": 0.342, "step": 9911 }, { "epoch": 0.08914751341214985, "grad_norm": 2.762279748916626, "learning_rate": 1.3278542591266291e-06, "loss": 0.3239, "step": 9912 }, { "epoch": 0.0891565073097903, "grad_norm": 3.2236080169677734, "learning_rate": 1.3254384825839738e-06, "loss": 0.3311, "step": 9913 }, { "epoch": 0.08916550120743076, "grad_norm": 4.0655670166015625, "learning_rate": 1.3230248457226897e-06, "loss": 0.126, "step": 9914 }, { "epoch": 0.08917449510507121, "grad_norm": 12.530380249023438, "learning_rate": 1.3206133487609128e-06, "loss": 0.6507, "step": 9915 }, { "epoch": 0.08918348900271166, "grad_norm": 2.6560561656951904, "learning_rate": 1.3182039919166e-06, "loss": 0.0609, "step": 9916 }, { "epoch": 0.08919248290035212, "grad_norm": 3.429694652557373, "learning_rate": 1.3157967754074963e-06, "loss": 0.2929, "step": 9917 }, { "epoch": 0.08920147679799256, "grad_norm": 1.96184241771698, "learning_rate": 1.3133916994511775e-06, "loss": 0.451, "step": 9918 }, { "epoch": 0.08921047069563301, "grad_norm": 2.395451784133911, "learning_rate": 1.3109887642650003e-06, "loss": 0.4472, "step": 9919 }, { "epoch": 0.08921946459327347, "grad_norm": 2.027381420135498, "learning_rate": 1.3085879700661412e-06, "loss": 0.4472, "step": 9920 }, { "epoch": 0.08922845849091392, "grad_norm": 2.7809438705444336, "learning_rate": 1.306189317071585e-06, "loss": 0.3625, "step": 9921 }, { "epoch": 0.08923745238855436, "grad_norm": 2.4035494327545166, "learning_rate": 1.3037928054981163e-06, "loss": 0.5179, "step": 9922 }, { "epoch": 0.08924644628619481, "grad_norm": 2.4804413318634033, "learning_rate": 1.3013984355623315e-06, "loss": 0.7394, "step": 9923 }, { "epoch": 0.08925544018383527, "grad_norm": 2.888136386871338, "learning_rate": 1.2990062074806325e-06, "loss": 0.1801, "step": 9924 }, { "epoch": 0.08926443408147572, "grad_norm": 2.568143844604492, "learning_rate": 1.2966161214692212e-06, "loss": 0.2054, "step": 9925 }, { "epoch": 0.08927342797911617, "grad_norm": 2.2158071994781494, "learning_rate": 1.2942281777441168e-06, "loss": 0.1784, "step": 9926 }, { "epoch": 0.08928242187675663, "grad_norm": 5.913144588470459, "learning_rate": 1.2918423765211297e-06, "loss": 0.6527, "step": 9927 }, { "epoch": 0.08929141577439707, "grad_norm": 2.590395927429199, "learning_rate": 1.2894587180158957e-06, "loss": 0.3939, "step": 9928 }, { "epoch": 0.08930040967203752, "grad_norm": 2.2543251514434814, "learning_rate": 1.2870772024438427e-06, "loss": 0.303, "step": 9929 }, { "epoch": 0.08930940356967797, "grad_norm": 2.2613799571990967, "learning_rate": 1.2846978300202122e-06, "loss": 0.6233, "step": 9930 }, { "epoch": 0.08931839746731843, "grad_norm": 13.715725898742676, "learning_rate": 1.2823206009600491e-06, "loss": 0.6988, "step": 9931 }, { "epoch": 0.08932739136495887, "grad_norm": 20.558555603027344, "learning_rate": 1.279945515478198e-06, "loss": 0.3609, "step": 9932 }, { "epoch": 0.08933638526259932, "grad_norm": 9.16641902923584, "learning_rate": 1.277572573789329e-06, "loss": 0.6166, "step": 9933 }, { "epoch": 0.08934537916023978, "grad_norm": 12.682315826416016, "learning_rate": 1.2752017761078927e-06, "loss": 0.586, "step": 9934 }, { "epoch": 0.08935437305788023, "grad_norm": 2.8349103927612305, "learning_rate": 1.2728331226481704e-06, "loss": 0.6547, "step": 9935 }, { "epoch": 0.08936336695552068, "grad_norm": 13.558541297912598, "learning_rate": 1.2704666136242326e-06, "loss": 0.8202, "step": 9936 }, { "epoch": 0.08937236085316114, "grad_norm": 3.611884593963623, "learning_rate": 1.268102249249961e-06, "loss": 0.6892, "step": 9937 }, { "epoch": 0.08938135475080158, "grad_norm": 13.104924201965332, "learning_rate": 1.2657400297390488e-06, "loss": 0.5267, "step": 9938 }, { "epoch": 0.08939034864844203, "grad_norm": 2.233048915863037, "learning_rate": 1.2633799553049885e-06, "loss": 0.7472, "step": 9939 }, { "epoch": 0.08939934254608248, "grad_norm": 17.50445556640625, "learning_rate": 1.2610220261610793e-06, "loss": 0.5982, "step": 9940 }, { "epoch": 0.08940833644372294, "grad_norm": 2.6212215423583984, "learning_rate": 1.2586662425204287e-06, "loss": 0.4966, "step": 9941 }, { "epoch": 0.08941733034136339, "grad_norm": 2.159651041030884, "learning_rate": 1.2563126045959522e-06, "loss": 0.2535, "step": 9942 }, { "epoch": 0.08942632423900383, "grad_norm": 3.9016270637512207, "learning_rate": 1.2539611126003714e-06, "loss": 0.5088, "step": 9943 }, { "epoch": 0.08943531813664429, "grad_norm": 2.858403205871582, "learning_rate": 1.2516117667462025e-06, "loss": 0.2645, "step": 9944 }, { "epoch": 0.08944431203428474, "grad_norm": 24.76820945739746, "learning_rate": 1.2492645672457837e-06, "loss": 0.4652, "step": 9945 }, { "epoch": 0.08945330593192519, "grad_norm": 2.5000014305114746, "learning_rate": 1.246919514311251e-06, "loss": 0.1039, "step": 9946 }, { "epoch": 0.08946229982956563, "grad_norm": 2.5689568519592285, "learning_rate": 1.244576608154549e-06, "loss": 0.5062, "step": 9947 }, { "epoch": 0.0894712937272061, "grad_norm": 3.7535104751586914, "learning_rate": 1.2422358489874276e-06, "loss": 0.4043, "step": 9948 }, { "epoch": 0.08948028762484654, "grad_norm": 8.590354919433594, "learning_rate": 1.2398972370214367e-06, "loss": 0.4011, "step": 9949 }, { "epoch": 0.08948928152248699, "grad_norm": 2.0101735591888428, "learning_rate": 1.2375607724679438e-06, "loss": 0.5065, "step": 9950 }, { "epoch": 0.08949827542012745, "grad_norm": 17.259014129638672, "learning_rate": 1.2352264555381132e-06, "loss": 0.3978, "step": 9951 }, { "epoch": 0.0895072693177679, "grad_norm": 1.9631774425506592, "learning_rate": 1.2328942864429177e-06, "loss": 0.3739, "step": 9952 }, { "epoch": 0.08951626321540834, "grad_norm": 1.8749871253967285, "learning_rate": 1.2305642653931387e-06, "loss": 0.3924, "step": 9953 }, { "epoch": 0.08952525711304879, "grad_norm": 2.6960983276367188, "learning_rate": 1.2282363925993578e-06, "loss": 0.4497, "step": 9954 }, { "epoch": 0.08953425101068925, "grad_norm": 2.8083319664001465, "learning_rate": 1.2259106682719678e-06, "loss": 0.4434, "step": 9955 }, { "epoch": 0.0895432449083297, "grad_norm": 1.787717342376709, "learning_rate": 1.2235870926211619e-06, "loss": 0.327, "step": 9956 }, { "epoch": 0.08955223880597014, "grad_norm": 2.3980565071105957, "learning_rate": 1.2212656658569465e-06, "loss": 0.0536, "step": 9957 }, { "epoch": 0.0895612327036106, "grad_norm": 17.326602935791016, "learning_rate": 1.2189463881891266e-06, "loss": 0.6809, "step": 9958 }, { "epoch": 0.08957022660125105, "grad_norm": 3.042703628540039, "learning_rate": 1.2166292598273177e-06, "loss": 0.0738, "step": 9959 }, { "epoch": 0.0895792204988915, "grad_norm": 3.505164861679077, "learning_rate": 1.2143142809809411e-06, "loss": 0.4687, "step": 9960 }, { "epoch": 0.08958821439653196, "grad_norm": 2.2176003456115723, "learning_rate": 1.2120014518592155e-06, "loss": 0.4562, "step": 9961 }, { "epoch": 0.0895972082941724, "grad_norm": 15.530914306640625, "learning_rate": 1.2096907726711792e-06, "loss": 0.4126, "step": 9962 }, { "epoch": 0.08960620219181285, "grad_norm": 4.206121444702148, "learning_rate": 1.2073822436256655e-06, "loss": 0.3628, "step": 9963 }, { "epoch": 0.0896151960894533, "grad_norm": 3.26822566986084, "learning_rate": 1.2050758649313183e-06, "loss": 0.6643, "step": 9964 }, { "epoch": 0.08962418998709376, "grad_norm": 10.523087501525879, "learning_rate": 1.2027716367965847e-06, "loss": 0.6181, "step": 9965 }, { "epoch": 0.08963318388473421, "grad_norm": 2.630450487136841, "learning_rate": 1.2004695594297178e-06, "loss": 0.3889, "step": 9966 }, { "epoch": 0.08964217778237465, "grad_norm": 2.3422629833221436, "learning_rate": 1.1981696330387787e-06, "loss": 0.6167, "step": 9967 }, { "epoch": 0.08965117168001512, "grad_norm": 1.9235345125198364, "learning_rate": 1.1958718578316264e-06, "loss": 0.4163, "step": 9968 }, { "epoch": 0.08966016557765556, "grad_norm": 3.097200632095337, "learning_rate": 1.1935762340159417e-06, "loss": 0.372, "step": 9969 }, { "epoch": 0.08966915947529601, "grad_norm": 9.87475299835205, "learning_rate": 1.1912827617991895e-06, "loss": 0.5092, "step": 9970 }, { "epoch": 0.08967815337293646, "grad_norm": 2.7784252166748047, "learning_rate": 1.1889914413886593e-06, "loss": 0.299, "step": 9971 }, { "epoch": 0.08968714727057692, "grad_norm": 11.371780395507812, "learning_rate": 1.1867022729914357e-06, "loss": 0.56, "step": 9972 }, { "epoch": 0.08969614116821736, "grad_norm": 10.844256401062012, "learning_rate": 1.1844152568144085e-06, "loss": 0.6016, "step": 9973 }, { "epoch": 0.08970513506585781, "grad_norm": 2.055630922317505, "learning_rate": 1.1821303930642819e-06, "loss": 0.3954, "step": 9974 }, { "epoch": 0.08971412896349827, "grad_norm": 2.8900275230407715, "learning_rate": 1.1798476819475546e-06, "loss": 0.3445, "step": 9975 }, { "epoch": 0.08972312286113872, "grad_norm": 3.2931206226348877, "learning_rate": 1.1775671236705365e-06, "loss": 0.3597, "step": 9976 }, { "epoch": 0.08973211675877917, "grad_norm": 16.251659393310547, "learning_rate": 1.1752887184393407e-06, "loss": 0.4549, "step": 9977 }, { "epoch": 0.08974111065641963, "grad_norm": 2.180044174194336, "learning_rate": 1.1730124664598912e-06, "loss": 0.1775, "step": 9978 }, { "epoch": 0.08975010455406007, "grad_norm": 3.262704610824585, "learning_rate": 1.1707383679379124e-06, "loss": 0.4606, "step": 9979 }, { "epoch": 0.08975909845170052, "grad_norm": 3.2245888710021973, "learning_rate": 1.1684664230789316e-06, "loss": 0.2006, "step": 9980 }, { "epoch": 0.08976809234934097, "grad_norm": 6.32314920425415, "learning_rate": 1.1661966320882873e-06, "loss": 0.5841, "step": 9981 }, { "epoch": 0.08977708624698143, "grad_norm": 13.99137020111084, "learning_rate": 1.1639289951711153e-06, "loss": 0.5879, "step": 9982 }, { "epoch": 0.08978608014462187, "grad_norm": 7.673317909240723, "learning_rate": 1.161663512532371e-06, "loss": 0.763, "step": 9983 }, { "epoch": 0.08979507404226232, "grad_norm": 2.0001845359802246, "learning_rate": 1.1594001843767993e-06, "loss": 0.5497, "step": 9984 }, { "epoch": 0.08980406793990278, "grad_norm": 2.9888551235198975, "learning_rate": 1.1571390109089586e-06, "loss": 0.8306, "step": 9985 }, { "epoch": 0.08981306183754323, "grad_norm": 17.46289825439453, "learning_rate": 1.154879992333216e-06, "loss": 0.5398, "step": 9986 }, { "epoch": 0.08982205573518368, "grad_norm": 2.835038661956787, "learning_rate": 1.1526231288537332e-06, "loss": 0.7551, "step": 9987 }, { "epoch": 0.08983104963282412, "grad_norm": 18.975370407104492, "learning_rate": 1.1503684206744886e-06, "loss": 0.278, "step": 9988 }, { "epoch": 0.08984004353046458, "grad_norm": 2.733116626739502, "learning_rate": 1.1481158679992555e-06, "loss": 0.5364, "step": 9989 }, { "epoch": 0.08984903742810503, "grad_norm": 3.7068440914154053, "learning_rate": 1.1458654710316207e-06, "loss": 0.502, "step": 9990 }, { "epoch": 0.08985803132574548, "grad_norm": 2.392681360244751, "learning_rate": 1.1436172299749693e-06, "loss": 0.7055, "step": 9991 }, { "epoch": 0.08986702522338594, "grad_norm": 2.6282870769500732, "learning_rate": 1.1413711450324965e-06, "loss": 0.4187, "step": 9992 }, { "epoch": 0.08987601912102638, "grad_norm": 2.2963318824768066, "learning_rate": 1.1391272164072042e-06, "loss": 0.372, "step": 9993 }, { "epoch": 0.08988501301866683, "grad_norm": 3.620335102081299, "learning_rate": 1.1368854443018912e-06, "loss": 0.3405, "step": 9994 }, { "epoch": 0.08989400691630728, "grad_norm": 2.522714853286743, "learning_rate": 1.1346458289191702e-06, "loss": 0.1134, "step": 9995 }, { "epoch": 0.08990300081394774, "grad_norm": 2.557077407836914, "learning_rate": 1.132408370461452e-06, "loss": 0.6403, "step": 9996 }, { "epoch": 0.08991199471158819, "grad_norm": 10.932153701782227, "learning_rate": 1.1301730691309547e-06, "loss": 0.7046, "step": 9997 }, { "epoch": 0.08992098860922863, "grad_norm": 3.081770658493042, "learning_rate": 1.127939925129709e-06, "loss": 0.1634, "step": 9998 }, { "epoch": 0.0899299825068691, "grad_norm": 2.068683385848999, "learning_rate": 1.1257089386595366e-06, "loss": 0.4944, "step": 9999 }, { "epoch": 0.08993897640450954, "grad_norm": 1.9351937770843506, "learning_rate": 1.1234801099220787e-06, "loss": 0.4072, "step": 10000 }, { "epoch": 0.08994797030214999, "grad_norm": 2.8275182247161865, "learning_rate": 1.121253439118769e-06, "loss": 0.1903, "step": 10001 }, { "epoch": 0.08995696419979045, "grad_norm": 3.7449872493743896, "learning_rate": 1.1190289264508518e-06, "loss": 0.2216, "step": 10002 }, { "epoch": 0.0899659580974309, "grad_norm": 15.455436706542969, "learning_rate": 1.1168065721193804e-06, "loss": 0.7512, "step": 10003 }, { "epoch": 0.08997495199507134, "grad_norm": 5.7858052253723145, "learning_rate": 1.1145863763252024e-06, "loss": 0.7429, "step": 10004 }, { "epoch": 0.08998394589271179, "grad_norm": 2.5350396633148193, "learning_rate": 1.1123683392689848e-06, "loss": 0.2738, "step": 10005 }, { "epoch": 0.08999293979035225, "grad_norm": 2.014319658279419, "learning_rate": 1.1101524611511816e-06, "loss": 0.2175, "step": 10006 }, { "epoch": 0.0900019336879927, "grad_norm": 14.46469783782959, "learning_rate": 1.1079387421720738e-06, "loss": 0.5187, "step": 10007 }, { "epoch": 0.09001092758563314, "grad_norm": 2.030540704727173, "learning_rate": 1.1057271825317266e-06, "loss": 0.7105, "step": 10008 }, { "epoch": 0.0900199214832736, "grad_norm": 7.944052696228027, "learning_rate": 1.1035177824300192e-06, "loss": 0.4722, "step": 10009 }, { "epoch": 0.09002891538091405, "grad_norm": 2.0630078315734863, "learning_rate": 1.101310542066636e-06, "loss": 0.1268, "step": 10010 }, { "epoch": 0.0900379092785545, "grad_norm": 15.319473266601562, "learning_rate": 1.0991054616410589e-06, "loss": 0.5889, "step": 10011 }, { "epoch": 0.09004690317619495, "grad_norm": 3.2701594829559326, "learning_rate": 1.096902541352593e-06, "loss": 0.6837, "step": 10012 }, { "epoch": 0.0900558970738354, "grad_norm": 16.8278865814209, "learning_rate": 1.0947017814003259e-06, "loss": 0.8669, "step": 10013 }, { "epoch": 0.09006489097147585, "grad_norm": 2.524794340133667, "learning_rate": 1.092503181983162e-06, "loss": 0.3075, "step": 10014 }, { "epoch": 0.0900738848691163, "grad_norm": 2.0118038654327393, "learning_rate": 1.0903067432998126e-06, "loss": 0.3483, "step": 10015 }, { "epoch": 0.09008287876675676, "grad_norm": 3.3263094425201416, "learning_rate": 1.0881124655487823e-06, "loss": 0.4207, "step": 10016 }, { "epoch": 0.09009187266439721, "grad_norm": 3.5140953063964844, "learning_rate": 1.085920348928396e-06, "loss": 0.7893, "step": 10017 }, { "epoch": 0.09010086656203765, "grad_norm": 3.7080302238464355, "learning_rate": 1.083730393636767e-06, "loss": 0.7105, "step": 10018 }, { "epoch": 0.09010986045967811, "grad_norm": 2.9532673358917236, "learning_rate": 1.0815425998718264e-06, "loss": 0.6977, "step": 10019 }, { "epoch": 0.09011885435731856, "grad_norm": 2.2178235054016113, "learning_rate": 1.0793569678313021e-06, "loss": 0.3429, "step": 10020 }, { "epoch": 0.09012784825495901, "grad_norm": 14.660715103149414, "learning_rate": 1.0771734977127328e-06, "loss": 0.9996, "step": 10021 }, { "epoch": 0.09013684215259946, "grad_norm": 2.776416301727295, "learning_rate": 1.074992189713453e-06, "loss": 0.6501, "step": 10022 }, { "epoch": 0.09014583605023992, "grad_norm": 2.9429616928100586, "learning_rate": 1.0728130440306071e-06, "loss": 0.6056, "step": 10023 }, { "epoch": 0.09015482994788036, "grad_norm": 2.4553322792053223, "learning_rate": 1.0706360608611493e-06, "loss": 0.3183, "step": 10024 }, { "epoch": 0.09016382384552081, "grad_norm": 2.9150187969207764, "learning_rate": 1.0684612404018301e-06, "loss": 0.0646, "step": 10025 }, { "epoch": 0.09017281774316127, "grad_norm": 2.1133222579956055, "learning_rate": 1.0662885828492036e-06, "loss": 0.4858, "step": 10026 }, { "epoch": 0.09018181164080172, "grad_norm": 2.771676540374756, "learning_rate": 1.0641180883996376e-06, "loss": 0.6064, "step": 10027 }, { "epoch": 0.09019080553844216, "grad_norm": 14.55605697631836, "learning_rate": 1.0619497572492948e-06, "loss": 0.6573, "step": 10028 }, { "epoch": 0.09019979943608261, "grad_norm": 2.2236814498901367, "learning_rate": 1.0597835895941516e-06, "loss": 0.1448, "step": 10029 }, { "epoch": 0.09020879333372307, "grad_norm": 3.126115322113037, "learning_rate": 1.0576195856299791e-06, "loss": 0.5943, "step": 10030 }, { "epoch": 0.09021778723136352, "grad_norm": 2.113910436630249, "learning_rate": 1.055457745552363e-06, "loss": 0.3269, "step": 10031 }, { "epoch": 0.09022678112900397, "grad_norm": 13.176977157592773, "learning_rate": 1.0532980695566851e-06, "loss": 0.5579, "step": 10032 }, { "epoch": 0.09023577502664443, "grad_norm": 2.0777640342712402, "learning_rate": 1.051140557838129e-06, "loss": 0.668, "step": 10033 }, { "epoch": 0.09024476892428487, "grad_norm": 1.9538826942443848, "learning_rate": 1.0489852105916993e-06, "loss": 0.5666, "step": 10034 }, { "epoch": 0.09025376282192532, "grad_norm": 2.427938222885132, "learning_rate": 1.0468320280121873e-06, "loss": 0.6693, "step": 10035 }, { "epoch": 0.09026275671956578, "grad_norm": 10.381492614746094, "learning_rate": 1.0446810102941962e-06, "loss": 0.5434, "step": 10036 }, { "epoch": 0.09027175061720623, "grad_norm": 16.333683013916016, "learning_rate": 1.042532157632134e-06, "loss": 0.5621, "step": 10037 }, { "epoch": 0.09028074451484668, "grad_norm": 12.921382904052734, "learning_rate": 1.0403854702202066e-06, "loss": 0.7078, "step": 10038 }, { "epoch": 0.09028973841248712, "grad_norm": 2.361595392227173, "learning_rate": 1.0382409482524363e-06, "loss": 0.8041, "step": 10039 }, { "epoch": 0.09029873231012758, "grad_norm": 9.123581886291504, "learning_rate": 1.0360985919226378e-06, "loss": 0.5101, "step": 10040 }, { "epoch": 0.09030772620776803, "grad_norm": 15.505912780761719, "learning_rate": 1.0339584014244397e-06, "loss": 0.4108, "step": 10041 }, { "epoch": 0.09031672010540848, "grad_norm": 2.574169397354126, "learning_rate": 1.0318203769512619e-06, "loss": 0.383, "step": 10042 }, { "epoch": 0.09032571400304894, "grad_norm": 16.203779220581055, "learning_rate": 1.0296845186963473e-06, "loss": 0.389, "step": 10043 }, { "epoch": 0.09033470790068938, "grad_norm": 2.3022377490997314, "learning_rate": 1.0275508268527246e-06, "loss": 0.3254, "step": 10044 }, { "epoch": 0.09034370179832983, "grad_norm": 8.941046714782715, "learning_rate": 1.0254193016132368e-06, "loss": 0.7144, "step": 10045 }, { "epoch": 0.09035269569597028, "grad_norm": 2.5488338470458984, "learning_rate": 1.02328994317053e-06, "loss": 0.4313, "step": 10046 }, { "epoch": 0.09036168959361074, "grad_norm": 14.841557502746582, "learning_rate": 1.021162751717053e-06, "loss": 0.4161, "step": 10047 }, { "epoch": 0.09037068349125119, "grad_norm": 2.600842237472534, "learning_rate": 1.0190377274450574e-06, "loss": 0.1956, "step": 10048 }, { "epoch": 0.09037967738889163, "grad_norm": 2.318016767501831, "learning_rate": 1.0169148705466036e-06, "loss": 0.5974, "step": 10049 }, { "epoch": 0.0903886712865321, "grad_norm": 3.2138423919677734, "learning_rate": 1.014794181213552e-06, "loss": 0.5623, "step": 10050 }, { "epoch": 0.09039766518417254, "grad_norm": 2.3871195316314697, "learning_rate": 1.0126756596375686e-06, "loss": 0.5124, "step": 10051 }, { "epoch": 0.09040665908181299, "grad_norm": 10.581353187561035, "learning_rate": 1.010559306010117e-06, "loss": 0.6938, "step": 10052 }, { "epoch": 0.09041565297945343, "grad_norm": 2.52357816696167, "learning_rate": 1.0084451205224803e-06, "loss": 0.451, "step": 10053 }, { "epoch": 0.0904246468770939, "grad_norm": 2.2721621990203857, "learning_rate": 1.006333103365728e-06, "loss": 0.5493, "step": 10054 }, { "epoch": 0.09043364077473434, "grad_norm": 15.354623794555664, "learning_rate": 1.004223254730749e-06, "loss": 0.7349, "step": 10055 }, { "epoch": 0.09044263467237479, "grad_norm": 4.1990180015563965, "learning_rate": 1.0021155748082266e-06, "loss": 0.2661, "step": 10056 }, { "epoch": 0.09045162857001525, "grad_norm": 2.0320701599121094, "learning_rate": 1.0000100637886477e-06, "loss": 0.3864, "step": 10057 }, { "epoch": 0.0904606224676557, "grad_norm": 10.55405330657959, "learning_rate": 9.9790672186231e-07, "loss": 0.8056, "step": 10058 }, { "epoch": 0.09046961636529614, "grad_norm": 2.5024466514587402, "learning_rate": 9.958055492193087e-07, "loss": 0.7139, "step": 10059 }, { "epoch": 0.0904786102629366, "grad_norm": 3.060526132583618, "learning_rate": 9.937065460495471e-07, "loss": 0.6528, "step": 10060 }, { "epoch": 0.09048760416057705, "grad_norm": 2.430326223373413, "learning_rate": 9.91609712542732e-07, "loss": 0.1757, "step": 10061 }, { "epoch": 0.0904965980582175, "grad_norm": 1.999621033668518, "learning_rate": 9.895150488883702e-07, "loss": 0.3049, "step": 10062 }, { "epoch": 0.09050559195585794, "grad_norm": 2.787027359008789, "learning_rate": 9.87422555275777e-07, "loss": 0.6898, "step": 10063 }, { "epoch": 0.0905145858534984, "grad_norm": 2.1148922443389893, "learning_rate": 9.853322318940706e-07, "loss": 0.4361, "step": 10064 }, { "epoch": 0.09052357975113885, "grad_norm": 8.419410705566406, "learning_rate": 9.83244078932169e-07, "loss": 0.4603, "step": 10065 }, { "epoch": 0.0905325736487793, "grad_norm": 3.594088315963745, "learning_rate": 9.811580965787965e-07, "loss": 0.4541, "step": 10066 }, { "epoch": 0.09054156754641976, "grad_norm": 2.9248764514923096, "learning_rate": 9.790742850224882e-07, "loss": 0.4178, "step": 10067 }, { "epoch": 0.0905505614440602, "grad_norm": 2.06826114654541, "learning_rate": 9.769926444515714e-07, "loss": 0.2705, "step": 10068 }, { "epoch": 0.09055955534170065, "grad_norm": 14.742314338684082, "learning_rate": 9.749131750541817e-07, "loss": 0.6594, "step": 10069 }, { "epoch": 0.0905685492393411, "grad_norm": 2.4120185375213623, "learning_rate": 9.728358770182633e-07, "loss": 0.6466, "step": 10070 }, { "epoch": 0.09057754313698156, "grad_norm": 2.2193009853363037, "learning_rate": 9.707607505315547e-07, "loss": 0.4923, "step": 10071 }, { "epoch": 0.09058653703462201, "grad_norm": 2.854531764984131, "learning_rate": 9.686877957816143e-07, "loss": 0.5017, "step": 10072 }, { "epoch": 0.09059553093226246, "grad_norm": 3.1635231971740723, "learning_rate": 9.666170129557838e-07, "loss": 0.4729, "step": 10073 }, { "epoch": 0.09060452482990292, "grad_norm": 2.658665657043457, "learning_rate": 9.645484022412166e-07, "loss": 0.6955, "step": 10074 }, { "epoch": 0.09061351872754336, "grad_norm": 2.5849695205688477, "learning_rate": 9.624819638248822e-07, "loss": 0.3615, "step": 10075 }, { "epoch": 0.09062251262518381, "grad_norm": 12.184783935546875, "learning_rate": 9.604176978935343e-07, "loss": 0.6224, "step": 10076 }, { "epoch": 0.09063150652282427, "grad_norm": 1.9635200500488281, "learning_rate": 9.5835560463374e-07, "loss": 0.1999, "step": 10077 }, { "epoch": 0.09064050042046472, "grad_norm": 2.0620625019073486, "learning_rate": 9.562956842318726e-07, "loss": 0.5439, "step": 10078 }, { "epoch": 0.09064949431810516, "grad_norm": 2.079639434814453, "learning_rate": 9.542379368741055e-07, "loss": 0.614, "step": 10079 }, { "epoch": 0.09065848821574561, "grad_norm": 2.7216289043426514, "learning_rate": 9.521823627464121e-07, "loss": 0.3454, "step": 10080 }, { "epoch": 0.09066748211338607, "grad_norm": 10.50622844696045, "learning_rate": 9.501289620345716e-07, "loss": 0.6912, "step": 10081 }, { "epoch": 0.09067647601102652, "grad_norm": 16.312999725341797, "learning_rate": 9.480777349241742e-07, "loss": 0.5397, "step": 10082 }, { "epoch": 0.09068546990866697, "grad_norm": 14.763509750366211, "learning_rate": 9.460286816006026e-07, "loss": 0.4015, "step": 10083 }, { "epoch": 0.09069446380630743, "grad_norm": 2.223224639892578, "learning_rate": 9.439818022490526e-07, "loss": 0.4123, "step": 10084 }, { "epoch": 0.09070345770394787, "grad_norm": 2.7298262119293213, "learning_rate": 9.419370970545155e-07, "loss": 0.6757, "step": 10085 }, { "epoch": 0.09071245160158832, "grad_norm": 6.6825785636901855, "learning_rate": 9.398945662017877e-07, "loss": 0.6422, "step": 10086 }, { "epoch": 0.09072144549922877, "grad_norm": 13.574831008911133, "learning_rate": 9.378542098754745e-07, "loss": 0.5117, "step": 10087 }, { "epoch": 0.09073043939686923, "grad_norm": 3.3486578464508057, "learning_rate": 9.358160282599809e-07, "loss": 0.5463, "step": 10088 }, { "epoch": 0.09073943329450967, "grad_norm": 20.136802673339844, "learning_rate": 9.337800215395154e-07, "loss": 0.8434, "step": 10089 }, { "epoch": 0.09074842719215012, "grad_norm": 2.006225347518921, "learning_rate": 9.31746189898089e-07, "loss": 0.1071, "step": 10090 }, { "epoch": 0.09075742108979058, "grad_norm": 19.563758850097656, "learning_rate": 9.297145335195212e-07, "loss": 0.6606, "step": 10091 }, { "epoch": 0.09076641498743103, "grad_norm": 1.9075000286102295, "learning_rate": 9.276850525874237e-07, "loss": 0.7154, "step": 10092 }, { "epoch": 0.09077540888507148, "grad_norm": 2.0322697162628174, "learning_rate": 9.256577472852219e-07, "loss": 0.3681, "step": 10093 }, { "epoch": 0.09078440278271192, "grad_norm": 8.58722972869873, "learning_rate": 9.23632617796144e-07, "loss": 0.7323, "step": 10094 }, { "epoch": 0.09079339668035238, "grad_norm": 2.761979103088379, "learning_rate": 9.216096643032163e-07, "loss": 0.7358, "step": 10095 }, { "epoch": 0.09080239057799283, "grad_norm": 3.129180669784546, "learning_rate": 9.195888869892727e-07, "loss": 0.7277, "step": 10096 }, { "epoch": 0.09081138447563328, "grad_norm": 2.1610405445098877, "learning_rate": 9.175702860369478e-07, "loss": 0.6854, "step": 10097 }, { "epoch": 0.09082037837327374, "grad_norm": 12.942185401916504, "learning_rate": 9.155538616286791e-07, "loss": 0.4368, "step": 10098 }, { "epoch": 0.09082937227091419, "grad_norm": 3.635099172592163, "learning_rate": 9.135396139467151e-07, "loss": 0.5667, "step": 10099 }, { "epoch": 0.09083836616855463, "grad_norm": 14.390472412109375, "learning_rate": 9.115275431730936e-07, "loss": 0.5842, "step": 10100 }, { "epoch": 0.0908473600661951, "grad_norm": 3.661224603652954, "learning_rate": 9.095176494896663e-07, "loss": 0.6531, "step": 10101 }, { "epoch": 0.09085635396383554, "grad_norm": 16.58531951904297, "learning_rate": 9.075099330780906e-07, "loss": 0.319, "step": 10102 }, { "epoch": 0.09086534786147599, "grad_norm": 2.4569737911224365, "learning_rate": 9.055043941198105e-07, "loss": 0.3869, "step": 10103 }, { "epoch": 0.09087434175911643, "grad_norm": 3.143734931945801, "learning_rate": 9.035010327960947e-07, "loss": 0.4292, "step": 10104 }, { "epoch": 0.0908833356567569, "grad_norm": 9.513731002807617, "learning_rate": 9.014998492880011e-07, "loss": 0.5235, "step": 10105 }, { "epoch": 0.09089232955439734, "grad_norm": 2.779907703399658, "learning_rate": 8.995008437763936e-07, "loss": 0.2052, "step": 10106 }, { "epoch": 0.09090132345203779, "grad_norm": 3.139160633087158, "learning_rate": 8.975040164419385e-07, "loss": 0.3348, "step": 10107 }, { "epoch": 0.09091031734967825, "grad_norm": 2.622713804244995, "learning_rate": 8.955093674651139e-07, "loss": 0.6129, "step": 10108 }, { "epoch": 0.0909193112473187, "grad_norm": 11.701874732971191, "learning_rate": 8.935168970261864e-07, "loss": 0.7356, "step": 10109 }, { "epoch": 0.09092830514495914, "grad_norm": 2.117100715637207, "learning_rate": 8.915266053052373e-07, "loss": 0.5763, "step": 10110 }, { "epoch": 0.09093729904259959, "grad_norm": 1.8555357456207275, "learning_rate": 8.895384924821448e-07, "loss": 0.3627, "step": 10111 }, { "epoch": 0.09094629294024005, "grad_norm": 3.0023720264434814, "learning_rate": 8.875525587365929e-07, "loss": 0.2819, "step": 10112 }, { "epoch": 0.0909552868378805, "grad_norm": 19.129331588745117, "learning_rate": 8.855688042480714e-07, "loss": 0.7096, "step": 10113 }, { "epoch": 0.09096428073552094, "grad_norm": 5.007407188415527, "learning_rate": 8.835872291958674e-07, "loss": 0.263, "step": 10114 }, { "epoch": 0.0909732746331614, "grad_norm": 2.8342387676239014, "learning_rate": 8.816078337590711e-07, "loss": 0.0634, "step": 10115 }, { "epoch": 0.09098226853080185, "grad_norm": 3.589451789855957, "learning_rate": 8.796306181165836e-07, "loss": 0.524, "step": 10116 }, { "epoch": 0.0909912624284423, "grad_norm": 1.9905812740325928, "learning_rate": 8.776555824470983e-07, "loss": 0.6298, "step": 10117 }, { "epoch": 0.09100025632608276, "grad_norm": 3.4595210552215576, "learning_rate": 8.756827269291224e-07, "loss": 0.2706, "step": 10118 }, { "epoch": 0.0910092502237232, "grad_norm": 19.89447593688965, "learning_rate": 8.737120517409548e-07, "loss": 0.6923, "step": 10119 }, { "epoch": 0.09101824412136365, "grad_norm": 2.30637788772583, "learning_rate": 8.717435570607058e-07, "loss": 0.3755, "step": 10120 }, { "epoch": 0.0910272380190041, "grad_norm": 12.40904712677002, "learning_rate": 8.697772430662859e-07, "loss": 0.8161, "step": 10121 }, { "epoch": 0.09103623191664456, "grad_norm": 10.961104393005371, "learning_rate": 8.678131099354059e-07, "loss": 0.5007, "step": 10122 }, { "epoch": 0.09104522581428501, "grad_norm": 15.617749214172363, "learning_rate": 8.658511578455846e-07, "loss": 0.1761, "step": 10123 }, { "epoch": 0.09105421971192545, "grad_norm": 12.38858699798584, "learning_rate": 8.638913869741388e-07, "loss": 0.4918, "step": 10124 }, { "epoch": 0.09106321360956592, "grad_norm": 7.131653785705566, "learning_rate": 8.619337974981934e-07, "loss": 0.4995, "step": 10125 }, { "epoch": 0.09107220750720636, "grad_norm": 2.3855037689208984, "learning_rate": 8.599783895946761e-07, "loss": 0.4158, "step": 10126 }, { "epoch": 0.09108120140484681, "grad_norm": 3.009195327758789, "learning_rate": 8.580251634403041e-07, "loss": 0.3562, "step": 10127 }, { "epoch": 0.09109019530248726, "grad_norm": 1.9213486909866333, "learning_rate": 8.560741192116168e-07, "loss": 0.5719, "step": 10128 }, { "epoch": 0.09109918920012772, "grad_norm": 8.625975608825684, "learning_rate": 8.541252570849451e-07, "loss": 0.4686, "step": 10129 }, { "epoch": 0.09110818309776816, "grad_norm": 3.375809669494629, "learning_rate": 8.52178577236426e-07, "loss": 0.3341, "step": 10130 }, { "epoch": 0.09111717699540861, "grad_norm": 2.165456771850586, "learning_rate": 8.502340798419939e-07, "loss": 0.1509, "step": 10131 }, { "epoch": 0.09112617089304907, "grad_norm": 3.9734323024749756, "learning_rate": 8.48291765077397e-07, "loss": 0.2784, "step": 10132 }, { "epoch": 0.09113516479068952, "grad_norm": 2.3147683143615723, "learning_rate": 8.463516331181753e-07, "loss": 0.2676, "step": 10133 }, { "epoch": 0.09114415868832997, "grad_norm": 2.376091718673706, "learning_rate": 8.444136841396777e-07, "loss": 0.3078, "step": 10134 }, { "epoch": 0.09115315258597041, "grad_norm": 2.209789991378784, "learning_rate": 8.424779183170528e-07, "loss": 0.2585, "step": 10135 }, { "epoch": 0.09116214648361087, "grad_norm": 3.3826873302459717, "learning_rate": 8.405443358252524e-07, "loss": 0.4052, "step": 10136 }, { "epoch": 0.09117114038125132, "grad_norm": 4.2342705726623535, "learning_rate": 8.38612936839031e-07, "loss": 0.4862, "step": 10137 }, { "epoch": 0.09118013427889177, "grad_norm": 2.2390458583831787, "learning_rate": 8.366837215329521e-07, "loss": 0.1741, "step": 10138 }, { "epoch": 0.09118912817653223, "grad_norm": 2.776228189468384, "learning_rate": 8.347566900813675e-07, "loss": 0.2969, "step": 10139 }, { "epoch": 0.09119812207417267, "grad_norm": 14.685256004333496, "learning_rate": 8.328318426584463e-07, "loss": 0.354, "step": 10140 }, { "epoch": 0.09120711597181312, "grad_norm": 2.309504508972168, "learning_rate": 8.309091794381524e-07, "loss": 0.3519, "step": 10141 }, { "epoch": 0.09121610986945358, "grad_norm": 2.1203465461730957, "learning_rate": 8.289887005942548e-07, "loss": 0.1355, "step": 10142 }, { "epoch": 0.09122510376709403, "grad_norm": 2.9693045616149902, "learning_rate": 8.270704063003232e-07, "loss": 0.177, "step": 10143 }, { "epoch": 0.09123409766473448, "grad_norm": 4.773801326751709, "learning_rate": 8.251542967297299e-07, "loss": 0.2852, "step": 10144 }, { "epoch": 0.09124309156237492, "grad_norm": 15.065667152404785, "learning_rate": 8.232403720556531e-07, "loss": 0.4728, "step": 10145 }, { "epoch": 0.09125208546001538, "grad_norm": 2.3336410522460938, "learning_rate": 8.213286324510738e-07, "loss": 0.7127, "step": 10146 }, { "epoch": 0.09126107935765583, "grad_norm": 2.493847608566284, "learning_rate": 8.194190780887678e-07, "loss": 0.5542, "step": 10147 }, { "epoch": 0.09127007325529628, "grad_norm": 18.64592170715332, "learning_rate": 8.175117091413165e-07, "loss": 0.3798, "step": 10148 }, { "epoch": 0.09127906715293674, "grad_norm": 2.614415168762207, "learning_rate": 8.156065257811124e-07, "loss": 0.6899, "step": 10149 }, { "epoch": 0.09128806105057719, "grad_norm": 1.9277855157852173, "learning_rate": 8.137035281803429e-07, "loss": 0.3792, "step": 10150 }, { "epoch": 0.09129705494821763, "grad_norm": 10.531218528747559, "learning_rate": 8.118027165109926e-07, "loss": 0.6458, "step": 10151 }, { "epoch": 0.09130604884585808, "grad_norm": 1.9823702573776245, "learning_rate": 8.09904090944863e-07, "loss": 0.5175, "step": 10152 }, { "epoch": 0.09131504274349854, "grad_norm": 3.104403018951416, "learning_rate": 8.080076516535445e-07, "loss": 0.209, "step": 10153 }, { "epoch": 0.09132403664113899, "grad_norm": 2.3995299339294434, "learning_rate": 8.061133988084391e-07, "loss": 0.2872, "step": 10154 }, { "epoch": 0.09133303053877943, "grad_norm": 18.537376403808594, "learning_rate": 8.042213325807457e-07, "loss": 0.3632, "step": 10155 }, { "epoch": 0.0913420244364199, "grad_norm": 2.3009517192840576, "learning_rate": 8.023314531414638e-07, "loss": 0.5859, "step": 10156 }, { "epoch": 0.09135101833406034, "grad_norm": 2.0223867893218994, "learning_rate": 8.00443760661404e-07, "loss": 0.1646, "step": 10157 }, { "epoch": 0.09136001223170079, "grad_norm": 2.2746119499206543, "learning_rate": 7.985582553111687e-07, "loss": 0.2573, "step": 10158 }, { "epoch": 0.09136900612934125, "grad_norm": 2.6787893772125244, "learning_rate": 7.966749372611715e-07, "loss": 0.2924, "step": 10159 }, { "epoch": 0.0913780000269817, "grad_norm": 2.2220981121063232, "learning_rate": 7.947938066816262e-07, "loss": 0.1815, "step": 10160 }, { "epoch": 0.09138699392462214, "grad_norm": 16.410903930664062, "learning_rate": 7.929148637425443e-07, "loss": 0.885, "step": 10161 }, { "epoch": 0.09139598782226259, "grad_norm": 2.205561876296997, "learning_rate": 7.910381086137425e-07, "loss": 0.3692, "step": 10162 }, { "epoch": 0.09140498171990305, "grad_norm": 3.7060108184814453, "learning_rate": 7.891635414648379e-07, "loss": 0.3469, "step": 10163 }, { "epoch": 0.0914139756175435, "grad_norm": 19.266834259033203, "learning_rate": 7.87291162465259e-07, "loss": 0.5621, "step": 10164 }, { "epoch": 0.09142296951518394, "grad_norm": 2.4919774532318115, "learning_rate": 7.854209717842231e-07, "loss": 0.2212, "step": 10165 }, { "epoch": 0.0914319634128244, "grad_norm": 2.584719181060791, "learning_rate": 7.83552969590759e-07, "loss": 0.1933, "step": 10166 }, { "epoch": 0.09144095731046485, "grad_norm": 17.447641372680664, "learning_rate": 7.816871560536925e-07, "loss": 0.4298, "step": 10167 }, { "epoch": 0.0914499512081053, "grad_norm": 1.8860012292861938, "learning_rate": 7.79823531341653e-07, "loss": 0.4527, "step": 10168 }, { "epoch": 0.09145894510574575, "grad_norm": 2.5758728981018066, "learning_rate": 7.779620956230804e-07, "loss": 0.1903, "step": 10169 }, { "epoch": 0.0914679390033862, "grad_norm": 2.8009963035583496, "learning_rate": 7.761028490661987e-07, "loss": 0.4654, "step": 10170 }, { "epoch": 0.09147693290102665, "grad_norm": 2.6039927005767822, "learning_rate": 7.742457918390511e-07, "loss": 0.1399, "step": 10171 }, { "epoch": 0.0914859267986671, "grad_norm": 3.27919602394104, "learning_rate": 7.723909241094756e-07, "loss": 0.0788, "step": 10172 }, { "epoch": 0.09149492069630756, "grad_norm": 2.9897713661193848, "learning_rate": 7.705382460451133e-07, "loss": 0.4001, "step": 10173 }, { "epoch": 0.09150391459394801, "grad_norm": 2.154838800430298, "learning_rate": 7.68687757813405e-07, "loss": 0.4121, "step": 10174 }, { "epoch": 0.09151290849158845, "grad_norm": 12.657393455505371, "learning_rate": 7.668394595816003e-07, "loss": 0.7273, "step": 10175 }, { "epoch": 0.0915219023892289, "grad_norm": 17.24180793762207, "learning_rate": 7.649933515167407e-07, "loss": 0.2962, "step": 10176 }, { "epoch": 0.09153089628686936, "grad_norm": 1.7124098539352417, "learning_rate": 7.631494337856788e-07, "loss": 0.4948, "step": 10177 }, { "epoch": 0.09153989018450981, "grad_norm": 3.164797067642212, "learning_rate": 7.613077065550644e-07, "loss": 0.4058, "step": 10178 }, { "epoch": 0.09154888408215026, "grad_norm": 3.070119857788086, "learning_rate": 7.594681699913536e-07, "loss": 0.5575, "step": 10179 }, { "epoch": 0.09155787797979072, "grad_norm": 2.0733954906463623, "learning_rate": 7.576308242607965e-07, "loss": 0.3067, "step": 10180 }, { "epoch": 0.09156687187743116, "grad_norm": 8.395702362060547, "learning_rate": 7.557956695294577e-07, "loss": 0.7682, "step": 10181 }, { "epoch": 0.09157586577507161, "grad_norm": 2.933793306350708, "learning_rate": 7.539627059631904e-07, "loss": 0.5432, "step": 10182 }, { "epoch": 0.09158485967271207, "grad_norm": 2.750368356704712, "learning_rate": 7.521319337276594e-07, "loss": 0.5284, "step": 10183 }, { "epoch": 0.09159385357035252, "grad_norm": 8.060389518737793, "learning_rate": 7.503033529883241e-07, "loss": 0.4662, "step": 10184 }, { "epoch": 0.09160284746799296, "grad_norm": 3.1312286853790283, "learning_rate": 7.484769639104549e-07, "loss": 0.5673, "step": 10185 }, { "epoch": 0.09161184136563341, "grad_norm": 2.4703125953674316, "learning_rate": 7.466527666591144e-07, "loss": 0.524, "step": 10186 }, { "epoch": 0.09162083526327387, "grad_norm": 13.77924633026123, "learning_rate": 7.448307613991734e-07, "loss": 0.5721, "step": 10187 }, { "epoch": 0.09162982916091432, "grad_norm": 34.17787551879883, "learning_rate": 7.430109482953029e-07, "loss": 0.4758, "step": 10188 }, { "epoch": 0.09163882305855477, "grad_norm": 2.6093392372131348, "learning_rate": 7.41193327511977e-07, "loss": 0.398, "step": 10189 }, { "epoch": 0.09164781695619523, "grad_norm": 10.857224464416504, "learning_rate": 7.39377899213467e-07, "loss": 0.4794, "step": 10190 }, { "epoch": 0.09165681085383567, "grad_norm": 2.7710554599761963, "learning_rate": 7.375646635638528e-07, "loss": 0.395, "step": 10191 }, { "epoch": 0.09166580475147612, "grad_norm": 14.870630264282227, "learning_rate": 7.357536207270088e-07, "loss": 0.6636, "step": 10192 }, { "epoch": 0.09167479864911657, "grad_norm": 14.900736808776855, "learning_rate": 7.339447708666181e-07, "loss": 0.6429, "step": 10193 }, { "epoch": 0.09168379254675703, "grad_norm": 2.4773354530334473, "learning_rate": 7.321381141461609e-07, "loss": 0.3864, "step": 10194 }, { "epoch": 0.09169278644439748, "grad_norm": 15.30113697052002, "learning_rate": 7.303336507289232e-07, "loss": 0.6109, "step": 10195 }, { "epoch": 0.09170178034203792, "grad_norm": 17.26283073425293, "learning_rate": 7.285313807779859e-07, "loss": 0.5314, "step": 10196 }, { "epoch": 0.09171077423967838, "grad_norm": 4.121975421905518, "learning_rate": 7.267313044562435e-07, "loss": 0.2652, "step": 10197 }, { "epoch": 0.09171976813731883, "grad_norm": 2.0933573246002197, "learning_rate": 7.249334219263799e-07, "loss": 0.5538, "step": 10198 }, { "epoch": 0.09172876203495928, "grad_norm": 2.431226968765259, "learning_rate": 7.231377333508843e-07, "loss": 0.2982, "step": 10199 }, { "epoch": 0.09173775593259974, "grad_norm": 3.2214791774749756, "learning_rate": 7.213442388920549e-07, "loss": 0.2898, "step": 10200 }, { "epoch": 0.09174674983024018, "grad_norm": 2.179311990737915, "learning_rate": 7.195529387119815e-07, "loss": 0.6463, "step": 10201 }, { "epoch": 0.09175574372788063, "grad_norm": 2.9209985733032227, "learning_rate": 7.177638329725567e-07, "loss": 0.3977, "step": 10202 }, { "epoch": 0.09176473762552108, "grad_norm": 2.360714912414551, "learning_rate": 7.159769218354873e-07, "loss": 0.3209, "step": 10203 }, { "epoch": 0.09177373152316154, "grad_norm": 3.7990031242370605, "learning_rate": 7.141922054622663e-07, "loss": 0.6333, "step": 10204 }, { "epoch": 0.09178272542080199, "grad_norm": 2.282663345336914, "learning_rate": 7.124096840141952e-07, "loss": 0.222, "step": 10205 }, { "epoch": 0.09179171931844243, "grad_norm": 1.725362777709961, "learning_rate": 7.106293576523731e-07, "loss": 0.3109, "step": 10206 }, { "epoch": 0.0918007132160829, "grad_norm": 2.079333543777466, "learning_rate": 7.088512265377101e-07, "loss": 0.3951, "step": 10207 }, { "epoch": 0.09180970711372334, "grad_norm": 14.483880996704102, "learning_rate": 7.070752908309052e-07, "loss": 0.4499, "step": 10208 }, { "epoch": 0.09181870101136379, "grad_norm": 1.9354444742202759, "learning_rate": 7.053015506924748e-07, "loss": 0.617, "step": 10209 }, { "epoch": 0.09182769490900423, "grad_norm": 3.059019088745117, "learning_rate": 7.035300062827182e-07, "loss": 0.6074, "step": 10210 }, { "epoch": 0.0918366888066447, "grad_norm": 15.067590713500977, "learning_rate": 7.017606577617519e-07, "loss": 0.2603, "step": 10211 }, { "epoch": 0.09184568270428514, "grad_norm": 2.335176467895508, "learning_rate": 6.99993505289484e-07, "loss": 0.3963, "step": 10212 }, { "epoch": 0.09185467660192559, "grad_norm": 12.340312957763672, "learning_rate": 6.982285490256285e-07, "loss": 0.784, "step": 10213 }, { "epoch": 0.09186367049956605, "grad_norm": 15.093340873718262, "learning_rate": 6.964657891297049e-07, "loss": 0.3373, "step": 10214 }, { "epoch": 0.0918726643972065, "grad_norm": 2.052055835723877, "learning_rate": 6.94705225761022e-07, "loss": 0.1195, "step": 10215 }, { "epoch": 0.09188165829484694, "grad_norm": 3.5023086071014404, "learning_rate": 6.929468590787053e-07, "loss": 0.3127, "step": 10216 }, { "epoch": 0.0918906521924874, "grad_norm": 2.82344913482666, "learning_rate": 6.911906892416664e-07, "loss": 0.6067, "step": 10217 }, { "epoch": 0.09189964609012785, "grad_norm": 2.1400508880615234, "learning_rate": 6.894367164086285e-07, "loss": 0.2197, "step": 10218 }, { "epoch": 0.0919086399877683, "grad_norm": 14.428130149841309, "learning_rate": 6.876849407381203e-07, "loss": 0.4713, "step": 10219 }, { "epoch": 0.09191763388540874, "grad_norm": 19.55944061279297, "learning_rate": 6.859353623884567e-07, "loss": 0.5618, "step": 10220 }, { "epoch": 0.0919266277830492, "grad_norm": 9.74146556854248, "learning_rate": 6.841879815177643e-07, "loss": 0.615, "step": 10221 }, { "epoch": 0.09193562168068965, "grad_norm": 1.9268678426742554, "learning_rate": 6.824427982839748e-07, "loss": 0.3351, "step": 10222 }, { "epoch": 0.0919446155783301, "grad_norm": 2.1858038902282715, "learning_rate": 6.806998128448094e-07, "loss": 0.3083, "step": 10223 }, { "epoch": 0.09195360947597056, "grad_norm": 2.4578614234924316, "learning_rate": 6.789590253578032e-07, "loss": 0.6972, "step": 10224 }, { "epoch": 0.09196260337361101, "grad_norm": 16.981767654418945, "learning_rate": 6.772204359802803e-07, "loss": 0.6037, "step": 10225 }, { "epoch": 0.09197159727125145, "grad_norm": 2.6648013591766357, "learning_rate": 6.754840448693789e-07, "loss": 0.1558, "step": 10226 }, { "epoch": 0.0919805911688919, "grad_norm": 12.876680374145508, "learning_rate": 6.73749852182029e-07, "loss": 0.2886, "step": 10227 }, { "epoch": 0.09198958506653236, "grad_norm": 2.4820704460144043, "learning_rate": 6.720178580749637e-07, "loss": 0.1803, "step": 10228 }, { "epoch": 0.09199857896417281, "grad_norm": 13.548033714294434, "learning_rate": 6.702880627047214e-07, "loss": 0.649, "step": 10229 }, { "epoch": 0.09200757286181326, "grad_norm": 2.004582643508911, "learning_rate": 6.685604662276413e-07, "loss": 0.3997, "step": 10230 }, { "epoch": 0.09201656675945372, "grad_norm": 3.3142783641815186, "learning_rate": 6.668350687998565e-07, "loss": 0.6019, "step": 10231 }, { "epoch": 0.09202556065709416, "grad_norm": 2.9585015773773193, "learning_rate": 6.65111870577309e-07, "loss": 0.4617, "step": 10232 }, { "epoch": 0.09203455455473461, "grad_norm": 2.0584659576416016, "learning_rate": 6.633908717157355e-07, "loss": 0.4087, "step": 10233 }, { "epoch": 0.09204354845237506, "grad_norm": 17.49829864501953, "learning_rate": 6.616720723706865e-07, "loss": 0.4345, "step": 10234 }, { "epoch": 0.09205254235001552, "grad_norm": 3.174543619155884, "learning_rate": 6.59955472697496e-07, "loss": 0.3968, "step": 10235 }, { "epoch": 0.09206153624765596, "grad_norm": 15.74765682220459, "learning_rate": 6.582410728513178e-07, "loss": 0.6156, "step": 10236 }, { "epoch": 0.09207053014529641, "grad_norm": 2.614473819732666, "learning_rate": 6.565288729870889e-07, "loss": 0.5314, "step": 10237 }, { "epoch": 0.09207952404293687, "grad_norm": 3.082075357437134, "learning_rate": 6.548188732595606e-07, "loss": 0.2103, "step": 10238 }, { "epoch": 0.09208851794057732, "grad_norm": 2.9937851428985596, "learning_rate": 6.531110738232815e-07, "loss": 0.5007, "step": 10239 }, { "epoch": 0.09209751183821777, "grad_norm": 15.544570922851562, "learning_rate": 6.514054748325976e-07, "loss": 0.3561, "step": 10240 }, { "epoch": 0.09210650573585823, "grad_norm": 1.9441697597503662, "learning_rate": 6.497020764416633e-07, "loss": 0.6313, "step": 10241 }, { "epoch": 0.09211549963349867, "grad_norm": 2.329266309738159, "learning_rate": 6.480008788044223e-07, "loss": 0.7635, "step": 10242 }, { "epoch": 0.09212449353113912, "grad_norm": 2.793558120727539, "learning_rate": 6.463018820746375e-07, "loss": 0.2336, "step": 10243 }, { "epoch": 0.09213348742877957, "grad_norm": 2.2764780521392822, "learning_rate": 6.446050864058556e-07, "loss": 0.1788, "step": 10244 }, { "epoch": 0.09214248132642003, "grad_norm": 2.698702573776245, "learning_rate": 6.429104919514317e-07, "loss": 0.2153, "step": 10245 }, { "epoch": 0.09215147522406048, "grad_norm": 15.696615219116211, "learning_rate": 6.412180988645239e-07, "loss": 0.3856, "step": 10246 }, { "epoch": 0.09216046912170092, "grad_norm": 2.447632074356079, "learning_rate": 6.395279072980848e-07, "loss": 0.2509, "step": 10247 }, { "epoch": 0.09216946301934138, "grad_norm": 16.470176696777344, "learning_rate": 6.378399174048755e-07, "loss": 0.7158, "step": 10248 }, { "epoch": 0.09217845691698183, "grad_norm": 2.363779067993164, "learning_rate": 6.361541293374518e-07, "loss": 0.4656, "step": 10249 }, { "epoch": 0.09218745081462228, "grad_norm": 2.6583635807037354, "learning_rate": 6.344705432481779e-07, "loss": 0.3308, "step": 10250 }, { "epoch": 0.09219644471226272, "grad_norm": 16.440711975097656, "learning_rate": 6.327891592892127e-07, "loss": 0.5698, "step": 10251 }, { "epoch": 0.09220543860990318, "grad_norm": 3.4546964168548584, "learning_rate": 6.311099776125151e-07, "loss": 0.3338, "step": 10252 }, { "epoch": 0.09221443250754363, "grad_norm": 2.1976816654205322, "learning_rate": 6.2943299836985e-07, "loss": 0.3641, "step": 10253 }, { "epoch": 0.09222342640518408, "grad_norm": 2.9919795989990234, "learning_rate": 6.277582217127792e-07, "loss": 0.3138, "step": 10254 }, { "epoch": 0.09223242030282454, "grad_norm": 2.7047014236450195, "learning_rate": 6.260856477926707e-07, "loss": 0.4983, "step": 10255 }, { "epoch": 0.09224141420046499, "grad_norm": 2.029925584793091, "learning_rate": 6.244152767606897e-07, "loss": 0.1322, "step": 10256 }, { "epoch": 0.09225040809810543, "grad_norm": 1.8446348905563354, "learning_rate": 6.227471087677961e-07, "loss": 0.4083, "step": 10257 }, { "epoch": 0.0922594019957459, "grad_norm": 10.931936264038086, "learning_rate": 6.210811439647663e-07, "loss": 0.7306, "step": 10258 }, { "epoch": 0.09226839589338634, "grad_norm": 2.0737762451171875, "learning_rate": 6.194173825021604e-07, "loss": 0.2557, "step": 10259 }, { "epoch": 0.09227738979102679, "grad_norm": 2.756702423095703, "learning_rate": 6.177558245303555e-07, "loss": 0.3048, "step": 10260 }, { "epoch": 0.09228638368866723, "grad_norm": 6.325134754180908, "learning_rate": 6.16096470199512e-07, "loss": 0.7478, "step": 10261 }, { "epoch": 0.0922953775863077, "grad_norm": 2.7517030239105225, "learning_rate": 6.144393196596071e-07, "loss": 0.3474, "step": 10262 }, { "epoch": 0.09230437148394814, "grad_norm": 3.9746758937835693, "learning_rate": 6.127843730604099e-07, "loss": 0.4322, "step": 10263 }, { "epoch": 0.09231336538158859, "grad_norm": 3.457892656326294, "learning_rate": 6.111316305514925e-07, "loss": 0.3645, "step": 10264 }, { "epoch": 0.09232235927922905, "grad_norm": 10.260929107666016, "learning_rate": 6.094810922822325e-07, "loss": 0.4635, "step": 10265 }, { "epoch": 0.0923313531768695, "grad_norm": 2.39544677734375, "learning_rate": 6.078327584017968e-07, "loss": 0.5957, "step": 10266 }, { "epoch": 0.09234034707450994, "grad_norm": 2.090984344482422, "learning_rate": 6.061866290591634e-07, "loss": 0.2969, "step": 10267 }, { "epoch": 0.09234934097215039, "grad_norm": 2.9129581451416016, "learning_rate": 6.045427044031105e-07, "loss": 0.1459, "step": 10268 }, { "epoch": 0.09235833486979085, "grad_norm": 2.695263147354126, "learning_rate": 6.029009845822082e-07, "loss": 0.7461, "step": 10269 }, { "epoch": 0.0923673287674313, "grad_norm": 16.216754913330078, "learning_rate": 6.012614697448376e-07, "loss": 0.6763, "step": 10270 }, { "epoch": 0.09237632266507174, "grad_norm": 2.953138828277588, "learning_rate": 5.99624160039175e-07, "loss": 0.5685, "step": 10271 }, { "epoch": 0.0923853165627122, "grad_norm": 2.368096351623535, "learning_rate": 5.979890556132017e-07, "loss": 0.2419, "step": 10272 }, { "epoch": 0.09239431046035265, "grad_norm": 2.3246922492980957, "learning_rate": 5.963561566146885e-07, "loss": 0.5305, "step": 10273 }, { "epoch": 0.0924033043579931, "grad_norm": 13.314128875732422, "learning_rate": 5.94725463191223e-07, "loss": 0.6542, "step": 10274 }, { "epoch": 0.09241229825563355, "grad_norm": 3.4433701038360596, "learning_rate": 5.930969754901843e-07, "loss": 0.8953, "step": 10275 }, { "epoch": 0.092421292153274, "grad_norm": 2.6519620418548584, "learning_rate": 5.914706936587494e-07, "loss": 0.5512, "step": 10276 }, { "epoch": 0.09243028605091445, "grad_norm": 2.6125543117523193, "learning_rate": 5.898466178439061e-07, "loss": 0.3032, "step": 10277 }, { "epoch": 0.0924392799485549, "grad_norm": 2.888674259185791, "learning_rate": 5.882247481924285e-07, "loss": 0.622, "step": 10278 }, { "epoch": 0.09244827384619536, "grad_norm": 2.350192070007324, "learning_rate": 5.866050848509081e-07, "loss": 0.5118, "step": 10279 }, { "epoch": 0.09245726774383581, "grad_norm": 2.149148464202881, "learning_rate": 5.849876279657219e-07, "loss": 0.4848, "step": 10280 }, { "epoch": 0.09246626164147626, "grad_norm": 2.89395809173584, "learning_rate": 5.833723776830558e-07, "loss": 0.4778, "step": 10281 }, { "epoch": 0.09247525553911672, "grad_norm": 10.52501392364502, "learning_rate": 5.817593341488959e-07, "loss": 0.5824, "step": 10282 }, { "epoch": 0.09248424943675716, "grad_norm": 5.463583946228027, "learning_rate": 5.801484975090254e-07, "loss": 0.7252, "step": 10283 }, { "epoch": 0.09249324333439761, "grad_norm": 3.258859872817993, "learning_rate": 5.785398679090337e-07, "loss": 0.3982, "step": 10284 }, { "epoch": 0.09250223723203806, "grad_norm": 2.372772455215454, "learning_rate": 5.769334454943016e-07, "loss": 0.4851, "step": 10285 }, { "epoch": 0.09251123112967852, "grad_norm": 13.416961669921875, "learning_rate": 5.753292304100183e-07, "loss": 0.6182, "step": 10286 }, { "epoch": 0.09252022502731896, "grad_norm": 2.9092326164245605, "learning_rate": 5.737272228011736e-07, "loss": 0.1723, "step": 10287 }, { "epoch": 0.09252921892495941, "grad_norm": 25.7786808013916, "learning_rate": 5.721274228125489e-07, "loss": 0.8933, "step": 10288 }, { "epoch": 0.09253821282259987, "grad_norm": 2.4770216941833496, "learning_rate": 5.705298305887397e-07, "loss": 0.506, "step": 10289 }, { "epoch": 0.09254720672024032, "grad_norm": 2.2147836685180664, "learning_rate": 5.689344462741275e-07, "loss": 0.375, "step": 10290 }, { "epoch": 0.09255620061788077, "grad_norm": 9.514796257019043, "learning_rate": 5.673412700129054e-07, "loss": 0.4854, "step": 10291 }, { "epoch": 0.09256519451552121, "grad_norm": 2.0608417987823486, "learning_rate": 5.657503019490667e-07, "loss": 0.4859, "step": 10292 }, { "epoch": 0.09257418841316167, "grad_norm": 1.9468642473220825, "learning_rate": 5.641615422263935e-07, "loss": 0.2526, "step": 10293 }, { "epoch": 0.09258318231080212, "grad_norm": 2.2003347873687744, "learning_rate": 5.62574990988482e-07, "loss": 0.31, "step": 10294 }, { "epoch": 0.09259217620844257, "grad_norm": 2.491312026977539, "learning_rate": 5.609906483787175e-07, "loss": 0.488, "step": 10295 }, { "epoch": 0.09260117010608303, "grad_norm": 11.127561569213867, "learning_rate": 5.594085145402967e-07, "loss": 0.9046, "step": 10296 }, { "epoch": 0.09261016400372347, "grad_norm": 2.196563482284546, "learning_rate": 5.578285896162106e-07, "loss": 0.5149, "step": 10297 }, { "epoch": 0.09261915790136392, "grad_norm": 2.0148627758026123, "learning_rate": 5.562508737492478e-07, "loss": 0.4356, "step": 10298 }, { "epoch": 0.09262815179900438, "grad_norm": 1.8347889184951782, "learning_rate": 5.546753670820054e-07, "loss": 0.4289, "step": 10299 }, { "epoch": 0.09263714569664483, "grad_norm": 3.0563971996307373, "learning_rate": 5.531020697568723e-07, "loss": 0.4379, "step": 10300 }, { "epoch": 0.09264613959428528, "grad_norm": 5.2119317054748535, "learning_rate": 5.515309819160403e-07, "loss": 0.2043, "step": 10301 }, { "epoch": 0.09265513349192572, "grad_norm": 5.800148010253906, "learning_rate": 5.49962103701504e-07, "loss": 0.6959, "step": 10302 }, { "epoch": 0.09266412738956618, "grad_norm": 6.160131454467773, "learning_rate": 5.483954352550585e-07, "loss": 0.5991, "step": 10303 }, { "epoch": 0.09267312128720663, "grad_norm": 16.555988311767578, "learning_rate": 5.46830976718296e-07, "loss": 0.5232, "step": 10304 }, { "epoch": 0.09268211518484708, "grad_norm": 16.99956703186035, "learning_rate": 5.452687282326091e-07, "loss": 0.5165, "step": 10305 }, { "epoch": 0.09269110908248754, "grad_norm": 2.5698118209838867, "learning_rate": 5.437086899391985e-07, "loss": 0.1486, "step": 10306 }, { "epoch": 0.09270010298012799, "grad_norm": 16.145360946655273, "learning_rate": 5.421508619790516e-07, "loss": 0.5616, "step": 10307 }, { "epoch": 0.09270909687776843, "grad_norm": 13.18730640411377, "learning_rate": 5.405952444929668e-07, "loss": 0.9326, "step": 10308 }, { "epoch": 0.09271809077540888, "grad_norm": 5.6510419845581055, "learning_rate": 5.390418376215373e-07, "loss": 0.2285, "step": 10309 }, { "epoch": 0.09272708467304934, "grad_norm": 1.6758183240890503, "learning_rate": 5.374906415051589e-07, "loss": 0.3497, "step": 10310 }, { "epoch": 0.09273607857068979, "grad_norm": 3.940402030944824, "learning_rate": 5.359416562840281e-07, "loss": 0.2844, "step": 10311 }, { "epoch": 0.09274507246833023, "grad_norm": 3.021186351776123, "learning_rate": 5.343948820981409e-07, "loss": 0.7934, "step": 10312 }, { "epoch": 0.0927540663659707, "grad_norm": 2.1988847255706787, "learning_rate": 5.328503190872913e-07, "loss": 0.3677, "step": 10313 }, { "epoch": 0.09276306026361114, "grad_norm": 3.3534555435180664, "learning_rate": 5.31307967391076e-07, "loss": 0.2699, "step": 10314 }, { "epoch": 0.09277205416125159, "grad_norm": 2.451180934906006, "learning_rate": 5.297678271488943e-07, "loss": 0.4355, "step": 10315 }, { "epoch": 0.09278104805889203, "grad_norm": 2.3583781719207764, "learning_rate": 5.282298984999351e-07, "loss": 0.5848, "step": 10316 }, { "epoch": 0.0927900419565325, "grad_norm": 15.787081718444824, "learning_rate": 5.266941815831983e-07, "loss": 0.3632, "step": 10317 }, { "epoch": 0.09279903585417294, "grad_norm": 2.126573324203491, "learning_rate": 5.25160676537484e-07, "loss": 0.6153, "step": 10318 }, { "epoch": 0.09280802975181339, "grad_norm": 3.1150996685028076, "learning_rate": 5.236293835013839e-07, "loss": 0.3316, "step": 10319 }, { "epoch": 0.09281702364945385, "grad_norm": 4.565044403076172, "learning_rate": 5.221003026132959e-07, "loss": 0.2054, "step": 10320 }, { "epoch": 0.0928260175470943, "grad_norm": 19.80307960510254, "learning_rate": 5.205734340114205e-07, "loss": 0.4554, "step": 10321 }, { "epoch": 0.09283501144473474, "grad_norm": 2.562147617340088, "learning_rate": 5.190487778337472e-07, "loss": 0.7529, "step": 10322 }, { "epoch": 0.0928440053423752, "grad_norm": 7.9344682693481445, "learning_rate": 5.175263342180769e-07, "loss": 0.9576, "step": 10323 }, { "epoch": 0.09285299924001565, "grad_norm": 9.568277359008789, "learning_rate": 5.160061033020053e-07, "loss": 0.3966, "step": 10324 }, { "epoch": 0.0928619931376561, "grad_norm": 2.4867539405822754, "learning_rate": 5.144880852229333e-07, "loss": 0.1828, "step": 10325 }, { "epoch": 0.09287098703529655, "grad_norm": 5.274918556213379, "learning_rate": 5.129722801180542e-07, "loss": 0.5955, "step": 10326 }, { "epoch": 0.092879980932937, "grad_norm": 2.2439379692077637, "learning_rate": 5.114586881243638e-07, "loss": 0.4483, "step": 10327 }, { "epoch": 0.09288897483057745, "grad_norm": 2.5845136642456055, "learning_rate": 5.099473093786611e-07, "loss": 0.5066, "step": 10328 }, { "epoch": 0.0928979687282179, "grad_norm": 2.447660446166992, "learning_rate": 5.084381440175423e-07, "loss": 0.2693, "step": 10329 }, { "epoch": 0.09290696262585836, "grad_norm": 2.121534585952759, "learning_rate": 5.069311921774039e-07, "loss": 0.4474, "step": 10330 }, { "epoch": 0.09291595652349881, "grad_norm": 2.276566505432129, "learning_rate": 5.054264539944398e-07, "loss": 0.7064, "step": 10331 }, { "epoch": 0.09292495042113925, "grad_norm": 3.1716089248657227, "learning_rate": 5.039239296046522e-07, "loss": 0.1684, "step": 10332 }, { "epoch": 0.0929339443187797, "grad_norm": 3.223259687423706, "learning_rate": 5.024236191438353e-07, "loss": 0.4272, "step": 10333 }, { "epoch": 0.09294293821642016, "grad_norm": 2.193173885345459, "learning_rate": 5.009255227475834e-07, "loss": 0.3667, "step": 10334 }, { "epoch": 0.09295193211406061, "grad_norm": 2.598322868347168, "learning_rate": 4.994296405512938e-07, "loss": 0.5235, "step": 10335 }, { "epoch": 0.09296092601170106, "grad_norm": 2.8046388626098633, "learning_rate": 4.979359726901639e-07, "loss": 0.4803, "step": 10336 }, { "epoch": 0.09296991990934152, "grad_norm": 3.0379273891448975, "learning_rate": 4.964445192991913e-07, "loss": 0.4223, "step": 10337 }, { "epoch": 0.09297891380698196, "grad_norm": 2.581973075866699, "learning_rate": 4.949552805131679e-07, "loss": 0.2872, "step": 10338 }, { "epoch": 0.09298790770462241, "grad_norm": 2.0220870971679688, "learning_rate": 4.93468256466692e-07, "loss": 0.5571, "step": 10339 }, { "epoch": 0.09299690160226287, "grad_norm": 2.2140257358551025, "learning_rate": 4.919834472941614e-07, "loss": 0.4082, "step": 10340 }, { "epoch": 0.09300589549990332, "grad_norm": 1.9740941524505615, "learning_rate": 4.905008531297661e-07, "loss": 0.2919, "step": 10341 }, { "epoch": 0.09301488939754377, "grad_norm": 3.330406665802002, "learning_rate": 4.890204741075072e-07, "loss": 0.4542, "step": 10342 }, { "epoch": 0.09302388329518421, "grad_norm": 2.2707061767578125, "learning_rate": 4.87542310361172e-07, "loss": 0.5376, "step": 10343 }, { "epoch": 0.09303287719282467, "grad_norm": 1.7529922723770142, "learning_rate": 4.86066362024365e-07, "loss": 0.3141, "step": 10344 }, { "epoch": 0.09304187109046512, "grad_norm": 7.772271156311035, "learning_rate": 4.845926292304737e-07, "loss": 0.6972, "step": 10345 }, { "epoch": 0.09305086498810557, "grad_norm": 2.519763946533203, "learning_rate": 4.831211121126944e-07, "loss": 0.3027, "step": 10346 }, { "epoch": 0.09305985888574603, "grad_norm": 27.458757400512695, "learning_rate": 4.816518108040236e-07, "loss": 0.7674, "step": 10347 }, { "epoch": 0.09306885278338647, "grad_norm": 3.016329288482666, "learning_rate": 4.801847254372494e-07, "loss": 0.2592, "step": 10348 }, { "epoch": 0.09307784668102692, "grad_norm": 18.862138748168945, "learning_rate": 4.787198561449713e-07, "loss": 0.5309, "step": 10349 }, { "epoch": 0.09308684057866737, "grad_norm": 1.9176340103149414, "learning_rate": 4.772572030595779e-07, "loss": 0.2579, "step": 10350 }, { "epoch": 0.09309583447630783, "grad_norm": 19.649471282958984, "learning_rate": 4.75796766313269e-07, "loss": 0.3325, "step": 10351 }, { "epoch": 0.09310482837394828, "grad_norm": 4.31743860244751, "learning_rate": 4.743385460380334e-07, "loss": 0.222, "step": 10352 }, { "epoch": 0.09311382227158872, "grad_norm": 2.295985698699951, "learning_rate": 4.7288254236565745e-07, "loss": 0.5638, "step": 10353 }, { "epoch": 0.09312281616922918, "grad_norm": 2.839708089828491, "learning_rate": 4.7142875542774414e-07, "loss": 0.313, "step": 10354 }, { "epoch": 0.09313181006686963, "grad_norm": 13.429010391235352, "learning_rate": 4.6997718535568e-07, "loss": 0.3541, "step": 10355 }, { "epoch": 0.09314080396451008, "grad_norm": 2.182997941970825, "learning_rate": 4.685278322806547e-07, "loss": 0.5714, "step": 10356 }, { "epoch": 0.09314979786215052, "grad_norm": 2.268744945526123, "learning_rate": 4.6708069633366047e-07, "loss": 0.2533, "step": 10357 }, { "epoch": 0.09315879175979098, "grad_norm": 13.959742546081543, "learning_rate": 4.656357776454873e-07, "loss": 0.3392, "step": 10358 }, { "epoch": 0.09316778565743143, "grad_norm": 1.8343124389648438, "learning_rate": 4.641930763467278e-07, "loss": 0.4074, "step": 10359 }, { "epoch": 0.09317677955507188, "grad_norm": 14.588099479675293, "learning_rate": 4.627525925677695e-07, "loss": 0.5363, "step": 10360 }, { "epoch": 0.09318577345271234, "grad_norm": 24.66094970703125, "learning_rate": 4.6131432643880533e-07, "loss": 0.5436, "step": 10361 }, { "epoch": 0.09319476735035279, "grad_norm": 2.609555959701538, "learning_rate": 4.5987827808981753e-07, "loss": 0.2351, "step": 10362 }, { "epoch": 0.09320376124799323, "grad_norm": 3.075176954269409, "learning_rate": 4.5844444765059945e-07, "loss": 0.5941, "step": 10363 }, { "epoch": 0.0932127551456337, "grad_norm": 9.261300086975098, "learning_rate": 4.570128352507391e-07, "loss": 0.9264, "step": 10364 }, { "epoch": 0.09322174904327414, "grad_norm": 3.1649057865142822, "learning_rate": 4.5558344101962194e-07, "loss": 0.1946, "step": 10365 }, { "epoch": 0.09323074294091459, "grad_norm": 2.8618555068969727, "learning_rate": 4.5415626508643896e-07, "loss": 0.5985, "step": 10366 }, { "epoch": 0.09323973683855503, "grad_norm": 11.82270622253418, "learning_rate": 4.5273130758017033e-07, "loss": 0.7375, "step": 10367 }, { "epoch": 0.0932487307361955, "grad_norm": 14.93346881866455, "learning_rate": 4.5130856862960756e-07, "loss": 0.4208, "step": 10368 }, { "epoch": 0.09325772463383594, "grad_norm": 20.13231658935547, "learning_rate": 4.4988804836333666e-07, "loss": 0.5297, "step": 10369 }, { "epoch": 0.09326671853147639, "grad_norm": 10.893341064453125, "learning_rate": 4.48469746909741e-07, "loss": 0.585, "step": 10370 }, { "epoch": 0.09327571242911685, "grad_norm": 19.14225959777832, "learning_rate": 4.4705366439700416e-07, "loss": 0.8786, "step": 10371 }, { "epoch": 0.0932847063267573, "grad_norm": 1.7560381889343262, "learning_rate": 4.4563980095310707e-07, "loss": 0.4652, "step": 10372 }, { "epoch": 0.09329370022439774, "grad_norm": 2.2213921546936035, "learning_rate": 4.4422815670584207e-07, "loss": 0.1296, "step": 10373 }, { "epoch": 0.09330269412203819, "grad_norm": 2.015347719192505, "learning_rate": 4.4281873178278475e-07, "loss": 0.5669, "step": 10374 }, { "epoch": 0.09331168801967865, "grad_norm": 2.424002170562744, "learning_rate": 4.4141152631131945e-07, "loss": 0.2905, "step": 10375 }, { "epoch": 0.0933206819173191, "grad_norm": 2.8269853591918945, "learning_rate": 4.4000654041862764e-07, "loss": 0.3207, "step": 10376 }, { "epoch": 0.09332967581495955, "grad_norm": 3.2095348834991455, "learning_rate": 4.386037742316912e-07, "loss": 0.3178, "step": 10377 }, { "epoch": 0.0933386697126, "grad_norm": 1.8850553035736084, "learning_rate": 4.3720322787728915e-07, "loss": 0.2991, "step": 10378 }, { "epoch": 0.09334766361024045, "grad_norm": 21.606624603271484, "learning_rate": 4.358049014820037e-07, "loss": 0.5745, "step": 10379 }, { "epoch": 0.0933566575078809, "grad_norm": 15.10933780670166, "learning_rate": 4.3440879517221425e-07, "loss": 0.3382, "step": 10380 }, { "epoch": 0.09336565140552136, "grad_norm": 2.7980380058288574, "learning_rate": 4.330149090740976e-07, "loss": 0.3847, "step": 10381 }, { "epoch": 0.09337464530316181, "grad_norm": 2.386357307434082, "learning_rate": 4.316232433136308e-07, "loss": 0.1805, "step": 10382 }, { "epoch": 0.09338363920080225, "grad_norm": 2.0832221508026123, "learning_rate": 4.30233798016591e-07, "loss": 0.3493, "step": 10383 }, { "epoch": 0.0933926330984427, "grad_norm": 2.9610562324523926, "learning_rate": 4.288465733085612e-07, "loss": 0.2848, "step": 10384 }, { "epoch": 0.09340162699608316, "grad_norm": 2.3298230171203613, "learning_rate": 4.2746156931490754e-07, "loss": 0.5916, "step": 10385 }, { "epoch": 0.09341062089372361, "grad_norm": 14.416666030883789, "learning_rate": 4.2607878616081344e-07, "loss": 0.4051, "step": 10386 }, { "epoch": 0.09341961479136406, "grad_norm": 2.429854393005371, "learning_rate": 4.2469822397124546e-07, "loss": 0.4517, "step": 10387 }, { "epoch": 0.09342860868900452, "grad_norm": 2.5568785667419434, "learning_rate": 4.2331988287098445e-07, "loss": 0.4301, "step": 10388 }, { "epoch": 0.09343760258664496, "grad_norm": 2.092238664627075, "learning_rate": 4.2194376298460017e-07, "loss": 0.3534, "step": 10389 }, { "epoch": 0.09344659648428541, "grad_norm": 2.4109652042388916, "learning_rate": 4.205698644364653e-07, "loss": 0.4376, "step": 10390 }, { "epoch": 0.09345559038192586, "grad_norm": 2.2286272048950195, "learning_rate": 4.1919818735075003e-07, "loss": 0.073, "step": 10391 }, { "epoch": 0.09346458427956632, "grad_norm": 11.63200569152832, "learning_rate": 4.1782873185143014e-07, "loss": 0.7205, "step": 10392 }, { "epoch": 0.09347357817720676, "grad_norm": 2.312091588973999, "learning_rate": 4.1646149806226777e-07, "loss": 0.2936, "step": 10393 }, { "epoch": 0.09348257207484721, "grad_norm": 3.4547860622406006, "learning_rate": 4.150964861068363e-07, "loss": 0.5041, "step": 10394 }, { "epoch": 0.09349156597248767, "grad_norm": 2.293710947036743, "learning_rate": 4.1373369610850646e-07, "loss": 0.3653, "step": 10395 }, { "epoch": 0.09350055987012812, "grad_norm": 3.403292179107666, "learning_rate": 4.1237312819044085e-07, "loss": 0.2608, "step": 10396 }, { "epoch": 0.09350955376776857, "grad_norm": 3.3784539699554443, "learning_rate": 4.1101478247561054e-07, "loss": 0.7478, "step": 10397 }, { "epoch": 0.09351854766540901, "grad_norm": 15.849844932556152, "learning_rate": 4.0965865908677846e-07, "loss": 0.371, "step": 10398 }, { "epoch": 0.09352754156304947, "grad_norm": 3.172250270843506, "learning_rate": 4.083047581465077e-07, "loss": 0.6248, "step": 10399 }, { "epoch": 0.09353653546068992, "grad_norm": 9.11272144317627, "learning_rate": 4.06953079777167e-07, "loss": 0.6702, "step": 10400 }, { "epoch": 0.09354552935833037, "grad_norm": 2.5581092834472656, "learning_rate": 4.0560362410091704e-07, "loss": 0.3438, "step": 10401 }, { "epoch": 0.09355452325597083, "grad_norm": 2.6747167110443115, "learning_rate": 4.0425639123972126e-07, "loss": 0.7768, "step": 10402 }, { "epoch": 0.09356351715361128, "grad_norm": 2.0659372806549072, "learning_rate": 4.029113813153379e-07, "loss": 0.7108, "step": 10403 }, { "epoch": 0.09357251105125172, "grad_norm": 3.8868675231933594, "learning_rate": 4.015685944493308e-07, "loss": 0.6183, "step": 10404 }, { "epoch": 0.09358150494889218, "grad_norm": 12.51449203491211, "learning_rate": 4.0022803076306126e-07, "loss": 0.5485, "step": 10405 }, { "epoch": 0.09359049884653263, "grad_norm": 2.301680564880371, "learning_rate": 3.9888969037768234e-07, "loss": 0.4617, "step": 10406 }, { "epoch": 0.09359949274417308, "grad_norm": 29.1734676361084, "learning_rate": 3.9755357341415835e-07, "loss": 0.5411, "step": 10407 }, { "epoch": 0.09360848664181352, "grad_norm": 2.4221177101135254, "learning_rate": 3.962196799932399e-07, "loss": 0.2659, "step": 10408 }, { "epoch": 0.09361748053945398, "grad_norm": 3.227947473526001, "learning_rate": 3.94888010235489e-07, "loss": 0.3299, "step": 10409 }, { "epoch": 0.09362647443709443, "grad_norm": 3.4247450828552246, "learning_rate": 3.935585642612566e-07, "loss": 0.3078, "step": 10410 }, { "epoch": 0.09363546833473488, "grad_norm": 2.0130863189697266, "learning_rate": 3.9223134219069645e-07, "loss": 0.4308, "step": 10411 }, { "epoch": 0.09364446223237534, "grad_norm": 14.744961738586426, "learning_rate": 3.909063441437627e-07, "loss": 0.5417, "step": 10412 }, { "epoch": 0.09365345613001579, "grad_norm": 2.577850818634033, "learning_rate": 3.8958357024020674e-07, "loss": 0.392, "step": 10413 }, { "epoch": 0.09366245002765623, "grad_norm": 11.36117935180664, "learning_rate": 3.882630205995802e-07, "loss": 0.8015, "step": 10414 }, { "epoch": 0.09367144392529668, "grad_norm": 2.275054931640625, "learning_rate": 3.869446953412292e-07, "loss": 0.1828, "step": 10415 }, { "epoch": 0.09368043782293714, "grad_norm": 1.878919243812561, "learning_rate": 3.8562859458430855e-07, "loss": 0.2134, "step": 10416 }, { "epoch": 0.09368943172057759, "grad_norm": 18.972970962524414, "learning_rate": 3.843147184477619e-07, "loss": 0.3855, "step": 10417 }, { "epoch": 0.09369842561821803, "grad_norm": 2.532954216003418, "learning_rate": 3.83003067050336e-07, "loss": 0.4683, "step": 10418 }, { "epoch": 0.0937074195158585, "grad_norm": 20.32674217224121, "learning_rate": 3.8169364051058034e-07, "loss": 0.3835, "step": 10419 }, { "epoch": 0.09371641341349894, "grad_norm": 1.905173897743225, "learning_rate": 3.803864389468337e-07, "loss": 0.5107, "step": 10420 }, { "epoch": 0.09372540731113939, "grad_norm": 1.8025925159454346, "learning_rate": 3.79081462477246e-07, "loss": 0.1515, "step": 10421 }, { "epoch": 0.09373440120877985, "grad_norm": 1.9178473949432373, "learning_rate": 3.7777871121975615e-07, "loss": 0.1277, "step": 10422 }, { "epoch": 0.0937433951064203, "grad_norm": 3.124072551727295, "learning_rate": 3.764781852921034e-07, "loss": 0.3364, "step": 10423 }, { "epoch": 0.09375238900406074, "grad_norm": 15.522039413452148, "learning_rate": 3.7517988481182984e-07, "loss": 0.4558, "step": 10424 }, { "epoch": 0.09376138290170119, "grad_norm": 2.5500950813293457, "learning_rate": 3.738838098962777e-07, "loss": 0.3772, "step": 10425 }, { "epoch": 0.09377037679934165, "grad_norm": 2.7796638011932373, "learning_rate": 3.7258996066258103e-07, "loss": 0.3914, "step": 10426 }, { "epoch": 0.0937793706969821, "grad_norm": 3.165266275405884, "learning_rate": 3.7129833722767404e-07, "loss": 0.5178, "step": 10427 }, { "epoch": 0.09378836459462254, "grad_norm": 1.9257030487060547, "learning_rate": 3.7000893970829676e-07, "loss": 0.2404, "step": 10428 }, { "epoch": 0.093797358492263, "grad_norm": 2.367077112197876, "learning_rate": 3.687217682209837e-07, "loss": 0.6033, "step": 10429 }, { "epoch": 0.09380635238990345, "grad_norm": 3.267767906188965, "learning_rate": 3.6743682288206403e-07, "loss": 0.6631, "step": 10430 }, { "epoch": 0.0938153462875439, "grad_norm": 2.3766796588897705, "learning_rate": 3.6615410380767544e-07, "loss": 0.0633, "step": 10431 }, { "epoch": 0.09382434018518435, "grad_norm": 15.89832878112793, "learning_rate": 3.6487361111374184e-07, "loss": 0.5187, "step": 10432 }, { "epoch": 0.0938333340828248, "grad_norm": 4.190877437591553, "learning_rate": 3.6359534491600123e-07, "loss": 0.4027, "step": 10433 }, { "epoch": 0.09384232798046525, "grad_norm": 11.408858299255371, "learning_rate": 3.62319305329975e-07, "loss": 0.5981, "step": 10434 }, { "epoch": 0.0938513218781057, "grad_norm": 2.2233734130859375, "learning_rate": 3.610454924709905e-07, "loss": 0.4881, "step": 10435 }, { "epoch": 0.09386031577574616, "grad_norm": 2.7475693225860596, "learning_rate": 3.597739064541777e-07, "loss": 0.3692, "step": 10436 }, { "epoch": 0.09386930967338661, "grad_norm": 3.0189902782440186, "learning_rate": 3.5850454739445596e-07, "loss": 0.6442, "step": 10437 }, { "epoch": 0.09387830357102706, "grad_norm": 2.6792209148406982, "learning_rate": 3.572374154065555e-07, "loss": 0.2984, "step": 10438 }, { "epoch": 0.09388729746866752, "grad_norm": 2.5050699710845947, "learning_rate": 3.5597251060499326e-07, "loss": 0.505, "step": 10439 }, { "epoch": 0.09389629136630796, "grad_norm": 15.131013870239258, "learning_rate": 3.5470983310409157e-07, "loss": 0.8153, "step": 10440 }, { "epoch": 0.09390528526394841, "grad_norm": 1.9795278310775757, "learning_rate": 3.5344938301797024e-07, "loss": 0.3502, "step": 10441 }, { "epoch": 0.09391427916158886, "grad_norm": 3.834540843963623, "learning_rate": 3.521911604605438e-07, "loss": 0.4806, "step": 10442 }, { "epoch": 0.09392327305922932, "grad_norm": 3.442615270614624, "learning_rate": 3.509351655455351e-07, "loss": 0.3036, "step": 10443 }, { "epoch": 0.09393226695686976, "grad_norm": 2.6167993545532227, "learning_rate": 3.496813983864533e-07, "loss": 0.2052, "step": 10444 }, { "epoch": 0.09394126085451021, "grad_norm": 2.8638782501220703, "learning_rate": 3.484298590966162e-07, "loss": 0.4973, "step": 10445 }, { "epoch": 0.09395025475215067, "grad_norm": 6.020667552947998, "learning_rate": 3.47180547789136e-07, "loss": 0.5657, "step": 10446 }, { "epoch": 0.09395924864979112, "grad_norm": 2.9882900714874268, "learning_rate": 3.459334645769252e-07, "loss": 0.391, "step": 10447 }, { "epoch": 0.09396824254743157, "grad_norm": 11.519808769226074, "learning_rate": 3.4468860957269087e-07, "loss": 0.6809, "step": 10448 }, { "epoch": 0.09397723644507201, "grad_norm": 6.836554050445557, "learning_rate": 3.4344598288894303e-07, "loss": 1.0593, "step": 10449 }, { "epoch": 0.09398623034271247, "grad_norm": 2.272571325302124, "learning_rate": 3.422055846379918e-07, "loss": 0.6215, "step": 10450 }, { "epoch": 0.09399522424035292, "grad_norm": 10.952095985412598, "learning_rate": 3.4096741493194197e-07, "loss": 0.6643, "step": 10451 }, { "epoch": 0.09400421813799337, "grad_norm": 7.336275577545166, "learning_rate": 3.3973147388269566e-07, "loss": 0.7678, "step": 10452 }, { "epoch": 0.09401321203563383, "grad_norm": 2.474224328994751, "learning_rate": 3.3849776160195243e-07, "loss": 0.4723, "step": 10453 }, { "epoch": 0.09402220593327427, "grad_norm": 23.91658592224121, "learning_rate": 3.3726627820122316e-07, "loss": 0.3157, "step": 10454 }, { "epoch": 0.09403119983091472, "grad_norm": 2.778959035873413, "learning_rate": 3.3603702379180204e-07, "loss": 0.0694, "step": 10455 }, { "epoch": 0.09404019372855517, "grad_norm": 1.8805420398712158, "learning_rate": 3.3480999848478635e-07, "loss": 0.4475, "step": 10456 }, { "epoch": 0.09404918762619563, "grad_norm": 2.168646812438965, "learning_rate": 3.3358520239107905e-07, "loss": 0.338, "step": 10457 }, { "epoch": 0.09405818152383608, "grad_norm": 1.956579327583313, "learning_rate": 3.3236263562137217e-07, "loss": 0.3503, "step": 10458 }, { "epoch": 0.09406717542147652, "grad_norm": 21.349279403686523, "learning_rate": 3.311422982861606e-07, "loss": 0.4316, "step": 10459 }, { "epoch": 0.09407616931911698, "grad_norm": 2.957038164138794, "learning_rate": 3.2992419049573676e-07, "loss": 0.454, "step": 10460 }, { "epoch": 0.09408516321675743, "grad_norm": 5.971503257751465, "learning_rate": 3.287083123601903e-07, "loss": 0.8741, "step": 10461 }, { "epoch": 0.09409415711439788, "grad_norm": 2.930605888366699, "learning_rate": 3.2749466398941384e-07, "loss": 0.0694, "step": 10462 }, { "epoch": 0.09410315101203834, "grad_norm": 18.423349380493164, "learning_rate": 3.262832454930947e-07, "loss": 0.4253, "step": 10463 }, { "epoch": 0.09411214490967879, "grad_norm": 12.908514022827148, "learning_rate": 3.2507405698072035e-07, "loss": 0.3915, "step": 10464 }, { "epoch": 0.09412113880731923, "grad_norm": 1.8565877676010132, "learning_rate": 3.238670985615755e-07, "loss": 0.2579, "step": 10465 }, { "epoch": 0.09413013270495968, "grad_norm": 3.969632625579834, "learning_rate": 3.2266237034474233e-07, "loss": 0.2165, "step": 10466 }, { "epoch": 0.09413912660260014, "grad_norm": 1.9156267642974854, "learning_rate": 3.2145987243910324e-07, "loss": 0.3708, "step": 10467 }, { "epoch": 0.09414812050024059, "grad_norm": 3.272609233856201, "learning_rate": 3.2025960495334075e-07, "loss": 0.3021, "step": 10468 }, { "epoch": 0.09415711439788103, "grad_norm": 2.6238443851470947, "learning_rate": 3.190615679959319e-07, "loss": 0.4083, "step": 10469 }, { "epoch": 0.0941661082955215, "grad_norm": 2.156137466430664, "learning_rate": 3.1786576167515403e-07, "loss": 0.4104, "step": 10470 }, { "epoch": 0.09417510219316194, "grad_norm": 12.223329544067383, "learning_rate": 3.1667218609908176e-07, "loss": 1.1013, "step": 10471 }, { "epoch": 0.09418409609080239, "grad_norm": 3.5830657482147217, "learning_rate": 3.154808413755928e-07, "loss": 0.4108, "step": 10472 }, { "epoch": 0.09419308998844284, "grad_norm": 25.408367156982422, "learning_rate": 3.142917276123564e-07, "loss": 0.636, "step": 10473 }, { "epoch": 0.0942020838860833, "grad_norm": 3.055053949356079, "learning_rate": 3.13104844916845e-07, "loss": 0.5028, "step": 10474 }, { "epoch": 0.09421107778372374, "grad_norm": 1.7858617305755615, "learning_rate": 3.1192019339632837e-07, "loss": 0.5512, "step": 10475 }, { "epoch": 0.09422007168136419, "grad_norm": 1.8298656940460205, "learning_rate": 3.107377731578709e-07, "loss": 0.2677, "step": 10476 }, { "epoch": 0.09422906557900465, "grad_norm": 1.8114864826202393, "learning_rate": 3.095575843083426e-07, "loss": 0.3482, "step": 10477 }, { "epoch": 0.0942380594766451, "grad_norm": 1.9248201847076416, "learning_rate": 3.083796269544026e-07, "loss": 0.4042, "step": 10478 }, { "epoch": 0.09424705337428554, "grad_norm": 24.16933822631836, "learning_rate": 3.0720390120251574e-07, "loss": 0.8955, "step": 10479 }, { "epoch": 0.094256047271926, "grad_norm": 1.8093535900115967, "learning_rate": 3.0603040715894704e-07, "loss": 0.3538, "step": 10480 }, { "epoch": 0.09426504116956645, "grad_norm": 2.7680258750915527, "learning_rate": 3.0485914492975054e-07, "loss": 0.4823, "step": 10481 }, { "epoch": 0.0942740350672069, "grad_norm": 2.8527462482452393, "learning_rate": 3.03690114620786e-07, "loss": 0.4552, "step": 10482 }, { "epoch": 0.09428302896484735, "grad_norm": 2.7008371353149414, "learning_rate": 3.0252331633770504e-07, "loss": 0.5318, "step": 10483 }, { "epoch": 0.0942920228624878, "grad_norm": 3.2528035640716553, "learning_rate": 3.013587501859677e-07, "loss": 0.351, "step": 10484 }, { "epoch": 0.09430101676012825, "grad_norm": 3.864752769470215, "learning_rate": 3.001964162708204e-07, "loss": 0.3874, "step": 10485 }, { "epoch": 0.0943100106577687, "grad_norm": 19.199308395385742, "learning_rate": 2.9903631469732075e-07, "loss": 0.6748, "step": 10486 }, { "epoch": 0.09431900455540916, "grad_norm": 2.4742989540100098, "learning_rate": 2.9787844557030987e-07, "loss": 0.3093, "step": 10487 }, { "epoch": 0.09432799845304961, "grad_norm": 3.121731996536255, "learning_rate": 2.967228089944402e-07, "loss": 0.2594, "step": 10488 }, { "epoch": 0.09433699235069005, "grad_norm": 20.386369705200195, "learning_rate": 2.955694050741531e-07, "loss": 0.6306, "step": 10489 }, { "epoch": 0.0943459862483305, "grad_norm": 2.37109375, "learning_rate": 2.944182339136958e-07, "loss": 0.3686, "step": 10490 }, { "epoch": 0.09435498014597096, "grad_norm": 2.23917293548584, "learning_rate": 2.9326929561710735e-07, "loss": 0.1796, "step": 10491 }, { "epoch": 0.09436397404361141, "grad_norm": 1.8674226999282837, "learning_rate": 2.921225902882241e-07, "loss": 0.3798, "step": 10492 }, { "epoch": 0.09437296794125186, "grad_norm": 2.265721082687378, "learning_rate": 2.9097811803069365e-07, "loss": 0.611, "step": 10493 }, { "epoch": 0.09438196183889232, "grad_norm": 2.598008155822754, "learning_rate": 2.898358789479444e-07, "loss": 0.3494, "step": 10494 }, { "epoch": 0.09439095573653276, "grad_norm": 2.816516637802124, "learning_rate": 2.886958731432132e-07, "loss": 0.347, "step": 10495 }, { "epoch": 0.09439994963417321, "grad_norm": 1.6858153343200684, "learning_rate": 2.875581007195316e-07, "loss": 0.5195, "step": 10496 }, { "epoch": 0.09440894353181366, "grad_norm": 3.229382038116455, "learning_rate": 2.8642256177973117e-07, "loss": 0.4092, "step": 10497 }, { "epoch": 0.09441793742945412, "grad_norm": 2.753706693649292, "learning_rate": 2.852892564264409e-07, "loss": 0.5195, "step": 10498 }, { "epoch": 0.09442693132709457, "grad_norm": 2.453273057937622, "learning_rate": 2.8415818476208444e-07, "loss": 0.3316, "step": 10499 }, { "epoch": 0.09443592522473501, "grad_norm": 3.498121500015259, "learning_rate": 2.8302934688889117e-07, "loss": 0.1724, "step": 10500 }, { "epoch": 0.09444491912237547, "grad_norm": 14.43619441986084, "learning_rate": 2.819027429088822e-07, "loss": 0.8425, "step": 10501 }, { "epoch": 0.09445391302001592, "grad_norm": 20.504350662231445, "learning_rate": 2.807783729238789e-07, "loss": 0.6884, "step": 10502 }, { "epoch": 0.09446290691765637, "grad_norm": 2.5684773921966553, "learning_rate": 2.796562370355027e-07, "loss": 0.5983, "step": 10503 }, { "epoch": 0.09447190081529683, "grad_norm": 3.0511820316314697, "learning_rate": 2.785363353451698e-07, "loss": 0.2652, "step": 10504 }, { "epoch": 0.09448089471293727, "grad_norm": 2.634070873260498, "learning_rate": 2.7741866795409356e-07, "loss": 0.2476, "step": 10505 }, { "epoch": 0.09448988861057772, "grad_norm": 11.84008502960205, "learning_rate": 2.763032349632877e-07, "loss": 0.4941, "step": 10506 }, { "epoch": 0.09449888250821817, "grad_norm": 1.718653917312622, "learning_rate": 2.751900364735688e-07, "loss": 0.3596, "step": 10507 }, { "epoch": 0.09450787640585863, "grad_norm": 11.270024299621582, "learning_rate": 2.740790725855397e-07, "loss": 0.482, "step": 10508 }, { "epoch": 0.09451687030349908, "grad_norm": 2.499061346054077, "learning_rate": 2.729703433996145e-07, "loss": 0.0783, "step": 10509 }, { "epoch": 0.09452586420113952, "grad_norm": 16.662935256958008, "learning_rate": 2.718638490159936e-07, "loss": 0.8919, "step": 10510 }, { "epoch": 0.09453485809877998, "grad_norm": 16.43417739868164, "learning_rate": 2.7075958953468316e-07, "loss": 0.8627, "step": 10511 }, { "epoch": 0.09454385199642043, "grad_norm": 2.0277934074401855, "learning_rate": 2.696575650554839e-07, "loss": 0.4882, "step": 10512 }, { "epoch": 0.09455284589406088, "grad_norm": 2.3156919479370117, "learning_rate": 2.6855777567799947e-07, "loss": 0.4172, "step": 10513 }, { "epoch": 0.09456183979170132, "grad_norm": 10.994976997375488, "learning_rate": 2.6746022150162254e-07, "loss": 0.5188, "step": 10514 }, { "epoch": 0.09457083368934179, "grad_norm": 2.3671655654907227, "learning_rate": 2.6636490262555157e-07, "loss": 0.4568, "step": 10515 }, { "epoch": 0.09457982758698223, "grad_norm": 11.69628620147705, "learning_rate": 2.652718191487824e-07, "loss": 0.5953, "step": 10516 }, { "epoch": 0.09458882148462268, "grad_norm": 2.194939136505127, "learning_rate": 2.641809711700999e-07, "loss": 0.2638, "step": 10517 }, { "epoch": 0.09459781538226314, "grad_norm": 2.6469621658325195, "learning_rate": 2.6309235878810014e-07, "loss": 0.5959, "step": 10518 }, { "epoch": 0.09460680927990359, "grad_norm": 2.4778261184692383, "learning_rate": 2.6200598210116833e-07, "loss": 0.6128, "step": 10519 }, { "epoch": 0.09461580317754403, "grad_norm": 2.896702527999878, "learning_rate": 2.6092184120748987e-07, "loss": 0.3984, "step": 10520 }, { "epoch": 0.0946247970751845, "grad_norm": 2.8841986656188965, "learning_rate": 2.598399362050502e-07, "loss": 0.2494, "step": 10521 }, { "epoch": 0.09463379097282494, "grad_norm": 16.17866325378418, "learning_rate": 2.587602671916295e-07, "loss": 0.6207, "step": 10522 }, { "epoch": 0.09464278487046539, "grad_norm": 14.64075756072998, "learning_rate": 2.5768283426480524e-07, "loss": 0.4075, "step": 10523 }, { "epoch": 0.09465177876810583, "grad_norm": 9.05252456665039, "learning_rate": 2.5660763752195507e-07, "loss": 1.0342, "step": 10524 }, { "epoch": 0.0946607726657463, "grad_norm": 1.8914000988006592, "learning_rate": 2.555346770602568e-07, "loss": 0.448, "step": 10525 }, { "epoch": 0.09466976656338674, "grad_norm": 13.912352561950684, "learning_rate": 2.544639529766829e-07, "loss": 0.6152, "step": 10526 }, { "epoch": 0.09467876046102719, "grad_norm": 3.049351215362549, "learning_rate": 2.533954653680004e-07, "loss": 0.4681, "step": 10527 }, { "epoch": 0.09468775435866765, "grad_norm": 2.2733728885650635, "learning_rate": 2.5232921433078483e-07, "loss": 0.3355, "step": 10528 }, { "epoch": 0.0946967482563081, "grad_norm": 2.1625187397003174, "learning_rate": 2.512651999613952e-07, "loss": 0.3919, "step": 10529 }, { "epoch": 0.09470574215394854, "grad_norm": 11.765480041503906, "learning_rate": 2.502034223560018e-07, "loss": 0.3951, "step": 10530 }, { "epoch": 0.09471473605158899, "grad_norm": 21.293954849243164, "learning_rate": 2.491438816105612e-07, "loss": 0.4439, "step": 10531 }, { "epoch": 0.09472372994922945, "grad_norm": 3.1875500679016113, "learning_rate": 2.4808657782084133e-07, "loss": 0.4324, "step": 10532 }, { "epoch": 0.0947327238468699, "grad_norm": 2.703941822052002, "learning_rate": 2.470315110823934e-07, "loss": 0.7283, "step": 10533 }, { "epoch": 0.09474171774451035, "grad_norm": 2.21935772895813, "learning_rate": 2.4597868149057736e-07, "loss": 0.4183, "step": 10534 }, { "epoch": 0.0947507116421508, "grad_norm": 2.507561206817627, "learning_rate": 2.4492808914054475e-07, "loss": 0.3378, "step": 10535 }, { "epoch": 0.09475970553979125, "grad_norm": 2.119502544403076, "learning_rate": 2.438797341272503e-07, "loss": 0.1246, "step": 10536 }, { "epoch": 0.0947686994374317, "grad_norm": 14.739083290100098, "learning_rate": 2.4283361654543755e-07, "loss": 0.6326, "step": 10537 }, { "epoch": 0.09477769333507215, "grad_norm": 2.4347622394561768, "learning_rate": 2.41789736489656e-07, "loss": 0.4492, "step": 10538 }, { "epoch": 0.09478668723271261, "grad_norm": 2.154097318649292, "learning_rate": 2.4074809405425225e-07, "loss": 0.4948, "step": 10539 }, { "epoch": 0.09479568113035305, "grad_norm": 3.0490543842315674, "learning_rate": 2.39708689333365e-07, "loss": 0.4422, "step": 10540 }, { "epoch": 0.0948046750279935, "grad_norm": 2.533897876739502, "learning_rate": 2.386715224209385e-07, "loss": 0.3211, "step": 10541 }, { "epoch": 0.09481366892563396, "grad_norm": 20.352174758911133, "learning_rate": 2.3763659341070886e-07, "loss": 0.9583, "step": 10542 }, { "epoch": 0.09482266282327441, "grad_norm": 14.02239990234375, "learning_rate": 2.3660390239620967e-07, "loss": 0.4498, "step": 10543 }, { "epoch": 0.09483165672091486, "grad_norm": 14.384698867797852, "learning_rate": 2.355734494707801e-07, "loss": 0.4247, "step": 10544 }, { "epoch": 0.09484065061855532, "grad_norm": 2.741610050201416, "learning_rate": 2.3454523472754562e-07, "loss": 0.562, "step": 10545 }, { "epoch": 0.09484964451619576, "grad_norm": 14.237570762634277, "learning_rate": 2.335192582594403e-07, "loss": 0.3427, "step": 10546 }, { "epoch": 0.09485863841383621, "grad_norm": 12.298928260803223, "learning_rate": 2.324955201591872e-07, "loss": 0.423, "step": 10547 }, { "epoch": 0.09486763231147666, "grad_norm": 3.2762880325317383, "learning_rate": 2.3147402051930943e-07, "loss": 0.2425, "step": 10548 }, { "epoch": 0.09487662620911712, "grad_norm": 3.2330830097198486, "learning_rate": 2.3045475943213602e-07, "loss": 0.1767, "step": 10549 }, { "epoch": 0.09488562010675756, "grad_norm": 2.0093069076538086, "learning_rate": 2.2943773698977932e-07, "loss": 0.1664, "step": 10550 }, { "epoch": 0.09489461400439801, "grad_norm": 2.170091152191162, "learning_rate": 2.284229532841603e-07, "loss": 0.6545, "step": 10551 }, { "epoch": 0.09490360790203847, "grad_norm": 8.272106170654297, "learning_rate": 2.2741040840699168e-07, "loss": 0.7468, "step": 10552 }, { "epoch": 0.09491260179967892, "grad_norm": 2.752206802368164, "learning_rate": 2.2640010244978637e-07, "loss": 0.1841, "step": 10553 }, { "epoch": 0.09492159569731937, "grad_norm": 7.12520170211792, "learning_rate": 2.2539203550385745e-07, "loss": 0.8446, "step": 10554 }, { "epoch": 0.09493058959495981, "grad_norm": 2.5365750789642334, "learning_rate": 2.2438620766030982e-07, "loss": 0.5665, "step": 10555 }, { "epoch": 0.09493958349260027, "grad_norm": 12.528582572937012, "learning_rate": 2.2338261901005132e-07, "loss": 0.7127, "step": 10556 }, { "epoch": 0.09494857739024072, "grad_norm": 11.00051498413086, "learning_rate": 2.2238126964378436e-07, "loss": 0.6393, "step": 10557 }, { "epoch": 0.09495757128788117, "grad_norm": 6.4314045906066895, "learning_rate": 2.213821596520116e-07, "loss": 0.553, "step": 10558 }, { "epoch": 0.09496656518552163, "grad_norm": 2.64270281791687, "learning_rate": 2.203852891250302e-07, "loss": 0.4131, "step": 10559 }, { "epoch": 0.09497555908316208, "grad_norm": 2.1872501373291016, "learning_rate": 2.193906581529348e-07, "loss": 0.5488, "step": 10560 }, { "epoch": 0.09498455298080252, "grad_norm": 2.738856315612793, "learning_rate": 2.1839826682562015e-07, "loss": 0.2079, "step": 10561 }, { "epoch": 0.09499354687844298, "grad_norm": 15.599671363830566, "learning_rate": 2.174081152327756e-07, "loss": 0.4145, "step": 10562 }, { "epoch": 0.09500254077608343, "grad_norm": 3.164151430130005, "learning_rate": 2.1642020346389625e-07, "loss": 0.2413, "step": 10563 }, { "epoch": 0.09501153467372388, "grad_norm": 9.639272689819336, "learning_rate": 2.1543453160826065e-07, "loss": 0.6972, "step": 10564 }, { "epoch": 0.09502052857136432, "grad_norm": 2.7117271423339844, "learning_rate": 2.144510997549587e-07, "loss": 0.4383, "step": 10565 }, { "epoch": 0.09502952246900478, "grad_norm": 1.8794254064559937, "learning_rate": 2.1346990799286926e-07, "loss": 0.3674, "step": 10566 }, { "epoch": 0.09503851636664523, "grad_norm": 10.935770988464355, "learning_rate": 2.124909564106714e-07, "loss": 0.8024, "step": 10567 }, { "epoch": 0.09504751026428568, "grad_norm": 2.415680408477783, "learning_rate": 2.1151424509684158e-07, "loss": 0.2992, "step": 10568 }, { "epoch": 0.09505650416192614, "grad_norm": 15.58050537109375, "learning_rate": 2.1053977413965364e-07, "loss": 0.6015, "step": 10569 }, { "epoch": 0.09506549805956659, "grad_norm": 12.793838500976562, "learning_rate": 2.0956754362718157e-07, "loss": 0.7581, "step": 10570 }, { "epoch": 0.09507449195720703, "grad_norm": 2.5143299102783203, "learning_rate": 2.0859755364729116e-07, "loss": 0.5761, "step": 10571 }, { "epoch": 0.09508348585484748, "grad_norm": 13.129387855529785, "learning_rate": 2.0762980428765122e-07, "loss": 0.6631, "step": 10572 }, { "epoch": 0.09509247975248794, "grad_norm": 2.261322021484375, "learning_rate": 2.0666429563572508e-07, "loss": 0.2916, "step": 10573 }, { "epoch": 0.09510147365012839, "grad_norm": 6.812705039978027, "learning_rate": 2.0570102777877632e-07, "loss": 0.7599, "step": 10574 }, { "epoch": 0.09511046754776883, "grad_norm": 19.92143440246582, "learning_rate": 2.0474000080386025e-07, "loss": 0.8265, "step": 10575 }, { "epoch": 0.0951194614454093, "grad_norm": 7.285543441772461, "learning_rate": 2.0378121479783796e-07, "loss": 0.8021, "step": 10576 }, { "epoch": 0.09512845534304974, "grad_norm": 2.085836410522461, "learning_rate": 2.028246698473596e-07, "loss": 0.1454, "step": 10577 }, { "epoch": 0.09513744924069019, "grad_norm": 3.6350855827331543, "learning_rate": 2.0187036603888098e-07, "loss": 0.3368, "step": 10578 }, { "epoch": 0.09514644313833064, "grad_norm": 7.538859844207764, "learning_rate": 2.0091830345864427e-07, "loss": 0.6518, "step": 10579 }, { "epoch": 0.0951554370359711, "grad_norm": 1.7916265726089478, "learning_rate": 1.999684821927028e-07, "loss": 0.6454, "step": 10580 }, { "epoch": 0.09516443093361154, "grad_norm": 2.6098132133483887, "learning_rate": 1.990209023268963e-07, "loss": 0.6022, "step": 10581 }, { "epoch": 0.09517342483125199, "grad_norm": 2.2706542015075684, "learning_rate": 1.9807556394686733e-07, "loss": 0.2851, "step": 10582 }, { "epoch": 0.09518241872889245, "grad_norm": 1.9561465978622437, "learning_rate": 1.9713246713805588e-07, "loss": 0.0376, "step": 10583 }, { "epoch": 0.0951914126265329, "grad_norm": 3.362926721572876, "learning_rate": 1.961916119856938e-07, "loss": 0.9092, "step": 10584 }, { "epoch": 0.09520040652417334, "grad_norm": 15.485942840576172, "learning_rate": 1.952529985748186e-07, "loss": 0.7748, "step": 10585 }, { "epoch": 0.0952094004218138, "grad_norm": 13.6810941696167, "learning_rate": 1.9431662699025965e-07, "loss": 0.3643, "step": 10586 }, { "epoch": 0.09521839431945425, "grad_norm": 2.7734146118164062, "learning_rate": 1.933824973166465e-07, "loss": 0.1993, "step": 10587 }, { "epoch": 0.0952273882170947, "grad_norm": 3.3004698753356934, "learning_rate": 1.9245060963840322e-07, "loss": 0.6996, "step": 10588 }, { "epoch": 0.09523638211473515, "grad_norm": 2.29561448097229, "learning_rate": 1.9152096403975418e-07, "loss": 0.5475, "step": 10589 }, { "epoch": 0.09524537601237561, "grad_norm": 2.821763515472412, "learning_rate": 1.9059356060471822e-07, "loss": 0.2975, "step": 10590 }, { "epoch": 0.09525436991001605, "grad_norm": 2.862636089324951, "learning_rate": 1.8966839941711446e-07, "loss": 0.4534, "step": 10591 }, { "epoch": 0.0952633638076565, "grad_norm": 2.483391284942627, "learning_rate": 1.887454805605565e-07, "loss": 0.1034, "step": 10592 }, { "epoch": 0.09527235770529696, "grad_norm": 16.076013565063477, "learning_rate": 1.8782480411845825e-07, "loss": 1.0279, "step": 10593 }, { "epoch": 0.09528135160293741, "grad_norm": 2.5077693462371826, "learning_rate": 1.8690637017402534e-07, "loss": 0.3059, "step": 10594 }, { "epoch": 0.09529034550057786, "grad_norm": 2.5156197547912598, "learning_rate": 1.8599017881027193e-07, "loss": 0.8094, "step": 10595 }, { "epoch": 0.0952993393982183, "grad_norm": 2.2158994674682617, "learning_rate": 1.8507623010999842e-07, "loss": 0.5414, "step": 10596 }, { "epoch": 0.09530833329585876, "grad_norm": 2.791837453842163, "learning_rate": 1.8416452415580544e-07, "loss": 0.4808, "step": 10597 }, { "epoch": 0.09531732719349921, "grad_norm": 1.9761961698532104, "learning_rate": 1.8325506103009372e-07, "loss": 0.4755, "step": 10598 }, { "epoch": 0.09532632109113966, "grad_norm": 2.272589683532715, "learning_rate": 1.8234784081505863e-07, "loss": 0.5778, "step": 10599 }, { "epoch": 0.09533531498878012, "grad_norm": 3.1003611087799072, "learning_rate": 1.8144286359269569e-07, "loss": 0.2676, "step": 10600 }, { "epoch": 0.09534430888642056, "grad_norm": 2.7594759464263916, "learning_rate": 1.8054012944479227e-07, "loss": 0.3575, "step": 10601 }, { "epoch": 0.09535330278406101, "grad_norm": 7.21886682510376, "learning_rate": 1.7963963845294142e-07, "loss": 0.5811, "step": 10602 }, { "epoch": 0.09536229668170147, "grad_norm": 2.138308048248291, "learning_rate": 1.7874139069852524e-07, "loss": 0.5332, "step": 10603 }, { "epoch": 0.09537129057934192, "grad_norm": 8.280406951904297, "learning_rate": 1.7784538626272607e-07, "loss": 0.5131, "step": 10604 }, { "epoch": 0.09538028447698237, "grad_norm": 2.215630054473877, "learning_rate": 1.7695162522652353e-07, "loss": 0.3341, "step": 10605 }, { "epoch": 0.09538927837462281, "grad_norm": 1.909396767616272, "learning_rate": 1.7606010767070024e-07, "loss": 0.3742, "step": 10606 }, { "epoch": 0.09539827227226327, "grad_norm": 8.931285858154297, "learning_rate": 1.7517083367582231e-07, "loss": 0.4996, "step": 10607 }, { "epoch": 0.09540726616990372, "grad_norm": 8.633070945739746, "learning_rate": 1.7428380332226435e-07, "loss": 0.6861, "step": 10608 }, { "epoch": 0.09541626006754417, "grad_norm": 12.243744850158691, "learning_rate": 1.7339901669019832e-07, "loss": 0.5891, "step": 10609 }, { "epoch": 0.09542525396518463, "grad_norm": 3.3821911811828613, "learning_rate": 1.7251647385958802e-07, "loss": 0.4696, "step": 10610 }, { "epoch": 0.09543424786282508, "grad_norm": 2.5483860969543457, "learning_rate": 1.7163617491019745e-07, "loss": 0.7523, "step": 10611 }, { "epoch": 0.09544324176046552, "grad_norm": 10.721990585327148, "learning_rate": 1.707581199215852e-07, "loss": 0.3973, "step": 10612 }, { "epoch": 0.09545223565810597, "grad_norm": 2.5953288078308105, "learning_rate": 1.6988230897310998e-07, "loss": 0.4228, "step": 10613 }, { "epoch": 0.09546122955574643, "grad_norm": 11.558123588562012, "learning_rate": 1.690087421439279e-07, "loss": 0.6201, "step": 10614 }, { "epoch": 0.09547022345338688, "grad_norm": 21.086566925048828, "learning_rate": 1.6813741951298978e-07, "loss": 0.8522, "step": 10615 }, { "epoch": 0.09547921735102732, "grad_norm": 2.455188751220703, "learning_rate": 1.6726834115904643e-07, "loss": 0.3667, "step": 10616 }, { "epoch": 0.09548821124866778, "grad_norm": 2.0811948776245117, "learning_rate": 1.6640150716064061e-07, "loss": 0.2695, "step": 10617 }, { "epoch": 0.09549720514630823, "grad_norm": 1.9941809177398682, "learning_rate": 1.65536917596118e-07, "loss": 0.1442, "step": 10618 }, { "epoch": 0.09550619904394868, "grad_norm": 2.193758964538574, "learning_rate": 1.646745725436216e-07, "loss": 0.1485, "step": 10619 }, { "epoch": 0.09551519294158912, "grad_norm": 2.6366920471191406, "learning_rate": 1.6381447208108358e-07, "loss": 0.4698, "step": 10620 }, { "epoch": 0.09552418683922959, "grad_norm": 2.8180079460144043, "learning_rate": 1.6295661628624447e-07, "loss": 0.3321, "step": 10621 }, { "epoch": 0.09553318073687003, "grad_norm": 2.4493541717529297, "learning_rate": 1.6210100523663397e-07, "loss": 0.2622, "step": 10622 }, { "epoch": 0.09554217463451048, "grad_norm": 2.871319055557251, "learning_rate": 1.6124763900958185e-07, "loss": 0.2685, "step": 10623 }, { "epoch": 0.09555116853215094, "grad_norm": 1.7282682657241821, "learning_rate": 1.6039651768221254e-07, "loss": 0.4038, "step": 10624 }, { "epoch": 0.09556016242979139, "grad_norm": 3.996229648590088, "learning_rate": 1.5954764133145062e-07, "loss": 0.215, "step": 10625 }, { "epoch": 0.09556915632743183, "grad_norm": 2.466295003890991, "learning_rate": 1.5870101003402084e-07, "loss": 0.7002, "step": 10626 }, { "epoch": 0.0955781502250723, "grad_norm": 2.260171413421631, "learning_rate": 1.578566238664314e-07, "loss": 0.2583, "step": 10627 }, { "epoch": 0.09558714412271274, "grad_norm": 2.8144121170043945, "learning_rate": 1.570144829050074e-07, "loss": 0.4339, "step": 10628 }, { "epoch": 0.09559613802035319, "grad_norm": 2.4209537506103516, "learning_rate": 1.5617458722585188e-07, "loss": 0.3736, "step": 10629 }, { "epoch": 0.09560513191799364, "grad_norm": 2.279158353805542, "learning_rate": 1.5533693690487904e-07, "loss": 0.384, "step": 10630 }, { "epoch": 0.0956141258156341, "grad_norm": 1.869589924812317, "learning_rate": 1.5450153201779504e-07, "loss": 0.2918, "step": 10631 }, { "epoch": 0.09562311971327454, "grad_norm": 3.0647315979003906, "learning_rate": 1.536683726401006e-07, "loss": 0.2671, "step": 10632 }, { "epoch": 0.09563211361091499, "grad_norm": 2.3422658443450928, "learning_rate": 1.5283745884709378e-07, "loss": 0.1407, "step": 10633 }, { "epoch": 0.09564110750855545, "grad_norm": 2.661769151687622, "learning_rate": 1.5200879071387564e-07, "loss": 0.3079, "step": 10634 }, { "epoch": 0.0956501014061959, "grad_norm": 2.525822877883911, "learning_rate": 1.5118236831534184e-07, "loss": 0.5811, "step": 10635 }, { "epoch": 0.09565909530383634, "grad_norm": 13.025613784790039, "learning_rate": 1.5035819172617705e-07, "loss": 0.6072, "step": 10636 }, { "epoch": 0.09566808920147679, "grad_norm": 8.168732643127441, "learning_rate": 1.4953626102087448e-07, "loss": 0.6485, "step": 10637 }, { "epoch": 0.09567708309911725, "grad_norm": 12.203253746032715, "learning_rate": 1.4871657627371916e-07, "loss": 1.5353, "step": 10638 }, { "epoch": 0.0956860769967577, "grad_norm": 3.323199510574341, "learning_rate": 1.4789913755878791e-07, "loss": 0.4307, "step": 10639 }, { "epoch": 0.09569507089439815, "grad_norm": 1.8338052034378052, "learning_rate": 1.4708394494996892e-07, "loss": 0.4575, "step": 10640 }, { "epoch": 0.0957040647920386, "grad_norm": 1.4585474729537964, "learning_rate": 1.4627099852093095e-07, "loss": 0.5045, "step": 10641 }, { "epoch": 0.09571305868967905, "grad_norm": 2.9087066650390625, "learning_rate": 1.4546029834514862e-07, "loss": 0.2518, "step": 10642 }, { "epoch": 0.0957220525873195, "grad_norm": 15.8721284866333, "learning_rate": 1.4465184449589664e-07, "loss": 0.6523, "step": 10643 }, { "epoch": 0.09573104648495996, "grad_norm": 2.4777159690856934, "learning_rate": 1.43845637046236e-07, "loss": 0.449, "step": 10644 }, { "epoch": 0.09574004038260041, "grad_norm": 2.046990394592285, "learning_rate": 1.4304167606903618e-07, "loss": 0.1836, "step": 10645 }, { "epoch": 0.09574903428024086, "grad_norm": 2.5263805389404297, "learning_rate": 1.4223996163695575e-07, "loss": 0.6776, "step": 10646 }, { "epoch": 0.0957580281778813, "grad_norm": 36.34952163696289, "learning_rate": 1.414404938224534e-07, "loss": 0.4661, "step": 10647 }, { "epoch": 0.09576702207552176, "grad_norm": 2.102116584777832, "learning_rate": 1.4064327269778244e-07, "loss": 0.3253, "step": 10648 }, { "epoch": 0.09577601597316221, "grad_norm": 2.430694818496704, "learning_rate": 1.3984829833499636e-07, "loss": 0.6312, "step": 10649 }, { "epoch": 0.09578500987080266, "grad_norm": 2.4803006649017334, "learning_rate": 1.39055570805946e-07, "loss": 0.0744, "step": 10650 }, { "epoch": 0.09579400376844312, "grad_norm": 3.1278135776519775, "learning_rate": 1.3826509018227128e-07, "loss": 1.078, "step": 10651 }, { "epoch": 0.09580299766608356, "grad_norm": 2.6210579872131348, "learning_rate": 1.3747685653542063e-07, "loss": 0.5145, "step": 10652 }, { "epoch": 0.09581199156372401, "grad_norm": 2.5668087005615234, "learning_rate": 1.3669086993663428e-07, "loss": 0.5411, "step": 10653 }, { "epoch": 0.09582098546136446, "grad_norm": 3.106771230697632, "learning_rate": 1.3590713045694425e-07, "loss": 0.6903, "step": 10654 }, { "epoch": 0.09582997935900492, "grad_norm": 2.069420576095581, "learning_rate": 1.351256381671856e-07, "loss": 0.609, "step": 10655 }, { "epoch": 0.09583897325664537, "grad_norm": 1.9462146759033203, "learning_rate": 1.343463931379907e-07, "loss": 0.4445, "step": 10656 }, { "epoch": 0.09584796715428581, "grad_norm": 2.0927348136901855, "learning_rate": 1.3356939543978652e-07, "loss": 0.6409, "step": 10657 }, { "epoch": 0.09585696105192627, "grad_norm": 2.0565452575683594, "learning_rate": 1.3279464514279472e-07, "loss": 0.2637, "step": 10658 }, { "epoch": 0.09586595494956672, "grad_norm": 2.35554575920105, "learning_rate": 1.320221423170398e-07, "loss": 0.5189, "step": 10659 }, { "epoch": 0.09587494884720717, "grad_norm": 11.95732307434082, "learning_rate": 1.3125188703233816e-07, "loss": 0.5965, "step": 10660 }, { "epoch": 0.09588394274484763, "grad_norm": 14.990621566772461, "learning_rate": 1.3048387935830352e-07, "loss": 0.4062, "step": 10661 }, { "epoch": 0.09589293664248807, "grad_norm": 2.446024179458618, "learning_rate": 1.2971811936434985e-07, "loss": 0.2486, "step": 10662 }, { "epoch": 0.09590193054012852, "grad_norm": 2.2286689281463623, "learning_rate": 1.289546071196801e-07, "loss": 0.4218, "step": 10663 }, { "epoch": 0.09591092443776897, "grad_norm": 3.71431303024292, "learning_rate": 1.2819334269330852e-07, "loss": 0.268, "step": 10664 }, { "epoch": 0.09591991833540943, "grad_norm": 2.9671640396118164, "learning_rate": 1.2743432615403283e-07, "loss": 0.5421, "step": 10665 }, { "epoch": 0.09592891223304988, "grad_norm": 2.188525676727295, "learning_rate": 1.2667755757045096e-07, "loss": 1.0246, "step": 10666 }, { "epoch": 0.09593790613069032, "grad_norm": 3.368112802505493, "learning_rate": 1.25923037010961e-07, "loss": 0.5058, "step": 10667 }, { "epoch": 0.09594690002833078, "grad_norm": 20.810365676879883, "learning_rate": 1.2517076454375277e-07, "loss": 0.4639, "step": 10668 }, { "epoch": 0.09595589392597123, "grad_norm": 11.454456329345703, "learning_rate": 1.2442074023681915e-07, "loss": 0.5611, "step": 10669 }, { "epoch": 0.09596488782361168, "grad_norm": 2.9250049591064453, "learning_rate": 1.2367296415794483e-07, "loss": 0.2478, "step": 10670 }, { "epoch": 0.09597388172125212, "grad_norm": 2.9752767086029053, "learning_rate": 1.229274363747146e-07, "loss": 0.3442, "step": 10671 }, { "epoch": 0.09598287561889259, "grad_norm": 15.313682556152344, "learning_rate": 1.2218415695450792e-07, "loss": 0.8897, "step": 10672 }, { "epoch": 0.09599186951653303, "grad_norm": 2.4862618446350098, "learning_rate": 1.214431259644988e-07, "loss": 0.6581, "step": 10673 }, { "epoch": 0.09600086341417348, "grad_norm": 2.5357112884521484, "learning_rate": 1.2070434347166426e-07, "loss": 0.3406, "step": 10674 }, { "epoch": 0.09600985731181394, "grad_norm": 3.919102191925049, "learning_rate": 1.199678095427731e-07, "loss": 0.6387, "step": 10675 }, { "epoch": 0.09601885120945439, "grad_norm": 2.645371198654175, "learning_rate": 1.192335242443915e-07, "loss": 0.4809, "step": 10676 }, { "epoch": 0.09602784510709483, "grad_norm": 17.5118350982666, "learning_rate": 1.185014876428886e-07, "loss": 0.5209, "step": 10677 }, { "epoch": 0.09603683900473528, "grad_norm": 2.0508921146392822, "learning_rate": 1.1777169980441705e-07, "loss": 0.4139, "step": 10678 }, { "epoch": 0.09604583290237574, "grad_norm": 2.7275092601776123, "learning_rate": 1.1704416079494074e-07, "loss": 0.4009, "step": 10679 }, { "epoch": 0.09605482680001619, "grad_norm": 22.64410972595215, "learning_rate": 1.1631887068020708e-07, "loss": 0.4715, "step": 10680 }, { "epoch": 0.09606382069765664, "grad_norm": 2.626894950866699, "learning_rate": 1.1559582952577752e-07, "loss": 0.489, "step": 10681 }, { "epoch": 0.0960728145952971, "grad_norm": 1.818824291229248, "learning_rate": 1.1487503739698869e-07, "loss": 0.4295, "step": 10682 }, { "epoch": 0.09608180849293754, "grad_norm": 3.547098159790039, "learning_rate": 1.1415649435899122e-07, "loss": 0.2518, "step": 10683 }, { "epoch": 0.09609080239057799, "grad_norm": 6.753252029418945, "learning_rate": 1.1344020047672765e-07, "loss": 0.7777, "step": 10684 }, { "epoch": 0.09609979628821845, "grad_norm": 2.4575366973876953, "learning_rate": 1.1272615581492952e-07, "loss": 0.4765, "step": 10685 }, { "epoch": 0.0961087901858589, "grad_norm": 1.9529733657836914, "learning_rate": 1.1201436043813684e-07, "loss": 0.2706, "step": 10686 }, { "epoch": 0.09611778408349934, "grad_norm": 2.2294857501983643, "learning_rate": 1.1130481441068152e-07, "loss": 0.5706, "step": 10687 }, { "epoch": 0.09612677798113979, "grad_norm": 2.6754467487335205, "learning_rate": 1.1059751779668448e-07, "loss": 0.5062, "step": 10688 }, { "epoch": 0.09613577187878025, "grad_norm": 16.172088623046875, "learning_rate": 1.0989247066007791e-07, "loss": 0.5357, "step": 10689 }, { "epoch": 0.0961447657764207, "grad_norm": 23.281824111938477, "learning_rate": 1.0918967306458027e-07, "loss": 0.6309, "step": 10690 }, { "epoch": 0.09615375967406115, "grad_norm": 2.733081579208374, "learning_rate": 1.0848912507371023e-07, "loss": 0.1592, "step": 10691 }, { "epoch": 0.0961627535717016, "grad_norm": 2.8209667205810547, "learning_rate": 1.07790826750781e-07, "loss": 0.4362, "step": 10692 }, { "epoch": 0.09617174746934205, "grad_norm": 7.030689239501953, "learning_rate": 1.0709477815890601e-07, "loss": 0.76, "step": 10693 }, { "epoch": 0.0961807413669825, "grad_norm": 1.946895956993103, "learning_rate": 1.0640097936099048e-07, "loss": 0.3867, "step": 10694 }, { "epoch": 0.09618973526462295, "grad_norm": 2.177426338195801, "learning_rate": 1.0570943041974257e-07, "loss": 0.3468, "step": 10695 }, { "epoch": 0.09619872916226341, "grad_norm": 20.4360294342041, "learning_rate": 1.0502013139766509e-07, "loss": 0.7577, "step": 10696 }, { "epoch": 0.09620772305990385, "grad_norm": 2.526332378387451, "learning_rate": 1.0433308235704986e-07, "loss": 0.1968, "step": 10697 }, { "epoch": 0.0962167169575443, "grad_norm": 10.993362426757812, "learning_rate": 1.036482833599972e-07, "loss": 0.8273, "step": 10698 }, { "epoch": 0.09622571085518476, "grad_norm": 2.792327404022217, "learning_rate": 1.0296573446839375e-07, "loss": 0.4371, "step": 10699 }, { "epoch": 0.09623470475282521, "grad_norm": 10.51991081237793, "learning_rate": 1.0228543574393179e-07, "loss": 0.8695, "step": 10700 }, { "epoch": 0.09624369865046566, "grad_norm": 23.083053588867188, "learning_rate": 1.0160738724809549e-07, "loss": 0.5313, "step": 10701 }, { "epoch": 0.09625269254810612, "grad_norm": 3.412219285964966, "learning_rate": 1.0093158904216637e-07, "loss": 0.366, "step": 10702 }, { "epoch": 0.09626168644574656, "grad_norm": 2.2473535537719727, "learning_rate": 1.0025804118722059e-07, "loss": 0.3824, "step": 10703 }, { "epoch": 0.09627068034338701, "grad_norm": 2.8686728477478027, "learning_rate": 9.958674374413168e-08, "loss": 0.4946, "step": 10704 }, { "epoch": 0.09627967424102746, "grad_norm": 2.386561632156372, "learning_rate": 9.891769677357332e-08, "loss": 0.6787, "step": 10705 }, { "epoch": 0.09628866813866792, "grad_norm": 2.9584569931030273, "learning_rate": 9.825090033601103e-08, "loss": 0.451, "step": 10706 }, { "epoch": 0.09629766203630837, "grad_norm": 2.2147140502929688, "learning_rate": 9.75863544917105e-08, "loss": 0.1776, "step": 10707 }, { "epoch": 0.09630665593394881, "grad_norm": 3.072395086288452, "learning_rate": 9.692405930073478e-08, "loss": 0.1436, "step": 10708 }, { "epoch": 0.09631564983158927, "grad_norm": 2.459864616394043, "learning_rate": 9.626401482293602e-08, "loss": 0.478, "step": 10709 }, { "epoch": 0.09632464372922972, "grad_norm": 1.9787861108779907, "learning_rate": 9.56062211179748e-08, "loss": 0.6067, "step": 10710 }, { "epoch": 0.09633363762687017, "grad_norm": 2.0995237827301025, "learning_rate": 9.495067824529525e-08, "loss": 0.8114, "step": 10711 }, { "epoch": 0.09634263152451061, "grad_norm": 11.613673210144043, "learning_rate": 9.429738626414996e-08, "loss": 0.7142, "step": 10712 }, { "epoch": 0.09635162542215107, "grad_norm": 2.791163921356201, "learning_rate": 9.364634523358062e-08, "loss": 0.9294, "step": 10713 }, { "epoch": 0.09636061931979152, "grad_norm": 2.604776620864868, "learning_rate": 9.299755521242903e-08, "loss": 0.3996, "step": 10714 }, { "epoch": 0.09636961321743197, "grad_norm": 10.333154678344727, "learning_rate": 9.235101625932885e-08, "loss": 0.5743, "step": 10715 }, { "epoch": 0.09637860711507243, "grad_norm": 12.15869140625, "learning_rate": 9.170672843271666e-08, "loss": 0.6035, "step": 10716 }, { "epoch": 0.09638760101271288, "grad_norm": 2.0060536861419678, "learning_rate": 9.106469179082088e-08, "loss": 0.5283, "step": 10717 }, { "epoch": 0.09639659491035332, "grad_norm": 16.028610229492188, "learning_rate": 9.042490639166734e-08, "loss": 0.3713, "step": 10718 }, { "epoch": 0.09640558880799377, "grad_norm": 9.811300277709961, "learning_rate": 8.978737229308198e-08, "loss": 0.3279, "step": 10719 }, { "epoch": 0.09641458270563423, "grad_norm": 2.4900519847869873, "learning_rate": 8.915208955268261e-08, "loss": 0.6185, "step": 10720 }, { "epoch": 0.09642357660327468, "grad_norm": 2.4333558082580566, "learning_rate": 8.851905822788442e-08, "loss": 0.5606, "step": 10721 }, { "epoch": 0.09643257050091512, "grad_norm": 15.334701538085938, "learning_rate": 8.788827837589996e-08, "loss": 0.6011, "step": 10722 }, { "epoch": 0.09644156439855558, "grad_norm": 2.8548638820648193, "learning_rate": 8.725975005374476e-08, "loss": 0.132, "step": 10723 }, { "epoch": 0.09645055829619603, "grad_norm": 2.674429416656494, "learning_rate": 8.663347331821503e-08, "loss": 0.511, "step": 10724 }, { "epoch": 0.09645955219383648, "grad_norm": 2.7171108722686768, "learning_rate": 8.600944822591828e-08, "loss": 0.4938, "step": 10725 }, { "epoch": 0.09646854609147694, "grad_norm": 2.0786452293395996, "learning_rate": 8.538767483325383e-08, "loss": 0.343, "step": 10726 }, { "epoch": 0.09647753998911739, "grad_norm": 1.9657721519470215, "learning_rate": 8.476815319641562e-08, "loss": 0.1829, "step": 10727 }, { "epoch": 0.09648653388675783, "grad_norm": 18.882793426513672, "learning_rate": 8.41508833713922e-08, "loss": 0.327, "step": 10728 }, { "epoch": 0.09649552778439828, "grad_norm": 3.2475156784057617, "learning_rate": 8.353586541397784e-08, "loss": 0.3541, "step": 10729 }, { "epoch": 0.09650452168203874, "grad_norm": 14.991649627685547, "learning_rate": 8.292309937975584e-08, "loss": 0.3407, "step": 10730 }, { "epoch": 0.09651351557967919, "grad_norm": 2.1196110248565674, "learning_rate": 8.231258532410135e-08, "loss": 0.253, "step": 10731 }, { "epoch": 0.09652250947731963, "grad_norm": 3.158534288406372, "learning_rate": 8.17043233022008e-08, "loss": 0.3962, "step": 10732 }, { "epoch": 0.0965315033749601, "grad_norm": 3.657785654067993, "learning_rate": 8.109831336901852e-08, "loss": 0.3197, "step": 10733 }, { "epoch": 0.09654049727260054, "grad_norm": 3.599865198135376, "learning_rate": 8.049455557933572e-08, "loss": 0.3159, "step": 10734 }, { "epoch": 0.09654949117024099, "grad_norm": 8.849898338317871, "learning_rate": 7.989304998771152e-08, "loss": 0.6213, "step": 10735 }, { "epoch": 0.09655848506788144, "grad_norm": 2.4310519695281982, "learning_rate": 7.929379664851077e-08, "loss": 0.1844, "step": 10736 }, { "epoch": 0.0965674789655219, "grad_norm": 18.606266021728516, "learning_rate": 7.869679561589293e-08, "loss": 0.4085, "step": 10737 }, { "epoch": 0.09657647286316234, "grad_norm": 8.208226203918457, "learning_rate": 7.810204694381762e-08, "loss": 0.4382, "step": 10738 }, { "epoch": 0.09658546676080279, "grad_norm": 1.8140814304351807, "learning_rate": 7.750955068603628e-08, "loss": 0.4167, "step": 10739 }, { "epoch": 0.09659446065844325, "grad_norm": 2.8638529777526855, "learning_rate": 7.691930689609771e-08, "loss": 0.6839, "step": 10740 }, { "epoch": 0.0966034545560837, "grad_norm": 9.860882759094238, "learning_rate": 7.633131562734541e-08, "loss": 0.4739, "step": 10741 }, { "epoch": 0.09661244845372415, "grad_norm": 1.8317703008651733, "learning_rate": 7.574557693292572e-08, "loss": 0.5372, "step": 10742 }, { "epoch": 0.0966214423513646, "grad_norm": 1.9852045774459839, "learning_rate": 7.51620908657713e-08, "loss": 0.1617, "step": 10743 }, { "epoch": 0.09663043624900505, "grad_norm": 1.994433879852295, "learning_rate": 7.458085747862331e-08, "loss": 0.1375, "step": 10744 }, { "epoch": 0.0966394301466455, "grad_norm": 3.0313844680786133, "learning_rate": 7.400187682400917e-08, "loss": 0.8853, "step": 10745 }, { "epoch": 0.09664842404428595, "grad_norm": 2.0555999279022217, "learning_rate": 7.342514895425646e-08, "loss": 0.3432, "step": 10746 }, { "epoch": 0.09665741794192641, "grad_norm": 2.601400136947632, "learning_rate": 7.285067392149015e-08, "loss": 0.449, "step": 10747 }, { "epoch": 0.09666641183956685, "grad_norm": 11.781649589538574, "learning_rate": 7.227845177762982e-08, "loss": 0.8169, "step": 10748 }, { "epoch": 0.0966754057372073, "grad_norm": 1.9788801670074463, "learning_rate": 7.170848257439522e-08, "loss": 0.1663, "step": 10749 }, { "epoch": 0.09668439963484776, "grad_norm": 2.9112775325775146, "learning_rate": 7.114076636329514e-08, "loss": 0.8373, "step": 10750 }, { "epoch": 0.09669339353248821, "grad_norm": 2.2927982807159424, "learning_rate": 7.057530319564409e-08, "loss": 0.4254, "step": 10751 }, { "epoch": 0.09670238743012866, "grad_norm": 1.8438622951507568, "learning_rate": 7.001209312254286e-08, "loss": 0.0532, "step": 10752 }, { "epoch": 0.0967113813277691, "grad_norm": 2.7279999256134033, "learning_rate": 6.945113619489796e-08, "loss": 0.506, "step": 10753 }, { "epoch": 0.09672037522540956, "grad_norm": 2.311075448989868, "learning_rate": 6.889243246340771e-08, "loss": 0.2582, "step": 10754 }, { "epoch": 0.09672936912305001, "grad_norm": 2.995574712753296, "learning_rate": 6.833598197856228e-08, "loss": 0.7709, "step": 10755 }, { "epoch": 0.09673836302069046, "grad_norm": 4.3128204345703125, "learning_rate": 6.778178479065756e-08, "loss": 0.4744, "step": 10756 }, { "epoch": 0.09674735691833092, "grad_norm": 2.3339693546295166, "learning_rate": 6.722984094978402e-08, "loss": 0.5086, "step": 10757 }, { "epoch": 0.09675635081597136, "grad_norm": 3.1689507961273193, "learning_rate": 6.668015050581844e-08, "loss": 0.2092, "step": 10758 }, { "epoch": 0.09676534471361181, "grad_norm": 11.694401741027832, "learning_rate": 6.613271350844608e-08, "loss": 0.8194, "step": 10759 }, { "epoch": 0.09677433861125226, "grad_norm": 2.2125396728515625, "learning_rate": 6.558753000714124e-08, "loss": 0.3409, "step": 10760 }, { "epoch": 0.09678333250889272, "grad_norm": 12.064775466918945, "learning_rate": 6.504460005117841e-08, "loss": 0.8825, "step": 10761 }, { "epoch": 0.09679232640653317, "grad_norm": 1.9521691799163818, "learning_rate": 6.450392368962943e-08, "loss": 0.2262, "step": 10762 }, { "epoch": 0.09680132030417361, "grad_norm": 15.282927513122559, "learning_rate": 6.396550097135523e-08, "loss": 0.4443, "step": 10763 }, { "epoch": 0.09681031420181407, "grad_norm": 2.0000112056732178, "learning_rate": 6.342933194501965e-08, "loss": 0.2302, "step": 10764 }, { "epoch": 0.09681930809945452, "grad_norm": 3.1112520694732666, "learning_rate": 6.289541665908394e-08, "loss": 0.4035, "step": 10765 }, { "epoch": 0.09682830199709497, "grad_norm": 14.122337341308594, "learning_rate": 6.236375516179838e-08, "loss": 0.6715, "step": 10766 }, { "epoch": 0.09683729589473543, "grad_norm": 3.27746319770813, "learning_rate": 6.183434750121619e-08, "loss": 0.2177, "step": 10767 }, { "epoch": 0.09684628979237588, "grad_norm": 9.228893280029297, "learning_rate": 6.130719372518523e-08, "loss": 0.6795, "step": 10768 }, { "epoch": 0.09685528369001632, "grad_norm": 2.9165308475494385, "learning_rate": 6.078229388134793e-08, "loss": 0.5656, "step": 10769 }, { "epoch": 0.09686427758765677, "grad_norm": 19.76356315612793, "learning_rate": 6.025964801714412e-08, "loss": 0.7958, "step": 10770 }, { "epoch": 0.09687327148529723, "grad_norm": 3.688864231109619, "learning_rate": 5.97392561798138e-08, "loss": 0.3136, "step": 10771 }, { "epoch": 0.09688226538293768, "grad_norm": 3.074702739715576, "learning_rate": 5.922111841638323e-08, "loss": 0.142, "step": 10772 }, { "epoch": 0.09689125928057812, "grad_norm": 2.1549861431121826, "learning_rate": 5.870523477368439e-08, "loss": 0.1749, "step": 10773 }, { "epoch": 0.09690025317821858, "grad_norm": 2.0674638748168945, "learning_rate": 5.819160529834111e-08, "loss": 0.1351, "step": 10774 }, { "epoch": 0.09690924707585903, "grad_norm": 16.05478286743164, "learning_rate": 5.768023003677736e-08, "loss": 0.7291, "step": 10775 }, { "epoch": 0.09691824097349948, "grad_norm": 2.6236495971679688, "learning_rate": 5.717110903520617e-08, "loss": 0.4224, "step": 10776 }, { "epoch": 0.09692723487113993, "grad_norm": 2.053647756576538, "learning_rate": 5.666424233964629e-08, "loss": 0.4452, "step": 10777 }, { "epoch": 0.09693622876878039, "grad_norm": 2.449167490005493, "learning_rate": 5.615962999590274e-08, "loss": 0.2007, "step": 10778 }, { "epoch": 0.09694522266642083, "grad_norm": 1.8996410369873047, "learning_rate": 5.565727204958626e-08, "loss": 0.4341, "step": 10779 }, { "epoch": 0.09695421656406128, "grad_norm": 2.5547664165496826, "learning_rate": 5.515716854609942e-08, "loss": 0.5127, "step": 10780 }, { "epoch": 0.09696321046170174, "grad_norm": 2.3925302028656006, "learning_rate": 5.4659319530636633e-08, "loss": 0.2967, "step": 10781 }, { "epoch": 0.09697220435934219, "grad_norm": 3.677001476287842, "learning_rate": 5.4163725048195225e-08, "loss": 0.3698, "step": 10782 }, { "epoch": 0.09698119825698263, "grad_norm": 2.584214210510254, "learning_rate": 5.3670385143569924e-08, "loss": 0.3407, "step": 10783 }, { "epoch": 0.0969901921546231, "grad_norm": 16.368074417114258, "learning_rate": 5.317929986134174e-08, "loss": 0.6443, "step": 10784 }, { "epoch": 0.09699918605226354, "grad_norm": 2.5952067375183105, "learning_rate": 5.2690469245900154e-08, "loss": 0.5793, "step": 10785 }, { "epoch": 0.09700817994990399, "grad_norm": 3.5613768100738525, "learning_rate": 5.2203893341423726e-08, "loss": 0.6721, "step": 10786 }, { "epoch": 0.09701717384754444, "grad_norm": 3.156369924545288, "learning_rate": 5.171957219188561e-08, "loss": 0.2924, "step": 10787 }, { "epoch": 0.0970261677451849, "grad_norm": 3.274749755859375, "learning_rate": 5.1237505841061904e-08, "loss": 0.4838, "step": 10788 }, { "epoch": 0.09703516164282534, "grad_norm": 12.769189834594727, "learning_rate": 5.075769433252053e-08, "loss": 0.4111, "step": 10789 }, { "epoch": 0.09704415554046579, "grad_norm": 1.8616174459457397, "learning_rate": 5.028013770962403e-08, "loss": 0.3524, "step": 10790 }, { "epoch": 0.09705314943810625, "grad_norm": 2.6696929931640625, "learning_rate": 4.9804836015535096e-08, "loss": 0.7691, "step": 10791 }, { "epoch": 0.0970621433357467, "grad_norm": 17.250652313232422, "learning_rate": 4.9331789293211026e-08, "loss": 0.4988, "step": 10792 }, { "epoch": 0.09707113723338714, "grad_norm": 2.1982035636901855, "learning_rate": 4.886099758540652e-08, "loss": 0.4744, "step": 10793 }, { "epoch": 0.09708013113102759, "grad_norm": 9.256131172180176, "learning_rate": 4.8392460934668096e-08, "loss": 0.7915, "step": 10794 }, { "epoch": 0.09708912502866805, "grad_norm": 3.2708027362823486, "learning_rate": 4.7926179383345204e-08, "loss": 0.4628, "step": 10795 }, { "epoch": 0.0970981189263085, "grad_norm": 20.521011352539062, "learning_rate": 4.746215297357637e-08, "loss": 0.3918, "step": 10796 }, { "epoch": 0.09710711282394895, "grad_norm": 11.771196365356445, "learning_rate": 4.700038174730303e-08, "loss": 0.754, "step": 10797 }, { "epoch": 0.0971161067215894, "grad_norm": 2.0202133655548096, "learning_rate": 4.654086574625571e-08, "loss": 0.3443, "step": 10798 }, { "epoch": 0.09712510061922985, "grad_norm": 9.68991756439209, "learning_rate": 4.608360501196785e-08, "loss": 0.3159, "step": 10799 }, { "epoch": 0.0971340945168703, "grad_norm": 2.374370813369751, "learning_rate": 4.5628599585767487e-08, "loss": 0.5307, "step": 10800 }, { "epoch": 0.09714308841451075, "grad_norm": 2.234515428543091, "learning_rate": 4.517584950877452e-08, "loss": 0.3024, "step": 10801 }, { "epoch": 0.09715208231215121, "grad_norm": 1.9921138286590576, "learning_rate": 4.4725354821911755e-08, "loss": 0.7296, "step": 10802 }, { "epoch": 0.09716107620979166, "grad_norm": 11.038368225097656, "learning_rate": 4.427711556588832e-08, "loss": 0.9471, "step": 10803 }, { "epoch": 0.0971700701074321, "grad_norm": 1.996293306350708, "learning_rate": 4.3831131781219006e-08, "loss": 0.3566, "step": 10804 }, { "epoch": 0.09717906400507256, "grad_norm": 2.675919771194458, "learning_rate": 4.338740350821324e-08, "loss": 0.2704, "step": 10805 }, { "epoch": 0.09718805790271301, "grad_norm": 2.553147077560425, "learning_rate": 4.2945930786972287e-08, "loss": 0.6795, "step": 10806 }, { "epoch": 0.09719705180035346, "grad_norm": 2.6418111324310303, "learning_rate": 4.250671365739756e-08, "loss": 0.3589, "step": 10807 }, { "epoch": 0.09720604569799392, "grad_norm": 2.6185994148254395, "learning_rate": 4.20697521591823e-08, "loss": 0.2989, "step": 10808 }, { "epoch": 0.09721503959563436, "grad_norm": 17.995094299316406, "learning_rate": 4.163504633182269e-08, "loss": 0.6295, "step": 10809 }, { "epoch": 0.09722403349327481, "grad_norm": 19.785747528076172, "learning_rate": 4.120259621460121e-08, "loss": 0.4826, "step": 10810 }, { "epoch": 0.09723302739091526, "grad_norm": 1.7818496227264404, "learning_rate": 4.07724018466088e-08, "loss": 0.6301, "step": 10811 }, { "epoch": 0.09724202128855572, "grad_norm": 2.82384991645813, "learning_rate": 4.034446326671992e-08, "loss": 0.6478, "step": 10812 }, { "epoch": 0.09725101518619617, "grad_norm": 2.7577829360961914, "learning_rate": 3.991878051361752e-08, "loss": 0.4123, "step": 10813 }, { "epoch": 0.09726000908383661, "grad_norm": 1.8911179304122925, "learning_rate": 3.949535362576806e-08, "loss": 0.4637, "step": 10814 }, { "epoch": 0.09726900298147707, "grad_norm": 9.875456809997559, "learning_rate": 3.9074182641446465e-08, "loss": 0.4427, "step": 10815 }, { "epoch": 0.09727799687911752, "grad_norm": 3.132779598236084, "learning_rate": 3.865526759871119e-08, "loss": 0.3307, "step": 10816 }, { "epoch": 0.09728699077675797, "grad_norm": 13.457146644592285, "learning_rate": 3.8238608535429157e-08, "loss": 0.3573, "step": 10817 }, { "epoch": 0.09729598467439841, "grad_norm": 2.9968996047973633, "learning_rate": 3.7824205489253587e-08, "loss": 0.7162, "step": 10818 }, { "epoch": 0.09730497857203887, "grad_norm": 2.7791388034820557, "learning_rate": 3.741205849764062e-08, "loss": 0.5225, "step": 10819 }, { "epoch": 0.09731397246967932, "grad_norm": 2.273871898651123, "learning_rate": 3.7002167597835477e-08, "loss": 0.3432, "step": 10820 }, { "epoch": 0.09732296636731977, "grad_norm": 3.5131893157958984, "learning_rate": 3.6594532826889075e-08, "loss": 0.4372, "step": 10821 }, { "epoch": 0.09733196026496023, "grad_norm": 2.463414192199707, "learning_rate": 3.61891542216386e-08, "loss": 0.525, "step": 10822 }, { "epoch": 0.09734095416260068, "grad_norm": 2.2324466705322266, "learning_rate": 3.578603181872697e-08, "loss": 0.4777, "step": 10823 }, { "epoch": 0.09734994806024112, "grad_norm": 2.9174864292144775, "learning_rate": 3.538516565458061e-08, "loss": 0.3574, "step": 10824 }, { "epoch": 0.09735894195788158, "grad_norm": 18.72479820251465, "learning_rate": 3.4986555765434413e-08, "loss": 0.7678, "step": 10825 }, { "epoch": 0.09736793585552203, "grad_norm": 2.469282865524292, "learning_rate": 3.459020218731512e-08, "loss": 0.4215, "step": 10826 }, { "epoch": 0.09737692975316248, "grad_norm": 9.256239891052246, "learning_rate": 3.419610495604131e-08, "loss": 0.5088, "step": 10827 }, { "epoch": 0.09738592365080292, "grad_norm": 2.106294631958008, "learning_rate": 3.380426410723447e-08, "loss": 0.1493, "step": 10828 }, { "epoch": 0.09739491754844339, "grad_norm": 2.1244988441467285, "learning_rate": 3.3414679676305184e-08, "loss": 0.2872, "step": 10829 }, { "epoch": 0.09740391144608383, "grad_norm": 2.630458354949951, "learning_rate": 3.302735169846416e-08, "loss": 0.3033, "step": 10830 }, { "epoch": 0.09741290534372428, "grad_norm": 3.526090383529663, "learning_rate": 3.264228020871951e-08, "loss": 0.4201, "step": 10831 }, { "epoch": 0.09742189924136474, "grad_norm": 1.9035688638687134, "learning_rate": 3.2259465241873954e-08, "loss": 0.3417, "step": 10832 }, { "epoch": 0.09743089313900519, "grad_norm": 15.018309593200684, "learning_rate": 3.1878906832524815e-08, "loss": 0.5403, "step": 10833 }, { "epoch": 0.09743988703664563, "grad_norm": 13.848657608032227, "learning_rate": 3.15006050150668e-08, "loss": 0.4837, "step": 10834 }, { "epoch": 0.09744888093428608, "grad_norm": 10.989568710327148, "learning_rate": 3.1124559823689226e-08, "loss": 0.5924, "step": 10835 }, { "epoch": 0.09745787483192654, "grad_norm": 12.61654281616211, "learning_rate": 3.075077129238158e-08, "loss": 0.6806, "step": 10836 }, { "epoch": 0.09746686872956699, "grad_norm": 14.77175521850586, "learning_rate": 3.037923945492238e-08, "loss": 0.5901, "step": 10837 }, { "epoch": 0.09747586262720744, "grad_norm": 2.040630340576172, "learning_rate": 3.000996434489311e-08, "loss": 0.5652, "step": 10838 }, { "epoch": 0.0974848565248479, "grad_norm": 3.288139581680298, "learning_rate": 2.964294599566708e-08, "loss": 0.293, "step": 10839 }, { "epoch": 0.09749385042248834, "grad_norm": 1.982471227645874, "learning_rate": 2.9278184440417744e-08, "loss": 0.3144, "step": 10840 }, { "epoch": 0.09750284432012879, "grad_norm": 2.1151909828186035, "learning_rate": 2.8915679712107622e-08, "loss": 0.7363, "step": 10841 }, { "epoch": 0.09751183821776924, "grad_norm": 9.731039047241211, "learning_rate": 2.8555431843504954e-08, "loss": 0.4288, "step": 10842 }, { "epoch": 0.0975208321154097, "grad_norm": 2.9558393955230713, "learning_rate": 2.8197440867161473e-08, "loss": 0.2624, "step": 10843 }, { "epoch": 0.09752982601305014, "grad_norm": 1.8875575065612793, "learning_rate": 2.7841706815437408e-08, "loss": 0.2735, "step": 10844 }, { "epoch": 0.09753881991069059, "grad_norm": 2.159536838531494, "learning_rate": 2.748822972048204e-08, "loss": 0.3576, "step": 10845 }, { "epoch": 0.09754781380833105, "grad_norm": 17.144332885742188, "learning_rate": 2.7137009614242038e-08, "loss": 0.2612, "step": 10846 }, { "epoch": 0.0975568077059715, "grad_norm": 15.148889541625977, "learning_rate": 2.6788046528461453e-08, "loss": 0.6106, "step": 10847 }, { "epoch": 0.09756580160361195, "grad_norm": 3.163529634475708, "learning_rate": 2.6441340494678945e-08, "loss": 0.3094, "step": 10848 }, { "epoch": 0.0975747955012524, "grad_norm": 1.7897125482559204, "learning_rate": 2.6096891544227788e-08, "loss": 0.5534, "step": 10849 }, { "epoch": 0.09758378939889285, "grad_norm": 2.1564829349517822, "learning_rate": 2.5754699708241402e-08, "loss": 0.3676, "step": 10850 }, { "epoch": 0.0975927832965333, "grad_norm": 2.1563827991485596, "learning_rate": 2.5414765017642284e-08, "loss": 0.3957, "step": 10851 }, { "epoch": 0.09760177719417375, "grad_norm": 12.855997085571289, "learning_rate": 2.5077087503158625e-08, "loss": 0.5053, "step": 10852 }, { "epoch": 0.09761077109181421, "grad_norm": 17.650911331176758, "learning_rate": 2.4741667195304907e-08, "loss": 0.4483, "step": 10853 }, { "epoch": 0.09761976498945465, "grad_norm": 2.1513326168060303, "learning_rate": 2.4408504124398545e-08, "loss": 0.4235, "step": 10854 }, { "epoch": 0.0976287588870951, "grad_norm": 17.52141571044922, "learning_rate": 2.4077598320551565e-08, "loss": 0.4733, "step": 10855 }, { "epoch": 0.09763775278473556, "grad_norm": 2.8524599075317383, "learning_rate": 2.3748949813670596e-08, "loss": 0.4823, "step": 10856 }, { "epoch": 0.09764674668237601, "grad_norm": 3.2618181705474854, "learning_rate": 2.3422558633454105e-08, "loss": 0.3701, "step": 10857 }, { "epoch": 0.09765574058001646, "grad_norm": 2.5880162715911865, "learning_rate": 2.309842480940627e-08, "loss": 0.3836, "step": 10858 }, { "epoch": 0.0976647344776569, "grad_norm": 2.4129557609558105, "learning_rate": 2.2776548370817552e-08, "loss": 0.4912, "step": 10859 }, { "epoch": 0.09767372837529736, "grad_norm": 2.0683887004852295, "learning_rate": 2.2456929346784117e-08, "loss": 0.3071, "step": 10860 }, { "epoch": 0.09768272227293781, "grad_norm": 2.880574941635132, "learning_rate": 2.213956776618842e-08, "loss": 0.459, "step": 10861 }, { "epoch": 0.09769171617057826, "grad_norm": 1.826324224472046, "learning_rate": 2.1824463657715842e-08, "loss": 0.4079, "step": 10862 }, { "epoch": 0.09770071006821872, "grad_norm": 2.8562850952148438, "learning_rate": 2.1511617049840836e-08, "loss": 0.6497, "step": 10863 }, { "epoch": 0.09770970396585917, "grad_norm": 2.5885238647460938, "learning_rate": 2.1201027970843556e-08, "loss": 0.3745, "step": 10864 }, { "epoch": 0.09771869786349961, "grad_norm": 2.5068962574005127, "learning_rate": 2.0892696448793214e-08, "loss": 0.5384, "step": 10865 }, { "epoch": 0.09772769176114007, "grad_norm": 2.6634609699249268, "learning_rate": 2.0586622511553633e-08, "loss": 0.0578, "step": 10866 }, { "epoch": 0.09773668565878052, "grad_norm": 2.954500913619995, "learning_rate": 2.0282806186791568e-08, "loss": 0.3729, "step": 10867 }, { "epoch": 0.09774567955642097, "grad_norm": 2.256135940551758, "learning_rate": 1.998124750196284e-08, "loss": 0.5448, "step": 10868 }, { "epoch": 0.09775467345406141, "grad_norm": 17.49842643737793, "learning_rate": 1.9681946484320644e-08, "loss": 0.793, "step": 10869 }, { "epoch": 0.09776366735170187, "grad_norm": 15.314598083496094, "learning_rate": 1.938490316092112e-08, "loss": 0.4181, "step": 10870 }, { "epoch": 0.09777266124934232, "grad_norm": 4.890852451324463, "learning_rate": 1.9090117558603904e-08, "loss": 0.2747, "step": 10871 }, { "epoch": 0.09778165514698277, "grad_norm": 1.87319815158844, "learning_rate": 1.8797589704017127e-08, "loss": 0.497, "step": 10872 }, { "epoch": 0.09779064904462323, "grad_norm": 2.4461374282836914, "learning_rate": 1.85073196235952e-08, "loss": 0.4074, "step": 10873 }, { "epoch": 0.09779964294226368, "grad_norm": 2.485321044921875, "learning_rate": 1.8219307343575464e-08, "loss": 0.9442, "step": 10874 }, { "epoch": 0.09780863683990412, "grad_norm": 2.061397075653076, "learning_rate": 1.793355288998433e-08, "loss": 0.4182, "step": 10875 }, { "epoch": 0.09781763073754457, "grad_norm": 2.5469160079956055, "learning_rate": 1.7650056288651127e-08, "loss": 0.792, "step": 10876 }, { "epoch": 0.09782662463518503, "grad_norm": 3.074612617492676, "learning_rate": 1.7368817565197037e-08, "loss": 0.4682, "step": 10877 }, { "epoch": 0.09783561853282548, "grad_norm": 1.871274709701538, "learning_rate": 1.7089836745040612e-08, "loss": 0.3509, "step": 10878 }, { "epoch": 0.09784461243046592, "grad_norm": 2.145906925201416, "learning_rate": 1.6813113853395014e-08, "loss": 0.2661, "step": 10879 }, { "epoch": 0.09785360632810639, "grad_norm": 1.8797953128814697, "learning_rate": 1.6538648915270792e-08, "loss": 0.248, "step": 10880 }, { "epoch": 0.09786260022574683, "grad_norm": 2.8489556312561035, "learning_rate": 1.626644195547311e-08, "loss": 0.4418, "step": 10881 }, { "epoch": 0.09787159412338728, "grad_norm": 3.1737210750579834, "learning_rate": 1.5996492998604508e-08, "loss": 0.0803, "step": 10882 }, { "epoch": 0.09788058802102774, "grad_norm": 2.7821784019470215, "learning_rate": 1.572880206906213e-08, "loss": 0.525, "step": 10883 }, { "epoch": 0.09788958191866819, "grad_norm": 2.487565040588379, "learning_rate": 1.5463369191037746e-08, "loss": 0.4332, "step": 10884 }, { "epoch": 0.09789857581630863, "grad_norm": 9.785359382629395, "learning_rate": 1.5200194388526045e-08, "loss": 0.4451, "step": 10885 }, { "epoch": 0.09790756971394908, "grad_norm": 3.028916597366333, "learning_rate": 1.4939277685308005e-08, "loss": 0.4155, "step": 10886 }, { "epoch": 0.09791656361158954, "grad_norm": 2.2725136280059814, "learning_rate": 1.468061910496754e-08, "loss": 0.5218, "step": 10887 }, { "epoch": 0.09792555750922999, "grad_norm": 19.181774139404297, "learning_rate": 1.4424218670880396e-08, "loss": 0.5693, "step": 10888 }, { "epoch": 0.09793455140687043, "grad_norm": 1.8172012567520142, "learning_rate": 1.4170076406219702e-08, "loss": 0.449, "step": 10889 }, { "epoch": 0.0979435453045109, "grad_norm": 16.111286163330078, "learning_rate": 1.3918192333953194e-08, "loss": 0.5086, "step": 10890 }, { "epoch": 0.09795253920215134, "grad_norm": 3.7883381843566895, "learning_rate": 1.3668566476848777e-08, "loss": 0.2742, "step": 10891 }, { "epoch": 0.09796153309979179, "grad_norm": 13.722253799438477, "learning_rate": 1.3421198857468952e-08, "loss": 0.7133, "step": 10892 }, { "epoch": 0.09797052699743224, "grad_norm": 1.9584861993789673, "learning_rate": 1.3176089498165289e-08, "loss": 0.3825, "step": 10893 }, { "epoch": 0.0979795208950727, "grad_norm": 8.636361122131348, "learning_rate": 1.2933238421092286e-08, "loss": 0.568, "step": 10894 }, { "epoch": 0.09798851479271314, "grad_norm": 2.255974769592285, "learning_rate": 1.2692645648201829e-08, "loss": 0.5551, "step": 10895 }, { "epoch": 0.09799750869035359, "grad_norm": 2.0465385913848877, "learning_rate": 1.2454311201234858e-08, "loss": 0.4647, "step": 10896 }, { "epoch": 0.09800650258799405, "grad_norm": 18.847827911376953, "learning_rate": 1.2218235101732478e-08, "loss": 0.5416, "step": 10897 }, { "epoch": 0.0980154964856345, "grad_norm": 16.179523468017578, "learning_rate": 1.1984417371030399e-08, "loss": 0.3905, "step": 10898 }, { "epoch": 0.09802449038327495, "grad_norm": 18.522647857666016, "learning_rate": 1.1752858030264491e-08, "loss": 0.4509, "step": 10899 }, { "epoch": 0.09803348428091539, "grad_norm": 2.6772940158843994, "learning_rate": 1.1523557100356908e-08, "loss": 0.8334, "step": 10900 }, { "epoch": 0.09804247817855585, "grad_norm": 1.906054139137268, "learning_rate": 1.1296514602038288e-08, "loss": 0.4029, "step": 10901 }, { "epoch": 0.0980514720761963, "grad_norm": 1.986649513244629, "learning_rate": 1.107173055582278e-08, "loss": 0.21, "step": 10902 }, { "epoch": 0.09806046597383675, "grad_norm": 2.0732619762420654, "learning_rate": 1.084920498202746e-08, "loss": 0.5279, "step": 10903 }, { "epoch": 0.09806945987147721, "grad_norm": 2.745117425918579, "learning_rate": 1.06289379007668e-08, "loss": 0.4823, "step": 10904 }, { "epoch": 0.09807845376911765, "grad_norm": 11.081366539001465, "learning_rate": 1.0410929331947095e-08, "loss": 0.5589, "step": 10905 }, { "epoch": 0.0980874476667581, "grad_norm": 2.0404975414276123, "learning_rate": 1.0195179295269252e-08, "loss": 0.2927, "step": 10906 }, { "epoch": 0.09809644156439856, "grad_norm": 13.293797492980957, "learning_rate": 9.981687810234341e-09, "loss": 0.652, "step": 10907 }, { "epoch": 0.09810543546203901, "grad_norm": 1.8948400020599365, "learning_rate": 9.770454896138038e-09, "loss": 0.315, "step": 10908 }, { "epoch": 0.09811442935967946, "grad_norm": 2.3651375770568848, "learning_rate": 9.561480572070625e-09, "loss": 0.8432, "step": 10909 }, { "epoch": 0.0981234232573199, "grad_norm": 4.179783344268799, "learning_rate": 9.354764856919773e-09, "loss": 0.5686, "step": 10910 }, { "epoch": 0.09813241715496036, "grad_norm": 2.253962993621826, "learning_rate": 9.150307769367761e-09, "loss": 0.2332, "step": 10911 }, { "epoch": 0.09814141105260081, "grad_norm": 1.6781617403030396, "learning_rate": 8.948109327891473e-09, "loss": 0.4164, "step": 10912 }, { "epoch": 0.09815040495024126, "grad_norm": 2.0865468978881836, "learning_rate": 8.74816955076796e-09, "loss": 0.4355, "step": 10913 }, { "epoch": 0.09815939884788172, "grad_norm": 2.1517326831817627, "learning_rate": 8.550488456066097e-09, "loss": 0.28, "step": 10914 }, { "epoch": 0.09816839274552216, "grad_norm": 11.383784294128418, "learning_rate": 8.355066061654927e-09, "loss": 0.442, "step": 10915 }, { "epoch": 0.09817738664316261, "grad_norm": 2.656196117401123, "learning_rate": 8.161902385189769e-09, "loss": 0.4242, "step": 10916 }, { "epoch": 0.09818638054080306, "grad_norm": 2.3299901485443115, "learning_rate": 7.970997444137208e-09, "loss": 0.7701, "step": 10917 }, { "epoch": 0.09819537443844352, "grad_norm": 3.612933874130249, "learning_rate": 7.782351255744558e-09, "loss": 0.4332, "step": 10918 }, { "epoch": 0.09820436833608397, "grad_norm": 2.240366220474243, "learning_rate": 7.59596383706207e-09, "loss": 0.5586, "step": 10919 }, { "epoch": 0.09821336223372441, "grad_norm": 4.525698661804199, "learning_rate": 7.411835204940154e-09, "loss": 0.0823, "step": 10920 }, { "epoch": 0.09822235613136487, "grad_norm": 19.085830688476562, "learning_rate": 7.2299653760127265e-09, "loss": 0.4132, "step": 10921 }, { "epoch": 0.09823135002900532, "grad_norm": 1.964813470840454, "learning_rate": 7.050354366724965e-09, "loss": 0.3341, "step": 10922 }, { "epoch": 0.09824034392664577, "grad_norm": 7.475910663604736, "learning_rate": 6.87300219330278e-09, "loss": 0.8756, "step": 10923 }, { "epoch": 0.09824933782428623, "grad_norm": 9.967905044555664, "learning_rate": 6.697908871780567e-09, "loss": 0.5255, "step": 10924 }, { "epoch": 0.09825833172192668, "grad_norm": 2.213815450668335, "learning_rate": 6.525074417979005e-09, "loss": 0.2371, "step": 10925 }, { "epoch": 0.09826732561956712, "grad_norm": 2.244016170501709, "learning_rate": 6.354498847521706e-09, "loss": 0.3848, "step": 10926 }, { "epoch": 0.09827631951720757, "grad_norm": 1.752897024154663, "learning_rate": 6.186182175824118e-09, "loss": 0.4953, "step": 10927 }, { "epoch": 0.09828531341484803, "grad_norm": 2.647512912750244, "learning_rate": 6.020124418099071e-09, "loss": 0.683, "step": 10928 }, { "epoch": 0.09829430731248848, "grad_norm": 15.591897964477539, "learning_rate": 5.85632558935123e-09, "loss": 0.5005, "step": 10929 }, { "epoch": 0.09830330121012892, "grad_norm": 4.80480432510376, "learning_rate": 5.694785704388195e-09, "loss": 0.3513, "step": 10930 }, { "epoch": 0.09831229510776938, "grad_norm": 15.414206504821777, "learning_rate": 5.535504777809397e-09, "loss": 0.4507, "step": 10931 }, { "epoch": 0.09832128900540983, "grad_norm": 3.6889686584472656, "learning_rate": 5.378482824011655e-09, "loss": 0.4741, "step": 10932 }, { "epoch": 0.09833028290305028, "grad_norm": 2.0445079803466797, "learning_rate": 5.223719857180842e-09, "loss": 0.3508, "step": 10933 }, { "epoch": 0.09833927680069073, "grad_norm": 3.188014507293701, "learning_rate": 5.07121589131132e-09, "loss": 0.33, "step": 10934 }, { "epoch": 0.09834827069833119, "grad_norm": 2.257406234741211, "learning_rate": 4.920970940180958e-09, "loss": 0.4109, "step": 10935 }, { "epoch": 0.09835726459597163, "grad_norm": 1.8172687292099, "learning_rate": 4.772985017370557e-09, "loss": 0.2913, "step": 10936 }, { "epoch": 0.09836625849361208, "grad_norm": 2.638300657272339, "learning_rate": 4.62725813625553e-09, "loss": 0.8257, "step": 10937 }, { "epoch": 0.09837525239125254, "grad_norm": 3.0135228633880615, "learning_rate": 4.483790310005898e-09, "loss": 0.3883, "step": 10938 }, { "epoch": 0.09838424628889299, "grad_norm": 2.145209312438965, "learning_rate": 4.342581551589065e-09, "loss": 0.5481, "step": 10939 }, { "epoch": 0.09839324018653343, "grad_norm": 4.0442047119140625, "learning_rate": 4.203631873767045e-09, "loss": 0.3298, "step": 10940 }, { "epoch": 0.09840223408417388, "grad_norm": 1.6987946033477783, "learning_rate": 4.06694128909646e-09, "loss": 0.4671, "step": 10941 }, { "epoch": 0.09841122798181434, "grad_norm": 10.63873291015625, "learning_rate": 3.932509809934093e-09, "loss": 0.6655, "step": 10942 }, { "epoch": 0.09842022187945479, "grad_norm": 1.9131672382354736, "learning_rate": 3.800337448425784e-09, "loss": 0.1824, "step": 10943 }, { "epoch": 0.09842921577709524, "grad_norm": 16.18643569946289, "learning_rate": 3.670424216520307e-09, "loss": 0.562, "step": 10944 }, { "epoch": 0.0984382096747357, "grad_norm": 10.031086921691895, "learning_rate": 3.5427701259582724e-09, "loss": 0.5073, "step": 10945 }, { "epoch": 0.09844720357237614, "grad_norm": 2.377021551132202, "learning_rate": 3.417375188274896e-09, "loss": 0.3626, "step": 10946 }, { "epoch": 0.09845619747001659, "grad_norm": 2.630786895751953, "learning_rate": 3.294239414808331e-09, "loss": 0.5275, "step": 10947 }, { "epoch": 0.09846519136765705, "grad_norm": 2.5689923763275146, "learning_rate": 3.1733628166802364e-09, "loss": 0.4465, "step": 10948 }, { "epoch": 0.0984741852652975, "grad_norm": 1.8861314058303833, "learning_rate": 3.0547454048235336e-09, "loss": 0.4109, "step": 10949 }, { "epoch": 0.09848317916293794, "grad_norm": 2.122084140777588, "learning_rate": 2.9383871899518745e-09, "loss": 0.3843, "step": 10950 }, { "epoch": 0.09849217306057839, "grad_norm": 2.3624427318573, "learning_rate": 2.8242881825846223e-09, "loss": 0.5859, "step": 10951 }, { "epoch": 0.09850116695821885, "grad_norm": 3.252140522003174, "learning_rate": 2.7124483930329737e-09, "loss": 0.6338, "step": 10952 }, { "epoch": 0.0985101608558593, "grad_norm": 3.70107364654541, "learning_rate": 2.6028678314055087e-09, "loss": 0.5449, "step": 10953 }, { "epoch": 0.09851915475349975, "grad_norm": 16.40505027770996, "learning_rate": 2.4955465076054173e-09, "loss": 0.5041, "step": 10954 }, { "epoch": 0.09852814865114021, "grad_norm": 2.0422050952911377, "learning_rate": 2.390484431333273e-09, "loss": 0.5515, "step": 10955 }, { "epoch": 0.09853714254878065, "grad_norm": 3.7918665409088135, "learning_rate": 2.287681612084258e-09, "loss": 0.4759, "step": 10956 }, { "epoch": 0.0985461364464211, "grad_norm": 2.495706558227539, "learning_rate": 2.1871380591509392e-09, "loss": 0.12, "step": 10957 }, { "epoch": 0.09855513034406155, "grad_norm": 3.793632745742798, "learning_rate": 2.0888537816149413e-09, "loss": 0.4859, "step": 10958 }, { "epoch": 0.09856412424170201, "grad_norm": 3.3100168704986572, "learning_rate": 1.9928287883663744e-09, "loss": 0.402, "step": 10959 }, { "epoch": 0.09857311813934246, "grad_norm": 2.650324821472168, "learning_rate": 1.8990630880788562e-09, "loss": 0.4365, "step": 10960 }, { "epoch": 0.0985821120369829, "grad_norm": 2.4900963306427, "learning_rate": 1.8075566892261641e-09, "loss": 0.4078, "step": 10961 }, { "epoch": 0.09859110593462336, "grad_norm": 2.0825424194335938, "learning_rate": 1.718309600082235e-09, "loss": 0.5679, "step": 10962 }, { "epoch": 0.09860009983226381, "grad_norm": 2.2773990631103516, "learning_rate": 1.6313218287128395e-09, "loss": 0.5063, "step": 10963 }, { "epoch": 0.09860909372990426, "grad_norm": 13.169318199157715, "learning_rate": 1.5465933829783563e-09, "loss": 0.5398, "step": 10964 }, { "epoch": 0.09861808762754472, "grad_norm": 12.313261032104492, "learning_rate": 1.4641242705365487e-09, "loss": 0.7296, "step": 10965 }, { "epoch": 0.09862708152518516, "grad_norm": 2.1145291328430176, "learning_rate": 1.383914498839789e-09, "loss": 0.3489, "step": 10966 }, { "epoch": 0.09863607542282561, "grad_norm": 13.603160858154297, "learning_rate": 1.3059640751378333e-09, "loss": 0.4415, "step": 10967 }, { "epoch": 0.09864506932046606, "grad_norm": 2.3268942832946777, "learning_rate": 1.2302730064778222e-09, "loss": 0.7978, "step": 10968 }, { "epoch": 0.09865406321810652, "grad_norm": 39.744747161865234, "learning_rate": 1.15684129969873e-09, "loss": 0.5907, "step": 10969 }, { "epoch": 0.09866305711574697, "grad_norm": 2.4798409938812256, "learning_rate": 1.0856689614396898e-09, "loss": 0.5807, "step": 10970 }, { "epoch": 0.09867205101338741, "grad_norm": 19.208757400512695, "learning_rate": 1.0167559981288931e-09, "loss": 0.4429, "step": 10971 }, { "epoch": 0.09868104491102787, "grad_norm": 2.872610569000244, "learning_rate": 9.501024159974669e-10, "loss": 0.4216, "step": 10972 }, { "epoch": 0.09869003880866832, "grad_norm": 2.121644973754883, "learning_rate": 8.857082210711464e-10, "loss": 0.2531, "step": 10973 }, { "epoch": 0.09869903270630877, "grad_norm": 2.2825682163238525, "learning_rate": 8.235734191647248e-10, "loss": 0.2504, "step": 10974 }, { "epoch": 0.09870802660394921, "grad_norm": 3.0773966312408447, "learning_rate": 7.636980158987062e-10, "loss": 0.3833, "step": 10975 }, { "epoch": 0.09871702050158968, "grad_norm": 3.1123545169830322, "learning_rate": 7.06082016682652e-10, "loss": 0.3943, "step": 10976 }, { "epoch": 0.09872601439923012, "grad_norm": 1.9091317653656006, "learning_rate": 6.507254267235085e-10, "loss": 0.5484, "step": 10977 }, { "epoch": 0.09873500829687057, "grad_norm": 3.558173418045044, "learning_rate": 5.976282510228303e-10, "loss": 0.4383, "step": 10978 }, { "epoch": 0.09874400219451103, "grad_norm": 5.901533126831055, "learning_rate": 5.467904943851077e-10, "loss": 0.6657, "step": 10979 }, { "epoch": 0.09875299609215148, "grad_norm": 17.370803833007812, "learning_rate": 4.98212161398337e-10, "loss": 0.5635, "step": 10980 }, { "epoch": 0.09876198998979192, "grad_norm": 2.818006753921509, "learning_rate": 4.5189325645622614e-10, "loss": 0.2651, "step": 10981 }, { "epoch": 0.09877098388743237, "grad_norm": 3.1137917041778564, "learning_rate": 4.078337837470914e-10, "loss": 0.2688, "step": 10982 }, { "epoch": 0.09877997778507283, "grad_norm": 2.665940046310425, "learning_rate": 3.6603374724830664e-10, "loss": 0.4636, "step": 10983 }, { "epoch": 0.09878897168271328, "grad_norm": 2.996216297149658, "learning_rate": 3.2649315074018137e-10, "loss": 0.3322, "step": 10984 }, { "epoch": 0.09879796558035372, "grad_norm": 1.9570175409317017, "learning_rate": 2.8921199779763373e-10, "loss": 0.3229, "step": 10985 }, { "epoch": 0.09880695947799419, "grad_norm": 2.018376588821411, "learning_rate": 2.5419029178741506e-10, "loss": 0.5015, "step": 10986 }, { "epoch": 0.09881595337563463, "grad_norm": 2.383944511413574, "learning_rate": 2.2142803587643647e-10, "loss": 0.2633, "step": 10987 }, { "epoch": 0.09882494727327508, "grad_norm": 2.6477298736572266, "learning_rate": 1.9092523302344234e-10, "loss": 0.6802, "step": 10988 }, { "epoch": 0.09883394117091554, "grad_norm": 8.00752067565918, "learning_rate": 1.6268188598733692e-10, "loss": 1.1784, "step": 10989 }, { "epoch": 0.09884293506855599, "grad_norm": 3.08449125289917, "learning_rate": 1.3669799732163313e-10, "loss": 0.5041, "step": 10990 }, { "epoch": 0.09885192896619643, "grad_norm": 2.3059043884277344, "learning_rate": 1.1297356937445269e-10, "loss": 0.25, "step": 10991 }, { "epoch": 0.09886092286383688, "grad_norm": 2.235987663269043, "learning_rate": 9.150860428852604e-11, "loss": 0.3056, "step": 10992 }, { "epoch": 0.09886991676147734, "grad_norm": 3.200597047805786, "learning_rate": 7.230310400396789e-11, "loss": 0.4354, "step": 10993 }, { "epoch": 0.09887891065911779, "grad_norm": 2.0202219486236572, "learning_rate": 5.5357070255501743e-11, "loss": 0.3481, "step": 10994 }, { "epoch": 0.09888790455675824, "grad_norm": 3.213041305541992, "learning_rate": 4.0670504578010916e-11, "loss": 0.3715, "step": 10995 }, { "epoch": 0.0988968984543987, "grad_norm": 1.9504659175872803, "learning_rate": 2.8243408298211927e-11, "loss": 0.2332, "step": 10996 }, { "epoch": 0.09890589235203914, "grad_norm": 2.6644089221954346, "learning_rate": 1.8075782534654472e-11, "loss": 0.39, "step": 10997 }, { "epoch": 0.09891488624967959, "grad_norm": 2.5969486236572266, "learning_rate": 1.0167628211599222e-11, "loss": 0.2552, "step": 10998 }, { "epoch": 0.09892388014732004, "grad_norm": 15.358250617980957, "learning_rate": 4.5189460423644694e-12, "loss": 0.7717, "step": 10999 }, { "epoch": 0.0989328740449605, "grad_norm": 13.970982551574707, "learning_rate": 1.129736534877246e-12, "loss": 0.5021, "step": 11000 }, { "epoch": 0.0989328740449605, "step": 11000, "total_flos": 1.6431132872624046e+18, "train_loss": 1.9227141628443178, "train_runtime": 121505.234, "train_samples_per_second": 0.724, "train_steps_per_second": 0.091 } ], "logging_steps": 1.0, "max_steps": 11000, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 1.6431132872624046e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }